the role of peroxiredoxin ii in radiation-resistant mcf-7 ... · radiation. (cancer res 2005;...

The Role of Peroxiredoxin II in Radiation-Resistant MCF-7

Breast Cancer Cells

Tieli Wang,1Daniel Tamae,

2Thomas LeBon,

2John E. Shively,

3Yun Yen,

2and Jian Jian Li

4

1Department of Chemistry, California State University, Carson, California; 2Department of Medical Oncology and 3Division of Immunology,City of Hope National Medical Center and Beckman Research Institute, Duarte, California; and 4Division of Molecular Radiobiology,Purdue University School of Health Sciences, West Lafayette, Indiana

Abstract

Although several signaling pathways have been suggested tobe involved in the cellular response to ionizing radiation, themolecular basis of tumor resistance to radiation remainselusive. We have developed a unique model system based uponthe MCF-7 human breast cancer cell line that became resistantto radiation treatment (MCF+FIR30) after exposure to chronicionizing radiation. By proteomics analysis, we found thatperoxiredoxin II (PrxII), a member of a family of peroxidases,is up-regulated in the radiation-derived MCF+FIR3 cells butnot in the MCF+FIS4 cells that are relatively sensitive toradiation. Both MCF+FIR3 and MCF+FIS4 cell lines are fromMCF+FIR30 populations. Furthermore, the resistance toionizing radiation can be partially reversed by silencing theexpression of PrxII by PrxII/small interfering RNA treatmentof MCF+FIR3 resistant cells, suggesting that PrxII is notthe sole factor responsible for the resistant phenotype. Therelevance of this mechanism was further confirmed by theincreased radioresistance in PrxII-overexpressing MCF+FIS4cells when compared with vector control cells. The up-regulation of the PrxII protein in radioresistant cancer cellssuggested that human peroxiredoxin plays an important rolein eliminating the generation of reactive oxygen species byionizing radiation. The present finding, together with theobservation that PrxII is also up-regulated in response toionizing radiation in other cell systems, strengthens thehypothesis that the PrxII antioxidant protein is involved inthe cellular response to ionizing radiation and functions toreduce the intracellular reactive oxygen species levels, result-ing in increased resistance of breast cancer cells to ionizingradiation. (Cancer Res 2005; 65(22): 10338-46)

Introduction

Clinical radiation resistance continues to be a major problem inthe treatment of cancer. The molecular mechanisms underlyingtumor-adaptive radioresistance remains to be elucidated. It hasbeen well documented that ionizing radiation produces freeradicals in biological tissues. Those radicals are the primary sourceof reactive oxygen species (ROS) and reactive nitrogen species(RNS). ROS/RNS, such as superoxide anion, hydrogen peroxide, andperoxynitrite, seem to be transiently produced in response toionizing radiation, growth factor, and cytokine stimulation (1, 2).They are involved in both the lethal clastogenic and mutagenic

effects of ionizing radiation (3). Radioprotective effects bymodifications of antioxidant enzyme expression or by addition offree-radical scavengers have been described (3–6). Accumulatingevidence show that ROS and RNS are not only toxic byproducts buthave been shown to act as signal transduction agents involved inregulating normal functions, including cell growth and differenti-ation, immune responses by phagocytic cells, and apoptosis (7).ROS and RNS play a significant but paradoxical role acting as a‘‘double-edged sword’’ to regulate cellular processes.In the past few years, nuclear DNA damage–sensing mechanisms

activated by ionizing radiation have been identified, includingataxia-telangiectasia mutated (ATM)/ATM and Rad-3 related andthe DNA-dependent protein kinase (8–10). Less is known aboutsensing mechanisms for cytoplasmic ionization events and howthese events influence nuclear processes. Several studies haveshown the importance of cytoplasmic signaling pathways incytoprotection and mutagenesis. For cytoplasmic signaling, radi-ation-stimulated ROS and RNS are essential activators of thecytoplasmic signaling pathways (11). Although as yet unproven,the nature and extent of the ROS and RNS insult may determinethe threshold of the cellular response, manifesting as proliferation,stress response, and damage repair or apoptosis (12, 13). Elevationof ROS in excess of the buffering capacity and enzymatic activitiesdesigned to modulate ROS levels results in potentially cytotoxicoxidative stress. Under these pro-oxidant conditions, highlyreactive radicals can damage DNA, RNA, protein, and lipid com-ponents, which may lead to cell death. Oxidative stress induced byROS/RNS, such as hydrogen peroxide, has been implicated in thepathogenesis of several neurodegenerative diseases and cancers(12–15). To cope with oxidative insult, the cell has developeda number of defense strategies, both at the level of oxidativedamage repair and ROS/RNS-scavenging mechanisms (2). Repairmechanisms that confer resistance to oxidative stress includevarious protein disulfide reductase enzymes as well as multifunc-tional DNA repair and thiol-reducing proteins. Antioxidantmechanisms comprise small proteins with redox-active sulfhydrylmoieties as well as enzymatic ROS-metabolizing systems,including superoxide dismutase, catalase, and peroxidase. Wehypothesize that radio adaptation could arise from the activationof damage repair and/or the reactions of the antioxidant defensesystems.To address the above questions, we have established an effective

model system to further explore the radiation resistance pheno-type. This model is based on the MCF-7 human breast cancer cellline that was exposed to chronic ionizing radiation and becameresistant to radiation treatment (16). Because most advanced andrecurrent tumors are resistant to chemotherapy, it is desirable toinvestigate persons or cells that express such abnormalities.Acquired radiation-resistant cancer cells (MCF+FIR30) wereobtained by exposure of MCF-7 breast cancer cells to fractionated

Requests for reprints: Tieli Wang, Department of Chemistry, California StateUniversity, Carson, CA 90072. Phone: 310-243-3388; Fax: 310-243-2593; E-mail:[email protected].

I2005 American Association for Cancer Research.doi:10.1158/0008-5472.CAN-04-4614

Cancer Res 2005; 65: (22). November 15, 2005 10338 www.aacrjournals.org

Research Article

Research. on November 17, 2020. © 2005 American Association for Cancercancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

ionizing radiation (FIR) with a total dose of 60 Gy of g-irradiation(2 Gy per fraction, five times per week for 6 weeks; ref. 16). Severalradiation-resistant clones were independently isolated and char-acterized. A radioresistant clone (MCF+FIR3) and a radiosensitiveclone (MCF+FIS4) isolated from the same MCF+FIR30 populationswere cultured for further study. Comparative proteomic analysiswas done to search for expression alterations between the twoclones. We found that the protein expression level of peroxiredoxinII (PrxII) is significantly up-regulated in radioresistant MCF+FIR3cells compared with that in radiosensitive MCF+FIS4 cells. Westernblot analysis of radioresistant MCF+FIR3 cells and radiosensitiveMCF+FIS4 cells confirmed the increased expression of PrxII proteinin MCF+FIR3 radioresistant breast cancer cells. We introduceddsRNA into MCF+FIR3 radioresistant breast cancer cells, whichleads to gene silencing in a sequence-specific manner. Clonogenicsurvival assays were done to show that silencing the PrxII genesensitized radioresistant MCF+FIR3 breast cancer cells. In addition,overexpression of PrxII in radiosensitive MCF+FIS4 cells results inincreased radioresistance in breast cancer cells. ROS are known tocontribute to the toxicity of ionizing radiation, and the intracellularscavenging of these species is an important modulator of ionizingradiation–induced damage. Our experimental results clearlyindicate that radiation-resistant cells contain elevated levels ofthe radical scavenger PrxII protein, a member of a family ofproteins that provide a protective role in redox damage. Inductionof PrxII protein in resistant breast cancer cells provides evidencefor the involvement of ROS in breast cancer cells resistance toradiation. Our study helps to understand and manipulate resistancein radiation therapy and may elucidate an alternative mechanismby which breast cancer can be sensitized to radiation.

Materials and Methods

Development radioresistance cell line. MCF-7 human breast cancercells were purchased from the American Type Culture Collection

(Manassas, VA). MCF+FIR30 cells were obtained from MCF-7 cells by

exposure to FIR with a total dose of 60 Gy of g-irradiation (2 Gy per fraction,

five times per week for 6 weeks). Radiation was delivered at roomtemperature at a rate of 5 Gy/1.15 minutes with a Cesium-137 (Cs-137)

Irradiator (JL Shepherd Mark I, San Fernando, CA). A moderate radio-

resistance [dose-modifying factor (DMF) = 1.6-2.3 at 10% isosurvival rate] inMCF-7 human breast cancer cells that was exposed to FIR for 20 fractions

(MCF+FIR20) or for 30 fractions (MCF+FIR30) was observed (16). These

FIR-derived MCF-7 populations showed increased radioresistant phenotype

for 7 to 10 passages after FIR treatment. There is no resistance observed incells after 10 passages without FIR. Several clones isolated from the

resistant MCF+FIR30 breast cancer cell population were individually

cultured. Clonogenic assays were used to determine the resistance level

among those clones. A radiation-resistant clone (MCF+FIR3) and aradiation-sensitive clone (MCF+FIS4) isolated from the MCF+FIR30 clonal

populations were used for comparative proteomic analysis.

Clonogenic assay. Experiments were done to compare the radio-

resistant level between the MCF+FIR3 and MCF+FIS4 cells. Radioresistancewas measured by clonogenic survival following exposure to IR. Cells were

plated into T-25 flasks and cultured for 6 hours before exposure to graded

dose of Cs-137. After irradiation, cells were cultured for 21 days, andcolonies having >50 cells were counted. Plating efficiencies were calculated

as colonies per 10 cells. Surviving fractions were normalized to the plating

efficiencies. DMFs were calculated using the following expression: DMF =

dose to reach the specified survival in resistant cells � dose to reach thesame survival in the control (1).

Silencing peroxiredoxin II expression by small interfering RNA.Small interfering RNA (siRNA) capable of targeting mRNAs encoding PrxII

was custom synthesized by Dharmacon Research (Lafayette, CO). The cells

were seeded to achieve 30% to 50% confluence on the day of transfection.Transient transfection of siRNAs was done using a protocol recommended

from the manufacturer (Dharmacon Research). In brief, the cells were

harvested by exposure to trypsin, diluted with fresh medium without

antibiotics, and transferred to T-25 flask. After incubation for 24 hours,the cells were transfected with the siRNAs for 2.5 days with the use of

Oligofectamine (Invitrogen, Carlsbad, CA). The Scramble RNA Duplex

(City of Hope, Duarte, CA) served as a control for siRNA activity. A mixture

of 20 AL of Opti-MEM medium (Invitrogen) and 5 AL of Oligofectaminewere incubated for 10 minutes at room temperature and were then

combined with 20 AL of 20 Amol/L siRNA diluted with 355 AL of Opti-MEM.

The resulting mixture (400 AL) was incubated for 20 minutes at room

temperature to allow complex formation and then overlaid onto each flaskcontaining the cells for a final volume of 3.6 mL/well. Approximately 105

MCF+FIS4 cells were transfected using Oligofectamine (Invitrogen) with

specific siRNAs. PrxII expression of the transiently transfected cells wasanalyzed by Western blot at 24, 48, 72, and 144 hours after transfection

using antibodies specific for PrxII (LabFrontier, Seoul, South Korea). All the

transfectants were maintained in 10% fetal bovine serum (FBS)/DMEM

(Mediatech, Inc., Herndon, VA) until collected for analysis.Sample preparation for proteomics analysis. Radioresistant

MCF+FIR3 and radiosensitive MCF+FIS4 cells were cultured for compar-

ative proteomics analysis. Exponentially growing cells were rinsed thrice

with ice-cold PBS and lysed in lysis buffer (8 mol/L urea, 4% CHAPS, and 40mmol/L Tris base) by sonication in a sonic bath three to four times for an

interval of 30 seconds. After removal of insoluble materials by centrifuga-

tion at 14,000� g for 30 minutes, cell lysates were mixed with 10 volumes ofrehydration buffer (8 mol/L urea, 2% CHAPS, 0.5% immobilized pH gradient

buffer, 20 mmol/L DTT, and 0.005% bromphenol blue) and loaded onto

immobilized pH gradient strips (pH 3-10, nonlinear and pH 4-7; Amersham

Biosciences, Piscataway, NJ). The protein concentration in the supernatantfraction was determined by bicinchoninic acid assay (Pierce Biotechnology,

Inc., Rockford, IL) using bovine serum albumin as a standard. Carrier

ampholytes (0.5% final concentration) were added, and the protein extracts

were stored at �20jC.Gel electrophoresis and analysis. Proteins were separated by two-

dimensional electrophoresis, using immobilized pH gradient strip (pH 3-10

and pH 4-7). The samples were loaded on the immobilized pH gradient strip

by in-gel sample rehydration, using a urea-thiourea mixture as solubilizing

agent. Total extract (240-450 Ag) was loaded on the first dimension gel.

Figure 1. Different sensitivities to radiation between cloned MCF+FIR30 celllines. The cloned cells (MCF+FIR3 and MCF+FIS4) were expanded in T-25flasks for two to four passages to get enough cells for the measurement ofclonogenic survival. Cloned cell lines were trypsinized, and clonogenicradiosensitivity was measured following irradiation with a range of single doses.Points, mean (n = 3); bars, SE.

Peroxiredoxin II in Radiation Resistance

www.aacrjournals.org 10339 Cancer Res 2005; 65: (22). November 15, 2005



Isoelectric focusing on an IPGPhor isoelectrofocusing unit (Amersham

Biosciences) and preparation (reduction and alkylation) of the immobilized

pH gradient strips for the second-dimension SDS-PAGE was carried out

according to the procedures recommended by the manufacturer. The

isoelectric focusing was done over 24 hours for a total of 55,000 V-h. SDS-

PAGE was conducted on 12% gels using an Amersham Biosciences SE600

vertical unit, and the protein spots were visualized by staining with colloidal

Coomassie (Bio-Rad, Hercules, CA). Colloidal Coomassie two-dimensional

gels were scanned and quantified with the GS 800 Bio-Rad densitometry

and PdQuest software. All two-dimensional gels were run a minimum of

three independent times under each condition to insure reproducibility.

Mass spectrometry analysis. Excised, chopped, destained, and dehy-

drated protein spots from gels were reduced in reducing buffer (10 mmol/L

DTT in 100 mmol/L NH4HCO3). The samples were then alkylated in 100

mmol/L iodoacetamide solution in the dark. Gel pieces were rehydrated on

ice for 45 minutes in 1 mmol/L HCl/100 mmol/L ammonium bicarbonate

containing 50 ng/AL sequencing grade pig trypsin (Promega, Madison, WI).

Digestion was carried out at 30jC for 15 hours. The resulting peptides were

extracted by sequential extraction for 20 minutes in 20 mmol/L ammonium

bicarbonate followed by 0.1% trifluoroacetic acid in 60% acetonitrile.

Combined extracts were concentrated in a Speed Vac to 20 AL. Samples

were loaded from a stainless steel chamber pressurized to 1,000 p.s.i. onto

Figure 2. Two-dimensional gel images.A, proteins extracted from MCF+FIR3radioresistant breast cancer cells wereseparated on an immobilized pH 3 to10 gradient strip followed by a 12%polyacrylamide gel, as stated in Materialsand Methods. The gels were stained withCoomassie blue. B, proteins extractedfrom the MCF+FIS4 radiosensitive breastcancer cells were run in the same conditionas in (A ). The two-dimensional images ofMCF+FIR3 and MCF+FIS4 cells wereanalyzed by PdQuest software. C, foldincreases for No.1 spot in radioresistantMCF+FIR3 cells compared with that inradiosensitive MCF+FIS4 cells. D, proteinsextracted from MCF+FIR3 cells wereseparated on an immobilized pH 4 to7 gradient strip followed by a 12%polyacrylamide gel, as stated in Materialsand Methods. The gels were stained withCoomassie blue. E, proteins extractedfrom the MCF+FIS4 cells were run inthe same condition as in (A ). Thetwo-dimensional images of MCF+FIR3and MCF+FIS4 cells were analyzed byPdQuest software. F, fold increases forNo.1 spot in MCF+FIR3 cells comparedwith that in MCF+FIS4 cells.

Cancer Research




fused silica capillary (150 Am, inner diameter) slurry packed to 75 mm with

Everest C18 beads (Michrom, Auburn, CA). The columnwas developedwith a

linear gradient of 5% to 50% acetonitrile in 0.4% acetic acid and 0.005%

hepta-fluorobutyric acid over 25 minutes at 300 to 400 nL/min. Electrospray

ionization was conducted by applying 1.25 kV to a Valco stainless steel union.

Capillary columns terminated inside the Valco union, and a fused silica

capillary (75 mm � 150 Am � 360 Am) tapered to f5-Am inner diameter

served as an emitter. The peptides eluting from the column were analyzed

directly on a TFinnigan LCQ Classic liquid chromatography tandem mass

spectrometry (LC/MS/MS) spectrometer equipped with an in-house built

microspray device at Core Facility of City of Hope. Data-dependent MS/MS

spectra were acquired automatically by an Instrument Control Language

procedure. Acquired MS/MS spectra were searched with SEQUEST against

the National Center for Biotechnology Information (NCBI) protein database.

Western blot analysis. For immunoblot analyses of Prx enzymes,

proteins on one-dimensional gels were transferred electrophoretically to a

nitrocellulose membrane, and the membrane was incubated with rabbit

antibodies against PrxII. Immunocomplexes were detected with horseradish

peroxidase–conjugated secondary antibodies and enhanced chemilumines-

cence reagents (Amersham Biosciences or Pierce Biotechnology). All

immunoblot and SDS-PAGE gels were scanned using a EPSON expression

800 flatbed with Adobe Photoshop software for quantification. Scanned

images were quantified by Kodak 1D image software. Excel software was

used for data analysis.

Overexpression of the PrxII gene in breast cancer cells. The coding

region of the PrxII gene was amplified by PCR with full-length human cDNA

as a template and subcloned into a mammalian expression vectorpcDNA3.1. The forward primer sequence is 5V-CCTTTGCCCACGAAGCTTT-CAGTCA, where the underlined A base pair was mutated from C to create

the HindIII enzyme digestion site. The reverse primer is 5V-CTCAC-TATCCGGGATCCAGCCTAATT, where the underlined A and C base pairs

were mutated from A residues to create the BamHI enzyme digestion site.

The PCR product was digested with HindIII and BamHI enzymes and

ligated into pcDNA3.1 vector digested by the same enzyme. The vectorcontains the neomycin phosphotransferase gene, a selectable marker for

mammalian cells. PrxII transfectants were obtained using published

method with LipofectAMINE reagent (Life Technologies, Inc., Gaithersburg,

MD). The transfected MCF+FIS4 cells were selected with the antibioticG418. A pair of cell lines was established by transfecting the recombinant

plasmid of PrxII and the empty vector control. Clonogenic assay was done

to compare the resistant level between PrxII-overexpressing cells andMCF+FIS4 vector control cells.

Results

Heterogeneity of radioresistant breast cancer cells. Radio-resistance developing during consecutive treatments with ionizingradiation is an extremely complex and poorly understoodphenomenon within the cells that develop radioresistance. Thereis tremendous heterogeneity as shown by their wide range ofresponses to different radiation doses. Tumor heterogeneity hasalso been well documented in the response to a variety of anti-cancer reagents. Previous experiments have shown that NSAIDsinduce heterogeneous changes in human tumor cells, includinggrowth arrest, apoptosis, and necrosis (17). To understand themechanism underlying the resistant phenotype, we acquired aradiation-resistant breast cancer cell population (MCF+FIR30) andused it as a model system to study the adaptive survival to chronicFIR (16). Clonogenic survival assays have revealed that theheterogeneity exists in the MCF+FIR30 population. Figure 1 showsthe results from clonogenic survival assays for the radioresistantMCF+FIR3 and a reverted radiosensitive MCF+FIR4 cloned cellline. This unique pair of cells were used for proteomics analysis.Changes in MCF+FIR3 proteomic profile. Radioresistant

MCF+FIR3 clone and radiosensitive MCF+FIS4 clones werecultured for four passages. MCF+FIR3 cells still retained theresistance phenotype, whereas MCF+FIS4 cells were relativelysensitive to ionizing radiation. Comparative proteomic analysis wasdone to search for expression alterations between the two clonesMCF+FIR3 and MCF+FIS4. Figure 2 shows the typical two-dimensional gels done on total cellular extracts for both MCF+FIR3and MCF+FIS4 cells. Figure 2A is the two-dimensional gel forMCF+FIS4 radiosensitive breast cancer cells, whereas Fig. 2B is forradioresistant MCF+FIR3 breast cancer cells. After colloidalCoomassie staining, 100 different spots focalized in the pH range3 to 10 used for the analysis of each sample. A pair of spots (Fig. 2Aand B , No. 1) was visible on the gels; one spot intensified inradioresistance MCF+FIR3 cells and diminished in intensity inradiosensitive MCF+FIS4 cells. PdQuest software analysis showedabout 4-fold increase in the area of the spot from the radioresistantMCF+FIR3 breast cancer cells compared with the one fromradiosensitive MCF+FIS4 breast cancer cells (Fig. 2C). To get a

Figure 2 Continued . G, partial two-dimensional gel images;protein corresponding to PrxII (arrow ). An electrospray LC/MS/MSof the single- and double-charged tryptic fragment 5 of PrxII.





clearer picture of them, proteins from both cells were furtheranalyzed in two-dimensional gels by narrowing down the pH rangefrom 3 to 10 to 4 to 7 (Fig. 2D and E). Figure 2D is the two-dimensional gel image for radiosensitive MCF+FIR4 breast cancercells, whereas Fig. 2E is the image for radioresistant MCF+FIR3breast cancer cells. PdQuest software analysis showed about thesame 4-fold increase in the area of the spot from the radioresistantMCF+FIR3 breast cancer cells versus MCF+FIS4 breast cancer cellsfor pH range between 4 and 7 as for pH range 3 to 10 (Fig. 2F). Thespots 1 from Fig. 2B and E were excised from the gel, digested withtrypsin, and analyzed by LC/MS/MS. Acquired MS/MS spectra weresearched with SEQUEST software against the NCBI proteindatabase. Peptide mass fingerprint analysis and nonredundantsequence database matching allowed the unambiguous identifica-tion of the analyzed species. The result showed that the spots fromMCF+FIR3 two-dimensional gels (pH 3-10 and pH 4-7) are PrxIIproteins. Table 1 reports the nature of the identified spot, themeasured two-dimensional coordinates, and relative sequencecoverage. The nature of the mass signals occurring in the spectraexcluded the simultaneous presence of different protein species inthe analyzed spot. Figure 2G shows an electrospray LC/MS/MS ofthe single and double charged tryptic fragment 5 of PrxII. UnlikeH2O2-treated cells, PrxII protein does not occur as differentisoforms with various isoelectric point values in the radioresistantMCF+FIR3 breast cancer cells compared with that in radiosensitiveMCF+FIS4 breast cancer cells.Peroxiredoxin II isoform is selectively up-regulated in

radioresistant breast cancer cells. The Western blot for all Prxscytosol isoforms showed that PrxII, PrxIII, PrxV, and PrxVI areconstitutively expressed higher (Fig. 3), whereas PrxIV is low. NoPrxIV isoform can be detected in MCF+FIR3 and MCF+FIS4 breastcancer cells, whereas PrxIV can be detected from HK18 cell lines(data not shown). Enhanced PrxII expression was seen in theradioresistant MCF+FIR3 cells compared with MCF+FIS4 cells. Aslightly increased expression of PrxI isozyme was also observed inradioresistant MCF+FIR3 cells. PrxI protein can be detected, butthe signal is not as strong as that of PrxII.PrxII protein is in the reduced form in both radioresistant

MCF+FIR3 cells and in radiosensitive MCF+FIS4 breast cancercells. Anti-SO3-PrxII antibody that recognizes both oxidized formsof cysteine residues, Cys-SO2H and Cys-SO3H, was used to detectoveroxidized PrxII enzyme. Our Western blot analysis did not showany oxidized PrxII protein in MCF+FIR3 and MCF+FIS4 cells (datanot shown). PrxII isozyme does not form large, insoluble,noncovalent aggregates when PrxII is present in high concentrationfrom the result of the Western blot. Our experimental results showselectivity for Prx isozymes overexpression in breast cancer–resistant cells. PrxII has been shown able to protect both DNA from

nicking and protein from inactivation mediated by oxidativedamage in mixed-function oxidation system (18, 19). The over-expression of PrxII in an endothelial line protects the cells fromoxidative stress caused by the heavy metal mercury, inorganichydrogen peroxides, and organic t-butyl-hydroperoxide (20–22). Ithas been shown to protect RBC membranes from peroxidation (23,24). PrxII has a wide variety of antioxidant functions throughoutthe cell.Peroxiredoxin II sequence alignment. We searched for

sequence homologues, using human PrxII sequence as a queryfor the Blast program in the NCBI (http://www3.ncbi.nlm.nih.gov/cgi-bin/BLAST/) databases. We identified four new homologoussequences (E = 10�73 to 10�108; see Fig. 4). This result supportsthe existence of the PrxII family. The PrxII family shows a broadtaxonomic distribution. Members of the family are present inEscherichia coli ; Saccharomyces cerevisiae ; yeast; Rattus norvegicus ;Mus musculus , mouse; Cricetulus griseus ; human, bovine; Bostaurus ; Brugia malayi ; Litomosoides sigmodontis ; Onchocercaochengi ; Onchocerca volvulus ; Echinococcus granulosus ; andSchistosoma mansoni (Fig. 4). Note that in most cases, copies ofmembers of the two-cysteine PrxI, PrxIII, and PrxIV are alsopresent. Members of the PrxII family have distinct structuralfeatures. They are slightly smaller than the PrxIII and PrxIVproteins: PrxII proteins possess 198 residues; PrxIII proteinspossess 238 residues; and PrxIV proteins possess 274 residues.Members of each family display a high degree of sequenceconservation (Fig. 4): amino acid residues conserved in alignedsequences (*); identical or closely similar residues (:). The twocysteines of the PrxII family proteins that form a transientdisulfide bridge are shown by Cp and CR. Isoleucine, leucine,methionine, and valine (those hydrophobic residues) were coloredgreen. Amino acid residues glycine, proline, serine, and threoninewere colored orange. Histine, lysine, and arginine were coloredred; phenylalanine, tryptophan, and tyrosine were colored blue.The previous report showed that bacterial two-cysteine Prxs aremuch less sensitive to oxidative inactivation than are eukaryotictwo-cysteine Prxs due to a three-residue insertion associated witha conserved Gly-Gly-Leu-Gly sequence (the GGLG motif) and aCOOH-terminal extension associated with a conserved Tyr-Phesequence (the YF motif). All yeast, parasite, and mammalianPrxII proteins have a three-residue insertion associated with aconserved sequence GGLG and a conserved Tyr-Phe sequence(the YF motif) identified by comparing the crystal structure ofa bacterial two-cysteine Prx with other Prx structures (25). Theyare all possibly sensitive to oxidation. Prokaryotic Prx enzymes,which do not contain the COOH-terminal GGLG motif and the YFmotif of their eukaryotic counterparts, are insensitive to oxidationinactivation (26).

Table 1.

Spot no. 2D gel pHrange

Protein Q score Sequencingcoverage (%)

pl M r (kDa)

1 3-10 Prx II human 57.74 55 5.41 21.8911 4-7 Prx II human 53.66 68 5.41 21.891

Abbreviation: 2d, two-dimensional.

Cancer Research




RNA interference leads to radiosensitization of radioresist-ant MCF+FIR3 cells. To inhibit protein expression in the highestfraction of cells, the siRNA transfection was repeated 72 hourslater. Transfection efficiency was monitored by Western blotanalyses using antibodies specific for PrxII (LabFrontier). All thetransfectants were maintained in 10% FBS/DMEM. PrxII geneexpression in PrxII/siRNA-treated MCF+FIR3 radioresistant breastcancer cells was silenced after 48 hours following the transfection(Fig. 5A), whereas control oligos with scramble sequence was notable to inhibit the PrxII gene expression (Fig. 5B). We wanted toexamine whether the down-regulation of PrxII increased cellsensitivity towards ionizing radiation. Figure 5 showed theclonogenic survival assays for MCF+FIS4 radiosensitive cells,MCF+FIR3 radioresistant cells, and PrxII/siRNA-treated MCF+FIR3cells. Control oligos were not able to sensitize MCF+FIR3 cells(data not shown). The DMFs at 10% isosurvival were found to be1.91 and 1.57 for MCF+FIR3 and PrxII/siRNA-treated MCF+FIR3cells, respectively. Clonogenic survival assay in Fig. 5 showed thatdown-regulation of PrxII could sensitize radioresistant breastcancer cells to ionizing radiation. However, the inhibition of PrxIIexpression did not completely reverse the radioresistance pheno-type of MCF+FIR3 cells. Therefore, it is reasonable to assume thatthere are other factors that contribute to radiation resistance otherthan the PrxII protein. These results conclude that the siRNAconstructs of PrxII could down-regulate the expression of PrxII andpartially reverse the resistant phenotype of breast cancer cells,suggesting that PrxII might play an important role in radiation-resistant cancer cells.Peroxiredoxin II overexpression results in increased radio-

resistance to ionizing radiation. In this study, we addressed the

question whether overexpression of the antioxidant enzyme PrxIImay modulate or interfere with ionizing radiation–induced ROSgeneration leading to increased radiation resistance of MCF+FIS4breast cancer cells. For this purpose, we have generated transienttransfected MCF+FIS4 breast cancer cell lines overexpressing PrxII.To study the effect of PrxII overexpression in radiosensitiveMCF+FIS4 breast cancer cells exposed to ionizing radiation, thePrxII-overexpressing MCF+FIS4 breast cancer cells were exposed toionizing radiation at a dose of 2, 4, 8, and 10 Gy. Clonogenic assayswere used to compare the radioresistance level between vector-transfected MCF+FIS4 cells and PrxII-overexpressing MCF+FIS4cells (Fig. 6). The DMF at 10% isosurvival was found to be 1.2 forPrxII-overexpressing MCF+FIR4 cells. We found that ionizingradiation led to an increase resistant level in PrxII-overexpressingMCF+FIS4 cells compared with vector-transfected controlMCF+FIS4 cells. The result shows that overexpression of the PrxIIgene confers a significant protection against ionizing radiation–induced cell death.These results also imply indirectly that PrxII overexpression

interferes with ionizing radiation–induced ROS generation, leadingto the increased radioresistance. It is highly possible that over-expression of PrxII reduce the intracellular ROS to a noncytotoxiclevel, resulting in increased resistance of breast cancer cellsto ionizing radiation.

Discussion

Anticancer therapy is generally more effective in the early stages,whereas advanced tumors are usually resistant to the sametreatments. The molecular basis for this observation is not entirelyunderstood. Therefore, studies involving strategies to overcomethis resistance to chemotherapy and radiation or to increasecellular sensitivity to chemotherapy and radiation therapy arecrucial to the development of more effective treatments. Thehypothesis that key proteins functioning in signal transductionpathways are associated with tumor cellular radiation sensitivity orresistance unifies radiation biology research projects. By identifyingsuch proteins, determining their mechanisms of action, andtargeting their modulation by chemical or biological means, newstrategies may be developed for improving radiation therapytherapeutic ratios. Other investigations related to DNA damage andrepair and carcinogenesis provide a broader base for investigationsand additional scientific ideas and goals.The studies presented here show that the PrxII protein is up-

regulated in radiation-resistant breast cancer cells compared withradiation-sensitive breast cancer cells and plays an importantprotective role from oxidative radical damage. Increased expres-sion of PrxII was observed in tissues isolated from the head andneck cancer patients who did not respond to radiation therapy,whereas PrxII expression was weak in tissues from the patientswith regressed tumors. Enhanced expression of PrxII in UMSCC-11A cells was also observed after treatment with g-radiation (27).This increased expression conferred radiation resistance to cancercells because overexpression of PrxII protected cancer cells fromradiation-induced cell death, suggesting that blocking PrxIIexpression could enhance radiation sensitivity. Our results showedthat silencing PrxII gene expression by PrxII/siRNA treatment ofMCF+FIR3 radioresistant breast cancer cells leads to partialradiosensitization of MCF+FIR3 cells. The relevance of thismechanism was further confirmed by the increased radioresistancein PrxII-overexpressing cells when compared with vector control

Figure 3. A, Western blot analysis of Prx proteins in MCF+FIR3 radioresistantbreast cancer cells and MCF+FIS4 radiosensitive breast cancer cells.h-Actins were used as control for the experiments. Kodak 1D image analysessoftware was used to quantify the gel bands. 5, MCF+FIR3 breast cancer cells.n, MCF+FIS4 radioresistant breast cancer cells.





cells. However, we found that inhibition of PrxII protein expressiondid not completely reverse the radioresistant phenotype ofMCF+FIR3 cells, suggesting that silencing PrxII gene expressiondid not completely abrogate radiation-induced DNA or proteindamage and cell death. There are other factors that have beenshown to contribute to radiation resistance other than PrxII. Anumber of researchers have shown that markers of apoptosis,reduced proliferation, and signal transduction proteins correlatewith breast cancer response to chemotherapy. Several signalingpathways have also been suggested to be involved in cellularradiation sensitivity and response to ionizing radiation (28–32).PrxII has been previously known as a natural killer–enhancing

factor B (19). It is induced by various oxidative stimuli and plays animportant protective role from oxidative radical damage by ROS/RNS (33–35). PrxII-overexpressing cells are more resistant to theoxidative damage caused by H2O2, t-butyl-hydroperxide, andmethyl mercury (22). Its expression is correlated with resistanceto apoptosis induced by radiation therapy or the anticancer drugcisplatin (27, 36), highlighting the potential clinical importance ofPrxII in radiation resistance in cancer.The up-regulation of PrxII in radioresistant breast cancer cells

suggested that human Prx plays an important role in eliminating

the generation of reactive oxygen species (37, 38). Prxs represent amajor group of detoxification enzymes that use redox-sensitivecysteine, the active site peroxidatic cysteine, to reduce peroxides(39–41). In addition, Prxs catalyze decomposition of RNS (26, 42,43). The importance of thiol-mediated detoxification of oxidativestress that produces ROS/RNS has been addressed over the pastfew years. There are six mammalian Prx enzymes. All Prxs containat least one essential cysteine per 21- to 25-kDa subunit. They donot contain metal cofactors or other noncysteine redox centersand share the same basic catalytic mechanism, in which an activesite cysteine is oxidized to a sulfenic acid by the peroxidesubstrate. The catalytic efficiency of mammalian Prxs indicated bythe Kcat/Km is significantly lower than that of catalase orglutathione peroxidase. However, catalase is mainly in peroxisomesand glutathione peroxidase, which is mainly in cytosol, exists inlow amounts in most tissues. In contrast, five of the sixmammalian peroxiredoxins are abundant (0.2-0.4% of total solubleprotein) in cytosol, whereas PrxIII is exclusively localized inmitochondria. In addition to its role as a peroxidase, a body ofevidence has accumulated to suggest that individual members alsoserve divergent functions, which are associated with variousbiological processes, such as the detoxification of oxidants, cell

Figure 4. Multiple sequence alignment of PrxII proteins. The alignment was constructed with the use of the ClustalX program. *, positions that have a single, fullyconserved residue; :, one of the following ‘‘strong’’ groups is fully conserved: STA, NEQK, NHQK, NDEQ, QHRK, MILV, MILF, HY, and FYW; ., one of the following‘‘weaker’’ groups is fully conserved: CSA, ATV, SAG, STNK, STPA, SGND, SNDEQK, NDEQHK, NEQHRK, FVLIM, and HFY. Secondary structure elements aredepicted as helices (cylinders ) or strands (arrows ), with structural differences as disconnected elements. Positions of the aligned regions in the respective proteinsequences (right-hand sites). Protein sequences are colored by residue as follows: orange, GPST; red, HKR; blue, FWY; green, ILMV. S. cere.,Saccharomycescerevisiae ; Rattus,Rattus norvegicus ; Mus,Mus musculus ; Crice.,Cricetulus griseus ; Bos,Bos taurus ; Bruma,Brugia malayi ; Lito.,Litomosoides sigmodontis ;Ochengi,Onchocerca ochengi ; Volvulus,Onchocerca volvulus ; ECHGR,Echinococcus granulosus ; Schist.,Schistosoma mansoni . Peroxidatic (Cp) and resolving (CR)cysteines.

Cancer Research




proliferation, differentiation, and gene expression. It seems likelythat the divergence is due to unique molecular characteristicsintrinsic to each member (37, 39).It seems probable that PrxII is regulated in vivo by ROS,

because not only are some of the most potent inducers capableof generating free radicals by redox cycling, but induction ofPrxII by ROS would seem to represent an adaptive response asthese enzymes detoxify some of the toxic peroxide- and epoxide-containing metabolites produced within the cell by oxidative

stress. PrxII isoenzyme is overexpressed in radiation-resistantbreast cancer cells, and increased level of this enzyme probablycontributes to the radiation-resistant phenotype. It is similarlyknown that large amounts of ROS are produced as byproductsof ionizing radiation exposure. Therefore, it is tempting to spec-ulate that the cellular pathway leading to PrxII up-regulationcould be triggered by ionizing radiation and ROS exposure in asimilar and/or common way. In fact, we found that PrxII wasup-regulated after chronic exposure to ionizing radiation,pointing to a role for this protein in stress protection in cancercells. From these results, we suggest that stress-induced over-expression of PrxII increased radiation resistance via protectionof cancer cells from radiation-induced oxidative damage.Induction of the ROS-scavenging proteins could protect cellsfrom oxidative cytolysis and reduce cellular damage by regu-lating cellular redox status (44, 45). PrxII is not the sole factorresponsible for the resistant phenotype, because inhibition ofPrxII gene expression did not completely reversed the resistantphenotype. However, PrxII is a potential key player resulting inincreased resistance of breast cancer cells to ionizing radiation.This mechanistic study may provide a basis for new targets oftherapy based on this contribution to radiation sensitivity andlead to the identification of biomarkers to predict radiationsensitivity.

Acknowledgments

Received 12/28/2004; revised 3/28/2005; accepted 5/18/2005.Grant support: Beckman Foundation Research Fellowship (T. Wang) and Office of

Science, U.S. Department of Energy grant DE-FG02-05ER63945 (J.J. Li).The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby marked advertisement in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Carlotta Glackin, Binghui Shen, Detlef Schuman, Terry Lee, Mary Young,Roger Moore, Helen Ge, and Jeffery Wong (City of Hope National Medical Center) forinsightful discussion; Mian Zhou (Dr. Binghui Shen’s lab) for the PrxII plasmidconstruct; and Kevin Clark for his critical reading of the article.

Figure 5. A, Western analysis of PrxII in PrxII/siRNA-transfected MCF+FIR3 radioresistant cells. B, Western analysis of PrxII in scrambled siRNA-transfectedMCF+FIR3 cells. C, clonogenic survival assays for MCF+FIS4, MCF+FIR3 and PrxII/siRNA-treated MCF+FIR3 breast cancer cells. n, MCF+FIS4 breast cancer cells.y, radiation-resistant MCF+FIR3 cells. ., PrxII/siRNA-treated MCF+FIR3 resistant breast cancer cells. Clonogenic survival was calculated relative to the correspondingMCF+FIS4 control cells. The DMFs were calculated from the radiation doses corresponding to 10% survival.

Figure 6. Survival of MCF+FIS4/PrxII PrxII-overexpressing cells andMCF+FIS4/vector control cells as measured by the clonogenic assay afterexposure to various doses of radiation. E, MCF+FIS4/PrxII PrxII-overexpressingcells. n, MCF+FIS4/vector control cells. Clonogenic survival was calculatedrelative to the corresponding MCF+FIS4/vector control. The DMF was calculatedfrom the radiation doses corresponding to 10% survival.





References1. Guo G, Yan-Sanders Y, Lyn-Cook BD, et al. Manganesesuperoxide dismutase-mediated gene expression inradiation-induced adaptive responses. Mol Cell Biol2003;23:2362–78.

2. Rhee SG, Kim KH, Chae HZ, et al. Antioxidant defensemechanisms: a new thiol-specific antioxidant enzyme.Ann N Y Acad Sci 1994;738:86–92.

3. Weiss JF aK, KS. Antioxidant mechanisms in radiationinjury and radioprotection. In: Cellular antioxidantdefense mechanisms. Chow CK, editor. Boca Raton(FL): CRC Press; 1998. p. 163–89.

4. Hirose K, Longo DL, Oppenheim JJ, et al. Over-expression of mitochondrial manganese superoxidedismutase promotes the survival of tumor cells exposedto interleukin-1, tumor necrosis factor, selected anti-cancer drugs, and ionizing radiation. FASEB J 1993;7:361–8.

5. Sun J, Chen Y, Li M, et al. Role of antioxidant enzymeson ionizing radiation resistance. Free Radic Biol Med1998;24:586–93.

6. Suresh A, Tung F, Moreb J, et al. Role of manganesesuperoxide dismutase in radioprotection using genetransfer studies. Cancer Gene Ther 1994;1:85–90.

7. Finkel T. Oxidant signals and oxidative stress. CurrOpin Cell Biol 2003;15:247–54.

8. Durocher D, Jackson SP. DNA-PK, ATM and ATR assensors of DNA damage: variations on a theme? CurrOpin Cell Biol 2001;13:225–31.

9. Khanna KK, Lavin MF, Jackson SP, et al. ATM, a centralcontroller of cellular responses to DNA damage. CellDeath Differ 2001;8:1052–65.

10. Shiloh Y, Kastan MB. ATM: genome stability,neuronal development, and cancer cross paths. AdvCancer Res 2001;83:209–54.

11. Mikkelsen RB, Wardman P. Biological chemistry ofreactive oxygen and nitrogen and radiation-inducedsignal transduction mechanisms. Oncogene 2003;22:5734–54.

12. Pelicano H, Carney D, Huang P. ROS stress in cancercells and therapeutic implications. Drug Resist Updat2004;7:97–110.

13. Vaquero EC, Edderkaoui M, Pandol SJ, et al. Reactiveoxygen species produced by NAD(P)H oxidase inhibitapoptosis in pancreatic cancer cells. J Biol Chem 2004.

14. Kinnula VL, Lehtonen S, Sormunen R, et al. Over-expression of peroxiredoxins I, II, III, V, VI in malignantmesothelioma. J Pathol 2002;196:316–23.

15. Yanagawa T, Iwasa S, Ishii T, et al. Peroxiredoxin Iexpression in oral cancer: a potential new tumormarker. Cancer Lett 2000;156:27–35.

16. Li Z, Xia L, Lee LM, et al. Effector genes altered inMCF-7 human breast cancer cells after exposure to

fractionated ionizing radiation. Radiat Res 2001;155:543–53.

17. Smith ML, Hawcroft G, Hull MA. The effect of non-steroidal anti-inflammatory drugs on human colorectalcancer cells: evidence of different mechanisms of action.Eur J Cancer 2000;36:664–74.

18. Sauri H, Butterfield L, Kim A, et al. Antioxidantfunction of recombinant human natural killer enhanc-ing factor. Biochem Biophys Res Commun 1995;208:964–9.

19. Shau H, Kim A. Identification of natural killerenhancing factor as a major antioxidant in humanred blood cells. Biochem Biophys Res Commun 1994;199:83–8.

20. Kim AT, Sarafian TA, Shau H. Characterization ofantioxidant properties of natural killer-enhancing fac-tor-B and induction of its expression by hydrogenperoxide. Toxicol Appl Pharmacol 1997;147:135–42.

21. Sarafian TA, Rajper N, Grigorian B, et al. Cellularantioxidant properties of human natural killer enhanc-ing factor B. Free Radic Res 1997;26:281–9.

22. Shau H, Kim AT, Hedrick CC, et al. Endogenousnatural killer enhancing factor-B increases cellularresistance to oxidative stresses. Free Radic Biol Med1997;22:497–507.

23. Cha MK, Kim IH. Thioredoxin-linked peroxidasefrom human red blood cell: evidence for the existence ofthioredoxin and thioredoxin reductase in human redblood cell. Biochem Biophys Res Commun 1995;217:900–7.

24. Lim YS, Cha MK, Kim HK, et al. Removals ofhydrogen peroxide and hydroxyl radical by thiol-specificantioxidant protein as a possible role in vivo . BiochemBiophys Res Commun 1993;192:273–80.

25. Wood ZA, Poole LB, Karplus PA. Peroxiredoxinevolution and the regulation of hydrogen peroxidesignaling. Science 2003;300:650–3.

26. Chen L, Xie QW, Nathan C. Alkyl hydroperoxidereductase subunit C (AhpC) protects bacterial andhuman cells against reactive nitrogen intermediates.Mol Cell 1998;1:795–805.

27. Park SH, Chung YM, Lee YS, et al. Antisense ofhuman peroxiredoxin II enhances radiation-induced celldeath. Clin Cancer Res 2000;6:4915–20.

28. Jung M, Dritschilo A. NF-n B signaling pathway as atarget for human tumor radiosensitization. SeminRadiat Oncol 2001;11:346–51.

29. McKenna WG, Muschel RJ, Gupta AK, et al. The RASsignal transduction pathway and its role in radiationsensitivity. Oncogene 2003;22:5866–75.

30. Wang T, Guo G, Gao Q, et al. Expression of ErbB2enhances radiation-induced NF-nB activation. Onco-gene 2004;23:535–45.

31. Wang T, Hu YC, Dong S, et al. Co-activation of ERK,NF-n B and GADD45h in response to ionizing radiation.J Biol Chem 2005.

32. Wang T, Zhang X, Li JJ. The role of NF-nB in theregulation of cell stress responses. Int Immunopharma-col 2002;2:1509–20.

33. Kowaltowski AJ, Netto LE, Vercesi AE. The thiol-specific antioxidant enzyme prevents mitochondrialpermeability transition. Evidence for the participationof reactive oxygen species in this mechanism. J BiolChem 1998;273:12766–9.

34. Kowaltowski AJ, Vercesi AE, Rhee SG, et al. Catalasesand thioredoxin peroxidase protect Saccharomycescerevisiae against Ca(2+)-induced mitochondrial mem-brane permeabilization and cell death. FEBS Lett 2000;473:177–82.

35. Netto LE, Kowaltowski AJ, Castilho RF, et al. Thiolenzymes protecting mitochondria against oxidativedamage. Methods Enzymol 2002;348:260–70.

36. Chung YM, Yoo YD, Park JK, et al. Increasedexpression of peroxiredoxin II confers resistance tocisplatin. Anticancer Res 2001;21:1129–33.

37. Fujii J, Ikeda Y. Advances in our understanding ofperoxiredoxin, a multifunctional, mammalian redoxprotein. Redox Rep 2002;7:123–30.

38. Rhee SG, Chang TS, Bae YS, et al. Cellularregulation by hydrogen peroxide. J Am Soc Nephrol2003;14:S211–5.

39. Butterfield LH, Merino A, Golub SH, et al. Fromcytoprotection to tumor suppression: the multifactorialrole of peroxiredoxins. Antioxid Redox Signal 1999;1:385–402.

40. Netto LES, Chae HZ, Kang SW, et al. Removal ofhydrogen peroxide by thiol-specific antioxidant enzyme(TSA) is involved with its antioxidant properties. TSApossesses thiol peroxidase activity. J Biol Chem 1996;271:15315–21.

41. Rhee SG, Kang SW, Netto LE, et al. A family of novelperoxidases, peroxiredoxins. Biofactors 1999;10:207–9.

42. Bryk R, Griffin P, Nathan C. Peroxynitrite reductaseactivity of bacterial peroxiredoxins. Nature 2000;407:211–5.

43. Wong CM, Zhou Y, Ng RW, et al. Cooperation of yeastperoxiredoxins Tsa1p and Tsa2p in the cellular defenseagainst oxidative and nitrosative stress. J Biol Chem2002;277:5385–94.

44. Gius D. Redox-sensitive signaling factors and anti-oxidants: how tumor cells respond to ionizing radiation.J Nutr 2004;134:3213–4S.

45. Spitz DR, Azzam EI, Li JJ, et al. Metabolic oxidation/reduction reactions and cellular responses to ionizingradiation: a unifying concept in stress response biology.Cancer Metastasis Rev 2004;23:311–22.

Cancer Research




2005;65:10338-10346. Cancer Res Tieli Wang, Daniel Tamae, Thomas LeBon, et al. Breast Cancer CellsThe Role of Peroxiredoxin II in Radiation-Resistant MCF-7

Updated version

http://cancerres.aacrjournals.org/content/65/22/10338

Access the most recent version of this article at:

Cited articles

http://cancerres.aacrjournals.org/content/65/22/10338.full#ref-list-1

This article cites 41 articles, 7 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/65/22/10338.full#related-urls

This article has been cited by 4 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. (CCC)Click on "Request Permissions" which will take you to the Copyright Clearance Center's

.http://cancerres.aacrjournals.org/content/65/22/10338To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/content/65/22/10338.full#ref-list-1

http://cancerres.aacrjournals.org/content/65/22/10338.full#related-urls

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



the role of peroxiredoxin ii in radiation-resistant mcf-7 ... · radiation. (cancer res 2005;...

Documents