1.2. Learning while we are asleep
One of the most audacious proposals throughout the history of psychology was the potential ability to learn while we sleep. The idea penetrated culture via sci-fi movies and inspired the invention of devices that claimed to teach foreign languages, facts, and even quit smoking by simply listening to audiocassettes or other devices during sleep. However, as demonstrated 60 years ago by Simon and Emmons (1956), humans cannot learn complex facts during sleep, and this dream was quickly discarded by the scientific community. Although the
sleeping brain is unable to acquire new complex information (i.e., words, images, facts, but see (Arzi et al., 2012) for learning new associations during sleep), we now know that it can be manipulated to strengthen the memory of recently acquired information. In recent years several approaches have been developed to intervene on the sleeping brain in order to modify the ongoing memory processing. Here, we provide an overview of the available techniques to modulate memory-related sleep physiology, including sensory, vestibular and electrical stimulation, as wellaspharmacological approaches.Thecurrent reviewisnotmeant to be an exhaustive literature search but to offer a general summary of possible interventions that may be used to stimulate the sleeping brain in order to shape memory consolidation.
2. Sensory stimulation
Thesleepingbrainisnotindifferent toexternalsensory information, which may modulate memory-related sleep physiology. Indeed, already in1939,Davisandcolleaguesobservedthatthepresentationofacoustic tones during sleep can elicit a SO/K-complex followed by slow (8Hz) or fast(14Hz)spindles.Also,olfactorystimulationduring sleepmodulates sleep physiology. Odor presentation (e.g., lavender) during NREM sleep induces greater SWS (Goel et al., 2005), and increases delta (0.5–4Hz) (Perl et al., 2016; Arzi et al., 2014) and slow spindle (9–12Hz) power (Perl et al., 2016). Interestingly, olfactory stimuli can be used to create new associations during sleep (Arzi et al., 2012), which can impact behaviors such as smoking habits (Arzi et al., 2014). Capitalizing on the ability of the sleeping brain to process external sensory information, Rasch and colleagues (2007) made a seminal discovery: the sleeping brain can be manipulated to strengthen the memory of specific, recently acquired information. During encoding, a sensory cue (i.e., the scent of a rose) was paired with the target information to learn (i.e., the position of two identical pairs in a grid of cards, as in the game “Memory”) and then the contextual cue (the scent) was re-presented during sleep. They observed that the odor stimulation during SWS, but not during wakefulness or during REM sleep, was able to enhance the memory for the pair of cards at morning recall. Moreover, the olfactory stimulation induced greater hippocampal activity during sleep compared to sleeping without the odor or with an odorless “vehicle” stimulation. This breakthrough study, which was the first to show that memory-related sleep physiology could be modified during sleep, had a major impact on the sleep field, leading to the development of new research paradigms, including targeted memory reactivation (TMR), rhythmic stimulation and closed-loop stimulation during sleep (described in the following paragraphs).
2.1. Stimulating with meaningful sensory cues: targeted memory reactivation (TMR)
Targetedmemoryreactivation(TMR)isawell-established paradigm that employs sensory stimulation to modulate memory consolidation during sleep (Cellini and Capuozzo, 2018). It consists of matching a sensory cue (e.g., an odor or a sound) with a target (e.g., a picture, a word) during wakefulness, and then re-presenting the cue alone during sleep (Fig. 1). This process facilitates the consolidation of the targeted information. Studies have shown that TMR can improve visual (Rasch et al., 2007) and verbal memories (Schreiner and Rasch, 2014b), enhance motor skills (Antony et al., 2012) and fear extinction (Hauner et al., 2013), and even modify implicit social biases (Hu et al., 2015). The idea behind TMR is that the sensory cue can induce a reactivation of the cued-target information, prioritizing its consolidation compared to uncued stimuli (i.e., encoded items in which the cue was not represented during sleep). TMR can be performed with different types of sensory stimuli. As mentioned in the previous paragraph, Rasch et al. (2007) were able to enhance visuospatial memories (object-location task) using an olfactory stimulus (i.e., the scent of a rose delivered via olfactometer and nasal
mask). In a series of experiments, they showed that the effect of stimulation was evident only when the odor was presented
mask). In a series of experiments, they showed that the effect of stimulation was evident only when the
mask). In a series of experiments, they showed that the effect of stimulationwasevidentonlywhentheodorwaspresentedduringSWS(in a 30-s on/30-s off sequence), and only when the odor was previously matched with the object-location to be learned. Several studies have replicated and extended these results (for recent reviews see (Cellini and Capuozzo, 2018; Schouten et al., 2017)), showing for example that the olfactory cues during SWS can induce a strong stabilization of memory traces making this information resistant to subsequent interference learning (e.g., learning new card-pair locations) (Diekelmann et al., 2011). At the physiological level, the presentation of a sensory cue during sleep increases frontal delta (1.5–4.5Hz) and parietal fast spindle (13–15Hz) activity (Rihm et al., 2014), which are presumed to coordinate reactivation andconsolidation ofdeclarative memoriesfrom the hippocampus to the cortical networks (Genzel et al., 2014; Rasch and Born, 2013). The beneficial effect of olfactory TMR is hypothesized to be due to the particular nature of the olfactory system, which projects directly to the hippocampus and the amygdala (Zelano and Sobel, 2005), along with its connections to the mediodorsal nucleus of the thalamus (Courtiol and Wilson, 2013). Studies using olfactory stimulation during sleep have been extended to other memory domains, showing positive effects on emotional memories and creativity skills, although findings on procedural memories and fear conditioning are less clear (Cellini and Capuozzo, 2018). These mixed results are likely due to the different paradigms used, and/or to strong individual differences, which may limit the benefits of olfactory TMR. Buildingontheinitial findingsfromRaschetal.(2007),Rudoyetal. (2009) introduced acoustic stimulation during sleep, which allowed to overcome some of the limitations of olfactory cuing. For example, the limitsoftheolfactorysystemmadeitdifficulttouseseveralodorsatthe same time to target individual items. Also, olfactory stimuli cannot be delivered in a temporally precise fashion. Instead, acoustic stimulation can be delivered with a high temporal accuracy, and different sounds can be used in the same experiment without impairing the auditory
system. Moreover, acoustic cues can be easily manipulated in order to create cues that are semantically related to individual items. Rudoy et al. (2009) asked participants to learn the location of 50 pictures of animals/objects displayed on a computer screen. Each picture was associated with a unique and semantically related sound (e.g., a picture of a cat with a meow). During sleep (a 60–80min nap), subjects were then presented with half of the auditory cues (with intensity ∼38dB SPL), which resulted in higher memory accuracy for the sleepcued objects, compared with the non-sleep-cued objects. Other studies have replicated these findings with different types of memories (e.g., verbal, visuospatial, procedural), whereas others have failed to find a behavioral effect (reviewed in (Cellini and Capuozzo, 2018)). At the physiological level, imaging studies report that auditory cueing increased activity in hippocampal and parahippocampal cortices (van Dongen et al., 2012; Hauner et al., 2013), as well as in the occipital cortex when the stimulation was performed either in NREM (Berkers et al., 2017) or in REM sleep (Sterpenich et al., 2014). Also, it has been observed that auditory stimulation may increase activity in theta and sigma frequency bands just after the cue presentation (Creery et al., 2015; Farthouat et al., 2016; Fuentemilla et al., 2013; Schreiner and Rasch, 2014a; Schreiner et al., 2018). As shown by Schreiner and colleagues (2018) and by studies comparing acoustic against sham stimulation (see next sections), the theta and sigma activity enhancement may be related to cue-evoked K-complexes, which drive a stronger physiological response compared to spontaneous K-complexes. Interestingly, a very recent paper suggests that theta oscillations (at 5Hz) may play a key role in orchestrating the reactivation of information both during sleep and wakefulness (Schreiner et al., 2018)
odor was present edduring SWS(in a 30-s on/30-s off sequence), and only when the odor was previously matched with the object-location to be learned. Several studies have replicated and extended these results (for recent reviews see (Cellini and Capuozzo, 2018; Schouten et al., 2017)), showing for example that the olfactory cues during SWS can induce a strong stabilization of memory traces making this information resistant to subsequent interference learning (e.g., learning new card-pair locations) (Diekelmann et al., 2011). At the physiological level, the presentation of a sensory cue during sleep increases frontal delta (1.5–4.5Hz) and parietal fast spindle (13–15Hz) activity (Rihm et al., 2014), which are presumed to coordinate reactivation and consolidation of declarative memories from the hippocampus to the cortical networks (Genzel et al., 2014; Rasch and Born, 2013). The beneficial effect of olfactory TMR is hypothesized to be due to the particular nature of the olfactory system, which projects directly to the hippocampus and the amygdala (Zelano and Sobel, 2005), along with its connections to the mediodorsal nucleus of the thalamus (Courtiol and Wilson, 2013). Studies using olfactory stimulation during sleep have been extended to other memory domains, showing positive effects on emotional memories and creativity skills, although findings on procedural memories and fear conditioning are less clear (Cellini and Capuozzo, 2018). These mixed results are likely due to the different paradigms used, and/or to strong individual differences, which may limit the benefits of olfactory TMR. Building on the initial findings from Raschetal.(2007),Rudoyetal. (2009) introduced acoustic stimulation during sleep, which allowed to overcome some of the limitations of olfactory cuing. For example, the limits of the olfactory system made it difficult to use several odors at the same time to target individual items. Also, olfactory stimuli cannot be delivered in a temporally precise fashion. Instead, acoustic stimulation can be delivered with a high temporal accuracy, and different sounds can be used in the same experiment without impairing the auditory
system. Moreover, acoustic cues can be easily manipulated in order to create cues that are semantically related to individual items. Rudoy et al. (2009) asked participants to learn the location of 50 pictures of animals/objects displayed on a computer screen. Each picture was associated with a unique and semantically related sound (e.g., a picture of a cat with a meow). During sleep (a 60–80min nap), subjects were then presented with half of the auditory cues (with intensity ∼38dB SPL), which resulted in higher memory accuracy for the sleepcued objects, compared with the non-sleep-cued objects. Other studies have replicated these findings with different types of memories (e.g., verbal, visuospatial, procedural), whereas others have failed to find a behavioral effect (reviewed in (Cellini and Capuozzo, 2018)). At the physiological level, imaging studies report that auditory cueing increased activity in hippocampal and parahippocampal cortices (van Dongen et al., 2012; Hauner et al., 2013), as well as in the occipital cortex when the stimulation was performed either in NREM (Berkers et al., 2017) or in REM sleep (Sterpenich et al., 2014). Also, it has been observed that auditory stimulation may increase activity in theta and sigma frequency bands just after the cue presentation (Creery et al., 2015; Farthouat et al., 2016; Fuentemilla et al., 2013; Schreiner and Rasch, 2014a; Schreiner et al., 2018). As shown by Schreiner and colleagues (2018) and by studies comparing acoustic against sham stimulation (see next sections), the theta and sigma activity enhancement may be related to cue-evoked K-complexes, which drive a stronger physiological response compared to spontaneous K-complexes. Interestingly, a very recent paper suggests that theta oscillations (at 5Hz) may play a key role in orchestrating the reactivation of information both during sleep and wakefulness (Schreiner et al., 2018)
duringSWS(in a 30-s on/30-s off sequence), and only when the odor was previously matched with the object-location to be learned. Several studies have replicated and extended these results (for recent reviews see (Cellini and Capuozzo, 2018; Schouten et al., 2017)), showing for example that the olfactory cues during SWS can induce a strong stabilization of memory traces making this information resistant to subsequent interference learning (e.g., learning new card-pair locations) (Diekelmann et al., 2011). At the physiological level, the presentation of a sensory cue during sleep increases frontal delta (1.5–4.5Hz) and parietal fast spindle (13–15Hz) activity (Rihm et al., 2014), which are presumed to coordinate reactivation and consolidation memories from the hippocampus to the cortical networks (Genzel et al., 2014; Rasch and Born, 2013). The beneficial effect of olfactory TMR is hypothesized to be due to the particular nature of the olfactory system, which projects directly to the hippocampus and the amygdala (Zelano and Sobel, 2005), along with its connections to the mediodorsal nucleus of the thalamus (Courtiol and Wilson, 2013). Studies using olfactory stimulation during sleep have been extended to other memory domains, showing positive effects on emotional memories and creativity skills, although findings on procedural memories and fear conditioning are less clear (Cellini and Capuozzo, 2018). These mixed results are likely due to the different paradigms used, and/or to strong individual differences, which may limit the benefits of olfactory TMR. Building on the initial findings from Raschetal.(2007),Rudoyetal. (2009) introduced acoustic stimulation during sleep, which allowed to overcome some of the limitations of olfactory cuing. For example, the limits of the olfactory system made it difficult to use several odors at the same time to target individual items. Also, olfactory stimuli cannot be delivered in a temporally precise fashion. Instead, acoustic stimulation can be delivered with a high temporal accuracy, and different sounds can be used in the same experiment without impairing the auditory
system. Moreover, acoustic cues can be easily manipulated in order to create cues that are semantically related to individual items. Rudoy et al. (2009) asked participants to learn the location of 50 pictures of animals/objects displayed on a computer screen. Each picture was associated with a unique and semantically related sound (e.g., a picture of a cat with a meow). During sleep (a 60–80min nap), subjects were then presented with half of the auditory cues (with intensity ∼38dB SPL), which resulted in higher memory accuracy for the sleep cued objects, compared with the non-sleep-cued objects. Other studies have replicated these findings with different types of memories (e.g., verbal, visuospatial, procedural), whereas others have failed to find a behavioral effect (reviewed in (Cellini and Capuozzo, 2018)). At the physiological level, imaging studies report that auditory cueing increased activity in hippocampal and para hippocampal cortices (van Dongen et al., 2012; Hauner et al., 2013), as well as in the occipital cortex when the stimulation was performed either in NREM (Berkers et al., 2017) or in REM sleep (Sterpenich et al., 2014). Also, it has been observed that auditory stimulation may increase activity in theta and sigma frequency bands just after the cue presentation (Creery et al., 2015; Farthouat et al., 2016; Fuentemilla et al., 2013; Schreiner and Rasch, 2014a; Schreiner et al., 2018). As shown by Schreiner and colleagues (2018) and by studies comparing acoustic against sham stimulation (see next sections), the theta and sigma activity enhancement may be related to cue-evoked K-complexes, which drive a stronger physiological response compared to spontaneous K-complexes. Interestingly, a very recent paper suggests that theta oscillations (at 5Hz) may play a key role in orchestrating the reactivation of information both during sleep and wakefulness (Schreiner et al., 2018)