Introduction:

The mammalian circadian timekeeping system (CTS) coordinates the physiological adaptation of an organism to the daily rhythms of the external world. Any alteration of this internal temporal order leads to pathophysiology induced by circadian dysfunction (reviewed in Salgado-Delgado et al. 2011). The light–dark cycle is the major cue that entrains the biological clock (Jagota 2006); in addition, a number of non-photic cues like food availability cycles (Damiola et al. 2000), socio-professional routines (Lévi et al. 2010) etc. have been shown to exhibit the ability to entrain the circadian clock of an organism. The principal circadian clock resides in cells grouped in two suprachiasmatic nuclei (SCN), just above the optic chiasm in the anterior hypothalamus (Welsh et al. 2010; Jagota 2012). Rhythmic expression of mammalian clock genes are found in SCN as well as peripheral tissues indicating existence of both central and peripheral oscillators (Partch et al. 2013). The nutritional status, adiposity, glucose homeostasis, sympathetic and parasympathetic balance of the body is some of the key inputs in the regulation of circadian clock (Laposky 2008; Bechtold and Loudon 2013). When nocturnal animals are fed under restricted daytime feeding regimen, it misaligns the sleep-wake and fasting-feeding cycles of the organism to the light–dark cycle showing a shift in the activity to day time with a diurnal food anticipatory activity (FAA) which has been conceptualized by many workers as circadian rhythm driven by food entrained circadian oscillator (FEO) separate from light entrained SCN (Caba et al. 2008; Boulos and Terman 1980).

The central pacemaker has been suggested to entrain the organism to the light–dark cycle whereas the peripheral clocks respond to the feeding fasting cycles leading to an internal desynchrony (Kalsbeek et al. 2011). Circadian desynchrony promotes metabolic disruption (Barclay et al. 2012; Yoon et al. 2012). Leptin, a product of the obese (ob) gene is secreted into blood stream primarily by white adipose tissue which acts on several hypothalamic sites including dorsal medial hypothalamic (DMH) and median preoptic area (MPOA) (Zhang et al. 2011; Zhao et al. 2013). It is also a key controller of food intake and energy homeostasis as it acts as a bridge between energy metabolism and circadian clock (Froy 2010). Leptin levels decrease during fasting (Boden et al. 1996) and rise after food intake (Dallongeville et al. 1998). Leptin receptors (Ob-Rs) exist in multiple forms (Ob-Ra_f) (reviewed in Martin et al. 2008) however leptin receptor Ob-Rb present in hypothalamus (Mercer et al. 1996), was reported to be mediating the satiety effects of leptin (Lee et al. 1996).

In addition to regulating the photic cycles, SCN regulates different hypothalamic regions involved in energy homeostasis, which includes medial preoptic area (MPOA) (Choi et al. 1998). MPOA regulates the adipose tissue activity and leptin production (Froy 2010). MPOA is located in the periventricular regions of the anterior hypothalamus covering the organum vasculosum of the lamina terminalis (OVLT). OVLT has been reported to be connected to MPOA (Polston and Simerly 2006). It has been shown earlier that leptin can phase advance the SCN circadian clock in a dose-dependent manner in isolated in vitro brain slices (Prosser and Bergeron 2003).

Aging is the progressive deterioration in the behavioral, biochemical and physiological functions of an organism (Jagota 2005; Rattan 2012). With aging the circadian clock properties and functions are altered leading to the desynchronization of rhythms (Doi et al. 2011; Yu and Weaver 2011). We reported earlier age induced alterations in daily rhythms of serotonin (5-HT) levels in brain as well as SCN starting at middle age (Jagota and Kalyani 2008, 2010). In addition we have reported age related loss of sensitivity to melatonin in restoration of serotonin (Jagota and Kalyani 2010), lipid peroxidation and antioxidant enzymes (Manikonda and Jagota 2012).

Thus, in order to understand age induced alteration in the food entrained regulation of biological clock, we studied the effect of timed restricted feeding (RF) on food intake, body weight, gross locomotor activity rhythms and daily leptin-ir rhythms in SCN and MPOA region of male Wistar rats in various age groups.

Materials and methods

Male Wistar rats of three age groups (1–3) [3, 12 and 24 months (m)] (with n = 44 in each age group) were maintained in light–dark conditions (LD 12:12), lights on: 06:30 AM i.e. Zeitgeber time (ZT-0) and lights off: 6:30 PM (ZT-12) at 20 ± 2° C with relative humidity of 55 ± 6 %, for 2 weeks prior to experiment (Fig. 1). All rats were kept individually in polypropylene cages contained within well ventilated light proof environmental cabinets isolated in animal facility. During light phase animals were exposed to 300 lux, automatically controlled by a 24 h timer. Dim red light was used for handling the animals in the dark. Cage changing was done at random intervals (Jagota and Reddy 2007).

Fig. 1
figure 1

Diagrammatic representation of the various age groups of animals used in the study. Three age groups (1–3) 3, 12 and 24 months (m) were further divided into two groups: A, ad libitum fed (AL) with food and water provided ad libitum for 2 weeks; B, restricted feeding (RF) (received food only during the light period zeitgeber times (ZT) 0 to ZT12

Full size image

Feeding schedules, food and body weight measurement

The rats in three age groups (1–3) were separated further into two groups—(A) ad libitum fed (AL) with food and water provided ad libitum for 2 weeks (n = 24) (Group 1A, 2A and 3A) (B) restricted feeding (RF) (received food only during the light period, i.e., from 6:30 AM (ZT-0) to 6:30 PM (ZT-12) for 3 weeks) (Bodosi et al. 2004) (n = 20 each) (Group 1B, 2B and 3B). All animals were fed with standard rat diet. AL group with n = 24 was further divided into three groups of (i) n = 4, (ii) n = 4, (iii) n = 16. RF group was further divided into two groups (i) n = 4 and (ii) n = 16. Food intake was recorded every 6 h interval for both AL [1A (i), 2A (i), and 3A (i)] and RF [1B (i), 2B (i) and 3B (i)] groups by taking weight of the unconsumed food. In addition body weights were also recorded every third day till the end of the experiment for these groups. All experiments were performed as per Institutional Animal Ethics.

Gross locomotor activity

Animals of three age groups in AL group (1A (ii), 2A (ii) and 3A (iii) (n = 4 in each group) were housed individually in cages equipped with infra-red (I.R) motion detector sensors (IR 28 kit, INT, India). The gross locomotor activity was recorded using Chronobiology Kit (Stanford Software Systems, USA) (Schumann et al. 2005; Mammen and Jagota 2011) for 10 days. This was followed by RF and the gross locomotor activity was recorded for three weeks.

Data analysis

Actograms and mean activity profiles were prepared using the Chronobiology Kit and analyzed using the Kit Analyze software (Stanford Software systems, USA). The duration of the active period (α), circadian period (τ) and percentage diurnality were calculated (Schumann et al. 2005; Mammen and Jagota 2011). α started with activity onset and ended with activity offset. Results were represented as mean ± standard error (SE) (Table 1).

Table 1 Effect of restricted feeding on various circadian rhythm parameters
Full size table

Leptin—immunohistochemistry

For studying daily leptin rhythms, animals 1A (iii), 2A (iii), and 3A (iii) and 1B (iii), 2B (iii), 3B (iii) with n = 16 in each group were sacrificed after anesthetizing and transcardial perfusing with heparinized 0.9 % saline at variable time points (ZT-0, 6, 12 and 18). The brains were removed rapidly and post-fixed immediately in 4 % paraformaldehyde in 0.1 M phosphate buffer (pH 7.4) for 2–4 h. Brains were then transferred into a gradient of sucrose solutions (20 and 30 % sucrose prepared with 0.1 M phosphate buffer (pH 7.4)) subsequently and stored at 4° C till it sinks to the bottom in both the solutions. 20 μm sections from these brains were taken using cryostat (Leica CM 1850). The sections were transferred into phosphate buffer saline (PBS) and left for 1 h at room temperature to remove sucrose. These sections were then incubated in 3 % H2O2 for 30 min followed by blocking using normal goat serum. The sections were incubated in primary antibody (dilution 1:350) (Rabbit polyclonal IgG primary antibody [Ob (Y-20) Sc-843, Santa Cruz Biotechnology, Inc., Santa Cruz, CA] at 37° C for 3 h and then with corresponding secondary antibody (dilution 1:450) (Goat anti-rabbit IgG-HRP: HP03, Bangalore Genei, India) at 37° C for 4 h. Finally the sections were incubated in substrate solution containing 0.1 % DAB (Diamino Benzedene) and 3 % H2O2 in PBS for about 10 min at 37° C till color develops. The reaction was stopped using 2 M HCl for 30 s and washed immediately with PBS. Sections were dehydrated using absolute alcohol and cleared by methyl benzoate and mounted using D.P.X. All sections were then arranged in a rostro-caudal axis, photographed using DP-12 digital camera attached to Olympus microscope (BX-41). Leptin-ir levels were compared using densitometric analysis with Image Pro AMS software (Media Cybernetics, USA) (Mammen and Jagota 2011).

Statistical analysis

Statistical Analysis Data was analyzed using Jandel Scientific Sigma stat software by Student’s t test and one way ANOVA followed by Post hoc Duncan’s test for multiple comparisons.

Results

The food intake increased with age as observed in various AL age groups. Food intake was highest in 24 m old rats (37.76 ± 0.06 g per day) but significantly less in 3 m old rats (28.86 ± 0.25 g per day). Food intake showed two peaks at ZT-6 and 18 respectively in 24 m AL rats. In RF animals, the food intake was more between ZT-0 and ZT-6, i.e. immediately after the food was provided. In both 3 and 12 m RF rats the food intake significantly decreased compared to respective AL group; however in 24 m RF rats food intake significantly increased to 38.56 ± 0.01 g when compared to AL group (Fig. 2a, b). Body weights of 3 and 12 m old RF rats significantly decreased under RF when compared to their respective AL group by approximately 12 and 5.5 % respectively. Whereas, the body weights of 24 m old RF rats significantly increased approximately by 3 % compared to their AL group (Fig. 2c).

Fig. 2
figure 2

Food intake profile in ad libitum (AL) fed and food restricted (RF) male Wistar rats in variable age groups: 3, 12 and 24 months (m) (n = 4 in each age group) at various zeitgeber times (ZT) 0, 6, 12 and 18. a Food intake; b mean food intake over 24 h; c body weights (measured every third day for 21 days). *, *1, and *2 indicate significant difference (p ≤ 0.05) in comparison to AL in the same age group, in 3 and 12 m in the same experimental group respectively

Full size image

Studies of gross locomotor activity rhythms

The locomotory activity profile of all the 3, 12 and 24 m old AL animals was restricted largely to the night time with 82.49 ± 0.27, 92.46 ± 0.18 and 88.88 ± 0.56 % nocturnality, respectively (Fig. 3; Table 1). The α of all the three age groups of AL animals was approximately 12 h. Interestingly all the age groups of animals have shown onset of bouts of activity mostly after the food was provided in RF condition. In RF 3 m old animals, the activity increased during the early feeding periods ZT 0-6, but declined after ZT-6 time point (mid-day) and then the onset of activity started ~3–4 h before the normal onset which starts at ZT-12. From ZT-12 onwards the activity in RF animals was similar with decreased robustness compared to AL animals. In RF 12 m old rats the activity was observed throughout the day and night. In RF 24 m animals the activity profile was distributed more during the day (ZT-0 to 12) than in night (ZT-18 to 0). Under RF, the % nocturnality decreased to 68.53 ± 0.21, 49.42 ± 0.22, and 18.42 ± 0.39 in 3, 12 and 24 m old rats respectively and % diurnality increased significantly (p ≤ 0.05) from 17.48 ± 0.24, 7.92 ± 0.26 and 11.44 ± 0.17 to 31.8 ± 0.31, 29.16 ± 0.13 and 72.75 ± 0.21 in 3, 12 and 24 m old rats, respectively. In RF 12 m old rats, α increased significantly (p ≤ 0.05) (18.09 ± 0.45 h) when compared to RF 3 m (15.37 ± 0.62 h) but in RF 24 m old rats, α decreased significantly compared to 3 and 12 m old rats (p ≤ 0.05) (13.03 ± 0.3 h). In 24 m old rats the activity was shifted more towards the light phase and the night time activity decreased. The τ was approximately 24 h in 3, 12 and 24 m AL rats, i.e. 24.16 ± 0.16, 23.97 ± 0.22 and 23.95 ± 0.02 h, respectively. But in RF rats the τ significantly (p ≤ 0.05) increased from 23.97 ± 0.22 to 24.72 ± 0.06 h in 12 m old rats whereas in 24 m old rats τ significantly (p ≤ 0.05) decreased from 23.95 ± 0.02 to 23.55 ± 0.1 h however no significant change was observed in the τ between AL and RF in 3 m old rats.

Fig. 3
figure 3

A Double plotted actograms of gross locomotor activity in ad libitum (AL) and food restricted (RF) condition: in 3 (Ai), 12 (Aii) and 24 (Aiii) m male Wistar rats. B Activity profiles of animals in 3 (Bi), 12 (Bii) and 24 (Biii) m (n = 4) under AL and RF conditions. Dashed lines on activity profile indicates lights on and off for animals in both groups. Additionally black and white bar at the bottom of mean profiles indicates food availability and non availability period for RF rats

Full size image

Studies of daily leptin rhythms

During the preliminary screening, all the coronal sections of SCN were screened for the leptin immunoreactivity (-ir), but leptin-ir could not be detected in this region (Fig. 4A), however when scanned serially from SCN towards optic chiasm, leptin-ir was detected in OVLT region located in the anteroventral region of the third ventricle (AV3V) in all the three age groups, i.e. AL rats (Fig. 4Bi, Ci, Di).

Fig. 4
figure 4

Leptin-ir in the coronal brain sections (25 μm) of 3 m old male Wistar rats killed under various zeitgeber times (ZT) 0, 6, 12 and 18: A no leptin-ir was observed in the SCN region (scale bar = 1 mm); B, C, D Leptin-ir in the organovasculosum of the lamina terminalis (OVLT) of 3, 12 and 24 m old ad libitum (AL) fed rats (Bi, Ci, Di); food restricted (RF) rats (Bii, Cii, Dii) (scale bar = 100 μm) and their densitometric analysis using Image Pro AMS software (n = 4 at each time point) (Biii, Ciii, Diii); E mean 24 h leptin-ir p a ≤ 0.05; p b ≤ 0.05; p c ≤ 0.05; p d ≤ 0.05 (a, b, c and d refers to comparison with ZT-0, 6, 12 and 18); p w ≤ 0.05 (w refers to comparison at same time point in AL and RF in same age group). *, *1, *2 indicate significant difference (p ≤ 0.05) in comparison to AL in the same age group; 3 and 12 m in the same experimental group

Full size image

The leptin-ir in OVLT region was found to be highest in 3 m AL animals when compared to 12 and 24 m AL animals (Fig. 4Biii, Ciii, Diii). The leptin-ir showed daily rhythm pattern both in 3 and 12 m though there was decrease in robustness in later. Interestingly leptin-ir was maximum at ZT-0 in all the groups studied but was minimum at ZT-12 in 3 and 12 m though at ZT-18 in 24 m. In 24 m old rats the leptin-ir further decreased in comparison to 3 and 12 m with a significant change in the pattern of the rhythm with two peaks at ZT-0 and ZT-12 though there was no significant difference in leptin-ir levels between 12 and 24 m old rats.

The variable RF age groups showed differential leptin-ir daily rhythm profile in OVLT region. RF in 3 m old animals resulted in decrease in mean leptin-ir though rhythmicity persisted but the maximum levels appeared delayed by 6 h at ZT-6 as compared to ZT-0 in AL. In 12 m, however there was significant decrease in leptin-ir maximum though minimum and mean (24 h) levels did not show significant change in OVLT with RF. In 24 m RF, the rhythmicity in leptin-ir persisted but there was significant decrease in mean levels with the maximum levels at ZT-0 and minimum at ZT-12 (Fig. 4E).

Discussion

The amount of food intake increased from 3 to 24 m in AL rats. The amount of food intake and body weights significantly decreased in 3 and 12 m RF rats when compared to AL groups. In 24 m AL rats, two peaks of food intake appeared at ZT-6 and 18 which could be comparable to anticipation of two daily meals and are related with two oscillator structure (Jagota et al. 2000) described by Mistlberger et al. 2012. In 24 m RF rats the amount of food intake and the body weights were significantly high compared to 24 m AL group. Interestingly some researchers have reported increased food intake and body weights of 9 weeks old mice fed under high fat diet during the light phase for 6 weeks demonstrating role of food on such parameters (Arble et al. 2009).

Locomotor activity in the present study is synchronized to feeding time in RF rats. We demonstrate here that altered meal timing switch the animals towards diurnality with a phase shift in the locomotor activity towards the day differentially with age. FAA was observed 3–4 h before the normal onset of activity in 3 m AL animals. In all the three RF age groups of animals the % diurnality increased and the % nocturnality decreased. The FAA in nocturnal rodents has been related to availability of food restricted to few hours in daytime (Mistlberger et al. 2012). This is in correlation with the earlier studies showing food deprivation and caloric restriction results in the nocturnal animals becoming partially diurnal (Challet 2010). Such synchronization of daily rhythms of locomotor activity to the feeding time was also shown previously in Glithead seabream fish (Sparus aurata) (Montoya et al. 2010) and Wistar rats (Carneiro and Araujo 2011). In 24 m old RF animals the gross locomotor activity decreased significantly and became arrhythmic. This is similar to the finding of some researchers for Mongolian gerbils when fed during their inactive phase under LD14:10 conditions (Karakas et al. 2006). In addition, in mice, switching from nocturnal to diurnal with availability of food only during the day time has been reported (Chabot et al. 2012). We demonstrate here that such food entrained switching from % nocturnality to % diurnality is quite robust with 6.5 folds increase in 24 m RF rats.

The timing of nursing and FAA in rabbit pups has been reported to differentially affect Per1 expression in SCN and DMH (Caba et al. 2008). Some researchers have related nocturnal to diurnal switch change in input pathway to SCN (Doyle et al. 2008). In the present study also there appears to be change from photic to non-photic entrainment with FEO. Interestingly leptin has been reported to stimulate DMH neurons (Lee et al. 2013).

The leptin-ir levels in SCN, MPOA and OVLT were studied but, to our surprise we could not find any leptin-ir in the SCN and MPOA region. However, we scanned all the coronal sections serially for any leptin-ir up to the end of the optic chiasm and interestingly we localized leptin-ir in OVLT region.

The mean 24 h leptin-ir levels in OVLT region decreased with age in AL groups as highest levels were observed in 3 m old rats. The peak leptin-ir levels in OVLT region of all the three AL age groups were observed at ZT-0 representing the beginning of the resting phase of these nocturnal animals, also a post feeding phase for these rats. Plasma leptin in diurnal humans has also been reported to exhibit a strong diurnal rhythm with a peak occurring at near midnight during their resting phase. In addition the plasma leptin levels have been reported to be entrained to meal timings and a shift of 6.5 h of meal timings was shown to result in shift of plasma leptin rhythm by about 5.7 h indicating a rapid phase alteration in the leptin rhythms with shift in the meal timings (Schoeller et al. 1997).

Interestingly, the leptin-ir levels in the OVLT region in 3 and 12 m AL rats increased from ZT-12 to ZT-24 after the onset of feeding peak which starts from ZT-12 onwards (i.e. onset of night) and leptin peak is 12 h after the onset of the feeding or at the offset of feeding (i.e. onset of day), i.e. off set of active period (ZT-0/24). In RF 3 m animals, the leptin-ir levels showed peak 6 h after onset of feeding, i.e. at ZT-6. In 12 and 24 m RF rats the onset of feeding and leptin levels were parallel. Thus, our results demonstrate that 3 m rats show more entrainment towards non-photic cue restricted feeding. In control rats upon aging the mean leptin-ir levels significantly decreased from 3 to 12 m but no significant difference was observed between 12 and 24 m old rats.

In 24 m old AL animals the locomotor activity increased during the natural inactive phase of the animals. This increased day time activity and increased body weight is correlating with the sleep loss human subjects having relatively low serum leptin levels and increased ghrelin and body mass index (BMI) (Taheri et al. 2004). Interestingly RF has also been related with shifts of corticosterone, temperature, activity and brain amine periodicity (Krieger 1974) in addition to decreased plasma insulin (Díaz-Muñoz et al. 2000).

The decreased mean 24 h leptin-ir levels in OVLT in 3 and 24 m old RF rats is in correlation with the decreased serum and hypothalamic leptin levels after 12 h of food deprivation in rats (Vujovic et al. 2011). This decrement in mean 24 h leptin-ir levels is also in correlation with the decreased plasma leptin levels in food deprived sheep under short and long days (Marie 2001). The robust increase in diurnality in aged rats might in addition to decreased leptin levels would have resulted in the increased circadian desynchrony responsible for increased food intake and body weight gain in 24 m old rats.

In the mammalian circadian clock function serotonin is implicated in the photic and non-photic regulation of circadian rhythms (Jiang et al. 2000). Serotonin has been reported to regulate ghrelin levels which influence leptin levels (Takeda et al. 2013). In addition, as serotonin innervation and varicosities into OVLT regulating circulating factors have been reported earlier by some researchers (Bosler and Descarries 1988). We have previously reported progressive decline with age in mean 5-HT and reduced amplitude of daily rhythmicity with disintegration at middle age (Jagota and Kalyani 2008, 2010). Such a decrease in the serotonin levels with age thus could be related to decrease in leptin levels and decrease in amplitude of leptin pulse in 12 and 24 m in OVLT region. Reduced leptin levels have been related by some researchers to obesity due to insatiable appetite leading to increased food intake (Korkmaz et al. 2009). Decrease in leptin as well as circadian misalignment could lead to reduced sleep efficacy (Scheer et al. 2008). Some researchers have reported that monoamines could play an important role in addition to neuropeptide Y (NPY) in mediating many of leptin’s central and neuroendocrine effects (Clark et al. 2006). These alterations in the male rat can be related to changes in middle age female rats such as serotonin uptake, serotonin transporter (SERT) binding sites, maintenance of normal cyclic release of LH, alpha-1 adrenergic receptors, decreased sensitivity of GnRH neurons to VIP input (i.e. decreased c-fos expression) and changes in glucose utilization (Cohen and Wise 1988; Krajnak et al. 2003). These changes have been related to mood, memory and sleep (Krajnak et al. 2003). Smith et al. 2005 have related onset of menopause with the alterations in the biological rhythms and CNS function due to reduced exposure to estradiol.

Interestingly OVLT is also a circumventricular organ (CVO) where blood brain barrier (BBB) is weak (Johnson and Loewy 1990). In addition OVLT has been reported to contain angiotensin II (AGII) and is rich in Gonadotropin-Releasing Hormone (GnRH) cells and fibers and plays important role in regulation of body fluid and cardiovascular function. Leptin has been reported to indirectly regulate GnRH neuronal function (Quennell et al. 2009). The decline in leptin-ir levels in the RF groups is parallel to decrease in the night time activity and increase in the day time activity with aging in the present study. Thus our work will be a step towards understanding the role of RF acting as a non-photic cue in entraining the daily leptin levels and rhythms in the OVLT region as well as the gross locomotor activity rhythms in the aging rats.

We have reported earlier age induced change in circadian alignment in lipid peroxidation and antioxidant enzymes (Manikonda and Jagota 2012) which could be the result of reduced leptin (energy metabolism regulatory hormone) levels and attenuated rhythms with aging as leptin has been reported to play important role in inflammation (Lee et al. 2013). Age induced decrease in leptin levels and daily rhythms in OVLT could be responsible for age induced increase in inflammation.