Scheduled feeding restores memory and modulates c-Fos expression in the suprachiasmatic nucleus and septohippocampal complex

Abstract

Disruptions in circadian timing impair spatial memory in humans and rodents. Circadian-arrhythmic Siberian hamsters (Phodopus sungorus) exhibit substantial deficits in spatial working memory as assessed by a spontaneous alternation (SA) task. The present study found that daily scheduled feeding rescued spatial memory deficits in these arrhythmic animals. Improvements in memory persisted for at least 3 weeks after the arrhythmic hamsters were switched back to ad libitum feeding. During ad libitum feeding, locomotor activity resumed its arrhythmic state, but performance on the SA task varied across the day with a peak in daily performance that corresponded to the previous daily window of food anticipation. At the end of scheduled feeding, c-Fos brain mapping revealed differential gene expression in entrained versus arrhythmic hamsters in the suprachiasmatic nucleus (SCN) that paralleled changes in the medial septum and hippocampus, but not in other neural structures. These data show that scheduled feeding can improve cognitive performance when SCN timing has been compromised, possibly by coordinating activity in the SCN and septohippocampal pathway.

Figure 1

Schematic of behavior (left) and c-Fos (right) experiments. Time of day is indicated by the black and white rectangles. Room lights were on from ZT0-16. Memory tests were performed at four zeitgeber times (ZT; 17, 23, 5, 11). ZT17 and 23 indicate early and late night, respectively. The timeline for the project runs from top to bottom (red). The number of days for each section of the study is indicated by arrows (blue). The periods of scheduled feeding (SF) and ad libitum feeding are shaded in gray. SF began at ZT5 and was 8 h long on the first day, and then decreased by 30 min on successive days until it was 4 h in duration. SF lasted 21 days. The days on which memory was tested, as well as the times of day testing occurred, are indicated by filled circles (blue). Those blue circles show that memory was tested in the same group of animals at multiple time points with 2 days in between each test. Memory was tested twice at ZT5 in DPS-arrhythmic (ARR) animals as indicated by the blue circle outlined in black. The day and time of tissue collection for c-Fos studies is indicated by a filled circle (green).

Figure 2

Scheduled feeding rescued spatial working memory in circadian arrhythmic hamsters and phase-delayed the rhythm of test performance in entrained animals. (A,B) Time of food availability (red outline) increased food anticipatory activity (FAA) prior to the time of feeding in both entrained (n = 9) and arrhythmic animals (n = 12). Activity returned to baseline levels when animals resumed ad libitum feeding. Brackets indicate portions of the actograms subjected to periodogram analysis. Peaks in blue above the black lines indicate statistically significant rhythms as determined by chi-square periodogram analysis. (C,D) The mean (±SE) daily FAA ratio was calculated during SF (red, green circles) and ad libitum feeding (gray circles). Gray bars represent the alternation scores (right y-axis) during baseline and on day 21 of SF. *Indicates significantly different from chance (i.e., 50%); P < 0.01. (E,F) Waveforms for total (mean ± SE) hourly locomotor activity during each phase of the study (BL = baseline, SF = scheduled feeding, AL = ad libitum) were constructed with the last 5 days of each condition (i.e., the last 5 days in panels C,D during baseline, SF, and AL). Data for each hour were averaged and plotted at the end of that hour (e.g., ZT17 is the mean of data from ZT16-17). Black and white rectangles indicate the times of night and day, respectively. Time of SF indicated by vertical black dotted lines. (G,H) Alternation scores during baseline and SF conditions at four different zeitgeber times (ZT; *P < 0.01). ZT17 and ZT23 represent early and late night, respectively. Night and day are indicated by black and white rectangles in panels E,F. Sample sizes (n) for panels G,H: entrained (SF = 9; AL = 10), arrhythmic (SF = 12; AL = 9). All animals remained in a 16:8 light-dark cycle throughout the study.

Figure 3

Induction of c-Fos expression in the SCN occurred primarily in the dorsomedial region along the mid to caudal SCN axis. (A) Representative tissue sections showing typical patterns of c-Fos expression in entrained (ENT) and in DPS-arrhythmic (ARR) animals that underwent either the scheduled feeding (SF) regimen or were fed ad libitum (AL). (B) Representative tissue sections from the mid (upper section) and caudal (lower section) SCN showing patterns of immunoexpression of vasopressin-neurophysin (green) and vasoactive intestinal polypeptide (red). These sections were used as templates for demarcating dorsomedial and ventrolateral SCN boundaries for the cell counts. (C) Number of cells counted at sequential levels of the SCN along its rostrocaudal axis. Sections were taken at intervals of 120 μm (* indicates significant differences between AL and SF conditions; P < 0.05, n = 6 per group). (D) Data from C given as density of c-Fos IR cells. (E) Number of c-Fos IR cells in the dorsomedial and ventrolateral subregions of the SCN, as well as total cell counts (* indicates P < 0.05 compared to AL condition).

Figure 4

SF equalized c-Fos expression among ENT and ARR animals in specific non-SCN brain regions. Bar graphs show mean (±SE) number of c-Fos IR cells in each group for the (A) medial septum, (B) granule cell layer of the dentate gyrus, and (C) central nucleus of the amygdala (*P < 0.05 compared to AL condition). The middle column of panels shows representative c-Fos labeling in brain sections from all four groups of animals. Red areas in hamster coronal sections on the right column indicate the areas in which cells were counted.

Figure 5

SF had no effect on c-Fos expression in these brain regions. (A) lateral septum, (B) polymorphic layer of the dentate gyrus, (C) basolateral nucleus of the amygdala, (D) ventral region of the dorsomedial hypothalamic nucleus (DMH), or (E) dorsal region of the DMH, and (F) compact nucleus of the DMH.