MS 213.qxp

MS 213.qxp

3/14/2005

1:56 PM

Page 395

CIRCADIAN RHYTHMS

Deletion of the Mammalian Circadian Clock Gene BMAL1/Mop3 Alters Baseline Sleep Architecture and the Response to Sleep Deprivation Aaron Laposky, PhD1; Amy Easton, PhD1; Christine Dugovic, PhD1; Jacqueline Walisser, PhD2; Christopher Bradfield, PhD2; Fred Turek, PhD1 1Neurobiology

and Physiology, Northwestern University, Evanston, IL; 2McArdle Center for Cancer Research, University of Wisconsin, Madison, WI

Study objectives: The finding that deletion or mutation of core circadian clock genes in both mice and flies induce unexpected alterations in sleep amount, sleep architecture and the recovery response to sleep deprivation, has led to new insights into functions of the circadian system that extend beyond its role as a regulator of the timing of the sleep-wake cycle. A key transcription factor in the transcriptional/translational feedback loop of mammalian circadian genes is BMAL1/Mop3, a heterodimeric partner to CLOCK. It was previously shown that mice deficient in the BMAL1/Mop3 gene become immediately arrhythmic in constant darkness and have reduced locomotor activity levels under entrained and constant conditions. In this study, we tested the hypothesis that the mammalian BMAL1/Mop3 gene would have regulatory effects on sleep-wake patterns. Design: In mice with targeted deletion of the BMAL1/Mop3 gene, EEG/EMG sleep-wake patterns were recorded under entrained and free-

running conditions as well as following acute (6-hrs) sleep deprivation. Measurements and results: Mice homozygous for the BMAL1/Mop3 deletion showed an attenuated rhythm of sleep and wakefulness distribution across the 24-hr period. In addition, these mice showed increases in total sleep time, sleep fragmentation and EEG delta power under baseline conditions, and an attenuated compensatory response to acute sleep deprivation. Conclusions: These new data strengthen the hypothesis that molecular components of the circadian system play a central role in the generation of sleep and wakefulness beyond just the timing of these behavioral vigilance states. Key Words: Mice, sleep, BMAL1/Mop3, clock, circadian genes, flies Citation: Laposky A; Easton A; Dugovic C. et al. Deletion of the mammalian circadian clock gene BMAL1/Mop3 alters baseline sleep architecture and the response to sleep deprivation. SLEEP 2005;28(4):395-409.

INTRODUCTION

servation of molecular components of the clock between species as diverse as fruit flies and mice.8,9 The discovery of circadian genes and the use of animal genetic models have been instrumental in understanding how molecular components of the circadian clock influence the multiple output rhythmic systems under their control. In particular, recent studies in both mice and flies have found that deletion or mutation in core clock genes induces unexpected alterations in sleep amount, sleep architecture, and compensatory responses to sleep deprivation.10-13 These results have led to the hypothesis that circadian clock genes may also be central to sleep-regulatory processes beyond just the circadian timing of sleep. For example, mutation in the mammalian circadian gene, Clock (mClock), induces a decrease in non-rapid eye movement (NREM) sleep time and an attenuated rapid eye movement (REM) recovery following sleep deprivation.10 A mutation in the fly homolog of this gene, dClock, reduces consolidated rest time and leads to alterations in the response to rest deprivation.11,13 Cryptochrome (Cry)1,2 doubleknockout mice have increases in baseline amounts of NREM sleep and consolidation of NREM episodes and NREM delta power over wild-type control levels, and they lack the normal compensatory response in sleep amount and NREM delta power following sleep deprivation.12 Interestingly, a mutation in the drosophila gene, timeless, which may function in a similar manner as the mammalian Cry gene, leads to an impaired recovery response to short-term rest deprivation.11,13 In mice, CLOCK dimerizes with BMAL1 (also called Mop3), while in flies, CLOCK combines with CYCLE to form key transcription factors controlling the expression of downstream circadian genes.8,9 BMAL1/Mop3 and dCycle are structural homologues, and deletion of these genes leads to arrhythmicity of locomotor activity in both species.14 The recent finding that deletion of dCycle also leads to a decrease in consolidated rest time and altered recovery responses to rest deprivation raises the possibility that the func-

THE BILATERALLY PAIRED SUPRACHIASMATIC NUCLEI (SCN) OF THE ANTERIOR HYPOTHALAMUS REPRESENT THE MASTER CIRCADIAN PACEMAKER THAT regulates most, if not all, 24-hour rhythms at the cellular, tissue, system, and behavioral levels in mammals.1,2 A critical role of the circadian system has been demonstrated in the control of sleep and wakefulness, whereby the circadian-timing signal interacts with a sleep-regulatory process to determine the propensity, duration, and intensity of sleep.3 It has been proposed that the mechanisms underlying the circadian and sleep-regulatory processes are completely independent, in that SCN lesions abolish the sleep-wake rhythm but leave 2 important measures of sleep control unaltered: the daily amount of sleep and the compensatory response to sleep deprivation.4-6 Based on data from the SCN-ablation model in rats, early reports suggested that the circadian clock only provides a time cue for sleep-wake expression, but serves little, if any, direct functional role in controlling the amount or homeostatic drive for sleep. The astonishing progress in uncovering mammalian circadian clock genes has led to the discovery of transcriptional/translational molecular feedback loops that form the core of endogenous rhythmicity within the cell.7 There are remarkable parallels in the conDisclosure Statement This was not an industry supported study. Drs. Laposky, Easton, Dugovic, Walisser, Bradfield, and Turek have indicated no financial conflicts of interest. Submitted for publication June 2004 Accepted for publication January 2005 Address correspondence to: Aaron Laposky, PhD, Neurobiology and Physiology, Northwestern University, 2205 Tech Drive Hogan 2-160, Evanston, IL, 60208-3520; Tel: (847) 467-7698; Fax: (847) 467-4065; E-mail: [email protected] SLEEP, Vol. 28, No. 4, 2005

395

Deletion of BMAL1/Mop3 and Sleep—Laposky et al

MS 213.qxp

3/14/2005

1:56 PM

Page 396

stricted movement throughout the recording cage. After 5 days of adaptation to the recording environment, a 48-hour baseline EEG/EMG recording was collected. Next, animals were manually sleep deprived for 6 hours during the last half of the light phase (noon-6:00 PM) by a gentle-handling procedure. To keep the animals awake, an experimenter observed EEG/EMG recordings for signs of sleep and then used a progression of stimuli (cage tapping, cage shaking, gentle contact) to awaken the animal. Beginning at dark onset (6:00 PM), sleep deprivation was stopped, and EEG/EMG recordings continued during an 18-hour recovery opportunity, consisting of the 12-hour dark phase (6:00 PM-6:00 AM) and the following 6-hour light phase (6:00 AM-noon).

tion of BMAL1/Mop3 extends beyond circadian control and is also involved in the regulation of the amount and homeostatic drive for sleep in mammals.13,15 To test this hypothesis, we monitored the sleep-wake cycle in mice with deletion of the BMAL1/Mop3 gene under both entrained and free-running conditions. METHODS The generation of C57Bl/6J mice carrying a null allele for the BMAL1/Mop3 gene has been previously described.14 In the present study, male (3 months of ages), wild-type (BMAL1/Mop3+/+), heterozygous (BMAL1/Mop3+/-) and homozygous (BMAL1/Mop3-/-) mice were used. All protocols and procedures were approved the Northwestern University Animal Care and Use Committee.

Infrared Activity Monitoring After sleep recordings in 12L:12D, animals remained in entrained conditions and underwent activity monitoring using infrared motion sensors (A1 Securing and Electrical Ltd., Huyton Merseyside, England). Activity was monitored for 10 to 14 days in 12L:12D followed immediately by monitoring for 10 to 14 days in constant darkness using ClockLab (Actimetrics; Evanston, Ill) software. The free-running period and the onset for the activity phase of the circadian cycle were identified for each animal.

Overall Procedure The general procedures in this experiment were conducted in the following order: (1) surgical implantation of electroencephalogram (EEG)/electromyogram (EMG) electrodes and body-temperature transmitters, (2) baseline and sleep-deprivation recordings in entrained 12-hour light and 12-hour dark (12L:12D) conditions, (3) locomotor activity monitoring for 10 to 14 days in 12L:12D conditions, (4) locomotor activity monitoring for 10 to 14 days in constant darkness, and (5) baseline sleep recordings in constant darkness. The same mice were used throughout the experiment. Male mice were used in this study primarily because this is the standard sex in rodent sleep studies. Also, we wanted to compare our data to previous sleep studies using circadian mutant mice, which have all used male mice. The details of these procedures are provided below.

Sleep Recording Under Constant Conditions In this phase of the experiment, 1 BMAL1/Mop3+/+, 1 BMAL1/Mop3+/-, and 3 BMAL1/Mop3-/- were excluded from the study because of deteriorating EEG/EMG signals. Therefore, the sample sizes were 7, 6, and 6 for these genotypes, respectively. Following locomotor-activity recordings in constant darkness, 48-hours of baseline EEG/EMG data were collected in constant darkness beginning at the onset of the activity phase for each animal. Sleep data were expressed and analyzed in circadian hours; therefore, an animal with an activity period of, for example, 23.7 hours had sleep times expressed as a percentage of 23.7 hours. Because BMAL1/Mop3-/- mice were arrhythmic in constant conditions, their sleep data was based on a 24-hour period.

Surgical Procedures At 3 months of age, BMAL1/Mop3+/+ (n = 8, 30.2 ± 0.8 g), BMAL1/Mop3+/- (n = 7, 29.8 ± 1.0 g), and BMAL1/Mop3-/- (n = 9, 27.1 ± 1.4 g) mice were anesthetized (ketamine 80mg/kg and xylazine 8mg/kg) and surgically implanted with EEG and EMG electrodes for polysomnographic recording.10 For monitoring EEG signals, stainless-steel screw electrodes (Small Parts, Inc. Miami Lakes, Fla #000-120) were positioned at 2 locations contralateral to each other on the skull surface (1.0 mm anterior to Bregma/0.5 right of the central suture and 0.5 mm posterior to lambda/1.0 mm left of the central suture). EMG activity was monitored using stainless-steel, Teflon-coated wires placed bilaterally in the nuchal muscle. All of the electrodes were connected to a prefabricated 1- x 4pin grid-array head implant (Plastics One, Ronake, Virg) that was secured to the skull using cyanoacrylimide adhesive. Finally, an implantable transducer (PDT-4000 E-Mitter by Mini-Mitter, Inc., Bend, Ore.) was placed in the peritoneal cavity to record body temperature. Buprinex (buprenorphrine, 2 mg/kg subcutaneously) was administered for 12 to 24 hours following the operation to control for pain during the recovery process.

Data Analysis of Sleep-Wake Recordings EEG signals were amplified 10,000 times with -6dB/oct highpass, and low-pass filters were set at 1.0 and 30.0 Hz, respectively. EMG signals were amplified 5000 times, with high- and low-pass filters at 30 and 100 Hz. Both signals were digitized at 100 Hz per channel by an analog-to-digital converter (model DT01EZ; Data Translation Inc., Marlboro, Mass) and stored on an IBM AT-compatible computer, using specialized software for acquiring and processing sleep data in rodents (Multilevel; Actimetrics, Evanston, Ill). Each 10-second epoch of the EEG/EMG tracing was assigned a score, based on visual inspection of either wake, NREM or REM sleep, using previously described criteria.10 Postscoring analysis, using custom-designed software (SleepReport, Actimetrics, Evanston, Ill) allowed for determination of sleep-structure parameters, including amount, distribution, consolidation, and EEG spectral analysis. The distribution of sleep-wake was assessed by averaging and graphing data in 2-hour intervals across the recording period and by determining the ratio of sleep-wake amounts between the 12-hour light and 12-hour dark periods (L:D ratio). Sleep consolidation was determined by the number of arousals (at

Sleep Recording During Entrained Conditions Following 2 weeks of recovery from surgery, animals were housed individually in sleep-recording cages, under light(12L:12D, lights on 6:00 AM), temperature- (23.0°C ± 1.0°C), and sound-controlled conditions with free access to food and water. The animals were connected to a wire tether/commutator system (Plastics One, Roanoke, Virg) for the collection of EEG/EMG signals. This swivel system allowed the animal unreSLEEP, Vol. 28, No. 4, 2005

396


MS 213.qxp

3/14/2005

1:56 PM

Page 397

least 2 consecutive sleep epochs interrupted by at least 2-consecutive wake epochs), stage shifts (number of transitions between 10second epochs of wake, NREM, and REM), sleep-wake bouts (at least 20 seconds of the respective stage until interruption by 20 consecutive seconds of another stage), and the average duration (minutes) of individual sleep-wake bouts. For quantitative analysis of the EEG signal, each 10-second scoring epoch was divided into five 2-second intervals and subjected to fast Fourier transformation, which included a range of 1 to 25 Hz with a frequency resolution of 0.1 Hz. For all epochs of NREM sleep, the EEG power in the 1.0 to 4.0, 6.0 to 10.0, and 11.0 to 15.0 frequency ranges (Hz) were calculated, yielding measurements of NREM delta, theta, and sigma power, respectively. Any 10-second epoch visually identified to contain EEG artifact was eliminated from quantitative analysis. By convention, we used NREM delta power as a quantitative measure of sleep homeostatic drive. In order to compare genotypes on the circadian distribution of NREM delta power, values were normalized for each animal and expressed as a percentage of the individual 24-hour mean value. In addition, an analysis of EEG activity was conducted by comparing genotypes on the absolute levels of delta, theta, and sigma power in all sleep-wake states.

BMAL1/Mop3+/- mice, as depicted in Figure 1. The body temperature rhythm was significantly attenuated in BMAL1/Mop3-/mice; however, there was no overall 24-hour body-temperature difference between genotypes (Figure 1). Inaddition, BMAL1/Mop3-/mice did not show an increase in wakefulness or body temperature at the transition to dark onset compared to the normal pattern of entrainment demonstrated by BMAL1/Mop3+/+ and BMAL1/Mop3+/- mice (Figure 1). In BMAL1/Mop3-/- mice, individual sleep-wake cycles were less consolidated, as indicated by a high level of sleep fragmentation. The number of arousals from sleep, stage shifts, and the number of NREM and REM sleep bouts were significantly elevated in BMAL1/Mop3-/- mice, due to significant increases during the dark phase (Table 1). In association with the increased number of bouts, BMAL1/Mop3-/- mice had overall reduced REM-bout durations, and a trend for decreased NREM-bout durations (Table 1). Sleep Time is Increased in BMAL1/Mop3-/- Mice Over the 24-hour baseline period, total sleep time differed between genotypes (F = 11.1, P < .001), with BMAL1/Mop3-/sleeping 6.3% (+ 90.7 minutes) more than BMAL1/Mop3+/+ and 5.6% (+ 80.6 minutes) more than BMAL1/Mop3+/- mice (Figure 1). The increased total sleep time in BMAL1/Mop3-/- mice resulted from elevated NREM (F = 6.3, P < .01) and REM (F = 8.2, P < .01) sleep during the 12-hour dark phase and a nonsignificant decrease in sleep times during the 12-hour light phase (Figure 1).

Body Temperature Monitoring Body temperature was monitored simultaneously with EEG/EMG recordings. Body temperature measurements were taken from a transducer surgically implanted in the peritoneal cavity (PDT-4000 E-Mitter by Mini-Mitter). These biotelemetry transducers were pre-calibrated to emit radiofrequency signals with an accuracy of 0.1°C. The transducers were powered by an induction coil, and output signals were detected by a radiofrequency receiver placed under each mouse cage. A body-temperature measurement was collected every 10 seconds. Data were collected using a software package developed in our laboratory (Multilevel, Actimetrics, Evanston, Ill).

NREM Delta Power Is Increased in BMAL1/Mop3-/- Mice The diurnal pattern of sleep drive was determined by normalizing NREM delta power for each baseline 2-hour interval as a percentage of the 24-hour NREM delta power (Figure 2). NREM delta power represents a quantitative EEG measure of sleep intensity or sleep drive. In BMAL1/Mop3+/+ and BMAL1/Mop3+/- mice, NREM delta power was elevated at light onset and showed the expected decline across this phase as sleep time accumulated. NREM delta power showed a typical elevation during the dark phase in association with higher levels of wakefulness. In sharp contrast, BMAL1/Mop3-/- mice lacked the normal diurnal pattern of sleep drive as evidenced by the flat distribution of NREM delta power and a number of significant genotype differences across the recording period (Figure 2). Additional analyses were performed to determine the absolute values of delta (1-4 Hz), theta (6-10 Hz), and sigma (11-15 Hz) activity in each sleep-wake state (Figure 3). Values were averaged over the 12-hour light and 12-hour dark phase for each genotype. In BMAL1/Mop3-/- mice, absolute delta power was significantly elevated in NREM and REM sleep in both the light and dark phases compared to the other genotypes. These were the only measures in which genotype differences occurred (Figure 3). Therefore, the changes in NREM and REM delta power in BMAL1/Mop3-/- were not accompanied by alterations in the waking delta activity or theta and sigma power.

Statistical Analyses A 1-way analysis of variance was used for all comparisons involving the 3 genotype groups. Posthoc testing (Tukey HSD) was used to follow-up analysis of variance main effects, when indicated. In addition, dependent sample t tests were utilized for some analyses of sleep deprivation and recovery data. Significance levels were set at P < .05 for all comparisons. There were no differences in sleep between the first and second baseline recordings for any genotype; therefore, the second baseline recording was used for data analyses. RESULTS Twenty-Four Hour Baseline During Entrained Conditions Sleep Consolidation is Attenuated in BMAL1/Mop3-/- Mice In BMAL1/Mop3+/+ and BMAL1/Mop3+/- mice, sleep was concentrated to the light phase and wakefulness to the dark phase, reflecting a normal consolidation pattern under entrained conditions (Figure 1). In contrast, BMAL1/Mop3-/- mice showed an attenuated distribution of sleep-wakefulness across the 24hour period, with significantly reduced L:D ratios for wake (F = 48.4, P < .001), NREM (F = 30.1, P < .001), and REM (F = 10.3, P < .001) amounts compared to BMAL1/Mop3+/+ and SLEEP, Vol. 28, No. 4, 2005

Sleep Deprivation Animals were sleep deprived for 6 hours during the second half of the light phase (noon-6:00 PM). At dark onset (6:00 PM), sleep deprivation ended, and animals were allowed to sleep ad lib for the next 18 hours, which we refer to as the recovery period, as 397


MS 213.qxp

3/14/2005

1:56 PM

Page 398

of REM sleep lost during this 6-hour interval was greater in BMAL1/Mop3+/+ (26.9 ± 1.6 minutes) and BMAL1/Mop3+/- (23.7 ± 1.5 minutes) compared to BMAL1/Mop3-/- (17.7 ± 2.3 minutes) mice (F = 5.5, P < .05). Similarly, for NREM sleep, the overall amount lost during deprivation was greater in BMAL1/Mop3+/+ (184.2 ± 10.8 minutes) and BMAL1/Mop3+/- (179.9 ± 6.9 min-

depicted in Figure 4. Six-Hour Sleep-Deprivation Period During the 6-hour sleep-deprivation period, all genotypes were completely deprived of REM sleep (Figure 4). However, due to different baseline characteristics between genotypes, the amount

Figure 1—Baseline sleep patterns in BMAL1/Mop3+/+ (n = 8, open), BMAL1/Mop3+/- (n = 7, grey), and BMAL1/Mop3-/- (n = 9, black) mice under entrained (12-hour light:12-hour dark) conditions. Left: Sleep-wake amounts and body temperature averaged (mean ± SEM) in 2-hour intervals. The dark phase is indicated on the x-axis. Right: Sleep-wake amounts and body temperature averaged (mean ± SEM) over the 12-hour light, 12-hour dark, and total 24-hour recording period. Between-group comparisons were made at each time interval using 1-way analysis of variance. To indicate significance of posthoc comparisons (Tukey HSD), asterisks left of the hash mark (*/) represent BMAL1/Mop3-/- vs BMAL1/Mop3 +/+, and those to the right of the hash mark (/*) represent BMAL1/Mop3-/- vs BMAL1/Mop3+/- (* P < .05, ** P < .01, *** P < .001). NREM refers to non-rapid eye movement sleep; REM, rapid eye movement.

SLEEP, Vol. 28, No. 4, 2005

398


MS 213.qxp

3/14/2005

1:56 PM

Page 399

genotypes (F = 5.5, P < .01), being significantly smaller in BMAL1/Mop3-/- mice compared to BMAL1/Mop3+/+ (P < .01) and BMAL1/Mop3+/- (P < .05) mice (Figure 5). In order to look at more cumulative rebound effects, the recovery period was divided into the following intervals, hours 1 to 6 (first half of dark phase), 7 to 12 (second half of dark phase), 13 to 18 (first half of light phase), 1 to 12 (entire dark period), and 1 to 18 (entire recovery period), as shown in Figure 6. In this analysis, we plotted recovery sleep as a percentage change of corresponding baseline levels for each time interval. First, we determined for each genotype whether recovery sleep at each time interval significantly differed from corresponding baseline levels (within-group comparisons). Second, we made genotype comparisons of the recovery response at each time interval (betweengroup comparisons). During hours 1 to 6 of the recovery period, all genotypes had significantly more NREM sleep compared to baseline levels (Figure 6). The degree to which NREM sleep increased over baseline levels was smaller but not statistically different in BMAL1/Mop3-/- mice compared to the other genotypes (Figure 6). Over the 12-hour dark phase (hours 1-12) and 18-hour total recovery periods (hours 1-18), NREM sleep was greater than baseline for BMAL1/Mop3+/+ and BMAL1/Mop3+/- mice, whereas NREM sleep time was not significantly increased in BMAL1/Mop3-/- mice. Consequently, the recovery response for these time intervals was significantly smaller in BMAL1/Mop3-/mice compared to the other genotypes (Figure 6). Following sleep deprivation, BMAL1/Mop3+/+ and BMAL1/Mop3+/mice had increased REM sleep over baseline during hours 1 to 6 and 7 to 12 of the dark phase, as well as cumulatively over the 12hour dark and 18-hour total recovery periods (Figure 6). In contrast, BMAL1/Mop3-/- mice failed to exhibit an increase in REM sleep over baseline levels during any interval of the recovery period (Figure 6). Therefore, significant genotype differences

utes) compared to BMAL1/Mop3-/- (146.1 ± 11.1 minutes) mice (F = 4.5, P < .05). BMAL1/Mop3-/- mice were notably more difficult to keep awake during the sleep-deprivation procedure, requiring more interventions to prevent sleep onsets. The average duration of individual wake bouts during the deprivation procedure was shorter in BMAL1/Mop3-/- mice (12.4 ± 5.3 minutes) compared to BMAL1/Mop3+/+ (43.1 ± 11.7 minutes) and BMAL1/Mop3+/- (45.6 ± 10.7 minutes) mice (F = 4.0, P < .05), indicating reduced propensity for sustained wakefulness in the BMAL1/Mop3-/- group. All mice achieved some sleep during the deprivation period; however, BMAL1/Mop3-/- mice tended to sleep more, losing 87.6% ± 4.1% of baseline NREM sleep compared to 93.7% ± 1.7% and 95.2% ± 1.9% in BMAL1/Mop3+/+ and BMAL1/Mop3+/- mice, respectively (P = .15). Recovery Versus Baseline The response to sleep deprivation was assessed separately for each genotype by comparing recovery and corresponding baseline recordings in 2-hour blocks, as shown in Figure 4. In response to sleep deprivation, both BMAL1/Mop3+/+ and BMAL1/Mop3+/- mice had a significant increase in NREM time in hours 3 and 4 of the recovery period, whereas BMAL1/Mop3-/mice showed an increase during hours 1 and 2 (Figure 4). REM sleep was significantly increased after sleep deprivation in hours 3 to 8 in BMAL1/Mop3+/+ and BMAL1/Mop3+/- mice; however, BMAL1/Mop3-/- mice did not increase REM sleep above baseline in any of the 2-hour intervals (Figure 4). NREM delta power was elevated during the first 2-hour interval of recovery sleep in BMAL1/Mop3+/+ (+53%, P < .000), BMAL1/Mop3+/- (+49%, P < .000), and BMAL1/Mop3-/- (+23%, P < .000) mice compared to corresponding baseline levels. However, the magnitude of this response was different between

Table 1—Baseline Sleep-Wake Fragmentation During Entrained Conditions

Arousals Light Dark

+/+

+/-

-/-

+/+ vs -/-

99.4 (8.0) 56.9 (7.1)

85.7 (8.6) 51.7 (7.6)

112.2 (7.6) 94.1 (6.7)

***

440.6 (19.0) 274.4 (25.8)

427.4 (20.3) 247.0 (27.6)

503.7 (17.9) 439.6 (24.4)

***

185.4 (7.4) 118.0 (11.3)

178.3 (7.9) 106.7 (12.1)

196.4 (7.0) 172.1(10.7)

***

42.8 (2.8) 14.9 (2.1)

43.3 (3.0) 13.6 (2.3)

44.6 (2.7) 41.1 (2.0)

***

***

2.2 (0.1) 2.0 (0.2)

2.3 (0.1) 2.3 (0.2)

1.9 (0.1) 1.9 (0.2)

-

-

-

-

1.3 (0.1) 1.3 (0.1)

1.3 (0.1) 1.4 (0.1)

1.1 (0.1) 1.1 (0.1)

** **

** **

-

+/- vs -/-

***

SS Light Dark

-

-

***

NREM (#) Light Dark

-

-

***

REM (#) Light Dark

-

-

NREM (min) Light Dark REM (min) Light Dark

Sleep consolidation was compared between BMAL1/Mop3+/+ (n = 8), BMAL1/Mop3+/- (n = 7), and BMAL1/Mop3-/- (n = 9) mice. Values for arousals, stage shifts (SS), number of sleep bouts (#), and duration (min) of individual sleep bouts in minutes are presented for non-rapid eye movement (N) and rapid eye movement (R) as mean ± SEM for both 12-hour light and 12-hour dark phases. One-way analysis of variance revealed a number of differences in sleep consolidation between genotypes. Posthoc test (Tukey HSD) results are indicated for BMAL1/Mop3+/+ vs BMAL1/Mop3-/- and BMAL1/Mop3+/- vs BMAL1/Mop3-/- comparisons (**P < .01; ***P < .001).

SLEEP, Vol. 28, No. 4, 2005

399


MS 213.qxp

3/14/2005

1:56 PM

Page 400

Figure 2—Baseline non-rapid eye movement (NREM) delta power in BMAL1/Mop3+/+ (n = 8, open), BMAL1/Mop3+/- (n = 7, grey), and BMAL1/Mop3-/- (n = 9, black) mice under entrained (12-hour light:12-hour dark) conditions. NREM delta power (mean ± SEM) in 2-hour intervals normalized as a percentage of 24-hour baseline. The dark bar on the x-axis indicates the 12-hour dark phase. A repeated-measures analysis of variance was used to determined group x time interactions. To indicate significance of posthoc comparisons (Tukey HSD), asterisks left of the hash mark (*/) represent BMAL1/Mop3-/- vs BMAL1/Mop3 +/+, and those to the right of the hash mark (/*) represent BMAL1/Mop3-/- vs BMAL1/Mop3+/- (* P < .05, ** P < .01, *** P < .001).

Figure 3—Electroencephalogram spectral analysis during baseline sleep in BMAL1/Mop3+/+ (n = 8, open), BMAL1/Mop3+/- (n = 7, grey), and BMAL1/Mop3-/- (n = 9, black) mice under entrained (12-hour light:12-hour dark) conditions. The values (mean ± SEM) for absolute delta (1-4 Hz), theta (6-10 Hz), and sigma (11-15 Hz) power were determined individually for wake, non-rapid eye movement (NREM), and rapid-eye movement (REM) states. These values were averaged across the 12-hour light (L) and 12-hour dark (D) periods, and genotype comparisons were made during each period using 1way analysis of variance. Posthoc comparisons (Tukey HSD) represent differences between BMAL1/Mop3-/- mice and both BMAL1/Mop3+/+ and BMAL1/Mop3+/- animals (**P < .01, ***P < .001).

SLEEP, Vol. 28, No. 4, 2005

400


MS 213.qxp

3/14/2005

1:56 PM

Page 401

deprivation (recovery sleep gained over corresponding baseline levels / amount of sleep lost during sleep deprivation). The recordings were divided into hours 1 to 6 (first half of dark phase), 7 to 12 (second half of dark phase), 13 to 18 (first half of light phase), 1 to 12 (entire dark period), and 1 to 18 (entire recovery period) as shown in Figure 7. The proportion of NREM sleep recovered during the first half of the dark phase (hours 1-6) was similar between genotypes. During hours 7 to 12 and 13 to 18 of recovery, BMAL1/Mop3-/mice tended to have less sleep than during corresponding baseline intervals and, therefore, failed to show any further recovery

occurred in the recovery response in REM sleep, particularly during hours 1 to 6, 1 to 12, and 1 to 18, as depicted in Figure 6. Sleep Lost Versus Sleep Gained The direct genotype comparison of recovery sleep is complicated when the genotypes have different amounts of sleep deprivation and different levels of baseline sleep. For this reason, we attempted to normalize the recovery response by determining the amount of sleep regained during the recovery period as a percentage of the amount of sleep lost during the 6 hours of sleep

Figure 4—Baseline vs recovery sleep in BMAL1/Mop3+/+ (n = 8), BMAL1/Mop3+/- (n = 7), and BMAL1/Mop3-/- (n = 9) mice. Data represent percentage of recording time (mean ± SEM) for each 2-hour interval during corresponding baseline (open circles) and sleep-deprivation/recovery (dark circles) times. The left panel presents non-rapid eye movement (NREM) sleep, and the right panel rapid eye movement (REM) sleep for each genotype. Animals were sleep deprived for 6 hours (noon-6:00 pm) followed by an 18-hour recovery period beginning at dark onset. The dark horizontal bar on the x-axis represents the 12-hour dark phase. For statistical analyses, comparisons were made between baseline and recovery at each 2-hour interval using dependent sample t tests. *P < .05, **P < .01.

SLEEP, Vol. 28, No. 4, 2005

401


MS 213.qxp

3/14/2005

1:57 PM

Page 402

differed between genotypes (F = 8.1, P < .01) and was higher in BMAL1/Mop3-/- by 6.2% compared to BMAL1/Mop3+/+ and by 5.3% compared to BMAL1/Mop3+/- mice. These differences were mainly due to NREM sleep (F = 7.9, P < .01), which was elevated in BMAL1/Mop3-/- mice (Figure 8). REM sleep was nonsignificantly higher in BMAL1/Mop3-/- mice (Figure 8). The body-temperature rhythm was apparent in BMAL1/Mop3+/+ and BMAL1/Mop3+/- mice, with expected higher levels during the active phase compared to the rest phase (Figure 8). In contrast, BMAL1/Mop3-/- mice had an equal distribution of body temperature across the recording period, with increased levels in the rest phase (F = 6.9, P < .01) and decreased levels in the active phase (F = 14.6, P < .001) compared to BMAL1/Mop3+/+ and BMAL1/Mop3+/- animals. Over the entire recording period, average body temperature was slightly reduced in BMAL1/Mop3-/- mice compared to the other genotypes (F = 9.0, P < .01) (Figure 8). Sleep fragmentation was elevated in BMAL1/Mop3-/- compared to BMAL1/Mop3+/+ and BMAL1/Mop3+/- mice (Table 2). Specifically, BMAL1/Mop3-/- mice had an increased numbers of arousals (F = 13.0, P < .001), stage shifts (F = 12.0, P < .001), and NREM (F = 9.1, P < .01) and REM (F = 9.3, P < .01) sleep bouts. The average duration of individual NREM (F = 4.2, P < .05) and REM (F = 8.4, P < .01) bouts were decreased in BMAL1/Mop3-/- mice. These results verify that the sleep phenotype in BMAL1/Mop3-/- mice is largely consistent in both entrained and free-running conditions and is not determined by the exogenous light:dark cycle.

or accumulation of NREM sleep time. Therefore, over the 18hour recovery period, BMAL1/Mop3-/- mice had a significantly smaller amount of NREM regained (+3.2% ± 7.8%) compared to BMAL1/Mop3+/- (+26.3% ± 6.4%, P < .05) and BMAL1/Mop3+/(+25.1% ± 6.9%, P < .05) mice. BMAL1/Mop3-/- mice regained a significantly smaller percentage of lost REM sleep during hrs 1 to 6 of the recovery period (Figure 7). The amount of REM sleep regained over the 18-hr recovery period was smaller but not significantly different in BMAL1/Mop3-/- (+45.1% ± 21.2%) compared to BMAL1/Mop3+/+ (+68.2% ± 11.7%), and BMAL1/Mop3+/(+74.7% ± 15.1%) mice. Baseline Sleep During Constant Conditions In constant darkness, BMAL1/Mop3-/- exhibited an arrhythmic locomotor activity pattern (data not shown), as previously described.14 The free-running period was 23.7 ± 0.1 hours in BMAL1/Mop3+/+ mice and 23.7 ± 0.2 hours in BMAL1/Mop3+/mice, similar to previous results.14 Similarly, BMAL1/Mop3-/mice had an equal distribution of sleep-wake states across the rest and active phases, whereas the other genotypes showed a clear consolidation of wakefulness and sleep to the rest and active phases, respectively (Figure 8). Similar to the pattern under entrained conditions, BMAL1/Mop3-/- mice lacked the normal diurnal rhythm of normalized NREM delta power in constant darkness that was demonstrated by BMAL1/Mop3+/+ and BMAL1/Mop3+/- mice, as evidenced by genotype differences across the recording period (Figure 9). BMAL1/Mop3-/- mice had significantly more total sleep time (F = 47.4, P < .001), NREM sleep (F = 28.9, P < .001), and REM sleep (F = 31.1, P < .001) during the active phase and less total sleep time (F = 14.7, P < .001), NREM sleep (F = 8.6, P < .01), and REM sleep (F = 11.7, P < .001) during the rest phase compared to BMAL1/Mop3+/and BMAL1/Mop3+/+ mice (Figure 8). Overall total sleep time

DISCUSSION In addition to abolishing the diurnal rhythm of the sleep-wake cycle, deletion of the BMAL1/Mop3 gene leads to profound differences in the amount of sleep, sleep architecture, and some responses to sleep deprivation. BMAL1/Mop3-/- mice had a

Figure 5—Non-rapid eye movement (NREM) delta power in response to sleep deprivation in BMAL1/Mop3+/+ (n = 8), BMAL1/Mop3+/- (n = 7), and BMAL1/Mop3-/- (n = 9) mice. For each genotype, NREM delta power (mean ± SEM) was determined for each 2-hour interval of the recovery period and then converted to a percentage of 24-hour baseline NREM delta power. Because BMAL1/Mop3-/- mice have a significantly higher level of NREM delta power during baseline sleep, as previously described, the normalization allowed a direct genotype comparison of the magnitude of the NREM delta power during recovery sleep. Note, the first time point is the beginning of the recovery period, beginning at dark onset (6:00 PM). The dark bar on the x-axis marks the 12-hour dark phase. Genotype comparisons were made using a 1-way analysis of variance at each 2-hour intervals with follow-up posthoc testing (Tukey HSD) **P < .01.

SLEEP, Vol. 28, No. 4, 2005

402


MS 213.qxp

3/14/2005

1:57 PM

Page 403

were notably more difficult to keep awake, consistent with their increase in NREM delta power during baseline sleep, possibly indicating that they are chronically under a high sleep pressure. This deficiency in maintaining wakefulness during the sleepdeprivation period was indicated by elevated NREM sleep time and a shorter duration of individual wake bouts during sleep deprivation. Because of the difficulty in interpreting recoverysleep patterns between genotypes with different baseline sleep patterns, we analyzed sleep deprivation and recovery data 2 ways. First, the traditional approach of comparing baseline to recovery sleep patterns was used. In the initial hours following sleep deprivation, BMAL1/Mop3-/- mice exhibited an increase in NREM sleep time and NREM delta power over corresponding baseline levels; however, the magnitude of the NREM delta power response was smaller compared to BMAL1/Mop3+/+ and BMAL1/Mop3+/- mice. In the middle and later hours of the recovery period (hours 7-18), BMAL1/Mop3-/- mice had a small negative rebound in NREM sleep, and, therefore, the recovery response over the entire 18-hour recovery period was significant-

notably attenuated diurnal distribution of sleep stages and NREM delta power under entrained conditions, as well as significant increases in 24-hour NREM and REM baseline sleep amounts. In addition to changes in sleep time, BMAL1/Mop3-/- mice had increased levels of absolute NREM delta power throughout the entire 24-hour baseline period. The concurrent increase in both sleep time and NREM delta power rules out the possibility that one parameter was enhanced as a response to, or compensation for, a decrease in the other. Delta power in REM sleep was also increased in BMAL1/Mop3-/- mice, although there was no overall change in REM sleep time. In constant conditions, the circadian sleep-wake rhythm was abolished, NREM sleep time and delta power were increased, and REM sleep time was nonsignificantly elevated in BMAL1/Mop3-/- mice compared to the other genotypes. Therefore, the sleep phenotype in BMAL1/Mop3-/mice was largely similar in both entrained and free-running conditions, indicating the changes were not accounted for by masking effects of the light:dark cycle. During the sleep-deprivation procedure, BMAL1/Mop3-/- mice

Figure 6—Recovery sleep (rec) normalized as a percentage change from baseline levels (bsln) in BMAL1/Mop3+/+ (n = 8), BMAL1/Mop3+/- (n = 7), and BMAL1/Mop3-/- (n = 9) mice. The 18-hour recovery period (beginning at dark onset, 6:00 PM) was divided into the following time intervals: hours 1 to 6 (first half of dark period), hours 7 to 12 (second half of dark period), hours 13 to 18 (first half of light period), hours 1 to 12 (total dark period), and hours 1 to 18 (total recovery period), as noted on the x-axis. For each time interval, the amount of non-rapid eye movement (NREM) and rapid eye movement (REM) sleep achieved during recovery was calculated as a percentage change (mean ± SEM) from corresponding baseline levels (recovery/baseline) for each genotype. At each interval, the recovery response was compared to baseline levels using dependent sample t tests for each genotype (*P < .05, **P < .01). Next, a 1-way analysis of variance was used to compare the recovery response (percentage change from baseline) between genotypes at each time interval (horizontal bars represent P < .05).

SLEEP, Vol. 28, No. 4, 2005

403


MS 213.qxp

3/14/2005

1:57 PM

Page 404

hours following sleep deprivation, NREM sleep rebound is similar between genotypes, whereas REM sleep rebound is significantly attenuated in BMAL1/Mop3-/- mice. Over the course of the 18-hour recovery period, BMAL1/Mop3-/- mice tended to surmount a smaller recovery response in NREM and REM sleep, in terms of percentage change from baseline levels and with respect to percentage of sleep lost during the deprivation procedure. The combination of circadian disruption, increased propensity or drive for sleep in baseline conditions, reduced ability to sustain wakefulness during sleep deprivation, and alterations in sleep recovery in BMAL1/Mop3-/- mice indicates that the BMAL1/Mop3 gene functions as a common element participating in both circadian and sleep regulatory processes. The manner in which the BMAL1/Mop3 gene participates in sleep-wake control may occur through (1) its role in regulating the circadian system, which inputs into sleep-wake pathways; (2) direct activity in sleep-related nuclei; and/or (3) pleiotropic effects in other processes, such as metabolism, that feed back to influence sleep-wake states. In the discussion below, we address each of these possibilities. A long-standing model of sleep-wake regulation proposes that sleep is controlled by 2 interacting processes, consisting of circadian and homeostatic drives.3 These 2 components are thought to be mechanistically independent of each other, in part because of studies in rats showing that SCN lesions disrupt the normal timing

ly smaller compared to the other genotypes. In BMAL1/Mop3-/mice, REM sleep was similar between baseline and recovery conditions, even though there was a robust REM recovery in BMAL1/Mop3+/+ and BMAL1/Mop3+/- mice. One possible reason for the attenuated recovery in BMAL1/Mop3-/- mice is that they already have a high sleep drive under baseline conditions, which may be difficult to surpass during the sleep-recovery opportunity. Also, BMAL1/Mop3-/- mice lost a smaller amount of sleep during the deprivation period compared to BMAL1/Mop3+/+ and BMAL1/Mop3+/- mice and, consequently, may have built up less of a sleep drive from which to recover. In order to account for the differences in the degree of sleep deprivation between genotypes, our second approach was to normalize the recovery response for each genotype by calculating the amount of sleep gained during recovery as a percentage of sleep lost during the deprivation period. These data showed that in the first 6 hours of recovery, all genotypes regained a similar amount of NREM sleep when normalized to the amount lost during sleep deprivation. Over the course of the 18-hour recovery period, BMAL1/Mop3-/- mice regained a smaller percentage of lost sleep compared to the other genotypes. The percentage of REM sleep regained during recovery was significantly smaller in BMAL1/Mop3-/- mice during the first 6 hours of recovery and nonsignificantly reduced over the entire 18-hour recovery period. Taken together, these 2 analyses indicate that, during the initial

Figure 7—Sleep regained during recovery (recov) versus sleep lost during the deprivation period in BMAL1/Mop3+/+ (n = 8), BMAL1/Mop3+/- (n = 7), and BMAL1/Mop3-/- (n = 9) mice. The 18-hour recovery period (beginning at dark onset, 6:00 PM) was divided into the following time intervals: hours 1 to 6 (first half of dark period), hours 7 to 12 (second half of dark period), hours 13 to 18 (first half of light period), hours 1-12 (total dark period), and hours 1 to 18 (total recovery period), as noted on the x-axis. At each time interval, the amount of non-rapid eye movement (NREM) and rapid eye movement (REM) sleep gained during recovery (minutes during recovery - minutes during baseline [bsln]) was calculated as a percentage of the amount of sleep lost during the deprivation period (minutes gained/minutes lost) (mean ± SEM) for each genotype. Genotype comparisons were made at each time interval using 1way analysis of variance with posthoc testing (Tukey HSD), when indicated (horizontal bars indicate P < .05).

SLEEP, Vol. 28, No. 4, 2005

404


MS 213.qxp

3/14/2005

1:57 PM

Page 405

of sleep but do not alter total sleep time or the compensatory response to sleep deprivation,4-6,16-17 although a modest increase in baseline sleep time following SCN lesions has recently been reported in young, middle-age, and old rats maintained in constant dim light.21 A recent study in mPer2 and mPer1/mPer2 mutant mice showed that total sleep time did not change between entrained conditions versus when these same animals became arrhythmic in constant darkness, indicating the independence of circadian and sleep homeostatic regulation.18 In contrast, SCN lesions in the diurnal squirrel monkey lead to increased total sleep time, indicating the SCN may produce an alerting factor.19-20 In addition, Siberian hamsters that become arrhythmic after a 5-hour phase shift have significantly more baseline NREM and REM sleep time compared to hamsters that maintain circadian rhythmicity after the same phase-shift stimulus.22 Interestingly, the arrhythmic hamsters were more difficult to keep awake during a 2-hour and 6-hour sleep-deprivation procedure but did express a rebound in NREM delta power during the recovery opportunity.22 In addition to the BMAL1/Mop3 gene, there is now accumulated evidence for a critical role of other mammalian circadian clock genes in sleep-wake regulation that goes beyond just the circadian timing of sleep. For example, in Clock mutant mice, NREM sleep is reduced by 1 to 2 hours per day under both light-dark and dark-dark conditions, and the rebound of REM sleep following

sleep deprivation is attenuated.10 Interestingly, Cry1,2 doubleknockout mice have a severely attenuated sleep-wake rhythm and increased NREM baseline sleep time and NREM delta power, similar to BMAL1/Mop3-/- mice in the present study, as well as decreased compensatory responses following sleep deprivation.12 In other circadian clock mutant-deletion mice models, changes in sleep and wake have mimicked the underlying circadian phenotype rather than reflecting a primary change in the amount or intensity of sleep. For example, while total sleep time is preserved in mPer1 and mPer2 mutant mice, the 24-hour distribution of sleep is affected by the mutation, reflecting changes in the phase of the activity and body temperature rhythms relative to the lightdark cycle in mutant animals.18,23 In NPAS2-deficient mice, wake time is significantly increased in the dark phase, reflecting the enhanced level and consolidation of locomotor activity, although the response to sleep deprivation was not assessed in this experiment.24 Deletion of the d-albumin binding protein gene, a target of the core circadian clock, leads to increased sleep fragmentation and an altered distribution of NREM delta power.25 These studies suggest that one way in which the BMAL1/Mop3 gene may be involved in sleep-wake regulation is by altering properties of the circadian system, which then results in changes in the nature of circadian input into sleep-wake pathways. An alternative possibility is that circadian clock genes directly influence cellular pro-

Figure 8—Baseline sleep patterns in BMAL1/Mop3+/+ (n = 7, open), BMAL1/Mop3+/- (n = 6, grey), and BMAL1/Mop3-/- (n = 6, black) mice under freerunning (constant darkness) conditions. Sleep and wake amounts were determined as a percentage of recording time (mean ± SEM) over the rest phase, active phase and total free-running period for each genotype. One-way analysis of variance was used to compare genotypes at each time interval. To indicate significance of posthoc comparisons (Tukey HSD), asterisks left of the hash mark (*/) represent BMAL1/Mop3-/- vs BMAL1/Mop3 +/+ and those to the right of the hash mark (/*) represent BMAL1/Mop3-/- vs BMAL1/Mop3+/- (*P < .05, **P < .01, ***P < .001). NREM refers to non-rapid eye movement; REM, rapid eye movement.

SLEEP, Vol. 28, No. 4, 2005

405


MS 213.qxp

3/14/2005

1:57 PM

Page 406

role in mechanisms involved in regulating sleep intensity and homeostatic sleep drive. In addition to Clock, BMAL1/Mop3 has also been found to dimerize with a homolog of the Clock gene, NPAS-2.26 The Clock: BMAL1/Mop3 and NPAS2: BMAL1/Mop3 heterodimers occur predominantly in different areas of the brain26 and may potentially be involved in different aspects of sleep-wake regulation, such as continuity, duration, or intensity. It will be possible to test more-specific properties of circadian genes when tissue-specific genetic models are developed. The BMAL1/Mop3 null allele could lead to changes in sleep by directly altering SCN function. It has been proposed that the SCN is a component of the wake-regulatory pathway and that efferent signals from the SCN have arousal-promoting effects.19,20 The SCN are connected with important sleep-wake nuclei in the hypothalamus and brainstem,27 influence activity of locus coeruleus neurons,20 and display sleep-wake state-specific firing patterns.28 A primary change in SCN function could result in removal of an alerting signal, resulting in an impaired ability to maintain wakefulness across the dark phase.19 Alternatively, sleep changes resulting from circadian gene manipulation may not be specific to activity within the SCN. Importantly, the highwake phenotype in Clock mutant mice is not changed following SCN lesions, indicating that the mutation is operative in extraSCN tissue (unpublished observations). The finding that circadian clock genes are expressed in a variety of brain regions, includ-

cesses within neurons involved in sleep-wake regulation either through circadian molecular machinery in these cells or through noncircadian cellular events. Regardless of the mechanism of action, the variety of sleep phenotypes among the genetic models for clock genes suggests unique roles for individual circadian genes in the sleep-regulatory system. Determining whether the mechanisms linking the circadian and sleep systems are specific to the circadian clock feedback loop or extend to posttranslational and noncircadian pathways remains to be determined. A prominent feature of sleep in BMAL1/Mop3-/- mice is disrupted consolidation of individual sleep cycles, indicated by a high number of arousals, stage shifts, and sleep bouts. The increase in sleep time in BMAL1/Mop3-/- mice could be attributed as compensation for the highly fragmented sleep pattern. Normally, sleep fragmentation and intensity are inversely related, whereas, in BMAL1/Mop3-/- mice, they are both increased. It is possible that BMAL1/Mop3-/- mice accumulate and dissipate NREM delta power with a faster time constant, as previously shown in Cry1,2/- mice12 and, therefore, can remain at higher and more constant levels, even in the context of fragmented sleep. The increase in NREM delta power in BMAL1/Mop3-/- and Cry1,2-/- mice cannot be accounted for simply by a loss of circadian control, since sleep intensity has not been reported to change in other rat, mouse, and hamster models of arrhythmicity.10,18,22 This could possibly mean that BMAL1/Mop3 and Cry genes serve a circadian-independent

Figure 9—Baseline non-rapid eye movement (NREM) delta power in BMAL1/Mop3+/+ (n = 7, open), BMAL1/Mop3+/- (n = 6, grey), and BMAL1/Mop3(n = 6, black) mice under free-running (constant darkness) conditions. NREM delta power (mean ± SEM) was normalized across the active (indicated by dark bar on x-axis) and rest phases as a percentage of the total NREM delta power for the entire recording period. At each time interval, genotypes were compared using 1-way analysis of variance with posthoc testing (Tukey HSD) to clarify main effects. To indicate significance of posthoc comparisons, asterisks left of the hash mark (*/) represent BMAL1/Mop3-/- vs BMAL1/Mop3 +/+ and those to the right of the hash mark (/*) represent BMAL1/Mop3-/- vs BMAL1/Mop3+/- (*P < .05, **P < .01, ***P < .001).

/-

SLEEP, Vol. 28, No. 4, 2005

406


MS 213.qxp

3/14/2005

1:57 PM

Page 407

not take measurements of daily food intake or other measures of fuel metabolism. The trend for lower body weight in BMAL1/Mop3-/mice cannot be explained by increased energy expenditure due to physical activity because these mice have low activity levels. We found that, under entrained conditions, the average daily temperature was similar between genotypes, indicating that BMAL1/Mop3-/- mice were not hypothermic or hyperthermic, and gross alterations in temperature regulation, such as fever, are not responsible for their sleep phenotype. However, the bodytemperature rhythm was abolished in Mop3-/- mice, resulting in increased temperature across the light phase and decreased temperature across the dark phase compared to the other genotypes. Because body temperature is closely related to sleep-wake state in normal animals, the lack of a temperature rhythm in BMAL1/Mop3-/- mice reflects the flattened distribution of sleep and wakefulness across the 24-hour cycle. In constant darkness, average daily body temperature is decreased in BMAL1/Mop3-/mice. The reason for this is unclear and is not directly explained by the animals sleeping more or by having lower activity levels. In fact, the increase in sleep time in BMAL1/Mop3-/- mice is comparable between entrained and constant conditions, and there is no indication that their activity levels further decrease in constant conditions.14 The alterations in activity level, body weight, and body temperature, as well as observations that BMAL1/Mop3-/- mice appear to be unhealthy,35 represent important aspects of their behavior/metabolic phenotype, in addition to the notable changes in sleep-wake patterns and circadian profiles. Again, characterization of the health status of these animals will require systematic studies of their metabolic function and possibly lifespan. While, the Clock mutant mouse was the first mammalian example of a genetic alteration of sleep and circadian disruption, it has only recently been discovered that these mice have notable changes in energy balance and endocrine function.30 These animal models stimulate important questions about the role of circadian genes in multiple physiologic processes and more generally about the interaction between sleep, circadian, and metabolic systems. For example, since normal sleep and circadian rhythmicity are thought to be important for the health and well-being of the organism, then any disruption of normal sleep or rhythmicity would be expected to impact the health of the organism, which in turn could impair sleep or rhythmicity even further. While the phenotypes of these animal models can become complex, they are providing exciting evidence about the expanded or multiple roles of sleep and circadian rhyth-

ing some that are involved in sleep regulation,29 raises the interesting possibility that changes in circadian clock-gene expression, independent of SCN-timing functions, may underlie their wide-ranging effects on sleep architecture. Circadian genes are expressed and oscillate in brain regions outside of the SCN as well as in peripheral tissues.7,29 The wide distribution of circadian genes may represent a mechanism by which behavior and physiologic systems respond to the output of the SCN, generate independent rhythms, or serve noncircadian functions. Along these lines, an accumulation of data is pointing to interesting relationships between the control of circadian rhythms, sleep, and metabolism. For example, there are indications of both sleep and metabolic phenotypes in Clock mutant mice: these mice have increased amounts of daily wake time, become obese, and develop other symptoms of metabolic dysfunction.30 We have also found that in a genetic model of obesity, the ob/ob mouse, sleep-wake patterns are notably fragmented, overall NREM sleep time is increased, and the rhythm of locomotor activity is attenuated.31 A high-fat diet has been shown to increase NREM sleep time in mice, and food restriction decreases NREM and REM sleep.32,33 Therefore, it is possible that some of the effects of circadian genes on sleep, including the BMAL1/Mop3 gene, could be mediated by their interactions with other physiologic systems, such as metabolic control, which then may feed back to influence sleep. Specific information about the metabolic status of BMAL1/Mop3-/- mice will be of particular interest in light of the growing evidence of interactions between sleep and metabolic control. It has previously been shown that levels of wheel-running activity were significantly reduced in BMAL1/Mop3-/- mice.14 We replicated these findings using infrared activity monitors, indicating that the low-activity phenotype in BMAL1/Mop3-/- mice is not specific to the presence of a running wheel. Comparable deficiency in locomotor activity has not been described in other circadian clock-gene mutant animals, although slightly reduced activity levels have been reported in albumin D-binding protein deficient mice.34 Therefore, it is unlikely that the low activity in BMAL1/Mop3-/- mice is completely related to the role of the BMAL1/Mop3 gene in circadian regulation. The BMAL1/Mop3 gene is expressed in muscle tissue and could possibly influence activity levels through involvement in the mechanics of muscle contraction or energy utilization in muscle. In the context of metabolism, low activity levels could represent an adaptive adjustment of the organism to conserve energy, particularly in response to starvation or weight loss. In our study, body weight tended to be decreased in BMAL1/Mop3-/- mice, although we did Table 2—Baseline Sleep-Wake Fragmentation During Constant Conditions

Arousals

+/+

+/-

-/-

+/+ vs. -/-

+/- vs. -/-

97.8(9.5)

102.5 (10.3)

163.2(10.3)

***

**

SS

554.3(37.7)

565.3(40.7)

799.5(40.7)

**

**

NREM (#)

236.7(17.3)

239.8(18.7)

334.8(18.7)

**

**

REM (#)

45.9(3.5)

58.2(3.8)

68.3(3.8)

**

NREM (min)

2.7(0.2)

2.8(0.2)

2.0(0.2)

REM (min)

1.5(0.1)

1.4(0.2)

1.3(0.1)

* **

*

Sleep consolidation was compared between BMAL1/Mop3+/+ (n = 7), BMAL1/Mop3+/- (n = 6), BMAL1/Mop3-/- (n = 6) mice in constant darkness. Values for arousals, stage shifts (SS), number of sleep bouts (#) and duration (min) of individual sleep bouts in minutes are presented for non-rapid eye movement (NREM) and rapid eye movement (REM) sleep as mean ± SEM for the overall recording period. Results of posthoc tests are indicated by letters representing BMAL1/Mop3+/+ vs BMAL1/Mop3-/- and BMAL1/Mop3+/- vs BMAL1/Mop3-/- (*P < .05; **P < .01; ***P < .001). SLEEP, Vol. 28, No. 4, 2005

407


MS 213.qxp

3/14/2005

1:57 PM

Page 408

micity for maintaining health and well being. Drosophila has emerged as an important model for studying the molecular and genetic regulation of sleep. The identification of rest-activity patterns in flies that parallel properties of sleep-wake states in mice, including compensatory responses to rest deprivation, has led to evidence for the role of circadian clock genes in the homeostatic regulation of rest-activity in flies.9,11,13,15 In view of the many common molecular elements involved in the generation of circadian rhythms in flies and mice, it is of great interest to establish the function of genes that have structural conservation between these genetic models. With regard to baseline characteristics, consolidated rest time is disturbed in dClock (ClkJrk) mutants and impaired to an even greater extent in cycle01 null (cyc01) flies compared to per and tim null flies and wild-type controls.13 Following various durations of rest deprivation, wild-type flies recovered about 30% to 40% of the rest that was lost, compared to nearly 100% recovery in per deficient and ClkJrk flies. Male cyc01 mutants have failed to exhibit a rebound following rest deprivation in either LD or DD conditions, whereas female cyc01 mutants have had a dramatically exaggerated recovery response both in magnitude and duration.11,13 These sleep changes were not detected, however, in male cyc02 flies.35 While it is difficult to precisely compare rest time in flies with sleep time in mice, due to differences in what is being measured, scoring criteria, and methodologies, there is an interesting similarity in rest-sleep fragmentation and attenuated compensatory responses to rest-sleep deprivation in male cycle null flies and BMAL1/Mop3-/- mice. Similarly, timeless-deficient flies did not rebound following mild (3-6 hours) sleep deprivation, and Cry1,2-/- mice have an attenuated rebound after 6 hours of sleep loss.12 We have recently reported sex differences in the sleep of male and female C57Bl/6J mice.36 Since our present study involved only male mice, it is not known whether sex differences also occur in male and female BMAL1/Mop3-/- animals. A sex difference has been shown in the sleep phenotype of cycle mutant flies.13 More generally, it will be of interest to determine what role the circadian system may play in sex differences in sleep. A body of literature in now accumulating in flies and mice that supports the hypothesis that homologous genes identified as core elements of the circadian clock in these diverse species also influence sleep-regulatory processes that are involved in the amount and consolidation of sleep and wake bouts, amount of sleep, and sleep architecture, as well as the rebound response to sleep deprivation. The use of genetically altered flies and mice is expected to lead to new insights into the mechanisms governing circadian and sleep-regulatory processes and how these processes have remained integrated with one another to regulate the overall temporal organization of diverse animals over millions of years of evolutionary time.

4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

14. 15. 16. 17. 18.

19. 20. 21. 22. 23.

REFERENCES 1. 2. 3.

24.

Herzog ED, Schwartz WJ. Functional genomics of sleep and circadian rhythm invited review: a neural clockwork for encoding circadian time. Appl Physiol 2002;92:401-8. Kalsbeek A, Buijs RM. Output pathways of the mammalian suprachiasmatic nucleus: coding circadian time by transmitter selection and specific targeting. Cell Tissue Res 2002;309:109-18. Daan S, Beersman DGM, Borbely AA. Timing of human sleep: recovery process gated by a circadian pacemaker. Am J Physiol 1984;246:161-78.

SLEEP, Vol. 28, No. 4, 2005

25. 26.

408

Ibuku N, Nihonmatsu I, Sekiguchi S. Sleep-wakefulness rhythms in mice after suprachiasmatic nucleus lesions. Waking Sleeping 1980:2:167-73. Mistleberger RE, Bergmann BM, Waldenar W, Rechtschaffen A. Recovery sleep following sleep deprivation in intact and suprachiasmatic nuclei-lesioned rats. Sleep 1983;6:217-33. Tobler I, Borbely AA, Groos G. The effect of sleep deprivation on sleep in rats with suprachiasmatic lesions. Neurosci Lett 1983;42:49-54. Wang GK, Sehgal A. Signaling components that drive circadian rhythms. Curr Opin Neurobiol 2002;12:331-8. Panda S, Hogenesch JS, Kay SA. Circadian rhythms from flies to human. Nature 2002;417:329-35. Hendricks JC. Invited review: sleeping flies don’t lie: the use of drosophila melanogaster to study sleep and circadian rhythms. J Appl Physiol 2003;94:1660-72. Naylor E, Bergmann BM, Krauski K, et al. The circadian clock mutation alters sleep homeostasis in the mouse. J Neurosci 2000;20:8138-43. Shaw PJ, Tononi G, Greenspan RJ, Robinson DF. Stress response genes protect against lethal effects of sleep deprivation in drosophila. Nature 2002;417:287-91. Wisor JP, O’Hara BF, Terao A, et al. A role for cryptochromes in sleep regulation. BMC Neurosci 2002;3:20-34. Hendricks JC, Lu S, Kume K, Yin JCP, Yang Z, Seghal A. Gender dimorphism in the role of cycle (BMAL1) in rest, rest regulation, and longevity in drosophila melanogaster. J Biol Rhythms 2003;18:12-25. Bunger MK, Wilsbacher LD, Moran SM, et al. Mop3 is an essential component of the master circadian pacemaker in mammals. Cell 2000;103:1009-17. Shaw P. Awakening to the behavioral analysis of sleep in drosophila. J Biol Rhythms 2003;18:4-11. Mouret J, Coindet J, Debilly G, Chouvet G. Suprachiasmatic nuclei lesions in the rat: alterations in sleep circadian rhythms. Electroencephalogr Clin Neurophysiol 1978;45:402-8. Eastman CI, Mistlberger RE, Rechtschaffen A. Suprachiasmatic nuclei lesions eliminate circadian temperature and sleep rhythms in the rat. Physiol Behav 1984;32:357-68. Shiromani PJ, Zu M, Winston EM, Shiromani SN, Gerashchenko D, Weaver DR. Sleep rhythmicity and homeostasis in mice with targeted disruption of mPeriod genes. Am J Physiol (in press, Epub ahead of print). Edgar CM, Dement WC, Fuller CA. Effect of SCN lesions on sleep in squirrel monkeys: evidence for opponent processes in sleepwake regulation. J Neurosci 1993;13:1065-79. Aston-Jones G, ChenS, Zhu Y, Oshinsky ML. A neural circuit for circadian regulation of arousal. Nat Neurosci 2001;4:732-8. Mendelson WB, Bergmann BM, Tung A. Baseline and post-deprivation recovery sleep in SCN-lesioned rats. Brain Res 2003;980):185-90. Larkin JE, Yokogawa T, Heller HC, Franken P, Ruby NF. Homeostatic regulation of sleep in arrhythmic Siberian hamsters. Am J Physiol (in press, Epub ahead of print). Kopp C, Albercht U, Zheng B, Tobler I. Homeostatic sleep regulation is preserved in mPer1 and mPer2 mutant mice. Europ J Neurosci 2002;16:1009-17. Dudley CA, Erbel-Sieler C, Estill SJ, et al. Altered patterns of sleep and behavioral adaptability in NPAS2-deficient mice. Science 2003;301:379-83. Franken P, Lopez-Molina L, Marcacci L, Schibler U, Tafti M. The transcription factors DBP affects circadian sleep consolidation and rhythmic EEG activity. J Neurosci 2000;20:617-25. Reick M, Garcia JA, Dudley C, McKnight SL. NPAS2:an analog of clock operative in the mammalian forebrain. Science 2001;294:506-9. Deletion of BMAL1/Mop3 and Sleep—Laposky et al

MS 213.qxp

3/14/2005

1:57 PM

Page 409

27. Pace-Schott EF, Hobson JA. The neurobiology of sleep: genetics, cellular physiology and subcortical networks. Nature Rev Neurosci 2002;3:591-605. 28. Deboer T, Vansteensel MJ, Detari L, Meijer JH. Sleep states alter activity of suprachiasmatic nucleus neurons. Nat Neurosci 2003;6:1086-90. 29. Abe M, Herzog ED, Yamazake S, Straume M, Tei H, Sakaki Y, Menaker M, Block GD. Circadian rhythms in isolated brain regions. J Neurosci 2002;22:350-6. 30. Kohsaka A, Joshu C, McDearmon E, et al. Role of the transcription factor, CLOCK, in feeding and energy balance. Ninth Meeting of the Society for Research on Biological Rhythms (9th SRBR), June 24-26, 2004, Whistler, BS, Canada. 31. Shelton J, Kosaka A, Perrino N, Joshu C, Bass J, Tuerk FW, Laposky AD. Role of leptin in sleep consolidation and circadian rhythmicity. Sleep 2004;27:A31. 32. Jenkins JB, Omori T, Guan Z, Fang J. Sleep is increased in mice with obesity induced by high fat food. Sleep 2004;27:A20. 33. Minet-Ringuet J, LeRuyet PM, Tome D, Even PC. A tryptophanrich protein efficiently restores sleep after food deprivation in the rat. Behav Br Res 2004;152:335-40. 34. Lopez-Molina L, Conquet F, Dubois-Dauphin M, Shibler U. The DBP gene is expressed according to a circadian rhythm in the suprachiasmatic nucleus and influences circadian behavior. EMBO 1997;16(22):6762-71. 35. Ward CP, Shaw PJ, Dauphin LJ, McKenna JT, McCarley RW, Strecker RE. Comparison of sleep homeostasis in BMAL1 KO mice and drosophila cyc01 and cyc02 mutants. Sleep 2004;27:A378. 36. Paul KN, Carlin KM, Turek FW, Laposky AD. Mice display sex differences in wake time, sleep time and recovery from sleep deprivation. Sleep 2004;27:A378.

SLEEP, Vol. 28, No. 4, 2005

409


3/14/2005

1:57 PM

Page 410

Affordable, accurate, dependable...

Accessories

IM Systems Offers Quality.

ActiTrac ¨

with IM Systems' ActiTrac actigraphy monitor, you can

Efficiency. Value.

Marker Button $50 Light Sensor $50 Smart Download Cable $99 PC Display & Analysis Software $399

Now with Light & Marker

Activity & Light Analysis

measure ambient light and record event marks, document and identify circadian patterns and

ActiTrac V8.19 File

Screen Setup

250

Activity

analyze sleep/wake efficiency: administer, monitor and evaluate treatment response; and objectively validate patient diaries.

Scale Data

Raster

Monitor

Export Data

Activity Scoring Reports

1

2

5

Help

480

200 360

Lux

MS 213.qxp

150 240 100

Circadian Raster Plots

120

50 11/2

0 13:20 11/3

13:30

13:40

Selected Activity Total: 50455

13:50

Average Activity: 28

14:00

14:10

0

Window Epochs: 1800

11/4

11/5

11/6 12 AM

3 AM

6 AM

9 AM

12 PM

3 PM

6 PM

9 PM

12 AM

● FDA # K992410 ● CE compliant ● 15,000 hr. battery life ● Programmable start

PAM-RL Recorder

Leg Activity & Body Position Monitor

● No expensive interface required ● Windows-based software ● Comprehensive report summarizes 33 parameters

High Speed Smart Cable $199 PC Display & Analysis Software $399

Featuring Body Position Sensor

Uniquely designed for monitoring leg activity, IM Systems' PAM-RL offers PSG-quality movement recording with integrated body position sensing at a fraction of the cost. Windows-based software provides position, time & intensity values for RLS/PLMS activity. For objective measurement & quantification of candidate leg movements & monitoring treatment response, PAM-RL is your optimum choice.

● FDA # K010997

PAM-RL Leg Activity Report #

Time

Strength

Duration

Interval

Position

350

06:24:05

203

1.1

108.7

Up

351

06:24:22

445

2.4

17

Up

352

06:24:35

896

6

13.1

Up

353

07:16:29

313

0.9

3113.9

Up

354

22:07:13

199

1

53443.3

Down

355

22:07:39

357

1.8

26.2

Down

356

22:07:54

795

8.9

15.5

Down

357

22:09:25

208

1.6

90.4

Down

358

22:09:55

286

2

30.7

Down

359

22:12:58

205

1

182.6

Down

360

22:20:04

464

4.5

426.4

Down

361

22:20:34

400

5.0

770

Down

www.imsystems.net

● CE compliant ● Tri-axial (X,Y,Z) leg motion sensors ● LCD displays total leg activity ● Programmable start ● No expensive interface required

[email protected]

IM Systems • 1055 Taylor Ave. • Suite 300 • Baltimore, MD 21286 • (888) 513-5969 • Fax (410) 321-0643