05, one-way ANOVA, Kruskal-Wallis with Dunn’s multiple comparison test). However, the response during all waking periods remained significantly different from wake 1 (p < 0.01, one-way ANOVA, Dunnett’s multiple comparison test). The overall mean amplitude for a first wake episode (n = 13, from 4 different GSK-3 inhibitor cats) was 0.297mV ± 0.176mV and it was 0.328mV ± 0.177mV during wake 2; this difference

was highly significant (Figure 1G, p < 0.001, Wilcoxon matched-pairs signed-rank test). Even relatively short periods of SWS were sufficient to enhance responses in the second wake episode (Figure 1H). We obtained six intracellular recordings from somatosensory cortical neurons in which we recorded evoked responses to

medial lemniscus stimuli during two consecutives wake episodes separated by a period of SWS, although it was not always the first and second wake episodes (Figure 3). As expected, membrane potential recordings showed the presence of prolonged silent states during SWS, which were absent during wake (Figure 3A). Responses were highly variable during SWS (data not shown), while they were stable during wake (Figures 3B and 3D). In five out of the six intracellular recordings in which we were able to obtain stable recording throughout wake-SWS-wake transitions (85 recording sessions, near 200 wake-SWS-wake transitions, 2 animals), the response amplitude was increased in the second wake episode (post-SWS) as compared to the first episode (pre-SWS); however, due to the small number of recordings, the difference was not significant (Figure 3C, p = 0.2, Wilcoxon matched-pairs see more signed-rank test). Therefore, intracellular recordings support field potential observations. To characterize the mechanisms implicated in the potentiation of evoked responses, we performed whole-cell recordings from layer II/III pyramidal regular-spiking neurons in vitro. From in vivo LFP recordings, we extracted the timing of a single unit firing recorded Bay 11-7085 during SWS (Figure 4A) and during wake (Figure 5A) and used that timing to build sleep-like ( Figure 4B) and wake-like ( Figure 5B) patterns

of synaptic stimulation; the timing of slow waves was also detected to build the intracellular hyperpolarizing current pulse stimulation pattern replicating hyperpolarizing (silent) states of SWS ( Figure 4B, see Experimental Procedures). The stimulation protocols are detailed in Figure S2. The mean membrane potential of neurons recorded in vitro was maintained to about −65mV to mimic the membrane potential of cortical neurons during wake or active phases of SWS. Minimal intensity stimuli were applied in the vicinity of recorded neurons. The sleep-like pattern of synaptic stimulation induced a transient facilitation only ( Figure 4C; 0.640mV ± 0.245mV in control versus 0.817mV ± 144mV in the first minute after conditioning, p < 0.05, Mann-Whitney test).