Theoretical and experimental investigations of an inter-segmental CPG-network model and its application to multi-legged locomotion

Martyna Grabowska Msc., University of Cologne, Germany
Nov 25 2014 - 1:00pm
Event type: 
Dynamical Systems & Nonlinear Science Seminar
Fine 224

In legged animals, inter-segmental coordination is required to perform locomotion. This is true for insects, e.g. the stick insect, as well as for invertebrates with more than six walking legs, like crabs and crayfish. In all of these animals, inter-segmental neural connections, as well as local sensory signals, are assumed to play a crucial role (Barnes 1974; Sillar, Clarac and Bush 1987; Cruse and Müller 1986; Clarac 1982). Daun-Gruhn and Tóth (2011) designed an inter-segmental network model to mimic some basic aspects of the inter-leg coordination in the stick insect. As an initial step, they focused on the local segmental networks with central pattern generators (CPG) as their core elements that generate rhythmic activity of the thoraco-coxal motoneuron pools, i.e. the protractor-retractor neuro-muscular system (ThC-CPG), in three adjacent segments. The three ipsilateral segmental ThC-CPGs are connected rostro-caudally, and the connections are modulated by excitatory sensory signals from anterior segments and by local inhibitory ones from the same segment. It was found that, in order to obtain smooth transitions between different coordination patterns, a neural connection from the meta-thoracic CPG to the pro-thoracic had to be assumed, making the connections cyclic. We present experimental evidence for this caudo-rostral inter-segmental neural connection in the stick insect. A walking hind leg was found to be able to entrain a pilocarpine-induced rhythm in the ThC-CPG of the pro-thoracic segment. Hence, the related assumption in the model could be confirmed. In addition, we used this model to test whether it could serve as basic module of an inter-segmental neuronal control network of n-legged (n>6) walking animals. To this end, we extended the model by additional segmental CPGs that had the same properties as the existing ones in the original model. We changed phase relations between the CPGs by varying appropriate system variables (e.g. excitatory and inhibitory sensory inputs), or model parameters (e.g. drive to the CPGs). In this way, we could simulate different coordination patterns, as well as transitions between them, that were similar to those observed in 8-legged animals. These results show that the inter-segmental network model, which is supported by experimental data gathered in the stick insect walking system, might serve as a fundamental module to simulate walking in animals with more than six legs.