442

Resting Membrane Potential, Extracellular PotassiumActivity, and Intracellular Sodium Activity during AcuteGlobal Ischemia in Isolated Perfused Guinea Pig Hearts

Andre G. KleberFrom the Department of Physiology, University of Berne, Berne, Switzerland

SUMMARY. Transmembrane potentials, extracellular potassium activity, and intracellular sodiumactivity were determined during acute global ischemia in Langendorff perfused guinea pig ventriclesby microelectrode techniques. Resting membrane potential decreased with a sigmoidal time coursefrom —82 mV to —49.5 ± 2.7 mV (SD, n = 6) and extracellular potassium activity increased from 4to 5 mM to 14.7 ± 1.3 ITIM (n = 8) during 15 minutes of ischemia. The estimated potassiumequilibrium potential was 7 mV more negative than resting membrane potential prior to occlusion,but approached resting membrane potential during ischemia. An increase in extracellular potassiumaccumulation occurred when heart rate was increased abruptly from 60 to 170 beats/min. Afterrapid stimulation, a transient decrease of extracellular potassium activity occurred which wasabolished in the presence of 10~6 M strophanthidin. If the preparations were paced before and afteraortic occlusion at a constant rate, potassium accumulation was independent of heart rate within arange of 50-170 beats/min. Intracellular sodium activity was 8.8 ± 2.8 mM (n = 8) prior to occlusionand decreased slightly to values between 4.7 and 7.6 mM after 10-15 minutes of ischemia. The resultssuggest that relative potassium permeability largely predominates over relative sodium permeabilityduring the decrease of resting membrane potential after interruption of aortic flow. Furthermore,active sodium-potassium exchange compensates for the rate-dependent fraction of potassium effluxand maintains a low intracellular sodium activity. For reasons of electroneutrality, the potassiumefflux underlying extracellular potassium accumulation must be balanced by an equivalent chargemovement which is not carried by sodium. The most probable hypothesis regarding the chargecarriers is that net potassium efflux occurs secondary to efflux of phosphate and lactate generatedduring ischemia. (Circ Res 52: 442 -450, 1983)

CHANGES of cardiac electrical activity occur withina few minutes of coronary occlusion: both restingpotential and action potential amplitude decrease,upstroke velocity slows, and the action potentialshortens (for ref. see Janse and Kleber, 1981). Theserapid changes in electrical activity are accompaniedby the first phase of a triphasic accumulation ofextracellular potassium. This initial phase is rapidlyreversible if reperfusion occurs within 15 minutes ofocclusion (Harris et al., 1954; Wiegand et al., 1979;Hill and Gettes, 1980; Hirche et al., 1980; Weiss andShine, 1982). Following the rapid and reversible ac-cumulation, extracellular K+ concentration ([K+]o) re-mains almost constant for 10-20 minutes before afinal irreversible increase occurs. This last phaseseems to be due to eel! damage, since intracellularwater and sodium increase (Schwartz et al., 1973).

The interdependence of metabolic changes, K+ ac-cumulation, shifts of other ions, and the changes ofelectrical activity in the early phase of ischemia is notfully understood. In a closed system such as globalischemia, total potassium content would be expectedto remain constant because no additional potassiumcan enter, or escape from, the system after circulatory

arrest. Thus, the observation that [K+]o increases dur-ing global ischemia implies a net loss of K+ from thecells and/or a diminution of the extracellular volumedue to a shift of water from the extra- to the intracel-lular space. The latter may play a role because os-motically active intracellular metabolites are knownto be increased during ischemia (Tranum-Jensen etal., 1981). However, the major component of theincrease in [K+]o probably is attributable to a net lossof cellular potassium. This, in turn, must result fromdecreased K+ influx, or increased K+ efflux, or both.If ischemia were to interfere with Na+/K+ pumpingand thereby decrease K+ influx, both [K+]o and[Na+]; would be expected to rise.

This study explores the relationship between rest-ing membrane potential (RMP) and extracellular po-tassium activity during the initial phase of acute is-chemia. We evaluated the mechanism of K+ accu-mulation during this phase by determining the effectsof several interventions (rapid stimulation; applicationof strophanthidin) on the changes in aj< and by meas-uring intracellular Na+ activity (aka). The results dem-onstrate that the resting membrane potential approx-imates the potassium equilibrium potential during

by guest on June 17, 2018http://circres.ahajournals.org/

Dow

nloaded from

Kieber/ Resting Membrane Potential, aR and aka in Ischemia 443

early acute ischemia. Furthermore, the results showthat aka remains low and Na+/K+ pump activity per-sists during early ischemia.

Methods

Perfused Heart Preparation

Guinea pigs (weight 900 g) were anesthetized with etherand stunned by a blow to the head. The heart was removedrapidly and transferred to a tissue chamber. The aorta wascannulated and the heart perfused with cold Tyrode's so-lution within 150 seconds after removal. Draining cannulaswere inserted in both ventricles, and the heart was con-nected to the final perfusion apparatus and placed horizon-tally in a recording chamber (Fig. 1). The Langendorffperfusion apparatus incorporated a roller pump to collectthe returning venous blood and propel it through an oxy-genator and a thermostat back to the heart in a closed loop.The priming volume of the perfusion system was 400 ml.The hearts were perfused at constant pressure (45-55 mmHg); myocardial flow rates varied between 100 and 150 ml/min per 100 g. The heart was superfused concurrently withthe same solution, except for a small epicardial area (di-ameter approximately 7 mm) which was kept in open air tofacilitate measurements with floating microelectrodes.Global ischemia was produced by interrupting the aorticflow with a clamp placed a few centimeters proximal to theaortic root (Fig. 1). Aortic occlusion did not affect the flowof superfusate, which was maintained throughout the ex-periment. The superfusion prevented a temperature dropafter the interruption of aortic flow. No visible drying ofthe surface exposed to air was observed, even after inter-ruption of perfusion. Temperature was monitored by athermistor placed on the epicardium close to the measuringsite. Temperature varied between 31°C and 33°C, but re-

Superfusion

Perfusion

vM

Stim A o r t ' c

NJ . Occlusion

Id.Hemp.It

Outj

FIGURE 1. Experimental arrangement. The perfused heart wasplaced horizontally in a perspex chamber and also superfusedexcept at the measuring site (area of 7 X 7 mm) which was kept inopen air.

mained constant during individual experiments. The oxy-genator was gassed with a mixture of oxygen and CO2; theflow rate of CO2 was adjusted to yield a pH between 7.34and 7.37. A conventional glass electrode was used to meas-ure pH of the perfusate close to the aortic inlet (see Fig. 1).Arterial oxygen content of the perfusate varied between 13and 15 ml O2 per 100 ml. A filter (pore size 40 /im) wasintroduced into the perfusion tubing to prevent embolismby corpuscular fibrin aggregates. Normal perfusion fluidwas composed of a mixture of washed bovine erythrocytes(hemoglobin concentration, 8 g/100 ml), dextran (mol wt,70,000; 4 g/100 ml), insulin (1 U/liter), heparin (400 U/liter) and Tyrode's solution of the following composition(mM): Na+, 149; K+, 4.5; Mg++, 0.49; Ca++, 1.8; Cl", 145.8;HCCV, 11.9; H2PCV, 0.4; and glucose, 20. Final [K+] inthe perfusion fluid varied within a ±0.5 mM range amongdifferent experiments, due to potassium uptake by eryth-rocytes during storage and/or hemolysis. Therefore, in eachexperiment, [K+] was determined by flame photometry orby an ion-selective probe.

Preparation and Calibration of Electrodes

Floating electrodes for recording conventional and ion-sensitive signals were constructed by means of differenttechniques. Intracellular floating electrodes for the meas-urement of transmembrane potentials were pulled fromborosilicate glass. DC tip resistances were 15-20 MQ whenfilled with 3 M KCl. The tips of similar borosilicate glassmicroelectrodes were broken to an outer diameter of 30 /imto prepare floating electrodes for measurements of extra-cellular reference electrograms. The broken tips were fire-polished to produce a smooth surface with an opening of3-4 /im. This procedure prevented injury to the subepicar-dial cells of the contracting preparation. Extracellular mi-croelectrodes were filled with a mixture of agar and 0.5 MKCl, except those which served as a reference for extracel-lular K+-sensitive electrodes which contained agar and 0.5M Na2SC>4.

Potassium-sensitive electrodes were prepared by twodifferent methods. Floating extracellular K+-sensitive elec-trodes for some of the measurements (shown in Figs. 2 and3) were made from fire-polished micropipettes silanized inan oven (200°C for 30 minutes) with trimethylsilyldimeth-ylamine (TMSDA; Fluka, Switzerland) and backfilled withpotassium sensor (Corning 477317). Calibration curves wereobtained from measurements of the potential differencebetween the K+-sensitive electrode and a reference electrodein perfusion fluid and in Tyrode's solution containing vary-ing [K+] but constant total [Na+] + [K+], In other measure-ments, K+-sensitive electrodes were made from polyethyl-ene tubing pulled to an inner tip diameter of 70-100 /im.These tubes were filled with 0.5 M KCl and sealed at the tipby a Valinomycin-PVC matrix membrane (valinomycin 4%).The fabrication and calibration of these electrodes were thesame as described by Hill and Gettes (1978). The Valino-mycin electrodes proved to be superior because they ex-hibited better DC stability and could be stored up to 2weeks with minimal loss of sensitivity (<10%). An activitycoefficient of 0.748 was assumed to evaluate voltages re-flecting extracellular K+ activity (Bates et al., 1970).

Floating Na+-sensitive microelectrodes were pulled fromborosilicate glass and beveled to a tip diameter of 1 /im.The electrodes were silanized in an oven (200°C for 30minutes) using TMSDA and stored in a desiccator forseveral hours to several days. Subsequently, the electrodes

by guest on June 17, 2018http://circres.ahajournals.org/

Dow

nloaded from

444 Circulation Research/Vo/. 52, No. 4, April 1983

were backfilled with sodium sensor [10% neutral ligandETH 227 + 0.5% sodium tetraphenylborate dissolved in o-nitrophenyloctylether, kindly provided by Dr. Ammann(Steiner et al., 1979)] and calibrated with a set of solutionscontaining varying [Na+] but constant total [Na+] -I- [K+].The ionic composition of the calibration solutions was madesimilar to that of the sarcoplasm (in min): Na+ + K+, 142;MgCI2/ 2; Tris, 10; pH 7.2. The DC shift of the Na+-sensitiveelectrodes was less than 3 mV/hour and the selectivity forNa+ vs. K+ was 50:1 (for details see Cohen et al., 1982). Anactivity coefficient for Na+ of 0.764 was used (Bates et al.,1970) to evaluate the calibration records in terms of Na+

activity. Both ion-selective and conventional microelec-trodes were mounted on a fine AgCl-coated silver wire(diameter 20 (im) to provide more stable records from bothextra- and intracellular locations.

Recording Procedure

Measurements of extra- and intracellular potentials weremade on the air-exposed part of the anterior left ventricular

-40

-50

-60

-70

•80-

-90

-100

(mM

)vi

ty

X

16-

14

12

10

8

6

4

- ?9 7 9

0 5 10 15 20

Time After Aortic Occlusion (min)

FIGURE 3. Time course of resting membrane potential (filledsquares) and a£ (filled circles) during the first 15 minutes after aorticocclusion. Mean values from nine experiments with bars indicatingstandard deviation. Number of measurements given near eachpoint. Smooth curve was fitted by eye to the ale values. Dotted linecorresponds to time course of calculated EAT.

2 - 5 0

10

53

2 3 4 5 6

Time After Aortic Occlusion (min)

£ ? B_ -40

| -50

^ -60

£ -70

| -80cn

• S -90

*. -100

tr

^ • 2 0

^ 15

race

llu o

X

^ 5

• ••

. /

Aortic5

Occlusion

Time

10

(mm)

15Reperfusion

20

FIGURE 2. Panel A: simultaneous recording of membrane potentialand aft impalement maintained between 2 and 6 minutes afteraorticocclusion. Resting membrane potential declined and action poten-tial amplitude decreased. Arrow indicates brief period of electricalalternans before unresponsiveness occurred at a resting membranepotential of —60 mV. Panel B: time course of resting membranepotential (filled squares) and aj|- (filled circles) during and after a15-minute period of ischemia; same experiment as panel 2A.Smooth curve was fitted by eye to a# values. Dotted line representspotassium equilibrium potential Eg, calculated from ale, as describedin Methods, using an assumed value of 100 min for initial intracel-lular potassium activity.

surface. Extracellular electrodes touched the epicardial sur-face, whereas intracellular electrodes penetrated the epicar-dium to an estimated depth of 125-200 fim 0anse et al.,1978; Tranum-Jensen et al., 1982). Recording electrodeswere connected to high-input impedance preamplifiers(Burr-Brown 3225 for conventional extra- and intracellularelectrodes; Analog Devices 515 for ion-sensitive electrodes)and the signals amplified by a differential instrumentationamplifier. The recorded potentials were stored on an analogtape recorder (Ampex PR 2200) and displayed either on aUV fiberglass recording system (Electronics for MedicineVR 12) or on a pen recorder. K+-Sensitive recordings werefiltered for display at low speed (low pass, corner frequency1 Hz). A bipolar electrode was apposed to the surface of theventricle for delivery of electrical stimuli that controlled theventricular rate. In some experiments, the effect of stimu-lation frequency on extracellular K+ activity was investi-gated by superfusing the sinoatrial node with cold solutionto lower the spontaneous heart rate.

Transmembrane potentials (Vm) were measured as thepotential difference between an intracellular microelectrodeand an extracellular reference microelectrode placed as closeas possible to the impalement site. Extracellular K+ activity(a?t) was measured from the potential difference (Vdirr)between the extracellular K+-sensitive electrode (VKE) andan extracellular reference electrode (Vref). Recordings ofVdifr showed a small bipolar extracellular electrogram dueto the finite distance (<200 /im) between the K+-sensitiveelectrode and the intracellular reference. The measurementsof Vdiff were taken during the T-Q segment of the extracel-lular electrogram. Intracellular Na+ activity (aisia) was meas-ured from the potential difference (VNBE — Vm) between aNa+-sensitive intracellular electrode (VN8E) and a conven-tional 3 M KG intracellular electrode (Vm). A single extra-cellular floating micropipette served as extracellular refer-ence for both intracellular measurements (Vm and VN8E).Both intracellular electrodes were impaled within a distanceof 200 /xm. In preliminary experiments, an attempt wasmade to employ double-barreled floating electrodes in thehope of obtaining measurements from single cells. Unfor-tunately, the inertia of the large double-barreled tips didnot allow stable intracellular measurements in the contract-ing preparation due to cell damage.

by guest on June 17, 2018http://circres.ahajournals.org/

Dow

nloaded from

KJeber/ Resting Membrane Potential, a^ and aka in Ischemia 445

Several possible problems with the measurement tech-niques were addressed. The decline of P02 from the air-surface interface into the subepicardial cell layers was cal-culated because oxygen diffusion from the air to the site ofrecording theoretically could affect the results during ische-mia. P02 declines as a parabolic function of distance de-pending upon oxygen consumption and the permeationcoefficient for oxygen (see Tsacopoulos et al., 1981). Partialpressure of oxygen is expected to fall from a level of 147mm Hg at the surface to below 5 mm Hg at a depth of 90jum. Since the estimated depth of electrode penetration isgreater, it seems unlikely that oxygen diffusion into themyocardium could have affected the results during ische-mia. In addition, several experiments were performed todetermine whether the epicardial connective tissue sheathis a diffusion barrier for K+ ions. The thin layer of tissuewas removed and a5< measured. No significant differencewas noted between aj; with the surface removed or intact.

Calculation of Potassium Equilibrium Potential

The potassium equilibrium potential (EK) was calculatedfrom the measured values of extracellular potassium activityas follows:

EK = 60.3 mV X logak — f X

where 60.3 mV = Nernst slope for a temperature of 31°C;al< represents an assumed value of 100 ITIM for initialintracellular potassium activity (see discussion); Aaj< is thedifference between the actual aj< and the extracellular K+

activity prior to occlusion (i.e., the extra potassium accu-mulated); and f is the extra- to intracellular volume ratioassumed to be 0.33 (Page, 1962).

Results

Relationship between Changes in RestingMembrane Potential and Extracellular PotassiumActivity

In nine experiments, an was measured simultane-ously with resting membrane potential (RMP) duringa 15-minute period of ischemia. The ventricular ratewas 150 to 180 per minute in all experiments. Figure2A shows the result of an experiment in which trans-membrane potential and a?c were recorded continu-ously from an epicardial site between 2 and 6 minutesafter coronary occlusion. The gradual increase of anwas accompanied by a decrease in RMP and actionpotential amplitude. Electrical alternans, which hasbeen described by several investigators (see Janse andKleber, 1981), occurred between 4.5 and 5.5 minutesafter occlusion, just prior to unresponsiveness (RMP—60 mV). The results of the entire experiment (com-bining the data in Figure 2A with other membranepotential records obtained within a 300 jiim radius ofthe K+-sensitive electrode) are shown in Figure 2B.Prior to occlusion, the resting potential was —82 mV.In the first 8 minutes after aortic occlusion, RMPdeclined rapidly to —54 mV, followed by slowerdepolarization to —46 mV after 15 minutes of ische-mia. Successive impalements within the area of studygave no indication of more than one level of restingmembrane potential. Changes in extracellular K+ ac-

tivity showed a similar time course to the changes inRMP. After an initially steep increase, a^ rose moreslowly to a value of 17.7 ITIM after 15 minutes ofocclusion. The dotted line in Figure 2B represents thepotassium equilibrium potential calculated from themeasured values of aj; (see Methods section). Beforeocclusion, EK was 4 mV more negative than the restingmembrane potential; during the subsequent depolar-ization, EK approached RMP as a?; increased. Therelationship between EK and RMP during reperfusionwas less clear, probably reflecting the technical diffi-culty of maintaining stable impalements during earlyreperfusion due to the increased strength of contrac-tions. Furthermore, heart rate often was uncontrolledduring early reperfusion because of ventricular tachy-arrhythmias. This was a significant problem becausesudden increases in heart rate (i.e., during the ven-tricular tachyarrhythmia) resulted in a transient in-crease in a?c. This effect could be mimicked by asudden increase in stimulation rate during reperfusion(see Fig. 4).

The time course and magnitude of changes in RMPand aK during ischemia were similar in all nine ex-periments (see Fig. 3). Resting membrane potentialwas —81.7 ± 2.0 mV (SD) prior to occlusion. After 8minutes, RMP was —55.3 ± 3.0 mV and by 15 min-utes, —49.5 ± 2.7 mV. Extracellular K+ began to risewithin the first minute of occlusion and reached av-erage values of 11.0 ± 2.0 mM and 14.7 ± 1.3 ITIM after

40,

3 20

E

{ A

cti

n5Z)

aj

12

10

3

6

U

10 0 5t

Aortic Occlusion

Time (mm)

10t

Reperfusion

15

FIGURE 4. Effects of an abrupt change in heart rate on a£ duringthree different time intervals: prior to occlusion, during occlusion,and during reperfusion. Upper trace: original recording of potentialsensitive to a'k (\',t,n)- Lower trace: a£- calculated from V,/,//. Thespontaneous heart rate (60 beats/min) was increased abruptly to170 beats/min for the times indicated by the bars. In the perfusedpreparation, the time course of reversion of a'k to the steady statelevel during and after stimulation (left half of figure) suggesteddelayed activation and inactivation of the Na+/K* pump. Rapidstimulation during ischemia (right haif of figure) caused a morepronounced increase in K+ accumulation. The decline of a'k afterstimulation implies that Na*/K* pump activity resulted in a netuptake of K+ from the extracellular space.

by guest on June 17, 2018http://circres.ahajournals.org/

Dow

nloaded from

446 Circulation Research/Vol. 52, No. 4, ApriJ 1983

8 and 15 minutes, respectively. Prior to occlusion, theaverage potassium equilibrium potential was calcu-lated to be —89.2 mV, 7.5 mV more negative thanRMP. EK approached RMP after 5 minutes of ischemiaat an an of 9 HIM. Fifteen minutes after the onset ofischemia, EK averaged —49.4 mV, virtually identicalto the average RMP. However, this apparently accu-rate relationship should be interpreted with cautionsince the calculated EK depended on the assumedinitial value of 100 min for ak (see Discussion).

Effects of Heart Rate on Extracellular PotassiumAccumulation

Abrupt changes in heart rate lead to fluctuations ofextracellular potassium activity in different isolatedcardiac tissues, caused by the delay between the rate-dependent change of K+ efflux and active, pumpedK+ influx. Therefore, changes in heart rate provide ameans to study activation and inactivation of the Na+/K+ pump (Kunze, 1977; Kline and Morad, 1978). Inthree experiments, spontaneous heart rate was low-ered to 60/min by local supervision of the rightatrium with cold perfusate. As shown in Figure 4, asudden increase in heart rate to 170/min during nor-mal perfusion resulted in a transient increase of a?;followed by a decrease toward the former steady statelevel. A transient undershoot of the steady level of a$<was observed after the train of rapid stimuli, consist-ent with delayed inactivation of active K+ influx. Boththe time course and magnitude of changes in aK weresimilar to those reported from isolated cardiac tissue(Kunze, 1977). After aortic occlusion, rapid stimula-tion led to a pronounced K+ accumulation. The rateof accumulation was fastest immediately after theonset of stimulation followed by a gradual slowing. Atransient decline of aK was again present after rapidstimulation. Since extracellular washout was pre-vented during the period of coronary occlusion, thistransient depletion (1.2 ITIM) is assumed to reflect anet potassium movement from the extra- to the intra-cellular space.

In two experiments, strophanthidin (10~6 M) wasadded to the perfusion fluid in order to assess thecontribution of changes in active K+ influx to thevariation of a?; before and after coronary occlusion(Fig. 5). Measurements were performed 30 minutesafter the administration of the cardiac steroid. Theexperimental procedure was the same as describedfor Figure 4. The left part of Figure 5 shows that theinitial increase of a?c after the onset of stimulation(170/min) was not followed by a delayed decrease.Instead, a& remained at a level 1 mM higher than a?cof the perfusion fluid. After rapid stimulation, aS<declined to the former baseline level without transientundershoot. During ischemia (Fig. 5, right), rapidstimulation (170/min) caused an immediate increasein the rate of K+ accumulation and there was nodistinct decline in the rate of K+ accumulation duringrapid stimulation, in contrast to experiments withoutstrophanthidin (e.g., Fig. 4). Furthermore, there wasno decline in aS< following cessation of rapid stimu-

| 12

0 10 p 5Aortic Occlusion

10Reperfusion

15

Time (min)

FIGURE 5. Effect of a change in heart rate in presence of 10 "6 Mstrophanthidin. Rapid stimulation (170 beats/min) indicated byhorizontal bars. During strophanthidin perfusion, block of the Na */K+ pump is suggested by the absence of a decrease in ale during,and of an undershoot in a£ following, rapid stimulation (170 beats/min). During ischemia, rapid stimulation resulted in a steadyincrease in K+ accumulation; after stimulation, ale did not decline.

lation in both experiments conducted in the presenceof strophanthidin.

Variations of aK during ischemia brought about bychanges in frequency appeared to be superimposedon a process of K+ accumulation which was more orless independent of rate. Three experiments wereperformed to investigate this possibility. In Figure 6,values for aK from two successive occlusions 30 min-utes apart are superimposed. Heart rate was con-trolled by ventricular stimulation. The solid line (filledcircles) shows the results obtained when the prepa-ration was stimulated at 50/min before and duringthe first period of ischemia. The broken line (filledsquares) shows the results obtained when the heart

i i o

0 S I 10 T 15Aortic Occlusion Reperfusion

Time (mm)

FIGURE 6. Effects of heart rate on a% prior to and during aorticocclusion. Superimposed are values of a£ before and during twosuccessive occlusions. Rates of stimulation are given in the upperpanel above the experimental results. Filled circles and smoothcurve: stimulation at 50 beats/min. Filled rectangles and dashedcurve: stimulation at 170 beats/min. Note during stimulation at 170beats/min before aortic occlusion, a£ had declined (dashed curve)to approximately its steady state level at 50 beats/min (smoothcurve).

by guest on June 17, 2018http://circres.ahajournals.org/

Dow

nloaded from

Kleber/ Resting Membrane Potential, a j and aka in Ischemia 447

rate was increased abruptly to 170/min, 5 minutesprior to occlusion. In the latter case, extracellular K+

activity initially rose but then declined to approachthe former steady state level before occlusion (cf. Fig.4). During ischemia, the time course of ax was almostidentical at both rates. This result, confirmed in twoother experiments, suggests that potassium accumu-lation during early ischemia is largely independent ofheart rate (within a 50 to 170/min range), providingactive K+ influx is allowed to reach a steady stateprior to occlusion, and heart rate during occlusion isconstant.

Intracellular Sodium Activity during AcuteIschemia

Intracellular sodium activity (aka) was measuredbefore and after aortic occlusion in eight experiments.These experiments tested the hypothesis that the netmovement of potassium ions from the intra- to theextracellular space is compensated by an inwardmovement of sodium ions. Intracellular Na+ activitywas determined by means of two separate microelec-trodes impaled within a distance of 200 fim. With thistechnique, measurements were obtained before andat different times after interruption of aortic flow.Figure 7 shows original recordings of transmembranepotential (Vm) and potential measured by the Na+-sensitive electrode (VNSE)- TWO simultaneous recordsbefore aortic occlusion are given in panel A. Restingmembrane potential (RMP) was —81 mV; 35 sec afterthe impalement VNHE was —136 mV. The potentialdifference (Vdiff) between VNSE and RMP was —55mV, corresponding to an aNa of 9.9 mM. In panel B,recordings from the standard intracellular microelec-trode (upper trace) and the Na+ electrode (lower trace)are displayed at a rapid paper speed. The action

E - 5 0

-100

_ 0

1 -503-100

> -150

0

30sec 500msec 30sec

FIGURE 7. Original intracellu/ar recordings of membrane potential(Vm, upper trace) and potential recorded by the Na """-sensitiveelectrode (V,VnK, lower trace). Panel A: recordings from the perfusedheart. Resting level of Vm was —81 mV, resting level of V,va£ - 136mV. Vdiff of -55 mV corresponded to a'/vo of 9.9 mM. Panel B:recordings of Vm and V,wae during normal perfusion displayed athigher speed. Note distortion of action potential recorded by theNa +-sensitive electrode. Pancel C: recordings from the same exper-iment as in panels A and B, 15 minutes after aortic occlusion. Vm

was -52 mV, VlVo£ was -112 mV. Vdiff of -60 mV correspondedto a !vn of 7.6 min. The upstroke of the action potential (Vm, panel B)was retouched for clarity.

•fea.

brar

en

-50-

-6 -

-70-

-80-

• T

-90J

15-

S\+T

o• a

0 5 10 15Time After Aortic Occlusion (min)

FIGURE 8. Resting membrane potential and a'xa before and atdifferent times after aortic occlusion. Different symbols representindividual measurements from eight different experiments.

potential recorded by the Na+ electrode was distortedby its low frequency response. The potential reacheda steady resting level only 400 msec after the upstroke.This poor electrical response time limited the rangeof heart rates over which a^a could be measured. Inpanel C, simultaneous recordings of Vm and VNSE aredepicted, from the same experiment, after 15 minutesof ischemia. Vm was —52 mV and VNSE —112 mV; Vdiffof —60 mV corresponded to an a^a of 7.6 mM, i.e., 2.3mvi less than that prior to aortic occlusion. The com-bined results of eight experiments (in seven prepara-tions) are summarized in Figure 8. Heart rate variedbetween 70 and 120/min in the different experiments.Resting membrane potentials and aka are given foreach experiment; up to four successful simultaneousimpalements were obtained in a single experiment.The time course of RMP was comparable to the datapresented in Figure 3. RMP decreased from a value of—79 ± 3.6 mV to values ranging from —47 to —54 mVbetween 10 and 15 minutes after aortic occlusion.During normal perfusion, the mean a^,, was 8.8 ± 2.8min (SD, n = 8). During ischemia, no increase or aslight decrease in akja was recorded. Between 10 and15 minutes after aortic occlusion, when resting mem-brane potential was approximately —50 mV, eightmeasurements (five experiments) were obtained,ranging from 4.7 to 7.6 mM. Mean a^a before occlusionfrom these five experiments was 7.9 mM.

From the resting membrane potentials shown inFigure 8 and the relationship between RMP and aS<(Fig. 3), a theoretical estimate was made for the in-

by guest on June 17, 2018http://circres.ahajournals.org/

Dow

nloaded from

448 Circulation Research/Vo/. 52, No. 4, April 1983

crease of alj, which should occur if extracellular K+

accumulation was fully compensated by intracellularNa+ accumulation. Thus, depolarization to —50 mV(see Fig. 3) was accompanied by an increase in aj; of11 mM. Therefore, assuming an extra- to intracellularspace ratio of 0.33, a 3.6 mM increase of aha shouldhave occurred after 10 to 15 minutes of ischemia. Forthe type of Na+ electrodes employed, an increase ofaha of this magnitude (i.e., from 8 to 11 mvi) would beexpected to change Vditr by approximately 8 mV. Apotential difference of that size certainly should havebeen detectable with these methods but was not ob-served.

Discussion

Intracellular potentials and extracellular K+ activitywere measured simultaneously in guinea pig hearts toevaluate the relationship between resting potentialand the potassium equilibrium potential after cessa-tion of myocardial perfusion. A model of global is-chemia was used to reduce spatial differences thatoccur with regional ischemia. The results demonstratea magnitude and time course for changes in a?t duringischemia similar to those reported for intramural a?;during ischemia in vivo (Hill and Gettes, 1980). Thisclose agreement for ax suggests that local variationsin ischemic metabolism, evaporation from the epicar-dial surface, and O2 diffusion into the air-exposedventricle were unlikely to have affected the resultssignificantly. In addition, the decrease in resting po-tential described qualitatively in the perfused pigheart (Downar et al., 1977) and in isolated measure-ments during regional ischemia (Kleber et al., 1978)has been verified and more precisely defined by thepresent results. Moreover, the results reveal thatchanges in resting potential and aj< both follow asimilar sigmoidal time course, suggesting (not sur-prisingly) a close relationship between the transsar-colemmal K+ distribution and the resting potential.

This relationship can be evaluated by comparingRMP with estimates of EK. The results demonstrate acloser approximation of RMP to EK as ischemia pro-gressed (the two values concurring around —50 mV).The close correspondence between EK and RMP arisespartly because the initial al< was assumed to be 100mM. This was not an unreasonable assumption, sincemeasured values for guinea pig papillary muscle (99mM, Wier, 1978; 106 mM, Baumgarten et al., 1981)and the mean of other reported values for aj< (86 mM,Lee and Fozzard, 1975; 116 mM, Cohen et al., 1982)are close to this value. Deviations of initial ak (±10mM) would have shifted EK by ±2 to 3 mV but wouldnot have affected the conclusion that EK is closer toRMP during ischemia than during normal perfusion.In addition, since total tissue K+ is constant in theabsence of extracellular washout, changes in ak duringischemia were estimated (3-4 mM after 15 minutes ofocclusion) and allowed for in the calculation of EK. Asimilar relation between RMP and EK at elevated[K+]o has been described in normoxic myocardium(Lee and Fozzard, 1975).

The implication of a close relation between EK andRMP during ischemia is that relative K+ permeability(PK) predominates over relative Na+ permeability(PNE)- The PK:PN8 ratio may be even higher in ischemicthan normal myocardium at elevated a?;- Comparisonof upstroke velocities and action potential amplitudesduring ischemia and during perfusion with increased[K+]o suggests that potassium conductance is in-creased and/or fast inward sodium current is de-creased in myocardial ischemia (see Janse and Kleber,1981). Additionally, Vleugels et al. (1980) have meas-ured an increase in slope conductance at EK and time-independent K+ outward current during hypoxic per-fusion.

A major emphasis of the present study was toinvestigate the mechanism underlying the reversibleincrease in aK that occurs during the first 15 minutesof ischemia. The current findings, together with thoseof Tranum-Jensen et al. (1981), would suggest that theincrease in a?c can be attributed largely to a net effluxof K+ from the myocardial cells. Certainly, a shift ofwater from the extra- to the intracellular space mustbe discussed as a possible factor contributing to theincrease in aK. Such a shift is expected because newosmotically active particles (e.g., lactate, phosphate)are formed in the intracellular compartment duringischemic metabolism. On the other hand, quantitativeconsiderations argue that transsarcolemmal watermovements influence aj to a rather small extent.Thus, Tranum-Jensen et al. (1981) measured a 20mOstn increase in extracellular osmolality after 15minutes of ischemia in Langendorff perfused pighearts. In the worst case (i.e., no new osmoticallyactive particles added to the extracellular space), thewater shift required to account for that increase inextracellular osmolality would amount to a 6% loss ofextracellular fiber water and a concomitant 2% gainof intracellular fiber water. This calculation certainlyargues against a significant effect of extra- to intra-cellular water movement, since an 80% loss of extra-cellular volume would be required to explain a rise ina5< from 3 to 14.7 mM. Thus, the rise in aft duringischemia must have resulted from a net loss of K+

from the cells.The relative contributions of changes in passive K+

efflux and in active K+ influx to the potassium accu-mulation were studied by measuring the effect ofheart rate on aK under two conditions: (1) during andafter rapid stimulation and (2) during constant stim-ulation at different rates. The first setting alloweddiscrimination between instantaneous changes of K+

efflux and delayed variations in active K+ influx(Kunze, 1977; Kline and Morad, 1978). Rapid stimu-lation during ischemia resulted in an immediate in-crease in the rate of change of aft, reflecting anincrease of efflux at the high rate of stimulation. Theslowing in the rate of rise of a?; that developed duringstimulation and the subsequent K+ depletion afterstimulation strongly suggest that the mechanism foractive K+ inward transport, i.e., the Na+/K+ pump,remains at least partially functional during ischemia.This interpretation is supported by the blocking effect

by guest on June 17, 2018http://circres.ahajournals.org/

Dow

nloaded from

Kieber/ Resting Membrane Potential, an and aka in Ischemia 449

of 10 6 M strophanthidin on these changes. Similareffects of rapid stimulation on a?; have been reportedrecently by Weiss and Shine (1982).

In a second series of experiments, K+ accumulationwas measured at constant stimulation rates of 50 and170 beats/min. If K+ influx (at 170 beats/min) wasallowed to reach a steady state prior to occlusion, a?cwas found to be independent of heart rate duringischemia. This result suggests that, even during ische-mia, active K+ inward transport is still capable of fullycompensating for the frequency-dependent increasein K+ efflux between rates of 50-170 beats/min, andit is in accordance with measurements reported fromin situ pig and dog hearts (Wiegand et al., 1979; Hilland Gettes, 1980). However, the findings differ fromthe results of Weiss and Shine (1982), who found anincrease in extracellular K+ accumulation if heart ratewas changed from 60-120 beats/min prior to circu-latory arrest and kept constant during the ischemicperiod. Any explanations for this discrepancy arenecessarily speculative, but it seems possible that therelatively low O2 content of their perfusate (Tyrode'ssolution, no oxygen carrier) could have limited themaintenance of oxygen delivery during the increasedoxygen consumption expected at higher rates. Such asituation might lead to changes in intracellular metab-olism prior to interruption of the circulation (e.g.,decrease in levels of energy-rich phosphates, intracel-lular acidosis) which could then influence extracellu-lar K+ accumulation during the subsequent ischemicperiod.

In the final series of experiments, intracellular Na+

activity determined prior to occlusion (8.8 ± 2.8 mivt)agreed with values obtained in sheep and guinea pigventricular strands measured with the same type ofNa+-sensitive ligand (7.9 mM, pacing at 0.5 Hz; Cohenet al., 1982). The observation that aNa remained low(4.7-7.6 mM, after 10 to 15 minutes of ischemia)suggests that the normal regulation of active Na+/K+

exchange by extracellular K+ and intracellular Na+ islittle affected in early ischemia. In normally perfusedtissue, changes in the activity of the Na+/K+ pumpare due mainly to variations in aka because pumpactivation by [K+]o becomes saturated at [K+]o above4-5 mM (Glitsch et al., 1976; Gadsby, 1980; Glitsch etal., 1981). Therefore, when aj< is elevated, only thosechanges in K+ efflux linked to concomitant changesin Na+ influx involve a change in the activity of theNa+/K+ pump (e.g., rate-dependent fraction of K+

efflux, see Fig. 6). Depolarization of the cells possiblymay contribute to a low aj\ia during ischemia. A changeof RMP by +30 mV after 15 minutes of ischemiareduces the electrochemical gradient for Na+ (ENS —Em, assumed EN8 +60 mV). This, in turn, diminishesthe energy requirement by 22% for maintenance ofaka in the static situation, i.e., when changes in Na+

influx are neglected. Measurements of energy-richphosphate compounds indicate that ATP levels de-crease relatively slowly during the initial phase ofischemia. Furthermore, ADP is largely converted toATP and AMP, the thermodynamic buffering systemsand glycolysis tending to maintain a high ATP poten-

tial (Jennings et al., 1981). In addition, the effect of apotential-dependent change of Na+ influx on a ashould be considered. This has been studied in sheepPurkinje fibers where depolarization induced a de-crease in Na+ influx and, hence, in ak'a (Eisner et al.,1981).

For reasons of electroneutrality, the net efflux ofpotassium ions must be balanced either by an equiv-alent influx of cations or by an equivalent efflux ofanions. The present results exclude the possibility ofa net influx of sodium, but they do not provide directinformation concerning the identity of the chargeswhich do maintain electroneutrality. For instance, anincrease in intracellular Ca++ cannot be excluded.However, even inward movement of all Ca++ ionsavailable in the extracellular space would not besufficient to compensate for the millimoles of extra-cellular K+ accumulated. Moreover, a significant shiftof Ca++ ions into the intracellular compartment wouldbe expected to increase resting tension, an effect notobserved by Weiss and Shine (1982). Therefore, themost probable hypothesis is that net K+ efflux occursconcomitantly with efflux of anions. In fact, such amechanism is suggested from experiments in whichanaerobic metabolism was induced by severe restric-tion of coronary flow. Analysis of venous effluent inthis situation demonstrated that the time course ofthe increase in potassium, phosphate, and lactate wasidentical. The quantity of lactate and phosphate evenexceeded that of potassium, probably by the amountof protons released (Mathur and Case, 1973). Loss ofpotassium associated with efflux of anions also couldhave its counterpart in skeletal muscle brought intoa state of fatigue at low extracellular pH. Under suchconditions, the resulting loss of intracellular lactateions is greater than the efflux of H+ (Mainwood andWorsley-Brown, 1975), and an increase in net K+

efflux also occurs (Mainwood and Lucier, 1972).Mainwood and Lucier (1972) provided a hypothesiswhich might explain the relationship between intra-cellular metabolic acidosis and net potassium effluxduring ischemia. Generation of weak acids inside thecells by breakdown of energy-rich phosphates andglycolysis will result in binding of the protons tonegative charges of intracellular proteins which mayhave a buffer capacity as high as 60 mM per unit pHand per kg sarcoplasmic fluid (Heisler and Piiper,1971). A loss of permeant anions so formed, associ-ated with an efflux of cations, is then to be expectedas a result of the decrease in fixed negative chargesand the rise in freely movable anions (Boyle andConway, 1941). Decrease of myocardial intracellularpH during ischemia, determined indirectly by nuclearmagnetic resonance (Garlick et al., 1979), showed thesame time course as the rise in ak, further supportingthe above hypothesis of anion loss.

J thank Dr. ML. Pressler for help with the manuscript, and 1.Schonberg, C. Cigada, and A. Meyer for technical assistance.

This work was supported by the Swiss National Science Foun-

by guest on June 17, 2018http://circres.ahajournals.org/

Dow

nloaded from

450 Circulation Research/VoJ. 52, No. 4, April 1983

dation (Grant No. 3.830-0.79) and the Swiss Foundation of Car-diology.

Address for reprints: Dr. A.G. KJeber, Department of Physiol-ogy, University of Berne, Buhlplatz 5, 3012 Berne, Switzerland.

Received July 22, 1982; accepted for publication January 28,1983.

ReferencesBates RG, Staples BR, Robinson RA (1970) Ionic hydration and

single ion activities in unassociated chlorides at high ionicstrengths. Anal Chem 42: 867-871

Baumgarten CM, Cohen CJ, McDonald TF (1981) Heterogeneity ofintracellular potassium activity and membrane potential in hy-poxic guinea pig ventricle. Circ Res 49: 1181-1189

Boyle PJ, Conway EJ (1941) Potassium accumulation in muscle andassociated changes. J Physiol (Lond) 100: 1-63

Cohen CJ, Fozzard HA, Sheu SS (1982) Increase in intracellularsodium ion activity during stimulation in mammalian cardiacmuscle. Circ Res 50: 651-662

Downar E, Janse MJ, Durrer D (1977) The effect of acute coronaryartery occlusion on subepicardial transmembrane potentials inthe intact porcine heart. Circulation 56: 217-224

Eisner DA, Lederer WJ, Vaughan-Jones RD (1981) The effects ofrubidium ions and membrane potential on the intracellular so-dium activity of sheep Purkinje fibres. J Physiol (Lond) 317: 189-205

Gadsby DC (1980) Activation of electrogenic Na+/K+-exchange byextracellular K+ in canine cardiac Purkinje fibers. Proc Natl AcadSci USA 77: 4035-4039

Garlick PB, Radda GK, Seeley PJ (1979) Studies of acidosis in theischemic heart by phosphorous magnetic resonance. Biochem J184: 547-554

Glitsch HG, Pusch H, Venetz K. (1976) Effects of Na and K ions onthe active Na transport in guinea-pig auricles. Pfluegers Arch365: 29-36

Glitsch HG, Kampmann W, Pusch H (1981) Activation of activeNa transport in sheep Purkinje fibres by external K or Rb ions.Pfluegers Arch 391: 28-34

Harris AS, Bisteni A, Russell RA, Brigham JC, Firestone JE (1954)Excitatory factors in ventricular tachycardia resulting from myo-cardial ischemia: Potassium a major excitant. Science 199: 200-203

Heisler N, Piiper J (1971) The buffer value of rat diaphragm muscletissue determined by Pco2 equilibration of homogenates. RespPhysiol 12:169-178

Hill JL, Gettes LS (1980) Effect of acute coronary artery occlusionon local myocardial extracellular K+ activity in swine. Circulation61: 768-778

Hill JL, Gettes LS, Lynch MR, Hebert NC (1978) Flexible valino-mycin electrodes for on-line determination of intravascular andmyocardial K+. Am J Physiol 235: H455-H459

Hirche HJ, Franz C, Bos L, Bissig R, Lang R, Schramm M (1980)Myocardial extracellular K+ and H+ increase and noradrenalinerelease as a possible cause of early arrhythmias following acutecoronary artery occlusion in pigs. J Mol Cell Cardiol 12: 579-593

Janse MJ, Kleber AG (1981) Electrophysiological changes andventricular arrhythmias in the early phase of regional myocardialischemia. Circ Res 49: 1069-1081

Janse MJ, Tranum-Jensen J, Kleber AG, van Capelle FJL (1978)Techniques and problems in correlating cellular electrophysiol-ogy and morphology in cardiac nodal tissues. In The Sinus Node,

Structure, Function and Clinical Relevance, edited by FIM Bonke.The Hague, Martinus Nijhoff

Jennings RB, Reimer KA, Hill ML, Mayer SE (1981) Total ischemiain dog hearts, in vitro. 1. Comparison of high energy phosphateproduction, utilization, and depletion, and of adenine nucleotidecatabolism in total ischemia in vitro vs. severe ischemia in vivo.Circ Res 49: 892-900

Kleber AG, Janse MJ, van Capelle FJL, Durrer D (1978) Mechanismand time course of S-T and T-Q segment changes during regionalmyocardial ischemia in the pig heart determined by extracellularand intracellular recordings. Circ Res 42: 603-613

Kline RP, Morad M (1978) Potassium efflux in heart muscle duringactivity: Extracellular accumulation and its implications. J Physiol(Lond) 280: 537-558

Kunze DL (1977) Rate-dependent changes in extracellular potas-sium in the rabbit atrium. Circ Res 41: 122-127

Lee CO, Fozzard HA (1975) Activities of potassium and sodiumions in rabbit heart muscle. J Gen Physiol 65: 695-708

Mainwood GW, Lucier GE (1972) Fatigue and recovery in isolatedfrog sartorius muscles: The effects of bicarbonate concentrationand associated potassium loss. Can J Physiol Pharmacol 50: 132-142

Mainwood GW, Worsley-Brown P (1975) The effects of extracel-lular pH and buffer concentration on the efflux of Iactate fromfrog sartorius muscle. J Physiol (Lond) 250: 1-22

Mathur PP, Case RB (1973) Phosphate loss during reversible myo-cardial ischemia. J Mol Cell Cardiol 5: 375-393

Page E (1962) Cat heart muscle in vitro. III. The extracellular space.J Gen Physiol 46: 201-213

Schwartz A, Wood JM, Allen JC, Bornet EP, Entmann ML, Gold-stein MA, Sordahl LA, Suzuki M, Lewis RM (1973) Biochemicaland morphologic correlates of cardiac ischemia. I. Membranesystems. Am J Cardiol 32: 46-61

Steiner RA, Oehme M, Ammann D, Simon W (1979) Neutralcarrier sodium ion-selective microelectrode for intracellular stud-ies. Anal Chem 51: 351-353

Tranum-Jensen J, Janse MJ (1982) Fine structural identification ofindividual cells subjected to microelectrode recording in perfusedcardiac preparations. J Mol Cell Cardiol 14: 233-247

Tranum-Jensen J, Janse MJ, Fiolet JWT, Krieger WJG, Naumannd'Alnoncourt Ch, Durrer D (1981) Tissue osmolality, cell swell-ing, and reperfusion in acute regional myocardial ischemia in theisolated porcine heart. Circ Res 49: 364-381

Tsacopoulos M, Poitry S, Borsellino A (1981) Diffusion and con-sumption of oxygen in the superfused retina of the drone (Apismellifera) in darkness. J Gen Physiol 77: 601-628

VIeugels A, Vereecke J, Carmeliet E (1980) Ionic currents duringhypoxia in voltage-clamped cat ventricular muscle. Circ Res 47:501-508

Weiss J, Shine KI (1982) Extracellular K+ accumulation duringmyocardial ischemia in isolated rabbit heart. Am J Physiol 242:H619-H628

Wiegand V, Guggi W, Meesmann W, Kessler M, Greitschus F(1979) Extracellular K+ activity changes in the canine myocar-dium after coronary occlusion and the influence of beta blockade.Cardiovasc Res 13: 297-302

Wier WG (1978) Ionic currents and intracellular potassium inhypoxic myocardial cells (abstr). Biophys J 21: 166a

INDEX TERMS: Acute myocardial ischemia • Resting membranepotential • Extracellular potassium activity • Intracellular sodiumactivity

by guest on June 17, 2018http://circres.ahajournals.org/

Dow

nloaded from

A G Kléberactivity during acute global ischemia in isolated perfused guinea pig hearts.

Resting membrane potential, extracellular potassium activity, and intracellular sodium

Print ISSN: 0009-7330. Online ISSN: 1524-4571 Copyright © 1983 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.52.4.4421983;52:442-450Circ Res. 

http://circres.ahajournals.org/content/52/4/442World 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 Research Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:

by guest on June 17, 2018http://circres.ahajournals.org/

Dow

nloaded from