Sharp wave ripples during learning stabilize the hippocampal spatial map

Abstract

Cognitive representation of the environment requires a stable hippocampal map but the mechanisms maintaining map representation are unknown. Because sharp wave-ripples (SPW-R) orchestrate both retrospective and prospective spatial information, we hypothesized that disrupting neuronal activity during SPW-Rs affects spatial representation. Mice learned daily a new set of three goal locations on a multi-well maze. We used closed-loop SPW-R detection at goal locations to trigger optogenetic silencing of a subset of CA1 pyramidal neurons. Control place cells (non-silenced or silenced outside SPW-Rs) largely maintained the location of their place fields after learning and showed increased spatial information content. In contrast, the place fields of SPW-R-silenced place cells remapped, and their spatial information remained unaltered. SPW-R silencing did not impact the firing rates or the proportions of place cells. These results suggest that interference with SPW-R-associated activity during learning prevents the stabilization and refinement of the hippocampal map.

The content of hippocampal “replay”

“…While this may be partly explained by the fact that the open field environment was highly familiar in the open field study, the predictable reward location was novel every session, and it is reasonable to expect strong release of salience signals at that position (Schultz et al, 1997). Instead, these data suggest that reverse replay may be primarily utilized in the initial formation of a mental map of a novel environment (Carr et al, 2011;Roux et al, 2017), serving to synaptically couple neurons that represent adjacent locations; once formed, other processes may flexibly assign value to specific positions within that cognitive map. A prediction from this hypothesis is that replay in a novel open arena should predominantly encode the rat's prior behavioral path, even if the rat is performing a familiar, goaldirected task that was learned in a separate environment.…”

“…Importantly, there is no current evidence indicating that forward and reverse replays are generated via distinct network mechanisms; therefore, identifying how the hippocampal circuitry selects for the expression of forward versus reverse replay (Ambrose et al, 2016) may provide clues as to how the same circuitry can represent specific paths during open field replay (Pfeiffer & Foster, 2013). Sleep and wake replay also seem fundamentally different, as blocking sleep-based replays impairs long-term memory without impacting place field representation (Girardeau et al, 2009;EgoStengel & Wilson, 2010;Kovacs et al, 2016), while preventing awake replays impacts working spatial memory and place field stability without globally impacting long-term memory (Jadhav et al, 2012;Roux et al, 2017). In addition, early reports failed to identify reverse replay in sleep-based ripples (Lee & Wilson, 2002) and recent studies observe significantly more forwards-ordered than reverse-ordered replay in sleep (Wikenheiser & Redish, 2013;Grosmark & Buzs aki, 2016), raising the intriguing possibility that forward replay itself can serve two purposes depending on the behavioral state of the animal (Diekelmann et al, 2011): memory retrieval during the awake state and memory consolidation during sleep (Carr et al, 2011;Lewis & Durrant, 2011).…”

Three brain states in the hippocampus and cortex

“…On the one hand, these two patterns can be seen as particularly different—replay sequences appear more flexible than theta sequences and furthermore these two patterns occur in different brain states (SWR and locomotor, respectively). More generally, the two‐stage hypothesis and related experimental results articulate major functional differences between SWR and theta activity ([Girardeau, Benchenane, Wiener, Buzsaki, & Zugaro, ; Ego‐Stengel & Wilson, ; Kovacs et al, ; van de Ven, Trouche, McNamara, Allen, & Dupret, ; Roux, Hu, Eichler, Stark, & Buzsaki, ]; discussed and reviewed in Buzsaki ()).…”

“…In agreement with the original two‐stage hypothesis, numerous studies have shown that neural firing during sleep SWRs replays prior spatial experience (Buzsaki, ). Furthermore, experiments disrupting hippocampal activity at the time of SWRs suggests a role of SWR‐associated hippocampal neural activity in consolidation of spatial memory (Girardeau et al, ; Ego‐Stengel & Wilson, ; Nokia et al, ; van de Ven et al, ; see also (Maingret, Girardeau, Todorova, Goutierre, & Zugaro, ; Roux et al, ).…”

The HexMaze: A previous knowledge and schema task for mice

“…4A). This simple learning effect, while often not considered as previous knowledge effect, does affect session performance and thus, must be considered even in experiments which just focus on each session individually, as seen in most electrophysiological experiments (Lopes-Dos-Santos et al, 2018;Michon et al, 2019;Roux et al, 2017).…”