Calmodulin kinase II regulates atrial myocyte late sodium current, calcium handling and atrial arrhythmia

Abstract

Background

Atrial fibrillation (AF) is the most common type of arrhythmia. Abnormal atrial myocyte Ca²⁺ handling promotes aberrant membrane excitability and remodeling important for atrial arrhythmogenesis. The sequence of molecular events leading to loss of normal atrial myocyte Ca²⁺ homeostasis is not established. Late Na⁺ current (I_Na,_L) is increased in atrial myocytes from AF patients together with an increase in activity of Ca²⁺/calmodulin-dependent kinase II (CaMKII).

Objective

To determine whether CaMKII-dependent phosphorylation at Ser571 on Na_v1.5 increases atrial I_Na,L, leading to aberrant atrial Ca²⁺ cycling, altered electrophysiology and increased AF risk.

Methods

Atrial myocyte electrophysiology, Ca²⁺ handling, and arrhythmia susceptibility were studied in wildtype and Scn5a knock-in mice expressing phosphomimetic (S571E) or phosphoresistant (S571A) Na_v1.5 at Ser571.

Results

Atrial myocytes from S571E but not S571A mice displayed an increase in I_Na,L and action potential duration, and with adrenergic stress have increased DADs. Frequency of Ca²⁺ sparks and waves was increased in S571E atrial myocytes compared to WT. S571E mice showed an increase in atrial events induced by adrenergic stress and AF inducibility in vivo. Isolated S571E atria were more susceptible to spontaneous atrial events, which were abrogated by inhibiting sarcoplasmic reticulum Ca²⁺ release (RyR2), CaMKII, or the Na⁺/Ca²⁺ exchanger. Expression of phospho-Na_v1.5 at Ser571 and autophosphorylated CaMKII were increased in atrial samples from human AF patients.

Conclusion

These studies identify CaMKII-dependent regulation of Na_v1.5 as an important upstream event in Ca²⁺ handling defects and abnormal impulse generation in the setting of AF.

Introduction

Atrial fibrillation (AF) is the most common arrhythmia in the United States, with 3 million affected individuals in this country alone.¹ Furthermore, the prevalence of AF is expected to increase 2.5 fold by the year 2050.² The clinical burden of AF is tremendous with high mortality and morbidity, including increased risk of stroke and heart failure.¹ Although the etiology of AF is complex, a common observation across multiple forms of the disease is that defects in atrial myocyte Ca²⁺ cycling promote abnormal electrical impulse formation and electrical/structural remodeling required for initiation and maintenance of AF.

Ca²⁺/calmodulin-dependent protein kinase II (CaMKII) is a multifunctional signaling molecule important for regulation of cardiac electrical and mechanical function.³ CaMKII targets several sarcolemmal and sarcoplasmic reticulum (SR) membrane ion channels, exchangers, and pumps essential for normal Ca²⁺ cycling. Importantly, increased CaMKII activity has been associated with dysregulation of intracellular Ca²⁺ handling proteins and arrhythmia in patients with AF and animal models.^3–7 While CaMKII-induced hyperactivity of the sarcoplasmic reticulum ryanodine receptor Ca²⁺ release channel (RyR2) is an important step leading to atrial arrhythmia,^5–8 it is unclear how the atrial myocyte maintains sarcoplasmic reticulum Ca²⁺ load in the face of abnormal RyR2 activity to drive AF. Furthermore, reduced L-type Ca²⁺ current coupled with decreased action potential (AP) duration, both observed in AF, would be expected to decrease overall myocyte Ca²⁺ load. Therefore, the precise sequence of molecular events responsible for precipitating atrial myocyte Ca²⁺ dysregulation and arrhythmogenesis remains unclear.

Mounting data support CaMKII-dependent regulation of voltage-gated Na⁺ channels (Na_v) as a driver for abnormal Ca²⁺ homeostasis and membrane excitability in cardiac arrhythmia and disease.^9–13 Notably, our group and others have shown that CaMKII phosphorylates the predominant cardiac Na_v alpha subunit Na_v1.5 to increase pathogenic late Na⁺ current (I_Na,L).^{9, 10, 13–15} Importantly, I_Na,L is upregulated in animal models and patients with chronic AF^16–18 and drugs that target I_Na,L have emerged as potential therapeutics for AF.¹⁹ Based on these combined studies, we hypothesized that CaMKII-dependent phosphorylation of Na_v1.5 is an important determinant of increased atrial myocyte I_Na,L, Ca²⁺ dysregulation and aberrant membrane excitability with implications for AF. Specifically, we sought to determine whether CaMKII, Na_v1.5, and RyR2 constitute a pro-arrhythmic feedback loop linking stress stimuli to altered intracellular ion homeostasis. To test our hypothesis and dissect the proposed pro-arrhythmia feedback loop, we used knock-in mouse models involving constitutive activation or ablation of CaMKII phosphorylation sites on Na_v1.5 (Ser571) and RyR2 (Ser2814). We report that CaMKII-dependent phosphorylation of Na_v1.5 increases atrial myocyte I_Na,L, alters Ca²⁺ homeostasis, RyR2 activity, and promotes atrial arrhythmogenesis. Importantly, our new data provide evidence for a synergistic relationship between atrial Na_v1.5 and RyR2 function mediated in part by changes in CaMKII activity. Together, these findings elucidate an important mechanism for sustained atrial myocyte Ca²⁺ dysregulation and aberrant excitability with implications for human AF.

Methods

Animal studies

All studies were performed according to protocols approved by the Institutional Animal Care and Use Committees of the Ohio State University and Baylor College of Medicine conforming to the Guide for the Care and Use of Laboratory Animals published by the U.S. National Institutes of Health (NIH Publication No. 85–23, revised 1996). Both male and female mice were used and randomly assigned to treatment groups. Mice between the ages of 2 and 6 months were used for this study.

Human heart tissue

Atrial tissue was obtained through approved relationship with Lifeline of Ohio. The Institutional Review Board of The Ohio State University provided approval for the use of human subjects. Age and sex were the only identifying information acquired. This investigation conforms with the principles outlined in the Declaration of Helsinki.

Additional methods are provided in the Supplemental Methods.

Results

Atrial arrhythmia propensity

To test the hypothesis that CaMKII-dependent phosphorylation of Na_v1.5 contributes to increased I_Na,L, aberrant ion homeostasis and arrhythmia triggers in AF, subsurface electrocardiograms were first measured from anesthetized WT mice and Scn5a knock-in mice with the CaMKII phosphorylation site Ser571 replaced with either: 1) glutamic acid (Na_v1.5-S571E) to mimic constitutive phosphorylation; or 2) alanine (Na_v1.5-S571A) to prevent phosphorylation ⁹. Na_v1.5-S571E mice showed no evidence of abnormal atrial excitability (RR, PR interval no difference between groups) or arrhythmias on the electrocardiogram at baseline compared to WT or Na_v1.5-S571A (Figure 1A–C). However, the majority of Na_v1.5-S571E mice (63%) displayed atrial arrhythmia events (premature atrial contractions) following β-adrenergic stimulation (3 mg/kg isoproterenol intraperitoneal) compared to just 17% of WT mice and 0% of Nav1.5-S571A mice (Figure 1D–E). The class IB drug mexiletine (25 mg/kg intraperitoneal) significantly reduced the incidence of isoproterenol-induced atrial arrhythmias in Na_v1.5-S571E mice (Figure 1D–E), while the neuronal Na⁺ channel inhibitor riluzole (10 mg/kg intraperitoneal) had no effect (Supplemental Figure I), supporting a role for Na⁺ current carried by Na_v1.5 in observed differences, although involvement of off target effects of mexiletine cannot be ruled out. Intracardiac programmed electrical stimulation (PES) studies were next performed to explore the role of CaMKII-dependent regulation of Na_v1.5 in AF inducibility. While spontaneous atrial arrhythmias were not observed in any group, the incidence of pacing-induced AF was significantly higher in Na_v1.5-S571E mice compared to either WT or Na_v1.5-S571A (reproducible AF was induced in 83% of Na_v1.5-S571E mice compared to 18% of WT and 20% of Na_v1.5-S571A, P<0.05) (Figure 2A–B). Arrhythmia burden was low with or without PES in both WT and Nav1.5-S571A.

Figure 1.

CaMKII-dependent phosphorylation of Na_v1.5 regulates susceptibility to atrial arrhythmias in vivo. (A-C) Representative surface electrocardiograms and summary data for RR and PR intervals from anesthetized WT, Na_v1.5-S571E (S571E), and Na_v1.5-S571A (S571A) mice at baseline. P=NS; numbers of animals are indicated in corresponding bars. (D-E) Representative surface electrocardiogram recordings and summary data on percent of animals showing atrial arrhythmia events within 10 minutes following isoproterenol injection (3 mg/kg, IP). A subset of S571E animals were treated with mexiletine (25 mg/kg IP) prior to isoproterenol injection. Red arrow denotes an atrial event (premature atrial contraction). #P<0.05 vs. S571A; Numbers of animals analyzed are provided in corresponding bars.

Figure 2.

CaMKII-depedent phosphorylation of Na_v1.5 regulates atrial fibrillation inducibility. (A) Representative electrocardiograms (top panel), intracardiac atrial (middle) and ventricular (bottom) electrograms during controlled pacing and (B) incidence of pacing-induced atrial fibrillation from WT, Na_v1.5-S571E (S571E), and Na_v1.5-S571A (S571A) mice. *P<0.05 vs. WT; numbers of animals analyzed are provided in corresponding bars.

Atrial myocyte electrophysiology

To gain insight into the molecular mechanism linking CaMKII-dependent phosphorylation of Na_v1.5 and susceptibility to AF, electrophysiological studies were performed on isolated atrial myocytes. Na_v1.5-S571E atrial myocytes showed a significant increase in I_Na,L (expressed as percent of peak or current density) compared to WT or Na_v1.5-S571A at baseline without any change in I_Na steady-state inactivation or recovery from inactivation (Figure 3A–B, Supplemental Figures II–III). Although not different between WT and Na_v1.5-S571E myocytes, peak I_Na was increased in S571A atrial myocytes, consistent with previous observations in ventricular myocytes (contributes to smaller I_Na,L differences between S571A and other groups when I_Na,L is expressed as current density vs. percent of peak, compare Figure 3B and Supplemental Figure III).⁹ Action potential (AP) measurements showed a significant prolongation of AP duration (APD) in Na_v1.5-S571E atrial myocytes compared to WT or Na_v1.5-S571A without any difference in resting V_m, AP amplitude or upstroke velocity (Figure 3C–D, Supplemental Figure IV). While afterdepolarizations were not observed in any genotype at baseline, Na_v1.5-S571E myocytes showed an increased susceptibility to development of delayed afterdepolarizations and triggered activity in response to pacing in the presence of caffeine and isoproterenol (9.1% vs. 81.8% of cells with afterdepolarizations in WT and Na_v1.5-S571E, respectively, P<0.05; Figure 3E–F). Functional changes in Na_v1.5 occurred without changes to the expression or localization of the Na_v macromolecular complex members, including Na_v1.5, CaMKII, β_IV-spectrin and ankyrin-G (Supplemental Figure V), similar to previous findings in ventricular myocytes.⁹ These data demonstrate that CaMKII-dependent phosphorylation of Na_v1.5 at Ser571 induces a specific increase in I_Na,L, and altered membrane excitability in atrial myocytes with potential implications for atrial arrhythmogenesis.

Figure 3.

CaMKII-dependent phosphorylation of Na_v1.5 regulates atrial myocyte late Na⁺ current and membrane excitability. (A-B) Representative voltage-gated sodium current (I_Na) traces and summary data at baseline from WT, Na_v1.5-S571E (S571E), and Na_v1.5-S571A (S571A) atrial myocytes. Summary data showing late sodium current (I_Na,L) in atrial myocytes during test pulses to −35, −30, and −25 mV (measured as average current from 50–150 ms following peak). I_Na,L was normalized to peak I_Na. (C-D) Representative action potentials (APs) and summary data of AP duration (APD) in atrial myocytes from WT, S571E, and S571A. APD at 90%, 75%, and 50% for atrial myocytes is shown (pacing = 1Hz). (E-F) Representative AP trace and summary data on incidence of spontaneous activity [delayed afterdepolarizations (black arrow in E) or triggered APs (red arrow in E)] in WT and S571E atrial myocytes during a pause following pacing (0.5–5 Hz) in the presence of isoproterenol (100 nmol/L) and caffeine (0.5 μmol/L). *P<0.05 vs. WT, ^#P<0.05 vs. S571A; numbers of cells analyzed are provided in corresponding bars in B, D, and F.

Ca²⁺ handling and dysregulation

The increase in I_Na,L and triggered activity in Na_v1.5-S571E atrial myocytes led us to ask whether CaMKII targeting of Na_v1.5 produced defects in intracellular ion homeostasis (Na⁺ and Ca²⁺ overload), leading ultimately to altered excitability and arrhythmia. Specifically, we hypothesized that an increase in intracellular Na⁺ would promote intracellular Ca²⁺ accumulation (presumably via the Na⁺/Ca²⁺ exchanger, NCX) and therefore enhance spontaneous Ca²⁺ release, precursors to arrhythmias. As a first step in testing this hypothesis, intracellular Ca²⁺ dynamics were assessed in WT and Na_v1.5-S571E atrial myocytes using linescan confocal microscopy. The frequency of spontaneous Ca²⁺ release events (Ca²⁺ sparks) was significantly higher in atrial myocytes from Na_v1.5-S571E mice compared to WT mice without any difference in sarcoplasmic reticulum (SR) Ca²⁺ load (Figure 4A–B and Supplemental Figure VI). There were no significant differences in Ca²⁺ spark properties such as spark amplitude, width, duration or time to peak between Na_v1.5-S571E and WT atrial myocytes (Supplemental Table I). In addition, Na_v1.5-S571E atrial myocytes displayed a significant increase in the incidence of spontaneous diastolic Ca²⁺ waves compared to WT (Figure 4C–D). Interestingly, genetic ablation of the CaMKII phosphorylation site on RyR2 by crossing Na_v1.5-S571E mice with RyR2-S2814A (S571E×S2814A) normalized diastolic Ca²⁺ handling defects observed in S571E (Figure 4A–D). These data indicate that CaMKII-dependent phosphorylation of Na_v1.5 induces RyR2 dysfunction and Ca²⁺ dysregulation, which depends in part on CaMKII-dependent phosphorylation of RyR2.

Figure 4.

CaMKII-dependent phosphorylation of Na_v1.5 regulates atrial myocyte Ca²⁺ homeostasis. (A-B) Representative Ca²⁺ sparks and summary data (spark frequency) and (C-D) representative diastolic Ca waves and summary data (percent of cells) in WT, Na_v1.5-S571E (S571E), and S571E crossed with RyR2-S2814A (S571E×S2814A) atrial myocytes. *P<0.05; **P<0.01; numbers of cells and animals analyzed are provided under corresponding bars in B and D (116, 345 and 403 total sparks were analyzed for WT, S571E and S571E×S2814A, respectively).

Inducibility for Ca²⁺-mediated arrhythmias

The functional relationship between I_Na,L, abnormal Ca²⁺ homeostasis and atrial arrhythmias was explored in more detail using optical mapping of isolated left atrial preparations from WT and Na_v1.5-S571E mice (Figure 5A). Overdrive pacing in the presence of isoproterenol (200 nmol/L) promoted frequent spontaneous Ca²⁺ transients following termination of pacing in Na_v1.5-S571E atria but not WT (Figure 5B–C). Na_v1.5-S571E atria with pharmacological CaMKII inhibition using autocamtide 2-related inhibitory peptide (S571E+AIP, 1 μmol/L) showed a decrease in spontaneous events compared to Na_v1.5-S571E control (Figure 5B–C). In preparations where spontaneous Ca²⁺ transients were observed, S571E+AIP showed a delay in the latency to the first spontaneous event compared to Na_v1.5-S571E control (Figure 5D). Pharmacological inhibition of RyR2 with dantrolene (10 μmol/L) completely normalized the incidence of spontaneous Ca²⁺ transients in Na_v1.5-S571E atria. Genetic ablation of the CaMKII phosphorylation site on RyR2 (S571E×S2814A) reduced the number and latency spontaneous events compared to Na_v1.5-S571E control (Figure 5B–D), although not to the extent observed with CaMKII or RyR2 inhibition (non-significant trend). Finally, inhibition of NCX with SN-6 (10 μmol/L) reduced spontaneous events and delayed latency to the first spontaneous event in S571E atria (Figure 5B–D). Together these data support that I_Na,L acts through NCX, CaMKII and RyR2 to increase atrial arrhythmia events. While phosphorylation of RyR2 via feedback on CaMKII influences I_Na,L-induced arrhythmia, it is clearly not the sole determinant. Presumably, equally important is an increase in RyR2 open probability induced by SR Ca²⁺ overload.

Figure 5.

Na_v1.5 acts synergistically with CaMKII and the ryanodine receptor Ca²⁺ release channel (RyR2) to mediate atrial arrhythmogenesis. (A) Representative isolated left atrial preparation used for optical mapping of spontaneous Ca²⁺ activity ex vivo. (B) Representative Ca²⁺ transients in isolated left atrium from WT, Na_v1.5-S571E (S571E), S571E+AIP (1 μmol/L), S571E crossed with RyR2-S2814A (S571E×S2814A), S571E+dantrolene (dant, 10 μmol/L),S571E+SN6 (10 μmol/L) in the presence of isoproterenol (200 nmol/L). Arrows denote spontaneous Ca²⁺ oscillations preceding triggered Ca²⁺ transients. (C-D) Summary data on inducibility of spontaneous Ca²⁺ transients and latency to first triggered event. *P<0.05 vs. S571E; numbers of independent preparations analyzed are provided in corresponding bars (71, 9, 23, and 13 ectopic beats were analyzed for S571E, S571E+AIP, S571E×S2814A, and S571E+SN6 respectively).

Expression of phospho-Na_v1.5 in AF patients

To determine a potential role for CaMKII-dependent phosphorylation of Na_v1.5 in human disease, the phosphorylation status of Na_v1.5 at Ser571 [phospho-Na_v1.5(Ser571)] was evaluated in atrial samples from human patients with AF or sinus rhythm (SR) [average age = 53.3±2.0 and 58.3±1.2 for sinus rhythm (SR) and AF, respectively; sex (M/F) = 2/5 and 3/3 for SR and AF, respectively]. Consistent with reports of increased CaMKII activity and I_Na,L in human AF, we observed a significant increase in phospho-Na_v1.5(Ser571) in atrial lysates from human AF patients compared to SR (Figure 6). These data indicate that CaMKII-dependent phosphorylation of Na_v1.5 at Ser571 is relevant to human AF, and may be linked with the upregulation of I_Na,L.

Figure 6.

Patients with atrial fibrillation have increased CaMKII-dependent phosphorylation of the voltage-gated cardiac Na⁺ channel, Na_v1.5. (A) Representative immunoblots and (B) summary densitometric data showing levels of phosphorylated (relative to total) Na_v1.5 and CaMKII (Ca²⁺/calmodulin-dependent protein kinase II) in right atrial samples from human patients with atrial fibrillation (AF) compared to patients in sinus rhythm (SR). *P<0.05 vs. SR; numbers of independent subjects analyzed are provided in corresponding bars [average age = 53.3±2.0 and 58.3±1.2 for SR and AF, respectively; sex (M/F) = 2/5 and 3/3 for SR and AF, respectively].

Discussion

In this study, we examined the role of CaMKII-dependent phosphorylation of Na_v1.5 in regulating atrial myocyte I_Na,L, Ca²⁺ homeostasis, abnormal excitability and arrhythmias. Using a variety of knock-in mouse models, we examined the role of this regulatory pathway in the development of atrial arrhythmogenesis. Specifically, we demonstrated that CaMKII-dependent phosphorylation of Na_v1.5 is important for enhancing atrial myocyte I_Na,L, leading to defects in intracellular Ca²⁺ handling, membrane excitability, and ultimately atrial arrhythmias. Importantly, our study uncovered evidence for a CaMKII-dependent synergy between Na_v1.5 and RyR2, which provides a vehicle for exacerbation of atrial arrhythmogenesis. Finally, we showed that Na_v1.5 phosphorylation at the CaMKII site Ser571 is increased in atrial samples from patients with permanent AF.

CaMKII has emerged as an important regulatory node in the development of AF. Previous studies reported that CaMKII expression/activity is upregulated in animal models and patients with AF.^{5–7, 20} Previously, our group and others have shown that CaMKII phosphorylates Na_v1.5 to control the amplitude of I_Na,L (although not demonstrated until now in atrial myocytes).^{14, 15} At the same time, increased I_Na,L has been reported (together with increased CaMKII activity) in animal models and patients with AF, despite reports of either no change or a decrease in peak Na⁺ current.^{16, 21} Here, we report that Ser571 phosphorylation is increased in patients with permanent AF. Furthermore, phosphomimetic Na_v1.5-S571E mice display altered atrial Ca²⁺ handling, membrane excitability and an increase in AF inducibility, while Na_v1.5-S571A mice are protected from atrial arrhythmia. Together, these new data provide a potential mechanistic link between dysregulation of CaMKII, I_Na,L, Ca²⁺ dysregulation and altered myocyte excitability during AF. An important caveat to these findings is that they come mostly from the mouse with very different electrophysiology and Ca²⁺ handling properties compared to human.²² Similarly, the importance of I_Na,L in human AF is unresolved, with conflicting data about its expression in atrial myocytes from human AF patients.^{16, 23} While our studies support that dysregulation of Na_v1.5 occurs in human AF, they do not rule out that in common forms of AF, changes in CaMKII-dependent phosphorylation of Na_v1.5 may be an effect rather than cause of arrhythmia.

Previous work has shown that increased I_Na,L disrupts intracellular Ca²⁺ handling.^{24, 25} Furthermore, mathematical modeling and experimental studies have hinted at the existence of a pro-arrhythmic synergistic relationship between I_Na,L and CaMKII ^26–28 Here, we use a host of transgenic mouse to dissect the functional significance of the proposed feedback loop in atrial myocytes and provide direct experimental evidence that increased I_Na,L potentiates Ca²⁺ mishandling, in part, through feedback on CaMKII activity (and subsequent hyperphosphorylation of RyR2). At the same time, we find that I_Na,L is capable of inducing defects in Ca²⁺ handling independent of CaMKII or RyR2 phosphorylation. Thus, it is likely that passive loading of the SR, independent of CaMKII-dependent feedback, is an important factor driving Ca²⁺ defects. These results highlight the complex pathway for potentiation of CaMKII-dependent signals in atrial myocytes with important implications for atrial arrhythmogenesis. While this study focuses on Na_v1.5 as a source for pathogenic late current, non-cardiac Na⁺ channels (e.g. Na_v1.1, Na_v1.2, Na_v1.6, Na_v1.8) have been shown to generate a component of late Na⁺ current in cardiac myocytes with ability to disrupt the normal balance of intracellular Na⁺ and Ca²⁺.^{29, 30} Furthermore, genetic variants in SCN10A have been linked to AF risk in the human population.³¹ Together, these studies point to a potential role for Na_v family members aside from Na_v1.5 in increased late current, loss of normal ion homeostasis and arrhythmias.

Given the synergy between atrial myocyte CaMKII, Na_v1.5 and RyR2, it would be expected that perturbation of any constituent member would feed forward to produce dramatic changes in phosphorylation/activity of the entire network, leading ultimately to major impairment in Ca²⁺ cycling and arrhythmia. Although the phosphomimetic S571E mouse shows increased AF susceptibility, it does not display widespread defects in atrial myocyte excitability at baseline or spontaneous AF, consistent with the fact that few patients with long QT type 3 syndrome (LQT3; human variants that generally increase I_Na) develop AF. Similarly, while mouse models of LQT3 have shown increased susceptibility for pacing-induced AF, spontaneous AF is rare. Taken together, this evidence suggests that atrial myocytes possess a robust system of “brakes” to antagonize the CaMKII-dependent feedback loop and prevent large scale defects in excitability given perturbation in I_Na,L. Logical candidates for protective “brakes” are homeostatic ion pumps and exchangers, which have capacity to adapt to increased Na⁺ or Ca²⁺ influx to limit large scale changes in ion concentrations. Similarly, Ca²⁺ buffers in the cytosol and SR are critical for the cellular adaptation to perturbations in ion transport. Protein phosphatases may also serve as protective antagonists for CaMKII-dependent signaling in atrial myocytes. Recent studies from our group show that PP2A regulates Nav1.5 phosphorylation and late current.³² It is possible that development of spontaneous AF requires impairment of one or more components of the endogenous “brake” on CaMKII signaling in atrial myocytes. While PP1 and PP2A have been shown to have increased total activity during chronic AF,³³ local activity has been shown to be compromised.³⁴ Therefore, there may be regional heterogeneity in phosphatase activity, allowing local proteins and ion channels to become hyperphosphorylated and contribute to Na⁺ and Ca²⁺ overload. Further insights are needed to determine how CaMKII hyperactivity affects the progressive nature of AF.

Effective treatment of AF remains a challenge for the field, as over 40% of AF patients experience recurrence with existing therapies.³⁵ Furthermore, therapy is limited by procedural complications, non-cardiovascular toxicity, and efficacy. Class I drugs (e.g. flecainide) block peak Na⁺ current and have been used to suppress paroxysmal AF, but are associated with deleterious pro-arrhythmic side effects. Recently, another class of drugs have been used to treat AF. Ranolazine preferentially blocks I_Na,L, and lacks the pro-arrhythmic side effects of Class I drugs. However, recurrence remains an issue even with high doses of ranolazine.¹⁹ Limitations of ranolazine may stem from its promiscuous nature with a large number of off target effects on ion channels, including delayed rectifier K⁺ channels, L-type Ca²⁺ channels and RyR2. Moreover, effects of ranolazine on peak Na⁺ current (vs. late component) are enhanced in atrial myocytes.²¹ Taken together, these affects may limit anti-arrhythmic potential of ranolazine in atria. As our study has shown, careful targeting of I_Na,L and RyR2 may be the key to anti-arrhythmic therapies. Furthermore, given the feedback mechanisms that exist via CaMKII, targeting both players concurrently with selective drugs may ameliorate atrial arrhythmias and later recurrence.