Physiology& Behavior,Vol. 50, pp. 373-378. ©PergamonPress plc, 1991. Printedin the U.S.A.
0031-9384/91 $3.00 + .00
Influence of Running Wheel Activity on Free-Running Sleep/Wake and Drinking Circadian Rhythms in Mice DALE M. EDGAR,1 T H O M A S S. KILDUFF, CONNIE E. M A R T I N AND W I L L I A M C. D E M E N T
Sleep Research Center, Department of Psychiatry & Behavioral Sciences Stanford University School of Medicine, Stanford, CA 94305 Received 16 January 1991 EDGAR, D. M., T. S. KILDUFF, C. E. MARTIN AND W. C, DEMENT. Influence of running wheel activity on free-running sleep~wake and drinking circadian rhythms in mice. PHYSIOL BEHAV 50(2) 373-378, 1991.--Previous studies have indicated that manipulationof activity levels can modify characteristics of sleep/wake and activity rhythms. The generalityof these observations was evaluated by simultaneouslymeasuringdrinking and sleep/wake rhythms while mice had free or no access to a running wheel in constant conditions (DD). Robust circadian rhythms in all parameters were observed in the "wheel free" (unrestricted) condition. When wheels were locked, the peak amplitude of the sleep/wake circadian rhythm decreased by approximately 50% without affecting the amplitude of the drinking rhythm. Total wake time decreased 11% per circadian day when wheels were locked with increases in both NREM and REM sleep. Whereas the amplitude of the drinking waveform was unaffected, wheel restriction caused an equivalent increase in period length ('r) for both rhythms. These results indicate that, unlike the generalized effects of activity on "r, activity restrictioninfluenceson rhythm amplitudedo not generalizeto all behavioraland/or physiological variables. This work also supports the notion that activity influenceson sleep/wake rhythm amplitude reflect behavioral "maaking" rather than a fundamentalchange in the direct couplingmechanismsof the biologicalclock. Activity
Exercise
Drinking
Circadianrhythm
Sleep
THE behavior and physiology of mammals are thought to be temporally organized through a biological clock located in the suprachiasmatic nuclei (SCN) of the hypothalamus. The SCN generate circadian rhythms which are revealed when animals are exposed to constant environmental (free-running) conditions. Lesions of the SCN disrupt circadian rhythms of activity, sleep/ wakefulness, drinking and other physiological and neurochemical variables. These observations, along with evidence for in vivo and in vitro electrophysiological, metabolic, and transplant studies, have supported the notion that the SCN is the primary pacemaker in the circadian timekeeping system of mammals (13,17). Circadian rhythms can be influenced by a number of periodic environmental cues including photoperiod, temperature, and food availability. Environmental factors influence the timing and observed waveform of circadian rhythms in two general ways. Tonic effects can result in modification of the amplitude or period of an observed rhythm (3). Environmental cues can also exhibit phasic effects either by acting as a zeitgeber to cause entrainment of the observed rhythm to the period of the external oscillation (2), or by superimposing modulations on the underlying endogenous oscillation (5, 6, 9). Passively driven oscillations in the manifest circadian rhythm (e.g., modulations in the level of physiological or behavioral variables which are not the product of direct circadian control) are generally referred to as "masking" effects.
Masking
Mice (Mus musculus)
Although photic input has been the most thoroughly documented zeitgeber in mammals, an increasing amount of evidence suggests that activity itself can influence the expression of circadian rhythms. Wever (24) demonstrated that as much as 50% of the human body temperature rhythm amplitude could be ascribed to activity-dependent masking, and activity restriction can markedly fragment sleep/wake patterns in the mouse (23). In addition, induced locomotion can produce phase shifts of activity rhythms in hamsters (14) as can activity induced by pharmacological treatment (15,21). Opportunities to exercise in running wheels has also been shown to influence free-running rhythm period (4, 11, 25). Recent evidence also suggests that providing mice scheduled opportunities for voluntary access to a running wheel can entrain circadian rhythms (10). These observations suggest that the expression of an overt rhythm may feed back onto the biological clock and alter subsequent timing of that rhythm. The extent of activity-dependent masking and the generality of these effects on circadian rhythms have not been thoroughly investigated. The purpose of the present study was to evaluate the extent of activity-dependent masking and the generality of these effects on free-running circadian rhythms by examining two biological variables, the circadian rhythms of drinking and sleep/wake, while the opportunity for exercise was either freely available or restricted. We find that activity restriction can af-
~Requests for reprints should be addressed m Dale M. Edgar, Ph.D., Sleep Research Center, Stanford University School of Medicine, 701 Welch Rd. #2226, Palo Alto, CA 94304. 373
374
EDGAR, KILDUFF, MARTIN AND DEMENT
fect the amplitude of the free-running sleep/wake circadian rhythm without affecting the drinking rhythm, suggesting that its manipulation may result in differential masking of sleep/wakefulness. METHOD
Experimental Animals and Housing Eight male mice (Mus musculus; C57BL/6Nnia) age 3-6 months were used in this study. Each animal was individually housed in a polycarhonate cage (Nalgene 20 x 34 cm floor) equipped with a continuous watering spout (Lixit), a microisoladon filter top, and an exercise wheel. Cages were located within separately ventilated and sound attenuated compartments of an animal recording chamber. Throughout the study the animals were maintained in constant darkness (DD), ambient temperature was 24-26°C, and food and water were available ad lib. Animal health was assessed daily using EEG, sleep/wake, drinking, and/or activity indices recorded by a computerized sleep/ wake and physiological monitoring system (see the Data Collection section). Animals were visually inspected and cages changed under dim red illumination once a week.
Surgical Preparations At 3-4 months of age, mice were anesthetized (sodium pentobarbital, 70 mg/kg) and surgically prepared with a cranial implant that permitted continuous electroencephalogram (EEG) and electromyogram (EMG) recording using procedures previously described (20). Briefly, stainless steel screws (#00X1/8, J. I. Morris Co.) were positioned bilaterally in the rostral margin of the frontal bone and caudal margin of the parietal bone and served as epidural cortical recording electrodes. EMG electrodes made of flexible teflon-coated stainless steel wires were positioned subcutaneously on the neck musculature. All leads were soldered to a shielded multiconductor recording cable which passed through the center of a 4 mm diameter × 8 mm polystyrene dowel which was affixed to the skull with cyanoacrylate and dental acrylic. A connector at the distal end of the recording cable was attached to a low torque electrical commutator (Biela Engineering, Irvine, CA) mounted at the top of the cage, permitting mice freedom to move throughout the cage. One month of postoperative recovery was permitted prior to study.
Data Collection Sleep and waking states, drinking, and running wheel activity were continuously monitored from animals using SCORE, a microcomputer based automated sleep/wake and physiological monitoring system. A comprehensive description of the sleep scoring algorithm and the performance of this system (agreement between automated scoring results and human assessments of mouse polygraph recordings) are thoroughly described elsewhere (20). EEG used for automated arousal state determinations was monitored using bipolar electrodes positioned at the frontal and parieto-occipital cortices. EEG input was amplified 10,000 times and bandpass filtered (1.0-30 Hz) using a Grass Model 12 Neurodata amplification system (Grass Instrument Co., Quincy, MA). EMG input was amplified 20,000 times, and was integrated using an RMS integrator (Barrows Co., Woodside, CA). Sleep/ wake determinations were performed every 10 s and scored as either awake (wake), non-REM sleep, REM sleep, or thetadominated wake (TDW). Drinking was detected using a low current (
0031-9384/91 $3.00 + .00
Influence of Running Wheel Activity on Free-Running Sleep/Wake and Drinking Circadian Rhythms in Mice DALE M. EDGAR,1 T H O M A S S. KILDUFF, CONNIE E. M A R T I N AND W I L L I A M C. D E M E N T
Sleep Research Center, Department of Psychiatry & Behavioral Sciences Stanford University School of Medicine, Stanford, CA 94305 Received 16 January 1991 EDGAR, D. M., T. S. KILDUFF, C. E. MARTIN AND W. C, DEMENT. Influence of running wheel activity on free-running sleep~wake and drinking circadian rhythms in mice. PHYSIOL BEHAV 50(2) 373-378, 1991.--Previous studies have indicated that manipulationof activity levels can modify characteristics of sleep/wake and activity rhythms. The generalityof these observations was evaluated by simultaneouslymeasuringdrinking and sleep/wake rhythms while mice had free or no access to a running wheel in constant conditions (DD). Robust circadian rhythms in all parameters were observed in the "wheel free" (unrestricted) condition. When wheels were locked, the peak amplitude of the sleep/wake circadian rhythm decreased by approximately 50% without affecting the amplitude of the drinking rhythm. Total wake time decreased 11% per circadian day when wheels were locked with increases in both NREM and REM sleep. Whereas the amplitude of the drinking waveform was unaffected, wheel restriction caused an equivalent increase in period length ('r) for both rhythms. These results indicate that, unlike the generalized effects of activity on "r, activity restrictioninfluenceson rhythm amplitudedo not generalizeto all behavioraland/or physiological variables. This work also supports the notion that activity influenceson sleep/wake rhythm amplitude reflect behavioral "maaking" rather than a fundamentalchange in the direct couplingmechanismsof the biologicalclock. Activity
Exercise
Drinking
Circadianrhythm
Sleep
THE behavior and physiology of mammals are thought to be temporally organized through a biological clock located in the suprachiasmatic nuclei (SCN) of the hypothalamus. The SCN generate circadian rhythms which are revealed when animals are exposed to constant environmental (free-running) conditions. Lesions of the SCN disrupt circadian rhythms of activity, sleep/ wakefulness, drinking and other physiological and neurochemical variables. These observations, along with evidence for in vivo and in vitro electrophysiological, metabolic, and transplant studies, have supported the notion that the SCN is the primary pacemaker in the circadian timekeeping system of mammals (13,17). Circadian rhythms can be influenced by a number of periodic environmental cues including photoperiod, temperature, and food availability. Environmental factors influence the timing and observed waveform of circadian rhythms in two general ways. Tonic effects can result in modification of the amplitude or period of an observed rhythm (3). Environmental cues can also exhibit phasic effects either by acting as a zeitgeber to cause entrainment of the observed rhythm to the period of the external oscillation (2), or by superimposing modulations on the underlying endogenous oscillation (5, 6, 9). Passively driven oscillations in the manifest circadian rhythm (e.g., modulations in the level of physiological or behavioral variables which are not the product of direct circadian control) are generally referred to as "masking" effects.
Masking
Mice (Mus musculus)
Although photic input has been the most thoroughly documented zeitgeber in mammals, an increasing amount of evidence suggests that activity itself can influence the expression of circadian rhythms. Wever (24) demonstrated that as much as 50% of the human body temperature rhythm amplitude could be ascribed to activity-dependent masking, and activity restriction can markedly fragment sleep/wake patterns in the mouse (23). In addition, induced locomotion can produce phase shifts of activity rhythms in hamsters (14) as can activity induced by pharmacological treatment (15,21). Opportunities to exercise in running wheels has also been shown to influence free-running rhythm period (4, 11, 25). Recent evidence also suggests that providing mice scheduled opportunities for voluntary access to a running wheel can entrain circadian rhythms (10). These observations suggest that the expression of an overt rhythm may feed back onto the biological clock and alter subsequent timing of that rhythm. The extent of activity-dependent masking and the generality of these effects on circadian rhythms have not been thoroughly investigated. The purpose of the present study was to evaluate the extent of activity-dependent masking and the generality of these effects on free-running circadian rhythms by examining two biological variables, the circadian rhythms of drinking and sleep/wake, while the opportunity for exercise was either freely available or restricted. We find that activity restriction can af-
~Requests for reprints should be addressed m Dale M. Edgar, Ph.D., Sleep Research Center, Stanford University School of Medicine, 701 Welch Rd. #2226, Palo Alto, CA 94304. 373
374
EDGAR, KILDUFF, MARTIN AND DEMENT
fect the amplitude of the free-running sleep/wake circadian rhythm without affecting the drinking rhythm, suggesting that its manipulation may result in differential masking of sleep/wakefulness. METHOD
Experimental Animals and Housing Eight male mice (Mus musculus; C57BL/6Nnia) age 3-6 months were used in this study. Each animal was individually housed in a polycarhonate cage (Nalgene 20 x 34 cm floor) equipped with a continuous watering spout (Lixit), a microisoladon filter top, and an exercise wheel. Cages were located within separately ventilated and sound attenuated compartments of an animal recording chamber. Throughout the study the animals were maintained in constant darkness (DD), ambient temperature was 24-26°C, and food and water were available ad lib. Animal health was assessed daily using EEG, sleep/wake, drinking, and/or activity indices recorded by a computerized sleep/ wake and physiological monitoring system (see the Data Collection section). Animals were visually inspected and cages changed under dim red illumination once a week.
Surgical Preparations At 3-4 months of age, mice were anesthetized (sodium pentobarbital, 70 mg/kg) and surgically prepared with a cranial implant that permitted continuous electroencephalogram (EEG) and electromyogram (EMG) recording using procedures previously described (20). Briefly, stainless steel screws (#00X1/8, J. I. Morris Co.) were positioned bilaterally in the rostral margin of the frontal bone and caudal margin of the parietal bone and served as epidural cortical recording electrodes. EMG electrodes made of flexible teflon-coated stainless steel wires were positioned subcutaneously on the neck musculature. All leads were soldered to a shielded multiconductor recording cable which passed through the center of a 4 mm diameter × 8 mm polystyrene dowel which was affixed to the skull with cyanoacrylate and dental acrylic. A connector at the distal end of the recording cable was attached to a low torque electrical commutator (Biela Engineering, Irvine, CA) mounted at the top of the cage, permitting mice freedom to move throughout the cage. One month of postoperative recovery was permitted prior to study.
Data Collection Sleep and waking states, drinking, and running wheel activity were continuously monitored from animals using SCORE, a microcomputer based automated sleep/wake and physiological monitoring system. A comprehensive description of the sleep scoring algorithm and the performance of this system (agreement between automated scoring results and human assessments of mouse polygraph recordings) are thoroughly described elsewhere (20). EEG used for automated arousal state determinations was monitored using bipolar electrodes positioned at the frontal and parieto-occipital cortices. EEG input was amplified 10,000 times and bandpass filtered (1.0-30 Hz) using a Grass Model 12 Neurodata amplification system (Grass Instrument Co., Quincy, MA). EMG input was amplified 20,000 times, and was integrated using an RMS integrator (Barrows Co., Woodside, CA). Sleep/ wake determinations were performed every 10 s and scored as either awake (wake), non-REM sleep, REM sleep, or thetadominated wake (TDW). Drinking was detected using a low current (
