Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Cardiac hypertrophy induced by pressure or volume overload is associated with alterations in the electrophysiologic properties of the heart (Hart, 1994; Lynch et al., 1992; Tomaselli et al., 1994). These alterations include prolongation of the APD (Aronson, 1980; Bril et al., 1991; Hart, 1994; Nordin et al., 1989), decrease in the resting membrane potential (Cameron et al., 1983), after depolarizations and triggered spontaneous activity (Aronson, 1991) and a decrease in the ventricular fibrillation threshold (Kohya et al., 1988; Kohya et al., 1995). Prolongation of the APD is the abnormality that has been observed most consistently in hypertrophied myocardium (Aronson, 1991; Hart, 1994; Tomaselli et al., 1994). This abnormality has been ascribed to altered characteristics of the L-type Ca⁺⁺ current (Keung, 1989; Kleiman and Houser, 1988; Nordin et al., 1989), of I_K (Brooksby et al., 1993; Kleiman and Houser, 1989) and of I_to (Beuklemann et al., 1993; Tomita et al., 1994; Wettwer et al., 1994; Xu and Best, 1991). Abnormalities of I_K1 have also been described in hypertrophied cardiac myocytes (Brooksby et al., 1993; Kleiman and Houser, 1989). I_K1 and I_to are present in human cardiac tissue (Beukelmann et al., 1993; Nabauer et al., 1993; Shibata et al., 1989; Wettwer et al., 1994) and are altered in humans with left ventricular failure (Beukelmann et al., 1993; Nabauer et al., 1993; Wettwer et al., 1994). I_K is believed to make an important contribution to repolarization in humans, although this current in human ventricular cells has been reported to be small or absent (Beukelmann et al., 1993; Konarzewska et al., 1995). I_K1, I_K and I_to also contribute to ventricular repolarization in rabbit ventricle (Carmeliet, 1993; Fedida and Giles, 1991; Giles and Imaizumi, 1988). Atrial and ventricular arrhythmias occur frequently in the setting of cardiac hypertrophy (Lynch et al., 1992; McLenachan and Dargie, 1990; Myerberg et al., 1992; Tomaselli et al., 1994), and antiarrhythmic drug therapy may frequently be prescribed. It is possible that the abnormalities of ion channel function in cardiac hypertrophy are associated with changes in the responsiveness to specific ion channel blockers (Hart, 1994; Kowey et al., 1991; Myerberg et al., 1992). Although the densities of I_K1, I_K and I_to are altered in ventricular hypertrophy, the effects of specific blockers of I_K1, I_K and I_to on ventricular repolarization in the setting of ventricular hypertrophy are unknown.

Increases in preload and afterload may alter ventricular repolarization (Dean and Lab, 1990; Lerman et al., 1985). Whether the magnitude of changes in repolarization under varying conditions of preload or afterload differs between normal and hypertrophied myocardium has not been extensively studied (Gillis et al., 1996b). Furthermore, it is not known whether specific K⁺ ion channel blockers alter the effects of afterload on repolarizations.

We hypothesized that the effects of some selective potassium channel blockers might be altered in cardiac hypertrophy and that these effects might be modulated in part by afterload. Accordingly, the purposes of the present study were 1) to characterize the electrophysiologic abnormalities associated with left ventricular pressure overload in the working rabbit heart, 2) to compare the effects of the I_K1 blocker barium, the I_Kr blocker dofetilide and the I_to blocker 4-AP on ventricular electrophysiologic properties in hypertrophied and normal rabbit hearts and 3) to compare the effects of increased afterload on cardiac electrophysiologic parameters in control and hypertrophied hearts at base line and during perfusion of barium, of dofetilide and of 4-AP.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Animals and surgical procedure. Male New Zealand White rabbits weighing 2.5 to 3.0 kg were used in this study. Left ventricular pressure overload was induced by partial ligation of the abdominal aorta. On the day of surgery, the rabbits were sedated with acepromazine (1.0-1.5 mg/kg) administered i.v. General anesthesia was then induced with an oxygen/halothane (2.5%) mixture administered via an anesthesia mask at a rate of 1 l/min. The abdominal aorta was exposed just above the renal arteries and looped with a 3-0 silk suture. The suture was tied against a 2.1-mm probe, which was then withdrawn. The incision was closed, and the animals were kept in the animal care center until the day of the study. Sham-operated rabbits served as controls. These animals underwent the abdominal laparotomy, but coarctation of the abdominal aorta was not induced. The day before the study, the animals were sedated with diazepam. A 22-gauge catheter was inserted percutaneously into an ear artery for determination of the arterial blood pressure. The catheter was then removed. All procedures conformed to the guiding principle of the Canadian Council on Animal Care.

Working heart preparation. Because we hypothesized that differences in the electrophysiologic effects of the selective ion channel blockers between hypertrophied hearts and controls would be most likely to be observed if electrophysiologic changes had already developed, the present studies were conducted in animals after the development of hypertrophy associated with prolongation of the APD. Animals were studied 43 ± 8 days after abdominal surgery. In preliminary studies, significant prolongation of APD was not observed if animals were studied less than 30 days after abdominal surgery. Prolongation of the APD and significant left ventricular hypertrophy had developed in animals undergoing abdominal aortic constriction compared with sham controls (P < .05, table 1). On the day of the study, the rabbits were pretreated with heparin sulfate (75 U/kg; Wyeth Ayerst, Montreal, Que.) and then anesthetized with sodium pentobarbital (35 mg/kg; MTC Pharmaceuticals, Cambridge, Ont.). Hearts were rapidly removed through a median sternotomy incision and perfused retrogradely via the aorta. Hearts were initially perfused with a modified Krebs-Henseleit buffer consisting of (mM): NaCl 118, KCl 3.3, CaCl₂ 2.0, MgSO₄ 1.2, KH₂PO₄ 1.2, NaHCO₃ 24, glucose 10, Na pyruvate 2.0 and albumin 20 mg/l (BDH Inc., Toronto, Ont.). The buffer was equilibrated with 95% O₂-5% CO₂, pH 7.4, temperature 37°C. The left atrium was cannulated for antegrade perfusion via a pulmonary vein. The venae cavae were ligated. A 4F Millar catheter was inserted in the left ventricle via a left atriotomy incision for monitoring of left ventricular pressure and the derivative of left ventricular pressure (dP/dt). Hearts were atrially paced at a cycle length of 400 msec (pulse width 2.0 msec; twice diastolic threshold intensity) via a bipolar platinum electrode positioned in the region of the sinus node. A bipolar platinum electrode was also positioned on the lateral wall of the left ventricle for ventricular stimulation protocols. Platinum electrodes were positioned on the epicardial surface of the heart for monitoring of the ECG. Monophasic action potentials were recorded from the epicardial surface of the posterior left ventricle wall via contact electrode catheters.

Control experiments. To determine the electrophysiologic and hemodynamic stability of the working heart preparation, as well as to determine the optimum cardiac output for these studies, we conducted time-dependent control experiments in six sham-operated animals. Hearts were perfused at an afterload of 30 cm H₂O and a cardiac output of 30 ml/min for 10 min. The cardiac output was increased to 60 ml/min for 10 min. The hearts were then perfused at an afterload of 100 cm H₂O and a cardiac output of 30 ml/min for 10 min. The cardiac output was then increased to 60 ml/min for 10 min. Afterload and cardiac output were then decreased to 30 cm H₂O and 30 ml/min, and perfusion was continued for 30 min. The perfusion sequence was then repeated at each afterload and cardiac output.

Barium experiments. The perfusion sequences for the experiments are shown in figure 1. The concentrations of BaCl₂ evaluated in the present study have been shown to block I_K1 selectively in rabbit ventricle and to prolong the APD without affecting other repolarizing and depolarizing currents (Giles and Imaizumi, 1988; Wang et al., 1993). The perfusion protocol was carried out in eight sham-operated hearts and eight hypertrophied hearts. The cardiac output was 60 ml/min throughout the study. Hearts were perfused at an afterload of 30 cm H₂O for 10 min. Afterload was then increased to 100 cm H₂O by increasing the height of the aortic ejection column, and perfusion with the blood-buffer mixture was continued for 10 min. Perfusion with 0.025 mM BaCl₂ in the blood-buffer perfusate was then initiated at an afterload of 30 cm H₂O and continued for 25 min, followed by a 15-minute equilibration period and a 10-min period of recordings at steady-state barium concentration. The afterload was increased to 100 cm H₂O, and 0.025 mM BaCl₂ perfusion was maintained for an additional 10 min. The concentration of BaCl₂ was increased to 0.050 mM, and this protocol was repeated. The blood-buffer mixture containing BaCl₂ was recirculated in a second oxygenated reservoir throughout all experiments.

View larger version (16K): [in this window] [in a new window]   Fig. 1.   Schema of sequence of experimental interventions for the barium chloride, dofetilide or 4-AP experiments. The ordinate represents time, and differences in ventricular afterload are represented on the abscissa. The 4-AP perfusion sequences were 5 min longer than the barium perfusion sequences.

Dofetilide experiments. The concentration of dofetilide used in the present study has been shown to block I_Kr selectively (Jurkiewicz and Sanguinetti, 1993). The perfusion protocol was similar to the BaCl₂ experiments except that only one drug concentration was evaluated (fig. 1). Electrophysiologic data were collected at base line and during perfusion with dofetilide (15 nM) at low and high ventricular afterloads in eight sham and eight hypertrophied hearts.

4-AP experiments. The concentrations of 4-AP evaluated in the present study have been shown to block selectively one component of I_to in rabbit ventricle (Campbell et al., 1993; Fedida and Giles, 1991). A perfusion protocol similar to that of the BaCl₂ experiments was followed for the 4-AP experiments (fig. 1). Electrophysiologic data were collected at base line and during perfusion with 4-AP (concentrations of 0.2 mM followed by 0.5 mM) at low and high ventricular afterloads in eight sham and eight hypertrophied hearts.

Electrophysiologic measurements. The ECG and monophasic APD recordings, the left ventricular pressure and the derivative of left ventricular pressure (dP/dt) were digitally acquired. Signals taken from Gould amplifiers (models 13-G 4615-58, 13-4615-50 and 13-4615-71) were acquired at 1000 Hz/channel on a Compaq 386 system using a Data Translation (DT 2821) analog and digital input-output board. This board provides a 12-bit resolution, 50-Hz throughput and software-programmable gains. Monophasic APD was measured at base line (t = 0) and at 5 and 10 min during each hemodynamic intervention. The APD recorded by the monophasic action potential catheters was measured from the steepest part of the action potential upstroke to the level of 25%, 50% and 90% repolarization. We defined the distance from the diastolic base line to the crest of the plateau as the total action potential amplitude (Gillis et al., 1996a).

Hemodynamic and morphologic measurements. Left ventricular end diastolic pressure, peak left ventricular systolic pressure and the peak positive first derivative of left ventricular pressure (dP/dt) were continuously monitored. Coronary flow was measured as the coronary sinus effluent ejected via the pulmonary artery. Hearts were blotted and weighed at the end of each experiment. The ventricles were isolated and weighed. Transverse sections of the heart were cut at the midventricular lead at the end of each study and fixed in 10% buffered formalin for light microscopic study. Sections of a paraffin block were stained either with hematoxylin and eosin for evaluating histology or with Gomori trichrome for distinguishing muscle and interstitial connective tissue (Gillis et al., 1991). Other sections were stained by PAS with or without diastase to outline the cell membranes. Histologic sections were imaged on a television monitor ×1000 magnification. The diameter across the nucleus of longitudinally oriented myocytes located in the intraventricular septum was measured. The diameter of the myocyte was measured semiautomatically by marking two points with a lightpen using a Bioquant 4 Image Analyses System (Rand M Biometrics, Inc., Nashville, TN) (Hoshino et al., 1983; Schoen et al., 1984). The diameters of 80 myocytes were averaged for each experiment.

Whole-cell voltage-clamp experiments. Solutions for patch-clamp studies were made using millipore-filtered water. The modified extracellular Tyrode's solution consisted of (mM): NaCl 130, KCl 5.5, MgCl₂ 0.8, NaH₂PO₄ 1.0, glucose 10, HEPES 10, pH adjusted to 7.4 with NaOH. The KB solution consisted of (mM): KCl 40, KH₂PO₄ 20, MgCl₂ 5.0, KHCO₃ 0.5, potassium glutamate 50, potassium aspartate 20, ethyleneglycol-bis-(-aminoethylether) N,N,N',N'-tetraacetic acid (EGTA) 0.1, glucose 10, HEPES 10, taurine 20, bovine serum albumin 0.02%, pH adjusted to 7.4 with KOH. HEPES-buffered solution consisted of (mM): NaCl 130, KCl 5.5, MgCl₂ 0.8, CaCl₂ 2, HEPES 10, glucose 10, pH adjusted to 7.4 with NaOH. The internal pipette solution consisted of (mM): potassium aspartate 110, Na₂ATP 4, MgCl₂ 4, CaCl₂ 1, KCl 10, EGTA 10, HEPES 5, pH adjusted to 7.2 with KOH.

Data analysis. APD₉₀, left ventricular pressures and dP/dt were measured semi-automatically using a custom-designed software program. An average of five complexes were measured at each sampling time. Electrophysiologic and hemodynamic parameters were compared between the sham and hypertrophied hearts. The effects of increasing afterload and the effects of barium and 4-AP were compared within groups as well as between groups. Whole-cell patch-clamp data were analyzed using the CLAMPFIT software (Axon Instruments).

Statistical analysis. Data are presented as mean ± S.D. Differences between groups were compared using an unpaired t test or a factorial analysis of variance for repeated measures with a split-plot design where appropriate (Montgomery, 1991). Multiple linear regression analysis was applied to identify predictors of APD₉₀ in association with hypertrophy. Differences were considered statistically significant at P < .05.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Time-dependent control experiments. Electrophysiologic and hemodynamic data measured 5 min after the initiation of a change in afterload or cardiac output are shown in table 2. Increasing cardiac output from 30 to 60 ml/min resulted in significant increases in peak systolic left ventricular pressure and dP/dt at both low and high ventricular afterload (P < .01). Increasing the cardiac output did not result in any significant changes in electrophysiologic parameters. However, when afterload was increased, APD₉₀ decreased significantly (P < .05). The electrophysiologic and hemodynamic parameters were found not to change significantly over time when the same hemodynamic states were compared (P = N.S.).

Base-line electrophysiologic measures in sham and hypertrophied hearts. APD was significantly longer in the hypertrophied hearts compared with the sham-operated hearts at both low and high afterload (fig. 2, P < .05). APD decreased significantly in both groups when afterload was increased (fig. 2, P < .01), but the magnitude of APD shortening was similar in both groups. The relationship between APD₉₀ and mean arterial pressure was linear (R = 0.37, P = .015, fig. 2). The number of days after aortic constriction (P = .07) and the mean arterial pressure (P = .08) tended to be independent predictors of APD.

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Upper panel: Monophasic APD90 recorded from the posterior wall of the left ventricle during the base-line perfusion in 24 sham hearts () and 24 hypertrophied hearts () at low (30 cm H2O) and high (100 cm H2O) afterload. Data are mean ± S.D. Lower panel: Relationship between APD90 and mean arterial pressure was linear.

Sham-operated and hypertrophied hearts. The characteristics of the sham-operated and hypertrophied groups are shown in table 1. The animals were similar in weight. The systolic, diastolic and mean arterial pressures measured in vivo the day before the study were significantly higher in the banded rabbits than in the sham-operated animals (P < .001). The mean heart weights and weights of the ventricles measured on completion of the experiments were significantly greater in the hypertrophied hearts than in the sham hearts (P < .01). Although heart mass tended to be greater in the barium subgroup than in the dofetilide or 4-AP subgroup, these differences were not significant (P = N.S.). Significant fibrosis was not observed by histologic assessment of the myocardium in either the sham or the hypertrophied hearts. The mean cell diameters tended to be greater in the hypertrophied cells than in the sham controls, but these differences were statistically significant only in the barium subgroup (table 1, P < .01). Significant differences in hemodynamic measurements were not observed between the sham and the hypertrophied hearts during the perfusion protocols.

BaCl₂ experiments. The effects of BaCl₂ on APD₉₀ are shown in figure 3. During BaCl₂ perfusion, APD₉₀ increased significantly in both hypertrophied and sham hearts (P < .001). BaCl₂ significantly prolonged APD₉₀ in the hypertrophied hearts compared with the control hearts (P < .01), and these effects were independent of afterload. The interaction of barium's effects on APD₉₀ and the presence of hypertrophy was highly statistically significant by factorial analysis of variance (P = .001). The change in APD₉₀ during BaCl₂ perfusion at low afterload tended to be greater in the hypertrophied hearts than in the control hearts at 0.025 mM (35 ± 14 msec vs. 20 ± 11 msec) and 0.05 mM (37 ± 5 msec vs. 22 ± 13 msec, P = .09). The change in APD₉₀ during BaCl₂ perfusion at high afterload was greater in the hypertrophied hearts than in the control hearts at 0.025 mM BaCl₂ (30 ± 6 msec vs. 14 ± 8 msec, P < .05) and 0.05 mM BaCl₂ (40 ± 13 msec vs. 20 ± 14 msec, P < .05). BaCl₂ did not significantly alter the effects of increasing afterload on the change in APD₉₀ in the control or the hypertrophied hearts (table 3).

View larger version (24K): [in this window] [in a new window]   Fig. 3.   Monophasic APD90 recorded from the posterior wall of the left ventricle in eight sham hearts () and eight hypertrophied hearts () before and during BaCl2 perfusion. Low afterload was 30 cm H2O, and high afterload was 100 cm H2O. Cardiac output was constant at 60 ml/min. Data are mean ± S.D.

Dofetilide experiments. The effects of dofetilide on APD₉₀ are shown in figure 4. During dofetilide perfusion, APD₉₀ increased significantly in both hypertrophied and sham hearts (P < .001). APD₉₀ remained greater during dofetilide perfusion in the hypertrophied hearts than in the sham hearts at low (P < .05) and high afterload (P < .05). However, a significant interaction between dofetilide's effects on APD₉₀ and the presence of hypertrophy was not identified. The change in APD₉₀ during dofetilide perfusion was similar in the sham hearts and in the hypertrophied hearts at low afterload (45 ± 18 msec vs. 40 ± 27 msec, P = N.S.) and high afterload (47 ± 18 msec vs. 49 ± 19 msec, P = N.S.). Dofetilide did not alter the effects of increasing afterload on the change in APD₉₀ in the control or the hypertrophied hearts (table 3).

View larger version (20K): [in this window] [in a new window]   Fig. 4.   Monophasic APD90 recorded from the posterior wall of the left ventricle in eight sham hearts () and eight hypertrophied hearts () before and during dofetilide perfusion. The format is the same as in figure 3. Data are mean ± S.D.

4-AP experiments. The effects of 4-AP on APD₉₀ are shown in figure 5. The low concentration of 4-AP caused slight prolongation of APD₉₀ in both sham and hypertrophied hearts (P = N.S.). The higher concentration of 4-AP caused significant prolongation of APD₉₀ in the sham and the hypertrophied hearts (P < .05). However, APD₉₀ was similar in both groups, and the magnitude of change in APD₉₀ was similar in the sham and the hypertrophied hearts at low afterload (25 ± 24 msec vs. 23 ± 17 msec) and high afterload (17 ± 17 msec vs. 24 ± 30 msec), respectively (P = N.S.). During 4-AP perfusion, an increase in afterload was associated with less shortening of APD₉₀ compared with base line in both sham and hypertrophied hearts, but the differences were not statistically significant (table 3).

View larger version (22K): [in this window] [in a new window]   Fig. 5.   Monophasic APD90 recorded from the posterior wall of the left ventricle in eight sham hearts () and in eight hypertrophied hearts () before and during 4-AP perfusion. The format is the same as in figure 3. Data are mean ± S.D.

Voltage-clamp experiments. The whole-cell capacitance was significantly greater in myocytes from hypertrophied hearts (199.5 ± 80.5 pF) compared with control hearts (131.7 ± 48.4 pF, P < .002). The average current-voltage relationships for I_K1 recorded in ventricular cells isolated from the control hearts and the hypertrophied hearts are shown in figure 6A. The average amplitude of I_K1, normalized to the cell capacitance, measured at a membrane potential of 90 mV was greater in control myocytes (4.62 ± 1.65 pA/pF, n = 17) than in hypertrophied myocytes (3.24 ± 1.02 pA/pF, n = 8, P < .05). The average amplitude of the outward portion of I_K1 is shown in the insert. I_K1 measured at a membrane potential of 70 mV was greater in control myocytes (3.09 ± 0.74 pA/pF) than in hypertrophied myocytes (2.48 ± 0.58 pA/pF, P < .05; see the insert).

View larger version (19K): [in this window] [in a new window]   Fig. 6.   A) IK1 current-voltage (I-V) relationships. Plot of I-V relationships in control myocytes (, n = 17) and hypertrophied myocytes (, n = 8). The insert shows the I-V relationship of the outward portion of IK1 at increased scale. Pulse protocol involved a holding potential of 50 mV followed by hyperpolarizing or depolarizing step pulses in 10-mV increments from 110 mV to 40 mV. After each 1-sec step pulse, the cell was returned to the holding potential. Data are mean ± S.D. *P < .05. B) IKr current-voltage relationships. Plot of I-V relationships in control myocytes (, n = 15) and hypertrophied myocytes (, n = 5). The pulse protocol involved a holding potential of 40 mV followed by depolarizing step pulses in 10-mV increments from 30 mV to +60 mV. Each pulse duration was 1 sec. Data are mean ± S.D. C) Ito current-voltage relationships. Plot of I-V relationships in control (, n = 13) and hypertrophied (, n = 5) myocytes. The insert shows IKsus. Ito was determined by subtracting IKsus from the peak outward current. The pulse protocol involved a holding potential of 80 mV followed by an initial depolarizing pulse of 40 mV followed by 10-mV decrements to 70 mV. Each pulse duration was 500 msec. Data are mean ± S.D., *P < .05.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

This present study yielded four major observations. 1) Prolongation of the ventricular APD in rabbits with hypertension secondary to aortic constriction is dependent on the mean arterial pressure. 2) BaCl₂ at concentrations that selectively block I_K1 in rabbit ventricle (Giles and Imaizumi, 1988; Wang et al., 1993) produces a greater prolongation of the monophasic APD in hypertrophied ventricle than in sham-operated control hearts. 3) Dofetilide at concentrations that selectively block I_Kr in rabbit ventricle (Jurkiewicz and Sanguinetti, 1993) and 4-AP at concentrations that block the voltage-sensitive Ca⁺⁺-independent component of I_to in rabbit ventricle (Campbell et al., 1993) produce similar prolongation of APD in hypertrophied and control hearts during pacing at a cycle length of 400 msec. 4) The increased effects of BaCl₂ on APD prolongation in hypertrophied hearts are associated with decreased current densities of I_K1 and I_to in hypertrophied myocytes compared with control hearts.

Abnormalities of repolarization in hypertrophy. Prolongation of APD is the electrophysiologic abnormality most frequently reported in cardiac hypertrophy (Aronson, 1991; Hart, 1994; Tomaselli et al., 1994). APD prolongation has been observed in spontaneously hypertensive rats (Cerbai et al., 1994; Hart, 1994; Kohya et al., 1988; Kohya et al., 1995), in cats with renovascular hypertension (Keung, 1989), in cats with aortic constriction (Kowey et al., 1991; Tomita et al., 1994) and in cats with right ventricular hypertrophy secondary to pulmonary vascular hypertension (Kleiman and Houser, 1989). Ventricular APD prolongation has also been described in rabbits with heart failure secondary to aortic insufficiency and aortic constriction (Bril et al., 1991). In the present study, we observed APD prolongation in isolated perfused working rabbit hearts with left ventricular hypertrophy secondary to aortic constriction. Although increasing afterload shortened APD, this change was similar at base line in both control and hypertrophied hearts. In spontaneously hypertensive rats (Cerbai et al., 1994) and in cats with renovascular hypertension (Kowey et al., 1991; Rials et al., 1995), the APD prolongation observed in left ventricular hypertrophy increases over time. In the present study, mean arterial pressure was a univariate predictor of APD and tended to be an independent predictor of APD (P = .08). The number of days after aortic banding (P = .07) also tended to be an independent predictor of APD. The variation in the degree of hypertrophy observed among the three groups probably reflects differences in mean arterial pressure (table 1).

Outward currents in ventricular hypertrophy. The APD prolongation observed in ventricular hypertrophy has been attributed to abnormalities of the inward calcium current (Keung, 1989; Kleiman and Houser, 1988; Nordin et al., 1989) and to altered outward potassium currents (Hart, 1994; Tomaselli, 1994). However, in moderate hypertrophy the calcium current density is probably unchanged (Hart, 1994). Decreased I_to density (Benitah et al., 1993; Tomita et al., 1994; Xu and Best, 1991), I_K density (Furukawa et al., 1994; Kleiman and Houser, 1989) and I_K1 density (Brooksby et al., 1993) have been described in ventricular hypertrophy. The importance of these repolarizing currents probably depends on the species studied (Hart, 1994). At present, it is not known whether changes in one of these currents contribute predominantly to the APD prolongation observed in left ventricular hypertrophy. In rabbit ventricular tissue, I_to, I_K1 and I_K contribute to repolarization (Carmeliet, 1993; Giles and Imaizumi, 1988; Veldkamp et al., 1993). We therefore chose to evaluate the effects of compounds that selectively block I_K1, I_Kr and I_to. BaCl₂ in concentrations < 500 µM selectively blocks I_K1 (Giles and Imaizumi, 1988; Mugelli et al., 1983). In keeping with this effect, BaCl₂ at concentrations  50 µM did not alter pacing thresholds, nor did we observe spontaneous depolarizations. Dofetilide at concentrations  1 µM selectively blocks I_Kr (Jurkiewicz and Sanguinetti, 1993). 4-AP at concentrations  1.0 mM blocks the voltage-sensitive Ca⁺⁺-independent component of I_to in rabbit ventricle (Campbell et al., 1993; Fedida and Giles, 1991).

Pharmacodynamics of potassium channel blockers. The effects of an ion channel blocker are determined by the surface density of the ion channels on the cell membrane, the dominance of the current, the concentration of the drug, the kinetics of the drug with the channel and state-specific interactions of the drug with the channel. In results consistent with previous studies, we have observed a reduction in the density of I_K1 in left ventricular hypertrophy (Brooksby et al., 1993). Thus it is not surprising that in the present study, the effects of BaCl₂ on APD prolongation were greater in the hypertrophied left ventricle than in the normal ventricle. I_K1 and I_to are the dominant currents that contribute to repolarization in rabbit ventricle (Giles and Imaizumi, 1988). A concomitant reduction in the density of other repolarizing currents (e.g., I_to) in hypertrophy compared with I_K1 would result in I_K1 being the dominant residual repolarizing current. Therefore, blockade of this dominant residual current would be expected to produce an exaggerated response (Haynh et al., 1992; Wang et al., 1996). In the present study, I_to current amplitude was reduced in hypertrophied ventricular myocytes compared with controls. Thus the increased effect of BaCl₂ on prolongation of APD might be due in part to a reduction in the contribution of I_to to ventricular repolarization in cardiac hypertrophy.

Contraction-excitation feedback. Changes in myocardial stretch or load have been association with changes in cardiac electrophysiologic characteristics. Increasing afterload causes shortening of VERP and APD, an effect that is due to the associated increase in preload (Franz, 1996; Hansen, 1993). The cellular mechanisms of this effect have been attributed to changes in intracellular calcium (Franz, 1996). Little is known about the effect of potassium outward currents or left ventricular hypertrophy on afterload/preload-induced changes in APD. In the present study, the presence of hypertrophy did not significantly alter shortening of APD after an increase in afterload. None of the selective potassium current blockers caused a significant change in the afterload-induced shortening of APD₉₀.

Limitations. The degree of hypertrophy in our experimental model develops more rapidly than would be expected in clinical left ventricular hypertrophy. Therefore, the changes observed might differ if the degree of hypertension and/or the timing of the experiments were altered. It is possible that different effects of these ion channel blockers would be observed at different stages of cardiac hypertrophy. Furthermore, small differences in the degree of hypertrophy might have influenced the magnitude of the effects of the ion channel blockers. Different effects on APD might have been observed at different pacing rates, so we cannot exclude the possibility that 4-AP effects might have differed between hypertrophied and control hearts at slower heart rates (Antzelevitch et al., 1991). Because these were isolated perfused hearts, the role of the autonomic nervous system could not be assessed.

Conclusions. The effects of selective I_K1, I_Kr and I_to blockers on APD differ in left ventricular hypertrophy. The effects of an I_K1 blocker, but not those of I_Kr or I_to blockers, are increased at rapid heart rates by the presence of hypertrophy. These effects occur in association with a reduction in I_K1 and I_to, but not in I_Kr, density in hypertrophied myocytes compared with control myocytes. Thus the increased effects of BaCl₂ on APD in hypertrophy may be secondary to a reduction in I_K1 amplitude and, in part, to a change in the balance of repolarizing currents in ventricular hypertrophy.