Endothelin 1-dependent neurovascular dysfunction in chronic intermittent hypoxia

Abstract

Obstructive sleep apnea, a condition resulting in chronic intermittent hypoxia (CIH), is an independent risk factor for stroke and dementia, but the mechanisms of the effect are unknown. We tested the hypothesis that CIH increases cerebrovascular risk by altering critical mechanisms regulating cerebral blood flow thereby lowering cerebrovascular reserves. Male C57Bl6/J mice were subjected to CIH (10% O(2) for 90 seconds/room air for 90 seconds; during sleep hours) or sham treatment for 35 days. Somatosensory cortex blood flow was assessed by laser Doppler flowmetry in anesthetized mice equipped with a cranial window. CIH increased mean arterial pressure (from 74±2 to 83±3 mm Hg; P<0.05) and attenuated the blood flow increase produced by neural activity (whisker stimulation; -39±2%; P<0.05) or neocortical application of endothelium-dependent vasodilators (acetylcholine response: -41±3%; P<0.05). The cerebrovascular dysfunction was associated with oxidative stress in cerebral resistance arterioles and was abrogated by free radical scavenging or NADPH oxidase inhibition. Furthermore, cerebrovascular dysfunction and free radical increase were not observed in mice lacking the NOX2 subunit of NADPH oxidase. CIH markedly increased endothelin 1 in cerebral blood vessels, whereas cerebrovascular dysfunction and oxidative stress were abrogated by neocortical application of the endothelin type A receptor antagonist BQ123. These data demonstrate for the first time that CIH alters key regulatory mechanisms of the cerebral circulation through endothelin 1 and NADPH oxidase-derived radicals. The ensuing cerebrovascular dysfunction may increase stroke risk in patients with sleep apnea by reducing cerebrovascular reserves and increasing the brain's susceptibility to cerebral ischemia.

Figure 1

Chronic intermittent hypoxia (CIH) increases mean arterial pressure (MAP; A) and attenuates the increase in cerebral blood flow (CBF) produced by whisker stimulation (B), acetylcholine (ACh; C), bradykinin (D), and A23187 (E) but not adenosine (F). Note that the cerebrovascular effects are already present at 14 days, when MAP is not elevated. *P<0.05 from respective sham group, ANOVA and Tukey test; n=5 per group; □, 14 days; ■, 35 days.

Figure 2

Chronic intermittent hypoxia (CIH; 35 days) increases reactive oxygen species (ROS) in blood vessels (CD31⁺) and neurons (NeuN⁺) of the somatosensory cortex (A and B). The ROS increase is not observed in NOX2^−/− mice and is suppressed by gp91ds-tat (gp91ds) or BQ123 but not by a scrambled gp91ds peptide (s-gp91ds; C); *P<0.05 from sham group; n=5 per group. □, sham; ■, CIH.

Figure 3

The reactive oxygen species (ROS) scavenger Mn(III)tetrakis(4-benzoic acid)porphyrin Chloride reverses the chronic intermittent hypoxia (CIH)–induced suppression of the cerebral blood flow (CBF) response to whisker stimulation (A) or acetylcholine (ACh; B). CIH does not increase mean arterial pressure (C) and does not alter CBF responses to whisker stimulation (D) or ACh (E) in NOX2^−/− mice. The NADPH oxidase peptide inhibitor gp91ds-tat (gp91ds) but not its scrambled control (s-gp91ds) reverses the CIH-induced attenuation of CBF response to whisker stimulation (F) or ACh (G). *P<0.05 from respective sham group; n=5 per group. A and B, □, vehicle; ■, Mn(III)tetrakis(4-benzoic acid)porphyrin Chloride. C through E, □, wild-type (WT); ■, NOX2^−/−; F and G, □, vehicle; ▩, gp91ds; ■, s-gp91ds.

Figure 4

Losartan does not rescue the chronic intermittent hypoxia (CIH)–induced attenuation of the cerebral blood flow (CBF) response to whisker stimulation (A, □, vehicle; ■, losartan) or acetylcholine (ACh; B, □, vehicle; ■, BQ123), whereas BQ123 rescues the dysfunction in full (C and D). *P<0.05 from respective sham group; n=5 per group.

Figure 5

Chronic intermittent hypoxia (CIH) increases endothelin (ET) 1 immunoreactivity in CD31⁺ pial arterioles of the somatosensory cortex (A). CIH increases ET1, assessed by ELISA (B), as well as endothelin-converting enzyme 1 (ECE-1; C), ET_A (D), and ET_B (E) receptor mRNA in pial vessels preparations. *P<0.05 from sham group; n=5 per group.