Calcium‐mediated cellular triggered activity in atrial fibrillation

Abstract

Although atrial fibrillation (AF) is the most commonly encountered cardiac arrhythmia, the basic mechanisms underlying this disorder remain incompletely understood. During the past decade or so, it has become clear that alterations in intracellular Ca²⁺ handling may play a role in the pathogenesis of AF. Studies in small and large animal models, as well as atrial samples from patients with different forms of AF, have implicated ryanodine receptor type 2 (RyR2) dysfunction and enhanced spontaneous Ca²⁺ release events from the sarcoplasmic reticulum (SR) as a potential cause of proarrhythmic cellular ectopic (triggered) activity in AF. The molecular mechanisms leading to RyR2 dysfunction and SR Ca²⁺ leak depend on the clinical stage of AF or specific animal model studied. This review focuses on the mechanisms and role of calcium‐mediated cellular triggered activity in AF, and addresses some of the current controversies in the field.

Abbreviations

AF

atrial fibrillation

cAF

long‐standing persistent (chronic) AF

CaMKII

Ca²⁺/calmodulin‐dependent protein kinase II

CPVT

catecholaminergic polymorphic ventricular tachycardia

DAD

delayed afterdepolarization

EAD

early afterdepolarization

JPH2

junctophilin 2

LTCC

L‐type Ca²⁺ channel

NCX

Na⁺/Ca²⁺‐exchanger

pAF

paroxysmal AF

PP1

protein phosphatase 1

PKA

protein kinase A

RyR2

ryanodine receptor type 2

SERCA2a

sarco/endoplasmic reticulum Ca²⁺‐ATPase 2a

SR

sarcoplasmic reticulum

Atrial fibrillation (AF) is the most common type of cardiac arrhythmia with a rising incidence and prevalence in the developed world (Andrade et al. 2014). AF is a progressive disease that can be classified into paroxysmal AF (pAF), which converts spontaneously to sinus rhythm within 7 days (usually less than 48 h), persistent AF, which lasts for more than 7 days, long‐standing persistent (chronic) AF (cAF), which lasts for over 1 year, and permanent forms, for which no further attempts are made to restore sinus rhythm (Kirchhof et al. 2016). Although our knowledge about the pathophysiology of AF has expanded greatly during the last 20 years, particularly about the cellular and molecular basis of AF‐promoting atrial remodelling, the precise mechanisms leading to AF have remained elusive (Heijman et al. 2014a).

Independent of the underlying cause, which may be clinically diverse, it is believed that abnormal impulse formation (focal ectopic activity) and reentry are the two major determinants of AF initiation and maintenance, although the precise clinical correlates of these fundamental mechanisms are not uniformly validated and are somewhat controversial (Schotten et al. 2016). In the clinical setting atrial ectopy may occur as a result of autonomic nervous system imbalance/altered innervation, acute ischaemia, wall stress (mechanosensing), and other macroscopic determinants of arrhythmogenesis (Andrade et al. 2014). A growing body of evidence suggests that abnormal intracellular Ca²⁺ handling may play a role in both the initiation of AF episodes and cellular remodelling processes that drive AF progression to more persistent forms. This review article will provide a critical examination of the experimental evidence that cellular ryanodine receptor type 2 (RyR2)‐mediated triggered activity may contribute to the pathophysiology of AF.

Basic mechanisms of triggered activity in AF

Triggered activity can be caused at the cellular level by early or delayed afterdepolarizations (EADs and DADs, respectively). EADs involve a secondary depolarization before final repolarization of the primary action potential as a result of reopening of L‐type Ca²⁺ channels and in rare cases by reactivation of fast Na⁺ currents. On the other hand, DADs occur after completion of repolarization of the primary action potential as a result of spontaneous Ca²⁺ release events from the sarcoplasmic reticulum (SR). These SR Ca²⁺ release events through ryanodine receptor/intracellular Ca²⁺ release channels (RyR2) activate the Na⁺/Ca²⁺‐exchanger (NCX), which brings three Na⁺ ions in for each Ca²⁺ ion pumped out of the cell, producing a depolarizing transient inward current that can cause a triggered action potential leading to focal ectopic firing (Voigt et al. 2012; Heijman et al. 2014a). Repetitive focal ectopic activity may sustain AF even in the absence of an AF‐maintaining substrate. On the other hand, focal ectopic firing can also initiate a reentrant circuit that can maintain AF. Reentry is considered the main mechanism for AF maintenance, and can occur around anatomical obstacles or can be functional in nature (Heijman et al., 2012, 2016).

Role of calcium in triggered activity in AF

Contraction of the atria is initiated by electrical depolarization of the myocytes, which initiates opening of voltage‐dependent L‐type Ca²⁺ channels. The influx of Ca²⁺ into the cytosol then triggers an amplified release of Ca²⁺ from the SR via RyR2 Ca²⁺ release channels (McCauley & Wehrens, 2011). The elevated cytosolic Ca²⁺ levels induce sarcomere shortening and atrial contraction. Relaxation occurs when cytosolic Ca²⁺ concentrations are brought back to diastolic levels by pumping Ca²⁺ back into the SR via the SR Ca²⁺‐ATPase 2a (SERCA2a), or by extruding Ca²⁺ through NCX on the plasma membrane. The amplitude of Ca²⁺‐induced Ca²⁺ release depends on proper alignment of L‐type Ca²⁺ channel (LTCC) on the plasma membrane and RyR2 on the SR by a structural protein known as junctophilin‐2 (JPH2) (Landstrom et al. 2014). In addition, JPH2 binding to RyR2 prevents abnormal diastolic SR Ca²⁺ leak (Beavers et al. 2013). The open probability of RyR2 is also regulated by phosphorylation by protein kinase A (PKA) and Ca²⁺/calmodulin‐dependent protein kinase II (CaMKII) and dephosphorylation by protein phosphatase 1 (PP1) in atrial myocytes.

Abnormal SR Ca²⁺ leak has emerged as a potential contributor to cellular triggered activity in right atrial cardiomyocytes of both pAF and cAF patients (Voigt et al. 2012, 2014; Heijman et al. 2014b; Abstract figure). In patients with pAF, an increased frequency of spontaneous SR Ca²⁺ release events was observed even though sinus rhythm was present at the time of tissue collection excluding rapid atrial rate‐induced remodelling as an underlying cause (Beavers et al. 2013; Voigt et al. 2014). The underlying molecular substrate was found to involve enhanced SR Ca²⁺ loading resulting from increased SERCA2a activity (Voigt et al. 2014). In addition, increased protein expression and open probability of RyR2 channels contributed to the increased frequency of spontaneous SR Ca²⁺ release events and related DADs in pAF (Voigt et al. 2014). Increased expression of RyR2 has been attributed to reduced levels of the inhibitory microRNA‐106b‐25 cluster in pAF patients (Chiang et al. 2014a). Moreover, a reduction in the ratio between RyR2‐stabilizing protein JPH2 and RyR2 may contribute to an increased RyR2 channel open probability, although the precise mechanisms of SR Ca²⁺ leak in pAF patients remain incompletely understood (Beavers et al. 2013). Further research is needed to elucidate the precise molecular mechanisms underlying the dysfunction of RyR2 and the related propensity to cellular triggered activity in pAF patients, and to validate whether these potentially proarrhythmic cellular events indeed contribute to the initiation and recurrence of AF episodes in pAF patients in vivo.

In patients with cAF, Ca²⁺‐handling abnormalities and DAD‐mediated cellular triggered activity have also been demonstrated in right atrial cardiomyocytes from patients undergoing cardiac surgery (Hove‐Madsen et al. 2004; Liang et al. 2009; Neef et al. 2010; Voigt et al. 2012). At the molecular level, enhanced phosphorylation or oxidation of RyR2 along with an increased RyR2 open probability have been proposed to cause Ca²⁺‐handling abnormalities in cAF patients (Vest et al. 2005; Voigt et al. 2012; Fischer et al. 2015; Xie et al. 2015). Studies in a mouse model of AF revealed a potential role for CaMKII in the development of triggered activity in AF (Chelu et al. 2009). The critical role of dysfunctional RyR2 for cellular triggered activity was confirmed in several other small and large animal models (Neef et al. 2010; Nishida et al. 2011; Li et al. 2012; Shan et al. 2012; Kettlewell et al. 2013; Purohit et al. 2013; Faggioni et al. 2014; Guo et al. 2014; Macquaide et al. 2015; Zhang et al. 2015), as well as human atrial cardiomyocytes isolated from cAF patients (Neef et al. 2010; Voigt et al. 2012). In addition, RyR2 may be more leaky due to increased PKA‐mediated phosphorylation although there is less evidence supporting this notion (Vest et al. 2005, Voigt et al. 2012; Li et al. 2012). Finally, reduced activity of protein phosphatases that dephosphorylate RyR2 can also contribute to enhanced phosphorylation of RyR2 and increased SR Ca²⁺ leak (Chiang et al. 2014b, 2015). Thus, extensive evidence from animal models and patients support a potential role of SR Ca²⁺ leak and associated cellular DAD‐mediated triggered activity in patients with persistent AF.

Finally, there is emerging evidence that SR Ca²⁺ leak may directly promote the progression of AF to more persistent forms. In a mouse model of spontaneous progressive AF, inhibition of RyR2 phosphorylation (and thus, enhanced RyR2‐mediated SR Ca²⁺ leak) was shown to prevent the development of a substrate for AF maintenance despite the presence of atrial triggered activity (Li et al. 2014). These findings suggest that RyR2‐mediated SR Ca²⁺ leak drives Ca²⁺‐dependent remodelling pathways (such as the calcineurin/NFAT pathway) that contribute to the formation of AF‐promoting atrial remodelling and an arrhythmogenic substrate for AF persistence. The role of Ca²⁺‐dependent remodelling in AF progression has been discussed in greater detail in another recent review and still needs to be evaluated in large animal models and patients (Heijman et al. 2014a).

Controversies regarding the role of SR Ca²⁺ leak in AF pathophysiology

An ideal way to test the hypothesis that SR Ca²⁺ leak though RyR2 is responsible for triggered activity in patients with AF would be to conduct a clinical trial in which highly specific pharmacological RyR2 channel modulators would be tested. To the best of our knowledge, such clinical trials have not been conducted to date. Several groups have developed RyR2 channel modulators – some more RyR2‐specific than others – that have been tested in animal models and tissue samples from patients with heart disease (Nakaya et al. 2000; Kumagai et al. 2003; Wehrens et al. 2004; Loughrey et al. 2007; Toischer et al. 2010; Sadrpour et al. 2015). For example, 1,4‐benzothiazepine derivative K201 (JT519) was shown to terminate atrial flutter and AF in canine models of sterile pericarditis (Kumagai et al. 2003; Sadrpour et al. 2015). This particular compound has many off‐target effects in addition to RyR2 inhibition. Further studies are needed to test whether the aforementioned and newer generation RyR2 inhibitors can actually suppress triggered activity in patients with AF.

There is strong clinical evidence that genetic gain‐of‐function defects in RyR2 – in patients with ‘catecholaminergic polymorphic ventricular tachycardia’ (CPVT) and otherwise normal hearts – develop cardiac arrhythmias including AF (Paavola et al. 2007). Although initially neglected, it is now established that individuals with CPVT have a higher propensity to both ventricular and atrial arrhythmias including AF. Most important, DADs were recorded in a patient with CPVT, which demonstrated for the first time that genetic RyR2 dysfunction might directly cause atrial triggered activity in the clinical setting (Paavola et al. 2007). Future studies are needed to elucidate whether gain‐of‐function defects of RyR2 in patients with non‐genetic forms of AF also contribute to DADs and ectopic activity.

Because of better clinical availability, most of the work dealing with RyR2 dysfunction was performed in right atrial cardiomyocytes from patients with AF (Hove‐Madsen et al. 2004; Vest et al. 2005; Neef et al. 2010; Voigt et al. 2012; Beavers et al. 2013; Voigt et al. 2014). Although it has been a widely accepted notion that the right atrium is not a major source for AF induction and maintenance, this assumption may be incorrect. Right atrium‐located AF drivers/rotors are detectable in up to 30–50% of patients with AF and in ex vivo perfused human hearts and in some patients only ablation of such right atrium‐located drivers stops the arrhythmia (Hocini et al. 2010; Hasebe et al. 2016; Li et al. 2016; Spitzer et al. 2017). This is solid evidence that the right atrium can be a relevant source for clinical AF and accordingly studies performed in human right atrial tissue clearly provide valuable insights into the molecular basis of AF. In addition, RyR2‐mediated SR Ca²⁺ leak was recently demonstrated in left atrial cardiomyocytes from AF patients (Fischer et al. 2015), making it likely that the RyR2 abnormalities detected in right atrial cardiomyocytes from patients with AF are not restricted to the right atrium only (Hove‐Madsen et al. 2004; Vest et al. 2005; Neef et al. 2010; Voigt et al. 2012; Beavers et al. 2013; Voigt et al. 2014). Additional work in human left atrial tissue is clearly needed to further validate the role of RyR2 abnormalities for AF pathophysiology.

A few recent studies of human tissue samples and animal models have suggested that silencing of intracellular Ca²⁺ handling rather than increased SR Ca²⁺ leak contribute to AF pathogenesis (Greiser et al. 2014). Greiser et al. (2014) showed that there was no increase in SR Ca²⁺ leak despite PKA hyperphosphorylation of serine 2808 on RyR2 in rabbits with atrial tachycardia remodelling but otherwise normal atria. In their model, 5 days of atrial tachypacing at 10 Hz led to a severe downregulation of RyR2 protein levels, which according to computational modelling might offset the effects of RyR2 hyperphosphorylation on SR Ca²⁺ leak. However, such a downregulation of RyR2 has not been reported in AF patients, suggesting that this model has limited clinical relevance or represents a disease stage not previously characterized in patients or healthy individuals. In addition, the reduction in intracellular [Na⁺] observed in this rabbit model and dogs with atrial tachycardia (Akar et al. 2003) is inconsistent with findings in human atrial cardiomyocytes from cAF patients, in which a significantly (3‐fold) elevated intracellular [Na⁺] has been reported (Hammer et al. 2016), although this finding needs further independent validation. Finally, rabbits with chronic myocardial infarction‐induced atrial remodelling and dogs with congestive heart failure show DAD‐mediated triggered activity and AF (Yeh et al. 2008; Kettlewell et al. 2013), which is in line with the phenotype of patients with AF outlined above and the emerging evidence that atrial cardiomyopathy is a critical determinant of the increased propensity to pAF and more persistent forms of AF (Goette et al. 2017).

Some authors have reported reduced, rather than enhanced, spontaneous arrhythmic activity in multicellular atrial trabeculae from patients with cAF compared with patients with sinus rhythm (Sossalla et al. 2010; Christ et al. 2014). However, the study of Sossalla et al. (2010) was not specifically designed to study DAD‐mediated triggered activity and one possible explanation for the opposite results of Christ et al. (2014) is that they incubated the trabeculae with phenoxybenzamine, a non‐selective, irreversible α‐adrenoceptor blocker which inhibits calmodulin and L‐type Ca²⁺ channels, among other things (Gengo et al. 1984; Cimino & Weiss, 1988). Irreversible inhibition of calmodulin could impact intracellular Ca²⁺ release due to inhibition of various channels, including L‐type Ca²⁺ channels, Na⁺ channels and RyR2. Therefore, the fact that abnormal aftercontractions (as used as an index of arrhythmias) were not more prevalent in atrial trabeculae from cAF patients as compared to sinus rhythm patients does not disprove the importance of DAD‐mediated triggered activity (Christ et al. 2014). If anything, it is very likely that these experiments rather proved that calmodulin inhibition is a more effective antiarrhythmic strategy in multicellular preparations from cAF patients compared to patients in sinus rhythm. Moreover, contraction of multicellular atrial preparations is probably not a suitable readout for proarrhythmic events, particularly when atrial contraction itself is strongly impaired as commonly observed in AF patients (Schotten et al. 2001; Wettwer et al. 2004).

In their study, Christ et al. (2014) did not observe an increased rate of spontaneous SR Ca²⁺ release events between the regular L‐type Ca²⁺‐current (I _Ca,L)‐stimulated Ca²⁺ transients (CaTs), even under conditions of increased SR Ca²⁺ load with β‐adrenoceptor activation, an observation which is in line with the results of Voigt et al. (2012) and Beavers et al. (2013). In order to increase the susceptibility to spontaneous CaT generation, Voigt et al. (2012) used 5 mm instead of 2 mm extracellular Ca²⁺ (which increases SR Ca²⁺ load) and clamped the cardiomyocytes at −80 mV (mimicking diastolic conditions) after SR Ca²⁺ loading with regular I _Ca,L stimulation. These conditions – which simulate the situation of a pause after a run of atrial tachycardia – unmasked the higher propensity to spontaneous SR Ca²⁺ release events in atrial cardiomyocytes of cAF patients compared to controls. Moreover these findings were also validated in the current‐clamp configuration (with physiological extracellular Ca²⁺) where DADs more frequently occurred in atrial cardiomyocytes from cAF patients compared to controls. Clearly independent verification in additional cohorts of AF patients using similar patch clamp protocols is needed to further validate the presence and role of cellular triggered activity in human AF.

Conclusions

There is a growing body of evidence that abnormal RyR2 activity and enhanced SR Ca²⁺ leak may cause DADs and cellular triggered activity in animal models of and patients with AF. The molecular mechanisms underlying SR Ca²⁺ leak, in particular RyR2 dysfunction, appear to depend on the clinical stage of AF or the specific animal model studied. Further assessment of single RyR2 channel properties, comprehensive mapping of post‐translational modifications of RyR2, recording of cellular DADs and evaluation of multicellular preparations from patients with different forms of AF are needed to gain a better understanding of the molecular basis of triggered activity in AF and to validate the role and importance of focal ectopic firing for atrial arrhythmogenesis. Ultimately, the only way to test the contribution of dysfunctional RyR2 channels to macroscopic AF mechanisms in suitable cohorts of AF patients is to employ highly selective RyR2 inhibitors, which are being developed by several groups but to the best of our knowledge have not been tested yet in patients with AF. Work in perfused intact human atria as successfully employed in a very recent study could be the next intermediate step in the translation of findings obtained in a cellular context to the clinical setting (Li et al. 2016).

Additional information

Competing interests

X.H.T.W. is a founding partner of Elex Biotech, a start‐up company that develops drug molecules that target ryanodine receptors for the treatment of cardiac arrhythmia disorders. D.D. is consultant for OMEICOS Therapeutics, which develops drug molecules targeting the ω‐fatty acid metabolism as an antiarrhythmic therapeutic strategy.

Author contributions

Both authors have approved the final version of the manuscript and agree to be accountable for all aspects of the work. All persons designated as authors qualify for authorship, and all those who qualify for authorship are listed.

Funding

D.D. is supported by DZHK (German Centre for Cardiovascular Research) the German Research Foundation (DO 769/4‐1), and a National Institutes of Health grant (NIH R01‐HL131517); X.H.T.W. is supported National Institutes of Health grants R01‐HL089598, R01‐HL091947, R01‐HL117641 and R41‐HL129570, and American Heart Association grant 13EIA14560061.

Biography

Xander Wehrens did his MD and PhD degrees at Maastricht University in the Netherlands, and a postdoctoral fellowship with Andrew Marks at Columbia University. In 2005, he moved as a faculty member to Baylor College of Medicine in Houston, TX, where has been a full professor since 2011, and director of the Cardiovascular Research Institute since 2012. His research lab focuses on mechanisms underlying altered intracellular calcium handling in heart disease. Dobromir Dobrev did his MD degree at Department of Pharmacology, Dresden University of Technology, Germany. In 2010, he moved to Heidelberg University (Germany) to become full professor and chair of Division of Experimental Cardiology. Since 2012 he is full professor of pharmacology and toxicology and director of Institute of Pharmacology, West German Heart and Vascular Center, University Duisburg‐Essen, Essen, Germany. His research focuses on the molecular mechanisms of arrhythmias, particularly of atrial fibrillation.