Brain Advance Access originally published online on October 17, 2006
Brain 2006 129(12):3277-3289; doi:10.1093/brain/awl287
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Altered expression of 
3-containing GABAA receptors in the neocortex of patients with focal epilepsy
1 Institute of Pharmacology and Toxicology, University of Zurich Zurich, Switzerland 2 Department of Neurosurgery, University Hospital Zurich Zurich, Switzerland 3 Department of Neurology, University Hospital Zurich Zurich, Switzerland 4 Department of Neurology, University Hospital and Medical School of Geneva Geneva, Switzerland 5 INSERM Faculty of Medicine Strasbourg, France 6 Service de Neurochirurgie, Hôpitaux Universitaires de Strasbourg Strasbourg, France 7 Present addresses: Swiss Epilepsy Center Zurich, Switzerland 8 The Mental Retardation Research Center, David Geffen School of Medicine, University of California Los Angeles, CA, USA
Correspondence to: Dr Fabienne Loup, Swiss Epilepsy Center, Bleulerstrasse 60, CH-8008 Zurich, Switzerland E-mail: loupf{at}pharma.unizh.ch
 |
Summary |
|---|
 Top Summary IntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Impaired transmission in GABAergic circuits is thought to contribute to the pathogenesis of epilepsy. Although it is well established that major reorganization of GABAA receptor subtypes occurs in the hippocampus of patients with medically refractory temporal lobe epilepsy (TLE), it is unclear whether this disorder is also associated with alterations in GABAA receptor subtypes in the neocortex. Here we have investigated immunohistochemically the subunit composition and neocortical distribution of three major GABAA receptor subtypes using antibodies specifically recognizing the subunits
1, 2, 3, ß2/3 and 
2. Cortical tissue was obtained at surgery from patients with TLE and hippocampal sclerosis (HS; n = 9), TLE associated with neocortical lesions (non-HS; n = 12) and frontal lobe epilepsy (FLE; n = 5), with post-mortem samples serving as controls (n = 4). A distinct laminar and neuronal expression pattern of the -subunit variants was found across the neocortical regions examined in the temporal and frontal lobes in both control and patient tissue samples. In the five patients with FLE, GABAA receptor subunit staining was unchanged as compared to controls. In patients with TLE we observed a marked decrease in 3-subunit staining in the superficial neocortical layers (I–III), but no change in the deep layers (V and VI) or in the expression pattern of the 1 and 2-subunits. Reduced expression in 3-containing GABAA receptors was detected in six out of nine patients of the HS group and four out of twelve patients of the non-HS group. Histopathological changes were present in eight out of the ten patients with decreased 3-subunit staining. The selective reduction in 3-containing GABAA receptors was confirmed using semiquantitative measurements of optical density (OD). The specific changes unique to 3-subunit expression in the superficial neocortical layers of patients with TLE suggest that this subtype is of particular significance in the reorganization of cortical GABAergic systems in focal epilepsy.
Key Words: cerebral cortex; GABA; human; seizures; temporal lobe epilepsy
Abbreviations: ECoG, electrocorticography; FLE, frontal lobe epilepsy; HS, hippocampal sclerosis; OD, optical density; TLE, temporal lobe epilepsy
Received May 30, 2006. Revised September 10, 2006. Accepted September 11, 2006.
|
Introduction |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
A variety of evidence indicates that inhibitory brain circuits, which depend primarily on signalling through GABAA receptors, are functionally impaired in epilepsy. Thus, blocking GABAA receptors pharmacologically promotes the generation of seizures in animals and in humans, whereas drugs that enhance GABAA receptor function are effective in treating seizures. This straightforward notion becomes more complex, however, when taking into account the wide array of structurally and functionally distinct GABAA receptor subtypes found in the central nervous system. Nineteen GABAA receptor subunits (
1–6, ß1–3, 1–3, 
, 
, 
, 
, 
1–3) have been identified and cloned from the mammalian CNS (Simon et al., 2004
), which theoretically can assemble into a vast number of distinct pentameric receptors. In fact, only a few dozen of these potential combinations are found in the brain, most of which include at least one of each of the , ß and subunit class (Fritschy and Möhler, 1995; Pirker et al., 2000). Furthermore, these distinct receptor subtypes are preferentially expressed in specific regions and neuronal populations and they exhibit different sensitivities to modulators including neurosteroids, benzodiazepines, ethanol and barbiturates (Sieghart and Sperk, 2002; Fritschy and Brünig, 2003).
Temporal lobe epilepsy (TLE) is the most common form of focal epilepsy in adults, and, when associated with hippocampal sclerosis (HS), is the most refractory to pharmacotherapy (Wieser and ILAE Commission on Neurosurgery of Epilepsy, 2004). Although abundant human data are available regarding reorganization of GABAergic interneuron circuits in the neocortex and hippocampus of TLE patients (DeFelipe, 1999; Sperk et al., 2004; Maglóczky and Freund, 2005), most studies aimed at characterizing the alterations of individual GABAA receptor subtypes in human TLE have focused on the hippocampus (Brooks-Kayal et al., 1999; Loup et al., 2000; Pirker et al., 2003; Porter et al., 2005). Previous immunohistochemical investigations of GABAA receptors in the neocortex of patients with refractory focal epilepsy were restricted to the 1-subunit (Wolf et al., 1994, 1996b). Results from autoradiographic investigations in the neocortex of patients with pharmacoresistant epilepsy yielded divergent conclusions, reporting no changes, decreases or increases in GABAA receptors (Olsen et al., 1992; Burdette et al., 1995; Nagy et al., 1999; Zilles et al., 1999; Sata et al., 2002). In vivo imaging studies using 11C-flumazenil, an antagonist at the benzodiazepine–GABAA receptor complex, have shown both focal increases and decreases of flumazenil binding in the neocortex of patients with partial epilepsy (Theodore, 2002; Koepp and Woermann, 2005). Although flumazenil-PET studies are indispensable for tracking changes in living subjects, the resulting images provide low spatial resolution of GABAA receptor distribution and flumazenil fails to distinguish between the different GABAA receptor subtypes. In the present study, we have used an immunohistochemical approach to investigate alterations in GABAA receptor subtype organization in neocortex removed at surgery from patients with medically intractable focal epilepsy. Tissue was processed following a protocol based on microwave irradiation to visualize the major GABAA receptor subunits 1, 2, 3, ß2/3 and 2 with subunit-specific antisera (Loup et al., 1998). Other potentially relevant subunits could not be studied as the corresponding antisera effective in human brain tissue are not yet available.
|
Material and methods |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Patient selection, intraoperative electrocorticography
Twenty-six patients undergoing surgery for medically intractable focal epilepsy were included in this study. Brain tissue was obtained from the neurosurgical units of the University Hospitals of Zurich, Geneva and Strasbourg. All procedures were performed with the informed consent of the patients or legal next of kin and were approved by the ethics committees of the respective institutions according to the Declaration of Helsinki. Presurgical assessment comprised high-resolution MRI, PET with 18fluoro-2-deoxyglucose and/or ictal and interictal single photon emission computed tomography for all cases, functional MRI in six cases, and magnetic resonance spectroscopy (MRS) in five cases. All patients underwent scalp EEG and, where indicated, invasive or semi-invasive EEG recordings were obtained (Zumsteg and Wieser, 2000
). Based on the histopathological findings in conjunction with neuroimaging, EEG and clinical data, patients were categorized into those with frontal lobe epilepsy (FLE) (n = 5, mean age at surgery ± SD: 22.6 ± 12.3 years, range 5–34 years) and those with TLE. The latter group was further subdivided into those with HS (n = 9, mean age at surgery ± SD: 39.6 ± 14.4 years, range 12–56 years) and those with neocortical lesions (non-HS; n = 12, mean age at surgery ± SD: 28.3 ± 12.2 years, range 9–49 years). Patients with focal cortical dysplasia (FCD) alone were specifically excluded. Relevant clinical data are summarized in Table 1.
|
Intraoperative electrocorticography (ECoG) was performed in 18 of 26 patients with grids of 4 x 8 electrodes and/or with strips of four electrodes (Pt/Ir) embedded in Silastic sheets (Ad-Tech, Racine, WI). In FLE patients, electrodes were placed over the exposed prefrontal and polar cortex and positioning of the grid and strip electrodes was modified to record from neighbouring areas where indicated. In TLE patients, the 4 x 8 grid was placed over the anterior and middle temporal lateral and inferior cortex and the strip electrode was positioned on the intraventricular hippocampus allowing for simultaneous recording from lateral and inferior temporal neocortex and from hippocampus. Centre-to-centre inter-electrode distance was 1 cm, the diameter of each electrode was 1.2 mm. Recordings were sampled at 400 Hz, with a bandwidth of 1–70 Hz over a minimum period of 25 min, using a 32-channel Nihon-Kohden EEG system. Spiking areas were defined where electrodes showed spikes and/or sharp waves with a mean frequency greater than 1 spike/min (+, 1–10; ++, 11–20; +++, > 20 spikes/min). Non-spiking areas were defined where electrodes showed no epileptiform graphoelements. Degree of spiking (+, ++, +++) was taken into account, as well as propagation if present. ECoG was carried out before resection and repeated during the surgical procedure as necessary. After surgical ablation, ECoG was performed again and showed either spike-free activity or residual spiking at the resection border. Residual spiking was anatomically localized and its degree was rated (+, ++, +++).
All tissue blocks were from resections performed for strictly therapeutic purposes. The surgical procedures included corticectomy or lesionectomy in the frontal or temporal lobe or anterior temporal lobectomy. Sixteen patients underwent amygdalo-hippocampectomy and HS was confirmed histopathologically in nine of these cases. The neocortical tissue samples originated from the anterolateral temporal neocortex and the frontopolar area, the frontal lateral region, and the orbital part of the inferior gyrus of the frontal lobe. In the patients with a circumscribed pathology, neocortical samples were collected within and adjacent to the lesion. Finally, control neocortical tissue from four subjects (mean age ± SD: 62.5 ± 11.1 years, range 48–73 years) with no known history of neurological or psychiatric disorders was collected at autopsy (mean post-mortem interval ± SD: 11.2 ± 3.4 h). Results on the hippocampi of these four subjects were reported in a previous study (Loup et al., 2000). The left, and, in two cases, also the right hemispheres were cut into coronal slabs of 1–1.5 cm thickness. One to three tissue blocks per subject were dissected from parts of the anterolateral temporal neocortex, which corresponded to the areas removed in those patients undergoing surgery for epilepsy. In one case, three blocks of control tissue were also dissected from the frontal lobe.
Tissue preparation
Immediately upon resection in the operating room or after dissection at autopsy, tissue blocks were rinsed in phosphate-buffered saline (PBS) at pH 7.4. They were then immersion-fixed for 6–8 h at 4°C under constant agitation in a mixture of 4% freshly dissolved paraformaldehyde and 15% saturated picric acid in 0.15 M phosphate buffer at pH 7.4 (Somogyi and Takagi, 1982) or in 4% paraformaldehyde alone. Following fixation, tissue blocks were pre-treated using a modified antigen-retrieval method based on microwave irradiation as described previously (Loup et al., 1998). Tissue blocks were cryoprotected in 10, 20 and 30% sucrose in PBS over a period of 3–4 days, frozen at –28°C in isopentane, and stored at –80°C. Series of 40 µm thick sections were subsequently cut in a cryostat and collected in ice-cold PBS. These were either processed for immunohistochemistry (see below) or transferred to antifreeze solution and stored at –20°C until use. This procedure allowed up to 12 different specimens to be processed in parallel. Staining for each of the five GABAA receptor subunits was always carried out on five consecutive sections with an interseries space of 720–800 µm. For histopathological examination, two additional adjacent series of sections were stained for Nissl with cresyl violet and with antibodies against the neuron-specific nuclear protein NeuN (Wolf et al., 1996a). In selected cases, we also used antibodies against glial fibrillary acid protein (GFAP) and an antiserum that recognizes specifically a non-phosphorylated epitope of neurofilament protein (SMI-32) and labels a subpopulation of pyramidal cells (Campbell and Morrison, 1989).
Immunohistochemistry
The following subunit-specific antibodies were used: mouse monoclonal antibodies bd-24 and bd-17 recognizing the human GABAA receptor 1-subunit and both the ß2 and ß3-subunits, respectively (Schoch et al., 1985; Ewert et al., 1990), and polyclonal guinea-pig antisera recognizing the 2, 3 and 2-subunits. The specificity of these antibodies has been extensively documented (Fritschy and Möhler, 1995; Loup et al., 1998; Waldvogel et al., 1999). The dilutions of the subunit-specific antibodies were: 1-subunit (monoclonal antibody bd-24), 0.14 µg/ml; 2-subunit (affinity-purified), 1.3 µg/ml; 3-subunit (crude serum), 1:3000; ß2/3-subunit (monoclonal antibody bd-17), 3.8 µg/ml; and 2-subunit (crude serum), 1:1500. Further, antibodies used were NeuN 1:1000 (MAB377; Chemicon, Temecula, CA), GFAP 1:100 000 (MAB360, Chemicon, Temecula, CA), and SMI-32 1:5000 (Sternberger Monoclonals Inc.; Covance Research Products, Berkeley, CA). Series of free-floating sections were pre-incubated in 1.5% H2O2 in PBS for 10 min at room temperature to block endogenous peroxidase activity. They were then washed three times for 10 min in PBS and processed for immunoperoxidase staining (Hsu et al., 1981) as described previously (Loup et al., 1998, 2000).
Data analysis
Sections were analysed with a Zeiss Axioskop 2 (Jena, Germany) equipped for bright-field microscopy. For display, images were digitized using a high-resolution camera (AxioCam; Zeiss, Jena, Germany) with Zeiss camera software (AxioVision version 4.4). Images were modified for contrast only and comparison images were adjusted uniformly (Adobe Photoshop version 7.0; Adobe Systems Incorporated, San José, CA). No other manipulation of images was performed. Illustrations were composed in Adobe Illustrator (version 10.0; Adobe Systems Incorporated, San José, CA).
Densitometric measurements
The intensity of labelling for the GABAA receptor subunits 1, 2, and 3 was measured by densitometry in sections from controls (n = 4), FLE (n = 4), HS (n = 8), and non-HS cases (n = 10) as described previously (Loup et al., 2000). Optical density (OD) measurements were recorded in the superficial (layers II and III) and deep layers (layers V and VI) of the neocortex. The average OD per section was calculated from measurements in four rectangles with an area of 400 x 200 µm each in the superficial as well as in the deep layers and all measurements were repeated twice in sets of adjacent sections for each patient. In four patients, insufficient tissue was available to perform a complete quantitative analysis of the subunits.
Statistical analysis
Densitometric measurements were analysed for statistical significance using the Kruskal–Wallis test [non-parametric analysis of variance (ANOVA); GraphPad Prism, GraphPad Software, San Diego, CA]. Data were further compared between individual groups (at P < 0.05) with a multiple comparisons test. In the cases where more than one block was available, values were first subjected to statistical analysis to ensure that interblock variations were not significant.
|
Results |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
The distribution of the GABAA receptor subunits
1, 2, 3, ß2/3 and 2 was analysed in neocortical specimens from three different groups of patients with medically intractable focal epilepsy. Tissue obtained at autopsy from patients with no evidence of neurological disease was used for controls as previous data indicated that staining patterns for GABAA receptor subunits in autopsy and surgical samples are comparable (Loup et al., 2000).
Patient histories and histopathological evaluation
Table 1 provides a summary of the relevant clinical data for each patient. The duration of epilepsy ranged from 4 to 29 years in the FLE group, from 10 to 44 years in the HS group and from 1 to 32 years in the non-HS group. Mean epilepsy duration was 13.4 years in the FLE group, 30.7 years in the HS group and 10.6 years in the non-HS group. The mean age of onset was 9.2 years in the FLE group, 8.9 years in the HS group and 17.6 years in the non-HS group. In the HS group, but not in the FLE or non-HS groups, an initial precipitating event was documented as described previously (French et al., 1993; Mathern et al., 1995), including febrile convulsions (n = 4), infantile meningitis/encephalitis (n = 2) and neonatal anoxia/ischaemia (n = 1). The overall mean postsurgical follow-up period was 57.3 months. Seizure-free status (class I; Engel, 1987) was achieved for all patients of the HS group, 9 out of 12 patients in the non-HS group and 1 patient of the FLE group.
With respect to the patients with FLE, histopathological examination revealed discrete focal Chaslin's subpial gliosis (n = 1), leptomeningeal venous angiitis with no abnormalities in the brain parenchyma (n = 1), a low grade tumour (n = 1), and a glial scar (n = 2). In the patients with histopathologically confirmed HS, examination of the anterolateral temporal neocortex showed mild Chaslin's subpial gliosis (n = 3), gliosis (n = 4) and/or white matter changes. Thus, in two cases, small aggregates of heterotopic neurons were detected in the subcortical white matter. According to Palmini et al. (2004), such abnormalities are classified as type II mild malformations of cortical development (mild MCD II). In contrast, in neocortical grey matter, staining for Nissl, NeuN or SMI-32 revealed a normal cytoarchitecture with no apparent neuronal cell loss (see Fig. 4A and D). The non-HS group consisted of patients with vascular cavernous malformations (n = 3), low grade tumours (n = 8) and a glial scar secondary to cranial trauma (n = 1). All lesions were located in the anterolateral temporal neocortex. In 4 of the 12 patients with non-HS we also found histopathological changes in tissue adjacent to the lesion. In particular, dyslamination and large neurons were observed in two cases, one with a dysembryoplastic neuroepithelial tumour (DNT) and the other with a ganglioglioma. According to Palmini et al. (2004), the abnormalities described above are consistent with type IB (FCD IB). Neocortical tissue from autopsy controls displayed a normal cytoarchitecture.
GABAA receptor subtypes in control grey matter
We first examined the distribution of the GABAA receptor subunits 1, 2, 3, ß2/3 and 2 in temporal neocortex at low power magnification (Fig. 1). Adjacent sections were stained for the neuronal marker NeuN to identify laminar borders. All subunit antibodies revealed specific patterns of immunoreactivity. In normal grey matter, the laminar pattern for the subunits 1, 2 and 3 was distinct, whereas the laminar pattern was similar for the subunits ß2/3 and 2 (Fig. 1). A comparable organization was also described in rodent brain where the -subunit variants represent largely distinct subtypes with a specific pharmacological profile while the ß2/3 and 2-subunits are common to most GABAA receptor subtypes (Fritschy and Möhler, 1995). Furthermore, the distribution pattern for the 1-subunit was nearly identical to that of the ß2/3 and 2-subunits, indicating that the 1-subunit is the most abundant of the -subunit variants in the human neocortex. Unless otherwise mentioned, the ubiquitously present subunits ß2/3 and 2, although analysed, will not be further described.
|
Figure 2 shows at higher magnification the distinct and specific pattern of laminar distribution in neocortex of each of the
-subunit variants tested. Staining for the 1-subunit was particularly intense in the lower part of layer III and in layer IV (Fig. 2B). A weaker labelled band corresponding to the upper/mid layer IV in adjacent NeuN and SMI-32-stained sections was often observed (Fig. 2A and B) and was best seen at low power magnification (Fig. 1). For the 2-subunit, labelling was pronounced in the superficial layers and weak in the deep layers (Figs. 1 and 2C). A transition in labelling intensity generally was apparent in layer IV. Immunoreactivity for the 3-subunit was most intense in layer II and the upper part of layer III, gradually decreasing toward the lower part of layer III. Layer IV was only lightly labelled (Figs. 1 and 2D). The deep layers were moderately labelled. Even at this power of magnification, apical dendrites originating from layers V and VI could be seen to extend toward the pial surface (Fig. 2D). For each subunit, similar patterns of staining were observed in temporal and frontal neocortical tissue of all autopsy cases (data not shown).
|
At the cellular level, staining for the
2-subunit was present in the neuropil of each layer with the neuronal cell bodies appearing lightly labelled against the background (Fig. 3A). 3-Subunit immunoreactivity was also observed in the neuropil, but additionally in the soma and dendrites of individual neurons, including in pyramidal cells predominantly situated in the lower part of layer III (Fig. 3B). In layers V and VI, a subpopulation of neurons was labelled that displayed the typical morphology of pyramidal cells with a few basal dendrites and long apical dendrites coursing toward the superficial layers (Fig. 3C). Figure 3D shows a high magnification image of a layer V pyramidal cell immunoreactive for the 3-subunit. For both the 2 and 3-subunits, intense immunoreactivity was also detected on the axon initial segment of pyramidal cells, as shown previously (Loup et al., 1998). Staining for the 1-subunit, and less prominently for the ß2/3 and 2-subunits, was not only seen in the neuropil, but also in numerous non-pyramidal cells. In particular layer II and the upper part of layer III displayed a high density of intensely stained interneurons with somata of small size and several, often radially oriented, dendrites (Fig. 3E).
|
GABAA receptor subtypes in the grey matter of patients with focal epilepsy
Neocortical tissue from patients with FLE, HS or non-HS was analysed for changes in GABAA receptor subunit immunoreactivity. In the patients with a circumscribed lesion in the neocortex (see Table 1), the tissue samples used for study were from the periphery of the lesion. Two major observations were made. First, in all three groups the staining pattern for the subunits
1 and 2, as well as ß2/3 and 2, was largely similar to that of controls in terms of laminar distribution and intensity (Fig. 4B, C and E). Secondly, a subset of patients (10/26) was found to exhibit markedly decreased 3-subunit staining in the superficial layers, whereas the deep layers appeared unchanged and thus provided an internal control within each section for each case (Figs. 4F and 5). In the patients with reduced 3-subunit staining, no apparent neuronal cell loss was observed in adjacent sections stained for SMI-32 or NeuN (Fig. 4A and D). The data for the 3-subunit in the superficial neocortical layers for each patient are summarized in Table 1.
|
|
In the group of patients with FLE (n = 5), no changes in
3-subunit immunoreactivity were observed as compared to controls. In the HS group (n = 9), 3-subunit immunoreactivity was unchanged in three and decreased in five cases. In the remaining case, regions of unchanged and decreased 3-subunit staining were seen in the superficial layers in two different samples while no alterations were observed in 3-subunit staining in the deep layers. In the non-HS group (n = 12), 3-subunit labelling was unchanged in seven cases, decreased in the superficial layers in four cases, and could not be determined in one case. Fig. 5 shows low power images of the laminar distribution pattern of the 3-subunit from six different cases (five HS, one non-HS), three of which had no changes and three of which had decreased 3-subunit staining in the superficial neocortical layers.
Semiquantitative densitometric analysis was performed to assess differences in staining intensity between the control, FLE, HS and non-HS groups for the subunits 1, 2 and 3 in the superficial and the deep layers of temporal or frontal neocortical tissue (Fig. 6). The following observations were made: (i) no significant differences in OD were found for the 1 and 2-subunits in the superficial or the deep layers between the different groups of patients (Fig. 6A–D). (ii) Differences in OD for the 3-subunit were observed in the superficial layers of a subset of patients in both the HS and the non-HS groups (Fig. 6E). When compared to the FLE group, 3-subunit OD was significantly decreased in both the HS and the non-HS groups (P < 0.01 and P < 0.05, respectively). Comparison within the HS group and within the non-HS group showed that 3-subunit OD was significantly decreased in a subset of patients (P < 0.0001 and P < 0.0001, respectively). (iii) No significant difference in OD was found for the 3-subunit in the deep layers for all patients (Fig. 6F).
|
At the cellular level, sections stained for the subunits
1 and 2 revealed similar patterns in all specimens when compared to controls (data not shown). For the 3-subunit, Fig. 7A shows a section from a case with an unchanged pattern of staining similar to what was seen in control tissue. In the HS and non-HS patients with changes in 3-subunit immunoreactivity in the superficial neocortical layers, a striking decrease in staining was present throughout layers I, II and III (Fig. 7B–D). Moreover, in layer II only, we observed pyramidal cells immunopositive for the 3-subunit, which had apical dendrites extending into layer I (Fig. 7C and D). These neurons possessed numerous long dendrites, which at times formed an intricate predominantly horizontal network with neighbouring layer II neurons (Fig. 7C). Some cells had a prominently labelled soma (Fig. 7D). Changes at the cellular level as depicted in Fig. 7 were seen to a variable degree in the temporal neocortex of all patients with reduced 3-subunit staining.
|
Intraoperative ECoG, changes in
3-containing GABAA receptors and histopathologyECoG was performed during surgery in 18 of 26 patients and results were analysed in a semiquantitative manner as described in Material and methods. As can be seen in Table 1, no consistent correlation was found between the degree of spiking activity recorded before resection and the changes in
3-subunit staining or the histopathology. In particular, areas with high spiking activity (+++) exhibited either unchanged or decreased 3-subunit expression. The postresection ECoG recordings showed that no or only little residual spiking was present at the border of the resection except in one FLE patient with rhythmic discharges in the contralateral frontal neocortex. |
Discussion |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Our first principal finding is that in the normal human neocortex the expression of five of the major GABAA receptor subunits (
1, 2, 3, ß2/3 and 2) displayed remarkable laminar and neuronal specificity. Moreover, the regional distribution of the GABAA receptor subunits was similar across all the neocortical regions examined in the temporal and frontal lobes. Second, in a subset of patients with TLE we observed a decrease in 3-subunit staining in the superficial neocortical layers, usually accompanied by histopathological changes. In contrast, the distribution and the intensity of labelling of the subunits 1, 2, ß2/3 and 2 were unchanged in patients with focal epilepsy.
Methodological considerations
The feasibility of this study depended on the utilization of high affinity antibodies and the quality and processing of the tissue. We verified the specificity of the subunit-specific antisera in human brain tissue in previous studies with competition experiments, by replacing primary antibodies with non-immune serum and in western blots (Loup et al., 1998; Waldvogel et al., 1999). Moreover, to achieve high specificity of staining with a low background, we used an antigen-retrieval microwave procedure adapted to human brain tissue (Loup et al., 1998).
A number of potential pitfalls must be considered in the interpretation of our findings. (i) Variability, because tissue samples originated from different epilepsy centres. However, tissue processing was performed according to a common protocol (Loup et al., 1998, 2000) and the specific reduction in 3-subunit staining was found in tissue samples from all three centres. (ii) A sampling problem, in that areas compared between different patients were in fact not homologous. This is rather unlikely as the location of the resected neocortical areas was carefully documented and the distribution of the subunits 1 and 2, which served as a reference, was unchanged across samples. (iii) Effects of confounding biographical/clinical data, in particular differences in pharmacotherapy or the occurrence of presurgical seizures (Bouvard et al., 2005). Thorough review of the clinical data from each patient failed to identify major discrepancies among patients. Furthermore, it is unlikely that the changes in 3-subunit expression are secondary to drug treatment for the following reasons: a decrease in 3-subunit expression was observed in tissue both from patients taking GABAergic drugs and from patients not taking GABAergic drugs. In other words, there was no correlation. Also, at least four patients who were not treated with GABAergic drugs at the time of surgery and with no history of prior exposure nevertheless exhibited decreased 3-subunit staining. Moreover, in patients in whom 3-subunit expression was decreased in temporal neocortex, the expression in the entorhinal cortex was unchanged (F. Loup, unpublished data), whereas all five subunits studied including the 3-subunit were differentially altered in the hippocampus (Loup et al., 2000). (iv) Staining artefacts. This possibility can be ruled out based on our uniform processing in parallel of the various tissue samples. In addition, immunohistochemical staining in the cases with decreased 3-subunit labeling was repeated several times and, finally, staining in the deep layers was unchanged, thus providing an intrasection control.
GABAA receptor subtype expression in normal grey matter
Among the three GABAA receptor subtypes identified in our study, the 1-subtype was most abundant, with highest staining in lower layer III and layer IV, followed by the 2 and the 3-subtypes. This is in line with the results of previous autoradiographic GABAA receptor binding studies (Young and Kuhar, 1979; Zezula et al., 1988) reporting high densities of benzodiazepine receptor sites, especially in layers III and IV. Using in situ hybridization for six GABAA receptor subunit mRNAs (1, 2, 5, ß1, ß2 and 2), Akbarian et al. (1995) found three principal patterns of laminar expression in normal human prefrontal cortex, with the 1-subtype dominating. The laminar distribution pattern that they reported for the 1, 2, ß2 and 2-subunit mRNAs is largely similar to that described in the present immunohistochemical study. Moreover, the GABAA receptor subunits 1, ß2/3 and 2 show a similar laminar distribution in the human visual cortex with immunoreactivity greatest in layer IV (Hendry et al., 1994). Finally, electrophysiological data from a study in human temporal cortical neurons demonstrated a benzodiazepine sensitivity profile consistent with a preponderance of 1-subunit expression (Gibbs et al., 1996).
To our knowledge, the present study provides the first description of 3-subunit immunoreactivity in the human neocortex, revealing a remarkably specific laminar distribution pattern. This subunit is especially interesting because, in contrast to the other subunits, its expression differs from that described in the rodent brain (Fritschy and Möhler, 1995; Pirker et al., 2000). In rodent neocortex, 3-subunit expression is mainly located in the deep layers, while in the human neocortical frontal and temporal regions examined, 3-subunit staining was also pronounced in the superficial layers, especially in layer II. Our results further demonstrate a cell-specific distribution of the five major GABAA receptor subunits. The 1 together with the ß2/3 and the 2-subunits were localized not only in pyramidal cells but also in numerous small interneurons, especially in the superficial layers. Layer II and the upper part of layer III of the human temporal neocortex contain smaller and a higher density of GABA-positive somata than the other layers (Kisvárday et al., 1990). In contrast, the 2 and 3-subunits were localized solely in pyramidal cells, most prominently at the axon initial segment (Loup et al., 1998; Volk et al., 2002).
Altered GABAA receptor subtype expression in the grey matter of patients with focal epilepsy
We found a selective reduction of the 3-subunit in the neocortex of a subset of TLE patients in the absence of a change of the 1, 2, ß2/3 and 2-subunits. Data from animal studies indicate, however, that functional 3-containing GABAA receptors generally co-assemble with ß and 2-subunits (Sieghart and Sperk, 2002; Fritschy and Brünig, 2003). The reason that we did not detect an associated decrease in ß2/3 and 2-subunit labelling probably relates to the ubiquitous distribution of these subunits. The ß and 2-subunits are included in the vast majority of GABAA receptors, most frequently in combination with the 1 or 2-subunit. Thus, a decrease in the fraction of ß and 2-subunits co-expressed with the 3-subunit, which in rodent brain constitutes only 10–20% of the total number of GABAA receptors (Sieghart and Sperk, 2002; Fritschy and Brünig, 2003), would result in a change too low to be detected by immunohistochemical methods.
Among all known subunits the 1 is the only one previously analysed immunohistochemically in the temporal neocortex of patients with pharmacoresistant focal epilepsy (Wolf et al., 1994, 1996b). In their first study where the primary focus was the hippocampal formation, Wolf et al. (1994) used the temporal neocortex as a reference and reported no changes in 1-subunit GABAA receptor immunoreactivity, which is confirmed by our data. Autoradiographic analysis of binding at GABAA receptors or at the benzodiazepine–GABAA receptor complex in the temporal neocortex of patients with medically refractory focal epilepsy has provided conflicting findings. Thus Zilles et al. (1999) using 3H-muscimol reported a downregulation of GABAA receptor density in four out of nine cases with non-HS and variable upregulation in the remaining cases. In TLE patients with HS, Olsen et al. (1992) found no significant difference in flumazenil binding between epileptic temporal neocortex and control tissue, whereas Burdette et al. (1995) reported a significant increase in flumazenil binding in layers V and VI of epileptic temporal neocortex. Taken together these studies show non-uniform changes in GABAA receptor binding in tissue from TLE patients. Similarly, imaging studies using 11C-flumazenil in patients with refractory focal epilepsy due to diverse pathologies have reported not only focal decreases but also focal increases of flumazenil binding either in temporal or extratemporal neocortex (Theodore, 2002; Koepp and Woermann, 2005). Until subtype-specific ligands become available it will not be possible to determine how the changes in 3-subunit expression observed in the present work relate to the changes described in autoradiography and PET studies.
In our study, the non-HS group consisted of patients with a focal lesion located in the anterolateral temporal neocortex. When we examined the neocortex adjacent to the lesion, we observed that it was histologically normal in 8 of the 12 cases where we also did not find any changes in 3-subunit staining compared to controls. In the other four patients with decreased 3-subunit expression, we found histopathological changes including gliosis and type IB FCD. This type of mild MCD was present in one patient with a DNT and one patient with a ganglioglioma. Both types of tumours may be associated with surrounding dysplastic cortical regions as described previously (Prayson et al., 1993; Daumas-Duport et al., 1999; Palmini et al., 2004). Wolf et al. (1996b) reported decreases as well as increases in 1-subunit immunoreactivity in the perilesional tissue in a subset of patients with neocortical lesions. We did not find alterations in 1-subunit staining in our study, but overall these results show that GABAA receptor changes can occur in the vicinity of focal lesions.
The HS group comprised nine patients, three with no change in 3-subunit labelling, five with decreased labelling and one with a decrease in one of two tissue samples. Mild Chaslin's subpial gliosis was detected in two of the patients with no change in 3-subunit labelling and in the patient where there was reduced staining in one of the two samples. In contrast, in four of the five cases with decreased 3-subunit staining, we observed gliosis and/or white matter changes, two of whom exhibited type II mild MCD (Palmini et al., 2004). Thus, decreased 3-subunit expression in patients with HS or with non-HS was associated in 8 out of 10 patients with histopathological changes, in particular mild forms of MCD. We did not, however, observe neuronal loss in TLE specimens, in agreement with previous studies (Babb et al., 1984; Bothwell et al., 2001). These findings suggest that the reduction in 3-subunit staining is related primarily to a downregulation of this subunit in pyramidal cells rather than to a loss of neurons expressing this subunit. Furthermore, decreased 3-subunit expression was not accompanied by a reduction in the 1 or 2-subunits, which would be expected were the changes a result of extensive neuronal loss.
We also studied five patients with FLE, where we did not find any changes in GABAA receptor subunit staining. Whether this lack of change reflects the small sample size, the type of pathology (absence of a circumscribed lesion except in one case) or other factors remains unclear. It is however interesting to note that the laminar distribution and intensity of GABAA receptor subunits was basically similar to that in the autopsy specimens and TLE samples.
In the patients with altered 3-subunit expression, the laminar pattern was remarkably stereotyped with a lack of change in deep layers and a marked decrease in staining in superficial layers. At the cellular level it is not possible to determine whether the more darkly stained, and therefore visible, pyramidal cells were the last cells to express the 3-subunit or whether these neurons expressed this subunit de novo. Nevertheless, the selective localization in layer II of intensely stained neurons with an extensive dendritic arborization confined to layers I and II suggests a reorganization of the 3-subtype in these layers. A recent study using complementary DNA microarrays and immunostaining reported a pattern of persistent gene activation in epileptic neocortex of patients with focal epilepsy mainly in layers II and III (Rakhade et al., 2005). As the superficial layers are primarily involved in processing activity from other cortical areas (Jones, 1984) and electrophysiological studies have shown the critical role of layers II and III in the generation of synchronous population events in human epileptic neocortex (Köhling et al., 1999), downregulation of 3-containing GABAA receptors in the superficial layers may contribute to decreased functional inhibition.
Other GABAA receptor subunits, for which antibodies effective in human tissue do not yet exist, might be upregulated and may thus compensate for a reduction in 3-subunit expression. However, a recent study using transcriptome profiling in human epileptic neocortex reported a prominent downregulation of the 5-subunit gene and other GABA system transcripts in both the pre-synaptic and the post-synaptic compartments in spiking samples (Arion et al., 2006). Further, functional and morphological alterations in GABAergic circuits in the neocortex of TLE patients have been described previously (DeFelipe, 1999; Avoli et al., 2005). In particular, the group of DeFelipe found patterns of decreased immunoreactivity for parvalbumin (PV) and glutamate decarboxylase in human epileptogenic neocortex (DeFelipe et al., 1993, 1994; Marco et al., 1996). The most important change was a focal loss of PV-positive chandelier cells (among other interneurons), which are thought to be the most powerful cortical inhibitory cells. Using quantitative electron microscopy, the same group found a loss of inhibitory synapses in the regions with marked decrease in staining for PV (Marco and DeFelipe, 1997). Interestingly, the changes in synaptic density in their study were most pronounced in the superficial layers, where we observed downregulation of 3-containing GABAA receptors. Taken together, these results suggest a prominent reorganization of the GABAergic circuitry, which could contribute to the genesis or the maintenance of seizure activity in human focal epilepsy.
|
Acknowledgements |
|---|
The authors are grateful to Franziska Parpan and Corinne Sidler for excellent technical assistance. The authors thank Drs N. de Tribolet, J.-G. Villemure and A. L. Benabid for providing surgical specimens and Dr A. Aguzzi for supplying post-mortem tissue. The authors also thank Dr H. Möhler for his generous support, Dr C. Marescaux for fruitful discussions during the initial phase of the project, Drs M. Seeck and T. Landis for help with clinical data collection and Dr U. Gerber for comments on the manuscript. This work was funded by the Swiss National Science Foundation and the NCCR on Neural Plasticity and Repair.
|
References |
|---|
|
TopSummaryIntroductionMaterial and methodsResultsDiscussionReferences |
|---|
Akbarian S, Huntsman MM, Kim JJ, Tafazzoli A, Potkin SG, Bunney WE Jr, et al. (1995) GABAA receptor subunit gene expression in human prefrontal cortex: comparison of schizophrenics and controls. Cereb Cortex 5:550–60.
Arion D, Sabatini M, Unger T, Pastor J, Alonso-Nanclares L, Ballesteros-Yáñez I, et al. (2006) Correlation of transcriptome profile with electrical activity in temporal lobe epilepsy. Neurobiol Dis 22:374–87.[CrossRef][ISI][Medline]
Avoli M, Louvel J, Pumain R, Köhling R. (2005) Cellular and molecular mechanisms of epilepsy in the human brain. Prog Neurobiol 77:166–200.[CrossRef][ISI][Medline]
Babb TL, Brown WJ, Pretorius J, Davenport C, Lieb JP, Crandall PH. (1984) Temporal lobe volumetric cell densities in temporal lobe epilepsy. Epilepsia 25:729–40.[ISI][Medline]
Bothwell S, Meredith GE, Phillips J, Staunton H, Doherty C, Grigorenko E, et al. (2001) Neuronal hypertrophy in the neocortex of patients with temporal lobe epilepsy. J Neurosci 21:4789–800.
Bouvard S, Costes N, Bonnefoi F, Lavenne F, Mauguiere F, Delforge J, et al. (2005) Seizure-related short-term plasticity of benzodiazepine receptors in partial epilepsy: a [11C]flumazenil-PET study. Brain 128:1330–43.
Brooks-Kayal AR, Shumate MD, Jin H, Lin DD, Rikhter TY, Holloway KL, et al. (1999) Human neuronal -aminobutyric acidA receptors: coordinated subunit mRNA expression and functional correlates in individual dentate granule cells. J Neurosci 19:8312–8.
Burdette DE, Sakurai SY, Henry TR, Ross DA, Pennell PB, Frey KA, et al. (1995) Temporal lobe central benzodiazepine binding in unilateral mesial temporal lobe epilepsy. Neurology 45:934–41.[Abstract]
Campbell MJ and Morrison JH. (1989) Monoclonal antibody to neurofilament protein (SMI-32) labels a subpopulation of pyramidal neurons in the human and monkey neocortex. J Comp Neurol 282:191–205.[CrossRef][ISI][Medline]
Daumas-Duport C, Varlet P, Bacha S, Beuvon F, Cervera-Pierot P, Chodkiewicz JP. (1999) Dysembryoplastic neuroepithelial tumors: nonspecific histological forms—a study of 40 cases. J Neurooncol 41:267–80.[CrossRef][Medline]
DeFelipe J. (1999) Chandelier cells and epilepsy. Brain 122:1807–22.
DeFelipe J, Garcia Sola R, Marco P, del Río MR, Pulido P, Ramón y Cajal S. (1993) Selective changes in the microorganization of the human epileptogenic neocortex revealed by parvalbumin immunoreactivity. Cereb Cortex 3:39–48.
DeFelipe J, Huntley GW, del Río MR, Sola RG, Morrison JH. (1994) Microzonal decreases in the immunostaining for non-NMDA ionotropic excitatory amino acid receptor subunits GluR 2/3 and GluR 5/6/7 in the human epileptogenic neocortex. Brain Res 657:150–8.[CrossRef][ISI][Medline]
Engel JJ. (1987) Outcome with respect to epileptic seizures. In Engel JJ (Ed.). Surgical treatment of the epilepsies(Raven Press, New York) pp. 553–72.
Ewert M, Shivers BD, Luddens H, Möhler H, Seeburg PH. (1990) Subunit selectivity and epitope characterization of mAbs directed against the GABAA/benzodiazepine receptor. J Cell Biol 110:2043–8.
French JA, Williamson PD, Thadani VM, Darcey TM, Mattson RH, Spencer SS, et al. (1993) Characteristics of medial temporal lobe epilepsy: I. Results of history and physical examination. Ann Neurol 34:774–80.[CrossRef][ISI][Medline]
Fritschy JM and Möhler H. (1995) GABAA-receptor heterogeneity in the adult rat brain: differential regional and cellular distribution of seven major subunits. J Comp Neurol 359:154–94.[CrossRef][ISI][Medline]
Fritschy JM and Brünig I. (2003) Formation and plasticity of GABAergic synapses: physiological mechanisms and pathophysiological implications. Pharmacol Ther 98:299–323.[CrossRef][ISI][Medline]
Gibbs JW III, Zhang YF, Kao CQ, Holloway KL, Oh KS, Coulter DA. (1996) Characterization of GABAA receptor function in human temporal cortical neurons. J Neurophysiol 75:1458–71.
Hendry SH, Huntsman MM, Vinuela A, Möhler H, de Blas AL, Jones EG. (1994) GABAA receptor subunit immunoreactivity in primate visual cortex: distribution in macaques and humans and regulation by visual input in adulthood. J Neurosci 14:2383–401.[Abstract]
Hsu SM, Raine L, Fanger H. (1981) Use of avidin-biotin-peroxidase complex (ABC) in immunoperoxidase techniques: a comparison between ABC and unlabeled antibody (PAP) procedures. J Histochem Cytochem 29:577–80.[Abstract]
Jones EG. (1984) Laminar distribution of cortical effect cells. In Peters A and Jones EG (Eds.). Cerebral cortex(Plenum, New York) Vol. 1: pp. 521–53.
Kisvárday ZF, Gulyas A, Beroukas D, North JB, Chubb IW, Somogyi P. (1990) Synapses, axonal and dendritic patterns of GABA-immunoreactive neurons in human cerebral cortex. Brain 113:793–812.
Koepp MJ and Woermann FG. (2005) Imaging structure and function in refractory focal epilepsy. Lancet Neurol 4:42–53.[CrossRef][ISI][Medline]
Köhling R, Qu M, Zilles K, Speckmann EJ. (1999) Current-source-density profiles associated with sharp waves in human epileptic neocortical tissue. Neuroscience 94:1039–50.[CrossRef][ISI]