ICANCER RESEARCH 46, 1966-1971, April 1986]

ABSTRACT

The Ca2@-and phospholipid-dependentprotein kinase (protein kinaseC) has been purified to electrophoretic homogeneityfrom mouse braincytosol. 120-3Hlphorbol 12,13-dibutyrate binding activity was found tocoelute quantitatively with the protein kinase activity throughout thepurification procedure. The crude extract was first run over a DES2column.Fractions containingpeak activities were then chromatographedusing a fast protein liquid chromatography system with a Mono Q columnfollowed by chromatography on the same column run in the presence ofI mM adenosine triphosphate. The adenosine triphosphate specificallyshifted the elution position of the Ca2@-and phospholipid-dependentprotein kinase, providinga powerful step in the purification procedure.The remaining minor contaminants were removedby hydrophobicchromatographyon a TSK-phenyl-5-PWcolumn.This purificationprocedurerequired less than 2 days after the initial large hatch DES2 columnchromatography. The molecular weight of the purified receptor wasestimated to be 81,000 by its mobility on sodium dodecyl sulfate/polyacrylamide gels, in agreement with the published values. Optimal conditions for the storage of the purified receptor were sought. Both proteinkinase and phorbol ester binding activities were stable for 2 mo whenstored in the presence of 0.01% Triton X-l00 at —70°C.Polyclonalantibodies to the purified receptor have been prepared from rabbits.These antibodies recognized the purified receptor in electroblotting assays and were able to immunoprecipitatethe purifiedreceptor.

INTRODUCTION

Phorbol ester tumor promoters have been shown to play animportant role in skin carcinogenesis (1—3)as well as to affectmany other biological responses in in vitro systems (4—6).Thehigh potency of the phorbol esters and their structure-activityrelationships for the induction of biological responses (7)strongly suggested that the phorbol esters functioned throughaction at a specific receptor. Such receptors were demonstratedfirst in chick embryo fibroblasts(8) and subsequently in a varietyof tissue preparations and intact cells (see Ref. 9 for a review)using [3HIPDBu.2 The receptor was found to share many properties in common with a protein kinase C first identified byNishizuka (10). These properties included tissue distribution,high activities in brain, evolutionary conservation, high Ca2@sensitivity, phospholipid association, and interaction with thephorbol esters (1 1, 12). Evidence for their identity was that thebinding and the protein kinase C activities coeluted both uponpartial purification (13—16)and with the homogeneously purifled protein (17, 18).

Considerable recent effort has been directed at characterizingthe substrate specificity of protein kinase C, its interaction withlipids and its subcellular localization (10, 19). A major im

Received 7/30/85; revised 12/9/85; accepted I 2/18/85.The costs of publication of this article were defrayed in part by the payment

of page charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

I Present address: Neuroscience/Cardiovascular Section, CIBA-GEIGY Corp.,

556 Morris Avenue, Summit, NJ 07901. To whom requests for reprints shouldbeaddressed.

2 The abbreviations used are: [3HJPDBu, [20-3Hlphorbol 12,13-dibutyrate;

BSA, bovine serum albumin; FPLC, fast protein liquid chromatography; proteinkinase C, Ca2@-and phospholipid-dependentprotein kinase;SDS, sodium dodecyl sulfate; EGTA, ethyleneglycol bis(ft-aminoethyl ether)-N,N,N',N'-tetraaceticacid; DTT, dithiothreitol.

pediment to these studies has been difficulties in obtainingadequate quantities ofthe purified enzyme. Following the initialreport on purification to homogeneity of protein kinase C (20),a variety of other protocols have been reported (13, 18, 2 1—26).General problems of the purification included: (a) the proteinwas not purified to homogeneity as judged by SDS/polyacrylamide gel electrophoresis (23, 24, 26); (b) only a small quantityof the purified protein was obtained (18, 20, 25); and (c) thepurified protein was not stable (21, 22).

Our laboratory has been involved in the study of the molecular mechanisms of tumor promotion. To understand the process of tumor promotion in the mouse, in which two-stage skincarcinogenesis has been extensively characterized, it is important to clarify the specifics of phorbol ester interaction in thisspecies. The one report on purification using mouse tissuefound a molecular weight of 70,000 for purified protein kinaseC from membranes (13), which was appreciably lower than theMr 82,000 for the rat brain enzyme obtained by Kikkawa et aL(20). It was not clear whether this lower molecular weightreflected a species difference, proteolytic fragmentation, or adifference between the membrane-bound and cytosolic forms ofthe protein. In this paper, we report a purification scheme ofthe receptor for the phorbol esters from mouse brain cytosol toelectrophoretic homogeneity. The purification procedure takesless than 2 days after the large batch DE52 step, and the purifiedreceptor is stable under our conditions of storage.

MATERIALS AND METhODS

Materials. Preswollen diethylaminoethyl cellulose (DE52) and cellulose phosphate P81 papers were obtained from Whatman (Clifton,NJ). Mono Q columns (HR 5/5 and 10/10), FPLC system, and electrophoresis molecular weight calibration kit were purchased from Pharmacia (Piscataway, NJ). 120-3Hlphorbol 12,13-dibutyrate ([3H]PDBu,10.8 to 13.4 Ci/mmol) and [‘y-32PJATPwere from New England Nuclear (Boston, MA). Bovine ‘y-globulins and fatty acid-free BSA wereobtained from Sigma (St. Louis, MO). Bio-gel TSK-phenyl-5-PWcolumn (75 x 7.5 mm), SDS, protein dye reagent, DEAE-Affi-Gel Blue,and electrophoresis grade acrylamide were purchased from Bio-Rad(Richmond, CA). ATP and Pansorbin were products of Calbiochem(LaJolla, CA).

Buffers Used in the Purification of Protein Kinase C. The buffersolutions used for the purification were: buffer A, 2 mM EGTA, 5 mMEDTA, 0.3 M sucrose, 5 mM DTT, and 60 mM Tris HC1 for thepreparation of crude extract; buffer B, 0.5 mM EGTA, 0.5 mM EDTA,5 mM DTT, and 20 mM Tris-HCI for DE52 column chromatography;buffer C, 0.5 mM EGTA, 0.5 mM EDTA, I m@sDTT, and 20 m@iTrisHCI for Mono Q column fractionation; buffer D, 0.5 mM EGTA, 0.5mM EDTA, 1 mM DTT, 1.5 M NaCl, and 20 mr@iTris-HC1 for TSKphenyl-5-PW column elution; buffer E, 0.125 mM EGTA, 0.125 mMEDTA, 0.25 mM DTT, and 5 mM Tris-HC1, also for TSK-phenyl-5-PW column elution. Buffers were adjusted to pH 7.4 at room temperature (for those used in Mono Q and TSK-phenyl-5-PW columnchromatography) or 4'C (for DE52 column chromatography).

Preparation of Cytosol from Mouse Brain. In general, 150 to 300female CD-l mice from Charles River (Portage, MI) aged between 4and 6 wk were decapitated and brains were quickly removed and chilledin buffer A. Brains were weighed and homogenized in buffer A (2 ml/g wet wt) using a Potter-Elvehjem homogenizer. The homogenate was

1966

Purification of Stable Protein Kinase C from Mouse Brain Cytosol by Specific

Ligand Elution Using Fast Protein Liquid Chromatography

Arco Y. Jeng,' Nancy A. Sharkey, and Peter M. Blumberg

Molecular Mechanisms of Tumor Promotion Section, Laboratory ofCellular Carcinogenesis and Tumor Promotion, NIH, Bethesda, Maryland 20892

Association for Cancer Research. by guest on August 29, 2020. Copyright 1986 Americanhttps://bloodcancerdiscov.aacrjournals.orgDownloaded from

PURIFICATION OF PROTEIN KINASE C

centrifuged at 40,000 rpm for 60 mm in Beckman 50.2 Ti rotors. Thesupernatant was pooled and diluted with 4 ml of chilled 5 mM DTT/gof brain. This diluted cytosol was then loaded on the DE52 column inthe next step of purification.

DE52 Column Fractionation. DE52 resin (50 g/150 mice) was prepared according to the manufacturer's instruction and was equilibratedin buffer B overnight. The diluted cytosol was loaded and the columnwas washed with 7- to 10 bed vol of buffer B. Proteins were then elutedwith a linear gradient containing 0 to 300 mM NaC1 in buffer B (totalgradient volume, 500 ml for cytosol from 150 mice). One hundredfractions were collected and I3HIPDBu binding or protein kinase Cactivity of the fractions was determined. The fractions containing highprotein kinase or (3HJPDBu binding activity were pooled and thenconcentrated and desalted with Amicon stirred cells (Danvers, MA)using PM3O membranes. The final protein concentration was about 25mg/mi in buffer B.

Mono Q Column Chromatography. The concentrated DE52 peakmaterial was further fractionated by a Mono Q (HR 10/10) columnusing a Pharmacia high resolution FPLC system. The Mono Q columnwas equilibrated with buffer C and about 80 mg of protein was loadedonto the column in each of the several repetitive column runs. Thecolumn was washed with 5 bed vol of buffer C and proteins were elutedwith a linear gradient containing 0 to 300 mM NaC1 in buffer C (totalgradient volume, 160 ml). The flow rate was 4 mI/mm; 2-mi fractionswere collected and the protein kinase activity was assayed. The peakkinase activity was pooled and again concentrated and desalted.

The desalted Mono Q pool was then run over the same Mono Qcolumn under identical conditions except that 1 mM ATP was includedin all of the buffers. In order to minimize the interference of ATP inthe protein kinase assay, protein samples from each fraction werediluted at least 1000-fold in the final reaction mixture. The peak proteinkinase activity was pooled.

TSK-phenyl-5-PW Column Elution. The salt concentration of theprotein pool from the previous Mono Q/ATP column run was adjustedto 1.5 M NaCI and then loaded onto a hydrophobic column, TSKphenyi-5-PW, preequilibrated in buffer D using the Pharmacia FPLCsystem. The column was washed with 5 vol of buffer D and proteinswere eluted with a linear gradient containing 10 ml each of buffer D(initial buffer) and buffer E (final buffer). The column was finallywashed with 10 ml of buffer E. The flow rate was 1 mI/mm and 0.5-mifractions were collected. Fractions containing high protein kinase activitywere pooled.

Preparation of PolyclonalAntibodiesagainst Purified Protein KinaseC. The purified protein (100 @zg)was emulsified with an equal volumeofcomplete Freunds adjuvant and injected i.m. into New Zealand whiterabbits (Denver, PA) at 3 to 4 spots. The same procedure was repeatedat 7-day intervals except that the incomplete Freunds adjuvant wasused. Rabbits were bled weekly after the third administration of antigen.The titer of antibodies was checked using the Bio-Rad lmmun-Blot(goat antirabbit-horseradish peroxidase) assay kit. -y-lmmunoglobulinsfrom preimmune serum and antiserum were purified using DEAE-AffiGel Blue.

Immunoprecipitationof Purified Protein Kinase C by PolycionalAntibodies. Pansorbin was prepared according to the manufacturer's instruction and was finally resuspended at a concentration of 20% inBSA, 2 mg/mi. The purified protein kinase C was incubated withpreimmune serum or antiserum in the presence of BSA, I .4 mg/mi,and 40 mM Tris-HC1, pH 7.4, for 2 h at 4T. Pansorbin (200 @l)wasthen added to each tube and further incubated for I h at 4C withconstant mixing. The samples were then centrifuged for 15 mm in aBeckman Microfuge centrifuge at 11,600 x g. Aliquots of the supernatant were assayed for I3HIPDBu binding (30 @zl)and protein kinase(5 @l)activities.To testwhetherthepolycionalantibodiescouldinteractdirectly with the site of I3HIPDBu binding or the catalytic site for theprotein kinase activity, BSA, 2 mg/mI, was used instead of Pansorbin.

Biochemical Assays. Protein was determined by the method of Bradford (27) using bovine -y-globulins as standard. (3H]PDBu binding wasmeasured according to the published protocol (19). Protein kinase Cassays were carried out as described (28) using phosphocellulose paper(29). The concentrations of Ca2@and phosphatidylserine used in the

incubation mixtures of both [3HJPDBu binding and protein kinase Cassays were 0.1 mM and 100 ,@g/ml, respectively. SDS/polyacrylamidegels (8.5%) were run according to the method of Laemmli (30), and theprotein bands were visualized by silver staining (31). Electroblottingwas done by the method ofTowbin et a!. (32). NH2-terminaland aminoacid composition analyses were carried out by the Protein SequencingFacility, University of Michigan. The content of cysteic acid wasdetermined using duplicate samples oxidized with performic acid beforehydrolysis.

RESULTS

Purification of Protein Kinase C. In previously publishedprotocols, chromatography on DEAE-cellulose has proven tobe a useful, high capacity step for the purification of proteinkinase C. Accordingly, mouse brain cytosol was loaded onto aDE52 column, the column was washed with buffer B, andproteins were eluted by a 0 to 300 mrviNaCl gradient in bufferB. Fig. 1 shows a typical elution profile. The major Ca2'@-andphospholipid-dependent protein kinase activity was eluted between 80 and 120 mr@iNaCl. Only minimal protein kinaseactivity was detected in the absence of Ca2@and phosphatidylserine (Fig 1). I3HIPDBu binding activity was found to coelutewith the Ca2―-and phospholipid-dependent protein kinase activity, and additional peaks were not seen within the limits ofthe elution gradient (data not shown). Protein concentration ineach fraction was measured (Fig. 1), and fractions containingthe peak ratio of the protein kinase activity per mg of proteinwere pooled.

The pool from the DE52 column was loaded onto a Mono Qcolumn. Fig. 2A shows the elution profile from the Mono Qcolumn using a 0 to 300 m@iNaCl gradient in buffer C. Peakprotein kinase C activity was eluted consistently at 180 mMNaCI (Fig. LI). [3HIPDBu binding activity was also found tocoelute with protein kinase C activity (results not shown). Sinceprotein kinase C interacts with ATP, we reasoned that inclusionof ATP in the elution buffer might selectively shift the elutionposition of the kinase, leaving most of the contaminating proteins unaffected. Given the high resolution of the Mono Qcolumn, a relatively small shift in elution position would sufficefor a substantial separation. Fractions for the Mono Q columncontaining high protein kinase activities were therefore pooled,desalted, and rerun over the same Mono Q column in thepresence of 1 m@iATP and 3.75 mM magnesium acetate. Underthis new condition, the peak protein kinase C activity wasshifted to 150 mM NaCl (Fig. 2B). A shift was not observed inthe presence of 1 mM GTP or in 1 mrvi ATP but without

6

F0

FRACTIONNUMBER

Fig. 1. Elution profile of mouse brain cytosoi run over a DE52 column.Protein kinaseC activitieswere measuredin the presence(•)and absence(0) ofCa2@and phosphatidylserine.

1967

Association for Cancer Research. by guest on August 29, 2020. Copyright 1986 Americanhttps://bloodcancerdiscov.aacrjournals.orgDownloaded from

Table I Amino acid composition ofihe purifiedreceptorTheamino acid composition shown was the average fromtwodifferentpreparations.

The content ofcysteic acid was determined using duplicatesamplesoxidizedwith performic acid before hydrolysis.Residues,

%Amino

acid Mouse brain PigspleenaAlanine

5.1 ±0•3b7.8Cysteicacid 2.0 ±0.40.3Aspartic

acid 9.4 ±0.18.6Glutamicacid 10.1 ±0.310.3Phenylalanine

5.1 ±0.16.0Glycine9.0±0.510.2Histidine2.4 ±0.22.0Isoleucine4.8 ±0.14.4Lysine6.6 ±0.14.6Leucine7.4 ±0.310.5Methionine1.9±0.21.5Proline6.0 ±0.26.1Arginine6.5±0.65.7Serine7.4 ±0.86.0Threonine5.2 ±0.45.4Valine6.2 ±0.28.1Tyrosine5.7 ±2.4 3.0

0.15

0.10

0.06

2IL'0

0

PURIFICATION OF PROTEIN KINASE C

75

0.25

0@

z0

2U,

I _____JI

Ii::.

—

@— .

ibcdI fFig. 3. SDS/polyacrylamide gel electrophoresis ofsamples from different steps

of the purification. Lane a. mouse brain cytosol; lane b, DE52 peak fractions;lanes c and d, peak fractions of Mono Q column run in the absence and presenceof I m@iATP, respectively; lane e, TSK-phenyl-5-PW column purified receptor,lanef, molecular weight standards. K, thousands.

— 94K

.4,, 67K0

@ACTlONNUMBER

100

0.3

0.2

0.1

0

4

25

0

FRACTIONNUMBER

Fig. 2. FPLC Mono Q column chromatography. Fractions from the DE52column containing high protein kinase C specific activity were run over a MonoQ column equilibrated in buffer C and eluted with a 0 to 300 msi NaCI lineargradient in buffer C (A). Peak protein kinase C activity from (A) was rerun overthe same Mono Q column under the same conditions except that I m@iATP and3.75 msi magnesium acetate were included in all of the buffers (B). Protein kinase(•activities were measured in the presence of Ca2@ and phosphatidylserine.Negligible protein kinase activities were detected in the absence of Ca2@andphosphatidylserine.

magnesium acetate (results not shown). The UV absorbingmaterial shown in Fig. 2B was predominantly due to the ATP.This shift ofthe elution position ofthe protein kinase C activityfrom the Mono Q column provided a major step in the purification of the protein as evidenced by the SDS/polyacrylamidegel electrophoresis (Fig. 3). A dominant band was observedwith three minor contaminants (Fig. 3, !ane ti). These minorcontaminants could be removed by a subsequent hydrophobicTSK-phenyl-5-PW column run (Fig. 3, !ane e). The proteinkinase C peak was eluted at 0% NaCI in buffer E (results notshown). Compared with the molecular weight standards, themolecular weight of the purified protein was calculated to be8 1,000 using log (molecular weight) versus mobility plots (resuits not shown). The composition of amino acids of the purifled protein is shown in Table I and is compared with thatreported for a 68,000-molecular weight form of protein kinaseC from spleen (33). The purified protein was found to have ablocked NH2-terminus, consistent with that observed by Nishizuka.3 The nature of the NH2-terminal blockage has not beendetermined.

A summary of the purification procedure is shown in Table2. [3H]PDBu binding to the cytosol was 28 ±5 (SD) pmol/mgof protein (n = 8). The purified receptor from the TSK-phenyl5-PW column gave 5500 ±900 pmol/mg of protein (n = 4),which represented a 200-fold purification. Since there wereinhibitors of protein kinase C activity in the cytosol (results notshown), it was not possible to express the fold purification ofthe protein kinase C activity with respect to that of the cytosol.Based on the pool from the DE52 column, the fold purificationof protein kinase C activities paralleled that of the [3H]PDBubinding activities at each step of the purification (Table 2). Thespecific binding activity of the purified receptor was approxi

aTakenfromRef.33.b Mean ± range.

mately one-half of that, 12,300 pmol/mg, predicted from themolecular weight. The specific enzymatic activity, however, wasslightly greater than that (1.1 ,@mol/mg/min) reported by Kikkawa et a!. (20).

A major loss of the receptor appears to be due to the firstMono Q column run (Table 2). The Mono Q column used inthis report was a preparative column. Previous runs with analytical Mono Q columns gave a consistent 50% total recovery3 V. Nishizuka, personal communication.

1968

43K

— 30K

Association for Cancer Research. by guest on August 29, 2020. Copyright 1986 Americanhttps://bloodcancerdiscov.aacrjournals.orgDownloaded from

Table 2 Summary ofpurijlcationprocedurePurificationKinase

activityBindingactivityFoldFoldstepProtein.

ca@tmol/mg/min purification% Yieldpmol/mgpurification%YieldDE521000•047@0•016b(12f

I100160±50(11)1(6)―100MonoQ9.8±4.4(4)0.063±0.024(4)113±3(4)150±25(3)1(5)10±4(3)Mono

Q/ATP0.48 ±0.38 (4)0.75 ±0.47 (4) 165.8 ±2.9 (4)1900 ±760 (3)12 (68)4.7 ±2.1(3)TSK-phenyl-5-PW0.06±0.02 (4)1.5 ±0.6 (4) 322.0 ±1.4 (4)5500 ±900 (4)34 (200)1.3 ±0.6 (3)

Table 3 Stability of@HJPDBu binding activity to the purified receptor inthepresenceofcarrier proteins, protease inhibitors, ordetergentThe

purified receptor was stored in 50% glycerol at —20T for 4 days intheabsenceor presence of the protective agents indicated. Specific I3HIPDBubindingwas

the difference between the total binding and the nonspecific binding determined in the presence of 30 pM nonradioactive phorbol I 2,13-dibutyrate.Bindingactivity

measured prior to the storage was set at 100%.I3HIPDBu

binding,Addition% ofcontrolNone

69Bovine@-globulins(0.4 mg/mI)84Soybeantrypsin inhibitor (0.2 mg/mI)720.2%

Triton X-lOO670.O5CeTritonX-I00100Leupeptin

(0.05 mg/mI) 54

PURIFICATION OF PROTEIN KINASE C

aProteinloadedvariedfromI30to450mg;theyieldofthepurifiedproteinwasI10toll0 @zg.b Mean ± S.D.

C Numbers in parentheses. number of experiments.

d Numbers in parentheses, fold purification relative to the cytosolic binding activity, 28 ± 5 (8) pmol/mg of protein.

A

\80@

60@

40

80

I

and a 30% recovery within the peak fractions (data not shown).The low yield of the first Mono Q column shown in Table 2represented a problem in the scaling-up process. Due to thereproducible nature of the Mono Q column, we have been ableto circumvent this problem by using repetitive Mono Q columnruns using the smaller HR 5/5 Mono Q column on an automated FPLC system and assaying only occasional runs toconfirm the peak position of the protein kinase activity. Underthis condition, the recovery of the protein kinase activity fromthe Mono Q column was 28 ±0.5% (n = 81). The purity of theprotein at the subsequent steps appeared to be the same as thatshown in Fig. 3 (results not shown). Preliminary results gavean overall yield of 7%.

Stability of the Purified Receptor. The partially purified receptor was found to be very stable when stored in 50% glycerolat —20°C.No significant decrease of protein kinase C and[@HJPDBu binding activities was observed for several months(results not shown). In contrast, the purified receptor was notstable. Stored in 50% glycerol at —20'C for 10 days resulted inan 80% loss ofthe phorbol ester binding activity. Various agentswere added to prevent the loss of [3H]PDBu binding activitydue to either surface adsorption or possible minor proteasecontaminants. Table 3 shows that the inclusion of proteaseinhibitors (leupeptin and soybean trypsin inhibitor) did notstabilize [3H]PDBu binding activity satisfactorily, suggestingthat the major loss of binding activity was not due to proteasecontamination. On the other hand, in the presence of bovine @yglobulins (0.4 mg/ml) or low concentrations of Triton X-100for 4 days in 50% glycerol at —20°C,binding activity was 84 or100%, respectively, of the original value (Table 3), indicatingthat the loss of activity might be due to surface absorption.

Further investigation revealed that a low concentration ofTriton X-lOO alone was sufficient to stabilize both kinase andbinding activities for prolonged periods if samples were firstfrozen with liquid nitrogen then kept at —70°C.The optimalconcentrations of Triton X-100 was between 0.01 and 0.05%,and 0.01 % Triton X-100 was routinely used. Storage at —70°Cfor 1 mo caused no apparent loss of the binding activity whilethere was only a 28% decrease in the kinase activity (results not

0.1 2 3 456

Rabbit Serum (@AI)

Fig. 4. Removal of (3HIPDBu binding and protein kinase activities by immunoprecipitation. The purified receptor was incubated with preimmune serum (0,t@)or antiserum (O, A, 0, I) and then precipitated by Pansorbin (0, •, @,A) asdescribed in @‘Materialsand Methods.―Aliquots of the supernatant were assayedfor protein kinase (0, I) and [3H]PDBu binding (@, A) activities. Parallelmeasurements of the protein kinase (@) and (3HJPDBu binding (0) activitiescarried out without precipitationby Pansorbinare also shown.Points,averageofduplicate experiments. The specific I3HIPDBu bound was the difference betweenthe total L3HIPDBu bound and the nonspecific I'HIPDBu bound measured in thepresence of 30 MMnonradioactive phorbol 12,13-dibutyrate.

shown). The effects of rapid freeze and thaw were also studied.No loss of [3HJPDBu binding activity was observed after twocycles of freeze and thaw during a period of storage for 2.5 moand about one-half of the kinase activity was still retained after4 mo (results not shown).

Immunoprecipitation of the Purified Receptor by PolyclonalAntibodies. Polyclonal antibodies were prepared as described in“Materialsand Methods.―These antibodies were able to recognize the purified receptor. Fig. 4 shows that the antibodyreceptor complex could be precipitated by Pansorbin and resulted in a decrease of the available receptor for [3HIPDBubinding and protein kinase activities remaining in the supernatant. This decrease was dose dependent and the preimmuneserum at the highest dose tested did not complex with thereceptor (Fig. 4). There was little decrease in either activitywhen the precipitation by Pansorbin was omitted, indicatingthat the majority of the antibodies was not directed against thesite of I3HIPDBu binding or the catalytic site for the proteinkinase activity (Fig. 4).

@y-Immunoglobulins obtained from antiserum recognized purifled protein kinase C (Fig. 5, !ane d). Among the numerousprotein bands shown in the pooled material from DE52 columnchromatography (Fig. 5, lane b), these ‘y-immunoglobulinsreacted strongly with a double band of molecular weight of about80,000 and a band of 60,000 (Fig. 5, lane c). In addition, one

1969

Association for Cancer Research. by guest on August 29, 2020. Copyright 1986 Americanhttps://bloodcancerdiscov.aacrjournals.orgDownloaded from

PURIFICATION OF PROTEIN KINASE C

(see Ref. 34 for an early review). Le Peuch et al. (22) used asimilar approach with Ultrogel AcA34 in EGTA in the purification of the protein kinase C. The peak activity from the firstcolumn was pooled and rerun on the same column in thepresence of phosphatidylserine and Ca2@ which shifted theprotein kinase C peak to the void volume. However, the proteinkinase activity obtained through this protocol was highly unstable. In addition, it was difficult to remove phosphatidylserinefor biochemical studies of protein kinase C. In the purificationscheme reported here, the binding of ATP to protein kinase Cwould have been expected both to give an additional negativecharge to the complex and also to induce a conformationalchange in the protein. Since the inclusion of ATP resulted inthe elution of protein kinase C at a lower salt concentration,this latter effect appeared to predominate. Our method wascarried out under nonequilibrium conditions. ATP was accumulated in the Mono Q column and eluted by salt in the courseof the gradient (Fig. 2B). Nevertheless, the shift in the elutionposition in the presence of ATP was specific, because GTP,which should not interact specifically with protein kinase C,did not cause a shift in the elution profile (data not shown).

A constant ratio of protein kinase activity to phorbol esterbinding activity was found at different stages of purificationand matched that of the purified receptor. Technically, thisresult means that either assay can be used to predict activity inthe other assay and that the choice of assay is a matter ofconvenience. In addition, this result and the observation thatboth activities could be immunoprecipitated to the same extentin the presence of polyclonal antibodies (Fig. 4) agree with thefindings of others that both kinase and binding activities residein the same protein entity. For mouse brain cytosol, this enzymeappears to be of Mr 81,000, larger than that reported for theenzyme from mouse brain membranes (13) but similar in sizeto the enzyme from rat brain cytosol. The basis for this difference is not presently known.

Although the polyclonal antibodies we obtained recognizedthe purified receptor with a good titer (1:1500 dilution) in theelectroblotting assay and immunoprecipitated the purified receptor in a dose-response manner (Fig. 4), they did not reactuniquely with the intact receptor with molecular weight of81,000. In addition to a double band of molecular weight ofabout 80,000, the purified @y-immunoglobins from the antiserum recognized in the DE52 pool another major band of60,000 and three other minor bands. Recently, the preparationof polyclonal antibodies to Ca2@/phospholipid-dependent protein kinase reported by Girard et aL (35) also showed a dominant recognition of a Mr 67,000 protein from the brain extract.Whether the Mr 60,000 protein is a degradation product of thepurified receptor remains to be investigated.

ACKNOWLEDGMENTS

We are indebted to Sandy White for typing and to Dr. Ulrike Lichtifor critical reading of the manuscript.

REFERENCES

I . Diamond, L., O'Brien, T. G., and Baird, W. M. Tumor promoters and themechanism of tumor promotion. Adv. Cancer Res., 32: 1—74,1980.

2. Hecker, E., Fusenig, N. E., Kunz, W., Marks, F., and Thielmann, H. W.(eds.) Carcinogenesis—A Comprehensive Survey, Vol. 7. New York: RavenPress,1982.

3. Slaga, T. J. (ed.) Mechanism of Tumor Promotion, Vol. 2. Boca Raton, FL:CRC Press, 1984.

4. Blumberg, P. M. In vitro studies on the mode of action of the phorbol esters,potent tumor promoters: Part I. CRC Crit. Rev. Toxicol., 8: 153—197,1980.

5. Blumberg, P. M. In vitro studies on the mode of action of the phorbol esters,

a b cdFig. 5. Immunoblot of pooled material from DE52 column chromatography

and purified protein kinase C. Samples were run on a 12.5% SDS/polyacrylamidegel and electroblotting was carried out at 60 V overnight. Proteins were stainedeither by Amido-Black (lanesa and b)orby Bio-Rad Immun-Blot(goat antirabbithorseradish peroxidase) assay using ‘y-immunoglobulinspurified from antiproteinkinase C antiserum (lanes c and d). Lane a, molecular weight standards; lanes band c, pooled material from DE52 column chromatography; lane d, purifiedprotein kinase C. K, thousands.

minor band of molecular weight of about 42,000 and two minorbands of 36,000 were also observed (Fig. 5, lane c).

DISCUSSION

The recent convergence of two originally independent research fields, Ca2'-activated phospholipid-dependent proteinkinase and phorbol ester tumor promoter receptor, has hadprofound impact on our understanding of the mechanism ofaction of the tumor promoting phorbol esters. However, thedifficulty in obtaining adequate quantities of purified stablereceptor had seriously impeded study of numerous aspects ofthe biochemistry of the phorbol ester receptor. We have therefore devoted effort to the methodological development pf asuitable purification protocol. The purification scheme reportedin this paper yielded up to 170 @tgofthe purified receptor within2 days after the initial DE52 column chromatography. In addition, conditions for stabilization of the purified receptorweredetermined. Under these conditions, protein kinase C andphorbol ester binding activities were not significantly lost in 2mo.

The advantages of the purification protocol presented in thispaper include the purity of the isolated protein and the reproducibility of the methodology. The SDS-polyacrylamide gelelectrophoresis indicated electrophoretic homogeneity of thepurified receptor and the amino acid compositions obtainedfrom two different preparations gave consistent results. Thefact that the purified receptor had a blocked NH2-terminusfurther substantiates its purity. Based on the average stoichiometry of binding, 0.45:1, we think it most probable that thepurified receptor has been partially inactivated during thecourse of purification. Although it is not likely, the possibilitythat more than one protein with blocked NH2-terminus andwith identical mobility in the SDS/polyicrylamide gel can notbe excluded. Sequence analysis should provide definitive evidence.

A methodological concept in the development of the purification scheme is the shift of elution position from the samecolumn in the presence of a ligand which interacts specificallywith the protein of interest. This approach is a variant of theusual technique of specific substrate elution chromatography

1970

Association for Cancer Research. by guest on August 29, 2020. Copyright 1986 Americanhttps://bloodcancerdiscov.aacrjournals.orgDownloaded from

PURIFICATION OF PROTEIN KINASE C

potent tumor promoters:Part 2. CRC Crit. Rev.Toxicol.,8: 199—234,1981.6. Diamond, L. Tumor promoters and cell transformation. Pharmacol. Ther.,

26:89—145,1984.7. Hecker, E. Structure-activity relationships in diterpene esters irritant and

cocarcinogenic to mouse skin. in: T. J. Slaga, A. Sivak, and R. K. Boutwell(eds.), Carcinogenesis—A Comprehensive Survey, Vol. 2, pp. 11—48.NewYork: Raven Press, 1978.

8. Driedger, P. E., and Blumberg, P. M. Specific binding ofphorbolester tumorpromoters. Proc. NatI. Acad. Sci. USA, 77: 567—571,1980.

9. Blumberg, P. M., Dunn, J. A., Jaken, S., Jeng, A. Y., Leach, K. L., Sharkey,N. A., and Yeh, E. Specific receptors for phorbol ester tumor promoters andtheir involvement in biological responses. In: T. J. Slaga (ed.), Mechanismsof Tumor Promotion, Vol. 3, pp. 143—184.Boca Raton, FL: CRC Press,1984.

10. Nishizuka, Y. The role of protein kinase C in cell surface signal transductionand tumor promotion. Nature (Lond.), 308: 693—698,1984.

I I. Blumberg, P. M., Sharkey, N. A., Konig, B., Jaken, S., Leach, K. L., andJeng, A. Y. Phorbol ester receptors—insights into the initial events in themechanism of action of the phorbol esters. In: H. Fujiki, E. Hecker, R. E.Moore, T. Sugimura, and I. B. Weinstein (eds.), Cellular Interactions byEnvironmental Tumor Promoters, pp. 75—87.Tokyo: Japan Scientific Societies Press, 1984.

12. Castagna, M., Takai, Y., Kaibuchi, K., Sano, K., Kikkawa, U., and Nishizuka,Y. Direct activation of calcium-activated,phospholipid-dependentproteinkinase by tumor promoting phorbol esters. J. Bioi. Chem., 257: 7847—7851,1982.

13. Ashendel, C. L., Staller, J. M., and Boutwell, R. K. Protein kinase activityassociated with a phorbol ester receptor purified from mouse brain. CancerRes., 43: 4333—4337,1983.

14. Leach, K. L., James, M. L., and Biumberg, P. M. Characterization of aspecific phorbol ester aporeceptor in mouse brain cytosol. Proc. NatI. Acad.Sci. USA, 80: 4208—4212,1983.

15. Niedel, J. E., Kuhn, L. J., and Vandenbark, G. R. Phorbol diester receptorcopurifies with protein kinase C. Proc. NatI. Acad. Sci. USA, 80: 36—40,1983.

16. Sando, J. J., and Young, M. C. Identification of high-affinity phorbol esterreceptor in cytosol of EL4 thymoma cells: requirement for calcium, magnesium, and phospholipids. Proc. NatI. Acad. Sci. USA, 80: 2642—2646,1983.

17. Kikkawa, U., Takai, Y., Tanaka, Y., Miyake, R., and Nishizuka, Y. Proteinkinase C as a possiblereceptor protein of tumor-promotingphorbol esters.J. Biol. Chem., 258: 11442—11445, 1983.

18. Parker, P. J., Stabel, S., and Waterfield, M. D. Purification to homogeneityof protein kinase C from bovine brain—identity with the phorbol esterreceptor. EMBO J., 3: 953—959,1984.

19. Konig, B., DiNitto, P. A., and Blumberg, P. M. Phospholipid and Ca@dependency ofphorbol ester receptors. J. Cell. Biochem., 27: 255—265,1985.

20. Kikkawa, U., Takai, Y., Minakuchi, R., Inohara, S., and Nishizuka, Y.Calcium-activated, phospholipid-dependent protein kinase from rat brain.

Subcellular distribution, purification, and properties. J. Biol. Chem., 257:13341—13348,1982.

21. Inagaki, M., Watanabe, M., and Hidaka, H. N-(2-Aminoethyl)-5-isoquinolinesulfonamide, a newly synthesized protein kinase inhibitor, functions as aligand in affinity chromatography. Purification of Ca2@-activated, phospholipid-dependent and other protein kinases. J. Biol. Chem., 260: 2922—2925,1985.

22. Le Peuch, C. J., Ballester, R., and Rosen, 0. M. Purified rat brain calciumand phospholipid-dependent protein kinase phosphorylates ribosomal protein S6. Proc. NatI. Acad.Sci. USA,80: 6858—6862,1983.

23. Uchida, T., and Filburn, C. R. Affinity chromatography ofprotein kinase C-phorbol ester receptor on polyacrylamide-immobilized phosphatidylserine.J.Biol.Chem.,259:12311—12314,1984.

24. Walsh, M. P., Valentine, K. A., Ngai, P. K., Carruthers, C. A., and Hollenberg, M. D. Ca2@-dependent hydrophobic-interaction chromatography. Isolation of a novel Ca2'-binding protein and protein kinase C from bovinebrain. Biochem. J., 224: 117—127,1984.

25. Wise, B. C., Raynor, R. L., and Kuo, J. F. Phospholipid-sensitive Ca2@-dependentprotein kinase from heart. I. Purificationand generalproperties.J. Biol.Chem.,257: 8481—8488,1982.

26. Wolf, M., Sahyoun, N., LeVine, H., III, and Cuatrecasas, P. Protein kinaseC: rapid enzymepurificationand substrate-dependenceof the diacylglyceroleffect.Biochem.Biophys.Res.Commun., 122: 1268—1275,1984.

27. Bradford, M. M. A rapid and sensitive method for the quantitation ofmicrogramquantitiesofprotein utilizingthe principleofprotein-dyebinding.Anal. Biochem.,72:248—254,1976.

28. Leach, K. L., and Blumberg, P. M. Modulation of protein kinase C activityand l3HlphorboI12,13-dibutyratebinding by various tumor promoters inmouse brain cytosoL Cancer Res., 45: 1958—1963,1985.

29. Witt, J. J., and Roskoski, R., Jr. Rapid protein kinase assay using phosphocellulose-paper absorption. Anal. Biochem., 66: 253—258,1975.

30. Laemmli, U. K. Cleavage of structural proteins during the assembly of thehead ofbacteriophage T4 Nature (Lond.),227:680—685,1970.

31. Wray, W., Boulikas,T., Wray, V. P., and Hancock, R. Silver staining ofproteins in polyacrylamide gels. Anal. Biochem., 118: 197—203,1981.

32. Towbin, H., Stahelin, T., and Gordon, J. Electrophoretic transfer of proteinsfrom polyacrylamide gels to nitrocellulose sheets. Proc. NatI. Acad. Sci.USA, 76:4350—4354,1979.

33. Schatzman, R. C., Raynor, R. L., Fritz, R. B., and Kuo, J. F. Purification tohomogeneity, characterization and monoclonal antibodies of phospholipidsensitive Ca2'@.dependentprotein kinase from spleen. Biochem. J., 209: 435—443, 1983.

34. Pogell, B. M. Enzymepurificationby specificelution procedureswith substrate. Methods Enzymol., 9: 9—15,1966.

35. Girard, P. R., Mazzei, G. J., Wood, J. G., and Kuo, J. K. Poiyclonalantibodies to phospholipid/Ca2@-dependent protein kinase and immunocytochemical localizationof the enzyme in rat brain. Proc. Nati. Acad. Sd.USA,82: 3030—3034,1985.

1971

Association for Cancer Research. by guest on August 29, 2020. Copyright 1986 Americanhttps://bloodcancerdiscov.aacrjournals.orgDownloaded from