Anteromedial thalamus gates the selection and stabilization of long-term memories.

Memories initially formed in hippocampus gradually stabilize to cortex over weeks-to-months for long-term storage. The mechanistic details of this brain re-organization remain poorly understood. We recorded bulk neural activity in circuits that link hippocampus and cortex as mice performed a memory-guided virtual-reality task over weeks. We identified a prominent and sustained neural correlate of memory in anterior thalamus, whose inhibition substantially disrupted memory consolidation. More strikingly, gain amplification enhanced consolidation of otherwise unconsolidated memories. To gain mechanistic insights, we developed a technology for simultaneous cellular-resolution imaging of hippocampus, thalamus, and cortex throughout consolidation. We found that whereas hippocampus equally encodes multiple memories, the anteromedial thalamus preferentially encodes salient memories, and gradually increases correlations with cortex to facilitate tuning and synchronization of cortical ensembles. We thus identify a thalamo-cortical circuit that gates memory consolidation and propose a mechanism suitable for the selection and stabilization of hippocampal memories into longer-term cortical storage.

Regulated Intron Removal Integrates Motivational State and Experience.

Myriad experiences produce transient memory, yet, contingent on the internal state of the organism and the saliency of the experience, only some memories persist over time. How experience and internal state influence the duration of memory at the molecular level remains unknown. A self-assembled aggregated state of Drosophila Orb2A protein is required specifically for long-lasting memory. We report that in the adult fly brain the mRNA encoding Orb2A protein exists in an unspliced non-protein-coding form. The convergence of experience and internal drive transiently increases the spliced protein-coding Orb2A mRNA. A screen identified pasilla, the fly ortholog of mammalian Nova-1/2, as a mediator of Orb2A mRNA processing. A single-nucleotide substitution in the intronic region that reduces Pasilla binding and intron removal selectively impairs long-term memory. We posit that pasilla-mediated processing of unspliced Orb2A mRNA integrates experience and internal state to control Orb2A protein abundance and long-term memory formation.

Dopamine Neurons Encoding Long-Term Memory of Object Value for Habitual Behavior.

Dopamine neurons promote learning by processing recent changes in reward values, such that reward may be maximized. However, such a flexible signal is not suitable for habitual behaviors that are sustained regardless of recent changes in reward outcome. We discovered a type of dopamine neuron in the monkey substantia nigra pars compacta (SNc) that retains past learned reward values stably. After reward values of visual objects are learned, these neurons continue to respond differentially to the objects, even when reward is not expected. Responses are strengthened by repeated learning and are evoked upon presentation of the objects long after learning is completed. These "sustain-type" dopamine neurons are confined to the caudal-lateral SNc and project to the caudate tail, which encodes long-term value memories of visual objects and guides gaze automatically to stably valued objects. This population of dopamine neurons thus selectively promotes learning and retention of habitual behavior.

Amyloidogenic Oligomerization Transforms Drosophila Orb2 from a Translation Repressor to an Activator.

Memories are thought to be formed in response to transient experiences, in part through changes in local protein synthesis at synapses. In Drosophila, the amyloidogenic (prion-like) state of the RNA binding protein Orb2 has been implicated in long-term memory, but how conformational conversion of Orb2 promotes memory formation is unclear. Combining in vitro and in vivo studies, we find that the monomeric form of Orb2 represses translation and removes mRNA poly(A) tails, while the oligomeric form enhances translation and elongates the poly(A) tails and imparts its translational state to the monomer. The CG13928 protein, which binds only to monomeric Orb2, promotes deadenylation, whereas the putative poly(A) binding protein CG4612 promotes oligomeric Orb2-dependent translation. Our data support a model in which monomeric Orb2 keeps target mRNA in a translationally dormant state and experience-dependent conversion to the amyloidogenic state activates translation, resulting in persistent alteration of synaptic activity and stabilization of memory.

Temporal multiplexing of perception and memory codes in IT cortex.

A central assumption of neuroscience is that long-term memories are represented by the same brain areas that encode sensory stimuli1. Neurons in inferotemporal (IT) cortex represent the sensory percept of visual objects using a distributed axis code2-4. Whether and how the same IT neural population represents the long-term memory of visual objects remains unclear. Here we examined how familiar faces are encoded in the IT anterior medial face patch (AM), perirhinal face patch (PR) and temporal pole face patch (TP). In AM and PR we observed that the encoding axis for familiar faces is rotated relative to that for unfamiliar faces at long latency; in TP this memory-related rotation was much weaker. Contrary to previous claims, the relative response magnitude to familiar versus unfamiliar faces was not a stable indicator of familiarity in any patch5-11. The mechanism underlying the memory-related axis change is likely intrinsic to IT cortex, because inactivation of PR did not affect axis change dynamics in AM. Overall, our results suggest that memories of familiar faces are represented in AM and perirhinal cortex by a distinct long-latency code, explaining how the same cell population can encode both the percept and memory of faces.

Spatial transcriptomics reveal neuron-astrocyte synergy in long-term memory.

Memory encodes past experiences, thereby enabling future plans. The basolateral amygdala is a centre of salience networks that underlie emotional experiences and thus has a key role in long-term fear memory formation1. Here we used spatial and single-cell transcriptomics to illuminate the cellular and molecular architecture of the role of the basolateral amygdala in long-term memory. We identified transcriptional signatures in subpopulations of neurons and astrocytes that were memory-specific and persisted for weeks. These transcriptional signatures implicate neuropeptide and BDNF signalling, MAPK and CREB activation, ubiquitination pathways, and synaptic connectivity as key components of long-term memory. Notably, upon long-term memory formation, a neuronal subpopulation defined by increased Penk and decreased Tac expression constituted the most prominent component of the memory engram of the basolateral amygdala. These transcriptional changes were observed both with single-cell RNA sequencing and with single-molecule spatial transcriptomics in intact slices, thereby providing a rich spatial map of a memory engram. The spatial data enabled us to determine that this neuronal subpopulation interacts with adjacent astrocytes, and functional experiments show that neurons require interactions with astrocytes to encode long-term memory.

Epigenetic priming of memory updating during reconsolidation to attenuate remote fear memories.

Traumatic events generate some of the most enduring forms of memories. Despite the elevated lifetime prevalence of anxiety disorders, effective strategies to attenuate long-term traumatic memories are scarce. The most efficacious treatments to diminish recent (i.e., day-old) traumata capitalize on memory updating mechanisms during reconsolidation that are initiated upon memory recall. Here, we show that, in mice, successful reconsolidation-updating paradigms for recent memories fail to attenuate remote (i.e., month-old) ones. We find that, whereas recent memory recall induces a limited period of hippocampal neuroplasticity mediated, in part, by S-nitrosylation of HDAC2 and histone acetylation, such plasticity is absent for remote memories. However, by using an HDAC2-targeting inhibitor (HDACi) during reconsolidation, even remote memories can be persistently attenuated. This intervention epigenetically primes the expression of neuroplasticity-related genes, which is accompanied by higher metabolic, synaptic, and structural plasticity. Thus, applying HDACis during memory reconsolidation might constitute a treatment option for remote traumata.

Cortical-hippocampal coupling during manifold exploration in motor cortex.

Systems consolidation-a process for long-term memory stabilization-has been hypothesized to occur in two stages1-4. Whereas new memories require the hippocampus5-9, they become integrated into cortical networks over time10-12, making them independent of the hippocampus. How hippocampal-cortical dialogue precisely evolves during this and how cortical representations change in concert is unknown. Here, we use a skill learning task13,14 to monitor the dynamics of cross-area coupling during non-rapid eye movement sleep along with changes in primary motor cortex (M1) representational stability. Our results indicate that precise cross-area coupling between hippocampus, prefrontal cortex and M1 can demarcate two distinct stages of processing. We specifically find that each animal demonstrates a sharp increase in prefrontal cortex and M1 sleep slow oscillation coupling with stabilization of performance. This sharp increase then predicts a drop in hippocampal sharp-wave ripple (SWR)-M1 slow oscillation coupling-suggesting feedback to inform hippocampal disengagement and transition to a second stage. Notably, the first stage shows significant increases in hippocampal SWR-M1 slow oscillation coupling in the post-training sleep and is closely associated with rapid learning and variability of the M1 low-dimensional manifold. Strikingly, even after consolidation, inducing new manifold exploration by changing task parameters re-engages hippocampal-M1 coupling. We thus find evidence for dynamic hippocampal-cortical dialogue associated with manifold exploration during learning and adaptation.

The DDHD2-STXBP1 interaction mediates long-term memory via generation of saturated free fatty acids.

The phospholipid and free fatty acid (FFA) composition of neuronal membranes plays a crucial role in learning and memory, but the mechanisms through which neuronal activity affects the brain's lipid landscape remain largely unexplored. The levels of saturated FFAs, particularly of myristic acid (C14:0), strongly increase during neuronal stimulation and memory acquisition, suggesting the involvement of phospholipase A1 (PLA1) activity in synaptic plasticity. Here, we show that genetic ablation of the PLA1 isoform DDHD2 in mice dramatically reduces saturated FFA responses to memory acquisition across the brain. Furthermore, DDHD2 loss also decreases memory performance in reward-based learning and spatial memory models prior to the development of neuromuscular deficits that mirror human spastic paraplegia. Via pulldown-mass spectrometry analyses, we find that DDHD2 binds to the key synaptic protein STXBP1. Using STXBP1/2 knockout neurosecretory cells and a haploinsufficient STXBP1+/- mouse model of human early infantile encephalopathy associated with intellectual disability and motor dysfunction, we show that STXBP1 controls targeting of DDHD2 to the plasma membrane and generation of saturated FFAs in the brain. These findings suggest key roles for DDHD2 and STXBP1 in lipid metabolism and in the processes of synaptic plasticity, learning, and memory.

Impairments to Consolidation, Reconsolidation, and Long-Term Memory Maintenance Lead to Memory Erasure.

An enduring problem in neuroscience is determining whether cases of amnesia result from eradication of the memory trace (storage impairment) or if the trace is present but inaccessible (retrieval impairment). The most direct approach to resolving this question is to quantify changes in the brain mechanisms of long-term memory (BM-LTM). This approach argues that if the amnesia is due to a retrieval failure, BM-LTM should remain at levels comparable to trained, unimpaired animals. Conversely, if memories are erased, BM-LTM should be reduced to resemble untrained levels. Here we review the use of BM-LTM in a number of studies that induced amnesia by targeting memory maintenance or reconsolidation. The literature strongly suggests that such amnesia is due to storage rather than retrieval impairments. We also describe the shortcomings of the purely behavioral protocol that purports to show recovery from amnesia as a method of understanding the nature of amnesia.

Mrj is a chaperone of the Hsp40 family that regulates Orb2 oligomerization and long-term memory in Drosophila.

Orb2 the Drosophila homolog of cytoplasmic polyadenylation element binding (CPEB) protein forms prion-like oligomers. These oligomers consist of Orb2A and Orb2B isoforms and their formation is dependent on the oligomerization of the Orb2A isoform. Drosophila with a mutation diminishing Orb2A's prion-like oligomerization forms long-term memory but fails to maintain it over time. Since this prion-like oligomerization of Orb2A plays a crucial role in the maintenance of memory, here, we aim to find what regulates this oligomerization. In an immunoprecipitation-based screen, we identify interactors of Orb2A in the Hsp40 and Hsp70 families of proteins. Among these, we find an Hsp40 family protein Mrj as a regulator of the conversion of Orb2A to its prion-like form. Mrj interacts with Hsp70 proteins and acts as a chaperone by interfering with the aggregation of pathogenic Huntingtin. Unlike its mammalian homolog, we find Drosophila Mrj is neither an essential gene nor causes any gross neurodevelopmental defect. We observe a loss of Mrj results in a reduction in Orb2 oligomers. Further, Mrj knockout exhibits a deficit in long-term memory and our observations suggest Mrj is needed in mushroom body neurons for the regulation of long-term memory. Our work implicates a chaperone Mrj in mechanisms of memory regulation through controlling the oligomerization of Orb2A and its association with the translating ribosomes.

Climate-driven flyway changes and memory-based long-distance migration.

Millions of migratory birds occupy seasonally favourable breeding grounds in the Arctic1, but we know little about the formation, maintenance and future of the migration routes of Arctic birds and the genetic determinants of migratory distance. Here we established a continental-scale migration system that used satellite tracking to follow 56 peregrine falcons (Falco peregrinus) from 6 populations that breed in the Eurasian Arctic, and resequenced 35 genomes from 4 of these populations. The breeding populations used five migration routes across Eurasia, which were probably formed by longitudinal and latitudinal shifts in their breeding grounds during the transition from the Last Glacial Maximum to the Holocene epoch. Contemporary environmental divergence between the routes appears to maintain their distinctiveness. We found that the gene ADCY8 is associated with population-level differences in migratory distance. We investigated the regulatory mechanism of this gene, and found that long-term memory was the most likely selective agent for divergence in ADCY8 among the peregrine populations. Global warming is predicted to influence migration strategies and diminish the breeding ranges of peregrine populations of the Eurasian Arctic. Harnessing ecological interactions and evolutionary processes to study climate-driven changes in migration can facilitate the conservation of migratory birds.

Dopamine-based mechanism for transient forgetting.

Active forgetting is an essential component of the memory management system of the brain1. Forgetting can be permanent, in which prior memory is lost completely, or transient, in which memory exists in a temporary state of impaired retrieval. Temporary blocks on memory seem to be universal, and can disrupt an individual's plans, social interactions and ability to make rapid, flexible and appropriate choices. However, the neurobiological mechanisms that cause transient forgetting are unknown. Here we identify a single dopamine neuron in Drosophila that mediates the memory suppression that results in transient forgetting. Artificially activating this neuron did not abolish the expression of long-term memory. Instead, it briefly suppressed memory retrieval, with the memory becoming accessible again over time. The dopamine neuron modulates memory retrieval by stimulating a unique dopamine receptor that is expressed in a restricted physical compartment of the axons of mushroom body neurons. This mechanism for transient forgetting is triggered by the presentation of interfering stimuli immediately before retrieval.

The effects of mnemonic variability and spacing on memory over multiple timescales.

The memory benefit that arises from distributing learning over time rather than in consecutive sessions is one of the most robust effects in cognitive psychology. While prior work has mainly focused on repeated exposures to the same information, in the real world, mnemonic content is dynamic, with some pieces of information staying stable while others vary. Thus, open questions remain about the efficacy of the spacing effect in the face of variability in the mnemonic content. Here, in two experiments, we investigated the contributions of mnemonic variability and the timescale of spacing intervals, ranging from seconds to days, to long-term memory. For item memory, both mnemonic variability and spacing intervals were beneficial for memory; however, mnemonic variability was greater at shorter spacing intervals. In contrast, for associative memory, repetition rather than mnemonic variability was beneficial for memory, and spacing benefits only emerged in the absence of mnemonic variability. These results highlight a critical role for mnemonic variability and the timescale of spacing intervals in the spacing effect, bringing this classic memory paradigm into more ecologically valid contexts.

Parallel vector memories in the brain of a bee as foundation for flexible navigation.

Insects rely on path integration (vector-based navigation) and landmark guidance to perform sophisticated navigational feats, rivaling those seen in mammals. Bees in particular exhibit complex navigation behaviors including creating optimal routes and novel shortcuts between locations, an ability historically indicative of the presence of a cognitive map. A mammalian cognitive map has been widely accepted. However, in insects, the existence of a centralized cognitive map is highly contentious. Using a controlled laboratory assay that condenses foraging behaviors to short distances in walking bumblebees, we reveal that vectors learned during path integration can be transferred to long-term memory, that multiple such vectors can be stored in parallel, and that these vectors can be recalled at a familiar location and used for homeward navigation. These findings demonstrate that bees meet the two fundamental requirements of a vector-based analog of a decentralized cognitive map: Home vectors need to be stored in long-term memory and need to be recalled from remembered locations. Thus, our data demonstrate that bees possess the foundational elements for a vector-based map. By utilizing this relatively simple strategy for spatial organization, insects may achieve high-level navigation behaviors seen in vertebrates with the limited number of neurons in their brains, circumventing the computational requirements associated with the cognitive maps of mammals.

A double dissociation between savings and long-term memory in motor learning.

Memories are easier to relearn than learn from scratch. This advantage, known as savings, has been widely assumed to result from the reemergence of stable long-term memories. In fact, the presence of savings has often been used as a marker for whether a memory has been consolidated. However, recent findings have demonstrated that motor learning rates can be systematically controlled, providing a mechanistic alternative to the reemergence of a stable long-term memory. Moreover, recent work has reported conflicting results about whether implicit contributions to savings in motor learning are present, absent, or inverted, suggesting a limited understanding of the underlying mechanisms. To elucidate these mechanisms, we investigate the relationship between savings and long-term memory by experimentally dissecting the underlying memories based on short-term (60-s) temporal persistence. Components of motor memory that are temporally-persistent at 60 s might go on to contribute to stable, consolidated long-term memory, whereas temporally-volatile components that have already decayed away by 60 s cannot. Surprisingly, we find that temporally-volatile implicit learning leads to savings, whereas temporally-persistent learning does not, but that temporally-persistent learning leads to long-term memory at 24 h, whereas temporally-volatile learning does not. This double dissociation between the mechanisms for savings and long-term memory formation challenges widespread assumptions about the connection between savings and memory consolidation. Moreover, we find that temporally-persistent implicit learning not only fails to contribute to savings, but also that it produces an opposite, anti-savings effect, and that the interplay between this temporally-persistent anti-savings and temporally-volatile savings provides an explanation for several seemingly conflicting recent reports about whether implicit contributions to savings are present, absent, or inverted. Finally, the learning curves we observed for the acquisition of temporally-volatile and temporally-persistent implicit memories demonstrate the coexistence of implicit memories with distinct time courses, challenging the assertion that models of context-based learning and estimation should supplant models of adaptive processes with different learning rates. Together, these findings provide new insight into the mechanisms for savings and long-term memory formation.

Item-specific neural representations during human sleep support long-term memory.

Understanding how individual memories are reactivated during sleep is essential in theorizing memory consolidation. Here, we employed the targeted memory reactivation (TMR) paradigm to unobtrusively replaying auditory memory cues during human participants' slow-wave sleep (SWS). Using representational similarity analysis (RSA) on cue-elicited electroencephalogram (EEG), we found temporally segregated and functionally distinct item-specific neural representations: the early post-cue EEG activity (within 0 to 2,000 ms) contained comparable item-specific representations for memory cues and control cues, signifying effective processing of auditory cues. Critically, the later EEG activity (2,500 to 2,960 ms) showed greater item-specific representations for post-sleep remembered items than for forgotten and control cues, indicating memory reprocessing. Moreover, these later item-specific neural representations were supported by concurrently increased spindles, particularly for items that had not been tested prior to sleep. These findings elucidated how external memory cues triggered item-specific neural representations during SWS and how such representations were linked to successful long-term memory. These results will benefit future research aiming to perturb specific memory episodes during sleep.

Astrocyte-neuron lactate transport is required for long-term memory formation.

We report that, in the rat hippocampus, learning leads to a significant increase in extracellular lactate levels that derive from glycogen, an energy reserve selectively localized in astrocytes. Astrocytic glycogen breakdown and lactate release are essential for long-term but not short-term memory formation, and for the maintenance of long-term potentiation (LTP) of synaptic strength elicited in vivo. Disrupting the expression of the astrocytic lactate transporters monocarboxylate transporter 4 (MCT4) or MCT1 causes amnesia, which, like LTP impairment, is rescued by L-lactate but not equicaloric glucose. Disrupting the expression of the neuronal lactate transporter MCT2 also leads to amnesia that is unaffected by either L-lactate or glucose, suggesting that lactate import into neurons is necessary for long-term memory. Glycogenolysis and astrocytic lactate transporters are also critical for the induction of molecular changes required for memory formation, including the induction of phospho-CREB, Arc, and phospho-cofilin. We conclude that astrocyte-neuron lactate transport is required for long-term memory formation.

Dynamics of retrieval strategies for remote memories.

Prevailing theory suggests that long-term memories are encoded via a two-phase process requiring early involvement of the hippocampus followed by the neocortex. Contextual fear memories in rodents rely on the hippocampus immediately following training but are unaffected by hippocampal lesions or pharmacological inhibition weeks later. With fast optogenetic methods, we examine the real-time contribution of hippocampal CA1 excitatory neurons to remote memory and find that contextual fear memory recall, even weeks after training, can be reversibly abolished by temporally precise optogenetic inhibition of CA1. When this inhibition is extended to match the typical time course of pharmacological inhibition, remote hippocampus dependence converts to hippocampus independence, suggesting that long-term memory retrieval normally depends on the hippocampus but can adaptively shift to alternate structures. Further revealing the plasticity of mechanisms required for memory recall, we confirm the remote-timescale importance of the anterior cingulate cortex (ACC) and implicate CA1 in ACC recruitment for remote recall.

eIF2α controls memory consolidation via excitatory and somatostatin neurons.

An important tenet of learning and memory is the notion of a molecular switch that promotes the formation of long-term memory1-4. The regulation of proteostasis is a critical and rate-limiting step in the consolidation of new memories5-10. One of the most effective and prevalent ways to enhance memory is by regulating the synthesis of proteins controlled by the translation initiation factor eIF211. Phosphorylation of the α-subunit of eIF2 (p-eIF2α), the central component of the integrated stress response (ISR), impairs long-term memory formation in rodents and birds11-13. By contrast, inhibiting the ISR by mutating the eIF2α phosphorylation site, genetically11 and pharmacologically inhibiting the ISR kinases14-17, or mimicking reduced p-eIF2α with the ISR inhibitor ISRIB11, enhances long-term memory in health and disease18. Here we used molecular genetics to dissect the neuronal circuits by which the ISR gates cognitive processing. We found that learning reduces eIF2α phosphorylation in hippocampal excitatory neurons and a subset of hippocampal inhibitory neurons (those that express somatostatin, but not parvalbumin). Moreover, ablation of p-eIF2α in either excitatory or somatostatin-expressing (but not parvalbumin-expressing) inhibitory neurons increased general mRNA translation, bolstered synaptic plasticity and enhanced long-term memory. Thus, eIF2α-dependent mRNA translation controls memory consolidation via autonomous mechanisms in excitatory and somatostatin-expressing inhibitory neurons.