Division of Pharmacokinetics, Toxicology and Targeting, Department of Pharmacy, University of Groningen, Antonius Deusinglaan 1, 9713 AV Groningen, The Netherlands.
Chem Res Toxicol. 2012 Sep 17;25(9):1938-47. doi: 10.1021/tx300248j. Epub 2012 Aug 16.
Idiosyncratic drug-induced liver injury (IDILI) has been the top reason for withdrawing drugs from the market or for black box warnings. IDILI may arise from the interaction of a drug's reactive metabolite with a mild inflammation that renders the liver more sensitive to injury resulting in increased toxicity (inflammatory stress hypothesis). Aiming to develop a robust ex vivo screening method to study inflammatory stress-related IDILI mechanisms and to find biomarkers that can detect or predict IDILI, mouse precision-cut liver slices (mPCLS) were coincubated for 24 h with IDILI-related drugs and lipopolysaccharide. Lipopolysaccharide exacerbated ketoconazole (15 μM) and clozapine (45 μM) toxicity but not their non-IDILI-related comparators, voriconazole (1500 μM) and olanzapine (45 μM). However, the other IDILI-related drugs tested [diclofenac (200 μM), carbamazepine (400 μM), and troglitazone (30 μM)] did not cause synergistic toxicity with lipopolysaccharide after 24 h of incubation. Lipopolysaccharide further decreased the reduced glutathione levels caused by ketoconazole or clozapine in mPCLS after 24 h of incubation, which was not the case for the other drugs. Lipopolysaccharide significantly increased nitric oxide (NO), cytokine, and chemokine release into the mPCLS media, while the treatment with the drugs alone did not cause any substantial change. All seven drugs drastically reduced lipopolysaccharide-induced NO production. Interestingly, only ketoconazole and clozapine increased the lipopolysaccharide-induced granulocyte colony-stimulating factor (G-CSF) and granulocyte-macrophage colony-stimulating factor (GM-CSF) release. Pilot experiments showed that diclofenac and troglitazone, but not carbamazepine, demonstrated synergistic toxicity with lipopolysaccharide after a longer incubation of 48 h in mPCLS. In conclusion, we have developed an ex vivo model to detect inflammatory stress-related liver toxicity and identified ketoconazole, clozapine, troglitazone, and diclofenac as drugs that showed synergistic toxicity with lipopolysaccharide. Reduced glutathione, G-CSF, and GM-CSF were identified to be potential biomarkers for IDILI-inducing drugs mediated by inflammatory stress, and mPCLS appear to be a promising screening tool to further unravel the mechanism of IDILI.
药物诱导的肝损伤（DILI）是药物撤市或黑框警告的首要原因。DILI 可能源于药物的反应性代谢物与轻度炎症的相互作用，使肝脏对损伤更敏感，从而导致毒性增加（炎症应激假说）。为了开发一种稳健的体外筛选方法来研究炎症应激相关的 DILI 机制，并找到可以检测或预测 DILI 的生物标志物，我们将与 DILI 相关的药物和脂多糖共同孵育 24 小时的小鼠精密切割肝切片（mPCLS）。脂多糖加剧了酮康唑（15 μM）和氯氮平（45 μM）的毒性，但没有加剧其非 DILI 相关对照物伏立康唑（1500 μM）和奥氮平（45 μM）的毒性。然而，测试的其他 DILI 相关药物[双氯芬酸（200 μM）、卡马西平（400 μM）和曲格列酮（30 μM）]在孵育 24 小时后与脂多糖没有协同毒性。脂多糖进一步降低了酮康唑或氯氮平孵育 24 小时后 mPCLS 中还原型谷胱甘肽的水平，但其他药物则没有。脂多糖显著增加了 mPCLS 培养基中一氧化氮（NO）、细胞因子和趋化因子的释放，而单独使用这些药物并没有引起实质性的变化。这七种药物都显著降低了脂多糖诱导的 NO 产生。有趣的是，只有酮康唑和氯氮平增加了脂多糖诱导的粒细胞集落刺激因子（G-CSF）和粒细胞-巨噬细胞集落刺激因子（GM-CSF）的释放。初步实验表明，双氯芬酸和曲格列酮，但不是卡马西平，在 mPCLS 中孵育 48 小时后，与脂多糖表现出协同毒性。总之，我们开发了一种体外模型来检测与炎症应激相关的肝毒性，并确定酮康唑、氯氮平、曲格列酮和双氯芬酸是与脂多糖表现出协同毒性的药物。还原型谷胱甘肽、G-CSF 和 GM-CSF 被鉴定为炎症应激介导的 DILI 诱导药物的潜在生物标志物，mPCLS 似乎是一种很有前途的筛选工具，可以进一步揭示 DILI 的机制。
