Role of slow oscillations, sleep spindles and sharp-wave ripples in memory consolidation

The article reviews what is known about the interplay of cortical slow oscillations, sleep spindles, and ripples in hippocampus during sleep. The role of those processes in memory consolidation is discussed.

Keywords: memory consolidation, slow oscillations, sleep spindles, ripples, spindle-ripple events, hippocampus.

Статья обозревает известные данные про взаимодействие медленных осцилляций, сонных веретен и рипплов гиппокампа во время сна. Обсуждается роль данных процессов в консолидации памяти.

Ключевые слова: консолидация памяти, медленноволновая активность, сонные веретена, рипплы, гиппокамп.

I. Introduction

Memory consolidation is not a new topic in neurophysiology, yet it seems that its popularity among scientists has been hugely revived only in recent years. Scientists of the 19 ^th century and the first half of the 20 ^th century were mainly dedicated to studying memory with psychological and neuropsychological methods (one can instantly think of Alexander Luria or Brenda Milner). It was already clear back then that the temporal lobes and, specifically, hippocampi located inside them play an important role in memory processes [1,2]. However, neuroscientists only started to carefully study physiological and cellular mechanisms underlying memory consolidation in a few recent decades.

Sleep’s role in memory is another intriguing topic that can shed some light on memory processes in general. Already back in 1885, the first psychologist who studied memory experimentally, Hermann Ebbinghaus, noticed that forgetting is reduced when sleep occurred in the retention interval [3]. In modern studies, polysomnography and selective sleep deprivation, combined with psychological testing, are common methods for studying sleep effects on memory function. However, for many years research in this area remained focused only on slow-wave sleep (SWS) and rapid-eye-movement sleep (REM), missing out on the possible influence of sleep spindles and other intricate EEG phenomena on memory consolidation. In recent years, new data offered some insights into how these electrophysiological events are connected with memory formation and intertwined with the processes happening in wakefulness.

II. Slow oscillations

Slow oscillations are EEG waves with the frequency of 0.5–4.0 Hz that are occurring during non-REM (NREM) slow-wave sleep (SWS) in a healthy subject [4]. A slow wave itself consists of up (depolarization) and down (hyperpolarization) states [5].

Slow oscillations during sleep have been persistently associated with memory and learning. Learning and active exploring in animals correlates with the increasing of slow oscillations in the subsequent sleep [6,7]. Analogous results were obtained in humans: intense learning of word pairs enhanced amplitudes of the slow oscillations [8]. Transcranial stimulation of slow oscillations during NREM sleep, too, correlated with the better performance in word pairs test after sleep [9], and selective suppression of slow oscillation can worsen the performance in memory tests [10]. Conversely, an immobilized arm during the day resulted in reduced slow oscillations in contralateral motor cortex during the subsequent sleep [11].

III. Sleep spindles

Sleep spindles are EEG events lying in the 11–15 Hz sigma-frequency band that are most prominent in the stage 2 of NREM sleep. [15] Sleep spindles are generated by thalamus through the activation of calcium ion channels [7].

Classical “half-night” experiments with sleep deprivation happening either in the beginning or the end of a night usually result in memory impairments. Authors often conclude that such effects are causally connected with deprivation of SWS and REM sleep. However, it is possible that many studies conducted in this paradigm missed the causality of sleep spindles and memory consolidation. To illustrate, in a study conducted in 2009 authors did not detect significant effects of REM or SWS deprivation but found that parameters of sleep spindles correlated with the successful performance in the morning declarative memory testing [16].

The problem with focusing research on sleep spindles is that it is not technically possible to suppress them. In order to fall into slow-wave sleep or REM sleep, a person has to first go through the NREM2 stage, which is characterized by sleep spindles and K-complexes. Hence, a deprivation of NREM2 would also mean a deprivation of NREM3 and REM, and such results would not be informative. However, correlation studies and induction of sleep spindles via stimulation remain possible.

There are 2 types of sleep spindles: fast and slow ones. Fast spindles are associated with a general memory improvement after sleep [17,18], whereas slow spindles more often correlate with such an improvement only in especially gifted volunteers [18].

Training in declarative memory tasks, such as memorizing word pairs or exploring a virtual maze, results in an increased amount of sleep spindles in the subsequent sleep, and only people with that increasement show better performance in the morning recall testing [19,20]. Experiments with rats also showed that learning in wakefulness increases the amount of sleep spindles later on in the sleep [21]. Overall, sleep spindles are consistently described as connected with consolidation of declarative memories.

There are data pointing to spindles’ connection with reactivations in hippocampus and neocortex that happen after learning [22]. It is not unreasonable to assume that sleep spindles are involved in conducting newly acquired information from the hippocampus to the neocortex. If that is the case, spindles must be crucial for the process of active consolidation.

IV. Sharp-wave ripples

Sharp-wave ripples (SW-R) are EEG events found in the hippocampus. Depolarizing sharp waves are generated in CA3 subfield and are superimposed by ripple activity generated by CA1 subfield. At the behavioral level, SW-R can be observed during quiet wakefulness and slow-wave sleep. [7]

Long-term potentiation is a well-studied phenomenon considered to be a cellular form of learning and memory [23]. Interestingly, hippocampal stimulation protocols that induce LTP also facilitate the generation of SW-Rs in the hippocampus, and SW-Rs during sleep can be initiated by neurons potentiated during preceding wakefulness [24].

V. Interplay of mentioned EEG phenomena in memory consolidation

Slow oscillations are generated in neocortex, which has a top-down control over sleep spindles and hippocampal SW-Rs. Slow oscillations have a synchronizing effect on most structures involved in memory consolidation, including thalamus and hippocampus, with the latter generating ripples and the former generating spindles. [7]

Prior learning appears to strengthen the top-down control of slow oscillations on spindles and ripples. In human studies, memorizing new words had a stimulating effect on fast spindle activity, with this increase concentrating on the up-states of the slow wave, whereas no changes were observed in hyperpolarizing down-states [8,26]. Furthermore, SW-Rs seem to be temporally coupled to sleep spindles [27].

All the aforementioned led to the emergence of the term ‘spindle-ripple events’, meaning a single ripple that is temporally linked to a fast sleep spindle [28]. This event may be a mechanism of transfer a labile memory trace from hippocampus to neocortex. Presumably, it can be described as a cycle: spindles generated in thalamus facilitate the appearance of hippocampal ripples that, once again, contribute to spindle activity [13] that then reaches the cortex during the depolarizing up-state of slow waves. This way, the interplay of those EEG phenomena may lead to memory consolidation and, therefore, a long-term storage of information in neocortical neural circuits.

References:

1. Smith, M. L. The role of the right hippocampus in the recall of spatial location / M. L. Smith, B. Milner. — Текст: непосредственный // Neuropsychologia. — 1981. — № 19 (6). — С. 781–793.
2. Neuropsychology of the temporal lobe: Luria's and contemporary conceptions / T. A. Aversi-Ferreira, B. H. Tamaishi-Watanabe, M. P., de Fátima M, R. A. G.M. F. Aversi-Ferreira — Текст: непосредственный // Dementia & Neuropsychologia. — 2019. — № 13 (3).
3. Ebbinghaus, H. Über das Gedächtnis: Untersuchungen zur experimentellen Psychologie / H. Ebbinghaus. — Leipzig, Germany: Duncker & Humblot, 1885. — 169 c. — Текст: непосредственный.
4. Achermann, P. Temporal evolution of coherence and power in the human sleep electroencephalogram / P. Achermann, AA Borbély. — Текст: непосредственный // Journal of Sleep Research. — 1998. — № 7 (1). — С. 36–41.
5. Contreras, D. Mechanisms of long-lasting hyperpolarizations underlying slow sleep oscillations in cat corticothalamic networks. / D. Contreras, I. Timofeev, M. Steriade. — Текст: непосредственный // Journal of Physiology. — 1996. — № 494. — С. 251–264.
6. Borbély, K. D. Effect of unilateral somatosensory stimulation prior to sleep on the sleep EEG in humans / K. D. Borbély. — Текст: непосредственный // Journal of Sleep Research. — 1994. — № 3. — С. 159 –164.
7. Rasch, B. About sleep's role in memory / B. Rasch, J. Born. — Текст: непосредственный // Physiological Reviews. — 2013. — № 93. — С. 681–766.
8. The influence of learning on sleep slow oscillations and associated spindles and ripples in humans and rats / M. Mölle, O. Eschenko, S. Gais [и др.]. — Текст: непосредственный // European Journal of Neuroscience. — 2009. — № 29. — С. 1071–1081.
9. Transcranial direct current stimulation during sleep improves declarative memory / L. Marshall, M. Molle, M. Hallschmid, J. Born. — Текст: непосредственный // Journal of Neuroscience. — 2004. — № 24. — С. 9985–9992.
10. Sleep-dependent improvement in visuomotor learning: a causal role for slow waves / E. C. Landsness, D. Crupi, B. K. Hulse [и др.]. — Текст: непосредственный // Sleep. — 2009. — № 32. — С. 1273–1284.
11. Arm immobilization causes cortical plastic changes and locally decreases sleep slow wave activity / R. Huber, MF Ghilardi, M. Massimini [и др.]. — Текст: непосредственный // Nature Neuroscience. — 2006. — № 9. — С. 1169 –1176.
12. Spontaneous high-frequency (10–80 Hz) oscillations during up states in the cerebral cortex in vitro / A. Compte, R. Reig, V. F. Descalzo [и др.]. — Текст: непосредственный // Journal of Neuroscience. — 2008. — № 28. — С. 13828 –13844.
13. Mölle, M. Slow oscillations orchestrating fast oscillations and memory consolidation / M. Mölle, J. Born. — Текст: непосредственный // Progress in Brain Research. — 2011. — № 193. — С. 93–110.
14. Communication between neocortex and hippocampus during sleep in rodents / A. Sirota, J. Csicsvari, D. Buhl, G. Buzsáki. — Текст: непосредственный // Proceedings of the National Academy of Sciences of the United States of America. — 2003. — № 100. — С. 2065–2069.
15. Characterizing sleep spindles in 11,630 individuals from the National Sleep Research Resource / S. M. Purcell, D. S. Manoach, C. Demanuele [и др.]. — Текст: непосредственный // Nature Communications. — 2017. — № 8.
16. Slow wave sleep and REM sleep awakenings do not affect sleep dependent memory consolidation / L. Genzel, M. Dresler, R. Wehrle [и др.]. — Текст: непосредственный // Sleep. — 2009. — № 32 (3). — С. 302–310.
17. Sleep spindle-related activity in the human EEG and its relation to general cognitive and learning abilities / M. Schabus, ̈. K. Ho, G. Gruber [и др.]. — Текст: непосредственный // European Journal of Neuroscience. — 2006. — № 23. — С. 1738–1746.
18. Interindividual sleep spindle differences and their relation to learning-related enhancements / M. Schabus, K. Hoedlmoser, T. Pecherstorfer [и др.]. — Текст: непосредственный // Brain Research. — 2007. — № 1191. — С. 127–135.
19. Learning-dependent increases in sleep spindle density / S. Gais, M. Mölle, K. Helms, J. Born. — Текст: непосредственный // Journal of Neuroscience. — 2002. — № 22. — С. 6830–6834.
20. Sleep spindles and their significance for declarative memory consolidation / M. Schabus, G. Gruber, S. Parapatics [и др.]. — Текст: непосредственный // Sleep. — 2004. — № 27. — С. 1479 –1485.
21. Fogel, S. M. Evidence for 2-stage models of sleep and memory: learning-dependent changes in spindles and theta in rats / S. M. Fogel, C. T. Smith, R. J. Beninger. — Текст: непосредственный // Brain Research Bulletin. — 2009. — № 79. — С. 445– 451.
22. Sleep spindle-related reactivation of category-specific cortical regions after learning face-scene associations / T. O. Bergmann, M. Mölle, J. Diedrichs [и др.]. — Текст: непосредственный // Neuroimage. — 2012. — № 59. — С. 2733–2742.
23. Nicoll, R. A. A brief history of long-term potentiation / R. A. Nicoll. — Текст: непосредственный // Neuron. — 2017. — № 93 (2). — С. 281–290.
24. Induction of sharp wave-ripple complexes in vitro and reorganization of hippocampal networks / C. J. Behrens, L.P. van den Boom, L. Hoz de [и др.]. — Текст: непосредственный // Nature Neuroscience. — 2005. — № 8. — С. 1560 –1567.
25. Sustained increase in hippocampal sharp-wave ripple activity during slow-wave sleep after learning / O. Eschenko, W. Ramadan, M. Molle [и др.]. — Текст: непосредственный // Learning & Memory. — 2008. — № 15. — С. 222–228.
26. Fast and slow spindles during the sleep slow oscillation: disparate coalescence and engagement in memory processing / M. Mölle, T. O. Bergmann, L. Marshall, J. Born. — Текст: непосредственный // Sleep. — 2011. — № 34. — С. 1411–1421.
27. Siapas, A. G. Coordinated interactions between hippocampal ripples and cortical spindles during slow-wave sleep / A. G. Siapas, M. Wilson. — Текст: непосредственный // Neuron. — 1998. — № 21 (5). — С. 1123–1128.
28. Fine-tuned coupling between human parahippocampal ripples and sleep spindles / Z. Clemens, M. Mölle, L. Eross [и др.]. — Текст: непосредственный // European Journal of Neuroscience. — 2011. — № 33. — С. 511–520.