Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Nonsedating H₁ receptor antagonists are a group of structurally diverse compounds that have been prescribed effectively in treatment of allergies and upper respiratory inflammation. Recently, however, adverse cardiac arrhythmogenic effects of some members of this group, especially terfenadine and astemizole, have been reported (Honig et al., 1993; MacConnell and Stammers, 1991). These effects have surfaced primarily in patients taking either excessive dosages of these drugs or using concomitant drugs that are known to inhibit cytochrome P-450 3A activity or in patients with impaired liver function (Craft, 1985; Monahan et al., 1990; Woosley, 1996). A number of reports suggest that the arrhythmogenic effect of terfenadine results from the prolongation of the QT interval secondary to suppression of the K⁺ channel (Berul and Morad, 1995; Crum et al., 1995; Woosley et al., 1993; Rampe et al., 1994; Salata et al., 1995). Because the individual members of this class of drugs have varying effects on QT interval in experimental animals (Honig et al., 1992, 1993) and on K⁺ channels of isolated ventricular myocytes (for review, see Woosley, 1996), it is critical that a detailed K⁺ channel-modulating profile of any novel H₁-R antagonist be determined to evaluate possible adverse effects of the drug. In this respect, H₁-R antagonists appear to fall into two major categories: those (e.g., terfenadine and astemizole) that suppress cardiac K⁺ channels (Berul and Morad, 1995; Rampe et al., 1994; Salata et al., 1995) as they prolong QT interval and those (e.g., loratadine and cetirizine) that have little or no effect on various members of K⁺ channel family (Ducic et al., 1997) and produce no significant effect on QT interval (Hey et al., 1996a). Ebastine, with a structure similar to terfenadine and equivalent antihistamine potency, appears to show significant QT prolongation in experimental models (Hey et al., 1996b). In this report, we examined the possible effects of ebastine on five members of the K⁺ channel family known to be expressed in human heart. These channels include the rat I_to (equivalent to human Kv4.2; Honore et al., 1994) and I_ped (the noninactivating component of I_to, equivalent to human Kv1.5 clone; Tamkun et al., 1991) and two members of the delayed rectifier K⁺ channel family, I_Kr and I_Ks, in guinea pig ventricular myocytes. I_Kr, in addition, was examined by expressing the human ether-a-go-go (HERG) gene associated with chromosome 7 (Sanguinetti et al., 1995) in Xenopus laevis oocytes. The I_K1 was examined in both rat and guinea pig myocytes. Such a comprehensive evaluation of possible effects of ebastine on K⁺ channels was undertaken not only because K⁺ channels have significant structural differences and are heterologously expressed in different parts of the heart but also because the prolongation of the different phases of the action potential (resulting from suppression of a particular K⁺ channel) is likely to contribute differentially to initiation of cardiac arrhythmogenesis. Our data suggest that ebastine had significant suppressive effects on the delayed rectifier family of K⁺ channels (I_Kr, I_ped and I_Ks) but was less effective in blocking I_to and I_K1. The suppressive effect of ebastine on K⁺ channels is somewhat weaker than that of terfenadine but much stronger compared with loratadine.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Cell isolation. Ventricular myocytes were obtained from male guinea pigs (200-300 g) and Wistar rats (100- 200 g) using a previously described rapid enzymatic isolation procedure (Mitra and Morad, 1985). Myocytes were dispersed and allowed to settle for 1 hr at room temperature (22-24°C) before their use. Animal experimentation was performed in accordance with the guidelines of Georgetown University Animal Care and Use Committee and the American Heart Association position statement on research animal use.

Solutions and technique. The whole-cell patch-clamp technique (Hamill et al., 1981) was used to evaluate individual ionic currents using a Dagan model 8900 patch-clamp amplifier (Dagan Corp., Minneapolis, MN) connected to an IBM computer with pCLAMP 5.5.1 software (Axon Instruments, Foster City, CA). Heat-polished borosilicate glass pipettes (World Precision Instruments, Sarasota, FL) with a tip resistance of 1.0 to 2.5 M were used to establish gigaohm seal and continuity with the intracellular medium. The dialyzing internal pipette solution contained (in mM) 135 KCl, 10 NaCl, 5 to 10 HEPES and 5 Mg-ATP, titrated with KOH to pH 7.20. Internal solutions were buffered using a combination of 5 mM EGTA and 1.9 mM Ca²⁺ to have 60 nM free Ca²⁺ activity according to the SPECS program (Fabiato, 1988). The control external perfusate was a modified Tyrode's solution and contained (in mM) 137 NaCl, 5.4 KCl, 1.0 MgCl₂, 2.0 Ca₂Cl, 10 HEPES and 10 glucose, titrated with NaOH to pH 7.4.

Expression of HERG channel. HERG cRNA for injection into oocytes was prepared from the HERG cDNA in the pSP64 plasmid (Sanguinetti et al., 1995; a gift from Dr. M. Keating, University of Utah, Salt Lake City, UT) using mCAP RNA capping kit (Stratagene) following linearization of the expression construct with EcoRI.

Drugs. Ebastine, terfenadine and loratadine were obtained from Schering-Plough Research Institute as base powders and dissolved in 100% DMSO (Sigma Chemical Co.). The drugs were stored in DMSO in 10 mM concentrations at 20°C. We believe that even when stock solutions were perfectly clear, the dilution of stock solution in Tyrode's might significantly decrease the solubility of the drug and therefore its effective concentration. In the experimental solutions, varying concentrations of drugs ranging from 0.03 to 3 µM were added to freshly prepared control Tyrode's solutions. In all experiments, the control solution always contained the concentration of DMSO equivalent to that required to dissolve the highest concentration of drug (e.g., 0.03% of DMSO at 3 µM of the drug). A multibarreled concentration clamp system (Vibraspec, Inc., Bear Island, ME) was used to exchange rapidly (~50 msec) the extracellular solution used to bathe the experimental cell and apply the desired drug concentrations in a concentration-clamp manner.

Ionic channels. Because the cardiac K⁺ channels that we examined (I_Kr, I_Ks, I_to, I_ped and I_K1) are differentially voltage-gated, previously established specific voltage-clamp protocols were used to activate them.

Statistics. The mean percent inhibition (percent block) and, in some experiments, the amount of recovery of the current on drug washout (percent recovery) were tabulated. Each cell served as its own control with data acquisition before, during and after drug application. Statistical analysis also included the two-tailed Student's t test for paired samples.

Tests and precautions. Even though it was possible to dissolve ebastine in 100% DMSO at 10 mM concentrations, there was little guarantee that after the dilution of the stocks into salt solution, especially at micromolar range, a significant amount of drug had not fallen out of solution. On occasion, we did find that ebastine fell out of solution, even in DMSO, on cooling the DMSO stock. The uncertainty of knowing exactly what concentration of drug the myocyte is exposed to in different experiments may provide for some of the divergence in the effect of the drug on K⁺ channels. It would be prudent, therefore, to assume that the effects of ebastine may be larger than that reported here. Furthermore, the high lipophilicity of such agents may allow cellular accumulation of these compounds, making their washout nearly impossible. Nevertheless, a fairly reliable dose response could be constructed on step increases of drug concentrations.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Modulation of I_Ks by Ebastine

I_Ks is a slowly activating member of the delayed rectifier family contributing a significant portion of the outward current in guinea pig ventricular myocytes. In the analysis below, we quantified I_Ks as the time-dependent current activating within 2-sec depolarizing pulses positive to 0 mV (final minus initial current).

Figure 1 illustrates the suppressive effects of 0.3 and 3.0 µM ebastine on I_Ks. The superimposed current tracings of figure 1A represent the time course of activation of I_Ks in control and ebastine-containing solutions. Ebastine suppressed the magnitude of I_Ks without noticeably affecting its kinetics of activation. From 100 to 150 sec were often required for the steady-state effects of ebastine to develop. The suppressive effect of ebastine on I_Ks was consistently observed at all membrane potentials tested (0 to +80 mV, fig. 1B). The drug showed little or no voltage dependence at either dose. Figure 1B shows the consistency of the ebastine effect at 0.3 and 3 µM concentrations on 5 to 9 cells from different guinea pig hearts. Note that I_Ks is uniformly suppressed at all potentials tested independent of membrane potential.

View larger version (26K): [in this window] [in a new window]   Fig. 1.   The effect of ebastine on IKs in guinea pig ventricular myocytes. A, Voltage pulse protocol and representative original currents obtained before and after exposure of the myocyte to 0.3 µM and 3 µM ebastine; only part of the currents corresponding to the voltage depicted by solid lines that activated IKs are shown. B, Averaged current density-voltage relations of control IKs () and IKs in the presence of 0.3 µM () and 3 µM () ebastine. Vertical bars illustrate S.E.M.; n equals the number of myocytes tested (P < .05 for ebastine groups vs. control at +40 mV). Inset, voltage pulse protocol and original current record with demonstration of how IKs was measured; IKs values were divided by cell capacitance to obtain current density (pA/pF). C, A dose-response relation for the effect of ebastine on IKs measured at +40 mV; symbols represent mean experimental values of the current in the presence of drug normalized to the control current in nine myocytes. Vertical bars illustrate S.E.M. Smooth curve represents the best fit of experimental points with Langmuir's isotherm: I/(Is + Ir) = A Kd/(Kd + [M]) + 1A, where A = Is/(Is + Ir), Is is drug-sensitive component of the current, Ir is drug-resistant component of the current and [M] is drug concentration (parameters of the fit: Kd = 0.8 µM and A = 0.6).

Dose-response experiments for ebastine were carried out at +40 mV in stable cardiac myocytes. Because the suppressive effect of the drug could not be completely reversed within 5 to 7 min, drug concentrations were given in progressively larger doses after steady-state effect at each dose was achieved. Unless multiple doses in a particular cell could be applied, the results were not included in the dose-response relation illustrated in figure 1C. The data were fit with Langmuir's isotherm (see legend to fig. 1C), giving a K_d value of 0.8 µM and the maximal block of 60% (i.e., proportion of drug-sensitive current, A = 0.6).

Effect of Ebastine on I_Kr in Guinea Pig Myocytes and Cloned HERG Channel Expressed in X. laevis Oocytes

Two different experimental approaches were chosen to quantify the effects of ebastine on I_Kr. In one set of experiments, we measured I_Kr in guinea pig myocytes as the time-dependent tail currents that activated on repolarizing the membrane from conditioning positive potential (fig. 2A, top). In another set of experiments, we expressed the HERG gene in X. laevis oocytes (Sanguinetti et al., 1995; Trudeau et al., 1995) and compared the expressed I_Kr sensitivity to ebastine and terfenadine.

View larger version (20K): [in this window] [in a new window]   Fig. 2.   The effect of ebastine on IKr current in guinea pig cardiomyocytes. A, Representative original currents obtained in a guinea pig cardiomyocyte in control solution and after exposure to 1 µM ebastine. Top, voltage protocol used to elicit the currents. IKr was measured as a time-dependent component of the tail current activated in response to membrane repolarization to different voltages from +60 mV. B, I-V relations of control IKr () and IKr in the presence of 1 µM ebastine (). C, A dose-response relationship for the effect of ebastine on Ikr. Symbols are mean experimental values of the IKr tail current in response to repolarization from +60 to 30 mV measured in the presence of drug and normalized to the control IKr tail current obtained from five myocytes. Smooth curve represents the best fit of experimental points with Langmuir's isotherm (see legend to Fig. 1C) with Kd = 0.14 µM and A = 0.74.

Native I_Kr in guinea pig myocytes. The 1-sec conditioning depolarizing pulses to +60 mV followed by test pulses to potentials between 20 and 50 mV were applied to activate I_Kr (fig. 2A, inset). This pulse protocol was used because I_Kr inactivates at positive potentials but recovers from inactivation rapidly on application of repolarizing pulses. The superimposed tracings of I_Kr at different potentials in the presence and absence of 1 µM ebastine are illustrated in figure 2A. Figure 2B quantifies the voltage dependence of I_Kr, showing its characteristic inwardly rectifying properties as described previously (Sanguinetti and Jurkiewicz, 1990; see also figs. 3 and 4). Ebastine at 3 µM concentrations uniformly suppressed I_Kr at all potentials. The dose dependence of ebastine effect was quantified at 30 mV (fig. 2C). The data points were fit with Langmuir's isotherm (see legend to fig. 1C), providing effective K_d value of the block of I_Kr by ebastine of 0.14 µM and suggesting maximal block of 74% (proportion of drug-sensitive current, A = 0.74).

View larger version (28K): [in this window] [in a new window]   Fig. 3.   Effects of ebastine and terfenadine on IKr induced in X. laevis oocytes by HERG cRNA. A, Representative original records of IKr obtained at indicated membrane potentials from the same HERG-expressing oocyte in control conditions, in the presence of 1 µM ebastine, after washout of ebastine and in presence of 1 µM terfenadine. Top, a step voltage-clamp protocol used to elicit the currents. B, I-V relations of IKr measured at the end of single-step depolarizations to different potentials in control (), in the presence of 1 µM ebastine (), after washout of ebastine () and in the presence of 1 µM terfenadine (). C, I-V plots of the tail currents activated in response to repolarization against depolarizing voltage in control (), in the presence of 1 µM ebastine (), after washout of ebastine () and in the presence of 1 µM terfenadine ().

View larger version (19K): [in this window] [in a new window]   Fig. 4.   Effects of ebastine and terfenadine on IKr induced in X. laevis oocytes by HERG cRNA. A, Representative original records of IKr obtained at indicated membrane potentials from the same HERG-expressing oocyte in control conditions, in the presence of 1 µM ebastine, after washout of ebastine and in presence of 1 µM terfenadine. Top, a voltage-clamp protocol consisting of the conditioning pulse to +40 mV followed by test repolarizations to different voltages used to elicit the currents. B, I-V plots of IKr activated in response to repolarizations vs. repolarizing voltage in control (circles), in the presence of 1 µM ebastine (squares), after washout of ebastine (diamonds) and in the presence of 1 µM terfenadine (triangles).

I_Kr expressed in X. laevis oocytes. In X. laevis oocytes, two voltage-clamp protocols were used to measure I_Kr induced by expression of HERG cRNA. During the first protocol, membrane potential was stepped to different voltages from a holding potential of 80 mV (fig. 3A, top). In the second protocol, test pulses to membrane potentials ranging between 120 to +20 mV were preceded by a 1.2-sec conditioning depolarization to +40 mV (fig. 4A, top). The unusual voltage dependence of I_Kr arises from fast inactivation of the channel at positive voltages, and the rapid removal of inactivation at negative holding potentials, resulting in larger tail currents compared with the currents activated during depolarizing pulses (fig. 3A) (Sanguinetti et al., 1995; Smith et al., 1996). The voltage dependence of I_Kr is therefore determined by a dual process of slow activation and fast inactivation, resulting in reduced current at positive potentials (figs. 3B and 4B).

View larger version (11K): [in this window] [in a new window]   Fig. 5.   Dose-response relationships for the effect of ebastine and terfenadine on IKr expressed in X. laevis oocytes by means of HERG cRNA. Symbols are mean experimental values of the IKr tail current in response to repolarization from +30 to 80 mV in the presence of drug normalized to the control IKr tail current obtained from four to six oocytes. Vertical bars illustrate S.E.M. Smooth curves represent the best fit of experimental points with Langmuir's isotherm (see legend to Fig. 1C) (parameters of the fit for ebastine, Kd = 0.3 µM, A = 0.46; for terfenadine, Kd = 0.4 µM, A = 0.8).

Effects of Ebastine on I_to

The transient component of I_to was studied by a pulse protocol illustrated in figure 6A (inset). I_to currents recorded from holding potential of 90 mV at different depolarizations are superimposed in the presence and absence of ebastine (fig. 6A). Ebastine only minimally suppressed I_to. Ebastine, at 3 µM concentrations, suppressed I_to by only 2% to 8% in a total of 9 rat ventricular myocytes (fig. 6B). Ebastine at such concentrations appeared also to enhance the kinetics of inactivation of I_to (fig. 6A), but this effect may result in part from strong suppression of the maintained component of K⁺ current (I_ped) (fig. 7). We often measured a 5-mV negative shift of the steady-state availability (inactivation) of I_to in some myocytes (data not shown). Figure 6C illustrates the dose-dependent suppression of I_to (measured at +75 mV) by ebastine in nine ventricular myocytes. The data points were fit with Langmuir's isotherm (see legend to fig. 1C), suggesting an effective K_d value of 1.1 µM and maximal block of only 10% (proportion of drug-sensitive current, A = 0.1). Thus, ebastine had only a minor effect on I_to. compared with I_Kr and I_Ks.

View larger version (25K): [in this window] [in a new window]   Fig. 6.   The effect of ebastine on Ito current in rat cardiomyocytes. A, Superimposed Ito traces obtained in control conditions and in the presence of 3 µM ebastine in response to voltage-clamp protocol (top). Dashed line, zero current. B, Averaged current density-voltage relations of the control Ito () and Ito in the presence of 3 µM ebastine (). Symbols represent mean values from nine myocytes. Vertical bars illustrate S.E.M. C, A dose-response relation for the effect of ebastine on Ito measured at +75 mV. Symbols represent mean experimental values of the current in the presence of drug normalized to the control current in nine myocytes. Vertical bars illustrate S.E.M. Smooth curve represents the best fit of experimental points with Langmuir's isotherm (see legend to Fig. 1C) with Kd = 1.1 µM and A = 0.1.

View larger version (27K): [in this window] [in a new window]   Fig. 7.   The effect of ebastine on maintained component (Iped) of Ito in rat ventricular myocytes. A, Superimposed traces of control Iped and Iped in the presence of 3 µM ebastine. Iped was activated from a holding potential of 90 mV by test pulses ranging from 10 to +70 mV, which were preceded by a conditioning prepulse of 10 mV to inactivate transient Ito (top). Dashed line, zero current. B, Averaged current density-voltage relations of the control Iped () and Iped in the presence of 3 µM ebastine (). Symbols represent mean values from nine myocytes. Vertical bars illustrate S.E.M. (P < .05 for ebastine group vs. control at +70 mV). C, A dose-response relation for the effect of ebastine on Iped measured at +70 mV. Symbols represent mean experimental values of the current in the presence of drug normalized to the control current in nine myocytes. Vertical bars illustrate S.E.M. Smooth curve represents the best fit of experimental points with Langmuir's isotherm (see legend to Fig. 1C) with Kd = 1.7 µM and A = 0.58.

Effect of Ebastine on I_ped

I_ped, a member of the rapidly activating delayed rectifier K⁺ channel family, is one of the predominant K⁺ channels in rat. In rat myocytes, this current can be best quantified by a pulse procedure (fig. 7A, inset) in which a conditioning depolarization from 90 to 10 mV first inactivates I_to, followed by pulses to more positive potentials that activate I_ped channels alone. Comparison of superimposed tracings of I_ped at potentials between 10 and +70 mV in presence and absence of ebastine shows significant suppression in the presence of the drug (fig. 7A). Ebastine appeared to be significantly more effective in suppressing I_ped than I_to at all potentials and doses tested (fig. 7, B and C). Analysis of the degree of suppression of I_ped by ebastine at membrane potentials between 30 to +75 mV suggested a small voltage dependence of the drug effect such that percent suppression was ~20% at 0 mV and ~35% at +60 mV. The data points of the dose-response relation for the effect of ebastine on I_ped (fig. 7C) were fit with Langmuir's isotherm (see legend to fig. 1C). The fit provided the K_d value of 1.7 µM and the level of maximal block of 58% (proportion of drug-sensitive current, A = 0.58).

Modulation of Rat and Guinea Pig I_K1 by Ebastine

The I_K1 channel is the dominant K⁺ channel in all types of mammalian cardiac cells at potentials negative to 60 mV. Suppression of I_K1 causes not only partial depolarization of resting potentials but also prolongation of phase 3 of the action potential, extending the vulnerable period of the cardiac action potential. Figure 8 shows that ebastine suppresses I_K1 by ~15% at only very high (3.0 µM) concentrations in rat ventricular myocytes. At lower concentrations, the drug effects were small and difficult to quantify. Ebastine had similar blocking potency with respect to I_K1 in guinea pig ventricular myocytes (data not shown): in the physiological potential range (90 to 60 mV); 3.0 µM ebastine suppressed I_K1 by 5% to 15% (n = 5), but the degree of suppression varied from cell to cell. Comparison of the relative effects of ebastine with two other H_l antagonists, loratadine and terfenadine, on I_K1 measured at 140 mV in rat ventricular myocytes (fig. 8) suggests that ebastine is less potent than terfenadine but more potent than loratadine in blocking I_K1 (loratadine data are summarized from another study using the same techniques and experimental protocol; Ducic et al., 1996, 1997).

View larger version (30K): [in this window] [in a new window]   Fig. 8.   Comparison of the suppressive effect of three nonsedating H1 antagonists (loratadine, ebastine and terfenadine) on IK1 in rat ventricular myocytes. A, Superimposed traces of control IK1 and IK1 in the presence of 3 µM loratadine, 3 µM ebastine and 3 µM terfenadine. Inset under the terfenadine records, voltage-clamp protocol used to elicit the currents. B, I-V plots of the control IK1 (squares) and IK1 in the presence of 3 µM loratadine (circles), 3 µM ebastine (triangles) and 3 µM terfenadine (*). C, Histogram showing relative potency of 3 µM loratadine, 3 µM ebastine and 3 µM terfenadine in inhibiting IK1 at 140 mV. Data are cumulative from nine cardiomyocytes. *, Significantly different group (P < .05) from loratadine group.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

The major finding of this study is that ebastine suppresses a number of cardiac K⁺ channels. However, the potency of the drug in suppressing the various members of K⁺ channel family differs, such that the drug suppresses the delayed rectifier family (I_Kr, I_Ks and I_ped) more effectively than shaker (I_to) or inward rectifier I_K1 channels. Among the delayed rectifiers, only I_Kr was prominently suppressed at a 0.3 to 1 µM range of drug concentrations. Figure 9 summarizes the effect of ebastine on I_to and I_ped in rat and on I_Ks and I_Kr in guinea pig ventricular myocytes at 3 µM drug concentrations. The effectiveness of ebastine suppressive effect was found to be I_Kr > I_Ks > I_ped  I_to. Comparing the blocking effects of ebastine with two other commonly used H₁ antagonists, terfenadine and loratadine (Ducic et al., 1997) suggested that the K⁺ channel-blocking property of ebastine is equivalent or somewhat weaker than that of terfenadine and much stronger than that of loratadine.

View larger version (45K): [in this window] [in a new window]   Fig. 9.   Comparison of the suppressive effect of three nonsedating H1 antagonists (loratadine, ebastine and terfenadine) on four K+ currents, Ito, Iped in rat and IKs and IKr in guinea pig ventricular myocytes. Histograms show relative potency of 3 µM loratadine, 3 µM ebastine and 3 µM terfenadine in inhibiting of Ito (A), Iped (B), IKs (C) and IKr (D). Data are cumulative from five to nine cells for each intervention. *, Significantly different groups (P < .05) from corresponding loratadine group.

Ebastines effect on I_Kr expressed in X. laevis oocytes. I_Kr appears to be expressed differentially in myocardium of different mammalian species (Sanguinetti et al., 1995). The quantification of I_Kr and its separation from other delayed rectifiers coexpressed in myocardium has been difficult and often required the use of pharmacological blockers (Sanguinetti et al., 1990). We did not use pharmacological blockers to isolate I_Kr, in guinea pig myocytes, in part because of the quantitative uncertainty of the data obtained using two K⁺ channel modulators together. Instead, expression of the "HERG" channel in oocytes was the alternate route chosen to evaluate the effect of ebastine and terfenadine. Sole expression of I_Kr in X. laevis oocyte made it easier to quantify the dose-response relation of the drug without possible contamination with other K⁺ currents. Ebastine and terfenadines effects were partially reversible, recovering often by 85% to 90% (figs. 3B and 4B). To get some insight into the possible voltage-dependent effects of ebastine and terfenadine on I_Kr, we used two different pulse protocols to measure primarily the steady-state activation or inactivation of the channels. The inactivation curves constructed from I_Kr tail currents in response to membrane repolarization from various positive potentials to the same holding potential in the presence and absence of the two H₁-R antagonist (fig. 3C) show increasing blocking potency at potentials positive to +20 mV. Such an effect may arise from either possible influence of the drug on gating kinetics of recovery from inactivation or increased fraction of blocked channels at more positive potentials. In the absence of significant change in the kinetics of recovery from inactivation in presence of ebastine or terfenadine as reflected by the tail currents (fig. 3A), we conclude that at highly positive potentials these drugs may become more accessible to the channel blocking site.

Possible mechanism of action of ebastine. Even though the experimental myocyte was exposed to the final drug concentration in <50 msec, the suppressive effect of ebastine required 100 to 150 sec to develop fully. Given the wide variance in the suppressive effects of ebastine on different K⁺ channel types (compare figs. 8 and 9) and the slow onset of the drug action, it is unlikely that these drugs interact with the highly conserved H₅ segment (the pore region) of the different channels. The suppressive effect of ebastine showed no significant voltage dependence on I_Ks and only a weak voltage dependence on I_Kr and I_ped. This may suggest that the drug neither senses the voltage field of the membrane nor interacts with the pore region of the channel.

View larger version (11K): [in this window] [in a new window]   Fig. 10.   Chemical structures of three nonsedating Hl antagonists. The three agents presented are loratadine (molecular weight = 382.89), ebastine (molecular weight = 469.67) and terfenadine (molecular weight = 471.69).

K⁺ channel blockade and QT interval prolongation. The gating parameters of the five K⁺ channels investigated in this study are both time and voltage dependent (i.e., the flow of current through the channel is both time and voltage dependent). During the time course of the action potential, therefore, various K⁺ channels contribute differentially to the shape and duration of the action potential. Even though suppression of any one of the K⁺ channels would lead to prolongation of the action potential, the resultant shape of the action potential would, in part, depend on which K⁺ channel is in fact modulated. For example, suppression of I_Ks would most likely prolong and lift the plateau, whereas the block of I_Kr or I_K1 would result in prolongation of phase 3 and triangulation of the action potential. Prolongation of the action potential, regardless of the channel type involved, would of course lead to the prolongation of QT interval in the entire heart. Theoretically, QT interval prolongation could serve as an antiarrhythmic device, a basis for development of class III antiarrhythmic drugs. Recent findings, however, suggest that in some cases QT interval prolongation leads to torsades de pointes and fibrillation (Roden et al., 1996). It may be that only those prolongations that involve lengthening of phase 3 and extending the vulnerable period of the action potential increase the propensity for early afterdepolarizations and torsades de pointes. Consistent with this idea, it appears that only those H₁-R antagonists that significantly suppress I_Kr and I_K1 (e.g., terfenadine and astemizole) induce arrhythmogenesis and torsades de pointes, whereas those that have minimal effects on I_Kr and I_K1 (e.g., loratadine) appear to lack this potential. Based on its K⁺ channel-suppressive profile and data from intact guinea pig heart (Hey et al., 1996b), it appears that ebastine may fall in the terfenadine category in producing arrhythmogenesis.