Store-Operated Ca²⁺ Entry as a Putative Target of Flecainide for the Treatment of Arrhythmogenic Cardiomyopathy

Abstract

Arrhythmogenic cardiomyopathy (ACM) is a genetic disorder that may lead patients to sudden cell death through the occurrence of ventricular arrhythmias. ACM is characterised by the progressive substitution of cardiomyocytes with fibrofatty scar tissue that predisposes the heart to life-threatening arrhythmic events. Cardiac mesenchymal stromal cells (C-MSCs) contribute to the ACM by differentiating into fibroblasts and adipocytes, thereby supporting aberrant remodelling of the cardiac structure. Flecainide is an I_c antiarrhythmic drug that can be administered in combination with β-adrenergic blockers to treat ACM due to its ability to target both Na_v1.5 and type 2 ryanodine receptors (RyR2). However, a recent study showed that flecainide may also prevent fibro-adipogenic differentiation by inhibiting store-operated Ca²⁺ entry (SOCE) and thereby suppressing spontaneous Ca²⁺ oscillations in C-MSCs isolated from human ACM patients (ACM C-hMSCs). Herein, we briefly survey ACM pathogenesis and therapies and then recapitulate the main molecular mechanisms targeted by flecainide to mitigate arrhythmic events, including Na_v1.5 and RyR2. Subsequently, we describe the role of spontaneous Ca²⁺ oscillations in determining MSC fate. Next, we discuss recent work showing that spontaneous Ca²⁺ oscillations in ACM C-hMSCs are accelerated to stimulate their fibro-adipogenic differentiation. Finally, we describe the evidence that flecainide suppresses spontaneous Ca²⁺ oscillations and fibro-adipogenic differentiation in ACM C-hMSCs by inhibiting constitutive SOCE.

1. Introduction

Arrhythmogenic cardiomyopathy (ACM) is a genetic disease that leads to the progressive replacement of ventricular myocardium with fibrofatty scar tissue, thereby predisposing the patient to ventricular arrhythmias and sudden cardiac death (SCD) [1,2]. ACM may affect the right ventricle (arrhythmogenic right ventricular cardiomyopathy, ARVC), the left ventricle (arrhythmogenic left ventricular cardiomyopathy, ALVC), or both cardiac ventricles and is the main cause of SCD in young people and athletes [1,2]. A crucial role in the pathogenic mechanism of ACM can be played by the fibro-adipogenic differentiation of cardiac mesenchymal stromal cells (C-MSCs), which renders the cardiac tissue more likely to develop the arrhythmic events leading to the patient’s death [2,3]. Intriguingly, a recent investigation showed that the I_c antiarrhythmic drug (AAD) flecainide inhibited fibro-adipose differentiation in human C-MSCs (C-hMSCs) expanded from ACM patients (ACM C-hMSCs) by blocking store-operated Ca²⁺ entry (SOCE) [4], which represents the most important Ca²⁺ entry pathway in non-excitable cells [5,6,7], including MSCs [8,9,10]. Herein, we first describe the genetic background and therapeutic options of ACM. Then, we briefly recapitulate the main molecular mechanisms targeted by flecainide to mitigate arrhythmic events. Subsequently, we describe the role of intracellular Ca²⁺ oscillations in determining MSC fate. Next, we discuss our recent work showing that spontaneous Ca²⁺ oscillations in ACM C-hMSCs are accelerated to stimulate their fibro-adipogenic differentiation. Finally, we describe the evidence that flecainide suppresses spontaneous Ca²⁺ oscillations and fibro-adipogenic differentiation in ACM C-hMSCs by inhibiting constitutive SOCE and interfering with endoplasmic reticulum (ER) Ca²⁺ refilling.

2. Pathological Background and Treatment of ACM

Recently, an evidence-based reevaluation of published ARVC, the most common form of ACM, genes was carried out by experts in the field and showed that only a small number of genes encoding for desmosomal proteins, such as plakophilin-2 (PKP2), plakoglobin (JUP), and desmoglein 2 (DSG2), were definitively or moderately associated with ARVC (Table 1) [11]. Other genes that encode adherens junction proteins, such as cadherin 2 (CDH2) or catenin α3 (CTNNA3), that are also associated with ACM (Table 1) [1,2], only showed a moderate association with ACM [11]. Mutations in genes encoding proteins that are unrelated to intercellular junctions, such as cytoskeletal proteins, ion channels, and transporters, were also reported (Table 1) [1,2]; however, they were also shown to be weakly associated with ARVC [11]. This Gene Curation Expert Panel refuted the gene encoding for type 2 ryanodine receptor (RyR2), which has long been associated with ACM [1,2], as an ARVC gene [11].

Besides maintaining the mechanical stability of the heart and enabling the electrical coupling between ventricular cardiomyocytes, desmosomes may also play a crucial role in intracellular signal transduction and gene expression [2]. For instance, when the desmosomal assembly is disrupted by genetically defective proteins, plakoglobin 2 translocates from the intercalated discs to the nucleus, where it may compete with β-catenin to suppress the Wnt/β-catenin signalling pathway and elicit a gene transcriptional switch from myogenesis to fibrogenesis and adipogenesis [1,2,12]. A series of recent studies demonstrated that desmosomal mutations may also affect c-MSCs [13,14], which support the structural and functional integrity of the myocardium [15,16]. C-MSCs represent the primary source of myofibroblasts and adipocytes in the heart of ACM patients [13,14]. The fibrofatty replacement of ventricular myocardium may favour life-threatening arrhythmias, such as ventricular fibrillation and sustained ventricular tachycardia (VT), that ultimately cause SCD [2,3]. Patients who develop symptoms or have been diagnosed with ACM after a genetic screening or an electrocardiogram can be treated with implantable cardioverter-defibrillators (ICDs), radiofrequency catheter ablation, heart transplantation, β-blockers, or AADs [3,17]. Currently, ICDs represent the only therapeutic approach that can effectively reduce patient mortality, although they can be recommended only in the presence of clear arrhythmic risks, such as ventricular fibrillation and sustained VT [17]. AADs can be administered to reduce the burden of ACM-associated arrhythmias in symptomatic patients with non-sustained VT; moreover, AADs can be recommended as adjunctive therapy to mitigate the morbidities associated with ICDs (e.g., frequent device discharges and VT recurrences) and to facilitate catheter ablation [3,17]. Sotalol and amiodarone represent the first-line AADs in clinical practice, alone or in conjunction with β-blockers; however, there is no evidence that AAD therapy is effective at protecting patients from SCD. Furthermore, long-term treatment with amiodarone could induce extracardiac toxicity, so its use should be balanced in terms of effective arrhythmia suppression or reduction [3,17]. New hope for the pharmacological treatment of ACM has been sparked by the class I_c AAD flecainide ((RS)-N-(piperidin-2-ylmethyl)-2,5-bis(2,2,2-trifluoroethoxy) benzamide); C₁₇H₂₀F₆N₂O₃) [18,19,20,21], which was approved by the Food and Drug Administration (FDA) in 1984 for the treatment of symptomatic sustained VT [18,20,22]. Flecainide targets a panel of ion channels and transporters that shape the cardiac action potential (AP; e.g., Na_v1.5, K_v1.5, and K_v11.1) or the depolarization-induced Ca²⁺ transient responsible for cardiomyocyte contraction (e.g., type 2 ryanodine receptor, or RyR2) [22,23]. In addition, a recent investigation showed that flecainide may also inhibit fibro-adipose differentiation in ACM C-hMSCs by targeting SOCE [4].

Table 1. Genes whose mutations are associated with ACM.

Gene	Encoded Protein	Estimated Frequency (%)	Chromosomal Location
Desmosomes
PKP2	Plakophilin-2	19–46	12p11.21
DSP	Desmoplakin	1–16	6p24.3
DSG2	Desmoglein-2	2.5–10	18q12.1
DSC2	Desmocollin-2	1–8	18q12.1
JUP	Junction plakoglobin	Rare	17q21.2
Adherens junctions
CTNNA3	Catenin-α3	Rare	10q21.3
CDH2	Cadherin 2	Rare	18q12.1
Cytoskeleton
LMNA	Lamin A/C	Rare	1q22
DES	Desmin	Rare	2q35
FLNC	Filamin C	Rare	7q32.1
TTN	Titin	Rare	2q31.2
Ion transport
SCN5A	NaV1.5	Rare	3p22.2
PLN	Phospholamban	Rare	6q22.31

For references, see: [1,17].

Table 1. Genes whose mutations are associated with ACM.

Gene	Encoded Protein	Estimated Frequency (%)	Chromosomal Location
Desmosomes
PKP2	Plakophilin-2	19–46	12p11.21
DSP	Desmoplakin	1–16	6p24.3
DSG2	Desmoglein-2	2.5–10	18q12.1
DSC2	Desmocollin-2	1–8	18q12.1
JUP	Junction plakoglobin	Rare	17q21.2
Adherens junctions
CTNNA3	Catenin-α3	Rare	10q21.3
CDH2	Cadherin 2	Rare	18q12.1
Cytoskeleton
LMNA	Lamin A/C	Rare	1q22
DES	Desmin	Rare	2q35
FLNC	Filamin C	Rare	7q32.1
TTN	Titin	Rare	2q31.2
Ion transport
SCN5A	NaV1.5	Rare	3p22.2
PLN	Phospholamban	Rare	6q22.31

For references, see: [1,17].

3. Anti-Arrhythmic Effects of Flecainide: From Na_v1.5 to RyR2

A number of excellent reviews have provided a comprehensive description of the molecular pharmacology of flecainide [22,23,24]. Briefly, flecainide was first synthesised in 1972 by incorporating fluorine in local anesthetics, thereby increasing their stability, and it is nowadays recommended among the first-line therapeutic options to manage arrhythmias in patients without severe hemodynamic stability or significant structural heart disorders [25]. Flecainide reduces that fast inward Na⁺ current responsible for phase 0 of the cardiac AP by inhibiting the underlying Na_v1.5 channel protein (Figure 1) [22,26,27]. In addition, flecainide may inhibit the late Na⁺ current (I_NaL) [26], which can arise because of recovery from inactivation and late re-opening of Na_v1.5 channels during the sustained plateau in Purkinje fibres [28].

Flecainide may also inhibit the rapid delayed rectifier K⁺ current (I_KR), which is carried by K_v11.1 channels (human Ether-à-go-go-Related Gene, or hERG) and is responsible for the late repolarization restoring the resting membrane potential in cardiomyocytes (Figure 2) [26,30]. Flecainide can further block the fast transient outward K⁺ current (I_tof) (Figure 2), which is mediated by K_v4.2 channels and contributes to the early depolarization phase, and the ultrarapid K⁺ current (I_KUR) (Figure 2), which is carried by K_v1.5 channels and is the major delayed rectifier K⁺ current in the atria [22,26].

At concentrations higher than 10 µM, flecainide inhibits K_v4.2, which mediates the fast transient outward K⁺ current (I_tof), thereby potentially increasing APD in atrial and ventricular cardiomyocytes. In accordance with this, this inhibitory effect is exerted at supra-clinical doses of flecainide (IC₅₀ = 15.2 µM) and, therefore, it is not expected to occur in the patients. Finally, flecainide may also inhibit K_v1.5 (IC₅₀ = 237.1 µM), which mediates the ultrarapid delayed rectifier K⁺ current (IK_UR) in the atria, where it is particularly abundant, and in the ventricles, thereby increasing the APD. This effect also occurs at supra-clinical doses and is unlikely to occur in patients.

Globally, flecainide prolongs the AP duration (APD) and AP refractoriness in ventricular and atrial cardiomyocytes, while they are both shortened in Purkinje fibres due to Na_v1.5 channel blockade [22,26]. However, in human atrial cardiomyocytes, the flecainide-induced increase in APD and refractoriness is more evident when the AP presents a long-lasting plateau preceded by a notch [26]. Intriguingly, the kinetics of flecainide unbinding from the Na_v1.5 channel pore during diastole are quite slow, thereby prolonging the refractoriness more than the APD (a phenomenon known as post-repolarization refractoriness), reducing excitability, and slowing intracardiac conduction, even at physiological heart rates, throughout all cardiac tissues [26]. As discussed in [20,26,31], the clinical concentration range of flecainide may selectively inhibit I_Na and I_KR. Therefore, Na_v1.5 and K_v11.1 channels represent the main molecular targets of flecainide in arrhythmic patients. Flecainide can indeed be used to prevent and treat various types of arrhythmias, such as paroxysmal atrial fibrillation, supraventricular tachycardia, and arrhythmic long QT syndromes (LQTS) [18,20,22]. Flecainide is particularly suitable for LQTS (LQTS3) type 3, which is associated with gain-of-function mutations in Na_v1.5 channels (Figure 1) [20,22], which increase I_NaL [28,29]. Recent studies further showed that flecainide can also block RyR2, thereby inhibiting spontaneous RyR2-mediated sarcoplasmic reticulum (SR) Ca²⁺ release, which leads to delayed afterdepolarization (DADs) and triggered activity (Figure 1). Spontaneous mobilisation of SR Ca²⁺ can indeed be compensated by the electrogenic Na⁺/Ca²⁺ exchanger (NCX), which exports 1 Ca²⁺ out for 3 Na⁺ in during each cycle (Figure 1) and, while doing so, brings about an inward, depolarizing current [22,23]. Therefore, flecainide can also be used to treat catecholaminergic polymorphic ventricular tachycardia (CPVT) (Figure 1), a genetic arrhythmogenic disorder associated with mutations in RyR2 (CPVT1) or in the SR Ca²⁺-binding protein calsequestrin (CPVT2), which both lead to dysregulated SR Ca²⁺ release through RyR2 in response to β-adrenergic stimulation [23,32,33]. Flecainide binds to multiple sites within the RyR2 channel pore in a voltage-dependent manner but may also interact with several cytoplasmic sites in a voltage-independent manner [23,33]. The IC₅₀ of flecainide-induced inhibition of RyR2 in transgenic mouse models of CPVT is ~2 µM [32,34], i.e., in the clinical therapeutic range of flecainide.

The use of flecainide has been prohibited in patients with myocardial infarction, left ventricular dysfunction, or structural heart disease based upon the increase in their mortality upon treatment with flecainide registered during the Cardiac Arrhythmia Suppression Trial (CAST) study [35]. The CAST study reported a significant increase in the mortality of patients treated with I_c AADs, including flecainide, as compared to subjects treated with placebo, with the primary causes of death being arrhythmias [35]. Nevertheless, a series of recent investigations [18,20,36,37,38], as well as a retrospective analysis by Rolland et al. [19], confirmed that flecainide is both effective and safe for reducing atrial fibrillation and symptomatic ventricular arrhythmias in patients with different structural heart disorders, including coronary artery disease (CAD). Lavalle and coworkers recently affirmed that the therapeutic use of flecainide is strongly limited by the anachronistic interpretation of the CAST since the introduction of more efficient diagnostic tools, such as cardiac magnetic resonance imaging (CMR) rather than transthoracic echocardiography, will be helpful to identify the type of structural heart disorder that may really prevent flecainide treatment [20]. In addition, it has been put forward that thrombolysis, as well as primary percutaneous coronary intervention, may enable the selection of CAD patients who do not present myocardial scar and residual ischaemia and for whom there is no straightforward evidence against the prescription of flecainide [20].

4. The Rationale for Using Flecainide to Treat Ca²⁺-Dependent Ventricular Arrhythmias in ACM

A consensus statement published by the Heart Rhythm Society (HRS) in conjunction with several organisations, including the American Heart Association (AHA), American College of Cardiology (ACC), and European Heart Rhythm Association (EHRA), recommended the use of flecainide in combination with β-blockers to treat ACM patients refractory to single-agent therapy or catheter ablation [19,21,39,40]. A pilot randomised clinical trial is currently evaluating the effectiveness of flecainide in reducing ventricular arrhythmia in ACM patients (Pilot Randomized Trial With Flecainide in ARVC Patients, NCT03685149, currently ongoing). Blockade of Na_v1.5 channels represents the most obvious mechanism to explain the anti-arrhythmic effects exerted by flecainide in ACM [26,38]. Emerging evidence, however, suggests that dysregulated RyR2-mediated SR Ca²⁺ release can also generate pro-arrhythmic Ca²⁺ events in ACM [41]. By using a cardiomyocyte-specific PKP2 knockout mouse model, Cerrone et al. found that PKP2 deletion leads to a complex remodelling of the Ca²⁺ handling machinery, involving a reduction in the SR Ca²⁺ leakage caused by the downregulation of calsequestrin and RyR2 proteins [42]. This causes an elevation in SR Ca²⁺ content, which increases the amplitude and duration of spontaneous RyR2-mediated SR Ca²⁺ release because of the increase in intraluminal Ca²⁺ concentration. As a consequence, PKP2 deletion rendered RyR2 more prone to release Ca²⁺ during excitation-contraction coupling and to generate pro-arrhythmic Ca²⁺ release events during β-adrenergic stimulation [42]. A follow-up investigation confirmed that, in right ventricle-derived PKP2-deficient cardiomyocytes, the frequency of Ca²⁺ sparks, i.e., the elementary events of SR Ca²⁺ release [43], and of pro-arrhythmic afterdepolarizations increased before the onset of the cardiomyopathy due to an increase in SR Ca²⁺ content [44]. Finally, the same group showed that physical exercise caused a dramatic increase in the frequency and amplitude of SR Ca²⁺ sparks in PKP2-deficient mouse hearts because of the hyperphosphorylation of phospholamban, which accelerates Ca²⁺ uptake into the SR via the SarcoEndoplasmic Reticulum Ca²⁺-ATPase, during β-adrenergic stimulation [45]. Interestingly, blocking RyR2 with flecainide reduced the arrhythmia burden induced by β-adrenergic stimulation in PKP2-knockout mice [42]. Furthermore, a recent report confirmed that the Ca²⁺ cycling machinery was also altered in homozygous Dsg2 mutant mice (Dsg2^mut/mut) [46]. Cytosolic Ca²⁺ overload during endurance exercise stimulates calpain-1 association with the mitochondria, thereby leading to the cleavage of mitochondrial-bound apoptosis-inducing factor (AIF). The truncated AIF translocated to the nucleus and triggered large-scale DNA fragmentation and cell damage, an effect that was exacerbated by mitochondrial-driven AIF oxidation. This induced myocyte necrosis in the heart of exercised Dsg2^mut/mut mice [46].

On the one hand, these findings confirm that, as reported in other arrhythmic disorders (e.g., CPVT, Timothy syndrome, Brugada syndrome, and long and short QT syndrome), ACM is also associated with the deregulation of the Ca²⁺ cycling machinery in cardiomyocytes [47,48]. On the other hand, they support the notion that flecainide could effectively treat ACM patients by targeting the Ca²⁺ signalling machinery. In accordance with this hypothesis, flecainide decreased the frequency of Ca²⁺ sparks and normalised the Ca²⁺ transients elicited by membrane depolarization in inducible pluripotent stem cell-derived cardiomyocytes isolated from an ACM patient bearing a mutation in the DSC2 gene [49]. It has long been known that oscillations in intracellular Ca²⁺ concentration ([Ca²⁺]_i) regulate MSC differentiation into a variety of cells belonging to multiple lineages [50,51,52]. Moreover, a recent study carried out in our laboratory showed that intracellular Ca²⁺ oscillations drive proliferation in C-hMSCs [53]. In the next paragraphs, we first discuss the mechanisms underpinning intracellular Ca²⁺ oscillations in MSCs. Then, we illustrate how an increase in the frequency and amplitude of spontaneous Ca²⁺ activity drives fibro-adipose remodelling in ACM C-MSCs. Finally, we describe how flecainide inhibits spontaneous Ca²⁺ waves and fibro-adipogenic differentiation by targeting SOCE in ACM C-hMSCs.

5. Intracellular Ca²⁺ Oscillations Regulate MSC fate

The signalling mechanisms and functions of intracellular Ca²⁺ oscillations have been primarily characterised in bone marrow (BM)-derived MSCs. A landmark series of studies conducted by Kawano et al. showed that, in BM-derived human MSCs (BM-hMSCs), spontaneous Ca²⁺ oscillations were driven by ATP autocrine/paracrine signalling [54,55,56]. ATP stimulated the Gq-protein coupled P2Y1 receptor to recruit phospholipase Cβ (PLCβ), which cleaved phosphatidylinositol-4,5-bisphosphate (PIP₂), a minor phospholipid component of the plasma membrane, into diacylglycerol and inositol-1,4,5-trisphosphate (InsP₃). InsP₃ triggers rhythmic Ca²⁺ release from the endoplasmic reticulum (ER), the most abundant intracellular Ca²⁺ reservoir, by gating the so-called InsP₃ receptors (InsP₃Rs) [54,55,56]. These are Ca²⁺-permeable, non-selective cation channels that are embedded within ER cisternae and whose opening can be finely tuned by cytosolic Ca²⁺ levels [57,58]. BM-hMSCs express type 1 and type 2 InsP₃Rs (InsP₃R1 and InsP₃R2, respectively), which are the most suitable InsP₃Rs to generate rhythmic Ca²⁺ spikes in response to the continuous production of InsP₃ because of their sensitivity to InsP₃ and Ca²⁺ [56,59]. Intriguingly, InsP₃-induced spontaneous Ca²⁺ spikes in BM-hMSCs fade away in the absence of extracellular Ca²⁺; therefore, Ca²⁺ influx is required to refill the ER Ca²⁺ pool and support the rhythmic Ca²⁺ activity over time [54], as widely reported in other cell types [60,61,62,63,64]. Electrophysiological recordings revealed that BM-hMSCs do not present voltage-gated Ca²⁺ currents, although transcripts encoding for the Cav1.1 and Cav3.2 α subunits of voltage-operated Ca²⁺ channels were detected. Conversely, these cells express a measurable inwardly-rectifying store-operated Ca²⁺ current [54], also known as I_CRAC (Ca²⁺ release-activated Ca²⁺ current) (Figure 3). Intriguingly, by recharging the ER Ca²⁺ store, SOCE could also sensitise InsP₃Rs via an increase in intraluminal Ca²⁺, which leads to periodic InsP₃R activation and enhances the frequency of InsP₃Rs-mediated Ca²⁺ release [57,58,65].

The I_CRAC is the primary Ca²⁺-entry pathway sustaining intracellular Ca²⁺ oscillations over time in non-excitable cells and mediates an influx of Ca²⁺ that is commonly termed store-operated Ca²⁺ entry (SOCE) [5,6,57,58]. SOCE is activated whenever the ER Ca²⁺ concentration ([Ca²⁺]_ER) in proximity of discrete sub-compartments of the peripheral ER falls below a threshold concentration because of ER Ca²⁺ release through InsP₃Rs [66,67]. SOCE is mediated by the physical interaction between two ubiquitously expressed proteins (Figure 3): STIM, which is the sensor of [Ca²⁺]_ER, and Orai, which contributes the Ca²⁺-selective channel protein on the plasma membrane [5,6,68]. STIM presents two isoforms, i.e., STIM1 and STIM2, whereas Orai displays three paralogues, i.e., Orai1, Orai2, and Orai3. Briefly, a reduction in [Ca²⁺]_ER stimulates STIM proteins to assemble into oligomers that redistribute to peripheral ER-plasma membrane junctions, known as puncta, while extending their cytosolic COOH-termini, which contain the STIM Orai-activating region/CRAC-activating domain (SOAR/CAD), towards the inner leaflet of the plasma membrane (Figure 3). The SOAR/CAD domain, in turn, binds to and gates the hexameric Orai channels, thereby activating the I_CRAC and inducing SOCE (Figure 3) [5,6,69]. STIM2 presents a lower Ca²⁺ affinity as compared to STIM1 and is, therefore, activated upon a modest fall in [Ca²⁺]_ER. It turns out that STIM2 is regarded as the most suitable isoform to tonically activate SOCE even in the absence of extracellular stimulation and to maintain ER Ca²⁺ levels [70]. Conversely, STIM1 is activated after a significant depletion of [Ca²⁺]_ER during InsP₃-induced Ca²⁺ release and sustains the Ca²⁺ response to extracellular cues [5,6]. Nevertheless, STIM2 can sustain SOCE elicited by agonist-induced sub-threshold depletion of the ER Ca²⁺ content [68,71] as well as support STIM1-dependent recruitment to the ER-plasma membrane during supra-threshold stimulation [68,72]. In addition, STIM1 can be activated even in the absence of an extracellular agonist when the resting [Ca²⁺]_ER is too low to prevent its constitutive activation [73,74,75]. Orai1 has long been regarded as the primary pore-forming subunit of CRAC channels [5,69]. However, recent evidence suggests that Orai2 and/or Orai3 can assemble into heteromultimeric channels with Orai1 and serve as negative modulators of the I_CRAC in naïve cells [76,77,78]. The studies conducted by Kawano et al. date back to the early decade of this century [54,55,56], when STIM and Orai were not known to mediate the I_CRAC. Therefore, we still do not know whether they support SOCE in BM-hMSCs. Nevertheless, genetic silencing of STIM1, Orai1, and/or Orai3 reduced SOCE in mouse BM-derived MSCs (BM-mMSCs) [79,80,81] and in human dental pulp-derived MSCs (DP-hMSCs) [82].

Spontaneous intracellular Ca²⁺ oscillations have been described not only in BM-hMSCs [54,83,84] but also also in human adipose tissue-derived MSCs (AD-hMSCs) [51,85]. Intracellular Ca²⁺ oscillations regulate stem cell differentiation by stimulating the expression of genes that control cellular fate and silencing those that are required for self-renewal [52,86,87,88,89,90]. The spatio-temporal profile of spontaneous Ca²⁺ activity can selectively change during hMSC differentiation in tissue-specific lineages by showing either an increase or a decrease in the spike frequency [52]. Spontaneous Ca²⁺ oscillations disappear during hMSC differentiation towards an adipogenic phenotype both when they derive from the BM [56,91] and from the adipose tissue [51]. The loss of spontaneous Ca²⁺ activity reduces the nuclear translocation of the Ca²⁺-sensitive transcription factor, nuclear factor of activated T-cells (NFAT), which is likely to prevent hMSC differentiation [56]. Conversely, BM-hMSCs undergoing neuronal differentiation show an increase in the frequency of spontaneous Ca²⁺ spikes, which become quite irregular in terms of amplitude and shape [90,92]. A more robust Ca²⁺ spiking could govern BM-hMSC differentiation toward a stable neuronal phenotype by promoting the phosphorylation of CREB (cAMP response element binding protein) [93], a transcription factor that plays a crucial role in neurogenesis and can be activated by the Ca²⁺/CaM-dependent protein kinase IV [94]. The requirement of spontaneous Ca²⁺ oscillations for hMSC commitment to a specific cellular lineage is so tight that the intracellular Ca²⁺ activity in BM-hMSCs can be physically manipulated, for instance by electrostimulation [95], to favour osteogenic [91] or neuronal [96] differentiation. Similarly, the addition of inductive soluble factors, such as the InsP₃-producing autacoid carbachol, has been successfully employed to increase the frequency of InsP₃Rs-mediated spontaneous Ca²⁺ oscillations and thereby promote neuronal differentiation in AD-hMSCs [97]. As described above, SOCE is crucial to maintaining spontaneous Ca²⁺ oscillations in hMSCs; an increase or decrease in the rate of Ca²⁺ entry can differentially affect their differentiation outcomes. Therefore, it is not surprising that pharmacological manipulation of SOCE has been put forward as an alternative strategy to control hMSC differentiation [10,80].

6. Flecainide Inhibits SOCE and Prevents Ca²⁺-Dependent Fibro-Adipogenic Differentiation in ACM C-hMSCs

A recent investigation carried out by our laboratory has shown that intracellular Ca²⁺ signalling drives C-hMSC proliferation by stimulating extracellular signal-regulated kinase (ERK) phosphorylation [53]. In addition, this study confirmed that a functional SOCE is expressed in C-hMSCs and is sensitive to InsP₃-induced depletion of the ER Ca²⁺ store [53]. Therefore, based upon the evidence provided by the Pompilio group that ACM C-hMSCs are more likely to undergo fibro-adipogenic differentiation as compared to C-hMSCs [13,14], we established a collaborative endeavour to understand whether spontaneous Ca²⁺ oscillations were dysregulated in ACM C-hMSCs. Then, we evaluated whether SOCE could be pharmacologically targeted to prevent aberrant lipid/fibrotic accumulation in the ACM heart [4].

6.1. Intracellular Ca²⁺ Oscillations Are Up-Regulated in ACM C-hMSCs

Single-cell imaging of the Ca²⁺-sensitive fluorophore, Fura-2, represents a widespread technique to analyse intracellular Ca²⁺ oscillations in tens to hundreds of non-excitable cells [62,63,64,98,99,100], including hMSCs [54,55,56]. This approach led to the observation that spontaneous Ca²⁺ oscillations in ACM C-hMSCs were significantly more robust than C-hMSCs in terms of percentage of oscillating cells, amplitude, and frequency (~2.2 mHz vs. ~1.2 mHz) of the single Ca²⁺ transients (Figure 4A,B) [4]. C-hMSCs are not amenable to genetic manipulation of the Ca²⁺ handling machinery [53]. However, a pharmacological approach confirmed that spontaneous Ca²⁺ oscillations in ACM C-hMSCs were driven by the Gq-protein-coupled P2Y1 receptors, triggered by rhythmic InsP₃-induced ER Ca²⁺ release, and maintained over time by SOCE [4]. Conversely, RyR2 was not expressed, whereas voltage-gated Ca²⁺ entry did not support spontaneous Ca²⁺ oscillations [4]. Interestingly, the Mn²⁺ quenching assay, which is widely employed to monitor basal Ca²⁺ entry in non-excitable cells [70,75,101], revealed that SOCE was constitutively activated in C-hMSCs and was significantly enhanced in ACM C-hMSCs (Figure 4C,D) [4]. Constitutive SOCE regulates resting [Ca²⁺]_i and maintains [Ca²⁺]_ER, thereby finely tuning the magnitude and duration of InsP₃-induced ER Ca²⁺ release [70,73,74,75,101,102]. In accordance, basal [Ca²⁺]_i and InsP₃-induced ER Ca²⁺ release were both up-regulated in ACM C-hMSCs [4]. These findings were supported by the molecular evidence that the expression of STIM1 and InsP₃R2 was enhanced at both transcript and protein levels [4]. In addition, SERCA2B, which sequesters cytosolic Ca²⁺ into the ER lumen in non-excitable cells, was also up-regulated in ACM C-hMSCs [4]. Intriguingly, intracellular Ca²⁺ spiking was enhanced in C-hMSCs transduced with lentiviral particles containing a short hairpin against PKP2, which caused an increase in the expression levels of STIM1 and SERCA2B proteins [4].

These data support a model according to which (Figure 3): (1) STIM1 protein, the sensor of [Ca²⁺]_ER that is activated in response to large falls in [Ca²⁺]_ER occurring during autocrine/paracrine ATP signalling, is up-regulated in ACM C-hMSCs; (2) this leads to the recruitment of more Orai1 hexamers in the plasma membrane and therefore increases constitutive SOCE in ACM C-hMSCs; (3) the larger influx of Ca²⁺ causes a remarkable increase in [Ca²⁺]_ER, also because of SERCA2B up-regulation, in ACM C-hMSCs; (4) the enhanced ER Ca²⁺ loading, on the one hand, sensitises InsP₃Rs to open at a higher frequency; on the other hand, it increases the availability of free intraluminal Ca²⁺ that can be released during each Ca²⁺ spike. In this context, it is worth clarifying that, although the SOCE signal driving the spontaneous Ca²⁺ oscillations in C-hMSCs has been termed constitutive [4], it is stimulated by ATP released in either an autocrine or paracrine manner via P2Y1 receptor activation. The term “constitutive” reflects the lack of any agonist in the perfusate when intracellular Ca²⁺ oscillations and basal Ca²⁺ entry are monitored.

6.2. Intracellular Ca²⁺ Oscillations in ACM C-hMSCs: Is There Any Role for Lysosomal Ca²⁺ Mobilization?

It is worth noticing that ER Ca²⁺ release through InsP₃Rs in C-hMSCs could also be activated by lysosomal Ca²⁺ release via two pore channels (TPCs) [53]. TPC1 and TPC2 are activated by the Ca²⁺-mobilizing second messenger, nicotinic acid adenine dinucleotide phosphate (NAADP) [103,104], and are both expressed in C-hMSCs, in which they trigger the intracellular Ca²⁺ oscillations by which foetal bovine serum stimulates proliferation [53]. Transmission electron microscopy revealed that lysosomal vesicles can establish quasi-synaptic membrane contact sites with the juxtaposed ER cisternae in C-hMSCs [53]. Herein, NAADP-induced lysosomal Ca²⁺ release via TPCs may activate InsP₃Rs via the CICR process [53]. This signalling pathway may trigger long-lasting oscillations in [Ca²⁺]_i not only in C-hMSCs but also in many other cell types [62,64,105,106]. In addition, NAADP-induced lysosomal Ca²⁺ release may contribute to cardiac hypertrophy induced by β-adrenergic stimulation [107,108] and ischemia-reperfusion injury [109]. Therefore, future investigations will have to assess whether NAADP and TPCs play any role in the spontaneous Ca²⁺ oscillations arising in C-hMSCs and whether their contribution, if any, is increased in ACM. Intriguingly, SOCE has also been shown to support lysosomal Ca²⁺ refilling, such that the increase in constitutive SOCE has the potential to increase lysosomal Ca²⁺ release in ACM C-hMSCs [110].

6.3. Intracellular Ca²⁺ Oscillations Drive Fibro-Adipogenic Differentiation in ACM C-hMSCs

Unexpectedly, we found that spontaneous Ca²⁺ oscillations in ACM C-hMSCs showed an increase in amplitude and frequency (up to 8 mHz) during adipogenic differentiation [4]. This finding was surprising since, both in BM- and AD-derived MSCs, the Ca²⁺ spiking activity decreases when they are committed to an adipogenic phenotype [51,56]. The pharmacological blockade of the Ca²⁺ signalling machinery that sustains the accelerated Ca²⁺ spikes, i.e., SOCE and InsP₃Rs, prevented ACM C-hMSCs from differentiating into either adipocytes or fibroblasts [4]. Therefore, it takes an increase in the amplitude and frequency of spontaneous Ca²⁺ oscillations to induce ACM C-hMSCs to undergo fibro-adipogenic differentiation. The multifunctional Ca²⁺/CaM-dependent protein kinase II (CaMKII) represents the most suitable decoder of intracellular Ca²⁺ oscillations not only in the brain but also in the heart [111,112,113]. CaMKII is a multimer of eight to twelve subunits that are present in four different isoforms (α, β, γ, and δ), each encoded by a distinct gene, arranged in two-stacked hexameric rings [114]. CaMKIIδ is the major isoform in the heart and exists as different spliced variants, e.g., CaMKIIδ_2/B, CaMKIIδ_2/C, and CaMKIIδ_2/C [111]. Briefly, in response to an increase in [Ca²⁺]_i, the Ca²⁺/CaM complex can bind to two neighbouring subunits, thereby enabling CaMKII to autophosphorylate on threonine 287 (Thr²⁸⁷). Autophosphorylation traps Ca²⁺/CaM and prolongs CaMKII activity even when the [Ca²⁺]_i returns to the baseline. Low-frequency Ca²⁺ transients are less effective than high-frequency Ca²⁺ transients to evoke autophosphorylation and Ca²⁺-independent activity since the Ca²⁺/CaM complex can dissociate from the holoenzyme between two consecutive Ca²⁺ spikes [115]. We found that CaMKIIδ and CaMKIIγ were the predominant CaMKII isoforms in C-hMSCs; however, there was no difference in CaMKII protein expression between C-hMSCs and ACM C-hMSCs. However, CaMKII protein in ACM C-hMSCs was hyperphosphorylated as compared to healthy cells, which is consistent with their greater eagerness to generate Ca²⁺ oscillations [4]. In agreement with this hypothesis, the pharmacological blockade of CaMKII activity with KN93 prevented fibro-adipogenic differentiation in ACM C-hMSCs [4]. Therefore, CaMKII is likely to translate the spontaneous Ca²⁺ oscillations that occur at a higher frequency in ACM C-hMSCs into a transcriptional programme that redirects their differentiation fate to fibroblasts or adipocytes. One could rightly notice that a frequency of 8 mHz might be too slow to enable the accumulation of Ca²⁺/CaM-bound CaMKII subunits sufficient for autophosphorylation and autonomous CaMKII activity [116]. Nevertheless, CaMKII does not only integrate the information encoded in the oscillation frequency but also in the length of the single Ca²⁺ pulses; an increase in the individual Ca²⁺ spike duration decreases the frequency threshold for CaMKII activation and autonomous activity [115]. The long duration of individual Ca²⁺ spikes in ACM C-hMSCs, ranging from ~200 s up to ~300 s (Figure 4B), enables coincident binding of CaM on neighbouring subunits and of Thr²⁸⁷ autophosphorylation even at 8 mHz [115,116]. For instance, the slow Ca²⁺ oscillations evoked by fertilisation in mouse oocytes were able to induce long-lasting oscillations in CaMKII activity that were crucial to resuming embryonic development [117]. The mechanism by which CaMKII stimulates the fibro-adipogenic differentiation of ACM-hMSCs is not clear [4]. However, CaMKII promoted adipogenesis in porcine BM-MSCs by increasing the expression of two adipogenic transcription factors, such as CCAAT/enhancer binding protein α (C/EBPα) and peroxisome proliferator activated receptor γ (PPARγ), by stimulating the PI3K/Akt-Fox01 signalling pathway [118].

6.4. Flecainide Abolishes Intracellular Ca²⁺ Oscillations and Fibro-Adipogenic Differentiation by Targeting SOCE

SOCE is either down-regulated or up-regulated in a growing number of disorders, including severe combined immunodeficiency [119], neurodegenerative diseases [120], genetic myopathies [121], pulmonary arterial hypertension [122], and cancer [123]. Preliminary evidence indicated that, even though Orai1 is weakly expressed in adult ventricular cardiomyocytes from humans and mice, its expression increases and leads to SOCE up-regulation in cardiac hypertrophy and heart failure [124]. Therefore, the pharmacological manipulation of SOCE with selective agonists or blockers of Orai1-mediated Ca²⁺ entry represents a promising therapeutic approach to rescue aberrant Ca²⁺-dependent processes in multiple cell types [65,122,125,126,127,128,129,130,131]. The evidence that flecainide may physically interact with many pro-arrhythmic ion channels led our group to hypothesise that it could also inhibit constitutive SOCE in ACM C-hMSCs. Consistently, single-cell Ca²⁺ imaging experiments carried out in our laboratory showed that 10 µM flecainide blocked constitutive SOCE (Figure 4E) and thereby inhibited InsP₃-induced ER Ca²⁺ release in these cells [4]. As expected, based upon these findings, flecainide also suppressed spontaneous Ca²⁺ oscillations (Figure 4F) and interfered with ACM C-hMSC fibro-adipogenic differentiation [4]. Of note, ACM C-hMSC cells do not express Na_V1.5; therefore, the inhibitory effect of flecainide is not related to the inhibition of voltage-gated inward Na⁺ currents [4]. In this view, it must be pointed out that Na_v1.5-mediated Na⁺ currents could favour intracellular Ca²⁺ oscillations only by switching the NCX into the reverse (i.e., Ca²⁺ in, Na⁺ out) mode [132,133]. However, the pharmacological blockade of the reverse-mode NCX activity did not affect the spontaneous Ca²⁺ oscillations in ACM C-hMSCs [4]. It turns out that flecainide-induced SOCE inhibition underlies the suppression of both the unsolicited Ca²⁺ spikes and fibro-adipogenic differentiation that take place in C-hMSCs expanded from ACM patients. The mechanisms by which flecainide inhibits constitutive SOCE remain to be elucidated (Figure 3). SOCE in ACM C-hMSCs is also sensitive to two established inhibitors of this Ca²⁺-entry pathway, namely the pyrazole compounds Pyr6 and BTP-2 (YM-58483) [7,123,134,135,136,137]. These drugs are known to selectively block Orai1-mediated SOCE [7,123,134,135,136,137]; however, they could also affect Orai2 [138,139]. Nevertheless, Orai2 is a negative modulator of Orai1 [76,77,78]. Therefore, the direct inhibition of Orai2 is predicted to enhance, not inhibit, SOCE. These observations support the view that Pyr6, BTP-2, and flecainide interfere with extracellular Ca²⁺ entry through Orai1.

The mechanism by which Pyr6 inhibits the I_CRAC and SOCE is still unclear, although it is unlikely to prevent the STIM-dependent Orai1 channel activation process [136]. BTP-2, in turn, could target a binding site that is located on the extracellular side of the Orai1 channel pore when administered at low micromolar doses [140], as throughout our investigations [4,53]. This hypothesis is supported by the evidence that BTP-2 does not block the I_CRAC when intracellularly applied [141] and does not prevent the physical interaction between STIM1 and Orai1 [142], as suggested for another established SOCE inhibitor, i.e., 2-Aminoethyldiphenyl Borate [143]. Future studies are mandatory to unveil how flecainide blocks constitutive SOCE in ACM C-hMSCs. Nevertheless, the evidence presented in [4] further supports the emerging view that this pleiotropic drug targets multiple signalling pathways that are involved in ACM pathogenesis: (1) Na_V1.5 and (2) RyR2 in cardiomyocytes, thereby rescuing the spontaneous electrophysiological events that lead to tachycardia and ventricular fibrillation; (3) SOCE in C-MSCs, thereby alleviating the aberrant remodelling of cardiac structure that predisposes the heart to generate arrhythmic events. Interestingly, we have previously shown that the β-blocker propranolol inhibits constitutive SOCE and reduces the vasculogenic activity of circulating endothelial colony forming cells in infantile hemangioma (IH) both in vitro and in children affected by IH, thereby resulting in a therapeutic benefit in IH patients [144,145]. The hypothesis that propranolol effectively targets ACM by inhibiting both β2-adrenergic receptors in sinoatrial fibres and cardiac myocytes and SOCE in C-MSCs is certainly worth of further investigation.

Notably, several Orai1 inhibitors have already entered Clinical Trials for the treatment of multiple diseases, including psoriasis (CM2489 from CalciMedica), severe acute pancreatitis (CM4620 from CalciMedica) (https://clinicaltrials.gov/ct2/show/NCT03401190 (accessed on 30 May 2023)), critical COVID-19 pneumonia (Auxora, from CalciMedica) (https://clinicaltrials.gov/ct2/show/NCT04661540 (accessed on 30 May 2023)), mild asthma (RP3128 from Rizen Pharmaceuticals) (https://clinicaltrials.gov/ct2/show/NCT02958982 (accessed on 30 May 2023)), and non-Hodgkin’s lymphoma (RP4010, also from Rizen Pharmaceuticals) (https://clinicaltrials.gov/ct2/show/NCT03119467 (accessed on 30 May 2023)) [134]. These studies confirmed that SOCE inhibitors can be administered orally, are safe, and do not induce serious adverse effects in human subjects [134]. In line with this evidence, several Food and Drug Administration (FDA)-approved drugs were recently found to inhibit both the I_CRAC and SOCE [146]: teriflunomide, leflunomide, roflumilast, tolvaptan, and lansoprazole. Therefore, targeting SOCE represents a promising therapeutic avenue for many diseases that still lack a valid therapeutic option, including ACM.

7. Future Directions

The mechanism by which flecainide inhibits SOCE requires further investigation. Likewise, the molecular determinants that increase constitutive SOCE may go beyond STIM1 overexpression and include the derangement of other regulatory factors, such as an increase in STIM1 expression or a reduction in Orai2 or Orai3 protein levels. Nevertheless, these preliminary findings strongly hint at flecainide as a pleiotropic drug that owes its efficacy to its ability to target multiple pro-arrhythmic signalling pathways in ACM. In vivo investigations will have to assess whether it does interfere with cardiac fibrofatty remodelling in transgenic mouse models of ACM. It will also be imperative to assess whether flecainide does interfere with CaMKII hyperactivation in ACM C-MSCs and, if so, to untangle the signalling pathways that are recruited by CaMKII to stimulate fibro-adipogenic differentiation. In accordance with this, Maione et al. did not assess whether inhibiting SOCE or InsP₃Rs reduced CaMKII hyperphosphorylation in ACM C-hMSCs. Addressing this issue is mandatory to provide a clear-cut landscape of the molecular mechanisms by which an increase in constitutive SOCE promotes C-hMSC fibro-adipogenic differentiation in ACM hearts. Currently, we cannot rule out the possibility that CaMKII in ACM-hMSCs is hyperphosphorylated because of a decrease in the expression and/or activity of protein phosphatases 1 and 2A, which dephosphorylate CaMKII when [Ca²⁺]_i returns to the baseline [147,148]. This would in turn lead to SOCE hyperactivation and ER Ca²⁺ overload since it has long been known that CaMKII is able to stimulate SOCE [149,150,151]. However, there is no doubt that the pharmacological blockade of SOCE inhibits both spontaneous Ca²⁺ oscillations and fibro-adipogenic differentiation in ACM C-hMSCs [4]. In this view, the evidence that flecainide suppresses fibro-adipogenic differentiation paves the way for future investigations assessing whether other FDA-approved SOCE inhibitors could also be used to treat ACM by preventing fibro-fatty remodelling in C-MSCs. Finally, it could also be evaluated whether flecainide is also effective to target other disorders that are associated with SOCE up-regulation, such as severe acute pancreatitis, chronic inflammation, and cancer.

8. Conclusions

Flecainide is a class I_c AAD that proved to be effective in the management of heart rate by targeting Na_V1.5, which mediates the fast inward Na⁺ current, and RyR2, which mediates aberrant SR Ca²⁺ release. Flecainide is emerging as a promising therapeutic tool to treat ACM, which involves a complex remodelling of the Ca²⁺-handling machinery in ventricular cardiomyocytes. In addition, studies conducted on C-MSCs isolated from ACM patients and transgenic mouse models of ACM revealed that intracellular Ca²⁺ dynamics is also dysregulated in the stromal compartment of the heart. SOCE, which represents the main intracellular Ca²⁺ entry pathway in non-excitable cells, is constitutively activated in ACM C-hMSCs and causes Ca²⁺ overload in the InsP₃-sensitive ER Ca²⁺ reservoir. This results in an increase in both the amplitude and frequency of spontaneous Ca²⁺ oscillations that finely tune MSC fate, thereby favouring fibro-adipogenic differentiation in ACM C-hMSCs. Therefore, SOCE up-regulation in C-hMSCs could play a pathogenic role in ACM by stimulating the cardiac structural remodelling that increases the incidence of arrhythmic events. Intriguingly, flecainide, when applied at doses known to inhibit Na_v1.2 and RyR2 channels, inhibited constitutive SOCE, thereby suppressing the spontaneous Ca²⁺ oscillations that occur during fibro-adipogenic differentiation and preventing ACM C-hMSC differentiation into fibroblasts or adipocytes. These pieces of evidence provide the first proof-of-concept that flecainide is also effective against SOCE and indicate that SOCE could be targeted for therapeutic purposes in ACM.