Biomarker Discovery In Valproic Acid Induced Hepatotoxicity Based On The Levels Of Toxic Metabolites And Metabolic Pathways And Relative Mechanism Study

Objectives:Valproic acid（Valproic acid,VPA）is a widely used anticonvulsant drug that can treat various types of epilepsy as monotherapy or adjunctive therapy.VPA is well tolerated but it may lead to fatal hepatotoxicity,potentially limiting its application in clinical.The mechanism of hepatotoxicity induced by VPA has been widely studied,but its exact molecular mechanism has not been identified.Also,there are no clear biomarkers that can predict the occurrence of hepatotoxicity.Considering that the liver injury caused by VPA is multi-pathway and multi-mechanism,the purpose of this study is to explore the biomarkers that can warn the hepatotoxicity of VPA from the level of toxic metabolite and metabolic pathway by UPLC-MS/MS and full quantitative targeted metabonomics technology from the epilepsy patients undergoing liver failure in VPA treatment in clinical.In addition,the related mechanisms are explored.The results of the study may provide new methods and strategies for the clinical safe use of VPA.Methods:1、An UPLC-MS/MS method for simultaneous determination of VPA and its toxic metabolite was established:VPA-D6 was used as the internal standard and the serum samples were treated with ice acetonitrile for protein precipitation.VPA and 4-ene-VPA were separated on the Agilent Eclipse Plus RHD column（2.1mm×100 mm,1.8μm）by acetonitrile-10mmol/L ammonium acetate aqueous solution.ESI negative ion mode and Pseudo-MRM was used in the mass spectrometry.The established method was verification and compared with the CMIA that was used in TDM.2、Correlation analysis of VPA and its toxic metabolite with abnormal liver function:Serum samples and the related medical records of epileptic patients treated in our hospital were collected.According to the different indexes of liver function,the samples were divided into abnormal liver function（ANLF）group and normal liver function（NLF）group.The concentrations of VPA and 4-ene-VPA were determined by UPLC-MS/MS method established above.Logistic regression analysis was used to analyze the risk factors of abnormal liver function induced by VPA.Odds ratio（OR）and 95%confidence interval（95%Cl）were used to evaluate the risk of various factors and abnormal liver function.ROC curve was used to evaluate the predictive performance of CDR_4-ene-VPA for abnormal liver function induced by VPA.In addition,the cut-off value of CDR_4-ene-VPAwas calculated.3、Screening of biomarkers by full quantitative targeted metabonomics:Serum samples and related medical records of epileptic children treated with VPA in our hospital were collected.According to the difference of liver function index,the samples were divided into severe abnormal liver function group（ANLF2）,mild liver function normal group（ANLF1）and normal liver function group（NLF）.The full quantitative targeted metabonomics based on Q300^TMtechnology was used to determine the content of 195endogenous metabolites.Statistical analysis was carried out by PCA,OPLS-DA and other statistical analysis methods.The potential biomarkers was screened according to the double standard of VIP>1 in multivariate analysis and p<0.05、|log2FC|>0 in univariate analysis.The predictive ability of the selected biomarkers was analyzed by ROC curve.The correlation between the selected biomarkers and the level of 4-ene-VPA was also evaluated.4、The role of SIRT1 in the abnormal liver function induced by VPA:The effects of VPA on the m RNA and protein expression of SIRT1,FXR and SHP were measured by Real-time PCR and Western Blot.The effect of VPA on cell proliferation and the expression of SIRT1,FXR and SHP were compared after pretreated with SIRT1 agonist（SRT1720）.The effect of VPA on the acetylation level of FXR was investigated by immunoprecipitation method.Also the role of SIRT1 in this process was exploredResults:1、An UPLC-MS/MS method for simultaneous determination of VPA and its toxic metabolite was established:VPA,4-ene-VPA and internal standard were separated well and the retention time was 3.91 min,3.04 min and 3.87 min,respectively.The linear range of VPA and 4-ene-VPA were 1.00-200.00μg/m L and 10.00-500.00 ng/m L,respectively.The lower limits of quantification are 1μg/m L and 10 ng/m L,respectively.Precision,accuracy,recovery,matrix effect,carry over,dilution reliability and stability test all met the requirements.The results of comparison with the CMIA method showed that there was good correlation between the two methods.The result of Deming regression was Y_UPLC-MS/MS=0.9868 X_CMIA+0.1280（r=0.9923）,the 95%confidence intervals of slope and intercept were（0.9907-0.9936）and（-0.5119-0.7679）.The concentration of VPA determined by UPLC-MS/MS method was 0.7μg/m L lower than that by CMIA method.2、Correlation analysis of VPA and its toxic metabolite with abnormal liver function:The levels of 4-ene-VPA,CDR_4-ene-VPA and CDR_VPA in ANLF group were significantly higher than those in NLF group,but there was no significant difference in VPA concentration between the two groups.The results of Logistic regression showed that the risk of abnormal liver function induced by VPA increased with the level of CDR_4-ene-VPAvalue.The results of ROC curve indicated that the AUC of CDR_4-ene-VPA was 0.769>0.750,with 95%Cl:0.713～0.825,p<0.05.So it couold be concluded that CDR_4-ene-VPA level may be used for predicting hepatotoxicity induced by VPA.The cut-off value of CDR_4-ene-VPAwas 9.26 ng/m L per mg/kg.3、Screening biomarkers by full quantitative targeted metabonomics:A total of 22 children with epilepsy were included in the study,including ANLF1 group（n=7）,ANLF2 group（n=7）and NLF group（n=8）.There was no significant difference in age,sex,height,weight and daily dose of 4-ene-VPA among the three groups,but there was significant difference in VPA concentrations among the three groups.The concentrations of VPA in ANLF2 group was significantly higher than that in ANLF1 group and NLF group.Based on the results of multivariate and univariate statistical analysis,21biomarkers were screened including 5 organic acids,4 short chain fatty acids,4 amino acids,3 fatty acids,3 benzenoids,1 carnitine and 1 carbohydrate.The AUC of all the 21biomarkers were>0.750,showing good prediction performance.Among the 21biomarkers,7 metabolites including 2-hydroxyglutaric acid,ketoleucine,propanoic acid,L-glutamine,methylsuccinic acid,4-hydroxy phenylpyruvic acid,benzenebutanoic acid were strongly correlated with the level of 4-ene-VPA.There were only three bile acid metabolites showing significant differences between ANLF1 and ANLF2 groups including taurodeoxycholic acid,taurochenodeoxycholic acid and deoxycholic acid.The contents of these three components increased at first and then decreased with the aggravation of liver injury,which may be used as characteristic biomarkers of progressive liver injury.4、The role of SIRT1 in the abnormal liver function induced by VPA:The proliferation of Hep G2 cells was inhibited by being treated with 2m M VPA for 48 hours.Also,the activities of ALT and AST were significantly increased,and the contents of total bile acid,triglyceride and total cholesterol were significantly increased.Moreover,VPA can down-regulate the m RNA and protein expression of SIRT1,FXR and SHP,and up-regulate the m RNA and protein expression of CYP7A1 and CYP8B1.Pretreatment with SRT1720,a SIRT1 agonist,could alleviate the inhibitory effect of VPA on the proliferation of Hep G2cells.Also,the down-regulation of SIRT and FXR and the up-regulation of CYP7A1,CYP8B1 m RNA and protein expression,and the severity of liver dysfunction induced by VPA could be alleviated.The results of co-immunoprecipitation showed that VPA could significantly increase the acetylation level of FXR.Conclusions:1、An UPLC-MS/MS method was developed and validated for simultaneous determination of VPA and its toxic metabolite,which can be used for routine determination of VPA concentrations,leading to potential improvements in patient care and laboratory management.CDR_4-ene-VPA showed good performance of predicting the abnormal liver function induced by VPA.The cut-off value of CDR_4-ene-VPAwas 9.26 ng/m L per mg/kg for the occurrence of VPA-induced abnormal liver function.2、21 potential biomarkers were screened including 5 organic acids,4 short chain fatty acids,4 amino acids,3 fatty acids,3 benzenoids,1 carnitine and 1 carbohydrate.The contents of taurodeoxycholic acid,taurochenodeoxycholic acid and deoxycholic acid increased at first and then decreased with the aggravation of liver injury,which may be used as characteristic biomarkers of progressive liver injury.3、VPA may increase the acetylation level of FXR by inhibiting the activity of SIRT1,and then inhibit the regulation of target genes.The activation of SIRT1 may be an effective approch to prevent abnormal liver function induced by VPA.