Background: Disturbed sleep is associated with cognitive decline in neurodegenerative diseases such as Alzheimer's disease (AD) and frontotemporal dementia (FTD). The progressive sequence of how neurodegeneration affects aspects of sleep architecture in conjunction with behavioural changes is not well understood. Methods: We investigated changes in sleep architecture, spectral power and circadian rhythmicity in the tet-off rTg4510 mouse overexpressing human P301L tau within the same subjects over time. Doxycycline-induced transgene-suppressed rTg4510 mice, tTa carriers and wild-type mice were used as comparators. Spectral power and sleep stages were measured from within the home cage environment using EEG electrodes. In addition, locomotor activity and performance during a T-maze task were measured. Results: Spectral power in the delta and theta bands showed a time-dependent decrease in rTg4510 mice compared to all other groups. After the initial changes in spectral power, wake during the dark period increased whereas NREM and number of REM sleep bouts decreased in rTg4510 compared to wild-type mice. Home cage locomotor activity in the dark phase significantly increased in rTg4510 compared to wild-type mice by 40 weeks of age. Peak-to-peak circadian rhythm amplitude and performance in the T-maze was impaired throughout the experiment independent of time. At 46 weeks, rTG4510 mice had significant degeneration in the hippocampus and cortex whereas doxycycline-treated rTG4510 mice were protected. Pathology significantly correlated with sleep and EEG outcomes, in addition to locomotor and cognitive measures. Conclusions: We show that reduced EEG spectral power precedes reductions in sleep and home cage locomotor activity in a mouse model of tauopathy. The data shows increasing mutant tau changes sleep architecture, EEG properties, behaviour and cognition, which suggest tau-related effects on sleep architecture in patients with neurodegenerative diseases.

Increasing vigilance without incurring the negative consequences of extended wakefulness such as daytime sleepiness and cognitive impairment is a major challenge in treating many sleep disorders. The present work compares two closely related mGluR2/3 antagonists LY3020371 and LY341495 with two well-known wake-promoting compounds caffeine and d-amphetamine. Sleep homeostasis properties were explored in male Wistar rats by manipulating levels of wakefulness via (1) physiological sleep restriction (SR), (2) pharmacological action, or (3) a combination of these. A two-phase nonlinear mixed-effects model combining a quadratic and exponential function at an empirically estimated join point allowed the quantification of wake-promoting properties and any subsequent sleep rebound. A simple response latency task (SRLT) following SR assessed functional capacity of sleep-restricted animals treated with our test compounds. Caffeine and d-amphetamine increased wakefulness with a subsequent full recovery of non-rapid eye movement (NREM) and rapid eye movement (REM) sleep and were unable to fully reverse SR-induced impairments in SRLT. In contrast, LY3020371 increased wakefulness with no subsequent elevation of NREM sleep, delta power, delta energy, or sleep bout length and count, yet REM sleep recovered above baseline levels. Prior sleep pressure obtained using an SR protocol had no impact on the wake-promoting effect of LY3020371 and NREM sleep rebound remained blocked. Furthermore, LY341495 increased functional capacity across SRLT measures following SR. These results establish the critical role of glutamate in sleep homeostasis and support the existence of independent mechanisms for NREM and REM sleep homeostasis.

BACKGROUND: Disturbed sleep is associated with cognitive decline in neurodegenerative diseases such as Alzheimer’s disease (AD) and frontotemporal dementia (FTD). The progressive sequence of how neurodegeneration affects aspects of sleep architecture in conjunction with behavioural changes is not well understood. METHODS: We investigated changes in sleep architecture, spectral power and circadian rhythmicity in the tet-off rTg4510 mouse overexpressing human P301L tau within the same subjects over time. Doxycycline-induced transgene-suppressed rTg4510 mice, tTa carriers and wild-type mice were used as comparators. Spectral power and sleep stages were measured from within the home cage environment using EEG electrodes. In addition, locomotor activity and performance during a T-maze task were measured. RESULTS: Spectral power in the delta and theta band showed a time-dependent decrease in rTg4510 mice compared to all other groups. After the initial changes in spectral power, wake during the dark period increased whereas NREM and number of REM sleep bouts decreased in rTg4510 compared to wild-type mice. Home cage locomotor activity in the dark phase significantly increased in rTg4510 compared to wild-type mice by 40 weeks of age. Peak-to-peak circadian rhythm amplitude and performance in the T-maze was impaired throughout the experiment independent of time. At 46 weeks rTG4510 mice had significant degeneration in the hippocampus and cortex whereas doxycycline treated rTG4510 mice were protected. Pathology significantly correlated with sleep and EEG outcomes, in addition to locomotor and cognitive measures. CONCLUSIONS: We show that reduced EEG spectral power precedes reductions in sleep and home-cage locomotor activity in a mouse model of tauopathy. The data shows increasing mutant tau changes sleep architecture, EEG properties, behaviour and cognition; which suggest tau related effects on sleep architecture in patients with neurodegenerative diseases.