acetyl coenzyme a carboxylase system of escherichia coli

THE JOURNAL OF B~o~oarcar. CHEMISTRY Vol. 249. No. 20, Issue of October 25, PP. 6653-6645, 1974

Printed in U.S.A.

Acetyl Coenzyme A Carboxylase System of Escherichia coli

PURIFICATION AND PROPERTIES OF THE BIOTIN CARBOXYLASE, CARBOXYLTRANSFERASE, AND CARBOXYL CARRIER PROTEIN COMPONENTS*

(Received for publication, December 20, 1973)

RAS B. GUCHHAIT, S. EFTHIMIOS POLAKIS, PETER DIMROTH,$ ERWIN STOLL,~ *JOEL iUoss, AND

M. DANIEL LANE

From the Department of Physiological Chemistry, The Johns Hopkins University School vj Medicine, Baltimore, Maryland 21205

SUMMARY

The three protein components (biotin carboxylase, carboxyltransferase, and the biotin-containing carboxyl carrier protein of the acetyl coenzyme A carboxylase system have been resolved and purified extensively or to homogeneity from cell-free extracts of Escherichia co2i B. Carboxylation of acetyl-CoA to form malonyl-CoA requires the presence of all three components. Biotin carboxylase, which catalyzes the first half-reaction,

CCP (carboxyl carrier protein)-biotin

Me*+ (1) + HCOI- + ATP ___L CCP-biotin-CO,- + ADP + P,,

has been purified to a homogeneous state and has been crystallized; earlier work in this laboratory showed the enzyme to be composed of two 50,000-dalton subunits. The carboxylation of free d-biotin which can substitute for car- boxy1 carrier protein-biotin as model substrate is markedly activated by certain organic solvents; an optimal rate enhancement of lo-fold is obtained with 15% (v/v) ethanol. Activation by ethanol affects Vrnax and is not accompanied by changes in K, values or the state of aggregation of the enzyme. Moreover, none of the carboxyl carrier protein-dependent reactions catalyzed by biotin carboxylase, e.g. acetyl-CoA carboxylation, ATP-[%]ADP exchange, and ATPm3*Pi exchange, are activated by organic solvents.

Carboxyltransferase, the catalyst for the second half-reaction,

CCP-biotin-CO1- + acetyl-CoA e CCP-biotin (II)

+ malonyl-CoA,

was purified to apparent homogeneity. Biotin carboxylase, although not required for the second half-reaction as measured by carboxyltransferase- and carboxyl carrier protein-

* This work was supported by research grants from the National Institutes of Health, United States Public Health Service (AM- 14574 and AM-14575), and the American Heart Association.

$ Present address, Institut fur Biochemie, Universitiit Iiegens- burg, UniversitBtstrasse, Regensburg, West Germany.

8 Present address, Institut fur Molekularbiologie und Bio- physik, Eidgenossiche Technische Hochschule, Zurich, Switzer- land.

dependent malonyl-CoA-[14C]acetyl-CoA exchange, activates this process indicating the existence of a ternary complex between the the three protein components. In addition to the above reaction (Reaction II), carboxyltransferase catalyzes: (a) net transcarboxylation from malonyl-CoA to free d- biotin derivatives in the absence of biotin carboxylase and carboxyl carrier protein, and (b) a slow biotin-independent decarboxylation of malonyl-CoA. The carboxyltransferase component has a molecular weight of 130,000 and is composed of nonidentical polypeptide chains of 30,000 and 35,000 daltons.

Carboxyl carrier protein has been purified extensively by a combination of conventional methods and affinity chromatography with Sepharose-avidin (monomer). Polyacryla- mide gel electrophoresis of purified carboxyl carrier protein revealed two major proteins both of which contain biotin and exhibit carboxyl carrier protein activity, i.e. carboxyl carrier protein- and carboxyltransferase-dependent malonyl-CoA- [i%]acetyl-CoA exchange.

The two catalytic components of the E. coli acetyl-CoA carboxylase system, biotin carboxylase and carboxyltransferase, are devoid of free or covalently bound biotin yet have the ability to carry out their respective model half-reactions utilizing free d-biotin derivatives in place of carboxyl carrier protein. Thus, in addition to possessing binding sites for their respective substrates, each catalytic component must contain a specific binding site for the biotinyl moiety of car- boxy1 carrier protein. It is evident that during the over-all sequence (Reactions I + II) the carboxylated biotinyl prosthetic group must undergo translocation from the carboxylation site on biotin carboxylase to the transfer site on carboxyl transferase while remaining attached to carboxyl carrier protein through its 14 A side chain.

Acetyl coenzyme A carboxylase catalyzes the initial committed step of de nooo fatty acid synthesis, namely the carboxylation of acetyl-CoA to form malonyl-CoA. Like other biotin-dependent carboxylation reactions (l), that carried out by Escherichia coli acetyl-CoA carboxylase can be partitioned into two half-reac-

6633

by guest on April 10, 2019

http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

6634

tions, the first involving carboxylation of the biotin prosthetic group by an MgATP-dependent process (Z-4) and the second, carboxyl transfer from the carboxylated prosthetic group to acetyl-CoA to yield malonyl-CoA (5, 6). Unlike its counter- parts in animal tissues (1, 7-12), the carboxylase system from E. co2i dissociates readily into three protein components, all of which are essential for acetyl-CoA carboxylation. As will be demonstrated in this and the accompanying papers, the resolved components retain the capacity to carry out their respective partial reactions, thus permitting studies on the carboxylation mechanism and its regulation not possible with unresolved animal carboxylases. Two of these components, i.e. biotin carboxylase and carboxyltransferase, possess catalytic centers for the first (Reaction 1) and second (Reaction 2) half-reactions, respectively, of acetyl-CoA carboxylation.

and, finally, 20 mM potassium phosphate, pH 7.0, containing 0.1 mM EDTA and 5.0 mM 2-mercantoethanol.

Sepharose 4B was obtained irom Pharmacia Fine Chemicals, nucleotides and CoA from PL Biochemicals, avidin (biotin- binding capacity 10 pg per mg of protein) from Worthington Biochemical Corporation, and ultrapure guanidine HCl and ammonium sulfate from Fchwarz-Mann. [l-14C]Acetic anhydride, NaH1*CO,, and BaWOa were purchased from New England Nuclear and d-[Y-“C]biotin from Amersham-Searle Corp. Ace- tyl-CoA was prepared as described by Simon and Shemin (28) and unlabeled and [2-14C]malonyl-CoA according to the method of Trams and Brady (29). [3-‘*C]Malonyl-CoA was enzymatically synthesized by the procedure of Gregolin et al. (8); CoA thioesters were purified as described earlier (8).

Protein concentration was determined as indicated by the methods of Lowry el ~2. (30), the spectrophotometric method of Warburg and Christian (31), or by the biuret method (32). Dialy- sis of carboxyl carrier protein was performed in tubing pretreated

MeZ+ CCP (carboxyl carrier protein) biotin + HO-CD- + ATP t biotin carboxylase ’ CCP-biotin-COz- + ADP -I- Pi(-OH)

Me* CCP-biotin-CO1 + acetyl-CoA , ’ CCP-biotin + malonyl-CoA

carboxyltransferaee

(1)

(‘4

Me2+ Net: HO-C02- + ATP + acetyl-CoA , ’ malonyl-CoA + ADP + Pi (3)

Evidence presented in an accompanying paper (13) and else- where (14-17) indicates that the biotin prosthetic group is covalently linked to the third component, carboxyl carrier protein. As with other biotin enzymes (l), the bicyclic ring of the prosthetic group resides at the distal end of a flexible 14 A side chain which allows it to act as a “mobile carboxyl carrier” between the two catalytic centers.

Although there is compelling evidence (10, 18,19) to show that fatty acid synthesis in animal cells is regulated by acetyl-CoA carboxylase, information on the mechanism of control of lipo- genesis in bacteria has been lacking. Recent investigations in this laboratory (20, 21) indicate that fatty acid synthesis per se

is under stringent control in E. coli and that the site of control is acetyl-CoA carboxylase. Amino acid starvation of stringent, but not relaxed, strains of E. coli activates ribosomal synthesis (22, 23) and the accumulation of guanosine 5’-diphosphate-3’- diphosphate (ppGpp) and guanosine 5’-triphosphate-3’-diphosphate (pppGpp), apparent mediators of the stringent control of RNA synthesis (24-26). Interestingly, ppGpp and pppGpp specifically inhibit the carboxyltransferase component of the acetyl-CoA carboxylase system.

The present paper describes the resolution, purification, and molecular properties of the biotin carboxylase, carboxyltransferase, and carboxyl carrier protein components of the acetyl-CoA carboxylase system from cell-free extracts of E. coli B.

EXPERIMENTAL PROCEDURE

Materials

Biotin Carboxylase: Spectrophotometric Assay-After Step 3 of the purification procedure (Table I), the spectrophotometric assay in which NADH oxidation is coupled to d-biotin-dependent ADP formation may be employed. A l.O-ml reaction mixture containing the same components and concentrations as above is used except that H’“C03 is replaced by unlabeled bicarbonate and 0.5 mM phosphoenolpyruvate, 0.2 mM NADH, 5 units of lactate dehydrogenase, and 3 units of pyruvate kinase are added. The reaction is initiated with biotin carboxylase and the d-biotin- dependent oxidation of NADH followed by 340 nm and 30”.

Full or 44 log phase E. coli B cells grown on enriched medium were purchased from Grain Processing Corp., Muscatine, Iowa. Calcium phosphate gel was prepared according to the method of Keilin and Hartree (27) except that the gel was washed more extensively with distilled water until the washes were chloride- free. Standard and type 20 DEAE-cellulose and standard phosphocellulose ion exchangers were obtained from the Schleicher and Schuell Company. DEAE-cellulose was equilibrated by washing with 0.5 M p&as&urn phosphate, pH 7.0, fofiom-ed by 0.02 M potassium nhosnhate. DH 7.0. containing 0.1 mM EDTA and 5 mM 2-merEaptoethanoi. Phosphocelluloie was equilibrated by suc- Carboxyltransjerase and Malonyl-CoA Decarboxylase: Radioac- cessive washing with 0.1 N NaOH, water (until neutral), 0.1 N tive AssaeCarboxyl transfer from malonyl-CoA to free d-biotin HCl, water (until neutral), 0.05 M potassium phosphate, pH 7.0, methyl ester is determined in a reaction mixture (total volume,

One unit of biotin carboxylase catalyzes the formation of 1 pmole of free carboxybiotin per min under the conditions described above. Both assays follow zero order kinetics and are proportional to enzyme concentration.

as described by Callanan et al. (33) to reduce the pore size and prevent the escape of carboxyl carrier protein.

Enzyme Assays

Biotin Carboxylase: ‘%-Bicarbonate Fixation Assay-To obviate high blanks due to l*C contaminant(s) in commercial NaH”C03: not removed by gassing with COZ, H11C03- was purified as follows Ba”C03 (0.088 mmole; 60 mCi per mmole) suspended in 2 ml of water is allowed to stand for 2 hours at room temperature. After centrifugation, the precipitate is washed again and transferred to a Warburg flask; 0.5 ml of 0.18 N NaOH (0.09 mmole) and 1 ml of 7yo perchloric acid are placed in the center well and side arm, respectively. After evacuating the flask, the acid is tipped in and ‘4CO2 transfer permitted to occur overnight. The center well contents are diluted with 0.1 M KHCO,. -

The carboxylation of free d-biotin is determined by following the incorporation of lY4C03 into a form (1’-N-[“Cjcarboxybiotin) stable to gassing with unlabeled COZ. The reaction mixture contains 100 mM triethanolamine (Cl-) buffer. DH 8.0: 1 mM ATP: 8 mM MgC12; 8 mM KHC03 (300 cpm pkr nmolkj; 50 m’M potassiu& d-biotin; 3 mM glutathione; 0.3 mg of bovine serum albumin; 0.05 ml of ethanol; and 0.1 to 1.0 milliunit of biotin carboxylase in a total volume of 0.5 ml. The reaction is initiated with enzyme, incubated for 10 min at 30”, and carboxylation is terminated by rapid transferral of a 0.4-ml aliquot to 1 ml of ice-cold water containing 2 drops of 1-octanol. CO2 is bubbled through the solution for 30 min at O-2” to remove the excess H14C0,. No significant losses of [14C]carboxybiotin occur (t 1,s = 192 min for decarboxylation under these conditions). After gassing, 0.1 ml of 0.1 N NaOH is added; the contents are transferred to a counting vial, scintillator added, and 14C activity determined.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

TABLE I TABLE II

Purijcation of biotin carboxylase from Escherichia coli B Purification of carboxyltransferase from Escherichia coli B

step

1. Cell-free extra&C (20,000 X g supernatant solution).

2. 2542y0 saturated (NHr)zSOd fractionation. .

3. Calcium phosphate gel fractionation

4. DEAE-cellulose chromatography .

5. Cellulose phosphate chromatography. . .

6. Crystallization (a) First. (b) Second.

Total ctivitya

units

334 1

175

196

137

73

53 40

-

Protein* Specific activity

w units/mg

11,000 0.0030

26,900 0.0065

5,630 0.035

38 3.6

12.0 6.1(3.8)

9.6 5.5(3.4) 5.3 5.9(3.7)

-

d

d

d

-

Yield

%

loo

52

59

41

22

16 12

a H”CO3 fixation biotin carboxylase assay. b Protein was determined by the biuret method (32) (Steps 1 to

4) and by the spectrophotometric method of Warburg and Chris- tian (31) (Steps 5 and 6).

c From 1 kg of packed E. coli B cells. d Specific activities in parentheses are calculated on the basis of

refractometrically determined protein (see “Experimental Pro- cedure”).

0.5 ml) containing 50 mM Tris (Cl-) buffer, pH 8.0; 85 PM [3-i4C]- or [2J*C]malonyl-CoA (4 to 6 X lo3 cpm per nmole); 10 mM d-biotin methyl ester; 0.3 mg of bovine serum albumin; and up to 1 milliunit of carboxyltransferase. After 5, 10, 15, and 20 min of incubation at 30”, O.l-ml aliquots are transferred to scintillation vials containing 0.1 ml of 6 h’ HCl. The mixture is taken to dryness at 95” in a forced-draft oven after which water and scintillator are added and residual “C acid-stable radioactivity determined. The [2J4C]acetyl group of acetyl-CoA, i4C02, or the [i4C]car- boxy1 group of N-carboxybiotin methyl ester generated during the reaction are volatile under these conditions, whereas [“Cl- malonyl-CoA is not. The amount of carboxyl transfer or decar- the reaction are volatile under these conditions, whereas [‘“Cl- malonyl-CoA is not. The amount of carboxyl transfer or decarboxylation equals the difference between acid-stable radioactivity at zero time and after incubation. Malonyl-CoA decarboxylase activity is estimated as described above but with the d-biotin methyl ester omitted; carboxyl transfer is equal to acid-stable radioactivity lost during incubation with d-biotin methylester minus that lost in the absence of the d-biotin derivative.

It has been established (5, 34) with the use of [3-14C]malonyl- CoA as carboxyl donor and a d-biotin derivative as acceptor that the acid-labile product stable to gassing with CO2 is the II-N- carboxybiotin derivative. Under the conditions described, the rate of carboxyl transfer is approximately 30 times faster than malonyl-CoA decarboxylation. The assay follows zero-order kinetics and activity is proportional to enzyme concentration.

Carboxyltransferase: Spectrophotometric Assay--This assay can be employed only with enzyme carried beyond Step 3 of Table II; it involves coupling the acetyl-CoA generated in the reaction to the combined citrate synthase-malate dehydrogenase reactions and following NAl>+ reduction spectrophotometrically. The reaction mixture contains 100 mM Tris (Cl-) buffer, pH 8.0, 0.1 mM malonyl-CoA, 5 mrw d-biotin methyl ester, 10 mM n-malate, 0.5 mM NAD+, 0.6 mg of bovine serum albumin, 18 units of malate dehydrogenase, 3.5 units of citrate synthase, and 1 to 10 milliunits of carboxyltransferase. The reaction is conducted at 30” and is initiated by the addition of d-biotin methyl ester; NADH formation is followed at 340 nm for 3 min.

One unit of carboxyltransferase catalyzes the formation of 1 pmole of free carboxybiotin methyl ester per min.

Carboxyl Carrier Protein: Malonyl-CoA-[lYT]Acetyl-CoA Ex- change Assay-Carboxyltransferase catalyzes an isotopic exchange between [i*C]acetyl-CoA and malonyl-CoA which is de-

6635

step

1. Cell-free extract*. . . . 2. 25-42y0 saturated am-

monium sulfate fractionation

3. Calcium phosphate gel fractionation. .

4. First DEAE-cellulose (standard) chromatography .

5. Phosphocellulose chromatography

Peak I. . . Peak II.. .

6. Second DEAE-cellulose (type 20) chromatog- raphyofPeakI......

-

1

--

.2

‘

-

Proteina Total

activity Specific

activity

alonyl de- rrboxylase specific activity

F !

10,000

uds

-c

43,000 8,500 0.2 0.008

5,176 6,910 1.33 0.050

360 4,564 12.7 0.52

20 900 45 1.80 30 260 8.67 0.33

6 570 96 3.20

-

-

Carboxyltransferase

(L Protein in Steps 1 to 4 was determined by the biuret method (32) and in Steps 5 and 6 by the method of Lowry et al. (30).

* From 2 kg of packed E. coli B cells. c Carboxyltransferase activity cannot be assayed accurately in

the cell-free extract due to the presence of an inhibitor; this inhibitor is removed by ammonium sulfate fractionation (Step 2).

pendent upon carboxyl carrier protein (see Reaction 2 and Befs. 5, 6, and 13). The reaction mixture contains 50 mM imidazole (Cl-) buffer, pH 6.5, 0.1 mM malonyl-CoA, 0.2 mM [l- or 2-14C]- acetyl-CoA (500 cpm per nmole), 1 mM dithiothreitol, 10 milliunits of carboxyltransferase, and carboxyl carrier protein in a total volume of 0.25 ml. After initiating the reaction (30”) with carboxyltransferase, 50-~1 aliquots are transferred at 2, 4, 6, and 8 min to scintillation vials containing 0.1 ml of 6 x HCl, the mixture taken to dryness at 95”, and the acid-stable ‘“C activity incorporated into malonyl-CoA determined as indicated for the carboxyltransferase assay (see preceding section). The rate of exchange is proportional to carboxyl carrier protein concentration up to a level of 0.25 nmole of covalently bound biotinyl prosthetic group at the specified level of carboxyltransferase. Since the exchange rate also depends upon the concentration of carboxyltransferase, the amount of purified transferase added must be adjusted to compensate for that present in the carboxyl carrier protein preparation. Following the first step in the purification of carboxyl carrier protein involving exposure to 6 M guanidine hydrochloride, the preparation is free of carboxyltransferase activity. Care must also be taken to avoid high salt concentrations since the exchange reaction is markedly inhibited by most salts at >O.l M.

Carboxyl Carrier Protein: Stoichiometric Carboxylation Assay- Carboxyl carrier protein may also be assayed by determining the extent of its carboxylation with Hi4C03 of known specific activity in the presence of Mg z+ ATP and excess biotin carboxylase (see Reaction 1). The extent of carboxylation of dialyzed preparations provides a stoichiometric measure of the amount of biotinyl prosthetic group present as carboxyl carrier protein. The assay mixture contains (final volume, 0.05 ml): 100 mM triethanolamine (Cl-) buffer, pH 8.0, 30 fig of bovine serum albumin, 1 mM ATP, 8 mM MgCl,, 8 mM KHi4C03 (20 X lo3 cpm per nmole), 10 to 20 milliunits of biotin carboxylase, and carboxyl carrier protein. The mixture is incubated for 10 min at 30”; 1.0 ml of cold water is added, followed by 3 drops of 1-octanol after which CO* is bubbled through the mixture at 0” for 30 min. The solution is transferred to a counting vial containing 0.1 ml of 0.1 N NaOH and the amount of ‘*C activity stable to gassing with COz (i.e. carboxy biotinyl carrier protein) determined.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

6636

Treatment of the carboxyl carrier protein with avidin abolishes its ability to support the malonyl-CoA-[Wlacetyl-Cob exchange reaction and its capacity to undergo carboxylation by biotin carboxylase; prior loading of avidin with free biotin renders it ineffective as inhibitor of both processes.

RESULTS

Isolation and Puri$cation of Biotin Carboxylase from E. coli B

The first three steps of the purification procedure for biotin carboxylase are identical with those for the purification of carboxyltransferase to be described later. All manipulations are carried out at 2-4” and all solutions employed contain 1 rnM EDTA and 5 mM 2-mercaptoethanol. The results of a typical biotin carboxylase purification are summarized in Table I. One kilogram of E. coli B cells, s/4 log phase, grown on enriched medium are suspended in 4 liters of 0.1 M potassium phosphate buffer, pH 7.0. The suspension is passed twice through a pre- cooled Manton-Gaulin submicron dispersor at 9000 p.s.i. The cell breaker is washed two or three times with 500 ml of buffer; the washes are combined with the extract, the pooled extract centrifuged at 20,000 x g for 30 min, and the supernatant solution retained.

Ammonium Sulfate Fractionation (25 to 42% Saturation)-The supernatant solution is brought to 25% saturation by addition of solid ammonium sulfate (144 g per liter) ; the pH was main- tained at 7.0 with dilute ammonium hydroxide. After 15 min, the mixture is centrifuged and the supernatant recovered and brought to 45% saturation with solid ammonium sulfate (125 g per liter). The precipitate is recovered by centrifugation and extracted with 4 liters of a 42% saturated ammonium sulfate solution containing 50 mM potassium phosphate buffer, pH 7.0. After removing the supernatant, the pellet is either stored at -90” or fractionated immediately with calcium phosphate gel

for the purification of biotin carboxylase (below) or carboxyltransferase.

Calcium Phosphate Gel Fractionation-The precipitated protein from the preceding step (approximately 27 g of protein) is dissolved in 200 ml of 5 mM KzHPOa containing 20% glycerol and dialyzed overnight against 10 liters of the same solution. The solution is diluted to give a protein concentration (spectrophotometric method) of 20 mg per ml. Sufficient calcium phosphate gel (30 mg dry weight per ml) is added to adsorb 30% of the protein; this usually requires bringing the gel to protein ratio to 1: 1. After stirring for 30 min, the mixture is centrifuged and the supernatant solution containing less than 10% of the original biotin carboxylase activity is discarded. The gel is washed once with 2 liters of 5 mM potassium phosphate buffer, pH 7.0. Biotin carboxylase activity is eluted by resuspending the gel precipitate in 2 liters of 0.12 M potassium phosphate buffer, pH 7.0, stirring for 30 min, and centrifuging; this process is repeated two additional times and the eluates combined and brought to 50% saturation with solid ammonium sulfate (313 g per liter). While all of the biotin carboxylase activity is recovered in this eluate, carboxyltransferase activity remains adsorbed to the gel. Approximately 6 g of protein are recovered in the 0.12 M buffer eluate and this can be stored frozen at -90” as a pellet after centrifugation or fractionated immediately by DEAE-cellulose chromatography.

DEAE-cellulose Chromatography-The precipitated enzyme is dissolved in 75 ml of 10 mM potassium phosphate buffer, pH 7.0, containing 20% glycerol. After dialysis against the same buffer and centrifugation, the supernatant (about 6 g of protein) is applied to a 3-liter DEAE-cellulose column (9 x 50 cm) pre-

viously equilibrated with dialysis buffer. Elution is accomplished with 8 liters of the same 10 mM potassium phosphate buffer and the eluate is collected fractionally. Biotin carboxylase activity appears with the first protein eluted from the column and precedes the major “breakthrough” protein peak. This step achieves a purification of at least loo-fold, giving rise to approximately 30 to 40 mg of protein in the biotin carboxylase peak; the active fractions are pooled and subjected immediately to chromatography on cellulose phosphate.

Cellulose Phosphate Chromatography-The pooled fractions from DEAE-cellulose chromatography (about 40 mg of protein) are applied directly to a 50-ml cellulose phosphate column (2 x 30 cm) previously equilibrated with 10 mM potassium phosphate buffer, pH 7.0, containing 20% glycerol. Elution is carried out with a l-liter linear potassium phosphate gradient (10 to 500 mM, pH 7.0, containing 20y0 glycerol). The effluent is collected fractionally and is monitored for biotin carboxylase activity and protein; carboxylase activity is eluted at a phosphate concentration of approximately 0.1 M. The fractions containing maximal activity are placed in dialysis bags, and dialyzed agaiust a solution containing 50 mM potassium phosphate buffer, pH 7.0, 20% glycerol, and sufficient ammoniutn sulfate to bring the solution to 60% saturation at equilibrium. After 1 to 2 days the floccu- lated protein is recovered by centrifugation and stored as a pellet at -90” (no loss of activity for several months) or is subjected immediately to the crystallization procedure.

Crystallization-The precipitated enzyme from the preceding step is dissolved in a minimal volume of 10 mM potassium phosphate buffer, pH 7.0, containing 1 mM EDTA and 2 mM dithiothreitol to produce a protein concentration of 4 to 6 mg per ml. The solution is then dialyzed against the same buffer and crystallization usually begins within 24 hours. As shown in Fig. 1 (inset), the crystals appear as prisms ranging from 0.01 to 0.025 mm. Crystallization can be accelerated by use of seed crystals or by 10% ethanol in the medium (35) ; crystals prepared in the presence of ethanol are larger and appear as plates which crack spontaneously.

Isolation and Purification of Carboxyltransjerase from E. coli B

The preparation of a cell-free extract of E. coli B cells and its fractionation with ammonium sulfate (25 to 42% saturated ammonium sulfate fraction) are carried out as described in the preceding section except that 2 kg of packed cells are used. All subsequent steps are performed at 2-4” with solutions containing 1 mM EDTA and 5 mM 2-mercaptoethanol. The results of a typical purification are summarized in Table II.

Calcium Phosphate Gel Fractionation-The precipitate obtained by ammonium sulfate fractionation (25 to 42y0 saturation) as described in the preceding section is dissolved in 250 ml of 2 mM KzHPOl and dialyzed against 25 liters of the same buffer overnight. The dialyzed enzyme is diluted with buffer to a protein concentration of 20 mg per ml and sufficient calcium phosphate gel (30 mg dry weight per ml) is added to adsorb 65% of the protein; this usually requires a gel-protein ratio of approximately 2.5:l.O. After stirring for 30 min, the mixture

is centrifuged and the supernatant solution which contains about 5% of the original carboxyltransferase activity is discarded or retained for the isolation of the carboxyl carrier protein as described later. The gel pellet is extracted three times with 2 liters of 0.12 M potassium phosphate buffer, pH 7.0. These eluates, which contain virtually no carboxyltransferase activity, are discarded or utilized for carboxyl carrier protein isolation. The enzyme is then eluted from the gel by extracting the pellet


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

6637

08

0.6

I

GEL BOTTOM

i 17

DYE MARKER

4 _ L

J I I I I v r T

3

DISTANCE , CM=

three times with 3 liters of 0.5 M potassium phosphate buffer, pH 7.0. The pooled 0.5 M eluates which contain 80 to 90% of the original carboxyltransferase activity are brought to 60% saturation with solid ammonium sulfate (390 g per liter) and allowed to stir for 20 min. After centrifugation, the precipitate is dissolved in 200 ml of 25 mM potassium phosphate buffer, pH 7.0, and then dialyzed against 30 liters of the same buffer overnight.

First DEAE-cellulose Chromatography-The dialyzed enzyme (about 5 g of protein) is applied to a X&liter DEAE-cellulose column (9 X 60 cm) previously equilibrated with 25 mM dialysis buffer. The column is washed with 4 liters of 25 mM potassium phosphate buffer, pH 7.0, after which the enzyme is eluted with an &liter linear phosphate gradient (25 to 500 mM, pH 7.0) ; carboxyltransferase is eluted at a phosphate concentration of approximately 0.2 M. The most active fractions are pooled and the enzyme precipitated by bringing the ammonium sulfate concentration to 60% saturation with solid ammonium sulfate. After collecting the precipitated protein by centrifugation, the pellet is dissolved in 40 ml of 25 mM phosphate buffer, pH 7.0, and then dialyzed overnight against 6 liters of the same buffer.

Phosphocellulose Chromatography-The dialyzed enzyme (about 360 mg of protein) is applied to a 410-ml phosphocellulose column (2.5 X 90 cm) previously equilibrated with 25 mM dialysis buffer, pH 7.0. After passing 200 ml of the same buffer through the column, the enzyme is eluted with a a-liter linear phosphate gradient (25 to 300 mM potassium phosphate, pH 7.0). The enzyme appears in the column eluate at a phosphate concentration of approximately 0.15 M. As shown in Fig. 2, two distinct peaks of carboxyltransferase activity are observed. The major activity peak (Peak I, at 1650 ml eluate volume) which comprised approximately 80% of the total activity is retained and further purified; although Peak II has not been further purified, its kinetic properties appear indistinguishable from those of

FIG. 1. Polyacrylamide gel electrophoresis pattern and crystals (inset) of purified biotin carboxylase. Acryl- amide gels (7.5%) were prepared (36) and equilibrated with electrophoresis buffers (cathode buffer, 50 mM Tris-glycine), pH 8.9, containing 10 mM P-mercaptoethanol; anode buffer, 100 mM Tris (Cl-), pH 8.1, for 16 hours at 4” with a constant current of 1 ma per gel. After replacing the buffers, 36 pg of purified biotin carboxylase (Step 5, Table I) in 50 mM potassium phosphate buffer, pH 7.0, containing 1 mM EDTA, 5 mM dithiothreitol, and 20% glycerol was layered on the gel surface and electrophoresis performed for 3 hours at 4 ma per gel at 4”. Gels were stained with Amido black and after destaining were scanned using a Gilford gel scanner. The inset shows crystalline biotin carboxylase prepared as described in the text from a solution containing 4 mg of enzyme per ml.

0 600 1200 moo 2400 3000 Volume ,ml

FIG. 2. Chromatographic purification of carboxyltransferase on cellulose phosphate. Carboxyltransferase (360 mg of protein; Table II, Step 4) was chromatographed on a cellulose phosphate column (2.5 X 90 cm) and assayed by the spectrophotometric assay as described in the text.

Peak I. The pooled fractions from Peak I (Fig. 2) are placed in dialysis bags and the enzyme is precipitated by dialysis against sufficient 65a/, saturated ammonium sulfate containing 50 mM phosphate buffer, pH 7.0, such that at equilibrium the percentage of saturation reaches 60. The precipitate is recovered by centrifugation, dissolved in 2 ml of 50 mM potassium phosphate buffer, pH 7.0, containing 1 mM EDTA and 2 mM of dithiothreitol, and is then dialyzed overnight against 1 liter of the same buffer.

Second DEAE-cellulose (Type .ZO) Chromatography-The dialyzed enzyme (about 20 mg of protein) from the preceding step is applied to a 120-ml DEAE-cellulose (type 20) column (1.5 X 90 cm) and the carboxyltransferase eluted with a 500-ml linear potassium phosphate, pH 7.0, gradient (50 to 250 mM). Car-


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

6633

boxyltransferase activity, which appears in the eluate when the phosphate concentration reaches about 150 mM, is exactly coin- cident with the protein peak. Fractions having maximal enzyme activity are pooled and precipitated by dialysis against 65$& saturated ammonium sulfate as described previously. After centrifugation the pellet is stored at -90“.

Isolation and Purification of Biotin-containing Carboxyl Carrier Protein(s) from E. coli B

Previous investigations in this laboratory (4, 5) with a biotin auxotroph (strain SA 283) of E. coli grown on d-[*4C]biotin revealed that >90% of the biotin present in the cell-free extract was recovered in the 25 to 42y0 saturated ammonium sulfate fraction routinely used for the preparation of biotin carboxylase (Table I) and carboxyltransferase (Table II). Moreover, > 80% of this protein-bound biotin appears in the gel supernatant from the subsequent steps of the fractionation of biotin carboxylase and carboxyltransferase on calcium phosphate gel. Although any of these fractions, i.e. the 25 to 42% ammonium sulfate fraction or the gel supernatants remaining after the adsorption of biotin carboxylase (Table I) or carboxyltransferase (Table II), can be used as starting material for the purification of the car- boxy1 carrier protein, the gel supernatant is generally employed. The carrier protein in this fraction is precipitated at a concentration of 80% saturated ammonium sulfate and the protein pellet, recovered by centrifugation, is redissolved in 50 mM potassium phosphate buffer, pH 7, containing 1 InM EDTA. This solution is brought to 6 M guanidine HCI with the crystalline salt, is stirred for 1 hour at O-2”, and then is dialyzed overnight against two changes of the 50 mM phosphate buffer. The volu- minous precipitate of denatured protein is removed by centrifugation and washed twice with the 50 mM phosphate buffer. The original supernatant solution and washings containing the car- boxy1 carrier protein are combined; protein is precipitated by the addition of sufficient solid ammonium sulfate to bring the concentration to 80 y0 saturation. The suspension is centrifuged and the protein pellet dissolved in 0.5 M KC1 containing 50 mM potassium phosphate, pH 7.0, and 0.1 mM EDTA. After centrifugation to clarify the solution, the supernatant is retained for purification of carboxyl carrier protein by affinity chromatography on Sepharose-avidin.

Preparation and Calibration of Sepharose-Avidin-Since the dissociation constant for the avidin tetramer-biotinh complex is unsuitably low (K, z 10-15 M) for affinity chromatography (l), whereas the affinity of the avidin monomer for biotin is much weaker (l), a procedure designed to produce Sepharose-avidin monomer was developed. Cyanogen bromide-activated Sepha- rose 4B is prepared (37) using 100 mg per ml of packed Sepha- rose. The coupling of avidin is performed at low pH (pH 5.5 in 0.1 M sodium acetate) to minimize the number of cross-links per avidin molecule in order that the subunits not covalently linked can be dissociated subsequently. One milliliter of avidin solution (1.5 mg per ml) in coupling buffer is used per ml of packed CNBr-activated Sepharose and coupling is allowed to proceed for 16 hours at 4”. The Sepharose-avidin tetramer is then washed with distilled water and treated with 1 M Tris base overnight; the tetramer form is washed with distilled water and stored in 0.2% NaN3 at 4”. It is recommended that Sepharose- avidin be stored in the tetramer form because it is considerably more stable than is the monomer form. Although the stability of the tetramer has not been studied rigorously, our observations suggest a 10 to 15% loss of activity per month.

The Sepharose-avidin monomer is generated as needed by

dissociating the subunits of the tetramer form not covalently linked to the Sepharose matrix with 6 M guanidine HCl. The appropriate amount of the tetramer is washed on a Millipore filter with distilled water and resuspended in 5 volumes of 6 M

guanidine HCl in 50 InM potassium phosphate, pH 7.0, containing 5 mM EDTA. The resuspended pellet is left for 30 min at 0” and the solution then removed by filtration; this step is repeated three additional times. After the fourth washing, the gel is washed repeatedly with distilled water and finally is resuspended in 0.1 M phosphate buffer, pH 7.0, containing 1 mM EDTA.

The biotin-binding capacity of Sepharose-avidin is determined by allowing an aliquot of the tetramer or monomer suspension to interact with an excess of [14C]biotin (approximately 70,000 cpm per nmole) for 20 min at 25” with occasional stirring. The Sepharose-avidin gel is collected on a Millipore filter, washed, and counted. Normally, 40 to 50y0 of the avidin added in the initial coupling reaction becomes covalently bound yielding the tetramer which contains about 0.7 mg of avidin per ml of gravity- packed gel (binding capacity, 7 pg of free d-biotin per ml). Treatment of the tetramer with 6 M guanidine HCl results in the dissociation of the bulk, 65 to 70%, of avidin subunits, the theoretical value being 75,%%. Thus, the residual covalently linked avidin subunits (monomer) have a biotin-binding capacity of 2 to 2.5 pg per ml of gravity-packed gel.

It became apparent that Sepharose-avidin monomer possessed multiple classes of biotin-binding sites which differ in their affinities for free biotin and the biotinyl prosthetic group. Thirty- six per cent of the bound [14C]biotin exchanges readily with unlabeled d-biotin (10 mM), whereas an additional 26% is eluted only with strong eluting agents, i.e. neutral (pH 7.0) or acidic (pH 1.5) 6 M guanidine HCl; moreover, another fraction appears to be bound almost irreversibly,

Afinity Chromatography of Carboxyl Carrier Protein on Sepha- rose-Avidin Monomer-Carboxyl carrier protein (1.22 g of protein in 100 ml of 0.5 M KCl-50 my phosphate buffer, pH 7.0) partially purified as described above by fractionation with ammonium sulfate, calcium phosphate gel, and guanidine HCl, is slurried with 70 ml of gravity-packed monomer gel (biotin binding capacity, about 2.5 pg per ml of gel). Essentially complete adsorption occurs within 1 to 2 hours. The suspension of Sepharose-avidin monomer-carboxyl carrier protein complex is then poured into a 4.2-cm diameter column (5.5 cm in height) ; the height to diameter ratio of the packed gel in the column should not exceed 3 : 1 in order to maintain an adequate flow rate. The column is eluted sequentially with 650 ml of 0.5 M KC1 containing 50 mM potassium phosphate, pH 7, and 1 rnhf EDTA; 300 ml 0.5 M KC1 containing 0.2 M sodium glycinate buffer, pH 9, and 1 IIlM EDTA; 200 ml of the preceding buffer containing 10 mM d-biotin; 200 ml of 6 M guanidine HCl containing 50 mM potassium phosphate buffer, pH 7, and 1 mM EDTA; and finally with 6 M guanidine HCl containing 50 mM HaPOa, pH 2, and 1 mM EDTA. As shown in Table III, the bulk (about 60%) of the carboxyl carrier protein activity, but only a small fraction of the protein applied to the column, is eluted by the latter three eluants (Fractions II, B, C, and D, Table III) ; the extent of purification by affinity chromatography ranges from 20 to 70 fold in the three fractions.

Properties of Acetyl-CoA Carboxylase System

Reconstitution of Acetyl-CoA Carboxylase Activity with Purified Biotin Carboxylase, Carboxyltransjerase, and Carboxyl Carrier Protein-The common initial steps for the purification of biotin


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

6639

TABLE III TABLE IV

Puri$cation of carboxyl carrier protein by afin.ity chromatography on Sepharose-avidin (monomer)

Reconstitution of Eech,erichia coli acetyl-CoA carboxylase system

The calcium phosphate gel supernat,ant fraction (18,800 mg of protein) after adsorption of biotin carboxylase (Step 3 of Table I, but beginning with 1.2 kg of packed Escherichia coli B cells) was treated with 6 M guanidine HCl and prepared for affinity chromatography as described in the text. The carboxyl carrier protein preparation (I220 mg of protein in 100 ml of 0.5 M KC1 containing 50 mM phosphate, pH 7, and 1 mM EDTA) was treated with 70 ml of gravity-packed Sepharose-avidin monomer (binding capacity, 7.4 nmoles or 180 rg of d-biotin). After allowing 2 hours with occasional stirring, the suspension was poured into a column and sequential elution accomplished as described in the text. Frac- tions II, A and B, were brought to 80% saturation with solid ammonium sulfate and the precipitated protein dissolved in and dialyzed against 50 mM potassium phosphate, pH 7, containing 2 mM dithiothreitol and 1 mM EDTA. Fractions II, C and I), were treated similarly except that guanidine HCl was removed by dialysis prior to precipitation of protein with ammonium sulfate. Malonyl-CoA-[“C]acetyl-CoA exchange activity was determined in the presence of carboxyltransferase as described under “En- zyme Assays.” Biotin content was determined either by the stoichiometric carboxylation assay (see “Enzyme Assays” for details) or by the avidin-[Wlbiotin binding assay described in Table II of the accompanying paper (13). Protein was determined by the method of Lowry et al. (30).

Acetyl-CoA carboxylation (forward reaction) rates were determined using a reaction mixture which contained: 60 mM imidazole (Cl-) buffer, pH 6.8; 0.5 mM ATP; 0.5 mM MnCl*; 14 mM KHWOI (280 cpm per nmole); 0.26 mM acetyl-CoA; 6 mM GSH; bovine serum albumin, 0.15 mg; crystalline biotin carboxylase, 10 pg; homogeneous carboxyltransferase, 10 pg; and purified carboxyl carrier protein (Fraction II, D, Table III), 7.5 pg, or as indicated in a total volume of 0.25 ml. After incubation at 30” for 3, 6, 9, and 12 min, aliquots were withdrawn for the determination of [Wlbicarbonate incorporated into malonyl-CoA as described previously (8). The rate of malonyl-CoA-dependent ATP formation (reverse rea.ction) was determined at 30’ with a reaction mixture (total volume, 0.05 ml) containing 50 mM potassium phosphate buffer, pH 7.0; 7.5 mM MgCl,; 3 mM GSH; bovine serum albumin, 0.6 mg per ml; 0.9 mM [S-W]AI)P (1000 cpm per nmole); 10 mM malate; 1.5 mM NAD+; 1 mM malonyl-CoA; malate dehydrogenase, 2.5 rg; citrate synthase, 2.5 pg; biotin carboxylase, 16 rg (38 milliunits) ; carboxyltransferase, 50 rg (1.5 milliunits) ; and carboxyl carrier protein, 6.9 pg (equivalent to 0.07 nmole of biotin). The reaction was started with malonyl-CoA, and it,s progress followed by sampling 5-111 aliquots at 2, 4, 6,8, 10,15, 20, 25, 30, 35, and 40 min of incubation. The samples were applied to polyethylenimine flexible thin layer chromatographic plates (Brinkmann) pre- spot.ted with 5.0 pl of a mixture containing 4 mM ATP, 4 mM ADP, 4 mM AMP, and 100 mM EDTA. The plates were developed with 0.2 M NHnHCOs, pH 7.8; the separated nucleotide spots were visualized under ultraviolet light, cut out, and counted.

-

Fraction Protein

I. Before chromatography.. II. After chromatography

(Sepharose-avidin monomer column eluates) : A. 0.5 M KCl, pH 7-9. B. 10 m&f d-biotin-0.5 M

KCl, pH 9. C. 6 M guanidine HCl, pH

i . D. 6 M guanidine HCl, pH

2.

mg

1220

1200 0.42

9.5 33

9.4 20

2.8 68

Specific activity

nalonyl-COI [Wlacetyl-

CoA exchange)

Biotin contenta

pg/nzg pro!&

0.1

0.04

3.8

2.2

7.0 (7.9)b

0 Determined by the stoichiometric carboxylation assay. b Determined by the avidin-[Wlbiotin binding assay

carboxylase, carboxyltransferase, and carboxyl carrier protein (through Step 2 of Tables I and II, 25 to 42% fractionation) yield a preparation which contains all three components and actively catalyzes the carboxylation of acetyl-CoA, i.e. 1 to 2 nmoles per min per mg of protein under the conditions described in Table IV. Nonetheless, the three components are largely resolved by the next step (calcium phosphate gel fractionation) and are completely separated by subsequent steps in the purification procedures. Thus it was found (4, 5) that chromatograph- ically purified biotin carboxylase and carboxyltransferase iso- lated from an E. coli biotin auxotroph (strain SA 283) grown on [2’-14C]biotin (7 x 107 cpm per pmole) were free of the [“Cl- biotin-containing carboxyl carrier protein. Both enzymes, prepared from [Wlbiotin-labeled cells as described above and carried through their respective purification procedures outlined in Tables I and II, contained <5 cpm per mg of protein. More- over, carboxyl carrier protein purified by affinity chromatog-

Additionsa FOlWXd RWWSe

reaction: reaction: acetyl-CoA malonyl-CoA-

carboxylated dependent ATP formation

n??des/min

BC or CT or CCP 0.00 0.00 BC+CTorBC+CCPorCT+CCP 0.00 0.00 BC + CT + CCP 9.50 1.25 BC + CT + CCP (1.5 pg) 2.70 BC + CT + lo-50 mM d-biotin methyl

ester, or biocytin 0.00

n BC, biotin carboxylase; CT, cnrboxyltransferase; and CCP, carboxyl carrier prot,ein.

raphy (Table III) contains no detectable biotin carboxylase or carboxyltransferase activity.

As shown in Table IV reconstitution of acetyl-CoA carboxylase activity is accomplished when all three purified protein components are present; any combination of two components is in- capable of catalyzing acetyl-CoA carboxylation or the reverse reaction. The rate of acetyl-CoA carboxylation is dependent upon carboxyl carrier protein concentration (Table IV), as well as the concentrations of the biotin carboxylase and carboxyltransferase components (results not shown). Although both biotin carboxylase and craboxyltransferase catalyze model partial reactions (3-5, 13, 34) in which free biotin derivatives, e.g. 10 to 50 mM d-biotin, biocytin, or d-biotin methyl ester, can substitute for carboxyl carrier protein, these compounds were unable to replace the carrier protein for acetyl-CoA carboxylation under the conditions described in Table IV.

The ease with which the three proteins of the acetyl-CoA carboxylase system can be resolved during purification suggests that if a ternary multienzyme complex functions in the overall- reaction, the complex must dissociate readily. Several findings suggest the participation of such a complex. As shown in Table


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

6640

TABLE V

Eflect of dilution and purified biotin carboxylase on carboxyltransferase- and carboxyl carrier protein-dependent malonyl-

CoA-[W]acetyl-CoA exchange

A, dialyzed ammonium sulfate fraction (Step 2, Tables I and II), 146 mg of protein per ml, was diluted with 0.05 M potassium phosphate buffer, pH 7.0, containing 1 mM EDTA and 5 mM fl-mercaptoethanol to the final protein concentrations indicated. The undiluted control and the diluted solutions were preincubat,ed for 1 hour at 30” before the exchange activity of 0.5 mg of protein was tested at 30” for 10 min as described under “Enzyme Assays.” Where indicated, the enzyme was concentrated by precipitation with solid ammonium sulfate at 80% saturation; the precipitated protein was dialyzed overnight at a protein concentration of 50 mg per ml and then assayed. B, malonyl-CoA-[‘4C]acetyl-CoA exchange activity was determined in a volume of 0.250 ml as described under “Enzyme Assays.” The biotin content of the car- boxy1 carrier protein (CCP) was 4.5 nmoles per mg and the specific activity of the carboxyltransferase (CT) was 1.1 pmoles per min per mg as measured with d-biotin at, pH 4.5 (5). Biotin carboxylase (BC) equivalent to 23.7 milliunits (1.5 units per mg, without ethanol) was added as indicated.

Conditions of preliminary incubation

A. 146 mg per ml, not preincubated 146 mg per ml, 1 hr at 30” 40 mg per ml, 1 hr at 30” 10 mg per ml, 1 hr at 30” 3 mg per ml, 1 hr at 30” 5 mg per ml, 64 hrs at 4” 5 mg per ml, 64 hrs at 4”, then concen-

trated to 50 mg per ml and assayed

Malonyl-CoA- [NJacetyl-CoA

exchange

nmole/min

0.75 0.85 0.72 0.50 0.24 0.12

0.60

Additions Without BC With BC

B. CT, 44 pg 0.0 0.0

CCP, 110 pg 0.0 0.0

CT, 44 /Ig + CCP, 110 pg 0.38 1.1 CCP, 220 /Ig 0.0 0.0 CT, 44 rg + CCP, 220 pg 1.1 1.2

V A the reversible loss by the unresolved carboxylase system of malonyl-CoA-[14C]acetyl-CoA exchange activity during a l-hour preliminary incubation is strongly dependent upon enzyme concentration. Moreover, this activity loss during prolonged incubation of a dilute enzyme preparation can be largely regained by concentrating the inactivated preparation. Another indication that a productive ternary complex is formed derives from the observation (Table V H) that the carboxyl carrier protein- dependent, carboxyltransferase-catalyzed malonyl-CoA-[r4C]- acetyl-CoA exchange is activated by the third component, biotin carboxylase, which does not participate in the reaction per se. The fact that biotin carboxylase activates at rate-limit- ing, but not at saturating, concentrations of the carrier protein, implicates a ternary complex between the protein components and suggests that the interaction between carboxyltransferase and carboxyl carrier protein is enhanced by the presence of biotin carboxylase in the complex.

Properties of Biotin Carboxylase Component and Carboxylation Reaction-Biotin carboxylase, obtained in good yield by the purification procedure outlined, has a specific activity about 2000-fold greater than that of the cell-free extract (Table I). Preparations carried through Step 6 are free of biotin (preceding

TABLE VI Requirements and stoichiometry of biotirh carboxylase-

catalyzed reaction

The complete reaction mixture was the same as described for biotin carboxylase (spectrophotometric assay) under “Enzyme Assays” except the unlabeled KHCO, was replaced by KHW03. The reaction was initiated with KHWOa and the d-biotin-dependent formation of ADP followed via the coupled oxidation of NADH at 340 nm and 30”. After a 5-min incubation, 0.1.ml aliquots were transferred to 1 ml of 10 mM triethanolamine (Cl-) buffer, pH 8.0, at 2”, 1 drop of I-octanol was added, and COz was bubbled through the solutions for 40 min. HWOI- fixed into a form stable to gassing with CO* at neutral pH was assessed by counting aliquots of the above solutions in a liquid scintillation counter. Acidification of the mixture to pH 2 led to a complete loss of ‘“C activity stable to gassing with CO* in all cases. Other additions to the reaction mixture were as follows: 50 mM I-biotin, 2 mM Mn2+, and 2 mM Cop+.

H”COa- fixed ADP formed

Complete. Minus d-biotin.. Minus d-biotin + I-biotin., Minus ATP. Minus HCO,-. Minus MgZf. Minus Mgz+ plus MnZ+. Minus Mgr+ plus Cozf. . Minus ethanol. Minus ethanol and d-biotin

0.0 16.7

4 0.0

24.6 4.3 0.0

a The small amount of carboxylase activity observed in the absence of added bicarbonate is due presumably to the unavoidable contamination of certain reaction mixture components, e.g. buffer, with traces of bicarbonate.

section) and crystallize readily from dilute potassium phosphate buffer in the form of elongated prisms (Fig. 1, inset). Although crystallization or recrystallization leads to no further improve- ment in specific activity, the stability of the enzyme is markedly increased. The crystalline enzyme appears homogeneous by several criteria; such preparations give rise to a single sediment- ing boundary in the analytical ultracentrifuge (Q~.~ = 5.7 S, Ref. 4) and to a single stained protein band by polyacrylamide gel electrophoresis (Fig. 1). Moreover, electrophoresis of the sodium dodecyl sulfate-dissociated enzyme on polyacrylamide gels containing sodium dodecyl sulfate (4) also yields a single stained protein band. On the basis of its subunit weight estimated by sodium dodecyl sulfate-acrylamide gel electrophoresis and the molecular weight of the native enzyme determined by sedimentation equilibrium and gel filtration, it was concluded (4) that biotin carboxylase is composed of two similar or identical polypeptide chains of 50,000 daltons.

In view of previous work on related oligomeric carboxylases (38, 39) which contain a covalently bound biotin prosthetic group, yet catalyze the carboxylation of free d-biotin, it was anticipated that the carboxylation product would be free carboxybiotin. This is consistent with the finding that the H14COa- fixation product of the biotin carboxylase reaction is acid-labile, but relatively stable at neutral pH and low temperature (Table VI), a property possessed by carboxybiotin (1). As reported in the accompanying paper (34), the pH dependence of the first order decarboxylation rate of the labeled carboxylation product is identical with that of l’-N-carboxy-d-biotin. Methylation


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

with diazomethane converted the acid-labile product to an acid- stable derivative which co-chromatographs (38-40) precisely with authentic l’-N-methoxycarbonyl-d-b&in methyl ester (40) (results not shown), That the l’-N position of biotin is the actual site of carboxylation is verified by the demonstration in the accompanying paper (34) that chemically synthesized l’-N- carboxy-d-biotin derivatives serve as substrates- for the reverse reaction catalyzed by biotin carboxylase, as well as for the model half-reaction catalyzed by carboxyltransferase.

The stoichiometry of the biotin carboxylase-catalyzed reaction was measured in the carbosylation direction by following H14C03 incorporation into carboxybiotin (H14C03- fixed into a form stable to gassing with CO2 at pH $:O, but which is labile in acid). AS shown in Table VI, approximately 1 mole of carboxybiotin (0.90 to 0.93) was synthesized per mole of ADP formed. Car- boxybiotin formation as assessed by one or both of these assay methods (Table VI) is absolutely dependent upon d-biotin, ATP, and divalent cation. Although the carboxylase-catalyzed reaction is activated 5- to 6-fold by 10% ethanol (4), the stoichiometry (Table VI) and the requirement for b&in, ATP, and Mg2+ are the same in the presence and absence of ethanol. The enzyme is highly specific in that I-biotin, the unnatural isomer, is completely inactive and a variety of other biotin derivatives are either inactive or less active than d-b&in (4, 13). The divalent cation requirement can be met by Mg*+, Co*+, or Mn2+; Co2+ and Mn*+ become inhibitory at concent,rations greater than 2 mM (i.e. at an ATP-metal ion concentration ratio > 2). A comparison of the pH optima for Mg2+ and NW+ shows the optimum for the former to be pH 8.0 and, for the latter, about pH 6.5. At the pH of the standard assay, i,e. pH 8, Co2+ is somewhat more active than Mg*+ or Mn*+.

As indicated previously (Table VI and Ref. 4), biotin carboxylase-catalyzed carboxylation of free d-b&in is markedly activated by ethanol. Activation is dependent on ethanol concentration reaching a maximum of lo-fold at approximately 15 volumes ri;, (Table VII) ; at higher concentrations, e.g. 20 volumes 7& inactivation of the enzyme occurs. It is evident (Table VII) f rom the diversity of organic solvents which proved effective that activation is not a specific property of ethanol.

6641

However, at comparable concentrations, ethanol produced rate enhancements greater than any solvent tested. Among the active solvents were methanol, ethanol, I- and Z-propanol, ethylene glycol monomethyl ether, acetone, 1 +dioxane, and tetrahydrofuran ; solvents having the higher dielectric constants were inactive. As with ethanol, several solvents caused irre- versible inactivation of the enzyme above a critical concentration.

Like the carboxylation of free biotin, the biotin carboxylase- catalyzed, carbamyl phosphate : ADP phosphotransferase reaction, which is also dependent on free d-biotin, is activated by ethanol (41) + However, in the latter case, the magnitude of activation (about 2%fold with 10 volumes G/ ethanol) is lower. In both instances (13), t’he principal kinetic effect of activation is on the maximal velocity and not on the K, values for substrates or cofactors. Thus, rate enhancement appears to be due to an effect on the reaction of bound substrate. Unlike the above model reactions which utilize free biotin as HCO, acceptor, two other biotin carboxylase-catalyzed reactions, i.e. the carboxylation of acetyl-CoA or of carboxyl carrier protein which involve subunit interaction with the carboxyl transferase or

carrier protein components, or both, respectively, are not activated by organic solvents.

Since changes in the hydrophobic character of the medium are known to alter the tertiary-quaternary structures of enzymes, the possibility btras considered that a productive con- formational change may be associated with aggregation of the enzyme. Thus, the effect of ethanol on the state of aggregation of the carbosylase was assessed by sucrose density gradient centrifugation under standard assay conditions, but at lower temperature. Preliminary experiments revealed that the homogeneous enzyme (10 pg per ml of assay mix) was completely stable for 24 hours in the complete assay reaction mixture containing 10 volumes 7C ethanol. In the density gradient experiments, avidin-[ 14C] biotin complex (avidin-biotin,) was included as in-

ternal marker to correct for the effect of ethanol in the viscosity and density of the medium which in turn affect sedimentation velocity. Ten volumes per cent ethanol was found to decrease the sedimentation rate of both avidin-biotin complex and biotin

T.,~~LF: VII

E$ert of orga~uk solvents on biotizl carboxylase activity

The biotin carboxyjase-clat~,zlyzed carboxylation of free d-biot)in was determined by the spectrophotometric method (see “Assay Methods”). Assays were completed within 3 min to minimize inactivation by the organic solvents; except as indicated, carboxylation rates were linear for at least 2 min.

Addi t iona

None (control) Methanol Ethanol

1-Propanol

2-Propanol

- Volumes per cent

10 r

1:

15 20

5 10

5 10 20

Relative activityb Addi t ionQ

%

100 330 200 680

1000 Inactivationc

310 Inactivationc

190 360

Inactivationc

Ethylene glycol monomethyl ether

Acetone 1,4-1Xoxane

Tetrahydrofuran

Volumes per cent

10 20 30 10 10 20 5 7

10

Relative activity b

%

210 480 540 300 200 460 360 460

Inactivationc

a 1,ZPropylene glycol, glycerol, and dimethyl formamide did not activate. b Activity relative to control with no addition. c Biotin carboxylase inactivated in “mock” preliminary incubation under these conditions, but without substrate.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

6642

carboxylase. Using the mobilities ratio method of Martin and Ames (42) and 4.58 S as the ~20,~ for avidin (l), the reference protein, sedimentation coefficients of 5.8 and 5.9 S, respectively, were calculated for biotin carboxylase in the presence and absence of ethanol. Therefore, ethanol at a concentration capable of promoting a g-fold activation causes neither polymerization nor depolymerization of the enzyme under assay conditions.

Properties of Carboxyltransjerase Component-The procedure outlined for purifying the transferase results in a 500-fold increase in specific enzyme activity from the ammonium sulfate stage (Table II, Step 2). The presence in the cell-free extract of an inhibitor which is removed by ammonium sulfate fractionation (Step 2), precludes use of the standard transferase assay until after this point in the fractionation. Based on a modified carboxyltransferase assay (pH 4.5 instead of pH 8.0 and the use of d-biotin in place of its methyl ester as substrate), it is estimated that the ammonium sulfate fractionated enzyme is purified approximately 3-fold over the cell-free extract. Thus, the over-all purification is approximately 1500-fold.

As indicated earlier, two peaks of carboxyltransferase activity are resolved by phosphocellulose chromatography (Fig. 2). Since Peak Z comprises greater than 80% of the total transferase activity, this fraction is used for further purification. A narrow cut of Peak Z fractions is made to avoid cross-contamination with Peck ZZ; hence, the yield at this step is not maximal. Ki- netic studies at pH levels of 4.5, 6.8, and 8.0 revealed no differ- ences in the ratios of activities nor in the ratios of carboxyltransferase to malonyl-CoA decarboxylase activities of the two fractions (Peaks Z and II). Both fractions are free of biotin carboxylase and carboxyl carrier protein and possess equal activity in reconstituting acetyl-CoA carboxylation in the presence of biotin carboxylase and carboxyl carrier protein. The sig- nificance of the occurrence of two forms of the transferase remains obscure.

Like the avian liver acetyl-CoA carboxylase (43), the carboxyltransferase component of the E. coli acetyl-CoA carboxylase system catalyzes a biotin-independent decarboxylation of malonyl-CoA. As illustrated in Table II, the ratio of carboxyltransferase to malonyl-CoA decarboxylase activity remains essentially constant, about 25 to 30, during purification. In view of the fact that the E. coli carboxyltransferase does not contain biotin and that avidin does not block malonyl-CoA decarboxylation catalyzed by the liver carboxylase (43), it is evident that labiliza- tion of the cr-carboxyl group of malonyl-CoA does not require participation of the biotinyl prosthetic group as previously be- lieved (1).

Several criteria indicate that carboxyltransferase purified by the procedure described is homogeneous. First, the elution pro- file from the terminal DEAE-cellulose chromatographic step in the purification shows close correspondence between transferase activity and protein. Also, as illustrated in Fig. 3 (Gel B), acrylamide gel electrophoresis of the enzyme yields a single- stained protein band which exactly coincides with carboxyltransferase activity. A further indication that the transferase preparation is monodisperse lies in the fact that plots of log, Rayleigh fringe displacement versus radius2 from sedimentation equilibrium experiments are linear.

The molecular weight of the native transferase was assessed both by gel filtration and sedimentation equilibrium methods. Prior to the molecular weight determinations, the enzyme was dialyzed for at least 30 hours against 50 mM potassium phosphate buffer, pH 7.0, containing 0.1 mM EDTA and 5 mM fl-mercaptoethanol. Gel filtration was carried out using a Sephadex G-200

FIG. 3. Polyacrylamide gel electrophoresis patterns of sodium dodecyl sulfate-dissociated (A) and native (B) carboxvltrans- ferase: For A, carboxyltransferase samples were incubated at 30” for 3 hours in 0.01 M sodium nhosnhate buffer. nH 7.0. containing 1% sodium dodecyl sulfate and 1% 2-mercaptoethanol. After incubation the enzyme solutions were dialyzed overnight against 1 liter of 0.01 M sodium phosphate buffer, pH 7.0, containing 0.1% sodium dodecvl sulfate and 0.1% 2-mercantoethanol at room temperature. IX”alyzed enzyme (21, pg) in ‘a solution containing tracking dye, 2-mercaptoethanol, and glycerol was applied to 10% acrylamide gels (44). After applying a constant current of 8 ma per gel for 454 hours at room temperature, the gels were stained with 0.25yo Coomassie brilliant blue. The subunit masses of the two polypeptide chains of the carboxyltransferase (Gel A) were estimated to be 35,000 and 30,600 by comparing their electrophoretic mobilities with those of standard proteins as described by Weber and Osborn (44). The subunit masses of the marker proteins used were: ovalbumin dimer, 90,000, and monomer, 45,000; bovine serum albumin, 68,000; ribulose diphosphate carboxylase heavy chain, 55,000; and light chain, 12,000. For B, approximately 40 pg of carboxyltransferase were applied to each of two 6% acrylamide gels (36). After electrophoresis with 50 mM potassium phosphate buffer, pH 7.5, for 7 hours at a constant current of 4 ma per gel, one gel was stained with Amido black and its companion gel was sliced into uniform segments to locate carboxyltransferase activity. Each slice was eluted overnight at 4” with 0.2 ml of 50 mM triethanolamine (Cl-) buffer. uH 8.0. containing 2 mM dithiothreitol and 0.24 me of bovine serum albumin. Carboxyltransferase activity was determined by the radioactive assay method.

column previously equilibrated with dialysis buffer and cali- brated with proteins of known molecular weight. A molecular weight of 145,000 was estimated from plots of molecular weight versus K,, (45). This value agrees reasonably well with that (Mm = 130,000 + 3,000) determined by the more accurate sedimentation equilibrium method. Sedimentation equilibrium runs were conducted in duplicate at 10,000 and 12,000 rpm with a Spinco model E analytical ultracentrifuge equipped with elec- tronic speed control. Experiments were performed at 20” with protein concentrations of 0.5 to 0.7 mg per ml; calculations were made according to Roark and Yphantis (46) from Rayleigh interference patterns obtained after 24 hours at 10,000 rpm and after an additional 12 hours at 12,000 rpm using an estimated partial specific volume of 0.73. Base-line corrections of interference patterns for cell distortion were made at speed after replacing the sample with the dialysis buffer without disassembling the cell.

The dissociation of homogeneous transferase with sodium dodecyl sulfate followed by electrophoresis on sodium dodecyl sulfate-polyacrylamide gels gives rise to two stained protein bands (Bands A and B, left and right, respectively) as shown in Fig. 3 (Gel A). This indicates that the enzyme is composed


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

6643

of two polypeptide chains of different size. The intensities of the stained bands were analyzed using a Gilford gel scanner and the relative mass ratio estimated from the areas under the two peaks on gel scan traces. The mass ratio estimated in this manner was approximately 1 to 0.78 for Peaks A and B, respectively. Comparison of the electrophoretic mobilities of the two traneferase polypeptides on sodium dodecyl sulfate-10% acrylamide gels with those of marker polypeptides (see Fig. 3) shows their molecular weights to be 35,000 (Band A) and 30,000 (Band B), respectively. In view of the relatively good agreement between the mass ratio of 1:0.78 (A: B) and the polypeptide weight ratio of 1:0.86 (A: B), it appears that the two subunits are present in nearly equivalent amounts. Thus, a subunit composition for native carboxyltransferase (130,000 daltons) of A~BB is suggested.

Purity and Properties of Carboxyl Carrier Protein-Affinity chromatography on Sepharose-avidin (monomer) of E. coli car- boxy1 carrier protein preparations that had been partially purified by fractionation with ammonium sulfate, calcium phosphate gel, and guanidine HCl yielded several fractions of high purity (Table III). Since Sepharose-avidin (monomer) possesses multiple types of binding sites which differ in their affinities for biotin, it was anticipated that the extent of release of the carrier protein from the affinity column would depend upon the nature of the eluant ; thus three fractions of carrier protein were obtained (Fractions II B, C, and D, Table III). The fraction eluted with acidic guanidine HCl (Fraction II D) and employed in most experiments described in the accompanying paper (13)) possessed the highest specific activity in the malonyl-CoA-[‘4C]acetyl-CoA exchange assay and contained approximately 30 nanoequiva- lents of active site (biotinyl prosthetic groups) per mg of protein. There is good agreement between these parameters (malonyl- CoA-acetyl-CoA exchange activity and biotin content) when applied to preparations of greatly differing purity (Table III). Moreover, the number of biotinyl prosthetic groups determined by direct carboxylation of the carrier protein preparation in the presence of Mg*+ATP and [14C]bicarbonate (29 nmoles per mg of protein; Table III, Fraction II D) closely approximates the biotin content assessed by the differential avidin-biotin binding method (32 nmoles per mg of protein; Table III, Fraction II D). Electrophoresis of Fraction II D on 14.5(r, polyacrylamide gels (Fig. 4, Gel E) revealed two stained protein bands which coin- cided precisely with carboxyl carrier protein activity measured on a companion gel. That both major bands contained covalently bound biotin is evidenced by the fact that prior treatment of the carrier protein preparation with avidin caused the dis- appearance of the two bands upon electrophoresis, as well as the appearance of a new band(s) at the top of the gel (Fig. 4, Gel D). Avidin, being a basic protein, would be expected to retard the rate of electrophoretic migration of carboxyl carrier protein(s) in this system; avidin alone, with or without bound biotin (Fig. 4, Gels A and B), moves to the cathode in this system and, therefore, does not appear on the gels. Importantly, the electrophoretic mobility of carboxyl carrier protein is un- affected by exposure to avidin previously treated with excess d-biotin (Fig. 4, Gel C). This indicates that the interaction between avidin and the two major stained bands is specific for the biotinyl prosthetic group.

I 2 3 4 5 6 7 8 9 IO Distance from Gel Top, cm

FIG. 4. Polyacrylamide gel electrophoresis of purified Esche- richia coli carboxyl carrier protein. The following were applied to polyacrylamide gels (14.5y0 acrylamide) in 0.1 ml containing 10% glycerol after a 30-min preliminary incubation: Gel A, 162 pg of avidin pretreated with 40 nmoles of d-biotin; Gel B, 162 fig of avidin; Gel C, 27 fig of carboxyl carrier protein (Fraction II, D, Table III) plus 162 pg of avidin pretreated with d-biotin as above; Gel D, the same as Gel C, but with d-biotin omitted; and Gel E (in duplicate), 27 pg of carboxyl carrier protein (Fraction II, D, Table III). After electrophoresis with 50 mM Tris-glycine buffer, pH 8.9, at a constant current of 3 ma per gel, one of the E gels was sliced into uniform segments for carboxyl carrier protein assays. The gel slices were eluted overnight at 4” in imidazole (Cl-) buffer. nH 6.8. containing 2 mM dithiothreitol and 0.6 mg of bovine serum albumin per ml and was then assayed. The assav reaction mixture contained (total volume. 0.26 ml): 50 rnM imid”azole (Cl-) buffer, pH 6.8; 5 mM MnC12; 1 mM ATP; 13.5 mM KHCOI (11,000 cpm per nmole); 0.3 mM acetyl-CoA; biotin carboxylase, 4.9 pg; and carboxyltransferase, 9 pg. The reaction was stopped after 10 min at 30” by the addition of 0.1 ml of 6 N

HCl and [Wlbicarbonate incorporated into acid-stable form (malonyl-CoA) determined as previously described (8). was accomplished with Amido black.

Staining

in approximately equal amounts as judged by the intensity of staining. Presumably the two carboxyl carrier proteins cor- respond to the lO,OOO- and 22,000-dalton biotin-containing proteins described by Vagelos (14-16), the smaller of the two being derived from the larger via proteolysis by a protease in the cell- free extract. This is also consistent with the finding (15, 16) that the larger and apparently native form of carboxyl carrier protein is considerably more active than the smaller form in the acetyl-CoA carboxylation assay. As will be demonstrated in the accompanying paper (13) the purified carboxyl carrier protein(s), in addition to being essential for acetyl-CoA carboxylation (Table IV) and malonyl-CoA-[r4C]acetyl-CoA exchange (Table V), is required for ATP-[“C]ADP- and ATP-[32P]Pi- exchange catalyzed by biotin carboxylase.

It is estimated that the slower-moving protein band is 10 to 20 times more active than the rapidly migrating band in support- ing acetyl-CoA carboxylation in the presence of purified biotin carboxylase and carboxyltransferase (see activity plot in Fig. 4) ; this is based on the fact that both carrier proteins are present

Although Fraction II B and C (Table III) have somewhat lower specific activities than Fraction II D, both exhibit the same relative amounts of the two major biotin-containing bands upon polyacrylamide gel electrophoresis. In addition, several other protein bands of slower mobility which do not contain biotin (results not shown) are detectable.

DISCUSSION

The acetyl-CoA carboxylases of animal cells (10) and E. coli (2-6), are multisubunit enzymes whose component polypeptide


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

6644

chains differ. Thus, Gregolin et al. (9) found that the basic 410,000-dalton protomer of the avian liver carboxylase to be comprised of four subunits of similar molecular weight, i.e. about 110,000. However, it appeared that these subunits were nonidentical since only one biotinyl prosthetic group, one covalent bicarbonate-loading site, one acetyl-CoA binding site, and one tight citrate binding site were present per protomer (410,000 daltons). &Ioreover, it was established by affinity chromatography (47) that each protomeric unit possessed a single biotinyl prosthetic group; therefore, the protomers per se are most likely iden tical. It became evident using electrophoresis of sodium dodecyl sulfate-dissociated carboxylase on lightly loaded acrylamide gels (5 pg of protein per gel) that two subunit weight classes differing by only about 5,000 to 10,000 daltons could be observed with the rat liver carbosylase (12). More recently, with high resolution acrylamide gel clectrophoresis in 6 M urea- 0.1 y0 sodium dodecyl sulfate, the avian liver carboxylase has been resolved into subunits of three weight classes (48), t,he approximate mole ratios of the three species, i.e. 117,000, 130,000, and 140,000 daltons, being 2: 1: 1, respectively. The covalently bound biotin prosthetic group is associated with the 117,000- dalton subunit(s) (48). This and cross-linking analysis of the avian liver carboxylase protomer by the dimethylsuberimidate method (49) show that this enzyme is composed of four different types of polypeptide chains (48). Unfortunately, attempts to assign function to the individual polypeptide chains of the animal carboxylases have met with little success; in no case has it been possible to detect or restore catalytic activity for any partial reaction following dissociation or resolution, or both.

In contrast to the avian liver carboxylase whose subunits are dissociated only by drastic treatment (6 M guanidine HCl or sodium dodecyl sulfate) (9, 10, 12), those of t,he E. coli acetyl- CoA carboxylase dissociate readily. Moreover, unlike the dissociated subunits of the avian enzyme which lose activity irreversibly by this treatment, the resolved components of the E. coli system (i.e. biot,in carboxylase, carboxyltransferase, and carboxyl carrier protein) retain enzymatic activity allowing the assignment of function to each. The three protein components of the E. coli acetyl-Cob carbosylase have been obtained in homogeneous form and their respective molecular characteristics determined : biotin carboxylase and carboxyltransferase in our laboratory (Refs. 4 and 5, a,nd this paper) and carboxyl carrier protein in Vagelos’ laboratory (14-17). These properties are summarized in Table VIII. Biotin carboxylase, the catalytic

TABLE VIII

Molecular properties of protein cornpo~e~~ts of Escherichia coli acetyl-CoA carboxylase system

Biotin carboxylase”

Molecular weight. ...... 98,000

Subunit(s) weight A ..................... 51,ooo B .....................

Subunits/molecule. ..... Afl Biotin content. ......... None

Carboxyl t ransf erasea

130,000

35,000 30,000 AZ& None

- Carboxyl carrier

proteid

44,ooO (or higher order aggre - gates)

22,000

A2 1.0 per 22,OOO dal-

tons

a Results presented in this paper or Refs. 4 and 5. b Results from Vagelos’ laboratory (14-16).

element responsible for the first half-reaction (Reaction 1), is a dimer composed of apparently identical 51 ,OOO-dalton subunits. The catalyst for the second half-reaction (Reaction 2), carboxyltransferase, appears to be a tetramer having an A& structure and subunits with molecular weights of 30,000 and 35,000. Finally, the carboxyl carrier peptide which possesses the covalently bound prosthetic group has been found in Vagelos’ laboratory (15-17) to have a molecular weight of 22,000. Thus, at least four polypeptide chains are implicated in the carboxylation of acetyl-CoA and perhaps the regulation of this process in E. coli (21). This is interesting in view of the finding that four apparently nonidentical subunits comprise the basic protomeric unit of the avian liver acetyl-CoA carboxylase (10) l In the case of the carboxylase system from K co&, it has not been possible to demonstrate unequivocally the existence of a complex of the protein components (biotin carboxylase, carboxyltransferase, and carboxyl carrier protein), whereas all of the acyl-CoA carboxylases from animal cells have stable quaternary structures (1, lo),

Tentative evidence for an interaction of the three components of the E. coli carbosylase system, i.e. in a ternary complex, derives from the observation (Table V) that biotin carboxylase, although not an essential participant in the carboxyl carrier protein- and carboxyltransferase-dependent malonyl-CoA-[‘4C]- acetyl-CoA exchange, activates this reaction. This suggests that the binding of biotin carboxylase to either carboxyltransferase or the carboxyl carrier protein stabilizes a ternary complex of the three components.

Acknowledgment-We thank Mr. Eberhard Zwergel for out- standing technical assistance.

1. Moss J., AND LANK, M, 1). (1971) Advan. Enxymol. 36, 321 2. ALIWRTS, A. W., ;IND V:IGE:LOS, P. I<. (1968) Proc. Nat. Acad.

Sci. U. S. A. 59, 561 3. hIH5ItTS, A. W., &XVI, A. M., AND VAGIILOS, P. R. (1969)

Proc. Nat. Acad. Sci. U. S. A. 63, 1319 4. DIMROTH, P., GUCHHMT, R. B., STOLL, fs., AND L,~NE, M. L).

(1970) Proc. Nat. Acad. Sci. U. S. A. 67, 1353 5. GUCHHAIT, 11. B., Moss, J., SOTCOLSKI, W., .~ND LANE, M. D.

(1971) Proc. iVat. Acad. Sci. LT. S. A. 68, 653 6. ALHERTS, A. W., GORDON, S. G., AND VAGI~LOS, P. R. (1971)

Proc. Nat. Acad. Sci. U. S. A. 68, 1259 7. GRKGOLIN, C., RYDER, E., W~~RNII:R, I<. C., KLIXNSCHMIDT,

A. K., AND LANE, M. I). (1966) Proc. ilrat. Acad. Sci. U. S. A. 66, 1751

8. GRXGOLIN, C., RYDER, E., AND LANE, M. 11. (1968) J. Biol. Chem. 243, 4227

9. GREGOLIN, C., RYDER, E., Wm~m, It. C., KLEINSCHMIDT, A. K., CHANG, H.-C., AND LANE, M. I). (1968) J. BioZ. Chem. 243, 4236

10. LANE, M. T)., Moss, J., AND POLAKIS, s* E. (1974) Current Topics in Cellular Regulation (HORIXICF:K, B. L., MD STADT- MAN, E. IL, eds) Vol. 8, p, 139, Academic Press, New York

11. Moss, J., YAMAGISHI, M., KLEINSCHMIDT, A, K., AND LANK, M. I>. (1972) Biochemistry 11, 3779

12. INOUN, I-I., AMI LOWENSTHN, J. M. (1972) J. Biol. Chem. 247, 4825

13. POLAKIS, 8. E., GUCHHATT, It. B., ZWI':RGEL, 13. II:., LANE, M, D., AND COOPER, T. G. (1974) J. BioZ. Chem. 249, 6657- 6667

14. NERVI, A. M., ALBERTS, A. W., ,~ND VAGELOS, P. R. (1971) Arch. Biochem. Biophys. 143, 401

15. FALL, It. It., NERVI, A. M., ALBERTS, A. W., AND VAGELOS, P. Xt. (1971) Proc. hbt. Acad. Sci, U. S. A, 68, 1512

16. FALL, R. R., AND VAGELOS, P. R. (1972) J. Viol. Chem. 247,8005 17. FALL, R. It., AND VAGELOS, P. R. (1973) J. i?ioZ. Chem. 248,

2078 18. LANE, M. D., AND Moss, J. (1971) Advan. Enzyme Regul. 9,237


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

6645

19. LANE, M. I)., AND Moss, J. (1971) in Metabolic Pathways 34. (VOGEL, H. J., ed) 3rd Ed, Vol. 5, p. 23, Academic Press,

GIJCHHAIT, R. B., POLAKIS, S. E., HOLLIS, D., FF,NSELAU, C., AND LANIC, M. D. (1974) J. Biol. Chem. 249,6646-6656

New York 35. DIMROTH, P., GUCHHAIT, R. B., AND LANE, M. D. (1971) Z. 20. POLAKIS. S. E., GUCHHAIT, R. B., AND LANE, M. D. (1973) Physiol. Chem. 362, 351

Abstraks of ihe 16th Int&natio&l Conference on the Bio- chemistry of Lipids, B, Nottingham, England

POLAKIS, S. E., GUCHHAIT, R. B., AND LANE, M. D. (1973) J. Biol. Chem. 248, 7957

36. 37.

21.

22. HASELTINE, W A., BLOCK, R., GILBERT, W., AND WEBER, K. (1972) Nature 238, 381

23. HASXLTINE, W. A., AND BLOCK, R. (1973) Proc. Nat. Acad. Sci. U. S. A. 70, 1564

24. 25. 26.

CASHEL. M.. AND GALLANT. J. 0969) Nature 221. 838 CASHEL: M.‘(1969) J. Biol.‘Chek. 244. 3133 LAZZARINI, R. A., CASHEL, M., AND GALLANT, J. (1971) J.

Biol. Chem. 246, 4381 27. KF,ILIN, D., AND HARTREN, E. F. (1938) Proc. Roy. Sot. Lond.

Ser. Biol. Sci. B. 124, 397 28. SIMON, E. J., AND SHEMIN, 11. (1953) J. Amer. Chem. Sot.

76, 2520 29. TR.4MS. E.. AND BRADY. R. 0. (1960) J. Amer. Chem. Sot. 82,

2972’ ’ 30. LOWRY, 0. H., ROSEBROUGH, N. J., FARR, A. L., AND RANDALL,

R. J. (1951) J. Biol. Chem. 193, 265 31. 32. 33.

WARHIJR~, O., AND CHRISTIAN, W. (1942) Biochem. 2.310,384 LAYNE, E. (1967) Methods Enzymol. 3, 447 CALLANAN, M. J., CARROLL, W. It., AND MITCHELL, E. R.

(1957) J. Biol. Chem. 229, 279

38.

39.

40.

41.

42.

43. 44. 45. 46.

47.

48.

49.

DAVIS, B. J. (1964) Ann. N. Y. Acad. Sci. 121,404 CUATRECASAS, P., AND ANFINSEN, C. B. (1971) Methods En-

2 ymol. 22, 345 LYNEN, F., KNAPPE, J., LORCH, E., J~~TTING,G., RINGELMANN,

E., AND LACHANCF,, J.-P. (1961) Biochem. Z. 336, 123 STOLL, E., RYDER, E., EDWARDS, J. B., AND LANE, M. I>.

(1968) Proc. Nat. Acad. Sci. U. S. A. 60,986 KNAPPE, J., RINGELMANN, E., AND LYNEN, F. (1961) Biochem.

Z. 336, 168 POLAKIS, S. E., GUCHHAIT, R. B., AND LONE, M. D. (1972)

J. Biol. Chem. 247, 1335 MARTIN, R. G., AND AMES, B. N. (1961) J. Biol. Chem. 236,

1372 Moss, J., AND LANE, M. 1). (1972) J. Biol. Chem. 247,4962 WEI)ER, K., AND OSBORN, M. (1969) J. Biol. Chem. 244, 4406 ANDREWS, P. (1970) Methods Biochem. Anal. 18, 1 ROARK, D. E., AND YPHANTIS, D. A. (1969) Ann. N. Y. Acad.

Sci. 164, 246 LANE, M. D., EDWARDS, J., STOLL, E., AND Moss, J. (1970)

Vitamins Hormones 28, 345 GUCHHAIT, R. B., ZWERGEL, E., AND LANDS;, M. 1). (1974) J.

Biol. Chem. 249, 4776-4780 DAVIES, G. E., AND STARK, G. R. (1970) Proc. Nut. Acad. Sci.

U. S. A. 66, 651


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Daniel LaneRas B. Guchhait, S. Efthimios Polakis, Peter Dimroth, Erwin Stoll, Joel Moss and M.

COMPONENTSCARBOXYLTRANSFERASE, AND CARBOXYL CARRIER PROTEIN

AND PROPERTIES OF THE BIOTIN CARBOXYLASE, : PURIFICATIONEscherichia coliAcetyl Coenzyme A Carboxylase System of

1974, 249:6633-6645.J. Biol. Chem.

http://www.jbc.org/content/249/20/6633Access the most updated version of this article at

Alerts:

When a correction for this article is posted•

When this article is cited•

to choose from all of JBC's e-mail alertsClick here

http://www.jbc.org/content/249/20/6633.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/content/249/20/6633

http://www.jbc.org/cgi/alerts?alertType=citedby&addAlert=cited_by&cited_by_criteria_resid=jbc;249/20/6633&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/249/20/6633

http://www.jbc.org/cgi/alerts?alertType=correction&addAlert=correction&correction_criteria_value=249/20/6633&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/249/20/6633

http://www.jbc.org/cgi/alerts/etoc

http://www.jbc.org/content/249/20/6633.full.html#ref-list-1

http://www.jbc.org/

acetyl coenzyme a carboxylase system of escherichia coli

Documents