Metabolic stress-induced cardiomyopathy is caused by mitochondrial dysfunction due to attenuated Erk5 signaling

Abstract

The prevalence of cardiomyopathy from metabolic stress has increased dramatically; however, its molecular mechanisms remain elusive. Here, we show that extracellular signal-regulated protein kinase 5 (Erk5) is lost in the hearts of obese/diabetic animal models and that cardiac-specific deletion of Erk5 in mice (Erk5-CKO) leads to dampened cardiac contractility and mitochondrial abnormalities with repressed fuel oxidation and oxidative damage upon high fat diet (HFD). Erk5 regulation of peroxisome proliferator-activated receptor γ co-activator-1α (Pgc-1α) is critical for cardiac mitochondrial functions. More specifically, we show that Gp91phox activation of calpain-1 degrades Erk5 in free fatty acid (FFA)-stressed cardiomyocytes, whereas the prevention of Erk5 loss by blocking Gp91phox or calpain-1 rescues mitochondrial functions. Similarly, adeno-associated virus 9 (AAV9)-mediated restoration of Erk5 expression in Erk5-CKO hearts prevents cardiomyopathy. These findings suggest that maintaining Erk5 integrity has therapeutic potential for treating metabolic stress-induced cardiomyopathy.The mechanistic link between metabolic stress and associated cardiomyopathy is unknown. Here the authors show that high fat diet causes calpain-1-dependent degradation of ERK5 leading to mitochondrial dysfunction, suggesting the maintenance of cardiac ERK5 as a therapeutic approach for cardiomyopathy prevention and/or treatment.

Fig. 1

Decreased expression and phosphorylation of Erk5 in the obesity/diabetic hearts. a Immunoblot analyses (antibody recognizing C-terminus of Erk5) showed a decrease of Erk5 expression in the hearts from C57BL/6J mice with 25-week HFD, db/db mice, ob/ob mice, or rhesus monkeys with metabolic syndrome (MS). Quantification of Erk5/Actin ratio is presented in the bar graph. Data are means  ± SD (*P < 0.05, vs. control groups, n = 6 animals per group). b Immunoblot analyses substantiated that expression level of Erk5 (antibody recognizing N-terminus of Erk5), Mef2a and Mef2d was decreased in the obesity/diabetic hearts. c The phosphorylation levels of Erk5, Erk1/2, Jnk, p38, Ampk, Creb, and Mek5 in the hearts from mice with 16-week or 25-week HFD were examined by immunoblot analyses. Actin is the protein loading control

Fig. 2

Impaired cardiac function in CKO mice with 16-week HFD. a Body weights of Flox and CKO mice fed with chow diet or HFD were recorded at 4-week intervals. (*P < 0.05, vs. chow groups, n = 11 mice per group). b Quantitative analysis of inguinal fat-pad mass. (n = 6 mice per group). c Diastolic and systolic blood pressure, d fasting blood glucose, e insulin, and f cholesterol were gradually increased with HFD for 16 weeks in both Flox and CKO mice. *P < 0.05, vs. chow groups, n = 9 mice per group. g Echocardiographic assessment of FS%, dPW, LVEDd, and LVESd by M-mode images and quantification demonstrated the impaired cardiac function in HFD-CKO mice. n = 12 mice per group. Data are presented as means  ± SD

Fig. 3

Mitochondrial morphological and functional aberrations in HFD-CKO hearts. a The content of mtDNA was measured by the ratio of Cox1 to cyclophilin A. b Mitochondria was examined by TEM (×1400). Representative images showing reduced and disorganized mitochondria cluttering in the cytosol of HFD-CKO hearts (scale bar: 2 µm). The quantitative analysis of mitochondria number is provided in bar graphs. c Higher magnification (×4800) of TEM images show disarrayed cristae, vacuoles and a reduced electron density of the matrix in the HFD-CKO mitochondria (scale bar: 0.5 µm), M symbolize mitochondria, arrows indicate lipid droplets. Quantitative analysis of cristae area, disorganized cristae in mitochondria, lipid droplet number, respectively, is provided in the bar graphs (n = 6 mice per group). d Respiratory rates of mitochondrial ETC complexes were measured in saponin-permeabilized myocardial fibers. With glutamate/malate as substrates, complex I respiratory rate was found to be decreased in HFD-CKO hearts. e Significantly reduced ATP production in HFD-CKO hearts was determined. Consistent with this, the respiratory control ratio (RCR) was lower in HFD-CKO hearts. f Analysis of complex I in mitochondria homogenates using palmitoylcarnitine as a substrate showed that ERK5-deficient mitochondria of HFD-fed mice had a reduction in FA-driven respiration. g Mitochondrial fuel oxidation was suppressed in HFD-CKO hearts evidenced by activity measurement of β-hydroxylacyl CoA dehydrogenase and pyruvate dehydrogenase (Pdh) (n = 7 mice per group). Data are presented as means  ± SD

Fig. 4

Loss of Erk5 caused cellular injury in lipids overloaded hearts. a FFA extracted from heart tissues, and b Heart sections stained with Oil Red O, indicated that neutral lipids were increased in HFD-CKO hearts (scale bar: 20 μm). c Measurement of DAG content demonstrated an accumulation of its level in HFD-CKO hearts. d Triple quadrupole mass spectrometer analysis of ceramide species showed an altered ceramide species profile in CKO hearts. ^# P < 0.05 vs. chow-Flox, ^§ P < 0.05 vs. HFD-Flox, *P < 0.05 vs. chow-CKO, n = 7 mice per group. e DHE staining determined more superoxide generation in HFD-CKO hearts (left panel, scale bar: 20 μm). Fluorescence intensity was quantified (right panel). f TUNEL assay by triple staining with DAPI (blue), anti-α-actinin antibody (red), and TUNEL (green) detected apparent apoptosis in HFD-CKO hearts (left panel, scale bar: 20 μm), arrows indicated TUNEL positive nuclei. The quantification of TUNEL positive nuclei is shown in bar graphs (right panel), n = 6 mice per group. g Immunoblot analysis showed increased active caspase-3 in the hearts of HFD-CKO mice. The quantifications are represented by the bar graphs. h Immunoblot analysis showed decreased Irs1 in HFD-CKO hearts. The quantification is represented by the bar graphs, n = 6 mice. Actin is the protein loading control. Data are presented as means  ± SD

Fig. 5

Erk5 was required for mitochondrial function-related gene expression programmes. a Quantitative real-time PCR (qPCR) showed a decreased expression of Ppargc1a, Ppargc1b, Ppara, and Nrf1 in the hearts of HFD-CKO mice. The mRNA level of genes involved in FAO b, glycolytic action c, or mitochondrial biogenesis and OXPHOS d were assayed by qPCR. e The mRNA level of antioxidants was decreased in HFD-CKO hearts. Data are plotted as means  ± SD (*P < 0.05, vs. HFD group, n = 7 mice per group)

Fig. 6

Molecular basis underlying Erk5 regulation of Pgc-1α and Erk5 degradation. a Immunoblot analyses revealed the phosphorylation level and total expression of Erk5, Mef2a, Mef2c, Mef2d, and Pgc-1α in the hearts of Flox and CKO mice fed with chow or HFD for 16 weeks. b Immunoblot analyses showed increased activation of Erk5 in NRCMs treated with palmitate acid (PA, 500 µM) for 2 h; while the total expression of Erk5 is significantly decreased by PA for 8 h. c NRCMs were stained with DAPI (blue), anti-α-actinin antibody (green), and Mef2a (red) (scale bar: 20 μm), fluorescence intensity of Mef2a in nuclei was quantified. d Increased Pgc-1α reporter luciferase activity was detected with overexpression of Erk5 or MEef2a. The increased reporter activity induced by palmitate was blunted by knockdown of Erk5 in NRCMs. e ChIP analysis demonstrated that enhanced binding of Mef2a to the Ppargca (Pgc-1α) promoter region at two sites (TSS, transcriptional starting site, -1506 and -2868, marked in red) in HL-1 cardiomyocytes is Erk5-dependent. Sequence analysis of the purified PCR fragment bound by anti-Mef2a antibody confirmed the Mef2 consensus binding sequence. f Immunoblot analyses showed decreased Erk5 in ARCMs treated with 8 h PA or stearate acid (SA, 500 µM), but not by unsaturated FFA linoleate acid (LA, 500 µM) or oleate acid (OA, 500 µM). qPCR analysis showed no reduction in mRNA expression of Erk5 in PA or SA-treated ARCMs. g Immunoblot analyses showed that degradation of Erk5 was blocked by pretreatment of E-64D (25 μM, 4 h), but not by MG132 (10 μM, 4 h). Erk5 degradation was caused by calpain-1 and its expression could be restored by calpeptin (20 μM, 6 h), MDL-28170 (30 μM, 16 h) or Tempol (10 nM, 16 h), but not by cathepsin K inhibitor II (1 μM, 6 h). h Calpain-1 expression and its activity were measured. i Increased calpain activity, Nadph oxidase activity and DHE fluorescence intensity induced by PA were inhibited by pretreatment of apocynin (10 µM, 1 h) or gp91phox knockdown. Application of apocynin or Gp91phox knockdown restored Erk5 expression despite palmitate stimulation, n = 5–6 independent experiments per group. Actin is the protein loading control. Data are presented as means  ± SD

Fig. 7

Prevention of Erk5 degradation rescued mitochondrial functions in ARCMs. a Prevention of Erk5 reduction by calpain-1 knockdown or MDL-28170 (30 µM for 16 h) was able to retrieve the mRNA level of key mitochondrial genes reduced by 8 h PA treatment. * or ^# P < 0.05, vs. PA-treated group. b Mitochondrial oxidation activities of β-hydroxylacyl CoA dehydrogenase and pyruvate dehydrogenase (Pdh) were restored when calpain-1 was blocked. c FFA-induced oxidative damage was ameliorated followed by calpain-1 knockdown or inhibition, indicated by decreased protein carbonyl amount or active caspase-3 expression, respectively. n = 6 independent experiments per group. Actin is the protein loading control. Data are presented as means  ± SD

Fig. 8

Erk5 protects against metabolic stress-induced cardiomyopathy. a The expression of Erk5, Mef2a, and Mef2d were examined in the hearts from CKO mice injected with AAV9-GFP or AAV9-Erk5 followed by 16-week chow or HFD feeding. Actin is the protein loading control. b DHE staining indicated less ROS production in AAV9-Erk5-HFD hearts (left panel, scale bar: 20 μm). Fluorescence intensity was quantified (right panel). c TUNEL assay by triple staining with DAPI (blue), anti-α-actinin antibody (red), and TUNEL (green) determined fewer apoptosis in AAV9-Erk5 injected HFD-CKO hearts (left panel, scale bar: 20 μm), arrows indicated TUNEL positive nuclei. Quantification of TUNEL positive nuclei is represented by the bar graphs (right panel). d Masson’s trichrome staining detected less interstitial fibrosis in AAV9-Erk5 injected HFD-CKO hearts (left image, scale bar: 50 μm). Quantification of the fibrosis area is shown in the bar graphs (right panel). Restoration of Erk5 in CKO hearts improved e cardiac function and mitochondrial function evidenced by f increased the content of mtDNA, g enhanced complex I respiration rate, h elevated ATP production along with RCR, and i improved mitochondrial oxidation capability indicated by β-hydroxylacyl CoA dehydrogenase and pyruvate dehydrogenase (Pdh). n = 6 mice per group. Data are presented as means ± SD

Fig. 9

Schematic model of Erk5 protection against metabolic stress. FFA induces escalated ROS production from Gp91phox, which activates caplain-1, leading to a breakdown of Erk5. Loss of Erk5 and resultant Pgc-1α downregulation impose detrimental impacts on mitochondrial functions through accumulation of toxic lipids and vicious cycle of ROS. Prevention of Erk5 degradation or Erk5 restoration by AAV9 approach revives mitochondrial functions