Topographically organized representation of space and context in the medial prefrontal cortex

Abstract

Spatial tuning of neocortical pyramidal cells has been observed in diverse cortical regions and is thought to rely primarily on input from the hippocampal formation. Despite the well-studied hippocampal place code, many properties of the neocortical spatial tuning system are still insufficiently understood. In particular, it has remained unclear how the topography of direct anatomical connections from hippocampus to neocortex affects spatial tuning depth, and whether the dynamics of spatial coding in the hippocampal output region CA1 such as remapping in novel environments is transmitted to the neocortex. Using mice navigating through virtual environments we addressed these questions in the mouse medial prefrontal cortex, which receives direct input from the hippocampus. We find a rapidly emerging prefrontal representation of space in the absence of task rules, which discriminates familiar from novel environments and is reinstated upon re-exposure to the same familiar environment. Topographical analysis revealed a dorso-ventral gradient in the representation of the own position, which runs opposite to the innervation density of hippocampal inputs. Jointly, these results reveal a dynamically emerging and topographically organized prefrontal place code during spontaneous locomotion. Significance StatementThe neocortex is composed of areas with specialized function (e.g. sensory vs. associational). Despite this functional diversity, emerging evidences suggest that the encoding of space might be a universal feature of cortical circuits. Here, we identified a gradient of spatial tuning depth along the dorso-ventral axis. A complex topography of spatial tuning properties might support a division of labor among medial prefrontal cortical subnetworks to support local circuit computation relevant for the execution of context-dependent behavioral outcomes.