The offline brain: understanding memory consolidation and reconsolidation

Sleep strengthens and stabilizes newly acquired memories in an active process of system consolidation. This process is assumed to rely on covert reactivations of new memories that occur spontaneously after learning, mainly during slow-wave sleep (SWS), but can also be externally triggered by learning-associated memory cues. In a series of experiments, we show that the application of learning-associated odor cues during sleep resulted in an immediate stabilization of new memories, whereas similar odor reactivations in the wake state induced memory labilization. Functional magnetic resonance imaging revealed different activation patterns following reminder presentation during sleep and wakefulness. Moreover, different types of reminders (complete/incomplete) exerted different effects on memory during sleep and wakefulness. While in the wake state, only incomplete reminders but not complete reminders labilized the memory traces, both the incomplete reminder and the complete reminder stabilized memories during sleep. This evidence collectively suggests that reactivation has different effects on memory during wakefulness and sleep, presumably serving different functions for the dynamic long-term storage of memory.

Dr Susanne Diekelman, University of Tubingen, Germany

Dr Susanne Diekelman, University of Tubingen, Germany

Dr Susanne Diekelmann is a Post-Doc and Research Associate at the Institute of Medical Psychology and Behavioural Neurobiology at the University of Tübingen. She studied Psychology and Philosophy at the Dresden University of Technology and at King’s College London. In 2011, she obtained her PhD at the University of Lübeck under the supervision of Jan Born. Her main research interest is on the functional role of sleep for processes of learning and memory formation, with a particular focus on memory reactivation during sleep. Her research has been published in a number of high-impact journals, including Nature Reviews Neuroscience, Nature Neuroscience, and Journal of Neuroscience. Beyond her research on sleep and memory, Susanne has a strong interest in philosophical and ethical questions arising from neuroscience research, particularly concerning cognitive enhancement.

The majority of sleep behavioural research is conducted in humans, results which are complemented by extensive animal experiments in rodents. Here we present neural and behavioural evidence for a role of sleep in sensorimotor vocal learning in juvenile and adult songbirds, and for perceptual learning in adult songbirds (the latter mostly behavioural work). We show deep similarities of the behavioural results in songbirds and humans, highlight similarities and differences between the songbird and rodent results, and discuss the implications of these similarities and differences. We show preliminary evidence that within Aves, the complex pattern of sleep observed in songbirds extends at least to parrots. Knowing the extent to which the behavioural and neurophysiological results apply broadly, beyond mammals and perhaps far across the animal kingdom, can open opportunities for developing new advantageous model systems to gain deeper insights into the neurobiology of sleep. Our results, in complement with recent data from reptiles and other poikilotherms and invertebrates, suggest the possibility of a broad diversity of species that exhibit sleep dependent processing. This places constraints on neuronal network models to explain these behaviours.

Professor Daniel Margoliash, University of Chicago, USA

Professor Daniel Margoliash, University of Chicago, USA

Dr Margoliash is interested in brain mechanisms of behaviour, focusing on plasticity and vocal behaviour in songbirds, and more recently in humans. One set of studies has explored the role of sleep in developmental song learning and adult song maintenance. Single cell neuronal replay during sleep is a feature of certain song system forebrain neurons, providing a rare opportunity to examine the neurobiology of sleep dependent motor skill learning. Vocal production is efficiently described in a non-linear dynamical systems framework. Ongoing work examines neuronal representations and plasticity in this framework, for example the recently described novel form of information representation, such that a bird's song is mapped onto uniform intrinsic properties in a population of neurons, and feedback errors rapidly inducing inhomogeneity in that representation.

Dr Margoliash received his BSc, MSc and PhD from Caltech, completing his doctoral research on auditory responses of song system neurons in the laboratory of Masakazu (Mark) Konishi.  He then conducted postdoctoral studies on echolocating bats with Nobuo Suga in Washington University (St. Louis) before joining the faculty at the University of Chicago. He is professor in the Department of Organismal Biology and Anatomy and the Department of Psychology.

The hippocampus and the amygdala are two structures required for emotional memory. The hippocampus, through place cells, is believed to encode the spatial or contextual part of the memory. During slow-wave sleep, the activity of place cells is replayed in the same order as during the preceding learning epoch. These reactivations specifically occur during local field potential (LFP) short oscillatory events associated with highly synchronous neuronal activity called “ripples”. The group have shown previously that the specific suppression of ripples during sleep impairs performance on a spatial task, underlying their crucial role in memory consolidation. On the other hand, the amygdala processes the emotional valence of an event. Disrupting activity in the basolateral amygdala (BLA) immediately after training impairs performance on emotional tasks like contextual fear conditioning. However, how the amygdala and the hippocampus interact to consolidate an emotional event is yet unknown. To study that, the group designed a new task where the rats learn the location of an aversive stimulus. Using large scale simultaneous neuronal ensemble recordings in the hippocampus and BLA, we found coordinated reactivations between the two structures during sleep following training. Moreover, these reactivations specifically involve a small subset of BLA neurons that are modulated during hippocampal ripples and this modulation increases during sleep ripples following training. Hippocampal ripples during sleep thus emerge as a crucial time windows for intra-hippocampus and cross-structure reactivations sustaining the consolidation of spatial and emotional memories.

Dr Gabrielle Girardeau, NYU Langone Medical Center, USA

Dr Gabrielle Girardeau, NYU Langone Medical Center, USA

Gabrielle Girardeau is a post-doctoral fellow in Dr Buzsaki's lab at the NYU Langone Medical Center in New York, USA. Her research focuses on the network mechanisms of memory consolidation, with a specific emphasis on sleep-dependent processes. In her current work, she is dissecting neuronal interactions between the amygdala and the hippocampus to better understand the emotional component of memory consolidation. She uses large-scale in-vivo recordings of neural ensembles in freely moving rodents to study how intra and cross-structural coordinated neural activity sustains the long-term stabilization of memories. Dr Girardeau is currently a Charles H. Revson Foundation Biomedical Fellow.

Dr Girardeau received her PhD from the College de France/University Pierre et Marie Curie, (Paris, France), where she contributed to unravel a key mechanism of hippocampus-dependent memory consolidation during slow-wave sleep. For this work, she was awarded the Best Thesis Award from the French Society for Neuroscience, The Major Advances in Biology Award from the French Academy of Science and a MaxPlanck/CNRS Neuroscience Award. Dr Girardeau is a member of the Society for Neuroscience (SfN) and the French Society for Neuroscience.

Due to their opposite behavioural outcomes, memory reconsolidation and extinction have usually been considered as separate entities, despite being induced by similar reactivation protocols. Based on modelling and pharmacological data, the group will argue that the two processes nevertheless seem to involve similar sets of plasticity mechanisms. One of these sets seems to be pharmacologically similar to the one involved in initial consolidation and Hebbian plasticity, while the second appears to be more involved with the labilization of existing memories and synaptic changes. Moreover, memory labilization mechanisms resemble those involved in forms of homeostatic plasticity, such as synaptic scaling. With this in mind, the group will discuss whether memory destabilization during reactivation might be a consequence of homeostatic-like plasticity induction, and why mismatch between learning and re-exposure could lead this to happen. On this basis, the group will argue that the field of memory updating might benefit from a paradigm shift in which reconsolidation and extinction, as well as online and offline reactivation, are viewed not as fully distinct processes but as different instantiations of common plasticity systems.

Dr Olavo Amaral, Federal University of Rio de Janeiro, Brazil

Dr Olavo Amaral, Federal University of Rio de Janeiro, Brazil

Olavo Amaral is a professor at the Federal University of Rio de Janeiro, Brazil, holding a medical degree and a PhD in Biochemistry from the Federal University of Rio Grande do Sul. As a memory researcher, he is particularly interested in understanding memory labilization processes occurring in reconsolidation, extinction and forgetting. For this means, he works with behavioural studies in rodents and with simple computational models of neural networks, trying to put together findings at the molecular and circuit levels. He is also interested in initiatives to improve the reproducibility of behavioural neuroscience studies, particularly through the use of systematic reviews and meta-analysis. Finally, he is interested in the relationships between neuroscience and psychiatric nosology, and currently works on a non-fiction book dealing with the shifting frontiers between health and illness in contemporary society.

Neurogenesis persists throughout life in the hippocampus, and there is a lot of interest in how the continuous addition of new neurons impacts hippocampal memory function. Behavioural studies have shown that artificially elevating hippocampal neurogenesis often facilitates new memory formation. However, since the integration of new neurons remodels existing hippocampal circuits, it has been hypothesized that hippocampal neurogenesis may also promote the degradation (or forgetting) of memories already stored in those circuits. Consistent with this idea, we have recently discovered that elevating rates of hippocampal neurogenesis after memory formation leads to forgetting. This finding changes the way we think about how hippocampal neurogenesis contributes to memory function, suggesting that it regulates a balance between encoding new memories and clearing out old memories.

Professor Paul Frankland, Program in Neuroscience and Mental Health, Hospital for Sick children, Toronto, Canada

Professor Paul Frankland, Program in Neuroscience and Mental Health, Hospital for Sick children, Toronto, Canada

Paul Frankland is a Senior Scientist in the program in Neurosciences & Mental Health at the Hospital for Sick Children. He holds a Canada Research Chair in Cognitive Neurobiology, and is appointed as an Associate Professor in the Department of Psychology, Department of Physiology and Institute of Medical Science at the University of Toronto. His research focuses on modelling cognitive function and dysfunction in genetically-engineered mice. In these studies he hopes to characterize the roles of different proteins in neuronal plasticity, how different brain regions contribute to distinct cognitive processes, and how these are altered in disease states.

For years it was believed that after fear memory has been established, the memory trace is engraved into the physical structure of the brain. But the burgeoning field of neuroscience has convincingly overturned this view by demonstrating that once (fear) memories are retrieved, they may enter a labile, protein synthesis dependent state, rendering it amenable to change. This process of memory reconsolidation may provide a window of opportunity to target emotional memory in patients suffering from anxiety disorders and other related disorders. In this talk a series of experimental studies in humans will be presented demonstrating that disrupting the process of memory reconsolidation by a noradrenergic β-blocker weakens or even eliminates the expression of fear memory. Although the process of memory reconsolidation has the potential to actually neutralize excessive fear memory, we cannot observe the molecular and cellular processes in the human brain underlying either the presence or absence of memory expression. The group can only infer the underlying neurobiological processes from the behavioural, cognitive, physiological or neural read-outs of fear memory. Because there is no one-to-one relationship between these memory expressions and the underlying neurobiological processes, erasure of memory traces cannot be proven. However, by critically testing hypotheses that follow from the reconsolidation conjecture, we may gradually unveil the conditions that govern the transition from simply retrieval, to reconsolidation, to extinction learning.

Professor Merel Kindt, University of Amsterdam, The Netherlands

Professor Merel Kindt, University of Amsterdam, The Netherlands

Merel Kindt (1967) is full Professor of Experimental Clinical Psychology at the University of Amsterdam. It is her mission to understand mechanisms of change, and to build on fundamental brain and cognition science in order to shape clinical interventions. In 2008, she started an innovative research programme on the plasticity of fear memory, which challenged the prevailing cognitive behavioural hypothesis that emotional memories are fixated in the physical architecture of the brain. She has used and developed a vast array of experimental methods to study the formation, consolidation and plasticity of emotional memory in humans. The potential of her research program lies in the bidirectional translational approach. It builds on fundamental insights from animal and human neuroscience literature, as well as clinical observations. These insights and observations provide operational tools to test novel hypotheses at different levels of analysis.

Spatial learning requires hippocampal NMDA receptors activity. Although studies have shown that rats treated with AP5, an NMDA receptor antagonist, do not learn a spatial task, few reports described the role of these receptors on spatial memory reconsolidation. This study explored the idea that NMDAR participate on spatial memory reconsolidation when animals retrieve the trace under circumstances in which new encoding is likely to occur. Using the Morris watermaze task, the group compared the impact AP5 in two conditions: when memory retrieval occurred in a reference memory task over several days, and after a comparable extent of training of a delayed matching-to-place task in which new memory encoding was required each day. Sensitivity to intra-hippocampal AP5 was observed only in the protocol involving new memory encoding at the time of retrieval. The group’s results suggest the existence of a reconsolidation process dependent of NMDA receptors that operates to update retrieved memories.

Dr Janine I. Rossato, Federal University of Rio Grande do Norte, Brazil

Dr Janine I. Rossato, Federal University of Rio Grande do Norte, Brazil

Janine Rossato, Ph.D. is an Associate Professor in the Department of Physiology at Federal University of Rio Grande do Norte. Her work is centred on memory reconsolidation studies with special attention to how previous knowledge can modulates the occurrence of reconsolidation.

Dr Rossato received in 2008 the L’Oreal Prize for young women in science. Dr Rossato received her PhD in Biochemistry with emphasis in Neuroscience from the Federal University of Rio Grande do Sul, Brazil (2006) under the mentorship of Dr Martín Cammarota. In 2015, Dr Rossato completed a Postdoctoral Research in the Centre for Cognitive neural Systems at The University of Edinburgh. Dr Rossato has 2535 citations and her h-index is 26.

Synaptic consolidation is a critical process in stabilising the neural representation of episodic experience in the human brain.  Synaptic consolidation is not perfectly effective, however, and not all plasticity triggered by experience results in lasting, durable memories.  Here the group examined the intriguing possibility that the probability of consolidation may depend, in part, on high-level mental processes that systemically alter the environment within which consolidation takes place.  Prior work in humans has shown that intentionally suppressing memory retrieval (retrieval stopping) reduces hippocampal activity via control mechanisms mediated by the lateral prefrontal cortex. In the current work, the group found that when people suppress retrieval given a reminder of an unwanted memory, they are considerably more likely to forget unrelated experiences from periods surrounding suppression. This amnesic shadow follows a dose-response function, becomes more pronounced after practice suppressing retrieval, exhibits characteristics indicating disturbed hippocampal function, and is predicted by reduced hippocampal activity. These findings indicate that stopping retrieval engages a suppression mechanism that broadly compromises hippocampal processes and that hippocampal stabilization processes can be interrupted strategically.  This process may operate via a top-down increase in tonic GABAergic inhibition in the hippocampus that compromises plasticity.

Professor Michael C. Anderson, MRC Cognition and Brain Sciences Unit, University of Cambridge, UK

Professor Michael C. Anderson, MRC Cognition and Brain Sciences Unit, University of Cambridge, UK

Michael Anderson received his PhD in cognitive psychology from the University of California, Los Angeles. After completing a post-doctoral fellowship in cognitive neuroscience at the University of California, Berkeley, he joined the psychology faculty at the University of Oregon, where he was director of the Memory Control Laboratory through 2007. Anderson is now Professor of Cognitive Neuroscience at the MRC Cognition and Brain Sciences Unit at the University of Cambridge. His research investigates the roles of inhibitory control processes as a cause of forgetting in long-term memory. Anderson’s recent work has focused on inhibitory control as a model of adaptive forgetting, with an emphasis on control mechanisms by which memory traces of unwelcome past experiences can be disrupted. This work begins to specify the mechanisms by which people adapt the functioning of their memories in the aftermath of traumatic experience.

Sleep is thought to play an important role in the consolidation of hippocampal-dependent forms of memory. Much less, however, is known about how sleep influences the consolidation of experience-dependent plasticity in the cerebral cortex. The group have shown that sleep enhances synaptic plasticity in the visual cortex in a manner consistent with consolidation, as has been described in the hippocampus. The underlying mechanisms are also shared with those known to be important in memory consolidation. These include the activation of several protein kinases and protein synthesis. Collectively, the groups findings indicate that waking experience creates temporary, labile synaptic changes in the cortex that are then made more permanent during sleep.

Professor Marcos Frank, Department of Biomedical Sciences, Elson S. Floyd College of Medicine, Washington State University

Professor Marcos Frank, Department of Biomedical Sciences, Elson S. Floyd College of Medicine, Washington State University

Marcos Frank received his PhD in Neuroscience at Stanford University and conducted post-doctoral research in the Department of Physiology at the University of California, San Francisco. He was then an Assistant and Associate Professor at the University of Pennsylvania) from 2002-2013. In 2014, Dr Frank joined WSU (now the Elson S. Floyd College of Medicine) as a Full Professor and is currently the Chair of the Department of Biomedical Sciences. Among other discoveries, the Frank lab has shown that sleep consolidates experience-dependent cortical plasticity during critical periods of development. The underlying mechanisms are similar to those that mediate consolidation in the hippocampus. The Frank lab has also shown that similar sleep-dependent processes occur in the adult brain. Work along these lines has appeared in the Journals Neuron, ScienceAdvances, Cerebral Cortex, Current Biology, Proceedings of the National Academy of Sciences, Journal of Neuroscience and SLEEP. Dr Frank is also the author of numerous review articles and book chapters on the topics of sleep function and brain plasticity.