Abstract
BACKGROUND: Human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) are in vitro models with the clear advantages of their human origin and suitability for human disease investigations. However, limitations include their incomplete characterization and variability reported in different cell lines and laboratories. OBJECTIVE: The purpose of this study was to investigate in silico ionic mechanisms potentially explaining the phenotypic variability of hiPSC-CMs in long QT syndrome type 3 (LQT3) and their response to antiarrhythmic drugs. METHODS: Populations of in silico hiPSC-CM models were constructed and calibrated for control (n = 1,463 models) and LQT3 caused by I(NaL) channelopathy (n = 1,401 models), using experimental recordings for late sodium current (I(NaL)) and action potentials (APs). Antiarrhythmic drug therapy was evaluated by simulating mexiletine and ranolazine multichannel effects. RESULTS: As in experiments, LQT3 hiPSC-CMs yield prolonged action potential duration at 90% repolarization (APD(90)) (+34.3% than controls) and large electrophysiological variability. LQT3 hiPSC-CMs with symptomatic APs showed overexpression of I(CaL), I(K1), and I(NaL), underexpression of I(Kr), and increased sensitivity to both drugs compared to asymptomatic LQT3 models. Simulations showed that both mexiletine and ranolazine corrected APD prolongation in the LQT3 population but also highlighted differences in drug response. Mexiletine stops spontaneous APs in more LQT3 hiPSC-CMs models than ranolazine (784/1,401 vs 53/1,401) due to its stronger action on I(Na). CONCLUSION: In silico simulations demonstrate our ability to recapitulate variability in LQT3 and control hiPSC-CM phenotypes, and the ability of mexiletine and ranolazine to reduce APD prolongation, in agreement with experiments. The in silico models also identify potential ionic mechanisms of phenotypic variability in LQT3 hiPSC-CMs, explaining APD prolongation in symptomatic vs asymptomatic LQT3 hiPSC-CMs.