Generation, Amplified excitatory postsynaptic potential (EPSP) leads to Occurrence, A paroxysmal depolarizing shift

It has been proposed that as the potentiated EPSP begins to depolarize the neuron, a threshold is reached for the development of a slowly inactivating Na⁺ current that amplifies the depolarization. As depolarization continues, the low threshold Ca²⁺ current may turn on to further depolarize the neuron, while NMDA-mediated excitatory synapses become more effective. Eventually, both higher threshold Na⁺ and Ca²⁺ currents are activated, and the neuron discharges with a burst of action potentials and an additional slow depolarization (Herron et al. 1985; Dingledine et al. 1986). This hypothesis involves the interplay of both synaptic and voltage-dependent intrinsic events that occur in normal central neurons.

An alternative hypothesis for PDS generation focuses more on changes in the intrinsic properties of neurons resulting in the development of burst firing independent of a primary change in synaptic interactions (Dichter and Ayala 1987).

For the first hypothesis, Higashida and Brown (1986) and Madison et al. (1986) have demonstrated that epilepsy occurs when the usual balance of these normal events is altered by a change in synaptic efficacy or a change in the control of intrinsic membrane currents. The reason that any given form of epilepsy may develop in a given brain region may depend on (i) differences in densities and locations of channels on various neurons, (ii) the interaction of intrinsic currents with one another and with synaptic currents under physiological conditions, (iii) the local synaptic organization of a given area, and (iv) the liberation of endogenous synaptic modulators that may alter the various voltage-dependent membrane currents through second messenger pathways.

For the alternative hypothesis, many studies have shown that in epilepsy models, this is most readily accomplished by inhibiting K⁺ currents and by allowing the slower Ca²⁺ currents to be expressed (see review by Dichter and Ayala (1987)). Under more "natural" circumstances, a variety of possible mechanisms may contribute to the development of endogenous burst propensity in the absence of exogenous epileptogenic agents: (i) neuromodulators (acetylcholine, norepinephrine, and peptides) can reduce K⁺ currents (although stimulation of endogenous pathways, even intensely, has not been shown to produce sufficient inhibition of K⁺ currents to result in burst firing); (ii) both elevation of extracellular K ([K⁺]) and reduction of extracellular Ca ([Ca²⁺]) can change membrane characteristics and induce burst firing modes; and (iii) anatomical distortion and redistribution of channels after injury and partial denervation as seen in chronic epileptic foci may induce burst firing.

In addition to the above two hypotheses with empirical evidence, some investigators have proposed that neurons with endogenous bursting characteristics must act as a pacemaker in order for epileptiform activity to develop (see review by Dichter and Ayala (1987)). Such neurons would be the CA2 and CA3 pyramidal cells in the hippocampus, layer IV and superficial layer V neocortical pyramidal cells, or the abnormally burst-firing neurons in chronic neocortical foci. This hypothesis is supported by the demonstration of the lower threshold for the induction of interictal discharges by epileptogenic agents in CA2 and CA3 and layer IV, the spread of abnormal activity from these areas to nearby areas in some experimental foci, and by the correlation of the number of bursting cells with the seizure frequency in chronic foci.

However, this hypothesis has been challenged on theoretical grounds by models that demonstrate that a system with either positive or negative feedback elements does not require unstable individual elements in order to develop oscillating behavior. There is also experimental evidence against the obligatory involvement of neurons with endogenous burst-firing characteristics. Studies of in vivo hippocampal penicillin epilepsy and in vitro low Ca²⁺-high K⁺ models of epilepsy indicate that area CAl is able to develop spontaneous IDs and seizures independent of areas CA2 and CA3. In addition, neocortical and spinal cord cultures, in which individual neurons do not discharge with intrinsic bursts, become organized into small synaptic networks that show synchronized "burst" behavior-all as a result of synaptic interactions. Thus it appears that endogenous, Ca²⁺-dependent bursts are not strictly necessary for the development of synchronous bursting activity in a neural network, although their presence may be facilitatory and CNS regions containing such burst-firing neurons may have a particularly high epileptiform potential.