Search

An animal’s nervous system monitors the actions of the body using its sense of proprioception. This information is used for precise motor control and to enable coordinated interaction with the animal’s surroundings. Proprioception is a multimodal sense that includes feedback about limb movement and loading from various peripheral sense organs. The sensory information from distinct sense organs must be integrated by the network to form a coherent representation of the current proprioceptive state and to elicit appropriate motor behavior. By combining intra- and extracellular electrophysiological recording techniques with precise mechanical sensory stimulation paradigms, I studied multimodal proprioceptive integration in the sensorimotor network of the stick insect leg. The findings demonstrate where, when, and how sensory feedback from load-sensing campaniform sensilla (CS) is integrated with movement information from the femoral chordotonal organ (fCO) in the sensorimotor network controlling movement of the femur-tibia (FTi) joint. Proprioceptive information about distinct sensory modalities (load / movement) and from distinct sense organs of the same sensory modality (trochanterofemoral CS (tr/fCS) / tibial CS (tiCS)) was distributed into one network of local premotor nonspiking interneurons (NSIs). The NSIs’ processing of fCO, tr/fCS, and tiCS was antagonistic with respect to a given NSI’s effect on the motor output of extensor tibiae motor neurons (ExtTi MNs). Spatial summation of load and movement feedback occurred in the network of premotor NSIs, whereas temporal summation was shifted between sensory modalities. Load feedback (tr/fCS / tiCS) was consistently delayed relative to movement signals (fCO) throughout the sensorimotor pathways of sensory afferents, premotor NSIs, and ExtTi MNs. The connectivity between these neuron types was inferred using transmission times and followed distinct patterns for individual sense organs. At the motor output level of the system, the temporal shift of simultaneously elicited load and movement feedback caused load responses to be superimposed onto ongoing movement responses. These results raised the hypothesis that load could alter movement signal processing. Load (tiCS) affected movement (fCO) signal gain by presynaptic afferent inhibition. In postsynaptic premotor NSIs, this led to altered movement parameter dependence and nonlinear summation of load and movement signals. Specifically, the amplitude dependence of NSIs opposing ExtTi MN output was increased, and, consistently, the movement response gain of the slow ExtTi MN was decreased. Movement signal processing in the premotor network was altered depending on the proprioceptive context, i.e. the presence or absence of load feedback. Lateral presynaptic interactions between load (tiCS) and movement (fCO) afferents were reciprocal, i.e. existed from fCO to tiCS afferents and vice versa, and also occurred between sensory afferents of the same sense organ. Additionally, a new type of presynaptic interaction was identified. Load signals increased the gain of directional movement information by releasing unidirectionally velocity- or acceleration-sensitive fCO afferents from tonic presynaptic inhibition. Paired double recordings showed lateral connectivity also at the level of the premotor NSI network. NSIs interacted via reciprocal excitatory connections. Additionally, the activity of individual NSIs was correlated in the absence of external stimuli, and specific types of NSIs showed rhythmic 30 Hz oscillations of the resting membrane potential, indicating an underlying mechanism of network synchronization. Taken together, the results of this dissertation provide an understanding of the integration of multimodal proprioceptive feedback in the sensorimotor network by identifying neuronal pathways and mechanism underlying spatial and temporal signal summation. The local network uses multimodal signal integration for context-dependent sensory processing, thereby providing insights into the mechanism by which a local network can adapt sensory processing to the behavioral context. Initial results clearly highlight the necessity to consider lateral connections along sensorimotor pathways to unravel the complex computations underlying proprioceptive processing and motor control. The findings on the integration of proprioceptive signals, obtained in the resting animal, broaden our understanding of sensorimotor processing and motor control not only in the stationary, but also in the walking animal.