The dentate gyrus efficiently converges LEC and MEC inputs into multimodal, highly specific and reliable environmental representations.

The entorhinal cortex-hippocampal network plays a key role in the processing, storage and retrieval of contextual information. However, how convergent multimodal information provided by the lateral and medial entorhinal cortex (LEC and MEC) is represented by granule cells (GCs), the principal cells of the dentate gyrus (DG), remains poorly understood. In this study, we employed two-photon calcium imaging of LEC and MEC projections in the DG together with GC population activity in mice navigating familiar and novel virtual environments over five consecutive days. We found that LEC inputs primarily convey olfactory information, whereas MEC inputs provide rich context-specific information to the DG. Although environmental representations rapidly emerged in LEC and MEC projections upon exposure to novel environments and remained stable over time, representations in the DG improved gradually and required repeated exposure to stabilize. Our findings suggest that the convergence of rapidly emerging, high-dimensional LEC and MEC inputs into a sparse and slowly evolving GC population code provides an energy-efficient mechanism for generating multimodal, modality-specific and context-specific representations in the DG.