Investigating weakly coupled oscillators in the stick insect locomotor system using Dynamic Causal Modeling

Nils Rosjat, Biocenter, University of Cologne
Mar 18 2015 - 3:00pm
Event type: 
Applied Dynamical Systems
Fine 224

Walking results from a complex interplay of central pattern generating networks (CPGs), local sensory feedback signalizing position, velocity and forces generated in the legs, and coordinating signals from neighboring limbs. In the stick insect, the neural basis of inter-segmental coordination, and the precise effects of sensory information on the central networks in establishing coordinated motor output are largely unknown. The antagonistic muscles of each leg joint are driven by one CPG that can be activated by the muscarinic acetylcholine agonist pilocarpine. Sensory information plays a crucial role in coordinating the different CPGs of one leg and appears to play a major role in inter-segmental coordination of the different legs, too. However, precious little is known about the interactions between the different CPG networks. Büschges et al. 1995 could show that there appear to be episodes when CPGs of different segments tended to be active in phase. In this study, we aim to uncover putative coordination in and between the three thoracic segmental ganglia by analyzing the activity at the CTr-joint of the different segments in the completely deafferented nervous system of the stick insect, Carausius Morosus. Rhythmic motor activity was induced by bath application of the muscarinic receptor agonist pilocarpine. The activity of the motoneuron pool was recorded by means of extracellular electrodes from the lateral nerves C2 in each thoracic ganglion. We used two different methods to investigate the coupling between the different ganglia. At first we used a descriptive approach to (i) identify time intervals of coupling between the Motorneuron-output of the different CTr-joint CPGs, (ii) provide a measure for coupling strength and (iii) determine if coupling was present with preferred phase relations between the different CTr-joint CPGs. In the second approach we used Dynamic Causal Modeling (DCM) to investigate the coupling structure and strength between two ganglia. We set up different coupling schemes for DCM and compared them using Bayesian model selection methods. There was a clear preference to models with lateral connections in each segment and ipsilateral connections on both sides. Both approaches indicate for the first time a high probability for the existence of ipsilateral inter- and contralateral intra-segmental coupling between the CPGs controlling the coxatrochanteral joint musculature in the stick insect thoracic nerve cord. Additionally they show a stronger intra-segmental coupling in the meso-thoracic ganglion and weaker inter-segmental coupling between meso- and meta-thoracic ganglia.