| BackgroundAs an important excretion organ of human body,kidney is one of the important target organs of exogenous compounds.Due to the long period and species differences,it is time-consuming and labor intensive to evaluat nephrptoxicity of exogenous compounds used animal experiments.With the introduction of Toxicity Testing in the 21st Century:A Vision and Strategy,the focus of toxicity testing has shifted from animal experiments to in vitro models used human resource cells or cellular components.Therefore,it has become one of the urgent problems to find an alternative method with short cycle,high sensitivity,strong specificity and good prediction efficiency.Characterised by various transporters and metabolic enzymes which involved in activating exogenous compounds,the proximal tubular epithelial cells are important targets for many exogenous toxins.Therefore,it is of great significance to establish a sensitive and specific in vitro cell model for nephrotoxicity screening.A good nephrotoxicity test model should select appropriate indicators to reflect the effects of renal tubular epithelial cells when exposed to exogenous compounds.Using the endpoints which can reflect different mechanisms of damage caused by exogenous compounds could find the key events,and provide ideas for the adverse outcome pathways(AOPs).Due to the powerful compensatory function of kidney,the changes of serum creatinine and urea nitrogen,which are the basis of preclinical drug evaluation,often appear when the kidney is seriously damaged.The four biomarkers including kidney injury molecule-1(KIM-1),clusterin(CLU),osteopontin(OPN)and Neutrophil Gelatinase-Associated Lipocalin(NGAL),could early respond to renal tubular injury in rats experimental and clinical epidemiological studies.However,whether the expression of these biomarkers in vitro cell models can accurately reflect kidney injury has not been clear so far.Objective1.To explore the scientificity and feasibility of HK-2 as a cell model for nephrotoxicity evaluation of exogenous compounds,and to evaluate the sensitivity and specificity of all detection indicators.2.To evaluate the predictive ability of the HK-2 cell model by nephrotoxicants and non-nephrotoxicants confirmed by the international database.3.To evaluate the nephrotoxicity of the rare earth elements,lanthanum and gadolinium,and provide a scientific evidence for risk assessment of these food contaminants.Methods1.Establishment of a nephrotoxicity test model based on HK-2The model of HK-2 renal toxicity was established by detecting the classic cell injury indexes(cell activity,apoptosis,oxidative stress,mitochondrial injury)and early renal injury markers(KIM-1,NGAL,OPN and CLU)by using 5 reference compounds(positive reference:cisplatin,gentamycin,cadmium chloride;negative reference:levodopa and carboplatin)The lowest observed effect concentration (LOEC)was used to compare the sensitivity of each index,and the highest sensitivity index was selected for model validation.2.Validation of HK-2 cell modelIn addition,15 kinds of nephrotoxic compounds and 8 kinds of non-nephrotoxic compounds from international database with definite evidence were used to validate the HK-2 cell model.The effect of 23 compounds on the viability and OPN protein level in the supernatant of HK-2 cells was detected.The ROC curve was used to analyze the diagnostic efficacy of IC50 and fold changes of OPN as the detection endpoint,and logistics regression analysis was used to combine IC50 and fold changes of OPN as the detection endpoint.3.Evaluation of rare earth elements used HK-2 cell modelHK-2 cells were exposed to lanthanum and gadolinium,and the activity of HK-2 cells and OPN protein level in supernatant were measured.The results were substituted into logistic regression model to detect renal toxicity of La and Gd,and compared the results with the international database and the existing references.Results1.Establishment of a nephrotoxicity test model based on HK-2The cell viability results showed that cisplatin(30～60 μmol/L)and gentamicin(16～64 mmol/L)significantly reduced the viability of HK-2 cells(P<0.05),both in a dose-dependent manner.Lower concentrations of cadmium chloride(<60 μmol/L)had no effect on cell viability,while high concentrations of cadmium chloride significantly reduced the cell viability of HK-2(P<0.05).No effect of carboplatin and levodopa on cell viability of HK-2 was observed.The results of oxidative stress and mitochondrial function indicators showed that exposure to cisplatin(5μmol/L)and cadmium chloride(80 μmol/L)significantly increased the expression of ROS in HK-2 cells(P<0.05),while gentamicin,carboplatin and levodopa had no effect on ROS expression at any does investigated.Exposure to cisplatin,gentamicin and cadmium chloride inhibited the levels of SOD and ATP in HK-2 cells(P<0.05).However,carboplatin and levodopa had no effect on SOD and ATP levels in HK-2 cells.The results of cell apoptosis showed that cisplatin(10μmol/L)and cadmium chloride(60 μmol/L)caused early apoptosis of HK-2 cells(P<0.05),while gentamicin,carboplatin and levodopa did not cause HK-2 cell apoptosis at any does investigated.The results of early kidney injury biomarkers showed that,cisplatin(20μmol/L),gentamicin(4 mmol/L)and cadmium chloride(60μmol/L)significantly increased the mRNA levels of OPN and CLU in HK-2 cells(P<0.05),and in a dose-dependent manner.Among them,OPN was not affected by carboplatin and levodopa,while the mRNA level of CLU increased significantly in the middle and high dose group of carboplatin(P<0.05).Secretion protein of OPN level was consistent with its mRNA level.The CLU protein level was only significantly increased in the highest dose group of cisplatin(P<0.05).The mRNA level and protein level of NGAL only increased in the cisplatin treatment group(P<0.05).The mRNA level of KIM-1 did not show a significant increase or decrease in the five compounds,but its protein level increased significantly in the cadmium chloride treatment group(P<0.05).Comparing the lowest observable effect concentration of all the detection indicators,it was found that the LOEC values of OPN and CLU indicators are generally lower than other indicators.However,due to the false positive results of CLU indicators,OPN was selected as a representative specific biomarker in validation panel.2.Validation of HK-2 cell modelThe result of validation of HK-2 cell model show that the sensitivity and specificity of IC50 value were 50%and 100%,respectively.And the maximum Youden index of IC50 value was 0.5.When fold changes of OPN was used as the endpoint,the sensitivity and specificity of IC50 value were 72.22%and 100%,respectively.And the maximum Youden index of IC50 value was 0.72.Using IC50 and fold changes of OPN as the detection endpoint,the sensitivity and specificity were 72.22%and 100%,respectively.And the maximum Youden index is 0.72.The logistics regression model formula was:Y=-2.798+2.46×fold changes of OPN +0.011 × IC50(mmol/L).3.Evaluation of rare earth elements used HK-2 cell modelThe nephrotoxicity of rare earth elements,lanthanum and gadolinium was evaluated.According to the regression equation of IC50 and fold changes of OPN,the predicted values of lanthanum and gadolinium were 1.426 and 1.345 respectively.Compared to the cut-off value(0.519),it was judged that the two substances may be nephrotoxic.Conclusions1.Nephrotoxicants and non-nephrotoxicants could be preliminary identified by HK-2 cell model.The biomarkers of renal tubular injury(OPN and CLU)are more sensitive than the classic toxicity indicators.Among them,OPN has better sensitivity and specificity.2.Validation of HK-2 cell model show that the sensitivity of fold changes of OPN was higher than IC50 value.When OPN and IC50 were combined as the detection indicators,the detection efficiency of the HK-2 nephrotoxicity test model was improved.3.The rare earth elements,lanthanum and gadolinium,were toxic to HK-2 cells at 100 μmol/L and 1 mmol/L,respectively. |
