Development And Validation Of An In Vitro Hepatotoxicity Prediction Model For Dietary Phenolics

Dietary phenolics in fruits and vegetables have been recommended as health-promoting components in daily diet,due to their beneficial bioacitivities such as anti-inflammatory and antioxidant effects.However,there is increasing evidence for hepatotoxic potential of phenolic compounds in recent researches.Food or drug-induced liver injury is highly associated with the initiation and development of various liver diseases.Thus,there is a strong demand for development of rapid and effective hepatotoxicity prediction methods to conduct a non-targeted detection of potential liver toxicants.To develop a hepatotoxicity prediction model,16 phenolic compounds were applied in rat epithelium cell line Clone-9 and rat hepatoma cell line MH1C1.Five hepatotoxic endpoints consisiting of ROS production,mitochondrial membrane potential,cellular DNA content,cellular ATP content and LDH leakgage were assayed,and an overall assessment for hepatotoxicity of each compound was obtained and a statistical hepatotoxicity prediction model was developed in each cell line using principal component analysis(PCA).According to hepatotoxicity prediction model,tested phenolics were classified into non/little-hepatotoxicity,middle hepatotoxic potential or high hepatotoxic groups,respectively.Moreover,the hepatotoxicity prediction model was validated in primary rat hepatocytes.According to hepatotoxic prediction model developed in Clone-9 and MH1C1 cell line,14 and 13 out of the 16 phenolic compounds got consistent hepatotoxic conclusion wih that in primary rat hepatocytes,indicating the effectiveness of hepatocyte cell line as an alternative to primary hepatocytes.In addition,a second set of 6 phenolic compounds and 6 botanical extractions rich in phenolics were chosen to further evaluate the established prediction model.Most results obtained in present study were in agreement with literatures,indicating the effectiveness of developed model in hepatotoxicity predction of phenolic compounds and botanical extractions rich in phenolics.In conlusion,the hepatotoxicity prediction model based on PCA developed in rat hepatocyte cell line is a rapid,convenient and cost-saving method,which is expected to effectively predict hepatotoxic potential of phenolics.The results from the present study provided information for development of hepatotoxic prediction models on the basis of in vitro hepatocyte cell line and statistical tools,contributing to the detection of non-targeted toxicants in early stage and prevention of liver damage induced by chemicals in food processing systems.