Introduction

Top Abstract Introduction Methods Results Discussion References

The mechanism of action of the widely used antiarrhythmic agent amiodarone is thought to involve use-dependent block of Na⁺, Ca⁺⁺, and K⁺ channels (see Nattel et al., 1992, for review). However, when studying electrophysiological effects of amiodarone, the Na⁺-K⁺ pump should also be considered. The pump directly, or indirectly through the operation of secondary active ion transport mechanisms, maintains the transmembrane ionic gradients. These gradients in turn drive the channel currents that are modulated by amiodarone. In addition, the electrogenic Na⁺-K⁺ pump current, arising from the 3Na⁺:2K⁺ exchange ratio, contributes to repolarization of the cardiac action potential (Gadsby et al., 1985).

Treatment with amiodarone influences physicochemical properties of the cell membrane, including fluidity and cholesterol content (Chatelain et al., 1986). Because the Na⁺-K⁺ pump is affected by such changes in membrane properties (Cornelius, 1991), it seems reasonable to speculate that amiodarone may alter pump function. Amiodarone might also alter pump function through an alternative mechanism. The drug is believed to induce a state of cellular hypothyroidism with chronic use (Patterson et al., 1986; Singh et al., 1983). Thyroid status in turn is an important determinant of synthesis of Na⁺-K⁺ pump units (Doohan et al., 1995; Kjeldsen et al., 1986).

Several previous studies using isolated myocardial membrane or microsomal preparations (Broekhuysen et al., 1972; Dzimiri and Almotrefi, 1991) have suggested that acute exposure to amiodarone decreases Na⁺-K⁺/ATPase activity, the enzymatic equivalent of the Na⁺-K⁺ pump. Although demonstrating that amiodarone has the potential to affect pump activity, the physiological relevance of such studies is unclear. The drug concentrations used were much higher than those encountered clinically, and the studies were performed using K⁺ and Na⁺ concentrations expected to saturate binding sites rather than at physiologically relevant concentrations. In addition, Na⁺-K⁺/ATPase studies require the pump molecule to be isolated from the native lipid membrane and purified. As the membrane lipid environment modulates pump activity (Cornelius, 1991), it is unclear how these results relate to the activity of in situ pumps.

In the present study, we examined the effects of both acute and chronic amiodarone exposure on the Na⁺-K⁺ pump in intact cardiac myocytes. We used the whole-cell patch-clamp technique to measure I_p. This approach allows independent control of ligands for the Na⁺-K⁺ pump on both sides of the intact native membrane, as well as control of membrane potential, variables that are important determinants of pump function.

	Methods

Top Abstract Introduction Methods Results Discussion References

Treatment protocols. Male New Zealand White rabbits, weighing 2.5 to 3.0 kg, were used in the study. Amiodarone was supplied in its hydrochloride powder form. In experiments examining the effects of chronic treatment, gelatin capsules containing appropriate quantities of amiodarone were made up at two weekly intervals. A single daily amiodarone capsule was administered to rabbits orally in a dose of 80 mg/kg/day for 4 weeks. This protocol was based on a previous study in rabbits demonstrating clinically relevant changes in electrophysiological and electrocardiographic parameters with 50 to 100 mg/kg oral amiodarone daily (Kodama et al., 1992). Control rabbits received empty gelatin capsules in the same manner.

Measurement of I_p. Single ventricular myocytes were isolated as described previously(Hool et al., 1995). After isolation, myocytes were stored at room temperature until used for experimentation. Myocytes were used on the day of isolation only, and pump currents were measured 2 to 10 hours after the heart was excised. Myocytes were voltage clamped with wide-tipped (4-5 µm) borosilicate glass pipettes made as described previously (Whalley et al., 1993). In experiments measuring I_p at a fixed membrane potential of 40 mV, pipettes were filled with a solution containing (in mM) 70 potassium glutamate, 1 KH₂PO₄, 5 HEPES, 5 EGTA, 2 MgATP and sodium glutamate plus TMA·Cl 90. The Na⁺ concentration in the solution was either 10 mM, which is near the physiological intracellular level, or 80 mM, a level expected to nearly saturate intracellular pump sites. The solution was titrated to pH 7.05 ± 0.01 at 35°C with 1 M KOH. In experiments designed to examine the relationship between I_p and membrane voltage pipettes were filled with a solution containing (in mM) 10 sodium glutamate, 1 KH₂PO₄, 5 HEPES, 5 EGTA, 2 MgATP, 60 TMA·Cl, 20 tetraethylammonium chloride, 70 CsCl and 50 aspartic acid. To examine the I_p-V_m relationship at a high intracellular Na⁺ level, the solution was similar except that sodium glutamate was 80 mM, CsCl was 65 mM, aspartic acid was 45 mM, and TMA·Cl was omitted. The solution was titrated to pH 7.05 ± 0.01 at 35°C with 1 M HCl. When filled with the above solutions, patch pipettes had resistances of 0.8 to 1.2 M.

Dose of ouabain. A previous study in noncardiac tissue suggested that amiodarone may competitively inhibit ouabain binding to the Na⁺-K⁺ pump (Prasada Rao et al., 1986). To ensure that the concentration of ouabain used in experimental protocols (100 µM) was sufficient to completely block the Na⁺-K⁺ pump in the presence of amiodarone, we performed a series of preliminary experiments on control myocytes and myocytes exposed to amiodarone in vitro. We found that after the pump had been inhibited by 100 µM ouabain, there was no additional shift in holding current on increasing the ouabain concentration to 500 µM, indicating that 100 µM ouabain caused complete pump inhibition. Unless otherwise indicated, I_p was identified by the shift in holding current induced by exposure of myocytes to 100 µM ouabain (Sigma Chemical, St. Louis, MO).

Statistics. Results are expressed as mean ± S.E.M. Statistical comparisons are made using Student's t test for paired or unpaired observations. Dunnett's test was used when the same control group was used for more than one comparison. P < .05 is regarded as significant in all comparisons. Nonlinear regression was used to fit the Hill equation to data.

	Results

Top Abstract Introduction Methods Results Discussion References

Effects of chronic amiodarone on ECG and thyroid function. Amiodarone is known to prolong the QT interval on the surface ECG and affects thyroid hormone levels by inhibiting peripheral conversion of T₄ to T₃. The effects of chronic treatment on thyroid function, heart rate, and QT_c are summarized in table 1. There was no significant difference in any of these parameters between the control and treated groups of rabbits at base line. Amiodarone treatment resulted in a significant increase in serum T₄ and a significant prolongation of QT_c. The treated rabbits had a serum amiodarone level of 1.14 ± 0.15 µmol/liter, which is similar to the concentration found during chronic treatment in humans (Debbas et al, 1984; Ikeda et al., 1984; Weinberg et al., 1993). The treatment protocol was well tolerated. One rabbit lost weight during the treatment period and was found to have a toxic serum amiodarone level of 3.9 µmol/liter. Its surface ECG showed slow (80 bpm) broad QRS complexes. Results from this rabbit were not included in the analysis.

Effect of chronic amiodarone treatment on I_p. Myocytes from 5 control rabbits and 5 rabbits treated with amiodarone were voltage-clamped at 40 mV, and I_p was measured using pipettes with Na⁺ concentrations ([Na]_pip) of 10 or 80 mM. Results from these experiments are summarized in figure 1A. When [Na]_pip was 10 mM, mean I_p of 14 myocytes from rabbits treated with amiodarone was 0.24 ± 0.03 pA/pF, whereas mean I_p of 10 myocytes from rabbits given placebo was 0.36 ± 0.03 pA/pF. This 33% difference was statistically significant. Mean I_p in myocytes from the rabbits given placebo was similar to mean I_p in myocytes from untreated rabbits reported previously (Gray et al., 1997). When [Na]_pip was 80 mM, amiodarone treatment resulted in a significant, 25% reduction in mean I_p (1.46 ± 0.07 pA/pF, n = 7, vs. 1.94 ± 0.05 pA/pF, n = 6). We conclude that chronic amiodarone treatment causes a reduction in overall Na⁺-K⁺ pump capacity.

View larger version (12K): [in this window] [in a new window]   Fig. 1.   A, Effect of chronic amiodarone (amio) treatment on Na+-K+ pump current (Ip), measured when the Na+ concentration in the patch pipette ([Na]pip) was 10 mM (open bars) and 80 mM (solid bars). Numbers in parentheses indicate the number of myocytes studied in each group. Amiodarone treatment significantly reduced Ip when [Na]pip was 10 mM and when [Na]pip was 80 mM. B, Effect of extracellular K+ ([K]o) on Ip in myocytes from rabbits chronically treated with amiodarone (, dashed line) and from control rabbits (, solid line). Ip has been normalized to the current measured when [K]o is 5.6 to eliminate variability of data arising from measurement of cell membrane capacitance.

Effect of chronic amiodarone on the affinity of the Na⁺-K⁺ pump for extracellular K⁺. Because it has been suggested that the effect of amiodarone on Na⁺-K⁺/ATPase is K⁺ dependent (Almotrefi and Dzimiri, 1991), we examined the effect of chronic amiodarone treatment on the apparent affinity of the pump for extracellular K⁺ concentrations ([K]_o). We measured changes in I_p in response to changes in [K]_o. To facilitate the detection of small pump currents at low [K]_o we used [Na]_pip of 80 mM to induce near-maximal activation of the pump at intracellular sites.

Effect of chronic amiodarone on the I_p-V_m relationship. Because amiodarone is an amphiphile and amphiphiles can alter the voltage dependence of the Na⁺-K⁺ pump (Läuger, 1991), we examined whether chronic amiodarone treatment affects the I_p-V_m relationship. After establishing the whole-cell configuration, we superfused myocytes with Ca⁺⁺-free modified Tyrode's solution, and we voltage-clamped the myocytes at 40 mV. We then applied 320-msec voltage steps in 20-mV increments to test potentials ranging from 100 to +60 mV. Averaged membrane currents at each test potential after superfusion of ouabain were subtracted from the respective averaged membrane currents before ouabain exposure to derive I_p at each test potential. Experiments were performed at [Na]_pip of both 10 and 80 mM. Figure 2A illustrates the voltage step protocol and an example of the resulting membrane currents for a myocyte studied using [Na]_pip of 10 mM. Details of the experimental protocol and data analysis have been published previously (Gray et al., 1997).

View larger version (11K): [in this window] [in a new window]   Fig. 2.   A, Voltage-clamp protocol and holding currents before and after exposure of a myocyte to 100 µM ouabain. The myocyte was from a rabbit treated with amiodarone. The current traces shown are the averaged currents from three applications of the voltage-clamp protocol. [Na]pip was 10 mM. B, Normalized Ip-Vm relationships for 13 myocytes isolated from rabbits treated with amiodarone () and 8 myocytes from control rabbits (). [Na]pip was 10 mM. C, Ip-Vm relationships for 7 myocytes from amiodarone-treated rabbits and 8 myocytes from control rabbits when [Na]pip was 80 mM. Amiodarone had no significant effect on the Ip-Vm relationships.

Effect of chronic amiodarone on intracellular Na⁺. If chronic amiodarone treatment inhibits Na⁺-K⁺ pump activity, it may be expected to alter steady-state intracellular Na⁺ activity (aⁱ_Na). However, the direction and magnitude of any such change in aⁱ_Na cannot be assumed because amiodarone may also influence Na⁺ influx. We therefore examined the effect of chronic amiodarone treatment on aⁱ_Na. We determined aⁱ_Na in papillary muscles from 11 control rabbits and 11 rabbits treated with amiodarone for 4 weeks. We used ion-sensitive microelectrodes to measure aⁱ_Na as described previously (Hool et al., 1995). One or two determinations of aⁱ_Na were made with separate microelectrode impalements in tissue from each rabbit. Where two measurements were made, the mean value was used for statistical comparisons. Determinations of aⁱ_Na were made only after microelectrode recordings had been stable for 20 minutes. The aⁱ_Na in control rabbits was 8.1 ± 0.4 mM. In amiodarone-treated rabbits, aⁱ_Na was 9.8 ± 0.8 mM. This difference was not statistically significant (P = .07). Amiodarone is expected to reduce Na⁺ influx via Na⁺ channels (Mason et al., 1984); because amiodarone reduces cardiac metabolic rate (Charlier et al., 1968), treatment may have reduced Na⁺ influx via the Na⁺-H⁺ exchanger. One may speculate that such decreases in Na⁺ influx partially offset the rise in aⁱ_Na expected from pump inhibition.

Acute effect of amiodarone on I_p. We next examined whether acute exposure of myocytes to amiodarone affected the Na⁺-K⁺ pump. For these experiments, myocytes were isolated from untreated rabbits and exposed to amiodarone in the tissue bath. After establishing the whole-cell configuration, we superfused myocytes with Ca⁺⁺-free Tyrode's solution that contained 2 mM BaCl₂ and either amiodarone in 1% ethanol or 1% ethanol alone. The cells were exposed to this superfusate for a minimum of 6 minutes before changing to a superfusate that was similar except that it contained 100 µM ouabain. The duration of exposure was adopted from previous studies on Na⁺ channels (Follmer et al., 1987) and Ca⁺⁺ channels (Nishimura et al., 1989) in cardiac myocytes. These studies indicated that steady state effects were achieved after 6-min exposure to amiodarone in vitro.

View larger version (9K): [in this window] [in a new window]   Fig. 3.   A, Effect of acute exposure to amiodarone on Ip, measured using [Na]pip of 10 mM (open bars) and 80 mM (solid bars). Numbers in parentheses indicate how many myocytes were studied in each group. Exposure of myocytes to amiodarone significantly reduced Ip when [Na]pip was 10 mM. 1% ethanol alone did not affect Ip. Amiodarone had no effect on Ip when [Na]pip was 80 mM. B, Effect of [K]o on Ip in myocytes exposed to amiodarone in 1% ethanol in vitro (, dashed lines) and in myocytes exposed to 1% ethanol alone (, solid line). Ip has been normalized to the current measured when [K]o is 5.6 mM. C, Normalized Ip-Vm relationships for 9 myocytes exposed to amiodarone in 1% ethanol () and for 8 myocytes exposed to 1% ethanol alone () in vitro.

Effect of acute amiodarone on the affinity of the Na⁺-K⁺ pump for extracellular K⁺. To examine the effect of acute amiodarone exposure on K⁺ affinity, we measured I_p at different levels of [K]_o. We used a [Na]_pip of 80 mM. After establishing the whole-cell configuration, we inactivated the Na⁺-K⁺ pump by superfusing myocytes with K⁺-free Tyrode's solution. This superfusate also contained amiodarone. Myocytes were then voltage-clamped at 40 mV, and after a minimum of 6-min exposure to the amiodarone-containing superfusate, the pump was reactivated by exposing myocytes to a superfusate that was similar except that it contained different [K]_o values. In contrast to experiments examining the effects of chronic amiodarone treatment, each myocyte was exposed in a random sequence to all of the following [K]_o values (in mM):1, 2, 3, 5.6 and 15. Results were normalized relative to 5.6 mM [K]_o. Each exposure to a [K]_o concentration was bracketed by return to the K⁺-free superfusate until I_p had returned to its base-line level. Control myocytes were maintained in the whole-cell configuration for the same duration as amiodarone-exposed myocytes before I_p was measured.

Effect of acute amiodarone on the I_p-V_m relationship. We next examined whether acute exposure of myocytes to amiodarone affects the I_p-V_m relationship. After establishing the whole-cell configuration, the superfusate was changed to Ca⁺⁺-free Tyrode's solution that contained either amiodarone in 1% ethanol or 1% ethanol alone. The myocytes were voltage-clamped at 40 mV; after a minimum exposure time of 6 min to this superfusate, we applied the voltage step protocol illustrated in figure 2A before and after exposure to 100 µM ouabain as outlined previously. Because in vitro exposure had no effect on I_p when [Na]_pip was 80 mM, these experiments were only performed at [Na]_pip of 10 mM.

	Discussion

Top Abstract Introduction Methods Results Discussion References

The major finding of our study is that both chronic and acute exposure to amiodarone decreased Na⁺-K⁺ pump activity in intact cardiac myocytes. After chronic treatment, pump activity was reduced at both low and high [Na]_pip. This pattern suggests a decrease in overall pump capacity. Acute exposure to amiodarone in vitro decreased pump activity only when [Na]_pip was near the physiological level for intracellular Na⁺ and had no significant effect at levels expected to saturate intracellular pump sites. This pattern is consistent with a decrease in the apparent affinity of the pump for intracellular Na⁺. A difference in effects between acute and chronic exposure is also well recognized for the clinical use of the drug (Ikeda et al., 1984) and for its effects on ion channel function (Varró et al., 1996). The [Na]_pip-dependent difference in effects between chronic and acute exposure suggests that amiodarone affects the pump via two different mechanisms. With chronic exposure, amiodarone might affect the pump via both mechanisms: one related to a decrease in the abundance of Na⁺-K⁺ pumps and one due to a direct effect of the drug in the membrane. However, amiodarone induced a similar degree of inhibition with chronic and acute exposure when [Na]_pip was 10 mM. This similarity might be explained by wash-out of amiodarone while cells from treated animals were maintained in amiodarone-free solutions for 2 to 10 hr before I_p was measured, with a loss of a direct effect of the drug bound to the membrane.

Two previous studies have examined the effect of chronic amiodarone treatment on the Na⁺-K⁺ pump. Prasada Rao et al. (1986) administered amiodarone parenterally to rats for 6 weeks. They found that amiodarone had no effect on activity of brain synaptosome Na⁺-K⁺/ATPase activity. To our knowledge, a study by Hensley et al. (1994) provides the only previous data on cardiac tissue. They examined the effects of 3- to 6-week parenteral amiodarone therapy on protein expression and mRNA levels for the alpha and beta subunits of the Na⁺-K⁺ pump in rat ventricular muscle. At 3 weeks, abundance of the alpha-1, alpha-2 and beta-1 subunits was significantly decreased. This was accompanied by a 30% to 33% decrease in serum T₃ and T₄ levels consistent with a systemic amiodarone-induced hypothyroid state. Because hypothyroidism reduces synthesis of Na⁺-K⁺ pump subunits in the heart (Horowitz et al., 1990, Kamitani et al., 1992), the decrease in subunit abundance at 3 weeks might be due to the effect of amiodarone on thyroid status. After 6-week therapy, T₃ and T₄ levels and alpha-1 and beta-1 subunit abundance had returned to control values, whereas the abundance of the alpha-2 subunit remained significantly depressed. Alpha-2 subunits do not appear to be quantitatively important in the rat, and Hensley et al. (1994) emphasized that the functional significance of altered alpha-2 expression was uncertain. The present study does not identify effects of amiodarone on specific subunits; however, it does indicate that chronic treatment has a functionally significant inhibitory effect on the sarcolemmal Na⁺-K⁺ pump.

Our study indicates that chronic treatment with amiodarone induces a decrease in I_p when the pump is maximally stimulated. Maximal I_p is highly correlated with the number of Na⁺-K⁺ pump units expressed in the plasmalemma of voltage-clamped oocytes (Jaunin et al., 1992). If the same applies to pump units in the sarcolemmal membrane, our results suggest that amiodarone induces a decrease in the number of functional Na⁺-K⁺ pump units. The amiodarone-induced decrease in pump function occurred in the absence of systemic hypothyroidism. However, this does not rule out the possibility that amiodarone exerted its effect on the pump via interference with thyroid hormone action at the intracellular level because amiodarone inhibits the binding of T₃ to its nuclear receptors (Drvota et al., 1994; Latham et al., 1987). The pattern of pump inhibition is similar to that reported by Doohan et al. (1995) in rabbits with experimentally induced hypothyroidism. It is also of interest to note that some of the electrophysiological effects of amiodarone, including prolongation of action potential duration, QT interval and sinus cycle length, are at least partly dependent on thyroid hormone activity and are greatly diminished or abolished by concurrent hypothyroidism (Polikar et al., 1986; Talajic et al., 1989).

Several previous studies have reported inhibition of purified cardiac Na⁺-K⁺/ATPase, the enzymatic equivalent of the Na⁺-K⁺ pump, during acute exposure to amiodarone in vitro (Broekhuysen et al., 1972; Dzimiri and Almotrefi., 1991). Although these studies suggest that acute amiodarone exposure may inhibit the pump, their significance is unclear for several reasons. Isolation of Na⁺-K⁺/ATPase from the sarcolemmal membrane necessarily involves removing the pump molecule from its native lipid environment. Because the lipid environment is an important regulator of pump activity (Cornelius, 1991) and may itself be influenced by amiodarone (Chatelain et al., 1985, 1986), the response of isolated Na⁺-K⁺/ATPase to amiodarone may not be representative of the behavior of the pump in the intact myocyte. In the studies by Broekhuysen et al. (1972) and Dzimiri and Almotrefi (1991), Na⁺-K⁺/ATPase was examined using saturating concentrations of Na⁺ and K⁺ to produce maximal pump activity. They therefore provide no information about the effect of amiodarone on the activity of the pump at physiological ligand concentrations. Finally, the concentrations of amiodarone necessary to induce significant ATPase inhibition were substantially higher than those encountered clinically.

In the only published study on intact cardiac muscle, Aomine (1989) indirectly determined the acute effects of amiodarone on pump activity in guinea pig papillary muscles using conventional microelectrode techniques. He found that exposure to 44 µM amiodarone decreased the postoverdrive hyperpolarization attributed to pump activity. Somewhat paradoxically, no effect was observed at the higher concentration of 440 µM.

Our in vitro study differs from previous studies in several aspects. We examined the effect of 0.3 to 1.2 µM amiodarone, a concentration range similar to that encountered clinically, and we examined the pump in its intact membrane while varying extracellular and intracellular ligand concentrations. In contrast to the studies on isolated ATPase, we found no effect on I_p measured under conditions expected to maximally stimulate the pump. However, pump inhibition could be demonstrated when we used a [Na]_pip near physiological levels for intracellular Na⁺. This is consistent with a decrease in the apparent affinity of the pump for intracellular Na⁺ without an influence on overall pump capacity.

Although the mechanism for modulation of the Na⁺ affinity of the pump cannot be determined from our study, one may speculate that the effect of amiodarone could be due to modification of the physicochemical properties of the lipid membrane. Amiodarone is known to insert into the hydrocarbon core of the lipid bilayer and decrease membrane lipid mobility and membrane fluidity (Chatelain et al., 1986). Changes in membrane fluidity, in turn, may affect pump activity by impairing lateral movement of the pump molecule and hence conformational changes during the pump cycle (see Cornelius, 1991, for review). An alternative explanation may be provided by the cationic amphiphilic nature of amiodarone. Cationic amphiphiles have been shown to decrease the apparent Na⁺ affinity of Na⁺-K⁺/ATPase reconstituted into liposomes (Cornelius, 1995).

Chronic treatment with amiodarone reduced I_p by 33% when [Na]_pip was 10 mM. This inhibitory effect is similar to that estimated for clinical use of digoxin (Rasmussen et al., 1990; Schmidt et al., 1991). This finding may have implications for our understanding of the effect of the drug when used clinically. It prolongs action potential duration and myocardial refractoriness, effects that are believed to be largely mediated via block of voltage-gated K⁺ channels (Balser et al., 1991). The Na⁺-K⁺ pump generates a hyperpolarizing membrane current that contributes to repolarization. This effect becomes more significant as intracellular Na⁺ increases, for example, during tachyarrhythmias (Gadsby, 1982). The inhibitory effect of amiodarone on the pump is therefore expected to contribute, at least in part, to the class III effects of the drug.