In vitro models comprising various cellular elements of brain have been established in order to identify the mechanisms involved in the observed in vivo tolerance to ischemia of the brain pretreated either by tumor necrosis factor-alpha (TNF-alpha) or by hypoxia/ ischemia. TNF-alpha pretreatment of rat cortical astrocytes and brain microvascular endothelial cells (BMEC) renders these cells unresponsive to TNF activation 24 hours later in terms of up-regulation and expression of ICAM-1 (analyzed by FACS fluorescent cell ELISA as well as by Northern blot analysis of ICAM-1 mRNA). The role of ceramide, a lipid messenger implicated in TNF signaling, has been examined in this model. The ceramide analogue, C-2 ceramide, when added to the cells for 4 hours to substitute for TNF pretreatment, failed to induce tolerance to subsequent TNF activation in both cortical astrocytes and BMEC. However, addition of C-2 ceramide 30 min before or even 1 hour after TNF treatment resulted in inhibition of ICAM-1 up-regulation of both protein and mRNA levels. HPLC determination of endogenous ceramide concentrations in TNF-activated cortical astrocytes and BMEC demonstrated two peaks of ceramide release: an early 1.7-fold increase at 15-20 minutes, and a late 4-fold increase at 18-21 hours after TNF addition. These results strongly suggest that late ceramide release is a mediator of TNF-induced tolerance. Primary cultures of neonatal cortical neurons were subjected to either 60 minutes of hypoxia or 60 minutes hypoxia plus glucose deprivation, and reoxygenated (with restoration of glucose concentrations) for various times. Cell death was quantitated using the ethidium iodide exclusion fluorescence method. The number of dead cells gradually increased over 8 hours of reoxygenation reaching 21.9% and 32.1% at 8 hours of hypoxia and hypoxia/glucose deprivation, respectively. Although glucose deprivation exacerbated the effect of hypoxia, glucose deprivation alone for 60 minutes did not cause neuronal death: after 8 hours the number of dead cells was 8.6% versus 8.0% in control cultures. Pretreatment of neuronal cultures with short hypoxia (15 minutes) 24 hours prior to 60 minutes of hypoxia, protected neurons completely against hypoxia (number of dead cells was 9.4%) and partially against hypoxia/glucose deprivation (number of dead cells was 20.5% versus 3% in non-pretreated cultures).
