Decreased ability to manage increases in reactive oxygen species may underlie susceptibility to arrhythmias in mice lacking Scn1b

Abstract

Scn1b plays essential roles in the heart, where it encodes β₁-subunits that serve as modifiers of gene expression, cell surface channel activity, and cardiac conductivity. Reduced β₁ function is linked to electrical instability in various diseases with cardiac manifestations and increased susceptibility to arrhythmias. Recently, we demonstrated that loss of Scn1b in mice leads to compromised mitochondria energetics and reactive oxygen species (ROS) production. In this study, we examined the link between increased ROS and arrhythmia susceptibility in Scn1b^−/− mice. In addition, ROS-scavenging capacity can be overwhelmed during prolonged oxidative stress, increasing arrhythmia susceptibility. Therefore, we isolated whole hearts and cardiomyocytes from Scn1b^−/− and Scn1b^+/+ mice and subjected them to an oxidative challenge with diamide, a glutathione oxidant. Next, we analyzed gene expression and activity of antioxidant enzymes in Scn1b^−/− hearts. Cells isolated from Scn1b^−/− hearts died faster and displayed higher rates of ROS accumulation preceding cell death compared with those from Scn1b^+/+. Furthermore, Scn1b^−/− hearts showed higher arrhythmia scores and spent less time free of arrhythmia. Lastly, we found that protein expression and enzymatic activity of glutathione peroxidase is increased in Scn1b^−/− hearts compared with wild type. Our results indicate that Scn1b^−/− mice have decreased capability to manage ROS during prolonged oxidative stress. ROS accumulation is elevated and appears to overwhelm ROS scavenging through the glutathione system. This imbalance creates the potential for altered cell energetics that may underlie increased susceptibility to arrhythmias or other adverse cardiac outcomes.

NEW & NOTEWORTHY Using an oxidative challenge, we demonstrated that isolated cells from Scn1b^−/− mice are more susceptible to cell death and surges in reactive oxygen species accumulation. At the whole organ level, they were also more susceptible to the formation of cardiac arrhythmias. This may in part be due to changes to the glutathione antioxidant system.

INTRODUCTION

The central function of the heart is to act as a pump, driving oxygenated blood throughout the body. The systolic phase, or period where the heart contracts, is induced via electrical excitation of cardiomyocytes; a cardiac action potential originates in the sinoatrial node and spreads throughout the heart to ventricular cardiomyocytes, which signals these cells to contract. Cardiac arrhythmias occur when the electrical function of the heart is disrupted leading to deviation from this standard cardiac conductivity. One genetic element implicated in the pathogenesis of arrhythmias is mutations in genes encoding subunits of voltage-gated sodium channels (VGSCs). In addition, it has been suggested that VGSC mutations could play a role in modulating pharmacological responses to certain sodium channel-blocking antiarrhythmics (1). VGSCs are the primary channel involved in the upstroke (depolarization) phase and propagation of the cardiac action potential. The molecular architecture of VGSCs includes a large, α-subunit ion-conducting pore flanked bound by 1–2 β-subunits (2, 3). Although β-subunits are typically described as an accessory subunit to α-subunit, β-subunits influence a variety of electrical parameters of the cell membrane, including VGSC α-subunit cell surface expression, localization, kinetics, and gating (2). Recently, we have shown that mice lacking the Scn1b subunit have compromised mitochondrial function and increased mitochondrial ROS production (4).

Pathogenic loss-of-function mutations in one β-subunit gene, Scn1b, are implicated in multiple cardiac diseases with arrhythmic phenotypes (2, 3, 5, 6). Scn1b encodes two β-subunit subtypes, which are splice variants of each other, the β₁- and β_1B-subunit. β₁ is the primary subunit that interacts with VGSC α-subunits, whereas β_1B, which lacks a transmembrane domain, may interact with VGSCs, but is largely secreted by the cell as an extracellular ligand (2, 5). β_1B secretion by neurons is essential to brain development, where it promotes neuronal maturation (7). Therefore, Scn1b mutations have also been implicated in diseases presenting with both neuronal and cardiac symptomology. Homozygous loss-of-function mutations in Scn1b can cause Dravet Syndrome (DS), a severe pediatric-onset epilepsy disorder characterized by pharmaco-resistant seizures, cognitive impairment, and developmental delay. Patients with DS are at an increased risk of Sudden Unexpected Death in Epilepsy (SUDEP). Studies of Scn1b^−/− mouse models indicate increased susceptibility to cardiac arrhythmias, which may contribute to SUDEP pathophysiology (8, 9). Overall, a deeper knowledge of the molecular pathways involved in the linkage between Scn1b and cardiac electrical dysfunction could improve pharmaceutical and clinical strategies for preventing arrhythmias in patients with Scn1b mutations.

In the heart, the β₁-subunit plays an important role in cell-cell coupling, gene transcription, and electrical excitability (10–12). β₁ subunits contain large, immunoglobulin extracellular domains, allowing them to function as cell adhesion molecules (5). Extracellular domains of β₁-subunits may participate in trans-homophilic interactions (13) and may also form heterophilic interactions with other cell adhesion molecules (14). Furthermore, β₁-subunits are capable of binding to extracellular matrix proteins, which may aid in VSGC localization (15). Under normal conditions, Scn1b expression is localized to intercalated disk regions, where it is expressed alongside the VGSC Na_V1.1 (16). Abnormal control of VGSC localization may contribute to disease states associated with a lack of β₁-expression. In Scn1b null tissue, tetrodotoxin-sensitive Na⁺ channels like Na_V1.1 are directed away from their normal position at the intercalated disk and abundantly expressed in the T-tubular region, where they are speculated to underlie abnormal intracellular Ca²⁺ release events (8).

β₁-Subunits are also substrates for regulated intramembrane proteolysis by β-secretases and γ-secretases (11). Cleavage at an extracellular site by β-secretases leads to shedding of the extracellular immunoglobulin domain, leaving the C-terminal embedded in the membrane. The cleaved extracellular domain is believed to function as a soluble cell adhesion molecule essential for neuronal development (7). The resulting COOH-terminal from β-secretase cleavage may be sequentially targeted by γ-secretases, generating a soluble intracellular fragment that migrates to the nucleus and contributes to transcriptional regulation (17). Signaling by the intracellular domain has been suggested to be repressive (11) as deletion of Scn1b in mice leads to upregulation of several transcripts, including genes involved in cellular potassium and calcium handling, and VGSC genes (8, 9, 11). These transcriptional changes may underlie alterations to cardiac excitability observed in Scn1b knockout models. Furthermore, β₁-subunits have been observed to be able to suppress activity of specific VGSC subtypes to reduce temporal variations in Na⁺ current in the cell (18), an effect that may be lost in Scn1b knockout and contribute to disease states associated with the loss of β₁-signaling.

In addition to its influences on cell surface expression, β₁-subunits modulate Na⁺ current density (I_Na) by regulating channel gating (2). Dissociated ventricular cardiomyocytes from Scn1b null mice show increases in peak and persistent I_Na (8, 9), largely because of an increased influence of tetrodotoxin-sensitive channels (8). The decreased decay of I_Na may underlie changes in conduction velocity observed in Scn1b null mice, including a prolonged repolarization phase and overall action potential duration (9). These electrophysiological abnormalities may contribute to increased arrhythmia susceptibility in Scn1b knockout models. ECG traces from Scn1b null mice show extended R-R and QTc intervals (9). In addition, Scn1b^−/− null mice demonstrate increased susceptibility to ventricular arrhythmia, including triggered beats, and an elevated incidence of abnormal action potential depolarizations (8). Although the prolongation of the QT interval due to changes to I_Na and conduction velocity may account in part for an increased risk of sudden death in Scn1b-linked DS, it likely does not tell the whole story. Although these substrates can predispose hearts to fatal arrhythmias, how stressors may act as triggers to these arrhythmias is unknown.

We recently showed that in a Scn1b knockout mouse, mitochondria from Scn1b^−/− hearts had decreased respiration through complex I-linked pathways involved in ATP synthesis. This suggests that complex I-mediated dysfunction may play a role in the increased arrhythmia risk in these mice. Furthermore, mitochondrial reactive oxygen species (ROS) production was significantly increased via reverse electron transport at complex I, which could be a source of oxidative stress in Scn1b^−/− cardiomyocytes (4). Therefore, this study was performed to investigate if imbalances between ROS accumulation and scavenging exist and may predispose the Scn1b^−/− mouse model of DS to cardiac arrhythmias. Elevated levels of mitochondrial ROS can consume cellular antioxidant resources, predisposing to cardiac disease. Our previous work has established this link between ROS accumulation and cardiac arrhythmias (19–23). In this study, we performed experiments to determine if there were deficiencies in ROS scavenging and inherent susceptibility to cardiac arrhythmias in mice lacking the Scn1b subunit. Our results indicate that the loss of the β₁-subunit in Scn1b^−/− mice results in decreased cardiac ROS scavenging by the glutathione system. Combined with our previous results suggesting increased ROS production via the mitochondrial respiratory chain, this study indicates a discrepancy exists between ROS buffering and accumulation in the Scn1b^−/− model. This imbalance may negatively impact cell energetics and contribute to lethal arrhythmias.

MATERIALS AND METHODS

Mice

All animal procedures were approved by East Tennessee State University’s Institutional Animal Care and Usage Committee and conform with the National Institutes of Health policy on Humane Care and Use of Laboratory Animals (NIH Publication No. 85-23, Revised 1996). Both male and female mice underwent euthanasia via CO₂ at ages between postnatal days 10 and 20 (exact age at experimentation noted in results). Scn1b^−/− mice, congenic on the C57BL/6J background for over 20 N generations, were a generous gift from Dr. Lori L. Isom and were maintained as previously described (4, 6, 9, 24).

Isolated Heart Perfusion

Langendorff heart preparations were made from hearts isolated from Scn1b^+/+ (n = 8) and Scn1b^−/− (n = 6) mice, similarly as previously described (19, 20). After an initial 15-min recording period to collect baseline data, the perfusate was switched to a solution of Krebs buffer with 80 µM diamide for 30 min. In a subset of experiments, hearts were perfused with 5 mM N-acetyl-l-cysteine (NAC) for 10 min following the initial stabilization period. Immediately after NAC treatment, hearts were then delivered Krebs buffer with 80 µM diamide for 30 min. Two independent blinded reviewers analyzed EKG traces for arrhythmia susceptibility and severity using previously established arrhythmia scoring systems (21, 22).

Myocyte Isolation and Imaging

Hearts from Scn1b^+/+ (n = 8) and Scn1b^−/− (n = 7) mice were mounted on a reverse perfusion cannula, and ventricular myocytes dissociated similarly to as previously described (19). Myocytes were loaded with 500 nM 5-(6)-chloromethyl-2,7-dichlorodihydrofluorescein diacetate (CM-DCF; Invitrogen) as we have previously done (22). Scn1b^+/+ (n = 29) and Scn1b^−/− (n = 28) myocytes were loaded on a coverslip of an inverted fluorescent microscope and images were acquired every 30 s. After 10 min of baseline imaging, 80 µM of diamide was added to the recording chamber. Images were collected until cell death occurred.

qPCR Analysis

Left ventricles were isolated from P10 and P17 Scn1b^−/− and Scn1b^+/+ mice (n = 3 in each group). Taqman assays for Sod1 (Mm01344232_g), Sod2 (Mm01313000_m1), Gpx (Mm00656761_g1), and Gsr (Mm00439154_m1) (Life Technologies) were used to detect the expression of select genes among the resultant cDNA.

Enzyme Activity Assays

Gsr and Gpx activity in hearts were measured as we have previously described (20, 22). Citrate synthase was measured using established protocols (25). Left ventricles were isolated from P10 and P17 Scn1b^−/− and Scn1b^+/+ mice (n = 3 in each group).

Western Blot Analysis

Isolated left ventricular sections from P17 Scn1b^−/− (n = 6) and Scn1b^+/+ (n = 5) were homogenized and prepared with the NuPage LDS sample buffer (ThermoFisher). Proteins were separated using NuPage 4–12% Bis-Tris mini protein gels (ThermoFisher) and then transferred to a PVDF membrane using the Trans-Blot Semi-Dry system (Bio-Rad). The membrane was blocked in 4% nonfat dairy milk (NFDM) in TBST (0.1% Tween) at room temperature for 1 h. Following blocking, the primary antibody was given a dilution of 1:500 (GPx, Cat. No. PA5-26323; Invitrogen) or 1:1,000 (GAPDH, Cat. No. PA1-988; Invitrogen) in 4% NFDM in 0.1% TBST at 4°C overnight. The primary antibody was detected using horseradish peroxidase (HRP)-conjugated secondary antibody (Cat. No. 31460; Invitrogen) diluted at 1:5,000 in 4% NFDM in 0.1% TBST followed by the SuperSignal West Pico PLUS chemiluminescent substrate (ThermoFisher). Western blot images were captured using a Chemi-Doc imager (Bio-Rad), and subsequent densiometric analysis was performed using ImageJ (Fiji).

Statistical Analysis

All data are presented as means ± SE. Male and female mice were used in this study. After it was determined that there were no sex differences (two-way ANOVA; sex × genotype), data were pooled. Student’s t test was used for comparisons between the two groups. Between groups, a comparison for incidence of cellular H₂O₂ burst and whole heart VT/VF was determined using Fisher’s exact test. The comparison of cellular survival (Kaplan–Meier curve) was analyzed using a log-rank test. P < 0.05 was considered statistically significant.

RESULTS

INCREASED MORTALITY in Scn1b^−/− Mice

Monitoring of Scn1b^−/− life span revealed mortality is significantly increased compared with wild type (P < 0.001) (Fig. 1). On average, animals died at ∼20.3 days of age (standard deviation ± 1.79 days), with no differences in mortality between male and female Scn1b^−/− mice (P = 0.53) (Fig. 1B). These values are in accordance with previously published studies on life span in Scn1b^−/− mice (26).

Figure 1.

Survival curve of Scn1b^+/+ and Scn1b^−/− mice. A: life span monitoring in Scn1b^+/+ and Scn1b^−/− animals revealed significantly increased mortality in Scn1b^−/− animals (average Scn1b^−/− life span = 20.33 days) (P < 0.001). B: mortality data were not different between Scn1b^−/− mice of the male or female sex (t = 19.75 vs. 20.44; P = 0.53). Log-rank Mantel–Cox (A and B). ****P < 0.0001. Scn1b^+/+ (N = 15) and Scn1b^−/− (N = 15, 7 females and 8 males). N = animal number.

Increased ROS Accumulation in Scn1b^−/− Ventricular Cardiomyocytes

As we previously observed elevated levels of mitochondrial H₂O₂ production in Scn1b^−/− hearts (4), we investigated if this increase may lead to ROS accumulation at the cellular level (Fig. 2). Similar to our previous work (19, 22), we chose to use diamide as an oxidative stressor, along with the fluorescent ROS probe CM-DCF, to determine if ROS accumulation in Scn1b^−/− cells is linked to an impaired ability to scavenge ROS. Diamide primarily targets the glutathione (GSH) scavenging system, which is the major pathway for H₂O₂ reduction in cardiomyocytes (27, 28). After the addition of 80 µM diamide, GSH is oxidized to GSSG, which gradually depletes the GSH:GSSG ratio. This decreases the ability of glutathione peroxidase (Gpx) to reduce H₂O₂, causing ROS accumulation indicated by surges in the DCF signal (27) and cell death. We found that cells isolated from Scn1b^+/+ hearts had significantly improved survival (Fig. 2B) compared with cells isolated from Scn1b^−/− hearts (P = 0.03). The first Scn1b^−/− cells began dying within 1 min after diamide administration, whereas the first death in Scn1b^+/+ cells did not occur until 5.5-min postdiamide. On average, Scn1b^+/+ cells survived ∼4.5 min longer following diamide treatment (14.3 ± 1.5 vs. 9. 6 ± 1.2 min; P = 0.02). In addition, we found that death in Scn1b^−/− cells was more often immediately preceded by sharp increases in the DCF signal (Fig. 2D) compared with Scn1b^+/+ cells (62% vs. 24% of cells; P < 0.0001), indicating that cells from Scn1b^−/− hearts more often experience rapid accumulations of ROS preceding death. These ROS surges increase the likelihood of cellular dysfunction and arrhythmia.

Figure 2.

Response of cardiomyocytes isolated from Scn1b^+/+ and Scn1b^−/− hearts to a prolonged oxidative challenge with 80 µM diamide. A: Kaplan–Meier survival curves comparing survival in Scn1b^+/+ and Scn1b^−/− cells show that cells isolated from Scn1b^−/− animals had significantly improved survival rates compared with cells isolated from Scn1b^+/+ mice (P = 0.03) after diamide treatment. B: average time to cell death in Scn1b^−/− cells (t = 9.56 min) was significantly faster than the average time to cell death in cells from Scn1b^+/+ (t = 14.29 min; P = 0.02) after exposure to diamide. C: death in cardiomyocytes isolated from Scn1b^−/− mice (62%) was more often preceded by sharp increases in the DCF signal (indicating ROS accumulation) compared with cells isolated from Scn1b^+/+ (24%) (P = 0.03). D: representative DCF trace in diamide-treated cells undergoing a ROS burst (black line) or no burst (gray line) before cell death. Log-rank Mantel–Cox (B), t test (C), and Fisher’s exact test (D). *P ≤ 0.05; ****P < 0.0001. Scn1b^+/+ (N = 8, n = 29) and Scn1b^−/− (N = 7, n = 28). N = animal number; n = number of cellular replicates. Animals were taken between P15 and P20 for experimentation. ROS, reactive oxygen species.

Increased Susceptibility to Cardiac Arrhythmias in Isolated Hearts

To determine if our results in isolated cardiomyocytes translate to the whole heart, we next performed experiments to determine if Scn1b^−/− mouse hearts have an increased susceptibility to cardiac arrhythmias during diamide perfusion (Fig. 3). Isolated, denervated, hearts from Scn1b^−/− and Scn1b^+/+ mice were perfused with 80 µM diamide. Arrhythmia susceptibility and severity were assessed for 30 min of perfusion. Scn1b^−/− hearts displayed a significantly increased sensitivity to diamide perfusion. Scn1b^−/− hearts had substantially higher arrhythmia scores (Fig. 3B) (1.3 ± 0.3 vs. 3.4 ± 0.5 for Scn1b^+/+ and Scn1b^−/− respectively; P = 0.002). Overall, hearts from Scn1b^−/− mice experienced a higher incidence of ventricular arrhythmia as defined as observed ventricular tachycardia (VT) and/or fibrillation (VF), regardless of length (Fig. 3C) (25% vs. 86% for Scn1b^+/+ and Scn1b^−/−, respectively; P < 0.0001). In addition, the time spent before the onset of arrhythmias was significantly shorter in Scn1b^−/− hearts (Fig. 3D) (mean survival of 7.25 for Scn1b^−/−; P = 0.01). In an additional set of experiments, hearts from Scn1b^−/− mice were perfused with 5 mM of N-acetyl-l-cysteine (NAC) before the onset of diamide treatment, a glutathione precursor, to determine if glutathione supplementation could rescue the arrhythmogenic phenotype observed in this model. In NAC-perfused Scn1b^−/− hearts, arrhythmia scores were significantly decreased compared with Scn1b^−/− with no pretreatment and were reverted to wild-type values (Fig. 3E) (mean score, 2.33 without NAC v.1 in hearts administered NAC, P = 0.002). Arrhythmia incidence was overall decreased in Scn1b^−/− hearts perfused with NAC, with just one heart undergoing VT and no instances of VF observed (Fig. 3F).

Figure 3.

Effect of 80 µM diamide perfusion on Scn1b^+/+ and Scn1b^−/− hearts. A: ECG signal in a diamide-perfused heart showing the transition to ventricular tachycardia (VT) from sinus rhythm. B: Scn1b^−/− hearts showed significantly higher arrhythmia scores than Scn1b^+/+ hearts after 30 min (P < 0.0001) of diamide perfusion. C: average time to arrhythmia onset (t = 7.35 min) is significantly lower in the hearts of Scn1b^−/− mice than in the hearts of Scn1b^+/+ animals (28.38 min, P = 0.01). D: Scn1b^−/− hearts experience a significantly higher incidence of ventricular fibrillation (VF) and ventricular tachycardia (VT) than hearts from Scn1b^+/+ mice after 30 min of diamide perfusion (25% vs. 86%; P = 0.02). E: Scn1b^−/− hearts pretreated with 5 mM of a glutathione precursor, N-acetyl-l-cysteine, before the onset of diamide perfusion have significantly reduced arrhythmia scores compared with untreated hearts (P = 0.002). F: incidence of VF and VT in Scn1b^–/– hearts with and without NAC pretreatment. t test (A, C, and E) and Fisher’s exact test (B and F). *P ≤ 0.05; **P ≤ 0.01; ****P < 0.0001. Scn1b^+/+ (N = 8) and Scn1b^−/− (N = 6) for (A–D). Scn1b^−/− (N = 6 each, ± NAC) for (E and F). Animals were taken between P15 and P20 for experimentation. NAC, N-acetyl-l-cysteine.

Altered Expression and Activity of Key Antioxidant Genes in Scn1b^−/− Hearts

Our prior results suggest that Scn1b^−/− hearts have increased ROS production and a decreased ability to buffer this accumulating ROS. Hence, we next sought to determine the mechanism underlying the compromised ROS scavenging in Scn1b^−/− hearts. Scn1b^−/− mice begin experiencing severe seizure episodes at approximately P10–P12 that persist throughout their life span, until sudden death around P20 (9). To determine if changing antioxidant profiles during this phase may potentially underlie reduced ROS scavenging and an arrhythmogenic phenotype, we measured expression of the key antioxidant genes glutathione peroxidase (Gpx), glutathione reductase (Gsr), superoxide dismutase 1 (Sod1), and superoxide dismutase 2 (Sod2) at both P10 (before seizure onset) and P17 (after seizure onset) in left ventricular samples (Fig. 4). Our results indicate no differences in the expression of these genes between Scn1b^−/− and Scn1b^+/+ hearts at P10 (P = 0.49 and 0.72). Expression of Gpx was significantly decreased on P17 in Scn1b^−/− hearts compared with age-matched Scn1b^+/+ hearts (P = 0.005). This decrease in Gpx at P17 was also significantly lower than Gpx expression on P10 (0.28-fold; P = 0.01). Gsr showed a similar trend; expression of Gsr was decreased in Scn1b^−/− hearts on P17 compared with age-matched Scn1b^+/+ hearts (0.49-fold; P = 0.043). Interestingly, in Scn1b^+/+ hearts, expression of Sod1 (1.49-fold; P = 0.04) and Sod2 (1.47-fold; P = 0.05) increased with age but did not change in Scn1b^−/− hearts over time (0.07 and 0.09-fold; P = 0.52 and 0.72 for Sod1 and Sod2 respectively). There was a strong, but insignificant trend for decreased expression of Sod1 in Scn1b^−/− hearts on P17 (0.52-fold; P = 0.06) while expression of Sod2 was significantly decreased when comparing Scn1b^−/− to Scn1b^+/+ hearts at P17 (0.49-fold; P = 0.04). Lastly, we examined protein expression of GPx in left ventricular samples, to determine if protein levels were affected similarly to gene expression. Despite the results of our qPCR data indicating decreased expression of the GPx gene, we found that GPx protein content (relative to GAPDH) was significantly increased in Scn1b^−/− hearts relative to Scn1b^+/+ (2.09-fold; P = 0.049).

Figure 4.

Quantification of antioxidant expression profiles in Scn1b^+/+ and Scn1b^−/− hearts pre- and postseizure development. A: superoxide dismutase catalyzes the enzymatic reduction of mitochondrial superoxide to H₂O₂, a reactive compound that is neutralized to H₂O by the glutathione redox couple. B: qPCR analysis of antioxidant gene expression of Sod1 (P = 0.06) and Sod2 (P = 0.04) in Scn1b^−/− hearts. Gpx expression is significantly decreased in Scn1b^−/− hearts compared with Scn1b^+/+ at P17 (P = 0.005) (C), as is the expression of Gsr (P = 0.043) (D). E: expression of the GPx protein is significantly increased in Scn1b^−/− hearts compared with samples from Scn1b^+/+ (P = 0.04). Two-way ANOVA (genotype × age) with Fisher’s LSD post hoc (A–D), t test (E). *P ≤ 0.05, and **P ≤ 0.01. For qPCR experiments, N = 3 mice/genotype + age. Animals were taken at P10 and P17 for experimentation as noted. For evaluation of GPx protein expression, hearts were collected at P17 from Scn1b^−/− (N = 6) and Scn1b^+/+ (N = 5) mice.

Increased Activity of Glutathione Peroxidase in Scn1b^−/− Hearts

We have previously shown that the activity of antioxidant enzymes, such as glutathione reductase, can be altered by posttranslational modifications (21). Therefore, we next established if the decreases in Gsr and Gpx expression we observed correspond to a decrease in enzyme activity as well. We measured the enzymatic action of glutathione peroxidase (GPx) and glutathione reductase (GR) in left ventricular tissue isolated from P10 and P17 Scn1b^−/− and Scn1b^+/+ mice (Fig. 5). Contrary to our qPCR experiments, but similar to our Western blot data, the results of our activity assays indicated that GPx activity is increased significantly at both P10 (10 ± 6 vs. 42 ± 7 mU/mg; P = 0.01) and P17 (15 ± 5 vs. 55 ± 8 mU/mg; P = 0.003) in Scn1b^−/− hearts compared with age-matched Scn1b^+/+. GR activity showed no differences between groups at either time point (P = 0.36 and 0.46 for P10 and P17 respectively).

Figure 5.

Evaluation of glutathione reductase and glutathione peroxidase activity in Scn1b^+/+ and Scn1b^−/− hearts pre- and postseizure development. A: left ventricular samples isolated from P10 and P17 Scn1b^−/− mice have significant increases in glutathione peroxidase (GPx) activity at P10 (P = 0.01) and P17 (P = 0.003). B: activity of glutathione reductase (GR) is increased, but not significantly elevated compared with Scn1b^+/+. Two-way ANOVA (genotype × age) with Fisher’s least significant difference post hoc (A and B). **P ≤ 0.01. n = 3 mice/genotype + age. Animals were taken at P10 and P17 for experimentation.

DISCUSSION

In this study, we investigated if ROS accumulation and deficits in antioxidant buffering may underly arrhythmogenic triggers in Scn1b^−/− mice. This work builds upon our previous investigations showing that loss-of-function mutations in Scn1b lead to increased ROS production by the mitochondrial respiratory chain (4). By investigating a cellular mechanism adjacent to mitochondrial function and ROS production and antioxidant buffering, we isolated a supplementary mechanism that may underly cardiac arrhythmias in Scn1b^−/− hearts. The results of this study indicate that loss of β₁-subunit function in the heart leads to increased ROS accumulation, decreased ability to buffer increases in ROS production, and a higher incidence and severity of cardiac arrhythmias when challenged with an oxidative burden. Overall, these deficits may underlie the mechanisms by which Scn1b^−/− hearts are more susceptible to fatal cardiac arrhythmias. Taken together, our findings suggest that Scn1b may play an important role not just in the regulation of voltage-gated sodium channels but also in cellular responses to oxidative stress. To date, there are little data on the interplay between how the loss of Scn1b function affects the ability to buffer surges in ROS production, underscoring this study’s importance. Targeting sources of oxidative stress in the heart may be a novel therapeutic approach to decrease the risk of cardiac arrhythmias and sudden cardiac death.

A major pathway by which oxidative stress mediates cardiac arrhythmias through is via the mitochondria. During the function of the mitochondrial respiratory chain, H⁺ is translocated from the mitochondrial matrix to the inner membrane space by complexes I, III, and IV. This generates an electrochemical gradient across the inner mitochondrial membrane known as the mitochondrial membrane potential (ΔΨ_m). In the heart, the central ROS buffer is the glutathione system (28). To regenerate reduced glutathione, glutathione reductase catalyzes the reduction of GSSG to GSH using the cofactor NADPH. NADPH is also a cofactor of other major thiol buffer systems in the heart, thioredoxin, and peroxiredoxin. Therefore, in the event of ROS accumulation, NADPH demand increases substantiality in the heart. Once overwhelmed, the heart can be more susceptible to cardiac arrhythmias.

The ΔΨ_m, together with the pH difference across the mitochondrial inner membrane, creates the proton motive force (pmf) that is harnessed by ATP synthase to drive the synthesis of ATP. ATP is required to drive the function of sarcolemmal channels such as K_ATP channels and the Na⁺/K⁺ ATPase, as well as intracellular ion transporters, such as the sarcoendoplasmic Ca²⁺ ATPase (SERCA). Reduced function of these ion-handling pathways may encourage proarrhythmic substrates within the heart (19, 29, 30). Depolarization of the ΔΨ_m decreases ATP production and supports the appearance of metabolic sinks in the myocardium that support arrhythmias (19, 29). Thus, ΔΨ_m is an essential moderator of cardiac electrical activity. Previous work, by us and others, has demonstrated that the ability to stabilize ΔΨ_m is associated with reduced ROS accumulation, cell death, and arrhythmia incidence (19–23).

Our previous work demonstrated that ROS production via the mitochondria is increased in Scn1b^−/− mice (4). We demonstrated that mitochondrial function is compromised in Scn1b^−/− mice, primarily via changes to complex I. This also appears to be the site of increased ROS production. However, we did not investigate if this increase in reactive oxygen species ultimately led to differences at the whole organ level that may predispose mice to cardiac arrhythmias. Our results herein demonstrate that under conditions of oxidative stress, Scn1b^−/− cardiomyocytes are more prone to bursts of ROS accumulation at the cellular level and cardiac arrhythmias. Overall, the increased ROS production and decreased mitochondrial function observed in our previous manuscript, coupled with our results in this study, further emphasize the importance of the glutathione system in detoxifying ROS and preventing arrhythmias in Scn1b^−/− mice. In support of this, by supplementation with the glutathione precursor NAC, we were able to decrease the cardiac arrhythmia burden.

During oxidative stress, ROS may accumulate when there is an imbalance between ROS production and inherent scavenging by antioxidant systems. When the balance favors ROS accumulations, changes in cellular excitability and cell death can occur (19, 21, 23). ROS accumulation has been shown to underlie ischemic tissue damage in multiple cardiac diseases, including heart attack and stroke (31). In accompaniment to increased ROS accumulation, we also found that cardiac tissue from Scn1b null mice display heightened sensitivity to oxidative stress, which may be due to deficits in their glutathione system. We found that diamide-treated cells isolated from Scn1b^−/− mice die significantly faster and more often experience a surge in ROS accumulation. Previous research has demonstrated that once the glutathione system is stressed to criticality, ROS rapidly accumulate, inducing oscillations in ΔΨ_m and irreversible collapses in the mitochondrial network (19, 27). Furthermore, it has been suggested that the onset of arrhythmia coincides with the point where ROS scavenging by the glutathione system is depleted (22). Consequently, given our observations of reduced ROS buffering capacity in isolated cardiomyocytes from Scn1b^−/− hearts, it is unsurprising that our experiments in whole hearts perfused with diamide indicated that Scn1b^−/− hearts experience increased susceptibility to and severity of arrhythmias.

Upon experiencing an oxidative challenge, Scn1b^−/− hearts were more prone to potentially fatal arrhythmias and had a shorter time to arrhythmia onset. These results suggest that the glutathione system is less resistant to oxidative stress in this model, hence insufficient antioxidant buffering may be involved in arrhythmia pathogenesis in hearts lacking substantial Scn1b. Our results also indicated there may be imbalances in antioxidant enzyme activity in Scn1b^−/− hearts, as we found GPx protein expression and activity are increased significantly, despite no changes in Gsr catalysis. Furthermore, as Scn1b^−/− animals show increased mitochondrial ROS production (4), non-thiol-sensitive pathways may act as compensatory when challenged by an oxidized environment and is more prone to arrhythmogenic attacks. Defects in mitochondria commonly coincide with glutathione depletion, as excessive ROS generated by a dysfunctional electron transport chain is buffered at the expense of intracellular antioxidant stores. Besides its influence on cardiac electromechanical function, persistent decreases in cardiac glutathione levels associated with loss of Scn1b may have broader implications on cardiovascular wellness. Decreased glutathione buffering allows ROS to accumulate intracellularly, where they can react with various lipids and proteins in the cell, as well as damage electron transport chain components (32) and support apoptotic and necrotic death of cardiomyocytes (33). Our previous work has shown that glutathione depletion does not only support arrhythmogenesis but also exacerbates the effects of ischemia-reperfusion injury on the heart (19, 22).

Interestingly, our results demonstrated there is a disconnect between GPx gene expression compared with protein abundance and enzyme activity. The reduced resistance to arrhythmia in Scn1b hearts could be due to post-translational modifications of glutathione system enzymes. In another model, we have previously demonstrated that redox-dependent changes to the glutathione system as a whole have the potential to be protective from arrhythmia susceptibility (21) and redox-sensitive post-translational modification in GR can occur (21). Discrepancies between expression and activity profiles of glutathione peroxidase may be explained by post-translational modifications that affect degradation of the protein. It has been suggested that under conditions of oxidative stress, GPx can undergo carbonylation which may decrease degradation and reinforce the enzymatic reaction of GPx (34). Others have demonstrated that cardiac GPx activity is stimulated by phosphorylation in response to electrophilic stress (35) and phosphorylation may decrease the degradation of GPx (36). In addition, as Scn1b^−/− animals show increased mitochondrial ROS production (4), non-thiol-sensitive pathways may act as compensatory when challenged by an oxidized environment and are more prone to arrhythmogenic attacks.

Overall, in this study, we showed that enhanced susceptibility to oxidative stress may be underlying molecular mechanisms that contribute to an arrhythmogenic phenotype in Scn1b null hearts. In addition, altered physiological profiles of key antioxidant genes may contribute to metabolic alterations in Scn1b^−/− mouse hearts. The activity of glutathione reductase and glutathione peroxidase is critical to maintaining cardiac antioxidant buffering, as there is minimal new synthesis of glutathione in the heart (37). Our results show that in Scn1b null hearts, there is a significant decrease in Gpx gene expression, but protein levels and activity of GPx are greatly upshifted. Furthermore, we detected significant decreases in the expression of Sod1, an essential ROS-scavenging enzyme that has been implicated in protecting against ischemic injury in the heart (38). Our observations of altered antioxidant expression profiles in Scn1b^−/− mice are unsurprising, given the functional role of β₁-subunits in gene expression, and the lack of functional β1-mediated signaling in Scn1b^−/− mice. β₁-Subunits undergo regulated proteolysis, generating a soluble intracellular domain that translocates to the nucleus and regulates gene expression (17). In Scn1b^−/− mice, the absence of β1-mediated signaling is associated with epileptic phenotypes in the brain (39) and significant shifts in gene expression profiles in the heart (11).

Conclusions

In this study, we investigated if compromised reactive oxygen species scavenging may underlie arrhythmia susceptibility in the Scn1b^−/− mouse model of DS. Pathogenic loss-of-function of Scn1b is associated with electromechanical dysfunction in a variety of diseases with cardiac manifestations. In Scn1b^−/− mice, mitochondrial H₂O₂ production is increased and may overwhelm cardiac ROS-scavenging defenses. The imbalance between ROS production and scavenging may increase the potential for increased arrhythmia susceptibility and adverse cardiac outcomes. Our findings evidenced that when subject to an oxidative challenge with the glutathione oxidant diamide, cell survival decreased significantly, and there was an increased severity and susceptibility to arrhythmia in Scn1b^−/− mice. This indicated severe deficits in thiol-based antioxidant buffering systems in Scn1b^−/− hearts. Altered expression and activity patterns of antioxidant enzymes, namely, glutathione peroxidase, may underlie shortfalls in the glutathione system of Scn1b^−/− mice. As glutathione redox status and cellular ROS levels are linked to ATP production through ΔΨ_m, cell energetics are influenced by its redox status. Decreased ATP generation may negatively affect cardiac conduction, further supporting arrhythmogenic events. More research is needed to elucidate further the relationships between altered cardiac bioenergetic parameters, responses to oxidative stress, and increased prevalence of dangerous arrhythmias in Scn1b^−/− mice. Targeting these mechanisms may be a potential therapeutic approach to reduce the incidence of adverse cardiac events associated with decreased Scn1b function.

DATA AVAILABILITY

Data will be made available upon reasonable request.

GRANTS

This work was supported by National Institute of Neurological Disorders and Stroke Grant R21NS116647 and a grant from the Research Development Committee at East Tennessee State University (to C.R.F.).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

J.L.A. and C.R.F. conceived and designed research; J.L.A., E.D.A., A.A.F., and C.R.F. performed experiments; J.L.A. and C.R.F. analyzed data; J.L.A. and C.R.F. interpreted results of experiments; J.L.A. and C.R.F. prepared figures; J.L.A. and C.R.F. drafted manuscript; J.L.A., E.D.A., A.A.F., and C.R.F. edited and revised manuscript; J.L.A., E.D.A., A.A.F., and C.R.F. approved final version of manuscript.