mechanism of purinergic activation of endothelial nitric oxide synthase in endothelial cells

Mechanism of Purinergic Activation of Endothelial NitricOxide Synthase in Endothelial Cells

Cleide Gonçalves da Silva, PhD; Anke Specht, MS; Barbara Wegiel, PhD;Christiane Ferran, MD, PhD; Elzbieta Kaczmarek, PhD

Background—Decreased endothelial nitric oxide (NO) synthase (eNOS) activity and NO production are criticalcontributors to the endothelial dysfunction and vascular complications observed in many diseases, including diabetesmellitus. Extracellular nucleotides activate eNOS and increase NO generation; however, the mechanism of thisobservation is not fully clarified.

Methods and Results—To elucidate the signaling pathway(s) leading to nucleotide-mediated eNOS phosphorylation atSer-1177, human umbilical vein endothelial cells were treated with several nucleotides, including ATP, UTP, and ADP,in the presence or absence of selective inhibitors. These experiments identified P2Y1, P2Y2, and possibly P2Y4 as thepurinergic receptors involved in eNOS phosphorylation and demonstrated that this process was adenosine independent.Nucleotide-induced eNOS phosphorylation and activity were inhibited by BAPTA-AM (an intracellular free calciumchelator), rottlerin (a protein kinase C� inhibitor), and protein kinase C� siRNA. In contrast, blockade of AMP-activatedprotein kinase, calcium/calmodulin-dependent kinase II, calcium/calmodulin-dependent kinase kinase, serine/threonineprotein kinase B, protein kinase A, extracellular signal-regulated kinase 1/2, and p38 mitogen-activated protein kinasedid not affect nucleotide-mediated eNOS phosphorylation.

Conclusions—The present study indicates that extracellular nucleotide–mediated eNOS phosphorylation is calcium andprotein kinase C� dependent. This newly identified signaling pathway opens new therapeutic avenues for the treatmentof endothelial dysfunction. (Circulation. 2009;119:871-879.)

Key Words: endothelium � nitric oxide synthase type III � nucleotides � protein kinase C � receptors, purinergic P2

Nitric oxide (NO) is an important signaling molecule,regulating inflammation, angiogenesis, and thrombosis.1

However, in certain diseases, including atherosclerosis anddiabetes mellitus, endothelial NO synthase (eNOS) expressionand activity are strongly attenuated, impairing NO production.2

NO generated and released from the endothelium promotesvasorelaxation, protects endothelial cells (ECs) from apoptosis,inhibits vascular smooth muscle cell proliferation, blocks plate-let activation, and enhances vascular growth. NO synthesisrequires the activation of a family of enzymes called NOSs.Three NOS isoforms have been identified: 1 inducible form and2 constitutive forms, neuronal NOS and eNOS. The expressionand function of eNOS are regulated at multiple levels, namelytranscriptional, posttranscriptional (mRNA stability), and post-translational (O-glycosylation, myristoylation, palmitoylation,and phosphorylation). Additional regulation occurs througheNOS dimerization and interaction with regulatory proteins suchas calmodulin, caveolin-1, and heat-shock protein 90, as well aseNOS subcellular targeting to caveolae, other cellular mem-branes, or the cytoplasm.3,4

Clinical Perspective p 879

eNOS is activated by various stimuli, including thrombin,bradykinin, vascular endothelial growth factor (VEGF), tu-mor necrosis factor-�, histamine, insulin, acetylcholine,endothelin-1, and angiotensin II. Agonist stimulation of thespecific receptors activates phospholipase C�, which is fol-lowed by an increase in diacylglycerol and an increase inintracellular free calcium concentration ([Ca2�]i), activatingcalmodulin. Ca2�/calmodulin binds to the calmodulin-bindingdomain in eNOS and allosterically activates eNOS. In addi-tion, calmodulin can activate calcium/calmodulin-dependentkinase (CaMK) II, which further activates eNOS by phos-phorylation of Ser-1177.5 Although the main signal transduc-tion pathway of agonist-stimulated eNOS activation dependson Ca2�/calmodulin, eNOS activity can be additionally reg-ulated via its phosphorylation by various other kinases,including phosphoinositol 3-kinase (PI3K)/Akt, AMP-acti-vated protein kinase (AMPK), extracellular signal-regulatedkinase 1/2 (ERK), p38 mitogen-activated protein kinase

Received January 9, 2008; accepted November 25, 2008.From the Center for Vascular Biology Research and the Division of Vascular Surgery, Department of Surgery (C.G.d.S., C.F., E.K.), Department of

Medicine (A.S.), and Transplantation Institute, Department of Surgery (B.W.), Beth Israel Deaconess Medical Center, Harvard Medical School, Boston,Mass. A. Specht is now at the Institute of Virology, University of Ulm, Ulm, Germany.

Correspondence to Elzbieta Kaczmarek, Beth Israel Deaconess Medical Center, Harvard Medical School, RN280A, 99 Brookline Ave, Boston, MA02215. E-mail [email protected]

© 2009 American Heart Association, Inc.

Circulation is available at http://circ.ahajournals.org DOI: 10.1161/CIRCULATIONAHA.108.764571

871

Vascular Medicine

by guest on June 28, 2015http://circ.ahajournals.org/Downloaded from

http://circ.ahajournals.org/

(MAPK), protein kinase C (PKC), and protein kinase A(PKA).6–12 Classically, eNOS phosphorylation at Ser-1177 ismandatory for its activation, whereas phosphorylation of Thr-495 by PKC�II has been linked to loss of eNOS activity.13,14

Interestingly, extracellular nucleotides also have been impli-cated in activation of eNOS, leading to NO generation andvasodilation.15 Nucleotides are released into extracellular fluidsby exocytosis from nucleotide-containing granules, byefflux through a membrane transport system in response tocell activation, or as a consequence of cell death. Extra-cellular nucleotides function as paracrine or autocrinemediators via activation of their respective purinergic P2receptors.16 On the basis of molecular structure and func-tion, P2 receptors are classified into 2 main groups: P2X,ligand-gated ion channels, and P2Y, G protein– coupledreceptors.17 Seven P2X receptors subtypes (P2X1 throughP2X7) and 8 P2Y receptor subtypes (P2Y1, P2Y2, P2Y4,P2Y6, and P2Y11 through P2Y14) have been characterizedto date. Although P2X receptors facilitate the entry ofextracellular Ca2�, activation of P2Y receptors results inCa2� release from intracellular stores.18 P2X receptors areexclusively activated by ATP, whereas P2Y receptorsrespond to both purine (ATP, ADP) and pyrimidine (UTP,UDP) nucleotides. Specifically, ADP is the preferentialligand of P2Y1 and P2Y12 receptors; ATP is the ligand forP2Y2, P2Y11, and P2Y13 receptors; UTP is the ligand ofP2Y2 and P2Y4 receptors; UDP is the ligand of P2Y6; andUDP-glucose is the ligand of P2Y14.17 Most cells expressmultiple P2 receptor subtypes. In particular, ECs expressP2Y1, P2Y2, P2Y4, P2Y6, and P2Y11 receptors.19

Activation of the various P2 receptors by nucleotidesinitiates multiple signaling pathways, including the Akt andfocal adhesion kinase/paxillin pathways, as well as theactivation of several kinases such as AMPK, ERK, p38, andJun N-terminal kinase.20–22 This leads to cytoskeletal changesand modulation of expression of several genes involved in theregulation of cell proliferation and apoptosis.

The aim of the present study was to elucidate the signalingpathway(s) involved in nucleotide-mediated phosphorylationof eNOS at Ser-1177. Our results unravel a novel nucleotide-triggered pathway that phosphorylates eNOS by increasing[Ca2�]i and activating PKC�.

MethodsReagentsATP, ADP, UTP, UDP, 2�,3�-O-(4-benzoylbenzoyl)-ATP (BzATP), 1-[N,O-bis (5-isoquinolinesulfonyl)-N-methyl-L-tyrosyl]-4-phenylpiperazine(KN62), 2-(4-morpholinyl)-8-phenyl-1(4H)-benzopyran-4-one hydrochlo-ride (LY294002), wortmannin, adenosine 5�-(�,�-methylene)diphosphate(AOPCP), and 7-Oxo-7H-benzimidazo[2,1-a]benz[de]isoquinoline-3-carboxylic acid (STO-609) were purchased from Sigma-Aldrich Co (StLouis, Mo). Compound C, bisindolylmaleimide I (BIM), myristoylatedPKA inhibitor (amide 14-22 cell permeable), staurosporine, 2�-amino-3�-methoxyflavone (PD98059), NG-nitro-L-arginine methyl ester (L-NAME),4-(4-fluorophenyl)-2-(4-methylsulfinylphenyl)-5-(4-pyridyl)1H-imidazole(SB203580), rottlerin, Gö6983, and 1,2-bis(2-aminophenoxy)ethane-N,N,N�,N�-tetraacetic acid tetraacetoxymethyl ester (BAPTA-AM) werepurchased from Calbiochem/EMD Biosciences, Inc (San Diego, Calif).Recombinant human VEGF was from R&D Systems (Minneapolis, Minn),and the enhanced chemiluminescence kit was from Perkin Elmer LifeScience (Boston, Mass).

Cell CultureHuman umbilical vein ECs (Cambrex, Walkersville, Md) werecultured in an EGM-2 Bullet Kit medium containing human epider-mal growth factor, hydrocortisone, human fibroblast growth factor,VEGF, ascorbic acid, gentamicin, amphotericin-B, human insulin-like growth factor, heparin, and 5% (vol/vol) FBS at 37°C in a 5%CO2 humidified air incubator. Confluent cells at passage 4 or 5 wereused in all experiments.

Cell Treatment and Cell Lysate PreparationCells were incubated with ATP, ADP, and UTP (1 to 250 �mol/L)for various periods of time (1 to 10 minutes). In some experi-ments, cells were preincubated with specific inhibitors for 20minutes. After incubation, cells were placed on ice, washed withice-cold Tris-buffered saline (20 mmol/L Tris, pH 7.5,150 mmol/L NaCl), suspended in cell lysis buffer (20 mmol/LTris, pH 7.5, 100 mmol/L NaCl, 1% [vol/vol] Nonidet P40,1 mmol/L sodium orthovanadate, 100 mmol/L sodium fluoride, 2�g/mL aprotinin, 1 �g/mL leupeptin, 1 mmol/L phenylmethyl-sulfonyl fluoride, 2.5 mmol/L EDTA, and 1 mmol/L EGTA),scraped, incubated on ice for 20 minutes, and centrifuged at�14 000g for 5 minutes at 4°C. Supernatants were kept at �80°Cuntil used for Western blot analysis. Protein concentration wasmeasured by the modified method of Lowry et al,23 using adetergent-compatible protein assay kit (Bio-Rad, Hercules, Calif).

Western BlotCell lysates (30 �g protein per well) were analyzed underreducing conditions by SDS-PAGE performed according to Lae-mmli.24 Proteins were separated on 4% to 15% polyacrylamidegel and transferred to polyvinylidine difluoride membrane bysemidry electroblotting. Membranes were blocked with 5% (wt/vol) nonfat dry milk and probed overnight at 4°C with specific primaryantibodies. Antibodies against human eNOS and phospho-eNOS (Ser-1177), human phospho-AMPK� (Thr-172) (Upstate Laboratories, SanDiego, Calif), human phospho-Akt (Ser-473; Cell Signaling Technol-ogy, Beverly, Mass), human phospho-ERK1/2 (Tyr-204; Santa CruzBiotechnology, Inc, Santa Cruz, Calif), and GAPDH (Calbiochem/EMD Biosciences, Inc) were used. Membranes were incubated withsecondary donkey anti-rabbit– or goat anti-mouse–horseradish peroxi-dase–conjugated antibody (Pierce Biotechnology, Rockford, Ill) for 1hour at room temperature. Protein bands were detected with enhancedchemiluminescence followed by exposure to autoradiography film.Immunoblots were scanned and analyzed with ImageJ 1.33u software(National Institutes of Health, Bethesda, Md).

EC Transfection With siRNACells were transfected with PKC�_7_HP–validated siRNA or anAllStars negative control siRNA (Qiagen, Valencia, Calif) usingLipofectamine 2000 transfection reagent (Invitrogen, Carlsbad,Calif). Briefly, 50% to 60% confluent ECs grown in 6-well cultureplates were transfected in 1 mL Opti MEM I–reduced serum mediumwith 5 �L Lipofectamine 2000 and 100 pmol of the indicated siRNAper well. The cells were incubated at 37°C for 4 to 6 hours, andmedium was changed to growth medium. After 48 hours, cells werestimulated with 100 �mol/L ATP or UTP for 1 minute, harvested,and analyzed by Western blotting.

Measurement of Intracellular cGMPECs were incubated with 100 �mol/L ATP or UTP for 20 minutes at37°C, and intracellular cGMP was measured by a competitiveenzyme immunoassay (Cayman Chemical, Ann Arbor, Mich), ac-cording to the manufacturer’s instructions.

Statistical AnalysisResults are presented as mean�SE. Data were analyzed by 1-wayANOVA followed by the posthoc Duncan multiple range test whenF was significant. Concentration-dependent effects were tested by

872 Circulation February 17, 2009



regression analysis. Differences between groups were rated signifi-cant at values of P�0.05.

The authors had full access to and take full responsibility for theintegrity of the data. All authors have read and agree to themanuscript as written.

ResultsExtracellular Nucleotides Induce eNOSPhosphorylation at Ser-1177ECs were incubated for 1 to 10 minutes with selectednucleotides, ATP, UTP and ADP at concentrations rangingfrom 1 to 250 �mol/L. Results demonstrate that ATP, ADP,and UTP stimulated eNOS phosphorylation in time- andconcentration-dependent manners, with maximal phosphory-lation occurring within 1 to 2 minutes (Figure 1A and 1B).ADP effects were relatively weaker compared with ATP andUTP. In contrast to these nucleotides, similar concentrationsof UDP did not induce eNOS phosphorylation (data notshown).

To evaluate a contribution of adenosine originating fromextracellular nucleotides hydrolyzed by combined actionof ectoenzymes, nucleoside triphosphate diphosphohydro-lases, and 5�-nucleotidase to eNOS phosphorylation, weused an inhibitor of 5�-nucleotidase, AOPCP. Preincuba-tion of ECs with AOPCP (100 �mol/L for 20 minutes) didnot decrease nucleotide-induced phosphorylation of eNOS(Figure 2).

Increase in [Ca2�]i Is Required for eNOSPhosphorylation After Treatment WithExtracellular NucleotidesBecause eNOS is a calcium/calmodulin-dependent enzyme, itis activated by various agonists that increase [Ca2�]i.25 Be-cause one of the first signals of P2 receptor stimulation is an

Figure 1. Extracellular nucleotides phosphorylate eNOS in atime- and concentration-dependent manner in ECs. RepresentativeWestern blots of ECs stimulated with (A) 100 �mol/L ATP, UTP, orADP for 1 to 10 minutes or (B) 1 to 250 �mol/L ATP, UTP, or ADPfor 1 minute. Cell lysates were immunoblotted with anti–phospho-eNOS (P-eNOS; Ser-1177) and anti-eNOS. Results are presentedas mean�SEM of 2 to 5 independent experiments. C indicatesuntreated control cells. *P�0.05, **P�0.01 vs control.

Figure 2. Nucleotide-induced phosphorylation of eNOS is notadenosine dependent. Representative Western blots of ECspretreated for 20 minutes with an inhibitor of 5�-nucleotidase,AOPCP (100 �mol/L), and incubated with or without nucleo-tides (100 �mol/L for 1 minute). Cell lysates were immuno-blotted with anti–phospho-eNOS (P-NOS; Ser-1177) and anti-eNOS antibodies. Results were analyzed statistically andexpressed as mean�SEM of 3 to 4 independent experiments.*P�0.05, **P�0.01 vs control cells.

Gonçalves da Silva et al Purinergic Activation of eNOS 873



increase in [Ca2�]i, we investigated the role of Ca2� in thenucleotide-induced eNOS phosphorylation. ECs were prein-cubated with a chelator of intracellular Ca2�, BAPTA-AM(10 �mol/L for 20 minutes), followed by stimulation with100 �mol/L ATP or UTP for 1 minute. BAPTA-AM signif-icantly attenuated nucleotide-induced phosphorylation ofeNOS (Figure 3A), suggesting that an increase in [Ca2�]i

released from intracellular stores plays an important role inthis activation.

Agonist-induced increase in [Ca2�]i activates CaMK,CaMK II, and CaMK kinase (CaMKK), which could mediateor contribute to eNOS phosphorylation. However, a 20-minute preincubation with KN62 (10 �mol/L), an inhibitor ofCaMK II, or STO-609 (1 �g/mL), a specific inhibitor ofCaMKK, had no effect on nucleotide-induced phosphoryla-

tion of eNOS, demonstrating that these kinases are notinvolved in purinergic activation of eNOS in ECs (Figures 3Band 4A, respectively).

AMPK, Akt, ERK, and p38 MAPK Are NotImplicated in P2 Receptor–MediatedPhosphorylation of eNOSWe have recently shown that extracellular nucleotides acti-vate AMPK in ECs.21 AMPK has been reported as anupstream kinase in the eNOS pathway.7 Accordingly, we

Figure 3. Calcium flux is required in extracellular nucleotide–induced eNOS phosphorylation in ECs. Representative Westernblots of ECs stimulated with ATP or UTP (1 minute at 100�mol/L) following 20-minute pretreatments with (A) an intracellu-lar calcium chelator, BAPTA-AM (10 �mol/L), or (B) an inhibitorof CaMK II, KN62 (10 �mol/L). Cell lysates were immunoblottedwith antibodies to phospho-eNOS (P-eNOS; Ser-1177) andeNOS. Results were analyzed statistically and expressed asmean�SEM of 3 to 4 independent experiments. **P�0.01 vscontrol cells (C).

Figure 4. Extracellular nucleotide–induced phosphorylation ofeNOS is independent of AMPK. Representative Western blotsof ECs pretreated for 20 minutes with either an inhibitor ofCaMKK, STO-609 (1 �g/�L), or an inhibitor of AMPK, com-pound C (CC; 20 �mol/L). Cell lysates were immunoblottedwith anti–phospho-eNOS (P-eNOS; Ser-1177), anti–phospho-AMPK (P-AMPK; Thr-172), and anti-GAPDH antibodies.Results were analyzed statistically and expressed asmean�SEM of 3 independent experiments. *P�0.05,**P�0.01 vs control (C).




evaluated a possible role of AMPK in purinergic activation ofeNOS. Inhibition of the AMPK upstream activator CaMKKwith STO-609 (1 �g/mL for 20 minutes) or AMPK withcompound C (20 �mol/L for 20 minutes) did not affectnucleotide-induced phosphorylation of eNOS (Figure 4A and4B). The inhibitory effects of STO-609 and compound Cwere confirmed by lack of phosphorylation of AMPK (Figure4A and 4B).

Akt is another potential kinase involved in eNOSphosphorylation at Ser-1177 in ECs.6 However, in ECs,nucleotide-induced eNOS phosphorylation usually pre-cedes that of Akt (1 to 2 minutes versus 5 to 10 minutes),which would argue against the implication of Akt in thissystem. The lack of involvement of Akt was directlyvalidated by our data showing that preincubation of ECswith LY294002 (10 �mol/L for 20 minutes), a selectiveinhibitor of the Akt upstream kinase, PI3K, did not affectphosphorylation of eNOS by ATP or UTP (Figure 5A).

Other signaling molecules, including ERK and p38 MAPKthat have been implicated in eNOS phosphorylation, also areactivation targets of P2 receptor signaling.22 ECs were preincu-bated for 20 minutes with the ERK and p38 MAPK inhibitorsPD98059 (50 �mol/L; inhibits upstream ERK kinase) andSB203580 (10 �mol/L), respectively. None of these inhibitorsdecreased eNOS phosphorylation in response to extracellularnucleotides (Figure 5B and data not shown).

The PKC� Isoform Is Responsible forNucleotide-Mediated eNOS Phosphorylation andActivation in ECPKC has been shown to increase eNOS phosphorylation andactivity in ECs.10,11 P2Y receptors activate phospholipase C,which hydrolyzes phosphatidylinositol bisphosphate into IP3and diacylglycerol to induce PKC activation. We investigatedthe role of PKCs in nucleotide-mediated eNOS phosphoryla-tion using pharmacological inhibitors of the various PKCisoforms.

Staurosporine (200 nmol/L; 20-minute preincubation),which originally was recognized as a PKC inhibitor (IC50, 5nmol/L), completely inhibited nucleotide-induced eNOSphosphorylation (Figure 6A). However, staurosporine has amuch broader specificity and may inhibit PKA (IC50, 15nmol/L), protein kinase G (PKG; IC50, 18 nmol/L), and/orCaMK II (IC50, 20 nmol/L). We excluded any role for PKAand PKG in nucleotide-induced phosphorylation of eNOS byusing their respective inhibitors, 14-22 amide (PKA inhibitor;10 �mol/L) and PKG inhibitor (100 �mol/L), which did notattenuate eNOS phosphorylation (Figure 6B and data notshown). In contrast, we confirmed the participation of PKC ineNOS phosphorylation by using BIM, a non–subtype-selective PKC inhibitor. Preincubation with BIM (5 �mol/Lfor 20 minutes) significantly decreased eNOS phosphoryla-tion in ECs (Figure 6C).

Having confirmed PKC involvement in nucleotide-induced phosphorylation of eNOS, we sought to identifythe PKC isoform that was responsible. We focused onPKC� and PKC�, which have previously been implicatedin eNOS activation in ECs.10,11 ECs were preincubated

with rottlerin (5 �mol/L for 20 minutes), a PKC�-selectiveinhibitor, and Gö6976 (1 �mol/L for 20 minutes), a PKC�-specific inhibitor. Rottlerin significantly reduced nucleotide-mediated phosphorylation of eNOS, whereas Gö6976 had noeffect (Figure 6D and 6E). In addition, we verified that extra-cellular nucleotides induced phosphorylation of PKC� at Thr-505, a residue localized in the activation loop of the catalyticdomain of this enzyme (Figure 6E, inset).

Finally, to confirm participation of PKC� in nucleotide-mediated eNOS phosphorylation, we used siRNA specific toPKC�. Transfection of ECs with PKC� siRNA significantly

Figure 5. Extracellular nucleotide–induced eNOS phosphory-lation is independent of PI3K/Akt and ERK. RepresentativeWestern blots of ECs pretreated for 20 minutes with (A) aninhibitor of PI3K, LY294002 (LY; 10 �mol/L), or (B) an ERKinhibitor, PD 98059 (50 �mol/L), and stimulated for 1 minutewith nucleotides (100 �mol/L). Cell lysates were immunoblot-ted with antibodies to phospho-eNOS (P-eNOS; Ser-1177),phospho-Akt (P-Akt; Ser-473), phospho-ERK (P-ERK), andGAPDH. Results were expressed as mean�SEM of 3 to 5independent experiments. **P�0.01 vs control (C).




reduced extracellular nucleotide–induced eNOS phosphory-lation, whereas nontarget, negative control siRNA had noeffect (Figure 6F).

To examine whether eNOS phosphorylation in responseto nucleotides corresponds to eNOS activation and NOgeneration, we measured NO-induced cGMP production

by ELISA 20 minutes after stimulation with nucleotides.ATP/UTP treatment increased intracellular cGMP, reach-ing levels comparable to those obtained after addition ofVEGF (50 ng/mL), a well-known eNOS activator in ECs(Figure 6G). cGMP accumulation in control and nucleo-tide-stimulated ECs was abolished by preincubation with

Figure 6. Extracellular nucleotide–in-duced eNOS phosphorylation and acti-vation are mediated by PKC� in ECs.Representative Western blots of ECspretreated for 20 minutes with (A) stau-rosporine (ST; 200 nmol/L), (B) inhibitorof PKA (PKAi; 5 �mol/L), (C) PKC inhibi-tor (BIM; 5 �mol/L), (D) isoform-specificPKC� inhibitor (Gö6976; 1 �mol/L), or(E) the PKC�-specific inhibitor rottlerin(ROT; 5 �mol/L) and incubated with orwithout nucleotides (100 �mol/L for 1minute). Activation of PKC� by extra-cellular nucleotides was confirmed byits phosphorylation at Thr-505 (E,inset). Cell lysates were immunoblottedwith anti–P-eNOS (Ser-1177), anti-eNOS, anti–phospho-PKC� (P-PKC�;Thr-505), and anti-PKC� antibodies.Results were statistically analyzed andexpressed as mean�SEM of 3 to 7independent experiments. �P�0.05 vsROT. F, Representative Western blotsof nontransfected (NT) ECs and ECstransfected for 48 hours with PKC�-specific or negative control (�) siRNAand incubated with or without nucleo-tides (100 �mol/L for 1 minute). Cell lysateswere immunoblotted with antibodies toP-eNOS (Ser-1177), eNOS, PKC�, andGAPDH (loading control). Results were sta-tistically analyzed and expressed asmean�SEM of 4 independent experiments.G, eNOS activity, expressed as cGMP lev-els, is PKC� dependent. ECs were incu-bated for 20 minutes with 100 �mol/L ATPor UTP, with and without L-NAME (100�mol/L) or rottlerin (ROT; 5 �mol/L), andcGMP production was measured by theenzyme immunoassay. VEGF (50 ng/mL)was used as a positive control. Results wereanalyzed statistically and expressed asmean�SEM of 2 to 4 independent experi-ments. *P�0.05, **P�0.01 vs control.




the eNOS inhibitor L-NAME (100 �mol/L for 20 minutes),indicating that cGMP is a proper marker of NO production.PKC� inhibition by rottlerin significantly attenuated nu-cleotide-induced cGMP production (Figure 6G).

DiscussionThe participation of extracellular nucleotides in vessel relax-ation has been described and accounted for by NO generationin response to a conformational activation of eNOS byCa2�/calmodulin.26,27 Besides conformation-based activation,eNOS activity is modulated by its phosphorylation at Ser-1177. Phosphorylation of this residue is recognized as apositive regulator of eNOS activity.6 The impact of extracel-lular nucleotides on eNOS phosphorylation has not beenclearly defined. In this study, we investigated the mecha-nism(s) of P2 receptor signaling leading to eNOS phosphor-ylation and activation in ECs.

We demonstrate that P2 receptor–mediated eNOS phos-phorylation at Ser-1177 in ECs is calcium and PKC� depen-dent. We also report that AMPK, Akt, PKA, PKG, CaMK II,CaMKK, p38, and ERK, which are activated by extracellularnucleotides and potential upstream kinases for eNOS, do notparticipate in this process.

Our results demonstrate that the extracellular nucleotidesATP, UTP, and ADP induce eNOS phosphorylation in time-and concentration-dependent manners (Figure 1). Further-more, we established that the prompt nucleotide-mediatedeNOS phosphorylation is strictly nucleotide and not adeno-sine dependent (Figure 2).

One of the first cellular responses to extracellularnucleotides is an increase in [Ca2�]i, which can originateeither from the cell milieu (by P2X receptor signaling) orfrom the intracellular stores (by P2Y receptor signaling).In previous work, we demonstrated that in ECs Ca2� isreleased from intracellular stores (endoplasmic reticulum)after stimulation with extracellular nucleotides.20 This ledus to conclude that in ECs extracellular nucleotides signalmostly via P2Y receptors. Accordingly, extracellular nu-cleotides are likely to phosphorylate eNOS by engagingP2Y receptors. This conclusion is additionally supportedby data showing that nucleotide-induced and NO-dependent vasodilation in human vessels is associated withP2Y and not P2X receptors.27 Here, we examined theeffects of ATP (ligand for P2Y2 and P2Y11 receptors),ADP (ligand for P2Y1 and P2Y12 receptors, the latter notexpressed in ECs), UTP (ligand for P2Y2 and P2Y4receptors), and UDP (ligand for P2Y6 receptors) on eNOSphosphorylation. Our results indicate that P2Y1, P2Y2,and possibly P2Y4 receptors are the main receptors in-volved in purinergic activation of eNOS in ECs. Weexcluded P2Y6 and P2Y11 receptors because their respec-tive ligands, UDP and BzATP, did not induce eNOSphosphorylation (data not shown).

Our results also demonstrate a crucial role for Ca2� innucleotide-induced eNOS phosphorylation. Indeed, eNOS phos-phorylation was totally inhibited by chelating intracellular Ca2�

(Figure 3A). These results suggest that increased [Ca2�]i isnecessary for the activation of an upstream kinase involved in

eNOS phosphorylation and/or that eNOS cannot be phosphory-lated unless calcium-related conformational changes have takenplace. Other studies showing that mobilization of [Ca2�]i is acrucial step of eNOS activation support our conclusion.25

Increased [Ca2�]i is usually associated with the activationof CaMK II and CaMKK, which, as stated earlier, could actas upstream kinases for eNOS.5 However, in our experimen-tal system, these enzymes did not participate in the phosphor-ylation of eNOS (Figures 3B and 4A), even though weconfirmed in previous work that extracellular nucleotides doactivate CaMKK in ECs.21

There are contradictory results with regard to the role ofAMPK in eNOS phosphorylation in ECs. AMPK has beenreported as a direct upstream activator of eNOS phosphor-ylation in ECs stimulated with thrombin and histamine.7

However, AMPK-independent activation of eNOS alsowas described in ECs treated with thrombin.11,28 Recently,we have shown that extracellular nucleotides activateAMPK in ECs.21 Therefore, we hypothesized that AMPKcould act as an upstream activator of eNOS in our system.Our results disproved this hypothesis. Inhibition of AMPKactivity with its pharmacological inhibitor, compound C, atten-uated phosphorylation of AMPK but had no effect on nucleo-tide-induced eNOS phosphorylation (Figure 4B). Likewise,inhibition of CaMKK, an upstream AMPK activator, withSTO-609 did not affect eNOS phosphorylation (Figure 4A).Moreover, dominant-negative AMPK�2 did not decrease eNOSphosphorylation by extracellular nucleotides (data not shown).Although we have documented in a previous study that extra-cellular nucleotides do activate AMPK in ECs,21 in our experi-mental system, AMPK does not participate in the phosphoryla-tion of eNOS.

Similarly, the participation of Akt in phosphorylation ofhuman eNOS at Ser-1177 was shown in ECs treated withvarious agonists, including VEGF, thrombin, and histamine.6

However, here again, nucleotide-induced eNOS phosphory-lation was not decreased by pretreatment of ECs with thePI3K inhibitors LY294002 (Figure 5A) and wortmannin (datanot shown), excluding any participation of the PI3K and Aktkinases in this pathway. We also excluded ERK, p38, PKA,and PKG from any involvement in nucleotide-induced eNOSphosphorylation (Figure 5B and data not shown).

To scrutinize other potential kinases that are involved innucleotide-induced eNOS phosphorylation, we investi-gated participation of the PKC isoforms in purinergicsignaling. PKC is a family of serine-threonine–specifickinases, which is divided into 3 groups: the conventionalPKCs (�, �I, �II, and �), the novel PKCs (�, �, �, and �)and the atypical PKCs ( and ).29 The conventional PKCsdepend on calcium and are activated by diacylglycerol; thenovel PKCs are calcium independent (but some physiolog-ical concentration of Ca2� is needed) and activated bydiacylglycerol; the atypical PKCs are calcium independentand not activated by diacylglycerol.30 Different PKCisoforms have been shown to exert opposite effects oneNOS activity. Although PKC�II decreases eNOS activityby phosphorylating Thr-495,13,31 PKC� and PKC� increaseeNOS phosphorylation and activity in ECs in response toagonists such as thrombin.10,11




In this study, we used an array of various pharmacologicalinhibitors of PKCs: staurosporine, a selective blocker of PKC;BIM, a PKC inhibitor; rottlerin, a PKC� inhibitor; and Gö6976,a PKC� inhibitor. Although staurosporine, BIM, and rottlerinsignificantly inhibited nucleotide-mediated phosphorylation ofeNOS, with rottlerin also diminishing cGMP levels, Gö6976 didnot affect eNOS phosphorylation (Figure 6). In addition, siRNAto PKC� significantly attenuated eNOS phosphorylation inducedby nucleotides. These combined data indicate that PKC� is themain kinase activating eNOS in ECs exposed to extracellularnucleotides.

Accordingly, we propose that on coupling of P2Y1 and P2Y2

receptors to phospholipase C via G�q proteins, diacylglycerol isgenerated and the level of [Ca2�]i is increased, enabling calmod-ulin and PKC� activation and subsequent eNOS phosphoryla-tion, as depicted in Figure 7. Our demonstration of this pathwayclarifies the mechanisms involved in nucleotide-induced eNOSactivation through combined Ca2�-dependent conformationalmodification, as well as PKC�-dependent phosphorylation ofeNOS.

ConclusionsOur observations strongly suggest that the phosphorylation ofeNOS that follows the treatment of ECs with ATP, UTP, andADP occurs via an increase in the intracellular calcium levelsand the activation of PKC�. We suggest that P2 receptors, byactivating eNOS and increasing NO generation, represent noveltherapeutic targets for the prevention and/or treatment of endo-thelial cell dysfunction associated with vascular diseases.

Sources of FundingThis work was supported in part by the National Institutes of Health(HL66167 to Dr Kaczmarek and HL080130 to Dr Ferran) and theJuvenile Diabetes Research Foundation (5-2007–736 to Dr Kacz-marek, 3-2006-617 to Dr Gonçalves da Silva, and 5-2005-1276 and1-2007-567 to Dr Ferran).

DisclosuresNone.

References1. Loscalzo J, Welch G. Nitric oxide and its role in the cardiovascular

system. Prog Cardiovasc Dis. 1995;38:87–104.2. Sessa WC. Regulation of endothelial derived nitric oxide in health and

disease. Mem Inst Oswaldo Cruz. 2005;100(suppl 1):15–18.3. Dudzinski DM, Michel T. Life history of eNOS: partners and pathways.

Cardiovasc Res. 2007;75:247–260.4. Fulton D, Gratton JP, Sessa WC. Post-translational control of endothelial

nitric oxide synthase: why isn’t calcium/calmodulin enough?J Pharmacol Exp Ther. 2001;299:818–824.

5. Schneider JC, El Kebir D, Chereau C, Lanone S, Huang XL, De BuysRoessingh AS, Mercier JC, Dall’Ava-Santucci J, Dinh-Xuan AT.Involvement of Ca2�/calmodulin-dependent protein kinase II in en-dothelial NO production and endothelium-dependent relaxation.Am J Physiol Heart Circ Physiol. 2003;284:H2311–H2319.

6. Dimmeler S, Fleming I, Fisslthaler B, Hermann C, Busse R, Zeiher AM.Activation of nitric oxide synthase in endothelial cells by Akt-dependentphosphorylation. Nature. 1999;399:601–605.

7. Thors B, Halldorsson H, Thorgeirsson G. Thrombin and histaminestimulate endothelial nitric-oxide synthase phosphorylation at Ser1177via an AMPK mediated pathway independent of PI3K-Akt. FEBS Lett.2004;573:175–180.

8. Zheng J, Wen Y, Chen DB, Bird IM, Magness RR. Angiotensin IIelevates nitric oxide synthase 3 expression and nitric oxide productionvia a mitogen-activated protein kinase cascade in ovine fetoplacentalartery endothelial cells. Biol Reprod. 2005;72:1421–1428.

9. Anter E, Thomas SR, Schulz E, Shapira OM, Vita JA, Keaney JF Jr.Activation of endothelial nitric-oxide synthase by the p38 MAPK inresponse to black tea polyphenols. J Biol Chem. 2004;279:46637– 46643.

10. Partovian C, Zhuang Z, Moodie K, Lin M, Ouchi N, Sessa WC, Walsh K,Simons M. PKCalpha activates eNOS and increases arterial blood flow invivo. Circ Res. 2005;97:482–487.

11. Motley ED, Eguchi K, Patterson MM, Palmer PD, Suzuki H, Eguchi S.Mechanism of endothelial nitric oxide synthase phosphorylation andactivation by thrombin. Hypertension. 2007;49:577–583.

12. Sessa WC. eNOS at a glance. J Cell Sci. 2004;117:2427–2429.13. Matsubara M, Hayashi N, Jing T, Titani K. Regulation of endothelial

nitric oxide synthase by protein kinase C. J Biochem. 2003;133:773–781.14. Fleming I, Busse R. Molecular mechanisms involved in the regulation of

the endothelial nitric oxide synthase. Am J Physiol Regul Integr CompPhysiol. 2003;284:R1–R12.

15. Buvinic S, Briones R, Huidobro-Toro JP. P2Y(1) and P2Y(2) receptorsare coupled to the NO/cGMP pathway to vasodilate the rat arterialmesenteric bed. Br J Pharmacol. 2002;136:847–856.

16. Dubyak GR, el-Moatassim C. Signal transduction via P2-purinergicreceptors for extracellular ATP and other nucleotides. Am J Physiol.1993;265:C577–C606.

17. Burnstock G. Purine and pyrimidine receptors. Cell Mol Life Sci. 2007;64:1471–1483.

18. Brake AJ, Wagenbach MJ, Julius D. New structural motif forligand-gated ion channels defined by an ionotropic ATP receptor. Nature.1994;371:519–523.

19. Wang L, Karlsson L, Moses S, Hultgardh-Nilsson A, Andersson M,Borna C, Gudbjartsson T, Jern S, Erlinge D. P2 receptor expressionprofiles in human vascular smooth muscle and endothelial cells. J Car-diovasc Pharmacol. 2002;40:841–853.

20. Kaczmarek E, Erb L, Koziak K, Jarzyna R, Wink MR, GuckelbergerO, Blusztajn JK, Trinkaus-Randall V, Weisman GA, Robson SC.Modulation of endothelial cell migration by extracellular nucleotides:involvement of focal adhesion kinase and phosphatidylinositol3-kinase-mediated pathways. Thromb Haemost. 2005;93:735–742.

21. da Silva CG, Jarzyna R, Specht A, Kaczmarek E. Extracellular nucleo-tides and adenosine independently activate AMP-activated protein kinasein endothelial cells: involvement of P2 receptors and adenosine trans-porters. Circ Res. 2006;98:e39–e47.

22. Erb L, Liao Z, Seye CI, Weisman GA. P2 receptors: intracellular sig-naling. Pflugers Arch. 2006;452:552–562.

23. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein mea-surement with the Folin phenol reagent. J Biol Chem. 1951;193:265–275.

24. Laemmli UK. Cleavage of structural proteins during the assembly of thehead of bacteriophage T4. Nature. 1970;227:680–685.

25. Michel JB, Feron O, Sacks D, Michel T. Reciprocal regulation of endo-thelial nitric-oxide synthase by Ca2�-calmodulin and caveolin. J BiolChem. 1997;272:15583–15586.

26. Mathie RT, Ralevic V, Alexander B, Burnstock G. Nitric oxide is themediator of ATP-induced dilatation of the rabbit hepatic arterial vascularbed. Br J Pharmacol. 1991;103:1602–1606.

27. Kalinowski L, Dobrucki LW, Szczepanska-Konkel M, Jankowski M,Martyniec L, Angielski S, Malinski T. Third-generation beta-blockersstimulate nitric oxide release from endothelial cells through ATP

Figure 7. Proposed signaling pathway of the extracellular nucle-otide–induced phosphorylation of eNOS in ECs.




efflux: a novel mechanism for antihypertensive action. Circulation.2003;107:2747–2752.

28. Stahmann N, Woods A, Carling D, Heller R. Thrombin activates AMP-activated protein kinase in endothelial cells via a pathway involvingCa2�/calmodulin-dependent protein kinase kinase beta. Mol Cell Biol.2006;26:5933–5945.

29. Parker PJ, Murray-Rust J. PKC at a glance. J Cell Sci. 2004;117:131–132.

30. Kheifets V, Mochly-Rosen D. Insight into intra- and inter-molecularinteractions of PKC: design of specific modulators of kinase function.Pharmacol Res. 2007;55:467–476.

31. Lin MI, Fulton D, Babbitt R, Fleming I, Busse R, Pritchard KA Jr, Sessa WC.Phosphorylation of threonine 497 in endothelial nitric-oxide synthase coor-dinates the coupling of L-arginine metabolism to efficient nitric oxide pro-duction. J Biol Chem. 2003;278:44719–44726.

CLINICAL PERSPECTIVEDecreased endothelial nitric oxide (NO) synthase activity and NO production are critical contributors to the endothelialdysfunction and vascular complications observed in many diseases, including atherosclerosis, diabetes mellitus, andhypertension. Here, we report a new pathway initiated by extracellular nucleotides and P2 purinergic receptors leading toendothelial NO synthase activation and NO generation. This signaling pathway does not involve activation of PI3K/Akt,extracellular signal-regulated kinase 1/2, or AMP-activated protein kinase, to name only a few kinases, the roles of whichin endothelial NO synthase activation are well established. Instead, extracellular nucleotide–mediated endothelial NOsynthase phosphorylation is calcium and protein kinase C� dependent. Extracellular nucleotides play a significantbiological role in many tissues and cell types as signaling molecules that regulate cellular functions under both normal andpathophysiological conditions. Extracellular nucleotides exert their biological action via specific purinergic P2 receptorsthat are classified into 2 main groups: P2X, ligand-gated ion channels, and P2Y, G protein–coupled receptors. Most cells,including endothelial cells, express multiple P2 receptor subtypes. Many laboratories are developing chemical agonists andantagonists of P2 receptors to be used as potential pharmacological agents. Clopidogrel, a blocker of platelet P2Y12receptor, has been used successfully as an antithrombotic drug. Similarly, denufosol tetrasodium, a nucleotide analogdesigned to treat cystic fibrosis, is in clinical trials. This indicates that P2 receptors, despite their ubiquitous distribution,constitute useful pharmacological targets for the management of various ailments. On the basis of our data, we propose thatP2 receptors also can be new pharmacological targets for the treatment of endothelial dysfunction.




KaczmarekCleide Gonçalves da Silva, Anke Specht, Barbara Wegiel, Christiane Ferran and Elzbieta

CellsMechanism of Purinergic Activation of Endothelial Nitric Oxide Synthase in Endothelial

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 2009 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation doi: 10.1161/CIRCULATIONAHA.108.764571

2009;119:871-879; originally published online February 2, 2009;Circulation.

http://circ.ahajournals.org/content/119/6/871World Wide Web at:

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

http://circ.ahajournals.org//subscriptions/

is online at: Circulation Information about subscribing to Subscriptions:

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

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

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

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


http://circ.ahajournals.org/content/119/6/871

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

http://www.lww.com/reprints

http://circ.ahajournals.org//subscriptions/


mechanism of purinergic activation of endothelial nitric oxide synthase in endothelial cells

Documents