HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) All Versions of this Article: 556/3/773    most recent jphysiol.2003.058248v1 Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Bouchard, R. Articles by Giles, W. R. Search for Related Content PubMed PubMed Citation Articles by Bouchard, R. Articles by Giles, W. R.

¹ Institute of Cardiovascular Sciences, University of Manitoba, St Boniface Research Centre, 351 Taché Avenue, Winnipeg, Manitoba, Canada, R2H 2A6² Department of Physiology and Biophysics, The University of Calgary, 3330 Hospital Drive N.W., Calgary, Alberta, Canada, T2N 4 N1³ Department of Bioengineering, University of California San Diego, 9500 Gilman Drive, La Jolla, CA 92093-0412, USA

	Abstract

Top Abstract Introduction Methods Results Discussion References

 
The mechanisms underlying the inotropic effect of reductions in [K⁺]_o were studied using recordings of membrane potential, membrane current, cell shortening and [Ca²⁺]_i in single, isolated cardiac myocytes. Three types of mammalian myocytes were chosen, based on differences in the current density and intrinsic voltage dependence of the inwardly rectifying background K⁺ current I_K1 in each cell type. Rabbit ventricular myocytes had a relatively large I_K1 with a prominent negative slope conductance whereas rabbit atrial cells expressed much smaller I_K1, with little or no negative slope conductance. I_K1 in rat ventricle was intermediate in both current density and slope conductance. Action potential duration is relatively short in both rabbit atrial and rat ventricular myocytes, and consequently both cell types spend much of the duty cycle at or near the resting membrane potential. Rapid increases or decreases of [K⁺]_o elicited significantly different inotropic effects in rat and rabbit atrial and ventricular myocytes. Voltage-clamp and current-clamp experiments showed that the effects on cell shortening and [Ca²⁺]_i following changes in [K⁺]_o were primarily the result of the effects of alterations in I_K1, which changed resting membrane potential and action potential waveform. This in turn differentially altered the balance of Ca²⁺ efflux via the sarcolemmal Na⁺–Ca²⁺ exchanger, Ca²⁺ influx via voltage-dependant Ca²⁺ channels and sarcoplasmic reticulum (SR) Ca²⁺ release in each cell type. These results support the hypothesis that the inotropic effect of alterations of [K⁺]_o in the heart is due to significant non-linear changes in the current–voltage relation for I_K1 and the resulting modulation of the resting membrane potential and action potential waveform.

(Received 17 November 2003; accepted after revision 24 February 2004; first published online 27 February 2004)
Corresponding author W. R. Giles: Department of Bioengineering, University of California San Diego, 9500 Gilman Drive, La Jolla, CA 92093-0412, USA. Email: wgiles{at}bioeng.ucsd.edu

	Introduction

Top Abstract Introduction Methods Results Discussion References

 
Hypokalaemia occurs when plasma [K⁺]_o falls from the normal range of 3.5–4.5 mM to < 3 mM (Braunwald, 2001), and hyperkalaemia develops when plasma [K⁺] rises to >6 mM. In both cases, changes in plasma [K⁺] levels result in marked electrophysiological and inotropic effects in mammalian myocardium, and these may be associated with a range of pathophysiological symptoms, including ischaemia, acidosis, and cardiac arrhythmias (Braunwald, 2001). However, despite the well-known clinical manifestations and treatment regimes for changes in plasma [K⁺], the cellular mechanisms responsible for these symptoms are not well understood.

In the laboratory, increases in the concentration of extracellular K⁺ ([K⁺]_o) have a negative inotropic effect in cat ventricle and guinea-pig ventricle and atrium (Kavaler et al. 1972; Ku et al. 1975; Ng et al. 1987; Herzig, 1992). Other studies, however, have shown a positive or ‘anomalous’ inotropic effect of increasing [K⁺]_o in rat atrium (Ng et al. 1987) or when contractile activity is allowed to re-equilibrate following the change in [K⁺]_o in cat ventricle and guinea-pig atrium (Kavaler et al. 1972; Ku et al. 1975). On the other hand, reduction of [K⁺]_o has been observed to elicit a positive inotropic effect in human ventricle, guinea-pig ventricle and atrium, and bullfrog atrium (Gettes et al. 1962; Godfraind & Ghysel-Burton, 1980; Christe, 1983; White & Terrar, 1991) or oscillations in force development in human Purkinje fibres (Christe, 1983). The inotropic effect resulting from alteration of [K⁺]_o is therefore heterogeneous, being both tissue (atrium versus ventricle) and species dependent. Differences in the inotropic response to [K⁺]_o of this nature have been interpreted in terms of variability in various cardiac tissues of Na⁺–K⁺ pump activity or the dependence of contraction on intracellular Na⁺ concentration (Ng et al. 1987; Aronson & Nordin, 1988; Nakao & Gadsby, 1989; Hayashi et al. 1994; Despa et al. 2002; Bers et al. 2003).

In addition to mechanical effects, changes in [K⁺]_o also result in significant alterations in the electrophysiological properties of cardiac cells. Based on the significant differences in the properties of voltage-gated K⁺, Na⁺ and Ca²⁺ channels from one species to the next and from one region of the myocardium to the next even within a given species (Antzelevitch & Dumaine, 2002), it is reasonable to speculate that changes in the electrophysiological profiles of the myocardium may explain some of the inotropic responses in different tissues. For example, reduction of [K⁺]_o is associated with hyperpolarization of the resting membrane potential in ventricular and atrial tissues of most mammals, while opposite effects are seen when [K⁺]_o is increased. Although the effect of increasing or decreasing [K⁺]_o to depolarize or hyperpolarize the resting potential is fairly consistent from one cell type to the next, the degree to which the membrane potential shifts away from the calculated Nernst potential for K⁺ (E_K) varies significantly from one cell type to the next (Whalley et al. 1994; Bailly et al. 1998; Baczko et al. 2003), depending upon the slope conductance for K⁺ and input resistance (R_input) of the cell (Aronson & Nordin, 1988; Bridge et al. 1990; McCullough et al. 1990; Whalley et al. 1994; Golod et al. 1998). The effect of [K⁺]_o on action potential waveform is similarly heterogeneous. Elevation of [K⁺]_o has been shown to increase (Ng et al. 1987) or decrease (Herzig, 1992) action potential duration in rat and rabbit atrial and ventricular muscle. Similar discrepancies have been observed following a reduction of [K⁺]_o in rabbit, human and bullfrog ventricular and atrial tissues (Gettes et al. 1962; Goto et al. 1977; Christe, 1983; White & Terrar, 1991).

The inwardly rectifying background K⁺ current, I_K1, is the main ionic current responsible for setting the resting membrane potential in mammalian heart cells and it can also influence the late phase of repolarization (Hume & Uehara, 1985; Giles & Imaizumi, 1988; Nichols et al. 1996). Moreover, it is known that these inward rectifier K⁺ channels differ in density and intrinsic voltage dependence from tissue to tissue both within and between species (Hume & Uehara, 1985; Giles & Imaizumi, 1988; Shimoni et al. 1992). Because changes in either resting membrane potential or action potential waveform can strongly modulate excitation–contraction coupling (E–C coupling) in cardiac tissues (Wood et al. 1969; for review, see Bers, 2001), we evaluated the hypothesis that the inotropic response to alteration of [K⁺]_o in various cardiac tissues is mediated, in part, by I_K1-induced changes in resting membrane potential and/or action potential waveform. Rat ventricular and rabbit atrial and ventricular myocytes were selected for study based on reported differences in the voltage dependence of I_K1. Our experiments showed that changes in [K⁺]_o had effects on cell shortening, intracellular Ca²⁺ concentration ([Ca²⁺]_i), resting membrane potential and action potential waveform and demonstrated that these varied substantially between cell types. These effects were due to a functional coupling of I_K1-induced changes in resting membrane potential and action potential waveform to the mechanisms responsible for modulating sarcolemmal Ca²⁺ fluxes and intracellular Ca²⁺ loading and release.

	Methods

Top Abstract Introduction Methods Results Discussion References

 
Myocyte preparation

Adult rabbits (1–2 kg) and adult male Sprague–Dawley rats (250–350 g) were anaesthetized with ether and killed by cervical dislocation. Ventricular and atrial myocytes were prepared from rabbit heart using a dissociation procedure similar to that of Giles & Imaizumi (1988). Myocytes from right ventricle of rat hearts were prepared as previously described (Bouchard et al. 1993a,b). Experimental protocols were approved by the University of Calgary institutional animal care committee, according to the guidelines of The Canadian Council of Animal Care.

Solutions and drugs

Standard Hepes-buffered Tyrode solution contained (mM): NaCl, 140; KCl, 5; CaCl₂, 1 or 2, as indicated; MgCl₂, 1; Hepes, 10; glucose, 5.5. The pH was titrated to 7.4 with 1 M NaOH. In experiments where the concentration of K⁺ in the Tyrode solution was changed, the amount of KCl was altered, without ionic or osmotic compensation. In some experiments, CsCl (3 mM), 4-aminopyridine (4-AP, 3 mM) and tetrodotoxin (TTX, 15 µM) were added to the standard Tyrode solution.

Two different pipette filling solutions were used. For the majority of experiments, the pipette solution contained (mM): potassium aspartate, 120; KCl, 20; MgCl₂, 1; Na₂ATP, 5; Hepes, 10. pH was adjusted to 7.2 with 1 M KOH. For experiments designed to isolate Ca²⁺ influx through voltage-gated Ca²⁺ channels (Figs 5 and 7), the pipette solution contained (mM): aspartic acid, 120; CsCl, 30; MgCl₂, 1, Na₂ATP, 5; Hepes, 10. pH was adjusted to 7.2 with CsOH solution. Analar grade chemicals (BDH Chemicals Ltd, Poole, UK) were used to prepare all solutions. TTX, 4-AP, BaCl₂ and Na₂ATP were obtained from Sigma Chemical Co. (St Louis, MO, USA). CsOH solution was obtained from Aldrich Chemical Co. (Milwakee, WI, USA).

View larger version (19K): [in this window] [in a new window]   Figure 5.  Effects of changes in holding potential on [Ca2+]i in voltage-clamped rat and rabbit ventricular myocytes A, membrane currents and indo-1 fluorescence transients from a rat ventricular myocyte. The cell was held at –80, –60 and –130 mV for 2 min at each potential, and membrane currents and [Ca2+]i transients were elicited by a 200 ms depolarizing step to +20 mV from each holding potential. The patch-clamp pipette solution (140 mM K+; see Methods) contained 100 µM Indo-1. B, membrane currents and indo-1 fluorescence transients from a rabbit ventricular myocyte. The patch-clamp pipette solution (140 mM Cs+; see Methods) contained 50 µM Indo-1. Extracellular solution contained 3 mM Cs+.

View larger version (24K): [in this window] [in a new window]   Figure 7.  Effects of changes in diastolic membrane potential and action potential waveform on net Ca2+ influx and unloaded cell shortening in rat (A, B) and rabbit ventricular (C, D) myocytes For all experiments, the pipette solution contained Cs+ in place of K+, and Tyrode solution contained TTX, Cs+ and 4-AP. Ca2+-dependent membrane currents were obtained by subtracting the current remaining after replacement of extracellular Ca2+ by Mg2+ from the control current (see Methods). A, depolarizing steps (top) were applied from a holding potential of either –80 (a) or –115 mV (b). Ca2+ current (middle) was not altered by membrane hyperpolarization, but unloaded cell shortening (bottom) was significantly reduced. B, a rat ventricular myocyte was voltage clamped with action potential waveforms (top) recorded in 5 mM (a) or 1 mM (b) [K+]o. Ca2+ current (middle) was slightly shorter in duration during the 1 mM[K+]o waveform compared with the 5 mM[K+]o waveform, while unloaded cell shortening (bottom) was reduced in amplitude by about 40%. Stimulation frequency was 0.2 Hz, and [Ca2+]o was 1 mM in both A and B. C, effects of 5 mM (a) and 1 mM (b) [K+]o action potential waveforms (top) on Ca2+ current (middle) and cell shortening (bottom) in a voltage-clamped rabbit ventricular myocyte. D, effect of oscillations in membrane potential typically recorded on prolonged exposure of myocytes to 1 mM[K+]o on Ca2+ current (middle) and unloaded cell shortening (bottom). Oscillations in cell length typically followed the time course of Ca2+ influx during oscillations in membrane potential and were abolished following a switch from control solution () to solution where all of the extracellular Ca2+ was substituted with Mg2+ (). Stimulation frequency was 0.2 Hz, and [Ca2+]o was 2 mM in C and D.

Standard whole-cell patch-clamp techniques were used to current and voltage-clamp single myocytes. Patch pipettes had DC resistances of 2–4 M. Recorded membrane potentials were corrected by –10 mV to compensate for liquid junction potentials between bath and pipette solutions. Myocytes were placed in a 200 µl recording chamber on the stage of an inverted microscope (Diaphot; Nikon, Tokyo, Japan), and superfused by gravity at about 0.5 ml min^–1. In some experiments, a multibarrelled local superfusion pipette (Bouchard et al. 1993a,b) was used. This allowed solution changes to be made within 1 s. Temperature was 20–22°C in all experiments.

Membrane potential and currents were sampled at 1 kHz using a 12-bit A/D convertor board (DT2801A; Data Translation, Marlborough, MA, USA) and stored in a microcomputer using custom software. In some experiments cell shortening signals were low-pass filtered (DC–100 Hz) and digitized simultaneously with membrane current and potential. Cell shortening was measured with a video edge detection circuit (Steadman et al. 1988). Digitized signals were analysed using a custom software package, and a commercial plotting program (‘Sigmaplot’, Jandel Scientific, Corta Madera, CA, USA).

In some experiments, rat and rabbit ventricular myocytes were voltage clamped with action potential-shaped command waveforms (Bouchard et al. 1995). These action potentials were recorded in normal or low [K⁺]_o, digitized at 20–50 kHz and stored in a separate microcomputer. When used as voltage-clamp command signals, these stored waveforms were directed to the D/A converter, and the output signal was low-pass filtered at 4 kHz before it was applied to the command input of the patch-clamp amplifier.

Ca²⁺-dependent membrane current was identified in some voltage-clamp experiments by subtracting the current records obtained in the absence of extracellular calcium [Ca²⁺]_o from those records obtained in the presence of normal [Ca²⁺]_o. For these experiments, CaCl₂ was replaced by MgCl₂. Solution changes in these experiments were made using the local superfusion pipette described above. To eliminate time- and voltage-dependent K⁺ and Na⁺ currents these experiments were also carried out with a pipette solution containing Cs⁺ in place of K⁺, and an external solution containing CsCl, 4-AP and TTX (see above). In experiments where the concentration of extracellular Na⁺ ([Na⁺]_o) was reduced, external Na⁺ was replaced with equimolar Li⁺.

Intracellular Ca²⁺ measurements

The methods for measuring fluorescence of the Ca²⁺-indicator dye Indo-1 have been described in detail elsewhere (Bouchard et al. 1995). Briefly, cells were loaded with the K⁺ salt of Indo-1 (Molecular Probes, Eugene, OR, USA) by adding 50 or 100 µM of the dye to the patch pipette solution. Cells were illuminated with UV light (365 ± 10 nm bandpass filter) via the epifluorescence port of the inverted microscope. Emitted fluorescence was collected with a x40 Fluo objective lens (NA 1.3), passed through a long-pass dichroic reflector with a half-transmission wavelength of 450 nm, and collected by two photomultipliers at wavelengths of 410 ± 20 nm and 500 ± 20 nm. Background fluorescence at each wavelength was subtracted after a gigaohm seal was made on a cell, but before the patch was ruptured and the cell was filled with indo-1. The ratio of background-corrected photomultiplier current at 410 nm to that at 500 nm (i.e. 410/500) was determined with an analog divider circuit and then digitized for storage in a microcomputer. Changes in this ratio were taken as a measure of changes in [Ca²⁺]_i. Membrane currents and [Ca²⁺]_i transients were elicited by 200 ms voltage-clamp steps (at 0.2 Hz) to +20 mV from a series of holding potentials. Eight successive current and [Ca²⁺]_i records were averaged.

	Results

Top Abstract Introduction Methods Results Discussion References

 
Effect of changes in [K⁺]_o on I_K1

Figure 1 compares typical action potential waveforms in myocytes from rabbit ventricle, rabbit atrium, and rat ventricle recorded in 5 mM[K⁺]_o Hepes-buffered Tyrode solution. Action potentials in rabbit ventricular (Fig. 1A) and atrial (Fig. 1B) myocytes were strikingly different in time course and amplitude. Ventricular action potentials typically had a small early repolarization phase and a prolonged plateau at 0 mV, which was then followed by a rapid final repolarization to the resting membrane potential (Fig. 1A). In contrast, atrial action potentials had a very prominent early phase of repolarization, followed by a slow repolarization to the resting potential (Fig. 1B). Consequently at membrane potentials corresponding to 25% and 50% repolarization, the ventricular action potential was more than 10-fold longer than the atrial action potential (Tables 1 and 2). Action potentials from rat ventricle (Fig. 1C) had a very prominent early phase of repolarization (denoted APD₂₅ and APD₅₀, respectively), which was followed by a much slower phase of final repolarization. Action potential parameters for each of the three cell types are summarized in Table 2.

View larger version (18K): [in this window] [in a new window]   Figure 1.  Comparison of action potentials (top) and inwardly rectifying K+ current (IK1; bottom) from rabbit ventricle (A), atrium (B) and rat ventricle (C) Stimulus frequency for action potentials was 0.2 Hz, [K+]o was 5 mM, and [Ca2+]o was 2 mM in A and B, and 1 mM in C. Action potentials and IK1 were recorded from different myocytes in each case. IK1 current–voltage curves in rabbit ventricular (A) and atrial (B) myocytes were generated using a 6 s, linear ramp voltage-clamp command from +60 to –140 mV. IK1 was the difference current obtained by subtracting the current remaining following exposure of the myocyte to solution containing 1 mM K+ and 0.3 mM Ba2+ from corresponding controls recorded in 1, 2, 5 or 10 mM K+ solution. Note the difference in current scales between rabbit ventricular and atrial cells. For rat ventricular myocytes (C), 1 s voltage-clamp steps were applied from a holding potential of –80 mV in the presence and absence of 0.3 mM Ba2+. IK1 was taken as the Ba2+-sensitive difference current at the end of the depolarizing step.

The magnitude and current–voltage relation for I_K1 was dependent on [K⁺]_o. In both rabbit and rat ventricular cells, reducing [K⁺]_o from 5 to 1 mM (Fig. 1) shifted the reversal potential, the potential at which peak outward current occurred, and the negative slope region of the current–voltage relation to more negative membrane potentials. Increasing [K⁺]_o had opposite effects on reversal potential and region of negative slope. In addition, the magnitude of the peak outward current increased with increased [K⁺]_o, which resulted in a ‘cross-over’ of the current–voltage relations in different [K⁺]_o. Changes in [K⁺]_o shifted the reversal potential by amounts consistent with I_K1 being K⁺-selective. In rabbit atrial cells, changes in [K⁺]_o affected the magnitude of outward I_K1 much less than in ventricular cells, resulting in a much smaller ‘cross-over’ of the current–voltage relations (Giles & Imaizumi, 1988; Whalley et al. 1994; Golod et al. 1998). In all three cell types, changes in [K⁺]_o shifted the reversal potential of I_K1 by amounts expected for K⁺-selective channels.

Due to the dominance of I_K1 in setting the resting membrane potential and in shaping the final phase of action potential repolarization, we investigated whether the observed species-dependent differences in I_K1 could differentially impact on the inotropic effects of changes in [K⁺]_o in these three cell types.

Effect of changes in [K⁺]_o on cell shortening and membrane potential

Reduction of [K⁺]_o resulted in a significant negative inotropic effect in rat ventricular myocytes. Figure 2A shows that in a rat ventricular myocyte stimulated at a frequency of 0.2 Hz, a sudden reduction in [K⁺]_o from 5 to 1 mM resulted in an immediate hyperpolarization of resting membrane potential by about 35 mV, which was followed by a slower, gradual reduction in the amplitude of cell shortening. After about 2 min in 1 mM[K⁺]_o, peak shortening had decreased to about 55% of the control value. The negative inotropic effect of [K⁺]_o reduction was reversed by briefly exposing the cell to a Tyrode solution where the concentration of extracellular Na⁺ ([Na⁺]_o) was reduced from 140 to 70 mM by substituting external Na⁺ with Li⁺. On return to 5 mM[K⁺]_o, the resting membrane potential immediately depolarized to –80 mV, but the negative inotropic effect reversed slowly. These results are consistent with previous data reported from our laboratory under voltage-clamp conditions (Bouchard et al. 1993a, b), where changes in holding potential or action potential duration effected cell shortening primarily through slow changes in intracellular Ca²⁺ loading and release. As such, the data are consistent with previous observations that the positive inotropic effect of increasing [K⁺]_o in rat atrial muscle can be blocked following exposure to either D600 or caffeine (Ng et al. 1987; Herzig, 1992), both of which inhibit Ca²⁺-induced Ca²⁺ release (CICR).

View larger version (38K): [in this window] [in a new window]   Figure 2.  Effect of [K+]o reduction on resting membrane potential, action potential wave form and contraction of a rat ventricular myocyte A, continuous chart recording of resting membrane potential (Vm; top) and peak unloaded cell shortening (CS; bottom) of a current-clamped rat ventricular cell. The cell was stimulated at 0.2 Hz through the recording pipette. Action potential amplitude was truncated at 0 mV. Rapid reduction of [K+]o from 5 to 1 mM resulted in an immediate membrane hyperpolarization and a slow decrease in peak shortening. The negative inotropic effect was reversed by exposing the cell to solution where [Na+]o was reduced from 140 to 70 mM by Li+ substitution. Return to 5 mM[K+]o resulted in a rapid depolarization of the cell, and a slow increase in peak cell shortening. B, effect of [K+]o reduction from 5 mM (a) to 1 mM (b) on action potential waveform (top) and unloaded cell shortening (bottom) in a rat ventricular myocyte. Peak shortening was reduced to 33% of control after 3 min exposure to 1 mM[K+]o.

The effects of changes in [K⁺]_o on action potentials and cell shortening of current-clamped rabbit ventricular and atrial cells are compared in Fig. 3. Reduction of [K⁺]_o from 5 to 1 mM resulted in hyperpolarization of the resting potential of a rabbit ventricular cell by about 35 mV, a shortening of the initial phase of repolarization, and a very marked prolongation of final repolarization that was often accompanied by large afterdepolarizations (Fig. 3A). These prolonged action potentials resulted in an increase in the magnitude of cell shortening, and smaller, secondary contractions. In some rabbit ventricular cells, prolonged reduction of [K⁺]_o resulted in oscillations of both membrane potential and cell shortening (Fig. 7D). The effects of increasing [K⁺]_o from 5 to 10 mM in a rabbit ventricular cell are shown in Fig. 3B. Increasing [K⁺]_o led to an 18 mV depolarization of the resting membrane potential, loss of the initial notch during early repolarization of the action potential and a small decrease in action potential duration. These changes in action potential were accompanied by a small negative inotropic effect.

View larger version (30K): [in this window] [in a new window]   Figure 3.  Effects of [K+]o changes on action potentials and unloaded cell shortening of rabbit ventricular (A, B) and atrial (C, D) cells A, reduction of [K+]o from 5 mM (a) to 1 mM (b) and return to 5 mM (c) in a rabbit ventricular cell. B, increase of [K+]o from 5 mM (a) to 10 mM (b) and return to 5 mM (c) in a rabbit ventricular myocyte. C, reduction of [K+]o from 5 mM (a) to 1 mM (b,c) and back to 5 mM (d) in a rabbit atrial myocyte. Continued exposure (c; 5 min) to I mM [K+]o resulted in membrane depolarization to about –50 mV, and complete loss of membrane excitability and contractions (c). D, increase of [K+]o from 5 mM (a) to 10 mM (b) and back to 5 mM in a rabbit atrial myocyte. Note that all electrophysiological and mechanical effects observed following changes in [K+]o were reversible on return to 5 mM[K+]o (A–D). Stimulation frequency was 0.2 Hz, and [Ca2+]o was 2 mM.

Effect of changes in membrane potential on cell shortening and [Ca²⁺]_i

Figures 2 and 3 illustrate that altering [K⁺]_o results in substantial changes in resting membrane potential and contractility, and that the inotropic effects of lowering [K⁺]_o were reversed on reduction of [Na⁺]_o. It has been shown previously that changes in the holding potential of voltage-clamped myocytes can have a significant impact on contractility of cardiac myocytes due to alterations in Na⁺–Ca²⁺ exchange activity (Wier & Beuckelmann, 1989; Bouchard et al. 1993a). Since both rat and rabbit cardiac myocytes spend much of the duty cycle at 0.5 Hz stimulation in diastole (Table 2), we speculated that a portion of the inotropic effects of changes in [K⁺]_o may have resulted from changes in diastolic membrane potential.

The inotropic effect of changes in resting membrane potential on rat ventricular and rabbit ventricular and atrial cells is shown in Fig. 4. Cells were voltage clamped at various holding potentials and contractions were elicited by 200 ms depolarizing steps to +20 mV. Figure 4A–C illustrates typical examples of the membrane currents and contractions recorded from rat ventricular, rabbit ventricular and rabbit atrial myocytes at a series of holding potentials. As illustrated by the pooled data in Fig. 4D, the dependence of cell shortening on holding potential varied significantly in rat ventricular and rabbit atrial and ventricular myocytes. Hyperpolarization of the holding potential from –80 mV resulted in negative inotropic effects of similar magnitude in all three cell types. However, depolarization of rat ventricular and rabbit atrial cells in the range –80 to –60 mV produced significantly larger positive inotropic effects compared with rabbit ventricular cells. For example, depolarization of the holding potential from –80 to –60 mV produced a 70% increase in peak shortening in rat ventricular cells, compared with only a 10% increase in rabbit ventricular myocytes (Fig. 4D). Thus, the dependence of cell shortening on holding potential was substantially different in rat and rabbit ventricular cells.

View larger version (30K): [in this window] [in a new window]   Figure 4.  Inotropic effects of changes in holding potential of voltage-clamped rabbit and rat cardiac myocytes Rat ventricular (A), rabbit atrial (B) and rabbit ventricular (C) myocytes were voltage clamped for 2 min at each of four holding potentials: a, –60 mV; b, –80 mV; c, –100 mV; d, –120 mV. A 200 ms depolarization to +20 mV (top) elicited membrane currents (middle) and contractions (bottom). Stimulation frequency was 0.2 Hz, [Ca2+]o was 2 mM. D, pooled data from rat ventricular (), rabbit ventricular (•) and rabbit atrial () cells showing effect of diastolic membrane potential on peak cell shortening. Contractions were recorded 2 min after changing the holding potential, in order to ensure steady-state conditions. For each cell, peak shortening at each potential was normalized to its magnitude at a holding potential of –80 mV. Data points are mean ±S.E.M.; n= 10 rat ventricular cells, n= 4 rabbit ventricular cells, n= 7 rabbit atrial cells.

Effect of changes in [K⁺]_o on cell shortening at constant membrane potential

Figure 6 shows the effects of changing [K⁺]_o under conditions where membrane potential was held constant using voltage clamp. Figure 6A shows the effect of reducing [K⁺]_o from 5 to 1 mM on unloaded cell shortening and membrane currents of a rat ventricular cell that was voltage clamped at –80 mV, close to its resting potential. Reduction of [K⁺]_o slightly increased the magnitude of the transient outward component of membrane current during the depolarizing step, and produced an outward shift in the holding current. This shift in holding current is consistent with the difference in current–voltage relations for I_K1 in 5 and 1 mM[K⁺]_o shown in Fig. 1A. In contrast to the inotropic effects of low [K⁺]_o observed under current-clamp conditions (Fig. 2), reduction of [K⁺]_o resulted in a small (< 10%) increase in the amplitude of unloaded cell shortening of voltage-clamped myocytes. In 15 myocytes held at –80 mV, peak shortening in 1 mM[K⁺]_o increased by 9.1 ± 3.9% compared with control (P < 0.017). Figure 6B shows that this small positive inotropic effect was independent of the holding potential over the range –60 to –120 mV. As shown in figure 6C and D similar results were obtained in rabbit ventricular and atrial myocytes. Reduction of [K⁺]_o from 5 to 1 mM resulted in a small positive inotropic effect, an approximately 12% increase in peak cell shortening in ventricular cells and a 9% increase in atrial cells. Conversely, an increase in [K⁺]_o (to 10 mM) resulted in a small negative inotropic effect in both cell types (Fig. 6D).

View larger version (23K): [in this window] [in a new window]   Figure 6.  Effects of changes in [K+]o on membrane current and unloaded cell shortening in voltage-clamped rat (A, B) and rabbit (C, D) cardiac cells A, effect of reducing [K+]o from 5 mM (a) to 1 mM (b) on membrane currents (middle) and peak shortening (bottom) of a voltage-clamped rat ventricular myocyte. Currents and contraction were elicited by 200 ms voltage-clamp steps to +20 mV (top) applied at 0.2 Hz. The changes in membrane current and contraction were completely reversible on return to 5 mM[K+]o (c). B, change in peak shortening of voltage-clamped rat ventricular myocytes after reduction of [K+]o from 5 to 1 mM at selected holding potentials between –120 and –60 mV. The increase in peak shortening was not significantly different at all potentials over this range. Data points are mean ±S.E.M., n= 3–8. [Ca2+]o was 1 mM. C, effects of changing [K+]o from 5 mM (a) to 1 mM (b) and 10 mM (c) on membrane currents (middle) and unloaded cell shortening (bottom) of a voltage-clamped rabbit ventricular cell. D, effect of changes in [K+]o on peak shortening of voltage-clamped rabbit ventricular and atrial myocytes (holding potential, –80 mV). The plot shows the percentage change in peak cell shortening in ventricular () and atrial (•) cells following reduction of [K+]o from 5 to 1 mM, and in ventricular () and atrial () cells following an increase in [K+]o to 10 mM. Data points are mean ±S.E.M., n= 4. Values were not significantly between ventricular and atrial cells (P > 0.2) at either [K+]o.

Effects of changes in resting membrane potential on Ca²⁺influx

Previous studies have reported that changes in action potential waveform can significantly affect the mechanisms underlying E–C coupling in rat and rabbit cardiac cells (Bouchard et al. 1995; Yuan et al. 1996; Sah et al. 2003). Accordingly, we investigated whether [K⁺]_o-induced changes in diastolic membrane potential and action potential waveform resulted in accompanying changes in the integral of Ca²⁺ entry through voltage-gated Ca²⁺ channels (I_Ca) in rabbit and rat ventricular myocytes.

Figure 7A shows the effects of membrane hyperpolarization on Ca²⁺-dependent membrane currents and cell shortening of a voltage-clamped rat ventricular cell. Ca²⁺-dependent difference currents consisted of a large, rapidly activating inward current that inactivated during the depolarizing step and a small, slowly declining inward current which followed repolarization of the cell to the holding potential. The large initial component results primarily from current through L-type Ca²⁺ channels, while the slow ‘tail’ of current is due to electrogenic Na⁺–Ca²⁺ exchange (Bridge et al. 1990; Bouchard et al. 1995). Changing the holding potential from –80 to –115 mV had no detectable effect on the influx of Ca²⁺, as measured by the integral of current during the depolarizing step. I_Ca was 19.4 pC at –80 mV and 20 pC at –115 mV. These values appear to be identical (within measurement error), and hence, in this experiment, changes in Ca²⁺ influx at different holding potentials cannot explain the substantial decrease in peak cell shortening that was observed.

The possibility that the changes in action potential waveform which accompany [K⁺]_o reduction in rat ventricle (Fig. 2) may be responsible in part for the negative inotropic effect was tested with action potential voltage-clamp experiments. Action potentials were recorded from the same rat ventricular cell in 1 and 5 mM[K⁺]_o and used as voltage-clamp command waveforms. Figure 7B shows that the Ca²⁺ currents produced by the action potential waveforms recorded in 5 and 1 mM[K⁺]_o had the same amplitude, but the duration of the current was slightly shorter for the 1 mM[K⁺]_o waveform than for the 5 mM[K⁺]_o waveform. I_Ca for the 5 mM[K⁺]_o waveform was 14.5 pC, compared with 12 pC for the 1 mM[K⁺]_o waveform, a reduction of about 17%. Peak unloaded cell shortening was reduced to about 38%. In four myocytes, I_Ca with the 1 mM[K⁺]_o action potential waveform was 82 ± 4% of that of the 5 mM[K⁺]_o waveform, while the corresponding peak shortening was only 58 ± 6%. Thus, it is unlikely that the small reduction in Ca²⁺ influx resulting from changes in action potential waveform is the primary mechanism responsible for the negative inotropic effect of reducing [K⁺]_o.

Figure 7C shows the effects of switching from a 5 mM[K⁺]_o action potential waveform to a more prolonged 1 mM[K⁺]_o action potential waveform in a rabbit ventricular myocyte. The 1 mM[K⁺]_o waveform resulted in a slightly shorter resting cell length and an increase in peak cell shortening. Both the peak of the Ca²⁺-dependent membrane current and its duration were increased. I_Ca for the 5 mM[K⁺]_o waveform was 14.4 pC, while that for the 1 mM[K⁺]_o waveform was increased to 16.2 pC. Thus, I_Ca was increased by 12.5% while peak cell shortening was enhanced by 32%.

Prolonged exposure of rabbit ventricular myocytes to low [K⁺]_o often led to oscillations in membrane potential that were followed closely by oscillations in cell length. Figure 7D shows that these oscillations were due to Ca²⁺ entry through voltage-gated Ca²⁺ channels. An action potential from a rabbit ventricular cell which displayed periodic oscillations in membrane potential in 1 mM[K⁺]_o was used as a voltage-clamp command signal. As shown in Fig. 7D, oscillations in membrane potential were associated with the development of a Ca²⁺-dependent membrane current, which had an apparent threshold for activation of about –40 mV, similar to that observed for voltage-dependent Ca²⁺ channels in rabbit ventricular myocytes (Yuan et al. 1996). Substitution of extracellular Ca²⁺ with Mg²⁺ resulted in a complete loss of oscillations in cell shortening. I_Ca during one cycle of the oscillation was 9.9 pC in this cell, suggesting that significant entry of Ca²⁺ through voltage-gated Ca²⁺ channels can occur during membrane potential oscillations in the presence of low [K⁺]_o.

	Discussion

Top Abstract Introduction Methods Results Discussion References

 
Working hypothesis

The working hypothesis for our experimental design and data analysis is illustrated in Fig. 8. In this scheme, alteration of [K⁺]_o results in two primary electrophysiological effects in cardiac myocytes; changes in resting membrane potential and alterations in action potential waveform. As shown in Fig. 1B, both of these changes are dependent on the properties of inwardly rectifying K⁺ current (Fig. 1A, Tables 1 and 2). Consequently, changes in [K⁺]_o may produce distinct effects on both resting membrane potential and action potential waveform in various cardiac tissues. This is consistent with previous reports of action potential heterogeneity observed in atrial and ventricular tissues in a wide range of species (Hume & Uehara, 1985; Giles & Imaizumi, 1988; Shimoni et al. 1992; Clark et al. 1993; Antzelevitch & Dumaine, 2002).

View larger version (13K): [in this window] [in a new window]   Figure 8.  Working hypothesis for the effects of changes in [K+]o on selected Ca2+ influx and Ca2+ efflux pathways in mammalian heart modulated by changes in the inwardly rectifying K+ current IK1 Alterations of [K+]o result in changes in both resting membrane potential (Em) and action potential shape (AP shape) due to IK1. As shown in Fig. 1, the resulting changes in Em and AP shape are dependent on the antecedent effects of changes in [K+]o on IK1 in a given cell type. The inotropic state of the cell is a function of the intracellular Ca2+ concentration ([Ca2+]i), which in turn depends on the functional properties in a given tissue of the sarcolemmal Na+–Ca2+ exchanger (NCX), Ca2+ influx through sarcolemmal Ca2+ channels (ICa) and the ability of the SR to take up and release Ca2+ via Ca2+-induced Ca2+ release (CICR). Therefore, perturbation of [K+]o can either increase or decrease sarcolemmal Ca2+ influx and/or intracellular Ca2+ release, and the corresponding inotropic effect is a function of the relative effects of changes in Em and AP shape on NCX, ICa and CICR.

Changes in resting membrane potential

The present study demonstrates that the inotropic effect of changing [K⁺]_o is strongly tissue-dependent, with additional variability between species and within a given species. Independently of whether contractility was increased or decreased in response to changes in [K⁺]_o, one feature common to all myocytes studied was that alteration of [K⁺]_o resulted in marked changes in resting membrane potential. Reduction of [K⁺]_o led to a rapid and marked hyperpolarization of diastolic potential in rat ventricular cells (Fig. 2) and in rabbit atrial and ventricular myocytes (Fig. 3). This was associated with a negative inotropic effect in rat ventricular and rabbit atrial myocytes and a positive inotropic effect in rabbit ventricular myocytes.

The sarcolemmal Na⁺–Ca²⁺ exchanger is known to be the major mechanism responsible for removing Ca²⁺ from cardiac cells (Hryshko, 2002; Bers et al. 2003). We have shown previously (Bouchard et al. 1993a) that hyperpolarization of diastolic potential results in a marked negative inotropic effect in rat ventricular myocytes. Because this effect was abolished following exposure of myocytes to a so-called ‘exchange inhibitory peptide’ XIP (Li et al. 1991), the negative inotropic effect of membrane hyperpolarization was deemed to be due to modulation of diastolic Ca²⁺ efflux via Na⁺–Ca²⁺ exchange. This inhibition of Na⁺–Ca²⁺ exchange increased the amount of SR Ca²⁺ loading and release during repetitive stimulation (Bouchard et al. 1993b, 1995; Sah et al. 2003). The results in Fig. 4 extend this observation to rabbit atrial and ventricular myocytes, and show that changes in cell shortening induced by changes in holding potential are accompanied by perturbation of [Ca²⁺]_i, both under resting conditions and during Ca²⁺ transients (Fig. 5).

Figure 2 shows that reduction of [Na⁺]_o reversed the negative inotropic effect following reduction of [K⁺]_o in rat ventricular myocytes. This finding provides indirect evidence in favour of the involvement of diastolic Na⁺–Ca²⁺ exchange under these conditions. Consistent with this hypothesis, the inotropic effects observed with changes in [K⁺]_o following were abolished in rat and rabbit ventricular and atrial myocytes when the holding potential was held constant under whole-cell voltage-clamp conditions (Fig. 6). Similar observations have been made in guinea-pig ventricular myocytes (White & Terrar, 1991). Combined with the negative inotropic effect of membrane hyperpolarization (Fig. 4), altered Na⁺–Ca²⁺ exchange activity coupled to an I_K1-induced shift in the diastolic potential is the mechanism likely to be responsible for much of the negative inotropic effect following reduction of [K⁺]_o in rat ventricular and rabbit atrial myocytes. By contrast, the positive inotropic effect of lowering [K⁺]_o in rabbit ventricular myocytes appears to be related to the increased Ca²⁺ influx during prolonged action potentials (Fig. 7), which apparently can offset the increase in Ca²⁺ efflux mediated by the Na⁺–Ca²⁺ exchanger at negative membrane potentials. These effects will be influenced by the compliment of K⁺ channels expressed in each of the three types of myocytes, as will determine the action potential waveform in each tissue. For example, according to the APD₉₀ data in Table 2, rat ventricle, rabbit atrium and rabbit ventricle will spend 98.5, 95 and 90% of their respective duty cycles in diastole under these experimental conditions. This underscores the importance of diastolic Ca²⁺ fluxes mediated by the Na⁺–Ca²⁺ exchanger in modulating cardiac contractility.

Elevation of [K⁺]_o led to depolarization of the resting potential, and a negative inotropic effect and cessation of mechanical activity in rabbit ventricular and atrial myocytes, respectively. Of interest was the observation that depolarization of the resting potential was far less substantial in rabbit ventricle than for rabbit atria, as was the negative inotropic effect. As illustrated in Fig. 4, the latter effect is likely to be due in part to differences in the cell shortening–voltage relationship in the two cell types. The relationship between holding potential and cell shortening is much flatter at positive potentials in rabbit ventricle compared with either rabbit atrium or rat ventricle. Thus, even though elevation of [K⁺]_o, will result in an increase in the amount of time at positive diastolic potentials in all three types of myocytes, the differences in the voltage dependence of cell shortening at more positive holding potentials (though still negative to the reversal potential for Na⁺–Ca²⁺ exchange, E_NaCa) in rabbit ventricle compared with rabbit atrial or rat ventricular cells would probably result in a comparative reduction in SR Ca²⁺ loading and release. This effect may be enhanced by the greater dependence of contraction on sarcolemmal Ca²⁺ influx compared to SR Ca²⁺ release in rabbit ventricle compared with rat ventricle (Bassani et al. 1994) and by the apparently limited ability of the Na⁺–Ca²⁺ exchanger to move Ca²⁺ into the cell at depolarized potentials via reverse mode exchange in rabbit compared with rat ventricular myocytes (Su et al. 1999; Bers, 2001).

Changes in action potential waveform

Differences in action potential waveforms amongst the three types of myocytes result from differences in the type and magnitude of other voltage-gated K⁺ currents in addition to I_K1. In rat ventricular myocytes, a large transient outward K⁺ current (I_to) underlies the rapid initial repolarization phase of the action potential, while a more slowly activating delayed rectifier K⁺ current contributes to the slower late phase of repolarization (Apkon & Nerbonne, 1991). Rabbit atrial and ventricular myocytes both express I_to currents which contribute to the initial phase of repolarization but the density of this current is much larger in atrial than in ventricular cells (Giles & Imaizumi, 1988), which accounts in part for the small initial phase of repolarization of ventricular cells. Both cells also have a rapidly activating delayed rectifier current, I_Kr (Muraki et al. 1995; Mitcheson & Hancox, 1999). This current plays an important role in repolarization during the plateau phase of the ventricular action potential, but is less important in atrial cell repolarization. The response of the action potential waveform to changes in [K⁺]_o for each cell type will differ, depending on the relative magnitudes of the different types of currents and their response to changes in [K⁺]_o.

In rabbit ventricular myocytes, a decrease in [K⁺]_o led to a positive inotropic effect that was accompanied by a substantial prolongation of the action potential (Fig. 3). This prolongation may have resulted partly from a decrease in the magnitude of I_Kr, which like I_K1, is reduced by reduction in [K⁺]_o (Yang & Roden, 1996). In some experiments, oscillations in membrane potential developed, and these were accompanied by oscillations in cell shortening (Fig. 3A). Even though the initial phase of repolarization remained unaffected, the marked prolongation of the later stages of action potential was associated with an increase in Ca²⁺ influx (Fig. 7C). This would be expected to gradually increase SR Ca²⁺ loading and release (Bouchard et al. 1995; Delbridge et al. 1996, 1997; Sah et al. 2003), and thus contractility. The complete loss of oscillations in cell length following substitution of external Ca²⁺ with Mg²⁺ (Fig. 7D), suggests that enhancement of Ca²⁺ influx under these conditions was probably due to an increase in Ca²⁺ influx through voltage-gated Ca²⁺ channels.

In contrast to rabbit ventricular cells, decreasing [K⁺]_o resulted in a so-called ‘anomalous’ inotropic effect in rat ventricular (Fig. 2) and rabbit atrial (Fig. 3) myocytes. This effect was accompanied by hyperpolarization of the resting potential but very little effect on action potential waveform. Indeed, Fig. 7 shows that Ca²⁺ influx in rat ventricular myocytes was slightly reduced in low [K⁺]_o and that this change was relatively small compared with the accompanying decrease in cell shortening. This discrepancy between Ca²⁺ influx and peak shortening was observed irrespective of whether cells were stimulated in voltage clamp with step depolarizing pulses (Fig. 7A) or with action potential waveforms recorded in low [K⁺]_o solutions (Fig. 7B). Thus, in distinction to rabbit ventricular cells, reducing [K⁺]_o either had minimal effects or slightly decreased Ca²⁺ influx in rat ventricular myocytes. A reduction of Ca²⁺ influx, along with an increase in Ca²⁺ efflux via Na⁺–Ca²⁺ exchange at negative potentials, could account for the negative inotropic effect of lowering [K⁺]_o in rat ventricle under the present conditions. Given that reduction of [K⁺]_o produces a small decrease in action potential duration in rabbit atrial myocytes (Fig. 3C), it is reasonable to conclude that the effect of low [K⁺]_o on Ca²⁺ influx is similar to that in rat ventricular myocytes.

Figure 3 shows that increasing [K⁺]_o results in a small negative inotropic effect in rabbit ventricular myocytes. The negative inotropic effect in rabbit ventricle was accompanied by small reductions in both the late phase of repolarization as well as in the initial phase of repolarization. Thus, Ca²⁺ influx via voltage-gated Ca²⁺ channels is mostly likely depressed during the action potential due to the concomitant decrease in the number of available Ca²⁺ channels following depolarization of the resting potential expected under these conditions.

Input resistance

An important consideration relating to the effects of changes in [K⁺]_o on contractility mediated by I_K1 is the effects of this manipulation on membrane input resistance and K⁺ slope conductance. The degree to which an ionic current will charge the membrane capacitance and alter membrane potential is inversely related to R_input and positively related to the K⁺ slope conductance, both of which are dependent on [K⁺]_o. If the slope conductance is high, as observed for rabbit and rat ventricular cells with increased [K⁺]_o (Fig. 1), then the amount of inward current required to depolarize the membrane potential will be relatively large (Aronson & Nordin, 1988; Golod et al. 1998; Mendez & Hernandez, 2001). This is due in large part to the dominance of I_K1 in setting the resting membrane potential and the shape of the total current–voltage relationship in cardiac cells (Hume & Uehara, 1985; Giles & Imaizumi, 1988). Opposite effects would be expected for decreased [K⁺]_o.

Table 1 shows that there is almost a 20-fold difference between the largest (rabbit atrium) and smallest (rabbit ventricle) R_input values in the three cell types studied. Given the expected effects of K⁺ slope conductance on the amount of current required to generate a given change in membrane voltage, the rank order for the largest membrane potential deviation that would be produced subsequent to injection of a given ionic current would be rabbit atrium >> rat ventricle > rabbit ventricle. This can be seen in the response of rabbit ventricular myocytes to reducing [K⁺]_o (Fig. 3A), where action potential duration is substantially increased due in part to the development of strong afterdepolarizations. This may be a function of a larger inward Na⁺ current due to an increase in Na⁺ channel availability at negative potentials or to increased Ca²⁺ influx via voltage-gated Ca²⁺ channels by the same process. Some evidence for the latter mechanism is shown in Fig. 7C, which illustrates that the Ca²⁺-dependent membrane current is enhanced and prolonged when myocytes are voltage-clamped with an action potential waveform collected under low [K⁺]_o conditions.

Similarly, a very small K⁺ slope conductance is the most likely explanation for entry of rabbit atrial myocytes into a stable depolarized resting potential (–50 mV) and cessation of mechanical activity following a sustained increase of [K⁺]_o (Fig. 3D). This is due to the fact that the slope conductance for I_K1 is very small and changes little with changes of [K⁺]_o (Fig. 1B). Under all conditions examined, very little inward current was required to effect large depolarizing changes in resting membrane potential. Similar results have been reported in human ventricular myocytes, which display stable resting potentials of –78 or –45 mV following a return to 4 mM from higher [K⁺]_o (McCullough et al. 1990). The authors found that entry into the depolarized state was prevented by blockers of voltage-gated Ca²⁺ channels but not by block of sarcolemmal Na⁺ channels or Na⁺–Ca²⁺ exchange, suggesting that inward current carried by membrane Ca²⁺ channels was able to depolarize the resting potential in conjunction with a low slope conductance of inwardly rectifying K⁺ channels, as shown previously in human ventricular cells (Bailly et al. 1998). Thus, the relationship between I_K1, R_input and Ca²⁺ homeostasis can be an important determinant of cardiac function. Previously, it has been shown that reduced I_K1 density in conjunction with up-regulated Na⁺–Ca²⁺ exchange current increases the incidence of delayed after-depolarizations in the context of heart failure or following inhibition of the Na⁺ pump with ouabain (Aronson & Nordin, 1988; Pogwizd et al. 2001).

	References

Top Abstract Introduction Methods Results Discussion References

 
Antzelevitch C & Dumaine R (2002). Electrical heterogeneity in the heart: physiological, pharmacological and clinical implications. In Handbook of Physiology, section 2, The Cardiovascular System, vol. 1, The Heart, ed. Solaro PF, pp. 654–692. American Physiological Society, New York.

Apkon M & Nerbonne JM (1991). Characterization of two distinct depolarization-activated K⁺ current in isolated adult rat ventricular myocytes. J General Physiol 97, 973–1011.[Abstract/Free Full Text]

Aronson RS & Nordin C (1988). Arrhythmogenic interaction between low potassium and ouabain in isolated guinea-pig ventricular myocytes. J Physiol 400, 113–134.[Abstract/Free Full Text]

Baczko I, Giles WR & Light PE (2003). Resting membrane potential regulates Na⁺–Ca²⁺ exchange-mediated Ca²⁺ overload during hypoxia-reoxygenation in rat ventricular myocytes. J Physiol 550, 889–898.[Abstract/Free Full Text]

Bailly P, Mouchoniere M, Benitah JP, Camilleri L, Vassort G & Lorente P (1998). Extracellular K⁺ dependence of inward rectification kinetics in human left ventricular cardiomyocytes. Circulation 98, 2753–2759.[Abstract/Free Full Text]

Bassani JW, Bassani RA & Bers DM (1994). Relaxation in rabbit and rat cardiac cells: species-dependent differences in cellular mechanisms. J Physiol 476, 279–293.[Abstract/Free Full Text]

Bers DM (2001). Excitation-Contraction Coupling and Cardiac Contractile Force, vol. 1. Kluwer Academic Publishers, The Netherlands.

Bers DM, Barry WH & Despa S (2003). Intracellular Na⁺ regulation in cardiac myocytes. Cardiovasc Res 57, 897–912.[Abstract/Free Full Text]

Beuckelmann DJ & Wier WG (1988). Mechanism of release of calcium from sarcoplasmic reticulum of guinea-pig cardiac cells. J Physiol 405, 233–255.[Abstract/Free Full Text]

Blaustein MP & Lederer WJ (1999). Sodium/calcium exchange: its physiological implications. Physiol Rev 79, 763–854.[Abstract/Free Full Text]

Bouchard RA, Clark RB & Giles WR (1993a). Regulation of unloaded cell shortening by sarcolemmal sodium-calcium exchange in isolated rat ventricular myocytes. J Physiol 469, 583–599.[Abstract/Free Full Text]

Bouchard RA, Clark RB & Giles WR (1993b). Role of sodium-calcium exchange in activation of contraction in rat ventricle. J Physiol 472, 391–413.[Abstract/Free Full Text]

Bouchard RA, Clark RB & Giles WR (1995). Effects of action potential duration on excitation-contraction coupling in rat ventricular myocytes. Action potential voltage-clamp measurements. Circ Res 76, 790–801.[Abstract/Free Full Text]

Braunwald E (2001). Harrison's Principles of Internal Medicine., vol. xxxii, 2630. McGraw-Hill, Medical Publishing Division, New York.

Bridge JH, Smolley JR & Spitzer KW (1990). The relationship between charge movements associated with I_Ca and I_Na-Ca in cardiac myocytes. Science 248, 376–378.[Abstract/Free Full Text]

Christe G (1983). Effects of low [K⁺]_o on the electrical activity of human cardiac ventricular and Purkinje cells. Cardiovasc Res 17, 243–250.[Abstract/Free Full Text]

Clark RB, Bouchard RA, Salinas-Stefanon E, Sanchez-Chapula J & Giles WR (1993). Heterogeneity of action potential waveforms and potassium currents in rat ventricle. Cardiovasc Res 27, 1795–1799.[Abstract/Free Full Text]

Delbridge LM, Bassani JW & Bers DM (1996). Steady-state twitch Ca²⁺ fluxes and cytosolic Ca²⁺ buffering in rabbit ventricular myocytes. Am J Physiol 270, C192–C199.[Medline]

Delbridge LM, Satoh H, Yuan W, Bassani JW, Qi M, Ginsburg KS, Samarel AM & Bers DM (1997). Cardiac myocyte volume, Ca²⁺ fluxes, and sarcoplasmic reticulum loading in pressure-overload hypertrophy. Am J Physiol 272, H2425–H2435.[Medline]

Despa S, Islam MA, Pogwizd SM & Bers DM (2002). Intracellular [Na⁺] and Na⁺ pump rate in rat and rabbit ventricular myocytes. J Physiol 539, 133–143.[Abstract/Free Full Text]

Eisner DA, Lederer WJ & Vaughan-Jones RD (1984). The quantitative relationship between twitch tension and intracellular sodium activity in sheep cardiac Purkinje fibres. J Physiol 355, 251–266.[Abstract/Free Full Text]

Gettes LS, Surawicz B & Shiue JC (1962). Effect of high K, and low K quinindine on QRS duration and ventricular action potential. Am J Physiol 203, 1135–1140.[Abstract/Free Full Text]

Giles WR & Imaizumi Y (1988). Comparison of potassium currents in rabbit atrial and ventricular cells. J Physiol 405, 123–145.[Abstract/Free Full Text]

Godfraind T & Ghysel-Burton J (1980). Independence of the positive inotropic effect of ouabain from the inhibition of the heart Na⁺/K⁺ pump. Proc Natl Acad Sci U S A 77, 3067–3069.[Abstract/Free Full Text]

Golod DA, Kumar R & Joyner RW (1998). Determinants of action potential initiation in isolated rabbit atrial and ventricular myocytes. Am J Physiol 274, H1902–H1913.[Medline]

Goto M, Tsuda Y & Yatani A (1977). Two mechanisms for positive inotropism of low-K Ringer solu