Inhibition of N-type calcium channels in cardiac sympathetic neurons attenuates ventricular arrhythmogenesis in heart failure

Abstract

Aims

Cardiac sympathetic overactivation is an important trigger of ventricular arrhythmias in patients with chronic heart failure (CHF). Our previous study demonstrated that N-type calcium (Cav2.2) currents in cardiac sympathetic post-ganglionic (CSP) neurons were increased in CHF. This study investigated the contribution of Cav2.2 channels in cardiac sympathetic overactivation and ventricular arrhythmogenesis in CHF.

Methods and results

Rat CHF was induced by surgical ligation of the left coronary artery. Lentiviral Cav2.2-α shRNA or scrambled shRNA was transfected in vivo into stellate ganglia (SG) in CHF rats. Final experiments were performed at 14 weeks after coronary artery ligation. Real-time polymerase chain reaction and western blot data showed that in vivo transfection of Cav2.2-α shRNA reduced the expression of Cav2.2-α mRNA and protein in the SG in CHF rats. Cav2.2-α shRNA also reduced Cav2.2 currents and cell excitability of CSP neurons and attenuated cardiac sympathetic nerve activities (CSNA) in CHF rats. The power spectral analysis of heart rate variability (HRV) further revealed that transfection of Cav2.2-α shRNA in the SG normalized CHF-caused cardiac sympathetic overactivation in conscious rats. Twenty-four-hour continuous telemetry electrocardiogram recording revealed that this Cav2.2-α shRNA not only decreased incidence and duration of ventricular tachycardia/ventricular fibrillation but also improved CHF-induced heterogeneity of ventricular electrical activity in conscious CHF rats. Cav2.2-α shRNA also decreased susceptibility to ventricular arrhythmias in anaesthetized CHF rats. However, Cav2.2-α shRNA failed to improve CHF-induced cardiac contractile dysfunction. Scrambled shRNA did not affect Cav2.2 currents and cell excitability of CSP neurons, CSNA, HRV, and ventricular arrhythmogenesis in CHF rats.

Conclusions

Overactivation of Cav2.2 channels in CSP neurons contributes to cardiac sympathetic hyperactivation and ventricular arrhythmogenesis in CHF. This suggests that discovering purely selective and potent small-molecule Cav2.2 channel blockers could be a potential therapeutic strategy to decrease fatal ventricular arrhythmias in CHF.

Graphical Abstract

1. Introduction

Chronic heart failure (CHF), the most common type of heart disease, affects approximately 6.5 million people in the USA.¹ Despite numerous clinical therapies, mortality in patients with CHF remains high.² Among various causes of sudden cardiac death (SCD), fatal ventricular arrhythmias account for ∼60% of mortality in patients with CHF,³ and management of ventricular arrhythmias is one of the clinical challenges in these patients. Although various interventions, including antiarrhythmic medications, catheter ablation, and implanted cardioverter-defibrillators are widely used to decrease ventricular arrhythmias in patients with CHF, such treatments can have serious side effects, with a subset of CHF patients being refractory to these therapies. For example, as first-line therapy for catecholaminergic polymorphic ventricular arrhythmia, β-blockers have been shown to have a high success rate. Unfortunately, however, syncope and aborted cardiac arrest continue to occur while patients are on prescribed β-blockers,⁴ and some patients are either intolerant or refractory to this therapy.^4,5

Besides structural and electrophysiological remodelling in the heart,⁶ neuronal remodelling of the autonomic nervous system plays a critical role in the occurrence of ventricular arrhythmias in CHF.^7,8 Acritical contribution of autonomic nervous system to ventricular arrhythmias is recently highlighted in state-of-the-art review by Herring et al.⁹ Since sympathetic hyperactivation is typical in patients with CHF and is an important contributor to ventricular arrhythmia and SCD in these patients,^9,10 selective modulation of cardiac sympathetic activity would be an effective approach for managing recurrent ventricular arrhythmias. Modulation of sympathetic input in a failing heart via cardiac sympathetic denervation (CSD) has again recently been proposed as an alternative in managing refractory ventricular arrhythmias and has been shown substantial clinical utility with a low complication rate.^11,12 Despite the efficacy of CSD was demonstrated in clinic, patients with CSD still showed a high risk of recurrences¹³ and rigorous clinical trials demonstrating the safety of CSD are still lacking.¹⁴ Additionally, CSD-related adverse complications such as Horner syndrome, hyperhidrosis, paraesthesia, differential hand temperatures, facial flush, and sympathetic fight/fight response loss are emerged post-operatively,^15,16 which limit use of the procedure and lead to the de-emphasis of CSD use as a last resort after other interventions fail. Thus, there is an urgent need to develop novel therapies against ventricular arrhythmias and for reducing mortality in CHF patients.

Voltage-dependent calcium channels (VDCCs) play an essential role in a wide variety of cell functions through mediating calcium influx.¹⁷ Among various types of VDCCs, N-type calcium (Cav2.2) channels, predominantly expressed in the nervous system, play an important role in modulating neurotransmitter release at sympathetic nerve terminals.^17,18 Our previous study demonstrated that Cav2.2 currents and the cell excitability of cardiac sympathetic post-ganglionic (CSP) neurons located in stellate ganglia (SG) are increased in coronary artery ligation-induced CHF rats.¹⁹ We also found cardiac sympathetic overactivation¹⁹ and ventricular arrhythmias²⁰ in the same animal CHF model. It has been reported that genetic elimination of CACNA1B, the gene encoding the α1 subunit of Cav2.2 channels, caused functional deterioration of the sympathetic nervous system.¹⁷ However, it remains unclear whether an increase in Cav2.2 currents in CSP neurons contributes to cardiac sympathetic overactivation and ventricular arrhythmogenesis in CHF. In the current study, we tested our hypothesis that normalizing CHF-increased Cav2.2 currents in CSP neurons could reduce cardiac sympathetic activation and decrease ventricular arrhythmias in CHF.

2. Methods

All experimental procedures were approved by the University of Nebraska Medical Center (UNMC) Institutional Animal Care and Use Committee and were carried out in accordance with the National Institutes of Health (NIH Publication No. 85-23, revised 1996). In the present study, urethane and α-chloralose were only used in a non-recovery procedure. Detailed information and procedures in Methods are available in Supplementary material online.

2.1 Study design, timeline, and interventions

In the present study, 133 male Sprague-Dawley rats (6–7 weeks of age) were randomly assigned to sham and CHF groups. At 12 weeks after coronary artery ligation-induced myocardial infarction (MI), CHF rats were further assigned to three subgroups for different treatments, including CHF with saline, CHF transfected with scrambled shRNA, and CHF transfected with Cav2.2-α shRNA. Saline, shRNA scramble control, or lentiviral Cav2.2-α shRNA with green fluorescent protein (GFP) (2 µL) was microinjected into bilateral SGs at 12 weeks of post-MI. Cardiac function was determined by echocardiography before MI (pre-MI), and at 1, 3, 8, 12 (before treatment), and 14 weeks (after treatment) of post-MI. Heart rate variability (HRV) from 24-h electrocardiogram (ECG) telemetry recording was evaluated before and after treatments in all groups. Terminal experiments including measurement of inducibility of ventricular arrhythmia, recording of cardiac sympathetic nerve activity (CSNA), haemodynamic and morphological measurements were performed at 14 weeks of post-MI (Figure 1). After in vivo experiments were performed, rats were euthanized with 0.39 mL/kg of Fatal-Plus euthanasia solution (about 150 mg/kg pentobarbital, i.p.) and then SGs and hearts were removed to perform RT-PCR, western blot, immunofluorescence staining, and patch-clamp recording for N-type calcium currents and action potentials (APs) in CSP neurons.

Figure 1.

Study design, timeline, and interventions. At the beginning of these experiments, rats were randomly assigned to sham or CHF group. CHF rats underwent surgical LAD ligation, and sham underwent the same surgery without LAD ligation. Implantation of radiotelemetry was performed at 11 weeks after LAD ligation. CHF rats were then randomly assigned to one of three subgroups, including CHF, scrambled shRNA+CHF, and Cav2.2-α shRNA+CHF at 12 weeks after LAD ligation. Saline, scrambled shRNA, or Cav2.2-α shRNA was microinjected into SGs at 12 weeks after LAD ligation. A 24-h radiotelemetry ECG recording was performed before and at 9 days after microinjection of SGs to measure ventricular arrhythmias and HRV in conscious rats in all groups. Echocardiography was performed before, and at 1, 3, 8, 12, and 14 weeks after LAD ligation to determine cardiac function. Terminal experiments were performed at 14 weeks after LAD ligation.

2.2 CHF rat model

CHF rats were anaesthetized with 2% isoflurane inhalation for surgical ligation of the left anterior descending coronary artery (LAD), and sham rats underwent the same surgery without LAD ligation. CHF was confirmed by multiple morphological and haemodynamic parameters.

2.3 Microinjection of lentiviral Cav2.2-α shRNA or scrambled shRNA in SGs

Under anaesthetized condition (2% isoflurane) and mechanical ventilation, the rat was kept in lateral recumbent position. Bilateral thoracotomy was performed in the second intercostal spaces. After SGs were identified (Figure 2A), lentiviral Cav2.2-α shRNA with GFP or scrambled shRNA (2 µL, 2 × 10⁷ TU/mL) was microinjected into bilateral SGs by glass micropipette. Nine days after transfection, transfection efficacy of lentiviral Cav2.2-α shRNA was verified by real-time RT-PCR and western blot analysis.

Figure 2.

Transfection efficiency and silencing effect of lentiviral Cav2.2-α shRNA in SGs. (A) Microinjection of SGs under rat left (a) and right (b) posterolateral thoracotomy. LSG and RSG indicate left and right SG, respectively. Scale bar: 2 mm. (B) Expression of GFP in SG neurons after in vivo transfection. Almost all SG neurons were transfected with Cav2.2-α shRNA expressing with GFP. Scale bar: 100 µm. (C) Expression of Cav2.2-α mRNA in SGs, measured by real-time RT-PCR; n = 6 rats per group. (D) Representative images (a) and quantitative data (b) showing the expression of Cav2.2-α protein in SGs, measured by western blot; n = 8 rats per group. Data are means ± SEM. Statistical significance was determined by one-way ANOVA with post hoc Dunnett’s test. *P < 0.05 vs. sham; ^†P < 0.05 vs. CHF.

2.4 ECG telemeter recording and HRV measurement in conscious rats

Implantation of the ECG telemeter was performed as described previously.^21,22 The rat was anaesthetized with 2% isoflurane. A radiotelemetry ECG transmitter was placed into the abdominal cavity. After 1 week of surgery, a 24-h continuous ECG recording in the conscious rat was performed. The number of premature ventricular contractions (PVCs) and cumulative duration of ventricular tachycardia/ventricular fibrillation (VT/VF) were quantified from the ECG recording. HRV parameters, including low frequency power (LF, 0.2–0.75 Hz), high frequency power (HF, 0.75–2.5 Hz), and LF/HF ratio; and ventricular arrhythmogenesis-related ECG markers, including QT and corrected QT (QTc) intervals, QT and QTc dispersions, and T-peak to T-end (Tpe) interval,²³ were also calculated from the ECG recording.

2.5 Measurement of inducibility of ventricular tachyarrhythmia

Under anaesthetized condition (800 mg/kg urethane combined with 40 mg/kg α-chloralose, i.p.) and mechanical ventilation, the heart was exposed and inducibility of ventricular tachyarrhythmia was measured by programmed electrical stimulation (PES), as described previously.²⁴ PES-triggered ventricular tachyarrhythmias was quantified by a quotient of ventricular arrhythmia score.²⁵

2.6 Recording of CSNA

CSNA was recorded as described previously.²⁶ Under anaesthesia (800 mg/kg urethane combined with 40 mg/kg α-chloralose, i.p.), a left thoracotomy was performed to expose the left cardiac sympathetic nerve, and then CSNA was recorded by a bipolar electrode.

2.7 Echocardiography

Rats were imaged by the MX201 transducer on the Vevo 3100 ultrasound machine under anaesthesia (1–1.5% isoflurane). Left ventricular end-diastolic diameter (LVDd) and left ventricular end-systolic diameter (LVDs) were measured. Then ejection fraction (EF), fractional shortening (FS), left ventricular end-diastolic volume (LVd Vol), and left ventricular end-systolic volume (LVs Vol) were calculated using standard formulas from the VisualSonics VevoLab software.

2.8 Labelling and isolation of SG neurons, and patch-clamp recording for VDCC currents and APs

CSP neurons were retrograde-labelled by a fluorescent dye (red colour DiI). Rats were anaesthetized with 2% isoflurane. After the left thoracotomy, 20 injections (2 µL per injection) were made sub-epicardially into the left and right atria and ventricles. One week post-surgery, the neurons were isolated by a two-step enzymatic digestion protocol as described previously.¹⁹ VDCC currents and APs only in DiI-labelled SG neurons (i.e. CSP neurons) were recorded by the whole cell patch-clamp technique.¹⁹

2.9 Real-time RT-PCR, western blot analysis, and immunofluorescence staining

Total mRNA was extracted from SGs. Based on the cDNA sequences of RPL19 (housekeeping gene) and calcium channel subunits (Cav1.2, Cav1.3, Cav2.1, Cav2.2, and Cav2.3), the primers (Supplementary material online, Table S1) were designed. Data were analysed by the 2^−ΔΔCt method. Target protein (Cav2.2-α) and housekeeping protein (β-actin) in SGs were analysed by immunoblotting. Target specificity of Cav2.2-α antibody was verified as described previously.²⁷

2.10 Haemodynamic and morphological measurement

Under anaesthesia (800 mg/kg urethane combined with 40 mg/kg α-chloralose, i.p.), blood pressure and heart rate (HR) were monitored by a cannulation of the left femoral artery, and left ventricular pressure was measured by a Millar pressure transducer. After in vivo experiments were performed, the heart was removed for measurement of infarct size.

2.11 Statistical analysis

All data are presented as means ± SEM. Statistical significance was determined by one-way ANOVA with post hoc Dunnett’s test for most parameters. Statistical significance was determined by paired t-test for comparison between before and after treatments. Statistical significance was determined by a χ² test for incidence of ventricular arrhythmias. Normal distribution of data was confirmed with the Kolmogorov–Smirov test and equal variance with the Levene’s test. Statistical significance was accepted when P < 0.05.

3. Results

3.1 Haemodynamic and morphological characteristics in sham and CHF rats

At 14 weeks of post-MI or sham-operation, the haemodynamic and morphological characteristics were measured (Supplementary material online, Figure S1). CHF rats presented a dense scar in the anterior ventricular wall with infarct size ranging 35–50% (41.7 ± 0.9%) of the left ventricle (Supplementary material online, Figure S1A and E). Heart weight, heart weight/body weight (HW/BW), lung weight, and lung weight/body weight (LW/BW) were higher in CHF rats than in sham rats, suggesting cardiac hypertrophy and substantial pulmonary congestion in CHF (Supplementary material online, Figure S1F–I). Additionally, left ventricular systolic pressure was decreased to 98.3 ± 1.7 mmHg, whereas left ventricular end diastolic pressure was increased to 19.4 ± 0.6 mm Hg in CHF rats, compared to sham rats (120.2 ± 1.5 and 1.6 ± 0.1 mmHg, respectively; Supplementary material online, Figure S1J and K). There was no significant difference in BW, mean arterial pressure, or HR between sham and CHF rats (Supplementary material online, Figure S1B–D).

3.2 Transfection efficiency and silencing effect of lentiviral Cav2.2-α shRNA in the SG from CHF rats

Lentiviral Cav2.2-α shRNA with GFP (2 µL/side, 2 × 10⁷ TU/mL) was in vivo microinjected into bilateral SGs at 12 weeks of post-MI (Figure 2A). By detecting GFP expression in the SG, we evaluated transfection efficiency of Cav2.2-α shRNA and found that Cav2.2-α shRNA with GFP was expressed in almost all SG neurons (Figure 2B). We also measured the expression of Cav2.2-α mRNA and protein in the SG to verify the silencing effect of Cav2.2-α shRNA (Figure2C and D). There was no significant difference in the expression of Cav2.2-α mRNA and protein in the SG between sham and CHF rats. Cav2.2-α shRNA reduced the expression of Cav2.2-α mRNA and protein (0.54 ± 0.05 and 0.23 ± 0.01 in Cav2.2-α shRNA + CHF group vs. 1.06 ± 0.16 and 0.46 ± 0.02 in CHF group, P < 0.05, Figure2C and D) in the SG from CHF rats. However, scrambled shRNA had no effect on the expression of Cav2.2-α mRNA and protein in the SG from CHF rats (Figure2C and D). Supplementary material online, Figure S2 shows original gels of western blot analysis for Cav2.2-α and β-actin protein. Additionally, in vivo transfection of Cav2.2-α shRNA or scrambled shRNA did not affect the expression of L-type (Cav 1.2, Cav 1.3), P/Q-type (Cav 2.1), and R-type (Cav 2.3) Ca²⁺ channel α-subunit mRNA in the SG from CHF rats (Supplementary material online, Figure S3).

3.3 The effects of Cav2.2-α shRNA on N-type Ca²⁺ currents and cell excitability of CSP neurons in CHF rats

We recorded VDCC currents and APs in DiI-labelled SG neurons (i.e. CSP neurons, Figure 3). A specific N-type Ca²⁺ channel blocker, ω-conotoxin GVIA, was used to separate N-type Ca²⁺ currents from total Ca²⁺ currents. N-type Ca²⁺ currents were obtained by subtracting Ca²⁺ currents under treatment of ω-conotoxin GVIA from total Ca²⁺ currents (Figure 3Aa). Compared to sham rats, CHF significantly increased total Ca²⁺ currents, N-type Ca²⁺ currents and cell excitability (including frequency of APs, current threshold inducing AP, and threshold potential) in CSP neurons (Figure 3 and Supplementary material online, Figures S4A and B and S5A), which is consistent with data from our previous study.¹⁹ Although Cav2.2-α shRNA had no effect on other types of Ca²⁺ currents in CSP neurons from CHF rats (Supplementary material online, Figure S5B), it normalized CHF-increased total Ca²⁺ currents (Supplementary material online, Figure S5A), N-type Ca²⁺ currents (24.5 ± 2.6 pA/pF in Cav2.2-α shRNA + CHF group vs. 49.4 ± 2.0 pA/pF in CHF group, P < 0.05, Figure 3A), and cell excitability (Figure 3B and Supplementary material online, Figure S4A and B) in CSP neurons. However, scrambled shRNA did not affect all types of Ca²⁺ currents and cell excitability in CSP neurons from CHF rats (Figure 3 and Supplementary material online, Figures S4 and S5). Additionally, there were no significant differences in input resistance, resting membrane potential and cell membrane capacitance among groups (Supplementary material online, Figures S4C and D and S5C).

Figure 3.

Effect of Cav2.2-α shRNA on N-type Ca²⁺ currents and cell excitability in CSP neurons in CHF. (A) Original recording of Ca²⁺ currents (a), current–voltage (I–V) curve (b), and quantitative data (c) of N-type Ca²⁺ currents in CSP neurons. (B) Original recording of APs (a), frequency of APs (b), and current threshold inducing APs (c) in CSP neurons. ω-Conotoxin GVIA: a specific N-type Ca²⁺ channel blocker. Data are means ± SEM; n = 8 neurons from 6 rats per group. Statistical significance was determined by one-way ANOVA with post hoc Dunnett’s test. *P < 0.05 vs. sham; ^†P < 0.05 vs. CHF.

3.4 Transfection of Cav2.2-α shRNA into CSP neurons improved cardiac sympathetic overactivation in both conscious and anaesthetized CHF rats

Our previous studies showed that cardiac sympathetic overactivation was accompanied by increases of N-type Ca²⁺ currents and cell excitability in CSP neurons from CHF rats.¹⁹ Our current study provided direct evidence to clarify the relationship between the alteration of N-type Ca²⁺ channels in CSP neurons and cardiac sympathetic overactivation in CHF. Transfection of Cav2.2-α shRNA in SGs markedly reduced the CSNA in CHF rats towards the levels seen in sham rats (34.7 ± 5.8% in Cav2.2-α shRNA + CHF group vs. 64.5 ± 3.4% in CHF group, P < 0.05, Figure 4A). However, the CSNA was still kept at a high level in CHF rats after transfection of scrambled shRNA in SGs (Figure 4Ab).

Figure 4.

Transfection of Cav2.2-α shRNA into CSP neurons improved CHF-induced cardiac sympathetic overactivation in CHF rats. (A) Representative tracings (a) and quantitative data (b) showing CSNA recorded in anaesthetized rats. (B) Representative (a and b) and quantitative (c and e) data of HRV recorded in conscious rats. Spectral power was quantified for LF from 0.2 to 0.75 Hz and HF from 0.75 to 2.5 Hz. Data are means ± SEM; n = 6 rats per group. Statistical significance was determined by one-way ANOVA with post hoc Dunnett’s test. *P < 0.05 vs. sham; ^†P < 0.05 vs. CHF.

We further used power spectral analysis of HRV from a 24-h ECG recording to test the effect of Cav2.2-α shRNA on cardiac sympathetic overactivation in conscious CHF rats. In power spectral analysis of HRV, HF comprises cardiac parasympathetic oscillation, while LF/HF ratio shows cardiac sympathetic modulation.^28,29 LF and LF/HF ratio were increased, whereas HF was decreased in conscious CHF rats, compared to sham rats (Figure 4B). Cav2.2-α shRNA significantly lowered LF (5.58 ± 0.81 ms² in Cav2.2-α shRNA + CHF group vs. 8.88 ± 0.83 ms² in CHF group, P < 0.05, Figure 4Bc) and LF/HF ratio (1.06 ± 0.07 ms² in Cav2.2-α shRNA + CHF group vs. 2.12 ± 0.05 ms² in CHF group, P < 0.05, Figure 4Be), but it had no significant effect on HF in CHF rats (5.14 ± 0.57 ms² in Cav2.2-α shRNA + CHF group vs. 4.19 ± 0.36 ms² in CHF group, P > 0.05, Figure 4Bd). Scramble shRNA did not show any effect on LF, HF, and LF/HF ratio in CHF rats (Figure 4B).

When HRV parameters were compared between before and after treatments, we found that there were no significant differences in HRV parameters in sham plus saline, CHF plus saline, and CHF plus scrambled shRNA groups (Supplementary material online, Table S2). However, a significant decrease in LF and LF/HF ratio (P < 0.05) and a slight increase in HF (P > 0.05) were found after transfection of Cav2.2-α shRNA in SGs of CHF rats, compared with before treatment (Supplementary material online, Table S2). Additionally, comparing HRV parameters between day and night, we found that LF and LF/HF ratio were higher, whereas HF was lower at night than during the day in all groups (Supplementary material online, Figure S6). Transfection of Cav2.2-α shRNA but not scrambled shRNA in SGs of CHF rats markedly reduced LF and LF/HF ratio during both day and night (Supplementary material online, Figure S6). The above data directly demonstrated that inhibition of N-type Ca²⁺ channels in SGs normalized CHF-induced cardiac sympathetic overactivation.

3.5 Transfection of Cav2.2-α shRNA into CSP neurons restored heterogeneity of ventricular electrical activities in CHF rats

Heterogeneity of ventricular electrical activities (including QT and QTc intervals, QT and QTc dispersions, and Tpe interval) is a key factor for ventricular arrhythmogenesis in CHF.²³ Data from radiotelemetry conscious ECG recording demonstrated that CHF prolonged QT and QTc intervals, QT and QTc dispersions, and Tpe interval, compared to sham rats (Figure 5A). In vivo transfection of Cav2.2-α shRNA into SGs significantly attenuated CHF-induced prolongation of QT and QTc intervals (70.0 ± 5.24 ms and 169.10 ± 6.32 ms in Cav2.2-α shRNA + CHF group vs. 87.85 ± 3.08 ms and 204.97 ± 6.7 ms in CHF group, P < 0.05, Figure 5Ba,b), QT and QTc dispersions (33.63 ± 2.83 ms and 73.44 ± 6.79 ms in Cav2.2-α shRNA + CHF group vs. 50.82 ± 3.67 ms and 102.48 ± 6.17 ms in CHF group, P < 0.05, Figure 5Bc,d), and Tpe interval (37.77 ± 2.48 ms in Cav2.2-α shRNA + CHF group vs. 56.65 ± 3.63 ms in CHF group, P < 0.05, Figure 5Be). Transfection of scrambled shRNA in SGs had no effect on these ventricular arrhythmia-related ECG markers in CHF rats (Figure 5B).

Figure 5.

Transfection of Cav2.2-α shRNA into CSP neurons attenuated heterogeneity of ventricular electrical activities in conscious CHF rats. (A) Representative tracings for QT and Tpe intervals in CHF and Cav2.2-α shRNA+CHF rats. The T_peak (Tp) is defined as the intersection point between the tangent of the descending limb and ascending limb of the T wave (the red point), and the T_end (Te) is defined as the intersection point between the tangent of the ascending limb of the T wave and the isoelectric line (the green point). (B) Quantitative data for QT interval (a), QTc interval (b), QT dispersion (c), QTc dispersion (d), and Tpe (e) in all groups. Data are means ± SEM; n = 6 rats per group. Statistical significance was determined by one-way ANOVA with post hoc Dunnett’s test. *P < 0.05 vs. sham; ^†P < 0.05 vs. CHF.

A comparison of ventricular arrhythmia-related ECG markers between before and after treatments or between day and night in all groups was shown in Supplementary material online, Table S2 or Supplementary material online, Figure S7. There were no significant differences in ventricular arrhythmia-related ECG markers between before and after treatments in sham plus saline, CHF plus saline, and CHF plus scrambled shRNA groups (Supplementary material online, Table S2). However, these ventricular arrhythmia-related ECG markers were shortened after transfection of Cav2.2-α shRNA in SGs of CHF rats, compared with before treatment (Supplementary material online, Table S2). We also found that ventricular arrhythmia-related ECG markers were longer at night than during the day in all groups (Supplementary material online, Figure S7). Transfection of Cav2.2-α shRNA but not scrambled shRNA in SGs markedly blunted CHF-induced prolongation of these ventricular arrhythmia-related ECG markers during both day and night (Supplementary material online, Figure S7).

3.6 Transfection of Cav2.2-α shRNA in CSP neurons decreased ventricular arrhythmias in both conscious and anaesthetized CHF rats

Spontaneous ventricular arrhythmia was monitored by radiotelemetry ECG recording in conscious rats (Figure 6A). In sham rats, neither PVCs nor VT/VF were observed. In CHF rats, 100% (6/6) and 83% (5/6) of rats had PVCs and VT/VF, respectively. The number of PVCs and cumulative duration of VT/VF were markedly increased in CHF rats (Figure 6A). In vivo transfection of Cav2.2-α shRNA in SGs significantly reduced the number of PVCs (234 ± 68 beats/h in Cav2.2-α shRNA + CHF group vs. 683 ± 89 beats/h in CHF group, P < 0.05, Figure 6Ac), and incidence and cumulative duration of VT/VF (50% and 2.3 ± 1.3 s/h in Cav2.2-α shRNA + CHF group vs. 83% and 10.0 ± 2.8 s/h in CHF group, P < 0.05, Figure 6Ad,e) in CHF rats, although it had no effect on incidence of PVCs (Figure 6Ab). Contrary to Cav2.2-α shRNA, scrambled shRNA failed to affect the occurrence of spontaneous ventricular arrhythmias in conscious CHF rats (Figure 6A).

Figure 6.

Transfection of Cav2.2-α shRNA into CSP neurons decreased ventricular arrhythmias in both conscious and anaesthetized CHF rats. (A) Raw ECG recordings for PVCs and VT/VF in conscious CHF and Cav2.2-α shRNA+CHF rats (a), and mean data for incidence of PVCs (b), the number of PVCs (c), incidence of VT/VF (d), and cumulative duration of VT/VF (e) in conscious rat groups. (B) Raw data for PES-evoked VT/VF in anaesthetized CHF and Cav2.2-α shRNA+CHF rats (a), mean data for incidence (b), and inducibility quotient (c) of PES-evoked VT/VF in anaesthetized rat groups. Data are means ± SEM; n = 6 rats per group. Statistical significance was determined by a χ² test for data presented in panels Ab, Ad, and Bb. Statistical significance was determined by one-way ANOVA with post hoc Dunnett’s test for data presented in panels Ac, Ae, and Bc. *P < 0.05 vs. sham; ^†P < 0.05 vs. CHF.

The occurrence of spontaneous ventricular arrhythmias was compared between before and after treatments in all groups of conscious rats to eliminate the influence derived from difference between animal groups (Supplementary material online, Table S2). In CHF plus Cav2.2-α shRNA group, the number of PVCs and incidence and cumulative duration of VT/VF were significantly decreased after transfection of Cav2.2-α shRNA to SGs, compared to before treatment. However, there were no differences in all parameters of spontaneous ventricular arrhythmias before and after treatments in sham, CHF, and CHF plus scrambled shRNA groups (Supplementary material online, Table S2). Additionally, when all parameters of spontaneous ventricular arrhythmias were compared between day and night, we found that the number of PVCs and incidence and cumulative duration of VT/VF were higher or longer at night than during the day in CHF, CHF plus scrambled shRNA, and CHF plus Cav2.2-α shRNA groups, except sham group (Supplementary material online, Figure S8). Transfection of Cav2.2-α shRNA in SGs from CHF rats significantly decreased the number of PVCs, and incidence and cumulative duration of VT/VF during both day and night (Supplementary material online, Figure S8).

To further clarify the influence of Cav2.2-α shRNA in ventricular arrhythmogenesis, inducibility of ventricular arrhythmias, a widely used parameter for studying susceptibility to ventricular arrhythmias, was detected in anaesthetized rats. An inducibility quotient was used to quantify inducibility of ventricular arrhythmias triggered by a PES protocol in anaesthetized rats. PES-induced fatal VT/VF with a high inducibility quotient in CHF rats, compared to sham rats (Figure 6B). Transfection of Cav2.2-α shRNA in SGs significantly decreased incidence of malignant VT/VF and inducibility quotient in anaesthetized CHF rats (60% and 1.5 ± 0.76 in Cav2.2-α shRNA + CHF group vs. 100% and 5.12 ± 0.54 in CHF group, P < 0.05, Figure 6Bb,c). Scrambled shRNA did not show any effect on PES-triggered ventricular arrhythmias in anaesthetized CHF rats (Figure 6B). In this part of the experiments, a sustained VT or an episode of VF induced by PES was not treated with direct current defibrillation because PES-induced ventricular arrhythmias including VT/VF terminated spontaneously in all animals and there was no animal died of PES-induced VT/VF.

3.7 Transfection of Cav2.2-α shRNA in CSP neurons failed to improve the performance of the failing heart

In the progression of heart failure, echocardiographic data in CHF rats demonstrated a significant reduction in cardiac performance from 1 week to 12 weeks of post-MI, compared to sham rats (Supplementary material online, Figure S9). In vivo transfection of Cav2.2-α shRNA in SGs had no ability to improve CHF-induced cardiac contractile dysfunction, including EF, FS, LVDd, LVDs, LVd Vol, and LVs Vol, which were measured at 14 weeks of post-MI (Figure 7). When these parameters of cardiac contractile function were compared between before and after treatments, there was no significant difference in all groups (Supplementary material online, Table S2).

Figure 7.

Transfection of Cav2.2-α shRNA into CSP neurons failed to improve performance of the failing heart. (A) Representative B-model (upper) and M-mode (bottom) echocardiography images in the left ventricles of Sham (a), CHF (b), scrambled shRNA+CHF (c), and Cav2.2-α shRNA+CHF rats (d). (B) Quantitative data for EF (a), FS (b), LVDd (c), LVDs (d), LVd Vol (e), and LVs Vol (f) calculated from M-mode images of parasternal long axis view in all groups. Data are means ± SEM; n = 6–7 rats per group. Statistical significance was determined by one-way ANOVA with post hoc Dunnett’s test. *P < 0.05 vs. sham.

Supplementary material online, Figure S1 summarizes haemodynamic and morphological characteristics in all groups. As described above, CHF rats presented a huge infarct size in the left ventricle, left ventricular contractile dysfunction, cardiac hypertrophy, and substantial pulmonary congestion (Supplementary material online, Figure S1). In vivo transfection of scrambled shRNA or Cav2.2-α shRNA in SGs failed to improve haemodynamic and morphological parameters in CHF rats (Supplementary material online, Figure S1).

4. Discussion

4.1 Major findings

Our current study reports a major contribution of N-type Ca²⁺ channels in CHF-induced cardiac sympathetic overactivation and ventricular arrhythmogenesis. We demonstrate for the first time that inhibition of N-type Ca²⁺ channels through in vivo transfection of Cav2.2-α shRNA was found to restore CHF-induced increases in N-type Ca²⁺ currents and cell excitability in CSP neurons. In vivo transfection of Cav2.2-α shRNA in CSP neurons markedly improved cardiac autonomic function through restoring cardiac sympathetic overactivation in CHF rats. Of greater importance, inhibition of N-type Ca²⁺ currents in CSP neurons improved the CHF-induced heterogeneity of ventricular electrical activity, reduced incidence and duration of VT/VF in conscious rats, and decreased susceptibility to ventricular arrhythmias in anaesthetized CHF rats. However, in vivo transfection of Cav2.2-α shRNA into CSP neurons failed to improve CHF-induced cardiac contractile dysfunction. These data directly demonstrated that overactivation of N-type Ca²⁺ channels in CSP neurons contributes to cardiac sympathetic overactivation and ventricular arrhythmogenesis in CHF.

4.2 Sympathetic overactivation and ventricular arrhythmogenesis

CHF is characterized by sympathetic overactivation and parasympathetic withdrawal. Sympathetic overactivation demonstrates a wide variety of cardiovascular actions (including increased cardiac contractility and HR acceleration), which initially compensate for depressed myocardial function and maintain sufficient circulation during the progression of CHF.¹¹ However, long-term sympathetic overactivation has deleterious effects on cardiac structure and function and leads to the progression of CHF and SCD.^11,30 Widespread evidence has already demonstrated the critical role of sympathetic overactivation in ventricular arrhythmogenesis, particularly the contribution of the SG.⁸ Ventricular arrhythmias in CHF are usually triggered by the sympathetic burst recorded in SGs.⁷ In a canine model of SCD, VT and SCD were immediately preceded by high sympathetic discharges from the SG.³¹ The above findings are consistent with the data from our present study, in which cardiac sympathetic overactivation shown by increased CSNA and LF/HF ratio was accompanied by ventricular arrhythmias in CHF rats.

Considering the critical contribution of sympathetic excitation in ventricular arrhythmogenesis and SCD, the neuromodulation of sympathetic input in the failing heart is becoming the most important strategy for suppressing ventricular arrhythmias in patients with CHF. As a therapeutic strategy, targeting neuronal β-adrenoceptor signalling in SG neurons to reduce peripheral sympathetic input in the pathophysiological state was recently explored by Bardsley et al.³² It is the first time to demonstrate the presence of β-receptors in both human and rat SG neurons, and suggest that sympathetic overactivation-increased norepinephrine release might provide positive feedback on neuronal β-receptors to promote further release of neurotransmitter in sympathetic nerve terminals. Additionally, β-blockers are serving as first-line therapy in the management of ventricular arrhythmias and the prevention of SCD related to sympathetic activation.³³ However, such pharmacological treatments may not be ideal. Some studies, for example, have demonstrated that β-blockers do not provide satisfactory protection against SCD, and it has been suggested that some patients are either intolerant or refractory to this therapy.^4,5 The possible explanation for the incomplete protection of β-blocker therapy is that β-blockers only antagonize β-adrenergic receptors, but they do not normalize CHF-enhanced cardiac sympathetic activity and affect co-neurotransmitters (such as neuropeptide Y) released from cardiac sympathetic nerve terminals that are also involved in arrhythmogenesis and SCD.^34–36 Moreover, despite being effective for reducing cardiac sympathetic activity and ventricular arrhythmias, thoracic sympathectomy or CSD has adverse complications (including hyperhidrosis, Horner’s syndrome, and paraesthesia) that limit use of the procedure in patients with CHF^15,37 and lead to the de-emphasis of thoracic sympathectomy as a last resort after other interventions fail.

4.3 Heterogeneity of ventricular electrical activities associate with ventricular arrhythmogenesis in CHF

Ventricular arrhythmogenesis is attributed to changes in cardiac mechanical, morphological, metabolic, and electrophysiological properties in the failing heart.³⁸ Although it is unclear what mechanisms account for cardiac sympathetic activation-triggered ventricular arrhythmias in CHF, a study in normal Yorkshire pigs demonstrates that sympathetic nerve stimulation prolonged Tpe interval, a marker of transmural dispersion of ventricular repolarization.²³ Our previous²⁰ and current studies report CHF-induced prolongation of QT, QTc, and Tpe intervals, and augmentation of QT and QTc dispersions. More important, our current study showed that Ca²⁺ channel shRNA-induced normalization of cardiac sympathetic activation attenuated the heterogeneity of ventricular electrical activities (including QT, QTc, and Tpe intervals, and QT and QTc dispersions), and subsequently decreased ventricular arrhythmias in CHF rats. Optogenetic approach-inhibited neural activity of left SG similarly decreased ventricular arrhythmias in canines with acute MI.³⁹ These data indicate that cardiac sympathetic overactivation triggers ventricular arrhythmias through inducing heterogeneity of ventricular electrical activities (including QT, QTc, and Tpe intervals, and QT and QTc dispersions) in CHF. However, our current study found that N-type Ca²⁺ channel shRNA-induced normalization of cardiac sympathetic activation did not improve the abnormality of cardiac morphology and contractile function (Figure 7 and Supplementary material online, Figure S1, i.e. other substrates of ventricular arrhythmias in CHF³⁸), although clinicians and cardiologists hope to decrease ventricular arrhythmias through reducing infarct size and improving cardiac contractile function in failing hearts.

4.4 Targeting N-type calcium channels in CSP neurons is an effective intervention against ventricular arrhythmias

VDCCs are involved in cell excitability, intracellular Ca²⁺ level, and neurotransmitter release in the central and peripheral nervous systems. Thus far, four types (L, N, P/Q, and R) of VDCCs have been functionally characterized in CSP neurons.¹⁹ Our previous study demonstrated that only N-type Ca²⁺ currents were increased in CSP neurons from CHF rats, compared to sham rats.¹⁹ N-type Ca²⁺ currents account for about 60–70% of all Ca²⁺ currents in CSP neurons,¹⁹ and neurotransmitter release from cardiac sympathetic nerve terminals is triggered by Ca²⁺ influx via N-type but not L- and P/Q-type Ca²⁺ channels.¹⁸ Yamada et al. demonstrated that genetic deletion of N-type Ca²⁺ channels improved autonomic dysfunction and prevented ventricular arrhythmias in CHF mice.¹⁰ In our current study, Cav2.2-α shRNA transfection-induced decrease in N-type Ca²⁺ channels in CSP neurons not only normalized cardiac sympathetic overactivation, but also reduced ventricular arrhythmogenesis in CHF rats.

These data provide a potential therapeutic strategy for CHF-increased cardiac sympathetic activity and ventricular arrhythmias, and are crucial for improving prognosis of CHF. However, some issues remain to be considered. First, it is a significant challenge to transfect Cav2.2-α shRNA into CSP neurons in patients with CHF. Second, selective N-type Ca²⁺ channel blockers such as ω-conotoxin GVIA and ω-conotoxin MVIIA (ziconotide) are relatively large polypeptides with limited distribution and permeation in ganglionic tissue during in vivo treatment. Third, some available small-molecular drugs (such as cilnidipine) are dual N- and L-type Ca²⁺ channel blockers with direct negative inotropic effects, although cilnidipine decreases ventricular arrhythmias in CHF mice.¹⁰ Fourth, N-type Ca²⁺ channels are only expressed in neurons,⁴⁰ and the haemodynamic and echocardiographic data in our current study demonstrate that inhibition of N-type Ca²⁺ channels in CSP neurons did not present negative inotropic and chronotropic effects in CHF rats (Figure 7 and Supplementary material online, Figure S1). Fifth, our previous study demonstrated that N-Type Ca²⁺ currents and cell excitability of CSP neurons located in the SG are increased, whereas N-Type Ca²⁺ currents and cell excitability of cardiac parasympathetic post-ganglionic (CPP) neurons located in intracardiac ganglia are decreased in CHF rats.¹⁹ Although the mechanisms responsible for the opposite electrophysiological changes between CSP and CPP neurons in CHF are unclear, we clearly understand that ubiquitous distribution of N-type Ca²⁺ channel blockers will cause a worsening of cardiac parasympathetic activation and dampen its therapeutic effects on cardiac sympathetic overactivation in CHF. Therefore, discovering pure and potent small-molecule N-type Ca²⁺ channel blockers with local (CSP neurons) targeting drug delivery system could be an effective new intervention to improve outcomes for patients with systolic CHF.

Our previous¹⁹ and current studies (Figures 2and3) demonstrated that CHF-increased N-type Ca²⁺ currents but not affected expression of Cav2.2-α mRNA and protein in CSP neurons. These results suggest that CHF-increased N-type Ca²⁺ currents are not due to transcriptional and translational modulations of Cav2.2-α subtypes in CSP neurons from CHF rats. Overactivation of N-type Ca²⁺channels is possibly induced by some post-translational modulations, such as phosphorylation of N-type Ca²⁺ channels, alteration of auxiliary subunits associating with N-type Ca²⁺ channels, and allosteric modulation of N-type Ca²⁺ channels.^41,42 Considering the sake of clarity and focus in this study, and many possibilities for post-translational modulation of N-type Ca²⁺ channels, further study is needed to explore the mechanisms responsible for CHF-increased N-type Ca²⁺ currents in CSP neurons.

4.5 Study limitations

There are several limitations in our present study. First, spectral analysis of HRV was used to determine cardiac autonomic function in conscious rats in the current study. However, there is specific limitation of HRV application for estimating cardiac sympathetic activation in CHF. The value of LF component in HRV analysis is dependent on numerous factors including baroreflex function, cardiac beta-receptor sensitivity, post-receptor transduction, and parasympathetic modulation.⁴³ Thus, using the LF component of HRV alone to estimate cardiac sympathetic activation is likely weak, especially in the CHF state where post-synaptic function may be altered.^43,44 Indeed, the LF component in HRV analysis is markedly reduced in patients with advanced CHF.⁴⁵ Nevertheless, LF/HF used in our present study is consistently thought to represent sympathovagal balance or to reflect sympathetic modulation.²⁸ Considering the limitation of HRV analysis, we also measured the CSNA in anaesthetized rats to further confirm cardiac sympathetic overactivation in our present study because the CSNA is not affected by post-synaptic mechanisms. Second, microinjection of Cav2.2-α shRNA into SGs cannot specifically target the CSP neurons in our present study because specifically targeting the CSP neurons requires the development of advanced techniques not yet available. Nevertheless, gene-silencing N-type Ca²⁺ channels in SGs not only reduced N-type Ca²⁺ currents and cell excitability in CSP neurons and CSNA towards normal (sham) levels, but also decreased ventricular arrhythmias in CHF rats. These data provide the potential therapeutic target for decreasing ventricular arrhythmias in patients with CHF. Third, our current study did not directly address electrophysiology in cardiac sympathetic nerve terminals, because the small size and complex architecture of sympathetic nerve terminals embedded in the myocardium does not allow at present for direct measurement. Nevertheless, these limitations motivated us to develop an alternative preparation of the isolated soma of CSP neurons for electrophysiological recording. The data from immunofluorescence staining demonstrated that in vivo transfection of Cav2.2-α shRNA into SGs also reduced the expression of Cav2.2-α protein in cardiac sympathetic nerve terminals (Supplementary material online, Figure S10), similar to that in the cell somata of SGs. Therefore, it is acceptable to test the functional role of N-type Ca²⁺ channels in regulating cardiac sympathetic activation by comparing data from isolated CSP neurons with results from in vivo measurements of CSNA and LF/HF ratio of HRV in all groups. Fourth, Tpe interval used in current study is thought to represent a marker of transmural dispersion of repolarization. In the experimental model, repolarization of the epicardium (shortest AP) coincides with the peak of the T wave and repolarization of the M cells (longest APs) with the end of the T wave, so Tpe interval provides a measure of transmural dispersion of repolarization.⁴⁶ However, a debate continues as to whether Tpe interval indicates the total dispersion of repolarization or transmural dispersion of repolarization of the heart.⁴⁷ On the one hand, a few papers published by Opthof et al.⁴⁷ and Izumi et al.⁴⁸ suggested that Tpe interval does not reflect transmural dispersion. They measured Tpe interval in open chest dogs under the anaesthetized condition.^47,48 In general, anaesthesia might suppress voltage-gated Na⁺ currents in a variety of cells including M cells.⁴⁷ Inhibition of Na⁺ currents plus the use of a pacing rate of 130 b.p.m.,⁴⁷ could result in no transmural dispersion of repolarization.⁴⁹ On the other hand, Tpe interval is thought to represent a marker of transmural dispersion of repolarization of the heart in most recent studies.^23,46,50 Nevertheless, Tpe interval reflects the repolarization heterogeneity of ventricular myocardium.⁵⁰

4.6 Perspectives

Cardiac sympathetic overactivation is a major feature in CHF. The role of cardiac sympathetic hyperactivity in CHF is highlighted by use of β-adrenergic receptor blockers and CSD as the selectable approaches to the current therapy of ventricular arrhythmias in CHF. However, some previous studies have demonstrated that β-blockers do not provide satisfactory protection against SCD, and even that some patients are either intolerant or refractory to this therapy.^4,5 Additionally, adverse complications after the procedure of CSD—limit the use of the procedure in patients with CHF, and this procedure is only chosen as a last resort after other interventions fail. Our present study demonstrates that inhibition of N-type Ca²⁺ channels in CSP neurons attenuates CHF-induced cardiac sympathetic overactivation and ventricular arrhythmias. The clinical significance of this study is to open a new avenue in therapeutics working against lethal ventricular arrhythmias in patients with CHF. N-type Ca²⁺ channels in CSP neurons could be a new therapeutic target for cardiac sympathetic overactivation and ventricular arrhythmias in CHF. Exploring specific, small-molecule N-type Ca²⁺ channel blockers with local targeting drug delivery system could translate to clinical trials and applications that improve outcomes for CHF patients with lethal ventricular arrhythmias.

5. Conclusions

In vivo transfection of Cav2.2-α shRNA reduces activation of N-type Ca²⁺ channels and cell excitability in CSP neurons from CHF rats. It further decreases CHF-induced cardiac sympathetic overactivation and ventricular arrhythmias. These data clearly demonstrate that hyperactivation of N-type Ca²⁺ channels in CSP neurons successively increases calcium influx, calcium-related NE release, and cardiac sympathetic activation in CHF. Finally, cardiac sympathetic overactivation causes lethal ventricular arrhythmias through triggering the heterogeneity of ventricular electrical activities in CHF. From these data, we conclude that inhibition of N-type Ca²⁺ channels in CSP neurons attenuates CHF-induced cardiac sympathetic overactivation and ventricular arrhythmias. Our present data provide a potential therapeutic strategy to decrease ventricular arrhythmia in the CHF state.

Supplementary material

Supplementary material is available at Cardiovascular Research online.

Author contributions

D.Z. and Y.L.L. conceived and designed the experiments. D.Z., H.T., C.W., L.C., W.H., B.T.H., and Y.L.L. performed the experiments. D.Z., H.T., B.T.H., R.L.M., M.C.W., and Y.L.L. analysed the data. Y.L.L., R.L.M., and M.C.W. contributed reagents/materials/analysis tools. D.Z. and Y.L.L. wrote the article.

Conflict of interest: none declared.

Funding

This work was supported by the National Institute of Health’s National Heart, Lung, and Blood Institute [R01HL-098503, R01HL-137832, and R01-HL144146 to Y.L.L.], and the American Heart Association [15GRANT24970002 to Y.L.L.].

 

Time for primary review: 19 days

 

Translational perspectives

Our present study demonstrates that inhibition of N-type Ca²⁺ channels in CSP neurons attenuates CHF-induced cardiac sympathetic overactivation and ventricular arrhythmias. The clinical significance of this study is to open a new avenue in therapeutics working against lethal ventricular arrhythmias in patients with CHF. N-type Ca²⁺ channels in CSP neurons could be a new therapeutic target for cardiac sympathetic overactivation and ventricular arrhythmias in CHF. Exploring specific, small-molecule N-type Ca²⁺ channel blockers with local targeting drug delivery system could translate to clinical trials and applications that improve outcomes for patients with CHF.