The effects of reactivation during sleep on the neural representations of episodic memories.

Poster Session E - Monday, April 15, 2024, 2:30 – 4:30 pm EDT, Sheraton Hall ABC

Sarvia Aquino Argueta^1,2 (ssaquino@uci.edu), Andrew Lazarus³, Laura K. Shanahan⁴, Kenneth A. Norman^5,6, Lila Davachi^7,8, Thorsten Kahnt⁹, Ken A. Paller³, Eitan Schechtman^1,2,3; ¹Department of Neurobiology and Behavior, University of California, Irvine, Irvine, CA, ²Center for the Neurobiology of Learning and Memory, University of California, Irvine, Irvine, CA, ³Department of Psychology, Northwestern University, Evanston, IL, ⁴Department of Psychology, Rhodes College, Memphis, TN, ⁵Princeton Neuroscience Institute, Princeton University, Princeton, NJ, ⁶Department of Psychology, Princeton University, Princeton, NJ, ⁷Department of Psychology, Columbia University, New York, NY, ⁸Nathan Kline Institute for Psychiatric Research, New York, NY, ⁹National Institute on Drug Abuse Intramural Research Program, Baltimore, MD

Sleep involves the reactivation of recently acquired memories, thereby shaping the neural representations supporting them. An essential feature of episodic memory is the link between a specific memory item (e.g., a chocolate cake) and the context in which it is embedded (e.g., your tenth birthday party). However, the possible role of sleep in binding memories to their contexts remains poorly understood. To investigate how reactivation during sleep impacts this binding process, we instructed participants to form stories, each linking together four objects (e.g., a gong) in unique contexts (e.g., a hike in the woods). As a measure of object-context binding, we used functional MRI to measure the overlap between neural representations for objects linked together by a story. Then, during non-REM sleep, we unobtrusively presented object-congruent sounds to selectively reactivate a subset of the object memories, a technique termed targeted memory reactivation. Sleep was followed by an additional functional MRI scan, identical to the previous one. We hypothesized that reactivating memories during sleep would either selectively strengthen object-context binding for the targeted objects, or, alternatively, keep the items segregated, prioritizing specificity. We found that targeted reactivation during sleep promoted specificity in items’ neural representation in the left ventromedial prefrontal cortex, as reflected by a decrease in neural overlap for reactivated vs. non-reactivated memories. The effect of selective reactivation counteracted a general increase in within-context similarity that we observed across pre- and post-nap functional MRI sessions. These results suggest that reactivation during sleep decontextualizes memories rather than strengthening item-context binding.

Topic Area: LONG-TERM MEMORY: Episodic