Navigation is the process by which an agent plans and follows routes within an environment. Animals can navigate their environments in ways that do not seem to result from simply learning a set of stimulus-response associations that result in reward. Rather, their navigation is consistent with some internal representation of the animal’s location within the environment, or a “cognitive map”. In mammals, this is thought to be supported by various spatially sensitive neurons in the hippocampal formation, including grid and place cells. The cognitive map also flexibly adjusts previous representations to reflect changes in a familiar environment.

Several studies have demonstrated continuous spatial and temporal adjustments in how the cognitive map represents the animal’s location. For example, in a rescaled version of a familiar environment, grid and place cells’ firing fields shift position throughout the trial depending on which wall the animal encountered last. These dynamic adjustments may reflect the process by which an animal “redraws” its cognitive map of an environment, switching between different walls as reference points as it relearns how they are arranged. Furthermore, maps are not simply used to locate oneself, but also to plan and execute routes. Accordingly, multiple studies have shown that grid and place cells often represent locations just behind/ahead of the animal, which may reflect the system recalling and planning routes through the environment. How this “time shift” varies between different cells, and throughout a trial, is unclear.

This project investigates the principles that organise these dynamic adjustments in the location represented by the cognitive map. We used data from experiments in which rats foraged in four similar asymmetrically deformed enclosures, only differing in the configuration of their western wall(s). Previous work shows that grid and place cell fields near the deformed side of each enclosure shift by larger distances than those on the stable side.

We look for direction-dependent changes in grid cells’ firing fields, and whether these occurred in a pattern consistent with the asymmetrical distortions observed. We used the rat’s head direction as a proxy for which cues the rat was relying on. We find that most grid patterns show statistically significant changes in phase and average firing rate based on the animal’s current head direction, with no clear systematic variation with enclosure shape. We do not find any clear spatial organisation in these direction-dependent phase shifts. This suggests that grid cell rate maps do shift depending on which cues the animal encounters, but that this is separate to the mechanism allowing grid cell maps to distort asymmetrically.

We also examine the time shifts in grid and place cells’ firing in these experiments. We find that both cell types mostly fire prospectively, on average signalling places ~150 ms ahead of the animal’s current location. The size of each grid cell’s average time shift correlates with its scale. Place cells’ time shifts similarly correlate with their firing field size. Cells’ time shifts also vary over the course of each trial, tending to contract as an animal approaches the enclosure walls. We also find that cells’ optimum time shifts are related to the LFP theta phases at which they tend to fire. These findings suggest that different grid modules together represent a spectrum of locations ahead of the animal, organised by the theta cycle.

To summarise, this work shows that grid cell maps appear to continuously adjust their location readout even in familiar environments; however, these do not appear to contribute to their distorted appearance in deformed enclosures. We also show that grid and place cells primarily represent space prospectively as rats explore open fields. These time shifts are organised by each cell’s spatial scale, which hints at how the anatomical organisation of cells by firing scale may be functionally significant. Time shifts are also related to the theta cycle, suggesting how memory encoding and retrieval may be organised in the hippocampal network. Overall, these analyses help isolate and explain subtle functional signals in what is normally considered noise in the cognitive map’s representation of space. This helps us understand how the brain incorporates new information about a familiar spatial context and how it processes it to plan and select appropriate actions.