Tumor Necrosis Factor Induces Early-Onset Endothelial Adhesivity by Protein Kinase C-Dependent...

Tumor Necrosis Factor-� Induces Early-Onset EndothelialAdhesivity by Protein Kinase C�–Dependent Activation of

Intercellular Adhesion Molecule-1Kamran Javaid, Arshad Rahman, Khandaker N. Anwar, Randall S. Frey,

Richard D. Minshall, Asrar B. Malik

Abstract—We tested the hypothesis that TNF-� induces early-onset endothelial adhesivity toward PMN by activating theconstitutive endothelial cell surface ICAM-1, the �2-integrin (CD11/CD18) counter-receptor. Stimulation of humanpulmonary artery endothelial cells with TNF-� resulted in phosphorylation of ICAM-1 within 1 minute, a response thatwas sustained up to 15 minutes after TNF-� challenge. We observed that TNF-� induced 10-fold increase in PMNadhesion to endothelial cells in an ICAM-1–dependent manner and that this response paralleled the rapid time courseof ICAM-1 phosphorylation. We also observed that the early-onset TNF-�–induced endothelial adhesivity was proteinsynthesis–independent and associated with cell surface ICAM-1 clustering. Pretreatment of cells with the pan-PKCinhibitor, chelerythrine, prevented the activation of endothelial adhesivity. As PKC�, an atypical PKC isoformabundantly expressed in endothelial cells, is implicated in signaling TNF-�–induced ICAM-1 gene transcription, wedetermined the possibility that PKC� was involved in mediating endothelial adhesivity through ICAM-1 expression. Weobserved that TNF-� stimulation of endothelial cells induced PKC� activation and its association with ICAM-1.Inhibition of PKC� by pharmacological and genetic approaches prevented the TNF-�–induced phosphorylation and theclustering of the cell surface ICAM-1 as well as activation of endothelial adhesivity. Thus, TNF-� induces early-onset,protein synthesis–independent expression of endothelial adhesivity by PKC�-dependent phosphorylation of cell surfaceICAM-1 that precedes the de novo ICAM-1 synthesis. The rapid ICAM-1 expression represents a novel mechanism forpromoting the stable adhesion of PMN to endothelial cells that is needed to facilitate the early-onset transendothelialmigration of PMN. (Circ Res. 2003;92:1089-1097.)

Key Words: tumor necrosis factor-� � intercellular adhesion molecule-1 � endothelium� polymorphonuclear leukocyte adhesion � protein kinase C�

Polymorphonuclear leukocytes (PMN) play an essentialrole as the first line of defense against the invading

microorganisms. Recruitment of PMN to the site of infectionis a multistep process involving sequential activation ofadhesive proteins on endothelial cells and their counter-receptors on the surface of leukocytes.1–3 Intercellular adhe-sion molecule-1 (ICAM-1), an adhesive protein expressed onthe endothelial cell surface, is a member of the immunoglob-ulin supergene family4 that serves as a counter-receptor forleukocyte �2-CD11/CD18 integrins.5 The interaction betweenICAM-1 and �2-integrins is required for firm adhesion andthe stable arrest of PMN on the endothelial cell surface5

before PMN can migrate across the endothelial barrier.3

Although ICAM-1 is normally present on the endothelialcell surface, its expression can be induced by proinflamma-tory mediators such as tumor necrosis factor-� (TNF-�) orthrombin.6–10 Time course studies showed that TNF-�–

induced ICAM-1 expression in endothelial cells was detectedwithin 1 hour and thereafter increased progressively over thenext 24 hours.6,7 The rapid expression of ICAM-1 is requiredto promote the early migration of PMN, which can bemaximal within 1 hour after activation by proinflammatorystimuli.10 Although the mechanisms regulating de novo ex-pression of ICAM-1 and the role of ICAM-1 in mediatingPMN-endothelial interactions have been studied in somedetail,3,11,12 little is known about the role of the constitutivecell surface ICAM-1 in promoting the rapid expression ofendothelial adhesivity. Studies showed significant ICAM-1expression and ICAM-1–dependent PMN adhesion to endo-thelial cells within minutes after exposure to an inflammatorysignal at a time when ICAM-1 would not be expected to besynthesized.10 It was postulated that rapid ICAM-1 expres-sion involves phosphorylation-dependent activation of thecell surface protein.10 Because the cytoplasmic domain of

Original received November 19, 2002; revision received April 10, 2003; accepted April 10, 2003.From the Department of Pharmacology, University of Illinois College of Medicine, Chicago, Ill. Current affiliation of A.R. is Department of Pediatrics,

University of Rochester School of Medicine, Rochester, NY.Correspondence to Asrar B. Malik, Distinguished Professor and Head, Department of Pharmacology, University of Illinois, College of Medicine, 835

S Wolcott Ave, Chicago, IL 60612. E-mail [email protected]© 2003 American Heart Association, Inc.

Circulation Research is available at http://www.circresaha.org DOI: 10.1161/01.RES.0000072971.88704.CB

1089

Molecular Medicine

by guest on February 11, 2016http://circres.ahajournals.org/Downloaded from

http://circres.ahajournals.org/

ICAM-1 has threonine residues at positions 521, 527, and530, we addressed the possibility that protein kinase C(PKC), a family of serine/threonine kinases,13,14 signalsICAM-1 activation. We have recently shown that TNF-�induces NADPH oxidase activation and ICAM-1 gene tran-scription in endothelial cells through a PKC�-dependentpathway.8,15 Because PKC� is critical in signaling severalcomponents of the inflammatory response,8,15,16 we addressedthe possibility that PKC�-dependent phosphorylation regu-lates the rapid induction of endothelial cell surface ICAM-1.We provide evidence herein that TNF-� induces early-onsetendothelial adhesivity toward PMN involving a qualitativealteration in cell surface ICAM-1 that is dependent onphosphorylation by PKC�.

Materials and MethodsMaterials125I and 32P were purchased from ICN Pharmaceuticals Inc. CellTracker dye and goat anti-mouse Alexa-488/568 IgG were fromMolecular Probes Inc. PKC� and PKC� peptide inhibitors wereobtained from Biosource International. ICAM-1 blocking mAb,CD-18 blocking mAb, nPKC� rabbit polyclonal Ab, goat anti-mousehorseradish peroxidase-linked (HRP) IgG, and goat anti-rabbit HRPIgG were from Santa Cruz Biotechnology. Phospho-PKC� antibody,which detects PKC� when it’s phosphorylated at Thr-410, waspurchased from Cell Signaling. Recombinant human TNF-� wasobtained from Promega Corp. Polyvinylidene difluoride (PVDF)membrane was from Millipore Corp, a protein assay kit fromBio-Rad Laboratories, Sephadex G-25 beads and Phosphate freeMEM from Sigma Aldrich Inc, and Polymorph prep was fromAccurate Chemical and Scientific Corp. All other materials werepurchased from Fisher Scientific Company.

Cell CultureHuman pulmonary arterial endothelial cells (HPAECs; BioWhit-taker, Walkersville, Md) were cultured as described.15 Confluentmonolayers were serum starved for 4 hours in EBM-2 before TNF-�challenge.

PMN Adhesion AssayPMN adhesion assay was performed as described.17 Briefly,HPAECs grown over 12-mm circular cover slips were labeled with3 �mol/L fluorescent (red) Cell-Tracker dye for 30 minutes beforestimulation with TNF-�, and then washed extensively to removeresidual TNF-�. Freshly isolated human neutrophils (PMN) werestained with 5 �mol/L fluorescent (green) Cell-Tracker dye, coin-cubated with endothelial cells for 20 minutes, washed with PBS, andvisualized using a fluorescent microscope. The adherent PMN werecounted and expressed as PMN/0.8 mm2 of endothelial cell.

Immunoblotting and ImmunofluorescenceImmunoblotting and immunofluorescence studies were performed asdescribed.15

Preparation of Cytosolic and Membrane FractionsCytosolic and membrane fractions were prepared as described.15

Briefly, cells were washed with ice-cold TBS and lysed in(in mmol/L) 10 Tris-HCl (pH 7.5), 1 MgCl2, 5 EDTA, 10 EGTA, 1NaVO4, and a cocktail of protease inhibitors (Sigma P-8340).Lysates were sonicated for 10 seconds and then centrifuged at100 000g for 1 hour at 4°C. Supernatants were collected anddesignated cytosolic fraction. The remaining pellets were resus-pended in the above lysis buffer containing 1% Triton X-100,sonicated, and incubated for 30 minutes at 4°C. These lysates weremicrofuged at 4°C, and the supernatants were designated membranefraction.

125I-labeled Antibody Binding AssayThe binding activity of cell surface ICAM-1 was determined by thespecific binding of 125I-labeled anti–ICAM-1 mAb to HPAECs asdescribed.10 Briefly, cells were fixed with paraformaldehyde (2%PFA for 15 minutes at 4°C), and then incubated with 10 �g/mL of125I-labeled ICAM-1 Ab for 1 hour at 4°C. The cells were washedthrice with EBM-2 containing 10% FBS and then lysed overnightwith 1N NaOH, and the lysates were counted for radioactivity. Thespecific binding of ICAM-1 antibody was determined by preincu-bating the cells with unlabeled antibody.

Recombinant AdenovirusesRecombinant adenovirus containing cDNA of dominant-negativePKC� (Ad-PKC� K281R) was a kind gift from Dr ViswanathanNatarajan (Johns Hopkins University School of Medicine, Balti-more, Md). Confluent HPAEC monolayers were infected withAd-DN-PKC� at a concentration of 4�107 plaque-forming units/mL,and adenoviruses carrying LacZ gene (Ad-LacZ; Clontech) wereused as controls as described.18 After incubation for 5 hours at 37°C,5% CO2, and 95% humidity, the medium was replaced with freshEBM-2–containing10% FBS for 48 hours before performing theexperiments.

Phosphorylation of ICAM-1Confluent HPAEC monolayers were labeled with 32P (75 �Ci/mLovernight, or 150 �Ci/mL for 4 hours) in phosphate-free medium at37°C. After treatment, cells were washed 3 times with ice-cold PBSand lysed with 1 mL/plate RIPA buffer [150 mmol/L NaCl,50 mmol/L Tris-HCl (pH 7.2), 1% deoxycholic acid, 1% TritonX-100, 0.25 mmol/L EDTA (pH 8.0), 5 mmol/L NaF supplementedwith 100 �mol/L sodium orthovanadate, 1 mmol/L PMSF, 5 �g/mLleupeptin, 5 �g/mL aprotinin, 1 �g/mL pepstatin]. Lysates wereprecleared with normal mouse IgG (0.25 �g/mL) and then immu-noprecipitated with anti–ICAM-1 Ab (2 �g/mL, 4°C, overnight).Immunocomplexes were washed and resolved by SDS/PAGE. Gelswere transferred to PVDF membrane, and the phosphorylated formof ICAM-1 was detected by autoradiography. The membrane wassubsequently immunoblotted with an antibody against ICAM-1.19

The bands were quantified densitometrically using Scion Image, andthe values were normalized to total protein in each lane.

Transient TransfectionThe plasmid pGreen Lantern-1 containing green fluorescence protein(GFP) gene was purchased from GIBCO-BRL. The expressionvector pcDNA3 containing tagged kinase-defective PKC� isoformwas a gift from Dr J.W. Soh (Columbia University, New York,NY).20 Endothelial transfections were performed with Superfect(Qiagen) as described.19

PKC� Kinase AssayPKC� kinase activity was assayed as described.19 Briefly, cell lysateswere immunoprecipitated with an antibody against PKC� usingprotein A/G plus agarose (Santa Cruz). The immunocomplexes werewashed twice with ice-cold PBS and once with kinase buffer(25 mmol/L Tris-HCl [pH 7.4], 5 mmol/L MgCl2, 0.5 mmol/LEGTA, 1 mmol/L DTT) and resuspended in 30 �L of kinase buffercontaining 2.5 �g of histone-H1, 0.5 mmol/L cold ATP, and 20 to 30�Ci of [�-32P]ATP. The reaction was incubated for 20 minutes at RTand terminated by addition of SDS-sample buffer. Proteins wereanalyzed by SDS-PAGE, and the phosphorylated form of histone-H1was detected by autoradiography.

ResultsTNF-� Induces Rapid Endothelial Adhesivity inan ICAM-1–Dependent MannerWe determined the ability of TNF-� to induce early-onset,protein synthesis–independent endothelial adhesivity towardPMN. Time course experiments showed that endothelial

1090 Circulation Research May 30, 2003



adhesivity increased �10-fold within 1 minute, and theresponse sustained up to 30 minutes, and increased further(�20-fold) at 4 hours after TNF-� challenge (Figure 1A).The rapidity of the response suggests that the modification ofpreexisting protein is involved in the mechanism of endothe-lial adhesivity. As ICAM-1 is constitutively expressed inendothelial cells, we addressed its role in mediating the rapidexpression of endothelial adhesivity. Preincubation ofHPAECs with anti–ICAM-1-mAb prevented TNF-�–inducedendothelial adhesivity toward PMN (Figure 1B). Anti–CD18mAb also prevented the TNF-�–induced endothelial adhesiv-ity, but the isotype-matched control IgG had no effect. Todemonstrate that the early-onset endothelial adhesivity ismediated by constitutive cell surface ICAM-1 and does notrequire de novo ICAM-1 expression, we determined the timecourse of ICAM-1 expression induced by TNF-�. Westernblot analysis showed that ICAM-1 is constitutively expressedin endothelial cells and that its expression is induced onlyafter 1 hour of TNF-� challenge (Figure 2A) consistent withthe delayed secondary expression of endothelial adhesivity(Figure 1A). Moreover, the early adhesive response wasinsensitive to the protein synthesis inhibitor cycloheximide

(CHX) (Figure 2B). These data show that TNF-� induces arapid expression of endothelial adhesivity toward PMN andthat this response is mediated by alteration in the constitutivecell surface ICAM-1.

TNF-� Induces Rapid Binding of Anti-ICAM mAbto Cell Surface ICAM-1We addressed the possibility that the expression of rapidendothelial adhesivity is secondary to activation of theconstitutive cell surface ICAM-1. Thus, we assessed thebinding of anti–ICAM-1 mAb to the ICAM-1 followingTNF-� exposure of endothelial cells. We used 125I-labeledanti–ICAM-1 mAb to determine the binding activity of cellsurface ICAM-1. Results showed that TNF-� induced bind-ing of anti–ICAM-1 mAb to the cell surface ICAM-1 withinminutes, paralleling the expression of endothelial adhesivity(Figures 3A and 1A). In a competition experiment, binding of

Figure 1. TNF-� induces rapid endothelial adhesivity towardPMN in an ICAM-1–dependent manner. Confluent HPAECmonolayers were treated with TNF-� (1000 U/mL) for indicatedtime periods, and adhesion assay was performed as describedin Materials and Methods. A, TNF-� challenge resulted in a�10-fold increase in PMN adhesion to EC within 1 minute ofstimulation, and increased further (�20 fold) after 4 hours. B,After TNF-� (1000 U/mL) challenge for 15 minutes, HPAECmonolayers were incubated with blocking antibodies (0.2�g/mL) for 15 minutes before incubation with PMN. This rapidadhesivity was blocked by ICAM-1 and CD18 antibodies, butisotype-matched normal IgG failed to prevent the response.Results are representative of 3 separate experiments. Bars indi-cate mean�SEM. *Difference from control (P�0.05). #Differencefrom TNF-�–stimulated control (P�0.05).

Figure 2. TNF-�–induced rapid endothelial adhesivity is inde-pendent of de novo ICAM-1 expression. A, Confluent HPAECmonolayers were treated with TNF-� (200 U/mL) for indicatedtime periods. Total cell lysate (15 �g/lane) was separated by10% SDS-PAGE and immunoblotted with an antibody againstICAM-1 and actin. Chart shows the ratio of ICAM-1 to actin. B,Cells were pretreated with cycloheximide (CHX, 0.1 �g/mL, 15minutes) before TNF-� (1000 U/mL, 15 minutes) challenge andthen incubated with PMN to determine adhesion as described inMaterials and Methods. Results are representative of 3 separateexperiments. Bars indicate mean�SEM. *Difference from control(P�0.05).

Javaid et al TNF-� Induces Rapid Endothelial Adhesivity via PKC� 1091



125I-labeled anti–ICAM-1 mAb to endothelial cells was effec-tively competed by the presence of excess unlabeled ICAM-1antibody (Figure 3A), thus indicating the specificity of theresponse. In another experiment, we failed to detect any increasein the membrane expression of ICAM-1 after TNF-� challenge(Figure 3B). These data exclude the possibility that TNF-�–induced increase in ICAM-1 binding activity is the result ofrapid translocation of a putative cytoplasmic pool of ICAM-1 tothe membrane. We next determined if the increased ICAM-1

binding activity involves qualitative changes in the preexistingICAM-1. Analysis by confocal microscopy showed that TNF-�induced a distinct punctate ICAM-1 staining suggestive of cellsurface ICAM-1 clustering (Figure 3C).

TNF-� Induces Phosphorylation of theConstitutive ICAM-1We determined the phosphorylation status of ICAM-1 afterTNF-� challenge of HPAECs to ascertain if the ICAM-1

Figure 3. A, TNF-� induces the binding activity of constitutive cell surface ICAM-1. HPAECs grown in 48-well plates were stimulatedwith TNF-� (200 U/mL) for indicated time periods before incubation with 125I-labeled ICAM-1 Ab (10 �g/mL) at 4°C. After 1 hour, cellswere washed, lysed, and counted for radioactivity as described in Materials and Methods. In some experiments, the specific binding ofICAM-1 antibody was determined by preincubating the cells with unlabeled antibody. Values are shown as mean�SEM; n�6 for eachcondition. *Difference from control (P�0.05). #Difference from 15 minutes TNF-� stimulation without unlabeled ICAM-1 mAb (P�0.05).B, TNF-�–induced binding activity does not involve membrane translocation of ICAM-1. Confluent HPAEC monolayers were treatedwith TNF-� (200 U/mL) for indicated time periods and cytoplasmic and membrane fractions (15 �g/lane) were separated by 10% SDS-PAGE and immunoblotted for ICAM-1. C, TNF-� induces clustering of constitutive cell surface ICAM-1. HPAECs were stimulated with200 U/mL TNF-� for 15 minutes and 30 minutes (inset), cells were then fixed using 2% PFA and stained with an ICAM-1 antibody incombination with 4�,6-diamidino-2-phenylindole (DAPI) to view the nucleus as described in Materials and Methods. Slides weremounted and analyzed by confocal microscopy. Results are representative of 3 experiments. Magnification: �60; Inset, �100.




activation is attributable to its phosphorylation. We found thatTNF-� induced ICAM-1 phosphorylation in a time-dependent manner; ie, the phosphorylated form of ICAM-1was detected as early as 1 minute and was sustained up to 15minutes after TNF-� challenge (Figure 4). This time courseparallels the TNF-�–induced ICAM-1 clustering, ICAM-1 bind-ing activity, and expression of ICAM-1–dependent endothelialadhesivity (Figure 4 versus Figures 1A, 3C, and 3A).

Inhibition of PKC� Prevents Expression ofEarly-Onset Endothelial AdhesivityWe addressed the possibility that PKC�, an atypical PKCisoform abundantly expressed in endothelial cells,8 mediatesthe rapid induction of endothelial adhesivity through phos-phorylation of ICAM-1. We used both general and isoform-specific inhibitors to evaluate the involvement of PKC�.Pretreatment of HPAECs with chelerythrine, a pan-PKCinhibitor, prevented the early-onset endothelial adhesivitytoward PMN induced by TNF-� (Figure 5A). As PKC�signals TNF-�–induced oxidant generation and ICAM-1 genetranscription,8,15 we next addressed the possibility that PKC�mediates the TNF-�–induced early-onset endothelial adhe-sivity. We used a myristoylated membrane-permeable peptideantagonist corresponding to the pseudosubstrate region ofPKC� that specifically inhibits this PKC isoform,15,21 toevaluate its role in the response. Results showed that inhibi-tion of PKC� prevented the expression of the rapid endothe-

Figure 4. TNF-� induces rapid ICAM-1 phosphorylation.HPAECs were metabolically labeled with 32P-orthophosphate,treated for the indicated times with TNF-� (200 U/mL), andICAM-1 was immunoprecipitated from cell lysates. A, ICAM-1immunoprecipitates were separated by 10% SDS-PAGE, trans-ferred onto the PVDF membranes, and the phosphorylation formof ICAM-1 was detected by autoradiography. B, Membrane in Awas immunoblotted with an antibody against ICAM-1 to indicateequal loading of the protein. C, Timeline showing the ratio ofphosphorylated ICAM-1 to total ICAM-1. Corrected density ofeach band in A was calculated by subtracting the background,and the value thus obtained was divided by the density of cor-responding band in B. Results are representative of 2 separateexperiments.

Figure 5. Inhibition of PKC� prevents TNF-�–induced endotheli-al adhesivity. HPAECs were pretreated with chelerythrine (10�mol/L, 20 minutes, 37°C) (A) or PKC� peptide inhibitor (1�mol/L, 20 minutes, 37°C) or PKC� peptide inhibitor (1 �mol/L,20 minutes, 37°C) (B) before stimulation with TNF-� (1000U/mL, 15 minutes). Expression of endothelial adhesiveness wasdetermined by PMN adhesion assays as described in Materialsand Methods. Results are representative of 3 separate experi-ments. Bars indicate mean�SEM. *Difference from control(P�0.05). #Difference from TNF-� stimulation without inhibitorpretreatment (P�0.05). C, Inhibition of PKC� prevents TNF-�–induced binding activity of constitutive cell surface ICAM-1.HPAECs were pretreated with PKC� or PKC� peptide inhibitor (1�mol/L, 20 minutes, 37°C) and then stimulated with TNF-� (200U/mL). After 15 minutes, 125I-labeled ICAM-1 Ab (10 �g/mL) wasincubated for 1 hour at 4°C, and binding activity was measuredas described in Materials and Methods. Values are shown asmean�SEM; n�6 for each condition. *Difference from control(P�0.05). #Difference from TNF-� stimulation without inhibitorpretreatment (P�0.05).




lial adhesivity in response to TNF-� challenge (Figure 5B). Incontrol experiments, we found that pretreatment of cells withmyristoylated peptide antagonist specific for PKC� failed toinhibit this response (Figure 5B).

Inhibition of PKC� Prevents TNF-�–InducedBinding Activity of Constitutive CellSurface ICAM-1We next determined the effects of inhibition of PKC� onTNF-�–induced binding activity of the constitutive cellsurface ICAM-1. We observed that inhibition of PKC� by thespecific peptide antagonist prevented the binding activity ofthe cell surface ICAM-1 induced by TNF-�, whereas inhibi-tion of PKC� had no effect (Figure 5C). These data are inaccordance with the effects of inhibition of PKC� on theTNF-�–induced expression of rapid endothelial adhesivity(Figure 5B).

Because the increased ICAM-1 binding activity is associ-ated with the clustering of cell surface ICAM-1 (Figure 3), wedetermined the requirement of PKC� in the mechanism of thisresponse. In these experiments, we used a kinase-defectivemutant of PKC� (PKC�K281R) to inhibit the function ofendogenous PKC�. We observed that expression ofPKC�K281R prevented TNF-�–induced ICAM-1 clustering

(Figure 6C). We also showed that expression of constitutivelyactive PKC� mutant induced ICAM-1 clustering in theabsence of TNF-� challenge (Figure 6D).

Inhibition of PKC� Prevents TNF-�–InducedPhosphorylation of ICAM-1The involvement of PKC� in TNF-�–induced ICAM-1 clus-tering and the resultant endothelial adhesivity led us toinvestigate if PKC� phosphorylates ICAM-1. The effect ofinhibition of PKC� on ICAM-1 phosphorylation was deter-mined by infecting HPAECs with nonreplicating adenoviralvectors containing either kinase-inactive dominant-negativemutant of PKC� (Ad-DN-PKC�) or �-galactosidase (Ad-�-Gal). We found that expression of Ad-DN-PKC� preventedTNF-�–induced phosphorylation of ICAM-1, whereas ex-pression of Ad-�-Gal failed to inhibit the response (Figure 7).

We next determined if PKC� directly associates withICAM-1 or requires participation of an intermediate proteinto phosphorylate ICAM-1. We first determined if TNF-�induces activation of PKC�. Results showed that TNF-�induced PKC� activation in a time-dependent manner andthat this activation was associated with a significant translo-cation of PKC� to the membrane (Figures 8A and 8B). Wecarried out coimmunoprecipitation studies to determine if

Figure 6. Expression of kinase-defective PKC� mutant prevents TNF-�–induced ICAM-1 clustering. HPAECs were cotransfected withgreen fluorescence protein (GFP) in combination with pcDNA3HA (A and B), kinase-defective PKC� (PKC�K281R) mutant (C), or consti-tutively active PKC� (PKC�CAT) mutant (D). At 24 hours, cells were left untreated (A and D) or stimulated with 200 U/mL TNF-� for 15minutes (B and C). Cells were fixed and stained with the anti–ICAM-1 antibody and images were acquired with the Zeiss LSM 510 con-focal microscope as described in Materials and Methods. Results are representative of 2 experiments. Panels: top left, ICAM-1 staining(red); top right, DAPI nuclear staining (blue); bottom left, GFP (green); and bottom right, composite.




PKC� translocated to the membrane in order to associate withICAM-1. Results showed increased amount of PKC� in theICAM-1 immunoprecipitates from TNF-�–challenged cellscompared with control untreated cells (Figure 8C). Theassociation of PKC� with ICAM-1 was rapid (occurringwithin 1 minute) and was sustained up to 30 minutes ofTNF-� challenge (Figure 8C), similar to the time course ofICAM-1 phosphorylation and increase in endothelial adhe-sivity (Figures 4 and 1A). These data indicate that ICAM-1 isa direct substrate for phosphorylation by PKC� in response toTNF-� challenge.

DiscussionICAM-1, an inducible endothelial counter-receptor for �2-integrins (CD11/CD18), mediates PMN-endothelial interac-tions, and thus it is essential for the mechanism of stablePMN adhesion and the consequent transendothelial PMNmigration. Interaction of ICAM-1 with �2 integrins enablesPMN to adhere firmly and stably to the vascular endotheliumand to migrate across the endothelial barrier. As ICAM-1 isalso constitutively expressed in endothelial cells and PMNare rapidly recruited as the first line of defense to site ofinfection,11,22 we reasoned that the constitutively expressedcell surface ICAM-1 may have an important role in mediatingthe rapid-onset of endothelial adhesivity toward PMN. In thepresent study, we provide evidence that the proinflammatorycytokine TNF-� induces early endothelial adhesivity towardPMN in an ICAM-1–dependent manner. Our data establishthat this adhesive response does not involve de novo ICAM-1protein synthesis, but it is the result of activation of the

constitutive ICAM-1. This was demonstrated by the increasedbinding activity of the cell surface ICAM-1 to anti–ICAM-1mAb. We further showed that PKC� rapidly associates withICAM-1 to phosphorylate ICAM-1, and that this event wasessential for the early-onset TNF-�–induced endothelialadhesivity.

We analyzed the time course of TNF-�–induced endothelialadhesivity. Results showed that TNF-� activated endothelialadhesivity toward PMN within 1 minute after TNF-� challenge.Moreover, this response was insensitive to the protein synthesisinhibitor cycloheximide. The rapidity of the response and itsinsensitivity to cycloheximide excluded the involvement of aprotein synthesis–dependent mechanism. The rapid time coursealso suggested a role of the preexisting ICAM-1 in mediating theendothelial adhesivity induced by TNF-�. To establish theinvolvement of constitutive cell surface ICAM-1 in the response,

Figure 7. Inhibition of PKC� by DN-PKC� prevents TNF-�–in-duced ICAM-1 phosphorylation. Confluent HPAECs infectedwith recombinant adenoviruses expressing DN-PKC� (Ad-PKC�-DN) and �-galactosidase (Ad-�-Gal) were metabolically labeledwith 32P-orthophosphate and then stimulated with TNF-� (200U/mL). ICAM-1 was immunoprecipitated, and phosphorylationstatus of ICAM-1 was determined as described in Figure 4. A,Phosphorylated form of ICAM-1. B, Total ICAM-1. C, Bar graphshowing the ratio of phosphorylated ICAM-1 to total ICAM-1was obtained as described in Figure 4C. Results are representa-tive of 2 separate experiments. Figure 8. A, TNF-� induces PKC� activation. HPAECs were

stimulated with TNF-� (200 U/mL) as indicated. Cell lysateswere immunoprecipitated with an antibody against PKC�, and invitro kinase assay was performed on immunoprecipitates usinghistone H1 as the exogenous substrate. Proteins were analyzedby SDS-PAGE, and phosphorylated form of histone H1 wasdetermined as a measure of PKC� activity. Bar graph representsthe relative density of bands. B, TNF-� induces membranetranslocation of PKC�. HPAECs were treated with TNF-� (200U/mL) for different time periods. Cells were lysed and mem-brane fractions were prepared as described in Materials andMethods. Membrane fractions (15 �g/lane) were subjected to10% SDS-PAGE and immunoblotted for PKC� and ICAM-1. C,TNF-� induces association of PKC� with ICAM-1. HPAECs werechallenged with TNF-� (200 U/mL) as indicated. ICAM-1 immu-noprecipitates were separated by 10% SDS-PAGE and immu-noblotted for PKC� and ICAM-1. Results are representative of 3experiments.




we determined the effect of the specific blocking antibodyagainst ICAM-1 on PMN adhesion to endothelial cells. Prein-cubation of HPAECs with anti–ICAM-1 antibody preventedTNF-�–induced endothelial adhesivity toward PMN, whereasan isotype-matched control Ab failed to alter this early response.The involvement of ICAM-1 in the mechanism of early-onsetendothelial adhesivity was further supported by the ability of theantibody against CD18, the counter-receptor for ICAM-1, tosimilarly inhibit PMN adhesion to endothelial cells. Becauseendothelial monolayers were extensively washed to removeresidual TNF-� before the addition of freshly isolated PMN, theincreased PMN adhesion cannot be ascribed to PMN activationby any residual TNF-�.

We addressed the possibility that there may be a cytoplas-mic pool of ICAM-1 that rapidly translocates to the mem-brane in response to TNF-�, and thus can increase the cellsurface ICAM-1, resulting in increased PMN adherence to theendothelial cells. However, we failed to detect any increase inmembrane expression of ICAM-1 after TNF-� challenge. Inrelated studies, we used 125I-labeled anti–ICAM-1 mAb toevaluate if TNF-� promotes endothelial adhesivity by in-creasing the binding activity of constitutive cell surfaceICAM-1. We observed that TNF-� indeed induced anti–ICAM-1 mAb binding of cell surface ICAM-1 with a timecourse similar to the expression of endothelial adhesivity.These data clearly demonstrate that TNF-�–induced early-onset endothelial adhesivity is secondary to activation of thepreexisting cell surface ICAM-1.

Hermanowski-Vosatka et al23 have shown that clusteringof cell surface receptors allows them to bind their ligandmore avidly than the same number of receptors in randomarray. These studies led us to assess the possibility thatincreased binding activity involves a qualitative change inthe constitutive ICAM-1 expressed on the cell surface. Ourdata showed a pattern of surface clustering of ICAM-1 asevident by punctate staining of the ICAM-1 after TNF-�stimulation. These data are consistent with the findingsshowing the clustering of monocyte binding receptorsincluding ICAM-1 as a mechanism for increased monocyteadhesion.24 Interestingly, both our study and that ofWojciak-Stothard et al24 showed clustering of ICAM-1,although in contrast to our study, Wojciak-Stothard et alfocused on the role of de novo expression of ICAM-1. Insupport of ICAM-1 clustering being a critical determinantof endothelial adhesivity, Jun et al25 have recently shownthat dimerized ICAM-1 has �1.5- to 3-fold greater affinitytoward the I-domain of integrin �L�2 (LFA-1) than mono-meric ICAM-1. These studies also postulated that dimersand W-shaped tetramers form oligomers of ICAM-1,which could have important implications for regulatingcell adhesion.25 These results collectively point to animportant role of TNF-� in promoting ICAM-1 clustering,and the increased binding activity of cell surface ICAM-1to PMN.

We addressed the mechanisms by which TNF-� promotesearly-onset ICAM-1–dependent endothelial adhesivity. Asphosphorylation is critical for activation of other adhesionproteins, eg, CD18,26,27 we investigated the possibility thatcell surface activation of ICAM-1 could occur by a mecha-

nism involving the phosphorylation of ICAM-1. Analysis ofphosphorylation status showed that ICAM-1 was phosphor-ylated within 1 minute of TNF-� challenge, a time-courseparalleling the increase in endothelial adhesivity. Clues to thekinase responsible for ICAM-1 phosphorylation came fromthe experiments in which we determined the effects ofinhibition of PKC. These results showed that pretreatment ofcells with chelerythrine, a pan-PKC inhibitor,15 prevented theinduction of rapid endothelial adhesivity. As PKC�, anabundant PKC isoform in endothelial cells, is implicated inthe mechanism of TNF-�–induced oxidant signaling viaNADPH oxidase activation as well as ICAM-1 gene tran-scription,8,15 we reasoned that PKC� may phosphorylate thecell surface ICAM-1 to induce its activation and endothelialadhesivity. We used pharmacological and genetic approachesto address the role of PKC� in the mechanism of TNF-�–induced endothelial adhesivity. Pretreatment of HPAECswith a myristoylated membrane-permeable peptide antagonistcorresponding to the pseudosubstrate region of PKC�, knownto inhibit PKC� activity,15,21 prevented the TNF-�–inducedbinding activity of the cell surface ICAM-1. Inhibition ofPKC� by this approach also prevented TNF-�–induced en-dothelial adhesivity. Moreover, we showed that expression ofkinase-defective PKC� mutant prevented TNF-�–inducedphosphorylation and clustering of the cell surface ICAM-1.We observed that expression of the constitutively activemutant of PKC� induced ICAM-1 clustering in the absence ofTNF-� challenge. Thus, our results show an important role ofPKC� in mediating TNF-�–induced ICAM-1 clustering andthe early-onset of endothelial adhesivity. These results, how-ever, do not exclude the involvement of other PKC isoformsin the mechanism of endothelial adhesivity. Indeed, we haveshown that PKC� can signal ICAM-1 expression and therebyendothelial adhesivity in response to thrombin, which acti-vates 7 transmembrane G protein–coupled receptors.19,28

Other studies have demonstrated the involvement of PKC� inthe mechanism of ICAM-1 expression and monocyte adhe-sion in response to TNF-� and interferon-� (IFN-�) challengeof epithelial cells.29,30

This specific involvement of PKC� in signaling ICAM-1phosphorylation, ICAM-1 clustering, and ICAM-1–depen-dent rapid endothelial adhesivity led us to examine whetherPKC� can directly phosphorylate ICAM-1 or it activates anintermediate kinase that is responsible for phosphorylation.Coimmunoprecipitation studies showed that TNF-� inducedthe association of PKC� with ICAM-1 and that the kinetics ofthis event was similar to the activation and membranetranslocation of PKC�. The present data show that ICAM-1 isa direct substrate of PKC�; however, it is possible that PKC�may also phosphorylate ICAM-1 through other kinases suchas MAP kinases (ERK1/2 and p38), known to signal down-stream of PKC�.31,32

In summary, the present study implicates the constitutivecell surface ICAM-1 as a critical determinant of early-onsetof endothelial adhesivity induced by TNF-�. Our data showthat induction of binding activity of the cell surface ICAM-1is secondary to its clustering, which in turn is the result ofphosphorylation of ICAM-1 by PKC�. Inhibition of PKC�activity by pharmacological and genetic approaches pre-




vented the TNF-�–induced ICAM-1 clustering and the early-onset of endothelial adhesivity. As PKC� also regulatesICAM-1 gene expression in endothelial cells8 and ICAM-1upregulation contributes to the amplification of PMN adhe-sion response,29 strategies aimed at preventing TNF-�–in-duced PKC� activation may be useful in controlling inappro-priate PMN sequestration and PMN-mediated vascularendothelial injury.

AcknowledgmentsThis work was supported by NIH grants HL46350, HL45638, andHL67424.

References1. Lawrence MB, Springer TA. Leukocytes roll on a selectin at physiologic

flow rates: distinction from and prerequisite for adhesion throughintegrins. Cell. 1991;65:859–873.

2. von Andrian UH, Chambers JD, McEvoy LM, Bargatze RF, Arfors KE,Butcher EC. Two-step model of leukocyte-endothelial cell interaction ininflammation: distinct roles for LECAM-1 and the leukocyte �2-integrinsin vivo. Proc Natl Acad Sci U S A. 1991;88:7538–7542.

3. Springer TA. Traffic signals for lymphocyte recirculation and leukocyteemigration: the multistep paradigm. Cell. 1994;76:301–314.

4. Staunton DE, Marlin SD, Stratowa C, Dustin ML, Springer TA. Primarystructure of ICAM-1 demonstrates interaction between members of theimmunoglobulin and integrin supergene families. Cell. 1988;52:925–933.

5. Smith CW, Marlin SD, Rothlein R, Toman C, Anderson DC. Cooperativeinteractions of LFA-1 and MAC-1 with inter-cellular adhesionmolecule-1 in facilitating adherence and trans-endothelial migration ofhuman neutrophils in vitro. J Clin Invest. 1989;83:2008–2017.

6. Wertheimer SJ, Myers CL, Wallace RW, Parks TP. Intercellular adhesionmolecule-1 gene expression in human endothelial cells: differential reg-ulation by TNF-� and phorbol myristate acetate. J Biol Chem. 1992;267:12030–12035.

7. Ritchie AJ, Johnson DR, Ewenstein BM, Pober JS. Tumor necrosis factorinduction of endothelial cell surface antigens is independent of proteinkinase-C activation or inactivation: studies with phorbol myristate acetateand staurosporine. J Immunol. 1991;146:3056–3062.

8. Rahman A, Anwar KN, Malik AB. PKC� mediates TNF-�–inducedICAM-1 gene transcription in endothelial cells. Am J Physiol CellPhysiol. 2000;279:C906–914.

9. Rahman A, Bando M, Kefer J, Anwar KN, Malik AB. Protein kinase-C-activated oxidant generation in endothelial cells signals ICAM-1 genetranscription. Mol Pharmacol. 1999;55:575–583.

10. Sugama Y, Tiruppathi C, Janekidevi K, Andersen TT, Fenton JW II,Malik AB. Thrombin-induced expression of endothelial P-selectin andintercellular adhesion molecule-1: a mechanism for stabilizing neutrophiladhesion. J Cell Biol. 1992;119:935–944.

11. Malik AB, Lo SK. Vascular endothelial adhesion molecules and tissueinflammation. Pharmacol Rev. 1996;48:213–229.

12. Doerschuk CM. Mechanisms of leukocyte sequestration in inflamedlungs. Microcirculation. 2001;8:71–88.

13. Nishizuka Y, Nakamura S. Lipid mediators and protein kinase-C forintracellular signaling. Clin Exp Pharmacol Physiol Suppl. 1995;22:S202–203.

14. Dekker LV, Palmer RH, Parker PJ. The protein kinase-C and proteinkinase-C related gene families. Curr Opin Struct Biol. 1995;5:396–402.

15. Frey RS, Rahman A, Kefer JC, Minshall RD, Malik AB. PKC� regulatesTNF�–induced activation of NADPH oxidase in endothelial cells. CircRes. 2002;90:1012–1019.

16. Leitges M, Sanz L, Martin P, Duran A, Braun U, Garcia JF, Camacho F,Diaz-Meco MT, Rennert PD, Moscat J. Targeted disruption of the �PKCgene results in the impairment of the NF-�B pathway. Mol Cell. 2001;8:771–780.

17. Bhunia AK, Arai T, Bulkley G, Chatterjee S. Lactosylceramide mediatesTNF�–induced ICAM-1 expression and the adhesion of neutrophil inhuman umbilical vein endothelial cells. J Biol Chem. 1998;273:34349–34357.

18. Suzama K, Takahara N, Suzuma I, Isshiki K, Ueki K, Leitges M, AielloLP, King GL. Characterization of protein kinase-C� isoform’s action onretinoblastoma protein phosphorylation, vascular endothelial growth fac-tor–induced endothelial proliferation, and retinal neovascularization.Proc Natl Acad Sci U S A. 2002;99:721–726.

19. Rahman A, Anwar KN, Uddin S, Xu N, Ye RD, Platanias LC, Malik AB.Protein kinase-C� regulates thrombin-induced ICAM-1 gene expressionin endothelial cells via activation of p38 mitogen-activated protein kinase.Mol Cell Biol. 2001;21:5554–5565.

20. Soh JW, Lee EH, Prywes R, Weinstein IB. Novel roles of specificisoforms of protein kinase-C in activation of the c-fos serum responseelement. Mol Cell Biol. 1999;19:1313–1324.

21. House C, Kemp BE. Protein kinase-C contains a pseudosubstrate pro-totype in its regulatory domain. Science. 1987;238:1726–1728.

22. Panes J, Perry MA, Anderson DC, Manning A, Leone B, Cepinskas G,Rosenbloom CL, Miyasaka M, Kvietys PR, Granger DN. Regional dif-ferences in constitutive and induced ICAM-1 expression in vivo. Am JPhysiol. 1995;269:H1955–H1964.

23. Hermanowski-Vosatka A, Detmers PA, Gotze O, Silverstein SC, WrightSD. Clustering of ligand on the surface of a particle enhances adhesion toreceptor-bearing cells. J Biol Chem. 1988;263:17822–17827.

24. Wojciak-Stothard B, Williams L, Ridley AJ. Monocyte adhesion andspreading on human endothelial cells is dependent on Rho-regulatedreceptor clustering. J Cell Biol. 1999;145:1293–1307.

25. Jun CD, Carman CV, Redick SD, Shimaoka M, Erickson HP, SpringerTA. Ultrastructure and function of dimeric, soluble ICAM-1. J BiolChem. 2001;276:29019–29027.

26. Buyon JP, Philips MR, Merrill JT, Slade SG, Leszczynska-Piziak J,Abramson SB. Differential phosphorylation of the �2-integrinCD11b/CD18 in the plasma and specific granule membranes of neu-trophils. J Leukoc Biol. 1997;61:313–321.

27. Valmu L, Fagerholm S, Suila H, Gahmberg CG. The cytoskeletal asso-ciation of CD11/CD18 leukocyte integrins in phorbol ester–activatedcells correlates with CD18 phosphorylation. Eur J Immunol. 1999;29:2107–2118.

28. Rahman A, True AL, Anwar KN, Ye RD, Voyno-Yasenetskaya TA,Malik AB. G�q and G�� regulate PAR-1 signaling of thrombin-inducedNF-�B activation and ICAM-1 transcription in endothelial cells. CircRes. 2002;91:398–405.

29. Chen C, Chou C, Sun Y, Huang W. TNF�–induced activation of down-stream NF-�B site of the promoter mediates epithelial ICAM-1expression and monocyte adhesion: involvement of PKC�, tyrosinekinase, and IKK2, but not MAPKs, pathway. Cell Signal. 2001;13:543–553.

30. Chang YJ, Holtzman MJ, Chen CC. Interferon-�–induced epithelialICAM-1 expression and monocyte adhesion: involvement of PKC-dependent c-Src tyrosine kinase activation pathway. J Biol Chem. 2002;277:7118–7126.

31. Kampfer S, Uberall F, Giselbrecht S, Hellbert K, Baier G, Grunicke HH.Characterization of PKC isozyme specific functions in cellular signaling.Adv Enzyme Regul. 1998;38:35–48.

32. Mizukami Y, Kobayashi S, Uberall F, Hellbert K, Kobayashi N, YoshidaK. Nuclear mitogen-activated protein kinase activation by PKC� duringreoxygenation after ischemic hypoxia. J Biol Chem. 2000;275:19921–19927.




and Asrar B. MalikKamran Javaid, Arshad Rahman, Khandaker N. Anwar, Randall S. Frey, Richard D. Minshall

Dependent Activation of Intercellular Adhesion Molecule-1−ζC Induces Early-Onset Endothelial Adhesivity by Protein KinaseαTumor Necrosis Factor-

Print ISSN: 0009-7330. Online ISSN: 1524-4571 Copyright © 2003 American Heart Association, Inc. All rights reserved.is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation Research

doi: 10.1161/01.RES.0000072971.88704.CB2003;92:1089-1097; originally published online April 24, 2003;Circ Res.

http://circres.ahajournals.org/content/92/10/1089World Wide Web at:

The online version of this article, along with updated information and services, is located on the

http://circres.ahajournals.org//subscriptions/

is online at: Circulation Research Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document. Permissions and Rights Question and Answer about this process is available in the

located, click Request Permissions in the middle column of the Web page under Services. Further informationEditorial Office. Once the online version of the published article for which permission is being requested is

can be obtained via RightsLink, a service of the Copyright Clearance Center, not theCirculation Researchin Requests for permissions to reproduce figures, tables, or portions of articles originally publishedPermissions:


http://circres.ahajournals.org/content/92/10/1089

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://circres.ahajournals.org//subscriptions/


Tumor Necrosis Factor Induces Early-Onset Endothelial Adhesivity by Protein Kinase C-Dependent...

Documents

Transcript of Tumor Necrosis Factor Induces Early-Onset Endothelial Adhesivity by Protein Kinase C-Dependent...