Jane A. Bursill

Inhibition of the human ether-a-go-go-related gene (HERG) potassium channel by cisapride : affinity for open and inactivated states

Cisapride is a prokinetic agent which has been associated with QT prolongation, torsades de pointes and cardiac arrest. The cellular mechanism for these observations is high affinity blockade of IKr (encoded by HERG). In a chronic transfection model using CHO‐K1 cells, cisapride inhibited HERG tail currents after a step to +25 mV with similar potency at room and physiological temperatures (IC50 16.4 nM at 20–22°C and 23.6 nM at 37°C). Channel inhibition exhibited time‐, voltage‐ and frequency‐dependence. In an envelope of tails test, channel blockade increased from 27±8% after a 120 ms depolarizing step to 50±4% after a 1.0 s step. These findings suggested affinity for open and/or inactivated channel states. Inactivation was significantly accelerated by cisapride in a concentration‐dependent manner and there was a small (−7 mV) shift in the voltage dependence of steady state inactivation. Channel blockade by cisapride was modulated by [K+]o, with a 26% reduction in the potency of channel blockade when [K+]o was increased from 1 to 10 mM. In conclusion, HERG channel inhibition by cisapride exhibits features consistent with open and inactivated state binding and is sensitive to external potassium concentration. These features may have significant clinical implications with regard to the mechanism and treatment of cisapride‐induced proarrhythmia.

Inhibition of HERG potassium channels by the antimalarial agent halofantrine

Halofantrine is a widely used antimalarial agent which has been associated with prolongation of the ‘QT interval’ of the electrocardiogram (ECG), torsades de pointes and sudden death. Whilst QT prolongation is consistent with halofantrine‐induced increases in cardiac ventricular action potential duration, the cellular mechanism for these observations has not been previously reported. The delayed rectifier potassium channel, IKr, is a primary site of action of drugs causing QT prolongation and is encoded by the human‐ether‐a‐go‐go‐related gene (HERG). We examined the effects of halofantrine on HERG potassium channels stably expressed in Chinese hamster ovary (CHO‐K1) cells. Halofantrine blocked HERG tail currents elicited on repolarization to −60 mV from +30 mV with an IC50 of 196.9 nM. The therapeutic plasma concentration range for halofantrine is 1.67–2.98 μM. Channel inhibition by halofantrine exhibited time‐, voltage‐ and use‐dependence. Halofantrine did not alter the time course of channel activation or deactivation, but inactivation was accelerated and there was a 20 mV hyperpolarizing shift in the mid‐activation potential of steady‐state inactivation. Block was enhanced by pulses that render channels inactivated, and channel blockade increased with increasing duration of depolarizing pulses. We conclude that HERG channel inhibition by halofantrine is the likely underlying cellular mechanism for QT prolongation. Our data suggest preferential binding of halofantrine to the open and inactivated channel states.

Inhibition of HERG channels stably expressed in a mammalian cell line by the antianginal agent perhexiline maleate.

Perhexiline has been used as an anti‐anginal agent for over 25 years, and is known to cause QT prolongation and torsades de pointes. We hypothesized that the cellular basis for these effects was blockade of IKr. A stable transfection of HERG into a CHO‐K1 cell line produced a delayed rectifier, potassium channel with similar properties to those reported for transient expression in Xenopus oocytes. Perhexiline caused voltage‐ and frequency‐dependent block of HERG (IC50 7.8 μM). The rate of inactivation was increased and there was a 10 mV hyperpolarizing shift in the voltage‐dependence of steady‐state inactivation, suggestive of binding to the inactivated state. In conclusion, perhexiline potently inhibits transfected HERG channels and this is the probable mechanism for QT prolongation and torsades de pointes. Channel blockade shows greatest affinity for the inactivated state.

Solution structure of CnErg1 (Ergtoxin), a HERG specific scorpion toxin.

The three‐dimensional structure of chemically synthesized CnErg1 (Ergtoxin), which specifically blocks HERG (human ether‐a‐go‐go‐related gene) K+ channels, was determined by nuclear magnetic resonance spectroscopy. CnErg1 consists of a triple‐stranded β‐sheet and an α‐helix, as is typical of K+ channel scorpion toxins. The peptide structure differs from the canonical structures in that the first β‐strand is shorter and is nearer to the second β‐strand rather than to the third β‐strand on the C‐terminus. There is also a large hydrophobic patch on the surface of the toxin, surrounding a central lysine residue, Lys13. We postulate that this hydrophobic patch is likely to form part of the binding surface of the toxin.

Blockade by N-3 polyunsaturated fatty acid of the Kv4.3 current stably expressed in Chinese hamster ovary cells.

The Kv4.3 gene is believed to encode a large proportion of the transient outward current (Ito), responsible for the early phase of repolarization of the human cardiac action potential. There is evidence that this current is involved in the dispersion of refractoriness which develops during myocardial ischaemia and which predisposes to the development of potentially fatal ventricular tachyarrhythmias. Epidemiological, clinical, animal, and cellular studies indicate that these arrhythmias may be ameliorated in myocardial ischaemia by n‐3 polyunsaturated fatty acids (n‐3 PUFA) present in fish oils. We describe stable transfection of the Kv4.3 gene into a mammalian cell line (Chinese hamster ovary cells), and using patch clamp techniques have shown that the resulting current closely resembles human Ito. The current is rapidly activating and inactivating, with both processes being well fit by double exponential functions (time constants of 3.8±0.2 and 5.3±0.4 ms for activation and 20.0±1.2 and 96.6±6.7 ms for inactivation at +45 mV at 23°C). Activation and steady state inactivation both show voltage dependence (V1/2 of activation=−6.7±2.5 mV, V1/2 of steady state inactivation=−51.3±0.2 mV at 23°C). Current inactivation and recovery from inactivation are faster at physiologic temperature (37°C) compared to room temperature (23°C). The n‐3 PUFA docosahexaenoic acid blocks the Kv4.3 current with an IC50 of 3.6 μmol L−1. Blockade of the transient outward current may be an important mechanism by which n‐3 PUFA provide protection against the development of ventricular fibrillation during myocardial ischaemia.

Alternative hypothesis for efficacy of macrolides in acute coronary syndromes

The figure shows our findings. Roxithromycin, but not tetracycline, produced a concentration-dependent, significant reduction in whole-cell conductance, which for these cells is almost entirely determined by the KOR channel. Both 30 mol/L and 100 mol/L concentrations of roxithromycin produced significant reductions in whole-cell conductance of about 35% and 50% respectively, and this effect reversed on washing off the antibiotic. Tetracycline on the other hand in concentrations up to 100 mol/L had no effect on the activity of the KOR channel. In clinical use, typical peak plasma concentrations seen with roxithromycin are about 10 mol/L, but the drug acts intracellularly and is known to be highly concentrated in polymorphonuclear leucocytes and macrophages, where concentrations 30-fold higher have been reported. The concentrations at which we have shown KOR blockade could be applied to clinical use. For tetracycline on the other hand, peak plasma concentrations are generally around 4–5 mol/L and we were unable to observe any effect on KOR at concentrations up to 100 mol/L. An alternative mechanism for the suggested benefit of roxithromycin in acute coronary syndromes might be suppression of macrophage activity via blockade of the KOR channel. This hypothesis would be amenable to a clinical trial in which patients were randomised to roxithromycin or tetracycline. If suppression of C pneumoniae is the major mechanism for benefit then both agents should have a similar efficacy; if our suggestion is correct, then it would be expected that roxithromycin, but not tetracycline would be effective. Alternative hypothesis for efficacy of macrolides in acute coronary syndromes

Comparative Study of the Effects of Erythromycin and Roxithromycin on Action Potential Duration and Potassium Currents in Canine Purkinje Fibers and Rabbit Myocardium

Background: Erythromycin and roxithromycin are macrolide antibiotics in common clini cal use. Erythromycin occasionally produces life-threatening arrhythmias (torsades de pointes) by blocking the outward potassium current responsible for repolarization of the cardiac action potential. Methods and Results: We used standard cellular electrophysiological and whole-cell patch- clamping techniques to compare the relative efficacy of erythromycin and roxithromycin in prolonging cardiac action potential in canine Purkinje fibers and in blocking individual out ward potassium currents in isolated rabbit ventricular myocytes. We demonstrated signifi cant prolongation of action potential duration in canine Purkinje fibers by erythromycin but not roxithromycin at a concentration of 100 μM. The delayed rectifier, the outward potas sium current thought to be most sensitive to modulation by drugs, was significantly depressed by both agents at concentrations of ≥30 μM in isolated rabbit ventricular myo cytes. Both drugs had similar potencies (26% and 21 % reduction by 30 μM erythromycin and roxithromycin, respectively, and 50% and 36% reduction by 100 μM erythromycin and roxithromycin). Neither agent significantly blocked other potassium currents (including the transient outward current). Conclusions: Taking into account normally observed peak blood concentrations of these agents in clinical use and the fact that roxithromycin is not normally administered intrave nously, we conclude that the risk of proarrhythmia during normal clinical use of oral roxithromycin is extremely remote.

DIFFERENTIAL EFFECTS OF ANTIARRHYTHMIC AGENTS ON POST‐PAUSE REPOLARIZATION IN CARDIAC PURKINJE FIBRES

1. Marked action potential duration (APD) prolongation with agents such as quinidine is often a precursor of early afterdepolarizations and triggered activity, thought to be the underlying mechanism of torsade de pointes. Episodes of torsade de pointes commonly occur following a pause.

Quinidine but not disopyramide prolongs cardiac Purkinje fiber action potentials after a pause

Summary: Prolongation of the cardiac action potential (AP), leading eventually to early afterdepolarizations (EADs), is believed to underlie drug-induced long QT syndromes and torsade de pointes. Episodes of torsade de pointes frequently occur after a prolonged pause. We studied the effects of quinidine and disopyramide on AP duration (APD) in canine cardiac Purkinje fibers after pauses of 2,000–10,000 ms. Standard intracellular micro- electrode techniques were used to record APs from canine Purkinje fibers at an interstimulus interval (ISI) of 1,000 ms. Pauses of 2,000–10,000 ms were introduced into the basic drive cycle in the presence and absence of sub-therapeutic and therapeutic concentrations of quinidine and disopyramide. We observed a biphasic response in APD to quinidine and disopyramide at ISI = 1,000 ms. Quinidine but not disopyramide produced a marked dose-and time-dependent additional prolongation of APD immediately after the pauses. This effect was highly statistically significant. We conclude that disopyramide and quinidine have qualitatively different effects on APD after a pause and that this observation may cast some light on the apparently greater frequency of torsade de pointes occurring with quinidine than with disopyramide. Possible mechanisms include differential drug effects on outward potassium or inward sodium channels.

Modulation of the Electrophysiologic Actions of E-4031 and Dofetilide by Hyperkalemia and Acidosis in Rabbit Ventricular Myocytes

Background: E-4031 and dofetilide are new class III antiarrhythmic agents that inhibit the rapid component of the delayed rectifier potassium channel (IKr); however, the effectiveness of many antiarrhythmic drugs in ischemic conditions is uncertain. Methods and Results: We modeled two components of ischemia, hyperkalemia (9.6 mM) and acidosis (pH 6.8), in voltage-clamped single rabbit ventricular myocytes to help deter mine the effect of ischemia on the action of these two drugs. In physiologic solution both E- 4031 and dofetilide blocked IKr and significantly reduced total outward current. In hyper kalemic solution, both E-4031 and dofetilide showed significantly reduced blockade of IKr, while in acidotic solution dofetilide showed significantly reduced blockade of IKr and E-4031 showed a trend to reduced blockade. Neither drug significantly reduced total outward current in hyperkalemic or acidotic solutions. Conclusions: In these conditions, E-4031 and dofetilide demonstrate reduced blockade of I Kr, resulting in loss of class III effect. Furthermore, the complete loss of blocking effect on total outward current during simulated ischemia suggests increases of other repolarizing cur rents also contribute to loss of class III effect.