Circuit mechanisms shaping place- and grid-cell representations under environmental deformation

Place cells in the hippocampus and grid cells in the medial entorhinal cortex jointly support spatial representations, yet how interactions between these populations shape representational changes following environmental deformation remains unclear. We developed a biologically grounded circuit model that integrates boundary vector cells, place cells, and modular grid-cell networks through plastic boundary-to-place, place-to-grid, and grid-to-place projections. During initial exposure to compressed or elongated enclosures, grid cells exhibited module-dependent grid rescaling and boundary-conditioned phase shifts, while place cells showed heterogeneous changes in place-field size and position. Transfer-entropy analysis revealed movement- and boundary-axis-dependent shifts in the directional balance of predictive dependence between place- and grid-cell populations relative to the familiar baseline. During departure, this balance shifted toward grid-to-place and place-to-grid predictive dependence along the deformed and undeformed axes, respectively. With repeated familiarization, different representational features followed distinct, geometry-dependent trajectories rather than uniformly reverting to their familiar-environment configurations. Selective blockade of plasticity in individual pathways revealed distinct effects on familiarization: place-to-grid plasticity supported grid-scale recovery and gains in spatial information, boundary-to-place plasticity modestly promoted grid-scale recovery, and grid-to-place plasticity helped constrain rescaling overshoot. Together, these findings suggest that environmental deformation elicits a multidimensional reorganization of spatial representations, shaped by dynamic place–grid interactions and pathway-specific plasticity, rather than a single uniform transformation.