Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Amiodarone is widely believed to be the most effective antiarrhythmic drug available at present. For example, it is the only antiarrhythmic agent that has been shown to reduce arrhythmic death in patients with frequent ventricular ectopy postmyocardial infarction (Cairns et al., 1997) and appears to have superior efficacy to other antiarrhythmic drugs for sinus rhythm maintenance in patients with atrial fibrillation (Nattel, 1995). The cellular actions of amiodarone were first described by Singh and Vaughan Williams in 1970, who noted that chronic amiodarone administration increases action potential duration (APD) without affecting maximum upstroke velocity in rabbit ventricle (Singh and Vaughan Williams, 1970). They remarked on the similarity between the actions of amiodarone and hypothyroidism on the heart and showed that thyroxine reversed the effects of amiodarone, leading them to conjecture that amiodarone may exert its effects by interfering with the action of thyroid hormone on the heart.

Since the classical studies of Singh and Vaughan Williams, many investigators have evaluated possible relations between amiodarone and thyroid actions on the heart. Several investigators have noted the similarity between various cardiac effects of amiodarone and hypothyroidism on APD (Singh et al., 1970), on ventricular refractoriness (Patterson et al., 1986; Liu et al., 1996), on ventricular fibrillation threshold (Liu et al., 1996), and on ventricular myosin enzyme isoforms (Wiegand et al., 1986). Hypothyroidism has been found to prevent amiodarone effects on heart rate and Q-T interval (Talajic et al., 1989) and on K⁺ currents in cultured rat cardiomyocytes (Guo et al., 1997). A variety of studies suggest that amiodarone and/or its desethyl metabolite inhibit binding of triiodothyronine to its nuclear receptor and produce a variety of cardiac metabolic changes that resemble those of hypothyroidism (Latham et al., 1987; Gotzsche, 1993; Gotzsche and Orskov, 1994; Drvota et al., 1995). On the other hand, mimicking at least one of the hypothyroid actions of amiodarone, prevention of conversion of thyroxine to triiodothyronine, did not reproduce the effects of the drug on heart rate and the Q-T interval (Stäubli and Studer, 1986).

Most of the work studying the relationship between the actions of amiodarone and those of hypothyroidism and the interaction between amiodarone and thyroid effects has been performed in whole-animal and standard microelectrode studies. Very little is known about the actions and interactions of chronic amiodarone therapy and hypothyroidism at the level of the ionic currents that control repolarization. The goal of the present studies was to establish the effects of chronic amiodarone administration and hypothyroidism, separately and in combination, on K⁺ currents in guinea pig ventricular myocytes. Complementary information, in terms of effects on the ECG in vivo and action potentials of isolated cells, was also obtained. In this way, we hoped to establish the extent to which the ionic effects of amiodarone can be attributed to an interaction with thyroid effects on the heart.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Experimental Groups. Adult male Hartley albino guinea pigs were assigned to one of the following groups: a control group (n = 25), a hypothyroid group (n = 14), an amiodarone-treated group (n = 9), a hypothyroid amiodarone-treated group (n = 6), and a dimethyl sulfoxide (DMSO)-treated group (n = 4). The procedures followed were in accordance with the guidelines of the Montreal Heart Institute Animal Ethics Committee and the Canadian Council on Animal Care. Animals assigned to the hypothyroid group were thyroidectomized by Charles River (St. Constant, QC, Canada) after anesthesia with xylazine (5 mg/kg, i.m.; Miles Canada, Inc.) and ketamine (40 mg/kg, i.m; Rogar/STP, Inc.). Subsequently, these guinea pigs were treated with 5-propylthiouracil (PTU; dissolved in the drinking water at a concentration of 0.05%) (Sigma Chemical Co., St. Louis, MO) for 6 to 8 weeks. In addition, CaCl₂ was added to the drinking water of hypothyroid animals at a concentration of 1% to avoid the possibility of hypocalcemia due to parathyroid damage at the time of thyroidectomy. Control animals were followed in the same way and for the same period of time as hypothyroid animals. In the animals of the amiodarone group, the drug was administered i.p. at a daily dose of 80 mg/kg body weight (dissolved in DMSO) over 7 days. Animals assigned to the hypothyroid amiodarone-treated group were subjected to both thyroidectomy/PTU treatment and amiodarone injections, the latter begun when hypothyroid effects reached steady state (8 weeks) and continued for 7 days. DMSO-treated animals were injected i.p. daily for 7 days with DMSO, the vehicle for amiodarone administration, in the same quantities as used for amiodarone administration. Weight and ECGs were obtained on a weekly basis. At the end of the observation period, animals were sacrificed by cervical dislocation, and their hearts were removed for cell isolation.

ECG Recordings. Six-lead electrocardiographic recordings (leads I, II, III, aVL, aVR, and aVF) were obtained (MT 95000, Astro-Med Inc.; platinum subdermal needle electrodes) after sedation with acepromazine (0.1 mg/kg, i.m.; Ayerst Laboratories, New York, NY) and ketamine (40 mg/kg, i.m; Rogar/STP, Inc.). A paper speed of 200 mm/s was used to achieve a measurement accuracy of ± 2.5 ms. The average of three successive measurements was used to determine the R-R, P-R, Q-R-S, and Q-T intervals. ECGs were recorded at baseline and once a week thereafter.

Cell Isolation and Solutions. Guinea pigs were sacrificed by cervical dislocation, and the hearts were quickly excised and mounted on a Langendorff apparatus. The hearts were retrogradely perfused via the aorta with oxygenated (100% O₂, pH adjusted to 7.35 with NaOH) Tyrode's solution containing: 136 mmol/liter NaCl, 5.4 mmol/liter KCl, 2.0 mmol/liter CaCl₂, 1.0 mmol/liter MgCl₂, 0.33 mmol/liter NaH₂PO₄, 5 mmol/liter HEPES, and 10 mmol/liter glucose at 37°C. When clear of blood, the perfusate was changed to a nominally Ca²⁺-free Tyrode's solution until contraction had ceased completely. Perfusion was then continued with the same solution containing 0.03% collagenase (Type II, Worthington Biochemical) and 1% BSA (Sigma Chemical Co.) until left ventricular tissue was softened. Small pieces of tissue were removed with forceps and mechanically dissociated by trituration. The isolated cells were kept in a storage solution containing 20 mmol/liter KCl, 10 mmol/liter KH₂PO₄, 25 mmol/liter glucose, 40 mmol/liter mannitol, 70 mmol/liter L-glutamic acid, 10 mmol/liter -hydroxybutyric acid, 20 mmol/liter taurine, and 10 mmol/liter ethylene glycol bis(-aminoethyl ether)-N,N,N,'N'-tetraacetic acid, along with 1% albumin (pH adjusted to 7.35 with KOH).

Voltage-Clamp Technique. Only quiescent, rod-shaped cells with clear cross striations were studied. Ionic currents were recorded with the whole-cell configuration of the voltage-clamp technique. Borosilicate glass electrodes (outer diameter, 1.0 mm) with resistances from 2.5 to 6 megaohms when filled with pipette solution were connected to a patch clamp amplifier (Axopatch 200A, Axon Instruments, Burlingame, CA). Data were sampled with an analog to digital converter (Digidata 1200, Axon Instruments) and stored on the hard disk of a computer for subsequent analysis. Recordings were low-pass filtered at 2 kHz.

Data Analysis. Group data are expressed as mean ± S.E.M. Statistical comparisons between the different experimental groups were obtained with ANOVA. Differences with a two-tailed P < 0.05 were considered statistically significant. Comparisons between multiple group means were performed with a Bonferroni-corrected t test for all pairwise group comparisons. A nonlinear least-square curve-fitting program (CLAMPFIT in pCLAMP 6.0 or Sigma Plot) was used to perform curve-fitting procedures.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Effects on ECG. In hypothyroid animals, the first electrocardiographic changes were noted after 4 weeks of PTU exposure, and the ECG stabilized by the end of 8 weeks. We have shown previously that amiodarone effects on the ECG of the guinea pig reach steady state after 1 week of amiodarone therapy (Talajic et al., 1989). Examples of typical ECGs of one animal/group at the end of the study period in each group are shown in Fig. 1. Animals in the amiodarone, hypothyroid, and amiodarone-hypothyroid group had significantly increased R-R and Q-T intervals compared with controls, with no change in P-R or R-R intervals. DMSO itself had no effect on the ECG. Table 2 shows mean R-R and Q-T intervals at the time of study for each group of animals.

View larger version (132K): [in this window] [in a new window]   Fig. 1.   Electrocardiographic recordings from guinea pigs of the five different experimental groups. ECGs were obtained at a paper speed of 200 mm/s 1 day before the patch-clamp experiments were performed. The figure shows a representative example of lead II for each experimental group. Individual R-R and Q-T intervals are also given for each recording. Please note that DMSO treatment did not alter heart rate or Q-T intervals. In contrast, hypothyroidism, amiodarone treatment, and the combination of both substantially prolonged the R-R and the Q-T interval.

Effects on Delayed Rectifier Current (I_K). I_K was recorded with the use of 3-s depolarizing pulses (0.1 Hz) from a holding potential of 50 mV to test potentials from 40 to +70 mV, followed by a 2-s repolarizing pulse to 40 mV to record tail current. Initial measurements were performed 15 min after cell-membrane rupture, and the protocol was run at least three times in each cell in 10-min intervals to detect rundown of I_K. In cells with a stable I_K (<10% rundown over 20 min), 1 µM dofetilide was added to the bath to block I_Kr, and after 10 min, the protocol was repeated. I_Kr was evaluated on the basis of dofetilide-sensitive I_K, and I_Ks was determined from the dofetilide-resistant component (Lei and Brown, 1996; Salata et al., 1996). Washout of dofetilide was obtained in seven cells, and a mean reversal of 92.4% in drug effect was observed. Cells with rundown of >10% (6.3% of cells) were rejected. Examples of currents recorded before and after dofetilide, as well as dofetilide-sensitive currents obtained by digital subtraction, are shown in Fig. 2, A, B, and C, respectively.

View larger version (25K): [in this window] [in a new window]   Fig. 2.   Delayed rectifier potassium current in guinea pig ventricle. A, typical example of delayed rectifier currents recorded at 36°C with 3-s depolarizing steps to test potentials between 40 and +70 mV from a holding potential of 50 mV. Tail currents were obtained upon repolarization to 40 mV. The bath solution contained 5 µM nifedipine to block ICa. The components of IK were separated pharmacologically with the use of 1 µM dofetilide, which specifically inhibits the rapid component IKr of IK. B, dofetilide-resistant component (IKs) at potentials between 40 and +70 mV. These recordings were obtained after a 10-min exposure of the cell to dofetilide. C, the dofetilide-sensitive current (IKr), obtained by digital subtraction of currents in B from currents in A. For reasons of clarity, only test potentials between 30 and +30 mV are shown in C. 0 indicates zero current level.

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Dofetilide-resistant component (IKs) of the delayed rectifier potassium current. A, representative recordings obtained with the protocol shown at 36°C, in the presence of 1 µM dofetilide to block IKr. B, mean IKs density-voltage relations recorded during 3-s voltage steps from 50 mV to the test potentials (TP) indicated. The results shown are step current densities for 8, 9, 8, 7, and 5 cells for control, hypothyroid, amiodarone, amiodarone plus hypothyroid, and DMSO groups, respectively. *, **P < .05, 0.01 versus control, respectively.

View larger version (23K): [in this window] [in a new window]   Fig. 4.   Dofetilide-sensitive component (IKr) of the delayed rectifier potassium current. IKr step currents were obtained by digital subtraction of delayed rectifier current in the presence of 1 µM dofetilide from those recorded before dofetilide infusion. A, representative recordings from each group. B, IKr step current density-voltage relations for 8, 9, 8, 7, and 5 cells for control, hypothyroid, amiodarone, amiodarone plus hypothyroid, and DMSO groups, respectively. **P < .01 versus control; ++P < .01 versus hypothyroid group.

Voltage and Time Dependence of I_K Activation. The voltage-dependent activation of I_Kr and I_Ks was analyzed by normalizing dofetilide-sensitive and dofetilide-resistant tail currents at each test potential (obtained with the voltage protocols illustrated in Figs. 3 and 4) to the current at the most positive voltage. A Boltzmann function was used to fit the activation curves of I_Kr and I_Ks. In all groups, I_Kr activated at negative voltages, with a half-activation voltage (V_h) of 17.1 ± 3.1 mV and a slope factor (k) of 9.0 ± 1.2 mV under control conditions and no significant differences among groups. These values are similar to results reported previously in normal guinea pig ventricle (Sanguinetti and Jurkiewicz, 1990). I_Ks activated at more positive potentials. For example, V_h under control conditions averaged 22.6 ± 2.4 mV, and k was 13.4 ± 1.1 mV. Similar to the results for I_Kr, the I_Ks activation voltage dependence was not altered in any treatment group (Table 4) and was similar to values reported previously in normal guinea pig ventricular myocytes (Sanguinetti and Jurkiewicz, 1990).

Inward Rectifier K⁺ Current. Representative recordings of I_K1 from various experimental groups are shown in Fig. 5. Cells from animals treated with amiodarone showed smaller currents than cells from the other three groups. Mean I_K1 density-voltage relations for all groups (cell numbers were 13, 20, 19, 7, and 9 for control, hypothyroid, amiodarone, hypothyroid/amiodarone, and DMSO, respectively) are shown in Fig. 6A. The smaller outward currents are shown on an expanded scale in Fig. 6B. Amiodarone significantly reduced both inward and outward components of I_K1, whether in the presence or absence of hypothyroidism, whereas hypothyroidism and DMSO did not alter I_K1 compared with control. For example, at 100 mV, current density for control cells averaged 36 ± 4 pA/pF, compared with 32 ± 3 pA/pF in DMSO-treated cells and 34 ± 2 pA/pF in hypothyroid cells (P was not significant for each group versus control). In contrast, the values at 100 mV were 20 ± 1 pA/pF for amiodarone treatment alone and 22 ± 2 pA/pF in hypothyroid amiodarone-treated animals. The outward component of I_K1 at voltages positive to the reversal potential of 70 mV was similarly affected. For example, at 40 mV, control cells had an average current density of 5.3 ± 0.4 pA/pF, which was not altered by DMSO (5.9 ± 0.5 pA/pF) or by hypothyroidism (5.6 ± 0.3 pA/pF), but was substantially reduced in the amiodarone (1.7 ± 0.2 pA/pF) and hypothyroid/amiodarone (2.3 ± 0.4 pA/pF)-treated groups.

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Amiodarone treatment but not hypothyroidism decreases IK1. Inward rectifier currents recorded from representative cells in each group at 36°C. Test pulses were 200-ms steps at 0.5 Hz to test potentials shown. 0 indicates zero current level.

View larger version (19K): [in this window] [in a new window]   Fig. 6.   IK1 density-voltage relations (n = 13, 20, 19, 7, and 9 cells for control, hypothyroid, amiodarone, hypothyroid plus amiodarone, and DMSO groups, respectively). A, IK1 density-voltage relations at all test potentials studied. B, an enlargement of the inset in A to show clearly the outward currents carried by IK1. Where error bars are absent, they fell within the symbol for mean. **P < .01 versus control; ++P < .01 versus hypothyroid cells.

Action Potential Characteristics. Representative examples of action potentials of cells from each group at a stimulation frequency of 4 Hz are shown in Fig. 7A. Amiodarone treatment, hypothyroidism, and the combination of both interventions were associated with a significant prolongation of the APD. Action potential shape, resting membrane potential, and action potential amplitude were similar in all experimental groups. APD to 90% repolarization (APD₉₀) is plotted as a function of stimulation frequency in Fig. 7B. In cells from hypothyroid animals, APD₉₀ was significantly longer than under control conditions for all stimulation frequencies. For example, at 4 Hz, the values were 87 ± 5 ms under control (n = 13) and 140 ± 6 ms under hypothyroid conditions (n = 15; P < .01 versus control). Amiodarone-treated cells had an APD₉₀ that was in the range of the hypothyroid group (e.g., at 4 Hz, 149 ± 6 ms, n = 23; P < .01 versus control). Amiodarone treatment in hypothyroid animals was associated with an additional substantial AP prolongation that was most prominent at slow frequencies but remained significant at higher frequencies. At 4 Hz, APD₉₀ averaged 193 ± 9 ms in hypothyroid amiodarone-treated cells (n = 9; P < .001 versus control and P < .01 versus amiodarone or hypothyroidism alone). DMSO treatment had no effect on APD₉₀ (e.g., at 4 Hz, 102 ± 7 ms, n = 9; P was not significant, versus control).

View larger version (24K): [in this window] [in a new window]   Fig. 7.   A, action potentials recorded in representative cells from each group at a stimulation frequency of 4 Hz. B, mean APD90 for stimulation frequencies between 0.1 and 4 Hz, as recorded in 13 control cells, 15 cells from hypothyroid guinea pigs, 23 cells from amiodarone-treated guinea pigs, 9 cells from amiodarone-treated hypothyroid animals, and 9 cells from DMSO-treated guinea pigs. **P < .01 versus control; ++P < .01 versus hypothyroid group; §§P < .01 versus amiodarone group.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

In this study, we examined the effects of chronic amiodarone therapy, hypothyroidism, and the combination of amiodarone and hypothyroidism on K⁺ currents in guinea pig ventricular myocytes. Whereas hypothyroidism only affected I_Ks, amiodarone altered I_Ks (albeit to a lesser extent than hypothyroidism), I_Kr, and I_K1. Combining amiodarone with hypothyroidism led to a pattern characteristic of both effects individually, with strong reductions in I_Ks along with decreases in I_Kr and I_K1. These observations suggest that amiodarone has actions on repolarizing current that are independent of thyroid effects, a concept supported by the additional changes in APD observed in the presence of amiodarone and hypothyroidism compared with either intervention alone.

Comparison with Previous Studies of Actions of Amiodarone on K⁺ Currents. The first published voltage-clamp studies of the ionic actions of amiodarone noted reductions in "total outward current" (predominantly I_K) in frog atrial muscle preparations at concentrations above 10 µM (Néliat et al., 1982). Balser et al. (1991) studied the effects of acute amiodarone administration (10 µM) on guinea pig ventricular myocytes and observed a reduction in the La²⁺-resistant component but no effect on the La³⁺-sensitive component, compatible with principal actions on I_Ks. On the other hand, Kamiya et al. (1995) and Varró et al. (1996) observed that acute amiodarone administration decreased I_K in rabbit ventricular myocytes, a species in which I_Kr is the predominant component under usual experimental conditions (Colatsky et al., 1990). I_to was unaffected by 10 µM amiodarone in rabbit myocytes (Kamiya et al., 1995; Varró et al., 1996). Sato et al. (1994) noted that amiodarone causes a decrease in I_K1 whole-cell current in guinea pig ventricular cells, along with an increase in interburst interval at the single channel level. Guo et al. (1997) showed that acute amiodarone exposure decreases the density of both I_to and end-pulse current in neonatal rat ventricular myocytes.

Comparison with Previous Studies of Amiodarone-Thyroid Interaction. As indicated in the Introduction, many lines of evidence point to the possibility of thyroid-mediated actions of amiodarone on cardiac repolarization; however, the findings available in the literature are insufficient to determine whether the APD-prolonging effects of amiodarone can be attributed to an inhibition of the cardiac actions of thyroid hormone. In the present study, we found qualitative differences between the changes in K⁺ currents observed in hypothyroid compared with amiodarone-treated animals. Furthermore, guinea pigs exposed to both thyroidectomy and amiodarone therapy showed a combination of the effects of either alone and had significantly longer action potentials than animals exposed to either intervention alone. These observations suggest that chronic amiodarone therapy has effects on repolarization that are not due to inhibition of thyroid action alone and may well be due to direct inhibitory effects on cardiac K⁺ channels, in agreement with previous observations of the effects of acute amiodarone administration (Néliat et al., 1982; Balser et al., 1991; Sato et al., 1994; Kamiya et al., 1995; Varró et al., 1996).

Potential Limitations. Any study of the effects of chronic processes like hypothyroidism and maintained amiodarone therapy must be performed in parallel and separate groups of animals, losing the statistical power of using each animal as its own control. Distortions can also occur if cells are not well distributed across hearts for each determination. We were careful to study a similar number of cells for each current measurement in each heart of each group, to avoid biasing the results. The similarity under Results obtained for control and DMSO groups is also reassuring in terms of the robustness of the data.

Potential Significance. Amiodarone is an important and widely used antiarrhythmic drug, with unique properties that are likely to result in increased use in the short term (Nademanee et al., 1993; Podrid, 1995; Singh, 1995; Link et al., 1996). It is therefore important to understand its fundamental mechanisms of action. Furthermore, there is considerable interest in developing new antiarrhythmic agents that share the beneficial properties of amiodarone without some of its drawbacks, including hypothyroidism, a variety of systemic toxicities, and very slow elimination after chronic use. Two candidate compounds have already been developed (Finance et al., 1995; Manning et al., 1995a,b; Raatikainen et al., 1996). A better understanding of the mechanisms of the properties of amiodarone, particularly its important class III actions (Singh et al., 1970), is likely to be a necessary component of this effort.