MEDIAL SEPTAL POPULATION DYNAMICS ENCODE SPATIAL VARIABLES AND PREDICT HIPPOCAMPAL PLACE-CELL ACTIVITY

The medial septum (MS) is a key generator and modulator of hippocampal theta oscillations, providing a scaffold for spatial navigation and memory. Consistent with this role, damage or inactivation of the MS impairs spatial memory and path integration. Navigating space depends on the integration of internal and external cues—including information about position, direction, and movement speed— to estimate displacement and changes in orientation. While previous work has provided limited insight into how these variables are represented at the single-cell level within the MS, their organization at the population level remains largely unexplored.
Here, we used large-scale Neuropixels recordings simultaneously targeting the MS and hippocampus, together with pupillometry as a proxy for arousal state, during a spatial alternation task on a theta-shaped maze. By analyzing MS population activity with CEBRA, we decoded the animal’s position and direction, its running speed, as well as combinations of these variables. Decoding accuracy was consistently higher for the labelled data compared to shuffled controls across all auxiliary variables. Moreover, oscillatory principal components extracted from the MS population activity predicted hippocampal place-cell firing rates more accurately than hippocampal theta LFP, underscoring the role of the MS in organizing place-cell assemblies. These findings motivate future optogenetic manipulations aimed at causally linking MS population dynamics to spatial navigation and hippocampal place-cell organization.