Importance of Species Selection in Arrythmogenic Models of Q-T Interval Prolongation

In a recent article by Ohtani and colleagues (4), prolongation of the Q-T interval in an anesthetized-rat electrocardiogram (ECG) model was used to determine the arrhythmogenic potential of macrolide antibiotics. Prolongation of the Q-T interval on an ECG has been linked to the rare (but potentially fatal) ventricular arrhythmia known as Torsade de Pointes. In vitro studies suggest that most Q-T-prolonging drugs block the delayed rectifier potassium current (iKr) carried by the pore-forming subunit encoded by hERG (human ether-a-go-go-related gene) in humans (6). While hERG-like mRNA and functional iKr current have been found in many species including dogs, guinea pigs, and rabbits (3), little (if any) functional iKr or hERG-like current is found in the rat ventricle (8). Therefore, we were surprised at the sensitivity of this rat ECG model to the selected macrolide antibiotics, given that these drugs do not elicit iKr blocking until supratherapeutic concentrations are achieved (7). To clarify these issues, we chose to compare the effects of erythromycin and two established iKr-blocking drugs on repolarization of rat ventricular muscle, because Q-T prolongation in vivo is reflective of action potential prolongation. 
 
We used standard microelectrode techniques (2) to evaluate the effects of erythromycin (at a concentration achieved by Ohtani et al. in plasma [3.6 μM]) as well as two standard iKr blockers, dofetilide (10 nM) and E-4031 (1.0 μM), at concentrations 2.5-fold greater than those shown to block 50% of native iKr current (1, 5). These concentrations have also been shown to significantly prolong the action potential duration (APD) in other species prominently expressing iKr (5). In brief, male CD rats (300 to 350 g; Charles River Labs) were anesthetized, hearts were removed, and ventricular papillary muscles were excised and placed in a warmed (37°C) superfusion chamber perfused at the rate of 8 to 10 ml/min with Tyrode's solution containing (millimolar concentrations) NaCl, 131; NaHCO3, 18; NaH2PO4, 1.8; MgCl2, 0.5; dextrose, 5.5; KCl, 4; and CaCl2, 2 (aerated with 95% O2-5% CO2 [pH = 7.4] at room temperature). Muscle preparations were field stimulated at a basic cycle length of 180 ms (corresponding to a typical rat heart rate of approximately 330 beats per min) with a biphasic waveform at a twofold threshold using platinum electrodes in the bath floor. Preparations were paced and equilibrated for a minimum of 1 h before recording. Preparations were considered suitable for study if the maximum diastolic potential was more negative than −70 mV and the APD was greater than 20 ms. Records were obtained during a 10-min drug-free control period and ensuing 25-min period of drug exposure. Impalements were maintained throughout each experiment. APD was measured from the upstroke until repolarization reached 10 mV positive to the maximum diastolic potential. All drugs were made up fresh daily as stocks in 100% dimethyl sulfoxide and then diluted in Tyrode's solution to achieve a final dimethyl sulfoxide concentration of 0.1%. Values are presented as means ± standard errors of the means. Erythromycin lactobionate and dofetilide were synthesized in-house; E-4031 was purchased from Wako Chemicals. 
 
The typical effects of erythromycin, E-4031, and dofetilide on rat ventricular muscle action potentials are illustrated in Fig. ​Fig.1.1. Rat papillary muscle repolarization is very sensitive to erythromycin (Fig. 1B) but not the iKr-blocking drug E-4031 (Fig. 1A). Figure 1C summarizes results obtained from 15 preparations (10 hearts); drugs are arranged on the abscissa in the order of increasing potency for native-iKr blockade. Whereas erythromycin elicited 76.6% ± 6.4% APD prolongation at concentrations substantially lower than those required for iKr blockade, the iKr blockers dofetilide and E-4031 elicited much smaller prolongation (2.4% ± 2.9% and 1.8% ± 2.1%, respectively) at concentrations 2.5-fold the 50% inhibitory concentration for iKr blockade. The greater APD prolongation of rat papillary muscle with erythromycin compared to potent iKr-blocking drugs suggests that iKr is not the current predominantly affected by erythromycin in the rat papillary muscle. 
 
 
 
FIG. 1. 
 
Effects of select drugs on rat ventricular repolarization. (A and B) Action potential traces from rat papillary muscle preparations. In each panel, pairs of superimposed traces show effects obtained during continuous recordings before and after drug exposure. ... 
 
 
 
In conclusion, this in vitro study corroborates the findings of Ohtani et al. regarding the sensitivity of the rat in vivo ECG model for erythromycin. However, the lack of effects of high concentrations of known iKr-blocking drugs on the APD corroborates the reported paucity of iKr in the rat ventricular muscle and suggests that currents other than iKr are responsible for delaying repolarization with erythromycin. Thus, the present studies suggest that the rat ECG model is inappropriate for evaluating the proarrhythmic potential of noncardiovascular drugs that prolong the Q-T interval by blocking iKr. In addition, this highlights the need to consider differences in the ionic currents expressed in preparations employed to evaluate cardiac safety. Further studies are necessary to determine the ionic channel(s) responsible for delayed repolarization in the rat myocardium seen with macrolide antibiotics.


Importance of Species Selection in Arrythmogenic Models of Q-T Interval Prolongation
In a recent article by Ohtani and colleagues (4), prolongation of the Q-T interval in an anesthetized-rat electrocardiogram (ECG) model was used to determine the arrhythmogenic potential of macrolide antibiotics. Prolongation of the Q-T interval on an ECG has been linked to the rare (but potentially fatal) ventricular arrhythmia known as Torsade de Pointes. In vitro studies suggest that most Q-T-prolonging drugs block the delayed rectifier potassium current (iK r ) carried by the poreforming subunit encoded by hERG (human ether-a-go-go-related gene) in humans (6). While hERG-like mRNA and functional iK r current have been found in many species including dogs, guinea pigs, and rabbits (3), little (if any) functional iK r or hERG-like current is found in the rat ventricle (8). Therefore, we were surprised at the sensitivity of this rat ECG model to the selected macrolide antibiotics, given that these drugs do not elicit iK r blocking until supratherapeutic concentrations are achieved (7). To clarify these issues, we chose to compare the effects of erythromycin and two established iK r -blocking drugs on repolarization of rat ventricular muscle, because Q-T prolongation in vivo is reflective of action potential prolongation.
We used standard microelectrode techniques (2) to evaluate the effects of erythromycin (at a concentration achieved by Ohtani et al. in plasma [3.6 M]) as well as two standard iK r blockers, dofetilide (10 nM) and E-4031 (1.0 M), at concentrations 2.5-fold greater than those shown to block 50% of native iK r current (1,5). These concentrations have also been shown to significantly prolong the action potential duration (APD) in other species prominently expressing iK r (5). In brief, male CD rats (300 to 350 g; Charles River Labs) were anesthetized, hearts were removed, and ventricular papillary muscles were excised and placed in a warmed (37°C) superfusion chamber perfused at the rate of 8 to 10 ml/min with Tyrode's solution containing (millimolar concentrations) NaCl, 131; NaHCO 3 , 18; NaH 2 PO 4 , 1.8; MgCl 2 , 0.5; dextrose, 5.5; KCl, 4; and CaCl 2 , 2 (aerated with 95% O 2 -5% CO 2 [pH ϭ 7.4] at room temperature). Muscle preparations were field stimulated at a basic cycle length of 180 ms (corresponding to a typical rat heart rate of approximately 330 beats per min) with a biphasic waveform at a twofold threshold using platinum electrodes in the bath floor. Preparations were paced and equilibrated for a minimum of 1 h before recording. Preparations were considered suitable for study if the maximum diastolic potential was more negative than Ϫ70 mV and the APD was greater than 20 ms. Records were obtained during a 10min drug-free control period and ensuing 25-min period of drug exposure. Impalements were maintained throughout each experiment. APD was measured from the upstroke until repolarization reached 10 mV positive to the maximum diastolic potential. All drugs were made up fresh daily as stocks in 100% dimethyl sulfoxide and then diluted in Tyrode's solution to achieve a final dimethyl sulfoxide concentration of 0.1%. Values are presented as means Ϯ standard errors of the means. Erythromycin lactobionate and dofetilide were synthesized inhouse; E-4031 was purchased from Wako Chemicals.
The typical effects of erythromycin, E-4031, and dofetilide on rat ventricular muscle action potentials are illustrated in Fig. 1. Rat papillary muscle repolarization is very sensitive to erythromycin (Fig. 1B) but not the iK r -blocking drug E-4031 (Fig. 1A). Figure 1C summarizes results obtained from 15 preparations (10 hearts); drugs are arranged on the abscissa in the order of increasing potency for native-iKr blockade. Whereas erythromycin elicited 76.6% Ϯ 6.4% APD prolongation at concentrations substantially lower than those required for iK r blockade, the iK r blockers dofetilide and E-4031 elicited much smaller prolongation (2.4% Ϯ 2.9% and 1.8% Ϯ Drugs are arranged on the abscissa in the order of increasing potency for blocking native iK r . While erythromycin significantly prolongs the APD at a concentration far below that which blocks iK r (IC 50 ϭ 100 M [7]), high concentrations of iK r -blocking agents E-4031 and dofetilide (2.5-fold above the 50% inhibitory concentration for blocking native iK r [1,5]) minimally prolong repolarization. n ϭ 5 per group; the data are means Ϯ standard errors of the means. Statistical significance was defined as P Ͻ 0.05 ‫)ء(‬ (paired t tests).
2.1%, respectively) at concentrations 2.5-fold the 50% inhibitory concentration for iK r blockade. The greater APD prolongation of rat papillary muscle with erythromycin compared to potent iK r -blocking drugs suggests that iK r is not the current predominantly affected by erythromycin in the rat papillary muscle.
In conclusion, this in vitro study corroborates the findings of Ohtani et al. regarding the sensitivity of the rat in vivo ECG model for erythromycin. However, the lack of effects of high concentrations of known iK r -blocking drugs on the APD corroborates the reported paucity of iK r in the rat ventricular muscle and suggests that currents other than iK r are responsible for delaying repolarization with erythromycin. Thus, the present studies suggest that the rat ECG model is inappropriate for evaluating the proarrhythmic potential of noncardiovascular drugs that prolong the Q-T interval by blocking iK r . In addition, this highlights the need to consider differences in the ionic currents expressed in preparations employed to evaluate cardiac safety. Further studies are necessary to determine the ionic channel(s) responsible for delayed repolarization in the rat myocardium seen with macrolide antibiotics.