Alda-1 attenuation of binge alcohol-caused atrial arrhythmias through a novel mechanism of suppressed c-Jun N-terminal Kinase-2 activity

Abstract

Holiday Heart Syndrome (HHS) is caused by excessive binge alcohol consumption, and atrial fibrillation (AF) is the most common arrhythmia among HHS patients. AF is associated with substantial morbidity and mortality, making its prevention and treatment of high clinical interest. This study defines the anti-AF action of Alda-1 (an established cardioprotective agent) and the underlying mechanisms of the action in our well-characterized HHS and cellular models. We found that Alda-1 effectively eliminated binge alcohol-evoked Ca²⁺ triggered activities (Ca²⁺ waves, prolonged Ca²⁺ transient diastolic decay) and arrhythmia inducibility in intact mouse atria. We then demonstrated that alcohol impaired human RyR2 channels (isolated from organ donors' hearts). The functional role of alcohol-caused RyR2 channel dysfunction in Ca²⁺ triggered arrhythmic activities was evidenced in a unique transgenic mouse model with a loss-of-function mutation (RyR2^E4872Q+/-). Alda-1 is known to activate aldehyde dehydrogenase 2 (ALDH2), a key enzyme in alcohol detoxification. However, we found an increased level of ALDH2 and a preserved normal balance of pro- vs anti-apoptotic signaling in binge alcohol exposed hearts and H9c2 differentiated myocytes, which suggests that the link of alcohol-ALDH2-apoptosis is unlikely to be a key factor leading to binge alcohol-evoked arrhythmogenicity. We have previously reported that binge alcohol-activated stress response kinase JNK2 causatively drives Ca²⁺-triggered atrial arrhythmogenicity. Here, we found that JNK2-specific inhibition in either isolated human RyR2 channels or intact mouse atria abolished alcohol-evoked RyR2 channel dysfunction and Ca²⁺ triggered arrhythmic activities, suggesting a strong alcohol-JNK2-RyR2 interaction in atrial arrhythmogenicity. Furthermore, we revealed, for the first time, that Alda-1 suppresses JNK2 (but not JNK1) enzyme activity independently of ALDH2, which in turn alleviates binge alcohol-evoked Ca²⁺ triggered atrial arrhythmogenesis. Our findings provide novel mechanistic insights into the anti-arrhythmic action of Alda-1 and suggest that Alda-1 represents a potential preventative agent for AF management for HHS patients.

Keywords: Alda-1; Atrial arrhythmias; Calcium transients; Cardiac protection; Holiday heart syndrome; c-Jun N-terminal kinase-2.

Fig. 1.

Alda-1 eliminated pacing-induced atrial arrhythmias in HHS mice. A) Summarized data of pacing induced incidence of atrial tachycardia (AT) and AF (quantified as incidence of AT/AF per burst pacing challenge per animal) in HHS mice and sham controls. Alda-1 treatment in HHS mice completely eliminated pacing-induced AT/AF episodes (far right bar). B) Summarized data of average duration of pacing-induced AT/AF in HHS mice and sham controls. C). Representative electrogram traces showing burst pacing-induced AF in an HHS heart (upper panel), while Alda-1 treated HHS heart showing self-restored normal sinus rhythm after burst pacing. Individual data points were shown alongside mean ± interquartile range with statistical analysis by Kruskal-Wallis one-way ANOVA and Dunn’s test for multiple comparison.

Fig. 2.

Binge alcohol increases ALDH2 abundance but does not affect apoptotic signaling markers. A-B) Representative immunoblotting images and summarized quantification data of ALDH2 protein expression in 24 h alcohol treated H9c2 myocytes with and without the presence of Alda-1 compared to sham controls. C-E) Pooled immunoblotting data and example images of unchanged apoptosis marker Bax (C & E, upper panel) and the ratio of pBcl2/tBcl2 (D & E, middle panels) in alcohol treated H9c2 myocytes with and without the presence of Alda-1 compared to sham controls. F-G) summarized immunoblotting data show unaltered levels of the apoptosis marker BAD and unchanged ratio of pBcl2/tBcl2 in HHS mouse hearts compared to sham controls, while Alda-1 treatment reduced the ratio of pBcl2/tBcl2 in HHS hearts compared to controls. Three technical replicates (immunoblotting) were performed. Individual data points were shown alongside mean ± interquartile range with statistical analysis by Kruskal-Wallis one-way ANOVA and Dunn’s test for multiple comparison.

Fig. 3.

Either Alda-1 treatment or JNK2 inhibition eliminates binge alcohol-evoked Ca²⁺ mishandling in intact mouse atria. A) Representative confocal images showing increased pacing-induced diastolic Ca²⁺ wave (yellow arrows) frequency in HHS intact atria, while there are no pacing-induced Ca²⁺ waves in sham controls and Alda-1-treated HHS atria. B) Summarized data showing increased frequency of Ca²⁺ waves in binge alcohol-exposed mouse atria, while either Alda-1 treatment or JNK2 inhibition in JNK2dn transgenic mice with cardiac-specific overexpression of inactive dominant negative (dn) JNK2 precluded binge alcohol-prompted Ca²⁺ waves. C) HHS mouse atria exhibited prolonged Ca²⁺ decay constant, τ, compared to sham controls. However, either Alda-1 treatment in HHS-WT mice or cardiac JNK2 inhibition in JNK2dn mice normalized τ of Ca²⁺ decay as seen in sham controls. Individual data points were shown alongside mean ± interquartile range with statistical analysis by Kruskal-Wallis one-way ANOVA and Dunn’s test for multiple comparison. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 4.

Alcohol drives RyR2 channel dysfunction in a JNK2-dependent manner. A) Summarized bar graphs showing alleviated binge alcohol-evoked Ca²⁺ waves frequency in HHS RyR2^E4872Q+/− (EQ^+/−) with loss-of-function of RyR2 channels, while HHS WT (littermates, WT-Lit) mouse atria exhibited significantly increased spontaneous Ca²⁺ wave frequency (sinus rhythm (SR) before burst pacing) and 10 Hz pacing-induced Ca²⁺ wave frequency during the recovery period (when burst pacing was stopped) in comparison to that of WT young controls. B) Pooled data showing precluded arrhythmic Ca²⁺ waves in JNK activator anisomycin (Aniso)-treated EQ^+/− mice. In contrast, Aniso-treated WT littermates showed significantly increased Ca²⁺ waves during sinus rhythm (SR) and in response to burst-pacing in Aniso-treated WT young mouse atria. Single channel recordings were performed on isolated human atrial RyR2 channels (hRyR2). C) Sample human single RyR2 channel recordings showing before (control) and after cytosolic addition of 39 mM alcohol (without and with pre-treatment of JNK2I-IX. The zero current levels are indicated by a dash). D-E) summarized bar graphs showing the mean single RyR2 Po and mean opening time after alcohol treatment alone or with JNK2 inhibitor JNK2I present. Individual data points were shown alongside mean ± interquartile range with statistical analysis by Kruskal-Wallis one-way ANOVA and Dunn’s test for multiple comparison.

Fig. 5.

Alda-1 suppresses binge alcohol-activated JNK2 activity, but not JNK1. (A-B) An in vitro kinase activity assay shows that immunoprecipitated JNK2 (IPed: JNK2 with JNK2 specific antibody; A) but not JNK1 (IPed: JNK1 using JNK1 specific antibody; B) activity is significantly elevated in alcohol-exposed H9c2 myocytes. C) Summarized data showing a significantly increased level of JNK-P in HHS mice, while Alda-1 pre-treated HHS mice showed suppressed JNK2 activity. D) Pooled immunoblotting data demonstrating unchanged levels of JNK2 (D) and JNK1 (E) expression in HHS mouse hearts with and without Alda-1 treatment. F) Representative immunoblotting images of JNK-P, JNK2, JNK1 bands in alcohol-treated H9c2 myocytes with and without the presence of Alda-1. G) Immunoblotting images showing comparable level of phosphorylated JNK (JNK-P) in HEK293T cells co-transfected with FLAG-tagged ALDH2 and turbo(t)-GFP-tagged active JNK2 (tGFP-JNK2) vectors (lanes 3–4) compared to tGFP-JNK2-only transfected cells (lanes 1–2), while a JNK-P band is absent in the FLAG-ALDH2-only transfected cells (lane 5). H) A summarized bar graph showing significantly reduced enzyme activity of active JNK2 pure protein in the presence of Alda-1 compared to the baseline JNK2 enzyme activity in the absence of Alda-1. At least three technical replicates (immunoblotting & enzyme activity assay) were performed. Individual data points were shown alongside mean ± interquartile range with statistical analysis by Mann-Whitney test for two-group comparison and Kruskal-Wallis one-way ANOVA and Dunn’s test for multiple comparison.