Gabapentin Prevents Progressive Increases in Excitatory Connectivity and Epileptogenesis Following Neocortical Trauma.

Neocortical injury initiates a cascade of events, some of which result in maladaptive epileptogenic reorganization of surviving neural circuits. Research focused on molecular and organizational changes that occur following trauma may reveal processes that underlie human post-traumatic epilepsy (PTE), a common and unfortunate consequence of traumatic brain injury. The latency between injury and development of PTE provides an opportunity for prophylactic intervention, once the key underlying mechanisms are understood. In rodent neocortex, injury to pyramidal neurons promotes axonal sprouting, resulting in increased excitatory circuitry that is one important factor promoting epileptogenesis. We used laser-scanning photostimulation of caged glutamate and whole-cell recordings in in vitro slices from injured neocortex to assess formation of new excitatory synapses, a process known to rely on astrocyte-secreted thrombospondins (TSPs), and to map the distribution of maladaptive circuit reorganization. We show that this reorganization is centered principally in layer V and associated with development of epileptiform activity. Short-term blockade of the synaptogenic effects of astrocyte-secreted TSPs with gabapentin (GBP) after injury suppresses the new excitatory connectivity and epileptogenesis for at least 2 weeks. Results reveal that aberrant circuit rewiring is progressive in vivo and provide further rationale for prophylactic anti-epileptogenic use of gabapentinoids following cortical trauma.

Dendritic mechanisms underlying penicillin-induced epileptiform activity.

The action of penicillin on synaptically evoked dendritic activity was examined with the use of hippocampal slice preparations. Orthodormic activation of CA1 pyramidal neurons produced an excitatory-inhibitory postsynaptic potential sequence recorded intracellularly in the dendrites. Treatment with penicillin resulted in the appearance of spontaneous and synaptically evoked multipeaked field potentials and associated depolarization shifts and spike burst generation in CA1 cells. Intracellular recordings revealed that penicillin produced no detectable change in passive membrane properties of the postsynaptic dendrites. However, the inhibitory postsynaptic potential was suppressed by penicillin, resulting in the release of intrinsic dendritic burst firing during synaptic activation. These findings emphasize the role of normal patterns of dendritic burst generation in the production of intense neuronal discharge during penicillin-induced epileptiform activities.

Dye coupling and possible electrotonic coupling in the guinea pig neocortical slice.

Iontophoretic injection of the fluorescent dye Lucifer Yellow CH into single neurons of guinea pig neocortical slices resulted in the staining of more than one cell. Dye-coupled neuronal aggregates were found only in the superficial cortical layers and were often organized in vertical columns. Antidromic stimuli evoked all-or-none, subthreshold depolarizations in some superficial cells. These potentials were not eliminated by manganese and did not collide with spikes originating in the soma, suggesting that they arose from electrotonic interaction between superficial cortical neurons.

Cholinergic switching within neocortical inhibitory networks.

Differential actions of acetylcholine on the excitability of two subtypes of interneurons in layer V of the rat visual cortex were examined. Acetylcholine excited low-threshold spike (LTS) cells through nicotinic receptors, whereas it elicited hyperpolarization in fast spiking (FS) cells through muscarinic receptors. Axons of LTS cells were mainly distributed vertically to upper layers, and those of FS cells were primarily confined to layer V. Thus, cortical cholinergic activation may reduce some forms of intralaminar inhibition, promote intracolumnar inhibition, and change the direction of information flow within cortical circuits.

Thalamocortical relay neurons: antidromic invasion of spikes from a cortical epileptogenic focus.

Thalamocortical relay neurons whose axons project into a penicillininduced cortical epileptogenic focus generate bursts of action potentials during spontaneous interictal epileptiform discharges. These bursts originate in intracortical axons and propagate antidromically into thalamic neurons. Repetitive spike generation in cortical axons and presynaptic terminals could produce a potent excitatory drive and contribute to the generation of the large depolarization shifts which are seen in cortical elements during focal epileptogenesis.

Antiepileptogenic repair of excitatory and inhibitory synaptic connectivity after neocortical trauma.

The "final common path" to epileptogenesis induced by cortical trauma and disease processes ultimately depends on changes in relative weights of excitatory and inhibitory synaptic activities in neuronal networks. Results of two experiments summarized here provide proof in principle that prophylaxis of posttraumatic epileptogenesis can result when antiepileptogenic treatments are focused on basic underlying synaptic mechanisms. (1) Brief gabapentin treatment after injury limits new excitatory synapse formation by preventing binding of thrombospondins to α2δ-1 receptors, resulting in long-lasting effects that limit aberrant excitatory connectivity and decrease epileptogenesis. (2) Fast-spiking (FS) interneurons are structurally and functionally abnormal in the partial cortical isolation and other models of epileptogenesis. Brain-derived neurotrophic factor (BDNF) supports growth and maintenance of GABAergic neurons during brain development, leading to the hypothesis that it might favorably affect injured interneurons. Partial activation of BDNF TrkB receptors with a small molecule reverses structural abnormalities in FS interneuronal terminals, increases the frequency of mIPSCs, and increases probability of GABA release. These changes are associated with significantly reduced spontaneous and evoked epileptiform bursts in vitro and increased threshold for pentylenetetrazole-induced seizures in vivo. Each of these treatments offers a potential promising approach to prophylaxis of injury-induced cortical epileptogenesis.

Heterogeneity of rat corticospinal neurons.

In order to examine the degree of diversity within a population of cortical projection neurons, rat corticospinal cells were retrogradely labeled in vivo by injecting rhodamine-tagged microspheres into the cervical spinal cord, and subsequently studied electrophysiologically and anatomically in neocortical slices maintained in vitro, by use of standard current clamp techniques and a double-labeling protocol (Tseng et al., J. Neurosci. Meth. 37:121-131, 1991). Three different subgroups were distinguished on the basis of their spiking behavior: (1) Adapting cells had a marked fast (50 ms) and slow phase (200 ms) of spike frequency adaptation; (2) regular spiking (RS) cells had only a period of fast adaptation; (3) some regular spiking neurons had prominent depolarizing afterpotentials (DAPs) and could generate bursts of spikes, often in repetitive fashion (RSDAP cells). Subgroups of RSDAP cells had different patterns of burst responses to depolarizing current pulses, suggesting differences in the types and/or sites of underlying ionic conductances. Adapting cells had a slightly higher membrane input resistance and more prominent slow hyperpolarizing afterpotentials than RS and RSDAP neurons; however, the activation of presumed anomalous rectifier current by intracellular hyperpolarizations was less prominent in adapting neurons. Orthodromic stimulation in layer I evoked presumed excitatory and inhibitory postsynaptic potentials (EPSPs and IPSPs)in all three types of cells, but prominent short-latency IPSPs were found in a higher percentage of adapting neurons. The morphology of electrophysiologically characterized corticospinal neurons was studied following intracellular injection of biocytin. All three spiking types were typical layer V pyramids with apical dendrites reaching layer 1, basal dendrites in infragranular layers, and deep-directed axons that had a moderate density of local collaterals in lower cortical layers. The profuseness of dendrites, examined by Sholl's analysis of two-dimensional, camera lucida-reconstructed neurons was comparable in the three neuronal subgroups, although a smaller somatic area and more slender apical dendritic trunk were found in adapting neurons. Our results suggest that corticospinal cells in rats are a heterogeneous population of projection neurons with respect to their spiking behavior, membrane properties, synaptic connections, and, to a lesser extent, their morphology. This diversity revealed in vitro adds new complexity to the classification of corticospinal neurons.

Burst generating and regular spiking layer 5 pyramidal neurons of rat neocortex have different morphological features.

Intracellular recordings were obtained from pyramidal neurons in layer 5 of rat somatosensory and visual cortical slices maintained in vitro. When directly depolarized, one subclass of pyramidal neurons had the capacity to generate intrinsic burst discharges and another generated regular trains of single spikes. Burst responses were triggered in an all-or-none manner from depolarizing afterpotentials in most bursting neurons. Regular spiking cells responded to electrical stimulation of ascending afferents with a typical EPSP-IPSP sequence, whereas IPSPs were hard to detect in bursting cells. Orthodromic activation of the latter evoked a prominent voltage-dependent depolarization that could trigger a burst response. Intracellularly labelled bursting and regular spiking cells were located in layer 5b, but had distinctly different morphologies. Bursting neurons had a large pyramidal soma, a gradually emerging apical dendrite, and an extensive apical and basal dendritic tree. Their axonal collateral arborization was predominantly limited to layers 5/6. In contrast, regular spiking cells had a more rounded soma with abruptly emerging apical dendrite, a smaller dendritic arborization, and 2 to 8 ascending axonal collaterals that arborized widely in the supragranular layers. Both bursting and regular spiking cells had main axons that entered the subcortical white matter. These data show that some subgroups of pyramidal neurons within the deeper parts of layer 5 of rat cortex are morphologically and physiologically distinct and have different intracortical connections. Bursting cells presumably function to amplify and synchronize cortical outputs, whereas regular spiking output neurons provide excitatory feedback to neurons at all cortical levels and receive a more effective orthodromic inhibitory input. These data support the hypothesis that differences in gross neuronal structure, perhaps even the subtle differences that distinguish subclasses of neurons in a given lamina, are predictive of underlying differences in the type and distribution of ion channels in the nerve cell membrane and connections of cells within the cortical circuit.

Heterogeneous axonal arborizations of rat thalamic reticular neurons in the ventrobasal nucleus.

The gamma-aminobutyric acid (GABA)-containing neurons of the thalamic reticular nucleus (nRt) are a major source of inhibitory innervation in dorsal thalamic nuclei. Individual nRt neurons were intracellularly recorded and labelled in an in vitro rat thalamic slice preparation to investigate their projection into ventrobasal thalamic nuclei (VB). Camera lucida reconstructions of 37 neurons indicated that nRt innervation ranges from a compact, focal projection to a widespread, diffuse projection encompassing large areas of VB. The main axons of 65% of the cells gave rise to intra-nRt collaterals prior to leaving the nucleus and, once within VB, ramified into one of three branching patterns: cluster, intermediate, and diffuse. The cluster arborization encompassed a focal region averaging approximately 25,000 mu m2 and contained a high density of axonal swellings, indicative of a topographic projection. The intermediate structure extended across an area approximately fourfold greater and also contained numerous axonal swellings. The diffuse arborization of nRt neurons covered a large region of VB and contained a relatively low density of axonal swellings. Analysis of somatic size and shape revealed that diffuse arborizations arose from significantly smaller, fusiform-shaped somata. Cytochrome oxidase reactivity or parvalbumin immunoreactivity was used to delineate a discontinuous staining pattern representing thalamic barreloids. The size of a cluster arborization closely approximated that of an individual barreloid. The heterogeneous arborizations from nRt neurons may reflect a dynamic range of inhibitory influences of nRt on dorsal thalamic activity.

The role of extracellular potassium in hippocampal epilepsy.

It has been proposed that the notable capacity for epileptogenesis in the hippocampus may be related to potassium accumulation in extracellular spaces. To investigate this hypothesis more directly, we measured changes in extracellular potassium concentration ([K+]o) during focal hippocampal epilepsy using potassium-sensitive microelectrodes. Interictal and ictal electroencephalographic events were accompanied by increases in [K+]o that varied systematically with depth from the ependymal surface and lateral distance from the focus. Maximal [K+]o changes during interictal and ictal discharges occurred in the stratum pyramidale. Initiation of ictal activity did not correlate with a particular "threshold" [K+]o. Comparing these results with similar data from neocortex, we observed that interictal K+ responses in hippocampus lasted longer and had slower rise times, and that peak interictal and ictal [K+]o values were consistently lower. Increases in [K+]o cannot be the sole explanation for regional variations in seizure susceptibility, interictal-ictal transitions, or termination of ictal episodes.

Changes of alpha-amino-3-hydroxy-5-methyl-4-isoxazole-propionate receptors in layer V of epileptogenic, chronically isolated rat neocortex.

In vivo chronic partial isolation of neocortical islands results in epileptogenesis that involves pyramidal neurons of layer V. To test whether an alteration in glutamate receptors might contribute to the epileptiform activity, we analysed the time-course of light microscopic changes in expression of alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionate receptors using subunit-specific antibodies. The isolation caused a rapid down-regulation of immunoreactivity for GluR1 and GluR2/3 subunits in deep layer V pyramidal neurons within the neocortical island which was evident 24h post-lesion, and within three days was reduced to about 40-60% of the control level. Many pyramidal cells in deep layer V completely lacked GluR2. Between one and four weeks of survival, down-regulation of GluR2/3 and GluR2 involved the majority of pyramidal layer V neurons, except for cells in the upper part of layer V, and those within narrow areas of all sub-laminae of layer V ("micro-islands"). Initial down-regulation was also observed one to three days post-lesion for subunits 1 and 2 of the N-methyl-D-aspartate receptor, but in contrast to GluR2/3 immunoreactivity, NMDAR2A/B immunoreactivity was enhanced three weeks post-lesion. The present data provide evidence for plastic changes in glutamate receptors in neurons of partially isolated neocortical island. A sub-population of layer V neurons remains relatively unaffected, and would presumably be capable of generating fast glutamatergic synaptic potentials necessary for the development of synchronous epileptiform activity.

Cholinergic pharmacology of mammalian hippocampal pyramidal cells.

Responses of CAl pyramidal cells to cholinergic compounds were recorded with intracellular microelectrodes in guinea-pig hippocampal slices. Perfusion of slices with medium containing the muscarinic antagonists atropine or scopolamine (10(-7)-10(-6)M) blocked all actions of acetylcholine. Properties of control neurons and those from separate populations of neurons impaled in slices exposed to muscarinic blocking agents were compared. 1-2 h of perfusion with atropine-containing media significantly decreased membrane input resistance from 37.6 +/- 8.7 (S.D.) M omega (n = 74) to 21.9 +/- 7.7 (S.D.) M omega (n = 24) without producing significant changes in membrane potential. Muscarinic antagonists also reduced or eliminated the anomalous inward rectification normally seen in hippocampal pyramidal neurons. Exposure of slices to 10(-5)-10(-6)M eserine for about 1 h produced changes in neuronal membrane input resistance and potential and slow after hyperpolarizations similar to those elicited by application of acetylcholine. Bethanechol mimicked the actions of acetylcholine but was effective at lower concentrations and had longer lasting effects on afterhyperpolarizations. Nicotine produced an excitatory response in only one of 7 neurons. These experiments demonstrate that the actions of acetylcholine on hippocampal CAl neurons result from interaction with muscarinic receptors. Acetylcholine has modulatory effects on cell membrane properties which may be mediated through tonic release mechanisms.

Light microscopic study of GluR1 and calbindin expression in interneurons of neocortical microgyral malformations.

Rat neocortex that has been injured on the first or second postnatal day (P0-1) develops an epileptogenic, aberrantly layered malformation called a microgyrus. To investigate the effects of this developmental plasticity on inhibitory interneurons, we studied a sub-population of GABAergic cells that co-express the alpha-amino-3-hydroxy-5-methylisoxazole-4-propionic acid receptor GluR1 subunit and the calcium-binding protein, calbindin (CB). Both malformed and control cortex of adult (P40-60) animals contained numerous interneurons double-stained for CB and GluR1. Immunoreactivity (IR) for CB was up-regulated in perikarya of interneurons within supragranular layers of control cortex between P12 and P40. However, in malformed adult (P40) cortex, CB-IR levels were significantly lower than in adult controls, and fell midway between levels in immature and adult control animals. Between P12 and P40, GluR1-IR was down-regulated in perikarya of interneurons in control cortex. Somatic GluR1-IR levels in malformed adult (P40) cortex were not different from adult controls. These neurons formed a dense plexus of highly GluR1-positive spiny dendrites within layer II. The dendritic plexus in the malformation was more intensely GluR1-immunoreactive than that in layer II of control cortex. This was due to apparent changes in thickness and length of dendrites, rather than to significant changes in the number of interneuronal perikarya in the microgyral cortex. Results indicate that the population of GluR1/CB-containing interneurons is spared in malformed microgyral cortex, but that these cells sustain lasting decreases in their somatic expression of calbindin and alterations of dendritic structure. Potential functional implications of these findings are discussed.

Differential effects of petit mal anticonvulsants and convulsants on thalamic neurones: calcium current reduction.

1. Succinimide derivatives can be either convulsant (tetramethylsuccinimide (TMS)), or anticonvulsant (ethosuximide (ES); alpha-methyl-alpha-phenylsuccinimide (MPS)). ES, an anticonvulsant succinimide, has previously been shown to block calcium currents of thalamic neurones, while the convulsant succinimide TMS blocks gamma-aminobutyric acid (GABA) responses in a similar fashion to the convulsant pentylenetetrazol (PTZ). 2. Using voltage-clamp techniques, we analysed the effects of the anticonvulsant succinimides ES and MPS and the convulsants TMS and PTZ on calcium currents of acutely isolated thalamic relay neurones of the rat. 3. MPS and ES reduced low-threshold calcium current (LTCC) in a voltage-dependent manner, without affecting steady-state inactivation. MPS was less potent than ES (IC50 of 1100 vs 200 microM) but greater in efficacy (100% maximal reduction vs 40% for ES). 4. PTZ had no effect on calcium currents, and TMS only reduced LTCC at very high concentrations, and did not occlude MPS effects when applied concurrently. 5. These results, which demonstrate that anticonvulsant, but not convulsant, succinimides block LTCC, provide additional support for the hypothesis that LTCC reduction is a mechanism of action of the anticonvulsant succinimides related to their effects in petit mal epilepsy.

Differential effects of petit mal anticonvulsants and convulsants on thalamic neurones: GABA current blockade.

1. Currents evoked by applications of gamma-aminobutyric acid (GABA) to acutely dissociated thalamic neurones were analysed by voltage-clamp techniques, and the effects of the anticonvulsant succinimides ethosuximide (ES) and alpha-methyl-alpha-phenylsuccinimide (MPS) and the convulsants tetramethylsuccinimide (TMS), picrotoxin, pentylenetetrazol (PTZ), and bicuculline methiodide were assessed. 2. TMS (1 microM-10 microM) reduced responses to iontophoretically applied GABA, as did picrotoxin (0.1-100 microM), PTZ (1-100 mM) and bicuculline (1-100 microM). 3. ES, in high concentrations (1-10 mM), reduced GABA responses to a lesser extent, and also occluded the reductions in GABA-evoked currents produced by TMS, picrotoxin, and PTZ. ES did not occlude the effects of bicuculline on GABA responses. Therefore, we propose that ES acts as a partial agonist at the picrotoxin GABA-blocking receptor. 4. MPS had no effect on GABA responses (at a concentration of 1 mM), and, like ES, occluded the GABA-blocking actions of TMS, apparently acting as a full antagonist. 5. The anticonvulsant actions of ES and MPS against TMS and PTZ-induced seizures may thus involve two independent mechanisms: (1) the occlusion of TMS and PTZ GABA-blocking effects; and (2) the previously described specific effect of ES and MPS on low-threshold calcium current of thalamic neurones. The latter cellular mechanism may be more closely related to petit mal anticonvulsant activity.

Double-labelling with rhodamine beads and biocytin: a technique for studying corticospinal and other projection neurons in vitro.

Corticospinal neurons retrogradely labelled with rhodamine-labelled latex microspheres (RLMs) in vivo were studied intracellularly in a slice preparation up to 13 months later with electrodes containing biocytin. The physiological properties of these double-labelled corticospinal neurons were indistinguishable from those of comparable neurons which were impaled with biocytin-containing electrodes without prior RLM-labelling, and neurons studied with potassium acetate-filled electrodes in similar areas. Thus, neither labelling with RLMs nor injection of biocytin affected neuronal properties. This important advantage of RLMs makes them suitable for prelabelling projection neurons in vivo for subsequent studies that take advantage of the versatility of a brain slice preparation. In addition to its lack of effects on neuronal properties, intracellular labelling with biocytin also provides high-quality morphological details ideal for anatomical analysis. The compatibility of retrograde labelling with RLMs and intracellular staining with biocytin make this a useful combined technique for tracking electrophysiological and anatomical changes in identified projection neurons over time.

Effects of TEA on hippocampal neurons.

The effects of the potassium channel blockers, tetraethylammonium (TEA) and tetramethylammonium (TMA) on behavior of hippocampal CA1 neurons were studied in the in vitro slice preparation. Intracellular injections of TEA produced marked spike broadening, increased input resistance, and a prolongation of a putative slow calcium-activated afterhyperpolization (AHP). Little effect was seen on synaptic potentials. Bursting activity induced by TEA in hippocampal neurons was qualitatively different from the epileptiform bursting induced by raising extracellular potassium concentrations or by including penicillin in the bathing medium. Effects of TMA were qualitatively similar to those of TEA, but much slower and less dramatic. The results suggested that there were at least two different potassium channels in the membranes of hippocampal CA1 neurons, and that both were important in determining normal neuronal activity. TEA appeared to preferentially block a fast, voltage-dependent potassium conductance (gk); however, it also modified the slow, calcium-activated gk. Although cell bursting activity may have resulted from blockade of a gk produced by TEA treatment, the epileptiform bursts produced by penicillin exposure appeared to be independent of this mechanism.

Human epileptic neurons studied in vitro.

The in vitro neocortical brain slice technique was used to study electrophysiological properties of neurons from brain biopsies in 10 patients undergoing neurosurgical treatment for a variety of conditions, including focal epilepsy. The principal finding was the occurrence of orthodromically evoked depolarization shifts (DSs) and burst discharges in a proportion of neurons in slices from epileptogenic cortex. These evoked depolarizations and bursts had a number of properties in common with those from experimental epileptogenic foci in neocortex, including large amplitude and prolonged duration; long and variable latencies; and all or none, threshold type behavior, dependent on the parameters of orthodromic stimulation. Also DSs could not be evoked by intracellular stimulation, or blocked by hyperpolarizing current pulses once they had been orthodromically evoked. Responses of DSs to current thus differed markedly from those of neurons in epileptogenic guinea pig hippocampal slices. The results of these experiments suggest that intracellular events in human neurons involved in epileptogenesis are similar in appearance to those in various animal models. Neurons in chronic epileptogenesis are similar in appearance to those in various animal models. Neurons in chronic epileptogenic foci retain some of their abnormal properties within brain slices maintained in vitro.

Cellular and field potential properties of epileptogenic hippocampal slices.

Epileptogenic activity was induced in hippocampal slices by addition of penicillin (2.0mM) to the bathing medium. Field potential epileptiform events were recorded and single cell bursts studied with intracellular electrodes. Epileptogenic activity was seen in areas CA1 and CA3 of the slice, with bursts in CA3 always leading CA1 bursts; a cut between CA1 and CA3 abolished spontaneous bursting in CA1 but not in CA3. Increased (Mg2+) and decreased (Ca2+) abolished epileptiform discharge, thus demonstrating its dependence on synaptic activity; burst occurrence was also sensitive to (K+). Measurements of single cell resting potentials, resistance, and time constant in CA1 cells revealed no difference between cells in normal medium and cells made epileptogenic by penicillin. Depolarization shifts in CA1 neurons during epileptogenesis did not behave like 'giant EPSPs' but rather were complexes to which depolarizing spike after-potentials, fast prepotentials, and underlying slow depolarizing events all contributed.