- Email: [email protected]
Scheduled wheel access during daytime: A method for studying conflicting zeitgebers
Physiology & Behavior 88 (2006) 459 – 465
Scheduled wheel access during daytime: A method for studying conflicting zeitgebers R. Dallmann ⁎, N. Mrosovsky Department of Zoology, University of Toronto, Ontario, Canada Received 24 February 2006; received in revised form 8 April 2006; accepted 24 April 2006
Abstract It is often stated that light is the primary environmental cue (zeitgeber) for entrainment of circadian clocks. Here, we use a new conflict test design in Syrian hamsters comparing the strength of a photic zeitgeber to that of a non-photic cue, i.e. wheel availability. Re-entrainment to an inverted LD cycle was significantly slowed down in the nocturnal hamster by restricting wheel access to the light phase of the inverted LD cycle. This effect is more pronounced if the illuminance level of the entraining lights is 0.1 lx compared to 6 lx. In this conflict design, the hamsters did not re-entrain to an inverted LD cycle for up to four weeks (when the experiment ended), but voluntarily ran during the light phase. This approximates the situation in people subjected to shift work or jet lag. © 2006 Elsevier Inc. All rights reserved. Keywords: Circadian rhythms; Conflict; Entrainment; Hamster; Jet lag; Non-photic; Photic; Shift work; Wheel access; Zeitgeber
1. Introduction Endogenous circadian clocks, which can be found in most living organisms, are of no use unless they can synchronize or entrain the organism to the cyclic environment. Most prominent of these daily changes is the light/dark cycle caused by the earth's rotation around its axis. Many papers on biological rhythms state that light is the primary or principal entraining agent, or zeitgeber [1–8]. But what exactly is meant by primary? A variety of non-photic cues have been found that can also act as zeitgebers, e.g. temperature [9–11], food availability [12–14], social cues [15,16], sound [17] and wheel running [18]. Saying that light is primary could mean that light is the entraining agent most used by animals in natural circumstances in the wild, or it could mean that light is a stronger entraining agent than non-photic cues. The latter possibility deserves further study in mammals given that, in Syrian hamsters, a species in which non-photic clock resetting has been studied intensively, the amplitude of non-photic ⁎ Corresponding author. Department of Neurobiology, University of Massachusetts Medical School, 364 Plantation Street, Worcester, MA 01605, USA. Tel.: +1 508 856 8488. E-mail address: [email protected] (R. Dallmann). 0031-9384/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.physbeh.2006.04.022
phase response curves (PRCs) can be as great as that of photic PRCs, suggesting comparable abilities to phase shift [18]. However, this may not be a valid point as double-pulse experiments show that phase shifts cannot be treated in an additive way [19,20]. A better approach to the question of relative zeitgeber strength is to set up competitions between zeitgebers. Setting up competitive situations, however, will not enable general statements to be made about whether photic or non-photic zeitgebers, as a class, are more powerful. This is because the units for specifying a photic stimulus, such as lux, are of a different quality as those specifying nonphotic clock resetting events. Indeed, with non-photic stimuli involving locomotion and/or arousal, it is not even known what are the critical variables leading to resetting of the clock. For example, Duffy et al. [21] found that light pulses facilitate adjustment to an inverted rest-activity schedule, but that social stimuli and meal schedules do not. The light pulses were 5 h of illumination of 7000–13,000 lx and given at times known from previous work to produce shifts. The social contact consisted of a technician entering the room to administer tests and provide meals or a urinal. Moreover, the social events occurred during a 16-h wake period that included 5 h of darkness in the middle. It is not known whether this procedure was optimal for eliciting non-photic
460
R. Dallmann, N. Mrosovsky / Physiology & Behavior 88 (2006) 459–465
shifts in people, and whether by having the dark period in the middle of the social stimulus advance and delay shifts might have cancelled each other out. Therefore, it might be premature to conclude that “inversion of the sleep–wake rest–activity and social contact cycles provide relatively minimal drive for resetting the human circadian pacemaker” [21]. Nevertheless, it is still valuable to study the effects of particular co-occurring photic and non-photic stimuli. Although it has been widely recognized that non-photic events play an important role in clock resetting, the interactions between the two types have been much less explored. Some of these interactions were not predicted, and were of significant magnitude [18,22]. As in real life situations photic and nonphotic stimuli are likely to interact, it is desirable to continue empirical investigations in which the relationship of non-photic events to LD cycles is altered (e.g. [23]). In this general context, the present paper explores an interaction between photic and non-photic cues in a conflict situation. The overall strategy was first to entrain hamsters to a light/dark cycle with 14 h of light per day (LD 14:10), with a running wheel being available at all times; then, after stable entrainment, to invert the LD cycle but now restrict the wheel availability to 10 h in the middle of the light portion of the LD cycle. Dominance of light over wheel availability was assessed by whether and how fast locomotion activity measured with passive infrared detectors entrained to the new LD cycle. The power of non-photic events was assessed by whether and for how long wheel access only during the light phase prevented or slowed down re-entrainment to a LD cycle inversion. Specifically, we asked two questions: (1) If the wheel is only available in the light phase after an LD cycle inversion, is the time to reentrain invariant of the light intensity? (2) Can wheel running in the middle of the light phase (i.e. the “wrong” time of day for nocturnal mammals) slow down re-entrainment after an LD cycle inversion?
installed for recording general activity as a second phase marker for the animals' clock. The PID was placed over the side of the cage at the opposite end to the running wheel. Both, the wheel revolutions and the signals from the PIDs were continuously recorded by a computerized data acquisition system (DataQuest III, Mini Mitter, Bend, Oregon, USA). A hook shaped brass rod hanging into the cage could block the wheel automatically when activated by a motor outside the cage (Fig. 1). The motors were controlled by an interface board, which could individually control up to 16 motors; this board was connected to the parallel port of a personal computer. The timing of the motor movements was set by custom-made software. The incandescent light sources were dimmed to appropriate light levels by addition of neutral density filters (Cinegel #3404, Rosco, UK). Illumination at the level of the cage floor was measured with a EX2 lux meter (B Hagner AB, Sweden). 2.3. Experiment 1 Prior to the onset of Experiment 1, the first batch of hamsters (n = 12) were held in a dim LD 14:10 for 21 days. The illumination at the cage floor was 5–6 lx. Then, the running wheels were blocked for 14 h starting at the onset of the light phase. Simultaneously, the LD cycle was inverted by having 1 day in constant light, i.e. 26 h of light starting at the onset of the light phase, with the new dark phase beginning 2 h after the former lights on (Fig. 2). Thus, the dark phase was delayed by 12 h. Starting on the first inverted day, the wheels were unblocked only throughout 10 h centered in the middle of the new light phase, which was identical with the former dark phase (Fig. 2). After 17 days of re-entrainment to the new conditions, animals were transferred to constant darkness (DD) at the onset of the last night in LD in order to assess the phase of
2. Materials and methods
A 2.1. Animals We used two batches of animals each consisting of 12 male Syrian hamsters (Mesocricetus auratus) obtained from Harlan (Indianapolis, USA). Upon arrival from the breeder, the animals were 6 weeks old and were maintained in a sound attenuated and temperature controlled room at 22 ± 1.5 °C. Animals were held individually in polypropylene cages (44 × 23 × 20 cm) equipped with a plastic mesh [24] surrounding the running wheel (17.5 cm in diameter) on standard bedding (Beta Chip, USA) with food (LabDiet #5001, PMI, USA) and tap water available ad libitum. 2.2. Apparatus design and data acquisition In addition to the wheel, a passive infrared motion detector (PID, HVW Technologies, Calgary, Alberta, Canada) was
Blocked
B
Open Fig. 1. Schematic drawing of the blocking apparatus and running wheel in (A) a blocked and (B) an open position.
R. Dallmann, N. Mrosovsky / Physiology & Behavior 88 (2006) 459–465
461
Experiment 1 6 lux 0.1 lux
Day -1
Day 0
Day 1
Day 2
Experiment 2 Scheduled Ad libitum No wheel
Day -1
Day 0
Day 1
Day 2
Fig. 2. Schedule of lighting conditions and wheel access Experiments 1 and 2. In Experiment 1 (upper panel), all animals had scheduled access to the wheel (top row), but either 6 lx (middle row) or 0.1 lx (bottom row) during L were used. In Experiment 2 (lower panel), all animals were subjected to 0.1 lx during L (1st row), but had either scheduled (2nd row), ad libitum (3rd row) or no wheel (4th row) access. Black bar is darkness, gray bar represents 0.1 lx and white bar 6 lx of light. Circles depict wheel availability.
their endogenous wheel running rhythm; the wheels were available ad libitum during the DD, i.e. they were left unblocked after the 10 h of running on the last day in LD. After 2 days in free-running conditions, the animals were placed on an LD cycle with ad libitum wheel access again and remained on that schedule for 11 days. We then repeated the procedure with a lower light level (0.1 lx), starting from the inversion of the LD cycle until the free-run. This time, however, the animals were put into free-running conditions (DD and ad libitum wheel access) after 28 days exposure to the inverted LD cycle. Due to technical problems, the records of only 10 hamsters could be analyzed.
phase. To assess this, the free-run was started when the animals were not yet re-entrained to the new LD cycle, i.e. 9 days after the inversion of the LD cycle. On arrival the second batch of hamsters were entrained for 17 days to a LD 14:10 cycle of which the first 11 days were in bright light (> 500 lx) and only the last 6 days in 0.1 lx during the L phase. Then, the LD cycle was inverted as in Experiment 2 with the difference that the hamsters were randomly assigned into only two groups (n = 6 per group). Group 1: Wheel running restricted to 10 h/day starting 2 h after lights on in the inverted LD cycle. Group 2: Ad libitum access to the running wheel.
2.4. Experiment 2 The animals from Experiment 1 were given 12 days for entrainment to LD 14:10 h with 0.1 lx during L. The LD cycle was then inverted as in Experiment 1. The animals were randomly assigned into three groups (n = 4 per group, Fig. 2).
Seven days after the inversion, the illumination level was elevated from 0.1 lx to 6 lx for 2 days until, finally, free-running conditions were introduced for both groups 9 days after inversion of the LD cycle. 2.6. Statistical analysis
Group 1: Wheel running restricted to 10 h/day starting 2 h after lights on in the inverted cycle. This was a repeat of the situation in Experiment 1. Group 2: Ad libitum access to the running wheel. Group 3: Wheel access was blocked at all times. After 21 days in the inverted conditions, the animals were released into free-run as in Experiment 1 (DD and ad libitum wheel access). 2.5. Experiment 3 This experiment was carried out to confirm that the activity recorded by the PIDs was an accurate measure of the clock
All data are given as mean ± standard error of the mean (S.E.M.). Differences between the groups in Experiment 2 were analyzed using one-way analyses of variance (ANOVA) and Scheffé's post hoc test where applicable. Actograms were plotted using ClockLab's (Actimetrics Software, Wilmette, IL, USA) scaled option set at a maximum of 80 for the running wheel data and the auto value for the PID data. In Figs. 3 and 4, single plotted actograms of running wheel (left panel) and PID recordings (right panel) were combined to one double plotted actogram, meaning running wheel and PID data for the same day were offset by 1 day. The onset of activity was determined using the built-in function of ClockLab set at a 12-h window. If, for
462
R. Dallmann, N. Mrosovsky / Physiology & Behavior 88 (2006) 459–465
DD B
-5
6 lx
0
Wheel revolutions / day
A
5
Days
10 15 20 25
0.1 lx
30 0
6
12
18
24
104
103
0.1 lx
102
6 lx 101
100
0
6 lx
D
0 0
45
45
0
10
0
0
5
0
40
40
0
10 10
DD
5
5
0
20
60
60
5
20 20
65
65
0
30 30
0
30
5
0 5
20
0.1 lx
0
0
10
Time (days)
Time of day C
DD
5
0
0
DD 5
0
10
DD 20
Fig. 3. Mean (±S.E.M.) (A) activity onsets and (B) wheel revolutions for each day after light cycle inversion in Experiment 1. Open circles 6 lx (n = 12) and black squares 0.1 lx (n = 10). The black bar marks D before the inversion. Day 0 marks LD cycle inversion. (C and D) Double plotted actogram composed from single plotted actograms of wheel running (left) and passive infrared detector (right) are arranged as one (C) for two hamsters in 6 lx and (D) for one hamster in 0.1 lx illumination during the L phase of the LD cycle. Light regimes before and after inversion are given at the top (black D, 0.1 lx gray, 6 lx white) and wheels were available only during the pre-inversion dark phase. Arrows mark transfer to free-running conditions. Light inversion occurs at day 0. Arrows mark transfer to free-running conditions. Note: In order to conservatively emphasize the difference between the two light levels, (C) shows the hamster which took the longest time to re-entrain under 6 lx (upper panel) and a typical example (lower panel), while (D) shows the hamster with the shortest time to re-entrain.
1 day, the onset of activity was more than 6 h different from the preceding and the following day, the day was excluded from the analyses. Overall, less than 5% of all days had to be
excluded from the analyses (this number also includes gaps in the data due to technical problems). The number of days to re-entrainment was calculated using the following criterion:
R. Dallmann, N. Mrosovsky / Physiology & Behavior 88 (2006) 459–465
B
-5
Wheel revolutions / day
A
0
Days
5 10 15
Scheduled Ad libitum None
20
4
10
Ad libitum 3
10
2
10
Scheduled
1
10
None
0
25 0
6
12
18
463
10
24
0
Time of Day
5
10
15
Days
20
DD
Wheel availability C
L only
None
Ad libitum
0
10
20
DD
DD
DD
Fig. 4. Mean (±S.E.M.) (A) activity onsets and (B) wheel revolutions for each day after light cycle inversion in Experiment 2 for scheduled (black squares, n = 4), ad libitum (gray circles, n = 4) and no wheel (open triangles, n = 4)access group. The black bar marks D before the inversion. (C) Representative double plotted actogram composed from two single plotted actograms of wheel running (left) and passive infrared detector (right) for all three groups of Experiment 2, i.e. wheel access during L only, no wheel access at all and ad libitum wheel access. Light inversion occurs at day 0. Light regimes before and after inversion are given at the top (black D, 0.1 lx gray). Arrows mark transfer to free-running conditions. Day 0 marks LD cycle inversion.
the first day on which the onset of activity in the PID data was within 1 h of the onset of darkness. All statistical tests were two-tailed and the level of significance was set to α = 0.05.
12 animals stopped running completely after 6 days. In contrast, in the 0.1 lx illumination, the number of wheel revolutions 13 days after the inversion was still at 6126 ± 608 representing more than 47% of baseline running on the day before the inversion, which was 12,922 ± 718 wheel revolutions (Fig. 3B).
3. Results 3.2. Experiment 2 3.1. Experiment 1 With illumination of 6 lx during the L phase of the LD cycle, the animals re-entrained after 5.7 ± 1.6 days (n = 12) after the LD cycle inversion. In the lower light level (0.1 lx), it took more than 28 days to re-entrain to the new LD cycle (Fig. 3A, C and D). Although all animals re-entrained much faster in the brighter light, there were 2 out of 12 animals that took more than 2 weeks to re-entrain (Fig. 3C). This result is reflected in the amount of running during the 10 h in the middle of the L phase of the light cycle. In the 6 lx illumination, the animals reduced their daily running drastically from 17,125 ± 369 on the day before the inversion to 316 ± 277 revolutions after 13 days in the inverted LD cycle, and 9 out of
Group 1 (scheduled wheel access) confirmed the results for the same light intensity in Experiment 1. None of the four hamsters in this group were re-entrained after 22 days. In contrast, all animals without wheel access (n = 4) as well as all animals with ad libitum wheel access (n = 4) re-entrained to the inverted light dark cycle within a maximum of 4 days (Fig. 4A). This, again, is reflected in the amount of wheel running in groups 1 and 2 (Fig. 4B). Similar to Experiment 1, the hamsters with the restricted access showed a steady decline in the amount of wheel running from 9874 ± 1104 revolutions at baseline to only 162 ± 93 revolutions 13 days after the inversion. This was only 2% of baseline revolutions. On the first day of DD free-run, this group was back to baseline running levels (9353 ± 1767).
464
R. Dallmann, N. Mrosovsky / Physiology & Behavior 88 (2006) 459–465
Percent of daily events per hour
PID scheduled Wheel scheduled 16
PID ad libitum Wheel ad libitum 12
8
4
0 4
8
12
16
20
24
Time of day Fig. 5. Mean percentage of general activity (squares) and wheel revolutions (circles) per hour for the ad libitum (open symbols) and scheduled access (black symbols) groups from Experiment 3 on the first day under free-running conditions (n = 6 per group). The black bar marks D on the last day in the inverted LD cycle.
The hamsters with unlimited access to their running wheel, however, did not differ from baseline (baseline: 8848 ± 1361 vs. day 13: 7177 ± 1008) in the amount of wheel revolutions except on the day of the inversion and the first day after the inversion. 3.3. Experiment 3 None of the hamsters in either group started to re-entrain to the inverted LD cycle until the illumination was brightened to 6 lx. For group 1, with scheduled wheel access, this is consistent with the results of Experiments 1 and 2. For group 2, with ad libitum wheel access, it took longer to re-entrain than expected. This might be explained by the light history of these hamsters. Unlike the first batch used in Experiments 1 and 2, these animals had only 6 days in the 0.1 lx dim light. Thus, long-term changes in the photoreceptor system might account for the apparent discrepancy. A similar phenomenon has been reported for melatonin suppression in humans [25] and, at the molecular level, the up-regulation of melanopsin by rod and cone input might be responsible [26]. In order to accelerate the process, after 7 days in the inverted LD cycle, the illumination level was changed to 6 lx. In the brighter light, consistent with the other experiments, group 2 was re-entrained rapidly to the inverted LD cycle, whereas the group 1 had the most of the activity still in the L phase. This difference in phase between the groups was confirmed after 2 days in the brighter light by the release into DD and ad libitum wheel access for both groups. Furthermore, it was confirmed that the PIDs and wheel running were exchangeable measures of clock phase. Except for the different amplitudes, the waveforms for both measures were closely similar (Fig. 5). 4. Discussion Wheel access confined to the “wrong” time of day (i.e. the light phase after an LD cycle inversion) significantly slowed
down rate of re-entrainment to the inverted LD cycle. Importantly, it took hamsters longer to re-entrain if the wheel was available only during the light phase than if it was not available at all after the inversion (Fig. 4). This implies that wheel availability only at this time actually has a retarding effect on the rate of re-entrainment. Earlier work suggests that that might be the case in people, too. Moog and Hildebrandt [27] showed that exercise can prevent adjustment to nightshift schedules in certain subjects, although their study suffered from a small sample size. Furthermore, exercise in people is known to result in phase delays and advances under certain conditions [28,29]. The retarding effect of wheel availability in the L phase was greater if the light level was 0.1 lx than if it was 6 lx; at both these illumination levels, the animals took longer to reach the entrainment criterion than did the animals with ad libitum access to the wheel in Experiment 2 and Experiment 3. This suggests that light competes with the non-photic cue rather than being all dominant. When the light was dimmer, re-entrainment was slowed down more, suggesting that the relative strength of the non-photic zeitgeber was increased. Besides the fact that brighter light yields bigger phase shifts [30], a second phenomenon seems to contribute to the present results: masking. The magnitude of wheel running induced phase shifting is positively correlated with the number of wheel revolutions [31]. Thus, less running would lead to a faster reentrainment to light, which was actually the case in Experiment 1. In 6 lx, running wheel activity in hamsters was suppressed to less than 2% of the baseline running but in 0.1 lx activity levels still reached 47% of baseline (Fig. 3B). Thus, masking might have accounted for low wheel running which in turn enabled faster re-entrainment in the brighter light. Although masking probably contributed to the results on rate of re-entrainment, it is not sufficient to explain all of the results. Otherwise the animals would re-entrain faster if they ran less. In Experiment 2, however, group 3, which had no access to the wheel at any time, took the same number of days to re-entrain as group 2 with ad libitum access, but much shorter than group 1 with the scheduled wheel access. Since the amount of running was obviously lowest in group 3, intermediate in group 1 and highest in group 2 (Fig. 4B), this suggests that not only the physical activity of running in a wheel but also a general arousing effect of the wheel availability as such was involved. It cannot be excluded, however, that the hamsters were more active in the cage because the PIDs cannot quantify general activity accurately. Our findings are relevant attempts to use non-photic inputs to ameliorate jet lag and rhythm disturbances associated with shift work. There are various reports of non-photic phase shifting in people, which involve physical activity in one way or another [29,32]. The present findings emphasize the importance of scheduling non-photic stimuli at an appropriate phase of the cycle. Activity or arousal related events at an inappropriate time may not only fail to have a beneficial effect, but could actually retard rate of re-entrainment. On the other hand, in certain circumstances this very retarding effect could be useful. If for example, one has to
R. Dallmann, N. Mrosovsky / Physiology & Behavior 88 (2006) 459–465
travel back and forth between time zones without adequate intervening periods for recovery, or is on a rapidly rotating shift work schedule, then non-photic inputs might help maintain a stable internal time and prevent internal desynchronization. It has been reported that frequent travel with short recovery periods in between is associated with longer reaction time and poorer performance on a delayed matching to sample task if the subjects had less than 5 days to recover between changing back to the original time zone [33]. In conclusion, because the units of measurement for light and non-photic inputs are different, it is necessarily difficult to compare strengths of these zeitgebers in general. At particular values (in their different units), photic and non-photic inputs can be compared to each other when both are presented in a conflict test situation such as the one described here. The present experimental design reveals that non-photic events, although often labeled as weak zeitgebers, can greatly retard reentrainment of locomotor activity to new LD cycles. However, much more remains to be discovered about this situation. Is the re-entrainment of temperature and hormonal rhythms also slowed? Different peripheral organs and even different parts of the SCN might differ in the rate of re-entrainment thereby causing internal desynchronization. The present test procedure is offered as a way of investigating such questions. Acknowledgement We thank Peggy Salmon and Stefanie LaZerte for animal care, Eckhard Glockmann and Trung Luu for development of the blockers, as well as Gunther Lemm for the design and advice on the programming of the control unit. Support came from the Canadian Institutes of Health Research. References [1] Gamble KL, Ehlen CJ, Albers HE. Circadian control during the day and night: role of neuropeptide Y Y5 receptors in the suprachiasmatic nucleus. Brain Res Bull 2005;65:513–9. [2] Klerman EB, Rimmer DW, Dijk DJ, Kronauer RE, Rizzo III JF, Czeisler C. Nonphotic entrainment of the human circadian pacemaker. Am J Physiol 1998;274:R991–6. [3] Rusak B, Zucker I. Neural regulation of circadian rhythms. Physiol Rev 1979;59:449–526. [4] Roenneberg T, Merrow M. Circadian clocks: Omnes viae Romam ducunt. Curr Biol 2000;10:R742–5. [5] Ashmore LJ, Sehgal A. A fly's eye view of circadian entrainment. J Biol Rhythms 2003;18:206–16. [6] Bae K, Weaver DR. Light-induced phase shifts in mice lacking mPER1 or mPER2. J Biol Rhythms 2003;18:123–33. [7] Brzezinski JAI, Brown NL, Tanikawa A, Bush RA, Sieving PA, Vitaterna MH, et al. Loss of circadian photoentrainment and abnormal retinal electrophysiology in Math5 mutant mice. Invest Ophthalmol Vis Sci 2005;46:2540–51. [8] Goel N. Late-night presentation of an auditory stimulus phase delays human circadian rhythms. Am J Physiol 2005;289:R209–16. [9] Liu Y, Merrow M, Loros JJ, Dunlap JC. How temperature changes reset a circadian oscillator. Science 1998;281:825–9.
465
[10] Rajaratnam SM, Redman JR. Entrainment of activity rhythms to temperature cycles in diurnal palm squirrels. Physiol Behav 1998;63:271–7. [11] Hoffmann K. Temperaturcyclen als Zeitgeber der circadianen Periodik. Verh Dtsch Zool Ges Innsbr 1968:266–73. [12] Challet E, Solberg LC, Turek FW. Entrainment in calorie-restricted mice: conflicting zeitgebers and free-running conditions. Am J Physiol 1998;274:R1751–61. [13] Mistlberger RE. Circadian food-anticipatory activity: formal models and physiological mechanisms. Neurosci Biobehav Rev 1994;18:171–95. [14] Weber DN, Spieler RE. Effects of the light–dark cycle and scheduled feeding on behavioral and reproductive rhythms of the cyprinodont fish, Medaka, Oryzias atipes. Experientia 1987;43:621–4. [15] Levine JD, Funes P, Dowse HB, Hall JC. Resetting the circadian clock by social experience in Drosophila melanogaster. Science 2002;298:2010–2. [16] Mrosovsky N. Phase response curves for social entrainment. J Comp Physiol A 1988;162:35–46. [17] Menaker M, Eskin A. Entrainment of circadian rhythms by sound in Passer domesticus. Science 1966;154:1579–81. [18] Mrosovsky N. Locomotor activity and non-photic influences on circadian clocks. Biol Rev Camb Philos Soc 1996;71:343–72. [19] Steinlechner S, Stieglitz A, Ruf T. Djungarian hamsters: a species with a labile circadian pacemaker? Arrhythmicity under a light–dark cycle induced by short light pulses. J Biol Rhythms 2002;17:248–58. [20] Mrosovsky N. Double-pulse experiments with nonphotic and photic phase-shifting stimuli. J Biol Rhythms 1991;6:167–79. [21] Duffy JF, Kronauer RE, Czeisler CA. Phase-shifting human circadian rhythms: influence of sleep timing, social contact and light exposure. J Physiol 1996;495:289–97. [22] Knoch ME, Gobes SM, Pavlovska I, Su C, Mistlberger RE, Glass JD. Short-term exposure to constant light promotes strong circadian phaseresetting responses to nonphotic stimuli in Syrian hamsters. Eur J Neurosci 2004;19:2779–90. [23] Murphy HM, Wideman CH, Nadzam GR. A laboratory animal model of human shift work. Integr Physiol Behav Sci 2003;38:316–28. [24] Mrosovsky N, Salmon PA, Vrang N. Revolutionary science: an improved running wheel for hamsters. Chronobiol Int 1998;15:147–58. [25] Smith KA, Schoen MW, Czeisler CA. Adaptation of human pineal melatonin suppression by recent photic history. J Clin Endocrinol Metab 2004;89:3610–4. [26] Sakamoto K, Liu C, Kasamatsu M, Pozdeyev NV, Iuvone PM, Tosini G. Dopamine regulates melanopsin mRNA expression in intrinsically photosensitive retinal ganglion cells. Eur J Neurosci 2005;22:3129–36. [27] Moog R, Hildebrandt G. Adaptation to shift work-experimental approaches with reduced masking effects. Chronobiol Int 1989;6:65–75. [28] Barger LK, Wright Jr KP, Hughes RJ, Czeisler CA. Daily exercise facilitates phase delays of circadian melatonin rhythm in very dim light. Am J Physiol Regul Integr Comp Physiol 2004;286:R1077–84. [29] Buxton OM, Lee CW, L'Hermite-Baleriaux M, Turek FW, Van Cauter E. Exercise elicits phase shifts and acute alterations of melatonin that vary with circadian phase. Am J Physiol 2003;284:R714–24. [30] Peirson SN, Thompson S, Hankins MW, Foster RG. Mammalian photoentrainment: results, methods, and approaches. Methods Enzymol 2005;393:697–726. [31] Bobrzynska KJ, Mrosovsky N. Phase shifting by novelty-induced running: activity dose–response curves at different circadian times. J Comp Physiol A 1998;182:251–8. [32] Honma K, Hashimoto S, Nakao M, Honma S. Period and phase adjustments of human circadian rhythms in the real world. J Biol Rhythms 2003;18:261–70. [33] Cho K. Chronic ‘jet lag’ produces temporal lobe atrophy and spatial cognitive deficits. Nat Neurosci 2001;4:567–8.
Scheduled wheel access during daytime: A method for studying conflicting zeitgebers R. Dallmann ⁎, N. Mrosovsky Department of Zoology, University of Toronto, Ontario, Canada Received 24 February 2006; received in revised form 8 April 2006; accepted 24 April 2006
Abstract It is often stated that light is the primary environmental cue (zeitgeber) for entrainment of circadian clocks. Here, we use a new conflict test design in Syrian hamsters comparing the strength of a photic zeitgeber to that of a non-photic cue, i.e. wheel availability. Re-entrainment to an inverted LD cycle was significantly slowed down in the nocturnal hamster by restricting wheel access to the light phase of the inverted LD cycle. This effect is more pronounced if the illuminance level of the entraining lights is 0.1 lx compared to 6 lx. In this conflict design, the hamsters did not re-entrain to an inverted LD cycle for up to four weeks (when the experiment ended), but voluntarily ran during the light phase. This approximates the situation in people subjected to shift work or jet lag. © 2006 Elsevier Inc. All rights reserved. Keywords: Circadian rhythms; Conflict; Entrainment; Hamster; Jet lag; Non-photic; Photic; Shift work; Wheel access; Zeitgeber
1. Introduction Endogenous circadian clocks, which can be found in most living organisms, are of no use unless they can synchronize or entrain the organism to the cyclic environment. Most prominent of these daily changes is the light/dark cycle caused by the earth's rotation around its axis. Many papers on biological rhythms state that light is the primary or principal entraining agent, or zeitgeber [1–8]. But what exactly is meant by primary? A variety of non-photic cues have been found that can also act as zeitgebers, e.g. temperature [9–11], food availability [12–14], social cues [15,16], sound [17] and wheel running [18]. Saying that light is primary could mean that light is the entraining agent most used by animals in natural circumstances in the wild, or it could mean that light is a stronger entraining agent than non-photic cues. The latter possibility deserves further study in mammals given that, in Syrian hamsters, a species in which non-photic clock resetting has been studied intensively, the amplitude of non-photic ⁎ Corresponding author. Department of Neurobiology, University of Massachusetts Medical School, 364 Plantation Street, Worcester, MA 01605, USA. Tel.: +1 508 856 8488. E-mail address: [email protected] (R. Dallmann). 0031-9384/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.physbeh.2006.04.022
phase response curves (PRCs) can be as great as that of photic PRCs, suggesting comparable abilities to phase shift [18]. However, this may not be a valid point as double-pulse experiments show that phase shifts cannot be treated in an additive way [19,20]. A better approach to the question of relative zeitgeber strength is to set up competitions between zeitgebers. Setting up competitive situations, however, will not enable general statements to be made about whether photic or non-photic zeitgebers, as a class, are more powerful. This is because the units for specifying a photic stimulus, such as lux, are of a different quality as those specifying nonphotic clock resetting events. Indeed, with non-photic stimuli involving locomotion and/or arousal, it is not even known what are the critical variables leading to resetting of the clock. For example, Duffy et al. [21] found that light pulses facilitate adjustment to an inverted rest-activity schedule, but that social stimuli and meal schedules do not. The light pulses were 5 h of illumination of 7000–13,000 lx and given at times known from previous work to produce shifts. The social contact consisted of a technician entering the room to administer tests and provide meals or a urinal. Moreover, the social events occurred during a 16-h wake period that included 5 h of darkness in the middle. It is not known whether this procedure was optimal for eliciting non-photic
460
R. Dallmann, N. Mrosovsky / Physiology & Behavior 88 (2006) 459–465
shifts in people, and whether by having the dark period in the middle of the social stimulus advance and delay shifts might have cancelled each other out. Therefore, it might be premature to conclude that “inversion of the sleep–wake rest–activity and social contact cycles provide relatively minimal drive for resetting the human circadian pacemaker” [21]. Nevertheless, it is still valuable to study the effects of particular co-occurring photic and non-photic stimuli. Although it has been widely recognized that non-photic events play an important role in clock resetting, the interactions between the two types have been much less explored. Some of these interactions were not predicted, and were of significant magnitude [18,22]. As in real life situations photic and nonphotic stimuli are likely to interact, it is desirable to continue empirical investigations in which the relationship of non-photic events to LD cycles is altered (e.g. [23]). In this general context, the present paper explores an interaction between photic and non-photic cues in a conflict situation. The overall strategy was first to entrain hamsters to a light/dark cycle with 14 h of light per day (LD 14:10), with a running wheel being available at all times; then, after stable entrainment, to invert the LD cycle but now restrict the wheel availability to 10 h in the middle of the light portion of the LD cycle. Dominance of light over wheel availability was assessed by whether and how fast locomotion activity measured with passive infrared detectors entrained to the new LD cycle. The power of non-photic events was assessed by whether and for how long wheel access only during the light phase prevented or slowed down re-entrainment to a LD cycle inversion. Specifically, we asked two questions: (1) If the wheel is only available in the light phase after an LD cycle inversion, is the time to reentrain invariant of the light intensity? (2) Can wheel running in the middle of the light phase (i.e. the “wrong” time of day for nocturnal mammals) slow down re-entrainment after an LD cycle inversion?
installed for recording general activity as a second phase marker for the animals' clock. The PID was placed over the side of the cage at the opposite end to the running wheel. Both, the wheel revolutions and the signals from the PIDs were continuously recorded by a computerized data acquisition system (DataQuest III, Mini Mitter, Bend, Oregon, USA). A hook shaped brass rod hanging into the cage could block the wheel automatically when activated by a motor outside the cage (Fig. 1). The motors were controlled by an interface board, which could individually control up to 16 motors; this board was connected to the parallel port of a personal computer. The timing of the motor movements was set by custom-made software. The incandescent light sources were dimmed to appropriate light levels by addition of neutral density filters (Cinegel #3404, Rosco, UK). Illumination at the level of the cage floor was measured with a EX2 lux meter (B Hagner AB, Sweden). 2.3. Experiment 1 Prior to the onset of Experiment 1, the first batch of hamsters (n = 12) were held in a dim LD 14:10 for 21 days. The illumination at the cage floor was 5–6 lx. Then, the running wheels were blocked for 14 h starting at the onset of the light phase. Simultaneously, the LD cycle was inverted by having 1 day in constant light, i.e. 26 h of light starting at the onset of the light phase, with the new dark phase beginning 2 h after the former lights on (Fig. 2). Thus, the dark phase was delayed by 12 h. Starting on the first inverted day, the wheels were unblocked only throughout 10 h centered in the middle of the new light phase, which was identical with the former dark phase (Fig. 2). After 17 days of re-entrainment to the new conditions, animals were transferred to constant darkness (DD) at the onset of the last night in LD in order to assess the phase of
2. Materials and methods
A 2.1. Animals We used two batches of animals each consisting of 12 male Syrian hamsters (Mesocricetus auratus) obtained from Harlan (Indianapolis, USA). Upon arrival from the breeder, the animals were 6 weeks old and were maintained in a sound attenuated and temperature controlled room at 22 ± 1.5 °C. Animals were held individually in polypropylene cages (44 × 23 × 20 cm) equipped with a plastic mesh [24] surrounding the running wheel (17.5 cm in diameter) on standard bedding (Beta Chip, USA) with food (LabDiet #5001, PMI, USA) and tap water available ad libitum. 2.2. Apparatus design and data acquisition In addition to the wheel, a passive infrared motion detector (PID, HVW Technologies, Calgary, Alberta, Canada) was
Blocked
B
Open Fig. 1. Schematic drawing of the blocking apparatus and running wheel in (A) a blocked and (B) an open position.
R. Dallmann, N. Mrosovsky / Physiology & Behavior 88 (2006) 459–465
461
Experiment 1 6 lux 0.1 lux
Day -1
Day 0
Day 1
Day 2
Experiment 2 Scheduled Ad libitum No wheel
Day -1
Day 0
Day 1
Day 2
Fig. 2. Schedule of lighting conditions and wheel access Experiments 1 and 2. In Experiment 1 (upper panel), all animals had scheduled access to the wheel (top row), but either 6 lx (middle row) or 0.1 lx (bottom row) during L were used. In Experiment 2 (lower panel), all animals were subjected to 0.1 lx during L (1st row), but had either scheduled (2nd row), ad libitum (3rd row) or no wheel (4th row) access. Black bar is darkness, gray bar represents 0.1 lx and white bar 6 lx of light. Circles depict wheel availability.
their endogenous wheel running rhythm; the wheels were available ad libitum during the DD, i.e. they were left unblocked after the 10 h of running on the last day in LD. After 2 days in free-running conditions, the animals were placed on an LD cycle with ad libitum wheel access again and remained on that schedule for 11 days. We then repeated the procedure with a lower light level (0.1 lx), starting from the inversion of the LD cycle until the free-run. This time, however, the animals were put into free-running conditions (DD and ad libitum wheel access) after 28 days exposure to the inverted LD cycle. Due to technical problems, the records of only 10 hamsters could be analyzed.
phase. To assess this, the free-run was started when the animals were not yet re-entrained to the new LD cycle, i.e. 9 days after the inversion of the LD cycle. On arrival the second batch of hamsters were entrained for 17 days to a LD 14:10 cycle of which the first 11 days were in bright light (> 500 lx) and only the last 6 days in 0.1 lx during the L phase. Then, the LD cycle was inverted as in Experiment 2 with the difference that the hamsters were randomly assigned into only two groups (n = 6 per group). Group 1: Wheel running restricted to 10 h/day starting 2 h after lights on in the inverted LD cycle. Group 2: Ad libitum access to the running wheel.
2.4. Experiment 2 The animals from Experiment 1 were given 12 days for entrainment to LD 14:10 h with 0.1 lx during L. The LD cycle was then inverted as in Experiment 1. The animals were randomly assigned into three groups (n = 4 per group, Fig. 2).
Seven days after the inversion, the illumination level was elevated from 0.1 lx to 6 lx for 2 days until, finally, free-running conditions were introduced for both groups 9 days after inversion of the LD cycle. 2.6. Statistical analysis
Group 1: Wheel running restricted to 10 h/day starting 2 h after lights on in the inverted cycle. This was a repeat of the situation in Experiment 1. Group 2: Ad libitum access to the running wheel. Group 3: Wheel access was blocked at all times. After 21 days in the inverted conditions, the animals were released into free-run as in Experiment 1 (DD and ad libitum wheel access). 2.5. Experiment 3 This experiment was carried out to confirm that the activity recorded by the PIDs was an accurate measure of the clock
All data are given as mean ± standard error of the mean (S.E.M.). Differences between the groups in Experiment 2 were analyzed using one-way analyses of variance (ANOVA) and Scheffé's post hoc test where applicable. Actograms were plotted using ClockLab's (Actimetrics Software, Wilmette, IL, USA) scaled option set at a maximum of 80 for the running wheel data and the auto value for the PID data. In Figs. 3 and 4, single plotted actograms of running wheel (left panel) and PID recordings (right panel) were combined to one double plotted actogram, meaning running wheel and PID data for the same day were offset by 1 day. The onset of activity was determined using the built-in function of ClockLab set at a 12-h window. If, for
462
R. Dallmann, N. Mrosovsky / Physiology & Behavior 88 (2006) 459–465
DD B
-5
6 lx
0
Wheel revolutions / day
A
5
Days
10 15 20 25
0.1 lx
30 0
6
12
18
24
104
103
0.1 lx
102
6 lx 101
100
0
6 lx
D
0 0
45
45
0
10
0
0
5
0
40
40
0
10 10
DD
5
5
0
20
60
60
5
20 20
65
65
0
30 30
0
30
5
0 5
20
0.1 lx
0
0
10
Time (days)
Time of day C
DD
5
0
0
DD 5
0
10
DD 20
Fig. 3. Mean (±S.E.M.) (A) activity onsets and (B) wheel revolutions for each day after light cycle inversion in Experiment 1. Open circles 6 lx (n = 12) and black squares 0.1 lx (n = 10). The black bar marks D before the inversion. Day 0 marks LD cycle inversion. (C and D) Double plotted actogram composed from single plotted actograms of wheel running (left) and passive infrared detector (right) are arranged as one (C) for two hamsters in 6 lx and (D) for one hamster in 0.1 lx illumination during the L phase of the LD cycle. Light regimes before and after inversion are given at the top (black D, 0.1 lx gray, 6 lx white) and wheels were available only during the pre-inversion dark phase. Arrows mark transfer to free-running conditions. Light inversion occurs at day 0. Arrows mark transfer to free-running conditions. Note: In order to conservatively emphasize the difference between the two light levels, (C) shows the hamster which took the longest time to re-entrain under 6 lx (upper panel) and a typical example (lower panel), while (D) shows the hamster with the shortest time to re-entrain.
1 day, the onset of activity was more than 6 h different from the preceding and the following day, the day was excluded from the analyses. Overall, less than 5% of all days had to be
excluded from the analyses (this number also includes gaps in the data due to technical problems). The number of days to re-entrainment was calculated using the following criterion:
R. Dallmann, N. Mrosovsky / Physiology & Behavior 88 (2006) 459–465
B
-5
Wheel revolutions / day
A
0
Days
5 10 15
Scheduled Ad libitum None
20
4
10
Ad libitum 3
10
2
10
Scheduled
1
10
None
0
25 0
6
12
18
463
10
24
0
Time of Day
5
10
15
Days
20
DD
Wheel availability C
L only
None
Ad libitum
0
10
20
DD
DD
DD
Fig. 4. Mean (±S.E.M.) (A) activity onsets and (B) wheel revolutions for each day after light cycle inversion in Experiment 2 for scheduled (black squares, n = 4), ad libitum (gray circles, n = 4) and no wheel (open triangles, n = 4)access group. The black bar marks D before the inversion. (C) Representative double plotted actogram composed from two single plotted actograms of wheel running (left) and passive infrared detector (right) for all three groups of Experiment 2, i.e. wheel access during L only, no wheel access at all and ad libitum wheel access. Light inversion occurs at day 0. Light regimes before and after inversion are given at the top (black D, 0.1 lx gray). Arrows mark transfer to free-running conditions. Day 0 marks LD cycle inversion.
the first day on which the onset of activity in the PID data was within 1 h of the onset of darkness. All statistical tests were two-tailed and the level of significance was set to α = 0.05.
12 animals stopped running completely after 6 days. In contrast, in the 0.1 lx illumination, the number of wheel revolutions 13 days after the inversion was still at 6126 ± 608 representing more than 47% of baseline running on the day before the inversion, which was 12,922 ± 718 wheel revolutions (Fig. 3B).
3. Results 3.2. Experiment 2 3.1. Experiment 1 With illumination of 6 lx during the L phase of the LD cycle, the animals re-entrained after 5.7 ± 1.6 days (n = 12) after the LD cycle inversion. In the lower light level (0.1 lx), it took more than 28 days to re-entrain to the new LD cycle (Fig. 3A, C and D). Although all animals re-entrained much faster in the brighter light, there were 2 out of 12 animals that took more than 2 weeks to re-entrain (Fig. 3C). This result is reflected in the amount of running during the 10 h in the middle of the L phase of the light cycle. In the 6 lx illumination, the animals reduced their daily running drastically from 17,125 ± 369 on the day before the inversion to 316 ± 277 revolutions after 13 days in the inverted LD cycle, and 9 out of
Group 1 (scheduled wheel access) confirmed the results for the same light intensity in Experiment 1. None of the four hamsters in this group were re-entrained after 22 days. In contrast, all animals without wheel access (n = 4) as well as all animals with ad libitum wheel access (n = 4) re-entrained to the inverted light dark cycle within a maximum of 4 days (Fig. 4A). This, again, is reflected in the amount of wheel running in groups 1 and 2 (Fig. 4B). Similar to Experiment 1, the hamsters with the restricted access showed a steady decline in the amount of wheel running from 9874 ± 1104 revolutions at baseline to only 162 ± 93 revolutions 13 days after the inversion. This was only 2% of baseline revolutions. On the first day of DD free-run, this group was back to baseline running levels (9353 ± 1767).
464
R. Dallmann, N. Mrosovsky / Physiology & Behavior 88 (2006) 459–465
Percent of daily events per hour
PID scheduled Wheel scheduled 16
PID ad libitum Wheel ad libitum 12
8
4
0 4
8
12
16
20
24
Time of day Fig. 5. Mean percentage of general activity (squares) and wheel revolutions (circles) per hour for the ad libitum (open symbols) and scheduled access (black symbols) groups from Experiment 3 on the first day under free-running conditions (n = 6 per group). The black bar marks D on the last day in the inverted LD cycle.
The hamsters with unlimited access to their running wheel, however, did not differ from baseline (baseline: 8848 ± 1361 vs. day 13: 7177 ± 1008) in the amount of wheel revolutions except on the day of the inversion and the first day after the inversion. 3.3. Experiment 3 None of the hamsters in either group started to re-entrain to the inverted LD cycle until the illumination was brightened to 6 lx. For group 1, with scheduled wheel access, this is consistent with the results of Experiments 1 and 2. For group 2, with ad libitum wheel access, it took longer to re-entrain than expected. This might be explained by the light history of these hamsters. Unlike the first batch used in Experiments 1 and 2, these animals had only 6 days in the 0.1 lx dim light. Thus, long-term changes in the photoreceptor system might account for the apparent discrepancy. A similar phenomenon has been reported for melatonin suppression in humans [25] and, at the molecular level, the up-regulation of melanopsin by rod and cone input might be responsible [26]. In order to accelerate the process, after 7 days in the inverted LD cycle, the illumination level was changed to 6 lx. In the brighter light, consistent with the other experiments, group 2 was re-entrained rapidly to the inverted LD cycle, whereas the group 1 had the most of the activity still in the L phase. This difference in phase between the groups was confirmed after 2 days in the brighter light by the release into DD and ad libitum wheel access for both groups. Furthermore, it was confirmed that the PIDs and wheel running were exchangeable measures of clock phase. Except for the different amplitudes, the waveforms for both measures were closely similar (Fig. 5). 4. Discussion Wheel access confined to the “wrong” time of day (i.e. the light phase after an LD cycle inversion) significantly slowed
down rate of re-entrainment to the inverted LD cycle. Importantly, it took hamsters longer to re-entrain if the wheel was available only during the light phase than if it was not available at all after the inversion (Fig. 4). This implies that wheel availability only at this time actually has a retarding effect on the rate of re-entrainment. Earlier work suggests that that might be the case in people, too. Moog and Hildebrandt [27] showed that exercise can prevent adjustment to nightshift schedules in certain subjects, although their study suffered from a small sample size. Furthermore, exercise in people is known to result in phase delays and advances under certain conditions [28,29]. The retarding effect of wheel availability in the L phase was greater if the light level was 0.1 lx than if it was 6 lx; at both these illumination levels, the animals took longer to reach the entrainment criterion than did the animals with ad libitum access to the wheel in Experiment 2 and Experiment 3. This suggests that light competes with the non-photic cue rather than being all dominant. When the light was dimmer, re-entrainment was slowed down more, suggesting that the relative strength of the non-photic zeitgeber was increased. Besides the fact that brighter light yields bigger phase shifts [30], a second phenomenon seems to contribute to the present results: masking. The magnitude of wheel running induced phase shifting is positively correlated with the number of wheel revolutions [31]. Thus, less running would lead to a faster reentrainment to light, which was actually the case in Experiment 1. In 6 lx, running wheel activity in hamsters was suppressed to less than 2% of the baseline running but in 0.1 lx activity levels still reached 47% of baseline (Fig. 3B). Thus, masking might have accounted for low wheel running which in turn enabled faster re-entrainment in the brighter light. Although masking probably contributed to the results on rate of re-entrainment, it is not sufficient to explain all of the results. Otherwise the animals would re-entrain faster if they ran less. In Experiment 2, however, group 3, which had no access to the wheel at any time, took the same number of days to re-entrain as group 2 with ad libitum access, but much shorter than group 1 with the scheduled wheel access. Since the amount of running was obviously lowest in group 3, intermediate in group 1 and highest in group 2 (Fig. 4B), this suggests that not only the physical activity of running in a wheel but also a general arousing effect of the wheel availability as such was involved. It cannot be excluded, however, that the hamsters were more active in the cage because the PIDs cannot quantify general activity accurately. Our findings are relevant attempts to use non-photic inputs to ameliorate jet lag and rhythm disturbances associated with shift work. There are various reports of non-photic phase shifting in people, which involve physical activity in one way or another [29,32]. The present findings emphasize the importance of scheduling non-photic stimuli at an appropriate phase of the cycle. Activity or arousal related events at an inappropriate time may not only fail to have a beneficial effect, but could actually retard rate of re-entrainment. On the other hand, in certain circumstances this very retarding effect could be useful. If for example, one has to
R. Dallmann, N. Mrosovsky / Physiology & Behavior 88 (2006) 459–465
travel back and forth between time zones without adequate intervening periods for recovery, or is on a rapidly rotating shift work schedule, then non-photic inputs might help maintain a stable internal time and prevent internal desynchronization. It has been reported that frequent travel with short recovery periods in between is associated with longer reaction time and poorer performance on a delayed matching to sample task if the subjects had less than 5 days to recover between changing back to the original time zone [33]. In conclusion, because the units of measurement for light and non-photic inputs are different, it is necessarily difficult to compare strengths of these zeitgebers in general. At particular values (in their different units), photic and non-photic inputs can be compared to each other when both are presented in a conflict test situation such as the one described here. The present experimental design reveals that non-photic events, although often labeled as weak zeitgebers, can greatly retard reentrainment of locomotor activity to new LD cycles. However, much more remains to be discovered about this situation. Is the re-entrainment of temperature and hormonal rhythms also slowed? Different peripheral organs and even different parts of the SCN might differ in the rate of re-entrainment thereby causing internal desynchronization. The present test procedure is offered as a way of investigating such questions. Acknowledgement We thank Peggy Salmon and Stefanie LaZerte for animal care, Eckhard Glockmann and Trung Luu for development of the blockers, as well as Gunther Lemm for the design and advice on the programming of the control unit. Support came from the Canadian Institutes of Health Research. References [1] Gamble KL, Ehlen CJ, Albers HE. Circadian control during the day and night: role of neuropeptide Y Y5 receptors in the suprachiasmatic nucleus. Brain Res Bull 2005;65:513–9. [2] Klerman EB, Rimmer DW, Dijk DJ, Kronauer RE, Rizzo III JF, Czeisler C. Nonphotic entrainment of the human circadian pacemaker. Am J Physiol 1998;274:R991–6. [3] Rusak B, Zucker I. Neural regulation of circadian rhythms. Physiol Rev 1979;59:449–526. [4] Roenneberg T, Merrow M. Circadian clocks: Omnes viae Romam ducunt. Curr Biol 2000;10:R742–5. [5] Ashmore LJ, Sehgal A. A fly's eye view of circadian entrainment. J Biol Rhythms 2003;18:206–16. [6] Bae K, Weaver DR. Light-induced phase shifts in mice lacking mPER1 or mPER2. J Biol Rhythms 2003;18:123–33. [7] Brzezinski JAI, Brown NL, Tanikawa A, Bush RA, Sieving PA, Vitaterna MH, et al. Loss of circadian photoentrainment and abnormal retinal electrophysiology in Math5 mutant mice. Invest Ophthalmol Vis Sci 2005;46:2540–51. [8] Goel N. Late-night presentation of an auditory stimulus phase delays human circadian rhythms. Am J Physiol 2005;289:R209–16. [9] Liu Y, Merrow M, Loros JJ, Dunlap JC. How temperature changes reset a circadian oscillator. Science 1998;281:825–9.
465
[10] Rajaratnam SM, Redman JR. Entrainment of activity rhythms to temperature cycles in diurnal palm squirrels. Physiol Behav 1998;63:271–7. [11] Hoffmann K. Temperaturcyclen als Zeitgeber der circadianen Periodik. Verh Dtsch Zool Ges Innsbr 1968:266–73. [12] Challet E, Solberg LC, Turek FW. Entrainment in calorie-restricted mice: conflicting zeitgebers and free-running conditions. Am J Physiol 1998;274:R1751–61. [13] Mistlberger RE. Circadian food-anticipatory activity: formal models and physiological mechanisms. Neurosci Biobehav Rev 1994;18:171–95. [14] Weber DN, Spieler RE. Effects of the light–dark cycle and scheduled feeding on behavioral and reproductive rhythms of the cyprinodont fish, Medaka, Oryzias atipes. Experientia 1987;43:621–4. [15] Levine JD, Funes P, Dowse HB, Hall JC. Resetting the circadian clock by social experience in Drosophila melanogaster. Science 2002;298:2010–2. [16] Mrosovsky N. Phase response curves for social entrainment. J Comp Physiol A 1988;162:35–46. [17] Menaker M, Eskin A. Entrainment of circadian rhythms by sound in Passer domesticus. Science 1966;154:1579–81. [18] Mrosovsky N. Locomotor activity and non-photic influences on circadian clocks. Biol Rev Camb Philos Soc 1996;71:343–72. [19] Steinlechner S, Stieglitz A, Ruf T. Djungarian hamsters: a species with a labile circadian pacemaker? Arrhythmicity under a light–dark cycle induced by short light pulses. J Biol Rhythms 2002;17:248–58. [20] Mrosovsky N. Double-pulse experiments with nonphotic and photic phase-shifting stimuli. J Biol Rhythms 1991;6:167–79. [21] Duffy JF, Kronauer RE, Czeisler CA. Phase-shifting human circadian rhythms: influence of sleep timing, social contact and light exposure. J Physiol 1996;495:289–97. [22] Knoch ME, Gobes SM, Pavlovska I, Su C, Mistlberger RE, Glass JD. Short-term exposure to constant light promotes strong circadian phaseresetting responses to nonphotic stimuli in Syrian hamsters. Eur J Neurosci 2004;19:2779–90. [23] Murphy HM, Wideman CH, Nadzam GR. A laboratory animal model of human shift work. Integr Physiol Behav Sci 2003;38:316–28. [24] Mrosovsky N, Salmon PA, Vrang N. Revolutionary science: an improved running wheel for hamsters. Chronobiol Int 1998;15:147–58. [25] Smith KA, Schoen MW, Czeisler CA. Adaptation of human pineal melatonin suppression by recent photic history. J Clin Endocrinol Metab 2004;89:3610–4. [26] Sakamoto K, Liu C, Kasamatsu M, Pozdeyev NV, Iuvone PM, Tosini G. Dopamine regulates melanopsin mRNA expression in intrinsically photosensitive retinal ganglion cells. Eur J Neurosci 2005;22:3129–36. [27] Moog R, Hildebrandt G. Adaptation to shift work-experimental approaches with reduced masking effects. Chronobiol Int 1989;6:65–75. [28] Barger LK, Wright Jr KP, Hughes RJ, Czeisler CA. Daily exercise facilitates phase delays of circadian melatonin rhythm in very dim light. Am J Physiol Regul Integr Comp Physiol 2004;286:R1077–84. [29] Buxton OM, Lee CW, L'Hermite-Baleriaux M, Turek FW, Van Cauter E. Exercise elicits phase shifts and acute alterations of melatonin that vary with circadian phase. Am J Physiol 2003;284:R714–24. [30] Peirson SN, Thompson S, Hankins MW, Foster RG. Mammalian photoentrainment: results, methods, and approaches. Methods Enzymol 2005;393:697–726. [31] Bobrzynska KJ, Mrosovsky N. Phase shifting by novelty-induced running: activity dose–response curves at different circadian times. J Comp Physiol A 1998;182:251–8. [32] Honma K, Hashimoto S, Nakao M, Honma S. Period and phase adjustments of human circadian rhythms in the real world. J Biol Rhythms 2003;18:261–70. [33] Cho K. Chronic ‘jet lag’ produces temporal lobe atrophy and spatial cognitive deficits. Nat Neurosci 2001;4:567–8.