Engram Stability and Maturation During Systems Consolidation Underlies Remote Memory

Abstract

Remote memories play an important role in how we perceive the world and are rooted in ensembles in the CA1 and ACC, however the evolution of these components during systems consolidation has not yet been comprehensively addressed. By applying transgenic approaches for ensemble identification, CLARITY, retro-AAV and rabies virus for circuit mapping, and chemogenetics for functional interrogation, we addressed the dynamics of CA1-ACC ensembles and their connectivity as well as the contribution of astrocytes to the process. We found that the CA1 engrams remain stable between recent and remote recall, and, the inhibition of the engram for recent recall during remote recall functionally impairs memory. We also found that the new cells in the remote recall engram in the CA1 are not added randomly, but differ according to their connections: First, the anterograde CA1toACC engram cell projection grows larger. Second, in the retrograde projections, the ACC reduces input to CA1 engram cells, while input from the entorhinal cortex and paraventricular nucleus of the thalamus increases. Finally, we found that activating CA1 astrocytes during acquisition improves recent but not remote recall, and that CA1toACC projecting cells are recruited earlier when the astrocytes are stimulated. Our results shine fresh light on systems consolidation by providing a deeper understanding of engram stability and maturation in the transition from recent to remote memory.

Single field evolution rule governs the dynamics of representational drift in mouse hippocampal dorsal CA1 region

“…While synaptic plasticity undoubtedly plays a critical role in representation stability, neuronal impacts should also be considered. Research on memory engrams [35][36][37][38][39][40] suggests that stable coding may be confined to particular neuronal ensembles, potentially distinguished by genetic traits [38][39][40][41] or excitability to neuroplastic changes 34 . This raises a critical question: to what extent do dCA1 pyramidal neurons affect the stability of their place fields?…”

Divergent Recruitment of Developmentally-Defined Neuronal Ensembles Supports Memory Dynamics

“…Their properties, such as specificity, accessibility, or susceptibility to disruption, evolve after encoding, and their content can be updated upon changes in external or internal contingencies (25,(31)(32)(33)(34)(35)(36)(37)(38)(39)(40)(41)(42). Likewise, activity patterns in brain areas where memory ensembles are established upon learning, like the hippocampus, evolve over time and experience (2,14,27,(43)(44)(45)(46)(47)(48)(49). New sets of neurons are recruited into memory ensembles over multiple learning sessions, while a neuron's probability of reactivation upon successive exposures to the same experience decreases as a function of time elapsed between them (a phenomenon known as "representational drift") (43,(50)(51)(52)(53)(54)(55)(56)(57).…”

“…Similarly, the c-Fos+ neuronal ensemble that in rodents is established during a memory's acquisition overlaps with the one established during its recall, with the extent of this overlap being associated with retrieval success (9,10,14,16,(23)(24)(25)(26). Furthermore, the reactivation of the c-Fos+ ensemble established at acquisition is both necessary for memory retrieval in physiological conditions, and sufficient to induce artificial recollection of a specific memory at later times (11,13,14,24,27). Together, these data have supported a reactivation-centred model, according to which memory encoding leads to the establishment of a distributed network of memory ensembles in multiple cortical and subcortical structures, with hippocampal ensembles serving as an index for the reactivation of this memory network during recall (8,10,12,17,(28)(29)(30).…”