P21-Activated Kinase 2 as a Novel Target for Ventricular Tachyarrhythmias Associated with Cardiac Adrenergic Stress and Hypertrophy

Abstract

Ventricular arrhythmias associated with cardiac adrenergic stress and hypertrophy pose a significant clinical challenge. We explored ventricular anti-arrhythmic effects of P21-activated kinase 2 (Pak2), comparing in vivo and ex vivo cardiomyocyte-specific Pak2 knockout (Pak2^cko) or overexpression (Pak2^ctg) murine models, under conditions of acute adrenergic stress, and hypertrophy following chronic transverse aortic constriction (TAC). Pak2 was downregulated 5 weeks following the latter TAC challenge. Cellular physiological, optical action potential and Ca²⁺ transient, measurements, demonstrated increased incidences of triggered ventricular arrhythmias, and prolonged action potential durations (APD) and altered Ca²⁺ transients with increases in their beat-to beat variations, in Pak2^cko hearts. Electron microscopic, proteomic, and molecular biological methods revealed a mitochondrial localization of stress-related proteins on proteomic and phosphoproteomic analyses, particularly in TAC stressed Pak2^cko mice. They further yielded accompanying evidence for mitochondrial oxidative stress, increased reactive oxygen species (ROS) biosynthesis, reduced mitochondrial complexes I-V, diminished ATP synthesis and elevated NADPH oxidase 4 (NOX4) levels. Pak2 overexpression and the novel Pak2 activator JB2019A ameliorated these effects, enhanced cardiac function and decreased the frequencies of triggered ventricular arrhythmias. Pak2 activation thus protects against ventricular arrhythmia associated with cardiac stress and hypertrophy, through unique mechanisms offering potential novel therapeutic anti-arrhythmic targets.

Keywords: Ca2+ handling; NADPH oxidase; P21‐activated kinase 2; cardiac arrhythmia; mitochondrial oxidative stress.

Figure 1

Pak2 deficiency aggravates the inducibility of cardiac arrhythmias in mice challenged with acute isoproterenol (ISO) or chronic (5 weeks) TAC challenge in vivo. A) The protocol scheme applied to Pak2^f/f and Pak2^cko mice under acute isoproterenol (ISO) or chronic TAC challenge. B) Representative in vivo cardiac electrophysiological recordings from Pak2^f/f and Pak2^cko mice with different frequencies of burst stimulation. C) VT inducibility during in vivo electrocardiogram measurements in Pak2^f/f and Pak2^cko mice before and after isoproterenol challenge. D) Quantification of VT occurrence following the application of isoproterenol in Pak2^f/f and Pak2^cko mice at different frequencies of burst stimulation. E) Representative in vivo cardiac electrophysiological recordings from Pak2^f/f and Pak2^cko mice at 5 weeks after sham or TAC surgery. F) Occurrence of ventricular ectopic beats of Pak2^f/f and Pak2^cko mice at 5 weeks after sham or TAC surgery. G) Survival curves of four groups of mice after operation. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 2

Pak2 deficiency aggravates the inducibility of VT and alternans in mice under chronic (5 weeks) TAC challenge in ex vivo hearts. A) Flow chart of optical mapping scheme. B) VT inducibility in ex vivo Langendorff hearts after TAC challenge in Pak2^f/f and Pak2^cko mice. C) Heat maps (indicating the spectral intensity differences of Ca²⁺ transients) showing frequency‐dependent CaT alternans during progressive decrements in cycle length between 100–70 ms. The corresponding CaT traces are also exhibited from the selected region within the boxes. D) Calculation of CaT alternans ratio for each group (n = 6 hearts for each group). Large alternans ratio (>20%) are counted for evaluation of the severity of CaT alternans. E) Typical AP traces during VT induced by progressive pacing; the activation map shows that the initiation positions for ectopy and reentry overlap with regions harboring the most severe AP alternans. F) Typical consecutive phase maps during VT in a Pak2^cko mouse heart, showing the fast re‐entry rotors meandering on the epicardium without losing the coupling of AP and CaT, along with the corresponding AP and CaT traces. *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 3

Cardiac specific Pak2 overexpression attenuates isoproterenol or TAC‐induced cardiac arrhythmias and Ca²⁺ alternans. A) The protocol scheme for Pak2^WT (Rosa26^{CAG‐LSL‐Pak2}) and Pak2^ctg mice with acute isoproterenol (ISO) or chronic TAC challenge. B) VT inducibility during in vivo electrocardiogram measurement in Pak2^WT and Pak2^ctg mice before and after isoproterenol challenge with rapid pacing for 100 stimuli each. C) Quantification of VT duration in the application of isoproterenol in Pak2^WT and Pak2^ctg mice in vivo. D) Calculation of ventricular ectopic beats after TAC challenge for 5 weeks in WT and Pak2^ctg mice in vivo with isoproterenol and caffeine challenge. E) Heat maps (indicating the spectral intensity difference of Ca²⁺ alternans) showing frequency‐dependent CaT alternans during progressive decrements of pacing cycle length between 100–70 ms. Corresponding CaT traces are also exhibited from the selected region within the boxes. F) Calculation of CaT alternans ratio for each group (n = 4‐7 hearts). Large alternans ratio (>20%) are counted for evaluation of the severity of CaT alternans. G) Typical AP traces during VT in a Pak2^WT mouse heart after TAC for 7 weeks induced by a progressive pacing; activation map shows the initiation positions of ectopy and reentry overlap with a region harboring the most severe AP alternans. H) Typical consecutive phase maps during VT, showing the fast reentry rotors meandering on the epicardium without losing the coupling of AP and CaT, along with the corresponding AP and CaT traces. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 4

Pak2 deletion aggravates isoproterenol and TAC‐induced disruptions of cardiac intracellular calcium homeostasis. A) Calcium fluorescence was measured in isolated ventricular myocytes from Pak2^f/f and Pak2^cko mice after acute isoproterenol treatment. B) The pie diagrams of occurrence of premature Ca²⁺ transients (PCT) and Ca²⁺ oscillation (CO) from the indicated groups. C) The statistical histogram of severity of Ca²⁺ transient abnormality from the indicated groups. D) Calcium fluorescence was measured in isolated cardiomyocytes from Pak2^f/f and Pak2^cko mice after sham or TAC surgery for 5 weeks. E) The pie diagrams of occurrence of PCT and CO from the indicated groups. F) The statistical histogram of severity of Ca²⁺ transient abnormality from the indicated groups. G) The statistical histogram of τ of Ca²⁺ transient decays from the indicated groups. *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 5

Pak2 deletion aggravates TAC‐induced cardiac oxidative stress. A) Heat map showing differentially expressed protein with proteomics in heart tissue from Pak2^f/f and Pak2^cko mice with sham or TAC surgery for 5 weeks. B) KEGG enrichment analysis of differentially expressed protein from Pak2^f/f and Pak2^cko mice at 5 weeks after TAC surgery. C) GO enrichment analysis of proteomics in heart tissue from different treatments. D) Heat map showing differentially expressed protein with phos‐proteome in heart tissue from different treatments. E) KEGG enrichment analysis of phos‐proteomics in heart tissue from different treatments. F) GO enrichment analysis of phos‐proteomics. G) Venn graph showing the significantly different proteins and phosphoprotein for Pak2^cko/TAC versus Pak2^f/f/TAC, and Pak2^cko/TAC versus Pak2^cko/sham respectively. H) Comparing the correlation between the proteomic and phosphoproteomic changes for TMT and PP proteomics results of each sample between Pak2^cko/TAC and Pak2^f/f/TAC. I) the relevant proteins enriched in the oxidative phosphorylation and tricarboxylic acid cycle pathways were analyzed for correlation between the phosphoproteomics and proteomics. K,M) The WGCNA method used to identify protein modules associated with sample phenotypes for proteomic (K) and phosphoproteomic arrays (M). J,L) The turquoise module, which is most associated with the sample phenotype, was subjected to KEGG and GO pathway enrichment analysis for proteomic (J) and phosphoproteomic arrays (L).

Figure 6

Pak2 deletion aggravates cardiac dysfunction by impairing mitochondrial function. A) Representative fluorescence images shows ROS levels of Pak2^f/f and Pak2^cko mice at 5 weeks after sham or TAC surgery. Scale bar 100 µm. B) Quantification of panel A. C) Representative transmission electron microscopy images of Pak2^f/f and Pak2^cko mice at 5 weeks after sham or TAC surgery. Scale bar 1 µm. D) ATP content in Pak2^f/f and Pak2^cko mice at 5 weeks after sham or TAC surgery. E) Pie diagrams showing the subcellular location of differentially expressed protein in Pak2^f/f and Pak2^cko mice at 5 weeks after TAC surgery in the comparison between the Pak2^cko/TAC and Pak2^f/f/TAC groups. F) Heat map showing differentially expressed mitochondrial respiratory electron transport chain protein in Pak2^f/f and Pak2^cko mice at 5 weeks after sham or TAC surgery. G) Representative immunoblotting images showing NDUFB6, NDUFS8, NDUFA8 protein from indicated groups. H) Quantification of G (n = 4 mice for each group). I) Statistical graph of relative activity of mitochondrial complex I. J) Representative immunoblotting images showing NOX4, NOX2, ox‐CaMKII, p‐CaMKII and t‐CaMKII protein from indicated groups. K) Quantification of J (n = 4 mice for each group). L) Representative fluorescence images shows ROS levels of Pak2^f/f and Pak2^cko mice after 30 min isoproterenol treatment. Scale bar 100 µm. M) Quantification of L. N,O) Representative immunoblotting images and statistical graph showing the effect of acute isoproterenol on p‐CaMKII in Pak2^f/f and Pak2^cko mice (n = 4 mice for each group). P,Q) OCRs were analyzed by seahorse extracellular flux analyses involving sequential injection of ADP, oligomycin (OA), trifluorocarbonyl cyanide phenylhydrazone (FCCP), and antimycin A (AA). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 7

Cardiac‐specific Pak2 overexpression attenuates TAC or isoproterenol‐induced cardiac oxidative stress and abnormalities in mitochondrial structure. A,B) Representative fluorescence images (A) and statistical graph (B) shows ROS levels of Pak2^WT and Pak2^ctg mice at 7 weeks after sham or TAC surgery. Scale bar 100 µm. C) Representative transmission electron microscopy images of Pak2^WT and Pak2^ctg mice at 7 weeks after sham or TAC surgery. Scale bar 1 µm. D) Representative immunoblotting images showing NOX4, ox‐CaMKII, p‐CaMKII and t‐CaMKII protein in Pak2^WT and Pak2^ctg mice at 7 weeks after sham or TAC surgery. E) Quantification of D (n = 4 mice for each group). F,G) Representative fluorescence images (F) and statistical graph (G) shows ROS levels of Pak2^WT and Pak2^ctg mice at 7 weeks after sham or TAC surgery. Scale bar 100 µm. H) Representative immunoblotting images and statistical analysis showing p‐CaMKII and t‐CaMKII protein in Pak2^WT and Pak2^ctg mice after sham or acute isoproterenol treatment (n = 4 mice for each group). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 8

Pak2 activator JB2019A attenuates TAC‐induced cardiac hypertrophy and susceptibility to ventricular arrhythmia. WT mice were subjected to sham or TAC surgery for 5 weeks. DMSO or JB2019A (80 mg kg⁻¹/d, i.p.) was injected for 4 weeks from 1 weeks after sham or TAC surgery. A) Chemical structure of Pak2 activator JB2019A. Docking model of Alphafold2 predicted Pak2 structure and its activator JB2019A. Key residues involved in JB2019A binding are labeled. Hydrogen bonds and distances are highlighted in yellow; B) Representative immunoblotting images showing p‐Pak2 protein level from WT mice with DMSO or JB2019A treatment (n = 4 mice for each group). C) Representative in vivo cardiac electrophysiological studies of Pak2^f/f and Pak2^cko mice with different frequencies of burst stimulation. D) Representative in vivo electrophysiological recordings from WT mice at 5 weeks after sham or TAC surgery. Cardiac arrhythmia was induced with isoproterenol (2 mg kg⁻¹) and caffeine (160 mg kg⁻¹). Arrows indicate the ventricular ectopic beats (EB). Scale bar: 1 s or 0.2 s. E) statistical graph showing the VT occurrence of C. F) Statistical graph of ventricular ectopic beats in the four groups of mice injected with isoproterenol and caffeine. G) Representative anatomic images and cardiac longitudinal morphology of WT heart at 5 weeks after sham or TAC surgery with DMSO or JB2019A treatment. Scale bar, 5 mm (Upper panel). Scale bar, 2.5 mm (Lower panel). H) WGA staining of heart tissue from indicated groups. Scale bar, 50 µm. I) HW/BW (mg/g) ratio and cell cross‐sectional area of WT heart at 5 weeks after sham or TAC surgery with DMSO or JB2019A treatment (n = 4‐7 mice). J) Echocardiographic analysis of WT heart at 5 weeks after sham or TAC surgery with DMSO or JB2019A treatment. K) Quantification of echocardiography parameters in J (n = 3‐7 mice). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Figure 9

Pak2 activator JB2019A attenuates TAC‐induced cardiac oxidative stress and abnormalities in mitochondrial structure. A) Representative fluorescence images shows ROS levels of WT mice 5 weeks after sham or TAC surgery with DMSO or JB2019A treatments. Scale bar 100 µm. B) Quantification of A. C) Representative transmission electron microscopy images of WT mice at 5 weeks sham or TAC surgery with DMSO or JB2019A treatments. Scale bar 1 µm. D,E) Representative immunoblotting images and statistical graph showing ox‐CaMKII and t‐CaMKII protein expressed 5 weeks after sham or TAC surgery with DMSO or JB2019A treatments (n = 4 mice for each group). F) Representative immunoblotting images showing Pak2 and NOX4 protein levels in the myocardium of control patients and patients with atrial fibrillation. G) Quantification of F. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.