Skill Memory: Mind the Ever-Decreasing Gap for Offline Processing
2019; Elsevier BV; Volume: 29; Issue: 8 Linguagem: Inglês
10.1016/j.cub.2019.03.007
ISSN1879-0445
Autores Tópico(s)Photoreceptor and optogenetics research
ResumoSkills continue to be enhanced even once practice has ceased. Such offline improvements were for a time regarded as the sole preserve of sleep, but recent work shows that they can occur within only a few seconds. Skills continue to be enhanced even once practice has ceased. Such offline improvements were for a time regarded as the sole preserve of sleep, but recent work shows that they can occur within only a few seconds. A memory continues to be processed by the brain following its formation. It was not so long ago that such offline processing was considered the strict preserve of sleep. Improvements following practice would only develop during sleep [1Walker M.P. Brakefield T. Morgan A. Hobson J.A. Stickgold R. Practice with sleep makes perfect: sleep-dependent motor skill learning.Neuron. 2002; 35: 205-211Abstract Full Text Full Text PDF PubMed Scopus (929) Google Scholar]; (for a review see [2King B.R. Hoedlmoser K. Hirschauer F. Dolfen N. Albouy G. Sleeping on the motor engram: The multifaceted nature of sleep-related motor memory consolidation.Neurosci. Biobehav. Rev. 2017; 80: 1-22Crossref PubMed Scopus (95) Google Scholar]). Subsequent work showed that improvements of similar magnitude could develop during wakefulness over the hours following the acquisition of a novel skill [3Robertson E.M. Pascual-Leone A. Press D.Z. Awareness modifies the skill-learning benefits of sleep.Curr. Biol. 2004; 14: 208-212Abstract Full Text Full Text PDF PubMed Google Scholar, 4Mosha N. Robertson E.M. Unstable memories create a high-level representation that enables learning transfer.Curr. Biol. 2016; 26: 100-105Abstract Full Text Full Text PDF PubMed Scopus (23) Google Scholar]. Recently, offline improvements were shown to develop within only a matter of minutes after practicing a novel skill [5Hotermans C. Peigneux P. Maertens de Noordhout A. Moonen G. Maquet P. Early boost and slow consolidation in motor skill learning.Learn. Mem. 2006; 13: 580-583Crossref PubMed Scopus (85) Google Scholar]. Now, a new study by Bönstrup et al. [6Bönstrup M. Iturrate I. Thompson R. Cruciani G. Censor N. Cohen L.G. A rapid form of offline consolidation in skill learning.Curr Biol. 2019; 29: 1346-1351Abstract Full Text Full Text PDF Scopus (39) Google Scholar], reported in this issue of Current Biology, convincingly shows that offline improvements can develop within only seconds. The ever-decreasing gap necessary for offline improvements gives fresh insight not only into offline processing, but also into the nature of learning more broadly. In the new work [6Bönstrup M. Iturrate I. Thompson R. Cruciani G. Censor N. Cohen L.G. A rapid form of offline consolidation in skill learning.Curr Biol. 2019; 29: 1346-1351Abstract Full Text Full Text PDF Scopus (39) Google Scholar], human participants learnt a novel skill —a sequence of finger movements — whilst simultaneously having their brain rhythms recorded using magnetoencephalography (MEG). Short, ten-second blocks of practice were interleaved with equally short intervals of rest. During practice there was little or no performance improvements. By contrast, in the rest interval between practice blocks there were substantial improvements. Individually these offline improvements were quite small, so called micro-offline improvements; however, cumulatively over the entire period of early practice these offline improvements were substantial. In fact, their cumulative total accounted for the amount of skill that would normally be attributed to practice. As a consequence, the defining characteristic of these offline improvements is not so much their small magnitude; instead, it is the speed of their development — they are ultra-fast offline improvements. The emergence of these ultra-fast offline improvements could in principle be due to the dissipation of fatigue during rest. During practice, a skill is being acquired and simultaneously fatigue both physical and mental may be accumulating (Figure 1A). Performance can be seen as the result of the interaction between skill and fatigue: on the one hand, skill, improving performance, and on the other hand, fatigue, impairing performance. As a consequence, during rest, when fatigue dissipates, the skill acquired during practice is expressed. This would explain the development of the ultra-fast offline improvements during rest. Rather than fatigue dissipating, the ultra-fast offline improvements could alternatively develop as a result of skill enhancement during rest. Patterns of neuronal activity present during the formation of a skill memory are played out, or replayed, during subsequent rest (Figure 1B) [7Ramanathan D.S. Gulati T. Ganguly K. Sleep-dependent reactivation of ensembles in motor cortex promotes skill consolidation.PLoS Biol. 2015; 13: e1002263Crossref PubMed Scopus (94) Google Scholar]. These replay events have been linked to the offline processing of memories, their consolidation, and specifically to their enhancement [8Genzel L. Robertson E.M. To replay, perchance to consolidate.PLoS Biol. 2015; 13: e1002285Crossref PubMed Scopus (19) Google Scholar]. A curious and poorly understood feature of replay activity is that it is often a much-accelerated (time-compressed) version of the activity during memory formation: whilst the pattern of activity present during memory formation is preserved it occurs far faster during subsequent replay [9Euston D.R. Tatsuno M. McNaughton B.L. Fast-forward playback of recent memory sequences in prefrontal cortex during sleep.Science. 2007; 318: 1147-1150Crossref PubMed Scopus (376) Google Scholar]. As a consequence, a sufficient number of replay events may occur even over the 10 second interval used in the Bönstrup et al. [6Bönstrup M. Iturrate I. Thompson R. Cruciani G. Censor N. Cohen L.G. A rapid form of offline consolidation in skill learning.Curr Biol. 2019; 29: 1346-1351Abstract Full Text Full Text PDF Scopus (39) Google Scholar] study to drive the formation of the ultra-fast offline improvements. During these 10 second intervals there is a decrease in the amplitude of brain oscillations specifically within the frequency range of 16–22 Hz — so-called β power — which has been linked to replay [10Spitzer, B., and Haegens, S. (2017). Beyond the status quo: a role for beta oscillations in endogenous content (re)activation. eNeuro 4, https://doi.org/10.1523/ENEURO.0170-17.2017.Google Scholar]. Overall, replay, with its accelerated pattern of firing, is ideally suited to driving offline improvements over short intervals, and a marker of replay (a β power decrease) occurs during these intervals. Together, these new findings from Bönstrup et al. [6Bönstrup M. Iturrate I. Thompson R. Cruciani G. Censor N. Cohen L.G. A rapid form of offline consolidation in skill learning.Curr Biol. 2019; 29: 1346-1351Abstract Full Text Full Text PDF Scopus (39) Google Scholar] converge with earlier work to make replay a potential candidate for creating ultra-fast offline improvements. Another possibility is that ultra-fast offline improvements emerge as a result of the loss of an inhibitory drive during the short rest intervals. An inhibition within the motor cortex prevents the development of offline improvements over the subsequent hours of wakefulness [11Tunovic S. Press D.Z. Robertson E.M. A physiological signal that prevents motor skill improvements during consolidation.J. Neurosci. 2014; 34: 5302-5310Crossref PubMed Scopus (59) Google Scholar]. When this inhibition is removed, offline improvements develop over wakefulness (Figure 1C). The ultra-fast offline improvements may similarly be associated with a decrease in inhibition: they are inversely correlated with β power, a measure of inhibition. As a consequence, ultra-fast offline improvements emerge as this inhibitory measure decreases, which demonstrates that the removal of an inhibitory drive may, at least in part, be responsible for the development of ultra-fast offline improvements during rest. These various scenarios could explain the emergence of ultra-fast offline improvements. However, they are not mutually exclusive. Each could make a contribution to ultra-fast offline improvements developing. In fact, they could be making complementary contributions. For instance, a decrease in the inhibition of motor circuits may be necessary for replay to take place in those circuits during rest, and so critical for the subsequent development of offline improvements. As a consequence, teasing apart and determining the precise contributions of each of these potential mechanisms to the development of ultra-fast offline improvements is non-trivial. Nonetheless, substantial progress could be made in determining the mechanistic basis for ultra-fast offline improvements. One experimental approach would be to disrupt neuronal activity during the 10 second rests. This could be achieved in humans using either single pulses or short bursts of transcranial magnetic stimulation (TMS), while a similar high temporal resolution disruption can be achieved in other animals using optogenetics [12Robertson E.M. Theoret H. Pascual-Leone A. Studies in cognition: the problems solved and created by transcranial magnetic stimulation.J. Cogn. Neurosci. 2003; 15: 948-960Crossref PubMed Scopus (285) Google Scholar, 13Yizhar O. Fenno L.E. Davidson T.J. Mogri M. Deisseroth K. Optogenetics in neural systems.Neuron. 2011; 71: 9-34Abstract Full Text Full Text PDF PubMed Scopus (1341) Google Scholar]. Disrupting neural activity would prevent replay, or other processes driving plastic changes within brain circuits, and so prevent ultra-fast improvements from developing. By contrast, the dissipation of fatigue would continue to allow the emergence of ultra-fast offline improvements despite the disruption to neuronal activity. Thus, a simple experimental approach could determine the contributions of different mechanisms to ultra-fast offline improvements. The same basic mechanisms have been proposed to support a diverse range of offline improvements. On occasion, offline improvements are state-dependent: only developing over sleep, or having different qualitative properties depending on whether they develop over sleep or wakefulness [1Walker M.P. Brakefield T. Morgan A. Hobson J.A. Stickgold R. Practice with sleep makes perfect: sleep-dependent motor skill learning.Neuron. 2002; 35: 205-211Abstract Full Text Full Text PDF PubMed Scopus (929) Google Scholar, 14Cohen D.A. Pascual-Leone A. Press D.Z. Robertson E.M. Off-line learning of motor skill memory: a double dissociation of goal and movement.Proc. Natl. Acad. Sci. USA. 2005; 102: 18237-18241Crossref PubMed Scopus (203) Google Scholar]; while, on other occasions, they develop over hours, minutes, or even, as elegantly demonstrated in the new study by Bönstrup et al. [6Bönstrup M. Iturrate I. Thompson R. Cruciani G. Censor N. Cohen L.G. A rapid form of offline consolidation in skill learning.Curr Biol. 2019; 29: 1346-1351Abstract Full Text Full Text PDF Scopus (39) Google Scholar], in as little as a few seconds [3Robertson E.M. Pascual-Leone A. Press D.Z. Awareness modifies the skill-learning benefits of sleep.Curr. Biol. 2004; 14: 208-212Abstract Full Text Full Text PDF PubMed Google Scholar, 5Hotermans C. Peigneux P. Maertens de Noordhout A. Moonen G. Maquet P. Early boost and slow consolidation in motor skill learning.Learn. Mem. 2006; 13: 580-583Crossref PubMed Scopus (85) Google Scholar, 6Bönstrup M. Iturrate I. Thompson R. Cruciani G. Censor N. Cohen L.G. A rapid form of offline consolidation in skill learning.Curr Biol. 2019; 29: 1346-1351Abstract Full Text Full Text PDF Scopus (39) Google Scholar]. Yet, in biology, the same simple rules, or mechanisms, can give rise to an enormous diversity — for example, in the pattern and shape of limbs — and this same principle may also apply to offline processing, and more broadly behaviour [15Green J.B. Sharpe J. Positional information and reaction-diffusion: two big ideas in developmental biology combine.Development. 2015; 142: 1203-1211Crossref PubMed Scopus (222) Google Scholar, 16Turing A.M. The chemical basis of morphogenesis.Philos. Trans. R. Soc. Lond. B. 1952; 237: 37-72Crossref Google Scholar]. Practice leads to skill, but at least in the new work [6Bönstrup M. Iturrate I. Thompson R. Cruciani G. Censor N. Cohen L.G. A rapid form of offline consolidation in skill learning.Curr Biol. 2019; 29: 1346-1351Abstract Full Text Full Text PDF Scopus (39) Google Scholar], the skill was not acquired during practice; instead, it developed during the intervening period of rest. Acquiring skill predominately through offline mechanisms may alter its properties, and consequently, explain the greater skill retained following interleaved, as opposed to a single block, of practice, when skill is acquired 'online' during practice [17Schmidt R.A. Bjork R.A. New conceptualizaitons of practice: common principles in three paradigms suggest new concepts for training.Psycholog. Sci. 1992; 3: 207-217Crossref Scopus (1187) Google Scholar]. However, skill may not ever be acquired during practice: it may always be acquired following practice. Practice may trigger a set of offline processes which, seconds later, lead to the acquisition of skill. Rather than occurring during practice, skill learning occurs several seconds after practice. Learning lags practice and so occurs offline after practice has ceased. This lag may be the inevitable consequence of implementing learning within a biological substrate, dependent upon time-consuming processes such as protein synthesis [18Robertson E.M. Memory instability as a gateway to generalization.PLoS Biol. 2018; 16: e2004633Crossref PubMed Scopus (25) Google Scholar]. Until now lengthy blocks of practice have perhaps hidden this short temporal delay between practice and subsequent skill acquisition. The skill is acquired after practice, but, more practice is being performed, and so the skill is assumed (misattributed) to come from that subsequent practice. Only when practice is massively extended, and there is no subsequent need for offline processing — with prolonged practice a skill memory does not require subsequent offline stabilisation — does it become apparent that offline processing is occurring latently during practice [19Shibata K. Sasaki Y. Bang J.W. Walsh E.G. Machizawa M.G. Tamaki M. Chang L.H. Watanabe T. Overlearning hyperstabilizes a skill by rapidly making neurochemical processing inhibitory-dominant.Nat. Neurosci. 2017; 20: 470-475Crossref PubMed Scopus (90) Google Scholar]. By elegantly distinguishing between the contributions of practice and rest to skill acquisition, Bönstrup et al. [6Bönstrup M. Iturrate I. Thompson R. Cruciani G. Censor N. Cohen L.G. A rapid form of offline consolidation in skill learning.Curr Biol. 2019; 29: 1346-1351Abstract Full Text Full Text PDF Scopus (39) Google Scholar] have challenged the idea that skill is acquired during practice, and shown instead that much, if not all, our skill is acquired offline during rest. A Rapid Form of Offline Consolidation in Skill LearningBönstrup et al.Current BiologyMarch 28, 2019In BriefBönstrup et al. take an unprecedented close look at the time course of online motor-skill learning. They find that relevant performance improvements occur during short periods of rest. Frontoparietal beta oscillatory activity predicts those micro-offline gains. This rapid form of consolidation substantially contributes to early skill learning. Full-Text PDF Open Archive
Referência(s)