Early brain-wide disruption of sleep microarchitecture in amyotrophic lateral sclerosis
Christina Lang, Simon J. Guillot, Dorothee Lule, Luisa T. Balz, Antje Knehr, Patrick Weydt, Johannes Dorst, Katharina Kandler, Hans-Peter Muller, Jan Kassubek, Laura Wassermann, Sandrine Da Cruz, Francesco Roselli, Albert C. Ludolph, Matei Bolborea, Luc Dupuis
Christina Lang, Simon J. Guillot, Dorothee Lule, Luisa T. Balz, Antje Knehr, Patrick Weydt, Johannes Dorst, Katharina Kandler, Hans-Peter Muller, Jan Kassubek, Laura Wassermann, Sandrine Da Cruz, Francesco Roselli, Albert C. Ludolph, Matei Bolborea, Luc Dupuis
View: Text | PDF
Clinical Research and Public Health Genetics Neuroscience

Early brain-wide disruption of sleep microarchitecture in amyotrophic lateral sclerosis

Abstract

BACKGROUND Amyotrophic lateral sclerosis (ALS), the major adult-onset motor neuron disease, is preceded by an early period unrelated to motor symptoms, including altered sleep, with increased wakefulness and decreased deep nonrapid eye movement (NREM). Whether these alterations in sleep macroarchitecture are associated with — or even precede — abnormalities in sleep-related EEG features remains unknown.METHODS Here, we characterize sleep microarchitecture using polysomnography for patients with ALS (n = 33) and controls (n = 32) as well as for asymptomatic carriers of superoxide dismutase 1 (SOD1) or C9ORF72 mutations (n = 57) and noncarrier controls (n = 30). Patients and controls with factors that could confound sleep structure, including respiratory insufficiency, were prospectively excluded. The results were complemented in 3 ALS mouse models (Sod1G86R, FusΔNLS/+, and TDP-43Q331K).RESULTS We observed a brain-wide reduction in the density of sleep spindles, slow oscillations, and K-complexes in patients with early-stage ALS and in presymptomatic gene carriers. These defects in sleep spindles and slow oscillations correlated with cognitive performance in both cohorts, particularly with scores on memory, verbal fluency, and language function. Alterations in sleep microarchitecture were replicated in 3 mouse models, and decreases in sleep spindles were rescued following intracerebroventricular supplementation of melanin-concentrating hormone (MCH) or by oral administration of a dual orexin receptor antagonist.CONCLUSION Sleep microarchitecture was associated with cognitive deficits and causally linked to aberrant MCH and orexin signaling in ALS.FUNDING Agence Nationale de la Recherche (ANR); Fondation Thierry Latran; Association Francaise de Recherche sur la sclérose latérale amyotrophique; Association Française contre les myopathies; TargetALS; and Joint Program on Neurodegenerative Diseases Research (JPND).

Authors

Christina Lang, Simon J. Guillot, Dorothee Lule, Luisa T. Balz, Antje Knehr, Patrick Weydt, Johannes Dorst, Katharina Kandler, Hans-Peter Muller, Jan Kassubek, Laura Wassermann, Sandrine Da Cruz, Francesco Roselli, Albert C. Ludolph, Matei Bolborea, Luc Dupuis

×

Figure 5

Slow oscillation alterations in presymptomatic ALS gene carriers.

Options: View larger image (or click on image) Download as PowerPoint
Slow oscillation alterations in presymptomatic ALS gene carriers. (A) Re...
(A) Representative slow oscillations across all electrodes for a healthy individual, 1 presymptomatic SOD1, and 1 C9ORF72 gene carrier. (B) Quantification of slow oscillation density, slope, PAC between sleep spindles and slow oscillations, and PSP in controls and SOD1 and C9ORF72 presymptomatic gene carriers. *Padj. < 0.05, **Padj. < 0.01, and ****Padj. < 0.0001, 1-way ANOVA with FDR-BKY correction. (C) Topographic maps across all electrodes of slow oscillation density, slope, PAC, and PSP in controls and SOD1 and C9ORF72 presymptomatic gene carriers. (D) Quantification of slow oscillation density, slope, PAC, and PSP across F3/F4 and C3/C4 electrodes in controls, SOD1 and C9ORF72 presymptomatic gene carriers as indicated. *Padj. < 0.05, ***Padj. < 0.001, and ****Padj. < 0.0001, by Kruskal-Wallis test with Dunn’s multiple tests adjusted with FDR-BKY correction. Results with P > 0.05 are not indicated. Data are presented as the median and IQRs. Corrected P values are shown.