Method for Selective REM Sleep Deprivation using The Selective Nose Flicking for Forgetting REM (SNiFFer)

Summary

We present a novel protocol for the selective deprivation of Rapid Eye Movement (REM) sleep in rodents.

Abstract

Sleep is broadly categorized into Rapid Eye Movement (REM) and non-REM stages, and various sleep deprivation techniques have been developed to study their functions. Traditional methods—such as mechanical stimulation or dropping the animal during REM sleep—often lead to rapid habituation and unintended stress effects, making interpretation difficult. Here, we introduce a novel, mild, and gentle method for REM sleep deprivation in rodents: SNiFFer. Rodents are known to be highly sensitive to nasal stimuli. In this method, a soft paintbrush gently strokes the animal's nose during REM sleep, reliably triggering arousal. To selectively apply this technique during REM sleep, we utilized an AI-based real-time Electroencephalography (EEG) sleep-stage classification system. The intervention was triggered only during REM episodes, and a yoked control group was included, receiving the same stimulation pattern regardless of sleep state. We further applied SNiFFer in a Trace Fear Conditioning paradigm to test whether REM sleep deprivation within 6 h after learning interferes with memory consolidation. Compared to yoked controls, SNiFFer-treated animals showed a significant decrease in total REM episode and REM episode duration, indicating the method's specificity and efficacy. In conclusion, the SNiFFer technique offers a highly selective, minimally invasive tactile approach for REM sleep deprivation in mice. It enables precise control of experimental conditions and opens new possibilities for investigating REM sleep's role in memory and cognition—with a ticklish twist.

Introduction

Rapid eye movement (REM) sleep plays a pivotal role in cognitive processes such as memory consolidation^1,2. To investigate the functional significance of REM sleep in the brain, REM sleep deprivation is commonly employed as an experimental approach. Currently, four primary methods are widely used to induce REM sleep deprivation in rodents: the rotating disk method¹, the flowerpot method³, the air-puff method⁴, and physical stimulation techniques^5,6. While the rotating disk and flowerpot methods are highly effective in eliminating REM sleep, they often introduce substantial stress to the animal, such as water exposure or physical fatigue⁷, which can confound the interpretation of behavioral and physiological outcomes. Similarly, the air-puff and conventional physical stimulation methods tend to deliver broad, non-specific stimuli to the body, resulting in variability in stimulus intensity and limited choice of control groups^4,8.

We conducted a comprehensive literature comparison spanning more than six decades of REM sleep deprivation research which is provided in Supplementary File 1 and Supplementary Table 1. We incorporated foundational historical, mechanistic, neurochemical, hormonal, and behavioral studies from the major research lines, including Jouvet, Rechtschaffen, Bergmann, Tufik, Andersen, and Borbély. Supplementary Table 1 organizes key papers chronologically. It traces the methodological evolution of REM sleep deprivation techniques and summarizes their principal biological, neurochemical, hormonal, immune, thermoregulatory, and behavioral findings. Since the present manuscript is structured as a methods article rather than a comprehensive historical review, this has been added as supplementary information.

To address these limitations, we developed a novel method offering high selectivity for REM sleep deprivation. This technique is designed to selectively deprive rodents of REM sleep in a precisely controlled manner. In this approach, we limit the stimulus to the animal’s nose—one of the most sensitive areas in rodents⁹. In addition, we employ a real-time sleep stage classification system by AI to detect and interrupt REM sleep episodes with temporal specificity¹⁰. This AI system refers to the EEG signal in the last 4 s epoch from a single channel and evaluates the current mouse sleep stages. This setup also enables the implementation of yoked control groups, allowing for temporally matched intervention. Although categorized as a physical stimulation-based REM sleep deprivation technique, our protocol differs from conventional methods. First, we applied real-time sleep stage analysis using AI for accurate and consistent REM-specific intervention. Second, it limits the stimulation region to only the top of the nose, which is one of the most sensitive areas in mice⁶, resulting in minimal variation in stimulation duration and strength. Lastly, we newly set up a yoked control group that received identical amounts of stimulation to control for confounding factors. This protocol proposes a method for investigating functions specific to REM sleep.

To evaluate the utility of this method, we applied it to investigate the role of REM sleep in trace fear memory consolidation. Previous studies have reported that REM sleep deprivation impairs hippocampus-dependent fear memory consolidation, including the delayed fear memories^11,12. However, there is conflicting evidence regarding the impact of REM sleep deprivation on memory processes that engage distributed and complex brain circuits^5,13. Trace fear memory consolidation requires the association of a footshock (unconditioned stimulus), an auditory cue (conditioned stimulus), and the trace interval between them—a process known to recruit multiple brain regions, including the hippocampus, medial prefrontal cortex, and amygdala¹⁴. Therefore, we deprived mice of REM sleep for 6 h immediately following trace fear conditioning¹⁵, a time window known to be critical for memory consolidation¹⁶. Using this protocol, we assessed the effect of REM sleep deprivation on the consolidation of memory traces that depend on the coordinated activity of distributed neural systems. This article presents a novel method describing a detailed experimental protocol testing the effects.

Protocol

All animal experiments were approved by the University of Tsukuba Institutional Animal Care and Use Committee. Male and female C57BL/6J mice (11–13 weeks old, 20–35 g) were used in this study. The complete list of materials and instruments used in this study is provided in the Table of Materials.

1. Surgical implantation of EEG electrodes¹⁷

1. Anesthesia and preparation of the animal
   1. Induce anesthesia with 4% isoflurane mixed with air in 3 min. Adequate anesthesia was confirmed by the absence of the pedal withdrawal reflex and muscle tone, along with stable respiratory patterns.
   2. Shave the dorsal head and neck region from the nape to the area between the eyes using an electric shaver. When a mouse indicates wakefulness such as whisker movement, induce anesthesia again.
2. Head fixation and skull exposure
   1. Secure the animal in a stereotaxic frame, ensuring nasal and cranial alignment. From this time to the end of the whole surgical procedure, anesthesia was continuously maintained via an isoflurane vaporizer pump system.
   2. Apply ophthalmic ointment to both eyes to prevent corneal drying. Remove any remaining hair with a depilatory cream.
   3. Make a midline scalp incision and gently retract the skin to expose the skull from the posterior ear line to anterior eye line. Remove the mucosal tissue with cotton swab.
3. Landmark identification and leveling
   1. Identify bregma and lambda; adjust the stereotaxic arm so that these points are level within ±0.05 mm. Verify that the lateral coordinates ±3.0 mm from bregma are also level (±0.05 mm).
4. Drilling and screw placement
   1. Mark the following coordinates relative to bregma. Drill four pilot holes as these coordinates. Anterior–Posterior / Medial–Lateral: +1.5 mm / ±1.5 mm, –3.0 mm / ±1.7 mm.
   2. Implant EEG recording screws into each hole and secure them with cyanoacrylate-based gel glue. Bundle and cover the exposed skull and leads with a thin layer of dental cement.
5. Postoperative care
   1. Administer 0.5 mL of 5% glucose solution and 0.13 mL of ibuprofen (or equivalent analgesic).
   2. Return the mouse to its home cage, a cylindrical Plexiglas cage (21.9 cm in diameter, 31.6 cm in height) placed inside an insulated chamber (45.7 x 50.8 x 85.4 cm). The chamber is maintained at a temperature of 23.5 ± 2.0 °C, a relative humidity of 50.0% ± 10.0%, and a 12 h light/dark cycle (light phase: 9:00 AM to 9:00 PM).
   3. Place the animal in a warmed recovery cage. Wait in front of the cage until a mouse gets awake. Following surgery, animals were maintained in individual housing and were not returned to group housing for the duration of the experiment.

2. Handling and habituation

1. Schedule for experimentation
   1. Begin handling during light phase 24–48 h after surgery. Continue for four consecutive days.
2. Operative procedure
   1. On day 1, put a mouse on hand for 2 min for 2x. Place the mouse on the experimenter’s open palm with fingers extended and the palm facing upward. Position the palm slightly higher than the forearm, as mice naturally tend to climb to elevated surfaces. Allow the mouse to move freely without restraint; if the mouse attempts to climb onto the forearm, gently return it to the palm by lightly holding the base of the tail. Each interval is more than 20 min. This should be conducted during light phase.
      NOTE: If a mouse jumps from the hand or bites it, use cloth or tissue paper to cover the hand only on day one. Clothes can be reused and should be cleaned every 2 weeks.
   2. On day 2–4, put a mouse on hand for 2 min for 3x. Each interval and the conducted period are the same as day 1. Assess habituation behaviorally: non-habituated mice typically attempt to escape by jumping or bite forcefully, whereas habituated mice actively explore the palm with minimal escape attempts and little to no forceful biting.
   3. At the end of the final handling session, connect the mouse to the EEG recording cable.
      NOTE: When bedding material is excessive to hinder a mouse's nose, pick some of them manually and remove them during the handling session or at the connection time.

3. Baseline EEG recording

1. Start from postoperative days 5 and 6. Record continuous EEG from 11:00 to 17:00 each day. If bedding material obstructs access to the mouse’s nose, gradually remove excess bedding to ensure clear nasal access while minimizing disturbance to the animal for further process.

4. Fear conditioning^18,19

1. Begin postoperative day 7 (or 8) between 10:00 and 11:00. As preparation, Spray 70% ethanol into the conditioning chamber (30 x 25 x 24 cm), which consists of clear acrylic walls and a stainless-steel grid floor (2.0 mm diameter rods spaced 6.0 mm apart).
2. Training protocol (total session 1,100 s)
   1. Deliver five auditory cues (20 s ascending sound, 5–20 kHz at 1.25 Hz, 0.8 s duration) at 180 s, 400 s, 620 s, 840 s, and 1060 s. Pair each sound-cue with foot shock (1.0 mA, 2 s) delivered at 218 s, 438 s, 658 s, 878 s, and 1098 s.

5. REM sleep deprivation

1. REM sleep deprivation group (RSD group)
   1. Immediately following fear conditioning (ZT2-ZT8), monitor EEG in real time by activating the AI-based sleep stage classification system to enable real-time detection of REM sleep. Upon detection of REM sleep, gently tickle the animal’s nose with a soft paintbrush until it re-arouses, with an average latency from stimulus onset to arousal of 0.89 ± 0.063 s (mean ± SEM).
   2. During the experiment, arousal was defined as a behavioral response to the tactile stimulus, specifically whisker movement following stimulation. Record each tickle’s timestamp for later analysis.
      NOTE: The tickling movement was performed without a fixed directional pattern (e.g., back-and-forth, upward, or circular), while maintaining consistent light tactile stimulation limited to the top of the nose. If the nose is obstructed, direct stimulation toward the vibrissal (whisker) region, which is typically exposed. If the mouse is facing away from the paintbrush, rotate the cage as needed to ensure unobstructed access to the nasal region. When using a cage with an open top, stimulation may also be applied from above to ensure consistent nasal or vibrissal access.
2. Yoked control group (Yoked group)
   1. Pair each yoked mouse to a REM-deprived counterpart. Apply nose tickling to the yoked mouse in synchrony with its partner’s REM episodes, matching both onset and duration. Each mouse uses the same dedicated paintbrush throughout the REM sleep deprivation experiment. After each use, spray the brush with weakly acidic water and wipe it with tissue. Allow at least a 24 h interval between uses to ensure that the cleaned paintbrush is completely dry.

6. Memory recall tests

1. Contextual recall test (Test A)
   1. Start from 24 h (±3 h) after sleep deprivation. Return the mouse to the original conditioning chamber (with 70% ethanol odor). Record freezing behavior for 5 min.
2. Sound-cued recall test (Test B)
   1. Conduct the session 3 h after Test A.
   2. Context preparation: Spray water in context B (triangular prism (30 cm in height) consisting of three walls with side lengths of 31 cm, 24 cm, and 24 cm. Two of the walls are covered with white paper, and the third wall is made of transparent acrylic through which a blue rectangle is visible in the background. The structure has a plastic floor and a transparent acrylic ceiling).
   3. Record freezing behavior for 6 min. Deliver a single 180 s auditory cue identical to training at 180 s.

7. Data analysis

1. Sleep Stage Scoring (Figure 1C-G)¹⁷
   1. For each Wake/NREM/REM, Calculate total amount, episode number, episode mean duration, and sleep transition number based on human expert-scored data. Criteria of each stages a following.
      Wake: Low-amplitude, high-frequency (β/γ)
      NREM: High-amplitude, low-frequency (δ)
      REM: Low-amplitude, theta-dominant
2. Deprivation quality (Figure 1B)
   1. Evaluate the sleep stage during which each tickle stimulus is applied by comparing its timestamp with EEG data annotated by a human expert.
3. Freezing behavior (Figure 1H,J-K)
   1. Use FreezeFrame4 (Actimetrics) to quantify immobility (excluding respiration) during fear conditioning and recall tests.
      1. Launch FreezeFrame 4 software and select Recorder mode. From the File menu (upper left corner), specify the destination folder for saving the recording files. In the four white input fields on the left panel, enter unique identifiers corresponding to each mouse, ensuring alignment with the four file display windows on the right side of the screen.
      2. Click the Protocol panel at the bottom of the interface and define the recording parameters, including total recording duration and the timing of auditory cues and foot shocks according to the experimental design.
      3. Click Reference in the center of the screen to acquire the background reference image. Place the mouse into the conditioning chamber, then click Start to begin recording.
4. Discrimination index analysis (Figure 1L)
   1. Calculate discrimination index by the following formula:
      freezing in 5 min average in test A - freezing in first 3 min average in test B)/max freezing in 5 min average in test A, freezing in first 3min average in test B.
5. Shock reactivity index analysis (Figure 1I)
   1. Use track by hand to measure the two-dimensional movement of animal 2s before and after shock startpoint in fear conditioning session. Calculate shock reactivity index by the following formula:
      ​movement during shock - movement preshock / movement during shock + movement preshock

Results

Surgical implantation of EEG electrodes
Mice should be fully anesthetized during the surgery. If they exhibit signs of failure of anesthesia, such as struggling, they should be removed from the protocol. Similarly, if bleeding lasts for more than 10 s, it may cause severe damage to the mouse’s brain, and removal from experimentation is recommended. After the surgery, most mice typically wake up within 1 h. Their gait and eye movement should return to normal, similar to pre-surgical conditions.

Discussion

This protocol enables selective REM sleep deprivation in mice with yoked control mice. Unlike traditional methods such as the flowerpot and rotating disk techniques, this system allows the implementation of finely matched control groups, in which animals receive identical stimulation at the same circadian time. Furthermore, this approach minimizes variability in stimulation location, a common limitation of conventional gentle handling and air puff methods. As a result, this technique may permit a more precise isolation o...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
5% glucose	Otsukba	35061410
6-pin header	Hirose	172-0989
Adhesive	Konishi	16351
Anaesthetic machine	Shinano Seisakusho	SN-487-0T
Crimp housing	Hirose	688-9095
Crimp socket	Hirose	688-8982
Freezeflame4	Med Associates	RRID:SCR_014429	Version 4.104
GraphPad Prism	GraphPad	RRID:SCR_002798	Version 10.5.0
Ibuprofen	Tokyo Chemical Industry	I0415
Isoflurane	Viatris	901036504
Loctite454	Henkel	UFI:  MAK2-7WRP-G202-F235
Mouse (C57BL/6J)	The Jackson Laboratory	RRID: IMSR_JAX:000664	11.0 - 13.0 week of age at fear conditioning session
Paintbrush	Namurataiseidou	4943668000448	SDflat No.8, Tip length: 8.2 mm; width: 18.8 mm
Provinice liquid	Shofu	21400BZZ00451000	250 mL
Provinice powder	Shofu	21400BZZ00451000	250 g
SleepSignRecorder	KISSEI COMTEC	RRID:SCR_018200	Version 1.2.10
Stainless steel screw	Yamazaki	N/A	φ1.0 × 2.0
Standard Stereotaxic Instruments	ALA Scientific	68038
Sunflower oil	Sigma-Aldrich	S5007-250ML
White petrolatum	Taiyo Pharmaceutical	K09-TS