Neuroscience Letters 482 (2010) 151–155
Contents lists available at ScienceDirect
Neuroscience Letters journal homepage: www.elsevier.com/locate/neulet
Glutamine-mediated protection from neuronal cell death depends on mitochondrial activity E.V. Stelmashook a , E.R. Lozier b , E.S. Goryacheva b , P. Mergenthaler c , S.V. Novikova a , D.B. Zorov b , N.K. Isaev a,b,∗ a b c
Institute of Neurology, Department of Brain Research, Russian Academy of Medical Sciences, 105064 Moscow, Russia A.N. Belozersky Institute of Physico-Chemical Biology, Moscow State University, 119992 Moscow, Russia Department of Experimental Neurology, Center for Stroke Research, Charite Universitetsmedizin Berlin, Germany
a r t i c l e
i n f o
Article history: Received 7 April 2010 Received in revised form 12 July 2010 Accepted 12 July 2010 Keywords: Glutamine Neurons Mitochondria
a b s t r a c t The specific aim of this study was to elucidate the role of mitochondria in a neuronal death caused by different metabolic effectors and possible role of intracellular calcium ions ([Ca2+ ]i ) and glutamine in mitochondria- and non-mitochondria-mediated cell death. Inhibition of mitochondrial complex I by rotenone was found to cause intensive death of cultured cerebellar granule neurons (CGNs) that was preceded by an increase in intracellular calcium concentration ([Ca2+ ]i ). The neuronal death induced by rotenone was significantly potentiated by glutamine. In addition, inhibition of Na/K-ATPase by ouabain also caused [Ca2+ ]i increase, but it induced neuronal cell death only in the absence of glucose. Treatment with glutamine prevented the toxic effect of ouabain and decreased [Ca2+ ]i . Blockade of ionotropic glutamate receptors prevented neuronal death and significantly decreased [Ca2+ ]i , demonstrating that toxicity of rotenone and ouabain was at least partially mediated by activation of these receptors. Activation of glutamate receptors by NMDA increased [Ca2+ ]i and decreased mitochondrial membrane potential leading to markedly decreased neuronal survival under glucose deprivation. Glutamine treatment under these conditions prevented cell death and significantly decreased the disturbances of [Ca2+ ]i and changes in mitochondrial membrane potential caused by NMDA during hypoglycemia. Our results indicate that glutamine stimulates glutamate-dependent neuronal damage when mitochondrial respiration is impaired. However, when mitochondria are functionally active, glutamine can be used by mitochondria as an alternative substrate to maintain cellular energy levels and promote cell survival. © 2010 Elsevier Ireland Ltd. All rights reserved.
Glutamatergic neurons in the central nervous system are highly dependent on glutamate synthesis from glutamine. However, under pathological conditions such as hypoxia and ischemia, the production of glutamate from glutamine can be increased significantly [10]. Subsequent overstimulation of glutamate receptors by the excitatory amino acid glutamate is an important pathogenic factor for neuronal cell death in cerebral ischemia [5,19,21]. Glutamine increases hypoxic damage of cultured cortical neurons [8], and induction of neuronal cell death by glutamine is also observed under other pathological conditions associated with glutamate toxicity. The toxic action of paraquat on CGNs is not revealed in cultivation medium not containing glutamine [29]. On the other hand, neurons can also use glutamine as an alternative energy substrate during hypoglycemia or after ischemia/reoxygenation, and this promotes neuronal survival [22,25,26]. Thus, glutamine shows Janus-faced properties by either promoting neuronal damage
∗ Corresponding author. Tel.: +7 495 939 59 44; fax: +7 495 939 31 81. E-mail address:
[email protected] (N.K. Isaev). 0304-3940/$ – see front matter © 2010 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.neulet.2010.07.022
through transformation into glutamate mediated by mitochondrial glutaminase or by functioning as an energy substrate for mitochondria to promote neuronal survival. In the present study we investigated whether utilization of glutamine by mitochondria is dependent on their functional activity. Primary neuronal cultures were prepared from cerebella of 6–7-day-old Wistar rats as described [30]. On the second day of cultivation (5% CO2 , 36.5 ± 0.5 ◦ C) cells were supplemented with fresh medium containing 25 mM KCl. Cultures were utilized for experiments after 3–4 days of cultivation. All experiments were carried out in accordance with the European Communities Council Directive of 24 November 1986 (86/609/EEC). For glucose deprivation (GD), neuronal cultures were washed twice in balanced salt solution (BSS) containing 154 mM NaCl, 25 mM KCl, 2.3 mM CaCl2 , 1 mM MgCl2 , 3.6 mM NaHCO3 , 0.35 mM Na2 HPO4 , and 10 mM HEPES at pH 7.3, and they were incubated in this medium. Control cultures were incubated in BSS supplemented with 1 g/L glucose (BSS1). Rotenone (2 M), ouabain (1 mM), or N-methyl-d-aspartate (NMDA, 50 M) and antagonists of ionotropic glutamate receptors 2-amino-5-phosphono-
152
E.V. Stelmashook et al. / Neuroscience Letters 482 (2010) 151–155
valerate (APV, 0.25 mM), (+)-5 methyl-10,11-dihydro-5Hdibenzo(a,d)cyclohepten 5-,10-imine maleate (MK-801, 10 M), and 2,3-dioxo-6-nitro-1,2,3,4-tetrahydrobenzoquinoxaline-7sulfonamide (NBQX, 10 M), and 2 mM glutamine were added directly to BSS [1]. For the analysis of neuronal viability, cultures were fixed with ethanol–formaldehyde–acetic acid (7:2:1) mixture and stained with trypan blue. The percentage of surviving neurons was estimated by counting the intact nuclei in the CGNs in five fields of view. The viability of untreated control cultures was taken as 100%, and the viability of treated cells was expressed as a percentage of the control. Changes in mitochondrial membrane potential were analyzed by loading cells with 0.2 M tetramethylrhodamine ethyl ester (TMRE, excitation 530 nm, emission 640 nm) for 15 min at 36.5 ± 0.5 ◦ C. For the analysis of [Ca]i , cells were loaded with 5 M Fluo-4 AM (excitation 485 nm, emission >530 nm) for 30 min at 36.5 ± 0.5 ◦ C followed by triple washing with BSS. Fluorescence intensities of Fluo-4 and TMRE were quantified using a Zenyth 3100 Multimode Detector (Anthos, Austria). One-way ANOVA with Newman–Keuls post hoc test was performed for statistical analysis. Levels of p < 0.05 were considered as statistically significant. The results are given as means ± SEM. All data were obtained using at least eight cultures from three independent experiments. Unless otherwise noted, all media and supplements used in cell culture were purchased from Biochrom KG Berlin, Germany. Fluo-4 AM and TMRE were from Molecular Probes (USA). All other reagents were from Sigma Chemicals (Germany). Neuronal cell death was induced by incubating CGNs in BSS1 supplemented with rotenone with or without glutamine for 2 h. Quantitative analysis of fixed cultures revealed a direct link of neuronal cell death and availability of glutamine. When treated with rotenone, neuronal survival without glutamine was 41 ± 4.8%, and it was 10 ± 1.3% in the presence of glutamine. Blockade of ionotropic glutamate receptors almost fully protected CGNs from the toxic effect of rotenone treatment either with or without glutamine (Fig. 1a). The absence of glucose in the culture medium potentiated rotenone toxicity, decreasing neuronal survival to 11 ± 0.8%, a level close to that observed with rotenone and glutamine. Glucose deprivation also significantly decreased the protective effect of the antagonists of ionotropic glutamate receptors in rotenone-induced cell death, while GD alone did not cause cell death under similar conditions (Fig. 1a). It is worth mentioning that lower efficacy of ionotropic glutamate receptors blockers under glucose deprivation has been demonstrated earlier [22]. Intracellular calcium concentration in CGNs increased substantially after 20-min exposure to rotenone (Fig. 1b). Glutamine did not have a significant effect on rotenone-induced increase in [Ca2+ ]i . Ionotropic glutamate receptor antagonists almost fully prevented the effect of rotenone on changes in [Ca2+ ]i in CGNs (Fig. 1b), and in the absence of glucose it potentiated the increase in [Ca2+ ]i . GD decreased the efficiency of ionotropic glutamate receptor antagonists to prevent the increase in [Ca2+ ]i , while GD alone did not affect [Ca2+ ]i (Fig. 1b). Neuronal death was induced by incubating CGNs in BSS1 supplemented with ouabain (Na/K-ATPase inhibitor) for 2 h. Ouabain treatment alone caused only slight toxicity in cultured CGNs. The survival of ouabain-treated neurons was 85 ± 3.7% versus 89 ± 3.6% with glutamine added. However, in the absence of glucose ouabain caused severe neuronal damage, decreasing cell survival to 34 ± 2.7%. Neurons were rescued from cell death by substitution of glutamine (survival: 80.5 ± 3.2%). Blockade of ionotropic glutamate receptors also almost fully protected CGNs from the toxic effect of ouabain (Fig. 2a).
Fig. 1. Effect of glutamine (Gln, 2 mM) and blockade of ionotropic glutamate receptors on (a) cell death of cerebellar granule neurons induced by rotenone (Rt, 2 M, 2 h) and on (b) intracellular calcium (20 min). *Statistically significant difference from control (normal glucose or without glucose), p < 0.05. # Statistically significant difference from Rt values (normal glucose or without glucose), p < 0.05, (a) n = 32–45, where n is the number of fields, (b) n = 9, where n is the number of cell cultures. Black bars, normal glucose level (BSS1); white bars, glucose deprivation (GD, BSS0).
Measurements of intracellular calcium concentration showed that treatment of CGNs with ouabain for 20 min induced a significant increase in [Ca2+ ]i (Fig. 2b). While treatment with glutamine did not change ouabain-induced [Ca2+ ]i increase under normal conditions, it dampened increase in [Ca2+ ]i under GD.
Fig. 2. Effect of glutamine (Gln, 2 mM) and blockade of ionotropic glutamate receptors on (a) cell death of cerebellar granule neurons induced by ouabain (Ou, 1 mM, 2 h) and (b) on intracellular calcium (20 min). *Statistically significant difference from control values (normal glucose or without glucose), p < 0.05. # Statistically significant difference from Ou values (normal glucose or without glucose), p < 0.05, (a) n = 45, where n is the number of fields, (b) n = 9–12, where n is the number of cell cultures. Black bars, normal glucose level (BSS1); white bars, glucose deprivation (GD, BSS0).
E.V. Stelmashook et al. / Neuroscience Letters 482 (2010) 151–155
The measured [Ca2+ ]i was returned to baseline by treatment with ionotropic glutamate receptor antagonists under all conditions. Cell death of CGNs was induced by the addition of BSS1 supplemented with 50 M NMDA for 2 h. Under normal and hypoglycemic conditions NMDA caused neuronal death which could be prevented by a competitive antagonist of Glu-receptors APV (Fig. 3a and b), although APV itself slightly lowered the survival of neurons under normal conditions. Under these, normal conditions, only moderate NMDA-induced cell damage was induced, resulting in 70 ± 3.7%
153
survival, and 82 ± 5.6% when BSS1 contained glutamine. Under GD, NMDA-induced cell death was potentiated and neuronal survival significantly decreased to 22 ± 2.8%. The addition of glutamine under GD increased neuronal survival to 58 ± 4.6% (Fig. 3a). After 20-min exposure to NMDA, [Ca2+ ]i increased, and the absence of glucose potentiated NMDA-induced increase in [Ca2+ ]i . During GD, glutamine caused a significant dampening of NMDAinduced increase in [Ca2+ ]i (Fig. 4a). While under normal conditions after NMDA and GD + NMDA exposures [Ca2+ ]i rises, with MK-801 and NBQX these values are significantly lower (Fig. 5a).
Fig. 3. Effect of glutamine (Gln, 2 mM) and APV 0,25 mM on cerebellar granule neurons death induced by NMDA (2 h): (a) cerebellar granule cells in 3-day cultures. Pyknotic nuclei are indicated by arrowheads. Cultures were fixed with ethanol–formaldehyde–acetic acid (7:2:1) mixture and stained with trypan blue. Scale bar 15 m. (b) quantitative evaluation of neuronal survival. *Statistically significant difference from control values (normal glucose or without glucose), p < 0.05. # Statistically significant difference from NMDA values (normal glucose or without glucose), p < 0.05, (a) n = 45, where n is the number of fields,.
154
E.V. Stelmashook et al. / Neuroscience Letters 482 (2010) 151–155
Fig. 4. Effect of glutamine (Gln, 2 mM) and NMDA 0,05 mM on (a) intracellular calcium (20 min) and (b) mitochondrial membrane potential (20 min). *Statistically significant difference from control values (normal glucose or without glucose), p < 0.05. # Statistically significant difference from NMDA values (normal glucose or without glucose), p < 0.05, (a) n = 8–9 and (b) n = 12, where n is the number of cell cultures. Black bars, normal glucose (BSS1); white bars, GD (BSS0).
Fig. 5. Effect of MK-801 (10 M) and NBQX (10 M) on 0.05 mM NMDA-induced (a) intracellular calcium (20 min) and (b) mitochondrial membrane potential (20 min). *Statistically significant difference from control values (normal glucose or without glucose), p < 0.05. # Statistically significant difference from NMDA values, p < 0.05, (a) n = 8–9 and (b) n = 9, where n is the number of cell cultures. Black bars, normal glucose (BSS1); white bars, GD (BSS0).
A 20-min exposure to NMDA gradually decreased mitochondrial membrane potential under normal conditions, and this was more profound under GD. When glutamine was supplemented in the incubation medium at the beginning of NMDA treatment loss of mitochondrial membrane potential was partially prevented (Fig. 4b). Incomplete protective effect of glutamine is apparently due to that glutamine is less efficient oxidative substrate than glucose/pyruvate [33], and on the other hand, overstimulation of Glu-receptors may cause calcium-dependent mitochondrial deenergization in CGNs even in the presence of substrate [12] which essentially determines NMDA-induced cell death [4]. Similarly, after NMDA and GD + NMDA exposures TMRE fluorescence dropped, with MK-801 and NBQX these values were significantly lower (Fig. 5b). Overstimulation of glutamate receptors by excitatory amino acids is one of the main causes of neuronal damage during neurodegenerative diseases. Increased extracellular glutamate under pathological conditions is not only due to an increased release, but also due to its synthesis from glutamine by glutaminase [8,23,24,29]. Permanent maintenance of optimal balance of glutamine and glutamate in nervous tissue is extremely important for the normal functioning of the brain. To a high extent this balance is supported by astrocytes removing glutamate excess. In these cells, cytosolic enzyme, glutamine synthase catalyzes the conversion of glutamate into glutamine. The activation of glutamine synthesis by glutamine synthase in a rat brain under ischemia and in postischemic period may be very important factor of normalization of glutamate level and considered as a defense against neurotoxic effect of excitatory amino acids [3,27]. This point of view has strong support by the data that inhibition of glutamine synthase potentiates neurotoxic effect of kainate and NMDA [16]. Glutamine serves not only as a glutamate precursor, but it can also be used as an energy substrate in the Krebs cycle [26]. However, in neurons the latter function is of minimal importance under normal conditions. Under conditions of substrate deprivation, for example after ischemia or during hypoglycemia, glucose metabolism in the brain is decreased, and this deficiency can be compensated by an increase in glutamine oxidation in neurons [25,33]. Utilization of glutamine in mitochondria as an energy substrate partially depends on oxidative phosphorylation, which ceases during anoxia/ischemia, but it can proceed during reoxygenation or GD [25,33]. In the present work we studied the influence of glutamine on the detrimental effect of rotenone and ouabain. We used dissociated immature cell cultures obtained from cerebella of 6–7-day-old rats. The population of glutamatergic granule neurons in these cultures, which are glutamate- and aspartate-responsive, is about 95% [7,18,20,28,32]. Under normal and pathological conditions CGNs synthesize glutamate from glutamine and release it more intensively than cerebral cortical neurons [10]. Thus, CGNs are a convenient model for studying the toxicity mediated by glutamate synthesis and release. The effect of glutamine on the death of neurochemically mature cultivated CGNs under conditions of the induction by a long-term glucose deprivation or by inhibition of the mitochondrial respiratory chain, which we have earlier explored in detail [31]. To explore this problem using immature neurons is vital due to that in spite of low tolerance of developing neurons to excitory amino acids still they are involved in perinatal brain injuries [14,6]. Rotenone is a specific inhibitor of the electron transport chain which blocks complex I, whereas the primary toxic effect of ouabain is not directly connected with inhibition of respiration, but it is mediated by the blockade of Na/K-ATPase and thereby causes an intracellular imbalance of univalent ions. However, our experiments indicate the involvement of glutamate toxicity for both drugs, as the effect of both rotenone and ouabain treatment could substantially be prevented by competitive antagonists of glutamate receptors. Therefore, we also studied the
E.V. Stelmashook et al. / Neuroscience Letters 482 (2010) 151–155
effect of glutamine supplementation on NMDA-induced neuronal damage. Rotenone treatment leads to disturbed mitochondrial function, which ultimately leads to decreased mitochondrial membrane potential. This decrease is substantially aggravated during glucose deficiency due to impaired generation of ATP by glycolysis as well as due to lack of oxidative phosphorylation (ischemic conditions). Abrupt decrease in ATP in neurons leads to accumulation of extracellular glutamate, resulting in overstimulation of glutamate receptors and increased intracellular calcium [13]. Under conditions of mitochondrial dysfunction, glutamine cannot be used by mitochondria as an alternative energy substrate, indicating that the major pathway of glutamine utilization in this context is mediated by phosphate-activated mitochondrial glutaminase, which in its active form is located on the outer surface of the inner mitochondrial membrane [17]. This metabolic pathway leads to the synthesis of glutamate and ammonium. However, under mitochondrial dysfunction and calcium overload, glutaminase activity can be increased as it can be activated by phosphate and calcium ions [2,9]. Contrary to treatment with rotenone, treatment with ouabain and NMDA resulted in minor damage to CGNs during normoglycemia. However, when these drugs were added during GD, substantial neuronal cell death was induced. Normal function of aerobic ATP generation warrants compensation of changes in ion balance in the cytosol by means of ATP-dependent transport systems with glutamine as the energy substrate. This was confirmed in our experiments where glutamine prevented decreased mitochondrial membrane potential and calcium overload caused by NMDA during hypoglycemia. It is worth mentioning that glutamine reduced only slightly the increase in intracellular calcium induced by ouabain + GD, as compared to the effect of MK-801 + NBQX. However, glutamine protected cells from cell death, as efficiently as the NMDA receptor antagonist, APV. Possibly, this effect was due to not glutamine rather than to a dual effect of ouabain on CGNs. From one side, as we demonstrated in this study ouabain is toxic for granule neurons under conditions with glucose deprivation, while from another in much lower concentrations it is able to protect the cell from apoptosis [11,15]. Finally, we demonstrated that under conditions of complete arrest of ATP generation in mitochondria, the presence of glutamine stimulated glutamate-dependent neuronal damage. However, when mitochondria are functionally active, glutamine can be used as a substrate by mitochondria in CGNs as an alternative to pyruvate for the maintenance of cellular energetics and promotion of neuronal survival. Acknowledgements This study was supported by RFBR # 08-04-00762-a, # 08-04-01667-a, # 09-04-01096-a and the Deutsche Forschungsgemeinschaft to PM. References [1] N. Andreeva, B. Khodorov, E. Stelmashook, E. Cragoe Jr., I. Victorov, Inhibition of Na+ /Ca2+ exchange enhances delayed neuronal death elicited by glutamate in cerebellar granule cell cultures, Brain Res. 548 (1991) 322–325. [2] A.M. Benjamin, Control of glutaminase activity in rat brain cortex in vitro: influence of glutamate, phosphate, ammonium, calcium and hydrogen ions, Brain Res. 208 (1981) 363–377. [3] D.N. Dao, M. Ahdab-Barmada, N.F. Schor, Cerebellar glutamine synthetase in children after hypoxia or ischemia, Stroke 22 (1991) 1312–1316. [4] F. Dessi, Y. Ben-Ari, C. Charriaut-Marlangue, Ruthenium red protects against glutamate-induced neuronal death in cerebellar culture, Neurosci. Lett. 201 (1995) 53–56.
155
[5] U. Dirnagl, C. Iadecola, M.A. Moskowitz, Pathobiology of ischaemic stroke: an integrated view, Trends Neurosci. 22 (1999) 391–397. [6] D. Fernбndez-Lypez, J. Marthnez-Orgado, I. Casanova, B. Bonet, J.C. Leza, P. Lorenzo, M.A. Moro, I. Lizasoain, Immature rat brain slices exposed to oxygen-glucose deprivation as an in vitro model of neonatal hypoxic-ischemic encephalopathy, J. Neurosci. Methods 145 (2005) 205–212. [7] V. Gallo, M.T. Ciotti, F. Aloisi, G. Levi, Selective release of glutamate from cerebellar granule cells differentiating in culture, Proc. Natl. Acad. Sci. U.S.A. 79 (1982) 7919–7923. [8] M.P. Goldberg, H. Monyer, D.W. Choi, Hypoxic neuronal injury in vitro depends on extracellular glutamine, Neurosci. Lett. 94 (1988) 52–57. [9] W.G. Haser, R.A. Shapiro, N.P. Curthoys, Comparison of the phosphatedependent glutaminase obtained from rat brain and kidney, Biochem. J. 229 (1985) 399–408. [10] R. Huang, L. Hertz, Effect of anoxia on glutamate formation from glutamine in cultured neurons: dependence on neuronal subtype, Brain Res. 660 (1994) 129–137. [11] N.K. Isaev, E.V. Stelmashook, A. Halle, C. Harms, M. Lautenschlager, M. Weih, U. Dirnagl, I.V. Victorov, D.B. Zorov, Inhibition of Na(+), K(+)-ATPase activity in cultured rat cerebellar granule cells prevents the onset of apoptosis induced by low potassium, Neurosci. Lett. 283 (2000) 41–44. [12] N.K. Isaev, D.B. Zorov, E.V. Stelmashook, R.E. Uzbekov, M.B. Kozhemyakin, I.V. Victorov, Neurotoxic glutamate treatment of cultured cerebellar granule cells induces Ca2+ -dependent collapse of mitochondrial membrane potential and ultrastructural alterations of mitochondria, FEBS Lett. 392 (1996) 143–147. [13] D. Jabaudon, M. Scanziani, B.H. Gдhwiler, U. Gerber, Acute decrease in net glutamate uptake during energy deprivation, Proc. Natl. Acad. Sci. U.S.A. 97 (2000) 5610–5615. [14] M.V. Johnston, Excitotoxicity in perinatal brain injury, Brain Pathol. 15 (2005) 234–240. [15] L.V. Karpova, E.R. Bulygina, A.A. Boldyrev, Different neuronal Na(+)/K(+)-ATPase isoforms are involved in diverse signaling pathways, Cell Biochem. Funct. 28 (2010) 135–141. [16] H. Kollegger, G.J. McBean, K.F. Tipton, The inhibition of glutamine synthetase in rat corpus striatum in vitro by methionine sulfoximine increases the neurotoxic effects of kainate and N-methyl-d-aspartate, Neurosci. Lett. 130 (1991) 95–98. [17] E. Kvamme, I.A. Torgner, B. Roberg, Kinetics and localization of brain phosphate activated glutaminase, J. Neurosci. Res. 66 (2001) 951–958. [18] G. Levi, V. Gallo, C. Giovannini, R. Suergin, Modulation of glutamate and GABA release by excitatory amino acid receptor agonists in cultured cerebellar cells, in: G. Hertting, H.C. Spatz (Eds.), Modulation of Synaptic Transmission and Plasticity, Springer-Verlag, Berlin, 1988, pp. 25–37. [19] M.P. Mattson, M. Gleichmann, A. Cheng, Mitochondria in neuroplasticity and neurological disorders, Neuron 60 (2008) 748–766. [20] P.P. McCaslin, W.W. Morgan, Cultured cerebellar cells as an in vitro model of excitatory amino acid receptor function, Brain Res. 417 (1987) 380–384. [21] P. Mergenthaler, U. Dirnagl, A. Meisel, Pathophysiology of stroke: lessons from animal models, Metab. Brain Dis. 19 (2004) 151–167. [22] H. Monyer, D.W. Choi, Glucose deprivation neuronal injury in vitro is modified by withdrawal of extracellular glutamine, J. Cereb. Blood Flow Metab. 10 (1990) 337–342. [23] R. Newcomb, A.R. Pierce, T. Kano, W. Meng, P. Bosque-Hamilton, L. Taylor, N. Curthoys, E.H. Lo, Characterization of mitochondrial glutaminase and amino acids at prolonged times after experimental focal cerebral ischemia, Brain Res. 813 (1998) 103–111. [24] R. Newcomb, X. Sun, L. Taylor, N. Curthoys, R.G. Giffard, Increased production of extracellular glutamate by the mitochondrial glutaminase following neuronal death, J. Biol. Chem. 272 (1997) 11276–11282. [25] J.M. Pascual, F. Carceller, J.M. Roda, S. Cerdn, Glutamate, glutamine, and GABA as substrates for the neuronal and glial compartments after focal cerebral ischemia in rats, Stroke 29 (1998) 1048–1056. [26] L. Peng, L. Gu, H. Zhang, X. Huang, E. Hertz, L. Hertz, Glutamine as an energy substrate in cultured neurons during glucose deprivation, J. Neurosci. Res. 85 (2007) 3480–3486. [27] C.K. Petito, M.C. Chung, L.M. Verkhovsky, A.J. Cooper, Brain glutamine synthetase increases following cerebral ischemia in the rat, Brain Res. 569 (1992) 275–280. [28] G.M. Shepherd, The Synaptic Organization of the Brain, Oxford Univ. Press, New York, Oxford, 1979. [29] E.V. Stelmashook, N.K. Isaev, D.B. Zorov, Paraquat potentiates glutamate toxicity in immature cultures of cerebellar granule neurons, Toxicol. Lett. 174 (2007) 82–88. [30] E.V. Stelmashook, N.K. Isaev, E.Y. Plotnikov, R.E. Uzbekov, I.B. Alieva, B. Arbeille, D.B. Zorov, Effect of transitory glucose deprivation on mitochondrial structure and functions in cultured cerebellar granule neurons, Neurosci. Lett. 461 (2009) 140–144. [31] E.V. Stelmashook, S.V. Novikova, N.K. Isaev, Influence of glutamine on the death of cultured cerebellar granule neurons induced by glucose deprivation and chemical hypoxia, Biochemistry (Mosc) 75 (2010) 1150–1156. [32] T.W. Stone, Glutamate as the neurotransmitter of cerebellar granule cells in the rat: electrophysiological evidence, Br. J. Pharmacol. 66 (1979) 291–296. [33] G.R. Sutherland, R.L. Tyson, R.N. Auer, Truncation of the Krebs cycle during hypoglycemic coma, Med. Chem. 4 (2008) 379–385.