Search

Both hypoxia and bilirubin are common risk factors in newborns, which may act synergistically to produce anatomical and functional disturbances of the CNS. Using primary cultures of neurons from the fetal rat brain, it was recently reported that neuronal apoptosis accounts for the deleterious consequences of these two insults. To investigate the influence of hypoxia, bilirubin, or their combination on the outcome of neuronal cells of the immature brain, and delineate cellular mechanisms involved, 6-d-old cultured neurons were submitted to either hypoxia (6 h), unconjugated bilirubin (0.5 microM), or to combined conditions. Within 96 h, cell viability was reduced by 22.7% and 24.5% by hypoxia and bilirubin, respectively, whereas combined treatments decreased vital score by 34%. Nuclear morphology revealed 13.4% of apoptotic cells after hypoxia, 16.2% after bilirubin, and 22.6% after both treatments. Bilirubin action was specifically blocked by the glutamate receptor antagonist MK-801, which was without effect on the consequences of hypoxia. Temporal changes in [(3)H]leucine incorporation rates as well as beneficial effects of cycloheximide reflected a programmed phenomenon dependent upon synthesis of selective proteins. The presence of bilirubin reduced hypoxia-induced alterations of cell energy metabolism, as reflected by 2-D-[(3)H]deoxyglucose incorporation, raising the question of free radical scavenging. Measurements of intracellular radical generation, however, failed to confirm the antioxidant role of bilirubin. Taken together, our data suggest that low levels of bilirubin may enhance hypoxia effects in immature neurons by facilitating glutamate-mediated apoptosis through the activation of N:-methyl-D-aspartate receptors.

Among glutamate receptor subtypes, the N-methyl-D-aspartate (NMDA) receptor plays a key role in brain development and cognitive processes, and mediates excitotoxic injury. To test the hypothesis that prolonged seizures may affect NMDA receptor characteristics in the developing brain, a 30-min episode of generalized seizures was induced in rats at 5, 10, 15 and 25 d of age by i.p. administrations of bicuculline, NMDA receptors were analyzed using specific binding of [3H]-labeled (+)-5-methyl-10,11-dihydro-5H-dibenzo-[a,d]-cycloheptene-5,10-imin e maleate (MK-801) in brain membrane preparations, and allosteric regulation was studied by addition of glutamate (10 microM) and glycine (10 microM). In control pups, total number of binding sites increased between 5 and 25 d, Bmax values varying from 1032 +/- 93 to 2311 +/- 449 fmol/mg protein, whereas receptor affinity decreased with age, the affinity constant (Kd) changing from 20.9 +/- 2.0 to 29.1 +/- 2.0 nM. Activation of NMDA receptors by glutamate and glycine led to age-dependent decreases in Kd values, from 30% at 5 d to 72% at 25 d. Seizures altered receptor density only at 5 d (by 40%). Receptor affinity was increased after seizures at 5, 15 and 25 d (from 12 to 60%). The capacity of receptor activation by glutamate and glycine was significantly reduced by seizures at 5 d. There was no change either in density nor affinity of receptors at 10 d. Therefore, as previously shown for central adenosine and benzodiazepine receptors, sustained seizures are able to alter the characteristics of NMDA receptors in a specific way depending on the maturational stage, suggesting developmental changes in the mechanisms of brain response to seizures.

Benzodiazepines are psychoactive substances classically used for their anticonvulsant properties in neonates as well as in adults. In a previous work, we have shown that seizures lead to an age-dependent upregulation of central benzodiazepine binding sites measured in isolated rat cerebral membranes. However, information concerning regional changes in the receptor density was lacking. In our present study, the effects of bicuculline-induced seizures were investigated by quantitative autoradiography of central benzodiazepine receptors in developing rats and in adults. Animals were killed 30 min after an intraperitoneal injection of either saline or a convulsive dose of bicuculline. Benzodiazepine binding sites in brain sections were labeled by [3H]flunitrazepam. Generalized seizures induced a widespread increase in benzodiazepine receptors, with a marked enhancement in structures that mediate seizure activity, such as substantia nigra, amygdala, septum, and hippocampus. The addition of exogenous gamma-aminobutyric acid to the incubation medium increased benzodiazepine binding by the same order of magnitude whether rats were given saline or bicuculline, suggesting that additional benzodiazepine sites are also functionally linked to gamma-aminobutyric acid receptors. The age-related postictal increase in benzodiazepine receptors might reflect a compensatory response for protection against recurrent seizures, especially in newborns.

Continuous generation of new neurons has been demonstrated in the adult mammalian brain, and this process was shown to be stimulated by various pathologic conditions, including cerebral ischemia. Because brain oxygen deprivation is particularly frequent in neonates and represents the primary event of asphyxia, we analyzed long-term consequences of transient hypoxia in the newborn rat. Within 24 h after birth, animals were exposed to 100% N(2) for 20 min at 36 degrees C, and temporal changes in the vulnerable CA1 hippocampus were monitored. Cell density measurements revealed delayed cell death in the pyramidal cell layer reflecting apoptosis, as shown by characteristic nuclear morphology and expression levels of Bcl-2, Bax, and caspase-3. Neuronal loss was confirmed by reduced density of neuron-specific enolase (NSE)-labeled cells, and peaked by 1 wk post insult, to reach 27% of total cells. A gradual recovery then occurred, and no significant difference in cell density could be detected between controls and hypoxic rats at postnatal d 21. Repeated injections of bromodeoxyuridine (50 mg/kg) showed that newly divided cells expressing neuronal markers increased by 225% in the germinative subventricular zone, and they tended to migrate along the posterior periventricle toward the hippocampus. Therefore, transient hypoxia in the newborn rat triggered apoptosis in the CA1 hippocampus followed by increased neurogenesis and apparent anatomical recovery, suggesting that the developing brain may have a high capacity for self-repair.

In neuronal cultures from the forebrain of 14-d-old rat embryos, transient hypoxia (95% N2/5% CO2, 37 degrees C) for 6 h has been shown to trigger delayed apoptotic death through sequential changes in protein synthesis, whereas preconditioning by a brief episode of hypoxia can rescue neurons. Because hypothermia has been reported to be neuroprotective, the present study was designed to test the influence of reduced temperature on the consequences of lethal hypoxia in our culture model, and cellular mechanisms involved were compared with those underlying preconditioning effects. After 6 d in vitro, cultures were subjected to hypoxia for 6 h. They were either placed at 32 degrees C concomitantly with hypoxia for 6 h or preconditioned the day before by a 1-h episode of hypoxia. The hypoxic insult decreased cell viability by 38% at 96 h after reoxygenation, and 23% of the neurons showed morphologic features of apoptosis. Both hypothermia and preconditioning prevented neuronal death and reduced apoptosis. Preconditioning led to time-dependent changes in leucine incorporation, with persistent overexpression of the survival proteins Bcl-2 and heat-shock protein 70. It also increased thymidine incorporation, in line with induction of the cofactor for DNA polymerase, proliferating cell nuclear antigen. Hypothermia reduced basal apoptosis and necrosis, but did not affect thymidine incorporation, and abolished hypoxia-associated protein synthesis. Therefore, both treatments were protective against neuronal injury consecutive to hypoxia in developing brain neurons in vitro. Whereas preconditioning activated a program that stimulated the expression of anti-apoptotic gene products and regulatory components of the cell cycle, hypothermia did not trigger active processes, but depressed cell activity, which in turn may impair the apoptotic phenomenon.

Brain reperfusion and/or reoxygenation may be of particular importance in the etiology of neuronal damage after hypoxic-ischemic insult in neonates, especially with reference to the generation of free radicals. To investigate this issue, the influence of either standard reoxygenation or transient hyperoxia was studied on the consequences of severe hypoxia in a model of cultured neurons isolated from the fetal rat brain. Culture dishes were exposed for 6 h to hypoxia (95% N2/5% CO2). They were then placed under normoxia (95% air/5% CO2) or hyperoxia (95% O2/5% CO2) for 3 h, and finally returned to normoxia. Control cultures were kept under normoxic conditions. Cell morphology, protein concentrations, lactate dehydrogenase leakage, energy metabolism, as reflected by specific transport and incorporation of 2-D-[3H]deoxyglucose, as well as superoxide radical formation were analyzed as a function of time. Po2 values in the cell incubating medium were decreased by 78% by hypoxia and increased by 221% by hyperoxia. No morphologic alteration could be noticed before 72 h posthypoxia, when cell degeneration became apparent, with a concomitant reduction in protein contents. Hypoxia-reoxygenation induced a transient cellular hypermetabolism, as shown by a 36% increase in 2-D-[3H]deoxyglucose uptake 24 h after hypoxia, and then a 23% decrease below control values at 72 h. It also led to a sharp increase in the formation of superoxide radicals (+108%). Transient hyperoxia during reoxygenation did not exacerbate these events, and thus would not enhance their deterimental effects on cell integrity.

Effects of Hypothermia on Hypoxia-Induced Apoptosis in Neuronal Cells from the Developing Brain: Comparison with Preconditioning

Medium- and Long-term Modifications of Brain NMDA Receptors following Repeated Seizures in Developing Rats

Developmental Changes in Excitotoxic Properties of Glutamate in Cultured Central Neurons † 257

Differential Expression of AP-1-related Gene Products after Hypoxia/Reoxygenation in Cultured Central Neurons † 1859

It has been previously shown that a 6-h hypoxia leads to delayed apoptotic cell death in cultured neurons isolated from fetal rat forebrain. Conversely, exposure to hypoxia for 3h increases the number of living neurons scored by the end of a 96h-period of reoxygenation, when compared to controls maintained in normoxia (Pediatr Res, 1997,41:289A). Such vital changes are associated to selective alterations in the synthesis of specific proteins. In order to test the hypothesis that a preceding period of hypoxia may induce tolerance to the effects of a long-lasting insult, cultured neurons after 6 days in vitro were submitted to hypoxia (95% N2-5% CO2) for 6h (H6 group) or to the same hypoxia preceded the day before by a transient exposure to hypoxia for 1h (P1H6 group). As a function of time following reoxygenation, cell viability was measured by the tetrazolium salt MTT, DNA synthesis was assessed by [3H]thymidine incorporation, and DNA integrity was analyzed using the fluorescent dye DAPI. Whereas changes in the synthesis of total proteins were monitored by [3H]leucine incorporation, the expression profile of specific proteins such as Bcl-2, HSP70 as well as PCNA, a cofactor for DNA polymerase, was analyzed by immunohistochemistry. At 96h post-reoxygenation, cell viability decreased by 36% (n=15) in H6 group, while it was significantly enhanced (14%, n=15) in P1H6 group as compared to normoxic controls. The use of DAPI revealed the presence of apoptotic neurons (23%, n=15) in the H6 group, while P1H6 neurons did not exhibit any sign of apoptosis but showed a significant stimulation of their mitotic activity. In parallel, thymidine incorporation increased by 20% (n=12) in the latter group, and decreased by 12% (n=12) below basal values in H6. After a biphasic increase of total protein synthesis, the final rates of leucine incorporation were significantly reduced at 96h post-reoxygenation (43%, n=12) in H6, whereas they remained in the range of basal values in P1H6. Finally, immunohistochemical analysis showed overexpression of PCNA, Bcl-2 and HSP70 starting from 24h after reoxygenation in P1H6 group, the expression of these proteins being unaffected in H6 neurons compared to normoxic controls. These data suggest that a preceding period of brief hypoxia promotes molecular events which render developing brain neurons more resistant to severe hypoxic stress. Adaptative mechanisms may involve anti-apoptotic gene products as well as related proteins which regulate the cell cycle.

Medium- and Long-term Cerebral Metabolic Consequences of Bicuculline-Induced Recurrent Seizures in Developing Rats † 1862

GLUTAMATE-RELATED CELL ALTERATIONS IN IMMATURE VERSUS MATURE NEURONS IN CULTURE. COMPARISON WITH THE EFFECTS OF HYPOXIA † 1715

We showed previously that a 6h-hypoxia followed by reoxygenation for 96h induces a delayed, mainly apoptotic cell death in cultured neurons from the fetal rat forebrain, whereas shorter hypoxia (3h) increases cell viability. To elucidate cellular events involved, the expression of specific proteins, either effector of apoptosis (Bax) or potentially related to cell recovery(Bcl-2, HSP70, HSP105), was analysed. After 6 days in vitro, neurons were exposed for 3 or 6h to hypoxia (95% N2-5% CO2) that decreased extracellular PO2 by 75-80%. Cells were then reoxygenated (95% air-5% CO2) for 96h. As a function of time, neuronal viability was assessed by the MTT procedure, while apoptosis was studied by the analysis of DNA integrity by gel electrophoresis and nuclear incorporation of the fluorescent dye DAPI. Protein synthesis was measured by incorporation of [3H]leucine, then neuronal proteins were separated by SDS-PAGE. Expression of specific proteins was investigated by immunohistochemistry and Western blot. All data reported are compared to controls kept in normoxia and are statistically significant(p

CHANGES IN RNA AND PROTEIN SYNTHESIS FOLLOWING HYPOXIA/REOXYGENATION IN CULTURED CENTRAL NEURONS. 2225

The formation of oxygen free radicals, especially superoxide and hydroxyl radicals, leads to hypoxic neuronal damage. In an attempt to investigate the infiuence of post-hypoxia reoxygenation in a model of primary culture of neurons from fetal rat brain, the production of superoxide radicals (SR) was assessed over 24h in the culture medium. After 8 days in vitro, culture medium was replaced by 1ml Krebs-Ringer buffer with acetyl-cytochrome c (ACc). A first set of cultures (n = 32) was incubated for 6h in a gas mixture of 95%air-5%CO, (normoxia), and a second set (n = 32) was incubated with 95%N2-5%CO2 (hypoxia) which reduced the PO2 by 77% in the culture medium. Half of the two sets of culture dishes was then placed in normoxic atmosphere, and the remaining dishes were incubated for 3h with 95% O2-5% CO2 (hyperoxia) which increased the PO2 by 250%. All cultures were finally returned to normoxia for 15h. The rate of SR formation was quantified spectrophotometrically by measuring the reduction of ACc at 550 nm. In control cultures (normoxia), about 4% and 8% of ACc per 100 μg protein were reduced after 6h and 24h, respectively, indicating a low basal production of SR under standard culture conditions. Hypoxia induced a significant increase in the reduction of ACc (18.7±3.1 %) which was even higher 3h after return to normoxia (44.3 ± 3.5%). Hyperoxia did not enhance the SR production over basal values and neither aggravaied hypoxia-induced production (20.3±5.55% vs 18.7±3.1%). The following day, while there was no apparent cell alteration, hypoxia-induced level of SR remained stable (40.8 ± 1.6%) as well as the SR accumulation induced by the sequence hypoxia-hyperoxia (18.6±3.9%). Thus, it appears that hypoxia induces an over-production of SR in cultured neurons which may contribute to cell damage. Unexpectedly, post-hypoxia hyperoxygenation lends to limit the SR formation in vitro.

In adult animals, S induce important alterations in CBF, which have not been evaluated in the neonatal period. Thus, we measured CBF before, during and 15 min after S induced by IV bicuculline (.6 to 1.2 mg/kg) with 15 μm radioactive microspheres, in 12 urethan-anesthetized spontaneously breathing piglets (2 to 9 days). Six piglets received IP PB 10 mg/kg, 6 hrs before S, and 6 did not (controls (C)). Mean arterial blood pressure (MABP), PaO2 and ce rebral tissue pH (CtpH) were measured. The duration of S was significantly decreased by PB (mean ± SEM) (10.8 ± 1.8 vs 21.8 ± 4.1 min, p < .05). S induced a significant increase in MABP in both C (83.7 ± 6.3 vs 59.7 ± 7.2 mmHg, p < .01) and PB group (92.0 ± 8.7 vs 68.0 ± 2.4 mmHg, p < .01). PaO2, PaCO2 and CtpH were not significantly modified. CBF was increased during S, mostly in basal gan glia (98 and 106 % respectively for C and PB group), less in brain stem (96 and 45%) and cerebellum (72 and 35%), least in cortex (13 and 30 %) and was unchanged in the choroid (.4 and -.8%). Fifteen min after S, CBF returned to pre S values. There was no significant difference in CBF at any time between C and PB group. CBF changes during S were independent of changes in PaCO2 and PaO2, but were positively correlated with changes in MABP (r=.753, p < .01) and CtpH (r=.570, p