Endotoxin-, glutamate- and drug-induced inflammation and cytotoxicity with emphasis on signal transduction mechanisms

Abstract: Inflammatory cells such as monocytes and glial cells play an important role in the pathogenesis of ischemic- or drug-induced brain or liver injury. Excessive concentrations of glutamate is toxic to neural cells, and triggers a series of transcriptional events beginning with the expression of the immediate early genes like c-fos and c-jun, which in turn affect the expression of other genes necessary for the development of brain injury. Glial cells are important mediators in such response as they are responsible for the secretion of proinflammatory cytokines and chemokines which in turn will cause recruitment of immune cells into the brain. Using primary rat glial cell cultures we found that lipopolysaccharide (LPS)- induced c-fos expression was p38 MAPK-dependent and occurred via the activation of the SRE or the CRE elements in the promoter. In contrast to what has been shown in neurons, we found that glutamate-induced c-fos expression in glial cells involves a novel calciumdependent pathway. This mechanism requires the participation of mGluR5, mobilization of ER-calcium and de-repression of DREAM at the DRE element in the c-fos promoter. Similar mechanisms of inflammation as seen in the brain also occur in the liver, where Kupffer cells play a similar role functioning as the hepatic macrophages in their ability to release pro-inflammatory cytokines. Drug-induced hepatotoxicity is a major problem in drug development since preclinical in vitro as well as in vivo animal models usually are of little value for prediction of hepatotoxicity in humans. Pure hepatocyte cultures are generally not a sensitive enough model system to predict drug-induced cytotoxicity. We therefore developed a novel in vitro system containing both monocytes and hepatocytes. Ximelagatran (thrombin inhibitor) as well as the PPAR-gamma agonists troglitazone (hepatotoxic) and rosiglitazone (not hepatotoxic) were used as model compounds. Studies in single cultures of monocytes (THP-1) showed a ximelagatran dependent release of pro-inflammatory chemokines and decreased cell viability, which was shown by inhibitors to involve the JNK- and ERK-signal transduction pathways. A novel human in vitro co-culture model system containing THP-1 and hepatocytes (Huh-7) was established where the cells were separated by a permeable membrane. In such cocultures troglitazone-induced cytotoxicity was more apparent and observed earlier than using single cultures of either Huh-7 or THP-1 cells, whereas rosiglitazone showed no cytotoxicity in either system. The troglitazone effect was accompanied by a much greater expression of genes encoding pro-inflammatory cytokines, chemokines and several other stress-related genes using the co-culture system as compared to single cell type cultures. Conditioned medium from troglitazone-treated THP-1 cells decreased the viability of Huh-7 cells indicate the release of monocyte-derived mediators. It is concluded that such co-culture system might constitute a valuable tool for predictions of drug-induced hepatotoxicity.