Oleic Acid Promotes Changes in the Subcellular Distribution of Protein ...

THEJOURNAL OF BIOLOGICAL CHEMISTRY

Vol. 266, No. 35, Issue of December 15, pp. 23568-23576.1991

0 1991 by The American Society for Biochemistry and Molecular Biology, Inc

Printed in U.S.A.

Oleic Acid Promotes Changesin the Subcellular Distributionof Protein Kinase C in Isolated Hepatocytes* (Received for publication, June 14, 1991)

Maria J. M. Diaz-Guerra, Maria Junco, and Lisardo BoscaS From the Instituto deBioquimica, Facultad de Farmacia, Universidad Complutense, 28040-Madrid, Spain

The effect of oleate on the subcellular distribution of protein kinase C (PKC) was studied in isolated hepatocytes and in perfused rat liver in the presence of physiological concentrations of serum albumin. A timeand dose-dependent translocation of PKC from the cytosol towards the membranes was observed at oleate concentrations that fell withinthe range of concentrations reached under several physiological conditions. Analysis of the membrane-bound isoenzymes of PKC by hydroxylapatite chromatography revealed that the @ isoenzyme was preferentially translocated to this compartment in hepatocytes incubated with oleate. Activation of PKC after incubation of hepatocytes with oleate involved at least three different effectors of the enzyme: the fatty acid itself, the diacylglycerol synthesized from oleate, and the rise in the cytosolic calcium concentration elicited by oleate. As a result of PKC activation, protein phosphorylation of intact hepatocytes in response to oleate exhibited an enhancement in the phosphate content of a protein of 82 kDa, similar to that phosphorylated in the presence of phorbol dibutyrate.

pathway of PKC activation is mediated by the rise in the concentrations of both diacylglycerol and Ca2+ resulting from phosphoinositide breakdown. In addition, DAG coming from other glycerolipids may also supportPKCactivity (7-9). However, less attention has been paid to the activation of PKC by phosphoinositide-independentmechanisms, among which is the activation of PKC by unsaturated fatty acids (10-12). This activationof PKC, althoughsomewhat dependent on the nature of the fattyacid, has been described forthe enzyme isolated from various animals and tissues. Stimulation of PKC by fatty acids requires the presence of phosphatidylserineandCa2+,althoughCa2+-independentactivation of PKC by oleatehas also beenreported (10). The specific responses to oleateof PKC isoenzymes isolated by hydroxylapatite chromatography have beendescribed (11). Recently it has been shown that platelet PKC is activatedby oleate by a mechanismthatisdistinct fromphosphatidylserine/DAG (12). Fatty acids may be available to the cell either from the extracellular environment or from the action of phospholipases that cleave the fatty acid moieties of the glycerolipids (13, 14). Concerning liver, it iswell known that oleate metabolism is involved in glycerolipid synthesis and diacylglycerol is one of its reaction products. Therefore, as a result of the activation by fatty acidsof the enzymes that control glyceroPhospholipid and calcium-dependent protein kinase (PKC)’ is involved in theregulation of physiological responses lipid metabolism (15, 16), an increased synthesis of diacylt o different stimuli, ranging fromcellular proliferation to the glycerol at theendoplasmic reticulum level has been reported may activate hepatic control of neural activity(1-3). Current interestregarding the (17). Thus, oleate and related fatty acids of specific isoenzymes mechanism of PKC activation has concentrated mostly on its PKC both through the direct activation sensitive to this metabolite possibly and through the increased capacity to be translocated from the cytosol to the plasma membrane (4, 5 ) . This subcellular redistribution of the en- diacylglycerol synthesis triggered by fatty acids. It is shown in this study that oleate, in the presence of zyme is aconsequence of theincreasedphosphoinositide turnover elicited by the occupancy of receptors from trans- physiological concentrations of BSA, induced the translocamembrane signaling pathways in which activation of phos- tion of PKC from the cytosol towards the membranes. In phatidylinositol-specific phospholipase C is involved (6). This addition, the capacity of the diacylglycerols present in the hepatocyte plasma membranes functioning as activators of *This work was supported by Grant PM88-0025 from CICYT, PKC is examined. A search for possible substrates of phosSpain. The costs of publication of this article were defrayed in part phorylation of PKC in response to oleate was made both in by the paymentof page charges.This article must therefore hereby he vivo and ina cell-free system.It is alsoshown that fattyacids marked “aduertisement” in accordance with 18 U.S.C. Section 1734 produce small changes in the intracellular calcium concentrasolely to indicate this fact. $ To whom correspondence should be addressed Instituto de Bio- tion that may participate in the activationof PKC either by glycerolipids. quimica, Facultad de Farmacia, 28040-Madrid, Spain. Fax: 341-549- the fattyacid itself or through its corresponding

46116. The abbreviations used are: PKC, protein kinase C; BSA, bovine serum albumin; PKM, proteolysis-generated catalytic fragment of PKC; SDS-PAGE,sodium dodecyl sulfate-polyacrylamide gel electrophoresis; TPA, 12-tetradecanoate phorbol 13-acetate; Bodipy-TPA, 12-(5,7-dimethyl Bodipy’”-1-dodecanoyl) phorbol13-acetate; DAG, diacylglycerol; InsP,,inositol 3,4,5-trisphosphate;PDBu,phorbol 12.13-dibutyrate; EGTA, [ethylenebis(oxyethylenenitrilo)]tetraacetic acid; PIPES, 1,4-piperazinediethanesulfonic acid;Hepes,4-(2-hydroxyethy1)-1-piperazineethanesulfonic acid. The fatty acid abbreviations used are: elaidic, trans-9-octadecenoic; oleic, cis-9-octadecenoic; arachidonic, cis-5,8,11,14-eicosatetraenoic;palmitic, hexadecanoic.

EXPERIMENTALPROCEDURES

Chemicals [y3’P]ATP (3000 Ci/mol), [32P]phosphate, andradioactive phorbols were from New England Nuclear. Histone H1, phosphatidylserine, dioctanoylglycerol, phorbol dibutyrate, fattyacids, and ATPwere from Sigma. Chemicals for electrophoresiswere from Bio-Rad. Other chemicals were from Merck or Boehringer. Fatty acid-freebovine albumin (fraction V) was from Boehringer and contained less than 0.1% of fatty acids in a molar ratio. Monoclonal anti-PKC antibody (clone MC5) and Inspa assay system were from Amersham. Mono-

23568

Oleate-induced Redistribution of PKC in Hepatocytes clonal antibodies to PKC isoenzymes (clones MC2a and 3a) were from Seikagaku. Escherichia coli diacylglycerol kinase was from Calbiochem. Fura 2AM and fluorescent phorbols were from Molecular Probes. Preparation of Hepatocytes-Hepatocytes were isolated by collagenase digestion of the livers from adult male Wistar rats ("200 g) fed ad libitum (18).To avoid membrane contamination, the cells were washed three times by careful centrifugation (50 X g for 2 min in a Minifuge T centrifuge, Haereus) with Krebs-Henseleit bicarbonate buffer saturated with 0z/COp (19:l). The hepatocyte suspension was filtered through a nylon membrane of 50-pm mesh and incubated in the presence of 0.5 mM CaCl, for 10 min a t 37 "C under continuous gassing. The cell viability was always higher than 90% as assessed by the trypan blue exclusion criteria. Cells were aliquoted at 0.2 g/ml (wet weight) in 2.5 ml of incubation volume and, unless stated otherwise, in the presence of 32-45 mg/ml (0.5-0.7 mM) fatty acidfree BSA and several ligands. When specifically indicated, equimolar complexes of palmitate-BSA were used prior to theaddition of oleate. Fatty acids were incubated with BSA prior to the exposure to the hepatocytes. The incubation was stopped by centrifugation at 1800 X g for 40s at 4 "C and thecell pellets frozen on liquid nitrogen. Oleate and other lipids were added to the incubation medium in half the final incubation volume of fresh Krebs-Ringer bicarbonate medium containing BSA. Fatty acids were prepared in KOH with a 15% molar excess of alkali. Other lipids were dispersed by sonication. Perfusion of Rat Liuer-Livers were perfused with Krebs-Henseleit bicarbonate buffer in a nonrecirculating system for the first 15 min after which the circuit was closed and completed to 0.2 liter of buffer containing 45 mg/ml fatty acid-free BSA, 0.5 mM CaC12,oleate, and continuous gassing with carbogen (19). The perfusion was maintained for 20 min a t a flow rate of about 30-40 mllmin, and themedium was infused through the portal vein. At appropriated times lobuli were ligated and portions of liver (0.5-0.7 g) were cut off and immediately homogenized. Preparation of Homogenates-Hepatocytes were homogenizedin 2 ml of ice-cold 0.25M sucrose, 10 mM &mercaptoethanol, 5 mM EGTA, 2 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, 10 pg/ml leupeptin, and 20 mM PIPES, pH 6.6 (buffer A). For perfused liver, the tissue was homogenized in four volumes of buffer A. After centrifugation at 1800 X g for 10 min the supernatant wasremoved and centrifuged at 40,000 X g for 15 min. The supernatant, containing the microsomes and the cytosol, was centrifuged at 105,000 X g for 30 min to yield a pellet (microsomes) and a soluble fraction (cytosol). The initial 1800 X g pellet was fractionated in buffer A following the procedure of Fleisher and Kervina (ZO), and the plasma membranes were purified by centrifugation in a two-layered step gradient of sucrose (1.6/1.45 M). The same protocol wasfollowed for the purification of nuclei and mitochondria. To extract the PKC activity from the particulate fractions, the pellets containing the plasma membrane, nuclei, mitochondria, andthe endoplasmic reticulum membranes were homogenized in one original volume of buffer A supplemented with 0.1% Nonidet P-40 and incubated at 4 'C for 30 min. Cytosolic and particulatefractions were partially purified through a DE52 column (0.5 X 2 cm) prior to the measurement of PKC activity. These conditions provided maximal recovery of PKC from particulate fractions. The distribution among these fractions of lactate dehydrogenase, 5"nucleotidase and rotenone-insensitive NADH-cytochrome c reductase was used to estimate the purity of the preparations of cytosol, plasma membrane, and endoplasmic reticulum, respectively (21). Purification of Rat Liuer PKC and PKM-Rat liver PKC was purified following the procedure of Woodgett and Hunter (22) developed for the purification of the brain enzyme. The specific activity was 2007 units/mg of protein. The purity of this preparation was estimated by SDS-PAGE (23) to be higher than 90%. The enzyme was very sensitive to the activation by Caz+and phospholipids, and the resolution of the isoenzymatic subspecies by hydroxylapatite chromatography showed only the presence of p and a isoenzymes (24). PKM, the kinase produced after proteolysis of PKC and containing the catalytic domain of the protein, was purified by DE52 chromatography after washing the column with buffer A supplemented with 150 mM NaCl. The enzyme eluted at 280 mM NaCl was independent of Ca2' and lipids and was characterized by its capacity to phosphorylate histone H1 and the inhibition by H7, but not by staurosporine, two characteristic inhibitors of PKC (25, 26). This kinase was recognizedby clones MC 2a and 3a of anti-PKC antibodies (Seikagaku) but not by clone MC5 antibodies (Amersham Corp.) raised against a peptide contained in the regulatory domain of PKC.

23569

Isolation of Isoenzymes from Particulate Fractions-The isoenzymatic pattern of the plasma membrane-bound PKC was characterized after solubilization of the paticulate enzyme by using 0.1% Nonidet P-40 in buffer A, followed by chromatography on DE52 (1 X 7 cm). The column was washed with 20 mM NaCl, 10 mM 0-mercaptoethanol, 0.5 mM EGTA, 0.2 mM EDTA, 1mM phenylmethylsulfonyl fluoride, 10 mM Hepes, pH 7.4 (buffer B), and the enzyme was then eluted with buffer B supplemented with 120 mM NaC1. Fractions containing activity were pooled and applied to a hydroxylapatite column (0.7 X 12 cm) and the isoenzymes resolved by using a potassium phosphate gradient (from 30 to 300 mM, pH 7.4) in buffer B supplemented with 5% glycerol in a total volume of 200 ml (24). Western Blot Analysis of PKC-Hepatocytes incubated with several fatty acid/BSA ratios were homogenized with 1.2 ml of buffer A. After centrifugation at 105,000 X g for 30 min, the enzyme from the supernatant and that extracted from the membranes were partially purified by DE52 chromatography (120 mM NaCl elution) and concentrated by ultrafiltration through a 20-kDa cut-off membrane (Lida Corp. Kenosha, WI). Proteins of these concentrated samples were separated by SDS-PAGE (10% acrylamide gel). After electrophoresis, the proteins were transferred (0.5 mA/cm2; 15 h) to nitrocellulose membranes and processed by Western blot analysis as recommended by the supplier of the PKC antibody (clones MC5 from Amersham and clones MC 2a and 3a from Seikagaku). Phorbol Diester Binding to PKC-In vivo binding of phorbol diesters to PKC was assessed following the fluorescence of Bodipy-TPA on a FACScan flow cytometer (Becton and Dickinson) equipped with a 15-milliwatt argon laser (488 nm). Hepatocytes (1 X IO6) were incubated with BSA-fatty acid complexes for 10 min at 37 'C as previously described and were immediately centrifuged at 100 X g for 2 min. After resuspension of the cells with 250 p1 of ice-cold PBS containing several concentrations of Bodipy-TPA, the incubation was continued in the dark a t 4 "C for 30 min, after which the fluorescence emission at 530nmwas collected. Results are expressed asthe difference between the fluorescence in the presence of Bodipy-TPA and the fluorescence of cells incubated with 2 p M TPA for 5 min prior to the addition of the fluorescent probe (unspecific binding). The [3H]PDBu binding to hydroxylapatite-purified PKC from perfused rat liver was measured as previously described (27, 28) as criteria to quantify the ability of oleate to translocate the a and p isoenzymes of PKC, in addition to the measurement of the enzyme activity. Basically, the eluate from the DE52 column containing PKC was incubated for 10 min at 4 "C in the presence of 20 pg/ml leupeptin, 2 p M CaCI,, 50 nM [3H]PDBu (2 pCi/ml) and in the presence or absence of 5 p~ PDBu (2 pCi/ml), followedby hydroxylapatite chromatography in two columns run in parallel. The column contain) used to determine ing the nonradioactive PDBu excess (5 p ~ was the nonspecific binding of the radioactive phorbol to the proteins. The results are expressed as the difference between the radioactivity eluted in the absence and presence of the nonradioactive phorbol excess. Isolation of Glycerolipids-Hepatocytes were incubated in the presence or absence of fatty acids for 10 min at 37 "C in a Krebsbicarbonate buffer containing 0.7 mM fatty acid-free BSA. Atthe end of the incubation period, the cells were centrifuged and the pellets homogenized in buffer A, centrifuged for 10 min at 2000 X g, and the plasma membrane was purified by sucrose step gradients as previously described. Membrane lipids were extracted following the method of Folch et al. (29) as follows; the plasma membrane pellets were resuspended in 1 mlof water followedby the addition of 2 mlof methanol/chloroform (21, v/v), and centrifuged at 2700 X g for 2 min. Supernatants were separated, and 1 ml each of water and chloroform were added. After thorough shaking and centrifugation, the chloroform layer was removed, dried under Np atmosphere and redissolved in 50 pl of methanol/ml of original medium. Samples of 50 p1 were applied to a silica gel TLC (Kieselgel 60, Merck) and developed in chloroform/methanol (1945, v:v). The positions of standards were determined by exposure to iodine vapor. The Rfvalues for oleic acid, oleoylacetyl glycerol,dioactanoyl glycerol, and trioleine were measured. The corresponding areas of sample and standards were scraped off the plate and the lipids extracted in 1 mlof the developing medium. The diacylglycerol recoveryof the standards and samples was quantified using E. cofi diacylglycerol kinase according to Preisset al. (30).These DAG from control or oleate-exposed hepatocytes were used to test their capacity to activate the (Y and p isoenzymes from partially purified PKC from hepatocytes. Phosphorylation of Intact Hepatocytes and Cytosolic ProteinsIsolated hepatocytes were washed with phosphate-free Krebs-Hen-

23570

Oleate-induced Redistribution of PKC in Hepatocytes

seleit bicarbonate buffer and incubated in the presence of 10 mM glucose for 10 min. After centrifugation at 80 X g for 2 min, the hepatocyte pellet was resuspended a t 100 mgof cells/ml in the same buffer but in the presence of [32P]phosphate(0.3 mCi/ml). A steady state of phosphorylation of cellular proteins was obtained after 15 min of incubation as determined by measuring the amount of trichloroacetic acid-insoluble radioactivity. After addition of the ligands, samples of 500 pl of cell incubation were collected, centrifuged 20 s in anEppendorf centrifuge, and thepellets chilled on liquid N,. After rapid thawing, cells were homogenized in 100 p1of 1 mM ATP, 0.1% SDS, 25 mM Tris, 200 mM glycine, 30 mM NaF, pH 8.3, followed by centrifugation for 10 min at 12,000 X g. The supernatants were submitted to SDS-PAGE by using 10% (w/v) acrylamide (23). The gel was stained with Coomassie Blue, dried, and exposed to Agfa xray film for autoradiography. Oleate-dependent phosphorylation of cytosolic proteins from hepatocytes was determined after incubation of a Sephadex G-25filtered homogenate from control hepatocytes, with purified rat liver PKC, [y-32P]ATP and several ligands in amedium containing 30 mM KF, 5 mM 8-mercaptoethanol, 0.1 mM EGTA, and 20mM Hepes, pH 7.4. The incubates were analyzed by SDS-PAGE and autoradiography as previously described. Assay of PKC and PKM-PKC was followed by its histone kinase activity, measured at 30 "C in an incubation volume of 150 pl as previously described (28). In final concentrations, the assay mixture contained 25 p~ ATP (0.4 pci), 10 mM magnesium acetate, 5 mM 8mercaptoethanol, 50 pg of histone H1,50 pl of sample, 20 mM Hepes, pH 7.5, and, unless otherwise indicated, 0.6 mM CaCl,,10pgof phosphatidylserine, and 2 pg ofdioctanoyl glycerol. The incorporation of [3ZP]phosphate intohistone was linear for at least 20 min. The reaction was stopped by the addition of 2 ml of ice-cold 5% trichloroacetic acid, 10 mM &PO,. The radioactivity retained on GF/C glass fiber filters after filtration was determined by counting the oven-dried filters in 2 ml of scintillation fluid. PKC activity was determined after subtractingthe incorporation in theabsence of Caz+ and phospholipids. PKM was measured as described for PKC in the absence of Ca2+and lipids. One unit of PKC or PKM was defined as incorporating 1nmol of phosphate/min into histone H1. Endogenous phosphorylation was negligible. Mixed Micelle Assay of PKC-The effect of endogenous glycerolipids on PKC activity was assayed in Triton X-100 mixed micelles containing 0-5% endogenous diacylglycerol and 0-15% of phosphatidylserine in amolar ratio (31,32).Micelles wereprepared as follows. Lipids were dissolved in chloroform and dried under N, atmosphere. The lipids were solubilized in 3%Triton X-100 followed byvortexing (four periods of 2 min) and incubation at 37 "C for 20 min with continuous shaking. Aliquots of these micelle preparations, instead of the sonicated lipids, were added to the PKC assay previously described. The final concentration of Triton X-100 in the assay was 0.1% (w:v). Assay of Glycogen Phosphorylase-Samples for the determination of glycogen phosphorylase a activity were collected after centrifugation of the hepatocyte suspension in an Eppendorf centrifuge (15 s), and the pellets were immediately chilled on liquid N,. The activity was measured following the method of Stalmans and Hers (33) by measuring the release of phosphate from glucose 1-phosphate (50 mM) in the presence of 100 mM NaF, 1%glycogen, and 0.5 mM caffein a t pH 6.1. Determination of InsP3-InsP3 concentration was determined by using a specific binding protein for the inositol 3,4,5-trisphosphate isomer, following the recommendations of the supplier. Suitable internal standards were used to determine the accuracy of the assay. Determination of the Intracellular Ca2' Concentration-To measure the intracellular calcium concentration, the hepatocytes were loaded with fura 2AM (final concentration 5 p ~ at) 37 "C under continuous shaking (60 cycles/min) for 10 min. After washing and centrifugation (50 X g for 2 min) twice in Krebs-Ringer buffer, the hepatocytes were transferred at 20-40 mg/ml to phosphate-buffered saline containing fatty acid-free BSA adjusted to a final pH of7.4. The fluorescence was recorded at 510 nm in a L50 Perkin-Elmer spectrofluorimeter using a dual excitation source at 320 and 380 nm. The maximal fluorescence was determined a t the end of the assay by adding 10 pl of 10% SDS. The minimal fluorescence was obtained by adding 15 p1 of a solution containing 0.5 M EGTA, pH 9.0 (27). Protein was determined as described by Bradford (34), with bovine serum albumin as standard.

RESULTS

Subcellular Distribution of PKC-The distribution of PKC was analyzed in the cytosol, endoplasmic reticulum, mitochondria, nuclei, and plasma membrane fractions from hepatocyte homogenates (3-5 g of cells). The PKC activity present in each compartment was measured after extraction of the enzyme from the particulate fraction. As shown in Table I, the activity present in the plasma membrane, microsomes, and cytosol represented more than 91% of the total cellular activity. Table I also summarizes the subcellular distribution of enzyme markers associated with the different structures obtained following the separation procedure used in the present work. Effect of Fatty Acids-The ability of several fatty acids to promote aredistribution of PKC activity was studied in isolated hepatocytes in the presence of concentrations of BSA in the physiological range (0.5-0.7 mM). As shown in Table 11, cis-unsaturated fatty acids, such as oleate and arachidonate, induced a translocation of PKC from the cytosol towards the plasma membrane. However, trans-unsaturated (elaidate), saturated (palmitate), or esterified fatty acids (linoleylethyl ester) lost the ability to produce changes in the distribution of the enzyme. This is in agreement with the effects of these fatty acids on the catalytic activity of PKC (10,ll). Moreover, when hepatocytes were preincubated with 50 p~ H7 to inhibit PKC activity, oleate retained its capacity to induce a redistribution of the enzyme. However, this was not the case for staurosporine which, in agreement with previous reports (35), translocated PKC to the plasma membrane even in the abTABLEI Subcellular distribution of PKC andmarker enzymes in isolated hepatocytes Cells (3-5 g) were homogenized in buffer A and the subcellular particulate fractions were analyzed for their content in PKC, lactate dehydrogenase, 5'-nucleotidase and rotenone-insensitive cytochrome c oxidase. Results are means of duplicates of two independent preparations. LDH, lactate dehydrogenase. Enzymatic activity Fraction

protein PKC LDH

5.6 9.0 11.4 19.2 54.8

Cytochrome c reductase

units/g of cells

%

Plasma membrane Nuclei Mitochondria Microsomes Cytosol

5' Nucleotidase

30 1.1 0.4 1.2 6.1 0.4 25.6 1.0 98.6 75.7

0.3 0.2 0.3 12.8 0.05

3.2 0.02 0.09 1.17 0.04

TABLE I1 Effect of fatty acids and PKCinhibitors on the subcellular distribution of PKC in isolated hepatocytes Cells were incubated for 10 min in the presence of the indicated ligands. Fatty acids were prepared in Krebs-bicarbonate buffer containing 0.7 mM fatty acid-free BSA. Results are means S.E. of three different experiments.

+

PKC activity Addition

Cytosol

Microsomes

Plasma membrane

unitslg of cells

87f7 None 27+4 Oleate, 2 mM 1272++33 Arachidonate, 0.5 mM 2832++55 Elaidate, 2 mM 27 38 f+33 Linoleylethyl ester, 2 mM 80+9 Palmitate, 2 mM 91+8 H7,50 p M 22+3 Staurosporine, 10 p~ Oleate. 2 mM H7. 50 uM 2 5 + 4

+

25+4 21f2

22f4 27+2 21+1 22+3

31 + 4 88f6 73 8 39 4 38 + 4 40 + 6 36 5 76 + 6 96 f 10

+ +

+

Oleate-induced Redistribution

of

PKC in Hepatocytes

23571

U D a 1 2 3 4 5 6 7 8 9 1 0 sence of oleate. According to these results, the oleate-induced translocation of PKC to the plasma membrane did not require 18the expression of the catalytic activity of the enzyme, and suggests that only cis-unsaturatedfatty acids areableto translocate PKC to the membrane (35, 36). FIG.2. Immunoblot analysis of PKC distribution in hepaThe effect of oleate on thesubcellular distribution of PKC tocytes incubated with oleate. Cells were incubated with 0.5 mM was studied in isolated hepatocytes incubated with an equiBSA and the indicated additions: none (lanes 1 and 2 ) , 5 min with molar concentrationof BSA and palmitate(0.5 mM) to mimic 0.5 mM palmitate (lanes 3 and 4 ) , 5 min (lanes 5 and 6), or 15 min the physiological conditions of saturation of BSA by fatty (lanes 7-10) with 0.5 mM palmitate, 1 mM oleate. Extracellular Ca2+ acids. As shown in Fig. 1, increasing concentrationsof oleate was replaced by 50 PM EGTA in samples corresponding to lanes 9 in the BSA-palmitate-oleate complexinducedadoseand and 10. Even and odd lanes are given for cytosol and plasma membrane, respectively. time-dependent decrease in the cytosolic activity of PKC in these cells. Fig. lA shows an increase in the plasma membrane-bound enzyme that parallels the decreasein the cytoA I solic activity of PKC in hepatocytes incubated for 10 min with several concentrations of oleate. The timecourse for the translocation of PKC is shown in Fig. 1B. The half-maximal translocation to the membraneswas obtained 8-10 min after the exposure of the hepatocytes to0.6 mM oleate. The PKC present in the endoplasmic reticulum membranes exhibiteda transient but significant increase in extractable enzyme at concentrations of oleate about0.2 mM, followed by areduction in the activity present in this compartment a t higher oleate concentrations. When the same experiment was carried out with BSA lacking palmitate, the results were qualitatively the same, but the concentration of oleate required to produce the half-maximal increase in the plasma membrane fractionwas at least twice the value obtained in the presenceof palmitate (results not shown). Thus, the concentration of oleate re0 > quired to induce a given effect on PKC distribution in hepatocytes is proportional to the concentrationof free fatty acid binding-sites available in the serum albumin. As a n additional Vcriteria to follow the subcellular distribution of PKC elicited E 2 40 by oleate, the amountof enzyme present incytosol and plasma membrane was quantified by Western blot afterSDS-PAGE. As Fig. 2 shows, the staining of the bands was in agreement with theactivity distribution reported inFig. 1. It is important t o note that, in the absence of calcium in the extracellular medium, oleate failed to produce changes in the distribution of PKC (Fig. 2, lunes 9 and 10). The relevance of the fattyacid/BSA ratio on the subcellular distribution of PKC was evidenced when hepatocytes were incubated for 15 min in the absence or presence of 1.2 mM oleate, and concentrations of BSA were varied to obtained molar ratios for oleate/BSA from 10 to0.4. As Fig. 3A shows, the recovery of the enzyme in the plasma membrane fraction of hepatocytes incubated with oleate accounted for the de0

II

A

0

,,loo

II a

1.5

0.5

1

Oleate, lmMl

0

10

1

2

4

10

Olorto/BSA. molar rrtlo

FIG.3. PKC and PKM activity in hepatocytes incubated with different BSA/oleate ratios. Cells were incubated for 15 min in theabsence or presence of 1.2 mM oleate and appropriatedconcentrations of BSA. A, PKCactivity from cytosol ( 0 ) and plasma membrane (W); B, densitometric analysis after Western blot of the 52-kDa fragment recognized by PKC antibodies (clones MC 2b and 2c, Seikagaku); and C, PKM activityafterpartial purification by DE52 chromatography. Results are means 2 S.E. of three cell preparations.

20

Time. min

FIG.1. Effect of oleate on the subcellular distribution of PKC in isolated hepatocytes incubated with 0.5 mM BSA, 0.5 mM palmitate. A, dose-dependent distribution of PKC in hepatocytes incubated for 10 min with several concentrations of oleate. B, time-dependent distribution in hepatocytes incubated with 0.6 mM oleate. Cytosolic (W), plasma membrane (O),and microsomal activity of PKC (A), are expressed as percentage of the total control activity (88, 23, and 31 units/g of cells, respectively). Results are means S.E. of three different experiments.

*

creased activityin the cytosol at fattyacid/BSA ratios higher than 0.5 and was followed by the disappearanceof the activity a t oleate/BSA ratios higher than 2. The possibility of an enhanced proteolysis of the membrane boundenzyme, induced by oleate,mayexplain this unexpected fall in the recovery of the enzyme that follows after exposure of hepatocytes to 1.5 mM oleate but in the presence of decreasing


23572

concentrations of BSA. The enzyme that proteolyzes PKC has been described for other tissues (37-39), and yields two fragments of approximately 50 and 30 kDa, respectively. The 50-kDa enzyme contains the catalytic activity of PKC and is usually denominated PKM, whereas the 30-kDa fragment contains the regulatory sites for calcium, phospholipids and diacylglycerol. To test the occurrence of PKC proteolysis in the mechanism of action of oleate when used at high fatty acid/BSA molar ratios, two independent criteria were used (a) Western blot quantitation of the amount of catalytic fragment produced after proteolysis, using a monoclonal antibody (Seikagaku, clones MC 2a and 3a) that recognized a 52-kDa fragment of this molecule (Fig. 3B), and(b) measurement of the calcium and phospholipid independent PKC activity (PKM), produced after proteolysis and recognized by its elution profile from DE52 column at 280 mM NaCl (Fig. 3C). Fig. 3 ( B and C) shows Western blot analysis using PKC antibodies and PKMactivity in samples corresponding to the experiment reported in panel A . The decrease in the activity of PKC from both cytosol and membranes showed a good correlation with the increase in thestaining of a proteinof 52 kDa and in the activity of PKM. These results indicate that proteolysis is the mechanism responsible for the decrease of PKC activity in hepatocytes incubated with oleate/BSA ratios higher than 2. The use of leupeptin as inhibitor of the proteinase activity that converts PKC into PKM was not effective in preventing the degradation of the enzyme in isolated hepatocytes (40), probably due to thelow permeability of the hepatocyte plasma membrane to this low molecular weight peptide. In addition to isolated hepatocytes, the effect of oleate on the subcellular distribution of PKC was studied in perfused rat liver. In thissystem, oleate also promoted a redistribution of PKC from the cytosol to the plasma membrane. Table I11 summarizes the time-dependent distribution of PKC in livers perfused with a 0.5 mM BSA-palmitate-oleate complex (1:l:l) either in the presence or absence of 0.3 mM Ca2+ in the medium. Again, the results obtained in this experiment suggest that Ca2+is required for the translocation to be observed and indicate that the effect of oleate on PKC distribution cannot be attributedto processes related to the isolation procedure of the hepatocytes. Isoenzyme Mobilizationby Oleate-Since PKC is a multienzyme system with important regulatory differences among the TABLE I11 Subcellular distribution of PKC in perfused rat liver The tissue was perfused for 10 min with Krebs-bicarbonate buffer (with or without 0.3 mM Ca2+),and containing 0.5 mM fatty acid-free BSA and 0.5 mM palmitate prior to the addition of oleate (1 mM). Samples of 0.5-1 g of tissue were collected and immediately homogenized. Results are means f S.E. of at least three different experiments. PKC activity" Time

0 mM extracellular Ca2+ Plasma membrane

0.3 mM extracellular Ca2+ cytOsO'

Plasma membrane

rnin

0 5

1 1 2 f 14 108& 9

3 21 f063 f 8 35 f 9

10

NDd

ND

15 30

1 0 2 k 11 105 f 10

34f 7 30 & 8

89 f gb 6 79 3f 1 f 90 ' 66 & 5 63 f 8

In units/g of liver. Significantly different from control ( p < 0.01). Significantly different from control (p < 0.025). ND, not determined.

37 k 3 59 f ' 8 73 f 9 62 f 9

isoenzymes, the specific pattern of the oleate-induced translocation of PKC to the hepatocyte plasma membranes was analyzed by hydroxylapatite chromatography after purification on a DE52 column. As Fig. 4 shows, the majorform present in the plasma membrane of hepatocytes incubated with 0.5 mM BSA corresponded to the a isoenzyme. After incubation for 10 min with 0.5 mM BSA, 0.5 mM palmitate, and 1 mM oleate, a clear increase in the isoenzyme eluted at 90 mM potassium phosphate was evident (3.2-fold), whereas the enzyme eluted at 140 mM was increased by 30%. These two isoenzymes corresponded to the p and CY subspecies, as deduced by its chromatographic and kinetic properties (Ca2+dependent activation; data notshown). Moreover, to ascertain that the changes in the plasma membrane activity of PKC corresponded to differences in the amount of enzyme present in this compartment, the specific binding of [3H]PDBu to PKC was measured. As Fig. 4B also shows, the PDBu binding profile after hydroxylapatite chromatography was similar to the changes in enzyme activity. Hence, oleate promotes a differential translocation of PKC isoenzymes to the membrane. Moreover, in agreement with previous results on the characterization of the biochemical properties of the isoenzymes (1, l l ) , t h eform preferentially translocated by oleate was less dependent on calcium than the a isoenzyme (data not shown). Stimulation of PKC by Glycerolipids-The hepatic metabolism of oleate produces an increase in the diglyceride and triglyceride pools bothin the endoplasmic reticulum and plasma membrane fractions (16, 17). To test the capacity of these glycerolipids to modulate hepatic PKC activity, hepatocytes (2.5 g) were incubated for 10 min in the absence or presence of 1.7 mM oleate and 0.7 mM BSA. After isolation of the plasma membrane, the lipidic fraction was extracted and aliquots of the lipids resolved byTLC. The fractions were characterized by their respective RFcompared with appropriate lipid standards. Identical areas of the corresponding fractions of TLC from controls and oleate-incubated hepatocytes and quantified standards were excised (1.5-3 cm'), the lipids extracted, and the concentration ofDAG determined (30). The capacity of these lipidic fractions to stimulate partially purified soluble PKC activity from hepatocytes was measured by using a mixed micelle assay for PKC. The Triton X-100 micelle composition included varying amounts of phosphatidylserine and endogenous diacylglycerol from the hepatocyte membranes. Micelles prepared with Nonidet P-40 turned out to be less efficient in promoting PKC activation by diacylglycerol, despite the very close values of the critical micellar concentration for both detergents. Fig. 5 shows the relative stimulation of p and a PKC isoenzymesby the diacylglycerol lipidic fractions from control and oleate-incubated hepatocytes. These lipids banded with an RF value of 0.62 (Fig. 5, A and B, respectively) and the presence of diacylglycerol was determined and quantified by its conversion to phosphatidate by diacylglycerol kinase. The concentration of diacylglycerol was 120 and 570 nmol/g of cells for control and oleate-incubated hepatocytes. This fraction represented at least 70% of the DAG contained in the lipidic fraction. In thepresence of 0.3 mM Ca2+and at least 5 mol % of phosphatidylserine/micelle,a clear enhancement of PKC activity was observed when the plasma membrane diacylglycerols from oleate-incubated hepatocytes wereused. The corresponding lipidic fractions from control cells were markedly less effective in promoting activation of PKC, regardless of the type of isoenzyme assayed (aor p isoforms). Moreover, despite the high increase in the triglyceride pool (3-4-fold), this fraction failed to reproduce the activation

23573


-

I..

Fraction number

FIG. 4. Membrane-bound profile of PKC isoenzymes. Hepatocytes were incubated for 10 min in the presence of 0.5 mM BSA, 0.5 mM palmitate (O), or the same medium supplemented with 1 mM oleate (0).The particulate enzyme was extracted and resolved by DE52 and hydroxylapatite chromatography using a potassium phosphate gradient (dotted line).A, PKC activity profile; B, [3H]PDBubinding to samples processed as in panel A.

a-isoenzyme

0

2

4

0

2

0

4

0

Dlacylglycerol, (X. mold FIG. 5. Activation of soluble rat liver PKC by diacylglycerols isolated from the plasma membrane. Hepatocytes were incubated with 0.5 mM BSA and 0.5 mM palmitate, in the absence (open symbols) or in the presence (filled symbols) of 1 mM oleate. Plasma membrane lipids were fractionated by TLC, and thecorresponding

diacylglycerol fraction was assayed as activator of partially purified rat liver PKC under identical conditions, in a Triton X-lOO/lipids mixed micelle system and in the presence of 0.3 mM Ca2+.Micelles contained 5 mol % of phosphatidylserine. A control of oleoylacetylglycerol (0.5 pg) was processed by TLC as hepatic lipids, and the molar ratio in the micelles was 5% (U).Results are mean f S.E. for three independent preparations of lipids.

observed in the case of the diacylglycerol fractions (data not shown). To show that oleate effectively produces metabolites that interact with the diacylglycerol binding site of PKC, the binding of fluorescent phorbols (Bodipy-TPA) to intact hepatocytes was measured by flow cytometry. As shown in Fig. 6, incubation of hepatocytes with a BSA-oleate complex containing 0.5 mM albumin and 1 mM oleate decreased the binding of Bodipy-TPA to the cells, which indicates that metabolites coming from oleate (possibly DAG) are bound to PKC. Phosphorylation of Proteins-According to our results, oleate is able to activate PKC activity in hepatocytes either by its direct action on the enzyme or through the rise in the synthesis of diacylglycerols and related glycerolipids. To determine whether this activation of the enzyme may produce a specific pattern of protein phosphorylation in hepatocytes, two experimental approaches were used ( a ) in uiuo phosphorylation of proteins from hepatocytes labeled with ["PI phosphate and ( b ) in uitro phosphorylation of cytosolic proteins with purified rat liver PKC. Incubation with phorbol &butyrate was used as positive control for the phosphorylation of hepatic proteins by PKC. As a negative control for the in vivo phosphorylation of proteins, hepatocytes were labeled

1 .o

0.8

0.6

0.4

0.2

"." nn 0

50

100

150

200

1

250

Bodipy-TPA, ng/mi

FIG. 6. Binding of fluorescent phorbols to intact hepato-

cytes. Cells were incubated in the absence or presence of 1mM oleate for 10 min. After centrifugation (50 X g for 4 rnin), the hepatocytes were incubated at 4 "C for 30 min in the absence or presence of 2 p~ TPA prior to the addition of the indicated concentration of BodipyTPA. Results were referred to the maximal fluorescence intensity of the controls after subtracting the fluorescence of samples incubated with TPA. Results aremeans k S.E. of three experiments. with [32P]phosphate andincubated for 2 min with the PKC inhibitor staurosporine,prior to theaddition of oleate in order to prevent oleate-stimulated PKC phosphorylation. Samples were collected after 5 min of incubation with several ligands

23574

of

Oleate-induced Redistribution

PKC in Hepatocytes

and analyzed by SDS-PAGE and autoradiography.As shown dependent increasein free cytosolic Ca2+. Themaximal effect in Fig. 7A, the phosphorylation patternof intact hepatocytes was obtained a t concentrations of oleate between 1.5 and 2 exhibits a specific oleate-induced phosphorylation of an 82- mM, and produced a change in Ca2+ from80 to about280 nM. kDa protein, inhibited when the cells were preincubated with In contrast to oleate, elaidate produced only a small change staurosporine. The same situation was observed after incu- in Ca2+ (Fig. 8 B ) . T o ascertainthatoleate effectively inobserved bation with phorbol dibutyrate. Fig. 7 B shows the in vitro creased the intracellularCa", and that the changes phosphorylation by PKC of hepatocyte cytosolic proteins in fura 2 fluorescence could not be attributed to an effect of incubated with several ligands for 5 min. Oleate required the the fatty acid on the Ca2+ probe, glycogen phosphorylase a presence of Ca2+ to stimulate the phosphorylation of a t least activity, a n enzyme very sensitive to changesin Ca2+concentwo proteins(69 and 70 kDa, respectively), as did those which tration, was also measured. Fig. 9 shows the time course of with were phosphorylated in the presence of the phorbol. In this glycogen phosphorylase activity in hepatocytes incubated system, staurosporine also blocked this reaction. The differ- 1.5 mM oleate or 100 nM vasopressin, a Ca2+-mobilizing horence in substrate phosphorylation between intact hepatocytes mone for liver cells. The maximal activation was obtained and the in vitro assay suggests that other activators in addi- after 1-3 min of incubation and representedonly 40% of that conditions. The rise in tion to oleate are involved in the substrate phosphorylation obtained with vasopressin in the same Ca2+ by oleate may be related to a net entry from the extrain uiuo, in agreement with the distinct mechanisms of PKC cellular medium rather than to mobilization fromintracellular activation elicited by DAG/phosphatidylserine andby unsatstores since the effect of oleate was not observed in hepatourated fatty acids, respectively (12). of Ca2+ in the incubation Intracellular Calcium Mobilization-Since oleate produced cytesincubatedintheabsence medium (Figs. 2 and 8C and Table 111). In agreement with a specific pattern of phosphorylation of proteins in intact theseresults,nochangesinInsPsconcentration were obhepatocytes, and in vitro phosphorylation of hepaticsub9), a t least during served in cells incubated with oleate (Fig. strates required Ca", it was decided to determine whether the the time at which phosphorylase was activated. incubation of the cells with oleatewas able toproduce changes in the intracellular Ca2+ concentration. As Fig. 8 shows, in DISCUSSION the presence of 0.7 mM BSA, oleate produced concentrationa In thiswork we studied theeffect of oleate and related fatty acids on hepatocyte PKC subcellular distribution and actiA B 1 2 3 4 1 2 3 4 5 6 7 8 vation,using physiological concentrations of albumin and I fatty acids. The protocol followed to analyze the enzyme Kd Kd -70 content in the soluble and particulate fractions included the 02 - 6 - 69 presence of ion chelators(EDTAandEGTA),as well as protease inhibitors to prevent possible artifactual redistribuY tions of PKC in the course of the subcellular fractionation. % 40% of the enzyme was found Under these conditions at least '.' in the particulate fractions. However, after homogenization b: in the absenceof chelators more than 80% of the enzyme was FIG. 7. Oleate-dependentphosphorylation of hepatocyte recovered in particulate fractions, and this may explain the proteins. A, in uiuo phosphorylation of intact hepatocytes labeled high percentage of membrane-bound enzyme reported in a with ["Plphosphate and incubated for 5 min in the presence of 0.7 mM BSA with: 1, none; 2, 2 mM oleate; 3, 10 p~ staurosporine plus 2 previous work on rat liver PKC (41). The high percentage of low salt mM oleate; 4, 160 nM PDBu. B, phosphorylation of soluble hepatocyte membrane-bound enzyme cannot be attributed to the proteins by partially purified rat liver PKC. Lane 1, purified PKC, in concentration used duringthehepatocyte homogenization the absence of activators. Remaining lanes contained purified PKC since in the presence of physiological salt concentrations(150 (5 units/assay), G-25 filtered cytosolicproteins (20 mg/ml),and mM KCl) the distribution pattern of the enzyme was not phosphatidylserine (1pglml). Lane 2, 0.5 mM EGTA; 3, 0.5 mM affected (control not shown). This continuous presence of oleate; 4, 0.5 mM oleate plus 10 p~ staurosporine; 5, 10 p~ staurosporine; 6, 0.5 mM oleate plus 20 p~ Ca2+;7, 20 PM Ca2+;8, 160 nM membrane bound activity aisstriking featureof hepatic PKC PDBu. The incubation volume was 200 PI. shared also by the majorityof the metabolically active tissues OCEATE. ( f i )

2 1

0

U

5 3

1"

VASBRESSIN,

100 nH

1 nin

FIG.8. Intracellular Ca2+ mobilizationby unsaturated fatty acids. Hepatocytes were loaded with fura 2 prior to the addition of 0.7 mM BSA. Fatty acids were complexed with RSA, and the fluorescence was recorded at 510 nm after dual excitation at 320/380 nm. The extracellular Ca" was 1.5 mM (panels A, R, and n),or 9) mM (panel C ) . Arrows indicate thetime for the addition of fatty acids or vasopressin, used as control of Ca2' mobilization.
