The Journal of Toxicological Sciences, 213 Vol.30, No.3, 213-227, 2005 IDENTIFICATION OF OXIDATIVE STRESS-RELATED PROTEINS FOR PREDICTIVE SCREENING OF HEPATOTOXICITY USING A PROTEOMIC APPROACH Toshinori YAMAMOTO, Rie KIKKAWA, Hiroshi YAMADA and Ikuo HORII Worldwide Safety Sciences, Pfizer Global Research & Development, Nagoya Laboratories, Pfizer Inc., 5-2 Taketoyo, Aichi 470-2393, Japan (Received January 15, 2005; Accepted April 19, 2005) ABSTRACT — We investigated the effects of three hepatotoxicants, acetaminophen (APAP), amio- darone (AD) and tetracycline (TC), on protein expression in primary cultured rat hepatocytes with toxi- coproteomic approach, which is two-dimensional gel electrophoresis (2DE) and mass spectrometry. The objectives of this study were to search for alternative toxicity biomarkers which could be detected with high sensitivity prior to the appearance of morphological changes or alterations of analytical conventional biomarkers. The related proteins in the process of cell degeneration/necrosis such as cell death, lipid metabolism and lipid/carbohydrate metabolism were mainly affected under exposure to APAP, AD and TC, respectively. Among the differentially expressed proteins, several oxidative stress-related proteins were clearly identified after 24-hr exposure, even though they were not affected for 6-hr exposure. They were glutathione peroxidase (GPX) as a down-regulated protein as well as peroxiredoxin 1 (PRX1) and peroxiredoxin 2 (PRX2) as up-regulated proteins, which are known to serve as antioxidative enzymes in cells. These findings suggested that the focused proteins, GPX and PRXs, could be utilized as biomarkers of hepatotoxicity, and they were useful for setting high throughput screening methods to assess hepato- toxicity in the early stage of drug discovery. KEY WORDS: Rat hepatocyte, Proteomics, Oxidative stress, Glutathione peroxidase, Peroxiredoxin, Biomarker INTRODUCTION the early stage of drug discovery. Several approaches for the estimation of gene expression have been made Recent advances in new technologies such as to search for the appropriate genomic biomarkers combinatorial chemistry and high throughput pharma- (Jessen et al., 2003; Boess et al., 2003; Longueville et cological assay require high throughput screening al., 2003). However, there have been only a few reports procedures on safety evaluations in the early stage of on the protein expression related to hepatotoxicity drug discovery. In this context, the use of in vitro cyto- (Fountoulakis et al., 2002; Ruepp et al., 2002). In this toxicity assays to predict general toxicity in vivo is study, we have investigated protein expression as a reli- usually requested (Fry et al., 1990; Paillard et al., 1999; able biomarker in primary cultured rat hepatocytes Bugelski et al., 2000; Luber-Narod et al., 2001). exposed to several hepatotoxicants. We used acetami- The liver is one of the important organs for toxi- nophen (APAP), amiodarone (AD) and tetracycline cological evaluation from the aspects of the major (TC), whose mechanisms of hepatotoxicity are well metabolism site of xenobiotics as a primary target of known [APAP: hepatic necrosis is mediated by the drugs (Fautrel et al., 1991; Alden et al., 1999). More toxic metabolite, N-acetyl-p-benzoquinone imine sensitive and reliable endpoints as toxicologically (NAPQI) (Holme et al., 1984; Nelson, 1990); AD: responsible biomarkers were required for setting the phospholipidosis caused by phospholipid accumulation high throughput screening system of hepatotoxicity in (Dake et al., 1985; Sirajudeen et al., 2002); TC: steato- Correspondence: Toshinori YAMAMOTO Vo l . 3 0 N o . 3 214 T. YAMAMOTO et al. sis induced by accumulation of fat in the liver (Breen et (Eugene, OR, USA). HPLC grade solvents and 3-[(3- al., 1975; Deboyser et al., 1989)]. cholamidopropyl) dimethylammonio]-1-propane- In this paper, we have investigated the effects of sulphonate (CHAPS) were purchased from Wako Pure the hepatotoxicants on protein expression in the pri- Chemicals. Sequencing grade-modified trypsin was mary cultured rat hepatocytes. Among the differentially obtained from Promega (Madison, WI, USA). An LC- µ expressed proteins, we have focused on oxidative Packings PepMap C18 -precolumn cartridge and a stress-related proteins, and discussed the feasibility of PepMap C18 capillary column were obtained from these protein biomarkers for application to high Dionex (San Francisco, CA, USA). throughput screening systems of hepatotoxicity. As a result, it is suggested that the focused proteins would Preparation of primary cultures of rat hepatocytes be utilized as biomarkers of hepatotoxicity for setting Primary hepatocytes were isolated from Sprague- high throughput screening methods to assess hepato- Dawley male rats (Charles River Japan, Yokohama, toxicity in the early stage of drug discovery. Japan) using the two-step collagenase perfusion method (Seglen, 1976; Kikkawa et al., 2005). Rat MATERIALS AND METHODS hepatocytes were cultured in Williams’ E culture medium (WEM, Invitrogen, Carlsbad, CA, USA) sup- Experimental design plemented with 10 mM 2-[4-(2-hydroxyethyl)-1- Primary cultured rat hepatocytes were exposed to piperazinyl]ethanesulfonic acid (HEPES, Invitrogen) three hepatotoxicants, APAP, AD and TC, for 6 and 24 buffer,pH7.4,10%fetalbovineserum(FBS,Invitro- hr. Concentrations of the test compounds (APAP: 10 gen), 100 U/mL penicillin, 100 µg/mL streptomycin mM, AD: 50 µM, TC: 500 µM) were determined based (Invitrogen), and insulin-transferring-selenium-A sup- on our previous cytotoxicity study in primary rat hepa- plement (Invitrogen). Rat hepatocytes were seeded on tocytes (Kikkawa et al., 2005). Protein expression was six-well type-I collagen coated plates (BD Bio- examined by two-dimensional gel electrophoresis sciences, Bedford, MA, USA) at a density of 105 cells/ (2DE), and proteins that expressed differentially in cm2. Prior to treatment with test compounds, hepato- ° control and treated cells were identified by mass spec- cytes were incubated at 37 Cundera5%CO2/95% air- trometric analysis following In-gel tryptic digestion. humidified atmosphere for 3 hr to attain cell attach- Eventually, our goal was to address several proteins ment. Viability of cultured rat hepatocytes was tested that were affected in common among the treatments of by LDH leakage (Roche Diagnostics GmbH, Man- compounds, because these proteins were potential can- nheim, Germany), mitochondrial respiration ability didates as hepatotoxicity markers. [water-soluble tetrazolium salts (WST-1) reduction assay, Roche Diagnostics GmbH] and morphological Test compounds observations. Acetaminophen (APAP) was purchased from Sigma (St. Louis, MO, USA). Amiodarone hydrochlo- Treatment conditions ride (AD) was obtained from ICN Biomedicals Inc. The test compounds were dissolved in dimethyl (Aurora, OH, USA). Tetracycline hydrochloride (TC) sulfoxide (DMSO, Wako Pure Chemicals), and the was purchased from Wako Pure Chemicals (Osaka, final concentration of DMSO in medium (FBS(−)) was Japan). adjusted to 1% for test solutions. After pre-incubation, the medium was replaced with a medium containing Chemicals and reagents test compounds, and hepatocytes cultured in 6-well Electrophoresis grade reagents [acrylamide, aga- plates were exposed to test compounds for 6 and 24 hr. rose, ammonium persulphate (APS), bromophenol Based on the results of our previous cytotoxicity study blue (BPB), glycerol, N,N,N',N'-tetramethylethylene- (Kikkawa et al., 2005), the concentrations of the test diamine (TEMED), sodium dodecyl sulfate (SDS) and compounds were set at 10 mM, 50 µM and 500 µMfor urea] and Immobiline DryStrip (pH3-10 L, 24 cm) APAP, AD and TC, respectively. At the end of the were purchased from Amersham Biosciences exposure periods, hepatocytes were harvested using a (Uppsala, Sweden). Pharmalyte (3-10) and TRIZMA clean cell scraper and recovered into 100 mM HEPES base were obtained from Sigma. SyproRuby protein buffer. They were frozen and stored in liquid nitrogen gel stain was purchased from Molecular Probes until protein extraction for 2DE experiments. Vo l . 3 0 N o . 3 215 Oxidative stress-related proteins for hepatotoxicity screening in rat hepatocytes. Two-dimensional gel electrophoresis (2DE) They were then rehydrated in 20 ng/mL of TPCK- Total rat hepatocytes proteins were precipitated modified trypsin at 4°C. In-gel tryptic digestion was with 10% trichloroacetic acid (TCA)/acetone, and then carried out at 37°C overnight, and digested peptides proteinpelletswereresolubilizedin2D-lysisbuffer[7 were then recovered into 5%(v/v) aqueous formic acid M urea, 2 M thiourea (Invitrogen), 4%(v/v) CHAPS, (Wako Pure Chemicals). DeStreak reagent, 1%(v/v) Pharmalyte (pH3-10) and trace BPB] after completely washing with ice-cold MS/MS analysis and protein identification acetone. The hepatocytes lysates from individual wells Tryptic peptides from 2D-gel protein gel plugs (n=3) were pooled, and then their protein concentra- were subjected to peptide sequencing tags (PSTs) for tion was determined with a modified Lowry’s method protein identification using MS/MS spectra derived (Lowry et al., 1951) using RC DC Protein Assay Kit from capillary liquid chromatography coupled to a (Bio-Rad Laboratories, Hercules, CA, USA). One hun- hybrid quadrupole orthogonal acceleration time-of- dred micrograms of proteins was applied to immobilized flight mass spectrometer (CapLC and Q-TOF Ultima pH gradient gels (pH 3-10 L, 24 cm) in triplicate using API, Micromass,