Integrative electrophysiological studies of cardiotoxic agents using human stem cell-derived cardiomyocytes

Abstract

Even though several in vitro and in vivo QT screening systems (e.g. hERG assay, telemetry in conscious animals) are currently used as standardized assays for cardiotoxicity, these testing models are deficient. The major reason for their poor predictive powers is that they cannot replicate human cardiac electrophysiology. With the development of stem cell technologies, in vitro assays using human pluripotent stem cell-derived cardiomyocytes (hPSC-CMs) is suggested as an effective candidate of drug screening platform. The aims of this study are to explore the necessity of integrative interpretation with multiple types of cardiac ion channels for the cardiotoxicity test, and to evaluate the usefulness of adopting hPSC-CMs for the electrophysiological study. Chapter 1: Comparison of electrophysiological effects of calcium channel blockers on cardiac repolarization1 It is anticipated that Ca2+ channel blockers (CCBs) would shorten action potential duration (APD), which could lead to tachycardia. Nevertheless, CCBs are widely prescribed to hypertension without serious problems of cardiac arrhythmia. Here I investigated the electrophysiological effects of dihydropyridine class of CCBs, nicardipine (NIC), isradipine (ISR), and amlodipine (AML). All the three CCBs inhibited the L-type Ca2+ currents (ICa) whereas the shortening of APD was observed only with ISR. In addition, interestingly, NIC and AML also inhibited voltage-gated K+ channels currents (IKr and IKs) at micromolar ranges while ISR did not. I interpret that the concomitant K+ channel inhibition by NIC and AML might have compensated the AP shortening effects induced by the ICa inhibition. Chapter 2: Integrative analysis of cardiac ion channel modulation by SARI class antidepressants in human stem cell-derived cardiomyocytes2 The potential usefulness of human stem cell-derived cardiomyocytes in drug toxicity testing is drawing attention to the pharmaceutical industry recently. Here I evaluated the usefulness of commercialized human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs). The cardiac three types of action potentials, nodal (N)-, atrial (A)-, and ventricular (V)-type, and ion channels related cardiac AP (IKr, IKs, IK1, If, INa, ICa) were recorded in the cells. Additionally, hiPSC-CMs effectively recapitulate the electrophysiological behaviors of the major ion channel blockers, (E-4031 for hERG channel, tetrodotoxin for Na+ channel, nifedipine for calcium channel), confirming the plausibility of a platform for preclinical drug safety assessment. Then I analyzed the cardiotoxic effects of trazodone and nefazodone, serotonin antagonist and reuptake inhibitor (SARI) class antidepressants, using hiPSC-CMs or HEK293 cells overexpressing cardiac ion channel. Both drugs induced APD prolongation and early afterdepolarizations (EADs) and reduced the upstroke velocity in a dose-dependent manner. Consistent with the changes in the AP parameters, nefazodone and trazodone inhibited IKr, IKs, INa, and ICa, among them especially IKr and INa, but nefazodone had a higher inhibitory potency than trazodone. Chapter 3: Differentiation period-dependent changes in the electrophysiological properties of human stem cell-derived cardiomyocytes The assessment of functionality of human embryonic stem cell derived cardiomyocytes (hESC-CMs) at early developmental stages is essential for determining the appropriate differentiation stage for cardiotoxicity screening. In this study, to determine more suitable stage of differentiation required for the reliable pharmacological and toxicological testing, I characterized 2 week (2W) and 4 week (4W) differentiated hESC-CMs and compared their electrophysiological phenotypes and functional maturation using patch-clamp technique. The densities of functional ion channels currents, INa, ICa, IKr, IKs, and IK1, tended to increase in the 4W hESC-CMs while not significantly. In the AP recordings, the 2W hESC-CMs displayed only A-type (87.5%) and N-type (12.5%) without V-type of APs. However, the 4W hESC-CMs revealed 3 types of AP with the majority of cells revealed V-type APs (69%). The pharmacological responses for anti-arrhythmic drugs revealed that quinidine and amiodarone (Na+ and K+ channel blockers, respectively) prolonged APD at 90% (APD90) in the 4W hESC-CMs while not in the 2W hESC-CMs. Nifedipine significantly shortened APD90 only in the 4W hESC-CMs. Taken together, this study demonstrated the drug-induced cardiotoxicity has to be estimated with overall effects on multiple ion channels because of their compensatory effects between depolarizing and repolarizing currents. The hiPSC-CMs could be a valuable testbed for evaluating the proarrhythmic liability of trazodone and nefazodone

electrophysiological properties of hiPSC-CMs and their responses faithfully reflected the changes of individual ion channel current. The hiPSC-CMs can be an effective model for detection of early drug-induced cardiotoxicity beyond the current standard assay of hERG K+ channels. However, to use stem cell-derived cardiomyocytes in drug screening, at least 4 weeks of differentiation period is required for the reliable pharmacological and toxicological testing.