Our group is interested in the function of the hippocampal formation and , in particular, its role in spatial behaviour and spatial memory. In 1971, Dostrovsky and I discovered that the major correlate of hippocampal pyramidal cells was the animal's location (place cells) and suggested that these might form the basis for a spatial mapping system. In 1978, Lynn Nadel and I (www.cognitivemap.net) expanded this notion and suggested that in addition to the place cells, the hippocampal formation might contain information about direction and distance. Together with the place cells this would allow the construction of a cognitive map of the environment. A cognitive map is a device for representing the current environment, the animal's location within it, and the location of desirable objects and threats to be avoided. Its outputs direct the animal's behaviour on the basis of distances and directions towards desired goals or away from undesirable objects and the locations. In addition the cognitive mapping system detects the absence of representations of novel environments and changes in maps of familiar environments and uses these mismatches to trigger and control exploration. We further elaborated on how the cognitive map theory could be expanded by the inclusion of a linear sense of time and the storage of language narratives in the left hippocampus to explain the episodic memory deficit in patients with hippocampal damage. Subsequent work has revealed cells in the hippocampal formation signaling the two types of information necessary to construct maps: in the 1980s, Ranck, Taube, Muller and colleagues discovered head direction cells in the presubiculum, and recently Haftig, Fyhn and the Mosers have found grid cells in the entorhinal cortex which could provide the spatial distance metric.
Over the past 15 years, Neil Burgess and I have constructed computational models of the cognitive map and its components, and have used these to generate theoretically-driven predictions which inform our empirical investigations of hippocampal function. Our experimental attack on these questions /predictions uses various combinations of behavioural, electrophysiological, pharmacological and genetic approaches. Our primary approach is to record from groups of individual hippocampal neurones during spatial navigation, exploration of novel environments, and foraging for food in familiar environments. Questions we are actively exploring/addressing include:
How is an environment and the animal's location within it represented by the firing pattern at the level of single hippocampal place cells and entorhinal grid cells and at the level of networks of such cells? Colin Lever and Tom Wills have found that hippocampal place cells can learn to discriminate between geometrically similar environments in 2 ways. Depending on the degree of similarity between the environments and animal's experience in them, place cells can either learn to differentiate between them as a set of individuals where each cell represents the environment independently of its neighbors (Lever) or alternatively a group of cells can act collectively as an network, taking each others behaviour into account, perhaps operating in ways similar to an attractor network (Wills).
Is learning about the properties of a novel environment dependent on the NMDA receptor and associated intracellular signaling pathways? By studying the place cells of mutant mice with modifications of their glutamatergic NMDA channels ( with R. Schoepfer) or the intracellular signaling molecule alphaCaMKI (with K-P. Giese of Kings College, London), we are trying to understand the synaptic and molecular mechanisms through which place fields develop and are maintained in novel environments. In the latter animal, Cacucci found that spatial tuning fails to show experience-dependent increase over days...