UCLA

The hippocampus is a brain structure critical for the formation of new memories. Case studies have demonstrated that a properly functioning hippocampus is required to form new episodic memories. But hippocampal cellular electrophysiology has traditionally studied hippocampal function by quantifying the role of the hippocampus in navigation and spatial localization through “place cells.” The mechanisms through which allocentric spatial representations arise from egocentric sensory cues are not well-understood, nor are the ways in which these representations are used to navigate or relate to episodic memory. This dissertation addresses three key issues in understanding the mechanisms of episodic memory.

First, in vivo electrophysiology has traditionally focused on somatic spiking as the output of the brain structures under study, completely ignoring the activity in the massive dendritic arbors that neurons maintain, and which have been proposed to have important contributions to neural computation. Here we describe development of a novel recording method and report the first long-term recordings during unrestrained, natural behavior of cortical dendritic activity representing egocentric motion. Second, confounds between different sensory or behavioral parameters can lead to problems in interpreting analyses when traditional methods of determining neural receptive fields are used. This work develops and applies powerful statistical analyses known as generalized linear models to identify the independent contributions of different sensory modalities to hippocampal tuning. Finally, in vivo electrophysiology has never been performed in the Morris Water Maze, one of the most widely-used behavioral tests of spatial memory, largely due to the incompatibility of water with recording electronics. By training rats to do an analogous task in virtual reality, we demonstrate that hippocampal pyramidal cells, which function as place cells in most environments, do not encode allocentric position in this task, but rather encode episodic distance within trials.

This multiple-level investigation of episodic memory, from cortical dendrites to hippocampal soma, is facilitated by novel technologies and methods. Each level of analysis reveals important new details of the computational principles utilized by neural circuits, and lead us to a more complete understanding of episodic memory.