Atrial cardiomyopathy resulting from loss of plakophilin-2 expression: Response to adrenergic stimulation and implications for the exercise response

Abstract

Atrial arrhythmias occur in 20-40% of patients with arrhythmogenic right ventricular cardiomyopathy (ARVC) and are associated with an increased risk of sustained ventricular arrhythmias and inappropriate implantable cardioverter-defibrillator shocks. The pathophysiology of atrial arrhythmias in ARVC remains unclear. Most cases of gene-positive ARVC are linked to pathogenic variants in the desmosomal gene plakophilin-2 (PKP2). Here, we test the hypothesis that loss of PKP2 expression leads to pro-arrhythmic changes in atrial cardiomyocytes. Atrial cells/tissue were obtained from a cardiac-specific, tamoxifen-activated model of PKP2 deficiency (PKP2cKO). By contrast to PKP2cKO ventricular myocytes, PKP2cKO atrial cardiomyocytes presented no significant differences in intracellular calcium (Ca²⁺ _i) transient dynamics, sarcoplasmic reticulum load or action potential morphology. PKP2cKO atrial cardiomyocytes showed elevated reactive oxygen species levels, increased frequency and amplitude of Ca²⁺ sparks, and increased diastolic [Ca²⁺]_i compared to control; the latter two parameters were further increased by isoproterenol exposure and reversed by exposure to ryanodine receptor blocker dantrolene. We speculate that these isoproterenol-dependent effects may impact on the exercise-related atrial arrhythmia risk in ARVC patients. Despite absence of changes in Ca²⁺ _i transient dynamics, PKP2cKO atrial cardiomyocytes showed enhanced sarcomere shortening and impaired sarcomere relaxation. Orthogonal transcriptomic analysis of human(GTEx) and PKP2cKO atrial tissue led to identification of 41 transcripts depending on PKP2 expression. Biochemical follow-up confirmed reduced abundance of sarcomeric protein myosin binding protein C, potentially playing a role in cellular shortening and relaxation changes observed. Our findings provide novel insights into the role of PKP2 in atrial myocardium with potential implications to therapeutic management of atrial fibrillation in patients with PKP2-related ARVC. KEY POINTS: Atrial arrhythmias occur in a large group of patients with arrhythmogenic right ventricular cardiomyopathy (ARVC), a cardiac disease mostly caused by pathogenic variants in the desmosomal gene plakophilin-2 (PKP2). Exercise is considered to be an independent risk factor for arrhythmias consequent to PKP2 deficiency. We show that loss of PKP2 expression affects cellular calcium handling and electrophysiology differently in left atrial vs. ventricular myocardium and causes extensive atrial fibrosis. PKP2-deficient atrial cardiomyocytes present increased spontaneous sarcoplasmic reticulum calcium release events, further enhanced by isoproterenol exposure and reversible by a ryanodine receptor blocker (dantrolene). In addition, PKP2-deficient atrial myocytes exhibit impaired relaxation and enhanced sarcomere shortening, most probably related to reduced abundance of myosin binding protein C. We speculate that cellular effects reported upon isoproterenol impact on the exercise-related atrial arrhythmia risk in ARVC patients. We further propose that therapeutic approaches aimed at mitigating ventricular damage may be effective to treat the atrial disease in ARVC.

Keywords: adrenergic stimulation; arrhythmogenic right ventricular cardiomyopathy; atrial fibrillation; oxidative stress; plakophilin‐2; ryanodine receptor.

Figure 1:. Action potential duration and sodium current properties recorded from control and plakophilin-2 conditional knockout (PKP2cKO) atrial myocytes.

A) Bar graphs depicting action potential duration at 25% (APD25), 50% (APD50) and 90% (APD90) of repolarization, B) action potential amplitude (APA) and C) resting membrane potential (RMP) in control (black bar) and PKP2cKO (red bar). D) Average peak sodium current density as a function of voltage command. Peak current amplitude at −40 mV. E) Voltage-dependence of I_Na steady-state inactivation curves. Voltage for half-maximal inactivation (V 1/2). Data is presented as mean ± SD. Number of cells noted in the corresponding plots; 5 mice per group. Test for clustering versus validity of assumption of independence between data obtained from more than one mouse within a group, as in Sikkel et al (Sikkel et al., 2017). Hierarchical statistics were used for all parameters. Normality of distribution was tested by Shapiro-Wilk and Kolgomorov-Smirnov tests.

Figure 2:. Calcium transient dynamics and the extent of calcium sparks in control and plakophilin-2 conditional knockout (PKP2cKO) atrial cardiac myocytes; response to beta adrenergic stimulation.

A) Representative example of the line scan positioning during calcium handling recordings by confocal microscopy in atrial myocytes. Scale bar: 5μm. B) Calcium transient amplitude and C) Tau of decay in atrial control and PKP2cKO cardiac myocytes, at baseline or exposed to exposed to 100 nmol/L isoproterenol (ISO). Data is presented as mean ± SD. Number of cells noted in the corresponding bars; 4 mice per group. Tested for clustering versus validity of assumption of independence between data obtained from more than one mouse within a group, as in Sikkel et al (Sikkel et al., 2017). Hierarchical statistics were used for both calcium transient amplitude and tau decay. Normality of distribution was tested by Shapiro-Wilk and Kolgomorov-Smirnov tests.

Figure 3:. Extent of calcium sparks and calcium spark amplitude in atrial myocytes from control, plakophilin-2 conditional knockout (PKP2cKO) and Dantrolene-treated PKP2cKO mice; response to beta adrenergic stimulation.

A) Confocal line-scan images of calcium sparks in atrial cardiac myocytes from PKP2cKO mice treated Dantrolene, at baseline or exposed to 100 nmol/L ISO. B-C) Quantitative analysis of calcium spark frequency (B) and amplitude (C) in atrial cardiac myocytes form control, PKP2cKO and Dantrolene-treated PKP2cKO mice, at baseline or exposed to 100 nmol/L ISO. Number of cells noted in the corresponding bars; 4 mice per group. Tested for clustering versus validity of assumption of independence between data obtained from more than one mouse within a group, as in Sikkel et al (Sikkel et al., 2017). Hierarchical statistics were used for both parameters. Normality of distribution was tested by Shapiro-Wilk and Kolgomorov-Smirnov tests.

Figure 4:. Width and duration of Ca²⁺ sparks from control and plakophilin-2 conditional knockout (PKP2cKO) atrial myocytes.

A) Full-width at half maximum (FWHM; μm) and B) Full-duration at half maximum (FDHM; ms) of Ca²⁺ sparks recorded from atrial myocytes isolated from control (black bars) and PKP2cKO mice (red bars) 21 days post-TAM, at baseline and upon 100 nM isoproterenol (ISO) exposure. Data are presented as mean ± SD. N=4 mice per group. Test for clustering versus validity of assumption of independence between data obtained from more than one mouse within a group, as in Sikkel et al (Sikkel et al., 2017). Hierarchical statistics on animal -and cell level were used for both FWHM and FDHM analysis. Normality of distribution was tested by Shapiro-Wilk and Kolgomorov-Smirnov tests.

Figure 5:. Effect of isoproterenol (ISO) on calcium handling parameters in PKP2cKO atrial myocytes.

Percentage of change in Ca²⁺ spark frequency, Ca²⁺ spark amplitude, SR load and diastolic Ca²⁺ levels. Normalized to values obtained in the absence of ISO (100%). Data is presented as mean ± SD. Number of cells noted in the corresponding bars; 4 mice per group, unpaired Students T-test.

Figure 6:. Intracellular calcium content and SR load in atrial cardiac myocytes from control, plakophilin-2 conditional knockout (PKP2cKO) and PKP2cKO mice treated with Dantrolene; response to beta adrenergic stimulation.

A) Diastolic calcium levels and B) SR calcium load in atrial cardiac myocytes form control, PKP2cKO Dantrolene-treated PKP2cKO mice, at baseline or exposed to 100 nmol/L isoproterenol (ISO). SR load is recorded upon a pulse of caffeine (10 mmol/L). Intracellular calcium levels and SR load were detected by a ratiometric method (F_Fluo-3/F_{Fura Red}). Data is presented as mean ± SD. Number of cells noted in the corresponding bars; 4 mice per group. Tested for clustering versus validity of assumption of independence between data obtained from more than one mouse within a group, as in Sikkel et al (Sikkel et al., 2017). Hierarchical statistics were used for both parameters. Normality of distribution was tested by Shapiro-Wilk and Kolgomorov-Smirnov tests.

Figure 7:. Post-translational modifications in the atrial myocardium from plakophilin-2 conditional knockout (PKP2cKO) mice.

A) Mass spectrometry–based investigation of ryanodine receptor 2 (Ryr2) phosphorylation state in control (black bars) and PKP2cKO (red bars) atria. Measured peptide intensities are shown for phosphorylated peptides covering the indicated Ryr2 phosphorylation sites. Amino acid positions for phosphorylation sites falling within the identified hot spot of Ryr2 are highlighted in red. Mean ± SD, atria from 6 mice per group. Evaluation of abundance differences between the two groups was conducted using Student’s t-test and corrected for multiple hypothesis testing by Benjamini-Hochberg. There were no significant differences. B) Quantitative analysis of MitoSOX intensity (AU) in control and PKP2cKO atrial cardiac myocytes. C) Representative images of atrial cardiac myocytes labelled with MitoSOX and MitoTracker. Data is presented as mean ± SD. Number of cells noted in the corresponding bars; 3 mice per group. Tested for clustering versus validity of assumption of independence between data obtained from more than one mouse within a group, as in Sikkel et al (Sikkel et al., 2017) and hierarchical statistics were used. Normality of distribution was tested by Shapiro-Wilk and Kolgomorov-Smirnov tests.

Figure 8:. Sarcomere shortening dynamics in control and plakophilin-2 conditional knockout (PKP2cKO) atrial myocytes.

A) Representative examples of time course of sarcomere shortening (in control and PKP2cKO atrial cardiac myocytes during 1 Hz field stimulation. B-E) Quantitative analysis of (B) Baseline sarcomere length, (C) percentage of sarcomere shortening, (D) upstroke time and (E) Tau of decay in control (black bars) and PKP2cKO (red bars) atrial cardiac myocytes. Data is presented as mean ± SD. Number of cells noted in the corresponding bars; 4 mice per group. Tested for clustering versus validity of assumption of independence between data obtained from more than one mouse within a group, as in Sikkel et al (Sikkel et al., 2017). Hierarchical statistics were used for all four parameters. Normality of distribution was tested by Shapiro-Wilk and Kolgomorov-Smirnov tests.

Figure 9:. Structural remodeling on the macro and micro scale in atria from plakophilin-2 conditional knockout (PKP2cKO) mice.

A) Representative examples of Massons’s Trichrome staining in left atria from control and PKP2cKO mice. Per genotype, left panels represent the entire atria and panels on the right zoom-ins throughout the atrium. B) Quantitative analysis of percentage of collagen in the left atrium from control and PKP2cKO mice. Data is presented as mean ± SD; 7 mice per group; Student’s Test. Normality of distribution was tested by Shapiro-Wilk and Kolgomorov-Smirnov tests. C) Single frame of a serial block face-scanning electron microscopy (SBF-SEM) - acquired set showing the ultrastructure of the intercalated disc from control (left panel) and PKP2cKO (right panel) left atrial tissue. Scale bar: 1 μm. A complete 3D volume is presented in supplemental movies 1 and 2.

Figure 10:. Transcripts correlated with plakophilin 2 (PKP2) expression in the human and murine atrial myocardium.

A) Volcano plot showing the differential transcriptome analysis in left atria from control versus plakophilin-2 conditional knockout (PKP2cKO) mice. Thresholds of | log2FC | >0.25 and 5% false discovery rate (FDR < 0.05). Notice the interruption in the ordinate axis to include the data corresponding to Pkp2 and to Esr1. B) Volcano plot showing the association analysis of the human atrial transcriptome in relation to PKP2 expression. Blue dots and red dots: transcripts significantly downregulated and upregulated, respectively, in atrial PKP2cKO RNAseq dataset. Selected gene names are indicated in the figure. Dots on top, and marked with an infinity sign in the ordinates, represent transcripts with a −log10 padj values larger than 30.

Figure 11:. Linear regression of normalized Gene A expression vs normalized Pkp2 expression.

Linear regression between each transcript and Pkp2 normalized expression was used to obtain the regression coefficient (β). Regression coefficient and p-value were obtained from line of best fit. P-value was corrected by Benjamini-Hochberg method.

Figure 12:. KEGG pathway analysis on human atrial GTEx dataset, of genes significantly correlated to Pkp2 expression.

Highlighted are pathways linked to A) the pathophysiology of Arrhythmogenic Right Ventricular Cardiomyopathy (ARVC) and B) neurodegenerative diseases. Thresholds of | regression coefficient | >0.5 and 5% P-adjusted (P-adj < 0.001) for significant correlation.

Figure 13:. Down-regulation of myosin binding protein C in atrial myocardium and atrial myocytes from plakophilin-2 conditional knockout (PKP2cKO) mice.

A) RT-PCR analysis to examine transcript levels of MyBPC3 and PKP2 in control and PKP2cKO left atria. Values are normalized to transcript levels in control mice. Data is presented as mean ± SD; 3 mice per group. B) Quantitative analysis (left panel) and gel images (right panel) of MyBPC3 protein detected in left atria from control and PKP2cKO mice. Values are normalized to GAPDH levels in corresponding samples. Molecular weight of MyBPC3 is 140 kDa. Data is presented as mean ± SD; 3 mice per group C) Left panel: Quantitative analysis of MyBPC3 intensity (AU) in control and PKP2cKO atrial cardiac myocytes. Right panel: Representative images of atrial cardiac myocytes immunostained for MyBPC3. Data is presented as mean ± SD. Number of cells noted in the corresponding bars; 3 mice per group. Tested for clustering versus validity of assumption of independence between data obtained from more than one mouse within a group, as in Sikkel et al (Sikkel et al., 2017); hierarchical statistics were used. Normality of distribution was tested by Shapiro-Wilk and Kolgomorov-Smirnov tests.

Figure 14:

Overview of outcomes on electrophysiological parameters measured in plakophilin-2 conditional knockout (PKP2cKO) ventricular and atrial myocytes, versus control. Based on data described in this manuscript, as well as in (Cerrone et al., 2017, Pérez-Hernández et al. 2021, Cerrone et al., 2022). SS: sarcomere shortening, Ca²⁺: calcium, = : no change, − : reduction, + : increase.