The paradox of sleeper’s disconnection from outside world

Get Complete Project Material File(s) Now! »

The function of sleep disconnection

Disconnection as conservation

Sleep may render prey more vulnerable to their predators. Yet, it has also been suggested that immobility could actually make the sleeper less detectable than during wakefulness. Behavioral inactivity during sleep can be compared to tonic immobility, a widespread form of adaptive behavior that feigns death in the face of danger (Overeem et al., 2002). Shared mechanisms have even been proposed for tonic immobility and REM sleep, which are both characterized by muscular atonia (Tsoukalas, 2012). Thus, immobility may help to escape predation (Meddis, 1975). Overall, this hypothesis explains why sleep might be linked to behavioral inactivity but does not explain why sleep would be homeostatically regulated.
Sleep has been compared to hibernation, a period of disengagement towards the external environment to save energy (Figure 2). This parallel is buttressed by the similarity of the physiological phenomena happening in both states such as reduction of temperature, elevated arousal thresholds and reduction in cerebral energy consumption (Maquet et al., 1990; Nofzinger et al., 2002). In this respect, sleep has been described as a form of adaptive inactivity (Siegel, 2009). The fact that humans sleep at night time fits with this hypothesis as our reliance on daylight vision makes sensory processing at night costly (Hobson & Friston, 2012).
More drastically, prolonged activity has been demonstrated to be deadly. Sleep deprivation experiments during which animals are maintained awake, for example by using electric shocks to prevent them from falling sleep, revealed that a few days of sustained wakefulness leads to death by metabolic and immunological failures (Bergmann et al., 1989; Everson et al., 1989). It has been argued that the dramatic effect of sleep restriction is not tied to sleep privation per se, but rather to the stress induced by sleep deprivation protocols (Rial et al., 2007; but see Rechstschaffen & Bergmann, 2002; Rattenborg et al., 2007). Thus, it has been argued that the function of sleep would be nothing but rest. Being biologically programmed and forced upon organism, sleep would avoid a deadly risk of energy overconsumption by inducing the regular occurrence of a period of inactivity, notably linked to a reduced responsiveness to external stimuli (Moruzzi, 1969). Following this hypothesis, unresponsiveness accomplishes an essential function for survival by preventing exhaustion.
Homeostasis and disconnection are thus explained and intrinsically linked.
Figure 2 Regulation of EEG activity and brain temperature in diverse inactivity states. A continuum is outlined between sleep, torpor and hibernation based on its effects on brain temperature and EEG activity. Torpor is a daily phenomenon associated with food scarcity, while hibernation is a seasonal phenomenon associated with winter survival and reproduction opportunity. Importantly, torpor is followed by a sleep rebound, suggesting that these functions do not involve core features of sleep that are associated with homeostatic regulation. Actually, these functions might come at the expense of sleep in extreme conditions of survival. From Siegel, 2009 and Harding et al., 2019

Disconnection as restoration

Even if sleep and waking rest share behavioral similarities, fundamental differences can be observed upon looking into internal activity. Widespread slow-wave activity is typically observed during sleep but not during rest. Reinforcing the amplitude and occurrence of slow-waves through auditory stimulations in normal sleep benefits immune functions (Besedovsky et al., 2017). SWA present both during sleep and anesthesia has also been related to the metabolic clearance of the brain (Xie et al., 2013). These evidences indicate that sleep does not passively reduce overactivity during wakefulness. It additionally includes mechanisms that restore awake functions and actively protect from the consequences of prolonged wakefulness.
Sleep deprivation leads to consequences on the organism and also on cognitive abilities. Attention and memory functions are typically impaired after prolonged wakefulness and are restored after a recuperation sleep (Killgore, 2010). To explain how sleep might serve the homeostatic regulation of cognitive functions, Tononi and Cirelli have proposed the synaptic homeostatic hypothesis (SHY) (Tononi & Cirelli, 2003). By reducing synaptic strength during sleep and resetting it at a baseline level, the metabolic needs associated with neural functioning are reduced and the efficiency of synaptic transmission is enhanced (Figure 3A). Slow waves have been proposed to play a key role in this process (Tononi & Cirelli, 2003).
In support to this hypothesis, activity in specific brain networks during wakefulness results in an amplification in SWA over the same brain region during the subsequent night (Huber et al., 2004). Conversely, less active brain regions present lower SWA during the subsequent night (Huber et al., 2006). Using a sleep deprivation protocol, a build-up of SWA can be observed even during wakefulness while recovery sleep was associated with a gradual decrease in SWA (Vyazovskiy et al., 2011; Figure 3B). This highlights SWA as a marker of homeostatic processes associated with activity in wakefulness and restoration during sleep. SWA is characterized by synchronized periods of neural silencing (Steriade et al., 1993; Figure 3B). After prolonged wakefulness, the intrusion of neuronal silencing in brain networks supporting sensory processing has been causally related to a drop in cognitive performance in response to stimuli (Nir et al., 2017). The presence of sustained SWA would thus cause the disruption of the cognitive processes supporting the behavioral responsiveness to external stimuli during sleep. Sleep disconnection is thus here a consequence of the brain processes involved in sleep homeostatic functions.
Figure 3 Synaptic homeostasis hypothesis. (A) Schematic evolution of synaptic strength across one wake-sleep cycle. Awake activity is tied to the potentiation of certain synapses which supports learning. During sleep, the strength of synapses is uniformly reduced. The global synaptic strength is normalized at a baseline level (here W=100). Weak synapses are eliminated while the relative strength of remaining synapses is preserved. From (Diekelmann & Born, 2010) (B) Homeostatic regulation of brain activity during and after sleep deprivation in rats. Prolonged wakefulness leads to the increase of low-voltage activity and the presence of concurrent periods of neural silencing in cortical activity. These electrophysiological events manifest the intrusion of sleep like patterns during wakefulness. Conversely, SWA activity is reduced over the course of the recovery night, demonstrating the link between SWA and homeostatic processes. LFP, local field potential; MUA, multi-unit activity. From Vyazovskiy et al., 2011

Disconnecting as reprocessing

Other functions may opportunistically benefit from sleep disconnection. Secretion of certain hormones are increased during sleep, e.g. the growth hormone (GH) (Takahashi et al., 1968). GH has a synergic effect on the selection of immunological cells. Yet, GH secretion is not consistently associated with SWA across the life span (Feinberg, 2000). It is actually rather associated with sleep onset rather than SWA per se (Born, 1988). Moreover, GH levels are not elevated when slow waves are enhanced with auditory stimulations (Besedovsky et al., 2017). This suggests that GH secretion benefits from sleep rather than being tied to the specific occurrence of SWA. The action of GH would be indeed optimally performed during sleep, since anti-inflammatory stress hormones active in wakefulness inhibit immune selection (Besedovsky et al., 2012).
Similarly, memory functions have been proposed to opportunistically benefit from the reduction of sensory inputs during sleep (Buszaki, 1989; see Mednick et al., 2011 for a review). It was first proposed that sleep disconnection would help memory recall by reducing the amount of new memories formed during sleep that could interference with pre-existing memory traces (Jenkins & Dallenbach, 1924). Yet, sleep also protects memory from interferences occurring after awakening. This process, called memory consolidation, relies on the transfer of memory traces temporarily encoded in hippocampal activity to the cortex (McGaugh, 2000; Rasch & Born, 2013, Figure 4A). Models predict that consolidation would be optimally performed “offline”, that is in absence of sensory inputs that can alter memory traces during this transfer (McClelland et al., 1995). It is supported by the synchronized activity of brain rhythms between hippocampal ripples, thalamo-cortical spindles, and cortical slow waves (Peyrache et al., 2009; Staresina et al., 2015, Figure 4B). While memory consolidation also happens during waking rest, it would optimally benefit from the sensory disconnection of NREM sleep.
REM sleep has also been demonstrated to be beneficial for memory reprocessing, especially at the emotional level (Walker & van der Helm, 2009; Boyce et al., 2016), as well as for creativity (Wagner et al., 2004). A continuity between wakefulness and dream has been proposed insofar as dreams would consist in an intensified form of mind-wandering (Fox et al., 2013). Similarly, as dreams, mind-wandering has been associated with emotional memory processing and creativity (Baird et al., 2012; Killingsworth & Gilbert, 2010). Yet, during wakefulness, mind-wandering leads to behavioral impairment (Smallwood et al., 2007). Memory reprocessing and creative thinking happening during REM sleep would thus be optimally performed when the brain is disconnected as they could lead to behavioral deficits if performed during wakefulness.
Disconnection is central in the realization of sleep functions, explaining why it has been selected during evolution. In the following section, we thus hypothesize that sleep disconnection can be informative regarding the sleep process. We will study variations of vigilance defined as the ability to have sustained attention towards external signal and potential threats (Oken et al., 2006). Vigilance can be broken down as a process involving:
– Stimulus encoding, allowing to detect external signals
– Stimulus selection, allowing to prioritize informative signals
– Stimulus integration, allowing to track regularities and discontinuities in sensory information We will limit the scope of our investigation to human sleep.
Figure 4 System consolidation during NREM sleep.
(A) Schematic model of memory consolidation. During wakefulness, cellular consolidation, i.e. the reinforcement of synaptic strength forming a memory trace, ensure the storage of memory traces. During sleep, traces stored in the hippocampus are transferred to the neocortex, forming a trace that is resistant to interferences. Adapted from Mednick et al., 2011
(B) Oscillatory correlates of the hippocampal-thalamo-neocortical dialogue supporting systems consolidation. A hierarchical nesting of oscillation allows the effective synchronization of brain regions involved in the transfer of memory information. From Rasch & Born, 2013

READ  p-robust multilevel and domain decomposition methods with optimal stepsizes for mixed nite element discretizations of elliptic problems 

The structure-function of sleep disconnection

Sensory detection

First investigation of sleep disconnection showed that louder sounds are needed for waking up a sleeper in the beginning of the night as compared to the end of the night (Kohlschütter, 1863; Basner, 2010). Elevated arousal threshold was hypothesized to be associated with the restorative functions of sleep (Michelson, 1899) and was typically associated with SWA (Rechtschaffen et al., 1966). Periods of elevated arousal thresholds during sleep have been referred to as deep sleep, while periods during which it is easier to wake up the sleeper have been called light sleep. A division of labor between light and deep sleep has been proposed. While deep NREM sleep, also called slow-wave sleep (SWS), would thus be rather dedicated to the function of sleep recovery, such as synaptic downscaling, light NREM sleep would be dedicated to more opportunistic functions of sleep such as memory consolidation and dreaming activity, linking thus the realization of sleep functions to variations of vigilance (Genzel et al., 2014).
A similar division of labor for REM sleep has been recently proposed (Simor et al., 2020). REM sleep contains period of eye movement activity called phasic REM sleep (pREM). pREM has been associated with enhanced and more vivid dreaming activity, as compared to periods devoid of eye movements called tonic REM sleep (tREM) (Berger & Oswald, 1962; Goodenough et al., 1959; Molinari & Foulkes, 1969). Arousal thresholds are elevated in pREM as compared to tREM (Price & Kremen, 1980; Ermis et al., 2010). Thus, it was proposed that tREM would favor sensory alertness while pREM would favor inner processes at the expense of sensory processing (Simor et al., 2020). Accordingly, neuroimaging revealed that responses to sounds were conserved in tREM sleep but reduced in pREM sleep (Wehrle et al., 2007). Distinguishing between pREM and tREM shows the importance of taking into account micro-physiology to go beyond classical sleep stage classification when studying sleep disconnection.
During NREM sleep, sensory stimulations are known to elicit K-complexes, a large deflection in cerebral activity within the thalamocortical network (Davis et al., 1939; Bastien & Campbell, 1992; Figure 5). A K-complex consists in an early activation of sensory areas, called an up-state, that is associated with sensory encoding (Figure 5). Then, a large negative deflection called a down-state follows and is associated with the silencing of cortical neurons (Cash et al., 2009). Finally, a second up-state is observed and promotes information processing (Halász, 2005). K-complexes differs from slow waves insofar as down states are isolated during K-complexes while they alternate with up states during slow waves (Cash et al., 2009). During slow waves, sensory processing critically depends on the timing of stimulation with the ongoing phase of the slow oscillation. Indeed, stimuli played during up-states are encoded in cortical activity and can boost memory consolidation (Shimizu et al., 2018; Göldi et al., 2019). Conversely, stimuli played during down states are unreliably processed and fail to influence memory consolidation (Schabus et al., 2012; Batterink et al., 2016). Additionally, sleep spindles, a fast-oscillatory thalamo-cortical activity in the 12-16 Hz range, have been involved in processes of memory consolidation (Schabus et al., 2004; Cairney et al., 2018). Spindles have been associated to an enhanced tolerance to external sounds and reduced auditory processing (Dang-Vu et al., 2010; Schabus et al., 2012). External stimuli thus affect internal sleep processes but this effect crucially depends on the presence of micro-physiological markers of sleep functions.

Table of contents :

1.1 The paradox of sleeper’s disconnection from outside world
1.1.1 The limits of introspective approaches of sleep
1.2 The function of sleep disconnection
1.2.1 Disconnection as conservation
1.3 The structure-function of sleep disconnection
1.3.1 Sensory detection
2.1 Experimental procedure
2.1.1 Task and stimuli
2.2 Studying sleep physiology
2.2.1 Neuroimaging of sleep cognition
2.3 Studying sensory processing during sleep
2.3.1 Inducing a task during sleep
3.1 Introduction
3.1.1 Is sleep learning real?
3.2 Methods
3.2.1 Stimuli
3.3 Results
3.3.1 Behavioural results
3.4 Discussion
3.4.1 Dynamics of sleep learning
4.1 Introduction
4.1.1 Sensory disconnection during REM sleep
4.2 Methods
4.2.1 Material and procedure
4.3 Results
4.3.1 Maintenance of selective processing during REM sleep
4.4 Discussion
4.4.1 Mechanism of the gating of external information during REM sleep
5.1 Introduction
5.1.1 Sensory processing during deep sleep
5.2 Methods
5.2.1 Procedure
5.3 Results
5.3.1 Behavioural results
5.4 Discussion
5.4.1 Evidence for sensory integration
6.1 Insights into sleep functions
6.1.1 Sensory processing across vigilance states
6.2 Breaking through sleep disconnection
6.2.1 Using complex stimuli to probe vigilance
6.3 Outlook
6.3.1 Consciousness science


Related Posts