Mice fed a single daily meal at intervals within the circadian range exhibit food anticipatory activity. Previous investigations strongly suggest that this behaviour is regulated by a circadian pacemaker entrained to the timing of fasting/refeeding. The neural correlate(s) of this pacemaker, the food entrainable oscillator (FEO), whether found in a neural network or a single locus, remain unknown. This study used a canonical property of circadian pacemakers, the ability to continue oscillating after removal of the entraining stimulus, to isolate activation within the neural correlates of food entrainable oscillator from all other mechanisms driving food anticipatory activity. It was hypothesized that continued anticipatory activation of central nuclei, after restricted feeding and a return to ad libitum feeding, would elucidate a neural representation of the signaling circuits responsible for the timekeeping component of the food entrainable oscillator. Animals were entrained to a temporally constrained meal then placed back on ad libitum feeding for several days until food anticipatory activity was abolished. Activation of nuclei throughout the brain was quantified using stereological analysis of c-FOS expressing cells and compared against both ad libitum fed and food entrained controls. Several hypothalamic and brainstem nuclei remained activated at the previous time of food anticipation, implicating them in the timekeeping mechanism necessary to track previous meal presentation. This study also provides a proof of concept for an experimental paradigm useful to further investigate the anatomical and molecular substrates of the FEO.
Citation: Blum ID, Waddington Lamont E, Rodrigues T, Abizaid A (2012) Isolating Neural Correlates of the Pacemaker for Food Anticipation. PLoS ONE 7(4): e36117. https://doi.org/10.1371/journal.pone.0036117
Editor: Shin Yamazaki, Vanderbilt University, United States of America
Received: August 12, 2011; Accepted: March 30, 2012; Published: April 27, 2012
Copyright: © 2012 Blum et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This project was funded by grants from the Natural Sciences and Engineering Research Council of Canada (NSERC, www.nserc-crsng.gc.ca/) and the Canada Foundation for Innovation (CFI, www.innovation.ca/) awarded to AA. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
All mammals exhibit circadian (daily) patterns of behaviour and physiology regulated by a complex network of endogenous clocks and coordinated by the master pacemaker in the suprachiasmatic nucleus of the hypothalamus (SCN) in order to phase align to the solar day , . Since zeitgeber (German, from zeit ‘time’ and geber ‘giver’) information may not necessarily reach every tissue in the body directly, there is a necessity for a hierarchical organization of information processing from the master pacemaker to central and peripheral cellular clock networks . Furthermore, it has been experimentally demonstrated that there is an SCN-independent circadian system responsible for the rhythmic behavioural, tissue level, cellular, and molecular processes in response to patterns of food intake . Although the exact mechanism(s) responsible for the timekeeping properties of this pacemaker are presently unknown , this system is generally referred to as the food entrainable oscillator (FEO). The behavioural output of this FEO exhibits canonical properties of circadian pacemaker control including: the requirement of a zeitgeber for initial entrainment, a limited range of entrainment falling within a circadian timeframe (∼22–29 hrs) , , , persistence of precisely timed oscillations for several cycles during sustained food deprivation , , , and transient rather than immediate resetting after a change in meal time . Although many physiological processes undergo temporal realignment under this type of feeding schedule, the output of the FEO typically used as a behavioral measure is food anticipatory locomotor activity (FAA), which has a robust, high amplitude oscillation that is easily recorded automatically with minimal disturbance to the subject.
Accumulated evidence has lead several groups to propose that the FEO may be a distributed system of coupled central and peripheral oscillators, unlike the anatomically circumscribed system found in the SCN , , , . Indeed, one of the key differences between light-entrained circadian rhythms and the food entrainable system relies on a fundamental difference between the sensory mechanisms that provide their respective zeitgeber inputs. The SCN is entrained by light impinging on the retinal ganglion cells , , . In contrast, the food-entrainable system relies on the integration of food-derived cues that act at a myriad of peripheral and central sites throughout the nervous system, using various independent mechanisms .
In determining the brain regions that contribute to a distributed FEO, a major challenge is to dissociate neural activity of the FEO from those of the metabolic homeostat and the appetitive/affective centers, all of which interact and converge to produce behavioural FAA under restricted feeding (RF) conditions. In order meet this challenge, this study investigated the neural correlates of the FEO under conditions that minimize the secondary effects of restricted feeding, those that do not rely on circadian-driven anticipation of food, but rather are activated in response to metabolic challenge. This was accomplished by entraining animals to a daytime meal then eliminating the influence of a negative metabolic state by allowing three days of ad libitum feeding, a timeframe shown to be sufficient to abolish FAA in rats , , ,  and further confirmed in these mice. Returning animals to ad libitum feeding does not disrupt the phase or function of the FEO, as previous studies have shown that FAA can be reinstated by a 24-hour fast, even after many, many days of ad libitum feeding , . However, allowing ad libitum feeding reduced any neural activation due to the homeostatic, appetitive, and affective drives observed either during RF or a subsequent fasting period, thus allowing an examination of brain regions involved in the maintenance of food-related circadian rhythms, without the confound of a negative metabolic state.
Materials and Methods
All manipulations were conducted under the guidelines set forth by the Canadian Council on Animal Care and approved by Carleton University's Animal Care Committee.
Twenty-four CD1 adult male mice (Charles River Laboratories, St. Constant, QC; 20–25 gm upon arrival) were single-housed in standard mouse cages (Nalgene) and received ad libitum chow and water in a temperature and humidity controlled vivarium with a 12-hour light-dark (L∶D) cycle starting at 8:00 AM (Zeitgeber Time 0 (ZT0)). An additional 8 mice were housed in metabolic chambers (TSE Systems, Inc., Chesterfield, MO, USA) under identical environmental conditions in order to automatically assess rhythms of feeding and drinking.
In the first experiment, mice were randomly assigned to one of three groups: Ad libitum fed (AL, n = 8), restricted feeding schedule (RF, n = 8) or restricted feeding schedule then ad libitum fed (RF+AL, n = 8). Whereas inn the metabolic chambers, 3 mice were assigned to the AL condition, and 5 were placed in the RF+AL condition. Animals in the AL condition (negative control) received free access to food throughout the study. Animals in the RF group (positive control) had restricted access to food to 4 hours during the day (from ZT 4-ZT8), and were sacrificed after 14 days at approximately ZT2–3, prior to food access. Finally, mice in the RF+AL group were exposed to the same restricted feeding protocol as those in the RF group, but then were allowed ad libitum access to food for three days in order to abolish FAA before tissue collection at the time of previous meal anticipation (See Figure 1). In our second study, RF+AL animals in metabolic chambers had 7 days of scheduled feeding before returning to ad libitum food access for 3 days.
A, typical restricted feeding paradigm used in mice; including ad libitum baseline (AL), restricted feeding with food available only for a short temporal window during the light phase leading to entrainment (RF), ad libitum refeeding to abolish FAA (AL), and finally a two day fast to verify circadian entrainment (F). B, groups and treatments employed in this study. Restricted feeding occurred between ZT 4 and 8 during RF with all body weight and food consumption measures taken at ZT8.
Locomotor activity was measured in the animals' home cage using a Superflex HC infrared locomotor activity monitoring module connected to a computer (Accuscan Instruments, Columbus, OH). Ambulatory activity data, defined by beam breaks, were continually collected in 6 minute bins over 24 hours using Fusion HC v1.5 software (Accuscan Instruments, Columbus, OH). Several measures were calculated including food anticipatory activity (activity in the 2 hours preceding the presentation of scheduled meal), total daily activity, nocturnal activity, and nocturnality (defined as nocturnal activity as a percentage of total 24 hr activity). Daily food consumption and body weight were recorded at ZT8 using an SI-2002 digital scale (Denver Instruments, Bohemia, NY). The data was divided into each feeding condition for analysis 1) entrainment, days 1 through 14 and 2) ad libitum feeding, days 15–17. Repeated Measures ANOVAs (α = 0.05) and planned comparisons between experimental groups (AL vs. RF vs. RFAL) were performed using PASW statistics v17.0 (SPSS Worldwide, Chicago, IL) to analyze the effect of experimental group, time and any interaction between these two factors.
On the last day of behavioural recordings between ZT3–4, mice were injected with an overdose of sodium pentobarbital. A syringe with a 23G needle was then used to collect a blood sample from the atrium, followed by intra-cardial perfusion of 100 ml of saline (0.9%), then 200 ml of paraformaldehyde 4%. The animals were then decapitated and brains were collected and post-fixed overnight in the same fixative. Brains were then transferred into vials containing a 30% sucrose solution until sliced in their entirety using a Cryotome FSE cryostat (Thermo Fisher Scientific, Waltham, MA). Sections were divided into 4 serial sets of 60 µm slices and were immediately stored in a cryo-protectant solution at −20°C until processed for immunocytochemistry .
To examine neuronal activation, sections were processed for cFOS immunocytochemistry as described previously . In short, brain sections were rinsed with shaking 5×5 min in cold phosphate buffer (0.1 M PB; pH 7.3), incubated for fifteen minutes in a quenching solution (1% H2O2 in 0.1 M PB), rinsed three times in cold 0.1 M PB, followed by a blocking solution (0.3% Triton-X detergent and bovine serum albumen at 0.01 g/mL in 0.1 M PB) for thirty minutes at room temperature (RT), and finally c-FOS (Anti-c-FOS Ab-5 4–17 Rabbit pAb, Calbiochem) diluted 1∶20 000 in a solution identical to the blocking solution and agitated for forty-eight hours at 4°C. After incubation with the primary antibody, sections were rinsed (5×5 min. in 0.1 M PB) and incubated for one hour with a an anti-rabbit secondary antibody (Biotinylated goat Anti-Rabbit IgG H+L 1∶250; Vector Labs) diluted in 0.1 M PB containing Triton-X and BSA. After incubation with the secondary antibody, the sections were rinsed (3×5 min. in 0.1 M PB) at RT for one hour with the Avidin-Biotinylated Enzyme Complex (ABC Complex; Vector Labs). Sections were again rinsed (3×5 min. in 0.1 M PB) and incubated in a solution of 3,3′-diaminobenzidine (DAB; 1 mL per 16 mL of buffer; Sigma) and Cobalt Chloride (600 µL per 17 mL of 0.1 M PB), and developed for 4 min. using 25 µL of a 0.1% H2O2 solution. After final rinsing, sections were wet-mounted onto gelatin-coated slides, dehydrated in three consecutive series of alcohols and coverslipped with Permount (Fisher).
Quantification of immunocytochemistry.
Sections were examined using an Olympus BX51 brightfield microscope (Olympus Canada, Markham, ON) and associated digital capture device DVC-2000C (DVC Company, Austin, TX) connected to a desktop computer with Windows XP operating system (Microsoft Corp., Redmond, WA). Three-dimensional stereological reconstructions using every fourth brain slice across the entire extent of each ROI were used to estimate cell density and analyzed to determine changes in c-FOS IR. Due to various stochastic factors including: missing sections for some animals or failure to meet an appropriate Gunderson's coefficient of error (described below), the first 5 regions (out of the eight total), passing such criterion per group, per brain region, were retained for analysis using the optical fractionator probe within Stereo Investigator v 9 (MBF bioscience, Williston, VT). Regions of interest (ROI), chosen a priori based on previous investigations observing changes in c-FOS under RF , , , , ,  were identified using a mouse brain atlas (see Table 1 for a full list of regions, their associated map coordinates, and the sampling parameters used). Cell counting was performed using sampling parameters determined to be sufficient to produce a Gunderson's coefficient of error (GCE, m = 1) less than 0.1. The GCE (m = 1) has been determined to be the most appropriate coefficient of error estimate for all non-homogeneous biological tissues including cell populations within the brain , . Furthermore, a histogram analysis of the distribution of cells throughout the extent of each slice was performed while assessing other sampling parameters to ensure that no “dead zones” caused by sub-optimal antibody penetration were apparent. This analysis revealed an even distribution of cells along the z-axis when the uppermost and lowermost 4 µm of tissue were excluded. Sampling parameters for each ROI (listed in Table 1) were chosen to ensure that the estimated cell counts did not include undue stochastic variability (as represented by the GCE), and once determined, were held constant across all three experimental groups. Each ROI was outlined digitally (4× magnification) using the software provided while referring to the mouse brain atlas  for comparison. This information was then analyzed to create a volumetric estimate of each ROI measure in µm3. Each nucleus sampled was individually examined at high magnification using an oil immersion lens (60× magnification) and the uppermost tip of c-FOS positive nuclei were labeled with a digital marker. The software provided used both the planar (XY) and depth (Z) information recorded for each marker in the analysis of cell population estimates thereby creating a three dimensional model of the distribution of cells within each section from which an estimate of the total population throughout the entire extent of the ROI is created. Cellular population estimates were then divided by the estimated volume in order to create an estimate of the density of cells expressing c-FOS such that comparisons between animals take into account the intrinsic variability between the shape and size of ROIs. For each ROI, differences in mean cell densities between groups were analyzed using one-way ANOVAs and followed by planned comparison (LSD) tests with an α = 0.05 critical level. These analyses were conducted using PASW statistics v 17.0 software (SPSS Worldwide, Chicago, IL).
Quantification of PER2 positive cells in the stomach.
Paraformaldhyde-fixed stomachs were embedded in paraffin and sliced at a thickness of 15 µm using a vertical microtome, (n = 6 for each group), then mounted individually on glass slides using a heated waterbath. Each slice was then dewaxed and rehydrated before processing for DAB immunochemistry using an anti-PER2 primary antibody (1∶3,000, Alpha Diagnostic Inc., San Antonio, TX). PER2 expression was quantified by calculating the relative optical density (ROD). To this end, an Olympus BX51 brightfield microscope (Olympus Canada, Markham, ON) and associated digital capture device DVC-2000C (DVC Company, Austin, TX) were used to observe stomach slices at 100× magnification while holding the exposure time (25 ms) and irradiance (full power with ND6 and ND25 filters engaged) constant. Images were digitized using the image capture function within Stereo Investigator v 9 (MBF bioscience, Williston, VT) and saved as 16bit grayscale TIFF images. Relative optical density was calculated using the measure function within Image J v 1.4 (National Institute of Health, USA). For each slice, the polygon selection tool was used to outline both the oxyntic mucosa and a blank area of the slide in order to measure the mean grey values of each. This returned a value that represents the average of all grey values (with a possible value of 1 for black to 256 for white). In order to determine the optical density of each region, these values were then transformed to fit a linear standard curve where 100 represented the highest optical density and 0 represented the lowest [(100-(Y/256*100)), where Y represents the mean grey value]. Finally, the ROD was calculated by subtracting the optical density of the background from that of the oxyntic mucosa. Values were then analyzed using PASW statistics v 17.0 (SPSS Worldwide, Chicago, IL) to perform a one-way ANOVA with posthoc LSD test (α = 0.05) and any group differences between experimental conditions were observed.
Ghrelin Enzyme Linked Immunoassay (ELISA).
Blood samples obtained at the time of sacrifice (about 400 µl each) were collected in EDTA laced tubes containing 5 µl of 0.1 M HCl and 5 µl of p-hydroxymercuribenzoic acid (PHMB: Fluka) at a concentration of 100 nM to slow the denaturing of the active form of ghrelin. Samples were then centrifuged (300 rpm for 5 min. at 4°C) to separate plasma. Plasma was stored at −80°C until processing for ELISA. Samples from 3 mice in the AL group, 2 in the RF group and 1 in the RFAL group were lost during collection. Plasma content of active (acylated) ghrelin from the remaining samples was determined using a commercially available ELISA kit (EZRGRA-90K, Millipore). All samples were assayed in duplicate and differences were evaluated using a one-way between groups ANOVA. Interassay variability was less than 10%.
Scheduled feeding was associated with changes in food intake and body weight. Mice in the RF groups consumed less food during the time they had access to it than control ad lib fed mice (F(2,24) = 4.972, p<0.05, Fisher's LSD). However, animals in the RFAL group ate more than ad lib controls once they were again allowed ad lib access (F(1,14) = 34.892, p<0.05, LSD; see Figure 2). Furthermore, there were no overall significant differences in body weight between groups during or after the entrainment period (p>0.05).
A, average body weights for each group at the end of each feeding treatment. B, average daily food consumption for each group across days. Note the sharp peak in rebound feeding after a return to ad libitum. However, both the initial reduction and subsequent rebound in feeding had no effect on body weight across conditions suggesting internal metabolic compensation/protection against food restriction. C, average daily values for ambulatory activity during the time of meal anticipation (ZT4–8). Note that FAA develops slowly over several days as the animals entrain to their new feeding schedule but sharply declines back to baseline values on day 15 after being returned to ad libitum feeding. D, nocturnality scores for animals during RF and subsequent AL feeding. Nocturnality is defined as the total ambulatory activity at night divided by total daily activity. E, average waveforms of all animals during the last week of restricted feeding. Note the sharp increase of FAA prior to meal presentation. F, average waveform during the third day of ad libitum feeding, The anticipatory peak and the increase in night time activity was completely abolished after a return to ad libitum feeding, although daytime activity did appear to remain slightly higher than AL controls. Note: 1 hr MA represents a smoothed curve where the moving average, the average score for all data points within the hour surrounding each time point, was plotted. Gray bars indicate the time of food availability and black triangles indicate the first day of ad libitum feeding.
As expected, restricted feeding schedules produced significant alterations in daily patterns of locomotor activity. Specifically, anticipatory locomotor activity was elevated in the RF and RFAL groups relative to AL controls (F(6,120) = 4.125, p.<0.05 LSD; see Figure 2). Nocturnality was similarly reduced in the RF and RFAL groups as compared to controls (F(6,120) = 0.240, p = 0.788, p.<0.05, LSD; see Figure 2). During the subsequent ad libitum feeding period, there was a significant change in FAA but not nocturnality across days (F(2,26) = 10.229, p = 0.004 and F(2,26) = 1.639, p = 0.218 respectively) according to the Greenhouse-Geisser criterion. However, both the expression of FAA and the nocturnality score were not significantly different from the AL group by the third day (F(1,13) = 1.351, p = 0.266 and F(1,13) = 4.292, p = 0.059 respectively; see Figure 2). These effects, as well as the temporal structure of activity across days are also visible in the representative actograms provided (Figure 3).
Note both the quantity and temporal structure of FAA during the restricted feeding paradigm and its loss after the RFAL animals are placed back on ad libitum feeding. Gray bars indicate the time of food access for the RF and RFAL groups.
Cell density estimates of c-FOS positive cells revealed a distinct pattern of neuronal activation as compared to AL controls for both the RF and RFAL groups (see Figure 4) for bar graphs and Table 2 for accompanying ANOVA values). The following areas exhibited significant changes in the number of c-FOS expressing cells between groups: AcbC, AcbSH, Cg1, VTA, SCN, LH, PVN, VMH, ARC, DMH, DR, PB, NTS, LC, BNSTov, CeA, BLA, and DG. Planned comparison analysis revealed a significant increase in the number of c-FOS expressing cells in the RF group, as compared to AL negative controls, for all regions examined with only four exceptions: the PFC and VTA remained at AL levels (although the VTA does exhibit a statistical trend), while the SCN and PVN were lower than their respective AL counterparts (Figure 4 and Figure 5). Interestingly, animals in the RFAL group still showed RF-like patterns of c-FOS expression in certain hypothalamic and extra-hypothalamic brain regions including the ARC, DMH, DR, PB and NTS (Figure 4 and Figure 6). Expression of c-FOS in all other regions, however, returned to levels similar to or less than the AL group.
A, Affective system nuclei. B, Appetitive system nuclei C, Hypothalamic nuclei D, Brainstem nuclei. Stars indicates a significant difference from AL controls (p<0.05), crosses indicate significant difference from RF group (p<0.05). Hollow symbols denote differences approaching significance (p<0.1).
Notice that c-FOS appears preferentially suppressed in the core nucleus of the SCN but not the shell, whereas the PVN is suppressed in both the parvocellular and magnocellular nuclei.
Note the increased c-FOS expression in the RF group as compared to AL controls and that this increase is maintained after returning to ad libitum conditions (RF+AL), even after FAA is abolished.
A possible alternative explanation for the observation of elevated c-FOS expression in certain brain regions of RF mice, even after being placed back on ad libitum feeding conditions, is that they began eating during the morning after lights-on, during the period when they previously showed FAA. If animals had just consumed a large meal immediately prior to being sacrificed, satiety signals from the gut or circulating hormones could cause c-FOS activation in areas such as the ARC and NTS . In order to address this possibility, separate groups of animals were run through a similar feeding protocol within metabolic chambers, where their food and water intake could be monitored continuously without disturbing the animals. This also had the benefit of avoiding a confounding timing cue i.e. the presence of a researcher at previous meal time, even in the absence of food removal or presentation after a return to ad libitum feeding. We found that animals returned to ad libitum feeding conditions did not eat more than the continuously AL animals during the period between lights on and the time of previous food presentation (ZT0–4), when FAA occurs under RF (Figure 7, middle panels). During this 4 hour period, AL animals ate an average of 0.35 g of chow (±0.071 g), which did not significantly differ from the food intake of RF+AL animals (T = 0.86, p>0.05). Interestingly, confirming previous reports from a similar feeding paradigm conducted in rats , RF+AL animals did continue to eat a daytime meal during the former food presentation period (ZT4–8; see Figure 7, bottom panels), consuming an average of 1.8 g of chow (±0.50 g), significantly more than AL animals (0.30 g of chow ±0.15 g, T = 2.8, p<0.05). This confirms that RF+AL animals did not eat a large meal prior to sacrifice, and instead supports the notion that the FEO continues to operate and entrain feeding, even in the absence of anticipatory behavior mediated by homeostatic and appetitive drives.
Average waveforms representing food consumed over twenty-four hours by both AL and RF+AL groups under baseline, restricted feeding, and return to ad libitum conditions. Bar graphs to the right of each 24-hr profile show the amount of food consumed from lights on to the scheduled meal time (ZT0–4) and the amount of food consumed during the 4 hours of food availability (ZT4–8) for each condition.
As expected, analysis of plasma active ghrelin concentrations determined that there was a significant group difference (F(2, 17) = 12.994, p<0.001). Planned comparison analysis revealed that while animals in the RF group developed elevated acylated ghrelin concentrations in anticipation of the scheduled meal (p = 0.001), this was no longer observable after a return to ad libitum feeding, i.e. the RFAL group did not significantly differ from AL controls (p = 0.824). This pattern was also accompanied by elevated levels of the clock protein PER2 within the stomach in the RF group (p = 0.023) and a return to AL levels after several days of ad libitum feeding (p = 0.940) (see Figure 8).
A, Average RODs of PER2 in the oxyntic mucosa of the stomach. B, Average concentrations of circulating active ghrelin. C, Representative images of PER2 staining. Note: stars indicates a significant difference from AL controls (p<0.05), crosses indicate significant difference from RF group (p<0.05).
The current project examined whether neuronal activity stimulated by scheduled meals would persist in a sub-population of brain structures following the termination of the restricted feeding schedules, thereby highlighting the potential brain structure(s) housing the pacemaker of the food entrainable oscillator. Our results show that a number of hypothalamic and brain stem structures continue to show elevated c-FOS expression days after the termination of the restricted feeding schedule and return to ad libitum feeding.
Confirmation of food entrainment is generally tested by assessing the re-emergence of anticipatory behaviours in animals that are food deprived following a protracted ad libitum feeding period after entrainment by restricted feeding. We reasoned that if the FEO is a true pacemaker entrained by restricted feeding, then a neural representation of this timekeeping mechanism should persist, observable as persistent neuronal activity, following the restricted feeding schedule, much in the same way as the SCN continues to exhibit rhythmic neural activation after removal of the zeitgeber (i.e. under constant darkness conditions) . This process would facilitate the appropriate timing of food anticipation the next time an animal encountered a shortage of food. Similarly, if the FEO is a true pacemaker then, once entrained, it must also produce self-sustained oscillations that are observable as a pattern of neural activity with stable peak-to-peak intervals. Comparing neuronal activation at the same circadian time in animals before, during, and after the restricted feeding schedule, was an attempt to sample the presumptive peak activation within these oscillating brain regions. We also reasoned that this neural representation should continue to be observed regardless of the metabolic state of the animal. This canonical pacemaker property, continued oscillations in the absence of a zeitgeber, was exploited in order to avoid c-FOS expression due to metabolic demand specifically in regions of the brain that are critically implicated in the homeostasis of food intake and metabolism. In examining these regions, we observed persistent c-FOS expression in select hypothalamic and caudal brainstem sites critical for nutrient sensing and energy balance, in the absence of any observable anticipatory locomotor activity. For instance, persistently increased c-FOS expression was detected in the DMH, ARC, DR, PB and NTS of mice in the RFAL group see (Figure 4). Individually, all of these regions have been investigated to determine whether they contain the FEO. Lesions destroying each of these nuclei individually, or targeted mutations to genes expressed by neurons in these regions, have been reported to transiently alter the ability of rats and mice to entrain to feeding schedules, and/or to show attenuated food anticipatory activity , . To date, however, there is no single brain nucleus that has been pinpointed as both necessary and sufficient for the entrainment to feeding schedules . In light of the findings presented here, this is not surprising, as all regions identified are sensitive to peripheral signals conveying information about metabolic state, and destruction of any of these individually may not be sufficient to prevent information about energy availability from reaching the brain and ultimately entraining biological rhythms to the time in which food is available.
Persistent c-FOS expression in the DMH further supports the importance of this hypothalamic region in the mechanisms underlying expression of FAA. It is still unclear whether the DMH is capable of intrinsic rhythmicity. Recent reports of circadian oscillations at the tissue level observed that rhythms of the DMH damped out very quickly in vitro, suggesting that it is not capable of self-sustained rhythms. However analysis at the single cell level suggests that it is a lack of synchrony between cells and not a lack of intrinsic rhythmicity which is the culprit. This strongly suggests that the DMH is capable of self-sustained oscillation at the cellular level, but that it requires an extrinsic coupling factor (possibly feeding-related) in order for continued coordination at the tissue level . Neuronal activity in the DMH is increased in anticipation of scheduled meals, but it is also elevated around the previous meal time in rats that are placed in restricted feeding schedules where food is provided at random times during the day . The rhythms of at least one clock gene, PER1, also tracks approximate meal presentation time, even in the absence of FAA , but peaks differentially according to meal duration . This suggests an underlying sensitivity to metabolic demand, rather than a precisely timed mechanism tuned to food availability. Furthermore, inhibition of the SCN by the DMH has emerged as a mechanism whereby SCN neuronal activity is suppressed, allowing anticipatory locomotor activity to be expressed during a time when nocturnal animals are normally inactive. It was demonstrated that this suppression is necessary for the full expression of FAA  but a recent study suggests that this is only necessary if the meal presentation occurs outside of the normal period of arousal . The DMH therefore represents one of several nuclei which disengage the SCN in order to allow the appropriate expression of FAA , while the observation of persistent activation within this region further suggests that it is integrated into a spatially distributed FEO pacemaker or, at the very least, that it is preferentially sensitive to FEO, rather than SCN, outputs.
Interestingly, the ARC, an area critical for energy balance , also remains active in animals previously exposed to a restricted feeding schedule (RF+AL group). Within the ARC, neurons release peptides critical for the regulation of metabolism, with α-MSH being important for increasing metabolic rate and decreasing food intake, and the agouti related peptide producing orexigenic responses and decreases in energy expenditure to promote obesity . The effects of both these peptides are mediated by the same receptors, the MC3/4 receptors, located throughout the hypothalamus, and with very dense concentrations in the PVN, DMH and LH. Interestingly, mice with targeted deletions of these receptors show attenuated food anticipatory responses when exposed to scheduled meals , . It is therefore possible that activation of the ARC during restricted feeding schedules is selective to neurons secreting agouti related peptide, and that these remain active in spite of the animals being returned to ad lib conditions. This idea is supported indirectly in our data by the observation of significantly elevated rebound feeding response once the mice are allowed to have free access to food (See Figure 2b) and it is also consistent with the role of Neuropeptide Y as an SCN suppressing signal when considering the extensive co-localization, and co-release, of these two neuropeptides within ARC neurons of animals under metabolic demand .
As expected, c-FOS expression was increased in other hypothalamic regions implicated in food intake and metabolism including the VMH and LH. However this activation was likely due to metabolic demand and/or the result of outputs signals of the FEO, or release from inhibitory inputs from the SCN, as it was only observable during RF. In contrast, the SCN and PVN of mice under the restricted feeding schedule showed decreased c-FOS expression, with neuronal activity in these areas being restored to baseline levels by a return to ad libitum feeding (Figure 4 and Figure 6). This is in agreement with similar observations in rats during FAA, at least for the SCN , . In terms of both evolutionary adaptation, as well as economy of function, the necessity for SCN-uncoupling during FAA is not surprising, and has been proposed as a mechanism facilitating entrainment to feeding schedules , , , , . A recent investigation has implicated the DMH in the suppression of SCN activity, specifically within the ventrolateral core, and this occurs via direct axonal projections from food sensitive neurons within this nucleus .
While c-FOS activation of the PVN was not expected, the observed decrease in c-FOS-IR in RF animals has not been widely reported, but was found in rats simply fasted for 64 hours . That we observe this decrease may be due to the power of the stereological approach but may also be related to the extreme fragility of mice under metabolic demand. Inhibition of activity within the PVN in food-restricted mice may also be related to increases in activity in other brain regions including the DMH and the BNSTov. Like the SCN, the PVN is directly innervated by both glutamatergic and GABAergic projections from the DMH , . Interestingly, even though both excitatory and inhibitory cells are found in the DMH, lesions of the DMH lead to increased hypothalamic-pituitary axis responses to stressors  and pharmacological stimulation of DMH neurons produces inhibitory post-synaptic potentials in neuroendocrine cells of the PVN . This is consistent with a strictly inhibitory role of the DMH on the PVN but implies an ability of the DMH outputs to switch from inhibitory to excitatory depending on the intrinsic rhythms of the animal, nocturnal or diurnal, as has been recently proposed . Similarly the PVN is heavily innervated by the BNSTov  via GABAergic projections , . Ablation of the posterior BNST, where the oval nucleus is situated, increases c-FOS expression and CRH release within the PVN, indicating that the BNSTov plays a significant role in the suppression of these two mechanisms within the neuroendocrine system . It is proposed by Cullinan et al.  that this represents a braking mechanism by which afferents of the forebrain limbic system can modulate the stress response, and this agrees with the results achieved in this experiment. While acute food deprivation has been shown to increase the stress response, repeated exposure to a predictable meal schedule should ideally decrease the stress response due to an ability to anticipate the near-future availability of food and the BNSTov likely plays a major role in the neural mechanism responsible.
Neurons in the NTS and PB are important targets for both circulating metabolic hormones, as well as visceral signals conveyed through the vagus nerve , , . Because of this, both of these structures have been suspected in participating in the processes underlying food entrainment, although lesions to PB attenuated FAA but did not appreciable affect entrainment . Similarly vagal deafferentation has no effect on FAA . Yet, persistent activation of these brain regions may reflect the action of endocrine and/or afferent visceral signals that continues to act in spite of ad lib food availability. Alternatively, continued activation of these brain stem nuclei may be tied to food consumed just prior to the animals being sacrificed. Indeed, signals like leptin and cholecystokinin (CCK) act directly and synergistically to stimulate these brain stem regions and induce satiety . In order to address this, our second experiment examined patterns of food consumption in animals under the various schedules. We observed continued daytime feeding behaviour, even after a return to ad libitum food availability (see Figure 7), however this occurred entirely during the time of previous meal presentation (ZT4–8) indicating that our RF+AL mice would not have consumed a meal prior to sacrifice and that the c-FOS expression observed is most likely a neural representation maintained by exposure to a previous restricted feeding schedule. Furthermore, if the animals had indeed eaten prior to sacrifice we would expect that mice in the RF+AL condition would have much greater c-FOS expression in the NTS and PB as a result of that meal compared to mice in either the AL or RF group . It is important to add that visual inspection of the stomach at the time of sacrifice provided no clear indication that the mice in the RF+AL group had consumed more food than animals in the other groups just prior to sacrifice, again supporting the idea that persistent c-FOS expression in the NTS and PB is related to the previous restricted feeding schedule and not to recent caloric intake. Critically, this hypothesis is further supported by the return of clock protein PER2 expression in the oxyntic mucosa of the stomach - putatively linked to rhythmic food intake  - to ad libitum levels after only three days. This last finding fits with PER1 transcription timing from a previous in vitro experiment in which per1:luciferase rhythms from rat stomachs were entrained to RF schedules and similarly returned to their original phase within days of a return to ad libitum feeding .
Observations of increased activation in the brainstem in anticipation of a meal, as well as the continued activation of the NTS and PB after a return to ad libitum feeding, conflicts with the findings of two other groups conducted in rats. In the first study , the authors maintain that there was no anticipatory c-FOS expression observed in the NTS, PB or the Area Postrema as compared to ad libitum fed controls, and activation was only observed after the feeding bout. While it was true that there was increased activation (almost four-fold) observed after feeding in all three of these brain stem nuclei related to visceral signals of satiety, a closer inspection of their results prior to the expected presentation of food showed apparent increases in activation of the medial NTS and medial PB during food entrainment and the medial PB during fasting after entrainment. These time points and nuclei correspond directly with this study. These changes appear statistically significant, although this was not tested in their study, but were much smaller in magnitude to the drastically increased activation observed after feeding. This small change may represent a subpopulation of cells within these regions and if this is the case our stereological analysis would be much more likely to resolve this for the following reason: Angeles-Castellanos et al., used a single 40 µm brain slice for cell counting in each region of interest. By comparison, the present study performed a stereological investigation of each region, creating an estimate of the activation over the entire extent of each ROI and therefore had greater power to detect small but significant changes both in anticipation of the meal (RF) and at the time of previous anticipation (RFAL). In the second study , Poulin et al. observed no anticipatory increase of c-FOS mRNA within the NTS. The reasons for this may be two-fold. Measurements of optical density from in situ hybridization reflects a direct correlation of the cfos mRNA levels. In contrast, our method of counting positively labeled c-FOS protein expressing cells does not correlate directly with the amount of c-FOS protein expression. The number of immunopositive cells is determined by the specific conditions and method used to visualize immunoreactivity, including the threshold of detection of our antibody at the concentration used, the amount of time the DAB reaction was developed, etc. From this, a binary analysis is then performed whereby the observer answers a simple question. Is protein visible compared to background, i.e. can a nucleus be visualized. If the answer is yes, the cell is counted. This creates a situation whereby even small changes in the protein level of each cell can quite significantly alter the total number of c-FOS positive cells. A more direct measurement of the relative amount of c-FOS protein can be done using a Western blot, but this does not allow for precise anatomical localization of the protein. Secondly, while protein levels typically follow up-regulation of mRNA, it is more difficult to compare timing of expression, as mRNA translation and protein transcription can be differentially regulated. There are also post-translational events which can regulate protein expression independently of transcription as has been previously shown for c-FOS . Any of these could account for a discrepancy between the results suggested by in situ hybridization vs. immunocytochemisry.
Finally, the dorsal raphé nucleus showed increased c-FOS expression in anticipation of the scheduled meal and this activity persisted in animals that were allowed to return to ad lib feeding. While there is little data linking the raphé nuclei with the generation of food entrainment, this region has been implicated as a potential photic and non-photic modulator of circadian rhythms of activity via it projections to the SCN , , , , , . Interestingly, the raphé of rats exposed to restricted feeding schedules (food available for two hours during the day for 15 days early in the light phase of the light dark cycle) show increased c-FOS expression during food search (i.e. prior to meal presentation) in comparison with ad lib controls and animals on the same schedule that are allowed to eat until satiation . Our study confirms these data in mice, and shows that the raphé continues to be active in anticipation of a previously scheduled meal.
An increase in c-FOS across brain regions along mesolimbic and mesostriatal dopaminergic reward pathways was observed in anticipation of a scheduled meal. These data confirms previous reports implicating reward processes in food anticipation , , , , . However, neuronal activity in these regions return to baseline, indicating that the effects of the restricted feeding schedule on the reward circuits do not persist beyond the restriction paradigm. Similarly, brain circuit associated with arousal (LH, LC) as well as those mediating emotional responses (BNSTov, hippocampus, amygdala) follow the same pattern. From this we infer either a diminished or secondary role of these other nuclei in the maintenance of food availability timekeeping while still supporting their necessity for the appropriate expression of food anticipatory behaviours and physiology.
Finally, a similar same pattern was observed for plasma concentration of acylated ghrelin, consistent with a return of PER2 levels in the stomach to baseline. It has previously been suggested that ghrelin may be either a tuning signal from the periphery  or a coupling mechanism between pacemaking components of the FEO in the stomach and brain . The present findings confirm that levels of the active form of ghrelin does not persist after a return to ad libitum feeding, nor does the increased level of PER2 in the stomach, thus the circadian oscillators in the stomach are not critically involved in the pacemaking mechanism of the FEO. Therefore, ghrelin signaling may act as a tuning mechanism, and possibly as one of many zeitgebers, conveying information about food availability from the periphery to numerous brain sites sensitive its signal—including many of those investigated in this study—in order to promote neural activation necessary for the appropriate expression of FAA.
In the present study, we have revealed brain areas which remain active in mice previously exposed to a restricted feeding schedule and then allowed free access to food, in comparison to ad lib fed mice (negative controls) and in mice that remained under the restricted feeding schedule (positive controls). Our findings, while not comprehensive, confirm that a number of hypothalamic and brain stem regions persistently activate at the time of prior meal anticipation including the DMH, ARC, DR, PB, and NTS. Furthermore this persistent neuronal activity was not ubiquitous; suggesting that any region with continued activation at the time of the previous scheduled meal may be directly involved in maintaining a neural representation necessary for timekeeping of previous food availability. A number of groups have proposed that the FEO may indeed be a network of interconnected brain structures , , , ,  and this proposition has recently received a thorough review . It is our belief that this is indeed the case, and while a correlative study such as the one presented here is incapable of incontrovertibly demonstrating that all structures identified are both necessary and sufficient for the pacemaker function of the FEO, both the number of well documented interconnections between the regions identified here, as well as the sheer number of regions identified as persistently active further suggests that this is the case. Our assertion is also based on a principle assumption of this study: that after a return to ad libitum conditions ONLY the regions that are necessary for timekeeping would remain preferentially activated. We also assert that there is no conceivable reason why either upstream effectors (as there is no food, peripheral or other cues expected/presented at this timepoint) nor downstream affectors (which are reset to normal, SCN driven, light/activity/food consumption/body temperature rhythms for appropriate control of behaviour and physiology under ad libitum conditions) should be influencing the levels of c-FOS expression in regions not critically involved in the timekeeping mechanism of the FEO.
This study thereby contributes to the search for a neural system tasked with the circadian entrainment to daily meals (the FEO) by narrowing down a large list of previously proposed candidates and presents an experimental paradigm useful for future studies in which it will be necessary to dissociate ancillary activation due to the various homeostatic, appetitive, and affective components driving FAA.
Conceived and designed the experiments: IDB EWL AA. Performed the experiments: IDB EWL TR. Analyzed the data: IDB EWL AA. Contributed reagents/materials/analysis tools: AA. Wrote the paper: IDB EWL TR AA.
- 1. Golombek DA, Rosenstein RE (2010) Physiology of Circadian Entrainment. Physiol Rev 90: 1063–1102.DA GolombekRE Rosenstein2010Physiology of Circadian Entrainment.Physiol Rev9010631102
- 2. Mistlberger RE (2011) Neurobiology of food anticipatory circadian rhythms. Physiology & behavior 104: 535–545.RE Mistlberger2011Neurobiology of food anticipatory circadian rhythms.Physiology & behavior104535545
- 3. Hirota T, Fukada Y (2004) Resetting mechanism of central and peripheral circadian clocks in mammals. Zoolog Sci 21: 359–368.T. HirotaY. Fukada2004Resetting mechanism of central and peripheral circadian clocks in mammals.Zoolog Sci21359368
- 4. Mistlberger RE (1994) Circadian food-anticipatory activity: formal models and physiological mechanisms. Neurosci Biobehav Rev 18: 171–195.RE Mistlberger1994Circadian food-anticipatory activity: formal models and physiological mechanisms.Neurosci Biobehav Rev18171195
- 5. Blum ID, Lamont EW, Abizaid A (2012) Competing clocks: Metabolic status moderates signals from the master circadian pacemaker. Neuroscience and biobehavioral reviews 36: 254–270.ID BlumEW LamontA. Abizaid2012Competing clocks: Metabolic status moderates signals from the master circadian pacemaker.Neuroscience and biobehavioral reviews36254270
- 6. Bolles RC, Stokes LW (1965) Rat's anticipation of diurnal and a-diurnal feeding. J Comp Physiol Psychol 60: 290–294.RC BollesLW Stokes1965Rat's anticipation of diurnal and a-diurnal feeding.J Comp Physiol Psychol60290294
- 7. Stephan FK, Swann JM, Sisk CL (1979) Anticipation of 24-hr feeding schedules in rats with lesions of the suprachiasmatic nucleus. Behav Neural Biol 25: 346–363.FK StephanJM SwannCL Sisk1979Anticipation of 24-hr feeding schedules in rats with lesions of the suprachiasmatic nucleus.Behav Neural Biol25346363
- 8. Stephan FK, Swann JM, Sisk CL (1979) Entrainment of circadian rhythms by feeding schedules in rats with suprachiasmatic lesions. Behav Neural Biol 25: 545–554.FK StephanJM SwannCL Sisk1979Entrainment of circadian rhythms by feeding schedules in rats with suprachiasmatic lesions.Behav Neural Biol25545554
- 9. Rosenwasser AM, Pelchat RJ, Adler NT (1984) Memory for feeding time: possible dependence on coupled circadian oscillators. Physiol Behav 32: 25–30.AM RosenwasserRJ PelchatNT Adler1984Memory for feeding time: possible dependence on coupled circadian oscillators.Physiol Behav322530
- 10. Boulos Z, Rosenwasser AM, Terman M (1980) Feeding schedules and the circadian organization of behavior in the rat. Behavioural Brain Research 1: 39–66.Z. BoulosAM RosenwasserM. Terman1980Feeding schedules and the circadian organization of behavior in the rat.Behavioural Brain Research13966
- 11. Clarke JD, Coleman GJ (1986) Persistent meal-associated rhythms in SCN-lesioned rats. Physiol Behav 36: 105–113.JD ClarkeGJ Coleman1986Persistent meal-associated rhythms in SCN-lesioned rats.Physiol Behav36105113
- 12. Stephan FK (1984) Phase shifts of circadian rhythms in activity entrained to food access. Physiol Behav 32: 663–671.FK Stephan1984Phase shifts of circadian rhythms in activity entrained to food access.Physiol Behav32663671
- 13. Verwey M, Amir S (2009) Food-entrainable circadian oscillators in the brain. Eur J Neurosci 30: 1650–1657.M. VerweyS. Amir2009Food-entrainable circadian oscillators in the brain.Eur J Neurosci3016501657
- 14. Carneiro BTS, Araujo JF (2009) The Food-entrainable Oscillator: a network of interconnected brain structures entrained by humoral signals? Chronobiology International 26: 1273–1289.BTS CarneiroJF Araujo2009The Food-entrainable Oscillator: a network of interconnected brain structures entrained by humoral signals?Chronobiology International2612731289
- 15. Escobar C, Cailotto C, Angeles-Castellanos M, Delgado RS, Buijs RM (2009) Peripheral oscillators: the driving force for food-anticipatory activity. Eur J Neurosci 30: 1665–1675.C. EscobarC. CailottoM. Angeles-CastellanosRS DelgadoRM Buijs2009Peripheral oscillators: the driving force for food-anticipatory activity.Eur J Neurosci3016651675
- 16. Provencio I, Jiang G, De Grip WJ, Hayes WP, Rollag MD (1998) Melanopsin: An opsin in melanophores, brain, and eye. Proc Natl Acad Sci U S A 95: 340–345.I. ProvencioG. JiangWJ De GripWP HayesMD Rollag1998Melanopsin: An opsin in melanophores, brain, and eye.Proc Natl Acad Sci U S A95340345
- 17. Hattar S, Lucas R, Mrosovsky N, Thompson S, Douglas R, et al. (2003) Melanopsin and rod-cone photoreceptive systems account for all major accessory visual functions in mice. Nature 424: 76–81.S. HattarR. LucasN. MrosovskyS. ThompsonR. Douglas2003Melanopsin and rod-cone photoreceptive systems account for all major accessory visual functions in mice.Nature4247681
- 18. Schmidt TM, Kofuji P (2008) Novel insights into non-image forming visual processing in the retina. Cellscience 5: 77–83.TM SchmidtP. Kofuji2008Novel insights into non-image forming visual processing in the retina.Cellscience57783
- 19. Dibner C, Schibler U, Albrecht U (2010) The Mammalian Circadian Timing System: Organization and Coordination of Central and Peripheral Clocks. Annual Review of Physiology 72: 517–549.C. DibnerU. SchiblerU. Albrecht2010The Mammalian Circadian Timing System: Organization and Coordination of Central and Peripheral Clocks.Annual Review of Physiology72517549
- 20. Coleman GJ, Harper S, Clarke JD, Armstrong S (1982) Evidence for a separate meal-associated oscillator in the rat. Physiology & behavior 29: 107–115.GJ ColemanS. HarperJD ClarkeS. Armstrong1982Evidence for a separate meal-associated oscillator in the rat.Physiology & behavior29107115
- 21. Mistlberger RE (2009) Food-anticipatory circadian rhythms: concepts and methods. Eur J Neurosci 30: 1718–1729.RE Mistlberger2009Food-anticipatory circadian rhythms: concepts and methods.Eur J Neurosci3017181729
- 22. Watson RE, Wiegand SJ, Clough RW, Hoffman GE (1986) Use of cryoprotectant to maintain long-term peptide immunoreactivity and tissue morphology. Peptides 7: 155–159.RE WatsonSJ WiegandRW CloughGE Hoffman1986Use of cryoprotectant to maintain long-term peptide immunoreactivity and tissue morphology.Peptides7155159
- 23. Abizaid A, Mezei G, Thanarajasingam G, Horvath TL (2005) Estrogen enhances light-induced activation of dorsal raphe serotonergic neurons. European Journal of Neuroscience 21: 1536–1546.A. AbizaidG. MezeiG. ThanarajasingamTL Horvath2005Estrogen enhances light-induced activation of dorsal raphe serotonergic neurons.European Journal of Neuroscience2115361546
- 24. Challet E, Jacob N, Vuillez P, Pevet P, Malan A (1997) Fos-like immunoreactivity in the circadian timing system of calorie-restricted rats fed at dawn: daily rhythms and light pulse-induced changes. Brain Res 770: 228–236.E. ChalletN. JacobP. VuillezP. PevetA. Malan1997Fos-like immunoreactivity in the circadian timing system of calorie-restricted rats fed at dawn: daily rhythms and light pulse-induced changes.Brain Res770228236
- 25. Choi S, Wong LS, Yamat C, Dallman MF (1998) Hypothalamic ventromedial nuclei amplify circadian rhythms: do they contain a food-entrained endogenous oscillator? J Neurosci 18: 3843–3852.S. ChoiLS WongC. YamatMF Dallman1998Hypothalamic ventromedial nuclei amplify circadian rhythms: do they contain a food-entrained endogenous oscillator?J Neurosci1838433852
- 26. Ribeiro AC, Sawa E, Carren-LeSauter I, LeSauter J, Silver R, et al. (2007) Two forces for arousal: Pitting hunger versus circadian influences and identifying neurons responsible for changes in behavioral arousal. Proc Natl Acad Sci U S A 104: 20078–20083.AC RibeiroE. SawaI. Carren-LeSauterJ. LeSauterR. Silver2007Two forces for arousal: Pitting hunger versus circadian influences and identifying neurons responsible for changes in behavioral arousal.Proc Natl Acad Sci U S A1042007820083
- 27. Takase LF, Nogueira MI (2008) Patterns of fos activation in rat raphe nuclei during feeding behavior. Brain research 1200: 10–18.LF TakaseMI Nogueira2008Patterns of fos activation in rat raphe nuclei during feeding behavior.Brain research12001018
- 28. Angeles-Castellanos M, Mendoza J, Diaz-Munoz M, Escobar C (2005) Food entrainment modifies the c-Fos expression pattern in brain stem nuclei of rats. Am J Physiol Regul Integr Comp Physiol 288: R678–684.M. Angeles-CastellanosJ. MendozaM. Diaz-MunozC. Escobar2005Food entrainment modifies the c-Fos expression pattern in brain stem nuclei of rats.Am J Physiol Regul Integr Comp Physiol288R678684
- 29. Angeles-Castellanos M, Aguilar-Roblero R, Escobar C (2004) c-Fos expression in hypothalamic nuclei of food-entrained rats. Am J Physiol Regul Integr Comp Physiol 286: R158–165.M. Angeles-CastellanosR. Aguilar-RobleroC. Escobar2004c-Fos expression in hypothalamic nuclei of food-entrained rats.Am J Physiol Regul Integr Comp Physiol286R158165
- 30. Gundersen HJ, Jensen EB, Kieu K, Nielsen J (1999) The efficiency of systematic sampling in stereology–reconsidered. J Microsc 193: 199–211.HJ GundersenEB JensenK. KieuJ. Nielsen1999The efficiency of systematic sampling in stereology–reconsidered.J Microsc193199211
- 31. Gundersen HJ, Jensen EB (1987) The efficiency of systematic sampling in stereology and its prediction. J Microsc 147: 229–263.HJ GundersenEB Jensen1987The efficiency of systematic sampling in stereology and its prediction.J Microsc147229263
- 32. Paxinos G, Franklin KBJ (2001) The Mouse Brain in Stereotaxic Coordinates. New York, NY: Academic Press. G. PaxinosKBJ Franklin2001The Mouse Brain in Stereotaxic CoordinatesNew York, NYAcademic Press
- 33. Woods SC (2004) Gastrointestinal satiety signals I. An overview of gastrointestinal signals that influence food intake. American journal of physiology Gastrointestinal and liver physiology 286: G7–13.SC Woods2004Gastrointestinal satiety signals I. An overview of gastrointestinal signals that influence food intake.American journal of physiology Gastrointestinal and liver physiology286G713
- 34. Mori T, Nagai K, Nakagawa H (1983) Dependence of memory of meal time upon circadian biological clock in rats. Physiology & Behavior 30: 259–265.T. MoriK. NagaiH. Nakagawa1983Dependence of memory of meal time upon circadian biological clock in rats.Physiology & Behavior30259265
- 35. Guido ME, de Guido LB, Goguen D, Robertson HA, Rusak B (1999) Daily Rhythm of Spontaneous Immediate-Early Gene Expression in the Rat Suprachiasmatic Nucleus. Journal of Biological Rhythms 14: 275–280.ME GuidoLB de GuidoD. GoguenHA RobertsonB. Rusak1999Daily Rhythm of Spontaneous Immediate-Early Gene Expression in the Rat Suprachiasmatic Nucleus.Journal of Biological Rhythms14275280
- 36. Davidson AJ (2009) Lesion studies targeting food-anticipatory activity. Eur J Neurosci 30: 1658–1664.AJ Davidson2009Lesion studies targeting food-anticipatory activity.Eur J Neurosci3016581664
- 37. Challet E, Mendoza J, Dardente H, Pevet P (2009) Neurogenetics of food anticipation. Eur J Neurosci 30: 1676–1687.E. ChalletJ. MendozaH. DardenteP. Pevet2009Neurogenetics of food anticipation.Eur J Neurosci3016761687
- 38. Guilding C, Hughes A, Brown T, Namvar S, Piggins H (2009) A riot of rhythms: neuronal and glial circadian oscillators in the mediobasal hypothalamus. Molecular Brain 2: 28.C. GuildingA. HughesT. BrownS. NamvarH. Piggins2009A riot of rhythms: neuronal and glial circadian oscillators in the mediobasal hypothalamus.Molecular Brain228
- 39. Escobar C, Martínez-Merlos MT, Angeles-Castellanos M, Miñana MDC, Buijs RM (2007) Unpredictable feeding schedules unmask a system for daily resetting of behavioural and metabolic food entrainment. European Journal of Neuroscience 26: 2804–2814.C. EscobarMT Martínez-MerlosM. Angeles-CastellanosMDC MiñanaRM Buijs2007Unpredictable feeding schedules unmask a system for daily resetting of behavioural and metabolic food entrainment.European Journal of Neuroscience2628042814
- 40. Verwey M, Lam GY, Amir S (2009) Circadian rhythms of PERIOD1 expression in the dorsomedial hypothalamic nucleus in the absence of entrained food-anticipatory activity rhythms in rats. Eur J Neurosci 29: 2217–2222.M. VerweyGY LamS. Amir2009Circadian rhythms of PERIOD1 expression in the dorsomedial hypothalamic nucleus in the absence of entrained food-anticipatory activity rhythms in rats.Eur J Neurosci2922172222
- 41. Verwey M, Amir S (2011) Nucleus-specific effects of meal duration on daily profiles of Period1 and Period2 protein expression in rats housed under restricted feeding. Neuroscience 192: 304–311.M. VerweyS. Amir2011Nucleus-specific effects of meal duration on daily profiles of Period1 and Period2 protein expression in rats housed under restricted feeding.Neuroscience192304311
- 42. Acosta-Galvan G, Yi C-X, van der Vliet J, Jhamandas JH, Panula P, et al. (2011) Interaction between hypothalamic dorsomedial nucleus and the suprachiasmatic nucleus determines intensity of food anticipatory behavior. Proceedings of the National Academy of Sciences 108: 5813–5818.G. Acosta-GalvanC-X YiJ. van der VlietJH JhamandasP. Panula2011Interaction between hypothalamic dorsomedial nucleus and the suprachiasmatic nucleus determines intensity of food anticipatory behavior.Proceedings of the National Academy of Sciences10858135818
- 43. Landry GJ, Kent BA, Patton DF, Jaholkowski M, Marchant EG, et al. (2011) Evidence for Time-of-Day Dependent Effect of Neurotoxic Dorsomedial Hypothalamic Lesions on Food Anticipatory Circadian Rhythms in Rats. PLoS ONE 6: e24187.GJ LandryBA KentDF PattonM. JaholkowskiEG Marchant2011Evidence for Time-of-Day Dependent Effect of Neurotoxic Dorsomedial Hypothalamic Lesions on Food Anticipatory Circadian Rhythms in Rats.PLoS ONE6e24187
- 44. Valassi E, Scacchi M, Cavagnini F (2008) Neuroendocrine control of food intake. Nutr Metab Cardiovasc Dis 18: 158–168.E. ValassiM. ScacchiF. Cavagnini2008Neuroendocrine control of food intake.Nutr Metab Cardiovasc Dis18158168
- 45. Ramos EJ, Meguid MM, Campos AC, Coelho JC (2005) Neuropeptide Y, alpha-melanocyte-stimulating hormone, and monoamines in food intake regulation. Nutrition 21: 269–279.EJ RamosMM MeguidAC CamposJC Coelho2005Neuropeptide Y, alpha-melanocyte-stimulating hormone, and monoamines in food intake regulation.Nutrition21269279
- 46. Sutton GM, Begriche K, Kumar KG, Gimble JM, Perez-Tilve D, et al. (2010) Central nervous system melanocortin-3 receptors are required for synchronizing metabolism during entrainment to restricted feeding during the light cycle. FASEB J 24: 862–872.GM SuttonK. BegricheKG KumarJM GimbleD. Perez-Tilve2010Central nervous system melanocortin-3 receptors are required for synchronizing metabolism during entrainment to restricted feeding during the light cycle.FASEB J24862872
- 47. Sutton GM, Perez-Tilve D, Nogueiras R, Fang J, Kim JK, et al. (2008) The melanocortin-3 receptor is required for entrainment to meal intake. J Neurosci 28: 12946–12955.GM SuttonD. Perez-TilveR. NogueirasJ. FangJK Kim2008The melanocortin-3 receptor is required for entrainment to meal intake.J Neurosci281294612955
- 48. Hahn TM, Breininger JF, Baskin DG, Schwartz MW (1998) Coexpression of Agrp and NPY in fasting-activated hypothalamic neurons. Nat Neurosci 1: 271–272.TM HahnJF BreiningerDG BaskinMW Schwartz1998Coexpression of Agrp and NPY in fasting-activated hypothalamic neurons.Nat Neurosci1271272
- 49. Mistlberger RE (2006) Circadian rhythms: PERturbing a food-entrained clock. Curr Biol 16: R968–969.RE Mistlberger2006Circadian rhythms: PERturbing a food-entrained clock.Curr Biol16R968969
- 50. Moriya T (2009) The dorsomedial hypothalamic nucleus is not necessary for food-anticipatory circadian rhythms of behavior, temperature or clock gene expression in mice. European Journal of Neuroscience 27: 1447–1460.T. Moriya2009The dorsomedial hypothalamic nucleus is not necessary for food-anticipatory circadian rhythms of behavior, temperature or clock gene expression in mice.European Journal of Neuroscience2714471460
- 51. Acosta-Galvan G, Yi C-X, van der Vliet J, Jhamandas JH, Panula P, et al. (2011) Interaction between hypothalamic dorsomedial nucleus and the suprachiasmatic nucleus determines intensity of food anticipatory behavior. G. Acosta-GalvanC-X YiJ. van der VlietJH JhamandasP. Panula2011Interaction between hypothalamic dorsomedial nucleus and the suprachiasmatic nucleus determines intensity of food anticipatory behavior.Proceedings of the National Academy of Sciences. Proceedings of the National Academy of Sciences.
- 52. Sanchez E, Singru PS, Acharya R, Bodria M, Fekete C, et al. (2008) Differential Effects of Refeeding on Melanocortin-Responsive Neurons in the Hypothalamic Paraventricular Nucleus. Endocrinology 149: 4329–4335.E. SanchezPS SingruR. AcharyaM. BodriaC. Fekete2008Differential Effects of Refeeding on Melanocortin-Responsive Neurons in the Hypothalamic Paraventricular Nucleus.Endocrinology14943294335
- 53. Herman JP, Tasker JG, Ziegler DR, Cullinan WE (2002) Local circuit regulation of paraventricular nucleus stress integration: glutamate-GABA connections. Pharmacol Biochem Behav 71: 457–468.JP HermanJG TaskerDR ZieglerWE Cullinan2002Local circuit regulation of paraventricular nucleus stress integration: glutamate-GABA connections.Pharmacol Biochem Behav71457468
- 54. Ziegler DR, Cullinan WE, Herman JP (2002) Distribution of vesicular glutamate transporter mRNA in rat hypothalamus. J Comp Neurol 448: 217–229.DR ZieglerWE CullinanJP Herman2002Distribution of vesicular glutamate transporter mRNA in rat hypothalamus.J Comp Neurol448217229
- 55. Bealer SL (1986) Corticosteroids and plasma restitution after hemorrhage and hypothalamic lesions. Am J Physiol 250: R18–23.SL Bealer1986Corticosteroids and plasma restitution after hemorrhage and hypothalamic lesions.Am J Physiol250R1823
- 56. Boudaba C, Szabo K, Tasker JG (1996) Physiological mapping of local inhibitory inputs to the hypothalamic paraventricular nucleus. J Neurosci 16: 7151–7160.C. BoudabaK. SzaboJG Tasker1996Physiological mapping of local inhibitory inputs to the hypothalamic paraventricular nucleus.J Neurosci1671517160
- 57. Kalsbeek A, Fliers E, Hofman MA, Swaab DF, Buijs RM (2010) Vasopressin and the output of the hypothalamic biological clock. J Neuroendocrinol 22: 362–372.A. KalsbeekE. FliersMA HofmanDF SwaabRM Buijs2010Vasopressin and the output of the hypothalamic biological clock.J Neuroendocrinol22362372
- 58. Dong H-W, Petrovich GD, Watts AG, Swanson LW (2001) Basic organization of projections from the oval and fusiform nuclei of the bed nuclei of the stria terminalis in adult rat brain. The Journal of Comparative Neurology 436: 430–455.H-W DongGD PetrovichAG WattsLW Swanson2001Basic organization of projections from the oval and fusiform nuclei of the bed nuclei of the stria terminalis in adult rat brain.The Journal of Comparative Neurology436430455
- 59. Cullinan W, Ziegler D, Herman J (2008) Functional role of local GABAergic influences on the HPA axis. Brain Structure and Function 213: 63–72.W. CullinanD. ZieglerJ. Herman2008Functional role of local GABAergic influences on the HPA axis.Brain Structure and Function2136372
- 60. Cullinan W, Helmreich DL, Watson SJ (1996) Fos expression in forebrain afferents to the hypothalamic paraventricular nucleus following swim stress. The Journal of Comparative Neurology 368: 88–99.W. CullinanDL HelmreichSJ Watson1996Fos expression in forebrain afferents to the hypothalamic paraventricular nucleus following swim stress.The Journal of Comparative Neurology3688899
- 61. Choi DC, Furay AR, Evanson NK, Ostrander MM, Ulrich-Lai YM, et al. (2006) Subnuclei of the bed nucleus of the stria terminalis differentially regulate PVN activation in response to restraint stress. Frontiers in Neuroendocrinology 27: 51–52.DC ChoiAR FurayNK EvansonMM OstranderYM Ulrich-Lai2006Subnuclei of the bed nucleus of the stria terminalis differentially regulate PVN activation in response to restraint stress.Frontiers in Neuroendocrinology275152
- 62. Faulconbridge LF, Grill HJ, Kaplan JM, Daniels D (2008) Caudal brainstem delivery of ghrelin induces fos expression in the nucleus of the solitary tract, but not in the arcuate or paraventricular nuclei of the hypothalamus. Brain Research 1218: 151–157.LF FaulconbridgeHJ GrillJM KaplanD. Daniels2008Caudal brainstem delivery of ghrelin induces fos expression in the nucleus of the solitary tract, but not in the arcuate or paraventricular nuclei of the hypothalamus.Brain Research1218151157
- 63. Hayes MR, Bradley L, Grill HJ (2009) Endogenous Hindbrain Glucagon-Like Peptide-1 Receptor Activation Contributes to the Control of Food Intake by Mediating Gastric Satiation Signaling. Endocrinology 150: 2654–2659.MR HayesL. BradleyHJ Grill2009Endogenous Hindbrain Glucagon-Like Peptide-1 Receptor Activation Contributes to the Control of Food Intake by Mediating Gastric Satiation Signaling.Endocrinology15026542659
- 64. Skibicka KP, Egecioglu E, Dickson SL (2010) Parabrachial nucleus is a new target of orexigenic action of ghrelin. Endocrine Journal 57: S518–S518.KP SkibickaE. EgeciogluSL Dickson2010Parabrachial nucleus is a new target of orexigenic action of ghrelin.Endocrine Journal57S518S518
- 65. Davidson AJ, Cappendijk SL, Stephan FK (2000) Feeding-entrained circadian rhythms are attenuated by lesions of the parabrachial region in rats. Am J Physiol Regul Integr Comp Physiol 278: R1296–1304.AJ DavidsonSL CappendijkFK Stephan2000Feeding-entrained circadian rhythms are attenuated by lesions of the parabrachial region in rats.Am J Physiol Regul Integr Comp Physiol278R12961304
- 66. Davidson AJ, Stephan FK (1998) Circadian food anticipation persists in capsaicin deafferented rats. J Biol Rhythms 13: 422–429.AJ DavidsonFK Stephan1998Circadian food anticipation persists in capsaicin deafferented rats.J Biol Rhythms13422429
- 67. Williams DL, Baskin DG, Schwartz MW (2009) Hindbrain leptin receptor stimulation enhances the anorexic response to cholecystokinin. American Journal of Physiology - Regulatory, Integrative and Comparative Physiology 297: R1238–R1246.DL WilliamsDG BaskinMW Schwartz2009Hindbrain leptin receptor stimulation enhances the anorexic response to cholecystokinin.American Journal of Physiology - Regulatory, Integrative and Comparative Physiology297R1238R1246
- 68. Johnstone LE, Fong TM, Leng G (2006) Neuronal activation in the hypothalamus and brainstem during feeding in rats. Cell metabolism 4: 313–321.LE JohnstoneTM FongG. Leng2006Neuronal activation in the hypothalamus and brainstem during feeding in rats.Cell metabolism4313321
- 69. LeSauter J, Hoque N, Weintraub M, Pfaff DW, Silver R (2009) Stomach ghrelin-secreting cells as food-entrainable circadian clocks. Proceedings of the National Academy of Sciences of the United States of America 106: 13582–13587.J. LeSauterN. HoqueM. WeintraubDW PfaffR. Silver2009Stomach ghrelin-secreting cells as food-entrainable circadian clocks.Proceedings of the National Academy of Sciences of the United States of America1061358213587
- 70. Davidson AJ, Poole AS, Yamazaki S, Menaker M (2003) Is the food-entrainable circadian oscillator in the digestive system? Genes Brain Behav 2: 32–39.AJ DavidsonAS PooleS. YamazakiM. Menaker2003Is the food-entrainable circadian oscillator in the digestive system?Genes Brain Behav23239
- 71. Poulin A, Timofeeva E (2008) The dynamics of neuronal activation during food anticipation and feeding in the brain of food-entrained rats. Brain Research 1227: 128–141.A. PoulinE. Timofeeva2008The dynamics of neuronal activation during food anticipation and feeding in the brain of food-entrained rats.Brain Research1227128141
- 72. Tsurumi C, Ishida N, Tamura T, Kakizuka A, Nishida E, et al. (1995) Degradation of c-Fos by the 26S proteasome is accelerated by c-Jun and multiple protein kinases. Molecular and Cellular Biology 15: 5682–5687.C. TsurumiN. IshidaT. TamuraA. KakizukaE. Nishida1995Degradation of c-Fos by the 26S proteasome is accelerated by c-Jun and multiple protein kinases.Molecular and Cellular Biology1556825687
- 73. Block M, Zucker I (1976) Circadian rhythms of rat locomotor activity after lesions of the midbrain raphe nuclei. J Comp Physiol 109: 235–247.M. BlockI. Zucker1976Circadian rhythms of rat locomotor activity after lesions of the midbrain raphe nuclei.J Comp Physiol109235247
- 74. Glass JD, Grossman GH, Farnbauch L, DiNardo L (2003) Midbrain raphe modulation of nonphotic circadian clock resetting and 5-HT release in the mammalian suprachiasmatic nucleus. J Neurosci 23: 7451–7460.JD GlassGH GrossmanL. FarnbauchL. DiNardo2003Midbrain raphe modulation of nonphotic circadian clock resetting and 5-HT release in the mammalian suprachiasmatic nucleus.J Neurosci2374517460
- 75. Jacobs BL, Foote SL, Bloom FE (1978) Differential Projections of Neurons within Dorsal Raphe Nucleus of Rat - a Horseradish-Peroxidase (Hrp) Study. Brain Research 147: 149–153.BL JacobsSL FooteFE Bloom1978Differential Projections of Neurons within Dorsal Raphe Nucleus of Rat - a Horseradish-Peroxidase (Hrp) Study.Brain Research147149153
- 76. Moga MM, Moore RY (1997) Organization of neural inputs to the suprachiasmatic nucleus in the rat. J Comp Neurol 389: 508–534.MM MogaRY Moore1997Organization of neural inputs to the suprachiasmatic nucleus in the rat.J Comp Neurol389508534
- 77. Varcoe TJ, Kennaway DJ (2008) Activation of 5-HT2C receptors acutely induces Per1 gene expression in the rat SCN in vitro. Brain Res 1209: 19–28.TJ VarcoeDJ Kennaway2008Activation of 5-HT2C receptors acutely induces Per1 gene expression in the rat SCN in vitro.Brain Res12091928
- 78. Webb IC, Patton DF, Landry GJ, Mistlberger RE (2010) Circadian clock resetting by behavioral arousal: neural correlates in the midbrain raphe nuclei and locus coeruleus. Neuroscience 166: 739–751.IC WebbDF PattonGJ LandryRE Mistlberger2010Circadian clock resetting by behavioral arousal: neural correlates in the midbrain raphe nuclei and locus coeruleus.Neuroscience166739751
- 79. Verwey M, Amir S (2012) Variable Restricted Feeding Disrupts the Daily Oscillations of Period2 Expression in the Limbic Forebrain and Dorsal Striatum in Rats. Journal of Molecular Neuroscience 46: 258–264.M. VerweyS. Amir2012Variable Restricted Feeding Disrupts the Daily Oscillations of Period2 Expression in the Limbic Forebrain and Dorsal Striatum in Rats.Journal of Molecular Neuroscience46258264
- 80. Verwey M, Khoja Z, Stewart J, Amir S (2008) Region-specific modulation of PER2 expression in the limbic forebrain and hypothalamus by nighttime restricted feeding in rats. Neurosci Lett 440: 54–58.M. VerweyZ. KhojaJ. StewartS. Amir2008Region-specific modulation of PER2 expression in the limbic forebrain and hypothalamus by nighttime restricted feeding in rats.Neurosci Lett4405458
- 81. Waddington Lamont E, Harbour VL, Barry-Shaw J, Renteria Diaz L, Robinson B, et al. (2007) Restricted access to food, but not sucrose, saccharine, or salt, synchronizes the expression of Period2 protein in the limbic forebrain. Neuroscience 144: 402–411.E. Waddington LamontVL HarbourJ. Barry-ShawL. Renteria DiazB. Robinson2007Restricted access to food, but not sucrose, saccharine, or salt, synchronizes the expression of Period2 protein in the limbic forebrain.Neuroscience144402411
- 82. Mistlberger RE, Mumby DG (1992) The limbic system and food-anticipatory circadian rhythms in the rat: ablation and dopamine blocking studies. Behav Brain Res 47: 159–168.RE MistlbergerDG Mumby1992The limbic system and food-anticipatory circadian rhythms in the rat: ablation and dopamine blocking studies.Behav Brain Res47159168
- 83. Hood S, Cassidy P, Cossette M-P, Weigl Y, Verwey M, et al. (2010) Endogenous Dopamine Regulates the Rhythm of Expression of the Clock Protein PER2 in the Rat Dorsal Striatum via Daily Activation of D2 Dopamine Receptors. J Neurosci 30: 14046–14058.S. HoodP. CassidyM-P CossetteY. WeiglM. Verwey2010Endogenous Dopamine Regulates the Rhythm of Expression of the Clock Protein PER2 in the Rat Dorsal Striatum via Daily Activation of D2 Dopamine Receptors.J Neurosci301404614058
- 84. Blum ID, Patterson Z, Khazall R, Lamont EW, Sleeman MW, et al. (2009) Reduced anticipatory locomotor responses to scheduled meals in ghrelin receptor deficient mice. Neuroscience 164: 351–359.ID BlumZ. PattersonR. KhazallEW LamontMW Sleeman2009Reduced anticipatory locomotor responses to scheduled meals in ghrelin receptor deficient mice.Neuroscience164351359
- 85. Escobar C, Diaz-Munoz M, Encinas F, Aguilar-Roblero R (1998) Persistence of metabolic rhythmicity during fasting and its entrainment by restricted feeding schedules in rats. Am J Physiol 274: R1309–1316.C. EscobarM. Diaz-MunozF. EncinasR. Aguilar-Roblero1998Persistence of metabolic rhythmicity during fasting and its entrainment by restricted feeding schedules in rats.Am J Physiol274R13091316
- 86. Landry GJ, Simon MM, Webb IC, Mistlberger RE (2006) Persistence of a behavioral food-anticipatory circadian rhythm following dorsomedial hypothalamic ablation in rats. Am J Physiol Regul Integr Comp Physiol 290: R1527–1534.GJ LandryMM SimonIC WebbRE Mistlberger2006Persistence of a behavioral food-anticipatory circadian rhythm following dorsomedial hypothalamic ablation in rats.Am J Physiol Regul Integr Comp Physiol290R15271534
- 87. Davidson AJ (2006) Search for the feeding-entrainable circadian oscillator: a complex proposition. American Journal of Physiology - Regulatory, Integrative and Comparative Physiology 290: R1524–R1526.AJ Davidson2006Search for the feeding-entrainable circadian oscillator: a complex proposition.American Journal of Physiology - Regulatory, Integrative and Comparative Physiology290R1524R1526