Elevated EZH2 in Ischemic Heart Disease Epigenetically Mediates Suppression of Na_V1.5 Expression

Abstract

Suppression of the cardiac sodium channel Na_V1.5 leads to fatal arrhythmias in ischemic heart disease (IHD). However, the transcriptional regulation of Na_V1.5 in cardiac ischemia is still unclear. Our studies are aimed to investigate the expression of enhancer of zeste homolog 2 (EZH2) in IHD and regulation of cardiac Na_V1.5 expression by EZH2. Human heart tissue was obtained from IHD and non-failing heart (NFH) patients; mouse heart tissue was obtained from the peri-infarct zone of hearts with myocardial infarction (MI) and hearts with a sham procedure. Protein and mRNA expression were measured by immunoblotting, immunostaining, and qRT-PCR. Protein-DNA binding and promoter activity were analyzed by ChIP-qPCR and luciferase assays, respectively. Na⁺ channel activity was assessed by whole-cell patch clamp recordings. EZH2 and H3K27me3 were increased while Na_V1.5 expression was reduced in IHD hearts and in mouse MI hearts compared to the controls. Reduced Na_V1.5 and increased EZH2 mRNA levels were observed in mouse MI hearts. A selective EZH2 inhibitor, GSK126 decreased H3K27me3 and elevated Na_V1.5 in HL-1 cells. Silencing of EZH2 expression decreased H3K27me3 and increased Na_V1.5 in these cells. EZH2 and H3K27me3 were enriched in the promoter regions of Scn5a and were decreased by treatment with EZH2 siRNA. GSK126 inhibited the enrichment of H3K27me3 in the Scn5a promoter and enhanced Scn5a transcriptional activity. GSK126 significantly increased Na⁺ channel activity. Taken together, EZH2 is increased in ischemic hearts and epigenetically suppresses Scn5a transcription by H3K27me3, leading to decreased Na_V1.5 expression and Na⁺ channel activity underlying the pathogenesis of arrhythmias.

1. Introduction

Ischemic heart disease (IHD) is a leading cause of fatal cardiac arrhythmias including ventricular tachycardia (VT) and ventricular fibrillation (VF)[1]. Patients with severely depressed left ventricular ejection fraction are at a high risk of developing VT/VF and sudden cardiac death (SCD)[2]. In the ischemic myocardium, altered expression and function of ion channels affect cardiac electrical excitability, conduction, and automaticity[3]. In infarcted canine hearts, Na⁺ current density is significantly decreased in the epicardial border zone, leading to slow conduction velocity[4]. Encoded by SCN5A, the pore-forming Na_V1.5 α-subunit of the cardiac Na⁺ channel governs cardiac depolarization[5]. Loss-of-function of Na_V1.5 is associated with arrhythmogenicity in patients with cardiac ischemia and end-stage heart failure[6,7]. The decrease in Na⁺ current density is caused by reduced Na_V1.5 protein levels on the cell membrane and/or altered Na⁺ channel kinetics[4]. The decrease of Na_V1.5 expression in ischemic myocardium from mice with myocardial infarction (MI) suggests that the repression of Scn5a gene transcription contributes to the downregulation of Na_V1.5 expression, suppression of Na⁺ channel activity, and cardiac arrhythmias[8].

Transcription factors, including nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB), T-box transcription factor 5 (TBX5), and forkhead box protein O1 (FoxO1), are crucial regulators of Na_V1.5 expression[8–11]. We previously showed that enhanced β-catenin/T-cell factor 4 (TCF-4) signaling transcriptionally suppresses Na_V1.5 expression[12–14] and increases arrhythmogenicity in mice[13]. High-throughput studies have revealed an essential role of epigenetic modulation, including DNA methylation and histone post-translational modification (PTM), underlying cardiovascular homeostasis[15]. Enhancer of zeste homolog 2 (EZH2), the functional enzymatic component of polycomb repressive complex 2 (PRC2), promotes heterochromatin formation via catalyzing trimethylation of histone 3 lysine 27 (H3K27me3), which leads to transcriptional suppression, serving as a potential barrier for cardiac regeneration[16,17]. In ischemic cardiomyopathy, EZH2 is elevated and regulates cardiac gene expression[18]. EZH2 functions as a methyltransferase to mediate Lamin A/C mutation-induced suppression of SCN5A transcription in human cardiac stem cells[19]. Interestingly, EZH2 interacts with β-catenin/TCF-4 to suppress target gene expressions, thereby regulating endothelial cell function[20]. Whether EZH2 epigenetically regulates Na⁺ channel expression in IHD warrants further investigation.

In this study, we report that EZH2 expression is increased in human IHD and mouse MI hearts. This upregulation correlates with increased H3K27me3 and decreased Na_V1.5 expression. Inhibition of EZH2 methyltransferase activity or knockdown of EZH2 increases Na_V1.5 expression at mRNA and protein levels in HL-1 cardiomyocytes. Also, we found that EZH2 is enriched in the promoter of Sen5a and is associated with a high level of H3K27me3. Both EZH2 expression and function are required to suppress the promoter activity of Scn5a. Moreover, GSK126, a selective EZH2 inhibitor, significantly increases Na⁺ channel activity in HL-1 cardiomyocytes. These findings demonstrate the indispensable role of EZH2 in the epigenetic downregulation of Na_V1.5, highlighting EZH2 inhibition as a potential antiarrhythmic therapy in IHD.

2. Materials and methods

2.1. Experimental protocol

Human heart tissue samples were composed of failing hearts with end-stage IHD hearts (n=3) and non-failing hearts (n=3). These samples were kindly offered by Dr. Shin Lin (Division of Cardiology, Department of Medicine, University of Washington). We obtained the institutional review board (IRB) approval at the University of Washington at Seattle. The patients’ information was de-identified. Mice were handled according to the National Institutes of Health’s Guide for the Care and Use of Laboratory Animals. Experimental procedures were approved by the Institutional Animal Care and Use Committee of the University of Washington. Detailed materials and methods in this study are included in the Online Data Supplement; the primers used in the studies are summarized in Supplemental Table 1.

2.2. Statistical analysis

Data are presented as mean±SEM (standard error of the mean). The unpaired 2-tailed Student’s t-test was used to analyze the difference between two groups; one-way ANOVA followed by post hoc Tukey’s test was used to compare the data from multiple groups. A value of P<0.05 was considered statistically significant. Statistical analysis was performed using Prism software (Graphpad, version 7.0)

3. Results

3.1. Na_V1.5 expression is decreased while EZH2 expression and H3K27me3 levels are increased in human hearts with IHD and mouse hearts with MI

We first measured the expression of EZH2, H3K27me3, and Na_V1.5 in the hearts from patients with end-stage heart failure secondary to IHD and in human samples from non-failing hearts (NFHs). Consistent with previous reports[18], we found that the protein levels of EZH2 were increased in IHD hearts compared to NFHs. In IHD hearts, we also found increased H3K27me3 levels, a repressive chromatin mark of genes. Histone H3 expression was similar between IHD and NFH groups. We detected a significant decrease of Na_V1.5 protein levels in IHD hearts, which was negatively correlated with EZH2 and H3K27me3 (Fig. 1A and B). One week after the ligation of the left anterior descending coronary artery (LAD) in MI mice, the peri-infarct zone (PIZ, 2.0 mm surrounding the infarcted zone) and the corresponding ventricular myocardium region of the sham-operated hearts were sampled for experiments. Western blotting demonstrated increased EZH2 and H3K27me3, and unchanged Histone H3 with decreased Na_V1.5 protein in the PIZ compared to the sham control (Fig. 1C–D). Increased EZH2 and decreased Na_V1.5 mRNA levels were observed in the PIZ (Fig. 1E). These results indicated that enhanced EZH2 expression and H3K27me3 levels are correlated with the downregulation of Na_V1.5 in ischemic hearts.

Fig. 1. EZH2 expression and H3K27me3 levels are increased while Na_V1.5 is decreased in the myocardium from patients with ischemic heart disease (IHD) and from the peri-infarct zone (PIZ) of mice with myocardial infarction (MI).

(A) Western blotting of the protein extract from the left ventricular myocardium of patients with IHD and those with non-failing hearts (NFHs) were collected. (B) Immunoblotted bands were normalized to GAPDH (n=3 for NFH and n=3 for IHD). *P<0.05, **P<0.01. (C) Western blots were performed on the protein extracts from the PIZs of mouse MI hearts and the areas of mouse hearts with sham procedure using antibodies against Na_V1.5, EZH2, H3K27me3, Histone H3 and GAPDH. (D) Quantification of immunoblotted bands by densitometry normalized to GAPDH (n=3 for sham and n=3 for MI). (E) Quantitative real-time qRT-PCR analysis of Na_V1.5 and EZH2 mRNA normalized to GAPDH mRNA of mouse PIZ and sham (n=4 for sham; n=3 for MI). Columns are mean±SEM and data were analyzed by Student’s t-test. *P<0.05, **P<0.01 (B, D and E).

3.2. Inhibition of EZH2 activity or decrease of EZH2 expression reduces H3K27me3 and increases expression of Na_V1.5 in HL-1 cardiomyocytes

To understand whether the inhibition of EZH2 catalytic activity or decrease of EZH2 expression affects Na_V1.5 expression, we first inhibited the activity of EZH2 histone methyltransferase using GSK126, a selective EZH2 inhibitor[21]. Inhibition of EZH2 was confirmed by a concentration-dependent decrease of H3K27me3 levels and unchanged histone H3 in HL-1 cells 72 hours (hrs) after treatment with 1 μM and 3 μM GSK126 (Fig. 2A and B). We found that protein levels of Na_V1.5 were significantly elevated in HL-1 cardiomyocytes treated with GSK126 compared to cells with vehicle control (DMSO) (Fig. 2A and B). GSK126 produced a remarkable upregulation of Na_V1.5 mRNA expression in HL-1 cells (Fig. 2C). The protein levels of EZH2 were not significantly different between DMSO- and GSK126-treated HL-1 cells (Fig. 2B). Immunofluorescence staining was performed on the HL-1 cells treated with 1 μM GSK126 using antibodies against Na_V1.5, EZH2 and H3K27me3. Decreased nuclear H3K27me3 in HL-1 cardiomyocytes treated with GSK126 was observed without alteration in EZH2 nuclear levels (Fig. 3A and C). Fluorescence intensity quantification using imageJ (National Institutes of Health, NIH) showed that GSK126 significantly increased Na_V1.5 signal (Fig. 3B); GSK126 significantly increased nuclear H3K27me3 fluorescence signal without altering EZH2 staining intensity (Fig. 3B and D). Therefore, the methyltransferase activity of EZH2 is indispensable for the suppression of Na_V1.5.

Fig. 2. Inhibition of EZH2 methyltransferase activity by GSK126 reduces H3K27me3 level and increases Na_V1.5 expression in HL-1 cardiomyocytes.

(A) Western blots were performed on the protein extract from HL-1 cardiomyocytes treated with the EZH2 inhibitor GSK126 (1 μM or 3 μM) or DMSO (vehicle) for 72 hours and the membranes were immunoblotted using antibodies against Na_V1.5, EZH2, H3K27me3, Histone H3 and GAPDH. (B) Quantification of immunoblotted bands by densitometry normalized to GAPDH (n=3 for DMSO, n=3 for GSK126 1 μM and n=3 for GSK126 3 μM). (C) Quantitative real-time RT-PCR analysis of Na_V1.5 mRNA normalized to GAPDH mRNA (n=10 for DMSO, n=10 for GSK126 1 μM and n=7 for GSK126 3 μM). Columns are mean±SEM and data were analyzed by one-way ANOVA with Tukey’s post hoc comparison test. *P<0.05, **P<0.01 (B and C).

Fig. 3. Inhibition of EZH2 methyltransferase activity by GSK126 increases the expression of Na_V1.5 and reduces the nuclear enrichment of H3K27me3 in HL-1 cardiomyocytes.

(A) Representative immunofluorescence images of Na_V1.5 (green), EZH2 (red), and 4′,6-diamidino-2-phenylindole (Hoechst 33342, blue) staining in HL-1 cardiomyocytes treated with GSK126 (1 μM) or DMSO for 72 hrs. (B) Quantitation of global Na_V1.5 and nuclear EZH2 fluorescence intensity in GSK126 group normalized to DMSO’s in HL-1 cells using imageJ (NIH) (n=7 cells for DMSO and n=6 cells for GSK126). (C) Representative immunofluorescence images of H3K27me3 (green), EZH2 (red), and Hoechst 33342 (blue) staining in HL-1 cells treated with GSK126 (1 μM) or DMSO. (D) Quantitation of nuclear H3K27me3 and EZH2 fluorescence intensity normalized to DMSO’s in HL-1 cells (n=9 HPFs [high-power fields] for DMSO and n=10 HPFs for GSK126). Columns are mean±SEM and data were analyzed by Student’s t-test (B and D). *P<0.05, **P<0.01 (B and D).

To explore whether a decrease of EZH2 expression correlates with the upregulation of Na_V1.5, we utilized siRNA to knockdown EZH2 in HL-1 cardiomyocytes. After 48 hrs transfection of siRNA, there was a significant reduction in EZH2 protein in HL-1 cells. Elevated Na_V1.5 expression was observed and was accompanied by reduced H3K27me3 levels and unchanged histone H3 in HL-1 cells transfected with siRNA targeting EZH2, compared to those treated with the scrambled siRNA (SC siRNA) (Fig. 4A and B). Real time qRT-PCR demonstrated that EZH2 siRNA significantly reduced the mRNA levels of EZH2 but increased that of Na_V1.5 (Fig. 4C). Immunofluorescence staining was performed on the HL-1 cells treated with SC siRNA and EZH2 siRNA for 48 hrs using antibodies against Na_V1.5, EZH2 and H3K27me3. Images showed that EZH2 siRNA had an effect similar to GSK126 on Na_V1.5 and H3K27me3 signals (Fig. 5A and C) and decreased nuclear EZH2 staining intensity (Fig. 5A and C). ImageJ (NIH) analyses showed that decreased EZH2 by siRNA (Fig. 5B and D) significantly increased Na_V1.5 signal (Fig. 5B) and decreased H3K27me3 fluorescence intensity (Fig. 5D) compared to the data obtained from SC siRNA. The results showed that both expression and methyltransferase activity of EZH2 play a pivotal role in the regulation of Na_V1.5 expression.

Fig. 4. Knockdown of EZH2 by siRNA reduces H3K27me3 levels and increases Na_V1.5 expression in HL-1 cardiomyocytes.

(A) Western blots were performed on the protein extract from HL-1 cardiomyocytes treated with the EZH2 siRNA (30 nM) or the scrambled (SC) siRNA for 48 hours, and the membranes were immunoblotted using antibodies against Na_V1.5, EZH2, H3K27me3, Histone H3 and GAPDH. (B) Quantification of immunoblotted bands by densitometry normalized to GAPDH (n=3 for SC siRNA and n=3 for EZH2 siRNA). (C) Quantitative real-time qRT-PCR analysis of Na_V1.5 mRNA normalized to GAPDH mRNA (n=6 for SC siRNA and n=6 for EZH2 siRNA). Columns are mean±SEM and data were analyzed by the Student’s t-test (B and C). **P<0.01 (B and C).

Fig. 5. Knockdown of EZH2 by siRNA increases the expression of Na_V1.5 and reduces the nuclear enrichment of H3K27me3 in HL-1 cardiomyocytes.

(A) Representative immunofluorescence images of Na_V1.5 (green), EZH2 (red), and Hoechst 33342 (blue) staining in HL-1 cardiomyocytes treated with EZH2 siRNA (30 nM) or the SC siRNA. (B) Quantitation of global Na_V1.5 and nuclear EZH2 fluorescence intensity normalized to DMSO’s in HL-1 cells (n=8 cells for SC siRNA; n=9 cells for EZH2 siRNA). *P<0.05, **P<0.01. (C) Representative immunofluorescence images of H3K27me3 (green), EZH2 (red), and Hoechst 33342 (blue) staining in HL-1 cells treated with EZH2 siRNA or SC siRNA. (D) Quantitation of nuclear H3K27me3 (n=6 cells for EZH2 siRNA, n=6 cells For SC siRNA) and EZH2 fluorescence intensity normalized to DMSO’s in HL-1 cells (n=8 HPFs for EZH2 siRNA, n=8 HPFs for SC siRNA). Columns are mean±SEM and data were analyzed by Student’s t-test (B and D). *P<0.05, **P<0.01 (B and D).

3.3. EZH2 is recruited in the promoter of Scn5a and inhibits the promoter activity by increasing H3K27me3

We have shown that EZH2 expression is negatively correlated with Na_V1.5 expression in ischemic hearts and HL-1 cardiomyocytes. These findings raise the possibility of EZH2 regulating the Scn5a promoter through H3K27me3. We screened for potential binding sites and confirmed the enrichment of EZH2 or H3K27me3 at conserved regions in the promoter region of human SCN5A and mouse Scn5a (Fig. 6A and B)[16,22,23]. Next, we determined the recruitment of EZH2 to the Scn5a promoter in HL-1 cardiomyocytes. ChIP-qPCR demonstrated the enrichment of EZH2 in the mouse Scn5a promoter region. A concomitant enrichment of H3K27me3 was detected and significantly reduced by the EZH2 inhibitor, GSK126, which did not alter the enrichment of EZH2 (Fig. 6C). Silencing of EZH2 expression by siRNA reduced the enrichment of both EZH2 and H3K27me3 in the Scn5a promoter region in HL-1 cells (Fig. 6D). A luciferase assay showed that inhibited EZH2 activity by GSK126 or decreased EZH2 expression by siRNA significantly increased the Scn5a promoter activity in HL-1 cells (Fig. 6E). These results suggest that EZH2 is a negative controller of Scn5a transcription.

Fig. 6. Enrichment of EZH2 and H3K27me3 in mouse Scn5a and human SCN5A promoter region and their regulation of Scn5a promoter activity in HL-1 cardiomyocytes.

(A) The enrichment of EZH2 and H3K27me3 (blue bars) in mouse cardiac Scn5a (light yellow box) promoter region was obtained by data mining using the Cistrome Data Browser and WashU Epigenome Browser. The CpG island (orange box) and gene conservation with human SCN5A (violet bands) are displayed. The original datasets are GSM1902476 and GSM742287. (B) The enrichment of H3K27me3 (blue bars) in mouse Scn5a (light yellow box) promoter region was obtained by data mining using the Cistrome Data Browser and WashU Epigenome Browser. The CpG island (orange box) and gene conservation with human cardiac SCN5A (violet bands) are displayed. The original dataset is GSM621450. (C) ChIP-qPCR analysis of EZH2 and H3K27me3 in the promoter region of Scn5a in HL-1 cells treated with GSK126 (1 μM) or DMSO (n=3 for IgG, n=3 for DMSO and n=3 for GSK126). **P<0.01. (D) ChIP-qPCR analysis of EZH2 and H3K27me3 in the promoter region of Scn5a in HL-1 cells treated with EZH2 siRNA (30 nM) or SC siRNA (30 nM) (n=3 for IgG, n=3 for SC siRNA and n=3 for EZH2 siRNA). (E) Transcriptional activity of Scn5a promoter analyzed by the luciferase assay in HL-1 cells treated with GSK126 (1 μM) (n=5 for DMSO and n=6 for GSK126) or EZH2 siRNA (n=5 for SC siRNA and n=5 for EZH2 siRNA). Columns are mean±SEM and data were compared using Student’s t-test (C, D and E). *P<0.05, **P<0.01 (C, D, and E).

3.4. Inhibition of EZH2 activity increases Na⁺ channel activity in HL-1 cardiomyocytes

Whole-cell Na⁺ currents were recorded from HL-1 cells after treatment with GSK126 (1 μM) or DMSO (Fig. 7A) for 48 hrs. The values of cell capacitances were not significantly different between cells with DMSO treatment (34.9±1.9, n=14) and cells with GSK126 treatment (37.2±3.8, n=12). The current-voltage curve illustrated that inhibiting EZH2 activity by GSK126 markedly augmented the peak Na⁺ current densities at depolarization voltages from −45 mV to +35 mV (Fig. 7B). We did not detect a significant impact of GSK126 on Na⁺ channel kinetics (Fig. 8A–G and Supplemental Table 2). Taken together, an EZH2 inhibitor increases Na⁺ channel activity without affecting the channel’s kinetics.

Fig. 7. EZH2 inhibitor, GSK126 increases Na⁺ channel activity in HL-1 cardiomyocytes.

(A) Characteristic voltage-gated Na⁺ currents recorded at depolarization voltage levels from −80 mV to +50 mV using a whole-cell patch clamp in HL-1 cells treated with GSK126 (1 μM) or the vehicle (DMSO) for 72 hrs. (B) I-V curves of peak Na⁺ current densities obtained from the peak currents divided by individual capacitances in HL-1 cells (black: DMSO, n=13 and blue: GSK126, n=11). Peak current densities are mean±SEM and data were analyzed by Student’s t-test. *P<0.05, **P<0.01.

Fig. 8. Na⁺ channel kinetics is not changed by EZH2 inhibitor, GSK126 in HL-1 cardiomyocytes.

(A and B) Time constants of activation and inactivation wereobtained from single exponential fit on the individual current traces at the voltages from −40 mV to −10 mV, and they were not significantly different between DMSO (n = 13) and GSK126 (1 μM) (n = 11). (C, D and E). Voltage-dependent steady-state activation (C, DMSO, n=13 and GSK126, n=11) and inactivation (D and E, DMSO, n=9 and GSK126, n=11) of Na⁺ channel were not changed by GSK126, and V_1/2 and k slope factor were not significantly different as described in table 1 in the supplemental data. (F) Typical Na⁺ current traces of recovery from inactivation at voltage levels from −120 mV to −40 mV in HL-1 cells treated with DMSO (black) and 1 μM GSK126 (blue) for 72 h, respectively. (G) Normalized recovery data for peak I_Na were plotted, and they were well described by a single exponential function. T and k slope were not significantly different between DMSO (n=17) and GSK126 (n=12) groups as described in table 1 in the supplemental data.

4. Discussion

In this study, we showed that elevated EZH2 expression and H3K27me3 levels were associated with decreased Na_V1.5 expression in human IHD and mouse MI hearts. Inhibiting EZH2 activity or decreasing EZH2 expression increased Na_V1.5 expression in HL-1 cardiomyocytes. EZH2 and H3K27me3 were enriched in the Scn5a promoter region, suppressing the promoter activity of Scn5a. GSK126 significantly increased the Na⁺ channel activity in HL-1 cells. The findings indicated that EZH2 mediated suppression of Na_V1.5 channel underlies cardiac arrhythmogenesis in patients with IHD.

We found that EZH2 expression was increased in the myocardium of IHD as reported by another group[18]. How EZH2 expression is regulated in ischemic hearts is unknown. EZH2 expression is regulated by various oncogenic transcription factors such as Myc and E2F and tumor suppressor miRNAs[24]. Recent work from Guo A. et al demonstrated that cardiac stress-induced N-terminal truncate of excitation-contraction (E-C) coupling protein junctoophilin-2 (JP2NT) functions as a transcription factor leading to cardioprotective transcriptional reprograming and specifically, it attenuates hypertrophic remodeling and the progression of heart failure. Overexpression of JP2NT significantly caused modulation of of multiple genes in adult mouse cardiomyocytes and among them. EZH2 expression was significantly decreased, indicating that JP2NT may regulate E2H2 transcription[25]. Further experiments are warranted to determine if transcription factors such as JP2NT and c-Myc mediate EZH2-suppressing Na_V1.5 expression in the myocardium of IHD.

A recent study combining genome-wide methylation and RNA-seq analysis revealed that enhanced EZH2 epigenetically inhibits KLF15 expression by coordinating with differential DNA methylation in its promoter, contributing to myocardial metabolic reprogramming in human ischemic cardiomyopathy[18]. The emerging role of EZH2 led us to explore its potential effect on the epigenetic control of cardiac expression of Na_V1.5, the downregulation of which predisposes to cardiac arrhythmias. We detected an increase in H3K27me3, an established marker of gene suppression[16]. In-silico promoter analysis unveiled the enrichment of EZH2 and H3K27me3 in the proximal region of human SCN5A and mouse Scn5a promoters [16,22]. Na_V1.5 expression was decreased in human IHD. The increase of Na_V1.5 expression and Na⁺ current density by inhibiting EZH2 suggested that histone methylation by EZH2 contributes to the suppression of Scn5a transcription. The dynamic control of SCN5A promoter activity by EZH2 may dictate the change of Na_V1.5 and the development of arrhythmias in IHD.

We found that enhanced β-catenin/TCF-4 signaling transcriptionally suppress expression of Na_V1.5 [12–14]. Interestingly, β-catenin contributes to the PRC2 complex to enable EZH2 binding to promoter regions of claudin-5. PTP and VWF thereby, affecting endothelial cell differentiation and vessel maturation[20]. Also. β-catenin/TCF-4 interacts with EZH2 to affect glioma tumorigenesis as well as survival and metastasis in chemoresistant triple-negative breast cancer[26,27]. These findings raise the possibility that β-catenin/TCF4 and EZH2 interplay to transcriptionally suppress Na_V1.5 expression. In future research plans, we will use mouse models with cardiac-specific loss of EZH2 expression, gain of β-catenin function and/or heart infarction to obtain direct evidence.

Genetic analysis of the human SCN5A promoter revealed mutations and polymorphisms associated with various arrhythmia phenotypes[28]. Changes in epigenetic modulation, as well as transcription factor interaction, may contribute to the genetic disorders. A remote SNP rs6801957 in a cardiac enhancer that spatially interacts with the promoter of SCN5A is associated with reduced Na_V1.5 expression and slowed conduction in Brugada syndrome. Decreased methylation of the SCN5A promoter has been associated with increased Na_V1.5 providing a rescuing phenotype in the probands[29]. One study showed that a K219T-Lamin A/C mutation has been suggested to recruit PRC2 to the SCN5A promoter and inhibits Na_V1.5 expression in a laminopathy[19]. By direct interaction with DNA methyltransferase, enriched EZH2 during cardiac ischemia may mediate the methylation of unmasked CpG island at the SCN5A promoter region, which is spatially overlapped with EZH2 enriched sites in the genome [16]. Further studies are necessary to determine the comprehensive role of EZH2 in regulating SCN5A promoter activity.

Compared to cancer studies, a much lower concentration of GSK126 was sufficient to increase Scn5a promoter activity, Na_V1.5 expression, and Na⁺ channel function in our experiments. GSK126 was well tolerated by participants in the Phase I study[30]. Our findings raised a potential therapeutic implication of GSK126 for cardiac arrhythmias. GSK126 may be considered as an adjunct therapy in IHD patients, especially in certain scenarios when β-blockers and other antiarrhythmic medicine are contraindicated[31]. Interestingly, late embryonic deletion of EZH2 conveyed a mild effect in cardiac structure and function, indicating its safety as a therapeutic target in developed heart[16]. Our observation showed only a low level of EZH2 expression in the NFHs. These findings suggest that EZH2 as a feasible antiarrhythmic target by GSK126 in IHD. The role of EZH2 as a candidate antiarrhythmic target will be investigated in future studies using larger animal MI models.

Our studies indicate that suppression of SCN5A transcription is an important mechanism of decreased cardiac Na⁺ channel activity in ischemic hearts. However, mRNA splicing variants should not be neglected. In the human ischemic cardiomyopathy, splicing factors LUC7L3 and RBM25 are elevated[32] and they cause SCN5A mRNA splicing variants[32,33]. The two SCN5A mRNA variants reach greater that >50% of the total SCN5A mRNA and do not produce functional Na⁺ channels[32,33]; in fact, they function in a dominant negative manner[32,33]. Therefore, both decreased transcriptional regulation and mRNA splicing variants of SCN5A contribute to the decrease of Na⁺ channel activity in ischemic cardiomyopathy.

5. Conclusions

Here we show that EZH2 is increased in IHD and plays an important role in the regulation of cardiac Na_V1.5 expression and Na⁺ channel function. Our findings may shed light on the epigenetic regulation of Na_V1.5 expression underlying the development of arrhythmias related to cardiac ischemia, thereby providing a new candidate molecular target for antiarrhythmic therapy.

Highlights.

• EZH2 and H3K27me3 are increased while Na_V1.5 expression is reduced in ischemic hearts.

• Silencing expression or suppressing activity of EZH2 leads to a decrease of H3K27me3 and an elevation of Na_V1.5.

• Suppressing activity of EZH2 enhances Scn5a promoter activity by decreasing H3K27me3 in the Scn5a promoter.

• Suppressing activity of EZH2 increases Na⁺ channel activity.

• EZH2-mediated suppression of Na_V1.5 is one of the mechanisms underlying arrhythmias in patients with IHD.