Late sodium current in synergism with Ca²⁺/calmodulin-dependent protein kinase II contributes to β-adrenergic activation-induced atrial fibrillation

Abstract

Atrial fibrillation (AF) is frequently associated with β-adrenergic stimulation, especially in patients with structural heart diseases. The objective of this study was to determine the synergism of late sodium current (late I_Na) and Ca²⁺/calmodulin-dependent protein kinase (CaMKII)-mediated arrhythmogenic activities in β-adrenergic overactivation-associated AF. Monophasic action potential, conduction properties, protein phosphorylation, ion currents and cellular trigger activities were measured from rabbit-isolated hearts, atrial tissue and atrial myocytes, respectively. Isoproterenol (ISO, 1–15 nM) increased atrial conduction inhomogeneity index, phospho-Na_v1.5 and phospho-CaMKII protein levels and late I_Na by 108%, 65%, 135% and 87%, respectively, and induced triggered activities and episodes of AF in all hearts studied (p < 0.05). Sea anemone toxin II (ATX-II, 2 nM) was insufficient to induce any atrial arrhythmias, whereas the propensities of AF were greater in hearts treated with a combination of ATX-II and ISO. Ranolazine, eleclazine and KN-93 abolished ISO-induced AF, attenuated the phosphorylation of Na_v1.5 and CaMKII, and reversed the increase of late I_Na (p < 0.05) in a synergistic mode. Overall, late I_Na in association with the activation of CaMKII potentiates β-adrenergic stimulation-induced AF and the inhibition of both late I_Na and CaMKII exerted synergistic anti-arrhythmic effects to suppress atrial arrhythmic activities associated with catecholaminergic activation.

This article is part of the theme issue ‘The heartbeat: its molecular basis and physiological mechanisms’.

1. Introduction

Atrial fibrillation (AF) is the most prevalent, progressive tachyarrhythmia and is associated with high morbidity and mortality [1]. Rhythm control strategies for AF include anti-arrhythmic drugs, ablation, surgery and upstream therapy. However, there is still a great need for novel therapeutic drugs owing to suboptimal effects of these therapies, including limited efficacies for all these strategies, serious adverse effects potentially for anti-arrhythmic drugs, and high recurrences rate and potential complications for AF-ablation [2,3]. In patients with heart failure, ischaemic heart disease, or other cardiovascular or chronic renal diseases, the prevalence of AF is high and is more commonly related to the over-activation of the sympathetic nervous system [4]. Excessive activation of the autonomic nervous system is an arrhythmogenic trigger and may be involved in the progression of AF from paroxysmal to persistent [5]. However, β-adrenergic receptor antagonists are insufficient for rhythm control. Inhibition of late sodium current (late I_Na) by ranolazine (RAN) and a more selective inhibitor of late I_Na eleclazine (ELEC) conferred protection against ischemia- and acetylcholine-induced AF in intact animal models, without significant proarrhythmic effects [6,7].

Late I_Na was found to be increased in persistent AF and contributed to the pathogenesis of AF in animal models [8,9]. In addition, Ca²⁺/calmodulin-dependent protein kinase (CaMKII)-dependent phosphorylation of Na_V1.5 resulted in dysregulation of I_Na, i.e. reduced peak I_Na and enhanced late I_Na, to enhance AF risk was demonstrated in patients with sleep-disordered breathing [10]. A vicious circle of [Ca²⁺]_i-CaMKII-late I_Na-[Na⁺]_i is reported to promote the ventricular and atrial arrhythmias [11,12]. Enhanced CaMKII activity augments late I_Na by directly phosphorylating Na_v1.5 [10,13,14]. Theoretically, targeting any part of the circle may break the arrhythmogenic mechanism both in ventricle and atrium [14]. Thus, further evidence of the interaction between late I_Na and CaMKII on electrical abnormalities and arrhythmogenesis with triggers is important both in research and clinic. The objective of this study was to determine the role of the β activation-CaMKII-late I_Na pathway in the pathogenesis of AF underlying the β-adrenergic overactivity, and to explore the synergistic effect of combined inhibitions of late I_Na and CaMKII in alleviation of atrial electrophysiological malfunctions and severity of AF in denervated rabbit hearts and atrial myocytes.

2. Methods

(a) . Isolated rabbit heart model

New Zealand White female rabbit (weighing about 2.5–3.5 kg, aged about 3–6 months old) hearts were isolated and perfused in retrograde Langendorff mode with modified Krebs–Henseleit (K–H) solution as previously described [13]. Rabbits were initially anaesthetized with sodium pentobarbitone (50 mg kg⁻¹) through the marginal ear vein. After midsternal incision and opening of the pericardium, rabbits were euthanized via exsanguination by excising the hearts for further ex vivo study. Modified K–H solution contained (mM): 118 NaCl, 2.8 KCl, 1.2 KH₂PO₄, 2.5 CaCl₂, 0.5 MgSO₄, 2.0 sodium pyruvate, 5.5 glucose, 0.57 Na₂EDTA and 25 NaHCO₃ (adjusted to pH 7.4, bubbled with 95% O₂ and 5% CO₂, and warmed to 37°C). Bipolar Teflon-coated electrodes were placed on the epicardium near the sinoatrial node to pace the right atrium at a fixed 4.5 Hz frequency. AF was invoked by programmed stimulation with a 3 ms pulse width and threefold diastolic threshold delivered from a Grass-S88X stimulator (Astro-Med, West Warwick, RI, USA). After initiation of pacing, we had a 10–20 min equilibration period before we began experiments.

(b) . Monophasic action potential and electrocardiogram recording

A pressure-contact Ag-AgCl monophasic action potential (MAP) electrode was placed on the endocardial surface of the left atrial appendage to obtain atrial MAP, and a pseudo 12-lead electrocardiogram (ECG) was recorded using a circular Einthoven-Goldberger ECG electrode system simultaneously (Harvard Apparatus, Inc., Holliston, MA, USA). Electrical signals were digitized in real time by Biopac Wilson MAP and ECG amplifiers (Biopac MP 150, Goleta, CA, USA) [13].

(c) . Electrophysiological experiment protocols

We tested dose-dependent changes in electrophysiological (EP) parameters using various concentrations of isoproterenol (ISO, 1–15 nM) in the same heart. Programmed stimulation (eight S1 stimulations at a cycle length (CL) of 200 ms followed by a premature S2 stimulation at a progressively prolonged CL from the atrial effective refractory period (aERP)) was applied to induce AF. EP measurements include (i) aERP, defined as the shortest S1–S2 interval that resulted in a propagated response after S1 [15]; (ii) AF-related parameters: AF incidence, inducible window (difference between the longest and the shortest S1–S2 interval that successfully invoked AF) and burden (the sum of the duration of inducible AF within the AF window when the pacing CL increased stepwise from the shortest to the longest by 2 ms). Hearts were infused with ISO at a submaximal concentration (15 nM) for 15 min to achieve steady-state EP parameters and induce AF. To test the anti-arrhythmic effects, ISO-infused hearts were treated with RAN and ELEC (MedChem Express, Monmouth Junction, NJ, USA) or KN-93 (Selleckchem, Houston, TX, USA) for 8–15 min at each concentration until a steady-state maximal effect was observed.

(d) . Epicardial activation mapping

Atrial epicardial activation mapping in isolated Langendorff-perfused rabbit hearts was performed by two multi-electrode arrays that contain 32 separated electrodes (arranged in a 4, 6, 6, 6, 6 and 4 grid within an 8 × 8 mm configuration, a 0.1 mm electrode diameter and a 1.6 mm interelectrode distance) connected to a 64-channel amplifier and data acquisition system (EMS64-USB-1002, MappingLab Inc., City of Industry, CA, USA). The inhomogeneity index (IHI) was calculated at the fixed rate pacing and at the first beat of AF [16]. Isochrones were derived by an off-line analysis program (EMapScope 3.0, MappingLab Inc., City of Industry, CA, USA). Data were sampled at 10 kHz per channel.

(e) . Immunoprecipitation and Western blot analysis

Left atrial tissue was homogenized using a tissue lyser. Levels of phospho-CaMKII at T286 (no. 12716S, Cell Signaling Technology, Danvers, MA, USA) and human cardiac voltage-gated sodium channel (no. sc-271255-1, NaV1.5, Santa Cruz Biotechnology, Santa Cruz, CA, USA) were determined by immunoprecipitation and Western blot [13]. In detail, non-specific proteins were removed by mixing the protein sample and IgG of the same species used in the subsequent immunoprecipitation reaction and Protein G Plus/Protein A-Agarose (no. IP-10-10mlCN, Millipore, MA, USA). Then tissue lysates were incubated with magnetic beads covalently conjugated with anti-Na_v1.5 Ab and anti-phospho-protein kinase A (PKA) substrate Ab (RRXS*/T*, no. 9624, Cell Signaling Technology) at 4°C overnight. Sample buffer mixed with the washed beads was heated at 95°C for 5 min and subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The proteins were blotted with rabbit anti-Na_v1.5 Ab and anti-phospho-PKA substrate Ab, respectively. The relative intensity of individual bands was visualized using an electrochemiluminescence detection system (Millipore, Darmstadt, Hessen, Germany) and quantified using ImageJ software. The ratio for the control was assigned a value of 1.

(f) . Whole-cell patch-clamp measurements

Atrial myocytes were enzymatically dissociated. Whole-cell voltage and current patch-clamps (EPC-10, Heka Electronic, Lambrecht, Pfalz, Germany) were used to measure late I_Na and action potentials (APs) as previously described (filtered at 1 kHz, digitized at 10 kHz) [13]. All experiments were conducted at room temperature. We began data acquisition after 5 min of drug exposure.

For late I_Na recording, bath solution contained (in mM) 135 NaCl, 5.4 CsCl₂, 1.8 CaCl₂, 1 MgCl₂, 0.3 BaCl₂, 0.33 NaH₂PO₄, 10 HEPES, 10 glucose and 0.001 nicardipine (adjusted to pH 7.2 with CsOH). The pipette solution contained (in mM) 120 CsCl₂, 1 CaCl₂, 5 MgCl₂, 5 Na₂ATP, 10 TEA-Cl, 10 EGTA and 10 HEPES (adjusted to pH 7.3 with CsOH). Late I_Na was recorded by a 300 ms depolarizing pulse to −20 mV from a holding potential of −90 mV, and measured at 200 ms after initiation of the depolarization step. Measured currents were normalized to the membrane capacitance.

For AP recordings, bath solution contained (mM) 144 NaCl, 5.6 KCl, 1.2 MgCl₂, 1.8 CaCl₂, 5 HEPES and 11 glucose (adjusted to pH 7.4 with NaOH), and the pipettes were filled with (in mM) 5 NaCl, 30 KCl, 10 K-aspartate, 5 creatine phosphate, 10 HEPES, 10 EGTA, 0.05 cAMP and 5 Mg-ATP (adjusted to pH 7.2 with KOH). APs were continuously elicited by square current pulses of 200–300 pA amplitude and 10 ms duration at a stimulation frequency of 1 Hz. The AP durations (APD) and incidence of early afterdepolarizations (EADs) and delay afterdepolarizations (DADs) were measured.

(g) . Statistical analysis

All statistical analyses were performed using GraphPad Prism (version 6.0, GraphPad Software, Inc., San Diego, CA) and SPSS (version 19.0, SPSS Inc., Chicago, IL). The normality of the continuous variable distribution was evaluated using the Shapiro–Wilk test. Continuous data were expressed as the mean ± s.e.m. and non-normally distributed data were expressed as median with interquartile range. Normally distributed values were analysed using Student's t-tests or one- or two-way repeated measures ANOVA, followed by the Student–Newman–Keuls test, while non-normally distributed values were compared using Mann–Whitney tests. Categorical variables were summarized by absolute numbers or percentages and compared with χ² tests. A two-sided p < 0.05 was considered statistically significant.

3. Results

(a) . Propensity of isoproterenol-induced atrial fibrillation was exacerbated in sea anemone toxin II pretreated hearts

As expected, we did not observe AF when programmed atrial stimulation was administered under control conditions (figure 1a). Relative to baseline, perfusion with increasing concentrations of ISO (1–15 nM) decreased aERP (from 79.4 ± 2.1 to 52.0 ± 3.8 ms, p < 0.05, not shown), increased the incidence, inducible window, and burden of stimulation-induced AF, in a dose-dependent manner (n = 12, p < 0.05; figure 1b and figure 2a–c(i)). To test the effect of enhanced late I_Na on ISO-induced AF, we perfused hearts with ISO in the presence of sea anemone toxin II (ATX-II). Pretreatment with ATX-II (2 nM), which was insufficient to induce any atrial arrhythmias, prolonged aERP from 79.4 ± 3.1 to 84.7 ± 4.0 ms (p < 0.05; electronic supplementary material, figure S1A) and accentuated the propensities of ISO-induced AF (n = 6, p < 0.05; figure 1c,d and figure 2a–c(i)). The incidence of AF was increased to 100% with 15 nM ISO alone, and with a lower concentration of ISO (6 nM) in the presence of ATX-II (figure 2a(i)). As the concentration of ISO increases, curves of AF inducible window and burden shifted upwards and leftwards with application of 2 nM ATX-II. The AF inducible window and burden were increased by 7.2 ± 1.5 ms and 14.4 ± 2.1 s, respectively, relative to hearts treated with 15 nM ISO alone (p < 0.05; figure 2b,c(i)). Moreover, the EC₅₀ values for ISO to induce AF incidence, inducible window and burden were significantly lower in the presence than in the absence of ATX-II (1.1 versus 3.0, 4.2 versus 5.3, and 8.2 versus 9.0 nM, respectively, n = 6, p < 0.05). Taken together, these data show that enhancement of late I_Na accentuated ISO-induced AF.

Figure 1.

Representative records of AF in the absence (control) and presence of isoproterenol (ISO; 15nM) treated with either ranolazine (RAN), eleclazine (ELEC), or KN-93 in rabbit-isolated hearts. Hearts were paced at 4.5 Hz in right atrium. MAP (upper recordings in each panel) and an ECG (lower recordings in each panel) were recorded simultaneously. The MAP and ECG traces under programmed stimulation of hearts treated with either vehicle or drugs as shown on the top of each panel.

Figure 2.

Concentration–response relationships of drugs on AF incidence (a), inducible window (b) and burden (c) induced by isoproterenol (ISO) in the absence (n = 12) and presence of sea anemone toxin II (ATX-II; 2 nM, n = 6) and treated with either ranolazine (RAN; n = 8), eleclazine (ELEC; n = 8), KN-93 (n = 8), or KN-92 (n = 6) in rabbit-isolated hearts. (i) Concentration–response relationships of ISO on AF in the absence and presence of ATX-II (2 nM); (ii) concentration–response relationships of RAN, ELEC, KN −93 or KN-92 on AF induced by ISO (15 nM). 0, ISO alone; *p < 0.05 versus control (CTL); **p < 0.05 versus ISO alone.

Infusion with 15 nM ISO increased the IHI from 1.2 ± 0.03 to 1.7 ± 0.03 at constant pacing rate (4.5 Hz, n = 6, p < 0.05, figure 3a). Compared with an orderly spread of excitation beginning with localized stimulation-induced activation, the excitation pattern at the first beat of AF resulted in an inhomogeneous mode, with the IHI increased by 108% (from 1.2 ± 0.03 to 2.6 ± 0.05 ms) (n = 6, p < 0.05, figure 3).

Figure 3.

Effects of drugs on inhomogeneity of atrial conduction. (a) The index of inhomogeneity (IHI) at a fixed rate (baseline) and at the first beat of AF in hearts treated with either isoproterenol (ISO; 15 nM) alone or ISO + ranolazine (RAN; 10 µM, n = 6), ISO + eleclazine (ELEC; 10 µM, n = 6) or ISO + KN-93 (3 µM, n = 6). (b) Representative examples of epicardial conduction properties of left atrium by multi-electrode array in the absence (i) and presence (ii) of ISO at baseline, and treated with ISO (iii) and ISO + ELEC (10 µM) (iv) at the first beat of AF. *p < 0.05 versus control (CTL); **p < 0.05 versus ISO at baseline; ***p < 0.05 versus ISO + drugs at the first beat of AF.

(b) . Inhibition of late sodium current by ranolazine and eleclazine suppressed isoproterenol-induced atrial fibrillation

Late I_Na inhibitors RAN and ELEC were used to evaluate the effects on ISO-induced AF. RAN significantly inhibited the incidence, inducible window, and burden of ISO-induced AF, with IC₅₀ values of 1.0, 0.3 and 0.2 µM, respectively (n = 8; figure 2a–c(ii)). Similar anti-arrhythmic effects were also observed for ELEC, with IC₅₀ values of 1.5, 0.7 and 0.5 µM, respectively (n = 8; figure 2a–c(ii)). At a concentration of 10 µM, both RAN and ELEC abolished AF in all hearts studied (figure 1e,f), suggesting that late I_Na may participate in ISO-induced AF. Compared with ISO alone, the prolongation of aERP by 10 µM RAN and ELEC were 29.8 ± 2.6 and 16.9 ± 2.4 ms, respectively (p < 0.05; electronic supplementary material, figure S1B).

Furthermore, both RAN and ELEC at 10 µM decreased IHI from 1.7 ± 0.03 to 1.2 ± 0.04 and 1.3 ± 0.02 at fixed rate pacing and from 2.6 ± 0.05 to 1.6 ± 0.04 and 1.7 ± 0.04 at the first beat of AF, respectively (n = 6, p < 0.05; figure 3).

(c) . Inhibition of Ca²⁺/calmodulin-dependent protein kinase by KN-93 abolished isoproterenol-induced atrial fibrillation

In hearts treated with ISO (15 nM), KN-93 (0.01–10 µM) decreased the incidence, inducible window and burden of AF in a concentration-dependent manner, with IC₅₀ values of 0.4, 0.1 and 0.06 µM, respectively (n = 8; figure 2a–c(ii)). KN-93 at 6 µM abolished AF in all hearts studied (figure 1g). KN-93 at 10 µM also substantially prolonged aERP by 28.3 ± 2.1 ms in the presence of ISO (p < 0.05; electronic supplementary material, figure S1B). By contrast, inactive analogue KN-92 had no effect on these parameters (n = 6; figure 2a–c(ii)). Moreover, with the perfusion of 3 µM KN-93 in the presence of ISO, IHI decreased by 0.4 ± 0.02 at fixed rate pacing and 1.0 ± 0.03 at the first beat of AF, respectively (n = 6, p < 0.05; figure 3).

(d) . Attenuation of the isoproterenol-enhanced activity of Na_v1.5 and Ca²⁺/calmodulin-dependent protein kinase in atria by ranolazine, eleclazine and KN-93

ISO dramatically increased phospho-Na_V1.5 at serine 573 and threonine 17 loci and phospho-CaMKII protein expression levels in atria by 65% and 135%, respectively (n = 5, p < 0.05; figure 4a). By contrast, RAN (10 µM), ELEC (10 µM) and KN-93 (6 µM) attenuated these effects, reducing phospho-Na_V1.5 by 62%, 48% and 70%, respectively, and of phospho-CaMKII by 67%, 66% and 75%, respectively (p < 0.05; figure 4a). Representative protein bands expression in each group are displayed in figure 4a(i).

Figure 4.

Reverse effects of late I_Na or CaMKII inhibitors on phosphorylation of Na_v1.5 and CaMKII (a) and late I_Na (b) treated with vehicle (CTL), isoproterenol (ISO; 15 nM) or ISO + ranolazine (RAN), eleclazine (ELEC) and KN-93 (3 µM) respectively. (a) Representative Western blots (i) and the relative levels (ii) phospho-Na_V1.5 (P-Na_v1.5; n = 5) and phospho-CaMKII (P-CaMKII-δ; n = 5, normalized to GAPDH and expressed relative to normal levels) protein expression are presented. (b) Representative records (i, n = 5) and summarized absolute values (ii, n = 5) of late I_Na are presented. *p < 0.05 versus control (CTL); **p < 0.05 versus ISO alone.

(e) . Is oproterenol-induced augmentation of late sodium current and trigger activities were attenuated by either late sodium current or Ca²⁺/calmodulin-dependent protein kinase inhibitors

ISO (15 nM) significantly increased late I_Na in isolated atrial myocytes by 87%, from an amplitude of 0.57 ± 0.01 to 1.04 ± 0.02 pA pF⁻¹ (n = 5, p < 0.05; figure 4b). The increases in late I_Na induced by the continued presence of ISO were reversed by RAN (10 µM), ELEC (10 µM) or KN-93 (3 µM) by 49%, 41% and 48%, respectively (n = 5, p < 0.05; figure 4b).

There were no substantial changes in APD after ISO infusion in atrial myocytes (201.7 ± 24.6 ms versus 213.3 ± 30.1 ms, n = 10, p > 0.05), as well as the AP morphology (figure 5a). ISO elicited EADs/DADs with an incidence of 7.8 ± 0.2 in 5 min (n = 5, p < 0.05; figure 5). Co-treatment with RAN (10 µM), ELEC (10 µM) or KN-93 (3 µM) suppressed almost all cellular trigger activities (n = 5, p < 0.05; figure 5b).

Figure 5.

Late I_Na and CaMKII inhibitors reduced early/delay afterdepolarizations (EADs/DADs) in atrial myocytes induced by isoproterenol (ISO; 15 nM). Original records of action potential ((a): EAD (a), DAD (b), n = 10) and summarized data ((b), n = 5) are presented. *p < 0.05 versus control (CTL); **p < 0.05 versus ISO alone.

(f) . The combined effect of late sodium current and Ca²⁺/calmodulin-dependent protein kinase inhibition on isoproterenol-induced atrial fibrillation

When used alone, 0.1 µM RAN, 0.3 µM ELEC or 0.03 µM KN-93 attenuated the ISO-induced AF incidence, inducible window and burden by 8–30% (figure 6). The combination of KN-93 and RAN inhibited the ISO-induced increase in AF incidence, inducible window and burden by 75%, 81.7% and 85.9%, respectively (n = 5, p < 0.05 versus individual effects or their sum; figure 6). A similar result was obtained in hearts treated with the combination of KN-93 and ELEC (n = 5, p < 0.05 versus individual effects or their sum; figure 6). Consistent with the electrophysiological data, the combination of KN-93 and either RAN or ELEC at the mentioned concentrations generated greater inhibitory effects on phospho-Na_V1.5 and phospho-CaMKII protein expression levels and late I_Na than each inhibitor alone or their sum (n = 5; electronic supplementary material, figure S2), suggesting that the inhibitors of late I_Na and CaMKII synergistically act to inhibit ISO-induced AF.

Figure 6.

The combined used of late I_Na and CaMKII inhibitors had synergistic inhibitive effects on isoproterenol (ISO)-induced AF. Combined use of KN-93 (0.03 µM) with either ranolazine (RAN; 0.1 µM) or eleclazine (ELEC; 0.3 µM) were superior to that when each drug was used alone or the sum of the effects of two drugs (either KN-93 and RAN or KN-93 and ELEC, n = 5) in reducing the incidence, inducible window and burden of AF in the continued presence of ISO. *p < 0.05 versus individual effects or the sum (not shown) of the effects of two agents when they were used individually.

4. Discussion

The main findings of this study are (i) β-adrenergic activation caused AF and abnormal atrial electrical activity, and late I_Na enhancer ATX-II promoted proarrhythmic activities of ISO in atria; (ii) the underlying mechanisms of ISO-induced AF may be attributed, at least in part, to augmentation of late I_Na by phosphorylation of Na_V1.5 and CaMKII-δ, and (iii) low concentrations of a late I_Na inhibitor (either RAN or ELEC) with a CaMKII inhibitor, KN-93, synergistically exerted anti-arrhythmic effects.

In Langendorff-perfused animal hearts [17] and patients with paroxysmal AF [18], AF can be induced more often via programmed stimulation after administration of ISO. The proarrhythmic effects of ISO are principally mediated by activation of a β-adrenergic receptor, resulting in increased I_Ca,L and elevation of [Ca²⁺]_i [19]. ISO can also activate CaMKII and PKA to enhance late I_Na through proteins activated by cAMP-dependent pathways directly and through increasing [Ca²⁺]_i indirectly [20]. The positive feedback circle of [Ca²⁺]_i-CaMKII-late I_Na-[Na⁺]_i could progressively facilitate the genesis of DADs via activation of a sodium-calcium exchanger, leading to transient inward currents (I_ti) and EADs via enhancement of I_Ca,L [2,21], thus promoting atrial hypertrophy, dilatation and AF progression [22]. In addition, ISO shortened aERP in Langendorff-perfused rat hearts, thereby increasing the availability of reactivation, the inhomogeneity of conduction and the origination of re-entry circuits [23]. Taken together, in line with our data, the underlying arrhythmogenic mechanisms of ISO are involved in focal ectopic activity and reentry, and CaMKII plays a crucial role in the initiation of β-adrenergic stimulation-induced AF and might also contribute to the maintenance of AF [24].

(a) . Late sodium current and atrial fibrillation

Both endogenous and enhanced late I_Na has been suggested as an important modulator of arrhythmias in patients with a variety of conditions [25]. We found that ISO significantly augmented late I_Na by twofold and promoted AF. These effects were attenuated by late I_Na inhibitors. Thus, the enhancement of late I_Na contributed to the arrhythmogenesis induced by ISO. ATX-II (2 nM) increased the baseline level of late I_Na and thereby reduced repolarization reserve but was insufficient to induce AF. Results in the present study are in agreement with the findings of previously reported studies [13,26,27] and extend these results by revealing that ATX-II pretreatment accentuated the propensities of ISO-induced AF to a greater extent than that observed with ISO alone. These results indicate that the integrative effect of ATX-II and ISO on late I_Na caused AF, suggesting that enhancement of late I_Na was a potential proarrhythmic factor. When the repolarization reserve was reduced (e.g. increases in late I_Na), ISO increased the occurrence of arrhythmias.

Increased late I_Na potentially promotes arrhythmogenesis via both triggered and re-entrant mechanisms [28]. The increases of inhomogeneity of conduction, abnormal cellular trigger activities and AF caused by ISO were attenuated or completely abolished by RAN and ELEC at concentrations that late I_Na was selectively inhibited. The incomplete reversal by agents that inhibit late I_Na may be owing to the involvement of other ion currents (e.g. peak I_Na, I_Kr and I_K1 and L-type calcium channels were affected by ISO directly) in hearts treated with ISO [29].

(b) . Ca²⁺/calmodulin-dependent protein kinase and late sodium current

The proarrhythmic mechanism underlying AF entails increased late I_Na, which probably contributes to triggered AF by increasing [Ca²⁺]_i secondary to the increase in intracellular sodium concentration [30], especially in some animal models of AF and in patients with chronic AF [8,31]. The interaction between late I_Na and CaMKII has been demonstrated in calcium-related ventricular arrhythmias [13] and ATX-II-induced AF [27].

In this study, phospho-Na_V1.5 and phospho-CaMKII protein expression levels were increased by ISO. CaMKII overactivity increased late I_Na by phosphorylating Na_v1.5 in rabbit ventricular myocytes [13]. On the other hand, increased late I_Na disrupts intracellular Ca²⁺ handling [14] which consequently increases the activity of CaMKII [27,28]. Furthermore, an increase in late I_Na has been shown to elevate intracellular Na⁺ and Ca²⁺, which can be attenuated with late I_Na inhibitors [32]. Consistent with these previous findings, our study demonstrated that increased levels of the phospho-Na_V1.5 and phospho-CaMKII proteins were significantly reduced by late I_Na inhibitors. Our results suggest that the activation of CaMKII induced by ISO resulted in the increase in late I_Na, and consequently promoted the proarrhythmic effect of CaMKII in hearts perfused by ISO.

(c) . Therapeutic potential with inhibition of both Ca²⁺/calmodulin-dependent protein kinase and late sodium current

Effective treatment of AF remains a challenge in this field. Increasing evidence has shown that late I_Na inhibition may be a promising, anti-arrhythmogenic, alternative strategy to improve AF with CaMKII hyperactivation [33]. A promiscuous nature with a large number of off-target effects on ion channels and β-blocker activity may limit anti-arrhythmic potential of RAN in atria [34]. ELEC, with minimal effects on other cardiac ion channels, confers protection against ischemia-induced AF during adrenergic stimulation without negative inotropic effects [6].

Previous studies have indicated the existence of a proarrhythmic, synergistic relationship between late I_Na and CaMKII [12]. In the present study, RAN or ELEC and KN-93 at low concentrations alone modestly inhibited ISO-induced AF in isolated hearts. The key finding was that a combination of KN-93 with either ELEC or RAN is superior to the individual effects of each inhibitor on the suppression of arrhythmic activities, which also suggests an interaction between ISO-enhanced CaMKII and late I_Na. Enhancement of late I_Na by ATX-II to induce AF was previously reported by Liang et al. in rat atrial preparations [27]. The results in this study confirmed that β-adrenergic activations to induce AF were associated with a synergistic effect between increased late I_Na and the activation of CaMKII with possible underlying mechanisms of a vicious circle of the β-receptor activation-CaMKII-late I_Na pathway. Activation of CaMKII lead to the phosphorylation of Na_v1.5 and enhancement of late I_Na, which forms the positive feedback cycle between activated CaMKII and enhanced late I_Na with the presence of a relatively low concentration of ATX-II to induce AF. CaMKII lead to the phosphorylation of Na_v1.5 and enhancement of late I_Na, which forms the positive feedback cycle between activated CaMKII and enhanced late I_Na to induce AF. Additionally, β-adrenergic stimulation may augment late I_Na via other mechanisms independent of CaMKII [29]. In this scheme, both CaMKII and late I_Na were activated by β-adrenergic stimulation, and augmented late I_Na was also mediated by activation of CaMKII, providing a mechanism for the combined effect of inhibitors of CaMKII and late I_Na on AF prevention. Using a combination of CaMKII and late I_Na inhibitors at lower dosages to achieve the same outcomes as those provided by mono-therapy can potentially reduce the risk of unexpected drug side effects. This finding may provide a new therapeutic strategy for β-adrenergic stimulation-related AF.

5. Limitations

The ex vivo effects of ISO may be different from the activation of the sympathetic nervous system in AF in vivo. ISO may cause phosphorylation and/or activation of multiple downstream targets which were not investigated individually. Abnormal intracellular calcium and sodium concentrations caused by either ISO or late I_Na enhancement were not measured in this study.

6. Conclusion

Enhancement of late I_Na represents an important factor in mediating β-adrenergic stimulation-mediated AF, which may be associated with the phosphorylation of CaMKII and Na_V1.5. Inhibitions of CaMKII-late I_Na are effective in synergistic mode in suppressing AF associated with catecholaminergic activation.

Ethics

This investigation conformed to the ‘Guide for the Care and Use of Laboratory Animals' (National Institutes of Health publication no. 85-23, revised 2011), and the animal use protocol was approved by the Institutional Animal Care and Use Committee (IACUC) of Peking University First Hospital (J201428).

Data accessibility

The data are available from the Dryad Digital Repository: https://doi.org/10.5061/dryad.1ns1rn8xn [35].

Data are also provided in the electronic supplementary material [36].

Authors' contributions

X.L.: conceptualization, formal analysis, investigation, methodology, resources, visualization, writing—original draft, writing—review and editing; L.R.: data curation, investigation, validation and visualization; S.Y.: data curation, investigation, resources, validation and visualization; G.L.: data curation, investigation, resources, validation and visualization; P.H.: formal analysis, investigation, resources and supervision; Q.Y.: data curation, investigation, resources, validation and visualization; X.W.: formal analysis, investigation, resources and supervision; P.N.T.: writing—original draft, writing—review and editing; L.W.: conceptualization, funding acquisition, investigation, methodology, project administration, resources, supervision, visualization, writing— review and editing; Y.H.: conceptualization, funding acquisition, investigation, methodology, project administration, resources, supervision, visualization, writing—review and editing.

All authors gave final approval for publication and agreed to be held accountable for the work performed therein.

Conflict of interest declaration

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Funding

This work was supported by the National Natural Science Foundation of China (grant nos. 81270253, 81641015 and 81770325) and the Natural Science Foundation of Beijing (grant no. 7132214).