CD39/Ectonucleoside Triphosphate Diphosphohydrolase 1 Provides Myocardial Protection During Cardiac Ischemia/Reperfusion Injury

CD39/Ectonucleoside Triphosphate Diphosphohydrolase 1Provides Myocardial Protection During Cardiac

Ischemia/Reperfusion InjuryDavid Köhler, PhD*; Tobias Eckle, MD*; Marion Faigle, BS; Almut Grenz, MD;

Michel Mittelbronn, MD; Stefanie Laucher, BS; Melanie L. Hart, PhD; Simon C. Robson, MD;Christa E. Müller, PhD; Holger K. Eltzschig, MD, PhD

Background—Extracellular adenosine, generated from extracellular nucleotides via ectonucleotidases, binds to specificreceptors and provides cardioprotection from ischemia and reperfusion. In the present study, we studied ecto-enzymaticATP/ADP-phosphohydrolysis by select members of the ectonucleoside triphosphate diphosphohydrolase (E-NTPDase)family during myocardial ischemia.

Methods and Results—As a first step, we used a murine model of myocardial ischemia and in situ preconditioning andperformed pharmacological studies with polyoxometalate 1, a potent E-NTPDase inhibitor (Na6[H2W12O40]). Polyoxo-metalate 1 treatment increased infarct sizes and abolished beneficial effects of preconditioning. To define relativecontributions of distinct E-NTPDases, we investigated transcriptional responses of E-NTPDases 1 to 3 and 8 topreconditioning. We noted robust and selective induction of E-NTPDase 1 (CD39) transcript and protein. Histologicalanalysis of preconditioned myocardium localized CD39 induction to endothelia and myocytes. Cd39�/� mice exhibitedlarger infarct sizes with ischemia (cd39�/� 43.0�3.3% versus cd39�/� 52%�1.8; P�0.05), and cardioprotection wasabrogated by preconditioning (cd39�/� 13.3%�1.5 versus cd39�/� 50.5%�2.8; P�0.01). Heightened levels of injuryafter myocardial ischemia and negligible preconditioning benefits in cd39�/� mice were corrected by infusion of themetabolic product (AMP) or apyrase. Moreover, apyrase treatment of wild-type mice resulted in 43�4.2% infarct sizereduction (P�0.01).

Conclusions—Taken together, these studies reveal E-NTPDase 1 in cardioprotection and suggest apyrase in the treatmentof myocardial ischemia. (Circulation. 2007;116:1784-1794.)

Key Words: adenosine � endothelium � enzymes � myocardial infarction � reperfusion

Several studies have found a pivotal role of extracellularadenosine signaling in tissue protection, particularly dur-

ing conditions of limited oxygen availability.1–4 Duringhypoxia, extracellular adenosine stems mainly from increasedphosphohydrolysis of precursor nucleotides (ATP/ADP/AMP) and contributes to cardioprotection from ischemia andreperfusion injury. Ischemic preconditioning (IP)5 and post-conditioning events6 involve protective cellular adaptive phe-nomena in the heart associated with protein kinase C activa-tion, a component of the mitochondrial K(ATP) signalingcascade that combats ischemic stress. In such pathophysio-

logical settings, the myocardial cellular phenotype alters tobecome more resistant to subsequent ischemia and tissueinjury. This polygenic response involves mediators includingadenosine,7 bradykinin,8 opioids,9 erythropoietin,10 adrener-gics, and muscarinics.11

Clinical Perspective p 1794

Our group has been interested in the increases in extracel-lular adenosine that are critical in cardioprotection duringIP.12 For example, studies measuring interstitial adenosineconcentrations in swine or perfused rabbit hearts via micro-

Received January 12, 2007; accepted August 17, 2007.From the Department of Anesthesiology and Intensive Care Medicine (D.K., T.E., M.F., S.L., H.K.E.), Department of Pharmacology and Toxicology

(A.G.), and Institute of Brain Research (M.M.), University Hospital, Tübingen, Germany; Center for Experimental Therapeutics and Reperfusion Injury,Department of Anesthesiology, Perioperative and Pain Medicine, Brigham and Women’s Hospital (M.L.H.), and Liver and Transplant Centers (S.C.R.),Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, Mass; Pharmaceutical Institute, PharmaceuticalSciences Bonn, University of Bonn, Bonn, Germany (C.E.M.); and Mucosal Inflammation Program, Department of Anesthesiology, University ofColorado Health Science Center, Denver (H.K.E.).

*The first 2 authors contributed equally to this work.Guest Editor for this article was Thomas F. Luscher, MD.The online-only Data Supplement, consisting of Methods and tables, is available with this article at http://circ.ahajournals.org/cgi/

content/full/CIRCULATIONAHA.107.690180/DC1.Correspondence to Holger K. Eltzschig, MD, PhD, Mucosal Inflammation Program, Department of Anesthesiology and Perioperative Medicine,

University of Colorado Health Sciences Center, 4200 E Ninth Ave, Campus Box B113, Denver, CO 80262. E-mail [email protected]© 2007 American Heart Association, Inc.

Circulation is available at http://circ.ahajournals.org DOI: 10.1161/CIRCULATIONAHA.107.690180

1784

Molecular Cardiology

by guest on June 1, 2016http://circ.ahajournals.org/Downloaded from by guest on June 1, 2016http://circ.ahajournals.org/Downloaded from by guest on June 1, 2016http://circ.ahajournals.org/Downloaded from

http://circ.ahajournals.org/



dialysis noted a 6- or 12-fold increase in extracellularadenosine with IP, respectively.7,13 Other studies in mice genetargeted for individual adenosine receptors also provideconvincing evidence for adenosine signaling in cardioprotec-tion by IP.12 Increases in extracellular adenosine predomi-nantly reflect enhanced extracellular adenosine generationfrom nucleotides. For example, studies of renal IP founddramatic increases of adenosine levels with IP treatment inthe kidneys that were blunted in gene-targeted mice for cd73(conversion of AMP to adenosine)14 or cd39 (conversion ofAMP to adenosine).15 Similarly, cardiac studies have showna critical role of CD73 in the elevation of cardiac adenosinelevels during IP.12,16,17 On the basis of these studies and thefact that extracellular levels of ATP and ADP are increaseddramatically during preconditioning,18 we presumed a contri-bution of extracellular ATP/ADP phosphohydrolysis in car-dioprotection from myocardial ischemia and reperfusioninjury.

Extracellular ATP/ADP-phosphohydrolysis is mainlyachieved enzymatically by ecto-nucleoside triphosphatediphosphohydrolases (E-NTPDases), a recently describedfamily of ubiquitously expressed membrane-bound en-zymes.19,20 The catalytic sites of plasma membrane–ex-pressed E-NTPDases 1 to 3 and 8 are exposed to theextracellular milieu, and the others are intracellular.20 Thepresumptive biological role of plasma membrane–boundE-NTPDases (E-NTPDases 1 to 3 and 8) is to fine-tuneextracellular nucleotide levels. For example, E-NTPDase 1(CD39) plays an important role in vascular endothelialfunction by blocking platelet aggregation via the phosphohy-drolysis of ATP and ADP from the blood to maintain vascularintegrity.21,22 At the same time, E-NTPDase 1 is also impor-tant in the maintenance of platelet functionality by preventingplatelet P2Y1-receptor desensitization. As such, mice genetargeted for E-NTPDase 1 (cd39�/� mice) show prolongedbleeding time with minimally perturbed coagulation param-eters.23 Of significant physiological relevance to the presentstudy is the fact that the E-NTPDase end product, AMP,serves as the major metabolic substrate for CD73-dependentgeneration of extracellular adenosine.1 Thus, E-NTPDaseexpression and function are key regulators of extracellularadenosine signaling. Therefore, we addressed the role ofATP/ADP-nucleotide phosphohydrolysis in cardioprotectionby applying a murine model of myocardial ischemia and insitu IP. This model uses a hanging-weight system for coro-nary artery occlusion, thus eliminating the necessity ofintermittently occluding the coronary artery with a knottedsuture.24

MethodsMiceAll animal protocols were in accordance with the German guidelinesfor use of living animals and were approved by the InstitutionalAnimal Care and Use Committee of the University Hospital Tübin-gen and the Regierungspräsidium Tübingen. C57BL/6J and C57BL/6x129Svj mice were obtained from Charles River (Sulzfeld, Ger-many). Mice deficient in cd39 on the C57BL/6x129Svj strain weregenerated, validated, and characterized as described previously.23

Murine Model of Myocardial Ischemia and IPC57BL/6x129Svj strain, cd39�/� mice, or littermate controls werematched in age, gender, and weight. Cardiac IP was performed withthe use of a hanging-weight system as described previously (seeMethods in the online-only Data Supplement).24

Transcriptional AnalysisTo assess the influence of IP on cd39 transcript level, IP wasperformed, and the area at risk was delineated by Evan’s bluestaining and excised at indicated time periods, followed by isolationof RNA and quantification of transcript levels by real-time reversetranscription polymerase chain reaction (iCycler; Bio-Rad Labora-tories, Munich, Germany), as described previously.12

Western Blots for CD39C57BL/6x129Svj mice were euthanized, cardiac IP was performed,and the area at risk was excised at 30, 60, 90, and 120 minutes afterIP and immediately frozen at �80°C (remaining blood was removedbefore). In subsets of experiments, we determined CD39 proteincontent from the area at risk, as described previously.12

ImmunohistochemistryIn subsets of experiments, we determined CD39 protein content fromthe area at risk, as described previously.12

Adenosine MeasurementsTissue adenosine and AMP levels were determined via high-performance liquid chromatography, as described previously.12

Measurement of CD39 Enzyme ActivityTo measure cardiac CD39 activity, we adopted a previously de-scribed technique (for details, see the online-only DataSupplement).1

Data AnalysisFor comparison of 2 groups, the nonparametric Mann-Whitney testwas performed. For comparison of �2 groups, the Kruskal-Wallistest with a Dunn posttest was performed. Statistical significance wasaccepted at a level of P�0.05. All values are expressed asmean�SEM from 6 animals per condition.

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.

ResultsPharmacological Inhibition of E-NTPDases Resultsin Increased Myocardial Infarct Size andAbolished Cardioprotection by IPOn the basis of the hypothesis that ATP/ADP phosphohy-drolysis attenuates myocardial ischemia/reperfusion injuryand contributes to cardioprotection by IP, we first sought toinhibit extracellular phosphohydrolysis pharmacologically.However, previously proposed E-NTPDase inhibitors are alsostrong inhibitors of ATP receptors, thus making it impossibleto selectively study E-NTPDase–dependent phosphohydroly-sis in vivo.25 In contrast, we recently identified polyoxometa-lates as a novel class of E-NTPDase inhibitors withoutactivity on purinergic receptors.26 A screen of differentpolyoxometalates revealed polyoxotungstate (Na6[H2W12O40],POM-1; Figure 1A) as a highly potent E-NTPDase 1 andE-NTPDase 3 inhibitor with Ki values of 2.58 and 3.26�mol/L, respectively, and an �10-fold lower inhibitoryactivity for E-NTPDase 2 (Ki�28.8 �mol/L) (Figure 1B). Todemonstrate a biological effect of polyoxometalate 1

Köhler et al CD39 in Myocardial Ischemia/Reperfusion Injury 1785

by guest on June 1, 2016http://circ.ahajournals.org/Downloaded from


(POM-1) on ATP/ADP phosphohydrolysis in vivo, we firstmeasured the effect of intravascular ATP treatment on heartrate changes. Application of an intravascular ATP bolus (50�L, 8 mg/mL) results in a heart rate reduction from 480 to120 bpm, lasting for 150 seconds. To confirm that this effectis mediated by adenosine, we subjected previously describedmice gene targeted for the adenosine A1 receptor (A1AR�/�

mice)27 to the same regimen. In contrast to wild-type mice, nochange in heart rate was observed in A1AR�/� mice (data notshown), suggesting that heart rate changes elicited by ATPinfusion are mediated by adenosine signaling. After intra-ar-terial POM-1 infusion (3 mg/kg), the observed reduction inheart rate was similar in degree but recovered after asignificantly shorter time period (50 seconds; P�0.05), sug-gesting that POM-1 treatment results in decreased extracel-lular adenosine generation from ATP. POM-1 treatment alonedid not alter heart rate. Measurements of cardiac adenosineand AMP levels in POM-1–treated mice revealed attenuated

increases in cardiac adenosine and AMP with preconditioning(data not shown).

After having shown attenuation of nucleotide-phosphohydrolysis with POM-1 treatment, we used thispharmacological regimen in a recently described model of insitu myocardial ischemia and preconditioning.24 We thussubjected mice (C57BL/6x129Svj) to 60 minutes of leftcoronary artery occlusion followed by 120 minutes of reper-fusion with or without prior IP treatment (4 cycles, 5 minutesof ischemia, 5 minutes of reperfusion). All mice survived thisexperiment. Heart rate and blood pressure did not differbetween POM-1–treated and –untreated mice (data notshown). As shown in Figure 1D, POM-1 treatment resulted insignificantly larger infarct sizes and complete inhibition ofcardioprotection by IP. These data provide novel pharmaco-logical evidence for a critical role of E-NTPDase–dependentnucleotide phosphohydrolysis in attenuating myocardial is-chemia and cardioprotection by IP.

POM-1Na6[H2W12O40]

-IP +IP -IP +IP0

10

20

30

40

50

60

Infa

rct S

ize

[% A

rea

at R

isk]

+ POM-1- POM-1 p<0.01

p<0.05

p<0.05

A

0

50

100

150

ATP

Indu

ced

Tim

e of

Bra

dyca

rdia

[s]

B

DC

p<0.05

10-7 10-6 10-5 10-4 10-3 10-20

20

40

60

80

100

120E-NTPDase3E-NTPDase2

E-NTPDase1

[POM-1], M

% N

TPD

ase

Act

ivity

Figure 1. E-NTPDase inhibition is associ-ated with increased myocardial infarctsizes and abolished cardioprotection by IP.A, Chemical structure of the E-NTPDaseinhibitor POM-1. B, Concentration-dependent inhibition of rat E-NTPDases 1to 3 by POM-1 (representative curves of 3separate experiments determined in dupli-cate for each enzyme). The following Ki

values were calculated from the deter-mined IC50 values: E-NTPDase 1, 2.58�mol/L; E-NTPDase 2, 28.8 �mol/L;E-NTPDase 3, 3.26 �mol/L. C, ATP-induced time of bradycardia is attenuatedafter POM-1 treatment. C57BL/6x129Svjmice received an intraarterial infusion afterPOM-1 (3 mg/kg per hour) or vehicle con-trol 30 minutes before application of anintravascular ATP bolus (50 �L, 8 mg/mL).Displayed is the time of bradycardia(reduction from 480 to 120 bpm). D,Increased infarct size and abolished car-dioprotection by IP with POM-1 treatment.C57BL/6x129Svj mice were anesthetizedand treated with POM-1 (3 mg/kg perhour) or vehicle control via a catheterplaced into the carotid artery. Mice weresubjected to 60 minutes of ischemia alone(�IP) or IP (4�5 minutes) before 60 min-utes of ischemia (�IP). Mice were eutha-nized after 120 minutes of reperfusion, andinfarct sizes were measured by doublestaining with Evans’s blue and triphenyltet-razolium chloride. Infarct sizes areexpressed as percentage of the area atrisk (mean�SEM; n�6).

1786 Circulation October 16, 2007



Cardiac E-NTPDase 1 (CD39) Is SelectivelyInduced by IPAfter having shown that pharmacological inhibition ofE-NTPDases results in increased susceptibility to myocardialischemia and reperfusion injury and abolished cardioprotec-tion by IP, we next sought to define the contribution of

individual ecto-NTDPases (E-NTPDases 1 to 3 and 8) tocardioprotection. For this purpose, we studied transcriptionalresponses of cardiac E-NTPDase expression to IP. We thusperformed 4 cycles of intermittent left coronary artery occlu-sion and reperfusion (5 minutes of ischemia, 5 minutes ofreperfusion) and harvested preconditioned myocardial tissues

A

B

c

NTPDase 1 (CD39)C 30 60 90 120

C

Actin

C 30 60 900

5

10

Time after IP [min]

NTP

Das

e 8

Tran

scrip

t

C 30 60 900

5

10

Time after IP [min]

NTP

Das

e 2

Tran

scrip

t

C 30 60 900

5

10

Time after IP [min]

p<0.001

p<0.05

NTP

Das

e 1

Tran

scrip

t

C 30 60 900

5

10

Time after IP [min]

NTP

Das

e 3

Tran

scrip

t

Figure 2. Cardiac CD39 is induced by IP. A, Murine model of cardiac IP. The IP protocol consisted of 4 cycles of ischemia/reperfusion(5 minutes each), followed by indicated times of reperfusion (A indicates anesthesia induction; T, thoracotomy). B, CD39 (E-NTPDase 1)mRNA is selectively induced by IP. After indicated time periods, the area at risk was excised, total RNA was isolated, and E-NTPDases1 to 3 and E-NTPDase 8 mRNA levels were determined by real-time reverse transcription polymerase chain reaction. Data were calcu-lated relative to an internal housekeeping gene (�-actin) and are expressed as fold change compared with control (no IP) �SEM ateach indicated time (n�6). C, E-NTPDase 1 (CD39) protein is induced by IP. Tissue from the area at risk was excised at the indicatedtime points, flash frozen, and lysed, and proteins were resolved by SDS-PAGE and transferred to nitrocellulose. Membranes wereprobed with an anti–E-NTPDase 1 antibody. The same blot was probed for �-actin expression as a control for protein loading. Onerepresentative experiment of 3 is shown.




at indicated time points after IP treatment for real-timereverse transcription polymerase chain reaction (Figure 2A).Baseline expression levels relative to �-actin revealed lowestexpressional rates of E-NTPDase 8, followed by E-NTPDase3 and E-NTPDase 2, with highest expression of E-NTPDase1 (data not shown). In addition, we found a robust andselective induction of E-NTPDase 1 (CD39) mRNA (eg, 90minutes after cardiac IP, 8.0.�1.5-fold; P�0.001; Figure2B). Analysis of other E-NTPDases (E-NTPDases 2, 3, and8) revealed no significant induction with IP (Figure 2B).Western blot analysis confirmed that CD39 protein is inducedby IP. As shown in Figure 2C, cardiac CD39 protein issignificantly induced as early as 90 minutes after IP treat-ment. To localize CD39 induction by IP to specific cardiactissues, we performed an immunohistochemical staining ofthe preconditioned myocardial tissue for CD39. In controltissues without IP, CD39 was expressed mainly on endotheliawith hardly any staining of cardiomyocytes (Figure 3A). Incontrast, tissues exposed to IP showed dramatic increases inCD39 expression, on both endothelia and myocytes (Figure

3B; isotype controls in Figure 3C and 3D). Taken together,these data provide strong evidence that CD39 is induced byIP on cardiac endothelia and myocytes.

Cardioprotection by IP Is Abolishedin cd39�/� MiceOn the basis of the pharmacological evidence of E-NTPDasesin cardioprotection and the observation of selective inductionof CD39 by IP, we next evaluated functional roles of CD39 inmediating cardioprotection. For this purpose, we used previ-ously described gene-targeted mice for cd39.23 We initiallyinvestigated ATP-mediated effects on cardiovascular param-eters. As shown in Figure 4A, ATP-induced time of brady-cardia was significantly shortened in cd39�/� mice comparedwith wild-type littermates, suggesting decreased extracellularadenosine generation from ATP in cd39�/� mice. We nextperformed IP in cd39�/� mice and littermate controls. Con-sistent with our pharmacological studies with POM-1, infarctsizes caused by 60 minutes of ischemia were significantlyincreased in cd39�/� mice, and cardioprotection by IP was

-IP +IPA B

C D

Isotype control Isotype control

-IP +IP

Figure 3. Cardiac E-NTPDase 1 (CD39)induction occurs on endothelia and myo-cytes. C57BL/6x129Svj mice were sub-jected to IP (�IP), and cardiac tissue fromthe area at risk was harvested 120 min-utes after IP, sectioned, and stained witha polyclonal CD39 antibody. Tissue froma perfused but nonpreconditioned wild-type mouse served as a control. Whereasnonpreconditioned myocardium showedonly weak immunoreactivity on both myo-cytes (arrow) and endothelia (dot) (�IP, A),staining for CD39 was significantlyincreased after preconditioning (�IP, B),predominantly on endothelia (dot) andmyocytes (arrow) (magnification �300 forthe large image and �1000 for photomi-crographs). C and D, Isotype controlmonoclonal antibodies were used to esti-mate the nonspecific binding of target pri-mary antibodies to cell surface antigens.Isotype controls were used at concentra-tions and staining conditions identical tothose of the target primary antibodies forCD39 immunohistochemistry staining.




abolished in cd39�/� mice (Figure 4B and 4C). In additionalstudies, we used 240 minutes of reperfusion time, whichconfirmed results similar to those with 120 minutes ofreperfusion (infarct size without and with IP after 120 and240 minutes of reperfusion, respectively: �IP 52�1.8%versus �IP 51�2.0% and �IP 50.5�2.8% versus �IP51.3�2.3%). Taken together, these studies provide geneticevidence of a cardioprotective role of CD39 in myocardialischemia.

Increases in Cardiac AMP and Adenosine With IPAre Attenuated in cd39�/� MiceOn the basis of the aforementioned findings of CD39 induc-tion and abolished cardioprotection by IP in cd39�/� mice, wehypothesized that increases in cardiac AMP and adenosinelevels with IP are attenuated in cd39�/� mice. Consistent withprevious studies,17 cardiac adenosine and AMP concentra-tions measured immediately after IP (Figure 5A and 5B) were

increased, suggesting that AMP is produced during IP treat-ment. In contrast, cd39�/� mice exhibited attenuated adeno-sine levels in conjunction with decreased AMP concentra-tions after IP treatment. Taken together, these studiesdemonstrate that CD39 functions to increase myocardialAMP/adenosine during preconditioning.

Cardiac CD39 Activity Is Increased After IPBecause cardiac CD39 induction of transcript and proteinoccurs only 90 minutes after IP, these transcriptional effectscannot account for the increased AMP/adenosine productionand cardioprotection that occur immediately after IP. Toconfirm that the increased adenosine/AMP production imme-diately after IP reflects CD39 enzyme activity, we adopted apreviously described technique to measure cardiac CD39activity by assessing the conversion of etheno-ATP to etheno-AMP.1 As a first step, we measured baseline CD39 activity inhearts from cd39�/� mice or corresponding littermate con-

0

50

100

150

ATP

Indu

ced

Tim

e of

Bra

dyca

rdia

[s] p<0.05

cd39-/-cd39+/+A

-IP

+IP

cd39+/+ cd39-/-

B

C

-IP +IP -IP +IP0

10

20

30

40

50

Infa

rct S

ize

[% A

rea

at R

isk]

p<0.05

p<0.01

p<0.01

Figure 4. Larger infarct size and abolished cardio-protection in cd39�/� mice. A, ATP-induced time ofbradycardia is attenuated in cd39�/� mice. cd39�/�

mice or littermate controls received an intravascu-lar ATP bolus (50 �L, 8 mg/mL). Displayed is thetime of bradycardia (reduction from 480 to 120bpm). B, Myocardial ischemia and IP in cd39 �/�

mice. cd39�/� and littermate controls (cd39�/�)were subjected to 60 minutes of ischemia alone(�IP) or 4�5 minutes IP (�IP) before ischemia.Infarct sizes were measured by double stainingwith Evan’s blue and triphenyltetrazolium chlorideafter 120 minutes of reperfusion (mean�SEM;n�6). C, Representative images of myocardial sec-tions of infarcts from the experiment in B are dis-played (blue/dark, retrograde Evan’s blue staining;red and white, area at risk; white, infarcted tissue).




trols, showing �50-fold higher CD39 activity levels incontrols (Figure 5C). Consistent with the aforementionedstudies of changes in tissue AMP and adenosine, cardiacCD39 activity was elevated immediately after IP (Figure 5D).In contrast, cd39�/� mice exhibited attenuated CD39 activityand no additional increase in activity with IP treatment. Thesestudies suggest that CD39 activity is elevated immediatelyafter IP.

Reconstitution of cd39�/� MiceAs proof of principle and to demonstrate that the absence ofcardioprotection by IP in cd39�/� mice reflects lack of extracel-lular AMP, we reconstituted extracellular AMP levels viaintra-arterial infusion (100 �L/h, AMP 8 mg/mL), a dose we

previously determined not to induce hypotension or bradycar-dia (data not shown). This treatment of cd39�/� mice wasassociated with decreased infarct sizes in nonpreconditionedanimals and with complete reconstitution of cardioprotectionby IP in preconditioned mutant mice (Figure 6A). In addi-tional experiments, we reconstituted cd39�/� mice via intra-peritoneal application of soluble potato (apyrase) NTPDase(80 U/kg). Similar to the aforementioned results, solubleapyrase treatment was associated with attenuated infarct sizesand reconstitution of cardioprotective effects of IP in cd39�/�

mice (Figure 6B). These studies confirm an important role ofCD39 in attenuating myocardial ischemia and reperfusioninjury after coronary artery occlusion.

Therapeutic Effects of Apyrase or AMPTreatment During Myocardial Ischemia andReperfusion in Wild-Type MiceAfter having demonstrated that pharmacological or geneticinhibition of extracellular nucleotide-phosphohydrolysis isassociated with increased myocardial infarct sizes and abol-ished cardioprotection by IP, we hypothesized that increasing

-IP +IP -IP +IP0

10

20

30

40

50 p<0.05

p<0.05

p<0.05

p<0.05

Ade

nosi

ne [n

mol

/g w

et w

eigh

t]cd39+/+ cd39-/-

-IP +IP -IP +IP0

100

200

300

400

500

p<0.05

p<0.05

p<0.05

AM

P [n

mol

/g w

et w

eigh

t]

BA

-IP +IP -IP +IP0

50

100

150

200

p<0.05

p<0.01

p<0.01

CD

39 e

nzym

e ac

tivity

[nm

ol/h

/mg]

cd39+/+ cd39-/-0

25

50p<0.01

CD

39 e

nzym

e ac

tivity

[nm

ol/h

/mg]

C D

Figure 5. Increased adenosine, AMP, and CD39 enzyme activitywith IP are attenuated in cd39�/� mice. cd39�/� mice andmatched littermate controls (cd39�/�) were subjected to IP. Pre-conditioned myocardium was snap-frozen immediately after IPtreatment (�IP), whereas littermates were sham operated (�IP).Cardiac adenosine (A), AMP levels (B), and CD39 enzyme activ-ity (C and D) were measured via high-performance liquid chro-matography. Data are presented as mean�SEM (n�6).

A B

-IP +IP -IP +IP0

10

20

30

40

50

60

cd39-/-

Infa

rct S

ize

[% A

rea

at R

isk]

+ Apyrase

- Apyrase

-IP +IP -IP +IP0

10

20

30

40

50

60

cd39-/-In

farc

t Siz

e [%

Are

a at

Ris

k]

+ AMP

- AMP

p<0.01

p<0.01

p<0.01

p<0.01

Figure 6. Reconstitution of cd39�/� mice. A, Treatment ofcd39�/� mice with AMP. cd39�/� mice were administered AMP(�AMP, 100 �L/h of 8 mg/mL AMP) or saline alone (�AMP) viaintra-arterial infusion before 4�5 minutes of IP (�IP) followed by60 minutes of ischemia or 60 minutes of ischemia alone (�IP).Infarct size was determined by Evan’s blue and triphenyltetrazo-lium chloride double staining after 120 minutes of reperfusion(mean�SEM; n�6). B, Treatment of cd39�/� mice with apyrase.cd39�/� mice were reconstituted intraperitoneally with solubleapyrase (�Apyrase, 80 U/kg in sterile saline) or saline alonebefore preconditioning (�IP) followed by 60 minutes of ischemiaor 60 minutes of ischemia alone (�IP). Infarct size was deter-mined by Evan’s blue and triphenyltetrazolium chloride doublestaining (mean�SD; n�6).




extracellular AMP generation ab initio by treatment withsoluble apyrase or AMP infusion may provide a therapeuticapproach. As shown in Figure 7A and 7C, soluble apyrase orAMP treatment of wild-type mice without IP resulted in a43�4.2% or 41�3.7% reduction of infarct size, respectively.To confirm that 120 minutes of reperfusion would be suffi-cient to show a therapeutic effect of AMP/apyrase, we alsodetermined infarct sizes after 240 minutes of reperfusion.Consistent with previous studies, no differences between 120or 240 minutes of reperfusion time were observed (apyrase/AMP-treated mice without IP after 120 minutes and 240minutes of reperfusion, respectively: 23.1�1.9 versus22.5�2.1 and 33.1�2.6 versus 34.1�1.3). Figure 7B and 7Ddepicts representative and corresponding photographs ofmyocardial tissues after double staining with Evan’s blue andtriphenyltetrazolium chloride in soluble apyrase/AMP-treat-ed/untreated wild-type mice without IP. Taken together, thesedata reveal significant cardioprotective effects by treatmentwith AMP or soluble apyrase.

Cardiac Adenosine Is Increased WithApyrase TreatmentWe hypothesized that the cardioprotective effects of apyrasetreatment are related to elevation of cardiac adenosine.

Therefore, we measured cardiac adenosine after apyrasetreatment in wild-type mice with or without prior IP treat-ment. Apyrase treatment alone was associated with a degreeof adenosine elevation similar to that in mice with IPtreatment (Figure 8). However, apyrase treatment in micepretreated with IP was not associated with a “hyperelevation”of adenosine. These findings might explain the fact thatapyrase treatment does not provide additional cardioprotec-tive effects in IP-treated animals (Figure 7A).

DiscussionIn the present study we pursued the contribution of extracel-lular ATP/ADP-phosphohydrolysis to myocardial ischemiaand cardioprotection by IP. As the first step in these experi-ments, we made the observation that pharmacological inhi-bition of E-NTPDases results in decreased myocardial resis-tance to ischemia and abolished cardioprotection by IP. Onthe basis of these results, we next pursued transcriptionalresponses of E-NTPDases by IP. We found that E-NTPDase1/CD39 is selectively induced by preconditioning and thatsuch increases are localized to endothelia and cardiac myo-cytes. Therefore, we subjected gene-targeted mice forE-NTPDase 1 (cd39�/� mice) to myocardial ischemia. Con-

-IP +IP -IP +IP0

10

20

30

40

50

Infa

rct S

ize

[% A

rea

at R

isk]

-IP +IP -IP +IP0

10

20

30

40

50

Infa

rct S

ize

[% A

rea

at R

isk]

A

WT + ApyraseWT - Apyrasep<0.01

p<0.05

p<0.05

- Apyrase

Wildtype no IP

+ Apyrase

Wildtype no IP

- AMP

+ AMP

CWT+ AMPWT - AMP

p<0.01

p<0.05

p<0.05D

B

Figure 7. Therapeutic effects of apyrase or AMPtreatment during myocardial ischemia. A, Treat-ment of wild-type (WT) mice with apyrase.C57BL/6J mice were treated intraperitoneally withsoluble apyrase (�Apyrase, 80 U/kg) or saline 30minutes before 4�5 minutes of IP followed by 60minutes of myocardial ischemia (�IP) or 60 min-utes of ischemia alone (�IP). After 2 hours ofreperfusion, infarct size was determined by Evan’sblue and triphenyltetrazolium chloride double stain-ing (mean�SEM; n�6). B, Representative imagesof myocardial sections from the experiment proce-dure in A are displayed (blue/dark, retrogradeEvan’s blue staining; red and white, area at risk;white, infarcted tissue). C, Treatment of wild-typemice with AMP. C57BL/6J mice were treated withan intra-arterial AMP infusion (starting 30 minutesbefore the experimental procedure, 100 �L/h of 8mg/mL AMP) or normal saline. Mice were thenexposed to 4�5 minutes of IP (�IP) followed by 60minutes of ischemia or 60 minutes of ischemiaalone (�IP). After 2 hours of reperfusion, infarctsize was determined by Evan’s blue and triphenyl-tetrazolium chloride double staining and expressedas percentage of the area at risk (mean�SEM;n�6). D, Representative images of myocardial sec-tions from the experiment procedure in C are dis-played (blue/dark, retrograde Evan’s blue staining;red and white, area at risk; white, infarcted tissue).




sistent with our pharmacological studies, we found increasedinfarct sizes at baseline and abolished cardioprotection by IP.Reconstitution of cd39�/� mice with AMP or soluble NTP-Dases corrected the deleterious phenotype. Similar treatmentin wild-type mice was associated with cardioprotection fromischemia, imitating the cardioprotective effects of IP. Takentogether, these studies provide pharmacological and geneticevidence for a contribution of E-NTPDase 1/CD39–depen-dent ATP/ADP phosphohydrolysis to cardioprotection fromischemia.

These studies are consistent with previous work in animalmodels of cardiac transplantation.28,29 These involved modelsof cardiac transplantation using cd39�/� mice or other miceoverexpressing human CD39. As such, deletion of CD39rendered mice extremely sensitive to vascular injury, andcardiac xenografts of cd39�/� mice rapidly failed withvascular-type rejection in the setting of enhanced plateletaggregation, increased P-selectin expression, and fibrin(ogen)deposition.23 Conversely, upregulation of CD39 in thesetransgenic models by somatic gene transfer by recombinantadenoviruses or administration of soluble apyrase had majorbenefits in prolonging cardiac graft survival and blockingplatelet-mediated thrombosis.20,29 Moreover, another studyexamining the role of CD39 in ischemic brain found in-creased cerebral infarct volumes and reduced postischemicperfusion in a different cd39�/� mouse line.21 In addition,pharmacological reconstitution with soluble CD39 restoredpostischemic cerebral perfusion and rescued these mice fromcerebral injury.

Although these studies have demonstrated a protective rolefor CD39 in preventing thrombus formation, it appearsunlikely in the present study that the observed protective

effects of CD39 during myocardial ischemia are due to amodulation of platelet activation or function. Although extra-cellular nucleotides (particularly ADP) have been implicatedas strong activators of purinergic receptors on platelets,thereby initiating platelet activation and thrombus forma-tion,30 the cd39�/� mice used in the present study had asignificantly prolonged bleeding time, resulting in hemor-rhagic shock, whereas no substantial differences in plasmaP-selectin concentrations could be observed. In addition,cd39�/� mice showed substantial delays in platelet plugformation in an in vivo model of arterial injury and thrombusformation.23 In view of other studies showing potentiation ofischemia/reperfusion injury with platelet activation31 and theknown platelet dysfunction and prolonged bleeding time ofthese cd39�/� mice, it appears unlikely that the presentobservations of increased myocardial infarction and abol-ished cardioprotection by IP are caused by the antithromboticstate of the cd39�/� mice. In addition to a thromboregulatoryrole, E-NTPDase 1–dependent generation of extracellularAMP represents the main substrate for extracellular adeno-sine generation by CD73. It is important to note that otherstudies found that ATPase and ADPase activities weresubstantially decreased in cardiac tissues cultured fromcd39�/� mice, indicating that the contribution of other nucle-otidases to cardiac hydrolysis of extracellular adenine nucle-otides is minimal.23 These findings are also supported by arecent study demonstrating that gene-targeted mice lackingthe major extracellular pathway of ATP/ADP phosphohy-drolysis in the kidney (CD39/ENTPdase 1) are not protectedfrom renal ischemia by IP. In this study, IP-elicited improve-ments of creatinine clearance, urinary flow rate, or histolog-ical tissue damage were completely abolished in cd39�/�

mice.15

It is surprising that baseline infarct sizes were increasedafter inhibition or deletion of CD39. In contrast, prior studiesusing adenosine receptor antagonists or adenosine destructionby deaminase found no difference in infarct size per se. Forexample, an elegant study of enflurane-anesthetized pigsfound that treatment with adenosine deaminase was notassociated with increased myocardial infarct sizes at baseline.In this study, a bradykinin B2 receptor blocker was also used.The authors found that although bradykinin is essentialduring IP of shorter duration, adenosine appears to be moreimportant during IP of longer duration.8 Similarly, a study onthe involvement of endogenous adenosine in IP in swinefound that whereas IP was associated with attenuated infarctsizes and increased interstitial adenosine concentrations,adenosine deaminase treatment per se did not alter infarctsizes.7 Although it is currently not clear why the present studyfound larger baseline infarct sizes with inhibition ofE-NTPDases or after genetic deletion of CD39, these differ-ences may reflect details of the anesthetic management orpreconditioning protocol or could reflect differences betweenthe species that were studied (eg, murine versus porcinestudies). Moreover, it is unclear why in the present studythere appears to be no residual cardioprotection with IP afterE-NTPDase inhibition or CD39 gene deletion. Given thelimitations of any animal model, the authors appreciate thatadenosine is not the only mechanism contributing to precon-

-IP +IP -IP +IP0

5

10

15

20

25

30

35

40

p<0.05

p<0.01A

deno

sine

[nm

ol/g

wet

wei

ght]

WT+ Apyrase

WT - Apyrase

Figure 8. Increased adenosine concentrations after apyrasetreatment. C57BL/6J mice were treated with apyrase intraperito-neally with (�IP, 4�5 minutes) and without IP (�IP). Myocardi-um was snap-frozen, and cardiac adenosine levels were mea-sured via high-performance liquid chromatography. Data arepresented as mean�SEM (n�6). WT indicates wild type.




ditioning. However, these studies suggest that adenosinecontributes significantly to cardioprotection by IP and pointtoward the extracellular metabolism of ATP/ADP as animportant source of adenosine.

The pathophysiological basis underlying the release ofnucleotides and nucleosides during ischemia remains unclear.Previous reports have suggested increased nucleotide release(particularly ATP) during conditions of inflammation orhypoxia.1,32 In addition, studies using large-animal modelscould show that after IP treatment, extracellular adenosinelevels are increased �4-fold.17 In this study, the authorscollected blood samples from the coronary veins, therebyconfirming extracellular localization of adenosine. Moreover,treatment with the CD73 inhibitor �-�-methylene-diphosphate completely blocked the observed increases inextracellular adenosine. These studies highlight that the majorpathway of adenosine generation during cardiac IP is viaectonucleotidase-dependent phosphohydrolysis of precursornucleotides. This is consistent with the present findingsshowing increased adenosine concentrations with IP treat-ment, which are abolished in mice gene targeted for cd39 orcd73.12 Taken together, these studies confirm that during IP,extracellular adenosine mainly stems from upstream metab-olism of nucleotides as opposed to adenosine released fromischemic cells.17

Increased E-NTPase mRNA and protein expression 90minutes after IP cannot be causal for the observed protection(reduced infarct size) for simple temporal reasons. However,increased CD39 enzyme activity and elevated adenosinelevels immediately after IP suggest that additional, nontran-scriptional mechanisms are involved in CD39-dependentcardioprotection. With regard to CD73, the key enzyme ofextracellular adenosine generation, a recent study found animmediate increase of enzyme activity after IP.12 For exam-ple, translocation of preformed enzyme to the cell surfacecould be responsible for the initial increase in enzymeactivity.

Taken together, our data indicate an important contributionof CD39 in mediating cardioprotection during myocardialischemia and reperfusion injury. In addition, the present studyindicates a role of soluble apyrase in the treatment ofmyocardial ischemia. On the basis of the fact that thepresented data are all derived from murine studies, furtherchallenges will include the translation from mice to humans.In addition, it will be important to address pharmacokineticsand additional effects of apyrase treatment, eg, on coagula-tion, blood pressure, or pulmonary function, before suchstudies can be implemented and tested in a clinical setting.Moreover, studies using different time points of treatment(eg, before, during, or after myocardial ischemia) will have todefine the therapeutic time window of apyrase treatment inmyocardial ischemia.

AcknowledgmentsWe gratefully acknowledge Stephanie Zug and Matthias Nagel fortechnical assistance and Jürgen Schnermann and Keiichi Enjyoji forkindly providing mice gene targeted for A1AR and cd39,respectively.

Sources of FundingThis work was supported by a Fortune grant 1416-0-0, Interdiscipli-nary Centre for Clinical Research (IZKF) Verbundprojekt 1597-0-0from the University of Tübingen, and German Research Foundation(DFG) grant EL274/2–2 to Dr Eltzschig and IZKF Nachwuchs-gruppe 1605-0-0 to Dr Eckle.

DisclosuresM. Faigle, S. Laucher, and Drs Eckle, Grenz, Hart, Köhler, Mittel-bronn, and Eltzschig are employees of Tübingen University Hospital.Use of apyrase is currently under consideration for a patent in thetreatment of myocardial ischemia by Tübingen University Hospital.The other authors report no conflicts.

References1. Eltzschig HK, Ibla JC, Furuta GT, Leonard MO, Jacobson KA, Enjyoji K,

Robson SC, Colgan SP. Coordinated adenine nucleotide phosphohy-drolysis and nucleoside signaling in posthypoxic endothelium: role ofectonucleotidases and adenosine A2B receptors. J Exp Med. 2003;198:783–796.

2. Thompson LF, Eltzschig HK, Ibla JC, Van De Wiele CJ, Resta R,Morote-Garcia JC, Colgan SP. Crucial role for ecto-5-nucleotidase(CD73) in vascular leakage during hypoxia. J Exp Med. 2004;200:1395–1405.

3. Eltzschig HK, Thompson LF, Karhausen J, Cotta RJ, Ibla JC, Robson SC,Colgan SP. Endogenous adenosine produced during hypoxia attenuatesneutrophil accumulation: coordination by extracellular nucleotide metab-olism. Blood. 2004;104:3986–3992.

4. Eckle T, Fullbier L, Wehrmann M, Khoury J, Mittelbronn M, Ibla J,Rosenberger P, Eltzschig HK. Identification of ectonucleotidases CD39and CD73 in innate protection during acute lung injury. J Immunol.2007;178:8127–8137.

5. Yellon DM, Downey JM. Preconditioning the myocardium: from cellularphysiology to clinical cardiology. Physiol Rev. 2003;83:1113–1151.

6. Philipp S, Yang XM, Cui L, Davis AM, Downey JM, Cohen MV.Postconditioning protects rabbit hearts through a protein kinase C-aden-osine A2b receptor cascade. Cardiovasc Res. 2006;70:308–314.

7. Schulz R, Rose J, Post H, Heusch G. Involvement of endogenous aden-osine in ischaemic preconditioning in swine. Pflugers Arch. 1995;430:273–282.

8. Schulz R, Post H, Vahlhaus C, Heusch G. Ischemic preconditioning inpigs: a graded phenomenon: its relation to adenosine and bradykinin.Circulation. 1998;98:1022–1029.

9. Post H, Heusch G. Ischemic preconditioning: experimental facts andclinical perspective. Minerva Cardioangiol. 2002;50:569–605.

10. Bullard AJ, Govewalla P, Yellon DM. Erythropoietin protects the myo-cardium against reperfusion injury in vitro and in vivo. Basic Res Cardiol.2005;100:397–403.

11. Sanada S, Kitakaze M. Ischemic preconditioning: emerging evidence,controversy, and translational trials. Int J Cardiol. 2004;97:263–276.

12. Eckle T, Krahn T, Grenz A, Kohler D, Mittelbronn M, Ledent C,Jacobson MA, Osswald H, Thompson LF, Unertl K, Eltzschig HK.Cardioprotection by ecto-5-nucleotidase (CD73) and A2B adenosinereceptors. Circulation. 2007;115:1581–1590.

13. Headrick JP. Ischemic preconditioning: bioenergetic and metabolicchanges and the role of endogenous adenosine. J Mol Cell Cardiol.1996;28:1227–1240.

14. Grenz A, Zhang H, Eckle T, Mittelbronn M, Wehrmann M, Kohle C,Kloor D, Thompson LF, Osswald H, Eltzschig HK. Protective role ofecto-5-nucleotidase (CD73) in renal ischemia. J Am Soc Nephrol. 2007;18:833–845.

15. Grenz A, Zhang H, Hermes M, Eckle T, Klingel K, Huang DY, MullerCE, Robson SC, Osswald H, Eltzschig HK. Contribution of E-NTPDase1(CD39) to renal protection from ischemia-reperfusion injury. FASEB J.2007:epub ahead of print.

16. Kitakaze M, Hori M, Morioka T, Minamino T, Takashima S, Sato H,Shinozaki Y, Chujo M, Mori H, Inoue M, et al. Alpha 1-adrenoceptoractivation mediates the infarct size-limiting effect of ischemic precondi-tioning through augmentation of 5-nucleotidase activity. J Clin Invest.1994;93:2197–2205.

17. Kitakaze M, Hori M, Morioka T, Minamino T, Takashima S, Sato H,Shinozaki Y, Chujo M, Mori H, Inoue M, et al. Infarct size-limiting effectof ischemic preconditioning is blunted by inhibition of 5-nucleotidase




activity and attenuation of adenosine release. Circulation. 1994;89:1237–1246.

18. Ninomiya H, Otani H, Lu K, Uchiyama T, Kido M, Imamura H. Com-plementary role of extracellular ATP and adenosine in ischemic precon-ditioning in the rat heart. Am J Physiol. 2002;282:H1810–H1820.

19. Zimmermann H. Extracellular metabolism of ATP and other nucleotides.Naunyn Schmiedebergs Arch Pharmacol. 2000;362:299–309.

20. Robson SC, Wu Y, Sun X, Knosalla C, Dwyer K, Enjyoji K. Ectonucle-otidases of CD39 family modulate vascular inflammation and thrombosisin transplantation. Semin Thromb Hemost. 2005;31:217–233.

21. Pinsky DJ, Broekman MJ, Peschon JJ, Stocking KL, Fujita T, RamasamyR, Connolly ES Jr, Huang J, Kiss S, Zhang Y, Choudhri TF, McTaggartRA, Liao H, Drosopoulos JH, Price VL, Marcus AJ, Maliszewski CR.Elucidation of the thromboregulatory role of CD39/ectoapyrase in theischemic brain. J Clin Invest. 2002;109:1031–1040.

22. Marcus AJ, Broekman MJ, Drosopoulos JHF, Islam N, Alyonycheva TN,Safier LB, Hajjar KA, Posnett DN, Schoenborn MA, Schooley KA, GayleRB, Maliszewski CR. The endothelial cell ecto-ADPase responsible forinhibition of platelet function is CD39. J Clin Invest. 1997;99:1351–1360.

23. Enjyoji K, Sevigny J, Lin Y, Frenette PS, Christie PD, Esch JS II, ImaiM, Edelberg JM, Rayburn H, Lech M, Beeler DL, Csizmadia E, WagnerDD, Robson SC, Rosenberg RD. Targeted disruption of cd39/ATPdiphosphohydrolase results in disordered hemostasis and thromboregu-lation. Nat Med. 1999;5:1010–1017.

24. Eckle T, Grenz A, Kohler D, Redel A, Falk M, Rolauffs B, Osswald H,Kehl F, Eltzschig HK. Systematic evaluation of a novel model for cardiacischemic preconditioning in mice. Am J Physiol. 2006;291:H2533–H2540.

25. Brunschweiger A, Muller CE. P2 receptors activated by uracil nucleo-tides: an update. Curr Med Chem. 2006;13:289–312.

26. Muller CE, Iqbal J, Baqi Y, Zimmermann H, Rollich A, Stephan H.Polyoxometalates: a new class of potent ecto-nucleoside triphosphatediphosphohydrolase (NTPDase) inhibitors. Bioorg Med Chem Lett. 2006;16:5943–5947.

27. Sun D, Samuelson LC, Yang T, Huang Y, Paliege A, Saunders T, BriggsJ, Schnermann J. Mediation of tubuloglomerular feedback by adenosine:evidence from mice lacking adenosine 1 receptors. Proc Natl Acad SciU S A. 2001;98:9983–9988.

28. Imai M, Takigami K, Guckelberger O, Enjyoji K, Smith RN, Lin Y,Csizmadia E, Sevigny J, Rosenberg RD, Bach FH, Robson SC. Modu-lation of nucleoside triphosphate diphosphohydrolase-1 (NTPDase-1)cd39 in xenograft rejection. Mol Med. 1999;5:743–752.

29. Dwyer KM, Robson SC, Nandurkar HH, Campbell DJ, Gock H,Murray-Segal LJ, Fisicaro N, Mysore TB, Kaczmarek E, Cowan PJ,d’Apice AJ. Thromboregulatory manifestations in human CD39transgenic mice and the implications for thrombotic disease and trans-plantation. J Clin Invest. 2004;113:1440–1446.

30. Dorsam RT, Kunapuli SP. Central role of the P2Y12 receptor in plateletactivation. J Clin Invest. 2004;113:340–345.

31. Xu Y, Huo Y, Toufektsian MC, Ramos SI, Ma Y, Tejani AD, French BA,Yang Z. Activated platelets contribute importantly to myocardial reper-fusion injury. Am J Physiol. 2006;290:H692–H699.

32. Eltzschig HK, Eckle T, Mager A, Kuper N, Karcher C, Weissmuller T,Boengler K, Schulz R, Robson SC, Colgan SP. ATP release from acti-vated neutrophils occurs via connexin 43 and modulates adenosine-dependent endothelial cell function. Circ Res. 2006;99:1100–1108.

CLINICAL PERSPECTIVENovel pharmacological approaches to myocardial ischemia are urgently needed. For example, perioperative myocardialinfarctions are among the leading causes of morbidity and mortality of surgical patients, despite decades of clinical effortstoward diagnostic and treatment strategies. A very promising approach to myocardial infarct size reduction is ischemicpreconditioning, first described in the 1980s, with pretreatment with short time periods of intermittent myocardial ischemiaresulting in a robust reduction of infarct sizes. However, it has been difficult to develop pharmacological approaches toutilize such mechanisms and translate the cardioprotection observed in experimental animals into patient treatment.Although previous studies have shown a pivotal role of extracellular adenosine signaling in cardioprotection by ischemicpreconditioning, the extracellular source for adenosine remained unclear. In the present study, we pursued the hypothesisthat extracellular phosphohydrolysis of ATP or ADP constitutes an important source for adenosine generation incardioprotection by ischemic preconditioning. In fact, pharmacological and genetic approaches in mice suggest that the keyenzyme in this response is CD39, an ectonucleoside-triphosphate diphosphohydrolase, which rapidly converts extracellularATP or ADP to AMP, which in turn is hydrolyzed to adenosine. In fact, soluble apyrase provides pharmacological activitysimilar to that of CD39. Thus, pretreatment with soluble apyrase is associated with increased myocardial adenosine levelsand a degree of cardioprotection similar to that achieved with experimental ischemic preconditioning. To realize thesepossibilities, our results will have to be translated from mice to humans, and the pharmacokinetics, time window fortreatment, and other effects of apyrase (eg, on platelet function or blood pressure) require further investigation.




Laucher, Melanie L. Hart, Simon C. Robson, Christa E. Müller and Holger K. EltzschigDavid Köhler, Tobias Eckle, Marion Faigle, Almut Grenz, Michel Mittelbronn, Stefanie

Protection During Cardiac Ischemia/Reperfusion InjuryCD39/Ectonucleoside Triphosphate Diphosphohydrolase 1 Provides Myocardial

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 2007 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation doi: 10.1161/CIRCULATIONAHA.107.690180

2007;116:1784-1794; originally published online October 1, 2007;Circulation.

http://circ.ahajournals.org/content/116/16/1784World Wide Web at:

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

http://circ.ahajournals.org/content/116/18/e514.full.pdfAn erratum has been published regarding this article. Please see the attached page for:

http://circ.ahajournals.org/content/suppl/2007/10/05/CIRCULATIONAHA.107.690180.DC2.html http://circ.ahajournals.org/content/suppl/2007/10/01/CIRCULATIONAHA.107.690180.DC1.html

Data Supplement (unedited) at:

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/116/16/1784

http://circ.ahajournals.org/content/116/18/e514.full.pdf

http://circ.ahajournals.org/content/suppl/2007/10/01/CIRCULATIONAHA.107.690180.DC1.html

http://circ.ahajournals.org/content/suppl/2007/10/05/CIRCULATIONAHA.107.690180.DC2.html

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

http://www.lww.com/reprints

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


In the version of the article, “CD39/Ectonucleoside Triphosphate Diphosphohydrolase 1 ProvidesMyocardial Protection During Cardiac Ischemia/Reperfusion Injury,” by Köhler et al that wasposted online on October 1, 2007 (DOI: 10.1161/CIRCULATIONAHA.107.690180), an authorname was misspelled.

In the byline of the article, “Marion Faiglem, BS” should have been “Marion Faigle, BS.”

The error has been corrected in the final print version of the article in the October 16, 2007,issue of the journal (Circulation. 2007;117:1784–1794) and in the current online version. Thepublisher regrets the error.

DOI: 10.1161/CIRCULATIONAHA.107.187466

(Circulation. 2007;116:e514.)© 2007 American Heart Association, Inc.

Circulation is available at http://circahajournals.org

e514

Correction

CD39/Ectonucleoside Triphosphate Diphosphohydrolase 1 Provides Myocardial Protection During Cardiac Ischemia/Reperfusion Injury

Documents