The cardiomyocyte firestarter—RyR clusters ignite their neighbours after augmentation of Ca2+ release by β‐stimulation

Adrenergic stimuli augment cardiomyocyte contraction through modulation of intracellular Ca2+ transient properties. These Ca2+ transients are generated through the process of excitation-contraction coupling (ECC) and arise primarily via Ca2+ release from the sarcoplasmic reticulum (SR) through ryanodine receptor (RyR) channels. Ca2+ release at RyR clusters gives rise to Ca2+ sparks, which summate in space and time to produce the cell-wide Ca2+ transient. The organization of RyRs within clusters and the distribution of these clusters relative to each other and the sarcolemma are key to the rapid and synchronized release of Ca2+ in response to cell depolarization. In this issue of Acta Physiologica, by imaging Ca2+ release from GFP-labelled RyR clusters, Nolla-Colomer et al provide new insights into the relationship between RyR cluster architectures and Ca2+ spark properties.1 Further, they determine that adrenergic stimuli elicit their inotropic and arrhythmic effects through enhancing Ca2+ release at individual clusters and recruitment of neighbouring Ca2+ release sites (Figure 1). Since proposed as the fundamental unit of SR Ca2+ release during ECC, the mechanism by which RyR contribute to the generation of the complex hierarchy of Ca2+ signals ranging from an undetectable leak to a Ca2+ spark to a Ca2+ transient in cardiomyocytes, has been much investigated. In this regard, a particular focus has been on the organization of RyRs in clusters and how they function and interact within these clusters and with neighbouring clusters to modulate SR Ca2+ release. RyR clusters are primarily distributed at regions of the SR (junctional SR, jSR) located within 15-20 nm of the surface or T-tubular regions of the sarcolemma, forming a dyad (also known as a couplon).2 These clusters sit astride the Z-line of the sarcomere, optimally placing them to generate Ca2+ signals for induction of cell contraction. The localization of RyR clusters to the dyadic cleft further provides the necessary configuration for their local control by a trigger of Ca2+ arising from juxtaposed L-type Ca2+ channels (LTCC) on the sarcolemma. This local control of Ca2+ release allows for graded Ca2+ release through the stochastic recruitment of RyRs as a function of trigger size.3 The large size of the RyR (type 2 isoform in cardiomyocytes consisting of a tetramer of RyRs with mass 2 mDa and 30 nm across) makes enumeration of these channels in electron microscopy (EM) or immunostained images achievable. Early EM studies indicated that RyRs formed clusters in excess of 100 channels, often arranged in a checkerboard pattern2, 4 whilst confocal imaging of immunostained receptors suggested clusters comprising between 74 and 285 channels5; although these sizes form part of a continuous distribution from 1 to several hundred channels. Notably, despite substantial variation in RyR clusters size, Ca2+ sparks are highly stereotypical, showing little variation between species and heart regions. Indeed, whilst as few as 10 RyRs are required for Ca2+ spark generation, larger clusters are predominant. This stereotypical nature of Ca2+ sparks breaks down, however, under conditions of neurohormonal stimulation and in hypertrophic or diseased cardiac myocytes.6-8 Obtaining a clear understanding of how Ca2+ release is remodelled at the level of the RyR/RyR cluster has been hampered by the lack of capacity to match a Ca2+ release event with its underlying RyR clusters, thereby preventing the assignment of causality. Mathematical modelling of channel gating and their organization in clusters shows that clusters of 10-20 RyRs are optimal for the generation of sparks, although larger clusters show a greater propensity to release Ca2+.9, 10 Modelling analysis also shows a positive relationship between numbers of RyRs in a cluster and spark amplitude. This relationship reaches a plateau, however, with Ca2+ depletion of the SR leading to spark inactivation. Further, larger clusters exhibit greater sensitivity to release. These kinds of studies have also suggested that smaller clusters of RyRs have a higher threshold for activation.9, 11 In parallel with these modelling approaches, super-resolution microscopical imaging has been applied to deeply analyse RyR clusters. The increased resolution this provides has led to a downward revision of the numbers of channels within a cluster, with new data reporting numbers ranging between 10 and 25 channels.4, 7 Further, super-resolution imaging has revealed that RyR clusters are not necessarily a single entity. Large clusters observable by diffraction-limited confocal microscopy (resolution ~250 nm) often comprise several smaller clusters.7 Providing these clusters are less than 100-150 nm apart, they may function as a single Ca2+ release unit (CRU; aka a supercluster)4 (Figure 1). Importantly, these nanoscale approaches have also identified alterations in RyR cluster geometries under conditions of adrenergic stimulation and in hypertrophic or diseased cardiomyocytes that could explain the alterations in spark morphologies and their likelihood of generation.6-8 In particular, fragmentation of RyR clusters and decreased density of channels within a cluster has been described in disease, which leads to a decrease in the fidelity of Ca2+ spark generation (the likelihood that stochastic release at one RyR channel results in recruitment of enough RyRs to generate a spark) and to a reduction in the kinetics of Ca2+ release.7, 9 In addition, the increase in small RyR clusters in disease is likely associated with enhanced Ca2+ leak, via incompetency in spark generation. Alteration in RyR configuration in clusters was also observed in response to PKA-mediated phosphorylation of single RyRs, suggesting a mechanism for altered cluster properties.6 Direct demonstration of the relationship between the number of RyR clusters or RyRs within a cluster, and the properties of generated Ca2+ sparks has been a challenge. To co-localize Ca2+ release with release sites, an approach using heavy buffering with a slow Ca2+ chelator was first developed. Using this technique, discrete electrically-evoked Ca2+ release events termed Ca2+ spikes were recorded, supporting the notion that propagation of these events leads to the Ca2+ transient. Moreover, this study revealed that β-adrenergic stimulation led to increased recruitment of release sites, thereby contributing to the expected augmentation of Ca2+ release.12 Whilst insightful, the use of heavy buffering may however preclude extrapolation of these findings to more physiological settings. To fill the missing link between RyRs and Ca2+ release, approaches have been developed to either target Ca2+ sensors to the dyad/the Ca2+ release site or, as in the paper by Nolla-Colomer et al presented in this issue, by simultaneously imaging Ca2+ sparks and the distribution of RyR clusters comprising fluorescently labelled RyRs.1 To record Ca2+ release at the dyad, genetically encoded Ca2+ indicators (GCaMP) have been there targeted by either fusion with dyadic proteins eg junctin/triadin or FKBP12.6 immunophilin that associate with the RyR.13, 14 These indicators were first used to analyse Ca2+ transients where they showed the existence of substantial gradients between junctional and global Ca2+ levels. These studies did not however provide insights into Ca2+ release at individual release sites. Re-deployment of triadin-GCaMP to investigate the influence of neurohormones on dyadic Ca2+ release, revealed for the first time during ECC in intact cells that β-adrenergic stimulation augmented Ca2+ release by recruitment of additional dyadic release sites as well their synchronization.15 Endothelin, which acts via InsP3 generation, also synchronized Ca2+ release between dyads. In this issue of Acta Physiologica, by imaging Ca2+ release from GFP-tagged RyR clusters, Nolla-Colomer et al provided new insights into the relationship between Ca2+ sparks and the number of RyR clusters engaged and how cluster-to-cluster propagation could lead to the initiation of larger (possibly arrhythmogenic) Ca2+ release events (Figure 1).1 In particular, under baseline conditions, Ca2+ sparks identified by a single centre of mass in the fluorescent signal of the Ca2+ indicator, predominantly arose from a single RyR cluster (91%), with 8% from 2 cluster events the greatest majority from the same Z-line. Notably, these 2 cluster Ca2+ sparks exhibited centres of mass with delayed peaks, consistent with activation of the second cluster and spark by Ca2+ diffusion from the first. These two cluster events had a larger signal mass than single cluster events. Without stratifying sparks according to the underlying cluster number as carried out in this study, the categorization of sparks of different masses would not be possible, especially given their low proportion. Whilst it could be hypothesized that the 2 cluster events were activated by diffusion from a release event with greater signal mass, this did not seem to be the case. The presence of these larger events contrasts with a second recent study in which RyRs were labelled using a fluorescent FKBP12.6 that was washed into the cardiomyocytes after permeabilization.11 In this aforementioned study, Ca2+ spark frequency but not amplitude was increased with cluster size, although no 2 cluster events were reported. The authors noted however that FKBP12.6 suppressed basal activity, and have previously reported that FKBP12.6 expression leads to increased spark amplitude and a modest increase in SR Ca2+ load. As such, differences between small and large clusters in Ca2+ release may be lost. In a parallel modelling study from the same authors, spark amplitude was however reported to be related to cluster size.10 Contrastingly, when Nolla-Colomer et al examined the effects of β-adrenergic stimulation, where the proportion of 2 cluster events was greater, Ca2+ signal mass at the initiating RyR cluster was larger.1 Further, they determined that β-adrenergic stimuli enhance Ca2+ release from the SR via increasing Ca2+ release from individual clusters, which acts to recruit and activate Ca2+ release at neighbouring clusters. The greater Ca2+ release at RyR clusters following β-adrenergic stimulation was brought about through an increased SR Ca2+ load combined with sensitization of RyR by CaMKII and PKA phosphorylation. Under adrenergic conditions, cluster-to-cluster propagation, even between Z-lines was enhanced. Together, these findings provide a mechanism by which adrenergic agonists modulate Ca2+ dynamics to elicit their inotropic effect. This activity was also suggested to mimic the hyper-adrenergic conditions in diseased cardiomyocytes. Increased Ca2+ release was however associated with a greater Ca2+ load, which is often absent in cardiomyocytes from diseased hearts. Whether RyR phosphorylation would be sufficient for enhanced cluster-to-cluster propagation was not determined. Significantly, comparison of spark masses between human atrial cardiomyocytes with and without atrial fibrillation (AF) with the data generated in mice, revealed a high correlation between the spark properties in AF with the 2 cluster events identified in adrenergic stimulated GFP-RyR2 mouse ventricular cardiomyocytes.1 These data suggest that increased RyR phosphorylation in AF promotes cluster to cluster propagation (the probability of 2 cluster events) and the generation of arrhythmogenic Ca2+ waves. Given that the majority of atrial cardiomyocytes lack T-tubules, care should be taken in interpreting comparisons between mouse ventricular cardiomyocytes and human atrial cardiomyocytes during normal ECC and generation of Ca2+ transients. In the mouse ventricular cardiomyocytes, Ca2+ release from each cluster would ordinarily be activated by triggering Ca2+ via LTCC, thereby limiting the capacity for inter-cluster coupling and promotion of the arrhythmic activity. In atrial cardiomyocytes, SR Ca2+ release is substantially activated via Ca2+ diffusing across the cytosol. Under these conditions, a far greater role for cluster-to-cluster coupling in arrhythmic activity could be proposed. A similar situation could be envisaged in cardiomyocytes of larger mammals, where T-tubules are at a lower density and less than 50% of RyR are junctional. It must be noted that this study by Nolla-Colomer et al provides a diffraction-limited view of Ca2+ release and the underlying machinery. The imaged RyR clusters are likely superclusters comprising a number of smaller clusters. RyR clusters described here and elsewhere, lie 0.4-0.8 µm apart when on the same Z-line and 1-8–2.0 µm apart when not.1, 2, 4, 7 For both situations, because of the level of Ca2+ required for RyR activation, their distance apart precludes their acting as a single Ca2+ release unit.3 In conclusion, the approach and innovative methodologies used by these authors uncovered new insights into the relationships between RyR, RyR clusters and Ca2+ release as well as how Ca2+ release is dysregulated in pathology. The simultaneous analysis of the Ca2+ spark and its underlying molecular machinery is an important step towards a deeper understanding of cardiomyocyte Ca2+ release mechanisms. It is important, however, to take these approaches and examine RyR cluster activation and crosstalk in the context of Ca2+ transients during ECC. The authors declare no conflict of interest.