Sleepsign software

After 8–10 days of post-operative recovery, the mice were placed in experimental cages for a 4-day habituation/acclimatization period and connected with counterbalanced recording leads. Baseline EEG/EMG was recorded for 24 h starting onset of dark phase. All mice that were subjected to EEG/EMG recordings received vehicle and multiple doses of Ashwagandha leaf extracts or pure TEG (10 mg, 20 mg, and 30 mg/mouse) on different days and any two administrations were separated by, at least, 2 days. Each EEG/EMG recording and drug administration was performed per os (p.o.), at the onset of dark phase (17:00 h). Cortical EEG and EMG signals were amplified and filtered (EEG, 0.5–30 Hz; EMG, 20–200 Hz), then digitized at a sampling rate of 128 Hz, and recorded by using SleepSign software (Kissei Comtec, Nagano, Japan) as previously described [19 ]. Polysomnographic recordings were scored with automated analysis, off-line, in 10-s epochs as wakefulness, rapid eye movement (REM) and non-REM (NREM) sleep by SleepSign software, using standard criteria [19 , 20 (link)]. The defined sleep–wake stages were visually examined and corrected, wherever necessary. Spectral analysis of EEG by fast Fourier transformation (FFT) was performed, and the EEG power densities of each 0.5- Hz bin were averaged by calculating the percentage of each bin with respect to the total power in the range of 0.5–35 Hz.

All of the EEG/EMG signals during optogenetic experiments were recorded and analyzed by the SleepSign software (Kissei Comtec). The amplified EEG/EMG data were collected at 128 Hz sampling rate. In the experiment exploring the effect of optogenetics on cortical EEG under 0.8% isoflurane anesthesia, the EEG frequency band (δ, 0.5–4.0 Hz; θ, 4.0–7.0 Hz; α, 8.0–15.0 Hz; β, 16.0–30.0 Hz) used fast Fourier transform to calculate the relative change in the total power of BF-GABAergic neurons within 120 s before, during, and after photostimulation. The burst suppression ratio (BSR) is a widely adopted method for quantifying burst suppression. For the optogenetic experiments of burst suppression oscillation, raw EEG data recorded by SleepSign software were converted to text format for further analysis of BSR using MATLAB R2019b (Bao et al., 2021 (link)). Finally, the BSR changes in BF-GABAergic neurons within 120 s before, during, and after photostimulation were calculated. The details of BSR analysis are provided in the Extended Data 1.

For EEG and EMG recordings, mice were implanted with a head mount (Pinnacle Technology Inc., Laurence, KS, USA) parallel to the sagittal suture. Four stainless steel screws were placed 1.5 mm lateral to the sagittal suture, 2.0 mm anterior of bregma, and 4.0 mm posterior to bregma to record EEG signals. EMG signals were acquired by a pair of multi-stranded stainless steel wires inserted into the neck extensor muscles. One week after surgery, animals were habituated for 3 days before EEG/EMG signals were acquired with a 3-channel EEG/EMG tethered system (Pinnacle Technology Inc.) and digitalised by SIRENIA SLEEP PRO^® software (Pinnacle Technology). EEG and EMG data were recorded in 10 s epochs and automatically scored by Sleep Sign^® software (Kissei Comtec, Matsumoto, Japan). Automatically scored data were manually inspected and corrected if required.

The EEG/EMG signals were amplified and filtered (EEG: 0.5–64 Hz, EMG: 16–64 Hz), then digitised at a sampling rate of 128 Hz, and recorded using SLEEPSIGN software^{47 (link)} (Kissei Comtec). In addition, locomotor activity was recorded with an infrared photocell sensor (Biotex). The vigilance states were scored offline by characterising 10-s epochs into three stages, including awake, SWS and REM sleep, according to standard criteria⁴⁶ . As a final step, defined vigilance stages were examined visually, and corrected when necessary. Spindles during SWS were analysed as previously reported^{48 (link)}. In brief, EEG was bandpass-filtered (10–13 Hz) to visually identify sleep spindles from the raw EEG signals.

The EEG/EMG signals were amplified and filtered (EEG: 0.5–64 Hz, EMG: 16–64 Hz), then digitised at a sampling rate of 128 Hz, and recorded using SLEEPSIGN software (Kohtoh et al., 2008 (link)) (Kissei Comtec). In addition, locomotor activity was recorded with an infrared photocell sensor (Biotex). The vigilance states were scored offline by 10 s epochs into three stages, including waking, SWS, and REM sleep, according to standard criteria (Oishi et al., 2016 (link)). As a final step, defined vigilance stages were examined visually, and corrected when necessary.

Sleep stage analysis by human experts was conducted based on visual characteristics of EEG and EMG waveforms with the help of FFT and video surveillance of mouse movement. The EEG dataset consisted of 214 recordings in total. In most cases, two recordings on two consecutive days were obtained from each mouse. Recordings were divided into 10-s epochs. FFT analysis was performed using Sleep Sign software (KISSEI COMTEC).
Wakefulness was defined by continuous mouse movement or de-synchronized low-amplitude EEG with tonic EMG activity. NREM was defined by dominant high-amplitude, low-frequency delta waves (1–4 Hz) accompanied by less EMG activity than that observed during wakefulness. REM was defined as dominant theta rhythm (6–9 Hz) and the absence of tonic muscle activity. If a 10-s epoch contained more than one sleep stage (i.e., NREM, REM, or wakefulness), the most represented stage was assigned for the epoch. At least two experts agreed on the classification of each recording.