SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH
Bachelor Degree Project in Cognitive Neuroscience Basic level 15 ECTS
Spring term 2015
Pernilla Andersson
Supervisor: Oskar MacGregor Examiner: Judith Annett
SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 1
Abstract
Throughout the history of sleep research, many theories have been put forward in an attempt to explain the function of the state of sleep. The most popular and well-accepted theory among sleep researchers today is that sleep plays an important role in learning and memory by affecting mechanisms of synaptic plasticity. At the present time, there are two main competing views regarding how sleep affects memory – the active system consolidation model and the synaptic homeostasis hypothesis (SHY). The former proposes that sleep is a time for the reactivation and redistribution of memory traces, while the latter emphasizes the role of sleep in improving signal-to-noise ratios and learning by down-regulating synaptic strength. In recent years, SHY has been widely discussed by sleep researchers. It has also been criticized for, for example, being too vague, ignoring contradictory evidence and misrepresenting available evidence. The aim of this review is to critically analyze SHY by considering recently published empirical evidence as well as criticism raised against the theory. The conclusion drawn from the analysis is that although SHY seemingly hold up well against recent contradictory empirical evidence, it faces serious problems with the criticism raised against it.
Key words: Sleep; Memory; Learning; Synaptic Homeostasis Hypothesis; SHY
SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 2
Table of Contents
Abstract ...... 1
Introduction ...... 4
Background ...... 9
The Physiology of Sleep ...... 9
The Function of Sleep ...... 11
Memory, Synaptic Plasticity, and Sleep ...... 14
The Synaptic Homeostasis Hypothesis (SHY) ...... 21
Wakefulness and Synaptic Potentiation ...... 22
Synaptic Potentiation and Slow-Wave Homeostasis ...... 24
Slow-Wave Homeostasis and Synaptic Downscaling ...... 26
Synaptic Downscaling and Performance ...... 29
In Support of SHY ...... 32
Critique of SHY ...... 37
The usage of random methods...... 37
The significance of the supporting evidence...... 38
Contradictory evidence ...... 39
The Vagueness of SHY ...... 41
Misunderstandings of the theory...... 42
Alternative Mechanisms Accounting for Synaptic Renormalization during Sleep ...... 43
Brain temperature...... 43
Glucocorticoid concentrations...... 44 SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 3
Controlling temperature...... 45
Inconclusive evidence...... 45
Discussion ...... 46
Future directions ...... 51
References ...... 53
SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 4
Introduction
Sleep has been the subject of scientific investigation for approximately a century and this period has generated several important findings and insights regarding the state (Pelayo &
Guilleminault, 2009). Sleeping has been observed in distantly related animal species, ranging from fruit flies to humans (Cirelli & Tononi, 2008), and occupies a non-trivial proportion of the day in all of these animal species (Abel, Havekes, Saletin, & Walker, 2013). Behaviorally, the state is characterized by immobility, a site preference, a specific posture, rapid reversibility, a higher arousal threshold, homeostatic control (more sleep after sleep deprivation), and a loss of or an altered state of consciousness (Rasch & Born, 2013; Wang,
Grone, Colas, Appelbaum, & Mourrain, 2011).
In the early 1950s, sleep researcher Nathanial Kleitman and his student Eugene
Aserinsky discovered periods of distinct rapid eye-movements during sleep (Aserinsky &
Kleitman, 1953), which initiated the division of mammalian sleep into two distinct states: rapid eye-movement (REM) sleep and non-rapid eye-movement (NREM) sleep. These two states are distinguished from each other by their respective pattern of brain activity measured by electroencephalogram (EEG) (Iber, Ancoli-Israel, Chesson, & Quan, 2007; Carskadon &
Dement, 2011). Observations using this measuring technique in humans has revealed an alteration between NREM and REM sleep every ~90 min during the sleeping period, with
NREM-periods being dominantly expressed during the earlier cycles and REM-periods being more present in the later cycles (Carskadon & Dement, 2011).
Before the second half of the 20th century, sleep was believed to be a passive state, a time during which the brain was “turned off”, which caused it to be regarded as fairly uninteresting (Dement, 1998). After Aserinsky and Kleitman’s (1953) discovery of REM sleep and its association with dreaming the interest in sleep began to increase. The following decades were coupled with several discoveries paramount in establishing the field of sleep SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 5 medicine, such as the discovery of sleep apnea in 1965 (Dement, 1998). Today, pervasive sleep deprivation and undiagnosed sleep disorders are considered one of the largest health problems in the world (Dement, 1998).
Several studies conducted during the latter half of the 20th century revealed that sleep deprivation has many negative effects on an animal. A series of experiments on sleep deprivation in rats, led by sleep researcher Allan Rechtschaffen, revealed that sleep deprivation resulted in serious physiological consequences such as weight loss, highly increased food intake, a decreased body temperature, skin lesions and even death
(Rechtschaffen, Bergmann, Everson, Kushida, & Gilliland, 1989). Similar results have been found in fruit flies, who died as a consequence of sleep deprivation (Shaw, Tononi,
Greenspan, & Robinson, 2002), and in humans suffering from fatal familial insomnia, a rare disease characterized by progressive insomnia, which eventually leads to death (Fiorino,
1996).
But despite a long line of research on sleep indicating that the state serves some need for animals engaging in it, the fundamental function of sleep has not yet been discovered (Abel et al., 2013; Cirelli & Tononi, 2008; Frank & Cantera, 2014). Recent decades have, for instance, seen several different theories emphasizing a role for sleep in regulating physiological functions, such as temperature (Rechtschaffen & Bergmann, 1995), metabolism (Cauter,
Spiegel, Tasali, & Leproult, 2008) or energy conservation (Berger & Phillips, 1995).
However, it has since been shown that these proposed functions, and others of the same nature, can be achieved independently of sleep and that they cannot account for all aspects of sleep (Hobson, 2005). This discovery led to the view that the function of sleep is mainly for the benefit of the brain (Hobson, 2005).
Today, it is widely agreed among sleep researchers that the function of sleep, at least so far as the brain is concerned, has to do with learning and memory (Abel et al., 2013; Ribeiro, SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 6
2012; Wang et al., 2011). This idea is not new. It was developed already in the 1920s, in a study demonstrating the beneficial effects of sleep on word recall rates (Jenkins &
Dallenbach, 1924). However, since sleep was, at that time, given a passive role in memory, it was hypothesized that sleep merely protects memories from interference (Jenkins &
Dallenbach, 1924). Today, a more active role in memory is attributed to sleep, as it is believed to be involved in memory consolidation (Diekelmann & Born, 2010). This idea is supported by evidence demonstrating the beneficial effects of sleep on the consolidation of both declarative and non-declarative (Cohen, Pascual-Leone, Press, & Robertson, 2005;
Ellenbogen, Hulbert, Stickgold, Dinges, & Thompson-Schill, 2006) memories.
The foundation of memory formation is thought to consist of long-term changes in the effectiveness, or strength, of synaptic transmission (Siegelbaum & Kandel, 2013), also called synaptic plasticity. Thus, modern theories of sleep and memory hold that sleep plays an important role in regulating synaptic activity, thereby affecting memory consolidation
(Ribeiro, 2012). The question of precisely how sleep affects synaptic activity is, however, still a matter of debate among sleep researchers today.
One of the most popular account of how sleep affects learning and memory today is the synaptic homeostasis hypothesis (SHY). The primary motivation behind the formulation of
SHY was considerations about how neuronal function would be negatively affected by learning and memory (Tononi & Cirelli, 2012, 2014). One important aspect of neuronal functioning is that they are selective in their firing, which means that they fire more in response to novel event in the environment (Tononi & Cirelli, 2014). SHY predicts that the plastic changes that occur during wakefulness are biased towards potentiation because the theory considers wakefulness as the primary time for learning and predicts that important events increase the strength of synaptic connections rather than decrease it which would reduce the selectivity of neurons and, subsequently, saturate the capacity for plasticity and SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 7 additional learning. Furthermore, because plasticity is an important part in the development of neuronal circuitry as well as the adaptation to a changing environment, the theory argues that if the synaptic changes occurring during wakefulness reduces neuronal selectivity and saturates the capacity for plasticity, these changes would need to be counteracted in some way in order for additional learning to be possible (Tononi & Cirelli, 2003, 2006, 2014).
Subsequently, SHY proposes that sleep has the function of renormalizing the overall synaptic strength in the brain, thus restoring the capacity for learning and benefitting memory.
According to SHY, sleep is an optimal time for synaptic renormalization because the animal is not interacting with the environment and therefore neurons can obtain a sample of the overall knowledge in the brain. This would avoid the selective weakening of synapses underused during a particular day, which would result in the loss of older memories (Tononi
& Cirelli, 2003, 2006, 2014).
SHY differs from other popular, contemporary accounts of the effects of sleep on learning and memory because it emphasizes the beneficial effects of the state on memory encoding (Tononi & Cirelli, 2003, 2006, 2014) while other theories are more focused on how sleep affects memory consolidation (Tononi & Cirelli, 2012, 2014; Diekelmann & Born,
2010; Guiditta, 2014; Rasch & Born, 2013). Moreover, the primary aim of SHY is being a theory which encompasses the function of sleep in all animal species engaging in it, not only humans (Tononi & Cirelli, 2012, 2014), which has not been expressed in other theories about sleep (Diekelmann & Born, 2010; Guiditta, 2014; Rasch & Born, 2013). Due to this, SHY is an interesting and intriguing theory about sleep, because it challenges the current view of the state and has the potential to influence the future of the field of sleep research.
During recent years, SHY has been the subject of great discussion in the field of sleep research (e.g., Abel et al., 2013; Frank, 2012, 2013, 2014; Guiditta, 2014; Rasch & Born,
2013). It has also received much criticism from other sleep researchers for, among other SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 8 things, being too vaguely formulated (Frank, 2013), lacking significant evidential support
(Frank, 2012; Timofeev, 2011), being incompatible with some of the evidence on the effects of sleep on memory (Born & Feld, 2012; Diekelmann & Born, 2010; Timofeev, 2011), and misrepresenting some of the evidence supporting it (Frank, 2012, 2013, 2014). Although, the sleep researchers behind SHY, Guilio Tononi and Chiara Cirelli, have responded to some of this critique, the majority of the criticism still needs to be addressed.
As the predictions of SHY contrast those claimed by other theories about the function of sleep, which are widely accepted by several sleep researchers, these predictions needs to be examined and the validity of the theory needs to be objectively evaluated in order to determine if the present view of the role of sleep in memory needs to be revised. The aim of this thesis is therefore to critically analyze SHY, to determine where the theory stands in the light of experimental findings and criticism published since its original formulation in 2003
(Tononi & Cirelli, 2003). In order to perform this analysis, the thesis will consider evidence providing support for the theory, evidence contradicting it, as well as the criticism raised against it. The thesis begins with a presentation of the relevant background knowledge about sleep, memory and synaptic plasticity. Following this, the core predictions of SHY and the evidence initiating the formulation of the theory will be thoroughly described. Then the evidential support for the theory will be addressed, before describing the existing evidence against the theory. Thereafter, some of the most prominent criticism against SHY is described, as well as Tononi and Cirelli’s response to some of this critique. The thesis concludes with an evaluation of where SHY currently stands in relation to recently published evidence and criticism, as well as some future directions concerning how to accommodate some of the persistent criticism. The conclusion drawn from the analysis is that SHY does not hold up well against the criticism raised against the theory since its original formulation. Even though the theory is supported by a large body of evidence, Tononi and Cirelli still needs to SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 9 properly address some important points of criticism before SHY can be considered a strong scientific theory about the essential, universal function of sleep.
Background
The Physiology of Sleep
Sleep is a state in which the organism is perceptually disengaged from and unresponsive to, the external environment (Carskadon & Dement, 2011). The different stages of sleep are defined by different patterns of brain activity detected using an electroencephalogram (EEG) in animals with developed neocortex (including mammals and birds) (Wang et al., 2011).
Following this definition normal human sleep is divided into two distinct states referred to as rapid eye-movement (REM) sleep and non-rapid eye-movement (NREM) sleep (Carskadon &
Dement, 2011, see Figure 1 & 2). The two sleep-states are differentiated from each other by their respective EEG-activity patterns as well as by eye-movement measured using electrooculogram (EOG) and muscle tone measured using electromyogram (EMG)
(Carskadon & Dement, 2011). REM-sleep is defined by the occurrence of rapid eye- movements, a loss of muscle activity and a low amplitude, mixed frequency EEG-pattern, usually ranging between 2-6 Hz (i.e., sawtooth waves) (Iber et al., 2007; Keenan &
Hirshkowitz, 2011, see Figure 1).
Figure 1. The EEG-pattern of REM sleep. Figure adapted from Carskadon & Dement, 2011.
NREM sleep is further divided into three different stages labelled N1, N2, and N3.
During stage N1, the EEG pattern shows low amplitude activity with mixed frequencies, SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 10 predominantly frequencies between 4-7 Hz (Iber et al., 2007, see figure 2). Stage N2 also shows low voltage, mixed frequency activity but is defined by the occurrence of negative sharp waves that are immediately followed by a positive component that stands out from the background EEG, also referred to as K-complexes, and trains of distinct waves with frequencies between 11-16 Hz, referred to as sleep spindles (Iber et al., 2007, see figure 2).
Stage N3 is defined by high amplitude and low frequency activity (delta activity: < 4 Hz) in the EEG-recordings. The stage is characterized by at least 20% of the activity being waves of frequencies ranging between 0.5-2 Hz often referred to as slow-wave activity (SWA; Iber et al., 2007, see figure 2) and therefore, stage N3 is also often termed slow-wave sleep (SWS)
(Carskadon & Dement, 2011).
Figure 2. The EEG-patterns of the different stages on NREM sleep. Figure adapted from (Carskadon
& Dement, 2011).
In adult humans, the sleeping episode begins with a brief period of N1 sleep then progresses through the deeper stages of NREM-sleep and after about 80-100 minutes, the first
REM-period appears. Then sleep alternates between REM and NREM-periods approximately every 90 minutes throughout the rest of the sleep episode (Carskadon & Dement, 2011). Stage SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 11
N3 of NREM sleep is more highly present in the earlier cycles of sleep while the REM- periods are prolonged towards the end of the sleeping episode (Carskadon & Dement, 2011).
To cite sleep researchers Mary Carskadon and William Dement (the founder of the world’s first sleep laboratory), the specific activity-patterns defining NREM sleep can been summarized as, “…a relatively inactive yet actively regulated brain in a moveable body” and the activity-patterns defining REM-sleep can be summarized as, “… an active brain in a paralyzed body” (Carskadon & Dement, 2011, p. 16).
The Function of Sleep
From an evolutionary perspective sleeping is a highly perplexing state due to it being present in the lives of distantly related animal species and occupying a substantial portion of the day in all of these animals (Abel et al., 2013). One of the most perplexing aspects of sleep appears from looking at the fitness of the animal through an evolutionary perspective. From this point of view, a behavior that has been preserved throughout evolutionary history should be adaptive and have beneficial effects on the fitness of the animal or else it would have been selected against (Abel et al., 2013). Due to sleep being regarded as a universal state present in almost all animal species, it is assumed to have been preserved throughout evolutionary history. Because the behavior renders the organism disengaged from the external environment, it subsequently renders the animal unable to gather resources, reproduce, nourish social bonds, and most importantly, increases the animal’s vulnerability to predatory attacks (Abel et al., 2013). Thus, from an evolutionary perspective, it seems highly probable that sleep serves considerable functional benefits which compensate for the several proposed functional drawbacks that the state imposes on the animal (Abel et al., 2013).
Despite a very long history of research on sleep and a strong indication that sleep has an important function which benefits the fitness of the animal engaging in it, the core function of SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 12 the sleeping state has not yet been identified (Abel et al., 2013; Cirelli & Tononi, 2008; Frank
& Cantera, 2014).
Throughout the history of sleep research many theories have been put forward in an attempt to unravel the core function of sleep. For example, sleep has been proposed to serve a metabolism-regulating function, important for the normal functioning of metabolic processes, hormonal processes and appetite regulation (Cauter et al., 2008); a thermoregulatory function, regulating the temperature in the body (Rechtschaffen & Bergmann, 1995); and an energy- conserving function, in which the decreased body temperature during sleep is claimed to promote the conservation of energy (Berger & Phillips, 1995). Theories of this nature were later disregarded as it was demonstrated that these proposed functions could be achieved through quiet wakefulness (occurring prior to sleep onset) and could not account for the reduced responsiveness to the external environment or the loss of consciousness caused by the induction of the state. This led to the view that the function of sleep is primarily important for the brain (Hobson, 2005).
Several theories have been proposed concerning what exact role sleep might play in the brain. Some have proposed that there is an imbalance between protein synthesis and degradation in the brain during wakefulness, with a greater amount of degradation occurring, and subsequently that sleep has a restorative effect on this imbalance, promoting protein synthesis and thereby helping to restore degraded tissue (Oswald, 1980). Another proposed function of sleep has been that it is energy-conserving, stating that wakefulness is associated with a higher consumption of energy, for example glycogen, and that sleep is a time of protein synthesis that compensates for the high energy-demands of waking (Scharf, Naidoo,
Zimmerman, & Pack, 2008). Others have also proposed that sleeping is a time for removing cerebral free radicals from the brain which have accumulated during wakefulness, by increasing the efficiency of antioxidant mechanisms (Reimund, 1994). SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 13
Among sleep researchers today the most widely accepted hypothesis regarding what the beneficial effects of sleep is on the brain is that the state plays an important role in learning and memory (Abel et al., 2013; Ribeiro, 2012; Wang et al., 2011). The fundamental idea behind this hypothesis is that sleep has an important involvement in the consolidation of memories, whereas wakefulness is of primary importance for memory encoding and retrieval
(Diekelmann & Born, 2010). This postulation is supported by a large body of evidence indicating beneficial effects of sleeping on the consolidation of several different forms of memory (Cohen et al., 2005; Fenn, Nusbaum, & Margoliash, 2003; Ellenbogen, Hulbert,
Jiang, & Stickgold, 2009; Ellenbogen et al., 2006; Gais, Lucas, & Born, 2006; Payne et al.,
2012; Stickgold, James, & Hobson, 2000; Walker, Brakefield, Morgan, Hobson, & Stickgold,
2002).
The idea of a relationship between sleep and memory has been around for a long time.
Some of the earliest evidence for a role of sleep in memory was provided by Jenkins and
Dallenbach (1924), in a study of sleep and forgetting. In this study the participants learned nonsense syllables which they would then try to recall after a period of sleep or wakefulness.
The results showed an increased recall-rate among the participants after periods of sleep as opposed to periods of wakefulness. Jenkins and Dallenbach (1924) interpreted these results as sleep serving a passive role in memory by protecting them from interference.
Since Jenkins and Dallenbach’s study, great progress has been made in understanding both sleep and memory. Researchers have discovered that sleep is not a unified state but is divided into two stages, NREM and REM, as noted above (Carskadon & Dement, 2011). In memory research there has been great progress towards defining the different forms of memory as well as the underlying molecular and cellular processes believed to be involved in memory formation (Kandel & Schwartz, 2013). SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 14
Memory, Synaptic Plasticity, and Sleep
Learning is a change in the behavior of an animal that is caused by the acquiring of knowledge about the world and memory is the process following learning in which this knowledge is encoded, stored and later retrieved. Both these processes are of high importance to the functioning and survival of both humans and other animals (Schacter & Wagner, 2013).
The past several decades of memory research has resulted in significant progress towards understanding the fundamental mechanisms and processes underlying learning and memory.
Among the most important discoveries are the discovery that there are different forms of learning and memory with distinctive cognitive properties, which are mediated by different specific brain systems and the discovery that memory formation is divided into discrete processes of encoding, consolidation and retrieval (Schacter & Wagner, 2013).
Memory research on patients with memory deficiencies has demonstrated that long-term memory includes both unconscious forms of memory, also referred to as implicit or non- declarative memory, such as priming, skill learning, conditioning and habit learning (Schacter
& Wagner, 2013), and conscious forms of memory, also referred to as explicit or declarative memory, such as memories of past events and factual knowledge about oneself or the world
(Schacter & Wagner, 2013).
The progress in memory research during the past decades has also resulted in the indication that learning and memory emerges from elementary properties of synapses which use chemical transmitters to communicate (Siegelbaum, Kandel, & Südhof, 2013). The effectiveness, or strength, of these synapses can be modified for shorter or longer periods of time, something called synaptic plasticity. Synaptic strength can be modified both in the presynaptic neurons, by affecting the release of a neurotransmitter, or in the postsynaptic neuron, by regulating the response to the transmitter, or both (Siegelbaum, Kandel, & Südhof,
2013). Long-term changes in these sorts of pre- and postsynaptic mechanisms are of great SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 15 importance for the development and learning of an organism, and therefore the foundation of memory formation, and thus learning, is currently believed to consist of changes in the effectiveness of neural connections (Siegelbaum, Kandel, & Südhof, 2013).
The storage of information in long-term memory is currently believed to be the result of long-lasting changes in the strength of synaptic connections caused by changes in the structure or function of neurons which affects the effectiveness of the communication between them such as increased/decreased transmitter release, an increased/decreased number of receptors, and dendritic growth or retraction (Siegelbaum & Kandel, 2013).
In 1973, Bliss and Lømo conducted an experiment in which they repetitively stimulated the perforant pathway terminating in the dentate gyrus of the hippocampus in anaesthetized rabbits, and then measured the population response to this stimulation by recording the activity of the granule cells in the dentate gyrus (Bliss & Lømo, 1973). They found that a train of high-frequency stimulation of the perforant pathway increased the efficiency of the synaptic transmission of these synapses, and also increased the excitability of the granule cells in the hippocampus (Bliss & Lømo, 1973). Initially, the authors entitled the observed synaptic changes long-lasting potentiation but later they renamed the process long-term potentiation, or LTP.
Following this, and similar observations, LTP has become one of the most well-studied processes of synaptic plasticity and one of the most well-characterized molecular mechanisms implicated in learning and memory today. Even though LTP has been observed in several brain areas, it has been most thoroughly studied in the hippocampus (Siegelbaum & Kandel,
2013). Studies in the different pathways in the hippocampus have shown that LTP is not a single form of synaptic plasticity but that there are different forms of LTP which can be induced in different brain areas or even in the same synapse (Siegelbaum & Kandel, 2013).
Despite there being some differences between these forms of LTP, they all share the function SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 16 of increasing the strength of the synaptic transmission between neurons by causing a long- lasting increase in the amplitude of the excitatory postsynaptic potentials (EPSPs; making the postsynaptic neuron more likely to fire an action potential in response to stimulation), they all occurs at glutamatergic synapses (synapses using the neurotransmitter glutamate), they all involve the glutamate-receptor N-methyl-D-aspartate, or NMDA, and they can all can cause changes persisting from days to several weeks (Siegelbaum & Kandel, 2013).
The Schaffer collateral is a pathway in the hippocampus which ascends from area CA3 and terminates in area CA1, in which LTP has been extensively studied and well- characterized (Siegelbaum & Kandel, 2013). In the Schaffer collateral pathway, the pore of the NMDA receptors are blocked by extracellular Mg2+ when the membrane of the neuron is at a resting potential or when the neuron is only modestly depolarized (Figure 3; Siegelbaum
& Kandel, 2013). When LTP is induced there is a large burst of strong synaptic activity which opens the α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) glutamate receptors, which generates an EPSP sufficient to trigger an action potential in the postsynaptic neuron. This action potential, in turn, causes a depolarization which removes the blockage of the NMDA receptor. The NMDA receptors are highly permeable to Ca2+, which entails that the opening of these channels causes the influx of Ca2+ in the postsynaptic neuron (Figure 3;
Siegelbaum & Kandel, 2013). The increased Ca2+ activates downstream signaling pathways such as calcium/calmodulin-dependent protein kinase II (CaMKII), which lead to the induction of LTP (Figure 3; Siegelbaum & Kandel, 2013). The LTP in the synapse is then expressed through long-lasting synaptic changes that increases the effectiveness of the synaptic transmission, for example, by creating additional AMPA-receptors (Figure 3;
Siegelbaum & Kandel, 2013).
SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 17
Figure 3. An illustration of (A) normal synaptic transmission, (B) the induction of LTP, and (C) the expression of LTP in the Schaffer collateral pathway of the hippocampus. Adapted from Siegelbaum & Kandel, 2013.. SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 18
The induction of LTP in the Schaffer collateral pathway have several properties which are interesting when considering the role of LTP in learning and memory (Figure 4; Siegelbaum
& Kandel, 2013):
1. Cooperativity. The induction of LTP requires an almost simultaneous activation of a
large number of axons because a large depolarization of the postsynaptic neuron is
needed to remove the blockage of the NMDA-receptors.
2. Associativity. If a weak synaptic input, normally insufficient to cause the required
amount of depolarization needed, is paired with a strong synaptic input it can induce
LTP.
3. Synapse specificity. If a particular synapse is not activated when there is strong
synaptic stimulation, LTP will not be induced in that synapse even if there is strong
postsynaptic depolarization.
These properties underlies essential parts of memory storage. Cooperativity ensures that only events with high significance will be stored in memory; associativity allows events with less significance to be stored in memory if they occur near simultaneously with more significant event, such as the case of Pavlovian conditioning; and synapse specificity ensures that input that are not related to a specific event will not be a part of the memory (Siegelbaum
& Kandel, 2013).
Figure 4. Illustration of three properties of LTP-induction in the Schaffer collateral pathway which have direct relevance to memory storage. Adapted from Siegelbaum & Kandel, 2013. SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 19
If it was only possible for synapses to increase in strength, the synapses would likely increase too much in strength, reaching some sort of end-point beyond which no more strengthening is possible (Siegelbaum & Kandel, 2013). Because regulations of synaptic strength are implicated in learning such a scenario would saturate the ability to learn. But people are able to learn and store new memories throughout their lives which indicates that neurons should also be able to undergo processes during which the synaptic function is downregulated or weakened (Siegelbaum & Kandel, 2013). This sort of mechanism, entitled long-term depression (LTD) has been identified, which results in a weakening of the synaptic strength. LTD is induced by longer periods of low-frequency (~1 Hz) stimulation of neurons, and, like LTP, involves NMDA receptors, but because of the lower frequency stimulation, the postsynaptic depolarization produced by the activity is insufficient to relieve the block of the
NMDA receptors. Subsequently, LTD does not activate CaMKII, but is thought to activate an enzyme complex called calcium-dependent phosphatase calcineurin, which leads to a reduction in the number of postsynaptic receptors, thus reducing the size of the EPSP, and thereby the strength of the synapse (Siegelbaum & Kandel, 2013).
These forms of synaptic plasticity are believed to be involved in the consolidation of both declarative and non-declarative memories, and are some of the best described molecular mechanisms thought to be important for learning and memory (Siegelbaum & Kandel, 2013).
In accordance with the progress that has been made in both sleep research and memory research, there is a wide consensus among sleep researchers today that sleep plays an important part in processes of synaptic plasticity (Ribeiro, 2012). But even though the idea that sleep implicates processes of synaptic plasticity in the brain and thereby plays an important role in memory is largely agreed upon in the field of sleep research, the nature of these changes is not (Diekelmann & Born, 2010; Rasch & Born, 2013; Ribeiro, 2012) and neither is the question of which stages of memory formation are affected by sleep nor which SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 20 forms of memories sleep benefits. At the present time, the two most popular accounts of the role of sleep in memory are the active system consolidation model and the synaptic homeostasis hypothesis (SHY).
The active system consolidation hypothesis is a collective term referring to several, similar, contemporary accounts of the function of sleep. Generally, these accounts of beneficial effects of sleep on memory consolidation postulate that recently acquired memory representations are repeatedly reactivated during NREM sleep, which mediates the redistribution of the memory representations and integrates them with preexisting long-term memories. The memory representations are then hypothesized to be stabilized during the following REM sleep (Diekelmann & Born, 2010; Mednick, Cai, Shuman, Anagnostaras, &
Wixted, 2011; Rasch & Born, 2013).
In contrast to this view on sleep, the primary emphasis of SHY is on the beneficial effects of sleep on memory encoding. SHY postulates that sleep has the function of renormalizing the overall synaptic strength in the brain, which has accumulated during wakefulness, to some baseline level and that this process benefits the ability of the brain to acquire new memories during subsequent wakefulness (Tononi & Cirelli, 2003, 2006).
Furthermore, the renormalization process is hypothesized to improve the signal-to-noise ratios in the brain, thereby also benefitting memory consolidation (Tononi & Cirelli, 2003, 2006).
In the following section, SHY will be looked at in more detail through a description of the main predictions that the theory makes, as well as a description of the empirical and conceptual evidence serving as the foundation of each of these predictions.
SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 21
The Synaptic Homeostasis Hypothesis (SHY)
SHY was originally formulated in 2003 by Tononi and Cirelli. The theory emphasizes a role of sleep in renormalizing the synaptic strengthening occurring during wakefulness, as a result of learning, to a baseline level (Tononi & Cirelli, 2003). Thus, wakefulness results in an overall synaptic strengthening while sleep, by renormalizing the balance between potentiation and depression, results in an overall decrease in synaptic strength (sometimes also referred to as synaptic weight in the literature) compared to the end of the waking period. Furthermore, this renormalization process has beneficial effects on memory and learning by promoting the capacity for memory encoding after the sleeping period as well as aiding memory consolidation during sleep (Tononi & Cirelli, 2003, 2006).
The formation of SHY was partly influenced by Borbély’s two-process model of sleep
(Rasch & Born, 2013; Tononi & Cirelli, 2006) according to which sleep is regulated by both a circadian process and homeostatic sleep pressure (Borbély, 2001; Borbély & Achermann,
2011). The circadian process is dependent on the daily light/dark cycle and ensures that the animal gets sleepy at the same time each day, while the homeostatic pressure rises with wakefulness and decreases during sleep and, thus, is dependent on the length of the previous waking period (Borbély, 2001; Borbély & Achermann, 2011). The two-process model predicts that SWA during NREM sleep in mammals is regulated by the homeostatic process.
In accordance with the two-process model, studies on both humans and animals has demonstrated the presence of a homeostatic process of the SWA in NREM sleep (Borbély,
2001).
SHY builds upon the two-process model by connecting the homeostatic process to processes of synaptic plasticity (Rasch & Born, 2013; Tononi & Cirelli, 2003, 2006), proposing an overall increase in synaptic potentiation during wakefulness and a renormalization of synaptic strength during the SWA in the following sleep-episode (Tononi SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 22
& Cirelli, 2003, 2006). The theory is built up of four main predictions concerning wakefulness and sleep, which depend upon each other and are strongly interconnected
(Tononi & Cirelli, 2003, 2006):
1. Wakefulness is accompanied by an overall increase in synaptic potentiation;
2. There is a relationship between the amount of synaptic potentiation and the amount
of the following SWA during NREM sleep;
3. The SWA is involved in the renormalization of synaptic strength;
4. The synaptic renormalization has beneficial effects on performance, especially
concerning the learning of new information (Tononi & Cirelli, 2003, 2006).
In the four sections following below, each of the predictions will be elaborated separately in the order they are presented above. The sections will be concerned with describing the predictions of SHY more thoroughly by providing a description of the fundamental theoretical background, including an overview of the experimental results serving as the foundation of each prediction as well as an outline of the evidence supporting each of them.
Wakefulness and Synaptic Potentiation
SHY postulates that wakefulness is accompanied by the learning of new information, which is characterized by long-lasting changes in the strength or number of synaptic connections between neurons, such as LTPs (Tononi & Cirelli, 2003, 2006). Furthermore, these plastic changes are hypothesized to occur during much of the waking state, rather than being restricted to when the animal is specifically engaged in an experimental learning paradigm. This is because, to synapses and neurons, the depolarization caused by the animal engaging in a specific learning task is not distinguished from the depolarization following spontaneous activity. Synapses and neurons are only able to respond to strong presynaptic firing accompanied by postsynaptic depolarization or firing in the presence of appropriate SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 23 levels of neuromodulators, something that should occur quite frequently during wakefulness
(Tononi & Cirelli, 2003, 2006).
Furthermore, the theory postulates that the synaptic changes occurring during wakefulness lead to LTP more often than LTD, resulting in an overall increase in the synaptic strength of connections between cortical neurons (Tononi & Cirelli, 2003, 2006) This activity causes, at least in the cerebral cortex, an imbalance between synaptic potentiation and depression and an increase in the overall synaptic strength in the brain (Tononi & Cirelli,
2003, 2006).
This claim originates from evidence demonstrating that several molecular mechanisms which are thought to be important in inducing LTP, and thus, in the process of acquiring new long-term memories, expresses higher levels during wakefulness than in sleep (Cirelli &
Tononi, 2000). Furthermore, the firing of the noradrenergic system, which involves neurons releasing the neurotransmitter noradrenaline (or norepinephrine), has been shown to be high during wakefulness, and low or absent during sleep (Aston-Jones & Bloom, 1981), and after damage to the noradrenergic system the levels of several different LTP-related molecular mechanisms decreases towards those seen in sleep even if the animal is still awake and the waking EEG is fairly unchanged (Cirelli & Tononi, 2000; Cirelli, Pompeiano, & Tononi,
1996). This evidence was interpreted to indicate that the noradrenergic system plays an important role in regulating the expression of different molecular mechanisms believed to be implicated in the formation of long-term memories through affecting the induction of LTP and thus, because the level of noradrenaline is high in waking and low in sleep, the expression of LTP-related molecular mechanisms is also high in waking and low in sleep (Tononi &
Cirelli, 2001). Subsequently, because the noradrenergic system shows greater firing during wakefulness than in sleep, and because this system has been shown to be important for regulating the expression of several molecular mechanisms that have been linked to synaptic SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 24 potentiation, SHY postulates that learning should primarily take place while we are awake and not while we are asleep.
The postulation that the waking state has a preference for synaptic potentiation over synaptic depression, was not derived from a substantial body of evidence (Tononi & Cirelli,
2003, 2006). It relies on unit recordings indicating that the cerebral cortex engages itself in sensory, motor or cognitive tasks by having some chosen groups of neurons strongly increase their firing rate, rather than decreasing it when there is low spontaneous activity (Tononi &
Cirelli, 2003). One study, in which a single whisker of a mouse was stimulated for a period of
24 h, showed an overall increase in the synaptic density on cortical neurons in the corresponding brain area (Knott, Quairiaux, Genoud, & Welker, 2002), which would be expected if synaptic potentiation occurs during spontaneous wakefulness and not only when the animal is specifically engaging in learning.
The claim was also developed through reasoning from an evolutionary perspective, from which it would seem more logical that learning (i.e. synaptic potentiation) should occur while we are awake because wakefulness exposes us to the external environment, and not while we are asleep, having a higher arousal threshold and neural activity generally disengaged from the external environment (Tononi & Cirelli, 2003, 2006).
Taken together, both the results from the studies and taking on an evolutionary point of view supports the notion that wakefulness is generally accompanied by the learning of new information through increases in synaptic strength, both when the animal is specifically engaged in a learning task and during spontaneous wakefulness.
Synaptic Potentiation and Slow-Wave Homeostasis
SHY predicts a relationship between the amount of synaptic potentiation occurring during wakefulness and the amount of SWA occurring during the subsequent period of sleep
(Tononi & Cirelli, 2003, 2006). More specifically, the hypothesis predicts that the higher the SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 25 amount of synaptic potentiation during wakefulness, the higher the amount of SWA during the following sleep-session (Tononi & Cirelli, 2003, 2006).
The hypothesis claims that it is not wakefulness per se that is responsible for operating the homeostatic increase in the amount of SWA during sleep, but rather the induction of LTP- related molecular mechanisms, which are expressed at higher levels during wakefulness, that is the cause of the increase. Furthermore, when synaptic potentiation is particularly strong in a specific brain area, SHY predicts that the subsequent period of sleep should show increased
SWA in the corresponding area, a process referred to as local intensification of sleep (Tononi
& Cirelli, 2003, 2006).
This part of the hypothesis stems primarily from the observation that SWA increases as a result of an increased duration of the previous waking period, and then decreases progressively during the course of the sleeping period (Borbély, 2001). Additional evidence that contributed to the foundation of this prediction came from studies of rats with damaged noradrenergic systems, in which the levels of molecules related to LTP in the cerebral cortex decreased during wakefulness (Cirelli, et al., 1996; Cirelli & Tononi, 2000). The results of these studies showed that even though the amount and timing of the sleeping period was unchanged, the peak in SWA during NREM sleep, which is normally seen after the active phase, and before the sleeping phase, disappeared (Tononi & Cirelli, 2003, 2006).
The foundation for the claim that stronger synaptic potentiations in a specific brain area should lead to a disproportionate increase of SWA in that area during subsequent sleep, was provided by a study by Huber, Ghilardi, Massimini and Tononi (2004). The study employed a visuomotor task known to activate specific parietal and motor areas. The authors compared the amount of SWA during sleep between the participants that had engaged in the visuomotor task and the participant that had engaged in the control task and found that the visuomotor SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 26 task elicited a substantial, localized increase in the amount of SWA during sleep in the brain areas predicted by the demands of the task (Huber et al., 2004).
Slow-Wave Homeostasis and Synaptic Downscaling
The postulation that wakefulness results in synaptic potentiation more often than in synaptic depression implicates that the state leads to an overall imbalance between the amount of synaptic potentiation and depression (Tononi & Cirelli, 2003, 2006). In order to account for this imbalance, as well as to explain the fact that SWA increases in correlation with increased wakefulness (i.e. synaptic potentiation), SHY proposes that the SWA during NREM sleep has the function of promoting a generalized depression or synaptic renormalization (Tononi &
Cirelli, 2003, 2006). During renormalization, the synaptic strength of each synapse converging onto a neuron is proportionately reduced, meaning that all synapses shed an equal proportion of their strength, thereby avoiding the loss of memory traces by still preserving differences in synaptic strength (Tononi & Cirelli, 2003, 2006).
The function of the synaptic renormalization is to return the total synaptic strength to some baseline level at which additional synaptic changes, and thus learning, is possible, thereby acting as a form of homeostatic regulator of the overall synaptic strength.
Consequently, while wakefulness is assumed to result in an overall increase in the total synaptic strength, sleep is assumed to renormalize synaptic strength, thereby returning the total synaptic strength to a baseline level, effecting a kind of synaptic homeostasis (Tononi &
Cirelli, 2003, 2006).
The specific mechanisms underlying sleep-dependent renormalization is not known, but whichever the mechanisms involved may be, SHY postulates that the function of the renormalization-process is to ensure that the synaptic input to cortical neurons is balanced
(Tononi & Cirelli, 2006). The authors do not specify which mechanisms should, or should not, be involved in driving this proposed function; they restrict their prediction to the resulting SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 27 renormalization-process, thereby leaving the underlying mechanisms up for debate (Tononi &
Cirelli, 2003, 2006).
Some support for this claim comes from computational models of synaptic plasticity, in which it has been recognized that the synaptic strength needs rescaling after learning in order to maintain a constant level of synaptic input without losing memory traces (Miller &
MacKay, 1994). Additional support underlying this prediction was provided by evidence indicating the existence of synaptic scaling mechanisms acting on synapses. In order for synaptic connections to remain responsive to the input they receive, and in order for activity to strengthen or weaken these connections, there is a need for mechanisms promoting a stability in neuronal firing rates by regulating the neuronal excitability, the synaptic strengths and synaptic stabilization (Turrigiano, 1999). Activity-dependent synaptic scaling is one such scaling mechanisms, with the function to stabilize the firing rates of cortical neurons by scaling up excitatory synaptic strengths in response to decreased firing rates and scaling down excitatory synaptic strengths in response to increase firing rate of all synapses depending on a neuron (Desai, Cudmore, Nelson, & Turrigiano, 2002; Turrigiano, 1999).
Activity-dependent synaptic scaling differs from the renormalization-process proposed by SHY in that it involves both synaptic upscaling and synaptic downscaling and because it is primarily concerned with regulating firing rates while the proposed renormalization process mainly involves synaptic downscaling and is primarily concerned with regulating synaptic strength (Tononi & Cirelli, 2003, 2006). However, similar to activity-dependent synaptic scaling, the renormalization process is postulated to affect most, or all, synapses in the brain, which distinguishes it from other processes that decreases synaptic strength such as LTD
(which only affects specific groups of synapses), and depotentiation (which affects only recently acquired synapses). However, SHY proposes that despite the fact that SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 28 renormalization differs from LTD and depotentiation, it is highly possible that this process uses many of the same molecular mechanisms.
The foundation of this prediction relies on the existence of a cellular phenomenon underlying the SWA in the EEG-recordings during NREM sleep, called slow oscillations
(Steriade, 2003). The slow oscillations are seen in just about every cortical cell and are thought to organize the SWA in the cortex during sleep. Slow oscillations occur at <1 Hz frequency (Steriade, 2003), which is a frequency perfectly suitable for inducing
LTD/depotentiation (Siegelbaum & Kandel, 2013). One crucial aspect of the induction of
LTD is changes in calcium dynamics (Siegelbaum & Kandel, 2013), something that is also likely to occur during SWS. Also, NREM sleep is accompanied by low levels of several neuromodulators (e.g. acetylcholine, noradrenaline, and serotonin) and lower levels of the protein brain-derived neurotrophic factor (BDNF), which all are important in regulating the induction of LTD (Tononi & Cirelli, 2003, 2006). Probably the most important aspect of
NREM sleep for promoting the downscaling of synaptic strength is hypothesized to be the sequence making up the slow oscillations. The slow oscillations are the result of a sequence of cellular depolarization (the up-phase) and hyperpolarization (the down-phase) (Steriade,
2003). The up-phase and the down-phase are temporally close which may, according to
Tononi & Cirelli (2003, 2006) cause synapses to interpret the activity of the pre-synaptic neuron as not being effective enough to cause post-synaptic activity, which is a fundamental part of inducing LTD in synapses (Siegelbaum & Kandel, 2013).
Moreover, as mentioned above, NREM sleep is accompanied by lower levels of the expression of LTP-related molecules. A study examining the differences in gene expression during sleep and wakefulness in rats found that there are not only lower levels of LTP-related molecules in sleep, but also molecules believed to be related to LTD showed higher levels of expression during sleep as compared to wakefulness (Cirelli, Gutierrez, & Tononi, 2004). SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 29
Taken together, Tononi & Cirelli interpreted this evidence to demonstrate that sleep, at least the molecular aspect of the state, do not just favor synaptic depotentiation/LTD but might altogether be specifically suited for inducing these kinds of synaptic plasticity (Tononi &
Cirelli, 2003, 2006).
Synaptic Downscaling and Performance
The postulation that wakefulness results in a net increase of synaptic strength, which is related to the amount of SWA during subsequent sleep, and that the SWA has the function of renormalizing the level of synaptic strength to a baseline level leads to the fourth, and last, prediction of SHY. Namely that the proposed function of SWA during sleep as causing synaptic renormalization in order to bring about synaptic homeostasis would have several beneficial effects on the performance of the organism (Tononi & Cirelli, 2003, 2006).
One beneficial effect of such a mechanism would be regarding energy conservation
(Tononi & Cirelli, 2006). The cerebral cortex is the body’s most metabolically expensive tissue, and a large portion of the energy that the cerebral cortex consumes is used for repolarization following postsynaptic potentials. The renormalization of synapses would serve to interrupt the synaptic growth caused by the waking state, thereby preventing synaptic overload and preserving energy resources (Tononi & Cirelli, 2006). Some evidence exists, indicating a decrease in cerebral blood flow in the brain during NREM sleep as compared to wakefulness and REM sleep (Braun et al., 1997).
Another beneficial effect of synaptic renormalization during sleep would be regarding space requirements (Tononi & Cirelli, 2006). Synaptic strengthening is generally accompanied by morphological changes, such as increases in the size of the axon terminals and dendritic spines, and an increasing number of synapses. Because the brain is tightly packed, space is very precious, and therefore synaptic downscaling during sleep would be important for not only controlling the metabolic costs of strengthened synapses, but also how SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 30 much space they consume by downscaling the morphological changes caused by synaptic potentiations (Tononi & Cirelli, 2006).
Finally, sleep-dependent renormalization of synaptic strength would have beneficial effects on learning and memory. If the synaptic plasticity occurring during wakefulness would go uninterrupted, the ability of the brain to acquire new information (learning) would be rapidly limited due to the energy and space costs it demands, if there was no effect counterbalancing it, such as downscaling (Tononi & Cirelli, 2006). Furthermore, synaptic renormalization would promote synaptic competition because it would preserve connections between strongly correlated neurons and eliminate others (Tononi & Cirelli, 2006). In adults, downscaling could also benefit learning and memory consolidation by improving the signal- to-noise (STN) ratios in the relevant brain circuits, due to the fact that noise-synapses, which do not contribute to the signal and which have a lower synaptic strength, would stop interfering with the activity (Tononi & Cirelli, 2006). Several studies have found evidence indicating that sleep benefits learning (Fenn, et al., 2003; Gais, Plihal; Mednick, Nakayama,
& Stickgold, 2000; Stickgold et al., 2000; Walker et al., 2002).
Summary
SHY predicts that synapses are primarily strengthened during wakefulness, leading to an overall imbalance towards synaptic potentiation, which creates higher demands on the brain in terms of energy and space consumption. Furthermore, if synapses would continue to be strengthened without restriction, the selectivity of synapses would be weakened and the ability to learn would be saturated. To account for this imbalance, and the negative consequences following from it, SHY proposes that SWS has the function of renormalizing the balance between synaptic potentiation and depression, through some mechanism(s) with the function of globally downscaling synaptic strength, to a baseline level where additional learning is possible while older memories are still preserved. The core claim of SHY, thus, is SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 31 that wakefulness results in a net increase in synaptic strength and that subsequent sleep has the function of causing a net decrease in synaptic strength in order to relieve the brain from the energetically and spatially demanding environment caused by wakefulness, as well as avoid the saturation of the ability to learn anew (Tononi & Cirelli, 2006, see figure 5). To cite the authors themselves: “Sleep is the price we pay for plasticity” (Tononi & Cirelli, 2014, p.12).
Figure 5. A schematic depiction of SHY. Adapted from Tononi & Cirelli, 2006.
Since the formulation of SHY (Tononi & Cirelli, 2003, 2006) the theory has received great amounts of attention in the scientific community. Supporting and contradictory evidence has both been published during the past decade or so, and both positive and negative opinions and voices has been raised, not only regarding the theory as a whole, but also concerning the evidence supporting it.
Following below, some of the recently published evidential support for SHY will be presented before continuing on to describe some of the contradictory findings retrieved over the years. SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 32
In Support of SHY
Following the publication of the first theoretical outline of SHY in 2003 (Tononi &
Cirelli), several sleep researchers have put a lot of time and energy into investigating the core claim of the theory – that wakefulness leads to a net increase in synaptic strength and that this strengthening is counterbalanced by sleep, which results in a net decrease in synaptic strength
(Tononi & Cirelli, 2003, 2006). This has resulted in a substantial body of evidence supporting the theory on both the molecular, electrophysiological and structural level. Below follows a description of some of most relevant and well-cited (both by the authors themselves and other sleep researchers) evidential support for SHY.
As was described previously (see section “The Synaptic Homeostasis Hypothesis”), all of the predictions made by SHY were derived from existing evidence supporting each of these claims. The prediction of SHY, postulating that wakefulness is accompanied by a net increase in synaptic strength, was formulated, in part, as a result of experiments showing higher levels of LTP-related molecular mechanisms and noradrenaline in waking than in sleep (Cirelli &
Tononi, 2000; Aston-Jones & Bloom, 1981).The predicted relationship between the amount of synaptic potentiation occurring in wakefulness and the amount of SWA in the following period of sleep, was put forward, inter alia, due to the observation of increased SWA in response to an increased preceding period of wakefulness, and the observation that the amount of SWA decreases along the course of the sleeping period (Borbély, 2001). The prediction of SHY, postulating that the SWA promotes a renormalization of the synaptic strength to a baseline level, was not based on strong experimental findings but was hypothesized, in part, as a result of computational models of synaptic plasticity indicating the need for a rescaling of synaptic strength following learning (Miller & MacKay, 1994), the finding of the activity-dependent synaptic scaling mechanism (Desai et al., 2002; Turrigiano,
1999), and collective evidence indicating that NREM sleep poses an environment perfectly SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 33 suited to induce processes such as LTD and depotentiation, which causes synaptic weakening
(Cirelli et al., 2004; Steriade, 2003). The predicted beneficial effects of the proposed renormalization process on the energy and space consumption of the brain as well as the capacity of additional learning, relied on the assumption that greater synaptic strength consumes more energy and space in the brain, which already has a limited capacity, and the claim that synaptic renormalization would improve the signal-to-noise ratios, thereby benefitting learning and memory consolidation (Tononi & Cirelli, 2003, 2006). The prediction was provided support from observations of a decrease in cerebral blood flow during NREM sleep compared to wakefulness (Braun et al., 1997), and evidence indicating beneficial effects of sleep on learning and memory consolidation (e.g., Fenn et al., 2003; Gais et al., 2000).
In addition to the experimental results serving as the foundation of SHY, recent years has provided further evidential support in favor of the theory. In 2008, Vyazovskiy, Cirelli,
Pfister-Genskow, Faraguna, and Tononi performed an experiment in rats employing a combined molecular and electrophysiological approach, in order to investigate whether wakefulness caused a net synaptic strengthening and sleep a net synaptic weakening/depression, as predicted by SHY. The target of the study was a specific type of receptor, whose delivery to specific sites of the neuron has been implicated in LTP and learning, and whose removal from these sites is associated with LTD (Vyazovskiy et al.,
2008). At the molecular level, the authors measured the protein levels of the receptors in these neuronal sites and found that the level of receptors were 30-40% higher after waking than after sleep, consistent with SHY’s prediction of a net synaptic strengthening after wakefulness and a net synaptic weakening following sleep (Vyazovskiy et al., 2008). To investigate the differences between wakefulness and sleep at the electrophysiological level, the rats’ brains were implanted with electrodes. The slope and amplitude of the local field potential (LFP) responses to electrical stimulation, a classical in vivo measurement of synaptic strength SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 34
(Tononi & Cirelli, 2014), were then recorded using polysomnography (EEG, EOG, EMG). An increase in the LFP slope is believed to be associated with LTP while a decrease has been linked to LTD (Vyazovskiy et al., 2008). The recorded responses showed an increase in the both the slope and amplitude of the LFPs after wakefulness and a decrease after sleep, in accordance with SHY (Vyazovskiy et al., 2008).
In line with this evidence, a recent study using transcranial magnetic stimulation (TMS) to stimulate cortical neurons in humans and EEG to measure the electrical response, demonstrated similar results (Huber et al., 2012). The study investigated the slope and amplitude of the response during baseline wakefulness, one night of sleep deprivation and one night of recovery sleep, and found a significant increase in the slope and amplitude of the
TMS-evoked potentials (TEPs) after sleep deprivation and a return to baseline after a night of recovery sleep. Furthermore, results showed a progressive increase in the slope and amplitude of TEPs during baseline wakefulness, with significantly larger TEPs in the afternoon as compared with the previous evening (Huber et al., 2012).
Further evidence supporting a net increase in synaptic strength after wakefulness and a net decrease after sleep was provided by a study measuring miniature excitatory postsynaptic currents (mEPSCs), the flow of ions that causes an (EPSP), and examined the changes in frequency and amplitude in brain slices from male rats and mice after spontaneous wakefulness and sleep deprivation (Liu, Faraguna, Cirelli, Tononi, & Gao, 2010). Changes in the frequency of mEPSCs are believed to be the result of modifications in the presynaptic component of synaptic transmission and changes in the amplitude are thought to be indicative of changes in the postsynaptic component (Liu et al., 2010). The results of the study showed both increased frequency and amplitude of mEPSCs in the animals who had been awake and a decrease in those that had been asleep, indicating that the changes in synaptic strength across waking and sleep occurs both pre-and postsynaptically (Liu et al., 2010). SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 35
Several studies have also been conducted to investigate the neuronal need for synaptic homeostasis proposed by SHY, which predicts that increased synaptic strength should lead to an increased amount of SWA during subsequent sleep. For example, one study from 2007 used TMS to stimulate the motor cortex in humans, in order to induce synaptic potentiation in that area and measured the responses using EEG (Huber et al., 2007). The results showed an increase in the amount of SWA during the following sleep, in the area that had been stimulated, indicative of a relationship between synaptic plasticity and SWA, consistent with the predictions of SHY (Huber, Esser, et al., 2007). Similar results were obtained from an experiment where rats had one hemisphere injected with a gene that has been linked to plasticity, which was then followed by EEG-measurements of the amount of SWA during the subsequent sleep period (Faraguna, Vyazovskiy, Nelson, Tononi, & Cirelli, 2008). The results showed greater amounts of SWA in the injected hemisphere compared to the non-injected, again indicating a link between synaptic plasticity and SWA (Faraguna et al., 2008). Similar observations have also been seen in an experiment in which the exploratory behavior of rats was manipulated, increasing the expression of plasticity-related genes (Huber, Tononi, &
Cirelli, 2007), an experiment in which rats were trained on a task known to elicit LTP in the motor cortex (Hanlon, Faraguna, Vyazovskiy, Tononi, & Cirelli, 2009), and an experiment in which human subjects were trained on a visuomotor task (Määttä et al., 2010).
In a series of three companion experiments in 2007, the prediction of SHY that the overall synaptic strength decreases during sleep were investigated in both a computer model
(Esser, Hill, & Tononi, 2007), rats (Vyazovskiy, Riedner, Cirelli, & Tononi, 2007), and humans (Riedner, Vyazovskiy, Huber, Massimini, Esser, Murphy, & Tononi, 2007). The results of these studies were consistent with the prediction of SHY, by observations of a decrease in SWA in response to downscaling synaptic strength (Esser et al., 2007), and the SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 36 observations of higher amounts of SWA during the early parts of sleep, and lower amounts of
SWA during the late parts of sleep (Riedner et al., 2007; Vyazovskiy et al., 2007).
Moreover, the prediction of SHY stating that sleep benefits the ability to learn has also received some support since the theory’s initial formulation. In a study in 2007 Yoo, Hu,
Gujar, Jolesz and Walker (2007) investigated the effects of a single night of sleep deprivation on hippocampal functioning, and found that after 35 h of sleep deprivation the ability of the human subjects to form memories was compromised (Yoo et al., 2007). Additional support for the importance of prior sleep in learning came from a study in 2009, demonstrating that a local increase in SWS plays a causal role in improving the performance on a visuomotor task
(Landsness et al., 2009).
Several experiments conducted in Drosophila melanogaster (fruit flies) has provided additional evidence in support of the core claim of SHY – that wakefulness results in a net increase in synaptic strength and that subsequent sleep progressively reduces the strength of synapses causing an overall decrease in synaptic strength. A study of Drosophila in 2009
(Gilestro, Tononi, & Cirelli, 2009) examined the effects of waking and sleep on synaptic markers in the flies by measuring changes in protein levels of several pre- and postsynaptic components. The results of the experiments demonstrated increased expression of the synaptic markers following wakefulness or sleep deprivation relative to sleep (Gilestro et al., 2009).
Further experiments in Drosophila have shown that wakefulness and sleep deprivation cause morphological synaptic changes, such as growth of dendrites, and increases the number and strength of synapses, as well as a reversal of these changes after the flies are allowed to sleep
(Bushey, Tononi, & Cirelli, 2011). Complementary findings were provided by another study in this fly species, in which the authors successfully controlled the timing and duration of the flies’ sleep by identifying a circuit for sleep regulation (Donlea, Thimgan, Suzuki, Gottschalk,
& Shaw, 2011). The study found that sleep caused a net decrease in the overall synaptic SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 37 strength in the brains of the flies, and also promoted the flies’ capacity to form of long-term memories (Donlea et al., 2011).
Critique of SHY
It seems reasonably justified to argue that SHY is supported by a considerable amount of empirical and conceptual findings. Even though a large amount of the experiments conducted since SHY was put forward have resulted in evidence supporting the theory (see section “In Support of SHY”), recent years have also resulted in a substantial amount of critique directed toward the theory. The critique brought up against SHY includes, among other aspects, criticism of the evidence cited in its support, findings contradicting the theory, criticism of the theory’s structure and arguments that alternative mechanisms could account for the supportive findings. Below follows a discussion of this criticism, in the order presented above, as well as Tononi and Cirelli’s response to some of this critique.
Criticism of Supporting Evidence
The evidence supporting SHY has received criticism from several sleep researchers who have all pointed out several different weaknesses of the material (Frank, 2012; Heller, 2014; Born
& Feld, 2012; Guiditta, 2014).
The usage of random methods. The methods used in the studies cited in support of
SHY is one aspect of the criticism which has been directed toward the evidence supporting the theory. Several different studies cited in support of SHY (Liu et al., 2010; Vyazovskiy et al., 2008; Vyazovskiy et al., 2009) have been criticized for using random methods to detect changes in net synaptic strength across wakefulness and sleep (Born & Feld, 2012; Guiditta,
2014). Furthermore, these random methods cannot reveal the origin of the changes observed, that is, they are unable to tell whether the results are caused by changes in potentiated or normal synapses and which memories are supported by the renormalized synapses (Guiditta,
2014). Because these studies examined sleep as a whole, including both NREM and REM SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 38 sleep, a role for REM sleep in contributing to the synaptic renormalization process cannot be overlooked (Born & Feld, 2012). Subsequently, Tononi and Cirelli have also been criticized for ignoring a role of REM sleep in their proposed role of sleep in synaptic plasticity
(Guiditta, 2014), even though some evidence indicates that REM sleep contributes to the renormalization process (Grosmark, Mizuseki, Pastalkova, Diba, & Buzsáki, 2012).
The significance of the supporting evidence. The significance of the evidence supporting SHY has also been discussed and criticized by several different sleep researchers.
For example, in an editorial in SLEEP, sleep researcher Craig Heller stated, concerning SHY, that “what has been lacking are experimental designs with the potential of disproving the hypothesis” (Heller, 2014, p.1157), indicating that the existing evidence in support of SHY did not have this potential.
In response to Heller, Cirelli and Tononi (2015) stated that all of the studies cited in support of SHY could have resulted in observations contradicting the theory, and thus had the potential of disproving it. If the results would have shown the opposite changes or if there has been no observable changes, the very same studies would have proved the theory wrong
(Cirelli & Tononi, 2015).
Sleep researcher Marcos Gabriel Frank, at the University of Pennsylvania, has also expressed concerns regarding the significance of the evidence supporting SHY (2012), by stating that “SHY is supported by an impressive number of findings…mostly reported by the same group” (Frank, 2012, p.1) and that:
The proponents of SHY have … amassed an impressive set of supportive findings, but
these have yet to be pursued in depth. … In the absence of a clearly proposed
mechanism …, the empirical support of SHY are hard to interpret. Therefore the
significance of SHY – and what it may one day reveal about sleep and synaptic
plasticity – remains elusive. (Frank, 2012, p.9). SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 39
In a review article published in Neural Plasticity, Tononi and Cirelli (2012) responded to Frank’s criticism, by stating that SHY is, first and foremost, a hypothesis about the universal, fundamental function of sleep and not about which mechanisms underlie that function. Different mechanisms might be involved in driving this proposed function in different brain structures and different species (Tononi & Cirelli, 2012). Whichever the mechanisms involved are, is not of main importance, Tononi and Cirelli continue. What matters for SHY to be true or false is if synaptic homeostasis is needed and if this process requires sleep (Tononi & Cirelli, 2012).
Furthermore, concerning the remark, made by Frank (2012, p. 9), that the evidential support for SHY is primarily reported by the same group of researchers, Tononi and Cirelli responded that even though there is some truth to that statement, several of the findings have been supported by independent evidence from other research groups (e.g., Donlea et al.,
2011). They agreed, however, that the main predictions of the theory need further testing both in different species, different brain structures and during different periods in development
(Tononi & Cirelli, 2012), in order to establish how well these predictions can be generalized.
Contradictory evidence
Most of the evidence contradicting the predictions of SHY has involved the observation of synaptic strengthening occurring during sleep. For example, in a study from 2012,
Chauvette, Seigneur, and Timofeev employed EOG, EMG and LFP to record electrographic signals from different cortical areas in cats. In order to study the effects of SWS on synaptic plasticity, the authors used a 1 Hz stimulation in the somatosensory cortex, and recorded the responses (Chauvette, Seigneur & Timofeev, 2012). The results showed an increased LFP response, indicative of the occurrence of LTP during SWS, contrary to the predictions of SHY
(Chauvette et al., 2012). SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 40
Additional evidence indicating synaptic strengthening during NREM sleep was provided by experiments demonstrating that NREM sleep promotes neural protein synthesis
(Czikk, Sweeley, Homan, Milley, & Richardson, 2003; Nakanishi et al., 1997; Vazquez, Hall,
Witkowska, & Greco, 2008;), which is an important part of persistent forms of synaptic potentiation, such as LTP, and that sleep deprivation reduces the level of several different proteins implicated in LTP (Basheer, Brown, Ramesh, Begum & McCarley, 2005; Guan,
Peng & Fang, 2004; Neuer-Jehle, Rhyner & Borbély, 1995).
Further support for the claim that sleep may favor synaptic potentiation, was provided by Aton, Suresh, Broussard, and Frank (Aton, Suresh, Broussard & Frank, 2014) in an experiment measuring orientation-specific response potentiation (OSRP) in mice. OSRP is a form of plasticity which enhances cortical responses to a visual stimuli of the same orientation as a previous briefly presented one, and is considered an in vivo form of LTP of glutamatergic synapses in the primary visual cortex (V1) (Aton et al., 2014). The study measured changes in
OSRP-associated responses in V1 neurons during sleep or sleep deprivation, using LFP/EEG.
The results showed that sleep deprivation impaired OSRP while sleep was important for
OSRP consolidation, which the authors considered an indication of sleep promoting synaptic potentiation (Aton et al., 2014).
Similar results have been seen in studies of ocular dominance plasticity (ODP), which refers to changes in visual cortical circuits following monocular deprivation (MD) or other changes in patterned vision, favoring the eye still receiving visual information (Frank, 2014).
Experiments inducing MD in cats has shown that sleep not only consolidates ODP (Frank,
Issa & Stryker, 2001), but furthermore, that it does so by strengthening the cortical responses to stimuli presented to the non-deprived eye, indicating the occurrence of synaptic strengthening during sleep (Aton et al., 2009). SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 41
In a study from 2012, Grosmark, Mizuseki, Pastalkova, Diba, and Buzsáki examined the firing rates of hippocampal neurons during NREM and REM sleep in rats, in order to test
SHY’s claim that sleep, more specifically NREM sleep, reduces firing rates and neuronal excitability. The results showed a modest increase in firing rates during NREM sleep, which was then overcome by larger decreases during the subsequent REM episode, resulting in an overall decrease in firing rates during sleep (Grosmark et al., 2012). Although SHY predicts an overall decrease in firing rates during sleep, the theory attributes these changes to NREM sleep, which contradicts the results of this study.
The Vagueness of SHY
Besides the criticism directed toward the evidence supporting SHY and the observations of contradictory results, the structure of the theory has also been criticized for being too vague. In his review of SHY in 2012, Frank criticized the theory for equating learning with LTP induction by stating that this is an oversimplification of learning and that learning can take place also through other molecular mechanisms (e.g., LTD) (Frank, 2012).
He then went on to criticize SHY for oversimplifying the effects of plasticity-related molecules on synaptic strength, by pointing out that, most often, these molecules have multiple effects on synaptic strength, mediating both LTP and LTD (Frank, 2012).
Furthermore, Frank also argues that the downscaling (i.e., renormalization) function of
NREM sleep, proposed by SHY, does not reflect what is currently known about synaptic scaling mechanisms (Frank, 2012). The principle of synaptic scaling, Frank states, is that decreases in neuronal activity upscale synapses and increases downscale synapses. Thus, following this principle, the down-phases occurring during SWA should upscale synapses, which contradicts the function of the renormalization-process proposed by SHY (Frank,
2012). SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 42
Misunderstandings of the theory. In response to this criticism, Tononi and Cirelli argued (2012) that Frank had misunderstood SHY’s predicted relationship between wakefulness, learning, and synaptic strengthening as well as the proposed renormalization process. SHY does not, the authors argue, claim that the plasticity occurring during wakefulness is exclusively caused by LTP, but rather that “wake is associated with a net increase in synaptic strength, irrespective of the particular mechanisms involved” (Tononi &
Cirelli, 2012, p. 3). Regarding the misrepresentation of the available synaptic scaling mechanisms, Tononi and Cirelli (2012), responded that the proposed renormalization process differs from other synaptic scaling mechanisms because it is postulated to primarily regulate synaptic strength, rather than neuronal firing rates. Furthermore, they state that they did not equate the renormalization process with activity-dependent synaptic scaling, but rather proposed that processes of synaptic scaling might be involved in synaptic homeostasis during sleep (Tononi & Cirelli, 2012). It is also possible that activity-dependent LTD is involved in the proposed function, provided that the depression is generalized to most, or all, of the synaptic connections. Finally, Tononi and Cirelli (2012) stated that the specific mechanisms contributing to synaptic homeostasis may vary, as long as the end result is a net synaptic depression.
The following year, Frank (2013) published an article in Neural Plasticity, in which he responded to Tononi and Cirelli by arguing against their defense of SHY. He claimed that, in their formulation of SHY, Tononi and Cirelli were very liberal with their usage of words such as “LTP-like”, making it difficult to interpret the theory in any other way than it emphasizing the induction of LTP during wakefulness and learning (Frank, 2013). Continuing, he pointed out that in their response Tononi and Cirelli stated that the renormalization process proposed by SHY is focused on synaptic strength, unlike activity-dependent synaptic scaling which focuses on firing rates, yet they cite changes in firing rates and mEPSCs (e.g., Liu et al., 2010; SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 43
Vyazovskiy et al., 2009) as support for SHY (Frank, 2013). Furthermore, Frank explains that he did not imply that SHY equated the proposed renormalization process with activity- dependent synaptic scaling, but rather, he argued that the renormalization process that the theory propose should be examined based on what is currently known about synaptic scaling mechanisms (Frank, 2013).
Finally, Frank (2013) accuses Tononi and Cirelli of ignoring evidence contradicting
SHY, such as observations of increased synaptic strength during sleep (see for example: Aton et al., 2014) as well as evidence indicating that SWA increases synaptic strength (Chauvette et al., 2012).
Alternative Mechanisms Accounting for Synaptic Renormalization during Sleep
Another point of criticism directed toward SHY is regarding the theory’s lack of specificity concerning the mechanisms underlying the global synaptic renormalization process during SWS (Frank, 2012). The absence of a clearly defined mechanism for this process means that, the critical argument goes, any evidence of synaptic weakening after sleeping can be interpreted as support for the theory, but it also means that the supporting evidence possibly can be accounted for by other alternative processes/mechanisms that are unrelated to sleep (Frank, 2012).
Brain temperature. Changes in brain temperature is one alternative mechanism which has been proposed to possibly be able to account for the evidential support of SHY, especially the evidence retrieved from studies in Drosophila (Frank, 2012, 2014). Ectothermic insects, such as Drosophila, do not possess the ability to internally regulate their core/brain temperature, but rather regulate their temperature behaviorally by selecting warmer environments or engaging in activity (Stevenson, 1985). Warmer environments cause several changes in Drosophila neurons, such as axonal branching (Peng et al., 2007; Zhong & Wu,
2004), and the outgrowth of dendrites and axons (Peng et al., 2007). SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 44
In mammals brain temperature peaks in the active phase and decreases during sleep
(Glotzbach & Heller, 2000). The effects of a higher brain temperature on synapses will be greater during wakefulness and then decline during the sleeping period. Some of the effects of a lowered brain temperature, as seen in sleep, that have been observed in mammals includes reductions in dendritic spines (Roelandse & Matus, 2004), reductions of EPSPs and reversal of LTP (Bittar & Muller, 1993), and reductions in synaptic strength (measured by mEPSCs;
Simkus & Stricker, 2002).
Thus, the supporting evidence which indicates a net synaptic potentiation after prolonged wakefulness or sleep deprivation may be accounted for by changes in brain temperature of the animal, rather than by the waking state as such (Frank, 2012, 2014).
Glucocorticoid concentrations. Changes in glucocorticoid concentrations in the brain are another proposed alternative explanation for the evidential support of SHY (Frank, 2012,
2014). Just like brain temperature, the glucocorticoid concentrations in the brain rises with wakefulness and declines in sleep (Van Cauter, 2005). Changes in glucocorticoid concentrations have been shown to affect synaptic efficacy in several different ways. For example, the stress hormone corticosterone increases the frequency (Olijslagers et al., 2008) and amplitude of mEPSCs (Karst & Joëls, 2005) in the mouse hippocampus. Also, the activation of glucocorticoid receptors increases the synaptic strength of glutamatergic synapses at dopamine neurons in the ventral tegmental area (Daftary, Panksepp, Dong, &
Saal, 2009).
The effects of increased or decreased corticosterone release, occurring in wake and sleep respectively, has been proposed to explain several different finding cited in support of
SHY (Frank, 2012), including observations of differences in EFPs (Vyazovskiy et al., 2008) and mEPSCs (Liu et al., 2010) due to these studies examining the animals during circadian times when corticosterone release is high (wakefulness) or low (sleep) as well as studies SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 45 demonstrating differences in plasticity-related molecules and proteins between wakefulness and sleep (Cirelli & Tononi, 2000; Vyazovskiy et al., 2008).
Controlling temperature. Responding to this critique, Tononi and Cirelli (2012) stated that in the Drosophila studies, the insects were kept in chambers with a constant temperature of 20C. Furthermore, the authors continue, the presynaptic structural changes observed are unlikely to be caused by the flies engaging in motor activities because the effects were found in flies who had been kept in small glass tubes which did not allow for much movement
(Tononi & Cirelli, 2012). The evidential support for plasticity-related molecular changes between wakefulness and sleep (Vyazovskiy et al., 2008), were obtained from rats with a very low temperature increase (0.3-0.4C), and the EFP responses recorded after sleep and wakefulness in the this study, were all collected during quiet wakefulness, thus eliminating the possibility of the circadian time influencing the results (Tononi & Cirelli, 2012).
Inconclusive evidence. Concerning the effects of glucocorticoids on synaptic plasticity,
Tononi and Cirelli (2012) argued that although there is evidence of glucocorticoids enhancing synaptic efficacy, there is also evidence of an opposite effect. For example, stress has been shown to reduce synaptic efficacy in the medial prefrontal cortex and dorsal hippocampus
(Caudal, Godsil, Mailliet, Bergerot, & Jay, 2010), corticosterone has been shown to, under certain conditions, facilitate LTD induction (Martin et al., 2009), and chronic glucocorticoid excess has been shown to cause long-lasting net spine elimination in cortex (Liston & Gan,
2011). The effects of glucocorticoids on processes of synaptic plasticity, is thus, according to
Tononi and Cirelli (2012), inconclusive and therefore it is difficult to evaluate whether changes in glucocorticoid levels could account for the findings cited in support of SHY.
SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 46
Discussion
The aim of this review has been to evaluate where SHY stands in the light of evidence and criticism published in relation to the predictions of the theory since it was originally formulated. In order to perform this evaluation, the thesis has considered some of the most cited evidence published in support of, and in opposition of the predictions of SHY, as well as some of the strongest points of criticism directed toward the theory.
Recent years has been accompanied by a large amount of evidence published in support of SHY. We have seen evidence at both the molecular, electrophysiological and structural level, supporting SHYs notion that wakefulness results in a net increase in synaptic strength, followed by a net decrease in synaptic strength after subsequent sleep. Recent years have also seen empirical evidence contrasting the predictions of SHY, primarily by observations of synaptic strengthening during sleep. In relation to these contradicting evidence, Tononi and
Cirelli (2014) have argued that SHY does not exclude the possibility that synaptic potentiation may also occur during sleep. Furthermore, the authors expresses that SHY is a theory about the essential function of sleep, that is, renormalizing the synaptic strength in the brain in order to maintain the capacity for learning, and not a theory about which specific mechanisms is involved in driving the proposed function (Tononi & Cirelli, 2012). Subsequently, it seems that, according to Tononi and Cirelli, SHY cannot be refuted by observations of the occurrence of synaptic potentiation during sleep, but could be refuted by evidence demonstrating a net increase in synaptic strength after sleep or evidence indicating no changes in synaptic strength after sleep.
At a first glance, there seem to be strong evidential support for SHY and, moreover, the contradicting evidence do not appear inconsistent with the predictions that the theory presents.
However, the theory’s avoidance of refutation from contradictory findings seem to be the result of the relatively vague formulation of the theory as well as its’ exclusion of clearly SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 47 proposed mechanisms responsible for the postulated changes in synaptic strength across wakefulness and sleep.
The strength of SHY seems to lie in its simplicity. The theory is relatively easy to comprehend and the predictions are built up in a manner such that they appear to logically follow from each other. However, this is also the primary weakness of SHY, as its simplicity is largely the result of it being relatively unspecific regarding the details of how the predicted changes in synaptic strength are carried out in the brain. By keeping the theory at a more vague and unspecific level, Tononi and Cirelli achieves their goal of the theory being applicable to all species engaging in sleep, and also avoid any impact of recently published empirical evidence contradicting the predictions. But this also limits the impact of the evidence collected in support of the theory, the level at which research questions can be made and said to answer, and does not exclude the possibility that other changes in the brain may be responsible for regulating synaptic strength. The absence of clearly proposed mechanisms driving the renormalization process renders it difficult to determine whether the changes in synaptic strength after sleep is caused by the state itself or some other process(es) that happens to coincide with sleep, such as changes in brain temperature or glucocorticoid levels, as proposed by Frank (see section “Alternative Mechanisms Accounting for Synaptic
Renormalization during Sleep”). In extension, this means that the impact of the supporting evidence is limited as the possibility that other processes might account for these findings cannot be excluded.
In their most recent description of SHY, Tononi and Cirelli (2014) have tried to accommodate some of the criticism directed towards the theory by excluding the use of LTP- related word when discussing the postulated increase in synaptic strength during wakefulness.
This modification was probably carried out in order to clarify that SHY do not predict that the synaptic strengthening is exclusively caused by LTP. Furthermore, in the initial formulation SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 48 of SHY, the discussion about the renormalization process was largely focused on a proportional downscaling in which the strength of most, or all, synapses were downscaled by the same factor (Tononi & Cirelli, 2003, 2006). In contrast to this, the most recent description of the theory considers several different ways in which the renormalization could be achieved by discussing different computer models, including not only a proportional downscaling of synaptic strength, but also a process where stronger synapses may depress less than weaker ones as well as a process in which the strongest synapses may be protected from being depressed (Tononi & Cirelli, 2014).
An additional change in the structure of SHY, since it was originally formulated, is in terms of the discussions about the effects of the postulated function on of sleep on memory.
This discussion includes, for example, how the postulated renormalization of synaptic strength during sleep would benefit memory acquisition, memory consolidation and the integration of new memories with preexisting ones (Tononi & Cirelli, 2014). Compared to the initial formulation of SHY (Tononi & Cirelli, 2003, 2006), this discussion includes more aspects of memory and discusses each of these aspects in more detail.
The most recent description of SHY also includes a discussion of how the proposed renormalization process during sleep could be applied to different developmental periods
(Tononi & Cirelli, 2014). Even though SHY was originally put forward with an emphasis on predictions that would be testable in adult mammals, the theory needs to be able to explain how the postulated renormalization process would be applicable to all stages of development, and why it would be needed during these periods, in order for SHY to be regarded as a theory concerning the universal function of sleep. In a review from 2012, Tononi and Cirelli argues that the synaptic homeostasis function proposed by SHY would be important during developmental times characterized by a net synaptogenesis, in which many new synapses are formed. This is because, according to the authors, the massive increase in synapses during this SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 49 period would result in even greater demands on the brain than those caused by wakefulness in adult mammals, and that neurons would need a restoration of synaptic homeostasis in order to properly function (Tononi & Cirelli, 2012). Moreover, Tononi and Cirelli have recently considered that the synaptic homeostasis function of sleep may not just be needed for regulating the strength of synapses, but may perhaps also involve regulating the number and distribution of synapses, and might also involve regulations in the positive direction (Tononi
& Cirelli, 2014). However, the discussions of how, and why, synaptic homeostasis would be needed during development has not yet been properly integrated into the theory, but remains at the level of brief speculations.
Despite the fact that SHY has been somewhat restructured since it was originally formulated in 2003, any mechanisms driving the postulated renormalization process has still not been proposed, rendering the criticism concerning the lack of mechanisms to still prevail.
The only aspect of SHY which can be considered to be a proposed mechanism driving the postulated function of sleep, is that SHY still predicts that SWA in mammals and birds both offers an indication of changes in synaptic strength and contributes to the renormalization process (Tononi & Cirelli, 2003, 2006, 2012, 2014). However, the loyalty to the role of SWA in mammals and birds is perplexing considering the observations of increased firing rates during NREM sleep made by Grosmark, Mizuseki, Pastalkova, Diba, and Buzsáki (Grosmark et al., 2012). Furthermore, Tononi and Cirelli even cites this observation in support of SHY in their most recent review of SHY (Tononi & Cirelli, 2014, pp. 20-21), making the issue even more confusing. The authors do recognize, however, that even though some of the supporting evidence indicate a specific role for NREM sleep in the renormalization process, many of these observations were not restricted to NREM sleep, but were observed after total sleep, including REM sleep (Tononi & Cirelli, 2014). Moreover, Tononi and Cirelli also points out that, at the present time, there is no distinction between different sleep stages in some of the SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 50 animal species in which synaptic homeostasis during sleep has been observed, including
Drosophila Melanogaster (Tononi & Cirelli, 2014). However, this statement makes it confusing that Tononi and Cirelli defends their exclusion of mechanisms by stating that their intention is for the theory to be applicable to all species engaging in sleep. The authors are still using SWA in NREM sleep as some form of proxy for synaptic strength even though they, themselves, note that some species, do not show such activity during sleep. This is self- contradictory and begs the question if the argument of a universal theory is just some way of trying to avoid contradictory findings or perhaps avoid refutation through predicting incorrect mechanisms. Whether this is a way for the authors to avoid the findings contradicting the predictions of SHY or if it reflects their view of SHY as a theory about the universal function of sleep rather than a theory about the mechanisms driving the postulated function, is not clear. However, whatever the reason for this discrepancy, it needs to be addressed and explained by the authors.
Conclusion
Even though SHY seem to be supported by a large amount of empirical observations, the criticism directed towards the theory remains deeply problematic for the significance of these observations and for the explanatory power of SHY.
In conclusion, SHY does not hold up well against the criticism raised against the structure of the theory. However, this does not mean that the synaptic homeostasis function of sleep, postulated by SHY, can be said to be incorrect, but rather that the proposed function cannot be said to be either correct or incorrect before further specifications and, subsequently, further investigations have been made which can rule out other possible explanations, unrelated to sleep, accounting for the observed changes in synaptic strength after sleep. SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 51
Future directions
The criticism raised against SHY still needs to be addressed in future experimental studies examining the predictions of the theory. Future studies should aim at distinguishing between the two different sleep states, in order to further investigate the role of NREM and
REM sleep, respectively, in the proposed renormalization process. Additional experiments evaluating the possibility of alternative mechanisms, such as brain temperature and glucocorticoid-levels (see section “Alternative mechanisms”), which might also account for the evidence supporting SHY, should be performed. Also, as noted by Frank (2014), more direct examinations of the adaptive changes caused by the proposed renormalization process is needed, as well as direct experiments of the relationship between synaptic plasticity and sleep function, proposed by SHY. Another aspect which will need further investigation is how the proposed synaptic homeostasis function of sleep would be carried out during different developmental stages and in different species that has not yet been studied, including those who do not show SWA during sleep. If the aim of the theory is to encompass the universal function of sleep, it needs to include predictions of how the function is carried out across the developmental periods as well as in species lacking SWA.
By predicting mechanisms driving the proposed synaptic strengthening during wakefulness and the proposed synaptic renormalization during sleep, possible experimental observations consistent with these predictions would carry a much heavier weight. Evidential support is not just a matter of quantity but more importantly the evidence needs high quality.
The theory seem to have a good foundation and if the criticism is properly addressed by including more specific predictions regarding the proposed regulations of synaptic strength across wake and sleep, as well as predictions concerning the different developmental stages, and further experimental investigations are made, excluding the possible role of other processes than sleep in driving the proposed function of synaptic homeostasis, SHY has the SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 52 potential of being a very strong scientific theory about the function of sleep. Furthermore, if these critical points are properly addressed and future investigations results in support for the theory, SHY has the potential of achieving its goal of explaining the essential function of sleep across species, but for now the theory is limited in its explanatory capacity.
The question of what the essential function of sleep is, still remains to be resolved. With the fast speed of the development of new experimental techniques, the past couple of decades has resulted in great scientific advances and the next couple of decades of sleep research will indubitably get us closer to finding out the answer. Whether the predictions of SHY will turn out to be correct, and whether there actually is a universal function of sleep, remains to be seen. The one thing that is clear, is that the field of sleep research certainly has a very interesting and exciting future ahead.
SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 53
References
Abel, T., Havekes, R., Saletin, J. M., & Walker, M. P. (2013). Sleep, plasticity and memory
from molecules to whole-brain networks. Current Biology 23(17), R774–R778.
doi:10.1016/j.cub.2013.07.025
Aserinsky, E., & Kleitman, N. (1953). Regularly occurring periods of eye motility, and
concomitant phenomena, during sleep. Science 118(3062), 273–274. Retrieved from:
http://www.jstor.org/stable/1680525 .
Aston-Jones, G., & Bloom, F. E. (1981). Activity of norepinephrine-containing locus
coeruleus neurons in behaving rats anticipates fluctuations in the sleep-waking cycle.
The Journal of Neuroscience 8(1): 876–886.
Aton, S. J., Seibt, J., Dumoulin, M., Jha, S. K., Steinmetz, N., Coleman, T., Naidoo, N., &
Frank, M. G. (2009). Mechanisms of sleep-dependent consolidation of cortical
plasticity. Neuron 61, 454–466. doi:10.1016/j.neuron.2009.01.007
Aton, S. J., Suresh, A., Broussard, C., & Frank, M. G. (2014). Sleep promotes cortical
response potentiation following visual experience. SLEEP 37(7), 1163–1170.
doi:10.5665/sleep.3830
Basheer, R., Brown, R., Ramesh, V., Begum, S., & McCarley, R. W. (2005). Sleep
deprivation-induced protein changes in basal forebrain: Implications for synaptic
plasticity. Journal of Neuroscience Research 82, 650–658. doi:10.1002/jnr.20675
Berger, R. J., & Phillips, N. H. (1995). Energy conservation and sleep. Behavioral Brain
Research 69, 65–73. doi:10.1016/0166-4328(95)00002-B
Bittar, P., & Muller, D. (1993). Time-dependent reversal of long-term potentiation by brief
cooling shocks in rat hippocampal slices. Brain Research 620(2), 181–188.
doi:10.1016/0006-8993(93)90154-F SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 54
Bliss, T. V. P., & Lømo, T. Long-lasting potentiation of synaptic transmission in the dentate
area of the anaesthetized rabbit following stimulation of the perforant path. The Journal
of Physiology 232, 331–356.
Borbély, A. A. (2001). From slow waves to sleep homeostasis: New perspectives. Archives
Italiennes de Biologie 139, 53–61.
Borbély, A. A., & Achermann, P. (2011). Sleep homeostasis and models of sleep regulation.
In Kryger, M. H., Roth, T., & Dement, W. C (Eds.), Principles and practice of sleep
medicine, (5th edition), (pp. 431–444). Philadelphia, PA: Elsevier
Born, J., & Feld, G. B. (2012). Sleep to upscale, sleep to downscale: Balancing homeostasis
and plasticity. Neuron 75, 933–935. doi:10.1016/j.neuron.2012.09.007
Braun, A. R., Balkin, T. J., Wesensten, N. J., Carson, R. E., Varga, M., … & Herscovitch, P.
(1997). Regional cerebral blood flow throughout the sleep-wake cycle. Brain 120,
1173–1197.
Bushey, D., Tononi, G., & Cirelli, C. (2011). Sleep and synaptic homeostasis: Structural
evidence in Drosophila. Science 332, 1576–1581. doi:10.1126/science.1202839
Carskadon, M. A., & Dement, W.C. (2011). Normal human sleep: An overview. In Kryger,
M. H., Roth, T., & Dement, W. C (Eds.), Principles and practice of sleep medicine, (5th
ed.), (pp. 16–25). Philadelphia, PA: Elsevier
Caudal, D., Godsil, B. P., Mailliet, F., Bergerot, D., & Jay, T. M. (2010). Acute stress induces
contrasting changes in AMPA receptor subunit phorphorylation within the prefrontal
cortex, amygdala and hippocampus. PLoS ONE 5(12), e15282.
doi:10.1371/journal.pone.0015282. Retrieved from:
http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0015282 SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 55
Cauter, E. V., Spiegel, K., Tasali, E., & Leproult, R. (2008). Metabolic consequences of sleep
and sleep loss. Sleep medicine 9(1), 23–28. doi:10.1016/S1389-9457(08)70013-3
Chauvette, A., Seigneur, J., & Timofeev, I. (2012). Sleep oscillations in the thalamocortical
system induce long-term neuronal plasticity. Neuron 75, 1105–1113.
doi:10.1016/j.neuron.2012.08.034
Cirelli, C., Gutierrez, C. M., & Tononi, G. (2004). Extensive and divergent effects of sleep
and wakefulness on brain gene expression. Neuron 41, 35–43. doi:10.1016/S0896-
6273(03)00814-6
Cirelli, C., Pompeiano, M., & Tononi, G. (1996). Neuronal gene expression in the waking
state: A role for the locus coeruleus. Science 274, 1211–1215.
doi:10.1126/science.274.5290.1211
Cirelli, C., & Tononi, G. (2000). Differential expression of plasticity-related genes in waking
and sleep and their regulation by the noradrenergic system. The Journal of
Neuroscience 20 (24), 9187–9194.
Cirelli, C., & Tononi, G. (2008). Is sleep essential?. PLoS Biology 6(8), 1605–1611.
doi:10.1371/journal.pbio.0060216
Cirelli, C., & Tononi, G. (2015). Sleep and synaptic homeostasis. SLEEP 38(1), 161–162.
doi:10.5665/sleep.4348
Cohen, D. A., Pascual-Leone, A., Press, D. Z., & Robertson, E. M. (2005). Off-line learning
of motor skill memory: A double dissociation of goal and movement. PNAS 102(50),
18237–18241. doi:10.1073/pnas.0506072102
Czikk, M. J., Sweeley, J. C., Homan, J. H., Milley, J. R., & Richardson, B. S. (2003).
Cerebral leucine uptake and protein synthesis in the near-term ovine fetus: Relation to SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 56
fetal behavioral state. American Journal of Physiology-Regulatory, Integrative and
Comparative Physiology 284, R200–R2007. doi:10.1152/ajpregu.00190.2002
Daftary, S. S., Panksepp, J., Dong, Y., & Saal, D. B. (2009). Stress-induced, glucocorticoid
dependent strengthening of glutamatergic synaptic transmission in midbrain dopamine
neurons. Neuroscience Letters 452, 273–276. doi:10.1016/j.neulet.2009.01.070
Dement, W. C. (1998). The study of humans sleep: A historical perspective. Thorax 53, S2–
S7. Retrieved from: http://thorax.bmj.com/content/53/suppl_3/S2.full.pdf+html
Desai, N. S., Cudmore, R.H., Nelson, S.B., & Turrigiano, G.G. (2002). Critical periods for
experience-dependent synaptic scaling in visual cortex. Nature Neuroscience 5(8), 783–
789. doi:10.1038/nn878
Diekelmann, S., & Born, J. (2010). The memory function of sleep. Nature Reviews 11, 114–
126. doi:10.1038/nrn2762
Donlea, J. M., Thimgan, M. S., Suzuki, Y., Gottschalk, L., & Shaw, P. J. (2011). Inducing
sleep by remote control facilitates memory consolidation in Drosophila. Science 332,
1571–1576. doi:10.1126/science.1202249
Ellenbogen, J. M., Hulbert, J. C., Stickgold, R., Dinges, D. F., & Thompson-Schill, S. L.
(2006). Interfering with theories of sleep and memory: Sleep, declarative memory and
associative interference. Current Biology 16, 1290–1294.
doi:10.1016/j.cub.2006.05.024
Ellenbogen, J. M., Hulbert, J. C., Jiang, Y., & Stickgold, R. (2009). The sleeping brain’s
influence on verbal memory: Boosting resistance to interference. PLoS One 4(1), 1–4.
doi:10.1371/journal.pone.0004117 SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 57
Esser, S. K., Hill, S. L., & Tononi, G. (2007). Sleep homeostasis and cortical synchronization:
I. Modeling the effects of synaptic strength on sleep slow waves. SLEEP 30(12), 1617–
1630.
Faraguna, U., Vyazovskiy, V. V., Nelson, A. B., Tononi, G., & Cirelli, C. (2008). A causal
role for brain-derived neurotrophic factor in the homeostatic regulation of sleep. Journal
of Neuroscience 28(15), 4088–4095. doi:10.1523/JNEUROSCI.5510-07.2008
Fenn, K. M., Nusbaum, H. C., & Margoliash, D. (2003). Consolidation during sleep of
perceptual learning of spoken language. Nature 425, 614–616. doi:10.1038/nature01951
Fiorino, A. S. (1996). Sleep, genes and death: Fatal familial insomnia. Brain Research
Reviews 22, 258–264.
Frank, M. G. (2012). Erasing synapses in sleep: Is it time to be SHY? Neural Plasticity 2012,
article ID: 264378, 15 pages. doi:10.1155/2012/264378
Frank, M. G. (2013) Why I am not SHY: A reply to Tononi and Cirelli. Neural Plasticity
2013, article ID: 394946, 3 pages. doi:10.1155/2013/394946.
Frank, M. G. (2014). Sleep and synaptic plasticity in the developing and adult brain. Current
Topics in Behavioral Neuroscience. doi:10.1007/7854_2014_305
Frank, M. G., & Cantera, R. (2014). Sleep, clocks and synaptic plasticity. Trends in
Neurosciences 37(9), 491–501. doi:10.1016/j.tins.2014.06.005
Frank, M. G., Issa, N. P., & Stryker, M. P. (2001). Sleep enhances plasticity in the developing
visual cortex. Neuron 30, 275–287.
Gais, S., Lucas, B., & Born, J. (2006). Sleep after learning aids memory recall. Learning &
Memory 13, 259–262. doi:10.1101/lm.132106 SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 58
Gais, S., & Born, J. (2004). Declarative memory consolidation: Mechanisms acting during
human sleep. Learning & Memory 11, 679–685. doi:10.1101/lm.80504
Gais, S., Plihal, W., Wagner, U., & Born, J. (2000). Early sleep triggers memory for early
visual discrimination skills. Nature Neuroscience 3(12), 1335–1339.
Genzel, L., Kroes, M. C. W., Dresler, M., & Battaglia, F. P. (2014). Light sleep versus slow
wave sleep in memory consolidation: A question of global versus local processes.
Trends in Neurosciences 37(1), 10–19. doi:10.1016/j.tins.2013.10.002
Gilestro, G. F., Tononi, G., & Cirelli, C. (2009). Widespread changes in synaptic markers as a
function of sleep and wakefulness in Drosophila. Science 324(5923), 109–112.
Retrieved from: http://www.jstor.org/stable/20493647
Glotzbach, S. F., & Heller, H. C. (2000). Temperature regulation. In: Roth, C., Kryger, M., &
Dement, W. C. (Eds.), Principles and practice of sleep medicine (3rd Ed.), (pp. 289–
304). Philadelphia, PA: Saunders.
Grosmark, A. D., Mizuseki, K., Pastalkova, E., Diba, K., Buzsáki, G. (2012). REM sleep
reorganizes hippocampal excitability. Neuron 75, 1001–1007.
doi:10.1016/j.neuron.2012.08.015
Guan, Z., Peng, X., & Fang, J. (2004). Sleep deprivation impairs spatial memory and
decreases extracellular signal-regulated kinase phosphorylation in the hippocampus.
Brain Research 1018, 38-47. doi:10.1016/j.brainres.2004.05.032
Guiditta, A. (2014). Sleep memory processing: The sequential hypothesis. Frontiers in
Systems Neuroscience 8, 1–8. doi:10.3389/fnsys.2014.00219 SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 59
Hanlon, E. C., Faraguna, U., Vyazovskiy, V. V., Tononi, G., & Cirelli, C. (2009). Effects of
skilled training on sleep slow wave activity and cortical gene expression in the rat.
SLEEP 32(6), 719–729.
Heller, C. (2014). The ups and downs of synapses during sleep and learning. SLEEP 37(7),
1157–1158. doi:10.5665/sleep.3824
Hinard, V., Mikhail, C., Pradervand, S., Curie, T., Houtkooper, R. H., Auwerx, J., … & Tafti,
M. (2012). Key electrophysiological, molecular, and metabolic signatures of sleep and
wakefulness revealed in primary cortical cultures. The Journal of Neuroscience 32(36),
12506–12517. doi:10.1523/JNEUROSCI.2306-12.2012
Hobson, J. A. (2005). Sleep is of the brain, by the brain and for the brain. Nature 437(27),
1254–1256. doi:10.1038/nature04283
Huber, R., Esser, S. K., Ferrarelli, F., Massimini, M., Peterson, M. J., & Tononi, G. (2007).
TMS-induced cortical potentiation during wakefulness locally increases slow wave
activity during sleep. PLoS ONE 3, e276. doi:10.1371/journal.pone.0000276. Retrieved
from: plosone.org
Huber, R., Ghilardi, M. F., Massimini, M., & Tononi, G. (2004). Local sleep and learning.
Nature 430, 78–81. doi:10.1038/nature02663
Huber, R., Mäki, H., Rosanova, M., Casarotto, S., Canali, P., Casali, A. G., … & Massimini,
M. (2012). Human cortical excitability increases with time awake. Cerebral Cortex
2013(23), 332–338. doi:10.1093/cercor/bhs014
Huber, R., Tononi, G., & Cirelli, C. (2007). Exploratory behavior, cortical BDNF expression,
and sleep homeostasis. SLEEP 30(2), 129–139. SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 60
Iber, C., Ancoli-Israel, S., Chesson, A., & Quan, S. F. (2007). The AASM manual for the
scoring of sleep and associated events: Rules, terminology and technical specifications.
Westchester, IL: American Academy of Sleep Medicine.
Jenkins, J. G., & Dallenbach, K. M. (1924). Obliviscence during sleep and waking. The
American Journal of Psychology 35(4), 605-612.
Karst, H., & Joëls, M. (2005). Corticosterone slowly enhances miniature excitatory
postsynaptic current amplitude in mice CA1 hippocampal cells. Journal of
Neurophysiology 94(5), 3479–3486. doi:10.1152/jn.00143.2005
Keenan, S., & Hirshkowitz, M. (2011). Monitoring and staging human sleep. In Kryger, M.
H., Roth, T., & Dement, W. C (Eds.), Principles and practice of sleep medicine (5th
ed.), (pp.1602–1607). Philadelphia, PA: Elsevier
Knott, G. W., Quairiaux, C., Genoud, C., & Welker, E. (2002). Formation of dendritic spines
with GABAergic synapses induced by whisker stimulation in adult mice. Neuron 34,
265–273. doi:10.1016/S0896-6273(02)00663-3
Landsness, E. C., Crupi, D., Hulse, B. K., Peterson, M. J., Huber, R.,…& Tononi, G. (2009).
Sleep-dependent improvement in visuomotor learning: A causal role for slow waves.
SLEEP 32(10), 1273–1284.
Liston, C., & Gan, W-B. (2011). Glucocorticoids are critical regulators of dendritic spine
development and plasticity in vivo. Proceedings of the National Academy of Science of
the United States of America 108(38), 16074–16079. doi:10.1073/pnas.1110444108/-
/DCSupplemental. Retrieved from: http://www.pnas.org/content/108/38/16074 SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 61
Liu, Z-W., Faraguna, U., Cirelli, C., Tononi, G., & Gao, X-B. (2010). Direct evidence for
wake-related increases and sleep-related decreases in synaptic strength in rodent cortex.
The Journal of Neuroscience 30(25), 8671–8675. doi:10.1523/jneurosci.1409-10.2010
Maret, S., Faraguna, U., Nelson, A. B., Cirelli, C., & Tononi, G. (2011). Sleep and waking
modulate spine turnover in adolescent mouse cortex. Nature Neuroscience 14(11),
1418–1420. doi:10.1038/nn.2934
Martin, S., Henley, J. M., Holman, D., Zhou, M., Wiegert, O., van Spronsen, M.,…&
Krugers, H. J. (2009). Corticosterone alters AMPAR mobility and facilitates
bidirectional synaptic plasticity. PLoS ONE 4(3), e4714.
doi:10.1371/journal.pone.0004714.
Mednick, S. C., Cai, D. J., Shuman, T., Anagnostaras, S., & Wixted, J. T. (2011). An
opportunistic theory of cellular and systems consolidation. Trends in Neurosciences
34(10), 504–514. doi:10.1016/j.tins.2011.06.003
Mednick, S., Nakayama, K., & Stickgold, R. (2003). Sleep-dependent learning: A nap is as
good as a night. Nature Neuroscience 6(7), 697–698. doi:10.1038/nn1078
Miller, K. D., & MacKay, D.J.C. (1994). The role of constraints in Hebbian learning. Neural
Computation 6, 100–126. doi:10.1162/neco.1994.6.1.100
Määttä, S., Landsness, E., Sarasso, S., Ferrarelli, F., Ferreri, F., Ghilardi, M. F., & Tononi, G.
(2010). The effects of morning training on night sleep: A behavioral and EEG study.
Brain Research Bulletin 82, 118–123. doi:10.1016/j.brainresbull.2010.01.006
Nakanishi, H., Sun, Y., Nakamura, R. K., Mori, K., Ito, M., Suda, S., … & Sokoloff, L.
(1997). Positive correlations between cerebral protein synthesis rates and deep sleep in
macaca mulatta. European Journal of Neuroscience 9, 271–279. SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 62
Neuner-Jehle, M., Rhyner, T. A., & Borbély, A. A. (1995). Sleep deprivation differentially
alters the mRNA and protein levels of neurogranin in rat brain. Brain Research 685,
143–154.
Olijslagers, J. E., de Kloet, E. R., Elgersma, Y., van Woerden, G. M., Joëls, M., & Karst, H.
(2008). Rapid changes in hippocampal CA1 pyramidal cell function via pre- as well as
postsynaptic membrane mineralocorticoid receptors. European Journal of Neuroscience
27, 2542–2550. doi:10.1111/j.1460-9568.2008.06220.x
Oswald, I. (1980). Sleep as a restorative process: Human clues. Progress in Brain Research
53, 279-288.
Payne, J. D., Tucker, M. A., Ellenbogen, J. M., Wamsley, E. J., Walker, M. P., Schacter, D.
L., & Stickgold, R. (2012). Memory for semantically related and unrelated declarative
information: The benefits of sleep, the cost of wake. PLoS One 7(3), 1–7.
doi:10.1371/journal.pone.0033079
Pelayo, R., & Guilleminault, C. (2009). Sleep research and sleep medicine in historical
perspective. In Squire, L. R. (Eds.), Encyclopedia of Neuroscience, (pp. 1043–1048).
Berlin: Elsevier.
Peng, I-F., Berke, B. A., Zhu, Y., Lee, W-H., Chen, W., & Wu, C-F. (2007). Temperature-
dependent developmental plasticity of Drosophila neurons: Cell-autonomous roles of
membrane excitability, Ca2+ influx, and cAMP signaling. The Journal of Neuroscience
27(46), 12611–12622. doi:10.1523/JNEUROSCI.2179-07.2007
Rasch, B., & Born, J. (2013). About sleep’s role in memory. Physiological Reviews 93, 681–
766. doi:10.1152/physrev.00032.2012. SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 63
Rechtschaffen, A., & Bergmann, B. M. (1995). Sleep deprivation in the rat by the disk-over
water method. Behavioral Brain Research 69, 55–63. doi:10.1016/0166-
4328(95)00020-T
Rechtschaffen, A., Bergmann, B. M., Everson, C. A., Kushida, C. A., & Gilliland, M. A.
(1989). Sleep deprivation in the rat: X. Integration and discussion of the findings.
SLEEP 12(1), 68–87.
Reimund, E. (1994). The free-radical flux theory of sleep. Medical Hypotheses 43(4), 231–
233. doi:10.1016/0306-9877(94)90071-X
Ribeiro, S. (2012). Sleep and plasticity. Pfügers Archiv-European Journal of Physiology 463,
111–120. doi:10.1007/s00424-011-1031-5
Riedner, B. A., Vyazovskiy, V. V., Huber, R., Massimini, M., Esser, S., Murphy, M., &
Tononi, G. (2007). Sleep homeostasis and cortical synchronization: III. A high-density
EEG study of sleep slow waves in humans. SLEEP 30(12), 1643–1657.
Roelandse, M., & Matus, A. (2004). Hypothermia-associated loss of dendritic spines. The
Journal of Neuroscience 24(36), 7843–7847. doi:10.1523/JNEUROSCI.2872-04.2004
Schacter, D. L., & Wagner, A. D. (2013). Learning and memory. In Kandel, E. R., Schwartz,
J. H., Jessell, T. M., Siegelbaum, S. A., & Hudspeth, A. J. (Eds.), Principles of neural
science (5th ed.), (pp.1441–1460). New York: McGraw-Hill.
Scharf, M. T., Naidoo, N., Zimmerman, J. E., & Pack, A. I. (2008). The energy hypothesis of
sleep revisited. Progress in Neurobiology 86, 264–280. doi:
10.1016/j.pneurobio.2008.08.003
Shaw, P. J., Tononi, G., Greenspan, R. J., & Robinson, D. F. (2002). Stress response genes
protect against lethal effects of sleep deprivation in Drosophila. Nature 417, 287–291 SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 64
Simkus, C. R. L., & Stricker, C. (2002). Properties of mEPSCs recorded in layer II neurons of
rat barrel cortex. Journal of Physiology 545(2), 509–520.
doi:10.1113/jphysiol.2002.022095
Steriade, M. (2003). The corticothalamic system in sleep. Frontiers in Bioscience 8, 878–899.
Siegelbaum, S. A., & Kandel, E. R. (2013). Prefrontal cortex, hippocampus, and the biology
of explicit memory storage. In Kandel, E. R., Schwartz, J. H., Jessell, T. M.,
Siegelbaum, S. A., & Hudspeth, A. J. (Eds.), Principles of neural science (5th ed.), (pp.
1487–1524). New York: McGraw-Hill
Siegelbaum, Kandel, & Südhof. (2013). Transmitter release. In Kandel, E. R., Schwartz, J. H.,
Jessell, T. M., Siegelbaum, S. A., & Hudspeth, A. J. (Eds.), Principles of neural science
(5th ed.), (pp. 260–288). New York: McGraw-Hill.
Stevenson, R. D. (1985). Body size and limits to the daily range of body temperature in
terrestrial ectotherms. The American Naturalist 125(1), 102–117. Retrieved from:
http://www.jstor.org.e.bibl.liu.se/stable/2461610
Stickgold, R., James, L., & Hobson, J. A. (2000). Visual discrimination learning requires
sleep after training. Nature Neuroscience 3(12), 1237–1238. doi:10.1038/81756
Timofeev, I. (2011). Neuronal plasticity and thalamocortical sleep and waking oscillations.
Progress in Brain Research 193, 121–144. doi:10.1016/B978-0-444-53839-0.0009-0
Tononi, G., & Cirelli, C. (2001). Some considerations on sleep and neural plasticity. Archive
Italiennes De Biologie 139, 221–241.
Tononi, G., & Cirelli, C. (2003). Sleep and synaptic homeostasis: A hypothesis. Brain
Research Bulletin 62, 143–150. doi:10.1016/j.brainresbull.2003.09.004 SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 65
Tononi, G., & Cirelli, C. (2006). Sleep function and synaptic homeostasis. Sleep Medicine
Reviews 10, 49–62. doi:10.1016/j.smrv.2005.05.002
Tononi, G., & Cirelli, C. (2012). Time to be SHY? Some comments on sleep and synaptic
homeostasis. Neural Plasticity: 2012. doi:10.1155/2012/415250
Tononi, G., & Cirelli, C. (2014). Sleep and the price of plasticity: From synaptic and cellular
homeostasis to memory consolidation and integration. Neuron 81, 12–34.
doi:10.1016/j.neuron.2013.12.025
Turrigiano, G. G. (1999). Homeostatic Plasticity in Neuronal Networks: The more things
change, the more they stay the same. Trends in Neurosciences 22(5), 221–227.
doi:10.1016/S0166-2236(98)01341-1
Van Cauter, E. (2005). Endocrine physiology. In: Kryger, M., Roth, T., & Dement, W. C.
(Eds.), Principles and practice of sleep medicine (4th ed.), (pp. 266–282). Philadelphia,
PA: Elsevier.
Vazquez, J., Hall, S. C., Witkowska, H. E., & Greco, M. A. (2008). Rapid alterations in
cortical protein profiles underlie spontaneous sleep and wake bouts. Journal of Cellular
Biochemistry 105, 1472–1484. doi:10.1002/jcb.21970
Vyazovskiy, V., Borbély, A. A., & Tobler, I. (2000). Unilateral vibrissae stimulation during
waking induces interhemispheric EEG asymmetry during subsequent sleep in the rat.
Journal of Sleep Research 9, 367–371. doi:10.1046/j.1365-2869.2000.00230.x
Vyazovskiy, V. V., Cirelli, C., Pfister-Genskow, M., Faraguna, U., & Tononi, G. (2008).
Molecular and electrophysiological evidence for net synaptic potentiation in wake and
depression in sleep. Nature Neuroscience 11(2), 200–208. doi:10.1038/nn2035 SLEEP AND ITS EFFECTS ON SYNAPTIC STRENGTH 66
Vyazovskiy, V. V., Olcese, U., Lazimy, Y. M., Faraguna, U., Esser, S. K., Williams, J. C., …
& Tononi, G. (2009). Cortical firing rates and sleep homeostasis. Neuron 63, 865–878.
doi:10.1016/j.neuron.2009.08.024
Vyazovskiy, V. V., Riedner, B. A., Cirelli, C., & Tononi, G. (2007). Sleep homeostasis and
cortical synchronization: II. A local field potential study of sleep slow waves in the rat.
SLEEP 30(12), 1631–1642.
Walker, M. P., Brakefield, T., Morgan, A., Hobson, J. A., & Stickgold, R. (2002). Practice
with sleep makes perfect: Sleep-dependent motor skill learning. Neuron 35, 205–211
Wang, G., Grone, B., Colas, D., Appelbaum, L., & Mourrain, P. (2011). Synaptic plasticity in
sleep: Learning, homeostasis and disease. Trends in Neuroscience 34(9), 452–463.
doi:10.1016/j.tins.2011.07.005
Yoo, S., Hu, P. T., Gujar, N., Jolesz, F. A., & Walker, M. P. (2007). A deficit in the ability to
form new human memories without sleep. Nature Neuroscience 10(3), 385–392.
doi:10.1038/nn1851
Zhong, Y., & Wu, C-F. (2004). Neuronal activity and adenylyl cyclase in environment-
dependent plasticity of axonal outgrowth in Drosophila. The Journal of Neuroscience
24(6), 1439–1445. doi:10.1523/JNEUROSCI.0740-02.2004