Proarrhythmia Risk Assessment Using Electro-Mechanical Window in Human iPS Cell-Derived Cardiomyocytes

Abstract

Evaluation of drug-induced cardiotoxicity is still challenging to avoid adverse effects, such as torsade de pointes (TdP), in non-clinical and clinical studies. Numerous studies have suggested that human-induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) are a useful platform for detecting drug-induced TdP risks. Comprehensive in vitro Proarrhythmia Assay (CiPA) validation study suggested that hiPSC-CMs can assess clinical TdP risk more accurately than the human ether-a-go-go-related assay and QT interval prolongation. However, there were still some outliers, such as bepridil, mexiletine, and ranolazine, among the CiPA 28 compounds in the CiPA international multi-site study using hiPSC-CMs. In this study, we assessed the effects of the positive compound dofetilide, the negative compound aspirin, and several CiPA compounds (bepridil, mexiletine, and ranolazine) on the electromechanical window (E-M window), which were evaluated using multi-electrode array assay and motion analysis, in hiPSC-CMs. Similar to previous in vivo studies, dofetilide, which has a high TdP risk, decreased the E-M window in hiPSC-CMs, whereas aspirin, which has a low TdP risk, had little effect. Bepridil, classified in the high TdP-risk group in CiPA, decreased the E-M window in hiPSC-CMs, whereas ranolazine and mexiletine, which are classified in the low TdP-risk group in CiPA, slightly decreased or had little effect on the E-M window of hiPSC-CMs. Thus, the E-M window in hiPSC-CMs can be used to classify drugs into high and low TdP risk.

INTRODUCTION

In the evaluation of drug-induced cardiotoxicity in non-clinical and clinical studies, it is still challenging to avoid adverse effects, such as arrhythmia, contractility dysfunction, and hypertension.¹⁾ Among the drug-induced cardiotoxicities, Torsade de Pointes (TdP), a fatal ventricular arrhythmia, can be the leading cause of drug withdrawal from the market.²⁾ To assess the drug-induced TdP risk, in vitro and in vivo studies were carried out to assess ventricular repolarization, including human ether-a-go-go-related (hERG) and in vivo assays to assess QT prolongation using an electrocardiogram. As several drugs that inhibit hERG or prolong QT interval have a low TdP risk, there is room for improvement in the current TdP risk assessment.

Numerous studies suggest that human-induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) are a useful platform for detecting drug-induced TdP risks.^3–5) Based on the delayed repolarization effects and occurrence of arrhythmia-like events, such as early after depolarization (EAD), in hiPSC-CMs, the Comprehensive in vitro Proarrhythmia Assay (CiPA) and Japan iPS Cardiac Safety Assessment have conducted validation studies to assess drug-induced proarrhythmic risk across multiple experimental sites.^3,4,6) The CiPA study evaluated the effect of 28 drugs, which were divided into three clinical TdP risk groups (high, intermediate, and low) by expert consensus, and found that three predictors (occurrence of arrhythmia-like events, prolongation of repolarization at maximum doses tested, and prolongation of repolarization at clinical exposure) performed risk categorization with sufficient accuracy (area under the curve was 0.8).⁷⁾ These data suggest that hiPSC-CMs can assess clinical TdP risk more accurately than the hERG assay and QT interval prolongation.³⁾ However, there were still some outliers, such as bepridil, mexiletine, and ranolazine, among the CiPA 28 compounds. Therefore, it is important to investigate other endpoints to improve the predictability of the TdP risk in hiPSC-CMs.

Previous in vivo studies have reported the electromechanical window (E-M window), which is defined as the interval between the end of T wave and the end of the left ventricular pressure, as another endpoint to assess TdP risk.^8,9) In in vivo studies using guinea pigs, drug-induced shortening of the E-M window improved TdP risk assessment compared with QT interval prolongation or hERG inhibition alone. Bepridil decreased the E-M window, whereas ranolazine and mexiletine had little effect in guinea pigs.^9,10) In addition to the in vivo study, we have previously reported the E-M window, which was defined as the difference duration between the electrical and mechanical events, using a multi-electrode array (MEA) system and motion analysis in hiPSC-CMs. We showed that the known TdP risk, E-4031, decreased the E-M window, while moxifloxacin, which prolonged the QT interval without TdP,¹¹⁾ and had little effect.⁵⁾ These data suggest that the E-M window can distinguish between compounds with high TdP risk and those with low TdP risk.

Here, we investigated whether the E-M window in hiPSC-CMs assessed drug-induced TdP risk using outlier drugs in the CiPA validation study. Our findings suggest that the E-M window improves the predictability of drug-induced TdP risk using hiPSC-CMs.

MATERIALS AND METHODS

Chemicals

Dofetilide, aspirin, bepridil hydrochloride, ranolazine dihydrochloride, mexiletine hydrochloride, and fibronectin were purchased from Sigma-Aldrich (St. Louis, MO, U.S.A.). Dulbecco’s phosphate-buffered saline (PBS) was purchased from Gibco (Gaithersburg, MD, U.S.A.). All other reagents were obtained from commercial sources.

Cell Culture

iCell cardiomyocytes 2.0 were purchased from FUJIFILM Cellular Dynamics International (FCDI, Madison, WI, U.S.A.). To prepare MED probes (MED-PG515A, Alpha MED Sciences, Osaka, Japan), the recording areas were coated with 1–2 µL of fibronectin and dissolved in PBS to obtain a 50 µg/mL solution. The coated probes were incubated at 37 °C for more than an hour. The iCell cardiomyocytes 2.0 were thawed and suspended in iCell Cardiomyocyte Plating Medium (FCDI) and plated onto the MED probes at a density of 3.5 × 10⁴ cells in 2 µL of medium. After incubation for 4 h, the probes were filled with 1 mL of iCell cardiomyocyte maintenance medium (FCDI) containing 10% fetal bovine serum. The maintenance medium was replaced every 1–2 d, and the cells were analyzed from days 5 to 14.

Field Potential Recordings

Field potentials (FPs) were recorded as previously described.^4,5) Briefly, before measuring FPs, hiPSC-CMs were incubated for more than 30 min using a 5% CO₂ incubator at 37 °C, after replacement with a fresh medium. For stabilization of waveform, the probes were set on the MEA system (MED64, Alpha MED Scientific) and equilibrated for at least 30 min in a humidified 5% CO₂ atmosphere. Once the waveform of the FPs reached a stable condition, FP recording was carried out for 10 min before drug administration. The drugs were added cumulatively to the probe up to 0.5% dimethyl sulfoxide (DMSO). FP recordings were carried out for 10 min after treatment with each concentration.

Field Potential Analysis

FP data analysis was performed as described previously.^4,5) Briefly, FP duration (FPD_end) was defined as the duration from the 1st peak to the end of 2nd peak in FPs. The FPD_end was corrected by the beating rate (inter-spike interval (ISI)) using Fridericia’s formula [FPD_end-cF = FPD_end/(ISI/1000)^1/3]. First peak amplitude was defined as the difference between the maximum and minimum of FPs at 1st peak. The ISI, FPD_end, and 1st peak amplitude were automatically recorded and averaged from the last 30 beats or from 30 beats at the time of stable ISI and FPD_end.

Video Microscopy for Cell Motion

Movie images were recorded as described previously.^5,12) Briefly, movie images of beating hiPSC-CMs were recorded using an SI8000 cell motion imaging system (Sony Corporation, Tokyo, Japan). The focus of the movie images was auto-adjusted or manually adjusted if necessary. The probes were placed on a stage-top incubation system and incubated at 37 °C in 5% CO₂ for at least 30 min for stabilization. The drugs were treated cumulatively with each probe until the maximum concentration of 0.5% DMSO was reached. Video image recording was performed 10 min after drug administration at each concentration for 10 s.

Motion Vector Analysis

Movie images were analyzed as described previously.^12,13) Briefly, a motion waveform was obtained from the movie images using the SI8000 Analyzer software after each recording of the beating hiPSC-CMs. This software system analyzes the contractile movements of hiPSC-CMs using a block-matching algorithm.¹⁴⁾ Several contraction parameters were automatically obtained from the average of the motion waveforms obtained during 10 s measurements, that is, contraction velocity, relaxation velocity, and contraction–relaxation duration (CRD). CRD was corrected for the beat rate using Fridericia’s formula [CRDcF = CRD/(ISI/1000)^1/3].

Electro-Mechanical Window Analysis

The E-M window in hiPSC-CMs was calculated as previously described.⁵⁾ Briefly, the E-M window in hiPSC-CMs was calculated as the difference between the electrical and mechanical durations. FPD_end-cF, which was obtained from the FPs analysis, was used as the electrical duration. CRDcF, which was obtained from the motion vector analysis, was used as the mechanical duration. The E-M window in hiPSC-CMs was calculated as the difference between FPD_end-cF and CRDcF (E-M window = CRDcF–FPD_end-CF). The value of E-M window was shown as a ratio of 100% before the drug administeration.

Statistical Analysis

Results are presented as mean ± standard error of the mean (S.E.M.). One-way ANOVA, followed by Dunnett’s multiple comparison test, was used to analyze the data. Statistical significance was set at p < 0.05. Analyses were performed using GraphPad Prism 5.00 (GraphPad Software, La Jolla, CA, U.S.A.).

RESULTS

Effect of Positive or Negative Compound on E-M Window in hiPSC-CMs

To analyze the effect of a drug with a high TdP risk on the E-M window, we first examined the positive compound dofetilide using MEA assays and motion analysis in hiPSC-CMs. Figure 1A shows representative motion waveform of hiPSC-CMs before and after dofetilide administration. Dofetilide decreased the relaxation velocity from 4.10 ± 0.04 to 3.26 ± 0.17 µm/s at 1 nM and to 3.01 ± 0.15 µm/s at 3 nM in a concentration-dependent manner (Fig. 1C). In contrast, dofetilide had little effect on the contraction velocity (Fig. 1B). Moreover, dofetilide prolonged CRDcF from 0.637 ± 0.019 to 0.672 ± 0.020 s at 1 nM and to 0.770 ± 0.017 s at 3 nM in hiPSC-CMs (Fig. 1D). In the FPs analysis, dofetilide prolonged FPD_end-cF from 0.484 ± 0.004 to 0.554 ± 0.005 s at 1 nM and to 0.775 ± 0.013 s at 3 nM (Fig. 1E). Based on these data, dofetilide decreased the E-M window in a concentration-dependent manner by 25.2 ± 15.2% at 1 nM and 157.9 ± 14.0% at 3 nM in hiPSC-CMs (Fig. 1G). As 10 nM dofetilide induced EAD in hiPSC-CMs, we did not analyze the E-M window at a higher concentration.

Fig. 1. Effect of Dofetilide on the Contractility and E-M Window in hiPSC-CMs

(A) Representative motion waveform of hiPSC-CMs before and after dofetilide treatment. (B–D) The effect of dofetilide on a series of contractile parameters, including (B) contraction velocity, (C) relaxation velocity, and (D) contraction–relaxation duration (CRD)-cF. (E–G) The effect of dofetilide on (E) FPD_end-cF, (F) 1st peak amplitude, and (G) E-M window at 0.3, 1, and 3 nM in hiPSC-CMs. EAD was observed at 10 nM. All data represent mean ± S.E.M. (n = 3). * p < 0.05, ** p < 0.01, difference compared with 0 nM. Statistical comparison was performed using one-way ANOVA followed by Dunnett’s multiple comparison test.

Next, we assessed the effect of the negative compound aspirin on contractility and the E-M window in hiPSC-CMs. Figure 2A shows representative motion waveform of hiPSC-CMs before and after aspirin administration. As expected, aspirin had little effect on contractility parameters, including contraction velocity, relaxation velocity, and CRDcF at 1–100 µM (Figs. 2A–D). Moreover, aspirin had little effect on FPD_end-cF, 1st peak amplitude, and E-M window at 1–100 µM in hiPSC-CMs (Figs. 2E–G).

Fig. 2. Effect of Aspirin on the Contractility and E-M Window in hiPSC-CMs

(A) Representative motion waveform of hiPSC-CMs before and after aspirin treatment. (B–D) The effect of aspirin on a series of contractile parameters, including (B) contraction velocity, (C) relaxation velocity, and (D) contraction–relaxation duration (CRD)-cF. (E–G) The effect of aspirin on (E) FPD_end-cF, (F) 1st peak amplitude, and (G) E-M window at 1, 3, 10, 30, and 100 µM in hiPSC-CMs. All data represent mean ± S.E.M. (n = 3).

These data suggest that the E-M window of hiPSC-CMs can distinguish between drugs with high TdP risk and those with low TdP risk.

Effects of Bepridil, Ranolazine, or Mexiletine on E-M Window in hiPSC-CMs

As mentioned above, bepridil, ranolazine, and mexiletine were outliers among the CiPA 28 compounds in the CiPA international multi-site study using hiPSC-CMs.³⁾ We then assessed the effect of these drugs on E-M window in hiPSC-CMs. Figure 3A shows representative motion waveform of hiPSC-CMs before and after bepridil administration. Bepridil decreased the relaxation velocity from 7.75 ± 1.04 to 4.77 ± 0.66 µm/s at 3 µM (Fig. 3C), while it had little effect on contraction velocity at 0.1–3 µM (Fig. 3B). Moreover, bepridil prolonged CRDcF from 0.566 ± 0.004 to 0.599 ± 0.006 s at 1 µM, while it prolonged to 0.576 ± 0.007 s at 3 µM (Fig. 3D). In FPs analysis, bepridil prolonged FPD_end-cF from 0.473 ± 0.005 to 0.553 ± 0.004 s at 1 µM and to 0.662 ± 0.027 s at 3 µM (Fig. 3E), and decreased 1st peak amplitude by 33.8 ± 9.9% at 0.3 µM, 36.1 ± 2.6% at 1 µM, and 69.0 ± 10.9% at 3 µM (Fig. 3F). Based on MEA assay and motion vector analysis, bepridil decreased the E-M window by 50.3 ± 6.3% at 1 µM and 192.3 ± 6.1% at 3 µM (Fig. 3G). We did not analyze the E-M window of bepridil at concentrations of 10 µM or higher, because of the cardiac arrest.

Fig. 3. Effect of Bepridil on the Contractility and E-M Window in hiPSC-CMs

(A) Representative motion waveform of hiPSC-CMs before and after bepridil treatment. (B–D) The effect of bepridil on a series of contractile parameters, including (B) contraction velocity, (C) relaxation velocity, and (D) contraction–relaxation duration (CRD)-cF. (E–G) The effect of bepridil on (E) FPD_end-cF, (F) 1st peak amplitude, and (F) E-M window at 0.1, 0.3, 1, and 3 µM in hiPSC-CMs. Cardiac arrest was observed at 10 µM. All data represent mean ± S.E.M. (n = 3). * p < 0.05, ** p < 0.01, difference compared with 0 µM. Statistical comparison was performed using one-way ANOVA followed by Dunnett’s multiple comparison test.

Figure 4A shows representative motion waveform of hiPSC-CMs before and after ranolazine administration. Ranolazine decreased the relaxation velocity from 4.83 ± 0.10 to 4.23 ± 0.15 µm/s at 10 µM and to 3.77 ± 0.23 µm/s at 30 µM in a concentration-dependent manner (Fig. 4C). The contraction velocity and CRDcF did not change at 1–30 µM (Figs. 4B, D). In addition to FPs, ranolazine prolonged FPD_end-cF from 0.402 ± 0.009 to 0.433 ± 0.010 s at 10 µM and to 0.457 ± 0.013 s at 30 µM (Fig. 4E) and decreased 1st peak amplitude by 38.6 ± 4.6% at 3 µM, 56.4 ± 8.8% at 10 µM, and 82.4 ± 10.7% at 30 µM (Fig. 4F). Based on MEA assay and motion analysis, ranolazine slightly decreased the E-M window by 26.9 ± 7.9% at 30 µM in hiPSC-CMs (Fig. 4G). We did not analyze the E-M window using ranolazine at 100 µM, because of cardiac arrest.

Fig. 4. Effect of Ranolazine on the Contractility and E-M Window in hiPSC-CMs

(A) Representative motion waveform of hiPSC-CMs before and after ranolazine treatments. (B–D) The effect of ranolazine on a series of contractile parameters, including (B) contraction velocity, (C) relaxation velocity, and (D) contraction–relaxation duration (CRD)-cF. (E–G) The effect of ranolazine on (E) FPD_end-cF, (F) 1st peak amplitude, and (G) E-M window at 1, 3, 10, 30 µM in hiPSC-CMs. Cardiac arrest was observed at 100 µM. All data represent mean ± S.E.M. (n = 3). * p < 0.05, ** p < 0.01, difference compared with 0 µM. Statistical comparison was performed using one-way ANOVA followed by Dunnett’s multiple comparison test.

Figure 5A shows representative motion waveform of hiPSC-CMs before and after mexiletine administration. Mexiletine had little effect on a series of contractility parameters (contraction velocity, relaxation velocity, and CRDcF) at 0.3–10 µM (Figs. 5A–D). In FPs analysis, mexiletine slightly prolonged FPD_end-cF from 0.421 ± 0.013 to 0.448 ± 0.014 s at 10 µM (Fig. 5E) and decreased 1st peak amplitude by 33.5 ± 5.4% at 3 µM and 71.3 ± 6.9% at 10 µM (Fig. 5F). Based on MEA assay and motion analysis, mexiletine at 0.3–10 µM had little effect on the E-M window (Fig. 5G). We did not analyze the E-M window using mexiletine at 30 µM because of cardiac arrest.

Fig. 5. Effect of Mexiletine on the Contractility and E-M Window in hiPSC-CMs

(A) Representative motion waveform of hiPSC-CMs before and after mexiletine treatment. (B–D) The effect of mexiletine on a series of contractile parameters, including (B) contraction velocity, (C) relaxation velocity, and (D) contraction–relaxation duration (CRD)-cF. (E, F) The effect of mexiletine on (E) FPD_end-cF, (F) 1st peak amplitude, and (G) E-M window at 0.3, 1, 3, and 10 µM in hiPSC-CMs. Cardiac arrest was observed at 30 µM. All data represent mean ± S.E.M. (n = 3). * p < 0.05, ** p < 0.01 difference compared with 0 µM. Statistical comparison was performed using one-way ANOVA followed by Dunnett’s multiple comparison test.

Taken together, these data suggest that the E-M window in hiPSC-CMs can assess the TdP risk quantitatively.

Comparison of E-M Window in hiPSC-CMs with CiPA Study and hERG Assay

To investigate the predictability of the E-M window in hiPSC-CMs, we compared our E-M window data in hiPSC-CMs with the public TdP database CredibleMeds, hERG assay, CiPA study using hiPSC-CMs, and E-M window from in vivo studies (Table 1).

Table 1. Comparison of E-M Window in hiPSC-CMs with hERG Assay and Other Studies

Public TdP database: TdP risk classified by CredibleMeds. CiPA clinical TdP risk: Drugs were divided into three clinical TdP risk groups by the CiPA expert consensus.⁷⁾ hERG IC₅₀: The value was described with reference to the following studies.^15–17) CiPA international study using hiPSC-CMs: Risk were classified by three predictors (occurrence of arrhythmia-like events, prolongation of repolarization at maximum doses tested, and prolongation of repolarization at clinical exposure).³⁾ E-M window in hiPSC-CMs: The effect of drugs at 1 × C_max on the E-M window in hiPSC-CMs. E-M window in guinea pigs: A high risk indicates that the value, which is calculated by dividing the concentration of 10% of the E-M window decrease in guinea pigs by their therapeutic C_max, is under 30-fold (safety margin).¹⁰⁾

In the CredibleMeds, dofetilide and bepridil are classified as having a “known risk of TdP,” ranolazine is classified as having a “conditional risk of TdP,” and mexiletine is unlisted (Table 1). CiPA expert team classifies dofetilide and bepridil as high TdP-risk compounds and ranolazine and mexiletine as low TdP-risk compounds.⁷⁾ In the CiPA international multi-site study using hiPSC-CMs,³⁾ dofetilide was classified as a high-risk compound. In contrast, bepridil was classified as a low-risk compound and ranolazine was classified as an intermediate-risk compound, suggesting that these drugs were not adequately evaluated.

The IC₅₀ values of dofetilide, bepridil, ranolazine, and mexiletine against hERG have been shown to be approximately 0.03,¹⁵⁾ 0.16,¹⁵⁾ 11.5,¹⁶⁾ and 50 µM,¹⁷⁾ respectively. Our results showed that dofetilide and bepridil decreased the E-M window by 157.9% and 192.3% at 1 × therapeutics total C_max, respectively. In contrast, ranolazine slightly decreased the E-M window by 26.9% at 5 × C_max, whereas mexiletine had little effect at 1 × C_max (Fig. 6). Based on these results, dofetilide and bepridil were classified as high TdP-risk compounds by the E-M window in iPSC-CMs, whereas ranolazine and mexiletine were classified as low TdP-risk compounds (Table 1). Considering that the safety margin is generally set at 30-fold between the clinically free C_max and hERG IC₅₀,¹⁸⁾ these drugs are not considered to have a sufficient safety margin for the E-M window in hiPSC-CMs (dofetilide, 27; bepridil, 4.3; ranolazine, 10; and mexiletine, 13.1). Similar to our E-M window data, dofetilide and bepridil decreased the E-M window at the C_max, whereas ranolazine and mexiletine had little effect on the C_max in guinea pigs.¹⁰⁾

Fig. 6. Summary of the Change of E-M Window by Drugs

The effect of drugs on E-M window was examined in hiPSC-CMs. The value of E-M window was shown as a ratio of 100% before the drug was administered. The x-axis represented the value that the drug concentration in medium divided by the total C_max. The red line indicates high risk and the blue line indicated low risk.

Taken together, these data suggest that the E-M window of hiPSC-CMs can be used to predict the risk of TdP.

DISCUSSION

In this study, we assessed the effects of a positive compound, a negative compound, and several CiPA compounds, which were outliers in the hiPSC-CM validation study, on the E-M window in hiPSC-CMs using an MEA assay and motion analysis. Similar to previous in vivo studies, dofetilide, which has a high TdP risk, decreased the E-M window in hiPSC-CMs, whereas aspirin, which has a low TdP risk, had little effect. Bepridil, classified as a high TdP-risk compound in CiPA, decreased the E-M window in hiPSC-CMs, whereas ranolazine and mexiletine, which are classified as low TdP-risk compounds in CiPA, slightly decreased or had little effect on the E-M window of hiPSC-CMs. Thus, the E-M window in hiPSC-CMs can be used to classify the drugs as either high or low TdP-risk compounds.

Dofetilide at 1 and 3 nM reduced the E-M window in hiPSC-CMs by prolonging FPD_end-cF and slightly prolonging CRDcF (Fig. 1G). Dofetilide is known to inhibit hERG (IC₅₀,0.03 µM) but not I_CaL or I_Na (I_CaL IC₅₀, 26.7 µM; I_Na IC₅₀, 162.1 µM), and it prolongs FPD_end-cF by hERG inhibition.¹⁵⁾ As the hERG-selective inhibitor E-4031 prolonged CRD in hiPSC-CMs,¹⁴⁾ dofetilide prolonged CRDcF via hERG inhibition. Similar to our results, a previous study using guinea pigs reported that dofetilide prolonged QT interval by approximately 20–40% at 1–10 nM and slightly prolonged contraction duration by approximately 7%; consequently, the E-M window was reduced.¹⁹⁾ Thus, the reduction of the E-M window in hiPSC-CMs by dofetilide can be used to assess the TdP risk of dofetilide.

Bepridil (3 µM) reduced the E-M window in hiPSC-CMs (Fig. 3G). In addition to hERG inhibition (IC₅₀, 0.16 µM), bepridil was reported to inhibit I_CaL and I_Na (I_CaL IC₅₀, 1.0 µM; I_Na IC₅₀, 2.3 µM).¹⁵⁾ Similar to our data, bepridil has been reported to reduce the E-M window in guinea pigs.¹⁰⁾ The reduction in the E-M window by bepridil was due to the prolongation of FPD_end-cF and shortening of CRDcF at 3 µM. Bepridil has been shown to prolong the QT interval at C_max in guinea pigs.¹⁰⁾ However, the increase in intracellular Ca²⁺ concentrations may contribute to the shortening of the CRD.^14,20) As bepridil inhibits the forward mode Na⁺/Ca²⁺ exchanger (Ca²⁺ exit mode) and results in a subsequent increase in intracellular Ca²⁺,^21,22) the shortening of CRD may be caused by an increase in intracellular Ca²⁺ concentration. Motion analysis did not measure contraction force and evaluated contraction in an indirect manner.¹⁴⁾ Further studies are required to perform simultaneous recordings of motion analysis and intracellular Ca²⁺ levels to validate the data.

Ranolazine (30 µM) slightly reduced the E-M window in hiPSC-CMs (Fig. 4G). Ranolazine is a multi-ion channel inhibitor that inhibits hERG (IC₅₀, 11.5 µM) and late I_Na (IC₅₀, 5.9 µM), but not I_CaL (IC₅₀, 296 µM).¹⁶⁾ QT prolongation by ranolazine in humans is thought to be determined by the balance between QT prolongation by hERG inhibition and QT shortening by late I_Na inhibition. However, the late I_Na current in hiPSC-CMs is considered smaller than that in the human adult heart, suggesting a difference between hiPSC-CMs and adult hearts.²³⁾ QT interval prolongation was observed at the C_max in guinea pigs. Contraction duration was also prolonged.¹⁹⁾ Ranolazine has been reported to lower intracellular Na⁺ concentrations via late I_Na current inhibition and prevent Ca²⁺ influx into the cell by the Na⁺/Ca²⁺ exchanger.²⁴⁾ Previous studies have shown that the introduction of cardiac inwardly rectifying potassium current I_K1 induces the maturation of hiPSC-CMs and increases late I_Na currents.^25,26) Therefore, it is necessary to investigate the effects of ranolazine on contraction duration using mature hiPSC-CMs.

Mexiletine at 10 µM prolonged FPD_end-cF in hiPSC-CMs without altering the E-M window (Fig. 5G). Mexiletine is a potent inhibitor of hERG (IC₅₀, 50 µM) and late I_Na (IC₅₀, 17.6 µM).^17,27) Both FPD_end-cF and CRDcF were considered prolonged by hERG inhibition,^14,17) and the total E-M window did not change. In contrast, QT interval prolongation was not observed with mexiletine at the C_max in guinea pigs.¹⁰⁾ Prolongation of the contraction duration was induced by mexiletine at C_max in guinea pigs,¹⁰⁾ suggesting that the E-M window was increased in guinea pigs. The difference between hiPSC-CMs and guinea pigs might be due to the late I_Na channels.^23,24)

The CiPA clinical risk assessment team classified dofetilide and bepridil as high TdP-risk compounds and ranolazine and mexiletine as low-risk compounds.⁷⁾ Safety margins of 30-fold are generally calculated by dividing the hERG IC₅₀ values by the therapeutic free C_max.¹⁸⁾ However, the safety margins for the four drugs in the study were smaller than 30-fold, suggesting the importance of multi-ion channel assays such as CiPA.^15–17) We found that the E-M window in hiPSC-CMs can classify high and low risks of TdP. Thus, the E-M window of hiPSC-CMs may be a useful marker for multi-ion channel inhibitors. As we analyzed a limited number of compounds, future studies are required to examine whether other CiPA compounds can be classified as high- or low-risk compounds.

Dofetilide at 1 × C_max has been reported to prolong QT interval without a change of contraction duration in guinea pigs, which results in decrease in E-M window.¹⁹⁾ Bepridil at 5 × C_max has been reported to prolong QT interval without a change of contraction duration in guinea pigs, which results in decrease in E-M window.¹⁰⁾ E-M window in guinea pigs were similar with that in hiPSC-CMs. In contrast, ranolazine at 1 × C_max did not affect E-M window in guinea pigs by both QT prolongation and increase in Q-wave to the end of left ventricular pressure signal.¹⁹⁾ We did not observe the effects of ranolazine on CRDcF in hiPSC-CMs. Inhibition of late I_Na is considered to reduce intracellular sodium and attenuate Na⁺-Ca²⁺ exchanger-related increase in diastolic Ca.²⁴⁾ The difference between guinea pigs and hiPSC-CMs might be late I_Na. Due to the immature property, hiPSC-CMs are not sufficiently mature to detect TdP risks associated with inhibition of the late sodium current.³⁾ Future studies will be required to make hiPSC-CM mature by several approaches, such as 3D model and co-culture.²⁸⁾

One of the limitations of the E-M window is the occurrence of beating arrest in hiPSC-CMs. When hiPSC-CMs stopped beating, the FPD_end-cF was not calculated. Bepridil, ranolazine, and mexiletine caused cell cycle arrest at 10, 100, and 30 µM, respectively. It is difficult to examine higher concentrations of these drugs, up to a 30-fold safety margin. Previous studies have reported that cardiac arrest in hiPSC-CMs occurs primarily with a decrease in 1st peak amplitude of FPs.^4,5) We found that bepridil, mexiletine, and ranolazine decreased 1st peak amplitude, resulting in cardiac arrest. The mechanism by which drugs induce cardiac arrest in hiPSC-CMs is unclear. Further studies are required to examine the effects of these drugs on mature hiPSC-CMs.

In conclusion, our results showed that the E-M window, calculated by the MEA assay and motion analysis, improved the predictability of TdP risks by CiPA compounds in hiPSC-CMs. Thus, the E-M window of hiPSC-CMs can be useful in assessing drug-induced TdP risk in non-clinical settings.

Acknowledgments

This study was supported by the Research on Regulatory Science of Pharmaceuticals and Medical Devices of the Japan Agency for Medical Research and Development (AMED) (JP21mk0101189 to Y.K.), a JSPS KAKENHI Grant (21H02634 to Y.K.), and a grant from the Smoking Research Foundation (Y.K.).

Conflict of Interest

The authors declare no conflict of interest.

REFERENCES

• 1) Cook D, Brown D, Alexander R, March R, Morgan P, Satterthwaite G, Pangalos MN. Lessons learned from the fate of AstraZeneca’s drug pipeline: a five-dimensional framework. Nat. Rev. Drug Discov., 13, 419–431 (2014).
• 2) Stockbridge N, Morganroth J, Shah RR, Garnett C. Dealing with global safety issues: was the response to QT-liability of non-cardiac drugs well coordinated? Drug Saf., 36, 167–182 (2013).
• 3) Blinova K, Dang Q, Millard D, et al. International multisite study of human-induced pluripotent stem cell-derived cardiomyocytes for drug proarrhythmic potential assessment. Cell Reports, 24, 3582–3592 (2018).
• 4) Ando H, Yoshinaga T, Yamamoto W, Asakura K, Uda T, Taniguchi T, Ojima A, Shinkyo R, Kikuchi K, Osada T, Hayashi S, Kasai C, Miyamoto N, Tashibu H, Yamazaki D, Sugiyama A, Kanda Y, Sawada K, Sekino Y. A new paradigm for drug-induced torsadogenic risk assessment using human iPS cell-derived cardiomyocytes. J. Pharmacol. Toxicol. Methods, 84, 111–127 (2017).
• 5) Yanagida S, Satsuka A, Hayashi S, Ono A, Kanda Y. Comprehensive cardiotoxicity assessment of COVID-19 treatments using human-induced pluripotent stem cell-derived cardiomyocytes. Toxicol. Sci., 183, 227–239 (2021).
• 6) Kanda Y, Yamazaki D, Osada T, Yoshinaga T, Sawada K. Development of torsadogenic risk assessment using human induced pluripotent stem cell-derived cardiomyocytes: Japan iPS Cardiac Safety Assessment (JiCSA) update. J. Pharmacol. Sci., 138, 233–239 (2018).
• 7) Colatsky T, Fermini B, Gintant G, Pierson JB, Sager P, Sekino Y, Strauss DG, Stockbridge N. The Comprehensive in Vitro Proarrhythmia Assay (CiPA) initiative—update on progress. J. Pharmacol. Toxicol. Methods, 81, 15–20 (2016).
• 8) van der Linde HJ, Van Deuren B, Somers Y, Loenders B, Towart R, Gallacher DJ. The electro-mechanical window: a risk marker for Torsade de Pointes in a canine model of drug induced arrhythmias. Br. J. Pharmacol., 161, 1444–1454 (2010).
• 9) Guns PJ, Johnson DM, Van Op Den Bosch J, Weltens E, Lissens J. The electro-mechanical window in anaesthetized guinea pigs: A new marker in screening for Torsade de Pointes risk. Br. J. Pharmacol., 166, 689–701 (2012).
• 10) Morissette P, Regan C, Fitzgerald K, Gerenser P, Travis J, Wang S, Fanelli P, Sannajust F. Shortening of the electromechanical window in the ketamine/xylazine-anesthetized guinea pig model to assess pro-arrhythmic risk in early drug development. J. Pharmacol. Toxicol. Methods, 81, 171–182 (2016).
• 11) Shah RR, Maison-Blanche P, Duvauchelle T, Robert P, Denis E. Establishing assay sensitivity in QT studies: experience with the use of moxifloxacin in an early phase clinical pharmacology study and comparison with its effect in a thorough QT study. Eur. J. Clin. Pharmacol., 71, 1451–1459 (2015).
• 12) Yanagida S, Satsuka A, Hayashi S, Ono A, Kanda Y. Chronic cardiotoxicity assessment of BMS-986094, a guanosine nucleotide analogue, using human iPS cell-derived cardiomyocytes. J. Toxicol. Sci., 46, 359–369 (2021).
• 13) Kanda Y, Satsuka A, Hayashi S, Hagiwara-Nagasawa M, Sugiyama A. Assessment of contractility in human iPS cell-derived cardiomyocytes using motion vector analysis. Methods Mol. Biol., 2320, 151–160 (2021).
• 14) Hayakawa T, Kunihiro T, Ando T, Kobayashi S, Matsui E, Yada H, Kanda Y, Kurokawa J, Furukawa T. Image-based evaluation of contraction-relaxation kinetics of human-induced pluripotent stem cell-derived cardiomyocytes: Correlation and complementarity with extracellular electrophysiology. J. Mol. Cell. Cardiol., 77, 178–191 (2014).
• 15) Kramer J, Obejero-Paz CA, Myatt G, Kuryshev YA, Bruening-Wright A, Verducci JS, Brown AM. MICE models: superior to the HERG model in predicting torsade de pointes. Sci. Rep., 3, 2100 (2013).
• 16) Antzelevitch C, Burashnikov A, Sicouri S, Belardinelli L. Electrophysiologic basis for the antiarrhythmic actions of ranolazine. Heart Rhythm, 8, 1281–1290 (2011).
• 17) Mirams GR, Cui Y, Sher A, Fink M, Cooper J, Heath BM, McMahon NC, Gavaghan DJ, Noble D. Simulation of multiple ion channel block provides improved early prediction of compounds’ clinical torsadogenic risk. Cardiovasc. Res., 91, 53–61 (2011).
• 18) Redfern WS, Carlsson L, Davis A, Lynch WG, Mackenzie I, Palethorpe S, Siegl P, Strang I, Sullivan A, Wallis R, Camm AJ, Hammond TG. 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., 58, 32–45 (2003).
• 19) Skinner M, Hale E, Ceuppens P, Pollard C. Differentiating multichannel block on the guinea pig ECG: use of T_peak-T_end and J-T_peak. J. Pharmacol. Toxicol. Methods, 111, 107085 (2021).
• 20) Ton AT, Nguyen W, Sweat K, Miron Y, Hernandez E, Wong T, Geft V, Macias A, Espinoza A, Truong K, Rasoul L, Stafford A, Cotta T, Mai C, Indersmitten T, Page G, Miller PE, Ghetti A, Abi-Gerges N. Arrhythmogenic and antiarrhythmic actions of late sustained sodium current in the adult human heart. Sci. Rep., 11, 12014 (2021).
• 21) Hu HJ, Wang SS, Wang YX, Liu Y, Feng XM, Shen Y, Zhu L, Chen HZ, Song M. Blockade of the forward Na⁺ /Ca²⁺ exchanger suppresses the growth of glioblastoma cells through Ca²⁺ -mediated cell death. Br. J. Pharmacol., 176, 2691–2707 (2019).
• 22) Watanabe Y, Kimura J. Blocking effect of bepridil on Na⁺ /Ca²⁺ exchange current in guinea pig cardiac ventricular myocytes. Jpn. J. Pharmacol., 85, 370–375 (2001).
• 23) Lemoine MD, Mannhardt I, Breckwoldt K, Prondzynski M, Flenner F, Ulmer B, Hirt MN, Neuber C, Horváth A, Kloth B, Reichenspurner H, Willems S, Hansen A, Eschenhagen T, Christ T. Human iPSC-derived cardiomyocytes cultured in 3D engineered heart tissue show physiological upstroke velocity and sodium current density. Sci. Rep., 7, 5464 (2017).
• 24) Gupta T, Khera S, Kolte D, Aronow WS, Iwai S. Antiarrhythmic properties of ranolazine: a review of the current evidence. Int. J. Cardiol., 187, 66–74 (2015).
• 25) Vaidyanathan R, Markandeya YS, Kamp TJ, Makielski JC, January CT, Eckhardt LL. IK1-enhanced human-induced pluripotent stem cell-derived cardiomyocytes: An improved cardiomyocyte model to investigate inherited arrhythmia syndromes. Am. J. Physiol. Heart Circ. Physiol., 310, H1611–H1621 (2016).
• 26) Li M, Kanda Y, Ashihara T, Sasano T, Nakai Y, Kodama M, Hayashi E, Sekino Y, Furukawa T, Kurokawa J. Overexpression of KCNJ2 in induced pluripotent stem cell-derived cardiomyocytes for the assessment of QT-prolonging drugs. J. Pharmacol. Sci., 134, 75–85 (2017).
• 27) Gao Y, Xue X, Hu D, Liu W, Yuan Y, Sun H, Li L, Timothy KW, Zhang L, Li C, Yan GX. Inhibition of late sodium current by mexiletine: a novel pharmotherapeutical approach in timothy syndrome. Circ. Arrhythm. Electrophysiol., 6, 614–622 (2013).
• 28) Ahmed RE, Anzai T, Chanthra N, Uosaki H. A brief review of current maturation methods for human induced pluripotent stem cells-derived cardiomyocytes. Front. Cell Dev. Biol., 8, 178 (2020).