Resources
Authors & Affiliations
Ane Lautrup, Emilie R. Skytøen, Soledad Gonzalo Cogno, Edvard I. Moser, May-Britt Moser
Abstract
Extracting spatiotemporal information from experience is essential for episodic memory. It is known that temporal information is expressed in the activity of neural populations in the lateral entorhinal cortex (LEC). However, whether or how this code for episodic time is integrated with the spatial codes in the hippocampus remains to be determined. To answer this question, we implanted high-density silicon probes in the hippocampal CA1 region and LEC in mice. Activity in the two regions was simultaneously recorded while mice were either freely foraging in an open field arena or head-fixed running on a wheel in sensory-minimized conditions, i.e., in darkness and with neither scheduled rewards nor external stimuli informative for navigation. To determine the amount of temporal information encoded by neural activity in either region, we trained a linear classifier to decode temporal epochs from neural population activity. As expected, the strongest decoding of time was found in LEC in both tasks, consistent with its population activity drifting over a session and enabling a readout of episodic time. However, during running on the wheel in sensory-minimized conditions, there was a larger increase in drift and decoding accuracy in CA1 than in LEC. We are currently testing whether a similar increase from the open field to the wheel task is present in the MEC. Our results suggest that temporal information can be extracted from neural population activity not only in LEC but also in the hippocampus, but drift of hippocampal activity is attenuated by spatial landmarks when these are present.