These results indicate that enhanced GC apoptosis during feeding and postprandial period occurred in association with postprandial behaviors. Given that sleeping behavior was the most conspicuous postprandial behavior in the present analysis (Figure 3A) and that sleep plays a crucial role in brain plasticity (Buzsáki, BMS777607 1989 and Diekelmann

and Born, 2010), we next examined the contribution of sleeping behavior during the postprandial period to GC apoptosis. Postprandial sleep was evaluated using EEG and EMG recording in freely behaving mice. After implantation of recording electrodes, mice were subjected to restricted feeding and analyzed on day 10 (Figure 4A). During the initial 1 hr of feeding, mice engaged in continuous eating without resting or sleeping. The EEG of the occipital cortex and Torin 1 EMG of the neck muscles

were recorded during the following hour, and the mice were then perfusion fixed. Using video recordings of behavior, EEG, and EMG, the behavioral state was classified every 10 s into the waking, light sleep, slow-wave sleep, and REM sleep states (Radulovacki et al., 1984 and Tsuno et al., 2008; Figure 4B). Waking-sleeping behaviors during postprandial period were fragmented into episodes of short duration. Thus the total time of each state during the 1 hr postprandial period was calculated and evaluated. All mice examined showed various sleep states with various lengths (Table S1). The length of total sleep (sum of light, slow-wave, and REM sleep) positively correlated with apoptotic

GC number (Figure 4C). By state, the length of slow-wave PAK6 sleep correlated well with apoptotic GC number (Figure 4D). On the other hand, REM sleep was not necessarily observed during the postprandial period, and many mice without REM sleep showed an increase in GC apoptosis (Figure 4E). These results confirmed that most mice slept during the postprandial period and suggested that slow-wave sleep or total sleep promoted GC apoptosis. They also suggested that a brief period of sleep of 20–40 min exerted a potent effect in enhancing GC apoptosis (Figure 4C). We also confirmed in EEG- and EMG-recorded mice that the gentle handling efficiently inhibited sleep states during the postprandial period (data not shown), supporting the potent role of sleep in enhancing GC apoptosis. The occurrence of sleep during the postprandial period is in accord with the notion that satiety induces sleep (Mieda and Yanagisawa, 2002). One question is whether sleep per se has a potent role in enhancing GC apoptosis, or whether this is due to a combination of feeding and sleep. Continuous behavioral analysis of food-restricted mice showed that they also slept outside the postprandial period (Figure S3), whereas enhanced GC apoptosis was apparent only during the postprandial period (Figure 1).