Results

The parameters of neuron and synaptic models were selected to allow for generation of bursts of action potentials in a small network. The network was activated by applying random excitatory input to 4 selected neurons. Output from a typical simulation is shown in Fig. 3. Application of a depolarizing current below a certain threshold level to all neurons temporarily alters bursting activity (Fig. 4), but fails to terminate bursts. However, if the stimulus is strong enough (in this stimulation $I_{ext} > 1.3 mA$ the bursts are abruptly aborted (Fig. 5).

A similar but more sustained effect can be observed by applying random excitatory input to all neurons. If background activity is above threshold ($A_n=0.06$ and $\lambda =1$), the bursts are terminated (Fig. 6) and do not appear so long, as the background activity remains elevated.

Introduction of a long delay excitatory feedback loop changes behavior of the network significantly. In these simulations there is no background activity. Initially the network is quiet. After application of relatively weak stimulus ( $I_{ext}=0.01 mA$), the network generates bursts of regular intervals determined by the delay in the feedback loop (Fig. 7). The termination of the burst in each cell is regulated by the calcium regulated potassium current $I_K(Ca)$, which is responsible for cell adaptation to repetitive input excitation. After termination of a burst there is a period when excitability of the cell is diminished. Application of a depolarizing current changes timing of this period of diminished excitability. This terminates bursting in the network as illustrated in Fig. 8. In both Fig. 7 and 8 External stimulus current is the same $I_{ext}=0.01 mA$. This effect is not dependent on the presence of inhibitory connections and can be observed in a purely excitatory network as well (Fig. 9).