Selective Loss of Catecholaminergic Wake–Active Neurons in a Murine Sleep Apnea Model

Irreversibility of hypersomnolence and sleep propensity in mice after long-term exposure to hypoxia/reoxygenation, modeling sleep apnea. A , Despite a recovery opportunity of 6 months, total wake times in 24 h (dark hatched bar) remain significantly less than age-matched control (sham treated; light hatched bar). NREMS, Non-rapid eye movement sleep. B , Multiple sleep latency test values remain reduced across unperturbed sleep–wake cycles at the end of the rest period and also after 6 h of sleep deprivation for the same circadian time periods. The asterisks denote significant (p < 0.05) differences relative to sham hypoxia/reoxygenation-exposed mice.

Long-term intermittent hypoxia impairs the c-fos response to wakefulness in select groups of wake neurons. A , Immunofluorescence in the ventral periaqueductal gray of anti-tyrosine hydroxylase (green) and anti-c-fos (red) in sham LTIH control (top two panels) mice and mice exposed to 8 weeks of LTIH (bottom two panels) after 3 h of spontaneous sleep–wake activity (left) and after 3 h of enforced wakefulness (right). The arrows highlight several c-fos-positive nuclei in TH-labeled neurons. B , Immunofluorescence in the perifornicular region of the hypothalamus of anti-orexin A (green) and c-fos (red). Arrows delineate orexin-labeled neurons with c-fos immunoreactivity in nuclei. C , Immunolabeling in the ventral tuberomamillary nuclei of anti-histidine decarboxylase (DAB; brown) and c-fos (DAB/nickel; black). The arrows highlight double-labeled neurons. D , Percentage of neurons per region with nuclear c-fos labeling after 8 weeks exposure to IH or control (room air). Data are presented for mice allowed 3 h of spontaneous sleep activity before perfusion (sleep) and mice kept awake with enriched environment for 3 h before perfusion (wake). The asterisks and bars denote ANOVA (significance, p < 0.05; n = 5/group). Scale bars, 25 μm.

Long-term intermittent hypoxia increases cleaved caspase-3, and this effect is partially blocked by treatment across IH with apocynin, an inhibitor of the catalytic (Nox2) NADPH oxidase subunit. A , Immunofluorescence of anti-TH (green) and anti-CC3 (red) in the locus ceruleus across conditions of sham LTIH or LTIH for durations of 8 and 24 weeks (sham LTIH 8 weeks, sham LTIH 24 weeks, LTIH 8 weeks, LTIH 24 weeks). The arrows denote anti-TH and anti-CC3-labeled neurons. At 24 weeks, anti-CC3 was evident in nuclei of <10% of noradrenergic neurons. A similar effect of LTIH across duration was observed for the dopaminergic ventral periaqueductal gray neurons. The bottom two panels show perifornicular hypothalamus anti-orexin A neurons (green) and anti-CC3 (red) in the same mouse shown above for anti-TH neurons. B , Effects of vehicle control (veh) (dimethyl sulfoxide; top panel) and apocynin treatment (apo) throughout LTIH exposure. The arrows highlight anti-CC3 immunoreactivity in TH+ neurons in the locus ceruleus. Scale bar, 25 μm (for all images). C , Individual data for relative CC-3 immunofluorescence intensity.

Effects of LTIH on cellular morphology of dopaminergic and noradrenergic wake neurons. The top two panels show ventrolateral periaqueductal gray neurons anti-TH-labeled Vector SG (blue) and Nissl background. LTIH-exposed neuron (top right panel) shows characteristic vacuolization in soma adjacent to dendrites and dendritic beading (arrows). The middle two panels compare dendrites of anti-TH-labeled neurons in locus ceruleus. The left panel shows prominent medial dendrites in sham LTIH mouse. In contrast, LTIH-exposed mouse shows loss of medial dendrites. The bottom panel shows adjacent region in the same mice, with preservation of orexinergic terminals into the dendritic region of the locus ceruleus. Scale bars, 50 μm.

Immunolocalization of NADPH oxidase subunits in dopaminergic and noradrenergic wake neurons. Ventral periaqueductal gray neurons labeled for catalytic subunit (Nox2) with anti-gp91^phox, silver enhanced gold and anti-TH immunoperoxidase labeled (DAB; brown) in 350 μm sections in sham LTIH, with subtle presence in dendrite (arrow) and in LTIH increased perinuclear and dendritic immunoreactivity (arrows). Locus ceruleus neurons with immunoreactivity for anti-p67^phox in anti-TH labeled neurons in sham LTIH (left) and LTIH (right). The arrows highlight clustering of anti-p67^phox labeling. Higher magnification shows intraneuronal labeling only (Figs. 1, 7). The bottom two panels show anti-p47^phox immunoreactivity in TH-labeled locus ceruleus neurons. Scale bars, 25 μm.

Ultrastructural localization of NADPH oxidase subunits in catecholaminergic wake neurons. Electron microscopic photomicrographs of noradrenergic locus ceruleus and dopaminergic VPAG neurons and dendrites labeled with anti-TH (stained with DAB) and cytosolic NADPH oxidase subunit, anti-p67^phox (stained with silver-intensified gold). A , TH-immunoperoxidase-labeled LC neuron (TH-s) in sham LTIH-treated mouse with well-preserved cytoarchitecture, ER, mitochondria (mit), and Golgi complex (G). B , TH-immunoperoxidase-labeled LC neuron (TH-s) in LTIH-treated mouse with swollen mitochondria (mit), increased lysozymes (Ly), and clustered p67^phox staining in ER around mitochondria (arrows) or on neuronal membrane. C–E , TH-immunoperoxidase-labeled dendrites (TH-d) in sham IH and LTIH-treated mice, with presence of silver-intensified gold staining of p67^phox on mitochondria and membranes ( D , E ). F , In a mouse exposed to sham LTIH, well-preserved regions of RER with rare p67^phox labeling were typical. G , In contrast, in LTIH-exposed mice, the RER showed, in regions of increased p67^phox, a haziness of the ribosomes and membranes. The arrows highlight p67^phox localization on RER. Scale bars: A , B , 1 μm; C–G , 500 ηm.