Address correspondence and reprint requests to Holopainen I. E., MD, PhD, Department of Pharmacology and Clinical Pharmacology, PharmaCity, University of Turku, Itäinen Pitkäkatu 4, FI-20520 Turku, Finland. E-mail: email@example.com
Kainic acid-induced status epilepticus leads to structural and functional changes in inhibitory GABAA receptors in the adult rat hippocampus, but whether similar changes occur in the developing rat is not known. We have used in situ hybridization to study status epilepticus-induced changes in the GABAAα1–α5, β1–β3, γ1 and γ2 subunit mRNA expression in the hippocampus of 9-day-old rats during 1 week after the treatment. Immunocytochemistry was applied to detect the α1, α2 and β3 subunit proteins in the control and treated rats. In the saline-injected control rats, the α1 and α4 subunit mRNA expression significantly increased between the postnatal days 9–16, whereas those of α2, β3 and γ2 subunits decreased. The normal developmental changes in the expression of α1, α2, β3 and γ2 subunit mRNAs were altered after the treatment. The immunostainings with antibodies to α1, α2 and β3 subunits confirmed the in situ hybridization findings. No neuronal death was detected in any hippocampal subregion in the treated rats. Our results show that status epilepticus disturbs the normal developmental expression pattern of GABAA receptor subunit in the rat hippocampus during the sensitive postnatal period of brain development. These perturbations could result in altered functional and pharmacological properties of GABAA receptors.
Gamma-aminobutyric acid (GABA), the major inhibitory neurotransmitter in the adult CNS, exerts its inhibitory action mainly through GABAA receptors, which form a ligand-gated chloride channel (Sieghart 1995; Costa 1998). The receptor protein is a heteropentameric structure composed of multiple subunits of different families and isoforms (α1–6, β1–3, γ1–3, ρ1–3, δ, ɛ, θ, π). The subunit combinations vary in different brain regions, cell types and within a single cell, and thus determine the functional and pharmacological properties of the receptor (Sieghart 1995; Costa 1998; Korpi et al. 2002).
Reduced GABAergic inhibition, in particular in the hippocampus, is proposed to be a key determinant for seizure generation in epilepsy (Rice et al. 1996; Tsunashima et al. 1997; Brooks-Kayal et al. 1998b). Repeated seizures in adult animals lead to loss of subpopulations of GABAergic interneurones, which reduces the presynaptic inhibitory drive to excitatory principal cells (Bouilleret et al. 2000). The postsynaptic component of the inhibition is also affected, and at the receptor level, abnormalities in the epileptic brain presumably largely relate to the altered expression of genes encoding different GABAA receptor subunits (Rice et al. 1996; Schwarzer et al. 1997; Brooks-Kayal et al. 1998b; Loup et al. 2000). Consequently, both short- and long-term alterations have been detected in the number, function and pharmacology of GABAA receptors in adult experimental epileptic animals and in humans with epilepsy (Gibbs et al. 1997; Loup et al. 2000; Andre et al. 2001; Laurén et al. 2003). However, it is not known whether and to what extent the subunit expression is modulated by seizures in the developing brain. A recent study by Zhang et al. (2004) suggests age-dependent differences in the effects of status epilepticus (SE) on the expression of hippocampal GABAA receptors. In immature organotypic hippocampal slice cultures, neuronal activity can regulate GABAA receptor subunit mRNA expression in a region- and subunit-specific manner (Holopainen and Laurén 2003) suggesting that altered activity, i.e. prolonged seizures, could also affect subunit mRNA expression in the hippocampus of immature rats.
The present in situ hybridization study was undertaken to clarify the acute (6 h) and subacute (3 and 7 days) changes in the GABAA receptor subunit mRNA expression after kainic acid (KA)-induced SE in immature, 9-day-old rats. The α1–5, β1–3 and γ1–2 subunits were chosen for our study, since these subunits are abundantly expressed in the postnatal rat hippocampus (Laurie et al. 1992). Moreover, immunocytochemistry was used to verify the α1, α2 and β3 subunit expression in control and KA-treated rats 6 h and 1 week after SE. Our results show that KA-induced SE profoundly disturbs the normal developmental expression pattern of several GABAA receptor subunit mRNAs in the hippocampus during the critical second postnatal week. During this period GABA turns from the trophic, excitatory transmitter to the main inhibitory one (Ben-Ari 2001; Ben-Ari et al. 2004).
Kainate treatment of rats
SE was induced with a single dose of 2 mg/kg of KA (Tocris Cookson Ltd, Avonmouth, UK) given intraperitoneally (i.p.) to 9-day-old Sprague-Dawley rat pups, which were then carefully followed up to detect signs of seizures. Within 5 min after the injection, pups first developed deep breathing and salivation was increased. These signs were rapidly followed by scratching, forelimb and hindlimb clonus, and finally by generalized tonic/clonic seizures within 12–15 min. One hour after the KA injection, 0.3 mL/kg of paraldehyde (Fluka Chemie AG, Buchs, Switzerland) was injected i.p. to stop seizures. Shortly after that, 0.9% NaCl was injected (i.p.) to the pups at a dose of 2.5% of the body weight to help the recovery. After a careful follow-up of about 2 h, pups were taken back to their lactating mothers and used for the in situ hybridization studies 6 h, 3 days and 1 week (n = 4–5 in each age group) after KA injection at the age of 9, 12 and 16 days, respectively. Because the expression of GABAA receptor subunit mRNAs rapidly changes during the postnatal development in rats (Laurie et al. 1992), age-matched control rats were used for each group of KA-treated rats, i.e. 9-, 12- and 16-day-old rats for the 6-h, 3-day and 1-week groups, respectively. Control rats (n = 4–5 in each age group) received the same volume of 0.9% NaCl as those of KA-treated rats, but in order to minimize the discomfort of the experimental animals, they did not receive any further injections. The time of maternal separation and the handling of the animals after the injections until being killed were similar for the control and KA-treated rats. For the in situ hybridization, rats were quickly decapitated, brains removed, frozen in isopentane and stored at −70°C until cut into coronal sections (14 µm). Sections were mounted on poly-l-lysine-coated slides, dried and stored at −70°C until further processed. All animal procedures were conducted in accordance with the guidelines set by the European Community Council Directives 86/609/EEC and had the approval of the Animal Use and Care Committee of the University of Turku. All efforts were made to minimize the pain, discomfort and number of the experimental animals.
In situ hybridization
Sections were thawed, fixed in ice-cold 4% paraformaldehyde (Sigma, St Louis, MO, USA, prepared in phosphate-buffered saline, pH 7.0), washed in phosphate-buffered saline, dehydrated in 70% ethanol and stored in 95% ethanol at 4°C until used for the in situ hybridization, which was performed as recently described (Sinkkonen et al. 2001; Laurén et al. 2003). The 45-bases-long antisense oligonucleotide probes (Institute of Biotechnology, University of Helsinki, Finland) were complementary to rat cDNA sequences for the GABAA receptor subunits as follows: α1 (cDNA nucleotides 1240–1284; GenBank Accession No. L08490), α2 (1045–1089; L08491), α3 (1426–1470; L08492), α4 (133–177; L08493), α5 (1416–1460; LO8494), β1 (1296–1340; X15466), β2 (1292–1336; X15467), β3 (1283–1327; X15468), γ1 (1207–1251; X57514) and γ2 (1170–1214; L08497). The probes were labelled at their 3′ end with [α-33P]deoxy-ATP (NEN, Boston, MA, USA) and terminal transferase (Finnzymes Oy, Espoo, Finland). Unincorporated nucleotides were separated by ProbeQuantTM G-50 Micro columns (Amersham Biosciences, Piscataway, NJ, USA) and labelling efficiency was determined with a scintillation counter (Wallac 1410, PerkinElmer, Boston, MA, USA).
Sections were incubated (42°C, overnight) with 100 µL of hybridization buffer (50% formamide, 10% dextran sulphate and 4 × saline-sodium citrate) containing about 0.06 nm of the labelled probe (7–9 × 105 cpm). After hybridization, sections were washed in 1 × saline-sodium citrate at room temperature for 10 min, 1 × saline-sodium citrate at 55°C for 30 min and finally dehydrated by immersion through the following 3-min washing steps at room temperature: 1 × saline-sodium citrate, 0.1 × saline-sodium citrate, 70% ethanol, and 95% ethanol. The sections were allowed to air-dry before exposure to Biomax MR-1 film (Eastman Kodak, Rochester, NY, USA) with 14C-labelled standards (Amersham, Buckinghamshire, UK) at 4°C for 5–54 days. Signal specificity was assessed by competition experiments, in which the radiolabelled probes were hybridized in the presence of excess (100-fold) unlabelled probe. This resulted in completely blank autoradiographs.
Quantification of autoradiographic films
The intensity of the hybridization signal in film autoradiograms was quantified by measuring the optical density (OD) values using the MCID AIS image analysis devices and programs (Imaging Research, St. Catharines, Ontario, Canada). Tissue-equivalent hybridization values, in nCi/g, were calculated from the OD values of the calibrated 14C standards in four hippocampal regions; the pyramidal cell layers CA3c, CA3a/b and CA1, and the granule cell layer of the dentate gyrus (DG), which were carefully outlined, and examined for each subunit. The CA3c region was defined as the CA3 layer between the blades of the DG granule cell layers, and CA3a/b as the CA3 cell layer excluding the CA3c region. For each rat, the OD values for various regions are means of three to four sections. The hybridization signal values of the studied GABAA receptor subunit mRNAs did not differ between the right and left side in either control or KA-treated rats (both p > 0.05, all ages separately, Student's independent two-tailed t-test), and therefore data from both sides were combined and used for further analysis. Images were produced by scanning the films with an HP ScanJet 4c/T scanner using HP DeskScan II program (Hewlett Packard, Palo Alto, CA, USA), and further processed using Adobe PhotoShop (version 5.5, Adobe Systems, Mountain View, CA, USA), and Corel Draw programs (version 11.0, Corel Corporation, Ottawa ON, Canada).
Nine-day-old rat pups were used for the immunocytochemical study 6 h (P9) and 7 days (P16) after KA injection together with their respective age-matched saline controls (n = 3 in each group). The staining procedure has recently been described in detail (Lopez-Picon et al. 2003, 2004). Briefly, rats were deeply anaesthetized with an injection (i.p.) of 50 mg/kg of sodium pentobarbital, transcardially rinsed with 0.9% NaCl, and thereafter perfused with 4% paraformaldehyde in phosphate-buffered saline (pH 7.4). All these procedures were carried out at room temperature. After that, brains were rapidly removed, postfixed at 4°C, and then processed with an antigen retrieval protocol to reduce the non-specific staining and to enhance the detection of the epitopes as published by Fritschy et al. (1998). Thereafter, brains were cryoprotected in 30% sucrose in phosphate-buffered saline at 4°C, frozen and kept at −80°C until used. For the immunostaining, brains were cryosectioned in 40-µm slices, collected in Tris-saline (pH 7.4) and processed in a free-floating system. Slices were first incubated in the blocking solution containing 2% bovine serum albumin, 2% goat serum and 0.1% Triton X-100 in Tris-saline (pH 7.4) for 1 h at room temperature, and thereafter with the primary antibodies for 24 h at 4°C in blocking solution at the following dilutions: rabbit anti-α1 (1:10 000), guinea pig anti-α2 (1:2000), and rabbit anti-β3 (1:1000). Specific antibodies for α1 and α2 subunits were kindly provided by Professor J.-M. Fritschy, University of Zürich, and the specific antibody for β3 subunit was purchased from Abcam (Abcam Ltd, Cambridge, UK). After washing in Tris-saline-Triton X-100 (0.1%), slices were incubated with the biotin-SP-conjugated secondary antibodies (Jackson Immunoresearch Laboratories, Inc., West Grove, PA, USA), goat anti-rabbit for α1 (1:4000) and β3 (1:1000), and goat anti-guinea pig for α2 (1:2000) in blocking solution for 1 h at room temperature, rinsed in Tris-saline-Triton X-100, and incubated with the avidin-peroxidase conjugate (Vectastain ABC Kit, Vector Laboratories, Burlingame, CA, USA) in blocking solution for 1 h at room temperature. The staining was detected using 3,3′-diaminobenzidinetetrahydrochloride (Sigma) as a chromogen, and further processed as earlier described (Lopez-Picon et al. 2003). In each experiment, brain slices of all four rat groups (P9, control and KA-treated; P16, control and KA-treated) were processed simultaneously to minimize the possible interexperimental differences in the staining. In all experiments, three to four slices in which the primary antibody was omitted but were otherwise treated as indicated above, served as negative controls. Olympus U-TV1 X digital camera (Olympus Optical Co., Ltd, Tokyo, Japan) was used to capture pictures using Olympus BX60 microscope (Olympus), and images were further processed using Adobe Photoshop (version 6.0) and Corel Draw (version 11.0).
Fluoro-Jade B and thionin stainings
Fluoro-Jade B (FJB) staining was used to detect degenerating neurones (Schmued et al. 1997), and thionin staining to detect neuronal loss in brain sections taken from the rats 6 h, 3 days and 1 week after the treatment, and from their 9-, 12- and 16-day-old control rats, respectively. Rats were deeply anaesthetized with an injection (i.p.) of 50 mg/kg of sodium pentobarbital, transcardially flushed with 0.9% NaCl, and thereafter perfused with 4% paraformaldehyde in phosphate-buffered saline (pH 7.4). After that, brains were rapidly removed, postfixed, cryoprotected in 30% sucrose in phosphate-buffered saline at 4°C, frozen, and thereafter kept at −80°C until used. For the staining, brain sections (20-µm thick) were cut, and mounted on gelatin-coated glasses. The FJB staining was carried out as previously described in detail for cultured organotypic hippocampal slices (Holopainen et al. 2004) with minor modifications. Brain sections were rehydrated in alcohol series, transferred for 2–5 min to 0.06% potassium permanganate (KMnO4) prepared in distilled water, washed twice with distilled water, and transferred to 0.001% FJB solution for 30 min. After the staining, sections were washed with distilled water, dried, cleared in xylene and coverslipped with the DePeX mounting medium (BDH Laboratory Supplies, Poole, England).
For thionin staining, brain sections (20-µm thick) from the same perfusion-fixed rats as used for the FJB staining were first rehydrated, stained in 0.1% thionin for 15–20 s, washed in water to remove the excess colour, dehydrated in alcohol series, cleared in xylene and mounted with Permount. The digital camera (Olympus U-TV1 X, Olympus Optical, Tokyo, Japan) was used to capture images using the Olympus BX60 microscope (Olympus Optical), which were further processed using Adobe Photoshop (version 6.0) and Corel Draw (version 11.0).
All statistical analyses were carried out using Prism software (version 4.0, GraphPad Software, Inc, San Diego, CA, USA). Regional differences in the expression of each subunit in the control (n = 4–5), and KA-treated rats (n = 4–5) in the three experimental age groups were analysed using one-way anova with Tukey-Kramer Multiple Comparison as a post hoc test. Student's independent two-tailed t-test was used to assess differences between control and epileptic rats of the corresponding age in each hippocampal subregion separately. The level of significance was set at p < 0.05. All values are given as a mean (± SEM).
Subunit mRNA expression in control rats
In control rats, the expression levels of several GABAA receptor subunit mRNAs varied significantly during the postnatal development (from P9 to P16) in the CA3a/b, CA3c, CA1 and DG regions. The α1 mRNA expression significantly increased between P9 and P16 in CA3a/b (by 22%), CA3c (by 35%) and DG (by 52%), whereas α2 mRNA expression decreased in CA3c (by 52%), CA3a/b (by 47%), DG (by 33%) and CA1 (by 55%), most pronouncedly between P9 and P12. Also the expression of α4 mRNA increased in CA1 (by 53%) and DG (by 41%), but no significant developmental changes were detected in the expression of α3 and α5 subunit mRNAs in any subregion (Fig. 1). The expression of β1 and β2 mRNA remained relatively constant through P9 to P16, whereas that of β3 mRNA significantly decreased between P9 and P12 in all subregions (by about 50% in CA3a/b and CA3c, and by about 45% in DG and CA1), but did not significantly change between P12 and P16 (Fig. 2). Moreover, the γ2 mRNA expression significantly decreased between P12 and P16 in all subregions (in CA3a/b by 60%; CA3c and CA1 by 50%; and DG by 55%), but no significant age-related changes were detected in γ1 mRNA expression (Fig. 3).
In addition to developmental changes, the expression levels of α1 (p < 0.001), α2, α3 (both p < 0.05) and α4 (p < 0.01) mRNAs significantly varied between the hippocampal subfields in the 9-day-old control rats. The expression of α1 mRNA decreased in the order CA1 = CA3a/b > CA3c > DG, and that of α2 in the order CA3a/b > CA3c > CA1 = DG. The expression of α4 mRNA was highest in DG, whereas no significant regional variation was detected in the expression of mRNAs encoding α5, β1, β2, β3, γ1 and γ2 subunits.
Subunit mRNA expression after kainic acid-induced status epilepticus
The normal developmental expression patterns of α1, α2, α3, α4, α5, β3 and γ2 subunits were significantly altered after KA-induced SE in all (α2, α3, β3, γ2) or in some (α1, α4, α5) hippocampal subregions as compared to their age-matched control rats (Figs 1–3). The normal developmental down-regulation in the expression of mRNAs encoding α2 and β3 subunits, and the developmental up-regulation of α1 subunit mRNA in the CA3 pyramidal cell regions were not detected in KA-treated rats. The expression of α3 subunit mRNA, after remaining unchanged between P9 and P12 as in the control rats, was significantly up-regulated in all subregions 1 week after SE. Also significant, but less pronounced changes were detected in α4 mRNA expression in the CA3 regions, and α5 expression in CA3a/b and CA1. Moreover, the normal significant developmental down-regulation in γ2 expression was absent in all subregions.
The comparison of GABAA receptor subunit mRNA expression between KA-treated and their age-matched controls at various time-points revealed further significant differences. The α1 mRNA expression was down-regulated in CA3c 1 week after SE, whereas β1 mRNA expression was up-regulated 6 h, and β3 subunit mRNA 1 week after SE, both in CA1. On the contrary, γ2 mRNA expression was down-regulated in all subregions 6 h and 3 days after SE, and γ1 expression in the CA3c, CA1, and DG regions 6 h after SE. Representative in situ hybridization images of α1, α2, α4, and β3 mRNA expression in P9 and P16 control and KA-treated rats 6 h (P9) and 1 week (P16) after SE are shown in Fig. 4.
Figure 5 shows the results of the α1, α2 and β3 immunostainings in the hippocampus 6 h (KA 6 h) and 7 days (KA 1 week) after SE together with their age-matched controls, C P9 and C P16, respectively. The results of the immunostaining correlated well with the changes detected in the mRNA expression in the control rats during the development, and in the KA-treated rats. There were no clear-cut differences in the localization of the α1, α2, and β3 stainings between the treated and control rats either 6 h or 7 days after SE. In control rats, the α1 immunoreactivity increased (Figs 5a and c), whereas the α2 (Figs 5e and g), and β3 (Figs 5i and k) immunoreactivities decreased during the developmental maturation. In the KA-treated rats, the developmental changes in the α1 (Figs 5b and d) and α2 (Figs 5f and h) immunoreactivities were less pronounced than those in the control rats, in particular in the stratum oriens and radiatum of the CA3 and CA1 regions. In both control and KA-treated rats, the α1 immunostaining was mainly localized in the processes of the main cell types in CA1, CA3, and DG, whereas the α2 immunostaining was intense in cell bodies of CA3, CA1, and dentate granule cells. No clear differences were detected in the β3 immunostaining between the KA-treated and control rats either 6 h (Figs 5i and j) or 1 week (Figs 5j and l) after SE, although the developmental decrease in the staining intensity was clear. The moderate β3 immunostaining at P9 was localized in the cell processes within the stratum oriens and radiatum of CA3 and CA1, and in the molecular layers of the DG. No immunoreactivity was detected in cell bodies at both ages either in control or KA-treated rats. Figure 6 shows representative images of the α1 (Figs 6a–d) and α2 (Figs 6e–h) immunostainings in the CA3 and DG regions of P9 and P16 control rats at the higher magnification. During the development, the number and the staining intensity of α1 interneurones increased in the CA3 and DG regions. The α2 immunostaining was pronounced in CA3 pyramidal cell bodies (Fig. 6e) and dentate granule cells (Fig. 6g), but decreased during the development (Figs 6f and h). Also interneurones in the stratum oriens and radiatum of CA3 and DG were stained with α2 as pointed by the black arrows.
Fluoro-Jade B and thionin stainings
No degenerating neurones were detected after FJB staining in any time point after SE in any hippocampal subregion. This finding was further verified with thionin staining, which showed that the pyramidal CA3 and CA1 cell layers and the granule cell layers of the DG were well preserved in all age groups after KA-induced SE. Representative FJB staining images of the hippocampus of a KA-treated rat 3 days (Fig. 7a) and 1 week (Fig. 7b) after SE, and thionin stainings in a P16 control rat (Fig. 7c) and in a KA-treated rat 1 week (P16) (Fig. 7d) after SE are shown.
Two major findings emerged from our present in situ hybridization and immunocytochemical study, in which we clarified the acute (6 h) and subacute (up to 1 week) changes in the GABAA receptor subunit mRNA expression in immature rats after KA-induced SE. First, in the control rats, the expression of GABAA receptor α1, α2, α4, β3 and γ2 subunit mRNAs dramatically changed in the hippocampus during the short time period of postnatal brain development, i.e. between P9 and P16. These developmental changes in the α1, α2 and β3 subunit expression were also confirmed at the subunit protein level with the immunocytochemical method. Secondly, KA-induced SE profoundly altered in a time-dependent, subunit- and region-specific manner the normal developmental expression pattern of mRNAs encoding α1, α2 and β3 subunits, and to some extent also those of α3, α4, α5 and γ2 subunits in the subregions of the hippocampus.
GABAA receptor subunit expression in the developing hippocampus
The observed developmental expression patterns of mRNAs encoding GABAA receptor subunits are in good agreement with previous in situ hybridization and immunocytochemical studies in the developing hippocampus of non-handled rats (Laurie et al. 1992; Poulter et al. 1992; Fritschy et al. 1994; Lopez-Tellez et al. 2004; Ramos et al. 2004). However, our present study gives the most detailed information of the developmental changes occurring during the critical, second postnatal week in the rat hippocampus. Most importantly, our results show the very rapid reciprocal alterations in the expression of α1 and α2 subunit mRNAs and proteins. The expression of α1 mRNA increased in the CA3 and DG regions between P9 and P16, whereas that of α2 decreased, in particular between P9 and P12 in all subregions. The developmental increase in α1 subunit mRNA expression, decrease in α2, and their coexistence during a short postnatal period in many individual neurones (Fritschy et al. 1994), although at different cellular compartments, as shown in our present immunocytochemical study, indicate that GABAA receptors containing α2 subunit are gradually being replaced by receptors containing the α1 subunit. This switch has been proposed to indicate the appearance of a new, prevalent receptor subtype involved in synaptic inhibition (Fritschy et al. 1994; Hevers and Lüddens 2002), the process being mainly completed by the second postnatal week in rats (Ben-Ari 2001; Khazipov et al. 2004).
The developmental expression patterns of β3 mRNA were strikingly similar to that of α2 mRNA. This is in keeping with the immunocytochemical study, which has shown the colocalization of α2 and β2/β3 subunit immunoreactivity within single pyramidal and dentate granule cells thus indicating their coordinated expression in the developing rat hippocampus (Fritschy et al. 1994). Our results with the specific α2 and β3 antibodies, however, suggest that in spite of the similar developmental expression profiles of α2 and β3 subunits, their cellular localization seems to be different, the α2 immunoreactivity being mainly localized in the cell bodies, whereas β3 immunoreactivity is more intense in the processes. Moreover, although the intensity of staining decreased during the development, the subregional distribution of immunostaining did not change between P9 and P16. Thus, the maturation of at least some GABAA receptor subunits is a tightly regulated process, which is also preserved in cultured hippocampal slices (Holopainen and Laurén 2003), and in primary neurones (Brooks-Kayal et al. 1998a). It remains to be ascertained whether or not the regulatory elements in β3 and α2 subunit genes show similarities (see Steiger and Russek 2004).
Status epilepticus arrests the normal developmental expression pattern of GABAA receptor subunit mRNAs in the rat hippocampus
Our novel results show that KA-induced SE in 9-day-old rats abolished the normal developmental expression pattern of mRNAs encoding α1, α2, α3, β3 and γ2 GABAA receptor subunits in many hippocampal subregions during the follow-up period of 1 week. The absence of normal developmental alterations in the subunit expression, especially in those encoding α1 and α2 subunits, may have long-term consequences on many developmental processes. For example, it has been shown that seizure activity can disturb the maturation of the inhibitory synaptic transmission and its strength (Seil and Drake-Baumann 1994; McLean et al. 1996; Kotak and Sanes 2000; Marty et al. 2000). The α1 subunit expression is regulated in an activity-dependent manner, and at the inhibitory synapse, α1 subunit may regulate the strength of inhibitory synapses by controlling the duration of inhibitory postsynaptic currents (IPSCs) (Hevers and Lüddens 2002). Furthermore, deficiency of the α1 subunits in knockout mice impairs normal development of dendritic spines (Heinen et al. 2003), indicating that this subunit is needed for the development and maintenance of normal synaptic contacts. Conversely, the expression of α2 subunit is suggested to be down-regulated by the inhibitory synaptic activity (Hevers and Lüddens 2002). The enhanced excitation in our KA-treated rats may disturb the normal developmental α2 subunit down-regulation. Moreover, the abnormally high α2 subunit levels may disturb synaptogenesis, which is suggested to coincidence the developmental down-regulation of α2 subunit-containing receptors (Fritschy et al. 1994). The absence of β3 subunit mRNA down-regulation and the lack of any clear changes in the β3 immunostaining in our KA-treated rats further confirmed the coordinated expression of α2 and β3 subunits also in conditions with altered balance between excitation and inhibition.
The importance of the fine-tuned regulation of relative GABAA receptor α1 and α2 subunit expression is also observed in cerebellar and hypothalamic neurones, in which the kinetics of miniature inhibitory postsynaptic currents are determined by the ratios of the subunit expression (Brussaard et al. 1997; Vicini et al. 2001). The SE-induced arrest of the normal maturation pattern of GABAA receptor subunit expression detected now is surprisingly similar to that observed in a recent study, in which P9 rats were exposed to maternal separation with repeated handling (Hsu et al. 2003). This resulted in an immature GABAA receptor phenotype prevailing up to adulthood in the hippocampal DG of the affected rats. Moreover, long-term changes in the GABAergic synaptic activity have been induced by repetitive depolarizing pulses during a restricted period in the developing rats, i.e. between P0 and P8, but not after P12 (P9 was not studied) (Gubellini et al. 2001). These studies together with our present findings suggest the exquisite sensitivity of GABAA receptor subunit transcription to external stimuli during the early postnatal development. We should emphasize that in addition to the SE-induced regional changes in the subunit mRNA expression, the short-term alterations can be different from those of chronic ones. For example, in pilocarpine model of SE in P10 rats a significant increase in α1 subunit mRNA expression was detected 3 months after SE in the DG granule cells (Zhang et al. 2004), whereas at the present subacute phase this was not observed. It remains, however, unknown whether the observed alterations in GABAA receptor subunit expression in our study were due to SE alone, or whether and to what extent the abnormal excitatory activity might have continued after the paraldehyde injection and contributed to detected changes. We emphasize that no clinical signs of seizures were observed at the time when the pups were returned to their cages.
The significance of altered GABAA receptor subunit expression after status epilepticus
The early life seizures may have important long-term consequences in the maturation of GABAA receptors leading not only to functional changes but also to altered pharmacological profile of the receptor (Brooks-Kayal et al. 2001). The sedative and amnestic drug actions are primarily mediated by receptors containing a functional α1 subunit (Rudolph et al. 2001; Hevers and Lüddens 2002; Korpi et al. 2002), which is also important for seizure sensitivity to GABAA antagonists (Kralic et al. 2003) and for anticonvulsant efficacy of benzodiazepines (Rudolph et al. 2001). The α, γ and δ subunit variants can effect zinc sensitivity (Korpi et al. 2002). It has been shown that a decrease in α1 subunit mRNA expression together with increase in α4 and δ subunit mRNAs alters zinc sensitivity in DG granule cells (Brooks-Kayal et al. 1998b) and predisposes the adult hippocampal circuit to generate seizures (Cohen et al. 2003). However, the aberrant sprouting of zinc-containing mossy fibers in the DG, which first appears weeks after SE in adult rats, has infrequently been detected in immature epileptic rats (Haas et al. 2001; Bender et al. 2003). Thus the zinc-mediated collapse of inhibition observed in the pilocarpine model of epilepsy in adult rats (Cohen et al. 2003) may not operate in immature rats, as recently shown in isolated dentate granule cells (Zhang et al. 2004). The comprehensive pharmacological study, however, awaits to be performed in our model, although the earlier studies in adult rats suggest that alterations in the subunit expression in epileptic tissue affect both pharmacological profiles of GABAA receptors and seizure sensitivity (Fritschy et al. 1999; Cohen et al. 2003).
As a conclusion, SE in 9-day-old rats prevented the normal developmental expression pattern of several GABAA receptor subunits in the rat hippocampus. It has been shown that the K+-Cl– cotransporter 2 plays a central role in the excitation of GABA in immature tissues (Rivera et al. 1999; Ben-Ari et al. 2004), and that GABA itself is a critical maturation factor for the switch of the physiological and biochemical properties of GABA signalling (Ganguly et al. 2001). Based on our present results we suggest that SE-induced changes, especially in the α1 and α2 subunit mRNA expression, during the critical maturation period can profoundly disturb the normal development and lead to expression of altered receptor phenotypes with abnormal pharmacological profile. The question, whether and to what extent these changes persist into adulthood, needs long-term follow-up studies.
The financial support of the Sigrid Juselius Foundation, the Special State Grant for Clinical Research (EVO), the Foundation of University of Turku and Arvo and Lea Ylppó Foundation to IEH, the Svenska Kulturfonden, and the Research and Science Foundation of Farmos to HBL, and the Finnish Graduate School of Neuroscience to FL-P are gratefully acknowledged. The skilful Fluoro-Jade B staining by Tiina-Kaisa Kukko-Lukjanov, MSc is highly appreciated.