J Mol Cell Cardiol

23, 639-649

Modification

of Cardiac

(1991)

Ionic Currents By Photosensitizer-generated Reactive Oxygen

Merrill

Tarr

and Dennis

Paul

Valenzeno

Department of Physiology, University of Kansas Medical Center, College of Health Sciences and Hospital. Kansas City, Kansas 66103, USA (Received 20 April

1990, accepted in revised form 11 January

1991)

M. TARR AND D. P. VALENZENO. Modification of Cardiac Ionic Currents By Photosensitizer-generated Reactive Oxygen. Journal of A4olccular and Cellular Cardiology (1991), 23, 639-649. The effects of reactive oxygen species (ROS) generated by light and the photosensitizer Rose Bengal on ionic currents in single frog atria1 cells were investigated. The excitatory inward sodium and calcium currents were both suppressed by ROS as was the outward, delayed rectifier potassium current. The inactivation kinetics of the sodium current were slowed markedly whereas the kinetics of calcium current inactivation were much less affected and potassium current activation was not changed. The sodium current-voltage relationship was shifted in the depolarizing direction by ROS whereas the voltage-dependencies of both the calcium and potassium currents were not affected. In addition to suppressing the time- and voltage-dependent sodium, calcium, and potassium currents, ROS enhanced a time-independent current which was outwardly directed at positive membrane potentials. However, the induction of this time-independent current required longer ROS exposure than was required to significantly suppress the other currents. The rapid onset of ROS-induced suppression of calcium and potassium currents followed by a later enhancement of a time-independent current can explain ROS-induced changes in action potential duration. Brief ROS exposure increased action potential duration whereas longer exposure reduced action potential duration. KEY WORDS: Photosensitization; sium; Calcium

Reactive

Oxygen;

Cardiac

Introduction There is increasing evidence that reactive oxygen species (ROS) play a role in reperfusion injury. Accordingly, there is interest in elucidating the effects of ROS at the membrane, cellular, and whole organ level. Since photosensitizers produce ROS when activated by light of appropriate wavelength, they provide a useful tool for investigating ROS effects. Duration and intensity of ROS production by photosensitizers are controlled simply by altering illumination duration and intensity, respectively. Recently, several laboratories have begun using the photosensitizer Rose Bengal (RB) to investigate ROS effects in cardiac muscle. Ver Donck and coworkers have investigated contractures in isolated rat heart cells induced by RB generated ROS (Borgers et al., 1987; Ver

electrophysiology;

+ 11 $03.00/O

currents;

Sodium;

Potas-

Donck et al., 1988). Hearse and coworkers have investigated the induction of arrhythmias in intact rat heart (Hearse et al., 1989; Kusama et al., 1989). We have been investigating the effects of RB generated ROS on the action potential and ionic currents in isolated frog heart cells. We (Tarr and Valenzeno, 1989) reported recently that ROS generated by extracellular RB initially prolong the action potential and reduce its amplitude. This is soon followed by a reduction in duration such that the action potential becomes spike like. These effects occur without alteration in the cell’s resting potential. To understand these action potential modifications, we investigated ROS effects on ionic currents known to play prominent roles in the cardiac action potential. These include excitatory inward currents carried by

Please address all correspondence to: Merrill Tarr, Department of Physiology, College of Health Sciences and Hospital, Kansas City, Kansas 66103, USA Supported by a Grant-in-Aid from the American Heart Association. 0022-2828/91/050639

Ionic

University

Q

of Kansas

1991 Academic

Medical

Center,

Press Limitrd

640

M. Tarr

and D. P. Valenzeno

sodium (I,,) and calcium ions (I,,), and outward currents carried by the potassium delayed rectifier (Ix), and a time-independent current (I,&. This paper reports the effects of ROS generated by extracellular RB and light on the current-voltage (I/V) relationships of each of thesecurrents. Preliminary accounts of ROS effects on the magnitude of these currents have been presented previously (Valenzeno and Tarr, 1988 & 1989).

and Methods Frog atria1 cells were isolated from minced frog atria1 tissue by a method similar to that described previously (Tarr and Trank, 1976). The exception was that trypsin (0.33 mg/ml) alone, rather than a combination of trypsin and collagenase,was used to digest the tissue. Cells were harvested at 30 min intervals by decanting 5 ml of the digestion medium containing isolated cells. Calcium-containing Ringer’s solution (2 ml) was added to this 5 ml of fluid containing the isolated cells. The cells were stored in this manner at room temperature until just prior to experimentation. At that time !O drops of the cell containing medium were added to 3 ml of Ringer’s solution (see below) containing 0.5 PM RB (Aldrich Chem. Co.) in a plastic culture dish on the stage of an inverted microscope. Once ceils were placed in the RB containing medium, all procedures were carried out under very dim room light. A 650 nm bandpassfilter (10 nm half-maximum bandwidth) placed in the light path of the microscope allowed viewing of the cellswithout activating the sensitizer. The cellswere allowed to equilibrate for 6-9 min in the RB containing bathing medium. At that time the opening of the tip of a patch pipette, to be used to record intracellular potential and ionic currents (see below), was touched to the cell surface near the center of the cell. With application of a small negative pressure, a high resistanceseal formed between the pipette and the cell membrane, followed by rupture of the isolated patch of membrane. After patch rupture, an additional 6 min was allowed for exchange of patch pipette contents (e.g., Mg’+ or CAMP) with the cell interior. Rapid exchange was favored by placing the patch pipette near the Materials

center of the cell. The ionic currents stabilized during this 6 min time period. Once ionic current stabilization had occurred, a control I/V relationship was determined. This was accomplished by clamping the membrane potential at various levels from an appropriate holding potential. The duration of the voltage-clamp pulse, aswell as the interpulse interval, varied with the current being investigated. The cell and bathing medium were then illuminated with green light to cause RB to produce ROS. This was accomplished by removing the 650 nm bandpassfilter in the light path for a designated time period. With the filter removed, the cell and surrounding medium were exposed to light from the 100 W tungsten-halogen illuminator passedthrough a 525 nm broad bandpass filter (150 nm half-maximum bandwidth). Illumination intensity was monitored by a photocell placed at the focal plane of a side arm of the microscope; the desired intensity having been previously set by an appropriate set of neutral density filters. The photocell was calibrated against an Eppley thermopile to determine total illumination energy in milliwatts/square centimeter (mW/cm’); illumination intensity of 6.5 mW/cm’ was used in theseexperiments. Illumination with green light was terminated by reintroducing the 650 nm filter in the light path. An I/V relationship after the first illumination period was then determined. The illumination procedure was repeated and an I/V relationship was again determined after a secondillumination. The patch pipettes used in this study had resistancesranging from 2 to 8 MR: the very low resistance pipettes were used in experiments on Z,,. Electrode seriesresistancecompensation was used routinely to reduce series resistanceaslow aspossiblewhile maintaining voltage control stability. Membrane potential and currents were measured between the patch pipette and a reference electrode in the bathing medium. The patch pipette was connected to a Dagan 8900 patch clamp-whole cell clamp system through a Dagan 8910 whole cell probe. Analog signals related to membrane voltage and current were digitized by an analog to digital interface and stored on magnetic disk using a computer based data acquisition system,

ROS

Sensitizer-generated

The composition of both the bathing medium and pipette solution depended on the ionic current under investigation. Nominally, the bathing medium contained 111 rn~ NaCI, 5.4 rn~ KCI, 1.8 rn~ CaCl,, 10 rn~ tris (hydroxymethyl) aminomethane, 4 rn~ glucoseand was titrated to pH 7.3 with HCI. The pipette solution contained nominally 150 rn~ KCI, 10 rn~ tris (hydroxymethyl) aminomethane, and was titrated to pH 7.12 with HCl. The following modifications were made to thesesolutions according to the ionic current under investigation. Inward sodium current (INa): The bathing medium was diluted 25% with deionized water. This caused the cells to swell thereby reducing intracellular longitudinal resistance and dramatically improving voltage control. LaCls (10 PM) was added to block Zc, (Nathan et al., 1988). The pipette solution contained 20 rn~ MgCl, to block outward currents through potassium channels (Tarr et al., 1989). Inward calcium current (Zc-): Tetrodotoxin (0.1 PM) was added to the bathing medium to block ZNa and 30 IXIM tetraethylammonium chloride (TEA) was added to reduce outward potassiumcurrents. The following were added to the pipette solution: 30 rn~ TEA, 0.5 I'tlM ~WJCL, and 3 x 10-O M adenosine 3’ : 5’

cyclic-monophosphate (CAMP). TEA was used to reduce outward potassium currents whereas CAMP was used to enhance and stabilize I,, (Tarr et al., 1986). Delayed reactifier potassium current (Zk): Tetrodotoxin (0.1 PM) and LaCI, (10 PM) were added to the bathing medium to block Z,, and Zc,, respectively. MgCl, (0.5 mu) and CAMP (3 x 10e4~) were added to the pipette soIution to stabilize Ix. Time-independent current (ZuJ: Tetrodotoxin (0.1 PM) and LaCI, (10 PM) were added to the bathing medium to block ZNaand I,,, respectively. In some experiments CsCl (20 mM) was also added to suppressthe inwardly rectifying potassium current (i.e., I,,) (Argibay et al,, 1983). The pipette solution contained 20 rn~ MgCI, to block outward Ix1 and Z, currents (Matsuda et al., 1987; Tarr et al., 1989). Results

Figure 1 presentsin summary form the effects RB-generated ROS have on the ionic currents characterized in this investigation. Figure 1(A) presentsZ,, elicited by depolarizing voltage pulsesto -20 mV from a holding potential of - 100 mV before (0 s) and after 2 s and

(A)

0

(8)

4

8 Time

12

0

16

40

60

Time

LiEI OS 2s

80

600

4

400

-

Leak 20

r

4s

1000

100

(ms)

(D)

800

Potassium

3000

20

(ms)

(Cl 4OcO

641

,

s

IO s

2coL

u-

0

0

12345

Time

(s)

too

! ! ! 200 xx, 400 Time (ms)

FIGURE 1. Effects ofROS on IN, (A), f, (B), IK (C) and Z,uk (D). Cumulative current trace are indicated. Currents are in picoamperes (PA). See text for further

/OS

I 500

light exposures discussion.

associated

with

each

M. Tarr

642

and D. P. Valemzeno

4 s of illumination. Figure l(B) presents I,, elicited by depolarizing voltage pulsesto 0 mV from a -70 mV holding potential; again before and after 2 s and 4 s of illumination. Figure l(C) presents IK elicited by voltage pulsesto +40 mV from a - 70 mV holding potential; again at 0, 2 and 4 s of illumination. Figure 1(D) presentsIleakelicited by depolarizing voltage pulsesto + 40 mV from a - 70 mV holding potential before (0 s) and after 10 s and 20 s of illumination. Illumination periods longer than that required to suppress INe, I,, and I, were required to enhance I,eak. The data presented in Figure 1 illustrate the nature of the total ionic current recorded under the different experimental protocols usedto isolate specific current components, as well as, the overall effect of ROS on each current component. For example, while ROS suppress both INa and I,,, the inactivation kinetics of INa are affected rather dramatically whereas those of I,-, are changed relatively little. In particular, the rate of INainactivation is slowed markedly. Note that prior to illumination INa is almost completely inactivated within 5 ms of depolarizing the cell to -20 mV. By comparison, significant I,, remains at

16 ms after illumination. Such a dramatic effect on the inactivation kinetics of I,-, is not apparent. ROS also suppress IK without affecting its kinetics, but ROS enhance the magnitude of a time-independent current [see Fig. 1(D)]. For lack of a better nomenclature, we have chosen to refer to this timeindependent current as Ileak. It may be a composite of several time-independent currents; the nature of which remain to be clarified. Figure 2 demonstratesthe effect of ROS on the sodium I/V relationship. For this analysis INa was taken as the difference between the peak inward current and the current just prior to termination of each 500 ms depolarizing voltage pulse used to elicit INa. The I/V relationships prior to (0 s) and after 2 s and 4 s of illumination are presented in Figure 2(A). In addition to suppressing INa, ROS shift the sodium I/V relationship on the voltage axis in the depolarizing direction. Thus, as illustrated in Figure 2(B), the more negative the membrane potential the greater is the ROSinduced suppressionof I,,. For example, at -57 mV only 45% of the initial INa remains after 2 s of illumination compared to 74% at

(A)

I -sooo_80

(B)

I -60

I -4’0

I -20

I 0

I 20 Potential

(mV)

FIGURE 2. Effects of ROS on IN.. The sodium I/V relationships prior to (closed square) and after 2 s (open square) and 4 s (triangle) of cumulative light exposure are presented in A. In B, the fraction of initial I,, at each potential after 2 s and 4 s ofcumulative light exposure is presented. Mean values f one SE are presented in each case. In some cases, the SE was too small to illustrate. Data obtained on eleven cells.

Sensitizer-generated - 26 mV and 95% at +2 mV. Although a greater suppression of Z,, occurs at all potentials after 4 s of illumination, the fraction of initial Z,, vs. membrane potential relationship has a similar shape as that obtained with 2 s of illumination. Figure 3 demonstrates the effect of ROS on the calcium I/V relationship. For this analysis, Ica was taken as the peak inward current during each depolarizing voltage pulse. Although the TEA added to both the bathing medium and pipette solution markedly suppressed potassium currents, it did not suppress these sufficiently to allow a subtraction procedure to be used to obtain Zc-. However, ZK is very small early during a depolarizing voltage pulse when Zca is prominent, especially at membrane potentials negative to 0 mV where Z, is not markedly activated (see below). Thus, the calcium I/V relationship is probably not affected by Z, contamination over the voltage range of -40 to about +20 mV. However, separation of Zca from ZK does become more serious at potentials positive to about +20 mV. Nevertheless, it is obvious that the ROS-induced suppression of Zc, is not accompanied by a voltage shift of the I/V relationship. As illustrated in Figure 3(B) for

ROS

both illumination times the fractional block of Z is nearly constant over the range of membcrane potentials tested. Figure 4 demonstrates the effect of ROS on the potassium I/V relationship. For this analysis, Z, was taken as the difference between the outward current just prior to termination of‘ each 5 s depolarizing voltage pulse and the minimum outward current during the pulsr. ROS suppression of Z, also occurs without a shift in the I/V relationship on the voltageaxis. As illustrated in Figure 4(B), the fraction of initial Z, after 2 s and 4 s of illumination is constant over the range of membrane potentials tested. Figure 5 presents I/V relationships relative to zleaL prior to and following 10 s and 20 s of illumination. For this analysis, the current just prior to termination of each voltage pulse was used as a measure of ZlceL. Figure 5 (A) presents I/V relationships in the presence of ZKI. In Figure 5(B), Z,, is blocked with 20 mM extracellular cesium. In the presence of Z,, [Fig. 5(A)], outward current is enhanced by 10 s of illumination whereas inward current is not affected. However, both inward and outward currents are enhanced after 20 s of illumination. In contrast, in the absence of Z,, [Fig.

(B)

(A)

1.0

0

0.8 t -1000

3

2 ,Q E k c:

.c=

0.6

-

i

,A-f+

t

-2ooc

0.2

-3cOc

I 3-40

I -20

I 0

I 20

L-J o-0

I 40

-40

Potent101 FIGURE 3. Effect ofROS and 4 s (triangle) of cumulative 2 s and 4 s ofcumulative light SE was too small to illustrate.

-

-20

0

20

40

(mV)

on I,,. The calcium I/V relationships prior to (closed square) and after 2 s (open squarei light exposure are presented in A. In B, the fraction of initial I,, at each potential after exposure is presented. Mean values f one SE are presented in each case. In some cases. the Data obtained on seven cells.

M. Tax-r and D. P. valennulo

644 (A) 4ooor---l

0.8

3000

cr ..; 0.6 B .-6 5 0.4 E

2 ,a 5 2000 k 3 1000

0 -50

0.2 -

0

50

0.0 ' -200

I Potential

' ' ' ' 20 40 60 80 I

10

(mV)

FIGURE 4. Effect of ROS on I,. The potassium I/V relationships prior to (closed square) and after 2 s (open square) and 4 s (triangle) of cumulative light exposure are presented in A. In B, the fraction of initial I, at each potential after 2 s and 4 s of cumulative light exposure is presented. Mean values f one SE are presented in each case. In some cases, the SE was too small to illustrate. Data obtained on ten cells.

5(B)] both inward and outward currents are enhanced by 10 s as well as 20 s of illumination. The apparent lack of effect on inward current with 10 s illumination in the presence of Z,, may be related to an ROS-induced suppression of Zx, occurring concomitantly

with enhancement of another current. There has been one preliminary report that ROS generated by xanthine (X) plus xanthine oxidase (X0) reduce Ix, in guinea-pig heart cells (Coetzee and Opie, 1988). Other results also suggest an ROS-induced suppression of Z,,.

(8) 800 ,-

(A)

600

-500 ‘i’ ’ ’ ’ -1000 -120 -80 -40

0

-6OO-120 -80 -40

40 Potential

0

40

I

3

(mV)

FIGURE 5. Effects of ROS on &,s in the presence (A) and absence (B) of 1,t. The I/V relationships present total currents prior to (closed square) and after 10 s (open square) and 20 s (triangle) of cumulative light exposure. Mean values + one SE are presented in each case. In some cases the SE was too small to illustrate. Data obtained on ten cells in A and eight cells in B.

645 (A)

(8) Duratton

I 0123456

0.0 Time

I 0

I I 123456

I

I

I

I

(mln)

FIGURE 6. Effects of ROS on action potential amplitude and duration. Data obtained on individual cells exposed to 4, 10 or 12 s of light as indicated. In A, the fraction of the initial action potential amplitude (peak minus resting potential) as a function of time is presented. In B, the fraction of the initial action potential duration (time from stimulus to repolarization within IO mV of resting potential) as a function of time is presented. In both cases, zero time indicates the time of illumination.

We routinely monitored inward current at potentials negative to -80 mV in the Zk experiments discussed previously. In those experiments, briefer illuminations (4 s) did reduce the inward current at - 110 mV by about 14%. We reported previously (Tarr and Valenzeno, 1989) that continuous ROS production initially causesthe action potential duration to increase but this is soon followed by a reduction in duration such that the action potential becomes spike like. Considerably lower RB concentration (0.125 PM) and illumination intensity (1.5 mW/cm2) were usedin those experiments than in the present experiments (0.5pM RB and 6.5 mW/cm2), and continuous rather than intermittent (2 to 20 s) illumination was used to produce ROS. To allow a better comparison of ROS effects on the various I/V relationships to ROS-induced modifications of the action potential, we assessedthe action potential modifications using the same RB concentration and illumination protocols used in the I/V determinations. Figure 6 compares the action potential modifications produced by 4 s of illumination to those following 10 and 12 s of illumination. Figure 6(A) presents the effects on action potential amplitude and Figure 6(B) presents

the effects on action potential duration. Although data from only 3 cellsare presented, they are representative of data obtained on six cells with 4 s of illumination and seven cells with illuminations of 1O-l 2 s. Clearly 1O-l 2 s of illumination produced more reduction in action potential amplitude than 4 s of illumination. Brief illumination (4 s) produced consistent effects on the action potential duration in that action potential duration increased and stayed prolonged, albeit somerecovery of duration did occur. In contrast, longer illumination (10-12 s) produced variable initial effects on action potential duration. In some cells there was an initial prolongation in duration (e.g., 10 s data asshown), whereasin other cells there was an initial reduction in action potential duration (e.g., 12 s data as shown). In both cases,however, the action potential duration declined rapidly and the action potential duration stayed well below the control value. Although data are not presented, the cell’s resting potential was not affected by either 4 s or 10-12 s of illumination. Discussion

The data presented in this paper illustrate the effects of ROS generated by extracellular RB

646

M. Tarr

and D. P. Valenzeno

on the I/V relationships of ionic currents which play dominant roles in the generation of the frog cardiac action potential. Our results illustrate that ROS suppress INa, Zc, and Zk but enhance IteaL. The ROS-induced suppression of I,, is accompanied by a voltage shift of its I/V relationship. This is not the case for the other currents. Discussions regarding photosensitizer-generated ROS effects on specific ionic currents and the action potential, as well as, the relevance of our results to ROSinduced effects in ischemia-reperfusion injury are presented below. Sodium current ROS suppress Z,, and shift its I/V relationship in the depolarizing direction. This I/V shift results in a greater current suppression of Z,, by ROS at more negative membrane potentials. It is well recognized that caution must be used in interpreting apparent voltage shifts in I/V relationships involving large, rapid inward currents such as Z,,. Loss of voltage control can produce an artifactual steepening of the negative conductance limb of the I/V relationship (see Beeler and McGuigan, 1978). Current reduction may, therefore, improve voltage control producing an apparent shift of the I/V relationship in the depolarizing direction. To rule out this possibility, we compared the effects of ROS and partial inactivation on the sodium I/V relationship. Inactivation which produced greater than 50% suppression of Z,, did not shift the I/V relationship. In contrast, a similar ROS-induced suppression of Z,, in the same cell produced at least a 20 mV shift in the I/V relationship. Thus, the ROS-induced shift in the sodium I/ V relationship is not an artifact related to voltage control. Similar shifts in the sodium I/ V relationship occur in nerve axons photomodified by acridine orange (Pooler, 1968) or Eosin Y (Oxford et al., 1977). ROS also alter the kinetics of INa. The most obvious effect is a slowing of the rate of inactivation of Z,, [Fig. l(A)]. The delay in inactivation cannot be due to an ROSinduced reduction in outward potassium current (i.e., Zk), since Zx was blocked with high internal Mg *+ . It should be noted that an ROS-induced slowing of inactivation of ZNa was obvious only at potentials positive to - 40

mV. A slowing of Z,, inactivation also occurs in nerve axons photomodified by acridine orange (Pooler, 1968) or Eosin Y (Oxford et al., 1977). Calcium current ROS also suppress Zc, but do not shift its I/V relationship. In addition, the rate of inactivation of Zca is not markedly affected as is the case for Z,,. These results are similar to those reported recently by Goldhaber et al. (1989) demonstrating that oxygen free radical generating systems such as X + X0 and/or hydrogen peroxide (H202) reduce Zc, in single guinea pig ventricular myocytes. They also reported that H,02 does not affect Zc, kinetics or the voltage dependency of Zc,. Potassium current Suppression of Zk by ROS occurs without apparent alteration in Zx kinetics. Also, with 224 s periods of illumination, there is no shift in the I/V relationship. However, preliminary experiments indicate that illuminations longer than 4 s produce a shift in the potassium I/V relationship toward more positive membrane potentials. Further investigation of this result is warranted. Leak current It is clear that ROS enhance a timeindependent current. Recently, Goldhaber et al. (1989) reported that free radical generating systems such as X + X0 and H,O, enhance a time-independent current which they identified as the ATP-sensitive potassium also current (ZK(ATP-~~~~)). 1s our Lk ZK(ArP-scns)? The following observations suggest that it is not. First, in the presence of extracellular cesium the IteaL reversal potential of - 40 mV [Fig. 5(B)] is positive to that expected for ZK(ATP-sens)* Second, the increase in Ztealr is not affected by the level of intracellular Mg* + . We observed similar increases in Zrcal with 0.5 mM and 20 mM MgCl, in the pipette. Horie et al. (1987) reported that increasing intracellular Mg * + from 0.5 to 10 mM markedly suppressed I K(Arp-scns). Third, in preliminary experiments we have found a ROS-induced enhancement cesium of Lk in the presence of extracellular

Sensitizer-generated

(20 mM) or internal cesium (150 mM) alone or in combination. However, cesium ion is a known blocker of ZK(ArP-scnsj (Quayle et al., 1987; Arena and Kass, 1989). Fourth, in preliminary experiments we found that during long ROS exposures the reversal potential of Ileekgradually shifts towards zero mV. While, our results do not clarify the ionic nature of / leak> they do indicate it is not ZK(ArP-scns,. Action potential

There is reasonable agreement between the ROS-induced changesin the I/V relationships and the action potential modifications we observed. For example, brief (2 to 4 s illuminations) ROS exposure primarily affects ZNa, Zc, and Zk but has little effect on Ztcak.Accordingly, the alterations in the action potential amplitude and duration following brief ROS exposure reflect ROS-induced changes in Zc, and Zk. For example, 4 s of illumination produces a slight decreasein action potential amplitude most Iikely related to a reduced Zc-, and also produces a significant prolongation of the action potential most likely related to a suppressionof Zk. Longer periods of illumination ( IO- 12 s), produce a greater reduction in action potential amplitude as a result of further suppressionof I,,, but produce variable initial effects on action potential duration. Some cells showed an initial prolongation followed by a rapid reduction in duration. Other cells showed only a reduction in duration. The variability in action potential duration with longer illumination is probably related to counteracting effects of a reduced Zk and enhanced Zieak. Relevance of photomodijication

results

While there is increasing evidence that ROS play a role in reperfusion-induced arrhythmias, there is presently limited information regarding the effects of ROS on cardiac ionic currents and the cardiac action potential. As stated previously, since photosensitizers produce ROS when activated by light of appropriate wavelength, they provide a tool for investigating ROS effects. Photosensitizers offer the advantage of precise and rapid control over the timing of ROS production. However, the advantages afforded by

ROS

647

photosensitizer-generated ROS are of little use if these ROS effects are not representative of those resulting from ROS generated by cellular mehanisms. It is encouraging that our results are similar to those reported by other investigators using a variety of free radical generating systems including X f X0, H,O,, and DHF (dihydroxyfumaric acid, a superoxide generator). Reductions in I,-, and action potential amplitude have been reported (Goldhaber et al., 1989) for both X + X0 and H202, but increasesin Zc, and action potential amplitude have also been reported (Coetzee and Opie, 1988; Barrington et al., 1988) for X0 and DHF. Effects of ROS on ZK, other than our results, have not been reported. But increases in action potential duration have been reported (Barrington el al., 1988; Coetzee and Opie, 1988; Firek and Beresewicz, 1989; Hayashi et al., 1989) for X + X0, HzO, and DHF asmight be expected if these agents also reduce Zk. It should be noted, however, that Goldhaber et al., (1989) observed that X0 and H,Oz produced onIy a decrease in action potential duration. There have been no other reports of ROS effects on ZNain cardiac tissue although it has been reported (Pallandi et al., 1987) that X + X0 reduces the maximum rate of depolarization of the action potential, a measure of ZN,. As discussedpreviously, Goldhaber et al., ( 1989) reported that both X + X0 and H,Oz a time-independent curenhance IK(ATP-~~~~J, rent. We also observed an ROS-induced increasein a time-independent current but we do not think it is ZK(ArP-sensr Lastly, our results suggestthat ROS may suppressZki and there has been one preliminary report that XOgenerated ROS also reduce Zk, (Coetzee and Opie, 1988). Thus, while there doesappear to be variability in the ROS-induced electrophysiological modifications in cardiac tissue, this variability is not related simply to the method of ROS generation (i.e., photosensitizer, X + X0, H,O,, DHF). In fact there are many striking similarities between photosensitizergenerated ROS effects and those related to ROS production by H,Oz and/or xanthine plus xanthine oxidase. Our present and previously published results illustrate an important aspect regarding ROS-induced action potential modifications. The net result dependson both thr level

M. Tarr

648

and D. P. Valemzeno

and duration of ROS exposure. For example, we reported previously (Tarr and Valenzeno, 1989) that 30 s of illumination in the presence of 0.125 PM extracellular RB produced a sustained increase in action potential duration. Much briefer (2-4 s) illuminations at higher intensity produced this result with 0.5 PM RB. We also reported previously that a reduction in action potential duration occurred after about 3 min of illumination with 0.125 PM RB. Again, briefer (lo-12 s), higher intensity illuminations produced this result with 0.5 PM RB. Our results relative to the effects of intensity and duration of ROS exposure on action potential modifications are similar to those reported recently by Firek and Beresewicz ( 1989). These investigators examined the effect of HzOz on the action potential of guinea-pig ventricular muscle. They reported that 0.6 mM HzO, produced initially a slow increase in action potential duration which was followed after about 14-16 min by a rapid decrease in duration. This sequence of changes was accelerated by increasing H,O, concentration (3 mM to 18 mM). Hayashi et al. (1989) also reported a similar sequence (i.e., increased followed by decrease action potential duration) in guinea-pig papillary muscle exposed to 10 mM HzO,. Thus, ROS may produce either an increase or decrease in action potential duration depending on the intensity and/or duration of tissue exposure to

ROS. Conceivably, the dependency of action potential modification on level and intensity of ROS exposure could contribute to inhomogeneities in action potential waveform in reperfused cardiac tissue. Regions exposed to high levels of ROS may have shortened action potentials compared to regions exposed to lower levels of ROS. Such action potential during inhomogeneities may contribute to the re-entrant arrhythmias which occur upon reperfusion of the ischemic myocardium. Added note: During the review of our manuscript a paper appeared reporting the effects of tert-butyl hydroperoxide (t-BHP), an initiator of free radical chain reactions, on ZNa, I,,, and Z,, (Bhatnagar et al., Circ Res 67: 535-549, 1990). In some regards, the effects of t-BHP are similar to our data in that t-BHP suppressed ZNa and slowed its inactivation. However, t-BHP did not shift the sodium I/V relationship. It also did not suppress either Zc, nonor b, nor did it produce large outward specific leakage currents as we observed.

Acknowledgements The authors thank Joanne Tarr for her expert technical assistance. We also thank Dr Steve Crockett, who provided the equipment and expertise to perform the measurements of spectral intensity of the light source.

References ARENA

JP, KASS RS (1989)

Enhancement

of potassium-sensitive

current

in heart

cells by pinacidil.

Circ

Res 65:

436-445. JA, DUTEY P, ILDEFONSE M, OJEDA C, ROUGIER 0, TOURNEUR Y (1983) Block by Cs of K current i,, and of induced K current i,,, in frog atrium. Pfliigers Arch 597: ‘295299. BARRINGTON PL, MEIER CF Jr, WEGLICKI WB (1988) Abnormal electrical activity induced by free radical generating systems in isolated cardiocytes. J Mol Cell Cardiol 20: 1163-l 178. BEELER GW, MCGUIGAN JAS (1978) Voltage clamping of multicellular myocardial preparations: capabilities and limitations of existing methods. Prog Biophys Mol Biol 31: 219-254. BURGERS M, VER DONCK L, VANDEPLASSCKE G (1987) Pathophysiology of cardiomyocytes. Ann NY Acad Sci 522: 433-453. COETZEE WA, OPIE LH (1988) Electrophysiological effects of free oxygen radicals on guinea pig ventricular myocytes (abstr.) J Mol Cell Cardiol20 (Suppl V): S.17. FIREK L, BERESEWICZ A (1989) Electrophysiological effects of HsO, on guinea-pig ventricular muscle. Permissive role of iron (abstr). J Mol Cell Cardiol 21 (Suppl IV): S.39. GOLDHABER JI, SCOTT SJ, LAMP ST, WEISS JN (1989) Effects of exogenous free radicals on electromechanical function and metabolism in isolated rabbit and guinea pig ventricle; implications for ischemia and reperfusion injury. J Clin Invest 83: 180&1809. HAYAS~I H, MIYATA H, WATANABE H, KOBAYASHI A, YAMAZAKI N (1989) Effects of hydrogen peroxide on action potentials and intracellular Ca*+ concentration of guinea-pig heart. Cardiovasc Res 23: 767-773. ARGIBAY

carbachol

Sensitizer-generated

ROS

649

DJ, KUSAMA Y, BERNIER M (1989) Rapid electrophysiological changes leading to arrhythmias in the aerobic rat photosensitization studies with Rose Bengal-derived reactive oxygen intermediates. Circ Res 65: 146153. HORIE M, IRISAWA H, NOMA A (1987) Voltage-dependent magnesium block of adenosine-triphosphate-sensitive potassium channel in guinea-pig ventricular cells. J Physiol (Lond) 381: 251-272. KUSAMA Y, BERNIER M, HEARSE DJ (1989) Singlet oxygen-induced arrhythmias: dose-response and light-response studies for photoactivation of Rose Bengal in rat heart. Circulation 80: 1432-1448. MATSUDA H, SAIGUSA A, IRISAWA H (1987) Ohmic conductance through the inwardly rectifying K channel and blocking by internal Mgs+. Nature (Lond) 325: 156159. NATHAN RD, KANAI K, CLARK RB, GILES W (1988) Selective block of calcium current by lanthanum in single bullfrog atria1 cells. J Gen Physiol 4: 549572. OXFORD GS, POOLER JP, NARAHASHI T (1977) Internal and external application ofphotodynamic sensitizers on squid giant axons. J Memb Biol36: 159173. PALLANDI RT, PERRY MA, CAMPBELL TJ (1987) Proarrhythmic effects of an oxygen-derived free radical generating system on action potentials recorded from guinea pig ventricular myocardium: A possible cause of reperfusion-induced arrhythmias. Circ Res 61: 56-54. POOLER J (1968) Light-induced changes in dye-treated lobster giant axons. Biophys J 8: 100910’26. QUAYLE J, STANDEN N, STANFIELD P (1987) The voltage-dependent block of ATP-sensitive potassium channels of frog skeletal muscle by cesium and barium ions. J Physiol (Lond) 382: 213-236. TARR M, TRANK JW (1976) Preparation of isolated single cardiac cells from adult frog atrial tissue. Exprrientia 32: 338-339. TARR M. TRANK JW, GOERTZ KK (1986) Voltage-tension relations in single frog atria1 cardiac cells. Circ Res 59: 447-455. TARR M, TRANK JW, GOERTZ KK (1989) Intracellular magnesium affects Ix in single frog atria1 cells. Am J Physiol 257 (Heart Circ Physiol 26): H1663-H1669. TARR M, VALENZENO DP (1989) Modification ofcardiac action potential by photosensitizer-generated reactive oxygen. J Mel Cell Cardiol 21: 539543. VALENZENO DP, TARR M (1988) Photomodification ofcell membranes: ionic currents in single frog atria1 cells iabstr.1 Photochem Photobiol47: 76s. VALENZENO DP, TARR M (1989) Photomodifications of cell membranes: sodium currents in single frog atria1 cells (abstr). Photochem Photobiol49 69s. VER DONCK L, VAN REEMPT~ J, VANDEPLASSCHE G, BURGERS M (1988) A new method to study activated oxygen sprcics induced damage in cardiomyocytes and protection by Ca2+ -antagonists. J Mel Cell Cardiol 20: 81 I-823. HEARSE

heart: