The peroxisome proliferator-activated receptor agonist troglitazone (TRO) was used for treatment of non-insulin-dependent diabetes until its removal from the market because of its severe hepatotoxicity. However, the mechanism for its hepatotoxicity is still poorly understood. In this study, we investigated whether TRO caused cell death by altering signaling pathways associated with cell damage and survival in human hepatoma cells. Our data reveal that TRO caused time-and concentrationdependent apoptosis of HepG2 and Chang liver human hepatoma cells, as evidenced by DNA fragmentation and staining with Hoechst 33342. In contrast, 50 or 100 M rosiglitazone, a structural analog of TRO, did not cause apoptosis in these hepatoma cells. TRO activated both c-Jun N-terminal protein kinase (JNK) and p38 kinase about 5-fold between 0.5 and 8 h before they returned to control levels at 16 h in HepG2 cells. In contrast, TRO failed to activate the extracellular signal-regulated kinase. Furthermore, TRO increased the levels of proapoptotic proteins, Bad, Bax, release of cytochrome c, and cleavage of Bid in a time-dependent manner. The antiapoptotic Bcl-2 protein level decreased in hepatoma cells treated with TRO. Pretreatment of hepatoma cells with a selective JNK inhibitor, anthra[1,9-cd]pyrazol-6(2H)-one (SP600125), significantly reduced the rate of TRO-induced cell death, whereas 4-(4-fluorophenyl)-2-(4-methylsulfinylphenyl)-5-(4-pyridyl)-1H-imidazole (SB203580), an inhibitor of p38 kinase, had little effect on apoptosis. Pretreatment with SP600125 also prevented JNK activation and c-Jun phosphorylation. In addition, rosiglitazone, which is not as toxic to hepatoma cells as TRO, did not stimulate JNK activity. Transfection of cDNA for the dominant-negative mutant JNK-KR (Lys3 Arg) or SEK1-KR (Lys3 Arg), an immediate upstream kinase of JNK, significantly reduced TRO-induced JNK activation and cell death rate. Furthermore, SP600125 pretreatment effectively prevented the TRO-mediated changes in Bad, Bax, Bid cleavage, and cytochrome c release. These data strongly suggest that hepatotoxic TRO causes apoptosis by activating the JNK-dependent cell death pathway accompanied by increased Bid cleavage and elevation of proapoptotic proteins.

The by-product of lipid peroxidation, 4-hydroxynonenal (HNE), was shown to cause apoptosis in PC12 cells. In this study, we investigated the molecular mechanism of HNE-induced apoptosis in these cells. Specifically, we determined the effect of HNE on the activities of mitogen-activated protein (MAP) kinases involved in early signal transduction. Within 15 to 30 min after HNE treatment, c-Jun N-terminal protein kinase (JNK) was maximally activated, before it returned to control level at 1 h post-treatment. In contrast, activities of extracellular signal-regulated kinase and p38 MAP kinase remained unchanged from their baseline levels. Stress-activated protein kinase kinase (SEK1), an upstream kinase of JNK, was also activated within 5 min after HNE treatment and remained activated for up to 60 min. Marked activation of the JNK pathway through SEK1 and apoptosis signal-regulating kinase 1 (ASK1), an upstream kinase of SEK1, was demonstrated by the transient transfection of cDNA for wild-type SEK1 or ASK1 together with JNK into COS-7 cells. Furthermore, significant reductions in JNK activation and HNE-induced cell death were observed when either of the dominant negative mutant of SEK1 or ASK1 was cotransfected with JNK. Pretreatment of PC12 cells with a survival-promoting agent, 8-(4-chlorophenylthio)-cAMP, prevented both the HNE-induced JNK activation and apoptosis. Nonaldehyde, a nontoxic aldehyde, neither caused apoptosis nor JNK activation. Pretreatment of PC12 cells with SB203580, a specific inhibitor of p38 MAP kinase, had no effect on HNE-induced apoptosis. All these data suggest that the selective JNK activation by HNE is critical for the apoptosis of PC12 cells and that the HNE-mediated apoptosis is likely to be mediated through the activation of the ASK1-SEK1-JNK pathway without activation of extracellular signal-regulated kinase or p38 MAP kinase.

Background: Drug candidates often cause an unwanted blockage of the potassium ion channel of the human ether-ago go related gene (hERG). The blockage leads to long QT syndrome (LQTS), which is a severe life-threatening cardiac side effect. Therefore, a virtual screening method to predict drug-induced hERG-related cardiotoxicity could facilitate drug discovery by filtering out toxic drug candidates. Result: In this study, we generated a reliable hERG-related cardiotoxicity dataset composed of 2130 compounds, which were carried out under constant conditions. Based on our dataset, we developed a computational hERG-related cardiotoxicity prediction model. The neural network model achieved an area under the receiver operating characteristic curve (AUC) of 0.764, with an accuracy of 90.1%, a Matthews correlation coefficient (MCC) of 0.368, a sensitivity of 0.321, and a specificity of 0.967, when tenfold cross-validation was performed. The model was further evaluated using ten drug compounds tested on guinea pigs and showed an accuracy of 80.0%, an MCC of 0.655, a sensitivity of 0.600, and a specificity of 1.000, which were better than the performances of existing hERG-toxicity prediction models. Conclusion: The neural network model can predict hERG-related cardiotoxicity of chemical compounds with a high accuracy. Therefore, the model can be applied to virtual high-throughput screening for drug candidates that do not cause cardiotoxicity. The prediction tool is available as a web-tool at http://ssbio.cau.ac.kr/CardPred.