GABAergic drugs become neurotoxic in cortical neurons pre-exposed to brain-derived neurotrophic factor

www.elsevier.com/locate/ymcne Mol. Cell. Neurosci. 37 (2008) 312 – 322

GABAergic drugs become neurotoxic in cortical neurons pre-exposed to brain-derived neurotrophic factor Gemma Molinaro,a Giuseppe Battaglia,a Barbara Riozzi,a Marianna Storto,a Sergio Fucile,a,b Fabrizio Eusebi,a,b Ferdinando Nicoletti,a,b and Valeria Brunoa,b,⁎ a

Istituto Neurologico Mediterraneo Neuromed, 86077 Pozzilli, Italy Department of Human Physiology and Pharmacology, University of Rome “La Sapienza”, 00185 Rome, Italy

b

Received 15 February 2007; revised 12 October 2007; accepted 16 October 2007 Available online 23 October 2007 A 24-h pretreatment with BNDF enhanced excitotoxic neuronal death in cultured mouse cortical cells challenged with NMDA in the presence of extracellular Mg2+. The GABAA receptor antagonist, bicuculline, enhanced NMDA toxicity in control cultures but, unexpectedly, became neuroprotective in cultures pretreated with BDNF. In contrast, drugs that activate GABAA receptors (e.g. muscimol, benzodiazepines, or phenobarbital) or drugs that indirectly enhance GABAergic transmission were protective in control cultures but amplified NMDA toxicity after pretreatment with BDNF. The atypical behaviour of GABAergic drugs in cultures pretreated with BDNF depended on changes in the anion reversal potential because (i) increases in extracellular Cl− concentrations abolished the neurotoxic action of muscimol; (ii) muscimol stimulated 36Cl− efflux after pretreatment with BDNF; and (iii) exposure to BDNF reduced the expression of the neuronal K+/Cl− co-transporter, KCC2. Our data raise the concern that GABAergic drugs may become neurotoxic under conditions associated with increases in brain BDNF levels. © 2007 Elsevier Inc. All rights reserved. Keywords: BDNF; Neuroprotection; GABAA receptors

Introduction Excitotoxicity contributes to neuronal death in a variety of acute and chronic neurodegenerative disorders, including stroke, brain trauma, status epilepticus, Parkinson's disease, Alzheimer's disease, and Huntington's disease (Choi, 1988). Neuronal death is largely mediated by an increased influx of extracellular calcium through the NMDA-gated ion channel, when the Mg2+ blockade of the channel is relieved by membrane depolarization (Mayer et al., 1984; Mayer and Westbrook, 1987). Hence, pharmacological activation of GABAA receptors, which form ligand-gated Cl− channels, is potentially one of the most powerful neuroprotective strategies ⁎ Corresponding author. Department of Human Physiology and Pharmacology, P.le Aldo Moro, 5, 00185 Roma, Italy. Fax: +39 0865 927575. E-mail address: [email protected] (V. Bruno). Available online on ScienceDirect (www.sciencedirect.com). 1044-7431/$ - see front matter © 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.mcn.2007.10.009

because the influx of extracellular Cl− produces membrane hyperpolarization. This prediction, however, is not translated into clinical practice where GABAergic drugs, such as benzodiazepines or barbiturates, have little value as neuroprotective agents in animal models (Westermaier et al., 2000; Kuhmonen et al., 2002; Zausinger et al., 2003). There are different explanations for the lack of success of these drugs. For example, neuroprotection may be achieved only by doses of GABAergic drugs that induce a general depression of the CNS (as occurs for barbiturates for the acute treatment of stroke), or activation of GABAA receptors in inhibitory neurons may produce disinhibition and enhance neuronal death. An interesting possibility is that the effect of GABAergic drugs is context-dependent, i.e. is influenced by environmental factors that regulate the intracellular concentrations of Cl− and other anions, such as HCO−3 . During early postnatal life, activation of GABAA receptors in hippocampal or cortical neurons induces membrane depolarization by promoting Cl− efflux instead of Cl− influx. Reduced expression of the KCC2 neuronal K+/Cl− cotransporter, which normally transports Cl− outward, increases intracellular Cl− concentrations and result in Cl− efflux via GABAA receptors (Thompson and Gahwiler, 1989; Kaila, 1994; Payne et al., 2003; DeFazio et al., 2000; Kakazu et al., 2000; Hubner et al., 2001). Changes in KCC2 are not restricted to brain development, but also occur in the adult brain under pathological conditions. In the hippocampal subiculum of patients with temporal lobe epilepsy, for example, activation of GABAA receptors becomes excitatory because of a reduced expression of KCC2 (Palma et al., 2006). Expression of KCC2 is down-regulated by brain-derived neurotrophic factor (BDNF), which is produced by astrocytes and microglia and activates TrkB receptors in neurons (Rivera et al., 2002, 2004). In dorsal horn neurons, BDNF released from activated microglia causes a depolarizing shift in the anion reversal potential, which inverts the polarity of GABA currents contributing to the pathophysiology of neuropathic pain (Coull et al., 2005). Production of BDNF is increased in response to a variety of psychotropic drugs, in response to dietary restriction, or under pathological conditions, such as neuroinflammatory disorders, epilepsy, and brain ischemia (Comelli et al., 1993; Simonato et al., 1998; Mizuta et al., 2001; Hashimoto et al., 2002; Duan et al., 2003; Kim et al., 2005;

G. Molinaro et al. / Mol. Cell. Neurosci. 37 (2008) 312–322

313

Fig. 1. A 24-h pretreatment with BDNF enhances NMDA toxicity in cultured cortical cells. In panel A, control (Ctrl) cultures and cultures pretreated with BDNF were challenged with increasing concentrations of NMDA. Neuronal degeneration was assessed by measuring both LDH release and trypan blue staining in the same cultures. After trypan blue staining, the number of dead neurons was counted from 3 random microscopic fields/well. Values are means ± S.E.M. of 6 determinations from 2 independent experiments. ⁎p b 0.05 (Student's t-test) vs. the respective control values. In panel B, BDNF or NGF were applied 24 h before the NMDA pulse, combined with NMDA, or applied just after the NMDA pulse. Toxicity produced by 60 μM NMDA was set as 100%. Values are means ± S.E. M. of 6–9 values from 2–3 independent experiments. ⁎p b 0.05 (One-way ANOVA ± Fisher's PLSD) vs. NMDA alone.

Ploughman et al., 2005; Pang et al., 2006). In addition, BDNF delivery to the brain is considered as a valuable strategy for the treatment of neurodegenerative disorders, such as Parkinson's disease or Huntington's disease (Yoshimoto et al., 1995; Larsson et al., 1999; Howells et al., 2000; Nagatsu et al., 2000; Siegel and Chauhan, 2000; Ohta et al., 2004; Liang et al., 2002; Borrell-Pages et al., 2006). It is important to know how GABAergic drugs affect the viability of neurons exposed to BDNF. We have addressed this issue in cultured neurons undergoing excitotoxic degeneration.

normally limit NMDA receptor activation by increasing membrane potential as a result of Cl− influx (Staley et al., 1995). Our cultures contain about 10–15% of GABAergic neurons, which release GABA in response to depolarizing agents or NMDA (Battaglia et al., 2001). A 10-min exposure to 60 μM NMDA increased GABA release by about 2 fold both in control cultures and in cultures pretreated with BDNF for 24 h, with no significant change between the two conditions (Fig. 2). The GABAA receptor antagonist, bicuculline, amplified NMDA toxicity in control cultures, but reduced the enhancement of neuronal

Results For the induction of excitotoxic neuronal death, we used mixed cultures of mouse cortical cells containing both neurons and astrocytes. Cultures at 14 days in vitro (DIV) were challenged with NMDA for 10 min under “physiological” conditions, i.e. in buffer containing 0.8 mM Mg2+ ions (physiological concentrations of extracellular Mg2+ are about 1 mM). NMDA induced neuronal death with an EC50 value of about 50–60 μM, and killed more than 80% of neurons at saturating concentrations (N 100 μM). The toxic action of NMDA was prevented by the selective NMDA channel blocker, MK-801 (1 μM; not shown). A 24-h pretreatment with 25 ng/ml of BDNF increased the potency without changing the efficacy of NMDA in causing neuronal death (Fig. 1A). BDNF was instead neuroprotective when either combined with NMDA or applied immediately after the NMDA pulse (Fig. 1B). As opposed to BDNF, nerve growth factor (NGF, 100 ng/ml) did not affect NMDA toxicity in any of these conditions (Fig. 1B). As the NMDA pulse was carried out in the presence of 0.8 mM Mg2+, we thought that pre-exposure to BDNF could affect endogenous mechanisms that regulate the Mg2+ blockade of NMDA receptors. We therefore focused on GABAA receptors, which

Fig. 2. NMDA stimulates GABA release to the same extent in control cultures and in cultures pretreated with BDNF for 24 h. Values are means± S.E.M. of 10–12 determinations. ⁎p b 0.05 (Student's t-test) vs. the respective basal values. Ctrl= control cultures.

314

G. Molinaro et al. / Mol. Cell. Neurosci. 37 (2008) 312–322

Fig. 3. Modulation of NMDA toxicity by bicuculline (A) and muscimol (A, B) in control cultures and in cultures pretreated with BDNF (25 ng/ml, 24 h). Bicuculline and muscimol were combined with NMDA during the excitotoxic pulse. Values are means ± S.E.M. of 12–18 determinations from 4–6 independent experiments. ⁎p b 0.05 (One-way ANOVA ± Fisher's PLSD) vs. the respective values obtained with NMDA alone.

death observed in cultures pre-exposed to BDNF (Fig. 3A). In contrast, the orthosteric GABAA receptor agonist, muscimol, was neuroprotective in control cultures, but further enhanced NMDA toxicity in cultures pretreated with BDNF (Fig. 3A). Muscimol produced both effects in the same range of concentrations (Fig. 3B). The dual action of muscimol was confirmed in additional studies in which a fixed concentration of the drug (10 μM) was combined with increasing concentrations of NMDA (Fig. 4). There was a tight temporal relationship between the potentiation of NMDA toxicity by BDNF and the further amplification of toxicity produced by muscimol in cultures pretreated with BDNF. When BDNF was applied to the cultures for only 3 h, it failed to affect NMDA toxicity, and, under these conditions, muscimol retained its classical neuroprotective activity. A pre-exposure to BDNF for 6, 12 and 24 h enhanced NMDA toxicity, and switched the action of muscimol from neuroprotective to neurotoxic (Fig. 5A). In addition, muscimol

was again neuroprotective when a 24-h pre-exposure to BDNF was followed by 24 h of washout, a condition in which BDNF pretreatment did not affect NMDA toxicity by itself (Fig. 5B). We extended the analysis to additional drugs that influence GABAergic transmission using cultures pretreated with BDNF for 24 h and then challenged with NMDA. The GABAA receptor positive allosteric modulators, diazepam, lorazepam, and phenobarbital, behaved like muscimol in further amplifying NMDA toxicity in cultures pretreated with BDNF; in control cultures, however, only phenobarbital was neuroprotective, whereas the two benzodiazepines failed to affect NMDA toxicity (Figs. 6A, B). We also used the GABA uptake inhibitor, SKF 89976A, which enhances extracellular GABA, and a number of ligands of mGlu1 metabotropic glutamate receptors. In cultured cortical cells, the mGlu1 receptor antagonists, LY367385 and CPCCOEt, are protective against NMDA toxicity by enhancing GABA release and indirectly

Fig. 4. Modulation of excitotoxicity by muscimol in control (Ctrl) cultures and in cultures pretreated with BDNF (25 ng/ml, 24 h) challenged with increasing concentrations of NMDA. Results are expressed as in Fig. 1A. Values are means ± S.E.M. of 6 determinations from 2 independent experiments. ⁎p b 0.05 (Student's t-test) vs. the respective values obtained in the absence of muscimol.

G. Molinaro et al. / Mol. Cell. Neurosci. 37 (2008) 312–322

315

Fig. 5. Time-dependent relationship between the ability of BDNF to enhance NMDA toxicity and the ability of muscimol to further amplify neuronal death. In panel A, cultures were pretreated with BDNF for 3, 6, 12 or 24 h, and then challenged with NMDA in the absence or presence of muscimol. In panel B, cultures were pretreated with BDNF for 24 h, and then additional 24 h were allowed before the NMDA pulse (with or without muscimol). Wo = washout. Data in panels A and B are means ± S.E.M. of 6–9 determination from 2–3 independent experiments. ⁎p b 0.05 (One-way ANOVA ± Fisher's PLSD) vs. the respective NMDA values in control cultures (⁎) or vs. the respective NMDA values in culture pretreated with BDNF (#).

activate GABAA receptors (Battaglia et al., 2001). SKF 89976A, LY367385, and CPCCOEt were protective in control cultures, as expected (Battaglia et al., 2001), but further enhanced NMDA toxicity in cultures pretreated with BDNF (Figs. 7A, B). In contrast, activation of mGlu1 receptors with the non-selective agonist, DHPG, enhanced NMDA toxicity in control cultures and reduced toxicity in cultures pretreated with BDNF (Fig. 7B). Taken together, these data indicate that BDNF causes a shift in the ability of GABAergic drugs to influence NMDA toxicity. Activation of GABAA receptors, which is notoriously neuroprotective, became neurotoxic in neurons pre-exposed to BDNF. We found no changes

in the expression of the α1, α2, and γ2 GABAA receptor subunits and in the expression of the NR1, NR2A and NR2B subunits of NMDA receptors in cultures pretreated with BDNF for 24 h (Figs. 8A, B). We therefore speculated that the switch in the GABAergic modulation of NMDA toxicity could involve mechanisms other than changes in the subunit composition of GABAA or NMDA receptor. In hippocampal and spinal neurons, exposure to BDNF changes the anion reversal potential in such a way that activation of GABAA receptors leads to the efflux of intracellular Cl− promoting membrane depolarization (Rivera et al., 2004; Coull et al., 2005). To examine whether a similar mechanism might have occurred in

Fig. 6. Modulation of NMDA toxicity by benzodiazepines (A) or phenobarbital (B) in control cultures and in cultures pretreated with BDNF (25 ng/ml, 24 h). Diazepam, lorazepam, and phenobarbital were combined with NMDA during the excitotoxic pulse. Values are means ± S.E.M. of 6 determinations from 2 independent experiments. ⁎p b 0.05 (One-way ANOVA ± Fisher's PLSD in panel A, and Student's t-test in panel B) vs. the respective value with NMDA alone.

316

G. Molinaro et al. / Mol. Cell. Neurosci. 37 (2008) 312–322

Fig. 7. Modulation of NMDA toxicity by the GABA uptake inhibitor, SKF 89976A (A), or by mGlu1 receptor ligands (B) in control cultures and in cultures pretreated with BDNF (25 ng/ml, 24 h). Values are means ± S.E.M. of 6–9 determinations from 2–3 independent experiments. ⁎p b 0.05 (One-way ANOVA ± Fisher's PLSD) vs. the respective values obtained with NMDA alone.

Fig. 8. A 24-h pretreatment with BDNF does not change the expression of α1, α2 and γ2 subunits of GABAA receptors (A) or NR1, NR2A and NR2B subunits of NMDA receptors (B) in cortical cultures. Representative immunoblots are shown in panels A and B. Densitometric data are shown in the graphs, where values are means ± S.E.M. of 4 determinations.

G. Molinaro et al. / Mol. Cell. Neurosci. 37 (2008) 312–322

317

regulating the expression of KCC2 by siRNA. We tested three KCC2 siRNAs for their ability to reduce the expression of KCC2. Cultures were incubated with siRNAs for 5 h, and then 12 h were allowed before the assessment of NMDA toxicity. One of the three siRNAs (named “A” in Fig. 12) partially lowered KCC2 expression, whereas the two additional siRNAs (“B” and “C”) were inactive (Fig. 12A). In cultures treated with siRNA “A”, muscimol was no longer protective against NMDA toxicity, although it did not enhance NMDA toxicity as occurred in cultures pretreated with BDNF (Fig. 12B). Discussion

Fig. 9. Modulation of NMDA toxicity by increasing concentrations of choline chloride in control cultures and in cultures pretreated with BDNF (25 ng/ml, 24 h). Choline chloride and muscimol were combined with NMDA during the pulse. Values are means ± S.E.M. of 9 determinations from 3 independent experiments. ⁎p b 0.05 (One-way ANOVA ± Fisher's PLSD) vs. the respective values obtained in the absence of choline chloride.

our cultures, we first assessed NMDA toxicity after increasing Cl− concentrations by 5, 10, or 20 mM during the NMDA pulse. Cl− ions were added as choline chloride. Increasing concentrations of Cl− were protective against NMDA toxicity in control cultures, and abolished the amplifying activity of muscimol on NMDA toxicity in cultures pretreated with BDNF (Fig. 9). In the absence of muscimol, only increases of Cl− concentrations of 20 mM reduced the enhancing activity of BDNF on NMDA toxicity (Fig. 9). We then measured the efflux of 36Cl− under steady state conditions and in response to muscimol. 36Cl− (0.3 μCi/well) was applied as Na36Cl for 30 min, and 36Cl− efflux was assessed both 5 and 15 s after having extensively washed out extracellular radioactivity. In control cultures, muscimol substantially reduced 36Cl− efflux at both times (Fig. 10), an effect that likely results from the inward movement of extracellular 36Cl− through the GABA-gated ion channel. In cultures pretreated with BDNF, basal 36Cl− efflux was lower at 5 s, and returned back to normal at 15 s. Addition of muscimol to these cultures did not reduce, but rather enhanced 36Cl− efflux at 5 s (Fig. 10). These data suggest that a 24-h exposure to BDNF changes the anion reversal potential in cortical neurons by increasing intracellular Cl− concentrations. To unravel the underlying mechanism(s), we measured the expression of the K+/Cl− co-transporter, KCC2, which transports Cl− outwords in neurons, and is regulated by BDNF in other neuronal populations. Immunocytochemistry showed that KCC2 was exclusively expressed in neurons, both in the cell soma and neurites (Fig. 11A), and was absent in astrocytes (identified by GFAP immunostaining in Figs. 11B, C). Immunoblot analysis of KCC2 showed a single band at the expected molecular size (140 kDa). Expression of KCC2 was significantly reduced in cultures pretreated with BDNF for 24 h (Figs. 11D, E). Taken together, these data suggest that activation of GABAA receptors amplified excitotoxicity in cultures pretreated with BDNF because of a reduced expression of KCC2, which leads to increases in intracellular Cl− concentrations and an outward Cl− current in response to receptor activation. To further support this hypothesis, we assessed the action of muscimol on NMDA toxicity after down-

BDNF is an “established” neuroprotective agent, and strategies aimed at increasing BDNF levels in the CNS are currently under development for the experimental treatment of neurodegenerative disorders (Yoshimoto et al., 1995; Larsson et al., 1999; Howells et al., 2000; Nagatsu et al., 2000; Siegel et al., 2000; Liang et al., 2002; Wu et al., 2004; Ohta et al., 2004; Husson et al., 2005; Nomura et al., 2005; Gonzalez et al., 2005; Borrell-Pages et al., 2006). BDNF is protective against excitotoxicity, and its induction mediates the phenomenon of “preconditioning”, which is the ability of subthreshold concentrations of NMDA to protect neurons against a successive excitotoxic insult (Soriano et al., 2006). There are exceptions, however, Koh et al. (1995) have shown for the first time that BDNF (50–100 ng/ml) and other neurotrophins (but not NGF) enhance necrotic death of cortical neurons when applied 48 h prior to a toxic pulse with NMDA or to oxygen–glucose deprivation (Koh et al., 1995). This unexpected toxic effect of BDNF is specific because it depends on the activation of TrkB receptors (Kim et al., 2003), but the underlying mechanism has not been clarified, as yet. Using mixed cultures of cortical cells we confirmed that a prolonged exposure to BDNF amplifies NMDA toxicity, showing that this paradoxical effect was mediated, at least

Fig. 10. Modulation of 36Cl− efflux by muscimol in control (Ctrl) cultures and in cultures pretreated with BDNF (25 ng/ml, 24 h). Values are means ± S. E.M. of 9 determinations from 3 independent experiments. p b 0.05 (Oneway ANOVA ± Fisher's PLSD) vs. the respective control values (⁎), or vs. the value obtained in cultures pretreated with BDNF without the addition of muscimol (pre-BDNF alone) (#).

318

G. Molinaro et al. / Mol. Cell. Neurosci. 37 (2008) 312–322

Fig. 11. KCC2 (A) and GFAP (B) immunostaining in mixed mouse cortical cultures. The lack of KCC2 in astrocytes is shown in panel C. Scale bar = 35 μm. A representative immunoblot of KCC2 in control cultures (Ctrl) and in cultures pretreated with BDNF (25 ng/ml, 24 h) is shown in panel D. Densitometric analysis is shown in panel E, where data are means ± S.E.M. of 9 determinations from 3 independent experiments. ⁎p b 0.05 (Student's t-test) vs. values obtained in control cultures.

Fig. 12. Activation of GABAA receptors with muscimol is no longer protective against NMDA toxicity in cultures with a reduced expression of KCC2 induced by siRNA. The effect of three different KCC2 siRNAs on KCC2 expression is shown in panel A. Densitometric values are means ± S.E.M. from 3–4 individual determinations. Toxicity data obtained in cultures treated with the effective siRNA (named “A”), an ineffective siRNA (named “B”) or with the transfection vehicle (lipofectamine) (Ctrl) are shown in panel B. Values are means ± S.E.M. from 6 determinations from 2 individual experiments. p b 0.05 (One-way ANOVA ± Fisher's PLSD) vs. the respective basal values (⁎) or vs. values obtained with NMDA alone (#).

G. Molinaro et al. / Mol. Cell. Neurosci. 37 (2008) 312–322

in part, by endogenous GABAergic transmission. Although a relatively low percentage of GABAergic neurons (about 10–15%) is present in our cultures (see Battaglia et al., 2001), these neurons are large, highly branched, and release GABA in response to NMDA (Battaglia et al., 2001; see also present data). The enhancement of NMDA toxicity depended on the activation of GABAA receptors because the selective GABAA receptor antagonist, bicuculline, which was otherwise neurotoxic, became protective in cultures pretreated with BDNF. In addition, drugs that activate GABAA receptors (muscimol, benzodiazepines, phenobarbital) or increase the availability of GABA at the synapses (SKF 89976A), enhanced NMDA toxicity in cultures pretreated with BDNF. Data obtained with mGlu1 receptor ligands were fully consistent with the “GABAergic hypothesis” of BDNF action. mGlu1 receptor antagonists, such as CPCCOEt and LY367385, are known to protect cortical and hippocampal neurons by enhancing GABAergic transmission (Battaglia et al., 2001; Cozzi et al., 2002), and instead amplified NMDA toxicity in cultures pretreated with BDNF. The opposite was found with the mGlu1/5 receptor agonist, DHPG (reviewed by Schoepp et al., 1999), which was neurotoxic in control cultures but reduced NMDA toxicity in cultures pretreated with BDNF. This finding supports the hypothesis that the role of mGlu1 receptors in neurodegeneration/neuroprotection is context-dependent (reviewed by Nicoletti et al., 1999). Activation of GABAA receptors normally causes membrane hyperpolarization because the low intracellular Cl− concentrations typical of mature neurons sets the Cl− reversal potential at a level lower than the resting potential, and this drives Cl− influx through the GABA-gated ion channel. Intracellular Cl− concentrations are kept low by the neuron-specific KCC2 co-transporter, the expression of which increases during development (Owens et al., 1996; Rivera et al., 2002, 2004; DeFazio et al., 2000; Kakazu et al., 2000; Hubner et al., 2001; Ganguly et al., 2001). Expression of KCC2 is constitutively low in immature neurons (Plotkin et al., 1997; Rivera et al., 1995; Lee et al., 2005), where activation of GABAA receptors causes membrane depolarization and synaptic excitation (Mueller et al., 1983; Ben-Ari et al., 1989; Ben-Ari, 2002; Luhmann and Prince, 1991). In mature neurons, KCC2 is down-regulated in response to increases in intracellular Ca2+ (Fiumelli et al., 2005) or under pathological conditions, e.g. in hippocampal neurons subjected to oxygen–glucose deprivation (Galeffi et al., 2004), and in hippocampal neurons from epileptic patients (Palma et al., 2006). BDNF is a key regulator of KCC2 expression. In developing neurons, BDNF increases KCC2 expression, thus contributing to the functional maturation of GABAergic synapses (Aguado et al., 2003). In contrast, BDNF down-regulates KCC2 in mature neurons through the activation of TrkB receptors (Rivera et al., 2002, 2004, 2005; Coull et al., 2005). Neurons in our mature cultures at 14 DIV expressed substantial amounts of KCC2, and the resulting low intracellular Cl− concentrations explain why GABAergic drugs were neuroprotective in control cultures. The reduced expression of KCC2 produced by a pretreatment with BDNF rendered GABAergic drugs neurotoxic. This was likely due to changes in Cl− reversal potential because (i) small increases in extracellular Cl− concentrations abolished the neurotoxic activity of muscimol, and (ii) muscimol enhanced 36Cl− efflux in cultures pretreated with BDNF. The resulting depolarization might have contributed to relieve the Mg2+ blockade of the NMDA channel (Mayer and Westbrook, 1987), thus explaining the neurotoxic action of muscimol, benzodiazepines, phenobarbital, and the other drugs that indirectly

319

enhance GABAergic transmission. This neurotoxic component mediated by GABAA receptors was so powerful to counterbalance the intrinsic neuroprotective activity of BDNF, which was regularly observed when BDNF was either combined with NMDA or applied immediately after the NMDA pulse. We also attempted to downregulate KCC2 in our cultures using a series of siRNAs. Only one siRNA partially reduced KCC2 expression, and muscimol lost its neuroprotective activity (although it was not amplified NMDA toxicity) in cultures treated with this particular siRNA. It is possible that a real switch from neuroprotection to neurotoxicity requires a down-regulation of KCC2 stronger than that produced by our siRNA. It could have been interesting to examine the action of GABAergic drugs on excitotoxic death in immature neurons, which express low levels of KCC2 (see above). However, we could not address this issue because NMDA did not induce neuronal death in immature mixed cortical cultures (6 DIV) (not shown). These findings may be relevant to human pathology because increases in BDNF levels are found in experimental animal models of ischemic brain damage (Comelli et al., 1993; Ploughman et al., 2005; Pera et al., 2005), in surviving neurons of the Alzheimer's disease brain (Siegel and Chauhan, 2000), and in the hippocampus of patients with mesial temporal lobe epilepsy associated with Ammon's horn sclerosis (Takahashi et al., 1999; Murray et al., 2000) where its expression correlates with neuronal loss and mossy fiber sprouting (Mathern et al., 1997). In addition, drugs that are currently used for the treatment of amyotrophic lateral sclerosis and Parkinson's disease, such as riluzole and cabergoline, enhance BDNF production in cultured astrocytes (Mizuta et al., 2001; Ohta et al., 2004), and the ubiquitine/proteasome inhibitors, cystamine and cysteamine, increase brain levels of BDNF in Huntington's disease (Borrell-Pages et al., 2006). Our data raise the concern that GABAergic drugs may facilitate neuronal death under all these conditions. A critical issue that remains to be solved is whether, and under which condition, endogenously produced BDNF is sufficient to down-regulate KCC2 and render GABAergic drugs potentially neurotoxic. Experimental methods Materials Bicuculline, 7-hydroxyiminocyclopropan[b]chromen-1a-carboxylic acid ethyl ester (CPCCOEt), (E)-2-methyl-6-stryrylpyridine(−)-2-oxa-4aminocyclo[3.1.0]hexane-4,6-dicarboxylic acid (LY367385), SKF 89976A, 3,5-dihydroxyphenylglycine (DHPG) were purchased from Tocris Cookson Ltd. (Bristol, UK). All other chemicals were purchased from Sigma (Milano, Italy).

Mixed cortical cultures Mixed cortical cell cultures containing both neurons and astrocytes were prepared from fetal mice at 14–16 days of gestation, as described previously (Rose et al., 1992). Briefly, dissociated cortical cells were plated in 15 mm multiwell vessels (Falcon Primaria, Lincoln Park, NJ, USA) on a layer of confluent astrocytes (7–14 days in vitro), using a plating medium of MEMEagle's salts (supplied glutamine-free) supplemented with 5% heat-inactivated horse serum, 5% fetal bovine serum, glutamine (2 mM), and glucose (final concentration 21 mM). Cultures were kept at 37 °C in a humidified 5% CO2 atmosphere. After 3–5 days in vitro, non-neuronal cells division was halted by 1–3 days exposure to 10 μM cytosine arabinoside, and cultures were shifted to a maintenance medium identical to plating medium but lacking fetal serum. Subsequent partial medium replacement was carried out twice a week. Only mature cultures (13–14 days in vitro) were used for the experiments.

320

G. Molinaro et al. / Mol. Cell. Neurosci. 37 (2008) 312–322

Exposure to excitatory amino acids Mixed cortical cell culture (cell density: 1.5 × 105/well) were exposed to NMDA (10, 30, 60, 100, 200 and 300 μM), at room temperature in a HEPESbuffered salt solution containing (in mM): NaCl, 120; KCl, 5.4; MgCl2, 0.8; CaCl2, 1.8; HEPES, 20; and glucose, 15. Afterwards, cultures were incubated at 37 °C for the following 24 h in MEM-Eagle's supplemented with 15.8 mM NaHCO3 and 25 mM glucose. Neurotrophic factors, such as BDNF (25 ng/ml), NGF (100 ng/ml) were either applied for 24 h prior to the NMDA pulse, combined with NMDA or applied just after the NMDA pulse. Timedependent relationship experiments of BDNF pretreatments were performed at 3, 6, 12 and 24 h prior to the NMDA pulse. After pretreatment with BDNF, cell cultures were treated with the GABAA receptor agonist muscimol (0.01– 100 μM), the GABAA receptor antagonist bicuculline (100 μM), or drugs that activate the GABAA receptor, such as diazepam (10 μM), lorazepam (10 μM), and phenobarbital (50 μM), or drugs that enhance GABA release, such as the mGlu1 receptor antagonists, CPCCOEt (30 μM) or LY367385 (10 μM), drugs that activate the mGlu1 receptor such as the non-selective agonist, DHPG (30 μM), or with the inhibitor of the high affinity GABA transporter, SKF 89976A (10 μM). Mixed cell cultures were also treated with choline chloride (5, 15, 20 mM) co-added with NMDA.

Blots were washed three times with TTBS buffer and then incubated for 1 h with secondary antibody (peroxidase-coupled anti-rabbit, GE Healthcare, Milano, Italy, 1:10,000). Immunostaining was revealed by enhanced ECL Western Blotting analysis system (GE Healthcare, Milano, Italy). Immunocytochemistry of the K+/Cl− synporter KCC2 in cortical cultures Mixed cortical cell cultures were washed with PBS and fixed for 30 min in 2% paraformaldehyde in PBS. After fixation, the cell membranes were permeabilized with Triton X-100 0.1% for 30 min at room temperature, followed by blocking with 0.1% BSA, 0.5% saponin and 2% goat serum for 1 h at room temperature. Cells were incubated with rabbit polyclonal antiKCC2 antibody (1:200, Upstate Biotechnology, Lake Placid, NY, USA), overnight at 4 °C. Cells were then incubated with goat anti-rabbit cy3conjugated secondary antibody (1:300, Chemicon, Temecula, CA, USA) for 2 h. For double staining, cells were incubated with anti-GFAP antibody (1:400, Sigma, St Louis, MO, USA) overnight at 4 °C, followed by incubation with horse antimouse FITC-conjugated secondary antibody (Vector Laboratories, Burlingame, CA, USA). Measurements of

Assessment of in vitro neuronal injury Neuronal injury was measured in all experiments by examination of cultures with phase-contrast microscopy at 20×, 18–20 h after the insult, when the process of cell death was largely complete. Neuronal damage was quantitatively assessed by counting dead neurons stained with trypan blue. Stained neurons were counted from three random fields per well. Neuronal damage was also assessed by measuring lactate dehydrogenase (LDH) released in the medium using a commercial available kit (Roche Diagnostic GmbH, Mannheim, Germany). Measurement of [3H]-GABA release in mixed cortical cultures Control mixed cortical cell cultures or cultures pretreated with BDNF, 25 ng/ml, for 24 h, were pre-incubated with [3H]-GABA (0.5 μCi/well, specific activity: 65 Ci/mmol, GE Healthcare, Milano, Italy) for 1 h in the presence of the GABA transaminase inhibitor, vigabatrin (100 μM). After extensive washing with HEPES-buffered salt solution, cells were incubated with NMDA (60 μM) for 10 min, in the presence of the GABA transporter inhibitor, SKF 89976A (10 μM), and vigabatrin. After the 10 min pulse with NMDA, the medium was collected and [3H]-GABA and measured by scintillation spectrometry. The radioactivity remaining in the cells was extracted with 500 μl of 0.1 N NaOH at room temperature overnight prior to assay. The [3H]-GABA release was quantified as % of the total amount of radioactivity. Western blot analysis in cortical cultures Cells were washed twice with PBS and lysed in RIPA buffer (1% NP40, 150 mM NaCl, 50 mM Tris–HCl, pH 7.5, 5 mM EDTA, 1 mM PMSF, 1 μg/ ml aprotinine, pepstatine, leupeptine) for the detection of the K+/Cl− synporter, KCC2. Proteins were separated on SDS polyacrylamide gel (8%) and transferred on Immuno PVDF membrane (Biorad, Milano, Italy) for 1 h. Filters were blocked overnight in TTBS (100 mM Tris–HCl, 0.9% NaCl, 1% Tween 20, pH 7.5) containing 5% not-fat dry milk. Blots were then incubated for 1 h at room temperature with a primary polyclonal antibody (0.5 μg/ml, Upstate Biotechnology, Lake Placid, NY, USA). For the detection of α1, α2 or γ2 subunits of GABAA receptors and NR1, NR2A or NR2B subunits of NMDA receptors, blots were incubated for 1 h at room temperature with primary anti-α1 (1:1000, Upstate), anti-α2 (1:1000, Chemicon, Temecula, CA, USA), anti-γ2, anti-NR1 (1:500, Upstate), antiNR2A (1:500, Upstate) or NR2B (1:500, Upstate) polyclonal antibodies.

Cl− efflux in mixed cortical cultures

36

Mixed cortical cell cultures were pretreated with BDNF (25 ng/ml, for 24 h). After extensive washing with HEPES-buffered salt solution, cells were loaded with Na36Cl− (300 nCi/well, specific activity: 0.125 mCi/ml, GE Healthcare, Milano, Italy) for 30 min at 37 °C. After extensive washing, muscimol (100 μM) was added and medium collected at 5 and 15 s and radioactivity was measured by scintillation spectrometry. siRNA studies Mixed cortical cell cultures were cultured for 12 days. On day 12, cells were incubated with lipofectamine and 90 pmol of 3 different siRNAs targeting the transcript of mouse KCC2 (Invitrogen, Milan, Italy), according to manufacturer instructions. Twelve h after transfection, cell cultures were used to confirm down-regulation of KCC2 expression by Western blot and for excitotoxicity experiments. The following siRNA duplexes were used: siRNA(A), GCCAUGCUCAUUGCCGGACUCAUUU (sense), siRNA(B), GGUCACCAAGAAUGUUUCCAUGUUU (sense), and siRNA(C) CCAUUCGGAGGAAGAAUCCAGCCAA (sense).

References Aguado, F., Carmona, M.A., Pozas, E., Aguilo, A., Martinez-Guijarro, F.J., Alcantara, S., Borrell, V., Yuste, R., Ibanez, C.F., Soriano, E., 2003. BDNF regulates spontaneous correlated activity at early developmental stages by increasing synaptogenesis and expression of the K+/Cl− cotransporter KCC2. Development 130, 1267–1280. Battaglia, G., Bruno, V., Pisani, A., Centonze, D., Catania, M.V., Calabresi, P., Nicoletti, F., 2001. Selective blockade of type-I metabotropic glutamate receptors induces neuroprotection by enhancing GABAergic transmission. Mol. Cell. Neurosci. 17, 1071–1083. Ben-Ari, Y., 2002. Excitatory actions of gaba during development: the nature of the nurture. Nat. Rev., Neurosci. 3, 728–739. Ben-Ari, Y., Cherubini, E., Corradetti, R., Gaiarsa, J.L., 1989. Giant synaptic potentials in immature rat CA3 hippocampal neurones. J. Physiol. 416, 303–325. Borrell-Pages, M., Canals, J.M., Cordelieres, F.P., Parker, J.A., Pineda, J.R., Grange, G., Bryson, E.A., Guillermier, M., Hirsch, E., Hantraye, P., Cheetham, M.E., Neri, C., Alberch, J., Brouillet, E., Saudou, F., Humbert, S., 2006. Cystamine and cysteamine increase brain levels of BDNF in Huntington disease via HSJ1b and transglutaminase. J. Clin. Invest. 116, 1410–1424. Choi, D.W., 1988. Glutamate neurotoxicity and diseases of the nervous system. Neuron 1, 623–634.

G. Molinaro et al. / Mol. Cell. Neurosci. 37 (2008) 312–322 Coull, J.A., Beggs, S., Boudreau, D., Boivin, D., Tsuda, M., Inoue, K., Gravel, C., Salter, M.W., De Koninck, Y., 2005. BDNF from microglia causes the shift in neuronal anion gradient underlying neuropathic pain. Nature 438, 1017–1021. Comelli, M.C., Guidolin, D., Seren, M.S., Zanoni, R., Canella, R., Rubini, R., Manev, H., 1993. Time course, localization and pharmacological modulation of immediate early inducible genes, brain-derived neurotrophic factor and trkB messenger RNAs in the rat brain following photochemical stroke. Neuroscience 55, 473–490. Cozzi, A., Meli, E., Carla, V., Pellicciari, R., Moroni, F., PellegriniGiampietro, D.E., 2002. Metabotropic glutamate 1 (mGlu1) receptor antagonists enhance GABAergic neurotransmission: a mechanism for the attenuation of post-ischemic injury and epileptiform activity? Neuropharmacology 43, 119–130. DeFazio, R.A., Keros, S., Quick, M.W., Hablitz, J.J., 2000. Potassiumcoupled chloride cotransport controls intracellular chloride in rat neocortical pyramidal neurons. J. Neurosci. 20, 8069–8076. Duan, W., Guo, Z., Jiang, H., Ware, M., Li, X.J., Mattson, M.P., 2003. Dietary restriction normalizes glucose metabolism and BDNF levels, slows disease progression, and increases survival in huntingtin mutant mice. Proc. Natl. Acad. Sci. U. S. A. 100, 2911–2916. Fiumelli, H., Cancedda, L., Poo, M.M., 2005. Modulation of GABAergic transmission by activity via postsynaptic Ca2+-dependent regulation of KCC2 function. Neuron 48, 773–786. Galeffi, F., Sah, R., Pond, B.B., George, A., Schwartz-Bloom, R.D., 2004. Changes in intracellular chloride after oxygen–glucose deprivation of the adult hippocampal slice: effect of diazepam. J. Neurosci. 24, 4478–4488. Ganguly, K., Schinder, A.F., Wong, S.T., Poo, M., 2001. GABA itself promotes the developmental switch of neuronal GABAergic responses from excitation to inhibition. Cell 105, 521–532. Gonzalez, S.L., Labombarda, F., Deniselle, M.C., Mougel, A., Guennoun, R., Schumacher, M., De Nicola, A.F., 2005. Progesterone neuroprotection in spinal cord trauma involves up-regulation of brain-derived neurotrophic factor in motoneurons. J. Steroid Biochem. Mol. Biol. 94, 143–149. Hashimoto, R., Takei, N., Shimazu, K., Christ, L., Lu, B., Chuang, D.M., 2002. Lithium induces brain-derived neurotrophic factor and activates TrkB in rodent cortical neurons: an essential step for neuroprotection against glutamate excitotoxicity. Neuropharmacology 43, 1173–1179. Howells, D.W., Porritt, M.J., Wong, J.Y., Batchelor, P.E., Kalnins, R., Hughes, A.J., Donnan, G.A., 2000. Reduced BDNF mRNA expression in the Parkinson's disease substantia nigra. Exp. Neurol. 166, 127–135. Hubner, C.A., Stein, V., Hermans-Borgmeyer, I., Meyer, T., Ballanyi, K., Jentsch, T.J., 2001. Disruption of KCC2 reveals an essential role of K– Cl cotransport already in early synaptic inhibition. Neuron 30, 515–524. Husson, I., Rangon, C.M., Lelievre, V., Bemelmans, A.P., Sachs, P., Mallet, J., Kosofsky, B.E., Gressens, P., 2005. BDNF-induced white matter neuroprotection and stage-dependent neuronal survival following a neonatal excitotoxic challenge. Cereb. Cortex 15, 250–261. Kaila, K., 1994. Ionic basis of GABAA receptor channel function in the nervous system. Prog. Neurobiol. 42, 489–537. Kakazu, Y., Uchida, S., Nakagawa, T., Akaike, N., Nabekura, J., 2000. Reversibility and cation selectivity of the K(+)–Cl(−) cotransport in rat central neurons. J. Neurophysiol. 84, 281–288. Kim, H.J., Hwang, J.J., Behrens, M.M., Snider, B.J., Choi, D.W., Koh, J.Y., 2003. TrkB mediates BDNF-induced potentiation of neuronal necrosis in cortical culture. Neurobiol. Dis. 14, 110–119. Kim, M.W., Bang, M.S., Han, T.R., Ko, Y.J., Yoon, B.W., Kim, J.H., Kang, L.M., Lee, K.M., Kim, M.H., 2005. Exercise increased BDNF and trkB in the contralateral hemisphere of the ischemic rat brain. Brain Res. 1052, 16–21. Koh, J.Y., Gwag, B.J., Lobner, D., Choi, D.W., 1995. Potentiated necrosis of cultured cortical neurons by neurotrophins. Science 268, 573–575. Kuhmonen, J., Lukkarinen, J., Grohn, O., Jolkkonen, J., Sivenius, J., 2002. Diazepam does not reduce infarct size in rats subjected to transient occlusion of the middle cerebral artery when normothermia is maintained. J. Pharm. Pharmacol. 54, 1565–1569.

321

Larsson, E., Nanobashvili, A., Kokaia, Z., Lindvall, O., 1999. Evidence for neuroprotective effects of endogenous brain-derived neurotrophic factor after global forebrain ischemia in rats. J. Cereb. Blood Flow Metab. 19, 1220–1228. Lee, H., Chen, C.X., Liu, Y.J., Aizenman, E., Kandler, K., 2005. KCC2 expression in immature rat cortical neurons is sufficient to switch the polarity of GABA responses. Eur. J. Neurosci. 21, 2593–2599. Liang, X.B., Liu, X.Y., Li, F.Q., Luo, Y., Lu, J., Zhang, W.M., Wang, X.M., Han, J.S., 2002. Long-term high-frequency electro-acupuncture stimulation prevents neuronal degeneration and up-regulates BDNF mRNA in the substantia nigra and ventral tegmental area following medial forebrain bundle axotomy. Brain Res. Mol. Brain Res. 108, 51–59. Luhmann, H.J., Prince, D.A., 1991. Postnatal maturation of the GABAergic system in rat neocortex. J. Neurophysiol. 65, 247–263. Mayer, M.L., Westbrook, G.L., 1987. Permeation and block of N-methyl-Daspartic acid receptor channels by divalent cations in mouse cultured central neurones. J. Physiol. 394, 501–527. Mayer, M.L., Westbrook, G.L., Guthrie, P.B., 1984. Voltage-dependent block by Mg2+ of NMDA responses in spinal cord neurones. Nature 309, 261–263. Mathern, G.W., Babb, T.L., Micevych, P.E., Blanco, C.E., Pretorius, J.K., 1997. Granule cell mRNA levels for BDNF, NGF, and NT-3 correlate with neuron losses or supragranular mossy fiber sprouting in the chronically damaged and epileptic human hippocampus. Mol. Chem. Neuropathol. 30, 53–76. Mizuta, I., Ohta, M., Ohta, K., Nishimura, M., Mizuta, E., Kuno, S., 2001. Riluzole stimulates nerve growth factor, brain-derived neurotrophic factor and glial cell line-derived neurotrophic factor synthesis in cultured mouse astrocytes. Neurosci. Lett. 310, 117–120. Mueller, A.L., Chesnut, R.M., Schwartzkroin, P.A., 1983. Actions of GABA in developing rabbit hippocampus: an in vitro study. Neurosci. Lett. 39, 193–198. Murray, K.D., Isackson, P.J., Eskin, T.A., King, M.A., Montesinos, S.P., Abraham, L.A., Roper, S.N., 2000. Altered mRNA expression for brainderived neurotrophic factor and type II calcium/calmodulin-dependent protein kinase in the hippocampus of patients with intractable temporal lobe epilepsy. J. Comp. Neurol. 418, 411–422. Nagatsu, T., Mogi, M., Ichinose, H., Togari, A., 2000. Changes in cytokines and neurotrophins in Parkinson's disease. J. Neural Transm., Suppl. 60, 277–290. Nicoletti, F., Bruno, V., Catania, M.V., Battaglia, G., Copani, A., Barbagallo, G., Ceña, V., Sanchez-Prieto, J., Spano, P.F., Pizzi, M., 1999. Group-I metabotropic glutamate receptors: hypotheses to explain their dual role in neurotoxicity and neuroprotection. Neuropharmacology 38, 1477–1484. Nomura, T., Honmou, O., Harada, K., Houkin, K., Hamada, H., Kocsis, J.D., 2005. I.V. infusion of brain-derived neurotrophic factor gene-modified human mesenchymal stem cells protects against injury in a cerebral ischemia model in adult rat. Neuroscience 136, 161–169. Ohta, K., Fujinami, A., Kuno, S., Sakakimoto, A., Matsui, H., Kawahara, Y., Ohta, M., 2004. Cabergoline stimulates synthesis and secretion of nerve growth factor, brain-derived neurotrophic factor and glial cell linederived neurotrophic factor by mouse astrocytes in primary culture. Pharmacology 71, 162–168. Owens, D.F., Boyce, L.H., Davis, M.B., Kriegstein, A.R., 1996. Excitatory GABA responses in embryonic and neonatal cortical slices demonstrated by gramicidin perforated-patch recordings and calcium imaging. J. Neurosci. 16, 6414–6423. Palma, E., Amici, M., Sobrero, F., Spinelli, G., Di Angelantonio, S., Ragozzino, D., Mascia, A., Scoppetta, C., Esposito, V., Miledi, R., Eusebi, F., 2006. Anomalous levels of Cl− transporters in the hippocampal subiculum from temporal lobe epilepsy patients make GABA excitatory. Proc. Natl. Acad. Sci. U. S. A. 103, 8465–8468. Pang, T.Y., Stam, N.C., Nithianantharajah, J., Howard, M.L., Hannan, A.J., 2006. Differential effects of voluntary physical exercise on behavioral and brain-derived neurotrophic factor expression deficits in Huntington's disease transgenic mice. Neuroscience 141, 569–584.

322

G. Molinaro et al. / Mol. Cell. Neurosci. 37 (2008) 312–322

Payne, J.A., Rivera, C., Voipio, J., Kaila, K., 2003. Cation-chloride cotransporters in neuronal communication, development and trauma. Trends Neurosci. 26, 199–206. Pera, J., Zawadzka, M., Kaminska, B., Szczudlik, A., 2005. Neurotrophic factor expression after focal brain ischemia preceded by different preconditioning strategies. Cerebrovasc. Dis. 19, 247–252. Plotkin, M.D., Snyder, E.Y., Hebert, S.C., Delpire, E., 1997. Expression of the Na–K–2Cl cotransporter is developmentally regulated in postnatal rat brains: a possible mechanism underlying GABA's excitatory role in immature brain. J. Neurobiol. 33, 781–795. Ploughman, M., Granter-Button, S., Chernenko, G., Tucker, B.A., Mearow, K.M., Corbett, D., 2005. Endurance exercise regimens induce differential effects on brain-derived neurotrophic factor, synapsin-I and insulin-like growth factor I after focal ischemia. Neuroscience 136, 991–1001. Rivera, C., Voipio, J., Payne, J.A., Ruusuvuori, E., Lahtinen, H., Lamsa, K., Pirvola, U., Saarma, M., Kaila, K., 1995. The K+/Cl− co-transporter KCC2 renders GABA hyperpolarizing during neuronal maturation. Nature 397, 251–255. Rivera, C., Li, H., Thomas-Crusells, J., Lahtinen, H., Viitanen, T., Nanobashvili, A., Kokaia, Z., Airaksinen, M.S., Voipio, J., Kaila, K., Saarma, M., 2002. BDNF-induced TrkB activation down-regulates the K+–Cl− cotransporter KCC2 and impairs neuronal Cl− extrusion. J. Cell Biol. 159, 747–752. Rivera, C., Voipio, J., Thomas-Crusells, J., Li, H., Emri, Z., Sipila, S., Payne, J.A., Minichiello, L., Saarma, M., Kaila, K., 2004. Mechanism of activity-dependent downregulation of the neuron-specific K–Cl cotransporter KCC2. J. Neurosci. 24, 4683–4691. Rivera, C., Voipio, J., Kaila, K., 2005. Two developmental switches in GABAergic signalling: the K+–Cl− cotransporter KCC2 and carbonic anhydrase CAVII. J. Physiol. 562, 27–36. Rose, K., Goldberg, M.P., Choi, D.W., 1992. Cytotoxicity in murine neocortical cell cultures. In: Tyson, C.A., Frazier, J.M. (Eds.), Methods in Toxicology, vol. 1. San Diego Academics, San Diego, pp. 46–60. Schoepp, D.D., Jane, D.E., Monn, J.A., 1999. Pharmacological agents acting at subtypes of metabotropic glutamate receptors. Neuropharmacology 38, 1431–1476. Siegel, G.J., Chauhan, N.B., 2000. Neurotrophic factors in Alzheimer's and Parkinson's disease brain. Brain Res. Brain Res. Rev. 33, 199–227.

Simonato, M., Molteni, R., Bregola, G., Muzzolini, A., Piffanelli, M., Beani, L., Racagni, G., Riva, M., 1998. Different patterns of induction of FGF-2, FGF-1 and BDNF mRNAs during kindling epileptogenesis in the rat. Eur. J. Neurosci. 10, 955–963. Soriano, F.X., Papadia, S., Hofmann, F., Hardingham, N.R., Bading, H., Hardingham, G.E., 2006. Preconditioning doses of NMDA promote neuroprotection by enhancing neuronal excitability. J. Neurosci. 26, 4509–4518. Staley, K.J., Soldo, B.L., Proctor, W.R., 1995. Ionic mechanisms of neuronal excitation by inhibitory GABAA receptors. Science 269, 977–981. Takahashi, M., Hayashi, S., Kakita, A., Wakabayashi, K., Fukuda, M., Kameyama, S., Tanaka, R., Takahashi, H., Nawa, H., 1999. Patients with temporal lobe epilepsy show an increase in brain-derived neurotrophic factor protein and its correlation with neuropeptide Y. Brain Res. 818, 579–582. Thompson, S.M., Gahwiler, B.H., 1989. Activity-dependent disinhibition. II. Effects of extracellular potassium, furosemide, and membrane potential on ECl− in hippocampal CA3 neurons. J. Neurophysiol. 61, 512–523. Yoshimoto, Y., Lin, Q., Collier, T.J., Frim, D.M., Breakefield, X.O., Bohn, M.C., 1995. Astrocytes retrovirally transduced with BDNF elicit behavioral improvement in a rat model of Parkinson's disease. Brain Res. 691, 25–36. Westermaier, T., Zausinger, S., Baethmann, A., Steiger, H.J., SchmidElsaesser, R., 2000. No additional neuroprotection provided by barbiturate-induced burst suppression under mild hypothermic conditions in rats subjected to reversible focal ischemia. J. Neurosurg. 93, 835–844. Wu, X., Zhu, D., Jiang, X., Okagaki, P., Mearow, K., Zhu, G., McCall, S., Banaudha, K., Lipsky, R.H., Marini, A.M., 2004. AMPA protects cultured neurons against glutamate excitotoxicity through a phosphatidylinositol 3-kinase-dependent activation in extracellular signal-regulated kinase to upregulate BDNF gene expression. J. Neurochem. 90, 807–818. Zausinger, S., Westermaier, T., Plesnila, N., Steiger, H.J., Schmid-Elsaesser, R., 2003. Neuroprotection in transient focal cerebral ischemia by combination drug therapy and mild hypothermia: comparison with customary therapeutic regimen. Stroke 34, 1526–1532.