Sorafenib-Induced Hepatocellular Carcinoma Cell Death ...Molecular Medicine in Practice...

Molecular Medicine in Practice

Sorafenib-Induced Hepatocellular Carcinoma Cell DeathDependsonReactiveOxygenSpeciesProduction InVitroandIn Vivo

RomainCoriat1,2,3,CaroleNicco1,ChristianeCh�ereau1,OlivierMir2,3, J�eromeAlexandre1,2,3, StanislasRopert3,Bernard Weill1,2, Stanislas Chaussade2,4, Francois Goldwasser1,2,3, and Fr�ed�eric Batteux1,2

AbstractSorafenib is presently the only effective therapy in advanced hepatocellular carcinoma (HCC). Becausemost

anticancer drugs act, at least in part, through the generation of reactive oxygen species, we investigated

whether sorafenib can induce an oxidative stress. The effects of sorafenib on intracellular ROS production and

cell death were assessed in vitro in human (HepG2) and murine (Hepa 1.6) HCC cell lines and human

endothelial cells (HUVEC) as controls. In addition, 26 sera from HCC patients treated by sorafenib were

analyzed for serum levels of advanced oxidation protein products (AOPP). Sorafenib significantly and dose-

dependently enhanced in vitro ROS production by HCC cells. The SODmimic MnTBAP decreased sorafenib-

induced lysis ofHepG2 cells by 20%andofHepa1.6 cells by 75%comparedwithHCCcells treatedwith 5mg/L

sorafenib alone. MnTBAP significantly enhanced by 25% tumor growth in mice treated by sorafenib. On the

other hand, serum levels of AOPP were higher in HCC patients treated by sorafenib than in sera collected

before treatment (P < 0.001). An increase in serumAOPP concentration�0.2 mmol/L chloramine T equivalent

after 15 days of treatment is a predictive factor for sorafenib responsewith higher progression free survival (P <0.05) and overall survival rates (P < 0.05). As a conclusion, sorafenib dose-dependently induces the generation

of ROS in tumor cells in vitro and in vivo. The sera of Sorafenib-treated HCC patients contain increased AOPP

levels that are correlatedwith the clinical effectiveness of sorafenib and can beused as amarker of effectiveness

of the drug. Mol Cancer Ther; 11(10); 2284–93. �2012 AACR.

IntroductionHepatocellular carcinoma (HCC) is one of the most

common lethal malignancies, accounting for 626,000 newcases per year worldwide (1). The incidence of HCC isincreasing in the United States and Europe and it is thethird highest cause of cancer-related death globally (1).

HCC is generally diagnosed at an advanced stage orwith progression after locoregional therapy and has adismal prognosis, owing to the underlying liver diseaseand the lack of effective therapeutic options (2, 3). Sor-afenib (BAY-43-9006 Nexavar, Bayer PharmaceuticalsCorp. and Onyx Pharmaceuticals Inc.), an oral kinaseinhibitor of theVEGF, PDGF receptors, andRaf, decreases

tumor growth and inhibits angiogenesis in advancedHCC (4, 5). Sorafenib exerts a direct effect on tumor-cellproliferation and survival but mainly acts on endothelialcells to inhibit tumor angiogenesis (6, 7). Finally, it is theonly nonsurgical and nonradiological treatment thatimproves median overall survival and progression-freesurvival in patients with advanced HCC (5).

Sorafenib inhibits tumor cell proliferation and inducescell death through the inhibition of the RAF/MEK/ERKpathways (6). Human HCC-lesions display both highexpression and enhanced activity ofMAP kinase (MAPK)compared with adjacent non-neoplastic liver (8). TreatingHCC cells with aMEK inhibitor reduces cell proliferationand induces apoptosis (9). Several studies have linked theoncogenic potential of ERK to the activation of anti-apo-ptotic proteins, such as Bcl-2, Bcl-XL, Mcl-1, IAP, and therepression of pro-apoptotic proteins, such as Bad and Bim(10). Moreover, the overactivation of theMEK/ERK path-way in liver tumor cells confers resistance to TGF-b-induced cell death through impairing NOX4 up-regu-lation and subsequent production of reactive oxygenspecies (ROS),which is required for an efficientmitochon-drial-dependent apoptosis (11, 12). The role of ROS intumor biology has been extensively studied (13, 14). ROSare chemically reactive molecules that fulfill essentialfunctions in living organisms. A moderate increase in

Authors' Affiliations: 1Laboratoire d'immunologie, EA 1833, 2Universit�eParis Descartes, Sorbonne Paris-Cit�e, Facult�e deM�edecine, 3CERIA (Cen-tre for Research on Angiogenesis Inhibitors), Medical Oncology Depart-ment, and 4Service de gastroent�erologie, Hopital Cochin, AssistancePublique-Hopitaux de Paris (AP-HP), Paris, France

Corresponding Authors: Fr�ed�eric Batteux, Laboratoire d'Immunolo-gie, UPRES EA 1833, 27 rue du faubourg St Jacques 75679 Pariscedex 14, France. Phone: 33-0-1-58-41-20-07; Fax: 33-0-1-79-73-48-81; E-mail: [email protected]; and Romain Coriat,[email protected]

doi: 10.1158/1535-7163.MCT-12-0093

�2012 American Association for Cancer Research.

MolecularCancer

Therapeutics

Mol Cancer Ther; 11(10) October 20122284

on May 10, 2020. © 2012 American Association for Cancer Research. mct.aacrjournals.org Downloaded from

Published OnlineFirst August 17, 2012; DOI: 10.1158/1535-7163.MCT-12-0093

http://mct.aacrjournals.org/

ROScanpromote cell proliferation anddifferentiation (15,16), whereas excessive amounts of ROS can cause oxida-tive damages to lipids, proteins and DNA and lead to celldeath (13, 14). Several chemotherapeutic drugs exert theircytotoxic effects through the generation of ROS (17). Assorafenib inhibits the MEK/ERK pathway that controlsROS production in HCC, one of the mechanisms throughwhich sorafenib could exert its cytotoxic effects is therestoration of the susceptibility of HCC cells to ROS-induced lysis.To determine whether ROS production is essential to

sorafenib cytotoxicity, we studied the levels of ROS inHepG2humanandHepa 1.6murineHCCcells exposed tosorafenib in vitro. We then investigated the consequencesof the selective inhibition of various ROS in vitro and in anin vivo mouse model of HCC. Finally, we showed ex vivothat the elevation of advanced oxidized protein productsin patients’ sera 15 days following the initiation of sor-afenib chemotherapy is correlated with the effectivenessof the drug.

Materials and MethodsAnimalsC57/BL6 female mice between 6 and 8 weeks of age

were used in all experiments (Harlan). Animals receivedhuman care in compliance with institutional guidelines(Accreditation number for animal research establishment:A-75-14-05, agreement number ofCaroleNicco to conductanimal research: 75-1302).

Chemicals, cell lines, and cultureAll chemicals were from Sigma (Saint Quentin

Fallavier) except for sorafenib (BAY-43-9006 Nexavar,Bayer Pharmaceuticals Corp. and Onyx Pharmaceuti-cals Inc.). Sorafenib (4-(4-(3-[4-chloro-3-(trifluoromethyl)-phenyl]ureido)phenoxy)-N2-methylpyridine-2-carboxa-mide-4-methylbenzenesulfonate) has a molecular formu-la of C21H16ClF3N4O3C7H8O3S and a molecular weight of637.0 (18). The chemical structure of sorafenib is presentedin Fig. 1. Hepa 1.6 (mouse HCC cell line), HepG2 (humanhepatocellular liver carcinoma cell line) were obtainedfromAmerican TypeCulture Collection.HUVEC (humanumbilical venous endothelial cells) were obtained bydigestion of umbilical cords with 0.1% collagenase aspreviously described (19). All cell lines were cultured in

DMEM/Glutamax-I supplemented with 10% heat-inacti-vated FCS and antibiotics (Life Technologies). All celllines were immediately expanded and frozen down suchthat all cell lines could be restarted every 3 months from afrozen vial of the same batch of cells. No further authen-tication was done in our lab. All cell lines were routinelytested to rule out Mycoplasma infection of cells.

Cellular Production of O2��, H2O2, and NO

Cells (2 � 104 cells per well) were seeded in 96-wellplates (Costar, Corning, Inc.) and incubated for 48 hourswith various concentrations of sorafenib alone or in asso-ciation with either 400 mmol/LN-acetylcysteine (NAC), aprecursor of reduced glutathione (GSH), or 100 mmol/Lmanganese (III) tetrakis (4-benzoic acid) porphyrin(MnTBAP), a mimic of superoxide dismutase (SOD), or4 mmol/L diethyldithiocarbamate (DDC), a SOD inhibi-tor, or 40 mmol/L diphenyleneiodonium chloride (DPI),an inhibitor of flavonoid-containing enzymes likeNADPH oxidase, or 40 mmol/L rotenone (inhibitor ofmitochondrial complex I) or 40 mmol/L antimycine, aninhibitor ofmitochondrial complex III, orwith 40mmol/Lallopurinol, an inhibitor of xanthine oxidase, or withculture medium alone (19). Levels of intracellular super-oxide anion O2

��, hydrogen peroxide H2O2 and nitricoxideNOwere assessed spectrofluorometrically (PackardBioscience) by oxidation of dihydroethidium (DHE,Molecular Probess) 20,70-dichlorodihydrofluorescein dia-cetate (H2DCFDA, Molecular Probes) 4,5-diaminofluor-escein diacetate (DAF2-DA), respectively (20). The levelsof O2

��, H2O2, and NOwere calculated in each sample asfollows: ROS rate (OD/min/106 cells) ¼ (OD60 minutesafter drug first contact – initial OD)/60 minutes/numberof viable cells as measured by the crystal violet assay.

Intracellular GSH measurementLevels of intracellular GSH were measured spectro-

fluorimetrically by monochlorobimane staining (21).Briefly, cells (2 � 104 cells per well) were seeded in 96-well plates (Costar, Corning, Inc.) and incubated for 48hours with various concentrations of sorafenib alone or inassociation with eitherMnTBAP orNAC or culturemedi-um alone. Cells were then washed once with PBS andincubated with 50 mmol/L monochlorobimane. Fluores-cence intensity was measured after 15 minutes at 37�C.Excitation and emission wavelengths were 380 and 485nm, respectively. IntracellularGSH levelwas expressedasarbitrary units of fluorescence intensity.

In vitro cell proliferation and viability assaysHepG2,Hepa 1.6, andHUVEC (2� 104 cells/well)were

seeded into 96-well plates and incubated for 48 hoursin complete DMEM medium with increasing amounts ofsorafenib alone, or in association with either 400 mmol/LNAC, or with 100 mmol/L MnTBAP or 200 mmol/L nitricoxide synthase inhibitorNG-nitro-L-argininemethyl ester(L-NAME). Cell proliferation was determined by pulsingthe cells with [3H]thymidine (1 mCi/well) during the lastFigure 1. Molecular structure of sorafenib.

Sorafenib Induces Reactive Oxygen Species Production

www.aacrjournals.org Mol Cancer Ther; 11(10) October 2012 2285




16 hours of culture. Cell viability was evaluated by thecrystal violet assay (20). Results were expressed as per-centages of viable cells compared with untreated cells(that represent 100% viability).

In vivo antitumor activity of sorafenib treatmentHepa 1.6 cells (2 � 106) were injected subcutaneously

into the back of the neck of C57/BL6 mice. When thetumor reached a mean size of 200 to 500 mm3, mice wererandomized (around day 10) into each experimental andcontrol groups according to tumor size, to start the treat-ment with a similar mean size in each group. Mice weretreatedwith either sorafenibbyoral route (50mg/kg/day,n¼ 7), or intraperitoneal MnTBAP (15 mg/kg three timesa week, n¼ 7), or PBS as control group, n ¼ 7. Tumor sizewasmeasuredwith a calliper rule every 2 days. The tumorvolumewas calculatedas follows:TV (mm3)¼ (L�W2)/2,where L is the longest and W the shortest radius of thetumor in millimeters. Results were expressed as means oftumor volumes � SEM (n ¼ 7 mice in each group).

Patients and seraTwenty-six patients with advanced-stage HCC were

enrolled in the study and treated with sorafenib at avalidated dose of 400 mg twice daily until there wasevidence of disease progression. Patients were classifiedas having advanced disease if theywere not eligible for orhad disease progression after surgical or locoregionaltherapies (5). All patients gave written informed consent.The studywas in accordance with Declaration of Helsinkiand was approved by the Comit�e de protection des per-sonnes d’Ile de France.All patients hadaChild-Pugh classand aMELDscore evaluation at baseline. TheMELDscorewas 3.8� loge(bilirubin [mg/dL])þ 11.2� loge(INR)þ 9.6� loge(creatinine [mg/dL]) þ 6.4 � (etiology: 0 if chole-static or alcoholic, 1 otherwise) as previously published(22). Sorafenib was administered at 50% of the planneddose if any severe adverse event related to the drugoccurred, and in frail patients (5). Treatment interruptionsand up to two dose reductions (first to 400 mg once daily,and then to 400 mg every 2 days) were decided in case ofdrug-related adverse effect. Treatment continueduntil theoccurrence of either radiologic progression as defined byRECIST criteria (23), or symptomatic progression. A ret-rospective reviewof electronicmedical records of patientswith advanced HCC treated with sorafenib was carriedout in our institution fromOctober 2008 toDecember 2010.Further analysis of clinical toxicity and interpretation ofbody composition from computed tomography imageswere approved by the Institutional Review Board. Serawere collected fromeachpatient 7 days before and 14daysafter sorafenib start.

Patients treatment and toxicity assessmentAdult patients with advanced HCC received sorafenib

at a starting dose of 200 mg twice a day or 400 mg twice aday according to their ECOG PS and comorbidities, atthe discretion of the treating physician. Patients were

treated in an outpatient setting, and toxicity wasassessed at visits on days 14 and 28 after the initiation ofsorafenib (or before if clinically indicated), then monthly.Sorafenib was reduced to 200 mg twice a day in caseof severe toxicity (grade 3 or 4 toxicity according to theNCI-CTCAE v3.0), except for patients with grade 3hypertension in whom anti-hypertensive drugs wereintroduced according to current guidelines (24). If toxicitywas not thereby resolved, treatment was terminated andif so, patients were returned to the initially scheduleddose. A dose-limiting toxicity was defined as any toxicityleading to dose reduction or to termination of treatment.Only dose-limiting toxicities occurring during the firstmonth of treatment were examined for the presentanalysis.

Assay of advanced oxidation protein products in seraAdvanced oxidation protein products (AOPP) were

measured by spectrophotometry as previously described(25). The assay was calibrated using chloramine-T. Theabsorbance was read at 340 nm on a microplate reader(Fusion, PerkinElmer). AOPP concentrations were expre-ssed as mmol/L of chloramine-T equivalents.

Assay of nitrite and nitrate levels in seraNitrite levels were measured by direct Griess reaction

(26), which is the simplest and most commonly usedassay. Nitrate levels were measured by subtraction of thenitrite levels from total NO measured by indirect Griessreaction. Nitrate serum levels reflect NO synthesis inpatients treated by sorafenib.

Statistical analysisThe statistical significance of differences between

experimental treated groups and untreated controls wasanalyzed by Student’s t test for comparison of means. Alevel ofP < 0.05was accepted as significant. �,P< 0.05; ��,P< 0.02; ��, P < 0.01; ��, P < 0.001 versus controls.Descriptive statistics were used to report patients anddisease characteristics [median, ranges, 95%CI], and treat-ment-induced toxicities. PFS andOSweremeasured fromthe date of first administration of sorafenib to the date ofdisease progression or death for the former, and the dateof death for the latter. Survival analyses were carried outusing the Kaplan–Meier method with the log-rank test.Toxicity and activity data were collected until December15, 2011.Calculationswere donewithNCSS2007 software(NCSS).

ResultsROS production by sorafenib-stimulated HCC celllines is dose-dependent

Sorafenib increased the production of O2��, NO andH2

O2 in HepG2, Hepa 1.6 and HUVEC cell lines (Fig. 2). Atthe highest concentration of sorafenib tested, O2

�� levelswere increased by 33% (P< 0.05), 300% (P< 0.001) and 40%(P < 0.05) in HepG2, Hepa 1.6 and HUVEC cell lines,respectively. Adding 400 mmol/L NAC to sorafenib did

Coriat et al.

Mol Cancer Ther; 11(10) October 2012 Molecular Cancer Therapeutics2286




not modify O2�� production in HepG2, Hepa 1.6 and

HUVEC cells (Fig. 2A). Adding 100 mmol/L MnTBAP to5 mg/L sorafenib, inhibited O2

�� production in the threecell lines tested (P < 0.05 for HepG2, Hepa 1.6 andHUVEC). Adding 40 mmol/L DPI to 5 mg/L sorafenibinhibited O2

�� production in HepG2 cells, Hepa 1.6 andHUVEC cells (P < 0.05 for HepG2, Hepa 1.6 andHUVEC).

Superoxide anions can originate from the mitochondri-al complex I or III of the respiratory chain, or from thecytosolic enzymesNAD(P)H oxidase or xanthine oxidase.Adding rotenone andantimycin that inhibit themitochon-drial respiratory chain complexes I and III, respectively,increased the O2

�� production in HepG2 cells, Hepa 1.6and HUVEC cells (P < 0.05) without modifying the

Figure 2. ROS production byHepG2, Hepa 1.6, and Huvec cells.HepG2, Hepa 1.6, and HUVEC wereincubated for 48 hours withincreasing concentrations ofsorafenib alone or in associationwith N-acetylcysteine, MnTBAP,DPI, allopurinol, DDC, rotenone, orantimycine. Measurements ofsuperoxide anion (O2

��; A and B),hydrogen peroxide (C), and nitricoxide (D) were carried out byspectrofluorometry as described inMaterial and Methods. Results areexpressed as mean OD � SEM.Data fromat least three independentexperiments have been pooled.Statistical significance: �, P < 0.05versus culture medium alone (basallevel). Error bars represent SEM.A.U., arbitrary unit.






production of superoxide anion induced by sorafenibbecause the O2�- increase in cells treated with sorafenibwas constant whether rotenone and antimycinwas addedor not (Fig. 2B). Adding 40 mmol/L allopurinol, a xan-thine oxidase inhibitor decreased the basal production ofO2

�� in HepG2 cells, Hepa 1.6 andHUVEC cells (P < 0.05)without modifying the production of superoxide anioninduced by sorafenib (Fig. 2B) Adding 40mmol/LDDC, aSOD inhibitor, enhancedO2

�� production inHepG2 cells,Hepa 1.6 and HUVEC cells (P < 0.01) without modifyingthe production of superoxide anion induced by sorafenib.

Adding400mmol/LNACto 5mg/Lsorafenib inhibitedH202production in the cell lines tested (P<0.05 forHepG2,P < 0.001 for Hepa 1.6 and P < 0.001 for HUVEC, versus 5mg/L sorafenib alone; Fig. 2C). Adding 100 mmol/LMnTBAP still augmented the generation of H202 in thecell lines (at 5 mg/L sorafenib: P < 0.001 for HepG2, P <0.05 for Hepa 1.6 and P < 0.05 for HUVEC, versus 5 mg/Lsorafenib in association with NAC. NO productioninduced by sorafenib was dose-dependent (Fig. 2D) andincreased with the concentration of sorafenib (at 5 mg/Lsorafenib: P < 0.05 for HepG2, Hepa 1.6 and HUVEC).

ROS production induced by sorafenib appears in thefirst hours. Sorafenib increased the production of O2

��

and H2O2 in HepG2, Hepa 1.6 and HUVEC cell lines(Fig. 3). At 2.5 mg/L of sorafenib, O2

�� levels reach asteady state after two hours in HepG2, Hepa 1.6 andHUVEC cell lines. Adding 100 mMMnTBAP to 2.5 mg/L

sorafenib, reduced O2�- production in the three cell lines

tested caused by O2�- dismutation by MnTBAP SOD

activity (P < 0.05 for HepG2, Hepa 1.6 and HUVEC).Adding 40 mmol/L DPI to 2.5 mg/L sorafenib inhibitedO2

�� production starting after 2 hrs in HepG2 cells,Hepa 1.6 and HUVEC cells (P < 0.05 for HepG2, Hepa1.6 and HUVEC). Adding 400 mmol/L NAC to 2.5 mg/Lsorafenib inhibited H202 production starting after 6 hrsin HepG2 and Hepa1.6 and after 12 hrs in HUVEC cells(P < 0.05 for HepG2, Hepa 1.6 and HUVEC, versussorafenib alone; Fig. 3B). Adding 100 mmol/L MnTBAPpotentiated the generation of H2O2 in the cell linesstarting after 2 hrs (P < 0.05 for HepG2 and Hepa 1.6and P < 0.001 for HUVEC, versus sorafenib).

Sorafenib exerts dose-dependent cytostatic andcytotoxic effects in vitro

Incubation of HepG2, Hepa 1.6 cells or HUVEC withincreasing amounts of sorafenib dose-dependentlydecreased their proliferative rates and their viability. At5 mg/L, sorafenib decreased the proliferative rates by98%, 90% and 26%, respectively (P < 0.001 for HepG2 andHepa 1.6, P < 0.05 for HUVEC; Fig. 4A), and their viabilityrates by 19%, 24% and 18%, respectively (P < 0.05 forHepG2, Hepa 1.6 and HUVEC; Fig. 4B). To determine themetabolic pathways through which sorafenib exerts itscytostatic and cytotoxic effects of sorafenib, the moleculewas associated with MnTBAP, or NAC, or L-NAME).

Figure 3. Kinetic analysis of ROSproduction by HepG2, Hepa 1.6,and Huvec cells. HepG2, Hepa1.6, and HUVEC were incubatedwith 2.5mg/L sorafenib alone or inassociation with 400 mmol/L N-acetylcysteine, 100 mmol/LMnTBAP, or 40 mmol/L DPI.Measurements of superoxideanion (O2

��; A) and hydrogenperoxide (B) were realized byspectrofluorometry as describedin Material and Methods. Resultsare expressed as mean OD �SEM. Data from at least threeindependent experiments havebeen pooled. Statisticalsignificance: �, P < 0.05 versusculture medium alone (basallevel). Error bars represent SEM.

Coriat et al.





Adding 100 mmol/L MnTBAP to 5 mg/L sorafenib sig-nificantly inhibited the cytostatic effects (P < 0.001 forHepG2 and Hepa 1.6, P < 0.05 for HUVEC; Fig. 4A) andcytotoxic effects of sorafenib (P < 0.001 for HepG2, Hepa1.6 and HUVEC; Fig. 4B). By contrast, neither 400mmol/LNAC nor L-NAME altered the cytostatic or cytotoxiceffects of sorafenib (Fig. 4A, B). In cells exposed to 5mg/L sorafenib for 48 hours, the GSH pool was depletedby 40%, 53% and 44% in HepG2, Hepa 1.6 and andHUVEC cells, respectively (P <0.001 for HepG2, P <0.001 for Hepa 1.6, P < 0.001 for HUVEC; Fig. 4C). Asexpected,NACpartially reversed thedepletion ofGSH (at5mg/L sorafenib:P< 0.05 forHepG2,P< 0.05 forHepa 1.6and P <0.01 for HUVEC), and MnTBAP worsened the

depletion ofGSH inducedby 5mg/L sorafenib (at 5mg/Lsorafenib: P < 0.001 for HepG2, P < 0.001 for Hepa 1.6and P < 0.001 for HUVEC), (Fig. 4C); As MnTBAP butneither NAC nor L-NAME could abrogate the cytostaticand cytotoxic effects of sorafenib on the cell lines tested,O2

�� is the main ROS induced by sorafenib and respon-sible, at least in part, for its cytostatic and cytotoxicproperties.

Effect of sorafenib on tumor growth in vivoMice bearing Hepa 1.6 tumors were treated with either

sorafenib alone or in association with MnTBAP (Fig. 5).Mice treated with sorafenib alone presented significantlysmaller tumors than untreated mice (-79%, P < 0.001 at

Figure 4. Proliferative responses (A)and viability (B) of HepG2, Hepa 1.6,and HUVEC cells incubated withvarious concentrations of sorafenib.Cell proliferation was expressed aspercent�SEM versus cells in culturemedium alone (100% proliferationrate). Viability was expressed aspercent�SEM versus cells in culturemedium alone (100% viability). Datafrom at least three independentexperiments have been pooled.Differences between untreated andsorafenib-treated cells were alwayssignificant for at least twoconcentrations of sorafenib.Proliferation rateofHepG2,Hepa1.6,and HUVEC cell lines are expressedfollowing incubation with increasingconcentrations of sorafenibalone, in association with either400 mmol/L N-acetylcysteine, or200 mmol/L L-NAME, or 100 mmol/LMnTBAP. Cellular proliferation wasmeasured by thymidineincorporation. Cell viability wasmeasured by the crystal violet assay.Levels of intracellular reducedglutathione (C) in HepG2, Hepa 1.6,and Huvec. The cells were incubatedfor 48 hours with increasingconcentrations of sorafenib eitheralone or in association with400 mmol/L N-acetylcysteine or with100 mmol/L MnTBAP. GSHlevels were determined byspectrofluorometry as described inthe Materials and Methods.Intracellular GSH level wasexpressed as arbitrary units offluorescence intensity. Results areexpressed as mean OD� SEM. Datafrom at least three independentexperiments have been pooled.Statistical significance: �, P < 0.05versus culture medium alone (basallevel). Error bars represent SEM.






day 22). MnTBAP did not inhibit tumor growth whenadministered alone (-11% at day 22, P ¼ NS versusuntreated mice). Tumor growth was still significantlyslower when sorafenib was associated with MnTBAP(-51%atday22,P<0.001), butMnTBAPreduced sorafenibeffect on tumor growth by 28% at day 22 compared withmice treated with sorafenib alone (P < 0.05).

Assayofnitrate andAOPP levels in sera frompatientswith HCC treated by sorafenib

Twenty-six patients treated by sorafenib for HCC wereincluded in the present study. Patients’ characteristics arepresented in Table 1. No difference was observed innitrate serum levels between the beginning of the treat-ment (day 0) and day 15 (mean: 22.8 versus 22.6; P¼NS).As nitrite and/or nitrate serum reflectNOsynthesis, theseresults confirm that NO does not mediate sorafenib anti-tumor effectiveness. Serum AOPP levels in patients trea-ted by sorafenib were higher after 15 days’ treatment thanbefore treatment (P < 0.05; Fig. 6A). The patients weredistributed into twogroups: one groupwith an increase inAOPP0.2 mmol/L chloramine T equivalent (<0.2 AOPP,n ¼ 10), and one group with �0.2 mmol/L chloramine Tequivalent (�0.2 AOPP, n ¼ 16). The latter group com-prised patients in whom sorafenib effectively generatedROS responsible for protein oxidation. No difference inthe adverse effects of sorafenibwas observed between thetwo groups and nodifference in limiting toxicity (P¼NS),diarrhea (P ¼ NS), hypertension (P ¼ NS), hand footsyndrome (P¼NS) nor asthenia (P¼NS; SupplementaryTable 1). Best response at 3 months was significantlyhigher in �0.2 AOPP group than in <0.2 AOPP groupconsidering mRECIST criteria or Choi criteria (mRECISTcriteria: 30%versus 13%;P<0.05;Choi criteria: 60%versus13%; P < 0.01)(Fig. 6B). Median progression-free survival(Fig. 6C)was significantly higher in the� 0.2AOPPgroup(174 days [range: 34 to 482]) than in the0.2 AOPP group(50 days [range: 9 to 136], P < 0.05). Median overall freesurvival (Fig. 6D) was significantly higher in the �0.2

AOPP group (315 days [range: 138 to 1075]) than in the0.2 AOPP group (191 days [range: 43 to 705], P < 0.05).

DiscussionThis study reports for the first time that the cytotoxic

and cytostatic effects of sorafenib are mediated via O2��.

Sorafenib has a well-documented effectiveness in HCC

Figure 5. MnTBAP abrogated the antitumor effects of sorafenib in miceimplanted with Hepa 1.6 tumor cells. Mice were injected with HCC cells(Hepa 1.6) and then received either sorafenib alone or sorafenib inassociation with MnTBAP, MnTBAP alone, or vehicle alone.

Table 1. Baseline characteristics ofhepatocellular carcinoma patients treated withsorafenib. The patients were distributed into twogroups: one group with an increase in AOPP0.2 mmol/L chloramine T equivalent betweenday 0 and day 15 (0.2 AOPP), and one groupwith �0.2 mmol/L chloramine T equivalent(�0.2 AOPP)

AOPPd <0.2group

AOPPd �0.2group

Patients 10 16

Age, years, Average(SD)

63.1 (15.9) 65.6 (11.0)

Sex, Male/Female (%) 80/20 88/12ECOG 0/1/� 2, (%) 30/50/20 44/50/6Specific metastaticsites, n

Lung 4 5Liver 10 15Other 4 7Duration of disease,months, Average(SD)

11.2 (9.3) 13.4 (11.9)

Cause of Cirrhosis, nHepatitis C only 4 5Hepatitis B only 3 7Alcohol only 2 1Other 1 3

Child PughclassificationA/B/C, n

8/2/0 11/5/0

MELD scorea 10 (4–16) 12 (6–19)Weight (kg), Average(SD)

66.8 (13.7.8) 75.6 (9.8)

BMI, Average (SD) 23.9 (5.8) 25.5 (4.0)Biologic ResultsAlbuminemia (g/dL) 31.8 (22–36) 37.7 (31–44)Pre-albuminemia 0.10 (0.05–0.18) 0.18 (0.09–0.28)Bilirubinemia 31.0 (4–81) 12.7 (4–22)ALT 80.9 (31–143) 48 (15–127)AST 125.8 (24–418) 77 (21–168)

Abbreviations: SD, standard deviation; ECOG, EasternCooperative Oncology Group Criteria Performance Status;BMI, body mass index (weight/height2).aValues reported are median (25th percentile to 75thpercentile).

Coriat et al.





cell lines and is presently the only validated treatment forHCC (6, 27). As previously published, we confirm thatsorafenib exerts dose-dependent cytostatic and cytotoxiceffects in vitro (6) and report that HCC cell lines producevarious ROS following exposure to sorafenib in vitro.Furthermore, the concentration of AOPP in the sera frompatients treated by sorafenib is a predictivemarker of bothprogression-free survival and overall survival.Sorafenib exerts cytostatic and cytotoxic effects onHCC

cell lines and on endothelial cells. Concomitantly, sorafe-nib can enhance the production of superoxide anion,hydrogen peroxide, and nitric oxide in those cells. Theuse of specific ROSmodulators shows that dismutation ofsuperoxide anions induced by a SOD mimic (MnTBAP;ref. 28) prevents sorafenib cytotoxicity whereas neitherthe H2O2 scavenger NAC (29) nor the NO synthase inhib-itor L-NAME (30) have any effect on sorafenib cytotoxic-ity. Those data highlight the prominent role of superoxideanion in sorafenib antiproliferative and cytotoxic effects.Superoxide anion is the first radical produced withinthe cells. It can interact with various substrates such asproteins, DNA or lipids and cause irreversible cellulardamages (31). Direct antiproliferative and cytotoxiceffects of superoxide anion have already been observedin various types of tumor cells such as, for example, B16mouse melanoma (32). Superoxide anion can originatefrom the cytosolic enzymes NAD(P)H oxidase, the xan-thine oxidase or the complexes I and III of the mitochon-drial respiratory chain. The xanthine oxidase inhibitor,

allopurinol, and the complexes I and III of the mitochon-drial respiratory chain inhibitors, antimycine and rote-none, had no significant impact on sorafenib-inducedsuperoxide anion production. The use of DPI that inhibitsthe NAD(P)H oxidase clearly shows that in sorafenib-treated cells O2

�� is produced by NAD(P)H oxidase.These data are in line with the finding that the treatmentof HCC cells withNAD(P)H oxidase inhibitors, includingDPI, significantly blunts both the generation of ROS andthe induction of apoptosis (33). This suggests thatNAD(P)H oxidase is the major source of ROS in human HCC cells(34). The activation of NADP(H) oxidase can be directlycaused by the molecule sorafenib. It can also be strength-ened in vivo by the VEGF-mediated effects of sorafenib.Indeed, while inhibiting VEGF signalling and reducingneoangiogenesis (35), sorafenib leads to tumor hypoxiaand triggers the production of ROS (35).

Our work emphasizes the involvement of the oxidativestress in the treatment of HCC patients by sorafenib andprovides a new serum marker to evaluate the effective-ness of sorafenib. As sorafenib induces ROS that canmodulate tumor cell death and proliferation (36, 37), wequantified the concentration of oxidized proteins in thesera of HCCpatients. Proteins or lipids or DNA can be thetargets of the oxidative burst (38). It has been shown thatthe concentration of advanced oxidation protein productcan be used to monitor the in vivo oxidative stress inhumans (39) as it reflects the intensity of the oxidativeinsult in various pathologic conditions. In our study, the

Figure 6. A, measurement ofadvanced oxidation protein products(AOPP) levels in 26 sera frompatientseither before sorafenib or after 15days of treatment. AOPP weremeasured by spectrophotometry aspreviously described (25). B, bestresponse rates at 3 monthsconsidering mRECIST criteria orChoi criteria. Best response ratecorrespond to either completeresponse, partial response, or stabledisease. Progression-free survival(C) andoverall free survival (D) inHCCpatients receiving sorafenib therapy.






initial level ofAOPPbefore any treatment is not predictivefor sorafenib effectiveness (data not shown).However, theelevation of serum AOPP after 15 days’ treatment issignificantly correlated with sorafenib clinical effective-ness. Indeed, a serum level of AOPP � 0.2 mmol/L chlo-ramine T equivalent (�0.2 AOPP) correlates with bothprogression-free survival and overall survival suggestingan efficient dose exposure. Accordingly, given the inter-individual variability of the values of the sorafenib areaunder the plasma concentration–time curve over 12 hours(AUC), the lack of effect observed in patients with 0.2AOPP may be linked to an underexposure (40). To date,the bulk of data on dose exposure to sorafenib is growingand suggests a potential benefit for sorafenib dose-esca-lation strategies in renal cell carcinoma (41). Consideringthose results and the sorafenib pharmacokinetic analysis,the assay of serumAOPPmight become a useful and earlyindicator of sorafenib effectiveness.

In conclusion, sorafenib effectiveness is mediated byROS and particularly by O2

�� production. The effect ofthis production on protein oxidation is an early predictivemarker of sorafenib effectiveness. In response to sorafenib,HCC cancer cells produce massive amounts of ROS that inreturn induce tumor cell death (6). No or weak ROSproduction predicts a lack of effectiveness. Thus, wehypothesize that the effectiveness of sorafenib in HCCpatients depends on the balance between the ability oftumor cells to produce ROS and their ability to tolerate an

exposure to sorafenib. Therefore, monitoring ROS produc-tion induced by sorafenib in the absence of limiting toxicitycould allow planning a dose escalation of the molecule.Clinical studies should be designed to best this hypothesis.

Disclosure of Potential Conflicts of InterestF. Goldwasser has acted as a paid consultant for Bayer. No potential

conflicts of interest were disclosed by the other authors.Our research article involved patients and written informed consent

was obtained from all patients. The study was in accordance with theDeclaration ofHelsinki andwas approvedby theComit�e de protection despersonnes d’Ile de France.

Authors' ContributionsConception and design: R. Coriat, O. Mir, F. BatteuxDevelopment of methodology: R. Coriat, C. Nicco, F. BatteuxAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): R. Coriat, C. Nicco, C. Chereau, O. Mir,J. Alexandre, S. Ropert, F. Goldwasser, F. BatteuxAnalysis and interpretation of data (e.g., statistical analysis, biostatis-tics, computational analysis): R. Coriat, C. Nicco, O. Mir, J. Alexandre,S. Chaussade, F. Goldwasser, F. BatteuxWriting, review, and/or revision of the manuscript: R. Coriat, C. Nicco,O. Mir, B. Weill, S. Chaussade, F. Goldwasser, F. BatteuxAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): C. Nicco, C. ChereauStudy supervision: B. Weill, F. Goldwasser, F. Batteux

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

Received February 9, 2012; revised June 22, 2012; accepted July 17, 2012;published OnlineFirst August 21, 2012.

References1. ParkinDM,BrayF, Ferlay J, Pisani P.Global cancer statistics, 2002.CA

Cancer J Clin. 2005;55:74–108.2. Llovet JM, Burroughs A, Bruix J. Hepatocellular carcinoma. Lancet

2003;362:1907–17.3. Bruix J, Sherman M, Llovet JM, Beaugrand M, Lencioni R, Burroughs

AK, et al. Clinical management of hepatocellular carcinoma. Conclu-sions of the Barcelona-2000 EASL conference. European Associationfor the Study of the Liver. J Hepatol 2001;35:421–30.

4. Wilhelm SM, Carter C, Tang L, Wilkie D, McNabola A, Rong H, et al.BAY 43–9006 exhibits broad spectrum oral antitumor activity andtargets the RAF/MEK/ERK pathway and receptor tyrosine kinasesinvolved in tumor progression and angiogenesis. Cancer Res2004;64:7099–109.

5. Llovet JM, Ricci S, Mazzaferro V, Hilgard P, Gane E, Blanc JF, et al.Sorafenib in advanced hepatocellular carcinoma. N Engl J Med 2008;359:378–90.

6. Liu L, Cao Y, Chen C, Zhang X, McNabola A, Wilkie D, et al. Sorafenibblocks the RAF/MEK/ERK pathway, inhibits tumor angiogenesis, andinduces tumor cell apoptosis in hepatocellular carcinoma model PLC/PRF/5. Cancer Res 2006;66:11851–8.

7. Carmeliet P, Dor Y, Herbert JM, Fukumura D, Brusselmans K,Dewerchin M, et al. Role of HIF-1a in hypoxia-mediated apoptosis,cell proliferation and tumour angiogenesis. Nature 1998;394:485–90.

8. Schmidt CM, McKillop IH, Cahill PA, Sitzmann JV. Increased MAPKexpression and activity in primary human hepatocellular carcinoma.Biochem Biophys Res Commun 1997;236:54–8.

9. Wiesenauer CA, Yip-Schneider MT, Wang Y, Schmidt CM. Multipleanticancer effects of blocking MEK-ERK signaling in hepatocellularcarcinoma. J Am Coll Surg 2004;198:410–21.

10. Balmanno K, Cook SJ. Tumour cell survival signalling by the ERK1/2pathway. Cell Death Differ 2009;16:368–77.

11. Caja L, Sancho P, Bertran E, Iglesias-Serret D, Gil J, Fabregat I.Overactivation of the MEK/ERK pathway in liver tumor cells confersresistance to TGF-{beta}-induced cell death through impairing up-regulation of the NADPH oxidase NOX4. Cancer Res 2009;69:7595–602.

12. Carmona-Cuenca I, Roncero C, Sancho P, Caja L, Fausto N, Fernan-dezM, et al. Upregulation of the NADPHoxidase NOX4 by TGF-beta inhepatocytes is required for its pro-apoptotic activity. J Hepatol2008;49:965–76.

13. Szatrowski TP, Nathan CF. Production of large amounts of hydrogenperoxide by human tumor cells. Cancer Res 1991;51:794–8.

14. Kawanishi S, Hiraku Y, Pinlaor S, Ma N. Oxidative and nitrative DNAdamage in animals and patients with inflammatory diseases inrelation to inflammation-related carcinogenesis. Biol Chem 2006;387:365–72.

15. Boonstra J, Post JA.Molecular events associatedwith reactive oxygenspecies and cell cycle progression in mammalian cells. Gene2004;337:1–13.

16. Schafer FQ, Buettner GR. Redox environment of the cell as viewedthrough the redox state of the glutathione disulfide/glutathione couple.Free Radic Biol Med 2001;30:1191–212.

17. Alexandre J, Batteux F, Nicco C, Chereau C, Laurent A, Guillevin L,et al. Accumulation of hydrogenperoxide is an early andcrucial step forpaclitaxel-induced cancer cell death both in vitro and in vivo. IntJ Cancer 2006;119:41–8.

18. Kane RC, Farrell AT, Saber H, Tang S, Williams G, Jee JM, et al.Sorafenib for the treatment of advanced renal cell carcinoma. ClinCancer Res 2006;12:7271–8.

19. Grange PA, Chereau C, Raingeaud J, Nicco C, Weill B, Dupin N, et al.Production of superoxide anions by keratinocytes initiates P. acnes-induced inflammation of the skin. PLoS Pathog 2009;5:e1000527.

Coriat et al.





20. Alexandre J, Nicco C, Chereau C, Laurent A, Weill B, Goldwasser F,et al. Improvement of the therapeutic index of anticancer drugs by thesuperoxide dismutase mimic mangafodipir. J Natl Cancer Inst2006;98:236–44.

21. Rice GC, Bump EA, Shrieve DC, Lee W, Kovacs M. Quantitativeanalysis of cellular glutathione by flow cytometry utilizing monochlor-obimane: some applications to radiation and drug resistance in vitroand in vivo. Cancer Res 1986;46:6105–10.

22. Kamath PS, Wiesner RH, Malinchoc M, Kremers W, Therneau TM,KosbergCL, et al. Amodel topredict survival inpatientswith end-stageliver disease. Hepatology (Baltimore, Md 2001;33:464–70.

23. Therasse P, Arbuck SG, Eisenhauer EA, Wanders J, Kaplan RS,Rubinstein L, et al. New guidelines to evaluate the response totreatment in solid tumors. European Organization for Research andTreatment of Cancer, National Cancer Institute of the United States,National Cancer Institute of Canada. J Natl Cancer Inst 2000;92:205–16.

24. Izzedine H, Massard C, Spano JP, Goldwasser F, Khayat D, Soria JC.VEGF signalling inhibition-induced proteinuria: Mechanisms, signifi-cance and management. Eur J Cancer 2010;46:439–48.

25. Witko-Sarsat V, Friedlander M, Capeillere-Blandin C, Nguyen-Khoa T,Nguyen AT, Zingraff J, et al. Advanced oxidation protein products as anovel marker of oxidative stress in uremia. Kidney Int 1996;49:1304–13.

26. Stichtenoth DO, Frolich JC. Nitric oxide and inflammatory joint dis-eases. Br J Rheumatol 1998;37:246–57.

27. Tai WT, Cheng AL, Shiau CW, Huang HP, Huang JW, Chen PJ, et al.Signal transducer and activator of transcription 3 is a major kinase-independent target of sorafenib in hepatocellular carcinoma. JHepatol2011.

28. Brandes RP, Koddenberg G, Gwinner W, Kim D, Kruse HJ, Busse R,et al. Role of increased production of superoxide anions by NAD(P)Hoxidase and xanthine oxidase in prolonged endotoxemia. Hyperten-sion 1999;33:1243–9.

29. LeeAC, Fenster BE, ItoH, TakedaK, BaeNS,Hirai T, et al. Rasproteinsinduce senescence by altering the intracellular levels of reactiveoxygen species. J Biol Chem 1999;274:7936–40.

30. Cotter G, Kaluski E, Milo O, Blatt A, Salah A, Hendler A, et al. LINCS: L-NAME (a NO synthase inhibitor) in the treatment of refractory cardio-genic shock: a prospective randomized study. Eur Heart J 2003;24:1287–95.

31. Bowler C, Slooten L, Vandenbranden S, De Rycke R, Botterman J,Sybesma C, et al. Manganese superoxide dismutase can reducecellular damage mediated by oxygen radicals in transgenic plants.Embo J 1991;10:1723–32.

32. Valverde P, Manning P, McNeil CJ, Thody AJ. Activation of tyrosinasereduces the cytotoxic effects of the superoxide anion in B16 mousemelanoma cells. Pigment Cell Res 1996;9:77–84.

33. Chandel NS, Maltepe E, Goldwasser E, Mathieu CE, Simon MC,Schumacker PT. Mitochondrial reactive oxygen species triggerhypoxia-induced transcription. Proc Natl Acad Sci U S A 1998;95:11715–20.

34. Choi SI, Jeong CS, Cho SY, Lee YS. Mechanism of apoptosis inducedby apigenin in HepG2 human hepatoma cells: involvement of reactiveoxygen species generated by NADPH oxidase. Arch Pharm Res2007;30:1328–35.

35. Mirzoeva S, Kim ND, Chiu K, Franzen CA, Bergan RC, Pelling JC.InhibitionofHIF-1 alpha andVEGFexpressionby the chemopreventivebioflavonoid apigenin is accompanied by Akt inhibition in humanprostate carcinoma PC3-M cells. Mol Carcinog 2008;47:686–700.

36. Nicco C, Laurent A, Chereau C, Weill B, Batteux F. Differential mod-ulation of normal and tumor cell proliferation by reactive oxygenspecies. Biomed Pharmacother 2005;59:169–74.

37. Laurent A, Nicco C, Chereau C, Goulvestre C, Alexandre J, Alves A,et al. Controlling tumor growth bymodulating endogenous productionof reactive oxygen species. Cancer Res 2005;65:948–56.

38. Legendre L, Heinstein PF, Low PS. Evidence for participation of GTP-binding proteins in elicitation of the rapid oxidative burst in culturedsoybean cells. J Biol Chem 1992;267:20140–7.

39. Servettaz A, Guilpain P, Goulvestre C, Chereau C, Hercend C, NiccoC, et al. Radical oxygen species production induced by advancedoxidation protein products predicts clinical evolution and responseto treatment in systemic sclerosis. Ann Rheum Dis 2007;66:1202–9.

40. Strumberg D, Clark JW, Awada A, Moore MJ, Richly H, Hendlisz A,et al. Safety, pharmacokinetics, and preliminary antitumor activity ofsorafenib: a review of four phase I trials in patients with advancedrefractory solid tumors. Oncologist 2007;12:426–37.

41. Escudier B, Szczylik C, Hutson TE, Demkow T, Staehler M, Rolland F,et al. Randomized phase II trial of first-line treatment with sorafenibversus interferon Alfa-2a in patients with metastatic renal cell carci-noma. J Clin Oncol 2009;27:1280–9.






2012;11:2284-2293. Published OnlineFirst August 17, 2012.Mol Cancer Ther Romain Coriat, Carole Nicco, Christiane Chéreau, et al.

In Vivo and In Vitroon Reactive Oxygen Species Production Sorafenib-Induced Hepatocellular Carcinoma Cell Death Depends

Updated version

10.1158/1535-7163.MCT-12-0093doi:

Access the most recent version of this article at:

Material

Supplementary

http://mct.aacrjournals.org/content/suppl/2012/08/16/1535-7163.MCT-12-0093.DC1

Access the most recent supplemental material at:

Cited articles

http://mct.aacrjournals.org/content/11/10/2284.full#ref-list-1

This article cites 40 articles, 14 of which you can access for free at:

Citing articles

http://mct.aacrjournals.org/content/11/10/2284.full#related-urls

This article has been cited by 6 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]

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

Permissions

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

.http://mct.aacrjournals.org/content/11/10/2284To request permission to re-use all or part of this article, use this link



http://mct.aacrjournals.org/lookup/doi/10.1158/1535-7163.MCT-12-0093

http://mct.aacrjournals.org/content/suppl/2012/08/16/1535-7163.MCT-12-0093.DC1

http://mct.aacrjournals.org/content/11/10/2284.full#ref-list-1

http://mct.aacrjournals.org/content/11/10/2284.full#related-urls

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

mailto:[email protected]

http://mct.aacrjournals.org/content/11/10/2284


Sorafenib-Induced Hepatocellular Carcinoma Cell Death ...Molecular Medicine in Practice...

Documents

Transcript of Sorafenib-Induced Hepatocellular Carcinoma Cell Death ...Molecular Medicine in Practice...