Sleep deprivation is a powerful tool to investigate sleep function and regulation. We describe a protocol to sleep deprive Drosophila using the Sleep Nullifying Apparatus, and to determine the extent of rebound sleep induced by deprivation.
Sleep homeostasis, the increase in sleep observed following sleep loss, is one of the defining criteria used to identify sleep throughout the animal kingdom. As a consequence, sleep deprivation and sleep restriction are powerful tools that are commonly used to provide insight into sleep function. Nonetheless, sleep deprivation experiments are inherently problematic in that the deprivation stimulus itself may be the cause of observed changes in physiology and behavior. Accordingly, successful sleep deprivation techniques should keep animals awake and, ideally, result in a robust sleep rebound without also inducing a large number of unintended consequences. Here, we describe a sleep deprivation technique for Drosophila melanogaster. The Sleep Nullifying Apparatus (SNAP) administers a stimulus every 10s to induce negative geotaxis. Although the stimulus is predictable, the SNAP effectively prevents >95% of nighttime sleep even in flies with high sleep drive. Importantly, the subsequent homeostatic response is very similar to that achieved using hand-deprivation. The timing and spacing of the stimuli can be modified to minimize sleep loss and thus examine non-specific effects of the stimulus on physiology and behavior. The SNAP can also be used for sleep restriction and to assess arousal thresholds. The SNAP is a powerful sleep disruption technique that can be used to better understand sleep function.
Sleep is near universal in animals, yet its function remains unclear. Sleep homeostasis, the compensatory increase in sleep following sleep deprivation, is a defining property of sleep, that has been used to characterize sleep states in a number of animals1,2,3,4,5.
Sleep in the fly has many similarities with human sleep, including a robust homeostatic response to sleep loss4,5. Numerous studies of sleep in the fly have used sleep deprivation both to infer sleep function by examining the adverse consequences that accrue from extended waking, and to understand sleep regulation by determining the neurobiological mechanisms controlling the homeostatic regulation of sleep. Thus sleep deprived flies were shown to exhibit impairments in learning and memory6,7,8,9,10,11,12, structural plasticity13,14,15, visual attention16, recovery from neuronal injury17,18, mating and aggressive behaviors19,20, cell proliferation21, and responses to oxidative stress22,23 to name a few. Further, investigations into the neurobiological mechanisms controlling rebound sleep have yielded critical insights into the neuronal machinery that constitutes the sleep homeostat8,9,23,24,25,26,27,28,29. Finally, in addition to revealing fundamental insights into sleep function in healthy animals, sleep deprivation studies have also informed insights into sleep function in diseased states30,31.
While sleep deprivation is undeniably a powerful tool, with any sleep deprivation experiment, it is important to distinguish phenotypes that result from extended waking, from those induced by the stimulus used to keep the animal awake. Sleep deprivation by hand deprivation or gentle handling, is generally regarded as setting the standard for minimally disruptive sleep deprivation. Here we describe a protocol for sleep depriving flies using the Sleep Nullifying Apparatus (SNAP). The SNAP is a device that delivers a mechanical stimulus to flies every 10s, keeping flies awake by inducing negative geotaxis (Figure 1). The SNAP efficiently deprives flies of >98% of night-time sleep, even in flies with high sleep drive8,32. The SNAP has been calibrated on bang sensitive flies, agitation of flies in the SNAP does not harm flies; sleep deprivation with the SNAP induces a rebound comparable with that obtained by hand deprivation7. The SNAP is thus a robust method to sleep deprive flies while controlling for the effects of the arousing stimulus.
1. Experimental preparation
2. Preparation of tubes for sleep recording
NOTE: Sleep is monitored using locomotor activity monitors. A monitor can hold 32 flies housed individually in 5 mm diameter tubes. Typically, genotypes are analyzed in groups of 16 or 32 flies.
3. Recording sleep
4. Sleep deprivation and recovery
Canton S (Cs) was used as a wild-type strain. Flies were maintained on a 12 h light: 12 h dark schedule, and sleep deprived for 12 hours overnight. Inspection of the sleep profiles of Cs flies on the baseline day (bs), sleep deprivation day (sd), and two recovery days (rec1 and rec2) (Figure 2A) suggests that flies were effectively sleep deprived in the SNAP, and recovered sleep during the day consistent with observed reports in the literature4,5. The effectiveness of the SNAP in keeping flies awake is also seen in the high activity (300-350 counts/h) exhibited by flies during sleep deprivation (Figure 2B). Indeed, monitoring the activity counts of flies during sleep deprivation can be a useful barometer of the effectiveness of the deprivation protocol and/or an indirect measure of sleep drive. When the sleep deprivation is ineffective, flies are not as active during the period of deprivation. Flies that are under high sleep drive quickly fall asleep after each stimulus and do not traverse the tube as much35. Both the angle of tilt of the apparatus, and the speed of the drop are critical to ensuring that flies are effectively kept awake without harming them. Each lab can optimize the angle and velocity by adjusting the spring (Figure 1B) and/or the size and shape of the cam (Figure 1C and Figure 1D, right).
To quantitatively estimate the effectiveness of sleep deprivation and of recovery, sleep lost during deprivation and then regained in the recovery days was calculated for each individual fly (Figure 2C). Importantly, there was no significant change in baseline sleep between the deprivation day and the baseline day (see 0-12 h in Figure 2C) indicating that sleep is stable in these flies. A large difference in sleep in this 12 hour period (e.g., ± 100 min) would suggest that sleep was not stable. The SNAP effectively deprived flies of >98% of their night-time sleep. Flies recovered ~20% of their sleep in the first 12 h and did not recover additional sleep during the night, as previously reported. However, flies began to recover sleep the following day such that they recovered ~36% of their sleep over 48 h of recovery (Figure 2D). 30 – 40% recovered sleep over 48 h is fairly typical for wild-type flies sleep deprived using the SNAP.
Sleep homeostasis is characterized both by increased sleep duration and by increased sleep depth during the recovery period following deprivation. Daytime sleep consolidation is commonly used as a readout of sleep depth. Sleep consolidation can be assessed as the average sleep bout duration over the entire day (Figure 2E). However, as sleep pressure is dissipated during recovery, the average sleep bout duration will be reduced as the day progresses. Thus, it is frequently helpful to also examine changes in the maximum sleep bout duration which can provide a more sensitive metric (Figure 2F).
Method of Sleep Deprivation | Total # of papers | % papers / technique | Avg recovery evaluated |
SNAP | 52 | 37.14% | 33 ± 3 |
Vortexer/Random Shaking | 49 | 35.00% | 18 ± 3 |
Hand-Deprivation | 9 | 6.43% | 36 ± 11 |
Thermogenetic SD | 15 | 10.71% | 36 ± 12 |
Unspecified | 15 | 10.71% | 29 ± 10 |
Table 1: Survey of different methods of sleep deprivation used in the literature. Only 116 /254 papers used sleep deprivation. The number of papers using each method = "Total # of papers". The fraction of papers using each method = "% papers / technique". The mean length of recovery evaluated for each method = "Avg recovery evaluated". SD – Sleep deprivation. SNAP – Sleep Nullifying Apparatus
Length of SD | Total studies |
< 6 h | 12 |
6 h | 23 |
>6 h & < 12 h | 17 |
12 h | 69 |
>12 h & <24 h | 7 |
24 h | 19 |
> 24 h | 9 |
Chronic SD | 4 |
Any SD | 160 |
Table 2. Length of sleep deprivation performed in different studies. SD – Sleep deprivation
Figure 1. The Sleep Nullifying APparatus (SNAP). A) Front view of the apparatus. The SNAP can accommodate 8 activity monitors in two rows; holder pins restrain the monitors in place. The legs can be adjusted to help position the apparatus at the correct orientation. B) Closeup view of the motor and spring that rock the apparatus back and forth. The motor turns a cam that tilts the apparatus back to the "up" position and compresses the spring. Release of the spring from compression snaps the apparatus back to the "down" position. C) Left – Side view of the apparatus in the "down" position. Holder pins restrain monitors; a monitor cord slot ensures that monitor cords are held in place. Pads help cushion the impact of the apparatus snapping to the 'down' position. Right – Close view up of the cam. D) Left Side view of the apparatus in the "up" position. Right – The counter-clockwise rotation of the cam tilts the apparatus into the 'up' position. Please click here to view a larger version of this figure.
Figure 2. Experimental results. A) Sleep plots of Cs flies for the four days of the experiment: the baseline day (bs), sleep deprivation day (sd), and two days of recovery (rec1 and rec2). B) Average locomotor activity counts of flies on the day of sleep deprivation. Flies were sleep deprived from hours 12-24. C) Time course of sleep deprivation and recovery. Cs flies were sleep deprived from hours 12 – 24, and allowed to recover from hour 24 – 72. The SNAP effectively deprived flies of >98% sleep, which was partially recovered over 48 h (n = 12 flies, Repeated Measures ANOVA for time, F [70,1470]=12.97, p < 10-15). D) Percentage of sleep recovered over 48 h. Flies recovered ~20% of their sleep over 12 h, and ~36% of their sleep over 48 h. E) Sleep consolidation for each day of the experiment as measured by average sleep bout duration during the day. Sleep is more consolidated on the first recovery day compared to baseline (p <0.05, t-test). F) Sleep consolidation for each day of the experiment as measured by maximum sleep bout duration during the day. Sleep is more consolidated on the first recovery day compared to baseline (p <0.05, t-test). Please click here to view a larger version of this figure.
Sleep in Drosophila was independently characterized in 2000, by two groups4,5. In these pioneering studies, flies were deprived of sleep by gentle handling (i.e., hand deprivation) and shown to exhibit a robust homeostatic response to overnight sleep deprivation. Importantly, with any sleep deprivation experiment it is crucial to control for potential confounding effects of the method used to keep the animal awake. Hand deprivation studies set the benchmark for studies of fly homeostasis as a minimally disruptive means of sleep depriving flies. The SNAP efficiently deprives flies of sleep of >98% of nighttime sleep, and importantly induces a sleep rebound comparable to that obtained with hand deprivation4,7.
Since the foundational studies defining sleep in flies, a number of methods have been developed to evaluate sleep homeostasis in flies in a high throughput manner7,9,39,40,41. We surveyed ~250 papers on sleep in flies and found ~46% of these published articles reported using sleep deprivation to evaluate sleep regulation or function (Table 1). A number of different methods effectively induced a sleep rebound in flies. Interestingly, of the studies that have evaluated sleep rebound, the protocols used for sleep deprivation and sleep rebound differed. Specifically, both the duration of sleep deprivation (Table 2) and duration for which rebound was evaluated (Table 1) varied substantially, potentially complicating comparisons of results obtained with different protocols. Sleep rebound in flies is known to persist for up to 48hours following sleep deprivation5. Accordingly, we think a thorough description of the effects of a given sleep manipulation on homeostasis are best obtained when homeostatic rebound is evaluated over a 48 h recovery period.
It is important to note that depriving flies of sleep during the day does not consistently increase sleep drive4. Hence, starting a 24 h sleep deprivation protocol at lights-on and continuing until the next day would not additionally enhance recovery sleep compared to a 12 h sleep deprivation protocol beginning at lights-off. In fact, the calculated sleep rebound may be lower since it will include non-homeostatically regulated daytime sleep in addition to nighttime sleep. The observation that daytime sleep deprivation does not induce a homeostatic rebound can however be used to control for potential confounding effects of the method of sleep deprivation. Thus, flies sleep deprived overnight in the SNAP are compared to flies that receive a comparable stimulus in the daytime7.
In addition to being used for total sleep deprivation, by changing the frequency of the stimulus, the SNAP can also be used to chronically restrict and fragment sleep7,42, thus mimicking conditions of chronic sleep loss in humans. Further, by delivering stimuli in steps of increasing frequency, the SNAP can also be used to measure arousal thresholds8. The SNAP is thus a facile way to effectively deprive and restrict sleep of flies, evaluate the homeostatic response, and measure other sleep characteristics.
The SNAP can fit in a standard laboratory fly incubator, but will definitely disturb flies in the incubator that are not part of the experiment. Fortunately, the SNAP can be placed in an isolated location to sleep deprive flies without disturbing other ongoing experiments. Since recovery sleep is fragile, care should be taken to ensure that recovery sleep takes place in a quiet location.
Complementing studies of sleep deprivation, genetic and pharmacological tools have been developed to enhance sleep in flies8,43,44. Thus, the ability to readily modulate sleep bidirectionally will allow fly sleep research to continue to provide deep insights into sleep regulation and function.
The authors have nothing to disclose.
This work was supported by NIH grants 5R01NS051305-14 and 5R01NS076980-08 to PJS.
Locomotor activity tubes | |||
Fisher Tissue Prep Wax | Thermo Fisher | 13404-122 | Wax used for sealing tubes |
Glass tubes | Wale Apparatus | 244050 | We cut 5mm diameter Pyrex glass tubes into 65mm long tubes to record sleep. Pre-cut tubes can also be purchased. |
Nutri Fly Bloomington Formulation fly food | Genesee Scientific | 66-113 | Labs might use their own fly food recipe. It is important that sleep be recorded on the same food that flies were reared in. |
Rotary glass cutting tool | Dremel Multi Pro | 395 | Used to cut 65mm long glass tubes |
Monitoring Sleep | |||
DAM System and DAMFileScan software | Trikinetics | Software used to acquire data from DAM monitors and save the acquired data in an appropriate format | |
Data acquisition computer | Lenovo | Idea Centre AIO3 | A equivalent computer from any manufacturer can substitute |
Drosophila Activity Monitors | Trikinetics | DAM2 | These monitors are used to record flies' locomotor activity |
Environment Monitor | Trikinetics | DEnM | Not essential, but an easy way to monitor environmental conditions in the chamber where sleep is recorded |
Light Controller | Trikinetics | LC4 | A convenient way to control the timing of when the SNAP is turned on and off |
Power Supply Interface Unit for DAM | Trikinetics | PSIU-9 | Required for data acquisition computers to record Trikinetics locomotor acitvity data |
RJ11 connector | 7001-64PC | Multicomp | DAM monitors accept RJ11 jacks |
Splitters | Trikinetics | SPLT5 | Used to connect upto 5 DAM monitors |
Telephone cable wire | Radioshack | 278-367 | Phone cables to acquire data from DAM monitors |
Sleep Deprivation | |||
Power supply | Gw INSTEK | GPS-30300 | Power supply for the SNAP |
Sleep Nullifying Apparatus | Washington University School of Medicine machine shop |