Fig. 3
a Dynamics of EEG spectra in the frontal derivation are shown for the transition from NREM sleep to REM sleep for the 24-h baseline. Only REM episodes > 1 min were included in this analysis (33.6 ± 1.2 and 28.8 ± 2.4 episodes per animal, WT and Gria1−/− respectively). 2 min prior to the transition ( = 0) and 1 min after the transition is illustrated based on 4-s epochs (x-axis: in seconds). Mean values (top panel: WT, n = 8, bottom panel: Gria1−/−, n = 6). EEG power density is colour-coded according to the log scale (µV2/0.25 Hz) on the right. Note a surge of EEG power in spindle-frequency range immediately prior to transition in WT mice, which is attenuated in Gria1−/− mice (circled). b The difference between WT and Gria1−/− is represented as %. For this analysis the average spectral values calculated for each 4-s epoch in WT mice are expressed as percentage from the corresponding spectral values in Gria1−/− mice. c Mean frontal EEG power density during the last 32 s of NREM sleep prior to transition to REM sleep. Mean values, SEM (WT: n = 8, Gria1−/−: n = 6) for each frequency bin expressed as % of mean EEG power over all artefact-free 4-s epochs in NREM sleep during 24 h in the corresponding bin (genotype*bin interaction: F(80,960) = 4.710, p < 0.001, two-way repeated-measures ANOVA). Triangles below the curves depict frequency bins, where EEG power during the last 32 s before NREM-REM transition differed significantly between genotypes (p < 0.05, post hoc Bonferroni test)
