Simultaneous Detection and Colocalization of Calcium Sparks and Ryanodine Receptor Clusters in Cardiac Myocytes

Abstract

Further understanding of calcium handling and excitation-contraction coupling in cardiac myocytes requires quantitative data analysis methods to characterize calcium release events in terms of structural properties of the cell. Such automatic methods provide a robust, consistent and reproducible characterization of physiological systems from observed experimental data. We present a multilevel analysis that simultaneously focuses on both the occurrence of spontaneous calcium sparks and the distribution of RyR clusters across a cardiac myocyte. The approach localizes the release events and determines the distance to the nearest ryanodine receptor cluster, therefore providing quantitative information on the spatio-temporal distribution of activation sites in the cell. The location of clusters takes into account motion artifacts produced by calcium waves or mini-waves. The method has been validated with line-scan confocal microscopy data from mouse ventricular myocytes and provides a full characterization of the spark morphology including its amplitude, baseline, decay time, upstroke time, Full-Width at Half-Maximum, Full-Duration at Half-Maximum and background noise.The processing is applied to linescan images of both RyR clusters and calcium fluorescence. It includes the following steps: i) Adaptive filtering of background fluorescence fluctuations using a robust estimation of the noise by means of the median absolute deviation of the basal fluorescence signal. ii) Individual sparks are detected by using a modified watershed segmentation method that includes stopping rules for both event size and shape. iii) Location of RyR clusters was implemented by thresholding a continuous wavelet transform of the cluster image (robust to motion and contraction). iV) Measurement of spark morphology properties, including the distance from each spark to the closest RyR cluster.