Ewa Matyja

Glutamate : A potential mediator of inorganic mercury neurotoxicity

Exposure to mercury vapor (Hgo) produces neurotoxic effects which are for the most part subsequent to its biotransformation in brain to the mercuric cation (Hg2+), which has an exceptionally strong affinity towards the SH groups in proteins. However, neurologic symptoms are often encountered in subjects in which Hg2+ concentration in the brain remains in the submicromolar range, markedly below the anticipated threshold for direct inhibition of cerebral metabolism and function. In this report we review biochemical and morphological evidence obtained in this and other laboratories in tissue culture studies suggesting that in such instances mercury neurotoxicity may be mediated by excitotoxic activity of glutamate (GLU). Mercuric chloride (MC) at 1 μM concentration (or less) inhibits GLU uptake and stimulates GLU release in cultured astrocytes, whichin vivo is likely to result in excessive GLU accumulation in the extracellular space of the CNS. Inhibition of GLU uptake and stimulation of GLU release by MC may be attenuated by addition to the cultures of a cell membrane-penetrating agent dithiothreitol (DTT) but not of glutathione (GSH), which is not transported to the inside of the cells. However, MC-stimulated release of GLU is suppressed when the intracellular GSH levels are increased by metabolic manipulation. The results indicate that the MC-vulnerable SH groups critical for GLU transport are located within the astrocytic membranes. Ultrastructural evidence for GLU-mediated MC neurotoxicity came from studies in an organotypic culture of rat cerebellum. We have shown that: 1) 1 μM MC lowers the threshold of GLU neurotoxicity, 2) the combined neurotoxic effect of GLU plus MC is attenuated by DTT but not by GSH, which is consistent with the involvement of impaired astrocytic GLU transport, and 3) neuronal damage induced by GLU plus MC becomes less accentuated in a medium with dizocilpine (MK-801), a noncompetitive NMDA receptor antagonist.

A comparative study of morphine stimulation and biphalin inhibition of human glioblastoma T98G cell proliferation in vitro

Biphalin is a new type of opioid peptide analog with high analgesic potency that is over 1000-fold greater than morphine. Because of its less addictive nature, biphalin has been suggested as a prospective new analgesic drug. Its high analgesic activity may be related to synergic interaction with all three types of opioid receptors (mu, delta, and kappa). Earlier data implicating involvement of opioid receptors, particularly MOR (mu opioid receptor) and KOR (kappa opioid receptor), in cell cycle regulation prompted us to investigate the effect of biphalin and morphine on human glioma T98G cell proliferation in vitro. We have documented an inhibitory effect of biphalin on tumor cell growth related to a decreased proliferation rate, decline of cell ability to form colonies, and modulation of the Ki-67 proliferation index. Morphine displayed the opposite effect and triggered stimulation of T98G cell proliferation. Our experiments have shown that biphalin might constitute an alternative solution for morphine application in anti-pain and anti-cancer therapy.

Transfection with liver-type glutaminase cDNA alters gene expression and reduces survival, migration and proliferation of T98G glioma cells.

Liver‐type glutaminase (LGA) is a glutaminase isoform that has been implicated in transcription modulation. LGA mRNA is absent from postoperative samples of primary gliomas and is low in cultured astrocytes. In this study, stable transfection of T98G cells with a vector carrying human LGA sequence increased the expression of LGA mRNA and protein, and the ability of the cells to degrade glutamine (Gln), as manifested by a three‐fold reduction of their steady‐state Gln content and a 2.5‐fold increase of their glutamate (Glu) content. The transfected cells (TLGA cells) showed a 40% decrease of cell survival as assessed by colony formation, well correlated with significant reduction of mitochondrial activity as demonstrated with MTT test. Also, a 45% reduction of cell migration and a 47% decrease of proliferation index (Ki67 immunostaining) were found as compared with sham‐transfected cells. Microarray analysis, which included over 47,000 transcripts, revealed a significantly altered expression of 85 genes in TLGA, but not in sham‐transfected or control cells (P < 0.005). Microarray data were confirmed with real‐time PCR analysis for eight genes potentially relevant to malignancy: S100A16, CAPN2, FNDC3B, DYNC1LI1, TIMP4, MGMT, ADM, and TIMP1. Of these changes, decreased expression of S100A16 and MGMT can be best reconciled with the current views on the role of their protein products in glioma malignancy. Malignancy‐reducing effect of newly inserted LGA mRNA in glioblastoma cells can be reconciled with a hypothesis that absence of such a modulatory mechanism in glia‐derived tumors deprived of LGA mRNA may facilitate some aspects of their progression.

Increased expression of a glutamine transporter SNAT3 is a marker of malignant gliomas.

Glutamine (Gln) is a growth determinant in neoplastic tissues. We analysed by RT-PCR the expression of mRNAs coding for the human variants of Gln transporters: ASCT2 (system ASC), SNAT1 [ATA1] (system A), SNAT3 [SN1] and SNAT5 [SN2] (system N), in samples of human malignant gliomas WHO grades III/IV (anaplastic astrocytoma and glioblastoma), glioma-derived cell cultures, brain metastases from peripheral organs, and control brain tissue. (Transporters designated according to the nomenclature recently proposed by Mckenzie and Ericksson [1]. Original designations in square brackets.) SNAT3 mRNA showed a 3–5 times stronger expression in gliomas than in metastases or control tissue, and was virtually absent from glioma cultures. Native glioblastoma immunostained positively with anti-SNAT3 antibody. The expression of ASCT2 mRNA, but not SNAT5 or SNAT1 mRNAs, was increased in all neoplastic tissues studied. Hence, increased expression of SNAT3 is a marker of primary malignant gliomas in situ.

Morphologic evidence of a primary response of glia to kainic acid administration into the rat neostriatum; Studied in vivo and in vitro

Glial changes that follow kainic acid administration were studied in the rat neostriatum at many different time intervals after the lesion, both in the animal model of Huntingtons chorea and in an organotypic culture of striatum. The glial reaction showed striking similarities between in vivo and in vitro conditions and resulted in extensive production and accumulation of gliofilaments leading to transformation of the protoplasmic type of astroglia into the fibrous type. The earliest ultrastructural study in vivo revealed severe swelling of the astrocytic cytoplasm and additional morphologic changes of cytoplasmic organelles, i.e., enlargement of mitochondria, dilation of rough endoplasmic reticulum, and presence of numerous vacuoles. The glial pathology progressed parallel to neuronal degeneration. The same reaction was observed in culture both in the explanted tissue in which neurons remained intact and in the distant outgrowth zone containing a pure population of glial cells. This study proved that kainic acid might act directly on astroglia cells and that glial changes were independent of neuronal damage. Because kainic acid is a structural analog of glutamate, the presented results may be interpreted to reflect changes in the metabolism of this amino acid occurring in astroglia independently of neuronal changes. This interpretation is consistent with the existence of two independent metabolic compartments of glutamate.

Excitotoxic neuronal injury in acute homocysteine neurotoxicity: Role of calcium and mitochondrial alterations

In this study we tested if calcium imbalance and mitochondrial dysfunction, which have been implicated in the conventional mechanisms of excitotoxicity induced by glutamate (Glu), are also involved in homocysteine (Hcy) neurotoxicity. Primary cultures of rat cerebellar granule cells were incubated for 30 min in the presence of 25 mM D,L-Hcy or 1mM Glu. At these concentrations both amino acids induced comparable neurodegeneration and chromatin condensation, evaluated after 24 h using the propidium iodide and Hoechst 33258 staining. These effects were partially prevented by cyclosporin A (CsA), but not FK506. Hcy-induced release of [(3)H]inositol phosphates and increase in intracellular calcium level (evaluated with fluo-3 fluorescent probe) were weakly expressed. Hcy- and Glu-induced mitochondrial swelling was visualized under electron microscope, and the release of Cytochrome c was evaluated using immunocytochemical method and confocal microscopy. Comparing to Glu, the effects of Hcy were slightly less expressed and less sensitive to CsA, while FK506 did not modify mitochondrial alterations. These data indicate that mitochondrial alterations play a similar role in acute Hcy and Glu neurotoxicity, although the mechanisms triggering Glu- and Hcy-evoked mitochondrial dysfunction seem to differ, Hcy toxicity being less dependent on calcium.

Homocysteine-Evoked 45Ca Release in the Rabbit Hippocampus Is Mediated by Both NMDA and Group I Metabotropic Glutamate Receptors: In Vivo Microdialysis Study

This in vivo microdialysis study compared the effects of NMDA and d,l-homocysteine (Hcy) administered via dialysis medium on 45Ca efflux from prelabeled rabbit hippocampus. Application of these agonists evoked dose-dependent, and sensitive to MK-801, opposite effects: NMDA decreased the 45Ca radioactivity in the dialysate, whereas Hcy induced the release of 45Ca. The latter effect was potentiated by glycine, inhibited by the antagonist of group I metabotropic glutamate receptors (mGluR) LY367385, and mimicked by t-ADA, an agonist of these receptors. Electron microscopic examination of pyramidal neurones in the CA1 sector of the hippocampus in the vicinity of the microdialysis probe after NMDA application demonstrated swelling of mitochondria, which was prevented by cyclosporin A. This study shows, for the first time, Hcy-induced activation of both group I mGluR and NMDA receptors, which may play a role in acute Hcy neurotoxicity. We present new applications of brain microdialysis in studies on excitotoxicity and neuroprotection.

Lack of expression of the liver-type glutaminase (LGA) mRNA in human malignant gliomas.

In the central nervous system (CNS), liver-type glutaminase (LGA) shows a unique nuclear localization suggesting its role in the regulation of transcription rather than in the cellular glutamine metabolism. RT-PCR analysis of RNA derived from postoperative tissue samples revealed the absence or only traces of LGA mRNA in all (9) cases of malignant gliomas (astrocytoma anaplasticum, AA, WHO grade III; glioblastoma multiforme, WHO grade IV) examined. The RNA was strongly expressed in the non-neoplastic tissue derived from the same patients (6 cases), and in most of the brain metastases from different organs (5 out of 7 cases). By contrast, the mRNAs coding for the kidney-type glutaminase (KGA) and its less ubiquitous isoform GAC, which catalyze degradation of the cytoplasmic pool of Gln, were expressed in all the tissues examined. The lack of LGA may be thus considered as a useful negative diagnostic marker of highly malignant gliomas in situ.

Relative expression of mRNAS coding for glutaminase isoforms in CNS tissues and CNS tumors.

Glutaminase (GA) in mammalian tissues occurs in three isoforms: LGA (liver-type), KGA (kidney-type) and GAC (a KGA variant). Our previous study showed that human malignant gliomas (WHO grades III and IV) lack expression of LGA mRNA but are enriched in GAC mRNA relative to KGA mRNA. Here we analyzed the expression of mRNAs coding for the three isoforms in the biopsy material derived from other central nervous system tumors of WHO grades I–III. Non-neoplastic resective epileptic surgery samples served as control, as did cultured rat astrocytes and neurons. The GAC mRNA/KGA mRNA expression ratio was as a rule higher in the neoplastic than in control tissues, irrespective of the cell type dominating in the tumor or tumor malignancy. LGA mRNA expression was relatively very low in cultured astrocytes, and very low to absent in astrocytoma pilocyticum, ependymoma and subependymal giant cell astrocytoma (SEGA), tumors of astrocytic origin. LGA mRNA expression was almost as high as that of KGA and GAC mRNA in cultured neurons and epileptic surgery samples which were enriched in neurons. LGA mRNA was also relatively high in ganglioglioma which contains a discernable proportion of neuronal cells, and in oligodendroglioma. The results show that low expression of LGA mRNA is a feature common to normal astrocytes and astroglia-derived tumor cells or ependymomas and can be considered as a cell-type, rather than a malignancy marker.

Light and electron microscopy characteristics of the muscle of patients with SURF1 gene mutations associated with Leigh disease

Aims: Leigh syndrome (LS) is characterised by almost identical brain changes despite considerable causal heterogeneity. SURF1 gene mutations are among the most frequent causes of LS. Although deficiency of cytochrome c oxidase (COX) is a typical feature of the muscle in SURF1-deficient LS, other abnormalities have been rarely described. The aim of the present work is to assess the skeletal muscle morphology coexisting with SURF1 mutations from our own research and in the literature. Methods: Muscle samples from 21 patients who fulfilled the criteria of LS and SURF1 mutations (14 homozygotes and 7 heterozygotes of c.841delCT) were examined by light and electron microscopy. Results: Diffuse decreased activity or total deficit of COX was revealed histochemically in all examined muscles. No ragged red fibres (RRFs) were seen. Lipid accumulation and fibre size variability were found in 14 and 9 specimens, respectively. Ultrastructural assessment showed several mitochondrial abnormalities, lipid deposits, myofibrillar disorganisation and other minor changes. In five cases no ultrastructural changes were found. Apart from slight correlation between lipid accumulation shown by histochemical and ultrastructural techniques, no other correlations were revealed between parameters investigated, especially between severity of morphological changes and the patient’s age at the biopsy. Conclusion: Histological and histochemical features of muscle of genetically homogenous SURF1-deficient LS were reproducible in detection of COX deficit. Minor muscle changes were not commonly present. Also, ultrastructural abnormalities were not a consistent feature. It should be emphasised that SURF1-deficient muscle assessed in the light and electron microscopy panel may be interpreted as normal if COX staining is not employed.