Search

The effects of arachidonic acid on glutamate and gamma-aminobutyric acid (GABA) uptake were studied in primary cultures of astrocytes and neurons prepared from rat cerebral cortex. The uptake rates of glutamate and GABA in astrocytic cultures were 10.4 nmol/mg protein/min and 0.125 nmol/mg protein/min, respectively. The uptake rates of glutamate and GABA in neuronal cultures were 3.37 nmol/mg protein/min and 1.53 nmol/mg protein/min. Arachidonic acid inhibited glutamate uptake in both astrocytes and neurons. The inhibitory effect was observed within 10 min of incubation with arachidonic acid and reached approximately 80% within 120 min in both types of culture. The arachidonic acid effect was not only time-dependent, but also dose-related. Arachidonic acid, at concentrations of 0.015 and 0.03 mumol/mg protein, significantly inhibited glutamate uptake in neurons, whereas 20 times higher concentrations were required for astrocytes. The effects of arachidonic acid were not as deleterious on GABA uptake as on glutamate uptake in both astrocytes and neurons. In astrocytes, GABA uptake was not affected by any of the doses of arachidonic acid studied (0.015-0.6 mumol/mg protein). In neuronal cultures, GABA uptake was inhibited, but not to the same degree observed with glutamate uptake. Lower doses of arachidonic acid (0.03 and 0.015 mumol/mg protein) did not affect neuronal GABA uptake. Other polyunsaturated fatty acids, such as docosahexaenoic acid, affected amino acid uptake in a manner similar to arachidonic acid in both astrocytes and neurons. However, saturated fatty acids, such as palmitic acid, exerted no such effect. The significance of the arachidonic acid-induced inhibition of neurotransmitter uptake in cultured brain cells in various pathological states is discussed.

The glycogen content of primary cultured astrocytes was approximately doubled by incubation with 1 mM L-glutamate or L-aspartate. Other amino acids and excitatory neurotransmitters were without effect. The increase in glycogen level was not blocked by the glutamate receptor antagonist kynurenic acid but was completely blocked by the glutamate uptake inhibitor threo-3-hydroxy-D,L-aspartate and by removal of Na+ from the medium. Incubation with radiolabeled glucose and glutamate revealed that the increased glycogen content was derived almost entirely from glucose. Glutamate at 1 mM was also found to cause a 53 +/- 12% decrease in glucose utilization and a 112 +/- 69% increase in glucose-6-phosphate levels. These results suggest that the glycogen content of astrocytes is linked to the rate of glucose utilization and that glucose utilization can, in turn, be affected by the availability of alternative metabolic substrates. These relationships suggest a mechanism by which brain glycogen accumulation occurs during decreased neuronal activity.

The effects were studied of short-term (1 week) versus long-term (2-3 weeks) fluoxetine treatment of primary cultures of mouse astrocytes, differentiated by treatment with dibutyryl cyclic AMP. From previous experiments it is known that acute treatment with fluoxetine stimulates glycogenolysis and increases free cytosolic Ca2+ concentration ([Ca2+]i]) in these cultures, whereas short-term (one week) treatment with 10 microM down-regulates the effects on glycogen and [Ca2+]i, when fluoxetine administration is renewed (or when serotonin is administered). Moreover, antagonist studies have shown that these responses are evoked by activation of a 5-HT2, receptor that is different from the 5-HT2A receptor and therefore at that time tentatively were interpreted as being exerted on 5-HT2C receptors. In the present study the cultures were found by RT-PCR to express mRNA for 5-HT2A and 5-HT2B receptors, but not for the 5-HT2C receptor, identifying the 5-HT2 receptor activated by fluoxetine as the 5-HT2B receptor, the most recently cloned 5-Ht2 receptor and a 5-HT receptor known to be more abundant in human, than in rodent, brain. Both short-term and long-term treatment with fluoxetine increased the specific binding of [3H]mesulergine, a ligand for alL three 5-HT2 receptors. Long-term treatment with fluoxetine caused an agonist-induced up-regulation of the glycogenolytic response to renewed administration of fluoxetine, whereas short-term treatment abolished the fluoxetine-induced hydrolysis of glycogen. Thus, during a treatment period similar to that required for fluoxetine's clinical response to occur, 5-HT2B-mediated effects are initially down-regulated and subsequently up-regulated.

Hypoxia caused injury and metabolic dysfunction of astrocytes, as indicated by a time-dependent loss of lactate dehydrogenase (LDH) activity and ATP content. The combination of 3.5 mM fructose-1,6-bisphosphate (FBP) and 7.5 mM glucose (GLC) reduced the decrease of ATP and prevented the loss of LDH. These data indicate that the combination of GLC + FBP protects astrocytes from hypoxia. The results also suggest that the maintainance of ATP concentration is the mechanism by which FBP prevents hypoxic injury.

This study demonstrates that virtually homogenous cultures of mouse cerebral neurons, obtained from 15-day-old embryos, differentiate at least as well as cultures which in addition contain astrocytes. This was indicated by glutamate decarboxylase activity which within 2 weeks rose from a negligible value to twice the level in the adult mouse cerebral cortex, and by a gamma-aminobutyric acid (GABA) uptake rate which quadrupled during the second week in culture and reached higher values than in brain slices. Within the same period, the GABA content increased four to five times to 75 nmol/mg protein, and a potassium-induced increase in [14C]GABA efflux became apparent. Although the development was faster than in vivo, optimum differentiation required maintenance of the cultures beyond the age of 1 week. Uptake and release rates for glutamate and glutamine underwent much less developmental alteration. At no time was there any potassium-induced release of radioactivity after exposure to [14C]glutamate, and the glutamate uptake was only slightly increased during the period of GABAergic development. This indicates that exogenous glutamate is not an important GABA precursor. Similarly, glutamine uptake was unaltered between days 7 and 14, although a small potassium-induced release of radioactivity after loading with glutamine suggests a partial conversion to GABA.

The metabolic fate of L-[U-14C]- and L-[1-14C]glutamate was studied in primary cultures of mouse astrocytes. Conversion of the uniformly labeled compound to glutamine and aspartate was followed by determination of specific activities after dansylation with [3H]dansyl chloride and subsequent thin layer chromatography of the dansylated amino acids. Metabolic fluxes were calculated from the alterations of specific activities and the pool sizes, which were likewise measured by a dansylation method. Formation of 14CO2 from [1-14C]glutamate was determined by the trapping of CO2 in hyamine hydroxide in a gas-tight chamber, which is, in the known absence of glutamate decarboxylase activity in the cultured astrocytes, an unequivocal expression of the metabolic flux via alpha-ketoglutarate to CO2 and succinyl-CoA. The metabolic fluxes determined by these procedures amounted to 2.4 nmol/min/mg protein for glutamine synthesis, 1.1 nmol/min/mg protein for aspartate production, and 4.1 nmol/min/mg protein for formation and subsequent decarboxylation of alpha-ketoglutarate. The latter process was unaffected by virtually complete inhibition of glutamate-oxaloacetic transaminase with aminooxyacetic acid, indicating that the formation of alpha-ketoglutarate occurs as an oxidative deamination rather than as a transamination. This suggests that the formation of alpha-ketoglutarate from glutamate represents a net degradation, not an isotopic exchange.

Uptake kinetics for glutamine were studied in several different neuronal preparations (perikarya prepared by gradient centrifugation, cultured cortical neurons, cultured, presumably glutamatergic cerebellar neurons, and brain prisms). In no case were any indications found of a high affinity uptake but a rather efficient low affinity uptake did occur. A similar, equally efficient low affinity uptake is, however, found in astrocytes. Thus, no preferential glutamine uptake occurs into neurons. It is, therefore, not likely that a net flow of glutamine takes place from astrocytes to neurons, compensating for the loss of TCA constituents when glutamate and GABA are released.

The effects of arachidonic acid and other polyunsaturated fatty acids (PUFAs) on both oxidative and metabolic perturbation were studied in primary cultures of rat cerebral cortical astrocytes. In the presence of 0.1 mM arachidonic acid, the rate of the reduction of nitroblue tetrazolium (NBT) to nitroblue formazan (NBF) was stimulated from 0.65 +/- 0.10 to 1.43 +/- 0.15 and from 0.092 +/- 0.006 to 0.162 +/- 0.009 nmol/min/mg protein in intact and broken cell preparations, respectively. The rate of superoxide radical formation, as measured by the superoxide dismutase (SOD)-inhibitable NBT reduction was 0.042 nmol/mg protein in broken cells and was negligible in intact cells. The latter is due to the impermeability of SOD into the intact cell preparation. NBF formation in intact astrocytes stimulated by arachidonic acid was both time- and dose-dependent. Other PUFAs, including linoleic acid, linolenic acid, and docosahexaenoic acid, were also effective in stimulating NBF formation in astrocytes, whereas saturated palmitic acid and monounsaturated oleic acid were ineffective. Similar effects of these PUFAs were observed in malondialdehyde formation in cells and lactic acid accumulation in incubation medium. These data indicate that both membrane integrity and cellular metabolism were perturbed by arachidonic acid and by other PUFAs. The sites of superoxide radical formation appeared to be intracellular and may be associated with membrane phospholipid domains, because liposome-entrapped SOD, which was taken up by intact astrocytes, reduced the level of superoxide radicals and lactic acid content, whereas free SOD was not effective.

The activity of the pyruvate carboxylase was determined in brains of newborn and adult mice as well as primary cultures of astrocytes, of cerebral cortex neurons, and of cerebellar granule cells. The activity was found to be 0.25 +/- 0.14, 1.24 +/- 0.07, and 1.75 +/- 0.13 nmol X min -1 X mg -1 protein in, respectively, neonatal brain, adult brain, and astrocytes. Neither of the two types of neurons showed any detectable enzyme activity (i.e., less than 0.05 nmol X min -1 X mg -1). It is therefore concluded that pyruvate carboxylase is an astrocytic enzyme.

To determine the effects of glucose and fructose-1,6-bisphosphate (FDP) on hypoxic cell damage, primary cultures of astrocytes were incubated for 18 h in an air-tight chamber that had been flushed with 95% N2/5% CO2for 15 min before it was sealed. Cultures containing 7.5 m M glucose without FDP or FDP without glucose showed evidence of significant cell injury after 18 h of hypoxia (increased lactate dehydrogenase content in the culture medium; cell edema and disruption by phase-contrast microscopy). Cultures exposed to glucose + FDP had normal lactate dehydrogenase concentrations and appeared normal microscopically. Maximal protection of hypoxic cells occurred at 6.0 m M FDP. Lactate concentrations of the culture medium of hypoxic cells increased 2.5 times above normoxic control values when glucose was present, but neither FDP alone nor glucose + FDP caused the lactate concentrations to increase further. This implies that anaerobic glycolysis was not increased by adding FDP to the medium. Cell volumes (water space) measured with [14C]-3-0-methyl-D-glucose were normal with glucose + FDP in the culture medium of hypoxic cells but were significantly larger than normal when glucose alone was present. Increases in cell volume paralleled changes in lactate dehydrogenase in the culture medium. Uptake of [14C]FDP occurred rapidly in normoxic cells and was maximal after 5 min of incubation. The data indicate that the presence of glucose + FDP in the culture medium protects primary cultures of hypoxic astrocytes from cell damage.

The metabolic fate of L-[14C]-glutamine was followed in cerebral cortical neurons in primary cultures, a GABAergic preparation. Part of the glutamine was converted to GABA (0.3 nmol/min per mg protein), which is consistent with the presence of glutaminase and glutamate decarboxylase activity in the cells and with findings by other authors in vivo or in brain slices. However, an even larger part (1.8 nmol/min per mg protein) was converted to CO2 and succinate via an oxidative deamination to alpha-ketoglutarate. This is not consistent with the concept that transfer of glutamine from astrocytes to neurons should replenish neuronal GABA stores quantitatively after release of GABA and its partial accumulation into astrocytes, but it is well compatible with the recent demonstration of a net glutamine uptake by the brain.

Compounds known to affect glycogen metabolism in vivo or in cell-free preparations were used to investigate the regulation of glycogen content in intact astrocytes cultured from newborn rat cortex. Compounds were added with fresh medium to culture dishes, and astrocyte glucose and glycogen content determined 24 h later. Increasing the medium glucose concentration from 7.5 mM to 30 mM increased cell glycogen content 80%. Addition of 2-deoxyglucose or 3-O-methyl glucose (2.5-10 mM) also increased cell glycogen content, 50-100%, suggesting a regulatory rather than mass action effect of glucose on astrocyte glycogen content. The phosphorylase b inhibitors 2,2',4,4',5,5'-hexabromobiphenyl and riboflavin had no effect on astrocyte glycogen content, consistent with negligible phosphorylase b activity in normal astrocytes. Phenobarbital and L-methionine-DL-sulfoximine (MSO) are both known to induce astrocyte glycogen accumulation in vivo. The addition of phenobarbital (2 mM) had no effect on the glycogen content of cultured astrocytes, suggesting an indirect mechanism for the in vivo effect. MSO at 1 mM, however, induced a 300% increase in glycogen content. The time course of glucose and glycogen content after MSO administration suggests this increase to be the result of slowed glycogenolysis rather than accelerated glycogen synthesis.

Rates of glutamine formation and of carbon dioxide production (as an indication of oxidative deamination of glutamate) were determined in primary cultures of astrocytes exposed to 50 microM labeled glutamate in the absence or presence of added ammonia (0.1-3 mM). Glutamine formation (1.7 nmol/min/mg protein) was unaffected by all concentrations of added ammonia. This probably reflects the presence of a low content of ammonia (0.1-0.2 mM), originating from degradation of glutamine, in the cells even in the absence of added ammonia, and it shows that pathophysiological concentrations of ammonia do not increase the formation of glutamine from exogenous glutamate. The carbon dioxide production rate was 5.9 nmol/min/mg protein, i.e., three to four times higher than the rate of glutamine formation. It was significantly reduced (to 3.5 nmol/min/mg protein) in the presence of 1 mM or more of ammonia. This is in keeping with suggestions by others that toxic levels of ammonia affect oxidative metabolism.

Uptake kinetics for glutamate, GABA, and glutamine were determined in primary cultures of cerebral neurons, a predominatly GABA-ergic cell population, and of cerebellar granule cells, a predominantly glutamatergic cell population. A specially high Vmax for GABA uptake into the former and for glutamate uptake into the latter cells suggests that considerable amounts of released transmitters may be reaccumulated into appropriate nerve terminals. Nervertheless, the glutamate uptake into the cerebellar granule cells was less intense than that previously observed into corresponding cultures of astrocytes, whereas GABA was accumulated more intensely into neurons than into astrocytes. This suggests that especially glutamatergic neurons may be depleted for their transmitter by accumulation into adjacent astrocytes. If a glutamine flow from astrocytes back to neurons served the purpose of balancing this transfer, it should be expected that glutamine accumulation was more intense in the glutamatergic than in the GABA-ergic cell population. This was not the case, suggesting that such a glutamate–glutamine cycle may not be operating to a major extent.