The NMDA receptor NR2B subunit contributes to epileptogenesis in human cortical dysplasia

Brain Research 1046 (2005) 10 – 23 www.elsevier.com/locate/brainres

Research report

The NMDA receptor NR2B subunit contributes to epileptogenesis in human cortical dysplasia Gabriel Mo¨ddela,d,*, Berit Jacobsona, Zhong Yinga, Damir Janigrob, William Bingamanb, Jorge Gonza´lez-Martı´nezb, Christoph Kellinghausa,d, Richard A. Praysonc, Imad M. Najma a Department of Neurology, The Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, OH 44195, USA Department of Neurosurgery, The Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, OH 44195, USA c Department of Pathology, The Cleveland Clinic Foundation, 9500 Euclid Avenue, Cleveland, OH 44195, USA d Klinik und Poliklinik fu¨r Neurologie, Universita¨tsklinikum Mu¨nster, Albert-Schweitzer-Strabe 33, 48149 Mu¨nster, Germany b

Accepted 15 March 2005 Available online 10 May 2005

Abstract Cortical dysplasia (CD) is often associated with pharmacoresistant epilepsy. Previous studies showed increased expression of the NMDA receptor subunit NR2B in dysplastic and epileptic human neocortex. We tested the hypothesis that differential increase of NR2B constitutes an epileptogenic mechanism in humans. Dysplastic neocortex and lateral temporal lobe regions resected for treatment of pharmacoresistant seizures were processed for electrophysiological, histological, and immunocytochemical studies. Assignment to the ‘‘dysplastic’’ (n = 8) and ‘‘non-dysplastic’’ (n = 8) groups was based on histology. Neurons in ‘‘dysplastic’’ samples differentially stained for NR2B. Western blot (n = 6) showed an immunoreactive band for NR2B in three out of four ‘‘dysplastic’’ samples. Epileptiform field potentials (EFP) were elicited in vitro by omission of magnesium from the bath. EFP in ‘‘dysplastic’’ slices were characterized by multiple afterdischarges, occurring at a significantly higher repetition rate than EFP in non-dysplastic slices. The NR2B-specific NMDA receptor inhibitor ifenprodil (10AM) suppressed EFP in dysplastic slices. In non-dysplastic slices, burst repetition rate did not change with ifenprodil application. In both dysplastic and non-dysplastic slices, EFP were suppressed by a non-specific NMDAR antagonist (APV) or AMPA receptor antagonist (CNQX). These results provide additional evidence that the differential expression of NR2B in dysplastic human neocortex may play a role in the expression of in-situ epileptogenesis in human CD. NR2B may constitute a target for new diagnostic and pharmacotherapeutic approaches. D 2005 Elsevier B.V. All rights reserved. Theme: Disorders of the nervous system Topic: Epilepsy: basic mechanisms Keywords: Epilepsy; Neocortex; Dysplasia; Human; Brain slice

1. Introduction Cortical dysplasia (CD) [42] is commonly associated with medically intractable epilepsy and is associated with a less favorable epilepsy surgery outcome than other patho* Corresponding author. Department of Neurology, Section of Epilepsy, Cleveland Clinic Foundation, 9500 Euclid Avenue, Desk S51, Cleveland, OH 44195, USA. E-mail addresses: [email protected], [email protected] (G. Mo¨ddel). 0006-8993/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.brainres.2005.03.042

logies [12,48]. CDs are developmental disorders caused by disruption of neuronal migration. They are characterized by changes in neocortical microarchitecture: disturbance of laminar organization, ectopic neurons, and cellular abnormalities such as cytomegalic or dysmorphic neurons [7]. The association of CD with epilepsy is well documented [13,30,35]. The processes involved in the generation of epileptiform activity in human CD remain unclear. Possible mechanisms include changes in network connectivity, excitatory and

G. Mo¨ddel et al. / Brain Research 1046 (2005) 10 – 23

inhibitory neurotransmitter function, as well as disturbances of glial physiology and the regulation of extracellular space and ion concentrations. One of the possible causes that may lead to abnormal neuronal synchronization is an imbalance between excitatory and inhibitory neurotransmitter receptor activation. Increased expression of excitatory amino acid (glutamate) receptors has been found in human epileptic cortex [51]. However, besides increased receptor quantity, changes in receptor subunit composition may also account for hyperexcitability, as these changes can dramatically alter physiologic properties of receptors [43]. NMDA receptors (NMDAR) are likely to play an important role because (1) the NMDAR channel is permeable to Ca2+ which acts as a second messenger in signaling cascades attributed to synaptic plasticity [19,26] and (2) NMDARs display slow kinetics with a long inactivation time constant. Three families of NMDAR subunits have been identified, termed NR1, NR2, and NR3. The NR1 subunit is a single gene product with eight different splice variants [52]. The four NR2 subunits (termed NR2A-D) are each encoded by separate genes. Native NMDA receptors are heterotetramers [22,38] or -pentamers [37] consisting of multiple NR1 and at least one NR2 subunit. NR1 – NR2A receptors dominate in the mature neocortex. In contrast, NR1 –NR2B receptors are physiologically expressed during fetal development and feature higher peak ionic currents and a six times slower inactivation time constant [31,32,43], resulting in an increased Ca2+ influx. Several reports have shown a differentially increased expression of NR2B (or NR2A/B) in tissue resected from patients with drug-resistant epilepsy [9,27,33,49,50]. Mathern et al. [27] found increased NR2B expression in the hippocampi of epileptic patients. Crino et al. [9] reported increased NR2B mRNA in dysplastic neocortical neurons in human CD. Our own group has demonstrated differential NR2B expression in dysplastic neocortex from patients with pharmacoresistant focal epilepsy [49] and showed that NR2B subunits are coexpressed with and colinked to NR1 subunits [50], indicating that they are likely to form functional receptors. Moreover, we correlated the density of immunocytochemical (ICC) staining for NR2B with in-situ epileptic activity, assessed by prolonged direct cortical recordings [33]. The goal of this study was to test the hypothesis that NR2B-NMDA receptors contribute to epileptogenicity in dysplastic lesions in vitro: (a) we compared in vitro epileptiform field potentials (EFP) activity in brain slices prepared from dysplastic human neocortex with EFP in non-dysplastic control slices, and (b) we investigated the effect of the NR2B-subunit-specific NMDA receptor antagonist ifenprodil [45,46] on EFP in slices from dysplastic and non-dysplastic human neocortical tissue.

11

2. Materials and methods 2.1. Human neocortical tissue Human neocortical tissue was obtained from 20 patients who underwent surgical treatment for pharmacoresistant focal epilepsy. All patients had undergone prolonged surface with or without subdural electrode video EEG monitoring and MRI studies prior to surgery. The tissue samples consisted of portions of neocortex with underlying white matter approximately 2 cm3 in size, which were excised as part of the planned surgical treatment. No tissue was resected for sole experimental purposes. One half of each sample was immediately cut into slices for electrophysiological experiments (see below), the other half was fixed in paraformaldehyde (PFA) for histological and immunocytochemical (ICC) studies and partly frozen on dry ice for Western immunoblot analysis. Dysplastic tissue was chosen from the most epileptogenic area as determined by chronic subdural grid or intraoperative electrocorticographic recordings or, in cases with a clear morphological abnormality detected on imaging studies, from the region corresponding to the MRI abnormality. The presence of CD was confirmed by histology. Non-dysplastic neocortical samples were obtained from the inferior or middle temporal gyrus as part of a standard temporal lobectomy in patients who presented either with hippocampal sclerosis as the sole imaging abnormality or with normal MRI; the absence of pathological changes in the neocortex was confirmed histologically (the histopathological diagnoses were: hippocampal sclerosis [n = 4], mild hippocampal gliosis [n = 2], and no abnormal findings [n = 2]). The material was freshly obtained during operations performed at the Cleveland Clinic Foundation between March and November 2003 (n = 20; 8 male, 12 female). Histopathological analysis of the tissue revealed cortical dysplasia in 8 out of 12 cases in the ‘‘dysplastic’’ group. In four cases with preoperatively suspected cortical dysplasia, histopathological analysis revealed neocortical pathologies different form CD: two patients had ischemic lesions, one had a cavernous angioma, and one had perivenous inflammatory changes consistent with the diagnosis of Rasmussen’s encephalitis. These four cases were excluded post-hoc. The use of human tissue was approved by the Cleveland Clinic Foundation Institutional Review Board. 2.2. Histological and ICC studies Histological and immunocytochemical (ICC) stainings were performed from blocks of neocortical tissue obtained during surgery from sites directly adjacent to the ones used for electrophysiological experiments. Additionally, representative specimens of the resected tissue were independ-

12

G. Mo¨ddel et al. / Brain Research 1046 (2005) 10 – 23

ently evaluated and interpreted at the Department of Pathology of the Cleveland Clinic Foundation. For histological studies, three sections (30 Am) from each block were collected for cresylecht violet (CV) Nissl staining. Slides were analyzed for evidence of dyslamination, abnormal neuronal morphology, and excessive intracellular Nissl substance staining [35]. The ICC stainings for the NR2B subunit of the NMDA receptor were performed using polyclonal NR2B antibodies (Anti-NMDAR2; AB1557P, Chemicon, Temecula, CA, USA). Representative cortical blocks were cut (30 Am sections) in the coronal plain. The ICC staining procedure has previously been described in detail [29,49]. In brief, sections were processed as follows: (a) 5 min in 3% H2O2/ 10% methanol in TBS; (b) 60 min in TBS containing 1.5% normal serum; (c) 18 h in primary antiserum (1:100) diluted in TBS containing 1% normal goat serum; (d) 45 min in diluted biotinylated goat anti-rabbit-immunoglobulin G (ABC kits, Vector Laboratories, Burlingame, CA, USA); (e) 60 min in a solution of excess avidin and biotinylated horse radish peroxidase (HRP; Vector Laboratories, Burlingame, CA, USA); (f) 8 min in 0.05% 3,3V-diaminobenzidine tetrahydrochloride and 0.01% H2O2; (g) the reaction was terminated in ice-cold TBS. TBS (pH 7.6) was used as the rinsing buffer throughout the procedure. Sections were mounted on slides, air dried, and coverslipped. Control experiments omitted the primary antibodies from the staining protocol described; no specific ICC staining was seen in this case. The intensity of NR2B-specific staining was semiquantitatively evaluated by two of the investigators (G.M. and Z.Y.) according to the scoring system proposed by Najm et al. [33]: The extent of NR2B specific staining of neurons was estimated as follows: 0: no receptor-labeled neurons in the section; 1: receptor-labeled neurons cover less than 20% of the section; 2: receptor-labeled neurons cover between 20% and less than 40% of the section; 3: receptor-labeled neurons cover between 40% and less than 70% of the section; 4: receptor-labeled neurons cover 70% or more of the section. The density of ICC staining was graded as follows: 0: no labeling; 1: faint; 2: medium; 3: dark. The final score was obtained by adding the grades for extent and density. 2.3. Western immunoblot analysis The tissue was prepared for Western immunoblot as previously described [11] with minor modifications. Briefly, cortex samples were homogenized in ice-cold dissection buffer [11] and centrifuged. Pellets were collected and suspended in solubilization buffer [41]. Samples were held at 37 -C for 30 min and insoluble proteins sedimented by centrifugation. For immunoblotting, 40 Ag of solubilized proteins per sample was boiled in Laemmli buffer (Bio-Rad, Hercules, CA, USA), applied to gradient 4– 20% sodium dodecyl sulfate (SDS)-polyacrylamide gels, and separated by electrophoresis. The blots were blocked in modified Tris-

buffered saline (MTBS; TBS containing 2% fat-free milk and 1% bovine serum albumine), incubated with polyclonal NR2B primary antibody, and subsequently with secondary antibody (Jackson ImmunoResearch, West Grove, PA, USA). Protein– antibody complexes were visualized with enhanced chemiluminescence reagents (ECL-PLUS; Amersham, Arlington Heights, IL, USA). To quantify the protein densities, blots were scanned, and the digitized images of the bands imported into NIH Image v.1.58 densitometry software. The digitized gray values of each band were used as a semiquantitative parameter for the amount of NR2B protein. 2.4. Slice preparation for electrophysiology To obtain the highest possible degree of comparability between individual experiments, all slicing procedures were done by the same experimenter, following the same sequence of steps. Samples were obtained as neocortical – subcortical tissue cubes (approximately 2 cm3 in size) directly from the neurosurgeon. The samples were excised from the area designated for resection before the resection margins were demarcated by coagulation of cortical vessels, in order to avoid tissue hypoxia. The edges of the sample blocks were not diathermia-coagulated but excised by using surgical blades only. Each sample was cut in two halves, one of which was processed for histological and immunocytochemical studies (see above), while the other one was designated for electrophysiological experiments. Those tissue blocks were immediately immersed in ice-cold oxycarbogenated (95% O2; 5% CO2) low-calcium artificial cerebrospinal fluid used for preincubation (pACSF; composition [in mM]: NaCl 124; KCl 3; CaCl2 1; MgCl2 1.4; NaHCO3 26; KH2PO4 1.25; glucose 10), then glued onto a vibratome stage with the pial surface facing towards the blade. Slices (500 Am thick) were cut perpendicular to the cortical surface. This was done within 5 min after resection of the sample by the surgeon. The first slice cut from the surface of the tissue cube was discarded. The second through seventh slices were used for experiments, the remainder of the tissue block was discarded. While the tissue was held on the vibratome stage, it was continuously immersed in an acrylic glass tank containing ice-cold oxycarbogenated pACSF. Immediately after slices were cut, they were transferred into a preincubation chamber containing 75 ml cold (4 -C) oxy-carbogenated pACSF at pH 7.4 using a paintbrush. The preincubation chamber was then transferred from the operating room to the electrophysiology laboratory, while its temperature was allowed to slowly approximate room temperature (20 -C). After 1 h of preincubation in low-calcium pACSF (60 min after the slicing procedure was done), slices were transferred into a chamber (75 ml) containing oxy-carbogenated artificial cerebrospinal fluid (ACSF; 2 mM Ca2+ concentration, otherwise same composition as pACSF, see above) to reach ion concentrations resembling in vivo conditions. Slices

G. Mo¨ddel et al. / Brain Research 1046 (2005) 10 – 23

were held in ACSF at pH 7.4 and room temperature for another 2 h before field potential recordings were begun. Slice preparation was described in detail elsewhere [14,24,39]. 2.5. Electrophysiological recordings Slices were transferred from the incubation chamber into a submerge recording chamber (2 ml), where the tissue was superfused with oxy-carbogenated ACSF at a constant flow rate of 2 –4 ml/min at room temperature. Extracellular field potentials were recorded from presumed neocortical layer II – III as estimated by measuring the position of the electrode 300 –600 Am below the pial surface [8] using borosilicate glass micropipettes (0.5 – 2 MV) filled with ACSF. Signals were acquired in direct current (DC) mode using an Axopatch 200A amplifier (Axon Instruments, Union City, CA, USA) and recorded on a PC using pCLAMP 8.0 software (Axon Instruments, Union City, CA, USA). In order to elicit epileptiform activity, magnesium was omitted from the bath solution (0 Mg2+ ACSF). Electrode position was adjusted to maximize epileptiform field potential (EFP) burst amplitude. This was done by repeatedly repositioning the electrodes in steps of 500 Am across the slice. 2.6. Drug application Once EFP activity was recorded for more than 1 h at a stable frequency, the NR2B-subunit-specific N-methyl-daspartate (NMDA) receptor inhibitor ifenprodil (10 AM; Sigma, St. Louis, MO, USA) dissolved in 0 Mg2+ ACSF was applied for 1 h followed by 90 min of drug washout with 0 Mg2+ ACSF. In a subset of experiments, the nonspecific NMDA receptor antagonist dl-2-amino-5-phosphonovalerate (APV, 100 AM; Sigma, St. Louis, MO, USA) and the a-amino-3-hydroxy-5-methylisoxazole-4-proprionic acid (AMPA) receptor antagonist 6-cyano-7-nitro-quinoxaline-2,3-dione (CNQX, 5 AM; Sigma, St. Louis, MO, USA) were applied. In two control cases displaying interictal-like EFP, separate slices were superfused with the potassium channel blocker barium (Ba2+, 2mM) after interictal-like EFP had been established by omission of Mg2+, and ifenprodil was subsequently applied. 2.7. Assessment of EFP burst parameters In order to characterize epileptiform bursting activity, the following measures were taken: EFP repetition rate (min 1), maximum burst amplitude (AV), burst duration (ms), the integral of individual EFP bursts (burst integral, BI, AVs, see Fig. 2D3), as well as the duration and amplitude of slow afterpotentials. Both burst repetition rate and the integral of individual bursts are frequently used as parameters for assessing the degree of epileptiform activity in brain slices, yet they tend to be inversely related. Therefore, we

13

calculated an activity index (AI, burst repetition rate * burst integral, AVs/min) which was felt to be a more valid measure for quantifying bursting activity than each repetition rate or burst integral alone. The effect of ifenprodil was quantified by calculating ratios of mean burst repetition rate, burst integral, and activity index for the time interval between 50 and 60 min after beginning of drug application, compared to the 10-min time interval immediately preceding the beginning of drug application. 2.8. Experimental groups and statistical methods The cases included in this study were divided into two groups (cortical dysplasia, n = 8; vs. control, n = 8) on the basis of the histological diagnosis. While electrophysiological experiments were performed, the examiner was blinded for the results of the histological evaluation (as they were performed and interpreted later). The differences between the mean values for burst repetition rate, burst integral, and activity index in the two experimental groups were tested for statistical significance using Mann –Whitney’s test. To assess the agreement between the two investigators (G.M. and Z.Y) on the ICC scores, Spearman’s correlation coefficient was calculated. As the scores given by both investigators were highly correlated (Rho = 0.859; P < 0.01), the median of both scores was used for statistical comparison by Mann – Whitney’s test. For all statistical comparisons, a significance level of 0.05 was accepted. All data are given as mean T standard error of the mean (SEM).

3. Results 3.1. Clinical characteristics The clinical, EEG, MRI, and pathology data for the cases included in this study are summarized in Table 1 (‘‘nondysplastic’’) and Table 2 (‘‘cortical dysplasia’’). 3.2. Histological characteristics Cresylecht-violet (CV)-stained sections from non-dysplastic cases showed normal cortical lamination and columnar organization patterns (Fig. 1A1) or minimal disorganization without cellular abnormalities. The principal neurons were pyramidal in shape, and their apical dendrites were oriented towards the pial surface. All dysplastic samples (CD) were of type 2A cortical dysplasia [34]. In sections resected from patients with cortical dysplasia (including patient No. 12), the histopathological changes included dyslamination, disrupted columnar disorganization, and dysmorphic, irregularshaped, darkly Nissl-stained neurons with dendritic processes pointing in all directions. These abnormal neurons

14

G. Mo¨ddel et al. / Brain Research 1046 (2005) 10 – 23

Table 1 Synopsis summarizing clinical, EEG, MRI, and pathology data for the patients included in the ‘‘non-dysplastic’’ group Case no.

Age/ handedness/ sex

Age of sz. onset (years)

Related conditions

Seizure semiology; seizure frequency

EEG: interictal (INT); ictal onset (ICT)

MRI

Anti-convulsive drugs at time of surgery

Resection

Post-op. pathological diagnosis

1

33/R/F

22

Migraine headaches

Oxcarbazepin

R temporal resection

Mild hippocampal gliosis

24/R/F

6

INT: spikes L and R temporal; ICT: R temporal INT: spikes L temporal; ICT: L temporal

Normal

2

L hippocampal volume loss

Levetiracetam Oxcarbazepin

L temporal resection

Hippocampal sclerosis

4

33/R/F

17

–

INT: sharp waves R temporal; ICT: R temporal

Normal

Carbamazepin

R temporal resection

Normal

6

36/R/F

27

–

Automotor > generalized tonic – clonic; 2 – 3/weeks Automotor > generalized tonic – clonic; 0.5 – 1/week Complex motor > generalized tonic – clonic; 1 – 3/week Automotor; 2 – 3/week

Normal

Levetiracetam Topiramate

L temporal resection

Normal

7

42/R/F

20

R hippocampal volume loss

Valproic acid Gabapentin

R temporal resection

9

45/R/M

26

Hx of febrile seizures –

Normal

Lamotrigine Levetiracetam

L temporal resection

Mild hippocampal gliosis Hippocampal sclerosis

10

15/R/M

7

INT: sharp waves L temporal; ICT: L temporal INT: spikes R temporal; ICT: R temporal INT: spikes L temporal; ICT: L hemisphere INT: normal; ICT: L temporal

L hippocampal volume loss

Lamotrigine Levetiracetam

L temporal resection

Hippocampal sclerosis

11

41/R/M

25

INT: spikes R temporal; ICT: R hemisphere

R hippocampal volume loss

Phenytoin, Carbamazepine, Topiramate, Levetiracetam

R temporal resection

Hippocampal sclerosis

Hx of febrile seizures

Hx of febrile seizures Hx of febrile seizures; head trauma (MVA)

Automotor; 5/week Complex motor; 2/week Automotor; 1/week Automotor; 2 – 3/month

R: right; L: left; M: male; F: female; sz.: seizure.

were scattered throughout the gray matter, including the molecular layer, and were also seen as ectopic neurons in the subcortical white matter. No balloon cells were found in any of the patients with CD.

90 T 14 for dysplastic compared to 41 T 13 for nondysplastic samples (Fig. 1F). 3.4. Features of 0-Mg2+-induced epileptiform field potentials

3.3. NR2B subunit expression In sections prepared from non-dysplastic tissue samples, the neurons showed no or only faint immunopositive staining for NR2B protein (Figs. 1B1 and C1). The mean immunoreactivity score [33] was 3.0 T 0.5. In sections prepared from dysplastic samples, neuronal somata and dendritic processes stained strongly positive for NR2B (Figs. 1B2 and C2) with a mean score of 5.3 T 0.4 (Fig. 1D). Immunoreactive neurons occurred abundantly throughout the gray matter of dysplastic cortex. Western blotting analysis was performed on tissue samples from six experiments, four of which were from the dysplastic group. Blots showed a 180 kD band which stained strongly positive for NR2B in three out of four dysplastic cases and only faintly in both non-dysplastic samples (Fig. 1E). The relative gray values of the 180 kD bands as determined by the NIH Image v.1.58 software were

After Mg2+ ions were removed from the bath solution, spontaneous EFP occurred with a latency of 61 T 8 min in dysplastic (n = 8) compared to a latency of 93 T 12 min in non-dysplastic neocortical slices (n = 8; P < 0.05). Burst repetition rate increased gradually and reached a plateau about 30 min after onset of activity. Activity was recorded for at least 1 h before any measurements of EFP characteristics were performed to ensure stable conditions. The characteristics of EFP activity in dysplastic compared to non-dysplastic slices are summarized in Figs. 2 and 3. In non-dysplastic samples, EFP occurred at a mean repetition rate of 1.0 T 0.2 min 1 (range between 0.3 and 2.1 min 1; Fig. 2A1). The discharges consisted of negative spikes with a mean peak amplitude of 181 T 24 AV (range between 109 AV and 288 AV) followed by a slow negative afterpotential of 20 to 135 AV (mean: 68 T 15 AV) amplitude and 350 to 4500 ms (mean: 2090 T 540 ms) duration (Fig.

G. Mo¨ddel et al. / Brain Research 1046 (2005) 10 – 23

15

Table 2 Synopsis summarizing clinical, EEG, MRI, and pathology data for the patients included in the ‘‘cortical dysplasia’’ group Case Age/ Age of Related no. handedness/ sz. onset conditions sex (years)

Seizure semiology; seizure frequency

EEG: interictal (INT); ictal onset (ICT)

MRI

Anti-convulsive Resection drugs at time of surgery

Post-op. pathological diagnosis

3

13/R/F

4

Tuberous sclerosis; S/p resection of R glioma

Automotor > generalized tonic – clonic; 4 – 5/day

INT: spikes L anterior temporal; ICT: L temporal

Oxcarbazepine

CD

5

23/R/F

3

–

Dialeptic or automotor; 3 – 4/week

12

6 months / – /F

Birth

Epidermal Hypomotor > nevus generalized syndrome; tonic; 50/day R hemiparesis

INT: spikes L and R temporal; ICT: R temporoparietal; L temporal INT: PLEDs L hemisphere; ICT: L hemisphere

Band-like hyperintense (T2) foci, R and L temporal and frontal cortices Loss of gray – white matter delineation L temporal pole

14

6/R/F

4

17

8 months /R/F

Birth

18

37/R/F

1

INT: spikes R occipital and generalized; ICT: R occipital INT: sharp L hemiparesis Hypomotor > waves R arm and R temporal – leg tonic; occipital; 7 – 10/day ICT: R centroparietal INT: sharp L hemiparesis L myoclonic; 12 clusters/day; waves R hemisphere; bilateral ICT: nonasymmetric localizable tonic; 3 – 4/week

19

7 months / – /F

10 days

L hemiparesis Infantile spasms; 2–6 clusters/day

INT: spikes R frontal; ICT: generalized

20

4/R/F

1.5

Tuberous sclerosis

INT: slow spike-andwave complexes, generalized; ICT: generalized

Premature birth; cleft palate

Dialeptic > L clonic > generalized clonic; 1 – 2/month

Epileptic spasms; 10 – 20 clusters/day

L temporal lobectomy

Felbamate; L temporal Oxcarbazepine; lobectomy Clonazepam

L hemispheric Valproic acid, polymicrogyria; Phenobarbital, agenesis of Levetiracetam corpus callosum

CD; mild hippocampal gliosis

L hemispherectomy CD; polymicrogyria; one heterotopic nodule in the subcortical white matter R fronto-centroCD temporal resection

Lateral ventricles enlarged bilaterally

Lamotrigin, Levetiracetam, Clonazepam

Wedge-shaped lesion R parietooccipital suggestive of CD

Phenobarbital, Levetiracetam, Clonazepam

R hemispherectomy CD; patchy gliosis and microcalcifications

Loss of gray – white matter delineation R frontal and temporal, suggestive of CD Pachygyria, loss of gray – white matter delineation R frontal and temporal R frontal CD

Levetiracetam, Gabapentin, Carbamazepin

R functional hemispherectomy

CD

Valproic acid, Phenobarbital

R frontal and temporal resection

Severe CD; microcalcifications

Vigabatrin, Valproic acid, Lorazepam

R frontal lobectomy

Mild CD; microcalcifications

R: right; L: left; M: male; F: female; sz.: seizure; CD: cortical dysplasia.

2B1). The mean burst integral was 295 T 140 AVs, and the mean activity index was 265 T 105 AVs/min. In slices prepared from dysplastic samples, the characteristics of EFP were more heterogenous. The repetition rate

of 0-Mg2+-induced EFP was significantly higher than in non-dysplastic slices (mean: 2.4 T 0.4 min 1; range: 0.4 to 3.4 min 1; Figs. 2C1 and 3A). The amplitude of the initial negative spike was not significantly different from the peak

16

G. Mo¨ddel et al. / Brain Research 1046 (2005) 10 – 23

Fig. 1. Differential expression of the NR2B subunit of the NMDA receptor in dysplastic neocortex. (A) Low magnification cresylecht violet (CV) stainings; pial surface oriented to the top. (A1) Normal cortical lamination and columnar organization. (A2) In dysplastic tissue, horizontal lamination and columnar organization of the cortex are disrupted. Note the presence of darkly stained dysmorphic neurons. (B, C) Immunocytochemical (ICC) stainings for NR2B protein, performed on tissue samples classified as normal (B1, C1) and dysplastic (B2, C2); same cases as shown in panel (A). Low magnification (B); high magnification (C). Pial surface oriented to the top. (C1) Cell bodies and apical dendrites in normal cortex stain very faintly. (C2) In dysplastic cortex, cell somata and processes stain strongly positive for NR2B. (D) Mean NR2B immunopositivity scores for dysplastic compared to non-dysplastic cases. In cortical dysplasia, neurons and their processes stained significantly stronger positive for NR2B than in non-dysplastic samples (P < 0.01). (E) Western blot analysis of non-dysplastic (E1, a, b) and dysplastic (E2, a – d) cortex samples. The 180 kD band stains strongly positive for NR2B in three out of four dysplastic cases and only faintly in non-dysplastic cases. (F) The average gray values of the 180 kD band were higher for dysplastic (90 T 14) compared to non-dysplastic (41 T 13) samples.

amplitude in non-dysplastic slices (mean: 245 T 35 AV; range: 108 to 422 AV). The initial spike was followed by a sustained negative potential shift of 45 to 250 AV (mean: 140 T 24 AV) amplitude and 550 ms to 28 s (mean: 6 T 3.4 s) duration. In six out of eight dysplastic slices, multiple (2 to 9; mean: 4 T 1) afterdischarges were recorded. The mean total duration of the discharges, including afterdischarges, was 10.5 T 4.3 s (range from 2.5 s to 40 s; Fig. 2D1). The mean burst integral was 1380 T 690 AVs (Fig. 3B), and the mean activity index was 1750 T 450 AVs/min (Fig. 3C).

3.5. Effects of NR2B-subunit-specific NMDA receptor inhibition After recording baseline epileptiform activity for 1 h, the NR2B-subunit-specific NMDA receptor inhibitor ifenprodil [45,46] (10 AM) was added to the bath solution for 60 min, while EFP were continuously recorded. In brain slices obtained from dysplastic tissue, activity was near-completely suppressed after 60 min (Fig. 2C2). Mean burst repetition rate was reduced to 12% of the baseline value

G. Mo¨ddel et al. / Brain Research 1046 (2005) 10 – 23

17

Fig. 2. Effect of the NR2B-subunit-specific N-methyl-d-aspartate (NMDA) receptor inhibitor ifenprodil (10 AM) on epileptiform field potentials (EFP). (A, C) Effect of ifenprodil on EFP repetition rate (10-min periods). (B, D) Effect of ifenprodil on single burst characteristics (potential fluctuations from the traces in panels A and C, displayed at a higher time resolution). (A, B) Field potential (FP) recording from a slice obtained from a case with normal neocortical histology (‘‘Non-dysplastic Cortex’’); (C, D) FP recording from a slice obtained from a specimen diagnosed as dysplastic (‘‘Cortical dysplasia’’). 1: baseline, superfusion with artificial cerebrospinal fluid devoid of magnesium ions (0 Mg2+ ACSF); 2: recordings 50 min to 60 min after ifenprodil was added to the bath solution; 3: recordings 90 to 100 min after beginning of drug washout. The insets in panel (B) show the same spikes as the main traces with a higher time and voltage resolution. In non-dysplastic neocortical samples, EFP consisted of a spike and a slow after-wave of short duration (B1). After application of ifenprodil, EFP repetition rate is unchanged (A2); burst peak amplitude (B2) as well as the amplitude and duration of slow afterpotentials (B2, inset) are only slightly reduced. In dysplastic slices, an initial spike is followed by a sustained negative potential shift and multiple afterdischarges, lasting up to 30 s (D1). EFP repetition rate is markedly reduced by ifenprodil (C2); burst peak amplitude is decreased; the sustained negative potential shift following the initial spike as well as afterdischarges are suppressed (D2). With drug washout, suppression is partly reversible (C3, D3). The shaded area in D3 indicates the burst integral.

(Fig. 4A). Mean burst integral (BI) and activity index (AI) were diminished to 3% (BI) and 0.5% (AI) of baseline (Figs. 4B and C). Remaining single EFP bursts consisted of brief spikes followed by a slow negative afterpotential of 450 ms to 11 s (mean: 2.7 T 1.7 s) duration and 10 to 100 AV (mean: 45 T 28 AV) amplitude, similar to the spikes seen in nondysplastic slices at baseline. Sustained negative potential shifts as well as afterdischarges were suppressed; the peak

amplitude of the initial negative spike was significantly reduced from 245 T 35 AV to 107 T 33 AV (Fig. 2D2). In non-dysplastic slices, mean burst repetition rate was virtually unchanged (Figs. 2A2, 4A). Mean BI was reduced to 49%, mean AI to 53% of baseline (Figs. 4B and C). The amplitude of the initial spike was unchanged (174 T 27 AV compared to 181 T 24 AV at baseline). The amplitude and duration of the slow negative afterpotentials were slightly

18

G. Mo¨ddel et al. / Brain Research 1046 (2005) 10 – 23

Fig. 3. Epileptiform bursting characteristics of brain slices obtained from dysplastic tissue samples (‘‘Cortical Dysplasia’’) compared to brain slices obtained from histologically normal samples (‘‘Non-dysplastic Cortex’’). Epileptiform activity was elicited by omission of Mg2+ ions from the bath solution. (A) Burst repetition rate; (B) burst integral; (C) activity index, defined as burst repetition rate  burst integral. Data are given as mean values; error bars represent standard error of the mean (SEM). Data were tested for significance using Mann – Whitney’s test; * indicates a statistically significant difference (significance level 0.05).

reduced from 68 T 15 AV to 51 T 15 AV and from 2090 T ms to 1230 T 360 ms, respectively (Fig. 2B2). Following drug washout (90 min), the suppressive effect of ifenprodil in dysplastic samples was partly reversible (Figs. 2C3 and D3). 3.6. Sensitivity to non-specific NMDA receptor antagonists and AMPA receptor antagonists The non-specific NMDA receptor antagonist dl-2amino-5-phosphonovalerate (APV, 100 AM) and the AMPA receptor antagonist 6-cyano-7-nitro-quinoxaline2,3-dione (CNQX, 5 AM) were subsequently applied to dysplastic (n = 2) and non-dysplastic slices (n = 3) followed by washout with 0 Mg2+ ACSF, respectively. EFP in both dysplastic and non-dysplastic slices were reversibly suppressed by either APV or CNQX (Fig. 5). 3.7. Discharge type does not confound ifenprodil sensitivity The ifenprodil-sensitive discharges recorded in dysplastic slices were mostly of the ictal type, characterized

by prolonged potential shifts and multiple afterdischarges, whereas non-dysplastic slices displayed interictal-like spikes of short duration. This difference in discharge pattern might confound the observed differences in the effect of ifenprodil. Therefore, we added 2 mM barium (Ba2+), which is known to block repolarizing potassium channels [2], to the bath solution in two non-dysplastic cases. The 0-Mg2+-induced interictal EFP were transformed into prolonged discharges resembling those recorded in dysplastic slices. Subsequent application of ifenprodil slightly reduced the burst duration but did not suppress EFP activity (Fig. 6).

4. Discussion This study demonstrates functional significance of differential expression of the NMDA receptor NR2B subunit in human dysplastic cortex. (1) The increased NR2B expression is correlated with the presence of prolonged, ictal-like epileptiform field potential (EFP) activity in human dysplastic neocortex in vitro; and (2) EFP are differentially

Fig. 4. Changes of zero-magnesium-induced epileptiform bursting characteristics after application of the NR2B-subunit-specific N-methyl-d-aspartate (NMDA) receptor inhibitor ifenprodil (10 AM). Comparison of changes of burst repetition rate (A), burst integral (B), and activity index (C) for brain slices obtained from dysplastic tissue (‘‘Cortical Dysplasia’’) and brain slices obtained from tissue diagnosed as histologically normal (‘‘Non-dysplastic Cortex’’). Data are given as mean values; error bars represent standard error of the mean (SEM). Data were tested for significance using Mann – Whitney’s test; * indicates a statistically significant difference (significance level 0.05).

G. Mo¨ddel et al. / Brain Research 1046 (2005) 10 – 23

19

Fig. 5. Sensitivity to non-specific NMDA receptor and AMPA receptor antagonists. (A) Application of the non-specific N-methyl-d-aspartate (NMDA) receptor antagonist dl-2-amino-5-phosphonovalerate (APV; 100 AM; A1) or of the a-amino-3-hydroxy-5-methylisoxazole-4-proprionic acid (AMPA) receptor antagonist 6-cyano-7-nitro-quinoxaline-2,3-dione (CNQX; 5 AM; A2) completely and reversibly suppressed zero-magnesium-induced EFP in non-dysplastic slices. (B) Application of APV (100 AM; B1) or CNQX (5 AM; B2) completely and reversibly suppressed zero-magnesium-induced EFP in dysplastic slices.

suppressed by selective NR2B receptor inhibition in dysplastic, but not in histologically normal human cortex. 4.1. Cortical dysplasia and NR2B overexpression Dysplastic samples were classified as cortical dysplasia type 2A [34] and displayed the histological features previously described [35,36,42]. Our ICC and Western blot results confirm our previous reports that showed a differential increase in NR2B or NR2A/B expression in dysplastic neurons and areas [33,49,50]. Cortical dysplasia constitutes a disorder characterized by disrupted cortical development, suggesting that the increased NR2B expression in CD might reflect the disturbed maturation of neurons in these lesions. NR2B is highly enriched in axonal growth cones and varicosities in the developing rodent neocortex and hippocampus [16,44]. This indicates a physiological role for NR2B during neuronal migration and differentiation. Monyer et al. [32] have shown in rats that NMDA receptors containing the NR2B subunit are expressed throughout the neocortex during fetal development and are postnatally replaced by NR2A receptors. The idea of NR2B overexpression in CD lesions representing cellular immaturity is intriguing. There are, however, some clues suggesting that increase in NR2B may be a secondary phenomenon induced by recurrent epileptic activity or other postnatal mechanisms. This is supported by our recent observation that NR2B expression is also increased in some cases of Rasmussen’s encephalitis, a perivenous inflammatory epileptogenic condition which develops postnatally and shows progression over time (own unpublished observations). Moreover, Williams et al. [47] found that induction of long-term potentiation (LTP) in

the hippocampus by high-frequency perforant path stimulation resulted in an increase in NR2B subunits in the dentate gyrus, whereas low-frequency stimulation, which induces long-term depression (LTD), decreased NR2B expression. This supports the notion that NR2B is upregulated in an activity-dependent manner. However, further research is needed to test whether these findings are also relevant for the neocortex. 4.2. Tissue sample selection The constitution of a suitable ‘‘control’’ group is a permanent problem in experimental epilepsy research employing fresh alive human brain tissue, since normal neocortex is rarely resected. This applies especially to the pediatric age group. As the best-possible solution for this dilemma, we opted for postero-lateral temporal neocortex from temporal lobe epilepsy patients who presented with solely mesial temporal EEG abnormalities, hippocampal sclerosis as the only pathological MRI finding, and absence of histological changes in the lateral temporal neocortex (‘‘non-dysplastic’’ group). In these patients, the mesial limbic structures can be considered the epileptogenic zone. The postero-lateral temporal neocortex is often part of the resected specimen for anatomical reasons and, in most cases, does not show any pathological changes. An estimated 30% of patients with hippocampal sclerosis have additional neocortical abnormalities (so-called ‘‘dual pathology’’), however, these abnormalities are very likely to be detected by either MRI or histopathological examination. To minimize the probability of error, we collected one half of each tissue sample in paraformaldehyde to obtain cresylecht violet (CV) histological staining of sections immediately adjacent to the ones used for electrophysiological experi-

20

G. Mo¨ddel et al. / Brain Research 1046 (2005) 10 – 23

Fig. 6. Discharge type does not confound ifenprodil sensitivity. Application of the voltage-gated-K+-channel inhibitor barium (Ba2+; 2 mM) transformed interictal-like EFP in non-dysplastic slices (A1, B1) into prolonged discharges (A2, B2) resembling the ictal-like EFP seen in dysplastic cortex. Subsequent addition of the NR2B-subunit-specific NMDA receptor antagonist ifenprodil (10 AM) slightly reduced burst duration (B3) but did not suppress EFP (A3). (A) 10-min periods; (B) single bursts from the traces in panel (A), displayed at higher time resolution.

ments. Unfortunately, patients with mesial temporal epilepsy usually present with intractable seizures as adults, much later than the majority of patients with cortical dysplasia. Therefore, it is virtually impossible to age-match ‘‘dysplastic’’ and ‘‘non-dysplastic’’ groups with this approach, unless very long study durations of several years are anticipated. However, the mean ages of both ‘‘dysplastic’’ (11 T 5 years) and ‘‘non-dysplastic’’ groups (34 T 4 years) were far beyond the age at which NR2B expression has been found in human cortex (between birth and 20 weeks of age for the subiculum) [23]. Besides developmental changes in NMDA subunit expression, regional differences in NR2B expression between cortical areas might play a role. Whereas all our non-dysplastic samples were taken from the temporal lobe, the dysplastic samples came from other cortical regions as well, including the postero-lateral frontal lobe (n = 2). Data on regionally specific expression of NR2A and NR2B in humans are sparse. Akbarian et al. [1] compared mRNA levels of NMDA receptor subunits in various cortical regions in post-mortem specimen from schizophrenic and normal patients. Although they found higher NR2B-mRNA levels in frontopolar than in temporo-parietal cortex of normal subjects, the NR2A/NR2B ratios were not different between these regions. Regionally specific NR2B expression is therefore not likely to influence our results, given that only two out of eight specimen which displayed ifenprodil-sensitive discharges originated from the frontal lobe. 4.3. Epileptiform field potentials in dysplastic and non-dysplastic brain slices Omission of Mg2+ from the superfusate [20,21,39] reliably elicited EFP in both dysplastic and non-dysplastic

cases. The EFP recorded from non-dysplastic slices consisted of short-duration (less than 2 s) sharp negative transients that occurred at a slow repetition rate of 1 T 0.2/ min. This discharge pattern has previously been described in human neocortex, occurring either spontaneously or after removal of Mg2+ from the superfusate [4,5,39,40] and was termed ‘‘interictal-like’’. In brain slices prepared from histopathologically confirmed dysplastic tissue, the EFP occurred at a significantly higher repetition rate of 2.4 T 0.4/min. They were prolonged in duration (longer than 2 s) and were characterized by an initial negative spike followed by a sustained negative potential shift of variable duration and multiple afterdischarges towards the end of the event. This discharge pattern has been described in human epileptic cortex due to various pathologies [4– 6,20,28] and has been referred to as ‘‘ictal-like’’. The occurrence of ictal-like discharges suggests that the mechanisms involved in epileptiform burst repolarization are impaired in dysplastic cortex. Using rat hippocampal slices, Domann et al. [10] showed that GABAA mediated inhibition and activation of voltage-dependent K+ channels are responsible for the termination of epileptiform bursts. Indeed, ictal-like discharges could be evoked by application of the voltage-dependent K+ channel inhibitor Ba2+ in non-dysplastic slices (Fig. 6). Other groups reported ictallike EFP induced by the voltage-dependent K+ channel blocker 4-aminopyridine [6,28] or GABAA receptor blocking substances such as bicuculline [15]. Yet, NMDA receptors are likely to play another important role for the generation of ictal-like discharges. Upon activation, NMDA receptors generate excitatory postsynaptic currents (EPSCs) with a long inactivation time-constant which is further dependent on their subunit composition. Prolonged NMDA-receptor-mediated depolarizations are normally masked by recurrent GABAergic inhibition. Luhmann

G. Mo¨ddel et al. / Brain Research 1046 (2005) 10 – 23

and Prince [25] showed that minimal disinhibition by application of low doses of bicuculline in adult rat neocortex can ‘‘unmask’’ these NMDA-receptor-mediated depolarizations and give rise to synchronized afterdischarges which are sensitive to the NMDA receptor antagonist dl-2-amino-5-phosphonovaleric acid (APV). In human cortical dysplasia, the increase in NR2B expression demonstrated in this and preceding studies [33,49,50] could well account for an imbalance between GABAergic inhibition and NMDA-receptor-mediated excitation, as NR2B-NMDA receptors display higher peak currents and slower inactivation than NR2A-NMDA receptors [43]. 4.4. Effects of NR2B-specific NMDA receptor inhibition by application of ifenprodil We found that selective inhibition of NMDA receptors containing the NR2B subunit by application of ifenprodil (10 AM) leads to an almost complete suppression of epileptiform discharges in dysplastic tissue, whereas EFP repetition rate was virtually unchanged in non-dysplastic slices. Our current results suggest that blockade of NMDA receptors can abolish bursting in tissue from CD patients in at least one model of induced epileptiform activity in vitro. As evoked field potentials and single-cell analysis of NMDA-mediated synaptic currents were not yet performed in this study, it is difficult to draw firm conclusions about how these receptor subunits contribute to epileptogenesis. Further studies are needed to answer these questions. Ifenprodil [45,46] is the best characterized NR2Bspecific NMDA receptor inhibitor. It binds to NR2Bcontaining NMDA receptors with high affinity and blocks receptor responses in a non-competitive, voltage-independent manner. The concentration of half-maximum effect (EC50) of ifenprodil blocking NMDA-induced currents is less than 1 AM for NR1/NR2B-heteromeric receptors and greater than 100 AM for NR1/NR2A-heteromeric receptors [46]. For our studies, we chose a concentration of 10 AM to ensure NR2B subunit selective inhibition. Ifenprodil suppressed ictal-like afterdischarges in dysplastic brain slices. The remaining EFP were of the interictal type, indicating that NR2B-NMDA-receptor-mediated responses are involved in the generation of ictal-like discharges. The interictal-like EFP recorded in non-dysplastic slices in 0 Mg2+ ACSF were not suppressed by ifenprodil: EFP repetition rate and mean peak amplitude of spikes did not change. Application of either a non-specific NMDA receptor antagonist (AVP) or of an AMPA receptor antagonist (CNQX) completely and reversibly suppressed EFP in both dysplastic and non-dysplastic samples. This has been previously described [3,17,18]. Thus, 0-Mg2+-induced spikes in normal human neocortex, although requiring both AMPA and NMDA receptor activation, are not dependent on NR2B-NMDA receptor function.

21

5. Conclusion Taken together, our results show (a) that human neocortical brain slices acquired from dysplastic areas show more severe epileptiform activity in vitro than slices prepared from histologically normal neocortex supporting the hypothesis that cortical dysplastic lesions are intrinsically epileptogenic and (b) that differential expression of NR2B-containing heteromeric NMDA receptors may play a role in the generation of epileptiform activity in dysplastic human neocortex. With NR2B being a receptor subunit which (1) is differentially expressed in epileptogenic cortical dysplastic lesions, but not in histologically normal neocortex, and (2) substantially contributes to hyperexcitability in CD, it may constitute a promising target for both improvement of diagnostic tools (e.g. functional imaging) and new pharmacotherapeutic approaches.

Acknowledgments This study was supported by the National Institute of Health (NIH 1R21 NS42354 and NIH K08 NS02046 to Imad M Najm; NIH 2RO1 HL51614, NIH RO1 NS43284, and NIH RO1 NS 38195 to Damir Janigro); by the Stiftung Neuromedizin (Neuromedical Foundation), University of Mu¨nster, Germany; and by the Zentrum fu¨r Innovative Medizinische Forschung (IMF), University of Mu¨nster, Germany (IMF scholarships MO 620202 to Gabriel Mo¨ddel and KE 620201 to Christoph Kellinghaus).

References [1] S. Akbarian, N.J. Sucher, D. Bradley, A. Tafazzoli, D. Trinh, W.P. Hetrick, S.G. Potkin, C.A. Sandman, W.E. Bunney Jr., E.G. Jones, Selective alterations in gene expression for NMDA receptor subunits in prefrontal cortex of schizophrenics, J. Neurosci. 16 (1) (1996) 19 – 30. [2] C.M. Armstrong, S.R. Taylor, Interaction of barium ions with potassium channels in squid giant axons, Biophys. J. 30 (3) (1980) 473 – 488. [3] M. Avoli, A. Olivier, Bursting in human epileptogenic neocortex is depressed by an N-methyl-d-aspartate antagonist, Neurosci. Lett. 76 (2) (1987) 249 – 254. [4] M. Avoli, J. Louvel, R. Pumain, A. Olivier, Seizure-like discharges induced by lowering [Mg2+]o in the human epileptogenic neocortex maintained in-vitro, Brain Res. 417 (1) (1987) 199 – 200. [5] M. Avoli, C. Drapeau, J. Louvel, R. Pumain, A. Olivier, J.G. Villemure, Epileptiform activity induced by low extracellular magnesium in the human cortex maintained in vitro, Ann. Neurol. 30 (4) (1991) 589 – 596. [6] M. Avoli, A. Bernasconi, D. Mattia, A. Olivier, G.G. Hwa, Epileptiform discharges in the human dysplastic neocortex: in vitro physiology and pharmacology, Ann. Neurol. 46 (6) (1999) 816 – 826. [7] C. Cepeda, R.S. Hurst, J. Flores-Herna´ndez, E. Herna´ndez-Echeagaray, G.J. Klapstein, M.K. Boylan, C.R. Calvert, E.L. Jocoy, O.K. Nguyen, V.M. Andre, H.V. Vinters, M.A. Ariano, M.S. Levine, G.W. Mathern, Morphological and electrophysiological characterization of

22

[8]

[9]

[10]

[11]

[12] [13]

[14]

[15]

[16]

[17]

[18]

[19]

[20]

[21]

[22] [23]

[24]

[25]

[26]

[27]

G. Mo¨ddel et al. / Brain Research 1046 (2005) 10 – 23 abnormal cell types in pediatric cortical dysplasia, J. Neurosc. Res. 72 (2003) 472 – 486. O.D. Creutzfeldt, Cortex Cerebri: Leistung, Strukturelle und Funktionelle Organisation der Hirnrinde, Springer Verlag, Berlin, 1983, pp. 7 – 42. P.B. Crino, C. Duhaime, G. Baltuch, R. White, Differential expression of glutamate and GABA-A receptor subunit mRNA in cortical dysplasia, Neurology 56 (2001) 906 – 913. R. Domann, C.H.A. Westerhoff, O.W. Witte, Inhibitory mechanisms terminating paroxysmal depolarizing shifts in hippocampal neurons of rats, Neurosci. Lett. 176 (1994) 71 – 74. A.W. Dunah, R.P. Yasuda, B.B. Wolfe, Developmental regulation of tyrosine phosphorylation of the NR2D NMDA glutamate receptor subunit in rat central nervous system, J. Neurochem. 71 (1998) 1926 – 1934. J.J. Engel, Surgery for seizures, N. Engl. J. Med. 334 (1996) 647 – 652. M.A. Farrell, M.J. De Rosa, J.G. Curran, D.L. Secor, M.E. Cornford, Y.G. Comair, W.J. Peacock, W.D. Shields, H.V. Vinters, Neuropathologic findings in cortical resections (including hemispherectomies) performed for the treatment of intractable childhood epilepsy, Acta Neuropathol. 83 (1992) 246 – 259. J.D. Fauteck, J. Bockmann, T.M. Bo¨ckers, W. Wittkowski, R. Ko¨hling, A. Lu¨cke, H. Straub, E.J. Speckmann, I. Tuxhorn, P. Wolf, Melatonin reduces low-Mg2+ epileptiform activity in human temporal slices, Exp. Brain Res. 107 (1995) 321 – 325. A. Gorji, J.M. Ho¨hling, M. Madeja, H. Straub, R. Ko¨hling, I. Tuxhorn, A. Ebner, P. Wolf, H.W. Panneck, F. Behne, R. Lahl, E.J. Speckmann, Effect of levetiracetam on epileptiform discharges in human neocortical slices, Epilepsia 43 (12) (2002) 1480 – 1487. M. Herkert, S. Rottger, C.M. Becker, The NMDA receptor subunit NR2B of neonatal rat brain: complex formation and enrichment in axonal growth cones, Eur. J. Neurosci. 10 (1998) 1553 – 1562. G.G. Hwa, M. Avoli, Excitatory synaptic transmission mediated by NMDA and non-NMDA receptors in the superficial/middle layers of the epileptogenic human neocortex maintained in vitro, Neurosci. Lett. 143 (1 – 2) (1992) 83 – 86. G.G. Hwa, M. Avoli, A. Olivier, J.G. Villemure, Bicuculline-induced epileptogenesis in the human neocortex maintained in vitro, Exp. Brain Res. 83 (2) (1991) 329 – 339. A. Kirkwood, S.M. Dudek, J.T. Gold, C.D. Aizenman, M.F. Bear, Common forms of synaptic plasticity in the hippocampus and neocortex in vitro, Science 260 (1993) 1518 – 1521. R. Ko¨hling, A. Lu¨cke, H. Straub, E.J. Speckmann, I. Tuxhorn, P. Wolf, H. Pannek, F. Oppel, Spontaneous sharp waves in human neocortical slices excised from epileptic patients, Brain 121 (1998) 1073 – 1087. R. Ko¨hling, M. Qu¨, K. Zilles, E.J. Speckmann, Current-source-density profiles associated with sharp waves in human epileptic neocortical tissue, Neuroscience 94 (4) (1999) 1039 – 1050. B. Laube, J. Kuhse, H. Betz, Evidence for a tetrameric structure of recombinant NMDA receptors, J. Neurosci. 18 (1998) 2954 – 2961. A.J. Law, C.S. Weickert, M.J. Webster, M.M. Herman, J.E. Kleinman, P.J. Harrison, Expression of NMDA receptor NR1, NR2A and NR2B subunit mRNAs during development of the human hippocampal formation, Eur. J. Neurosci. 18 (5) (2003) 1197 – 1205. A. Lu¨cke, R. Ko¨hling, H. Straub, D. Moskopp, H. Wassmann, E.J. Speckmann, Changes of extracellular calcium concentration induced by application of excitatory amino acids in the human neocortex in vitro, Brain Res. 671 (1995) 222 – 226. H.J. Luhmann, D.A. Prince, Control of NMDA receptor-mediated activity by GABAergic mechanisms in mature and developing rat neocortex, Brain Res. Dev. Brain Res. 54 (2) (1990) 287 – 290. A.B. Macdermott, M.L. Mayer, G.L. Westbrook, S.J. Smith, J.L. Barker, NMDA-receptor activation increases cytoplasmic calcium concentration in cultured spinal cord neurons, Nature 321 (1986) 519 – 522. G.W. Mathern, J.K. Pretorius, J.P. Leite, H.I. Kornblum, D. Mendoza,

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35]

[36]

[37]

[38] [39]

[40]

[41]

[42]

[43]

[44]

[45]

A. Lozada, E.H. Bertram III, Hippocampal AMPA and NMDA mRNA levels and subunit immunoreactivity in human temporal lobe epilepsy patients and a rodent model of chronic mesial limbic epilepsy, Epilepsy Res. 32 (1998) 154 – 171. D. Mattia, A. Olivier, M. Avoli, Seizure-like discharges recorded in human dysplastic neocortex maintained in vitro, Neurology 45 (1995) 1391 – 1395. N. Mikuni, T. Babb, Z. Ying, I. Najm, K. Nishiyama, C. Wylie, K. Yacubova, T. Okamoto, W. Bingaman, NMDA receptor 1 and 2A/B coassembly increased in human epileptic focal cortical dysplasia, Epilepsia 40 (1999) 1683 – 1687. P.S. Mischel, L.P. Nguyen, H.V. Vinters, Cerebral cortical dysplasia associated with pediatric epilepsy: review of neuropathologic features and proposal for a grading system, J. Neuropathol. Exp. Neurol. 54 (1995) 137 – 153. H. Monyer, R. Sprengel, R. Schoepfer, A. Herb, M. Higuchi, H. Lomeli, N. Burnashev, B. Sakmann, P.H. Seeburg, Heteromeric NMDA receptors: molecular and functional distinction of subtypes, Science 256 (1992) 1217 – 1221. H. Monyer, N. Burnashev, D.J. Laurie, B. Sakmann, P.H. Seeburg, Developmental and regional expression in the rat brain and functional properties of four NMDA receptors, Neuron 12 (1994) 529 – 540. I.M. Najm, Z. Ying, T. Babb, A. Mohamed, J. Hadam, E. LaPresto, E. Wyllie, P. Kotagal, W. Bingaman, N. Foldvary, H. Morris, H.O. Luders, Epileptogenicity correlated with increased N-methyl-d-aspartate receptor subunit NR2A/B in human focal cortical dysplasia, Epilepsia 41 (8) (2000) 971 – 976. A. Palmini, I.M. Najm, G. Avanzini, T. Babb, R. Guerrini, N. Foldvary-Schaefer, G. Jackson, H.O. Luders, R. Prayson, R. Spreafico, H.V. Vinters. Terminology and classification of the cortical dysplasias. Neurology 62 (6 Suppl. 3) (2004) S2 – S8. R.A. Prayson, M.L. Estes, Cortical dysplasia: a histopathologic study of 52 cases of partial lobectomy in patients with epilepsy, Hum. Pathol. 26 (1995) 493 – 500. R.A. Prayson, J.L. Frater, Cortical dysplasia in extratemporal lobe intractable epilepsy: a study of 52 cases, Ann .Diagn. Pathol. 7 (3) (2003) 139 – 146. L.S. Premkumar, A. Auerbach, Stoichiometry of recombinant Nmethyl-d-aspartate receptor channels inferred from single-channel current patterns, J. Gen. Physiol. 110 (1997) 485 – 502. C. Rosenmund, Y. Stern-Bach, C.F. Stevens, The tetrameric structure of a glutamate receptor channel, Science 280 (1998) 1596 – 1599. H. Straub, A. Lu¨cke, R. Ko¨hling, D. Moskopp, M. Pohl, H. Wassmann, E.J. Speckmann, Low-magnesium induced epileptiform activity in the human neocortex maintained in vitro: suppression by the organic calcium antagonist verapamil, J. Epilepsy 5 (1992) 166 – 170. H. Straub, J.M. Ho¨hling, R. Ko¨hling, A. Lu¨cke, I. Tuxhorn, A. Ebner, P. Wolf, H. Pannek, F. Oppel, E.J. Speckmann, Effects of nifedipine on rhythmic synchronous activity of human neocortical slices, Neuroscience 100 (3) (2000) 445 – 452. T. Takagi, R. Logan, L. Teves, M.C. Wallace, J.W. Gurd, Altered interaction between PSD-95 and the NMDA receptor following transient global ischemia, J. Neurochem. 74 (2000) 169 – 178. D.C. Taylor, M.A. Falconer, C.J. Bruton, J.A. Corsellis, Focal dysplasia of the cerebral cortex in epilepsy, J. Neurol. Neurosurg., Psychiatry 34 (1971) 369 – 387. S. Vicini, J.F. Wang, J.H. Li, W.J. Zhu, Y.H. Wang, J.H. Luo, B.B. Wolfe, D.R. Grayson, Functional and pharmacological differences between recombinant N-methyl-d-aspartate receptors, J. Neurophysiol. 79 (1998) 555 – 566. Y.H. Wang, T.Z. Bosy, R.P. Yasuda, D.R. Grayson, S. Vicini, T. Pizzorusso, B.B. Wolfe, Characterization of NMDA receptor subunitspecific antibodies: distribution of NR2A and NR2B receptor subunits in rat brain and ontogenic profile in the cerebellum, J. Neurochem. 65 (1995) 176 – 183. K. Williams, Ifenprodil discriminates subtypes of the N-methyl-

G. Mo¨ddel et al. / Brain Research 1046 (2005) 10 – 23

[46] [47]

[48]

[49]

d-aspartate receptor: selectivity and mechanisms at recombinant heteromeric receptors, Mol. Pharmacol. 44 (4) (1993) 851 – 859. K. Williams, Ifenprodil, a novel NMDA receptor antagonist: site and mechanism of action, Curr. Drug Targets 2 (3) (2001) 285 – 298. J.M. Williams, D. Guevremont, J.T. Kennard, Long-term regulation of N-methyl-d-aspartate receptor subunits and associated synaptic proteins following hippocampal synaptic plasticity, Neuroscience 118 (4) (2003) 1003 – 1013. E.C.Y. Wyllie, Y.G. Comair, P. Kotagal, J. Bulacio, W. Bingaman, P. Ruggieri, Seizure outcome after epilepsy surgery in children and adolescents, Ann. Neurol. 44 (1998) 740 – 748. Z. Ying, T. Babb, Y.G. Comair, W. Bingaman, M. Bushey, K. Touhalisky, Induced expression of NMDAR2 proteins and differential

23

expression of NMDAR1 splice variants in dysplastic neurons of human epileptic neocortex, J. Neuropathol. Exp. Neurol. 57 (1998) 47 – 62. [50] Z. Ying, T.L. Babb, N. Mikuni, I. Najm, J. Drazba, W. Bingaman, Selective coexpression of NMDAR2A/B and NMDAR1 subunit proteins in dysplastic neurons of human epileptic cortex, Exp. Neurol. 159 (2) (1999) 409 – 418. [51] K. Zilles, M.S. Qu¨, R. Ko¨hling, E.J. Speckmann, Ionotropic glutamate and GABA receptors in human epileptic neocortical tissue: quantitative in vitro receptor autoradiography, Neuroscience 94 (4) (1999) 1051 – 1061. [52] R.S. Zukin, M.V. Bennett, Alternatively spliced isoforms of the NMDARI receptor subunit, Trends Neurosci. 18 (7) (1995) 306 – 313.