Back

The Reinstatement of a Forgotten Infantile Memory

Lahr, M.; Imhof, F.; Mauro, L.; Ulmer, T.; Donato, F.

2025-09-29 neuroscience
10.1101/2025.09.27.678956 bioRxiv
Show abstract

Infantile memories present a striking paradox: while early-life experiences are typically forgotten, reflecting the phenomenon of infantile amnesia, traumatic events from infancy can profoundly shape adult cognition and behavior. How do memories that are seemingly inaccessible persistently influence cognitive processes and behaviors throughout life? Rodent studies have demonstrated that forgotten infantile memories remain encoded as latent "infantile memory engrams" (iEngrams) within neuronal circuits, capable of memory reinstatement under artificial experimental conditions. Still, the network mechanisms underpinning the natural reinstatement of a forgotten infantile memory are unknown. Here, we show that infantile memories, though physiologically irretrievable in adults, remain stored within hippocampal circuits and their engrams contribute to hippocampal network dynamics. Crucially, reinstating these memories requires a carefully orchestrated network process. An initial contextual reminder primes the hippocampal network to increase activity of the iEngram during a subsequent aversive reminder, which tags iEngram neurons for offline reactivation. This reactivation facilitates the integration of previously latent infantile memories with novel neuronal ensembles, reinstating behavior consistent with the original memory. These findings critically advance our understanding of the neuronal mechanisms underlying physiological memory encoding, retrieval, and reinstatement across development. Furthermore, they delineate the temporal boundaries and underlying physiological substrate of the process by which a latent representation becomes associated with a novel neuronal substrate, suggesting potential interventions to prevent the maladaptive reinstatement of traumatic infantile memories.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.