Electroencephalographic changes of REM sleep in chronically sleep restricted mice

Abstract

Chronic sleep restriction (CSR) has been shown to impair people’s health and cognitive functions, as well as change sleep homeostasis. Previous studies have shown that CSR may impair non-rapid eye movement (NREM) sleep generation in addition to reducing NREM delta power. However, little is known about how CSR alters rapid eye movement (REM) sleep or EEG power spectra during REM. Here, we used a high-density EEG method in freely behaving mice to assess the REM sleep response to CSR. Mice (N=9) were sleep deprived daily for 18-h using periodically rotating wheels, followed by 6-h of sleep opportunity that started at the beginning of each light period. This sleep restriction (SR) protocol was repeated for 5 consecutive days. High-density EEG was analyzed for only 2-h of each 6-h sleep opportunity (1-3h after the initiation of sleep opportunity), whereas conventional EEG data using skull screw electrodes was analyzed for the full 24-h experimental days. The REM sleep time during the daily 6-h sleep opportunities increased as the SR days progressed, compared to the corresponding baseline levels. A power spectral analysis of the high-density EEG revealed that low theta power (5-7 Hz) increased significantly in the frontal cortex on SR day 1, then continuously decreased on SR3 and SR5; high theta power (7-10 Hz) was persistently elevated throughout all SR days, especially in the centro-parietal cortex. A close examination of theta oscillation revealed a transition from unimodal to bimodal oscillation showing that a peak frequency at 7 Hz during baseline was split into two peak frequencies at 7 and 9 Hz. Regarding REM gamma power, a gradual but significant increase in low gamma power (30-50 Hz) was observed near the prefrontal cortex especially on SR3 and SR5, while robust increases in high gamma power (70-100 Hz) were observed most significantly in the centro-parietal cortex on SR3. Additionally, the analysis of the cross-frequency coupling between theta phase and gamma power showed that modulation of theta on gamma oscillation was not altered during CSR. Thus, CSR produces opposite effects on the low and high theta power of REM sleep in mice. In addition, this study indicates that CSR significantly increases REM gamma power while 18-h of acute sleep deprivation does not. Further studies are needed to determine if REM high theta power and gamma power actually correlates with the behavioral sleepiness and cognitive impairments observed in CSR.