Dynamical Consequences of Fast-Rising, Slow-Decaying Synapses in Neuronal Networks

this article, we will consider the role of recurrent synaptic excitation in creating bistability in networks. Bistability and recurrent excitation have been implicated in many aspects of cortical physiology. For example, epileptic activity is thought to be maintained by the strong recurrent excitation between neurons (Traub & Miles, 1991). In disinhibited cortical slices, recurrent excitation is responsible for the propagation of synaptically generated bursts (Golomb & Amitai, 1997). Delayed-response tasks as well as localized attractors have been modeled via neural nets by recurrent excitatory interactions between model cells (Amari, 1977; Lisman et al., 1998; Compte et al., 2000; Laing & Chow, 2001; Gutkin et al., 2001; Wang, 1999, 2001) Indeed, Lisman et al. (1998) use the nonlinear voltage dependence of the NMDA synapse to construct a bistable neuron model. Here, we will consider a simpler scenario, where just the slowness of the excitatory decay leads to the bistability in a recurrent networ