Pak2 Regulation of Nrf2 Serves as a Novel Signaling Nexus Linking ER Stress Response and Oxidative Stress in the Heart

Abstract

Endoplasmic Reticulum (ER) stress and oxidative stress have been highly implicated in the pathogenesis of cardiac hypertrophy and heart failure (HF). However, the mechanisms involved in the interplay between these processes in the heart are not fully understood. The present study sought to determine a causative link between Pak2-dependent UPR activation and oxidative stress via Nrf2 regulation under pathological ER stress. We report that sustained ER stress and Pak2 deletion in cardiomyocytes enhance Nrf2 expression. Conversely, AAV9 mediated Pak2 delivery in the heart leads to a significant decrease in Nrf2 levels. Pak2 overexpression enhances the XBP1-Hrd1 UPR axis and ameliorates tunicamycin induced cardiac apoptosis and dysfunction in mice. We found that Pak2 deletion and altered proteostasis render Nrf2 detrimental by switching from its antioxidant role to renin-angiotensin aldosterone system (RAAS) gene regulator. Mechanistically, Pak2 mediated Hrd1 expression targets Nrf2 for ubiquitination and degradation thus preventing its aberrant activation. Moreover, we find a significant increase in Nrf2 with a decrease in Pak2 in human myocardium of dilated heart disease. Using human-induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs), we find that Pak2 is able to ameliorate Nrf2 induced RAAS activation under ER stress. These findings demonstrate that Pak2 is a novel Nrf2 regulator in the stressed heart. Activation of XBP1-Hrd1 is attributed to prevent ER stress-induced Nrf2 RAAS component upregulation. This mechanism explains the functional dichotomy of Nrf2 in the stressed heart. Thus, Pak2 regulation of Nrf2 homeostasis may present as a potential therapeutic route to alleviate detrimental ER stress and heart failure.

Keywords: ER stress; RAAS and oxidative stress; gene therapy; heart failure; oxidative stress; proteostasis.

Figure 1

Pak2 cardiac deletion enhances ER stress induced cardiac dysfunction, defective UPR and oxidative stress. Oxidative stress analysis of heart tissue was performed using dihydroethidium (DHE) (A) (scale bar = 20 μm, n = 5), and reduced glutathione to oxidized glutathione (GSH/GSSG) ratio (B) in TAC-stressed hearts (n = 5–6). Oxidative stress analysis of heart tissue using DHE (C) (scale bar = 20 μm, n = 5) and GSH/GSSG ratio (n = 6) (D) in TM-injected mouse hearts. (E) TUNEL assay of hearts after 2 days of TM injection (scale bar = 20 μm), arrows indicate TUNEL positive nuclei (n = 7). (F) Immunoblots and quantification of UPR components in TAC-stressed hearts (n = 5). Student's t-test or 2-way ANOVA with Bonferroni correction for post-hoc comparisons were used for analyses. Data presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 2

Nrf2 expression is upregulated in Pak2 depleted hearts under sustained ER stress. (A) Differentially expressed genes by Affymetrix gene array from Tunicamucin treated mice. Gene expression profiling in Pak2-Flox vs. Pak2-CKO hearts was normalized to gray. Red represents decreased expression (n = 2). (B) Immunoblots and quantification from hearts of C57BL/6N mice subjected to 2 or 7 days TM intraperitoneal injection (n = 5). (C) Quantitative polymerase chain reaction (qPCR) analysiss of Nrf2 transcript in TM-stressed hearts (n = 5). (D) Echocardiographic analysis of mice after TM injection for 2 days and 1 week (n = 5). (E) Heart weight/tibia length ratio (HW/TL) determined an increase in cardiac hypertrophy following TM injection (n = 5). (F) Immunoblots and quantification of hearts from mice subject to TAC (n = 4). (G) qPCR analysis of Nrf2 transcript in hearts of mice subject to 1 or 5 week TAC (n = 4). Immunoblots (H) and qPCR (I) analysis of Nrf2 levels in TAC-stressed Pak2-CKO hearts (n = 5). 1-way or 2-way ANOVA with Bonferroni correction for post-hoc comparisons were used for analyses. Data presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 3

Activated Pak2 is protective against sustained ER stress and decreases Nrf2 accumulation in the heart. (A) M-mode images captured by echocardiography were used to determine cardiac function in AAV-9-Pak2-T402E-injected C57BL/6N mice. (B,C) Echocardiographic analysis after 1 week Tunicamycin (TM, 2 mg/kg) injection (n = 5). Detailed Echocardiographic data available in Supplementary Table 3. Cardiac hypertrophy was determined by assessing mice heart weight/tibia length ratio (HW/TL) (D) and myocyte cross-sectional areas (E) (scale bar = 20 μm, n = 5). Cell death in was evaluated by immunoblots and quantification of Chop and cleaved caspase 3 (F) and by TdT-mediated dUTP nick end labeling (TUNEL) assay (G) (scale bar = 20 μm) arrows indicate TUNEL positive nuclei (n = 5). (H) Immunoblots and quantification of Nrf2 and UPR components in Tunicamycin-stressed hearts. (I) qPCR analysis of Nrf2 transcript in AAV9-Pak2 -injected hearts (n = 5). 2-way ANOVA with Bonferroni correction for post hoc comparisons were used for analyses. Data presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 4

Pak2 depletion promotes aberrant Nrf2-dependent RAS gene upregulation under ER stress. (A) Immunoblots and quantification of phosphorylated Pak2 and Nrf2 in ARCMs subjected to tunicamycin (TM)-induced ER stress (n = 3). (B) Immunoblots of Pak2-knockdown (Ad-shPak2) ARCMs under TM treatment (n = 4). Nrf2 subcellular localization was assessed by immunoblot analysis using fractional protein preparation (C) and immunofluorescence analysis in Ad-shPak2 H9C2 cells under TM stress (D,E) (n = 3, Scale bar: 20μm). (F) Quantitative polymerase chain reaction (qPCR) was used to assess Nrf2 conventional transcriptional targets (n = 3). (G) Cell viability was determined by MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] reduction assay in WT (G) or Ad-shPak2 (H) cardiomyocytes under TM stress in combination with Nrf2 inhibition with ML385 (n = 4). (I, J) qPCR analyses was used to assess RAAS components Agt, Atr1, Mas1, and Ace1 in Ad-shPak2 cardiomyocytes under ER stress (n = 3). 2-way ANOVA with Bonferroni correction for post-hoc comparisons were used for analyses. Data presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 5

Pak2-dependent IRE-XBP1 activation leads to enhanced Hrd1-mediated Nrf2 ubiquitination and degradation. (A) Hrd1 reporter luciferase activity was detected in Pak2 knockdown (Ad-shPak2) H9C2 cells in response to Tunicamycin in XBP1s expressing (Flag-XBP1) cardiomyocytes (n = 4). (B) The effect of active Pak2 overexpression and Hrd1 knockdown (siHrd1) (C) on Nrf2 ubiquitination was determined by HA-ubiquitin (HA-ub) immunoprecipitation and Nrf2 immunoblot (n = 3). Association of endogenous Nrf2 with Hrd1 was observed by immunoprecipitation in H9C2 cells under basal conditions (D) and in Tunicamycin stressed cells (E). The effect of tunicamycin in Hrd1 and Nrf2 binding is expressed as the change in IP/input ratio for Hrd1 and Nrf2 (n = 3). Nrf2 degradation in cardiomyocytes was assessed by cyclohexamide (CHX, 100 μg/mL) chases. Immunoblots showed that Flag-Pak2-T402E (F) and Myc-Hrd1 (G) enhance the clearance of Nrf2 in ARCMs (n = 3). (H) Immunoblots showing the effect of Keap1 knockdown (Keap1 siRNA) on Pak2-mediated Nrf2 downregulation (n = 3). Student's t-test or 2-way ANOVA with Bonferroni correction for post-hoc comparisons were used for analyses. Data presented as mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 6

Pak2 activation alleviates ER stress-induced cell death in human cardiomyocites via Nrf2 downregulation. (A) Immunoblots and quantification of Nrf2, Hrd1 and cell death markers in tunicamycin stressed human-induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs) infected with adenovirus expressing active Pak2 (Ad-Pak2) (n = 3). (B) Cell viability of iPSC-CMs was determined by MTT assay (n = 4). Immunoblots and quantification of Nrf2, cell death markers (C), ATR1, Angiotensinogen, and Mas1 (E) in tunicamycin stressed Pak2 knockdown iPSC-CMs under Nrf2 inhibition with ML385 (n = 3). (D) The effect on cell viability was determined by MTT assay (n = 4). (F) Immunoblots and quantification of Pak2 and Nrf2 in heart samples from healthy donors and transplantation patients with dilated heart disease (n = 5). Student's t-test or 2-way ANOVA with Bonferroni correction for post-hoc comparisons were used for analyses. Data presented as mean ± SEM. (G) Proposed model for Pak2-mediated regulation of Nrf2 in response to ER stress. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.