Soc. Neurosci. Abstr. 371.17, 2006.

Theta-frequency dependence and phase-place correlations in models of hippocampal phase precession.

R. Schmidt, C. Leibold, K. Thurley, A. Gundlfinger, R. Ritz, D. Schmitz, R. Kempter.

Why does the firing phase of hippocampal pyramidal cells decrease with respect to the theta rhythm during the crossing of a place field? We present a theoretical model that explains phase precession through combination of short-term facilitation of the mossy fiber synapse with theta oscillations in the membrane potential of CA3 pyramidal cells. The model predicts for example that (1) phase precession depends on the length of the current theta period, and that (2) firing phase correlates best with the index of the theta cycle. A test of these predictions allows to separate the synaptic facilitation model from other models of phase precession. According to the synaptic facilitation model, phase precession is more linear for higher theta frequencies (short theta periods), while it is more bimodal for lower theta frequencies (long theta periods).

Previous experimental studies found that, on average over many trials, firing phase correlates better with place than with time since the animal entered the place field. If time is measured relative to the entering of the place field, this introduces a misalignment, e.g. due to different running speeds in different trials. We propose two approaches to overcome this alignment problem. First, phase precession should be analyzed on the basis of single trials. Second, a new alignment criterion, in which a correlation measure is used instead of the position of the animal, should be applied. For both, single trials and correlation-based alignment, we predict that the correlation between phase and theta cycle is better than phase-time or phase-place correlations.