Chocolate for breakfast prevents circadian desynchrony in experimental models of jet-lag and shift-work

Jet-lag results from a sudden change of the light-dark cycle due to transmeridional traveling, which leads to a misalignment between internal circadian rhythms and the external day-night cycle. The transitory days necessary for re-entrainment are associated with behavioral and physiological discomfort.

To determine whether a piece of chocolate can influence re-entrainment in jet-lag conditions, adult Wistar rats were exposed to a sudden 6 h phase advance (6PA) of their light-dark cycle and received simultaneously 5 grams of chocolate either at the beginning of the new night (CH-N; representing breakfast) or at the beginning of the previous night (CH-P; wrong phase breakfast); they were compared with rats undergoing the jet-lag without receiving chocolate (JL; Fig. 1A). Successful re-entrainment was defined as reaching the new acrophase, thus reaching similar values to the 6 h projected acrophase and maintaining this phase relation of at least 3 consecutive days.

After a sudden 6PA the JL rats required 7 days to reach the new acrophase for general activity and 9 days for core temperature (Fig. 1B–D). Rats with access to chocolate scheduled to the previous night (CH-P) were unable to reach the expected acrophases for activity and temperature (Fig. 1E–G), while rats with access to chocolate in the new night (CH-N) reduced to 4 days the amount of days needed to reach the new acrophase in general activity and reduced to 6 days for core temperature (Fig. 1H–J). The one way ANOVA indicated a significant difference between groups in the number of days required for re-entrainment for general activity (F(2,21) = 59.34; P < 0.0001) and for core temperature (F(2,15) = 38.67: P < 0.001). The posthoc test indicated significant difference between all groups (P < 0.01).

Interestingly, the 5 grams piece of chocolate was ingested within 15 min, and this brief event had similar synchronizing effects as observed in animals exposed to a 12 h scheduled food access coinciding with the new activity phase (Supplementary Fig.). Importantly, as observed in the CH-P rats, 12 h of food access scheduled to the previous night prevented re-synchronization (Supplementary Fig.). 1E–G 1B–D

In order to identify which element of chocolate may have favored re-entrainment, another series of rats exposed to the jet-lag protocol were exposed for five days to scheduled access to one of the individual elements of chocolate. On the sixth day animals were left in constant darkness (DD) and did not receive the corresponding element of chocolate in order to observe the endogenous phase achieved with each component in general activity and core temperature.

On day 6, in which rats were in DD, the control group exhibited a difference of 6.1 h between the current acrophase and the expected new acrophase (Fig. 2A), the chocolate group exhibited a difference of 2.4 h from the new acrophase (Fig. 2B), the cocoa butter had a difference of 4.8 h (Fig. 2C), gelatin of 5 h (Fig. 2D), sugar of 3.8 h (Fig. 2E) and cocoa flavor of 4.7 h (Fig. 2F). The Dunnet’s multiple comparisons test indicated a no significant difference between the current acrophase and the expected acrophase for the groups having access to chocolate and to sugar (P < 0.05).

In view of the fast entrainment observed in the CH-N rats, the effect of chocolate for breakfast on day-night pattern of c-Fos was explored in the SCN of rats exposed to experimental jet-lag (6PA).

Under baseline conditions and before the 6PA, control rats (CNT) exhibited a day-night difference of c-Fos immunoreactivity in the SCN (P < 0.01), characterized by high c-Fos levels early in the light phase (Zeitgeber time 1, ZT1) and low c-Fos levels in the early night (ZT13). One day after the 6PA, this day-night c-Fos difference was lost in the dorsal region of the SCN for the JL and the CH-P groups (Fig. 3A,B). Contrasting, chocolate intake coinciding with the onset of the new night (CH-N group) induced a day-night rhythm in c-Fos expression in the dorsal SCN (P < 0.01) as observed in the CNT rats. The two-way ANOVA indicated significant interaction for time of the day x chocolate treatment (F_(3,38) = 7.991; P = 0.0003). In all groups, the ventral SCN responded directly to the light onset with high c-Fos levels in the early day and no differential effect was observed associated with the chocolate timing (Fig. 3C). The two-way ANOVA indicated significant interaction for time of the day x chocolate treatment (F_(3,38) = 3.701; P = 0.0019).

Shift-work produces a chronic condition of internal desynchronization because individuals are forced to be active when their biological clock dictates rest^14,16. In a shift-work model, based on forced activity during the rest phase, rats develop circadian disruption associated with metabolic dysfunction and overweight¹⁷. Restricting food intake to the night, prevented internal desynchronization¹⁰. This same shift-work model, based on forced activity in the rest-phase, was used to test whether a daily piece of chocolate for breakfast, prevented circadian disruption (Fig. 4A).

After 4 weeks of experimental shift-work (WRK), rats developed internal desynchrony characterized by the loss of rhythmicity in circulating glucose and in body temperature and a shifted daily peak of triglycerides (TG, Fig. 4B–D left column), while the daily pattern of activity and melatonin were not shifted (Fig. 4E and F left column). A daily 5 g piece of chocolate scheduled to the onset of the nocturnal activity (breakfast, WCH-B) restored the daily peak of glucose, of TG and of core temperature, while a piece of chocolate at the beginning of the rest phase (dinner, WCH-D) maintained the disrupted circadian pattern observed in the WRK group. The two-way ANOVA for RM indicated significant effects for the interaction time x chocolate treatment for glucose (F_(21,126) = 3.91; P < 0.0001), for temperature (F_(72,864) = 2.72; P < 0.0001), for TG (F_(21,126) = 7.63; P = 0.0.0001), for general activity (F_(72,480) = 4.16; P < 0.0001) and melatonin (F_(24,144) = 5.375; P < 0.0001).

The shift-work protocol alone or combined with chocolate scheduled for dinner lead to an arrhythmic pattern and loss of the acrophases for glucose and body temperature (Fig. 4B,D middle column). The beneficial effect of chocolate for breakfast in the shift-work group (WCH-B) was also observed in the acrophases that were restored to the night phase to similar values as the control rats (represented by the vertical dotted line, Fig. 4 middle column). The one way ANOVA indicated a significant difference among groups for glucose (F_(3,16) = 358.3; P < 0.0001), for TG (F_(3,16) = 45.21; P < 0.0001), for body temperature (F_(3,34) = 5261; P < 0.0001), for general activity (F_(3,20) = 9.976; P = 0.0003), but no effects on melatonin (F_(3,18) = 0.42; P = 0.73). Also, chocolate for breakfast increased the amplitude to similar values as the control group for glucose and body temperature, while activity was decreased and melatonin was not affected (Fig. 4B–F right column). Importantly, in the shift-work model chocolate for dinner (WCH-D) affected daily rhythms in a similar way as the work schedule alone (WRK).

In non-worker rats, with no risk of circadian disruption, chocolate neither for breakfast nor for dinner influenced acrophases of daily rhythms. However, CH-B enhanced the nocturnal peak of glucose and TG (Supplementary Fig. 2B,C left panel). Both chocolate schedules reduced the amplitude of general activity (Supplementary Fig.). 2

The prevention of circadian disruption observed in WCH-B rats suggested direct effects on the SCN. Therefore, the day-night pattern of c-Fos was evaluated in the SCN of rats exposed to the shift-work protocol without chocolate and with access to chocolate either for breakfast or for dinner (Fig. 5).

After 4 weeks in the forced activity protocol the brains of the WRK and WCH-D rats exhibited a significant decreased activation in the ventral and dorsal regions of the SCN during the rest phase as compared with the CNT group (Fig. 5B,C). Chocolate for dinner induced a loss of rhythm in both SCN regions of the WCH-D rats. Contrasting, chocolate for breakfast induced in the WCH-B group a clear day-night pattern in the number of c-Fos positive cells in both regions of the SCN as observed in the CTL group (Fig. 5B,C). The two-way ANOVA indicated significant interaction for time of the day X chocolate treatment for the dorsal SCN (F_(3,39) = 12.83; P < 0.0001) and for the ventral SCN (F_(3,39) = 73.74; P < 0.0001).

In non-working rats, chocolate for dinner (CH-D) also induced a loss of day-night c-Fos rhythms in the dorsal SCN with low values during the rest-phase as compared to CNT rats. Rats receiving chocolate for breakfast (CH-B) exhibited similar day-night patterns in the ventral SCN as the CNT group (Supplementary Fig.). 3

An important consequence of a desynchronized circadian physiology is the disturbance in energy balance². After 3 weeks in the protocol the WRK and WCH-D groups had attained similar body weight gain, while WCH-B rats had gained less weight as compared with all groups (Fig. 6A). The two-way ANOVA indicated significant interaction for weeks of protocol x chocolate schedule (F_(9,108) = 2.13; P = 0.032). At the end of the protocol WRK and WCH-D groups has reached 7% and 4% more body weight gain than the CNT group, respectively (Fig. 6B). Importantly rats that received chocolate for breakfast had gained 17% less body weight than the CNT group and 24% less than the WRK group (Fig. 6B). The one-way ANOVA indicated significant difference between groups (F_(3,36) = 4.961; P = 0.005).

A similar effect was observed in control, non-worker rats. In non-worker rats chocolate for dinner significantly increased body weight gain to 16% more than the CNT group, while rats eating CH for breakfast attained a low body weight (Supplementary Fig.). 4A,B

In order to determine whether the presentation of chocolate at different phases induced a differential response in energy expenditure, the thermogenic effect of chocolate during the last week of the protocol was evaluated. The body temperature in the 30 min prior chocolate presentation was compared with the body temperature for the following 5 hours after chocolate intake.

The WRK and WCH-D groups exhibited a similar posprandial body temperature response, in spite that the WCH-D group had ingested chocolate (Fig. 6C), suggesting no posprandial response to chocolate during the rest phase. The two-way ANOVA indicated no significant effect for groups X time (F_(10,180) = 0.46; P = 0.908). Contrasting, in the active phase chocolate ingestion resulted in an increased thermogenic response of the WCH-B group as compared with the WRK group (Fig. 6D; F_(10,180) = 4.073; P < 0.001). Likewise, in non-worker rats CH-B induced postprandial thermogenesis while CH-D did not (Supplementary Fig. 4C,D).

The circadian system is a complex network of interacting organs and brain areas that are coordinated by the SCN¹⁸. Recent studies show that not only the light-dark cycle, but also other external and internal cues provide temporality to the circadian system^19,20. Specifically, feeding schedules are powerful time signals for peripheral oscillators involved in metabolic and digestive functions like liver, kidneys, gut and adipose tissue²¹. Also, brain areas involved in energy balance and motivation for food intake are entrained by food elicited signals^14,22,23. Importantly, food does not drive circadian rhythms for all systems and organs, because the signals provided by the light-dark cycle can be predominant for some systems and brain regions including the SCN^24,25. In this study we observed a different speed of re-entrainment between general activity and temperature daily rhythms which confirms the differential response to chocolate elicited signals by the regulatory systems involved in activity expression and temperature balance. Other studies have also reported similar difference in the speed of re-entrainment to feeding schedules in peripheral oscillators^24,26.

Due to this complex interaction between external time signals, the beneficial entrainment effects on circadian function are seen when food coincides with the activity phase^17,27 while adverse effects have been reported when food is provided in the rest phase^28,29. This study provides additional evidence of the relevant impact of the timing of food intake and specifically for palatable food intake as a strong entraing factor in order to keep circadian synchrony.

Different non-photic cues have been used to accelerate re-entrainment in experimental models of jet-lag. As early as 1987, Mrosovsky and Salmon demonstrated a fast re-entrainment to a sudden 12 h shift of the light-dark cycle by exposing hamsters for 3 h to a novel running wheel associated with the onset of the new night³⁰; this was confirmed by Yamanaka et al. (2013) in mice⁷ and by Christian and Harrington (2002) in hamsters³¹. Moreover, exposure to the wheel in the light phase or at the end of the night phase interfered with the re-entrainment efficiency in behavior and in peripheral oscillators^32,33. All together such findings suggest that the fast re-entrainment achieved with the daily restricted chocolate intake may be partly due to the increased arousal and excitement induce by chocolate arrival.

Food elicited signals have also shown to be powerful entraining factor for the circadian system. In recent years experimental and clinical studies have demonstrated that in order to maintain a coordinated circadian function, food intake needs to be synchronized to the light-dark cycle. And that the main effects are due to a direct synchronizing effect on peripheral clocks and brain oscillators^24,34. The time of food intake is now suggested as a chronotherapeutic strategy that can help to ameliorate the burden resulting from circadian disruption due to shift-work and jet-lag^10,11,35. Contrasting, feeding schedules coupled to the rest phase, trigger an internal desynchrony, resulting in the loss of homeostasis, and represent a risk for metabolic health^28,29. Here we show that palatable food scheduled for breakfast is sufficient to maintain circadian synchrony.

Palatable food has powerful reinforcing effects on the brain and behavior. High caloric diets rich in fat, sugar and/or salt are normally highly palatable for humans as well as for experimental animals. Palatability results from taste and from the energy provided by the diet. Animals exposed to palatable diets develop non-homeostatic feeding, driven mainly by hedonic aspects of food intake and by affecting the brain reward system^36,37. Previous studies have reported that a daily piece of chocolate can drive circadian oscillations in brain areas involved in the reward system^38,39 and can have entraining effects on behavior and on the SCN in rats under constant darkness¹⁵. Moreover, palatable food based on a sugar-rich as well as fat-rich food have proven to exert a strong effect as synchronizing factors on the circadian system⁴⁰. In this study elements contained in chocolate, like cocoa butter and flavor were also effective at advancing phase, although not with the efficiency as observed with chocolate. Contrasting sugar combined with the onset of the new night turned to be an effective element for accelerating re-entrainment. This highlights the relevance of palatability as entraining factor for the circadian system. Palatability is signaled by the ventral tegmental area to corticolimbic regions, moreover, a recent study demonstrated that stimulation of direct dopamine projections to the SCN accelerate re-entrainment after a 6 h phase shift⁴¹.

In normal light dark (LD) conditions, scheduled food access does not shift the phase of the SCN^25,42. However, the daily rhythms of vasopressin in the dorsomedial SCN are modulated by feeding schedules⁴³. Other studies indicate that the SCN is inhibited during food anticipation and during fasting as observed with c-Fos^19,44 or electrophysiological recordings^44–46, while the ventral SCN is activated after refeeding as well as by light⁴⁵.

Moreover Gallardo et al. (2012) described c-Fos activation in the SCN in anticipation to a high fat diet and to cheese, providing further evidence of the influence of palatable food on the SCN⁴⁷.

Recent findings indicate that the SCN can also respond to hedonic information via dopaminergic projections from the ventral tegmental area (VTA)⁴¹, which may be a pathway used by palatable food. Here we report that a palatable food scheduled for breakfast can influence the activation in the SCN, as seen with c-Fos, by setting a rhythmic pattern in the dorsomedial region, which is suggested to be the anatomical site for the biological clock. This rhythmic pattern in the dorsal SCN may lead to fast re-entrainment³⁵. Importantly, the beneficial synchronizing effects of palatable food on the SCN and overt circadian rhythms was gated by a time window, where chocolate in the active phase exerted strong entraining effects, while chocolate during the rest phase did not favor re-entrainment.

Chocolate for breakfast exerted beneficial effects on circadian function and the correction of this desynchrony was associated with improved body weight. This is in agreement with a previous study reporting in mice exposed to chronic jet-lag, that fixed 12 h restricted food access prevented overweight and metabolic disturbance. In this study however, the fixed 12 h food access did not synchronize activity rhythms because it was mainly presented out of phase of the LD cycle⁴⁸. In control rats eating chocolate in the rest phase had an adverse effect on body weight, while chocolate for breakfast prevented overweight. Our data agree with⁴⁹ who reported increased body weight in mice exposed to restricted acces to chocolate during the day. Other studies have also reported that in rodents the main meal scheduled during the first half of the active phase provides beneficial effects for body weight and metabolism^9,50. Moreover, restricted food access scheduled to the rest phase can disrupt overt circadian expression, favors obesity and leads to metabolic disorders²⁹.

Finally, present data indicate that the thermogenic effect to a high caloric food (chocolate) can have a differential outcome depending on the phase of ingestion. While a piece of chocolate produced a high postprandial thermogenic response during the active phase, this was not observed in the rest phase. This is in agreement with previous reports indicating that breakfast induces a strong postprandial thermogenesis leading to energy expenditure^51–53. Al together here we provide evidence that eating chocolate at the right phase prevents circadian desynchrony and the correction of the desynchrony may improve metabolic health.

All animals were maintained in a 12:12 h LD cycle [lights on at 07.00 h, Zeitgeber time (ZT0)], constant temperature (22 ± 1 °C) and free access to food (Rodent Laboratory Chow 5001) and water, unless otherwise stated. Experimental procedures used in this study were in strict accordance with the Mexican norms for animal handling, Norma Oficial Mexicana NOM-062-ZOO-1999, which conforms to international guidelines for animal handling, and were approved by the Ethics Committee (063/2016) in the Faculty of Medicine UNAM. Experiments conform to international guidelines on the ethical use of animals; procedures were aimed at minimizing the number of animals used and their suffering.

Rats were randomly assigned to a Control (CNT) or a Shift-work (WRK) condition. These groups were further subdivided in: the control group (CNT) receiving regular chow ad libitum during the entire experiment, the control-chocolate for dinner group (CH-D) receiving daily 5 g of chocolate immediately after lights on (ZT0) and the control-chocolate for breakfast group (CH-B) receiving daily 5 g of chocolate immediately after lights off (ZT12). In a similar way the WRK rats were subdivided in three groups: The WRK group received chow ad libitum during the entire experiment, the work rats receiving chocolate for dinner (WCH-D) received daily 5 g of chocolate 10 minutes before entering to the slow rotating wheel for (ZT2); the work rats receiving chocolate for breakfast (WCH-B) received daily 5 g of chocolate at ZT 12. Rats assigned to the shift-work protocol were placed in slow rotating drums (1 revolution / 3 min) that are normally used for sleep deprivation (33 cm diameter, 633 cm wide) from Monday to Friday for 8 h (from ZT2 to ZT10) during 4 weeks, as previously reported⁵⁴. The scheduled chocolate was only provided from Monday to Friday associated with the shift-work protocol. During weekends, all rats remained undisturbed in their home cages. Rats were weighed before starting the baseline and once / week during the first 3 week of the experimental manipulations. Body weight gain was calculated for this interval and for each group. At the end of the 3rd working week rats underwent surgery to implant a jugular cannula (n = 4–6 per group) in order to obtain blood samples for a 24 h cycle.

Rats were allowed to recover during the weekend and continued their work protocol during week 4. One series of rats was used to evaluate postprandial thermogenesis. At the end of week 4 the change in core temperature was compared between WRK rats and WRK rats receiving chocolate either for breakfast or dinner.

All rats were housed in individual cages (45 × 30 × 35 cm) placed on plates with tilt sensors, in soundproof lockers with controlled lighting conditions. General activity in the home cage was continuously monitored with the tilt sensors. Behavioral events were collected with a digitized system and automatically stored every minute in a PC for further analysis SPAD9 (Data processing system, 1.1.1 version; Omnialva SA. De CV. Mexico City, Mexico) based on MATLAB. Double plotted actograms were obtained for each animal by collecting the sum of activity for 15 min intervals. Daily activity counts were fitted to a nonlinear cosinor analysis for estimating daily individual acrophases (see cosinor analysis below). For each rat, normalized activity counts for the last day of the base line, and for the following days after the 6PA were analyzed individually with the cosinor analysis, to obtain the daily acrophases. In order to evaluate core body temperature, rats underwent a brief surgery to receive intra-abdominal temperature sensors (iButton Sensor- Temperature Logger; Maxim integrated products, USA) as previously described⁵⁵. Briefly, one week before starting baseline, rats were anaesthetized with an intramuscular dose of xylazine (Procin 8 mg/kg) and ketamine (Inoketam 40 mg/kg). Under anesthesia a small incision was performed in the abdominal cavity and the temperature sensor, previously sterilized, was introduced in the peritoneum. Abdominal muscles were sutured with absorbable catgut (000) and skin was sutured with surgical suture (Atramat, International Farmaceutica, SA. de CV. Mexico). Rats were left for 1 week to recover before starting the baseline. Temperature sensors were programmed to store samples every 30 min. Data for the baseline and for the 12 following days after the 6PA were analyzed with the cosinor test to obtain acrophases as described for general activity. The acrophase for the last day of the base line was compared with the acrophases following the 6PA a one way ANOVA for repeated mesurements followed by a Dunnett’s pos thoc test for general activity. For daily temperature profiles data in the shift-work model data are presented as average per hour. Postprandial thermogenesis in response to chocolate was evaluated at the end of week 4, in a normal work day. The postprandial response was compared between WRK rats and rats receiving chocolate (WCH-B and WCH-D). The temperature change was determined every 30 min. by comparing temperature prior chocolate presentation with the temperature of the 5 h after chocolate intake for WCH-B and WCH-D rats.

At the end of week 3 of the work protocol, animals in the shift-work protocol and their controls were anesthetized with an intramuscular dose of xylazine (Procin 8 mg/kg) and ketamine (Inoketam 40 mg/kg) and cannulated in the internal jugular vein with a polyethylene silicon tube (0.025 in. i.d. and 0.047 in. o.d.; Silastic Laboratory tubing; Dow Corning Corp., Midland, MI, USA) filled with heparin (500 U/ml) as anti-coagulant as previously reported⁵⁴. The outer end of the cannula was fixed in the back between both shoulder blades and clotted with a small needle. Rats were allowed to recover during the weekend and on Monday the work protocol was reinitiated. At the end of the 4th working week blood samples were obtained distributed in 2 days (Thursday and Friday) to cover a 24 h cycle with 3 h intervals; 4 samples were obtained one day at ZT0, ZT6, ZT12 and ZT18 and the other 4 samples were obtained the next day (ZT3, ZT9, ZT15, AND ZT21) Blood samples (500 ul) were collected in Eppendorf tubes (1.8 ml) containing a clot-activator gel and were centrifuged at 2500 r.p.m. during 10 min, serum was stored in 100 µl aliquots at −45 °C until assay. Aliquots were processed with colorimetric methods for determination of glucose, triglycerides (TG) and melatonin.

Glucose was measured using a commercial colorimetric kit (GPSL-0507, ELI Tech Clinical Systems). TG were assessed with a commercial kit (TGML-0427, ELI Tech Clinical Systems), melatonin was determined by the ELISA method, with a commercial kit (Enzyme immunoassay for the direct, quantitative determination of melatonin; of IBL International, Germany). A cosinor analysis was used to obtain daily acrophases of the metabolic variables/ individual (see cosinor analysis).

One series of the groups employed for the jet-lag study was used to obtain the brains in two time points. Brains were obtained 1 day after the 6PA either at their new ZT1 (one hour after the lights onset) or new ZT13 (one hour after the light offset). ZT13 also corresponded to one hour after chocolate intake for the CH-N group. For the CH-P group, night sampling was performed at ZT19, one hour after chocolate intake.

One series of rats for the shift-work protocol was used for brain collection and immunohistochemistry. On week 4 of the shift-work protocol, rats were perfused one hour after chocolate intake (ZT1) or 12 h later (ZT13) to obtain a day-night pattern.

Rats were anaesthetized with an overdose of sodium pentobarbital (Sedalphorte 65 mg/ml, Pisa, Mexico) and were perfused transcardially with 250 ml of 0.9% saline followed by 250 ml of fixative 4% paraformaldehyde in phosphate buffer saline (PBS, 0.1 M, pH 7.2).

For the night-time points rats were anesthetized under dim red light and their eyes covered. The time between anesthesia and the start of fixation was of 5–6 min. Brains were removed and post fixed for 24 h in paraformaldehyde 4%. After post fixation they were cryoprotected in 30% sucrose for 3–4 days. Brains were frozen and cut in sections of 40 μm at −18 °C. Sections were collected in four series; one series was processed for c-Fos as previously described³⁶. Free floating sections were incubated in c-Fos antibody raised in rabbit (1:2500; Santa Cruz biotechnology, Santa Cruz, CA, USA) in phosphate buffer 0.1 M, pH 7.2 with 0.9% saline 1% (PBS) 0.25% gelatin (G), and 0.3% Triton X-100 (PBSGT) for 72 h in 4 °C. This was followed by incubation in secondary antibody, goat anti-rabbit (Vector Laboratories, California USA), 1:200 in PBSGT for 2 h at room temperature, followed by incubation in avidin–biotin complex (0.9% avidin and 0.9% biotin solutions; Vector Laboratories, California, USA) in PBSGT for 2 h at room temperature. Between incubations sections were rinsed three times for 10 min in PBS. Tissues were reacted with diaminobenzidine (0.01%), nickel (0.05%) and hydrogen peroxide (35 μl,/ 100 mL 30% H2O2) to obtain a blue color. Tissues were mounted, dehydrated and cover slipped with microscopy Entellan (Merck, Darmstadt, Germany).

One medial section of the SCN per animal was selected, images were digitalized with Leica camera ICC50 HD attached to a Leica DM500. Picture were taken with an objective 20X and analyzed with ImageJ software by determining SCN ventral and dorsal regions bilaterally of the SCN. For counting c-Fos positive cells background was subtracted (50%); threshold was determined, and particle analysis was set for particles of 1.0–2.0 circularity and 20–150 pixels. The counts of the dorsal and the ventral regions were added to determine the total amount of c-Fos in the SCN. The results were graphed as total, dorsal and ventral cFos expressing cells and were compared with a two-way ANOVA for the factors groups and time (day-night).

The cosinor analysis was used to obtain daily acrophases for each individual, mean and standard error of the mean (SEM) were obtained / group. The cosinor analysis was performed with MATLAB version 5.3 using the least square method to fit a sine wave to a time series (in this case to 24 h). The formula used was: Y(t)=M + Acophase (2πt/τ + φ) + e(t); Y = collected data; M = mesor; A = amplitude; φ = acrophase; T = period; e = error at each time. The obtained acrophase/day was compared with the expected acrophase using the Student “t” test (significant values set at p < 0.05). Days when the acrophase was statistically different (indicated with asterisks) from the new expected acrophase were considered transitory cycles, while re-entrainment was achieved when no statistical difference between the mean daily acrophase and the expected new acrophase was indicated.

Graphics and statistical analysis were elaborated using the GraphPad Prism version 6.00 for macOS, GraphPad Software, La Jolla California USA, www.graphpad.com↗.

All data generated and analyzed as a part of this study are included within this article (and its supplementary information files).