Astrocyte-neurone communication following oxygen-glucose deprivation.

We looked at the possible interactions between astrocytes and neurones during reperfusion using an in vitro model of ischaemia-reperfusion injury, as a controlled environment that lends itself easily to manipulation of the numerous variables involved in such an insult. We constructed a chamber in which O2 can be lowered to a concentration of 1 microm and developed a primary cortical neuronal culture that is 99% pure and can survive to at least 10 days in vitro. We also established a novel system for the co-culture of astrocytes and neurones in order to study the communication between these cells in a manner that allows the complete separation of one cell type from another. Neurone cultures showed profound cell death following an ischaemic period of only 15 min. We co-cultured neurones that had been subjected to a 15-min ischaemic insult with either non-insulted astrocytes or astrocyte-conditioned medium during the reperfusion stage. Both astrocytes and astrocyte-conditioned medium enhanced neuronal survival. Our data also suggest that astrocyte-sourced neuronal glutathione synthesis may play a role in preventing neuronal death.

Alterations in expression of dopamine receptors and neuropeptides in the striatum of GTP cyclohydrolase-deficient mice.

The hph-1 mice have defective tetrahydrobiopterin biosynthesis and share many neurochemical similarities with l-dopa-responsive dystonia (DRD) in humans. In both, there are deficiencies in GTP cyclohydrolase I and low brain levels of dopamine (DA). Striatal tyrosine hydroxylase (TH) levels are decreased while the number of DA neurones in substantia nigra (SN) appears normal. The hph-1 mouse is therefore a useful model in which to investigate the biochemical mechanisms underlying dystonia in DRD. In the present study, the density of striatal DA terminals and DA receptors and the expression of D-1, D-2, and D-3 receptors, preproenkephalin (PPE-A), preprotachykinin (PPT), and nitric oxide synthase (NOS) mRNAs in the striatum and nucleus accumbens and nigral TH mRNA expression were examined. Striatal DA terminal density as judged by specific [3H]mazindol binding was not altered while the levels of TH mRNA were elevated in the SN of hph-1 mice compared to control (C57BL) mice. Total and subregional analysis of the striatum and nucleus accumbens showed that D-2 receptor ([3H]spiperone) binding density was increased while D-1 receptor ([3H]SCH 23390) and D-3 receptor ([3H]7-OH-DPAT) binding density was not altered. In the striatum and nucleus accumbens, expression of PPT mRNA was elevated but PPE-A mRNA, D-1, D-2 receptor, and nNOS mRNA were not changed in hph-1 mice compared to controls. These findings suggest that an imbalance between the direct strionigral and indirect striopallidal output pathways may be relevant to the genesis of DRD. However, the pattern of changes observed is not that expected as a result of striatal dopamine deficiency and suggests that other effects of GTP cyclohydrolase I deficiency may be involved.

Beta-amyloid fragment 25-35 causes mitochondrial dysfunction in primary cortical neurons.

Beta-amyloid deposition and compromised energy metabolism both occur in vulnerable brain regions in Alzheimer's disease. It is not known whether beta-amyloid is the cause of impairment of energy metabolism, nor whether impaired energy metabolism is specific to neurons. Our results, using primary neuronal cultures, show that 24-h incubation with A beta(25-35) caused a generalized decrease in the specific activity of mitochondrial enzymes per milligram of cellular protein, induced mitochondrial swelling, and decreased total mitochondrial number. Incubation with A beta(25-35) decreased ATP concentration to 58% of control in neurons and 71% of control in astrocytes. Levels of reduced glutathione were also lowered by A beta(25-35) in both neurons (from 5.1 to 2.9 nmol/mg protein) and astrocytes (from 25.2 to 14.9 nmol/mg protein). We conclude that 24-h treatment with extracellular A beta(25-35) causes mitochondrial dysfunction in both astrocytes and neurons, the latter being more seriously affected. In astrocytes mitochondrial impairment was confined to complex I inhibition, whereas in neurons a generalized loss of mitochondria was seen.

Beta-amyloid inhibits integrated mitochondrial respiration and key enzyme activities.

Disrupted energy metabolism, in particular reduced activity of cytochrome oxidase (EC 1.9.3.1), alpha-ketoglutarate dehydrogenase (EC 1.2.4.2) and pyruvate dehydrogenase (EC 1.2.4.1) have been reported in post-mortem Alzheimer's disease brain. beta-Amyloid is strongly implicated in Alzheimer's pathology and can be formed intracellularly in neurones. We have investigated the possibility that beta-amyloid itself disrupts mitochondrial function. Isolated rat brain mitochondria have been incubated with the beta-amyloid alone or together with nitric oxide, which is known to be elevated in Alzheimer's brain. Mitochondrial respiration, electron transport chain complex activities, alpha-ketoglutarate dehydrogenase activity and pyruvate dehydrogenase activity have been measured. Beta-amyloid caused a significant reduction in state 3 and state 4 mitochondrial respiration that was further diminished by the addition of nitric oxide. Cytochrome oxidase, alpha-ketoglutarate dehydrogenase and pyruvate dehydrogenase activities were inhibited by beta-amyloid. The K(m) of cytochrome oxidase for reduced cytochrome c was raised by beta-amyloid. We conclude that beta-amyloid can directly disrupt mitochondrial function, inhibits key enzymes and may contribute to the deficiency of energy metabolism seen in Alzheimer's disease.

Contrasting effects of N5-substituted tetrahydrobiopterin derivatives on phenylalanine hydroxylase, dihydropteridine reductase and nitric oxide synthase.

Tetrahydrobiopterin [(6R)-5,6,7,8-tetrahydro-L-biopterin, H(4)biopterin] is one of several cofactors of nitric oxide synthases (EC 1.14.13.39). Here we compared the action of N(5)-substituted derivatives on recombinant rat neuronal nitric oxide synthase with their effects on dihydropteridine reductase (EC 1.6.99.7) and phenylalanine hydroxylase (EC 1.14.16.1),the well-studied classical H(4)biopterin-dependent reactions. H(4)biopterin substituted at N(5) with methyl, hydroxymethyl, formyl and acetyl groups were used. Substitution at N(5) occurs at a position critical to the redox cycle of the cofactor in phenylalanine hydroxylase/dihydropteridine reductase. We also included N(2)'-methyl H(4)biopterin, a derivative substituted at a position not directly involved in redox cycling, as a control. As compared with N(5)-methyl H(4)biopterin, N(5)-formyl H(4)biopterin bound with twice the capacity but stimulated nitric oxide synthase to a lesser extent. Depending on the substituent used, N(5)-substituted derivatives were redox-active: N(5)-methyl- and N(5)-hydroxyl methyl H(4)biopterin, but not N(5)-formyl- and N(5)-acetyl H(4)biopterin, reduced 2,6-dichlorophenol indophenol. N(5)-Substituted H(4)biopterin derivatives were not oxidized to products serving as substrates for dihydropteridine reductase and,depending on the substituent, were competitive inhibitors of phenylalanine hydroxylase: N(5)-methyl- and N(5)-hydroxymethyl H(4)biopterin inhibited phenylalanine hydroxylase, whereas N(5)-formyl- and N(5)-acetyl H(4)biopterin had no effect. Our data demonstrate differences in the mechanism of stimulation of phenylalanine hydroxylase and nitric oxide synthase by H(4)biopterin. They are compatible with a novel, non-classical, redox-active contribution of H(4)biopterin to the catalysis of the nitric oxide synthase reaction.

Stimulation of the brain NO/cyclic GMP pathway by peripheral administration of tetrahydrobiopterin in the hph-1 mouse.

Mutations in GTP-cyclohydrolase I (GTP-CH) have been identified as causing a range of inborn errors of metabolism, including dopa-responsive dystonia. GTP-CH catalyses the first step in the biosynthesis of tetrahydrobiopterin (BH4), a cofactor necessary for the synthesis of catecholamines and serotonin. Current therapy based on monoamine neurotransmitter replacement may be only partially successful in correcting the neurological deficits. The reason might be that BH4 is also a cofactor for nitric oxide synthase. Using a strain of mutant GTP-CH-deficient (hph-1) mice, we demonstrate that in addition to impaired monoamine metabolism, BH4 deficiency is also associated with diminished nitric oxide synthesis in the brain (as evaluated by measuring the levels of cyclic GMP), when compared with wild-type animals. We have found a decline in the levels of BH4 with age in all animals, but no gender-related differences. We found a strong association between the levels of BH4 and cyclic GMP in hph-1 mice but not in wild-type animals. We also demonstrate that acute peripheral administration of BH4 (100 micromol/kg s.c.) in hph-1 mice significantly elevated the brain BH4 concentration and subsequently cyclic GMP levels in cerebellum, with peaks at 2 and 3 h, respectively. We suggest that BH4 administration should be considered in BH4 deficiency states in addition to monoamine replacement therapy.

beta-Amyloid fragment 25-35 selectively decreases complex IV activity in isolated mitochondria.

Defects in mitochondrial oxidative metabolism, in particular decreased activity of cytochrome c oxidase, have been demonstrated in Alzheimer's disease, and after the expression of the amyloid precursor protein (APP) in cultured cells, suggesting that mitochondria might be involved in beta-amyloid toxicity. Recent evidence suggests that the proteolysis of APP to generate beta-amyloid is at least in part intracellular, preceding the deposition of extracellular fibrils. We have therefore investigated the effect of incubation of isolated rat brain mitochondria with the beta-amyloid fragment 25-35 (100 microM) on the activities of the mitochondrial respiratory chain complexes I, II-III, IV (cytochrome c oxidase) and citrate synthase. The peptide caused a rapid, dose-dependent decrease in the activity of complex IV, white it had no effect on the activities on any of the other enzymes tested. The reverse sequence peptide (35-25) had no effect on any of the activities measured. We conclude that inhibition of mitochondrial complex IV might be a contributing factor to the pathogenesis of Alzheimer's disease.

Effect of postischaemic hypothermia on the mitochondrial damage induced by ischaemia and reperfusion in the gerbil.

In order to test the effect of hypothermia on mitochondrial function damage following cerebral ischaemia/reperfusion, Mongolian gerbils were submitted to 30 min bilateral carotid occlusion and 2 h of reperfusion at 37 degreesC or 30 degreesC. After normothermic (37 degreesC) ischaemia/reperfusion, significant decreases in mitochondrial state 3 (+ADP) oxygen consumption (-42.2%), complex II-III activity in synaptosomes (-31.7%) and complex IV were measured, in both free mitochondria and synaptosomes (-30.3% and -27. 8% respectively). However, following hypothermic (30 degreesC) reperfusion, both respiration rates and all enzyme activities remained at levels not significantly different from those in the sham operated controls.

Molecular cloning of rat mitochondrial 3-hydroxy-3-methylglutaryl-CoA lyase and detection of the corresponding mRNA and of those encoding the remaining enzymes comprising the ketogenic 3-hydroxy-3-methylglutaryl-CoA cycle in central nervous system of suckling rat.

We have investigated, by RNase protection assays in rat brain regions and primary cortical astrocyte cultures, the presence of the mRNA species encoding the three mitochondrially located enzymes acetoacetyl-CoA thiolase, mitochondrial 3-hydroxy-3-methylglutaryl-CoA synthase (mt. HMG-CoA synthase) and HMG-CoA lyase (HMG-CoA lyase) that together constitute the ketogenic HMG-CoA cycle. As a prerequisite we obtained a full-length cDNA encoding rat HMG-CoA lyase by degenerate oligonucleotide-primed PCR coupled to a modification of PCR-rapid amplification of cDNA ends (PCR-RACE). We report here: (1) the nucleotide sequence of rat mt. HMG-CoA lyase, (2) detection of the mRNA species encoding all three HMG-CoA cycle enzymes in all regions of rat brain during suckling, (3) approximately twice the abundance of mt. HMG-CoA synthase mRNA in cerebellum than in cortex in 11-day-old suckling rat pups, (4) significantly lower abundances of mt. HMG-CoA synthase mRNA in brain regions derived from rats weaned to a high-carbohydrate/low-fat diet compared with the corresponding regions derived from the suckling rat, and (5) the presence of mt. HMG-CoA synthase mRNA in primary cultures of neonatal cortical astrocytes at an abundance similar to that found in liver of weaned animals. These results provide preliminary evidence that certain neural cell types possess ketogenic potential and might thus have a direct role in the provision of fatty acid-derived ketone bodies during the suckling period.

Spatial training and high-frequency stimulation engage a common pathway to enhance glutamate release in the hippocampus.

We have measured depolarization-induced release of endogenous glutamate in synaptosomes prepared from the dentate gyrus after the induction of LTP by high-frequency stimulation in anesthetized rats, and after training in the water maze. Both spatial training and LTP in untrained rats were accompanied by an increase in glutamate release from dentate synaptosomes. The enhancement of synaptosomal glutamate release induced by high-frequency stimulation was abolished in well-trained rats, and was reduced in partially trained rats and in rats trained in a nonspatial task. However, the magnitude of LTP was similar in well-trained and untrained groups. These results indicate that spatial training activates a glutamate release pathway that converges with that activated in LTP, and demonstrate an unexpected dissociation between increased glutamate release and LTP.

Threshold effects in synaptosomal and nonsynaptic mitochondria from hippocampal CA1 and paramedian neocortex brain regions.

After a brief period of global ischemia, the hippocampal CA1 region is more susceptible to irreversible damage than the paramedian neocortex. To test whether primary differences in bioenergetic parameters may be present between these regions, respiration rates and respiratory control activities were measured. In synaptosomal and nonsynaptic mitochondria isolated from the hippocampal CA1 region, state 3 respiration rates and complex IV activities were significantly lower than those present in synaptosomal and nonsynaptic mitochondria from the paramedian neocortex. These results suggest that mitochondria from the CA1 hippocampal area differ in some properties of metabolism compared with the neocortex area, which may render them more susceptible to a toxic insult such as that of ischemia. In addition, when complex I and IV activities were titrated with specific inhibitors, thresholds in ATP synthesis and oxygen respiration became apparent. Complex I and IV activities were decreased by 60% in nonsynaptic mitochondria from the hippocampal CA1 region and paramedian neocortex before oxidative phosphorylation was severely compromised; however, in synaptosomes from these regions, complex I activities had a threshold of 25%, indicating heterogenous behaviour for brain mitochondria. Reduced complex I thresholds in mitochondria, in association with other constitutive defects in energy metabolism, may induce a decreased ATP supply in the synaptic region. The implications of these findings are discussed in relation to delayed neuronal death and processes of neurodegeneration.

Activity of mitochondrial respiratory chain enzymes after transient focal ischemia in the rat.

Previous results demonstrated that after 2-hour middle cerebral artery occlusion (MCAO) in the rat, 1- to 2-hour recirculation temporarily restored the bioenergetic state and mitochondrial function, but secondary deterioration took place after 4 hours. The authors measured the activity of mitochondrial respiratory chain complexes, citrate synthase, and glutamate dehydrogenase as possible targets of secondary damage. Focal and penumbral tissues were sampled in the control condition, after 2 hours of MCAO, and after 1, 2, or 4 hours of postischemic recirculation; two groups were treated with alpha-phenyl-N-tert-butyl-nitrone (PBN). Complex IV activity transiently decreased after MCAO, but after recirculation all measured activities returned to control values.

Inhibitors of PLA2 and NO synthase cooperate in producing amnesia of a spatial task.

Both nitric oxide and arachidonic acid have been suggested to play a role as a retrograde messenger in synaptic plastic changes which underlie memory formation. However, inhibitors of nitric oxide (NO) synthesis or of arachidonic acid release have produced contradictory results. We suggest a model which involves simultaneous release of both messenger types which can compensate for the loss of one type. To test this theory, rats were injected either with a drug that inhibits release of arachidonic acid, or a drug that inhibits synthesis of NO, or with both drugs. Acquisition of a water maze task was not different between groups. In the test sessions, however, animals injected with both drugs showed marked amnesic symptoms, while the groups injected with a low dose of nordihydroguaiaretic acid (NDGA) or with nitro-L-arginine (L-NARG) showed a trend towards amnesia. The effect of both drugs appears to be additive. The results support the proposed theory.

Long-term potentiation and glutamate release in the dentate gyrus: links to spatial learning.

Long-term potentiation (LTP) in the dentate gyrus of the anaesthetized rat is associated with a persistent increase in the concentration of glutamate in the extracellular compartment. At the in vitro level, this is mirrored by an increase in the ability of slices or synaptosomes from potentiated tissue to release glutamate in response to a depolarizing stimulus. In both cases, the activity-induced enhancement of glutamate release is dependent on the activation of the NMDA receptor. A similar increase in glutamate release in vitro is observed in the dentate gyrus prepared from rats trained in a variety of learning tasks, including classical conditioning and the Morris water maze. These results are consistent with the hypothesis that similar presynaptic mechanisms are engaged in LTP and learning.

LTP in the dentate gyrus is associated with a persistent NMDA receptor-dependent enhancement of synaptosomal glutamate release.

LTP was induced unilaterally in the dentate gyrus of the anaesthetized rat by tetanic stimulation of the perforant path. Two hours after tetanization synaptosomes were prepared from control and potentiated sides, and the depolarization-induced release of endogenous amino acid neurotransmitters was compared in the two groups, using 4-aminopyridine, potassium or veratridine as depolarizing agents. Glutamate release was greater in synaptosomes from potentiated than control tissue with all three depolarizing agents, but the highest release from control synaptosomes and the most significant enhancement of release from potentiated synaptosomes was seen with veratridine. The tetanus-induced enhancement of veratridine-induced glutamate release was prevented by pretreatment with MK-801. The results indicate that the induction of LTP in the dentate gyrus is accompanied by a persistent enhancement in the ability of synaptosomes to release glutamate in response to a depolarizing stimulus.

High-performance liquid chromatographic separation with electrochemical detection of amino acids focusing on neurochemical application.

Twenty-three amino acids and dipeptides, including compounds of neurochemical interest, are measured by high-performance liquid chromatography using electrochemical detection after precolumn derivatization with o-phthalaldehyde/beta-mercaptoethanol. The method uses a multistep polarity gradient system and the entire separation is performed in less than 23 min of analysis. The minimum detectable quantity was 0.66 pmol injected, corresponding to 50 nM concentration in the sample; the response was linear in the tested range of 1.33-1333 pmol (0.1-100 microM). Relative standard deviations ranged from 0.75 to 6.089% for area measurements (mean, 2.33) and from 0.209 to 0.779% for retention times (mean, 0.546). Examples of application of the method to analysis of biological samples such as rat brain homogenate and human cerebrospinal fluid are shown.

Glutamate and GABA levels in CSF from patients affected by dementia and olivo-ponto-cerebellar atrophy.

The modifications in the CSF content of glutamate and GABA in patients afflicted with primary degenerative dementia (PDD) and olivo-ponto-cerebellar atrophy (OPCA) have been evaluated. Control subjects (with disk herniation) were also included in the study. The amino-acids assays were carried out utilizing enzymatic-bioluminescence technique. GABA levels in controls were 803 +/- 98 (n = 7) and in demented patients 702 +/- 98 (n = 7) pmol/ml. Glutamate levels were 2067 +/- 244 (n = 10) in controls, 1190 +/- 81 (n = 16) pmol/ml (vs controls p less than 0.01) in demented patients, and 1116 +/- 146 (vs controls p less than 0.01) in OPCA patients. These results suggest that CSF glutamate levels in severely demented patients might be a result of generalized neuronal loss in the brain with a reactive gliosis.

Effects of calcium antagonists on glycolysis of rat brain synaptosomes.

The effects of calcium antagonists nimodipine, nicardipine and flunarizine on lactate production and specific activities of some enzymes regulating glycolytic flux have been evaluated in synaptosomes isolated from rat whole brain and submitted to in vitro chemical hypoxia induced by rotenone, an inhibitor of mitochondrial respiration. The following enzymes have been tested; hexokinase (ATP: D-hexose-6-phosphotransferase, EC2.7.1.1), phosphofructokinase (ATP: D-fructose-6-phosphate 1-phosphotransferase, EC 2.7.1.11) and pyruvate kinase (ATP: pyruvate 2-O-phosphotransferase, EC 2.7.1.40). The results show that rotenone increases by about eight times the production of lactate; nicardipine and nimodipine, starting from a concentration of 10(-4) M, were able to counteract the rotenone-induced stimulation of glycolysis, but flunarizine was without effect. The dihydropyridines but not flunarizine decreased the maximum activity of phosphofructokinase. This effect was already detectable at a concentration of 10(-5) M. Neither hexokinase nor pyruvate kinase were affected by any of the drugs studied.