Tobacco cigarette smoking induces cerebrovascular dysfunction followed by oxidative neuronal injury with the onset of cognitive impairment

Abstract

While chronic smoking triggers cardiovascular disease, controversy remains regarding its effects on the brain and cognition. We investigated the effects of long-term cigarette smoke (CS) exposure (CSE) on cerebrovascular function, neuronal injury, and cognition in a novel mouse exposure model. Longitudinal studies were performed in CS or air-exposed mice, 2 hours/day, for up to 60 weeks. Hypertension and carotid vascular endothelial dysfunction (VED) occurred by 16 weeks of CSE, followed by reduced carotid artery blood flow, with oxidative stress detected in the carotid artery, and subsequently in the brain of CS-exposed mice with generation of reactive oxygen species (ROS) and secondary protein and DNA oxidation, microglial activation and astrocytosis. Brain small vessels exhibited decreased levels of endothelial NO synthase (eNOS), enlarged perivascular spaces with blood brain barrier (BBB) leak and decreased levels of tight-junction proteins. In the brain, amyloid-β deposition and phosphorylated-tau were detected with increases out to 60 weeks, at which time mice exhibited impaired spatial learning and memory. Thus, long-term CSE initiates a cascade of ROS generation and oxidative damage, eNOS dysfunction with cerebral hypoperfusion, as well as cerebrovascular and BBB damage with intracerebral inflammation, and neuronal degeneration, followed by the onset of impaired cognition and memory.

Keywords: Cerebrovascular disease; free radicals; inflammation; nitric oxide; oxidative stress.

Figure 1.

Cigarette smoke exposure causes vascular oxidative damage, hypertension, vascular endothelial dysfunction, and decreased carotid blood flow. (a–b) Eight weeks of cigarette smoke (CS) exposure elevated systolic and diastolic blood pressure (BP) with progressive increase out to 60 weeks, compared to air-exposed controls (Air). (c–d) Carotid artery rings were pre-constricted with 1 µM phenylephrine (PE), and the concentration-dependent relaxation effect of acetylcholine was measured. 16 weeks of CS-exposure impaired vessel relaxation compared to Air, and this was more pronounced at 60 weeks. (e–f) Reduced carotid blood velocity and flow were seen with 30 weeks of CS exposure and further decreased with 60 weeks of exposure, compared to Air. (g–h) CS exposure caused time-dependent increase of protein carbonylation (PC) and malondialdehyde (MDA) levels in carotid artery homogenate, as measured by ELISA, compared to Air. Data are mean ± SD with n = 10 for (a–b), n = 6 for (c–d), n = 5 for (e–f), and n = 6 for (g–h). Analysis for (a–b), and (e–h) was done by two-way ANOVA followed by Bonferroni multiple comparison test or ANOVA with repeated measures for (c–d). Data show statistical significance from: Air (*) at p < 0.05; or from 8 weeks (#), 16 weeks (@), and 30 weeks of CS exposure ($) at p < 0.05.

Figure 2.

Cigarette smoke exposure decreases endothelial nitric oxide synthase (eNOS) and nitric oxide (NO) levels in the brain vessels and homogenate. Mice were exposed to cigarette smoke (CS) or filtered air (Air) for 60 weeks. (a) Western blotting measurement of eNOS in brain homogenate showing lower expression of eNOS in CS-exposed mice. (b) eNOS band density was quantitated and normalized to corresponding GAPDH bands. (c) Immunofluorescence staining of brain sections for eNOS (green) showing lower level in the brain vessels of CS-exposed mice compared to Air. DAPI (blue) was used to visualize the nuclei. A control in the absence of eNOS primary antibody is shown in the right panel. (d) Quantitation of eNOS fluorescence in (c). (e) Detection of NO in brain sections using DAF-FM fluorescent probe with brain vessels showing NO-derived green fluorescence. The right 2 panels show that the NO scavenger, PTIO, abolished the NO-derived fluorescence, while in the absence of DAF there was minimal background autofluorescence and (f) quantitation of NO-derived green fluorescence in (e). CS significantly decreased eNOS and NO levels in the brain, compared to Air. Scale bar (white line) is 20 μm with identical magnification in all panels. Data are mean ± SD with n = 3 for (b), and n = 6 for (d) & (f). Analysis was done by unpaired t-test. *denotes significance from Air at p < 0.05.

Figure 3.

Cigarette smoke exposure causes cerebrovascular and blood-brain barrier (BBB) dysfunction. Mice were exposed to cigarette smoke (CS) or filtered air (Air) for 60 weeks. (a) Brain tissue was sectioned and stained with Haematoxylin & Eosin. Left photo shows normal vessel structures in Air, compared to enlarged perivascular space (black arrows in middle and right photos) and congested vessels (dotted circle in left photo) following CS exposure. Scale bar in black is 20 μm with identical magnification in all panels. (b) Immunofluorescence staining of brain tissue. CS-exposed mice show increased extravasation of albumin and fibrinogen into the vessel wall and surrounding brain tissue. eNOS was used as an endothelial marker. Albumin and fibrinogen were co-localized with eNOS (Merge+DAPI). DAPI (blue) was used to visualize the nuclei. Scale bar in white is 20 μm with identical magnification in all panels. (c) Immunofluorescence staining of brain tissue showing that CS decreased the levels of the BBB tight junction proteins Occludin and Claudin-5 in the endothelium of cerebral small vessels compared to Air. CD31 was used as an endothelial marker. Occludin and Claudin-5 were co-localized with CD31 (Merge+DAPI). DAPI (blue) was used to visualize the nuclei. Scale bar in white is 20 μm with identical magnification in all panels. (d) Quantitation of Albumin and Fibrinogen fluorescence in (b). (e) Quantitation of Occludin, and Claudin-5 fluorescence in (c). (f–g) Western blotting showing decreased levels of Occludin (53 kDa) and Claudin-5 (22 kDa) in brain homogenate with band density quantitation and normalization to corresponding GAPDH bands. Data shown are mean ± SD with n = 6 for (d–e) and n = 3 for (f–g). Analysis was done by unpaired t-test. *: denotes significance from Air at p < 0.05.

Figure 4.

Cigarette smoke exposure causes reactive oxygen species generation and oxidative damage in the brain. Mice were exposed to cigarette smoke (CS) or filtered air (Air) for 16 or 60 weeks. (a–b) Show elevated levels of protein carbonyl (PC), and malondialdehyde (MDA) in CS-exposed mice as measured in brain homogenates by ELISA. (c) Sections of cortex and hippocampus show superoxide production (ROS; red), using dihydroethidium (DHE). Superoxide dismutase mimetic (SODm) was used to confirm the specificity of the DHE fluorescence from 60 weeks of CS exposure. Immunofluorescence staining of nitrotyrosine (e; green) and 8-hydroxyguanosine (g; 8-OHG; red). DAPI (blue) was used to visualize the nuclei. From series of experiments, fluorescence in (c), (e) & (g) were quantitated with data presented in (d), (f) & (h), respectively. CS increased PC and MDA levels in the brain with increased generation of ROS, nitrotyrosine, and 8-OHG in the cortex and hippocampus of 60 week CS-exposed mice; with similar but lower increase at 16 weeks of exposure. Scale bar shown corresponds to 50 μm with identical magnification in all panels. Data are mean ± SD with n = 6. Analysis was done by two-way ANOVA followed by Bonferroni multiple comparison test. Data show statistical significance from: Air (*); and 16 weeks of CS exposure (#) at p < 0.05.

Figure 5.

Cigarette smoke exposure increases microglial activation and astrocytosis in the brain. Mice were exposed to cigarette smoke (CS) or filtered air (Air) for 16 or 60 weeks. Brain sections were incubated with antibodies against microglial activation and astrocytosis markers Iba1 (a; red) and GFAP (c; green), respectively. DAPI (blue) was used to visualize the nuclei. Iba1 and GFAP images were analyzed, and fluorescence was quantitated and presented in (b) and (d), respectively. Scale bar corresponds to 50 μm with identical magnification in all panels. Data are mean ± SD with n = 6. Analysis was done by two-way ANOVA followed by Bonferroni multiple comparison test. Data show statistical significance from: Air (*) and 16 weeks of CS exposure (#) at p < 0.05.

Figure 6.

Cigarette smoke exposure induces amylogenesis and tau phosphorylation in the brain. Mice were exposed to cigarette smoke (CS) or filtered air (Air) for 16 or 60 weeks. Sections from cortex and hippocampus were incubated with antibodies to amyloid β (a; Aβ; green) or phosphorylated tau (f; P-tau; red). DAPI (blue) was used to visualize the nuclei. P-tau and Aβ images were analyzed, and fluorescence was quantitated and presented in (b) and (g), respectively. Amyloid fragments, Aβ40 and Aβ42, were quantitated in brain homogenate using ELISA (c) & (d), respectively. (e) Western blotting of P-tau (75 kDa) from brain homogenate of mice exposed to 60 weeks of CS or Air, with band density quantitation. (h) Protein aggregates (green) detected with Thioflavin S, with quantitation of the green fluorescence (i). CS caused a modest increase in p-tau and Aβ at 16 weeks, with greater elevation at 60 weeks that was accompanied by appearance of Aβ/p-tau protein aggregates. Scale bar corresponds to 50 μm with all accompanying images in a panel shown at the same magnification. Data are means ± SD with n = 6 for all panels except (e) with n = 3. Analysis was done by two-way ANOVA followed by Bonferroni multiple comparison test. Data show statistical significance from: Air (*) and 16 weeks of CS exposure (#) at p < 0.05.

Figure 7.

Cognitive impairment following cigarette smoke exposure. Mice were exposed to cigarette smoke (CS) or filtered air (Air) for 30 or 60 weeks. Spatial learning and memory were evaluated by Barnes maze (a–f). Contextual and cued fear memories were evaluated using contextual and cued fear conditioning test (g–h). (a–c) Acquisition phase at 60 weeks of CS. (d–f) Probe test at 30 and 60 weeks of exposure. (g) Freezing response during presentation of conditioning context (Context recall) and an auditory cue (Cue recall) on the second and third testing days as depicted in the bottom panel (h). Dotted line in (g) presents the freezing response of air-exposed control to the contextual and cued recall as 100%. CS data shown as bars is relative to the response shown as dotted line. At 30 weeks of exposure, there were no significant differences in all tested parameters. However, we observed cognitive impairment at 60 weeks of CS exposure, shown as longer escape latency (a) & (d), longer path length (b) & (e), higher number of errors and less time spent in proximity (c) & (f), and less times of freezing (g), compared to Air. Data are mean ± SD with n = 10 per group. Analysis was done by two-way ANOVA followed by Bonferroni multiple comparison test. While CS exposure of 30 weeks was not sufficient to cause impairment, with 60 weeks of CS exposure both learning and memory were impaired. Data show statistical significance from: Air (*) and 30 weeks of CS exposure (#) at p < 0.05.