Spike phase precession persists after transient intrahippocampal perturbation

Abstract

Oscillatory spike timing in the hippocampus is regarded as a temporal coding mechanism for space, but the underlying mechanisms are poorly understood. To contrast the predictions of the different models of phase precession, we transiently turned off neuronal discharges for up to 250 ms and reset the phase of theta oscillations by stimulating the commissural pathway in rats. After recovery from silence, phase precession continued. The phase of spikes for the first theta cycle after the perturbation was more advanced than the phase of spikes for the last theta cycle just before the perturbation. These findings indicate that phase advancement that emerges within hippocampal circuitry may be updated at the beginning of each theta cycle by extrahippocampal inputs.

Figure 1

Single-pulse stimulation of intrahippocampal afferents resets theta and silences spiking activity. (a) First (upper) panel: superimposed field activity recorded during successive runs in a typical recording session, before and after weak stimulation (time zero). Second panel: instantaneous distribution of phases across successive runs as a function of time. Vertical lines indicate troughs of averaged field activity (data not shown). Third panel: peri-event histogram of three simultaneously recorded interneurons. Fourth (bottom) panel: combined peri-event spiking activity of pyramidal cells (n = 20) and interneurons (n = 3). Color calibration, spikes per bin (fourth panel), probability (second panel). Note transient silence (<200 ms) and full recovery of global firing rate after stimulation. (b) Theta reset after single-pulse stimulation (n = 5 rats). Variance of theta peak occurrence times (vertical bars) in successive cycles after stimulation compared with the control condition after crossing of the photobeam. A smaller variance corresponds to higher phase coherence: *P < 0.01, F-test; colored asterisk indicates significant difference when compared with control; black asterisk indicates significant difference between strong and weak stimulation.

Figure 2

Phase precession is preserved after stimulation-induced perturbation. (a) Phase precession for a place cell during control, weak and strong stimulation runs. Note overall similarity of plots across conditions. Insets: peri-stimulus histograms of spikes (x axis, counts per bin; y axis, time in seconds). (b) For overall group statistics, theta phases were determined separately for each of the early, middle and late subfields and were expressed as deviations from the overall mean phase. Error bars, 95% confidence intervals. Insets: Peak firing rates (PFR) and theta phases (Phase) at peak firing rate locations. Color code for experimental conditions as in Figure 1.

Figure 3

Phase precession is preserved after stimulation-induced perturbation. (a) Firing field during control trials. Color calibration, spikes per second; white bar, 5 cm; arrow, stimulation site. (b) Field theta rhythm, place cell firing (ticks) and average spike phase per theta cycle (in degrees; closed circles) from representative single runs under each condition. Note similar slope of phase advancement in each trial, despite transient disruption of spiking and theta oscillation after stimulation. (c) Group data for 35 neurons. Mean phase difference between the last theta cycle before stimulation and the first theta cycle after stimulation. For the control, the last theta cycle before crossing the photobeam and the theta cycle closest in time to the recovery cycle in stimulation trials were analyzed. Vertical bars, 95% confidence intervals. Under all three conditions, the theta phase advanced significantly (P < 0.05). Control and stimulation conditions did not differ significantly.

Figure 4

Assembly coding is maintained after transient perturbation. (a) Cross-correlograms of two place cells with partially overlapping firing fields (distance between centers, 22 cm). Notice theta rhythm modulation of cross-correlograms. Successive peak occurrence times (vertical dotted lines) were determined for all overlapping pairs (n = 99) for group statistics. (b) Cross-correlation between peak occurrence times and distance between firing fields of neuron pairs. Color code for experimental conditions as in Figure 1.