Introduction

Both clinical and experimental data suggest that various forms of synaptic reorganization in brain tissue may contribute to evolution of epileptic foci. In particular, mossy fiber sprouting is a well-known phenomenon both in human temporal lobe epilepsy and in experimental models of epilepsy (Represa et al., 90, Sutula et al. 98). The axons of the granule cells, the mossy fibers, which normally terminate in the hilus and stratum lucidum of the CA3 subfield, sprout and project through the granule cell layer toward the molecular layer, forming an abnormal termination field in the inner molecular layer. Recent findings have also indicated that inhibitory nonprincipal cells are also involved in newly formed circuitries (Dudek et al., 94). In animal models, recurrent inhibition in the inner molecular layer increases after sprouting has occurred, which suggests that sprouting mossy fibers also form connections with inhibitory interneurons (Sloviter 92).

Our previous simulations (Kudela et al., 1999, Franaszczuk et al., 1999) have shown that randomly connected reduced neuron model networks can produce repetitive bursting activity and we also found that inhibition in a network is one of the major factors controlling these activities. Here we propose a simple model of the sprouting phenomenon in a randomly connected neural network. In this study we examine changes in network burst generation after increasing the number of synaptic connections. We increased both excitatory and inhibitory connections proportionally.