Improved Detection of Compound Arrhythmogenicity in 2D Cardiac Stem Cell Models via Advanced Maturation Strategies

Abstract

The adoption of stem cells as relevant models of in vitro human function is often hampered by their lack of phenotypic maturity. For example, it has been shown that cardiomyocytes differentiated from human induced pluripotent stem cells (hiPSC-CMs) can recapitulate some aspects of human toxicity in vitro, but fail to identify certain well-known safety liabilities. Bioengineering strategies hold great promise in advancing the maturity of a host of structural and functional phenotypes. This is especially relevant to cardiomyocytes, which are highly reliant on in vivo extracellular factors for their growth, development, and function. In this study, we hypothesized that advancing the in vitro maturation of hiPSC-CMs with a combination of exogenous cues will enable them to better detect known safety liabilities and better recapitulate in vivo biology based on industry-standard measures of assay fidelity. We used a novel microRNA-based maturation strategy that has been demonstrated to shift gene expression patterns in hiPSC-CMs to a more adult-like phenotype. We took these matured cells and treated them with several drugs from the CiPA formulary and benchmarked their measured risk score against unmatured cells. We measured the field potential duration and localized extracellular potential (LEAP) of the cells on MEAs. We tested bepridil, a known arrhythmogenic drug; in contrast to unmatured cells, matured myocytes showed clear FP prolongation (ddFPDc > 250 ms), clear Type-C arrhythmias in the field potential signal, and early afterdepolarizations in the LEAP signal. We also demonstrated that ranolazine, metoprolol, and mexiletine—which are safe in vitro but are typically scored as borderline dangerous in traditional myocyte assays—showed little to no toxicity (< 20 ms ddFPDc) and an absence of arrhythmias in matured preparations. We conclude that advancing hiPSC-CM maturation in vitro improves toxicity assay fidelity. The strategies presented are utilized in a cell, assay, and instrument-agnostic fashion and can be extended to nearly all hiPSC-CM models.