Relations between the theta rhythm and activity patterns of hippocampal neurons.

Abstract

This dissertation explores the relations between the hippocampal theta rhythm, the activity patterns of hippocampal neurons-considered both as individuals and as populations-and memory. The main focus is a series of studies designed to gain a deeper understanding of the phenomenon of phase precession of spike activity in CA1 pyramidal cells, as first observed by O'Keefe and Recce (1993). Making use of parallel recording techniques capable of recording simultaneously from dozens of single units and up to twelve channels of EEG, it is demonstrated that: 1) The population activity of CA1 pyramidal cells is modulated by the theta rhythm, with a depth of modulation of about 50%. The peak firing of interneurons in CA1 (theta cells) occurs about 60° in advance of pyramidal cells. 2) The first spikes emitted by a CA1 pyramidal cell, as the rat enters the cell's place field, come 90°-120° after the phase of maximal pyramidal cell population activity, near the phase where inhibition is least. As shown previously by others, as the rat passes through the place field, spike activity from the cell precesses to earlier phases of the theta cycle, for a net precession that often reaches 360°. 3) Phase precession occurs both on linear tracks and in two dimensional environments where the animal does not follow regular trajectories. 4) Spikes occurring near the end of the theta cycle give less information about the animal's spatial location than spikes that occur near the beginning of the theta cycle. 5) Granule cells of the fascia dentata show maximum population activity approximately 900 phase advanced with respect to the CA1 pyramidal cell population, and are nearly silent during the last quarter of the theta cycle. 6) Granule cells show phase precession similar to that seen in CA1 pyramidal cells, with the exception that the maximum total precession is approximately 2700 rather than 360°. 7) Phase precession also appears to occur during REM sleep, but the evidence for this is not conclusive. These results are shown to imply that portions of the temporal sequence of place fields are replicated repeatedly within individual theta cycles, in highly compressed form. It is hypothesized that this permits the use of long-term potentiation for onetrial learning of the sequence, thereby giving a temporal dimension to hippocampal memory traces. The possible implications of this are discussed, as well as future experiments that may yield insight into phase precession and its functional role.