THE HUMAN ETHER-A-GO-GO-RELATED GENE (hERG) CURRENT INHIBITION SELECTIVELY PROLONGS ACTION POTENTIAL OF MIDMYOCARDIAL CELLS
TO AUGMENT TRANSMURAL DISPERSION

Arrhythmias are serious side effects of both cardiac and non-cardiac drugs caused by the modulation of ion channel activity. Among these, Torsades-de-Pointes (TdP) is one of the most dangerous forms of arrhythmia. TdP is the result of long QT syndrome resulting from prolongation of action potential duration (APD) of myocytes in the ventricular wall. As APD is determined by a balance between inward and outward currents, various channels, exchangers, and pumps are involved in APD prolongation. However, inhibition of rapidly activating delayed rectifier potassium current (IKr), mediated by hERG channels, has been implicated in the majority of drug-induced QT prolongation and/or TdP cases (1).

IKr block gives rise to early or delayed afterdepolarization (EAD or DAD), thereby establishing a trigger for arrhythmia. However, for arrhythmia to develop and be sustained, transmural dispersion of repolarization (TDR) is also required as the substrate (2). Transmural electrophysiological heterogeneity is inherent to the ventricular wall, which is composed of a distinct species of myocytes; endocardial, mid-myocardial (M), and epicardial cells (2). Additionally, drug interventions and changes in heart rate can induce a greater APD prolongation in M-cells compared with epicardial and endocardial cells in a canine wedge preparation (3-5). Accordingly, for a more accurate evaluation of the arrhythmogenic risk of drugs, separate measurements of APD in cells from each layer are required. Although use of a ventricular wedge preparation or separate isolation of cells from each layer of the ventricular myocardium allows such measurement, (3, 6) these experiments are expensive and time-consuming. An alternative approach is the use of computational models of electrophysiology, kinetic parameters of which are obtained from conventional patch clamp measurements. However, the resultant behavior of a hERG current in such models depends heavily on the mathematical description of hERG channel kinetics. (7-9).

The action potential clamp technique is an established technique for studying ion currents in the physiological milieu. In contrast to conventional techniques using the square wave pulse, the action potential clamp technique can be used to record the behavior of specific ion currents directly without using mathematical models, as the physiological action potential waveform obtained experimentally or in simulations can be used instead (10-14). Specifically, the dynamic action potential clamp technique developed by Berecki and colleagues provides an opportunity to evaluate the effect of hERG current inhibition on APD, because of the real-time feedback of current to the action potential calculated using the cell model (15, 16).

In the present study, we examined the effect of disopyramide, a drug that has been associated with cases of QT prolongation and TdP (17, 18). Specifically, we used the dynamic action potential technique and applied it to heterogeneously expressed hERG channels in Chinese hamster ovary (CHO) cells. By coupling the measured hERG current with a human ventricular myocyte model, we examined the effect of disopyramide on APD to assess whether disopyramide-induced hERG inhibition prolongs the APD of M-cells more than that of epicardial and endocardial cells. In addition, the role of slowly activating delayed rectifier current (IKs) in repolarization reserve is discussed.

MATERIALS AND METHODS

Electrophysiological recording

Chinese hamster ovary (CHO) cells, that stably expressed hERG channels, were provided by the Channelopathy Foundation (Berne, Switzerland). Membrane currents were recorded in a whole-cell configuration using an automated patch clamp system (Port-a-Patch, Nanion Technologies, Munich, Germany) equipped with a patch clamp amplifier (EPC-10, HEKA Electronics, Lambrecht, Germany).

Measurements were performed at 36°C with internal (KCl 50 mM, NaCl 10 mM, KF 60 mM, EGTA 20 mM, HEPES 10 mM, pH 7.2, 285 mOsmol) and external (NaCl 140 mM, KCl 4 mM, MgCl2 1 mM, CaCl2 2 mM, D-glucose 5 mM, HEPES 10 mM, pH 7.4, 298 mOsmol) solutions.

Action potential clamp

The action potential clamp technique used was similar to the model-cell method of Berecki et al., (15) in which a computer simulation model of a ventricular myocyte and a CHO cell expressing hERG channels are electrically coupled. We used this system in both open-loop and closed-loop (dynamic action potential clamp) modes. In the open-loop mode, membrane potential calculated by the human ventricular cell model including IKr was used as the command voltage (Fig. 1A) (19, 20). In the closed-loop mode, IKr of the model was set to zero and replaced by the measured IKr for the calculation of membrane potential (Fig. 1B). Prior to the closed-loop experiments, we measured the IKr under the open-loop condition in each cell, and set the scaling factor for the following closed-loop experiment by assuming the measured IKr peak to be equal to that calculated by the model. The resulting calculated membrane potential was used as the command voltage for the voltage clamped CHO cells in the next time-step. We modelled endocardial, M, and epicardial cells by changing the conductance of IKs and the transient outward current (Ito), as proposed by the original ten Tusscher et al. model (19, 20). The control program was written in LabView 2012 software using the real-time module (National Instruments, Austin, TX, USA), and analog-to-digital and digital-to-analog conversions were performed at 10 kHz (PXI-1036 and 6251, National Instruments).

Fig. 1. Diagram of the action potential clamp experiments. (A) Open-loop condition. An action potential, derived from the ventricular cell model including IKr, was applied to a CHO cell expressing hERG channels and IKr was recorded. (B) Feedback condition. The recorded IKr was scaled and digitized, and then included in the membrane potential calculation of the ventricular cell model. The calculated action potential was then applied to the CHO cell.

Drugs

E-4031 and disopyramide were purchased from Sigma (Sigma, Tokyo, Japan).

Statistics

Data are expressed as mean ± S.E. A value of P < 0.05 was considered significant for ANOVA and for Student’s t-test, and the Bonferroni method was used to adjust P values for multiple comparisons.

RESULTS

Definition of IKr

The IKr was measured in 10 cells during an action potential in the presence and absence of 10 µM E4031 under the open-loop condition. As shown in the middle row of Fig. 2, we observed significant outward currents during IKr blockade by E4031 in all types of action potentials. By subtracting these currents, we defined IKr as an E4031-sensitive current with a similar shape to the calculated waveforms (bottom row of Fig. 2).

Fig. 2. IKr recordings under the open-loop condition using endocardial (A), M (B), and epicardial (C) cell models. Top row, action potential; Middle row, hERG currents with (gray line) and without (black line) E-4031. Bottom row, calculated IKr used for the feedback experiments. Representative data obtained from one cell for each cell type.

Although it would have been preferable to identify IKr in the closed-loop experiment, it was difficult to wash out E4031 while holding the cell under study. Thus, we estimated IKr as follows. As seen in Fig. 2, the amplitude of the shoulder point after the dip in hERG current measured without E4031 was very close to the plateau of the current measured with E4031 (“a” and “b”, middle panel of Fig. 2A). Thus, we measured IKr under the open-loop condition and determined the point “a” and the peak of the hERG current (“c” in the middle panel of Fig. 2A). Assuming that “c – a” is equal to the peak amplitude of IKr calculated by the model (IKrmodel), the scaling factor F = (c – a) / IKrmodel was determined. Finally, the IKr calculated by the following equation was fed back to the simulation model in each time-step:

IKr = ((measured current) – a) × F .

Effect of disopyramide

The effects of disopyramide on both the IKr and action potential during the action potential clamp under the closed-loop condition are shown for the three types of cells in Fig. 3. In all types of cells, the IKr decreased in a dose dependent manner resulting in prolongation of APD, although APD prolongation was highest in M-cells. The dose response of the averaged APD90 (n = 5 cells for each cell type) confirmed the difference among cell types (Fig. 4A, Table 1). Of note, compared with the modest and similar APD prolongation observed in epicardial and endocardial cells, the APD of M-cells increased progressively at the higher dose range with greater dispersion. To confirm this observation, we normalized APD with various doses of disopyramide under control conditions for each cell type (Fig. 4B). The differences among the normalized APDs were significantly different at higher doses (10 and 100 µM). By contrast, inhibition of peak current, similarly normalized under control conditions, did not differ among the cell types over the entire range of drug doses (Fig. 4C, Table 2).

Table 1. Drug concentration versus APD.

ADPs are shown as mean ± S.E. (msec). *P < 0.05 vs. 0 µM. ** P < 0.05 vs. 1 µM. ***P < 0.05 vs. 10 µM.

Fig. 3. Effects of disopyramide on APD (top row) and IKr (bottom row) recorded in endocardial (A), M (B), and epicardial (C) cell models under the feedback condition. The concentrations of disopyramide used were 0 µM (blue, control), 1 µM (red), 10 µM (green), and 100 µM (purple). Representative data obtained from one cell for each cell model.

Fig. 4. Effects of disopyramide on APD and IKr (n = 5 for each cell model). (A) Relationship between APD and disopyramide concentration for the three types of cell models. *P < 0.05 between M and endocardial cells. #P < 0.05 between M and epicardial cells. ANOVA followed by ad hoc Student’s t-test, with adjustment by Bonferroni method. (B) APDs were normalized by control values for each cell model and plotted as a function of disopyramide concentration. *P < 0.05 between M and endocardial cells. #P < 0.05 between M and epicardial cells. ANOVA followed by ad hoc Student’s t-test, with adjustment by Bonferroni method. (C) Relationship between peak IKr inhibition and disopyramide concentration. Peak IKr inhibition was calculated as follows: (peak IKr – peak IKr control) / peak IKr control.

Table 2. Drug concentration versus peak IKr inhibition.

Peak IKr inhibition rates are shown as mean ± S.E. [–]. *P < 0.05 vs. 0 µM. **P < 0.05 vs. 1 µM. ***P < 0.05 vs. 10 µM.

Instantaneous current-potential (I-V) relationships in different cell types

We compared the instantaneous I-V relationships in the absence (control) and presence of 10 µM disopyramide in the three cell types by normalizing the current using the peak value under control conditions (Fig. 5). In all cell types, disopyramide reduced the peak current by approximately 40% (epicardial, 56.5 ± 8.1; M, 61 ± 10; endocardial, 55.9 ± 9.9) and inhibited the current similarly over the potential range examined. Combined with the results shown in Fig. 4C, these data suggest that the augmented APD prolongation in M-cells with higher doses of disopyramide was not due to a pharmacological effect on hERG channels. In addition, the slower time course of membrane potential change of M-cells did not affect the instantaneous I-V relationship appreciably. These results are in clear contrast to the work by Hancox et al. demonstrating the dependence of I-V relationship on applied voltage waveform. Nevertheless, small differences in action potential waveforms between M-cells and other cell types may account for this discrepancy (11).

Fig. 5. Instantaneous I-V relationship during the repolarization phase for endocardial (A), M (B), and epicardial (C) cell models. Closed circle, control; open circle, 10 µM disopyramide. Averaged data from five cells in each cell model.

Kinetics of IKs during APD prolongation

Next, we investigated the cause of disproportionate APD prolongation in M-cells by studying the slowly activating delayed rectifier current (IKs) channel, another major participant in membrane repolarization. The density of IKs in M-cells is known to be low, and is modeled in the ten Tusscher et al. model by reducing the conductance values using the following formula: KKs = 0.062 (M-cell), 0.245 (epicardial and endocardial) nS/pF. (19).

τs, time constant, EKs, reversal potential

Such a low IKs density is the basis for reduced repolarization reserve in M-cells and explains the longer APD under control conditions, but does not account for augmented prolongation in the higher dose range of disopyramide. We hypothesized that the kinetics of IKs has an impact on the progressive loss of repolarization reserve in the higher dose range, and thus examined at the status of IKs gating xs2 over a wide range on action APD (Fig. 6A top panel). Accordingly, we applied various action potentials of epicardial (broken lines) and M-cells (solid lines) obtained in the experiments. We excluded endocardial action potentials from the analysis because the shape of phase I is different in this cell type, due to the distinct characteristics of the transient outward current. As APD became longer, the peak of xs2 became larger and reached approximately at APD50 (Fig. 6A, second panel). When the peak xs2 values were plotted as a function of APD50, the relationship was curvilinear and leveled off at the longer APD50 range (Fig. 6B, closed circle), suggesting saturation of repolarization reserve. To further clarify these findings, we applied linear regression to the data points in the physiological range (between the APD50 of epicardial cells and M-cells under control conditions, Fig. 6B, indicated by the vertical broken lines). Under the influence of disopyramide, the epicardial and endocardial APDs became longer in this region (white arrow), whereas the M-cell APD became prolonged above this region where the repolarization reserve levels off (black arrow). When the time constant (τs) was set at 50% of the control, xs2 reached higher values earlier (Fig. 6A third panel), but leveled off significantly (Fig. 6B, square). When we set ts at double the control, the peak xs2 increased linearly over the entire range, while the absolute peak values were smaller (Fig. 6A bottom panel, Fig. 6B, triangle).

Fig. 6. The responses of IKs channels with different kinetic properties. (A) Action potentials of varying duration observed in epicardial (broken lines) and M-cells (solid lines) (top panel), and corresponding IKs responses with a normal (second panel), 50% normal (third panel), and 200% normal (bottom panel) kinetic time constant. (B) Peak IKs gate values identified in A plotted as a function of APD50 for normal (solid circle), 50% normal (square), and 200% normal (triangle) kinetic time constants. Linear regression lines applied to the data in the physiological range (between the APD50 of epicardial cells and M-cells both under control conditions, indicated by the vertical broken lines) are also shown with arrows indicating the region where the epicardial APD (white arrow) or M cell APD reached under the influence of IKr inhibition.

DISCUSSION

Cardiotoxicity testing and action potential clamp

An in vitro assay of IKr through hERG channels and QT interval testing in animals and humans (by ECG QT study) are commonly used for the preclinical evaluation of the arrhythmogenic risk of drugs. However, these tests are imperfect in their predictive ability and often report false positives or negatives (21). Ignorance of the drug effect on multiple ion channels, other than hERG, is a major cause of flaws in current screening strategies. In fact, roles of IK1 channel, Na+/Ca2+ exchanger and Na+/H+ exchanger in cardiac arrhythmogenesis have been reported (22, 23). However, a revision of the hERG channel assay may improve the testing performance. First, although IKr is determined by the interaction between activation and inactivation kinetics of hERG channels controlled by the membrane potential, (24-26), detailed analyses of these properties are eliminated in most screening assays, and only the inhibition of the peak tail current is evaluated. Secondly, arrhythmia is the consequence of complex interplay between different ion channels, as well as myocytes with distinct characteristics (2). Investigating this complex interplay can only be achieved using expensive animal experiments. In the present study, we applied the dynamic action potential clamp technique to examine the effect of disopyramide on IKr, reflecting both activation and inactivation processes during the physiological action potential. The significance of action potential clamp technique has also been shown in the recent report evaluating the hERG channel blockade by a selective-serotonin reuptake inhibitor (27). We also addressed the issue of heterogeneous tissue properties using three different cell models. Our results clearly demonstrate a differential effect of disopyramide on these cell types, which may result in an increase in transmural dispersion of repolarization, and therefore, the risk of arrhythmia.

Mechanistic considerations

An effect of drugs and heart rate on APD prolongation in M cells has been reported in studies using canine ventricular wedge preparations (3-5). As IKs and IKr play important roles in membrane repolarization, M cells, which have a lower density of IKs compared with other cells, are expected to show greater APD prolongation when IKr is inhibited because of reduced repolarization reserve (28). However, augmentation of APD prolongation in the higher dose range of disopyramide was observed for the first time in this study, and cannot be solely explained by the low density of IKs in M cells. Our simulation results revealed the importance of gating kinetics in the regulation of action potential repolarization. As summarized in Fig. 6B, if the time constant of gating is short, the IKs channel actively responds to APD prolongation to compensate for the inhibition of IKr current, but such an effect levels off drastically once the APD is prolonged beyond the physiological range. In contrast, if the time constant is long, compensatory augmentation of IKs current functions over a wide range of APD, but its magnitude is small. In this sense, the kinetics of IKs is optimally tuned to maintain sufficient repolarization reserve over the physiological range and beyond of APD with an appropriate magnitude. However, if prolongation of APD exceeds the upper range by IKr inhibition, reserve is lost and transmural dispersion is augmented. At present, the importance of gating kinetics has been discussed in connection with the heart rate dependence of APD (28), although the contribution of IKs to cardiac repolarization may change dynamically, not only between beats (rate dependent), but also within a single beat, depending on the APD duration. This possibility has been shown for the first time in the present study.

Study limitations

Currently, several models of human ventricular action potential are available (19, 20, 29-31). However, the models are constructed differently, and yielding different action potential characteristics. We adopted a model by ten Tusscher et al. (19) as it is consistent with experimental results on IKs and IKr block (32). However, because the IKs density of the ten Tusscher et al. model is larger than other models, comparison of this model with other models should be undertaken in the future.

Out of 12 ion currents in the ventricular cell model of electrophysiology, only the IKr was replaced by experimental recordings in the current study. However, because disopyramide is a class Ia anti-arrhythmic drug blocking fast sodium channels, Ito and L-type Ca channel (ICaL) (33-35) measured APDs with disopyramide can be different from those evaluated in intact cells. To further elucidate this point, we evaluated the impact of INa and ICaL inhibitions on ADP by simulations. We created dose-inhibition curves for INa, ICaL and IKr using IC50 values and Hill coefficients (nH) adopted from the literature (35) and estimated the inhibition rate of each current at 100 µM of disopyramide (Fig. 7A). Because IC50 values of INa and ICaL are much larger than that of IKr (INa: IC50 = 168.4 µM, nH = 1.09; ICaL: IC50 = 1036.7 µM, nH = 1.0; IKr: IC50 = 14.4 µM, nH = 0.91), estimated inhibition rates of INa and ICaL were small compared with IKr (INa: 37%, ICaL: 9%, IKr: 85%). Using these inhibition rates, we performed action potential simulations at 100 µM of disopyramide in the absence and presence of INa and ICaL inhibitions. Similar to prior experiments, inhibition of IKr prolonged the APD of M-cell disproportionately (Fig. 7C) compared with controls (Fig. 7B), while the inclusion of INa and ICaL inhibitions did not induce additional changes to APDs of three cell types (Fig. 7C). These data indicate that the impact of inhibitions of INa and ICaL is minimal, at least at this concentration of disopyramide. However, dose-inhibition curves suggest that inhibition of ICaL is expected to contribute to APDs at even higher concentrations.

Fig. 7. Effects of other ion channels. (A) Dose inhibition relations of disopyramide for IKr (solid line), INa (dotted line) and ICaL (broken line). Inhibition rate was calculated as % inhibition = 1/(1 + 10(logIC50 – logx)*nH) × 100, where x is the drug concentration. IC50 values and Hill coefficients (nH) were adopted from the literature (INa: IC50 = 168.4 µM; nH = 1.09; ICaL: IC50 = 1036.7 µM; nH = 1.0; IKr: IC50 = 14.4 µM, nH = 0.91) (35). Simulated action potentials under control condition (B), with IKr inhibition (C) and inhibitions of IKr, ICaL and INa (D). Solid line: epicardial cell, dotted line: endocardial cell, broken line: M-cell.

Finally, in this study, we used CHO cells stably expressing hERG because of their availability. However, the use of other cell lines without expressing IKs channels could have facilitated experiments and analyses, and should be considered in future studies.

Summary

Using the action potential clamp technique coupled with the in silico model of cardiac electrophysiology, we studied the effect of the hERG inhibitor disopyramide on the action potential duration of epicardial-, M-, and endocardial cells in the ventricular wall. Disopyramide increased the action potential duration of M-cells disproportionately compared with the other two cell types, despite the similar inhibitory effect on hERG current for all three cell types. The resulting augmentation of the transmural dispersion of repolarization may underlie the arrhythmogenicity of this drug. The action potential clamp technique is useful for studying the mechanism of arrhythmogenesis by hERG inhibition under physiological conditions, which cannot be provided by conventional voltage clamp protocols.

Author contributions: C.Y., J.O., and S.S. conception and design of research; C.Y., and S.Y. performed experiments; C.Y. and S.Y. analyzed data; J.O., H.Y. and S.S. interpreted results of experiments; S.S. drafted manuscript; H.Y., T.H, and S.S. edited and revised manuscript; C.Y., S.Y., H.Y., J-I.O., T.H., and S.S. approved final version of manuscript.

Acknowledgements: Some parts of the electrophysiological data were generated using ion channel cells designed and owned by the Channelopathy Foundation.

This research was supported by the Japan Society for the Promotion of Science (JSPS) through its “Funding Program for World-Leading Innovative R&D on Science and Technology” (FIRST Program).

Conflict of interests: None declared.

REFERENCES

1. Redfern WS, Carlsson L, Davis AS, et al. Relationships between preclinical cardiac electrophysiology, clinical QT interval prolongation and torsade de pointes for a broad range of drugs: evidence for a provisional safety margin in drug development. Cardiovasc Res 2003; 58: 32-45.
2. Antzelevitch C. Drug induced channelopathies. In: Cardiac Electrophysiology: From Cell to Bedside. Zipes DP, Jalife J, (eds.), Philadelphia, W.B. Saunders; 2009, pp. 195-203.
3. Sicouri S, Antzelevitch C. A subpopulation of cells with unique electrophysiological properties in the deep subepicardium of the canine ventricle. The M cell. Circ Res 1991; 68: 1729-1741.
4. Shimizu W, Antzelevitch C. Sodium channel block with mexiletine is effective in reducing dispersion of repolarization and preventing Torsade de Pointes in LQT2 and LQT3 models of the long-QT syndrome. Circulation 1997; 96: 2038-2047.
5. Akar EG, Yan GX, Antzelevitch C, Rosenbaum DS. Unique topographical distribution of M cells underlies reentrant mechanism of Torsade de Pointes in the long-QT syndrome. Circulation 2002; 105: 1247-1253.
6. Stankovicova T, Szilard M, De Scheerder I, Spido KR. M cells and transmural heterogeneity of action potential configuration in myocytes from the left ventricular wall of the pig heart. Cardiovasc Res 2000; 45: 952-960.
7. Clancy CE, Rudy Y. Cellular consequences of hERG mutations in the long QT syndrome: precursors to sudden cardiac death. Cardiovasc Res 2001; 50: 301-313.
8. Suzuki S, Murakami S, Tsujimae K, Findlay I, Kurachi Y. In silico risk assessment for drug-induction of cardiac arrhythmia. Prog Biophys Mol Biol 2008; 98: 52-60.
9. Fink M, Noble D, Virag L, Varro A, Giles WR. Contributions of HERG K+ current to repolarization of the human ventricular action potential. Prog Biophys Mol Biol 2008; 96: 357-376.
10. Doerr T, Denger R, Trautwein W. Calcium currents in single SA nodal cells of the rabbit heart studied with action potential clamp. Pflugers Arch 1989; 413: 599-603.
11. Hancox JC, Levi AJ, Witchel HJ. Time course and voltage dependence of expressed HERG current compared with native “rapid” delayed rectifier K current during the cardiac ventricular action potential. Pflugers Arch 1989; 436: 843-853.
12. Linz KW, Meyer R. Modulation of L-type calcium current by internal potassium in guinea pig ventricular myocytes. Cardiovasc Res 1997; 33: 110-122.
13. Puglisi JL, Yuan W, Bassani JW, Bers DM. Ca(2+) influx through Ca(2+) channels in rabbit ventricular myocytes during action potential clamp: influence of temperature. Circ Res 1999; 85: e7-e16.
14. El Harchi A, Melgari D, Zhang YH, Zhang H, Hancox JC. Action potential clamp and pharmacology of the variant 1 Short QT Syndrome T618I hERG K channel. PLoS One 2012; 7: e52451.
15. Berecki G, Zegers JG, Verkerk AO, et al. hERG channel (dys)function revealed by dynamic action potential clamp technique. Biophys J 2005; 88: 566-578.
16. Verkerk AO, Zegers JG, van Ginneken ACG, Wilders R. Dynamic action potential clamp as a powerful tool in the development of a gene-based bio-pacemaker. Annual International Conference of the IEEE Engineering in Medicine & Biology Society, 2008.
17. Kurita T, Ohe T, Shimizu W, et al. Early afterdepolarization-like activity in patients with class 1A induced long QT syndrome and Torsades de pointes. Pacing Clin Electrophysiol 1997; 20: 695-705.
18. Furushima H, Niwano S, Chinushi M, Ohhira K, Abe A, Aizawa Y. Relation between bradycardia dependent long QT syndrome and QT prolongation by disopyramide in humans. Heart 1998; 79: 56-58.
19. ten Tusscher KH, Noble D, Noble PJ, Panfilov AV. A model for human ventricular tissue. Am J Physiol 2004; 286: H1573-H1589.
20. ten Tusscher KH, Panfilov AV. Alternans and spiral breakup in a human ventricular tissue model. Am J Physiol 2006; 291: H1088-H1100.
21. Chi KR. Revolution dawning in cardiotoxicity testing. Nat Rev Drug Discov 2013; 12: 565-567.
22. Bebarova M, Matejovic P, Pasek M, Simurdova M, Simurda J. Dual effect of ethanol on inward rectifier pottasium current IK1 in rat ventricular myocytes. J Physiol Pharmacol 2014; 65: 497-509.
23. Szepesi J, Acsai K, Sebok Z, et al. Comparison of the efficiency of Na+/Ca2+ exchanger or Na+/H+ exchanger inhibition and their combination in reducing coronary perfusion-induced arrhythmias. J Physiol Pharmacol 2015; 66: 215-226.
24. Smith PL, Baukrowitz T, Yellen G. The inward rectification mechanism of the HERG cardiac potassium channel. Nature 1996; 379: 833-836.
25. Liu S, Rasmusson RL, Campbell DL, Wang S, Strauss HC. Activation and inactivation kinetics of an E-4031 sensitive current (IKr) from single ferret atrial myocytes. Biophys J 1996; 70: 2704-2715.
26. Wang S, Liu S, Morales MJ, Strauss HC, Rasmusson RL. A quantitative analysis of the activation and inactivation kinetics of HERG expressed in Xenopus oocytes. J Physiol 1997; 502: 45-60.
27. Alexandrou AJ, Milnes JT, Sun SZ, et al. The human ether-a-go-go related gene (hERG) K+ channel blockade by the investigative selective-serotonin reuptake inhibitor CONA-437: limited dependence on S6 aromatic residues. J Physiol Pharmacol 2014; 65: 511-523.
28. Silva J, Rudy Y. Subunit interaction determines IKs participation in cardiac repolarization and repolarization reserve. Circulation 2005; 112: 1384-1391.
29. Priebe L, Beuckelmann DJ. Simulation study of cellular electric properties in heart failure. Circ Res 1998; 82: 1206-1223.
30. Grandi E, Pasqualini FS, Bers DM. A novel computational model of the human ventricular action potential and Ca transient. J Mol Cell Cardiol 2010; 48: 112-121.
31. O’Hara T, Virag L, Varro A, Rudy Y. Simulation of the undiseased human cardiac ventricular action potential: Model formulation and experimental validation. PLoS Comput Biol 2011; 7: e1002061.
32. ten Tusscher KH, Bernus O, Hren R, Panfilov AV. Comparison of electrophysiological models for human ventricular cells and tissues. Prog Biophys Mol Biol 2006; 90: 326-345.
33. Grant AO, John JE, Nesterenko VV, Starmer CF, Moorman JR. The role of inactivation in open-channel block of the sodium channel: studies with inactivation-deficient mutant channels. Mol Pharmacol 1996; 50: 1643-1650.
34. Virag L, Varro A, Papp JG. Effect of disopyramide on potassium currents in rabbit ventricular myocytes. Naunyn-Schmiedeberg’s Arch Pharmacol 1998; 357: 268-275.
35. Kramer J, Obejero-Paz CA, Myatt G, et al. MICE Models: Superior to the HERG model in predicting Torsade de Pointes. Sci Rep 2013; 3: 2100.