Overexpression of thioredoxin prevents acute hepatitis caused by thioacetamide or lipopolysaccharide in mice

Overexpression of Thioredoxin Prevents AcuteHepatitis Caused by Thioacetamide or

Lipopolysaccharide in MiceHiroaki Okuyama,1 Hajime Nakamura,2 Yasuyuki Shimahara,1 Shinichi Araya,2 Norifumi Kawada,3

Yoshio Yamaoka,1 and Junji Yodoi2

Thioredoxin (Trx) is a small redox-active protein with antioxidant and antiapoptotic effects. Trxtransgenic (Tg) mice are more resistant to cerebral infarction and survive longer than wild-type(WT) C57BL/6 mice. The aim of the present study was to investigate the protective role of Trx inacute hepatitis models. The expression of endogenous Trx was decreased in thioacetamide(TAA)-induced acute hepatitis. TAA (100 �g/g) was injected intraperitoneally in WT and Tgmice. Survival rate after TAA injection was higher in Tg mice than in WT mice. The level ofoxidative stress was significantly less in Tg mice than in WT mice, as shown by the proteincarbonylation assay and lipid peroxidation assay. Terminal deoxynucleotidyl transferase–medi-ated deoxyuridine triphosphate nick-end labeling (TUNEL)-positive cells were less in Tg micethan in WT mice, which was consistent with DNA laddering assay. Caspase-3 and caspase-9activities and cytochrome c release were significantly inhibited in Tg mice compared with thosein WT mice. In addition, lipopolysaccharide (LPS) plus D-galactosamine (GalN), or anti-Fasantibody (Jo2) were injected. Survival rate after LPS plus GalN injection was much higher in Tgmice than in WT mice. In contrast, there was no difference in survival rate after Jo2 injectionbetween WT and Tg mice. In conclusion, transgene of Trx attenuated TAA- or LPS-inducedacute lethal hepatitis. In addition to an antioxidant effect, Trx has the potential to protect acuteliver injury via an antiapoptotic effect, which mainly inhibits mitochondria-mediated apoptosissignaling. (HEPATOLOGY 2003;37:1015-1025.)

Fulminant hepatitis carries a very high mortality,resulting from acute hepatitis caused by virus infec-tion, alcohol, or drugs. Conventional medical ther-

apies can rescue only about 10% of patients with

fulminant hepatitis. Although liver transplantation hasimproved their mortality, about 40% of these patients diewhile waiting for liver transplantation.1

Oxidative stress contributes to the pathogenesis ofacute hepatitis induced by alcohol, virus infection, hemo-chromatosis, ischemia/reperfusion injury, toxic expo-sures, and drug abuse such as acetaminophen overdose.2-4

Free radicals are toxic to various cells, including hepato-cytes and initiate reactive oxygen species (ROS)-mediatedcascade causing hepatocyte cell death, leading to acutehepatitis. Therefore, radical scavengers such as pyrroli-dine dithiocarbamate have been tried to improve acuteliver injury via blocking ROS-mediated hepatocyte celldeath.5 However, the effects of most antioxidants againstacute lethal hepatitis are not satisfactory enough to applyto clinical use.

Thioredoxin (Trx) is an endogenous multifunctionalprotein with a redox-active disulfide/dithiol within theconserved active site sequence: Cys-Gly-Pro-Cys.6 Origi-nally, we cloned human Trx (hTrx) as adult T cell leuke-mia-derived factor (ADF) produced by human T cellleukemia virus type-I-transformed T cells.7,8 Trx is a

Abbreviations: ROS, reactive oxygen species; Trx, thioredoxin; hTrx, human Trx; ADF,adult T cell leukemia-derived factor; NF�B, nuclear factor �B; AP-1, activator protein-1;Ref-1, redox factor-1; Tg, transgenic; WT, wild-type; TAA, thioacetamide; LPS, lipopolysac-charide; GalN, D-galactosamine; mTrx, mouse Trx; SDS, sodium dodecyl sulfate; SDS-PAGE, SDS-polyacrylamide gel electrophoresis; HE, hematoxylin-eosin; TUNEL, terminaldeoxynucleotidyl transferase–mediated deoxyuridine triphosphate nick-end labeling; PBS,phosphate-buffered saline; EDTA, ethylenediaminetetraacetic acid; DNP, dinitrophenyl;MDA, malondialdehyde; TNF, tumor necrosis factor; ASK1, apoptosis signal-regulatingkinase 1; SAPKs, stress-activated protein kinases.

From the 1Department of Gastroenterological Surgery, Graduate School of Medicineand 2Department of Biological Responses, Laboratory of Infection and Prevention, In-stitute for Virus Research, Kyoto University, Kyoto; and the 3Department of Hepatology,Graduate School of Medicine, Osaka City University, Osaka, Japan.

Received November 20, 2002; accepted March 2, 2003.Supported by a grant-in-aid for Scientific Research from the Ministry of Education,

Culture, Sports, Science and Technology.Address reprintrequests to: JunjiYodoi,M.D.,Ph.D.,DepartmentofBiologicalResponses,

Institute for Virus Research, Kyoto University, 53, Shogoin, Kawahara-cho, Sakyo-ku, Kyoto606-8505, Japan. E-mail: [email protected]; fax: (81) 75-761-5766.

Copyright © 2003 by the American Association for the Study of Liver Diseases.0270-9139/03/3705-0010$30.00/0doi:10.1053/jhep.2003.50203

1015

stress-inducible protein the expression of which is en-hanced by various types of stresses, e.g., viral infection,exposure to ultraviolet light, x-ray irradiation, and hydro-gen peroxide.9 Trx scavenges singlet oxygen and hydroxylradicals by itself10 and hydrogen peroxide together withperoxiredoxin.11 Recombinant Trx has protective activityagainst ROS-mediated cytotoxicity.12 Moreover, Trx reg-ulates the activation of transcription factors, includingnuclear factor �B (NF-�B), activator protein-1 (AP-1),and redox factor-1 (Ref-1).13,14

Trx attenuates focal ischemic brain damage by scav-enging ROS in Trx transgenic (Tg) mice.15 Pancreaticbeta cell-specific expression of Trx prevents autoimmuneand streptozotocin-induced diabetes.16 Trx inhibits infec-tion by influenza virus and EB virus.17,18 In hepatitis Cvirus infection, serum level of Trx is elevated, serving as apossible indicator for the effectiveness of interferon ther-apy.19 Circulating Trx prevents neutrophil extravasationand suppresses inflammation.20

Based on these considerations, we hypothesized thatTrx could attenuate oxidative stress-mediated acute hep-atitis. To clarify the protective role of Trx for acute liverinjury, we subjected C57BL/6 wild-type (WT) and Tgmice to thioacetamide (TAA)-induced acute lethal hepa-titis, because TAA is known to be a hepatotoxin via gen-eration of ROS. Moreover, to examine whether Trx couldattenuate apoptotic liver failure, we subjected WT and Tgmice to lipopolysaccharide (LPS) plus D-galactosamine(GalN)- or anti-Fas antibody (Jo2)–induced acute liverfailure.

Materials and Methods

Reagents. The expression of hTrx and mouse Trx(mTrx) proteins was determined by using anti-hTrxmonoclonal antibody (ADF 11-mAb) and anti-mTrxpolyclonal antibody (RedoxBioscience, Inc., Kyoto, Ja-pan), respectively, as previously described.15 TAA waspurchased from Wako Pure Chemical Co. (Osaka, Ja-pan). Purified hamster anti-mouse Fas monoclonal anti-body (clone; Jo2) was purchased from BD Pharmingen(San Jose, CA). LPS from Escherichia coli 0111:B4 andGalN were purchased from Sigma-Aldrich Chemicals (St.Louis, MO). Unless specifically indicated, all other re-agents were purchased from Sigma (St. Louis, MO).

Animals. Tg mice were originally provided by Ajino-moto, Inc. (Kawasaki, Japan). WT (C57Bl/6) and Tgmice were housed at a constant temperature and suppliedwith laboratory chow and water ad libitum. Procedureswere performed according to the recommendations of theinstitutional animal care committee.

Acute Liver Failure Models. Male mice weighing 25to 30 g were used for in vivo liver injury models. TAA(100 �g/g) was injected intraperitoneally into WT (n �16) and Tg mice (n � 16). We observed the survival ofTAA-treated mice until 5 days. To estimate the patho-physiologic values of livers from WT and Tg mice, 24hours after TAA administration, mice were anesthetizedby diethylether, and the livers were removed. In addition,LPS (1 �g/body) plus GalN (10 mg/body) were injectedintraperitoneally into WT (n � 15) and Tg mice (n �15). Anti-Fas antibody Jo2 (20 �g/body) was injectedintraperitoneally into WT (n � 13) and Tg mice (n �14). We observed the survival of LPS plus GalN- or Jo2-treated mice until 5 days. To estimate the pathophysio-logic values of livers, these mice were killed 6 hours afterLPS plus GalN or Jo2 administration, and the livers wereremoved.

Western Blotting. To determine Trx (hTrx andmTrx), Bcl-2, Bcl-xL, and Bax, on total cell lysates, theliver samples were homogenized in 1 � sample buffer(100 �L/35-mm dish) (62.5 mmol/L Tris-HCl, pH 6.8,10% glycerol, 2% sodium dodecyl sulfate [SDS], 5%2-mercaptoethanol, 1 mmol/L Na3VO4). After heat de-naturation, the samples (10 �g of protein) were subjectedto SDS-polyacrylamide gel electrophoresis (SDS-PAGE)(12% or 15%) and then transferred onto an Immobilon Pmembrane (Millipore Corp., Bedford, MA). The mem-branes were subsequently treated with antibody againsthTrx (ADF-11mAb), mTrx, Bcl-2 (clone 3F11; Pharm-ingen), Bcl-xL (clone S-18; Santa Cruz, CA), and Bax(clone N-20; Santa Cruz). Cytochrome c determinationwas performed on the cytosolic (S100) and enriched mi-tochondrial fractions of mouse liver extracts, as describedby Yang et al.21 Proteins were electrophoresed on SDS-PAGE (15%), transferred to Immobilon P membranes,and incubated with anti–cytochrome c mAbs (clone7H8.2C12; Pharmingen).

Immunohistochemistry. Tissue samples were recov-ered, fixed in 10% buffered formalin, and embedded inparaffin. Sections (5 �m) were stained with hematoxylin-eosin (HE) for morphologic examination under micros-copy. Formalin-fixed tissues also were used to detect theexpression of mTrx and hTrx. Briefly, tissue sections weretreated with 3% hydrogen peroxide to inactivate endoge-nous peroxidase activity. The slides were incubated withprimary antibody or nonspecific immunoglobulin Governight at 4°C, followed by incubation with biotinyl-ated second antibody for 1 hour at room temperature.After 30 minutes of avidin-biotin amplification (ABCelite, Vector Laboratories, Burlingame, CA), they wereincubated with the substrate 0.1% 3�,3�-diaminobenzi-dine at room temperature for 10 minutes.

1016 OKUYAMA ET AL. HEPATOLOGY, May 2003

TUNEL Staining. Apoptosis was detected by termi-nal deoxynucleotidyl transferase–mediated deoxyuridinetriphosphate nick-end labeling (TUNEL) using theapoptosis detection kit (Wako Pure Chemical Co.). Fro-zen tissue samples were fixed with 4% formalin phos-phate-buffered saline (PBS) solution. TUNEL stainingwas performed according to the manufacturer’s instruc-tions.

DNA Fragmentation Assay. Liver tissue sampleswere minced into small pieces. The pellet was resus-pended in 1 mL of lysis buffer consisting of 10 mmol/LTris-HCl, pH 7.4, 10 mmol/L NaCl, 10 mmol/L ethyl-enediaminetetraacetic acid (EDTA), 100 �g/mL protein-ase K, and 0.5% SDS and incubated for 2 hours at 50°Cbefore being treated with ribonuclease overnight at 37°C.After extraction with phenol-chloroform twice and pre-cipitation with ethanol, the DNA was dissolved in TEbuffer (10 mmol/L Tris-HCl, pH 7.5, 1 mmol/L EDTA).The DNA was loaded onto 1.5% agarose gel containingethidium bromide, electrophoresed in Tris acetate/EDTA buffer for 2 hours at 50 V, and photographedunder ultraviolet illumination.

Caspase-3, -9, and -8 Activities. Caspase-3, -9, and-8 activities were determined using 200 �mol/L chromo-genic substrates N-acetyl-Asp-Glu-Val-Asp-p-nitroanilide(DEVD-pNA), N-acetyl-Leu-Glu-His-Asp-p-nitroanilide(LEHD-pNA), and N-acetyl-Ile-Glu-Thr-Asp-p-nitroani-lide (IETD-pNA) (Biovision, Mountain View, CA), respec-tively. Each substrate was incubated with 100 �g of cytosolicfraction (S100) in 100 �L of 100 mmol/L HEPES, pH 7.4,20% glycerol, 10 mmol/L dithiothreitol and protease inhib-itors for 1 hour. Absorbance was measured at 405 nm.

Detection of Oxidized Proteins. Oxidized proteinwas detected using an oxidized protein detection kit(Oxyblot; Oncor, Illkirch, France). The Oxyblot providesreagents for sensitive immunodetection of carbonylgroups, which is a hallmark of the oxidation status ofproteins. The carbonyl groups in the protein side chainsare derivatized to 2,4-dinitrophenylhydrazone (DNP-hy-drazone) by reaction with 2,4-dinitrohydrazine. TheDNP-derivatized protein samples are separated on a 15%SDS-PAGE followed by Western blotting. The filters areincubated with primary antibody, specific to the DNPmoiety of the proteins, which was followed by incubationwith a horseradish peroxides-antibody conjugate goat an-ti-rabbit immunoglobulin G. The amount of carbonylgroups extracted from 10 �g of sample proteins was mea-sured by an image analyzer compared with 2.5 �L of acontrol DNP-related protein mixture.

Statistical Analysis. Data presented as bar graphs arethe means � SD of 3 independent experiments except for

in vivo analysis of more than 5 independent experiments.Statistical analysis was performed by Student’s t test (P �.01 was considered significant). Kaplan-Meier survivalanalysis was used for survival data by log-rank test andWilcoxon test.

Results

Down-regulation of Endogenous Trx in Acute Hep-atitis. We checked the expression of endogenous mTrxin acute hepatitis. Immunohistochemistry showed thatthe expression of Trx was prominently decreased in thepericentral areas of TAA-induced acute hepatitis (Fig. 1).The expression of mTrx was decreased in parenchymalcells of TAA-induced acute hepatitis (Fig. 1). In contrast,the expression of Trx was maintained in nonparenchymalcells independently of TAA treatment (Fig. 1). These re-sults indicated that TAA-induced acute hepatitis showsthe intralobular heterogeneity of liver injury, that is he-patocellular injury, only in the pericentral areas. Westernblot also showed that the expression of mTrx was signifi-cantly less in TAA-treated livers than in normal livers(Fig. 2B and C; P � .05).

Attenuation of Acute Lethal Hepatitis in Tg Mice.We used Tg mice to check the protective role of Trx foracute hepatitis. Immunohistochemistry showed that there

Fig. 1. Down-regulation of endogenous Trx in TAA-induced acutehepatitis. Immunohistochemical staining for mTrx in normal mouse liver(upper left, original magnification �100; upper right, �400) andTAA-induced acute hepatitis liver (lower left, �100; lower right, �400).Endogenous Trx expression was prominently decreased in the pericentralareas of TAA-induced acute hepatitis (lower left). Trx expression inparenchymal cells was decreased, whereas Trx expression in nonparen-chymal cells (arrowhead) was maintained (upper right and lower right).Arrows indicate the central veins, and arrowheads indicate the non-parenchymal cells.

HEPATOLOGY, Vol. 37, No. 5, 2003 OKUYAMA ET AL. 1017

was an abundant expression of hTrx in parenchymal cellsof the livers of Tg mice (Fig. 2A), and Western blotshowed that Tg mice had the constitutive expression ofhTrx (Fig. 2B). We subjected both WT and Tg mice toTAA-induced acute lethal hepatitis. Survival rate afterTAA administration in WT mice (n � 16) was 68.7% (24hours) and 37.5% (48 hours), whereas survival rate in Tgmice (n � 16) was 93.5% (24 hours) and 93.5% (48hours). There was a significant difference in survival rateafter TAA administration between WT and Tg mice (P �0.0012 by Log-Rank test, P � 0.0018 by Wilcoxon test,Fig. 3). Twenty-four hours after TAA administration, theaspartate aminotransferase (AST) and alanine amino-transferase (ALT) levels were significantly lower in Tg

Fig. 2. hTrx expression in Tg mice. (A) Im-munohistochemical staining using antibodyagainst hTrx in WT (left) and Tg mice (right). Tgmice expressed hTrx diffusely in the liver. Orig-inal magnification �200. (B) Expression ofhTrx, mTrx, and �-actin in normal and TAA-treated livers of WT and Tg mice. Lane 1,nontreated WT mice; lane 2, nontreated Tgmice; lane 3, TAA-treated WT mice; lane 4,TAA-treated Tg mice. Equal amounts of celllysates (10 �g) were electrophoresed and sub-jected to immunoblot analysis using anti-hTrx,mTrx, and �-actin antibodies. (C) Densitometricanalysis of mTrx and �-actin expression. TAAsignificantly decreased mTrx expression in WTmice (*P � .05).

Fig. 3. Survival rates in WT and Tg mice were shown. Survival rateswere analyzed until 5 days after TAA administration (100 �g/g). Astatistically significant difference in survival was found between WT (n �16) and Tg mice (n � 16). Kaplan-Meier survival rates were calculatedby log-rank test and Wilcoxon test. P value by log-rank test is shown inthe figure.


mice than in WT mice (AST: 7,930 � 2,711 U/mL inWT mice versus 1,417 � 772 U/mL in Tg mice, P � .01;ALT: 10,933 � 4,305 U/mL in WT mice versus 1,885 �1,288 U/mL in Tg mice, P � .01; Fig. 4A). Histologicanalysis by HE staining showed that the destruction ofhepatic sinusoid with massive thrombosis was observed inthe pericentral areas of the livers in WT mice, whereas itwas attenuated in Tg mice (Fig. 4B).

Suppression of Oxidative Stress in Tg Mice. TAA-induced acute hepatitis is partly mediated by ROS.Hence, we checked whether Trx affected the level ofTAA-induced ROS by estimation of protein carbonyla-tion and malondialdehyde (MDA). TAA-induced car-bonylation of proteins in the livers of WT mice, whereasTAA-induced protein carbonylation was inhibited in Tg

mice (Fig. 5A). TAA induced accumulation of MDA inthe livers of WT mice, but not in Tg mice (0.233 � 0.103U/mg and 0.071 � 0.057 U/mg, respectively, P � .01;Fig. 5B). These results indicated that Trx attenuatesTAA-induced acute hepatitis through scavenging ROS.

Suppression of Apoptosis in Tg Mice. We investi-gated whether TAA-induced acute hepatitis is caused bynecrosis or apoptosis. TAA is known to cause pericentralnecrosis by ROS.22 However, we showed that a low doseof TAA (100 �g/g) induces apoptosis in the liver byTUNEL staining and DNA laddering (Fig. 6A, left, andFig. 7). TUNEL-positive cells were observed only in thepericentral areas (Fig. 6A). The distribution of apoptoticcells coincided with the inflammation areas in which thehepatic sinusoidal architecture was destructive, as shown

Fig. 4. Attenuation of acute hepatitis in Tgmice. (A) The levels of transaminases (left,AST; right, ALT) 24 hours after the administra-tion of TAA (100 �g/g) are shown. Significantdifferences were found in the levels of AST andALT between WT (n � 9) and Tg mice (n � 9);*P � .01. Data are expressed as means � SD.Statistical analysis is performed by using Stu-dent’s t test. (B) Histologic study of HE staining.Upper left, nontreated WT mice; upper right,nontreated Tg mice; lower left, TAA-treated WTmice; and lower right, TAA-treated Tg mice.TAA-induced acute hepatitis was observed onlyin the pericentral areas (lower left). Inflamma-tion area was smaller in Tg mice than in WTmice. Original magnification �100.


by TUNEL and HE staining of the serial sections (Fig.6B). However, TUNEL-positive cells were hardly de-tected in the livers of WT mice treated by a high dose ofTAA (500 �g/g), although inflammation caused strik-ingly the destruction of hepatic sinusoidal architecture inthe pericentral areas (Fig. 6C). These results indicatedthat a low dose of TAA induces hepatocellular injurymainly caused by apoptosis in the pericentral areas.

Further, we investigated whether Trx inhibited TAA-induced apoptosis. TUNEL-positive cells were signifi-cantly less in Tg mice than in WT mice (Fig. 6A and B).DNA laddering was strikingly detected in TAA-treatedlivers of WT mice, whereas it was hardly detected in TAA-treated livers of Tg mice (Fig. 7). These results indicatedthat Trx inhibits TAA-induced apoptosis in the pericen-tral areas.

Caspase activities generally are required for apoptosis.We checked whether TAA induces activation of caspasecascade and how Trx affects TAA-induced caspase activ-ities. TAA induced activation of caspase-3 in WT mice 24hours after administration, whereas it was significantlyinhibited in Tg mice (0.143 � 0.008 U/100 �g and

Fig. 5. Suppression of TAA-induced oxidative stress in Tg mice. (A)Changes in protein carbonyl content after the administration of TAA (100�g/g) in WT and Tg mice. The DNP-derivatized protein samples preparedfrom livers subjected to TAA administration were separated on a 15%SDS-PAGE followed by Western blot with primary antibody, specificmoiety of the proteins. Protein carbonylation was inhibited in TAA-treatedTg mice. (B) Lipid peroxidation was measured by the formation of thefluorescent adducts of MDA. The level of MDA in TAA-treated Tg mice(n � 3) was significantly lower than that in TAA-treated WT mice (n �3); *P � .01. Data are expressed as means � SD. Statistical analysisis performed by using Student’s t test.

Fig. 6. Trx inhibited TAA-induced apoptosis. (A) TUNEL stainingshowed that TUNEL-positive cells apparently were observed in thepericentral areas of TAA-treated WT mice (left), whereas these cellswere hardly detected in TAA-treated Tg mice (right). Original magni-fication �40. Arrows indicate the central veins, and arrowheadsindicate the portal veins. (B) Histologic examination of serial sectionsby TUNEL and HE staining. Upper left, TUNEL staining in TAA-treatedWT mice; upper right, HE staining in TAA-treated WT mice; lower left,TUNEL staining in TAA-treated Tg mice; lower right, HE staining inTAA-treated Tg mice. A low dose of TAA induced acute hepatitismainly because of apoptosis, and Trx inhibited TAA-induced apopto-sis. Original magnification �100. Arrows indicate the central veins,and arrowheads indicate the portal veins. (C) Histologic examinationof the livers from WT mice treated by a high dose of TAA (500 �g/g)for 18 hours. Left, TUNEL staining; right, HE staining. A high dose ofTAA (500 �g/g) induced acute hepatitis mainly because of necrosis.Original magnification �100. Arrows indicate the central veins, andarrowheads indicate the portal veins.


0.035 � 0.004 U/100 �g, respectively, P � .01; Fig. 8A).TAA induced activation of caspase-9 in WT mice 12hours after administration, whereas it was significantlyinhibited in Tg mice (0.122 � 0.006 U/100 �g and0.042 � 0.005 U/100 �g, respectively, P � .01; Fig. 8B).In contrast, TAA failed to induce activation of caspase-8in WT mice (control WT mice, 0.029 � 0.01 U/100 �gand TAA-treated WT mice, 0.053 � 0.01 U/100 �g, Fig.8C). We observed no significant difference of caspase-8activity between WT and Tg mice after TAA administra-tion (0.053 � 0.01 U/100 �g and 0.044 � 0.01 U/100�g, respectively, P � .105; Fig. 8C). These results indi-cated that TAA-induced apoptosis is mediated by the mi-tochondrial pathway, not via death receptors.23

The release of cytochrome c from the mitochondriainto the cytosol plays an important role in the mitochon-dria-mediated apoptosis pathway. We checked whetherTAA-induced apoptosis involves the release of cyto-chrome c from the mitochondria to the cytosol. The cy-tosolic level of cytochrome c increased significantly 12hours after TAA administration in WT mice, whereas therelease of cytochrome c was inhibited in Tg mice (Fig. 9).These results indicated that TAA-induced apoptosis ismediated by the release of cytochrome c, and that Trxsuppresses TAA-induced apoptosis partly by inhibition ofcytochrome c release into the cytosol.

Bcl-2 Family Proteins Are Not Related to Anti-Apoptotic Effect of Trx. Bcl-2 family proteins areknown to regulate cytochrome c release.24 Bcl-2 andBcl-xL are an antiapoptotic protein. Bax is a proapoptoticprotein. To see if susceptibility of Tg mice to TAA-in-duced apoptosis is caused by abnormal expression of Bcl-2family proteins, we examined the level of them in the liversamples of TAA-treated WT and Tg mice by Westernblot. There was no difference in the level of Bcl-2, Bcl-xL,and Bax 24 hours after TAA administration between WTand Tg mice (data not shown). These results indicated

that inhibition of cytochrome c release in Tg mice is notdue to the alteration of Bcl-2, Bcl-xL, and Bax expression.

Trx Inhibits LPS Plus GalN-Induced ApoptoticHepatitis. LPS plus GalN-induced liver injury is a well-characterized model for fulminant hepatitis and partlydependent on endogenously produced tumor necrosisfactor � (TNF-�).25 We subjected WT and Tg mice toLPS plus GalN–induced acute liver failure. Survival rateafter administration of LPS plus GalN in WT mice (n �15) was 26.6% (24 hours) and 13.3% (48 hours), whereas

Fig. 8. Caspase activities were inhibited in Tg mice. (A) The catalyticactivity of caspase-3 in liver homogenates of WT and Tg mice 24 hoursafter TAA administration were assayed using the specific substratesDEVD-pNA. Caspase-3 was activated in TAA-treated WT mice, whereas Trxinhibited activation of caspase-3 in Tg mice. (B) The catalytic activity ofcaspase-9 in liver homogenates of WT and Tg mice 12 hours after TAAadministration were assayed using the specific substrates LEHD-pNA.Caspase-9 was activated in TAA-treated WT mice, whereas Trx inhibitedactivation of caspase-9 in Tg mice. (C) The catalytic activity of caspase-8in liver homogenates of WT and Tg mice 12 hours after TAA administra-tion were assayed using the specific substrates IETD-pNA. TAA did notinduce activation of caspase-8 in WT and Tg mice.

Fig. 7. DNA laddering was detected in TAA-treated WT mice. Incontrast, it was negligibly detected in TAA-treated Tg mice.


survival rate in Tg mice (n � 15) was 73.3% (24 hours)and 46.6% (48 hours). There was a significant differencein survival rate after administration of LPS plus GalNbetween WT and Tg mice (P � .0103 by Log-Rank test,P � .0086 by Wilcoxon test; Fig. 10A). LPS plus GalN-induced accumulation of MDA in the liver in WT micewas significantly higher than that in Tg mice (0.221 �0.048 U/mg versus 0.106 � 0.016 U/mg, P � .05; Fig.10C).

Trx Is Not Protective Against Fas-MediatedApoptotic Hepatitis. Fas is an apoptosis-signaling cellsurface antigen that has been shown to trigger cell deathon specific ligand or antibody binding. Treatment of micewith anti-Fas antibody causes fulminant hepatic failurebecause of massive apoptosis.26 To test a protective effectof Trx for Fas-induced apoptotic hepatic failure, we sub-jected WT and Tg mice to Fas-induced acute liver failure.Survival rates after administration of anti-Fas antibody(Jo2) in WT mice (n � 13) was 64.2% (24 hours) and61.5% (48 hours), whereas survival rate in Tg mice (n �14) was 57.1% (24 hours) and 57.1% (48 hours). Therewas no difference in survival rate after administration ofanti-Fas antibody (Jo2) between WT and Tg mice (P �.8019 by Log-Rank test, P � .7856 by Wilcoxon test, Fig.10B). There was no difference in Fas-induced hepaticaccumulation of MDA between WT and in Tg mice(0.130 � 0.014 U/mg and 0.127 � 0.008 U/mg, respec-tively, P � .7838; Fig. 10C).

DiscussionTAA is a thiono-sulfur-containing compound, under-

going an extensive metabolism to acetamide and TAA-S-oxide. Whereas acetamide is devoid of liver-necrotizingproperties, TAA-S-oxide is further metabolized, at least inpart, by cytochrome P-450 monooxygenases to furtherproducts, including a polar product that is thought to bethe sulfene, TAA-S-dioxide, a very reactive compound.27

Therefore, TAA is a typical hepatotoxin and causes cen-trilobular necrosis by generation of ROS. Although TAAis known mainly to cause necrosis in the liver, it has beenreported recently that TAA induced apoptosis in rat liverwithin a few hours after its administration.22 The presentstudy showed that a low dose of TAA induced apoptosisin the pericentral areas, whereas a high dose of TAA in-

Fig. 9. Cytochrome c release was inhibited in Tg mice. Cytochrome cboth in mitochondrial fractions and cytosolic fractions was estimated 12hours after TAA administration by Western blot. TAA induced the releaseof cytochrome c from mitochondria to cytosol, whereas Trx inhibited therelease of cytochrome c in Tg mice.

Fig. 10. Effect of Trx in LPS-induced and Fas-induced acute liverfailure. (A) Survival rates were analyzed until 5 days after LPS (1�g/body) plus GalN (10 mg/body) administration. A significantdifference in survival rates was found between WT (n � 15) and Tgmice (n � 15). Kaplan-Meier survival rates were calculated bylog-rank test and Wilcoxon test. P value by Log-Rank test was shownin the figure. (B) Survival rates were analyzed until 5 days after Jo2(20 �g/body) administration. There was no difference between WT(n � 13) and Tg mice (n � 14). Kaplan-Meier survival rates werecalculated by log-rank test and Wilcoxon test. (C) Lipid peroxidationwas measured by the formation of the fluorescent adducts of MDA.Jo2- or LPS-treated mice were sacrificed 6 hours after intraperitonealinjections. There was no difference in the level of MDA betweenJo2-treated WT mice (n � 3) and Jo2-treated Tg mice (n � 3). Thelevel of MDA in LPS plus GalN-treated Tg mice (n � 3) wassignificantly less than that in LPS plus GalN–treated WT mice (n � 3);*P � .01. Data are expressed as means � SD. Statistical analysisis performed by using Student’s t test.


duced necrosis (Fig. 6). The volume of TAA determineswhether TAA causes necrosis or apoptosis, indicating thatlow level of ROS causes apoptosis, whereas excessive ROScauses necrosis. This result is consistent with our previousreport that 150 to 200 �mol/L of diamide, a thioloxidant,induces apoptosis in Jurkat T cells, whereas at concentra-tions over 400 �mol/L, diamide induces necrosis.28

Moreover, we previously showed that diamide-inducedsulfhydryl oxidation induces apoptosis in Jurkat T cellsand human peripheral blood lymphocyte blasts, whereoxidation of cellular Trx but not glutathione plays a keyrole in apoptosis.29

TAA-induced apoptosis is mediated by mitochondrialpathway (Fig. 8). In general, ROS can trigger mitochon-dria to release caspase-activating proteins, among whichare cytochrome c and possibly other proteins such asapoptosis-inducing factor and intramitochondrialcaspases.30 Trx-inhibited TAA induced apoptosis by theinhibition of cytochrome c release (Fig. 9). Because ROSitself is a trigger to release cytochrome c, antioxidant effectis essential for Trx to inhibit cytochrome c release. Indeed,Trx scavenges TAA-induced ROS (Fig. 5A and B). Wehave reported previously that although ethanol inducesapoptosis by generation of ROS, Trx inhibits ethanol-induced hepatocyte cell damage.31

Further, we examined whether any factors might con-tribute to inhibition of cytochrome c release in Tg mice.Recent evidence has shown that Bcl-2 family proteins actas regulators for apoptosis.32 Bcl-2 and Bcl-xL are anti-apoptotic proteins that function to maintain mitochon-drial membrane integrity and prevent the release ofcytochrome c. The apoptotic mitochondrial permeabilitytransition is mediated by opening of the permeabilitytransition pore complex. Proapoptotic Bax induces themitochondrial permeability transition and cytochrome crelease by directly interacting with the permeability tran-sition pores.33 Malassagne et al reported that mitochon-drial Bcl-2 content was significantly decreased in livers ofmice injected with anti-Fas mAb, and that the superoxidedismutase mimetic MnTBAP prevented Fas-treated liversfrom Bcl-2 down-regulation.34 However, the presentstudy showed that there was no difference in the expres-sion of Bcl-2, Bcl-xL, and Bax between WT and Tg mice(data not shown). Hence, inhibition of cytochrome c re-lease in Tg mice is not caused by Bcl-2, Bcl-xL, and Bax inTAA-induced apoptosis.

The present study showed that Trx inhibited LPS- orTAA-induced apoptosis, but not Fas-induced apoptosis,indicating that there is a different mechanism betweenLPS- or TAA- and Fas-induced apoptosis. LPS plusGalN–induced acute liver failure is partly mediated byTNF-�.25 TNF-� binds death receptor, leading to acti-

vation of caspase-8. Caspase-8 directly activates caspase-3.However, TNF-� also induces the generation of ROS.35

ROS induces cytochrome c release, resulting in activatingcaspase-9. That is, TNF-�–induced apoptosis is mediatedby not only direct activation of caspase cascades but alsoROS-mediated mitochondrial pathway. Indeed, many anti-oxidants and free-radical scavengers inhibit TNF-inducedcytotoxicity in tumor cells, glucocorticoid-, etoposide-,and irradiation-induced apoptosis in thymocytes and glu-tathione-depletion–induced apoptosis in neurons.36

These results indicate that ROS-mediated mitochondrialpathway predominantly contributes to TNF-�–inducedapoptosis.

In contrast, Fas-induced apoptosis is mediated only byactivation of caspase cascade, independent of ROS gener-ation.37 Indeed, many antioxidants and free-radical scav-engers do not inhibit apoptosis induced by Fas orstaurosporine.38 These results indicate that ROS does notcontribute to Fas-induced apoptosis. In Fas-inducedapoptosis, trimerization of the death receptor leads to di-rect activation of caspase-3 by caspase-8. The mitochon-drially mediated caspase-9 activation pathway alsoamplifies Fas signaling through caspase-8-mediated cleav-age of Bid and translocation into the mitochondria, inde-pendent of ROS. However, the contribution of thisamplification pathway to Fas-mediated apoptosis is mostlikely tissue-specific.39 Moreover, deficiency of caspase-9results in compensatory activation of alternate caspasesafter injection of Fas, and fails to protect mice against Fastoxicity.40

Apoptosis signal-regulating kinase 1 (ASK1) is a MAP-KKK that can activate both the stress-activated proteinkinases (SAPKs) and the p38s, promoting apoptosis.41

The carboxyl-terminal noncatalytic domain of ASK1 in-teracts in vivo with TNFR-associated factor 2, an adapterprotein that is required for coupling TNFR1 to theSAPKs. ASK1 also binds and is activated by Daxx, anadapter protein originally thought to relay signals fromFas to the SAPKs. TNF can stimulate the production ofmitochondrial ROS, and Trx is an endogenous ASK1inhibitor that directly binds to the ASK1 aminoterminalnoncatalytic domain and blocks activation of ASK1 byTNF. Trx binding to ASK1 is substantially reversed byROS, suggesting that stimuli such as TNF, which gener-ates ROS, activate ASK1 in part by promoting Trx disso-ciation.42 Hence, there is a possibility that Trx inhibitsROS-mediated apoptosis by blocking ASK1.

Because most overexpressed Trx exists in the cytosol,the present data indicate that Trx shows antioxidant effectin the cytosol and antiapoptotic effect on mainly mito-chondria-mediated pathways. Recently, a mitochondria-specific Trx named Trx-2 was identified.43 Trx-2 also


plays a crucial role in mitochondria-mediated apoptosis.We have recently reported that Trx-2 played a crucial rolein the scavenging ROS in mitochondria and regulatingthe mitochondrial apoptosis signaling pathways using theconditional Trx-2-deficient cells expressing a tetracycline(tet)-repressible TRX-2 transgene in DT 40 cell line.44

More interestingly, Trx-2 inhibits the release of cyto-chrome c through the direct binding to cytochrome c inmitochondria. Hence, the role of Trx-2 in Tg mice is to befurther investigated.

Alcoholic hepatitis also is associated with ROS. In vitroexperiments showed that ethanol induces apoptosis inprimary hepatocytes.45 Previously, we have reported thatrecombinant hTrx suppressed ethanol-induced cytotoxic-ity in cultured hepatocytes.31 Therefore, Trx is effectivefor ROS-mediated hepatitis including alcoholic hepatitis.

In summary, we present evidence showing that Trxinhibits TAA- and LPS-induced acute liver injury, butnot Fas-induced acute liver injury, indicating that Trx iseffective for apoptosis via ROS-mediated mitochondrialpathway. This qualifies Trx as a promising gene therapyand a drug candidate for the treatment and prevention offulminant hepatitis including alcohol and virus infection.

Acknowledgment: The authors thank NorihikoKondo (Institute for Virus Research, Kyoto University)for supplying FLAG-tagged expression vectors, andYong-Won Kwon (Institute for Virus Research, KyotoUniversity), and Dr. Hiroshi Masutani (Institute for Vi-rus Research, Kyoto University) for critical discussion.

References1. Shakil AO, Kramer D, Mazariegos G, Fung JJ, Rakela J. Acute liver failure:

clinical features, outcome analysis, and applicability of prognostic criteria.Liver Transpl 2000;6:163-169.

2. Shaw S, Jayatilleke E, Ross WA, Gordon E, Lieber CS. Ethanol inducedlipid peroxidation: potentiation by long-term alcohol feeding and attenu-ation by methionine. J Lab Clin Med 1981;98:417-424.

3. Kyle ME, Miccadei S, Nakae D, Farber JL. Superoxide dismutase andcatalase protect cultured hepatocytes from the cytotoxicity of acetamino-phen. Biochem Biophys Res Commun 1987;149:889-896.

4. Mayer M, Caselman WH, Schulter V, Schreck R, Hofschneider PH, Bae-uerle PA. Hepatitis B virus transactivator MHBs�: activation of NF-�B,selective inhibition of antioxidants and integral membrane localization.EMBO J 1993;11:2991-2998.

5. Bruck R, Hussein A, Schey R, Matas Z, Reifen R, Zaiger G, Hochman A,et al. Pyrrolidine dithiocarbamate protects against thioacetamide-inducedfulminant hepatic failure in rats. J Hepatol 2002;36:370-377.

6. Holmgren A. Thioredoxin. Annu Rev Biochem 1985;54:237-271.7. Tagaya Y, Maeda Y, Mitsui A, Kondo N, Matsui H, Hamuro J, Brown N,

et al. ATL-derived factor (ADF), an IL-2 receptor/Tac inducer homolo-gous to thioredoxin; Possible involvement of dithiol-reduction in the IL-2receptor induction. EMBO J 1989;8:757-764.

8. Yodoi J, Tursz T. ADF: A growth-promoting factor derived from adultT-cell leukemia and homologous to thioredoxin; Involvement in lympho-cyte immortalization by HTLV-I and EBV. Adv Cancer Res 1991;57:381-411.

9. Nakamura H, Nakamura K, Yodoi J. Redox regulation of cellular activa-tion. Annu Rev Immunol 1997;15:351-369.

10. Das KC, Das CK. Thioredoxin, a singlet oxygen quencher and hydroxylradical scavenger: redox independent functions. Biochem Biophys ResCommun 2000;277:443-447.

11. Chae HZ, Robison K, Poole LB, Church G, Storz G, Rhee SG. Cloningand sequencing of thiol-specific antioxidant from mammalian brain: alkylhydroperoxide reductase and thiol-specific antioxidant define a large fam-ily of antioxidant enzymes. Proc Natl Acad Sci USA 1994;91:7017-7021.

12. Nakamura H, Matsuda M, Furuke K, Kitaoka Y, Iwata S, Toda K, In-amoto T, et al. Adult T cell leukemia-derived factor/human thioredoxinprotects endothelial F-2 cell injury caused by activated neutrophils or hy-drogen peroxide. Immunol Lett 1994;42:75-80.

13. Hirota K, Matsui M, Iwata S, Nishiyama A, Mori K, Yodoi J. AP-1 tran-scriptional activity is regulated by a direct association between thioredoxinand Ref-1. Proc Natl Acad Sci USA 1997;94:3633-3638.

14. Hirota K, Murata M, Sachi Y, Nakamura H, Takeuchi J, Mori K, Yodoi J.Distinct roles of thioredoxin in the cytoplasm and in the nucleus. J BiolChem 1999;274:27891-27897.

15. Takagi Y, Mitsui A, Nishiyama A, Nozaki K, Sono H, Gon Y, HashimotoN, et al. Overexpression of thioredoxin in transgenic mice attenuates focalischemic brain damage. Proc Natl Acad Sci USA 1999;96:4131-4136.

16. Hotta M, Tashiro F, Ikegami H, Niwa H, Ogihara T, Yodoi J, Miyazaki J.Pancreatic � cell-specific expression of thioredoxin, an antioxidative andantiapoptotic protein, prevents autoimmune and streptozotocin-induceddiabetes. J Exp Med 1998;188:1445-1451.

17. Nakamura H, Tamura S, Watanabe I, Iwasaki T, Yodoi J. Enhancedresistancy of thioredoxin-transgenic mice against influenza virus-inducedpneumonia. Immunol Lett 2002;82:165-170.

18. Nishinaka Y, Nakamura H, Okada N, Okada H, Yodoi J. Redox control ofEBV infection: prevention by thiol-dependent modulation of functionalCD21/EBV receptor expression. Antioxid Redox Signal 2001;3:1075-1087.

19. Sumida Y, Nakashima T, Yoh T, Nakajima Y, Ishikawa H, Mitsuyoshi H,Sakamoto Y, et al. Serum thioredoxin levels as an indicator of oxidative stress inpatients with hepatitis C virus infection. J Hepatol 2000;33:616-622.

20. Nakamura H, Herzenberg LA, Bai J, Araya S, Kondo N, Nishinaka Y,Herzenberg LA, et al. Circulating thioredoxin suppresses lipopolysaccha-ride-induced neutrophil chemotaxis. Proc Natl Acad Sci U S A 2001;98:15143-15148.

21. Yang J, Liu X, Bhalla K, Kim N, Ibrado A, Cai J, Peng T, et al. Preventionof apoptosis by Bcl-2: release of cytochrome c from mitochondria blocked.Science 1997;275:1129-1132.

22. Ledda-Columbano GM, Coni P, Curto M, Giacomini L, Faa G, OliverioS, Piacentini M, et al. Induction of two different modes of cell death,apoptosis and necrosis, in rat liver after a single dose of thioacetamide. Am JPathol 1991;139:1099-1109.

23. Fulda S, Wick W, Weller M, Debatin KM. Smac agonists sensitize forApo2L/TRAIL- or anticancer drug-induced apoptosis and induce regres-sion of malignant glioma in vivo. Nat Med 2002;8:808-815.

24. Gross A, McDonnell JM, Korsmeyer SJ. Bcl-2 family members and themitochondria in apoptosis. Genes Dev 1999;13:1899-1911.

25. Oberholzer A, Oberholzer C, Bahjat FR, Edwards CK III, Moldawer LL.Genetic determinants of lipopolysaccharide and D-galactosamine-medi-ated hepatocellular apoptosis and lethality. J Endotoxin Res 2001;7:375-380.

26. Lacronique V, Mignon A, Fabre M, Viollet B, Rouquet N, Molina T,Porteu A, et al. Bcl-2 protects from lethal hepatic apoptosis induced by ananti-Fas antibody in mice. Nat Med 1996;2:80-86.

27. Hunter AL, Holscher MA, Neal RA. Thioacetamide-induced hepatic ne-crosis: Involvement of the mixed function oxidase enzyme system. J Phar-macol Exp Ther 1977;200:439-448.

28. Ueda S, Nakamura H, Masutani H, Sasada T, Yonehara S, Takabayashi A,Yamaoka Y, et al. Redox regulation of caspase-3 protease activity: regula-tory roles of thioredoxin and cytochrome c. J Immunol 1998;161:6689-6695.

29. Sato N, Iwata S, Nakamura K, Hori T, Mori K, Yodoi J. Thiol-mediatedredox regulation of apoptosis: Possible roles of cellular thiols other thanglutathione in T cell apoptosis. J Immunol 1995;154:3194-3203.


30. Green DR, Reed JC. Mitochondria and apoptosis. Science 1998;281:1309-1312.

31. Hirota K, Nakamura H, Arai T, Ishii H, Bai J, Itoh T, Fukuda K, et al.Geranylgeranylacetone enhances expression of thioredoxin and suppressesethanol-induced cytotoxicity in cultured hepatocytes. Biochem BiophysRes Commun 2000;275:825-830.

32. Gross A, McDonnell JM, Korsmeyer SJ. Bcl-2 family members and themitochondria in apoptosis. Genes Dev 1999;13:1899-1911.

33. Narita M, Shimizu S, Ito T, Chittenden T, Lutz RJ, Matsuda H, Tsuji-moto Y. Bax interacts with the permeability transition pore to inducepermeability transition and cytochrome c release in isolated mitochondria.Proc Natl Acad Sci USA 1998;95:14681-14686.

34. Malassagne B, Ferret PJ, Hammoud R, Tulliez M, Bedda S, Trebeden H,Jaffray P, et al. The superoxide dismutase mimetic MnTBAP preventsFas-induced acute liver failure in the mouse. Gastroenterology 2001;121:1451-1459.

35. Schulze-Osthoff K, Krammer PH, Droge W. Divergent signalling viaAPO-1/Fas and the TNF receptor, two homologous molecules involved inphysiological cell death. EMBO J 1994;13:4587-4596.

36. Sandstrom PA, Mannie MD, Buttke TM. Inhibition of activation-induceddeath in T cell hybridomas by thiol antioxidants: oxidative stress as amediator of apoptosis. J Leukoc Biol 1994;55:221-226.

37. Jacobson MD, Raff MC. Programmed cell death and Bcl-2 protection invery low oxygen. Nature 1995;374:814-816.

38. Ratan RR, Murphy TH, Baraban JM. Macromolecular synthesis inhibi-tors prevent oxidative stress-induced apoptosis in embryonic cortical neu-

rons by shunting cysteine from protein synthesis to glutathione. J Neurosci1994;14:4385-4392.

39. Hakem R, Hakem A, Duncan GS, Henderson JT, Woo M, Soengas MS,Elia A, et al. Differential requirement for caspase 9 in apoptotic pathwaysin vivo. Cell 1998;94:339-352.

40. Zheng TS, Hunot S, Kuida K, Momoi T, Srinivasan A, Nicholson DW,Lazebnik Y, et al. Deficiency in caspase-9 or caspase-3 induces compensa-tory caspase activation. Nat Med 2000;6:1241-1247.

41. Ichijo H, Nishida E, Irie K, ten Dijke P, Saitoh M, Moriguchi T, TakagiM, et al. Induction of apoptosis by ASK1, a mammalian MAPKKK thatactivates SAPK/JNK and p38 signaling pathways. Science 1997;275:90-94.

42. Saitoh M, Nishitoh H, Fujii M, Takeda K, Tobiume K, Sawada Y, Kawa-bata M, et al. Mammalian thioredoxin is a direct inhibitor of apoptosissignal-regulating kinase (ASK) 1. EMBO J 1998;17:2596-2606.

43. Spyrou G, Enmark E, Miranda-Vizuete A, Gustafsson J. Cloning andexpression of a novel mammalian thioredoxin. J Biol Chem 1997;272:2936-2941.

44. Tanaka T, Hosoi F, Yamaguchi-Iwai Y, Nakamura H, Masutani H,Ueda S, Nishiyama A, et al. Thioredoxin-2 (TRX-2) is an essential generegulating mitochondria-dependent apoptosis. EMBO J 2002;21:1695-1703.

45. Kurose I, Higuchi H, Miura S, Saito H, Watanabe N, Hokari R, HirokawaM, et al. Oxidative stress-mediated apoptosis of hepatocytes exposed toacute ethanol intoxication. Hepatology 1997;25:368-378.


Overexpression of thioredoxin prevents acute hepatitis caused by thioacetamide or lipopolysaccharide in mice

Documents