Crucial role for Nox2 and sleep deprivation in aircraft noise-induced vascular and cerebral oxidative stress, inflammation, and gene regulation

Abstract

Aims: Aircraft noise causes endothelial dysfunction, oxidative stress, and inflammation. Transportation noise increases the incidence of coronary artery disease, hypertension, and stroke. The underlying mechanisms are not well understood. Herein, we investigated effects of phagocyte-type NADPH oxidase (Nox2) knockout and different noise protocols (around-the-clock, sleep/awake phase noise) on vascular and cerebral complications in mice.

Methods and results: C57BL/6j and Nox2-/- (gp91phox-/-) mice were exposed to aircraft noise (maximum sound level of 85 dB(A), average sound pressure level of 72 dB(A)) around-the-clock or during sleep/awake phases for 1, 2, and 4 days. Adverse effects of around-the-clock noise on the vasculature and brain were mostly prevented by Nox2 deficiency. Around-the-clock aircraft noise of the mice caused the most pronounced vascular effects and dysregulation of Foxo3/circadian clock as revealed by next generation sequencing (NGS), suggesting impaired sleep quality in exposed mice. Accordingly, sleep but not awake phase noise caused increased blood pressure, endothelial dysfunction, increased markers of vascular/systemic oxidative stress, and inflammation. Noise also caused cerebral oxidative stress and inflammation, endothelial and neuronal nitric oxide synthase (e/nNOS) uncoupling, nNOS mRNA and protein down-regulation, and Nox2 activation. NGS revealed similarities in adverse gene regulation between around-the-clock and sleep phase noise. In patients with established coronary artery disease, night-time aircraft noise increased oxidative stress, and inflammation biomarkers in serum.

Conclusion: Aircraft noise increases vascular and cerebral oxidative stress via Nox2. Sleep deprivation and/or fragmentation caused by noise triggers vascular dysfunction. Thus, preventive measures that reduce night-time aircraft noise are warranted.

Figure 1

Effects of continuous noise exposure (24 h/day) on blood glucose, vascular function, systemic oxidative stress, and gene regulation (next generation sequencing data) in wild-type and Nox2^−/− (gp91phox^−/−) mice. Around the clock noise exposure (24 h/day) impaired endothelial function significantly (A and B) and increased blood glucose (C) in wild-type mice. The endothelium-independent relaxation (NTG response) is shown in Supplementary material online, Figure S1. Levels of 3-nitrotyrosine- and malondialdehyde-positive proteins were significantly increased in mouse plasma in wild-type mice (D and E). Circulating IL-6 in mouse plasma was significantly elevated in wild-type mice (F). Representative dot blots are shown in Supplementary material online, Figure S2. Vascular ROS formation as envisaged by DHE staining was increased in noise-exposed wild-type mice (G). None of the parameters were changed in around-the-clock noise exposed Nox2^−/− (gp91phox^−/−) mice. Noise-induced aortic superoxide formation as measured by HPLC analysis was higher by trend in wild type than in Nox2^−/− (gp91phox^−/−) mice (H). Representative HPLC chromatograms are shown in Supplementary material online, Figure S24. The pie chart and box/scatter plots summarize the number of genes that are significantly and tissue-independently regulated (I). Changes of transcriptional profiles were determined by next generation sequencing in aorta, heart, and kidney of mice exposed to aircraft noise and unexposed controls. Clustering of genes in heat maps revealed tissue-conserved transcriptional changes in the circadian clock pathway (J). Most significantly up- or down-regulated genes for all tissues can be found in Supplementary material online, Tables S1–S5. Data are presented as mean ± SD from n = 14–15 (A and B) or 6 (C) animals/group, at least three samples/group (pooled from 2 to 3 mice/sample) (D–F), 4–6 WT and 7–8 gp91phox^−/− (G), and 4 mice/group (H–J). (A and B) Statistical analysis was performed using two-way ANOVA comparing the values of the entire curves; *P < 0.05 vs. control without noise at the same ACh concentration (as indicated by colour code). (H) Unpaired t-test with Welch’s correction was used. For all other data: normality test passed and one-way ANOVA with Tukey’s correction was used. lfcMLE means Log2 fold change with the unshrunken maximum likelihood estimate (MLE).

Figure 2

Effects of continuous noise (24 h/day) on cerebral oxidative stress and inflammation in wild type and Nox2^−/− (gp91phox^−/−) mice. Around the clock noise (24 h/day) increased oxidative stress in the cortex of wild-type mice significantly, whereas Nox2^−/− (gp91phox^−/−) mice were protected on the first two exposure days (A). Modulation of the ROS signal by L-NAME was used to identify ROS formation coming from uncoupled NOS (B). Representative minimized cerebral cryo-stainings of DHE fluorescence microtopography are shown below the quantification. Supplementary material online, Figures S24 and S25 show the full images and higher magnification as well as stainings of the entire brain. Modulation of the ROS signal by ARL-17477 was used to identify ROS formation coming from uncoupled nNOS (C). Cerebral nNOS expression was quantified at the protein and mRNA level in wild type and Nox2 deficient mice (D and E). Representative original western blots are shown in Supplementary material online, Figure S24. Noise-induced superoxide formation in frontal cortex as measured by HPLC analysis was significantly higher in WT than in Nox2^−/− (gp91phox^−/−) mice and whole brain tissue without frontal cortex showed the same trend (F and G). Representative HPLC chromatograms and full images of the representative DHE stainings are shown in Supplementary material online, Figure S24, respectively. Markers of inflammation (iNOS, CD68, IL-6) were increased and catalase was decreased at the mRNA level by noise, all of which was prevented by Nox2 deficiency (H). Cryo-sections of the frontal cortex were stained with mitoSOX red (1 µM) for detection of mitochondrial ROS formation (I). Representative full size staining images for all groups are shown in Supplementary material online, Figure S10. Data are presented as mean ± SD from n = 8 (A and B), at least 5 (C and D), 4 (F–H), and 7–8 (I) mice/group. (I) Normality test failed and non-parametric Kruskal–Wallis test with Dunn’s correction was used. (C) Paired and (F, G) unpaired t-test was used. For all other data: normality test passed and one-way ANOVA with Tukey’s correction was used.

Figure 3

Effects of noise during the sleep or awake phase (12 h/day) on gene expression (NGS). The pie chart as well as box/scatter plots summarize the number of genes that are significantly and noise protocol-independently regulated in aortic tissue (A). NGS data for the three noise exposure protocols were used to identify genes that are similarly regulated in response to around-the-clock or sleep phase aircraft noise for 4 days (B). Most significantly up- or down-regulated genes (in aorta) for awake and sleep phase as well as around-the-clock noise can be found in Supplementary material online, Tables S1–S3 and S6–S9. Data are presented as mean ± SD from four samples in control group, five samples in sleep or awake group (each pooled from four mice) (B). Another 27 genes with high regulatory similarity between around-the-clock or sleep phase aircraft noise for 4 days and a summary of these comparisons can be found in Supplementary material online, Figures S12 and S13. (B, Ihh): Normality test failed and non-parametric Kruskal–Wallis test with Dunn’s correction was used. For all other data: normality test passed and one-way ANOVA with Tukey’s correction was used.

Figure 4

Effects of noise during the sleep or awake phase (12 h/day) on blood pressure, vascular function, endothelin-1 levels, and vascular oxidative stress. Systolic and diastolic blood pressure was increased after sleep and awake phase noise. The increase of blood pressure was more pronounced after sleep phase noise (A and B). Time-dependent changes of blood pressure for 1, 2, 3, and 4 days of noise are shown in Supplementary material online, Figure S14. Relaxation by the endothelium-dependent vasodilator acetylcholine (ACh) was impaired by sleep phase but not awake phase noise (C and D). The dashed lines are reproduced data of Figure 1A above as reference values of mice exposed to around-the-clock aircraft noise for 4 days. The endothelium-independent relaxation (NTG response) is shown in Supplementary material online, Figure S15. Noise during the sleep phase but not the awake phase caused an increase in the expression ET-1 in aortic tissue (E). Immunohistochemical analysis revealed that noise during the sleep phase but not the awake phase enhanced vascular ET-1 levels on Day 1, mainly within the endothelium (F). Representative stained images and original blots are shown in Supplementary material online, Figure S16. In vascular tissue, sleep but not awake phase noise increased ROS production as revealed by DHE cryo staining (G, all staining images and quantification in Supplementary material online, Figure S19). Noise-induced aortic superoxide formation as measured by HPLC analysis was higher in the sleep than in the awake phase noise group (H). Representative HPLC chromatograms are shown in Supplementary material online, Figure S26. ROS production in the endothelial cell layer was increased in aortic segments of animals exposed to sleep phase noise but not awake phase noise and treatment with the eNOS inhibitor L-NAME exerted opposing effects on endothelial superoxide production in the sleep (rather down) vs. awake (rather up) phase noise groups on all days (I). Sleep phase but not awake phase noise lead to a substantial increase in S-glutathionylation of eNOS (J). Representative stained images and western blots are shown below the densitometric quantification. Data are presented as mean ± SD from n = 5 (A and B), 15–20 (C and D) mice per group, 6–8 samples (pooled from 2 to 3 mice per sample) (E), and 6–8 (F), 4–6 (H), and 6–10 (I, J) mice/group. Two-way ANOVA (C and D); *P < 0.05 vs. control without noise at the same ACh concentration (as indicated by color code). (A and I) Normality test failed and non-parametric Kruskal–Wallis test with Dunn’s correction was used. (H) Unpaired t-test was used. For all other data: normality test passed and one-way ANOVA with Tukey’s correction was used.

Figure 5

Effects of sleep and awake phase noise (12 h/day for 4 days) on cerebral oxidative stress, inflammation, and gene regulation. Sleep phase noise caused a more pronounced increase in ROS formation in the frontal cortex than awake phase noise and this signal was blocked by Nox inhibition (DHE staining for cytoplasmic ROS and mitoSOX red for mitochondrial ROS, A). Representative stained images are shown together with the densitometric quantification. Noise-induced cerebral superoxide formation as measured by HPLC analysis was higher in the sleep than in the awake phase noise group (B). Representative HPLC chromatograms are shown in Supplementary material online, Figure S26. Astrocytes were activated in sleep phase noise exposed mice in the corpus callosum (C). Representative images for immunohistochemical staining for GFAP are shown at the level of the hippocampus sector CA1 [corpus callosum (CC), stratum oriens (Or), pyramidale (Py), and radiatum (Rad); 4 mice/group]. Markers of inflammation (iNOS, CD68 significant and IL-6, MCP-1 by trend) and ROS-producing Nox1 were increased (at least by trend), whereas the antioxidant/protective genes catalase, Foxo3, and nNOS were decreased at the mRNA level by sleep phase noise (D). Data are presented as mean ± SD from at least n = 6 (A), 9–10 (B), and at least 4 (D) mice/group. (D, Catalase): normality test failed and non-parametric Kruskal–Wallis test with Dunn’s correction was used. (B) Unpaired t-test was used. For all other data: normality test passed and one-way ANOVA with Tukey’s correction was used.

Figure 6

Oxidative stress markers in serum of noise-exposed human subjects. ELISA measurements revealed increased 3-nitrotyrosine (3NT)-positive proteins (A) and 8-isoprostane concentrations (B) in serum of human subjects. Multiplex measurements revealed increased markers of inflammation (IL-1β by trend, (C) IL-18 significant, (D)) in serum of human subjects. Data are presented as mean ± SD. Samples from n = 18 individuals (A and B) were compared before and after noise (aircraft noise for 6 h, mean SPL 47 dB(A)) using the Wilcoxon matched-pairs signed rank test. Cytokine measurements yielded results for n = 10–12 individuals using the unpaired t-test (without (IL-1β) and with Welch’s correction (IL-18)) (C and D). (E) Summarizing central scheme: around-the-clock and sleep phase noise triggers cerebral oxidative stress and a neuroinflammatory phenotype that translates the adverse effects of noise to the vascular and systemic level (e.g. by adverse stress hormone signalling and dysregulation of circadian clock inducing changes in key signalling pathways). Noise via neuronal pathways triggers vascular oxidative stress and inflammation with subsequent endothelial dysfunction, increases in blood pressure, all of which contributes the development and progression of cardiometabolic disease.