Therapeutics Cucurbitacin I Suppressed Stem-Like Property and … · Therapeutic Discovery Cucurbitacin I Suppressed Stem-Like Property and Enhanced Radiation-Induced Apoptosis in

Ther

CucRadCar

Yi-WeLing-M

Abst

Intro

Heaincludprevaly 8%of im

AuthorDentistMedicin5CanceOral an8Depart

Note: SCancer

Y-W. C

CorresResearSec. 2,Fax: 88

doi: 10

©2010

www.a

Dow

Published OnlineFirst November 9, 2010; DOI: 10.1158/1535-7163.MCT-10-0504

Molecular
Cancer
apeutics

apeutic Discovery

urbitacin I Suppressed Stem-Like Property and Enhancediation-Induced Apoptosis in Head and Neck Squamous

Ther

cinoma–Derived CD44+ALDH1+ Cells

i Chen1,5, Kuan-Hsuan Chen1, Pin-I Huang1,5, Yu-Chih Chen1,6, Guang-Yu Chiou1,6, Wen-Liang Lo2,7,
ing Tseng1,8, Han-Sui Hsu3,8, Kuo-Wei Chang2, and Shih-Hwa Chiou1,4,6
ractHea

activaHNSCstudieaim ofmunohigherfor HNigenictransitcan effresistaand domicroaCD44−

revealtasis a

to 10% ofproveme

s' Affiliatiry, School oe, and 4Instr Center, 6Dd Maxilloment of Sur

upplementTherapeuti

hen, K-H. C

ponding Ach and EduShih-Pai R6-2-2875-7

.1158/1535-

American A

acrjourna

nloaded

d and neck squamous cell carcinoma (HNSCC) is a prevalent cancer worldwide. Signal transducers andtors of transcription 3 (STAT3) signaling is reported to promote tumor malignancy and recurrence inC. Cucurbitacins, triterpenoid derivatives, are strong STAT3 inhibitors with anticancer properties. Recents have shownaldehydedehydrogenase 1 (ALDH1) tobe amarker of cancer stem cells (CSC) inHNSCC. Thethis study was to investigate the therapeutic effect of cucurbitacin I in HNSCC-derived CSCs. Using im-histochemical analysis, we firstly showed that CD44, ALDH1, and phosphorylated STAT3 (p-STAT3) werein high-gradeHNSCCs, and that triple positivity for CD44/ALDH1/p-STAT3 indicated aworse prognosisSCCpatients. Secondly, CD44+ALDH1+ cells isolated from sevenHNSCCpatients showed greater tumor-ity, radioresistance, and high expression of stemness (Bmi-1/Oct-4/Nanog) and epithelial-mesenchymal-ional (Snail/Twist) genes as p-STAT3 level increased. Furthermore, we found that cucurbitacin I (JSI-124)ectively inhibit the expression of p-STAT3 and capacities for tumorigenicity, sphere formation, and radio-nce in HNSCC-CD44+ALDH1+. Notably, 150 nmol/L cucurbitacin I effectively blocked STAT3 signalingwnstream survivin and Bcl-2 expression, and it induced apoptosis in HNSCC-CD44+ALDH1+. Moreover,rray data indicated that 100 nmol/L cucurbitacin I facilitated CD44+ALDH1+ cells to differentiate intoALDH1− and enhanced the radiosensitivity of HNSCC-CD44+ALDH1+. Xenotransplant experimentsed that cucurbitacin I combinedwith radiotherapy significantly suppressed tumorigenesis and lungmetas-nd further improved the survival rate in HNSCC-CD44+ALDH1+-transplanted immunocompromisedTaken together, our data show that cucurbitacin I, STAT3 inhibitor, reduces radioresistant, distant-
mice.
metastatic, and CSC-like properties of HNSCC-CD44+ALDH1+ cells. The potential of cucurbitacin I as a radio-sensitizer should be verified in future anti-CSC therapy. Mol Cancer Ther; 9(11); 2879–92. ©2010 AACR.

HNSCmargisurvivtumorPrinc

duction

d and neck squamous cell carcinoma (HNSCC),ing oral squamous cell carcinoma, is the sixth mostlent cancerworldwide and accounts for approximate-

all cancers in Southeast Asia (1, 2). In spitents in the diagnosis and management of

(CD4renewdehydrespotributstem cthat AmaligepitoHNSCbe a pcancertantlyhigh tserveSig

(STAT

ons: 1Institute of Clinical Medicine, 2Department off Dentistry, 3Institute of Emergency and Critical Careitute of Pharmacology, National Yang-Ming University,epartment of Education and Research, 7Division of

facial Surgery, Department of Stomatology, andgery, Taipei Veterans General Hospital, Taipei, Taiwan

ary material for this article is available at Molecularcs Online (http://mct.aacrjournals.org/).

hen, and P-I. Huang contributed equally to this work.

uthor: Shih-Hwa Chiou, Department of Medicalcation, Taipei Veterans General Hospital, No. 201,oad, Taipei 11217, Taiwan. Phone: 886-2-2875-7394;436. E-mail: [email protected]

7163.MCT-10-0504

ssociation for Cancer Research.

ls.org

on December 16, 2020. © 2mct.aacrjournals.org from

C, long-term survival rates have improved onlynally over the past decade (3). To increase the patiental rate, investigations elucidating the mechanisms ofigenicity in HNSCC are urgently needed. Recently,e et al. showed that the purified CD44-positive4+) population of HNSCC cells possesses the self-ing properties of cancer stem cells (4). Aldehyderogenase 1 (ALDH1), a cytosolic isoenzyme, is

nsible for oxidizing intracellular aldehydes and con-ing to the oxidation of retinol to retinoic acid in earlyell differentiation (5, 6). Visus et al. further suggestedLDH1A1 is a marker of HNSCC that distinguishesnant from premalignant cells and is also an essentialpe for developing ALDH1A1-based vaccines forC therapy (7). Recently, ALDH1 has been shown toutative marker of cancer stem cells (CSC) in breast, hepatoma, colon cancer, and HNSCC (5, 6). Impor-, CD44/ALDH1-coexpressing HNSCC cells displayumorigenic and radioresistant properties and mayas a reservoir for developing tumors andmetastasis (8).
nal transducer and activator of transcription 33), a transcription factor for cytokine signaling, is
2879

010 American Association for Cancer Research.

http://mct.aacrjournals.org/

constileukem(9–13)duce cand tp53 tan onto incpress(p21,(MMPPersisin epiabrogHNSChas bfor tubly, rSTAT3sensitsuppr20). Hthe seCSC iDue

cancesafe aconst(alsopenoiof druthe cubeenanti-inhavegrowrylati(21).ALDHsentedhighbioinfshowof CDgradewhethsensitmentsthe Cand sCD44

Mate

ReagCuc

purch

DMSOI wasfinal c

Isolaand nThi

Helsinvidedthe InBoarded cewere37°Cpositisamped, BTechnsortinturedplemhuma(bFGFof cella denthe mSigma

Irradγ-R

T-1001.1 GFor clation10 dastainevioletmean(PE) aPE =SF =

MicroTot

reage(Qiagversetranslabele

Chen et al.

Mol C2880

Dow


tutively activated in prostate cancer, breast cancer,ia, multiple myeloma, brain tumors, and HNSCC

. Bromberg et al. reported that STAT3 mutations in-ellular transformation and tumor formation in vivohat activation of STAT3 signaling further inhibitsranscriptional activity, fulfilling the definition ofcogene (14, 15). Oncogenic STAT3 activation leadsreased expression of downstream genes that sup-apoptosis (Bcl-xL), regulate cell cycle progressionc-Myc, and cyclin D1), mediate cellular invasion-9), and modulate angiogenesis (VEGF; ref. 16).tent STAT3 activation is associated with mutationsdermal growth factor (EGF) receptor (EGFR) andates growth factor dependence and apoptosis inC cells (13). The activation of STAT3 in HNSCCeen suggested to serve as a prognostic indicatormor growth and malignant progression (17). Nota-ecent reports have suggested that inhibition ofin cancer cell lines can significantly increase radio-

ivity and radiation-induced apoptosis and furtheress tumorigenicity as well as angiogenesis (18–owever, whether STAT3 plays a role in maintaininglf-renewal and radioresistance inHNSCC-associateds still an open question.to the pivotal role of STAT3 in HNSCC and other

rs, numerous studies have focused on identifyingnd effective therapeutic agents that can abrogateitutively active STAT3 signaling. Cucurbitacin Iknown as JSI-124), a natural cell-permeable triter-d compound, belongs to the cucurbitacin familygs isolated from various plant families, such ascurbitaceae and cruciferae. Cucurbitacins haveused as folk medicines for centuries due to theirflammatory and analgesic effects. Recent studiesreported that cucurbitacin I potently inhibits cellth via selectively repressing tyrosine phospho-on of STAT3 in various human cancer cell linesIn this study, we show that the subset of CD44+

1+ cells isolated from seven HNSCC patients pre-cancer stem-like properties, radioresistance, and

levels of p-STAT3. According to microarray andormatic analysis, the treatment of cucurbitacin Ied the potential to promote the CSC-like subset44+ALDH1+ cells shift toward ALDH1− and low-HNSCC cells. Furthermore, we investigateder targeting STAT3 signaling with cucurbitacin Iized the HNSCC-derived CSC to radiation treat-. Our results show that cucurbitacin I inhibitedSC-like properties, enhanced the radiosensitivity,
uppressed the lung metastatic ability of HNSCC- the co
probeing 1FluorsuredArrayusing

+ALDH1+ cells in vivo.

rials and Methods

ents
urbitacin I (JSI-124; Supplementary Fig. S1) wasased from Sigma Chemical Co. and dissolved in
Sprindistan

ancer Ther; 9(11) November 2010

on December 16, 2020. © 2mct.aacrjournals.org nloaded from

as a stock solution of 100 μmol/L. Cucurbitacinfurther diluted in culture medium to appropriateoncentrations just before use.

tion of CD44+ALDH1+ cell subsets from headeck tissuess research followed the tenets of the Declaration ofki. All samples were obtained after patients pro-informed consent. The study was approved bystitutional Ethics Committee/Institutional Reviewof Taipei Veterans General Hospital. The dissociat-lls derived from the samples of HNSCC patientssuspended at a concentration of 1 × 106/mL inDMEM with 2% FCS. We then identified CD44-ve and/or ALDH1-positive cells in the HNSCC cellles using CD44 antibody (phycoerythrin conjugat-ioLegend) and/or the Aldefluor assay (StemCellologies), followed by fluorescence-activated cellg analysis (FACS). CD44+ALDH1+ cells were cul-in serum-free DMEM/F12 (GIBCO) medium sup-ented with N2 supplement (R&D), 10 ng/mLn recombinant basic fibroblast growth factor; R&D), and 10 ng/mL EGF (22, 23). For evaluationproliferation, cells were seeded on 24-well plates atsity of 2 × 104 cells/well in medium, followed byethyl thiazole tetrazolium assay (MTT assay,-Aldrich Co.).

iation and clonogenic assayadiation was delivered by a Theratronic cobalt unit0 (Theratronic International, Inc.) at a dose rate ofy/minute (source-to-surface distance = 57.5 cm).onogenic assay, cells were exposed to different radi-doses (0, 2, 4, 6, 8, and 10 Gy). After incubation forys, colonies (>50 cells/colony) were fixed andd for 20 minutes with a solution containing crystaland methanol. Cell survival was determined by

s of colony formation assay. The plating efficiencynd survival fraction (SF) were calculated as follows:(colony number/inoculating cell number) × 100%;colonies counted/ (cells seeded × (PE/100)).

array analysis and bioinformaticsal RNA was extracted from cells using Trizolnt (Life Technologies) and the Qiagen RNAeasyen) column for purification. Total RNA was re--transcribed with Superscript II RNase H-reversecriptase (Gibco BRL) to generate Cy3- and Cy5-d (Amersham Biosciences Co.) cDNA probes forntrol and treated samples, respectively. The labeleds were hybridized to a cDNA microarray contain-0,000 gene clone immobilized cDNA fragments.escence intensities of Cy3 and Cy5 targets were mea-and scanned separately using a GenePix 4000BScanner (AxonInstruments).DataanalysiswasdoneGenePix Pro 3.0.5.56 (Axon Instruments) and Gene-
g GX 7.3.1 software (Agilent). The average-linkagece was used to assess the similarity between two
Molecular Cancer Therapeutics



groupThe dexprethe coof allof theparisopairwber ovolvecarrieto prople gr

QuanRea

done(1 μgμL reaII RTtimeThe a20 μLMgClgreenlutedand hof denfor 5 sdard ccentrathe enQuanLight3.3 (R

WestThe

analywas b10% SHyboantiboSTATcleavanti-hsurvivp21 (Sand rβ-actidetect

EnzymimmudeoxydUTPThe

enzym

and qries).The pdescriplexstainiwithand pnoreatinyl(Calbent fiwereferasmethoBoehr

In vitThe

ate filCell sof the100 μfilterstainecells wthe so6-wel[DMEthe boture (2 × 14 weecrystausing

In vivAll

withVetera(2 × 1nudedailycin I5 dayagingIllumsourcmor swaswidthPro-p

StatisThe

ysis w

Cucurbitacin I Effects in HNSCC Stem Cells

www.a

Dow


s of gene expression profiles as described below.ifference in distance between two groups of samplession profiles to a third was assessed by comparingrresponding average linkage distances [the meanpairwise distances (linkages) between memberstwo groups concerned]. The error of such a com-n was estimated by combining the SE (the SD ofise linkages divided by the square root of the num-f linkages) of the average-linkage distances in-d. Classical multidimensional scaling (MDS) wasd out using the standard function of the R programvide a visual impression of how the various sam-oups are related (24).

titative real-time reverse-transcriptase-PCRl-time reverse transcriptase-PCR (RT-PCR) wasas previously described (23). Briefly, total RNA) of each sample was reverse-transcribed in a 20-ction using 0.5 μg oligo(dT) and 200 U Superscript(Invitrogen). The primer sequences used for real-RT-PCR are shown in Supplementary Table S1.mplification was carried out in a total volume ofcontaining 0.5 μmol/L of each primer, 4 mmol/L2, 2 μL LightCycler-FastStart DNA Master SYBRI (Roche Molecular Systems), and 2 μL of 1:10 di-cDNA. PCR reactions were prepared in duplicateeated to 95°C for 10 minutes, followed by 40 cyclesaturation at 95°C for 10 seconds, annealing at 55°Ceconds, and extension at 72°C for 20 seconds. Stan-urves (cycle threshold values versus template con-tion) were prepared for each target gene and fordogenous reference (GAPDH) in each sample.tification of unknown samples was done usingCycler Relative Quantification Software versionoche Molecular Systems).

ern blot assayextraction of proteins from cells and Western blot

sis were done as described (23). A sample (15 μL)oiled at 95°C for 5 minutes and separated byDS-PAGE. The proteins were wet-transferred tond-ECL nitrocellulose paper (Amersham). Primarydies were as follows: rabbit anti-human phospho-3, rabbit anti-human STAT3, rabbit anti-humaned poly(ADP-ribose) polymerase (PARP), rabbituman cleaved caspase 3, and mouse anti-humanin (Cell Signaling Technology); rabbit anti-humananta Cruz Biotechnology); mouse anti-human Bcl-2abbit anti-human Bax (Upstate); and mouse anti-n (Chemicon). The reactive protein bands wereed by the ECL detection system (Amersham).

e-linked immunosorbent assay,nohistochemistry staining, and terminalnucleotidyl transferase–mediatednick end labeling assay
activities of caspase-3 was determined using ane-linked immunosorbent assay kit (R&D Systems)
two-wpropr

acrjournals.org


uantified at 490 nm (MRX, Dynatech Laborato-Each individual sample was analyzed in triplicate.rotocol of immunofluorescence staining has beenbed previously (23). Briefly, an avidin-biotin com-method was used for the immunofluorescenceng in the spheroid cells. Each slide was treatedantibodies for CD44 (DAKO), ALDH1 (Abcam),hospho-STAT3 (Cell Signaling Technology). Immu-ctive signals were detected with a mixture of bio-ated rabbit anti-mouse IgG and Fluoesaveiochem). Positive cells were counted in six differ-elds by microscopy. Furthermore, apoptotic cellsidentified by the terminal deoxynucleotidyl trans-e–mediated dUTP nick-end labeling (TUNEL)d (In situ Cell Death Detection Kit, POD, Rocheinger Mannheim Corp.; ref. 23).

ro cell invasion analysis and soft agar assay24-well plate Transwell system with a polycarbon-ter membrane (8 μm pore size; Corning) was used.uspensions were seeded in the upper compartmentTranswell chamber at a density of 1 × 105 cells inL serum-free medium. The opposite surface of themembrane, which faced the lower chamber, wasd with Hoechst 33342 for 3 minutes, and migratingere visualized under an inverted microscope. Forft agar assay, the bottom of each well (35 mm) of al culture dish was coated with a 2 mL agar mixtureM with 10% (v/v) FCS and 0.6% (w/v) agar]. Afterttom layer solidified, 2 mL top agar/medium mix-DMEM with 10% FCS and 0.3% agar) containing04 cells was added and incubated at 37°C forks. The plates were stained with 0.5 mL 0.005%l violet, and the number of colonies was counteda dissecting microscope.

o analysis of tumor growth and metastasisprocedures involving animals were in accordancethe institutional animal welfare guideline of Taipeins General Hospital. HNSCC-CD44+ALDH1+ cells05) were injected into the neck region of 8-week-oldmice (BALB/c strain; ref. 25) and then treated withi.p. injections of vehicle (10% ethanol) or cucurbita-(1 mg/kg JSI-124 in 10% ethanol) for a total ofs (25). In vivo green fluorescent protein (GFP) im-was done using an illuminating device [LT-9500atool TLS equipped with excitation illuminatinge (470 nm) and filter plate (515 nm); ref. 23]. Tu-ize was measured using calipers, and the volumecalculated according to the formula: (length ×2)/2 and subsequently analyzed using Imagelus software (23).

tical analysisresults are reported as mean ± SD. Statistical anal-as done using Student's t-test or a one-way or
ay ANOVA test followed by Turkey's test, as ap-iate. Survival was estimated by the Kaplan-Meier
Mol Cancer Ther; 9(11) November 2010 2881



methoconsid

Resu

Poorpositip-STATo

amoncancenohis111 Hshowtivelyto poosignifALDHand 0tivelyof CDwithanaly5-yeargraderesultthe AderabALDHwith lnosispatienhad asides,tumocoeffigradesuggep-STAof HNitive fwithp-STAcate tHNSCdiseas

IsolatcellsUsi

ed ALsue sdescriS2). Itin susserumfree mspher

ALDHsigniCD44Suppof tumMatrition aCD44and 3the Mleft). Sand Cwhenthe saWe fuNanogscriptgenesCD44right;effectradiatfive gtreatmand CALDHstudiethemin ALshowTyr70thanpmediplay sCD44

CucumaligALDHCuc

centlydetermthe CviabilMTTcentramentcoloncells (stemncells tscriptysis (Fprincthat Hhigh-

Chen et al.

Mol C2882

Dow


d and compared by the log-rank test. P < 0.05 wasered statistically significant.

lts

overall survival rate of patients with HNSCC isvely associated with CD44, ALDH1, andT3 expression

investigate whether there is a positive correlationg CD44, ALDH1, and p-STAT3 in head and neckrs, we studied the levels of these proteins by immu-tochemical staining of a panel of specimens fromNSCC patients. The immunohistochemistry resultsed that increased expression of ALDH1 was posi-correlated with the advanced stages and mediumr differentiation of HNSCC (Fig. 1A). There was aicantly high correlation between tumor grade and1/CD44/p-STAT3 (relative coefficiency R = 0.87.92 in low-grade and high-grade tumors, respec-; Fig. 1A). To determine the prognostic significance44, ALDH1, and p-STAT3 expression in patientsHNSCC, we carried out Kaplan-Meier survivalsis. First, we found a significant difference for thesurvival prognosis between high-grade and low-HNSCC patients (Fig. 1B; P < 0.001). Second, thes of Kaplan-Meier survival analysis showed thatLDH1-positive cases were associated with a consi-ly worse overall survival rate compared with1-negative ones (Fig. 1B; P < 0.001). Third, patientsower CD44 expression had a better survival prog-compared with the CD44-highly expressingts (Fig. 1B; P = 0.002). Fourth, p-STAT3+ patientsworse survival prognosis (Fig. 1B; P < 0.001). Be-there was a significantly high correlation betweenr grade and ALDH1+/CD44+/p-STAT3+ (relativeciency R = 0.87 and 0.92 in low-grade and high-tumors, respectively). Taken together, these resultsst that elevated expression of CD44, ALDH1, andT3 were strongly associated with advanced gradeSCC and worse prognosis. In addition, patients pos-or all three had the worst survival rate comparedother HNSCC patients (Fig. 1C; CD44+ALDH1+

T3+ versus other groups). Overall, these data indi-hat expression of CD44, ALDH1, and p-STAT3 inC patients could be a critical factor in predictinge progression and clinical outcomes.

ion and characterization of CD44+ALDH1+

from HNSCC tissuesng the Aldefluor assay and FACS analysis, we isolat-DH1+, ALDH1−, CD44+, and CD44− cells from tis-amples of seven HNSCC patients as previouslybed (ref. 8; Fig. 2A, left, and Supplementary Tablehas been reported that CSC-like cells can be culturedpension to generate floating spheroid-like bodies in-free medium with bFGF and EGF. In DF-12 serum-
edium with bFGF and EGF, the ability to formoid-like bodies and the proliferation rate in the
or noresult



1+-lineage (ALDH1+ and CD44+) cells were bothficantly higher than in parental, ALDH1−, and− cells (patients 1 to 3; P < 0.001; Fig. 2A, right;lementary Table S2). To evaluate the enhancementorigenicity of HNSCC-ALDH1+ cells, we employedgel/Transwell invasion and soft agar colony forma-ssays. Compared with ALDH1− or CD44+ALDH1−,+ALDH1+ cells derived from HNSCC patients 1, 2,showed higher invasion activity, as assessed byatrigel/Transwell invasion assay (P < 0.001; Fig. 2B,imilarly, the colony formation ability of ALDH1+

D44+ALDH1+ cells from HNSCC patients was enhancedcompared with the ALDH1− or CD44+ALDH1− ofme patient (P < 0.001; Supplementary Fig. S2A).rther found that the stemness genes (Oct-4A and), epithelium-mesenchymal transition (EMT) tran-ional factors (Snail and Twist), and the drug-resistant(MDR-1) were upregulated in both ALDH1+ and+ALDH1+ cells using real-time RT-PCR (Fig. 2B,Supplementary Fig. S2B). To further determine theof radiation on tumor growth rate, we used ionizingion doses from 0 to 10 Gy to treat the tumors in theseroups. As shown in Fig. 2C, after ionizing radiationent, both the survival rate and number of ALDH1+

D44+ALDH1+ cells were significantly higher than in1− and CD44+ALDH1− (P < 0.001). Because severals have shown that STAT3 activation is related toalignancy ofHNSCCs, the activation status of STAT3DH1+ and ALDH1− HNSCC cells was detected. Asn in Fig. 2D, the levels of activated STAT3 (p-STAT3-5) in CD44+ALDH1+ cells were significantly higherarental cells both in serum-free and serum-containingum. These findings suggest the p-STAT3 mayome roles in regulating the property of HNSCC-+ALDH1+.

rbitacin I inhibits the proliferation, tumornancy, and stemness signatures of CD44+1+ cellsurbitacin I, a specific STAT3 inhibitor, has been re-suggested to suppress tumor growth. It remains un-ined, however, whether cucurbitacin I can inhibit

SC properties of HNSCC-CD44+ALDH1+ cells. Theity of HNSCC-CD44+ALDH1+ cells determined byassay significantly decreased with increasing con-tions of cucurbitacin I (P < 0.05; Fig. 3A). Treat-with cucurbitacin I also significantly blocked they formation capability of HNSCC-CD44+ALDH1+

P < 0.05; Fig. 3B). To explore molecules governingess and tumorigenicity in HNSCC-CD44+ALDH1+

reated with cucurbitacin I, we examined their tran-ome profile using gene expression microarray anal-ig. 3C; Supplementary Tables S3 and S4). MDS andiple component analysis (PCA) further showedNSCC-CD44+ALDH1+ cells were more similar to

grade tissues of HNSCC than low-grade HNSCC
rmal oral tissues (Fig. 3D). In contrast, the MDSs showed that the expression patterns of ALDH1−,



Figureof immright, hin low-g(top lef(bottomas refer


www.a

Dow


1. Correlation of CD44, ALDH1, and p-STAT3 expression to the clinical grading and survival rate of HNSCC patients. A, left, representative resultsunohistochemical staining for ALDH1 (top), CD44 (middle), and p-STAT3 (bottom) in 111 HNSCC patients at different grades (left, low grade;igh grade). Right, significantly high correlation between tumor grade and ALDH1, CD44, and pSTAT3 (relative coefficiency R = 0.87 and 0.92rade and high-grade tumor, respectively). B, Kaplan-Meier analysis of overall survival in 111 HNSCC patients according to clinical histology gradingt; P < 0.001), single ALDH1 expression (bottom left; *, P < 0.001), single CD44 expression (top right; **, P = 0.002), single p-STAT3 expression

+ + + − − −
right; *, P < 0.001), and combined expression of CD44 ALDH1 p-STAT3 (C; ***, P < 0.001). Inset, CD44 ALDH1 p-STAT3 cells (C) usedence group for comparison.
Mol Cancer Ther; 9(11) November 2010acrjournals.org 2883

on December 16, 2020. © 2010 American Association for Cancer Research.mct.aacrjournals.org nloaded from


Figurevia FACmedium*, P < 0of indicALDH1and ser

Chen et al.

Mol C2884

Dow


2. Isolation and characterization of CD44+ALDH1+ cells from HNSCC tissues. A, 6.15% CD44+ALDH1+ cells were identified from HNSCC tissuesScan. DEAB, an inhibitor of ALDH, was used as negative control (left). The ability to form spheroid-like bodies in various groups in serum-freewith bFGF and EGF was evaluated (right). Bar, 100 μm. B, invasion ability was detected in the different ALDH1+ and ALDH1− groups (left).

.001, ALDH1+ versus parental and ALDH1−; #, P < 0.001, CD44+ALDH1+ versus ALDH1+. Q-RT-PCR results to quantify the amounts of transcriptsated genes (right). C, to determine the radiation effect on the tumor growth rate, ionizing radiation doses of 0 to 10 Gy were used to treat+ − +
and ALDH1 cells. D, protein levels of activated STAT3 (p-STAT3-Tyr705) and total STAT3 from ALDH1 and parental cells in serum-free (top)um-containing (bottom) medium were determined by Western blot.
ancer Ther; 9(11) November 2010 Molecular Cancer Therapeutics



Figurein 24-wassay.*, P < 0HNSCC


www.a

Dow


3. Microarray analysis reveals key cucurbitacin I–regulated transcriptomes. A, HNSCC- CD44+ALDH1+ cells from patients 1 to 3 were platedell plates and incubated for 48 hours with various concentrations of cucurbitacin I. At the end of treatment, cell viability was determined by MTT

+ +
B, treatment with cucurbitacin I in HNSCC-CD44 ALDH1 cells impeded the capability of colony formation as evaluated by soft agar assay..001. C, gene expression microarray analysis (gene tree) of the 987 genes that were differentially expressed in cucurbitacin I–treated (cu-)-CD44+ALDH1+ cells as compared with control cells as shown by a hierarchy heat map. The time-dependent changes in expression of the



cucurgradeNotabof 987tacinwhen(Fig. 3I in Hof theALDH

CucuapoptTo

pressecells wbitacisigniflike band FALDHcucur(Fig. 4bitacinin HNthe egenesof surWestesurvivproduThe aouslyTUNEweretion (pressactiva(Fig.maintthe inoptos

Cucuin HNTo

tumordiatiodosesHNSCsurvivcells wwithexploI–medthe 10rentaposed

colonspheruatedditioneffectviabilionizisignifcucurtivityHNSCplus itreatmthe efALDHeffect

CucuionizdistatransWe

and tALDHparencontaALDHHNSCnude(Fig.receivneck rnot sstron(Fig. 6tivelyGFPcucurradiacompto ionmore,ALDHHNSCbilitie(Fig. 6ALDHnumb6B, Su4 Gyin theductiosuggeergistALDHmore,

Chen et al.

Mol C2886

Dow


bitacin I-treated CD44+ALDH1+ cells, and low-HNSCC were more close to the normal oral tissues.ly, microarray analysis showed that the expressionprobe sets was significantly altered in the cucurbi-I–treated group compared with the control groupcompiled with the hierarchical clustering methodC). More importantly, the treatment of cucurbitacinNSCC-CD44+ALDH1+ cells resulted in an extensionaverage linkage distances among CD44+ALDH1+,1+, and high-grade HNSCC (P < 0.05; Fig. 3D).

rbitacin I promotes differentiation and inducesosis by blocking STAT3 signaling in HNSCCexamine whether CSC-like properties are sup-d via STAT3 inhibition, HNSCC-CD44+ALDH1+

ere incubated with 50, 100, or 150 nmol/L cucur-n I for 24 hours. Treatment with cucurbitacin Iicantly interfered with the formation of spheroid-odies (Fig. 4A). Importantly, the Aldefluor assayACS analysis showed that the quantities of1 and CD44 were dramatically decreased inbitacin I–treated HNSCC-ALDH1+CD44+ cellsB; P < 0.001). Western blotting showed that cucur-I induced a dose-dependent decrease of p-STAT3SCC-CD44+ALDH1+ cells (Fig. 4C). To elucidateffects of cucurbitacin I on STAT3 downstreamand cell survival–related genes, the expressionsvivin, Bcl-2, Bcl-xL, and Bax were examined byrn blot. As shown in Fig. 4C, Bcl-2, Bcl-xL, andin expression were downregulated, whereas Baxction was enhanced with cucurbitacin I treatment.mount of cleaved PARP and caspase 3 was obvi-increased (Fig. 4C). Consistent with this finding,L staining also revealed that apoptotic signalspositively correlated with cucurbitacin I concentra-Fig. 4D). We further found that cucurbitacin I sup-ed Stat3 and Janus-activated kinase 2 (JAK2)tion without influencing Src phosphorylation4C). These data suggest that the STAT3 pathwayains the stemness of ALDH1+CD44+ cells andhibition of STAT3 activation further promotes ap-is in HNSCC-CSC.

rbitacin I improves sensitivity to radiotherapySCC-CD44+ALDH1+ cells

further investigate the biological roles of STAT3 inigenicity of HNSCC-CD44+ALDH1+ cells under ra-n treatment, we applied varying ionizing radiationfrom 0 to 10 Gy to vehicle- or cucurbitacin I–treatedC-CD44+ALDH1+cells. As shown in Fig. 5A, theal rate of vehicle-treated HNSCC-CD44+ALDH1+

as significantly higher than that of cells treated100 nmol/L cucurbitacin I (P < 0.01). To furtherre the mechanism involved in the cucurbitaciniated radiosensitizing effect against HNSCC cells,0 or 150 nmol/L cucurbitacin I–treated HNSCC pa-
l cells or HNSCC-CD44+ALDH1+ cells were ex-to 4 Gy ionizing radiation. The capabilities of
levelsCD44



y formation (Fig. 5B), invasion (Fig. 5C, left), ande formation (Fig. 5C, right) were dramatically atten-by cucurbitacin I and by ionizing radiation. In ad-, the combination treatment showed a synergisticin abrogating these HNSCC cell capabilities. Cellity assays showed that the cytotoxic effect of 4 Gyng radiation on HNSCC-CD44+ALDH1+ cells wasicantly increased with the addition of 100 nmol/Lbitacin I (P < 0.01; Fig. 5D, left). Meanwhile, the ac-of caspase 3 was concomitantly increased inC-CD44+ALDH1+ cells treated with cucurbitacin Ionizing radiation compared with ionizing radiationent alone (Fig. 5D, right). These data indicate thatfectiveness of radiation treatment on HNSCC-CD44+

1+ cells can be improved with cucurbitacin I, anmediated by the suppression of STAT3 signaling.

rbitacin I presents the synergistic effects withing radiation to inhibit tumorigenicity andnt-metastatic ability in HNSCC-ALDH1+-planted immunocompromised micefurther investigated the role of the STAT3 signalinghe effects of cucurbitacin I in HNSCC-CD44+

1+ cells in vivo. HNSCC-ALDH1+/− and HNSCC-tal cells were transfected with a lentiviral vectorining GFP (22–24). We first injected 2 × 105 CD44+

1+-GFP, 2 × 105 ALDH1−-GFP cells, and 2 × 105

C-parental cells-GFP cells into the neck region ofmice that received different treatment protocols6). First, we found that the subset of nude miceing ALDH1−-GFP cells formed no tumors in theegion within 10 weeks of xenotransplantation (datahown). CD44+ALDH1+-GFP cells presented thegest ability to form tumors in transplanted miceA). Cucurbitacin I (1 mg/kg, i.p. for 5 days) effec-suppressed the proliferation of CD44+ALDH1+-

cells in the transplanted mice (Fig. 6A). Notably,bitacin I showed a synergistic effect with ionizingtion in CD44+ALDH1+-GFP–transplanted miceared with mice receiving the same cells but exposedizing radiation alone (P < 0.05; Fig. 6A). Further-our in vivo data show that the subset of CD44+

1+ cells, but not the ALDH1−-lineage cells orC-parental cells group, exhibited significant capa-s of invasion and distant metastasis to the lungsB). Importantly, cucurbitacin I treatment in CD44+

1+-GFP–transplanted mice effectively reduced theer of lung metastases and tumor size in vivo (Fig.pplementary Fig. S3). Further combination withionizing radiation showed a significant diminutionmultiple nodules of tumor formation and the re-n of tumor volume in the CD44+ALDH1+ group,sting that cucurbitacin I plus ionizing radiation syn-ically blocked the metastatic ability of CD44+

1+ cells (Fig. 6B, Supplementary Fig. S3). Further-immunohistochemistry showed that the expression
of p-STAT3 in the neck region tumors of HNSCC+ALDH1+– injected nude mice were highly



expreFig. 6nohisCD44ment

CD44combI had

FigureHNSCCpopulatBcl-xL in cucurbitacin I–treated HNSCC-CD44 ALDH1 cells were detected by Western blot. D, the presence of TUNEL-positive cells was determinedin HNS the m


www.a

Dow


ssed in comparison with the other groups (P < 0.05;C). Our result further showed that p-STAT3 immu-tochemistry levels were significantly decreased in

CC-ALDH1+ cells after treatment with cucurbitacin I. *, P < 0.01. Data are

+ALDH1+–xenotransplanted graft after the treat-of cucurbitacin I (P < 0.001; Fig. 6C). Moreover,

prolonALDH

acrjournals.org


+ALDH1+-GFP–transplanted mice treated with theination of 4 Gy ionizing radiation and cucurbitacina mean survival rate that was significantly

ean ± SD of three independent experiments.

4. Blocking STAT3 signaling suppresses HNSCC-CD44+ALDH1+ viability and induces apoptotic cell death. A, treatment with cucurbitacin I in-CD44+ALDH1+ cells impeded the capability to form spheroid-like bodies. *, P < 0.001. B, FACScan showed decreased ALDH1+ and CD44+

ions after cucurbitacin I treatment. C, p-STAT3, STAT3, survivin, Bcl-2, Bax, cleaved caspase 3, cleaved PARP, p-JAK2, JAK2, p-Src, Src, and+ +

ged compared with the control group and CD44+

1+-GFP–transplanted mice that received other




FigureradiothALDH1cells wirradiati(MTT asof cucu

Chen et al.

Mol C2888

Dow


5. Cucurbitacin I suppresses CSC-like properties of HNSCC-CD44+ALDH1+ cells by inhibiting STAT3 signaling and improves the sensitivity toerapy. A, to determine the effect of radiation on tumor growth rate, ionizing radiation (IR) doses from 0 to 10 Gy were used to treat HNSCC-CD44++ cells in combination with vehicle or cucurbitacin I (Cu). Colony formation (B) and the invasion abilities (C) of parental and HNSCC-CD44+ALDH1+

ere examined after treatment with either cucurbitacin I or 4 Gy ionizing radiation or both (C, left). Treatment with cucurbitacin I combined withon in parental and HNSCC-CD44+ALDH1+ cells impeded the capability to form spheroid-like bodies (C, right). *, P < 0.001. D, cell viability

+ +
say) and caspase 3 activity were determined in HNSCC-CD44 ALDH1 cells after treatment with 4 Gy ionizing radiation in the presence or absencerbitacin I. Data shown are the mean ± SD of three independent experiments.
ancer Ther; 9(11) November 2010 Molecular Cancer Therapeutics



FigureHNSCC36 micerevealemice trecucurbanalyze4 Gy ioeach grALDH1survivaionizing


www.a

Dow


6. Evaluation of in vivo tumorigenicity of HNSCC-CD44+ALDH1+ cells and survival time in a xenotransplanted animal model. A, a total of 2 × 105

-CD44+ALDH1+ and HNSCC-parental cells were injected s.c. into the necks of nude mice. Six mice in each group (n = 6 in each group; total) received daily i.p. injections of vehicle (10% ethanol) or drug [1 mg/kg cucurbitacin I (Cucur/Cu) in 10% ethanol]. After 4 weeks, in vivo GFP imagingd that transplanted HNSCC-CD44+ALDH1+-GFP cells grew solid tumors in the injection site. Tumor volumes in HNSCC-CD44+ALDH1+–transplantedated with cucurbitacin I (1 mg/kg, i.p. for 5 days) with ionizing radiation (4 Gy) were significantly lower than in those receiving ionizing radiation oritacin I only (P < 0.01). B, the number of metastatic foci (top left, fluorescence spots; arrows) and total volume of tumors in the lungs of mice wered by macroscopic and histologic examination (bottom left and middle; arrow, neovascularity and thrombosis). Cucurbitacin I only or combined withnizing radiation effectively reduced the number of lung metastases and tumor size in CD44+ALDH1+-GFP–transplanted mice (*, P < 0.01; n = 6 inoup; total 36 mice). C, immunohistochemistry for p-STAT3 indicated that cucurbacitin I effectively suppressed the level of p-STAT3 in CD44++– and parental cell–xenotransplanted immunocompromised mice. Bar, 20 μm. Data shown are the mean ± SD of three experiments. D, Kaplan-Meier

+ +
l analysis further indicated that the mean survival rate for animals receiving HNSCC-CD44 ALDH1 cells treated with cucurbitacin I combined withradiation was significantly prolonged compared with those receiving ionizing radiation or drug along (each group n = 10 mice).



treatmshowsbearinly imp

Discu

STAcytok(IL-6)cer tyby intgonuccells s(26). Bmor cto tuSTAT3py stALDHHNSCThe impositiworsewith tlevelsHNSCthe Cover,ALDHcharatanceALDHrivativand efects.couldduce ably, 1spherferent100 nCD44nohistivelyALDHin vivsurvivcells aicantl(Fig. 6that thing Ccucurand rRad

HNSCmor r

the thther stheraneedenalinggenesdiatiocombtivelysiRNAlaryngmanpressiSTATinhibiical sptry haand pto tretumunificanon–sfore, timpordiothbitacradiatsistanof p-Ssion (also ip-STArelateshowstaticwith 4cucurwithing radeedproapI on Cto impas advstudytenuadownthermbetweducedand uthe epTheseway wAG49adeno

Chen et al.

Mol C2890

Dow


ents (P < 0.05; Fig. 6D). Overall, this in vivo studythat the effectiveness of ionizing radiation in micegHNSCC-CD44+ALDH1+ tumors can be significant-roved with the addition of cucurbitacin I treatment.

ssion

T3, a transcription factor regulated by variousines and growth factors, especially interleuikin-6and EGF, is aberrantly activated in numerous can-pes, including HNSCC. Targeting STAT3 signalingroducing dominant-negative constructs, decoy oli-leotides, or antisense oligonucleotides to HNSCCignificantly inhibits growth and induces apoptosisecause persistent activation of STAT3 promotes tu-ell proliferation and survival, further contributingmor progression and migration, abrogation ofsignaling is emerging as a potential cancer thera-

rategy. Herein we found that levels of CD44,1, and p-STAT3 were greater in higher-gradeC tissues than in lower-grade samples (Fig. 1).munohistochemical analysis showed that triple

vity for CD44, ALDH1, and p-STAT3 indicated aprognosis for HNSCC patients (Fig. 1). Consistenthis observation, we found that the phosphorylationof STAT3 in CD44+ALDH1+ cells from sevenC patients were significantly higher than those inD44−, ALDH1−, and parental cells (Fig. 2). More-consistent with our previous findings (27), CD44+

1+ cells isolated from HNSCC patients sharedcteristics of CSC and showed greater radioresis-compared with HNSCC cells of the CD44− or1− lineage (Fig. 2). Cucurbitacins, triterpenoid de-es, are strong inhibitors of the JAK/STAT pathwayxhibit biopharmacologic activities in anticancer ef-Our data showed that 150 nmol/L cucurbitacin Ieffectively block STAT3 signaling and further in-poptosis in HNSCC-CD44+ALDH1+ (Fig. 3). Nota-00 nmol/L cucurbitacin I blocked formation ofoid-like bodies and facilitated CD44+ALDH1+ dif-iation into CD44−ALDH1− cells (Fig. 3). Moreover,mol/L cucurbitacin I enhanced radiosensitivity in+ALDH1+ cells (Fig. 3). Notably, the result of immu-tochemistry showed that cucurbitacin I can effec-inhibit the expression of p-STAT3 in CD44+

1+–xenotransplanted mice (Fig. 6C). Finally,o xenotransplant analysis indicated that the meanal rate of mice bearing HNSCC-CD44+ALDH1+

nd treated with ionizing radiation could be signif-y improved by adding cucurbitacin treatment). To our knowledge, this is the first study to showe STAT3 axis plays an important role in maintain-SC-like properties and that targeting STAT3 withbitacin I significantly suppresses tumorigenicityadioresistance in HNSCC-associated CSCs.iotherapy is the conventional treatment for
C. Radioresistance is one of the major causes of tu-ecurrence and metastasis in HNSCC. To improve
type,induc



erapeutic outcome of malignant HNSCC and fur-pecifically target HNSCC-associated CSCs, novelpeutic agents and radiosensitizers are urgentlyd. STAT3 is an important component of several sig-pathways, such as that of EGF, which activatesrequired for DNA repair and cell survival after ra-n treatment. Recent reports have suggested that aination of STAT3 inhibitor and radiotherapy effec-attenuated tumor cell growth. Introducing STAT3significantly enhanced radiosensitivity in humaneal squamous cell carcinoma cells (28) and in hu-

squamous cell carcinoma cells with EGFR overex-on (20). Chen et al. further showed that the IL-6/3 pathway is responsible for the resistance to EGFRtor and irradiation in pharyngeal cancer (29). Clin-ecimens further assessed by immunohistochemis-ve shown that higher levels of IL-6, IL-6 receptor,-STAT3 are correlated with lower response ratesatments and shorter survival (29). Moreover, pani-mab, a monoclonal antibody against EGFR, can sig-ntly augment the radiosensitivity of HNSCC andmall cell lung cancer in vitro and in vivo (30). There-he inhibition of IL-6/STAT3 or p-STAT3 may be antant therapy when combined with conventional ra-erapy. In this study, our data suggest that cucur-in I (JSI-124) significantly increased ionizingion–induced apoptosis and suppressed the radiore-ce in HNSCC-CSC, in part through the inactivationTAT3 and downstream survivin and Bcl-2 expres-Fig. 3). According to clinical follow up, our datandicated that the combined high expression ofT3, CD44, and ALDH1 in HNSCC was highly cor-d with the clinical radioresistant history (data notn). Importantly, the in vivo tumorigenic and meta-capabilities of HNSCC-CD44+ALDH1+ cells treatedGy ionizing radiation combined with 100 nmol/L

bitacin I were significantly decreased comparedthose of untreated CD44+ALDH1+ and only ioniz-diation–treated CD44+ALDH1+ cells (Fig. 5). In-, these findings indicate the antiproliferative,optotic, and radiosensitizing effect of cucurbitacinD44+ALDH1+ cells and could potentially be usedrove the clinical treatment of HNSCC-CSC as wellanced-stage HNSCC patients. In addition, a recentalso suggests that cucurbitacin could potentially at-te the proinflammatory effect of IL-6 throughregulation of IL-6/STAT3 signaling (31, 32). Fur-ore, recent evidence has reported the associationen STAT3 and EMT. Exogenous addition of EGF in-migratory phenotype, enhanced IL-6 production,

pregulated the level of N-cadherin and vimentin inithelial ovarian cancer lines OVCA 433 and SKOV3.enhancements were abrogated as the STAT3 path-as blocked by neutralizing IL-6R antibody and

0 (33). Likewise, ectopic IL-6 expression in the breastcarcinoma cell line MCF-7 promoted EMT pheno-
including impaired E-cadherin expression andtion of vimentin, N-cadherin, Snail, and Twist (34).



In thALDHof invwas nThe rPCR finhiband T(Suppcucurstatictranspthat Ssignalcancesynerapeutmolecis stilling ILor autthe tuHNSCIn s

axis mresistacucur

pathwHNSCI enhblockby iousingas a mtumor

Discl

No p

Grant

Nati075-00ER2-016), Yen(MinistDeveloment oof Exce(DOH9

Thepaymenadvertisthis fac

Refe1. Ka

moredwo

2. KhtobnecJ C

3. Al-Clicar

4. Priposqu

5. Maep

6. Shcannec

7. VisdetumCa

8. ChenhcelOra

9. BoOn

10. Bro113

11. Brosay


www.a

Dow


is study, our in vivo data showed that CD44+

1+-lineage cells exhibited significant capabilitiesasion and distant metastasis to the lungs, whichot found in CD44− or ALDH1− HNSCC cells (Fig. 6).esults of microarray analysis and quantitative RT-urther showed that the treatment of cucurbitacin Iited the constitutive mRNA expressions of Snailwist in treated HNSCC-CD44+ALDH1+ cellslementary Fig. S4). Notably, our data showed thatbitacin I presents the potential to suppress the meta-ability to lung organ in HNSCC-CD44+ALDH1+–lanted nude mice (Fig. 6). These findings suggestTAT3 pathway may be involved in the molecularing of EMTand distant metastasis in head and neckr in vivo. Moreover, some reports have shown thegistic effect of cucurbitacins with known chemother-ic agents, such as doxorubicin (35). Although theular mechanism of HNSCC-CSC or CD44+ALDH1+

unclear, there is a need to further investigate target--6/JAK/STAT3 signaling as well as other paracrineocrine factors and radiochemoattractants involved inmor microenvironment that biologically impactC progression.ummary, our data indicate that the STAT3 signalingay contribute to the CSC-like properties and radio-
nce of HNSCC-CD44+ALDH1+ cells. We identified
ve JAK/STAT3 signalingRece

publish

9–42.mberg JF. Activation of STAT proteins and growth control. BioEs-s 2001;23:161–9.

12. EpSTlym10

13. Alop

14. Broco

15. Niusio

16. Yuag

17. Kaog

18. Otacota

19. Kimacviv

20. Boofcin

21. BatumerRe

22. Chrad20

23. Kaincter

acrjournals.org


ay inhibitor, as a potent antitumor agent inC-CSC in vivo and in vitro. Notably, cucurbitacin

anced the inhibition of cancer stem-like property,age of invasion ability, and induction of apoptosisnizing radiation in HNSCC-CSC. The potential oftriple expression of CD44, ALDH1, and p-STAT3arker of radioresistance in HNSCC or other solids should be verified in future clinical trials.

osure of Potential Conflicts of Interest

otential conflicts of interest were disclosed.

Support

onal Science Council (NSC-97-3111-B-075-001-MY3/98-2314-B -8-MY3), Taipei Veterans General Hospital (V97B1-006, E1-008,8, ER3-005, F-001), the Joint Projects of UTVGH (VGHUST-98-G6--Tjing-Ling Medical Foundation, National Yang-Ming Universityry of Education, Aim for the Top University Plan), Technologypment Program for Academia (98-EC-17-A-19-S2-0107)/Depart-f Industrial Technology, Ministry of Economic Affairs, and Centerllence for Cancer Research at Taipei Veterans General Hospital9-TD-C-111-007), Taiwan.costs of publication of this article were defrayed in part by thet of page charges. This article must therefore be hereby markedement in accordance with 18 U.S.C. Section 1734 solely to indicatet.

ived 05/27/2010; revised 08/30/2010; accepted 09/15/2010;ed OnlineFirst 11/09/2010.
bitacin I (JSI-124), a selecti
rencesmangar F, Dores GM, Anderson WF. Patterns of cancer incidence,rtality, and prevalence across five continents: defining priorities touce cancer disparities in different geographic regions of therld. J Clin Oncol 2006;24:2137–50.awaja MR, Mazahir S, Majeed A, et al. Chewing of betel, areca andacco: perceptions and knowledge regarding their role in head andk cancers in an urban squatter settlement in Pakistan. Asian Pacancer Prev 2006;7:95–100.Swiahb JN, Chen CH, Chuang HC, Fang FM, Tasi HT, Chien CY.nical, pathological and molecular determinants in squamous cellcinoma of the oral cavity. Future Oncol 2010;6:837–50.nceME, Sivanandan R, Kaczorowski A, et al. Identification of a sub-pulation of cells with cancer stem cell properties in head and neckamous cell carcinoma. Proc Natl Acad Sci U S A 2007;104:973–8.ck B, Gires O. CD44s and CD44v6 expression in head and neckithelia. PLoS One 2008;3:e3360.eridan C, Kishimoto H, Fuchs RK, et al. CD44+/CD24- breastcer cells exhibit enhanced invasive properties: an early stepessary for metastasis. Breast Cancer Res 2006;8:R59.us C, Ito D, Amoscato A, et al. Identification of human aldehydehydrogenase 1 family member A1 as a novel CD8+ T-cell-definedor antigen in squamous cell carcinoma of the head and neck.ncer Res 2007;67:10538–45.en YC, Chang CJ, Hsu HS, et al. Inhibition of tumorigenicity andancement of radiochemosensitivity in head and neck squamousl cancer-derived ALDH1-positive cells by knockdown of Bmi-1.l Oncol 2010;46:158–65.wman T, Garcia R, Turkson J, Jove R. STATs in oncogenesis.cogene 2000;19:2474–88.mberg J. Stat proteins and oncogenesis. J Clin Invest 2002;109:

ling-Burnette PK, Liu JH, Catlett-Falcone R, et al. Inhibition ofAT3 signaling leads to apoptosis of leukemic large granularphocytes and decreased Mcl-1 expression. J Clin Invest 2001;7:351–62.Zaid Siddiquee K, Turkson J. STAT3 as a target for inducing ap-tosis in solid and hematological tumors. Cell Res 2008;18:254–67.mberg JF, Wrzeszczynska MH, Devgan G, et al. Stat3 as an on-gene. Cell 1999;98:295–303.G, Wright KL, Ma Y, et al. Role of Stat3 in regulating p53 expres-n and function. Mol Cell Biol 2005;25:7432–40.H, Jove R. The STATs of cancer-new molecular targets come ofe. Nat Rev Cancer 2004;4:97–105.lyankrishna S, Grandis JR. Epidermal growth factor receptor biol-y in head and neck cancer. J Clin Oncol 2006;24:2666–72.ero DC, Poli V, David M, Rickert RC. Cutting edge: inherent andquired resistance to radiation-induced apoptosis in B cells: a piv-l role for STAT3. J Immunol 2006;177:6593–7.KW, Mutter RW, Cao C, et al. Inhibition of signal transducer and

tivator of transcription 3 activity results in down-regulation of Sur-in following irradiation. Mol Cancer Ther 2006;5:2659–65.nner JA, Trummell HQ, Willey CD, Plants BA, Raisch KP. InhibitionSTAT-3 results in radiosensitization of human squamous cell car-oma. Radiother Oncol 2009;92:339–44.nerjee S, Byrd JN, Gianino SM, et al. The neurofibromatosis type 1or suppressor controls cell growth by regulating signal transduc-

and activator of transcription-3 activity in vitro and in vivo. Cancers 2010;70:1356–66.iou SH, Kao CL, Chen YW, et al. Identification of CD133-positiveioresistant cells in atypical teratoid/rhabdoid tumor. PLoS ONE08;3:e2090.
o CL, Huang PI, Tsai PH, et al. Resveratrol-induced apoptosis andreased radiosensitivity inCD133-positive cells derived fromatypicalatoid/rhabdoid tumor. Int J Radiat Oncol Biol Phys 2009;74:219–28.



24. Chmiislepri

25. YaHIF

26. JintracleCa

27. Chpucan

28. Matarsqu

29. ChnalepPh

30. Kruresca

31. Rioscrtio12

32. JaymaLe

33. CoEGch13

34. Suthece

Chen et al.

Mol C2892

Dow


iou SH, Chen SJ, Chang YL, et al. MafA promotes the reprogram-ng of placenta-derived multipotent stem cells into pancreaticts-like and insulin-positive cells. J Cell Mol Med. Epub ahead ofnt 2010 Feb 16.ng MH, Wu MZ, Chiou SH, et al. Direct regulation of TWIST by-1alpha promotes metastasis. Nat Cell Biol 2008;10:295–305.g N, Zhu Q, Yuan P, Li Y, Mao L, Tweardy DJ. Targeting signalnsducer and activator of transcription 3 with G-quartet oligonu-otides: a potential novel therapy for head and neck cancer. Molncer Ther 2006;5:279–86.en YC, Chen YW, Hsu HS, et al. Aldehyde dehydrogenase 1 is atative marker for cancer stem cells in head and neck squamouscer. Biochem Biophys Res Commun 2009;385:307–13.rur S, Forastiere AA. Challenges of integrating chemotherapy andgeted therapy with radiation in locally advanced head and neckamous cell cancer. Curr Opin Oncol 2010;22:206–11.en CC, Chen WC, Lu CH, et al. Significance of interleukin-6 sig-
ing in the resistance of pharyngeal cancer to irradiation and theidermal growth factor receptor inhibitor. Int J Radiat Oncol Biolys 2010;76:1214–24.
35. Sacoco



ser TJ, Armstrong EA, Ghia AJ, et al. Augmentation of radiationponse by panitumumab in models of upper aerodigestive tractncer. Int J Radiat Oncol Biol Phys 2008;72:534–42.s JL, Recio MC, Escandell JM, Andujar I. Inhibition of tran-iption factors by plant-derived compounds and their implica-ns in inflammation and cancer. Curr Pharm Des 2009;15:12–37.aprakasam B, Seeram NP, Nair MG. Anticancer and antiinflam-tory activities of cucurbitacins from Cucurbita andreana. Cancertt 2003;189:11–6.lomiere M, Ward AC, Riley C, et al. Cross talk of signals betweenFR and IL-6R through JAK2/STAT3 mediate epithelial-mesen-ymal transition in ovarian carcinomas. Br J Cancer 2009;100:4–44.llivan NJ, Sasser AK, Axel AE, et al. Interleukin-6 induces an epi-lial-mesenchymal transition phenotype in human breast cancerlls. Oncogene 2009;28:2940–7.dzuka Y, Hatakeyama H, Sonobe T. Enhancement of doxorubicinncentration in the M5076 ovarian sarcoma cells by cucurbitacin E
-treatment. Int J Pharm 2010;383:186–91.



2010;9:2879-2892. Published OnlineFirst November 9, 2010.Mol Cancer Ther Yi-Wei Chen, Kuan-Hsuan Chen, Pin-I Huang, et al.

Cells+ALDH1+Derived CD44−Carcinoma Radiation-Induced Apoptosis in Head and Neck Squamous

Cucurbitacin I Suppressed Stem-Like Property and Enhanced

Updated version

10.1158/1535-7163.MCT-10-0504doi:

Access the most recent version of this article at:

Material

Supplementary

http://mct.aacrjournals.org/content/suppl/2010/11/11/1535-7163.MCT-10-0504.DC1

Access the most recent supplemental material at:

Cited articles

http://mct.aacrjournals.org/content/9/11/2879.full#ref-list-1

This article cites 34 articles, 9 of which you can access for free at:

Citing articles

http://mct.aacrjournals.org/content/9/11/2879.full#related-urls

This article has been cited by 5 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. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://mct.aacrjournals.org/content/9/11/2879To request permission to re-use all or part of this article, use this link

on December 16, 2020. © 2010 American Association for Cancer Research.mct.aacrjournals.org Downloaded from


http://mct.aacrjournals.org/lookup/doi/10.1158/1535-7163.MCT-10-0504

http://mct.aacrjournals.org/content/suppl/2010/11/11/1535-7163.MCT-10-0504.DC1

http://mct.aacrjournals.org/content/9/11/2879.full#ref-list-1

http://mct.aacrjournals.org/content/9/11/2879.full#related-urls

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

mailto:[email protected]

http://mct.aacrjournals.org/content/9/11/2879


Documents

Therapeutics Cucurbitacin I Suppressed Stem-Like Property and … · Therapeutic Discovery Cucurbitacin I Suppressed Stem-Like Property and Enhanced Radiation-Induced Apoptosis in