THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1991 by The American Society for Biochemistry and Molecular Biology, Inc

Vol. 266, No. 14, Issue of May 15, pp. 9285-9294,1991 Printed in U. S. A.

Protein Kinase C Isotypes and Signaling in Neutrophils DIFFERENTIAL SUBSTRATE SPECIFICITIES OF A TRANSLOCATABLE, CALCIUM- AND PHOSPHOLIPID- DEPENDENT @-PROTEIN KINASE C AND A NOVEL CALCIUM-INDEPENDENT, PHOSPHOLIPID- DEPENDENT PROTEIN KINASE WHICH IS INHIBITED BY LONG CHAIN FATTY ACYL COENZYME A*

(Received for publication, July 30, 1990)

S. Majumdar, M. W. RossiS, T. Fujiki, W. A. Phillips$, S. Disa, C. F. Queen, R. B. Johnston, Jr., 0. M. Rosent, B. E. Corkeyll, and H. M. Korchak From the Departments of Pediatrics and BiochernistrylBiophysics, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104 and the llDepartment of Medicine, Boston University Medical School, Boston, Massachusetts 02118

Neutrophils possess a classical Ca2+, phosphatidyl serine (PS) and diglyceride (DG)-dependent protein kinase C (8-PKC) which was translocatable from cy- tosol to membrane in response to elevated Ca2+ in the physiologic range or to pretreatment with phorbol my- ristate acetate (PMA). The translocatable 8-PKC was purified from neutrophil membranes prepared in the presence of Ca2+, eluted with EGTA and subjected to hydroxyapatite chromatography. An 80-kDa protein possessing Ca/DG/PS-dependent histone phosphoryl- ating activity was recognized by a monoclonal anti- body to @-PKC but not to a-PKC or y-PKC. A cytosolic kinase activity remaining after Ca2+-induced translo- cation of 8-PKC was dependent on PS and DG but did not require Ca”. This novel Ca2+-independent, PS/DG- dependent kinase, termed nPKC, eluted from hydrox- yapatite between a-PKC and B-PKC, ran as a 76-kDa band on sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and was reactive to a polyclonal con- sensus antibody but not to monoclonal antibodies to a- PKC, 8-PKC, or y-PKC. Long chain fatty acyl-CoA, but not the corresponding free fatty acids, inhibited nPKC in the 1-10 MM range. The chemotactic peptide met-Leu-Phe triggered prompt but transient in- creases in neutrophil long chain fatty acid acyl-CoA, suggesting that nPKC is regulated by fatty acyl-CoA as well as DG during neutrophil activation. Purified 8- PKC phosphorylated a number of cytosolic proteins in a Ca2+-dependent manner, including a major 47-kDa cytosolic protein, which may be implicated in super- oxide anion generation. In contrast, nPKC did not phosphorylate the 47-kDa protein, but phosphorylated numerous cytosolic proteins in a Ca2+-independent manner, including a 66-kDa protein which was not phosphorylated by 8-PKC. Differences in location, sub- strate specificity, and cofactor dependence between nPKC and 8-PKC suggest these kinases may play se- lective roles in the activation sequence of the neutro- phil.

* This work was supported by Grants AI 24840, AI 24782, NS17752, and DK35914 from the National Institutes of Health. 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: 3M Pharmaceuticals, St. Paul, MN 55144. 3 Present address: Royal Melbourne Hospital, Melbourne, Victoria,

t Deceased. Australia.

Protein kinase C (PKC)’ has become a focus of studies on signaling in many cell types (1-3) including the neutrophil (4-8). A role for PKC as a positive signal for the activation sequence of neutrophils has been proposed, since activators of PKC such as phorbol esters and permeant diacylglycerols (9, 10) activate a plasmalemma1 NADPH oxidase to generate 0;. PKC can also act as a negative signal and can down- regulate ligand-induced signaling (11). The ability of PKC to trigger positive and negative pathways in neutrophils suggests the activity of multiple PKC isotypes.

Molecular cloning and sequencing have revealed that PKC exists as a family of multiple isozymes (12-14). The classical Ca2+, phosphatidylserine (PS) and diglyceride (DG)-depend- ent isotypes are referred to as a-PKC, PI-PKC, 011-PKC, and yPKC; more recent work has demonstrated 6 - , e - , and {“PKC isotypes which lack Ca2+ sensitivity (12-20). Tissue distribu- tion is different for each isotype suggesting functional speci- ficity. It has been proposed that different PKC subspecies, having different cofactor and substrate specificities, could play different roles in signaling pathways.

Activation of PKC can serve as a link between the cell responses and ligand-induced hydrolysis of phosphoinositides and other phospholipids by phospholipase C or D to form DG. PKC is activated by DG in the presence of PS and Ca2+. Generation of DG and elevation of cytosolic Ca2+ has been observed in neutrophils activated by the chemotactic peptide met-Lcu-Phe (21, 22). The demonstration that cis-unsatu- rated fatty acids such as arachidonic acid (23) and its metab- olite lipoxin A (24) can activate PKC, also links ligand- initiated triggering of phospholipase A2 to protein phosphoryl- ation. In addition to activation of PKC by cofactors such as Ca2+ and DG, PKC may also be subject to negative regulation. Specific types of glycolipids, especially sphingolipids (25) and long chain fatty acyl-CoA (26, 27) inhibit PKC in uitro. Inhibition may be of importance in the sequential activation of different PKC isoforms.

In the present study, we have demonstrated that @-PKC, the major Ca/PS/DG-dependent PKC isotype in neutrophils (28), is translocated from cytosol to membrane in the presence of elevated Ca2+ in the physiological range or in response to pretreatment with phorbol esters. The Ca2’-elicited translo-

The abbreviations used are: PKC, protein kinase C; met-Leu- Phe, N-formyl-methionyl-leucyl-phenylalanine; O;, superoxide an- ion; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electro- phoresis; HEPES, N-2-hydroxyethylpiperazine-N’-2-ethanesulfonic acid; PMA, phorbol myristate acetate; EGTA, [ethylene- bis(oxyethylenenitrilo)]tetraacetic acid; PS, phosphatidylserine; DG, diglyceride; CoASH, coenzyme A; PMSF, phenylmethylsulfonyl flu- oride; BSA, bovine serum albumin.

9285

9286 Cofactor and Substrate Specificities for Phospholipid-dependent Protein Kinases in Neutrophils

cation of P-PKC from the cytosol to the membrane, leaves a Ca2+-independent, but PS/DG-dependent kinase in the cyto- sol. This novel nontranslocated protein kinase, which we have termed nPKC, is PS/DG-dependent but is independent of Ca2+. The substrate specificity of nPKC was different from that of @-PKC. nPKC phosphorylated a number of cytosolic proteins of molecular mass 45, 58, and 80 kDa plus a 66-68- kDa protein, but not a 47-kDa band. In contrast, @-PKC phosphorylated a 47 kDa but not the 66-kDa band. In addition nPKC, but not @-PKC, was inhibited by micromolar levels of fatty acyl-CoA, but not by free fatty acids. The sensitivity of nPKC to fatty acyl-CoA could play a role in the activation sequence since Met-Leu-Phe elicited a prompt but transient increase in long chain fatty acyl-CoA. The differences in cofactor specificity between nPKC and @-PKC provides a possible mechanism for sequential activation of these differ- ent kinases. Variation in localization and substrate specificity between P-PKC and nPKC suggests selective roles for these kinases in the signaling for neutrophil responses.

MATERIALS AND METHODS

Reagents-Reaction buffer A (Ca" depleted) consisted of 50 mM Tris-HCI, pH 7.5, 50 mM EGTA, 1 mM phenylmethylsulfonyl fluoride (PMSF), anti-protease mixture, and 50 mM 2-mercaptoethanol. Anti- protease mixture consisted of 0.33 mM leupeptin, 0.35 mM antipain, 0.24 mg/ml chymostatin, 0.35 mM pepstatin, and 4.8 TIU/ml apro- tinin. Extraction buffer B (Ca'+ present) consisted of 50 mM Tris- HC1, pH 7.5, plus CaC12, pH 7.5, to reach a final concentration of 500 nM Ca", 1 mM PMSF, anti-protease mixture, and 50 mM 2-mercap- toethanol. The Caz+ buffer B was prepared using stock solutions of 50 mM EGTA (pH 7.5, 25 "C) and of 50 mM EGTA with 50 mM CaCI', pH 7.5, 25 "C (29). Final Ca'+ concentrations were calculated by iterative calculation of the equilibria for H', Mg'+ and Ca" using the dissociation constants given in Martell and Smith (30). Bovine serum albumin (BSA), EGTA, PMSF, and 3,3'-diaminobenzidine were purchased from Sigma. Horseradish peroxidase-conjugate goat anti-rabbit IgG and horseradish peroxidase-conjugate sheep anti- mouse IgG were purchased from Cappel.

Preparation of Human Neutrophils and Cytoplasts-Neutrophils were prepared using standard isolation techniques (31) followed by dextran sedimentation and hypotonic lysis to remove red cells. Cells were suspended in a HEPES buffer, pH 7.5, having the composition Na+ 150 mM, K+ 5 mM, Ca2+ 1.29 mM, M$+ 1.2 mM, C1- 155 mM, and HEPES 10 mM (32). Cytoplasts were prepared from neutrophils, as previously described (33,34).

Preparation of Subcellular Fractions-Neutrophils or neutrophil cytoplasts were resuspended to a concentration of 5 X 106/ml in the appropriate ice-cold extraction buffer and sonicated for 5-10 s at 4 "C (whole sonicate fraction), followed by centrifugation at 100,000 X g for 1 h at 4 "C. The supernatant (cytosolic fraction) was collected and the pellet resuspended by sonication (5 s) at 4 'C to the original

tion studies, 0.2% Triton X-100 was added to the pellet before volume in extraction buffer (particulate fraction). For the transloca-

sonication. Fractions were stored at 4 "C and assayed within 48 h. Protein Kinase CAssay-PKC was assayed in the presence of Ca2+,

PS, and DG by measuring the incorporation of 32P into histone type III-S as previously described (4, 7). Ca'+-independent PKC (nPKC) was assayed in the presence of PS and DG, in the absence of Ca'+, using the same technique. Aliquots equivalent to fractions from 1.25

35 mM Tris-HC1, pH 7.5,0.01% Triton X-100, 10 mM 2-mercaptoeth- X lo6 cells were incubated in a 250-pl reaction mixture consisting of

anol, 0.2 mM PMSF, 5 pg/ml leupeptin, 0.4 mM EGTA, 10 mM MgCL, 20 pg/ml phosphatidylserine (Avanti Polar Lipids), 2 pg/ml of the specific DG, 1,2-dioleoyl-rac-glycerol (Sigma), 160 pg/ml histone type III-S (Sigma), and 50 p~ (1 pCi) [y3'P]ATP (ICN Radiochemicals) in the presence or absence of 0.6 mM CaCI2. Incubation was at 30 "c for either 30 min (cytosol fraction) or 60 min (particulate fraction). The reaction was stopped by addition of 1 ml of ice-cold 25% trichlo- roacetic acid, followed by 25 p l of BSA (20 mg/ml) (Sigma fraction V) as a carrier. Precipitates were collected on 0.45 pm (type HA) Millipore filters by vacuum filtration. The filters were washed with 12% cold trichloroacetic acid and counted for "P using liquid scintil- lation spectroscopy. Activity was measured as picomoles of 32P incor- porated/minute/1O7 cell equivalents. The non-phospholipid-depend-

ent phosphorylation was subtracted from the total amount of 32P incorporated in order to determine the total specific Ca-dependent PKC activity or nPKC.

Electrophoresis and Immunoblotting-Whole sonicate, cytosol, and particulate fractions from neutrophil cytoplasts, partially purified translocatable Ca'+-dependent protein kinase C or purified non- translocatable Ca'+-independent nPKC were separated on 7.5% SDS gel under reducing conditions (35) and transferred to nitrocellulose paper (36). The nitrocellulose paper was blocked overnight with 3% BSA in Tris-saline buffer, pH 7.5. We have used two panels of anti- PKC antibodies. Anti-peptide polyclonal antibodies to PKC consen- sus (ILKKDVIVQDDDVD, residues 381-394 in the C3 region), cy- PKC (AGNKVISPSEDRRQ, residues 313-326 in the V3 region), B- PKC (GPKTEEKTANT-ISKFD, residues 313-329 in the V3 region) and y-PKC (NYPLELYERVRTG, residues 306-318 in the V3 region)

AYQPYGKSVD, residues 528-537 in the C4 region, was obtained (37). A second PKC consensus antipeptide antibody against

from Oncogene Science. Specificity of the antipeptide antiserum was checked by solid-phase radioimmunoassay. PKC purified by DEAE column chromatography from cytosol of rat brain synaptosomes gave one 80-kDa hand on immunoblots using the consensus, CY-, 8-, and y- PKC antipeptide antibodies (results not shown). The nitrocellulose blot was incubated for 2 h at room temperature with a 1 in 200 dilution of antipeptide antibody followed by incubation for 45 min with horseradish peroxidase conjugate/goat anti-rabbit IgG. A panel of murine monoclonal antibodies to the rat brain PKC isotypes, CY-

PKC, B-PKC, and y-PKC were purchased from Seikagaku America Inc. For studies with these monoclonal antibodies, the nitrocellulose blot was incubated for 2 h at room temperature with 2.5 pg/ml monoclonal antibody followed by incubation for 45 min with horse- radish peroxidase conjugate/sheep antimouse IgG, then extensive washing with Tris-saline buffer, pH 7.5. Immunoreactive bands were visualized with 3,3'-diaminohenzidine as a chromogenic substrate for horseradish peroxidase.

Elution of PKC from Gels-PKC activity was demonstrated in the 80-kDa polypeptide band from SDS-PAGE. Cytosol (2.8 mg of pro- tein) was run on a preparative gel, the band sliced from the gel, and eluted overnight at 4 "C in a buffer containing 20 mM Tris-HCI, pH 7.5, 50 mM NaC1, and 50 mM 2-mercaptoethanol (38). This eluate was dialyzed with the same buffer to remove SDS, concentrated in an Amicon PM-10 filter and assayed for Ca/PS/DG-dependent his- tone phosphorylating activity.

Purification of Translocatable PKC-Neutrophils (lo' cells) were sonicated in a Ca2+-containing buffer (buffer B), in order to achieve translocation of the PKC to the membrane fraction (7), and the sonicate centrifuged at 100,000 X g for 60 min at 4 "C. The pellet fraction was resuspended in 5 mM EGTA (buffer A) to elute the translocated PKC, recentrifuged, and the supernatant (crude 8-PKC) subjected to hydroxyapatite chromatography. The column was equil- ibrated with 20 mM phosphate buffer, pH 7.5, containing 10% glyc- erol, 0.5 mM EGTA, 0.5 mM EDTA, 10 mM 2-mercaptoethanol (39). The crude PKC fraction was eluted with a continuous gradient of 20- 320 mM phosphate buffer, pH 7.5, containing 10% glycerol, 0.5 mM EGTA, 0.5 mM EDTA and 10 mM 2-mercaptoethanol. Fractions were pooled in steps of 10 mM phosphate, and assayed for Ca/PS/DG- dependent histone phosphorylating activity. Fractions were kept at -80 "C with excess of protease inhibitors to avoid proteolytic cleav- age. Stored material was stable for at least 3 weeks.

Purification of Nontranslocatable, Ca'+-independent Protein Ki- nase-Neutrophils (IO') were sonicated in the Ca'+-containing buffer (buffer B), in order to achieve translocation of the Ca'+-dependent p- PKC to the membrane fraction, and the sonicate was centrifuged at 100,000 X g for 60 min. The supernatant was subjected to DEAE

20 mM Tris, pH 7.5, 0.5 mM EGTA, 0.5 mM EDTA, 10 mM 2- (DE52) column chromatography. The column was equilibrated with

mercaptoethanol, and 10% glycerol (39), and eluted with a gradient of 20-200 mM NaCI. The Ca'+-independent nPKC eluted at 110 mM NaCI. The DEAE-purified kinase was further subjected to hydroxy- apatite chromatography (HPLC). The column was equilibrated with 20 mM phosphate buffer, pH 7.5, containing 10% glycerol, 0.5 mM EGTA, 0.5 mM EDTA, 10 mM 2-mercaptoethanol (391, and eluted with a gradient of 20-320 mM phosphate buffer, pH 7.5, containing 10% glycerol, 0.5 mM EGTA, 0.5 mM EDTA, and 10 mM 2-mercap- toethanol. Fractions were pooled in steps of 10 mM phosphate and assayed for PS/DG/-dependent PKC activity; appropriate fractions were stored at -80 "C with excess protease inhibitors.

Phosphorylation of Endogenous Substrates by Purified, Ca2+/Phos- pholipid-dependent, Translocatable PKC and Ca2+-independent, Phos-

Cofactor and Substrate Specificities for Phospholipid-dependent Protein Kinases in Neutrophils 9287 pholipid-dependent Protein Kinase, nPKC-Neutrophil cytoplasts or neutrophils were sonicated in EGTA (buffer A) and centrifuged a t 100,000 X g for 60 min a t 4 "C. The supernatant and the pellet fraction were used as the endogenous substrate for purified Ca- dependent p-PKC and Ca-independent nPKC. The reaction mixture contained 20 mM Tris-HC1, pH 7.5,6 mM MgCI,, 0.25 mM EGTA, 1 mM NaF, 0.1 mM sodium vanadate, 20 pg/ml phosphatidylserine, 2 pg/ml 1,2-dioleoyl-rac-glycerol, 50 p~ [y-"'P]ATP, 10-15 pg of en- dogenous substrate and purified PKC (4-5 pg of protein of specific activity 132 pmol/min/mg protein for 8-PKC and 262 pmol/min/mg protein for nPKC) in the presence or absence of 0.6 mM CaCI2, in a total volume of 100 PI. Incubation was carried out for 30 min a t 30 "C (7). The reaction was stopped by addition of 100 pl of ice-cold 25% trichloroacetic acid together with 5 pl of 20 mg/ml BSA as carrier protein. After standing on ice for 30 min, the mixture was centrifuged at 4 "C for 5 min (40). The pellet was solubilized with Laemmli buffer, boiled for 5 min, separated on SDS-PAGE (7.5%). and subjected to autoradiography.

Acyl-CoA Analysis-Suspensions of neutrophils were acidified with ice-cold trichloroacetic acid (0.6 N final concentration) a t designated times after addition of fMet-Leu-Phe. Supernatants containing free CoASH and acid soluble CoA esters up to a chain length of about eight carbons were separated from precipitated protein and long chain acyl-CoA esters by centrifugation for one min a t 12,000 X g (41). The supernatant was neutralized and trichloroacetic acid removed by extraction with ether. The mixtures of equal volumes of ether and trichloroacetic acid were vortexed about 30 s, centrifuged briefly to separate the phases, and the trichloroacetic acid-containing ether phase completely removed by suction. The extraction procedure was repeated five times until neutrality was reached (42). The neutralized extract and pellet fractions were dried and stored a t -70 "C until analyzed. Just prior to analysis, the CoA esters were hydrolyzed in 1 mM dithiothreitol by adjusting the pH to 11.5 with KOH and incu- bating for 10 min a t 55 "C. Correction was made for loss of CoA during extraction using a palmitoyl-CoA standard treated in the same manner as the samples. Total CoA was determined in each of the fractions by enzymatic analysis using a-ketoglutarate dehydrogenase (41). The total CoA pool was 27.0 f 0.8 pmol/106 cells and did not change during the course of the experiments. Experiments were carried out in duplicate or triplicate with each sample containing 400 X 10" PMN. Results are expressed as the mean f S.E. Protein was determined by the method of Lowry (43).

RESULTS

Translocation of Ca/PS/DG-dependent Histone Phospho- rylating Activity in Response to Elevated Ca" and to Phorbol Myristate Acetate-Cytoplasts were used to demonstrate re- versible Ca2+-induced PKC translocation, and irreversible phorbol ester-induced PKC translocation from cytosol to membrane. PKC activity (Ca/PS/DG-dependent histone (111s) phosphorylation) of unstimulated cytoplasts disrupted in Ca"-depleted buffer (buffer A) was found predominantly in the cytosolic fraction, with little activity in the pellet fraction (Fig. 1). In contrast, extraction of cytoplasts in the presence of 500 nM Ca'' (buffer B) resulted in elevated PKC activity in the particulate fraction with a concomitant de- crease in cytosolic activity (Fig. l). A Ca2+-independent but PS/DG-dependent histone phosphorylating activity remained in the cytosol after Ca2+-induced translocation of classical PKC. Pretreatment of cytoplasts with 1 pg/ml PMA a t 37 "C for 5 min, followed by disruption in EGTA (buffer A) induced a reduction in cytosolic PKC activity, as compared to resting cytoplasts, and an increase in PKC activity in the particulate fraction (Fig. 1). These findings suggest that elevated free Ca2+ in the physiologic range (500 nM) was effective in trans- locating PKC activity from cytosol to membrane. In addition, PMA-induced translocation was not reversible and was ob- served in the absence of Ca2+ in the extraction medium, whereas Ca"-induced translocation was readily reversible in the presence of Ca'+ chelators such as EGTA (7).

Translocation of PKC Immunoreactivity in Response to El- evated Ca2+ and to Phorbol Myristate Acetate-To demon-

1254 T T

100

- .er !? 75 C 0

50

25

0 Ca PMA - -

- + - - + - - + - + - - + " + Whole Cytosol Pellet

Sonicate

FIG. 1. Effect of Ca2+ and PMA on the subcellular distribu- tion of protein kinase C activity in neutrophil cytoplasts. Neutrophil cytoplasts were sonicated in the presence of EGTA (ex- traction buffer A) or in 500 nM Ca2+ (buffer R); cytoplasts were pretreated with 1 pg/ml of PMA (5 min a t 37 "C) and sonicated in a Ca"-depleted buffer (buffer A). The whole sonicate, cytosol, and particulate fractions were assayed in the presence of PS, dioleoyl- glycerol, and sufficient Ca2+ to override the Ca2+ buffering (see "Materials and Methods"). Data expressed as percentage of whole sonicate activity prepared in Ca"-depleted buffer, where 100% value = 73.9 f 9.8 (n = 5) pmol of "P/min/lO' cell equivalents.

L A N E 1 2 3 4 1 6 7 0 9

PKC-; I

i 39 LDa

7- TS LDa

!- SO LDo

CO*+ - PMA - -

+ - - + - - + - -

+ - + - - + I"

Whole Cytosol Partleulot. Sonlcote Fraction

FIG. 2. Calcium and phorbol ester-dependent distribution of protein kinase C in cytoplasts demonstrated by immuno- reactivity to a consensus (C3) peptide sequence antibody. Resting cytoplasts were sonicated in extraction buffer depleted of Ca2+ (buffer A) or in the presence of 500 nM Ca'+ (buffer B). Pretreatment of cytoplasts with 1 pg/ml of PMA was carried out for 5 min a t 37 "C, followed by extraction with buffer A. Fractions were run on 7.5% reduced SDS-PAGE and transferred to nitrocellulose (see "Materials and Methods" for details of blotting procedures). Each lane contains sample equivalent to 9 X lo6 cytoplasts. Whole sonicate, cytosol, and particulate fraction were prepared in the pres- ence and in the absence of Ca'+ or after PMA pretreatment as indicated. Representative experiment of n = 3. Molecular mass stand- ards are shown on the right margin and PKC is indicated by an arrow at the left.

strate that translocation of histone phosphorylating activity was due to physical translocation of PKC, we used a consensus (C3) antipeptide antibody to PKC. Whole cytoplast sonicates, cytosol and particulate fractions from cytoplasts, were pre- pared in the presence and in the absence of Ca" and proteins were separated by electrophoresis. Immunoblots of these gels demonstrated the presence of an immunoreactive species of approximately 80 kDa in resting cytoplasts extracted in the presence or in the absence of Ca", as well as in cytoplasts that had been pretreated with PMA (Fig. 2). The level of immunoreactivity in the whole cytoplast sonicates was not significantly different between the different treatments, in- dicating that elevated Ca'+ or PMA pretreatment did not deplete the cytoplasts of PKC. Sonication of cytoplasts in a Ca"-containing buffer (500 nM) caused loss of an 80-kDa immunoreactive band from the cytosol and an increase in an

9288 Cofactor and Substrate Specificities for Phospholipid-dependent Protein Kinases in Neutrophils

80-kDa band in the particulate fraction (Fig. 2). When cyto- plasts were pretreated with PMA and extracted in EGTA (buffer A), there was also loss of a cytosolic, 80-kDa band accompanied by an increase in an 80-kDa band in the partic- ulate fraction (Fig. 2).

To confirm that the 80-kDa translocatable band contained PKC activity, the 80-kDa band was cut from the gel, eluted with Tris buffer, and dialyzed to remove the SDS. The di- alyzed eluate contained PKC measured as Ca/DG/PS-de- pendent ability to phosphorylate histone 111. PKC activity of 61.4 f 5.33 ( n = 3) pmol/mg protein/min was eluted from the gel. In contrast, elution of the 50-kDa band yielded zero histone phosphorylating activity.

Distribution of PKC Zsoforms-To demonstrate which PKC isotypes were present in neutrophil cytoplasts and to identify which isotype(s) were translocated in response to elevated Cat+ or to PMA, we used antipeptide antiserum and mono- clonal antibodies to a-PKC, P-PKC, and y-PKC. Immuno- blots of gels using the antipeptide antibody to a-PKC dem- onstrated a prominent 50-kDa band, which was observed in both cytosol and particulate fractions; elution of this 50-kDa band from a preparative band gave no histone phosphorylat- ing activity (results not shown). A faint immunoreactive band of 80 kDa was observed in the cytosol (Fig. 3A). The distri- bution of these a-PKC immunoreactive bands did not change in response to elevated Ca2+ or to PMA pretreatment.

By using antipeptide antibody to P-PKC, a major immu- noreactive band of 80 kDa was demonstrated in whole soni- cates of cytoplasts (Fig. 3B); the intensity of this 80-kDa band was not altered by elevated Ca'+ or by pretreatment with PMA. The 80-kDa p-PKC immunoreactive band was the predominant band in cytosol prepared in the absence of Ca". Extraction of cytoplasts in the presence of Ca'+ (500 nM) buffer elicited a decrease in the 80-kDa band in the cytosol and an increase in the intensity of the 80-kDa band in the particulate fraction. When cytoplasts were pretreated with PMA (1 pg/ml) and extracted in EGTA (buffer A), a decrease in the 80-kDa, P-PKC immunoreactive band in the cytosol together with an increase in this immunoreactive band in the particulate fraction (Fig. 3B) . The lower M , bands may be due to some proteolytic cleavage, in spite of the presence of numerous antiproteases, including leupeptin. I t is also possi- ble that alternate splicing could yield multiple immunoreac- tive species of PKC having different molecular weights (18, 39).

Immunoblots with antipeptide antibody to y-PKC showed multiple faint bands in the whole sonicate but no 80-kDa band indicating a lack of y-PKC in neutrophil cytoplasts (Fig. 3C). Bands of 90 and 50 kDa were predominantly found in the cytosolic fraction but did not redistribute in response to elevated Ca2+ or to pretreatment with PMA.

Immunoblots using monoclonal antibodies to a-, P-, and y- PKC confirmed that the major isotype was P-PKC, with small amounts of cytosolic 80-kDa a-PKC and no y-PKC; the translocatable isotype was p-PKC (Fig. 30 ) .

Purification of the Translocatable PKC-Neutrophils were sonicated in Ca2+-containing buffer (buffer B) in order to induce translocation of PKC to the membrane fraction (7) and centrifuged at 100,000 X g for 1 h at 4 "C. The pellet was resuspended in 5 mM EGTA to elute the translocatable PKC, and recentrifuged; DEAE chromatography of the supernatant yielded PKC activity eluting at 120 mM NaCl. Purification by hydroxyapatite column chromatography using a phosphate concentration step gradient yielded a peak protein concentra- tion and PKC activity at 130-140 mM phosphate (Fig. 4A). Translocatable neutrophil PKC ran in a similar fashion on

LIS... . - -

- 130 LDa - 7 5 LDa - 50 koa - 39 LDa

COZ+ PYA

- + - - + " + - " + - - + - - +

Whole cg1oso1 Porilculoie Sonlcoie Fracllon

D 1 2 3 4 5 6

a F - 1 ~~

+ PKC

P M A - - + - - + cytosol Particulate

FIG. 3. Calcium and phorbol ester-dependent distribution of protein kinase C isotypes in cytoplasts demonstrated by immunoreactivity with polyclonal antipeptide and mono- clonal antibodies to a-PKC, j3-PKC, and -y-PKC. Resting cyto- plasts were sonicated in extraction buffer depleted of Ca'+ (buffer A) or in the presence of 500 nM Ca'+ (buffer B). Pretreatment of cytoplasts with 1 pg/ml of PMA was carried out for 5 min at 37 'C, followed by extraction with buffer A. Fractions were run on 7.5% reduced SDS-PAGE (see "Materials and Methods" for details of blotting procedures). Each lane contains sample equivalent to 9 X 10' cytoplasts. Whole sonicate, cytosol, and particulate fraction were prepared in the presence and in the absence of Ca'+ or after PMA pretreatment as indicated. Molecular mass standards are shown on the right margin and PKC is indicated by an arrow at the left. Representative experiments of n = 3. A, immunoblot using an anti- peptide antibody to a-PKC. R immunoblot using an antipeptide antibody to P-PKC. C, immunoblot using an antipeptide antibody to -y-PKC. D, immunoblots using monoclonal antibodies to a-PKC, 8- PKC, and -y-PKC.

Fraction

hydroxyapatite to p-PKC from rat brain (Fig. 4B). The peak PKC fraction from neutrophils ran on SDS-PAGE as an 80- kDa band by silver stain and was immunoreactive to a poly- clonal consensus (C4) antibody (Fig. 5). Immunoblots with monoclonal antibodies to a-, P-, and y-PKC showed an 80- kDa band that was reactive to the antibody for 0-PKC but not to those for a- or y-PKC (Fig. 5).

A Ca"-independent, Phospholipid-dependent Kinase,

Cofactor and Substrate Specificities for Phospholipid-dependent Protein Kinases in Neutrophils 9289

h

0.250

60

40

0.200 g 0.150 E 0.100

0

*ou 0.050

0 0.000 0.080 0.160 0.240

Phosphote concentration [MI

0.080 0.1 60 0.240 Phosphate concentration [MI

FIG. 4. Purification of the translocatable protein kinase C by hydroxyapatite column chromatography.A, neutrophils were sonicated in 500 nM Ca'+ (buffer B), then centrifuged for 1 h a t 100,000 X g. The pellet was resuspended in 5 mM EGTA (buffer A) in order to elute the PKC, recentrifuged, and the supernatant loaded onto a hydroxyapatite column. Fractions were eluted with a phosphate buffer gradient (see "Materials and Methods"). PKC activity was assayed as Ca/PS/DG-dependent histone phosphorylating ability. B, the elution pattern for rat brain PKC was obtained by loading cytosol from rat brain synaptosomes onto a hydroxyapatite column and eluting with a phosphate buffer gradient as detailed above. Repre- sentative experiments of rz = 3.

+ 97KDa

f 66KDa

f 43KDa

Silver I C s a p Y ] Stain

lmmunoblot FIG. 5. Isotype specificity of purified Ca2+-translocatable

protein kinase. PKC was purified by successive Ca2+-induced trans- location to membranes and elution with EGTA, DEAE chromatog- raphy, and hydroxyapatite chromatography. The PKC fraction (5 pg/ lane) was run on 7.5% SDS-PAGE, transferred to nitrocellulose, and blotted with a polyclonal antibody to a consensus (C4) peptide se- quence (CS), and with monoclonal antibodies to rat brain n-PKC, B- PKC, and r-PKC (see "Materials and Methods" for details of im- munoblots). Representative experiment of n = 3.

nPKC, in Human Neutrophils-Neutrophils were disrupted in the presence of a physiological level of Ca2+ (500 nM) to elicit translocation of the Ca/PS/DG-dependent kinase to the membrane fraction. The kinase activity remaining in the cytosol, measured as histone phosphorylating activity, was dependent on the presence of PS and DG for optimal activity (Fig. 6). However, this histone-phosphorylating activity was not dependent on the presence of Ca2+ for optimal activity. Indeed, addition of 0.6 mM Ca2+ to the assay medium, an

PS - + + DG - + + CO2+ - - +

FIG. 6. Ca2+-independent, PS/DG-dependent histone-phos- phorylating activity in neutrophil cytosol. Neutrophils were sonicated in the presence of Ca2+ (buffer B) in order to translocate the Ca/PS/DG-dependent kinase to the membrane fraction. Kinase activity remaining in the cytosol was assayed in the presence of PS (20 pglml), DG (2 pg/ml), and in the presence and in the absence of 0.6 mM Ca2+. Data are expressed as cpm ['"PI incorporated/min/lO' cells, mean & S.E. (n = 4).

optimal Ca2+ concentration for p-PKC, caused a slight but not significant decrease in histone phosphorylation (Fig. 6). This novel Ca"-independent, PS/DG-dependent histone phosphorylating activity was designated nPKC.

Purification and Characterization of nPKC-To purify nPKC, neutrophil cytosol was prepared in the presence of 500 nM Ca2+ and applied to a DEAE column. The PS/DG-de- pendent histone phosphorylating activity eluting a t 110 mM NaCl was applied to a hydroxyapatite column. A peak of PS/ DG-dependent histone-phosphorylating activity eluted a t 120 mM phosphate (Fig. 7A). The peak activity ran on SDS- PAGE as a band of 76 kDa that was reactive with a polyclonal consensus (C-4) antibody (Fig. 7B). Western blot analysis using murine monoclonal antibodies to rat brain CY-, p-, and y-PKC revealed no reactive bands (Fig. 7B).

Cofactor Requirements for Purified nPKC and p-PKC- Cofactor requirements were assessed using nPKC and p-PKC purified by hydroxyapatite chromatography. Histone phos- phorylation was measured in the presence of 2 pg/ml DG, 0.6 mM CaCl,, and varying concentrations of Ps; optimal activity was obtained in the presence of 20 pg/ml PS for both nPKC and p-PKC (Fig. 8A). When histone phosphorylation was assayed in the presence of 20 pg/ml PS, 0.6 mM CaC12, and varying concentrations of DG, optimal activity was obtained in the presence of 2 pg/ml DG (Fig. 8B). Thus nPKC dem- onstrates the same requirements for PS and DG as the clas- sical Ca2+-dependent PKC. In contrast, when nPKC- and p- PKC-catalyzed histone phosphorylation was assayed in the presence of 2 pg/ml DG, 20 pg/ml PS, and varying concentra- tions of Ca", no dependence on Ca2+ was observed for nPKC activity whereas 0-PKC showed a marked dependence on Ca2+ (Fig. 8C).

Phosphorylation of Endogenous Substrates by Purified, Ca2+-dependent, Translocatable PKC and Ca2+-independent nPKC-To demonstrate the substrate specificities of purified p-PKC and nPKC, neutrophil cytoplasts were sonicated in buffer containing 5 mM EGTA (buffer A), and cytosol and particulate fractions were prepared to serve as endogenous substrates. Cytosol or particulate fraction, together with [y- "'PIATP, PS, and DG, were added to purified @-PKC or nPKC in the presence or in the absence of Ca2+ (see "Materials and Methods"). The phosphorylated proteins were separated on SDS-PAGE and the gels subjected to autoradiography. Neu- trophil p-PKC strongly phosphorylated 45-, 47-, 54-, 58-, and 97-kDa bands in the cytosol fraction, in a Ca2+-dependent

Cofactor and Substrate Specificities for Phospholipid-dependent Protein Kinases in Neutrophils

300 0.025

250 A 0.020

1504 / f \ \ t 0.01 5

l o 0 l 50 Il’ r+o.olo --0.005

0 0.050 0.100 0.150 0.200 0.250 0.300

- ro.000

2oot Ah \ “ 0.01 5

100 “0.010

50 --0.005

0.050 0.100 0.150 0.200 0.250 0.300 0.000

Phosphate concentration [MI

+ 116KDa

+ 97KDa

nPKC+

S K D a - + 116KDa

+ 97KDa

+ 66KDa

1 J I U cs a p Y I Silver

Slain lmrnuncblol

FIG. 7. Purification and characterization of a Ca’+-inde- pendent, PS- and DG-dependent protein kinase (nPKC). A, Purification of nPKC by hydroxyapatite column chromatography. Cytosol was prepared in the presence of 500 nM Ca” and run on a DEAE column. The fraction eluting a t 110 mM NaCl was applied to a hydroxyapatite column and eluted with a phosphate buffer concen- tration gradient (see “Materials and Methods”). Protein kinase activ- ity was assayed as PS/DG-dependent histone phosphorylation; activ- ity in the absence of PS and DG was subtracted. H , SDS-PAGE and immunohlots of nPKC. The nPKC purified by DEAE and hydroxy- apatite chromatography (10-12 pg/lane) was run on 7.5% SDS-PAGE (see “Materials and Methods”). Molecular mass standards are shown on the right margins. Immunoblots were prepared using a polyclonal antibody to a consensus (C4) peptide sequence (CS), and murine monoclonal antibodies to n-PKC, P-PKC, and y-PKC. Representa- tive experiments of n = 4.

manner (Fig. 9). Weaker phosphorylation of 80- and 210-kDa bands was also observed (Fig. 9). Little phosphorylation of these bands was observed when either the @-PKC or Ca‘+ was omitted from the incubation medium. Identical results were obtained when cytosol prepared from intact neutrophils was used as the endogenous substrate (data not shown). In the presence of nPKC, phosphorylation of 45-, 58-, 66-, and 80- kDa bands was observed (Fig. 9). This pattern of phosphoryl- ation was markedly different from that for @-PKC, where a prominent 47-kDa band was phosphorylated and no phos- phorylation of a 66-kDa band was observed. In contrast to the marked phosphorylation of numerous cytosolic bands catalyzed by @-PKC and nPKC, no enhanced phosphorylation of particulate proteins was observed (Fig. 9). Indeed, addition of @-PKC to the particulate fraction decreased endogenous phosphorylation, and in the presence of nPKC a band of 58 kDa showed less phosphorylation (Fig. 9). This decrease in phosphorylation was observed in spite of the presence of the phosphatase inhibitors, fluoride (1 mM), and vanadate (100 PM).

Fatty Acyl-CoA and nPKC Actiuity-In addition to PS and DG, free fatty acids and their CoA thioesters have been implicated as cofactors in modulating PKC activity (12, 23, 26, 27). The effect of oleic acid and oleoyl-CoA on nPKC activity was tested using crude cytosolic nPKC, i.e. cytosol prepared in the presence of 500 nM Ca’+. Increasing concen- trations of oleic acid from 0 to 10 PM, a concentration that remains below the critical micelle concentration, showed no ability to modify nPKC activity measured as Ca”-independ-

a 100 x

0 I

V X

FIG. 8. Cofactor dependence of 8-PKC and nPKC-catalyzed histone 111s phosphorylation. nPKC and (j-PKC purified from neutrophils by hydroxyapatite chromatography were used to deter- mine the phosphatidylserine ( P S ) , DG, and Ca” requirements for histone phosphorylation. The specific activity of nPKC was 312 pmol/ min/mg protein and the specific activity of 8-PKC was 183 pmol/ min/mg protein. Representative experiments of n = 3. A, PS depend- ence of nPKC and 8-PKC activity. Histone phosphorylation was assayed in the presence of 0.6 mM Ca’+, 2 pg/ml DG, and varying concentrations of PS. R, DG dependence of nPKC and (j-PKC activity. Histone phosphorylation was assayed in the presence of 0.6 mM Ca”, 20 pg/ml PS, and varying concentrations of DG. C, Ca’+ dependence of nPKC and (j-PKC activity. Histone phosphorylation was assayed in the presence of 2 pg/ml DG, 20 pg/ml PS, and varying concentrations of Ca’+ obtained using a Ca” buffer system (see “Materials and Methods”).

ent, PS/DG-dependent histone phosphorylation (Fig. 10A). However, oleoyl-CoA over the same concentration range caused a significant dose-dependent inhibition of nPKC ac- tivity (Fig. 10A). Activity of nPKC was inhibited 21.0% by 1 PM oleoyl-CoA and 59.2% by 10 PM oleoyl-CoA. In contrast to the significant inhibition of nPKC by 10 PM oleoyl-CoA, no inhibition of @-PKC by 10 PM oleoyl-CoA was observed (results not shown).

The ability of fatty acyl-CoA thioesters of different chain lengths and degrees of unsaturation to inhibit nPKC activity was next assessed. Free coenzyme A and decanoyl-CoA (C100) at a dose of 10 PM did not significantly inhibit nPKC activity (Fig. 10B). Increasing chain length of the saturated fatty acyl-CoA derivatives, lauryl-CoA (C12:0), myristoyl-

Cofactor and Substrate Specificities for Phospholipid-dependent Protein Kinases in NeutrophiLq 9291

97KDa "+ .L BOKDa - +

SBKOa

54KDak: 8

4 200KDa

4 r r 6 K O a

4 9 7 K D a

v 4 6 6 K D a * BOKDa - + 66KDa -+, %Koa-+

+ -

+ ZOOKDa

4 116KDa 4 9 7 K D a

L.

@ 4 6 6 K D a

DPKC- + - + - + - + nPKC - + - + Ca2" - + + - - + +

FIG. 9. Phosphorylation of endogenous substrates by puri- fied Ca'+-independent, PS/DG-dependent nPKC. Cytosol or particulate fractions prepared from cytoplasts were incubated with MgC12, PS, DG, and [-y-:"P]ATP, in the presence (+) and in the ahsence (-) of Ca"', purified 8-PKC (5 pg of protein), or purified nPKC ( 5 yg of protein) as indicated (see "Materials and Methods"). Proteins were precipitated by addition of trichloroacetic acid, solu- bilized in Laemmli buffer, and subjected to SDS-PAGE followed by autoradiography (representative experiment of 3). Molecular weight markers are indicated on the right margin and bands phosphorylated by 8-PKC and nPKC on the left margin.

CoA (C14:0), and palmitoyl-CoA (C160), caused a progressive increase in the inhibition of nPKC (Fig. 10B). Inhibition of nPKC activity by oleoyl-CoA (C181) was approximately equivalent to that of palmitoyl-CoA (C16:0), whereas arachi- donyl-CoA (C20:4) was a relatively poor inhibitor of nPKC (Fig. 10B).

Fatty acyl-CoA esters also inhibited the activity of nPKC purified by hydroxyapatite chromatography (Fig. lOC). Oleoyl-CoA and palmitoyl-CoA a t 1 p~ inhibited nPKC ac- tivity to 53 and 44% of control, respectively; a t a dose of 10 p~ oleoyl-CoA caused 100% inhibition of nPKC while pal- mitoyl-CoA caused an inhibition of 76% (Fig. lOC).

Effect of fMet-Leu-Phe on Levels of Long Chain Fatty Acyl- CoA-Since long chain fatty acyl CoAs were effective in regulating nPKC activity, it was important to determine if ligands such as Met-Leu-Phe could trigger changes in cell- associated levels of fatty acyl-CoA. Levels of long chain and short chain fatty acyl-CoA esters were measured enzymati- cally. The level of long chain fatty acyl-CoA (C-10 and greater) in resting neutrophils was 5.4 k 1.7 p M ( n = 3), a concentration that was within the active range for inhibition of nPKC activity. Neutrophils were stimulated with 10" M met-Leu-Phe, a dose that triggers responses such as degran- ulation, generation of superoxide anion, and cell-cell aggre- gation. Addition of 10" M Met-Leu-Phe triggered a prompt increase in long chain fatty acyl-CoA (C-10 and higher). The level of long chain fatty acyl-CoA was elevated from 6.8 k 0.3 pmol of fatty acyl-CoA/106 neutrophils in resting cells to 12.5 -+ 1.5 pmol of fatty acyl-CoA/lO" (i.e. 9.7 pM) neutrophils by 5 s after addition of the stimulus (Fig. 11A). This prompt increase in the level of long chain fatty acyl-CoA was followed by a return toward resting levels by 120 s (Fig. 11A). Soluble CoA levels, i.e. free CoASH plus fatty acyl-CoA of chain length C-8 and less, showed a reciprocal decline after addition of met-Leu-Phe followed by a recovery to resting values. Levels of soluble CoA declined to 15.1 f 3.1 pmol/lO" neutro- phils by 5 s, followed by a recovery to resting levels by 120 s (Fig. 11B). Total levels of CoA remained constant at 27.3 * 1.6 ( n = 3) pmol/lO" neutrophils throughout the stimulation and recovery process. Thus, elevation of absolute levels of long chain fatty acyl-CoA was an early and transient event in neutrophils activated by Met-Leu-Phe.

0 5 10 [Inhibitor] &4

Control I I CoA

Decanoyl CoA

0

Oleoyl CoA Palmitoyl CoA

FIG. 10. Effect of fatty acyl-coenzyme A and free fatty acids on nPKC activity. A, effect of oleoyl-coenzyme A and oleic acid on crude nPKC activity. Cytosol prepared in the presence of Cay+ was used as a source of crude nPKC. The effect of 0, 1, 5 , and 10 y~ oleoyl-CoA and oleic acid on histone phosphorylation by nPKC was determined in the presence of PS, DG, MgCI?, histone 111s and [-y- '"PIATP (see "Materials and Methods"). Data are expressed as % inhibition where the control value was 90.2 & 10.1 pmol/min/lO'cells (n = 4). R, effect of acyl chain length on the inhibition of nPKC activity by fatty acyl-CoA esters. The histone phosphorylating activ- ity of crude nPKC was determined in the presence of 10 PM free coenzyme A or of different fatty acyl-CoA thioesters. Data are ex- pressed as % inhibition where the control value was 102.2 & 9.6 pmol/ min/l0' cells (n = 4). C, effect of oleoyl-CoA and palmitoyl-CoA on the histone phosphorylating activity of purified nPKC. The nPKC was purified by hydroxyapatite chromatography and assayed in the presence of 0, 1, or 10 PM fatty acyl-CoA, PS, DG, MgCI,, [-y-'"PI ATP, and histone 111s. Data shown are the mean of four experiments and are expressed as % inhibition, where the control activity of the purified nPKC was 262 pmol/min/mg protein.

DISCUSSION

We report here that neutrophils possess two distinct phos- pholipid-dependent kinases, one a classical Ca/PS/DG-de- pendent 0-PKC and the other, termed nPKC, Ca-independent but PS/DG-dependent PKC. Translocation of classical Ca/ PS/DG-dependent PKC from the cytosol to the plasma- lemma, the site of DG generation during ligand-induced cell activation, is considered an essential step for initiation of PKC-specific protein phosphorylation. PKC translocation was studied using neutrophil cytoplasts, vesicles of plasma- lemma enclosing cytosol, as a model system. PKC activity (Ca/PS/DG-dependent histone phosphorylating activity) was translocated reversibly from cytosol to cytoplast membranes, in the presence of elevated Ca2+ (500 nM) and irreversibly after PMA pretreatment, in agreement with previous findings

9292 Cofactor and Substrate Specificities for Phospholipid-dependent Protein Kinases in Neutrophils

A 1 5 1 I 2 5 ,

seconds seconds

Long chain CoA Soluble CoA

FIG. 11. Time course of changes in long chain fatty acyl- coenzyme A and soluble coenzyme A in neutrophils stimulated by Met-Leu-Phe. Changes in levels of long chain fatty acyl-CoA esters and soluble coenzyme A were measured by an enzymatic method (see “Materials and Methods”) in neutrophils stimulated for varying times with M fMet-Leu-Phe. Long chain fatty acyl- coenzyme A esters are defined as the acid-insoluble fraction having chain lengths of C-10 and greater. The soluble coenzyme A pool consisted of free coenzyme A plus esters of C-8 and less. Data are expressed as pmol of fatty acyl-CoA/106 cells (mean * S.E., n = 3). A, time course of changes in long chain fatty acyl coenzyme A. B, time course of changes in soluble coenzyme A.

in intact cells (4, 6, 7, 44-46). PKC activity lost from the cytosol in response to elevated Ca2+ or to PMA pretreatment was recovered in the membrane fraction of cytoplasts. Such complete recovery of activity in the membrane fraction was not demonstrable in intact cells (7) possibly due to an inhib- itor present in specific granules (47). Physical translocation of PKC was demonstrated as reactivity with an antibody to a consensus peptide sequence of PKC (7). An 80-kDa immu- noreactive protein was translocated from cytosol to membrane in response to elevated Ca2+ or to PMA pretreatment. No loss of total 80 kDa immunoreactivity was observed in the pres- ence of elevated Ca2+ or subsequent to treatment with PMA, indicating that proteolytic cleavage of PKC was not an inte- gral part of the translocation process (5). It has been proposed that translocation of PKC to the membrane initiates proteo- lytic processing of PKC to the cofactor-independent PKM (5). However, the present study demonstrates that in the short term, i.e. after 5-10 min, neither elevated Ca2+ nor PMA treatment results in degradation/loss of PKC measured either as reactivity to the consensus antibody or as histone phos- phorylating activity. These findings also imply that conver- sion of PKC to PKM is not an essential step in the activation sequence since cytoplasts can generate 0; during the same time period in response to ligands such as Met-Leu-Phe and in response to PMA.

The P-PKC was the common isotype translocated in re- sponse to both Ca2+ and PMA, as shown by reactivity to an antipeptide antibody to p-PKC and to a monoclonal antibody to rat brain p-PKC. Only small amounts of reactivity to a monoclonal antibody to a-PKC was observed, and no 80-kDa band reactive to either the antipeptide antibody to y-PKC or to a monoclonal antibody to y-PKC was observed in cytosol or membrane fraction. Purification of translocatable p-PKC was carried out by sequential translocation of PKC to the membrane in the presence of 500 nM Ca2+ and elution of the translocated PKC from the membrane with EGTA, followed by DEAE, and hydroxyapatite chromatography. A Ca/PS/ DG-dependent histone-phosphorylating activity eluted at the same phosphate concentration as rat brain P-PKC, ran on SDS-PAGE as an 80-kDa band, and was reactive to a consen- sus PKC antibody and to a monoclonal antibody to P-PKC but not to antibodies to a- or y-PKC, confirming its identity as P-PKC.

Translocation of classical Ca/PS/DG-dependent P-PKC to the membrane elicited by elevated Ca2+ revealed the presence of a novel PS/DG-dependent, but Ca2+-independent, kinase activity in the cytosol which we have termed nPKC. Purifi- cation by column chromatography demonstrated that nPKC eluted between a-PKC and P-PKC from hydroxyapatite. Pu- rified nPKC ran as a 76-kDa band on SDS-PAGE and was not reactive to monoclonal antibodies to rat brain a-, p-, or y-PKC, but was immunoreactive to a polyclonal PKC consen- sus antibody. The requirement of nPKC for both DG and PS for optimal activity, in the same concentration range as for classical Ca2+-dependent PKC suggests that nPKC may be- long to the same family of enzymes. Sequence information will be required to substantiate this hypothesis. Ca2+-inde- pendent isoforms of PKC have been demonstrated in other cell types (7-10). The 6-, t-, and 1-isoforms of PKC are missing the C2 conserved domain, presumably the Ca2+-binding site, and do not require Ca2+ for optimal activity. A {-PKC has been demonstrated by immunochemical means in bovine neu- trophils, however, its molecular weight is greater than 80 kDa (20). nPKC may be one of these isotypes, or may be yet another member of the PKC family of enzymes (1, 13-20).

In addition to differences in localization and cofactor spec- ificity, the substrate specificity of the PKC isoforms present in a given cell will determine functional responses. Purified P-PKC phosphorylated numerous endogenous cytosolic pro- teins in a Ca/PS/DG-dependent manner. A major 47-kDa band was phosphorylated by p-PKC together with bands of 45, 54, 58, 80, 97, and 210 kDa. The 47-kDa band was of particular interest for neutrophil activation since a 47-kDa protein has been implicated in the assembly of an active NADPH oxidase (48-52). The phosphorylated 80-kDa band may represent PKC itself, since PKC can be autophosphoryl- ated (53), while phosphorylation of the 97-kDa @-subunit of the leukocyte integrin has been reported in neutrophils (54). Purified nPKC phosphorylated endogenous proteins in a pat- tern that was substantially different to P-PKC. Cytosolic proteins of 45, 58, 66, and 80 kDa were phosphorylated by purified nPKC in a Caz+-independent, phospholipid-depend- ent manner. In contrast, no net phosphorylation of membrane proteins was observed. Phosphorylation of a 66-68-kDa pro- tein has been observed in intact neutrophils stimulated with Met-Leu-Phe or with PMA (55). The identity of the 66-68- kDa band has not been established. Thus, the substrate specificity of nPKC was substantially different from that for p-PKC; notably nPKC phosphorylated a 66-68-kDa cytosolic protein but not a 47-kDa protein, whereas P-PKC phosphoryl- ated a 47-kDa cytosolic protein but not a 66-68-kDa protein. In intact neutrophils activated by Met-Leu-Phe, phosphoryl- ation of a 47-kDa band is an early event evident by 5 s, whereas phosphorylation of a 68-kDa band is a late event only evident 60 s after stimulation by Met-Leu-Phe (55). Thus, P-PKC may phosphorylate a substrate early in the activation sequence whereas nPKC phosphorylates a protein late in the sequence. Such kinetic differences in phosphoryl- ation by nPKC and P-PKC imply that further cofactor(s) in addition to Ca2+ and DG would be required to differentially regulate these kinases.

Fatty acyl-CoA thioesters were shown to inhibit nPKC activity at physiologically relevant concentrations, suggesting that levels of fatty acyl-CoA could act as a further cofactor for regulation of nPKC. Long chain fatty acyl-CoA in the concentration range of 1-10 p~ inhibited nPKC activity while no inhibition of P-PKC was observed. The inhibitory activity of long chain fatty acyl-CoA was enhanced with increasing chain length of the fatty acyl moiety; decanoyl-CoA (C1O:O)

Cofactor and Substrate Specificities for Phospholipid-dependent Protein Kinases in Neutrophils 9293

was inactive while increasing activity was obtained up to palmitoyl-CoA (C16:O). The cis-unsaturated oleoyl-CoA (C18:l) was equipotent with palmitoyl-CoA as an inhibitor, however, arachidonyl-CoA (C20:4) was less active. Long chain fatty acyl-CoA has been shown to inhibit PKC in neutrophils (25,26) and to inhibit platelet functions (56). Fatty acyl-CoA esters have also been shown to stimulate PKC activity in other cell systems (57, 58). We have found that cellular levels of fatty acyl-CoA were approximately 5 pM in neutrophils in comparison to 1 p~ in liver (59, 60), levels that could serve to regulate nPKC. Cis-unsaturated free fatty acids have also been implicated as activators of PKC, however, these fatty acids act a t levels of 10-300 p ~ , a nonphysiologic concentra- tion range that is above the critical micelle concentration (23, 61-64). Indeed, no effect of 10 p~ oleic acid was observed on neutrophil nPKC activity, demonstrating that the effect of the CoA esters is specific.

Activation of neutrophils with M Met-Leu-Phe trig- gered a prompt increase in absolute levels of long chain fatty acyl-CoA from 5 to approximately 10 p ~ , followed by a decline to base-line levels. Such ligand-elicited changes in levels of fatty acyl-CoA suggest that acyl-CoA could act as a cofactor to regulate nPKC during the activation sequence. A role for activation of phospholipase Az and release of fatty acids has been proposed by numerous investigators in neutrophil acti- vation (23, 65-69). However, under physiological conditions free fatty acids do not accumulate and are removed by con- version to the CoA derivatives and reincorporation into phos- pholipids and triglycerides. Previous studies have focused on the release of arachidonate, which can be further metabolized to the inflammatory mediators leukotriene Bq and lipoxins. We now propose that fatty acids in addition to arachidonate, when converted to the CoA ester, can play a dynamic role in regulating cell function.

Two different diglyceride-dependent protein kinases, nPKC and P-PKC, have been demonstrated in neutrophils. One is Ca2+ dependent and the other is Ca2+ independent but inhib- ited by long chain fatty acyl-CoA. Differential roles for these kinases in neutrophil activation is proposed based on differ- ences in cofactor requirements, kinetics of cofactor genera- tion, and differences in substrate specificity.

REFERENCES 1. Nishizuka, Y. (1988) Nature 334, 662-665 2. Kuo, J. F., and Turner, R. S. (1985) Adu. Znflamm. Res. 10, 63-

71 3. Takai, Y., Kaihuchi, K., Tsuda, T., and Hoshijima, M. (1985) J.

Cell. Biochem. 29, 143-155 4. Wolfson, M., McPhail, L. C., Nasrallah, V. N., and Snyderman,

R. (1985) J. Zmmunol. 135, 2057-2062 5. Melloni, E., Pontremoli, S., Michetti, M., Sacco, O., Sparatore,

B., Salamino, F., and Horecker, B. L. (1986) J. Biol. Chem.

6. Gennaro, R., Florio, C., and Romeo, D. (1985) FEES Lett. 180,

7. Phillips, W. A., Fujiki, T., Rossi, M. W., Korchak, H. M., and Johnston, R. B., Jr . (1989) J. Biol. Chem. 264, 8361-8365

8. Tauber, A. I. (1987) Blood 69,711-720 9. Fujita, I., Irita, K., Takeshige, K., and Minakami, S. (1984)

Biochem. Biophys. Res. Commun. 120, 318-324 10. COX, C. C., Dougherty, R. W., Ganong, B. R., Bell, R. M., Niedel,

J. W., and Snyderman, R. (1986) J. Immunol. 135,4611-4616 11. Della Bianca, V., Grzeskowiak, M., Casatella, M., Zeni, L., and

Rossi, F. (1986) Biochem. Biophys. Res. Commun. 135, 556- 565

12. Kikkawa, U., Kishimoto, A., and Nishizuka, Y. (1989) Annu. Reu. Biochem. 58, 31-44

13. Parker, P. J., Coussens, L., Totty, N., Rhee, L., Young, S., Chen, E., Stahel, S., Waterfield, M. D., and Ullrich, A. (1986) Science

261,4101-4105

185-190

233,853-859

14. Coussens, L., Parker, P. J., Rhee, L., Yang-Feng, T. L., Chen, E., Waterfield, M. D., Francek, U., and Ullrich, A. (1986) Science

15. Marais, R. M., and Parker, P. J. (1989) Eur. J. Biochem. 182,

16. Schaap, D., and Parker, P. J. (1990) J . Biol. Chem. 265, 7301-

17. Akita, Y., Ohno, S., Konno, Y., Yano, A,, and Suzuki, K. (1990)

18. Ono, Y., Fujii, T., Ogita, K., Kikkawa, U., Igarashi, K., and Nishizuka, Y. (1988) J. Biol. Chem. 263,6927-6932

19. Ohno, Y., Akita, Y., Konno, Y., Imajoh, S., and Suzuki, K. (1988) Cell 53, 731-741

20. Stasia, M. J., Strulovici, B., Daniel-Issakani, S., Pelosin, J. M., Dianoux. A. C., Chambaz, E., and Vignais, P. V. (1990) FEES

233,859-866

129-137

7307

J. Biol. Chem. 265, 354-362

Lett. 274,61-64 21. Honevcutt. P. J.. and Niedel. J. E. (1986) J. Biol. Chem. 261,

159’60-15905 ’

22. Korchak, H. M., Vosshall, L. B., Haines, K. A., Wilkenfeld, C., Lundquist, K. F., and Weissmann, G. (1988) J. Biol. Chem. 263, 11098-11105

23. McPhail, L. E., Clayton, C. C., and Snyderman, R. (1984) Science

24. Hansson, A., Serhan, C. N., Ingelman-Sundberg, M., and Samu- elsson, B. (1985) Biochem. Biophys. Res. Commun. 134, 1215- 1220

224,622-624

25. Hannun, Y. A,, and Bell, R. M. (1987) Science 235, 670-674 26. Wise, B. C., and Kuo, J. F. (1983) Biochem. Phurmucol. 32,1259-

1264 27. Stasia, M.-J., Dianoux, A,-C., and Vignais, P. V. (1987) Biochem.

Biophys. Res. Commun. 147,428-436 28. Pontremoli, S., Melloni, E., Sparatore, B., Michetti, M., Sala-

mino, F., and Horecker, B. L. (1990) J. Biol. Chem. 265, 706- 712

29. Grynkiewicz, G., Poenie, M., and Tsien, R. Y . (1985) J. Biol. Chem. 260,3440-3450

30. Martell, A. E., and Smith, R. M. (1974) Critical Stability Con- stants, Vol. 2, Plenum Publishing Corp., New York

31. Boyum, A. (1968) Scand. J. Clin. Lab. Znuest. 97, Suppl. 21, 77- 89

32. Korchak, H. M., Vienne, K., Rutherford, L. E., Wilkenfeld, C., Finkelstein, M. C., and Weissmann, G. (1984) J. Biol. Chem. 259,4076-4082

33. Roos, D., Voetman, A. A,, and Meerhof, L. J. (1983) J. Cell Biol.

34. Korchak, H. M., Roos, D., Giedd, K. N., Wynkoop, E. M., Vienne, K., Rutherford, L. E., Buyon, J . P., Rich, A. M., and Weiss- mann, G. (1983) Proc. Natl. Acad. Sci. U. S. A. 80,4968-4972

35. Laemmli, U. K. (1970) Nature 227, 680-685 36. Towbin, H., Staehelin, T., and Gorden, J . (1979) Proc. Nutl. Acad.

37. Makowske, M., Ballester, R., Cayre, Y., and Rosen, 0. M. (1988)

38. Fabro, D., Jungmann, R. A., and Eppenberger, U. (1985) Arch.

39. Sekiguchi, K., Tsukuda, M., Ase, K., Kikkawa, U., and Nishizuka,

40. Gschwendt, M., Kittstein, W., Horn, F., Leibersperger, H., and

41. Williamson, J. R., and Corkey, B. E. (1969) Methods Enzymol.

42. Corkey, B. E. (1988) Methods Enzymol. 166, 55-70 43. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J.

(1951) J. Biol. Chem. 193, 265-275 44. Wolf, M., Cuatrecasas, P., and Sahyoun, N. (1985) J. Biol. Chem.

260, 15718-15722 45. May, W. S., Jr., Sahyoun, N., Wolf, M., and Cuatrecasas, P.

(1985) Nature 317, 549-551 46. Horn, W., and Karnovsky, M. L. (1986) Biochem. Biophys. Res.

Comrnun. 139, 1169-1175 47. Balazovich, K. J., Smolen, J. E., and Boxer, L. A. (1986) J.

Zmmunol. 137, 1665-1673 48. Kramer, I. M., Verhoeven, A. J., van der Bend, R. L., Weening,

R. S., and Roos, D. (1988) J. Biol. Chem. 263,2352-2357 49. Heyworth, P. G., Karnovsky, M. L., and Badwey, J . A. (1989) J.

Biol. Chem. 264,14935-14939 50. Segal, A. W., Heyworth, P. G., Cockcroft, S., and Barrowman,

M. M. (1985) Nature 316, 547-549

97, 368-377

Sci. U. S. A . 76, 4350-4354

J . Biol. Chem. 263, 3402-3410

Biochem. Biophys. 239, 102-111

Y. (1988) J. Biochem. (Tokyo) 103, 759-765

Marks, F. (1989) J. Cell. Biochem. 40, 295-307

13,434-513

9294 Cofactor and Substrate Specificities for Phospholipid-dependent Protein Kinases in Neutrophils 51. Lomax, K. J., Leto, T. L., Nunoi, H., Gallin, J. I., and Malech, 60. Woldegiorgis, G., Spennetta, T., Corkey, B. E., Williamson, J. R.,

52. Clark, R. A., Malech, H. L., Gallin, J. I., Nunoi, H., VO~PP, B. D., 61. Seifert, R., Schachtele, C., and Schultz, G. (1987) Biochem. Bio-

62. Verkest, V., MacArthur, M., and Hamilton, S. (1988) Biochem.

63. Wooten, M. W., and Wrenn, R. W. (1988) Biochem. Biophys. Res.

H. L. (1989) Science 245,409-412 and Shrago, E. (1985) Anal. Biochem. 150,8-12

Pearrson, D. W., Nauseef, W. M., and Curnutte, J. T. (1989) N . Engl. J. Med. 321, 647-652

phys. Res. Commun. 149 , 762-768

53. Huang, K.-P., Nakabayashi, H., and Huang, F. L. (1986) Proc. Natl. Acad. Sci. U. S. A. 8 3 , 8535-8539

Biophys. Res. Commun. 152,825-829

54. Buyon, J. P., Slade, S. G., Reibman, J., Abramson, S. B., Philips, Commun. 153,67-73

(1988) Mol. Endocrinol. 2 , 1043-1048 M. R.9 Weissmann, G.3 and Winchester, R. (1990) J. hW7Wnol. 64. Naor, Z., Shearman, M. S,, Kishimoto, A., and Nishizuka, y. 144,191-197

55' R ~ ~ ~ ~ . ~ . K ~ ~ h ~ ~ i ~ ~ ~ ~ , v ~ s ~ ~ ~ 8 ~ ~ , ~ i ~ ~ ~ ~ , K ~ ~ ~ 65. Cockcroft, S., and Stutchfield, J. (1989) Biochem. J. 263 , 715- "0"

6322-6328 56. Lascu, I., Edwards, B., Cucuianu, M. P., and Deamer, D. W.

(1988) Biochem. Biophys. Res. Commun. 156, 1020-1025 57. Bronfman, M., Morales, M. N., and Orellana, A. (1988) Biochem.

Biophys. Res. Commun. 152,987-992 58. Shoyab, M. (1985) Arch. Biochem. Biophys. 236,435-440 59. Corkey, B. E. (1990) in Fatty Acid Oxidation: Clinical, Biochemical

and Molecular Aspects, pp. 217-232, Alan R. Liss, Inc.. New

66. Badwey, J. A., Curnutte, J. T., and Karnovsky, M. L. (1981) J.

67. Bauldry, S. A,, Wykle, R. L., and Basu, D. A. (1988) J. Biol.

68. Walsh, C. E., DeChatelet, L. R., Chilton, F. H., Wykle, R. L., and

69. Wynkoop, E. M., Broekman, M. J., Korchak, H. M., Marcus, A.

I LJ

Biol. Chem. 256,12640-12643

Chem. 263,16787-16795

Waite, M. (1983) Biochim. Biophys. Acta 750, 34-40 ~ . "

York J., and Weissmann, G . (1986) Biochem. J. 236,829-837