Aerobic exercise training protects against endothelial dysfunction by increasing nitric oxide and hydrogen peroxide production in LDL receptor-deficient mice

Abstract

Background: Endothelial dysfunction associated with hypercholesterolemia is an early event in atherosclerosis characterized by redox imbalance associated with high superoxide production and reduced nitric oxide (NO) and hydrogen peroxide (H2O2) production. Aerobic exercise training (AET) has been demonstrated to ameliorate atherosclerotic lesions and oxidative stress in advanced atherosclerosis. However, whether AET protects against the early mechanisms of endothelial dysfunction in familial hypercholesterolemia remains unclear. This study investigated the effects of AET on endothelial dysfunction and vascular redox status in the aortas of LDL receptor knockout mice (LDLr(-/-)), a genetic model of familial hypercholesterolemia.

Methods: Twelve-week-old C57BL/6J (WT) and LDLr(-/-) mice were divided into sedentary and exercised (AET on a treadmill 1 h/5 × per week) groups for 4 weeks. Changes in lipid profiles, endothelial function, and aortic NO, H2O2 and superoxide production were examined.

Results: Total cholesterol and triglycerides were increased in sedentary and exercised LDLr(-/-) mice. Endothelium-dependent relaxation induced by acetylcholine was impaired in aortas of sedentary LDLr(-/-) mice but not in the exercised group. Inhibition of NO synthase (NOS) activity or H2O2 decomposition by catalase abolished the differences in the acetylcholine response between the animals. No changes were noted in the relaxation response induced by NO donor sodium nitroprusside or H2O2. Neuronal NOS expression and endothelial NOS phosphorylation (Ser1177), as well as NO and H2O2 production, were reduced in aortas of sedentary LDLr(-/-) mice and restored by AET. Incubation with apocynin increased acetylcholine-induced relaxation in sedentary, but not exercised LDLr(-/-) mice, suggesting a minor participation of NADPH oxidase in the endothelium-dependent relaxation after AET. Consistent with these findings, Nox2 expression and superoxide production were reduced in the aortas of exercised compared to sedentary LDLr(-/-) mice. Furthermore, the aortas of sedentary LDLr(-/-) mice showed reduced expression of superoxide dismutase (SOD) isoforms and minor participation of Cu/Zn-dependent SODs in acetylcholine-induced, endothelium-dependent relaxation, abnormalities that were partially attenuated in exercised LDLr(-/-) mice.

Conclusion: The data gathered by this study suggest AET as a potential non-pharmacological therapy in the prevention of very early endothelial dysfunction and redox imbalance in familial hypercholesterolemia via increases in NO bioavailability and H2O2 production.

Keywords: Aerobic exercise training; Endothelial dysfunction; Familial hypercholesterolemia; Hydrogen peroxide; LDL receptor-deficient mice; Nitric oxide synthase; Superoxide dismutase.

Fig. 1

Concentration-response curves to acetylcholine (A), to sodium nitroprusside (B) and to hydrogen peroxide (H₂O₂, C) in thoracic aortic rings from sedentary (S) and exercise-trained (Ex) wild-type (WT) and LDLr knockout mice (LDLr^−/−). Data are mean ± SEM (n = 5–14 per group). 2-way ANOVA

Fig. 2

Concentration-response curves to acetylcholine before and after incubation with L-NAME (LN, A and B), 7-Nitroindazole (7-NI, D and E) or catalase (CAT, G and H) in thoracic aortic rings from sedentary (S) and exercise-trained (Ex) wild-type (WT) and LDLr knockout mice (LDLr^−/−). Bar graphs show the acetylcholine maximal vasodilator effect (E_max) in the presence or absence of L-NAME (C), 7-NI (F) and CAT (I). ANOVA: * p < 0.05 vs. WT S;^# p < 0.05 vs. LDLr^−/− S; + indicates p < 0.05, ++ indicates p < 0.001, and +++ indicates p < 0.0001 vs. without incubation ring. Data are mean ± SEM (n = 5–8 per group)

Fig. 3

Representative fluorographs (A, left panel) and quantitative analysis (A, bar graphic right) of NO production evaluated by DAF-2A fluorescence in transverse sections of thoracic aorta. DAF-2A fluorescence is expressed as the percentage of intensity per vessel area obtained in WT S group (bar scale = 100 μm). The bar graph (B) represents the fluorescence curve slope to Amplex Red estimating the production of H₂O₂ per minute in aortic rings from sedentary (S) and exercise-trained (Ex) wild-type (WT) and LDLr knockout mice (LDLr^−/−). Data are mean ± SEM. Two-way ANOVA: * p < 0.05 vs. WT S; ^$ p < 0.05 vs. WT Ex; ^# p < 0.05 vs. LDLr^−/− S. Numbers into the bars represent N of animals used in each group

Fig. 4

Representative blots (top) and quantitative protein expression (bottom) of total and phosphorylated eNOS (p-eNOS) at Ser1177 (A); dimerized eNOS (B); total nNOS (C) and catalase (D) in aorta from sedentary (S) and exercise-trained (Ex) wild-type (WT) and LDLr knockout mice (LDLr^−/−). Proteins expression were normalized to α-actin content in each sample, and protein expression of p-eNOS were normalized to total eNOS expression. eNOS dimerization was expressed as a ratio of dimer:monomer band intensity. The results were expressed as the percentage of the protein expression values obtained in WT S group. Data are mean ± SEM. Two-way ANOVA: * p < 0.05 vs. WT S; ^$ p < 0.05 vs. WT Ex; ^# p < 0.05 vs. LDLr^−/− S. Numbers into the bars represent N of animals used in each group

Fig. 5

Concentration-response curves to acetylcholine before and after incubation with diethyldithiocarbamate (DETCA, A and B), or apocynin (Apo, D and E) in thoracic aortic rings from sedentary (S) and exercise-trained (Ex) wild-type (WT) and LDLr knockout mice (LDLr^−/−). Bar graphs show the acetylcholine maximal vasodilator effect (E_max) in the presence or absence of DETCA (C) and Apo (F). ANOVA: * p < 0.05 vs. WT S;^# p < 0.05 vs. LDLr^−/− S; + indicates p < 0.05, ++ indicates p < 0.001, and +++ indicates p < 0.0001 vs. without incubation rings. Data are mean ± SEM (n = 5–8 per group)

Fig. 6

Representative blots (top) and quantitative protein expression (bottom) of Nox2 subunit of NADPH oxidase (A); CuZn-SOD (B); Mn-SOD (C) and EC-SOD (D) in aorta from sedentary (S) and exercise-trained (Ex) wild-type (WT) and LDLr knockout mice (LDLr^−/−). Proteins expression were normalized to α-actin content in each sample. The results were expressed as the percentage of the protein expression values obtained in WT S group. Data are mean ± SEM. Two-way ANOVA: * p < 0.05 vs. WT S; ^$ p < 0.05 vs. WT Ex; ^# p < 0.05 vs. LDLr^−/− S. Numbers into the bars represent N of animals used in each group

Fig. 7

Representative fluorographs (A) and quantitative analysis of reactive oxygen species production in transverse sections of aorta (B) evaluated by the ethidium-bromide-positive nuclei under basal conditions and after incubation with apocynin (30 mM), or with Mn(III) tetrakis1-methyl-4-pyridyl porphyrin pentachloride (MnTMPyP, 25 µM). Bar scale 100 μm. Sedentary and exercised (Ex) wild-type (WT) and LDLr knockout mice (LDLr^−/−). Data are mean ± SEM (n = 4–15 per group). 2-way ANOVA: * p < 0.05 vs. WT S; ^# p < 0.05 vs. LDLr^−/− S; + p < 0.05 apocynin vs. basal