Anchoring the Cognitive Map to the External World

Representations of space in the broader hippocampal formation are thought to neurally instantiate components of a cognitive map. For a navigator to make use of this map, these representations must be anchored to the external world. Here we address the cues and mechanisms which anchor two of these components – representations of heading and location – to the external world at multiple levels of explanation. In Chapter 2, we record from hippocampal place cells as disoriented mice freely explore or search for hidden rewards in environments of various shapes containing polarizing visual cues. We demonstrate that the hippocampal map is oriented by the spatial geometry alone, and that the orientation of this map predicts search behavior. In Chapter 3, we develop a network-level computational model of how input from border cells could shape the activity of grid and place cells during environmental deformations. From this model we derive a novel prediction: environmental deformations induce dynamic shifts in the phase of the grid location code through interactions with boundaries. We then reanalyze the two key datasets of grid cells recorded during environmental deformations and find clear evidence of these boundary-tethered shifts. Finally, in Chapter 4, we test whether boundary-tethered shifts might contribute to cross-species similarities in the behavior of the place cell representation of rodents and the spatial memory of humans. To this end, we first record from hippocampal place cells as oriented mice explored deformed versions of a familiar environment, and find evidence of boundary-tethered place field shifts as predicted. Next, we teach human participants the locations of objects in a familiar environment, after which they are to replace those objects while in deformed versions of the familiar environment. Across three experiments, we find that the replaced location of objects shifted dependent on the most recently contacted boundary as predicted, even when immersive visual and vestibular cues are available. Together, these results indicate that boundaries play a unique role in anchoring multiple components of the cognitive map. In Chapter 5, we close with a brief discussion of the implications of these findings.