Dynamics of Central Pattern Generating Networks: Locus of Control
چکیده
Oscillatory activity, paramount throughout the central nervous system, is of special significance in the production of rhythmic motor movement , such as locomotion, breathing, and chewing. Such movements are typically generated by networks of synaptically coupled neurons, known as central pattern generators (CPGs), which can produce rhythmic output even without patterned sensory or other extrinsic inputs. In an oscillatory network, participating neurons are commonly active at a specific phase of the network cycle, a phenomenon known as phase locking. The activity phase of neurons within a CPG corresponds to the order in which the involved muscles become active and is thus decisive in the production of meaningful motor behavior. Additionally, most motor rhythms operate across a range of frequencies—e.g., walking vs. run-ning—and the operation of a CPG thus depends on factors that maintain or change the activity phase of CPG components as a function of network frequency. Factors that determine phase in an oscillatory network have attracted considerable attention from theorists [4]. In particular, methods of dynamical systems, such as geometric singular perturbation theory and averaging, have been successfully applied to understand how coupled oscillatory components interact to produce phase-locked network oscillations. Extensive theory has been developed for weakly coupled networks , such as coupled oscillators with similar frequencies, but theoretical analysis of networks involving moderate to strong coupling of dis-parate components has lagged behind. This is so in part because the outputs of such networks are much more dependent on the coupling architecture , which cannot be averaged or universally approximated. It is commonly found that the generation of network oscillations results from dynamic interaction between properties of synapses and voltage -gated ionic currents. However, network oscillations may be more dependent on a subset of network parameters as opposed to others. Recent mathematical and computational studies have shown that networks in different states of activity—in different frequency ranges, for example— can be sensitive to synaptic parameters or to parameters affecting intrinsic properties of involved neurons, but not both. It is well known that most synapses can change in efficacy in a use-dependent manner, a phenomenon known as short-term synaptic plasticity (STSP), which can take the form of facilitation or depression, depending on whether the synaptic strength is increased or decreased. STSP is important in determining network dynamics because the time scales at which it operates are the same as the gating time scales of nonlinear ionic currents underlying intrinsic …