Abstract
In the striatum, interneurons tightly control the activity of output neurons, the striatal projection neurons (SPN). Among these striatal interneurons, parvalbumin-expressing (PV) and somatostatin-expressing (SOM) cells exhibit markedly distinct electrophysiological and anatomical properties, suggesting differential integration into striatal circuits and distinct forms of feedforward inhibition onto SPN. However, it remains unclear whether differences already emerge at the level of corticostriatal connectivity and if they shape specific interactions with cortical inputs to SPN. Here, we investigated the spatial organization of the projections from the primary and secondary somatosensory areas (S1 and S2), onto striatal PV and SOM interneurons and SPN. With a combination of anatomical tracing and functional circuit mapping, we characterized key features of corticostriatal connectivity, including input number, origin, spacing, overlap and strength. We found that input spacing and strength, at both the individual cell and population levels, were the principal features distinguishing PV and SOM interneurons. In contrast, connectivity differed primarily in input number and laminar organization across S1 and S2. Our findings suggest two potential modes of interaction between interneurons and SPN connectivity: one in which PV interneurons receive S1 inputs structured in large clusters, that densely overlap with those of neighboring SPN, providing an anatomical substrate for potential neuronal assemblies. And another in which SOM interneurons also receive abundant cortical inputs, but thinly disseminated across S1 and S2 and that overlap little with those of SPN, potentially shaping these cell integration through input cross-interaction. These cell-type principles of corticostriatal wiring may support distinct forms of striatal computation.