By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Please be advised that we experienced an unexpected issue that occurred on Saturday and Sunday January 20th and 21st that caused the site to be down for an extended period of time and affected the ability of users to access content on Wiley Online Library. This issue has now been fully resolved. We apologize for any inconvenience this may have caused and are working to ensure that we can alert you immediately of any unplanned periods of downtime or disruption in the future.
Effects of scheduled exposures to novel environment with a running-wheel were examined on re-entrainment to 8 h shifted light–dark (LD) cycles of mouse circadian rhythms in locomotor activity and clock gene, Per1, expression in the suprachiasmatic nucleus (SCN) and peripheral tissues. Per1 expression was monitored by a bioluminescence reporter introduced into mice. The animals were exposed to the novel environment for 3 h from the shifted dark onset for four cycles and released into constant darkness. In the phase-advance shift, the circadian rhythm in locomotor activity fully re-entrained in the exposed group, whereas it was in transients in the control. On the other hand, the circadian rhythm of Per1 expression in the SCN almost completely re-entrained in both the control and exposed groups. In the skeletal muscle and lung, the circadian rhythm fully re-entrained in the exposed group, whereas the rhythms in the control did not. In the phase-delay shift, the circadian rhythms in locomotor activity and Per1 expression almost completely re-entrained in both groups. These findings indicate that the scheduled exposures to novel environment with a running-wheel differentially accelerate the re-entrainment of the mouse peripheral clocks to 8 h phase-advanced LD cycles.
Circadian rhythms in mammals are regulated by the pacemaker in the suprachiasmatic nucleus (SCN) which entrains to a light–dark (LD) cycle (Rusak & Zucker 1979) and phase-controls the so-called peripheral clocks in a variety of tissues (Yoo et al. 2005). On the other hand, non-photic time cues such as timed wheel-running and exposure to novel environment are known to accelerate re-entrainment of circadian behavioral rhythms to phase-shifted LD cycles (Mrosovsky et al. 1989). Three hour wheel-running at the first dark onset of phase-advanced LD cycle dramatically increased the rate of re-entrainment of circadian wheel-running rhythm in hamsters (Mrosovsky & Salmon 1987). Furthermore, a single pulse of wheel-running or forced treadmill running produced a phase-dependent phase-shift of behavioral rhythm, indicating that wheel-running is a potential zeitgeber for the circadian clocks in several rodent species (Reebs & Mrosovsky 1989; Marchant & Mistlberger 1996). Wheel-running is also known to change the free-running period of circadian locomotor rhythm under constant darkness (DD) in rats (Yamada et al. 1988). Similar effects of non-photic time cues have been demonstrated on the circadian rhythms in humans (Van Reeth et al. 1994; Buxton et al. 1997; Miyazaki et al. 2001; Barger et al. 2004). These findings have been interpreted as the indication of feedback from the behavioral outputs to the circadian pacemaker in the SCN (Mistlberger & Holmes 2000). There are some indications that the clock genes such as Per1 and Per2 in the SCN are suppressed by non-photic stimuli (Maywood et al. 1999; Maywood & Mrosovsky 2001; Yannielli et al. 2002). However, it remains unknown whether or not non-photic time cues affect directly the circadian pacemaker in the SCN.
Yamazaki et al. (2000) demonstrated different rates of re-entrainment of rat circadian rhythms in various tissues following abrupt shifts of LD cycle. The circadian rhythm of Per1 expression in the SCN shifted immediately, whereas those in the liver, lung and skeletal muscle shifted slowly, taking several days for full re-entrainment. Immediate re-entrainment of the SCN circadian pacemaker was confirmed by several studies (Hara et al. 2001; Abe et al. 2002; Nagano et al. 2003). Accordingly, the acceleration of re-entrainment by wheel-running is unlikely due to the accelerated phase-shift of the circadian pacemaker in the SCN, since the pacemaker re-entrained to shifted LD cycles immediately, regardless of the presence or absence of non-photic time cues. Alternative hypotheses for the effect of non-photic time cues could be enhancement of phase-control by the SCN over the peripheral clocks, and direct resetting of the peripheral clocks.
The aim of present study was to determine whether or not scheduled exposures to novel environment with a running-wheel could accelerate the re-entrainment of circadian rhythms in mice to 8 h phase-shifted LD cycles. Here, we demonstrate differential effects of the scheduled exposures to the novel environment on re-entrainment of the circadian rhythms in locomotor activity and in the expression of clock gene, Per1, in peripheral tissues to phase-advanced LD cycle.
Accelerated phase-advance shifts of circadian locomotor rhythm by scheduled exposures to novel environment with a running-wheel
Figure 1 illustrates the experimental protocols in which mice were exposed to novel environment with a running-wheel from the dark onset of the LD cycle which was phase-advanced or phase-delayed by 8 h. Representative activity rhythms in mice subjected to two exposures in the phase-advance protocol were illustrated in Fig. 2, and those in mice subjected to four exposures in the phase-advance (left) and phase-delay (right) protocols in Fig. 3. Table 1 summarizes the phase of activity onset, amount of phase-shift and free-running period of locomotor activity rhythm.
Table 1. Activity onset, phase-shift and free-running period of circadian rhythms in locomotor activity.
In the phase-advance experiment with two exposures, neither the phase of activity onset nor the amount of phase-shift was significantly different between the baseline and shifted LD cycle or between the exposed and control groups (Table 1). In 3 of 10 mice in the exposed group, a large phase-advance shift up to 10 h was observed (Fig. 2). By contrast, in the phase-advance experiment with four exposures, a significant difference was found in the phase of activity onset between the baseline and shifted LD cycle in both the control and exposed groups (Table 1). And, the amount of phase-shift was significantly larger in the exposed group than in the control. In the phase-delay experiment with four exposures, the activity onset was significantly phase-shifted in both groups, and the amount of phase-shift was not different between the control and exposed groups.
The free-running period in DD was not significantly different between the control and exposed groups in these experiments (Table 1). However, a significant difference was detected in the free-running period between the phase-advanced and delayed groups. The mice in the phase-delay shift showed a longer period (23.79 ± 0.07 h; mean ± SD, n = 20) than in the phase-advance (23.46 ± 0.14 h, n = 16) (two-way anova with post-hoc t-test; P < 0.01).
There was no statistically significant correlation between the number of wheel-revolution and the amount of phase-shift. Some mice ran a wheel substantially but did not show an accelerated phase-shift.
Circadian rhythms in Per1 gene expression in the SCN and peripheral tissues in culture
Figures 4 and 5 illustrate representative Per1 expression rhythms in the cultured SCN and peripheral tissues (liver, lung and skeletal muscle) in mice subjected to four exposures in the 8 h phase-advance and phase-delay protocols, respectively. The Per1 expression rhythms were monitored continuously by means of a bioluminescence reporter. Table 2 demonstrates the circadian peaks of bioluminescence rhythms in the first cycle of culture.
Table 2. Circadian peaks of bioluminescence rhythms in the SCN, liver, lung and skeletal muscle
Values are expressed with the mean ± SD in local time. Numbers in the parentheses are the amount of phase-shift in hours (+, phase-advance; −, phase-delay). *P < 0.0167 vs. baseline;†P < 0.0167 vs. control (Bonferroni test).
In the baseline, the circadian peak in the SCN was detected at 12.5 ± 0.9 h (local time), 6.5 h after the time of light-on in the previous LD cycles. The circadian peaks in the peripheral tissues were 21.1 ± 1.4 h in the liver, 17.8 ± 1.3 h in the lung and 21.8 ± 1.2 h in the skeletal muscle, respectively.
In the phase-advance experiment, the circadian rhythms in the SCN and peripheral tissues of the control group differentially responded to the advanced LD cycles. The circadian peaks in the SCN, lung and skeletal muscle were significantly phase-shifted (Bonferroni test; P < 0.0167). On the other hand, the circadian peak in the liver was not phase-shifted (Fig. 6).
In the exposed group, one circadian peak in the skeletal muscle was extremely deviated (Smirnov–Grubbs test; P < 0.05), and excluded from further analyses. The circadian peaks in the SCN, lung and skeletal muscle were significantly phase-advanced (Bonferroni test; P < 0.0167), whereas the peak in the liver was not shifted (Fig. 6). When compared with the control group, the circadian peaks in the lung and skeletal muscle showed significantly larger phase-shifts, the phases of which were not different from the expected peak phases in full re-entrainment (Fig. 6). On the other hand, the phase-shifts in the SCN were not significantly different between the control and exposed groups.
In the phase-delay experiment, the circadian peaks in the SCN and the peripheral tissues in both groups were significantly phase-delayed (Bonferroni test; P < 0.0167) and were not different between the control and exposed groups (Table 2 and Fig. 6).
Scheduled exposures to the novel environment with a running-wheel accelerated the re-entrainment of mouse circadian rhythms to 8 h phase-shifted LD cycles. The effect was differential, depending on the direction of phase-shift and the overt circadian rhythms.
In the case of circadian locomotor rhythm, the re-entrainment to the phase-advanced LD cycle was almost completed by four exposures to the novel environment, whereas it was still on the half way without exposure. In some mice, a large phase-shift up to 10 h was observed even by two exposures. On the other hand, the re-entrainment to the phase-delay shift was completed by four exposures, regardless of whether the mice were exposed to the novel environment or not. The number of cycles needed for re-entrainment to a phase-shifted LD cycle has been reported to depend on the direction of phase-shift and the phase-marker of circadian rhythm (Aschoff 1981). In rats, the phase-shift of locomotor rhythm was slow in the phase-advance direction but rapid in the phase-delay, when the activity onset was used as a phase-marker (Honma et al. 1985; Takamure et al. 1991), which was similar to the present results in mice. The free-running period in DD was different between the phase-advance and delay experiment, which was reported as after-effects of phase-shifts (Pittendrigh & Daan 1976).
In the case of the circadian rhythm in Per1 expression, the rate of phase-shift to the phase-advanced LD cycle was different among tissues (Table 2, Fig. 6). The circadian rhythm in the SCN re-entrained almost completely in four cycles, regardless of whether the mice were exposed to the novel environment or not. Immediate phase-shifts of the SCN circadian pacemaker have been observed in response to a phase-shifted LD cycle (Yamazaki et al. 2000; Abe et al. 2002). In the exposed group, the re-entrainment was also facilitated in the lung and skeletal muscle by the novel environment. In both tissues, the circadian rhythms in the control group were still in transients of full re-entrainment, whereas those in the exposed group were fully re-entrained (Fig. 6). In contrast, the circadian rhythm in the liver did not phase-shift significantly, regardless of the presence or absence of the novel environment. The slow response of liver circadian clock to a phase-shifted LD cycle was also reported previously (Yamazaki et al. 2000; Hara et al. 2001). As a result, the so-called internal desynchronization occurred in the control group between the SCN circadian pacemaker and behavioral rhythms or peripheral clocks. Internal desynchronization is regarded as the indication of uncoupling of different circadian oscillators which have been coupled together (Aschoff & Wever 1981). The immediate phase-shift of the SCN pacemaker may be due to high responsiveness to lights, which is described as a type 0 phase-response curve (PRC), whereas the slow phase-shift of the peripheral clocks could be due to a low responsiveness of the oscillator(s) to lights, which is described as a type 1 PRC (Johnson 1999). Or more likely, the slow phase-shift reflect the slow process of re-establishment of an oscillatory coupling with the SCN pacemaker which is disturbed by a rapid phase-shift of the SCN circadian pacemaker.
The circadian oscillator(s) directly responsible for the behavioral rhythms is unknown. Recently, Nagano et al. (2003) demonstrated two regions in the SCN, where the circadian rhythms in Per1 gene expression responded differentially to 6 h phase-advanced LD cycles. The ventrolateral SCN shifted almost immediately, whereas the dorsomedial SCN shifted slowly. A similar dissociation of phase-shifts was also reported in the Per1, Per2 and Cry1 expression rhythms in the SCN (Reddy et al. 2002). Although the possibility is not excluded that the oscillator(s) directly responsible for the behavioral rhythms is located in the SCN, the behavioral oscillator(s) is more likely located outside the SCN, as suggested by the methamphetamine-induced or feeding-associated behavioral rhythms in the SCN lesioned animals (Stephan 1983; Honma et al. 1987). Recently, several structures in the brain were reported to show the SCN-independent circadian rhythms (Abraham et al. 2005) which could be a candidate of peripheral clock for behavioral rhythms.
The mechanism of the differential acceleration of re-entrainment is not known. Exposures to the novel environment may exert their effects not on the SCN circadian pacemaker but on somewhere in the output pathway from the pacemaker to the site of behavior expression, since the Per1 expression rhythm in the SCN almost completely re-entrained to the phase-shifted LD cycle with or without exposures to the novel environment. Recently, prokineticin 2 (PK2) and brain-derived neurotrophic factor (BDNF) were suggested to be released from the SCN and involved in the expression of behavioral rhythm (Cheng et al. 2002; Naert et al. 2006). Exposures to the novel environment could stimulate the release of these substances in the SCN and thereby strengthen the output signals to which the peripheral clock entrains. Alternatively, exposures to the novel environment may directly stimulate the peripheral clock(s) responsible for behavioral rhythms. This possibility has been suggested by the so-called prefeeding enhancement of wheel-running or other physiological measures under restricted daily feeding (Stephan 2002). The prefeeding peaks were detected even in the SCN lesioned animals (Krieger et al. 1977; Stephan 1983). In addition, the Per1 expression rhythm in the SCN did not response to restricted feeding but those in the peripheral tissues phase-shifted differentially by restricted feeding (Stokkan et al. 2001). The feeding schedule is likely to affect the peripheral clocks directly. The scheduled exposures to the novel environment may directly reset the peripheral clocks in the lung and skeletal muscle similar to restricted daily feeding.
In the control mice subjected to the phase-advanced shift, the peripheral clocks in the lung and skeletal muscle seemed to phase-delay instead of phase-advance, despite that the behavioral rhythm showed phase-advanced shifts (Figs 3 and 6). Such re-entrainment by partition has been observed in humans under temporal isolation (Aschoff & Wever 1981), and in passengers who crossed many time zones rapidly (Takahashi et al. 1999). The mechanism for re-entrainment by partition is not known. One possible explanation for this phenomenon is that the peripheral clocks in different tissues have different sensitivities to the time cue and show different PRC for it. Even if the same time cue simultaneously hits the circadian oscillators, the direction of response would be different depending on the shape of each PRC.
The mediator of time cue(s) in the novel environment is not known. A mouse sniffs, walks around and runs a wheel in a new cage. The number of wheel-revolution in individual animals did not correlate with the amount of phase-shift with a statistically significant level. Some mice ran a wheel substantially but did not show an accelerated phase-shift. Therefore, it is not clear whether or not the wheel-running is responsible for the acceleration of phase-shift. The endocrine systems such as the hypothalamic–pituitary–adrenocortical axis are known to be stimulated by novel environment. The plasma corticosterone level increased significantly in mice and rats after transferred to a new cage (Hennessy 1991; Yamazaki et al. 2005). Corticosterone was reported to reset the peripheral clocks and could be a mediator of the time cue (Balsalobre et al. 2000; Le Minh et al. 2001). The autonomic nervous system was also activated by novel environment, which may either enhance the output signal from the SCN pacemaker or directly reset the peripheral clocks. Timed and repeated injection of methamphetamine, a stimulant of the sympathetic nervous system, was reported to shift the circadian rhythms in clock gene expression in the peripheral tissues (Iijima et al. 2002). Catecholamines are the possible mediators of the time cue.
It is concluded that the scheduled exposures to the novel environment with a running-wheel differentially accelerate re-entrainment of the peripheral clocks in mice subjected to 8 h phase-advanced LD cycles.
Forty-nine male mice of C57BL6 strain and 50 males of Per1-luc transgenic mice of C57BL6 background (Inagaki et al. 2007) were used in the present in vivo and in vitro experiments, respectively. The mice were bred and reared in our animal quarter and maintained under an LD cycle of 12 h light and 12 h darkness (lights-on from 06.00 to 18.00 h) and with a light intensity of c. 300 lx. The animal rooms were air-conditioned at constant temperature (22–24 °C) and humidity (50–60%). They had free-access to food and water throughout the day. The animals were cared for according to the Guidelines for the Care and Use of Laboratory Animals in Hokkaido University.
Phase-shifts of LD cycle
At the age of 6–8 weeks, the mice were transferred to individual cages (15 × 25 × 15 cm) equipped with a running-wheel (10 cm diameter) to accustom them to the wheel for 10 days. A running-wheel was attached to the wire cover of the cage. Then, the animals were transferred individually to a light-proof chamber (40 × 50 × 30 cm) and housed in a cage without a running-wheel. The LD cycle in the chamber was the same as in the animal room. Locomotor activity was measured for at least 10 days to confirm entrainment of the circadian rhythm to the LD cycle, which was followed by an 8 h phase-shift of LD cycle by shortening (phase-advance) or lengthening (phase-delay) the light phase. The new light phase was from 22.00 to 10.00 h in the phase-advance shift, and from 14.00 to 02.00 h in the phase-delay shift, respectively (Fig. 1). The animals were kept under the new LD conditions for one or three cycles and then released into DD for 14 days.
Two experiments were performed with the phase-advance protocol. In one experiment, the animals were subjected to the new LD for one cycle with two exposures to the novel environment and then released into DD. In the other, they were subjected to the new LD for three cycles with four exposures and released into DD. In the phase-delay protocol, the mice were subjected to the new LD for three cycles with four exposures and released into DD. Animals were divided into two groups in each experiment. The mice in the exposed group were transferred to new individual cages with a running-wheel for 3 h from the dark onset in the new LD cycle and DD by exchanging cages outside the chamber under dim red light (ca. <0.01 lx) within 30 s (Fig. 1). The mice in the control group were treated with the same procedure except for cage exchange.
Measurement of locomotor activity
The spontaneous locomotor activity was recorded continuously by a thermal sensor and the number of movement was stored every 1 min by a computer system (The Chronobiology Kit, Stanford Software Systems, Stanford, CA). The frequency of movement (counts/5 min) was plotted in a standard double-plotted actogram for the visual estimation of daily activity onsets (CLOCKLAB software, Actimetrics, Evanston, IL). The amount of phase-shift of circadian locomotor rhythm was defined as the phase difference between the activity onset before the phase-shifts (baseline) and that on the second day of DD. The baseline phase in individual mice was defined as the mean of 10 successive activity onsets. The activity onset on the second day of DD was obtained individually from the regression line fitted to the first 10 successive activity onsets in DD. The number of wheel-revolution was counted with a micro-switch every 1 min.
Measurement of Per1 gene expression in the SCN and peripheral tissues
Per1-luc mice were used for monitoring the circadian rhythms in clock gene, Per1, expression in the SCN, liver, lung and skeletal muscle (Inagaki et al. 2007). The tissues were obtained from 12.00 to 17.00 h in the baseline, and from 16.00 to 21.00 h after the start of DD in the shifted groups. The brain and peripheral tissues were rapidly removed under dim red light (ca. <0.01 lx) and immersed in ice-cold Hanks’ balanced salt solution (pH 7.4, Sigma, St Louis, MO). The coronal brain slices of 300 µm thick were prepared with Microslicer (Dosaka, Osaka, Japan), and a pair of SCN was dissected with a surgical knife. The liver and lung were prepared in slices of 300 µm thick with Tissue Chopper (McIlwain Tissue chopper, The Mickle Laboratory, Surrey, UK). The rectus abdominis muscle was cut into fragments with a surgical knife. Each tissue was explanted on a culture membrane (Millicell CM, pore size 0.4 µm, Millipore, Bedford, MA) in a 35 mm Petri dish sealed with parafilm (American Can Co., Greenwich, CT) and incubated at 37 °C with 1.3 mL DMEM (Gibco-Invitrogen, Carlsbad, CA) supplemented with NaHCO3 (2.7 mm), HEPES (10 mm), kanamycin (20 mg/L, Gibco), insulin (5 µg/mL, Sigma), putrescine (100 mm, Sigma), human transferrin (100 mg/mL, Sigma), progesterone (20 nm, Sigma), sodium selenite 30 nm (Gibco) and 0.1 mm d-luciferin K salt (Dojindo, Gaithersburg, MD). Bioluminescence was monitored for 1 min at 10 min intervals with a dish type luminometer (AB2500 Kronos, ATTO; Lumicycle, Actimetrics, Evanston, IL) for 5 days.
Circadian phases of bioluminescence were analyzed with a modified method of Inagaki et al. (2007). Briefly, the original records were detrended by subtracting the 24 h moving average from them and smoothed by a five-point moving average method. The peak observed within the next day of explanation was defined as the first circadian peak.
Significant difference in the phase of activity onset between the baseline and shifted LD cycle was evaluated by paired t-test. The difference in the amount of phase-shift and that in free-running period between the control and exposed groups were compared by unpaired t-test. Differences in the first peak of circadian bioluminescence rhythm among the baseline, control and exposed groups were compared by one-way analysis of variance (anova) with the post-hoc Bonferroni test. Statistical analyses were performed with StatView 5.0 software (SAS Institute, Cary, NC).
This work was supported by Grant-in-Aid for Scientific Research of the Ministry of Education, Culture, Sports, Science and Technology of Japan (no. 15109004).