Pacing-induced heart failure causes a reduction of delayed rectifier potassium currents along with decreases in calcium and transient outward currents in rabbit ventricle

Abstract

Objective: Heart failure in patients and in animal models is associated with action potential prolongation of the ventricular myocytes. Changes in several membrane currents have been already demonstrated to underlie this prolongation. However, information on the two components (I_Kr and I_Ks) of the delayed rectifier potassium current (I_K) in rapid pacing induced heart failure is lacking. Methods and results: Action potentials and whole-cell currents, I_K, I_to1, I_K1, and I_Ca-L were recorded in apical myocytes of left ventricle from 10 rabbits subjected to left ventricular pacing at 350–380 beats/min for 3–4 weeks and 10 controls with sham operation. Action potential duration at 90% repolarization (APD₉₀) was prolonged in myocytes from failing hearts compared to controls at both cycle lengths of 333 and 1000 ms. Both E-4031-sensitive and -resistant components of I_K (I_Kr, I_Ks) in myocytes from failing hearts were significantly less than those of control hearts; tail current densities of I_Kr and I_Ks following depolarization to +50 mV were 0.62±0.05 vs. 0.96±0.12 pA/pF (P<0.05), and 0.27±0.08 vs. 0.52±0.08 pA/pF (P<0.05), respectively. There was no significant difference between control and failing myocytes in the voltage- and time-dependence of activation of total I_K, I_Kr and I_Ks. The peak of L-type Ca²⁺ current (I_Ca-L) was significantly reduced in myocytes from failing hearts (at +10 mV, −9.29±0.52 vs. −12.28±1.63 pA/pF, P<0.05), as was the Ca²⁺-independent transient outward current (I_to1; at +40 mV, 4.8±0.9 vs. 9.6±1.3 pA/pF, P<0.05). Steady state I–V curve for I_K1 was similar in myocytes from failing and control hearts. Conclusions: Decrease of I_K (both I_Kr and I_Ks) in addition to reduced I_to1, may underly action potential prolongation at physiological cycle length and thereby contribute to arrhythmogenesis in heart failure.

Arrhythmia (mechanisms), Heart failure, Membrane currents, K-channel

Time for primary review 34 days.

This article is referred to in the Editorial by S. Nattel (pages 188–190) in this issue.

1 Introduction

Congestive heart failure carries a grim prognosis. Five year survival is well below 40% [1] with half of deaths classified as sudden [2,3]. These deaths are probably caused by ventricular arrhythmias or by electromechanical dissociation [4–6]. Electrophysiological remodeling is associated with the progression of hypertrophy and heart failure [7]. Heart failure in patients and in several animal models is associated with action potential prolongation of the ventricular myocytes [8–10], albeit that this prolongation is less prominent at physiological cycle length [9]. Action potential prolongation in itself may be expected to decrease the propensity to (reentrant) arrhythmias [11]. However, this no longer holds when the prolongation becomes excessive, thereby leading to early afterdepolarizations [12], or when the prolongation is regional and thereby increases dispersion in action potential duration and refractoriness [13].

One of the available models of heart failure [9,14–16], the rapid pacing model, shows severe reduction in cardiac output, based on impaired contractile function, and increase of systemic vascular resistance, left ventricular wall stress, and of several neurohumoral factors [15,16]. Rapid pacing of the ventricles leads to heart failure within 2–3 weeks in rabbits [17,18], and within 3–4 weeks in pigs [19] and dogs [20,21]. Unfortunately, action potential prolongation associated with heart failure is often assessed at unphysiological long cycle length [18,20,21] and absent at physiological cycle length [18].

Changes in membrane currents underlying prolongation of the ventricular action potential are incompletely understood. The transient outward K⁺ current (I_to1) is downregulated by some 65% in rabbit [18] and in dog [21] and also substantially in myocytes from patients [22] with end-stage heart failure (review [23]). The L-type inward Ca²⁺ current (I_Ca-L) was found to be unchanged in rabbit [18] and in dog [21], whereas a reduction by 40% was demonstrated in the pig [19]. In all three species the responsiveness of I_Ca-L to β-adrenoceptor stimulation was blunted by 40% or more [18,19,21]. By and large, there is a tendency to a decrease in I_Ca-L in the more severe stages of hypertrophy and heart failure (reviews [7,8]). Measurement of membrane currents at room temperature and of action potentials at physiological temperature [19,21] hampers understanding of electrophysiological changes relevant for arrhythmias in heart failure.

The delayed rectifier (I_K), relevant for repolarization [24] and first described by Noble and Tsien [25] in 1969 and later dissected into rapid (I_Kr) and slowly activating components (I_Ks) by Sanguinetti and Jurkiewicz [26], has been demonstrated in guinea-pig [26], dog [27], rabbit [28] and human ventricular myocytes [29]. Information on I_Kr and I_Ks is lacking in pacing-induced models of heart failure [18–21] and in patients with end-stage heart failure [7]. In the present study we demonstrate for the first time downregulation of both I_Kr and I_Ks by about 50% in rabbits with pacing-induced heart failure in association with action potential prolongation at physiological stimulation frequency and at physiological temperature.

2 Methods

Experiments were performed on Japanese white rabbits weighing 2.0 to 3.0 kg. All rabbits were fed and housed according to institutional guidelines at Nagoya University. This investigation conforms with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85–23, revised 1996).

2.1 Animal model

Rabbits were anesthetized with a mixture (1 ml/kg im.) of ketamine hydrochloride (50 mg/kg) and xylazine (5 mg/kg). Supplemental anesthesia was provided by thiamyl sodium (25 mg/kg) iv. They were intubated and ventilated artificially with room air supplemented with oxygen (tidal volume 10 ml/kg; frequency 25 /min). After opening the left fourth intercostal space, part of the pericardium was cut to expose the anterior surface of the heart. An unipolar pacing lead (Medtronic, USA: 6491) fixed to the anterobasal region of the left ventricular free wall was connected to a pacemaker (Medtronic, USA: 5985), which was implanted subcutaneously in the back. The chest was closed in layers. Post-operatively, cefotiam hydrochloride (10 mg/kg) was given for 5–7 days. Two weeks after the operation, i.e. following full recovery from surgery, the left ventricle of each rabbit was paced at 350–380 beats/min, and pacing was continued for 3–4 weeks. The pacemaker implantation was carried out in a total of 26 rabbits; 16 were paced to induce heart failure, the remaining 10 were unpaced as controls. All 16 paced rabbits showed physical signs of heart failure 3–4 weeks after the initiation of pacing. Six rabbits died during the observation period probably due to serious pump failure. The remaining 10 rabbits were sacrificed for electrophysiological experiments. All 10 control rabbits survived the observation period.

To monitor the progress of heart failure during the observation period, two-dimensional echocardiography was carried out periodically together with recordings of scalar electrocardiograms (ECGs). The recordings were performed under anesthesia by a mixture (1 ml/kg im.) of ketamine hydrochloride (30 mg/kg) and xylazine (5 mg/kg). The pacing was turned off transiently for the recordings in sinus rhythm. Electrophysiological experiments in single ventricular myocytes from paced hearts were carried out as soon as fractional shortening (FS) was <22%.

2.2 Electrophysiological experiments

Myocytes were isolated enzymatically from the apex of the left ventricle by the same procedure as recently described [28]. In brief, the hearts were perfused on a Langendrorff apparatus with normal Tyrode's solution (gassed with 100% O₂ at 34°C) for 3–5 min, then with Ca²⁺-free Tyrode's solution for 10–15 min and finally with 0.12 mg/ml collagenase (Yakult, Japan) containing Ca²⁺-free solution for 15 min. The hearts were subsequently washed with high-K⁺ storage solution (Kraftbrühe solution: KB solution) for 5 min. The apical region of the left ventricle was separated and minced in the KB solution to obtain a suspension of myocytes. The standard whole-cell patch-clamp method was used for recording the membrane potential and currents.

An aliquot of the cell suspension was placed in a recording chamber on the stage of an inverted microscope (Diaphoto, Nikon Co., Tokyo). The cells were superfused with normal Tyrode's solution at 3 ml/min. The bath temperature was maintained at 34°C. Suction pipettes of borosilicate glass had a resistance of 3–5 MΩ after filling with the pipette solution. Junction potential of the pipette was 4–10 mV (5.6 mV in average). Resting membrane potential recorded by the current clamp mode was corrected for the value. The cell capacitance was measured by small voltage steps from a holding potential of −50 mV. The cell capacitance and series resistance (R_S) were compensated for by about 50–70%, giving rise to the values of R_S at 2–5 MΩ (4.2 MΩ in average) and the time constant of capacitive surge at 450–500 μs (479 μs in average). Command potentials generation and data acquisition were performed with a patch-clamp amplifier (Axopatch 200B) controlled by pCLAMP software (Axon Instruments) and a PC. Action potentials were recorded in current-clamp mode and were elicited by injecting 5 ms long rectangular pulses of depolarizing current through the pipette. Voltage signals were displayed on an oscilloscope. Current signals were filtered at 1 kHz, and digitized at a sampling frequency of 2 kHz.

When the delayed rectifier K⁺ current (I_K) was measured, the myocytes were superfused with Na⁺-free, K⁺-free solution including N-methyl-d-glucamine (NMG) to eliminate relatively large inward rectifier K⁺ current (I_K1). It has been proposed that the rapidly activating component of I_K (I_Kr) is greatly suppressed in extracellular K⁺-free condition in some animal species such as guinea pigs and dogs [27,30]. However, in our previous study [28] on I_Kr of rabbit ventricular myocytes, we showed that although the current density during depolarization in NMG solution was significantly larger than in Tyrode's solution, there was no significant difference in the tail current density between the two solutions. The two components of I_K, I_Kr and I_Ks, were pharmacologically separated by the application of a selective blocker, E-4031 (5 μM). The E-4031-sensitive component was obtained by digital subtraction of the E-4031-resistant component from the total I_K current. The E-4031-sensitive and -resistant components were measured as indexes of I_Kr and I_Ks, respectively. The E-4031-resistant component was confirmed as I_Ks by elimination with 30 μM chromanol-293B, as has been shown by Cheng et al [28] in rabbit ventricular myocytes.

I_to1 and I_K1 were measured in Tyrode's solution including 3 μM nisoldipine for blockade of I_Ca-L. In experiments to measure I_Ca-L, KCl in the external (Tyrode) solution and pipette solution was replaced by CsCl to avoid the contamination of K⁺ currents. For comparison of currents derived from control and failing hearts, current amplitudes were normalized to the cell capacitance.

2.3 Solutions

Tyrode's solution for cell isolation and the recording of action potentials was composed of (in mM): NaCl, 143; KCl, 5.4; CaCl₂, 1.8; MgCl₂, 0.5; NaH₂PO₄, 0.25; HEPES, 5.0; and glucose, 5.6, pH adjusted to 7.4 with NaOH. The Ca²⁺-free Tyrode's solution was as above without CaCl₂. The KB solution contained (in mM): L-glutamic acid, 50; KCl, 40; KH₂PO₄, 20; taurine, 20; HEPES, 10; MgCl₂, 3; glucose, 10; EGTA, 0.5, pH adjusted to 7.4 with KOH. The Na⁺-free, K⁺-free NMG solution contained (in mM): N-methyl-d-glucamine, 149; MgCl₂, 5; CaCl₂, 0.65; HEPES, 5 and nisoldipine, 0.003, pH adjusted to 7.4 with HCl. The internal pipette solution contained (in mM): KOH, 50; KCl, 80; aspartate, 40; HEPES, 5; EGTA, 10; MgATP, 5; sodium creatinine phosphate, 5; and CaCl₂, 0.65 (pH 7.2, pCa 8.0). The pipette solution for the measurement of I_Ca-L contained (in mM): CsCl, 80; CsOH, 60; aspartate, 40; HEPES, 5; EGTA, 10; MgATP, 5; sodium creatinine phosphate, 5; CaCl₂, 0.65 (pH 7.2, pCa 8.0).

2.4 Statistics

The pCLAMP program (Axon Instruments) was used in data analysis. Data were expressed as mean±SEM. Statistical analysis was performed using unpaired Student's t-test. Values of P<0.05 were considered significant.

3 Results

3.1 Characteristics of heart failure

Long-term (3–4 weeks) rapid ventricular pacing decreased fractional shortening, increased end-systolic and end-diastolic diameter, and left atrial diameter. It reduced the wall thickness of the interventricular septum and the posterior wall of the left ventricle (Table 1). Table 2 summarizes the ECG data obtained from the failing and control rabbits. The QT interval was corrected (QTc) according to the method of Carlsson et al [31] for rabbits with the formula QT_c=QT−0.175(RR-300). In the failing rabbits 3–4 weeks after pacing, the QT interval and the QT_c interval were significantly prolonged by 9.2% and 6.8%, respectively from the baselines before pacing. QRS duration was unchanged after pacing. Heart rate (HR) after pacing tended to be increased, but the difference from the baseline did not reach statistical significance. For comparison, the control rabbits showed no changes in the measured ECG parameters after the corresponding observation period. All failing rabbits had considerable amounts of ascites and pleural effusion. They showed increased heart, lung and liver weight both absolute and relative to body weight (Table 3).

Table 3

Animal characteristics of control and paced rabbitsa

	Control (n = 10)	Paced (n = 10)
BW (kg)	2.7±0.5	2.5±0.7	n.s.
HW (g)	13.6±3.9	18.8±4.8	*
HW/BW (g/kg)	5.02±0.71	7.12±1.21	**
Lung weight (g)	11.9±2.2	14.4±2.9	*
Lung weight/BW (g/kg)	4.62±0.87	5.89±1.05	*
Liver weight (g)	68.1±13.9	90.0±22.3	*
Liver weight/BW (g/kg)	27.12±5.68	36.82±8.35	*
Ascites	0/10	10/10	**
Pleural effusion	0/10	10/10	**

a

Values are mean±S.E.M. BW=body weight; HW=heart weight; HW/BW=heart weight to body weight ratio; n = number of rabbits.

*

P<0.05.

**

P<0.01 denotes significant difference from control.

Table 2

ECG data of paced and control rabbitsa

	Baseline	After 3–4 weeks
Paced rabbits (n = 10)
Heart rate (beats/min)	215±13	225±11	n.s.
QRS duration (ms)	40.3±1.0	40.2±0.7	n.s.
QT interval(ms)	140.7±3.7	153.7±3.6	*
QTc interval (ms)	145.8±3.6	155.7±3.3	*
Control (n = 10)
Heart rate (beats/min)	205±17	202±10	n.s.
QRS duration (ms)	40.1±0.5	40.0±0.5	n.s.
QT interval (ms)	141.3±3.6	141.4±2.9	n.s.
QTc interval (ms)	146.8±3.5	146.6±3.1	n.s.

a

Values are mean±S.E.M. QT_c=corrected; QT=interval; n = number of rabbits.

*

P<0.05 denotes significant difference from baseline.

Table 1

Echocardiographic findings in control and paced rabbitsa

	Control (n = 10)	Paced (n = 10)
FS (%)	59.1±0.1	20.4±0.1**
LVED diameter (mm)	14.5±0.2	17.0±0.6**
LVESdiameter (mm)	8.3±0.3	14.7±0.8**
IVS (mm)	2.9±0.1	2.3±0.1**
PW (mm)	3.0±0.1	2.3±0.2**
LA diameter (mm)	9.0±0.3	13.8±0.3**

a

Values are mean±S.E.M. FS=fractional shortening; LVED=left ventricular end-diastolic; LVES=left ventricular end systolic; IVS=interventricular septum; PW=posterior wall of left ventricle; LA=left atrium; n = number of rabbits.

**

P<0.01 denotes significant difference from control.

3.2 Action potential changes in heart failure

Action potentials in myocytes from failing hearts differed in shape and duration compared with control hearts. Fig. 1A shows traces at 333 ms and 1000 ms cycle length (CL). Control myocytes typically have a prominent notch in phase 1 at 333 ms and 1000 ms, which is markedly attenuated in myocytes from failing hearts. Action potential duration (APD₉₀), measured at 90% of repolarization, was prolonged in myocytes from failing hearts compared with control hearts at both 333 and 1000 ms (Fig 1A, Table 4). Fig. 1B shows the relationship between CL and APD₉₀. With increasing CLs, the difference in APD₉₀ between normal and failing myocytes became more pronounced. Action potential duration, measured at 50% repolarization, resting membrane potential and action potential amplitude did not differ between the two groups.

Fig. 1

Frequency response of APD prolongation A: Representative action potentials recorded in ventricular myocytes from a control (C) and a failing heart (HF). Records were obtained at 333 ms (left) and 1000 ms (right) of cycle length (CL). Action potentials recorded from myocytes isolating from failing hearts were significantly prolonged compared with control, with attenuation in the amplitude of the notch in phase 1 at both cycle lengths. The resting membrane potential was not changed. B: Relationship of cycle length and action potential duration at 90% of repolarization (APD₉₀). APD₉₀ in failing myocytes (●, n = 25 cells from 8 rabbits) was significantly increased compared with control (○, n = 20 cells from 6 rabbits) at all studied cycle length.

Table 4

Action potential characteristics of myocytes from control and failing heartsa

CL (ms)	Control (n = 20 cells from 6 rabbits)
	APD50 (ms)	APD90 (ms)	RMP (mV)	APA (mV)	NA (mV)
333	138.5±5.8	169.8±6.5	−78.7±1.4	125.2±1.8	8.0±1.2
1000	193.2±5.4	225.3±6.3	−79.2±0.9	127.2±0.7	9.2±0.7
	Heart failure (n = 25 cells from 8 rabbits)
	APD50 (ms)	APD90 (ms)	RMP (mV)	APA (mV)	NA (mV)
333	158.6±12.5	201.2±13.5*	−79.7±1.1	126.5±2.4	–
1000	212.3±16.3	259.6±18.8*	−78.8±1.4	128.0±1.3	–

a

Values are mean±S.E.M. CL=cycle length; RMP=resting membrane potential; APA=action potential amplitude; NA=notch amplitude; n = number of rabbits.

*

P<0.05 denotes significant difference from control.

The cell capacitance – an estimate for cell surface – was comparable in myocytes isolated from control (165±6.6 pF, n = 71) and failing (170±8.9 pF, n = 58) hearts.

3.3 Delayed rectifier in control and failing myocytes

We studied the two components of I_K in control and failing hearts. At first, total I_K was activated by applying voltage clamp steps for 1.5 s at 0.1 Hz from a holding potential of −50 mV to a range of depolarizing levels between −40 mV and +50 mV. E-4031 (5 μM) was then applied to see the E-4031-resistant component (I_Ks). The measurements of I_Ks were carried out twice at 10 and 15 min after application of E-4031 to monitor time-dependent rundown of I_Ks. Data obtained from myocytes showing I_Ks reduction >10% from 10 to 15 min were not included in the results. The E-4031-sensitive component was obtained as an index of I_Kr by digital subtraction of total I_K and I_Ks estimated 10 min after application of E-4031.

Representative tracings are shown in a control and a failing myocyte in Fig. 2. In the failing myocyte, amplitudes of the time-dependent outward current during depolarization (I_Kstep) and the tail current on repolarization (I_Ktail) of total I_K were smaller than those in the control myocyte. The corresponding amplitude of the E-4031-resistant component (I_Ks) and those of the E-4031-sensitive component (I_Kr) of the failing myocyte were also smaller than the control myocyte. The E-4031-sensitive current had relatively large I_Ktail compared with I_Kstep.

Fig. 2

Heart failure reduces the delayed rectifier K⁺ currents. Total I_K (upper panel) was elicited by applying depolarizing potentials from a holding potential of −50 mV to various levels ranging from −40 to +50 mV for 1500 ms. E-4031 (5 μM) was used to separate the two components. The E-4031-sensitive component (I_Kr, lower panel) was obtained by digital subtraction of the E-4031-resistant component (I_Ks, middle panel) from total I_K. Representative current tracings from a control (left) and a failing myocyte (right). Arrows indicates zero current level.

Fig. 3A shows the current–voltage (I–V) relationship of I_Kstep (upper panel) and I_Ktail (lower panel) of total I_K in control and failing myocytes. The current amplitude was normalized to cell capacitance and plotted as a function of test potentials. Densities of I_Kstep and I_Ktail in failing myocytes were both significantly lower than those in control myocytes from +10 mV to more positive potentials (at +50 mV, 0.67±0.14 vs. 1.34±0.24 pA/pF for I_Kstep; 0.87±0.12 vs. 1.47±0.20 pA/pF for I_Ktail). Fig. 3B illustrates the I–V relationship of I_Kstep and I_Ktail of the E-4031-resistant component (I_Ks). Densities of I_Kstep and I_Ktail in failing myocytes were significantly lower than those in control myocytes at +30 mV to more positive potentials (at +50 mV, 0.47±0.11 vs. 0.81±0.13 pA/pF for I_Kstep; 0.27±0.08 vs. 0.52±0.08 pA/pF for I_Ktail). Fig. 3C illustrates the I-V relationship of I_Kstep and I_Ktail of the E-4031-sensitive component (I_Kr). Densities of I_Kstep and I_Ktail in failing myocytes were significantly lower than those in control myocytes at +20 mV to more positive potentials (at +50 mV, 0.20±0.08 vs. 0.53±0.10 pA/pF for I_Kstep; 0.62±0.07 vs. 0.96±0.12 pA/pF for I_Ktail).

Fig. 3

The average current-voltage relation of delayed rectifier current (I_K) plotted for failing myocytes (●, n = 19 cells from 10 rabbits) and control (○, n = 14 cells from 10 rabbits). A: total I_K, B: E-4031-resistant component (I_Ks), C: E-4031-sensitive component (I_Kr). Upper and lower panels are time-dependent outward current (I_Kstep) and tail current (I_Ktail), respectively. *P<0.05 vs. control.

Voltage-dependence of activation was determined by normalizing I_Ktail at each test potential to the current at the most positive test potential, and analyzed by Boltzmann fits. Half activation voltage of total I_K was −7.6±3.5 mV in failing myocytes (n = 19), and −6.3±2.3 mV in controls (n = 14) (n.s.). Corresponding slope factors were 11.8±1.9 mV in failing myocytes, and 13.7±1.0 mV in controls (n.s.). Half activation voltage of the E-4031-resistant component (I_Ks) was 0.2±2.2 mV in the failing myocytes, and 1.2±1.9 mV in the controls (n.s.). Corresponding slope factors were 14.3±2.2 mV in the failing myocytes, and 12.4±1.5mV in the controls (n.s.). Half activation voltage of the E-4031-sensitive component (I_Kr) was −9.4±3.9 mV in the failing myocytes, and −7.1±1.5 mV in the controls (n.s.). Corresponding slope factors were 8.9±2.1 mV in the failing myocytes, and 11.7±1.5 mV in the controls (n.s.).

The activation of the E-4031-resistant component (I_Ks) at +50 mV was approximated by a single exponential function with a time constant of 495±64 ms in control (n = 6) and 513±50 ms in failing myocytes (n = 6, n.s.). Deactivation of I_Ks tail following the depolarization to +50 mV was fitted by a dual exponential function. The fast and slow time constants for the deactivation at −50 mV in failing myocytes (200±60 and 1100±180 ms, n = 6) were similar to those in control myocytes (185±45 and 952±189 ms, n = 6; both n.s.). The activation of the E-4031-sensitive component (I_Kr) at 0 mV was approximated by a single exponential function with a time constant of 401±62 ms in control (n = 6) and 385±42 ms in failing myocytes (n = 6, n.s.). Deactivation of I_Kr tail following the depolarization to +50 mV was fitted by a dual exponential function. The fast and slow time constants for the deactivation at −50 mV in failing myocytes (90±28 and 555±110 ms, n = 6) were significantly longer than those in control myocytes (45±11 and 350±74 ms, n = 6; both P<0.05).

3.4 L-type Ca²⁺ current (I_Ca-L)

I_Ca-L was examined by application of voltage-clamp steps for 200 ms from a holding potential of −50 mV to different depolarizing levels up to +40 mV at 0.1 Hz. Fig. 4 shows representative traces of I_Ca-L (upper panel) and the current-voltage relationship of the peak I_Ca-L density in the control and the failing myocytes (lower panel). I_Ca-L density from −20 mV to +20 mV was significantly decreased in the failing myocytes.

Fig. 4

I_Ca-L is decreased in heart failure. I_Ca-L was measured in Tyrode's solution, in which KCl was replaced by cesium chloride. It was elicited by applying depolarizing potentials from a holding potential of −50 mV to various levels ranging from −40 to +40 mV for 200 ms. Upper panel: representative current tracing of I_Ca-L in a control (left) and a failing myocyte (right). Lower panel: the average current-voltage relation of I_Ca-L plotted for failing myocytes (●, n = 10 cells from 8 rabbits) and control (○, n = 8 cells from 6 rabbits). *P<0.05 vs. control.

3.5 Transient outward current (I_to1) and inward rectifier K⁺ current (I_K1)

Whole-cell currents were elicited by application of 300 ms hyperpolarizing or depolarizing voltage steps between −100 mV and +40 mV from a holding potential of −50 mV to inactivate Na⁺ channel at 0.1 Hz. I_to1 was measured as the difference between the peak outward current and the sustained current at the end of the depolarizing pulse. Fig. 5 shows representative traces (upper panel) of I_to1 and the current-voltage relationship (lower panel) of I_to1 density in control and failing myocytes. I_to1 density was significantly decreased in failing myocytes from −10 mV to more positive voltage and I_to1 was reduced by about 50% at all voltages. Time course of I_to1 inactivation was similar for control and failing myocytes.

Fig. 5

I_to1 is decreased in heart failure. I_to1 was measured in Tyrode's solution, with 3 μM nisoldipine for blockade of I_Ca-L. It was elicited by applying depolarizing potentials from a holding potential of −50 mV to various levels ranging from −40 to +40 mV for 300 ms. Upper panel: representative current tracing of I_to1 in a control (left) and a failing myocyte (right). Lower panel: the average current-voltage relation of I_to1 plotted for failing myocytes (●, n = 10 cells from 8 rabbits) and control (○, n = 11 cells from 6 rabbits). *P<0.05 vs. control.

To elicit I_K1, 300 ms hyperpolarizing or depolarizing voltage steps between −100 and −20 mV were applied from a holding potential of −40 mV at 0.1 Hz (Fig. 6). The steady-state currents at the end of the clamp pulses were measured as an index of I_K1. There was no significant difference in the density of I_K1 between control and failing myocytes.

Fig. 6

I_K1 is unchanged in heart failure. To elicit I_K1, 300 ms hyperpolarizing or depolarizing voltage steps between −100 and −20 mV were applied from a holding potential of −40 mV. The steady-state currents at the end of the clamp pulses were measured as an index of I_K1. Upper panel: representative current tracing of I_K1 in a control (left) and a failing myocyte (right). Lower panel: the average current-voltage relation of I_K1 plotted for failing myocytes (●, n = 10 cells from 8 rabbits) and control (○, n = 7 cells from 5 rabbits).

4 Discussion

In the present study of pacing-induced heart failure in rabbits, we observed a substantial reduction in I_K density. We also observed a decrease in I_to1 and I_Ca-L densities, but noted unchanged I_K1 density. Both E-4031-resistent and -sensitive components of I_K (I_Ks and I_Kr) were reduced in failing myocytes compared with control myocytes. The changes in membrane currents were assessed at physiological temperature and were associated with prolonged action potentials at physiological cycle length as well as prolongation of QT and QTc intervals in ECG.

4.1 Downregulation of membrane currents in heart failure

Although a decrease in -unseparated- delayed rectifier current has been demonstrated in ventricular myocytes obtained from cats with hypertrophy resulting from pulmonary [32] and abdominal banding [33], our study is the first to show a substantial (about 50%) reduction in both components of the delayed rectifier in an animal model of heart failure based on rapid pacing. Especially the downregulation of I_Ks, which constitutes a current with marked regional dispersion even in normal hearts (see below), may contribute to increased dispersion in action potential duration, and related refractory periods and calcium transients [34] in heart failure. Downregulation of I_Kr was also previously demonstrated in cardiomyopathic hamsters, in which I_Ks was absent before and after the development of heart failure [35]. Interestingly, it was recently shown in a canine model with chronic atrioventricular block, hypertrophy and acquired Torsades de Pointes that the density of I_Ks was decreased by about 50% in isolated myocytes from left and right ventricles, compared to control. The density of I_Kr was decreased by 45%, but only in myocytes from the right ventricle [36]. In atrial myocytes isolated from dogs with pacing-induced heart failure, Li et al. [37] have shown a significant reduction of I_Ks density along with decreased I_Ca-L and I_to1 densities.

Our data on downregulation of I_Ca-L are in agreement with those in the pig [19]. Recently, it was demonstrated that this downregulation starts as early as after 1 week of rapid pacing [38]. There were three differences between our study and another study in the rabbit of Rozanski et al. [18] In our model we observed (i) a decrease in I_Ca-L density, (ii) prolongation of the terminal phase of the action potential at physiological cycle length and (iii) cardiomegaly, whereas all these three changes were absent in the study of Rozanski et al. [18] Probably the longer duration of the period of pacing in our study compared to the study of Rozanski et al. [18] caused these differences. The absence and presence of changes in I_Ca-L density in models with respectively less severe [18] and more severe heart failure (this study) is in line with previous data (reviews [7,8]).

4.2 Physiological implications

Both delayed rectifier currents have been described in dog ventricle [27,39], in rabbit ventricle [28], in guinea pig ventricle [40] and in human ventricle [29]. I_Kr has also been demonstrated at the single channel level in rabbit ventricle [41]. Interestingly, I_Ks distribution in the ventricle is regionally different and may contribute to the dispersion of repolarization [42]. First, transmural left ventricular differences in density of I_Ks have been demonstrated in dog with a weaker I_Ks in M-cells than in subendocardial and subepicardial cells [27] and in guinea pig with lower density in subendocardial compared to subepicardial cells [40]., Secondly, I_Ks is larger at the base than at the apex of the rabbit ventricle whereas I_Kr is larger at the apex than at the base [28]. Thirdly, I_Ks is about two times larger in right than left canine ventricular midmyocardial cells [39].

Animal studies are still very useful, because the picture in human studies is far from clear. In ventricular myocytes from normal donor hearts not used for transplantation, Beuckelmann et al. [22] were unable to show delayed rectifier current. Iost et al. [43] described I_Kr, but no I_Ks in normal human ventricular myocytes. The occurrence of I_Kr and I_Ks was described in myocytes from apparently normal right ventricles which were obtained from failing explanted hearts [30]. Whether or not these hearts may be considered normal is crucial for conclusions, because, Beuckelmann et al. [22] did show some delayed rectifier current in about half of myocytes from failing hearts. In cells from hearts with end-stage dilated and ischemic cardiomyopathy, I_Kr but no I_Ks, could be demonstrated at the single channel level [44]. This may be due to much smaller single channel conductance of I_Ks[45]. Because delayed rectifier current is very sensitive to cell isolation artifacts [46], definite conclusions with respect to heart failure in man cannot be made until more data are available.

Kääb et al. [21] showed previously that inhibition of I_to1 converts a ‘normal action potential’ into a ‘failing action potential’ in myocytes from normal canine ventricle, whereas current injection during a brief 8 ms period directly following the upstroke of an action potential -mimicking the effect of activation of I_to1 -reverts a ‘failing action potential’ into a ‘normal action potential’ in myocytes from failing canine ventricle. This suggests an important role for I_to1 in action potential changes during the progression of heart failure. Indeed, downregulation of I_to1 appears to present a common factor in all animal models of heart failure and also in end-stage heart failure in patients [7,18–23,27,29,35]. However, these observations are often made at very long cycle length and I_to1 is strongly rate dependent. At high heart rates it may fail to recover from inactivation almost completely [47] and its behavior is very much different at steady state cycle length and after sudden changes in cycle length [48]. Recovery from inactivation is actually so slow that in rabbit atrium a transition of 0.1 to 4 Hz will leave peak I_to current intact for only 3% [49]. We wish also to emphasize that Kääb et al. [21] noticed that despite the substantial downregulation of I_to1 in myocytes from canine ventricle, these cells were more sensitive to 4-aminopyridine, a blocker of I_to, than myocytes from normal hearts. This may be explained if the alternative currents for final repolarization in heart failure, I_Kr and I_Ks, are very small in heart failure.

4.3 Limitations

In rabbits, unlike guinea pigs and dogs, activation time constants of I_Kr and I_Ks are in a similar range (300–600 ms) [28]. Accordingly, it is difficult to discriminate the two components using a protocol of different pulse duration. Therefore, we used E-4031 to separate the two components; E-4031-resistant and -sensitive components were measured as indexes of I_Ks and I_Kr, respectively. Validity of this method depends on the stability of I_Ks and I_Kr during the 10 min exposure to E-4031. If I_Ks is running down substantially over the time, the E-4031-sensitive component would include not only real I_Kr but also the rundown of I_Ks. We cannot completely rule out this possibility, but it is unlikely for the following reason: I_Ks (E-4031-resistant component) was measured twice, 10 and 15 min after application of E-4031, and the data from the myocytes showing rundown >10% were not included in the results. It was observed in our different series of experiments in rabbit ventricular myocytes that I_Ks and I_Kr estimated by E-4031-resistant and -sensitive components of I_K using the same protocol as in the present study were virtually identical to I_Ks and I_Kr estimated by chromanol 293B (30 μM, 10 min bath application) -sensitive and -resistant components of I_K (unpublished data).

Although one simulation study [24] emphasized the significance of I_Kr and I_Ks for final repolarization, another simulation study shows that in heart failure a mere reduction in membrane currents will not mimic action potential prolongation unless deviations for calcium handling proteins such as the SR Ca²⁺-ATPase and the sarcolemmal Na⁺/Ca²⁺ exchanger are included in the program [50]. This may underscore the significance of altered calcium handling for action potential changes [10,51,52].

5 Conclusion

The downregulation of delayed rectifier potassium currents (I_Kr and I_Ks) in heart failure, may render these currents important for action potential prolongation at physiological cycle length and for arrhythmogenesis in heart failure.

Acknowledgements

We thank Medtronic Japan, Inc. for graciously supplying the pulse generators and pacing leads for this study and Dr. Marieke W.Veldkamp for critical reading of our manuscript.