ROS/ATP Ratio As Endpoint For Predicting Drug-induced Liver Injury

BackgroundDrug induced liver injury?DILI?is increasingly becoming a serious problem affecting public health and drug development.Preclinical drug safety assessments are designed to minimize potential risks to human health and reduce drug development costs.However,Because of the difference in species,conventional preclinical animal safety evaluation experiments have a poor correlation with human DILI.The development of in vitro hepatotoxicity prediction system based on human-derived cells can help solve the problem of species differentiation caused by conventional animal experiments.In addition,the in vitro test cycle is short,and high-throughput detection can be available,and the mechanism of drug induced hepatotoxicity can be explored.However,any single detection endpoint has proven to be incapable of accurately predicting human DILI.Therefore,cell-based in vitro multi-parameters combination prediction of human DILI has become a research trend for hepatotoxicity screening at drug development early stage.Research indicates that the ROS/ATP ratio as endpoint for detection drug hepatotoxicity.The results exhibited high sensitivity and specificity.Compared with the traditional single endpoint evaluation of drug toxicity,this composite endpoint took into account the multi-mechanistic toxicity of the drug,and the accuracy of the prediction has been improved.Compared with the existing multi-endpoint predictive system,the ratio-type endpoint simplifies the experimental procedure and reduces the research and development costs,and helps standardize the prediction of hepatic toxicity in vitro,which is of great significance for the large-scale screening of candidates in early development of new drugs.Therefore,this composite ratio-type endpoint provides a new idea for the standardized prediction of liver toxicity in vitro,and it is necessary and worth studying.ObjectiveThe purpose of this study is verification of the sensitivity and specificity of endpoint ROS/ATP ratio recognition for hepatotoxicity in the most commonly used HepG2 cells and human induced pluripotent stem cells?hiPSCs?derived hepatocytes,thus further confirmed the feasibility of ROS/ATP as an endpoint for liver toxicity testing in vitro.Methods and ResultsThis study is divided into four parts.Part I:Effect of 8 drugs on the viability of HepG2 cellsIn this part of the study,HepG2 cells were used as a model,8 drugs with different degrees of hepatotoxicity in clinic are used for detection.100-fold the therapeutic maximum plasma concentration(C_max)was used as the maximum detection concentration of each drug.The remaining concentrations are 50×C_max,25×C_max,12.5×C_max,and 6.25×C_max.The cells were treated with each drug for 24 h.Then the survival rate of the HepG2 cells was detected by a cell counting kit-8?CCK-8?to determine the concentration of the candidates for subsequent multi-parametric detection experiments.The results showed that for severely hepatotoxic drugs,such as amiodarone hydrochloride,benzbromarone,sodium valproate,and terbinafine hydrochloride,toxic effects were observed in the concentration range of 100×C_max,and the IC50/C_max ratio was less than 100.For non-hepatotoxic drug dexamethasone,there was no significant change in the concentration range of 100×C_max compared with the control group,and the IC₅₀/C_max value was far greater than 100;the prediction performance of the cell viability test results of the three moderately hepatotoxic drugs was poor.The toxicity of chlorpromazine hydrochloride was obvious,while cimetidine and captopril did not show obvious toxicity.The feasibility of using 100×C_max as the maximum concentration of candidates for subsequent multi-parameter detection was initially determined.Part II:Establishment of a multi-parameter hepatic toxicity test system based on HepG2 cellsIn this part of the experiment,HepG2 cells were used as a model to test the effect of the positive drug amiodarone hydrochloride on multiple endpoints in HepG2 cells.Amiodarone hydrochloride was set at 5 concentrations according to the method of Part I.After the cells were treated with the drug for 24 h,the content of intracellular ROS,ATP,caspase 3/7 and reduced glutathione?GSH?were detected with a fluorescence microplate reader.ROS values were corrected for cell viability.The results showed that the accuracy of ROS detection after cell viability correction was significantly improved,and the trend of change was closer to the test results by flow cytometry,and the detection sensitivity was higher than that by flow cytometry.All four endpoints have changed varying degrees in a dose-dependent manner.The detection conditions and the best detection methods for each parameter were determined and established standards for subsequent use of the detection system for in vitro hepatotoxicity screening.Part III:Multi-parameter detection system based on HepG2 cells for hepatotoxicity screeningIn this part of the study,HepG2 cells were used as a model,and select 27 drugs with clinically varying degrees of hepatotoxicity?severely hepatotoxicity,moderate hepatotoxicity,and non-hepatotoxicity?.The effects of 27 drugs on ROS,ATP,caspase3/7?abbreviated as caspase?,and GSH levels in HepG2 cells were detected based on the detection conditions and procedures in the part II.ROC curve method statistics the test results of single end-point ROS,ATP,caspase,GSH and composite ratio endpoint ROS/ATP,caspase/ATP,GSH/ATP.The area under the ROC curve was used to determine the performance of different endpoints identification hepatotoxic drugs at different concentrations.The results showed that the composite endpoint ROS/ATP had the best performance in discriminating between hepatotoxic and non-hepatotoxic drugs compared to other endpoints,and had the best ROC curve at concentrations of 25×C_max,50×C_max,and 100×C_max,0.9342,0.8816,and 0.9485,respectively.The best sensitivity and specificity were found at a concentration of 100×C_max,82.35%and 100%,respectively.In addition,ROS/ATP distinguishes between severely hepatotoxic drugs and non-severely hepatotoxic drugs.The area under the ROC curve at concentrations of25×C_max,50×C_max,and 100×C_max is 0.8580,0.8239,and 0.8867,respectively.The best sensitivity and specificity were found at 100×C_max concentrations,which were 100%and 73.33%,respectively.Part IV:Application of ROS/ATP ratio in hepatotoxicity detection of hiPSCs-derived hepatocytesAccording to the results of the previous study,this experiment used two hiPSC-derived hepatocyte-like cells?hiPSC-HLCs?from different sources as models for the ROS and ATP contents assay of 6 hepatotoxic drugs,to verify the predictive performance of the composite endpoint ROS/ATP ratio in different cell models.The four hepatotoxic drugs were ranked according to the results of ROS/ATP assay.The results showed that the ranking results of the two hiPSC-HLCs models were consistent with the ranking of the four drugs in clinically induced degrees of hepatotoxicity,namely amiodarone hydrochloride>Sodium valproate>Pyrazinamide>Captopril.Compared with single endpoint ROS and ATP,ROS/ATP has the best performance in distinguishing different degrees of hepatotoxic drugs in three cell models;the difference in ROS/ATP results of two hiPSC-HLCs compared to HepG2 cells is smaller,and the ROS/ATP levels of both hiPSC-HLCs are increased by each drug.Conclusions1.Detection the effect of drugs on the ratio of ROS/ATP ratio endpoint in HepG2cells at concentrations of 25-100×C_max is helpful to distinguish drugs with different degrees of hepatotoxicity.2.Detection the ROS/ATP ratio endpoint in hiPSC-HLCs at concentrations of25-100×C_max can help sort clinically drugs with different degrees of hepatotoxicity.3.Compared to HepG2 cells,the ROS/ATP endpoint of hiPSC-HLCs is more sensitive and may help to identify hepatotoxic drugs that are clinically responsible for non-hepatocyte injury.