Models of the epilepsies have been developed to address a variety of different issues. Although much of the current focus in epilepsy research is on the cellular and molecular mechanisms underlying abnormal activities of the central nervous system (CNS), the usefulness of such information becomes clear only within a larger context that is provided most valuably by intact-animal models of the epilepsies. Chapters 1 to 5 provide an introduction to some of the issues that can be addressed effectively using intact-animal models. The discussions concerning these models make it clear that we can learn a great deal simply from careful examination of the intact animal – from characterization of behavioral seizures, as well as from electroencephalographic (EEG) phenomenology. These discussions also illustrate a major advantage of studying ‘epilepsy’ in animal models, as opposed to examining the epilepsies directly in human clinical material – the ability to control what appear to be a large number of relevant variables.

Control of the stimuli that initiate the epileptogenic process is a feature of kindling that has made this model perhaps the most widely used of all the current intact animal approaches to studies of the epilepsies. McNamara, one of the leaders in investigation of the kindling model (McNamara et al, 1985), lays out a number of salient features of this model – both technical and conceptual (Chapter 1, this volume).

The kindling model of epilepsy

Morphologic rearrangements of kindled animals

Identification of neuronal rearrangements in the brain of kindled animals has provided a potential structural explanation for the hyperexcitability of the kindled brain. Sutula and colleagues were the first to demonstrate the sprouting and permanent reorganization of the mossy fiber axons of the dentate granule cells of the hippocampus of kindled animals (Sutula et al., 1988; Cavazos et al., 1991). They subsequently demonstrated a 40% loss of neurons in the dentate hilus of animals in which 30 kindled seizures had been evoked (Cavazos & Sutula, 1990). This latter observation is important for two reasons. It demonstrates that periodic seizures without overt cyanosis are sufficient to induce neuronal loss and at least part of the picture of Ammon's horn sclerosis; this, in turn, suggests that recurrent isolated complex partial seizures, not merely status epilepticus, may be sufficient to kill neurons in humans. It also suggests that part of the mechanism of the mossy fiber sprouting involves a denervation or loss of synaptic input into the inner third of the granule cell dendrites.

The functional consequences of the mossy fiber rearrangements in the kindled brain and in other models remain controversial. Whether the net effect of the synaptic rearrangements is an elevated or reduced seizure threshold has generated intense arguments.

The chapters in the previous section focused on questions that could be addressed with intact-animal models of the epilepsies. In Section 2, there is an attempt to deal more directly with mechanisms that may underlie epileptiform properties that give rise to, for example, the models discussed in Section 1. Two broad issues are addressed by the chapters in this section, (a) What are the features of epileptic brain, i.e. what are the mechanisms that might underlie the production of abnormal epileptiform activities? (b) What do we know about the processes of epileptogenesis itself, i.e., how does the epileptic brain become epileptic? Most of our understanding of underlying mechanisms is derived from studies on brain (or brain tissue) that is already epileptic; we have a still minimal insight into the process of epileptogenesis. Studies have begun to establish which brain attributes are correlated with epileptiform (i.e., abnormal) activities and are providing clues about the consequences of seizure activity; however, initial ‘cause’ is, in most cases (and models) still to be determined. It has been tricky to separate the underlying features of epileptogenesis from characteristics of the already epileptic tissue.

Concepts and models

There is little doubt that advances in epilepsy research are occurring at a rapid rate. For those of us who are interested in understanding the mechanisms underlying epileptiform activities – whether because of a basic interest in how the brain works, or driven by a concern for more effective treatments for the epilepsies – progress in the laboratory has been almost bewildering. It seems that in each new issue of each neuroscience publication there are new insights and possibilities that we must integrate into our old frameworks. Much of the ‘progress’ has been propelled by advances in technology. For example, at the electrophysiological level, new recording techniques such as patch-clamping have allowed investigators to gain much greater detailed understanding of singlecell properties. New antibodies and tract-tracing techniques have provided information about specific cell populations and about plasticity in neuronal interconnections. Molecular neurobiology is beginning to make a considerable impact on epilepsy research, providing techniques for studying the genetics of inherited epilepsies, as well as for examining the structure and expression of channels and receptors. Pharmacological and neurochemical methods now provide highly sensitive means for analyzing receptor populations, and for assaying transmitter systems (e.g., with microdialysis).

In Section 1 of this volume, many current concepts in epilepsy research were discussed within the context of intact animal models. In Section 2, the emphasis was on insights from the in-vitro examination of tissue from such models. In this last section, the experimental ‘model’ begins with normal tissue. The investigators have asked, ‘What are the basic cellular properties, present in normal cells and tissue, that could contribute to the generation of abnormal activity?’ These studies provide a lexicon of the ‘mights’ and ‘coulds’ with respect to various forms of epileptiform activities.

The discussions presented in this section provide examples of a variety of different levels of analysis. In the chapters by Wilson & Bragdon (Chapter 11) and by Connors & Amitai (Chapter 12), the emphasis is on local circuitry (the connectivity between neural elements) and a consideration of which elements are critical for the generation of abnormal activities. Connors & Amitai pursue this issue by examining the properties of a given cell type as it relates to the generation of epileptiform activities within the circuit. In Chapter 14 (McBain et al) the focus is on the influence of the extracellular milieu surrounding neurons – how changes in that environment might lead to the transformation from normal to abnormal activities, even in the absence of specific abnormalities within given cell populations.