Abstracts

Synchronized and recurrent neuronal bursting is a signature of epileptiform activity. Much work has been done on mechanisms underlying these pathological network behaviors. Several mechanisms including membrane after-hyperpolarization and depletion of synaptic transmitter have been recognized as being implicated in the mechanism of recurrent bursting. Using neuronal network models we demonstrate another plausible mechanism for recurrent bursting in disinhibited neuronal networks, caused by short-term (10-100 ms) synaptic depression., Simulated networks were composed of inhibitory (fast spiking) and excitatory (regular spiking) neurons. Excitatory neurons in response to injected external excitatory current exhibited no spike frequency adaptation. The synaptic weights in networks were dynamically modified with ongoing network activities. To simulate synaptic plasticity we used a phenomenological model in which the synaptic strength change depends on the history of pre-postsynaptic activity and is determined by the synaptic calcium residual level. The networks were activated by action potentials applied at random intervals on inputs of selected excitatory neurons (less then 25% of the total number of neurons). Neuronal bursting in these networks was induced by gradually reducing the excitatory drive to inhibitory neurons., Action potentials applied at random intervals to selected neurons activated the entire network but no bursting behavior was observed. When excitatory drive to inhibitory neurons in these networks was gradually reduced, excitatory neurons exhibited steady, synchronous, and periodic neuronal bursting. Single bursts in neurons lasted for up to 60 ms and were followed by [sim]100 ms long periods of inactivity. This temporary cessation of single trains of APs in neurons was mediated by synaptic depression. This was confirmed by the two observations: 1) the maximum of short-term synaptic depression of presynaptic connections overlapped with burst cessation in examined excitatory neurons in these networks, and 2) when short-term depression was not simulated there were no bursting behaviors in these networks. The recovery from synaptic depression was always followed by the onset of new bursts in examined neurons, and networks returned to bursting mode., Our results show that short-term plasticity can be a plausible mechanism contributing to the termination of neuronal bursting. In this network model the cessation of single trains result directly from synaptic depression. We specifically utilized a calcium-regulated model of synaptic plasticity. However, other forms of synaptic plasticity matching the time scale of tens of milliseconds might be also responsible for the recurrent cessation of bursting events, including synaptic depletion., (Supported by NIH NS 51382, NS 38958.)