FENS Abstr. 3, A152.25, 2006.

Hippocampal phase precession through mossy fiber facilitation.

K. Thurley¹*, C. Leibold¹ ², D. Schmitz² ³, and R. Kempter¹ ² ³.

¹Institute for Theoretical Biology, Department of Biology, Humboldt-Universität zu Berlin, 10115 Berlin, Germany.
²Neuroscience Research Center, Charite, Medical Faculty of Berlin, 10117 Berlin, Germany.
³Bernstein Center for Computational Neuroscience Berlin, 10115 Berlin, Germany.
*Current address: Department of Physiology, Universität Bern, Bern, Switzerland.

Hippocampal phase precession in pyramidal cells poses an example of a relational spike code. It is hypothesized to be important for learning of sequences of events, e.g. odors or spatial locations, which is a part of episodic and spatial memory. Pyramidal neurons in rat hippocampus exhibit spatially selective activity while the animal is exploring its environment. Such neurons are called place cells, and their receptive fields are called place fields. Furthermore, during exploratory behavior, the extracellular field potential shows oscillations in the theta band (4-12 Hz). O'Keefe and Recce (1993, Hippocampus 3) observed that during place field traversals action potentials of a place cell are emitted at progressively earlier phases with respect to the theta cycle. This phenomenon is called phase precession. The mechanisms underlying the generation of phase precession are only poorly understood. Here, we present a mechanism that explains phase precession as a result of the interplay between short-term synaptic plasticity of a hippocampal mossy fiber synapse and subthreshold theta oscillations of the membrane potential of a CA3 pyramidal cell. Consistent with experiments, our mechanism generates precession of spike phases over a range of up to one theta cycle. Our model also accounts for other features of in-vivo data like non-linear and bimodal distributions of spike phases (Yamaguchi et al., 2002, J Neurophysiol 87) and the variability of spike phases as a function of the animal's position in the place field. Our hypothesis predicts that granule cells in the dentate gyrus, which are presynaptic to mossy fiber synapses, preferentially fire in the first half of a theta cycle.