Ionic Remodeling of Cardiac Purkinje Cells by Congestive Heart Failure
Background Cardiac Purkinje cells (PCs) are important for the generation of triggered arrhythmias, particularly in association with abnormal repolarization. The effects of congestive heart failure (CHF) on the ionic properties of PCs are unknown.
Methods and Results PCs were isolated from false tendons of control dogs and dogs with ventricular tachypacing-induced CHF. CHF PCs were hypertrophied (capacitance, mean±SEM, 149±4 pF, n=130; versus 128±3 pF, n=150, control; P<0.001). Transient outward current density was reduced in CHF PCs without change in voltage dependence or kinetics. CHF also reduced inward-rectifier current density, with no change in form of the current-voltage relationship. Densities of L- and T-type calcium, rapid and slow delayed rectifier, and Na+-Ca2+ exchange currents were unaltered by CHF, but L-type calcium current inactivation was slowed at positive potentials. Purkinje fiber action potentials from CHF dogs showed decreased phase 1 amplitudes and elevated plateau voltages and demonstrated twice as much prolongation on exposure to the rapid delayed rectifier blocker E-4031 as control Purkinje fibers.
Conclusions CHF causes remodeling of important K+ and Ca2+ currents in cardiac PCs, decreasing repolarization reserve and causing an exaggerated repolarization delay in response to a class III drug. These results have important potential implications regarding ventricular arrhythmogenesis, particularly related to triggered activity in PCs, in patients with CHF.
Received May 17, 2001; revision received July 20, 2001; accepted July 25, 2001.
Congestive heart failure (CHF) predisposes to the generation of ventricular tachyarrhythmias1 and the occurrence of sudden death.2 In addition, CHF promotes drug-induced Torsades de Pointes arrhythmias.3 Abnormal repolarization, related to ion channel remodeling, is important in the arrhythmogenic potential of CHF.1,4,5
CHF-induced remodeling of ionic currents in ventricular5 and atrial6 myocytes has been studied in detail. Cardiac Purkinje cells (PCs) are believed to play important roles in the generation of ventricular arrhythmias, particularly those related to triggered activity.7–9 Ionic currents are altered in subendocardial PCs in regions of myocardial infarction10–12; however, almost nothing is known about PC ionic remodeling in CHF. The present study was designed to evaluate CHF-induced changes in ionic currents and action potentials (APs) in canine cardiac PCs.
CHF was produced as previously described6 by pacing the right ventricle at 240 bpm for 3 weeks followed by 2 weeks at 220 bpm. Dogs were then anesthetized with morphine (2 mg/kg SC) and α-chloralose (120 mg/kg IV load, 29.25 mg/kg per hour infusion), and a median sternotomy was performed. All animal care and handling procedures followed the guidelines of the Canadian Council on Animal Care. Excised hearts were immersed in Tyrode solution at room temperature. Free-running false tendons were excised into modified Eagle’s MEM (Gibco-BRL; pH 6.8, HEPES-NaOH) containing collagenase (800 to 900 U/mL, Worthington Type-II) and 1% BSA (Sigma), and single PCs were isolated as previously described.13,14
Standard Tyrode solution contained (in mmol/L) NaCl 136, KCl 5.4, MgCl2 1, CaCl2 1, NaH2PO4 0.33, HEPES 5, and dextrose 10; pH 7.4 (NaOH). High-K+ storage solution contained (in mmol/L) KCl 20, KH2PO4 10, dextrose 10, glutamic acid 70, β-hydroxybutyric acid 10, taurine 10, and EGTA 10; 0.1% BSA; pH 7.4 (KOH). Standard pipette solution contained (in mmol/L) K+ aspartate 110, KCl 20, MgCl2 1, Mg2ATP 5, HEPES 10, phosphocreatine 5, GTP 0.1, and EGTA 5; pH 7.2 (KOH). Solutions were equilibrated with 100% O2.
For K+ current measurement, the extracellular solution included 1 μmol/L atropine to eliminate muscarinic K+ currents and CdCl2 (200 μmol/L) or nimodipine (1 μmol/L, for IK studies) to block Ca2+ currents. Na+ current contamination was prevented by equimolar substitution of choline for extracellular Na+. For currents other than transient outward current (Ito), 1 mmol/L 4-AP was used to block Ito. Rapid delayed rectifier current (IKr) was studied as 5 μmol/L E4031-sensitive current and inward rectifier current (IK1) as 1 mmol/L Ba2+-sensitive current. Slow delayed rectifier current (IKs) was studied in the presence of 1 μmol/L dofetilide to eliminate IKr. For ICa recording, the bath solution contained tetraethylammonium chloride, CsCl, and CsOH in place of NaCl, KCl, and NaOH, respectively, and [CaCl2] was 2 mmol/L. The pipette for ICa recording contained (in mmol/L) CsCl 20, Cs-aspartate 110, MgCl2 1, EGTA 10, GTP 0.1, ATP-Mg 5, HEPES 10, and Na2 phosphocreatine 5; pH 7.2 (CsOH). Na+-Ca2+ exchange (NCX) current (INCX) was recorded with ramp pulses and extracellular (in mmol/L, NaCl 140, CaCl2 0 or 5, MgCl2 1, CsCl 5, HEPES 5, nimodipine 0.001, ouabain 0.01, and ryanodine 0.005; pH 7.2 CsOH) and pipette (in mmol/L, CsCl 90, NaCl 50, MgATP 5, MgCl2 3, EGTA 20, CaCl2 13, and HEPES 20; pH 7.2, CsOH) solutions designed to suppress K+ current, Na+-K+ ATPase, and sarcoplasmic reticulum Ca2+ release.15
Data Acquisition and Analysis
Whole-cell patch clamp was performed as previously described13,14 at 36.5°C. Compensated series resistance and capacitive time constants (τs) averaged 2.5±0.1 MΩ and 290±10 μs. Leakage compensation was not used. The capacitance of CHF cells was increased (149±4 pF, n=130, versus 128±3 pF in control, n=150; P<0.001), so currents are expressed in terms of density.
Standard microelectrode techniques were used to record action potentials (APs). The Tyrode solution contained (in mmol/L) NaCl 120, KCl 1.5, KH2PO4 1.2, MgSO4 0.1, NaHCO3 25, CaCl2 1.25, and dextrose 5; pH 7.4. Purkinje fiber false tendons were superfused with oxygenated (95% O2 and 5% CO2) Tyrode solution at 36°C and impaled with 3 mol/L KCl-filled glass microelectrodes (8 to 20 MΩ) connected to a high input-impedance amplifier.
Nonlinear least-square curve-fitting algorithms were used for curve fitting. Nonpaired t tests were used to compare CHF with control cells. P<0.05 was considered to indicate statistical significance. Group data are expressed as mean±SEM.
Changes in Ionic Currents
Figure 1 shows representative IK1 recordings (panels A and B). CHF significantly reduced IK1 (C), including the outward component (D), without altering the form of the IK1 current-voltage relation (Figure 1C, inset).
Figures 2A and 2B illustrate Ito recordings from control and CHF PCs. Ito density was significantly smaller in CHF PCs (Figure 2C); however, there were no differences in the form of the Ito-V relation (Figure 2D). The voltage dependence of Ito inactivation was tested with a 2-pulse protocol as described in Figure 2E. Activation voltage dependence was evaluated from the relation ITP=aTPGmax(VTP−VR), where ITP and aTP are current and activation variable at test potential VTP, VR is reversal potential, and Gmax is maximal conductance. Half-maximal voltage (V) and activation slope factor (Boltzmann fits) were 8.9±0.7 and 10.3±0.4 mV (control) and 8.9±1.0 and 10.9±1.0 mV (CHF, P=NS). VR obtained from the reversal of Ito tail currents after 2-ms activating pulses averaged −75.3±2.2 mV. Inactivation V and slope factor were −30±2 and 11±1 mV (control) and −31±2 and 11±1 mV (CHF, P=NS). Ito inactivation kinetics were biexponential, with τs unaltered by CHF (Figure 2F). Ito recovery (Figure 2G) was biexponential, with τs averaging 42±8 and 1391±67 ms (control) and 52±11 and 1486±108 ms (CHF, n=10 for each, P=NS). Ito frequency dependence was similarly unaffected by CHF (Figure 2H).
Figure 3 shows results for IKr (left) and IKs (right). Representative control recordings are shown at the top, with mean step I-V relations, which were unchanged by CHF, in the middle. Activation voltage dependence based on normalized tail currents is shown in bottom panels and was not altered by CHF for either IKr or IKs.
Representative L-type calcium current (ICa.L) recordings are shown in Figures 4A and 4B and point to slowed ICa.L decay in CHF. ICa.L density was not significantly different between control and CHF cells (Figure 4C). Inactivation τs were slowed significantly by CHF at voltages positive to +10 mV (Figure 4D). For example, at +20 mV, τfast increased from 6.0±0.3 to 8.2±0.9 ms (P<0.05) and τslow from 38±2 to 66±8 ms (P<0.01) in CHF PCs. In addition to a slowing of inactivation τs, CHF significantly increased the proportion of slow-phase inactivation at positive voltages (Figure 4E). For example, at +30 mV, τ2 accounted for 36±4% of inactivation in control PCs compared with 51±5% in CHF (P<0.05). ICa.L inactivation voltage dependence was assessed with a 2-pulse protocol (Figure 4F). Inactivation V and slope factor were −25±2 and −6±1 mV (control) and −26±1 and −7±1 mV (CHF, n=10 cells/group, P=NS). Activation voltage dependence was assessed according to the relation ITP=aTPGmax (VTP−VR), with VR obtained from a linear fit to the ascending portion of the I-V relation. Resulting mean V and slope factors were 2.0±1.4 and 6.6±0.2 mV (control) and 0.4±1.4 and 6.5±0.1 mV (CHF, n=10/group, P=NS). Reactivation kinetics at holding potential (HP) (−80 mV) close to the PC resting potential are shown in Figure 4G. Reactivation was monoexponential, with similar τs in control (44±7 ms) and CHF PCs (53±6 ms, n=10/group, P=NS). ICa.L showed little frequency dependence between 1 and 5 Hz at a HP of −80 mV, and there were no significant CHF-related differences in ICa.L frequency dependence (Figure 4H).
All PCs possessed a relatively large T-type calcium current (ICa.T) (Figures 5A and B). ICa.T was obtained as previously described16,17 by digital subtraction of ICa elicited at HP −50 mV from ICa at a HP of −90 mV. The mean ICa.T density voltage relation was not altered by CHF (Figure 5C). ICa.T inactivation V and slope factors averaged −61±2 and 4.1±0.4 mV (control) and −63±2 and 4.9±0.5 mV (CHF, n=10/group, P=NS) (Figure 5D). Activation V and slope factors were −32±1 and 10±1 mV (control) and −30±2 and 12±1 mV (CHF), respectively (n=10/group, P=NS). The time course of ICa.T inactivation was monoexponential, and there were no significant differences between groups; for example, at −20 mV (voltage of maximum ICa.T), τs were 4.5±0.2 ms (control, n=10) versus 4.6±0.3 ms (CHF, n=10, P=NS). ICa.T reactivation was assessed (Figure 5E) at a test potential (−30 mV) at which no ICa.L is activated (Figure 4C). ICa.T reactivation τs averaged 51±3 ms (control) versus 49±4 ms (CHF, n=10 cells/group, P=NS). The frequency dependence of ICa.T (Figure 4F) was greater than that of ICa.L but did not differ between groups.
Figure 6A shows INCX as determined from current during ramp depolarizations from −60 to +50 mV in the presence of 5 and 0 mmol/L [Ca2+]o. Reverse-mode INCX is substantial in the presence of 5 mmol/L Ca2+ and is absent in the presence of 0 mmol/L Ca2+, allowing INCX to be calculated from the difference between the two current recordings.15 The results in Figure 6B show that INCX amplitude was larger in CHF cells but that after correction for cell size (capacitance normalization), there were no significant differences.
Changes in APs
PC AP characteristics were first recorded at a total of 42 sites from 11 control dogs and 37 sites in 8 CHF dogs at 1 Hz. Phase 1 repolarization was less marked in CHF, and the plateau voltage was higher (Figure 7, top). There were no significant differences in resting potential, AP amplitude, or AP duration (APD) between control and CHF cells, but plateau-voltage was significantly more positive and phase 1 amplitude significantly smaller in CHF PCs (Table).
To evaluate the possibility that the response to an IKr-blocking class III drug may be altered in CHF PCs, APs were recorded from control and CHF PCs in free-running false tendons with standard microelectrodes before and after exposure to E-4031 (1 μmol/L). The results are illustrated in Figures 7C and 7D, and mean data are provided in Figures 7E and 7F. As in the first series of experiments, predrug APDs were not significantly different between control and CHF PCs. However, E-4031 had substantially larger effects on APD in CHF PCs, so that APD95 after the drug was significantly greater in CHF cells at all frequencies (P<0.001 for each). For example, E-4031-induced APD95 increases averaged 103±45 ms (25±10%) at 1 Hz in control preparations compared with 231±40 ms (54±10%) in CHF preparations.
We have evaluated the effects of CHF on K+ currents, Ca2+ currents, NCX, and AP properties of PCs from free-running ventricular false tendons. CHF decreased PC Ito and IK1 density and slowed the inactivation of ICa.L without altering its density. IKs, IKr, INCX, and ICa.T were unaffected. CHF reduced the amplitude of phase 1 repolarization, increased plateau voltage, and enhanced the APD-prolonging effect of E-4031.
Comparison with Previous Studies of CHF-Induced Ionic Remodeling
There have been extensive studies of the ionic remodeling of K+ and Ca2+ channels in ventricular myocytes of patients and experimental animals with CHF.5,18,19 Ito is quite consistently reduced5,20–22 by an average of ≈35% in ventricular myocytes of patients with CHF.18 IK1 also tends to be decreased by ≈25%.18 Although there is some variability in the results for ICa.L,5,23 overall there seems to be no change in ICa.L density or kinetics. Recent studies point to downregulation of IKs and possibly IKr in ventricular myocytes of rabbits with tachypacing-induced CHF.24 Atrial ionic remodeling in CHF has been studied to a lesser extent, but recent studies suggest that Ito, IKs, and ICa are all decreased and INCX is increased, with no change in kinetics or voltage dependence and no change in IK1 or IKr.6
Several aspects of PC ionic remodeling resemble previously reported findings in ventricular myocytes: Ito and IK1 were reduced and ICa.L density was unchanged. These observations suggest that the mechanisms leading to CHF-induced Ito and IK1 downregulation at the ventricular level also likely operate on free-running Purkinje fiber false tendons. The CHF-induced slowing of ICa inactivation that we observed has not, to our knowledge, been reported in ventricular myocytes. We did not observe changes in IKr or IKs, suggesting that PCs may be spared from the IK downregulation occurring with CHF in ventricular myocytes24 and possibly explaining why overall APD was not prolonged in CHF PCs. Unlike typical findings in ventricular5 and atrial6 myocytes, INCX density was not increased by CHF in PCs.
Relationship to Other Studies in PCs
Ito, ICa, IK1, and E-4031-sensitive current have been studied in PCs from the subendocardial Purkinje fiber network overlying 24- to 48-hour-old myocardial infarctions.10–12 Ito density was reduced by >50% in the infarct zone, with no change in voltage dependence but a slowing in reactivation.10 PCs from free-running false tendons had normal Ito properties. Both ICa.L and ICa.T were reduced in subendocardial PCs from the infarct zone, with no changes in voltage dependence or inactivation kinetics.11 IK1 was reduced in PCs from the infarct zone, and a very rapidly activating E-4031-sensitive current was increased, with no classical IK noted in either normal or infarct zone PCs.12 In general, the abnormalities noted in infarct zone PCs were more severe than those we observed. There are, however, some qualitative similarities in terms of decreases in IK1 and Ito. In contrast to our findings, myocardial infarction did not affect the electrophysiology in PCs from free-running false tendons, suggesting that myocardial infarction produces severe but localized ionic remodeling in infarct zone PCs whereas CHF produces more generalized but less severe remodeling.
The present study constitutes the first detailed analysis of CHF-induced ionic remodeling in PCs. PCs are believed to play an important role in ventricular arrhythmogenesis, particularly in the generation of early afterdepolarization-induced arrhythmias.7,8 The CHF-induced ionic abnormalities we noted in PCs may contribute to promoting the occurrence of arrhythmogenic afterdepolarizations in patients with CHF, particularly in response to interventions such as IKr-blocking antiarrhythmic drugs and hypokalemia that prolong APD. Downregulation of Ito, along with slowed inactivation of ICa, is likely responsible for the positive shift in the plateau voltage of PCs. The positive shift in plateau voltage and slowed ICa.L inactivation at positive voltages would act to promote the occurrence of ICa.L-dependent early afterdepolarizations under conditions that delay repolarization.25
The CHF-induced decreases in Ito and IK1 are likely to reduce the repolarization reserve, the ability of myocardial cells to repolarize when normal repolarizing currents are reduced by drugs, metabolic or electrolyte imbalances, or intercurrent diseases. Consistent with this notion, PCs from dogs with CHF showed twice as great APD prolongation in response to IKr blockade with E-4031 compared with control PCs. CHF significantly increases the risk of drug-induced Torsades de Pointes arrhythmias.3 The CHF-induced ionic remodeling that we observed in PCs could clearly play an important role in this clinically important phenomenon.
The isolation of PCs from free-running Purkinje fibers is technically challenging, requiring prolonged periods of bath exposure to cell-dissociating enzymes (chunk method). This likely explains the absence of studies of PC remodeling in CHF, despite the large number of studies that have evaluated ventricular myocyte remodeling. IK is particularly sensitive to isolation technique26 and has been difficult to record in isolated PCs.12,27 We were able to record robust IKs in isolated PCs, but IKr was smaller and more difficult to record; therefore, our results regarding IKr should be interpreted with caution. The properties of the currents we recorded from normal PCs were generally similar to those reported for studies with other preparations; however, we cannot exclude the possibility that cell isolation might have affected the currents we evaluated. Isolation of PCs uncouples them from other PCs and from ventricular myocytes, removing an important modulator of electrophysiologic function present in vivo. This consideration does not apply to the multicellular preparations we studied, which maintain attachments to both false tendon PCs and underlying muscle intact. We measured currents in the presence of heavily buffered [Ca2+]i. [Ca2+]i buffering was necessary to preserve cell viability under our recording conditions; however, CHF-induced changes in [Ca2+]i could importantly affect currents like the NCX. Such changes would not have been detected in the present study.
This work was supported by operating grants from the Canadian Institutes for Health Research and the Quebec Heart and Stroke Foundation. Dr Li was a Heart and Stroke Corporation of Canada/AstraZeneca Fellow. The authors thank Essai and Pfizer Pharmaceuticals for providing E-4031 and dofetilide, Chantal St-Cyr and Chantal Maltais for technical assistance, and Annie Laprade for secretarial help.
Nuss HB, Kaab S, Kass DA, et al. Cellular basis of ventricular arrhythmias and abnormal automaticity in heart failure. Am J Physiol. 1999; 277: H80–H91.
Lehmann MH, Hardy S, Archibald D, et al. Sex difference in risk of torsade de pointe with d,l-sotalol. Circulation. 1996; 94: 2535–2541.
Tomaselli GF, Marbán E. Electrophysiological remodeling in hypertrophy and heart failure. Cardiovasc Res. 1999; 42: 270–283.
Li D, Melnyk P, Feng J, et al. Effects of experimental heart failure on atrial cellular and ionic electrophysiology. Circulation. 2000; 101: 2631–2638.
El Sherif N, Caref EB, Yin H, et al. The electrophysiological mechanism of ventricular arrhythmias in the long QT syndrome: tridimensional mapping of activation and recovery patterns. Circ Res. 1996; 79: 474–492.
Berenfeld O, Jalife J. Purkinje-muscle reentry as a mechanism of polymorphic ventricular arrhythmias in a 3-dimensional model of the ventricles. Circ Res. 1998; 82: 1063–1077.
Jeck C, Pinto JM, Boyden PA. Transient outward currents in subendocardial Purkinje myocytes surviving in the 24 and 48 hr infarcted heart. Circulation. 1995; 92: 465–473.
Boyden PA, Pinto JM. Reduced calcium currents in subendocardial Purkinje myocytes that survive in the 24- and 48-hour infarcted heart. Circulation. 1994; 89: 2747–2759.
Han W, Wang Z, Nattel S. A comparison of transient outward currents in canine cardiac Purkinje cells and ventricular myocytes. Am J Physiol. 2000; 279: H466–H474.
Han W, Wang Z, Nattel S. Slow delayed rectifier current and repolarization in canine cardiac Purkinje cells. Am J Physiol. 2001; 280: H1075–H1080.
Hirano Y, Fozzard HA, January CT. Characteristics of L- and T-type Ca2+ currents in canine cardiac Purkinje cells. Am J Physiol. 1989; 256: H1478–H1492.
Tseng GN, Boyden PA. Multiple types of Ca2+ currents in single canine Purkinje cells. Circ Res. 1989; 65: 1735–1750.
Priebe L, Beuckelmann DJ. Simulation study of cellular electric properties in heart failure. Circ Res. 1998; 82: 1206–1223.
Kaab S, Nuss HB, Chiamvimonvat N, et al. Ionic mechanism of action potential prolongation in ventricular myocytes from dogs with pacing-induced heart failure. Circ Res. 1996; 78: 262–273.
Beuckelmann DJ, Nabauer M, Erdmann E. Alterations of K+ currents in isolated human ventricular myocytes from patients with terminal heart failure. Circ Res. 1993; 73: 379–385.
Tsuji Y, Opthof T, Kamiya K, et al. Pacing-induced heart failure causes a reduction of delayed rectifier potassium currents along with decreases in calcium and transient outward currents in rabbit ventricle. Cardiovasc Res. 2000; 48: 300–309.
Viswanathan PC, Rudy Y. Pause induced early afterdepolarizations in the long QT syndrome: a simulation study. Cardiovasc Res. 1999; 42: 530–542.
Yue L, Feng J, Li GR, et al. A characterization of transient outward and delayed rectifier currents in canine atrial myocytes: properties of currents and role of cell isolation methods. Am J Physiol. 1996; 270: H2157–H21683.