SUBCELLULAR BASIS OF SUPRAMODAL SPATIAL REPRESENTATIONS IN MOUSE ASSOCIATIVE CORTEX

Generalizing learned rules across sensory modalities is a hallmark of flexible cognition, yet the computations that support supramodal representations remain poorly understood. In previous work, we identified the mouse rostrolateral (RL) associative cortex as a key causal node for cross-modal generalization and proposed a model in which RL encodes a modality-invariant spatial variable in peripersonal space. Here, we ask how supramodal coding in RL is implemented within single neurons, focusing on synaptic convergence and dendritic integration as candidate mechanisms bridging systems-level function and cellular computation. We perform chronic two-photon imaging in head-fixed mice during passive visual, tactile, and combined visuo-tactile stimulation, targeting superficial RL. At the population level, we map how spatial tuning is distributed across RL ensembles and how tuning relates to multisensory responses. At the single-cell level, we longitudinally image layer 2/3 pyramidal neurons throughout their somatodendritic arbors to link somatic responses to local dendritic activity within the same identified neurons across sessions. Using sparse co-expression of genetically encoded glutamate and calcium indicators, we simultaneously monitor excitatory synaptic drive and postsynaptic calcium signals in dendritic shafts and spines. This dual readout provides a direct window onto where sensory inputs from different modalities converge along dendrites and how local integration is reflected in neuronal output. By resolving multisensory processing at synaptic and dendritic scales, this work will connect RL’s causal role in cross-modal generalization to dendritic mechanisms that support supramodal representations of peripersonal space.