Ca2+-permeable AMPA receptors and selective vulnerability of motor neurons

Journal of the Neurological Sciences 180 (2000) 29–34 www.elsevier.com / locate / jns

Ca 21 -permeable AMPA receptors and selective vulnerability of motor neurons Ludo Van Den Bosch*, Wim Vandenberghe, Hugo Klaassen, Elisabeth Van Houtte, Wim Robberecht Laboratory of Neurobiology, Department of Neurology, University of Leuven, Campus Gasthuisberg, Herestraat 49, B-3000 Leuven, Belgium

Abstract To evaluate the role of excitotoxicity in the pathogenesis of amyotrophic lateral sclerosis (ALS), we compared the sensitivity of motor neurons and that of dorsal horn neurons to kainic acid (KA). Short exposure to KA resulted in the death of motor neurons, while dorsal horn neurons were unaffected. This selective motor neuron death was completely dependent on extracellular Ca 21 and insensitive to inhibitors of voltage-operated Ca 21 or Na 1 channels. It was also completely inhibited by the specific AMPA antagonist LY300164 and by Joro spider toxin (JSTx), a selective blocker of AMPA receptors that lack the edited GluR2 subunit. KA selectively killed those motor neurons that stained positive for the Co 21 histochemical staining, a measure for the presence of Ca 21 -permeable AMPA receptors. These results suggest that Ca 21 entry via Ca 21 -permeable AMPA receptors is responsible for the selective motor neuron death. As the Ca 21 permeability of the AMPA receptor is regulated by its GluR2 subunit, we stained motor neurons for GluR2. Immunoreactivity was present in all motor neurons, albeit to a variable degree. However, double-staining experiments demonstrated that motor neurons clearly expressing GluR2, also expressed Ca 21 -permeable AMPA receptors. This indicates that despite the abundant expression of GluR2, this subunit is excluded from a subset of AMPA receptors and that the activation of these receptors is responsible for the selective motor neuron death.  2000 Elsevier Science B.V. All rights reserved. Keywords: ALS; Excitotoxicity; Kainate; Motor neuron; Spinal cord; AMPA receptor; GluR2

1. Introduction Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disorder, characterized by the progressive loss of motor neurons in brainstem, spinal cord and motor cortex. Its pathogenesis is poorly understood and why motor neurons are specifically vulnerable in ALS is enigmatic. One possible explanation is that motor neurons are particularly sensitive to glutamate-induced excitotoxicity mediated by the a-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) / kainic acid (KA)-type of receptor [1–3]. This susceptibility may be due to unusual Ca 21 -permeability of these receptors on motor neurons. Permeability of the AMPA receptor for Ca 21 is determined by the edited GluR2 subunit: it’s presence in the

*Corresponding author. Tel.: 132-16-345785; fax: 132-16-345699. E-mail address: [email protected] (L. Van Den Bosch).

heteromeric receptor complex drastically reduces the channel’s permeability for divalent ions [4]. Low abundance or even absence of this subunit in motor neurons may explain the high Ca 21 permeability of their AMPA receptors, and thus their higher sensitivity to excitotoxicity [5–7]. Several groups have investigated the presence of the GluR2 subunit in motor neurons, but results are not entirely consistent. Most studies found evidence for the presence of GluR2 [8–19], but others correlated a low expression or absence of GluR2 with the selective vulnerability of motor neurons [5–7]. To approach the problem of selectivity, we cultured motor neurons and dorsal horn neurons in identical conditions and we compared the sensitivity of these two cell types to KA-induced excitotoxicity. We found that motor neurons were selectively vulnerable to KA-induced cell death as compared to dorsal horn neurons, that this cell death was Ca 21 -dependent and that it was selective for motor neurons containing Ca 21 -permeable AMPA receptors. Furthermore, we investigated whether the presence or

0022-510X / 00 / $ – see front matter  2000 Elsevier Science B.V. All rights reserved. PII: S0022-510X( 00 )00414-7

30

L. Van Den Bosch et al. / Journal of the Neurological Sciences 180 (2000) 29 – 34

absence of the GluR2 subunit was the factor determining the selective vulnerability of the motor neurons.

2. Materials and methods

2.1. Cell cultures Motor neurons and glial cells were prepared as described before [16,20]. Briefly, spinal cords were dissected from 14-day-old Wistar rat embryos in Hanks’ Balanced Salt Solution (HBSS) and the meninges and the dorsal root ganglia were removed. Fragments of the ventral cord were digested for 15 min with 0.05% trypsin at 378C. After treatment with DNase, the tissue was further dissociated by trituration. The cell suspension obtained was layered on a 6.5% (w / v in L15) metrizamide cushion (two spinal cord equivalents per tube), and after centrifugation at 500 g for 15 min, a sharp band (fraction ‘‘F1’’) on top of the metrizamide cushion and a pellet (fraction ‘‘F2’’) were obtained. The fractions were resuspended and centrifuged for 20 min at 75 g on a 4% BSA cushion and were resuspended in their respective plating media. Glial feeder layers were prepared by plating F2 cells in 35 mm dishes (12 000 cells / cm 2 ) in L15 medium supplemented with glucose (3.6 mg / ml), sodium bicarbonate (0.2%), penicillin (100 IU / ml), streptomycin (100 mg / ml) and fetal bovine serum (10%). In these cultures, glial cells rapidly proliferate and reach confluency after 1–2 weeks in vitro. In 4-week-old feeder layers 95% of cells are GFAPpositive. After 4 weeks in vitro glial cell division was halted by exposure to 10 mM cytosine arabinoside for 2–3 days. After 24 h cocultures were established by seeding F1 cells on top of the feeder layers at a density of 3000 cells / cm 2 . The culture medium consisted of L15 supplemented with sodium bicarbonate (0.2%), glucose (3.6 mg / ml), progesterone (20 nM), insulin (5 mg / ml), putrescine (0.1 mM), conalbumin (0.1 mg / ml), sodium selenite (30 nM), penicillin (100 IU / ml), streptomycin (100 mg / ml), and horse serum (2%). Dorsal horn neuronal cultures were prepared starting from the dorsal half of spinal cords from 14-day-old rat embryos. A similar protocol as for the motor neurons was used except that the metrizamide gradient centrifugation was omitted. After trypsinization and trituration, the cells were centrifuged on a 4% BSA cushion and seeded on a glial feeder layer at a density of 1500 cells / cm 2 . Cultures were kept in a humidified 7% CO 2 incubator at 378C. Media were refreshed twice a week.

2.2. Co 21 staining and immunocytochemistry Kainate-activated Co 21 labeling was performed as described [21], with minor modifications [16,20]. Cells were rinsed twice with uptake buffer (139 mM sucrose, 57.5 mM NaCl, 5 mM KCl, 2 mM MgCl 2 , 1 mM CaCl 2 , 12

mM glucose, 10 mM HEPES, pH 7.6) and then exposed for 30 min to 100 mM KA dissolved in uptake buffer containing 2.5 mM CoCl 2 . After washing the cells with uptake buffer containing 3 mM EDTA, they were incubated with 0.05% (NH 4 ) 2 S for 5 min to precipitate intracellular Co 21 . Subsequently, cultures were washed three times in uptake buffer and fixed in 4% paraformaldehyde / PBS for 30 min. Enhancement of the CoS precipitate was performed by incubating the cells at 50–558C with 0.1% AgNO 3 in 292 mM sucrose, 15.5 mM hydroquinone, 42 mM citric acid for 40–60 min. The silver solution was changed every 15 min. For the combined GluR2 staining and Co 21 uptake, the cells were first subjected to KA stimulated Co 21 loading (15 min) and the Co 21 staining procedure was performed except for the enhancement step. Subsequently, the cells were treated for 30 min with PBS containing 3% serum. Cultures were then incubated overnight at 48C with primary GluR2 antibody (2 mg / ml; rabbit polyclonal; Chemicon International, Temacula, CA) in PBS containing 1% blocking serum. This antibody is selective for the GluR2 subunit [14]. The cells were rinsed three times with PBS, followed by incubation with biotinylated secondary antibody for 30 min at room temperature. After washing the cells three times with PBS, the stain was developed using the ABC complex, H 2 O 2 and diaminobenzidine. Selected immunostained fields were photographed and subsequently the CoS precipitate was enhanced with the silver staining.

2.3. KA exposure and assessment of neuronal survival For the KA exposures, cultures were incubated for 20 min at 378C in a modified Krebs solution (135 mM NaCl, 5.9 mM KCl, 1.5 mM CaCl 2 , 1.2 mM MgCl 2 , 11.6 mM HEPES and 11.5 mM glucose, pH 7.3) containing KA and MK801 (10 mM). The antagonists were added to the cells 15 min prior to the KA exposure and afterwards the cells were washed once with Krebs before returning them to normal culture medium. In order to obtain Ca 21 free Krebs, CaCl 2 was omitted and EGTA (2 mM) was included. Krebs containing 10 mM Ca 21 was obtained by replacing an equimolar amount of NaCl by CaCl 2 and in the Na 1 free Krebs, NaCl was replaced by an equimolar amount of N-methyl-D-glucamine. Neuronal survival was quantified by direct counting of unfixed neurons under phase contrast optics (Nikon, Tokyo, Japan) at 1003. All phase-bright cells of neuronal morphology, without vacuolar inclusions and with intact neurites longer than two cell body diameters were taken into account. Neurons were counted immediately before and 24 h after KA exposure within an identified region of 1 cm 2 (610% of the total area of the dish) using a grid. The percentage neuronal survival was determined as the ratio of the number of neurons 24 h after the KA exposure to the number of neurons before the treatment, and was

L. Van Den Bosch et al. / Journal of the Neurological Sciences 180 (2000) 29 – 34

normalized to controls. Neurons were counted by an observer blinded to the treatment protocol.

2.4. Materials and statistics Media and additives were obtained from Gibco BRL (Grand Island, NY); tetrodotoxin (TTX) was from Calbiochem (San Diego, CA); normal goat serum, biotinylated secondary antibody and ABC complex were from DAKO (Glostrup, Denmark); 6-cyano-7-nitroquinoxaline-2,3dione-disodium (CNQX) and MK-801 was from Tocris Cookson (Bristol, UK); Joro spider toxin (JSTx) was from Research Biochemicals International (Natick, MA). LY300164 was provided by Dr. J.D. Leander from EliLilly (Lilly Corporate Center, Indianapolis, IN). All other chemicals were from Sigma (St. Louis, MO). All results are expressed as mean6S.E.M. Student’s t-test and oneway or two-way ANOVA were used as indicated.

3. Results To compare the KA-induced excitotoxicity in neurons affected and those that are spared in ALS, motor neurons and dorsal horn neurons were cultured in identical conditions on a glial feeder layer. At day 8 in vitro, the cultures were exposed for 20 min to KA in the presence of MK801, an inhibitor of the NMDA receptor. As is shown in Fig. 1, there was a dose-dependent motor neuron death, while the dorsal horn neurons were not affected by treatment with KA, not even at higher concentrations. This lack of effect on dorsal horn neurons of short term KA

Fig. 1. Survival of motor neurons (black bars) and of dorsal horn neurons (open bars) after short KA exposures. The neuronal cultures were cultured during 8 days on a glial feeder layer. Subsequently, the cells were treated during 20 min with different KA concentrations dissolved in Krebs (Ca 21 e 510 mM). Two-way ANOVA showed P50.01 for the interaction between motor neurons and dorsal horn neurons (P-values for student’s t-test of differences between types of neurons are indicated, n53).

31

exposure is not due to the complete absence of AMPA receptors on these cells as these neurons are clearly sensitive to long term (24 h) KA exposure [20] and they are clearly immunoreactive for the AMPA receptor subunits, GluR1 and GluR2 (data not shown). The motor neuron death could be blocked by CNQX, an antagonist of AMPA and KA receptors (Fig. 2A). KAinduced cell death was undetectable in the absence of external Ca 21 and was dose-dependently enhanced by extracellular Ca 21 (Fig. 2B). In contrast, Na 1 entry through the receptor channel did not appear to contribute, as the effect of KA was maintained and even somewhat increased if the KA exposure was performed in Na 1 -free Krebs (Fig. 2B). To further evaluate whether depolarization was necessary for the motor neuron death observed, the effect of TTX, a blocker of Na 1 channels, was studied. Neither basal survival nor KA-induced motor neuron death were affected by TTX (1 mM; data not shown). Furthermore, the blocker of voltage operated Ca 21 channels, NiCl (1 mM; Fig. 2A), or the L-type Ca 21 channel blocker, nifedipine (10 mM; data not shown), did not affect KA-induced cell death. This indicates that Ca 21 entry via voltage-operated Ca 21 channels does not contribute to the KA-induced motor neuron death. To unequivocally demonstrate involvement of the AMPA receptor and to exclude a role for the KA receptor, we studied the effect of the selective AMPA antagonist LY300164 [22]. As shown in Fig. 3A, this antagonist dose-dependently inhibited the KA-induced motor neuron death. To further characterize the AMPA receptors involved, we studied the effect of Joro spider toxin (JSTx) on KA-induced cell death. This toxin is known to selectively block AMPA receptors which lack the GluR2 subunit and

1 Fig. 2. Effect of CNQX (A) and of Ca 21 (B) on e , NiCl 2 and Na e KA-induced motor neuron death. On day 8 in vitro, motor neuron cocultures were exposed for 20 min to 300 mM KA and survival was determined 24 h after the exposure. The effect of CNQX was evaluated in Krebs containing 10 mM Ca e21 (student’s t-test for CNQX effect, n55; one-way ANOVA for the effect of Ca 21 e ; student’s t-test showed no statistically significant effect of 1 mM NiCl 2 and of 0 mM Na 1 e , n53).

32

L. Van Den Bosch et al. / Journal of the Neurological Sciences 180 (2000) 29 – 34

Fig. 3. Effect of LY300164 (A) and of JSTx (B) on KA-induced motor neuron death. On day 8 in vitro, motor neurons were exposed for 20 min to 300 mM KA dissolved in Krebs (10 mM Ca 21 e ) in the absence and presence of increasing concentrations of LY300164 or JSTx (one-way ANOVA, n53).

therefore are highly Ca 21 -permeable [23]. JSTx completely abolished neuronal death induced by KA exposure (Fig. 3B). All together these data indicate that Ca 21 entry via Ca 21 -permeable AMPA receptors is responsible for KAinduced motor neuron death. The presence of Ca 21 -permeable AMPA receptors can be visualised by the Co 21 histochemical staining [21]. Those cells that stain positive for Co 21 contain Ca 21 permeable AMPA receptors. In the motor neuron cultures, 5062% of the cells showed KA-induced Co 21 -uptake, while in the dorsal horn cultures only 661% of neurons were positive (Fig. 4A). JSTx as well as LY300164 blocked the KA-induced Co 21 -uptake in motor neurons (data not shown) confirming that Co 21 is indeed entering

Fig. 4. (A) Relative number of Co 21 -positive motor neurons (MN) and dorsal horn neurons (DHN) after 8 days in vitro (student’s t-test, n53). Cells were exposed for 30 min to 100 mM KA in uptake buffer containing 2.5 mM CoCl 2 . Afterwards the CoS precipitate was visualised using a silver staining. (B) Effect of a short (20 min) exposure to KA (Ca 21 e 510 mM) on the relative number of Co 21 -positive neurons in the population of surviving motor neurons. The Co 21 staining was performed 24 h after the short KA exposure by exposing the motor neurons for 30 min to 100 mM KA in uptake buffer containing 2.5 mM CoCl 2 and the intracellular CoS was visualised with a silver staining (student’s t-test, n53).

Fig. 5. Double-staining of motor neurons on day 5 in vitro. The cells were first subjected to KA-stimulated Co 21 loading (15 min) and the Co 21 staining procedure was performed except for the enhancement step. Subsequently, the cells were stained using a specific GluR2 antibody (2 mg / ml) and an immunostained field was photographed using conventional light microscopy (A). Afterwards, the CoS precipitate was enhanced with the silver staining and a picture was taken of the same field (B). Bar550 mm.

the motor neurons via Ca 21 -permeable AMPA receptors. The Co 21 -positive motor neurons were found to be more vulnerable to excitotoxicity than the others, as staining of the surviving motor neurons 24 h after a short KA exposure significantly diminished the relative amount of Co 21 positive cells (Fig. 4B). To correlate the sensitivity of motor neurons to the presence or absence of the GluR2 subunit, motor neurons were stained for GluR2. GluR2 immunoreactivity was present in all motor neurons, as we have shown before [16]. Clearly, some neurons stained more intensively than others (Fig. 5A). This variability was most prominent after 5 days in vitro. To evaluate whether those neurons that poorly stained for GluR2 were the ones that expressed Ca 21 -permeable receptors (i.e. showed KA-induced Co 21 uptake), double-staining was performed. As is demonstrated in Fig. 5B, the neurons that were clearly GluR2positive were also clearly Co 21 -positive. This indicates that motor neurons with Ca 21 -permeable AMPA receptors express a high amount of the GluR2 subunits. Our results do not support the idea that the absence or low abundance of the GluR2 subunit is responsible for the presence of Ca 21 -permeable AMPA receptors in motor neurons.

L. Van Den Bosch et al. / Journal of the Neurological Sciences 180 (2000) 29 – 34

4. Discussion The vulnerability of motor neurons to AMPA / KA receptor-mediated excitotoxicity may contribute to the selective death of motor neurons in ALS [1,24]. The selectivity of cell death induced by short term KA exposure for motor neurons as compared to dorsal horn neurons suggests that this mechanism may indeed contribute to the selective vulnerability of motor neurons in ALS. Several authors have suggested that this susceptibility is due to the presence of Ca 21 -permeable AMPA / KA receptors on motor neurons, as treatment with KA selectively and Ca 21 dependently affects neurons expressing these receptors [2,3,6]. Our data confirm and extend these studies. In motor neuron-enriched cocultures, exposure to KA during a 20 min period selectively affected those neurons expressing Ca 21 -permeable AMPA / KA receptors, assessed by the histochemical Co 21 staining. Part of the dorsal horn neurons also contain Ca 21 -permeable AMPA / KA receptors as also described by others [25], but their number was small when the KA-induced Co 21 staining was performed in identical conditions as for motor neurons. That these dorsal horn neurons readily contain AMPA / KA receptors is shown by the finding that they are sensitive to chronic exposure to KA [20]. The KA-induced motor neuron death completely depended upon extracellular Ca 21 . Several of our findings indicate that Ca 21 is entering the motor neurons via Ca 21 permeable AMPA receptors. Inhibitors of voltage-operated Ca 21 or Na 1 channels, or omitting Na 1 from the medium did not affect the KA-induced cell death. This suggests that Ca 21 entry through voltage-operated Ca 21 channels, or secondary to Na 1 entry (through reversal of the Na 1 / Ca 21 exchange function) is unlikely to contribute. Furthermore, LY300164, a specific AMPA receptor antagonist [22], completely antagonized this excitotoxic cell death. The AMPA receptor is a tetra- or pentameric complex composed of various combinations of four subunits, named GluR1 to 4. The presence of the edited GluR2 subunit has major effects on the channel’s characteristics, one of which is a drastic reduction of its permeability to Ca 21 [4]. JSTx is a subunit-specific antagonist, only blocking those AMPA receptors that lack the edited GluR2 subunit [23]. Therefore we can conclude from the observation that JSTx antagonized KA-induced cell death that Ca 21 entry via GluR2 deficient AMPA receptors is responsible for the selective motor neuron death. Several studies have investigated whether GluR2 is present in motor neurons but results are not consistent. In motor neurons of neonatal and adult rodents, GluR2 mRNA and immunoreactivity has been demonstrated [8– 11,14,15,17–19]. In human motor neurons, two studies found GluR2 mRNA to be present [12,13], while another did not [5]. Our data not only confirm that GluR2 immunoreactivity is present in motor neurons in enriched cultures, but in addition demonstrate that those neurons

33

that show strong GluR2 immunoreactivity, also stain for Co 21 after stimulation by KA. This finding indicates that these motor neurons, in spite of significantly expressing the GluR2 subunit, contain Ca 21 -permeable AMPA receptors. The presence of Ca 21 -permeable AMPA receptors in motor neurons is not due to the presence of unedited GluR2 subunits as it was shown before that GluR2 mRNA from these motor neurons is virtually completely edited at the Q / R site [19]. In view of the stochiometrics of the receptor complex, a small number of GluR2-deficient receptors is still likely to occur, even in the abundant presence of edited GluR2 [4]. If the number of AMPA receptors on the surface of the motor neuron is high, the small subset of GluR2-deficient receptors that is bound to occur, seems to be sufficient to induce cell death. It is tempting to speculate that the motor neuron selectively targets these Ca 21 -permeable AMPA receptors, as has been demonstrated for hippocampal interneurons [26] and pyramidal neurons [27]. If located at a strategic site of the cell, a minute amount of such receptors may give rise to KA-induced Ca 21 entry into a critical compartment of the cell, maybe sufficient to activate cell death mechanisms.

Acknowledgements We are grateful to Dr. J.D. Leander (Lilly Corporate Center, Indianapolis) for providing us with LY300164. This work was supported by a grant from the Research Council of the University of Leuven and from the Fund for Scientific Research — Flanders (FWO). L.V.D.B. is a Postdoctoral Fellow, W.V. a Research Assistant and W.R. a Clinical Investigator of the FWO.

References [1] Rothstein JD, Jin L, Dykes Hoberg M, Kuncl RW. Chronic inhibition of glutamate uptake produces a model of slow neurotoxicity. Proc Natl Acad Sci USA 1993;90:6591–5. [2] Carriedo SG, Ying HZ, Weiss JH. Motor neurons are selectively vulnerable to AMPA / Kainate receptor-mediated injury in vitro. J Neurosci 1996;16:4069–79. [3] Terro F, Yardin C, Esclaire F, Ayer-Lelievre C, Hugon J. Mild kainate toxicity produces selective motoneuron death with marked activation of Ca 21 -permeable AMPA / kainate receptors. Brain Res 1998;809:319–24. [4] Jonas P, Burnashev N. Molecular mechanisms controlling calcium entry through AMPA-type glutamate receptor channels. Neuron 1995;15:987–90. [5] Williams TL, Day NC, Ince PG, Kamboj RK, Shaw PJ. Calciumpermeable a-amino-3-hydroxy-5 methyl-4-isoxazole proprionic acid receptors: a molecular determinant of selective vulnerability in amyotrophic lateral sclerosis. Ann Neurol 1997;42:200–7. [6] Bar-Peled O, O’Brien RJ, Morrison JH, Rothstein JD. Cultured motor neurons possess calcium-permeable AMPA / kainate receptors. NeuroReport 1999;10:855–9. [7] Shaw PJ, Williams TL, Slade JY, Eggett CJ, Ince PG. Low

34

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17]

L. Van Den Bosch et al. / Journal of the Neurological Sciences 180 (2000) 29 – 34 expression of GluR2 AMPA receptor subunit protein by human motor neurons. Neuroreport 1999;10:261–5. ¨ TR, Berthele A, Zieglgansberger ¨ Tolle W, Seeburg PH, Wisden W. The differential expression of 16 NMDA and non-NMDA receptor subunits in the rat spinal cord and periaqueductal gray. J Neurosci 1993;13:5009–28. Sato K, Kiyama H, Tohyama M. The differential expression patterns of messenger RNAs encoding non-N-methyl-D-aspartate glutamate receptor subunits (GluR1-4) in the rat brain. Neuroscience 1993;52:515–39. Pellegrini-Giampietro DE, Fan S, Ault B, Miller BE, Zukin RS. Glutamate receptor gene expression in spinal cord of arthritic rats. J Neurosci 1994;14:1576–83. Jakowec MW, Yen L, Kalb RG. In situ hybridization analysis of AMPA receptor subunit gene expression in the developing rat spinal cord. Neuroscience 1995;67:909–20. Tomiyama M, Rodriguez-Puertas R, Cortes R, Christnacher A, Sommer B, Pazos R, Palacios JM, Mengod G. Differential regional distribution of AMPA receptor subunit messenger RNAs in the human spinal cord as visualized by in situ hybridization. Neuroscience 1996;75:901–15. Virgo L, Samarasinghe S, de Belleroche J. Analysis of AMPA receptor subunit mRNA xpression in control and ALS spinal cord. Neuroreport 1996;7:2507–11. Petralia RS, Wang YX, Mayat E, Wenthold RJ. Glutamate receptor subunit 2-selective antibody shows a differential distribution of calcium-impermeable AMPA receptors among populations of neurons. J Comp Neurol 1997;385:456–76. Temkin R, Lowe D, Jensen P, Hatt H, Smith DO. Expression of glutamate receptor subunits in alpha-motoneurons. Mol Brain Res 1997;52:38–45. Vandenberghe W, Van Den Bosch L, Robberecht W. Glial cells potentiate kainate-induced neuronal death in a motoneuron-enriched spinal coculture system. Brain Res 1998;807:1–10. Morrison BM, Janssen WGM, Gordon JW, Morrison H. Light and electron microscopic distribution of the AMPA receptor subunit,

[18]

[19]

[20]

[21]

[22]

[23]

[24] [25]

[26]

[27]

GluR2 in the spinal cord of control and G86R mutant superoxide dismutase transgenic mice. J Comp Neurol 1998;395:523–34. Van Den Bosch L, Verhoeven K, De Smedt H, Wuytack F, Missiaen L, Robberecht W. Calcium handling proteins in isolated spinal motorneurons. Life Sci 1999;65:1597–606. Vandenberghe W, Robberecht W, Brorson JR. AMPA receptor calcium permeability, GluR2 expression and selective motoneuron vulnerability. J Neurosci 2000;20:123–32. Van Den Bosch L, Robberecht W. Difference between short and long exposures of cultured motor neurons to excitotoxicity. Brain Res Bull 2000;in press. Pruss RM, Akeson RL, Racke MM, Wilburn JL. Agonist-activated cobalt uptake identifies divalent cation-permeable kainate receptors on neurons and glial cells. Neuron 1991;7:509–18. Bleakman D, Ballyk BA, Schoepp DD, Palmer AJ, Bath CP, Sharpe EF, Woolley ML, Bufton HR, Kamboj RK, Tarnawa I, Lodge D. Activity of 2,3-benzodiazepines at native rat and recombinant human glutamate receptors in vitro: stereospecificity and selectivity profiles. Neuropharmacology 1996;35:1689–702. Blaschke M, Keller BU, Rivosecchi R, Hollmann M, Heinemann S, Konnerth A. A single amino acid determines the subunit-specific spider toxin block of alpha-amino-3-hydroxy-5-methylisoxazole-4propionate / kainate receptor channels. Proc Natl Acad Sci USA 1993;90:6528–32. Rothstein JD. Excitotoxicity hypothesis. Neurology 1996;47:S19– 25. Kyrozis A, Goldstein PA, Heath MJ, MacDermott AB. Calcium entry through a subpopulation of AMPA receptors desensitized neighbouring NMDA receptors in rat dorsal horn neurons. J Physiol 1995;485:373–81. Toth K, McBain CJ. Afferent-specific innervation of two distinct AMPA receptor subtypes on single hippocampal interneurons. Nat Neurosci 1998;1:572–8. Yin HZ, Sensi SL, Carriedo SG, Weiss JH. Dendritic localization of Ca(21)-permeable AMPA / kainate channels in hippocampal pyramidal neurons. J Comp Neurol 1999;409:250–60.