Noninvasive sleep monitoring in large-scale screening of knock-out mice reveals novel sleep-related genes

Shreyas Joshi, Mansi Sethi, Martin Striz, Neil Cole, James M. Denegre, Jennifer Ryan, Michael E. Lhamon, Anuj Agarwal, Steve Murray, Robert E. Braun, David W. Fardo, Vivek Kumar, Kevin D. Donohue, Sridhar Sunderam, Elissa J. Chesler, Karen L. Svenson, Bruce F. O’HarabioRxiv 2019 · DOI

TL;DR

This study leverages a high-throughput, non-invasive piezoelectric system (PiezoSleep) to screen over 6,000 single-gene knockout mice, identifying 122 novel genes that influence sleep duration and bout length. This large-scale approach reveals the complex genetic underpinnings of sleep and opens avenues for understanding its fundamental mechanisms and potential therapeutic targets.

What Problem Does This Paper Address?

Despite sleep's critical importance for mammalian health and function, its underlying genetic mechanisms and molecular pathways remain largely mysterious. Traditional methods for studying sleep in mice, such as EEG/EMG recordings, are time-consuming and expensive, limiting the scale of genetic screens. This bottleneck hinders the discovery of new genes and pathways involved in sleep regulation. The authors aim to overcome this by employing a high-throughput, non-invasive phenotyping system to conduct a large-scale genetic screen for sleep-related genes.

Methods

The core of this study is the use of the PiezoSleep system, a non-invasive, high-throughput method for monitoring sleep and wake parameters in mice. This system utilizes piezoelectric films embedded in the cage floor to detect subtle movements, including breathing patterns during sleep, which are used to differentiate between sleep and wake states. The system has been validated against EEG-based scoring with over 90% accuracy.

The researchers analyzed data from over 6,000 mice, representing 343 single-gene knockout lines generated on a C57BL/6N background. These mice were monitored for five days. The analysis involved:

  • Sleep-wake parameter measurement: Total sleep duration, average sleep bout lengths, and breath rate during sleep were recorded.
  • Statistical analysis: One-way ANOVA and Dunnett's post-hoc tests were used to identify significant differences between knockout strains and controls. Multifactorial ANOVA was employed to assess sex-specific effects and genotype-sex interactions.
  • Principal Component Analysis (PCA): Used to reduce the dimensionality of the sleep data and identify key patterns.
  • Multivariate outlier analysis (Mahalanobis distance): Applied to identify genes with the most extreme effects on sleep variables.
  • Gene association analysis: GeneMANIA was used to explore associations with known circadian and sleep-related genes.
  • Cross-trait analysis: The IMPC database was analyzed to identify coincident phenotypes for top candidate genes.

Datasets

The primary dataset used in this study consists of sleep and wake data collected from 6,350 mice across 343 different single-gene knockout (KO) strains, along with 1,884 control mice (918 female, 894 male). These mice were part of the International Mouse Phenotyping Consortium (IMPC) and the Jackson Laboratory's KOMP2 pipeline.

Key Insights and Ideas

  • Sleep is highly polygenic: The study's high "hit rate" (approximately one-third of tested knockouts showing altered sleep) suggests that a significant portion of the genome may influence sleep, challenging the idea of a small set of "sleep-specific" genes.
  • Circadian phase-specific gene effects: Many identified genes affect sleep only during specific times of day (light or dark phases), highlighting the temporal complexity of sleep regulation.
  • Sex differences are pervasive: Significant sex differences were observed in both normal and altered sleep behaviors, underscoring the importance of considering sex in sleep research.
  • Pleiotropy is common: Candidate genes often exhibit multiple phenotypes beyond sleep, suggesting that sleep is intricately linked with other physiological processes.
  • Non-invasive, high-throughput screening is powerful: The PiezoSleep system enabled a large-scale, unbiased screen, uncovering novel sleep-related genes that might have been missed by traditional methods.

Key Results

  • 122 novel genes identified: The study identified 122 genes influencing sleep duration and bout length that had not been previously implicated in sleep.
  • Significant hit rate: Approximately 35% of the tested gene knockouts showed some degree of altered sleep.
  • Sex differences in controls: Female control mice exhibited reduced sleep duration and shorter bout lengths compared to males.
  • Specific gene effects:
    • Pitx3 knockout mice showed markedly reduced sleep percent and bout length specifically during the light phase.
  • Ppp1r9b and Ap4e1 were identified as highly significant candidate genes affecting both sleep percent and bout length across all phases.
  • Kcnh3 (a potassium channel subunit) was significant for sleep percent in females and associated with known circadian pathway genes.
  • Breath rate variations: 24 genotypes showed significantly lower breath rates during sleep, with some also affecting sleep parameters.

Limitations and Future Work

  • Subtle effects may be missed: The extensive filtering of sleep signals might have excluded genes that affect sleep in very subtle ways or those with compensatory mechanisms.
  • Knockout limitations: The standard knockout method ablates genes in all tissues, which can lead to non-viable animals or mask essential gene functions. The use of conditional alleles could overcome some of these limitations.
  • Direct vs. indirect effects: It remains unclear whether observed sleep alterations are direct or indirect effects of the gene in question. Further investigation of multiple phenotypes for each knockout is suggested.
  • Unbiased screening utility: The study highlights the value of unbiased gene selection for phenotyping, as many identified genes were not initially selected for sleep-related research.
  • Future directions: The authors propose further investigation of the identified candidate genes, particularly Pitx3 and Ppp1r9b, to uncover new signaling pathways and gain a deeper understanding of sleep regulation. They also suggest exploring the role of eye morphology and function in sleep maintenance, given the co-occurrence of vision/eye phenotypes with sleep alterations in some candidate genes. The potential for developing novel therapeutic targets for sleep disorders is also emphasized.

Summary updated

Abstract

Abstract Sleep is a critical process that is well-conserved across mammalian species, and perhaps most animals, yet its functions and underlying mechanisms remain poorly understood. Identification of genes and pathways that can influence sleep may shed new light on these functions. Genomic screens enable the detection of previously unsuspected molecular processes that influence sleep. In this study, we report results from a large-scale phenotyping study of sleep-wake parameters for a population of single-gene knockout mice. Sleep-wake parameters were measured using a high throughput, non-invasive piezoelectric system called PiezoSleep. Knockout mice generated on a C57BL6/N (B6N) background were monitored for sleep and wake parameters for five days. By analyzing data from over 6000 mice representing 343 single gene knockout lines, we identified 122 genes influencing traits like sleep duration and bout length that have not been previously implicated in sleep, including those that affect sleep only during a specific circadian phase. PiezoSleep also allows assessment of breath rates during sleep and this was integrated as a supplemental tool in identifying aberrant physiology in these knockout lines. Sex differences were evident in both normal and altered sleep behavior. Through a combination of genetic and phenotypic associations, and known QTLs for sleep, we propose a set of candidate genes playing specific roles in sleep. The high “hit rate” demonstrates that many genes can alter normal sleep behaviors through a variety of mechanisms. Further investigation of these genes may provide insight into the pathways regulating sleep, functional aspects of sleep, or indirect potentially pathological processes that alter normal sleep.
BibTeX
@misc{joshi2019noninvasive,
  author = {Shreyas Joshi and Mansi Sethi and Martin Striz and Neil Cole and James M. Denegre and Jennifer Ryan and Michael E. Lhamon and Anuj Agarwal and Steve Murray and Robert E. Braun and David W. Fardo and Vivek Kumar and Kevin D. Donohue and Sridhar Sunderam and Elissa J. Chesler and Karen L. Svenson and Bruce F. O’Hara},
  title = {{Noninvasive sleep monitoring in large-scale screening of knock-out mice reveals novel sleep-related genes}},
  year = {2019},
  doi = {10.1101/517680},
  howpublished = {bioRxiv},
}

Loading related content...

Figure