Electrophysiologic Mechanism Underlying Action Potential Prolongation by Sevoflurane in Rat Ventricular Myocytes

SEVOFLURANE has been reported to prolong QTc interval in healthy adults1^,2 and children3^,4 during anesthesia induction. Furthermore, sevoflurane has been reported to markedly prolong the QT intervals in children with congenital long QT syndrome.5^,6 In in vitro animal preparations using guinea pig ventricular myocardium7 and myocytes,8^,9 sevoflurane prolonged the action potential (AP) duration. This seemed to be caused by inhibition of the slowly activating delayed outward K⁺ current with minimal effect on the inward rectifier K⁺ current (I_KI). In contrast to this prolonging effect, shortening of AP duration by sevoflurane has also been observed in guinea pig myocardium,10 canine,11 and rat ventricular myocytes.12 

Cardiac AP duration is determined by a balance between inward and outward membrane currents.13^,14 In most species, the transient outward K⁺ current (I_to) is responsible for the initial phase of repolarization, and the L-type Ca²⁺ current (I_{Ca, L}) is the main inward current during the plateau phase. The delayed outward K⁺ current (I_K) activation initiates repolarization near the end of the ventricular AP plateau, and the I_KI plays an important role in generating the resting membrane potential and in modulating the final repolarization phase of the ventricular AP.14 The contribution of these currents varies between species and is responsible for the characteristic differences in AP shapes. For example, rat ventricular myocytes possess a prominent I_to, the major outward current of the repolarization phase, but little I_K, and have a short AP duration. In contrast, guinea pig ventricular cells lack I_to, but have a slowly activating I_k, resulting in a long-lasting AP.15 

In rat ventricular myocytes, Rithalia et al.12 observed a shortening effect of sevoflurane on AP duration, attributed to the reduced I_{Ca, L}, with little effect on I_to. In another recent study9 using cloned human cardiac K⁺ channels, sevoflurane inhibited Kv4.3 cardiac K⁺ channel currents, suggesting inhibition of I_to. Based on the similar electrophysiologic characteristics of I_to between human and rat ventricular myocytes,16 we speculated that the I_to in rat heart cells may be inhibited by sevoflurane. During preliminary experiments, we observed concentration-dependent prolongation of AP duration in rat ventricular myocytes, which contradicted the results of Rithalia et al.12 Therefore, to elucidate the mechanisms of AP prolongation, we examined the effect of sevoflurane on I_to and inward rectifier K⁺ current (I_KI), as well as I_{Ca, L}.

According to a protocol approved by the Yonsei University College of Medicine Animal Research Committee (Seoul, Korea), the heart was quickly excised from rats (Sprague-Dawley, weighing 250–300 g) after halothane anesthesia. The excised heart was retrogradely perfused by using a Langendorff perfusion system for 5 min at 37°C. The perfusion was at a rate of 7 ml/min with modified Tyrode solution containing 143 mm NaCl, 5.4 mm KCl, 1.8 mm CaCl₂, 0.5 mm MgCl₂, 5 mm HEPES, and 0.18 mm glucose, pH 7.4. The perfusate was then switched to a nominally Ca²⁺-free Tyrode solution for 5 min, followed by perfusion with the same solution to which collagenase (0.4 mg/ml, Worthington type II; Worthington Biochemical Corporation, Lakewood, NJ) and hyaluronidase (0.4 mg/ml, Sigma type II; Sigma-Aldrich Co., St. Louis, MO) had been added. After 10–12 min of enzymatic treatment, a final perfusion was performed for 5 min with a Kraftbrühe solution (10 mm taurine, 10 mm oxalic acid, 70 mm glutamic acid, 35 mm KCl, 10 mm H₂PO₄, 11 mm glucose, 0.5 mm EGTA, and 10 mm HEPES, pH 7.4). The ventricles were then cut off, minced with scissors, and agitated in a small beaker of a Kraftbrühe solution. The resulting slurry was filtered through a 200-μm nylon mesh. The isolated ventricular cells were stored in a Kraftbrühe solution for 1 h at room temperature (21°–22°C), then kept at 4°C, and used within a period of 8 h. Only the rod-shaped cells with apparent striations that remained quiescent in the solution containing 2 mm CaCl₂ were used for the experiments. All experiments were performed at room temperature.

Isolated myocytes were allowed to settle to the bottom of a recording chamber, which was mounted onto an inverted microscope where the bathing solutions could be exchanged. The chamber was continuously perfused at a constant rate (2 ml/min). Standard whole cell voltage clamp methods were used.17 To establish a stable baseline, an interval of 4–6 min was allowed after initiating the whole cell recording configuration. Voltage clamp measurements were performed by using an Axopatch 200B patch clamp amplifier (Axon Instruments Inc., Foster City, CA). Patch electrodes were prepared from a borosilicate glass model KIMAX-51 (American Scientific, Charlotte, NC), which have a typically 2- to 3-MΩ resistance when filled with an internal solution. After fabricating the pipettes with a two-stage micropipette puller, the pipette tips were heat-polished by using a microforge. Data acquisition was performed by using a pCLAMP system version 6.0 (Axon Instruments Inc.) coupled with a Pentium-III personal computer.

The APS were elicited in current-clamp mode by 5-ms, 800-pA current injections at a frequency of 1 Hz.

To examine I_to, from a holding potential of −80 mV, the cells were depolarized to −40 mV for 50 ms to inactivate the Na⁺ current and then depolarized to test potentials of up to +60 mV in 10-mV increments for 300 ms. To obtain more information about the possible mechanism of sevoflurane-induced voltage blockade of K⁺ currents, the voltage dependence of steady state inactivation was determined by using a standard double-pulse protocol. The membrane potential was initially clamped at −80 mV and then stepped to different potentials ranging from −100 to 0 mV in 10-mV increments for 500 ms followed by a 200-ms test pulse to +80 mV. The voltage clamp protocol was repeated every 2 s. The steady state inactivation data were fitted by a Boltzmann distribution of the following equation: I/I_max = I/{1 + exp [(V − V_1/2)]/κ}, where I_max is the maximal current, V_1/2 is the membrane potential producing 50% inactivation, and κ is the slope factor.

To test whether the accelerated inactivation of I_to in the presence of sevoflurane was associated with time-dependent block of the open channel, the magnitude of current inhibition at various times after the initiation of the depolarizing pulse was determined at the membrane potential of +60 mV. The outward current in the presence of each concentration of sevoflurane, expressed as a proportion of the outward current observed in the control, was plotted as a function of time after the start of depolarization. A hyperbola function [y = B_max (x/(k_d + x))] was used to fit the rate of block of I_to at each concentration of sevoflurane. B_max is the maximum block at drug concentration; K_d is the dissociation constant.

The sustained outward current (I_sus) was recorded with the same voltage protocol after addition of 5 mm 4-aminopyridine, which preferentially blocks I_to, in modified Tyrode solution. The outward currents activated by depolarizing voltage steps in rat ventricular myocytes consist of a rapidly inactivating component, I_to, and a noninactivating, sustained component, I_sus.18^,19 The sustained outward currents, comprising I_K and a small time-independent outward current,18 remain in the presence of 4-aminopyridine. Whereas I_to underlies the initial, rapid repolarization phase of the AP, I_sus is responsible for the slower phase of AP repolarization back to the resting membrane potential in adult rat ventricular myocytes.20 

The effect on the I_KI was verified by measuring the I_KI by step depolarizations from −130 to 0 mV from a holding potential of −40 mV in 10-mV increments, using a 200-ms pulse applied at 5-s intervals.

The voltage-dependent I_{Ca, L} was evoked by step depolarizations that lasted 200 ms from a holding potential of −40 mV to +10 mV in one step at a frequency of 0.1 Hz.

After the baseline measurements, myocytes were exposed to 1.7% or 3.4% sevoflurane for 2–3 min, and recovery responses were measured after washes for 2–3 min to remove the drugs. A 2-min application of sevoflurane was sufficient to produce a stable and consistent effect in pilot experiments.

Before establishing the whole cell recording configuration, modified Tyrode solution, containing 140 mm NaCl, 5.4 mm KCl, 1 mm CaCl₂, 1 mm MgCl₂, 5 mm HEPES, and 10 mm glucose, adjusted to pH 7.4 with 1N NaOH, was used as an external bathing solution. For K⁺ current measurements, a patch pipette solution was used, containing 20 mm KCl, 110 mm K-aspartate, 10 mm EGTA, 10 mm HEPES, 1 mm MgCl₂, 5 mm K₂ATP, 1 mm CaCl₂, and 10 mm NaCl, adjusted to pH 7.2 with 3N KOH. To measure I_sus, 5 mm 4-aminopyridine was added to the modified Tyrode solution. As the pH of the modified Tyrode solution containing 4-aminopyridine revealed 9.04 ± 0.03 (n = 2), the pH was corrected to 7.4 with HCl before the experiment. To eliminate any confounding Ca²⁺ current, 0.2 mm CdCl₂ was added to the external solution after establishing the whole cell voltage clamp.

The inward Ca²⁺ current was measured by using a patch pipette solution containing 30 mm CsCl, 100 mm aspartic acid, 100 mm CsOH, 10 mm BAPTA, 10 mm HEPES, 10 mm phosphocreatine, 1 mm Na₂GTP, 5 mm Na₂ATP, 10 mm glucose, and 2 mm MgCl₂, adjusted to pH 7.2 with 1 m CsOH. Once whole cell recording was achieved, the bathing solution was exchanged to 140 mm NaCl, 5.4 mm CsCl, 2 mm CaCl₂, 1 mm MgCl₂, and 10 mm HEPES, adjusted to pH 7.4 with 1 m CsOH. Sevoflurane was purchased from the Abbott Korea Ltd. (Seoul, Korea), and all other chemicals were purchased from Sigma-Aldrich Co. (St. Louis, MO).

Before perfusion, sevoflurane was equilibrated in solution in one reservoir by passing 100% O₂ (flow rate: 0.2 l/min) for 15 min through a sevoflurane vaporizer (Sevotec 3; Ohmeda, West Yorkshire, United Kingdom). The end-tidal concentrations of sevoflurane were monitored by using a calibrated gas analyzer (Capnomac; Datex, Helsinki, Finland). As determined by gas chromatographic measurement, sevoflurane concentrations in the room-temperature Tyrode superfusate equilibrated for 15 min were 0.35 mm (n = 4) and 0.7 mm (n = 4) for 1.7% and 3.4% sevoflurane, respectively. With the Tyrode solution/gas partition coefficient of sevoflurane of 0.40 at 22°C,21 0.35 and 0.7 mm sevoflurane correspond to gas phase concentrations of 2.12% and 4.24%, respectively.

Repeated measures of analysis of variance followed by the Student-Newman-Keuls test were applied to test for significant differences among control, drug application, and washout. An unpaired t test was used to compare the differences in currents of I_to, I_KI, and I_{Ca, L} between 1.7% and 3.4% sevoflurane. All values are expressed as mean ± SEM. A P value less than 0.05 was considered significant.

Figure 1 shows concentration-dependent prolongation of AP duration observed in a rat ventricular myocyte. Sevoflurane at 0.35 mm (n = 6) prolonged the APD₅₀ and APD₉₀ to 129 ± 10% (P < 0.05) and 115 ± 2% (P < 0.05) of the control, respectively. Sevoflurane at 0.7 mm (n = 7) also prolonged the APD₅₀ andAPD₉₀ to 133 ± 3% (P < 0.05) and 139 ± 6% (P < 0.05) of the control, respectively. The AP amplitude and resting membrane potential remained unaltered at either concentration (table 1). The AP duration was completely restored to baseline after washout.

At a membrane potential of +60 mV, 0.35 mm sevoflurane reduced the control peak currents of I_to (3.82 ± 0.54 nA) by 18 ± 2% (n = 7; P < 0.05), and the plateau currents measured at the end of depolarization (1.51 ± 0.12 nA) by 10 ± 3% (n = 7; P < 0.05). Sevoflurane, 0.7 mm, reduced peak currents of I_to (3.64 ± 0.30 nA) by 24 ± 2% (n = 11; P < 0.05) and plateau currents (1.52 ± 0.19 nA) by 10 ± 2% (n = 11; P < 0.05) (figs. 2A–C). Complete recovery of peak and plateau currents was observed after washout of either concentration (figs. 2A and B).

Under control conditions, the voltage required for half-inactivation (V_1/2) and slope factor (κ) for the I_to were −31.53 ± 0.68 mV and −6.23 ± 0.60 mV, respectively (n = 7). Sevoflurane, 0.7 mm, did not shift the steady state inactivation curve (V_1/2 = −35.04 ± 0.84 mV, κ = −7.40 ± 0.74, n = 7; not significant) (fig. 3).

Because application of 0.7 mm sevoflurane in figure 2A indicated that sevoflurane accelerated the decay of current during the pulse, we evaluated the effect of sevoflurane on the kinetics of inactivation of I_to. Individual current records evoked by a test pulse to +60 mV were fitted with a double exponential function. Figures 4A and B show quality of fit of the double exponential function to the inactivation phase of I_to under control conditions and in the presence of 0.7 mm sevoflurane, respectively. The control τ₁ (47 ± 3 ms) was reduced to 27 ± 3 ms by 0.7 mm sevoflurane (n = 7; P < 0.05). Inactivation kinetics returned to baseline values after washout (47 ± 4 ms) (fig. 4C).

Figure 5 shows the inhibition of I_to by 0.35 and 0.7 mm sevoflurane during depolarization. The inhibition increased in a hyperbolic manner during depolarization. Both the magnitude of the maximum inhibition and the rate of development of the maximum inhibition seemed to be concentration dependent. The B_max values of 0.35 and 0.7 mm sevoflurane were 0.6 (n = 7) and 0.82 (n = 8), respectively. The K_d values of 0.35 and 0.7 mm sevoflurane were 19.66 ms (n = 7) and 15.84 ms (n = 8), respectively.

Representative tracings of isolated I_sus by application of 5 mm 4-aminopyridine are illustrated in figures 6A and B. Sevoflurane, 0.7 mm, had no effect on the I_sus (n = 6) (fig. 6C). Before sevoflurane exposure, the baseline value of I_sus at the end of the plateau during test potential of +60 mV was 1.31 ± 0.12 nA (n = 6).

At membrane potential ranges from −130 to 0 mV, 0.7 mm sevoflurane did not alter the I_KI (n = 7; not significant) (fig. 7B). The I_KI was measured at the end of the pulse duration. Before sevoflurane exposure, the baseline value during test potential of −130 mV was −2.92 ± 0.25 nA.

At a membrane potential of +10 mV, 0.35 and 0.7 mm sevoflurane reduced the I_{Ca, L} by 28 ± 3% (n = 8; P < 0.05) and 33 ± 1% (n = 7; P < 0.05), respectively (fig. 8B). The effect of sevoflurane on I_{Ca, L} was completely reversible after washout. The baseline values before exposure to 0.35 and 0.7 mm sevoflurane were 0.75 ± 0.09 and 0.81 ± 0.2 nA, respectively.

This study shows that clinically relevant concentrations of sevoflurane prolong the AP duration and significantly inhibit the I_to and I_{Ca, L} in isolated rat ventricular myocytes. Although sevoflurane did not shift the steady state inactivation curve, it significantly accelerated inactivation of I_to.

Prolongation of the ventricular AP duration by sevoflurane has been observed in guinea pigs,7,^–9 whereas other studies using canine ventricular cells11 or guinea pig papillary muscles10 have reported shortening of the AP duration. In a study using rat ventricular cells, sevoflurane caused modest but significant shortening of AP duration,12 a curious difference from our study despite use of the same animal species.

In rat ventricular myocytes, cardiac voltage-activated K⁺ current consists of I_to and I_K, sensitive to 4-aminopyridine and tetraethylammonium, respectively.20 In frog and guinea pig ventricular myocytes, the dominant voltage-activated outward K⁺ current is I_K.22^,23 However, in rat, dog, rabbit, and human myocytes, I_to is prominent and plays a significant role in the early repolarization phase of the AP.24,^–27 In several cardiac tissues, two types of I_to have been identified; one is voltage- and Ca²⁺-independent, and the other is Ca²⁺-dependent.28^,29 In rat ventricular myocytes, only a Ca²⁺-independent I_to has been identified.30 The Ca²⁺-independent I_to, carried predominantly by K⁺ ions, has been suggested to be a major determinant of the cardiac AP duration because of its large size and pronounced frequency dependence.29 

Rithalia et al.,12 in their study using rat ventricular subendocardial and subepicardial myocytes, observed no changes of I_to by application of 0.6 mm sevoflurane (30°C). In contrast, we found that sevoflurane caused a modest but significant depression of approximately 24% by 0.7 mm sevoflurane (22°C). In a recent study using cloned human cardiac K⁺ channels, 3 mm sevoflurane inhibited the Kv4.3 cardiac K⁺ channel currents by approximately 28% (35°C), suggesting inhibition of I_to.9 Kv4.3 channel has been reported to underlie a significant fraction of I_to in the heart of several species, including rat, canine, and human.31 Considering lower solubility at higher temperatures, the 0.6 mm sevoflurane at 30°C is estimated to be around 0.7 mm at 22°C.7 Despite application of a similar concentration in the same animal species, it remains unclear why a disparity exists between the result of Rithalia et al. and ours. In the above study using cloned human cardiac K⁺ channels, 0.43 mm sevoflurane at 37°C will be approximately 0.7 mm at 22°C.7 Although a 0.43 mm concentration of sevoflurane caused modest depression of the Kv4.3 cardiac K⁺ channel currents, significant acceleration of the rate of decay was shown at this concentration,9 indicating a possible significant open channel inhibition in clinically relevant concentrations of sevoflurane. Considering the similar electrophysiologic characteristics of I_to between human and rat ventricular myocytes,16 reduction of I_to in our results seems to correspond to that of Kang et al.9 The prolongation of AP duration as a result of inhibition of I_to has been demonstrated with tedisamil, a blocker of I_to and I_K in isolated rat ventricular myocytes,32 and was similar to that reported in isolated rat papillary muscles.33 Our findings of approximately 24% reduction of I_to appear as a small fractional reduction. However, considering the large current density in rat ventricular myocytes, 19.9 ± 2.8 pA/pF at the membrane potential of +60 mV,16 the effect on the plateau phase will be greater, resulting in AP prolongation.

The current study shows that sevoflurane decreased the peak I_to and also accelerated the rate of decay. The acceleration of the rate of decay could be explained by an anesthetic-induced acceleration of the normal conversion of open channels to an inactivated state with sustained depolarization. Sevoflurane could preferentially inhibit the open state of the channel. These findings are similar to that previously reported in rat ventricular I_to for quinidine (class Ia antiarrhythmic agent),34 clofilium,35 tedisamil (class III antiarrhythmic agent),32 bupivacaine,36 and imipramine in rabbit atrial I_to.37 

In the current study, whereas there was no inhibition of I_to at the onset of the depolarizing pulses, we observed the inhibition of I_to upon continued depolarization. These results may indicate that sevoflurane does not bind to the resting state of the channel in an inhibitory fashion, but rather inhibits the open channel in a hyperbolic manner during continued depolarization. The lack of any effect on the voltage dependence of the steady state inactivation suggests that sevoflurane does not bind to the inactivated state of the channel.

The sustained outward currents, comprising I_K and a small time-independent outward current,18 remain in the presence of 4-aminopyridine and tetraethylammonium. I_sus contributes to the overall repolarization process in rat ventricular myocytes. Whereas 0.7 mm sevoflurane has been reported to significantly depress the I_K in guinea pig ventricular cells, approximately 60%,7 I_sus was not affected by sevoflurane. Although suppression of I_K by sevoflurane in guinea pig ventricular myocytes is mainly responsible for the AP prolongation, I_sus does not likely contribute to the prolongation of AP duration by sevoflurane in rat ventricular myocytes. Presumably, little I_sus change by sevoflurane may be attributed to marked variations among cells in rat ventricle in the relative amplitudes of I_to and I_K (I_to/I_K),21 significant inhibition of I_K by 4-aminopyridine at concentrations above 1 mm,18 and/or species differences.

The I_KI is the primary current responsible for maintaining a stable cardiac resting membrane potential near the K⁺ equilibrium potential. Inhibition of I_KI can result in diastolic depolarization, which can increase cardiac excitability38 and lead to dysrhythmias and abnormal automaticity.39 Both sevoflurane and isoflurane have been reported to cause significant depression of inward component and enhancement of outward component of I_KI in guinea pig ventricular myocytes.40^,41 However, the actual differences from baseline values seemed to be modest, suggesting minimal effect on I_KI by these anesthetics. Modest effects on inward and outward components of I_KI by sevoflurane have been reported in guinea pig ventricular myocytes,7 and our results in rat ventricular myocytes also showed no depression of the inward component of I_KI by sevoflurane. The lack of change in the resting membrane potential after application of 0.7 mm sevoflurane in our results may also reflect the lack of effect on I_KI.

In the current study on steady state inactivation of I_to, approximately 80% of the channels seemed to be available for activation at −40 mV. This indicates that the outward component of I_KI above −40 mV may include I_to and thus can influence the outward component of I_KI in this preparation.

Our whole cell voltage clamp studies revealed a reduction of peak I_{Ca, L} despite the prolongation of AP duration. Suppression of I_{Ca, L} by sevoflurane has been reported in various animal preparations.7^,9^,11^,12 Inhibition of I_{Ca, L} can lead to a shortening of AP duration, however, the reduction of I_to would seem to have a greater effect, resulting in lengthening of the AP duration.

In human ventricular myocytes, although the current density of I_to has been reported to be two or three times smaller than that of the rat myocytes (8.2 ± 0.7 pA/pF at +60 mV), it is, nonetheless, a major outward current in human myocytes.16 Small changes of I_to during the early phase of the AP can profoundly affect the activation of other plateau currents, such as Ca²⁺ and the delayed outward K⁺ currents, influencing the AP duration and Ca²⁺ influx.27 

In conclusion, prolongation of AP duration, induced by clinically relevant concentrations of sevoflurane, appears due to the suppression of I_to in rat ventricular myocytes. Considering that the voltage and time dependence of I_to in human ventricular myocytes are similar to those found in rat heart cells16 and prolongation of AP duration in failing human heart seems to be prominently caused by a reduction in the level of Kv4.3 mRNA, resulting in down-regulation of Ca²⁺-independent I_to,42 suppression of I_to by sevoflurane may partly account for the clinical observation of QTc prolongation in humans. In addition, considering the presence of I_ks in healthy human ventricle,43^,44 inhibition of I_ks by sevoflurane8 may also contribute to QTc prolongation.