2025-03-20 アメリカ国立衛生研究所(NIH)
<関連情報>
- https://www.nih.gov/news-events/news-releases/study-illuminates-structural-features-memory-formation-cellular-subcellular-levels
- https://www.science.org/doi/10.1126/science.ado8316
マウスの海馬における記憶エングラムのシナプス構造 Synaptic architecture of a memory engram in the mouse hippocampus
Marco Uytiepo, Yongchuan Zhu, Eric Bushong, Katherine Chou, […], and Anton Maximov
Science Published:21 Mar 2025
DOI:https://doi.org/10.1126/science.ado8316
Ultrastructural analysis of the physical substrates of memory engrams in the mouse hippocampus.
(A) Overview of the experimental design. (B) A saturated 3D-EM reconstruction of excitatory synapses in the stratum radiatum of the dorsal area CA1. (C) Diagram illustrating the absence of preferential wiring among projection neurons activated during associative learning in areas CA3 and CA1 alongside the expansion of axonal networks of CA3 neurons. (D) Diagram depicting the expansion of axonal connectivity in initial engram ensembles through atypical MSBs. A 3D view of an individual MSB connecting to three dendritic branches (D1 to D3) is also shown.
Editor’s summary
How are memories formed and stored in the brain? Uytiepo et al. used three-dimensional electron microscopy coupled with chemogenetic tagging and behavioral analysis to determine the structural changes supporting the acquisition of short-term fear memories, focusing on the Schaffer collateral synapses. Short-term memory acquisition was found to be associated with a selective increase in multisynaptic boutons without the involvement of simultaneous activation of synaptically connected neurons. These results challenge the general applicability of the Hebbian model and expand our understanding of the mechanisms responsible for memory formation. —Mattia Maroso
Structured Abstract
INTRODUCTION
The architectures of brain circuits reflect their capacities for information processing, storage, and retrieval. Long-term memories of new experiences are thought to form through mechanisms involving the restructuring of both neuronal wiring and individual synapses. The classic Hebbian theory posits that neurons recruited for discrete learning epochs form stable networks by selectively strengthening connections among themselves. However, the generalizability of this principle is challenged by observations suggesting that the population coding of sensory stimuli is not consistently fixed over time and that neural ensembles engaged during memory acquisition and recall only partially overlap across different brain regions.
RATIONALE
Current insights into the nature and origins of experience-dependent events that shape the physical substrates of memory engrams are largely derived from optical imaging. Despite ongoing technical advancements, light-imaging methods alone remain insufficient for the detailed dissection of complex brain tissues because of their limited resolution and inability to provide unbiased analyses of diverse structural features. Consequently, the intricate organization of circuits that store memory traces remains poorly understood. To address this knowledge gap, we used three-dimensional electron microscopy (3D-EM). By integrating 3D-EM with temporally controlled chemogenetic tagging of behaviorally relevant neurons and artificial intelligence–based tools for image analysis, we aimed to uncover the structural correlates of long-term associative memory in the mouse hippocampus.
RESULTS
We performed nanoscale reconstructions of the hippocampal CA3-CA1 pathway, where projection neurons (PNs) recruited during Pavlovian fear conditioning were irreversibly labeled in a dual Fos– and drug-inducible manner using the engineered peroxidase APEX2. PNs with a remote history of activity coinciding with associative learning displayed no strong preference for wiring with one another. Instead, these PNs reorganized their connectivity through atypical multisynaptic boutons (MSBs) without altering the number or spatial distribution of isolated nerve terminals and dendritic spines. CA3 PNs expanded their axonal networks in CA1 by increasing both the relative abundance of MSBs and their structural complexity. This expansion was driven by presynaptic excitation elicited by negative valence stimuli, but not by neutral stimuli, and occurred independently of the coactivation state of postsynaptic partners. MSB-mediated rewiring of ensembles representing initial engrams was accompanied by spatially restricted, input-specific upscaling of individual synapses, remodeling of presynaptic mitochondria, redistribution of the postsynaptic spine apparatus, and enhanced interactions with astrocytes.
CONCLUSION
Our study elucidates the physical hallmarks of long-term memory at cellular and subcellular levels. High-resolution imaging of excitatory circuits and synapses allocated for an engram provides a structural basis for the cellular flexibility of information coding, as previously observed in brain regions critical for various sensory modalities and associative learning. Our findings highlight a mechanism by which projection neurons involved in memory acquisition increase their wiring complexity while maintaining the steady-state arrangements of individual synaptic sites. This mechanism may enhance a network’s associative capacity, facilitate ensemble sharing for memory generalization, and/or improve the efficiency of temporal coding. Collectively, our results establish a framework for advancing the understanding of how sensory experience–dependent structural plasticity in brain circuits affects their computational properties.
Abstract
Memory engrams are formed through experience-dependent plasticity of neural circuits, but their detailed architectures remain unresolved. Using three-dimensional electron microscopy, we performed nanoscale reconstructions of the hippocampal CA3-CA1 pathway after chemogenetic labeling of cellular ensembles recruited during associative learning. Neurons with a remote history of activity coinciding with memory acquisition showed no strong preference for wiring with each other. Instead, their connectomes expanded through multisynaptic boutons independently of the coactivation state of postsynaptic partners. The rewiring of ensembles representing an initial engram was accompanied by input-specific, spatially restricted upscaling of individual synapses, as well as remodeling of mitochondria, smooth endoplasmic reticulum, and interactions with astrocytes. Our findings elucidate the physical hallmarks of long-term memory and offer a structural basis for the cellular flexibility of information coding.
