Large-Scale, Automated Calcium Spark Analysis using iSpark Reveals Functional and Spatial Remodeling During Cardiac Hypertrophy

Abstract

Understanding of cardiac Ca signaling is driven by advancements in imaging and analysis tools. The availability of fast line scanning allowed the identification of Ca sparks that revolutionized our appreciation of cardiac Ca signaling in the healthy and diseased heart. Ultra-high-speed confocal imaging of three dimensional (2D over time) Ca spark properties yields novel high content data about such signals whose characteristics eventually determine the performance of the entire heart. The full potential of such data remains concealed owing to the lack of comprehensive, fully automated and unbiased analysis tools. We have developed an intelligent software tool (iSpark) employing fully automatic, self-adaptive and unbiased algorithms to investigate Ca sparks in cardiac myocytes from healthy animals and mice with tiered stages of hypertrophy. Long-term recording of Ca sparks with high speed 2D over time confocal imaging of permeabilized ventricular myocytes produced high content spark data. With iSpark we explored 670,000 individual events and revealed that their subcellular arrangement, amplitude and frequency were substantially altered with the severity of cardiac hypertrophy. While line scanning of our data failed to show any correlation, iSpark-based analysis demonstrated a malfunctional microscopic Ca signaling strongly correlating with the disorganization of T-tubules and the severity of hypertrophy. The availability of highly sophisticated analysis tools, such as iSpark, substantially fosters large-scale data exploration. iSpark-based analysis of large-scale Ca spark data revealed a causal contribution of distinct subcellular remodeling of Ca handling to the progression of cardiac hypertrophy.