Long Wake/Short Sleep Bouts and Hyperactivity with Advanced Age in a Mouse Model of Early Onset Alzheimers Disease

This study demonstrates that aged AppNL-G-F mice, a model of early-onset Alzheimer's disease, exhibit a progressive phenotype of hyperactivity, fragmented sleep with long wake bouts, and reduced NREM/REM sleep—particularly in females—suggesting that amyloid-beta deposition impairs state transition mechanisms and validating this strain for investigating early-stage interventions.

The Gist

The Big Picture: A Mouse Model for Alzheimer's Sleep Trouble

Imagine your brain is a bustling city. In a healthy city, there are clear schedules: the workers (neurons) rest at night and work during the day. But in Alzheimer's disease, the city gets clogged with trash (amyloid plaques), and the traffic lights start malfunctioning.

This study looked at a special group of mice called AppNL-G-F mice. Think of these mice as "genetic twins" to humans with early-onset Alzheimer's. They carry a specific genetic mutation that causes them to build up that "trash" (amyloid) in their brains much faster than normal mice.

The researchers wanted to answer a simple question: As these mice get older and their brains get "clogged," how does their sleep change?

The Main Discovery: The "Insomnia" Mice

The researchers found that as these Alzheimer's mice aged, they developed a very specific sleep problem that looks a lot like severe insomnia in humans.

The Analogy: The Broken Switch
Imagine a light switch in your bedroom.

• Normal Mice: When they get tired, they flip the switch to "Sleep." They stay asleep for a long time, then flip it back to "Wake" when they are rested.
• Alzheimer's Mice: Their switch is sticky. Once they flip it to "Wake," it gets stuck. They stay awake for much longer than they should. Conversely, when they try to flip it to "Sleep," it's hard to get it to stay on.

The Results:

• Long Wake, Short Sleep: The Alzheimer's mice spent way more time awake and way less time sleeping than their healthy cousins.
• Hyperactivity: The older Alzheimer's mice were like hamsters on a sugar rush. They were incredibly active, especially at night (when mice are supposed to be most active, but too active). This constant running around kept them awake.
• The "Short Bout" Problem: Instead of sleeping in long, deep chunks, the healthy mice took short naps. The Alzheimer's mice, however, had very long stretches of being awake and very short, fragmented stretches of sleep. It's like trying to sleep on a train that stops and starts every 30 seconds.

The Gender Gap: Why Females Suffer More

One of the most striking findings was the difference between male and female mice.

The Analogy: The Storm
Imagine a storm hitting two houses.

• Male Mice: The storm (Alzheimer's pathology) hit them, and their sleep got a little messy, but they mostly held together.
• Female Mice: The storm hit them harder. Their sleep completely fell apart. They were awake more often, slept less, and had more trouble staying asleep.

The study confirmed that female Alzheimer's mice had a much more severe sleep disorder than the males. This mirrors what we see in human medicine: women are more likely to develop Alzheimer's and often experience more severe sleep issues as the disease progresses.

Did They Feel Tired? (The Nap Test)

You might think, "If they are awake so much, they must be exhausted!"

The researchers gave the mice a "Nap Test" (similar to a human sleep study where you try to fall asleep in a quiet room). Surprisingly, the Alzheimer's mice were not actually sleepier than the healthy mice. They could fall asleep just as fast when given the chance.

The Takeaway: Their problem wasn't that they couldn't sleep; it was that their brains couldn't stay asleep. Their internal "sleep switch" was broken, causing them to wake up too easily and stay awake too long, even though they weren't necessarily exhausted.

The "Homeostasis" Check (The Sleep Bank)

To see if their "sleep bank" was working, the researchers kept the mice awake for 6 hours (sleep deprivation) and then let them sleep.

• The Result: Both the sick mice and the healthy mice tried to "pay back" their sleep debt. They slept deeper and longer to recover.
• The Meaning: This is good news! It means the basic machinery that says, "I need sleep," is still working in these mice. The problem isn't that they don't need sleep; the problem is that the disease is preventing them from getting it.

Why Does This Matter?

1. It's a Warning Sign: Sleep problems might be one of the very first signs of Alzheimer's, appearing even before memory loss gets bad.
2. Women Need Special Attention: Because female mice (and women) are hit harder by these sleep issues, treatments might need to be tailored specifically for them.
3. A New Target for Cures: Since the "sleep switch" seems to be the issue, doctors might be able to develop drugs that help fix the switch, allowing patients to sleep better. Better sleep might even slow down the progression of the disease itself.

Summary

This study tells us that in this mouse model of early Alzheimer's, the brain's ability to switch between "awake" and "asleep" gets broken. The mice become hyperactive and stay awake too long, especially the females. While their bodies still know they need sleep, their brains can't seem to hold onto it. This suggests that fixing sleep could be a powerful way to fight Alzheimer's in its early stages.

Technical Summary

1. Problem Statement

Alzheimer's Disease (AD) is characterized by a bidirectional relationship with sleep: sleep disturbances are both a symptom of AD pathology and a potential driver of disease progression (specifically $\beta$-amyloid accumulation). While previous studies have characterized sleep in younger AD mouse models (6–12 months), there is a critical gap in understanding how sleep/wake architecture evolves as the disease progresses into advanced age and how biological sex influences these phenotypes. Specifically, it is unclear if the "long wake/short sleep" phenotype observed in human AD patients is recapitulated in aged mouse models and if this phenotype is exacerbated in females, mirroring the higher prevalence and severity of AD in human women.

2. Methodology

The study utilized the App$^{NL-G-F}$ mouse model, a knock-in strain expressing three human APP mutations (Swedish, Arctic, Iberian) associated with early-onset AD, compared against wild-type (WT) littermates.

• Experimental Design: Two cohorts were studied using continuous 14-day EEG/EMG recordings to capture baseline sleep architecture, activity, and body temperature ($T_{sc}$).
  • Cohort 1 (Longitudinal): Mice were recorded at 14–18 months and again at 18–22 months to assess age-related progression. Following the second recording, mice underwent a Multiple Sleep Latency Test (MSLT) to assess sleepiness and a 6-hour Sleep Deprivation (SD) protocol to test sleep homeostasis.
  • Cohort 2 (Cross-sectional): Mice were recorded at 21–24 months. Data from Cohort 1 (at 18–22 months) and Cohort 2 were combined for a robust sex-comparison analysis (Total N=31: 17 App$^{NL-G-F}$ and 14 WT).
• Data Acquisition & Analysis:
  • Physiology: Subcutaneous telemetry measured locomotor activity and body temperature.
  • Scoring: Automated machine learning (Somnivore) classified 10-second epochs into Wake, NREM, and REM sleep. Bout durations and bout counts were analyzed.
  • Spectral Analysis: EEG power density was analyzed across frequency bands (Delta, Theta, Alpha, Beta, Gamma) for each state.
  • Histology: Post-mortem immunohistochemistry (82E1 for A$\beta$ plaques, Iba1 for microglia) quantified amyloid burden and neuroinflammation in five brain regions.
• Statistical Approach: Mixed-effects model ANOVAs were used to account for repeated measures, with factors including Genotype, Age, Sex, and Time of Day (Light vs. Dark phases).

3. Key Contributions

• Longitudinal Aging Profile: The study provides the first detailed longitudinal characterization of sleep in App$^{NL-G-F}$ mice from mid-life (14 months) to advanced age (24 months), demonstrating a progressive worsening of the sleep phenotype.
• Sex-Specific Disparities: It rigorously quantifies that the sleep disruption in this AD model is significantly more severe in females than in males, particularly regarding REM sleep loss and wakefulness, aligning with human epidemiological data.
• Mechanistic Insight into State Transitions: The data suggests that the primary defect is not sleepiness (homeostatic drive) but a failure in state transition mechanisms, characterized by prolonged wake bouts and fragmented sleep.
• Validation of Sleep Homeostasis: The study demonstrates that despite severe sleep loss, the sleep homeostat (response to deprivation) remains intact in aged App$^{NL-G-F}$ mice, distinguishing this model from models of advanced neurodegeneration where homeostasis fails.

4. Key Results

A. Pathology Confirmation

• At 22 months, App$^{NL-G-F}$ mice showed significant A$\beta$ deposition (highest in the retrosplenial cortex) and widespread microglial activation (Iba1) compared to WT, confirming the model's validity for early-stage AD pathology.

B. Sleep Architecture and Activity (The "Long Wake/Short Sleep" Phenotype)

• Increased Wakefulness: App$^{NL-G-F}$ mice spent significantly more time awake and less time in NREM and REM sleep compared to WT at both ages.
• Age Progression: At 14–18 months, deficits were primarily restricted to the dark phase. By 18–22 months, the phenotype expanded to include the light phase, indicating a loss of circadian rhythmicity in sleep regulation.
• Hyperactivity: App$^{NL-G-F}$ mice exhibited hyperactivity, which increased with age, particularly during the dark phase. This hyperactivity correlated with longer wake bouts.
• Bout Analysis:
  • App$^{NL-G-F}$ mice had fewer short wake bouts (<60 sec) and **significantly more long wake bouts** (>260 sec) compared to WT.
  • This suggests a difficulty in initiating sleep or terminating wakefulness rather than simple sleep fragmentation.

C. Sex Differences (18–24 Months)

• Females: Female App$^{NL-G-F}$ mice exhibited the most severe phenotype: significantly more wakefulness, less NREM, and significantly less REM sleep compared to female WT. They also showed reduced EEG spectral power in the theta range during REM.
• Males: Male App$^{NL-G-F}$ mice also showed increased wakefulness and reduced NREM/REM, but the deficit in REM sleep was less pronounced than in females. Interestingly, male App$^{NL-G-F}$ mice showed longer NREM and REM bouts (consolidated sleep) but fewer total bouts, whereas females showed more fragmentation.

D. Sleep Homeostasis and Sleepiness

• MSLT: No difference in sleep latency or sleep accumulation was found between App$^{NL-G-F}$ and WT mice, indicating that App$^{NL-G-F}$ mice are not sleepier despite sleeping less.
• Sleep Deprivation: Following 6 hours of SD, both strains showed a robust rebound in NREM sleep and increased Delta power, indicating that the sleep homeostat is functional even in aged, amyloid-laden mice.

E. EEG Spectral Changes

• NREM: Female App$^{NL-G-F}$ mice showed reduced spectral power in the 4.88–10.50 Hz range (alpha/beta) compared to WT.
• REM: Both sexes showed reduced theta power and a decreased theta/delta ratio, consistent with REM sleep instability.
• Wake: Spectral power in wakefulness declined with age in both strains, but App$^{NL-G-F}$ mice showed lower alpha power during wakefulness at 18–22 months.

5. Significance and Conclusion

• Model Validity: The App$^{NL-G-F}$ mouse is a robust model for the prodromal/early stages of AD. It recapitulates the "long wake/short sleep" phenotype and hyperactivity seen in human AD patients but lacks the severe EEG slow-wave reduction and tauopathy of late-stage AD.
• Sex as a Critical Variable: The study confirms that biological sex is a major determinant of AD severity in this model. The exacerbated phenotype in females supports the use of this strain to investigate why women are at higher risk for AD and to develop sex-specific interventions.
• Mechanism of Disruption: The findings suggest that A$\beta$ pathology impairs the mechanisms governing state transitions (specifically the ability to switch from wake to sleep and maintain REM), rather than causing a primary deficit in sleep drive.
• Therapeutic Implications: Since sleep homeostasis remains intact, interventions targeting the arousal systems (e.g., hypocretin/orexin or GABAergic pathways) or state transition mechanisms could potentially mitigate sleep disruption and, by extension, slow amyloid accumulation in early AD.

In summary, this paper establishes that aged App$^{NL-G-F}$ mice develop a progressive, sex-dependent insomnia-like phenotype driven by hyperactivity and impaired state transitions, providing a critical preclinical platform for testing sleep-based interventions in early Alzheimer's disease.