Hypocretin/orexin influences chronic sleep disruption injury in the hippocampus

Abstract

Chronic sleep disruption is a risk factor for Alzheimer's disease (AD), yet mechanisms by which sleep disturbances might promote or exacerbate AD are not understood. Short-term sleep loss acutely increases hippocampal amyloid β (Aβ) in wild type (WT) mice and long-term sleep loss increases amyloid plaque in AD transgenic mouse models. Both effects can be influenced by the wake-promoting neuropeptide, hypocretin (HCRT), but whether HCRT influences amyloid accumulation independent of sleep and wake timing modulation remains unclear. Here, we induced chronic fragmentation of sleep (CFS) in WT and HCRT-deficient mice to elicit similar arousal indices, sleep bout lengths and sleep bout numbers in both genotypes. We then examined the roles of HCRT in CFS-induced hippocampal Aβ accumulation and injury. CFS in WT mice resulted in increased Aβ₄₂ in the hippocampus along with loss of cholinergic projections and loss of locus coeruleus neurons. Mice with HCRT deficiency conferred resistance to CFS Aβ₄₂ accumulation and loss of cholinergic projections in the hippocampus yet evidenced similar CFS-induced loss of locus coeruleus neurons. Collectively, the findings demonstrate specific roles for orexin in sleep disruption hippocampal injury.

Significance statement: Chronic fragmentation of sleep (CFS) occurs in common conditions, including sleep apnea syndromes and chronic pain disorders, yet CFS can induce neural injury. Our results demonstrate that under conditions of sleep fragmentation, hypocretin/orexin is essential for the accumulation of amyloid-β and loss of cholinergic projections in the hippocampus observed in response to CFS yet does not influence locus coeruleus neuron response to CFS.

Keywords: amyloid; chronic sleep disruption; chronic sleep loss; degeneration; septohippocampal cholinergic system.

FIGURE 1

Behavioral state effects of HCRT deficiency and CFS effects on arousal frequency in HRCT–/– and WT mice. Sets of mice of both studied genotypes (wild type, WT and hypocretin-deficient, HCRT–/–) and both sleep conditions, rested and chronic fragmentation of sleep (CFS) underwent 24 h electroencephalography (EEG) and electromyography (EMG) recordings to confirm baseline sleep/wake abnormalities in HCRT–/– mice and to determine the effectiveness of rotor platform movement on sleep consolidation. (A) Hourly percent (%) wake time for Rest WT (blue dots) and Rest HCRT–/– (red triangles) mice (mean ± SE, n = 5/group). Repeated measures ANOVA identified differences at Zeitgeber hours 10–16 and 18 and 19 (black line). (B) Group data for Rest WT and HCRT–/– mice for the average duration of wake bouts in the dark (lights-off) period expressed in minutes (mean ± SE) analyzed with an unpaired t-test, ***p < 0.001. (C) Number of arousals/h of sleep (averaged across 24 h) group data (mean ± SE, n = 5/group) for WT and HCRT–/– mice exposed to Rested (dark blue) or chronic fragmentation of sleep (CFS, light blue), n = 5/group. Data were analyzed as two-way ANOVA with selected post hoc comparisons (Sidak’s test), **p < 0.01; ****p < 0.0001. (D) A representative raw tracing of the effects of rotor table movement on the EEG and EMG signals in a mouse. Rotor-elicited arousals looked similar in both WT and HCRT–/– mice. The blue line depicts the time of the shaker table movement. 4-s epochs, demarcated with vertical lines in the fragment, were scored as non-rapid-eye-movement sleep (NR) and wakefulness (W). Time bar, 2 s. Over the EEG tracing, green boxes denote computerized program detection of delta waves (0–4 Hz), and pink boxes denote computer detection of theta waves (6–10 Hz).

FIGURE 2

Sleep disturbances from chronic sleep fragmentation are NREMS specific for both genotypes. (A–C) Percentages of time/24 h spent in Wake (A), NREMS (B), and REMS (C). Data are presented as mean ± SE (n = 5/group), for Rested, dark blue and CFS, light blue across the two genotypes. (D–F) Behavioral state bout lengths expressed as mean ± SE, and (G–I) bout numbers/24 h for the three behavioral states. Data in (A–E,H,I) were analyzed as two-way ANOVA, corrected for selected post hoc comparisons (Sidak’s test). Non-normal data in (F,G) were analyzed with Kruskall-Wallis and Dunn’s post hoc comparisons. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

FIGURE 3

HCRT–/– mice confer resistance to CFS Aβ₄₂ accumulation. (A) Representative confocal images within the CA1 hippocampus across the genotype and sleep conditions immunolabeled for Aβ₄₂ (green) and microglial marker Iba-1 (purple). Calibration bar, 75 μm. Subregions of CA1 are labeled: Oriens, Or; pyramidal neuron somata, Pyr; and Strata Radiatum, Str Rad. (B) Group data for percentage area CA1 Aβ₄₂, expressed as mean ± SE (n = 5–7/group) for Rested (dark blue) and CFS (light blue) for WT and HCRT–/– mice. (C) Group data for percentage area CA1 Iba-1, expressed as mean ± SE (n = 5–7/group) for Rested (dark blue) and CFS (light blue) for WT and HCRT–/– mice, analyzed as two-way ANOVA for sleep condition and genotype with post hoc correction, Sidak’s. *p < 0.05; ****p < 0.0001.

FIGURE 4

HCRT-deficient mice confer resistance to CFS-induced loss of cholinergic projections in the CA1 region of the hippocampus. (A) Representative confocal images of vesicular acetylcholine transporter (VAchT, red) immunoreactivity in CA1 hippocampus across the genotype and sleep conditions. Subregions of CA1 are labeled: Oriens, Or; pyramidal neuron somata, Pyr; and Strata Radiatum, Str Rad. Calibration bar, 50 μm. (B) Group data for percentage area CA1 VAchT, expressed as mean ± SE (n = 5–7/group) for Rest (dark blue) and CFS (light blue) for WT and HCRT–/– mice. Data were analyzed with two-way ANOVA and Sidak’s post hoc tests, *p < 0.05; ***p < 0.001; ****p < 0.0001.

FIGURE 5

HCRT genotype-independent effects of CFS on locus coeruleus neuron counts. (A) Representative mid-locus coeruleus (LC) coronal sections immunolabeled for tyrosine hydroxylase detected with substrate blue (navy) and counterstained with Giemsa staining across the four groups. Surrounding landmarks: Superior Cerebellar Peduncle, SCP; Medial Parabrachialis nucleus (MPB; Mesencephalic Trigeminal, Mes V; and Barrington’s nucleus, Bar. Calibration bar, 50 μm. (B) Group data for stereological LC cell counts, expressed as mean ± SE (n = 5/group) for Rest (dark blue) and CFS (light blue) for WT and HCRT–/– mice, analyzed with two-way ANOVA and Sidak’s post hoc tests, *p < 0.05.