385

Acute Impairment of Endothelium-DependentRelaxations to Aggregating PlateletsFollowing Reperfusion Injury in

Canine Coronary ArteriesPaul J. Pearson, Hartzell V. Schaff, and Paul M. Vanhoutte

Experiments were designed to determine whether endothelial injury contributes to augmentedcoronary vascular tone seen during myocardial reperfusion. Canine left anterior descendingcoronary arteries were exposed to ischemia followed by reperfusion (60 minutes each). Rings(3-4 mm) of the reperfused artery and of normal left circumflex (control) coronary arterysegments were prepared. Rings were suspended for isometric force measurement in organ

chambers containing modified Krebs' Ringer bicarbonate solution (370 C, 95% 02-5% C02).Endothelium-independent contractions to KCI and prostaglandin F2a, were unaltered afterreperfusion. Endothelium-dependent relaxations to nitric oxide, sodium nitroprusside, andisoproterenol were comparable in control and reperfused arteries. However, reperfusedcoronary arteries contracted with prostaglandin F2, lost the ability to express endothelium-dependent relaxations to aggregating platelets. Reperfused arterial rings also exhibitedimpaired endothelium-dependent relaxations to acetylcholine, the calcium ionophore A23187,and the platelet-derived compounds ADP and serotonin. Quiescent (noncontracted) reperfusedarterial rings exhibited larger contractions than controls when exposed to aggregating platelets.In such quiescent rings, the endothelium-dependent increase in tension to hemoglobin was

unaltered after reperfusion. Thus, coronary reperfusion impairs the normal endothelium-dependent relaxations to aggregating platelets and vasoactive drugs. This impairment ofplatelet-mediated coronary relaxation could help explain the increased vascular tone andtendency toward vasospasm commonly observed after reperfusion of the coronary arteries.(Circulation Research 1990;67:385-393)

C oronary vasospasm can occur after reperfu-sion of previously ischemic coronaryarteries.1-3 The mechanism of this increased

vascular reactivity is not defined, but an impairmentof endothelium-dependent relaxation is one possibleexplanation. Indeed, if reperfusion-induced injury tothe coronary endothelium were to prevent it fromexerting its protective role against platelet-inducedspasm and thrombosis,4'5 this could play an importantrole in the expression of reperfusion injury to theheart. Altered endothelium-dependent responses inthe large coronary arteries, favoring the occurrenceof arterial spasm in particular in response to aggre-

From the Department of Physiology and Biophysics, Depart-ment of Surgery, Section of Thoracic and Cardiovascular Surgery,Mayo Clinic and Mayo Foundation, Rochester, Minn.

Supported in part by Heart, Lung, and Blood Institute grantsHL-31183 and 31547-04.Address for correspondence: Paul M. Vanhoutte, MD, PhD,

Director, Center for Experimental Therapeutics, Baylor College ofMedicine, One Baylor Plaza, Room 802E, Houston, TX 77030.

Received March 27, 1989; accepted March 27, 1990.

gating platelets, could severely impair reperfusion atthe level of the myocardium. While previous studieshave demonstrated that endothelium-dependentrelaxations to thrombin,6 acetylcholine,7 8 andbradykinin8 are impaired after reperfusion, no infor-mation seems available on the effect of reperfusionon endothelium-dependent responses to platelets.The present study was designed to examine theeffects of reperfusion injury to the canine coronaryartery on endothelium-dependent responses toaggregating platelets and platelet products.

MethodsAnimal Preparation

Heartworm-free mongrel dogs (25-30 kg) of eithersex were anesthetized with intravenous pentobarbitalsodium (30 mg/kg bolus injection; Fort Dodge Lab-oratories, Fort Worth, Tex.), intubated with a cuffedendotracheal tube, and ventilated with 100% oxygen.The respirator was adjusted to keep blood pH andPCO2 in the physiological range. The Po2 was always

by guest on Decem

ber 31, 2017http://circres.ahajournals.org/

Dow

nloaded from

386 Circulation Research Vol 67, No 2, August 1990

above 120 mm Hg. The blood gas values were moni-tored every 15 minutes during the experiments (pH/Blood Gas Analyzer 813, Instrumentation Labora-tory, Lexington, Mass.). The electrocardiogram wascontinuously monitored, and the animals' body tem-perature was maintained by hot water thermal blan-kets. A left lateral thoracotomy was performed in thefourth intercostal space to expose the heart. A 4-mmsegment of the left anterior descending coronaryartery (LAD) was carefully dissected free immedi-ately distal to its first diagonal branch. After a 30- to45-minute stabilization period following the dissec-tion, a control electrocardiogram was recorded alongwith arterial blood gas values. Next, the LAD wasoccluded with a small vascular clamp. Cessation ofblood flow was confirmed by darkening of the myo-cardium subtended by the artery, paradoxical wallmotion of the ischemic myocardium, and electrocar-diographic evidence of myocardial ischemia (typicallyST segment depression in leads II and III). Dysrhyth-mias that occurred were not treated. At 15, 30, and60 minutes after occlusion, the electrocardiogramwas recorded again to document myocardial isch-emia. After 60 minutes of ischemia, the occlusion clipwas gently removed, and the artery was reperfusedfor 60 minutes. Reperfusion was confirmed by vascu-lar and myocardial hyperemia, electrocardiographicchanges, and the resumption of normal myocardialwall motion. If an animal had extensive collateralblood flow and did not exhibit ischemic changes 15minutes after occlusion, it was excluded from thestudy (n =7). The time span for occlusion in thepresent study was chosen because 1) reperfusionaccelerates the disintegration of irreversibly injuredcells, and in the canine myocardium irreversibleinjury begins within 20 minutes after the onset ofpersistent ischemia9; and 2) previous investigators,using a similar model, have used comparable periodsof occlusion.6-8

After 60 minutes of reperfusion, the animal wasexsanguinated, and the beating heart was quicklyremoved and immersed in cold, oxygenated modifiedKrebs' Ringer bicarbonate solution of the followingmillimolar composition: NaCl 118.3, KCl 4.7, MgSO41.2, KH2P04 1.22, CaCl2 2.5, NaHCO3 25.0, CaEDTA0.016, and glucose 11.1 (control solution). The pro-cedures and the handling of the dogs were reviewedand approved by the Institutional Animal Care andUse Committee of the Mayo Foundation.

In Vitro ExperimentsThe LAD (reperfused) and left circumflex coro-

nary artery (control) were carefully dissected freeand placed in control solution. Care was taken toremove as much of the connective tissue as possible.A maximum of four rings (4-5 mm in length) of thereperfused LAD was taken at least 10 mm distal tothe occlusion site. Rings of mid to distal left circum-flex arteries were used as controls. Previous studies6,7and preliminary experiments (data not shown) dem-onstrated no difference in endothelium-dependent

responses between these vessels. In some rings, theendothelium was removed by gentle rubbing of theintimal surface with the tip of a pair of watchmaker'sforceps. Previous histological studies have shown thatthis procedure successfully removes the endothelialcells in large canine arteries while preserving theability of the vascular smooth muscle to contract.10,11The rings were suspended in (25-ml) organ cham-

bers filled with control solution maintained at 370 Cand aerated with 95% 02-5% CO2 (pH 7.4). Eachring was suspended by two stainless steel clips passedthrough the lumen. One clip was anchored to thebottom of the organ chamber; the other was con-nected to a strain gauge (Statham UC2, Gould,Cleveland) for the measurement of isometric force.Rings were placed at the optimal point of theirlength-tension relation by progressively stretchingthem until the contraction to KCl (20 mM), imposedat each level of distension, was maximal.4 In allexperiments, the presence or absence of endotheliumwas confirmed by the response to acetylcholine(10-6 M) of rings contracted with potassium ions(20 mM).10"12 After optimal tension was achieved, therings were allowed to equilibrate for 45 minutesbefore the administration of drugs.

ProtocolsRings of reperfused and control arteries (with and

without endothelium) from the same animal werestudied in parallel. Rings were used for one of sixprotocols (Table 1). Indomethacin (10-5 M) wasadded to the organ bath for 40 minutes before con-traction of rings with prostaglandin F2a to preventsynthesis of endogenous prostanoids. When concen-tration-response experiments to 5-hydroxytryptamine(5HT, serotonin) or platelets were perfused, the ringswere first incubated with the 5HT2-serotonergic antag-onist ketanserin (10`6 M) for 30 minutes to inhibitdirect 5HT2-serotonergic effects of the monoamine onthe vascular smooth muscle. An equilibration periodof at least 30 minutes occurred between each experi-ment in the specific protocols.

Drugs and PlateletsThe following drugs were used: acetylcholine chlo-

ride, ADP, the calcium ionophore A23187, catalase(bovine), hemoglobin (bovine), 5-hydroxytryptaminecreatinine sulfate, (±)-isoproterenol hydrochloride,indomethacin, KCl, prostaglandin F2a, sodium nitro-prusside, superoxide dismutase (bovine) (all fromSigma Chemical Co., St. Louis), ketanserin tartrate(Janssen Pharmaceutica Inc., Piscataway, N.J.), andnitric oxide (Union Carbide, Chicago). All drugswere prepared daily with distilled water except forindomethacin, which was dissolved in Na2CO3(10-5 M), and the calcium ionophore A23187, whichwas dissolved in dimethyl sulfoxide, with furtherdilutions obtained in distilled water. The concentra-tions are expressed as final molar concentration inthe organ chamber. A platelet-rich solution wasprepared by centrifugation as described previous-

by guest on Decem

ber 31, 2017http://circres.ahajournals.org/

Dow

nloaded from

Pearson et al Reperfusion and Endothelium-Dependent Responses

TABLE 1. Experimental Protocols Performed on Control and Reperfused Coronary Artery Rings With and WithoutEndothelium

Order of agonists

Protocol 1 2 3 4

I Acetylcholine* Isoproterenol* Platelets*(10-9-10-4 M) (10--_10-4 M) (1,000-100,000 platelets/,ul final

bath concentration)II ADP* Sodium nitroprusside* A23187*

(10-9_10-4 M) (10-9-_10-4 M) (10-9_10-4 M)III 5-Hydroxytryptamine* ADP*t KClt

(10-_-10-4 M) (10-_-10-4 M) (5-40 mM)IV Prostaglandin F2,4 Acetylcholine*t Platelets t A23187*t

(10-9_1o-5 M) (10- -_0-4 M) (single dose of 75,000 (10-9_10-6 M)platelets/,ul; final bathconcentration)

V Hemoglobintt(10-5 M)

VI Nitric oxide* ADP*(10-9-10-6 M) (10-9-10-4 M) in presence of

superoxide dismutase(150 units/ml) and catalase(1,200 units/ml)

*Relaxations to agonist in rings contracted with prostaglandin F2a (2x 10-6 M). The response to the Ca2 ionophoreA23187 is taken as last because previous experience has demonstrated that it is possibly irreversible.

tExperiment performed in the presence of indomethacin (10-5 M).tExperiment performed in the presence of ketanserin (10-6 M).

ly.413"14 Briefly, autologous blood (360 ml) was drawnfrom the left carotid artery of the dog into citrateanticoagulant to yield final concentrations of 9.3 mMsodium citrate, 0.7 mM citric acid, and 14 mMdextrose.13,14 The blood was centrifuged for 40 min-utes at 500 rpm (55g) at room temperature, and theplatelet-rich plasma was drawn off with a pipette. Anequal volume of cold citrate anticoagulant solution(in mM: sodium citrate 93, citric acid 7, dextrose 105,and KCl 5, pH 6.5) was added to the platelet-richplasma, and the mixture was centrifuged for 20minutes at 1,600 rpm (570g). The supernatant wasdiscarded, and the remaining platelet pellet wasresuspended in a small volume of the second citrateanticoagulant mixture. A platelet count of this sus-pension was then obtained (Coulter Electronics Inc.,Hialeah, Fla.). Platelet aggregation on exposure tothe collagen of the blood vessel wall and the calcium-containing modified Krebs' Ringer bicarbonate solu-tion was evidenced by clearing of the initially turbidsolution and formation of visible platelet clumps'516;under the experimental conditions of the presentstudy, it is likely that the subsequent vascularresponses to the platelet products are transmittedmainly through the adventitial wall. Earlier studieshave demonstrated the release of 5HT and throm-boxane B2 under these conditions.

Bovine hemoglobin (type 1) contains a mixture ofoxyhemoglobin and the oxidized derivative, methemo-globin. Oxyhemoglobin, which inactivates endothelium-derived relaxing factor,17 was prepared by adding600 mg bovine hemoglobin to 10 ml distilled watercontaining 70 mg sodium dithionite (Na2S204). Sodiumdithionite was then removed by dialysis in 15 1 distilledwater (containing 0.001% EDTA) for 2 hours at room

temperature. During dialysis, the water was bubbledwith nitrogen. The percentage of oxyhemoglobin wasdetermined spectrophotometrically.'7

Nitric oxide from a cylinder was used to fill a glassgas bulb fitted with a silicon injection septum. By usinga glass syringe, gas was removed from the bulb andinjected into another glass bulb, which had been filledwith 100 ml distilled water (which had been bubbledwith helium for 3 hours) to give stock solutions ofnitric oxide (4xl10-5 M, 4x10-4 M, and 4x10-3 M).18

Data AnalysisThe results are expressed as mean+SEM. In all

experiments, n refers to the number of animals fromwhich rings were taken. In rings contracted withprostaglandin F2a, responses are expressed as percentchanges from the contracted levels, and in quiescent(noncontracted) rings responses are expressed aspercent changes of the response to KCl (20 mM)unless otherwise stated. For relaxations, the negativelogarithm of the effective molar concentration ofagonist causing 50% inhibition of the contractions toprostaglandin F2a was calculated for concentration-response curves, and the mean of these values ispresented. Statistical evaluation of the data wasperformed by Student's t test for either paired orunpaired observations. Values of p<0.05 were con-sidered to be statistically significant.

ResultsPlatelets

Quiescent rings. In both control and reperfusedquiescent (noncontracted) rings without endothe-lium, aggregating platelets (single dose of 75,000

387

by guest on Decem

ber 31, 2017http://circres.ahajournals.org/

Dow

nloaded from

388 Circulation Research Vol 67, No 2, August 1990

With endothellum Without endothellumFIGURE 1. Response to aggregating platelets (75,000/p) ofquiescent control and reperfused coronary artery rings withand without endothelium (n=6). Data are mean +SEM andare expressed as percent of contraction to potassium ions (20mM). Astersk denotes significance between control and reper-fused rings with endothelium (Student's t test for pairedobservation; p<0.05).

platelets/,ul final organ chamber concentration)caused comparable contractions (47+21% and40±20% of the contraction to 20 mM KCl for controland reperfused rings, respectively). In control ringswith endothelium, the platelets induced a moderatecontraction (13+±7%), while a significantly greatercontraction (35± 11%) was induced in the reperfusedrings (Figure 1).

Contracted rings. In contracted control and reper-fused rings without endothelium, there was no sig-nificant difference in response to aggregating platelets(1-100,000 platelets/,l final organ bath concentra-tion). However, in control rings with endothelium,platelets caused concentration-dependent relaxationsand a return to baseline tension; these relaxationswere significantly greater than in control rings withoutendothelium (Figure 2).

In contracted reperfused rings with endothelium,the lower concentration of platelets produced addi-tional contraction (producing a 29±10% increase intension of the initial contraction to prostaglandinF2a). At higher concentrations, platelets causedconcentration-dependent relaxations. However, evenat the highest platelet concentration, reperfusedrings relaxed only an average of 41% of the initialprostaglandin contraction. There was no significantdifference in relaxation to aggregating platelets inreperfused rings with and without endothelium.

Platelet ProductsADP. ADP caused comparable concentration-

dependent relaxations in control and reperfusedrings without endothelium that did not reach thebaseline value. In control and reperfused rings withendothelium, ADP caused concentration-dependentrelaxations to baseline that were significantly greaterthan in rings without endothelium. However, in thereperfused rings with endothelium, there was a sig-

140120

100

8060

40

20

00 10 20 40 60 80 100

Platelets, thousands /.PLFIGURE 2. Concentration-response curves to aggregatingplatelets in control and reperfused coronary arterial rings withand without endothelium (n=6). Rings were contracted withprostaglandin F2a (PGF2J (2 x1-6 M) and incubated withthe 5-hydroxytryptamine serotonergic antagonist ketanserin(10-6 M, 30 minutes). Values are mean+±SEM. Asterisksdenote significance between control and reperfused rings withendothelium (Student's t test for paired observation; p<0.05).LCX, left circumflex coronary artery; LAD, left anteriordescending coronary artery.

nificant shift of the concentration-response curve tothe right compared with control rings with endothe-lium (Figure 3, Table 2.). In control rings withoutendothelium, indomethacin significantly attenuatedthe relaxation to 10` M ADP; in reperfused rings, asignificant attenuation was observed at 3 x i-` andi0` M ADP (Table 3). Indomethacin also attenu-ated relaxations to ADP in control arteries withendothelium (10`' to 3x 106 M). Indomethacin hadno effect on reperfused arteries with endothelium.

Endothelium:with without

control LCX * o

reperfused LAD a

100

e 80

60

200~

0k

9 8 7 6 5 4Adenosine diphosphate,-log M

FIGURE 3. Concentration-response curves toADP in controland reperfused coronary arterial rings with and without endo-thelium (n=7). Rings were contracted with prostaglandin F2a(PGF2oJ (2x10-6 M). Values are mean+±SEM. Asterisksdenote significance between control and reperfused rings withendothelium (Student's t testforpaired observation; p<0.05).LCX, left circumflex coronary artery; LAD, left anteriordescending coronary artery.

1 0Or

6OF.0

c

0c

(I)

cU)to

Endothelium:with without

control LCX . oreperfused LAD a o

01Control Reperfused

c0

0a-

c

c~J00)

10.Control Reperfuseda Af

by guest on Decem

ber 31, 2017http://circres.ahajournals.org/

Dow

nloaded from

Pearson et al Reperfusion and Endothelium-Dependent Responses 389

TABLE 2. Values for 50%o Inhibition (-Log M) of Control andReperfused Rings With and Without Endothelium

Control Reperfused

With endotheliumAcetylcholine 7.59±0.14 7.12±0.19*ADP 6.98±0.12 6.11±0.16*A23187 7.59±0.13 7.34±0.13*

Without endotheliumNitric oxide 6.78±0.18 6.91+0.20Sodium nitroprusside 7.78±0.14 7.47±0.15Isoproterenol 7.54±0.19 7.33±0.15

*Significant difference from control rings by Student's t test forpaired observation (p<0.05).

Superoxide dismutase and catalase (150 units/ml and1,200 units/ml, respectively, added together), poten-tiated endothelium-dependent responses in controland reperfused arteries but did not modify the dif-ference in response between control and reperfusedcoronary segments (n=3, data not shown).

5-Hydroxytryptamine. There was no significant dif-ference between control and reperfused rings with-out endothelium in response to 5HT (serotonin). Incontrol and reperfused rings with endothelium, sero-

tonin caused concentration-dependent relaxationsthat were significantly greater than in rings withoutendothelium. However, in reperfused rings with endo-thelium, there was a significant shift of the concentra-tion-response curve to the right, and the reperfusedrings relaxed to only 38+11% of the initial contractionto prostaglandin F2a as opposed to 11+4% for controlrings with endothelium (Figure 4).

Other endothelium-dependent dilators. Acetylcho-line (10-9-10-4 M) (Figure 5) and the calcium iono-phore A23187 (109-106 M) (Figure 6) causedendothelium-dependent, concentration-dependentrelaxations in control and reperfused arteries withendothelium. However, in the reperfused rings, theconcentration-response curves to these agonists wereshifted to the right (Table 2). Indomethacin had no

significant effect on relaxations to acetylcholine ineither control or reperfused arteries (Table 3). Indo-methacin attenuated relaxations to the calcium ion-ophore A23187 (10-8 to 3 x 10-7 M) in control but notin reperfused rings (Table 3).Basal release of endothelium-derived relaxing factor.

In quiescent coronary rings with endothelium, hemo-globin (10-5 M) caused comparable increases intension in control and reperfused vessels (14.0±4.8%

TABLE 3. Concentration-Dependent Relaxations in Control and Reperfused Coronary Arterial Rings With andWithout Endothelium in the Presence and Absence of Indomethacin

Drug concentration (-log M)9 8.5 8 7.5 7 6.5 6 5.5 5 4.5 4

ReperfusedWith endotheliumAch 100 98±2 90±9 74±10 42±11 17±8 8±4 4±2 4±2 3±2 3±2Ach+Indo 100 84±14 35±10 6±6 5±5 5±5 5±5 5±5 5±5A23187 100 87±6 42± 16 6±4 0.3±0.3 0A23187+Indo 100 99±1 98±1 67±8 23±13 8±8 0ADP 100 95±2 72±7 36±10 15±9 6±4 0.4±0.4 0.4±0.4ADP+Indo 100 98±1 85±6 62±11 33±11 17±11 7±5 6±5

Without endotheliumADP 100 99±1 99±1 99±1 99±1 99±1 93±5 84±9 69±12 59±11 43±11ADP+Indo 100 99±1 99±1 98±1 98±1 97±1 96±1 92±1 87±2* 77±4*

Control

With endotheliumAch 100 97±2 82±8 46±14 13±8 0Ach+Indo 100 99±1 64±8 1±1 0A23187 100 99±1 44±14 5±4 0A23187+Indo 100 99±1 95±2* 33±9* 0ADP 100 96±3 91±5 49±12 15±8 2±1 0

ADP+Indo 100 97±1* 65±9* 13±6 1±1 0

Without endotheliumADP 100 101±1 103±3 102±2 101±3 97±3 94±3 84±4 74±6 67±9 60±11

ADP+Indo 100 99±1 99±1 97±2 94±3 91±4* 87±5 84±5

Data are presented as percent of initial prostaglandin F2a contraction (2x 10-6 M); mean±SEM. ACh, acetylcholine;Indo, indomethacin.

*Significant difference between rings in the presence and absence of indomethacin to a particular agonist (Student'st test for unpaired observations; p<0.05).

by guest on Decem

ber 31, 2017http://circres.ahajournals.org/

Dow

nloaded from

390 Circulation Research Vol 67, No 2, August 1990

Endothelium:with without

control LCX * oreperfused L AD a o

*X

9 8 7 6 5 45-Hydroxytryptamine,-log M

FIGURE 4. Concentration-response curves to 5-hydroxy-tryptamine in control and reperfused coronary arterial ringswith and without endothelium (n=6). Rings were contractedwith prostaglandin F2a (PGF2a) (2X 1o6 M) and incubatedwith the 5-hydroxytryptamine serotonergic antagonist ketan-serin (10-6 M, 30 minutes). Values are mean±SEM. Asterisksdenote significance between control and reperfused rings withendothelium (Student's t testforpaired observation; p<0.05).LCX, left circumflex coronary artery; LAD, left anteriordescending coronary artery.

and 12.4±3.0% of the contraction to 20 mM KCl,respectively) (n=6, data not shown).Endothelium-independent dilators. Nitric oxide

(10-'-10-S M) caused comparable concentration-dependent relaxations in control and reperfusedrings without endothelium (Figure 7, Table 2). Inaddition, sodium nitroprusside (10-9_10-6 M) causedcomparable, concentration-dependent relaxations incontrol and reperfused rings without endothelium(Table 2). Also, in control and reperfused ringswithout endothelium, there was no difference in the

* control LCX* reperfused LAD

c 200

' 40

0gr 60

Acetylcholine, -log MFIGURE 5. Concentration-response curves to acetylcholinein control and reperfused coronary arterial rings with endothe-lium (n=9). Rings were contracted with prostaglandin F2a(PGF2a) (2x10`6 M). Values are mean+±SEM. Asterisksdenote significance (Student's t test for paired observation;p<O.OS). LCX, left circumflex coronary artery; LAD, leftanterior descending coronary artery.

* control LCX* reperfused LAD

Calcium ionophore A23187, - log M

FIGURE 6. Concentration-response curves to the calciumionophore A23187 in control and reperfused coronary arterialrings with endothelium (n= 7). Rings were contracted withprostaglandin F2a (PGF2J (2X10-6 M). Values are mean±+SEM. Asterisks denote significance (Student's t test forpairedobservation; p<0.05). LCX, left circumflex coronary artery;LAD, left anterior descending coronary artery.

relaxations induced by ADP, serotonin, and isopro-terenol (10-9-1O-4 M) (Figures 3, 4, and 8).

Contractions. KCl (5-40 mM) and prostaglandinF2a (10-9-10-5 M) caused comparable, concentration-dependent contractions in control and reperfusedrings with and without endothelium (Table 4). Therewas no difference in the maximal response to eitheragonist between the two groups. Within each group,there was no difference between rings with andwithout endothelium.

DiscussionThe present study was undertaken to examine

endothelium-dependent responses to aggregatingplatelets and vasoactive drugs in the canine coronaryartery after acute occlusion and reperfusion. Themajor findings were 1) contracted reperfused coronaryarteries with endothelium lose their ability to generateendothelium-dependent relaxations to aggregating

control LCX = oreperfused LAD=O

9 8 7 6Nitric oxide,-log M

5

FIGURE 7. Concentration-response curves to nitric oxide incontrol and reperfused coronary arterial rings without endo-thelium (n=5). Rings were contracted with prostaglandin F2a(PGF2a) (2x10`6 M). Values are mean+±SEM. LCX, leftcircumflex coronary artery; LAD, left anterior descendingcoronary artery.

c0

CL0

c.)

a-0

140120100

80

L 6040

200

100c0

._o3 80a

c 60U

L' 40(2,.- 20

09

by guest on Decem

ber 31, 2017http://circres.ahajournals.org/

Dow

nloaded from

Pearson et al Reperfusion and Endothelium-Dependent Responses 391

C Jo0-c- oEL |-

o c*0

*=control LCXO=reperfused LAD

100-

80-

60-

40-

20-

0-.-9 8 7 6 5 4

Isoproterenol, -log M

FIGURE 8. Concentration-response curves to isoproterenol incontrol and reperfused coronary arterial rings without endo-thelium (n=6). Rings were contracted with prostaglandin F2a(PGF2a) (2x106 M). Values are mean+±SEM. LCX, leftcircumflex coronary artery; LAD, left anterior descendingcoronary artery.

platelets and, as a consequence, exhibit additionalcontraction after exposure to lower numbers of aggre-gating platelets; 2) quiescent reperfused arteries withendothelium exhibit an exaggerated constrictorresponse to aggregating platelets; 3) reperfusionimpairs endothelium-dependent relaxations to ADP,serotonin, acetylcholine, and the calcium ionophoreA23187; 4) reperfusion apparently does not alter thebasal release of endothelium-derived relaxing factor;and 5) the ability of the vascular smooth muscle tocontract or relax is unaltered after 60 minutes ofischemia followed by 60 minutes of reperfusion.As reperfused coronary arteries exhibit a severe

attenuation of endothelium-dependent relaxations toaggregating platelets, reperfused arteries with endo-thelium act in a comparable manner to those without

endothelium. Potentiated contraction, as a conse-quence of attenuated relaxations to aggregatingplatelets and their products, is also apparent afterendothelial regeneration14 and is a common findingin other types of vascular disease such as hyperten-sion15 and atherosclerosis.19 The mechanism of thisaberrant response could be multifactorial. First,release of endothelium-derived relaxing factor inresponse to the platelet-mediated vasodilator couldbe impaired. In the canine coronary artery, ADP(and ATP) and serotonin are the platelet productsthat evoke endothelium-dependent relaxations.4,513In this species, the adenine nucleotides apparentlyexert the key role in mediating endothelium-dependent relaxation to aggregating platelets.13However, reperfusion impairs endothelium-depen-dent relaxations to both serotonin and ADP whileleaving the direct effect of these agonists on smoothmuscle unaltered. Thus, the impaired response toboth platelet products could cause the aberrantplatelet response. This is in agreement with datafrom atherosclerotic porcine blood vessels in whichdecreased responses to both of these platelet-mediated vasodilators are impaired.16 However, thisdiffers from the findings in porcine coronary arterieswith regenerated endothelium or in the aorta of thespontaneous hypertensive rat in which the alteredendothelium-dependent responses to serotonin arethe primary abnormality that underlies the abnormalresponse to the platelets.14,15,20Another possible mechanism of the impaired

platelet response would be the production of aconstricting prostanoid15'21 or decreased productionof a relaxing prostanoid22 by the endothelium. Exper-iments performed in the presence of indomethacin

TABLE 4. Concentration-Dependent Contractions to Potassium Ions and Prostaglandin F2. in Control and ReperfusedCanine Coronary Arterial Rings

With endothelium (n=5)Control Reperfused

Without endothelium (n=5)Control Reperfused

Potassium ions (mM)*510203040Maximal response (g)

Prostaglandin F2a (-Log M)7.576.565.5MMaximal response (g)

8±333±778±394±1100

12±2

1±0.75±3322±449±581±41009±2

7±328±678±394±1100

12±1

1±0.95±223±353±483±2100

10±2

9+433±876±596±2100

11±1

0.6±0.64±322±752±878±41009±1

7±426±771±393±110012±1

2±24±226±460±586±21008±1

Data are presented as percent of maximal response to potassium ions (40 mM) or prostaglandin F2a (10-5 M);mean±SEM.

*Concentration of potassium ions above control solution concentration.

by guest on Decem

ber 31, 2017http://circres.ahajournals.org/

Dow

nloaded from

392 Circulation Research Vol 67, No 2, August 1990

support the hypothesis that reperfused arteries havea decreased production of a relaxing prostanoidbecause its presence attenuated the endothelium-dependent responses to ADP and the calcium iono-phore in control but not reperfused rings.The endothelium-dependent responses to acetyl-

choline and the calcium ionophore A23187 were alsoattenuated by reperfusion. Such a general impair-ment of endothelium-dependent responses is alsoobserved in atherosclerosis.23 This differs fromchanges in porcine coronary arteries with regener-ated endothelium in which there is a selectivedecrease in endothelium-dependent response tosome agonists, while the response to others remainsintact.14There are several possible explanations for the

general impairment of endothelium-dependentrelaxations in reperfused arteries. First, the smoothmuscle of the arterial wall could lose its ability torespond to endothelium-derived relaxing factor.However, this is not the case as there was no differ-ence between control and reperfused arteries inresponse to nitric oxide, which has been identified asone major chemical mediator of endothelium-dependent relaxations.18'24 Also, there was no differ-ence between control and reperfused arteries inresponse to sodium nitroprusside which, likeendothelium-derived relaxing factor, activates gua-nylate cyclase in the vascular smooth muscle to causerelaxation.25,26 In addition, there was no difference inthe ability of control and reperfused arteries withoutendothelium to relax in response to either isoprotere-nol or ADP, suggesting that the potential of thesmooth muscle to relax is unaltered after reperfusion.Second, there could be a decreased production ofrelaxing factor(s) by the endothelial cell. This is pos-sible because other types of cells exhibit impairedsynthetic processes after sustaining reperfusioninjury.9 Third, there could be impairment of therelease of the factor from the endothelial cell becauseother cells also exhibit membrane transport and struc-tural dysfunction after reperfusion.27 Finally, therecould be an accelerated breakdown of endothelium-derived relaxing factor by the artery. As reperfusionpromotes the generation of oxygen-derived free radi-cals in other tissues,2829 and as superoxide anionsinactivate the relaxing factor,3031 it is possible thatradicals continuously produced by reperfused coro-nary arteries scavenge endothelium-derived relaxingfactor. However, this explanation is unlikely, becauseexperiments done in the presence of scavengers offree radicals failed to alter the difference in responsebetween control and reperfused arteries.

Clinical ImplicationsThe present study suggests that a decrease in the

ability of the endothelium to produce relaxing factorcould become a reason for an exaggerated vasocon-strictor response to platelet products. Endothelium-derived relaxing factor not only relaxes vascularsmooth muscle, but also inhibits platelet adhesion,32

and is itself a potent antiaggregatory substance.33,34 Ifthe release of endothelium-derived relaxing factor isreduced in reperfused arteries, platelet adhesion willnot be inhibited; since the antiaggregatory effect ofendothelium-derived relaxing factor would also becurtailed, platelet aggregation would be favored incoronary arteries damaged by the reperfusion. Thisinterpretation is strengthened by the demonstrationthat platelet deposition is promoted by damage to theendothelium and that reperfusion of the coronaryarteries leads to injury of the endothelial lining.7Thus, reperfusion could lead to unchecked plateletdeposition and aggregation in coronary arteries. Thepresent study demonstrates that the endothelium ofthe reperfused segments of the coronary artery is nolonger capable of breaking the constrictor response tothe platelet products, which should favor the occur-rence of vasoconstriction. In fact, in vivo a correlationexists between quantitative platelet deposition andlocalized vasoconstriction at the site of endothelialinjury secondary to trauma.34 The present resultssuggest that an augmented adhesion and aggregationof platelets, and constriction of the reperfused coro-nary arteries to the platelet products resulting fromendothelial dysfunction, could play a key role in thelocal vasoconstrictions. Aggregating platelets havebeen implicated in the potentiation of myocardialischemia,35 progressive reperfusion injury,36'37 and sur-gically induced myocardial ischemia.38

In conclusion, reperfusion impairs the normalrelaxation to aggregating platelets and vasoactivedrugs mediated by the coronary endothelium. Thisimpairment of relaxation could help explain theincreased vascular tone and tendency toward vaso-constriction commonly observed after reperfusion ofthe coronary vasculature.

AcknowledgmentsThe authors wish to thank Mr. R.R. Lorenz for

preparing the figures and Mrs. J. Krage, Ms. Kath-leen Kros, and Mrs. Cynthia Lindsay for secretarialassistance.

References1. Dorros G, Cowley MJ, Simpson J, Bentivoglio LG, Block PC,

Bourassa M, Detre K, Gosselin AJ, Gruntzig AR, Kelsey SF,Kent KM, Mock MB, Mullin SM, Myler RK, Passamani ER,Stertzer SH, Williams DO: Percutaneous transluminal coro-nary angioplasty: Report of complications from the NationalHeart, Lung, and Blood Institute PTCA Registry. Circulation1983;67:723-730

2. Lockerman ZS, Rose DM, Cunningham Jr, Lichstein E:Postoperative St-segment elevation in coronary artery bypasssurgery. Chest 1986;89:647-651

3. Cowley MJ, Dorros G, Kelsey S, VanRaden M, Detre K:Acute coronary events associated with percutaneous translu-minal coronary angioplasty. Am J Cardiol 1984;53:12c-16c

4. Cohen RA, Shepherd JT, Vanhoutte PM: Inhibitory role ofendothelium in the response of isolated coronary arteries toplatelets. Science 1983;221:273-274

5. Cohen RA, Shepherd JT, Vanhoutte PM: 5-Hydroxytryptaminecan mediate endothelium-dependent relaxation of coronaryarteries. Am J Physiol 1983;245:H1077-H1080

by guest on Decem

ber 31, 2017http://circres.ahajournals.org/

Dow

nloaded from

Pearson et al Reperfusion and Endothelium-Dependent Responses 393

6. Ku DD: Coronary vascular reactivity after acute myocardialischemia. Science 1982;218:576-578

7. VanBenthuysen KM, McMurtry IF, Horwitz LD: Reperfusionafter acute coronary occlusion in dogs impairs endothelium-dependent relaxation to acetylcholine and augments contrac-tile reactivity in vitro. J Clin Invest 1987;79:265-274

8. Mehta JL, Nichols WW, Donnelly WH, Lawson DL, SaldeenTGP: Impaired canine coronary vasodilator response to ace-

tylcholine and bradykinin after occlusion-reperfusion. Circ Res1989;64:43-54

9. Jennings RB, Reimer KA: Factors involved in salvaging isch-emic myocardium: Effect of reperfusion of arterial blood.Circulation 1983;68(suppl I):I-25-I-36

10. DeMey JG, Vanhoutte PM: Role of the intima ion cholinergicand purinergic relaxation of isolated canine femoral arteries.JPhysiol (Lond) 1981;316:347-355

11. DeMey JG, Vanhoutte PM: Heterogeneous behavior of thecanine arterial and venous wall: Importance of the endothe-lium. Circ Res 1982;51:437-439

12. Furchgott RF, Zawadzki JV: The obligatory role of endothe-lial cells in the relaxation of arterial smooth muscle byacetylcholine. Nature 1980;228:373-376

13. Houston DS, Shepherd JT, Vanhoutte PM: Adenine nucleo-tides, serotonin and endothelium-relaxations to platelets.Am J Physiol 1985;248:H389-H395

14. Shimokawa H, Aarhus LL, Vanhoutte PM: Porcine coronary

arteries with regenerated endothelium have a reducedendothelium-dependent responsiveness to aggregating plate-lets and serotonin. Circ Res 1987;61:256-270

15. Luscher TF, Vanhoutte PM: Endothelium-dependentresponses to platelets and serotonin in spontaneously hyper-tensive rats. Hypertension 1986;8(suppl II):II-55-II-60

16. Shimokawa H, Vanhoutte PM: Dietary cod-liver oil improvesendothelium-dependent responses in hypercholesterolemicand atherosclerotic porcine coronary arteries. Circulation1988;78:1421-1430

17. Martin W, Villani GM, Jothianandan D, Furchgott RF: Selec-tive blockade of endothelium-dependent and glyceryltrinitrate-induced relaxation by hemoglobin and methyleneblue in the rabbit aorta. J Phamacol Exp Ther 1985;232:708-716

18. Palmer RMJ, Ferrige AG, Moncada S: The release of nitricoxide by vascular endothelial cells accounts for the activity ofEDRF. Nature 1987;327:524-526

19. Heistad DD, Armstrong ML, Marcus ML, Piegors DJ, MarkAL: Augmented responses to vasoconstrictor stimuli in hyper-cholesterolemic and atherosclerotic monkeys. Circ Res 1984;54:711-718

20. Vanhoutte PM, Luscher TF: Vascular endothelium and hyper-tension. J Cardiovasc Pharmacol 1987;10(suppl 4):S1-S24

21. Miller VM, Vanhoutte PM: Endothelium-dependent contrac-tions to arachidonic acid are mediated by products of cyclo-oxygenase in canine veins. Am J Physiol 1985;248:H432-H437

22. Gryglewski FJ, Dembinska-Kiec A, Zumuda A, Gryglewska T:Prostacyclin and thromboxane A2 biosynthesis capacities ofheart, arteries, and platelets at various stages of experimentalatherosclerosis in rabbits. Atherosclerosis 1978;31:385-394

23. Frieman RC, Mitchell GG, Heistad DD, Armstrong ML,Harrison DG: Atherosclerosis impairs endothelium-dependentvascular relaxation to acetylcholine and thrombin in primates.Circ Res 1986;58:783-789

24. Ignarro LJ, Byrns RE, Buga GM, Wood KS: Endothelium-derived relaxing factor from pulmonary artery and vein pos-sesses pharmacologic and chemical properties identical tothose of nitric oxide radical. Circ Res 1987;61:866-879

25. Katusbi S, Arnold WP, Mittal CK, Murad F: Effects of sodiumnitroprusside, nitroglycerin and sodium oxide on levels ofcyclic nucleotides and mechanical activity of various tissues.J Cyclic Nucleotide Res 1984;55:575-579

26. Furchgott RF: Role of endothelium in responses of vascularsmooth muscle. Circ Res 1983;53:557-573

27. Frame LH, Lopez JA, Khaw BA, Fallon JT, Haber E, PowellWJ Jr: Early membrane damage during coronary reperfusionin dogs: Detection by radiolabeled anticardiac myosin (Fab')2.J Clin Invest 1983;72:535-544

28. Hammond B, Hess ML: The oxygen free radical system:Potential mediator of myocardial injury. Basic Concepts Car-diol 1985;6:215-220

29. Peterson DA, Asinger RW, Elsperger KJ, Homans DC, EatonJW: Species may cause myocardial reperfusion injury. BiochemBiophys Res Commun 1985;127:87-93

30. Rubanyi GM, Vanhoutte PM: Oxygen-derived radicals, endo-thelium and responsiveness of vascular smooth muscle. Am JPhysiol 1986;250:H815-H821

31. Gryzlewski FJ, Palmer RMHJ, Moncada S: Superoxide anionis involved in the breakdown of endothelium-derived vascularrelaxing factor. Nature 1986;320:454-456

32. Sneddon JH, Vane JR: Endothelium-derived relaxing factorreduces platelet adhesion to bovine endothelial cells. Proc NatlAcad Sci USA 1988;85:2800-2804

33. Azuma H, Ishikawa M, Sekizaki S: Endothelium-dependentinhibition of platelet aggregation. Br J Pharmacol 1986;88:441-445

34. Radomski MW, Palmer RMJ, Moncada S: Comparative phar-macology of endothelium-derived relaxing factor, nitric oxideand prostacyclin in platelets. Br J Pharmacol 1987;92:181-187

35. Rosen R, Dausch W, Beck E, Kaus W: Platelet inducedaggravation of acute ischemia in an isolated rabbit heartmodel. Cardiovasc Res 1987;21:293-298

36. Yee ES, Price DC, Aherne T, Ebert PA: Intracoronaryplatelet aggregation: Pattern of deposition after ischemia,cardioplegia, and reperfusion. J Surg Res 1986;40:499-503

37. Rosenbaum D, Levitsky S, Silverman N, Kohler J, Lipowski J,Le Breton G, Feinberg M: Cardioplegia does not preventreperfusion injury induced by intracoronary platelet deposi-tion. Circulation 1983;68(suppl II):II-102-II-106

38. Feinberg H, Rosenbaum DS, Levitsky S, Silverman NA,Kohler J, Le Breton G: Platelet deposition after surgicallyinduced myocardial ischemia: An etiologic factor for reperfu-sion injury. J Thorac Cardiovasc Surg 1982;84:815-822

KEY WORDS * ADP * ischemia * serotonin * thrombosis .

vasospasm by guest on D

ecember 31, 2017

http://circres.ahajournals.org/D

ownloaded from

P J Pearson, H V Schaff and P M Vanhouttereperfusion injury in canine coronary arteries.

Acute impairment of endothelium-dependent relaxations to aggregating platelets following

Print ISSN: 0009-7330. Online ISSN: 1524-4571 Copyright © 1990 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.67.2.3851990;67:385-393Circ Res. 

http://circres.ahajournals.org/content/67/2/385World Wide Web at:

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

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

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

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

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

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

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

by guest on Decem

ber 31, 2017http://circres.ahajournals.org/

Dow

nloaded from