SummaryThe mammalian hippocampus is important for normal memory function, particularly memory for places and events. Place cells, neurons within the hippocampus that have spatial receptive fields, represent information about an animal’s position. During periods of rest, but also during active task engagement, place cells spontaneously recapitulate past trajectories. Such ‘replay’ has been proposed as a mechanism necessary for a range of neurobiological functions, including systems memory consolidation, recall and spatial working memory, navigational planning, and reinforcement learning. Focusing mainly, but not exclusively, on work conducted in rodents, we describe the methodologies used to analyse replay and review evidence for its putative roles. We identify outstanding questions as well as apparent inconsistencies in existing data, making suggestions as to how these might be resolved. In particular, we find support for the involvement of replay in disparate processes, including the maintenance of hippocampal memories and decision making. We propose that the function of replay changes dynamically according to task demands placed on an organism and its current level of arousal.

Dominant theories of hippocampal function propose that place cell representations are formed during an animal's first encounter with a novel environment and are subsequently replayed during off-line states to support consolidation and future behaviour. Here we report that viewing the delivery of food to an unvisited portion of an environment leads to off-line pre-activation of place cells sequences corresponding to that space. Such ‘preplay’ was not observed for an unrewarded but otherwise similar portion of the environment. These results suggest that a hippocampal representation of a visible, yet unexplored environment can be formed if the environment is of motivational relevance to the animal. We hypothesise such goal-biased preplay may support preparation for future experiences in novel environments.DOI: http://dx.doi.org/10.7554/eLife.06063.001

During exploration, the activity of place 1 and grid cells 2 represent self-location. Together these cells have been hypothesized to support spatial memory 1,3 and navigation 3,4 . Hippocampal replay 5,6 -the reactivation of place cell sequences during immobility and sleep -has been proposed as a mechanism for consolidation 5 and route planning 7 , yet the involvement of grid cells remains unknown. Potentially, spatially coherent place and grid cell activity may emerge during replay as the hippocampus broadcasts memory traces to the cortex 8 . To study the involvement of grid cells in replay we recorded concurrently from rodent medial entorhinal cortex layers 5/6 (MECV&VI) and hippocampus (CA1) during trackrunning and subsequent rest. We report robust coherence between place and grid cell spatial representations during hippocampal replay.A total of 43 grid cells were recorded across 11 sessions from 6 rats. Concurrently, 34-72 place cells were recorded in each session, 592 in total (Fig. 1a-d, Supplementary Fig. 1,2, Supplementary Table 1). During subsequent rest, we identified putative replay events based on place cell activity (Figure 1e, see Online Methods). A Bayesian decoding algorithm 9 and a trajectory-fitting procedure was used to reconstruct position and score the replay 9 ( Supplementary Fig. 3). Robust replay events exhibiting clear, straight trajectories (each p<0.2 vs own shuffle) were used for further analyses. On average, during replay events, grid cell activity was higher than during non-replay periods 10 (2.19Hz (SD =1.74) vs 1.33Hz (SD=1.61), t(43)=3.24, p=0.00023), with peak grid cell activity lagging that of place cells by 10ms (Fig. 1f, Supplementary Fig. 4). To investigate grid-place cell spatial coherence, we assessed the similarity of the grid and place cell representations during replay events in which both were active. 2Specifically, we superimposed the trajectory derived from a hippocampal replay event onto the decoded representation from concurrently recorded grid cell spikes ( Supplementary Fig. 5). The position represented by grid cells during replay events was similar to that represented by the place cell trajectories (Fig. 2a, Supplementary Fig. 6), exceeding the coherence obtained by pairing a grid cell event with a random place cell event from the same session (p<0.0001 area under the curve (AUC) test, see Online Methods, Fig. 2b, Supplementary Fig. 5); comparisons against shuffled distributions generated by permuting grid cell ratemaps (p<0.0001 AUC) and spike times (p<0.001 AUC) corroborated this finding ( Supplementary Fig. 7,8). Importantly, grid-place coherence also exceeded chance levels when analyses was limited to just the strongest (p<0.025 vs own shuffle) place cell replay events (p<0.0001 AUC, Supplementary Fig. 9) and did not exceed chance levels for the least robust (p>0.5) place cell events (p=0.17 AUC, Supplementary Fig. 10).To confirm these results, we constructed an 'event-ratemap' for each grid cell using the spikes emitted during replay events and the posit...

No abstract