Action potential-based MEA platform for in vitro screening of drug-induced cardiotoxicity using human iPSCs and rat neonatal myocytes

Accepted Manuscript Action potential-based MEA platform for in vitro screening of drug-induced cardiotoxicity using human iPSCs and rat neonatal myocytes

Danny Jans, Geert Callewaert, Olga Krylychkina, Luis Hoffman, Francesco Gullo, Dimiter Prodanov, Dries Braeken PII: DOI: Reference:

S1056-8719(17)30045-X doi: 10.1016/j.vascn.2017.05.003 JPM 6450

To appear in:

Journal of Pharmacological and Toxicological Methods

Received date: Revised date: Accepted date:

21 February 2017 10 May 2017 20 May 2017

Please cite this article as: Danny Jans, Geert Callewaert, Olga Krylychkina, Luis Hoffman, Francesco Gullo, Dimiter Prodanov, Dries Braeken , Action potential-based MEA platform for in vitro screening of drug-induced cardiotoxicity using human iPSCs and rat neonatal myocytes, Journal of Pharmacological and Toxicological Methods (2017), doi: 10.1016/j.vascn.2017.05.003

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Action potential-based MEA platform for in vitro screening of drug-induced cardiotoxicity using human iPSCs and rat neonatal myocytes

Danny Jans, Geert Callewaert, Olga Krylychkina, Luis Hoffman, Francesco Gullo, Dimiter

PT

Prodanov, and Dries Braeken

[email protected]

AC

CE

PT E

D

MA

NU

SC

003216288046

RI

IMEC, Kapeldreef 75, B3000 Leuven

1

ACCEPTED MANUSCRIPT Abstract

Drug-induced cardiotoxicity poses a negative impact on public health and drug development. Cardiac safety pharmacology issues urged for the preclinical assessment of drug-induced ventricular arrhythmia leading to the design of several in vitro electrophysiological screening assays. In general, patch clamp systems allow for intracellular recordings, while multi-

PT

electrode array (MEA) technology detect extracellular activity. Here, we demonstrate a complementary metal oxide semiconductor (CMOS)-based MEA system as a reliable platform for non-invasive, long-term intracellular recording of cardiac action potentials at

RI

high resolution. Quinidine (8 concentrations from 10-7 to 2.10-5 M) and verapamil (7

SC

concentrations from 10-11 to 10-5 M) were tested for dose-dependent responses in a network of cardiomyocytes. Electrophysiological parameters, such as the action potential duration

NU

(APD), rates of depolarization and repolarization and beating frequency were assessed. In hiPSC, quinidine prolonged APD with EC50 of 2.2 10-6 M. Further analysis indicated a multifactorial action potential prolongation by quinidine: (1) decreasing fast repolarization

MA

with IC50 of 1.1 10-6 M; (2) reducing maximum upstroke velocity with IC50 of 2.6 10-6 M; and (3) suppressing spontaneous activity with EC50 of 3.8 10-6 M. In rat neonatal cardiomyocytes,

D

verapamil blocked spontaneous activity with EC50 of 5.3 10-8 M and prolonged the APD with EC50 of 2.5 10-8 M. Verapamil reduced rates of fast depolarization and repolarization with

PT E

IC50s of 1.8 and 2.2 10-7 M, respectively. In conclusion, the proposed action potential-based MEA platform offers high quality and stable long-term recordings with high information content allowing to characterize multi-ion channel blocking drugs. We anticipate application

Keywords

AC

safety.

CE

of the system as a screening platform to efficiently and cost-effectively test drugs for cardiac

cardiotoxicity; action potential; quinidine; verapamil; CMOS-MEA; intracellular recording; hiPSC

2

ACCEPTED MANUSCRIPT Introduction

Cardiac toxicity is an unfortunate side effect of several drug compounds increasing the risk for morbidity and mortality. Furthermore, discontinuation of approval or withdrawal of these drugs for clinical use imposes financial drawbacks to pharmaceutical companies. To improve

PT

drug performance and reduce costs for drug development, new methods that screen for

RI

cardiotoxic effects early in the discovery process have been introduced. Initially, the QT

SC

interval attracted considerable attention because of the proarrhythmic potential of some drugs as a result of the promiscuous behavior of Kv11.1 channels encoded by the human ether-à-

NU

go-go-related gene (hERG) (Sanguinetti and Tristani-Firouzi, 2006). These ion channels mediate the rapidly activating delayed rectifying potassium current (IKr) critical for cardiac

MA

repolarization. Mutations in hERG or drug-induced block of channel activity prolong the QT interval inducing the risk for developing fatal arrhythmia. In clinical settings a possibly life-

D

threatening disorder known as torsades de pointes may show up (Roden, 2004). At the

PT E

cellular level the QT interval can be assessed via the ventricular action potential duration (APD). Hence, in vitro measurement of either hERG activity or APD prolongation was

CE

accepted as a preclinical screening assay for estimating proarrhythmic risk of new drugs (Shah, 2002). Though the manual patch clamp technique offers the highest quality to evaluate

AC

these parameters, the assay is technically challenging and suffers from low throughput (Zhang et al., 2014). In addition, dilution of diffusible intracellular substances and buffering of intracellular ion concentrations occur during whole-cell recordings thereby affecting ion currents, especially Ca2+ currents, and severely limiting long-term studies (Inayat et al., 2013) (Hamill et al., 1981). Automated patch clamp technology exploited the whole-cell recording principle in a high-throughput system. Therefore, its applications mainly focused on noncardiac cell types expressing hERG channels heterologously (Möller and Witchel, 2011). The

3

ACCEPTED MANUSCRIPT technology focused on analyzing ion currents in single suspended cells (Franz et al., 2017), but struggled to analyze cells in a network. Two distinct approaches have been accomplished, of which the planar patch clamp technology established a higher success then the pipettebased system. Recently, voltage-gated Na+ and Ca2+ currents were successfully recorded from human iPSCs using a planar patch clamp platform (Obergrussberger et al., 2016).

PT

Alternatively, multi-electrode array (MEA)-based measurements were promoted successfully

RI

as an accessible system to measure APD. MEA systems are non-invasive, have medium

SC

throughput capabilities and extract basic electrophysiological parameters, such as beating frequency and field potential duration (Natarajan et al., 2011). Recent models combined

NU

impedance recordings at rapid data acquisition rates to evaluate contractility in addition to electrical activity (Obergrussberger et al., 2016; Millard et al., 2016). MEA assays handle

MA

cell populations and deliver accurate extracellular electrical signals but often lack the sensitivity to analyze the complex effects of drugs on cardiac electrophysiology.

D

In this paper, we demonstrate a ‘complementary metal oxide semiconductor (CMOS)’-based

PT E

MEA chip harboring a high density of subcellular-sized microelectrodes that interface with individual cardiac cells in a cellular network. Electrical stimulation using the electrode

CE

underneath a selected cell creates small (nano)pores in the membrane of the attached cell. The so-called ‘nanoporated-cell state’ allows to record intracellular potential changes at high

AC

resolution (Braeken et al., 2012). Because of its minimal invasiveness, the procedure is reproducible on the same cell and, therefore, allows for long-term recordings. In addition to key electrophysiological parameters, such as APD and beating frequency, these intracellular recordings allow to quantify the rates of fast depolarization and repolarization. To explore these features two compounds with multi-ion channel blocking properties, quinidine and verapamil, were tested using hiPSC and neonatal rat cardiomyocytes, respectively. Dose response curves based on the analyzed key parameters provide important insights into the

4

ACCEPTED MANUSCRIPT mechanisms by which both drugs may affect cardiac electrophysiology. We envision application of the CMOS-MEA as a screening platform that can be applied efficiently without intensive training to cost-effectively scan compounds for toxic behavior on cardiac cells.

RI

PT

Materials and methods

SC

Human iPSC-derived cardiomyocytes

Human-induced pluripotent stem cell (hiPSC)-derived cardiomyocytes (iCell®) and cell

NU

culture media (iCell Cardiomyocyte Maintenance Medium) were obtained from Cellular Dynamics International (CDI, Madison, WI, USA). Cells were thawed and handled according the

manufacturer’s

user

guide

(https://cellulardynamics.com/assets/CDI_iCell

MA

to

Cardiomyocytes2_UG.pdf) with the exception for coating where fibronectin (50 µg/ml)

D

replaced gelatin (adapted from Gibson et al., 2014). The cells were plated at 20K cells per

PT E

MEA and maintained at 37°C, 5% CO2, and 95% humidity. Cells began to beat spontaneously within 24–48 h of plating and by day 6 the entire population formed an

AC

days after plating.

CE

electrically connected syncytial layer. Electrophysiological recordings were performed 5 to 7

Rat primary neonatal cardiomyocytes Neonatal ventricular myocytes were harvested from two day old Wistar rats. Extracted ventricles were cut in small pieces, washed in Hanks balanced salt solution (HBSS) without Ca2+ and Mg2+ and incubated in trypsin (0.05%) overnight at 4°C (adapted from Louch et al., 2011). Thereafter, collagenase (Collagenase, Type II, ThermoFisher Scientific, 1 mg/mL, in HBSS) at 37°C for 15 min and intermittent mechanical trituration was used to digest the

5

ACCEPTED MANUSCRIPT tissue and release the cardiomyocytes from the tissues. The cell suspension was filtered through 40µm cell strainers and centrifuged at 300g for 5 min. The cell pellet was resuspended in 6% BSA in HBSS and again centrifuged at 8g for 20 min. The supernatant was discarded and the cell pellet was re-suspended in culture medium (Ham F10 containing 5% FCS, 1% PenStrep, Hepes, 0.5% ITS, 0.1 mM Norepinephrine, and 2 µg/ml vitamin B12).

PT

Cells were pre-plated in a tissue culture flask (T-75) for 90-120 min to allow for selective

RI

attachment of remaining fibroblasts. Finally, non-attached cells were centrifuged for 5 min at

SC

200g, re-suspended in cell culture medium with the optional addition of 10µM AraC and seeded on fibronectin (5 µg/ml) coated MEA substrates at a density of approximately 1000

NU

cells/mm2.

MA

CMOS-MEA

CMOS-MEA chips were designed and fabricated in Taiwan Semiconductor Manufacturing

D

Company (TSMC) using standard 0.18 µm 4-metal CMOS technology on 8-inch wafers

PT E

(Huys et al., 2012). Whereas conventional MEA systems typically implement a small number (<256) of large electrodes (>10 µm) and the recorded signals are amplified externally

CE

(passive system), the CMOS-MEA used in this study is an arrangement of 16,384 TiN subcellular-sized 3D-electrodes (1-4 µm in diameter)

that are circuited to individual

AC

amplifiers underneath (active system) (Fig. 1). Each electrode on the chip is addressable individually for electrical recording or stimulation. Chips were diced and packaged on a printed circuit board. A glass ring was glued on top of the substrate forming a culture chamber. Custom software was developed to allow selection of the electrodes and fast switching between recording and stimulation modes using the on-chip circuitry. Temporal intracellular access, referred to as the nanoporated-cell state, was created by applying small

6

ACCEPTED MANUSCRIPT voltage pulses that cause electroporation of a small membrane patch. All data were sampled

D

MA

NU

SC

RI

PT

at 22 kHz and stored in binary files.

PT E

Figure 1. CMOS-MEA chip. Upper left picture shows a CMOS-MEA chip with active area (blue dashed outline) and culture chamber on top; the upper right picture depicts an electron micrograph of

3D-

microelectrodes in the active area. The lower schematic illustrates a cardiac myocyte adhered to a 3D-

CE

microelectrode with integrated amplifier/filter-circuitry underneath. A close‐up is displayed (red dashed outline)

AC

illustrating the electrical equivalent circuit model for the nanoporated-cell state.

Drugs, Drug Assay and Data Analysis Quinidine anhydrous (Sigma, Q3625, Belgium) was prepared as a stock solution of 100 mM in DMSO. Verapamil hydrochloride (Sigma, V4629, Belgium) was dissolved in ethanol at a concentration of 100 mM. Drug dilutions and final working solutions were prepared in cellspecific cardiac medium. At the highest drug concentration, DMSO and ethanol were diluted

7

ACCEPTED MANUSCRIPT 5,000 and 10,000 times, respectively. Dilutions up to 1,000 had no effect on cardiac electrical activity (data not shown). Cells were continuously perfused with preheated solutions at 35°C using a temperature control system (Multichannel Systems, Germany). Before treatment with test compounds, intracellular recordings were obtained for each individual cell. Test compounds at the appropriate concentrations were then added and 10 minutes after exposure

PT

a second series of nanoporated-cell state was established. A negative control was included to

RI

check stability of the nanoporated-cell state over time.

SC

Intracellular recordings were aligned, averaged and analyzed using custom written MatLab scripts. To generate EC50 or IC50 values, normalized data were fitted to a logistic function

NU

(Graphpad Prism equation Y=Bottom + (Top - Bottom) / (1+10^((LogEC50 or LogIC50 -X) *

MA

Hillslope)).

Statistical Methods

D

Data are presented as mean ± standard deviation (SD) and ‘n’ denotes the number of cells

AC

CE

PT E

tested.

8

ACCEPTED MANUSCRIPT Results

Extracellular recordings from a single hiPSC on top of a selected electrode yielded inputreferred amplitudes of about 2 mV (Fig. 2A). Next, a voltage pulse sequence was used to locally electroporate the cell membrane, creating the nanoporated-cell state. In this state, a

PT

low-impedance connection is made to the intracellular space, enabling measurement of action

RI

potential waveforms with input-referred amplitudes of 25 ± 5 mV (n=20) (Fig. 2A). Success

SC

rates in obtaining intracellular recordings were high. For all 30 cells the nanoporated-cell state was successfully established at the first stimulation, 3 cells failed at the second

NU

stimulation, and all remaining cells but two were responsive to further electrical stimulation up to ten times with interval periods of 10 minutes (Fig. 2C, left panel). Although the

MA

nanoporated-cell state is short-lasting (93 ± 50 s, n=20), the ability to repetitively re-establish the nanoporated-cell state over extended periods of time (up to several days) affords the

D

possibility of studying long-term effects and is as such of great importance to drug screening.

PT E

To illustrate the stability of the intracellular recordings, the APD was evaluated in 5 cells over a period of 48 hrs. For each cell, action potentials were first monitored in the

CE

nanoporated-cell state every 15 minutes during a 1 hour period. Cells were then transferred back to culture conditions followed by a next examination at 24 and 48 hrs. Over time, no

AC

statistically significant differences in APD were observed (Fig. 2C, right panel). A plot illustrating the average action potential waveform recorded from a single hiPSC cell in the nanoporated-cell state is shown in figure 2B. To create the plot, intracellular recordings were time-aligned by their peaks prior to averaging. The grey shaded areas at either side of the action potential trajectory reflect standard errors. To characterize the electrical activity of the cardiac cells, the following parameters were quantified: duration of action potential (APD),

9

ACCEPTED MANUSCRIPT maximum rate of intracellular potential change during upstroke and final repolarization, and

PT E

D

MA

NU

SC

RI

PT

beating frequency.

Figure 2. CMOS-MEA intracellular recordings of hiPSCs. (A) Example of extracellular (grey) and intracellular recordings (black) from a single cardiomyocyte. Intracellular recording was obtained by applying

CE

biphasic voltage pulses (10- ms pulses of -1.65/+1.65 V at a frequency of 100 Hz – marked in green). (B) Averaged action potential waveform ± SD (grey shaded areas either side of the action potential trajectory) from

AC

a single cardiomyocyte. Action potential parameters analyzed included duration (3) and rates of intracellular potential changes (1 and 2). (C) Success rates and stability of intracellular recordings. Left panel: the firstattempt success rate was 100% and slightly declined with subsequent attempts (85% for ≥ 3 attempts). Right panel: action potential duration evaluated in a single cardiomyocyte over a period of 48 hours.

Dose-response to quinidine in hiPSC cardiomyocytes In a first approach to evaluate the applicability of the system for cardiac safety drug screening, hiPSC cardiomyocytes were subjected to quinidine, a well-characterized class 1

10

ACCEPTED MANUSCRIPT antiarrhythmic compound. Initially, a batch of 60 cells was randomly selected and each of them was analyzed in the nanoporated-cell state (control trace in Fig. 3). Next, quinidine was administered using increasing concentrations, ranging from 10-7 to 2.10-5 M. For each concentration, six cells were randomly chosen and tested in the nanoporated-cell state. Figure 3 (left panel) illustrates averaged traces of the intracellular recordings of five individual adult

PT

human iPSC cardiomyocytes at each concentration of quinidine. Quinidine caused

RI

concentration-dependent changes in action potential morphology and duration. Analysis of

SC

the dose-response curves yielded half-maximal concentrations of 1.1 and 2.6 10-6 M for the decrease in the rates of repolarization and depolarization, respectively (Fig. 3, right panel)

NU

and 2.2 and 3.8 10-6 M for the lengthening of the action potential and suppression of spontaneous beating frequency, respectively (Fig. 3, middle panel). In 5 out of 60 cells

CE

PT E

D

MA

analyzed, quinidine (2.10-7 M) induced EADs (data not shown).

Figure 3. Concentration-dependent effects of quinidine on action potentials recorded from hiPSC

AC

cardiomyocytes. Action potentials were recorded in control conditions (ctr) and in the presence of increasing concentrations of quinidine. Action potential parameters, including duration and rate of intracellular potential change, and spontaneous beating frequency were measured throughout the entire experiment. Data are represented as mean ± SD. EC50 or IC50 values are indicated by vertical lines.

Dose-response to verapamil in neonatal rat cardiomyocytes We further validated the CMOS-MEA screening platform using spontaneously beating rat neonatal cardiomyocytes in the presence of noradrenaline. For these experiments, the cells

11

ACCEPTED MANUSCRIPT were exposed to increasing concentrations of verapamil, a well-characterized class 4 antiarrhythmic compound. The same procedure was followed as outlined for quinidine. About 55 cells were randomly picked and intracellular recordings were obtained in the absence of the drug (control trace in Fig. 4). Next, verapamil was administered using increasing concentrations, ranging from 10-11 to 10-5 M. For each concentration, six cells

PT

were randomly chosen and tested in the nanoporated-cell state. Figure 4 (left panel) shows

RI

the averaged intracellular recordings of five individual rat neonatal cardiomyocytes.

of 5.3 10-8 M) (Fig. 4, middle panel)

SC

Verapamil decreased the beating rate at concentrations between 10-9 and 10-5 M (EC50 value and spontaneous activity was abolished at

NU

concentrations >10-5 M. The decrease in beating rate was associated with a lengthening of the action potential (EC50 values of 2.5 10-8 M) (Fig. 4, middle panel). At concentrations >10-8 M

MA

verapamil also reduced the rates of depolarization and repolarization (IC50 values of 1.8 and 2.2 10-7 M, respectively) (Fig. 4, right panel). Whole-cell patch-clamp recordings of an

AC

CE

PT E

beating rate (data not shown).

D

individual rat cardiomyocyte confirmed verapamil-induced APD lengthening and decrease in

Figure 4. Concentration-dependent effects of verapamil on action potential recorded from neonatal rat cardiomyocytes. Action potentials were recorded in control conditions (ctr) and in the presence of increasing concentrations of verapamil. Action potential parameters, including duration and rates of intracellular potential change, and spontaneous beating frequency were measured throughout the entire experiment. Data are

12

ACCEPTED MANUSCRIPT represented as mean ± SD. On the dose-response relationships the EC50 and IC50 values are indicated by vertical lines.

Discussion

PT

The current study demonstrates an innovative CMOS-based MEA platform to screen drugs for their impact on cardiac electrical activity. The system is based on the MEA principle in

RI

which flat electrodes capture extracellular electrical potential changes of cells cultured atop.

SC

The novelty lies in (a) the design of a dense array of subcellular-sized electrodes allowing to

NU

electrically communicate with a single cell in a cellular network, and (b) the integrated circuitry under each electrode, which establishes a two-way electrical communication with

MA

each cell enabling stimulation and recording with the same electrode. Notably, electrical stimulation of the cell on top of an electrode greatly reduces access resistance between cell

D

and electrode allowing high-resolution intracellular voltage recordings referred to as

PT E

‘intracellular’ recordings (Braeken et al., 2012). The procedure to obtain ‘intracellular’ recordings is simple, has no discernible effects on cell homeostasis and generates

CE

reproducible data. The major aim of the present study was to evaluate the capability of this platform to detect dose-dependent drug effects on myocardial action potentials. Two

AC

clinically documented drugs, quinidine and verapamil, were carefully analyzed in hiPSC and neonatal rat myocytes, respectively, for spontaneous beating frequency, action potential duration, and maximum rates of depolarization and repolarization. In hiPSC, 1 μM quinidine significantly reduced the rate of final repolarization and prolonged the action potential, while a reduction of the rate of fast depolarization and the beating frequency became only evident at higher concentrations. The EC50 value of 2.2 10-6 M obtained for the APD prolongation is in line with that reported for the prolongation of the field potential duration (3 10-6 M) (Mehta et al., 2012). Consistent with previous data, these

13

ACCEPTED MANUSCRIPT findings indicate that quinidine-induced prolongation of action potentials is mainly related to block of outward membrane currents particularly hERG K+ currents (Tsujimae et al., 2004). At higher concentrations quinidine also decreased the spontaneous beating frequency. This behavior may be largely related to a reduction in the rate of depolarization, reflecting the ability of quinidine to block fast Na+ channels (Salata and Wasserstrom, 1988).

PT

In rat neonatal cells, the beating rate decreased significantly with verapamil concentrations

RI

above 10-9 M whereas the concentrations required to reduce the rate of repolarization were

SC

one order of magnitude higher. These findings therefore indicate that the effect of verapamil on action potential duration is primarily due to suppression of spontaneous activity, rather

NU

than being mediated via changes in the rates of repolarization or depolarization. Various ion transporters, including L- and T-type Ca2+ channels, have been implicated in generating

MA

spontaneous activity in neonatal cardiomyocytes (Viatchenko-Karpinski et al., 1999). Our finding that verapamil significantly decreased beating rate would be compatible with its well-

D

known block of Ca2+ currents (probably mediated by the T-type channel) (Bergson et al.,

PT E

2011). The lack of action potential shortening with verapamil despite its well-known L-type Ca2+ channel block (Himmel et al., 2012) is likely because of counterbalancing effects of

CE

suppressed beating frequency (Grace and Camm, 2000). We speculate that changes in ionic concentrations and ion channel kinetics contributing to action potential prolongation on

AC

decreasing beating frequency, greatly attenuate the frequency-dependent L-type Ca2+ channel block by verapamil. The high sensitivity of the beating frequency supports the suppressive effect of verapamil on sinoatrial and atrioventricular nodal conduction and its clinical effectiveness for treating tachycardias (Durham and Worthley, 2002). The ICH S7B non-clinical guideline has proven adequate in reducing drug withdrawals (Colatsky et al., 2016). Many studies have demonstrated that restricting analyses to recording hERG activity (patch clamp systems) or reporting APD (MEA platforms) incorporates both

14

ACCEPTED MANUSCRIPT false positive and false negative outcomes (Kramer et al., 2013; Liang et al., 2013). Because many drugs exert their effect on multiple molecular targets they may mitigate or abolish potential arrhythmias. Therefore, a screening platform that detects drug effects on cardiac action potential parameters is likely to be more predictive. In summary, the proposed CMOS-MEA system offers high quality recordings with high

PT

information content that can be produced without extensive trainings. Future system

RI

development focuses at higher throughput and cell pacing functionalities in order to fully

AC

CE

PT E

D

MA

NU

SC

utilize the potential of this technology.

15

ACCEPTED MANUSCRIPT References

Bergson, P., Lipkind, G., Lee, S.P., Duban, M.-E., and Hanck, D.A. (2011). Verapamil block of T-type calcium channels. Molecular Pharmacology 79:411–419.

PT

Braeken, D., Jans, D., Huys, R., Stassen, A., Collaert, N., Hoffman, L., Eberle, W.,

RI

Peumans, P., and Callewaert, G. (2012). Open-cell recording of action

SC

potentials using active electrode arrays. Lab Chip 12:4397–4402.

NU

Colatsky, T., Fermini, B., Gintant, G., Pierson, J.B., Sager, P., Sekino, Y., Strauss, D.G., and Stockbridge, N., (2016). The comprehensive in vitro proarrhythmia assay

MA

(CiPA) initiative — Update on progress. Journal of Pharmacological and

D

Toxicological Methods 81:15-20.

PT E

Durham, D., and Worthley, L.I.G. (2002). Cardiac arrhythmias: diagnosis and

CE

management. The tachycardias. Critical Care and Resuscitation 4:35-53.

Franz, D., Olsen, H.L., Klink, O., and Gimsa, J. (2017). Automated and manual patch

AC

clamp data of human induced pluripotent stem cell-derived dopaminergic neurons. Scientific Data 4: (170056) 1 -11.

Gibson, J.K., Yue, Y., Bronson, J., Palmer, C. and Numann, R. (2014). Human stem cellderived cardiomyocytes detect drug-mediated changes in action potentials and ion currents. Journal of Pharmacological and Toxicological Methods 70(3): 255-267.

16

ACCEPTED MANUSCRIPT

Grace, A.A. and Camm, A.J. (2000). Voltage-gated calcium-channels and antiarrhythmic drug action. Cardiovascular Research 45:43-51.

Hamill, O.P., Marty, A., Neher, E., Sakmann, B. and Sigworth, F. J. (1981). Improved

PT

patch clamp techniques for high-resolution current recording from cells and

SC

RI

cell-free membrane patches. Pflügers Arch. 391, 85-100.

NU

Himmel, H.M., Bussek, A., Hoffman, M., Beckmann, R., Lohmann, H., Schmidt, M., and Wettwer, E. (2012). Field and action potential recordings in heart slices:

MA

correlation with established in vitro and in vivo models. British Journal of

D

Pharmacology 166(1):276-296

PT E

Huys, R., Braeken, D., Jans, D., Stassen, A., Collaert, N., Wouters, J., Loo, J., Severi, S. Vleugels, F., Callewaert, G., Verstreken, K., Bartic, C., and Eberle W. (2012).

CE

Single-cell recording and stimulation with a 16k micro-nail electrode array

AC

integrated on a 0.18 μm CMOS chip. Lab Chip 12:1274-1280.

Inayat, S., Pinto, L.H., and Troy J.B. (2013). Minimizing cytosol dilution in whole-cell patch-clamp experiments. Transactions on Biomedical Engineering 60, 7:2042-2051.

17

ACCEPTED MANUSCRIPT Kramer, J., Obejero-Paz, C.A., Myatt, G., Kuryshev, Y.A., Bruening-Wright, A., Verducci J.S., and Brown, A.M. (2013). MICE models: superior to the HERG model in predicting Torsade de Pointes. Scientific Reports 3:2100.

Liang, P., Lan, F., Lee, A.S., Gong, T., Sanchez-Freire, V., Wang, Y., Diecke, S., Sallam,

PT

K., Knowles, J.W., Wang, P.J., Nguyen, P.K., Bers, D.M., Robbins, R.C., and

RI

Wu, J.C. (2013). Drug screening using a library of human induced pluripotent

SC

stem cell–derived cardiomyocytes reveals disease-specific patterns of

NU

cardiotoxicity. Circulation 127:1677-1691.

Louch, W.E., Sheehan, K.A., and Wolska B.M. (2011). Methods in cardiomyocyte

MA

isolation, culture, and gene transfer. Journal of Molecular and Cellular

D

Cardiology 51:288-298.

PT E

Mégarbane, B., Karyo, S., Abidi, K., Delhotal-Landes, B., Aout, M., Sauder, P., and Baud, F.J. (2011). Predictors of mortality in verapamil overdose: usefulness of

CE

serum verapamil concentrations. Basic & Clinical Pharmacology &

AC

Toxicology, 108:385–389.

Mehta, A., Chung, Y.Y., Sequiera, G.L., Wong, P., Liew, R., and Sim, W. (2012). Pharmacoelectrophysiology of viral-free induced pluripotent stem cell-derived human cardiomyocytes. Toxicological Sciences, 131(2):458-469.

Millard, D.C., Strock, C.J., Carlson, C.B., Aoyama, N., Juhasz, K., Goetze T.A., StoelzleFeix, S., Becker, N., Fertig, N., January, C.T., Anson, B.D., and Ross, J.D.

18

ACCEPTED MANUSCRIPT (2016). Identification of drug–drug interactions in vitro: a case study evaluating the effects of sofosbuvir and amiodarone on hiPSC-derived cardiomyocytes. Toxicological Sciences, 154(1):174-182.

Möller, C., and Witchel, H. (2011). Automated electrophysiology makes the pace for

RI

PT

cardiac ion channel safety screening. Front. Pharmacol., 2, 73: 1-7.

SC

Natarajan, A., Stancescu, M., Dhir, V., Armstrong, C., Sommerhage, F., Hickman, J.J. and Molnar, P. (2011). Patterned cardiomyocytes on microelectrode arrays as

NU

a functional, high information content drug screening platform. Biomaterials

MA

32:4267-4274.

Obergrussberger, A., Brüggemann, A., Goetze, T.A., Rapedius, M., Haarmann, C., Rinke,

D

I., Becker, N., Oka, T., Ohtsuki, A., Stengel, T., Vogel, M., Steindl, J.,

PT E

Mueller, M., Stiehler, J., George, M., and Fertig, N. (2016). Automated patch clamp meets high-throughput screening: 384 cells recorded in parallel on a

CE

planar patch clamp module. Journal of Laboratory Automation 21(6):779-793.

AC

Roden, D.M. (2004). Drug-induced prolongation of the QT interval. The New England Journal of Medicine 350:1013-1022.

Salata, J.J., and Wasserstrom J.A. (1988). Effects of quinidine on action potentials and ionic currents in isolated canine ventricular myocytes. Circulation Research 62:324-337.

19

ACCEPTED MANUSCRIPT Sanguinetti, M.C., and Tristani-Firouzi, M. (2006). hERG potassium channels and cardiac arrhythmia. Nature, Nature Insight 440:463-469.

Shah, R.R. (2002). The significance of QT interval in drug development. British Journal

PT

of Clinical Pharmacology 54:188–202.

RI

Tsujimae, K., Suzuki, S., Yamada, M., and Kurachi, Y. (2004). Comparison of kinetic

NU

Journal of Pharmacology 493:29-40.

SC

properties of quinidine and dofetilide block of HERG channels. European

Viatchenko-Karpinski, S., Fleischmann, B.K., Liu, Q., Sauer, H., Gryshchenko, O., Ji,

MA

G.J., and Hescheler, J. (1999). Intracellular Ca2+ oscillations drive spontaneous contractions in cardiomyocytes during early development.

PT E

D

Proceedings of the National Academy of Science USA. 96:8259-8264.

Zhang, J., Qu, J., and Wang, J. (2014). Patch clamp apply in cardiomyocytes derived

CE

from patient’s iPS cells for individual anticancer therapy. International

AC

Journal of Clinical and Experimental Medicine 7:4475-4478.

Zhang, S., Zhou, Z., Gong, Q., Makielski, J.C., and January, C.T. (1999). Mechanism of block and identification of the verapamil binding domain to HERG potassium channels. Circulation Research 84:989-998.

20