If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Experimental Cardiology, Amsterdam UMC, Meibergdreef 15, 1105 AZ Amsterdam, The NetherlandsDepartment of Medical Biology, Amsterdam UMC, Meibergdreef 15, 1105 AZ Amsterdam, The Netherlands
Corresponding authors: Carol Ann Remme, MD, PhD, Simona Casini, PhD, Department of Experimental Cardiology, Academic Medical Center, Room K2-104.2; Tel:+31-20-5668544, Meibergdreef 9, 1105 AZ Amsterdam, The Netherlands. ;
Affiliations
Department of Experimental Cardiology, Amsterdam UMC, Meibergdreef 15, 1105 AZ Amsterdam, The Netherlands
The cellular mechanisms underlying progression from paroxysmal to persistent atrial fibrillation (AF) are not fully understood, but alterations in (late) sodium current (INa) have been proposed. Human studies investigating electrophysiological changes at the paroxysmal stage of AF are sparse, with the majority employing right atrial appendage cardiomyocytes (CMs). We here investigated action potential (AP) characteristics and (late) INa remodeling in left atrial appendage CMs (LAA-CMs) from patients with paroxysmal and persistent AF and patients in sinus rhythm (SR), as well as the potential contribution of the “neuronal” sodium channel SCN10A/NaV1.8.
Methods
Peak INa, late INa and AP properties were investigated through patch-clamp analysis on single LAA-CMs while qPCR was used to assess SCN5A/SCN10A expression levels in LAA tissue.
Results
In paroxysmal and persistent AF LAA-CMs, AP duration was shorter than in SR LAA-CMs. Compared to SR, peak INa and SCN5A expression were significantly decreased in paroxysmal AF, while they were restored to SR levels in persistent AF. Conversely, while late INa was unchanged in paroxysmal AF as compared to SR, it was significantly increased in persistent AF. Peak or late Nav1.8-based INa was not detected in persistent AF LAA-CMs. Similarly, expression of SCN10A was not observed in LAAs at any stage.
Conclusions
Our findings demonstrate differences in (late) INa remodeling in LAA-CMs from patients with paroxysmal versus persistent AF, indicating distinct cellular pro-arrhythmic mechanisms in different AF forms. These observations are of particular relevance when considering potential pharmacological approaches targeting (late) INa in AF.
. AF can be classified as paroxysmal when episodes terminate within seven days; persistent if the arrhythmia continues for more than 7 days; and long-standing persistent when AF persists for more than one year. It is defined as permanent when interventions aimed to restore sinus rhythm are no longer pursued
Hindricks G, Potpara T, Dagres N, et al. 2020 ESC Guidelines for the diagnosis and management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS): The Task Force for the diagnosis and management of atrial fibrillation of the European Society of Cardiology (ESC) Developed with the special contribution of the European Heart Rhythm Association (EHRA) of the ESC. Eur Heart J. 2021;42:373-498.
. AF typically starts as paroxysmal, progressing over time to the persistent and permanent forms. Since therapeutic approaches aimed at restoring sinus rhythm are generally more successful in patients with paroxysmal AF
, prevention of AF progression to more advanced stages is essential. Hence, a better understanding of the mechanisms underlying AF and its development is needed to optimize current therapies and establish novel therapeutic strategies.
Electrical remodeling in the setting of AF generally includes reduction of action potential (AP) duration due to down-regulation of L-type calcium (Ca2+) current and up-regulation of potassium (K+) currents, in addition to altered Ca2+ homeostasis
. Whether and to what extent sodium current (INa) function is also altered in the setting of AF remains unclear. Reduced peak INa may lead to conduction slowing, whereas increased late INa is associated with AP duration (APD) prolongation and dysregulation of intracellular sodium and Ca2+ homeostasis
. The functional relevance of alterations in peak and/or late INa during the various stages of AF has not been fully resolved. In isolated cardiomyocytes (CMs) from right atrial appendages (RAAs) of patients with “chronic”, i.e. (long-standing) persistent or permanent AF, late INa was shown to be increased
. However, characteristics of peak and/or late INa in CMs from patients with paroxysmal AF have not been previously investigated. Moreover, the functional relevance of sodium channel isoforms other than Nav1.5, such as the “neuronal” isoform SCN10A/NaV1.8, during the various stages of AF progression is as yet not fully elucidated. Surprisingly, there are also no studies investigating AP properties, late or peak INa in CMs from left atrial appendage (LAA) of AF patients, despite the fact that AF is increasingly considered a predominantly left atrial disease
. Indeed, rapidly firing foci initiating paroxysmal AF arise most commonly from the left atrial myocardial sleeves, which extend into the pulmonary veins
. In AF patients, gene expression patterns differ significantly between LAA and RAA, indicating that distinct mechanisms may underlie the development and progression of AF within the left and right atria
. Hence, studies performed in CMs isolated from human LAAs are highly relevant for understanding the pathophysiology of AF. We here therefore investigated AP, peak and late INa properties in LAA CMs isolated from patients with paroxysmal and persistent AF and patients in sinus rhythm (SR) and assessed the potential contribution of NaV1.8.
Methods
Study population
LAAs were obtained from patients in SR without a history of AF undergoing cardiac surgery (coronary bypass grafting or valve surgery)
Thoracoscopic video-assisted pulmonary vein antrum isolation, ganglionated plexus ablation, and periprocedural confirmation of ablation lesions: first results of a hybrid surgical-electrophysiological approach for atrial fibrillation.
Circulation. Arrhythmia and electrophysiology.2011; 4: 262-270
(for details, see Data Supplement). All patients gave written informed consent. Criteria of AF subtype used were in accordance with the 2020 ESC guidelines
Hindricks G, Potpara T, Dagres N, et al. 2020 ESC Guidelines for the diagnosis and management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS): The Task Force for the diagnosis and management of atrial fibrillation of the European Society of Cardiology (ESC) Developed with the special contribution of the European Heart Rhythm Association (EHRA) of the ESC. Eur Heart J. 2021;42:373-498.
LAAs were excised in the operation theatre and transported to the laboratory on ice. Single cells were obtained by an enzymatic isolation protocol as described in Casini et al.
Peak INa, late INa and APs were measured with the ruptured and perforated patch-clamp technique, respectively (for details, see Data Supplement). Details on number of patients and CMs used for each experiment including sex distribution are listed in Supplemental Tables S1 and S2.
Real time quantitative polymerase chain reaction (qPCR)
qPCR analysis was performed in the same LAA samples used for peak INa measurements, except for one paroxysmal AF patient, for whom not enough cDNA was obtained. For details, see Data Supplement.
Statistical analysis
Values are shown as mean±S.E.M. For details, see Data Supplement and relevant table and figure legends.
Results
Patient characteristics
Experiments were performed in LAA tissue and cardiomyocytes from 18 SR, 12 paroxysmal and 30 persistent AF patients. Patient characteristics are shown in Table 1. Paroxysmal AF patients were on average younger than patients in SR and persistent AF. Persistent AF patients generally displayed a higher heart rate, as well as a larger indexed LA volume. Moreover, the prevalence of vascular diseases and diabetes mellitus was higher in SR patients. Medication also differed between groups; patients in SR more often used Ca2+ channel blockers and antiplatelets, while flecainide use was more frequent in AF patients.
Table 1Patient characteristics
General Characteristics
Sinus Rhythm (N=18)
Paroxysmal AF (N=12)
Persistent AF (N=30)
P-value
Gender (m/f)
17/1
10/2
21/9
NS Contingency
Age (years)
66.2±1.7
56.3±2.7&
62.1±1.3
<0.01 One-way ANOVA
BMI (kg/m2)
27.6±0.1
27.3±0.1
27.8±0.1
NS One-way ANOVA
Heart rate (BPM)
70.1±6.3
61.7±4.8
81.2±4.7*
<0.01 Kruskal-Wallis
QTc interval (ms)
430.8±7.1
431.8±6.4
436.7±4.1
NS One-way ANOVA
LV ejection fraction (%)
51.7±2.6
56.7±1.8
49.2±2.5
NS One-way ANOVA
LA volume index (ml/m2)
30.7±2.3
33.9±3.2
44.1±3.1&
<0.01 One-way ANOVA
Hypertension (%)
12 (66.7)
7 (58.3)
15 (50.0)
NS Fisher-Freeman-Halton
Congestive HD (%)
0 (0)
1 (8.3)
4 (13.3)
NS Fisher-Freeman-Halton
Valve dysfunction (%)
6 (33.3)
2 (16.7)
8 (26.7)
NS Fisher-Freeman-Halton
Vascular diseases (%)
18 (100.0)
2 (16.7)
1 (3.3)
<0.0001 Fisher-Freeman-Halton
Diabetes mellitus (%)
4 (22.2)
2 (16.7)
0 (0)
<0.05 Fisher-Freeman-Halton
Medication
Digitalis (%)
0 (0)
1 (8.3)
7 (23.3)
NS Fisher-Freeman-Halton
Flecainide (%)
0 (0)
5 (41.7)
9 (30.0)
<0.01 Fisher-Freeman-Halton
β-blockers (%)
10 (55.6)
3 (25.0)
15 (50.0)
NS Fisher-Freeman-Halton
ACE inhibitors (%)
6 (33.3)
8 (66.7)
12 (40.0)
NS Fisher-Freeman-Halton
Ca2+channel blockers (%)
9 (50.0)
2 (16.7)
2 (6.7)
<0.01 Fisher-Freeman-Halton
ATII blockers (%)
6 (33.3)
1 (8.3)
4 (13.3)
NS Fisher-Freeman-Halton
Antiplatelets (%)
16 (88.9)
1 (8.3)
0 (0)
<0.0001 Fisher-Freeman-Halton
BMI, body mass index; Congestive HD, congestive heart disease; ACE, angiotensin-converting enzyme; ATII, angiotensin II. &p<0.01 vs SR (one-way ANOVA followed by Tukey test for post-hoc analysis); *p<0.01 vs paroxysmal AF (one-way ANOVA on Ranks (Kruskal-Wallis test) followed by Dunn’s test for post-hoc analysis).
Decreased AP duration in LAA CMs from paroxysmal AF patients
Figure 1A shows typical examples of AP traces recorded at physiological temperature (36ºC) and 1 Hz stimulation in LAA CMs from SR, paroxysmal and persistent AF patients. All measured APs displayed a morphology largely comparable to the so-called “Type A” AP (also named “Type 3”), consisting of a fast phase 1 repolarization without a dome, and a plateau at around -20 mV17. No significant differences in maximal AP upstroke velocity (Vmax) or AP amplitude (APA) were observed among groups (Figure 1B, Supplemental Table S3). Compared to SR, paroxysmal and persistent AF CMs showed a significantly more negative (hyperpolarized) resting membrane potential (RMP). Paroxysmal AF CMs displayed significantly shorter APD90, whereas APD50 and APD20 remained unchanged (Figure 1B, Supplemental Table S3). In persistent AF, APD20 was significantly prolonged compared to SR, while APD90 reduction was no longer statistically different from SR. Since sex distribution differed between groups (Supplemental Table S1), we assessed whether this may have impacted on the AP results. However, similar differences were observed between SR, paroxysmal and persistent AF when comparing AP parameters using CMs from male patients only (Supplemental Figure S1).
Figure 1Action potential (AP) properties in human left atrial appendage (LAA) cardiomyocytes (CMs) isolated from sinus rhythm (SR), paroxysmal and persistent AF patients. A, Typical AP recordings in CMs stimulated at 1 Hz. B, Average AP parameters. Vmax, maximal upstroke velocity; APA, AP amplitude; RMP, resting membrane potential; APD20, APD50 and APD90, AP duration at 20, 50, and 90% repolarization, respectively; n, number of CMs; N, number of patients. &p<0.05 paroxysmal AF vs SR, #p<0.05 persistent AF vs SR (one-way ANOVA followed by Tukey test for post-hoc analysis or one-way ANOVA on Rank (Kruskal-Wallis test) followed by Dunn’s test when data were not normally distributed).
Increased late INa in LAA CMs from persistent but not paroxysmal AF patients
To measure late INa in LAA CMs, we used a descending ramp protocol (see inset Figure 2A), which (in contrast to a single-step protocol) allows measurement of late INa across a dynamic voltage range simulating a plateau and repolarization phase of a cardiac AP
. Figure 2A shows typical examples of late INa recordings in CMs isolated from SR, paroxysmal and persistent AF patients. Late INa, measured as tetrodotoxin (TTX)-sensitive current, was obtained by subtraction of the current recorded in the presence of TTX from the current recorded in the absence of the drug (baseline). Late INa was undetectable in the majority of CMs from paroxysmal AF and SR patients and consequently no differences in average late INa density were observed between these two groups (Figure 2B). In contrast, late INa magnitude was significantly increased in persistent AF (at -10 mV, -0.20±0.04 pA/pF, n=25) as compared to SR (at -10 mV, -0.04±0.02 pA/pF, n=21) and paroxysmal AF (at -10 mV, 0.01±0.05 pA/pF, n=13, Figure 2B).
Figure 2Late sodium current (INa) is increased in LAA CMs from persistent AF patients. A, Representative late INa traces measured using a descending ramp protocol (top panel) in LAA CMs. Late INa was measured as TTX-sensitive current (bottom panels) obtained by subtraction of the current recorded in the presence of TTX (30 μM) from the current recorded earlier in the absence of the compound (baseline; middle panel). B, Average current-voltage (I-V) relationships for late INa measured as TTX-sensitive current. *p<0.05 persistent AF vs paroxysmal AF; #p<0.05 persistent AF vs SR (two-way repeated measures ANOVA followed by Holm-Sidak test for post-hoc analysis).
Decreased peak INa in LAA CMs from paroxysmal but not persistent AF patients
In addition to late INa, we also investigated peak INa density and its voltage dependency of activation and inactivation. Figure 3A shows representative peak INa recordings obtained in LAA CMs isolated from patients in SR, paroxysmal AF, and persistent AF. Compared to SR, average peak INa density was significantly reduced in paroxysmal AF (Figure 3B, Supplemental Table S4). However, peak INa in persistent AF was significantly increased compared to paroxysmal AF, returning to the same density as observed in SR. INa voltage dependence of activation and inactivation, assessed as the half voltage of (in)activation (V1/2) and the slope factor k, were unchanged in the three groups, indicating that INa gating properties remained unaffected during progression from SR to paroxysmal and persistent AF (Figure 3C, Supplemental Table S4). Although sex distribution differed between groups (Supplemental Table S1), similar results were observed for peak and late INa in SR, paroxysmal, and persistent AF using CMs isolated from male patients only (Supplemental Figure S2), indicating that sex differences did not impact on our findings.
Figure 3Peak INa is decreased in LAA CMs from paroxysmal AF patients. A, Representative peak INa traces recorded in LAA CMs. B-C, Average current-voltage (I-V) relationships (B) and voltage dependence of activation and inactivation (C). Dotted lines indicate Boltzmann fits through the average data. Insets: voltage protocols. &p<0.05 paroxysmal AF vs SR; *p<0.05 paroxysmal AF vs persistent AF (two-way repeated measures ANOVA followed by Holm-Sidak test for post-hoc analysis).
, we investigated the effect of the Nav1.8 channel blocker A-803467 on AP properties in persistent AF CMs. Figure 4A shows typical AP recordings obtained from LAA CMs at baseline conditions, in the presence of 100 nM A-803467 and after wash-out of the drug. On average, Vmax, APA, RMP, APD20 and APD50 were not affected by A-803467 exposure, while the compound induced a significant although small (8%) reduction in APD90 (Figure 4B, Supplemental Table S5).
Figure 4Effect of the NaV1.8 blocker A-803467 on AP properties in LAA CMs from persistent AF patients. A, Examples of AP traces recorded at 1 Hz under physiological conditions (baseline), after 5 minutes wash-in of 100 nM A-803467 and 5 minutes wash-out of the drug. B, Average data for Vmax, APA, RMP, APD20, APD50 and APD90 before (baseline) and after wash-in and wash-out of A-803467. *p<0.05 A-803467 vs baseline, $p<0.05 A-803467 vs wash-out (one-way repeated measures ANOVA followed by Holm-Sidak test for post-hoc analysis or one-way repeated measures ANOVA on Ranks (Friedman test) when data were not normally distributed).
We next investigated whether Nav1.8-dependent late INa contributes to the increased late INa observed in persistent AF. Late INa measurements were performed in each cell at baseline, after perfusion with A-803467 (100 nM), and following application of 30 μM TTX to block all NaV isoforms. Nav1.8-dependent late INa, measured as A-803467-sensitive current, was obtained by subtraction of the current recorded in the presence of A-803467 from the current in the absence of the compound (baseline; Figure 5A). Similarly, in the same cell, TTX-sensitive current (total late INa), was measured by subtraction of the current recorded in the presence of TTX from the baseline current. Average TTX-sensitive current was around -0.2 pA/pF while the A-803467-sensitive current was undetectable (Figure 5B). We also assessed the effects of A-803467 on peak INa density and voltage dependency of activation and inactivation in persistent AF CMs. Average INa density was unchanged after exposure to A-803467 (Figure 6A-B). A-803467 caused a small yet significant negative shift in V1/2 of activation and inactivation (-2.8 mV for activation and -2.0 mV for the inactivation curve; Figure 6C-D, Supplemental Table S6). However, wash-out of A-803467 failed to reverse these effects, and a further negative shift of V1/2 of (in)activation was in fact observed (Supplemental Figure S3), indicating a time-dependent rather than a A-803467-dependent effect on INa gating properties. Taken together, these experiments demonstrate the absence of functional NaV1.8-based peak and late INa in LAA CMs from persistent AF patients.
Figure 5NaV1.8 does not contribute to late INa in persistent AF CMs. A, Representative late INa traces recorded during a ramp protocol (see inset) in LAA CMs isolated from persistent AF patients. Nav1.8-based late INa and total late INa were obtained by subtraction of the current recorded in the presence of A-803467 (A-803467-sensitive current) and tetrodotoxin (TTX-sensitive current) from the current recorded earlier in the absence of the drugs (baseline). B, Average current-voltage relationship of NaV1.8-based late INa and total late INa measured as A-803467-and TTX-sensitive currents, respectively.
Figure 6Nav1.8 does not contribute to peak INa in persistent AF CMs. A, Representative peak INa traces recorded from LAA CMs before (baseline) and after 5 minutes wash-in of 100 nM A-803467. B-D, Average current-voltage relationships (B), voltage dependence of activation (C) and voltage dependence of inactivation (D) at baseline and in the presence of A-803467. Dotted lines indicate Boltzmann fits through the average data. Insets: voltage protocols.
SCN5A mRNA transcript levels are decreased in paroxysmal AF LAAs
We finally explored whether the observed changes in INa were associated with alterations in mRNA expression levels of SCN5A and SCN10A, encoding Nav1.5 and Nav1.8, respectively.
In line with our functional INa results (Figure 3B), qPCR analysis of LAA samples showed that SCN5A expression was significantly lower (48%) in paroxysmal AF compared to SR (Figure 7A), while in persistent AF, SCN5A levels were similar to SR values. Moreover, in accordance with the absence of functional NaV1.8-based peak and late INa in persistent AF CMs (Figures 5B, 6B), we could not detect the SCN10A full transcript in LAA samples. We also investigated expression levels of the recently described short SCN10A transcript (SCN10A-short) comprising the last 7 exons of the gene, which in itself does not encode for a functional conducting channel, but has been shown to increase NaV1.5 activity
. While there was a tendency towards a lower expression of SCN10A-short in paroxysmal AF (18% reduction versus SR) and persistent AF (9% reduction versus SR), these differences were not significant (Figure 7B).
Figure 7SCN5A expression is reduced in paroxysmal AF. A-B, SCN5A (A) and SCN10A-short (B) mRNA expression in LAAs isolated from SR, paroxysmal and persistent AF patients. Gene expression was normalized to the reference gene β-Glucuronidase (GUSB). N, indicates the number of patients. SCN10A-short RNA levels were not detectable in three persistent AF patients and were not included. &p<0.05 paroxysmal AF vs SR (one-way ANOVA on Ranks (Kruskal-Wallis test) followed by Dunn’s test for post-hoc analysis).
We here investigated for the first time AP and (late) INa characteristics in LAA CMs from patients with paroxysmal and persistent AF. Our results show that AP shortening already occurs in LAA CMs from paroxysmal AF, potentially contributing to pro-arrhythmia in this early stage of the disease. Moreover, peak INa is decreased in patients with paroxysmal AF as compared to those in SR, while late INa remained unchanged during this stage. In contrast, in persistent AF peak INa returned to SR values, whereas late INa significantly increased. We furthermore demonstrate that Nav1.8-based current does not contribute to these AF-related alterations in peak or late INa. Crucially, our measurements were for the first time performed in CMs isolated from LAAs. Our study protocols and surgical and thoracoscopic procedures only allowed for the collection of LAA tissue, precluding a direct comparison between RAA and LAA. Nevertheless, AF is increasingly considered a predominantly left atrial disease
, and hence our findings in LAA contribute significantly to the understanding of AF progression. As with all human studies, we cannot rule out the possibility that the observed electrical remodeling may have been influenced by pharmacological treatment and/or underlying cardiac disease. Nevertheless, our findings demonstrate distinct remodeling of (late) INa during the progression of AF, which may have significant consequences for treatment strategies aimed at different stages of the disease.
AP changes in paroxysmal versus persistent AF
Our results obtained in LAA CMs from persistent AF patients (APD90 reduction, APD20 prolongation, and more hyperpolarized RMP, as compared to SR) are in line with previous results in single CMs and trabeculae isolated from RAAs of persistent/permanent AF patients
Molecular basis of downregulation of G-protein-coupled inward rectifying K(+) current (I(K,ACh) in chronic human atrial fibrillation: decrease in GIRK4 mRNA correlates with reduced I(K,ACh) and muscarinic receptor-mediated shortening of action potentials.
. In contrast, our finding that APD90 shortening and more negative RMP already occurred in LAA CMs from paroxysmal AF patients differs from previous studies performed in human RAAs, where AP parameters remained unchanged during this stage
. Clearly, the early APD shortening is of functional relevance, since it is associated with a decrease in refractoriness which may increase the likelihood of re-entrant arrhythmias. The observed changes in APD and RMP are likely explained by alterations in Ca2+ or K+ currents, as previously demonstrated in various studies by other groups. A more hyperpolarized RMP is a well-known finding in RAAs from patients with advanced forms of AF
Molecular basis of downregulation of G-protein-coupled inward rectifying K(+) current (I(K,ACh) in chronic human atrial fibrillation: decrease in GIRK4 mRNA correlates with reduced I(K,ACh) and muscarinic receptor-mediated shortening of action potentials.
Molecular basis of downregulation of G-protein-coupled inward rectifying K(+) current (I(K,ACh) in chronic human atrial fibrillation: decrease in GIRK4 mRNA correlates with reduced I(K,ACh) and muscarinic receptor-mediated shortening of action potentials.
Caballero R, de la Fuente MG, Gomez R, et al. In humans, chronic atrial fibrillation decreases the transient outward current and ultrarapid component of the delayed rectifier current differentially on each atria and increases the slow component of the delayed rectifier current in both. J Am Coll Cardiol. 2010;55:2346-2354.
furthermore observed a reduced transient outward K+ current (Ito) in LAA CMs from chronic AF patients, which may underlie the observed APD20 prolongation. Moreover, downregulation of L-type Ca2+ current contributes to APD shortening in AF, in addition to upregulation of various K+ currents
Caballero R, de la Fuente MG, Gomez R, et al. In humans, chronic atrial fibrillation decreases the transient outward current and ultrarapid component of the delayed rectifier current differentially on each atria and increases the slow component of the delayed rectifier current in both. J Am Coll Cardiol. 2010;55:2346-2354.
Peak INa remodeling in paroxysmal versus persistent AF
The current study is the first to investigate INa characteristics in CMs from paroxysmal AF patients, showing a reduction in peak INa density during this early disease stage. Loss of peak INa is associated with conduction slowing, a well-established risk factor for arrhythmias. However, despite the decrease in peak INa, we did not observe changes in Vmax and APA likely because the AF-induced RMP hyperpolarization resulted in increased sodium channel functional availability for AP upstroke generation
, which counteracts the reduction in peak INa. Hence, the decreased INa density would only result in conduction slowing if sodium channel availability is not affected. Reduced conduction velocity was previously observed in left and right atria from paroxysmal AF patients
, the contribution of reduced INa to conduction slowing in paroxysmal AF remains as yet unclear.
Two studies have previously examined peak INa in CMs isolated from patients in permanent/persistent AF, but conflicting results have been reported. Similar to us, Bosch et al. found unchanged INa density in RAA CMs from patients with chronic AF
. On the one hand, discrepancies in peak INa results among studies could be due to intrinsic heterogeneities between LAA and RAA. On the other hand, the same clinical status, e.g. permanent AF, can include patients with different durations of AF. For instance, Bosch et al.
enrolled patients with chronic AF, defined as continuous AF of at least 1 month duration, while no time cut-off was given for the permanent AF patients enrolled in the study of Sosalla and colleagues
. This heterogeneity might underlie differences in electrical remodeling, that in turn could explain the discrepancies in peak INa and/or SCN5A/NaV1.5 expression among studies.
Our qPCR results showed a significant decrease in SCN5A expression in paroxysmal AF LAAs that may underlie, at least in part, the decreased INa observed at this stage. Interestingly, this reduced INa was no longer evident in persistent AF, and both INa density and SCN5A levels in this more advanced stage had seemingly returned to SR values. The exact mechanism underlying the apparent restoration of SCN5A expression and consequently peak INa in persistent AF remains unknown. While we cannot exclude that baseline clinical differences between the groups played a role, one could speculate that this restoration may reflect a compensatory effect to counterbalance the atrial conduction slowing occurring secondary to alterations in e.g. structural components and connexins. Besides the increase in SCN5A expression, post-translational modification of NaV1.5 could also contribute.
Increased late INa in persistent but not paroxysmal AF
While the arrhythmogenic role of increased late INa in acquired diseases, such as heart failure and myocardial ischemia, has been extensively investigated
, only few studies have assessed the role of late INa in AF. In experiments conducted at room temperature, enhanced late INa was shown in RAA myocytes from chronic AF patients
. Moreover, no studies have so far investigated late INa in CMs from paroxysmal AF patients. Hence, our current experiments are the first to show late INa measured at physiological temperature in LAA CMs, and clearly demonstrate that late INa is increased in the later stages of AF (i.e. persistent AF), but not in paroxysmal AF. Increased late INa is known to prolong APD, potentially evoking early afterdepolarizations (EADs) as triggers for arrhythmias, but the contribution of EADs to arrhythmogenesis in AF patients in the absence of repolarization-delaying mutations remains questionable
. Enhanced late INa furthermore leads to increased [Na+]i, which can cause Na+/Ca2+ exchanger (NCX)-mediated Ca2+ overload. Increased diastolic Ca2+ levels may in turn promote spontaneous release from the sarcoplasmic reticulum leading to delayed afterdepolarizations (DADs), which serve as a trigger for irregular APs and focal arrhythmias
. Ultimately, a complex interaction between various factors will determine arrhythmogenesis and AF susceptibility, which is further complicated by the common presence of structural and metabolic remodeling.
Our finding that late INa is specifically increased in persistent AF may be of relevance when considering potential therapeutic strategies. Indeed, several clinical studies have shown anti-AF properties for the late INa blocker ranolazine
, while the more selective and potent late INa inhibitors GS-458967 (GS967) and eleclazine (GS-6615) have shown to reduce AF vulnerability and suppress arrhythmias episodes in porcine models of autonomically and ischemia induced AF
Eleclazine, a new selective cardiac late sodium current inhibitor, confers concurrent protection against autonomically induced atrial premature beats, repolarization alternans and heterogeneity, and atrial fibrillation in an intact porcine model.
Inhibition of the cardiac late sodium current with eleclazine protects against ischemia-induced vulnerability to atrial fibrillation and reduces atrial and ventricular repolarization abnormalities in the absence and presence of concurrent adrenergic stimulation.
, the presence of Nav1.8 in atrial CMs and its contribution to AF is still a matter of debate. RNA sequencing of LAA and RAA from patients in permanent AF did not show the presence of SCN10A transcripts
. Also, Poulet et al. reported a very low SCN10A expression in human RA tissue, with the chronic AF samples showing significantly reduced SCN10A levels compared to SR, making the contribution of Nav1.8 current to AF unlikely
. In contrast, Pabel et al. detected the presence of SCN10A/Nav1.8 in atrial tissue from SR and AF patients, however both gene and protein expression levels did not differ between SR and AF
. In the current study, we could not detect any Nav1.8-based peak or late INa in LAA CMs from persistent AF patients, in accordance with our previous studies in non-diseased human LAA CMs, human induced pluripotent stem cell (hiPSC) derived CMs, murine and rabbit ventricular CMs
. Despite this absence of NaV1.8-based (late) INa, we still observed a small, yet significant and mostly reversible decrease in APD90 induced by A-803467, indicating an actual impact of the drug rather than a time-dependent effect. This reversible AP shortening is consistent with previous findings in murine CMs
. Recently, the same group has reported the presence of a short SCN10A transcript (SCN10A-short), comprising the last 7 exons, in human and mouse heart tissue. The RE1-deletion mouse model fails to express Scn10a-short and shows right atrial conduction slowing in vivo and reduced INa in isolated atrial cardiomyocytes. Interestingly, this was accompanied by only a slight reduction of right atrial Scn5a expression, indicating a potential direct effect of Scn10a-short on INa by modulating NaV1.5 function. Indeed, increased INa density was observed when Scn10a-short was transfected in HEK293 cells stably expressing Nav1.521. Similar to what has been reported by Man et al.,
qPCR analysis in the present study did not reveal the presence of the full-length SCN10A transcript in LAA tissue from SR or AF patients, while the SCN10A-short isoform was present. Interestingly, our qPCR data showed a 18% decrease of SCN10A-short in paroxysmal AF LAAs but only 9% reduction in persistent AF. While it is tempting to speculate that these changes in SCN10A-short expression contribute to the decrease of INa in paroxysmal AF and its restoration to SR values in persistent AF, additional evidence is required. Finally, it is important to note that a modulatory role for the full-length SCN10A/NaV1.8 in AF could also be due to a functional involvement in intracardiac neurons, since we and others have shown that NaV1.8 is specifically expressed in murine, canine and human cardiac neurons
suppressed AF most likely by inhibiting neuronal GP activity. Hence, while our current observations indicate that Nav1.8 (full length) does not contribute to alterations in (late) INa observed in LAA CMs of persistent AF patients, we cannot rule out a potential modulatory effect of SCN10A/NaV1.8 on AF through its functional involvement in intracardiac neurons. Further studies are needed to elucidate the exact role of SCN10A/NaV1.8 in AF.
Conclusions
In human LAA CMs, AP shortening occurs already at the paroxysmal AF stage, indicating the importance of investigating AF disease progression and mechanisms in left atria. Peak and late INa are differently remodeled during paroxysmal versus persistent AF, which is not explained by changes in Nav1.8-based current. In particular, peak INa reduction occurred already during paroxysmal AF whereas late INa was increased in persistent AF only. Differential (late) INa remodeling during different stages of AF may be associated with distinct pro-arrhythmic mechanisms, which is of potential relevance when considering novel therapeutic approaches for AF.
The authors thank Dr Phil Barnett for expert advice on molecular studies and Cees Schumacher for technical assistance.
Funding Sources: This study was funded by two Innovational Research Incentives Scheme Vidi grants from the Netherlands Organisation for Health Research and Development (ZonMw 91714371 to C.A.R. and ZonMw 016.146.310. to J.R.G); a ZonMw Priority Medicines (PM-Rare) grant (113303006 to C.A.R.); the Netherlands CardioVascular Research Initiative CVON (PREDICT2 CVON2018-30 to C.A.R.).
Hindricks G, Potpara T, Dagres N, et al. 2020 ESC Guidelines for the diagnosis and management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS): The Task Force for the diagnosis and management of atrial fibrillation of the European Society of Cardiology (ESC) Developed with the special contribution of the European Heart Rhythm Association (EHRA) of the ESC. Eur Heart J. 2021;42:373-498.
Thoracoscopic video-assisted pulmonary vein antrum isolation, ganglionated plexus ablation, and periprocedural confirmation of ablation lesions: first results of a hybrid surgical-electrophysiological approach for atrial fibrillation.
Circulation. Arrhythmia and electrophysiology.2011; 4: 262-270
Molecular basis of downregulation of G-protein-coupled inward rectifying K(+) current (I(K,ACh) in chronic human atrial fibrillation: decrease in GIRK4 mRNA correlates with reduced I(K,ACh) and muscarinic receptor-mediated shortening of action potentials.
Caballero R, de la Fuente MG, Gomez R, et al. In humans, chronic atrial fibrillation decreases the transient outward current and ultrarapid component of the delayed rectifier current differentially on each atria and increases the slow component of the delayed rectifier current in both. J Am Coll Cardiol. 2010;55:2346-2354.
Eleclazine, a new selective cardiac late sodium current inhibitor, confers concurrent protection against autonomically induced atrial premature beats, repolarization alternans and heterogeneity, and atrial fibrillation in an intact porcine model.
Inhibition of the cardiac late sodium current with eleclazine protects against ischemia-induced vulnerability to atrial fibrillation and reduces atrial and ventricular repolarization abnormalities in the absence and presence of concurrent adrenergic stimulation.
Since approaches aimed at restoring sinus rhythm are generally more successful in patients with paroxysmal atrial fibrillation (AF), better understanding of the mechanisms underlying progression towards persistent AF is needed. In left atrial appendage cardiomyocytes from AF patients, we observed action potential shortening and peak sodium current (INa) reduction in paroxysmal AF, while late INa was enhanced in persistent AF. These distinct cellular pro-arrhythmic mechanisms are of particular relevance when considering therapeutic approaches for different AF stages.
Atrial fibrillation (AF), the clinically most common form of arrhythmia, is usually classified based on its duration: ranging from paroxysmal AF, consisting of self-terminating episodes lasting typically less than 7 days, to persistent and long-lasting persistent or chronic AF, in which AF fails to self-terminate.1 In all the different forms, electrical remodelling occurs. This remodelling further increases ectopic-triggered activity and provides an electrical substrate even more prone to re-entry formation—the 2 main arrhythmogenic mechanisms in AF—thereby facilitating the occurrence and maintenance of AF, as well described in the term “AF begets AF” by the Allessie group in 1995.
01119-9&id=fx1.jpg){kind=link}
01119-9&id=gr1.jpg){kind=link}
01119-9&id=gr2.jpg){kind=link}
01119-9&id=gr3.jpg){kind=link}
01119-9&id=gr4.jpg){kind=link}
01119-9&id=gr5.jpg){kind=link}
01119-9&id=gr6.jpg){kind=link}
01119-9&id=gr7.jpg){kind=link}