Oxoglutarate dehydrogenase complex controls glutamate-mediated neuronal death

Abstract

Brain injury is accompanied by neuroinflammation, accumulation of extracellular glutamate and mitochondrial dysfunction, all of which cause neuronal death. The aim of this study was to investigate the impact of these mechanisms on neuronal death. Patients from the neurosurgical intensive care unit suffering aneurysmal subarachnoid hemorrhage (SAH) were recruited retrospectively from a respective database. In vitro experiments were performed in rat cortex homogenate, primary dissociated neuronal cultures, B35 and NG108-15 cell lines. We employed methods including high resolution respirometry, electron spin resonance, fluorescent microscopy, kinetic determination of enzymatic activities and immunocytochemistry. We found that elevated levels of extracellular glutamate and nitric oxide (NO) metabolites correlated with poor clinical outcome in patients with SAH. In experiments using neuronal cultures we showed that the 2-oxoglutarate dehydrogenase complex (OGDHC), a key enzyme of the glutamate-dependent segment of the tricarboxylic acid (TCA) cycle, is more susceptible to the inhibition by NO than mitochondrial respiration. Inhibition of OGDHC by NO or by succinyl phosphonate (SP), a highly specific OGDHC inhibitor, caused accumulation of extracellular glutamate and neuronal death. Extracellular nitrite did not substantially contribute to this NO action. Reactivation of OGDHC by its cofactor thiamine (TH) reduced extracellular glutamate levels, Ca²⁺ influx into neurons and cell death rate. Salutary effect of TH against glutamate toxicity was confirmed in three different cell lines. Our data suggest that the loss of control over extracellular glutamate, as described here, rather than commonly assumed impaired energy metabolism, is the critical pathological manifestation of insufficient OGDHC activity, leading to neuronal death.

Graphical abstract

Fig. 1

Correlation between intracranial glutamate and clinical outcome in patients suffering severe aneurysmal SAH. (a) Schematic presentation of SAH; (b) percentage of SAH patients with manifestation of cerebral infarction and (c) poor functional outcome 3 months after the bleeding used in this study. Images d represent two patients suffering severe aneurysmal SAH, both Hunt&Hess grade IV, revealed distinct clinical course. In case patient 1 (d), cerebral microdialysis (MD) showed glutamate levels of 1.4 μmol/L following the period of early brain injury (day 3 after SAH). In this patient, no secondary ischemic events on computed tomography (CT) scans were observed and the patient showed good recovery with a modified Rankin Scale (mRS) of 1 at 3 months. In case patient 2 (d) cerebral glutamate was as high as 58.5 μmol/L on day 3. In this patient, multiple cerebral infarctions occurred within the following days, resulting in poor functional outcome (mRS 5) 3 months following SAH. (e) Correlation between cerebral glutamate levels on day 3 following SAH and either secondary ischemic events or (f) neurological outcome. *p < 0.05. Statistical significance was evaluated by t – test.

Fig. 2

Effects of NO on the TCA cycle enzymes and mitochondrial respiration in rat cortex homogenates. (a) Glutamate-dependent segment of the TCA cycle examined in these experiments; (b) and (c) effect of NO on GDH and OGDHC activities, respectively; (d) effect of nitrite on OGDHC activity; (e) and (e-inset) experimental procedure for examination of mitochondrial function; (f), (g) effect of NO on the mitochondrial oxidative phosphorylation (OXPHOS) via complex I (f) and complex II (g); (h), (i) - effect of peroxynitrite (ONOO⁻) on OXPHOS via complex I (h) and complex II (i). Enzymatic activities were determined by generation of NADH in the presence of corresponding substrates. OXPHOS was determined by oxygen uptake upon addition of ADP to mitochondria respiring with 10 mM glutamate (complex I substrate) or 10 mM succinate (complex II substrate). St2 – state 2 respiration of glutamate; St-3-CI and St-3-CII, state 3 respirations determined for complex I (OXPHOS CI) and complex II (OXPHOS CII), respectively. See also methods section for details. The data were analyzed by either one-way ANOVA with Dunnett's multiple comparisons test or two-tailed t-test (for two groups). Data are presented as mean ± SEM, n ≥ 3, * - p < 0.05; /**p < 0.01/*** - p < 0.001 vs. 0 μM).

Fig. 3

The impact of RBC on the OGDHC activity. (a) Pathways of NO generation in brain tissue in the presence of RBC. (b) Determination of NO bound to hemoglobin (Hb-NO) and released in cortex homogenate determined by formation of nitrosyl complexes of hemoglobin and iron ions (Fe–NO). The arrows indicate the peaks corresponding to Fe–NO and Hb-NO, respectively. Concentrations of (c) Hb-NO (d) and Fe–NO in the brain homogenate (BH), effects of hypoxia and the presence of RBC and nitrite. (e) The OGDHC activity upon hypoxia, effect of RBC and nitrite. (f) RBC added after hypoxia do not influence the activity of OGDHC. The data are presented as mean ± SEM, n = 3–5. The data were analyzed by Kruskal-Wallis followed by Dunn's test, *p < 0.05, **p < 0.01, ***p < 0.001. Schematic representation created with biorender.com.

Fig. 4

Influence of glutamate accumulation on OGDHC activity and consequent toxicity in NG108-15 cells. (a) Effect of increasing glutamate concentrations on the release of LDH. (b) Effect of glutamate, NO, SP and TH on LDH release, (c) non-viable cell counts, (d) OGDHC activity, (e) extracellular glutamate concentrations and (f) PDH activity. Cells were treated with 5 mM glutamate, 1 mM TH, 0.5 mM NO-donor DETA-NONOate and 0.2 mM SP over 48 h. Cell pellets were collected for determination of viability, OGDHC and PDH activity; culture supernatants were collected for detection of extracellular glutamate concentration and LDH release. The data are presented as mean ± SEM. The data were analyzed by either one-way ANOVA with Tukey's multiple comparisons test, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 or by paired two tailed t-test, #p < 0.05, ###p < 0.001.

Fig. 5

Effect of the OGDHC enzyme precursor TH on LDH released from neurons (B35 cell line) treated with NO or glutamate. (a) Effect of TH on the LDH release in control cells and cells treated with NO by 4 h after treatment with NO; (b) effect of TH on the LDH release in control cells and cells treated with NO by 24 h after treatment with NO; (c) effect of TH on the LDH release in control cells and cells treated with glutamate by 4 h after treatment with glutamate; (d) effect of TH on the LDH release in control cells and cells treated with glutamate by 24 h after treatment with glutamate. One-way ANOVA with Fisher's LSD test. The data are presented as mean ± SEM; n = at least 6, *p < 0.05, **p < 0.01, ***p < 0.001.

Fig. 6

Number of dopaminergic (THir positive) neurons and neurite outgrowth in murine mesencephalic cultures. (a) Typical morphology of intact neurons. (b) Glutamate (15 min treatment) substantially reduced the number of neurons. (c) This effect was attenuated by addition of memantine (60 μM, 75 min in general, added 60 min before glutamate), an NMDA receptor antagonist. (d) Addition of thiamine (1 mM, 75 min in general, added 60 min before glutamate) remarkably increased the survival rate of primary neuronal cells. (e) The length of neurites had a trend to become shorter in glutamate-treated THir neurons (0.5 mM). (f) This shortening was significant at 0.5 mM of glutamate. (g) In the presence of memantine the length of neurons was increased at 0.01 mM of glutamate; no other changes or trends were observed upon this treatment. (h) Addition of thiamine did not affect the length of neurons. The data were tested by ROUT test (Q = 5%) for outliers and analyzed by one-way ANOVA with Holm-Sidak multiple comparisons test. Data are presented as mean ± SEM; *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Fig. 7

The effect of thiamine (TH) on the glutamate induced changes in intracellular free Ca²⁺ concentration ([Ca²⁺]_i). (a) Changes in [Ca²⁺]_i in individual neurons (thin grey curves) and average traces (bold lines) in control culture (glutamate without TH) (n = 101 cells; error bars correspond to SEM). (b) X-scale expansion of [Ca²⁺]_i changes during the first 2 min of glutamate administration ([Ca²⁺]_i signals were monitored at 0.8 Hz frequency; the rest of experiment the recordings were carried out every 30 s). (c) Average changes of [Ca²⁺]_i (Mean ± SEM) of neurons in response to glutamate (glutamate, 10 μM) application in the control culture (blue line) and in the presence of TH (1 mM, red line, n = 93). (d) Areas under the curves (AUC) of [Ca²⁺]_i changes in individual neurons during the first 2 min of glutamate application. Glutamate (10 μM) was applied in Mg²⁺-free buffer containing 10 μM of glycine. The changes in [Ca²⁺]_i are presented as the ratio of the fluorescence intensities Ca²⁺ indicator Fura-FF excited at 340 and 380 nm (F340/F380). TH (1 mM) presented in the buffer during whole experiment until addition of protonophore FCCP (1 μM). Significance was analyzed by two-tailed t-test, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 8

The effect of thiamine (TH) on the glutamate (Glu) induced changes in mitochondrial membrane potential (ΔΨm) in the cultured rat cortical neurons. The ΔΨm changes are presented as the fluorescence intensities ratio (F/Fo) of potential-sensitive probe Rh123 (excitation 500/24, emission 578/105 nm) and normalized to its value at the beginning of the experiment (Fo). (a) Changes of mitochondrial potential (ΔΨm) in individual neurons (thin grey curves) and average traces (bold lines) in control culture (glutamate without TH) (n = 101 cells; error bars correspond to SEM). (b) X-scale expansion of ΔΨm changes during the first 2 min of glutamate administration (ΔΨm signals were monitored at 0.8 Hz frequency; the rest of experiment the recordings were carried out every 30 s). (c) Average changes of ΔΨm (Mean ± SEM) of neurons in response to glutamate (10 μM, glycine 10 μM, Mg²⁺-free) application in the control culture (blue line) and in the presence of TH (1 mM, red line, n = 93). (d) Areas under the curves of ΔΨm changes in individual neurons during the first 2 min of glutamate application (AUC, left axis) and amplitudes of F/Fo ratio at 60 s after glutamate addition (right axis). Thiamine (1 mM) presented in the buffer during whole experiment until addition of protonophore FCCP (1 μM). Shown are the results of one of 4 experiments performed with cell cultures prepared on different days (total n > 700). Significance was analyzed by two-tailed t-test, ***p < 0.001. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Scheme 1

Working scheme illustrating OGDHC-mediated pathway regulating neuronal death. Detailed scheme comparing glutamate metabolism in physiological and in injured brain. One pathological state that is activated after an injury is neuroinflammation. Therefore, the increased production of nitric oxide (NO) can target important enzymes of the TCA-cycle in mitochondria (OGDHC) and reduces its activity. In addition, other factors that could be presented in critical brain injury patients, like higher accumulation of nitrite and subarachnoid hemorrhage leads to the production of NO, which could slightly decrease OGDHC activity but not sufficient to completely inhibit the enzyme. The dysfunction of OGDHC indirectly increases the excretion of glutamate into the extracellular space to toxic levels, mediating neuronal cell death. The addition of thiamine, an intracellular precursor of the oxoglutarate dehydrogenase complex (OGDHC) coenzyme, promotes the OGDHC function. In the physiological state, the OGDHC function and the glutamate fluxes are adjusted in such a way, that even higher glutamate concentrations in the intracellular space are not toxic.