LINKING IN VIVO RESPONSES, MOLECULAR IDENTITY, AND CONNECTIVITY OF CORTICAL NEURONS AT SCALE

Computations in the neocortex emerge from the interactions of functionally diverse neurons belonging to distinct cell types with different connectivity rules. Therefore, understanding the circuit mechanisms underlying sensory processing requires studying the synaptic connections of single neurons in relation to their functional properties and molecular identity. Previous studies have identified a like-to-like connectivity rule among layer 2/3 excitatory neurons in the primary visual cortex (V1), whereby neurons with similar tuning preferences are more likely to form strong, bidirectional connections. However, it remains unknown how functional wiring rules differ between neuronal cell types across the cortical column. To map synaptic inputs and molecular identities of functionally characterised neurons in mouse V1, we combined in vivo two-photon calcium imaging during visual stimulation with Barcoded Rabies In Situ Connectomics (BRISC). BRISC uses libraries of rabies viruses expressing random molecular barcodes to uniquely label hundreds of starter neurons in parallel, which then transmit their barcodes to their direct presynaptic partners. By matching barcodes between starter and presynaptic cells using in situ sequencing, we reconstruct input connectivity at single-cell resolution at a much higher throughput than traditional approaches. Registering in vivo imaging volumes with the in situ sequencing data enables us to match individual V1 neurons across these datasets, associating their connectivity and gene expression patterns with their visual response properties. Using this integrative approach, we are investigating how molecular identity and circuit connectivity shape sensory representations in V1 and contribute to cortical computations during perception.