Ursula Sonnewald

Cortical and cerebellar astrocytes were cultured in medium containing pentylenetetrazole (PTZ), a γ‐aminobutyric acid (GABA)_A receptor antagonist, for 3 weeks (up to 6 mM) or 2 hr (10 mM). Cells were incubated in medium containing [U‐¹³C]glutamate (0.5 mM) and unlabeled glucose (3 mM) for 2 hr and cell extracts and media were analyzed by ¹³C magnetic resonance (MR) spectroscopy and high‐performance liquid chromatography (HPLC). When cerebellar astrocytes were incubated with PTZ for 2 hr, the amount of glucose removed from the medium and glucose and [U‐¹³C]glutamate oxidation were decreased. Metabolism in cortical astrocytes was affected only slightly; amounts of glutathione and aspartate were decreased. When cerebellar and cortical cells were cultured in the presence of PTZ for 3 weeks, the amount of glucose removed from the medium and lactate formed were increased, indicating increased glycolytic activity. Despite the increased intracellular [U‐¹³C]glutamate concentration in both types of astrocytes cultured with PTZ, labeled glutamine and glutathione were unchanged, indicating intracellular compartmentation. The amount of cellular protein was decreased at 6 mM PTZ for cerebellar astrocytes and 1 mM for cortical astrocytes, indicating a differential sensitivity to the effects of PTZ. In conclusion, mitochondrial metabolism and glycolysis were decreased by short‐term incubation with PTZ in cerebellar astrocytes, whereas long‐term incubation affected both types of astrocytes, leading to increased glycolysis.

Neuronal function is highly influenced by the extracellular environment. To study the effect of the milieu on neurons from cerebellum and neocortex, cells from these brain areas were cultured under different conditions. Two sets of cultures, one neocortical and one cerebellar neurons, were maintained in media containing [U‐¹³C]glucose for 8 days at initial concentrations of 12 and 28 mM glucose, respectively. Other sets of cultures (8 days in vitro) maintained in a medium containing initially 12 mM glucose were incubated subsequently for 4 hr either by addition of [U‐¹³C]glucose to the culture medium (final concentration 3 mM) or by changing to fresh medium containing [U‐¹³C]glucose (3 mM) but without glutamine and fetal calf serum. ¹³C Nuclear magnetic resonance (NMR) spectra revealed extensive γ‐aminobutyric acid (GABA) synthesis in both cultured neocortical and cerebellar neurons after maintenance in medium containing [U‐¹³C]glucose for 8 days, whereas no aspartate labeling was observed in these spectra. Mass spectrometry analysis, however, revealed high labeling intensity of aspartate, which was equal in the two types of neurons. Addition of [U‐¹³C]glucose (4 hr) on Day 8 in culture led to a similar extent of labeling of GABA in neocortical and in cerebellar cultures, but the cellular content of GABA was considerably higher in the neocortical neurons. The cellular content of alanine was similar regardless of culture type. Comparing the amount of labeling, however, cerebellar neurons exhibited a higher capacity for alanine synthesis. This is compatible with the fact that cerebellar neurons could ameliorate a low alanine content after culturing in low glucose (12 mM) by a 4‐hr incubation in medium containing 3 mM glucose. A low glucose concentration during the culture period and a subsequent medium change were associated with decreases in glutathione and taurine contents. Moreover, glutamate and GABA contents were reduced in cerebellar cultures under either of these conditions. In neocortical neurons, the GABA content was decreased by simultaneous exposure to low glucose and change of medium. These conditions also led to an increase in the aspartate content in both types of cultures, although most pronounced in the neocortical neurons. Further experiments are needed to elucidate these phenomena that underline the impact of extracellular environment on amino acid homeostasis.

Culturing mouse cerebellar neurones (predominantly glutamatergic) in the presence of [1−¹³C]glucose for 7 days resulted in a surprisingly extensive labelling of the inhibitory neurotransmitter GABA, the average content and labelling of which were 20 ± 4 nmol/mg protein and 20 ± 4%, respectively. Cultures of neocortical neurones (predominantly GABAergic) had under similar conditions a GABA content and labelling of 32 ± 2 nmol/mg protein and 21 ± 2%. The cerebellar cultures contained only 6% glutamate decarboxylase (GAD)‐positive neurones when immunolabelled using a GAD67 antibody, while a dense network of neurones in the neocortical cultures stained positively for GAD67. Exposure of the cerebellar cultures to 50 µm kainic acid (KA) which is known to eliminate vesicular release of GABA, only marginally affected GABA labelling and cellular content. Likewise this treatment had no effect on the number of GAD67‐positive neurones but a massive punctate immunostaining observed in control cultures was essentially eliminated. Increasing the KA concentration to 0.5 mm in the culture medium for 7 days led to a reduction of GABA labelling and content compared to cerebellar cultures not exposed to KA. Although it is likely that this large capacity for GABA synthesis resides in the relatively few GAD‐positive neurones, it seems unlikely that they could account for the large average GABA content in the cultures. Therefore it must be concluded that the newly synthesized GABA is redistributed among the majority of the cells in these cultures, i.e. the glutamatergic neurones.

Nuclear magnetic resonance spectroscopy (NMRS) has been used extensively for several decades to elucidate metabolic pathways in biological systems and has led to new insights into cerebral metabolism. Many of these insights have been gained by NMRS using in vitro models, such as tissue extracts, cell cultures or brain slices. Extracts of cells or tissue provide an excellent basis for metabolic studies and facilitate the interpretation of in vivo spectra. ¹³C NMRS is at present mostly used for in vitro or animal studies. Conclusions from the epilepsy models presented in this review are that turnover of metabolites is time‐dependant in kainate‐injected rats with limbic seizures. Early and only temporarily enhanced astrocytic activity is followed by altered metabolism in neurons with an increased turnover of important amino acids such as GABA and glutamate. However, pentylenetetrazole (PTZ) kindling affects astrocytes in younger and glutamatergic neurons in older animals. In the presence of PTZ, phenobarbital decreases labeling of most metabolites in all cell types, except GABAergic neurons, from both labeled precursors in the younger animals. However, in older animals only GABAergic neurons are affected by phenobarbital, as indicated by an increase in GABA labeling. In kaolin‐induced hydrocephalus it was shown that astrocyte metabolism is disturbed in the early phase, particularly in the cerebrum. These alterations continue into the chronic period. Only then do the first signs of neuronal metabolic impairment appear, which might explain why dementia is a prominent clinical feature in patients with chronic hydrocephalus. Copyright © 2003 John Wiley & Sons, Ltd.

Seizures were induced in rats by intraperitoneal injection of pentylenetetrazole (PTZ, 70 mg/kg), followed, 30 min later, by injection of [1‐¹³C]glucose and [1,2‐¹³C]acetate. Analyses of extracts from cortex, subcortex and cerebellum were performed using ¹³C magnetic resonance spectroscopy and HPLC. It could be shown that PTZ affected different brain regions differently. The total amounts of glutamate, glutamine, GABA, aspartate and taurine were decreased in the cerebellum and unchanged in the other brain regions. GABAergic neurones in the cortex and subcortex were not affected, whereas those in the cerebellum showed a pronounced decrease of GABA synthesis. However, glutamatergic neurones in all brain regions showed a decrease in glutamate labelling and in addition a decreased turnover in cerebellum. It could be shown that this decrease was in the metabolic pool of glutamate whereas release of glutamate was unaffected since glutamine labelling from glutamate was unchanged. Aspartate turnover was also decreased in all brain regions. Changes in astrocyte metabolism were not detected, indicating that PTZ had no effect on astrocyte metabolism in the early postictal stage.

The proper diagnosis of the arrested or the progressive form of hydrocephalus has a critical impact on treatment, but remains difficult. The assessment of early changes in cerebral metabolism might help in the development of adequate non‐invasive diagnostic tools. This study examined the alterations in label incorporation in neurotransmitter amino acids and other compounds in kaolin‐induced progressive hydrocephalus in rats by means of magnetic resonance spectroscopy (MRS) combined with the administration of [1‐¹³C]glucose and [1,2‐¹³C]acetate. Some 2, 4 and 6 weeks after kaolin injection into the cisterna magna, cerebrum, brainstem and cerebellum were dissected. Interestingly, labelling of most amino acids derived from [1‐¹³C]glucose showed no alterations, whereas labelling from [1,2‐¹³C]acetate was affected. Two weeks after induction of hydrocephalus the taurine concentration was decreased, whereas the concentration of [1,2‐¹³C]lactate was increased in the cerebrum and that of [1,2‐¹³C]GABA in the brainstem. Furthermore, labelling from [1,2‐¹³C]acetate was significantly decreased in [4,5‐¹³C]glutamate, [1,2‐¹³C]glutamate and [1,2‐¹³C]GABA in cerebrum from 4 weeks after hydrocephalus induction. The concentration ofN‐acetylaspartate, a neuronal marker, was unchanged. However, labelling of the acetyl group from [1‐¹³C]glucose was decreased in cerebellum and brainstem at 6 weeks after the induction of hydrocephalus. As glucose is metabolized predominately by neurones, whereas acetate is exclusively taken up by astrocytes, these results indicate that mostly astrocytic, and only later neuronal, metabolism is disturbed in the kaolin model of hydrocephalus. If verified in patients usingin vivoMRS, impaired astrocyte metabolism might serve as an early indication for operative treatment.

Cerebellar granule neurons were incubated with or without glucose (3 mM) in the presence or absence of citrate (20 mM) using normoxic and/or hypoxic incubation conditions. During 4 h of hypoglycemia...