Iii

20 ms

Figure 16.2 Electrophysiological events in the development of an epileptic focus and EEG interictal spike. Intracellular recordings generally show that afferent stimulation of a normal cortical neuron produces one action potential superimposed on a small depolarisation (approx. 10 mV), the excitatory postsynaptic potential of the form drawn in (IA). In a focal type-I epileptic neuron, as found in the CA3 region, the same stimulus can produce a much larger depolarisation, the paroxysmal depolarising shift (PDS) and a burst of spikes (IB). Other neurons must then be recruited and this is shown to be possible in the intracellular recording from two mono-synaptically connected CA3 neurons in the hippocampal slice preparation in which each action potential in the presynaptic neuron (IIA) elicits an excitatory potential in the postsynaptic cell which eventually shows a burst of potentials (IIB). Once a number of neurons are recruited there is an almost synchronous discharge of cortical neurons which give rise to an EEG interictal spike. This can be seen from the extracellular recording made with a glass-coated tungsten microelectrode in the cortex of an anaesthetised rat after topical application of the GABA antagonist bicuculline (III). The burst shown in (i) gives rise to a large EEG spike while the other discharges (ii and iii) correspond to medium and small EEG spikes respectively. (II reproduced from Wong et al. 1986 and III from Neuropharmacology 30: Zia-Gharib and Webster 1991 with permission from Elsevier Science.) See also Fig. 2.14

ORIGIN OF FOCAL NEURONS (A in Fig. 16.3)

(i) Properties of focal neurons

Focal neurons must either possess inherent abnormal electrophysiological characteristics or develop them as a result of morphological changes induced in them or around them following some event. There is little evidence of any abnormality in the intrinsic electrophysiological properties of individual neurons studied in brain slices from human focal cortical or hippocampal tissue, although the possibility of some unidentified genetic change in the characteristics of certain ion channels remains possible. By contrast, in electrically kindled rats, NMDA receptors on dentate gyrus granule cells show some plasticity, which at the channel level is manifest by prolonged bursts, clusters and increased agonist potency. Although these changes persist through the kindled state and must therefore be transferred to new receptors, the molecular basis is not known (see Mody 1998). Brain damage can, however, modify neuron function and so possibly make some of them hyperexcitable and focal.

(ii) Reduced inhibition

It has been known for many years that inhibitory interneurons in the spinal cord are very vulnerable and easily destroyed by a reduction in blood supply and that in their absence motoneurons become much more excitable. So it is possible that localised ischemia or hypoxia in the brain could equally well cause a selective loss of GABA inhibitory interneurons and increased excitability of some pyramidal cells. Certainly there is morphological evidence for the loss of such interneurons from occlusion experiments in rodents, as well as a loss of GABA nerve terminals around a cortical alumina focus in monkeys and reduced GABA uptake, and probably therefore GABA nerve terminals, during brain dialysis in epileptic patients. Despite these findings, any neuronal loss reported in human epilepsy appears to be confined to the larger pyramidal neurons, and these do not release GABA.

(iii) Increased excitation

It is equally well known that if a neuron dies, or is destroyed, then any other neuron, which had been innervated by it, gradually becomes supersensitive to the NT it released. In the case of degenerating pyramidal cells this would be glutamate, the excitatory NT. Not surprisingly, undercutting the cortex in animals to produce a deafferentation of some of its neurons not only renders them more likely to show epileptic-like discharges but neurons in hippocampal slices from kindled rats and human focal cortex show supersensitivity to the excitatory amino acids. Such supersensitivity could make some neurons so easily activated that they become 'epileptic'.

The rate of development of such experimentally induced supersensitivity following denervation or hypoxia is similar to that seen in animals with focal (alumina) lesions but quicker than epileptogenesis following focal pathology (injuries) in humans. Also it must be remembered that although neurons may become supersensitive to glutamate this will no longer be released synaptically from the afferent terminals of the degenerating neurons although its release from others could produce inappropriate, disorganised and extended activation. Indeed there are some morphological changes that would support this.

Figure 16.3 Changes in neuronal function required for the development of epileptic seizures. The factors that may control or induce the changes in neuronal function that turn a normal neuron into a focal one (A) recruit other neurons (focal epileptogenesis) to produce an interictal EEG spike (B) and ensure the spread of activity (general epileptogenesis) to full ictal activity (C) are discussed in the text. They include alterations to various ion channels, especially those for Na+, a reduction in local inhibitory activity or an increase in local excitatory drive. The electrophysiological counterparts of some of the events involved are shown in Fig. 16.2

Figure 16.3 Changes in neuronal function required for the development of epileptic seizures. The factors that may control or induce the changes in neuronal function that turn a normal neuron into a focal one (A) recruit other neurons (focal epileptogenesis) to produce an interictal EEG spike (B) and ensure the spread of activity (general epileptogenesis) to full ictal activity (C) are discussed in the text. They include alterations to various ion channels, especially those for Na+, a reduction in local inhibitory activity or an increase in local excitatory drive. The electrophysiological counterparts of some of the events involved are shown in Fig. 16.2

The dendrites of neurons adjacent to those which degenerate also show extensive growth and sprouting which could facilitate abnormal and disorganised synaptic transmission and cause hyperactivity. It is also known that the dendrites of cells around an alumina focus in monkeys, as well as in human epileptic brain, lose their spinous processes, which might contribute to the paroxysmal discharge by facilitating the spread of depolarisation to the neuron soma. Certainly an increase in the number of Na+ channels on the dendrites of spinal motoneurons, which would facilitate the occurrence of reactive dendritic Na+ spikes, has been seen after axotomy.

ORIGIN OF INTERICTAL AND ICTAL SPIKES (B and C in Fig. 16.3)

There are many studies on the induction and spread of spiking in animals both in vivo and in isolated brain slices, generally initiated by the use of GABA antagonists or removal of Mg2+ ions (in vitro). Unfortunately since neither of these events is likely to occur in or around a human epileptic focus the results do not tell us much about how focal activity arises and spreads in humans. This needs to be achieved by the use of human epileptic tissue even though the procedures found to control experimentally induced spiking may well be applicable to humans.

There have been a number of observations which show increased excitation and/or reduced inhibition in slices prepared from human epileptic brain tissue. Thus burst discharges can be evoked with stimuli that would not do so in normal animal tissue and these can be blocked by NMDA receptor antagonists. The inhibitory postsynaptic currents (IPSCs) in hippocampal dentate granule cells in slices prepared from temporal lobe epileptic tissue are in fact reduced by stimulation that activates NMDA currents (Isokawa 1996), which are more prolonged than usual and show changes in slope conductance.

It is perhaps not surprising that NMDA and AMPA receptor mechanisms are important in epileptogenesis. The summation of EPSPs through activation of recurrent polysynaptic excitatory pathways is necessary to mediate the large depolarisation of neurons in and around a focus and the intense discharge and extracellular field potentials of the interactal EEG spike, although these may only occur if counteracting inhibition is reduced. There is in fact some evidence of morphological changes in human epileptic hippocampal tissue that would facilitate such excitatory circuits with aberrant networks of collaterals from axons of individual mossy fibre neurons ramifying through to the CA3 and other regions (Isokawa et al. 1993). Also the increase in extracellular K+ following increasing neuronal activity may itself reinforce the activity by directly depolarising nerve terminals and neurons. High extracellular K+ would also counteract K+ efflux and so initiate a prolonged low depolarisation that would facilitate repetitive firing.

From this survey it is clear that just as normal neuronal function requires appropriately balanced inhibitory and excitatory controls so the generation of interictal spikes depends on disturbances in both. Clearly activity cannot spread without the activation of excitatory circuits, in which NMDA receptors play an important role, but it will be much facilitated by reduced inhibition (Masukawa et al. 1989). These observations may help to explain the establishment of a focus and the development of the interictal spike, but why activity can only spread to seizure proportions, at certain times, is less clear. It will, however, again require overactivity of excitatory circuits inadequately controlled by inhibitory processes. Since these controls are mediated by

NTs it is now appropriate to consider what evidence there is for a malfunction of NT activity in epilepsy, particularly in those responsible for primary excitation and inhibition, i.e. the amino acids. Before doing so the epileptogenesis of absence seizures (petit mal) justifies separate consideration.

ORIGIN OF ABSENCE SEIZURES

There is much evidence that absence seizures originate in the thalamus probably due to some malfunction of neuronal Ca2+ channels. The sudden synchronous bilateral nature of the slow-wave discharge (SWD) in the EEG which typifies this condition was justifiably considered by Jasper (see Jasper and Drooglewer-Fortuyn 1997) to require a subcortical focus and he was able to reproduce them in anaesthetised cats by 3 Hz stimulation of the intralamina thalamus, which in conscious animals also produced absence-like behavioural symptoms such as staring and unresponsiveness. Also in rats with genetic absence epilepsy (GAER) such symptoms are not only accompanied by a synchronous 7-9 Hz SWD but this coincides with high-amplitude discharges in the lateral part of the thalamus, the lesion of which inhibits SWDs.

Within the thalamus the reticular nucleus, which contains predominantly GABA neurons, sends axons to all the other thalamic muclei and although it does not appear to directly drive any thalamic projection to the cortex it receives collaterals from both thalamo-cortical and cortico-thalamic pathways and is well positioned to influence cortico-thalamic activity. If its neurons are stimulated while slightly hyperpolarised they show repetitive burst discharges in rat brain slices followed by a marked afterhyperpolarisation, i.e. oscillatory activity (Avanzini et al. 1992). Pharmacological studies in vivo in the genetically prone rat show that this depends on the activity of certain Ca2+ and Ca2+-activated K+ conductances and that blocking Ca2+ channels just in the reticular nucleus reduces the cortical SWDs. In fact cloning studies in mutant mice strains with features of absence epilepsy show defects in the subunit structure of these channels (Fletcher et al. 1996), although why such an effect on channels that have a very widespread distribution should manifest itself in rhythmic activity only in thalamic neurons is uncertain. It may, however, depend on a particular inhibitory control and hyperpolarisation induced locally by GABA, which certainly invokes rhythmic activity when applied to firing neurons and potentiates SWDs in GAERs. In fact this response is probably mediated by GABAb rather than GABAa receptors since not only does baclofen (GABAB agonist) have a similar effect to GABA but when GABA is applied to thalamic neurons it produces a bicuculline-insensitive long-lasting but slight hyperpolarisation which is followed by a low-threshold calcium potential (LTCP) and spike. This T-type Ca2+ channel is common in GAERs and larger than normal in thalamic GABA neurons.

Continue reading here: Neurotransmitters In Epileptic Activity

Was this article helpful?

0 0