What Do We Learn about Hepatotoxicity Using Coumarin-Treated Rat Model? authors M. David Ho 1, Bob Xiong 1, S. Stellar 2, J. Proctor 2, J. Silva 2, H.K. Lim 2, Patrick Bennett 1, and Lily Li 1 Tandem Labs, New England 1, and Johnson & Johnson PRD 2 introduction Coumarin is a well documented hepatotoxicant in rats and the metabolism is well studied including the identification of 3-hydroxycoumarin, 7-hydroxycoumarin, o-hydroxyphenylethanol (o-hpe), and O-hydroxyphenylacetic acid (o-hpaa) as metabolites of coumarin. Two reactive metabolites, o-hydroxyacetaldehyde (o-ha) and coumarin 3, 4-epoxide (CE) were postulated to cause liver injuries in rodents. This model was used to identify potential biomarkers for hepatotoxicity with different mass spectroscopic analytical methods for correlation analysis with coumarin metabolites and histological data. This study provides a survey of the coumarin metabolite profiles in rat liver, plasma, red blood cells and urine. Two major conjugate metabolites were found in liver and urine. Histological changes were observed in liver slices from the treated rats. The data obtained through the Tandem biomarker screening program reveal significant up-regulation and down-regulation for certain endogenous molecules. These facts indicated that the lipid, amino acid, carbohydrate, and energy metabolisms in rats were altered 8 hours following the administration of coumarin. experimental Materials and Chemicals Rats were dosed with coumarin at 400 mg/kg orally and sacrificed at 8 hours post dose administration. Plasma, red blood cell (RBC), urine and liver samples were collected at zero time and 8 hours from both vehicle-treated control and coumarin-treated rats. Coumarin, 3-OH coumarin, 7-OH coumarin, and o-hydroxylphenylacetic acid (o-hpaa) standards were used as references to confirm coumain metabolites detected in rat. Amino acids, catcholamines, nucleotides, organic acids etc were spiked into rat plasma, urine, RBC, and liver homogenates, respectively, to be used as QC for the measurement of endogenous molecules. M. David Ho et al. (2009) 1
EXPERIMENTAL continued Sample Preparation Liver tissues were homogenized with phosphate buffer (1:3 v:v). Rat plasma, urine, red blood cells, and liver homogenates were precipitated with 100% acetonitrile with glyburide as an internal standard. The supernatants were evaporated under N 2 using a TurboVap 96. Prior to sample analysis, water containing 0.1% formic acid and 5 mm NH 4 OAc were used to reconstitute the dry extracts for the mass spectrometric data acquisition under positive and negative modes. LC-TOF MS Metabolite Profiling QSTAR XL (Applied Biosystem/Sciex) was used as a mass detector in connection with HPLC separation using an electrospray (ES) interface. The mass range used for LC-TOF/MS experiment was from m/z 100 to m/z 1000 under positive and negative modes. LC/MS/MS Biochemical Profiling API 4000 (Applied Biosystem/Sciex) was used as a mass detector in connection with HPLC separation using an electrospray (ES) interface. Two different gradients were applied onto Phenomenex Luna Phenylhexyl and Synergi Polar-RP columns. Endogenous molecules (e.g, amino acids, catcholamines, nucleotides, organic acids etc) were measured through multiple reaction monitoring (MRM) under a postive and a negative modes using a turbo ion spray source. Data Processing and Analysis LC-TOF MS Metabolite Profiling The LC-TOF data were collected and processed using Analyst QS 1.1. The TOF MS spectra of the coumarin treated and non-treated samples were compared at each retention time for differences. The ion chromatogram of each individual metabolite or endogenous substance found was extracted to further confirm the profile between control and treated samples. The peak area ratios were calculated against the internal standard and summarized. Analytical standards (e.g., coumarin, 7-OH courmarin, 3-OH coumarin, o-hpaa etc) were used to confirm the identified metabolites in plasma, urine, liver, and RBC. M. David Ho et al. (2009) 2
Data Processing and Analysis continued LC/MS/MS Biochemical Profiling LC/MS/MS data were processed using Analyst 1.4.2 (Quantitate Package). For each endogenous molecule, spiked analytical standards in a specific matrix were used to assign ID based on retention time and the transition of precusor/product ion pairs. The peak area ratios were calculated against the internal standard and used for biochemical profiling of control and treated samples. R Statistical Computing Program was used as a tool for statistical analysis. Identified potential biomarkers were selected at p-value of 0.01. Spotfire was applied to visualize the large quantity of LC/MS/MS data. Results and Discussion The metabolic profiles of coumarin were quite different in each of the collected biological matrices (Figure 1). Coumarin was mainly excreted in urine. The coumarin glutathione conjugate was predominantly present in liver tissue homogenates as compared to in plasma, RBC, and urine. The coumarin cysteine conjugate was also detected in the liver tissue which could be produced by further metabolism of the glutathione conjugate by γ-glutamytranspeptidase and aminopeptidase. The presence of the glutathione and cysteine conjugates of coumarin in liver tissue may be related to the cell necrosis/apoptosis observed in the histological study (Figure 2). The coumarin glucuronide conjugate was significantly detected in urine. No o-hpaa was found in liver tissues. 0.5% - 1% of o-hpaa was detected in urine, plasma, and RBC. Both 3-hydroxycoumarin and 7-hydroxycoumarin were detected in plasma but not in other three matrices. A metabolic pathway of coumarin in the rat is shown in Figure 3. In addition to metabolites, we observed large amounts of endogenous glutathione accumulated in urine, liver, RBC, and plasma (Figure 4). The over-expression of glutathione S-transferase activity and generation of the glutathione conjugate and its related metabolites may be correlated with the hepatotoxicity induced by coumarin. Through LC/MS/MS biochemical profiling, some of endogenous molecules were significantly elevated and others were down-regulated in coumarin-treated rats as shown in Figure 5. The relative abundances of individual endogenous molecules measured in rat liver, plasma, RBC, and urine were plotted in Figure 6. Lipid, amino acid, carbohydrate, and energy metabolisms (Figure 7) were altered after administration of coumarin (400 mg/kg) in rat. M. David Ho et al. (2009) 3
Figure 1: Relative Abundances of Coumarin Metabolites in Rat Figure 2: Histological Study of Single Liver Cell: Observation of Necrosis and Apotosis M. David Ho et al. (2009) 4
Figure 3: Coumarin Metabolic Pathways in Rat Figure 4: Glutathione Accumulated in Urine, Liver, Plasma, and RBC M. David Ho et al. (2009) 5
Figure 5: Up- and Down-Regulated Endogenous Molecules Observed in Coumarin-Treated Rat Figure 6: Relative Abundances of Coumarin Metabolites and Endogenous Molecules in Rat M. David Ho et al. (2009) 6
Figure 7: Alternated Biochemical Pathways in Courmarin-Treated Rats conclusions The coumarin metabolite profile in rat liver, plasma, red blood cells and urine showed that there were two major conjugate metabolites, coumarin glutathione and cysteine conjugates, present in the liver tissue. They may be related to the cell necrosis/apoptosis observed in the histological study. The over-expression of glutathione S-transferase activity generating excessive amount of glutathione in liver, plasma, red blood cells and urine may be adaptive response to the hepatotoxicity induced by coumarin. Some endogenous molecules were either significantly up-regulated or down-regulated, indicating that the metabolism of lipid, amino acid, and energy production in rats had been alternated critically 8 hours after the administration of coumarin. Acknowledgement We thank Ms. Qing Zhu for her hard work in tissue homogenization and all sample preparation for the LC-TOF/MS and LC/MS/MS analyses. M. David Ho et al. (2009) 7