To address this possibility,

mice habituated to restricted feeding were left without food at the presumptive feeding time (Figure 7A; no food). In contrast to mice that ate food, those without food continued to show exploratory behavior, without resting, sleeping, or extended periods of grooming, during the initial 2 hr of the presumptive feeding time (data not shown). In this period, there was no increase in apoptotic GC number (Figure 7B; 2 hr—no food). In addition, mice with restricted feeding that were allowed to smell food odor but were prevented from eating (Figure 7A; food odor) also showed continual exploratory and sniffing behaviors during the presumptive feeding time, and also exhibited no enhancement of GC apoptosis (Figure 7B; 2 hr—food odor). The observation period of the food-deprived mice S3I201 was then prolonged beyond the presumptive feeding time (Figure 7C). After many hours, the mice showed various behaviors including grooming, resting, and sleeping. When examined after showing sleeping behavior (Figure 7C, arrows),

some showed a several-fold increase in GC apoptosis (Figure 7D). Pazopanib mw This observation indicates that actual food intake is not an absolute requirement for enhanced GC apoptosis in food-restricted mice and also suggests that the postprandial period is a typical but not the only period in which GC apoptosis can be enhanced (see Discussion). The enhanced GC apoptosis observed so far might largely depend on the specific paradigm of food restriction. Alterations in body status such as hormonal levels and energy metabolism in long-term food-restricted mice (Gao and Horvath, 2007) may be important to the enhancement of GC apoptosis during the postprandial period. To examine whether GC apoptosis during the postprandial period is enhanced in mice without long-term food restriction, we designed a one-time food restriction paradigm. In this paradigm, food was abruptly removed only for 4 hr and 20 min in ad libitum feeding mice

and then made available again to efficiently induce feeding and postprandial behaviors (Figure 7E, middle bar). Food was removed during the early dark phase of the circadian cycle, because this was the period in which ad libitum feeding mice ate most extensively (data not shown; Zucker, 1971). Following food redelivery, the mice successfully showed feeding and subsequent postprandial behaviors, including grooming, resting, and sleeping. Under this paradigm, GC apoptosis was enhanced in mice with feeding and postprandial behaviors compared to mice before food supply (Figure 7F). Because under this condition the time of eating and postprandial behaviors after food redelivery varied widely among mice, the redelivery period was limited to 1 hr only (Figure 7E, bottom bar), which efficiently induced postprandial behaviors and enhanced GC apoptosis within 2.