THE JOURNAL OF Bmux~cn~ CHEMISTRY Vol. Z,54, No. 5, Issue of March 10, pp. 1586-1593. 1979 Printed in IJ.S.A.

Purification and Characterization of Cytochrome c Oxidase from Rat Liver Mitochondria*

(Received for publication, August 12, 1977, and in revised form, August 14, 1978)

Ralph J. Rascati.$ and Peter Parsons8

From the Department of Biochemistry, University of Massachusetts, Amherst, Massachusetts 01003

Cytochrome c oxidase has been purified from rat liver mitochondria using affinity chromatography. The preparation contains 10.5 to 13.4 nmol of heme a + a3 per mg of protein and migrates as a single band during polyacrylamide gel electrophoresis under nondissociat- ing conditions. It has a heme a/as ratio of 1.12 and is free of cytochromes b, c, and cl as well as the enzymes, NADH dehydrogenase, succinic dehydrogenase, coen- zyme Q-cytochrome c reductase, and ATPase. The en- zyme preparation consists of six polypeptides having apparent M, of 66,000,39,000,23,000,14,000,12,500, and 10,000 as determined by sodium dodecyl sulfate-poly- acrylamide gel electrophoresis. The peptide composi- tion is similar to those found for cytochrome c oxidases from other systems. The enzymatic activity of the pu- rified enzyme is completely inhibited by carbon mon- oxide or cyanide, partially inhibited by Triton X-100 and dramatically enhanced by Tween 80 or phospholip- ids.

Our studies have been prompted by an interest in the cellular site of synthesis of mitochondrial proteins, particularly in mammalian systems. Studies based on specific inhibition of either cytoplasmic protein synthesis or mitochondrial protein synthesis using whole cells of lower systems have now indi- cated that at least three mitochondrial proteins, cytochrome c oxidase, oligomycin-sensitive ATPase, and cytochrome b are synthesized in part on cytoplasmic ribosomes and in part on mitochondrial ribosomes (1). While these experiments have been done very carefully under well controlled conditions, there remains the possibility that erroneous results might occur due to nonspecific or indirect effects of such inhibitors (1, 2). We, therefore, carried out our studies using isolated mitochondria to avoid this problem and to determine whether the isolated organelle could carry out the synthesis of com- pleted mitochondrial peptides. Cytochrome c oxidase was studied because preliminary findings in our laboratory indi- cated that isolated rat liver mitochondria could synthesize polypeptides associated with this enzyme.

Our approach to the study of the biosynthesis of cytochrome c oxidase has been to: (a) purify and characterize the enzyme,

* This investigation was supported by grants from the National Institutes of Health (GM19763) and Research Corporation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “adver- tisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ Present address, Biology Division, Oak Ridge National Labora- tory, Oak Ridge, TN 37830.

0 Joint appointment: Faculty Research Associate, Department of Biochemistry, University of Massachusetts, Amherst, MA 01003, and Associate Professor of Biochemistry, Department of Biological Sci- ences, Mount Holyoke College, South Hadley, MA 01075.

(b) prepare specific antibodies to the enzyme, (c) incubate isolated mitochondria with radioactive leucine, (d) treat sol- ubilized mitochondrial proteins labeled in this manner with antibodies to cytochrome c oxidase, and (e) analyze the im- munoprecipitates by SDS’-polyacrylamide gel electrophoresis to determine whether any newly synthesized, radioactively labeled cytochrome c oxidase peptides are present.

This paper describes the purification and characterization of the cytochrome c oxidase. Purification procedures for this enzyme have been reported for bovine heart (3-g), yeast (3, lo), Neurospora crassa (ll), Locusta migratoria (12), and rat liver (13). However, in our hands, the procedures available when we initiated this work yielded preparations which were low in heme a + a3 content and exhibited a large number of bands on SDS-polyacryIamide gel electrophoresis. We have developed a procedure including affinity chromatography on cytochrome c-Sepharose which results in an active enzyme of very high heme a + u3 content. Ozawa et al. (14) have used a similar affinity column to obtain cytochrome c oxidase with high heme a content from beef heart mitochondria. Our enzyme has been charape2ized with respect to subunit com- position and various otl@r-ghysical, chemical, and enzymatic properties. Experiment; detailing the preparation of specific antiserum against cytochrome c oxidase and the mitochon- drial biosynthesis of some of the peptides of this enzyme are presented in the following paper (15). Some of this work has been reported previously in preliminary form (16).

EXPERIMENTAL PROCEDURES

Materials

White Sprague-Dawley rats of both sexes used for all experiments were obtained from Holtzmann Laboratories, Madison, WI. Horse heart cytochrome c type III, poly(L-lysine) (30,000 daltons), N,N,N’,N’-tetramethylethylenediamine, glycine, sodium dodecyl sul- fate, Tween 80, Triton X-100, and twice crystallized bovine serum albumin were obtained from Sigma. Sephadex G-15 and Sepharose 4B were obtained from Pharmacia. Methylenebisacrylamide and am- monium persulfate were obtained from Bio-Rad. Cyanogen bromide and acrylamide were purchased from Eastman Organic. Whatman DE52 DEAE-cellulose was purchased from Reeve Angel. Coenzyme Q10 was a generous gift from Merck, Sharpe and Dohme Laboratories. Dichloroindophenol and phenazinemethosulfate were kindly supplied by Dr. R. C. Fuller, Department of Biochemistry, University of Massachusetts, Amherst, MA.

Preparation of Chromatographic Resins

Cytochrome c-Sepharose-Sepharose 4B was activated by reaction with cyanogen bromide (CNBr) according to the procedure of Cua- trecasas (17). Twenty milliliters of packed resin was diluted with 20 ml of distilled water and 2 g of CNBr were added. The reaction was maintained at pH 11.0 by addition of 2 N NaOH, dropwise at 20°C until proton release stopped. The activated resin was washed rapidly

’ The abbreviations used are: SDS, sodium dodecyl sulfate; SMP, submitochondrial particles.

1586

by guest on June 6, 2020http://w

ww

.jbc.org/D

ownloaded from

Studies on Cytochrome c Oxidase from Rat Liver Mitochondria 15x7

by suction filtration with 20 volumes of 0.2 M sodium citrate, pH 6.5. Twenty milliliters of cytochrome c (4 mg/ml) in 0.2 M sodium citrate, pH 6.5, was added to the activated washed resin and the mixture was gently stirred at 0-4°C for 16 to 24 h to allow covalent linkage of the cytochrome c to the Sepharose.

DEAE-cellulose-Whatman DE52 microgranular, preswollen resin was equilibrated in 20 mM sodium phosphate, pH 7.0, containing 1% Triton X-100.

Poly(~-lysine)Sepharose-Sepharose 4B was prepared as outlined above for preparation of cytochrome c-Sepharose with the following exceptions: 0.1 M sodium carbonate, pH 8.5 was used to wash the CNBr-activated resin and the poly(L-lysine) (10 mg/ml) was dissolved in this same buffer.

Sephadex G-I5-Sephadex G-15 was swollen and equilibrated in 20 mM sodium phosphate, pH 7.0, containing 1% Tween 80.

Purification of Cytochrome c Oxidase

The following procedure is for 1 g of intact mitochondrial protein. All steps were performed at 0-4°C.

Preparation of Mitochondria-Mitochondria were prepared by the method of Parsons and Simpson (18) using a 4% homogenate to improve the yield of organelles. Mitochondria were washed four times with 0.25 M sucrose, 1 mM EDTA, pH 7.5, and suspended in 50 mM

potassium phosphate, pH 7.0, at a final concentration of 20 mg of mitochondrial protein/ml.

Preparation of Submitochondrial Particles-Submitochondrial particles (SMP) were prepared by sonication of mitochondria (10) for 1 min at maximum power in the Branson Sonifier-S-175. The broken mitochondria were centrifuged for 1 h at 100,060 x g in a Spinco type 30 rotor. The supernatant which contained no cytochrome oxidase was discarded and the pellet (SMP) was resuspended in 0.25 M sucrose at a final concentration of precisely 20 mg/ml.

Solubilization of Cytochrome Oxidase-Twenty per cent sodium cholate was added to a final concentration of 3 mg/mg of SMP protein. Ammonium sulfate (saturated at 0°C) was then added drop- wise in a ratio of 0.34 ml/ml of SMP solution which brought the final concentration of ammonium sulfate in the mixture to 25% saturation. Gentle stirring for 2 h followed by centrifugation for 15 min at 40,000 x g in a Sorvall SS-34 rotor yielded a supernatant (S1) which con- tained all the cytochrome oxidase.

Precipitation of Cytochrome Oxidase-The S1 fraction was brought to 45% saturation with ammonium sulfate by mixing directly with 0.37 ml of saturated ammonium sulfate (O’C) per ml of S,. This mixture was centrifuged for 15 min at 40,000 x g in a Sorvall SS-34 rotor. The supernatant (SZ) was discarded and the oily, dark green pellet (P2) was dissolved in a minimum volume of 20 mM sodium phosphate (pH 7.0), 1% Tween 80.

Desalting-The PP fraction was desalted on a Sephadex G-15 column (1.5 x 30 cm) equilibrated in the same buffer used to dissolve P2. All colored eluate fractions were combined and diluted lo-fold with elution buffer (G-15 eluate).

Cytochrome c-Sepharose Chromatography-The G-15 eluate was applied to a cytochrome c-Sepharose affinity column (1.5 x 25 cm) equilibrated with 20 mM sodium phosphate (pH 7.0), 1% Tween 80. After washing successively with 3 column volumes each of 20 mM

sodium phosphate (pH 7.0), 1% Tween 80, and 200 mM sodium phosphate (pH 7.0), 1% Tween 80, the cytochrome oxidase was eluted with 100 mM sodium phosphate (pH 7.0), 1% Triton X-100. The green- colored fractions were combined and diluted 5-fold with 1% Triton X- 100.

DEAE-cellulose Chromatography-The diluted eluate from the previous column was applied to a DEAE-cellulose column (1.5 x 7 cm) equilibrated with 20 mM sodium phosphate (pH 7.0), 1% Triton X-100. The column was washed successively with 2 column volumes of 20 mM sodium phosphate (pH 7.0), 1% Triton X-100, and 50 mM

sodium phosphate (pH 7.0), 1% Triton X-100. The latter buffer usually causes spreading of the green cytochrome c oxidase on the resin. Occasionally, some oxidase is eluted by this buffer but this is of low purity and is discarded. The enzyme is eluted with 100 mM sodium phosphate (pH 7.0), 1% Triton X-100. The eluted colored fractions are combined and diluted 5-fold with distilled water.

Poly(L-lysine)-Sepharose Chromatography-The diluted eluate from the previous column was applied to a poly(L-lysine)-Sepharose column (0.9 x 4 cm) equilibrated with 20 mM sodium phosphate (pH 7.0), 0.1% Triton X-100. The column was then washed successively with 3 column volumes each of 20 mM sodium phosphate (pH 7.0), 0.1% Triton X-100, and 40 mM sodium phosphate (pH 7.0), 0.1%

Triton X-100. Further washing was carried out with at least 1 column volume of 80 mM sodium phosphate (pH 7.0), 0.1% Triton X-100, and until the oxidase spread to approximately the middle of the column. The enzyme was eluted with 160 mM sodium phosphate (pH 7.0), 0.1% Triton X-100.

The purification scheme is outlined in Fig. 1.

Enzyme Assays

Cytochrome c Oxidase Actiuity-Cytochrome c oxidase activity was assayed at 23°C by a modification of the procedure of Smith and Conrad (19). A l-ml assay contained 40 mM potassium phosphate (except where otherwise noted), pH 7.0, 0.5% Tween 80, 30 pM

ferrocytochrome c, and enzyme. Ferrocytochrome c was prepared by adding a few grains of sodium dithionite to a 1% solution of cyto- chrome c in assay buffer followed by aeration to remove excess dithionite. The activity of the enzyme is expressed as the first order rate constant. An extinction coefficient, AE,,I ,l,“, of 18.5 mM-’ cm-l was assumed (19).

Assays of Enzymatic Impurities- 1. NADH dehydrogenase was measured by the procedure of

DeBernard (20). Reduction of ferricyanide was determined in a l-ml assay containing 40 mM potassium phosphate, pH 7.5, 0.5 mM ferri- cyanide, and 0.3 mM NADH. An extinction coefficient, AE,lo ,,m, of 1.0 mM-’ cm-’ was used (20).

2. Coenzyme Q-cytochrome c reductase was measured by a modi- fication of the procedure of Rieske (21). Reduction of cytochrome c was monitored in a l-ml assay containing 40 mM potassium phosphate. pH 7.0, 0.1 mM EDTA, 20 mM sodium azide, 0.08% bovine serum albumin, 50 pg of reduced coenzyme QIO, and 210 pM cytochrome c. An extinction coefficient, AEs5, ““,, of 18.5 mM--’ cm-’ was assumed.

3. Succinic acid dehydrogenase was assayed by the method of King (22). Dichloroindophenol reduction was measured in a l-ml assay containing 50 mM potassium phosphate, pH 7.8, 1.5 mM sodium cyanide, 20 mM sodium succinate, 0.5 mM dichloroindophenol, 1 mg of bovine serum albumin, and 1 mru phenazinemethosulfate. An extinction coefficient, AE 5w “,,,, of 2.0 rnM- ’ cm -’ was used (22).

4. ATPase activity was measured indirectly by the ability of anti- bodies to the most highly purified fraction of cytochrome c oxidase (see Ref. 15) to inhibit the conversion of [‘H]ATP to [:‘H]ADP by

FIG. 1. Purification scheme for cytochrome c oxidase. Abbrevia- tions used are: HSS, high speed supernatant; SMP, submitochondrial particles; S, supernatant; P, pellet; TW, Tween 80; TX, Triton X-100; Nap,, sodium phosphate.

by guest on June 6, 2020http://w

ww

.jbc.org/D

ownloaded from

1588 Studies on Cytochrome c Oxidase from Rat Liver Mitochondria

TABLE I Purification of cytochrome c oxidase

Enzyme activity, heme measurements, and abbreviations are described under “Experimental Procedures.” Cytochrome c, DEAE, and polylysine are eluates from the cytochrome c-Sepharose, DEAE-Sephadex, and poly(L-lysine)-Sepharose columns, respectively.

Fraction Protein Heme a + .j- Activity* Specific activity Heme recovery

w nmol nmol/mgprotein K (mini’) x IO-” K (min-’ mg-‘J K (min-’ nmot’) I, Mitochondria 1000 140 0.14 44.0 44 314.3 100 SMP 550 123 0.23 40.2 73 326.4 88 S, 292 152 0.56 49.3 169 324.6 108 Pa 285 160 0.58 37.3 131 233.1 114 G-15 200 124 0.62 38.0 190 306.4 88 Cytochrome c 15.6 76 5.00 11.4 729 149.6 54 DEAE 6.0 30 5.50 4.1 679 135.8 21 Polylysine 2.0 23 11.50 1.8 915 79.6 16.4

o Both methods for heme determination cited under “Experimental Procedures” gave similar results. ’ Activity is expressed as the first order rate constant = ln(AJAJ i protein (mg/ml).

sonicated or solubilized mitochondria. A l-ml assay contained 10 mM Tris, pH 7.4, 3 mM magnesium acetate, 20 ItIM KCl, 3 mM [“HIATP (0.8 Ci/mol), 7 to 8 mg (1 nmol of heme a + a& of sonicated mitochondria or mitochondria solubilized with 0.5% Brij 58 and 0.2 ml of antiserum. Incubation was carried out at 35’C. At zero time and after 15 min, 0.2.ml aliquots were treated with 40 pmol of EDTA to stop the reaction and the mixture was chromatographed on thin layer polyethyleneimine cellulose using 0.75 M potassium phosphate, pH 3.4, to separate ATP and ADP.

Analytical Procedures

Heme a + a3 Content-Oxidase preparations were analyzed for heme a + a3 content by two methods.

1. The difference in absorbance at 601 nm and 630 nm after reduction with a few grains of sodium dithionite. An extinction coefficient, AE ,601 nm 630 “In rrducedl, of 16.5 mM-’ cm-’ was assumed (19).

2. Pyridine hemochromogen formation as described by Horie and Morrison (23) and modified by Yubitsui (24). An extinction coefficient, AE,aLii ,,m _ 630 “,,, reduced), of 26 mM-’ cm-’ was used (23).

In addition, the heme a3 content was approximated from the difference in absorbance at 445 nm and 430 nm between reduced cytochrome oxidase and cytochrome oxidase reduced in the presence of carbon monoxide (CO). An extinction coefficient, AEc~~~ ““, ?:IO nm reduced iedr,Led + CO,, of 148 mM ’ cm I was assumed (25). The heme a/ a3 ratio = ((heme (a + aa) - heme azl)/(heme a:J).

Protein Determination-Protein was determined according to the procedure of Lowry et al. (26) in the presence of 0.4% sodium deoxycholate. Bovine serum albumin was used as a standard.

Determination of Total Iron and Non-Heme Iron-Total iron was determined by atomic absorption spectroscopy.” Non-heme iron was estimated as the difference between total iron and heme iron which was taken as equal to the concentration of heme a + ad.

Phospholipid Content-Cytochrome oxidase fractions obtained at various purification steps were dialyzed exhaustively against 10 mM Tris-HCl, pH 7.0, to remove all traces of phosphate buffer. Phospho- lipid was extracted with 2:l chloroform/methanol and samples were evaporated to dryness. Wet ashing was done by adding 0.05 ml of concentrated sulfuric acid and 0.05 ml of concentrated nitric acid and heating samples in a sand bath at 220°C for 1 h. Additional nitric acid was added until the samples were colorless. Samples were diluted to 1 ml with H20, neutralized with ammonium hydroxide, and analyzed for phosphate by the method of Chen et al. (27). The phospholipid equivalent was calculated by assuming a content of 1.25 pmol of phosphate/mg of phospholipid (28).

Polyacrylamide Gel Electrophoresis-Samples were dissociated by heating at 100°C for 5 min in the presence of 2.5% sodium dodecyl sulfate, O.l%l 2-mercaptoethanol, 10% glycerol, and bromphenol blue which was added as a tracking dye. They were analyzed at 20°C using cylindrical gels (0.6 x 10 cm) containing 12% acrylamide, 0.37% bisacrylamide, 0.3 M Tris-HCl, pH 8.8, and 0.5% SDS (10). A l-cm stacking gel containing 2.5% acrylamide, 0.15% bisacrylamide, 0.057 M Tris/phosphate, pH 7.2, and 0.5%~ SDS was routinely used. The electrophoresis buffer contained 0.052 M Tris/glycine, pH 8.8, and 0.03% SDS. Electrophoresis was carried out at 1 mA/tube until the

‘We wish to thank Dr. Gregory Dabkowski, Chemistry Depart- ment, University of Massachusetts, for these determinations.

tracking dye reached the running gel and 3 mA/tube thereafter. Gels to be stained were fixed overnight in 10% trichloroacetic acid, 20% methanol, stained for 2 h with 0.25% Coomassie brilliant blue R in 7.5% acetic acid, 50% methanol, and destained by diffusion in 7.5?’ acetic acid. The stained gels were scanned at 546 nm with a Gilford model 2410 linear transport attachment. M, were estimated using cytochrome c (lZ,SOO), lycozyme (14,500), hemoglobin (15,500), papain (23,500), chymotrypsinogen (27,500), aldolase (41,000), and catalase (60,000) as standards.

Electrophoresis in the absence of SDS was done essentially as just described except that SDS was omitted and all solutions contained 0.1% Triton X-100.

Isolectric Focusing-The p1 of cytochrome c oxidase was deter- mined by the procedure of Frater (29) using 3% (w/v) polyacrylamide gels containing 0.1% (v/v) Triton X-100. The oxidase was located by scanning gels at 424 nm.

RESULTS

Purification of Cytochrome c Oxidase

The results of the purification of cytochrome c oxidase according to the scheme given in Fig. 1 are presented in Table I. Purification of the enzyme will be discussed on the basis of its heme a + a3/protein ratio since heme is an obvious, relatively stable marker of cytochrome oxidase. Preparation of SMP from intact mitochondria as well as subsequent solu- bilization and ammonium sulfate fractionation of these parti- cles each results in a small purification of the enzyme; no heme a + a8 is lost up to this point. The use of cytochrome c- Sepharose as an affinity column resulted in a further S-fold purification and loss of about 50% of the heme a + a3. Elution from this resin required the presence of both detergent and salt, indicating both ionic and hydrophobic interactions are involved in the binding of enzyme. Oxidase could not be eluted by 20 mM sodium phosphate, 1% Triton X-100, but was slowly eluted by 50 mM sodium phosphate, 1% Triton X-100 and readily eluted by 100 mM sodium phosphate, 1% Triton X-100. However, it is curious that as much as 200 mM sodium phosphate could be used in the presence of Tween 80 without eluting the enzyme from the column. Chromatography on DEAE-cellulose results in only a small purification. This step appears to be an important one, however, since attempts to eliminate it result in preparations whose purity has never been higher than 8.6 nmol of heme a + a:</mg of protein. Occasionally, some heme a + a3 with very low heme/protein ratios elutes with the low salt wash buffers. This material may be enzyme which is somewhat altered in structure and must be removed in order to obtain the levels of purity normally observed upon subsequent poly(L-lysine) chromatography. Poly(L-lysine)-Sepharose was chosen as a resin for the purifi- cation of cytochrome oxidase because poly(L-lysine) is a known competitive inhibitor of the activity of this enzyme and therefore has binding affinity for the enzyme. The inter-

by guest on June 6, 2020http://w

ww

.jbc.org/D

ownloaded from

Studies on Cytochrome c Oxidase from Rat Liver Mitochondria 1589

action with this column appears to be primarily ionic since enzyme elution was accomplished by a stepwise salt gradient. The concentration of enzyme obtained in this final purification step is approximately 2 to 3 mg/ml.

The complete procedure results in an 82-fold purification of the enzyme using intact mitochondria as the starting material. The recovery of enzyme as judged by its heme content is approximately 16%. The final purity of the enzyme obtained from several preparations is 10.5 to 13.4 nmol of heme a + aa/ mg of protein.

Sonication during the preparation of SMP particles some- times results in the loss of as much as 50% of the oxidase activity. Further activity is lost upon introduction of Triton X-100. While the specific activity of the enzyme based on activity per mg of protein does increase at each purification step, the apparent enzyme activity per nmol of heme (turnover rate) decreases during purification so that the enzyme in the final fraction has an apparent activity which is 23 to 30% that of intact mitochondria.

When this preparation was analyzed by polyacrylamide gel electrophoresis under nondissociating conditions, a single pro- tein band was observed (Fig. 2). As little as 1 pg of protein can be detected by the staining method used. Therefore, since 200 pg of protein were applied to these gels, the preparation is judged to be greater than 99% pure. This, however, does not exclude the possibility that there are contaminants which co- migrate with the oxidase.

Subsequent analysis of this enzyme preparation by poly- acrylamide gel electrophoresis in the presence of SDS con- sistently revealed the presence of six polypeptide components having apparent M, of 66,000, 39,000, 23,500, 14,000, 12,500, and 10,000 (Fig. 3). In some preparations, there was a small amount of protein on the leading side of the 39,000 component which, when present, constituted less than 2% of the total protein. The three low molecular weight components are variable in the extent of their resolution. Enzymatic iodination of the enzyme preparation prior to gel electrophoresis signifi- cantly enhances resolution of these components and results in the appearance of three distinct components in this molecular weight range (see Ref. 15, Figs. 3 and 5~).

In summary, we have developed a procedure which results in an 82-fold purification of cytochrome c oxidase. The enzyme preparation has a heme a + &protein ratio of 10.5 to 13.4 nmol/mg, migrates as a single band during electrophoresis under nondissociating conditions and consists of six poly- peptide components.

0.4 c

MOBILITY (cm)

FIG. 2. Polyacrylamide gel electrophoresis of cytochrome c oxidase under nondissociating conditions. Polyacrylamide gels (5%) contain- ing 0.1% Triton X-100 were used. The enzyme was visualized by staining with Coomassie brilliant blue R. The arrow represents the d.ve front.

Properties of Cytochrome c Onidase

Enzymatic Properties-The enzymatic activity was deter- mined as described under “Experimental Procedures.” The effects of various parameters on this activity are discussed below.

pH-The enzyme was tested for its ability to oxidize cyto- chrome c at various pH values. The pH of the assay was varied from 3.0 to 12.0 by 0.5-pH unit increments. The results (Fig. 4) demonstrate that there is a very sharp optimum around pH 6.0.

Ionic Strength-The activity of the enzyme was tested with regard to the ionic strength of the assay buffer. The concen- tration of potassium phosphate, pH 7.0, was varied from 10 to 200 ITIM. It can be seen in Fig. 5 that the activity shows a broad optimum from 10 to 70 mM. At higher concentrations of phosphate, however, enzymatic activity is severely im- paired. This effect is not limited to potassium phosphate alone; equivalent concentrations of potassium chloride pro- duce the same effect.

Inhibitors-Table II shows the effect of cyanide and carbon monoxide on enzyme activity. The presence of 0.01 M KCN completely inhibits enzyme activity as does bubbling CO through the enzyme solution for 10 min prior to assay.

Detergents-Cytochrome c oxidase is a membrane-bound

Mobility (cm)

FIG. 3. Subunit composition of cytochrome c oxidase. Polyacr.vl- amide gels (12%) containing 0.5%’ SDS were used. Polypeptides were visualized as described in Fig. 2. Molecular weights were determined from the relative mobility of standards as described under “Experi- mental Procedures.”

3 4 5 6 7 8 9 10 I, I* PH

FIG. 4. Effect of pH on cytochrome c oxidase. Activity was meas- ured as described under “Experimental Procedures.”

by guest on June 6, 2020http://w

ww

.jbc.org/D

ownloaded from

Studies on Cytochrome c Oxidase from Rat Liver Mitochondria

I 1 / I I 1 1 1 I 1 I 0 20 40 60 80 100 120 140 160 180 200 210

POTASSIUM PHOSPHATE (mM)

FIG. 5. Effect of potassium phosphate concentration on cyto- chrome c oxidase activity. Activity was measured as described under “Experimental Procedures.”

TABLE II

Inhibition of cytochrome c oxidase by KCN and CO Enzyme activity was measured as described under “Experimental

Procedures.”

Conditions Specific activitv Inhibition

K (min-‘mg-‘) %

Control” 680.9 + 0.01 M KCN 0.0 100 + cob 0.0 100

a The enzyme fraction was from poly(L-1ysine)Sepharose; heme/ protein = 12.4 nmol/mg.

b Carbon monoxide was bubbled through the enzyme fraction for 10 min prior to assay. Twenty microliters of enzyme was used in a LO-ml assay volume.

enzyme. Presumably in situ, this binding involves membrane phospholipids which may be very important for enzyme activ- ity. As shown in Table III, each enzyme purification procedure which results in a decrease in the turnover rate of the enzyme is accompanied by the removal of significant amounts of phospholipid. Further experiments indicate that addition of phospholipid or Tween 80 to the enzyme results in significant increases in enzyme activity (Table IV). Unlike Tween 80 or phospholipids, Triton X-100 inhibits enzyme activity 46%. This is approximately equal to the 44% loss of enzyme activity (units per nmol of heme) which occurs during cytochrome c- Sepharose chromatography (when Triton X-100 is first intro- duced into the procedure) as shown in Table I.

Spectral Properties-The spectral properties of the purified enzyme are given in Figs. 6 and 7. Fig. 6 shows the spectra of the most highly purified cytochrome c oxidase fraction in the oxidized and reduced forms. The oxidized form shows a broad peak in the region of 595 nm and a sharp peak at 424 nm. The reduced form shows sharp peaks at 601 nm and 441 nm. No traces of contamination by any other cytochromes are evident as shown by the lack of absorption between 540 and 570 nm. Determination of the heme a + a3 content of enzyme from the difference in absorbance at 601 nm and 630 nm after dithionite reduction gave results which were in agreement with those obtained by the pyridine hemochromogen method (see “Experimental Procedures”). Fig. 7 shows the difference spectrum obtained between a sample of enzyme reduced with sodium dithionite and a sample through which carbon mon- oxide has been bubbled and then reduced with sodium dithi- onite. The difference in absorbance between 445 nm and 430

TABLE III

Phospholipid content of cytochrome c oxidase at various stages of purification

Enzyme activity and phospholipid content was measured as de- scribed under “Experimental Procedures.” Abbreviations are ex- plained in Table I.

Fraction Enzvme activitv” Phomholioid content

K (min-’ Tl~Ol-‘) mg/nmol heme 9oh

Mitochondria 44 314.3 1.720 18.00 SMP 73 326.4 0.920 21.00 P2 131 233.1 0.232 19.00 Cytochrome c 729 149.6 0.010 3.75 Polylysine 915 79.6 0.003 3.40

“Activity is expressed as the first order rate constant = In (A2/At,) + protein (mg/ml).

Per cent phospholipid = milligrams of phospholipid/mg of phos- pholipid + mg of protein.

TABLE IV

The effect of various detergents on cytochrome c oxidase activity Enzyme activity was measured as described under “Experimental

Procedures.”

Experimental conditions Specific activ- ity” Stimulation Inhibition

K (min-‘mg-‘) % %

Experiment 1 ControP 127.7 +P-lipid” 255.4 100 +0.5% Tween 80 680.9 533 +0.5% Tween 80, 740.5 580

+ P-lipid Experiment 2

Controld 159.6 +I% Triton X-100 86.0 46

a Activity is expressed as the first order rate constant = In (At/AI,) + protein (mg/ml).

The enzyme fraction was from poly(L-lysine)-Sepharose; heme/ protein = 12.4 nmol/mg. No Tween 80 present.

‘An amount of either commercial soybean phosphatides or rat liver mitochondrial phospholipids which gave a phospholipid/protein ratio approximating that of mitochondrial membrane was incubated with enzyme at 0°C for 30 min prior to assay. No Tween 80 present.

’ The enzyme fraction used was from Sephadex G-15.

>\ \\

,r-..__

I \\ - ---- -------- ,' -.._

I

400 440 480 520 560 600 640

h (nm)

FIG. 6. Spectral properties of cytochrome c oxidase. The spectra were taken using water as a reference solution. Reduction of the heme was carried out using a few grains of solid sodium dithionite.

content as described under “Experimental Procedures.” The heme o/as ratio calculated from this data for the purified enzyme is 1.12, in good agreement with active cytochrome c oxidase (25).

Contaminating Enzymes-Coenzyme Q-cytochrome c re- ductase. NADH dehvdroeenase. succinic acid dehvdronenase. nm in this spectrum was used in determining the heme a3 -.-.- .-.- -, ~~ II v ” ”

by guest on June 6, 2020http://w

ww

.jbc.org/D

ownloaded from

Studies on Cytochrome c Oxidase from Rat Liver Mitochondria 1591

and ATPase are mitochondrial inner membrane enzymes and are associated with cytochrome c oxidase as part of the electron transport chain. Because of this association and the fact that each of these enzymes also has some hydrophobic properties, the most highly purified fraction of cytochrome c oxidase was examined for possible contamination by these enzymes. Coenzyme Q-cytochrome c reductase, NADH de- hydrogenase, and succinic acid dehydrogenase were measured directly and in each case the enzyme was present in intact mitochondria but was completely absent from the most highly purified fraction of cytochrome oxidase (Table V). In the case of NADH dehydrogenase, which has been reported by Pen- niall (30,31) to be present in his preparation of cytochrome c oxidase from rat liver mitochondria, we have calculated that if it is present at all there is less than 1% contamination by this enzyme. Ninety-five per cent of the NADH dehydrogen- ase activity present in intact mitochondria fractionates into the 20 mM sodium phosphate, pH 7.0, 1% Tween 80 wash during cytochrome c-Sepharose chromatography. Thus, es- sentially all of this enzyme is separated from cytochrome c oxidase at this step. In the case of both coenzyme Q-cyto- chrome c reductase and succinic acid dehydrogenase, we have calculated that any contamination of the preparation by either of these enzymes would be less than 0.1%. The locations of these two enzyme activities during fractionation were not determined directly, but the majority of cytochromes b and

.03 c

$ ::- 4 (-.p+h- 0 ---- ----- -- --_-------------------

400 440 480 520 560 600 640

1 (nml

FIG. 7. The effect of carbon monoxide on the spectral properties of cytochrome c oxidase. Carbon monoxide was bubbled through the oxidase preparation for 10 min. The solution was then reduced with solid sodium dithionite and analyzed using reduced cytochrome oxi- dase as reference solution.

TABLE V Assay for contaminating enzymes in cytochrome c oxidase

Enzyme activities were determined as described under “Experi- mental Procedures.” Abbreviations used are: Mt, mitochondria; Po- lvlvsine. nolvk-1vsineLSeuharose eluate.

Contaminating enzyme Fraction assayed Specific ac- No. of de- tivity tennina-

lions

CoQHz-cytochrome c reductase

NADH dehydrogen- ase

Succinic acid dehy- drogenase

units”/mg protein

Mt 0.38 Polylysine* 0.00 2

Mt 3.10 3 Polylysine 0.00 Mt 2.40 Polylysine 0.00 2

a 1 unit = 1 ,umol of substrate converted to product per min. * Polylysine = poly(L-lysine)-Sepharose eluate: this fraction had a

heme a + as/protein = 12.4 nmol/mg.

TABLE VI

Inhibition of ATPase activity by antiserum prepared to highly purified cytochrome c oxidase.

Conditions” Activitv” Inhibition

Experiment 1 Sonicated mitochondria -Antiserum +Antiserum, no preincuba-

tion +Antiserum, 30-min prein-

cubation Experiment 2

Solubilized mitochondria -Antiserum +Antiserum, no preincuba-

tion +Antiserum, 30-min prein-

cubation

pm01 %

77.9 76.1

78.8 0

87.5 87.5

87.5 0

n Sufficient antiserum was added to cause 100% precipitation of heme a + a3 in 24 to 72 h. Preincubation, when used was 30 min and no precipitation occurs at this point.

* Activity is expressed as picomoles of ATP converted to ADP in 15 min, see “Experimental Procedures.”

cl, which are integral components of coenzyme Q-cytochrome c reductase were found in the 20 and 200 mM sodium phos- phate, pH 7.0, 1% Tween 80 washes during cytochrome c- Sepharose chromatography. We have not determined the location of succinic acid dehydrogenase. ATPase activity was determined indirectly by measuring the ability of antibodies to the most highly purified fraction of cytochrome c oxidase to inhibit the ATPase activity of sonicated or solubilized mitochondria. There was no inhibition of ATPase activity (Table VI) whereas cytochrome c oxidase activity was in- hibited by more than 85% (data not shown, see Ref. 15). This indicates that there are no antibodies to ATPase in the antiserum prepared to cytochrome c oxidase and suggests, therefore, that the most highly purified fraction of this enzyme is free of contamination by ATPase.

Non-heme Iron-NADH dehydrogenase, coenzyme Q-cy- tochrome c reductase, and succinic acid dehydrogenase each contain measurable quantities of non-heme iron (21, 32, 33). Since these enzymes are not detectable in our enzyme prepa- ration and since cytochrome c oxidase in other systems does not contain non-heme iron (6), one would not expect our most highly purified cytochrome c oxidase fraction to contain non- heme iron. Analysis indicates that the total iron concentration approximates that of the heme iron. Therefore, the enzyme is not heavily contaminated with non-heme iron.

DISCUSSION

Purification of cytochrome c oxidase from rat liver mito- chondria has yielded an enzyme which is suitable for the study of its biosynthesis by isolated mitochondria. The procedure includes two resins, cytochrome c-Sepharose and poly(~-ly- sine)-Sepharose, which we have attempted to use as affinity resins. While the precise interactions between the resins and the enzyme during purification are not known, some obser- vations are worth noting. First, ionic interactions seem to be involved during purification of the enzyme using cytochrome c-Sepharose. The p1 of the cytochrome c is 10.65 while that of the rat liver oxidase was found to be 5.6. Thus, upon chro- matography at pH 7.0, resin-linked substrate and enzyme have opposite net charges and, as expected, increases in salt concentration are needed to release bound enzyme. Also, the ability of this resin to bind cytochrome c oxidase in preference to cytochromes b and cl depends on the pH at which the Sepharose is activated and linked to cytochrome c. When

by guest on June 6, 2020http://w

ww

.jbc.org/D

ownloaded from

1592 Studies on Cytochrome c Oxidase from Rat Liver Mitochondria

prepared at pH 8.5 to 9.5, the cytochrome c-Sepharose binds all three hemoproteins in the presence of 100 InM phosphate, 1% Tween 80. However, when this resin is prepared at pH 6.5 to 7.5, it binds cytochrome c oxidase, but not cytochromes b or c1 in the presence of this same buffer. The preferential binding of enzyme under these conditions seems to involve a hydrophobic interaction since replacement of the Tween 80 by Triton X-100 results in ready elution of the cytochrome c oxidase from this resin. The significant difference in structure between these two detergents is that Triton X-100 contains an aromatic phenol moiety while Tween 80 does not. Thus, it may be that there is a significant aromatic interaction between the respective hemes of cytochrome c oxidase and the cyto- chrome c-Sepharose during binding to the resin and Triton X-100 may effect elution because its aromatic moiety inter- feres with this interaction.

The complete purification method provides an enzyme preparation of high purity based on several criteria. Its heme a + a8 content (IO.5 to 13.4 nmol/mg of protein) is one of the highest reported for cytochrome c oxidase of comparable peptide composition obtained from any biological system (3, 10-12, 14, 34-38). The preparation contains no detectable cytochrome b, c, or c,, as judged by spectral examination. It contains no measurable succinic acid dehydrogenase or coen- zyme Q-cytochrome c reductase activity, appears to be free of ATPase and in contrast to the preparation of Penniall(30,31) is also free of NADH dehydrogenase activity. Finally, the enzyme preparation migrates as a single band when analyzed by polyacrylamide gel electrophoresis under nondissociating conditions.

When the enzyme preparation is analyzed by SDS-poly- acrylamide gel electrophoresis, it is resolved into six polypep- tides having apparent M, of 66,000, 39,000, 23,500, 14,000, 12,500, and 10,000. Table VII shows that preparations of this enzyme from other sources are similar in that they all consist of six to seven polypeptides of comparable molecular weights. Their molecular weight profiles also agree well with that of the rat liver enzyme except that the latter contains a high molecular weight species of 66,000. Because of this molecular weight discrepancy, we have considered the possibility that this peptide is a contaminant which is not physically associ- ated with the oxidase, but, rather, merely co-purifies with it. However, analysis indicates that the known membrane en- zymes are absent and the enzyme preparation does migrate as a single protein band when analyzed by polyacrylamide gel electrophoresis under nondissociating conditions. While this

is not compelling evidence, at the present time, we consider the 66,000-dalton polypeptide to be a bona fide component of the enzyme. More information on this point can be obtained when antibodies to individual polypeptides of the enzyme become available. Polypeptides which simply co-purify with the enzyme but are not in physical association with it will not be precipitated by antibodies to individual polypeptides of the “true” enzyme complex. Thus, it is possible to detect such contaminants by comparing the polypeptide composition of the enzyme preparation with that of immunoprecipitates ob- tained from treating the enzyme preparation with antibodies to individual polypeptides of the enzyme. Such an approach has been very useful in studying the composition of yeast cytochrome c oxidase (39) as well as gaining insight into its topography (40).

While this procedure yields an enzyme preparation of high purity, the turnover rate of the enzyme decreases during purification to 23 to 30% of its value in intact mitochondria. Pertinent to this, Triton X-100 inhibits enzyme activity some 46% (Table IV). The extent of inhibition is very similar to the apparent 44% loss of enzyme activity obtained upon cyto- chrome c-Sepharose chromatography which is the step at which Triton X-100 is introduced into the purification proce- dure. This suggests that the enzyme activity at this step and all subsequent steps is inhibited by the presence of Triton X- 100. Another factor which appears to be important for enzyme activity is the enzyme-associated lipids. Purification proce- dures which result in decreased enzyme turnover rate also result in removal of significant amounts of phospholipid (Ta- ble III). These phospholipids may provide a somewhat hydro- phobic or quasi-membranous environment required for opti- mal activity. This suggestion is supported somewhat by the increased enzyme activity obtained when preliminary at- tempts to restore activity by adding phospholipids (or Tween 80) to the purified enzyme were carried out (Table IV). Similar results have been obtained with the yeast and beef heart enzymes (36, 41). Thus, while it is impossible at this point to quantitate the various parameters responsible for the appar- ent decrease in turnover rate of the enzyme during purification (Table I), it seems likely that inactivation due to the presence of Triton X-100 is very important and the removal of phos- pholipids may also contribute significantly.

An important question, as yet unanswered, is whether all of the polypeptides observed in the various oxidase preparations (Table VII) have a role in enzyme activity. This is a difficult question to answer at the moment, in part, because of our

TABLE VII Comparison of cytochrome c oxidase preparations from various sources

Saccharomyces cereuisiae Neurospora Rat liver Beef heart crassa Locusta XC?lU3/,US

-a migratorra -h lawis -/ h -i’ -F h d -f -p

daltons (X IO-“,

4.4 7.3 8.2 8.0

10.0 9.5 9.5 10.0 9.8 10.0 9.5 12.5 10.2 12.5 10.6 11.5 11.2 11.0 11.7 12.5 12.0 14.0 13.0 14.0 12.3 14.0 13.0 13.4 12.4 14.5

13.8 14.6 16.0 15.0 14.5 15.2 16.8 17.0 21.0 19.0

23.0 25.0 23.0 24.7 21.0 22.0 20.4 20.7 24.1 24.0 23.0 27.3 34.5 34.1 28.5 33.0

39.0 40.0 42.0 42.4 41.0 40.0 47.5 39.6 35.4 38.0 44.0 66.0

n This report. ‘Ref. 35. h Ref. 3. p Ref. 37. ’ Ref. 10. h Ref. 12. d Ref. 36. ’ Ref. 38.

by guest on June 6, 2020http://w

ww

.jbc.org/D

ownloaded from

Studies on Cytochrome c Oxidase from Rat Liver Mitochondria 1593

relative inexperience with multicomponent membrane-asso- ciated enzymes. In a single polypeptide enzyme, only a portion of its structure actually constitutes the catalytic site. However, the remainder of the molecule is clearly needed for the for- mation of that site and also can affect its activity. It is more difficult to delineate what constitutes the “complete enzyme” in a multicomponent enzyme because of the lack of defined limits such as is given by the covalent bonding of a single polypeptide enzyme. In the case of cytochrome oxidase, most experimental evidence suggests that the heme prosthetic groups are bound to the smaller (<20,000 daltons) polypep- tides (5, 36, 42, 43), although Tzagoloff et al. (44) reach a different conclusion. Poyton and Schatz (39) and Eytan and Schatz (40) have demonstrated that antibodies to two of the larger polypeptides of the yeast enzyme (>20,000 daltons) inhibit catalytic activity. These authors interpret this to mean that these polypeptides are involved with the catalytic site. Phan and Mahler (36, 43) have been able to remove at least two of the large polypeptide components of the enzyme and retain 73% of its activity. These authors emphasize, therefore, that these large polypeptides are not part of the catalytic site and suggest that they may be involved either in assembly of the catalytic polypeptides or as “regulatory-integrative” com- ponents of the enzyme. At present, then, the data do not rule out participation of the large subunits in catalytic activity, nor does it rule it in. Rather, the role of the various polypeptides of this enzyme with respect to its activity remains to be defined and probably will not be resolved until reconstitution of the enzyme can be accomplished from individual polypep- tides.

Acknowledgments-We wish to thank Mr. Alex Bonica for assist- ance in the preparation and characterization of the cytochrome c- Sepharose, Mr. David Thurlow for part time technical assistance during the investigation, Dr. Thomas Mason for his suggestion of the cytochrome c-Sepharose affinity column as well as very helpful dis- cussion during the course of this investigation, and Drs. Lucille Smith and Diana Beattie for critical evaluation of the manuscript.

REFERENCES

1. Mason, T. L., and Schatz, G. (1974) Anna Rev. Biochem. 43,51- 87

2. Timberlake, W. E., Hagen, G., and Griffin, D. H. (1972) Biochem. Biophys. Res. Commun. 48, 823-832

3. Rubin, M. S., and Tzagoloff, A. (1973) J. Biol. Chem. 248, 4269- 4274

4. Capaldi, R. A., and Hayashi, H. (1972) FEBS Lett. 26,261-263 5. Komai, H., and Capaldi, R. A. (1973) FEBS Lett. 30, 273-276 6. Wharton, D. C., and Tzagoloff, A. (1967) Methods Enzymol. 10,

245-250 7. Yonetani, T. (1960) J. Biol. Chem. 235, 845-852 8. Van Buuren, K. J. H., and Van Gelder, B. F. (1974) Biochim.

Biophys. Acta 333,209-217

9.

10.

11.

12.

13.

14.

15.

16. 17. 18. 19.

20. 21. 22. 23. 24.

25. 26.

27.

28. 29. 30. 31. 32.

33.

34.

35.

36.

37.

38. 39. 40. 41.

42.

43.

44.

Hartzell, C. R., and Beinert, H. (1974) Biochim. Biophys. Actn 368, 318-338

Mason, T. L., Poyton, R. O., Wharton, D. C., and Schatz, G. (1973) J. Biol. Chem. 248, 1346-1354

Weiss, H., Sebald, W., and Bucher, Th. (1971) Eur. ,I. Biochem. 22, 19-26

Weiss, H., Lorenz, B., and Kleinow, W. (1972) FEBS Lett. 25,49- 51

Jacobs, E. E., Andrews, E. C., Cunningham, W., and Crane, F. L. (1966) Biochem. Biophys. Res. Commun. 25,87-95

Ozawa, T., Okumura, M., and Yagi, K. (1975) Biochem. Biophys. Res. Commun. 65, 1102-1107

Rascati, R. J., and Parsons, P. (1979) J. Biol. Chem. 254, 1594- 1599

Rascati, R. J., and Parsons, P. (1974) J. Cell Biol. 59, 557a Cuatrecasas, P. (1970) J. Biol. Chem. 245, 3059-3065 Parsons, P., and Simpson, M. V. (1967) Science 155,91-93 Smith, L., and Conrad, H. (1956) Arch. Biochem. Biophys. 63,

403-413 De Bernard, B. (1957) Biochim. Biophys. Acta 23,510-515 Rieske, J. S. (1967) Methods Enzymol. 10, 239-245 King, T. E. (1963) J. Biol. Chem. 238,4032-4036 Horie, S., and Morrison, M. (1963) J. Biol. Chem. 238, 1855-1860 Yubitsui, T. (1968) Anna Rev. Biol. Works Fuc. Sci. Osaka

Univ. 16,35-60 Vanneste, W. H. (1966) Biochemistry 5, 838-848 Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J.

(1951) J. Biol. Chem. 193, 265-275 Chen, P. S., Toribara, T. Y., and Warner, H. (1956) Anal. Chem.

28, 1756-1758 Kopaczyk, K. C. (1967) Methods Enzymol. 10, 253-258 Frater, R. (1970) Anal. Biochem. 38, 536-538 Davidian, N., and Penniall, R. (1973) J. Cell Biol. 59, 70a Penniall, R. (1974) J. Cell Biol. 63, 263a Hatefi, Y., and Rieske, J. S. (1967) Methods Enzymol. 10, 235-

239 Ziegler, D., and Rieske, J. S. (1967) Methods Enzymol. 10, 231-

235 Sebald, W., Machleidt, W., and Otto, J. (1973) Eur. cJ. BLochem.

38, 311-324 Kuboyama, M., Yong, F. C., and King, T. E. (1972) J. Biol. Chhem.

247,6375-6383 Phan, S. H., and Mahler, H. R. (1976) J. Biol. Chem. 251, 257-

269 Downer, N. W., Robinson, N. C., and Capaldi, R. A. (1976)

Biochemistry 15, 2930-2936 Koch,G. (1976)J. Biol. Chem. 251,6097-6107 Poyton, R. O., and Schatz, G. (1975) J. Biol. Chem. 250, 762-766 Eytan, G. D., and Schatz, G. (1975) J. Biol. Chem. 250, 767-774 Yu, C., Yu, L., and King, T. E. (1975) J. Biol. Chem. 250, 1383-

1392 Schatz, G., Groot, G. S. P., Mason, T., Rouslin, W., Wharton, 1).

C., and Saltzgaber, J. (1972) Fed. Proc. 31, 21-29 Phan, S. H., and Mahler, H. R. (1976) J. Biol. Chem. 251, 270-

276 Tzagoloff, A., Akai, A., and Rubin, M. S. (1974) in Biogenesis of

Mitochondria (Kroon, A. M., and Saccone, C., eds) pp. 405- 421, Academic Press, New York

by guest on June 6, 2020http://w

ww

.jbc.org/D

ownloaded from

R J Rascati and P Parsonsmitochondria.

Purification and characterization of cytochrome c oxidase from rat liver

1979, 254:1586-1593.J. Biol. Chem. 

  http://www.jbc.org/content/254/5/1586Access 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/254/5/1586.full.html#ref-list-1

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

by guest on June 6, 2020http://w

ww

.jbc.org/D

ownloaded from