Local awakening: Regional reorganizations of brain oscillations after sleep

Highlights

• Local awakening concept was proved by regional specificity and sleep regulation.

• Between-network variability was shown in both connectivity and ALFF after sleep.

• Thalamo-cortical synchronizations were enormously enhanced after sleep.

• Slow-wave sleep leads to a stronger reduction of spectral power on awakening.

Abstract

Brain functions express rhythmic fluctuations accompanied by sleep and wakefulness each day, but how sleep regulates brain rhythms remains unclear. Following the dose-dependent local sleep concept, two succeeding questions emerge: (1) is the sleep regulation a network-specific process; and (2) is the awakening state dependent on the previous sleep stages? To answer the questions, we conducted simultaneous EEG and fMRI recordings over 22 healthy male participants, along pre-sleep, nocturnal sleep and awakening. Using paired comparisons between awakening and pre-sleep conditions, three scenarios of the regional specificity were demonstrated on awakening: (1) the default-mode and hippocampal networks maintained similar connectivity and spectral power; (2) the sensorimotor network presented reduced connectivity and spectral power; and (3) the thalamus demonstrated substantially enhanced connectivity to the neo-cortex with decreased spectral power. With regard to the stage effect, the deep sleep group had significant changes in both functional connectivity and spectral power on awakening, whereas the indices of light sleep group remained relatively quiescent after sleep. The phenomena implied that slow-wave sleep could be key to rebooting the BOLD fluctuations after sleep. In conclusion, the regional specificity and the stage effect were verified in support of the local awakening concept, indicating that sleep regulation leads to the reorganization of brain networks upon awakening.

Introduction

Previous descriptions of sleep have recognized it as an offline period of consciousness. However, studies have shown that the simple dichotomy between sleep and wakefulness is invalid (Edgar et al., 1993, Steriade et al., 2001). From the neuroscience perspective, the literature suggests that sleep is not a quiescent offline state with minimal brain activities, but is composed of intensive variations of spatio-temporal oscillations across the brain (Dang-Vu et al., 2010, Massimini et al., 2005). The distributed oscillating signals originate from the spontaneous molecular process, internally transferring essential information, forming synaptic plasticity and network reorganization founded by multiple sleep stages (Adam and Oswald, 1977, Karni et al., 1994, Walker and Stickgold, 2006). Detected by electroencephalography (EEG) and polysonography, the sleep stages were believed to facilitate the reorganization and restoration of cerebral physiology (Aserinsky and Kleitman, 1953, Iber, 2007). The spontaneous propagation of these temporal features, such as spindles, slow-wave activities, and rapid eye movement (REM) activity, during sleep leads to the energy conservation and reorganization of attention and memory (Hu et al., 2006, Walker and Stickgold, 2006). Although researchers have made advances in understanding what occurs during sleep, how much sleep is optimal and how to measure sleep regulation become primary challenges for further sleep investigations (Roenneberg, 2013). Once determined, the sleep quality measurement can provide positive and pervasive benefits to most patients with sleep disorders, or even to the healthy public.

Beyond investigating the intrinsic nature of sleep architecture, an alternative viewpoint for estimating sleep regulations is to detect the recovery condition after sleep (Hayashi et al., 2010, Voss, 2010). Awakening from sleep not only suggests the restoration of consciousness and body energy, but also reflects the synchronous integration of multiple neural assemblies (Mignot, 2008). Therefore, the brain activity on awakening has the potential to detect sleep efficiency and daytime arousal level (Hayashi et al., 2010, Jung et al., 2014). However, public attention on awakening remains focusing on the temporary decrement of subsequent cognitive performance immediately after sleep, or “sleep inertia” (Tassi and Muzet, 2000), so the importance of the awakening condition was underestimated. Ferrara observed an increased EEG power in the delta to low-alpha bands (1–9 Hz) and a decreased power in the beta range (18–24 Hz) on awakening (Ferrara et al., 2006). Marzano et al. confirmed an enhanced delta power in the posterior brain, reduced delta power in the frontal regions, and a generalized decrease of beta power on the overall scalp within 10 min after awakening (Marzano et al., 2011). These reports noted that brain functions on awakening possess regional specificity, implying the possibility of estimating sleep regulation effects on brain networks immediately after sleep.

Such regional specificity was also noted during sleep. The “local sleep” concept underpins that cortical activities during sleep reflect regional variations in brain activities while awake (Nir and Tononi, 2011, Vyazovskiy et al., 2011). Huber reported that local slow-wave activity (SWA) distribution in the brain has high correlations with the learning-associated synapse connection before sleep (Huber et al., 2004). Hung demonstrated that under consecutive 24-h language and visuomotor tasks, slow/theta power (2–6 Hz) locally increased within the task-related brain regions (Hung et al., 2013), suggesting that previous waking experiences affect regional sleep patterns. Echoing the local sleep notion, a local awakening concept is founded in the current study with the following two postulations: (1) sleep regulation has regional specificity, imposing the reorganizations of brain dynamics before and after sleep, and (2) cortical activity or network synchrony on awakening is mediated by stage characteristics during sleep. Previously, Ikeda and Hayashi detected arousal-level disparities existing between various task performances on awakening, providing a preliminary evidence for local awakening (Ikeda and Hayashi, 2008). For spatial reorganization, Balkin reported a possible shift in regional cerebral blood flow (rCBF) and brain interregional connectivity after awakening (Balkin et al., 2002). Though previous studies implied the regional changes upon awakening, the specific impact on the brain networks and the link between the awakening condition and the previous sleep contents remain question marks. Therefore, as the first trial to test the local awakening concept, we tested the regional specificity across networks and whether the existence of the slow-wave sleep affects the following awakening condition across local brain networks.

In the current study, two steps were adopted to test the hypothesis and the local awakening concept. The first step is to measure the regional changes of brain dynamics before and after sleep using the resting-state functional magnetic resonance imaging (RS-fMRI) technology for the 2 advantages (Biswal et al., 1995). First, the RS-fMRI technique measures the dynamic disparities across multiple brain networks within a single session (Biswal et al., 2003, Yang et al., 2007). Second, because participants are unable to respond to external stimuli during sleep (Larson-Prior et al., 2009, Sämann et al., 2011), the RS-fMRI technique is suitable for observing transient brain interactions associated with sleep while providing high spatial resolution. In practice, we measured both spectral power and interregional synchronizations before and after sleep across multiple sleep-related brain networks, including the default-mode network (Horovitz et al., 2009, Sämann et al., 2011), the sensorimotor network (Larson-Prior et al., 2009), the hippocampal network (Hu et al., 2006, Walker and Stickgold, 2006) and thalamo-cortical networks (Boveroux et al., 2010). The second step is to test the causal effect of previous stage-3 (N3) sleep on the awakening brain. We segregated the participants into 2 groups, with and without N3 sleep, and observed their BOLD spectral power and functional connectivity for how sleep architecture modulates cerebral organizations on awakening.