Intracellular calcium leak: a unifying mechanism and therapeutic target for heart failure and atrial fibrillation

Abstract

Calcium (Ca²⁺) is a fundamental second messenger in all cell types and is required for numerous critical cellular functions including cardiac and skeletal muscle contraction. Intracellular free [Ca²⁺] is regulated primarily by ion channels, pumps (ATPases), exchangers and Ca²⁺ binding proteins. Defective regulation of [Ca²⁺] is found in a diverse spectrum of pathological states and human disorders affecting all the major organs. In the heart we and others have reported abnormalities in cytosolic [Ca²⁺]_cyt and mitochondrial [Ca²⁺]_mito regulation in heart failure (HF) and atrial fibrillation (AF), the two most common forms of heart disease and leading contributors to mortality and morbidity. This review is focused on mechanisms that regulate the major SR Ca²⁺ release channel in the heart, the ryanodine receptor type 2 (RyR2), how the channel becomes dysfunctional in HF and AF and is a potential therapeutic target. Inherited RyR2 mutations, and/or stress-induced phosphorylation and oxidation, destabilize the closed state of the channel resulting in a pathological SR Ca²⁺ leak that both triggers arrhythmias and impairs contractility. Recent high-resolution cryo-electron microscopy (cryo-EM) RyR structures (Figs. 4–6) have clarified the architecture of multiple functional states, ligand-binding sites and gating mechanisms, and informed as to mechanisms that go awry in HF and AF. Based on new understandings of SR Ca²⁺ leak as a common Ca²⁺-dependent pathological mechanism in HF and AF, a new class of drugs developed in our laboratory called Rycals, that stabilize RyR2 channels and prevent SR Ca²⁺ leak, are undergoing clinical trials.

Introduction

Heart failure (HF) and atrial fibrillation (AF) are leading causes of mortality and morbidity in developed countries. Globally, HF affects an estimated 26 million people and contributes to more than 1 million hospitalizations annually¹. AF is the most common cardiac arrhythmia with a prevalence of 2.5–6.0 million in the United States, affecting 8% to 10% of individuals over the age of 80². HF and AF share common risk factors including hypertension, diabetes mellitus, and valvular heart disease which often occur together. The prevalence of AF in patients with symptomatic HF (NYHA class II-IV) is between 15 and 35% ^3,4.

During AF the atria exhibit disorganized, rapid contractions, at rates of 400–500 beats per minute ⁵. This results in atria that fibrillate instead of contracting uniformly, which leads to ventricular underfilling and reduced cardiac output. The absence of regular cardiac contractions increases the risk of thrombosis within the left atrial appendage, which can embolize to the brain and cause ischemic strokes ⁶. AF can be managed by rate control or rhythm control. Rate control is achieved with medications that target conduction through the atrio-ventricular (AV) node, resulting in a reduction in the number of ventricular beats ⁵. With rhythm control, an effort is made to maintain sinus rhythm using anti-arrhythmic medications, electrical cardioversion, and/or catheter ablation.

Studies comparing rate control to rhythm control suggest at best clinical equivalence⁷ and even possible harm with aggressive rhythm control ⁸. Potential benefits from AF therapies are offset by limited efficacy of anti-arrhythmic medications and intolerable side effects which impair medication compliance. A recent study comparing radiofrequency catheter ablation to medical rate control suggested clinical benefit for patients receiving catheter ablation to maintain sinus rhythm, however this benefit was not observed with intention-to-treat analysis⁹. Survival benefit for rhythm control in AF is better established in patients with HF^9,10.

Heart failure is a condition characterized by the inability to generate a cardiac output sufficient to meet the body’s metabolic needs. HF can be sub-classified as having either reduced ejection fraction (HFrEF) or preserved ejection fraction (HFpEF)¹¹. Common causes of HFrEF are ischemia, hypertension, and valvular heart disease. Medical management of HF focuses on reducing the effect of elevated sympathetic tone using beta-blockers and/or blocking angiotensin 2 production, and reducing vascular tone and stress on the heart. Despite these interventions, many patients have disease progression requiring cardiac transplantation or mechanical assist devices¹².

Although HF and AF are complex conditions with many contributing factors, alterations in intracellular Ca²⁺ handling in cardiac myocytes have been implicated in the pathogenesis of both conditions¹³. Ca²⁺ is a ubiquitous second messenger that controls numerous biological functions including muscle contraction, synapse transmission, hormone secretion, proliferation, and apoptosis¹⁴. Accumulating evidence indicates that defective cardiac Ca²⁺ signaling represents a central pathogenic mechanism underlying cardiac diseases, including HF and AF^15–21. This review discusses Ca²⁺ dysregulation in HF and AF, and includes data from high-resolution cryo-EM structures of RyRs with a focus on clinical and therapeutic implications. Based on a recent pubmed search there have been >44,000 reviews written about HF and ~12,000 on AF. There is no need for another review with a broad comprehensive scope. Instead the present review is focused on just one part of the story: the role that dysregulation of Ca²⁺ homeostasis ²¹ in ventricular and atrial cardiomyocytes plays in HF and AF. As such we have not addressed some important molecular mechanisms that contribute to HF as well as structural considerations such as T-tubule remodeling ²²; and those that contribute to AF as well as the roles of fibrosis and conduction defects, all of which are presented in detail in several excellent textbooks as well as reviews. Rather our intent is to focus on Ca²⁺ and in particular the role of leaky RyR2 channels. In so doing we do not address other equally important aspects of HF and AF. We have briefly summarized a few of the scientific disagreements in the field most of which have been addressed elsewhere ^13,23. Our purpose herein is to highlight a common mechanism linking HF and AF that is also a therapeutic target: diastolic SR Ca²⁺ leak.

The mechanisms underlying AF and HF have been obscure for many decades and the current therapies are largely aimed at reducing symptoms, but are not disease modifying. This review makes the point that diastolic SR Ca²⁺ leak via dysfunctional RyR2 channels is a common underlying mechanism in both AF and HF and potentially can be treated with a new class of drugs called Rycals that fix leaky RyR2 channels. A key player in the etiology of the leak is the channel subunit calstabin (Calcium Channel Stabilizing subunit). We first showed that calstabin1 (FKBP12) is a subunit of RyR1 ²⁴ and that it is required to stabilize the closed state and prevent aberrant Ca²⁺ leak through the channel ²⁵ and that calstabin2 (FKBP12.6) plays the same role in RyR2 ²⁶. Others have confirmed that calstabin2 plays an important physiological role in modulating RyR2 function ^27–34.

The RyR2 dysfunction in AF and HF is due to oxidation and PKA hyperphosphorylation (defined as PKA phosphorylation of 3 or 4 of the 4 PKA sites on each homotetrameric RyR2 channel) of the RyR2 channel which cause leak by destabilizing the closed state of the channel associated with depletion of the stabilizing subunit calstabin2 from the RyR2 channel macromolecular complex ^{19,20,26,35–37}. Rycals bind to RyR2 and prevent dissociation of calstabin2 via an allosteric mechanism. Of note a similar process occurs in skeletal muscle where leaky RyR1 channels cause impaired ECC and reduced exercise capacity (a major symptom in HF patients) ^37–39. Rycals can also fix this leak and HF really should be thought of as a generalized myopathy involving both types of striated muscle, skeletal and cardiac ^37–40. A rycal is now being tested in RyR1-myopathy patients in a clinical trial at the NIH (ClinicalTrials.gov Identifier: NCT04141670) that involves one month oral dosing of rycal and pre- and post- muscle biopsies to see whether the leaky RyR1 channels in this inherited form of muscular dystrophy caused by RyR1 mutations that make the channel leaky have been fixed.

1. Cardiac Excitation-contraction coupling

To understand Ca²⁺ signaling in the heart one must be familiar with cardiac excitation–contraction coupling (ECC). ECC which converts electrical signals into mechanical force, requires coordinated Ca²⁺ release and reuptake in cardiomyocytes. The Ca²⁺ release cycle is initiated by depolarization of the plasma membrane and the specialized invagination called the transverse tubule (T-tubule) ⁴¹. Voltage-gated L-type Ca²⁺ channels (LTCC) on the T-tubular membrane are activated by depolarization and allow a small inward Ca²⁺ current. This Ca²⁺ binds to and activates RyR2 channels, which release Ca²⁺ from the sarcoplasmic reticulum (SR) where it is stored at high concentration (high micromolar to low millimolar range).

The process by which Ca²⁺ activates RyR2 channels is known as Ca²⁺-induced-Ca²⁺-release (CICR) and was originally described by Fabiato and Fabiato ^42,43. Ca²⁺ released into the cytoplasm raises [Ca²⁺]_cyt ~10-fold (although local [Ca²⁺] is likely even higher) and binds to troponin C (and other Ca²⁺-sensitive proteins), allowing actin-myosin cross-bridging as the thick and thin filaments of the sarcomere slide past one another, shortening the sarcomere and causing muscle contraction (systole). Individual Ca²⁺ release events can be imaged as Ca²⁺ sparks ⁴⁴ the frequency of which is a measure of SR Ca²⁺ release and/or leak which is increased in failing hearts ²². Cardiac relaxation, or diastole, is initiated when Ca²⁺ release is terminated and Ca²⁺ is pumped back into the SR by the sarco-endoplasmic reticulum Ca²⁺ ATPase 2a (SERCA2a), extruded by the Na⁺/Ca²⁺ exchanger (NCX1) (Fig. 1&3). Dysfunctional Ca²⁺ handling may occur at different stages of the ECC process including Ca²⁺ entry to the cardiomyocytes¹⁵, intracellular Ca²⁺ release⁴⁵ and Ca²⁺ uptake and buffering⁴⁶. This review will focus on the role of RyR2 dysfunction in Ca²⁺ dysregulation in HF and AF. Of note, RyR2 is also expressed in other organs including the brain⁴⁷ and the pancreas⁴⁸ and it is likely that RyR2 dysfunction in these and other organs also contributes to HF as a systemic disorder.

Figure 1. Diastolic SR Ca²⁺ leak in ventricular myocytes from failing hearts.

Representative line scan of Ca²⁺ sparks recorded in ventricular cardiomyocytes from normal (A) and failing murine hearts (B). Representative RyR2 single-channel tracings under conditions that simulate diastole when the cytosolic Ca²⁺ concentration is low (e.g. 100–150 nM), in normal (C) and heart failure (D) cardiomyocytes ²⁴³. Schematic representation of intracellular Ca²⁺ release in EC coupling in normal (E) and heart failure (F) cardiomyocytes. There is little or no diastolic SR Ca²⁺ leak in ventricular cardiomyocytes from normal non-failing hearts (panels A,C and E). In contrast, ventricular cardiomyocytes from failing hearts exhibit increased spontaneous Ca²⁺ sparks, increased RyR2 open probability, diastolic SR Ca²⁺ leak and reduced SR Ca²⁺ stores, all consistent with a pathologic leak of SR Ca²⁺via RyR2 remodeled (including PKA hyperphosphorylation, oxidation, nitrosylation and dissociation of calstabin2 from the RyR2 channel complex) resulting in RyR2 channels that do not close properly during diastole. (panels B,D and F). This results in reduced SR Ca²⁺ content. AP, action potential; TT, transverse tubule; cAMP, cyclic adenosine monophosphate; Cav1.2, L-type Ca²⁺ channel; NCX, Na⁺/Ca²⁺ exchanger; RyR2, ryanodine receptor type-2; SERCA2a, sarco/endoplasmic reticulum ATPase type-2a; PLB, phospholamban; PKA, protein kinase A; CaMKII, Ca^2+/calmodulin-dependent protein kinase II; ROS, reactive oxygen species.

Figure 3. SR Ca²⁺ leak triggers AF in atrial myocytes.

Representative line scan of Ca²⁺ sparks recorded in atrial myocytes from normal (A) and atrial fibrillation (B)¹⁷. Representative RyR2 single-channel tracings recorded under conditions simulating diastole in atrial cardiomyocytes at a resting [Ca²⁺]_cyt ≃ 150 nM, in normal (C) and atrial fibrillation cardiomyocytes (D)¹⁰¹. Schematic representation of Ca²⁺ signaling in normal (E) and atrial fibrillation (F) atrial myocytes. Atrial myocyte exhibit increased spontaneous Ca²⁺ sparks, increased RyR2 open probability, Ca²⁺ leak and reduced SR Ca²⁺ stores in AF.

2). Stress-induced RyR2-mediated diastolic SR Ca²⁺ leak

RyR2 is the largest known ion channel with a molecular weight in excess of 2 million Daltons. The channel is a homotetrameric, macromolecular complex comprised of four RyR2 565 kDa protomers⁴⁹. RyR2 channel activity is regulated by kinases and phosphatases²⁶, phosphodiesterase⁵⁰, calmodulin^51–57 and a stabilizing subunit protein FK506-binding protein 12.6 (FKBP12.6 or Calstabin2)^25,26. Protein Kinase A (PKA) and Ca²⁺/calmodulin-dependent Kinase II (CaMKII) tether to RyR2 and phosphorylate the channel^26,58. PKA is activated by the adrenergic receptor pathway⁵⁹. Epinephrine and norepinephrine bind to the beta-adrenergic receptor in cardiomyocytes, activate adenylyl cyclase (AC) which produces cAMP resulting in downstream PKA activation that increases inotropy, lusitropy, and chronotropy⁶⁰. CaMKII is initially activated by Ca²⁺/calmodulin, and then remains active due to autophosphorylation ^58,61,62. We and others have identified Ser2808 ²⁶ and Ser2814 ^63,58 as the physiologic PKA and CaMKII phosphorylation sites, respectively. Both sites modulate RyR2 activity (Fig. 1–3).

Indicative of the importance of PKA phosphorylation of RyR2 is the finding that mice harboring an RyR2-S2808A mutation have blunted inotropic and chronotropic responses to catecholamines ^19,20,35. On the other hand mice with a phosphomimetic mutation of the RyR2 PKA phosphorylation site (RyR2-Ser2808Asp) have leaky RyR2 channels and an age dependent cardiomyopathy and arrhythmias ²⁰. Moreover, there is protection against HF progression in RyR2-Ser2808Ala knock-in mice ^{19,20,35–37}. Another group generated an RyR2-S2808A mouse and used it to report that RyR2-Ser2808 plays no role in the cardiomyocyte response to β-adrenergic stimulation ⁶⁴. These and other differences have been addressed in detail elsewhere ²³.

Inherited mutations linked to exercise-induced sudden cardiac death (Catecholaminergic Polymorphic Ventricular Tachycardiac, CPVT) ^59,65 as well as PKA phosphorylation, S-nitrosylation and Cys-oxidation of RyR2 can cause dissociation of calstabin2 from RyR2^20,36,37,66. Calstabin2 is tightly associated with RyR2 and modulates it function^25,59,67. Similarly, Calstabin1 is bound to RyR1 and regulates EC coupling gain as shown by Dirksen and colleagues ⁶⁸.

One molecule of calstabin2 binds to each protomer of RyR2 and stabilizes the channel closed state, increases mean open and closed times as well as facilitates coupled gating between neighboring RyR2 channels ⁶⁹. Stress-induced calstabin2 dissociation leads to unstable channels that leak Ca²⁺ as the channel is unable to close properly^25,26. Dysregulation of intracellular Ca²⁺ handling alters contractility and electrical stability. Chronic SR Ca²⁺ leak reduces SR Ca²⁺ stores as well as the Ca²⁺ wave amplitude necessary for optimal cardiomyocyte contraction during systole⁷⁰. RyR2 Ca²⁺ leak can be observed in isolated cardiomyocytes as short unsynchronized SR Ca²⁺ release events during diastole⁷¹.

The physiologic role of calstabin2 binding to RyR2, and the stress-induced dissociation of calstabin2 binding from RyR2 have been challenged. For example, one group reported that only 10—20% of RyR2 have calstabin2 in the channel macromolecular complex, and that PKA phosphorylation of RyR2 did not affect calstabin2 binding ⁷². Another group overexpressed calstabin2 in cells expressing recombinant RyR2 with a phosphomimetic residue in the PKA site (Ser2809Asp) and did not observe a decrease in binding compared to WT RyR2⁷³. Our assessment is that these differences are due to variances in methods and data interpretation and have been addressed in detail elsewhere ²³. Further supporting a role for PKA phosphorylation of RyR2 in dissociation of calstabin2, we observed that leaky RyR2-Ser2808Asp channels are depleted of calstabin2 and progressively oxidized both when heterologously expressed in HEK293 cells and in vivo in RyR20-Ser2808Asp mice, and that channel oxidation also contributes to the decrease in calstabin2 binding²⁰.

Isoproterenol-induced dissociation of calstabin2 from RyR2 has been reported by others²⁹ and overexpression of calstabin2 prevents SR Ca²⁺ leak³¹ and ventricular arrhythmias in mice exposed to isoproterenol and myocardial burst pacing²⁷. In HF there is a reduction in the amount of calstabin2 bound to RyR2^74,75 and beta-blockers improve calstabin2 binding to RyR2⁷⁶ by preventing PKA phosphorylation of RyR2. The mechanisms by which PKA phosphorylation and oxidation/nitrosylation dissociate calstabin2 from RyR2 have been confirmed by a number of groups ^{27,31,67,74–77}.

We first showed that PKA hyperphosphorylation of RyR2 depletes calstabin2 from the RyR2 complex ²⁶. Years later we found that oxidation and nitrosylation of RyR2 also cause depletion of calstabin2 from the RyR2 complex: together all three processes (oxidation, nitrosylation and phosphorylation of Ser2808) fully deplete calstabin2 from the channel complex ^19,20. RyR2-Ser2808Ala mice are protected from PKA phosphorylation-induced depletion of calstabin2 and RyR2-Ser2808Asp mice exhibit decreased levels of calstabin2 in the RyR2 complex ^19,20 as confirmed by others ^29,32.

Offering opposing views, Bers and colleagues, in disagreement with their own previous study which concluded that there is “an important SR-stabilizing effect of FKBP in intact rat ventricular myocytes” ⁷⁸, reported that FKBP12.6 (calstabin2) does not play a significant role in cardiac physiology ⁷². Using a fluorescent-labeled calstabin2 with 1 nM affinity for RyR2 ⁷² (compared to 160 nM reported by Fleischer ⁷⁹) they estimated that ≤ 20% of the calstabin2 binding sites on RyR2 are occupied and that calstabin2 is not an important RyR2 regulator. Fleischer had reported that 83% of the calstabin2 sites on RyR2 are occupied ⁸⁰.

Meissner showed that PKA phosphorylation of RyR2 does not cause calstabin dissociation from RyR2 using a molar excess of calstabin (see Figure 1B in ⁷³). Indeed, other groups have reported that PKA phosphorylation does not remove castabin2 from RyR2 ^{64,81,82,83–85} and/or that CaMKII is the more important kinase regulating the channel ^86–90. Since RyR channels have > 30 exposed free cysteines in each monomer (>120 per tetrameric channel) ⁹¹, some disagreements about calstabin2 binding can be explained by RyR2 oxidation as it is customary, for example, to add antioxidants such as DTT to samples.

Valdivia and colleagues have challenged the roles of PKA hyperphosphorylation, calstabin2 depletion and diastolic SR Ca²⁺ leak via RyR2 channels in HF ⁸⁵. We originally showed evidence for SR Ca²⁺ leak via HF RyR2 as manifested by increased open probability at diastolic [Ca²⁺]_cyt (~150 nM). Valdivia used systolic [Ca²⁺]_cyt (5 μM) and not surprisingly reported no leak since RyR2 is open at this activating [Ca²⁺]_cyt and leak has no meaning when the channel is already open ⁸⁵. Valdivia and colleagues further reported that HF RyR2 were not PKA hyperphosphorylated or depleted of calstabin, but did not normalize the amount of RyR2 protein (see Fig. 8A in ⁸⁵) or co-immunoprecipitate RyR2 resulting in contamination of their RyR2 samples with calstabin1 ⁸⁰.

Chen has proposed an alternative mechanism for RyR2 leak which he refers to as store overload induced calcium release (SOICR) ⁹². In the Chen model SOICR depends on a Ca²⁺ sensor ⁹² that turns out to be cytoplasmic based on the channel structure ^93,94.

CaMKII-dependent phosphorylation of RyR2 is increased in HF and alternative mechanisms involving CaMKIIδ have been proposed to explain SR Ca²⁺ leak in HF. ⁹⁵ Mice expressing AC3-I, the CaMKII inhibitory peptide, are protected against HF ⁹⁶, an RyR2-S2814A mutation that ablates CaMKII mediated RyR2 phosphorylation protects against HF after transverse aortic constriction (TAC) ⁸⁸, but not after myocardial infarction ⁵⁸. TAC is a hypertrophy model and CaMKII phosphorylation of RyR2 is associated with the hypertrophy. RyR2 CaMKII phosphorylation is not increased in a post-MI model but plays an important role in rate-related increase in contractility known as the Bowditch phenomenon ⁵⁸.

Diastolic SR RyR2 Ca²⁺ leak promotes an inward depolarizing current via NCX1 antiporters which leads to delayed afterdepolarizations (DADs) and the initiation of triggered activity ⁹⁷. At rest the leak is minimal, during adrenergic stress the cardiomyocyte becomes loaded with Ca²⁺ (PKA phosphorylation of DHPR, PLN and SERCA2a conspire to increase SR Ca²⁺ levels) and the leak is exacerbated, leading to arrhythmogenesis. RyR2-Ser2808 was the first residue identified to be phosphorylated in HF leading to calstabin2 dissociation, enhanced RyR2 open probability (P₀), and diastolic Ca²⁺ leak²⁶. Wehrens et al showed that preventing PKA hyper-phosphorylation of RyR2 inhibits progressive cardiac dysfunction in mice with ischemia-induced HF using genetically altered mice in which serine 2808 on RyR2 has been replaced by Ala (RyR2-Ser2808Ala mice)³⁵. Thus, there is an important role for RyR2-Ser2808 in RyR2 regulation by the sympathetic nervous system (SNS). These findings were independently validated by Walweel et al ⁹⁸ and are in line with the efficacy of beta-blockers in HF patients where the beta-antagonist reduces the incidence of cardiovascular events as well as the mortality in patients with chronic heart failure⁹⁹. Indeed, because the RyR2 of RyR2-Ser2808Ala mice cannot be PKA-phosphorylated, calstabin2 binding to RyR2 is not reduced as seen in HF mice.

Multiple animal models have been used to support a role for PKA hyperphosphorylation of RyR2 in HF progression ^{19,20,26,35–37,39,50}. As noted, genetically altered mice harboring RyR2 that cannot be PKA phosphorylated (RyR2-S2808A), are protected against calstabin2 depletion from the RyR2 complex and HF progression 4 week post-MI ³⁵. Others reported that they could not reproduce these results, but published data in strong support of ours showing preserved cardiac function in RyR2-S2808A mice compared to WT mice (see their Supplemental Table 1, fractional shortening second line from the bottom in Reference ⁶⁴) in a transverse aortic constriction (TAC) model of HF. These and other differences have been addressed in detail elsewhere ²³.

Additionally, PDE4D3 deficient mice develop an age-dependent cardiomyopathy and arrhythmias, RyR2 PKA hyperphosphorylation and calstabin2 depletion. Crossing the PDE4D3 deficient mice with the RyR2-S2808A mice resulted in protection against cardiomyopathy ⁵⁰. Genetically altered mice expressing mutant calstabin2-D37V, which binds to PKA phosphorylated RyR2 channels, are protected against post-MI HF ¹⁰⁰. In summary, there is an important role for chronic PKA hyper-phosphorylation of RyR2 in HF progression as demonstrated using RyR2-S2808A mice.

Recently, high-resolution RyR cryo-EM structures have provided further details supporting phosphorylation-mediated diastolic Ca²⁺ leak. The RyR2-Ser2808 PKA phosphorylation site was found on the P2 rim of the helical domain where limited transient PKA phosphorylation may weaken its contact with SPRY3 of the neighboring protomer and contribute to a flexible conformation of RyR2 channels and enhanced open state probability (Fig. 4). More extensive PKA phosphorylation for longer periods of time might disrupt critical domain interactions resulting in destabilization of the channel closed state and diastolic SR Ca²⁺ leak however this possibility remains to be explored experimentally with new structures of PKA phosphorylated channels. We have also reported RyR2 PKA hyperphosphorylation at Ser2808 in a canine model of AF¹⁰¹. Aberrant diastolic SR Ca²⁺ release and DADs in atrial myocytes are capable of triggering arrhythmia giving arise to AF⁶⁵.

Figure 4. RyR2 phosphorylation domain and CPVT-associated mutation hot spots.

CPVT-associated RyR2 mutations (http://www.uniprot.org/uniprot/Q92736) in the N-terminal domain (red spheres), bridging solenoid (blue spheres), core solenoid (purple spheres), transmembrane domain (orange spheres), and C-terminal domain (magenta), viewed from the cytosol (left) and from the plane of the SR membrane (right). Insets show the domain-domain interface between the NTD and the bridging solenoid at the ‘zipping’ point where DPc10 (black ribbon) is located, the newly revealed calmodulin binding site (green ribbon) and all 4 known ARVD-causing mutations (green spheres). On the right inset the interaction between CTD (cornflower blue ribbon) and core solenoid TAF motif also shows mutation clusters on both sides of this domain-domain interface. The location of the phosphorylation domain containing the PKA and CaMKII sites is shown.

The role of another potential RyR2 PKA phosphorylation site, Ser2030, in HF pathogenesis remains unclear. Ser2030 phosphorylation had no effect on RyR2 function (see their Fig. 7 in ¹⁰²), and our data concur³⁵. However, a recent publication of an RyR2Ser2030A mouse model reported a blunted response to catecholamines ⁸¹. Moreover, absence PKA phosphorylation of RyR2-Ser2808Ala in knock-in mice indicates that Ser2808 is the only RyR2 PKA phosphorylation site in-vivo ^35,63.

A distinct phosphorylation site, RyR2-Ser2814, is phosphorylated by CaMKII⁶³. However, mice engineered with channels that cannot be phosphorylated by CaMKII (RyR2-Ser2814Ala) are not protected in an ischemic model of HF⁵⁸. Reactive oxygen species (ROS) induce CaMKII activation by direct oxidation of the enzyme’s regulatory domain¹⁰³. Both oxidative stress and CaMKII have been linked to HF and AF. Oxidative stress is elevated in patients with AF¹⁰⁴, and CaMKII-dependent phosphorylation of RyR2 has been reported to cause SR Ca²⁺ leak that may contribute to AF⁶¹.

In contrast we found that CaMKII phosphorylation of RyR2 at RyR2-Ser2814 is required for rate-dependent increase in cardiac contractility thus providing a molecular mechanism for this phenomenon first described almost 150 years ago by Henry Bowditch in 1871⁵⁸! This is a critical physiological response that enables the heart to maintain or increase cardiac output as heart rate increases with exercise. An impaired or blunted positive force-frequency relationship results in diminished cardiac output at higher heart rates which blunts the ability to respond to stress (fight-or-flight response). While the role of CaMKII phosphorylation of RyR2Ser2814 provides a mechanism for the Bowditch effect, the role of CaMKII phosphorylation of RyR2Ser2814 in HF and AF remains uncertain³⁵. Although a link between adrenergic activation, the major trigger of cardiac arrhythmias, and CaMKII activation has been reported¹⁰⁵.

Reduced phosphodiesterase activity⁵⁰, causing impaired dephosphorylation of RyR2, contributes to SR Ca²⁺ leak in humans and animal models with cardiac disease^106,107. Three major serine-threonine phosphatases including PP1, PP2A, and PP2B modulate RyR2 function. Our group has shown that PP1 and PP2 are tethered to RyR2 via the leucine-isoleucine zipper motif of their targeting proteins spinophilin and PR130, respectively^108,109. In failing hearts, spinophilin is depleted from the RyR2 macromolecular complex resulting in reduced PP1, which promotes increased PKA hyper-phosphorylation of RyR2¹¹⁰. Decreased PP2A binding to RyR2 is also associated with RyR2 PKA hyper-phosphorylation²⁶. PDE4D3 plays a critical role in regulating the local cAMP levels around the RyR2⁵⁰. Reduced PDE4D3 activity in HF or inhibition of PDE4D3 with the drug rolipram can promote RyR2 leak and cardiac dysfunction and arrhythmias⁵⁰.

3). RyR2 oxidation and nitrosylation: role in HF and cardiac arrhythmias

Cardiac contractility is determined by the amplitude and kinetics of the Ca²⁺ transient. We showed that RyR2-mediated diastolic SR Ca²⁺ leak contributes to HF progression ²⁶ and fatal ventricular arrhythmias ^36,59,65,111. Before we reported the role of diastolic SR Ca²⁺ leak in HF, the dogma was that SR Ca²⁺ overload occurred in HF ^{112,113,114,115}. Others have confirmed both SR Ca²⁺ depletion and diastolic SR Ca²⁺ leak in HF ^{27,31,32,116,117}.

RyR2 is modified by superoxide anions, hydroxyl radicals and by reactive nitrogen species (RNS) such as nitric oxide (NO) and xanthine oxidase. Given the number of cysteine residues located in the same region as the Ser2808 phosphorylation site, it is plausible that excessive thiol modification in HF could potentiate RyR2 Ca²⁺ leak. HF is associated with increased ROS production from the uncoupled mitochondrial electron transport chain (ETC) as well as upregulation of nitric oxide synthase (NOS), xanthine oxidase (XO) and NADPH oxidase (NOX) resulting in increased generation of O₂^.- and decreased GSH/GSSH ratio¹¹⁸. In cardiac myocytes, mitochondria occupy 30–40% of the cellular volume and constitute the major source of intracellular ROS production. Up to 20% of the electrons in the ETC leak molecular oxygen to form superoxide anion and subsequently hydrogen peroxide (H₂O₂)^119,120. Excessive mitochondrial ROS production has been implicated in the pathogenesis of HF and AF.

Moreover, electron microscopy studies have shown that mitochondria are in close proximity to the RyR2 with distances varying between 40 and 180 nm in rat myocardium¹²¹. Thus, mitochondrial ROS production modulates local as well as global RyR2-mediated Ca²⁺ release ¹²². NOX2 and NOX4 are the predominant isoforms of NADPH oxidase expressed in cardiac myocytes and are upregulated in patients with heart failure^123,124. However, there are important differences between the two NOX isoforms with respect to their structure, localization, function, and pathophysiological significance¹²⁵. NOX2 is located on the plasma membrane and is usually dormant. It is activated either mechanically (e.g. aortic banding) or hormonally (e.g. angiotensin II) increased myocardial afterload¹²⁶. NOX4, on the other hand, is constitutively active and its precise localization (e.g. cytoplasm, mitochondria, SR/ER) remains controversial^127,128.

The contribution of NOX to SR Ca²⁺ leak is also supported by the observation that addition of NOX to microsomes isolated from cardiac muscle significantly enhances S-glutathionylation of RyR2 and CICR¹²⁹. Indeed, Sanchez. et al provided indirect evidence that NOX2 contributes to RyR2 glutathionylation during tachycardia¹²⁹. We have recently demonstrated a physical interaction between NOX4 and the skeletal muscle isoform of the RyR (RyR1) during cancer-induced muscle weakness¹³⁰. In this condition, the TGF-β pathway activates NOX4 in skeletal muscle causing RyR1 oxidation and increased intracellular Ca²⁺ leak, reducing SR Ca²⁺ and weakening muscle force production¹³⁰. However, regulation of RyR2 by NOX still requires further investigation.

Oxidative stress during HF is associated with diastolic SR Ca²⁺ leak via enhanced RyR2 activity^131,132. Cys-oxidation of RyR2 increases channel open probability causing diastolic Ca²⁺ leak, which leads to contractile dysfunction. In atrial myocytes this leak may trigger AF¹³¹. Importantly, reducing mitochondrial ROS production attenuates atrial diastolic SR Ca²⁺ leak and prevents AF. Combination of RyR2 PKA hyper-phosphorylation and Cys-oxidation contribute to calstabin2 dissociation, Ca²⁺ leak, and myocardial dysfunction ²⁰. This synergy could be due to PKA phosphorylation exposing additional cysteine residues to oxidation and/or Ca²⁺ leak from PKA hyper-phosphorylation causing mitochondrial Ca²⁺ overload and ROS production which causes RyR2 oxidation, perpetuating the RyR2 Ca²⁺ leak.

4). RyR2 oxidation and nitrosylation: role in atrial fibrillation

Elevated levels of reactive oxygen species occur in persistent AF and AF recurrence after radio frequency catheter ablation in paroxysmal AF patients ¹⁰⁴. The Framingham Study shows that HF is the strongest predictor for AF ¹³³. HF results in RyR2 PKA hyperphosphorylation, oxidation, nitrosylation and calstabin2 dissociation from RyR2 molecular complex.

Despite intense interest and research for more than a century – much remains to be understood about the most common cardiac arrhythmia, atrial fibrillation ^134,135, and treatment remains a challenge ^136–141. At a gross, anatomical level of understanding, atrial enlargement and fibrosis play a role in AF maintenance ¹⁴². But what triggers AF?

Since AF occurs in structurally normal hearts atrial enlargement and fibrosis cannot be considered de facto requirements for AF. Indeed, individuals with leaky RyR2 channels but structurally normal hearts, who suffer from exercise-induced sudden cardiac death., known as catecholaminergic polymorphic ventricular tachycardia (CPVT) ^143,144, also have AF ^59,145,146. Diastolic SR Ca²⁺ leak via RyR2 is a feature of AF ^61,101. Knock-in mice harboring leak-inducing RyR2 mutations are susceptible to AF ^16,17,147. A striking example is a 2-year-old child with a CPVT linked RyR2 mutation and AF¹⁴⁴

Oxidative stress is a feature of AF ^148–152, and antioxidant drugs have some benefit ^149,151,152. AF increases as does oxidative damage ^153,154. and RyR2 undergoes stress-induced oxidation ¹⁴⁶ in both ventricular and atrial myocytes ^{20,146,155–157,17}. We have used murine models of human CPVT (e.g. RyR2-R2474S^+/−, RyR2-R2386I^+/−, and RyR2-L433P^+/− mice) and mice expressing a phosphomimetic aspartic acid residue at position 2808 (RyR2-S2808D^+/+) leading to constitutively leaky channels as ideal models to explore the mechanisms of AF because they have structurally normal hearts. We found that mitochondrial free radicals oxidize atrial RyR2, associated with depletion of calstabin2 from the RyR2 channel complex resulting in diastolic SR Ca²⁺ leak and pacing-induced AF that can be rescued by overexpressing human catalase targeted to mitochondria (mCAT mice) or the rycal S107 ¹³¹.

The role of castabin2 binding to RyR2 in AF was first reported in myocytes from chronic AF patients ¹⁰¹ and Sood et al. reported pacing-induce AF in calstabin2 knockout mice ¹⁵⁸. We showed that intra-esophageal pacing protocol induced AF in calstabin2 knockout mice that was not prevented by S107 in calstabin2 deficient mice showing the requirement for calstabin2 in the mechanism of action of the rycal. Interestingly, in contrast to ventricular arrhythmias that are induced by exercise and epinephrine, AF was induced without catecholamines was not inhibited by β-blocker treatment suggesting that catecholamines and PKA phosphorylation of RyR2 may not be as important in triggering atrial arrhythmias as it is in ventricular arrhythmias.

Chelu et al. reported CaMKII phosphorylation of RyR2 in mice with AF ¹⁵⁹. The CaMKII inhibitor KN93 suppressed AF suggesting that CaMKII phosphorylation of Ca²⁺ cycling proteins including the L-type Ca²⁺ channel and phospholamban which modulates SERCA2a to regulate SR Ca²⁺ uptake may be involved as well as RyR2. We showed that RyR2-S2814A mice harboring RyR2 that cannot be CaMKII phosphorylated exhibit similar progression of heart failure after myocardial infarction compared to WT littermates ⁵⁸and similar incidences of AF arguing that CaMKII phosphorylation of RyR2 does not play a major role in triggering AF in HF.

Rat atrial myocytes have higher SR Ca²⁺ uptake, ~3-fold higher SR Ca²⁺ load compared to ventricular myocytes ¹⁶⁰ and increased Ca²⁺ spark frequencies in both WT and RyR2-R2474S^+/− atrial myocytes compared to their ventricular counterparts and may lower the threshold for induction of atrial arrhythmias induced by burst atrial pacing. The baseline Ca²⁺ spark frequencies of WT atrial myocytes and RyR2-R2474S^+/− ventricular myocytes were comparable (Figure 5, first and fourth bar), indicating comparable diastolic SR Ca²⁺ leak. This leak by itself is not sufficient to induce AF in vivo burst pacing stimulation in WT mice, or ventricular arrhythmias in RyR2-R2474S^+/− mice. Clinically, VT is observed during exercise in patients with the RyR2-R2474S mutation indicating the importance of sympathetic activation of the SR Ca²⁺ uptake pathway and loading of the SR to increase the amplitude of the leak. The exact reasons for these differences are still not well understood. However, it is well known that unlike ventricular fibrillation which leads to sudden cardiac death, AF is typically not lethal in the absence of a by-pass tract. Therefore, there might be less evolutionary pressure to maintain a higher threshold for arrhythmias in the atria.

Figure 5. Architecture of the open (left) and closed (right) states of RyR2.

Side view (top panels) and top down views (bottom panels) are shown for the closed state (right) and the open state (left). In the closed state RyR2Glu4873 (red spheres) are in close contact with RyR2Arg4875 (blue spheres) of the adjacent protomer preventing Ca²⁺ flux through the pore, while in the open state (left) Arg4875 rotates out of the pore to open the channel. One of the four monomers forming the homotetrameric channel pore region is shown in cyan. Dashed lines indicate the transmembrane region.

Although re-entry and multiple wavelets are observed in AF, the molecular events initiating AF remain uncertain. In agreement with previous reports implicating abnormal Ca²⁺ handling in AF ^61,161, we observed Ca²⁺ waves and indirectly Ca²⁺-activated inward current (DADs). In the setting of normal cardiac structure and function in CPVT mouse models, the diastolic SR Ca²⁺ leak via RyR2 leads to Ca²⁺ waves and DADs, wavelets and likely re-entry loops that trigger atrial tachycardia ^{17,101,131,162}. Thus, we propose that atrial diastolic SR Ca²⁺ leak is an essential contributing factor in the pathogenesis of AF ^16,17,147. Diastolic SR Ca²⁺ leak due to RyR2 dysfunction may be linked to multiple factors including phosphorylation, oxidation or pathological mutations of RyR2, all of which can result in channel dysfunction¹⁴⁶.

Accumulating evidence suggest that oxidative stress ¹⁰⁴ plays a pivotal role in the triggering and maintaining AF ^{148–150,163,164}. Our findings strongly support others who have suggested that AF is associated with myocardial oxidative stress ^148,163We have previously reported PKA hyperphosphorylation and calstabin2 dissociation from RyR2 in several models of chronic AF¹⁰¹. The importance of PKA phosphorylation in cardiac disease has been challenged by Valdivia’s group, who concluded that phosphorylation of Ser²⁸⁰⁸ plays no role in β-adrenergic cardiac response. The fact that our findings have been confirmed by multiple other groups is addressed in a recent review¹⁶⁵.

Chronic PKA phosphorylation of ventricular RyR2 is associated with oxidation of the channel and the combination of PKA phosphorylation and oxidation of RyR2 results in significantly more calstabin2 dissociation from RyR2 compared to each post-translational modification alone ²⁰. Furthermore, oxidation of atrial RyR2 has been shown to be a key contributor to diastolic SR Ca²⁺ leak and AF in animal models of CPVT¹⁷.

CaMKII is also a molecular target of oxidative stress, and oxidized CaMKII can phosphorylate RyR2 at Ser²⁸¹⁴ inducing intracellular Ca²⁺ leak ¹⁶⁶. KN-93 treatment reduced SR leak and arrhythmic activity in atrial cardiomyocytes ¹⁶⁷. Inhibition of late sodium current and CaMKII in right atrial tissue decreased arrhythmic activity ¹⁶⁸. Wehrens and colleagues showed that the phosphatase PP1 ¹¹⁰ regulates RyR2 locally by dephosphorylating the CaMKII phosphorylation site in RyR2. Decreased local PP1 regulation of RyR2 contributes to RyR2 hyperactivity and promotes AF susceptibility ¹¹⁰.

There are multiple sources of ROS, including mitochondria, NADPH oxidases and NOS uncoupling that contribute to AF ^138,141,169 and mitochondria are the major ROS source for long-term AF ¹⁵³.

Mitochondria and the SR are co-localized in the ‘mitochondrial microdomain’ ¹⁷⁰ and mitochondrial Ca²⁺ uptake via the mitochondrial Ca²⁺ uniporter is modulated by SR Ca²⁺ release ^170–172. In atrial myocytes from RyR2-S2808D^+/+ mice, leaky RyR2 channels were associated with an age-related increase in diastolic SR Ca²⁺ release without increase in systolic Ca²⁺ transient amplitudes. Pharmacological inhibition of RyR2 Ca²⁺ leak restored atrial mitochondrial morphology and function showing that mitochondrial Ca²⁺ overload plays a key role in AF pathophysiology ^17,131.

RyR2 mediated intracellular Ca²⁺ leak caused by constitutive PKA phosphorylation or CPVT mutations triggers a vicious cycle in which atrial myocyte SR Ca²⁺ leak in causes mitochondrial dysfunction and increased ROS production promoting RyR2 oxidation and further Ca²⁺ leak. Pharmacological inhibition of leaky RyR2 channels or genetically inhibiting mitochondrial ROS production prevents AF thus leaky RyR2 could be new therapeutic targets for AF.

Increased persistent Na⁺ current due to gain-of-function Na_V1.5 dysfunction leads to spontaneous and prolonged episodes of AF, left atrial enlargement and fibrosis ¹⁷³. A murine model of gain-of-function Na_V1.5 dysfunction is relevant to human AF. AF is increased in LQT3 patients who harbor SCN5A mutations altering Na_V1.5 ¹⁷⁴. Na⁺ current is increased 26% in atrial appendages of patients with permanent AF ¹⁷⁵. Ranolazine reduced atrial arrhythmias and new AF episodes ^176,177: Increased persistent Na⁺ current is observed in heart failure, hypoxia, inflammation, oxidative stress ¹⁷⁸.

5). Coupled gating of RyR2 channels: role in HF and cardiac arrhythmias

In cardiomyocytes, RyR2 channels are densely clustered on the surface of the SR^179,180, with physical contact between the large cytosolic domains. RyR2 channels open and close in a concerted manner^34,69, in a process known as coupled gating. Calstabin2 is required for coupled gating and this process contributes to the termination of Ca²⁺ release during ECC. Without Calstabin2, channels do not couple properly and exhibit instability and failure to close stably, manifested as sub-conductance states^26,69. Decoupling and sub-conductance states are observed in RyR2 isolated from failing hearts¹⁸¹.

6). Ca²⁺ uptake, extrusion, sequestration, and buffering: role in HF and AF

Diastole begins with SR Ca²⁺ re-uptake, Ca²⁺ extrusion from the cytosol via the NCX1. The ability of SERCA2a to bind Ca²⁺ and pump it into the SR is governed by binding to phospholamban (PLB)^182,183. AAV9 vector-mediated SERCA2a overexpression in a rabbit model of AF reduced diastolic cytosolic Ca²⁺ levels by increasing SR Ca²⁺ uptake and reversed pathologic electrical and structural remodeling¹⁸⁴. SERCA2a levels are reduced in HF, resulting in SR Ca²⁺ depletion, lower Ca²⁺ transient amplitude, and reduced cardiac contractility. Accumulation of cytosolic Ca²⁺ increases Ca²⁺ extrusion via NCX1, promoting triggered activity¹⁸⁵. Redox modification of SERCA2a by ROS has also been proposed to alter pump activity, impairing cardiomyocyte relaxation¹⁸⁶.

Increasing SERCA2a activity in cardiomyocytes may improve contractility in HF, however it can also increase the incidence of arrhythmias^187,188. An initial clinical trial testing viral-mediated SERCA2a overexpression in HF was not positive ¹⁸⁹. Patients with lone AF (AF in the absence of known precipitating factors) exhibit increased phosphorylation of PLB, reducing SERCA2a inhibition. This may be a compensatory mechanism by the cell to increase Ca²⁺ uptake in the SR in order to reduce cytosolic Ca²⁺ overload¹⁶¹.

During phase 0 of the cardiac action potential, the inflow of Na⁺ through voltage-gated sodium channels causes NCX1 to transport Ca²⁺ into the cell. This improves the efficiency of CICR by priming RyR2¹⁹⁰. During phases 2 and 3 of the action potential, NCX1 extrudes Ca²⁺ from the cytosol promoting membrane repolarization and Ca²⁺ transient decay ^191,192. In HF and AF, Ca²⁺ leak from RyR2 activates NCX1, generating inward I_NCX, promoting action potential prolongation, reactivation of L-type channels and generation of DADs^193–195. DAD-mediated triggered activity contributes to arrhythmogenesis. Na⁺ channel dysfunction manifested as maintained current has been linked to AF ^162,173.

The heart is a high-energy consumption organ, with 95% of its ATP being supplied by oxidative metabolism. Mitochondria, which occupy 30% of cardiomyocytes by volume, are intimately connected to the SR^196–198 and Ca²⁺ is essential for its optimal function through the stimulation of key enzymes in the tricarboxylic acid cycle (Krebs cycle) such as pyruvate dehydrogenase phosphatase, isocitrate dehydrogenase and α-ketoglutarate dehydrogenase^199,200. Mitochondrial Ca²⁺ uptake increases ATP production in response to the energy requirements imposed by the ATPase in excitation contraction coupling²⁰¹. However, excessive mitochondrial Ca²⁺ uptake contributes to cellular dysfunction, HF, and AF progression^131,202. Ca²⁺ influx into the mitochondria occurs primarily via the mitochondrial calcium uniporter (MCU), and efflux occurs primarily via the mitochondrial Na⁺/Ca²⁺ exchanger (mNCX)^203,204. Defective mitochondrial Ca²⁺ buffering has been implicated in mitochondrial dysfunction and increased ROS production in HF and AF^{131,205–208} but the main source of mitochondrial Ca²⁺ remained uncertain for several years. Since RyR2 is the main Ca²⁺ release channel in cardiac cells, it was postulated that RyR2 Ca²⁺ release during systole creates high local Ca²⁺ concentrations (~30μM) in the vicinity of the mitochondrial membrane, leading to mitochondrial Ca²⁺ uptake^209–212. We recently demonstrated mitochondrial Ca²⁺ overload in HF²⁰² coincided with RyR2 Ca²⁺ leak. Mice with constitutively leaky RyR2 channels exhibited altered mitochondrial structure, reduced ATP content, and increased mitochondrial ROS production²⁰².

In contrast, eliminating type 2 inositol 1,4,5-trisphosphate (IP3R2) channels in cardiac muscle cells did not attenuate HF progression, alter Ca²⁺ sparks, SR Ca²⁺ load, mitochondrial Ca²⁺ level, or ROS production suggesting that these channels do not contribute to dysfunctional Ca²⁺ handling by the mitochondria in HF²⁰².

As previously noted, excessive mitochondrial Ca²⁺ uptake increases mitochondrial ROS production and oxidative stress, a hallmark of HF and AF¹⁵³. ROS react rapidly with lipids, nucleic acids, and proteins including ion channels such as the RyR. RyR2 is highly redox sensitive ^91,55, with more than 90 cysteines residues per protomer, including 21 in the free thiol state and available for redox modifications^91,213. Oxidation of the cysteine residues in RyR2 plays a role in calstabin2 dissociation from RyR2 and diastolic SR Ca²⁺ leak and contributes to HF progression ²⁰ and AF ^17,101,131. Recent work suggests that in the presence of H₂O₂, RyR undergoes covalent disulfide cross-linking between the N-terminal domain of the four neighboring subunits, adopting an open state conformation and generating SR Ca²⁺ leak²¹⁴. Rebinding calstabin2 to RyR2 using Rycals such as S107 significantly reduces its open probability and the Ca²⁺ leak, despite the persistence of channel oxidation²¹⁵. Reducing mitochondrial ROS production using antioxidants or transgenic mouse models attenuates RyR2 oxidation, prevents calstabin dissociation, diastolic SR Ca²⁺ leak, and prevents HF and AF^131,202.

Thus, there is a feedback loop between the SR and mitochondria in HF and AF in which SR Ca²⁺ triggers mitochondrial dysfunction and ROS production, which in turn causes RyR2 oxidation and enhances intracellular Ca²⁺ leak, speeding the progress of HF and AF. Future studies will need to focus on the triggering events and identify which cysteine residues are central to the pathophysiology of HF and AF. Continuous advances in resolving the structure of RyR2 using single particle cryo-EM is a promising tool to answer this question.

7). Role of IP3R1 versus RyR2 in vascular smooth muscle during HF

IP3R1-mediated SR Ca²⁺ release regulates vascular smooth muscle (VSM) tone, impacting blood pressure control and afterload (force countering cardiac output)^216–217. In contrast, RyR2-mediated Ca²⁺ release in VSM causes vasorelaxation²¹⁸. This occurs via functional coupling of RyR2 with large, transient outward K⁺ (BK) channels such that a single Ca²⁺ spark generates a large, transient outward K⁺ current, which hyperpolarizes the plasma membrane, deactivating the voltage-dependent Ca²⁺ influx and causing VSM relaxation ²¹⁹. Targeted VSM deletion of RyR2 eliminates Ca²⁺ sparks, leading to increased myogenic tone and higher systemic blood pressure²²⁰. The role of IP3R1 in determining afterload in HF has been proposed ^216,221 and might provide additional novel therapeutic targets.

8). Structural bases for RyR2 pathology

Low resolution structures of RyRs using cryo-EM reconstructions originally provided images showing the beautiful four-fold symmetry of the channel ²²². With the advent of direct electron detector cameras coupled with improved microscopes and software it rapidly became possible to resolve high-resolution RyR structures using cryo-EM^49,223 which have revolutionized our understanding of the channel structure-function relationships (Figs. 4–6).

Figure 6. Contact sites formed by calstabin2 and the RY12 domain may promote coupled gating between neighboring RyR2 channels.

Possible domain-domain interactions between neighboring RyR2 channels in arrays as suggested by previous studies (Cryo-EM, 2D crystallography and freeze-fracture images of SR membranes). The interaction between RY12 domain (Red) and calstabin2 (yellow) of the neighboring RyR2 is a potential therapeutic target that may enhance coupled gating and prevent pathological SR Ca²⁺ leak.

In early 2015, 3D structures for the mammalian RyR1^224–226 resolved the closed state of the channel (Fig. 5). The gating mechanism and activation of the channel by Ca²⁺ became clearer with structures of the open and the ‘primed’ states (an intermediate state between closed and open channels in which the channel is Ca²⁺ bound but not yet fully activated) of RyR1^223,227,228 and RyR2²²⁹. Recent RyR2 structures provide high-resolution maps of RyR2 bound to apo- and Ca²⁺-calmodulin suggesting its modulation mechanism²³⁰. These structures revealed the exact location of most disease-causing mutations, providing clues to understanding the role of the Ca²⁺ leak in RyR2 in heart failure⁵⁹, catecholaminergic polymorphic ventricular tachycardia (CPVT)²³¹, and arrhythmogenic right ventricular dysplasia (ARVD) associated with RyR2 naturally accruing mutations and pathological post translation modifications.

The structure of RyR2 can be divided into three main regions: the dynamic cytosolic N-terminus and helical solenoid domain, the central core domain, and the transmembrane (TM) domain²²⁵. The cytosolic shell is the largest assembly, composed of the N-terminal domains NTD (residues 1–642), the three SPRY domains (643–1605), the first RYR motif (P1, 861–1066), and the large, helical solenoid domain (~35% of the sequence, 1642–3528) that contains the second RYR motif with the modulatory PKA and CaMKII phosphorylation sites (P2, 2701–2907) (Fig. 4). The cytoplasmic assembly is also the scaffold for many RyR2 modulatory proteins including calstabin2, calmodulin, kinases and phosphatases ^26,39,50.

The NTD harbors a ‘hot spot’ of disease-causing mutations located at the interfaces between the four protomers²³² (Fig 4). The NTD also forms an inter-domain interface with a second mutation hot-spot clustered on the helical solenoid domain located in close proximity to the calmodulin binding site (Fig 4). The dynamics between these two moving parts was shown to be coupled to the opening and closing of the pore^223,229. Stabilizing these inter-domain contact sites is required to keep the ‘upward’ shell conformation of the closed state^223,233. Disruption of this interface may destabilize the closed state, causing a diastolic SR Ca²⁺ leak. Interestingly, two mutation clusters are adjacent to the DPc10 peptide (Fig 4, black ribbon), a short peptide (Gly2460-Pro2495) that was shown to increase arrhythmia-triggering SR Ca²⁺ leak characteristic in CPVT ²³⁴. Intriguingly, all 4 known ARVD-causing mutations (Arg176Gln, Leu433Pro, Asn2386ILe, and Thr2504Met) are located in this interface between the two disease associated clusters (Figure 4 left inset, green spheres).

The second RYR motif repeat^235,236, or P2 (Ry3&4 in RyR1), is located at the ‘top’ of the cytosolic assembly of the helical solenoid assembly. P2 is the PKA and CaMKII phosphorylation site (Ser2808 and Ser2814, respectively) (Fig. 4). It is a major channel modulatory site and was suggested as the cause for Ca²⁺ leak in chronic adrenergic stimulation-induced heart failure^35,237. Given its location and that phosphorylation of this site increases open probability, addition of a phosphate group could promote the downward conformation of the cytosolic shell leading to a pathological SR Ca²⁺ leak, possibly by modifying and disrupting inter-domain interactions with the SPRY3 domain of the adjacent protomer, or by modifying protein-protein interactions with the voltage-gated Ca²⁺ channel in the T-tubule.

As with RyR1²²³, in the presence of activating ligands, the RyR2 outward rotation of the cytosolic shell is coupled to the dilation of the channel aperture²²⁹. The aperture dilation correlates with the degree of S6 helix bowing under the pressure created by the primed central solenoid U motif, or ‘thumb and forefinger’ (TaF) domain in RyR1, that engulfs the CTD. This interface between the central domain U motif and the CTD harbors the binding site for the activating Ca²⁺ and ATP and mediates the transmission of signals between the cytoplasmic assembly and the pore aperture. Multiple CPVT-causing mutations are located on both sides of this pivotal interaction site important to the channel gating apparatus. We can now speculate that disruption of this domain-domain interaction may contribute to diastolic SR Ca²⁺ leak.

Calstabin2 plays an important role in preventing Ca²⁺ leak and reducing HF progression^65,69. The recent cryo-EM structures show that Ca²⁺ rigidifies the triangular interface between the helical solenoid domain, SPRY1, and SPRY2 similar to the role of calstabin1 in RyR1^49,223,238. Cryo-EM structures of RyR2 in the absence of calstabin suggest a greater level of flexibility of the helical solenoid evident from the lower local resolution in the P1 domain^94,230 possibly due to a higher degree of freedom that calstabin2 prevents. Additionally, calstabin plays an important role in RyR2 and RyR1 coupled gating^69,108. Freeze-fracture experiments²³⁹ and 2D crystallization in lipid membranes²⁴⁰ suggest that the P1 domain and the helical domain from the adjacent protomer form the contact sites between neighboring RyRs in the array. For RyR2 the interactions between channels in solution occur at the P1 domain and the calstabin2–SPRY1 interface²⁴¹. In both cases calstabin plays a crucial role in stabilizing the “checkerboard” array of RyR channels (Fig. 6), either by directly interacting with the neighboring SPRY1 (Fig. 6) or by stabilizing the P1-helical domain interaction and promoting coupled gating between neighboring RyR channels ^34,69. Coupled gating, which depends on calstabin binding to the channel, provides an additional mechanism for preventing channel leak because when one channel in the RyR array closes the channels in the array close due to protein-protein coupling with their neighbors as shown in Fig. 6. This pivotal contact site can control and modulate Ca²⁺ release and leak in RyR2 arrays and may be an important drug target for treatment of HF and AF.

It is important to note that the high resolution structures of RyR channels are based on single particle reconstruction and lack some regulatory proteins including Cav1.2, Casq2, junctin, and triadin²⁴². Nevertheless, key insights are rapidly emerging based on new structures which have clearly opened the way for major advances in our understanding of cardiac contractility and arrhythmias.

9. Summary and conclusions

The notion that Ca²⁺ is important in the heart is nothing new. What has emerged in the past 20 years is a detailed understanding of the molecules that regulate cardiac Ca²⁺. This has made it feasible for the first time to discover basic mechanisms that cause HF and AF. However, there is still a lot of speculation about the causes of HF and AF. So students today are not significantly better off than they were 30 years ago in terms of understanding the underlying mechanisms of AF and HF. Why is that? Well, we are still uncovering the basic mechanisms of how the heart works. For example, 30 years ago it was not known that the ryanodine receptor was the SR Ca²⁺ release channel required for ECC. You could see RyRs – they were called “feet” bridging the gap between the terminal cisternae of the SR and the T-tubule. But they were not known to be Ca²⁺ channels. Cloning RyRs provided a big step forward, as it did for the other major Ca²⁺ handling proteins in the heart and skeletal muscle. The evolution of understanding of how the heart works pretty much paralleled the revolution in technology including advances in cloning, biophysics, generation of genetic animal models and imaging.

Today we are beginning to understand molecular signals that cause HF progression and trigger AF. We have the tools to advance these understandings at a rapid rate and no doubt the mechanisms that emerge will likely not be the ones we predict. Who would have thought 30 years ago that reduced systolic contractility in HF was due, at least in part, to a diastolic SR Ca²⁺ leak? Or that such a leak could be a therapeutic target for both HF and AF and might even improve exercise capacity in patients with heart failure?

The conclusion (slightly modified) from a review on HF written over 15 years ago ¹³. holds as well today: “The challenge remains, as always, to objectively assess new information as it comes forth in this complex field (Fig. 1–3) and to maintain a rigorous commitment to careful, controlled studies. Indeed, these requirements are not new in the field of medicine. The 12th century physician-philosopher Moses Maimonides wrote in his physician’s oath over 800 years ago, “Grant me the strength, time and opportunity always to correct what I have acquired, always to extend its domain; for knowledge is immense and the spirit of man can extend indefinitely to enrich itself daily with new requirements. Today he can discover his errors of yesterday and tomorrow he can obtain a new light on what he thinks himself sure of today.”

Figure 2. Reduced Ca²⁺ transients in ventricular myocytes from failing hearts.

Representative line scan showing Ca²⁺ waves (transients) recorded in ventricular cardiomyocytes from normal (A) and failing hearts (B)²⁴⁴. The Ca²⁺ transient amplitude is lower in HF ventricular cardiomyocytes primarily because of the reduced SR Ca²⁺ content due to RyR2 diastolic leak and reduced SERCA2a, and the Ca²⁺ transient duration is longer primarily due to slower re-uptake kinetics compared with normal ventricular cardiomyocytes resulting in impaired contractility in failing hearts. Schematic representation of Ca²⁺ signaling in normal (C) and heart failure (D) ventricular myocytes. Of note RyR1 in skeletal muscle are also leaky during HF. This contributes to impaired exercise capacity which is a common symptom in patients with advanced (Stage III and IV) HF ³⁹.

1. Cardiac ryanodine receptors/calcium release channels are required for excitation-contraction coupling in heart.

2. Pathological oxidation and PKA-hyperphosphorylation of ryanodine receptors in heart failure, or due to inherited mutations, substantially contribute to SR Ca²⁺ leak, contributing to heart failure progression, atrial fibrillation, and ventricular arrhythmias.

3. Inhibiting ryanodine receptor mediated SR Ca²⁺ leak using drugs that stabilize the interactions between calstabins (FKBPs) and ryanodine receptors normalize can reduce heart failure progression and cardiac arrhythmias including atrial fibrillation.