The first significant achievements in cellular cardiac electrophysiology date back to the beginning of the 1950s, when the two-microelectrode technique was applied to a variety of preparations, most notably “false tendons” of dog, kid, or lamb heart [1]. The Purkinje cells contained in these tissue samples present several important features, including a larger diameter (40-100 μm) and length and a lower contractility compared to ordinary ventricular fibres, enhancing multiple impalement and long stable recordings (over 10 minutes), an almost complete insensitivity to parasympathomimetic agents, and higher densities of voltage-dependent ion channels, required to achieve a speed of action potential propagation of several m/s. Beyond the remarkable contribution brought to the discovery of different ion current components and modeling the cardiac cell electrophysiology [2,3], they remained for decades a preferred system for in vitro cardiac pharmacology trials. It has been shown that, compared to papillary muscles or ventricular trabeculae, isolated rabbit Purkinje fibers are far more sensitive in detecting changes in action potential (AP) duration (APD90 – duration from the point of maximum upstroke velocity to 90% return from the peak depolarization to the resting potential) and presence of early afterdepolarizations (EAD) for a variety of drugs, including erythromycin, dofetilide, sertindole, and sparfloxacin [4]. Recently, this classical “golden standard” has been challenged by rapid progress in the field of embryonic stem cell (ESC) and especially induced pluripotent stem cell (iPSC) research. Since the first report in 2006 by the group of Shinya Yamanaka of iPSC generation via transfection of four “pluripotency” factors genes using retroviral vectors [5], multiple technological breakthroughs allowed for substantially higher yields of stem cells derived from differentiated cells like skin fibroblasts, peripheral blood lymphocytes, adipocytes or hair follicle keratinocytes, lower risks of tumorigenesis, as well as increasingly sophisticated protocols for redifferentiation of these stem cells into mature cardiomyocytes [6]. The painstaking in vitro redifferentiation process requires a stepwise application of conditioning media including three families of extracellular growth factors required for cardiac development in mammals: bone morphogenetic (BMP) factors like BMP4 and activin A, inhibitors of Wnt (wingless) signaling like DKK1 (dickkopf 1), and fibroblast growth factors, like βFGF [7]. However, once obtained, these cardiomyocytes are suitable for a variety of applications, including regenerative medicine approaches, in vitro models of diseases, and cardiac safety drug screening.

Multiple pharmacology assays using hiPSC-CM

More than a decade has passed since the first report describing spontaneous APs and contractility in cardiomyocytes within embryoid bodies grown from human embryonic stem cells [8], and recently we have witnessed an emergence of studies using either human embryonic stem cell-derived (hESC-CM) and/or human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CM) as standardized preparations for in vitro experiments. Recent progresses include refinement of cell culture protocols [9], use of antibiotic resistance tags to increase homogeneity of cell preparations [10], use of retro/lentiviral-free transfection methods in view of clinical applications [6]. Commercial stem cell-derived cardiomyocyte products are available from several suppliers, e.g. hESC-CMs from GE Healthcare (Piscataway, NJ; Cytiva™) or hiPSC-CMs from Cellular Dynamics International (Madison, WI; iCell^® Cardiomyocytes), Axiogenesis (Cologne, DE; Cor.4U^®), Pluriomics (Leiden, NL; StemCARDcells) or ReproCell (Yokohama, JP; ReproCardio2).

This scientific development led to novel high-throughput pharmacology screening equipments and techniques, based either on optical methods [11], using wide dynamic range Ca²⁺-sensitive dyes and improved cell loading with extracellular dye fluorescence quenching (e.g. FLIPR^® Calcium 6 assay) and transmembrane voltage-sensitive dyes (e.g. FLIPR^® Membrane Potential Assay Kit), or multi-electrode array (MEA) methods, allowing extracellular recordings of field potentials (FP), the cellular equivalent of the surface electrocardiogram, impedance measurements [12], and even two-dimensional local activation maps and conduction velocity studies [13]. Both optical and FP MEA-based screening assays present certain advantages, such as high throughput and simplicity of use, the possibility to assess AP propagation and effects of gap junction modulators due to use of confluent monolayers of mature highly differentiated iPSC-CM, and the ability to monitor “subacute” or “chronic” drug exposure effects. However, these systems suffer several shortcomings in exploring the mechanism of action and proarrhythmic effects of certain pharmacological compounds. This is related to the fact that arrhythmogenesis represents an emergent property at cell level, involving complex imbalances between multiple ion current components and intracellular ion handling and modulatory mechanisms [14], and optical/MEA-based iPSC-CM assays report only the sum of all drug-induced effects on the cell. The whole-cell patch clamp adds to this the proper reporting and control of the membrane potential. Recent modeling studies in combination with experimental approaches such as isolated heart preparations have evidenced new arrhythmogenesis mechanisms, such as synchronization of chaotic EADs in the genesis of torsades-des-pointes (TdP) and polymorphic ventricular tachycardia (PVT), a fascinating example of “deterministic chaos” at cellular level [15]. Moreover, the use of detailed cardiomyocyte cell electrophysiology models and advanced mathematical methods like dynamical systems linear stability and bifurcation analysis contributed to a detailed understanding of the mechanisms of cellular pace-making and ectopic focal arrhythmic activity in terms of damped or sustained oscillatory regimes achieved for certain combinations of parameters of a dynamical system, including a steady externally applied current [16-19], which is experimentally possible with whole-cell patch-clamp, but not with extracellular methods like MEA.

Performance of automated patch-clamp platforms for current-clamp AP recordings in hiPSC-CM

Thus, evidence of a pharmacological effect at the level of AP (FP or whole-cell) in spontaneous or induced pacing conditions represents only the first step in a complex process of arrhythmogenic mechanism assessment for that compound. In a simulation study performed with the LuoRudy dynamic model [20], using a large database of 45,000 APs generated with variable combinations of ion current surface densities, in the range of 0.25-2 times the default value, scanned against a test action potential obtained by modifying a single current component, we have shown that the simple AP shape matching criterion does not lead to a unique unambiguous identification of the modified current component [21]. The whole-cell patch-clamp approach offers superior versatility compared to both optical and MEA-based screening assays, because for each tested cell, beyond AP recordings in various pacing conditions, pharmacological effects on different current components can be thoroughly assessed via adequately designed voltage-clamp protocols. There are a few recent reports of hiPSC-CM experiments using planar automated patch-clamp platforms, consisting either in AP recordings in current-clamp mode [22] or voltage-clamp recordings of I_Na, I_Ca and I_Kr [10]. Our own experiments with the CytoPatch™ 2 equipment on iCell^® Cardiomyocytes (CDI, Madison, WI) indicate a good stability of AP duration and shape in the classical whole-cell configuration using physiological external and internal solutions, as shown in Figure 1, and the possibility to combine current-clamp and voltage-clamp protocols in complex assays. Previous studies have emphasized the capacity of automated patch-clamp platforms to use very small samples of cell suspension [23], and our special cytocentering technique [24] allows experiments with as few as 150 cells in a 3-μl sample, offering over 80% capture rates and high yields of stable tight seals, with both seal and membrane resistance over 1 GΩ. In addition, using special culture conditions, we obtained routinely high percentages of viable mature cardiomyocytes with predominantly ventricular phenotype (APD50/90>0.7).

Figure 1: AP shape and stability in hiPSC-CM (iCell^® Cardiomyocytes, CDI, 14-21 days in vitro). Typical whole-cell recordings of ventricular, atrial, and nodal-like cardiomyocytes (A). Stability plots of APD90 (B) and APD50/90 (C) for 5 iPSC-CM. All experiments were performed with CytoPatch™ 2 in current-clamp mode, with external pacing (0.5 ms, 2 nA, at 3-s intervals), at 21-22°C. External solution (in mM): 140 NaCl, 2.5 KCl, 2 MgCl₂, 2 CaCl₂, 10 HEPES, 10 Glucose, 15 Sucrose, pH 7.40, 320 ± 5 mOsm/kg.; internal solution (in mM): 100 K Gluconate, 20 KCl, 1 CaCl₂, 1 MgCl₂, 10 HEPES, 11 EGTA-KOH, 4 ATP-Mg²⁺, 3 Phosphocreatine-Na₂-H₂O, 9 Sucrose, pH 7.20, 290 mOsm/kg.

iPSC-CMs represent however imperfect electrophysiology models of mature cardiomyocytes of different types, due to variations in the levels of expression of multiple ion channel subunits and other morphological or functional features, such as a low density of T tubules, which leads to a completely different dynamics of Ca²⁺ between subcellular compartments during an AP [25]. Some of these unwanted features can be corrected using advanced stimulus protocols, e.g. a steady negative externally applied current can override the effects of excessive inward ion currents, bringing the resting potential from a depolarized state back to a value of -80 to -90 mV, specific for mature ventricular cardiomyocytes, and suppressing spontaneous APs, thus allowing pharmacological screening in ventricular-like settings using defined external pacing. Furthermore, a specific current component can be pharmacologically suppressed, while controlling the number of APs executed by the cell, to unravel the repolarization reserve [26], which is also very important in cardiac channelopathies [27].

Multiple ion channel effects and the CiPA paradigm

A paradigm that gains more and more support in the field of drug safety high-throughput screening is that of Multiple Ion Channel Effects (MICE), relying on assessment of pharmacological effects on three ion channels playing the major role in shaping the ventricular AP (Nav1.5, Cav1.2, and hERG1), and then evaluating the risks for different types of arrhythmias (e.g. the “torsadogenic” risk) [28]. Using a receiver operating characteristic (ROC) analysis for different risk prediction models, these authors have convincingly proved that the MICE approach offers a better sensitivity/specificity compromise compared to hERG screening alone, the classical paradigm in cardiac drug safety assessment. The new Comprehensive in vitro ProArrhythmia Assay (CiPA) guidelines, issued by a consortium of regulatory agencies and international research institutions, drive home the paramount importance of patch-clamp methods for cardiac safety pharmacology screening, proposing a combination of automated high-throughput or manual voltage-clamp assays and computer simulations with cellular cardiac electrophysiology models, in view of a complex proarrhythmic liability assessment of a drug candidate [29].

The importance of patch-clamp electrophysiology and pharmacology trials is even more obvious when it comes to in vitro studies using patient-derived iPSC-CMs as a diagnostic tool in cardiac channelopathies and other cardiac diseases with genetic inheritance [30-33]. Since the clinical picture is often unspecific, an accurate molecular diagnosis currently relies on extensive gene sequencing assays. The use of patch-clamp experiments on patient-derived iPSCCMs could change the diagnostic and therapeutic flowchart in the future, by pinpointing the faulty ion current component based on significant changes in cell surface density and/or kinetics, thus focusing and limiting the gene screening approach. In addition, patient-derived iPSC-CMs can be used in extensive patch-clamp pharmacology trials to explore in detail the arrhythmogenesis mechanisms, to generate an extended list of drugs incurring arrhythmia risks in an individual case, a list of well-tolerated drugs, and, in certain instances, even a patienttailored pharmacological treatment for arrhythmia prevention. A summary of our preliminary experiments with patient-specific hiPSCCM preparations can be found in an application note available at http://www.cytocentrics.com/en-us/newsmedia/downloadinformation.aspx.

In the context of extended electrophysiology and pharmacology trials on human (including patient-specific) iPSC-CMs, automated patch-clamp equipments, such as the CytoPatch™ 2, commercially available at Cytocentrics [24,34], in combination with advanced experimental protocols and computerized modeling studies, may prove invaluable tools in improving the prognosis and life quality of patients with such rare but severe diseases, as well as in routine quality assurance of batch production and daily use of cardiomyocytes in CiPA safety assays.

The authors gratefully acknowledge the entire Cytocentrics team for support, and Dr. Blake Anson (Cellular Dynamics International) for instructive discussions.