Proprioception, the sense of self-movement and position, is mediated by mechanosensory neurons that detect diverse features of body kinematics. Although proprioceptive feedback is crucial for accurate motor control, little is known about how downstream circuits transform limb sensory information to guide motor output. Here, we investigate neural circuits in that process proprioceptive information from the fly leg. We identify three cell-types from distinct developmental lineages that are positioned to receive input from proprioceptor subtypes encoding tibia position, movement, and vibration. 13Bα neurons encode femur-tibia joint angle and mediate postural changes in tibia position. 9Aα neurons also drive changes in leg posture, but encode a combination of directional movement, high frequency vibration, and joint angle. Activating 10Bα neurons, which encode tibia vibration at specific joint angles, elicits pausing in walking flies. Altogether, our results reveal that central circuits integrate information across proprioceptor subtypes to construct complex sensorimotor representations that mediate diverse behaviors, including reflexive control of limb posture and detection of leg vibration.

Dopaminergic neurons (DANs) drive learning across the animal kingdom, but the upstream circuits that regulate their activity and thereby learning remain poorly understood. We provide a synaptic-resolution connectome of the circuitry upstream of all DANs in a learning center, the mushroom body of Drosophila larva. We discover afferent sensory pathways and a large population of neurons that provide feedback from mushroom body output neurons and link distinct memory systems (aversive and appetitive). We combine this with functional studies of DANs and their presynaptic partners and with comprehensive circuit modeling. We find that DANs compare convergent feedback from aversive and appetitive systems, which enables the computation of integrated predictions that may improve future learning. Computational modeling reveals that the discovered feedback motifs increase model flexibility and performance on learning tasks. Our study provides the most detailed view to date of biological circuit motifs that support associative learning.