Sensory-motor integration in a nonspiking interneuron contributes to active sensor control in Drosophila

An animals nervous system enables it to detect and respond to stimuli to navigate its environment. To enhance sensory acquisition, animals can actively position sensors, altering how they extract information from the external world. However, active sensing, and movement in general, produces sensory feedback, requiring mechanisms for integrating predictive motor signals with externally generated sensations. Despite the importance of these mechanisms for guiding coordinated behavior, the cellular and circuit basis of motor control and sensory processing during active movements are not fully understood. Here, we investigate how mechanosensory information contributes to motor output in the Drosophila antenna, a sensor that can be actively positioned. We combine electrophysiology, quantitative behavior, optogenetics, and connectomics to characterize APN2, a nonspiking interneuron in Drosophila that encodes both mechanosensory stimuli and motor commands. We show that these neurons receive input from two classes of antennal mechanosensors, enabling responses to mechanical perturbations in both quiescence and flight. We then demonstrate how these neurons integrate higher-order motor commands with mechanosensory input to shape antennal movement. Together, this work reveals a previously uncharacterized sensory-motor circuit in the antennal mechanosensory system. These findings provide insight into how nervous systems integrate sensation with motor commands to guide the movement of active sensors, highlighting circuit mechanisms that may broadly support sensory acquisition during movement.