ISSN: 1524-4636 Copyright © 2011 American Heart Association. All rights reserved. Print ISSN: 1079-5642. Online

7272 Greenville Avenue, Dallas, TX 72514Arteriosclerosis, Thrombosis, and Vascular Biology is published by the American Heart Association.

DOI: 10.1161/ATVBAHA.110.216044 10, 2011;

2011;31;1169-1176; originally published online FebArterioscler Thromb Vasc BiolNg, Simon Siu Man Ng, Maik Gollasch, Xiaoqiang Yao and Yu Huang

Siu Ling Wong, Chi Wai Lau, Wing Tak Wong, Aimin Xu, Chak Leung Au, Chi FaiCyclooxygenase-2 Expression: A Link to Vascular Inflammation

Pivotal Role of Protein Kinase C{delta} in Angiotensin II-Induced Endothelial

http://atvb.ahajournals.org/cgi/content/full/ATVBAHA.110.216044/DC1Data Supplement (unedited) at:

  http://atvb.ahajournals.org/cgi/content/full/31/5/1169

located on the World Wide Web at: The online version of this article, along with updated information and services, is

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

journalpermissions@lww.com410-528-8550. E-mail:

Fax:Kluwer Health, 351 West Camden Street, Baltimore, MD 21202-2436. Phone: 410-528-4050. Permissions: Permissions & Rights Desk, Lippincott Williams & Wilkins, a division of Wolters 

http://atvb.ahajournals.org/subscriptions/Biology is online at Subscriptions: Information about subscribing to Arteriosclerosis, Thrombosis, and Vascular

at Chinese University of Hong Kong on April 22, 2011 atvb.ahajournals.orgDownloaded from

Pivotal Role of Protein Kinase C� in Angiotensin II–InducedEndothelial Cyclooxygenase-2 Expression

A Link to Vascular Inflammation

Siu Ling Wong, Chi Wai Lau, Wing Tak Wong, Aimin Xu, Chak Leung Au, Chi Fai Ng,Simon Siu Man Ng, Maik Gollasch, Xiaoqiang Yao, Yu Huang

Objective—The purpose of this study was to examine the hypothesis that angiotensin II (Ang II) induced endothelialcyclooxygenase-2 (COX-2) expression, which in turn mediated the generation of proinflammatory cytokines.

Methods and Results—Western blot analysis on primary rat endothelial cells showed Ang II induced COX-2 expression,which was abolished by cotreatment of p38 mitogen-activated protein kinase (SB 202190) and extracellularsignal–regulated kinase 1/2 (PD 98059) inhibitors. Protein kinase C� (PKC�) inhibitor (rottlerin) prevented extracellularsignal–regulated kinase 1/2 phosphorylation and COX-2 expression. The pivotal role of PKC� was further supported bya similar stimulatory effect of the PKC activator on COX-2 expression, signified by Ang II–stimulated translocation ofPKC� to the plasma membrane, and confirmed by PKC� phosphorylation at Tyr311. Small interfering RNA targetingPKC� diminished COX-2 expression, which was further abrogated by SB 202190. Human mesenteric arteries incubatedwith Ang II showed increased levels of endothelial COX-2 and monocyte chemoattractant protein-1; the former wasinhibited by SB 202190 plus rottlerin, whereas the latter was prevented by COX-2 inhibitor.

Conclusion—The present study pinpoints a novel role of PKC� in Ang II–induced endothelial COX-2 upregulation andidentifies a COX-2-dependent proatherosclerotic cytokine monocyte chemoattractant protein-1. The findings raise thepossibility of curtailing endothelial COX-2 expression as a means of limiting or preventing vascular inflammation.(Arterioscler Thromb Vasc Biol. 2011;31:1169-1176.)

Key Words: angiotensin II � endothelium � cyclooxygenase-2 � monocyte chemoattractant protein-1 � protein kinase C�

Angiotensin II (Ang II), the most prominent vasoactivepeptide in the renin-angiotensin system, causes vascular

dysfunction through an exaggerated production of reactiveoxygen species (ROS) and vascular hypercontractility throughstimulation of Ang II type 1 receptor (AT1R).1 Ang II is also apotent inducer of the expression of proinflammatory cytokines,such as tumor necrosis factor-�, and adhesion molecules such asvascular cell adhesion molecule-1.2–4 These factors are prereq-uisite for the initiation of atherosclerosis. In addition, Ang IIactivates matrix metalloproteinases, thereby promoting cell mi-gration and adverse vascular remodeling.5 Ang II is thus closelyassociated with diseases accompanied by vascular inflammation.

Cyclooxygenase-2 (COX-2) is minimally expressed inhealthy vascular tissues, but it is highly inducible on stimulationby growth factors, proinflammatory cytokines, and bacterialtoxins.6–8 A marked upregulation of COX-2 is reported ininflamed vascular tissues,9 in vascular remodeling of wire-injured mouse femoral arteries,10 and in human atherosclerotic

plaques.11,12 The expression of COX-2, prostanoid synthases,and prostaglandin receptors are upregulated in blood mononu-clear cells and plaques of patients with carotid atherosclerosis.11

The plasma level of prostaglandin E2, known to activate matrixmetalloproteinases, is augmented in patients with atherosclero-sis.11,12 Recent studies have shown that nonsteroidal antiinflam-matory drugs can reduce vascular inflammation3 and thatCOX-2 inhibition is beneficial in decreasing adhesion moleculeexpression in cancer cell lines.13

Both Ang II and COX-2 are associated with vascular inflam-mation and remodeling. However, it remains to be exploredwhether COX-2 could play a direct role as a downstreameffector in mediating Ang II–induced vascular pathogenesis. Thealtered endothelial cell function is the key contributor to vascularinflammation. In the present study, we aimed to test thehypothesis that Ang II could induce COX-2 expression inendothelial cells, which in turn is related to generation ofproinflammatory cytokines.

Received on: September 7, 2010; final version accepted on: January 31, 2011.From the Institute of Vascular Medicine, Li Ka Shing Institute of Health Sciences and School of Biomedical Sciences (S.L.W., C.W.L., W.T.W.,

C.L.A., X.Y., Y.H.) and Department of Surgery (C.F.N., S.S.M.N.), Chinese University of Hong Kong, Hong Kong Special Administrative Region,China; Department of Medicine and Department of Pharmacology and Pharmacy, University of Hong Kong, Hong Kong Special Administrative Region,China (A.X.); Medical Clinic for Nephrology and Internal Intensive Care, Charite University Medicine Berlin, Berlin, Germany (M.G.).

Correspondence to Yu Huang, PhD, School of Biomedical Sciences, Chinese University of Hong Kong, Shatin, N.T., Hong Kong SpecialAdministrative Region, China (E-mail yu-huang@cuhk.edu.hk); or Siu Ling Wong, PhD, School of Biomedical Sciences, Chinese University of HongKong, Hong Kong Special Administrative Region, China (E-mail christine_wong@cuhk.edu.hk).

© 2011 American Heart Association, Inc.

Arterioscler Thromb Vasc Biol is available at http://atvb.ahajournals.org DOI: 10.1161/ATVBAHA.110.216044

1169 at Chinese University of Hong Kong on April 22, 2011 atvb.ahajournals.orgDownloaded from

MethodsA Supplemental Methods section is available online athttp://atvb.ahajournals.org.

Cell Culture and TreatmentEndothelial cells were freshly cultured from rat thoracic aortae.14

The identity of the cells was confirmed by positive staining for anendothelial cell–specific marker, platelet endothelial cell adhesionmolecule-1 (Supplemental Figure I). On confluence, cells were serumdeprived for 24 hours and incubated with Ang II (100 nmol/L) for 8hours unless otherwise stated. When used, inhibitors were preincubatedwith the cells for 30 minutes before the addition of Ang II.

Western Blot Analysis and ReverseTranscription–Polymerase Chain ReactionProtein expressions and phosphorylation levels of COX-2, COX-1,p38 mitogen-activated protein kinase (MAPK), extracellular signal–regulated kinase 1/2 (ERK1/2), and protein kinase C (PKC) isoformswere determined by Western blotting. COX-2 mRNA stabilizationwas examined by reverse transcription–polymerase chain reaction.

Cellular Protein Fractionation for TranslocationStudy of PKC IsoformsEndothelial cells were incubated with 100 nmol/L Ang II for 1minute and quickly cooled on ice to quench cellular reactions.Cytosolic and membranous proteins were extracted with the Proteo-Extract Subcellular Proteome Extraction Kit (Calbiochem). Nuclearproteins were extracted with custom-made high-salt buffer. Equalamount of proteins were subjected to SDS-PAGE.

PKC� Knockdown With Small Interfering RNACells were transfected with either scramble siRNA or predesignedspecific small interfering RNA (siRNA) targeting PKC� transcriptsto detect COX-2 expression after Ang II stimulation.

Suspension Antibody Array-BasedMultiplex ImmunoassayConditioned medium from endothelial cells treated with Ang II (100nmol/L) for 24 hours were harvested, and the levels of interleukin-6,tumor necrosis factor-� and monocyte chemoattractant protein-1(MCP-1) were measured with MILLIPLEX MAP rodent cytokine/chemokine panel (Millipore) using the Bio-plex Suspension ArraySystem (Bio-Rad), according to the manufacturers’ instructions.

Ang II Infusion in RatsSprague-Dawley rats were infused with Ang II at 0.7 mg/kg per dayor vehicle (normal saline) for 9 days using an osmotic pump. Aortaeand renal arteries were preserved for immunohistochemical staining.Interlobal renal arteries were used for functional evaluation ofendothelium-dependent relaxations.

Tissue Culture of Human SmallMesenteric ArteriesHuman small mesenteric arteries were treated with Ang II(1 �mol/L) for 24 hours in Dulbecco’s modified Eagle’s mediumsupplemented with 10% fetal bovine serum and 1% penicillin/streptomycin at 37°C. Arteries were preserved for cryosectioningand immunofluorescence localization of COX-2 and MCP-1.

Immunohistochemical Staining andImmunofluorescence LocalizationLocalization of COX-2 in the aortae and renal arteries from Ang II–infusedrats was determined by immunohistochemistry. Human small mesentericarteries harvested after a 24-hour incubation protocol were preserved forimmunofluorescence to detect COX-2 and MCP-1 expression.

Functional Examination With MyographyEndothelium-dependent relaxations of the interlobal renal arteriesfrom Ang II–infused rats were determined in myographs.

ROS Detection WithDihydroethidium FluorescenceProduction of intracellular ROS was measured with dihydroethidium(Molecular Probes).

Data AnalysisResults represent means�SEM of 5 to 6 separate experiments. Statis-tical significance was determined by 2-tailed Student t test or 1-wayANOVA followed by Bonferroni post tests when more than 2 treat-ments were compared (GraphPad Software, San Diego, CA). A proba-bility value of less than 0.05 was regarded as statistically different.

ResultsAng II Induces COX-2 Expression inEndothelial CellsCOX-2 expression was undetectable in the untreated quies-cent endothelial cells and was not induced by culturing inserum-deprived medium. Ang II at 100 nmol/L increasedCOX-2 expression, reaching a maximum after an 8-hourincubation, and this effect was concentration dependent (3 to100 nmol/L) (Supplemental Figure IIA and IIB). Treatmentwith actinomycin-D (10 �mol/L), an inhibitor of RNAsynthesis, prevented Ang II–induced COX-2 expression(Supplemental Figure IIC) without affecting COX-1 expres-sion (Supplemental Figure IIIC). By contrast, COX-1 wasconstitutively expressed in endothelial cells, and its level wasunaffected by Ang II (Supplemental Figure IIIA and IIIB).

Ang II Upregulates COX-2 Expression via AT1RTreatment with losartan (3 �mol/L), an AT1R blocker, abolishedAng II–induced COX-2 expression, whereas PD 123319 (1 �mol/L), an AT2R blocker, was without effect (Supplemental Figure IIC).

p38 Mitogen-Activated Protein Kinase andERK1/2 Jointly Mediate Ang II–InducedCOX-2 ExpressionOf the 3 well-known kinase pathways, only p38 MAPK inhibitor(SB 202190, 10 �mol/L) and ERK1/2 inhibitor (PD 98059,20 �mol/L) reduced Ang II–induced COX-2 expression, each by�50%, whereas SP 600125 (10 �mol/L), a c-Jun-N-terminalkinase inhibitor, had no effect. Cotreatment with SB 202190 andPD 98059 produced additive inhibition (Figure 1A), suggesting aparallel involvement of p38 MAPK and ERK1/2 in mediating AngII–induced COX-2 expression. By contrast, these inhibitors did notalter the expression of COX-1 (Supplemental Figure IIID).

When the time course of ERK1/2 and p38 MAPK activa-tion by Ang II was examined, peak phosphorylation of bothkinases was noted �5 minutes after the addition of 100nmol/L Ang II (Figure 1B and 1C).

PKC Activation Is Upstream of ERK1/2 inCOX-2 InductionTreatment with GF 109203X (GFX, 2 �mol/L), an inhibitor fora broad spectrum of PKC isoforms, markedly reduced AngII–induced COX-2 expression. Rottlerin (PKC� inhibitor,10 �mol/L) but not Go 6976 (PKC�/� inhibitor, 1 �mol/L)

1170 Arterioscler Thromb Vasc Biol May 2011

at Chinese University of Hong Kong on April 22, 2011 atvb.ahajournals.orgDownloaded from

inhibited COX-2 expression (Figure 1D). These inhibitors didnot affect COX-1 expression (Supplemental Figure IIIE).

Next, the sequence of events involving PKC, ERK1/2, andp38 MAPK on Ang II stimulation was determined. PKCinhibition would be expected to prevent the activation ofERK1/2 and p38 MAPK if PKC is the upstream regulator of the 2latter pathways. ERK1/2 phosphorylation was inhibited by GFXand rottlerin but not Go 6976, with PD 98059 serving as a positivecontrol for ERK1/2 inhibition (Figure 1E). By contrast, p38 MAPKphosphorylation was prevented by losartan but unaltered by GFX,Go 6976, and rottlerin (Figure 1F), indicating that unlike ERK1/2,activation of p38 MAPK was independent of PKC� regulation.

Reverse transcription–polymerase chain reaction on cells preex-posed to Ang II for 4 hours and then treated with actinomycin-Drevealed the participation of these kinases in maintaining COX-2mRNA stability. Posttreatments with either SB 202190, PD 98059,or rottlerin partially reduced COX-2 mRNA levels, whereas thecombined treatments of SB 202190�PD 98059 and SB202190�rottlerin did not result in synergistic reduction in the levelsof transcribed COX-2 mRNA (Supplemental Figure IID), implyingthat unlike induction, these kinases may act in series to maintainCOX-2 mRNA stability posttranscriptionally.

PKC� Plays a Key Role With Little Involvementof Other PKC IsoformsStudies with the pharmacological inhibitor of PKC�, rottlerin,implied the participation of PKC� in Ang II–induced COX-2

expression. To confirm the involvement of PKC�, a time-dependentphosphorylation of PKC� at the activation sites was determined.PKC� was phosphorylated at Tyr311 in less than 1 minute after theaddition of Ang II (Figure 2A). In contrast, residues Thr505 andTyr332 were found to be nonphosphorylated throughout the 1-hourtime course studied (Supplemental Figure IV).

Cytoplasm to membrane translocation of various PKC iso-forms was examined by fractionating the cytosolic and mem-branous portion of total protein from endothelial cells. Thismethod was validated by a clear detection of large amount ofcytosolic GAPDH in the cytosol but a minimal amount in themembranous fraction, and vice versa for membrane-boundNOX-2 subunit of NADPH oxidase (Supplemental Figure VAand VB). The results showed that PKC� and PKC� wereactivated on 1-minute stimulation with Ang II, which wassignified by their increased levels in membranous fractions. Bycontrast, �, �, and � isoforms were not increased in themembranous portions of Ang II–stimulated cells compared withthat of control. PKC� was minimally expressed in endothelialcells (Figure 2B, Supplemental Figure V). PKC� translocationwas abolished by losartan (Figure 2C). SB 202190 and PD98059 failed to inhibit Ang II–induced PKC� translocation(Supplemental Figure VE), again confirming the notion thatERK1/2 can only be a downstream target of PKC�, but not viceversa, whereas p38 MAPK and PKC� are independent of eachother. Notably, PKC� translocation from cytosol to nucleus was

Figure 1. Ang II–induced COX-2 expression is mediated through ERK1/2, p38 MAPK, and PKC�. A, COX-2 expression was inhibited bySB 202190 (SB, 10 �mol/L) or PD 98059 (PD, 20 �mol/L) but not SP 600125 (SP, 10 �mol/L). Cotreatment with SB and PD furtherreduced the COX-2 expression. B and C, Time course for phosphorylation of ERK1/2 and p38 MAPK in response to Ang II (100 nmol/L). D,COX-2 expression was inhibited by GF 109203X (GFX, 2 �mol/L) or rottlerin (10 �mol/L) but not Go 6976 (1 �mol/L). E, ERK1/2 phosphoryla-tion was inhibited by GFX, rottlerin, or PD 98059 (PD) but not Go 6976. F, Phosphorylation of p38 MAPK was inhibited by losartan (3 �mol/L)but not PKC inhibitors. *P�0.05, **P�0.01, ***P�0.001 vs control; ###P�0.001 vs the Ang II–treated group.

Wong et al Ang II and Endothelial COX-2 Upregulation 1171

at Chinese University of Hong Kong on April 22, 2011 atvb.ahajournals.orgDownloaded from

not detected shortly after Ang II exposure (Supplemental FigureVIA; fractionation method was validated by strong positivesignals of GAPDH in the cytosolic portion and histone H3 in thenuclear portion, as shown in Supplemental Figure VIB andVIC), indicating that the action of PKC� mainly resided in thecytosol at its peak activation state. Cotreatment with SB 202190and rottlerin produced additive effects in suppressing the AngII–induced COX-2 expression, without affecting COX-1 expres-sion (Supplemental Figure VIIA and VIIB).

Transfection with siRNA targeting PKC� (siPKC�) furthersupports the key role of PKC�. Western blot analysis con-firmed that siPKC� successfully knocked down PKC� withoutaffecting PKC� (Figure 2D; Supplemental Figure VIIIA andVIIIB). In siPKC�-transfected cells, Ang II–induced COX-2expression was reduced by �50%, and the remaining portioncould almost be abolished by cotreatment with SB 202190,again indicating the operation of dual signal transductionpathways involving PKC� and p38 MAPK. By contrast,scramble siRNA had no effect on Ang II–induced COX-2expression compared with the nontransfected endothelialcells (Figure 2D, Supplemental Figure VIIIC and VIIID).

As shown in the study of PKC translocation, PKC� was alsoactivated by Ang II (Figure 2B, Supplemental Figure VF).However, the PKC� inhibitor peptide �V1-2 (10 �mol/L) did notprevent Ang II–induced COX-2 expression (Supplemental Fig-ure VIIC). siPKC� did not suppress PKC� levels (Figure 2D,Supplemental Figure VIIIB) but abolished COX-2 expression incombination with the p38 MAPK inhibitor SB 202190 (Figure2D, Supplemental Figure VIIID), thus suggesting that PKC� isunlikely to be one of the mediators of COX-2 expression.

The inducible role of PKC� in COX-2 expression was alsopinpointed using an exogenous PKC activator, phorbol 12-my-ristate 13-acetate (PMA). PMA at 1 �mol/L time-dependentlyincreased COX-2 expression, whereas its negative analog, 4�-PMA, had no effect (Figure 2E, Supplemental Figure VIIIE).Such increase in COX-2 expression was again sensitive toinhibition by GFX and rottlerin but not Go 6976 (Figure 2E,Supplemental Figure VIIIF).

Ang II–Induced Release of MCP-1 IsCOX-2 DependentThe level of MCP-1 in the conditioned medium from endo-thelial cells treated with 100 nmol/L Ang II for 24 hoursincreased (Figure 3), whereas the release of interleukin-6 andtumor necrosis factor-� fell below detectable levels (data notshown). The MCP-1 release was inhibited by losartan and 2structurally different COX-2 inhibitors, celecoxib and NS 398(both at 3 �mol/L) (Figure 3).

Renovascular Dysfunction and ElevatedEndothelial COX-2 Expression in AngII–Infused RatsSystolic blood pressure of rats rose from 104.3�4.6 to174.0�9.6 mm Hg during a 9-day period of Ang II infusion,and this increase was prevented by concomitant oral treat-ment with losartan (Figure 4A). Endothelium-dependent re-laxations were impaired in interlobal renal arteries from theAng II–infused rats, and this impairment was abolished bylosartan treatment (Figure 4B). The attenuated relaxationswere restored by acute treatment with the COX-2 inhibitor

Figure 2. PKC� is activated by Ang IIstimulation, acting as a major mediator ofCOX-2 expression. A, PKC� was phos-phorylated at Tyr311 within 1 minute ofAng II addition. B, Representative immu-noblots of translocation of various PKCisoforms stimulated by Ang II (100nmol/L) and PMA (1 �mol/L). C indicatescytosol; M, membrane. C, Ang II stimu-lated PKC� translocation to the mem-brane, which was prevented by losartan(3 �mol/L). Cyto indicates cytosol; mb,membrane. D, siRNA targeting PKC�

(siPKC�) significantly reduced Ang II–in-duced COX-2 expression, which was fur-ther abolished by SB 202190. E, PMAinduced COX-2 expression, which was inhib-ited by GFX (2 �mol/L) and rottlerin (10 �mol/L). **P�0.01, ***P�0.001 vs control;##P�0.01 vs the Ang II–treated group.

Figure 3. MCP-1 production as a consequence of COX-2 upregula-tion. MCP-1 release increased after a 24-hour incubation with Ang II(100 nmol/L), which was prevented by losartan (3 �mol/L) and theCOX-2 inhibitors celecoxib and NS 398 (each at 3 �mol/L).***P�0.001 vs control; ##P�0.01 vs the Ang II–treated group.

1172 Arterioscler Thromb Vasc Biol May 2011

at Chinese University of Hong Kong on April 22, 2011 atvb.ahajournals.orgDownloaded from

celecoxib (3 �mol/L) (Figure 4C) but not by the COX-1inhibitor sc-560 (data not shown). Immunohistochemicalstaining revealed that endothelial COX-2 expression (Figure4D and 4E, arrows) was augmented in both aortae and renalarteries of Ang II–infused rats, and the increased COX-2staining was attenuated by losartan.

Ang II–Induced COX-2 Expression andCOX-2-Dependent MCP-1 Expression in HumanMesenteric ArteriesAng II–induced expression of proinflammatory MCP-1 wasalso demonstrated in the endothelial layer of human mesen-teric arteries. The left columns of Figure 5A and 5B show thegreen autofluorescence from elastin of the internal elasticlamina, which delineated the artery into endothelial andsmooth muscle layers. The middle columns feature thefluorescence emitted from Alexa Fluor 546–conjugated sec-ondary antibodies, which were tagged to primary antibodiesagainst either COX-2 (Figure 5A) or MCP-1 (Figure 5B),appearing in reddish orange together with the noise fromautofluorescence. In the overlay images in the right column,the autofluorescence and noise are yellowish green, and theremaining reddish orange signifies the signals from COX-2 orMCP-1. Referring to the right column of Figure 5A, endo-thelial COX-2 expression was remarkably augmented after a24-hour exposure to Ang II (1 �mol/L); expression wasprevented by cotreatment with SB 202190 and rottlerin butremained unaffected in solvent control exposed to dimethylsulfoxide. It is noteworthy that this COX-2 expression maycontribute to the MCP-1 expression, as evidenced by an increase

of endothelial MCP-1 expression in the same arteries treatedwith Ang II, which was inhibited by preincubation with cele-coxib but not the solvent dimethyl sulfoxide (Figure 5B).

ROS Do Not Mediate COX-2 UpregulationDihydroethidium fluorescence in endothelial cells showedthat the ROS level rose �2-fold in �30 s after Ang IIaddition, which was inhibited by losartan, the NADPHoxidase inhibitor apocynin (100 �mol/L), and ROS scaven-gers, tiron (1 mmol/L) plus diethyldithiocarbamate(100 �mol/L) (Supplemental Figures IX and XA). Inhibitorsof ROS-producing enzymes, including apocynin (100 �mol/L), oxypurinol (xanthine oxidase inhibitor, 100 �mol/L),2-amino-5,6-dihydro-6-methyl-4H-1,3-thiazine hydrochlo-ride (inducible nitric oxide synthase inhibitor, 30 �mol/L),NG-nitro-L-arginine methyl ester (nonspecific nitric oxidesynthase inhibitor, 100 �mol/L), the ROS scavengers tempol,and tiron plus diethyldithiocarbamate, however, did notdecrease the Ang II–induced COX-2 expression (Supplemen-tal Figure XB), thus ruling out the participation from ROS.Indeed, exogenous H2O2 induced only a small increase(�3-fold) in COX-2 expression compared with the markedstimulation (�15-fold) by Ang II (Supplemental Figure XC).

DiscussionThe present study provides novel insights on the molecularmechanism and sequence of intracellular events on theupregulation of endothelial COX-2 expression by Ang II,pinpointing a novel role of PKC� activated at Tyr311 andjoint contributions of ERK1/2 and p38 MAPK. We alsodefined the Ang II–induced release or expression of the

Figure 4. COX-2 upregulation and vascular dysfunction in AngII–infused hypertensive rats. A, Systolic blood pressure increasedmarkedly in Ang II–infused rats, which was prevented by oraladministration of losartan (Losar, 10 mg/kg per day). B, Acetylcho-line (ACh)–induced endothelium-dependent relaxations were atten-uated in the interlobal renal arteries from Ang II–infused rats butnot in those orally administered losartan. C, Celecoxib acutelyrestored the impaired endothelium-dependent relaxations. D and E,Endothelial COX-2 expression (indicated by red arrows) was aug-mented in the aortae (D) and renal arteries (E) of Ang II–infusedrats (AII), and the expression was reduced in those from AngII–infused rats orally treated with losartan (L�AII). Similar obser-vations were each made in arteries from 4 rats. *P�0.05,**P�0.01 vs control; #P�0.05, ##P�0.01 vs the Ang II–infusedgroup. CTL indicates control; Endo, endothelium.

Figure 5. Ang II induces COX-2 expression and COX-2-dependentMCP-1 expression in human mesenteric arteries. Autofluorescence(Auto, left columns, in green) signified the elastin from internal elasticlamina (IEL) in the vessel wall. It delineated the layers of endothelialcells (EC) and vascular smooth muscle cells (VSMC). Detection forAlexa Fluor 546 (AF546, middle columns, in reddish orange) gave thesignals from both the elastin and the antibody-fluorophore complex.Right columns feature the overlay images of the other columns, withthe elastin appearing yellow. The reddish orange color in the right col-umn indicates the pure signal from antibodies. A, Ang II induced en-dothelial COX-2 expression (indicated by white arrows), which wassensitive to SB 202190 (SB) and rottlerin (ROT). B, Ang II elevatedendothelial MCP-1 expression, which was inhibited by celecoxib.dimethyl sulfoxide did not affect the expressions of either Ang II–in-duced COX-2 or MCP-1. When primary antibody was neglected (ie,negative control), the signals were purely from the elastin.

Wong et al Ang II and Endothelial COX-2 Upregulation 1173

at Chinese University of Hong Kong on April 22, 2011 atvb.ahajournals.orgDownloaded from

COX-2-dependent proatherosclerotic cytokine MCP-1. Thepresent results were first obtained from primary culture of ratendothelial cells, but they were also confirmed in an in vivoanimal model of Ang II–infused hypertensive rats and ex-tended to tissue culture of human arteries.

Ang II, a potent vasoconstrictive peptide, plays a central rolein vascular dysfunction associated with hypertension and diabe-tes, and AT1R blockers represent a major class of antihyperten-sive drugs that interrupt Ang II–initiated intracellular signaling.15

Although Ang II serves as a powerful ligand to trigger transcrip-tions of proatherosclerotic cytokines,2–4 the role of the highlyinducible proinflammatory enzyme COX-2 in vascular diseaseshas recently gained attention, as its overexpression is alsodetected in human atherosclerotic plaques.11 However, the rela-tionship between Ang II and COX-2 remains elusive as towhether COX-2 acts as a direct downstream effector of Ang IIto mediate vascular inflammation, especially in endothelial cells,which represent the first-line initiation point of atherosclerosis.The present study demonstrated that Ang II could markedlyupregulate COX-2 expression in endothelial cells through AT1Ractivation, agreeing with the preliminary findings of Kramer etal using Northern blot analysis.16 The COX-2 expression wasjointly mediated by p38 MAPK- and ERK1/2-dependent signal-ing pathways, as the inhibitor of either kinase, SB 202190 or PD98059, could only partially suppress the COX-2 expression,whereas their cotreatment resulted in additive inhibition thatalmost abolished COX-2 expression. It appeared that c-Jun-N-terminal kinase was unlikely to be involved, as its inhibitor didnot alter Ang II–induced COX-2 expression.

Because p38 MAPK and ERK1/2 are sensitive to redoxtriggers,17–19 we initially suspected that Ang II–induced oxida-tive stress might contribute to COX-2 expression. This possibil-ity, however, was excluded by 2 observations. First, AngII–stimulated ROS production was prevented by the NADPHoxidase inhibitor apocynin and by the ROS scavenger tiron plusdiethyldithiocarbamate; however, Ang II–induced COX-2 ex-pression was unaltered by these inhibitors and by inhibitorsagainst xanthine oxidase, inducible nitric oxide synthase, andnitric oxide synthase uncoupling. Second, H2O2, which wasreported to increase COX-2 expression in human umbilical veinendothelial cells,20 induced a mild increase in COX-2 expressioncompared with that stimulated by Ang II. Our results differ froma previous study showing that in human saphenous vein endo-thelial cells, interleukin-1�-induced COX-2 expression involvesboth NADPH oxidase-derived superoxide anion and H2O2.21

Indeed, our recent findings on bone morphogenic protein-4–induced COX-2 expression also showed that the kinase mediatorp38 MAPK is activated by bone morphogenic protein-4–stimulated NADPH oxidase-derived ROS.22 Our further exper-imentation, aimed at understanding why oxidative stress isinvolved in bone morphogenic protein-4–induced COX-2 in-duction but not that stimulated by Ang II, showed that bonemorphogenic protein-4 (20 ng/mL) induced a significantlyhigher ROS level compared with Ang II at 100 nmol/L (theconcentration used in the cell-based studies) or even 1 �mol/L(Supplemental Figure XI), implying that the level of oxidativestress induced by the agonist is an important determinant inwhether the redox-sensitive kinases are turned on by the radicalsand hence whether there is a role for the ROS to mediate COX-2

upregulation. The MAPKs are not preferentially activated by theAng II–induced ROS, possibly owing to the ineffective ROSlevel. Alternatively, the source and production site of ROS ortype of ROS produced may play differential roles in bothinduction of COX-2 expression and stimulation of COX-2activity. These need further investigation at subcellular levels.

PKC appears to be another possible upstream target becausePKC is associated with the G-protein-coupled AT1R. Herein wedefined a novel role of PKC�, activated at Tyr311, to mediateCOX-2 expression in Ang II–treated rat aortic endothelial cells.To our surprise, PKC� is not phosphorylated at the activationloop of residue Thr505 in response to Ang II. As pointed out byLiu et al, phosphorylation of the activation loop may not benecessary for PKC� activation and that PKC�-mediated apopto-sis of HEK293T cells does not require phosphorylation at theactivation loop.23 Indeed, earlier reports also found that mutantPKC� with activation loop dephosphorylated was still in fullactivity.24–26 As to the tyrosine residues, Tyr311 but not Tyr332was phosphorylated by Ang II, although both are concomitantlyphosphorylated in response to oxidative stress such as H2O2.27

As explained by Rybin et al, the agonist (Ang II or PMA)-dependent selective phosphorylation of tyrosine residues may pos-sibly be effected via the selective Tyr311 kinase c-Abl, whereasother kinases (sensitive to redox changes) may cause a concurrentphosphorylation in both residues.27 Although we cannot detail allthe phosphorylation residues that are involved in the activation ofPKC� in response to Ang II (because of a large number of possiblephosphorylation sites and the complex regulation of PKC�), ourpresent data suggest that at least Tyr311 is phosphorylated tomediate Ang II–induced PKC� activation, without the possibleinvolvement of residues Thr505 and Tyr332.

Pharmacologically, Ang II–induced COX-2 expression andPMA (PKC activator)-induced COX-2 expression were bothsensitive to inhibition by GFX (a broad spectrum of PKCinhibitor) and rottlerin (PKC� inhibitor). Of note, GFX androttlerin also inhibited Ang II–induced ERK1/2 phosphorylation,indicating that PKC� is likely the upstream activator of ERK1/2.Activation of p38 MAPK was likely independent of PKCregulation because its phosphorylation was unaffected by PKCinhibitors. Interestingly, posttreatments of these kinase inhibitorsalong with actinomycin-D in cells preexposed to Ang II dem-onstrated a partial reduction in Ang II–induced COX-2 mRNAlevels, indicating that p38 MAPK, ERK1/2, and PKC� may alsohelp to stabilize the COX-2 mRNA transcribed. In contrast to theresults from Western blot analysis in which cotreatments of SB202190�PD 98059 and SB 202190�rottlerin produced clearadditive effect that totally abolished Ang II–induced COX-2expression, neither SB 202190�PD 98059 nor SB 202190�rot-tlerin showed a significant synergistic suppression on the COX-2mRNA level, implying that the kinases may function in series inpreserving mRNA stability, which is different from COX-2induction, when they act jointly. Doller et al recently demon-strated in rat mesangial cells that PKC� take a critical role inshuttling the mRNA-stabilizing factor human antigen R andthereby posttranscriptionally maintaining COX-2 expressioninduced by Ang II.28 Of note, members of MAPK family canalso interact with human antigen R to stabilize the mRNAcoding for inflammatory mediators such as tumor necrosisfactor-� and COX-2.29,30 Ang II–induced COX-2 expression in

1174 Arterioscler Thromb Vasc Biol May 2011

at Chinese University of Hong Kong on April 22, 2011 atvb.ahajournals.orgDownloaded from

vascular smooth muscles is reported to be mediated by p38 MAPK-and ERK1/2-mediated mRNA stabilization of COX-2 mRNA.31

Nonetheless, taken together with the results from exogenous PKCactivator PMA which can induce rottlerin-sensitive COX-2 expres-sion, Ang II–stimulated activation of PKC� and the MAPKspossibly cause an upregulation of COX-2 expression by bothtranscription induction and mRNA stabilization.

To address the concern over the specificity of pharmacolog-ical inhibitors used, we studied the treatment-induced transloca-tion of PKC� from the cytoplasm to the plasma membrane andexcluded the participation of other PKC isoforms. Translocationof soluble cytosolic PKC to the particulated membranous form isa key feature that signifies PKC activation.32 Comparing theportion of membranous PKC after Ang II stimulation with thatof control, only PKC� and PKC� were found to be responsive toAng II. While PKC� was activated by PMA but not Ang II,PKC� and PKC� remained unresponsive to either treatments,and PKC� was not expressed in endothelial cells. Indeed,although PKC� was activated by PMA, treatments with thePKC�/� inhibitor Go 6976 failed to suppress the PMA-inducedCOX-2 expression, thus contrasting a complete abrogation byrottlerin. Likewise, Ang II activated both PKC� and PKC�, yetthe role of PKC� in mediating COX-2 expression appears to benegligible, as indicated by the following 2 observations. First,PKC� inhibitor �V1–2 did not affect Ang II–induced COX-2expression. Second, siPKC� markedly suppressed COX-2 ex-pression without affecting PKC� expression. If PKC� were to besignificantly involved in mediating Ang II action, its nonsup-pressed levels in siPKC�-transfected cells with or withoutconcomitant exposure to the p38 MAPK inhibitor SB 202190would have maintained the inducible response of COX-2 ex-pression as in the control or scramble siRNA-transfected cells.Taken together, these data indicate that even a number of PKCisoforms might be activated by Ang II PKC� is probably themain isoform that is essentially coupled with endothelial COX-2expression. It is noteworthy that inhibition of p38 MAPK by SB202190 in combination with PKC� knockdown cells preventedthe induction of COX-2 expression, thus proving a parallelinvolvement of the PKC�-ERK1/2 and p38 MAPK pathways.Such findings are in contrast to Ang II–induced COX-2 upregula-tion in vascular smooth muscle cells, which depends mainly on theERK1/2 activity without the participation of PKC or p38 MAPK.7

Pretreatment by inhibitor of either ERK1/2 or p38 MAPK did notaffect Ang II–induced PKC� translocation, again confirming thatPKC� is upstream of ERK1/2 activation and independent of p38MAPK regulation. Although PKC� can also move into the nucleusfrom the cytosol,33 our present results showed that PKC� is nottranslocated into the nucleus shortly (�1 minute) after Ang IIexposure, in contrast to its clear and rapid particulation at the plasmamembrane. Indeed, PKC� was found to be imported into thenucleus at least 30 minutes after Ang II stimulation.28 Our presentdata thus suggest that PKC�, shortly after Ang II stimulation, firsttranslocates to the membrane where it exerts its major effect onCOX-2 induction via regulation of ERK1/2 activity. COX-2mRNA stabilization by PKC� may occur as a consecutive eventfollowing its delayed translocation into the nucleus.

Upregulation of endothelial COX-2 by Ang II could be demon-strated not only in cultured cells. Indeed, our study clearly showedan augmentation of endothelial COX-2 protein expression in the

aorta and renal arteries of Ang II–infused rats, which was inhibitedby concomitant oral treatment of losartan. Impairment of endothe-lium-dependent relaxations was acutely restored by the COX-2inhibitor celecoxib, indicating that COX-2-derived vasoconstrictiveprostanoids may contribute to vascular dysfunctions, as documentedin diabetic, hypertensive, and aging animal models.34–36

Tissue culture experiments on human mesenteric arteriessubstantiate the findings in cultured rat endothelial cells. AngII–induced endothelial COX-2 expression was prevented bycotreatment of SB 202190 and rottlerin. More importantly,increased COX-2 expression was accompanied by a concomi-tant rise in MCP-1 expression, with both proteins being immu-nolocalized to the endothelium. This MCP-1 increase wassensitive to COX-2 inhibition by celecoxib, clearly suggesting acausal relationship between these 2 proinflammatory and proath-erosclerotic cytokines. Indeed, Ang II–induced MCP-1 release intothe conditioned medium of cultured rat endothelial cells was alsoprevented by the 2 specific COX-2 inhibitors celecoxib and NS 398.Taken together, endothelial COX-2-dependent release of MCP-1may represent a new concept in our understanding of the AngII–induced vascular inflammation and atherosclerotic plaque forma-tion involving macrophages. Which of the COX-2-derived prosta-noids may mediate MCP-1 release warrants further investigation.

Assembling the time course of maximal activation of signal-ing proteins, and studies using pharmacological inhibitors andsiRNA, we propose the following cellular pathways of AngII–induced endothelial COX-2 expression (Figure 6). Ang IIstimulates AT1R to activate PKC� in �1 minute and subse-quently ERK1/2 in �5 minutes, along with maximal PKC-independent p38 MAPK activation in �5 minutes. COX-2expression was effected via induction and mRNA stabilization;it was detected at 1 hour and reached the maximum at 8 hours.MCP-1 release detected at 24 hours was COX-2 dependent. Inview of the controversy regarding the use of COX-2 inhibitors

Figure 6. Postulated cellular mechanisms on Ang II–inducedCOX-2 expression in endothelial cells and time course ofevents. Ang II acts on AT1R, which then activates PKC� in �1minute. Activations of PKC�-dependent ERK1/2 and PKC�-independent p38 MAPK occur at �5 minutes. COX-2 begins tobe expressed at �1 hour via induction and mRNA stabilization,and the expression reaches a maximum at �8 hours. COX-2-dependent MCP-1 release was detected in 24 hours.

Wong et al Ang II and Endothelial COX-2 Upregulation 1175

at Chinese University of Hong Kong on April 22, 2011 atvb.ahajournals.orgDownloaded from

and considering that PKC� is preferentially activated in hyper-glycemia and diabetes (detailed in Significance of Study in theonline Supplement), the present study may suggest PKC� as amore promising target for therapeutic intervention in COX-2-mediated vascular complications.

Sources of FundingThis study was supported by the Hong Kong Research Grant Council(Grants CUHK465308M, CUHK466110, HKU2/07C, and HKU4/CRF10), Deutsche Forschungsgemeinschaft, Germany-Hong KongJoint Research Scheme, and Focused Investment Scheme of ChineseUniversity of Hong Kong.

DisclosuresNone.

References1. Landmesser U, Spiekermann S, Preuss C, Sorrentino S, Fischer D, Manes

C, Mueller M, Drexler H. Angiotensin II induces endothelial xanthineoxidase activation: role for endothelial dysfunction in patients withcoronary disease. Arterioscler Thromb Vasc Biol. 2007;27:943–948.

2. Cheng ZJ, Vapaatalo H, Mervaala E. Angiotensin II and vascular inflam-mation. Med Sci Monit. 2005;11:RA194–RA205.

3. Costanzo A, Moretti F, Burgio VL, Bravi C, Guido F, Levrero M, Puri PL.Endothelial activation by angiotensin II through NF�B and p38 pathways:involvement of NF�B-inducible kinase (NIK), free oxygen radicals, andselective inhibition by aspirin. J Cell Physiol. 2003;195:402–410.

4. Pueyo ME, Gonzalez W, Nicoletti A, Savoie F, Arnal JF, Michel JB.Angiotensin II stimulates endothelial vascular cell adhesion molecule-1via nuclear factor-�B activation induced by intracellular oxidative stress.Arterioscler Thromb Vasc Biol. 2000;20:645–651.

5. Yaghini FA, Zhang C, Parmentier JH, Estes AM, Jafari N, Schaefer SA,Malik KU. Contribution of arachidonic acid metabolites derived via cyto-chrome P4504A to angiotensin II-induced neointimal growth. Hypertension.2005;45:1182–1187.

6. Eligini S, Barbieri SS, Cavalca V, Camera M, Brambilla M, De FranceschiM, Tremoli E, Colli S. Diversity and similarity in signaling events leading torapid Cox-2 induction by tumor necrosis factor-� and phorbol ester in humanendothelial cells. Cardiovasc Res. 2005;65:683–693.

7. Hu ZW, Kerb R, Shi XY, Wei-Lavery T, Hoffman BB. Angiotensin IIincreases expression of cyclooxygenase-2: implications for the function ofvascular smooth muscle cells. J Pharmacol Exp Ther. 2002;303:563–573.

8. Rikitake Y, Hirata K, Kawashima S, Takeuchi S, Shimokawa Y, Kojima Y,Inoue N, Yokoyama M. Signaling mechanism underlying COX-2 induction bylysophosphatidylcholine. Biochem Biophys Res Commun. 2001;281:1291–1297.

9. Yogi A, Callera GE, Tostes R, Touyz RM. Bradykinin regulates calpain andproinflammatory signaling through TRPM7-sensitive pathways in vascularsmooth muscle cells. Am J Physiol Regul Integr Comp Physiol. 2009;296:R201–R207.

10. Ogawa M, Suzuki J, Hirata Y, Nagai R, Isobe M. A critical role of COX-2in the progression of neointimal formation after wire injury in mice.Expert Opin Ther Targets. 2009;13:505–511.

11. Gomez-Hernandez A, Martin-Ventura JL, Sanchez-Galan E, Vidal C, OrtegoM, Blanco-Colio LM, Ortega L, Tunon J, Egido J. Overexpression ofCOX-2, Prostaglandin E synthase-1 and prostaglandin E receptors in bloodmononuclear cells and plaque of patients with carotid atherosclerosis: regu-lation by nuclear factor-�B. Atherosclerosis. 2006;187:139–149.

12. Cipollone F, Prontera C, Pini B, Marini M, Fazia M, De Cesare D, Iezzi A,Ucchino S, Boccoli G, Saba V, Chiarelli F, Cuccurullo F, Mezzetti A.Overexpression of functionally coupled cyclooxygenase-2 and prostaglandinE synthase in symptomatic atherosclerotic plaques as a basis of prostaglandinE(2)-dependent plaque instability. Circulation. 2001;104:921–927.

13. Dianzani C, Brucato L, Gallicchio M, Rosa AC, Collino M, Fantozzi R.Celecoxib modulates adhesion of HT29 colon cancer cells to vascularendothelial cells by inhibiting ICAM-1 and VCAM-1 expression. Br JPharmacol. 2008;153:1153–1161.

14. Huang Y, Chan NW, Lau CW, Yao XQ, Chan FL, Chen ZY. Involvementof endothelium/nitric oxide in vasorelaxation induced by purified greentea (-)epicatechin. Biochim Biophys Acta. 1999;1427:322–328.

15. Schmieder RE. Mechanisms for the clinical benefits of angiotensin IIreceptor blockers. Am J Hypertens. 2005;18:720–730.

16. Kramer C, Sunkomat J, Witte J, Luchtefeld M, Walden M, Schmidt B, TsikasD, Boger RH, Forssmann WG, Drexler H, Schieffer B. Angiotensin IIreceptor-independent antiinflammatory and antiaggregatory properties of lo-sartan: role of the active metabolite EXP3179. Circ Res. 2002;90:770–776.

17. Hsu YH, Chen JJ, Chang NC, Chen CH, Liu JC, Chen TH, Jeng CJ, ChaoHH, Cheng TH. Role of reactive oxygen species-sensitive extracellularsignal-regulated kinase pathway in angiotensin II-induced endothelin-1gene expression in vascular endothelial cells. J Vasc Res. 2004;41:64–74.

18. Jaulmes A, Sansilvestri-Morel P, Rolland-Valognes G, Bernhardt F, Gaertner R,Lockhart BP, Cordi A, Wierzbicki M, Rupin A, Verbeuren TJ. Nox4 mediatesthe expression of plasminogen activator inhibitor-1 via p38 MAPK pathway incultured human endothelial cells. Thromb Res. 2009;124:439–446.

19. Matesanz N, Lafuente N, Azcutia V, Martin D, Cuadrado A, Nevado J,Rodriguez-Manas L, Sanchez-Ferrer CF, Peiro C. Xanthine oxidase-derivedextracellular superoxide anions stimulate activator protein 1 activity and hyper-trophy in human vascular smooth muscle via c-Jun N-terminal kinase and p38mitogen-activated protein kinases. J Hypertens. 2007;25:609–618.

20. Eligini S, Arenaz I, Barbieri SS, Faleri ML, Crisci M, Tremoli E, Colli S.Cyclooxygenase-2 mediates hydrogen peroxide-induced wound repair inhuman endothelial cells. Free Radic Biol Med. 2009;46:1428–1436.

21. Massaro M, Habib A, Lubrano L, Del Turco S, Lazzerini G, Bourcier T, WekslerBB, De Caterina R. The omega-3 fatty acid docosahexaenoate attenuates endo-thelial cyclooxygenase-2 induction through both NADP(H) oxidase and PKC�inhibition. Proc Natl Acad Sci U S A. 2006;103:15184–15189.

22. Wong WT, Tian XY, Chen Y, Leung FP, Liu L, Lee HK, Ng CF, Xu A, YaoX, Vanhoutte PM, Tipoe GL, Huang Y. Bone morphogenic protein-4 impairsendothelial function through oxidative stress-dependent cyclooxygenase-2upregulation: implications on hypertension. Circ Res. 2010;107:984–991.

23. Liu Y, Belkina NV, Graham C, Shaw S. Independence of protein kinaseC-� activity from activation loop phosphorylation: structural basis andaltered functions in cells. J Biol Chem. 2006;281:12102–12111.

24. Liu Y, Graham C, Li A, Fisher RJ, Shaw S. Phosphorylation of theprotein kinase C-theta activation loop and hydrophobic motif regulates itskinase activity, but only activation loop phosphorylation is critical to invivo nuclear-factor-�B induction. Biochem J. 2002;361:255–265.

25. Stempka L, Girod A, Muller HJ, Rincke G, Marks F, Gschwendt M, BossemeyerD. Phosphorylation of protein kinase C� (PKC�) at threonine 505 is not aprerequisite for enzymatic activity: expression of rat PKC� and an alanine 505mutant in bacteria in a functional form. J Biol Chem. 1997;272:6805–6811.

26. Stempka L, Schnolzer M, Radke S, Rincke G, Marks F, Gschwendt M.Requirements of protein kinase c� for catalytic function: role of glutamic acid500 and autophosphorylation on serine 643. J Biol Chem. 1999;274:8886–8892.

27. Rybin VO, Guo J, Gertsberg Z, Feinmark SJ, Steinberg SF. Phorbol12-myristate 13-acetate-dependent protein kinase C �-Tyr311 phosphor-ylation in cardiomyocyte caveolae. J Biol Chem. 2008;283:17777–17788.

28. Doller A, Gauer S, Sobkowiak E, Geiger H, Pfeilschifter J, Eberhardt W.Angiotensin II induces renal plasminogen activator inhibitor-1 and cycloox-ygenase-2 expression post-transcriptionally via activation of the mRNA-stabilizing factor human-antigen R. Am J Pathol. 2009;174:1252–1263.

29. Doller A, Pfeilschifter J, Eberhardt W. Signalling pathways regulatingnucleo-cytoplasmic shuttling of the mRNA-binding protein HuR. CellSignal. 2008;20:2165–2173.

30. Jin SH, Kim TI, Yang KM, Kim WH. Thalidomide destabilizes cyclo-oxygenase-2 mRNA by inhibiting p38 mitogen-activated protein kinaseand cytoplasmic shuttling of HuR. Eur J Pharmacol. 2007;558:14–20.

31. Ohnaka K, Numaguchi K, Yamakawa T, Inagami T. Induction of cyclo-oxygenase-2 by angiotensin II in cultured rat vascular smooth musclecells. Hypertension. 2000;35:68–75.

32. Salamanca DA, Khalil RA. Protein kinase C isoforms as specific targetsfor modulation of vascular smooth muscle function in hypertension.Biochem Pharmacol. 2005;70:1537–1547.

33. Humphries MJ, Ohm AM, Schaack J, Adwan TS, Reyland ME. Tyrosinephosphorylation regulates nuclear translocation of PKC�. Oncogene. 2008;27:3045–3053.

34. Wong SL, Leung FP, Lau CW, Au CL, Yung LM, Yao x, Chen ZY,Vanhoutte PM, Gollasch M, Huang Y. Cyclooxygenase-2-derived prosta-glandin F2� mediates endothelium-dependent contractions in the aortae ofhamsters with increased impact during aging. Circ Res. 2009;104:228–235.

35. Akamine EH, Urakawa TA, de Oliveira MA, Nigro D, de Carvalho MH, deCassia ATR, Fortes ZB. Decreased endothelium-dependent vasodilation indiabetic female rats: role of prostanoids. J Vasc Res. 2006;43:401–410.

36. Adeagbo AS, Zhang X, Patel D, Joshua IG, Wang Y, Sun X, Igbo IN, OriowoMA. Cyclo-oxygenase-2, endothelium and aortic reactivity during deoxycortico-sterone acetate salt-induced hypertension. J Hypertens. 2005;23:1025–1036.

1176 Arterioscler Thromb Vasc Biol May 2011

at Chinese University of Hong Kong on April 22, 2011 atvb.ahajournals.orgDownloaded from

Wong et al., 2011, Supplemental Material

P. 1

Pivoted Role of PKCδ in Angiotensin II-induced Endothelial Cyclooxygenase-2 Expression A Link to Vascular Inflammation

1Siu Ling Wong, 1Chi Wai Lau, 1Wing Tak Wong, 3Aimin Xu, 1Chak Leung Au, 2Chi Fai Ng, 2Simon Siu Man Ng,

4Maik Gollasch, 1Xiaoqiang Yao, 1Yu Huang

1Institute of Vascular Medicine, Li Ka Shing Institute of Health Sciences, and School of Biomedical Sciences, 2Department of Surgery, Chinese University of Hong Kong, HKSAR, China; 3Department of Medicine and Department of Pharmacology and Pharmacy, University of Hong Kong, HKSAR, China; 4Medical Clinic for

Nephrology and Internal Intensive Care, Charité University Medicine Berlin, Germany.

Supplemental Material

Methods This study was approved by the Animal Experimentation Ethics Committee, Chinese University of Hong Kong. Male Sprague Dawley rats (260-280 g) were supplied by Laboratory Animal Service Centre, Chinese University of Hong Kong. This investigation conformed to the Guide for the Care and Use of Laboratory Animals published by the US National Institute of Health (NIH Publication No. 85-23, revised 1996). The use of human samples for experimentation was approved by the Joint Chinese University of Hong Kong-New Territories East Cluster Clinical Research Ethics Committee. Human mesenteric arteries distant from the tumor were obtained during surgery for colon cancer with informed consent from patients. The mean age of the patients was 73 years old, ranging between 62 and 88. Only arteries from patients without hypertension (having systolic and diastolic blood pressure lower than 140 mmHg and 90 mmHg respectively) and without diabetes (having fasting glucose level <7.0 mmol/L) were employed in the present study. Cell culture Endothelial cells were freshly cultured from rat thoracic aortae as described elsewhere.1 The aorta was excised free, cleaned of perivascular adventitial adipose tissue and subjected to digestion in collagenase (Sigma, Type IA) at 37 oC for 15 min. Serum containing medium was added to quench the digestion and the mixture was centrifuged at 800 g for 15 min. The cells were re-suspended in RPMI 1640, supplemented with 10% fetal bovine serum (FBS) plus 1% penicillin/streptomycin (P/S) (GIBCO) and allowed to settle for 1 h upon which the medium was changed. The identity of the endothelial cells was confirmed by immunocytochemical staining for an endothelial cell specific marker, PECAM-1 (Supplemental Figure I). To avoid the influence of culture condition on endothelial cell phenotype, only cells from passage 1 were used in the present study. SDS-PAGE and Western blot analysis Cells were serum-deprived for 24 h and incubated with Ang II (100 nmol/L) for 8 h unless otherwise stated. When used, inhibitors were pre-incubated with the cells for 30 min before the addition of Ang II. After the incubation, cells were lysed in an ice-cold RIPA buffer with a cocktail of protease inhibitors (leupetin, 1 µg/mL; aprotonin, 5 µg/mL; PMSF, 100 µg/mL; sodium orthovanadate, 1 mmol/L; EGTA, 1 mmol/L; EDTA, 1 mmol/L; NaF, 1 mmol/L and β-glycerolphosphate, 2 mg/mL). The lysates were centrifuged at 20,000 g for 20 min and protein concentration was determined by the Lowry method (BioRad). Each protein sample (20 µg) was electrophoresed through the 10% SDS-polyacrylamide gels and then transferred to an Immobilon-P polyvinylidene difluoride (PVDF) membrane (Millipore).2 The membranes were blocked with 5% non-fat skimmed milk or 1% bovine serum albumin in Tween-20 phosphate buffer saline (PBST) and probed overnight at 4 oC with antibodies against COX-2 (Santa Cruz), COX-1 (Cayman), p38 MARK, ERK1/2 and PKC isoforms (Cell Signaling). After washes in PBST, the membranes were incubated with appropriate HRP-conjugated secondary IgG (DakoCytomation) for 1 h at room temperature. The membranes were developed

with an enhanced chemiluminescence detection system (ECL reagents; Amersham Pharmacia Biotech, Uppsala, Sweden) and exposed on X-ray films (Fuji). Densitometry was performed using a documentation program (GBOX-CHEM1-HR16, SynGene). GAPDH antibody (Ambion, Inc) was probed as a loading control. Reverse-transcription polymerase chain reaction (RT-PCR) The potential roles of p38 MAPK, ERK1/2 and PKCδ on COX-2 mRNA stabilization were examined with RT-PCR. Primary rat aortic endothelial cells were first exposed to Ang II (100 nmol/L) for 4 h, after which actinomycin-D (10 µmol/L) and the kinase inhibitors (SB 202190, PD 98059 and rottlerin) were introduced for another 4 h. The cells were lysed and the mRNA was extracted with Trizol reagent (Invitrogen) according to the manufacturer’s instruction. The extracted mRNA was reverse transcribed using the iScriptTM cDNA synthesis kit (BioRad), and PCR was performed with Taq DNA polymerase (Invitrogen, Carlsbad, CA, USA) with thermal cycles of 3-min 94 oC, 29 cycles of 45-s 94 oC, 30-s 56oC, 1.5-min 72 oC, finally followed by 10-min 72 oC using the forward and reverse primers of sequences 5’-TGGTGCCGGGTCTGATGATG-3’ and 5’-GCAATGCGGTTCTGATACTG-3’ respectively.3 The PCR products were then resolved by electrophoresis in a 1.5% agarose gel with ethidium bromide staining. Densitometry was determined in the gel documentation system (GBOX-CHEM1-HR16, SynGene) and normalized to the internal control of GAPDH. Cellular protein fractionation for translocation study of PKC isoforms Endothelial cells were incubated with 100 nmol/L Ang II for 1 min and quickly cooled on ice to quench cellular reactions. Cytosolic and membranous proteins were extracted with ProteoExtract® Subcellular Proteome Extraction Kit according to manufacturer’s instructions (Calbiochem). Cytosolic and nuclear proteins were extracted with tailor-made Buffers A (KCl, 10 mmol/L; DTT, 1 mmol/L; PMSF, 500 µmol/L; EDTA, 100 µmol/L; EGTA, 100 µmol/L; HEPES, 10 mmol/L, pH 7.9) and C (NaCl, 400 mmol/L; DTT, 1 mmol/L; PMSF, 1 mmol/L; EDTA, 1 mmol/L; EGTA, 1 mmol/L, HEPES, 20 mmol/L, pH 7.9). Briefly, cells harvested were first solubilized in Buffer A with vortexing. After incubating on ice for 15 min, 10% NP40 was added and the mixture was centrifuged at 12 000 rpm for 5 min at 4 oC. The supernatant was saved as the cytosolic portion and the pellet was resuspended in Buffer C with vigorous vortexing. The suspension was centrifuge at 12 000 rpm for 5 min at 4 oC after 15-min incubation on ice. The supernatant collected thereafter was labeled as nuclear fraction. Method validation of the extraction protocols were performed by cross-probing with markers for cytosolic, membranous and nuclear proteins (i.e. GAPDH, NOX-2 and histone H3, respectively). PKCδ knockdown with small interfering RNA (siRNA) When primary rat endothelial cells reached 80% confluence, the cells were transfected with siRNA by electroporation using Amaxa Basic Nucleofector Kit for primary mammalian endothelial cells (Lonza, Germany). Briefly, 2.5 x106 cells/mL were trypsinized and washed two times with PBS, and

at Chinese University of Hong Kong on April 22, 2011 atvb.ahajournals.orgDownloaded from

Wong et al., 2011, Supplemental Material

P. 2

resuspended in 100 µL basic nucleofector solution and transferred to a cuvette containing either 30 pmol scramble siRNA or pre-designed specific siRNA targeting PKCδ transcripts (Ambion). The cells were electroporated with the Amaxa NucleofactorTM apparatus, re-plated in 6-well plates containing pre-warmed complete RPMI medium and left undisturbed for 24 h. The cells were then serum-deprived for 24 h before an 8-h incubation with Ang II (100 nmol/L). Inhibitors, when use, were incubated for 30 min prior to Ang II addition. Angiotensin II infusion in rats Sprague-Dawley rats weighing ~250 g were anesthetized with 37.5% ketamine plus 25% xylazine in normal saline. Osmotic minipump (ALZET, model 2ML2, Alza Pharmaceutical, Palo Alto, CA, USA) filled with either Ang II or the vehicle (normal saline) was implanted subcutaneously in the dorsal region. Ang II was infused at 0.7 mg/kg/day for 9 days. Blood pressure was regularly monitored throughout the infusion period with a tail-cuff electrosphygmomanometer system, after which aortae and renal arteries were excised, dissected free from perivascular connective tissue and preserved in 4% paraformaldehyde for immunohistchemical staining. Interlobal renal arteries were used for functional evaluation of endothelium-dependent relaxations. Immunohistochemical staining Localization of COX-2 in the aortae and renal arteries from Ang II-infused rats were determined by immunohistochemistry. The tissues were fixed overnight in 4% paraformaldehyde, processed for embedding in wax and then cut into 5 µm-sections. Following re-hydration and treatment with 1.4% hydrogen peroxide in absolute methanol for 30 min at room temperature to inhibit endogenous peroxidase activity, antigen retrieval was performed by boiling the sections in 0.01 mol/L sodium citrate buffer (pH 6) for 30-60 s. After rinsing in PBS, the sections were blocked with 5% donkey serum and incubated overnight with COX-2 antibody diluted in PBS supplemented with 2% BSA in a humidified chamber at 4 oC. The sections were then incubated with biotinylated secondary antibodies for 1 h at room temperature, followed by 1 h incubation with peroxidase-conjugated streptavidin. DAB (Vector Laboratories, California, USA) was used for colour development according to the manufacturer’s instruction. Counter-staining of nuclei and cytoplasm was performed with haematoxylin and eosin respectively. Negative control was performed in the absence of primary antibody. Images were viewed and captured under Leica DMRBE microscope coupled to SPOT-RT cooled CCD color digital camera using the objective PL FLUOTAR 20x/0.50 and SPOT Advanced software (Version 3.5.5). Immunofluorescence localization Human small mesenteric arteries harvested after a 24-h incubation protocol were embedded in OCT compound (Sakura Finetek, the Netherlands) in aluminium cryomolds, snap frozen in isopentane pre-cooled in liquid nitrogen and cut into 10 µm thick cryostat sections. The thawed sections were air-dried, post-fixed in 4% paraformaldehyde for 30 min, and then briefly treated with 0.05% Triton X in PBS. The sections were blocked with 5% donkey serum for 1 h at room temperature. Primary antibodies against COX-2 or MCP-1 (Santa Cruz) were incubated overnight at 4 oC. The sections were then labeled with Alexa Fluor 546 donkey anti-goat IgG (Invitrogen, Molecular Probes, California, USA) for 1 h at room temperature. The sections were cover-slipped in anti-fade mounting medium (Vector Laboratories) and viewed under fluorescence microscope (Nikon Eclipse Ti-U) with mercury lamp Nikon Intensilight C-HGFI using the objective Nikon S Fluor 20x/0.75. Images were acquired with SPOT RT3 cooled 2 MP CCD scientific color digital camera and SPOT advanced software (Version 4.6). Functional examination with myography Endothelium-dependent relaxations of the interlobal renal arteries were determined in myographs. After subjecting the rats to a 9 day-infusion of Ang II, interlobal renal arteries were dissected and suspended in the myograph (Danish Myo

Technology, Aarhus, Denmark) with stainless steel wires. Each chamber was filled with 5 mL Krebs-Henseleit solution containing (mmol/L): NaCl 119, NaHCO3 25, MgCl2 1, KCl 4.7, KH2PO4 1.2, CaCl2 2.5, and D-glucose 11.1, and aerated with 95% O2-5% CO2 at 37 °C to give a pH of ~7.4. The arterial rings were stretched to a pre-determined optimal resting tension of 2 mN and allowed to stabilize at this basal tone for 60 min before experiments began. Acetylcholine-induced endothelium-dependent relaxations were evaluated in arteries from control rats and Ang II-infused rats with or without concomitant oral administration of losartan (10 mg/kg/day). The acute effects of celecoxib (3 µmol/L) and sc-560 (0.3 µmol/L) were tested on the relaxations in renal arteries from Ang II-infused rats. Reactive oxygen species (ROS) detection with dihydroethidium (DHE) fluorescence Production of intracellular ROS was measured with DHE (Molecular Probes), which binds to DNA upon oxidation to emit fluorescence. Following a 30-s treatment of Ang II, primary rat endothelial cells seeded on cover-slips were incubated in 5 µmol/L DHE for 20 min at 37 oC. After a rinse in PBS, fluorescence was observed under the confocal microscopy system Olympus Fluoview FV1000 (515-nm excitation; 585-nm long pass filter) using the objective UMPlanFI 10x/0.30. DHE fluorescence intensity was analyzed by Olympus Fluoview software (version 1.5; FV10-ASW1.5). Similar experiments were performed to compare the difference in ROS levels between 15-min exposure of Ang II and bone morphogenic protein-4 (BMP4). Data were expressed in fold change compared with untreated control. Chemicals Angiotensin II, losartan, PD 123319, actinomycin-D, SB 202190, PD 98059, SP 600125, L-NAME, AMT, GF109203X, Go 6976, NS 398 and phorbol 12-myristate 13-acetate were purchased from Tocris (Avonmouth, UK). Rottlerin was from Enzo Life Sciences (New York, USA) and εV1-2 from AnaSpec (California, USA). BMP4, oxypurinol, tiron, tempol, DETCA and 4α-phorbol 12-myristate 13-acetate were purchased from Sigma-Aldrich Chemical (St Louis, MO, USA). Apocynin were from Calbiochem, EMD Biosciences (La Jolla, CA, USA). Celecoxib was from Pfizer and sc-560 was a kind gift from Institut de Recherches Servier (Suresnes, France). Angiotensin II, PD 123319, NG-nitro-L-arginine methyl ester, AMT, tiron, tempol, DETCA and εV1-2 were dissolved in distilled water and the others in dimethyl sulphoxide (DMSO, Sigma). DMSO served as solvent control. Significance of Study Clinical use of COX-2 inhibitors in suppressing inflammatory responses remains a hot topic in recent years due to their reported adverse cardiovascular effects. While COX-2 inhibition was shown to increase blood pressure and thrombotic events,4, 5 it also produces benefits in patients with hypertension and coronary artery disease.6, 7 These seemingly contradictory findings may suggest COX-2 taking up a Janus role in producing different types of metabolites probably depending on the nature of triggers under physiological and pathological states. While the exact role of COX-2 remains unresolved and preferentially targeting COX-2 in particular systems is a challenging task, the present study provides a novel insight into how COX-2 action can be modulated by regulating its expression. In this study, we demonstrate an important role of PKCδ in mediating Ang II-induced COX-2 expression. Intriguingly, a recent study has shown that adiponectin, a cardioprotective adipokine, up-regulates endothelial COX-2 expression by activating calreticulin/CD91-PI3K/Akt and promotes revascularization after ischemia.8 It is noteworthy that Ang II-induced COX-2 expression does not involve PI3K pathways (data not shown), implying that COX-2 expression can be mediated by different agonist-dependent pathways. If PKCδ is selectively targeted in a strategic manner to suppress the up-regulation of pro-inflammatory COX-2, the expression of protective COX-2 can remain undisturbed and thus preventing the occurrence of the scenario in which non-selective inhibition towards these functionally different COX-2 abrogates the beneficial effects

at Chinese University of Hong Kong on April 22, 2011 atvb.ahajournals.orgDownloaded from

Wong et al., 2011, Supplemental Material

P. 3

from the protective ones. Indeed, PKCβ and PKCδ are preferentially activated in response to hyperglycemia and diabetes9 and PKC inhibitors such as ruboxistaurin targeting the β isoform has been introduced in clinical trials.10 Inhibiting COX-2 expression and thus subsequently down-regulating MCP-1 release may offer an alternative to retard the initiation of atherosclerosis. With the advancement in the development of phosphopeptide mimetics, specific peptide inhibitors of PKCδ may uncover novel therapeutic potential to combat vascular inflammatory disease. References 1. Huang Y, Chan NW, Lau CW, Yao XQ, Chan FL,

Chen ZY. Involvement of endothelium/nitric oxide in vasorelaxation induced by purified green tea (-)epicatechin. Biochim Biophys Acta. 1999;1427(2):322-328.

2. Wong SL, Leung FP, Lau CW, Au CL, Yung LM, Yao X, Chen ZY, Vanhoutte PM, Gollasch M, Huang Y. Cyclooxygenase-2-derived prostaglandin F2alpha mediates endothelium-dependent contractions in the aortae of hamsters with increased impact during aging. Circ Res. 2009;104(2):228-235.

3. Jones MK, Sasaki E, Halter F, Pai R, Nakamura T, Arakawa T, Kuroki T, Tarnawski AS. HGF triggers activation of the COX-2 gene in rat gastric epithelial cells: action mediated through the ERK2 signaling pathway. FASEB J. 1999;13(15):2186-2194.

4. Bresalier RS, Sandler RS, Quan H, Bolognese JA, Oxenius B, Horgan K, Lines C, Riddell R, Morton D, Lanas A, Konstam MA, Baron JA, Adenomatous Polyp Prevention on Vioxx Trial I. Cardiovascular events associated with rofecoxib in a colorectal adenoma chemoprevention trial. N Engl J Med. 2005;352(11):1092-1102.

5. Capone ML, Tacconelli S, Francesco LD, Petrelli M, Patrignani P. Cardiovascular effects of valdecoxib: transducing human pharmacology results into clinical read-outs. Expert Opin Drug Saf. 2008;7(1):29-42.

6. Widlansky ME, Price DT, Gokce N, Eberhardt RT, Duffy SJ, Holbrook M, Maxwell C, Palmisano J, Keaney JF, Jr., Morrow JD, Vita JA. Short- and long-term COX-2 inhibition reverses endothelial dysfunction in patients with hypertension. Hypertension. 2003;42(3):310-315.

7. Chenevard R, Hurlimann D, Bechir M, Enseleit F, Spieker L, Hermann M, Riesen W, Gay S, Gay RE, Neidhart M, Michel B, Luscher TF, Noll G, Ruschitzka F. Selective COX-2 inhibition improves endothelial function in coronary artery disease. Circulation. 2003;107(3):405-409.

8. Ohashi K, Ouchi N, Sato K, Higuchi A, Ishikawa TO, Herschman HR, Kihara S, Walsh K. Adiponectin promotes revascularization of ischemic muscle through a cyclooxygenase 2-dependent mechanism. Mol Cell Biol. 2009;29(13):3487-3499.

9. Meier M, King GL. Protein kinase C activation and its pharmacological inhibition in vascular disease. Vasc Med. 2000;5(3):173-185.

10. Geraldes P, King GL. Activation of protein kinase C isoforms and its impact on diabetic complications. Circ Res. 2010;106(8):1319-1331.

Supplemental Figure I. Confirmation of the endothelial identity in the primary culture of cells. Primary culture of cells from rat thoracic aortae were confirmed to be endothelial cells, which were positively stained with an endothelial cell-specific maker, PECAM-1. Fibroblasts served as a negative control to the staining.

Supplemental Figure II. AT1R-mediated COX-2 expression reaches a maximum in 8 h with 100 nmol/L Ang II and COX-2 mRNA is stabilized by p38 MAPK, ERK1/2 and PKCδ. A and B, COX-2 expression in primary rat endothelial cells increased in response to Ang II in (A) a time-dependent (1-8 h, 100 nmol/L) and (B) concentration-dependent (3-100 nmol/L, 8 h) manner. C, Ang II (100 nmol/L)-induced COX-2 expression was prevented by pre-treatment of actinomycin-D (10 µmol/L) and losartan (3 µmol/L), but unaffected by the presence of PD 123319 (1 µmol/L) and the solvent DMSO. D, Inhibitors of p38 MAPK (SB 202190, SB, 10 µmol/L), ERK1/2 (PD 98059, PD, 20 µmol/L) and PKCδ (rottlerin, ROT, 10 µmol/L) all partially reduced COX-2 mRNA stability in endothelial cells pre-exposed to Ang II (100 nmol/L) for 4 h with a subsequent treatment of actinomycin-D (AD, 10 µmol/L). Results represent means ± SEM of 4-6 experiments. **P<0.01, ***P<0.001 versus control; #P<0.05, ##P<0.01 and ###P<0.001 compared to Ang II (A-C) or Ang II + AD group (D).

PECAM-1

Endothelialcells

Fibroblasts

NegativeControl

200 µm

PECAM-1

Endothelialcells

Fibroblasts

NegativeControl

200 µm

0 3 10 30 1000

10

20

Ang II (nmol/L)

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

*** *** ******

##

COX-2GAPDH

0 3 10 30 1000

10

20

Ang II (nmol/L)

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

*** *** ******

##

0 3 10 30 1000

10

20

Ang II (nmol/L)

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

*** *** ******

##

COX-2GAPDHCOX-2

GAPDH

1 2 4 80

10

20 ControlAng II 100 nmol/L

Time (h)

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

****** ***

**

COX-2GAPDH

1 2 4 80

10

20 ControlAng II 100 nmol/L

Time (h)

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

****** ***

**

COX-2GAPDHCOX-2

GAPDH

Contro

l

Ang II

Actino

mycin-

D

Losa

rtan

PD 1233

19

DMSO

0

10

20Ang II

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

***

### ###

COX-2GADPH

Contro

l

Ang II

Actino

mycin-

D

Losa

rtan

PD 1233

19

DMSO

0

10

20Ang II

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

***

### ###

Contro

l

Ang II

Actino

mycin-

D

Losa

rtan

PD 1233

19

DMSO

0

10

20Ang II

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

***

### ###

COX-2GADPHCOX-2

GADPH

A B

0

5

10

15

20

CO

X-2

mR

NA

leve

l(F

old

of c

ontro

l)

COX-2GAPDH

Ang IIADSBPD

ROTDMSO

+ + + + + + + + ++ + + + + + + + + + + + + + + + ++ + + + + + + + ++ + + + + + + + ++ + + + + + + + +

###

#

###

**

0

5

10

15

20

CO

X-2

mR

NA

leve

l(F

old

of c

ontro

l)

COX-2GAPDH

Ang IIADSBPD

ROTDMSO

+ + + + + + + + ++ + + + + + + + + + + + + + + + ++ + + + + + + + ++ + + + + + + + ++ + + + + + + + +

###

#

###

**

C D

0 3 10 30 1000

10

20

Ang II (nmol/L)

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

*** *** ******

##

COX-2GAPDH

0 3 10 30 1000

10

20

Ang II (nmol/L)

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

*** *** ******

##

0 3 10 30 1000

10

20

Ang II (nmol/L)

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

*** *** ******

##

COX-2GAPDHCOX-2

GAPDH

1 2 4 80

10

20 ControlAng II 100 nmol/L

Time (h)

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

****** ***

**

COX-2GAPDH

1 2 4 80

10

20 ControlAng II 100 nmol/L

Time (h)

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

****** ***

**

COX-2GAPDHCOX-2

GAPDH

Contro

l

Ang II

Actino

mycin-

D

Losa

rtan

PD 1233

19

DMSO

0

10

20Ang II

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

***

### ###

COX-2GADPH

Contro

l

Ang II

Actino

mycin-

D

Losa

rtan

PD 1233

19

DMSO

0

10

20Ang II

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

***

### ###

Contro

l

Ang II

Actino

mycin-

D

Losa

rtan

PD 1233

19

DMSO

0

10

20Ang II

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

***

### ###

COX-2GADPHCOX-2

GADPH

A B

0

5

10

15

20

CO

X-2

mR

NA

leve

l(F

old

of c

ontro

l)

COX-2GAPDH

Ang IIADSBPD

ROTDMSO

+ + + + + + + + ++ + + + + + + + + + + + + + + + ++ + + + + + + + ++ + + + + + + + ++ + + + + + + + +

###

#

###

**

0

5

10

15

20

CO

X-2

mR

NA

leve

l(F

old

of c

ontro

l)

COX-2GAPDH

Ang IIADSBPD

ROTDMSO

+ + + + + + + + ++ + + + + + + + + + + + + + + + ++ + + + + + + + ++ + + + + + + + ++ + + + + + + + +

###

#

###

**

C D

at Chinese University of Hong Kong on April 22, 2011 atvb.ahajournals.orgDownloaded from

Wong et al., 2011, Supplemental Material

P. 4

Supplemental Figure III. Unaltered COX-1 expression in the presence of Ang II and various inhibitors. Constitutive expression of COX-1 remained changed in response to (A) Ang II across time (1-8 h, 100 nmol/L), (B) different concentrations of Ang II (3-100 nmol/L, 8h), and (C-F) various inhibitors.

Supplemental Figure IV. PKCδ was not phosphorylated at Threonine 505 (T505) or Tyrosine 332 (Y332) in response to Ang II (100 nmol/L). In contrast, these residues were phosphorylated by H2O2 (5 mmol/L) at 10 min, which served as a positive control. Supplemental Figure V. Translocation study (from cytosol to the membrane) of various PKC isoforms in response to Ang II and PMA. A and B, Validation on the extraction method of cytosolic and membranous protein by detecting cytosolic GAPDH and membrane-bound NOX-2. C through I, Translocation of various PKC isoforms in response to 1-min stimulation of Ang II (100 nmol/L) or PMA (1 µmol/L). Cyto, cytosol; M, membrane. **P<0.01 and ***P<0.001 compared to the (A) cytosolic protein or (B-H) membranous portion of control.

Supplemental Figure VI. PKCδ translocation study (from cytosol to the nucleus) in response to Ang II. A, PKCδ is not translocated into the nucleus after 1-min Ang II stimulation. B and C, Fractionation of cytosolic (cyto) and nuclear proteins was validated by the cytosolic (b, GAPDH) and nuclear (c, histone H3) markers respectively. ***P<0.001 versus cyto.

Supplemental Figure VII. Ang II-induced COX-2 expression is mediated by p-38 MAPK and PKCδ, without involvement of PKCε. A and B, Co-treatment of SB 202190 (SB, 10 µmol/L) and rottlerin (10 µmol/L) abolished Ang II-induced COX-2 expression, while these inhibitors did not affect COX-1 expression. C, εV1-2 (10 µmol/L) did not reduce COX-2 expression. ***P<0.001 compared to control; #P<0.05, ##P<0.01 and ###P<0.001 compared to Ang II-treated group.

Supplemental Figure VIII. PKCδ mediates COX-2 expression induced by Ang II or exogenous PMA. A and B, Small interfering RNA targeting PKCδ (siPKCδ) abolished expression of PKCδ but not PKCε. C and D, Ang II-induced COX-2 expression was significantly reduced in siPKCδ-transfected cells, treatment with SB 202190 (SB) further prevented COX-2 up-regulation by Ang II. E, COX-2 expression peaked at 4-h incubation of PMA (1 µmol/L), while its negative analogue 4α-PMA (1 µmol/L) did not induce COX-2 expression. F, PMA-induced COX-2 expression was inhibited by GFX (2 µmol/L) and rottlerin (10 µmol/L), but not Go 6976 (1 µmol/L). *P<0.05, **P<0.01 and ***P<0.001 compared to respective control; #P<0.05, ##P<0.01 and ###P<0.001 compared to Ang II- or PMA-treated group.

ControlAng II SB

Rottlerin

SB + RottlerinDMSO

0

10

20

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

#

##

***

###

COX-2

GAPDH

ControlAng II SB

Rottlerin

SB + RottlerinDMSO

0

10

20

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

#

##

***

###

ControlAng II SB

Rottlerin

SB + RottlerinDMSO

0

10

20

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

#

##

***

###

COX-2

GAPDH

COX-2

GAPDH

A B C

ControlAng II

V1-2ε

Losartan

0

10

20

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

***

##

COX-2

GAPDH

ControlAng II

V1-2ε

Losartan

0

10

20

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

***

##

ControlAng II

V1-2ε

Losartan

0

10

20

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

***

##

COX-2

GAPDH

COX-2

GAPDH

ControlAng II SB

Rottlerin

SB + RottlerinDMSO

0.0

0.5

1.0

1.5

CO

X-1

expr

essi

on(F

old

of c

ontro

l)

COX-1

GAPDH

ControlAng II SB

Rottlerin

SB + RottlerinDMSO

0.0

0.5

1.0

1.5

CO

X-1

expr

essi

on(F

old

of c

ontro

l)

COX-1

GAPDH

COX-1

GAPDH

Ang IIAng II

Ang II

ControlAng II SB

Rottlerin

SB + RottlerinDMSO

0

10

20

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

#

##

***

###

COX-2

GAPDH

ControlAng II SB

Rottlerin

SB + RottlerinDMSO

0

10

20

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

#

##

***

###

ControlAng II SB

Rottlerin

SB + RottlerinDMSO

0

10

20

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

#

##

***

###

COX-2

GAPDH

COX-2

GAPDH

A B C

ControlAng II

V1-2ε

Losartan

0

10

20

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

***

##

COX-2

GAPDH

ControlAng II

V1-2ε

Losartan

0

10

20

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

***

##

ControlAng II

V1-2ε

Losartan

0

10

20

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

***

##

COX-2

GAPDH

COX-2

GAPDH

ControlAng II SB

Rottlerin

SB + RottlerinDMSO

0.0

0.5

1.0

1.5

CO

X-1

expr

essi

on(F

old

of c

ontro

l)

COX-1

GAPDH

ControlAng II SB

Rottlerin

SB + RottlerinDMSO

0.0

0.5

1.0

1.5

CO

X-1

expr

essi

on(F

old

of c

ontro

l)

COX-1

GAPDH

COX-1

GAPDH

ControlAng II SB

Rottlerin

SB + RottlerinDMSO

0

10

20

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

#

##

***

###

COX-2

GAPDH

ControlAng II SB

Rottlerin

SB + RottlerinDMSO

0

10

20

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

#

##

***

###

ControlAng II SB

Rottlerin

SB + RottlerinDMSO

0

10

20

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

#

##

***

###

COX-2

GAPDH

COX-2

GAPDH

A B C

ControlAng II

V1-2ε

Losartan

0

10

20

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

***

##

COX-2

GAPDH

ControlAng II

V1-2ε

Losartan

0

10

20

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

***

##

ControlAng II

V1-2ε

Losartan

0

10

20

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

***

##

COX-2

GAPDH

COX-2

GAPDH

ControlAng II SB

Rottlerin

SB + RottlerinDMSO

0.0

0.5

1.0

1.5

CO

X-1

expr

essi

on(F

old

of c

ontro

l)

COX-1

GAPDH

ControlAng II SB

Rottlerin

SB + RottlerinDMSO

0.0

0.5

1.0

1.5

CO

X-1

expr

essi

on(F

old

of c

ontro

l)

COX-1

GAPDH

COX-1

GAPDH

Ang IIAng IIAng IIAng II

Ang II

0 3 10 30 1000.0

0.5

1.0

1.5

Ang II (nmol/L)C

OX-

1 ex

pres

sion

(Fol

d of

con

trol)

COX-1GAPDH

0 3 10 30 1000.0

0.5

1.0

1.5

Ang II (nmol/L)C

OX-

1 ex

pres

sion

(Fol

d of

con

trol)

COX-1GAPDHCOX-1

GAPDH

1 2 4 80.0

0.5

1.0

1.5 ControlAng II 100 nmol/L

Time (h)

CO

X-1

expr

essi

on(F

old

of c

ontro

l)

COX-1GAPDH

1 2 4 80.0

0.5

1.0

1.5 ControlAng II 100 nmol/L

Time (h)

CO

X-1

expr

essi

on(F

old

of c

ontro

l)

COX-1GAPDHCOX-1

GAPDH

Contro

l

Ang II

Apocy

nin

Oxypu

rinol

Tempo

l

Tiron +

DETCA

AMT

L-NAME

0.0

0.5

1.0

1.5

CO

X-1

expr

essi

on(F

old

of c

ontro

l)

COX-1GAPDH

Ang II

Contro

l

Ang II

Apocy

nin

Oxypu

rinol

Tempo

l

Tiron +

DETCA

AMT

L-NAME

0.0

0.5

1.0

1.5

CO

X-1

expr

essi

on(F

old

of c

ontro

l)

COX-1GAPDH

Ang II

Contro

l

Ang II

GFX

Go 697

6

Rottler

in

DMSO

0.0

0.5

1.0

1.5

CO

X-1

expr

essi

on(F

old

of c

ontro

l)

COX-1GAPDH

Ang II

Contro

l

Ang II

GFX

Go 697

6

Rottler

in

DMSO

0.0

0.5

1.0

1.5

CO

X-1

expr

essi

on(F

old

of c

ontro

l)

COX-1GAPDHCOX-1

GAPDHAng II

Contro

l

Ang II SB PD SP

SB + PDDMSO

0.0

0.5

1.0

1.5

CO

X-1

expr

essi

on(F

old

of c

ontro

l)

COX-1GAPDH

Ang II

Contro

l

Ang II SB PD SP

SB + PDDMSO

0.0

0.5

1.0

1.5

CO

X-1

expr

essi

on(F

old

of c

ontro

l)

COX-1GAPDHCOX-1

GAPDH

Ang II

Contro

l

Ang II

Actino

mycin-

D

Losa

rtan

PD 1233

19

DMSO

0.0

0.5

1.0

1.5

CO

X-1

expr

essi

on(F

old

of c

ontro

l)

COX-1GAPDH

Ang II

Contro

l

Ang II

Actino

mycin-

D

Losa

rtan

PD 1233

19

DMSO

0.0

0.5

1.0

1.5

CO

X-1

expr

essi

on(F

old

of c

ontro

l)

COX-1GAPDHCOX-1

GAPDH

Ang II

D E F

A B C

0 3 10 30 1000.0

0.5

1.0

1.5

Ang II (nmol/L)C

OX-

1 ex

pres

sion

(Fol

d of

con

trol)

COX-1GAPDH

0 3 10 30 1000.0

0.5

1.0

1.5

Ang II (nmol/L)C

OX-

1 ex

pres

sion

(Fol

d of

con

trol)

COX-1GAPDHCOX-1

GAPDH

1 2 4 80.0

0.5

1.0

1.5 ControlAng II 100 nmol/L

Time (h)

CO

X-1

expr

essi

on(F

old

of c

ontro

l)

COX-1GAPDH

1 2 4 80.0

0.5

1.0

1.5 ControlAng II 100 nmol/L

Time (h)

CO

X-1

expr

essi

on(F

old

of c

ontro

l)

COX-1GAPDHCOX-1

GAPDH

Contro

l

Ang II

Apocy

nin

Oxypu

rinol

Tempo

l

Tiron +

DETCA

AMT

L-NAME

0.0

0.5

1.0

1.5

CO

X-1

expr

essi

on(F

old

of c

ontro

l)

COX-1GAPDH

Ang II

Contro

l

Ang II

Apocy

nin

Oxypu

rinol

Tempo

l

Tiron +

DETCA

AMT

L-NAME

0.0

0.5

1.0

1.5

CO

X-1

expr

essi

on(F

old

of c

ontro

l)

COX-1GAPDH

Ang II

Contro

l

Ang II

GFX

Go 697

6

Rottler

in

DMSO

0.0

0.5

1.0

1.5

CO

X-1

expr

essi

on(F

old

of c

ontro

l)

COX-1GAPDH

Ang II

Contro

l

Ang II

GFX

Go 697

6

Rottler

in

DMSO

0.0

0.5

1.0

1.5

CO

X-1

expr

essi

on(F

old

of c

ontro

l)

COX-1GAPDHCOX-1

GAPDHAng II

Contro

l

Ang II SB PD SP

SB + PDDMSO

0.0

0.5

1.0

1.5

CO

X-1

expr

essi

on(F

old

of c

ontro

l)

COX-1GAPDH

Ang II

Contro

l

Ang II SB PD SP

SB + PDDMSO

0.0

0.5

1.0

1.5

CO

X-1

expr

essi

on(F

old

of c

ontro

l)

COX-1GAPDHCOX-1

GAPDH

Ang II

Contro

l

Ang II

Actino

mycin-

D

Losa

rtan

PD 1233

19

DMSO

0.0

0.5

1.0

1.5

CO

X-1

expr

essi

on(F

old

of c

ontro

l)

COX-1GAPDH

Ang II

Contro

l

Ang II

Actino

mycin-

D

Losa

rtan

PD 1233

19

DMSO

0.0

0.5

1.0

1.5

CO

X-1

expr

essi

on(F

old

of c

ontro

l)

COX-1GAPDHCOX-1

GAPDH

Ang II

D E F

A B C

Ang II0 1 2 3 4 5 10 15 30 45 60 10 (min)

H2O2

p-PKC(T505)p-PKC(Y332)

PKC

GAPDH

Ang II0 1 2 3 4 5 10 15 30 45 60 10 (min)

H2O2

p-PKC(T505)p-PKC(Y332)

PKC

GAPDH

Control Ang II012345 cyto nuclear

PKCδ

(Opt

ical

den

sity

)

PKCδ

Control Ang II012345 cyto nuclear

PKCδ

(Opt

ical

den

sity

)

PKCδ GAPDH

cyto nuclear0

2

4

6

8

GA

PDH

(Opt

ical

den

sity

)

***

GAPDH

cyto nuclear0

2

4

6

8

GA

PDH

(Opt

ical

den

sity

)

***cyto nuclear

0

2

4

6

8

His

tone

H3

(Opt

ical

den

sity

)

HistoneH3

***

cyto nuclear0

2

4

6

8

His

tone

H3

(Opt

ical

den

sity

)

HistoneH3

cyto nuclear0

2

4

6

8

His

tone

H3

(Opt

ical

den

sity

)

HistoneH3

***

A B C

Control Ang II012345 cyto nuclear

PKCδ

(Opt

ical

den

sity

)

PKCδ

Control Ang II012345 cyto nuclear

PKCδ

(Opt

ical

den

sity

)

PKCδ GAPDH

cyto nuclear0

2

4

6

8

GA

PDH

(Opt

ical

den

sity

)

***

GAPDH

cyto nuclear0

2

4

6

8

GA

PDH

(Opt

ical

den

sity

)

***cyto nuclear

0

2

4

6

8

His

tone

H3

(Opt

ical

den

sity

)

HistoneH3

***

cyto nuclear0

2

4

6

8

His

tone

H3

(Opt

ical

den

sity

)

HistoneH3

cyto nuclear0

2

4

6

8

His

tone

H3

(Opt

ical

den

sity

)

HistoneH3

***

A B C

Control

Scramble δ

siPKC

0.0

0.5

1.0 ControlAng II

CO

X-2/

GAP

DH

*** ***

*###

Control

Scramble δ

siPKC

0.0

0.5

1.0 ControlAng II

CO

X-2/

GAP

DH

*** ***

*###

Control Scramble siPKCδ

0.0

0.5

1.0

PKCδ e

xpre

ssio

n(F

old

of c

ontro

l)

***Control Scramble siPKCδ

0.0

0.5

1.0

PKCδ e

xpre

ssio

n(F

old

of c

ontro

l)

***

Control

Scramble δ

siPKC

+ SBδ

siPKC

0.0

0.5

1.0

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

******

#

Control

Scramble δ

siPKC

+ SBδ

siPKC

0.0

0.5

1.0

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

******

#

Control Scramble siPKCδ

0.0

0.5

1.0

PKCε e

xpre

ssio

n

0 0.5 1 2 4 8 80

5

10

15

Time (h)

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

**

**

***

******

PMA

4α-PMA

0 0.5 1 2 4 8 80

5

10

15

Time (h)

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

**

**

***

******

PMA

4α-PMA

A

C

E

Contro

lPMA

GFX

Go 697

6

Rottler

in

DMSO

0

5

10

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

PMA (4 h)***

####

Contro

lPMA

GFX

Go 697

6

Rottler

in

DMSO

0

5

10

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

PMA (4 h)***

####

B

D

FControl

Scramble δ

siPKC

0.0

0.5

1.0 ControlAng II

CO

X-2/

GAP

DH

*** ***

*###

Control

Scramble δ

siPKC

0.0

0.5

1.0 ControlAng II

CO

X-2/

GAP

DH

*** ***

*###

Control Scramble siPKCδ

0.0

0.5

1.0

PKCδ e

xpre

ssio

n(F

old

of c

ontro

l)

***Control Scramble siPKCδ

0.0

0.5

1.0

PKCδ e

xpre

ssio

n(F

old

of c

ontro

l)

***

Control

Scramble δ

siPKC

+ SBδ

siPKC

0.0

0.5

1.0

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

******

#

Control

Scramble δ

siPKC

+ SBδ

siPKC

0.0

0.5

1.0

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

******

#

Control Scramble siPKCδ

0.0

0.5

1.0

PKCε e

xpre

ssio

n

0 0.5 1 2 4 8 80

5

10

15

Time (h)

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

**

**

***

******

PMA

4α-PMA

0 0.5 1 2 4 8 80

5

10

15

Time (h)

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

**

**

***

******

PMA

4α-PMA

A

C

E

Contro

lPMA

GFX

Go 697

6

Rottler

in

DMSO

0

5

10

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

PMA (4 h)***

####

Contro

lPMA

GFX

Go 697

6

Rottler

in

DMSO

0

5

10

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

PMA (4 h)***

####

B

D

F

cyto mb0

2

4

6

NO

X-2

(Opt

ical

den

sity

)

**NOX-2

cyto mb0

2

4

6

NO

X-2

(Opt

ical

den

sity

)

**

cyto mb0

2

4

6

NO

X-2

(Opt

ical

den

sity

)

**NOX-2

cyto mb0

5

10

GAP

DH

(Opt

ical

den

sity

)

***

GAPDH

cyto mb0

5

10

GAP

DH

(Opt

ical

den

sity

)

***cyto mb

0

5

10

GAP

DH

(Opt

ical

den

sity

)

***

GAPDH

Contro

l

Ang II SB PD

DMSO

0

2

4

6cyto mb

Ang II

PKCδ

(Opt

ical

den

isty

)

PKCδ

**

Contro

l

Ang II SB PD

DMSO

0

2

4

6cyto mb

Ang II

PKCδ

(Opt

ical

den

isty

)

PKCδ

**

Control Ang II PMA0

2

4

6cyto mb

PKCδ

(Opt

ical

den

isty

)

PKCδ

** **

Control Ang II PMA0

2

4

6cyto mb

PKCδ

(Opt

ical

den

isty

)

PKCδ

** **

Control Ang II PMA0

2

4

6 cyto mb

PKCα

(Opt

ical

den

isty

)

PKCα

Control Ang II PMA0

2

4

6 cyto mb

PKCα

(Opt

ical

den

isty

)

PKCα

Control Ang II PMA0.0

0.2

0.4

0.6 cyto mb

PKCζ

(Opt

ical

den

isty

)PKCζ

Control Ang II PMA0.0

0.2

0.4

0.6 cyto mb

PKCζ

(Opt

ical

den

isty

)PKCζ

Control Ang II PMA0

1

2

3 cyto mb

PKCµ

(Opt

ical

den

isty

)PKCµ

Control Ang II PMA0

1

2

3 cyto mb

PKCµ

(Opt

ical

den

isty

)PKCµ

Control Ang II PMA01234 cyto mb

PKCθ

(Opt

ical

den

isty

)

PKCθ

Control Ang II PMA01234 cyto mb

PKCθ

(Opt

ical

den

isty

)

PKCθ

Control Ang II PMA01234 cyto mb

PKCε

(Opt

ical

den

isty

)PKCε

** **

Control Ang II PMA01234 cyto mb

PKCε

(Opt

ical

den

isty

)PKCε

** **

A B C

D E F

G H I

cyto mb0

2

4

6

NO

X-2

(Opt

ical

den

sity

)

**NOX-2

cyto mb0

2

4

6

NO

X-2

(Opt

ical

den

sity

)

**

cyto mb0

2

4

6

NO

X-2

(Opt

ical

den

sity

)

**NOX-2

cyto mb0

5

10

GAP

DH

(Opt

ical

den

sity

)

***

GAPDH

cyto mb0

5

10

GAP

DH

(Opt

ical

den

sity

)

***cyto mb

0

5

10

GAP

DH

(Opt

ical

den

sity

)

***

GAPDH

Contro

l

Ang II SB PD

DMSO

0

2

4

6cyto mb

Ang II

PKCδ

(Opt

ical

den

isty

)

PKCδ

**

Contro

l

Ang II SB PD

DMSO

0

2

4

6cyto mb

Ang II

PKCδ

(Opt

ical

den

isty

)

PKCδ

**

Control Ang II PMA0

2

4

6cyto mb

PKCδ

(Opt

ical

den

isty

)

PKCδ

** **

Control Ang II PMA0

2

4

6cyto mb

PKCδ

(Opt

ical

den

isty

)

PKCδ

** **

Control Ang II PMA0

2

4

6 cyto mb

PKCα

(Opt

ical

den

isty

)

PKCα

Control Ang II PMA0

2

4

6 cyto mb

PKCα

(Opt

ical

den

isty

)

PKCα

Control Ang II PMA0.0

0.2

0.4

0.6 cyto mb

PKCζ

(Opt

ical

den

isty

)PKCζ

Control Ang II PMA0.0

0.2

0.4

0.6 cyto mb

PKCζ

(Opt

ical

den

isty

)PKCζ

Control Ang II PMA0

1

2

3 cyto mb

PKCµ

(Opt

ical

den

isty

)PKCµ

Control Ang II PMA0

1

2

3 cyto mb

PKCµ

(Opt

ical

den

isty

)PKCµ

Control Ang II PMA01234 cyto mb

PKCθ

(Opt

ical

den

isty

)

PKCθ

Control Ang II PMA01234 cyto mb

PKCθ

(Opt

ical

den

isty

)

PKCθ

Control Ang II PMA01234 cyto mb

PKCε

(Opt

ical

den

isty

)PKCε

** **

Control Ang II PMA01234 cyto mb

PKCε

(Opt

ical

den

isty

)PKCε

** **

A B C

D E F

G H I

at Chinese University of Hong Kong on April 22, 2011 atvb.ahajournals.orgDownloaded from

Wong et al., 2011, Supplemental Material

P. 5

Supplemental Figure IX. Representative images of DHE fluorescence. The left panel of each pair of images showed the DHE fluorescence signal (in red) on the production of reactive oxygen species, while the right panel showed the brightfield images of the endothelial cells.

Supplemental Figure X. Reactive oxygen species are not involved in Ang II-induced COX-2 expression. A, DHE fluorescence showed that production of reactive oxygen species in the endothelial cells increased about 2-fold in 30 s after Ang II (100 nmol/L) addition, which were inhibited by losartan (3 µmol/L), apocynin (100 µmol/L) and tiron (1 mmol/L) plus DETCA (100 µmol/L). B, COX-2 expression was not inhibited in the presence of various ROS inhibitors. C and D, H2O2 only induced a very low level of COX-2 expression compared to Ang II, and COX-1 expression was not altered with either H2O2 or Ang II treatment. *P<0.05 and ***P<0.001 versus control; #P<0.05 compared with Ang II-treated group.

A

B Supplemental Figure XI. Different levels of oxidative stress were induced by Ang II and BMP4 in the endothelial cells. A, Representative images of DHE fluorescence in rat aortic endothelial cells in response to Ang II (0.1 and 1 µmol/L) and BMP4 (20 ng/mL). B, BMP4 induced a higher level of reactive oxygen species compared to Ang II. *P<0.05 versus control; #P<0.05 compared with Ang II-treated group.

0 1 2 4 8 80

10

20

Time (h)

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

COX-2

GAPDH

H2O2

*

***Ang II

0 1 2 4 8 80

10

20

Time (h)

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

COX-2

GAPDH

H2O2

*

***Ang II

Contro

l

Ang II

Losa

rtan

Apocy

nin

Tiron +

DETCA

DMSO

0

1

2

3

Rel

ativ

e flu

ores

cenc

e(F

old

of c

ontro

l)

***

# # #

Ang II

Contro

l

Ang II

Losa

rtan

Apocy

nin

Tiron +

DETCA

DMSO

0

1

2

3

Rel

ativ

e flu

ores

cenc

e(F

old

of c

ontro

l)

***

# # #

Contro

l

Ang II

Losa

rtan

Apocy

nin

Tiron +

DETCA

DMSO

0

1

2

3

Rel

ativ

e flu

ores

cenc

e(F

old

of c

ontro

l)

***

# # #

Ang II

0 1 2 4 8 8 0.0

0.5

1.0

Time (h)

CO

X-1

expr

essi

on(F

old

of c

ontro

l) H2O2 Ang II

COX-1

GAPDH

0 1 2 4 8 8 0.0

0.5

1.0

Time (h)

CO

X-1

expr

essi

on(F

old

of c

ontro

l) H2O2 Ang II

COX-1

GAPDH

Contro

l

Ang II

Apocy

nin

Oxypu

rinol

Tempo

l

Tiron +

DETCA

AMT

L-NAME

0

10

20

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

COX-2GAPDH

***Ang II

Contro

l

Ang II

Apocy

nin

Oxypu

rinol

Tempo

l

Tiron +

DETCA

AMT

L-NAME

0

10

20

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

COX-2GAPDHCOX-2

GAPDH

***Ang II

A B

C D

0 1 2 4 8 80

10

20

Time (h)

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

COX-2

GAPDH

H2O2

*

***Ang II

0 1 2 4 8 80

10

20

Time (h)

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

COX-2

GAPDH

H2O2

*

***Ang II

Contro

l

Ang II

Losa

rtan

Apocy

nin

Tiron +

DETCA

DMSO

0

1

2

3

Rel

ativ

e flu

ores

cenc

e(F

old

of c

ontro

l)

***

# # #

Ang II

Contro

l

Ang II

Losa

rtan

Apocy

nin

Tiron +

DETCA

DMSO

0

1

2

3

Rel

ativ

e flu

ores

cenc

e(F

old

of c

ontro

l)

***

# # #

Contro

l

Ang II

Losa

rtan

Apocy

nin

Tiron +

DETCA

DMSO

0

1

2

3

Rel

ativ

e flu

ores

cenc

e(F

old

of c

ontro

l)

***

# # #

Ang II

0 1 2 4 8 8 0.0

0.5

1.0

Time (h)

CO

X-1

expr

essi

on(F

old

of c

ontro

l) H2O2 Ang II

COX-1

GAPDH

0 1 2 4 8 8 0.0

0.5

1.0

Time (h)

CO

X-1

expr

essi

on(F

old

of c

ontro

l) H2O2 Ang II

COX-1

GAPDH

Contro

l

Ang II

Apocy

nin

Oxypu

rinol

Tempo

l

Tiron +

DETCA

AMT

L-NAME

0

10

20

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

COX-2GAPDH

***Ang II

Contro

l

Ang II

Apocy

nin

Oxypu

rinol

Tempo

l

Tiron +

DETCA

AMT

L-NAME

0

10

20

CO

X-2

expr

essi

on(F

old

of c

ontro

l)

COX-2GAPDHCOX-2

GAPDH

***Ang II

A B

C D

Ang II 0.1 µmol/L

Ang II 1 µmol/L

BMP4

Control

500 µm

Ang II 0.1 µmol/LAng II 0.1 µmol/L

Ang II 1 µmol/LAng II 1 µmol/L

BMP4BMP4BMP4

Control

500 µm

Control

500 µm

Control

500 µm500 µm

Controlmol/L)µ

Ang II (0.1 mol/L)

µ

Ang II (1 BMP4

0

1

2

DH

E flu

ores

cenc

e(F

old

of c

ontro

l)

* **NS#

Controlmol/L)µ

Ang II (0.1 mol/L)

µ

Ang II (1 BMP4

0

1

2

DH

E flu

ores

cenc

e(F

old

of c

ontro

l)

* **NS#

at Chinese University of Hong Kong on April 22, 2011 atvb.ahajournals.orgDownloaded from