UC San Diego

The hippocampus- the aptly named seahorse-shaped structure located in the medial temporal lobe of the human brain- has been central to the study of memory for over fifty years. Over this time period, memory research in humans and rodents has focused on the hippocampus’ role in memory differently, with the human work focused on declarative memories for events and facts while rodent work has emphasized the spatial nature of memories. While these two views are not necessarily incompatible, there are still many open questions as to why the rodent hippocampus appears to rely upon spatial aspects of memory far more than its human counterpart.

This dissertation was designed to bridge these two traditions using a spatial navigation task, path integration, which was first described in both humans and animals by Charles Darwin. We approached this first by studying whether working memory could support path integration. Working memory is independent of the hippocampus; however, its capacity is limited by the amount and complexity of information needed for recall. If the paths the subjects needed to recall were short and simple, presumably working memory could support them. Indeed, we found that patients with hippocampal damage were able to path integrate as well as normal subjects. However, rats with hippocampal damage were impaired relative to controls regardless of path complexity.

This finding may reflect the limited capacity of rodent working memory; it may not be possible for rodents to path integrate without using long-term memory. Rats with lesions of the hippocampus and the medial prefrontal cortex (the region commonly associated with working memory) were tested on a variety of spatial and nonspatial memory tasks. We concluded that the hippocampus is needed for spatial memory in rodents not because it is dedicated to spatial processing, but because spatial tasks inherently exceed the capacity of rodent working memory.

These studies, in addition to a third study on extrahippocampal regions involved in the path integration, help to connect the declarative and spatial memory views of the hippocampus and provide a cross-species interpretation of how the brain helps us remember.