Obstruction of ventricular Ca²⁺‐dependent arrhythmogenicity by inositol 1,4,5‐trisphosphate‐triggered sarcoplasmic reticulum Ca²⁺ release

Abstract

Key points

• Augmented inositol 1,4,5‐trisphosphate (IP₃) receptor (IP₃R2) expression has been linked to a variety of cardiac pathologies.

• Although cardiac IP₃R2 function has been in the focus of research for some time, a detailed understanding of its potential role in ventricular myocyte excitation–contraction coupling under pathophysiological conditions remains elusive.

• The present study focuses on mechanisms of IP₃R2‐mediated sarcoplasmic reticulum (SR)‐Ca²⁺ release in ventricular excitation–contraction coupling under IP₃R2‐overexpressing conditions by studying intracellular Ca²⁺ events.

• We report that, upon IP₃R2 overexpression in ventricular myocytes, IP₃‐induced Ca²⁺ release (IP₃ICR) modulates the SR‐Ca²⁺ content via “eventless” SR‐Ca²⁺ release, affecting the global SR‐Ca²⁺ leak. Thus, IP₃R2 activation could act as a SR‐Ca²⁺ gateway mechanism to escape ominous SR‐Ca²⁺ overload.

• Our approach unmasks a so far unrecognized mechanism by which “eventless” IP₃ICR plays a protective role against ventricular Ca²⁺‐dependent arrhythmogenicity.

Abstract

Augmented inositol 1,4,5‐trisphosphate (IP₃) receptor (IP₃R2) function has been linked to a variety of cardiac pathologies including cardiac arrhythmias. The functional role of IP₃‐induced Ca²⁺ release (IP₃ICR) within ventricular excitation–contraction coupling (ECC) remains elusive. As part of pathophysiological cellular remodelling, IP₃R2s are overexpressed and have been repeatedly linked to enhanced Ca²⁺‐dependent arrhythmogenicity. In this study we test the hypothesis that an opposite scenario might be plausible in which IP₃ICR is part of an ECC protecting mechanism, resulting in a Ca²⁺‐dependent anti‐arrhythmogenic response on the cellular scale. IP₃R2 activation was triggered via endothelin‐1 or IP₃‐salt application in single ventricular myocytes from a cardiac‐specific IP₃R type 2 overexpressing mouse model. Upon IP₃R2 overexpression, IP₃R activation reduced Ca²⁺‐wave occurrence (46 vs. 21.72%; P < 0.001) while its block increased SR‐Ca²⁺ content (∼29.4% 2‐aminoethoxydiphenyl borate, ∼16.4% xestospongin C; P < 0.001), suggesting an active role of IP₃ICR in SR‐Ca²⁺ content regulation and anti‐arrhythmogenic function. Pharmacological separation of ryanodine receptor RyR2 and IP₃R2 functions and two‐dimensional Ca²⁺ event analysis failed to identify local IP₃ICR events (Ca²⁺ puffs). SR‐Ca²⁺ leak measurements revealed that under pathophysiological conditions, “eventless” SR‐Ca²⁺ efflux via enhanced IP₃ICR maintains the SR‐Ca²⁺ content below Ca²⁺ spark threshold, preventing aberrant SR‐Ca²⁺ release and resulting in a protective mechanism against SR‐Ca²⁺ overload and arrhythmias. Our results support a so far unrecognized modulatory mechanism in ventricular myocytes working in an anti‐arrhythmogenic fashion.

Key points

• Augmented inositol 1,4,5‐trisphosphate (IP₃) receptor (IP₃R2) expression has been linked to a variety of cardiac pathologies.

• Although cardiac IP₃R2 function has been in the focus of research for some time, a detailed understanding of its potential role in ventricular myocyte excitation–contraction coupling under pathophysiological conditions remains elusive.

• The present study focuses on mechanisms of IP₃R2‐mediated sarcoplasmic reticulum (SR)‐Ca²⁺ release in ventricular excitation–contraction coupling under IP₃R2‐overexpressing conditions by studying intracellular Ca²⁺ events.

• We report that, upon IP₃R2 overexpression in ventricular myocytes, IP₃‐induced Ca²⁺ release (IP₃ICR) modulates the SR‐Ca²⁺ content via “eventless” SR‐Ca²⁺ release, affecting the global SR‐Ca²⁺ leak. Thus, IP₃R2 activation could act as a SR‐Ca²⁺ gateway mechanism to escape ominous SR‐Ca²⁺ overload.

• Our approach unmasks a so far unrecognized mechanism by which “eventless” IP₃ICR plays a protective role against ventricular Ca²⁺‐dependent arrhythmogenicity.

Introduction

Sudden cardiac death due to arrhythmia is responsible for 10–30% of deaths among patients with hypertrophic cardiomyopathy (Ullal et al. 2016). In cardiac hypertrophy, alterations in Ca²⁺‐induced Ca²⁺ release (CICR) and excitation–contraction coupling (ECC) have been associated with cardiac instability (Gusev et al. 2009). Thus understanding the underlying mechanisms essentially contributes to the development of novel treatment strategies.

Ca²⁺ release from the sarcoplasmic reticulum (SR) in ventricular cardiomyocytes is mediated by two Ca²⁺ release channels: the ryanodine receptor type 2 (RyR2) and the inositol 1,4,5‐trisphosphate receptor type 2 (IP₃R2; Bers, 2002; Hohendanner et al. 2014). While RyR2s play a central role in ventricular ECC, the contribution of functionally expressed IP₃R2s under physiological and pathophysiological conditions is less well understood. In contrast to RyR2s, which are exclusively activated by Ca²⁺, IP₃R2s are sensitive to Ca²⁺ and IP₃. The latter SR‐Ca²⁺ release mechanism has been established as IP₃‐induced Ca²⁺ release (IP₃ICR; Zima & Blatter, 2004; Harzheim et al. 2009). Intracellular IP₃ synthesis is triggered by a variety of vaso‐ and cardioactive hormones (e.g. endothelin‐1, angiotensin II) by stimulating G protein‐coupled receptors (GPCRs) that couple to phospholipase C, liberating IP₃ from phosphatidylinositol 4,5‐bisphosphate (Proven et al. 2006; Berridge, 2009).

Although the ratio of functionally expressed RyR2s versus IP₃R2s is approximately 100:1, a modulatory function of IP₃ICR in ventricular myocyte ECC has been suggested and is the subject of an ongoing controversial discussion (Lipp et al. 2000; Kockskämper et al. 2008). In contrast to the physiological condition, a more pronounced role of IP₃ICR in the framework of cardiac pathophysiology seems to be reasonable. It has been found that remodelling of ECC under various cardiac pathologies (e.g. cardiac hypertrophy, ischaemic dilated cardiomyopathy, atrial fibrillation, failing myocardium and hypertension) is associated with IP₃R2s overexpression and, in this context, enhanced IP₃ICR has been repeatedly linked to arrhythmogenicity (Zhang et al. 2002; Harzheim et al. 2009, 2010; Nakayama et al. 2010; Signore et al. 2013). Supporting evidence comes from the observation that endothelin‐1 and angiotensin II – both GPCR agonists – have similar positive inotropic effects in ventricular myocytes, suggesting a role of IP₃ICR in the regulation of cytosolic Ca²⁺ homeostasis (Zima & Blatter, 2004). In addition, RyR2–IP₃R2 local bi‐directional crosstalk mechanisms in cardiomyocytes have been recently described and may be involved in Ca²⁺‐dependent arrhythmogenicity at the cellular level (Wullschleger et al. 2017). Moreover, these studies support the concept of distinct RyR2‐IP₃R2 Ca²⁺ signalling compartments within cardiomyocytes – either in the cytosol and/or in the nucleus – during pathophysiological signalling driven by IP₃ and IP₃R2 (Wu & Bers, 2006; Drawnel et al. 2010). In contrast, it has also been suggested that increased synthesis of IP₃ as well as IP₃R2 expression, may be a common feature of cardiac pathologies but without having functional significance for the ECC mechanism (Cooley et al. 2013).

At the subcellular level, synchronized and functionally clustered RyR2 openings result in “Ca²⁺ sparks”, the building blocks of global Ca²⁺ transients in ventricular myocytes; analogously coordinated openings of clustered IP₃R2s give rise to “Ca²⁺ puffs” (Cheng et al. 1993; Yao et al. 1995). Although similar, the distinct spatio‐temporal characteristics of Ca²⁺ puffs and Ca²⁺ sparks and their pharmacological profile can be used for classification and separation of SR‐Ca²⁺ release events (Hohendanner et al. 2015). Openings of singular and/or “misplaced” RyR2s or IP₃R2s that do not belong to channel clusters may not generate detectable SR‐Ca²⁺ release events but can still contribute to the intracellular Ca²⁺ homeostasis (Lipp & Niggli, 1996; Sobie et al. 2006; Brochet et al. 2011). Analogously to recently described “rogue” RyR2s, “rogue” IP₃R2s may generate an “eventless” SR‐Ca²⁺ release flux which might have a significant impact on the total SR‐Ca²⁺ leak in ventricular myocytes without giving rise to localized Ca²⁺ events (Sobie et al. 2006).

Among other factors, the open probability (P _o) of both IP₃R2s and RyR2s and thus the SR‐Ca²⁺ leak, is critically controlled by SR luminal Ca²⁺ (Nunn & Taylor, 1992; Terentyev et al. 2002). Spontaneous increased RyR2 P _o under SR‐Ca²⁺ overload results in asynchronous SR‐Ca²⁺ release (e.g. Ca²⁺ waves) and may lead to delayed after‐depolarizations, a major cause of ventricular tachyarrhythmias and sudden cardiac death (Keller et al. 2007; Liu et al. 2013). Although largely outnumbered by RyR2s, overexpression of IP₃R2s within cardiac pathology may render the latter receptor a more significant SR‐Ca²⁺ regulatory function. The present study focuses on potential mechanisms by which IP₃R2 may have a Ca²⁺‐dependent anti‐arrhythmogenic function at the cellular level.

Our data in the present study reveal that stimulation of IP₃ICR in freshly isolated ventricular myocytes from a cardiac‐specific IP₃R2‐overexpressing mouse model leads to a decrease in asynchronous SR‐Ca²⁺ release, a decline in the SR‐Ca²⁺ content and subsequent decrease in pro‐arrhythmic Ca²⁺ wave occurrence, suggesting a potential anti‐arrhythmic role of IP₃ICR. Although under cellular remodelling conditions the SR‐Ca²⁺ content appears to be under active control of IP₃ICR, no additional IP₃R2‐mediated local SR‐Ca²⁺ events (e.g. Ca²⁺ puffs) were identified. Interestingly, study of the SR‐Ca²⁺ leak revealed a more prominent IP₃R2‐dependent component that controls the SR‐Ca²⁺ content. Thus, these results support the view that IP₃ICR contributes to the total SR‐Ca²⁺ leak via “eventless” SR‐Ca²⁺ release. The underlying mechanism actively stabilizes the SR‐Ca²⁺ content and limits luminal RyR2 sensitization below the critical threshold for spontaneous Ca²⁺ release thereby offering a protective mechanism against arrhythmias in ventricular myocytes.

Methods

Ethical approval

All experiments were approved by the State Veterinary Office of Bern (Switzerland), in accordance with Swiss Federal Animal Protection Law and were approved by the in‐house ethical committee (Ref. BE96/16, Ref. EFWW0443). For single myocyte isolation, mice were killed by cervical dislocation, hearts were extracted and perfused using the Langendorff perfusion technique. For transthoracic echochardiography measurements, mice were anaesthetized with 1–1.5% isoflurane, maintaining heart rate at 450–500 beats min⁻¹. Following data acquisition, mice were monitored until recovered from anaesthesia.

Cardiac‐specific IP₃R2‐overexpressing mouse model

Mouse lines FVB‐Tg(Myh6/tet0‐Itpr2)3.11Jmol/J (The Jackson Laboratory, Bar Harbor, ME, USA) and FVB.CG‐Tg(MyH6‐tTA)6Smbf/J (Jacksons) were crossbred to generate the cardiac‐specific IP₃R type 2 (IP₃R2) overexpressing mouse model (TG; Nakayama et al. 2010). In the following, wild‐type (WT) will be a synonym for ventricular myocytes isolated from WT mice and transgene (TG) for cells from IP₃R2‐overexpressing mice. Genotype was assessed via conventional polymerase chain reaction. Primers used include: for IP₃R2 5′‐CTTTCATCGGATGCCTGCCT‐3′ and 5′‐TTGGGAGGCCAAGGTGGGTA‐3′ forward and reverse, respectively; and for tetracycline transactivator (tTA) 5′‐CGCTGTGGGGCATTTTACTTTAG‐3′ and 5′‐CATGTCCAGATCGAAATCGTC‐3′ forward and reverse, respectively. Heart and body weights (mg g⁻¹) were determined at 5 months of age. All mice were fed with doxycycline‐containing food (0.2 g kg⁻¹, Envigo, Füllinsdorf, Switzerland) during pregnancy and until 4 weeks of age.

Echocardiography

Transthoracic echocardiography was performed using a 30 MHz probe and the Vevo 2100 Ultrasound machine (VisualSonics, Bothell, WA, USA). All mice were anaesthetized with 1–1.5% isoflurane. The heart was imaged in 2‐D mode on a parasternal long‐axis view from where an M‐mode cursor was positioned perpendicular to the interventricular septum and the posterior wall of the left ventricle. Echocardiographies were performed at the Cardiovascular Assessment Facility (CAF, Lausanne, Switzerland).

RNA and protein analysis

Animals were killed by cervical dislocation and tissues from ventricle, atria, aortic smooth muscle, femoral skeletal muscle and brain were excised, washed and snap‐frozen in liquid nitrogen. For western blot analysis, ventricular tissue was ground in extraction buffer (1x Protease Inhibitor (04693159001, Sigma‐Aldrich Chemie, Buchs, Switzerland) in NP40 extraction buffer, Invitrogen, Thermo Fisher Scientific, Reinach, Switzerland) for 30 min. Samples were centrifuged at 1400 g for 10 min at 4°C to obtain the supernatant. Following quantification by the bicinchoninic acid assay (BCA) method, 25 μg per lane of protein in loading solution (10% glycerol, 5% β‐mercaptoethanol, 43% SDS, 62.5 mM Tris, 6 M urea) were loaded in 4–20% pre‐cast gel (Bio‐Rad, Basel, Switzerland) and run for 55 min at 200 V. Samples were transferred to PVDF membranes at 100 mA and 4°C overnight. After PBS washing steps, membranes were blocked for 1 h at room temperature (RT) in blocking solution (5% milk in PBS). Membranes were incubated with primary antibodies anti‐IP₃R2 (1:5000 KM1083, kindly provided by Prof. Dr K. Mikoshiba; Sugiyama et al. 1994), anti‐RyR2 (1:2000 ab2827, Abcam, Cambridge, UK) or anti‐GAPDH (1:100,000 10R‐G109A, Fitzgerald, Acton, MA, USA) in primary antibody solution (PBS‐Tween, 0.5% milk). Secondary antibody anti‐mouse Alexa 790 (1:100,000, ThermoFisher Scientific, Reinach, Switzerland) was applied in secondary antibody solution (PBS‐Tween, 0.5% milk, 0.01% SDS) for 1 h at RT. Imaging was performed on an Odyssey set‐up and images were analysed using GelAnalyzer. Full western blot images are available upon request.

Total RNA was extracted from snap‐frozen mice tissues with NucleoSpin RNA Plus kit (Macherey‐Nagel, Oensingen, Switzerland) and concentrations were determined with a NanoDrop 2000c (ThermoFisher Scientific). RT‐qPCR was performed by loading 10 ng of sample in an Eco Real‐Time PCR system (Illumina, Cambridge, UK) using KAPA SYBR FAST One‐Step Universal kit (LabGene, Châtel‐Saint‐Denis, Switzerland) in accordance with the manufacturers’ instructions. Primers used include: ITPR2 primer pair (M_Itpr2_3, Sigma) and β‐2‐microglobulin primers (forward FM1_B2m, and reverse RM1_B2m, Sigma). ITPR2 gene relative expression was normalized to reference gene β‐2‐microglobulin. ΔΔC_q was calculated and used for statistical analysis. TG data are shown normalized to their respective WT tissue (norm.).

Ventricular myocyte isolation

Animals were killed at 5 months of age by cervical dislocation. Ventricular cardiomyocytes were isolated in Tyrode Ca²⁺‐free solution with the Langendorff perfusion technique with collagenase type II (Worthington) and Protease type XIV (Sigma). Extracellular Ca²⁺ was slowly increased to physiological concentration (1.8 mM).

Solutions and chemicals

Ventricular cardiomyocytes were permeabilized with 0.05% escin (Sigma). Cells were kept in final solution containing (mM): 120 potassium aspartate (KAsp), 3 K‐ATP, 3.734 MgCl₂, 0.5 EGTA, 10 phosphocreatine, 10 HEPES and 0.05 Fluo‐3 5K (Biotium), with the addition of 5 U L⁻¹ creatine phosphokinase (pH 7.4, 310 mOsm L⁻¹). Free‐[Ca²⁺] was measured using Indo‐1 AM (Molecular Probes, Eugene, OR, USA) in a NanoDrop fluorospectrometer and adjusted to 70 nM. Cells were plated using Cell‐Tak (Corning, Fisher Scientific, Schwerte, Germany)‐coated coverslips. Pharmacological interventions include 10 μM IP₃‐salt (IP₃, SiChem, Bremen, Germany), 2 μM 2‐aminoethoxydiphenyl borate (2‐APB, Calbiochem) and 20 mM caffeine (Sigma). Reports suggest that 2‐APB at higher concentrations than that used in the present study may have secondary effects in other intracellular components besides blocking IP₃R (Maruyama et al. 1997; Ascher‐Landsberg et al. 1999; Peppiatt et al. 2003). In our hands, application of 2 μM 2‐APB in WT cells – in the absence of IP₃ stimulation – returned no unspecific effects. To ensure that the responses seen with 2‐APB were due to IP₃R block, 3 μM xestospongin C (XeC, Focus Biomolecules) was used as an alternative IP₃R blocker. We are aware that both blockers used in our study do not have a 100% selectivity; however, the combination of both is currently the best option for a pharmacological separation of IP₃ effects.

Intact myocytes were loaded at RT with 5 μM Fluo‐3 AM (Biotium, Fremont, CA, USA) for 30 min, plated in Cell‐Tak‐coated coverslips and kept in extracellular solution containing (mM): 140 NaCl, 5 HEPES, 1.1 MgCl₂, 5.4 KCl, 10 glucose and 1.8 CaCl₂. For Na⁺ and Ca²⁺ free solution, NaCl was replaced by 140 LiCl and 0.5 EGTA. To reach steady state cells were field stimulated for 30 s at a frequency of 1 Hz. The following pharmacological agents were used: 100 nM endothelin‐1 (ET‐1, Sigma), 2 μM 2‐APB, 3 μM XeC, 1 mM tetracaine (tet, Sigma) and 10 mM caffeine (Sigma).

[Ca]²⁺ measurements and protocols

For the analysis of the SR‐Ca²⁺ leak/load relationship, intact cardiomyocytes were kept at 1.8 mM Ca²⁺ and paced for 30 s at 1 Hz to reach steady state. External solution was rapidly exchanged to a Na⁺ and Ca²⁺ free solution containing 1 mM NiCl₂ to ensure Na⁺/Ca²⁺ exchanger block (Kimura et al. 1987). After 20 s, 1 mM tetracaine was added to block RyR2. After 30 s, 10 mM caffeine was rapidly applied to determine the SR‐Ca²⁺ content. ΔCa²⁺ _cyt is estimated as the difference (ΔCa²⁺ _cyt = F _tet – F ₀) between fluorescence before (F ₀) and after RyR2 block (F _tet); as more Ca²⁺ is released to the cytosol via SR‐Ca²⁺ leak, the difference (ΔCa²⁺ _cyt) between baselines shrinks and the SR‐Ca²⁺ leak becomes more positive.

For single cell arrhythmogenicity studies, intact cells were kept at 1.8 mM Ca²⁺ external solution and were field stimulated for 30 s at 1 Hz prior to a 10 s rest period to study Ca²⁺ wave occurrence as an estimate of cell propensity to aberrant SR‐Ca²⁺ release. Finally, SR‐Ca²⁺ load was assessed by application of 10 mM caffeine. Ca²⁺ wave occurrence (%) is calculated as the portion of cells analysed that showed Ca²⁺ waves during the 10 s resting period.

Confocal Ca²⁺ imaging

To assess intracellular Ca²⁺ changes, Fluo‐3 was excited at 488 nm wavelength of an argon laser and fluorescence was collected at 525 nm in line‐scan mode (2 ms per line) with a photomultiplier tube (PMT) on a MicroRadiance laser scanning confocal microscope (Bio‐Rad, Watford, UK) or alternatively a FluoView 1000 (Olympus, Volketswil, Switzerland) with a solid‐state laser (Sapphire, Coherent, Santa Clara, CA, USA).

Data analysis and statistics

Raw line‐scan data from permeabilized myocytes were analysed using the ImageJ (NIH) plug‐in SparkMaster to obtain Ca²⁺ spark frequency (Ca²⁺ sparks (100 μm)^–1 s^–1), amplitude (Amp, change in fluorescence relative to the fluorescence background (ΔF/F ₀)), full‐width at amplitude half‐maximum (μm, FWHM), full duration at amplitude half‐maximum (ms, FDHM), full duration (ms), exponential time constant of the spark decay (τ_decay, ms) and time to peak (ms). Spark mass was calculated as Amplitude × 1.206 × (FWHM)³ (Hollingworth et al. 2001); Ca²⁺ released via Ca²⁺ sparks was estimated as the product of the spark frequency and spark mass. Spark frequency data are presented as 100 sparks μm⁻¹ s⁻¹ or normalized to the vector where indicated (norm.). Data analysis was performed using the softwares ImageJ (NIH, Bethesda, MD, USA) and Igor Pro (WaveMetrics, Lake Oswego, OR, USA).

Data are presented as box (25–75 percentiles) and whisker (10 and 90 percentiles) plots; outliers are highlighted. Data within tables are presented as mean ± standard deviation (SD). Normal distribution of the data was assessed by D'Agostino's K ² test. Statistical comparison was performed via one‐way ANOVA for multi‐parametric comparison followed by Dunnett's test or paired or unpaired Student's t test. Binary data were tested by the χ² test. Number of animals (N), cells (n), Ca²⁺ events (n _e) and statistical results are indicated in the figure legends.

Results

Cardiac IP₃R2 overexpression induces cardiac remodelling both at cellular and organ levels

A cardiac‐specific IP₃R2‐overexpressing mouse model (TG; Nakayama et al. 2010) has been used to mimic cardiac hypertrophy occurring in patients (Harzheim et al. 2009). In addition, the strategy to tune one element within a complex cellular remodelling situation to extreme could unmask an otherwise potentially invisible regulatory mechanism. The TG mice display a significantly higher heart weight to body weight ratio when compared to their wild‐type littermates (WT, 7.75 ± 0.92 mg g⁻¹ vs. TG 10.6 ± 1.69 mg g⁻¹; N = 5, 9; P = 0.001; Fig. 1 A). A cardiac hypertrophic phenotype was confirmed via transthoracic echocardiography (Table 1), which revealed greater interventricular septum (IVS, 0.78 ± 0.08 mm WT vs. 0.93 ± 0.09 mm TG; N = 4, 6; P = 0.039) and left ventricular posterior wall thickness (LVPW, 0.78 ± 0.07 mm WT vs. 0.94 ± 0.11 mm TG; P = 0.033) in TG hearts when compared to WT.

Figure 1. TG mice display cardiac remodelling at both cellular and organ levels.

A, heart weight (mg)/body weight (g) of WT (N = 5) and TG (N = 9) hearts at 5 months of age (^** P < 0.01; Student's t test). B, ITPR2 gene relative expression in various tissues depicted as ΔΔC_q and normalized to WT. β‐2‐Microglobulin was used as reference gene (ventricle and atria N = 4 each; brain, smooth and skeletal muscle N = 3 each). C, relative IP₃R2 and RyR2 protein expression in TG ventricular tissue when compared to WT and normalized to GAPDH. Box plots display IP₃R2 (left) and RyR2 (right) (^* P < 0.05, ^** P < 0.01; N = 4 each; Student's t test).

Table 1.

Systolic and diastolic parameters measured by transthoracic echocardiography in WT and TG mice at 5 months of age

	WT	TG
IVSd (mm)	0.78 ± 0.08	0.93 ± 0.09*
IVSs (mm)	1.07 ± 0.11	1.27 ± 0.13*
LVPWd (mm)	0.78 ± 0.07	0.94 ± 0.11*
LVPWs (mm)	1.09 ± 0.07	1.27 ± 0.16
%FS	33.35 ± 1.97	29.04 ± 4.06
%EF	62.45 ± 2.78	55.94 ± 6.21
E‐wave deceleration time (ms)	21.07 ± 6.84	12.38 ± 3.8*
ET (ms)	49.29 ± 3.6	58.81 ± 6.42*

TG mice display mild cardiac hypertrophy at 5 months of age. Values are means ± SD. Abbreviations correspond to: diastole (d), systole (s), interventricular septum (IVS), left ventricular posterior wall (LVPW), fractional shortening (%FS), ejection fraction (%EF) and ejection time (ET) (^* P < 0.05; N = 4, 7; Student's t test).

Figure 1 B shows an increase in the relative expression of ITPR2 mRNA (IP₃R2 encoding gene) levels in both ventricular and atrial tissue. A minor ITPR2 expression increase was detected in smooth and skeletal muscle but was absent in brain tissue. Accordingly, western blot analysis revealed a 12.72 (±2.97)‐fold increase in IP₃R2 protein synthesis (N = 4 each; P = 0.001; Fig. 1 C, left) in ventricle. In addition and in line with other reports (Zhao et al. 2007), RyR2 protein level was significantly decreased in TG ventricular tissue (61.73 ± 0.24%; N = 4 each; P = 0.012; Fig. 1 C, right).

Global Ca²⁺ wave occurrence in response to IP₃R2 pathway stimulation

Spontaneous Ca²⁺ waves are triggers for cardiac arrhythmias at the cellular level. To assess Ca²⁺‐dependent arrhythmogenicity in individual cells, IP₃ synthesis was elicited via activation of the phosphatidylinositol 4,5‐bisphosphate (PIP₂)–phospholipase C (PLC)–IP₃ signalling cascade by endothelin‐1 (ET‐1, 100 nM) in intact ventricular myocytes (Table 2). Cells were field stimulated for 30 s at 1 Hz to ensure comparable SR‐Ca²⁺ loading conditions prior to a 10 s rest period to assess the occurrence of Ca²⁺ waves (Fig. 2 A). In control conditions, without ET‐1 stimulation and when compared to WT, TG myocytes displayed higher Ca²⁺ transient amplitude (1.9 ± 0.4 ΔF/F ₀ WT vs. 2.32 ± 0.54 ΔF/F ₀ TG; n = 13, 14; P = 0.014; Fig. 2 B) under conditions of comparable SR‐Ca²⁺ load (3.5 ± 0.79 ΔF/F ₀ WT vs. 3.53 ± 0.78 ΔF/F ₀ TG; P = 0.46) and consequently the fractional SR‐Ca²⁺ release in TG cells was increased (0.56 ± 0.1 a.u. WT vs. 0.65 ± 0.12 a.u. TG; P = 0.025). Furthermore, TG myocytes displayed a higher spontaneous Ca²⁺ wave occurrence (5.5% WT vs. 46% TG; P < 0.001; Fig. 2 C).

Table 2.

Parameters of field stimulated WT and TG intact ventricular myocytes upon IP₃R2 pharmacological intervention

WT	Ca2+ transient amplitude (ΔF/F 0)	τdecay (ms)	T50 (ms)	Fractional SR‐Ca2+ release (a.u.)	SR‐Ca2+ content (ΔF/F 0)
Control (13, 4)	1.9 ± 0.4	156.92 ± 27.76	119.24 ± 19.25	0.56 ± 0.1	3.5 ± 0.79
ET‐1 (11, 3)	3.25 ± 1.17**	162.26 ± 20.73	124.88 ± 12.6	0.66 ± 0.14*	5.06 ± 1.81*
ET‐1 + 2‐APB (16, 3)	2.69 ± 0.84	176.78 ± 40.56	132.62 ± 23.49	0.63 ± 0.11	4.17 ± 0.92
2‐APB (16, 3)	2.09 ± 0.58	173.83 ± 42.17	129.04 ± 25.02	0.62 ± 0.15	3.47 ± 1.02
XeC (15, 3)	2.12 ± 0.73	168.44 ± 31.59	131.1 ± 21.19	0.62 ± 0.15	3.57 ± 1.46

TG	Ca2+ transient amplitude (ΔF/F 0)	τdecay (ms)	T50 (ms)	Fractional SR‐Ca2+ release (a.u.)	SR‐Ca2+ content (ΔF/F 0)
Control (14, 4)	2.32 ± 0.54‡	122.39 ± 22.3‡‡‡	97.01 ± 17.82‡‡	0.65 ± 0.12‡	3.53 ± 0.78
ET‐1 (16, 4)	2.63 ± 0.77	135.01 ± 35.5	106.82 ± 26.29	0.65 ± 0.13	4.07 ± 0.9
ET‐1 + 2‐APB (14, 3)	2.77 ± 0.86	126.68 ± 25.8	102.57 ± 20.22	0.71 ± 0.12	4.37 ± 0.65
2‐APB (14, 3)	3.04 ± 1.28*	131.9 ± 12.36	107.82 ± 13.74*	0.65 ± 0.18	4.57 ± 1.18**
XeC (11, 3)	2.95 ± 0.78*	164.2 ± 44.4	128.5 ± 24.78	0.72 ± 0.1	4.11 ± 1.08

Values are means ± SD. Spatio‐temporal characteristics analysed include Ca²⁺ transient amplitude (ΔF/F ₀), tau decay (τ_decay, ms), time to 50% amplitude (T50, ms), fractional SR‐Ca²⁺ release (a.u.) and SR‐Ca²⁺ content (ΔF/F ₀) from WT and TG (^* P < 0.05, ^** P < 0.01 vs. control; ^‡ P < 0.05, ^‡‡ P < 0.01, ^‡‡‡ P < 0.001 vs. WT). Number of animals (N) and cells (n) are indicated below each condition as (n, N).

Figure 2. IP₃R2 activation alters Ca²⁺ wave occurrence in intact cardiomyocytes.

A, representative line‐scans of fluo‐3‐loaded intact cardiomyocytes with the corresponding fluorescence profiles upon pharmacological intervention with 100 nM ET‐1, 2 μM 2‐APB or 3 μM XeC. Cells were field stimulated for 30 s at 1 Hz to reach steady state prior to a 10 s rest period to assess Ca²⁺ wave occurrence as a measure for arrhythmogenicity at the single cell level. SR‐Ca²⁺ load was assessed by application of 10 mM caffeine. B, Ca²⁺ transient amplitude (ΔF/F ₀) and SR‐Ca²⁺ content (ΔF/F ₀) under the tested pharmacological interventions in WT (n of groups 1–4 = 13, 11, 16, 15; N of groups 1–4 = 4, 3, 3, 3) and TG myocytes (n of groups 1–4 = 14, 16, 14, 11; N of groups 1–4 = 4, 4, 3, 3; ^* P < 0.05, ^** P < 0.01 vs. control; ANOVA followed by Dunnett's test). C, Ca²⁺ wave occurrence (%) calculated as the portion of cardiomyocytes displaying Ca²⁺ waves within the 10 s rest period (^* P < 0.05, ^*** P < 0.001 vs. control; ^‡‡‡ P < 0.001 vs. WT; N = 3–4, n = 9–21; χ² test).

Compared to control, ET‐1 (100 nM, after 7–10 min application) triggered an increase in Ca²⁺ transient amplitude (3.25 ± 1.17 ΔF/F ₀ ET‐1; n = 11; P = 0.001; Fig. 2 B, upper left), SR‐Ca²⁺ content (5.06 ± 1.81 ΔF/F ₀ ET‐1; P = 0.013; Fig. 2 B, upper right) and Ca²⁺ wave occurrence (27.78% ET‐1; P = 0.035; Fig. 2 C) in WT cardiomyocytes, but surprisingly this effect was absent in IP₃R2‐overexpressing cells (Fig. 2 B and C). In contrast, ET‐1 induced a 21.72% decrease in Ca²⁺ wave occurrence only in TG myocytes (Fig. 2 C; n = 16; P < 0.001).

In WT cardiomyocytes, block of IP₃R2 by 2‐aminoethoxydiphenyl borate (2‐APB, 2 μM) or xestospongin C (XeC, 3 μM) returned no detectable changes in the analysed Ca²⁺ responses. In contrast, under IP₃R2‐overexpressing conditions, blockade of IP₃R2 led to increases in Ca²⁺ transient amplitude (3.04 ± 1.28 ΔF/F ₀ 2‐APB and 2.95 ± 0.78 ΔF/F ₀ XeC; n = 14, 11; P = 0.033, 0.018; Fig. 2 B, lower left) with concomitantly higher SR‐Ca²⁺ content (4.57 ± 1.18 ΔF/F ₀ 2‐APB and 4.11 ± 1.08 ΔF/F ₀ XeC; P = 0.007, 0.08; Fig. 2 B, lower right). Finally, block of IP₃R2 by 2 μM 2‐APB abolished ET‐1‐mediated effects on cellular arrhythmogenicity both in WT and TG myocytes (Fig. 3). In other words, ET‐1 stimulation fails to alter the Ca²⁺ wave occurrence while IP₃R remains blocked (Fig. 3).

Figure 3. Block of IP₃R2 reverses ET‐1‐mediated arrhythmogenic effects.

Intact ventricular myocytes were stimulated as indicated in Fig. 2. Application of 2 μM 2‐APB with 100 nM ET‐1 reverses the latter mediated arrhythmogenic effects in Ca²⁺ wave occurrence both in WT (n of groups 1–4 = 13, 11, 16, 16; N of groups 1–4 = 4, 3, 3, 3) and TG myocytes (n of groups 1–4 = 14, 16, 14, 14; N of groups 1–4 = 4, 4, 3, 3; ^* P < 0.05, ^*** P < 0.001 vs. control; ^†† P < 0.01, ^††† P < 0.001 vs. ET‐1; ^‡‡‡ P < 0.001 vs. WT; χ² test).

To further study ET‐1‐mediated effects, estimation of sarco(endo)plasmic reticulum Ca²⁺‐ATPase (SERCA) and Na⁺–Ca²⁺ exchanger (NCX) relative contributions to the total Ca²⁺ transient decay was approached (Fig. 4). Fitting of a single exponential function to the action potential (AP)‐induced Ca²⁺ transient decay was used as a reference of the overall Ca²⁺ extrusion capacity of the cell, and to the Ca²⁺ transient (T) decay in the presence of 10 mM caffeine to assess NCX contribution (Fig. 4 A, left panel). SERCA and NCX (orange) contributions to the twitch decline (%) were calculated as: ((1/T _twitch) − (1/T _caffeine))/(1/T _twitch) and (1/T _caffeine)/(1/T _twitch), respectively. In control conditions (Fig. 4 A, right panel), TG myocytes display a higher SERCA contribution when compared to WT (^** P < 0.01; N = 4 each, n = 13, 14). An increase in SERCA contribution to the twitch Ca²⁺ decline upon 100 nM ET‐1 stimulation was seen only in WT ventricular myocytes (Fig. 4 Ba; ^‡‡ P < 0.01; N = 4, 3; n = 13, 11).

Figure 4. ET‐1 alters SERCA/NCX balance in the cytosolic Ca²⁺ decay in WT myocytes.

A, contribution to twitch decline of SERCA (%, white for WT and black for TG) and NCX (%, orange) in WT and TG myocytes (left panel). A single exponential function was fitted to the AP‐induced Ca²⁺ transient decay as a reference of the overall Ca²⁺ extrusion capacity of the cell; and to the Ca²⁺ transient decay in the presence of 10 mM caffeine to assess NCX contribution. SERCA and NCX contributions to twitch decline (%) were calculated as: ((1/T _twitch) − (1/T _caffeine))/(1/T _twitch) and (1/T _caffeine)/(1/T _twitch), respectively (^‡‡ P < 0.01; N = 4 each, n = 13, 14; Student's t test). B, contribution to twitch decline upon 100 nM ET‐1, 2 μM 2‐APB or 3 μM XeC intervention in WT (a; n of groups 1–4 = 13, 11, 16, 16; N of groups 1–4 = 4, 3, 3, 3) and TG ventricular myocytes (b; n of groups 1–4 = 14, 16, 14, 11; N of groups 1–4 = 4, 4, 3, 3; ^** P < 0.01 vs. control; ANOVA followed by Dunnett's test).

Taken together, under conditions of IP₃R2 overexpression, ET‐1 stimulation unexpectedly, and in contrast to the wild‐type situation, led to a decrease in Ca²⁺ wave frequency without altering Ca²⁺ transient amplitude nor SR‐Ca²⁺ content. However, block of IP₃R2 results in a significant increase in SR‐Ca²⁺ content, which indicates that IP₃ICR may play an intrinsic role in the regulation of the SR‐Ca²⁺ content and the propensity to arrhythmias.

Reduced SR‐Ca²⁺ content cannot be explained by Ca²⁺ sparks

The Ca²⁺ wave is a fundamental phenomenon that arises from unsynchronized SR‐Ca²⁺ release. Ca²⁺ sparks are elementary Ca²⁺ release events that can explain, at least in part, both the initiation and propagation of Ca²⁺ waves via the summation of individual Ca²⁺ release events and CICR. As was shown for atrial cells (Wullschleger et al. 2017), local bi‐directional crosstalk of IP₃R2 and RyR2 may trigger and boost ventricular CICR. Under this view, pronounced IP₃ICR may lead to an increased local Ca²⁺ events occurrence. Therefore, the effect of IP₃ICR under IP₃R2‐overexpressing conditions on the incidence of Ca²⁺ sparks was examined next.

Figure 5 A shows spontaneous Ca²⁺ sparks recorded in permeabilized cardiomyocytes upon diverse pharmacological interventions targeting the IP₃R2 pathway. Under resting conditions, IP₃R2‐overexpressing cells exhibit a higher spontaneous Ca²⁺ spark frequency (CaSpF) than WT (3.67 ± 1.8 sparks (100 μm)⁻¹ s⁻¹ TG vs. 1.76 ± 1 sparks (100 μm)⁻¹ s⁻¹ WT; n = 25 each; P < 0.001; Fig. 5 C, left panel) under similar SR‐Ca²⁺ loading conditions (Fig. 5 B, right, and C). In line with the concept that IP₃ICR can boost CICR (e.g. Ca²⁺ sparks), in WT ventricular cardiomyocytes IP₃‐salt (10 μM) evoked a significant increase in CaSpF (24.2%; n = 22; P = 0.006; Fig. 5 Da, left) together with a statistically significant decrease of the SR‐Ca²⁺ content (3.73 ± 0.89 ΔF/F ₀ control vs. 3.3 ± 0.69 ΔF/F ₀ IP₃; P = 0.041; Fig. 5 Da, right) with no further changes in Ca²⁺ spark mass (Table 3). This response was sensitive both to 2 μM 2‐APB and 3 μM XeC (Fig. 5 Da). In contrast to the native situation, in IP₃R2 overexpressing cardiomyocytes, IP₃ induced a 25.17% decrease of the CaSpF (P = 0.015; n = 25, 13; P = 0.015; Fig. 5 Db, right) along with a reduction of the SR‐Ca²⁺ content (4.17 ± 0.48 ΔF/F ₀ control vs. 3.56 ± 0.57 ΔF/F ₀ IP₃; P = 0.012; Fig. 5 Db, left) 2–4 min after application. Both effects were sensitive to XeC and 2‐APB.

Figure 5. IP₃ evokes a divergent response in CaSpF in WT and TG permeabilized myocytes.

A, representative line‐scan images of Ca²⁺ events in permeabilized cardiomyocytes in control conditions and in the presence of 10 μM IP₃‐salt and/or 2 μM 2‐APB, 3 μM XeC in WT and TG. After permeabilization and plating, each cell was scanned before and 2–4 min after pharmacological intervention. B, representative line‐scan images with the corresponding fluorescence profiles of de‐skewed caffeine (20 mM)‐induced Ca²⁺ transients under control conditions in WT and TG permeabilized myocytes. C, Ca²⁺ spark frequency (Ca²⁺ sparks (100 μm)⁻¹ s⁻¹) and SR‐Ca²⁺ content (ΔF/F ₀) in control conditions in WT and TG cells (^*** P < 0.001 vs. WT; n = 25 each, N = 8 each; Student's t test). D, normalized Ca²⁺ spark frequencies (norm.) and SR‐Ca²⁺ content (20 mM caffeine, ΔF/F ₀) after Ca²⁺ event occurrence assessment in control conditions and in the presence of IP₃‐salt and/or 2‐APB, XeC in WT (n of groups 1–4 = 25, 22, 20, 16; N of groups 1–4 = 8, 8, 3, 3) and TG myocytes (n of groups 1–4 = 25, 13, 11, 10; N of groups 1–4 = 8, 3, 3, 3; ^* P < 0.05, ^** P < 0.01 vs. control; ^† P < 0.05, ^†† P < 0.01 vs. IP₃; ANOVA followed by Dunnett's test).

Table 3.

Spatio‐temporal characteristics of Ca²⁺ events in permeabilized ventricular myocytes upon pharmacological intervention of the IP₃R2 pathway

WT	Amplitude (ΔF/F 0)	FDHM (ms)	Full duration (ms)	FWHM (μm)	Spark mass (a.u.)	τdecay (ms)	Time to peak (ms)
Control (25, 8)	1.33 ± 0.12	18.17 ± 1.35	45.62 ± 5.08	1.95 ± 0.16	13.41 ± 4.44	18.12 ± 2.9	10.8 ± 1.28
IP3 (22, 8)	1.34 ± 0.15	18.22 ± 2.32	44.71 ± 7.26	1.96 ± 0.16	14 ± 4.26	19.82 ± 8.38	10.76 ± 1.14
2‐APB + IP3 (20, 3)	1.37 ± 0.22	18.51 ± 2.15	48 ± 9	2.01 ± 0.14	15.3 ± 4.47	17.56 ± 4.54	11.5 ± 1.94
XeC + IP3 (16, 3)	1.27 ± 0.11	22.75 ± 3.41†††	50.69 ± 10.35†††	2.04 ± 0.2	16.55 ± 5.37	24.67 ± 9.48†	13.32 ± 3.31
2‐APB (12, 3)	1.32 ± 0.13	19.5 ± 3.53	50.25 ± 10.07	1.98 ± 0.2	13.96 ± 3.63	20.81 ± 8.87	11.83 ± 2.28
XeC (16, 3)	1.33 ± 0.25	22.54 ± 2.54***	55.28 ± 8.66***	2.03 ± 0.24	15.31 ± 5.02	22.34 ± 4.8**	13.9 ± 2.88***

TG	Amplitude (ΔF/F 0)	FDHM (ms)	Full duration (ms)	FWHM (μm)	Spark mass (a.u.)	τdecay (ms)	Time to peak (ms)
Control (25, 8)	1.32 ± 0.14	26.22 ± 6.53	62.13 ± 14.89	1.96 ± 0.22	13.89 ± 4.34	26.86 ± 8.38	13.6 ± 2.71
IP3 (13, 3)	1.41 ± 0.15	24.79 ± 7.5	57.94 ± 13.57	1.98 ± 0.15	16 ± 5.22	32.49 ± 17.1	13.38 ± 5.01
2‐APB + IP3 (11, 3)	1.32 ± 0.25	29.95 ± 12.8	68.63 ± 27.89	2 ± 0.12	15.04 ± 4.58	38.29 ± 27.57	15.64 ± 7.37
XeC + IP3 (10, 3)	1.31 ± 0.08†	27.16 ± 8.57	61.99 ± 16.85	1.91 ± 0.17	13.27 ± 4.76	36.91 ± 19.59	12.35 ± 1.57
2‐APB (17, 4)	1.31 ± 0.17	26.1 ± 5.27	57.12 ± 10.58	2.05 ± 0.23	16.34 ± 6.52	30.12 ± 11.52	13.03 ± 2.58
XeC (14, 3)	1.28 ± 0.16	26.93 ± 7.47	62.62 ± 14.65	2.02 ± 0.2	15.69 ± 5.17	42.3 ± 12.72	15.4 ± 4.73

Following cell permeabilization and plating, a representative line‐scan was recorded to assess Ca²⁺ events in control conditions. Pharmacological effects were assessed 2–4 min post‐intervention (10 μM IP₃‐salt, 2 μM 2‐APB, 3 μM XeC). Parameters analysed include Ca²⁺ spark amplitude (ΔF/F ₀), full duration at amplitude half‐maximum (FDHM, ms), full duration (ms), full width at amplitude half‐maximum (FWHM, μm), spark mass (a.u.), tau decay (τ_decay, ms) and time to peak (ms) for WT and TG (^** P < 0.01, ^*** P < 0.001 vs. control; ^† P < 0.05, ^††† P < 0.001 vs. IP₃‐salt; ANOVA followed by Dunnett's test). N and n are indicated below each condition as (n, N).

To elucidate how the SR‐Ca²⁺ content can be reduced even with a lower number of Ca²⁺ sparks, individual Ca²⁺ events were analysed. Spatio‐temporal Ca²⁺ event analysis revealed no differences in individual parameters between Ca²⁺ events observed in WT and TG cells upon direct IP₃ stimulation (Table 3). Recently, we reported that multi‐parametric analysis of single Ca²⁺ events upon similar pharmacological interventions as presented in the current study can be used as a tool to differentiate among Ca²⁺ sparks and Ca²⁺ puffs in permeabilized atrial myocytes (Wullschleger et al. 2017). Hence, we repeated the analysis for the data presented here. Figure 6 shows single Ca²⁺ events in control conditions and upon IP₃ application plotted according to their individual fluorescence amplitude and full duration at amplitude half‐maximum (FDHM). IP₃ application did not induce any detectable differences either in WT (Fig. 6 Aa; n _e = 195, 374) or TG myocytes (Fig. 6 Ab; n _e = 443, 269). The Ca²⁺ event analysis suggests that only one population of Ca²⁺ events was detected and can be identified as Ca²⁺ sparks. In other words, a reduced SR‐Ca²⁺ content in TG cardiomyocytes cannot be explained by RyR2 activity alone, given the reduced Ca²⁺ sparks occurrence.

Figure 6. Multi‐parametric analysis of single Ca²⁺ event properties failed to differentiate IP₃R2‐mediated Ca²⁺ events.

A, scatter plot of single Ca²⁺ events (from Fig. 3) according to amplitude (ΔF/F ₀) and full‐duration at half‐maximum amplitude (FDHM, ms) in WT (a, n _e for control and IP₃ = 195, 374, N = 8 each) and TG (b, n _e = 443, 269; N = 8, 3) myocytes. B, box and scatter plots display Ca²⁺ released via Ca²⁺ sparks (a.u.) estimated as the product of spark frequency and mass per cell in WT (a) and TG (b) myocytes in control conditions and upon stimulation with 10 μM IP₃‐salt (data from Fig. 3; ^* P < 0.05, ^** P < 0.01 vs. control; Student's t test).

In addition, we estimated the Ca²⁺ released via Ca²⁺ sparks under the tested conditions as the product of the Ca²⁺ spark mass and the CaSpF. Figure 6 B shows that IP₃ stimulation increased Ca²⁺ release via Ca²⁺ sparks in WT (Fig. 6 Ba), but evoked a reduction in TG cells (Fig. 6 Bb). Both effects are presumably due to the differences in CaSpF induced by the application of IP₃.

The discrepancy between reduced SR‐Ca²⁺ content and the total amount of Ca²⁺ released observed during IP₃ stimulation in TG cardiomyocytes can neither be explained by the number of detected events nor the spatio‐temporal characteristics of individual Ca²⁺ events.

Intrinsically active IP₃ICR modulates SR‐Ca²⁺ content via SR‐Ca²⁺ leak

Enhanced luminal Ca²⁺ increases RyR2 channel open probability (P _o; Györke & Györke, 1998). Thus aberrant SR‐Ca²⁺ overload may result in spontaneous SR‐Ca²⁺ release (Chen et al. 2014). Since a discrepancy between CaSpF and SR‐Ca²⁺ content was evident, we hypothesized a contribution of IP₃ICR via “eventless” SR‐Ca²⁺ release. We performed a series of SR‐Ca²⁺ leak protocols (Fig. 7 A) in the presence of RyR2 antagonists and/or IP₃R2 agonists/antagonists to assess a potential contribution of “eventless” IP₃ICR on the total SR‐Ca²⁺ leak (Shannon et al. 2002). Figure 7 B shows zoomed‐in examples with the corresponding fluorescence profiles for WT and TG myocytes (left and right, respectively) showing the RyR2 contribution to the total SR‐Ca²⁺ leak estimated as the difference (ΔCa²⁺ _cyt = F _tet − F ₀) between the fluorescence before (F ₀) and after RyR2 block (F _tet). In principle, an increase in the “tetracaine‐insensitive” SR‐Ca²⁺ leak (e.g. via IP₃ICR) may result in an apparent elevation in cytosolic [Ca²⁺] despite the presence of tetracaine (F _tet), thus the difference between baselines (ΔCa²⁺ _cyt) may become smaller, implying an increase in the total SR‐Ca²⁺ leak. Upon NCX block, the apparent ΔCa²⁺ _cyt significantly depends on the SR‐Ca²⁺ leak/load balance (e.g. SR‐Ca²⁺ release and re‐uptake) and steady state prior to RyR2 block (F ₀). In other words, the expected ΔCa²⁺ _cyt decrease may not fully reflect large increases in SR‐Ca²⁺ leak above steady state. This is seen in Figs 7 and 8.

Figure 7. TG intact myocytes display altered basal SR‐Ca²⁺ leak.

A, representative confocal line‐scan images of WT and TG cell responses (left and right, respectively). Intact ventricular myocytes were electrically paced for 30 s at 1 Hz to reach steady state. Tyrode solution was rapidly switched to a 0 Na⁺, 0 Ca²⁺ solution for 15 s prior to addition of 1 mM tetracaine (tet). Finally, SR‐Ca²⁺ content was assessed (10 mM caffeine, Caff). B, representative zoomed‐in confocal line‐scan images (from dotted box in A) of WT and TG cell responses (right) in control conditions with the corresponding fluorescence profiles (left). For illustrative purposes each plot is shown overlapping with a trace corresponding to the fit of a single exponential decay function. SR‐Ca²⁺ leak (as ΔCa²⁺ _cyt) was estimated by approaching the difference between fluorescence before (F ₀) and after RyR2 block (F _tet, ΔCa²⁺ _cyt = F _tet − F ₀). C, compared to WT, TG cells display a smaller ΔCa²⁺ _cyt (−0.16 ± 0.06 ΔF/F ₀ WT vs. −0.05 ± 0.07 ΔF/F ₀ TG; P < 0.001), larger Ca²⁺ transients (1.95 ± 0.78 ΔF/F ₀ WT vs. 3.18 ± 0.99 ΔF/F ₀ TG; P = 0.002) with enhanced fractional SR‐Ca²⁺ release (0.51 ± 0.13 a.u. WT vs. 0.69 ± 0.22 a.u. TG; P = 0.011) and SR‐Ca²⁺ content (3.84 ± 1.62 ΔF/F ₀ WT vs. 4.75 ± 1.39 ΔF/F ₀ TG; P = 0.075; ^* P < 0.05, ^** P < 0.01, ^*** P < 0.001 vs. WT; n = 10, 16, N = 4 each; Student's t test).

Figure 8. IP₃ICR modulation of the SR‐Ca²⁺ leak diverges upon IP₃R2 overexpression.

A, representative SR‐Ca²⁺ leak profiles in the tested conditions for WT and TG ventricular myocytes. Individual cells were separately treated with 100 nM ET‐1 and/or 2 μM 2‐APB to study the contribution of IP₃ICR in the total SR‐Ca²⁺ leak. For illustrative purposes each trace is fitted with a single exponential decay function shown as an overlapped coloured trace. Ba, illustrative representation of the SR‐Ca²⁺ leak protocol responses shown in A (WT and TG, above and below, respectively). Bb, illustrative montage of SR‐Ca²⁺ leak curves in the tested conditions showing the contribution of IP₃R and RyR Ca²⁺ leak to the total leak in WT and TG myocytes (above and below, respectively). Application of 1 mM tetracaine reveals the RyR‐mediated SR‐Ca²⁺ leak while addition of 2‐APB – in the presence of tetracaine – unmasks the IP₃R2 component. C, ΔCa²⁺ _cyt (ΔF/F ₀), Ca²⁺ transient amplitude (ΔF/F ₀) and SR‐Ca²⁺ content (ΔF/F ₀) of WT (a, n of groups 1–4 = 10, 20, 18, 21; N = 4 each) and TG (b, n of groups 1–4 = 16, 14, 21, 28; N of groups 1–4 = 4, 3, 3, 4) groups upon pharmacological intervention (^* P < 0.05, ^** P < 0.01 vs. control; ^† P < 0.05, ^††† P < 0.001 vs. ET‐1; ANOVA followed by Dunnett's test).

Figure 7 C shows that compared to WT, TG cells display a lower ΔCa²⁺ _cyt (−0.16 ± 0.06 ΔF/F ₀ WT vs. −0.05 ± 0.07 ΔF/F ₀ TG; n = 10, 16; P < 0.001), larger Ca²⁺ transients with enhanced fractional SR‐Ca²⁺ release and SR‐Ca²⁺ content (3.84 ± 1.62 ΔF/F ₀ WT vs. 4.75 ± 1.39 ΔF/F ₀ TG; P = 0.075).

Figure 8 A shows representative SR‐Ca²⁺ leak profiles in the tested conditions (control, ET‐1, ET‐1 + 2‐APB and 2‐APB) for WT and TG ventricular myocytes. Figure 8 Ba exhibits the aforementioned traces from the different conditions overlapped for each animal group (WT and TG, above and below, respectively). Finally, Fig. 8 Bb displays the fitted samples according to cell type (WT or TG, above and below, respectively) and pharmacological intervention (control vs. ET‐1, left and right, respectively) and indicates the SR‐Ca²⁺ leak portion corresponding to RyR or IP₃R.

In WT myocytes, 100 nM ET‐1 induced a 31.25% decrease in ΔCa²⁺ _cyt (n = 20; P = 0.011; Fig. 8 C, top left panel), which was sensitive to 2 μM 2‐APB. 2‐APB alone did not significantly affect any of the tested parameters. In addition, ET‐1 induced IP₃‐independent increases of the SR‐Ca²⁺ content but not the Ca²⁺ transient amplitude (Fig. 8 C, top middle and right panels, respectively).

Surprisingly, in TG cells, ET‐1 application resulted in no detectable ΔCa²⁺ _cyt changes (Fig. 8 C, bottom left panel) and non‐significant decreases of the Ca²⁺ transient amplitude and the SR‐Ca²⁺ content (Fig. 8 C, bottom middle and left panels, respectively). In contrast, IP₃R2 block by 2‐APB resulted in a large decrease of the ΔCa²⁺ _cyt both in the presence and absence of ET‐1 (Fig. 8 C, bottom left panel) together with increases in the SR‐Ca²⁺ content (Fig. 8 C, bottom right panel). No further changes were detected in Ca²⁺ transient amplitude (Fig. 8 C, bottom middle panel) upon IP₃R2 block. However, the “hidden” IP₃‐dependent ΔCa²⁺ _cyt effect in the leak protocol is unmasked by IP₃R2 block.

Taken together, in IP₃R2‐overexpressing myocytes IP₃ICR modulates the SR‐Ca²⁺ content via sustained SR‐Ca²⁺ leak driven by “eventless” IP₃ICR rather than by localized Ca²⁺ release events, e.g. Ca²⁺ puffs. In addition, under IP₃R2‐overexpressing conditions, IP₃ICR is intrinsically active and thereby fine‐tunes the SR‐Ca²⁺ load and cytosolic Ca²⁺, subsequently modulating Ca²⁺ wave occurrence.

Discussion

Our results obtained in ventricular cardiomyocytes reveal that, upon IP₃R2 overexpression, IP₃ICR stimulation leads to a decrease in RyR2 activity together with a decline in SR‐Ca²⁺ content, indicating that IP₃R2 activation can act as a SR‐Ca²⁺ gateway mechanism to escape ominous SR‐Ca²⁺ overload and subsequent unsynchronized SR‐Ca²⁺ release. Our approach unmasks a so far unrecognized mechanism by which IP₃ICR modulates SR‐Ca²⁺ content via “eventless” SR‐Ca²⁺ leak and effectively restores cardiomyocyte Ca²⁺ homeostasis.

These observations are based on a cardiac‐specific IP₃R2‐overexpressing mouse model that mimics at least in part cardiac hypertrophy in patients (Nakayama et al. 2010). Cardiac remodelling is a complex process that affects a broad range of intracellular Ca²⁺ signalling components. TG myocytes are characterized by increases in IP₃R2 expression together with RyR2 down‐regulation, a phenomenon known for various cardiac diseases (Go et al. 1995; Domeier et al. 2012; Zima et al. 2014).

IP₃ is known to significantly contribute to a variety of intracellular Ca²⁺ signalling mechanisms by functionally exercising as a secondary messenger that interconnects extracellular signals to numerous subcellular processes. In ventricular myocytes, IP₃Rs are preferentially located within the nuclear envelope and are found at a lower extent in the SR membrane (Bare et al. 2005). While nuclear IP₃ICR has been linked to the modulation of gene expression (Higazi et al. 2009), the contribution of SR‐IP₃Rs in ventricular ECC under physiological conditions remains controversial (Zhang et al. 2002; Harzheim et al. 2009, 2010; Nakayama et al. 2010; Cooley et al. 2013; Signore et al. 2013). Whether and in which direction enhanced IP₃ICR may effectively modulate cardiac ECC in pathophysiological conditions warrants further investigation.

Ca²⁺ wave occurrence is controlled by IP₃ICR

In the myocardium, the amplitude of global Ca²⁺ signals is locally controlled by Ca²⁺ influx through L‐type Ca²⁺ channels (LTCCs) and subsequently amplified by SR‐Ca²⁺ release via CICR. Because of its positive feedback, CICR has an inherent tendency to oscillate, particularly under pathological conditions. Based on spontaneous SR‐Ca²⁺ release, these oscillations have been found to play a fundamental role in the generation of cardiac arrhythmias. At the cellular level, Ca²⁺ oscillations originating from spontaneous SR‐Ca²⁺ release channel openings often manifest as Ca²⁺ waves that travel from release junction to release junction in an unsynchronized manner. This process can be described as a “fire‐diffuse‐fire” mechanism (Keizer et al. 1998). In this scenario, aberrant SR‐Ca²⁺ releases can substantially increase cytosolic Ca²⁺ levels and trigger delayed afterdepolarizations, a major source for ventricular tachyarrhythmias (Lipp & Niggli, 1993). A contribution of IP₃ICR to the genesis of Ca²⁺‐dependent arrhythmias has been suggested for some time (Zhang et al. 2002; Zima & Blatter, 2004; Kockskämper et al. 2008; Harzheim et al. 2009; Nakayama et al. 2010). Our observations in wild‐type cells (WT) are in agreement with this concept. Under control conditions we found an increase in SR‐Ca²⁺ release activity triggered by stimulation of the IP₃ pathway in the shape of Ca²⁺ waves or Ca²⁺ sparks.

The use of G‐protein coupled receptor (GPCR) agonists (e.g. ET‐1) leads to the intracellular synthesis of IP₃ together with activation of protein kinase C (PKC). Although the positive inotropic action of ET‐1 in ventricular myocytes is generally accepted, the underlying mechanisms are controversial, while numerous cellular targets are involved (Braz et al. 2004; Li et al. 2005; Domeier et al. 2012; Smyrnias et al. 2018). In our hands, the inotropic responses triggered by ET‐1 in WT cells (Figs 4 and 7) might be partially mediated by PKC during the ET‐1/GPCR pathway stimulation. In the presence of ET‐1, the SR‐Ca²⁺ leak increases together with the SR‐Ca²⁺ content; this discrepancy could be explained by the positive inotropic PKC‐related effects on the LTCCs. As a consequence, increased Ca²⁺ influx and SERCA activity during the pacing protocol may lead to the observed increased SR‐Ca²⁺ content (Fig. 2). This may also have some relevance for the leak protocol (Fig. 7). Interestingly, application of 2‐APB in the presence of ET‐1 restores the SR‐Ca²⁺ leak to basal levels, while the SR‐Ca²⁺ content remains elevated (Fig. 8). In other words, the leak effect trigged by ET‐1 is IP₃R dependent. In addition, the ET‐1‐triggered arrhythmogenic responses on a cellular level can be reversed by IP₃R2 block excluding significant PKC‐dependent side effects (e.g. LTCCs, NCX) and indicating that IP₃ICR seems to be the main mechanism responsible for the induced arrhythmic responses.

Analogously, the increase in Ca²⁺ spark and Ca²⁺ wave occurrence upon IP₃ stimulation may be partly explained by a functional bi‐directional interaction between IP₃R2 and RyR2 (Wullschleger et al. 2017). Therefore, IP₃ICR may directly trigger and/or amplify CICR resulting in the initiation and/or propagation of Ca²⁺ waves.

Under conditions of pathophysiological stress and the associated increased IP₃R2 expression, the scenario given above may be fundamentally altered. Neurohumoral GPCR stimulation (e.g. via ET‐1) serves as a positive inotropic reserve that partially functions via IP₃ICR activation. This may include sensitization of RyR2 to activating (“triggering”) cytosolic Ca²⁺ together with increases in SR‐Ca²⁺ load, events that cooperatively increase single RyR2 channel open probability (P _o) and hence the tendency to develop arrhythmogenic Ca²⁺ waves (Domeier et al. 2008). Hence activation of GPCR via ET‐1 unfolds a positive inotropic reserve at the expense of increased spontaneous SR‐Ca²⁺ release events. In line with this scenario, under increased IP₃R2 expression, an even higher probability of Ca²⁺ wave occurrence compared to the wild‐type situation would be expected. Surprisingly, the opposite response was evident: activation of IP₃ICR in conditions of IP₃R2 overexpression (TG) resulted in a decrease rather than an increase in Ca²⁺ wave occurrence. Furthermore, block of IP₃R2 increased the SR‐Ca²⁺ content, suggesting an inherent regulatory role of IP₃ICR in the control of the SR‐Ca²⁺ content. Thus, under conditions of enhanced IP₃ICR, we hypothesized a secondary regulatory mechanism operating through IP₃ICR activation and resulting in the reduction of Ca²⁺ wave occurrence.

Global cytosolic Ca²⁺ waves are composed of the temporal and spatial summation of individual CICR events (Ca²⁺ sparks) that occur by spontaneous but concerted openings of RyR2 clusters (Cheng et al. 1993). Analogously, IP₃R2s are organized in clusters prerequisite to the generation of local Ca²⁺ events known as Ca²⁺ puffs (Yao et al. 1995). Analysis of individual Ca²⁺ events may reveal the contribution of IP₃ICR (e.g. Ca²⁺ puffs) to the Ca²⁺ wave occurrence during IP₃R stimulation. Under equal cytosolic Ca²⁺ levels, TG cells displayed a higher Ca²⁺ spark frequency (Fig. 5 C) than cells from WT littermates, reflecting a higher basal RyR2 sensitization. In agreement with others (Domeier et al. 2008), IP₃ stimulation in WT cells induced an expected increase in RyR activity. Surprisingly, the effect was reversed in TG myocytes, in which IP₃R2 stimulation returned not an increase but a decrease in Ca²⁺ event occurrence and SR‐Ca²⁺ content. The results in CaSpF in permeabilized myocytes are thus in line with those obtained in Ca²⁺ wave occurrence within intact ventricular myocytes (Fig. 2), suggesting that (1) cell permeabilization did not disrupt the mechanism by which IP₃ may be altering SR‐Ca²⁺ release, and (2) discarding sarcolemmal components (e.g. LTCCs, Na⁺–Ca²⁺ exchanger) as the mechanisms directly involved in the reduction of Ca²⁺ wave occurrence.

IP₃ stimulation of TG permeabilized ventricular myocytes induced a decrease in Ca²⁺ event frequency together with a reduction of the SR‐Ca²⁺ content. To explain such discrepancy, we hypothesized that IP₃ may induce a change in single Ca²⁺ event properties, so that larger Ca²⁺ events carrying more Ca²⁺ may keep CaSpF steady but yet reduce the SR‐Ca²⁺ content. In this context, multi‐parametric analysis of Ca²⁺ event profiles has been recently used to distinguish between Ca²⁺ sparks and Ca²⁺ puffs (Hohendanner et al. 2015; Wullschleger et al. 2017). Recently we showed that, under similar experimental conditions, IP₃ application to permeabilized atrial myocytes triggered the appearance of a group of Ca²⁺ events that can be identified and classified – via multi‐parametric analysis of single Ca²⁺ event properties – as IP₃R2‐mediated Ca²⁺ puffs. In the present study, however, multi‐parametric analysis of Ca²⁺ event characteristics from permeabilized ventricular myocytes (Fig. 6 A) failed to identify and distinguish between two classes of events and thus the sole group detected was classified as RyR2 Ca²⁺ sparks. Hence, as no differences in Ca²⁺ event parameters were detectable and IP₃ application triggered a decrease in Ca²⁺ event occurrence, RyR2 Ca²⁺ sparks cannot account for the decrease in SR‐Ca²⁺ content.

SR‐Ca²⁺ leak and “eventless” SR‐Ca²⁺ release

The local SR‐Ca²⁺ release carried via stochastic opening of SR‐Ca²⁺ release channels independent from the beat‐to‐beat ECC mechanism is known as SR‐Ca²⁺ leak. The luminal SR‐Ca²⁺ has been shown to be a major regulatory factor controlling Ca²⁺ release channel P _o, hence the SR‐Ca²⁺ leak (Terentyev et al. 2002; Keller et al. 2007). Although SR‐Ca²⁺ release channels are grouped in clusters from which detectable Ca²⁺ events arise, a differential channel distribution has been proposed (Sobie et al. 2006). Individual RyR2s that do not belong in a cluster or small RyR clusters can still release Ca²⁺ from the SR in a manner that does not generate a Ca²⁺ event detectable by confocal microscopy (Lipp & Niggli, 1996; Brochet et al. 2011). Single or “rogue” RyR2s may modulate SR‐Ca²⁺ content with no apparent changes in Ca²⁺ event occurrence (Song et al. 2006). Analogously, singular or “rogue” IP₃Rs may modulate SR‐Ca²⁺ content in an “eventless” manner. Thus the SR‐Ca²⁺ leak mechanism includes both SR‐Ca²⁺ release events (Ca²⁺ sparks and Ca²⁺ puffs) and non‐detectable or “eventless” SR‐Ca²⁺ release. Our data in CaSpF showing a decrease in Ca²⁺ spark occurrence with no detectable changes in Ca²⁺ event characteristics suggest that the triggered IP₃ICR occurs in an “eventless” manner and thus contributes to the total SR‐Ca²⁺ leak without generating distinct Ca²⁺ events. In this scenario, we hypothesized that “eventless” IP₃ICR reduces SR‐Ca²⁺ content and ultimately modulates the cardiomyocytes’ propensity to Ca²⁺ waves. Indeed, by approaching an adapted SR‐Ca²⁺ leak/load protocol (Shannon et al. 2002) we revealed a significant IP₃R2‐dependent component of the SR‐Ca²⁺ leak suggesting that albeit being largely outnumbered by RyR2s, IP₃R2s in wild‐type ventricular myocytes can still contribute to the total SR‐Ca²⁺ efflux when recruited by GPCR agonists.

Regarding the leak/load protocol, interpretation of the data is less straightforward than expected. Strong evidence of pronounced SR‐Ca²⁺ leak mediated by IP₃ICR is shown by IP₃R block (2‐APB) in TG cells, where 2‐APB triggers a pronounced decay of the total SR‐Ca²⁺ leak, especially when compared to the antagonists’ effect in WT (Fig. 8 C). In the latter, no differences to the control are detected whereas in the former a strong decrease in the basal SR‐Ca²⁺ leak is seen, indicative of increased IP₃ICR under IP₃R2‐overexpression conditions. Now, in TG cells, an intuitively expected SR‐Ca²⁺ leak increase was virtually absent upon ET‐1 stimulation. When compared to WT, in TG cells, tetracaine revealed a much smaller basal RyR2‐dependent SR‐Ca²⁺ leak (Fig. 6). Under these specific conditions SERCA works close to the cytosolic steady‐state [Ca²⁺] (F ₀) and thus, additional or excess IP₃R‐mediated SR‐Ca²⁺ leak is immediately compensated by SERCA activity. Because of the closed system (NCX block), cytosolic Ca²⁺ levels above the steady‐state (F ₀) cannot be expected, as also shown by others (Shannon et al. 2002; Curran et al. 2007; Gonzalez et al. 2007; Santiago et al. 2010; Bers, 2014; Tzimas et al. 2017). Hence the ET‐1 effect is minor or not detectable in TG myocytes as it is masked by SERCA activity. On the contrary, in WT cells, as tetracaine application sharply decreases the cytosolic Ca²⁺ (F _tet) far from steady‐state levels (F ₀), extra SR‐Ca²⁺ release (triggered by ET‐1) is readily detectable. In other words, a direct comparison of ΔCa²⁺ _cyt WT versus TG is inapplicable and only feasible in combination with IP₃R2 blocking conditions, as suggested.

In addition, our data in TG cells show that the RyR2 protein down‐regulation detected concomitantly diminishes the receptors’ contribution to the basal SR‐Ca²⁺ leak. One would expect that the RyR2 down‐regulation seen under remodelling conditions goes in parallel with an increased ECC gain in order to stabilize the CICR as long as possible. Indeed, that was the case. The “remaining” RyR2 Ca²⁺ release units are more sensitive to the trigger Ca²⁺ resulting in a higher P _o. That would imply a significantly higher RyR2 Ca²⁺ leak and/or a significant increase in Ca²⁺ spark occurrence. However, that was not observed under IP₃R2 overxpression conditions and can be used as additional evidence that the observed SR‐Ca²⁺ leak was not due to increased RyR2 P _o alone. Moreover IP₃R block induces a large reduction of the total SR‐Ca²⁺ leak suggesting that IP₃R2 P _o compensates at least in part the loss of RyR2‐mediated SR‐Ca²⁺ leak. This view is supported by increases in the SR‐Ca²⁺ content only seen upon IP₃R inhibition in IP₃R2‐overexpressing cells.

Taken together our results indicate that the contribution of IP₃ICR to the total SR‐Ca²⁺ release is mainly driven by “eventless” SR‐Ca²⁺ leak rather than by localized Ca²⁺ events. We hypothesize that this may be because the low expression of IP₃R2s, even in pathologically high expressing conditions, may not suffice to group the receptors in clusters large enough to form discernible Ca²⁺ events (e.g. Ca²⁺ puffs). By this mechanism, activation of enhanced IP₃ICR reduces the SR‐Ca²⁺ content below the threshold for Ca²⁺ spark generation, which prevents aberrant SR‐Ca²⁺ release and subsequently stabilizes intracellular Ca²⁺ homeostasis. In conclusion, the data presented reveal a novel mechanism by which activation of IP₃R2s in a model of cardiac pathophysiology serves as an anti‐arrhythmic mechanism that fine‐tunes the SR‐Ca²⁺ content via “eventless” SR‐Ca²⁺ release.

Limitations of the study

At present, elucidating cardiac IP₃‐dependent regulatory mechanisms is mainly limited to observations at the level of global Ca²⁺ events (e.g. of Ca²⁺ waves or Ca²⁺ transients) or by indirect proof of interaction by counting changes in Ca²⁺ event frequencies (Yamasaki et al. 2013). The reasons for this are: (1) in contrast to RyR Ca²⁺ signals, the smaller and rare IP₃ICR events are difficult to identify and to discern among the larger Ca²⁺ responses that originate from CICR, and (2) identifying the specific contribution of IP₃R2s in local and/or global Ca²⁺ responses is impeded by a deficiency of highly selective pharmacological tools (Gafni et al. 1997; Bootman et al. 2002). Although studies have fostered a contribution of IP₃ICR on local Ca²⁺ signalling in cardiomyocytes, the observation of individual functional interactions of Ca²⁺ sparks and Ca²⁺ puffs was just recently successful (Wullschleger et al. 2017). In addition, immunocytochemistry largely failed to identify the precise localization of individual IP₃R2s in mouse cardiac preparations, presumably because of limited antibody specificity. This limitation thus impedes the confirmation of the “rogue” IP₃R2 function hypothesis (data not shown; Mohler et al. 2005; Kockskämper et al. 2008; Nakayama et al. 2010; Wullschleger et al. 2017).

Transgenic mouse models are sufficient and appropriate tools to examine regulatory mechanisms of CICR in cardiomyocytes. However, there is no doubt that a 1:1 extrapolation of findings emerging from animal studies to the human situation has to be critically evaluated. There are only a few studies comparing cellular remodelling situations in human cardiac disease with TG animal models used for specific aspects of cardiac remodelling both on global and cellular levels. Not only that, at present ECC under physiological conditions in human cardiomyocytes is only rudimentarily characterized (Signore et al. 2013). The TG mouse model used in the present study reveals cardiac hypertrophy with several similarities to a human pathology. However, it was used as an appropriate tool to study a potential regulatory function of IP₃ICR in CICR. A modulatory mechanism based on IP₃‐dependent SR‐Ca²⁺ leak has been unmasked by overexpression of IP₃R2 which would otherwise have been potentially overlooked. Although several studies focus on IP₃ICR as a potential pro‐arrhythmogenic factor (Mackenzie et al. 2002; Proven et al. 2006) this model shows that higher expression of the IP₃R2 might actually be beneficial for the preservation of ECC within pathology.

Additional information

Competing interests

None declared.

Author contributions

All experiments were performed at the Department of Physiology, University of Bern, Switzerland. M.E. conceived and designed the experiments. J.B. collected and analysed the experimental data. M.E. and J.B. drafted and revised the article. All authors approved the final version for publication and agree to be accountable for all aspects of the work. All persons designated as authors qualify for authorship, and all those who qualify for authorship are listed.

Funding

This work was supported by the Swiss National Science Foundation (31‐149301) and Novartis Research Foundation to M.E.

Biography

Joaquim Blanch i Salvador received his BS and MS degrees in Biomedical Sciences and Biotechnology from the University of Barcelona. After graduating, he joined Professor Dr M. Egger's group in the Department of Physiology of the University of Bern as a PhD student. Combining electrophysiology techniques with confocal Ca²⁺ imaging and molecular biology approaches, his research focuses on the study of intracellular Ca²⁺ oscillations within cardiomyocytes in a framework of cardiac pathology.