Filling the Gap Between Calcium Sparks and Waves: Automatic Detection and Classification of Local Calcium Releases in Cardiac Pacemaker Cells

Abstract

Local calcium releases (LCRs) observed in cardiac pacemaker cells have a complex spatiotemporal structure that has never been studied. We developed a computer algorithm for automatic detection and classification of LCRs in simulations of rabbit sinoatrial-node cells (using our recent 3D-model) to get new insights into pacemaker cell operation, specifically, the role of sarcoplasmic reticulum calcium pumping rate (Pup).Identified release events that share a common intensity level are categorized as a release cluster, i.e. a complex release with multiple intensity peaks. These complex LCRs tend to live longer and propagate farther via calcium-induced-calcium release, thus occupying larger areas. Release events that don’t share any intensity level with other events are calcium sparks that do not live for a long time and do not propagate. Collisions and splits of LCRs are handled as follows. When an LCR separates into different parts, all parts are still considered part of the LCR. On the other hand, when an LCR collides with another, the one with the weaker signal mass is considered dead and the one with the larger signal mass takes its signal mass as its own. An LCR may also die by stochastic attrition when all its components fade out.Under voltage clamp, LCR areas and signal masses were paradoxically smaller at larger Pup, likely reflecting uptake of cytosolic calcium before it can propagate. Under spontaneous beating conditions, however, higher Pup greatly increased diastolic LCR signal mass and beating rate as predicted by the coupled-clock theory. Interestingly, the total integral of all LCRs during diastolic depolarization in both cases remained almost the same as longer integration time with smaller events is comparable to shorter time with larger events.