Layer-specific control of inhibition by NDNF interneurons

Abstract

Neuronal processing of external sensory input is shaped by internally-generated top-down information. In the neocortex, top-down projections predominantly target layer 1, which contains NDNF-expressing interneurons, nestled between the dendrites of pyramidal cells (PCs). Here, we propose that NDNF interneurons shape cortical computations by presynap-tically inhibiting the outputs of somatostatin-expressing (SOM) interneurons via GABAergic volume transmission in layer 1. Whole-cell patch clamp recordings from genetically identified NDNF INs in layer 1 of the auditory cortex show that SOM-to-NDNF synapses are indeed modulated by ambient GABA. In a cortical microcircuit model, we then demonstrate that this mechanism can control inhibition in a layer-specific way and introduces a competition for dendritic inhibition between NDNF and SOM interneurons. This competition is mediated by a unique mutual inhibition motif between NDNF interneurons and the synaptic outputs of SOM interneurons, which can dynamically prioritise different inhibitory signals to the PC dendrite. NDNF interneurons can thereby control information flow in pyramidal cells by redistributing dendritic inhibition from fast to slow timescales and by gating different sources of dendritic inhibition, as exemplified in a predictive coding application. This work corroborates that NDNF interneurons are ideally suited to control information flow within cortical layer 1.