Excitotoxic mechanisms in hypoglycaemic hippocampal injury

Neuroprotective effects of silymarin on ischemia-induced delayed neuronal cell death in rat hippocampus

We examined the effects of silymarin, which was extracted from Silybum marianum, on delayed neuronal cell death in the rat hippocampus. Rats were divided into four groups: sham-operated rats (sham group), rats which underwent ischemic surgery (control group), rats which were treated with silymarin before and after ischemic surgery (pre group), and rats which were treated with silymarin after ischemic surgery only (post group). We performed the ischemic surgery by occluding the bilateral carotid arteries for 20min and sacrificed the rats one week after the surgery. Silymarin was administered orally at 200mg/kg body weight. Smaller numbers of delayed cell deaths were noted in the rat CA1 region of the pre- and post-groups, and no significant difference was observed between these groups. There were few apoptotic cell deaths in all groups. Compared to the control group, significantly fewer cell deaths by autophagy were found in the pre- and post-group. We concluded that silymarin exerts a preservation effect on delayed neuronal cell death in the rat hippocampus and this effect has nothing to do with the timing of administering of silymarin.

Cellular mechanisms of brain hypoglycemia

Data on intracellular processes induced by a low glucose level in nerve tissue are presented. The involvement of glutamate and adenosine receptors, mitochondria, reactive oxygen species (ROS), and calcium ions in the development of hypoglycemia-induced damage of neurons is considered. Hypoglycemia-induced calcium overload of neuronal mitochondria is suggested to be responsible for the increased ROS production by mitochondria.

The Phenomenon of “Pe-ischaemic Conditioning” in the Brain only Partly involves the NMDA Receptor: A Magnetic Resonance Study

We have investigated in more detail our previous observations on a form of ischaemic pre-conditioning "metabolic adaptation", i.e.--that sequential metabolic insults (hypoxia followed 40 min later by combined hypoxia + hypoglycaemia, or vice versa) are less injurious (monitored by increased [Ca2+]i and decreased PCr) than the immediate combined insult. We have now observed that the "adaptation" occurs between 10 and 20 min. Pre-treatment of the tissues with 10 microM-MK801 showed that it had no effect on the increase in [Ca2+]i caused by the sequential insult and only partially blocked the increase observed by exposure to the immediate combined insult. Exposure to both the delayed and immediate combined insults with low extracellular Ca2+ resulted in a two-fold increase in [Ca2+]i, similar to the increase observed with normal extracellular Ca2+ in the presence of MK801. The results are discussed in terms of the possible origins of the increases in [Ca2+]i.

Effects of fructose-1,6-bisphosphate on morphological and functional neuronal integrity in rat hippocampal slices during energy deprivation

D-fructose-1,6-bisphosphate, a high energy glycolytic intermediate, attenuates ischemic damage in a variety of tissues, including brain. To determine whether D-fructose-1,6-bisphosphate serves as an alternate energy substrate in the CNS, rat hippocampal slices were treated with D-fructose-1,6-bisphosphate during glucose deprivation. Unlike pyruvate, an endproduct of glycolysis, 10 mM D-fructose-1,6-bisphosphate did not preserve synaptic transmission or morphological integrity of CA1 pyramidal neurons during glucose deprivation. Moreover, during glucose deprivation, 10-mM D-fructose-1,6-bisphosphate failed to maintain adenosine triphosphate levels in slices. D-fructose-1,6-bisphosphate, however, attenuated acute neuronal degeneration produced by 200 microM iodoacetate, an inhibitor of glycolysis downstream of D-fructose-1,6-bisphosphate. Because (5S, 10R)-(+)-5-methyl-10, 11-dihydro-5H-dibenzo [a,d]cyclohepten-5,10-imine, an antagonist of N-methyl-D-aspartate receptors, exhibited similar protection against iodoacetate damage, we examined whether (5S, 10R)-(+)-5-methyl-10, 11-dihydro-5H-dibenzo [a,d]cyclohepten-5,10-imine and D-fructose-1,6-bisphosphate share a common neuroprotective mechanism. Indeed, D-fructose-1,6-bisphosphate diminished N-methyl-D-aspartate receptor-mediated synaptic responses and partially attenuated neuronal degeneration induced by 100-microM N-methyl-D-aspartate. Taken together, these results indicate that D-fructose-1,6-bisphosphate is unlikely to serve as an energy substrate in the hippocampus, and that neuroprotective effects of D-fructose-1,6-bisphosphate are mediated by mechanisms other than anaerobic energy supply.

Genomic responses of the brain to ischemic stroke, intracerebral haemorrhage, kainate seizures, hypoglycemia, and hypoxia

RNA expression profiles in rat brain were examined 24 h after ischemic stroke, intracerebral haemorrhage, kainate-induced seizures, insulin-induced hypoglycemia, and hypoxia and compared to sham- or untouched controls. Rat oligonucleotide microarrays were used to compare expression of over 8000 transcripts from three subjects in each group (n = 27). Of the somewhat less than 4000 transcripts called 'present' in normal or treated cortex, 5-10% of these were up-regulated 24 h after ischemia (415), haemorrhage (205), kainate (187), and hypoglycemia (302) with relatively few genes induced by 6 h of moderate (8% oxygen) hypoxia (15). Of the genes induced 24 h after ischemia, haemorrhage, and hypoglycemia, approximately half were unique for each condition suggesting unique components of the responses to each of the injuries. A significant component of the responses involved immune-process related genes likely to represent responses to dying neurons, glia and vessels in ischemia; to blood elements in haemorrhage; and to the selectively vulnerable neurons that die after hypoglycemia. All of the genes induced by kainate were also induced either by ischemia, haemorrhage or hypoglycemia. This strongly supports the concept that excitotoxicity not only plays an important role in ischemia, but is an important mechanism of brain injury after intracerebral haemorrhage and hypoglycemia. In contrast, there was only a single gene that was down-regulated by all of the injury conditions suggesting there is not a common gene down-regulation response to injury.

Neuroprotection of cultured foetal rat hippocampal cells against glucose deprivation: are GABAergic neurons less vulnerable or more sensitive to TCP protection?

In the rat brain, hippocampal neurons are particularly sensitive to secondary excitotoxic injury induced by ischaemia or hypoglycaemia. To determine some distinctive features of vulnerability among neuronal phenotypes in the hippocampus following a metabolic insult, we used an in vitro model of mild glucose deprivation. Primary cultures from the rat hippocampus (21 days in vitro) were deprived of glucose for 4 h and then were returned to the standard medium for 24 or 48 h. Survival of the GABAergic neuronal population was evaluated both by measuring [3H]GABA uptake and by counting GAD65-immunostained cells. This was compared with the survival of the total neuronal population evaluated by counting the neurofilament-200-immunostained cells. Glucose deprivation for 4 h followed by a recovery period of 48 h induced a decrease of 59% and 40% in the number of GAD65- and neurofilament-200-immunostained cells, respectively. Thus, GABAergic neurons were slightly more vulnerable to glucose deprivation than the other neurons in the hippocampal cell cultures. When the excitotoxic component of cellular death was blocked in the presence of TCP, an NMDA-antagonist, the survival of GABAergic neurons was almost complete after 48 h of recovery. In contrast, measurements of the release of lactate dehydrogenase in the medium indicated that TCP largely protected hippocampal cells after 24 h but was ineffective after 48 h. This observation was confirmed by immunostaining data which showed that after 48 h TCP did not significantly increase the survival of neurofilament-200-immunostained cells. These results indicate that after glucose deprivation and a recovery period of 48 h, GABAergic neurons in hippocampal cell cultures are not more resistant than other neurons but they are more sensitive to TCP protection.

Role of excitatory aminoacids in neonatal hypoglycemia

BACKGROUND

In many neurological disorders, injury to neurons may be due in part to overstimulation of the receptors for the excitatory amino acids glutamate and aspartate. The same excitotoxic mechanism and high aspartate levels in experimental studies led to this study of the concentrations of glutamate and aspartate and zinc, copper, and magnesium levels in the cerebrospinal fluid (CSF) of hypoglycemic newborns.

METHODS

Aspartate and glutamate were determined by high-performance liquid chromatography, and magnesium, zinc and copper by atomic absorption spectrophotometer.

RESULTS

The CSF levels of aspartate (3.98 +/- 1.77 mumol/L) and glutamate (1.7 +/- 1.05 mumol/L) in 20 hypoglycemic newborns were significantly higher when compared with the values of aspartate (2.19 +/- 0.6 mumol/L) and glutamate (0.77 +/- 0.34 mumol/L) of 10 control newborns. In the hypoglycemic patients, the concentration of zinc (0.57 +/- 0.13 microgram/mL), but not copper (0.39 +/- 0.40 microgram/mL) was significantly lower when compared with the control values. There was no difference in the magnesium levels between the two groups.

CONCLUSIONS

The higher levels of excitatory amino acids found in the CSF of hypoglycemic infants than in controls were consistent with previous animal studies, which may indicate the role of excitatory amino acids in the late biochemical effects of hypoglycemia in newborn brain metabolism.

Vulnerability of CA1 neurons to glutamate is developmentally regulated

Although it is well documented that glutamate receptor subtypes are differentially expressed during central nervous system development postnatally, how glutamate affects neurons during postnatal development is unclear. We therefore examined the development of the intrinsic neuronal response to glutamate receptor activation by studying single, hippocampal CA1 neurons that had been acutely dissociated from newborn (P1-3), 1 week old (P6-8), and 3 week old (P21-25) rats. Using laser scanning confocal microscopy and the calcium dye Fluo-3, we made time-lapse studies of the effects of glutamate stimulation on free intracellular calcium ([Ca2+]i) and simultaneous changes in neuronal morphology. In P21-25 neurons, glutamate increased [Ca2+]i fluorescence, and caused marked somal swelling, blebbing, and retraction of dendrites into the soma. These major morphological changes were followed by sudden loss of intracellular fluorescence, indicative of a loss of membrane integrity and cell death. In P6-8 neurons, glutamate increased [Ca2+]i to the same extent, but this increase was not followed by either major morphological changes or loss of membrane integrity. In P1-3 neurons, glutamate increased [Ca2+]i minimally, and no morphologic changes were observed. P1-3 neurons dissociated without enzymatic digestion demonstrated glutamate responses identical to responses seen in neurons dissociated with enzymatic digestion. In the presence of MK-801 (15 microM), glutamate still increased [Ca2+]i and caused cell death in P21-25 neurons, but the latency to these effects more than tripled. This late, MK-801-resistant [Ca2+]i increase was not eliminated by DNQX or Ni2+/Cd2+, suggesting that this increase is mediated by metabotropic receptors. These findings demonstrate that (1) hippocampal neurons from newborns are intrinsically less vulnerable to glutamate toxicity than neurons from 3 weeks old animals, and (2) multiple glutamate receptor subtypes affect the magnitude of the [Ca2+]i increase in response to glutamate in the neuronal microenvironment.

Kynurenate is neuroprotective following experimental brain injury in the rat

Pharmacologic inhibition of excitatory amino acid neurotransmission improves physiologic, metabolic, and neurobehavioral outcome following experimental brain trauma. However, no studies to date have demonstrated pharmacologically-induced attenuation of histopathological changes associated with experimental brain injury models. The present study examined the effects of kynurenate, an NMDA and non-NMDA receptor antagonist, on neuronal survival in the hippocampus after lateral fluid-percussion brain injury in the rat. Animals (n = 10/treatment) randomly received an intravenous injection of either kynurenate (300 mg/kg) or buffer (equal volume) 15 min following fluid-percussion brain injury of moderate severity. Two weeks after injury, animals were sacrificed and neuronal cell loss in the hippocampus was examined with Nissl staining. Selective loss of neurons in the CA3 region of the hippocampus, which has previously been characterized in this model of brain injury, was found to be significantly attenuated following kynurenate treatment (P < 0.05). These data suggest that pharmacologic compounds which are known to have beneficial effects on neurobehavioral and physiological outcome following brain injury may also significantly attenuate post-traumatic neuronal cell loss. Our results also support other recent data that pharmacological intervention with an excitatory amino acid receptor antagonist may be of therapeutic value in the treatment of brain injury.

Possible role of glutamatergic neurotransmission in regulating ethanol-evoked brain ascorbate release

It was found that systemic application of ethanol induced brain ascorbate (AA) release. In order to study the mechanism of ethanol-evoked AA release, the role of brain glutamatergic neurotransmission was investigated using in vivo voltammetry in the striatum of freely moving rats. Pretreatment with L-trans-pyrrolidine-2,4-dicarboxylate (PDC, 10 nmol, i.c.v.), a glutamate (Glu) uptake blocker, potentiated ethanol (1 g/kg, intraperitoneal injection, i.p.)-evoked release of brain AA. N-methyl-D-aspartate (NMDA, 1 nmol, i.c.v.) produced a fast transient increase in extracellular AA, whereas alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionate (AMPA, 1 nmol, i.c.v.) produced a decrease in extracellular AA (75.8 +/- 3% of control). Kainate (KA, 1 nmol, i.c.v.) produced an initial decrease (48.7 +/- 11.7% of control) then an increase (250 +/- 68.5% of control) in extracellular AA. These results suggest that systemic administration of ethanol may affect the release or uptake of brain glutamatergic neurotransmitters which appear to regulate brain AA release. The NMDA, but not the non-NMDA, type of Glu receptor may be responsible for this regulation.

Neither moderate hypoxia nor mild hypoglycaemia alone causes any significant increase in cerebral [Ca2+]i: only a combination of the two insults has this effect. A 31P and 19F NMR study

(1) The energy state and free intracellular calcium concentration ([Ca2+]i) of superfused cortical slices were measured in moderate hypoxia (approximately 65 microM O2), in mild hypoglycaemia (0.5 mM glucose), and in combinations of the two insults using 19F and 31P NMR spectroscopy. (2) Neither hypoxia nor hypoglycaemia alone caused any significant change in [Ca2+]i. Hypoxia caused a 40% fall in phosphocreatine (PCr) content but not in ATP level, and hypoglycaemia produced a slight fall in both (as expected from previous studies). These changes in the energy state recovered on return to control conditions. (3) A combined sequential insult (hypoxia, followed by hypoxia plus hypoglycaemia) produced a 100% increase in [Ca2+]i and a decrease in PCr level to approximately 25% of control. The reverse combined sequential insult (hypoglycaemia, followed by hypoglycaemia plus hypoxia) had the same effect. On return to control conditions there was some decrease in [Ca2+]i and a small increase in PCr content, but neither recovered to control levels. (4) Exposure of the tissue to the combined simultaneous insult (hypoxia plus hypoglycaemia) immediately after the control spectra had been recorded resulted in a fivefold increase in [Ca2+]i and a similar decrease in PCr level to 20-25% of control. There was little if any change of [Ca2+]i or PCr level on return to control conditions. (5) These results are discussed in terms of metabolic adaptation of some but not all of the cortical cells to the single type of insult, which renders the tissues less vulnerable to the combined insult.

Inhibition of [3H]-(+)-MK 801 binding to rat brain sections by CPP and 7-chlorokynurenic acid: an autoradiographic analysis

1. The regional binding of [3H]-(+)-5-methyl-10,11-dihydro-5H-dibenzo (a,d)cyclohepten-5,10-imine maleate ([3H]-(+)-MK 801) to sections of rat brain was measured by an in vitro quantitative autoradiographic technique. A heterogeneous distribution of binding sites was observed. 2. High values of binding were detected in the hippocampal formation and cerebral cortex, while very low binding was found in cerebellum. [3H]-(+)-MK 801 binding was not detectable in white matter tracts or in the brain stem. 3. [3H]-(+)-MK 801 binding was inhibited by increasing concentrations of both 7-chlorokynurenate (1-1000 microM) and ((+)-2-carboxypiperazine-4-yl)propyl-1-phosphonic acid (CPP) (0.1-100 microM). High concentrations of both drugs were able to inhibit completely specific [3H]-(+)-MK 801 binding. 4. IC50 values calculated for both 7-chlorokynurenate and CPP-induced [3H]-(+)-MK 801 binding inhibition were similar in all brain regions analyzed. 5. The inhibitory action of 7-chlorokynurenate and that of CPP on [3H]-(+)-MK 801 binding were reversed by addition of glycine and glutamate respectively. 6. It is concluded that activation of glycine and N-methyl-D-aspartate (NMDA) receptors is obligatory for the binding of [3H]-(+)-MK 801 to occur in all of the brain regions examined in the present study. Furthermore, on the basis of the similar regional sensitivities of [3H]-(+)-MK 801 binding to the inhibitory action of 7-chlorokynurenate and CPP, a single pharmacological classification of the NMDA receptor complex in brain is suggested. The cerebellum was not included in the study due to the very low level of [3H]-(+)-MK 801 binding detected under the experimental conditions used.

Ascorbic acid in the brain

Ascorbic acid is highly concentrated in the central nervous system. Measurement of the extracellular concentration of ascorbate in animals, mainly by the technique of voltammetry in vivo, has demonstrated fluctuation in release from neuropil, both spontaneously and in response to physical stimulation of the animal and to certain drugs. Although in the adrenal medulla ascorbate is co-released with catecholamines, release of ascorbate from brain cells is associated principally with the activity of glutamatergic neurones, mainly by glutamate-ascorbate heteroexchange across cell membranes of neurones or glia. This phenomenon is discussed in relation to a possible role of ascorbate as a neuromodulator or neuroprotective agent in the brain.

Neuropathology of the chronic epileptic syndrome induced by intrahippocampal tetanus toxin in rat: preservation of pyramidal cells and incidence of dark cells

A few nanograms of tetanus toxin injected into a rat hippocampus causes a chronic epileptic syndrome characterized by brief seizures that recur intermittently for about 6 weeks. Cognitive and other behavioural impairments persist after the seizures and other epileptic electrographic activity have remitted, and may be permanent. Our previous studies suggested that the behavioural changes following seizure remission were an indication of functional impairment associated with decreased neuronal excitability rather than with neuronal loss. The conclusion that neurons were preserved relied on qualitative histological observations and, indirectly, on electrophysiological measurements of the amplitudes of antidromic population spikes. Recently, gross histopathology has been described in a quantitative histological study of rats 7-10 days after they had received rather higher doses of intrahippocampal tetanus toxin. Here we report a quantitative histological study of hippocampi from rats which had gained remission from seizures induced by low doses of tetanus toxin. Adult Sprague Dawley rats received unilateral injections of 3-4 ng (about 6-8 mouse LD50) tetanus toxin, or vehicle, into the dorsal hippocampus. The first experiment confirmed that postsynaptic evoked responses recorded from pyramidal cells were depressed 10-19 weeks after injection. Unexpectedly, there also was a decrease of 20% in the antidromic response from CA3a contralateral to the injection. However, cell counts in these hippocampi revealed no change in pyramidal cell numbers. The second experiment used rats from two breeding colonies, prepared for histology 7 weeks after injection. Hippocampal pyramidal cell numbers were within the normal range in all but three of the 24 rats that had received tetanus toxin. These three had lesions of the CA1 pyramidal layer contralateral to the injection. The lesions were of the order of 2 mm in diameter, and were associated with glial proliferation. When these three cases were excluded, there remained a small increase in glial density in CA1 of the toxin-injected rats. In addition, toxin-injected rats from one of the colonies were susceptible to a pathology known as acidophylic or dark cell change. These occurred in 11 of 18 toxin-injected rats from this colony, in all divisions of the pyramidal layer, in both the injected and the contralateral hippocampus (where parallel studies revealed independent secondary epileptic foci). We conclude that loss of pyramidal neurons is not necessary for the persistent behavioural changes in this model.(ABSTRACT TRUNCATED AT 400 WORDS)

Cerebral polyamine metabolism in reversible hypoglycemia of rat: relationship to energy metabolites and calcium

Thirty minutes of insulin-induced reversible hypoglycemic coma (defined in terms of cessation of EEG activity) was produced in anesthetized rats. At the end of the hypoglycemic coma or after recovery for 3, 24, or 72 h induced by glucose infusion, the animals were reanesthetized and their brains frozen in situ. Two control groups were used: untreated controls without prior manipulations, and insulin controls, which received injections of insulin followed by glucose infusion to maintain blood glucose within the physiological range. The brains of these latter animals were frozen 3, 24, or 72 h after glucose infusion. Tissue samples from the cortex, striatum, hippocampus, and thalamus were taken to measure ornithine decarboxylase (ODC) activity, and putrescine and spermidine levels, as well as phosphocreatine (PCr), ATP, glucose, and lactate content. In addition, 20-microns thick coronal sections taken from the striatum and dorsal hippocampus were used for histological evaluation of cell damage and also stained for calcium. Insulin in the absence of hypoglycemia produced a significant increase in ODC activity and putrescine level but had no effect on the profiles of energy metabolites or spermidine. During hypoglycemic coma, brain PCr, ATP, glucose, and lactate levels were sharply reduced, as expected. Energy metabolites normalized after 3 h of recovery. In the striatum, significant secondary decreases in PCr and ATP contents and rises in glucose and lactate levels were observed after 24 h of recovery. ODC activity, and putrescine and spermidine levels were unchanged during hypoglycemic coma. After 3 h of recovery, ODC activity increased markedly throughout the brain, except in the striatum. After 24 h of recovery, ODC activity decreased and approached control values 2 days later. Putrescine levels increased significantly throughout the brain after reversible hypoglycemic coma, the highest values observed after 24 h of recovery (p less than or equal to 0.001, compared with controls). After 72 h of recovery, putrescine levels decreased, but still significantly exceeded control values. Reversible hypoglycemic coma did not produce significant changes in regional spermidine levels except in the striatum, where an approximately 30% increase was observed after 3 and 72 h of recovery (p less than or equal to 0.01 and p less than or equal to 0.05, respectively). Twenty-four hours after hypoglycemic coma, intense calcium staining was apparent in layer III of the cerebral cortex, the lateral striatum, and the crest of the dentate gyrus. After 72 h of recovery, the intense calcium staining included also cortical layer II, the septal nuclei, the subiculum, and the hippocampal CA1-subfield.(ABSTRACT TRUNCATED AT 400 WORDS)

Comparative behavioral and pharmacological studies with centrally administered kynurenine and kynurenic acid in rats

In the present study the effects of kynurenine and its metabolite kynurenic acid were compared in different behavioral and pharmacological tests. Kynurenic acid administered i.c.v. resulted in ataxia and stereotypy in a dose-dependent manner (0.025-1.6 mumol). Administration of 0.8 mumol of kynurenic acid resulted in sleeping and an approximate 25% mortality of the animals. At a dose of 1.6 mumol all of the animals died within 2-5 min from cardiorespiratory failure. One hour after lower doses of kynurenic acid the behavior of the rats appeared normal (neither stereotypy nor ataxia were observed in their familiar environemnt), but their exploratory activity (0.025-0.2 mumol) was significantly lower in a novel environment (open-field box) compared to the control group. Twenty four hours after the injection of kynurenic acid the exploratory activity of the animals did not differ from the control group. Kynurenine administered i.c.v. in equimolar doses did not result in stereotypy, ataxia, sleeping or mortality of the animals although, immediately after high doses short-lasting (1-2 min) immobility was observed. The rearing activity of the high dose kynurenine-treated animals was lower 1 h after injection, but this effect disappeared 24 h after the treatment. Post-trial injection of kynurenic acid (0.2 mumol) slightly, but not significantly, inhibited the learning ability of the rats in an active avoidance paradigm. Kynurenine administered in an equimolar dose had no effect on the speed of learning, but significantly attenuated the intertrial activity of the rats. Kynurenic acid (0.2 mumol, 0.4 mumol) did not significantly inhibit the passive avoidance latency of the animals after post-trial treatment.(ABSTRACT TRUNCATED AT 250 WORDS)

Changes in pyruvate dehydrogenase complex activity during and following severe insulin-induced hypoglycemia

The effect of severe insulin-induced hypoglycemia on the activity of the pyruvate dehydrogenase enzyme complex (PDHC) was investigated in homogenates of frozen rat cerebral cortex during burst suppression EEG, after 10, 30, and 60 min of isoelectric EEG, and after 30 and 180 min and 24 h of recovery following 30 min of hypoglycemic coma. Changes in PDHC activity were correlated to levels of labile organic phosphates and glycolytic metabolites. In cortex from control animals, the rate of [1-14C]pyruvate decarboxylation was 7.1 +/- 1.3 U/mg of protein, or 35% of the total PDHC activity. The activity was unchanged during burst suppression EEG whereas the active fraction increased to 81-87% during hypoglycemic coma. Thirty minutes after glucose-induced recovery, the PDHC activity had decreased by 33% compared to control levels, and remained significantly depressed after 3 h of recovery. This decrease in activity was not due to a decrease in the total PDHC activity. At 24 h of recovery, PDHC activity had returned to control levels. We conclude that the activation of PDHC during hypoglycemic coma is probably the result of an increased PDH phosphatase activity following depolarization and calcium influx, and allosteric inhibition of PDH kinase due to increased ADP/ATP ratio. The depression of PDHC activity following hypoglycemic coma is probably due to an increased phosphorylation of the enzyme, as a consequence of an imbalance between PDH phosphatase and kinase activities. Since some reduction of the ATP/ADP ratio persisted and since the lactate/pyruvate ratio had normalized by 3 h of recovery, the depression of PDHC most likely reflects a decrease in PDH phosphatase activity, probably due to a decrease in intramitochondrial Ca2+.

The neurotoxicity of ouabain, a sodium-potassium ATPase inhibitor, in the rat hippocampus

Intrahippocampal injection of 1 nmol ouabain, a sodium/potassium (Na+,K(+)-)ATPase inhibitor, produced a necrotic lesion within 4 days, characterised by a massive invasion by foaming macrophages. A lower dose of ouabain (0.1 nmol) produced a more discrete lesion of all groups of neuronal perikarya in the hippocampus, with only a minimal degree of glial infiltration. The neuronal perikaryal death produced in the subicular, CA1 and CA2 regions was only partially decreased by intraperitoneal injections of the anticonvulsants diazepam and MK-801; these drugs were without effect in the CA3 or hilar interneuronal regions. At neither dose of ouabain was there any indication of neuronal loss in brain regions outside the hippocampus, typically produced by prolonged seizure activity. It is suggested that ouabain has a two-fold action, a release of toxic acidic amino acids and a prolonged depolarization of neurons leading to osmolysis or calcium necrosis.

TCP shortens the latency of onset of isoelectricity in hypoglycaemia and fails to protect striatal neurones and dentate gyrus granule cells from hypoglycaemic injury in rats

Competitive N-methyl-D-aspartate (NMDA) receptor antagonists are known to protect neurones against hypoglycaemic damage. We tested N-[1-(2-thienyl)cyclohexyl]piperidine (TCP), a non-competitive NMDA antagonist, in a recovery model of hypoglycaemic coma in the rat. Administered concomitantly with insulin, TCP shortened the latency of onset of electrocerebral silence, and failed to prevent striatal and dentate gyrus hypoglycaemia-induced injury. This effect is probably related to an increase in glucose consumption of neurones: TCP enhances energy metabolism in several brain structures, which could facilitate, at low blood glucose levels, the onset of isoelectricity, and hamper a putative neuro-protective effect of the drug.

Selective degeneration of cerebellar cortical neurons caused by cycad neurotoxin, L-beta-methylaminoalanine (L-BMAA), in rats

Both the racemate and the L-form of BMAA (beta-methylaminoalanine), when injected intraperitoneally into young rats, produced acute signs of cerebellar dysfunction and degeneration of cerebellar stellate, basket, Purkinje and Golgi cells, but not granule cells. Degenerative changes were also occasionally seen in cerebellar roof nuclei which may be secondary in nature. No other changes were found in the remainder of the central nervous system. The doses of the L-form of BMAA producing these changes were from 6 to 14 mumols/g body weight, i.e. the lower and upper levels of the dose range used by Vega and Bell (1967) and equivalent to 75 and 183 mg/rat. Doses of 1 to 4 mg/g body weight of the racemate were given to young rats less than 100 g in weight, but no changes were apparent after daily doses of the racemate of 0.5 mg/g body weight. Damage to cerebellar neurons is considered to be the result of excitotoxic activity. All cells showing degeneration are GABAergic, although not all are known to possess N-methyl-D-aspartate (NMDA) receptors. The present finding of selective cerebellar neuron damage may not conflict with the earlier findings of others, but our results suggest that L-BMAA has unusual glutamate receptor binding properties.

Extracellular levels of quinolinic acid are moderately increased in rat neostriatum following severe insulin-induced hypoglycaemia

Extracellular concentrations of the brain metabolite quinolinic acid, an endogenous excitotoxin, were monitored by microdialysis in rat neostriatum and hippocampus/cortex during and following a 30-min period of insulin-induced hypoglycaemia. During hypoglycaemia-induced isoelectricity, extracellular levels of quinolinic acid in the striatum (basal value, 1.1 +/- 0.3 pmol per 30-microliters fraction) were elevated 1.7 times as compared to the control period. Thirty to ninety minutes following hypoglycaemia a significant increase in extracellular quinolinic acid to 2.2 times basal level was noted. After 2 h recovery, the beginning of neuronal necrosis was observed in the dorsolateral striatum. Implantation of the dialysis probe did not influence the extent of neuronal damage. No changes in extracellular quinolinic acid levels were observed in the hippocampus/cortex. The data indicate that following a severe hypoglycaemic insult vulnerable striatal cells are exposed to hyperphysiological extracellular quinolinic acid concentrations over an extended period of time. Considering the pronounced susceptibility of rat striatal neurons to the toxin, the small but prolonged elevation in the extracellular levels of quinolinic acid could be of significance for the development of delayed neuronal death in hypoglycaemia.

Brain and plasma quinolinic acid in profound insulin-induced hypoglycemia

Profound insulin-induced hypoglycemia is associated with early-onset neuronal damage that resembles excitotoxic lesions and is attenuated in severity by antagonists of N-methyl-D-aspartate receptors. Hypoglycemia increases L-tryptophan concentrations in brain and could increase the concentration of the L-tryptophan metabolite quinolinic acid (QUIN), an agonist of N-methyl-D-aspartate receptors and an excitotoxin in brain. Therefore, we investigated the effects of 40 min of profound hypoglycemia (isoelectric EEG) and 1-2 h of normoglycemic recovery on the concentrations of QUIN in brain tissue, brain extracellular fluid, and plasma in male Wistar rats. Plasma QUIN increased 6.5-fold by the time of isoelectricity (2 h after insulin administration). Regional brain QUIN concentrations increased two- to threefold during hypoglycemia and increased a further two- to threefold during recovery. However, no change in extracellular fluid QUIN concentrations in hippocampus occurred during hypoglycemia or recovery as measured using in vivo microdialysis. Therefore, the increases in brain tissue QUIN concentrations may reflect elevations of QUIN in the intracellular space or be secondary to the increases in QUIN in the vascular compartment in brain per se. L-Tryptophan concentrations increased more than twofold during recovery only. Serotonin decreased greater than 50% throughout the brain during hypoglycemia, while 5-hydroxyindoleacetic acid concentrations increased more than twofold during hypoglycemia and recovery. In striatum, dopamine was decreased 75% during hypoglycemia but returned to control values during recovery, while striatal 3,4-dihydroxyphenylacetic acid and homovanillic acid were increased more than twofold during both hypoglycemia and recovery.(ABSTRACT TRUNCATED AT 250 WORDS)

Regulation of glutamate and aspartate release from the Schaffer collaterals and other projections of CA3 hippocampal pyramidal cells

Excitatory synaptic transmission in the CNS can be modulated by endogenous substances and metabolic states that alter release of the transmitter, usually glutamate and/or aspartate. To explore this issue, we have studied the release of endogenous glutamate and aspartate from synaptic terminals of the CA3-derived Schaffer collateral, commissural and ipsilateral associational fibers in slices of hippocampal area CA1. These terminals release glutamate and aspartate in about a 5:1 ratio. The release process is modulated by adenosine, by the transmitters themselves and by nerve terminal metabolism. Adenosine inhibits the release of both amino acids by acting upon an A1 receptor. The transmitters, once released, can regulate their further release by acting upon both an NMDA and a non-NMDA (quisqualate/kainate) receptor. Activation of the NMDA receptor enhances the release of both glutamate and aspartate, whereas activation of the non-NMDA receptor depresses the release of aspartate only. Superfusion of CA1 slices with a glucose-deficient medium increases the release of both amino acids and reduces the glutamate/aspartate ratio. These results have implications for the regulation of excitatory synaptic transmission in the CA1 area and for the mechanism of hypoglycemic damage to CA1 pyramidal cells.

Hypoglycemic neurotoxicity in vitro: involvement of excitatory amino acid receptors and attenuation by monosialoganglioside GM1

Rat cerebellar granule cells, when subjected to a glucose-free environment for 4 h, developed extensive degeneration of neuronal cell bodies and their associated neurite network over the following 24 h. This neuronal damage was quantitated with a colorimetric assay using the metabolic dye 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide. Hypoglycemic neuronal injury could be markedly reduced by the presence of both competitive (3-(+/-)-2-carboxypiperazin-4-yl)-propyl-1-phosphonic acid) and non-competitive (phencyclidine) N-methyl-D-aspartate receptor antagonists, but not by kainate/quisqualate preferring antagonists 6-cyano-7-nitroquinoxaline-2,3-dione and 6,7-dinitroquinoxaline-2,3-dione. Glucose deprivation neuronal injury was also reduced by adding glutamate-degrading enzymes to the incubation medium. Monosialoganglioside GM1, but not its asialo derivative (lacking sialic acid), was also effective in protecting against hypoglycemic neurodegeneration when included during the period of glucose deprivation. These results suggest that the neuronal injury to cerebellar granule cells resulting from glucose deprivation is mediated predominantly by activation of the N-methyl-D-aspartate type of excitatory amino acid receptor, perhaps through the action of endogenously released glutamate. Furthermore, the monosialoganglioside GM1, a member of a class of naturally occurring sialoglycosphingolipids, is able to attenuate this neuronal injury--as already observed for glutamate neurotoxicity and anoxic neuronal death in cerebellar granule cells. Gangliosides may thus prove to be of therapeutic utility in excitatory amino acid-associated neuropathologies.

Effects of glucose deficiency on glutamate/aspartate release and excitatory synaptic responses in the hippocampal CA1 area in vitro

The effects of glucose deficiency on (1) the K+-evoked release of glutamate and aspartate and (2) excitatory synaptic transmission were studied in the Schaffer collateral-commissural-ipsilateral associational (SCCIA) projection to area CA1 of the rat hippocampal formation in vitro. Compared with 1 or 10 mM glucose, superfusion of CA1 slices with 0.1 mM glucose enhanced the K+-evoked release of both glutamate and aspartate, increased the ratio of aspartate release to glutamate release and did not affect the release of GABA. With both high and low glucose concentrations, the K+-evoked release of glutamate and aspartate originated predominantly from a Ca2+-sensitive store associated with the SCCIA projection. Superfusion with glucose-deficient medium abolished the inhibitory effect of adenosine on glutamate and aspartate release, but augmented the enhancing effect of the adenosine antagonist 8-phenyltheophylline. These results suggest that enough endogenous adenosine was released from the slices under these conditions to saturate the presynaptic A1 receptors. Despite its facilitatory effect on excitatory transmitter release, glucose-deficient medium inhibited transmission at Schaffer collateral-commissural synapses. Even when the postsynaptic response to a single electrical pulse was abolished, however, a substantial response could still be evoked through paired-pulse or frequency potentiation and the inhibition promptly reversed upon superfusion with 10 mM glucose. The increased ratio of aspartate release to glutamate release appears to reflect changes in the tissue content of these amino acids. The enhanced release of both excitants is suggested to result partly from a rise in intraterminal Ca2+ concentration and partly from inhibition of glutamate/aspartate uptake. Enhanced aspartate release may be particularly relevant to hypoglycemic damage in the CA1 area, because aspartate is a more potent hippocampal excitotoxin than glutamate.

Glucose deprivation neuronal injury in cortical culture

Murine cortical cell cultures deprived of glucose for 6-8 h developed extensive neuronal degeneration, apparent both morphologically and by efflux of lactate dehydrogenase to the bathing medium. This neuronal damage could be substantially reduced by addition of D-2-amino-5-phosphonovalerate (D-APV), in a concentration-dependent (IC50 about 2 microM) and stereospecific (D-APV more potent than L-APV) fashion. A similar neuron-protective effect could also be obtained with several other NMDA antagonists, 2-amino-7-phosphonoheptanoate, phencyclidine, MK-801, ketamine, and (+)-SKF 10,047, as well as with the broad spectrum glutamine antagonist kynurenate. In contrast, little protection could be obtained with gamma-D-glutamylaminomethyl sulfonate and L-glutamate diethyl ester, compounds which have been reported to act primarily at non-NMDA receptors. These observations support the hypothesis that glucose deprivation-induced cortical neuronal injury is largely mediated by NMDA receptors, and suggest that cell culture methodology can be useful in the quantitative characterization of that injury.

N-methyl-D-aspartate antagonists: ready for clinical trial in brain ischemia?

Antagonists of the N-methyl-D-aspartate (NMDA) subclass of glutamate receptors may offer a new approach for the treatment of ischemic brain injury. This strategy is supported by a well-developed scientific foundation and encouraging results in a variety of in vivo and in vitro experimental models. Several specific antagonists, including MK-801, dextrorphan, dextromethorphan, and ketamine, have already been used at low doses in humans for other indications and are potential candidates for Phase I clinical trials.

N-methyl-D-aspartate receptors in the cortex and hippocampus of baboon (Papio anubis and Papio papio)

In vitro autoradiography was used to examine the N-methyl-D-aspartate receptor in the brain of a baboon species, Papio anubis, and compared to that of Papio papio which exhibits a photosensitive epilepsy. The epilepsy originates in the frontal cortex and is accompanied by an enhanced sensitivity to N-methyl-D-aspartate. In both Papio anubis and Papio papio, the density of N-methyl-D-aspartate receptors was greatest in the hippocampus, followed by associational areas including frontal cortex, and low in primary sensory areas such as the visual cortex. The receptors were concentrated in the outer cortical layers I-III, very low in layer IV except in primary visual cortex, and of intermediate density in layer V. The density of binding sites was approximately two-fold lower than previously observed in the rodent brain, whereas the affinity of the receptor for [3H]L-glutamate was greater in the primate versus the rodent brain. Glycine potentiated the binding of [3H]L-glutamate in both cortex and hippocampus. No significant differences in the properties of N-methyl-D-aspartate receptors were observed between the two baboon species, suggesting that the photosensitivity of Papio papio is not due to alterations in the binding of L-glutamate to the N-methyl-D-aspartate receptor complex.

Morphinans attenuate cortical neuronal injury induced by glucose deprivation in vitro

The non-narcotic dextrorotatory morphinan, dextrorphan, as well as its levorotatory opioid enantiomer, levorphanol, and its O-methyl derivative, dextromethorphan, have recently been shown to antagonize N-methyl-D-aspartate receptor-mediated neurotoxicity. Consistent with in vivo data suggesting that this neurotoxicity contributes to the neuronal damage associated with hypoglycemia, micromolar concentrations of these morphinans markedly attenuated the injury of cultured mouse cortical neurons produced by acute glucose deprivation. These observations lend specific support to the possibility that morphinan compounds may prove to have clinical therapeutic utility in hypoglycemic encephalopathy.

Glucose depletion hyperpolarizes guinea pig hippocampal neurons by an increase in potassium conductance

Superfusion of guinea pig hippocampal brain slices with a glucose-free solution induced a membrane hyperpolarization and an increase in input conductance of neurons in the CA3 region. Under voltage clamp, glucose depletion induced an outward current with a reversal potential near the K+ equilibrium potential. The action of glucose depletion was different from the effect of ouabain, indicating that low-glucose-induced changes in the membrane conductance are primarily due to alterations in cell metabolism rather than due only to an inhibition of the Na+/K+ pump.

Kynurenate inhibition of cell excitation decreases stroke size and deficits

Pharmacological inhibition of excitatory neurotransmission attenuates cell death in models of global ischemia/reperfusion and hypoglycemia. The current investigations extend these observations to a model of focal ischemia. Kynurenic acid, a broad-spectrum antagonist at excitatory amino acid receptors, was used as treatment (300 mg/kg; 3 doses at 4-hour intervals) before and after focal cerebral ischemia in rats (n = 54). Preischemia but not 1 hour postischemia treatment with kynurenate attenuated infarction size (p less than 0.001) and improved neurological outcome (p less than 0.001) studied at 24 hours after injury. These data support the role of excitatory neurotransmission in acute neuronal injury and support pharmacological inhibition of cell excitation as a potential therapy for stroke.

Blood flow and metabolism in heterotopic cerebellar grafts during hypoglycemia

Hypoglycemia-induced disturbances of brain metabolism and neuronal injury exhibit a distinct predilection for forebrain structures, in particular the caudate-putamen, hippocampus and cerebral cortex, whereas the cerebellum is remarkably resistant. In an attempt to assess the biological basis of this differential regional vulnerability, we have used a neural transplantation technique to compare hemodynamic and metabolic changes in cerebellum during severe hypoglycemia with those in heterotopic cerebellar grafts. To this end, the cerebellar anlage of fetal rat brain (day 15 of gestation) was stereotactically transplanted into the vulnerable caudate-putamen. Following a differentiation period of 8 weeks the grafts had developed into an organotypic population of mature cells with laminar histoarchitecture. Host animals were then subjected to insulin-induced hypoglycemia. After 15 min of isoelectric EEG, blood flow was increased throughout the brain but residual glucose consumption was significantly higher in cerebellum (0.29 mumol/g per min) and cerebellar grafts (0.22 mumol/g per min) as a result of increased glucose extraction. Hypoglycemia caused a depletion of ATP in all brain structures except cerebellum where normal levels were maintained. Correlation of local ATP content and glucose utilization revealed a threshold-like decline of ATP at a glucose utilization rate of 0.27 mumol/g per min. ATP, in consequence, was normal in cerebellum but partially depleted in cerebellar grafts. It is concluded that the resistance of cerebellum to hypoglycemia is due to its capacity for higher glucose extraction at low blood glucose levels, and that this unique intrinsic property is preserved after heterotopic transplantation.