Signaling Pathways Involved in Adenosine Triphosphate-Induced Endothelial Cell Barrier Enhancement

Signaling Pathways Involved in AdenosineTriphosphate-Induced Endothelial Cell

Barrier EnhancementIrina A. Kolosova, Tamara Mirzapoiazova, Djanybek Adyshev, Peter Usatyuk, Lewis H. Romer,

Jeffrey R. Jacobson, Viswanathan Natarajan, David B. Pearse, Joe G.N. Garcia, Alexander D. Verin

Abstract—Endothelial barrier dysfunction caused by inflammatory agonists is a frequent underlying cause of vascular leakand edema. Novel strategies to preserve barrier integrity could have profound clinical impact. Adenosine triphosphate(ATP) released from endothelial cells by shear stress and injury has been shown to protect the endothelial barrier in somesettings. We have demonstrated that ATP and its nonhydrolyzed analogues enhanced barrier properties of culturedendothelial cell monolayers and caused remodeling of cell–cell junctions. Increases in cytosolic Ca2� and Erk activationcaused by ATP were irrelevant to barrier enhancement. Experiments using biochemical inhibitors or siRNA indicatedthat G proteins (specifically G�q and G�i2), protein kinase A (PKA), and the PKA substrate vasodilator-stimulatedphosphoprotein were involved in ATP-induced barrier enhancement. ATP treatment decreased phosphorylation ofmyosin light chain and specifically activated myosin-associated phosphatase. Depletion of G�q with siRNA preventedATP-induced activation of myosin phosphatase. We conclude that the mechanisms of ATP-induced barrier enhancementare independent of intracellular Ca2�, but involve activation of myosin phosphatase via a novel G-protein–coupledmechanism and PKA. (Circ Res. 2005;97:115-124.)

Key Words: endothelial barrier � extracellular adenosine triphosphate � G protein � myosin phosphatase

Inflammatory agonist-induced endothelial cell (EC) barrierdysfunction is associated with cytoskeletal remodeling,

disruption of cell–cell contacts, and the formation of para-cellular gaps.1 Less is known about the mechanisms of ECbarrier maintenance and protection. Some naturally occurringsubstances such as sphingosine 1-phosphate, angiotensin 1,and the second messenger cAMP are known to enhance theEC barrier. Recently, much attention has been given to thetherapeutic potential of purinergic agonists and antagonistsfor the treatment of cardiovascular and pulmonary diseases.2

Accumulated experimental data suggest that adenosinetriphosphate (ATP) and other purines are promising asphysiologically-relevant barrier-protective agents as they arereadily present in the surrounding EC microenvironment invivo, and they decrease transendothelial permeability in vitro.ATP can be released into the bloodstream from platelets3 andred blood cells.4 Extracellular ATP concentrations may tem-porarily exceed 100 �mol/L in blood.5 Furthermore, theendothelium provides a source of ATP locally within vascularbeds. ATP is released constitutively across the apical mem-brane of EC under basal conditions.6 Enhanced release ofATP was observed from EC in response to various stimuliincluding hypotonic challenge,6 calcium agonists,6 shear

stress,7 thrombin,7 ATP itself,8 and lipopolysaccharide.9 Oncereleased, ATP is degraded rapidly and its metabolites, aden-osine diphosphate (ADP) and adenosine, have also beencharacterized as signaling molecules, able to regulate variouscellular functions.10

Extracellular nucleotides and adenosine act via purinore-ceptors, which are divided into 2 classes: P1, or adenosinereceptors, and P2 receptors, that recognize extracellular ATP,ADP, uridine 5�-triphosphate (UTP), and uridine 5�-diphosphate (UDP).10 Four different P1 receptors have beenidentified and pharmacologically characterized: A1, A2A, A2B,and A3.11 The A2A and A2B receptors preferably interact withmembers of the Gs family of G proteins and the A1 and A3

receptors with Gi/o proteins.11 The P2 receptors are dividedinto 2 subclasses, X and Y. P2X receptors are ATP-gatednonselective cation channels.12 The P2Y receptors areG-protein coupled. P2Y1, 2, 4, 6, and 11 are coupled to Gq

and activate PLC�. P2Y12, 13, and 14 are coupled to Gi andinhibit adenylate cyclase.10

The expression of purinoreceptors in human EC is variableand dependent on the specific EC type. Among P2 receptors,P2X46,13,14 and P2Y213–15 are the most abundant and widelyexpressed in different EC. Wang et al14 also demonstrated

Original received January 5, 2005; revision received June 9, 2005; accepted June 20, 2005.From the Department of Medicine (I.A.K., T.M., D.A., P.U., J.R.J., V.N., D.B.P., J.G.N.G., A.D.V.), Division of Pulmonary and Critical Care

Medicine, and Departments of Anesthesiology (L.H.R.), Johns Hopkins University School of Medicine, Baltimore, Md.This manuscript was sent to Donald Heistad, Consulting Editor, for review by expert referees, editorial decision, and final disposition.Correspondence to Dr Alexander D. Verin, 5200 Eastern Ave, MFL Bldg, Center Tower, Room 660, Baltimore, MD 21224. E-mail [email protected]© 2005 American Heart Association, Inc.

Circulation Research is available at http://circres.ahajournals.org DOI: 10.1161/01.RES.0000175561.55761.69

115

Molecular Medicine

by guest on September 16, 2015http://circres.ahajournals.org/Downloaded from by guest on September 16, 2015http://circres.ahajournals.org/Downloaded from by guest on September 16, 2015http://circres.ahajournals.org/Downloaded from by guest on September 16, 2015http://circres.ahajournals.org/Downloaded from by guest on September 16, 2015http://circres.ahajournals.org/Downloaded from by guest on September 16, 2015http://circres.ahajournals.org/Downloaded from

http://circres.ahajournals.org/






that human umbilical vein endothelial cells (HUVEC) ex-press a high level of P2Y1 and P2Y11. The P2Y expressionprofile suggests that nucleotide signaling in EC is likelymediated specifically via G�q pathway. Recently P2Y12 hasbeen characterized in rat EC,16 implicating G�i-mediatednucleotide signaling. Adenosine receptors also have beenfound in human EC.17

Recently the barrier-protective properties of ATP havebeen reported in HUVEC, bovine, and porcine EC.18,19 Theexact nature of ATP-induced barrier augmentation is not welldefined. In this study we examined the mechanisms of ECbarrier enhancement caused by extracellular ATP using acombination of pharmacological and molecular approaches.We sought to define the molecular components couplingreceptor activation with barrier enhancement.

Materials and MethodsSources of reagents and details of procedures are provided in theexpanded Materials and Methods section in the online data supple-ment available at http://circres.ahajournals.org. Human and bovinepulmonary artery EC (HPAEC; Clonetics, Walkersville, Md andBPAEC; American Type Tissue Culture Collection, Rockville, Md,respectively), and human lung microvascular EC (HLMVEC; Clo-netics) were used in the study. siRNA-based protein depletion ofsmall GTPases were performed as described elsewhere.20 The barrierproperties of EC monolayers were characterized using electrical cellimpedance sensor system.21 Described immunostaining protocol wasused.22 The percentage of total cell surface area occupied byVE-cadherin labeled cell–cell junctions was quantitatively deter-mined using Openlab (4.0) software (Improvision). Concentrationsof cytosolic Ca2� were measured as described previously.23 cAMPconcentration in EC lysates was determined with TRK 432 radioas-say system (Amersham). PKA activity was measured using anonradioactive PKA Kinase Activity assay Kit (Stressgen). Myosin-enriched fraction of HPAEC was prepared as described previously.24

Ser/Thr Phosphatase Assay Kit (Upstate) was used to determinemyosin light chain phosphotase (MLCP) activity. MLC phosphory-lation was analyzed by either phosphospecific antibody or ureapolyacrylamide gel electrophoresis as previously described.24 Forbasic statistical analysis a GraphPad Prism Program was used. Datawere compared by a Student t test. Probability values �0.05 wereconsidered to be significant. Values are expressed as mean�SE.

ResultsATP Increases TransendothelialElectrical ResistanceATP increased the transendothelial electrical resistance(TER) of HPAEC monolayers in a concentration-dependent

manner (Figure 1A). ADP, another nonselective P2 receptoragonist, and the stable ATP analogs ATP-�-S and 2-MeS-ATP also increased TER. AMP-CCP, which is more specificfor the P2X1 and P2X3, receptors was completely inactive(Figure 1B). Other types of EC (HPAEC, HLMVEC, andBPAEC) demonstrated similar responses to ATP stimulationcharacterized by increased TER (online Table I).

ATP Affects Cell–Cell JunctionsImmunofluorescence studies revealed changes in distribu-tions of cell–cell junctional proteins after ATP treatment.VE-cadherin, a major component of endothelial adherensjunctions, was more pronounced at the cellular periphery,presumably at cell–cell contacts (Figure 2A). The calculatedpercentage of total cell surface area occupied by VE-cad-herin–labeled cell–cell junctions confirmed that ATP induceda significant increase in the surface area of cell–cell inter-faces as a percentage of total cell surface area (Figure 2B).Furthermore, whereas the tight junctional component zonulaoccludens-1 (ZO-1) had a thin and somewhat discontinuouspattern at cell–cell borders of unstimulated monolayers, athicker, more regular and continuous ZO-1 distribution wasobserved after ATP treatment (Figure 2C). This rearrange-ment of ZO-1 is consistent with a tightening of permeabilitybarrier.

ATP-Induced Increases in EC Barrier FunctionAre Independent of Changes in Cytosolic Ca2�

Purenergic stimulation is known to increase intracellular Ca2�

concentrations.10 However, inflammatory agonists that in-crease Ca2� compromise EC barrier function.25 It has beenpreviously reported that ATP decreases transendothelial al-bumin permeability despite increase on intracellular Ca2�.18

In our experiments, ATP increased Ca2� in HPAEC (Figure3A) in a dose-dependent fashion. 10 �mol/L 1,2-bis(0-Aminophenoxy)ethane-N,N,N�,N�-tetraacetic Acid Tet-ra(acetoxymethyl) Ester (BAPTA) completely inhibited theATP-induced Ca2� increase (Figure 3B). BAPTA itselfcaused a decrease in TER, followed by recovery, but did notaffect the TER response to ATP (Figure 3C). Enhancement ofVE-cadherin–mediated intercellular junctions by ATP wasalso unaffected (Figure 3D). Therefore ATP-induced en-

Figure 1. Effect of purinergic stimulation on permeability of HPAEC monolayer. A, Dose-dependent effect of ATP on TER. B, Agonistsof P2 receptors (50 �mol/L each) increase TER (n�4; *P�0.01, **P�0.05 compared with vehicle).

116 Circulation Research July 22, 2005

by guest on September 16, 2015http://circres.ahajournals.org/Downloaded from


hancement of endothelial barrier and adherens junctions isindependent of intracellular Ca2�.

Inhibition of Erk Phosphorylation Does notPrevent ATP-Induced Barrier EnhancementThe mitogen-activated protein kinase (MAPK) cascade is asignal transduction system, which is known to participate inmultiple cellular functions.26 It has been previously shownthat extracellular ATP induces Erk MAPK phosphorylation inEC.27 In our system time-dependent Erk phosphorylationoccurred after ATP stimulation (Figure 4A). To investigate theinvolvement of Erk in the ATP-induced barrier response, theupstream kinase (MEK) was inhibited with U0126. Pretreatmentof HPAEC with U0126 completely abolished ATP-induced Erkphosphorylation (Figure 4B), but had no effect on ATP-inducedTER increase (Figure 4C). These data suggest the absence of afunctional relationship between Erk phosphorylation and in-creased EC barrier function after ATP stimulation.

ATP Response Is Mediated Via G ProteinsTo determine the role of G proteins in ATP-induced barrierenhancement, HPAEC were treated either with a G protein–specific silencing RNA or with pertussis toxin (PTX). Deple-tion of either G�i (Figure 5A) or G�q (Figure 5B) with specificsiRNAs significantly attenuated the increased TER inducedby ATP, which confirms the involvement of both G�i and G�q

subunits. Depletion of G�12 had an opposite effect as itpotentiated the response to ATP (Figure 5C), whereas deple-tion of G�13 had no effect (Figure 5D). Based on thesensitivity to PTX, G proteins are grouped into 2 families.The Gi/Go family is sensitive to PTX whereas the Gq family isinsensitive to this toxin. Pretreatment of HPAEC with PTXblocked the ATP response (Figure 5E), suggesting the exclu-sive role of Gi/Go proteins in ATP-induced barrier enhance-ment. The apparent discrepancy between PTX and siRNAdata regarding the contribution of G�i and G�q into ATPresponse may be caused by additional effects of PTX,different from ADP ribosylation of G�i/o.28–30

ATP Induces PKA ActivationIn the P2Y family of purine receptors only P2Y11 has beenreported to activate adenylate cyclase. However, adenylatecyclase activation and cAMP production are established stepsin signal transduction via adenosine receptors. ATP did notsignificantly raise the intracellular cAMP level in HPAEC asopposed to adenosine receptor agonist NECA (Figure 6A),suggesting that the ATP effect is not mediated by adenosinereceptors. A known activator of cAMP/PKA forskolin wasused as a positive control in these experiments. ATP chal-lenge did, however, lead to a transient increase in PKAactivity, that could be inhibited by PKA inhibitor H89 (Figure6B). H89 and another PKA inhibitor KT5720 also signifi-cantly attenuated the ATP-induced increase of TER (Figure6C). These data implied that activation of PKA via acAMP-independent mechanism was a necessary componentof ATP-induced barrier enhancement.

Vasodilator-stimulated phosphoprotein (VASP) is a knownPKA effector protein which has been recently shown tolocalize to cell–cell junctions and participate in EC cytoskel-etal rearrangement leading to permeability changes.31 ATPinduced PKA-specific phosphorylation of VASP on Ser157

(Figure 6D, left) simultaneously with PKA activation (5minutes). It should be noted that phosphorylation of VASPpersisted later, when PKA was no more activated. This mayoccur because dephosphorylation requires increased phospha-tase activity, which may be inhibited or unchanged after ATPtreatment. VASP phosphorylation was insensitive to Ca2�

chelation with BAPTA, which correlates with the Ca2�-insensitivity of ATP-induced increase in TER (Figure 6D,middle). ATP-induced VASP phosphorylation was com-pletely inhibited by H89 (Figure 6, right).

To directly examine the role of VASP in ATP-induced barrierenhancement, we used specific silencing RNA. VASP-depletedHPAEC exhibited an appreciably more robust response to ATP(Figure 6E). Taken together, these results indicated that VASPmay serve as a negative regulator of barrier function, and that itsphosphorylation after ATP exposure eliminated this property,resulting in enhanced EC barrier.

Figure 2. Effect of ATP on cell–cell junc-tions in HPAEC. Cells were treated with50 �mol/L ATP for 20 minutes and werestained for VE-cadherin or ZO-1 as indi-cated. A, Appreciably more VE-cadherinis recruited to the area of cell–cell junc-tions after ATP treatment. Arrows indi-cate overlapping edges of neighboringcells. B, Quantification of the surfacearea of the cell–cell interface. The per-centage of total cell surface area occu-

pied by VE-cadherin–labeled cell–cell junctions was calculated for 20 cells in eachgroup. The graph demonstrates that ATP induced a significant increase in cell–cellinterface surface area as a percentage of total cell surface area (*P�0.001 comparedwith control). The box and whiskers plot shows the means (lines at box centers,13.3% and 31.6% for control and ATP-treated cells, respectively), seventy-fifth per-centile (tops of boxes, 15.2% and 34.1%, for control and ATP-treated cells, respec-tively), twenty-fifth percentile (bottoms of boxes, 10.2% and 28.3%, for control andATP-treated cells, respectively), and standard deviations for each group. C, Cellswere stained for ZO-1. More regular and continuous cortical ZO-1 distribution isobserved after ATP treatment.

Kolosova et al ATP-Induced Endothelial Cell Barrier Enhancement 117



Myosin-Associated Phosphatase Is Involved inATP-Induced Barrier EnhancementIt has been previously reported that ATP-induced decreasesin transendothelial albumin flux correlates with decreased

phosphorylation of MLC.32 In our experimental system ATPtreatment had a biphasic effect with initial stimulation ofMLC phosphorylation (at 5 minutes) followed by inhibition(at 30 minutes) and a return to baseline values by 1 hour

Figure 3. ATP increases barrier function of HPAEC independently from intracellular Ca2�. A, ATP increases intracellular Ca2� in a dose-dependent manner. Left, time course of [Ca2�]i after stimulation with ATP. Right, ATP dose dependence of maximal [Ca2�]i increase. B,BAPTA inhibits ATP-induced increase in intracellular Ca2�. Left, Time course of [Ca2�]i after ATP stimulation in the presence ofBAPTA. Right, BAPTA dose dependence of maximal [Ca2�]i increase after stimulation with 50 �mol/L ATP. C, BAPTA (10 �mol/L)causes a decrease in basal TER, but does not affect increased TER induced by ATP (50 �mol/L). ATP-induced TER increase was cal-culated as a difference between TER values at the point of ATP addition and 30 minutes after ATP addition. D, BAPTA (10 �mol/L)does not affect ATP-induced enhancement of cortical VE-cadherin.




(Figure 7A). These data differ from those of Noll et al,18 whodemonstrated that ATP caused a fast and sustained MLCdephosphorylation in porcine aortic EC. Pretreatment of cellswith BAPTA prevented the early transient increase in MLCphosphorylation but did not affect decreased MLC phosphor-ylation at later time points (Figure 7A). These results sug-gested that early MLC phosphorylation is associated withintracellular Ca2� elevation but is not causally related tobarrier enhancement. It also suggested that the later Ca2�-independent reduction of MLC phosphorylation may befunctionally related to the observed barrier response. Datashown on Figure 7B confirms the decreased amount ofphosphorylated form of MLC after 30 minutes of ATPtreatment using urea gel electrophoresis.

In the next set of experiments we attempted to clarify therole of MLCP in the ATP-induced reduction of MLC phos-phorylation. First, the effect of ATP on TER was dramaticallysuppressed by microcystin, an inhibitor of phosphatase type 1(PP1) and type 2 (PP2A), but not by fostriecin, an inhibitor ofPP2A, implicating the involvement of PP1 (online Table II).Second, treatment of HPAEC with ATP led to increasedmyosin-associated phosphatase activity (Figure 8A) in atime-dependent manner, which agreed with the time course ofATP-induced barrier enhancement (Figure 8B). Increasedassociation of the PP1� isoform with the myosin fraction wasalso observed after ATP challenge (Figure 8C and 8D).Furthermore, the increase of myosin-associated phosphataseactivity was completely abolished by calyculin (inhibitor ofPP1 and PP2A), but not by fostriecin (online Table III).Importantly, phosphatase activation (Figure 8A), increase inTER (Figure 8B), PP1� association (Figure 8C and 8D), andMLC dephosphorylation (Figure 7A) reached their maximumat approximately the same time (30 minutes), suggesting astrong correlation between these processes. Taken together,these results indicate that MLCP plays an important role inbarrier enhancement induced by ATP. Furthermore, there is aclear association between ATP-induced activation of MLCPand G proteins. Depletion of G�q, but not G�i2 with siRNAabolished ATP-induced increase in phosphatase activity

(0.34�0.04 pmoles of phosphate per mg protein comparedwith 1.22�0.04 pmoles of phosphate per mg protein incontrol cells.)

DiscussionThe inflammatory response of lung endothelium includesincreased transendothelial permeability, leading to extravasa-tion of fluid and blood cells and resulting in lung edema.Inflammatory mediators acting via G protein–coupled recep-tors trigger increased endothelial permeability by increasingintracellular Ca2� concentrations which in turn activate sig-naling pathways leading to cytoskeletal reorganization anddisassembly of VE-cadherin at adheren junctions.1 Extracel-lular nucleotides activate ion-channel P2X receptors and Gprotein–coupled P2Y receptors inducing apoptotic, proinflam-matory, and thrombotic changes in many tissues and cell types.10

However, unlike other inflammatory stimuli, ATP and its ana-logues do not compromise endothelial barrier function.

In this study we show that extracellular ATP, despiteinducing increases in intracellular Ca2�, acts as a potentbarrier-protective agonist on EC derived from different typesof lung blood vessels. ATP induced an increase in TER(Figure 1A) and rearrangement of VE-cadherin and ZO-1,suggesting a tightening of cell–cell contacts (Figure 2).Because ATP undergoes hydrolysis within minutes on thesurface of EC, producing ADP and adenosine,33 it is possiblethat both purinergic systems (P1 and P2) are involved inATP-induced endothelial barrier enhancement. As nonhydro-lyzed ATP analogues also enhanced endothelial barrier func-tion (Figure 1B) and, as has been previously published, theadenosine receptor antagonist 8-phenyltheophylline does notinhibit the barrier-protective effects of ATP,18 it is evidentthat ATP directly triggers barrier-protective mechanisms viaP2 receptors.

Many effects of ATP as an extracellular mediator have beenattributed to an increase in intracellular Ca2� via P2Y receptors(see reference 10 for review). Intracellular Ca2�, however, is notimportant for the ATP-induced decrease in albumin flux acrossporcine endothelial monolayers.18 Our study confirms that the

Figure 4. Erk MAPK is not involved in ATP-induced barrier enhancement. ATP (50 �mol/L) induced time-dependent Erk phosphorylation (A).U0126 pretreatment (5 �mol/L, 10 minutes) completely blocked ATP-induced Erk phosphorylation (B), but did not affect TER (C).




barrier-protective property of ATP is unrelated to intracellularCa2� concentrations. Although ATP caused a dose-dependentrise of intracellular Ca2�, the chelation of this Ca2� with BAPTAdid not affect either increased barrier function or enhancedVE-cadherin staining at the cell periphery (Figure 3).

ATP has been shown to activate Erk in different cell types,including endothelium.34,35 However, Erk activation appearsto be involved in endothelial barrier dysfunction, rather thanprotection.36 We demonstrate that ATP-induced Erk activa-tion does not play a functional role in barrier enhancement(Figure 4).

ATP-induced EC barrier enhancement occurs via a Gprotein–coupled mechanism because treatment of EC withsiRNA designed to target G�q and G�i2 markedly decreasedATP effect (Figure 5A and 5B). PTX also prevented ATP

effects on TER (Figure 5E), suggesting the involvement ofG�i. Interestingly, depletion of G�12 protein potentiated effectsof ATP on TER (Figure 5C). Previously published datasuggest that G�12 may contribute to a procontractile phenotypeof EC. Specifically, G�12 activates the small GTPase Rho.37

Rho activation ultimately leads to MLC phosphorylation, cellcontraction, and barrier disruption.1 Activation of G�12 in-creased paracellular permeability38 and disrupted tight andadherens junctions39 in epithelial cells. The introduction ofmutationally activated G�12 protein into K562 cells blockedcadherin-mediated cell adhesion.40 Considering these data wespeculate that activation of G�12 by ATP may negativelycontribute to TER measurements in control cells, whereasG�12 depletion eliminated this negative effect (Figure 5C),thereby potentiating the effect of ATP.

Figure 5. ATP increase endothelial barrier via G protein–coupledmechanism. HPAEC were treated either with G protein isoform-specific siRNA as indicated or nonspecific RNA for 48 hours. Theexpression of G protein isoforms was detected by immunoblottingand �-tubulin was stained as a loading control. TER was mea-sured after siRNA treatment. Cells were incubated in serum-freemedium for 1 hour followed by challenge with 50 �mol/L ATP.Depletion of G�q (A) and G�i2 (B) significantly attenuated the effectof ATP on TER. Depletion of G�12 (C) and G�13 (D) slightly

enhanced cellular response to ATP. C, Pretreatment of HPAEC with pertussis toxin (100 ng/mL, 4 hours) inhibited ATP-inducedincrease of TER at 1 hour after ATP treatment (n�4; *P�0.01).




Elevation of cAMP levels and activation of PKA areknown to be associated with barrier enhancement.41 Previousdata, however, suggest that ATP-induced barrier enhance-ment is cAMP-independent.18 In our experiments ATP chal-lenge did not produce an elevation of cAMP (Figure 6A) butdid increase PKA activity (Figure 6B). PKA activationindependent of cAMP has recently been described. Thissignaling pathway uses activation of PKA via anchoringproteins such as AKAP11042 and NF-�B.43 Despite an estab-lished role for PKA in barrier protection, PKA targetsinvolved in endothelial barrier function remain largely un-

known. The focal adhesion- and microfilament-associatedprotein VASP is a known PKA target. Because VASP hasbeen implicated in many actin-based processes,44 its involve-ment in barrier regulation seems plausible. It has beendemonstrated that PKA-dependent phosphorylation of VASPacts as a negative regulator of actin dynamics45 and occurs oncell spreading.46 VASP is abundantly expressed in EC,however its role in endothelial physiology is only starting tobe explored. VASP might participate in maintaining an openparacellular pathway, acting as a negative regulator of barrierfunction, whereas phosphorylation on Ser157 may be associ-

Figure 6. ATP activates PKA without increasing intracellularcAMP. A, Unlike NECA (50 �mol/L, 10 minutes) or forskolin(50 �mol/L, 10 minutes), ATP does not significantly increaseintracellular cAMP level. B, ATP induces a transient increase inPKA activity, which is abrogated by 30 minutes preincubationwith 10 �mol/L H89 (*P�0.01, **P�0.001 compared with vehicle;***P�0.001 compared with ATP 5 minutes). Forskolin (50 �mol/L,10 minutes) was used as a positive control. C, PKA inhibitionwith 10 �mol/L H89 and 10 �mol/L KT5720 (30 minutes preincu-bation) attenuates ATP-induced TER increase (*P�0.01). D, ATPtreatment induces phosphorylation of VASP, which is not inhib-ited by pretreatment with BAPTA (10 �mol/L, 10 minutes), but iseliminated by pretreatment with H89 (10 �mol/L, 30 minutes). E,Depletion of endogenous VASP with siRNA resulted in anincreased response to ATP stimulation as measured by TER(50 �mol/L).




ated with relaxation of the actin cytoskeleton and increasedbarrier function.31 Our data support this idea. First, extracel-lular ATP triggered PKA-dependent VASP phosphoryla-tion, which occurred in parallel with an increase in TERand cell spreading (Figure 6D) suggesting that nonphos-phorylated VASP is a negative regulator of barrier func-tion, and its phosphorylation diminishes that negative

effect. Second, depletion of total VASP also eliminates thenegative effect and strengthens the barrier (Figure 6E),even if the amount of the phosphorylated form should bereduced accordingly. Therefore, both VASP phosphoryla-tion and depletion are associated with EC barrier enhance-ment, implicating a role for unphosphorylated VASP as anegative regulator.

Figure 7. ATP affects MLC phosphorylation inHPAEC. A, Immunoblotting using SDS-polyacryl-amide gel electrophoresis and a diphospho-specific MLC antibody shows a time course ofMLC phosphorylation after ATP treatment in theabsence (left) and in the presence (right) of BAPTA.BAPTA inhibits MLC phosphorylation at early timepoints (5 to 10 minutes) but does not affectdecreased MLC phosphorylation at 30 minutes. B,Immunoblotting using urea gel and MLC-specificantibody reveals a decrease of phosphorylatedforms of MLC after 30 minutes of ATP stimulation.Bars denote an average ratio�SE of phospho-MLCsignal to total MLC signal (n�3; *P�0.001).

Figure 8. ATP-induced barrier enhancement correlates with increased MLCP activity in HPAEC. A, Time course of TER after addition ofATP at zero time point (n�3). B, ATP stimulates phosphatase activity in the myosin-enriched fraction in time-dependent manner(*P�0.01). C, Immunoblotting of myosin-enriched fractions after treatment of HPAEC with ATP for indicated time periods. ATP treat-ment stimulates binding of PP1� to the myosin fraction. D, Densitometric analysis of PP1� signal intensity (n�3; *P�0.01).




Another molecular mechanism of ATP effects on endothe-lial barrier function defined in the current work is MLCdephosphorylation. EC contraction driven by MLC phosphor-ylation is a key event in several models of agonist-inducedbarrier dysfunction.1 It is unclear, however, if the opposite istrue: the role of MLC dephosphorylation in endothelialbarrier protection has not been confirmed. Noticeable dephos-phorylation of MLC occurs at 30 minutes after ATP chal-lenge and coincides with the peak of barrier enhancement(Figure 7A and 8A). However, the activity of MLCP startsincreasing shortly after ATP treatment and completely corre-lates with the time course of barrier enhancement (Figure 8Aand 8B). Early lack of MLC dephosphorylation may beexplained by intracellular Ca2� elevation resulting in theincreased activity of MLC kinase, which in turn, leads to anincreased level of phosphorylated MLC. But it neither over-comes the effect of ATP nor is it causally related toATP-induced barrier enhancement.

Because the catalytic subunit of MLCP was identified as aPP1� isoform,47 we specifically studied the association ofPP1� with the myosin fraction (Figure 8C) and found it to beincreased (Figure 8C and 8D). This confirms the involvementof MLCP in ATP-induced enhancement of endothelial bar-rier. Furthermore, we found that ATP causes activation ofMLCP via a G protein–coupled mechanism. Interestingly,although both G�q and G�i2 are involved in barrier-enhancingATP signaling (Figure 5A and 5B), only G�q appears to beimportant for phosphatase activation. MLCP is regulated viaits regulatory subunit myosin-binding phosphatase targeting.Phosphorylation of myosin-binding phosphatase targeting byRho kinase leads to inhibition of its activity.48 Our dataprovide novel evidence of a positive regulation for MLCP viaa G protein–coupled mechanism. To our knowledge, apositive regulatory mechanism for MLCP has only beenshown in a study of cell division.49

Our study presents an attempt to clarify some of thesignaling pathways involved in ATP-induced endothelial cellbarrier enhancement. Lack of experimental data leaves roomfor many speculations regarding the similarity of ATP sig-naling to other known barrier-protective mechanisms. Forinstance, action of potent barrier protector sphingosine1-phosphate has been long investigated and involves activa-tion of small GTPase Rac followed by strong enhancement ofcortical actin.50 We have not studied these particular mecha-nisms in our work. Further studies are needed to fullycharacterize complex signaling machinery involved in ATP-induced enhancement of endothelial barrier. Based on ourdata, however, several signaling elements can be distin-guished. These include a G protein–coupled receptor (mostlikely P2Y type), G�q and G�i2, PKA, and MLCP. PKAactivation, however, occurs via a cAMP-independent mech-anism, possibly involving protein kinase A–anchoring pro-teins (AKAPs). Ca2� signaling and Erk activation are notinvolved in the effect of ATP on endothelial barrier.

Beneficial effects of ATP on barrier function suggest thatendothelium is a potential therapeutic target for purine-basedagonists. Further studies, using isolated lung and animalmodels, should clarify the possible use of such agonists in thetreatment of acute lung injury.

AcknowledgmentsThis work was supported by grants from National Heart, Lung, andBlood Institutes (HL67307, HL68062, and HL58064), and AmericanLung Association of Maryland Research Grant.

References1. Dudek SM, Garcia JG. Cytoskeletal regulation of pulmonary vascular

permeability. J Appl Physiol. 2001;91:1487–1500.2. Burnstock G. Potential therapeutic targets in the rapidly expanding field

of purinergic signalling. Clin Med. 2002;2:45–53.3. Beigi R, Kobatake E, Aizawa M, Dubyak GR. Detection of local ATP

release from activated platelets using cell surface-attached fireflyluciferase. Am J Physiol. 1999;276:C267–C278.

4. Bergfeld GR, Forrester T. Release of ATP from human erythrocytes inresponse to a brief period of hypoxia and hypercapnia. Cardiovasc Res.1992;26:40–47.

5. Coade S, Pearson J. Metabolism of adenine nucleotides in human blood.Circ Res. 1989;65:531–537.

6. Schwiebert LM, Rice WC, Kudlow BA, Taylor AL, Schwiebert EM.Extracellular ATP signaling and P2X nucleotide receptors in monolayersof primary human vascular endothelial cells. Am J Physiol. 2002;282:C289–C301.

7. Pearson JD, Gordon JL. Vascular endothelial and smooth muscle cells inculture selectively release adenine nucleotides. Nature 1979;281:384–386.

8. Bodin P, Burnstock G. ATP-stimulated release of ATP by human endo-thelial cells. J Cardiovasc Pharmacol. 1996;27:872–875.

9. Bodin P, Burnstock G. Increased release of ATP from endothelial cellsduring acute inflammation. Inflamm Res. 1998;47:351–354.

10. Ralevic V, Burnstock G. Receptors for purines and pyrimidines.Pharmacol Rev. 1998;50:413–492.

11. Fredholm BB, Ijzerman AP, Jacobson KA, Klotz KN, Linden J. Interna-tional Union of Pharmacology 2001: XXV. Nomenclature and classifi-cation of adenosine receptors. Pharmacol Rev. 2001;53:527–552.

12. North RA. Molecular physiology of P2X receptors. Physiol Rev. 2002;82:1013–1067.

13. Ray FR, Huang W, Slater M, Barden JA. Purinergic receptor distributionin endothelial cells in blood vessels: a basis for selection of coronaryartery grafts. Atherosclerosis. 2002;162:55–61.

14. 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.

15. Moore DJ, Chambers JK, Wahlin JP, Tan KB, Moore GB, Jenkins O,Emson PC, Murdock PR. Expression pattern of human P2Y receptorsubtypes: a quantitative reverse transcription-polymerase chain reactionstudy. Biochim Biophys Acta. 2001;1521:107–119.

16. Simon J, Filippov AK, Goransson S, Wong YH, Frelin C, Michel AD,Brown DA, Barnard EA. Characterization and channel coupling of theP2Y(12) nucleotide receptor of brain capillary endothelial cells. J BiolChem. 2002;277:31390–313400.

17. Olanrewaju HA, Qin W, Feoktistov I, Scemama JL, Mustafa SJ. Aden-osine A(2A) and A(2B) receptors in cultured human and porcine coronaryartery endothelial cells. Am J Physiol. 2000;279:H650–H656.

18. Noll T, Holschermann H, Koprek K, Gunduz D, Haberbosch W,Tillmanns H, Piper HM. ATP reduces macromolecule permeability ofendothelial monolayers despite increasing [Ca2�]i. Am J Physiol. 1999;276:H1892–H1901.

19. Gunduz D, Hirche F, Hartel FV, Rodewald CW, Schafer M, Pfitzer G,Piper HM, Noll T. ATP antagonism of thrombin-induced endothelialbarrier permeability. Cardiovasc Res. 2003;59:470–478.

20. Birukova AA, Birukov KG, Smurova K, Adyshev D, Kaibuchi K, AlievaI, Garcia JG, Verin AD. Novel role of microtubules in thrombin-inducedendothelial barrier dysfunction. FASEB J. 2004;18:1879–1890.

21. Tiruppathi C, Malik AB, Del Vecchio PJ, Keese CR, Giaever I. Electricalmethod for detection of endothelial cell shape change in real time:assessment of endothelial barrier function. Proc Natl Acad Sci U S A.1992;89:7919–7923.

22. Borbiev T, Verin AD, Birukova A, Liu F, Crow MT, Garcia JG. Role ofCaM kinase II and ERK activation in thrombin-induced endothelial cellbarrier dysfunction. Am J Physiol. 2003;285:L43–L54.

23. Usatyuk PV, Fomin VP, Shi S, Garcia JG, Schaphorst K, Natarajan V.Role of Ca2� in diperoxovanadate-induced cytoskeletal remodeling andendothelial cell barrier function. Am J Physiol. 2003;285:L1006–L1017.




24. Verin AD, Patterson CE, Day MA, Garcia JG. Regulation of endothelialcell gap formation and barrier function by myosin-associated phosphataseactivities. Am J Physiol. 1995;269:L99–L108.

25. Tiruppathi C, Minshall RD, Paria BC, Vogel SM, Malik AB. Role ofCa2� signaling in the regulation of endothelial permeability. VasculPharmacol. 2002;39:173–185.

26. Lewis TS, Shapiro PS, Ahn NG. Signal transduction through MAP kinasecascades. Adv Cancer Res. 1998;74:49–139.

27. Klingenberg D, Gunduz D, Hartel F, Bindewald K, Schafer M, Piper HM,Noll T. MEK/MAPK as a signaling element in ATP control of endothelialmyosin light chain. Am J Physiol. 2004;286:C807–C812.

28. Thom RE, Casnellie JE. Pertussis toxin activates protein kinase C and atyrosine protein kinase in the human T cell line Jurkat. FEBS Lett.1989;244:181–184.

29. Kindt RM, Lander AD. Pertussis toxin specifically inhibits growth coneguidance by a mechanisms independent of direct G protein inactivation.Neuron. 1995;15:79–88.

30. Garcia JG, Wang P, Liu F, Hershenson MB, Borbiev T, Verin AD.Pertussis toxin directly activates endothelial cell p42/p44 MAP kinasesvia a novel signaling pathway. Am J Physiol. 2001;280:C1233–1241.

31. Comerford KM, Lawrence DW, Synnestvedt K, Levi BP, Colgan SP.Role of vasodilator-stimulated phosphoprotein in PKA-induced changesin endothelial junctional permeability. FASEB J. 2002;16:583–585.

32. Noll T, Schafer M, Schavier-Schmitz U, Piper HM. ATP induces dephos-phorylation of myosin light chain in endothelial cells. Am J Physiol.2000;279:C717–C723.

33. Gordon EL, Pearson JD, Slakey LL. The hydrolysis of extracellularadenine nucleotides by cultured endothelial cells from pig aorta. Feed-forward inhibition of adenosine production at the cell surface. J BiolChem. 1986;261(33):15496–15507.

34. Graham A, McLees A, Kennedy C, Gould GW, Plevin R. Stimulation bythe nucleotides, ATP and UTP of mitogen-activated protein kinase inEAhy 926 endothelial cells. Br J Pharmacol. 1996;117:1341–1347.

35. Patel V, Brown C, Goodwin A, Wilkie N, Boarder MR. Phosphorylationand activation of p42 and p44 mitogen-activated protein kinase arerequired for the P2 purinoceptor stimulation of endothelial prostacyclinproduction. Biochem J. 1996;320:221–226.

36. Bogatcheva NV, Dudek SM, Garcia JG, Verin AD. Mitogen-activatedprotein kinases in endothelial pathophysiology. J Investig Med. 2003;51:341–352.

37. Kurose H. Galpha12 and Galpha13 as key regulatory mediator in signaltransduction. Life Sci. 2003;74:155–161.

38. Meyer TN, Schwesinger C, Denker BM. Zonula occludens-1 is a scaf-folding protein for signaling molecules. Galpha(12) directly binds to theSrc homology 3 domain and regulates paracellular permeability in epi-thelial cells. J Biol Chem. 2002;277:24855–24858.

39. Meyer TN, Hunt J, Schwesinger C, Denker BM. Galpha12 regulatesepithelial cell junctions through Src tyrosine kinases. Am J Physiol.2003;285:C1281–C1293.

40. Meigs TE, Fedor-Chaiken M, Kaplan DD, Brackenbury R, Casey PJ.Galpha12 and Galpha13 negatively regulate the adhesive functions ofcadherin. J Biol Chem. 2002;277:24594–24600.

41. van Nieuw Amerongen GP, van Hinsbergh VW. Targets for pharmaco-logical intervention of endothelial hyperpermeability and barrier function.Vascul Pharmacol. 2002;39:257–272.

42. Niu J, Vaiskunaite R, Suzuki N, Kozasa T, Carr DW, Dulin N, Voyno-Yasenetskaya TA. Interaction of heterotrimeric G13 protein with anA-kinase-anchoring protein 110 (AKAP110) mediates cAMP-independent PKA activation. Curr Biol. 2001;11:1686–1690.

43. Zieger M, Tausch S, Henklein P, Nowak G, Kaufmann R. A novelPAR-1-type thrombin receptor signaling pathway: cyclic AMP-independent activation of PKA in SNB-19 glioblastoma cells. BiochemBiophys Res Commun. 2001;282:952–957.

44. Krause M, Dent EW, Bear JE, Loureiro JJ, Gertler FB. Ena/VASPproteins: regulators of the actin cytoskeleton and cell migration. Annu RevCell Dev Biol. 2003;19:541–564.

45. Harbeck B, Huttelmaier S, Schluter K, Jockusch BM, Illenberger S.Phosphorylation of the vasodilator-stimulated phosphoprotein regulatesits interaction with actin. J Biol Chem. 2000;275:30817–30825.

46. Lawrence DW, Pryzwansky KB. The vasodilator-stimulated phospho-protein is regulated by cyclic GMP-dependent protein kinase duringneutrophil spreading. J Immunol. 2001;166:5550–5556.

47. Hartshorne DJ, Ito M, Erdodi F. Myosin light chain phosphatase: subunitcomposition, interactions and regulation. J Muscle Res Cell Motil. 1998;19:325–341.

48. Somlyo AP, Somlyo AV. Signal transduction by G-proteins, rho-kinaseand protein phosphatase to smooth muscle and non-muscle myosin II.J Physiol. 2000;522:177–185.

49. Totsukawa G, Yamakita Y, Yamashiro S, Hosoya H, Hartshorne DJ,Matsumura F. Activation of myosin phosphatase targeting subunit bymitosis-specific phosphorylation. J Cell Biol. 1999;144:735–744.

50. Garcia JG, Liu F, Verin AD, Birukova A, Dechert MA, Gerthoffer WT,Bamberg JR, English D. Sphingosine 1-phosphate promotes endothelialcell barrier integrity by Edg-dependent cytoskeletal rearrangement. J ClinInvest. 2001;108(5):689–701.




Kolosova et al.

MATERIALS AND METHODS

Materials. H89, U0126, pertussis toxin and VASP polyclonal antibody were purchased

from Calbiochem (San Diego, CA). ATP, ADP, ATPγS, MeS-ATP, AMP-CCP, total

myosin light chain (MLC) antibody and forskolin were purchased from Sigma.

Monoclonal VE-Cadherin and ZO-1 antibodies were purchased from BD Biosciences

(San Diego, CA). BAPTA, Texas Red-conjugated phalloidin, and goat anti-mouse Alexa

488 antibodies were purchased from Molecular Probes (Eugene, OR). Phospho-specific

(Thr18/Ser19) MLC antibody, phospho-p44/42 and total p44/42 MAPK antibodies were

purchased from Cell Signaling (Beverly, MA).

Cell Cultures. Human pulmonary artery EC (HPAEC), and human lung microvascular

EC (HLMVEC) were purchased from Clonetics (Walkersville, MD) and were cultured in

EBM-2 complete medium (Clonetics). HPAEC were utilized at passages 5–10 and

HLMVEC at passages 4-8. Bovine pulmonary artery endothelial cells (BPAEC) were

utlizied at 16th passage from American Type Tissue Culture Collection (Rockville, MD),

cultured in medium consisting of DMEM (Gibco BRL, Grand Island, NY) supplemented

with 20% fetal calf serum.

Depletion of endogenous proteins using small interfering RNA. To reduce the content

of endogenous Gai2, Gaiq, or VASP HPAEC were treated with specific small interfering

RNA (siRNA) duplexes, which guide sequence-specific degradation of the homologous

mRNA. The following duplexes of sense and antisense siRNA from Ambion (Austin,

TX) were used:

Target Sense Antisense

Gαi2 5’GGUGAAGUUGCUGCUGUUGtt3’ 5’CAACAGCAGCAACUUCACCtc3’

Kolosova et al.

Gαiq 5’GGAGAGAGUGGCAAGAGUAtt3’ 5’UACUCUUGCCACUCUCUCCtg3’

Gαi12 5’GGGCUCAAGGGUUCUUGUUtt3’ 5’AACAAGAACCCUUGAGCCCtt3’

VASP 5’GGAAAUAAGAUGCUGUAACtt3’ 5’GUUACAGCAUCUUAUUUCCtc3’

A non-specific RNA duplex Fl-Luciferase GL2 (Dharmacon Research, Lafayette,

CO) was used as a control treatment. HLMVEC were plated on 60 mm dishes to yield

70% confluence, and transfection of siRNA was performed using GeneSilencerTM

transfection reagent (Gene Therapy Systems, San Diego, CA) to achieve a final RNA

concentration of 100 nM. Forty-eight hours later, cells were lysed in SDS sample buffer

and specific protein depletion was analyzed by immunoblotting.

For transendothelial resistance measurement cells were plated in electrode wells

and transfected with siRNA as described.

Endothelial monolayer resistance measurement. The barrier properties of EC

monolayers were characterized using highly sensitive electrical cell impedance sensor

system (Applied Biophysics, Troy, NY) as previously described [21]. Transendothelial

resistance (TER) data were normalized to the initial voltage.

Immunofluorescent microscopy. Immunostaining protocol described previously [22] was

used in the study. Specific antibodies were used for detection of VE-cadherin 5 and ZO-

1. The percentage of total cell surface area occupied by VE-cadherin labeled cell-cell

junctions was quantitatively determined using a surface area measurement tool in

Openlab (4.0) software (Improvision, Lexington, MA).

Measurement of [Ca2+]i. Concentrations of cytosolic free Ca2+ were measured as

described previously [23].

Kolosova et al.

Measurement of intracellular cAMP level. HPAEC were grown to a concentration of

2x106 cells per well and stimulated with ATP or forskolin. Cells were lysed and total

cellular cAMP concentration in samples was determined with TRK 432 radioassay

system (Amersham, UK) according to the manufacturer’s instructions. Samples were

normalized to protein concentrations which were determined with BCA Protein Assay

Reagent Kit (Pierce, RockRockford, IL).

PKA Activity Assay. PKA activity was measured using a nonradioactive PKA Kinase

Activity assay Kit (Stressgen, Victoria, Canada) according to the manufacturer's

instructions. Assays were performed in triplicate. Samples were normalized to protein

concentrations as described above.

Myosin-associated Posphatase Activity Assay. Myosin-enriched fraction of HPAEC was

prepared as described previously [24]. Ser/Thr Phosphatase Assay Kit (Upstate, Lake

Placid, NY) was used to determine phosphatase activity in myosin fraction.

MLC phosphorylation profile. MLC phosphorylation was analyzed either by phospho-

specific antibody or urea PAGE as previously described [24]

Statistical Analysis. For basic statistical analysis a GraphPad Prism Program was used.

Data were compared by a Student’s t-test. P-values of less than 0.05 were considered to

be significant. Values are expressed as mean +SE. Experiments were repeated several

times as indicated on figure legends.

Online Table 1. Effect of ATP on TER of different endothelial cell lines.

Cell line ATP-induced increase of TER,

% of control

P

HPAEC 42.0+2.8 <0.001

HLMVEC 50.3+6.0 <0.001

BPAEC 20.0+2.5 <0.05

Shown are mean values + SE (n=3) of TER at 30 min after stimulation with 50 µM ATP. Online Table 2. ATP-induced TER of HPAEC is inhibited by 100 µM microcystin (an inhibitor of PP1 and PP2A) but not 1 µM foctriecin (an inhibitor of PP2A).

Treatment Normalized Resistance

Control 0.896+0.013

Microcystin 0.892+0.011

Fostriecin 0.964+0.025

ATP 1.388+0.075*

Microcystin+ATP 1.096+0.024**

Fostriecin+ATP 1.436+0.080 *

Shown are mean values+SE (n=3) of TER at 30 min after the addition of 50 µM ATP. *P<0.01 compared to control; **P<0.01 compared to ATP. Online Table 3. Calyculin (10 nM), but not fostriecin (1 µM) inhibited PPase activity in myosin-enriched fraction of HPAEC.

Treatment PPase activity, % of control

Control 100

Calyculin 30+9

Fostriecin 126+48

ATP 195+17 *

Calyculin+ATP 35+13 *, **

Fostriecin+ATP 160+48 *

Shown are mean values+SE (n=3) at 30 min after the addition of 50 µM ATP. *P< 0.01 compared to control; **P<0.01 compared to ATP.

D. VerinJeffrey R. Jacobson, Viswanathan Natarajan, David B. Pearse, Joe G.N. Garcia and Alexander

Irina A. Kolosova, Tamara Mirzapoiazova, Djanybek Adyshev, Peter Usatyuk, Lewis H. Romer,Enhancement

Signaling Pathways Involved in Adenosine Triphosphate-Induced Endothelial Cell Barrier

Print ISSN: 0009-7330. Online ISSN: 1524-4571 Copyright © 2005 American Heart Association, Inc. All rights reserved.is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation Research

doi: 10.1161/01.RES.0000175561.55761.692005;97:115-124; originally published online June 30, 2005;Circ Res.

http://circres.ahajournals.org/content/97/2/115World Wide Web at:

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

http://circres.ahajournals.org/content/suppl/2005/06/30/01.RES.0000175561.55761.69.DC1.htmlData Supplement (unedited) at:

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

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

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

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

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

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


http://circres.ahajournals.org/content/97/2/115

http://circres.ahajournals.org/content/suppl/2005/06/30/01.RES.0000175561.55761.69.DC1.html

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

http://www.lww.com/reprints

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


Signaling Pathways Involved in Adenosine Triphosphate-Induced Endothelial Cell Barrier Enhancement

Documents