Link to Pubmed [PMID] – 40592329
Link to HAL – hal-05407924
Link to DOI – 10.1016/j.neuron.2025.05.029
Dendrites critically influence single-neuron computations, but their role in neocortical GABAergic interneurons (INs) remains poorly understood. We found that the two major cortical IN subtypes-somatostatin (SST)- and parvalbumin (PV)-expressing cells-use distinct strategies for distributing and integrating excitatory synaptic inputs along their dendrites. SST-INs exhibit NMDAR-dependent supralinear integration and a uniform distribution of synapses, whereas PV-INs show sublinear integration with a higher density of synapses on proximal dendrites with low NMDAR expression. Compartmental modeling revealed that, while both strategies enhance synaptic efficacy, passive integration and proximally biased inputs enable precise tracking of fast-changing signals in PV-INs, whereas NMDARs in SST-INs promote broader temporal integration, supporting sustained activity tuned to slower input variations. Consistent with these predictions, in vivo measurements showed differentially shaped dynamic visual responses in PV- and SST-INs. Therefore, the heterogeneity of dendritic mechanisms strongly influences the spatiotemporal dynamics of IN-specific inhibition in cortical circuits.