Cellular electrophysiological basis for oxygen radical-induced arrhythmias. A patch-clamp study in guinea pig ventricular myocytes

E Cerbai, G Ambrosio, F Porciatti, M Chiariello, A Giotti and A Mugellipatch-clamp study in guinea pig ventricular myocytes

Cellular electrophysiological basis for oxygen radical-induced arrhythmias. A

ISSN: 1524-4539 Copyright © 1991 American Heart Association. All rights reserved. Print ISSN: 0009-7322. Online

72514Circulation is published by the American Heart Association. 7272 Greenville Avenue, Dallas, TX

1991, 84:1773-1782Circulation

http://circ.ahajournals.org/content/84/4/1773located on the World Wide Web at:

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

http://www.lww.com/reprintsReprints: Information about reprints can be found online at [email protected]: 410-528-8550. E-mail: Kluwer Health, 351 West Camden Street, Baltimore, MD 21202-2436. Phone: 410-528-4050. Permissions: Permissions & Rights Desk, Lippincott Williams & Wilkins, a division of Wolters http://circ.ahajournals.org//subscriptions/Subscriptions: Information about subscribing to Circulation is online at

by guest on July 10, 2011http://circ.ahajournals.org/Downloaded from

http://circ.ahajournals.org/

1773

Cellular Electrophysiological Basis for OxygenRadical-Induced Arrhythmias

A Patch-Clamp Study in Guinea Pig Ventricular Myocytes

Elisabetta Cerbai, PhD; Giuseppe Ambrosio, MD, PhD; Francesco Porciatti, MSc;

Massimo Chiariello, MD; Alberto Giotti, MD; and Alessandro Mugelli, MD

Background. Oxygen radicals have been implicated in the pathogenesis of reperfusionarrhythmias. However, the basic electrophysiological alterations accompanying the effects ofoxygen radicals on action potential (AP) are poorly understood.Methods and Results. We investigated the effects of oxygen radicals generated by dihydroxy-

fumarate (DHF, 5 mM) on AP parameters and on ionic currents in patch-clamped guinea pigventricular myocytes. DHF consistently caused a marked prolongation of AP duration, whichwas already significant after 60 seconds of exposure and continued to increase over time. Within5 minutes, the majority of cells developed early afterdepolarizations (EADs) or becameunexcitable. Both AP prolongation and occurrence of EADs were completely prevented in thepresence of the oxygen radical scavengers superoxide dismutase (SOD) and catalase (CAT).Prolongation of AP duration was accompanied by a marked decreased in time-dependentpotassium current (IK) and calcium current (LCa). The inward rectifier K current (IKI) wasunaffected, suggesting no widespread changes in membrane properties. 'K and lCa alterationswere also significantly reduced by SOD and CAT. In additional experiments, intracellularcalcium levels were kept constantly low by addition of 200 PM ethyleneglycol-bis(f-aminoethylether)-N,N,N',N'-tetra-acetic acid (EGTA) to the pipette solution. Under these conditions, theeffects of DHF on AP duration and the occurrence of EADs were largely prevented. However,EGTA did not prevent cells from becoming unexcitable, nor did it affect the decrease in both IKand IC. upon exposure to DHF.

Conclusions. Exposure to an exogenous source of oxygen radicals may induce majorelectrophysiological alterations in isolated myocytes, which might be related to changes inspecific ionic currents and in level of intracellular calcium. These alterations occur with a timecourse consistent with the rapid onset of(Circulation 1991;84:1773-1782)

I t has recently been shown that large quantities ofoxygen free radicals are generated in postisch-emic hearts at the time of reflow,1-5 and it has

been proposed that this phenomenon may be respon-sible for the occurrence of deleterious effects associ-ated with reperfusion, including complex ventriculararrhythmias.67 The hypothesis that oxygen radicalsare responsible for the occurrence of reperfusionarrhythmias is supported by the observation that the

From the Institute of Pharmacology, University of Ferrara(E.C., A.M.); the Centro di Ricerca Interuniversitario "Ipossie,"Firenze (E.C., A.G., A.M.); the Division of Cardiology, SecondSchool of Medicine, University of Napoli (G.A., M.C.); and theDepartment of Pharmacology, University of Firenze (F.P., A.G.).

Supported by grants from Ministero Pubblica Istruzione (40%).Address for correspondence: Alessandro Mugelli, MD, Institute

of Pharmacology, Via Fossato di Mortara 64/B, 44100 Ferrara,Italy.

Received March 15, 1991; revision accepted June 11, 1991.

ventricular arrhythmias in reperfused hearts.

incidence of ventricular tachycardia and/or ventricu-lar fibrillation upon reperfusion is drastically reducedby different "anti-free radical" interventions89 andthat complex ventricular arrhythmias can be inducedby exposing the heart to an exogenous source ofoxygen radicals.9"10The possible arrhythmogenic effects of oxygen radi-

cals have recently been investigated in vitro. In onestudy," exposure of guinea pig ventricular strips to anoxygen radical-generating system resulted in a reduc-tion of action potential (AP) amplitude, resting poten-tial, and upstroke velocity. Those effects developedslowly during incubation (20-30 minutes), and werenot associated with changes in action potential duration(APD). In contrast, a different sequence of events wasobserved in canine'2 and frog'3 myocytes. In thosestudies, the electrophysiological alterations were char-acterized by an increase in APD within minutes of



1774 Circulation Vol 84, No 4 October 1991

exposure to oxygen radicals, followed by developmentof both early and delayed afterdepolarizations (EADsand DADs), and then by the rapid occurrence of adecrease in APD and cell excitability. The differenteffects on APD suggest that multiple forms of injurymay be caused in cardiac cells, possibly depending onthe amount and type of free radical generated. On thewhole, these data suggest that oxygen radicals canaffect cardiac excitation, but they do not clearly de-scribe a specific arrhythmogenic effect. Furthermore,the basic electrophysiological alterations (such aschanges in membrane ionic currents) underlying possi-ble arrhythmogenic effects of oxygen radicals were notinvestigated in those studies.

Patch-clamp measurements in single myocytes rep-resent an accurate method for studying the effects ofan intervention on membrane currents. Using thistechnique, Shattock et al14 have recently describedchanges in the steady-state background current, ac-companied by activation of the oscillatory transientinward current (TI) during exposure to free radicalsgenerated by photoactivation of rose bengal. Thosefindings strongly suggest that free radicals may pro-mote arrhythmias through specific alterations inmembrane currents. However, no attempt was madein that study to prevent the development of theelectrophysiological changes by means of specificoxygen radical scavengers. In addition, while photo-activation of rose bengal generates predominantlysinglet oxygen,15 the oxygen metabolites formed dur-ing postischemic reperfusion are represented primar-ily by superoxide anions, hydrogen peroxide, andhydroxyl radicals.2-5The present study was designed to investigate the

electrophysiological effects of exogenously generatedsuperoxide anions and hydrogen peroxide. Changesin cardiac AP and membrane currents were recordedin single myocytes from guinea pig hearts exposed tooxygen radicals under control conditions, as well asduring administration of the specific scavenging en-zymes superoxide dismutase (SOD) and catalase(CAT).

Preliminary results have been presented in ab-stract form.16'17

MethodsExperimental Preparation

Isolated myocytes were obtained according to thetechnique described previously.18 Briefly, guinea pigs(200-300 g) were killed by a blow on the neck. Theirhearts were rapidly excised, mounted on a Langen-dorff apparatus, and perfused at 35°C for 3 minuteswith a nominally zero calcium solution of the follow-ing millimolar composition: NaCl 120, KCl 10,KH2P04 1.2, MgCl2 1.2, glucose 20, taurine 20, pyru-vate 5, pH 7.2 with HEPES/NaOH. The hearts werethen perfused with enzymatic solution (calcium-freesolution plus 30 mg/l pronase E, Serva) for 3 min-utes, and the ventricles were sliced and slowly stirredin enzymatic solution for about 1 hour. Isolated cells

that appeared in the supernatant were placed inTyrode's solution containing penicillin (50 IU/ml)and streptomycin (50 ,ug/ml) (Gibco), stored at roomtemperature, and used within the day.A drop of cells was placed in the experimental

chamber (0.2 ml) mounted on the plate of an in-verted microscope (Nikon TMS) connected to amonitor (Hitachi VM 122) by a camera (HitachiHM30).The whole-cell configuration of the patch-clamp

technique was used to measure currents and mem-brane potential.19 Patch pipettes (Corning Capillar-ies 7052, Garner Glass) with a resistance of 1-2 Mflwere pulled with a two-step puller (Hans Otchozki,Homburg), fire polished, and filled with a solution ofthe following millimolar composition: KCI 140,MgCl2 1, Na2ATP 3, HEPES 10, pH 7.2 with KOH. Insome experiments, the pipette solution contained 200,uM ethyleneglycol-bis(,8-aminoethyl ether)-N,N,N',N'-tetra-acetic acid (EGTA) to reduce theintracellular calcium activity (aCa,i), whose value wasestimated as 0.09 ,uM.20The electric signal was recorded by a patch ampli-

fier (Axopatch 1A, Axon), digitized (Labmaster TL100, Scientific Solutions), and displayed on the mon-itor of a personal computer. The cutoff frequency was2 kHz. Setting of experimental protocols, data acqui-sition, and elaboration were done automatically withpClamp software (Axon).

Protocols for current- and voltage-clamp measure-ments changed according to experimental necessity.APs were sampled at 1 kHz. Holding potential wasusually kept at -80 mV. Low-threshold calciumcurrent (ICa) was evoked by a 200-msec step to 0 mVand sampled at 5 kHz; a brief (10-msec) prestep to-40 mV was used to inactivate sodium current.Potential of half-maximal activation and inactivation(VH) and the slope factor k were obtained by fittingnormalized current values with the distribution:y.= 1/{1+exp[(V-VH)/k]}. Time-dependent K cur-rent (IK) was evoked by a 500-800-msec step to 40mV and sampled at 0.5 kHz.

Oxygen Free Radical-Generating SolutionsThe cells were superfused at a rate of 2 ml/min

using a two-flow line system controlled by electronicvalves to allow rapid change from control to experi-mental solutions. Control solution was a modifiedTyrode's solution containing (mM): NaCl 137, KCl5.4, CaCl2 5.4, MgCl2 1.2, glucose 10, HEPES 5, pHadjusted to 7.35 with NaOH, at a temperature of35°C. The oxygen radical-generating solution wasprepared just before use and consisted of 5 mMdihydroxyfumarate (DHF, Sigma) added to the Ty-rode's solution and adjusted to pH 7.35 with NaOH.The solutions were bubbled with 100% 02 (P02:700-750 mm Hg in the bottle and 400-500 mm Hg in thechamber). DHF is a well-established system for gen-erating oxygen radicals.21,22 Under our experimentalconditions of pH and oxygen tension, DHF under-goes redox cycling, with formation of superoxide



Cerbai et al Electrophysiological Effects of Oxygen Radicals

radicals.21,22 Hydrogen peroxide can also be formed,either directly or through spontaneous dismutationof superoxide radicals.21,22

In the experiments designed to determine theeffect of adding scavenging enzymes, a solution con-taining SOD (Cu/Zn form, from bovine erythrocytes,specific activity 3,000 IU/mg of protein, from Sigma)and CAT (from bovine liver, specific activity 65,000IU/mg of protein, from Boehringer Mannheim) wasinjected into the line of superfusion through a needleconnected to a peristaltic pump (Gilson, Minipulse2) at constant rate (150 ,l/min). SOD and CAT weredissolved in Tyrode's solution immediately beforeuse to achieve a final concentration in the chamber of300 IU/ml for each enzyme. To check whether per-fusion with SOD and CAT could affect the electricalactivity of the cells in the absence of exogenousoxygen radicals, superfusion with the scavenging en-zymes was started at least 2 minutes before switchingto the DHF solution.Data are expressed as mean+SEM. Effects of

DHF and DHF plus SOD and CAT were comparedby Student's t test for grouped data. A value ofp<0.05 was considered significant.

ResultsEffect of Oxygen Radicals on Action Potential

Superfusion with DHF resulted in a consistent andtime-dependent increase in APD. A typical example ofthis effect is shown in Figure 1A. In this cell, after 2minutes of superfusion with DHF, APD is clearlyprolonged and, after 5 minutes, early afterdepolariza-tions (EADs) developed, superimposed on the plateauphase. EADs delayed cell repolarization, which oc-curred after more than 1 second. APD consistentlyincreased in all preparations tested, whereas EADsdeveloped in seven out of 14 cells. In the remainingcells, unexcitability (n=3) or depolarization to -40/-30 mV (n=3) was observed. All these effects devel-oped within 10 minutes from the beginning of superfu-sion with DHF.The prolongation of APD occurred immediately

after switching to the DHF solution, increased withtime, and tended to a steady state within 5 minutes(Figure 2). The effects of exposure to oxygen radicalson AP profile and ionic currents were not reversedby washing out DHF, consistent with previousobservations.'2

Effects of Oxygen Radical ScavengersTo verify whether the prolongation of APD was

caused by the generation of free radicals due toauto-oxidation ofDHF in the presence of 02, specificscavenging enzymes were used.

In the presence of SOD (300 units/ml) and CAT(300 units/ml), superfusion with DHF failed to pro-long APD or to induce the appearance of EADs(Figure 1B). Development of unexcitability or depo-larization was also prevented. The results are sum-marized in Figure 2. It is apparent that in the

50 -

0-

mY

-50

-100

50

my

-50

-100

1 s

SOD&CAT +DHF 2 min 5 min

B0 - _ __ _

0.0.5 s

FIGURE 1. Plot showing effects of oxygen radicals generatedby dihydroxyfumarate (DHF) on action potential (AP) profileof isolated myocytes. Panel A shows a series of typical APsrecorded before (control) and after 2 and 5 minutes ofsuperfusion with 5 mM DHF. Note the appearance of earlyafterdepolarizations after 5 minutes of DHF superfusion.Panel B shows a different cell, in which the effects of DHFwere prevented by the addition of the scavenger enzymessuperoxide dismutase (SOD) and catalase (CAT). Time scaleis different between panels A and B.

presence of the scavengers, APD did not change withrespect to the control values. Differences in APDbetween cells receiving DHF alone and those treatedwith SOD plus CAT were already evident after 1minute of incubation (p<0.05, Figure 2).

Effect of Oxygen Radicals on Ionic CurrentsTo investigate the membrane ionic modifications

underlying the described effect on the AP profile,voltage-clamp experiments were performed.

Figure 3 shows the activation of the IK in responseto a long pulse to +40 mV. By return to -40 mV, IKtypically inactivates within a few hundred millisec-onds. Two minutes of superfusion with the oxygenradical-generating system (DHF, 5 mM) reduced IKboth during the depolarizing step and upon returning

1775




6 1 2 3 4 5min

to -40 mV (Figure 3A). Figure 4 summarizes theresults observed in five cells; the amplitude of acti-vated IK, measured at -40 mV as the differencebetween the peak outward current and the steady-state current (tail current), is here reported as afunction of the depolarizing step potential (I-Vcurve). After 2 minutes of superfusion with 5 mM

1000 -

500

0-

-_s0

-1000

A- In the presence of DHF

B- In the presence of DHF + SOD + CAT

40- 40 mV

mV-80

-14 1

DHF 5 mMA

DIF+SOD+CAT

FIGURE 2. Scatterplot of effect of di-hydroxyfumarate (DHF) on action po-tential duration (APD) in the absenceand presence of scavengers. APD wasmeasured at 90% repolarization(APD90). Arrow marks the tite at whichsuperfsion with 5mMDHF was started.APD could not be measured in cells thatdepolanzed or became unexcitable, andwas not measured in those developingearly afterdepolarization. Consequently,the number of cels that could be aver-aged in the absence of scavengers de-creased with time (as indicated in brack-ets). (*p<0.05, ** p<0.01 vs. DHFalone). SOD, superoxide dismutase;CAT, catalase.

DHF, IK is reduced at all voltages; at +50 mV, it was363±43 pA in the control and 199±49 pA in thepresence of DHF (p<O.OS). The inset of Figure 4shows for comparison the effect of DHF on thetime-independent background current, measured atthe end of 200 msec conditioning steps ranging from-100 to +50 mV (n=5). The current is only slightly

FIGURE 3. Plot showing effect of dihydroxyfumarate(DHF) on time-dependent potassium current (IK) as

1200 measured by patch-clamp technique in isolated myo-cytes. Oxygen radicals reduced IK in the absence but notin the presence ofscavengers. C, control tracing; DHF 2min, traces recorded after 2 minutes ofsuperfsion with5 mMDHF. The voltage-clamp protocol is reported inpanel B.

400-

1 300-

p 200-

100

(8

(14(14) (14

jT-A (8)

A A ** *

r

U 1 1

-

-

400 860 me 1



Cerbai et al Electrophysiological Effects of Oxygen Radicals

-20 0 20 40mV -100 -50 0 50

FIGURE 4. Plot showing current-voltage (I-V) relation for time-dependent potassium current (IK) under control conditions andafter 2 minutes ofsuperfusion with 5 mM dihydroxyfumarate (DHF). The holding potential (HP) was -80 mVV; the step potential(SP) varied between -30 and +50 mV and lasted for 500 msec (n=5). The inset shows the effect of oxygen radicals on

time-independent background current (HP= -80 mV, SP= -100 to +50 mVfor 200 msec) (n=4).

affected at potential more negative than -80 mV,and is completely unaffected between -80 and 0 mV.The slope of the curve is depressed at more positivepotentials, suggesting a rather selective blockade ofoutward currents by oxygen radicals.

Figure 3B shows the effect of 2 minutes' exposure

to DHF on IK recorded from another cell, in thepresence of SOD plus CAT. Under these conditions,the effects of DHF on lK were completely preventedby the addition of scavengers in four out of eightcells, and significantly reduced in the remaining cells(see below).

Ic, was also rapidly reduced by the superfusionwith 5 mM DHF (Figure SA). After 2 minutes ofexposure to the oxygen radical-generating system,the peak Ic, is decreased to about 70% of its initialvalue, without any appreciable modification of itskinetics. It is also evident that the steady-state cur-

rent at -40 mV, as well as the late current at 0 mV,is virtually unchanged, in agreement with the resultsreported in the inset of Figure 4. Again, the presenceof scavengers prevented ICa reduction by DHF (Fig-ure SB) almost completely.The effects of 2 minutes of incubation with DHF

on the I-V relation for ICa are shown in the graphs ofFigure 6. Oxygen radical generation induced a

marked reduction in the amplitude of the I-V rela-tion, from 1,366+230 pA in control to 892±+84 pA inDHF at +10 mV (p<0.05, n=4), while the overallshape was not modified. In fact, steady-state activa-tion and inactivation curves (insets of Figure 6) werenot significantly modified by DHF. Since both half-maximal activation and inactivation (VK) and slopefactor (k) were not modified (see legend of Figure 6),the decrease of peak Ica by DHF might be attributedto a reduction in the amount of available calcium

channels, rather than to a modification of the currentkinetics.The time course of the effects ofDHF on IK and ICa

in the absence and presence of scavengers are sum-

marized in Figure 7. The amplitude of the current(normalized with respect to the control value at timezero) is here expressed as a function of superfusiontime. The amplitude of IK (Figure 7A) was reduced to69±6% of baseline value (n = 10) as early as 1 minuteafter beginning superfusion with DHF and reached a

steady state by 5 minutes (52+4% of the zero-timevalue, n=4). Similarly, the amplitude of ICa (Figure7B) was decreased to 73+4% at 1 minute (n= 14) andto 62±+3% at 5 minutes (n=5) with respect to thepre-DHF value.

Pretreatment of the cells with the scavenging en-

zymes largely prevented the effect of DHF at 1minute, 'K and Ica amplitude being 103+4% (n=8)and 91±5% (n=8) of the control, respectively(p<0.02). The current amplitudes gradually de-creased as the superfusion of the cells with theoxygen radical-generating system progressed in time(Figure 7). However, this phenomenon was lesspronounced than in the absence of scavengers.

Experiments With Intracellular EGTAThe decrease of 'Ca might suggest that intracellular

calcium activity is reduced by oxygen radicals. How-ever, intracellular calcium is regulated by severalother mechanisms (either intracellular or sarcolem-mal), which might be altered by oxygen radicals.When intracellular calcium was kept constantly

low by addition of 200 I£M EGTA to the pipettemedium (aca=0.09 ,iM),20 the effect of DHF on APDwas greatly reduced. The APD in the control solutionwas 190±13 msec, comparable to the value in theabsence of EGTA. In Figure 8, typical APs recorded

400-

300

A 200-

100-

v'4-40n 1

1777




A- In the presence of DHF

ai

°-~~~~~~~-I-,-

-500 50 100 15

B-In the presence of DHF + SOD + CAT500

o0-

.500-

-1500-

-2000- ..50

FIGURE 5. Plot showing effect of oxygen radicals onM calcium current (Ica) in the absence and presence of

scavengers. Voltage protocol is shown in panel B. C,control tracing; DHF 2 min, after 2 minutes of super-fusion with 5 mM dihydroxyfumarate (DHF). SOD,superoxide dismutase; CAT, catalase.

100 me 150

in control and after 5 and 6 minutes of superfusionwith DHF were superimposed: at 5 minutes, theAPD is only slightly prolonged (206+±9 msec) withrespect to the previously described experiments inwhich APD almost doubled (Figures 1 and 2). Inter-estingly, intracellular EGTA did not prevent the cellfrom becoming unexcitable upon further exposure to

00R W4 1.0_

-500-

~~~~~~~~~~~~~~0-

0-1

DHF (Figure 8). Similar results were obtained in fiveout of five cells. Furthermore, in no instance did weobserved the appearance of EADs. The results aresummarized in Figure 9, which also shows the ampli-tude of IK and Ica recorded in the voltage-clampmode under these experimental conditions (n=5). Itis apparent that the presence of 200 ,M EGTA in

mAu

FIGURE 6. Plot showing effects ofoxygen radicals on current-voltage

(I-V) relation for calcium current (IcJ)(left panel). On the right, the steady-state activation (upper right panel)and inactivation curves (bottom rightpanel) are reported. Continuous lineswere obtained as described under

4OmV 'Methods." Data fitting gave, for theactivation curve [in control and in thepresence of 5 mM dihydroxyfumarate(DHF), respectively]: potential ofhalf-maximal activation and deactivation(VH, -7.0 and -7.8 mV; slope factor(k), -5.8 and -6.9 mVV; for the inac-tivation curve: VH, -21.9 and -26.0mV; k; 5.2 and 5.8 mV(differences notstatistically significant).

-80 -40 0 mv

c

_ 5M6

i1

200 me

0

rnV-40

DHF 2 min



Cerbai et al Electrophysiological Effects of Oxygen Radicals 1779

miin

FIGURE 7. Plot showing time course of the effect of oxygenradicals on time-dependent potassium current (IK and cal-cium current (ICa). Changes over time of the current in theabsence of dihydroxyfimarate (DHF) (control) are alsoshown for comparison. (* p<0.05, ** p<0.02 vs. DHFalone.) SOD, superoxide dismutase; CAT, catalase.

the pipette solution, while almost completely pre-venting APD prolongation, did not modify the effectsof the oxygen radical-generating system on IK andICa. After 5 minutes' exposure to DHF, IK wasreduced to 54±4% and ICa to 64±6% with respect tothe control value. These values were not statisticallydifferent from those measured in the absence ofEGTA (Figure 7).

mv

40

-40 C DHF 5 min

DHF 6 min

.80

DiscussionIn the present study, exposure of isolated ventric-

ular myocytes from guinea pig hearts to DHF-gener-ated oxygen radicals induced a progressive prolonga-tion of APD, which was already evident after 1minute and reached a maximum within 5 minutes ofincubation. Development of EADs, membrane depo-larization, and unexcitability were also observed inthe majority of cells. In contrast, when the specificoxygen radical scavengers SOD and CAT were pres-ent in the superfusion medium, the effects of DHFwere largely prevented. The alterations observed inAP were also accompanied by specific changes inmembrane currents, as determined by voltage-clampmeasurements. In particular, superfusion with DHFresulted in a decrease in time-dependent 'K and ICa,Again, these electrophysiological changes were sig-nificantly reduced by treatment with SOD and CAT.Taken together, these results demonstrate that exog-enously generated superoxide radicals and hydrogenperoxide can rapidly induce major electrophysiolog-ical alterations in ventricular myocytes.Our observation of a prolongation in APD dis-

agrees with previous reports by Pallandi et al.11However, our experimental model differs from thatof Pallandi et al in that they employed a complexsystem (i.e., superfused guinea pig ventricular strips),in which access of short-lived oxygen radicals tomyocytes might have been hampered. The findingthat in our study the electrophysiological alterationswere already evident after 1 minute of incubation,whereas in Pallandi's study they took 20-30 minutesto develop, lends further support to this speculation.On the other hand, the results of the present studyare in agreement with those of Barrington et al.12 Inan experimental model similar to ours (i.e., isolatedcanine ventricular myocytes), they observed a pro-gressive increase in APD and development of after-depolarizations within a few minutes following incu-bation with either DHF or xanthine/xanthine-oxidasesystem, which could be prevented by SOD and CAT.

FIGURE 8. Plot showing effect of dihydroxyfumarate(DHF) on action potential (AP) in the presence of intra-cellular ethyleneglycol-bis((-aminoethyl ether)N,N,N',N'-tetra-acetic acid (EGTA): the figure shows superimposedAPs recorded from a ventrcular myocyte before (C) andafter 5 and 6 minutes of exposure to the oxygen radical-generating system (5 mM DHF). EGTA (200 pM) waspresent in the pipette medium. Time scale is different fromFigure 1.

100 ms




FIGURE 9. Time course ofthe effect ofoxygen radicalson action duration potential measured at 90% repolar-ization (APD90), time-dependent potassium current(IK), and calcium current (ICa) in the presence ofintracellular ethyleneglycol-bis (/-aminoethylether)N,N,N',N'-tetra-acetic acid (EGTA) (200 pM).Values are expressed as percentage ofpre-dihydroxyfu-marate (DHF) value (zero time). (See also Figures 2and 7for comparison.)

In our study, we used a well-established system,which generates mostly superoxide radicals andsmaller amounts of hydrogen peroxide and hydroxylradicals.21,22 Significant electrophysiological alter-ations were observed within a few minutes afterexposure and were markedly attenuated by the use ofthe specific scavenger enzymes SOD and CAT. Thus,this study provides the first documentation that ex-posure to a system producing biologically relevantoxidant species can induce alterations in myocytemembrane currents, with a time course similar to thatobserved for the occurrence of reperfusion arrhyth-mias in postischemic hearts.Our results demonstrate that the electrophysiolog-

ical changes in AP profile (APD prolongation, appear-ance of EADs) are accompanied by modifications inpotassium and calcium currents. These data are inagreement with those reported by Shattock et al,14who observed a prolongation ofAPD and appearanceof membrane oscillations during the plateau of theprolonged AP in rabbit myocytes exposed to photoac-tivated rose bengal. However, they also observed theappearance of membrane potential oscillations at theresting potential level, that is, DADs. Both types ofoscillations were apparently associated with the ap-pearance of the TI. In our experimental conditions, wehave never recorded either DADs (as reported byBarrington et al'2 in canine myocytes exposed toDHF) or TI, which is known to underlie DADs. Theexplanation for such a discrepancy is not obvious. It isknown that myocytes isolated from the hearts ofdifferent species may have a different propensity todevelop TI and DADs, probably as a consequence ofa different susceptibility to spontaneous calcium re-lease from intracellular stores.23 Nor can it be ex-cluded that different oxidant species may exert differ-ent effects on the sarcolemma and/or intracellularly.As already pointed out, photoactivation of rose bengalyields singlet oxygen.'5'24 Preliminary results wouldsuggest that oxygen singlet but not other radicalspecies can inhibit calcium uptake and Ca,ATPaseactivity of cardiac sarcoplasmic reticulum.25

In our study, the reduction in both IK and Ica canbe attributed, for the most part, to the production of

oxygen radicals, since in the first 2 minutes of expo-sure to DHF the effects of oxygen radicals on ioniccurrents were significantly prevented by the scaven-gers. This finding appears of particular relevance inthe setting of reperfusion arrhythmias that occurimmediately after reflow.26 Furthermore, SOD andCAT prevented APD prolongation and the develop-ment of EADs: EADs are generally considered as anarrhythmogenic mechanism27; APD prolongationmay result in arrhythmias under certain conditions.Inhomogeneity in repolarization (and in APD) is awell-accepted arrhythmogenic stimulus, since itcauses a dispersion in the refractory period andenhances the likelihood for reentry.28 Moreover, theappearance of EADs has been associated with thedevelopment of reperfusion arrhythmias in vivo29 andin vitro.30The ionic mechanisms underlying EADs are not

completely clarified, although it has been establishedthat EADs can be induced in vivo and in vitro byvarious pharmacological or experimental interven-tions that cause AP prolongation.3'-34 The durationof AP is determined by the balance of several plateaucurrents35; among the outward currents, a key role isplayed by 1K. In the present study, the time course ofIK reduction by DHF closely paralleled the effects onAPD. Although we cannot directly prove a cause-and-effect relation between the two events, this find-ing is consistent with the notion that a decrease in theK current responsible for membrane repolarization(at potential values around the plateau phase) mayfavor EAD development.36 A decrease in membraneconductance during the plateau phase due to theblock of IK would make the cell more sensitive toeven small depolarizing currents such as the Na/Cacurrent, which is known to contribute to the normalAP of guinea pig ventricular myocytes.37

It has been proposed that the electrophysiologicaleffects of oxygen radicals are mediated mainly by anincrease in the level of intracellular calcium,38-4'possibly through a direct or indirect modification ofthe Na/Ca exchanger activity.'4 A role of increasedintracellular calcium in the action of oxygen radicalsis also suggested by our observation that APD pro-

nmil



Cerbai et al Electrophysiological Effects of Oxygen Radicals 1781

longation and the development of EADs were pre-vented by the intracellular addition of the calciumchelator EGTA. EADs have been associated withcalcium transients in isolated myocytes,42 and theirdevelopment was similarly prevented by EGTA in ratventricular myocytes exposed to an oxygen radical-generating system.38 Since Ica is reduced as an effectof oxygen radicals, its role in the prolongation ofAPD or development of EAD appears to be intrigu-ing. The contribution of ICa to the duration of theplateau phase is still discussed; recent evidence sug-gests that other inward currents (such as Na/Caexchanger current) may influence APD more direct-ly.37 Indeed, APD of atrial cells has been demon-strated to correlate with potassium-current ampli-tude better than calcium current amplitude.43 Therelative contribution of membrane currents to theeffects of oxygen radicals on AP is difficult to assess,although nonspecific membrane damage appears un-likely, since the lKl was only slightly affected by oursystem.

In the present study, we used both a superoxideradical scavenger (SOD) and a hydrogen peroxidescavenger (CAT) to prevent the effects of DHF.Thus, we have no direct evidence of the oxygenradical species involved. However, although bothsuperoxide radicals and hydrogen peroxide mightindependently exert toxic effects, it is widely acceptedthat their toxicity in biological systems is largely dueto the formation of the more reactive hydroxyl radi-cal, when these two species react in the presence ofsmall amounts of metal ions.44-48 In this respect,Barrington et a112 have shown in a similar model thatscavenging with either SOD or CAT was equallyeffective in reducing DHF-induced electrophysiolog-ical changes. This finding indicates that the oxidantspecies involved in this phenomenon was neither thesuperoxide nor hydrogen peroxide, but the productof their reaction, the hydroxyl radical. Similar resultswere obtained by other investigators with respect tothe effects of oxygen radicals on other indexes ofcardiac damage.49-52 In the present study we usedonly one system to generate oxygen radicals. There-fore, in principle it cannot be ruled out that DHF perse could cause direct electrophysiological alterations,independent of oxygen radical generation. However,this possibility seems unlikely, since the effects ofDHF were greatly reduced by SOD and CAT. Fur-thermore, in the study by Barrington et al,12 changesin APD similar to those observed in the present studywere reproduced by treatment with DHF, as well aswith the unrelated xanthine/xanthine-oxidase enzy-matic system, confirming that they were mediated byoxygen radicals, irrespective of the system employed.On the other hand, the use of different sources of

oxygen radicals should be investigated in futurestudies, as it may allow a better definition of severalimportant issues, such as the chemical nature of theoxidant species involved or the possible site (i.e.,extracellular versus intracellular) of the modifica-tions induced by oxygen radicals.

In conclusion, our data show that exposure to exog-enously generated oxygen radicals can induce mem-brane electrophysiological alterations in isolated myo-cytes. These effects develop rapidly, and areaccompanied by specific changes in ionic currents,providing strong support to the hypothesis that oxygenradicals generated at reflow are responsible for theoccurrence of reperfusion arrhythmias in intact hearts.

References1. Zweier JL, Flaherty JT, Weisfeldt ML: Direct measurement of

free radical generation following reperfusion of ischemicmyocardium. Proc Natl Acad Sci U S A 1987;84:1404-1407

2. Kramer JH, Arroyo CM, Dickens BF, Weglicki WB: Spin-trapping evidence that graded myocardial ischemia alterspost-ischemic superoxide production. Free Radic Biol Med1987;3:153-159

3. Bolli R, Patel BS, Jeroudi MO, Lai EK, McCay PB: Demon-stration of free radical generation in "stunned" myocardiumof intact dogs with the use of the spin trap alpha-phenylN-tert-butyl nitrone. J Clin Invest 1988;82:476-485

4. Zweier JL: Measurement of superoxide-derived free radicalsin the reperfused heart: Evidence for a free radical mechanismof reperfusion injury. J Biol Chem 1988;263:1353-1357

5. Zweier JL, Kuppusamy P, Williams R, Rayburn BK, Smith D,Weisfeldt ML, Flaherty JT: Measurement and characteriza-tion of post-ischemic free radical generation in the isolatedperfused heart. J Biol Chem 1989;264:18890-18895

6. Hess ML, Manson NH: Molecular oxygen: Friend and foe.The role of the oxygen free radical system in the calciumparadox, the oxygen paradox and ischemia/reperfusion injury.J Mol Cell Cardiol 1984;16:969-985

7. Becker LC, Ambrosio G: Myocardial consequences or reper-fusion. Prog Cardiovasc Dis 1987;30:23-44

8. Woodward B, Zakaria MNM: Effect of some free radicalscavengers on reperfusion induced arrhythmias in the isolatedrat heart. J Mol Cell Cardiol 1985;17:485-493

9. Bemier M, Hearse DJ, Manning AS: Reperfusion-inducedarrhythmias and oxygen-derived free radicals: Studies with"anti-free radical" interventions and a free radical-generatingsystem in the isolated perfused rat heart. Circ Res 1986;58:331-340

10. Hearse DJ, Kusama Y, Bernier M: Rapid electrophysiologicalchanges leading to arrhythmias in the aerobic heart: Photo-sensitization studies with rose bengal-derived reactive oxygenintermediates. Circ Res 1989;65:146-153

11. Pallandi RT, Perry MA, Campbell TJ: Proarrhythmic effectsof an oxygen-derived free radical generating system on actionpotentials recorded from guinea-pig ventricular myocardium:A possible cause of reperfusion-induced arrhythmias. Circ Res1987;61:50-54

12. Barrington PL, Meyer CF Jr, Weglicki WB: Abnormal electri-cal activity induced by free radical generating systems inisolated cardiocytes. J Mol Cell Cardiol 1988;20:1163-1178

13. Tarr M, Valenzeno DP: Modification of cardiac action poten-tial by photosensitizer-generated reactive oxygen. J Mol CellCardiol 1989;21:539-544

14. Shattock NJ, Hearse DJ, Matsuura H: Ionic currents under-lying oxidant stress-induced arrhythmias, in Vereecke J, VanBogaert PP, Verdonck F (eds): Ionic Currents and Ischemia.Leuven, University Press, 1990, pp 165-189

15. Kusama Y, Bemier M, Hearse DJ: Singlet oxygen-inducedarrhythmias: Dose and light-response studies for photoactiva-tion of rose bengal in the rat heart. Circulation 1989;80:1432-1448

16. Cerbai E, Visentin S, Ambrosio G, Mugelli A: Electrophysio-logical alterations caused by free radicals in isolated myocytes(abstract). J Mol Cell Cardiol 1990;22(suppl III):S79

17. Mugelli A, Ambrosio G, Cerbai E, Masini I, Visentin S:Electrophysiological effects caused by a free radical-gener-ating system in guinea-pig isolated myocytes, in Vereecke J,




van Bogaert PP, Verdonck F (eds): Ionic Currents and Isch-emia. Leuven, University Press, 1990, pp 288-290

18. Cerbai E, Kloeckner U, Isenberg G: Ca-antagonistic effects ofadenosine in guinea pig atrial cells. Am J Physiol 1988;255:H872-H878

19. Hamill OP, Marty A, Neher E, Sakmann B, Sigworth FJ:Improved patch clamp techniques for high resolution currentrecording from cells and cell-free membrane patches. PflugersArch 1981;391:85-100

20. Fabiato A, Fabiato F: Contractions induced by a calcium-triggered release of calcium from sarcoplasmic reticulum of a

single skinned cardiac cell. JPhysiol (Lond) 1975;249:469-49521. Tong Mak I, Misra HP, Weglicki WB: Temporal relationship

of free radical induced lipid peroxidation and loss of latentenzyme activity in highly enriched hepatic lysosomes. J BiolChem 1983;258:13733-13737

22. Kramer JH, Tong Mak I, Weglicki WB: Differential sensitivityof canine cardiac sarcolemmal and microsomal enzymes toinhibition by free radical induced lipid peroxidation. Circ Res1984;55:120-124

23. Fabiato A, Fabiato F: Calcium induced release of calciumfrom the sarcoplasmic reticulum of skinned cells from adulthuman, dog, cat, rabbit, rat and frog hearts and from fetal andnewborn rat ventricles. Ann N YAcad Sci 1978;307:491-522

24. Lee PCC, Rodgers MAJ: Laser flash photokinetics studies ofrose bengal sensitized photodynamic interactions of nucleo-tides and DNA. Photochem Photobiol 1987;45:79-86

25. Kukreja RC, Kearns AA, Zweier JL, Kuppusamy P, Hess ML:Singlet oxygen interaction with Ca2'-ATPase of cardiac sarco-plasmic reticulum (abstract). Circulation 1990;82(suppl III):III-269

26. Manning AS, Hearse J: Reperfusion-induced arrhythmias:Mechanism and prevention. J Mol Cell Cardiol 1984;16:497-518

27. Hoffman BF, Rosen MR: Cellular mechanism for cardiacarrhythmias. Circ Res 1981;49:1-15

28. Janse MJ, Wit AL: Electrophysiological mechanisms of ven-

tricular arrhythmias resulting from myocardial ischemia andinfarction. Physiol Rev 1989;69:1049-1169

29. Priori SG, Mantica M, Napolitano C, Schwartz PJ: Earlyafterdepolarizations induced in vivo by reperfusion of ischemicmyocardium: A possible mechanism for reperfusion arrhyth-mias. Circulation 1990;81:1911-1920

30. Hamra M, Rosen MR: a-Adrenergic receptor stimulationduring simulated ischemia and reperfusion in canine cardiacPurkinje fibers. Circulation 1988;78:1459-1502

31. El-Sherif N, Bekheit S-S, Henkin R: Quinidine-induced longQTU interval and torsade de pointes: Role of bradycardia-dependent early afterdepolarizations. JAm Coll Cardiol 1989;14:252-257

32. Davidenko JM, Cohen L, Goodrow R, Antzelevitch C: Quini-dine-induced action potential prolongation, early afterdepo-larizations, and triggered activity in canine Purkinje fibers:Effects of stimulation rate, potassium and magnesium. Circu-lation 1989;79:674-686

33. Levine JH, Spear JF, Guarnieri T, Weisfeldt ML, De LangenCDJ, Becker LC, Moore EN: Cesium chloride-induced longQT syndrome: Demonstration of afterdepolarizations andtriggered activity in vivo. Circulation 1985;72:1092-1104

34. Mugelli A, Amerini A, Piazzesi G, Giotti A: Barium-inducedspontaneous activity in sheep Purkinje fibers. J Mol CellCardiol 1983;15:697-712

35. Vassalle M: Electrogenesis of the plateau and pacemakerpotential. Annu Rev Physiol 1979;41:425-440

36. Wit AL, Rosen MR: Afterdepolarizations and triggered activ-ity, in Fozzard HA, Haber E, Jennings RB, Katz AM, MorganNE (eds): The Heart and Cardiovascular System. New York,Raven Press Publishers, 1986, pp 1449-1490

37. Egan TM, Noble D, Noble SJ, Powell T, Spindler AJ, TwistVW: Sodium-calcium exchange during the action potential inguinea pig ventricular cells. JPhysiol (Lond) 1989;411:639- 661

38. Ninomiya M, Silverman HS, duBell WH, Zweier JL, SternMD, Lakatta EG: Cytosolic Ca2' modulates early afterdepo-larizations induced by oxygen radicals (abstract). Circulation1990;82(suppl III):III-452

39. Burton KP, Morris AC, Massey KD, Busta LM, Hagler HK:Free radicals alter ionic calcium levels and membrane phos-pholipids in cultured rat ventricular myocytes. J Mol CellCardiol 1990;22:1035-1047

40. Ninomiya M, Silverman HS, duBell WH, Stern MD, LakattaEG: Voltage clamp prevents manifestation of calcium loadingby oxygen radicals in rat ventricular myocytes (abstract). J MolCell Cardiol 1990;22(suppl I):S38

41. Josephson RA, Silverman HS, Lakatta EG, Stern MD, ZweierJL: Study of the mechanisms of hydrogen peroxide andhydroxyl free radical induced cellular injury and calciumoverload in cardiac myocytes. JBiol Chem 1991;266:2359-2361

42. Szabo B, Sweidan R, Fugate R, Banyasz T, Lazzara R: Ca2transients during generation of early afterdepolarizations inCs+ treated canine ventricular myocytes (abstract). Circulation1990;82(suppl III):III-746

43. Visentin S, Wu S-N, Belardinelli L: Adenosine-inducedchanges in atrial action potential: Contribution of Ca and Kcurrents. Am J Physiol 1990;258:H1070-H1078

44. Aust SD, Morehouse LA, Thomas CE: Role of metals inoxygen radical reactions. Free Radic Biol Med 1985;1:3-25

45. Halliwell B: Superoxide-dependent formation of hydroxylradicals in the presence of iron salts is a feasible source ofhydroxyl radicals in vivo. Biochem J 1982;205:461-462

46. McCord JM, Day ED Jr: Superoxide-dependent production ofhydroxyl radicals catalyzed by the iron-EDTA complex. FEBSLett 1978;86:139-142

47. Starke PE, Farber JL: Ferric iron and superoxide ions arerequired for the killing of cultured hepatocytes by hydrogenperoxide: Evidence for the participation of hydroxyl radicalsformed by an iron-catalyzed Haber-Weiss reaction. J BiolChem 1985;260:10099-10104

48. Winterbourn CC, Sutton HC: Iron and xanthine oxidasecatalyze formation of an oxidant species distinguishable from'OH: Comparison with the Haber-Weiss reaction. Arch Bio-chem Biophys 1986;244:27-34

49. Ambrosio G, Zweier JL, Jacobus WE, Weisfeldt ML, FlahertyJT: Improvement of post-ischemic myocardial function andmetabolism by administration of deferoxamine at the time ofreflow: The role of iron in the pathogenesis of reperfusioninjury. Circulation 1987;77:906-915

50. Blaustein AS, Schine L, Brooks WW, Fanburg BL, Bing OHL:Influence of exogenously generated oxidant species on myo-cardial function. Am J Physiol 1986;250:H595-H599

51. Burton KP, McCord JM, Ghai G: Myocardial alterations dueto free-radical generation. Am J Physiol 1984;246:H776-H783

52. Miki S, Ashraf M, Salka S, Sperelakis N: Myocardial dysfunc-tion and ultrastructural alterations mediated by oxygen metab-olites. J Mol Cell Cardiol 1988;20:1009-1024

KEY WORDS * oxygen radicals * dihydroxyfumarate * earlyafterdepolarizations * reperfusion arrhythmias * ionic currents



Documents

Cellular electrophysiological basis for oxygen radical-induced arrhythmias. A patch-clamp study in guinea pig ventricular myocytes