Region- and pattern-specific effects of glutamate uptake blockers on epileptiform activity in rat brain slices.

Many epileptic syndromes develop into pharmaco-resistant forms, calling for the development of new anticonvulsant strategies. The transmitter glutamate serves a double role as excitatory transmitter and as precursor for GABA, thus interfering with glutamate uptake may therefore exert complex effects on excitation-inhibition-balance in epileptic networks. In the present study we tested the effect of two different glutamate uptake blockers on acutely induced epileptiform activity in hippocampal-entorhinal cortex slices from adult rats: dihydrokainate (DHK) which blocks predominantly glial glutamate uptake, and threo-beta-benzyloxyaspartic acid (TBOA) which blocks both glial and neuronal glutamate uptake. Three different models were used to induce epileptiform discharges: (i) increasing NMDA receptor-mediated excitation by omitting Mg(2+)-ions; (ii) blocking potassium channels by 4-aminopyridine; (iii) reducing GABA(A) receptor-mediated inhibition by penicillin. Application of DHK or TBOA markedly reduced the frequency of epileptiform discharges in CA1 in the low magnesium and the 4-AP model while pathological activity was increased in the penicillin-model. In contrast, frequency of epileptiform discharges in EC was consistently increased by DHK and TBOA. Effects of DHK were more easily reversible than those of TBOA. Thus glutamate uptake blockers exert variable effects on epileptiform activity, depending on brain region and on the mechanism of ictogenesis.

Age-related decline of functional inhibition in rat cortex.

Recent studies with functional magnetic brain imaging showed different task-related patterns of brain activation and deactivation in aged as compared to young healthy subjects. We hypothesized that these changes of brain activation patterns might be due to age-dependent changes of neuronal excitability. Therefore, we experimentally studied the functional cortical inhibition by paired pulse stimulation in brain slices of young adult (3 months), aged adult (24 months) and old (36 months) male rats. Field potentials were evoked by application of double pulses at layer VI/white matter and recorded in layer II/III. We also analyzed the regional distribution of five major gamma-aminobutyric acid A (GABA(A)) receptor subunits (alpha1, alpha2, alpha3, alpha5, and gamma2) by immunohistochemistry. A reduced functional inhibition in aged as compared to young animals associated with an altered composition of GABA(A)-receptors, especially a reduction of subunit alpha5 in aged animals, was shown. The present study suggests that the age-dependent functional activation patterns and possibly also the cognitive and motor abilities are at least partially modulated by an age-dependent alteration of functional inhibition in the neocortex.

Differential alterations of the inactivation properties of high voltage activated calcium currents in area CA1 and CA3 of the rat following photothrombotic lesion.

There are many indications that focal brain ischemia may alter the properties of remote brain tissue. We investigated whether changes of neuronal properties can be observed in the unlesioned ipsilateral hippocampus following cortical photothrombosis in the somatosensory cortex of rats. The whole-cell patch-clamp technique was used to investigate calcium current properties of hippocampal neurons (CA1 and CA3) 7 days after infarct induction. A significant alteration in the half-maximal potential of inactivation (V(h,i)) could be demonstrated, when comparing lesioned with sham operated animals, while other current parameters remained unchanged. The alterations of the V(h,i) in the CA1 and CA3 regions were of opposite directions: V(h,i) in CA1 neurons was shifted negatively by 5.6 mV, and positively by 5.0 mV in neurons from the CA3 region. It has been speculated that these differential alterations may be due to different subunit compositions of calcium channels in these two brain areas. The data indicate that small cortical lesions can lead to widespread alterations of the neuronal network's excitability in the hippocampal formation.

Relation between bicarbonate concentration and voltage dependence of sodium currents in freshly isolated CA1 neurons of the rat.

It recently has been shown that whole cell calcium and sodium currents are modulated by CO(2)/HCO(3)(-)-buffered saline. While the bicarbonate ion, but not CO(2), has been proven to modulate calcium currents, this information is lacking for sodium currents. Furthermore, it is not known whether the strength of modulation dependents on the bicarbonate concentration or whether it is an all-or-nothing phenomenon. To answer these questions, we used the whole cell voltage-clamp technique on freshly isolated hippocampal CA1 neurons from the rat. A voltage step from -130 to -20 mV elicited a sodium current with an amplitude of -5.1 +/- 0.5 nA (mean +/- SE, n = 17) when cells were superfused with HEPES-buffered saline. The amplitude of this current increased during a subsequent superfusion with solutions containing increasing amounts of bicarbonate and CO(2) (%CO(2)/mM HCO(3)(-): 2.5/5.6; 5.0/18; 10/37), with a maximal increment in 10% CO(2)/37 mM HCO(3)(-) of -6.9 +/- 0.8 nA. The increase in amplitude was associated with a linear negative shift (slope: -0.7 mV/mM HCO(3)(-)) of the potential of half-maximal activation (DeltaV(h,a): -19.4 +/- 1.8 mV in 10% CO(2)) but not with an alteration in the maximal conductance (g(max): HEPES: 203.1 +/- 21.0 nS and 10% CO(2)/37 mM HCO(3)(-): 207.3 +/- 21.3 nS). In addition, the potential of half-maximal inactivation (V(h,i)) shifted to more negative potentials (slope: -0.6 mV/mM HCO(3)(-)) with increasing amounts of bicarbonate and CO(2) (HEPES: -53.6 +/- 11.8 mV; 10% CO(2)/37 mM HCO(3)(-): -69.8 +/- 2.1 mV), making the amplitude of the current highly sensitive for small potential changes at resting membrane potential. The same negative shift in voltage dependence arose when cells were exposed to solutions with different amounts of bicarbonate (5.6; 18; 26 mM) but constant CO(2) (5%) with slope rates of -0.5 mV/mM HCO(3)(-) for V(h,a) and -0.5 mV/mM HCO(3)(-) for V(h,i). Again, there was no correlation between bicarbonate concentration and the size of g(max). When currents were evoked in solutions containing a constant concentration (18 mM) of bicarbonate but different amounts of CO(2) (2.5; 5.0 10%), no significant changes have been observed. The present data demonstrate that bicarbonate ions, and not CO(2), modulate voltage-gated sodium currents in a concentration-dependent manner. Because the amplitude of the sodium current becomes highly sensitive to membrane potential changes concomitant with increased bicarbonate amounts, this may be critical for the excitability of the neuronal network in situations (like metabolic acidosis, respiratoric alkalosis and hypercapnia) in which the concentration of this ion can alter.

Metabolic and electrophysiological alterations in an animal model of neocortical neuronal migration disorder.

Cortical migration disorders are a major cause for intractable epilepsy syndromes. High resolution MRI and PET are increasingly capable to identify cortical dysgenesis. In this study we used the rat freeze lesion model to investigate cortical morphological and functional changes in adult rats after induction of a cortical freeze lesion at postnatal day (p) 0. Autoradiographic measurements of basic cortical [14C]deoxyglucose metabolism showed a significant reduction up to 1 mm lateral to the lesion but no remote changes. Electrophysiological in vitro recordings revealed a significant reduction in the amplitude of stimulus-evoked field potential responses recorded lateral to the lesion as compared to medial recording sites. Our data provide further evidence that spatially restricted developmental alterations of cortical morphology cause functional changes in surrounding and histologically normal areas that need to be considered for a better understanding of the resulting pathophysiology.

Concentration dependence of bicarbonate-induced calcium current modulation.

High-voltage-activated calcium currents (HVA) of CA1 neurons are prominently attenuated following a switch from HEPES-buffered solution to one buffered with CO(2)/HCO(3)(-). In the present study we investigated whether bicarbonate ions or the dissolved CO(2) induce this alteration in current characteristic. The study was carried out on freshly isolated CA1 neurons using the whole cell patch-clamp technique. Maximal calcium conductance and the mean peak amplitude of the currents showed a concentration-dependent decrease when cells were consecutively bathed in solutions containing increasing amounts of bicarbonate and CO(2). This decrease is best described by the Hill equation, yielding a maximal attenuation of 69%, a half-maximal concentration (EC(50)) of 7.4 mM HCO(3-), and a Hill coefficient of 1.8. In parallel, the potentials of half-maximal activation (V(h,a)) and inactivation (V(h,i)) were linearly shifted in hyperpolarizing direction with a maximal shift, in the 10% CO(2)/37 mM HCO(3)(-) containing solution of 10 +/- 1 mV for V(h,a) (n = 23) and 17 +/- 1.4 mV for V(h,i) (n = 18). When currents were evoked in solutions containing equal concentrations of bicarbonate but different amounts of CO(2), only nonsignificant changes were observed, while marked alterations of the currents were induced when bicarbonate was changed and CO(2) held stable. The experiments suggest that bicarbonate is the modulating agent and not CO(2). This bicarbonate-induced modulation may be of critical relevance for the excitation level of the CNS under pathological situation with altered concentration of this ion, such as hyperventilation and metabolic acidosis.

Enhancement of whole cell calcium currents following transient MCAO.

Cerebral infarctions have been shown to cause widespread changes of neuronal excitability in non-infarcted tissue. Calcium currents are major determinants of neuronal behavior, and pathological modulation of Ca(2+)-channels is known to lead to altered excitability states in a variety of paradigms. In the present study we addressed the question to what extent whole cell calcium currents are altered after middle cerebral artery occlusion (MCAO) in both the ipsi- and contralateral sensory cortex. Transient middle cerebral artery occlusion was induced for 1 h in rats using the intraluminal thread model. After 7 or 28 days survival, whole cell patch clamp studies were carried out on freshly isolated neurons of the ipsi- and contralateral sensory cortex, and high voltage activated (HVA) calcium currents were examined. In lesioned animals, we found a significant increase of calcium current amplitude and maximal conductance in the sensory cortex contralateral to the infarcts. This was paralleled by a prominent positive shift of the potential of half-maximal activation (V(h,a)) in these cells. Changes were long-lasting and at least stable for the following 28 days. These alterations were present in animals with lesions of moderate size, but not in those with massive infarction, and only in the cortex contralateral to the lesion. Following cortical infarctions, changes of calcium current properties are selectively observed in neurons contralateral to the lesion. At the behavioral level, compensatory mechanisms involving the unaffected hemisphere may induce this alteration of calcium current properties.

Enlargement of cortical vibrissa representation in the surround of an ischemic cortical lesion.

It has been shown that cortical lesions are associated with an increase of excitability in surrounding brain regions, and with a downregulation of GABA(A) receptors. In the present study we investigated whether this increased excitability affects the cortical map of inputs represented in areas surrounding the lesioned brain area. Focal lesions with a diameter of 2-2.5 mm were induced photochemically in the hindlimb area at the border of the primary somatosensory cortex of the rat. One week after lesioning, the cortical representation of the B3 vibrissa was studied using 14C-deoxyglucose (DG) autoradiography. In all animals mechanical stimulation of the B3 vibrissa produced a column-shaped DG-labeling in the somatosensory cortex, corresponding to the B3-barrel with a maximum of the glucose uptake in layer IV. In control animals without cortical lesions (n=6), stimulation increased the glucose uptake rate by 50.8+/-10.5% in layer IV. In lesioned animals (n=6) maximum DG-uptake in layer IV (54.8+/-8.6%) did not differ significantly from that in controls. However, as compared to control animals, lesioned animals showed also increased glucose uptake within the activated column in layers II/II (51.+/-11.1%, lesioned animals; 31.8+/-11.2%, controls; P<0.05, lesioned vs. control) and V (47.5+/-5.8%, lesioned animals, 28.8+/-10.5%, controls; P<0.05, lesioned vs. control). The diameter of the metabolically activated B3-barrel area of layer IV was expanded from 461.8+/-77.6 microm in control animals to 785.5+/-103.6 microm; P<0.01) in lesioned animals. Lesioned animals also showed expansion of the activated area in layers II/III (890.4+/-134.8 microm, lesioned animals; 430.6+/-95.1 microm, controls; P<0.01) and layer V (1117.5+/-163.6 microm, lesioned animals; 648.7+/-114.1 microm, controls; P<0.01). The depth profile of the activation columns showed a maximum in layer IV in control animals, which was expanded towards layers II/III and layer V in lesioned animals. It is concluded that cortical lesions alter the representational area of neighboring afferent inputs through disinhibition or 'unmasking' of pre-existing silent or ineffectual intracortical synapses. The present observations raise the possibility that the brain supports recovery from lesions by decreasing GABAergic inhibition, thereby facilitating a remapping of the cortical representation in neighboring brain areas.

Electrophysiological transcortical diaschisis after middle cerebral artery occlusion (MCAO) in rats.

Remote changes in brain function following stroke are called diaschisis. These remote effects may contribute to the neurological deficit following brain infarction; in addition they may lead to post-stroke epilepsy and affect functional recovery. In the present study we addressed the question of whether an increase in excitability can be observed contralateral to middle cerebral artery (MCA) infarction. Permanent occlusion of the middle cerebral artery (MCAO) was induced experimentally in rats with an intraluminal silicon-coated filament. Seven days later, brain excitability was tested with extracellulare recording techniques in neocortical coronal brain slices using a paired-pulse stimulus protocol. In rats with MCAO, excitability was increased in the neocortex contralateral to the infarction compared with the control group. These alterations extended through wide parts of the contralateral neocortex. The study demonstrates that MCAO causes transcallosal electrophysiological diaschisis. Together with results obtained previously with photothrombotic cortical lesions, it can be concluded that these remote effects are not due to characteristics of the individual lesion model, but are common consequences of brain lesions.

Calcium currents in acutely isolated stellate and pyramidal neurons of rat entorhinal cortex.

Calcium currents were studied in morphologically identified pyramidal and stellate neurons acutely isolated from layer II/III of rat entorhinal cortex, using the whole-cell patch-clamp configuration. The peak amplitude of high-voltage activated current (HVA) measured at +10 mV was not different in both neuron populations with 0.94+/-0.08 nA for pyramidal and 1.03+/-0.08 nA for stellate cells. Stellate neurons had a larger capacitance (14.4+/-1. 1 pF) than pyramidal neurons (9.6+/-0.8 pF), indicating a 50% larger cell surface. Most striking was the difference between the current density in stellate (79+/-8 pA/pF) versus pyramidal neurons (113+/-13 pA/pF). The potential of half maximal inactivation was not different: -37+/-2 mV (pyramidals) and -37+/-3 mV (stellates). Half of the cells contained a low-voltage activated calcium current (LVA) with a peak amplitude that was twice as large in stellate as in pyramidal neurons (0.21+/-0.04 nA resp. 0.11+/-0.03 nA; at -50 mV). In contrast to the HVA component, the current density of the LVA component was not different between cell types (13+/-3 pA/pF vs. 13+/-2 pA/pF). This implies that the relative abundance of LVA and HVA currents in stellate and pyramidal neurons is different which could result in different firing characteristics. The potential of half maximal LVA inactivation was -88+/-4 mV (pyramidals) and -85+/-3 mV (stellates). The slope of the voltage dependent steady state inactivation was steeper in stellate (7+/-1 mV) than in pyramidal cells (10+/-2 mV).

Effects of tetanus toxin on functional inhibition after injection in separate cortical areas in rat.

Tetanus Toxin is widely used as a model of chronic focal epilepsy and is assumed to act by blocking neurotransmitter release with high selectivity for inhibitory synapses. However, the exact mechanisms are not fully understood, since, e.g., GABA release is only temporarily decreased although epileptiform activity persists pointing towards a change in the interplay of excitation and inhibition. Furthermore there have been reports about different effects of tetanus toxin after injection in separate brain areas. Therefore, we investigated the functional inhibition after injecting tetanus toxin either in the motor or sensory cortex of adult rats by using a paired-pulse paradigm as a measure of excitatory and inhibitory drive. Tetanus toxin injection into the motor cortex (n=10) induced a marked, long-lasting reduction in inhibition which was highly significant in most parts of the injected cortical area. Injections into the sensory cortex, however, showed less marked changes in inhibition which were more widespread and significant only in 3 of 14 animals injected. These results give further evidence for a prominent effect of tetanus toxin on functional inhibition and strengthen the idea of a differential effect in separate cortical areas. They may be accounted for by the different cytoarchitecture of cortical areas with variable inhibitory and excitatory intracortical connections.

Thalamocortical circuits causing remote hypometabolism during focal interictal epilepsy.

The functional circuit causing depression of cerebral glucose metabolism in brain areas remote from an epileptic focus was investigated in experiments on the cortex of the rat. Epileptic activity was induced by direct epicortical application of Na-penicillin onto the motor cortical area Fr1/Fr2. The increased neuronal activity was associated with an increase of metabolism in the focal area and a decrease in somatosensory cortical areas. Metabolism was also massively increased in the thalamus, predominantly in the posterior nucleus. Stereotactic radiofrequency lesioning of this nucleus, 30 days prior to the induction of the epileptic focus, restricted the area with increase of metabolism to the upper cortical laminae, and abolished the cortical hypometabolism in the sensory cortex. It is suggested that the primary functional circuit affected by the acute epileptic focus in the present model consists of the motor cortex, the thalamic nucleus posterior and the somatosensory cortex.

Uncoupling of blood flow and metabolism in focal epilepsy.

PURPOSE: Interictal measurements of cerebral blood flow are less helpful in localizing epileptic foci than are measurements of brain metabolism. This may be related to an uncoupling of blood flow and metabolism. In this study, brain metabolism and blood flow were compared in an acute experimental model of focal interictal epilepsy. METHODS: Interictal epileptic foci were induced by an epicortical application of penicillin in rats. After 1 h, stereotyped interictal activity was initiated, lasting until the end of the experiment. Brain metabolism was determined with [14C]deoxyglucose, and cerebral blood flow with [14C]iodoan-tipyrine autoradiography. RESULTS: In control experiments, metabolism and blood flow were coupled. In animals with focal interictal epileptic activity, the metabolism was strongly increased in the focus and reduced in areas lateral to the focus. In contralateral brain areas, blood flow and metabolism varied in a parallel fashion. Ipsilateral to the focus, however, blood flow and metabolism were altered disproportionately. In the focus, the increase of blood flow was less marked than the increase of metabolism, and the area with increased blood flow was larger than the area with increased metabolism. Lateral to the focus, in the area with a hypometabolism, blood flow was not concomitantly reduced. CONCLUSIONS: The experiments show that blood flow and metabolism in focal epilepsy may be uncoupled in widespread regions. This is due neither to structural abnormalities nor to the duration or discharge pattern of epileptic activity. The results explain why interictal metabolic investigations have a higher predictive value in presurgical epilepsy evaluation than do interictal measurements of blood flow.

Brain hypometabolism in a model of chronic focal epilepsy in rat neocortex.

PURPOSE: Metabolic mapping of the human brain has become a widely used method for identifying and localizing epileptic foci. A reduction of glucose consumption usually is found interictally in the area of the focus. By contrast, animal models of acute epilepsy show a hypermetabolism in the epileptic focus. Here we investigated how metabolism is altered in an animal model of chronic epilepsy caused by focal injection of tetanus toxin into rat neocortex. METHODS: A total of 27 male Wistar rats were anesthetized and injected into the motor or sensory cortex either with dissolved tetanus toxin or with the solvent only. Animals recovered for 7, 14, or 30 days and then were anesthetized again for quantitative 14C-deoxyglucose autoradiography. Data were analyzed with an imaging program, and regional cerebral glucose metabolism (rCMRGlc) was determined. RESULTS: Injection of tetanus toxin into the motor cortex caused a focal hypometabolism which was confined to the cytoarchitectonic boundaries of the injected area, whereas sensory cortex injection caused a more widespread hypometabolism in all sensory cortical and connected, areas. None of the animals displayed focal hypermetabolism and we observed no significant time-dependent alteration of brain metabolism. CONCLUSIONS: Tetanus toxin injection into the cortex of the rat induces chronic epileptic activity accompanied by a focal hypometabolism. The data suggest that the spread of the metabolic alterations depends on the connectivity of the injected cortical area.

Induction of FOS and JUN proteins during focal epilepsy: congruences with and differences to [14C]deoxyglucose metabolism.

fos and jun belong to multigene families coding for transcription factors. These cellular immediate-early genes (IEGs) are thought to be involved in coupling neuronal excitation to changes of target gene expression. Immunocytochemistry with specific antisera was used to assess regional levels of five IEG-encoded proteins (c-FOS, FOS B, c-JUN, JUN B and JUN D) in a rat model of penicillin-induced focal epilepsy. To assess whether brain regions with post-ictal de novo transcription factor synthesis correspond to those areas with increased glucose metabolism, IEG expression patterns were compared with [14C]deoxyglucose autoradiography performed in a subset of animals. The results demonstrated marked induction of c-FOS, FOS B, c-JUN and JUN B but not JUN D in the cortical epileptic focus. Thereby, individual IEG-encoded proteins exhibited differential temporal and spatial expression patterns. Within the epileptic focus, IEG expression correlated with increased glucose metabolism. In contrast, IEG induction was not observed in brain areas distant from the epileptic focus that also demonstrated increased glucose metabolism, such as homotopic contralateral motor cortex and ipsilateral thalamic nuclei. These findings indicate that in focal epilepsy changes of the genetic programme are restricted to neurons of the epileptic focus. In contrast, the increased [14C]deoxyglucose metabolism in contralateral motor cortex and ipsilateral thalamus seems to indicate functional changes.

Coupling of cortical and thalamic metabolism in experimentally induced visual and somatosensory focal epilepsy.

Focal epileptic activity induces widespread metabolic disturbances beyond the area of the electroencephalographically detectable focus. In order to find out whether the metabolic coupling between the epileptic focus and other brain regions depends on the localization of the focus, two groups of rats with epileptic foci at different sites were investigated. In the first group acute epileptic activity was induced by application of penicillin to the secondary visual cortex (Oc2), and in the second group to the primary somatosensory cortex (Par1). Metabolism was analyzed using the [14C]deoxyglucose autoradiographic method. In both groups of animals, hypermetabolism in the area of the focus and in specific functionally coupled thalamic nuclei was observed. Focal epileptic activity in the secondary visual cortex induced significant hypometabolism in remote ipsilateral cortical areas. In rats with epileptic foci in the primary somatosensory cortex hypometabolism in extrafocal ipsilateral cortical areas was less prominent. These findings provide further support for the integral involvement of the thalamus in modulating metabolism in remote cortical brain regions during focal epileptic activity. The extent of metabolic alterations may depend on the site of the epileptic focus and the connectivity of the recruited thalamic nuclei.

Polyunsaturated fatty acids modulate sodium and calcium currents in CA1 neurons.

Recent evidence indicates that long-chain polyunsaturated fatty acids (PUFAs) can prevent cardiac arrhythmias by a reduction of cardiomyocyte excitability. This was shown to be due to a modulation of the voltage-dependent inactivation of both sodium (INa) and calcium (ICa) currents. To establish whether PUFAs also regulate neuronal excitability, the effects of PUFAs on INa and ICa were assessed in CA1 neurons freshly isolated from the rat hippocampus. Extracellular application of PUFAs produced a concentration-dependent shift of the voltage dependence of inactivation of both INa and ICa to more hyperpolarized potentials. Consequently, they accelerated the inactivation and retarded the recovery from inactivation. The EC50 for the shift of the INa steady-state inactivation curve was 2.1 +/- 0.4 microM for docosahexaenoic acid (DHA) and 4 +/- 0.4 microM for eicosapentaenoic acid (EPA). The EC50 for the shift on the ICa inactivation curve was 2.1 +/- 0.4 for DHA and > 15 microM for EPA. Additionally, DHA and EPA suppressed both INa and ICa amplitude at concentrations > 10 microM. PUFAs did not affect the voltage dependence of activation. The monounsaturated oleic acid and the saturated palmitic acid were virtually ineffective. The combined effects of the PUFAs on INa and ICa may reduce neuronal excitability and may exert anticonvulsive effects in vivo.

Neuronal hyperexcitability and reduction of GABAA-receptor expression in the surround of cerebral photothrombosis.

Changes of neuronal excitability and gamma-aminobutyric acid (GABAA)-receptor expression were studied in the surround of photothrombotic infarcts, which were produced in the sensorimotor cortex of the rat by using the rose bengal technique. In a first series of experiments, multiunit recordings were performed on anesthetized animals 2-3 mm lateral from the lesion. Mean discharge frequency was considerably higher in recordings from lesioned animals (> 100 Hz in the first postlesional week) compared with control animals (mean, 15 Hz). These alterations were already present after 1 day but were most pronounced 3 to 7 days after lesion induction. Thereafter the hyperexcitability declined again, although it remained visible up to 4 months. In a second series of experiments, the GABAA-receptor expression was studied autoradiographically. This revealed a reduction of GABAA receptors in widespread brain areas ipsilateral to the lesion. The reduction was most pronounced in the first days after lesion induction and declined with longer intervals. It is concluded that cortical infarction due to photothrombosis leads to a long-lasting and widespread reduction of GABAA-receptor expression in the surround of the lesion, which is associated with an increased neuronal excitability. Such alterations may be responsible for epileptic seizures that can be observed in some patients after stroke and may contribute to neurologic deficits after stroke.