Functional genomics of sleep

High-throughput visual assessment of sleep stages in mice using machine learning

STUDY OBJECTIVES

Sleep is an important biological process that is perturbed in numerous diseases, and assessment its substages currently requires implantation of electrodes to carry out electroencephalogram/electromyogram (EEG/EMG) analysis. Although accurate, this method comes at a high cost of invasive surgery and experts trained to score EEG/EMG data. Here, we leverage modern computer vision methods to directly classify sleep substages from video data. This bypasses the need for surgery and expert scoring, provides a path to high-throughput studies of sleep in mice.

METHODS

We collected synchronized high-resolution video and EEG/EMG data in 16 male C57BL/6J mice. We extracted features from the video that are time and frequency-based and used the human expert-scored EEG/EMG data to train a visual classifier. We investigated several classifiers and data augmentation methods.

RESULTS

Our visual sleep classifier proved to be highly accurate in classifying wake, non-rapid eye movement sleep (NREM), and rapid eye movement sleep (REM) states, and achieves an overall accuracy of 0.92 +/- 0.05 (mean +/- SD). We discover and genetically validate video features that correlate with breathing rates, and show low and high variability in NREM and REM sleep, respectively. Finally, we apply our methods to non-invasively detect that sleep stage disturbances induced by amphetamine administration.

CONCLUSIONS

We conclude that machine learning based visual classification of sleep is a viable alternative to EEG/EMG based scoring. Our results will enable non-invasive high-throughput sleep studies and will greatly reduce the barrier to screening mutant mice for abnormalities in sleep.

Genetic dissection of sleep homeostasis

Sleep is a complex behavior both in its manifestation and regulation, that is common to almost all animal species studied thus far. Sleep is not a unitary behavior and has many different aspects, each of which is tightly regulated and influenced by both genetic and environmental factors. Despite its essential role for performance, health, and well-being, genetic mechanisms underlying this complex behavior remain poorly understood. One important aspect of sleep concerns its homeostatic regulation, which ensures that levels of sleep need are kept within a range still allowing optimal functioning during wakefulness. Uncovering the genetic pathways underlying the homeostatic aspect of sleep is of particular importance because it could lead to insights concerning sleep's still elusive function and is therefore a main focus of current sleep research. In this chapter, we first give a definition of sleep homeostasis and describe the molecular genetics techniques that are used to examine it. We then provide a conceptual discussion on the problem of assessing a sleep homeostatic phenotype in various animal models. We finally highlight some of the studies with a focus on clock genes and adenosine signaling molecules.

Obstructive Sleep Apnoea: From pathogenesis to treatment: Current controversies and future directions

Obstructive sleep apnoea (OSA) is a common disease, recognized as an independent risk factor for a range of clinical conditions, such as hypertension, stroke, depression and diabetes. Despite extensive research over the past two decades, the mechanistic links between OSA and other associated clinical conditions, including metabolic disorders and cardiovascular disease, remain unclear. Indeed, the pathogenesis of OSA itself remains incompletely understood. This review provides opinions from a number of leading experts on issues related to OSA and its pathogenesis, interaction with anaesthesia, metabolic consequences and comorbidities, cardiovascular disease, genetics, measurement and diagnosis, surgical treatment and pharmacotherapeutic targets.

What are microarrays teaching us about sleep?

Many fundamental questions about sleep remain unanswered. The presence of sleep across phyla suggests that it must serve a basic cellular and/or molecular function. Microarray studies, performed in several model systems, have identified classes of genes that are sleep-state regulated. This has led to the following concepts: first, a function of sleep is to maintain synaptic homeostasis; second, sleep is a stage of macromolecule biosynthesis; third, extending wakefulness leads to downregulation of several important metabolic pathways; and, fourth, extending wakefulness leads to endoplasmic reticulum stress. In human studies, microarrays are being applied to the identification of biomarkers for sleepiness and for the common debilitating condition of obstructive sleep apnea.

Drosophila melanogaster: an insect model for fundamental studies of sleep

In 2000, Drosophila melanogaster joined the ranks of vertebrates and invertebrates with a defined behavioral sleep state. The characterization of this sleep state revealed striking similarities to sleep in humans: sleep in flies has both circadian and homeostatic components, it is influenced by sex and age, and it is affected by pharmacological agents such as caffeine and antihistamines. As in mammals, arousal thresholds in flies increase with sleep deprivation. Furthermore, changes in brain electrical activity accompany the change from wake to sleep states. Not only do flies and vertebrates share these behavioral and physiological traits of sleep, but they are likely to share at least some genetic mechanisms underlying the regulation of sleep as well. This article reviews the methods currently used to identify and characterize the Drosophila sleep state. As these methods become more refined and our understanding of Drosophila sleep more detailed, the powerful techniques afforded by this organism are likely to unveil deep insights into the function(s) and regulatory mechanisms of sleep.

Unearthing the phylogenetic roots of sleep

Why we sleep remains one of the enduring unanswered questions in biology. At its core, sleep can be defined behaviorally as a homeostatically regulated state of reduced movement and sensory responsiveness. The cornerstone of sleep studies in terrestrial mammals, including humans, has been the measurement of coordinated changes in brain activity during sleep measured using the electroencephalogram (EEG). Yet among a diverse set of animals, these EEG sleep traits can vary widely and, in some cases, are absent, raising questions as to whether they define a universal, or even essential, feature of sleep. Over the past decade, behaviorally defined sleep-like states have been identified in a series of genetic model organisms, including fish, flies and worms. Genetic analyses in these systems are revealing a remarkable conservation in the underlying mechanisms controlling sleep behavior. Taken together, these studies suggest an ancient origin for sleep and raise the possibility that model organism genetics may reveal the molecular mechanisms that guide sleep and wake.

Identification of coding polymorphisms in human circadian rhythm genes PER1, PER2, PER3, CLOCK, ARNTL, CRY1, CRY2 and TIMELESS in a multi-ethnic screening panel

STUDY OBJECTIVE

In this study, the exonic regions of the circadian rhythm genes PER1, PER2, PER3, CLOCK, ARNTL, CRY1, CRY2 and TIMELESS were re-sequenced and coding changes identified in a panel of 95 individuals varying in ethnicity.

STUDY PARTICIPANTS

DNA screening panel consisting of 95 DNA samples (17 American Caucasians, 17 African Americans, 8 Ashkenazi Jews, 8 Chinese, 8 Japanese, 5 Mexican Indians, 8 Mexicans, 8 Northern Europeans, 8 Puerto Ricans, and 8 South Americans) selected from the Coriell Institute Human Variation Panel.

RESULTS

In addition to coding changes already identified in the database dbSNP, novel coding changes were identified, including PER1: Pro37Ser, Pro351Ser, Gln988Pro, Ala998Thr; PER2: Leu83Arg, Leu157Leu, Thre174Ile, Phe400Phe, Pro822Pro, Ala828Thr, Ala861Val, Phe876Leu, Val883Met, Val903Ile, Ala923Pro; PER3: Pro67Pro, Val90Ile, His638His, Ala820Ala, Leu929Leu; ARNTL: Arg166Gln, Ser459Phe; CLOCK: Ala34Ala, Ser208Cys, Phe233Phe, Ser632Thr, Ser816Ser; TIMELESS: Met870Val and CRY2: His35His. No coding polymorphisms were identified in CRY1.

CONCLUSIONS

Considerable genetic variation occurs within the coding region of the genes regulating circadian rhythm. Many of the non-synonymous coding polymorphisms could affect protein structure/function with the potential to affect molecular regulation of the sleep/wake cycle. Many of the potential functional effects could be ethnic group specific.

Circadian regulation of sleep in mammals: Role of the suprachiasmatic nucleus

Despite significant progress in elucidating the molecular basis for circadian oscillations, the neural mechanisms by which the circadian clock organizes daily rhythms of behavioral state in mammals remain poorly understood. The objective of this review is to critically evaluate a conceptual model that views sleep expression as the outcome of opponent processes-a circadian clock-dependent alerting process that opposes sleep during the daily wake period, and a homeostatic process by which sleep drive builds during waking and is dissipated during sleep after circadian alerting declines. This model is based primarily on the evidence that in a diurnal primate, the squirrel monkey (Saimiri sciureus), ablation of the master circadian clock (the suprachiasmatic nucleus; SCN) induces a significant expansion of total daily sleep duration and a reduction in sleep latency in the dark. According to this model, the circadian clock actively promotes wake but only passively gates sleep; thus, loss of circadian clock alerting by SCN ablation impairs the ability to sustain wakefulness and causes sleep to expand. For comparison, two additional conceptual models are described, one in which the circadian clock actively promotes sleep but not wake, and a third in which the circadian clock actively promotes both sleep and wake, at different circadian phases. Sleep in intact and SCN-damaged rodents and humans is first reviewed, to determine how well the data fit these conceptual models. Neuroanatomical and neurophysiological studies are then reviewed, to examine the evidence for direct and indirect interactions between the SCN circadian clock and sleep-wake circuits. Finally, sleep in SCN-ablated squirrel monkeys is re-examined, to consider its compatibility with alternative models of circadian regulation of sleep. In aggregate, the behavioral and neurobiological evidence suggests that in rodents and humans, the circadian clock actively promotes both wake and sleep, at different phases of the circadian cycle. The hypersomnia of SCN-ablated squirrel monkeys is unique in magnitude, but is not incompatible with a role for the SCN pacemaker in actively promoting sleep.

A genetic analysis of ambulatory cardiorespiratory coupling

This study assessed the heritability of ambulatory heart period, respiratory sinus arrhythmia (RSA), and respiration rate and tested the hypothesis that the well-established correlation between these variables is determined by common genetic factors. In 780 healthy twins and siblings, 24-h ambulatory recordings of ECG and thorax impedance were made. Genetic analyses showed considerable heritability for heart period (37%-48%), RSA (40%-55%), and respiration rate (27%-81%) at all daily periods. Significant genetic correlations were found throughout. Common genes explained large portions of the covariance between heart period and RSA and between respiration rate and RSA. During the afternoon and night, the covariance between respiration rate and RSA was completely determined by common genes. This overlap in genes can be exploited to increase the power of linkage studies to detect genetic variation influencing cardiovascular disease risk.

A whole-genome scan for 24-hour respiration rate: a major locus at 10q26 influences respiration during sleep

Identification of genes causing variation in daytime and nighttime respiration rates could advance our understanding of the basic molecular processes of human respiratory rhythmogenesis. This could also serve an important clinical purpose, because dysfunction of such processes has been identified as critically important in sleep disorders. We performed a sib-pair-based linkage analysis on ambulatory respiration rate, using the data from 270 sibling pairs who were genotyped at 374 markers on the autosomes, with an average distance of 9.65 cM. Uni- and multivariate variance-components-based multipoint linkage analyses were performed for respiration rate during three daytime periods (morning, afternoon, and evening) and during nighttime sleep. Evidence of linkage was found at chromosomal locations 3q27, 7p22, 10q26, and 22q12. The strongest evidence of linkage was found for respiration rate during sleep, with LOD scores of 2.36 at 3q27, 3.86 at 10q26, and 1.59 at 22q12. In a simultaneous analysis of these three loci, >50% of the variance in sleep respiration rate could be attributed to a quantitative-trait loci near marker D10S1248 at 10q. Genes in this area (GFRA1, ADORA2L, FGR2, EMX2, and HMX2) can be considered promising positional candidates for genetic association studies of respiratory control during sleep.

C57BL/6J and B6.V-LEPOB mice differ in the cholinergic modulation of sleep and breathing

Respiratory and arousal state control are heritable traits in mice. B6.V-Lep(ob) (ob) mice are leptin deficient and differ from C57BL/6J (B6) mice by a variation in the gene coding for leptin. The ob mouse has morbid obesity and disordered breathing that is homologous to breathing of obese humans. This study tested the hypothesis that microinjecting neostigmine into the pontine reticular nucleus, oral part (PnO), of B6 and ob mice alters sleep and breathing. In B6 and ob mice, neostigmine caused a concentration-dependent increase (P < 0.0001) in percentage of time spent in a rapid eye movement (REM) sleeplike state (REM-Neo). Relative to saline (control), higher concentrations of neostigmine increased REM-Neo duration and the number of REM-Neo episodes in B6 and ob mice and decreased percent wake, percent non-REM, and latency to onset of REM-Neo (P < 0.001). In B6 and ob mice, REM sleep enhancement by neostigmine was blocked by atropine. Differences in control amounts of sleep and wakefulness between B6 and the congenic ob mice also were identified. After PnO injection of saline, ob mice spent significantly (P < 0.05) more time awake and less time in non-REM sleep. B6 mice displayed more (P < 0.01) baseline locomotor activity than ob mice, and PnO neostigmine decreased locomotion (P < 0.0001) in B6 and ob mice. Whole body plethysmography showed that PnO neostigmine depressed breathing (P < 0.001) in B6 and ob mice and caused greater respiratory depression in B6 than ob mice (P < 0.05). Western blot analysis identified greater (P < 0.05) expression of M2 muscarinic receptor protein in ob than B6 mice for cortex, midbrain, cerebellum, and pons, but not medulla. Considered together, these data provide the first evidence that pontine cholinergic control of sleep and breathing varies between mice known to differ by a spontaneous mutation in the gene coding for leptin.