Distinct morphometric features of cardiomyocytes isolated from mouse hypertrophy models: An ImageJ analysis combined with machine learning algorithms.

This study aims at defining a standardized workflow based on a customized ImageJ macro combined with a machine-learning algorithm to analyze morphometric features of isolated cardiomyocytes using high-resolution/high-content photomicrographs and to identify key and specific morphological features of cardiomyocytes isolated from various murine cardiac hypertrophy models. For that purpose, we set up and optimized a Langendorff based protocol for isolating cardiomyocytes from mouse hearts. This optimized protocol yielded in a significantly high number of formaldehyde-fixed cardiomyocytes, with more than 97% of rod shaped cells. Moreover, our method allowed for reliable gene expression analysis and conservation of cell integrity through multiple freeze-thaw cycles. Next, we successfully applied our analytical workflow on formaldehyde-fixed cardiomyocytes isolated from various murine cardiac hypertrophy models and defined distinct morphological features in Angiotensin II, Isoproterenol, and age-induced hypertrophy. Taken together, our study provides an effective and standardized workflow for high-throughput morphological and molecular characterization of isolated cardiomyocytes, and could constitute a robust and reliable analytical tool to distinguish healthy versus diseased states and assess the ability of a potential therapeutic agent or strategy to reverse the situation.