High-throughput Multimodal Functional Phenotypic Assay for Human Cardiotoxicity Screening

About This Grant

PROJECT SUMMARY Safety pharmacology concerns are a major cause of drug attrition during preclinical testing, with cardiac arrhythmias accounting for most cases. The advent of human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) enables an unlimited supply of human cardiomyocytes, offering a human-relevant alternative to animal-based testing in preclinical cardiac safety assessment. Although 2D hiPSC-CM monolayers have been validated for safety pharmacology testing by FDA-led international committees, they lack the physiological 3D structure and function of the mature human heart. Another limitation is the reliance on mixed populations of hiPSC-CM subtypes, including nodal, atrial, and ventricular cells. Furthermore, most current assays are low throughput and rely on a single electrophysiological metric without assessing contractility or multimodal functional activity. To overcome these limitations, we propose to develop a high-throughput, multimodal human 3D engineered heart tissue (EHT) assay that enables simultaneous, real-time measurement of contractile function (via magnetic sensing) and electrophysiological parameters (via optical mapping) for in vitro toxicity screening (called MagOptiTox). This platform integrates key technological innovations to overcome barriers in scalability, reproducibility, and interpretability. Specifically, we will employ AI-guided laser purification to generate high-yield, chamber-specific atrial and ventricular cardiomyocytes with consistent purity, and automated robotic tissue casting to ensure uniform, consistent EHT formation with rigorous quality control in a 96-well format. Together, these approaches establish a scalable manufacturing pipeline that minimizes human variability and enables large-scale production of standardized cardiac microtissues for predictive drug screening. We will also implement a physiology-aware machine-learning pipeline that integrates multimodal electrophysiological, calcium, and contractile readouts for cardiotoxicity risk modeling. This framework will extract key functional phenotypes from synchronized contractile/electrophysiological signals to classify torsadogenic risk using FDA/CiPA-endorsed reference compounds at clinically relevant concentrations. By combining high- content experimental measurements with AI-driven analytics, the system will enable quantitative, mechanistic prediction of drug-induced cardiotoxicity across diverse compound classes. The MagOptiTox platform will be validated against conventional 2D monolayer assays using multiple patient-derived hiPSC-CMs screened against medications with distinct mechanisms of action and known arrhythmia risk (Aim 1) and extended to model ischemic heart failure in a patient- and chamber-specific manner (Aim 2). We will demonstrate that heart- failure EHTs generated under metabolic stress are more sensitive to cardiotoxic compounds and ischemic injury than paired healthy controls. Successful completion of this project will establish a scalable, multimodal, and physiologically relevant human in vitro cardiotoxicity screening assay, advancing preclinical safety pharmacology and accelerating the development of safer therapeutics.