Affinity Chromatography of Protein Kinase C-Phorbol Ester Receptor on Polyacrylamide-immobilized Phosphatidylserine”

An affinity column, prepared by immobilizing phos- phatidylserine and cholesterol in polyacrylamide, was utilized in the purification of protein kinase C. Protein kinase activity and phorbol ester binding were moni- tored by assaying Ca2+ plus phosphatidylserine-de-pendent phosphorylation of histone H1 and [3H]phorbol dibutyrate binding, respectively. Both activities were present in a cytosolic extract of rabbit renal cortex, eluted together from a DEAE-cellulose column, bound to the affinity column in the presence of Ca2+, and eluted symmetrically upon application of EGTA. Re- covery from the affinity column was high (30-50%) and resulted in as much as a 6000-7700-fold purifi- cation, depending on the region of the DEAE-cellulose peak that was applied. Following affinity column pu- rification, protein kinase and phorbol ester binding activity eluted symmetrically upon gel filtration, with a molecular weight of approximately 80 kDa. A protein of the same size was present in silver-stained gels following sodium dodecyl sulfate-polyacrylamide gel electrophoresis of affinity column purified samples from the DEAE-cellulose peak. From 2-4 other, smaller proteins were also present, their number and relative amounts depending on the region of the DEAE- cellulose peak used. These data indicate that Ca2+-dependent binding to a polyacrylamide-immobilized phospholipid provides a useful technique for purification

A protein kinase activated by a combination of phosphatidylserine and Ca2+ with sensitivity to Ca2+ regulated by diacylglycerol (1)(2)(3)(4)(5)(6)(7) has now been demonstrated in many tissues (1)(2)(3). Since both diacylglycerol generation and Ca2+ mobilization occur in response to many hormones and transmitters (8,9), this protein kinase, termed protein kinase C , appears likely to play a fundamental role in mediating the actions of these agents. In addition, the apparent identity of protein kinase C with a phorbol ester receptor (10)(11)(12), its responsiveness to phorbol esters (7,12), and the fact that phorbol * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
$ Present address, Department of Biochemistry, University of Gunma, School of Medicine, Mebashi 371, Japan.
§ To whom correspondence should be addressed.
"~_ _ _ _ esters cause a redistribution of the enzyme from cytosol to membranes in responsive cells (13) has heightened interest in this enzyme. Current methods of purification require several steps and suffer from low recovery (3,14). In addition, the purified enzyme is relatively unstable, further complicating studies of its regulation and function. In this communication, we report on the use of an affinity column, prepared by immobilizing phosphatidylserine in polyacrylamide, in purification of the protein kinase C-phorbol ester receptor from rabbit renal cortex. Caz+-dependent binding of proteins to the dispersed gel is highly specific, permitting rapid, high recovery of the highly purified protein kinase-phorbol ester receptor. Methods-Renal cortices dissected from kidneys of New Zealand White rabbits were minced and homogenized with a glass homogenizer with a motor-driven Teflon pestle in 9 volumes of 0.25 M sucrose, 20 mM Tris, pH 7.5, 5 mM dithiothreitol, 2 mM EGTA, 2 mM EDTA, and 0.1 mM phenylmethylsulfonyl fluoride, then further homogenized with three bursts of 15 s each on a Polytron homogenizer at setting 8. The homogenate was centrifuged for 30 min at 35,000 X g and the supernatant was used as starting extract.

Materials
DEAE-cellulose Chromatography-Cyclic AMP was added to the cytosolic extract to a final concentration of 1 pM and applied to a column (1.5 X 25 cm) of DEAE-cellulose equilibrated with homogenizing buffer minus sucrose but with 1 p~ cyclic AMP. After application of sample and further washing with column buffer, a 0-0.4 M gradient of KC1 was applied in a total of 500 ml and 5.0-ml fractions were collected.
Preparation of Affinity Column-Phosphatidylserine (5 mg) and cholesterol (25 mg) dissolved in chloroform were combined in a glass scintillation vial and evaporated under nitrogen. Ethanol (0.5 ml) was added and the vial was capped, placed in boiling water, and swirled until the lipids were dispersed. The vial was quickly removed, 5 ml of a solution of 15% acrylamide, 5% BIS was added and vigorously mixed, followed immediately by addition of 50 p1 of 140 mg/ml of ammonium persulfate, 2.5 p1 of TEMED, further mixing, and an additional 50 pl of ammonium persulfate. The mixture was transferred to a test tube (13 X 100 mm), covered with parafilm and aluminum foil, and allowed to fully polymerize usually overnight at room temperature.
Gel-containing tubes were broken and the rigid white gel was rinsed with water, minced with a razor blade, and homogenized in 20-30 ml of Hz0 with three passes of a loose-fitting (0.25-mm clearance) Dounce homogenizer. The homogenized gel was allowed to settle for 5 min, the supernatant was decanted, and this procedure was repeated at least twice after resuspension in water. Settled gel particles were resuspended in a column buffer containing 5 mM MKS, pH 6. activities. This mixture was pumped at 15 ml/h into a closed 0.5-ml mixing chamber, into which a solution containing 10 mM MES, pH 6.5, 14 mM CaC12, 5 mM dithiothreitol, 0.1 mM phenylmethylsulfonyl fluoride, 300 m M KC1 was also pumped at 15 ml/h. The resulting mixture was pumped at 30 ml/h directly onto the affinity column and, after the total sample was applied, was followed by 15 ml of column buffer, which was then followed by 45 ml of the same buffer but with CaC12 reduced to 0.1 mM. At this point, eluting buffer, identical to the column buffer but with 2 mM EGTA in place of CaC12, was applied and additional 3.0-ml fractions were collected. Plastic tubes were used to collect fractions. Fractions were assayed immediately after collection and bovine serum albumin was subsequently added, to a final concentration of 1 mg/ml, to a portion of the fraction to stabilize protein kinase activity. Gel Filtration-An aliquot of affinity column purified protein kinase C (100 pl) or a mixture of molecular weight markers (20 pl) was applied to a Spherogel-TSK 3000 SW column (7.5 X 300 mm) connected to a Pharmacia FPLC system. Fractions of 0.22 ml were collected and assayed. Molecular weight markers were monitored by Am.
SDS-PAGE-Samples for SDS-PAGE were prepared by adding 1 volume of 50% trichloroacetic acid to 9 volumes of affinity column eluate, letting stand 30 min on ice, then centrifuging 30 min at 35,000 X g. The acid supernatant was carefully aspirated, 100 pl of 0.0625 M Tris, pH 6.7,2% SDS, 10% glycerol, 40 mM dithiothreitol, and 0.002% bromphenol blue was added, and the samples were heated for 2 min at 100 "C. Samples were electrophoresed on a 10% polyacrylamide gel essentially as described by Laemmli (15) and were stained using a Bio-Rad silver stain kit.
Protein Kinase Assay-Protein kinase activity was assayed in a reaction mixture containing 25 mM Tris, pH 7.5,200 pg/ml of histone H1, 5 mM MgC12, 20 pM ATP, 1-2 X lo6 cpm [y-32P]ATP, 400 p~ EDTA (cytosol and DEAE-cellulose fractions) or 400 p~ EGTA (affinity column fractions), t 50 pg/ml of phosphatidylserine, 0.5 pg/ ml of diolein, and 0.5 mM free Ca2+ in a total volume of 100 pl. Phosphatidylserine and diolein were combined, evaporated under nitrogen, and dispersed in water or 10 mM Tris, pH 7.5, by sonication for 2-5 min at 60 watts with a Heat System-Ultrasonic sonicator. Reactions were initiated by addition of 10-20 p1 of enzyme, incubated for 1 min at 30 "C, and terminated by addition of 1 ml of 10% trichloroacetic acid, 5 mM NaH2,P04, 2 mM ATP, followed by 100 p l of 0.63% bovine serum albumin. Precipitated protein was sedimented by centrifugation, redissolved in 1 N NaOH, reprecipitated with stopping solution, collected and washed on glass fiber filters, and counted in a scintillation spectrometer. At least a 10-fold dilution of extract, and in some cases of column fractions, was necessary to assure proportionality of activity with amount of extract.
Phorbol Ester Binding Assay-Phorbol ester binding was assayed by a modification of a fiber glass filtration method (16). The reaction mixture contained 25 mM Tris, pH 7. PDB was separated from free [3H]PDB by adding 1 ml of 20 mM Tris, pH 7.5, 10 mM MgAc, 1 mM CaC12 and filtering the mixture through 2.4-cm Whatman GF/C glass filters by suction. The tubes and filters were washed five times with 1 ml of filtering solution. The filters were counted in 10 ml of Ready Solv MP (Beckman). As with protein kinase activity, some dilution of extract and of column fractions was necessary to achieve proportionality. Specific binding was calculated as total binding minus nonspecific binding observed in the presence of 3 pM 4p-phorbol 12P-myristate 13a-acetate, usually less than 2% of total counts.
Protein determinations on extracts, column fractions, and purified enzyme were done by a dye-binding assay (17) using 0.01% Serva Blue G in 1.6 N phosphoric acid, 0.8 M methanol with bovine serum albumin as standard.

RESULTS AND DISCUSSION
The protein kinase C activity present in the cytosol of the rabbit renal cortex eluted from DEAE-cellulose at approxi- Elution and assay conditions were as described for Fig. 1 and under "Experimental Procedures." mately 0.1 M KC1 (Fig. l), similar in position to the activities present in the brain (3) and heart (14). Phorbol ester binding activity coeluted with protein kinase C activity. Upon application of a portion of the combined fractions of protein kinase C-phorbol ester binding activity to the affinity column, negligible amounts of both activities passed through the column, while virtually all of the readily detectable protein passed directly through (Fig. 2). Upon eluting with EGTA, protein kinase C and [3H]PDB binding coeluted in a single, sharp peak. Addition of fatty acid free bovine serum albumin (1 mg/ ml final concentration) to plastic collection tubes before or immediately after collection was necessary to effectively stabilize an otherwise highly unstable preparation. The enzyme could not be frozen, but remained fully active and responsive for weeks if kept on ice.
The freshly isolated enzyme was almost totally dependent upon ea2+ and could be activated 5-lO-fold by diolein in the presence of phosphatidylserine and low Ca2+.' Phorbol ester C. R. Filburn and T. Uchida, manuscript in preparation.

Affinity Chromatography of
Protein Kinase C 12313 Affinity column a Numbers in parentheses indicate per cent recovery of activity applied to the column. The early and late fractions used were located two fractions before and after, respectively, the peak fraction of protein kinase-phorbol ester binding activities eluting from a DEAE-cellulose column similar to that shown in Fig.  1. binding exhibited high affinity, with a KO of 0.5-1 nM (not shown), in the range demonstrated for other preparations of phorbol ester receptor (10)(11)(12). While [3H]PDB binding was difficult to estimate accurately in crude extracts, purification from the DEAE-cellulose step through the affinity column step paralleled that of protein kinase C. An increase in purification of as much as 1000-fold was achieved for both activities (Table I). Binding capacity for ['HJPDB varied, depending on the source of phosphatidylserine (highest with phosphatidylserine from Avanti, lowest with Sigma) and on the DEAE-cellulose fractions applied to the affinity column (Table I).
Specific activity of both PDB binding and protein kinase activity was lowest in affinity column eluates derived from fractions on the ascending side of the peak of activity eluting from the DEAE-cellulose column, highest when fractions from the descending side of the peak were used ( Table I).
This varying specific activity appears to be a consequence of binding to the phosphatidylserine affinity gel of proteins that were present in the early fractions but absent in the later fractions. Electrophoresis of affinity column eluates from an early and a late fraction revealed a t least five proteins in the early fraction of 23,28,60,62 and 80 kDa, but only an 80 and a 60-62-kDa doublet in the late fraction (Fig. 3A). Upon gel filtration of an affinity column purified preparation, both [3H] PDB binding and protein kinase C activity eluted at a position of 80 kDa (Fig. 3B), in close agreement with previous reports on the purified enzyme (3,14). The binding capacity of similar late fractions was 3.7-5.0 pmollpg of protein, or 0.3-0.4 mol/ mol of protein, assuming a molecular weight of 80 kDa. Since some of this protein is not the enzyme itself (Fig. 3B), actual capacity is higher and compares well with the 0.9 mol/mol reported for a preparation from brain (11).
Since additional proteins bind to the affinity column when combined fractions from the DEAE-cellulose peak are used, it is apparent that some additional purification step is needed before or after the affinity column step in order to achieve purification of larger amounts of protein kinase C. The capacity and properties of the lipid-gel matrix itself present no barrier to larger scale use. Combined DEAE-cellulose fractions representing the protein kinase C from 15-20 g of renal taining protein kinase C activity. Molecular weight markers were glutamate dehydrogenase (290 kDa), lactate dehydrogenase (140 kDa). enolase (67 kDa), adenylate kinase (32 kDa), and cytochrome ELUTION VOLUME lmll FIG. 3. SDS-PAGE and gel filtration of protein kinase C purified by affinity chromatography. A , silver stain patterns after SDS-PAGE of preparations derived from an early fraction ( l a n e A, 2 fractions ureceding the ueak) and a late fraction ( l a n e E ) of the DEAE-celluloie peak i f activity (Fig. 1). Molecular weight markers were phosphorylase b (93 kDa), bovine serum albumin (67 kDa), and ovalbumin (45 kDa). E , gel filtration of affinity column eluate con-c (12.4 kDa). Protein Kinase C cortical tissue have been applied to an affinity column (1.6 X 5.0 cm), with only a minimal amount passing through the column. Recovery was usually reduced (10-25%), probably due in part to a more prolonged exposure of the kinase to a Ca2+-dependent protease present in the kidney and other tissues (18, 19). This protease is physically similar to protein kinase C, coelutes with the kinase from DEAE-cellulose (18, 20), and is very active against it (20). The lipid-gel matrix described here has, in fact, been used successfully in purification of milligram quantities of protein kinase C from kilogram amounts of bovine brain. 3 It should be appreciated that the total binding capacity of the final gel matrix, which may contain as much as 1.2 Fmol of phosphatidylserine/ml of gel, is likely to vary with the surface area of the particles used. Any increase in surface area by reduction in particle size will result in a decrease in maximal flow rate through packed particles. Particles prepared as described here have proven rigid enough to permit a high flow rate, but small enough to provide adequate interaction of the kinase with and binding to the trapped cholesterol-phosphatidylserine micelles. In addition, affinity columns prepared as described here have performed very reproducibly and have been regenerated by simple buffer washes and reused as many as 4-5 times with no loss in binding efficiency and only moderate reduction in enzyme recovery. Further investigation of various parameters pertaining to preparation and use of lipid-gel matrix may reveal conditions that further enhance its effectiveness. Simply increasing the ratio of cholesterol to phosphatidylserine beyond that described here appears to increase r e~o v e r y .~ In summary, the purification procedure described here provides a simple, rapid means of obtaining a highly purified preparation of protein kinase C-phorbol ester receptor. In addition, it illustrates the usefulness of a technique of embedding a lipid in polyacrylamide for preparation of an affinity gel. It is possible that this method may be extended to preparation of affinity gels with other lipids exhibiting liganddependent interactions with soluble cellular components. The demonstration that proteins other than protein kinase C bind in a Ca2+-dependent manner to such a phosphatidylserine affinity gel (Fig. 3A) raises interesting questions as to the function of these and possibly other proteins and to the role of Ca2+ in these protein-phospholipid interactions. It is possible that some of these proteins are protease-derived fragments of protein kinase C, a question that is currently under investigation.