- Email: [email protected]
Increased levels of cytosolic thioredoxin reductase activity and mRNA in rat liver nodules
Journal of Hepatology 35 (2001) 259±264
www.elsevier.com/locate/jhep
Increased levels of cytosolic thioredoxin reductase activity and mRNA in rat liver nodules Linda BjoÈrkhem, Habtemichael Teclebrhan, Emine Kesen, Jerker M. Olsson, Lennart C. Eriksson, Mikael BjoÈrnstedt* Department of Microbiology, Pathology and Immunology, Division of Pathology, Karolinska Institutet, Huddinge University Hospital, F 46, SE-141 86, Stockholm, Sweden
Background/Aims: Thioredoxin reductase, a redox active enzyme, is induced in several tumors. This study focuses on the presence of and subcellular localisation of thioredoxin reductase in a tumor model where neoplastic lesions are selected by their resistance to the toxic effects of the promotor. Methods: Liver nodules produced by intermittent feeding of 2-acetylamino¯uorene to male Wistar rats were analyzed for thioredoxin reductase (TrxR) activity and mRNA. Results: This activity was increased 3.5-fold in the cytosol but decreased 60% in the mitochondrial fraction compared to the liver of age-matched untreated animals. Only traces of activity were observed in the microsomal, plasma membrane and nuclear fractions from normal liver or nodules. The level of TrxR mRNA was 3-fold higher in nodules than in normal rat liver. Furthermore, the total level of SH groups in homogenates was 2-fold higher in the case of the nodules. Conclusions: These ®ndings indicate that the thioredoxin system makes an important contribution to the resistant phenotype of the neoplastic liver cell, which conveys a growth advantage of signi®cance for tumor progression. q 2001 European Association for the Study of the Liver. Published by Elsevier Science B.V. All rights reserved. Keywords: Thioredoxin reductase; Hepatocarcinogenesis; Rat liver nodules; 2-Acetylamino¯uorene
1. Introduction In the resistant hepatocyte model the carcinogenic process begins with the formation of a so-called initiated cell, which demonstrates resistance to the toxic and mitoinhibitory effects of a promoter. Thus, in the presence of a promoter, initiated cells are selected and a clone of cells exhibiting pronounced resistance to toxic factors such as oxidative stress is formed. This increased resistance provides the cell with a growth advantage, i.e. a higher potential for survival in a toxic environment. Nodules exhibiting these properties arise in response to intermittent feeding with 2-acetylamino¯uorene (2-AAF), using well-characterized protocols for inducing chemical hepatocarcinogenesis in rats [1,2]. The thioredoxin system, consisting of thioredoxin (Trx), Received 28 February 2001; received in revised form 5 April 2001; accepted 24 April 2001 * Corresponding author. Tel.: 146-8-58583809; fax: 146-8-58581020. E-mail address: [email protected] (M. BjoÈrnstedt).
thioredoxin reductase (TrxR) and NADPH functions as an ef®cient general reductase for protein disul®de bonds, thereby playing an important role in a variety of cellular processes, including the formation of precursors for DNA synthesis, defenses against oxidative stress and maintenance of the generally reducing intracellular environment [3,4]. Trx is a ubiquitous 12 kDa protein originally characterized as a hydrogen donor to ribonucleotide reductase but a large number of other functions for this protein have subsequently been discovered. Through the reversible formation of disul®de dithiols involving the sulfur atoms in the side chains of critical cysteine residues, Trx may regulate the activities of enzymes and transcription factors [4]. Such covalent modi®cation might thus be analogous to phosphorylation/dephosphorylation. The mammalian TrxR belong to a family of pyridine nucleotide-disul®de oxidoreductases demonstrating sequence homologies, e.g. a conserved Cys-Val-Asn-Val-GlyCys sequence at the catalytic center, a sequence also present in glutathione reductase [5]. Other members of this family
0168-8278/01/$20.00 q 2001 European Association for the Study of the Liver. Published by Elsevier Science B.V. All rights reserved. PII: S 0168-827 8(01)00113-1
260
L. BjoÈrkhem et al. / Journal of Hepatology 35 (2001) 259±264
include lipoamide dehydrogenase and mercuric ion reductase. All these enzymes are homodimeric proteins containing ¯avine adenine dinucleotide (FAD) as a prosthetic group, an NADPH-binding site and a catalytically active disul®de/dithiol group at the active center [5]. In contrast to the other pyridine nucleotide-disul®de oxidoreductases, each of the two subunits of mammalian TrxR contains a penultimate C-terminal selenocysteine residue [6]. Mammalian TrxR exhibits exceptionally broad substrate speci®cities, reducing not only Trx from the same but also distant species, as well as a variety of low molecular weight substrates, including selenium compounds, nitrosoglutathione, dithionitrobenzoicacid (DTNB) and hydroperoxides [7]. This uniquely broad substrate speci®city can be explained by the high nucleophilicity of the selenocysteine residue. This broad substrate speci®city, along with the fact that TrxR functions both with and without Trx as a general reductant in the cell to reduce harmful hydroperoxides [8], makes this protein highly interesting in connection with studies on resistant preneoplastic and neoplastic cells which arise during the process of chemical carcinogenesis. Thus, the aim of the present study was to characterize possible changes in the activity and corresponding mRNA expression as well as the subcellular localization of TrxR in such cells.
2. Materials and methods 2.1. Chemicals Trx from E. coli was purchased from Promega. Bovine TrxR (used for constructing standard curves) was obtained from IMCO Corp. (Sweden). All other chemicals employed for enzyme assays were procured from Sigma. The basal rat diet R36 was bought from Lactamin (Sweden). 2-AAF (Fluca, Switzerland) was mixed with the modi®ed Altromine1314 diet (Altromine, Germany) which resembles the basal diet R36. All molecular biological reagents and kits, with the exception of the ®rst-strand cDNA Synthesis Kit (Boehringer Mannheim), were purchased from Life Technologies.
2.2. Animal experiments and isolation of tissues The male Wistar rats (Mùllegaard Breeding Laboratory, Ejby, Denmark) weighed 100 g at the time treatment with 2-AAF (0.05%, w/w) was initiated. The 25 week ad libitum feeding schedule suggested by Epstein et al. [1] and modi®ed by Eriksson et al. [2] was employed in order to obtain persistent liver nodules. Prior to harvesting, the rats were fed basal diet free from 2-AAF for at least 4 weeks and then starved for 24 h. The rats were subsequently sacri®ced and homogeneous nodules were removed with a pair of tweezers. Livers from age-matched animals receiving basal diet alone during the feeding regimen were employed as the control tissue. The liver nodules are regarded as persistent, premalignant neoplastic lesions. They are not dependent on the presence of promoter to express their growth advantage but are not invasive malignant hepatocellular carcinomas. This study was approved by the Southern Stockholm Committee for Ethical Review of Animal Experiments.
2.3. Preparation of homogenates and subcellular fractions The different tissue samples were diced with a pair of scissors in cold 0.25 M sucrose and homogenized in a Potter-Elvehjelm glass-Te¯on homogenizer with 4 up-and-down strokes at a speed of 440 rev./min. More sucrose was then added to this homogenate to obtain 1 g tissue (wet weight)/10 ml. Cell debris was removed by centrifugation at 350 £ g for 10 min. The resulting supernatant was subsequently centrifuged at 2800 £ g for 20 min and the mitochondrial pellet thus obtained was washed three times with 0.25 M sucrose and ®nally resuspended in the same medium at a concentration corresponding to 1 g liver/ml [9]. Thereafter, microsomes and cytosolic fractions were prepared as described elsewhere [10]. The nuclear fraction was isolated according to Blobel and Potter [11]. For this purpose, rats were anesthetized by i.v. administration of Hypnorm (0.15 ml/100 g body weight) and Pentobarbital (0.15 ml/100 g body weight). The liver was then perfused in the retrograde direction with 0.25 M sucrose (22 ml/min in 48C) applied through the caval vein and leaving through the portal vein, until the organ was cold and devoid of blood. The liver was subsequently homogenized in a solution containing 0.25 M sucrose, 50 mM Tris±Cl (pH 7.5), 25 mM KCl and 5 mM MgCl2 (STKM) at 440 rev./min using ten up-and-down strokes and the homogenate was thereafter adjusted to contain 2 g liver/10 ml. One portion of this homogenate was mixed with twice as much 75% STKM solution and 28 ml of this mixture was layered on top of 10 ml 75% STKM in a centrifuge tube. After centrifugation in the swinging bucket SW 28 rotor at 25 000 rev./min for 60 min, the pellet was resuspended in STKM, this suspension was centrifuged again at 3000 rev./min in a JA 20 rotor for 10 min and ®nally the pellet thus obtained was suspended in STKM to give a ®nal concentration corresponding to 1 g liver/ml. Plasma membranes (PLM) were prepared as described earlier [12]. Protein concentrations were determined by the Biuret procedure using bovine serum albumin as standard.
2.4. Enzyme assays 2.4.1. TrxR activity
The activities of TrxR in homogenates and subcellular fractions was determined using insulin disul®des as substrate, essentially as described by Holmgren and BjoÈrnstedt [3]. Brie¯y, 50 mg of sample protein was incubated with 80 mM HEPES (pH 7.5), 0.9 mg/ml NADPH, 6 mM EDTA, 2 mg/ml insulin and 10 mM E. coli Trx at 378C for 20 min in a total volume of 120 ml. For each sample, a blank containing everything except Trx was incubated and treated in the same manner. The reaction was terminated by the addition of 500 ml DTNB (0.4 mg/ml) and 6 M guanidine hydrochloride in 0.2 M Tris±Cl (pH 8.0). The absorbance at 412 nm was subtracted from the corresponding absorbance of the sample. By comparison with a standard curve prepared employing puri®ed calf thymus TrxR (IMCO, Stockholm, Sweden), levels of TrxR in the various homogenates and subcellular fractions could be determined.
2.4.2. Redox status
In order to quantitate the total relative amounts of thiols (SH groups), both native and enzymatically formed, present in the homogenates of nodules and control liver, samples were incubated in the same reaction mixture as described above at 378C for 20 min. Thereafter, the reaction was terminated and the absorbance at 412 nm was measured also as described above. The concentration of SH groups was calculated using E412nm 13:6 mM 21.
2.5. Isolation of RNA Rat liver samples were removed and immediately frozen in liquid nitrogen. Subsequently, tissue samples weighing 50±100 mg were homogenized in a total volume of 1 ml TRIZOL Reagent using a POLYTRON, and total RNA was isolated as described in the TRIZOL Reagent (Life Technologies) Kit. To ensure inactivation of contaminating RNAses, metal objects and
L. BjoÈrkhem et al. / Journal of Hepatology 35 (2001) 259±264
Fig. 1. Active TrxR (ng/mg total protein) in homogenate from normal liver and liver nodules. The values are the mean ^ SD of six repeat assays performed on homogenate from three independent preparations. ***P , 0:001 compared to the control value. glassware were cleaned with detergent, rinsed thoroughly with RNAse-free water and ®nally heated at 2008C for 6 h. Plastic items were rinsed with 0.1 M NaOH containing 1 mM EDTA, followed by RNAse-free water, and thereafter heated at 1208C for 6 h. 0.2% diethyl pyrocarbonate (DEPC) was added to water and aqueous solutions. The quality and concentration of the RNA preparations thus obtained were ascertained by measurement of their absorbance at 260 and 280 nm. Possible contaminating remnants of genomic DNA were eliminated by treating these preparations with deoxyribonuclease I (ampli®cation grade) prior to RT-PCR ampli®cation.
2.6. RT-PCR Total RNA was subjected to oligo(dT)-primed ®rst-strand cDNA synthesis in a total volume of 20 ml using the 1st Strand cDNA Synthesis Kit, according to the protocol provided by the manufacturer (Boehringer Mannheim). The transcription products thus obtained were then subjected to PCR ampli®cation in a total volume of 100 ml using an automatic thermal cycler (HYBAID). The step-cycle program employed involved denaturation at 948C for 1 min, annealing at 648C for 1 min and extension at 728C for
261
1.5 min for a total of 35 cycles. Prior to initiation of this process, the mixture was denatured for 3 min at 948C; and following after the last cycle an extension reaction was performed for 7 min at 728C. In order to achieve semi-quantitative analysis, primer pairs for the `housekeeping gene' encoding b-actin were simultaneously ampli®ed under identical conditions for use as an internal standard. Equal volumes of the PCR products for TrxR and b-actin were mixed and co-electrophoresed on a 1.5% agarose gel, after which visualization was achieved by ethidium bromide staining. The PCR products were quantitated on the basis of the intensities of the bands, as photographed by a Kodak DC40 camera under UV-light and examined using a computerized digital image analysis program (Kodak Digital Science 1D Image analysis software, Eastman Kodak Co., Rochester, NY). The number of cycles, MgCl2 concentration and annealing temperatures employed were each optimized. Appropriate primer pairs for RT-PCR were constructed on the basis of the gene sequence for rat TrxR (GenBank accession no. U63923), i.e. nucleotides 1077±1100 (5 0 -CCTGTCACGGATGAGGAGCAGACC-3 0 ) were used as the forward primer and nucleotides 2105±2082 (5 0 -GCCATCTGACCCCCTTCCACACAG-3 0 ) were used as the reverse primer. The primers for RT-PCR of b-actin mRNA were designed to amplify 285 bp DNA and were purchased from SDS Promega. Statistical analysis was performed using Student's t-test.
3. Results 3.1. Activity of TrxR in homogenates and subcellular fractions The activity of TrxR in homogenates of liver nodules was 4-fold greater than in normal liver (Fig. 1). Corresponding measurements on subcellular fractions revealed the presence of TrxR activity in the cytosolic and mitochondrial fractions from both control and nodular hepatic tissues (Fig. 2). In the cytosolic fractions, these activities were 250 and 900 ng TrxR/mg protein, respectively. However, in the
Fig. 2. Active TrxR (ng/mg total protein) in cytosol, microsomes, mitochondrias, PLM, and nuclei from normal liver (C) and liver nodules (N). The values are the mean ^ SD of six repeat assays performed on each fraction from three independent preparations. ***P , 0:001 compared to the control value.
262
L. BjoÈrkhem et al. / Journal of Hepatology 35 (2001) 259±264
Fig. 3. Active TrxR (mg/g liver) in homogenate, cytosol, microsomes, mitochondrias, PLM, and nuclei from normal liver (C) and liver nodules (N). The values are the mean ^ SD of six repeat assays performed on each fraction from three independent preparations. ***P , 0:001 compared to the control value.
mitochondrial fractions TrxR activity was higher in control tissue than in nodules (230 versus 140 ng TrxR/mg protein, respectively). Only traces of TrxR activity could be detected in the microsomal, PLM and nuclear fractions from both types of tissue. Fig. 3 depicts the total activity of TrxR per gram of liver in the various subfractions, showing that, by far, most of the cellular TrxR activity is cytosolic. The mitochondrial activity is less than 20% of the total in normal liver and ,5% in liver nodules.
Fig. 4. Semi-quantitative analysis of the level of TrxR mRNA employing RT-PCR. Total RNA was isolated and electrophoresed on 1.5% agarose gel as described in Section 2. b-Actin was used as an internal standard. Lanes 1, 3, 5 and 7 are from normal liver and lanes 2, 4, 6 and 8 are from 2-AAF-induced nodules. Total RNA (100 ng) was applied to lanes 1±4 and 200 ng was applied to lanes 5±8. M, molecular weight markers separated by 250 bp.
3.2. Levels of TrxR mRNA When a semi-quantitative RT-PCR technique was employed to determine the levels of TrxR mRNA in liver nodules and normal liver [13], the nodular level was found to be signi®cantly higher (Fig. 4). When RT-PCR was performed on 100 ng total RNA, no TrxR mRNA was detectable in normal liver, whereas the nodular level was 80% of the level of b-actin mRNA. With 200 ng total RNA, the TrxR mRNA level in liver nodules was seen to be 3-fold higher than in normal liver. In this latter case, the intensity of the TrxR mRNA band in normal liver was 27% of the band intensity for b-actin mRNA, while the corresponding value for nodules was 79%.
Fig. 5. The level of thiol groups (nmol SH groups/mg homogenate protein) in homogenates from normal liver and liver nodules. The values shown are the mean ^ SD of six repeat assays performed on homogenate from three independent preparations. **P , 0:01 compared to the control value.
L. BjoÈrkhem et al. / Journal of Hepatology 35 (2001) 259±264
3.3. Redox status in normal liver and liver nodules The redox status in homogenates from normal liver and liver nodules was determined by measuring the total level of SH groups following a 20 min incubation in the presence of NADPH. The level of thiol groups in homogenate from nodules was double (840 nmol/mg protein) compared to control liver (420 nmol/mg protein) (Fig. 5).
4. Discussion In this study elevations in the activity and mRNA level of TrxR in rat liver nodules have been demonstrated. The activity of TrxR in homogenates from nodules was 4-fold greater than in control homogenate. The level of TrxR mRNA was also elevated in nodular cells compared to normal hepatocytes, indicating that the increase in TrxR activity is due to transcriptional up-regulation and/or mRNA stabilization. The advantage of the activity assay employed here (i.e. reduction of insulin disul®des in the presence of excess Trx) is that the total activity of all forms of TrxR is detected, as opposed to the detection of individual proteins using immunoblotting. In subcellular fractions, TrxR activity was increased in the cytosolic but decreased in the mitochondrial fraction from nodules. Mitochondrial TrxR is a distinct enzyme (referred to as TrxR2) and has been isolated from humans [14,15] and rats [16]. The primary function of TrxR2 is still unknown but a role in protecting against oxidative stress has been suggested [17]. Several reports have demonstrated elevated TrxR activity in tumor cells [6,18±20]. Furthermore, treatment of mouse skin with several different tumor promoters resulted in increased activity of TrxR as well as in increased levels of Trx and glutaredoxin in this tissue [21]. However, no distinction was made between the cytosolic TrxR and the mitochondrial TrxR2 in these earlier studies. Our present ®ndings of increased activity in the cytosol and decreased activity in the mitochondrias of nodules indicate that the cytosolic enzyme ®lls a function different than that of the mitochondrial enzyme in the resistant cell. Only traces of activity were present in the microsomes (endoplasmic reticulum, ER) and PLM and in these subfractions there were no differences between nodules and control liver. In a previous study in which TrxR in different organs of the adult rat was observed immunohistochemically, the ¯uorescence of anti-TrxR antibodies was reported to be primarily cytosolic [22]. In another study of rat hepatocytes, involving electron microscopic immunohistochemistry in combination with the immunogold technique, the TrxR protein was found to be localized not only in the cytosol but also in association with the granular ER, in the cistern of the Golgi complex, in the chromatin of the nucleus and at the periphery of mitochondria [23]. The different techniques employed as well as differences in speci®c TrxR activity in
263
various organelles probably explain the subtle discrepancies in the results obtained utilizing immunochemical procedures and enzymatic assays. It is expected that the protein is detected also where the protein is synthesized and processed and not only where it has its activity. 2-AAF, originally used as an insecticide, is a well-known hepatocarcinogen [24]. This molecule is highly hydrophobic and diffuses easily through biological membranes, as a result of which it is initially enriched in lipid-rich compartments of the body. The parent compound is metabolized by the microsomal cytochrome P450 system to various hydroxylated metabolites, of which the most reactive is N-OH-2acetylamino¯uorene. This compound is then conjugated enzymatically, primarily to glucuronic acid, glutathione and sulfate. Whereas the other conjugates are excreted, the sulfate ester of N-OH-2-acetylamino¯uorene is an especially labile compound that spontaneously forms the ultimate reactive metabolite, the nitrenium ion. The nitrenium ion radical reacts with cellular macromolecules to give rise to the toxic effects of the compound. Resistant nodular cells also contain high levels of antioxidants and are resistant to various sources of oxidative stress, including those unrelated to 2-AAF. This phenotype is constitutively expressed by persistent nodules even in the absence of promoter. Rather than being speci®c for 2-AAF, it is probable that increased TrxR activity is a part of the resistant phenotype that develops during tumor development and characterizes the process of tumor development in general. The mammalian TrxR reduce a wide variety of low molecular weight substrates and are above all ef®cient peroxidases, reducing hydrogen peroxide [25] and, even more interestingly, lipid hydroperoxides [7]. In addition, Trx itself reduces hydrogen peroxide, organic hydroperoxides and cellular disul®des formed in connection with oxidative stress. Since the Trx system plays an important role in DNA synthesis, increased TrxR activity might enhance the rate of ribonucleotide reduction. Oxidative stress has been suggested to activate the regulatory protein oxy-R, which results in overexpression of alkyl hydroperoxide reductase, a TrxR-like protein [26]. We suggest that TrxR activity is of great signi®cance for cellular resistance, since overexpression of this protein results in more ef®cient defense systems. The level of SH groups, both endogenous and enzymatically formed, in homogenates from nodules was 2-fold higher than the corresponding value for normal rat liver. This ®nding demonstrates that nodules possess a greater reducing potential and thereby a greater capacity to withstand oxidative stress. Our results clearly demonstrate that 2-AAF-induced resistant rat liver nodules contain an enhanced reducing potential and increased activity of TrxR, a key enzyme for defense and growth. Further experiments focused on the regulation of selenoproteins in connection with chemical carcinogenesis are currently in progress in our laboratory.
264
L. BjoÈrkhem et al. / Journal of Hepatology 35 (2001) 259±264
Acknowledgements The excellent technical assistance of Mrs Ulla-Britta Torndal is gratefully acknowledged. This investigation was supported by grants from the Swedish Cancer Society, The Swedish Medical Association and Karolinska Institutet. References [1] Epstein S, Ito N, Merkow L, Farber E. Cellular analysis of liver carcinogenesis: the induction of larger hyperplastic nodules in the liver with 2-¯ourenylacetamide or ethionine and some aspects of their morphology and glycogen metabolism. Cancer Res 1967;27:1702. [2] Eriksson LC, Torndal UB, Andersson GN. Isolation and characterization of endoplasmic reticulum and Golgi apparatus from hepatocyte nodules in male Wistar rats. Cancer Res 1983;43:3335±3347. [3] Holmgren A, BjoÈrnstedt M. Thioredoxin and thioredoxin reductase. Methods Enzymol 1995;252:199±208. Ê slund F, BjoÈrnstedt M, Zhong L, Ljung J, et [4] Holmgren A, ArneÂr E, A al. Redox regulation by the thioredoxin and glutaredoxin systems. In: Montagnier L, Olivier R, Pasquier C, et al., editors. Oxidative stress in cancer, aids, and neurodegenerative diseases, Paris: Marcel Dekker, 1996. pp. 229±246. [5] Williams CHJ. Lipoamide dehydrogenase, glutathione reductase, thioredoxin reductase and mercuric ion reductase ± a family of ¯avoenzyme transhydrogenase. In: MuÈller F, editor. Chemistry and biochemistry of ¯avoenzyme, 3. Boca Raton, FL: CRC Press, 1992. pp. 121±211. [6] Tamura T, Stadtman TC. A new selenoprotein from human lung adenocarcinoma cells: puri®cation, properties, and thioredoxin reductase activity. Proc Natl Acad Sci USA 1996;93:1006±1011. [7] BjoÈrnstedt M, Hamberg M, Kumar S, Xue J, Holmgren A. Human thioredoxin reductase directly reduces lipid hydroperoxides by NADPH and selenocystine strongly stimulates the reaction via catalytically generated selenols. J Biol Chem 1995;270:11761±11764. Ê kesson B, Holmgren A. The thior[8] BjoÈrnstedt M, Xue J, Huang W, A edoxin and glutaredoxin systems are ef®cient electron donors to human plasma glutathione peroxidase. J Biol Chem 1994;269:29382±29384. [9] Sottocasa JL, Kuylenstierna B, Ernster L, Bergstrand A. Separation and some enzymatic properties of the inner and outer membranes of rat liver mitochondria. Methods Enzymol 1967;10:448±463. [10] Dallner G. Isolation of rough and smooth microsomes ± general. Methods Enzymol 1974;31:191±201. [11] Blobel G, Potter VR. Nuclei from rat liver: isolation method that combines purity with high yield. Science 1966;154:1662±1665.
[12] Loten EG, Redshaw-Loten JC. Preparation of rat liver plasma membranes in a high yield. Anal Biochem 1986;154:183±185. [13] Serazin-Leroy V, Denis-Henriot D, Morot M, de Mazancourt P, Giudicelli Y. Semi-quantitative RT-PCR for comparison of mRNAs in cells with different amounts of housekeeping gene transcripts. Mol Cell Probes 1998;12:283±291. [14] Gasdaska PY, Berggren MM, Berry MJ, Powis G. Cloning, sequencing and functional expression of a novel human thioredoxin reductase. FEBS Lett 1999;442:105±111. [15] Miranda-Vizuete A, Damdimopoulos AE, Pedrajas JR, Gustafsson Ê , Spyrou G. Human mitochondrial thioredoxin reductase cDNA JA cloning, expression and genomic organization. Eur J Biochem 1999;261:405±412. [16] Lee SR, Kim JR, Kwon KS, Yoon HW, Levine RL, Ginsburg A, Rhee SG. Molecular cloning and characterization of a mitochondrial selenocysteine-containing thioredoxin reductase from rat liver. J Biol Chem 1999;274:4722±4734. [17] Miranda-Vizuete A, Damdimopoulos AE, Spyrou G. cDNA cloning, expression and chromosomal localization of the mouse mitochondrial thioredoxin reductase gene(1). Biochim Biophys Acta 1999;1447:113±118. [18] Sun X, Dobra K, BjoÈrnstedt M, Hjerpe A. Upregulation of 9 genes, including that for thioredoxin, during epithelial differentiation of mesothelioma cells. Differentiation 2000;66:181±188. [19] Berggren M, Gallegos A, Gasdaska JR, Gasdaska PY, Warneke J, Powis G. Thioredoxin and thioredoxin reductase gene expression in human tumors and cell lines, and the effects of serum stimulation and hypoxia. Anticancer Res 1996;16:3459±3466. [20] Powis G, Gasdaska JR, Gasdaska PY, Berggren M, Kirkpatrick DL, Engman L, Cotgreave IA, et al. Selenium and the thioredoxin redox system: effects on cell growth and death. Oncol Res 1997;9:303±312. [21] Kumar S, Holmgren A. Induction of thioredoxin, thioredoxin reductase and glutaredoxin activity in mouse skin by TPA, a calcium ionophore and other tumor promoters. Carcinogenesis 1999;20:1761± 1767. [22] Rozell B, Hansson HA, Luthman M, Holmgren A. Immunohistochemical localization of thioredoxin and thioredoxin reductase in adult rats. Eur J Cell Biol 1985;38:79±86. [23] Rozell B, Holmgren A, Hansson HA. Ultrastructural demonstration of thioredoxin and thioredoxin reductase in rat hepatocytes. Eur J Cell Biol 1988;46:470±477. [24] Floyd RA. Free radicals and cancer. New York: Marcel Dekker, 1982. [25] Zhong L, Holmgren A. Essential role of selenium in the catalytic activities of mammalian thioredoxin reductase revealed by characterization of recombinant enzymes with selenocysteine mutations. J Biol Chem 2000;275:18121±18128. [26] Storz G, Tartaglia LA, Ames BN. The OxyR regulon. Antonie Van Leeuwenhoek 1990;58:157±161.
www.elsevier.com/locate/jhep
Increased levels of cytosolic thioredoxin reductase activity and mRNA in rat liver nodules Linda BjoÈrkhem, Habtemichael Teclebrhan, Emine Kesen, Jerker M. Olsson, Lennart C. Eriksson, Mikael BjoÈrnstedt* Department of Microbiology, Pathology and Immunology, Division of Pathology, Karolinska Institutet, Huddinge University Hospital, F 46, SE-141 86, Stockholm, Sweden
Background/Aims: Thioredoxin reductase, a redox active enzyme, is induced in several tumors. This study focuses on the presence of and subcellular localisation of thioredoxin reductase in a tumor model where neoplastic lesions are selected by their resistance to the toxic effects of the promotor. Methods: Liver nodules produced by intermittent feeding of 2-acetylamino¯uorene to male Wistar rats were analyzed for thioredoxin reductase (TrxR) activity and mRNA. Results: This activity was increased 3.5-fold in the cytosol but decreased 60% in the mitochondrial fraction compared to the liver of age-matched untreated animals. Only traces of activity were observed in the microsomal, plasma membrane and nuclear fractions from normal liver or nodules. The level of TrxR mRNA was 3-fold higher in nodules than in normal rat liver. Furthermore, the total level of SH groups in homogenates was 2-fold higher in the case of the nodules. Conclusions: These ®ndings indicate that the thioredoxin system makes an important contribution to the resistant phenotype of the neoplastic liver cell, which conveys a growth advantage of signi®cance for tumor progression. q 2001 European Association for the Study of the Liver. Published by Elsevier Science B.V. All rights reserved. Keywords: Thioredoxin reductase; Hepatocarcinogenesis; Rat liver nodules; 2-Acetylamino¯uorene
1. Introduction In the resistant hepatocyte model the carcinogenic process begins with the formation of a so-called initiated cell, which demonstrates resistance to the toxic and mitoinhibitory effects of a promoter. Thus, in the presence of a promoter, initiated cells are selected and a clone of cells exhibiting pronounced resistance to toxic factors such as oxidative stress is formed. This increased resistance provides the cell with a growth advantage, i.e. a higher potential for survival in a toxic environment. Nodules exhibiting these properties arise in response to intermittent feeding with 2-acetylamino¯uorene (2-AAF), using well-characterized protocols for inducing chemical hepatocarcinogenesis in rats [1,2]. The thioredoxin system, consisting of thioredoxin (Trx), Received 28 February 2001; received in revised form 5 April 2001; accepted 24 April 2001 * Corresponding author. Tel.: 146-8-58583809; fax: 146-8-58581020. E-mail address: [email protected] (M. BjoÈrnstedt).
thioredoxin reductase (TrxR) and NADPH functions as an ef®cient general reductase for protein disul®de bonds, thereby playing an important role in a variety of cellular processes, including the formation of precursors for DNA synthesis, defenses against oxidative stress and maintenance of the generally reducing intracellular environment [3,4]. Trx is a ubiquitous 12 kDa protein originally characterized as a hydrogen donor to ribonucleotide reductase but a large number of other functions for this protein have subsequently been discovered. Through the reversible formation of disul®de dithiols involving the sulfur atoms in the side chains of critical cysteine residues, Trx may regulate the activities of enzymes and transcription factors [4]. Such covalent modi®cation might thus be analogous to phosphorylation/dephosphorylation. The mammalian TrxR belong to a family of pyridine nucleotide-disul®de oxidoreductases demonstrating sequence homologies, e.g. a conserved Cys-Val-Asn-Val-GlyCys sequence at the catalytic center, a sequence also present in glutathione reductase [5]. Other members of this family
0168-8278/01/$20.00 q 2001 European Association for the Study of the Liver. Published by Elsevier Science B.V. All rights reserved. PII: S 0168-827 8(01)00113-1
260
L. BjoÈrkhem et al. / Journal of Hepatology 35 (2001) 259±264
include lipoamide dehydrogenase and mercuric ion reductase. All these enzymes are homodimeric proteins containing ¯avine adenine dinucleotide (FAD) as a prosthetic group, an NADPH-binding site and a catalytically active disul®de/dithiol group at the active center [5]. In contrast to the other pyridine nucleotide-disul®de oxidoreductases, each of the two subunits of mammalian TrxR contains a penultimate C-terminal selenocysteine residue [6]. Mammalian TrxR exhibits exceptionally broad substrate speci®cities, reducing not only Trx from the same but also distant species, as well as a variety of low molecular weight substrates, including selenium compounds, nitrosoglutathione, dithionitrobenzoicacid (DTNB) and hydroperoxides [7]. This uniquely broad substrate speci®city can be explained by the high nucleophilicity of the selenocysteine residue. This broad substrate speci®city, along with the fact that TrxR functions both with and without Trx as a general reductant in the cell to reduce harmful hydroperoxides [8], makes this protein highly interesting in connection with studies on resistant preneoplastic and neoplastic cells which arise during the process of chemical carcinogenesis. Thus, the aim of the present study was to characterize possible changes in the activity and corresponding mRNA expression as well as the subcellular localization of TrxR in such cells.
2. Materials and methods 2.1. Chemicals Trx from E. coli was purchased from Promega. Bovine TrxR (used for constructing standard curves) was obtained from IMCO Corp. (Sweden). All other chemicals employed for enzyme assays were procured from Sigma. The basal rat diet R36 was bought from Lactamin (Sweden). 2-AAF (Fluca, Switzerland) was mixed with the modi®ed Altromine1314 diet (Altromine, Germany) which resembles the basal diet R36. All molecular biological reagents and kits, with the exception of the ®rst-strand cDNA Synthesis Kit (Boehringer Mannheim), were purchased from Life Technologies.
2.2. Animal experiments and isolation of tissues The male Wistar rats (Mùllegaard Breeding Laboratory, Ejby, Denmark) weighed 100 g at the time treatment with 2-AAF (0.05%, w/w) was initiated. The 25 week ad libitum feeding schedule suggested by Epstein et al. [1] and modi®ed by Eriksson et al. [2] was employed in order to obtain persistent liver nodules. Prior to harvesting, the rats were fed basal diet free from 2-AAF for at least 4 weeks and then starved for 24 h. The rats were subsequently sacri®ced and homogeneous nodules were removed with a pair of tweezers. Livers from age-matched animals receiving basal diet alone during the feeding regimen were employed as the control tissue. The liver nodules are regarded as persistent, premalignant neoplastic lesions. They are not dependent on the presence of promoter to express their growth advantage but are not invasive malignant hepatocellular carcinomas. This study was approved by the Southern Stockholm Committee for Ethical Review of Animal Experiments.
2.3. Preparation of homogenates and subcellular fractions The different tissue samples were diced with a pair of scissors in cold 0.25 M sucrose and homogenized in a Potter-Elvehjelm glass-Te¯on homogenizer with 4 up-and-down strokes at a speed of 440 rev./min. More sucrose was then added to this homogenate to obtain 1 g tissue (wet weight)/10 ml. Cell debris was removed by centrifugation at 350 £ g for 10 min. The resulting supernatant was subsequently centrifuged at 2800 £ g for 20 min and the mitochondrial pellet thus obtained was washed three times with 0.25 M sucrose and ®nally resuspended in the same medium at a concentration corresponding to 1 g liver/ml [9]. Thereafter, microsomes and cytosolic fractions were prepared as described elsewhere [10]. The nuclear fraction was isolated according to Blobel and Potter [11]. For this purpose, rats were anesthetized by i.v. administration of Hypnorm (0.15 ml/100 g body weight) and Pentobarbital (0.15 ml/100 g body weight). The liver was then perfused in the retrograde direction with 0.25 M sucrose (22 ml/min in 48C) applied through the caval vein and leaving through the portal vein, until the organ was cold and devoid of blood. The liver was subsequently homogenized in a solution containing 0.25 M sucrose, 50 mM Tris±Cl (pH 7.5), 25 mM KCl and 5 mM MgCl2 (STKM) at 440 rev./min using ten up-and-down strokes and the homogenate was thereafter adjusted to contain 2 g liver/10 ml. One portion of this homogenate was mixed with twice as much 75% STKM solution and 28 ml of this mixture was layered on top of 10 ml 75% STKM in a centrifuge tube. After centrifugation in the swinging bucket SW 28 rotor at 25 000 rev./min for 60 min, the pellet was resuspended in STKM, this suspension was centrifuged again at 3000 rev./min in a JA 20 rotor for 10 min and ®nally the pellet thus obtained was suspended in STKM to give a ®nal concentration corresponding to 1 g liver/ml. Plasma membranes (PLM) were prepared as described earlier [12]. Protein concentrations were determined by the Biuret procedure using bovine serum albumin as standard.
2.4. Enzyme assays 2.4.1. TrxR activity
The activities of TrxR in homogenates and subcellular fractions was determined using insulin disul®des as substrate, essentially as described by Holmgren and BjoÈrnstedt [3]. Brie¯y, 50 mg of sample protein was incubated with 80 mM HEPES (pH 7.5), 0.9 mg/ml NADPH, 6 mM EDTA, 2 mg/ml insulin and 10 mM E. coli Trx at 378C for 20 min in a total volume of 120 ml. For each sample, a blank containing everything except Trx was incubated and treated in the same manner. The reaction was terminated by the addition of 500 ml DTNB (0.4 mg/ml) and 6 M guanidine hydrochloride in 0.2 M Tris±Cl (pH 8.0). The absorbance at 412 nm was subtracted from the corresponding absorbance of the sample. By comparison with a standard curve prepared employing puri®ed calf thymus TrxR (IMCO, Stockholm, Sweden), levels of TrxR in the various homogenates and subcellular fractions could be determined.
2.4.2. Redox status
In order to quantitate the total relative amounts of thiols (SH groups), both native and enzymatically formed, present in the homogenates of nodules and control liver, samples were incubated in the same reaction mixture as described above at 378C for 20 min. Thereafter, the reaction was terminated and the absorbance at 412 nm was measured also as described above. The concentration of SH groups was calculated using E412nm 13:6 mM 21.
2.5. Isolation of RNA Rat liver samples were removed and immediately frozen in liquid nitrogen. Subsequently, tissue samples weighing 50±100 mg were homogenized in a total volume of 1 ml TRIZOL Reagent using a POLYTRON, and total RNA was isolated as described in the TRIZOL Reagent (Life Technologies) Kit. To ensure inactivation of contaminating RNAses, metal objects and
L. BjoÈrkhem et al. / Journal of Hepatology 35 (2001) 259±264
Fig. 1. Active TrxR (ng/mg total protein) in homogenate from normal liver and liver nodules. The values are the mean ^ SD of six repeat assays performed on homogenate from three independent preparations. ***P , 0:001 compared to the control value. glassware were cleaned with detergent, rinsed thoroughly with RNAse-free water and ®nally heated at 2008C for 6 h. Plastic items were rinsed with 0.1 M NaOH containing 1 mM EDTA, followed by RNAse-free water, and thereafter heated at 1208C for 6 h. 0.2% diethyl pyrocarbonate (DEPC) was added to water and aqueous solutions. The quality and concentration of the RNA preparations thus obtained were ascertained by measurement of their absorbance at 260 and 280 nm. Possible contaminating remnants of genomic DNA were eliminated by treating these preparations with deoxyribonuclease I (ampli®cation grade) prior to RT-PCR ampli®cation.
2.6. RT-PCR Total RNA was subjected to oligo(dT)-primed ®rst-strand cDNA synthesis in a total volume of 20 ml using the 1st Strand cDNA Synthesis Kit, according to the protocol provided by the manufacturer (Boehringer Mannheim). The transcription products thus obtained were then subjected to PCR ampli®cation in a total volume of 100 ml using an automatic thermal cycler (HYBAID). The step-cycle program employed involved denaturation at 948C for 1 min, annealing at 648C for 1 min and extension at 728C for
261
1.5 min for a total of 35 cycles. Prior to initiation of this process, the mixture was denatured for 3 min at 948C; and following after the last cycle an extension reaction was performed for 7 min at 728C. In order to achieve semi-quantitative analysis, primer pairs for the `housekeeping gene' encoding b-actin were simultaneously ampli®ed under identical conditions for use as an internal standard. Equal volumes of the PCR products for TrxR and b-actin were mixed and co-electrophoresed on a 1.5% agarose gel, after which visualization was achieved by ethidium bromide staining. The PCR products were quantitated on the basis of the intensities of the bands, as photographed by a Kodak DC40 camera under UV-light and examined using a computerized digital image analysis program (Kodak Digital Science 1D Image analysis software, Eastman Kodak Co., Rochester, NY). The number of cycles, MgCl2 concentration and annealing temperatures employed were each optimized. Appropriate primer pairs for RT-PCR were constructed on the basis of the gene sequence for rat TrxR (GenBank accession no. U63923), i.e. nucleotides 1077±1100 (5 0 -CCTGTCACGGATGAGGAGCAGACC-3 0 ) were used as the forward primer and nucleotides 2105±2082 (5 0 -GCCATCTGACCCCCTTCCACACAG-3 0 ) were used as the reverse primer. The primers for RT-PCR of b-actin mRNA were designed to amplify 285 bp DNA and were purchased from SDS Promega. Statistical analysis was performed using Student's t-test.
3. Results 3.1. Activity of TrxR in homogenates and subcellular fractions The activity of TrxR in homogenates of liver nodules was 4-fold greater than in normal liver (Fig. 1). Corresponding measurements on subcellular fractions revealed the presence of TrxR activity in the cytosolic and mitochondrial fractions from both control and nodular hepatic tissues (Fig. 2). In the cytosolic fractions, these activities were 250 and 900 ng TrxR/mg protein, respectively. However, in the
Fig. 2. Active TrxR (ng/mg total protein) in cytosol, microsomes, mitochondrias, PLM, and nuclei from normal liver (C) and liver nodules (N). The values are the mean ^ SD of six repeat assays performed on each fraction from three independent preparations. ***P , 0:001 compared to the control value.
262
L. BjoÈrkhem et al. / Journal of Hepatology 35 (2001) 259±264
Fig. 3. Active TrxR (mg/g liver) in homogenate, cytosol, microsomes, mitochondrias, PLM, and nuclei from normal liver (C) and liver nodules (N). The values are the mean ^ SD of six repeat assays performed on each fraction from three independent preparations. ***P , 0:001 compared to the control value.
mitochondrial fractions TrxR activity was higher in control tissue than in nodules (230 versus 140 ng TrxR/mg protein, respectively). Only traces of TrxR activity could be detected in the microsomal, PLM and nuclear fractions from both types of tissue. Fig. 3 depicts the total activity of TrxR per gram of liver in the various subfractions, showing that, by far, most of the cellular TrxR activity is cytosolic. The mitochondrial activity is less than 20% of the total in normal liver and ,5% in liver nodules.
Fig. 4. Semi-quantitative analysis of the level of TrxR mRNA employing RT-PCR. Total RNA was isolated and electrophoresed on 1.5% agarose gel as described in Section 2. b-Actin was used as an internal standard. Lanes 1, 3, 5 and 7 are from normal liver and lanes 2, 4, 6 and 8 are from 2-AAF-induced nodules. Total RNA (100 ng) was applied to lanes 1±4 and 200 ng was applied to lanes 5±8. M, molecular weight markers separated by 250 bp.
3.2. Levels of TrxR mRNA When a semi-quantitative RT-PCR technique was employed to determine the levels of TrxR mRNA in liver nodules and normal liver [13], the nodular level was found to be signi®cantly higher (Fig. 4). When RT-PCR was performed on 100 ng total RNA, no TrxR mRNA was detectable in normal liver, whereas the nodular level was 80% of the level of b-actin mRNA. With 200 ng total RNA, the TrxR mRNA level in liver nodules was seen to be 3-fold higher than in normal liver. In this latter case, the intensity of the TrxR mRNA band in normal liver was 27% of the band intensity for b-actin mRNA, while the corresponding value for nodules was 79%.
Fig. 5. The level of thiol groups (nmol SH groups/mg homogenate protein) in homogenates from normal liver and liver nodules. The values shown are the mean ^ SD of six repeat assays performed on homogenate from three independent preparations. **P , 0:01 compared to the control value.
L. BjoÈrkhem et al. / Journal of Hepatology 35 (2001) 259±264
3.3. Redox status in normal liver and liver nodules The redox status in homogenates from normal liver and liver nodules was determined by measuring the total level of SH groups following a 20 min incubation in the presence of NADPH. The level of thiol groups in homogenate from nodules was double (840 nmol/mg protein) compared to control liver (420 nmol/mg protein) (Fig. 5).
4. Discussion In this study elevations in the activity and mRNA level of TrxR in rat liver nodules have been demonstrated. The activity of TrxR in homogenates from nodules was 4-fold greater than in control homogenate. The level of TrxR mRNA was also elevated in nodular cells compared to normal hepatocytes, indicating that the increase in TrxR activity is due to transcriptional up-regulation and/or mRNA stabilization. The advantage of the activity assay employed here (i.e. reduction of insulin disul®des in the presence of excess Trx) is that the total activity of all forms of TrxR is detected, as opposed to the detection of individual proteins using immunoblotting. In subcellular fractions, TrxR activity was increased in the cytosolic but decreased in the mitochondrial fraction from nodules. Mitochondrial TrxR is a distinct enzyme (referred to as TrxR2) and has been isolated from humans [14,15] and rats [16]. The primary function of TrxR2 is still unknown but a role in protecting against oxidative stress has been suggested [17]. Several reports have demonstrated elevated TrxR activity in tumor cells [6,18±20]. Furthermore, treatment of mouse skin with several different tumor promoters resulted in increased activity of TrxR as well as in increased levels of Trx and glutaredoxin in this tissue [21]. However, no distinction was made between the cytosolic TrxR and the mitochondrial TrxR2 in these earlier studies. Our present ®ndings of increased activity in the cytosol and decreased activity in the mitochondrias of nodules indicate that the cytosolic enzyme ®lls a function different than that of the mitochondrial enzyme in the resistant cell. Only traces of activity were present in the microsomes (endoplasmic reticulum, ER) and PLM and in these subfractions there were no differences between nodules and control liver. In a previous study in which TrxR in different organs of the adult rat was observed immunohistochemically, the ¯uorescence of anti-TrxR antibodies was reported to be primarily cytosolic [22]. In another study of rat hepatocytes, involving electron microscopic immunohistochemistry in combination with the immunogold technique, the TrxR protein was found to be localized not only in the cytosol but also in association with the granular ER, in the cistern of the Golgi complex, in the chromatin of the nucleus and at the periphery of mitochondria [23]. The different techniques employed as well as differences in speci®c TrxR activity in
263
various organelles probably explain the subtle discrepancies in the results obtained utilizing immunochemical procedures and enzymatic assays. It is expected that the protein is detected also where the protein is synthesized and processed and not only where it has its activity. 2-AAF, originally used as an insecticide, is a well-known hepatocarcinogen [24]. This molecule is highly hydrophobic and diffuses easily through biological membranes, as a result of which it is initially enriched in lipid-rich compartments of the body. The parent compound is metabolized by the microsomal cytochrome P450 system to various hydroxylated metabolites, of which the most reactive is N-OH-2acetylamino¯uorene. This compound is then conjugated enzymatically, primarily to glucuronic acid, glutathione and sulfate. Whereas the other conjugates are excreted, the sulfate ester of N-OH-2-acetylamino¯uorene is an especially labile compound that spontaneously forms the ultimate reactive metabolite, the nitrenium ion. The nitrenium ion radical reacts with cellular macromolecules to give rise to the toxic effects of the compound. Resistant nodular cells also contain high levels of antioxidants and are resistant to various sources of oxidative stress, including those unrelated to 2-AAF. This phenotype is constitutively expressed by persistent nodules even in the absence of promoter. Rather than being speci®c for 2-AAF, it is probable that increased TrxR activity is a part of the resistant phenotype that develops during tumor development and characterizes the process of tumor development in general. The mammalian TrxR reduce a wide variety of low molecular weight substrates and are above all ef®cient peroxidases, reducing hydrogen peroxide [25] and, even more interestingly, lipid hydroperoxides [7]. In addition, Trx itself reduces hydrogen peroxide, organic hydroperoxides and cellular disul®des formed in connection with oxidative stress. Since the Trx system plays an important role in DNA synthesis, increased TrxR activity might enhance the rate of ribonucleotide reduction. Oxidative stress has been suggested to activate the regulatory protein oxy-R, which results in overexpression of alkyl hydroperoxide reductase, a TrxR-like protein [26]. We suggest that TrxR activity is of great signi®cance for cellular resistance, since overexpression of this protein results in more ef®cient defense systems. The level of SH groups, both endogenous and enzymatically formed, in homogenates from nodules was 2-fold higher than the corresponding value for normal rat liver. This ®nding demonstrates that nodules possess a greater reducing potential and thereby a greater capacity to withstand oxidative stress. Our results clearly demonstrate that 2-AAF-induced resistant rat liver nodules contain an enhanced reducing potential and increased activity of TrxR, a key enzyme for defense and growth. Further experiments focused on the regulation of selenoproteins in connection with chemical carcinogenesis are currently in progress in our laboratory.
264
L. BjoÈrkhem et al. / Journal of Hepatology 35 (2001) 259±264
Acknowledgements The excellent technical assistance of Mrs Ulla-Britta Torndal is gratefully acknowledged. This investigation was supported by grants from the Swedish Cancer Society, The Swedish Medical Association and Karolinska Institutet. References [1] Epstein S, Ito N, Merkow L, Farber E. Cellular analysis of liver carcinogenesis: the induction of larger hyperplastic nodules in the liver with 2-¯ourenylacetamide or ethionine and some aspects of their morphology and glycogen metabolism. Cancer Res 1967;27:1702. [2] Eriksson LC, Torndal UB, Andersson GN. Isolation and characterization of endoplasmic reticulum and Golgi apparatus from hepatocyte nodules in male Wistar rats. Cancer Res 1983;43:3335±3347. [3] Holmgren A, BjoÈrnstedt M. Thioredoxin and thioredoxin reductase. Methods Enzymol 1995;252:199±208. Ê slund F, BjoÈrnstedt M, Zhong L, Ljung J, et [4] Holmgren A, ArneÂr E, A al. Redox regulation by the thioredoxin and glutaredoxin systems. In: Montagnier L, Olivier R, Pasquier C, et al., editors. Oxidative stress in cancer, aids, and neurodegenerative diseases, Paris: Marcel Dekker, 1996. pp. 229±246. [5] Williams CHJ. Lipoamide dehydrogenase, glutathione reductase, thioredoxin reductase and mercuric ion reductase ± a family of ¯avoenzyme transhydrogenase. In: MuÈller F, editor. Chemistry and biochemistry of ¯avoenzyme, 3. Boca Raton, FL: CRC Press, 1992. pp. 121±211. [6] Tamura T, Stadtman TC. A new selenoprotein from human lung adenocarcinoma cells: puri®cation, properties, and thioredoxin reductase activity. Proc Natl Acad Sci USA 1996;93:1006±1011. [7] BjoÈrnstedt M, Hamberg M, Kumar S, Xue J, Holmgren A. Human thioredoxin reductase directly reduces lipid hydroperoxides by NADPH and selenocystine strongly stimulates the reaction via catalytically generated selenols. J Biol Chem 1995;270:11761±11764. Ê kesson B, Holmgren A. The thior[8] BjoÈrnstedt M, Xue J, Huang W, A edoxin and glutaredoxin systems are ef®cient electron donors to human plasma glutathione peroxidase. J Biol Chem 1994;269:29382±29384. [9] Sottocasa JL, Kuylenstierna B, Ernster L, Bergstrand A. Separation and some enzymatic properties of the inner and outer membranes of rat liver mitochondria. Methods Enzymol 1967;10:448±463. [10] Dallner G. Isolation of rough and smooth microsomes ± general. Methods Enzymol 1974;31:191±201. [11] Blobel G, Potter VR. Nuclei from rat liver: isolation method that combines purity with high yield. Science 1966;154:1662±1665.
[12] Loten EG, Redshaw-Loten JC. Preparation of rat liver plasma membranes in a high yield. Anal Biochem 1986;154:183±185. [13] Serazin-Leroy V, Denis-Henriot D, Morot M, de Mazancourt P, Giudicelli Y. Semi-quantitative RT-PCR for comparison of mRNAs in cells with different amounts of housekeeping gene transcripts. Mol Cell Probes 1998;12:283±291. [14] Gasdaska PY, Berggren MM, Berry MJ, Powis G. Cloning, sequencing and functional expression of a novel human thioredoxin reductase. FEBS Lett 1999;442:105±111. [15] Miranda-Vizuete A, Damdimopoulos AE, Pedrajas JR, Gustafsson Ê , Spyrou G. Human mitochondrial thioredoxin reductase cDNA JA cloning, expression and genomic organization. Eur J Biochem 1999;261:405±412. [16] Lee SR, Kim JR, Kwon KS, Yoon HW, Levine RL, Ginsburg A, Rhee SG. Molecular cloning and characterization of a mitochondrial selenocysteine-containing thioredoxin reductase from rat liver. J Biol Chem 1999;274:4722±4734. [17] Miranda-Vizuete A, Damdimopoulos AE, Spyrou G. cDNA cloning, expression and chromosomal localization of the mouse mitochondrial thioredoxin reductase gene(1). Biochim Biophys Acta 1999;1447:113±118. [18] Sun X, Dobra K, BjoÈrnstedt M, Hjerpe A. Upregulation of 9 genes, including that for thioredoxin, during epithelial differentiation of mesothelioma cells. Differentiation 2000;66:181±188. [19] Berggren M, Gallegos A, Gasdaska JR, Gasdaska PY, Warneke J, Powis G. Thioredoxin and thioredoxin reductase gene expression in human tumors and cell lines, and the effects of serum stimulation and hypoxia. Anticancer Res 1996;16:3459±3466. [20] Powis G, Gasdaska JR, Gasdaska PY, Berggren M, Kirkpatrick DL, Engman L, Cotgreave IA, et al. Selenium and the thioredoxin redox system: effects on cell growth and death. Oncol Res 1997;9:303±312. [21] Kumar S, Holmgren A. Induction of thioredoxin, thioredoxin reductase and glutaredoxin activity in mouse skin by TPA, a calcium ionophore and other tumor promoters. Carcinogenesis 1999;20:1761± 1767. [22] Rozell B, Hansson HA, Luthman M, Holmgren A. Immunohistochemical localization of thioredoxin and thioredoxin reductase in adult rats. Eur J Cell Biol 1985;38:79±86. [23] Rozell B, Holmgren A, Hansson HA. Ultrastructural demonstration of thioredoxin and thioredoxin reductase in rat hepatocytes. Eur J Cell Biol 1988;46:470±477. [24] Floyd RA. Free radicals and cancer. New York: Marcel Dekker, 1982. [25] Zhong L, Holmgren A. Essential role of selenium in the catalytic activities of mammalian thioredoxin reductase revealed by characterization of recombinant enzymes with selenocysteine mutations. J Biol Chem 2000;275:18121±18128. [26] Storz G, Tartaglia LA, Ames BN. The OxyR regulon. Antonie Van Leeuwenhoek 1990;58:157±161.