Offline Memory Reactivation Promotes the Consolidation Of Spatially Unbiased Long-Term Cognitive Maps

Abstract

Spatial memories which can last a lifetime are thought to be encoded during 'online' periods of exploration and subsequently consolidated into stable cognitive maps through their 'offline' reactivation. However, the mechanisms and computational principles by which offline reactivation stabilize long-lasting spatial representations remain poorly understood. Here we employed simultaneous fast calcium imaging and electrophysiology to track hippocampal place cells over weeks of online spatial reward learning behavior and offline resting. We describe that recruitment to persistent network-level offline reactivation of spatial experiences predicts cells' future multi-day representational stability. Moreover, while representations of reward-adjacent locations are generally more stable across days, reactivation-related stabilization is, conversely, most prominent for reward-distal locations. Thus, while occurring on millisecond time-scales, offline reactivation counter-balances the observed multi-day representational reward-adjacency bias, promoting the stabilization of cognitive maps which comprehensively reflect entire underlying spatial contexts. These findings suggest that post-learning offline-related memory consolidation plays complimentary and computationally distinct role in learning as compared to online encoding.

CA2 inhibition reduces the precision of hippocampal assembly reactivation

“…To examine how these changes in firing altered interactions between neurons, we next calculated pairwise correlations of the normalized peri-SWRs firing rate curve of each neuron for each session in the two groups. Following spatial experience a more structured and clustered pattern emerged in the controls during the Post-CNO compared to the Pre-CNO sessions, consistent with previous data finding experience-dependent changes in coordinated activity(Grosmark and Buzsaki, 2016; Grosmark, 2020; Wilson and McNaughton, 1994) ( Fig. 2c; Extended Data Fig.…”

“…We first examined SWRs in each CA region independently (intra-region), comparing how their properties changed between the first rest session and the last rest session of the day. It has been suggested that spatial experience can lead to increases in SWR power, length and occurrence in subsequent rest sessions (Eschenko et al, 2008; Grosmark and Buzsaki, 2016; Grosmark, 2020; Ramadan et al, 2009); consistent with this, the normalized change in CA1 ripple-band power between the Pre- and Post-CNO sessions increased in control mice, but not in the DREADD mice. However, SWR frequency, duration, and occurrence were similar in both groups across the sessions ( Fig.…”

Parallel processing of sensory cue and spatial information in the Dentate Gyrus

“…To calculate a measure for spatial information content for granule cells in Fig. 1g, we adapted a traditional method of spatial information assessment 23,38,57 to Ca 2+ imaging data. For each cell, we used the firing rate-by-position vector and shuffled null distribution computed above and calculated the spatial information content for each as described previously 38,57 .…”

“…1g, we adapted a traditional method of spatial information assessment 23,38,57 to Ca 2+ imaging data. For each cell, we used the firing rate-by-position vector and shuffled null distribution computed above and calculated the spatial information content for each as described previously 38,57 . To account for the fact that low firing rates artificially produce high spatial information scores, we subtracted the mean of the shuffled information per spike from observed information per spike, divided by the standard deviation of the shuffled values to determine the spatial variance for each cell.…”

2-D Neural Geometry Underpins Hierarchical Organization of Sequence in Human Working Memory

“…Interestingly, numeours studies show that replay during retention or in spontaneous activities mediates sequence memory, structure representation, and novel inference, or in a broad sense, forms cognitive maps [59][60][61][62] . How to reconcile 1-D replay with 2-D neural geometry?…”