Impaired Intracellular Calcium Buffering Contributes to the Arrhythmogenic Substrate in Atrial Myocytes From Patients With Atrial Fibrillation

Abstract

BACKGROUND:

Alterations in the buffering of intracellular Ca²⁺, for which myofilament proteins play a key role, have been shown to promote cardiac arrhythmia. It is interesting that although studies report atrial myofibrillar degradation in patients with persistent atrial fibrillation (persAF), the intracellular Ca²⁺ buffering profile in persAF remains obscure. Therefore, we aimed to investigate the intracellular buffering of Ca²⁺ and its potential arrhythmogenic role in persAF.

METHODS:

Transmembrane Ca²⁺ fluxes (patch-clamp) and intracellular Ca²⁺ signaling (fluo-3-acetoxymethyl ester) were recorded simultaneously in myocytes from right atrial biopsies of sinus rhythm (Ctrl) and patients with persAF, alongside human atrial subtype induced pluripotent stem cell–derived cardiac myocytes (iPSC-CMs). Protein levels were quantified by immunoblotting of human atrial tissue and induced pluripotent stem cell–derived cardiac myocytes. Mouse whole heart and atrial electrophysiology were measured on a Langendorff system.

RESULTS:

Cytosolic Ca²⁺ buffering was decreased in atrial myocytes of patients with persAF because of a depleted amount of Ca²⁺ buffers. In agreement, protein levels of selected Ca²⁺ binding myofilament proteins, including cTnC (cardiac troponin C), a major cytosolic Ca²⁺ buffer, were significantly lower in patients with persAF. Small interfering RNA (siRNA)–mediated knockdown of cTnC (si-cTNC) in atrial iPSC-CM phenocopied the reduced cytosolic Ca²⁺ buffering observed in persAF. Si-cTnC treated atrial iPSC-CM exhibited a higher predisposition to spontaneous Ca²⁺ release events and developed action potential alternans at low stimulation frequencies. Last, indirect reduction of cytosolic Ca²⁺ buffering using blebbistatin in an ex vivo mouse whole heart model increased vulnerability to tachypacing-induced atrial arrhythmia, validating the direct mechanistic link between impaired cytosolic Ca²⁺ buffering and atrial arrhythmogenesis.

CONCLUSIONS:

Our findings suggest that loss of myofilament proteins, particularly reduced cTnC protein levels, causes diminished cytosolic Ca²⁺ buffering in persAF, thereby potentiating the occurrence of spontaneous Ca²⁺ release events and atrial fibrillation susceptibility. Strategies targeting intracellular buffering may represent a promising therapeutic lead in persAF management.

Clinical Perspective.

What Is New?

• Here we provide the first in-depth analysis of cytosolic Ca²⁺ buffering in human atrial cardiac myocytes.

• We demonstrate that cytosolic Ca²⁺ buffering is reduced in persistent atrial fibrillation (persAF), which promotes the occurrence of arrhythmogenic Ca²⁺ waves and AF maintenance.

• By showing that myofibrillar degradation, particularly reduced expression of cTnC (cardiac troponin C), is a major contributor to altered Ca²⁺ buffering in persAF, we provide a novel mechanistic link between contractile dysfunction and the proarrhythmic substrate in atrial cardiac myocytes from patients with persAF.

What Are the Clinical Implications?

• Modulation of the intracellular Ca²⁺ buffering provides a novel target that could be exploited in the treatment of persAF.

• Clinically approved Ca²⁺ sensitizers such as levosimendan or omecamtiv mecarbil and nutritional supplements like taurine and β-alanine, possessing buffering modulatory properties, could be valuable additions to currently available therapeutics used in persAF management.

Editorial, see p 560

Atrial fibrillation (AF) is the most frequent cardiac arrhythmia and is associated with increased mortality and morbidity.¹ Despite recent advances in the understanding of the molecular mechanisms underlying AF pathophysiology, treatment remains challenging, particularly because of the self-promoting remodeling induced by AF.²

Altered intracellular Ca²⁺ handling is a key contributor to AF-associated remodeling.^3–5 In healthy cardiac myocytes, Ca²⁺ enters the cell during systole through voltage-gated L-type Ca²⁺ channels (I_Ca,L) and induces a much larger Ca²⁺ release from the sarcoplasmic reticulum (SR) through SR Ca²⁺ release channels (ryanodine receptors, RyR2 [cardiac ryanodine receptor type 2]). The released Ca²⁺ binds to cTnC (cardiac troponin C) and induces contraction of myofilaments. During diastole, Ca²⁺ is pumped back into the SR by SERCA (SR Ca²⁺-ATPase) and is extruded from the cell predominantly by the Na⁺-Ca²⁺ exchanger (NCX). In AF, normal physiological Ca²⁺ cycling is disturbed, and RyR2 fails to remain closed during diastole.^6–8 This leads to spontaneous diastolic Ca²⁺ releases (spontaneous Ca²⁺ release events [SCaEs]), including sparks and waves, which are thought to play a major role in AF initiation and maintenance.^3,9

It is important to note that only about 1% of cytoplasmic Ca²⁺ is free, whereas the remainder is bound to cytoplasmic Ca²⁺ buffers.¹⁰ Therefore, it is not surprising that even minor changes in the Ca²⁺ binding properties of Ca²⁺ buffers may have a huge impact on the size and kinetics of free Ca²⁺ transients (CaT). In addition, altered Ca²⁺ buffering has been suggested to be involved in the propagation of arrhythmogenic Ca²⁺ waves.^11,12

Because cTnC is one of the most important Ca²⁺ buffers^,10,13 and AF is associated with severe contractile dysfunction and degradation of myofilament proteins,^14–17 we hypothesize that Ca²⁺ buffering is reduced in atrial myocytes from patients with AF and that this could contribute to arrhythmogenesis in patients with AF.

In the present study, we quantified cytosolic Ca²⁺ buffering in atrial cardiac myocytes from right atrial samples of patients in sinus rhythm (control) and persistent AF (persAF). Using atrial subtype human induced pluripotent stem cell–derived cardiac myocytes (iPSC-CMs), we demonstrate that knockdown of cTnC leads to reduced Ca²⁺ buffering and increases incidence of SCaEs, thereby phenocopying the Ca²⁺ handling alterations observed in patients with AF.^3,4,7,8 Last, we use a new ex vivo mouse heart model to directly link reduced cytosolic Ca²⁺ buffering to increased atrial arrhythmogenesis. Taken together, we conclude that increasing cytosolic Ca²⁺ buffering may represent a novel therapeutic strategy to improve atrial contraction and reduce arrhythmogenesis in patients with AF.

METHODS

A detailed description of all methods is provided in the Supplemental Material.

Experimental protocols were approved by the ethics committees of Göttingen University (No. 10/9/15, 15/2/20 and No. 4/11/18) and conducted following the Declaration of Helsinki. Each patient gave written informed consent.

The data that support the findings of this study are available from the corresponding authors upon reasonable request.

Human Tissue Samples and Myocyte Isolation

Right atrial appendages were obtained from sinus rhythm patients (Ctrl) and patients in long-term persAF undergoing open-heart surgery (Tables S1 through S3). Excised right atrial appendages were either snap-frozen in liquid nitrogen for biochemical studies or subjected to a standard protocol^3,18 for myocyte isolation. Right atrial myocytes were suspended in EGTA-free storage solution for subsequent simultaneous measurement of cellular electrophysiology and [Ca²⁺]_i (Figure S1).

Cardiac Differentiation of Human iPSCs and Small Interfering RNA–Mediated cTnC Knockdown

Atrial iPSC-CMs were generated by subtype-directed differentiation of iPSCs from healthy donors, as previously described.^19–21 In brief, directed feeder-free cardiac differentiation of iPSC lines was achieved through canonical WNT modulation with small molecules CHIR (day 0) and IWP2, followed by metabolic selection with lactate. For atrial subtype specification, 1 µmol/L retinoic acid (Sigma Aldrich) was added between day 3 and day 6. Between day 27 and day 30, purified iPSC-CMs were digested with TrypLE (Thermo Fisher Scientific) and sparsely plated on 1:60 Matrigel-coated borosilicate glass 10-mm No. 0 round coverslips at a density of 15 000 cells/cm².

Human atrial iPSC-CMs in culture were transfected for 48 hours using predesigned small interfering RNAs (siRNAs) targeting the cardiac troponin C 1 gene (TNNC1; Thermo Fisher, s14273) or nonsilencing negative control siRNA (Thermo Fisher, 4390843) following provided manufacturer instructions. Transfected cells were used for subsequent downstream experiments.

Intracellular Calcium Measurement and Cellular Electrophysiology

Only rod-shaped myocytes with clear striations and defined margins, as observed in brightfield mode with the microscope ocular, were selected for measurements of [Ca²⁺]_i and cellular electrophysiology. [Ca²⁺]_i of right atrial myocytes was measured using the fluorescent Ca²⁺ indicator Fluo-3 AM, according to our previously published protocol.³ Simultaneously, the whole-cell ruptured patch-clamp technique was used to record membrane currents at 37 °C.³ Membrane currents were related to membrane capacitance and expressed in current density (pA/pF). Measurements of Ca²⁺ entry (integrated I_Ca,L) and SR Ca²⁺ content (integrated caffeine induced trainsient inward current) are expressed per liter total cell volume, which has been estimated based on a capacitance to volume relationship. Capacitance to volume relationship of atrial iPSC-CMs was estimated based on previously published data (4.57 pF/pL).²²

Ca²⁺ sparks in Fluo-4 AM–loaded atrial iPSC-CMs were measured in separate experiments using a LSM 5 confocal microscopy system (Carl Zeiss, Jena, Germany) with a 40× oil objective in line-scan mode (512 pixels, 37.5 μm, 1302 Hz, 10 000 cycles, pinhole 67 µm) and Zen 2009 acquisition software. Field stimulation (2 Hz) was applied to myocytes for approximately 20 s, after which confocal line scans were performed during rest.

Optical action potentials (AP) were measured under field stimulation in 0.1× Fluovolt (Thermo Scientific)–loaded atrial iPSC-CMs in a bath solution containing (in mmol/L) CaCl₂ 2, glucose 10, HEPES 10, KCl 4, MgCl₂ 1, NaCl 140; pH=7.35 was adjusted with NaOH, on the heated (37 °C) stage of an epifluorescence microscope (λ_ex=470 nm, λ_em=535 nm), optimized for high-speed photomultiplier signal capture.^23,24

Determination of Atrial Cell Volume

A modification of the approach used by Walden et al²⁵ was used to determine the capacitance to volume relationship in atrial myocytes, allowing Ca²⁺ fluxes and SR Ca²⁺ content to be expressed relative to total cell volume. Atrial myocytes from patients were loaded with the membrane-staining dye di-4-ANEPPS (2 μmol/L, 2 minutes) and imaged using serial z stacks (0.16 μm thickness, 63×1.2 NA water immersion objective, λ_ex=488 nm, and λ_em=500–783 nm). Cell volume was calculated by multiplying the cross-sectional area with the z interval and assuming an accessible fraction of 0.65.

Skinned Fiber Preparation and Force Measurements

Skinned fibers were prepared as described previously.²⁶ Resected muscle fibers were incubated for 24 hours in a 1% Triton X-100 solution for membrane permeabilization. After this skinning process, muscle strips were prepared (2–2.5 mm×0.3 mm).

For force measurements muscle fibers were installed in a force transducer system (Scientific Instruments, Heidelberg, Germany) and perfused with relaxation buffer containing (in mmol/L) imidazole 68, creatine phosphate 327, sodium azide 65, ethyleneglycol tetraacetic acid 380, MgCl₂ 203, dithioerythritol 154, ATP 605, and creatine kinase 400 U/mL. Free Ca²⁺ was increased stepwise according to Fabiato and Fabiato for measurement of contraction.²⁷

Biochemical Studies

Expression of myofilament proteins was quantified by immunoblot, as previously described,^3,28,29 and normalized to calsequestrin, which was unchanged in persAF compared with control samples (Tables S4 and S5).

Langendorff Experiments

All animal procedures were reviewed and approved by the Institutional Animal Care and Use Committee in compliance with European Union Directive 2010/63/EU, and with the current version of the German Law on the Protection of Animals. Mouse hearts were perfused retrogradely on a Langendorff system, and 3 electrodes were placed on the epicardial surface to measure whole-heart and atrial electrophysiology. In addition, a stimulating electrode was used. To reduce the threshold for atrial arrhythmia induction, hearts were perfused with solution containing decreasing concentrations of potassium (5.4, 3.7, and 2.0 mmol/L). Diazoxide (300 µmol/L)³⁰ was added to the perfusing solution (2.0 mmol/L potassium) as a final step. At each potassium concentration, various electrophysiological measurements were made, including the absolute refractory period and reaction to step burst pacing (400–4500 bpm) and shorter periods of high-frequency burst pacing (6000 bpm). Inducibility and duration of arrhythmic activity were measured and quantified.

Statistical Analysis

Normally distributed data (Shapiro-Wilk normality test) were compared using the unpaired 2-tailed Student t test. Differences between unpaired data with unequal variances were assessed using the Welch t test, which is indicated in the legends of all relevant figures. Nonnormally distributed data and all data sets with n<10 were compared with the Mann-Whitney U test. The Kruskal-Wallis test, followed by a Dunn post hoc test, was used to assess differences between 3 or more experimental groups. The simultaneous influence of 2 independent factors was appraised by using a 2-way ANOVA, followed by the Fisher Least Significant Difference post hoc test. Kaplan-Meier curves were compared using the Gehan-Breslow-Wilcoxon test. P<0.05 was considered statistically significant.

RESULTS

Reduced Ca²⁺ Buffering in Atrial Cardiac Myocytes From Patients With persAF

To quantify total cytosolic Ca²⁺ concentrations and cytosolic Ca²⁺ buffering, it is necessary to express sarcolemmal Ca²⁺ fluxes relative to total cell volume. Cell volume of atrial cardiac myocytes could be quantified from z-stack images, as described above (Figure 1A). The calculated volume was found to be higher in myocytes from patients with persAF, which appears to be mainly a result of cellular elongation, whereas cell width was not different (Figure 1B). Cell capacitance was measured at the beginning of electrophysiological experiments and was found to be comparable in both groups (persAF: 120.0±14.4 pF, n/N myocytes/patients=21/13 versus control: 100.5±9.4, n/N=26/16; P=0.25). We calculated the capacitance to volume ratio based on mean values of both parameters (persAF: 12.6±1.9 pF/pL, control: 14.9±2.2 pF/pL) and further analyzed the association between them, which we found to be strongly linear (Figure S2). These ratios were then applied to all individual capacitance measurements to estimate the volume of each individual cell undergoing electrophysiological measurement.

Figure 1.

Characterization of atrial myocytes isolated from patients without (Ctrl) and with persAF. A, Example 3-dimensional reconstruction of confocal z-stack images of atrial myocytes from Ctrl and persAF stained with di-4-ANEPPS. B, Mean±SEM cell dimensions and volume of control and persAF myocytes. C, I_Ca,L-triggered CaT in control and persAF atrial myocytes; representative simultaneous recordings of I_Ca,L (upper, inset, voltage-clamp protocol, 0.5 Hz) and triggered CaT (Fluo-3, lower). D, Mean±SEM peak I_Ca,L (left) and integrated I_Ca,L (right). E, Mean±SEM diastolic and systolic [Ca²⁺]_i (left) and resulting CaT amplitude (middle), and time constant (τ) of decay (right). *P<0.05, **P<0.01, ***P<0.001 vs control. n/N=number of myocytes/patients. Normality of data was determined by Shapiro-Wilk test, whereas comparison was made using the Student t test with Welch correction and Mann-Whitney U test for normally and nonnormally distributed data, respectively. CaT indicates Ca²⁺ transient; Ctrl, control; and persAF, persistent atrial fibrillation.

Intracellular Ca²⁺ handling was investigated by simultaneous electrophysiological (whole-cell ruptured patch) and epifluorescence measurements (Figure 1C; Figure S3). A voltage-step protocol (0.5 Hz stimulation) was used to induce I_Ca,L, and in agreement with previous findings, both peak I_Ca,L amplitude and I_Ca,L integral were smaller in persAF versus Ctrl (Figure 1D). Consistent with previous findings,³ the diastolic Ca²⁺ levels tended to be higher in persAF, whereas amplitude of the I_Ca,L-triggered CaT was found to be smaller (Figure 1E).

In subsequent experiments, SR Ca²⁺ content was quantified; myocytes were stimulated for 3 to 5 minutes using the same protocol, after which they were clamped at –80 mV, and 10 mmol/L caffeine was applied, causing complete SR Ca²⁺ release (Figure 2A). Interestingly, the amplitude of the caffeine-induced CaT (“free” Ca²⁺) was comparable in persAF versus control (Figure 2B). However “total” Ca²⁺, calculated from the integral of the resulting inward NCX current and normalized to cell volume, was found to be lower in persAF (Figure 2C). Not only does this finding indicate smaller SR Ca²⁺ content in persAF but, taken together with the comparable amplitude of the caffeine-induced CaT, it points towards altered Ca²⁺ buffering properties in persAF. To investigate this further, intracellular buffering was quantified by plotting total Ca²⁺ against cytosolic free Ca²⁺ during the caffeine-induced CaT, as shown in the representative traces of Figure 2D. The data were fitted with a Michaelis-Menten buffer curve:

Figure 2.

Caffeine-induced CaT and corresponding transient inward current (I_NCX) to assess SR Ca²⁺ content and buffering properties of atrial myocytes isolated from patients without (control) and with persAF. A, Representative caffeine-induced CaT (upper), associated I_NCX (middle) and integral of inward current, corrected for cell volume to give a measure of total Ca²⁺ (lower). B, Mean±SEM amplitude and time constant (τ) of decay of caffeine-induced CaT. C, Mean±SEM calculated total Ca²⁺. D, Representative buffer curves showing the relationship between cytosolic free Ca²⁺ and total Ca²⁺, fitted with a hyperbolic function. E, Mean±SEM maximum buffering capacity (B_max, left) and dissociation constant (K_d, right), determined from buffer curves. F, Mean (line) ±SEM (shaded) of calculated individual total buffer power curves as a function of free [Ca²⁺]_i. **P<0.01, ***P<0.001 vs control. n/N=number of myocytes/patients. Normality of data was determined by Shapiro-Wilk test and comparison was made using the Mann Whitney U test. Ctrl indicates control; and persAF, persistent atrial fibrillation.

$${\displaystyle \begin{array}{l}{\displaystyle {\left[C{a}^{2+}\right]}_{total}=\frac{B_{max}\cdot {\left[C{a}^{2+}\right]}_i}{K_d+{\left[C{a}^{2+}\right]}_i}}\end{array}}$$

Maximum buffering capacity (B_max) was found to be significantly lower in persAF versus control (Figure 2E), suggesting fewer cytosolic Ca²⁺ buffers, whereas the dissociation constant, K_d, was comparable in both groups. Figure 2F shows total buffer power (β) (see review by Smith and Eisner¹⁰), which is defined as the change of total Ca²⁺ divided by that of free Ca²⁺:

$${\displaystyle \begin{array}{l}{\displaystyle \beta =\frac{d\left[C{a}_T\right]}{d\left[C{a}^{2+}\right]}=\frac{B_{max}\cdot K_d}{{({\left[C{a}^{2+}\right]}_i+K_d)}^2}}\end{array}}$$

As previously described, total Ca²⁺ buffering represents the sum of the intrinsic Ca²⁺ buffering of the cardiac myocytes and the Ca²⁺ buffering provided by the added Fluo-3.³¹ However, B_max remained lower in AF after correcting for the contribution of Fluo-3 to intracellular Ca²⁺ buffering (Figure S4).

Reduced Expression of cTnC Contributes to Impaired Ca²⁺ Buffering in persAF

cTnC is a major cytosolic Ca²⁺ buffer; therefore, impairment of cTnC interaction with Ca²⁺ could alter the regulation of [Ca²⁺]_i and cardiac contraction. The expression and phosphorylation of key myofilament proteins, which influence cytosolic Ca²⁺-myofilament interaction, were determined (Figure 3A; Figure S5; Figure S6). It is interesting that expression of the cardiac troponins cTnC and cTnI (cardiac troponin I) was lower in persAF, whereas phosphorylation of cTnI was comparable between groups. Because the cardiac troponins, particularly cTnC, mediate Ca²⁺-binding to myofilaments, lower troponin expression likely contributes to the smaller B_max observed in persAF. The expression of cMyBP-C (cardiac myosin binding protein-C) was also lower in persAF versus Ctrl, with preserved phosphorylation levels (Figure 3A), which, together with reduced expression of cardiac troponins, points to a loss of myofilament proteins. Furthermore, expression of Myh6 (myosin heavy chain 6) and Tm1 (tropomyosin 1) were lower in persAF, whereas expression of Tm2 (tropomyosin 2) and α-actin were comparable with Ctrl (Figure S5). Skinned muscle fibers from persAF exhibited lower maximum force than control (Figure 3B), which may contribute to impaired fractional shortening of isolated cardiac myocytes from patients with persAF (Figure S3). Ca²⁺ sensitivity of force generation (pCa₅₀) was higher in atrial muscle fibers from patients with persAF and may represent a compensatory mechanism for the reduced maximum force in persAF. Phosphorylation of MLC2a (atrial isoform of myosin light chain) and desmin was increased in persAF (Figure S6) and may contribute to increased myofilament Ca²⁺ sensitivity.¹⁷

Figure 3.

Myofilament protein expression and contractile response to cytosolic Ca²⁺ in control and persAF. A, Immunoblots (upper left) and quantification (upper right) of cMyBP-C (cardiac myosin binding protein-C), its phosphorylated state (P-cMyBP-C), and cTnC (cardiac troponin-C) in atrial samples from controls and patients with persAF, normalized to CSQ (calsequestrin), except for P-cMyBP-C, which was normalized to total cMyBP-C. Immunoblots (lower left) and quantification (lower right) of cTnI (cardiac troponin I) and its phosphorylated state (P-cTnI) in atrial samples from controls and patients with persAF, normalized to CSQ and total cTnI, respectively. B, Absolute (left) and normalized (right) force-pCa relationship of skinned muscle fibers of controls and patients with persAF with mean±SEM of maximum force (F_max) and calcium sensitivity (pCa₅₀). *P<0.05 vs control. n=number of patients. Normality of data was determined by Shapiro-Wilk test, whereas comparison was made using the Student t test with Welch correction. Ctrl indicates control; and persAF, persistent atrial fibrillation.

In addition to cTnC, SERCA is also an important buffer of cytosolic Ca²⁺ in cardiac myocytes.¹⁰ To quantify SERCA activity independent of cytosolic Ca²⁺ buffering, the difference in the decay of both the total caffeine-induced CaT and total systolic CaT (calculated from buffering properties) was analyzed, as previously described.^23,32 Figure 4A shows example plots of the rate of decay of total Ca²⁺ against free Ca²⁺ concentration during the systolic CaT. The mean slope of this relationship was comparable in control and persAF. Decay of caffeine-induced CaT is SERCA-independent and results from sarcolemmal Ca²⁺ extrusion, mainly through NCX (Figure 2A). When plotting rate of decay of total Ca²⁺ against free Ca²⁺ concentration during the caffeine-induced CaT, the slope was higher in persAF, indicating faster removal of Ca²⁺ by NCX in persAF (Figure 4B). Contribution by SERCA could be ascertained by calculating the difference of the slope gradients in Figure 4A and 4B. It could be shown that SERCA contribution was comparable in control (60.7%±3.1%) and persAF (55.1%±2.6%; Figure 4C), indicating that altered buffering capacity in persAF is unlikely a result of changes in SERCA. This leaves the reduction in cTnC expression as an explanation for the reduced Ca²⁺ buffering capacity.

Figure 4.

Quantification of decay of total Ca²⁺ in atrial myocytes isolated from patients without (control) and with persAF. A, Representative rate of decay of total Ca²⁺ (-d[Ca²⁺]_T/dt) plotted during systolic CaT against free [Ca²⁺]_i (left) and slope of -d[Ca²⁺]_T/dt plotted against [Ca²⁺]_i (right). B, Representative rate of decay of total Ca²⁺ during caffeine-induced Ca²⁺ transient (-d[Ca²⁺]_T/dt) plotted against the corresponding free [Ca²⁺]_i (left) and slope of -d[Ca²⁺]_T/dt during caffeine plotted against corresponding [Ca²⁺]_i (right). C, Difference between slopes in A and B, indicating unaltered [Ca²⁺]_i dependence of SERCA-mediated Ca²⁺ removal. *P<0.05 vs control. n/N=number of myocytes/patients. Normality of data was determined by Shapiro-Wilk test, whereas comparison was made using the Student t test with Welch correction and Mann-Whitney U test for normally and nonnormally distributed data, respectively. Ctrl indicates control; and persAF, persistent atrial fibrillation.

siRNA-Mediated cTnC Knockdown in Atrial iPSC-CM Phenocopies Ca²⁺ Buffering Characteristics of persAF

To further explore whether reduced cTnC is responsible for the altered buffering observed in persAF, an atrial iPSC-CM model with reduced cTnC protein expression was used. Atrial iPSC-CMs were differentiated as previously described and confirmed for atrial-specific markers (MLC2a and I_K,ACh [acetylcholine activated inward rectifier K⁺ current]; Figure S7).^19–21 cTnC knockdown was mediated by siRNA targeting cTnC (si-cTnC), and for the control group, atrial iPSC-CMs were treated with nonsilencing siRNA (Figure S8). For simultaneous epifluorescence and electrophysiological measurement, atrial iPSC-CMs were loaded with Fluo-3 AM and stimulated by voltage-clamp control in the whole cell ruptured patch configuration (Figure 5A). I_Ca,L was induced by a voltage-step protocol (0.5 Hz stimulation), and peak current and current integral were found to be comparable in control and si-cTnC (Figure 5B), whereas the amplitude of the I_Ca,L-triggered CaT was larger in si-cTnC versus control (Figure 5C). iPSC-CMs were subsequently clamped at –80 mV, and 10 mmol/L caffeine was applied, causing total release of SR Ca²⁺ (Figure 5D). The amplitude of the caffeine-induced CaT was larger in si-cTnC; however, when the integral of the resulting inward NCX current was quantified and total Ca²⁺ was calculated, these parameters were found to be similar in si-cTnC and control, pointing to comparable SR Ca²⁺ load (Figure 5D through 5F). Intracellular buffering was quantified, as described previously, by plotting total Ca²⁺ against cytosolic free Ca²⁺ during the caffeine-induced CaT and fitting the data with a hyperbolic curve (Figure 5G). Analysis revealed that B_max was smaller in si-cTnC versus control, whereas K_d was comparable between both groups (Figure 5H; Figure S9), thus mimicking the Ca²⁺ buffering characteristics of persAF. In addition, buffer power was lower in si-cTnC compared with control (Figure 5I). Contribution of NCX and SERCA to cytosolic Ca²⁺ removal remained unaltered in si-cTnC iPSC-CMs (Figure S10).

Figure 5.

Ca2+ handling and Ca²⁺ buffering properties in atrial iPSC-CMs with normal (control) and reduced (si-cTnC) cTnC levels. A, Representative simultaneous recordings of I_Ca,L (upper, inset, voltage-clamp protocol, 1 Hz) and triggered CaT (lower). B, Mean±SEM peak I_Ca,L (left) and integrated I_Ca,L (right) in control (siRNA ns) and si-cTnC (siRNA cTnC) iPSC-CMs. C, Mean±SEM diastolic and systolic [Ca²⁺]_i (left) and resulting CaT amplitude (middle), and time constant (τ) of decay (right). D, Representative caffeine-induced CaT (upper), associated I_NCX (middle) and integral of inward current, corrected for cell volume to give a measure of total Ca²⁺ (lower). E, Mean±SEM amplitude (left) and time constant (τ) of decay (right) of caffeine-induced CaT. F, Mean±SEM calculated total Ca²⁺. G, Buffer curves showing the relationship between cytosolic free Ca²⁺ and total Ca²⁺, fitted with a hyperbolic function. H, Mean±SEM maximum buffering capacity (B_max, left) and dissociation constant (K_d, right), determined from buffer curves. I, Mean±SEM of calculated individual total buffer power curves as a function of free [Ca²⁺]_i. *P<0.05, **P<0.01 vs control. n=number of myocytes (2–4 differentiations). Normality of data was determined by Shapiro-Wilk test, whereas comparison was made using the Student t test and Mann-Whitney U test for normally and nonnormally distributed data, respectively. Ctrl indicates control; cTnC, cardiac troponin C; CaT, Ca²⁺ transient; iPSC-CM, induced pluripotent stem cell–derived cardiac myocyte; ns, nonsilencing; and siRNA, small interfering RNA.

Potential Proarrhythmic Consequences of Reduced cTnC Levels

It has previously been shown that increased Ca²⁺ leak from the SR provides a basis for arrhythmogenesis in persAF.^3,7,8 To investigate whether reduced cytosolic Ca²⁺ buffering, specifically in the form of reduced cTnC expression, could play a proarrhythmic role, diastolic SR Ca²⁺ release was measured and quantified in atrial si-cTnC versus control (siRNA nonsilencing) iPSC-CMs. To this end, confocal line scans were performed on atrial iPSC-CMs during rest, after a brief period (20 s) of field stimulation (2 Hz). Line scan analysis revealed significantly higher Ca²⁺ spark frequency in si-cTnC compared with control (Figure 6A and 6B), indicating increased SR Ca²⁺ leak. In addition, we found increased SR Ca²⁺ leak using the previously described tetracaine protocol³³ (Figure S11). Moreover, the leak-load relationship was shifted leftward in persAF, indicating increased SR Ca²⁺ leak at any given SR Ca²⁺ content. Immunoblot analysis revealed comparable expression levels of RyR2, CaMKII, and junctophilin-2, as well as phosphorylated RyR2 and CaMKII, in si-cTnC compared with control iPSC-CMs (Figure S12).

Figure 6.

Incidence of Ca²⁺ sparks and action potential (AP) alternans in atrial iPSC-CMs with normal (control) and reduced (si-cTnC) cTnC levels. A, Representative confocal line scans showing SR Ca²⁺ release in the form of Ca²⁺ sparks in control (siRNA ns) and si-cTnC (siRNA cTnC) iPSC-CMs. B, Mean±SEM Ca²⁺ spark frequency (CaSpF, left) and amplitude (right). C, Representative normalized traces of AP at 0.5 Hz (upper) and 2 Hz (lower) in control (left) and si-cTnC iPSC-CMs. D, AP duration at 90% repolarization (APD₉₀) at increasing diastolic intervals (AP restitution), fitted with a 1-phase association nonlinear function to determine maximum curve slope. E, Kaplan-Meier plot indicating the percentage of iPSC-CMs without alternans in relation to the respective pacing frequency. F, Mean±SEM alternans threshold frequency. Number of myocytes without AP alternans are shown in boxes above. ***P<0.001, *P<0.05 vs control. n=number of myocytes (2 or 3 differentiations). Comparison was made using the unpaired Student t test, the Mann-Whitney U test, and the Gehan-Breslow-Wilcoxon test (E). Ctrl indicates control; cTnC, cardiac troponin C; iPSC-CM, induced pluripotent stem cell–derived cardiac myocyte; ns, nonsilencing; and siRNA, small interfering RNA.

In accordance with our in vitro experiments, computational modeling revealed that reduction of cytosolic Ca²⁺ buffering increases incidence of Ca²⁺ waves and delayed after depolarizations (Figure S13).^4,34,35

Optical AP measurements revealed no difference in AP duration or restitution between control and si-cTnC iPSC-CMs (Figure 6C and 6D). However, the incidence of AP alternans at lower pacing frequencies was significantly larger in cells with reduced cTnC (Figure 6E and 6F). The maximum slope of the restitution curve did not exceed 1 in either group, indicating that voltage alternans of the AP is driven by intracellular Ca²⁺ aberrations.

Taken together, these data demonstrate that cTnC reduction is sufficient to reproduce the phenotype of impaired Ca²⁺ buffering observed in persAF, including increased SR Ca²⁺ leak and alternans, both of which are strongly arrhythmogenic.^3,7,8

To ascertain whether improving Ca²⁺ buffering can ameliorate increased SR Ca²⁺ leak in si-cTnC iPSC-CMs, si-cTnC iPSC-CMs were pretreated with the Ca²⁺ sensitizer EMD57033 (5 µmol/L, pretreatment for 5 minutes),²³ and confocal line scan analysis was repeated. Although the amplitude of Ca²⁺ sparks was similar between treated and nontreated groups, Ca²⁺ spark frequency in si-cTnC iPSC-CMs was significantly reduced by EMD57033 (Figure 7A and 7B). Furthermore, Ca²⁺ spark frequency in the EMD57033-treated group was similar to that in control iPSC-CMs, thus confirming that increasing buffering can normalize SR Ca²⁺ leak.

Figure 7.

Effect of Ca²⁺ sensitization on Ca²⁺ sparks in atrial iPSC-CMs with reduced (si-cTnC) cTnC levels. A, Representative confocal line scans of atrial iPSC-CMs with normal (control, siRNA ns) and reduced (si-cTnC, siRNA cTnC) cTnC levels, pretreated with EMD57033 (EMD, 5 µmol/L). B, Mean±SEM Ca²⁺ spark frequency (CaSpF, left) and amplitude (right). *P<0.05, **P<0.01 vs control and si-cTnC. n=number of myocytes (2 differentiations). Comparison was made using the Kruskal-Wallis test followed by the Dunn post hoc test. Ctrl indicates control; cTnC, cardiac troponin C; iPSC-CM, induced pluripotent stem cell–derived cardiac myocyte; ns, nonsilencing; and siRNA, small interfering RNA.

Desensitization of Myofilaments to Ca²⁺ Causes Atrial Arrhythmia

To determine whether reduced cytosolic Ca²⁺ buffering by myofilaments alone can induce arrhythmic activity in the atria, blebbistatin was applied to Langendorff-perfused mouse hearts to reduce myofilament Ca²⁺ sensitivity, during which atrial electrograms were recorded (Figure S14). Blebbistatin significantly increased the inducibility of atrial arrhythmic activity after burst-pacing (potassium in perfusing solution: 2 mmol/L, both with and without diazoxide, Figure 8A and 8B). However, the duration of atrial arrhythmic episodes was comparable even in the presence of blebbistatin (Figure 8C). Figure 8D shows a Kaplan-Meier curve of the percentage of hearts without atrial arrhythmic activity after burst pacing plotted against decreasing potassium concentrations. Blebbistatin significantly altered this curve, pointing to a higher susceptibility to atrial arrhythmic activity.

Figure 8.

Langendorff experiments in mouse heart. A, Representative atrial electrogram traces showing effect of burst pacing (100 Hz) in the absence (control, left) and presence (right) of blebbistatin (5 µmol/L) (2 mmol/L K⁺ in both). B, Grouped bar chart showing mean±SEM inducibility of atrial arrhythmic activity for each potassium level, differentiated by the presence or absence of blebbistatin. The chart highlights significant main effects of blebbistatin, F(1, 72)=20.06, P<0.0001, and potassium, F(3, 72)=14.91, P<0.0001. C, Mean±SEM arrhythmic episode duration for each potassium level, in the presence or absence of blebbistatin. The graph highlights the significant effect of potassium, F(3, 36)=3.52, P<0.05. D, Kaplan-Meier plot showing the percentage of hearts without arrhythmic activity. *P<0.05, ***P<0.001 vs control. n=number of hearts. Comparison using 2-way ANOVA followed by a Fisher Least Significant Difference post hoc test (B and C) and the Gehan-Breslow-Wilcoxon test (D). Ctrl indicates control.

DISCUSSION

In the present study, we observed impaired cytosolic Ca²⁺ buffering in atrial myocytes from patients with persAF and analyzed the underlying molecular substrate and its contribution to atrial arrhythmogenesis. Analysis of transmembrane Ca²⁺ fluxes during systolic and caffeine-induced CaT enabled estimation of total Ca²⁺ in relation to free cytosolic Ca²⁺. Our experiments revealed reduced total Ca²⁺ buffering capacity in persAF, likely because of degradation of myofilament proteins, which represent a major Ca²⁺ buffer in cardiac myocytes. Myofilaments play an important role in cytosolic Ca²⁺ handling, and therefore, altered myofilament expression may have direct consequences on cytosolic Ca²⁺ homeostasis. Here we demonstrate, for the first time, that reduction of cTnC expression decreases intracellular Ca²⁺ buffering and increases the incidence of both SCaEs and AP alternans in atrial iPSC-CMs, thereby phenocopying the arrhythmogenic Ca²⁺ handling phenotype observed in atrial myocytes from patients with persAF.^3,7,8

Reduced Ca²⁺ buffering leads to a higher change in free cytosolic Ca²⁺ per total Ca²⁺ released from the SR and therefore amplifies consequences of higher incidence of SCaEs as a mechanism of increased ectopic activity in persAF. Last, reducing cytosolic Ca²⁺ buffering in an in vitro mouse model increased susceptibility to pacing-induced atrial arrhythmia, validating the direct mechanistic link between impaired cytosolic Ca²⁺ buffering and atrial arrhythmogenesis in clinical persAF.

Impaired Contractile Function in persAF

It has been established that persAF is associated with Ca²⁺ handling abnormalities that contribute to impaired contractility and arrhythmogenesis; reduced I_Ca,L has been widely shown to be a characteristic hallmark of AF-associated remodeling and a major contributor to AP shortening in persAF, promoting maintenance of reentry.^3,36–38 In addition, reduced I_Ca,L triggers smaller Ca²⁺ release from the SR, thereby contributing to impaired contractile function of atrial myocytes from persAF.

Impaired contractility is a major hallmark of AF-associated remodeling.^15,39,40 During an AF episode, contractile function of the atria is mainly constrained because of the fast and uncoordinated atrial excitation. However, the impaired contractile function persists for several weeks even after the cardioversion of AF back to sinus rhythm, leading to a high risk of atrial thromboembolism and stroke, despite sinus rhythm maintenance.⁴¹ Similarly, force of contraction is clearly reduced in atrial preparations from patients with AF when stimulated in vitro with a constant frequency.⁴⁰ Impaired Ca²⁺ handling^40,42 and structural remodeling, including increased fibrosis, have been suggested to contribute to AF-associated contractile dysfunction.² Other studies of AF have shown that reduced contractility even persists in myofibril preparations lacking sarcolemma and Ca²⁺ handling machinery,^16,17 demonstrating that reduction in myofibrillar maximal active tension in persAF must be attributable to defects in the myofilaments themselves. Furthermore, and in accordance with our results, it has been shown that loss of myofilament proteins is a major mechanism contributing to impaired atrial contractility.^{14,15,17,39,40,43,44} Meanwhile, mechanistic studies have revealed that degradation of cardiac troponins mainly results from increased activation of the Ca²⁺-dependent protease calpain during high atrial stimulation frequencies,^14,43,45 suggesting that calpain inhibition may represent a future therapeutic target in preventing AF-associated contractile remodeling.

In contrast with impaired maximal force development associated with AF, our present investigation, in accordance with a previous study,¹⁷ demonstrated higher myofilament Ca²⁺ sensitivity in atrial preparations from patients with persAF. Increased phosphorylation of MLC2a is likely a major contributor to the increased Ca²⁺ sensitivity.^46,47 However, results on phosphorylation of MyBP-C (thought to increase myofilament Ca²⁺ sensitivity) are controversial, showing increased,¹⁷ decreased,⁴⁸ or unaltered (present study) phosphorylation levels.

Determinants of Increased Incidence of SCaEs in persAF

Increased incidence of SCaEs originating from the SR during diastole is a well-accepted mechanism underlying enhanced ectopic activity, triggering AF episodes and contributing to AF progression and maintenance.^3,5,8,9 Ca²⁺ removal from the cytosol by NCX brings 3 Na⁺ ions into the cell per extruded Ca²⁺ ion. This is an electrogenic process, leading to membrane depolarizations (delayed afterdepolarizations) which, if large enough, could trigger a new AP, resulting in ectopic activity. Several mechanisms have been identified to amplify consequences of leaky RyR2 channels: (1) higher SR Ca²⁺ load, for example as a result of increased excitation frequencies in AF, escalates SR Ca²⁺ leak, because of the exponential increase of leak-load relationship⁴⁹; (2) expression and activity of NCX are increased in persAF, resulting in increased arrhythmogenic transient inward current in response to a given diastolic Ca²⁺ release from the SR³; (3) the distance between RyR clusters is reduced in AF, thereby enhancing the propagation of Ca²⁺ waves^50,51; (4) impaired buffering of cytosolic Ca²⁺ promotes occurrence of atrial SCaEs in persAF, as demonstrated in this study.

Atrial Arrhythmias and Ca²⁺ Buffering

It has been established that alterations in cytosolic Ca²⁺ buffering contribute to ventricular arrhythmias.^13,23,52 To the best of our knowledge, there are currently no studies investigating the role of reduced Ca²⁺ buffering in atrial arrhythmogenesis in humans. It is often overlooked that only 1% of cytosolic Ca²⁺ is free and detected by conventional Ca²⁺ indicators such as Fluo-3, whereas the remainder is bound to Ca²⁺ buffers.¹⁰ Because the myofilament protein cTnC is one of the major cytosolic Ca²⁺ buffers in cardiac myocytes,¹⁰ it can be assumed that even minor changes in Ca²⁺ binding to cTnC can have major effects on free cytosolic Ca²⁺, which could play an important role in cardiac arrhythmias. Early computational modeling studies, for example, suggested that Ca²⁺ buffers critically limit the diffusion of locally released Ca²⁺ and hamper activation of neighboring Ca²⁺ release sites, thereby preventing the occurrence of arrhythmogenic Ca²⁺ waves.⁵³ Accordingly, experimental reduction of cytosolic Ca²⁺ buffering caused increased Ca²⁺ sparks and intracellular propagation of arrhythmogenic Ca²⁺ waves.⁵⁴ Here we provide evidence that genetic downregulation of cTnC leads to significant reduction of cytosolic Ca²⁺ buffering and consecutively increased Ca²⁺ spark frequency. Furthermore, Ca²⁺ spark frequency could be reduced by pharmacological treatment with a Ca²⁺ sensitizing agent, EMD57033 (Figure 7). It is important to note that in si-cTnC iPSC-CMs, we found no difference, compared with control, in expression levels of RyR2, CaMKII, and junctophilin-2, nor the phosphorylated levels of RyR2 and CaMKII. Although RyR2 hyperphosphorylation is a well-accepted mechanism underlying increased diastolic Ca²⁺ leak in AF, the unaltered expression levels we found in our si-cTnC iPSC-CM model allowed us to focus on the contribution of reduced buffering to altered Ca²⁺ handling and arrhythmogenic activity.

Although several previous studies have investigated Ca²⁺ buffering properties in various animal models of AF, there have been contradictory results. Greiser et al found increased cytosolic Ca²⁺ buffering in a rabbit model after 5 days of atrial pacing.¹¹ They hypothesized that their observation of reduced cTnI phosphorylation, causing increased Ca²⁺-cTnC binding, prevents detrimental effects of Ca²⁺ overload induced by rapid pacing. One could speculate that this may be an early response to high atrial stimulation frequencies, whereas loss of myofilament proteins, including cTnC, may represent a hallmark of late-stage remodeling, as shown by others.^14,50,55 Interestingly, mavacamten, a recently approved compound for hypertrophic cardiomyopathy treatment, reduces myofilament Ca²⁺ affinity and was found to increase incidence of AF in the PIONEER-HCM study (Phase 2 Open-label Pilot Study Evaluating Mavacamten in Subjects With Symptomatic Hypertrophic Cardiomyopathy and Left Ventricular Outflow Tract Obstruction),^56,57 suggesting that reduced Ca²⁺ buffering may indeed facilitate occurrence of atrial arrhythmias.

We observed AP alternans at low pacing frequencies in atrial iPSC-CMs with reduced Ca²⁺ buffering because of cTnC knockdown. Primary AP alternans is assumed to arise during pathological (steep) AP restitution, which was not observed here, pointing to underlying Ca²⁺-driven alternans, which can canonically arise because of slow removal of Ca²⁺ into the SR, or in conditions of increased RyR2-mediated Ca²⁺ release into the cytosol. Because we detected significantly increased incidence of diastolic Ca²⁺ leak in atrial iPSC-CMs with reduced cTnC, we suggest that alternans arises at lower pacing rates, predominantly because of increased SCaEs, secondary to reduced buffer availability. This is in line with a previous modeling study suggesting that reduced cTnC-related Ca²⁺ buffering can increase Ca²⁺ alternans and also AP alternans.⁵⁸

In addition to demonstrating the cellular arrhythmogenic phenotype caused by cTnC downregulation, we established the first ex vivo mouse heart model to investigate effects of reduced buffering on atrial arrhythmogenesis; we provided evidence at the whole-organ level that myofilament desensitization to Ca²⁺ increases susceptibility to atrial arrhythmia. Therefore, we hypothesize that reduced cytosolic Ca²⁺ buffering may play an important role in the progression and maintenance of AF in patients with persAF.

Potential Limitations

In our study, we collected samples from only 1 atrial region (right atrial appendage). Our findings may therefore not apply fully to other regions of the atria. Expression of α-actin and Tm2, for example, which was unaltered in our study of right atrial appendages, has been shown to be increased and decreased in left atrial biopsies from patients with AF, respectively.⁴⁴ The individuals from whom we obtained tissue samples included only patients who underwent coronary bypass or valve replacement surgery. Such individuals have numerous comorbidities, and the phenotype of atrial cardiac myocytes from our Ctrl patients may be different from nondiseased controls. Furthermore, it is unclear whether the mechanisms identified here also apply to patients with persAF without any heart disease. However, impaired contractile function is a common clinical finding throughout all atrial regions and in all patients with persAF.^2,40 Furthermore, ectopic activity occurs in right and left atria from patients with persAF, suggesting that increased incidence of SCaEs occurs throughout the whole atria. Here we demonstrate for the first time a direct mechanistic link between atrial hypocontractility and ectopic activity. Future clinical studies will be required to reveal whether there is a direct correlation between local hypocontractility and occurrence of ectopic activity in the different regions of the atria.

Our data suggest that loss of myofilament proteins and reduced expression of the major Ca²⁺ buffer cTnC are important contributors to reduced Ca²⁺ buffering in atrial cardiac myocytes from patients with persAF. There are numerous additional buffers that may contribute to impaired cytosolic Ca²⁺ buffering in AF, including the giant sarcomere protein titin,⁵⁹ as well as the important buffers SERCA and the sarcolemma.¹⁰ However, activity of SERCA was comparable in persAF versus control cardiac myocytes investigated in our cohort (Figure 4), and cell capacitance as a marker for total membrane area was also comparable, making altered Ca²⁺ binding to SERCA and sarcolemma unlikely. Furthermore, the contribution of mitochondrial Ca²⁺ uptake to cytosolic Ca²⁺ buffering is controversial.^60–63 However, several studies have observed no impact of mitochondrial Ca²⁺ uptake on the amplitude of cytosolic Ca²⁺ transients, even during β-adrenergic stimulation where relevant Ca²⁺ accumulation into the mitochondrial matrix (to stimulate Krebs cycle dehydrogenases) occurs.^64–66 These observations render a significant contribution of mitochondria to cytosolic Ca²⁺ buffering unlikely. It needs to be considered that the concentration of cytosolic Ca²⁺ buffers is 3 times higher in the atria than in the ventricles.¹⁰ This may also explain the relatively high buffer power observed in the present study.¹⁰

Because neither genetic mouse models of impaired myofilament Ca²⁺ binding nor specific pharmacological Ca²⁺ desensitizing agents are available, in the present study, we used the myosin ATPase inhibitor blebbistatin to indirectly modulate myofilament Ca²⁺ sensitivity in Langendorff-perfused mouse hearts. Blebbistatin is widely used as a contractile uncoupling agent for electrophysiological studies and has been shown to reduce Ca²⁺ sensitivity (pCa₅₀) of mouse skinned fiber preparations in a concentration-dependent manner.⁵² Although minimal or nonsignificant effects on cardiac ion channels and APs by blebbistatin have been reported,⁶⁷ unspecific effects on ion channels cannot be completely excluded, in particular under the experimental conditions used in our model. However, atrial refractory periods, recorded at each experimental step (Figure S14B), did not differ between blebbistatin-treated and untreated hearts across the various K⁺ concentrations. This suggests that increased susceptibility to atrial arrhythmia after blebbistatin treatment is indeed a result of impaired cytosolic Ca²⁺ buffering. Nevertheless, mouse models in general do not phenocopy all important aspects of human cardiac electrophysiology. Further extensive work in large-animal AF models will be required to investigate the time course and pathophysiological role of impaired Ca²⁺ buffering under conditions similar to those in patients who develop persAF.

Conclusions and Potential Significance

Here we provide evidence for a direct link between impaired atrial contractility in persAF and arrhythmogenesis. Our data show that loss of myofilament proteins, including reduced expression of cTnC, leads to reduced cytosolic Ca²⁺ buffering, which promotes the occurrence of SCaEs and increases susceptibility to atrial arrhythmia.

Because loss of myofilament proteins represents a rather late phenomenon in AF remodeling, this mechanism is likely to contribute to the chronification of the arrhythmia. Furthermore, because free cytosolic Ca²⁺ increase is believed to play a major role in electrical and structural remodeling,^45,68,69 reduced cytosolic Ca²⁺ buffering could therefore intensify remodeling in response to atrial tachycardia. In summary, our data suggest that, although the arrhythmogenic mechanisms of impaired Ca²⁺ buffering are not yet fully understood, the development of new strategies, specifically targeting intracellular Ca²⁺ buffering, may open novel therapeutic avenues to prevent progression and maintenance of AF. Already existing Ca²⁺ sensitizers, such as levosimendan and the indirectly acting omecamtiv mecarbil, may represent promising lead compounds.^70,71 Furthermore, nutrient supplements such as β-alanine and taurine, which have been demonstrated to increase intracellular Ca²⁺ buffering in cardiac myocytes, may be considered as valuable additions to currently available therapeutics used in AF management.^72,73

ARTICLE INFORMATION

Acknowledgments

The authors are grateful to Lucie Carrier (Hamburg, Germany) for providing the antibody against phosphorylated cMyBP-C, and they would also like to thank David A. Eisner (Manchester, United Kingdom) for helpful discussion. Furthermore, the authors thank Ines Müller and Brigitte Korff for technical assistance and Maren Dilaj for secretarial help.

Sources of Funding

This work was supported by grants from the Deutsche Forschungsgemeinschaft to N.V. (DFG, VO 1568/3-1, VO1568/4-1) and to N.V. and S.E.L. (IRTG1816, SFB1002 and under Germany’s Excellence Strategy-EXC 2067/1-390729940), from the Else-Kröner-Fresenius Foundation to N.V. (EKFS, 2016_A20), from the German Center for Cardiovascular Research to N.V. (DZHK, 81X4300102, 81X4300115, 81X4300112) and to A.E. (D81X4300123), and from the Bundesministerium für Bildung und Forschung (BMBF) subproject of the German Network for RASopathy Research (GeNeRARe) to G.K. (01GM1902D). This work was further supported by scholarships from the Göttingen Promotionskolleg für Medizinstudierende, funded by the Jacob-Henle-Programm and the Else-Kröner-Fresenius-Stiftung to V.M. and from the German Academic Exchange Service to I.S. and N.I. We thank the Clinic for Cardiology and Pneumology at the University Medical Center Göttingen and the Deutsche Stiftung für Herzforschung (F/13/20) for funding provided to A.E.

Disclosures

None.

Supplemental Material

Supplemental Methods

Tables S1–S5

Figures S1–S14

Reference 74

Supplementary Material

cir-150-544-s001.pdf ^{(1.3MB, pdf)}

cir-150-544-s002.pdf ^{(4.9MB, pdf)}