Heart-on-a-chip (HOC) platforms play a pivotal role in cardiac research, yet existing models suffer from mechanical mismatch between the soft myocardial tissue, sensors, and substrate, leading to impaired myocardial function and compromised data capture. Here, we introduce a mechanically matched HOC to address these challenges by mimicking the Young's modulus of the myocardial bilayer, including the elastic epicardium (30-70 kPa) and soft extracellular matrix (28-37 kPa). A process based on liquid-gas phase transition-induced porosification was developed, which introduces porosity into polydimethylsiloxane through controlled tetradecane phase transition, allowing for a tunable reduction in Young's modulus. This platform demonstrated excellent durability, withstanding over 1,000,000 stretch cycles, and allowed continuous electromechanical monitoring of cardiomyocyte behavior for 11 days. The mechanically matched platform promoted significant upregulation of key genes linked to cell adhesion, contraction, and electrical propagation (e.g., ITGA1, CACNA1C, SCN5A, and KCNH2) and enhanced excitation-contraction coupling by 128% compared to mismatched models. Additionally, the integration of machine learning into the HOC further improved drug classification accuracy, demonstrating the potential for advancing pharmacological evaluation.
Keywords: cardiomyocytes; excitation–contraction coupling; flexible sensor; heart-on-a-chip; mechanical matched.