integrin-linked kinase is a potential therapeutic target ... · anaplastic thyroid cancer and ilk...

Integrin-linked kinase is a potential therapeutictarget for anaplastic thyroid cancer

Maher N. Younes,1 Seungwon Kim,1

Orhan G. Yigitbasi,1 Mahitosh Mandal,1

Samar A. Jasser,1 Yasemin Dakak Yazici,1

Bradley A. Schiff,1 Adel El-Naggar,2

Benjamin N. Bekele,3 Gordon B. Mills,4

and Jeffrey N. Myers1,5

Departments of 1Head and Neck Surgery, 2Pathology,3Biostatistics, 4Molecular Therapeutics, and 5CancerBiology, The University of Texas M.D. AndersonCancer Center, Houston, Texas

AbstractWe investigated integrin-linked kinase (ILK), a focaladhesion serine-threonine protein kinase, as a newmolecular target for treating anaplastic thyroid cancer.ILK mediates cell growth and survival signals and isoverexpressed in a number of cancers. Therefore, wehypothesized that inhibition of ILK leads to growth arrestand apoptosis of thyroid cancer cells. According toWestern blotting, the level of ILK protein was highlyexpressed in one papillary (NPA187) and four of five(Hth74, DRO, ARO, KAT4, and K4) anaplastic thyroidcancer cell lines. Immunohistochemical analysis of ahuman tissue microarray revealed that ILK was highlyexpressed in anaplastic thyroid cancer but not in normalhuman thyroid tissue. Treating thyroid cancer cell lineswith a new ILK inhibitor, QLT0267, inhibited epidermalgrowth factor–induced phosphorylation of AKT, inhibitedcell growth, and induced apoptosis in the NPA187, DRO,and K4 cell lines. QLT0267 also inhibited the kinaseactivity of immunoprecipitated ILK in four of five cell lines.Tumor volumes in mice treated with QLT0267 weresignificantly reduced compared with those in untreatedmice. In immunohistochemical studies, QLT0267 sup-pressed phosphorylated p-AKT and angiogenesis (i.e.,reduced mean vascular density) and induced apoptosis inboth tumor cells and tumor-associated endothelial cells ofthe thyroid DRO xenografts. In summary, we found that

ILK expression and activity were elevated in humananaplastic thyroid cancer and ILK inhibition led to growtharrest and apoptosis in vitro and in vivo. Our resultsprovide preliminary evidence that ILK is a potentialtherapeutic target for treating anaplastic thyroid cancer.[Mol Cancer Ther 2005;4(8):1146–56]

IntroductionIn the United States, thyroid cancer accounts for about 1%of all new cases of cancer. Pathologically, 94% of thyroidcancers are differentiated thyroid carcinomas (i.e., papil-lary and follicular thyroid cancer; ref. 1). Another 5% aremedullary tumors of a neuroendocrine origin. Theremaining 1% of thyroid cancers is anaplastic (2).Although the incidence of anaplastic thyroid cancer islow, this disease has a major effect because the associatedmortality rate is nearly 100% (3–8). In a study of 15,700patients with thyroid cancer in the United States, theoverall 10-year relative survival rate, corrected for age andsex, was 13% for anaplastic disease (9). The mediansurvival time after diagnosis is 3 to 7 months, and a worseprognosis is associated with large tumors, distant metas-tasis, acute obstructive symptoms, and leukocytosis(10, 11). Despite the widespread use of multimodalitytreatment, survival rates have not improved much in thepast few decades (12, 13). Patients presenting withwidespread local invasion often develop distant metastasisin the lungs, pleurae, bone, and brain (1). Treatment ofanaplastic cancer is generally palliative only; thus, newtargets for molecular therapy of this cancer are needed.

Among the molecular targets that show promise in thefight against anaplastic cancer is integrin-linked kinase(ILK), a recently identified protein serine/threonine kinasethat was discovered through its interactions with the h1 andh3 integrin subunits (14). In general, integrin-mediatedinteractions of cells with components of the extracellularmatrix regulate cell survival, proliferation, differentiation,and migration (15). ILK has four noncatalytic ankyrinrepeats that aid in targeting it to focal adhesion complexesand linking it to growth factor receptor tyrosine kinasesignaling (16–18).

ILK has a number of oncogenic properties. First, inepithelial cells, forced expression of wild-type ILKsuppresses suspension-mediated apoptosis (i.e., anoikis)and stimulates anchorage-independent growth (19).Second, ILK overexpression tends to constitutively activateintegrin-signaling pathways so that the anchorage state ismimicked in the absence of cell-extracellular matrixinteractions (20). Third, ILK is involved in enhancingtumor cell invasion (21) and tumorigenicity in nude mice(22). Fourth, ILK is overexpressed in colon (23) andprostate tumors (24). Moreover, ILK expression is closely

Received 3/18/ 05; revised 6/6/ 05; accepted 6/14/ 05.

Grant support: The University of Texas M.D. Anderson Multi-DisciplinaryProgram in Thyroid Cancer and Golfers Against Cancer.

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 toindicate this fact.

Requests for reprints: Jeffrey N. Myers, Department of Head and NeckSurgery, The University of Texas M.D. Anderson Cancer Center, Unit 441,1515 Holcombe Boulevard, Houston, TX, 77030-4009. Phone: 713-792-6920; Fax: 713-794-4662. E-mail: [email protected]

Copyright C 2005 American Association for Cancer Research.

doi:10.1158/1535-7163.MCT-05-0078

1146

Mol Cancer Ther 2005;4(8). August 2005

Research. on November 10, 2020. © 2005 American Association for Cancermct.aacrjournals.org Downloaded from

http://mct.aacrjournals.org/

correlated with the invasion and metastasis of gastriccancer (25). In addition, ILK phosphorylates a majorprotein called protein kinase B or AKT kinase on Ser473

and activates it in PTEN mutant prostate cancer cells (26)and directly phosphorylates GSK-h, thereby inhibiting itsactivity (27). Recently, it has been shown that conditionalknockout of ILK results in almost complete inhibition ofphosphorylation on Ser473, significant inhibition of proteinkinase B or AKT kinase activity, and suppression ofphosphorylation of GSK-3h on Ser9 and cyclin inhibition ofD1 expression (28, 29). Moreover, ILK plays a key role inmediating tumor angiogenesis (30).

Until now, the role of ILK expression and activity inthyroid carcinomas has not been studied; the use of ILKinhibitors in cancer therapy is in its early stages. Identifyingthe role of ILK in thyroid cancer and the use of new ILKinhibitors in thyroid cancer treatment has the potential toidentify a new molecular and therapeutic target that isurgently needed in the face of the poor survival rates forpatients with anaplastic thyroid cancer.

In this study, we hypothesized that ILK is overexpressedin anaplastic and papillary thyroid cancers and that itsinhibition leads to cell cycle arrest and apoptosis. Weinvestigated the expression of ILK in a panel of thyroidcancer cell lines, xenografts, and human thyroid cancerspecimens. We also studied the effects of a novel ILKinhibitor on ILK kinase activity, cell growth, and survival ofthyroid cancer cells in vitro and on anaplastic thyroidcancer xenografts growth in vivo .

Materials andMethodsCell Lines and Culture ConditionsThe anaplastic thyroid carcinoma cell lines KAT-4, K18,

Hth74, ARO, and DRO were used. NPA187 thyroid cancercell line represented the less aggressive yet the mostprevalent form of thyroid cancer. PC3P, a prostate cancercell line, was used as a positive control for ILK expression.The cells were grown in DMEM supplemented with 10%fetal bovine serum, L-glutamine, penicillin, sodium pyru-vate, nonessential amino acids, and a 2-fold vitaminsolution (DMEM 10% fetal bovine serum; Life Techno-logies, Inc., Grand Island, NY). Adherent monolayercultures were maintained on plastic and incubated at37jC in 5% carbon dioxide and 95% air. The cultures werefound to be free of Mycoplasma species. The cultures weremaintained no longer than 12 weeks after recovery fromfrozen stocks.

ReagentsThe ILK inhibitor QLT0267 was obtained from QLT, Inc.

(Vancouver, British Columbia, Canada). QLT0267 wasdeveloped through the optimization of a lead compoundidentified in the high-throughput screening of a rationallydesigned small-molecule library against the target ILK.The initial lead compound was found to be a submicro-molar inhibitor of the phosphotransferase activity of ILKtowards a specific peptide substrate. An intensivemedicinal chemistry effort around this lead compound

has identified many structure/activity relationships andallowed for the identification of a bioactive core of themolecule. Further optimization of this molecular core toenhance both in vitro enzyme and cell-based potenciesand to improve the pharmacokinetic and pharmacody-namic properties of this class of compounds has lead tothe second-generation compound QLT0267. QLT0267 hasbeen shown to inhibit the kinase activity of ILK in cell-free assay at 26 nmol/L and preliminary experimentssuggest that it has f1,000-fold selectivity over otherkinases tested under similar conditions, including CK2,CSK, DNA-PK, PIM1, protein kinase B or AKT kinase,and PKC; f100-fold selectivity over extracellular signal-regulated kinase 1, GSK3h, LCK, PKA, p70S6K, and RSK1(QLT).6 Of those tested, CDK1, CDK2, and CDK5 showthe greatest inhibition by QLT0267 but the selectivitywindow is still at least 10-fold. The drug was prepared as10 mg/mL solution by adding PTE vehicle [66.6%polyethylene glycol 300/8.2% Tween 80/25% ethanol(95%)/0.2% citric acid (w/w)] to the powder, sonicatingfor 10 minutes, and vortexing until fully dissolved. Thedosing solutions were prepared in one batch and stored at�80jC, to be thawed on the day of dosing. For in vitroadministration, QLT0267 was dissolved in DMSO to aconcentration of 20 mmol/L. 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) was purchasedfrom Sigma-Aldrich Corp. (St. Louis, MO). The AKTkinase kit was purchased from Cell Signaling Technology(Beverly, MA).

MaterialsILK antibodies were purchased from Upstate Biosciences

(Waltham, MA). For Western blotting, an anti-ILKantibody was used at a dilution of 1:5,000. For immuno-precipitation, an ILK antibody was used. The anti-p-AKTSer473 antibody was purchased from Cell SignalingTechnology and was used at a dilution of 1:2,000. Forsecondary antibodies, we used goat anti-mouse secondaryantibody from Bio-Rad Laboratories (Hercules, CA) at adilution of 1:3,000 and a goat anti-rabbit antibody fromSanta Cruz Biotechnology (Santa Cruz, CA) at a 1:3,000dilution. Protein A beads and myelin basic protein werepurchased from Upstate Biotechnology, Inc. (LakePlacid, NY).

Measurement of CytotoxicityFive thousand cells were plated into 38-mm2 wells of 96-

well tissue culture plates. The cells were grown in DMEMsupplemented with sodium pyruvate, essential aminoacids, and 10% fetal bovine serum. After a 24-hourattachment period, the cells were refed with this medium(the negative control was refed with DMEM alone) ormedium containing the ILK inhibitor QLT0267. Afterincubation for 3 days, the number of metabolically activecells was determined by MTT assay: the conversion of MTTto formazan was measured by a 96-well microtiter platereader (MR-5000; Dynatech, Inc., Chantilly, VA) at an

6 D. Morrison, personal communication.

Molecular Cancer Therapeutics 1147




absorbance of 570 nm. Growth inhibition was calculatedwith the following formula: cytostasis (%) = [1 � (A/B)] �100, where A is the absorbance of treated cells and B is thatof control cells.

Measurement of ApoptosisCells were plated at a density of 5 � 105 in 38-mm2 six-

well plates (Costar, Cambridge, MA) and incubated at 37jCfor 24 hours before treatment with QLT0267. Seventy-twohours later, the extent of cell death was determined bypropidium iodide staining of hypodiploid DNA: 3 � 105

cells were resuspended in Nicoletti buffer (50 Ag/mL; SigmaChemical, St. Louis, MO), 0.1% sodium citrate, 0.1% TritonX-100, and 1 mg/mL of RNase (Roche, Basel, Switzerland)in PBS and then analyzed by fluorescence-activated cellsorter analysis (FACScan, Becton Dickinson, MountainView, CA). The fraction of cells with sub-G1 DNA contentwas assessed using the Lysis program (Becton Dickinson).The percentage of specific apoptosis was calculated bysubtracting the percentage of spontaneous apoptosis of therelevant controls from the total percentage of apoptosis.

Western ImmunoblottingCells were grown in 10-cm culture flasks (Corning, Life

Sciences, Inc., Big Flats, NY) in DMEM 10% fetal bovineserum. After the cells reached 70% to 80% confluency, themedium was discarded and adherent cells were washedtwice with ice-cold PBS. The cells were then solubilized inlysis buffer [1� concentration: 20 mmol/L Tris (pH 7.5),150 mmol/L sodium chloride, 1 mmol/L EDTA, 1 mmol/LEGTA, 1% Triton X-100, 2.5 mmol/L sodium pyrophos-phate, 1 mmol/L h-glycerolphosphate, 1 mmol/L Na3VO4,1 Ag/mL of leupeptin, and 1 mmol/L phenylmethylsul-fonyl fluoride] for 15 minutes on ice.

Lysates were clarified by centrifugation at 10,000 � g for15 minutes at 4jC. Equal amounts of protein (30–50 Ag)were resolved by 10% SDS-PAGE and transferred to anitrocellulose membrane (Millipore, Billerica, MA), whichwas blocked with 1% milk in TBS-T for 30 to 45 minutes atroom temperature. Afterwards, the primary antibody wasadded at the proper dilution and left overnight at 4jC.After the membrane was washed thrice with TBS-T for 10minutes each, the corresponding secondary antibody wasadded at the proper dilution for 1 hour at roomtemperature. Then, the membrane was washed thrice withTBS-T, and protein detection was carried out usinghorseradish peroxidase–conjugated anti-rabbit immuno-globulin G (IgG) from Santa Cruz Biotechnology or anti-mouse IgG from Sigma Chemical and an enhancedchemiluminescence kit (Amersham Pharmacia, Little Chal-font Buckinghamshire, United Kingdom).

ILKKinaseAssayLysate (250 Ag) was immunoprecipitated overnight with

1 Ag of rabbit polyclonal anti-ILK antibody (UpstateBiotechnology) or 1 Ag of polyclonal rabbit IgG (SantaCruz Biotechnology) as a negative control. We used thenonradioactive AKT kinase kit from Cell Signaling Tech-nology. In brief, 30 AL of protein A agarose beads wereadded to the immune complexes and rotated for 2 to 3

hours at 4jC. Afterwards, they were centrifuged for 30seconds at 4jC, and the pellets were washed twice with500 AL of 1� lysis buffer and twice with 500 AL of 1� kinasebuffer [25 mmol/L Tris (pH 7.5), 5 mmol/L h-glycerol-phosphate, 2 mmol/L DTT, 0.1 mmol/L Na3VO4, and10 mmol/L MgCl2].

For the cold kinase assay, the pellets were suspended in40 AL of 1� kinase buffer supplemented with 200 Amol/LATP and 1 Ag of GSK-3 fusion protein. After incubation at30jC for 30 minutes, the reaction was terminated using 20AL of 3� SDS sample buffer, mixed with a vortex mixer,and then spun in a microcentrifuge for 2 minutes. Thesupernatant was transferred to a new tube; the samplewas boiled for 5 minutes and loaded (15–30 AL) onto 10%SDS-polyacrylamide gel. Phosphorylation of the substratewas detected by Western blot analysis with the anti-GSK-3Ser21/9 antibody at 1:1,000 dilution (Cell SignalingTechnology).

AnimalsEight- to 12-week-old male athymic nude C57BL/6 mice

were purchased from the National Cancer Institute(Bethesda, MD) and housed in a specific pathogen-freeanimal facility. The animals were fed irradiated mouse foodand autoclaved reverse osmosis treated water. All animalprocedures were done in accordance with a protocolapproved by our institutional Animal Care and UseCommittee. All mice were euthanized after 6 weeks byasphyxiation with carbon dioxide.

Subcutaneous Xenografts to Assess In vivo ILKExpression

Thyroid cancer cells were harvested from subconfluentcultures by trypsinization and washed. One million cellsof DRO, Hth74, ARO, K4, and NPA187 were separatelyinjected s.c. using a 30-gauge needle into the flanks of theathymic nude mice (10 mice per cell line). The mice wereweighed and the tumors measured weekly using micro-calipers until the mice were euthanized 2 weeks after theinjection of tumor cells. For immunohistochemical androutine H&E staining, the tumor tissue was fixed informalin and embedded in paraffin.

Human Thyroid Tissue ArraysThyroid tumor tissue arrays representative of the entire

spectrum of benign and malignant neoplasms, includingATC constructed at the head and neck tissue care facility,were used to screen for epidermal growth factor receptorexpression. The arrays represented 25 papillary carcino-mas, 33 medullary carcinomas, 21 anaplastic carcinomas,55 follicular carcinomas, 20 Hurthle carcinomas, and eightsamples of normal nondiseased thyroid tissue. Two coresof each sample were placed differentially in the recipientblock. Two pathologists scored these blindly and indepen-dently on a scale of 0 to 3. Whenever a discrepancy inscoring was noted, both pathologists reexamined thesample in question and a consensus was reached.

Immunohistochemical Staining of MurineTumor andHumanTissue Sections

Antibody staining was done on 5-Am histologic sec-tions of formalin-fixed, paraffin-embedded specimens. All

Integrin-Linked Kinase in Anaplastic Thyroid Cancer1148




sections were deparaffinized and hydrated by exposure tothe following: xylene for 3 minutes; 100% ethanol for2 minutes, twice; 95% ethanol for 1 minute, twice; 80%ethanol for 1 minute; and PBS for 2 minutes, twice.Afterwards, antigen retrieval was done using Targetretrieval solution (DakoTarget; DAKO Corp., Carpinteria,CA). After incubation with 0.3% peroxide in methanol andincubation with normal blocking serum for 30 minutes atroom temperature, sections were incubated overnight at4jC with primary anti-ILK antibody at 1:100 dilution(Upstate Biotechnology). Immunodetection was donewith peroxidase-labeled secondary antibody diluted inblocking solution for 1 hour at room temperature usingDAP as the chromogen. Positive controls for ILK immuno-staining were formalin-fixed, paraffin-embedded surgicalspecimens of human prostate tissue. Negative controlsincluded sections stained with nonspecific rabbit IgG atthe same protein concentration as the primary anti-ILKantibody. All sections were counterstained with Gill’shematoxylin.

Tumor Xenograft Generation for the ILK InhibitorTreatment Study

To produce tumors, DRO cells were harvested fromsubconfluent cultures by a brief 2-minute exposure to0.25% trypsin and 0.02% EDTA. Trypsinization wasstopped with medium containing 10% fetal bovine serum.The cells were washed once in serum-free medium andresuspended in HBSS. Only suspensions consisting ofsingle cells with >90% viability were used for the injections.One million DRO cells were suspended in 0.1 mL of HBSSand injected s.c. into the right flank area of nude mice. After7 days, all mice were examined. Forty mice with homoge-neous tumors were identified and then randomized intofour groups (n = 10 each): group 1 was the untreatedcontrol group, group 2 was the group treated with PTEvehicle only, group 3 was the group treated with QLT0267(50 mg/kg) by oral gavage daily, and group 4 was thegroup treated with QLT0267 (100 mg/kg) by oral gavagedaily. Tumor size was measured twice a week in eachmouse with calipers. Tumor volume (in mm3) wascalculated using the formula AB2p/6, where A is thelength of the longest aspect of the tumor and B is the lengthof the tumor perpendicular to A.

The mice were euthanized once they became moribundor lost >20% of their recorded initial body weight beforesuccumbing to tumor progression. After the mice wereeuthanized, necropsies were done and the xenograftedtumors were obtained. For immunohistochemical androutine H&E staining, half of the tumor tissue was fixedin formalin and embedded in paraffin. The other half wasembedded in ornithine carbamyl transferase compound(Miles, Inc., Elkhart, IN), rapidly frozen in liquid nitrogen,and stored at �70jC.

Immunohistochemical Detection of CD31/Platelet/Endothelial Cell Adhesion Molecule 1, ILK, p-AKT, andTotal AKT

Frozen tissues were sectioned into 8- to 10-Am slices andused for detection of AKT/p-AKT and CD31/platelet/

endothelial cell adhesion molecule 1. The slices weremounted on positively charged Plus slides (Fisher Scien-tific, Pittsburgh, PA) and air-dried for 30 minutes; fixed incold acetone (5 minutes), 1:1 acetone/chloroform (v/v; 5minutes), and acetone (5 minutes); and then washed withPBS. Immunohistochemical procedures were done asdescribed previously. Dilutions of primary antibodieswere as follows: AKT, 1:100; p-AKT, 1:100; CD31/platelet/endothelial cell adhesion molecule 1, 1:400; andILK, 1:100.

Peroxidase-conjugated secondary antibody was used forimmunohistochemical analysis of CD31/platelet/endothe-lial cell adhesion molecule 1. A positive reaction for p-AKTwas visualized by incubating the slides for 1 hour with a1:600 dilution of conjugated secondary antibody. Bleachingof fluorescence was minimized by covering the slides with90% glycerol and 10% PBS. A positive reaction wasvisualized by incubating the slides with stable 3,3V-diaminobenzidine for 10 to 20 minutes for identificationof CD31/platelet/endothelial cell adhesion molecule 1. Thesections were rinsed with distilled water, counterstainedwith Gill’s hematoxylin for 1 minute, and mounted withUniversal Mount (Research Genetics, Huntsville, AL).Control samples that had not been exposed to primaryantibody showed no specific staining.

Staining for CD31/Terminal Deoxynucleotidyl Trans-ferase ^Mediated Nick-End Labeling and Double Immu-nofluorescenceAssays

For terminal deoxynucleotidyl transferase – mediatednick-end labeling (TUNEL) and double immunofluores-cence assays, frozen tissues were used. After beingmounted on slides and fixation with acetone as describedabove, the frozen samples were washed thrice with PBS,incubated with protein-blocking solution containing 5%normal horse serum and 1% normal goat serum in PBS for20 minutes at room temperature, and then incubated with a1:400 dilution of rat anti-mouse CD31 monoclonal antibody(human cross-reactive) overnight at 4jC. After the sampleswere rinsed thrice with PBS for 3 minutes each, the slideswere incubated for 1 hour at room temperature in the darkwith a 1:600 dilution of secondary goat anti-rat antibodyconjugated to Alexa Fluor 594 (red fluorescence). Thesamples were then washed thrice with PBS containing 0.1%Brij and once with PBS for 3 minutes.

TUNEL AssayA TUNEL assay was done using an apoptosis detection

kit (Promega, Madison, WI) with the following modifica-tions: samples were fixed with 4% paraformaldehyde(methanol free) for 10 minutes at room temperature,washed twice with PBS for 5 minutes, and incubated with0.2% Triton X-100 for 15 minutes at room temperature.After two 5-minute washes with PBS, the samples wereincubated with equilibration buffer for 10 minutes at roomtemperature. The equilibration buffer was drained, andreaction buffer, which contained 44 AL of equilibrationbuffer, 5 AL of nucleotide mix, and 1 AL of terminaldeoxynucleotidyl transferase (supplied in the kit), wasadded to the tissue sections and incubated in a humid





atmosphere at 37jC for 1 hour, avoiding exposure to light.The reaction was terminated by immersing the samples in2� SSC for 15 minutes. Samples were then washed thricefor 5 minutes to remove unincorporated fluorescein-dUTP.

Quantification of Mean Vascular Density and Apop-totic Endothelial Cells

For quantification analysis, five slides were prepared foreach group, and two areas were selected on each slide. Toquantify the expression in the TUNEL assays, the number ofpositively stained cells and total cells were also counted in10 random 0.159-mm2 fields of tumor area at �100 magnifi-cation, and the percentages of positively stained cells fromamong the total number of cells were calculated andcompared.

Immunofluorescence microscopy was done using a NikonMicrophot-FX (Nikon, Inc., Garden City, NY) equippedwith a HBO 100 mercury lamp and narrow band-pass filtersto individually select for green, red, and blue fluorescence(Chroma Technology Corp., Brattleboro, VT). Images werecaptured using a cooled charged coupled device Hama-matsu 5810 camera (Hamamatsu Corp., Bridgewater, NJ)and Optimas Image Analysis software (Media Cybernetics,Silver Spring, MD). Stained sections were examined in aNikon Microphot-FX microscope equipped with a three-chip charged coupled device color video camera (modelDXC990, Sony Corp., Tokyo, Japan). Photomontages wereprepared using Photoshop software (Adobe Systems, Inc.,San Jose, CA). Endothelial cells were identified by redfluorescence staining, and DNA fragmentation was detectedby localized green fluorescence within the nuclei ofapoptotic cells. Photomontages were printed in a Sonydigital color printer (model UPD7000).

To quantify mean vascular density (MVD), the areascontaining the higher number of tumor-associated bloodvessels were identified by scanning the tumor sections atlow microscopic power (�40). Vessels that were completely

stained with anti-CD31 antibodies were then counted in 10random 0.159-mm2 fields at �100 magnification. Quantifi-cation of apoptotic endothelial cells was expressed as theaverage of the ratios of apoptotic endothelial cells to thetotal number of endothelial cells in 10 random 0.011-mm2

fields at �400 magnification.

Statistical MethodsDifferences between various tumor types were analyzed

for their correlation with ILK expression. To test forstatistical significance between the intensities of ILKstaining in human tissues, associations between categoricalvariables were assessed via cross-tabulation and Fisher’sexact test. All computations were carried out on a personalcomputer using the Windows NT operating system (Micro-soft, Redmond, WA) using StatXact 4.0 software.

To assess the differences in tumor volumes between thefour groups of mice and to statistically quantitate the dif-ferences in TUNEL, MVD, and CD31-TUNEL staining,we used Wilcoxon’s test or the nonparametric Kruskal-Wallis test. The Kruskal-Wallis test was used to detectthe difference between the groups on each observation day,and Wilcoxon’s test was used to assess the differencebetween each treatment and control group on eachobservation day. All statistical computations were doneon a Dell 1,000-mHz personal computer using SASstatistical software (SAS Institute, Cary, NC). Ps < 0.05were considered statistically significant.

ResultsILK Protein Expression Is Elevated in HumanThyroid

Cancer SpecimensAfter verifying the overexpression of ILK in thyroid

cancer cell lines, we examined whether this observationtranslates to human thyroid cancer specimens. Usingimmunohistochemical techniques, we assessed the levelof ILK protein expression in a panel of human thyroid

Figure 1. Immunohistochemicalstaining for ILK on a human thyroidcancer tissue microarray. The inten-sity of the ILK staining was gradedas 0, 1+, 2+, and 3+. Mostnormal thyroid tissue stained slight-ly for ILK (+1). On the other hand,most of the anaplastic thyroid can-cer tissues stained intensely (+3 ).





tissue microarrays that contained normal thyroid tissue andspecimens of various thyroid cancer subtypes. AlthoughILK protein expression was detected in most of the normalthyroid tissue sections (75% of slides), the intensity ofstaining was significantly higher in the tumor sections. Thepercentages of slides staining positively at 2+ and 3+ levelswere as follows: normal (12.5%), papillary (12%), medullary(39.4%), follicular (27.3 %), anaplastic (57%), and Hurthlecell carcinoma (77.8%; Fig. 1; Table 1). Both anaplastic andHurthle cell (a subgroup of follicular) thyroid carcinomasstained significantly stronger than normal human thyroidtissue (P = 0.044 and 0.011, respectively).

ILK Protein Expression and Kinase Activity Are Ele-vated in a Panel of Thyroid Cancer Cell Lines

Using Western blotting, we assessed ILK proteinexpression in one papillary and five anaplastic thyroidcancer cell lines. ILK protein was expressed in all celllines and was highly so in the DRO, K4, KAT-4, andARO anaplastic cell lines. The anaplastic (Hth74)thyroid cancer and the papillary (NPA187) cell linesshowed lower expression of the ILK protein (Fig. 2A).In an in vitro ILK kinase assay, all six thyroid cancercell lines showed high levels of ILK kinase activity(Fig. 2B).

Table 1. Immunohistochemical analysis of ILK expression in a human papillary cancer tissue microarray

Tissue type No. tumors Staining intensity % Specimens staining2+ and 3+

% Specimens stainingpositive for ILK

0 1+ 2+ 3+

Normal 8 2 5 0 1 12.5 75Papillary 25 15 7 3 0 12 40Medullary 33 10 10 7 6 39.4 71.4Follicular 55 32 8 8 7 27.3 41.8Anaplastic 21 4 5 4 8 57* 81Hurthle cell 20 2 4 5 9 77.8* 81.8Total 162 65 39 27 31

*Statistically significant compared with normal (P = 0.044, anaplastic; P = 0.011, Hurthle; Fisher’s exact test).

Figure 2. Immunoblot demonstrat-ing ILK in six different thyroid cancercell lines: papillary cell line NPA187and anaplastic cell lines ARO, K4,DRO, KAT-4, and Hth74. ILK wasexpressed at a low level in NPA187cells but was highly expressed in allanaplastic cell lines. PC3P, a prostatecancer cell line, was the positivecontrol for ILK expression (A). ILKin vitro kinase assay (measured byGSK phosphorylation by ILK) for sixdifferent thyroid cancer cell linesshowing that ILK kinase activitywas present in all six tested thyroidcell lines; a nonspecific rabbit IgGwas used as a control (B). The resultsare representative of three indepen-dent experiments. C, results of im-munohistochemical testing for ILKprotein on a panel of thyroid cancerxenografts from the flanks of athy-mic nude mice. No ILK staining wasdetected in the normal thyroid glandof these mice, although strong stain-ing for ILK was detected in almost allthe tumors arising from the thyroidcancer cell lines. The exceptionswere the papillary (NPA187) andanaplastic (Hth74) thyroid cancer celllines, which stained only slightly (C).





ILK Is Overexpressed inThyroid Cancer XenograftsTo determine whether ILK expression in the previously

tested thyroid cancer cell lines exceeds that in normalthyroid cells, we s.c. injected one million cells each ofNPA187, ARO, DRO, K4, and Hth74 cell lines in nude mice.On immunohistochemical analysis, ILK was overexpressedin all tested cell line xenografts compared with normalthyroid tissue (Fig. 2C).

QLT0267 Inhibits Growth of Thyroid Cancer CellsIn vitro

QLT0267 was developed through the optimization of alead compound identified in the high-throughput screen-ing of a rationally designed small molecule library againstthe target ILK. The initial lead compound was found to be asubmicromolar inhibitor of the phosphotransferase activityof ILK towards a specific peptide substrate. An intensivemedicinal chemistry effort around this lead compound hasidentified many structure/activity relationships andallowed for the identification of a bioactive core of themolecule. Further optimization of this molecular core toenhance both in vitro enzyme and cell-based potencies andto improve the pharmacokinetic and pharmacodynamicproperties of this class of compounds has lead to thesecond-generation compound QLT0267. QLT0267 has beenshown to inhibit the kinase activity of ILK in cell-free assayat 26 nmol/L and preliminary experiments suggest that ithas f1,000-fold selectivity over other kinases tested undersimilar conditions, including CK2, CSK, DNA-PK, PIM1,protein kinase B or AKT kinase, and PKC; f100-foldselectivity over extracellular signal-regulated kinase 1,GSK3h, LCK, PKA, p70S6K, and RSK1 (QLT).6 Of thosetested, CDK1, CDK2, and CDK5, show the greatestinhibition by QLT0267 but the selectivity window is stillat least 10-fold. In an MTT assay, one papillary (NPA187)and two anaplastic cell lines (K4, a low expressor of ILKprotein and DRO, a high expresser of ILK protein) weresubjected to increasing concentrations of QLT0267 for 72hours. QLT0267 inhibited cell growth in all cell lines withan IC50 of f3 Amol/L (Fig. 3A).

QLT0267 Induces Apoptosis inThyroid Cancer CellsIn vitro

Using an apoptosis assay based on flow cytometry, weassessed the ability of QLT0267 to induce apoptosis in theDRO cell line. When cells were subjected to increasingconcentrations of QLT0267 for 48 hours, 10% of themunderwent apoptosis at a dose as low as 3 Amol/L, and amaximum of about 69% reached cell death at f12 Amol/L(Fig. 3B). Furthermore, when the NPA187, DRO, and K4cell lines were subjected to a 3 Amol/L concentration ofQLT0267 for 24 and 48 hours, all three cell linesunderwent apoptosis, as evidenced by the induction ofDNA fragmentation (Fig. 3C).

Figure 3. Effect of QLT0267 on cell growth, proliferation, and inductionof apoptosis in thyroid cancer cell lines. A, results of MTT assay on K4,DRO, and NPA187 thyroid cancer cell lines treated with QLT0267showing that reduction of cellular proliferation was first seen at aQLT0267 concentration of 1 Amol/L, and the IC50 was 3 Amol/L aftertreatment for 72 h. Near-maximal growth inhibition was seen at aconcentration of 15 Amol/L. B, propidium iodide apoptosis assay on theDRO cell line in the presence of 2% serum and treated for 48 h withincreasing concentrations of QLT0267 showing that apoptosis inductionwas induced at the 3 Amol/L concentration (10% cell death) and reachedits peak (69% cell death) at a concentration of about 12 Amol/L. y-axisscale, % cells undergoing apoptosis in the G0-G1 phase of growth. C,results of DNA fragmentation assay on all three thyroid cancer cell lines inthe presence of 2% serum and treated for various periods with 3 Amol/L ofQLT0267. Representative of three independent experiments.





QLT0267 Reduces ILK Kinase Activity in ThyroidCancer In vitro

In an in vitro kinase assay, 3 Amol/L QLT0267 reducedILK kinase activity in all five tested anaplastic thyroidcancer cell lines. In four of these cell lines (ARO, DRO,Hth74, and K4), ILK in vitro kinase activity was greatlydiminished compared with that of the respective DMSO-treated cells (Fig. 4A).

QLT0267 Inhibits Epidermal Growth Factor ^ InducedPhosphorylation of AKT inThyroid Cancer Cells In vitro

When Hth74 cells were starved of serum overnight,treated with as little as 1 Amol/L QLT0267 for a prolongedperiod of time, and then stimulated with epidermal growthfactor for 15 minutes, epidermal growth factor–inducedphosphorylation of AKT, as measured by Western blotting,was inhibited (Fig. 4B). There was a time dependentinhibition of the phosphorylation of AKT with a maximalinhibition reached at 6 hours.

QLT0267 Induces Tumor Growth Inhibition of DROXenografts In vivo

To assess the effects of QLT0267 on in vivo tumorgrowth, we injected DRO cells s.c. into athymic nude mice.

All mice were killed on day 16 after the start of treatmentbecause the mice in the untreated control and vehicle-only groups had large tumors or had become moribundowing to the tumor burden. As shown in Fig. 5, thetumor volumes in both groups of treated mice (50 and100 mg/kg) were significantly lower than the tumorvolumes in both the untreated control and vehicle-controlgroups (P < 0.005 at d 10 and P < 0.001 at day 16). Treat-ment with QLT0267 was well tolerated, as determined bythe maintenance of body weight of mice in both treatedgroups (data not shown).

QLT0267 Inhibitsp-AKTandInducesApoptosis In vivoTo evaluate the mechanism by which QLT0267 works

in vivo, we analyzed the DRO tumors using immunohis-tochemical and immunofluorescence analyses for markersof survival and apoptosis. No differences in the levelof ILK staining were detected in the four groups ofmice. Immunofluorescence analysis revealed that thelevels of total AKT in tumors were similar in the fourgroups. However, the levels of expression of phosphor-ylated AKT (Ser473) in the tumor tissues between theQLT0267-treated groups and the two control groupsdiffered greatly. As predicted, phosphorylated AKT wasexpressed in the two control groups but rarely expressedin the two treated groups (Fig. 6A) thus confirming thatQLT0267 treatment inhibits the phosphorylation of AKTin vivo .

To assess the effect of QLT0267 on the induction ofapoptosis in vivo , we used the immunofluorescenceTUNEL assay, which revealed that the apoptotic fractionin the tumor specimens was greater for mice treatedwith QLT0267 than it was for the two control groups ofmice (Fig. 6B). The numbers of TUNEL-positive cells perhigh-power field in the untreated control and vehicle-control groups were 3.30 F 0.82 and 3.76 F 0.88,respectively, compared with 23.87 F 2.23 in the grouptreated with 50 mg/kg of QLT0267 and 35.97 F 7.89 inthe group treated with 100 mg/kg of QLT0267 (P < 0.001for both when compared with control, Fig. 6C). Thesein vivo data are consistent with the in vitro resultsmeasured by the MTT and propidium iodide apoptosisassays.

QLT0267 Reduces MVD and Induces Tumor-Associated Endothelial Cell Apoptosis

To determine whether QLT0267 has an antiangiogeniceffect, the MVD was measured by staining the tumorspecimen with antibodies against CD31. Treatment withQLT0267 significantly reduced the number of bloodvessels from 7.78 F 1.85 in the tumors from untreatedcontrol mice and 7.53 F 1.1 in those from the vehicle-treated mice to 5.68 F 1.37 in the tumors from mice treatedwith 50 mg/kg of QLT0267 (P < 0.05) and 4.78 F 1.18 inthose from mice treated with 100 mg/kg of QLT0267(Fig. 6B, P < 0.001).

Finally, we investigated whether QLT0267 inducesapoptosis of tumor-associated endothelial cells. Endothe-lial cells of tumors resulting from injection of DRO cellsunderwent apoptosis, as shown by double staining for

Figure 4. A, immunoblot and bar graph showing the effect of the newILK inhibitor QLT0267 on the ILK in vitro kinase activity in five anaplasticthyroid cell lines. QLT0267 successfully reduced the ILK kinase activity (asmeasured by GSK phosphorylation) in four of the five thyroid cancer celllines. B,Western blot for p-AKT (Ser473) in the Hth74 cell line treated withQLT0267 (1 Amol/L) at various points showing that with the passage oftime, QLT0267 inhibited epidermal growth factor (EGF ) – induced phos-phorylation of AKT to baseline levels. Representative of three independentexperiments.





CD31 and TUNEL (Fig. 6B). The percentage of apoptoticendothelial cells was significantly higher in DRO xeno-grafts harvested 16 days after the initiation of treatmentwith QLT0267 than in those from the untreated controland vehicle-control groups (P < 0.01; Fig. 6C).

DiscussionIn this study, we showed that a panel of thyroid cancercell lines had elevated ILK protein expression and kinaseactivity. We also found that when these cell lines wereinjected s.c. into athymic nude mice, they retained theirhigh level of ILK protein expression compared with thatin the normal thyroid gland of nude mice. Furthermore,the intensity of ILK protein expression in a panel ofhuman anaplastic thyroid cancer specimens was higherthan it was in normal thyroid tissue. Treatment with anew ILK inhibitor, QLT0267, in three thyroid cancer celllines (NPA187, DRO, and K4) led to growth inhibitionand induction of apoptosis. Furthermore, treatment ofmice bearing DRO tumor xenografts with two experi-mental doses of QLT0267 significantly reduced tumorvolume by day 16. This tumor reduction was attributedto a reduction in survival signals (reduction in p-AKT),inhibition of angiogenesis (reduction in MVD), andinduction of apoptosis in both tumor cells and tumor-associated endothelial cells.

Most of these data are consistent with those fromprevious studies on ILK overexpression in various kindsof tumors (23–25, 31). For example, ILK expressionincreases with prostate tumor grade, and enhanced ILKexpression is inversely related to the 5-year patientsurvival rate patients with prostate cancer (24). Similarly,ILK expression is closely correlated with gastric cancerinvasion and metastasis (25). In addition, ILK over-expression has been linked to the progression of tumor

grade in ovarian cancer (31). In an ongoing study, wefound that ILK protein expression and activity are highlyelevated in squamous cell carcinoma of the head andneck.7

Although ILK protein expression in the tested thyroidcancer cell lines was related to their respective tumorxenografts in mice, NPA187 and Hth74 cell lines showed amuch lower expression of ILK in vivo. This finding maypoint to the fact that ILK protein expression in vivo iscontrolled by various factors that are not present in tissueculture (i.e., interactions with various components of theextracellular matrix).

Among the limitations of this study was the use ofa limited number (n = 5) of anaplastic thyroid cancercell lines. Despite this small number, the cell lines weused are among the most well established in the literature(30, 32, 33). We are currently working with a larger panelof cell lines to verify our findings.

Although our in vitro data indicated that the ILKinhibitor has direct antitumor activity, our immunohisto-chemical analysis of tumors that responded to ILKtreatment in nude mice indicated that ILK also couldinduce apoptosis in the tumor vasculature. Our findings ofdecreased MVD and tumor endothelial cell apoptosissubstantiate the ability of ILK to induce apoptosis in thetumor vasculature but do not explain whether the ILKinhibitor acts directly on the tumor endothelium orindirectly by changing the pattern of endothelial growthfactors elaborated by the tumor cells.

Other investigators have reported finding this linkbetween ILK and regulation of angiogenesis. Tan et al.(30) showed that ILK plays a pivotal role in mediatingvascular endothelial growth factor signaling in endothelialcells and induces hypoxia-inducible factor-1a-dependentvascular endothelial growth factor expression in prostatecancer cells. Our preliminary data showed that ouranaplastic thyroid cancer cell lines produced varying levelsof vascular endothelial growth factor (data not shown).In our animal experiments reported here, we detectedno change in the levels of vascular endothelial growthfactor expression, as determined immunohistochemically,between QLT0267-treated groups and the control groups.

To our knowledge, little has been published on thepharmacologic inhibition of ILK (i.e., only a few studieson the use of the specific ILK inhibitors KP-SD-1 and KP392; refs. 16, 21, 34, 35). More recently, the efficacy of asimilar ILK inhibitor QLT0254 on pancreatic carcinomagrowth in vitro and in vivo has been established (36). Wehave tested four different ILK inhibitors on a panel ofsquamous cell carcinoma of the head and neck cell linesand found that QLT0267 was the most effective inducerof growth arrest and apoptosis.8 QLT0267 has proven itsefficacy in reducing cellular growth and inducingapoptosis in various forms of thyroid cancer cells.

Figure 5. Effect of QLT0267 on DRO thyroid cancer xenografts. Graphdepicts increase in tumor volume over time in the four groups of mice:untreated control, vehicle-treated controls, and QLT0267-treated mice (50and 100 mg/kg). At the end of 16 d of treatment, only the treated miceshowed a significant reduction in tumor volume compared with the controlor vehicle-only mice (P < 0.005 at d 10 and P < 0.001 at day 16).

7 Unpublished data.8 Younes et al., in preparation.





Figure 6. Immunohistochemical analy-sis of ILK, AKT, p-AKT, CD31, TUNEL,and CD31/TUNEL staining in the fourmice groups. A, after 16 d of treatmentwith QLT0267, tumors were processedfor immunohistochemical analysis.Tumors from untreated control and vehi-cle-control mice expressed p-AKT,whereas those from mice in both treat-ment groups lost their p-AKT expression.No difference was found in total AKTexpression among the four groups. ILKexpression remained unaltered in tumorsfrom the four groups. B, in the controland vehicle groups, DRO tumors showedan increase in the number of bloodvessels (brown staining ), as marked byCD31 staining. Both treatment groupshad tumors that showed a reducednumber of blood vessels. Apoptosis asmarked by TUNEL staining was markedlyelevated in both treatment groups (greenstaining ) compared with control andvehicle tumors (B). When colocalizingfor both CD31-positive endothelial cellsand apoptotic cells, more endothelialcells were found to undergo apoptosisin the 100-mg/kg treatment arm of theexperiment than in the control, vehicle,and the 50-mg/kg treatment arms. C,quantitative analysis of the immunohis-tochemistry data. i, the number ofTUNEL+ cells per unit area in the controland mice receiving the vehicle were15.83 and 23.75, respectively, com-pared with 34.06 in the group given 50mg/kg QLT0267 and 73.51 in the groupgiven 100 mg/kg QLT0267 (P < 0.001).ii, MVD of the tumor was decreased from7.78 in the control tumors and 7.53 inthe vehicle tumors to 5.68 in the 50 mg/kg KP74728-treated tumors (P < 0.05)and 4.78 in the 100 mg/kg KP74728-treated tumors (P <0.001). iii, numbersof tumor-associated endothelial cells un-dergoing apoptosis were elevated in the100 mg/kg treatment arm (73.51; P <0.01) compared with the control andvehicle groups (15.83 and 23.75,respectively).





In summary, our findings shed some light on the biologyof anaplastic thyroid cancer, which is rare and nearlyalways fatal, and on the potential role that ILK might playin mediating tumorigenesis by regulating tumor cellgrowth and angiogenesis. The use of the novel ILKinhibitor QLT0267 on the tested cell lines was associatedwith cell growth arrest, induction of apoptosis in vitro ,and reduction of tumor volume in vivo.

References

1. Wartofsky L. Thyroid cancer: a comprehensive guide to clinicalmanagement, p. xx. Totowa (NJ): Humana Press; 2000. p. 515.

2. Sherman SI. Thyroid carcinoma. Lancet 2003;361:501–11.

3. Pierie JP, Muzikansky A, Gaz RD, Faquin WC, Ott MJ. The effect ofsurgery and radiotherapy on outcome of anaplastic thyroid carcinoma. AnnSurg Oncol 2002;9:57–64.

4. Xu G, Pan J, Martin C, Yeung SC. Angiogenesis inhibition in the in vivoantineoplastic effect of manumycin and paclitaxel against anaplasticthyroid carcinoma. J Clin Endocrinol Metab 2001;86:1769–77.

5. McIver B, Hay ID, Giuffrida DF, et al. Anaplastic thyroid carcinoma: a50-year experience at a single institution. Surgery 2001;130:1028–34.

6. Haigh PI, Ituarte PH, Wu HS, et al. Completely resected anaplasticthyroid carcinoma combined with adjuvant chemotherapy and irradiation isassociated with prolonged survival. Cancer 2001;91:2335–42.

7. Passler C, Scheuba C, Prager G, et al. Anaplastic (undifferentiated)thyroid carcinoma (ATC). A retrospective analysis. Langenbecks ArchSurg 1999;384:284–93.

8. Hundahl SA, Fleming ID, Fremgen AM, Menck HR. A National CancerData Base report on 53,856 cases of thyroid carcinoma treated in the U.S.,1985–1995 [see comments]. Cancer 1998;83:2638–48.

9. Gilliland FD, Hunt WC, Morris DM, Key CR. Prognostic factors forthyroid carcinoma. A population-based study of 15,698 cases from theSurveillance, Epidemiology and End Results (SEER) program 1973–1991.Cancer 1997;79:564–73.

10. Takashima S, Morimoto S, Ikezoe J, et al. CT evaluation of anaplasticthyroid carcinoma. AJR Am J Roentgenol 1990;154:1079–85.

11. Sugitani I, Kasai N, Fujimoto Y, Yanagisawa A. Prognostic factors andtherapeutic strategy for anaplastic carcinoma of the thyroid. World J Surg2001;25:617–22.

12. Teppo L, Hakulinen T. Variation in survival of adult patients withthyroid cancer in Europe. EUROCARE Working Group. Eur J Cancer 1998;34:2248–52.

13. Dinneen SF, Valimaki MJ, Bergstralh EJ, Goellner JR, Gorman CA,Hay ID. Distant metastases in papillary thyroid carcinoma: 100 casesobserved at one institution during 5 decades. J Clin Endocrinol Metab1995;80:2041–5.

14. Hannigan GE, Leung-Hagesteijn C, Fitz-Gibbon L, et al. Regulation ofcell adhesion and anchorage-dependent growth by a new h 1-integrin-linked protein kinase. Nature 1996;379:91–6.

15. Giancotti FG, Ruoslahti E. Integrin signaling. Science 1999;285:1028–32.

16. Tu Y, Li F, Goicoechea S, Wu C. The LIM-only protein PINCH directlyinteracts with integrin-linked kinase and is recruited to integrin-rich sites inspreading cells. Mol Cell Biol 1999;19:2425–34.

17. Hobert O, Moerman DG, Clark KA, Beckerle MC, Ruvkun G. Aconserved LIM protein that affects muscular adherens junction integrityand mechanosensory function in Caenorhabditis elegans. J Cell Biol1999;144:45–57.

18. Li F, Zhang Y, Wu C. Integrin-linked kinase is localized to cell-matrixfocal adhesions but not cell-cell adhesion sites and the focal adhesionlocalization of integrin-linked kinase is regulated by the PINCH-bindingANK repeats. J Cell Sci 1999;112:4589–99.

19. Radeva G, Petrocelli T, Behrend E, et al. Overexpression of theintegrin-linked kinase promotes anchorage-independent cell cycle progres-sion. J Biol Chem 1997;272:13937–44.

20. Attwell S, Roskelley C, Dedhar S. The integrin-linked kinase (ILK)suppresses anoikis. Oncogene 2000;19:3811–5.

21. Troussard AA, Costello P, Yoganathan TN, Kumagai S, Roskelley CD,Dedhar S. The integrin linked kinase (ILK) induces an invasive phenotypevia AP-1 transcription factor-dependent upregulation of matrix metal-loproteinase 9 (MMP-9). Oncogene 2000;19:5444–52.

22. Wu C, Keightley SY, Leung-Hagesteijn C, et al. Integrin-linked proteinkinase regulates fibronectin matrix assembly, E-cadherin expression, andtumorigenicity. J Biol Chem 1998;273:528–36.

23. Marotta A, Parhar K, Owen D, Dedhar S, Salh B. Characterisation ofintegrin-linked kinase signalling in sporadic human colon cancer. Br JCancer 2003;88:1755–62.

24. Graff JR, Deddens JA, Konicek BW, et al. Integrin-linked kinaseexpression increases with prostate tumor grade. Clin Cancer Res 2001;7:1987–91.

25. Ito R, Oue N, Zhu X, et al. Expression of integrin-linked kinase isclosely correlated with invasion and metastasis of gastric carcinoma.Virchows Arch 2003;442:118–23.

26. Persad S, Attwell S, Gray V, et al. Inhibition of integrin-linked kinase(ILK) suppresses activation of protein kinase B/Akt and induces cell cyclearrest and apoptosis of PTEN-mutant prostate cancer cells. Proc Natl AcadSci U S A 2000;97:3207–12.

27. Delcommenne M, Tan C, Gray V, Rue L, Woodgett J, Dedhar S.Phosphoinositide-3-OH kinase-dependent regulation of glycogen synthasekinase 3 and protein kinase B/AKT by the integrin-linked kinase. Proc NatlAcad Sci U S A 1998;95:11211–6.

28. Troussard AA, Mawji NM, Ong C, Mui A, St -Arnaud R, Dedhar S.Conditional knock-out of integrin-linked kinase demonstrates an essen-tial role in protein kinase B/Akt activation. J Biol Chem 2003;278:22374–8.

29. Persad S, Dedhar S. The role of integrin-linked kinase (ILK) in cancerprogression. Cancer Metastasis Rev 2003;22:375–84.

30. Tan C, Cruet-Hennequart S, Troussard A, et al. Regulation of tumorangiogenesis by integrin-linked kinase (ILK). Cancer Cell 2004;5:79–90.

31. Ahmed N, Riley C, Oliva K, Stutt E, Rice GE, Quinn MA. Integrin-linked kinase expression increases with ovarian tumour grade and issustained by peritoneal tumour fluid. J Pathol 2003;201:229–37.

32. Xu J, Moatamed F, Caldwell JS, et al. Enhanced expression ofnicotinamide N -methyltransferase in human papillary thyroid carcinomacells. J Clin Endocrinol Metab 2003;88:4990–6.

33. Ain KB, Tofiq S, Taylor KD. Antineoplastic activity of taxol againsthuman anaplastic thyroid carcinoma cell lines in vitro and in vivo . J ClinEndocrinol Metab 1996;81:3650–3.

34. Yoganathan N, Yee A, Zhang Z, et al. Integrin-linked kinase, apromising cancer therapeutic target: biochemical and biological properties.Pharmacol Ther 2002;93:233–42.

35. Rak J, Filmus J, Kerbel RS. Reciprocal paracrine interactions betweentumour cells and endothelial cells: the ‘angiogenesis progression’hypothesis. Eur J Cancer 1996;32A:2438–50.

36. Yau CY, Wheeler JJ, Sutton KL, Hedley DW. Inhibition ofintegrin-linked kinase by a selective small molecule inhibitor,QLT0254, inhibits the PI3K/PKB/mTOR, Stat3, and FKHR pathwaysand tumor growth, and enhances gemcitabine-induced apoptosis inhuman orthotopic primary pancreatic cancer xenografts. Cancer Res2005;65:1497–504.





2005;4:1146-1156. Mol Cancer Ther Maher N. Younes, Seungwon Kim, Orhan G. Yigitbasi, et al. anaplastic thyroid cancerIntegrin-linked kinase is a potential therapeutic target for

Updated version

http://mct.aacrjournals.org/content/4/8/1146

Access the most recent version of this article at:

Cited articles

http://mct.aacrjournals.org/content/4/8/1146.full#ref-list-1

This article cites 34 articles, 11 of which you can access for free at:

Citing articles

http://mct.aacrjournals.org/content/4/8/1146.full#related-urls

This article has been cited by 13 HighWire-hosted articles. Access the articles at:

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

Subscriptions

Reprints and

[email protected] at

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

Permissions

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

.http://mct.aacrjournals.org/content/4/8/1146To request permission to re-use all or part of this article, use this link



http://mct.aacrjournals.org/content/4/8/1146.full#ref-list-1

http://mct.aacrjournals.org/content/4/8/1146.full#related-urls

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

mailto:[email protected]



integrin-linked kinase is a potential therapeutic target ... · anaplastic thyroid cancer and ilk...

Documents