Influences of self-motion signals on the hippocampal neural code for space

Abstract

The experiments that makeup this dissertation are designed to test the theory that the hippocampus functions as a path integrator of optic, vestibular, and ambulatory self-motion information. Hippocampal neural recordings were made in eight subjects during manipulation of the self-motion cues. In the first experiment, rats were trained to drive a car between physical locations on a circular track, thereby eliminating ambulatory self-motion signals. This manipulation resulted in a multitude of changes in hippocampal neural activity. The spatial information content and firing rate of CA 1 pyramidal cells, the power of the hippocampal theta rhythm with its first harmonic, and the modulation of single-unit firing by the theta rhythm were significantly reduced, but not entirely eliminated, when ambulatory cues were eliminated. The amplitude of one of the two hypothesized current generators of the rhythm is shown to be dependent on the ambulatory velocity of the animal. Higher velocities were associated with more prominent "shoulders" in the theta wave. During driving, the theta wave was similar in shape and amplitude to the theta wave obtained during low (near 0)-velocity walking. The results indicate that distance information from locomotor activity is represented in the hippocampal theta rhythm and the firing rates of CAI pyramidal cells. In the second experiment, the entire maze was rotated around the animal during driving, a condition simulating movement optically when the animal was, in fact, stationary. Under these conditions, place specificity and firing rate were further reduced compared to the car driving condition. The theta rhythm that remained after the elimination of ambulatory cues was not affected by the additional elimination of vestibular self-motion cues. Overall, the data suggest that directional information from the vestibular system and distance information from ambulation are integrated in the hippocampal ensemble code for space. Thus, the results are consistent with a role of the hippocampus in path integration.