Glutamate Neurotoxicity in Rat Cerebellar Granule Cells Involves Cytochrome c Release from Mitochondria and Mitochondrial Shuttle Impairment

Authors

  • Anna Atlante,

    1. Centro di Studio sui Mitochondri e Metabolismo Energetico, CNR, Bari, Italy*Dipartimento di Biochimica e Biologia Molecolare, Università di Bari, Bari, ItalyIstituto di Neurobiologia, CNR, Roma, ItalyDipartimento di Scienze Animali, Vegetali e dell’Ambiente, Università del Molise, Campobasso, Italy
    Search for more papers by this author
  • Sabatina Gagliardi,

    1. Centro di Studio sui Mitochondri e Metabolismo Energetico, CNR, Bari, Italy*Dipartimento di Biochimica e Biologia Molecolare, Università di Bari, Bari, ItalyIstituto di Neurobiologia, CNR, Roma, ItalyDipartimento di Scienze Animali, Vegetali e dell’Ambiente, Università del Molise, Campobasso, Italy
    Search for more papers by this author
  • Ersilia Marra,

    1. Centro di Studio sui Mitochondri e Metabolismo Energetico, CNR, Bari, Italy*Dipartimento di Biochimica e Biologia Molecolare, Università di Bari, Bari, ItalyIstituto di Neurobiologia, CNR, Roma, ItalyDipartimento di Scienze Animali, Vegetali e dell’Ambiente, Università del Molise, Campobasso, Italy
    Search for more papers by this author
  • Pietro Calissano,

    1. Centro di Studio sui Mitochondri e Metabolismo Energetico, CNR, Bari, Italy*Dipartimento di Biochimica e Biologia Molecolare, Università di Bari, Bari, ItalyIstituto di Neurobiologia, CNR, Roma, ItalyDipartimento di Scienze Animali, Vegetali e dell’Ambiente, Università del Molise, Campobasso, Italy
    Search for more papers by this author
  • Salvatore Passarella

    1. Centro di Studio sui Mitochondri e Metabolismo Energetico, CNR, Bari, Italy*Dipartimento di Biochimica e Biologia Molecolare, Università di Bari, Bari, ItalyIstituto di Neurobiologia, CNR, Roma, ItalyDipartimento di Scienze Animali, Vegetali e dell’Ambiente, Università del Molise, Campobasso, Italy
    Search for more papers by this author

  • Lippincott Williams & Wilkins, Inc., Philadelphia

  • Abbreviations used: ADK, adenylate kinase; BME, basal medium with Eagle’s salts; CGCs, cerebellar granule cells; cyt c, cytochrome c; DHAP, dihydroxyacetone phosphate; e.u., enzymatic units; GDH, glutamate dehydrogenase; Glu-CGCs, glutamate-treated CGCs; GNT, glutamate neurotoxicity; G3P, glycerol 3-phosphate; G3PDH, glycerol-3-phosphate dehydrogenase; MAL, malate; MDH, malate dehydrogenase; MK-801, (5R, 10S)-(+)-5-methyl-10,11-dihydro[a,d]cyclohepten-5,10-imine hydrogen maleate; OAA, oxaloacetate; PBS, phosphate-buffered saline; TMPD, N,N,N′,N′-tetramethyl-p-phenylenediamine.

Address correspondence and reprint requests to Dr. S. Passarella at Centro di Studio sui Mitocondri e Metabolismo Energetico, CNR, Via Amendola, 165/A, 70126 Bari, Italy.

Abstract

Abstract: To gain some insight into the mechanism by which glutamate neurotoxicity takes place in cerebellar granule cells, two steps of glucose oxidation were investigated: the electron flow via respiratory chain from certain substrates to oxygen and the transfer of extramitochondrial reducing equivalents via the mitochondrial shuttles. However, cytochrome c release from intact mitochondria was found to occur in glutamate-treated cells as detected photometrically in the supernatant of the cell homogenate suspension. As a result of cytochrome c release, an increase of the oxidation of externally added NADH was found, probably occurring via the NADH-b5 oxidoreductase of the outer mitochondrial membrane. When the two mitochondrial shuttles glycerol 3-phosphate/dihydroxyacetone phosphate and malate/oxaloacetate, devoted to oxidizing externally added NADH, were reconstructed, both were found to be impaired under glutamate neurotoxicity. Consistent early activation in two NADH oxidizing mechanisms, i.e., lactate production and plasma membrane NADH oxidoreductase activity, was found in glutamate-treated cells. In spite of this, the increase in the cell NADH fluorescence was found to be time-dependent, an index of the progressive damage of the cell.

The exposure of cultured cerebellar granule cells (CGCs) to glutamate has been found to cause massive neuronal degeneration and death as revealed by both in vivo and in vitro experiments (Choi et al., 1987; Choi, 1988, 1994; Coyle and Puttfarcken, 1993). Although some aspects of glutamate neuronal excitotoxicity, including the statement that the initial event of this process is the extensive entry of Ca2+ into the cell, have been exhaustively investigated, the details of how cell death takes place remain to be established. In particular, in a number of studies aimed at identifying the cell components involved in glutamate neurotoxicity (GNT), the pivotal role of mitochondria has been highlighted: it was shown that dying neurons rapidly lose their mitochondrial membrane potential and energy charge (Ankarcrona et al., 1995). Further, Ca2+-dependent uncoupling was found to contribute to the mitochondrial oxidative stress initiated by glutamate exposure (Dugan et al., 1995) with a key role in Ca2+ homeostasis proposed for mitochondria (Budd and Nicholls, 1996; Schinder et al., 1996). Finally, we have shown the early and progressive inhibition of both glucose and succinate oxidation, in intact CGCs and cell homogenate, respectively, under GNT (Atlante et al., 1996). In light of these reports, the identification of the mitochondrial target(s) and processes involved in neurotoxicity is a worthwhile goal to be pursued.

As a first step in the elucidation of the role of mitochondria in the cascade of events that lead neuronal cells to necrosis and death, we have considered two processes that play a major role in the glucose oxidation, namely, the electron flow in the respiratory chain and the transport of reducing equivalent from cytosol to mitochondria, as mediated by the mitochondrial shuttles. We show that cytochrome c (cyt c) release from the mitochondria to the extramitochondrial phase and the progressive inhibition of mitochondrial shuttles occur within minutes as a result of glutamate exposure of CGCs.

MATERIALS AND METHODS

Materials

Materials were those reported by Atlante et al. (1996).

Cell cultures

Primary cultures of cerebellar granule neurons were obtained as described by Levi et al. (1984). Before each experiment, the culture medium was removed and the plated CGCs were washed with phosphate-buffered saline (PBS), pH 7.4, containing 138 mM NaCl, 2.7 mM KCl, 8 mM Na2HPO4, and 15 mM KH2PO4, and then collected by gentle scraping in a final volume of 4 ml of PBS per 9-cm diameter dish. Suspended granule cells showed full viability, even though they lacked the morphological organization present in culture dishes, such as cell-cell and cell-substrate contacts, as well as neurites. Cell integrity, which remains rather constant for 3-5 h, was assessed quantitatively by checking the inability of cells to oxidize externally added succinate, which cannot enter intact cells (Berry et al., 1991), by checking the ability of ouabain to block glucose transport in cells (Atlante et al., 1996), and by counting dead cells, identified as large phase-bright cell bodies, as described by Volontè et al. (1994). The final cell suspension contained routinely 85-95% intact cells.

Cell homogenate from cell suspension was obtained by ∼20 strokes with a Dounce Potter homogenizer at room temperature. With this procedure, lactate dehydrogenase cellular content is released and subsequent treatment with Triton X-100 does not cause further release. The integrity of mitochondria in the homogenate was assessed by measuring the activities of both adenylate kinase (ADK; EC 2.7.4.3) and glutamate dehydrogenase (GDH; EC 1.4.1.3) (see below), which are marker enzymes of the mitochondrial intermembrane space and matrix, respectively. The percentage of damaged mitochondria was <1.5%. Moreover, mitochondrial coupling, which reflects the ability of mitochondria to produce ATP, was checked by measuring the respiratory control ratio (oxygen uptake rate after ADP addition/oxygen uptake rate before ADP addition, with succinate used as a respiratory substrate). In agreement with previous results (Atlante et al., 1996), the respiratory control ratio was 5 ± 0.8 in the control cells. As expected, both the inhibitors of electron flow and atractyloside, a powerful inhibitor of the ADP/ATP translocator (La Noue and Schoolwerth, 1984), blocked ADP-stimulated oxygen uptake. Inhibition was caused by oligomycin, which can inhibit ATP synthase, and rapidly reversed by carbonyl cyanide p-(trifluoromethoxy)phenylhydrazone (FCCP), which stimulates the rate of oxygen consumption (Nicholls, 1982).

The cell protein assay was determined according to Waddel and Hill (1956), with bovine serum albumin as a standard.

Cell toxicity studies

Glutamate exposure was performed 7 days after plating. Primary cultures were exposed, usually for 30 min or for the times indicated in the legends, to glutamate (100 μM) at 25°C in Locke’s solution (154 mM NaCl, 5.6 mM KCl, 3.6 mM NaHCO3, 2.3 mM CaCl2, 5.6 mM glucose, 10 mM HEPES, pH 7.4) in the presence of 1 μM glycine added to activate fully the NMDA-sensitive glutamate recognition sites (Johnson and Ascher, 1987). Cells were then replenished with basal medium with Eagle’s salts (BME) containing 25 mM KCl, 2 mM glutamine, and 100 μg/ml gentamicin and put in the incubator. For the quantitative assessment of GNT, cell integrity and count were measured, as described above, after 12-24 h. Neurotoxicity was expressed as percentage of intact cells with respect to control cells kept under the same experimental conditions without the glutamate addition. In control experiments, 95-97% cell integrity was found in 24 h.

Oxygen uptake studies

O2 consumption was measured polarographically at 37°C, as described (Atlante et al., 1996), by means of a Gilson 5/6 oxygraph, using a Clark electrode. The instrument sensitivity was just set to a value that allowed rates of O2 uptake as low as 0.5 nanoatoms/min/mg of protein to be followed.

Cyt c assay

Cyt c was assayed as described by Errede et al. (1978). Either cell homogenate or the resulting supernatant (obtained by centrifuging homogenate at 15,000 g for 10 min), in the presence of 3 μM rotenone plus 0.8 μM antimycin A plus 6 μM myxothiazole, was first incubated with potassium ferricyanide [K3Fe(CN)6; 0.1 mM] to oxidize the reduced cyt c. Then potassium cyanide (KCN; 1 mM) was added to inhibit cyt c oxidase. The reduction of cyt c (ε = 21 mM-1 cm-1) was then obtained by adding a few grains of sodium dithionite and measured as an increase of absorbance by using the Shimadzu double wavelength (548 minus 540 nm) spectrophotometer model UV3000.

ADK and GDH assay

ADK activity was measured photometrically at 37°C essentially as reported by Atlante et al. (1998) in the presence of 0.2 μg of oligomycin plus 1 μM atractyloside. GDH activity was assayed photometrically at 37°C essentially as described by Bernt and Bergmeyer (1963). Given that no significant activity was usually measured owing to the integrity of the mitochondria, a further check was made in Triton X-100-treated cell homogenate.

Reconstruction of glycerol 3-phosphate (G3P)/dihydroxyacetone phosphate (DHAP) and malate (MAL)/oxaloacetate (OAA) shuttles

G3P/DHAP and MAL/OAA shuttles were reconstructed at 37°C, essentially as described (Dawson and Cooney, 1978; Passarella et al., 1984), by using the Perkin-Elmer Lambda 5 spectrophotometer. In brief, 2 ml of the cell homogenate suspended in a medium, containing 0.2 M sucrose, 10 mM KCl, 20 mM HEPES-Tris, pH 7.2, and 1 mM MgCl2, was placed in two cuvettes: one was used as a reference, whereas in the other 0.2 mM NADH was added and the change in absorbance monitored in time. Then either L-glycerol-3-phosphate dehydrogenase (G3PDH; EC 1.1.1.8) [1 enzymatic unit (e.u.)] with 1 mM G3P (in the case of G3P/DHAP shuttle reconstruction) or malate dehydrogenase (MDH; EC 1.1.1.37) (2 e.u.) with 1 mM MAL (in the case of MAL/OAA shuttle reconstruction) was added to both the cuvettes. The rates of change of absorbance were obtained as the tangent at the initial part of the experimental curve and expressed as nanomoles of NADH oxidized per minute per milligram of cell protein. ε340 value at 37°C measured for NADH under our experimental conditions was found to be 6.5 mM-1 cm-1.

Measurement of lactate production

Cells were plated in 9-cm diameter dishes (15 × 106 cells/dish), and after 7 days in vitro glutamate treatment was performed as described above. At the end of glutamate incubation, Locke’s medium was removed and plated CGCs were replenished with fresh serum-free BME (containing 25 mM KCl, 2 mM glutamine, and 100 μg/ml gentamicin) and put in the incubator. During or after glutamate treatment, aliquots of BME were collected and extracellular lactate concentration, which is a reliable measure of lactate production inside the cells (Walz and Mukerji, 1988), was assayed according to Brandt (1982).

Plasma membrane NADH oxidoreductase assay

Plasma membrane NADH oxidoreductase assay was performed essentially as reported by Crane et al. (1982). In brief, cells were carefully washed in PBS and then suspended in the same medium. The reaction was started by adding 0.25 mM ferricyanide to the cell suspension (3 × 106 cells/ml). Enzyme activity was assayed by measuring the decrease of absorbance at 420 nm due to Fe3+ (CN)6 reduction by means of a Perkin-Elmer Lambda 5 spectrophotometer.

RESULTS

Respiratory chain components in GNT

We first investigated whether GNT affects the electron flow through the respiratory chain. This was achieved by measuring the oxygen uptake caused by the respiratory substrates (added without ADP) that are oxidized by complex I (β-hydroxybutyrate), complex II (succinate), and complex IV [(ascorbate plus) N,N,N′,N′-tetramethyl-p-phenylenediamine (TMPD)] (Fig. 1). In a typical experiment, a decrease in β-hydroxybutyrate (5 mM) oxidation was found due to glutamate exposure [v = 8 and 6 nanoatoms of O2/min/mg of cell protein in control and glutamate-treated CGCs (Glu-CGCs) at 30 min after glutamate treatment, respectively]. As expected, externally added rotenone was found to block completely the oxygen uptake in both control and Glu-CGCs. As a result of succinate (5 mM) addition, mitochondrial respiration was restored in both samples; however, as already reported (Atlante et al., 1996), Glu-CGCs were found to oxidize succinate at a lower rate than that of control cells (18 and 25 nanoatoms of O2/min/mg of cell protein, respectively). In both cases, oxygen uptake was fully inhibited by adding antimycin A. Following the addition of ascorbate (5 mM) plus TMPD (0.2 mM), which reduce cyt c directly and then oxygen, via cyt c oxidase, mitochondrial respiration was restored with an oxygen consumption rate lower in neurons treated with glutamate (38 and 69 nanoatoms of O2/min/mg of cell protein in Glu-CGCs and control, respectively). Cyanide, an inhibitor of cyt c oxidase, was found to block oxygen uptake. A statistical analysis, carried out in six experiments by comparing, as a percentage with respect to the control, the different rates of oxygen uptake values, clearly shows that the differences between control and Glu-CGCs are significant (see table in Fig. 1).

Figure 1.

Oxygen uptake in either control or Glu-CGC mitochondria. Homogenates from either control (trace a) or Glu-CGCs (trace b) (∼0.2 mg of protein) were incubated at 37°C in 1.5 ml of PBS. The following additions were made at the indicated concentrations: β-hydroxybutyrate (β-OH; 5 mM), rotenone (ROT; 3 μM), succinate (SUCC; 5 mM), antimycin A (AA; 0.8 μM), ascorbate (ASC; 5 mM) plus TMPD (0.2 mM), and cyanide (CN-; 1 mM). Numbers along the traces are rates of oxygen uptake expressed as nanoatoms of O2/min/mg of cell protein. In the table are reported the values of the oxygen uptake rate ± SE (VO2± s.e.) obtained from six experiments, carried out by using different cell preparations.

FIG. 1.

To ascertain whether the differences between control and Glu-CGCs depend on the progressive impairment of respiratory chain complexes, the activities of three respiratory complexes were monitored photometrically in a parallel experiment (data not shown). The addition of 0.5 mM NADH to solubilized CGCs can reduce the added cyt c, via complex I + III activities, with no apparent difference between control and Glu-CGCs in either the absence or presence of rotenone. No changes were found in the activity of duroquinol oxidase (cyt c-dependent) checked by measuring the increase in absorbance due to the oxidation of 30 ≥M duroquinol in the presence of 10 μM cyt c. It was also determined that this reaction was inhibited almost completely by antimycin A, thus accounting for duroquinol spontaneous oxidation. Finally, cyanide-sensitive oxidation of reduced cyt c was monitored to measure activity of complex IV, with no changes observed between control and Glu-CGCs. It should be noted that no change in mitochondrial complex activities was also found in another model system (Canevari et al., 1997).

The observed impairment of the oxygen uptake caused by ascorbate plus TMPD under neurotoxicity in the absence of variation in cyt c oxidase activity raises the question as to whether the other mitochondrial component involved in the oxidation of ascorbate plus TMPD, i.e., cyt c, might have been somehow affected in Glu-CGCs. In light of a number of studies that show cyt c release under apoptosis (Liu et al., 1996; Kluck et al., 1997a; Yang et al., 1997; Bossy-Wetzel et al., 1998; Skulachev, 1998), such a possibility was checked under GNT condition. Direct evidence of cyt c release from mitochondria was obtained by assaying cyt c both in the cell homogenates and in the supernatants. This was achieved photometrically by measuring the increase of absorbance due to the dithionite-dependent reduction of the previously oxidized cyt c (Fig. 2A). In a typical experiment, either cell homogenate or the relative supernatants were added with potassium ferricyanide (0.1 mM) and then cyanide (see MATERIALS AND METHODS). As a result of the addition of a few grains of sodium dithionite to both control and Glu-CGCs, a fast increase of absorbance was found, which indicates the cyt c presence in the sample. In five experiments, carried out with different cell culture preparations, cyt c was found to be 266 ± 6 and 310 ± 5 pmol/106 cells in Glu-CGCs at 30 min of exposure and control homogenate, respectively, in fairly good agreement with previous reports (Schwerzmann et al., 1986; Bourgeron et al., 1992; Petit et al., 1998), and 45 ± 3 and 4 ± 0.5 pmol/106 cells in the supernatants. In contrast, no significant absorbance change, i.e., cyt c release, was found in the control samples, in Glu-CGCs added with 1 μM (5R, 10S)-(+)-5-methyl-10,11-dihydro[a,d]cyclohepten-5,10-imine hydrogen maleate (MK-801), a selective NMDA receptor antagonist, or in 1 mM EGTA, a Ca2+ complexing agent. In the same experiment, the activities of the mitochondrial ADK and GDH were assayed in either control or Glu-CGCs. A difference in the ADK activity, measured in control and glutamate-treated cell homogenate (42 ± 1.3 and 39 ± 1.8 nmol of NADP+ reduced/min/mg of cell protein, respectively), was not found as judged by a statistical analysis (p > 0.1), nor was significant activity present in the respective supernatants. Negligible GDH activity was found both in cell homogenate and in the respective supernatant, whereas it proved to be essentially similar [450 ± 22 and 433 ± 17 nmol of NAD(P)H oxidized/min/mg of cell protein, p > 0.3] following treatment with Triton X-100 of control and Glu-CGCs.

Figure 2.

Oxygen uptake dependent on cyt c release from mitochondria. A: Detection of cyt c released from mitochondria under GNT. Supernatants, resulting from the homogenate of control (trace a) and Glu-CGCs (trace b) at 30 min after treatment, were incubated at 37°C in the presence of 3 μM rotenone, 0.8 μM antimycin A, and 6 μM myxothiazole and assayed for cyt c as reported in Materials and Methods. The following additions were made at the indicated concentrations: potassium ferricyanide [K3Fe(CN)6; 0.1 mM], potassium cyanide (KCN; 1 mM), and sodium dithionite (Na2S2O4; a few grains). B: Oxygen consumption caused by externally added ascorbate. Homogenates from either control (trace a) or Glu-CGCs (trace b) (∼0.2 mg of protein) were incubated at 37°C in 1.5 ml of PBS in the presence of 3 μM rotenone, 0.8 μM antimycin A, and 6 μM myxothiazole. The following additions were made at the indicated concentrations: ascorbate (ASC; 5 mM), TMPD (0.2 mM), and cyanide (CN-; 1 mM). Numbers along the traces are rates of oxygen uptake expressed as nanoatoms of O2/min/mg of cell protein. In five experiments, carried out by using different cell preparations, variations of up to 5% were found. C: Dependence of ascorbate-dependent oxygen uptake versus duration of glutamate treatment. The dependence of the ascorbate-dependent oxygen uptake rate versus duration of glutamate treatment (before STOP) or following glutamate exposure (after STOP) was studied for control ([UNK]) and Glu-CGCs (○), as reported in B. Results are means ± SE of triplicate measurements and representative of at least five different experiments carried out with different cell preparations from different groups of animals.

FIG. 2.

The question arises as to the functional consequence of the observed release. To gain a first insight into this point, we investigated the mechanism by which glutamate exposure affects ascorbate-dependent oxygen consumption in either the absence or presence of TMPD (Fig. 2B). It should be noted that ascorbate cannot permeate the outer mitochondrial membrane (Alexandre and Lehninger, 1984). Comparison made between control and Glu-CGCs with respect to their ability to oxidize ascorbate (5 mM) shows an approximately threefold increase in the rate of oxygen uptake (13 and 5 nanoatoms of O2/min/mg of cell protein at 30 min after glutamate exposure and in the control, respectively, p < 0.05 in five experiments). In both cases, the addition of TMPD (0.2 mM) strongly stimulated the rate of oxygen uptake in both cells even though oxygen was reduced at a lower rate in Glu-CGCs with respect to the control (37 and 72 nanoatoms of O2/min/mg of cell protein at 30 min after glutamate exposure and in the control, respectively). The capability of CGC homogenate to oxidize ascorbate at a rate higher than that of the control is consistent with cyt c release from mitochondria: in fact, cyt c available in the extramitochondrial phase, once reduced by ascorbate, can reduce oxygen via cyt c oxidase (La Piana et al., 1998).

The percentage increase of the rate of oxygen uptake caused by ascorbate addition to either Glu-CGCs or control as a function of the glutamate exposure time is reported in Fig. 2C. It is interesting that a few minutes (5-10) of glutamate exposure causes a remarkable increase of ascorbate-dependent oxygen uptake with respect to control cultures, the increase of oxygen uptake rate being ∼240% after a 5 h-treatment in Glu-CGCs. When both glutamate-treated and control cells were incubated in the presence of either 1 μM MK-801 or 1 mM EGTA, ascorbate failed to increase the oxygen uptake (data not shown), clearly showing that ascorbate-dependent oxygen uptake is a result of the GNT.

Evidence in favor of the release of fully functional cyt c was obtained in a set of experiments in which the rate of NADH oxidation due to the cyt c-dependent NADH-b5 oxidoreductase (Marzulli et al., 1995; La Piana et al., 1998) was measured photometrically. Cell homogenate, containing intact mitochondria, proved to oxidize NADH (0.2 mM) at a rate increasing with exposure time (Fig. 3). These data strongly suggest that the oxidation of either ascorbate or NADH is a result of the same phenomenon.

Figure 3.

NADH oxidation dependent on cyt c release from mitochondria. Homogenates from either control ([UNK]) or Glu-CGCs (○) (∼0.2 mg of protein) were incubated at 37°C in 1.5 ml of PBS in the presence of 3 μM rotenone, 0.8 μM antimycin A, and 6 μM myxothiazole. The dependence of the NADH (0.2 mM) oxidation versus duration of glutamate treatment (before STOP) or following glutamate exposure (after STOP) was studied as shown in the inset. Results are means ± SE of triplicate measurements and representative of at least six different experiments carried out with different cell preparations from different groups of animals.

FIG. 3.

Mitochondrial shuttles in GNT

The above-reported NADH oxidation under GNT raises the question as to whether the mitochondrial shuttles devoted to oxidized cytosolic NADH in glucose metabolism are affected in Glu-CGCs. Thus, a first comparison was made between Glu-CGCs and control in terms of their ability to transfer cytosolic reducing equivalents to the respiratory chain via the shuttle systems.

Reconstruction of the G3P/DHAP and the MAL/OAA shuttles was done by using the whole cell homogenate (Fig. 4), essentially as reported by Dawson and Cooney (1978) and Passarella et al. (1984), respectively. To achieve this, cell homogenate was added with NADH in the presence of either G3PDH or MDH. In agreement with Fig. 3, Glu-CGCs oxidize NADH. As a result of the addition of either 1 mM G3P (trace a) or 1 mM MAL (trace c), a decrease of absorbance was observed at rates equal to ∼5 and 8 nmol of NADH oxidized/min/mg of cell protein, respectively (Fig. 4A). The explanation for the observed data is that (a) added α-G3P is oxidized by mitochondria, in the reaction catalyzed by the mitochondrial α-G3PDH (m-α-G3PDH), to DHAP, which, in turn, oxidizes NADH outside mitochondria via the added c-α-G3PDH, and (b) MAL enters mitochondria likely via the dicarboxylate carrier in exchange for Pi (Passarella et al., 1984) and once inside the mitochondria it is oxidized to OAA via intramitochondrial mMDH. OAA efflux occurs in exchange with further malate, via the oxodicarboxylate carrier. Outside mitochondria, OAA oxidizes NADH via the externally added MDH (see scheme, Fig. 4B).

Figure 4.

Impairment of G3P/DHAP and MAL/OAA shuttles under GNT condition and cell viability. A: Reconstruction of G3P/DHAP and MAL/OAA shuttles. Homogenates from either control (traces a and c) or Glu-CGCs (traces b and d) (∼0.2 mg of protein) were incubated at 37°C in 1.5 ml of reaction medium, containing 0.2 M sucrose, 10 mM KCl, 20 mM HEPES-Tris, pH 7.2, and 1 mM MgCl2, in the presence of NADH (0.2 mM). The following additions were made at the indicated concentrations: G3PDH (1 e.u.) and G3P (1 mM), MDH (2 e.u.) and MAL (1 mM). Either DHAP or OAA appearance was measured as described in MATERIALS AND METHODS. Numbers along the traces are the rates of NADH oxidation measured as tangent at initial part of progress curve and are expressed as nmol of NADH oxidized/min/mg of cell protein. B: Scheme showing the oxidation of NADH via shuttle mechanisms. C: Dependence of either G3P/DHAP and MAL/OAA shuttle activity or cell viability versus duration of glutamate treatment. The dependence of either G3P/DHAP (▴) and MAL/OAA ([UNK]) shuttle activity or cell viability (□) versus duration of glutamate treatment (before STOP) or time following glutamate exposure (after STOP) was studied as reported in A and in Materials and Methods, respectively. Both shuttle activity and cell viability are expressed as a percentage of control (%C) to which 100% value was given. Control values were 100 ± 10. Results are means ± SE of triplicate measurements and representative of six different experiments carried out with different cell preparations from different groups of animals.

FIG. 4.

In both shuttles, a significant decrease of the rate of NADH oxidation was measured in Glu-CGCs [∼60 and 20% at 30 min after glutamate treatment for G3P/DHAP (trace b) and MAL/OAA (trace d) shuttle, respectively].

The dependence of the activity of either G3P/DHAP or MAL/OAA shuttle as a function of the glutamate treatment time was studied (Fig. 4C). It is interesting that stimulation of the G3P/DHAP shuttle activity was observed in the first 10 min, up to a 25% maximal increase; however, this was followed, after 15 min of exposure, by inhibition (50% at 25 min). In contrast, early inhibition was observed when the MAL/OAA shuttle activity was monitored. The latter was 50% inhibited after 90 min. In both cases, no shuttle-mediated NADH oxidation was measured after 5 h of exposure. In a parallel experiment, carried out with the same cell preparation, cell viability was tested (Fig. 4C) by counting healthy cells as described by Volontè et al. (1994). This was found to be ∼90% after 5 h of glutamate treatment, a time at which NADH oxidation, via the investigated shuttles, was almost completely blocked, and decreased up to 2-5% at 24 h.

Nonetheless, the inability of oxidizing cytosolic NADH was confirmed by following the cell pyridine nucleotide redox state during GNT. Cell emission fluorescence spectra of both control and glutamate-treated cells (Fig. 5) were carried out. Glutamate-treated cell fluorescence spectra showed an increase of the fluorescence emission close to the wavelength of 456 nm due to an increase in intracellular NADH concentration. On the contrary, no time-dependent change in the NADH fluorescence emission spectra was found in control cells. Such results are consistent with the mitochondrial shuttle impairment.

Figure 5.

Fluorescence emission spectra of control and Glu-CGCs. Either control (trace a) or Glu-CGCs (traces b-e) were incubated at 37°C in 1.5 ml of PBS with protein concentration adjusted to 30 μg/ml. The spectra of fluorescence emission (λexc = 334 nm; 410 nm < λ < 550 nm) were obtained, as reported in Materials and Methods, at the following times: 15 min of glutamate treatment (trace b), 1.5 h (trace c), 3 h (trace d), 5 h (trace e), i.e., 30 min of glutamate treatment plus time following glutamate exposure. No significant changes of control spectra were found at the investigated times. The experiment was repeated four times with different cell preparations, giving values with 5-10% variation. a.u., arbitrary units.

FIG. 5.

Alternative NADH oxidation mechanisms in GNT

In another series of experiments, two alternative pathways in NADH oxidation were examined, namely, anaerobic glycolysis and NADH oxidation via the plasma membrane NADH oxidoreductase. In the first case, lactate formation was studied as a function of glutamate exposure time (Fig. 6). Lactate overproduction was found in Glu-CGCs with respect to control occurring in a time-dependent manner.

Figure 6.

Increase of lactate production in Glu-CGCs. The dependence of lactate production versus duration of glutamate treatment (before STOP) or following glutamate exposure (after STOP) was studied for control ([UNK]) and Glu-CGCs (○). At the times indicated, BME aliquots of either control or Glu-CGCs were collected and the lactate released from cells was assayed, as reported in Materials and Methods. Results are means ± SE of six different experiments carried out with different cell preparations from different groups of animals.

FIG. 6.

In a parallel experiment, plasma membrane NADH oxidoreductase activity (Crane et al., 1985) was investigated by following ferricyanide reduction. In early glutamate exposure, a significant increase of NADH oxidoreductase activity up to 60% in the first 5 min of treatment was found, but for longer exposure times a decrease was found in the enzyme activity up to 80% inhibition after 3 h of treatment (Fig. 7). MK-801 and EDTA completely prevented both lactate overproduction and the NADH oxidoreductase activity changes (data not shown).

Figure 7.

Plasma membrane NADH oxidoreductase activity under GNT condition. Either control ([UNK]) or Glu-CGCs (○) (∼0.2 mg of protein), incubated at 37°C in 1.5 ml of PBS, were assayed for NADH oxidoreductase activity (see MATERIALS AND METHODS). The dependence of NADH oxidoreductase activity (expressed as nmol of Fe3+ (CN)6 reduced/min/mg of cell protein) versus duration of glutamate treatment (before STOP) or time following glutamate exposure (after STOP) was studied. Results are means ± SE of triplicate measurements and representative of six different experiments carried out with different cell preparations from different groups of animals. LDH, lactate dehydrogenase. A scheme of the NADH oxidation mechanism is shown below.

FIG. 7.

DISCUSSION

In this article, we give evidence of two processes occurring under GNT, namely, cyt c release and mitochondrial shuttle impairment. These results will be discussed separately. Cyt c release from the mitochondria is shown directly by the spectroscopic assay carried out in the cell homogenates and in the relative supernatants (Fig. 1). Moreover, the increase in both ascorbate and NADH oxidation, via NADH-b5 oxidoreductase, is consistent with this conclusion.

It should be noted that cyt c release has already been reported to occur under apoptosis (Liu et al., 1996; Kluck et al., 1997a; Yang et al., 1997; Bossy-Wetzel et al., 1998). In those articles, cyt c release was suggested to activate the caspase cascade (Kluck et al., 1997b; Zou et al., 1997; Skulachev, 1998), as well as determining an impairment of mitochondrial respiration (Krippner et al., 1996).

In our experiments, the observed capability of the respiratory chain to allow for electron flow at a rate not limiting the rates of either β-hydroxybutyrate or succinate oxidation, or the low cyt c release percentage with respect to the inhibition of the oxidation of ascorbate plus TMPD (30% with respect to 45% at 30 min after treatment, respectively), rules out the possibility that, at least in early glutamate exposure, cyt c release is involved in the respiration impairment. On the other hand, we give evidence consistent with the participation of the released cyt c in the oxidation of cytosolic NADH (Fig. 3) via NADH-b5 oxidoreductase (La Piana et al., 1998).

How cyt c release occurs remains to be established. As the leakage due to damage of the mitochondrial outer membrane is ruled out by the absence of either ADK or GDH in the extramitochondrial phase, cyt c might be released during permeability transition, as previously proposed (Kantrow and Piantadosi, 1997; Petit et al., 1998), even though the permeability transition pore, when incorporated into artificial membranes, has been shown not to mediate the cyt c transfer outside mitochondria (Zamzami et al., 1998).

We demonstrate that the activity of the tested respiratory complexes is unaffected during glutamate exposure (data not shown), in agreement with that found in cells undergoing apoptosis (see Krippner et al., 1996). Nonetheless, in both cases, changes in the electron flow parameters, as well as in the H+ translocating properties of the mitochondrial complexes, cannot be excluded.

Indeed taking into consideration that GNT is mostly dependent on radical oxygen species formation (Lafon-Cazal et al., 1993; Dugan et al., 1995; Reynolds and Hastings, 1995; Atlante et al., 1997), proteins located at the outer face of the mitochondrial membrane could be possible targets of GNT. In this regard, both the reported sensitivity of the mitochondrial carriers to singlet oxygen (Atlante et al., 1986, 1989, 1990) and the early and progressive inhibition of both glucose and succinate oxidation found in Glu-CGCs (Atlante et al., 1996) are consistent with the direct involvement of external proteins in the GNT. In addition to cyt c, one of them could be the mitochondrial G3PDH that is probably impaired under neurotoxicity, thus causing the impairment of the G3P/DHAP shuttle. This study shows that CGCs can oxidize NADH via the G3P/DHAP and MAL/OAA shuttles that have been reconstructed in vitro. In this regard, the increase in the G3P/DHAP shuttle observed early in the GNT might be a result of the Ca2+-dependent activation of the mitochondrial α-G3PDH (Swierczynski et al., 1976). How the decrease of the rate of NADH oxidation in vitro, i.e., of the glucose oxidation in vivo, occurs is yet to be established. However, by assuming that in brain cells the rate-limiting step of this shuttle system is the rate of MAL/OAA exchange across the mitochondrial inner membrane (Passarella et al., 1984), we suggest that the mitochondrial carriers could be considered as additional targets of the GNT.

Indeed, the impairment in the transfer of reducing equivalents from cytosol to mitochondria, via the mitochondrial shuttles, apparently is counterbalanced in at least four different ways: the increase in (a) glucose transport (Minervini et al., 1997), (b) lactate formation, (c) plasma membrane NADH oxidoreductase activity, and (d) cyt c release-dependent NADH oxidation.

How lactate release from CGCs occurs is at present unknown. However, given that lactate levels appear to increase with time, the possibility that lactate-dependent metabolic acidosis (Phillips et al., 1995) is responsible should be taken into consideration when discussing the mechanism of cell damage and death (Gottlieb et al., 1996, and references therein). In light of the experimental findings reported in this article, the oxidation of cytosolic NADH, via the plasma membrane NADH oxidoreductase, plays a relevant role under GNT. It is interesting that this enzyme has already been proposed to compensate for the loss of mitochondrial respiratory chain activity (Crane et al., 1985; Larm et al., 1994).

The picture emerging from this and our previous articles is the following: during early glutamate exposure, certain cell processes are impaired probably due to reactive oxygen species formation (see Kuroda et al., 1996; Atlante et al., 1997). In particular, a severe reduction of glucose oxidation takes place (Atlante et al., 1996) due, at present, to not well known processes, including the impairment of the mitochondrial shuttles. To counterbalance the deficit in glucose oxidation, different processes are evoked that contribute to the maintenance of adequate cell ATP (see Liu et al., 1996), including NADH oxidation via mitochondrial NADH-b5 oxidoreductase, made possible by cyt c release, as well as plasma membrane NADH oxidoreductase and lactate production. In spite of these protective mechanisms, cell damage is almost complete in 5 h of treatment, although cell survival differently monitored at that time seems still to be very high.

In conclusion this study sheds some light on the mechanism by which excitotoxicity takes place (for references, see Choi, 1988; Obrenovitch and Richards, 1995).

Acknowledgements

The authors thank the following: Dr. Gianluigi La Piana for his advice and stimulating discussion concerning cyt c release experiments; Dr. Lidia De Bari, who participated as a student at the first stages of this work, for helpful cooperation; Dr. Steve Reshkin for critical review of the manuscript; and Mr. Vito Giannoccaro for excellent technical assistance. This work was partially financed by PRIN “Bioenergetica e Trasporto di Membrana” (MURST) fund, by Fondi di Ricerca di Ateneo del Molise to S.P., and by MURST fund to P.C.

Ancillary