Extended Data Fig. 3: Effects of closed-loop activation and inhibition of mPFC Pyr neurons on EEG and REM sleep duration.
(a) PSD of the EEG during REM sleep with (ON) and without laser stimulation (OFF) of mPFC Pyr neurons expressing ChR2. Two-way RM ANOVA with laser (P = 0.0057) and frequency band (δ, θ, or σ) as within-subjects factors (interaction, P = 0.0001); t-tests with Holm-Bonferroni correction; OFF versus ON: δ, P = 0.0094; θ, P = 0.0011; σ, P = 0.0005. n = 11 mice. Shadings, 95% CIs. (b) Laser-induced changes in the δ, θ, and σ power (Δ power) between REM sleep with and without laser stimulation in eYFP and ChR2 mice. Mixed ANOVA with frequency band as within- and virus (eYFP or ChR2) as between-subjects factor (interaction, P = 0.0000); t-tests with Holm-Bonferroni correction; eYFP versus ChR2: δ, P = 0.0041; θ, P = 0.0012; σ, P = 0.0330. ChR2, n = 11; eYFP, n = 8 mice. Bars, averages across mice; dots, individual mice; error bars, 95% CIs. (c) Duration of REM sleep episodes with (ON) or without laser (OFF) and REM sleep duration in baseline recordings without laser stimulation (Bsl) in ChR2 mice. One-way RM ANOVA (laser, P = 0.0001); t-tests with Holm-Bonferroni correction; Bsl versus OFF: P = 0.0253; Bsl versus ON: P = 0.0054; OFF versus ON: P = 0.0054. n = 6 mice. Box plots; lines, individual mice. (d) Left, heatmaps depicting expression of iC++-eYFP in the mPFC in four consecutive brain sections (n = 8 mice). Right, location of fiber tracts. Each pair of colored bars represents the location of two optic fibers for bilateral inhibition. Brain atlas images adapted with permission from ref. 61. (e) PSD of the EEG during REM sleep with and without laser stimulation of mPFC Pyr neurons expressing iC++. Two-way RM ANOVA (laser, P = 0.0001; interaction, P = 0.0000); t-tests with Holm-Bonferroni correction; OFF versus ON: δ, P = 0.1792; θ, P = 0.0001; σ, P = 0.0281. n = 8 mice. Shadings, 95% CIs. (f) Laser-induced changes in the δ, θ, and σ power (Δ power) between REM sleep with and without laser stimulation in eYFP and iC++ mice. Mixed ANOVA (interaction, P = 0.0000); t-tests with Holm-Bonferroni correction; eYFP versus iC++: δ, P = 0.8948; θ, P = 0.0000; σ, P = 0.0029. iC++, n = 8; eYFP, n = 8 mice. Bars, averages across mice; dots, individual mice; error bars, 95% CIs. (g) Duration of laser-on or laser-off REM sleep episodes and REM sleep duration in baseline recordings without laser stimulation in iC++ mice. One-way RM ANOVA (laser, P = 0.0000); t-tests with Holm-Bonferroni correction; Bsl versus OFF: P = 0.0120; Bsl versus ON: P = 0.0120; OFF versus ON: P = 0.0003. n = 8 mice. Box plots; lines, individual mice. (h) PSD of the EEG during phasic θ events and remaining REM sleep (tonic θ) with (ON) and without laser (OFF) in iC++ mice. Peak frequency, two-way RM ANOVA with type of REM (tonic or phasic, P = 0.0000) and laser (ON or OFF) as within-subjects factors (interaction, 0.2849); t-tests with Holm-Bonferroni correction, tonic versus phasic: OFF, P = 0.0000; ON, P = 0.0000. EEG θ power, two-way RM ANOVA (type, P = 0.0001; interaction, P = 0.0953); t-tests with Holm-Bonferroni correction; tonic versus phasic: OFF, P = 0.0003; ON, P = 0.0003. n = 8 mice. Lines, averages across mice; shadings, 95% CIs. (i) Heart rate during phasic θ events and remaining REM sleep (tonic θ) with (ON) and without laser (OFF) in iC++ mice. Two-way RM ANOVA (type, P = 0.0001; interaction, P = 0.4153); t-tests with Holm-Bonferroni correction, tonic versus phasic: OFF, P = 0.0182; ON, P = 0.0123; n = 5 mice. Box plots; lines, individual mice; bpm, beats per minute. (j) Frequency of phasic θ events as function of time in REM sleep (since REM sleep onset) in baseline recordings from ChR2 and iC++ mice without laser stimulation. For each mouse, we averaged the frequency of phasic θ events across all REM sleep episodes for each 25 s time bin on the x-axis. Each dot corresponds to the mean frequency of one animal. Robust regression; P = 0.0000; n = 14 mice. The line and shading indicate the robust regression fit and 95% CI. (k) Mean duration of all REM sleep episodes (averaged across laser-on and laser-off episodes) in closed-loop experiments for ChR2, iC++ mice, and eYFP controls for activation (eYFP1) and inhibition experiments (eYFP2). One-way ANOVA (P = 0.8359); eYFP1, n = 8; ChR2, n = 11; eYFP2, n = 8; iC++, n = 8 mice. Bars, averages across mice; dots, individual mice; error bars, 95% CIs. See Supplementary Table 1 for detailed statistical information. *P < 0.05, **P < 0.01, ***P < 0.001.
