[Cellular mechanisms of the epilepsies: In vitro studies on human tissue]. / Etudes in vitro sur tissu humain des mécanismes cellulaires des épilepsies.

BACKGROUND AND PURPOSE: Animal models have provided very valuable data to specify the physiopathological mechanisms of the various forms of epilepsy. However, the question arises of knowing which of these experimental results are relevant to the human epileptic brain. The development of epileptic surgery makes it possible to directly study the functional properties of human brain tissue in vitro and to analyze the mechanisms underlying seizures and epileptogenesis. We review some of the results obtained over the last few years in our laboratory based on electrophysiological, immunocytochemical and molecular experiments conducted on human brain tissue. RESULTS: This review covers a number of the mechanisms of neuronal synchronizations generating epileptiform discharges, including the role of electrical synapses connecting the inhibitory interneurons, particularly in Taylor-type focal cortical dysplasia and the functional lability of GABAergic inhibition in epileptogenic human cortical tissue, which may sustain triggering and propagation of seizures. Some of these mechanisms have not been described in animal models. CONCLUSIONS: Studies on human tissue, when carefully designed, are necessary to validate the data collected on animal models and will continue to provide us with new and important information on the cerebral changes related to epilepsy. Moreover, these studies allow development of a class of antiepileptic drugs that have a completely new mechanism of action, which could be effective in the treatment of drug-resistant epilepsies.

Effects in vitro and in vivo of a gap junction blocker on epileptiform activities in a genetic model of absence epilepsy.

We investigated the effects of carbenoxolone (CBX), a gap junctions (GJ) blocker, on epileptiform activities in vivo and in vitro. In a first series of experiments, i.p. CBX decreased the cumulative duration of cortical spike-wave discharges (SWD) in adult Genetic Absence Epilepsy Rats from Strasbourg (GAERS) without reduction in the SW amplitude or frequency. Since SWD are generated in thalamocortical networks, we studied the effect of CBX on thalamic and cortical activities elicited by 4-aminopyridine (4AP) in thalamocortical slices from GAERS or non-epileptic rats (NER). Spontaneous ictal-like activities (ILA) were recorded simultaneously in thalamus and somatosensory cortex. However, experiments where these structures were surgically separated showed that ILA were generated in the cortex and recorded by volume conduction in the thalamus. GABA-dependent negative field potentials were also recorded in the cortex, either isolated or initiating ILA. After bath-applying CBX (100 microM), the frequency and cumulative duration of ILA decreased but less rapidly in GAERS than in NER slices and they disappeared at a time point when GABA-dependent negative potentials remained. These data suggest that GJ do not mediate the 4AP induced interneuronal synchronisation but may be implicated in the spreading of the synchronised activities from interneuronal networks to principal neurones. Our results show that CBX exerts an antiepileptic action in vivo, and that GJ blockers limits spread of synchronised activities in vitro. They may represent an appropriate target for development of new antiepileptic drugs.

Role of gap junctions on synchronization in human neocortical networks.

Gap junctions (GJ) have been implicated in the synchronization of epileptiform activities induced by 4-aminopyrine (4AP) in slices from human epileptogenic cortex. Previous evidence implicated glial GJ to govern the frequency of these epileptiform events. The synchrony of these events (evaluated by the phase unlocking index, PUI) in adjacent areas however was attributed to neuronal GJ. In the present study, we have investigated the effects of GAP-134, a recently developed specific activator of glial GJ, on both the PUI and the frequency of the 4AP-induced epileptiform activities in human neocortical slices of temporal lobe epilepsy tissue. To delineate the impact of GJ on spatial spread of synchronous activity we evaluated the effects of carbenoxolone (CBX, a non-selective GJ blocker) on the spread in three axes 1. vertically in a given cortical column, 2. laterally within the deep cortical layers and 3. laterally within the upper cortical layers. GAP-134 slightly increased the frequency of the 4AP-induced spontaneous epileptiform activities while leaving the PUI unaffected. CBX had no effect on the PUI within a cortical column or on the PUI in the deep cortical layers. CBX increased the PUI for long interelectrodes distances in the upper cortical layers. In conclusion we provide new arguments toward the role played by glial GJ to maintain the frequency of spontaneous activities. We show that neuronal GJ control the PUI only in upper cortical layers.

Different pharmacology of N-desmethylclozapine at human and rat M2 and M 4 mAChRs in neocortex.

Cholinergic transmission plays a pivotal role in learning, memory and cognition, and disturbances of cholinergic transmission have been implicated in neurological disorders including Alzheimer's disease, epilepsy and schizophrenia. Pharmacological alleviation of these diseases by drugs including N-desmethylclozapine (NDMC), promising in animal models, often fails in patients. We therefore compared the effects of NDMC on glutamatergic and GABAergic transmission in slices from rat and human neocortex. We used carbachol (CCh; an established agonist at metabotropic muscarinic acetylcholine (ACh) receptors (mAChRs)) as a reference. Standard electrophysiological methods including intracellular and field potential recordings were used. In the rat neocortex, NDMC prevented the CCh-induced decrease of GABAA and GABAB receptor-mediated responses but not the CCh-induced increase of the paired-pulse depression. NDMC reduced neither the amplitude of the excitatory postsynaptic potentials (EPSP) nor antagonized the CCh-induced depression of EPSP. In the human neocortex, however, NDMC failed to prevent CCh-induced decrease of the GABAB responses and directly reduced the amplitude of EPSP. These data suggest distinct effects of NDMC in rat and human at M2 and M4 mAChRs underlying presynaptic modulation of GABA and glutamate release, respectively. In particular, NDMC might be a M2 mAChR antagonist in the rat but has no activity at this receptor in human neocortex. However, NDMC has an agonistic effect at M4 mAChR in the human but no such effect in the rat neocortex. The present study confirms that pharmacology at mAChRs can differ between species and emphasizes the need of studies in human tissue.

Waking selective neurons in the posterior hypothalamus and their response to histamine H3-receptor ligands: an electrophysiological study in freely moving cats.

Neurons which discharge selectively during waking (waking selective) have been found in the tuberomamillary nucleus (TM) and adjacent areas of the posterior hypothalamus. Although they share some electrophysiological properties with aminergic neurons, there is no direct evidence that they are histaminergic. We have recorded from posterior hypothalamic neurons during the sleep-wake cycle in freely moving cats, and investigated the effects on waking selective neurons of specific ligands of histaminergic H3-receptors, which autoregulate the activity of histaminergic neurons. Two types of neurons were seen. Waking selective neurons, termed "waking-on (W-on)," were located exclusively within the TM and adjacent areas, and discharged at a low regular rate during waking (1.71-2.97 Hz), decreased firing during light slow wave sleep (SWS), became silent during deep SWS and paradoxical sleep (PS) and resumed their activity on, or a few seconds before, awakening. "Waking-related" neurons, located in an area dorsal to the TM, displayed a similar, although less regular, low rate of firing (1.74-5.41 Hz) and a similar discharge profile during the sleep-wake cycle; however, unlike "W-on" neurons, they did not completely stop firing during deep SWS and PS. Intramuscular (i.m.) injection of ciproxifan (an H3-receptor antagonist, 1mg/kg), significantly increased the discharge rate of W-on neurons and induced c-fos expression in histamine-immunoreactive neurons, whereas i.m. injection of imetit (an H3-receptor agonist, 1mg/kg) or microinjection of alpha-methylhistamine (another H3-receptor agonist, 0.025-0.1 microg/0.2 microl) in the vicinity of these cells significantly decreased their discharge rate. Moreover, the effect of the antagonist was reversed by the agonists and vice versa. In contrast, "waking-related" neurons were unaffected by these H3-receptor ligands. These data provide evidence for the histaminergic nature of "W-on" neurons and their role in cortical desynchronization during waking, and highlight the heterogeneity of posterior hypothalamic neuronal populations, which might serve different functions during the wakefulness.

Muscarinic acetylcholine receptor-mediated effects in slices from human epileptogenic cortex.

Acetylcholine has been implicated in higher cortical functions such as learning, memory and cognition, yet the cellular effects of muscarinic acetylcholine receptor (mAChR) activation are poorly understood in the human cortex. Here we investigated the effect of the mAChR agonist carbachol (CCh) and various mAChR antagonists in human cortical slices (from tissue removed during neurosurgical treatment of epilepsy) by intracellular and extracellular recordings. CCh increased neuronal firing, which was antagonised by atropine (non-selective mAChR antagonist) and pirenzepine (M(1)/M(4) mAChRs antagonist) when applied before or after CCh application. AF-DX 116 (M(2)/M(4) mAChRs antagonist) had no effect on CCh-induced increase of firing. CCh also reduced evoked excitatory postsynaptic potentials (EPSP), and the CCh-induced depression of EPSP was fully reversed by atropine. Pirenzepine reversed the depression of CCh on EPSP, but failed to prevent the depression when applied before CCh. AF-DX 116 prevented the CCh-induced depression of evoked EPSP when applied before CCh. CCh also depressed GABAergic transmission and this effect was antagonised by AF-DX 116. Xanomeline (M(1)/M(4) mAChR agonist) increased neuronal firing and decreased EPSP, but had no effect on GABAergic transmission. Reduction (with linopirdine) and enhancement (with retigabine) of the M-current (mediated by K(V)7 channels), increased and decreased neuronal firing, respectively, but had marginal effects on the evoked EPSP. Our results indicate that three pharmacologically distinct mAChRs modulate neuronal firing, glutamatergic and GABAergic transmissions in the human epileptogenic neocortex. The data are discussed towards possible implications of altered mAChR signalling in hyperexcitability and cognitive functions in the human neocortex.

Effects of gap junction blockers on human neocortical synchronization.

Field potentials and intracellular recordings were obtained from human neocortical slices to study the role of gap junctions (GJ) in neuronal network synchronization. First, we examined the effects of GJ blockers (i.e., carbenoxolone, octanol, quinine, and quinidine) on the spontaneous synchronous events (duration = 0.2-1.1 s; intervals of occurrence = 3-27 s) generated by neocortical slices obtained from temporal lobe epileptic patients during application of 4-aminopyridine (4AP, 50 muM) and glutamatergic receptor antagonists. The synchronicity of these potentials (recorded at distances up to 5 mm) was decreased by GJ blockers within 20 min of application, while prolonged GJ blockers treatment at higher doses made them disappear with different time courses. Second, we found that slices from patients with focal cortical dysplasia (FCD) could generate in normal medium spontaneous synchronous discharges (duration = 0.4-8 s; intervals of occurrence = 0.5-90 s) that were (i) abolished by NMDA receptor antagonists and (ii) slowed down by carbenoxolone. Finally, octanol or carbenoxolone blocked 4AP-induced ictal-like discharges (duration = up to 35 s) in FCD slices. These data indicate that GJ play a role in synchronizing human neocortical networks and may implement epileptiform activity in FCD.