Divergent co-transmission by a predictive motor circuit modulates auditory processing.

During behavior, an animals movements often stimulate its own sensory receptors. For instance, whether during flight or the production of courtship song, when flies beat their wings they cause mechanosensory disturbances that activate primary auditory neurons in the antennae. However, flight and courtship occur under very different behavioral contexts and temporal-scales. As a consequence, the nervous system contains predictive motor circuits that modify sensory processing in anticipation of self-induced (re-afferent) stimulation. However, it is unclear how information about planned movements can be used to effectively tune specialized sensory processing circuits. Here we describe a predictive motor circuit that releases two transmitters to differentially modify auditory neurons that process sound under different behavioral contexts. In Drosophila, two metathoracic ascending histamine neurons (MtAHNs) become active prior to wing movement and project to the primary auditory network. We found that the MtAHNs co-transmit the neuropeptide Dh44 and histamine, and each suppresses the activity of auditory neurons that are integrated into distinct networks that support escape and courtship, respectively. Additionally, Dh44 has a greater impact on male auditory neurons and suppression by Dh44 occurs over a longer time course than histamine. Disrupting either histamine synthesis or neuropeptide release by the MtAHNs reduces courtship success with an increase in cessation during the wing extension step. Consistent with a role in auditory processing of courtship cues, Dh44 cotransmission is present in Drosophila species that have an auditory component to their courtship displays, but is absent in species that use either purely visual courtship or lack apparent courtship behavior. Finally, the projections of the MtAHNs into auditory centers were missing in snow flies, which are wingless. Our results indicate that divergent cotransmission by predictive motor circuits may represent an efficient means to differentially modulate sensory subnetworks impacted by reafferent feedback.