Mechanism of the increase in intracellular sodium during metabolic inhibition: Direct evidence against mediation by voltage-dependent sodium channels

J Mol

Cell

Cardiol

24, 1307-1320

Mechanism Metabolic

(1992)

of the Increase Inhibition: Direct Voltage-dependent Rafael

Division of Cardiolou,

in Intracellular Sodium During Evidence Against Mediation by Sodium Channels

Mejia-Alvarez

Department

and

Eduardo

Marban

of Medicine, The Johns Hopkins University School of Medicine, Baltimore, MD 21205, USA

(Received 13 April

1992, accepted in revised form 3 June 1992)

R. MEJ~A-ALVAREZ AND E. MARBAN. Mechanism of the Increase in Intracellular Sodium During Metabolic Inhibition: Direct Evidence Against Mediation by Voltage-dependent Sodium Channels. Journal of Molecular and Cellular Cardiology (1992) 24, 1307-1320. During ischemia or metabolic inhibition, intracellular Nat conccntration ([Na+],) increases considerably. Elevation of [Na+], figures critically in the mechanism of cellular injury by promoting Car’ influx via the Na+-Cap+ exchanger, but the exact mechanism of this intracellular Na’ accumulation remains unknown. To test directly the hypothesis that voltage-dependent Nat channels are involved, we measured Nat currents (c) in isolated guinea-pig ventricular myocytes using the patch-clamp technique. The cell-attached configuration was used in order to avoid disturbing the intracellular milieu. Metabolic inhibition was induced by exposing the cells to either iodoacetate (IAA, I mM) to inhibit glycolysis or 2,4-dinitrophenol (DNP, 0.2 mM) to uncouple oxidative phosphorylation. The amplitude ofIN,, was measured in multichannel patches before and during exposure to IAA or DNP, by depolarizing the cell to different membrane potentials from a holding potential of - 135 mV. Analysis of current-voltage relations before and during metabolic inhibition revealed a modest but significant reduction of peak INa at test potentials positive to - 40 mV with DNP; no change was observed with IAA. The voltage dependence of steady-state parameters of inactivation was not altered by either intervention; specifically, no steady-state (“window”) current was induced. Although we cannot exclude the possibility that other factors not explored here might lead to different conclusions during genuine &hernia, metabolic inhibition alone does not up-regulate the function of Na’ channels. Thus, we conclude that other mechanisms underlie the accumulation of intracellular Na’ observed during metabolic inhibition. KEY WORDS: Sodium

current;

Ventricular

myocytes;

Introduction Metabolic inhibition produces marked alterations of intracellular cation homeostasis, which are just as marked or even greater in ischemia. Intracellular sodium and calcium increase significantly during metabolic inhibition or during ischemia; the accumulation of intracellular calcium figures prominently in the pathogenesis of the associated cellular injury [I]. The increase in cell calcium comes about at least partially due to Ca2+ influx via Na+-Ca2+ exchange [2, 31, driven by the concomitant increase of intracellular Naf concentration ( [Na+li) [m. Although the

Patch-clamp;

Glycolysis;

1 I 1307 + 14 008.00/O

phosphorylation.

exact mechanism of the [Na+li accumulation is still unknown, two general possibilities need to be considered: decreased Na+ efflux or increased Na+ influx. During ischemia, ATP production and the free energy of ATP hydrolysis are diminished, so that decreased Na’ efflux might result simply from Na’/K+ ATPase inhibition. This mechanism seems unlikely to be primary in hypoxia or metabolic inhibition since the time course of ATP depletion lags behind that of [Na+], accumulation [A; furthermore, there is no simple correlation between ATP and [Na+]; in either metabolic inhibition or ischemia [5, 6, 81. i2s

Please address all correspondence to: E. Marban, Division of Cardiology, University, 720 N. Rutland Avenue, Baltimore, MD 21205, USA. ‘This work was performed with the support of NIH grant no. HL 44065. 0022-2828/92/

Oxidative

844 Ross Building,

Ql992

The Johns

Academic

Hopkins

Press Limited

1308

R. Mejia-Alvarez

regards an increased Na+ influx mechanism, a fall in intracellular pH ( pHi) might drive Na+ into the cells via Na’-H’ exchange [2, 7, 9, 201. Experiments with organic Na’-H+ exchange blockers support the notion that this exchanger attempts to neutralize intracellular acidosis by extruding protons from the cell at the expense of the normal Na+ gradient, both physiologically [II, 121 and during ischemia [IO] or hypoxia [9]. Despite the attractive features of Na+-H+ exchange as a possible pathway for Na + influx, the evidence for its involvement rests largely upon the effects of high concentrations of amiloride, which is a non-specific blocker [1.%15]. An alternative mechanism that has been put forth to explain an increase of Na+ influx is modification of voltage-dependent Na’ channel activity. The opening ofjust one Na+ channel can admit more than 10 million Na+ ions per s into the cytoplasm, and the density of these channels is known to be extremely high in heart cells [lq. Indeed, several pieces of evidence suggest the involvement of NaC channels: lidocaine, a Na+ channel blocker, inhibits the increase in [Na+]; during ischemia [18]; likewise, R-56865, a putative blocker of voltage-dependent Na+ channels, prevents the rise in [Na’li and the cellular damage produced by hypoxia in ventricular myocytes [19]; tetrodotoxin has been found to protect against neural injury provoked by hypoxia and reoxygenation [20]; finally, lysophosphatidylcholine, a toxin that reportedly accumulates in ischemia, increases dramatically the open probability (PO) of cardiac Na+ channels [21]. Nevertheless, the gradual depolarization that occurs during ischemia would tend to inactivate Na’ channels unless their gating properties are altered to favor steady-state current. A provocative precedent for such modification can be found in the work of Bhatnagar e! al. [22], who have demonstrated that the steady-state kinetics of Naf channels are modified during free radical-induced oxidative stress in frog myocardial cells. Under these circumstances, the noninactivating Na+ “window” current increases 12-fold in amplitude, and its peak is shifted 10 mV towards more negative potentials. Such modifications in Z,* would be expected to allow persistent Na+ entry into the cell at the resting potential during oxida-

and

E. Marban

tive stress. In principle, a similar mechanism could explain the increase in Na+ influx during or after metabolic inhibition, particularly since oxidative stress has been implicated as a component of reoxygenation-induced injury

v31. To investigate directly the idea that voltage-dependent Na+ channels are modified such that they might contribute to the increase in [Nat],, we studied the effects of metabolic inhibition on Z,, using the patchOur results indicate that clamp technique. during metabolic inhibition induced by exposing the cells to 2,4-dinitrophenol (DNP) or iodoacetic acid (IAA), ZNa is slightly reduced or unaffected. By exclusion, this observation provides an indirect argument in favor of Na+-H+ exchange (or another coupled transport mechanism) to explain the pathophysiologically important increase in [Na’],.

Methods Cell isolation Male guinea pigs weighing approximately 250 g were anesthetized by intraperitoneal injection of Na-pentobarbital (97 mg), after which hearts were excised rapidly. Single ventricular myocytes were isolated by a variant of the enzymatic dissociation technique described by Mitra and Morad [24]. Briefly, the heart was initially perfused with normal Tyrode solution (in mM, NaCl 140, KC1 5, MgCl, 1, CaCI, 1, glucose 10, HEPES 10, pH= 7.4 adjusted with NaOH) for 3-4 min. After the heart fully recovered its contractility, Ca ‘+-free Tyrode solution was perfused for about 6 min. After 7 min of enzymatic digestion (collagenase type I, 262.5 pug/ml; protease type XIV, 57.5,uglml; both from Sigma, St. Louis, MO, USA) the heart was rinsed with low Ca’+ (0.2 mM) Tyrode solution. Finally, the ventricular portion of the heart was placed in high K+ solution (in mM, KC1 20, K-glutamate 120, EGTA 0.1, glucose 10, HEPES 10, pH= 7.4 adjusted with KOH) and cut into small fragments. The cells were dispersed by gentle agitation and stored at room temperature. Until storage, the entire isolation process was carried out with O,saturated solutions at 37°C.

Effects

Electrical

of Metabolic

Inhibition

recordings

To study INa, the isolated ventricular myocardial cells were allowed to settle to the bottom of a small experimental chamber and continuously superfused with high K’ solution at a rate of l-4 ml/min. The cell-attached configuration [,%I was chosen for two reasons: first, to avoid dialysis of the intracellular milieu, which might bias the cellular response to metabolic inhibition; second, to enable the simultaneous evaluation of gating and unitary permeation properties (see Results). All experiments were performed within 4-6 h of cell isolation at room temperature (21°C). Patch-clamp pipettes were made of borosilicate glass by a Flaming-Brown programmable pipette puller (model P-87, Sutter Instruments Company, San Francisco, CA, USA), coated with Sylgard (Dow Corning Corporation, Midland, MI, USA) and firepolished. In order to increase the number of Na’ channels per patch to enable assessment of the “macroscopic” behavior of INa, fairly large pipettes (0.5-l MQ resistance) were favored. Ag/AgCl electrodes were used to connect electrically the pipette and the bath solution. An Axopatch-1D amplifier (Axon Instruments, Foster City, CA, USA) with an IH-1 integrating headstage was used. Currents were digitized at 10 kHz with a 12 bit A/D converter (model TL- 1 DMA Labmaster, Axon Instruments, Foster City, CA, USA) and lowpass filtered at 2 kHz (4-pole Bessel). Pipette capacitance was compensated by injection of approximately 5 pF of capacitive current. The cells were stimulated at 0.3 Hz by 70 ms depolarizing pulses to different membrane potentials from a holding potential (HP) of - 135 mV (unless otherwise indicated). The data were collected and stored for further analysis in an IBM-compatible personal computer using custom software. Electrical recordings were first collected 5-10 min after seal formation. This delay was sufficient for stabilization of the well-recognized shift of steady-state activation and inactivation to more negative potentials [16].

Data analysis To eliminate tial functions

capacitative were fitted

transients exponento the baseline and

on Na’

Channels

1309

subtracted from the single-channel records. Best-fit curves were calculated by a nonlinear, least squares method (Levenberg-Marquardt algorithm). Averaged data are expressed as mean f S.E.M. Statistical significance was assessed using paired samples I test and Wilcoxon test [Zq. Solutions and chemicals The bath solution used during the electrical recordings contained (in mM): KC1 20, Kglutamate 120, MgCl, 1, dextrose IO, EGTA 1, HEPES 10, pH= 7.4, adjusted with KOH. The isotonic K’ served to zero the cell membrane potential and thus enabled explicit quantification of the trans-patch potential. To record Na+ channels the pipette solution consisted of (in mM): NaCl200, BaCl, 1, CaCl, 2, HEPES 5, pH= 7.4, adjusted with NaOH; the high [Na’] increased the signal to noise ratio, while Ba2+ suppressed K+ current (notably I KATP [301).For K+ channel recordings the pipette contained (in mM): KC1 140, CaCl, 1, HEPES 10, pH= 7.4, adjusted with KOH. MgATP, IA.4 and DNP were purchased from Sigma.

Results

Metabolic

inhibition

induced hv Il.NP

or IA.4

In order to induce metabolic inhibition, we blocked either glycolysis or oxidative phosphorylation. Glycolysis was inhibited using IAA, which impairs the production of 1,3bisphosphoglycerate by inactivating the enzyme glyceraldehyde 3-phosphate dehydrogenase [27]. Oxidative phosphorylation was uncoupled by DNP [28] which dissipates the proton gradient across the inner membrane of the mitochondria [29]. To confirm the effectiveness of these two drugs as metabolic inhibitors under our experimental conditions, we monitored the activity of Z,..,, 1301. This channel provides a convenient bioassay for the subsarcolemmal ATP concentration, since its PO increases dramatically when [ATP] falls to micromolar levels [.%I. The result of such experiments is shown in Figure 1. In the control, patches were generally either devoid of activity or contained only the inward rectifier Zk,. With a variable latency after addition

1310

R. Mejia-Alvarez

(a)

and

E. Marban

Control

DNP

(b)

200

Membrane -100

-75

IAA I rw

PM

potential

(mV)

-50

-10 I K,ATP

Current

(PA)

of either 0.2 mM DNP [Fig. 1 (a), left] or 1 mM IAA [Fig. l(a), right] to the bath solution, a new type of single-channel activity became evident. This effect was observed 7-23 min (13.6 f 3.1 min; n = 5) after addition of either metabolic inhibitor. The new type of channel activity was identified as IK,ATP based on the observations that: (a) activity was completely blocked by 1 mM ATP (upon excision into the inside-out configuration, data not shown), and (b) the single-channel conductance equalled 95 pS [ 140 mM KC1 in the pipette solution, Fig. 1 (b)].

FIGURE 1. ATP sensitive K’ channel activity as an assessment of the efficacy of metabolic inhibition. Singlechannel activity was recorded in the cell-attached configuration, using 120 ms depolarizing pulses to -80 mV from - 100 mV of HP. (a) No single-channel activity was recorded in control conditions. After 20 min of exposure to 200 PM DNP in the extracellular solution (bottom, left) a new type of single-channel activity became evident. Single-channel opening events are displayed as downward deflections. The same phenomenon was observed when 1 rnM IAA was added to the extracellular solution (right). (b) The single-channel conductance calculated from the unitary current-voltage relation was 95 pS (140 rnM KC1 in the pipette solution). Filled circles reprrsent pooled data from five experiments. No single-channel openings were resolved at positive potentials.

The consistent activation of Z,,,,, within the first 23 min of exposure to either IAA or DNP provides strong evidence that Na+ channels would be exposed to similarly severe metabolic rundown in the subsarcolemmal region adjacent to the ionic channels within the time frame of our experiments. Efects of metabolic inhibition

on I,

amplitude

Figure 2(a) shows records of Na+ channel activity obtained in the cell-attached configuration from a ventricular myocyte.

Effects

of Metabolic

Inhibition

on Na’

Channels DNP

(a )

-

I

WM

_

-4OmV

-20

200

mV

5 PA L Elms

(b)

-120

Membrane

potentlal

-80

-40 +--..

m-0

Control

o-.-o

DNP 200 PM

(mV)

0

Current

40

+

(PA)

FIGCRL 2. Efiect of mrtaholic inhibition induced by DNP on Ix,,. (a) Na’ channel current rtwrdinqs nbtainvd with the cell-attached configuration. G~rrcnt~ ww viicited by 70 ms drpolarizin,q pulses to dilfcrcnt mcmbranc potentials (indicated hetwwn columns) from a HP 01 - 135 mV. The left hand column shows singlr-channl-I activity in control conditions. ‘Thr riqht hand , olnmrr shows the effect of 200~~ DNP. ihi 1-L’ cur\c
1312

R. Mejia-Alvarez

The currents were elicited from a HP of - 135 mV by 70 ms depolarizing pulses to different potentials. A lower limit for the number of Na+ channels per patch was derived by dividing the peak current amplitude, measured at membrane potentials where PO was maximal, by the single-channel amplitude at that particular potential. The number of channels per patch estimated in this way ranged between 10 and 300. The large number of channels present in each patch allowed us to study the macroscopic properties of ZNa without disturbing the intracellular milieu and without the need for signal averaging. The macroscopic properties of IN, are clearly displayed in Figure 2(a): the currents activate promptly upon depolarization, but eventually inactivate completely at all potentials. Although a rare non-inactivating mode of Na+ channel gating has been reported [17], such activity was absent under our experimental conditions. Interestingly, there was no obvious difference in any of these properties after addition of0.2 mM DNP [Fig. 2(a), right column] to the bath solution, even with 16 3 1 min of exposure to the drug (24.5 & 10 min, n= 5). Analysis of pooled data for the current-voltage relations [Fig. 2(b), control data are in filled circles] revealed a slight but significant reduction of the peak amplitude during exposure to DNP [Fig. 2(b), open circles] at test potentials between -30 and 10 mV (PcO.05, n=5). In two experiments, the effects of removal of metabolic inhibition were studied by washing out the DNP (for 15 or 26 min). In neither case did ZNa recover to control levels, nor did any kinetic modifications appear. The fact that currents did not recover even after washing out the DNP argues against a specific blocking effect of DNP. To estimate the voltage dependence of Z,, activation the following expression was used to fit [Fig. 2(b), smooth lines] the experimental I-V data:

and

E. Marban

where I’,,, is the activation mid-point; V, the membrane potential; K,,,, the activation steepNat permeabness factor; PNa,the maximum ility of the patch; [Na+],,, extracellular Na+ and z, concentration; ENa,reversal potential; R, and T have the usual thermodynamic meaning. This expression is the combination of two functions, the Boltzmann distribution for steady-state activation (first term) and the Goldman-Hodgkin-Katz constant field equation (second and third terms). The only parameter which decreased following metabolic inhibition induced by DNP was PNa(seeFig. 2 legend), indicating that the voltage dependence of the activation process of Na+ channels is not affected by metabolic inhibition. In order to investigate the possibility that Na’ channels might be directly influenced by cytoplasmic ATP sources, namely glycolysis, we explored the effect of IAA on INa. Despite the modest effects of inhibition of oxidative phosphorylation, glycolysis merits individual examination given the evidence that Z,,,,, is particularly sensitive to inhibition of this pathway [3Z]. Figure 3 shows the result of such experiments. 1X= was elicited with the same pulse protocol that we used for the experiments with DNP. No effect of IAA on Z,, properties was observed even after long periods of exposure [22-45 min, 29 f 8 min, n = 5; Fig. 3(a), right column]. The I-V curve pooled from 5 experiments confirmed this observation [Fig. 3(b)]. No significant difference was observed in the amplitude of ZNa at any membrane potential after addition of 1 mM IAA [Fig. 3(b), open circles]. As with DNP, the voltage dependence of the activation process was not modified by IAA. We took advantage of the patch-clamp technique to measure unitary current amplitude, and thus to exclude major changes in permeation properties (conductance and selectivity) as a result of metabolic inhibition. Single Na+-channel conductance was estim-

Effects

of Metabolic

Inhibition

on Na’

Channels

I 3 13 IAA

-80

mv

I rn~

I

-60mV lr----I------40mV {I------

l------20

mv

0 mV

iI60

(b)

pA

Membrone -120

-80

L8Ins poienilol

(mV)

-40

0

-. Control 0. --0 IAA I rn~ l

Current

(nA)

FIGURE 3. Elkct of metabolic inhibition induced t)! IAA on 2,“. (a! I,< &cited under the same conditions and with thr samr p&c protocol as in Figure. 2. b&xc 4rli column] and 28 min after additkm cd I m~ IA4 :ri$t column). fhr IL\. curve ronstructcd with data lkom live experiments. Experimental data WPW scaled using the same procrdurt as in Figure Z(b). ‘I‘he activation thwsk old was - 80 mV in both control and metabolic inhihitifm The maximal amplitude was ohswwd ;II conditions. - 38 mV. No significant ditferrncr in thr c urwnt amplitude was ohscrvcd aftcar addition of IAA. ‘l‘hv \moclth curws were obtained with rquation (1 j1 thv par;mwt~~s (11 the hest fit in control conditions wt‘w: f’,,,. 1.710.1 X IO ‘cm/s: 9 i 0.9 mv; ;111d h-r,, r;,,. - 51.4f I .5 m\‘. Aftm IAA those param~tm had the following valutx I’,,,, l.4fO.l x ICI ‘.,.m \. hi. 9.210.1 m\‘: &Id I’,,,, -53*o.L’mv.

(0)

Control

-60

DNP 200

IIM

-60

IAA I mM FIGURE 4. Effect of metabolic inbibition on Na+ channel conductance. (a) single-channel current recorded under control conditions (upper panel) with 70 ms depolarizing pulses to different membrane potentials from a HP of - 135 mV. Single-channel activity after addition of either 200~~ DNP (middle) or 1 mM IAA (lower panel). (b) Under control conditions the conductance (filled circles, left was panel) 21.5~kO.l pS and after addition of DNP 21.5fO.l pS (n=5). Under the same control conditions the conductance was: 20.5 f 0.1 pS and after IAA 2 1.6 f 0.1 pS (right panel, n= 5).

-60

-50

775~A

5 Ins Membrane

potentlcl

(mVi

(b) -80

0-0 Control D-.-S DNP 200

-60

-40

-20

0 I

-80

0-0

-60

-40

-20

I

0

Control

VM

Current

ipAi

Current

(pAi

Effects

of Metabolic

Inhibition

ated by measuring unitary events resolvable near the end of the INa decay phase, such as those shown in Figure 4(a). The conductance value calculated over the potential range from - 80 to - 20 mV was not modified by ATP depletion [Fig. 4(b)]. The unitary conductance values fell between 20 and 22 pS both in control conditions and during metabolic inhibition. Effect of metabolic inhibition

on I,,+, kinetics

Although we have found that the peak amplitude of ZYawas not changed, a modification of Na* channel kinetics favoring the development of steady-state current might still be capable of accounting for significant [Na’], accumulation during ischemia. Thus, we carefully examined the effect of metabolic inhibition on the voltage- and time-dependence of IX,. To determine the steady-state inactivation prepulses was of 4q,, the effect of conditioning studied before [Figs. 5(a) and (b), left colmetabolic inhibition umns] and during [Figs. 5(a) and (b) , right columns]. A conditioning prepulse (1 s) to different membrane potentials was followed by a test pulse (70 ms) to - 45 mV. Figures 5(c) and (d) illustrate the fraction of Na+ channels that was not inactivated as a function of conditioning prepulse potential before (filled circles) and during metabolic inhibition (open circles). The inactivation parameter (h,) was calculated as the ratio of peak current with and without the conditioning prepulse. The smooth curves ‘Figs. 5(c) and (d)] were described by the following expression:

h, = l/-l1 +exp[( r- ~‘,,,j2)/X;,lI

c4

where I’ is conditioning pulse potential; 1’,,,12, is the inactivation midpoint; and K,,, is the steepness factor [32]. None of these parameters was significantly changed as a result of metabolic inhibition. To investigate any possible effect of metabolic inhibition on the time course of IN., during depolarizing steps, we measured the activation and inactivation time constants (r,,, and r,, respectively) by fitting a single exponential to the fast component of each process. Figure 6 shows examples of the fitting process on the left, while the pooled results are plotted as a function of the membrane potential on

on Na+

Channels

I :<15

the right. Neither activation nor inactivation time constants were affected by ATP depletion. Thus, Na+ channel gating is not signihcantly altered by metabolic inhibition; in particular, there is no hint of the development of a window current. Discussion

Several lines of evidence suggest that the accumulation of [Na+]; that occurs during ischemia or metabolic inhibition can lead to an increase in intracellular Ca” via Na’ ca2+ exchange. An elevation of [Na’], is necessary to support the Na+-C:a’* exchange hypothesis and, indeed, such an increase in [Na+], has been extensively documented [d, 5, 6, 10, 29, 33, 341. Nevertheless, the mechanism of the increase in [Na+], has not yet been established. An increase of Na+ influx appears to be primarily responsible fbr [Nat], accumulation, at least during true ischemia when pH, falls by about one unit [I, 6, 8, /O] as a result of increased lactate production without metabolite washout. The resulting acidosis could activate Na+-H’ exchange and thereby promote [Na’], accumulation [9, 3.51. Direct evidence for this idea has been obtained by inhibition of Na’-H+ exchange with amiloride [ZO, 361, in which case [&a+], accumulation is significantly lower than in the drug-free controls. However, amiloride did not exacerbate the fall of pHi [ZO]: the absence of a change in pH, is surprising it amiloride is simply inhibiting Na- H+ eschange. Other effects are difficult to excludt given that amiloride is far from specifc, as a Na+-Ht exchange blocker; inhibition of’ C:a”+ [1.5] and Na’ channels [/S]. and of Na+-Cal’+ exchange [Iq], has alsc~ been reported. In the present study, we tla\.e explicitly considered the possibility that 1.01tagedependent Na+ channels contribute to Na + influx. The precedent that oxidative stress can enhance non-inactivating openings of ,%a channels [Z’] gave reason to wonder \z~hether such a mechanism might come into play during metabolic inhibition. The possible inv,olvement of Na+ channels is particularly important to evaluate given the high dcnsitv ot‘ these channels and their potential to tnediatc verv larpe ion fluxes.

(a 1

Control

DNP 200

-‘--~~

‘:,-I

-llomv ?J----7OmV

PM

L-p

IO pp.

8 ms (b)

Control

80 PA

IAA

I nw

L

4 ma

(d) DNP 200

,

,

I

UM O---O

Control a-e IAA

I WA 0-O 1

0.50

-

0.25

-

0.25

.

0.00 -140

-120

-100

-80

-60

-40 Membrane

potential

(mV)

Effects

of Metabolic

Inhibition

Using direct measurements of IX, in heart cells, we investigated the effects of metabolic inhibition on voltage-dependent Na+ channels. The amplitude of 1N,a, as measured in multichannel patches before and during metabolic inhibition, revealed a modest but significant reduction of peak Zxa at test potentials positive to -40 mV with DNP, but no change was observed with IAA. Of course, the modrst decrease seen with DNP is in the direction opposite to that which might have helped to explain the known changes in [Na- 1,. The voltage dependence of steadystate inactivation did not change with either intervention, nor did the time constants of activation and inacti\ration. Most importand), no window current was induced. In light of these experimental findings, the gradual depolarization of the resting membrane potential during ischemia would be predicted to inactivate Na- channels progressively, rendering them incapable of contributing to the accumulation of intrarellular sodium. The difference between the effects of IAA and thosr ofDiYP might relate to the fact that DNP induces more severe bulk ATP depletion than IAA [28], although this may not be the case in the immediate subsarcolemmal space near Na’ or I(* channels [31]. Our measuren1tmts of’ &.\,,, p rovide clear-cut evidence of subsarcolemmal ATP depletion with either IA;2 or DNP, but overall cellular integrit)

F’IGCRE 5. IXect of metabolic inhibition on the stead)-state inactivation of I,,C. The experimental cunditions wcrc the suw as previously described. (a) Effect of conditioning pulses on I,, before (left column) and ;Iftcr right rolumnl metabolic inhibition induced by 200,~~ DNP. One wcond depolarizing prepulscs wcrc drlivered to different membrane potentials (indicated at the leli of the traces). followed by a 70 ms depolarizing pulsr to - 40 mV, from a holding potential of - 135 mV. hl Same as in la). but using 1 rnM IAA as metabolic Inhibitor. ic / Plot of steady-state inactivation (IxJIyd,n,tx, h, as a function of the membrane potential, in control conditions (filled circksj and after DNP (open circles). ‘I‘he smooth curves were drawn with equation (2). In control conditions the best fit was obtained when I,= - 103.9f 1.8mV and &=9.8* 1.6mV. After 2OOfi~ DNP the value of these parameters were: I,= -105.lf2.2mV and K,=9.6+1.9mV. (d) Same ‘1s in (c! but using IAA. In control conditions the curve parameters WPTC: Vh= - 102+ 1.1 mV and h~=lOfl.OmV. After I rnM IAA, L’,,= -96.8f2.4mV ;u:d h;, = 10.7 * 2. I mV.

on Na’

Channels

I 3 I7

may also depend upon bulk ATP. Fvr ~L\XII the idea that the observed reduction of I,,, induced by DNP reflects cell damage rather than a specific block phenomenon, gi\.en that washout of DNP did not restore I,,, to its original value. At the single-channel le\,el WC did not observe any change in the number of channt+ (aYY)or in the unitary current amplitude (il. Thus, a decrease in PO must account ti)r the reduction of IX,. If [Na’], increases, a reduction in i might be expected due to ;I dccrrase in the driving force for Na’. entr!.. I.nder 0111 experimental conditions the lack ot‘ efyect of metabolic inhibition on i is explainc~d b! tht fact that WC used Ka+-free isotonic Ii’ in tht bath to null the cell membrane potential ;~nd thereby to enable explicit quantification of the potential across the patch of tnembrant under scrutiny. Llnder these circumstances [Na’], is near zero and cannot rixe during metabolic inhibition, since the onI\ source of‘ Na’ is in the patch pipette which ia in c’ontact with only few square microns of‘ X;ll.(.(Jlt’llilna. Our experiments were specificall) dc3igned to examine possible Xa’ channel ul~~-t-t~,~~~latio~l by metabolic inhibition, not to quantify the inhibition of &a+ flux throu,+ Na ’ chanrlt~ls that would be expected secondar) IO an increase in [Na’],. Nevertheless. it iy lvc,rth considering the likely chatiges in I that mi$t be obser\,ed under physiological conditions. The [Na+], increase recently rc1~ortr.d in \c,lltricular myocytes during relevant studies is (11) to five-fold [4, 19, 3.7, 34. 37j. ‘I‘lruz, if’ I+(‘ assume an initial Isa+], of 5 niM, tlit, masimum concentration during metat)olic. irlhit)ition that we should expect would be approximately 25 mM. :\t this c.orrc.rntratiorl, simulations using Goldman-HodgkiIlKatz assumptions (not shown! predict that the single-channel current amplitude ot)sc~r\.c~tl at negative potentials would remain practicalI) unaffected, even though the revrrsal potential would be significantly shifted to the left. In any case, the increase in [Na- J, that would occur under physiological ionic conditions would certainly not incl-east> Na * influx through voltage-dependent Na’ channels. Although we cannot exclude the possibility that other factors not explored here ~such as accumulation of lysophosphatid~-l~tl~)linr, 1~1: [2/j might come into play durirl ~vnuittt

1318

R. Mejia-Alvarez (a

1 0

.,,

_

and

E. Marban

r,=O.l7ms

r,=0.14

ms

+

100 pA I

(b

1 O’:. IAA I ITIM 0. --0

4,

50 pA -40

DNP 200

I

100 pp.

0

-20

0

pM

O----O

L

-60

(d

1

1

50

pA

I

?\k ‘e --i---4

--..I---

'6 0

-40 Membrane

-20 potential

0 (mV)

FIGURE 6. Effect of metabolic inhibition on the time course of I,,. A single exponential curve (continuous line) was fitted to the I,, rising phase recorded at - 30 mV (open circles represent experimental data under control conditions) before and after DNP [(a) dotted line] or IAA (b). Activation time constants were plotted as a function of the membrane potential (right panels, n= 5). A single exponential curve was fitted to the I,, decay phase before and after of DNP (c) or IAA (d). Inactivation time constants were plotted as a function of the membrane potential (right panels, R= 5). As in previous figures, control data are shown as filled circles and as open circles after the addition of the drug.

Effects

of Metabolic

Inhibition

ischemia, metabolic inhibition alone does not importantly alter the function of Na+ channels. Direct measurements of Z,, are impossible during true ischemia. Freshly dissociated ventricular myocytes may also present other important and undefined differences with the in vivo condition. Nevertheless, hypoxia and metabolic inhibition are known to increase

on Na+

Channels

1319

[Na+], in perfused hearts and in isolated myocytes 14, 7, 9, 19, 33, 34, 371; at the very least, our experiments enable us to reject the involvement of Na+ channels in these simple models. Thus, we conclude that other mechanisms must underlie the [Na’], accumulation observed during metabolic inhibition in heart cells.

References I LEE, J. 4., ALLEN, D. G. Mechanisms of acute ischemic contractile failure of the heart. Role or intracellular calcium. J Clin Invest 88, 361-367 (1991). 2 TANI, M., NEELY, J. R. Role of intracellular Na + in Ca2+ overload and depressed rc~overy of \rentricular function of reperfused ischcmic rat hearts. Circ Res 65, 1045-1056 (1989). 3 GRINWALD, P. M., BROSNAHAM, C. Sodium imbalance as a cause ofcalcium overload in post-hypoxic rcosygenation injury. J Mel Cell Cardiol 19, 487-495 (1987). 4 GUARNIERI, T. Intracellular sodium-calcium dissociation in early contractile failure in hypoxia tirret papillary muscle. J Physiol 388, 449465 ( 1987). 5 MALLOY, C. R., BUSTER, D. C., CASTRO, M. C. A., GERALDES, C. F. G. C., JEFFREY. F. M. H., SHERRY, .4. D. Influence of global ischemia on intracellular Na in the perfused rat heart. Magn Reson Med 15. 33 -44 I 19901. 6 PIKE, M. M., KITAKAZE, M., MARBAN, E. 23Na-NMR measurements of intracellular sodium in intact perfused ferret hearts during ischemia and reperfusion. Am J Physiol 259, H 1767-HI 773 i 1990) 7 NEUBAUER, S., BALSCHI, J, A., SPRINGER, C. S., SMITH, T. W., INGWALL, J. S. Intracellular Na’ accumulation in hypoxic us. ischemic rat heart: evidence for Nat-H+ exchange (Abstract). Circulation 76, IV-56 t, 1987 ). 8 KUSUOKA, H., PORTERFIELD, J. K., WEISMAN, H. F., WEIsFELpr, M. L., MARBAN, E. Pathophysiology and pathogenesis of stunned myocardium. Depressed Ca ‘+-activation as a u~nsequen~c of rept.rfilsioll-irrtl~l~,(,d cellular calcium overload in ferret hearts. J Clin Invest 79, 950-961 (1987). 9 ANDERSON, S.. MURPHY, E., STEENBERGEN, C., LONDON, R. E., CALA, I’. NapH exchange in myocardium: Eltfects ot hypoxia and acidification on Na and Ca. Am J Physiol 259, C940-C948 (1990). IO MURPHY, E., PERLMAN, M., LONDON, R. E., STEENBERGEN, C. Amiloride delays the isrhemia-induced rise in cytosolic free calcium. Circ Res 68, 1250-1258 (1991). I1 BORON, W. F., DE WEER, P. Intracellular pH transients in squid giant axons caused by CO,, XH, .rnd m~~tabolic inhibitors. J Gen Physiol 67, 91-l 12 (1976). 12 CINGOLANI, H. E., KORETSUNE, Y., MARBAN, E. Recovery of contractility and pH, during rrspirarr~r\ .*( idosi\ in ferret hearts: role of Na+-H’ exchange. Am J Physiol 259, H843-H848 ( 1990). 13 PAI.MER, L. Voltage-dependent block by amiloride and other monovalent cations of apical Na channels m thy toad urinary bladder. J Membr Biol 80, 153-165 (1984). 14 SCHELLENBERG, G. D., ANDERSON, L., SWANSON, P. Inhibition of Na+ ~Ca” exchange in rat brain I~\ nmilorid~. Mel Pharmacol 24, 251, (1983). 15 TANG, C., PRESSER, F., MORAD, M. Amiloride selectivity blocks the low threshold (‘1‘) calcium channel. St icncr 240. 213-215 (1988). 16 KIMITSUKI, T., .MITSUIYE, T., NOMA, A. Negative shift of cardiac Na’ channel kinetics in rell-atr,u hcd l)at~~h rerordings. Am J Physiol 258, 8247-8254 (1990). I7 PATLAK, J. B., ORTIZ, M. Slow currents through single sodium channels of the adult rat heart.J Gen Ph>>iol86. 89 104 (1985). 18 PIKE, M. M., KITAKAZE, M., MARBAN, E. Increase in intracellular free sodium conrcntration during irchemia revealed by “Na NMR in perfused ferret hearts. Circulation 78, II-56 I 1988). 19 HAIGNEY, M. C. P., VER DONCK, L., STERN, M. D., SILVERMAN, H. S. R-56865 prevents (:;1” overlor~d and (~11 death by limiting Na’ loading in hyposic cardiac myocytes. Circulation 84, II-666 (1991). 20 WAXMAN, S. G., RANSOM, B. R., STYS, P. K. Non-synaptir mechanisms of Ca” -mediated injury in (ZNS \\ hite matter. Trends Neurosci 14, 461-168 (1991). ‘21 BURNASHEV, N. A., ~NDROVINAS, A. I., FLEIDERVISH, I. A., ROSENSH.rRAuKH, L. 1’. Ischemic poison lyaophosphatidylcholine modifies heart sodium channels gating inducing long-lasting burst of openings. Pflrigers Arch 415, I L’4 126 (1989). 22 BHATNAGAR, A., SRIVATAVA, S. K., SZABO, G. Oxidative stress alter specific membrant currt’nt in isoiarrd c.udia( myocytes. Circ Res 67, 535-549 (1990). 23 HORI. M., GOTOH, K., KITAKASE, M., IWAI, K.. IWAKURA, K., SATO, H., KORETX.NE, Y.. INOUE, 11.. KIIABATAKO, A., KAMADA, ‘I‘. Role ofoxygen-derived free radicals in myorardial edema and isrhrmia in coronary mi( ~~~~as~~ulat cmbolization. Circulation 84, 828-840 (1991).

1320 24 25

26 27 28

29 30

31 32 33 34 35

36 37

R. Mejia-Alvarez

and

E. Marban

R., MORAD, M. A uniform enzymatic method for dissociation of myocytes from hearts and stomachs of vertebrates. Am J Physiol 249, Hl056-H1060 (1986). HAMILL, 0. P., MARTY, A., SAKMANN, B., SIGWORTH, F. J. Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches. Pfliigers Arch 391, 855100 (1981). WILCOXON, F. Individual comparisons by ranking methods. Biometrics Bull 1, 80 (1945). WEBB, J. L. Iodoacetate and iodoacetamide. In Eqymes and Metabolic Inhibition Vol3, pp. l-283 (1966). New York: Academic Press. WEISS, J. N., HILTBRAND, B. Functional compartmentation of glycolytic versus oxidative metabolism in isolated rabbit heart. J Clin Invest 75, 436-447 (1985). STRYER, L. Biochemistry. p. 421, (1988). New York: W. H. Freeman and Co. KAKEI. M., NOMA, A. Adenosine-5’-Triphosphate-sensitive single potassium channel in the atrioventrirular node cell of the rabbit heart. J Physiol 352, 265-284 (1984). WEISS, J. N., LAMP, S. T. Glycolysis preferentially inhibits ATP-sensitive K’ channels in isolated guinea-pig cardiac myocytes. Science 238, 67-69 (1987). HODGKIN, A. L., HUXLEY, A. F. A quantitative description of membrane currents and its application to conduction and excitation in nerve. J Physiol 117, 500-544 (1952). MURPHY, E., LEFURGEY, A., LIEBERMAN, M. Biochemical and structural changes in cultured heart cells induced by metabolic inhibition. Am J Physiol 253, C708C706 (1987). HIGNEY, M. C. P., MIYATA, H., LAKATTA, E. G., STERN, M. D., SILVERMAN, H. S. Sodium-linked Cay+ overload and cell death in reoxygenated rat cardiac myocytes. 64th Sci. Sessions of AHA. Circulation 84, II-664 (1991). LAZDUNSKI, M., FRELIN, C., VIGNE, P. The sodium/hydrogen exchange system in cardiac cells: Its biochemical and pharmacological properties and its role in regulating internal concentrations ofsodiuim and internal pH. J Mel Cell Cardiol 17, 1029-1042 (1985). GRINSTEIN, S., COHEN, S., ROTHSTEIN, A. Cytoplasmic pH regulation in thymic lymphocytes by an amiloridesensitive Na+/H+ antiport. J Gen Physiol 83, 341-369 (1984). MURPHY, J. G., SMITH, T. W., MARSH, J. D. Mechanisms of reoxygenation-induced calcium overload in cultured chick embryo heart cells. Am J Physiol 254, HI 133-HI 141 (1988). MITRA,