GABAA receptor subunits in the rat hippocampus II: Altered distribution in kainic acid-induced temporal lobe epilepsy

Pergamon

PII:

Neuroscience Vol. 80, No. 4, pp. 1001–1017, 1997 Copyright ? 1997 IBRO. Published by Elsevier Science Ltd Printed in Great Britain. All rights reserved 0306–4522/97 $17.00+0.00 S0306-4522(97)00145-0

GABAA RECEPTOR SUBUNITS IN THE RAT HIPPOCAMPUS II: ALTERED DISTRIBUTION IN KAINIC ACID-INDUCED TEMPORAL LOBE EPILEPSY C. SCHWARZER,* K. TSUNASHIMA,*‡ C. WANZENBO } CK,‡ K. FUCHS,† W. SIEGHART† and G. SPERK*§ *Department of Pharmacology, University of Innsbruck, Peter-Mayr-Strasse 1a, A-6020 Innsbruck, Austria †Section of Biochemical Psychiatry, University Clinic for Psychiatry, Wa¨hringer-Gu¨rtel 18-20, A-1090 Vienna, Austria ‡Department of Psychiatry, National Center for Neurological Psychiatry, Tokyo, Japan Abstract––Intraperitoneal injection of kainic acid in the rat represents a widely used animal model of human temporal lobe epilepsy. Injection of kainic acid induces acute limbic seizures which are accompanied by seizure-induced brain damage and late spontaneous recurrent seizures. There is considerable evidence for an altered transmission of GABA in human temporal lobe epilepsy and in the kainic acid model. We therefore investigated by immunocytochemistry the distribution of 13 GABA receptor subunits in the hippocampus of rats 12 h, 24 h, and two, seven and 30 days after injection of kainic acid. Within the molecular layer of the dentate gyrus, decreases in á2- and ä- and slight increases in á1-, â2- and â3-immunoreactivities were observed at early intervals (12 to 24 h) after kainic acid injection. These changes were succeeded by marked increases in á1-, á2-, á4-, á5-, â1-, â3-, ã2- and ä-immunoreactivities in the same area after seven to 30 days. Within the hippocampus proper, changes in expression of GABAA receptor subunits were demarcated by considerable neurodegeneration of CA1 and CA3 pyramidal neurons. All subunits present within dendritic areas of CA1 and CA3 were affected. These were á1, á2, á5, â1–â3, ã2 and á4 (present only in CA1). Decreases in these subunits were followed by increased expression of á2-, á5-, â3-, ã2- and ä-subunits in the hippocampus proper notably in CA3 at later intervals (up to 30 days). á1-, â2-, ã2- and ä-subunits were found in presumed GABA containing interneurons throughout the hippocampus. Their immunoreactivity was augmented after two to seven days. Some á4-, ã3- and ä-immunoreactivity was also found in astrocytes 48 h after kainic acid injection. Our data indicate an impairment of GABA-mediated neurotransmission due to a lasting loss of GABAA receptor containing cells after kainic acid-induced seizures. The seizure-induced loss in GABAA receptors within the hippocampus may in part be compensated by increased expression of GABAA receptor subunits within the molecular layer of the dentate gyrus and in pyramidal cells. ? 1997 IBRO. Published by Elsevier Science Ltd. Key words: benzodiazepines, dentate gyrus, epilepsy, limbic system, receptor adaptation, temporal lobe epilepsy.

Temporal lobe epilepsy is included in the category of partial epilepsies. Its rate of occurrence is estimated to be about 40% of the epilepsies in the adult. Typically the patients experience status epilepticus or prolonged febrile convulsions early in life and then after a latency period of several years begin to suffer from repetitive temporal lobe seizures. These seizures are often resistant to common antiepileptic treatment but can frequently be cured by unilateral surgical removal of parts of the entire hippocampus (for review see Ref. 53). The most consistent neuropathological finding is Ammon’s horn sclerosis, characterized by loss of neurons and gliosis in CA1, CA3 and the dentate hilus. A dearrangement of other limbic struc§To whom correspondence should be addressed. Abbreviations: IR, immunoreactivity; KA, kainic acid; PBS, phosphate-buffered saline.

tures especially the amygdala and the piriform and entorhinal cortices, is also commonly seen.1,10,28,42 Limbic seizures in the rat induced by intracerebral or systemic injection of the neurotoxin kainic acid (KA) represent a valuable animal model for temporal lobe epilepsy which closely reflects the clinical and neuropathological symptoms of the disease in humans.4,34,43 Upon injection of KA the rats experience severe limbic seizures with status epilepticus. The animals recover during the following days. They, however, expose a decreased threshold to seizureinducing compounds such as pentylenetetrazol.30 About three weeks later, the rats start to display spontaneous recurrent seizures which increase in frequency as time proceeds.3 After KA-induced seizures in rats the neuropathological changes are strikingly similar to those observed in patients with temporal lobe epilepsy. Pronounced damage

1001

1002

C. Schwarzer et al.

of pyramidal cells in CA3 and CA1 and loss of certain interneurons in the hilus of the dentate gyrus become manifest. In the animal model, morphological rearrangement of some neuronal connections and conspicuous changes in the expression of immediate early genes, neurotrophins, neuropeptides and of a variety of proteins implicated in the function of the hippocampus have been reported.12,13,19,20,31,43,45 The molecular mechanisms leading to severe initial seizures and subsequent manifestation of epilepsy are still not entirely understood. A primary cause for spread of the acute seizures may be a marked impairment of transmission of GABA, the main inhibitory neurotransmitter in the brain.15,17 Decrease in GABAergic function occurs primarily at the presynaptic level and results in a block of inhibitory postsynaptic potentials in the CA1 sector of the hippocampus. A recovery can be observed after some weeks.16 These electrophysiological findings are supplemented by neurochemical findings showing an initial decrease in glutamate decarboxylase activity in the hippocampus of KA-treated rats.22,44 This decrease in activity of the key enzyme for GABA synthesis may partially be due to the loss of a subpopulation of GABAergic neurons.45 At later intervals, however, increased activity of glutamate decarboxylase in the hippocampus and cortex9,30,44 and increased expression of the respective mRNAs indicate augmented GABAergic activity in surviving neurons.14,39 So far only limited studies have been conducted on GABAA receptors in the KA model.16,25 Franck et al.16 observed a slight increase in flunitrazepam binding in the CA1 sector one to two months after KA-induced seizures when also the presynaptic GABAergic transmission recovered. Evidence for increased GABAA receptor binding in the molecular layer of the dentate gyrus has been reported in rats after electrical kindling.36,49–51 In the accompanying papers we report on regional distribution of 13 GABAA receptor subunits in the hippocampus of untreated rats46 and on changes in expression of mRNAs encoding the different subunits of the GABAA receptors after KA-induced seizures.52 In this study we investigated by immunocytochemistry changes in the distribution of GABAA receptor subunits. Various time intervals after kainate injection were examined to allow some interpretation of a possible association of changes in GABAA receptor subunit expression with various clinical stages, e.g., acute seizure syndrome, occurrence of neurodegenerations and development of a chronic epileptic state (temporal lobe epilepsy).

saline.43,44 The resulting behavioural syndrome consisted of early staring, ‘‘wet dog shakes’’ and seizures ranging from mild forehead nodding to severe limbic convulsions with rearing and foam at the mouth. These acute behavioural changes were rated using a previously established rating scale.43 Only rats exposing the entire behavioural spectrum including sustained limbic seizures (rating 3–4, in more than 80% of the rats) were included in the study and were killed after 12, 24, 48 h, seven or 30 days (n=3–4 per interval). Saline-injected controls of the same litter were included at each experimental interval (total n=7).

EXPERIMENTAL PROCEDURES

The immunocytochemical distributions of the GABAA receptor subunits in control rats are described in detail in the accompanying paper.46 In accordance with previous observations, no á6-IR was detected in the hippocampus of controls18 and KA-treated rats.

Kainic acid injection Male Sprague–Dawley rats (250–350 g, Forschungsinstitut fu¨r Versuchstierzucht, Himberg, Austria) were injected with 10 mg/kg KA, i.p. (Sigma, St Louis, MI, U.S.A.) in buffered saline or with the corresponding amount of

Tissue preparation At various intervals (12 h, 24 h, and two, seven and 30 days) after injection of KA, rats were killed by injection of a lethal dose of thiopenthal (150 mg/kg, i.p.; Sanabo, Austria). Then they were perfused with 50 ml ice-cold phosphate-buffered saline (PBS; 50 mM phosphate buffer, pH 7.4 in 0.9% NaCl) followed by 200 ml chilled 4% paraformaldehyde in PBS through the ascending aorta. The brains were removed from the skulls and blocked into three parts by coronal cuts. These were postfixed in the same fixative for 90 min at 4)C, transferred to 20% sucrose in PBS and kept there for 24 h at 4)C. Thereafter blocks were rapidly frozen by immersion in "70)C isopentane for 3 min and, after evaporating the isopentane, stored in tightly sealed vials at "70)C. For immunocytochemistry of the á3-subunit a different protocol was followed. Without prior perfusion the brains were immersed into isopentane ("70C), 90 s), cut in a cryostat and the mounted sections were fixed in acetone (5 min, room temperature).55 Immunocytochemistry Coronal sections (40 µm) were obtained from the dorsal hippocampus and kept in Tris–HCl-buffered (50 mM, pH 7.2) saline containing 0.1% sodium azide (4–6)C). The indirect peroxidase-antiperoxidase technique of Sternberger was used for immunostaining free-floating sections.47 The procedure and the antibodies have been described in detail in the accompanying paper.46 In each experiment controls were included using primary antibodies preadsorbed with 10 µg/ml of the respective synthetic peptide (except for ã2 for which 50 µg/ml peptide were used) or fusion protein (24 h, 4)C). Slices incubated without the primary antibody were included. Glial fibrillary acid protein was detected by immunofluorescence using a Cy3-coupled monoclonal antibody (clone G-A-5, Sigma, St Louis, MI). RESULTS

Histopathology As shown in the accompanying paper52 the rats exposed a typical loss of pyramidal cells (CA1 and CA3, but not CA2) and of neurons in the hilus of the dentate gyrus. There were considerable variations between individual animals in respect to cell losses in CA1 and CA3 although only rats with full seizure rating had been included in the study. á-Subunits

GABAA receptor subunits in limbic seizures

á1-subunit. The á1-subunit was found throughout the dendritic areas of the hippocampus proper, in the strata oriens and radiatum CA1 to CA3, the stratum lacunosum moleculare, and in the molecular layer of the dentate gyrus. (Fig. 1a). A considerable amount of the á1-immunoreactivity (IR) was located in perikarya and fibres of interneurons including pyramidalshaped basket cells in the dentate gyrus (Fig. 2a). Twelve hours after KA-induced seizures increased á1-IR was observed in most parts of the hippocampus (Fig. 1c). These increases were most pronounced in the stratum oriens (Fig. 2e), in the stratum lacunosum moleculare, the middle and outer molecular layer and adjacent to the inner and outer surface of the granule cell layer (Figs 1c, 2b). Already after 24 h, likely due to regional neurodegeneration, a ‘‘patch-like’’ reduction of á1-IR could be seen in some parts of the hippocampus (Fig. 1e). At the later intervals (after 30 days), a general reduction of á1-IR was observed in the hippocampus proper related to the neurodegenerative processes in CA1 and CA3 (Fig. 1g). At the same time, á1-IR was increased in the molecular layer of the dentate gyrus. At high magnification intense staining of individual (surviving) fibres was detected (Fig. 2f,i). Pronounced (probably increased) staining of interneurons and their fibres located in the dentate hilus and in CA3 could be observed. Although not quantified, the number of neurons appeared to be reduced (Fig. 2f,i). á2-subunit. The á2-subunit was found in fibres of all hippocampal subfields (Fig. 1b). It was especially enriched in the molecular layer of the dentate gyrus and was more concentrated in CA3 than in CA1. Twelve to 48 h after KA-induced seizures, á2-IR was slightly reduced in all parts of the hippocampus (Fig. 1d,f). It was, however, enhanced after 30 days, especially in the molecular layer of the dentate gyrus and in surviving fibres of CA3c (Fig. 1h). á3-subunit. á3-IR was weak in controls (Fig. 3a). It appeared to be unchanged 12 to 24 h after the initial limbic seizures (Fig. 3d). After two days, clusters of darkly stained pyramidal-shaped perikarya were observed in the hilus of the dentate gyrus and, after seven to 30 days in the stratum pyramidale CA3 and CA1 (Fig. 3g,j). á4-subunit. á4-IR was concentrated in the molecular layer of the dentate gyrus. Faint staining was observed in the strata oriens and radiatum of CA1. In other parts of the hippocampus marginal staining was detected (Fig. 3b). Moderate changes in á4-IR were observed after KA-induced seizures. A slight increase in the immunoreactive protein was detected in the molecular layer of the dentate gyrus after 30 days (Fig. 3k). At the same time á4-IR decreased in the CA1 sector, likely due to progressing neurodegeneration. After two days, faint staining was seen

1003

throughout the hippocampus that was likely associated with astrocytes (Figs 4e, 10d). á5-subunit. The á5-subunit was present at high concentrations in most parts of the hippocampus proper especially in the strata radiatum and oriens of CA1 and CA2 (Figs 3c, 4a). Twelve to 48 h after KA-induced seizures, á5-IR was markedly enhanced in the dendritic areas of CA1 (Fig. 3f). Similarly, after 30 days intense á5-IR was observed in the dendritic fields of CA1 to CA3. This staining, however, had a ‘‘patch-like’’ appearance and was obviously restricted to dendrites originating from clusters of surviving pyramidal neurons (compare Fig. 3i and l, 30 days after KA). In some animals, extensive loss in á5-IR as observed in the CA1 sector, although staining was conserved (or even enhanced) in CA2 (Figs 3l, Fig. 4c). These striking decreases in á5-IR in CA1 coincided with losses in CA1 pyramidal neurons in these rats. Thirty days after KA-induced seizures, á5-IR was detected also in interneurons of the dentate gyrus and in processes of these cells projecting through the stratum granulosum (Figs 3i,l, 4d). â-Subunits â1-subunit. The â1-subunit was present throughout the dendritic areas of the hippocampus, the strata oriens and radiatum CA1 to CA3 and the molecular layer of the dentate gyrus (Fig. 5a). It was especially concentrated within the dendrites of the CA2 sector (arrowhead in Fig. 5a). Twelve hours after injection of KA a reduction of â1-IR was found in the strata radiatum and oriens of CA1 to CA3 (Fig. 5d). This change was especially evident because â1-IR was not reduced in the adjacent stratum lacunosum moleculare and in the molecular layer of the dentate gyrus. At later intervals (24 h to 30 days), presumably due to neurodegeneration, staining became distorted in the strata oriens, radiatum and lacunosum moleculare, revealing a ‘‘patchy’’ distribution of immunostaining (Fig. 5g,j). In the CA2 sector, â1-IR was preserved at all time intervals investigated. It was most prominent after 48 h (Fig. 5g) and slightly reduced at the other intervals. â1-IR became also slightly increased in the strata oriens and radiatum CA1 and CA3 two to 30 days after KA injection. It was, however, restricted to patches of surviving clusters of pyramidal cells (Fig. 5g). â2-subunit. â2 was the least abundant among the â-subunits in the hippocampus. Using a comparatively high concentration of the â2 antibody (15 µg/ ml) intense and still highly-specific immunostaining was observed in the hippocampus (Fig. 5b). Especially interneurons throughout the hippocampal formation were heavily labelled.46 Twelve hours after KA-induced seizures increased â2-IR was observed in the granule cell layer of the dentate gyrus (Fig. 5e). At later intervals increased staining was present in

1004

C. Schwarzer et al.

Fig. 1. á1- and á2-subunits-IR of the GABAA receptor in the hippocampus after kainic acid induced seizures. á1-IR: control (a), 12 h (c), 24 h (e) and 30 days (g) after KA injection. á2-IR: control (b), 12 h (d), 24 h (f) and 30 days (h) after KA injection. Note increased á1- and á2-IR in the molecular layer 30 days after KA (arrowheads in g and h). Scale bar=0.5 mm.

GABAA receptor subunits in limbic seizures

Fig. 2. á1-subunit immunoreactivity in presumed interneurons in hippocampal subfields. The high magnification photomicrographs show á1-IR close to the stratum granulosum (sg) of the dentate gyrus (a–c) and in the CA3a sector (d–i) in controls (a,d,g) and 12 h (b,e), 24 h (c,h) and 30 days (f,i) after KA-induced seizures. Note the staining for the á1-subunit in type I basket cells of the dentate gyrus and in associated processes (a) which appears to be increased at early intervals 12 h (b) to 24 h (c) after KA seizures. Numerous intensely-stained neurons are found within and adjacent to the stratum pyramidale of CA3 (d–i). The number of these presumptive á1-immunopositive interneurons appears to be somewhat reduced after KA seizures (e–f). Processes originating from these neurons may contribute to the dense fibre plexus seen in the strata oriens and radiatum CA3. Staining of these fibres appears to be increased 12 h (e) to 30 days (f) after KA injection. Scale bars: in c=20 µm (for a–c); in f=100 µm (for d–f) and in i=20 µm (for g–i).

1005

1006

C. Schwarzer et al.

Fig. 3. á3-, á4- and á5-subunits-IR of the GABAA receptor in the hippocampus after kainic acid-induced seizures. Whereas only faint staining was observed for the á3-subunit in the entire hippocampus of controls (a), moderate á4-IR was found in the molecular layer of the dentate gyrus and throughout the CA1 sector (b). For the á5-subunit, intense staining was present in CA1 to CA3 (being highest in CA2) but not in the dentate gyrus. Note the moderate increase in á4-IR in the molecular layer of the dentate gyrus after 30 days (k) and the increase in á5-IR in the stratum oriens and radiatum CA3 at early intervals after KA seizures (f, 48 h). Losses in á5-IR at later intervals (30 days) reflect neurodegeneration of pyramidal cells in CA3 (i) and CA1 (l). Note the different extent of neurodegeneration in CA1 and CA3 in different animals (i,l). Cells intensely stained for the á3-subunit can be seen after two days in the dentate gyrus and seven to 30 days in the stratum pyramidale (j, and arrowheads in g). Scale bars=0.5 mm in l (for a–l, except j); in j=50 µm.

the strata oriens and radiatum of the subiculum (after 48 h) and of CA1 and in the molecular layer of the dentate gyrus (30 days). The patchy structure of the staining indicated the location of surviving pyramidal cell dendrites (Fig. 5k). â3-subunit. The â3-subunit was similarly distributed as â1. It was, however, more concentrated in the hippocampus than the â1-subunit (Fig. 5c). The concentration of â3-IR was highest in the molecular

layer of the dentate gyrus and slightly less in the strata oriens and radiatum CA1 and CA3. The CA2 sector appeared to be spared from â3-IR. In contrast to the â1-subunit, â3-IR was enhanced in the strata oriens and radiatum CA1 after 12 h (Fig. 5f), whereas no change in â3-IR was observed in the stratum lacunosum moleculare of the same sections. In spite of a reduction in the dendritic field of CA3, darkly stained fibres were seen throughout the CA3 sector already after 12 to 24 h (Fig. 5f). At the later intervals

GABAA receptor subunits in limbic seizures

1007

Fig. 4. High magnification photomicrographs of á5- and á4-subunits-IR of the GABAA receptor after kainic acid seizures. á5-IR is shown in the area of the molecular layer and stratum radiatum CA1 of a control rat (a) and 30 days after KA (b). Note the á5-IR in presumptive type I basket cells (arrows in b) and fibers after 30 days (b). In c, á5-IR in the CA1/CA2 transition zone 30 days after KA seizures (detail of Fig. 3l) is shown. Note the intense staining in fibres of CA2 whereas immunoreactivity is decreased in CA1 due to neurodegeneration. In d the dentate gyrus of the same section as b (30 days after KA) is shown. Note staining in presumptive type l basket cells (arrows) and fibres projecting to the molecular layer of the dentate gyrus. In e, á4-IR is shown in the strata oriens and radiatum CA1 two days after KA injection. Note the staining of astrocyte-like structures. Abbreviations: sr, stratum radiatum; slm, stratum lacunosum moleculare; ml, molecular layer; sg, stratum granulosum. Scale bars=100 µm in d (for a–d) and 200 µm in e.

(two to 30 days), â3-IR, in the same way as â1-IR, recovered in the CA3 sector, notably in CA3c (Fig. 5i,l). â3-IR appeared to be increased in the molecular layer of the dentate gyrus. This increase in â3-IR was initially somewhat more pronounced in the outer and inner (but not in the middle) parts of the molecular layer (12 h, Fig. 5f). At later intervals, it was present throughout the entire molecular layer (30 days, Fig. 5l). ã-Subunits ã2-subunit. Among the ã-subunits, only ã2-IR was found at high concentrations in neuronal structures of the hippocampus. It was concentrated in dendrites

of the stratum radiatum CA1 to CA3, in the stratum lacunosum moleculare and in the molecular layer of the dentate gyrus (Fig. 6a). Prominent ã2-IR was also present in perikarya located throughout the dentate hilus (Figs 6a, 8a,c). Twelve to 48 h after KA-induced seizures, ã2-IR became markedly elevated in pyramidal neurons of CA1 to CA3 and in the respective dendrites projecting to the stratum radiatum (Figs 6c,e, 7). Due to neurodegeneration, ã2immunostaining of dendritic fibres in the stratum radiatum was interrupted by unstained islets after 48 h and at later intervals (Figs 6g, 7f), although intensely stained dendrites originating from the preserved pyramidal neurons were seen even after 30 days. At this interval, prominent ã2-IR was

1008

C. Schwarzer et al.

Fig. 5. â1-, â2- and â3-subunits-IR of the GABAA receptor in the hippocampus after kainic acid-induced seizures. (Controls in a–c; 12 h after KA, d–f; 48 h, g–i; and 30 days, j–l). â1-IR was found at low concentrations throughout the hippocampus of controls (a). It became only slightly enhanced in the molecular layer of KA treated rats (d, g, j). Under the experimental conditions (use of high concentration of â2-antibody), the â2-subunit was widely distributed in controls (b), especially in local circuit neurons. Staining was enhanced after 12 h especially in the granule cell layer (arrowheads in e) and after 48 h and 30 days after KA injection (h,k). Strong immunostaining was found for the â3-subunit in most parts of the hippocampus (c). After KA seizures this staining became enhanced in the molecular layer and parts of CA3 especially after 30 days (l). The ‘‘patchy’’ pattern of immunostaining for all â-subunits in the strata oriens and radiatum after KA seizures (f,g,i,j,k,l, arrows in j–l) may reflect loss of pyramidal cell dendrites due to neurodegeneration. Note that whereas highest concentrations of the â1-subunit are present in CA2 (arrowhead in a), no â3-subunit is found in this area (arrowhead in c). Scale bar=0.5 mm.

also depicted in the molecular layer and in hilar interneurons, notably in pyramidal shaped basket cells, of the dentate gyrus (Fig. 8d). ã1- and ã3-subunits. For the ã1- and ã3-subunits, only faint and diffuse immunoreactivity was detected in all parts of the hippocampus (Fig. 9a,b). ã3-IR was observed in fibres throughout the hippocampus es-

pecially in the terminal field of mossy fibres (Fig. 10g) and, to a lesser extent, in the molecular layer and the hilus of the dentate gyrus (Fig. 9b). ã1-IR was slightly increased in all subfields of the hippocampus 24 h to 30 days after KA injection and appeared to be associated with glial structures (Figs 9e,g, 10f). Staining with the ã3 antibody was enhanced in the molecular layer of the dentate gyrus and in the terminal field

Fig. 6. ã2-, and ä-IRs in the hippocampus after kainic acid-induced seizures. ã2-IR is highly concentrated in all parts of the hippocampus (a). Note increased ã2-staining in the stratum pyramidale CA1 after 12 h (see details in Fig. 7) in CA3 after 24 h (e) and in the stratum moleculare after 30 days (g). ä-IR is present in moderate concentrations throughout CA1 and in high concentrations in the molecular layer of the dentate gyrus of controls (b). Staining for the ä-subunit in the molecular layer becomes reduced after kainic acid treatment (d,f,h). Scale bar=0.5 mm.

1010

C. Schwarzer et al.

Fig. 7. High magnification photomicrographs of ã2-subunit of the GABAA receptor-IR in CA1 and CA3 after kainic acid seizures. Intense ã2-IR is present in dendrites of the stratum radiatum of CA1 (a) and in CA3 (d) of controls. High concentrations of ã2-immunoreactivity are present in pyramidal neurons and their dendrites in CA1 12 h (b), 24 h (c) and in CA3 24 h (e), 48 h (f) after KA seizures. Scale bars: in c (for a–c)=25 µm, and in f (for d–f)=100 µm.

of mossy fibres after 12 h to 30 days (Fig. 9d,f,h). After 30 days, ã3-IR was reduced in the molecular layer, whereas increased staining was seen in the terminal field of mossy fibres (Fig. 9h). In some animals, restricted to the two to seven day intervals, ã3-IR was observed also in astrocyte-like structures of the strata radiatum and oriens of CA1 and CA3 (Figs 9f, 10h). ä-Subunit Overall distribution of the ä-subunit largely matched that of á4. ä-IR was preferentially located

in the molecular layer of the dentate gyrus (Fig. 6b). Faint staining was observed in the stratum granulosum, in the strata radiatum and oriens of CA1 and in the pyramidal cell layer of CA1 to CA3. ä-IR was also observed in pyramidal-shaped basket cells in the hilus of the dentate gyrus and in numerous interneurons of CA3.46 Immunostaining of the ä-subunit was slightly reduced in the molecular layer of the dentate gyrus and in the CA1 sector at all time intervals after KA injection (Fig. 6d,f,h). Two days after KA-induced seizures immunoreactivity was also observed in astrocytes (Fig. 10j,l).

GABAA receptor subunits in limbic seizures

1011

Fig. 8. High magnification photomicrographs of ã2-subunit-IR in interneurons of the hilus of the dentate gyrus. In controls (a,c), high concentrations of ã2-IR are present, besides staining in the molecular layer and in CA3, in interneurons of the hilus. Staining in interneurons and their processes becomes intensified 24 h (b) to 30 days (d) after KA seizures. Scale bars: in b (for a,b)=100 µm and in d (for c,d)=25 µm.

Possible localization of GABAA receptor subunits in glia By far the largest portion of immunoreactive GABAA receptor subunits appears to be located on neuronal structures. After KA-induced seizures, however, transient expression of á4-,ã3-, ä- and possibly of ã1-IR is observed in astrocytes, being most prominent about two days after KA injection (Fig. 10). This staining is blocked by preincubation of the antibodies with the respective antigens, however, it has to be judged with caution. Other proteins expressed at high concentrations in glia could interfere in epileptic rats.11,33,35 Our data, however, lend some support to the work of others suggesting the expression of functioning GABAA receptors on astrocytes.5 Since these astrocytic receptors may be capable of binding benzodiazepines7 they may, however, contain also á-subunits other than á4.54

DISCUSSION

This study was designed to investigate changes in the immunocytochemical distribution of 13 GABAA receptor subunits at different stages of acquisition of epilepsy induced by injection of KA.43 Thus, the early time intervals (up to 24 h after KA injection)

reveal alterations in the subunit expression induced during or early after the acute seizures and/or the early neurodegenerative processes. After two and seven days most of the neurodegenerative processes have stabilized. Morphological adaptations and manifestation of epilepsy develop. Thirty days after the initial KA-induced status epilepticus the chronic epileptic state with recurrent seizures is manifest.43 Conditions were chosen that allowed optimal comparison of staining patterns at the different time intervals after KA treatment. Sections obtained at various stages were processed together with control sections. It has to be kept in mind, however, that neurodegenerative changes are rather heterogeneous although only rats with the same high seizure rating were used (see e.g., á5-IR in Fig. 3i and l; for detailed discussion refer to Ref. 43). Secondly, immunocytochemistry at the best provides a semiquantitative picture of the altered subunit content. The immunocytochemical data therefore may be interpreted in the context with changes in mRNA levels.52 Two major events seem to be reflected by the alterations of immunoreactive GABAA receptor proteins: i) Changes directly related to neurodegeneration, ii) functional adaptation in receptor expression, such as (iia) presumed down-regulation of receptor subunits in undamaged neurons, and (iib)

1012

C. Schwarzer et al.

Fig. 9. ã1- and ã3-subunit immunoreactivities in the hippocampus of controls (a,b) and in KA-induced seizures: after 12 h (c,d); ã1 after 24 h (e); ã3 after seven days (f); after 30 days (g,h). Note the intense ã3-IR in the stratum lucidum (c,d,f,h). Scale bar=0.5 mm.

Fig. 10. GABAA receptor subunit immunoreactivities in astrocytes after kainic acid seizures. Immunofluorescense labelling of glial fibrillary acidic protein (GFAP) in astrocytes 48 h after KA is shown in b and k. Photomicrographs of the CA3 area (Nissl stain in a) show controls (a,c,g,i) and KA-treated rats (b,d,f,h,j,k,l). Strong immunostaining was found for á4 (d) and ä(j,l) in astrocytes 48 h after KA. Considerably weaker staining was found for ã3 (k). Only marginal labelling of presumed astrocytes (arrows) was seen for ã1 (f) after 30 days in CA1. Scale bars: in j=100 µm (for a–j), in l=20 µm (for k, l).

GABAA receptor subunits in limbic seizures 1013

1014

C. Schwarzer et al.

enhanced expression of subunit-IR, after distinct intervals during and after acquisition of epilepsy. Changes directly related to neurodegeneration As described in detail in the accompanying article on mRNA expression of GABAA receptor subunits,52 prominent neurodegeneration is seen in hilar interneurons and in CA3 and CA1 pyramidal neurons. Whereas neurodegeneration approaches a maximal extent after one to two days in the hilus and CA3 sector, cell damage appears to progress also at later intervals in CA1. The extensive cell losses are generally reflected by losses in immunostaining for all receptor subunits as seen in the cornu Amonis for á1-, á2-, á4-, á5-, all â-, ã2- and ä-subunits. The decrease in immunostaining in CA3 (e.g., after 12 h) precedes that in CA1 which became manifest after two and 30 days (refer to á5- and â3-IR in Figs 3 and 5). This is consistent with observations in Nisslstained sections52 and has been described before.27 The extensive losses in GABAA receptor proteins document the massive impairment of the hippocampal GABAergic circuitry in this animal model of temporal lobe epilepsy, as is also reflected by previous binding studies using whole hippocampal tissue.25 In human temporal lobe epilepsy, compatible results were obtained by receptor autoradiography in neurosurgically obtained specimens from patients with temporal lobe epilepsy23,32 and by positron emission tomography.6 Early down-regulation of GABAA receptor subunits Our in situ hybridization data suggest an early (6 to 12 h after KA injection) down-regulation of some mRNA species, notably those encoding the á2-, á5-, â1-, â3-, ã2- and ä-subunits in the granular cell layer of the dentate gyrus.52 These changes are only paralleled by transient (12 to 24 h) decreases in á2-, á5-, and ä-IR. Down-regulation of â mRNAs were observed only at the earliest time intervals (notably after 6 h)47 and do not become manifest at the protein level after 6 h (data not shown) and 12 h. Similarly, only a marginal decrease is observed for ã2-IR after 12 h. Thus, it is debatable whether these marginal changes in the GABAA receptor subunit expression at the early intervals after KA-induced seizures may facilitate initial seizure activity. Enhanced expression of GABAA subunits Early increases in ã2-immunoreactivity in the CA1 and CA3 pyramidal cells. As shown in Figs 6 and 7, a pronounced increase in ã2-IR is detected throughout the pyramidal layers including perikarya of pyramidal cells (where ã2-IR is not detected in controls). This is in variance with our in situ hybridization data, indicating slight decreases or a lack of changes in the mRNA expression. Altered post-translational

mechanisms or impaired dendritic transport of the ã2-protein may be possible reasons. Changes in ã2-IR in the KA model, reported here, are consistent with alterations in ã2 mRNA expression and with receptor binding studies in the kindling model. During acquisition of epilepsy by kindling, an initial decrease in ã2 mRNA expression is followed by a lasting increase in ã2 message in the pyramidal cell layer.24 These changes in mRNA expression are accompanied by initially decreased and later (in fully kindled rats) increased benzodiazepine and muscimol binding in the CA1 sector.49,51 Enhanced expression of GABAA receptor subunit proteins in pyramidal neurons in the chronic state of kainic acid-induced epilepsy. In spite of the extensive cell losses as observed within the CA3 sector and evaluated more closely in the accompanying paper,52 considerable recovery in immunoreactivity of several subunits (notably of á2, á5, â3 and ã2) occurs at the late intervals (30 days) after KA injection. This is especially striking in the CA3c sector in which neurodegeneration is most extensive.43,52 In this area total immunostaining observed after 30 days sometimes even exceeds that of controls. This observation is in agreement with the binding studies and electrophysiological data by Franck et al.16 that indicate a recovery of GABAA receptor function in pyramidal neurons at a late interval after KA-induced seizures. In this respect it is noteworthy, that a large portion of GABAergic neurons is resistant to seizure-induced cell losses2,45 and that GABAergic neurons may exert even increased activity as suggested by enhanced expression of glutamate decarboxylase.14,30,39,45 In conjunction with these presynaptic mechanisms, the observed postsynaptic accumulation of GABAA receptor subunits implies potent compensatory mechanisms in the chronic epileptic state. Under epileptic conditions, GABA may not only be released from interneurons but also from (glutamatergic) mossy fibres upon CA3 neurons.39 Enhanced expression of GABAA receptor subunit proteins in the molecular layer of the dentate gyrus. The observed accumulation of nearly all GABAA receptor subunit proteins expressed within the molecular layer of the dentate gyrus (á1, á2, á4, á5, â1, â2, â3, ã2 and ä) 30 days after KA injection is consistent with increases in mRNA expression of several subunits (notably á1, á3, á4, â1 and â2) in the granule cell layer.52 Other mRNA species (á2, á5, and ã2) have recovered to control values at this time (after declining initially). The increased synthesis of GABAA receptor components within the dendritic field of the granule cells indicates augmented GABAA receptor function at the chronic epileptic state. In spite of some loss of GABA neurons within the hilus, surviving GABA neurons may be capable of enhanced synthesis of the transmitter.14,39 Our data obtained in the KA model are in perfect

GABAA receptor subunits in limbic seizures

agreement with observations after electrical kindling. Markedly enhanced expression of GABAA receptor subunit mRNAs in granule cells,24,26 and increased GABAA receptor binding in the molecular layer of the dentate gyrus36,40,51 have been reported together with considerable rises in conductance of miniature inhibitory postsynaptic currents37 and increased GABA-induced chloride fluxes.48 It is also important to emphasize that granule cells are resistant to KA-induced seizures. Thus, an unaffected population of excitatory mossy fibres is innervating a considerably reduced number of CA3 pyramidal neurons.52 To avoid progressing excitotoxic damage of CA3 neurons a drastic reduction of mossy fibre function is required. This may be the reason why a variety of mechanisms becomes recruited to reduce mossy fibre function in KA-induced epilepsy. Sprouting of mossy fibres and other mechanisms may be capable of augmenting GABAergic

1015

transmission in the dentate gyrus.14,29,39,41 Neuropeptide Y and Y2-receptors become expressed in granule cells in kainate-induced seizures29,38,45 and may mediate presynaptic inhibition of glutamate release from mossy fibres.8,21 And GABA synthesized in granule cells of epileptic rats39 may mediate an inhibitory action when presumably released from mossy fibres or granule cell dendrites. Thus, enhanced expression of GABAA receptor proteins within the molecular layer of the dentate gyrus and in CA3 neurons may be supportive of such protective mechanisms. Acknowledgements—We would like to thank E. Kirchmair for excellent assistance and C. Trawo¨ger for preparing the photographs. We also thank Dr U. Berresheim for reading the manuscript. For information concerning antibodies contact Dr W. Sieghart. The work was supported by grants of the Austrian Science Foundation (to G.S. and W.S.) and of the Japanese Foundation for Aging and Health (to K.T.).

REFERENCES

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.

Babb T. L. and Brown W. J. (1987) Pathological findings in epilepsy. In Surgical Treatment of the Epilepsies (ed. Engel J. Jr), pp. 511–540, Raven, New York. Babb T. L., Pretorius J. K., Kupfer W. R. and Crandall P. H. (1989) Glutamate decarboxylase-immunoreactive neurons are preserved in human epileptic hippocampus. J. Neurosci. 9, 2562–2574. Baran H., Lassmann H. and Hornykiewicz O. (1988) Behavior and neurochemical changes 6 months after kainic acid. Neurosci. Abstr. 14, 472. Ben-Ari Y. (1985) Limbic seizure and brain damage produced by kainic acid: mechanisms and relevance to human temporal lobe epilepsy. Neuroscience 14, 375–403. Bormann J. and Kettenmann H. (1988) Patch-clamp study of ã-aminobutyric acid receptor Cl" channels in cultured astrocytes. Proc. natn. Acad. Sci. U.S.A. 85, 9336–9340. Burdette D. E., Sakurai S. Y., Henry T. R., Ross D. A., Pennell P. B., Frey K. A., Sackellares J. C. and Albin R. L. (1995) Temporal lobe central benzodiazepine binding in unilateral mesial temporal lobe epilepsy. Neurology 45, 934–941. Bureau M., Laschet J., Bureau-Heeren M., Hennuy B., Minet A., Wins P. and Grisar T. (1995) Astroglial cells express large amounts of GABAA receptor proteins in mature brain. J. Neurochem. 65, 2006–2015. Colmers W. F., Lukowiak K. and Pittmann Q. J. (1988) Neuropeptide Y action in the hippocampal slice: site and mechanism of presynaptic inhibition. J. Neurosci. 8, 3827–3837. Davenport C. J., Brown W. J. and Babb T. L. (1990) Sprouting of GABAergic and mossy fiber axons in dentate gyrus following intrahippocampal kainate in the rat. Expl Neurol. 109, 180–190. Du F., Whetsell W. O. J., Abou-Khalil B., Blumenkopf B., Lothmann E. W. and Schwarcz R. (1993) Preferential neuronal loss in layer III of the entorhinal cortex in patients with temporal lobe epilepsy. Epilepsy Res. 16, 223– 233. Du F., Williamson J., Bertram E., Lothman E., Okuno E. and Schwarcz R. (1993) Kynurenine pathway enzymes in a rat model of chronic epilepsy: immunohistochemical study of activated glial cells. Neuroscience 55, 975–989. Dugich-Djordjevic M. M., Tocco G., Lapchak P. A., Pasinetti G. M., Najm I., Baudry M. and Hefti F. (1992) Regionally specific and rapid increases in brain-derived neurotrophic factor messenger RNA in the adult rat brain following seizures induced by systemic administration of kainic acid. Neuroscience 47, 303–315. Ernfors P., Bengzon J., Kokaia Z., Persson H. and Lindvall O. (1991) Increased levels of messenger RNAs for neurotrophic factors in the brain during kindling epileptogenesis. Neuron 7, 165–176. Feldblum S., Ackermann R. F. and Tobin A. J. (1990) Long-term increase of glutamate decarboxylase mRNA in a rat model of temporal lobe epilepsy. Neuron 5, 361–371. Fisher R. S. and Alger B. E. (1984) Electrophysiological mechanisms of kainic acid-induced epileptiform activity in the rat hippocampal slice. J. Neurosci. 4, 1312–1323. Franck J. E., Kunkel D. D., Baskin D. G. and Schwartzkroin P. A. (1988) Inhibition in kainate-lesioned hyperexcitable hippocampi: physiologic, autoradiographic, and immunocytochemical observations. J. Neurosci. 8, 1991–2002. Franck J. E. and Schwartzkroin P. A. (1985) Do kainate-lesioned hippocampi become epileptogenic? Brain Res. 329, 309–313. Fritschy J. M., Benke D., Mertens S., Oertel W. H., Bachi T. and Mo¨hler H. (1992) Five subtypes of type A ã-aminobutyric acid receptors identified in neurons by double and triple immunofluorescence staining with subunitspecific antibodies. Proc. natn. Acad. Sci. U.S.A. 89, 6726–6730. Gall C. and Lauterborn J. (1992) The dentate gyrus: a model system for studies of neurotrophin regulation. Epilepsy Res. 7, 171–185. Gall C., Murray K. and Isackson P. J. (1991) Kainic acid-induced seizures stimulate increased expression of nerve growth factor mRNA in rat hippocampus. Molec. Brain Res. 9, 113–123.

1016 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52.

C. Schwarzer et al. Greber S., Schwarzer C. and Sperk G. (1994) Neuropeptide Y inhibits potassium-stimulated glutamate release through Y2 receptors in rat hippocampal slices in vitro. Br. J. Pharmac. 113, 737–740. Heggli D.E. and Malthe-Sorenssen D. (1982) Systemic injection of kainic acid: effect on neurotransmitter markers in piriform cortex, amygdaloid complex and hippocampus and protection by cortical lesioning and anticonvulsants. Neuroscience 7, 1257–1264. Johnson E. W., deLanerolle N. C., Kim J. H., Sundaresan S., Spencer D. D., Mattson R. H., Zoghbi S. S., Baldwin R. M., Hoffer P. B., Seibyl J. P. and Innis R. B. (1992) Central and peripheral benzodiazepine receptors: opposite changes in human epileptogenic tissue. Neurology 42, 811–815. Kamphuis W., De Rijk T. C. and Lopes da Silva F. H. (1995) Expression of GABAA receptor subunit mRNAs in hippocampal pyramidal and granular neurons in the kindling model of epileptogenesis: an in situ hybridization study. Molec. Brain Res. 31, 33–47. Kish S. J., Sperk G. and Hornykiewicz O. (1983) Alterations in benzodiazepine and GABA receptor binding in rat brain following systemic injection of kainic acid. Neuropharmacology 22, 1303–1309. Kokaia M., Pratt G. D., Elmer E., Bengzon J., Fritschy J.-M., Kokaia Z., Lindvall O. and Mo¨hler H. (1994) Biphasic differential changes in GABAA receptor subunit mRNA levels in dentate gyrus granule cells following recurrent kindling-induced seizures. Molec. Brain Res. 23, 323–332. Magloczky Z. and Freund T. F. (1995) Delayed cell death in the contralateral hippocampus following kainate injection into the CA3 subfield. Neuroscience 66, 847–860. Margerison J. H. and Corsellis J. A. N. (1966) Epilepsy and the temporal lobes: a clinical, electroencephalographic and neuropathological study of the brain in epilepsy, with particular reference to the temporal lobes. Brain 89, 499–530. Marksteiner J., Ortler M., Bellmann R. and Sperk G. (1990) Neuropeptide Y biosynthesis is markedly induced in mossy fibers during temporal lobe epilepsy of the rat. Neurosci. Lett. 112, 143–148. Marksteiner J. and Sperk G. (1988) Concomitant increase of somatostatin, neuropeptide Y and glutamate decarboxylase in the frontal cortex of rats with decreased seizure threshold. Neuroscience 26, 379–385. Marksteiner J., Wahler R., Bellmann R., Ortler M., Krause J. E. and Sperk G. (1992) Limbic seizures cause pronounced changes in the expression of neurokinin B in the hippocampus of the rat. Neuroscience 49, 383–395. McDonald J. W., Garofalo E. A., Hood T., Sackellares J. C., Gilman S., McKeever P. E., Troncoso J. C. and Johnston M. V. (1991) Altered excitatory and inhibitory amino acid receptor binding in hippocampus of patients with temporal lobe epilepsy. Ann. Neurol. 29, 529–541. Mitchell J., Sundstrom L. E. and Wheal H. V. (1993) Microglial and astocytic cell responses in the rat hippocampus after an intracerebroventricular kainic acid injection. Expl Neurol. 121, 224–230. Nadler J. V. (1981) Kainic acid as a tool for the study of temporal lobe epilepsy. Life Sci. 29, 2031–2042. Niquet J., Jorquera I., Ben-Ari Y. and Represa A. (1994) Proliferative astrocytes may express fibrinonectin-like protein in the hippocampus of epileptic rats. Neurosci. Lett. 180, 13–16. Nobrega J. N., Kish S. J. and Burnham W. M. (1989) Autoradiographic analysis of benzodiazepine binding in entorhinal-kindled rat brains. Brain Res. 498, 315–322. Otis T. S., DeKoninck Y. and Mody I. (1994) Lasting potentiation of inhibition is associated with an increased number of ã-aminobutyric acid type A receptors activated during miniature inhibitory postsynaptic currents. Proc. natn. Acad. Sci. U.S.A. 921, 7698–7702. Ro¨der C., Schwarzer C., Vezzani A., Gobbi M., Mennini T. and Sperk G. (1996) Autoradiographic analysis of neuropeptide Y receptor binding sites in the rat hippocampus after kainic acid-induced limbic seizures. Neuroscience 70, 47–55. Schwarzer C. and Sperk G. (1995) Hippocampal granule cells express glutamic acid decarboxylase-67 (GAD67) after limbic seizures in the rat. Neuroscience 69, 705–709. Shin C., Pedersen H. B. and McNamara J. O. (1985) ã-aminobutyric acid and benzodiazepine receptors in the kindling model of epilepsy: a quantitative radiohistochemical study. J. Neurosci. 5, 2696–2701. Sloviter R. S. (1992) Possible functional consequences of synaptic reorganization in the dentate gyrus of the kainate-treated rats. Neurosci. Lett. 137, 91–96. Sommer W. (1880) Erkrankung des Ammonshorns als aetiologisches Moment der Epilepsie. Arch. Psychiat. Nervkrankh. 10, 631–675. Sperk G. (1994) Kainic acid seizures in the rat. Prog. Neurobiol. 42, 1–32. Sperk G., Lassmann H., Baran H., Kish S. J., Seitelberger F. and Hornykiewicz O. (1983) Kainic acid induced seizures: Neurochemical and histopathological changes. Neuroscience 10, 1301–1315. Sperk G., Marksteiner J., Gruber B., Bellmann R., Mahata M. and Ortler M. (1992) Functional changes in neuropeptide Y and somatostatin containing neurons induced by limbic seizures in the rat. Neuroscience 50, 831– 846. Sperk G., Schwarzer C., Tsunashima K., Fuchs K. and Sieghart W. (1997) GABAA-Receptor subunits in the rat hippocampus I: immunocytochemical distribution of thirteen subunits, Neuroscience 80, 987–1000. Sternberger L. (1979) Immunohistochemistry, 2nd edn, John Wiley and Sons, New York. Titulaer M. N. G., Ghijsen W. E. J. M., Kamphuis W., Derijk T. C. and Lopes da Silva H. L. (1995) Opposite changes in GABA(A) receptor function in the CA1–3 area and fascia dentata of kindled rat hippocampus. J. Neurochem. 64, 2615–2621. Titulaer M. N. G., Kamphuis W. and Lopes da Silva F. H. (1995) Long-term and regional specific changes in [3H]flunitrazepam binding in kindled rat hippocampus. Neuroscience 68, 399–406. Titulaer M. N. G., Kamphuis W. and Lopes da Silva H. L. (1995) Autoradiographic analysis of [35S]tbutylbicyclophosphorothionate binding in kindled rat hippocampus shows different changes in CA1 area and fascia dentata. Neuroscience 66, 547–554. Titulaer M. N. G., Kamphuis W., Pool C. W., Van Heerikhuize J. J. and Lopes da Silva F. H. (1994) Kindling induces time-dependent and regional specific changes in the [3H]muscimol binding in the rat hippocampus: a quantitative autoradiographic study. Neuroscience 59, 817–826. Tsunashima K., Schwarzer C., Kirchmair E., Sieghart W. and Sperk G. (1997) GABAA-Receptor subunits in the rat hippocampus III: Altered expression of their mRNAs in kainic acid-induced epilepsy, Neuroscience 80, 1019–1032.

GABAA receptor subunits in limbic seizures 53.

1017

Wieser H. G., Engel J. J., Williamson P. D., Babb T. L. and Gloor P. (1993) Surgically remediable temporal lobe epilepsy syndromes, In Surgical Treatment of the Epilepsies (ed. Engel J. Jr.) pp. 49-63, Raven Press, New York. 54. Wisden W., Herb A., Wieland H., Keinanen K., Lu¨ddens H. and Seeburg P. H. (1991) Cloning, pharmacological characteristics and expression pattern of the rat GABAA receptor á4 subunit. Fedn Eur. biochem. Socs Lett. 289, 227–230. 55. Zezula J., Fuchs K. and Sieghart W. (1991) Separation of á1-, á2- and á3-subunits of the GABAA benzodiazepine receptor complex by immunoaffinity chromatography. Brain Res. 563, 325–328. (Accepted 19 March 1997)