heat shock protein 90-sheltered overexpression of insulin-like … · the hitherto disappointing...

Cancer Therapy: Preclinical

Heat Shock Protein 90-Sheltered Overexpression of Insulin-LikeGrowth Factor 1 Receptor Contributes to Malignancy ofThymic Epithelial Tumors

Marco Breinig1, Philipp Mayer1, Andreas Harjung1, Benjamin Goeppert1, Mona Malz1, Roland Penzel1,Olaf Neumann1, Arndt Hartmann2, Hendrik Dienemann3, Giuseppe Giaccone4, Peter Schirmacher1,Michael Andr�e Kern1, Gabriela Chiosis5, and Ralf Joachim Rieker1,2

AbstractPurpose: The underlying molecular mechanisms of thymic epithelial malignancies (TEMs) are poorly

understood. Consequently, there is a lack of efficacious targeted therapies and patient prognosis remains

dismal, particularly for advanced TEMs. We sought to investigate protumorigenic mechanism relevant to

this understudied cancer.

Experimental Design: Recently established cell lines derived from thymic epithelial tumors were used

as a model system. The antitumor activity of specific heat shock protein 90 (Hsp90) inhibitors was

investigated by an analysis of cell viability, cell cycle, and apoptosis using MTT-assays and flow cytometry.

Western blotting was used to investigate the altered expression of Hsp90 clients. Pharmacological

inhibitors against select Hsp90 clients, as well as RNAi, were employed to test the relevance of each

client independently. Tissue microarray analysis was performed to match the in vitro findings with

observations obtained from patient-derived samples.

Results:Hsp90 inhibition significantly reduces cell viability of thymic carcinoma cells, induces cell cycle

arrest and apoptosis, and blocks invasiveness. Hsp90 inhibition triggers the degradation of multiple

oncogenic clients, for example insulin-like growth factor 1 receptor (IGF-1R), CDK4, and the inactivation

of PI3K/Akt and RAF/Erk signaling. Mechanistically, the IGF/IGF-1R–signaling axis contributes to the

establishment of the antiapoptotic phenotype of thymic cancer cells. Finally, IGF-1R is overexpressed in

advanced TEMs.

Conclusions: We have unraveled a novel protumorigenic mechanism in TEMs, namely Hsp90-

capacitated overexpression of IGF-1R, which confers apoptosis evasion in malignant thymic epithelial

cells. Our data indicate that Hsp90 inhibition, which simultaneously blocks multiple cancer hallmarks,

represents a therapeutic strategy in TEMs that may merit evaluation in clinical trials. Clin Cancer Res; 17(8);

2237–49. �2011 AACR.

Introduction

Thymic epithelial malignancies (TEMs) show a broadspectrum of clinical and histologic characteristics. Accord-ing to WHO, TEMs are principally classified as type A, AB,B1, B2, and B3 thymoma. Finally, various subtypes ofthymic carcinomas (formerly classified as type C thymo-mas) exist. Overall, type B3 thymomas and thymic carcino-mas show a particularly unfavorable outcome (1, 2).Advanced disease is often irresectable and response ratesto conventional chemotherapy are lower than 50% (3).Because of an inadequate knowledge regarding the onco-genic mechanisms of TEMs, very few targeted therapyapproaches have been discussed, for example inhibitionof the epidermal growth factor receptor (EGFR), or KIT.However, clinical trials have revealed that treatment withcetuximab or gefitinib (both targeting EGFR), as well asimatinib (targetingKIT) showonlyminimalactivity inTEMs(3). Consequently, new therapeutic strategies are needed.

Authors' Affiliations: 1Department of General Pathology, University Hos-pital Heidelberg, Heidelberg, 2Department of Pathology, University Hos-pital Erlangen, Erlangen, and 3Department of Thoracic Surgery,Thoraxklinik am Universit€atsklinikum Heidelberg, Heidelberg, Germany;4Medical Oncology Branch, National Cancer Institute, Bethesda, Mary-land; and 5Program in Molecular Pharmacology and Chemistry, Depart-ment of Medicine, Memorial Sloan-Kettering Cancer Center, New York,New York

Note: Supplementary data for this article are available at Clinical CancerResearch Online (http://clincancerres.aacrjournals.org/).

Current address for M. Breinig: German Cancer Research Center (DKFZ),Division of Signaling and Functional Genomics, Im Neuenheimer Feld 580,D-69120 Heidelberg, Germany. E-mail: [email protected]

Corresponding Author: Marco Breinig, Institute of Pathology, UniversityHospital Heidelberg, 69120 Heidelberg, Germany. Phone: þ49-6221-56-5409; Fax: þ49-6221-56-5251; E-mail: [email protected]

doi: 10.1158/1078-0432.CCR-10-1689

�2011 American Association for Cancer Research.

ClinicalCancer

Research

www.aacrjournals.org 2237

on March 12, 2020. © 2011 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst March 3, 2011; DOI: 10.1158/1078-0432.CCR-10-1689

http://clincancerres.aacrjournals.org/

The hitherto disappointing results related to targetedtherapies in this cancer might reflect the lack of represen-tative model systems for preclinical studies, as well as thenecessity to simultaneously block diverse oncogenicmechanisms to achieve potent antineoplastic activity inmost tumors (4). Regarding the first aspect, adequatemouse models that recapitulate this cancer are still missingbut advancements have been made by establishing stablecell lines derived from TEMs (5, 6). Regarding the secondaspect, preliminary studies have shown beneficial effects ofthe multikinase inhibitors sorafenib and sunitinib in TEMs(7). Besides the use of nonspecific multikinase inhibitors,specific mono-targeted inhibition of the molecular chaper-one Hsp90 represents a compelling strategy that allows asynchronized blockade of multiple malignancy drivingmechanisms (8).

Hsp90 generally regulates the stabilization and activa-tion of so-called "client" proteins, numerous of which arebona fide oncoproteins (9). Hence, Hsp90 controls signal-ing pathways involved in the establishment of all hallmarksof cancer (10). Hsp90 inhibition induces the simultaneousproteasomal degradation of various onco-clients. The ther-apeutic efficacy of Hsp90 inhibitors most likely relates tothe concurrent shutdown of multiple protumorigenic cir-cuitries, hence preventing possible "escape"mechanisms oftumor cells by activating compensatory signaling pathways(10). Reports from clinical trials with the geldanamycinclass of Hsp90 inhibitors (e.g., 17-AAG) demonstrate ben-eficial effects in several malignancies. However, this com-pound series has several limitations and an unfavorabletoxicity profile (11, 12). With the advent of novel Hsp90inhibitors that lack geldanamycin’s benzoquinone moiety,which is deemed accountable for most of the observed side

effects, hope has been raised that these compounds willimprove the clinical applicability of Hsp90 inhibition (13).

Here, we show that the Hsp90-sheltered activity of insu-lin-like growth factor 1 receptor (IGF-1R) contributes to theantiapoptotic phenotype of thymic epithelial cancer cells.We demonstrate that IGF-1R overexpression manifests thisnewly identified mechanism of malignancy in advancedTEMs. Hsp90 inhibition reverses IGF-1R-mediated antia-poptosis and induces multimodal antitumor activity. Ourresults suggest that Hsp90 inhibitors may represent pro-mising therapeutic agents for the treatment of TEMs.

Materials and Methods

Detailed methods are included as SupplementaryMaterial.

Reagents8-(6-Iodobenzo[d][1,3]dioxol-5-ylthio)-9-(3-(isopropy-

lamino)propyl)-9H-purin-6-amine (PU-H71) was synthe-sized as previously reported (14). 17-Allylamino-17-desmethoxygeldanamycin (17-AAG) and the EGFR inhibi-tor N-(3-chloro-4-fluoro-phenyl)-7-methoxy-6-(3-mor-pholin-4-ylpropoxy)quinazolin-4-amine (gefitinib) wereobtained from LC Laboratories. The IGF-1R inhibitor cyclo-lignan picropodophyllin (PPP), the CDK4 inhibitor 2-bromo-12,13-dihydro-5H-indolo[2,3-a]pyrrolo[3,4-c]car-bazole-5,7(6H)-dione, the Raf-1 inhibitor 5-Iodo-3-[(3,5-dibromo-4-hydroxyphenyl)methylene]-2-indolinone(Raf1 inhibitor I, GW5074), the Akt inhibitor 1,3-dihydro-1-(1-((4-(6-phenyl-1H-imidazo[4,5-g]quinoxalin-7-yl)ph-enyl)methyl)-4-piperidinyl)-2H-benzimidazol-2-one (Ak-ti-1/2, Akt inhibitor VIII), and the EGFR inhibitor 4-(3-chloroanilino)-6,7-dimethoxyquinazoline (AG1478) werepurchased from Calbiochem/Merck. The selective IGF-1Rinhibitor NVP-AEW541 (pyrrolo[2,3-d] pyrimidine deriva-tive) and the dual IGF-1R/insulin receptor (IR) inhibitorBMS-536924 (1H-benzoimidazol-2-yl)-1H-pyridin-2-one)were from Selleck Chemicals. The broad-spectrum caspaseinhibitor (Z-Val-Ala-Asp(OMe)-FMK) was from BioVision.The selective caspase-8 inhibitor (Z-IE(OMe)TD(OMe)-FMK)) and the selective caspase-9 inhibitor (Z-LE(OMe)HD(OMe)-FMK) were from Calbiochem/Merck. Thesecompounds were dissolved in DMSO and stock solutionswere stored at �20�C. Recombinant IGF-I and -II was fromR&D Systems and the neutralizing anti-IGF-1R antibody(aIR3) was from Calbiochem. These compounds weredissolved in sterile PBS and stored as stock solutions at�20�C. Purified mouse IgG was from Millipore.

Cell linesThe thymic carcinoma cell line, TC1889, and the thy-

moma cell line, T1682, were established, characterized andauthenticated as previously described (5). Cells were cul-tured in RPMI-media containing HEPES (PAA Labora-tories) supplemented with 10% fetal calf serum (FCS)and 1% penicillin/streptomycin (Sigma) in an atmospherecontaining 5% CO2. Cell lines were regularly checked for

Translational Relevance

Therapeutic approaches that improve the dismalsituation of patients suffering from advanced andaggressive thymic epithelial malignancies (TEMs) aremissing since the molecular mechanisms involved inthe establishment of this cancer are largely unknown.Here we demonstrate, by using newly established in vitromodels, that heat shock protein 90 (Hsp90) is a criticaland multimodal modifier of TEMs. Pharmacologicalinhibition of Hsp90 elicits marked antitumor activityin thymic carcinoma cells and robustly inhibits multiplecancer-relevant signaling pathways. Along this line, wemoreover show that the Hsp90 client IGF-1R is over-expressed in human thymic tumors and plays a role inestablishing the malignant antiapoptosis phenotype ofthymic cancer cells. Collectively, these findings for thefirst time uncover molecular mechanisms that are vitalto TEMs and that represent intervention points fortargeted therapies. Our studies suggest that targetingHsp90 or its client IGF-1R may warrant evaluation inclinical trials.

Breinig et al.

Clin Cancer Res; 17(8) April 15, 2011 Clinical Cancer Research2238




Mycoplasma infection using the VenorGEM MycoplasmaDetection Kit (Minerva Biolabs) according to the man-ufacturer’s instructions.

Cell viability measurementThe number of viable cells was determined using the

MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl-tetrazo-lium-bromide)-assay (EZ4U; Biomedica), as previouslydescribed (5).

Cell cycle analysis and detection of hypodiploidyDNA content was analyzed by flow cytometry of propi-

dium iodide–stained nuclei using a FACSCalibur flowcyt-ometer (Becton Dickinson) as previously described (5).Data were analyzed using CellQuest and ModFit LT 3.00software.

Western immunoblottingPreparation of total protein lysates from human thymo-

mas, thymic carcinomas, and normal thymi was performedas previously described (15). Preparation of total proteinlysates from cell cultures as well as subsequent SDS-PAGEandWestern immunoblotting was performed as previouslydescribed (16). See Supplementary Material for the anti-bodies employed.

Tissue microarrayThe tissue microarray (TMA) was designed as previously

described (15). Matching clinical data of the tissue arraysubjects including Masaoka stage (17) are given in Table 1.

ImmunohistochemistryParaffin sections, each 4 mm thick, were cut, de-paraffi-

nized in xylene and rehydrated with graded ethanol.IGF-1R staining. 1 mM EDTA buffer (pH 9) was added

to the slides and samples were heated by incubation in amicrowave oven at 600 W for 3� 5 min. After cooling, theslides were treated with hydrogen peroxidase in PBS toblock the endogenous peroxidase. Nonspecific antibodybinding was blocked using the avidin/biotin blocking Kitaccording to the manufacturer’s instructions (Dako Cyto-mation). Anti–IGF-1RO antibody (1:50, sc-713, SantaCruz) was applied and incubated in a moist chamber at4�C overnight. Signal detection was performed with bio-tinylated secondary antibody in a 1:200 dilution (DakoCytomation) followed by a streptavidin–biotin complexand visualized with AEC-solution (Dako Cytomation)according to the manufacturer’s instructions.IGF-I staining. Slides were subjected to 5 min heating

at 120�C in Tris-EDTA-Buffer (pH 8.5) in a pressure cooker.Endogenous peroxidase activity was quenched by 5 minincubation with Peroxidase-Blocking solution (DakoREAL, Dako Cytomation). Anti–IGF-I antibody (1:50, sc-7144, Santa Cruz) was applied for 30 min. Slides weresubsequently incubated with the HRP-labeled Polymeranti-goat Histofine (414161 F, Nichirei Cooperation) for30 min and visualized using DAB solution (Dako Cytoma-tion) according to the manufacturer’s instructions.

Generally, nuclear counterstaining was performed withMayer’s hemalum. The sections weremounted with Kaiser’sglycerine gelatine. Adequate negative controls wereobtained by replacing the primary antibody with antibodydilution buffer (Dako Cytomation). Nonepithelial com-partments of thymi were excluded from the analyses andstaining intensity was scored for each specimen on a scaleof 0 to 3, in which 0 represents negative-, 1 low-, 2medium-, and 3 high-staining intensity (Fig. 5) by 2independent investigators (B.G. and R.R.; interobserveragreement >0.7). In the cases of divergent scoring, a thirdobserver (M.A.K.) decided upon the final category.

Statistical analysisStatistical analyses of in vitro studies were performed with

the 2-sided Student’s t-test. Statistically significant changesare given as P values: *P < 0.05, **P � 0.01 and ***P �0.001. For TMA analyses, the differences in the frequencydistribution of the intensity were analyzed using the Krus-kal–Wallis test, with respect to stage and histologic subtype.The Jonckheere–Terpstra test was employed to analyzetrends. A difference was considered statistically significantif the P value of the corresponding statistical test was lessthan 5% or equal (P � 0.05). Kappa statistics were used toassess the degree of agreement between the ratings of the 2observers. All statistical analyses were performed using thestatistical software system SAS (SAS Institute).

Results

Hsp90 inhibition induces multimodalantitumorigenic activity in thymic epithelial tumorcells

To start, the effect of the nonquinone Hsp90 inhibitorPU-H71 on cell viability of a thymic carcinoma cell line(TC1889) was investigated using MTT-assays. Hsp90 inhi-bition significantly reduced cell viability in a concentra-tion- and time-dependent manner (Fig. 1A). After 72hours, the concentration that achieved half maximalreduction of cell viability (IC50) was �0.25 mM. Treat-ment with PU-H71 induced G2/M arrest (Fig. 1B),increased aneuploidy (data not shown) and resulted incell death, as evidenced by increased hypodiploidy (>30%of cells in subG1; Fig. 1C). Cell death was attributable toapoptosis since PU-H71 treatment increased caspase-3activity (data not shown) and induced a time-dependentcleavage of PARP (Fig. 1D). Along these lines, broad-spectrum caspase inhibition significantly reduced PU-H71 treatment-induced caspase-3 activity (data notshown) and hypodiploidy (Fig. 1E). Essentially the sameresults were obtained with the chemically distinctive Hsp90inhibitor 17-AAG (Supplementary Fig. S1A–E), hence sug-gesting a common antitumorigenic activity of Hsp90 inhi-bition in TC1889 cells. Related to Hsp90 inhibition-induced apoptosis, we observed that inhibition of cas-pase-8 and -9 reduced hypodiploidy associated with PU-H71 treatment (Fig. 1E). In addition, PU-H71 treatmentinduced mitochondrial depolarization (Fig. 1F), hence

Relevance of Hsp90 and IGF-1R in Thymic Carcinomas

www.aacrjournals.org Clin Cancer Res; 17(8) April 15, 2011 2239




suggesting that Hsp90 inhibition unleashes intrinsic apop-tosis in TC1889 cells.

Given that Hsp90 inhibition interferes with multiplemechanisms of malignancy (10), we investigated whetherthis strategy also blocks the cancer hallmark of tissueinvasion. PU-H71 treatment inhibited the invasive capacityof TC1889 cells in a time- and concentration-dependentmanner. A concentration of 1 mM was sufficient to reduceinvasion through Matrigel up to �75% after 72 hours ascompared to controls (Fig. 1G).

To further delineate the molecular basis for the observedantitumor activity of Hsp90 inhibitors in thymic cancercells, we investigated the expression of well-establishedHsp90 clients following Hsp90 inhibition. In TC1889 cells,PU-H71 treatment reduced the expression of EGFR, IGF-1R,cyclin-dependent kinase 4 (CDK4), Akt, RAF-1, phosphory-lated Akt, and phosphorylated Erk1/2 in a concentration-and time-dependentmanner (Fig. 1H). Amarked reductionwas achieved at PU-H71 concentrations >0.1 mM, whichcorrelates well with the concentrations necessary to

Table 1. Expression patterns for IGF-1R and IGF-I and additional clinical data

Patientnumber

Age Sexa Histology(WHO)

Staging(Masaoka)

IGF-1R IGF-I

1 10 months M Normal 2 12 1 week M Normal 1 13 2 months M Normal 1 14 68 years M A 2 3 15 71 years M A 2 3 16 57 years M A 2 2 37 58 years F A 2 1 18 78 years F A 2 2 19 56 years M A 2 3 110 47 years F AB 2 1 111 74 years F AB 2 3 112 51 years M AB 1 1 113 66 years F AB 2 3 114 53 years F AB 2 2 115 59 years F AB 1 3 116 58 years F B1 1 1 217 66 years F B1 2 1 118 52 years M B1 2 1 119 41 years F B1 1 1 120 71 years M B1 1 2 121 34 years M B1 2 3 122 55 years M B2 3 3 123 43 years M B2 4 2 124 49 years F B2 2 225 76 years F B2 2 3 126 75 years F B2 3 227 58 years M B3 3 3 228 62 years F B3 2 2 129 62 years F B3 4 3 130 63 years F B3 4 3 131 73 years M B3 2 3 332 57 years F Carcinoma 4 3 133 51 years M Carcinoma 4 3 234 58 years F Carcinoma 2 3 235 66 years M Carcinoma 3 3 136 76 years F Carcinoma 3 3 237 2.5 months M Normal 1 138 9 months M Normal 1 1

NOTE: Representative expression intensities for IGF-1R and IGF-I are depicted in Figure 5A and B, respectively.aSex: M, male; F, female.

Breinig et al.





Figure 1.Hsp90 inhibition inducesmultimodal antitumor activity in thymic epithelial tumor cells. A, cell viability reduction. TC1889 cells were incubatedwith PU-H71 and cell viability was analyzed using MTT-assays. Assays were performed in sextuple. Data are expressed as the mean � SD (n � 3). B, inductionof cell cycle arrest. TC1889 cells were incubated with PU-H71 and cell cycle distribution was assessed using FACS-analyses following PI-staining. Results ofcells in G2/M are shown. Data are expressed as the mean � SD (n ¼ 3). C, induction of cell death. TC1889 cells were incubated with PU-H71 and cell death(hypodiploid cells in subG1) was assessed using FACS-analyses following PI-staining. Data are expressed as themean� SD (n¼ 3). D, induction of apoptosis.TC1889cellswere incubatedwithPU-H71 (1mM)andPARPcleavagewas investigatedusingWestern immunoblotting. Representative results are shown (n�3).Actin served as a loading control. fl., full length; cl., cleaved. E, caspase-dependence of apoptosis. TC1889 cells were incubated with PU-H71 and abroad caspase inhibitor (Caspi), a caspase-8 inhibitor (Casp8i) and a caspase-9 inhibitor (Casp9i) and cell death was assessed after 72 hours using FACS-analyses following PI-staining. Data are expressed as themean� SD (n¼ 3). F, mitochondrial depolarization. TC1889 cells were incubatedwith PU-H71 (1 mM)and mitochondrial membrane potential was analyzed after 48 hours using FACS analysis following JC-1 staining. The uncoupler CCCP (100 mM) was usedas a positive control inducing amarked shift in JC-1 fluorescence as compared to untreated controls (CO). Assays were performed in duplicate. Representativeresults are shown (n ¼ 3). G, inhibition of invasiveness. TC1889 cells were pretreated for 24 hours with vehicle or PU-H71. Thereafter, viable cells able tomigrate throughMatrigel over a 48- and 72-hour periodwere quantified by phase contrastmicroscopy and data graphed. Data are expressed as themean�SD(n ¼ 3). H, simultaneous degradation and inactivation of potential oncoproteins. (Left) Concentration dependence. TC1889 cells were treated with vehicleor PU-H71 and the expression of the indicated proteins was investigated after 24 hours using Western immunoblotting. (Right) Time dependence. TC1889cells were treated with PU-H71 (1 mM) and the expression of the indicated proteins was investigated using Western immunoblotting. Representativeresults are shown (n � 3). Actin served as a loading control.






induce growth inhibition and apoptosis. In accordancewithprevious findings (16), expression of the inducible chaper-one Hsp70 was increased. Essentially the same results wereobtained with 17-AAG (Supplementary Fig. S1F), hencesuggesting a common mechanism of action of Hsp90 inhi-bition. Similarly, PU-H71 treatment significantly reducedcell viability in the previously established thymoma cell lineT1682, induced cell death as evidenced by increased hypo-diploidy, and resulted in the simultaneous degradation andinactivation of aforementioned clients (SupplementaryFig. S2).

Collectively, Hsp90 inhibition exerts multimodal anti-tumorigenic effects and simultaneously affects the expres-sion and activity of multiple potential oncoproteins inthymic epithelial tumor cells.

The Hsp90 client IGF-1R represents a noveltherapeutic target that contributes to apoptosisevasion in thymic cancer cells

Given the lack of insight into molecular mechanismrelevant to TEMs, we attempted to dissect the contributionof each of the investigated Hsp90 clients to the establish-ment of the malignant phenotype of TC1889 cells.

We first employed pharmacological inhibitors to inde-pendently block the function of the clients EGFR, IGF-1R,CDK4, RAF-1, and Akt (Supplementary Table S1). Tocorroborate our pharmacologically based strategy and totake liabilities like off-target effects of the inhibitors intoconsideration, we further employed RNA interference(RNAi) to silence the highly druggable and hence clinicallyrelevant clients IGF-1R, EGFR, and CDK4 (18–20) inde-pendently.

The different inhibitors employed reduced cell viabilityin a concentration- and time-dependent manner with theIGF-1R inhibitor PPP being the most potent compound(IC50: �0.25 mM; Fig. 2A). Regarding EGFR inhibition, wealso tested iressa/gefitinib, a drug previously tested inclinical trials with patients suffering from TEMs (3) andfound that none of the gefitinib concentrations employed(up to 50 mM) were sufficient to achieve a half maximalreduction of cell viability (Supplementary Fig. S3). Becausethe efficacy of EGFR inhibitors can be associated withspecific EGFR mutations (21), we tested whether thesemutations can be found in TC1889 cells. We did not detectmutations in exons 18 to 21 of the EGFR gene (data notshown). In support of the results obtained by our pharma-cological approach, RNAi experiments revealed that sub-stantial siRNA-mediated depletion of the clients analyzedwas associated with a significant reduction in cell viability.Again, effects were strongest for IGF-1R knockdown with a�40% decrease in viability for both siRNAs after 120 hours(Fig. 2B).

Given that Hsp90 inhibition induced apoptosis in thy-mic carcinoma cells, we next investigated the contributionof Hsp90 clients to the cancer hallmark of apoptosis eva-sion. Therefore, we examined the apoptosis phenotype ofTC1889 cells after treatment with inhibitors for each clientat low and high concentrations. Among the compounds

examined, only the IGF-1R inhibitor PPP potently inducedapoptosis at low concentrations as evidenced by increasedhypodiploidy (>25% of cells in subG1) and by PARPcleavage (Fig. 2C and D). In addition, IGF-1R inhibitionwith PPP induced cell death in a thymoma cell line, asindicated by increased hypodiploidy and cell viabilityreduction (Supplementary Fig. S4A and B).

To strengthenour results that suggested an important roleof the Hsp90 client IGF-1R and its potential to represent avalid therapeutic target in TEMs, we additionally analyzedthe effects of a neutralizing IGF-1R antibody (aIR3) and ofthe IGF-1R inhibitors NVP-AEW541 (NVP) and BMS-536924 (BMS). Notably, all of these strategies to interferewith IGF-1R signaling reduced TC1889 cell viability andsignificantly induced apoptosis (Fig. 3A and B).

Related to IGF-1R inhibition-induced apoptosis, we alsoobserved that inhibition of caspase-3, -8 and -9 reducedhypodiploidy associated with PPP treatment (Fig. 3C). Inaddition, PPP treatment induced mitochondrial depolar-ization (Fig. 3D), suggesting that, similar to Hsp90 inhibi-tion, IGF-1R inhibition also unleashes intrinsic apoptosisin TC1889 cells.

Together, these results suggest that among the Hsp90clients investigated, IGF-1R contributes to the malignantphenotype of thymic epithelial tumor cells by its antia-poptotic properties. Given its druggability (22), IGF-1Rmight represent a potential therapeutic target in this cancer.

The IGF/IGF-1R signaling axis contributes to thymiccancer cell survival

To further investigate the relevance of IGF-1R-signalingin TEMs, we went on to analyze whether stimulation withIGF-I and -II affects cell viability. In TC1889 cells, IGF-I aswell as IGF-II administration increased cell viability ascompared to unstimulated controls. This effect was evi-dent in cells grown in FCS-supplemented media but wasconsiderably robust under serum-starvation conditions(Fig. 4A). Correspondingly, IGF-I and -II administrationsignificantly reduced cell death as a consequence ofserum-starvation (Fig. 4B), hence further suggesting acontribution of IGF-signaling to the antiapoptoticphenotype.

A subsequent analyses of signal transduction pathwaysrevealed that, under serum-starvation conditions, IGF-Iadministration induced the activation of Akt, GSK3b,MEK1/2, and Erk as evidenced by respective phosphoryla-tions of conserved residues (Fig. 4C). IGF-1R inhibitionwith PPP, aIR3, BMS, as well as NVP reduced the IGF-I–induced activation of PI3K/Akt and MAPK/Erk signaling.However, inhibitory effects, particularly on Akt phosphor-ylation (Fig. 4C), were more marked with NVP and BMSwhich are both known to also inhibit the insulin receptor(22). Essentially the same results were obtained for IGF-Iand -II stimulation under normal growth conditions with10% FCS (Supplementary Fig. S5 and data not shown).

Surprisingly, in unstimulated nonstarved TC1889 cells,PPP treatment for longer time periods (24–72 hours)resulted in an increase in the activity ofMAPK/Erk signaling

Breinig et al.





Figure 2. Hsp90 clients contribute to the establishment of the malignant phenotype of thymic carcinoma. A, cell viability reduction. TC1889 cells were treatedwith pharmacological inhibitors against the Hsp90 clients indicated and cell viability was assessed using MTT-assays. Assays were performed in sextuple.Data are expressed as the mean � SD (n � 3). B, RNAi. TC1889 cells were transfected with the indicated siRNAs against the select Hsp90 clients andthe expression of respective proteins was investigated using Western immunoblotting. Representative results are shown (n ¼ 3). Actin served as a loadingcontrol.Cell viability following transfectionwith respective siRNAswasmeasuredusingMTT-assays.Assayswereperformed in sextuple anddata are expressedas the mean � SD (n ¼ 3). C, induction of cell death. TC1889 cells were treated with pharmacological inhibitors against select Hsp90 clients and cell deathwas evaluated by FACS-analyses following PI-staining. Data are expressed as the mean � SD (n ¼ 3). D, induction of apoptosis. TC1889 cells were treatedwith vehicle or the indicated pharmacological inhibitors against select Hsp90 clients and PARP cleavage was analyzed by Western immunoblotting.Representative results are shown (n ¼ 3). Actin served as a loading control. fl., full length; cl., cleaved.






as indicated by augmented expression of phosphorylatedErk and MEK1/2, whereas Akt phosphorylation was onlyslightly decreased (Fig. 4D). Essentially the same results,although less pronounced, were obtained by inhibition ofIGF-1R with the neutralizing antibody aIR3. Treatmentwith the IGF-1R/IR inhibitors NVP and BMS both induceda marked decrease in phsophorylated Akt, which corre-sponded to their aforementioned strong inhibitory effecton IGF-stimulated Akt signaling. Effects on MAPK/Erksignaling with these compounds were however heteroge-nous, with BMS rather decreasing Erk phosphorylation,

whereas NVP activated Erk (Fig. 4D). Corresponding find-ings were obtained for 48 hours compound treatment inserum-starved cells (Supplementary Fig. S6). RegardingIGF-1R inhibition with PPP, an activation of MAPK/Erksignaling was also detected following treatment with thisdrug in thymoma cells (Supplementary Fig. S4). Besides,we generally noticed that, in TC1889 cells, mono-targetedpharmacological inhibition of the Hsp90 clients EGFR,RAF-1, CDK4, and Akt resulted in differential activationof MAPK/Erk and/or PI3K/Akt signaling after 48 and 72hours (Supplementary Fig. S7).

Figure 3. IGF-1R contributes to thymic cancer cell survival. A, cell viability reduction. TC1889 cells were treated with pharmacological inhibitors against IGF-1R(NVP and BMS), as well as a neutralizing IGF-1R antibody (aIR3) and cell viability was assessed using MTT-assays. Mouse IgG1 antibody was employedas a reference control for the effects of aIR3 at respective concentrations. Assays were performed in sextuple. Data are expressed as themean� SD (n� 3). B,induction of cell death. TC1889 cells were treated with pharmacological inhibitors against IGF-1R as well as a neutralizing IGF-1R antibody and cell deathwas evaluated by FACS-analyses following PI-staining. Data are expressed as the mean � SD (n ¼ 3). C, caspase-dependence of apoptosis. TC1889cells were incubated with the IGF-1R inhibitor PPP and a broad caspase inhibitor, a caspase-8 and a caspase-9 inhibitor and apoptosis was assessed after 72hours using FACS-analyses following PI-staining. Data are expressed as the mean � SD (n ¼ 3). (D) Mitochondrial depolarization. TC1889 cells wereincubated with PPP (1 mM) and mitochondrial membrane potential was analyzed after 48 hours using FACS analysis following JC-1 staining. Theuncoupler CCCP (100 mM) was used as a positive control. Assays were performed in duplicate. Representative results are shown (n ¼ 3).

Breinig et al.





Figure 4. Relevance of the IGF/IGF-1R signaling axis in thymic carcinoma cells. A, cell viability increase. TC1889 cells were treated with IGF-I or -II undernormal growth conditions (10% FCS) as well as serum-starvation conditions (0% FCS) and cell viability was assessed using MTT-assays. Assays wereperformed in sextuple. Data are expressed as themean� SD (n� 3). B, apoptosis rescue. TC1889 cells were treated with vehicle, IGF-I or -II (100 ng/mL) for 72hours under serum-starvation conditions. Cell death was evaluated by FACS-analyses following PI-staining. Data are expressed as the mean � SD (n ¼ 2).C, TC1889 cells were serum-starved for 16 hours, treated with vehicle or the IGF-1R inhibitor PPP, NVP, BMS, or the neutralizing IGF1R antibody aIR3 for 2hours, and then stimulated with IGF-I (100 ng/mL) for 15 min. The activity of PI3K/Akt- and MAPK/Erk-signaling was investigated by an analysis of theexpression of phosphorylated Akt, GSK3b, MEK1/2, and Erk using Western immunoblotting. Representative results are shown (n ¼ 3). Actin served as aloading control. D, TC1889 cells were treated with vehicle or PPP, NVP, BMS, or aIR3. The activity of PI3K/Akt- and MAPK/Erk-signaling was investigated asmentioned earlier. Representative results are shown (n ¼ 3). Actin served as a loading control.






Taken together, these results confirm a growth-stimulat-ing, antiapoptotic property of the IGF-signaling axis inthymic cancer cells. Moreover, these findings demonstratethat IGF signaling is, at least partly, mediated via the PI3K/Akt and MAPK/Erk pathway. Finally, our findings revealthat the IGF-1R inhibitors employed elicit varying effectson IGF signaling.

IGF-1R is overexpressed in thymic tumorsGiven the potential relevance of IGF-1R for the malig-

nant antiapoptotic phenotype of thymic epithelial tumorcells, we finally investigated whether these functional invitro data are matched by expression data in human TEMs.Using Western immunoblot analyses, we initially foundthat IGF-1R expression in normal thymi samples washardly detectable, whereas IGF-1R was expressed in thevast majority of TEMs (Supplementary Fig. S8). To corro-borate and extend this finding and to take into account thatthe results may have been confounded by the presence oflymphocytic infiltrates, particularly in normal thymi sam-ples, we performed immunohistochemistry employing tis-sue micro arrays (TMA; Fig. 5 and Table 1). Considering

IGF-1R expression in the epithelial compartment only, alow expression intensity of IGF-1R was observed in normalthymi (Fig. 5A, arrow). Overall, a trend was observed forhigher IGF-1R expression in TEMs as compared to normalthymi (P ¼ 0.0005) and significant differences wereobserved between normal thymi and thymic tumors(P ¼ 0.006). In direct comparison to normal thymi, ahigher expression of IGF-1R was observed in type B2 (P¼ 0.032), and type B3 thymoma (P ¼ 0.008), as well asthymic carcinoma (P ¼ 0.008). Statistical analyses alsorevealed that there were significant differences regardingIGF-1R expression and Masaoka stages (P ¼ 0.044), withadvanced stages showing higher IGF-1R expression (Sup-plementary Tables S2 and S3). In addition, an analysis ofIGF-I expression (Fig. 5B and Table 1) revealed a trendtoward higher expression in thymic carcinoma (P¼ 0.012),however no statistically significant differences could becalculated when normal thymi were compared with TEMs.

Taken together, IGF-I shows a trend for higher expressionin TEMs and IGF-1R is significantly overexpressed inadvanced TEMs, hence suggesting that IGF-1R contributesto malignancy of thymic tumors.

Figure 5. The Hsp90 client IGF-R1 is overexpressed in thymic tumors. IGF-1R and IGF-I protein expression in normal thymus, different thymoma subtypes andthymic carcinomas evaluated by immunohistochemistry. Clinical data of each specimen and respective stagings are given in Table 1. A, IGF-1R: normalthymus (case no. 38, score: 1); thymomas: type A (case no. 8, score: 2), type AB (case no. 12, score: 1), type B1 (case no. 20, score: 2), type B2 (caseno. 23, score: 2), type B3 (case no. 30, score: 3); thymic carcinoma (case no. 36, score: 3). B, IGF-I: normal thymus (case no. 38, score: 1); thymomas: type A(case no. 8, score: 1), type AB (case no. 14, score: 1), type B1 (case no. 16, score: 2), type B2 (case no. 23, score: 1), type B3 (case no. 31, score: 3);thymic carcinoma (case no. 34, score: 2).

Breinig et al.





Discussion

Here, we dissected protumorigenic mechanisms relevantto TEMs and identified novel therapeutic targets. To ourknowledge, this is the first functional analysis of factorsrelevant to TEMs that therefore differentiates itself from thedescriptive nature of other studies, which are solely basedon expression analyses by immunohistochemistry or glo-bal gene expression arrays (e.g., 23–25). Importantly, weprovide evidence that the molecular chaperone Hsp90plays a vital role in the maintenance of the cancer pheno-type of thymic epithelial tumor cells. We demonstrate thatHsp90 "safeguards" the stability and activity of multiplemolecules, such as IGF-1R, CDK4, and EGFR, which wefound to differentially contribute to the survival of thymiccancer cells. In this regard, Hsp90 may rather supportmalignancy by capacitating various protumorigenic aber-rations (8), than act as a classical oncoprotein; a phenom-enon referred to as "nononcogene addiction" (26).Consequently, inhibition of Hsp90 interferes withmultiplemalignancy driving mechanisms and elicits multimodalantineoplastic activity (11). We show here that, in thymiccancer cells, this is reflected by the induction of cellcycle arrest and apoptosis, as well as the abrogation ofinvasiveness.Given the observed antitumor potency of Hsp90 inhibi-

tion, alongside its known mechanism of onco-clientdegradation, we hypothesized that characterizing the pro-tumorigenic role of specific Hsp90 clients in thymic carci-noma cells might enable us to obtain insights into thelargely unknownmechanisms of malignancy in this cancer.Using this biased approach, we show that IGF-1R contri-butes to the cancer hallmark of apoptosis evasion in thymicmalignancies. Subsequent expression analyses in humanthymic tumors moreover revealed that IGF-1R is overex-pressed particularly in highly aggressive, advanced stageTEMs. Recently, TMA analyses in a large panel of TEMs (n¼132) confirmed an overexpression of IGF-1R and foundcorrelations with WHO classification, stage, and relapse(27). Interestingly, loss of heterozygosity in the 6q23.3-25.3 region, harboring among others IGF-2R, has beenobserved to be themost frequent chromosomal aberrationsin TEMs (24). Absence of IGF-2R, which sequesters IGF,might theoretically lead to locally increased IGF concentra-tions (28). Moreover, comprehensive genomic analysesrevealed that IGF2BP3/IMP-3, a translational activator ofIGF mRNA (29), is highly overexpressed in thymic carci-nomas (25). Our initial analyses revealed that, although atrend towards higher IGF-I expression in thymic carcinomacould be observed, IGF-I was not statistically significantlyoverrepresented in TEMs when compared to normal thymi.Nevertheless, support that IGF-signaling might represent apotential oncogenic mechanism in TEMs stems from trans-genic mice overexpressing either IGF-II or the IGF-bindingprotein 4 (IGFBP4). Whereas thymic size is markedlyincreased in the former mouse models (30, 31), it isdecreased in the latter, being associated with apoptosis(32). Consequently, the IGF-signaling axis that plays a

relevant role in thymic epithelial cells under physiologicalconditions (33) might be exploited by an up-regulation ofIGF-1R in TEMs to manifest the malignant phenotype. Infact, prolonged disease stabilization was observed in apatient with metastatic thymoma in a phase 1 study ofIGF-1R monoclonal antibodies (34). Additional experi-ments are necessary to analyze the protumorigenic con-tribution of IGF-signaling in thymic tumors. Uncoveringthe effector mechanisms of IGF-1R–mediated signalingcascades that eventually contribute to apoptosis evasionin TEMs will very likely help in the identification of noveltreatment strategies. Meanwhile, a phase 2 trial with ananti–IGF-1R antibody is currently being conducted inpatients with TEMs (27).

Regarding other Hsp90 clients, our results support thenotion that CDK4 contributes to thymic carcinoma cellsurvival. Correspondingly, previous reports have shownthat inactivation of p16INK4A, which results in higherCDK4 activity (35), is involved in the progression ofTEMs (36). Regarding EGFR, which is frequently over-expressed in TEMs (23), mutations that alter its suscept-ibility towards different inhibitors are rare in this cancer(37–39). Accordingly, we observed a rather poor perfor-mance of EGFR inhibitors in TC1889 cells, which do notharbor relevant EGFR mutations (data not shown).Further analyses are compulsory in order to reveal theactual importance of CDK4 and EGFR signaling in themalignancy of TEMs.

Besides allowing an insight in protumorigenic mechan-isms of TEMs, our studies interestingly also revealed thatmono-targeted inhibition of every onco-client alone, asexemplified by IGF-1R inhibition, can be associated with aprolonged activation of vital signaling cascades, for exam-ple MAPK/Erk signaling. Off-target effects of the substancesemployed, as well as varying selectivity, for example, theIGF-1R inhibitors NVP and BMS are known to also inhibitthe insulin receptor (22), might contribute to these effects.Given that certain inhibitors elicit differential effectsdepending on the respective genetic background of cancercells (40) care has to be taken in interpreting our resultsrelated to signaling pathways modified by the substancesused. Additional studies are mandatory to comprehen-sively address this topic. Still, activation of potentiallycompensatory signaling pathways as a consequence oftargeted inhibition of one factor could generally be relatedto the plasticity and adaptability of signaling networks(41). As previously demonstrated in other malignancies,these mechanisms can contribute to the occurrence ofsecondary resistance against single agent approaches(41–43). Further analyses are warranted to delineatewhether compensatory pathway activation as a means ofacquired resistance also relates to our initial observations.So far, our attempts to develop TC1889 cells resistant toPPP have failed (data not shown).

Overall, a concurrent shutdown of multiple oncogenicmechanisms relevant to TEMs, including the parallel inhi-bition of PI3K/Akt- and MAPK/Erk-signaling has only beenobserved for Hsp90 inhibition, which was in contrast to all






other single agent approaches tested here. Consequently,acquired resistance against Hsp90 inhibitors is unlikely todevelop on the background of a dynamic switchingbetween signaling circuitries. All currently known de novoor acquired resistance mechanisms of cancer cells againstHsp90 inhibitors relate to the chemical structure of gelda-namycin analogs (44). Nonquinone Hsp90 inhibitors werenot found to share this liability (45–47). The hereinobserved multimodal antitumor activity of the purine-scaffold Hsp90 inhibitor PU-H71, which has previouslybeen reported to be well-tolerated in vivo (16, 48, 49) mayprovide a first rational for the initiation of clinical trialswith novel Hsp90 inhibitors for individuals with advancedTEMs.

In summary, our approach allowed us to obtain valuableinsights into protumorigenic mechanisms involved in aninadequately understood cancer. Strikingly, the findingsrelated to IGF-1R obtained in our in vitro model of thymiccarcinomas could be matched by observations frompatient-derived tumor samples hence underpinning thepotential value of our thymic carcinoma cell line model.Until the development of animal models with whichthymic epithelial tumorigenesis can be recapitulated, invitro studies like the one described here might assist in the

identification of potential therapeutic targets and treatmentstrategies for this peculiar cancer type.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

We thank Ms. Eva Eiteneuer and Mr. Rudolf Jung for excellent technicalassistance. We thank members of the Breuhahn lab for their valuablesupport. We are highly indebted to Dr. Josef H€ogel (University of Ulm)for his expert assistance with statistical analyses.

Grant Support

This work was supported by "Mr. William H. and Mrs. Alice Goodwinand the Commonwealth Foundation for Cancer Research" and "The Experi-mental Therapeutics Center of Memorial Sloan-Kettering Cancer Center" toG. Chiosis, by grants from the Deutsche Krebshilfe to M.A. Kern andP. Schirmacher (Ke107685), and by grants from the "Forschungsschwer-punktprogramm des Landes BadenW€urttemberg" (Kapitel. 1423;Tit.Gr 74).

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

Received June 28, 2010; revised December 1, 2010; accepted December10, 2010; published OnlineFirst March 3, 2011.

References1. Rieker RJ, Hoegel J, Morresi-Hauf A, Hofmann WJ, Blaeker H, Penzel

R, et al. Histologic classification of thymic epithelial tumors: compar-ison of established classification schemes. Int J Cancer 2002;98:900–6.

2. Strobel P, Marx A, Zettl A, Muller-Hermelink HK. Thymoma and thymiccarcinoma: an update of the WHO Classification 2004. Surg Today2005;35:805–11.

3. Rajan A, Giaccone G. Treatment of advanced thymoma and thymiccarcinoma. Curr Treat Options Oncol 2008;9:277–87.

4. Dancey JE, Chen HX. Strategies for optimizing combinations ofmolecularly targeted anticancer agents. Nat Rev Drug Discov2006;5:649–59.

5. Ehemann V, Kern MA, Breinig M, Schnabel PA, Gunawan B, SchultenHJ, et al. Establishment, characterization and drug sensitivity testingin primary cultures of human thymoma and thymic carcinoma. Int JCancer 2008;122:2719–25.

6. Inai K, Takagi K, Takimoto N, Okada H, Imamura Y, Ueda T, et al.Multiple inflammatory cytokine-productive ThyL-6 cell line estab-lished from a patient with thymic carcinoma. Cancer Sci 2008;99:1778–84.

7. Strobel P, Hohenberger P, Marx A. Thymoma and thymic carcinoma:molecular pathology and targeted therapy. J Thorac Oncol 2010;5:S286–90.

8. Whitesell L, Lindquist SL. HSP90 and the chaperoning of cancer. NatRev Cancer 2005;5:761–72.

9. Pearl LH, Prodromou C, Workman P. The Hsp90 molecular chaper-one: an open and shut case for treatment. Biochem J 2008;410:439–53.

10. Zhang H, Burrows F. Targeting multiple signal transduction pathwaysthrough inhibition of Hsp90. J Mol Med 2004;82:488–99.

11. Workman P, Burrows F, Neckers L, Rosen N. Drugging the cancerchaperone HSP90: combinatorial therapeutic exploitation of onco-gene addiction and tumor stress. Ann N Y Acad Sci 2007;1113:202–16.

12. Taldone T, Gozman A, Maharaj R, Chiosis G. Targeting Hsp90: small-molecule inhibitors and their clinical development. Curr Opin Phar-macol 2008;8:370–4.

13. Taldone T, Sun W, Chiosis G. Discovery and development ofheat shock protein 90 inhibitors. Bioorg Med Chem 2009;17:2225–35.

14. Rodina A, Vilenchik M, Moulick K, Aguirre J, Kim J, Chiang A, et al.Selective compounds define Hsp90 as a major inhibitor of apoptosisin small-cell lung cancer. Nat Chem Biol 2007;3:498–507.

15. Rieker RJ, Joos S, Mechtersheimer G, Blaeker H, Schnabel PA,Morresi-Hauf A, et al. COX-2 upregulation in thymomas and thymiccarcinomas. Int J Cancer 2006;119:2063–70.

16. Breinig M, Caldas-Lopes E, Goeppert B, Malz M, Rieker R, BergmannF, et al. Targeting heat shock protein 90 with non-quinone inhibitors: anovel chemotherapeutic approach in human hepatocellular carci-noma. Hepatology 2009;50:102–12.

17. Masaoka A, Monden Y, Nakahara K, Tanioka T. Follow-up study ofthymomas with special reference to their clinical stages. Cancer1981;48:2485–92.

18. Schwartz GK, Shah MA. Targeting the cell cycle: a new approach tocancer therapy. J Clin Oncol 2005;23:9408–21.

19. Hewish M, Chau I, Cunningham D. Insulin-like growth factor 1 recep-tor targeted therapeutics: novel compounds and novel treatmentstrategies for cancer medicine. Recent Pat Anticancer Drug Discov2009;4:54–72.

20. Zhang H, Berezov A, Wang Q, Zhang G, Drebin J, Murali R, et al. ErbBreceptors: from oncogenes to targeted cancer therapies. J Clin Invest2007;117:2051–8.

21. Sequist LV, Bell DW, Lynch TJ, Haber DA. Molecular predictors ofresponse to epidermal growth factor receptor antagonists in non-small-cell lung cancer. J Clin Oncol 2007;25:587–95.

22. Chitnis MM, Yuen JS, Protheroe AS, Pollak M, Macaulay VM. The type1 insulin-like growth factor receptor pathway. Clin Cancer Res2008;14:6364–70.

23. Henley JD, Koukoulis GK, Loehrer PJ, Sr. Epidermal growth factorreceptor expression in invasive thymoma. J Cancer Res Clin Oncol2002;128:167–70.

24. Zhou R, Zettl A, Strobel P, Wagner K, M€uller-Hermelink HK, Zhang S,et al. Thymic epithelial tumors can develop along two different patho-genetic pathways. Am J Pathol 2001;159:1853–60.

Breinig et al.





25. Girard N, Shen R, Guo T, Zakowski MF, Heguy A, Riely GJ, et al.Comprehensive genomic analysis reveals clinically relevant moleculardistinctions between thymic carcinomas and thymomas. Clin CancerRes 2009;15:6790–9.

26. Luo J, Solimini NL, Elledge SJ. Principles of cancer therapy: oncogeneand non-oncogene addiction. Cell 2009;136:823–37.

27. Zucali PA, Petrini I, Lorenzi E, Merino M, Cao L, Di Tommaso L, et al.Insulin-like growth factor-1 receptor and phosphorylated AKT-serine473 expression in 132 resected thymomas and thymic carcinomas.Cancer 2010;116:4686–95.

28. Pollak M. Insulin and insulin-like growth factor signalling in neoplasia.Nat Rev Cancer 2008;8:915–28.

29. Liao B, Hu Y, Herrick DJ, Brewer G. The RNA-binding protein IMP-3 isa translational activator of insulin-like growth factor II leader-3 mRNAduring proliferation of human K562 leukemia cells. J Biol Chem2005;280:18517–24.

30. Savino W, Cotta-de-Almeida V, van Buul-Offers SC, Koster JG,Dardenne M. Abnormal thymic microenvironment in insulin-likegrowth factor-II transgenic mice. Neuroimmunomodulation 2005;12:100–12.

31. van Buul-Offers SC, de Haan K, Reijnen-Gresnigt MG, Meinsma D,Jansen M, Oei SL, et al. Overexpression of human insulin-like growthfactor-II in transgenic mice causes increased growth of the thymus. JEndocrinol 1995;144:491–502.

32. Zhou R, Flaswinkel H, Schneider MR, Lahm H, Hoeflich A, Wanke R,et al. Insulin-like growth factor-binding protein-4 inhibits growth of thethymus in transgenic mice. J Mol Endocrinol 2004;32:349–64.

33. Chu YW, Schmitz S, Choudhury B, Telford W, Kapoor V, Garfield S,et al. Exogenous insulin-like growth factor 1 enhances thymopoiesispredominantly through thymic epithelial cell expansion. Blood2008;112:2836–46.

34. Haluska P, Shaw HM, Batzel GN, Yin D, Molina JR, Molife LR, et al.Phase I dose escalation study of the anti insulin-like growth factor-Ireceptor monoclonal antibody CP-751,871 in patients with refractorysolid tumors. Clin Cancer Res 2007;13:5834–40.

35. Malumbres M, Barbacid M. To cycle or not to cycle: a critical decisionin cancer. Nat Rev Cancer 2001;1:222–31.

36. Hirabayashi H, Fujii Y, Sakaguchi M, Tanaka H, Yoon HE, Komoto Y,et al. p16INK4, pRB, p53 and cyclin D1 expression and hypermethy-lation of CDKN2 gene in thymoma and thymic carcinoma. Int J Cancer1997;73:639–44.

37. Suzuki E, Sasaki H, Kawano O, Endo K, Haneda H, Yukiue H, et al.Expression andmutation statuses of epidermal growth factor receptorin thymic epithelial tumors. Jpn J Clin Oncol 2006;36:351–6.

38. Yoh K, Nishiwaki Y, Ishii G, Goto K, Kubota K, Ohmatsu H, et al.Mutational status of EGFR and KIT in thymoma and thymic carcinoma.Lung Cancer 2008;62:316–20.

39. Meister M, Schirmacher P, Dienemann H, Mechtersheimer G, Schna-bel PA, Kern MA, et al. Mutational status of the epidermal growthfactor receptor (EGFR) gene in thymomas and thymic carcinomas.Cancer Lett 2007;248:186–91.

40. Cichowski K, Janne PA. Drug discovery: inhibitors that activate.Nature 2010;464:358–9.

41. Amit I, Wides R, Yarden Y. Evolvable signaling networks of receptortyrosine kinases: relevance of robustness tomalignancy and to cancertherapy. Mol Syst Biol 2007;3:151:1–14.

42. Sos ML, Fischer S, Ullrich R, Peifer M, Heuckmann JM, Koker M, et al.Identifying genotype-dependent efficacy of single and combinedPI3K- and MAPK-pathway inhibition in cancer. Proc Natl Acad SciU S A 2009;106:18351–6.

43. Chakravarti A, Loeffler JS, Dyson NJ. Insulin-like growth factor recep-tor I mediates resistance to anti-epidermal growth factor receptortherapy in primary human glioblastoma cells through continued acti-vation of phosphoinositide 3-kinase signaling. Cancer Res 2002;62:200–7.

44. Workman P. Combinatorial attack on multistep oncogenesis byinhibiting the Hsp90 molecular chaperone. Cancer Lett 2004;206:149–57.

45. Gaspar N, Sharp SY, Pacey S, Jones C, Walton M, Vassal G, et al.Acquired resistance to 17-allylamino-17-demethoxygeldanamycin(17-AAG, tanespimycin) in glioblastoma cells. Cancer Res 2009;69:1966–75.

46. Eccles SA, Massey A, Raynaud FI, Sharp SY, Box G, Valenti M, et al.NVP-AUY922: a novel heat shock protein 90 inhibitor active againstxenograft tumor growth, angiogenesis, and metastasis. Cancer Res2008;68:2850–60.

47. Chandarlapaty S, Sawai A, Ye Q, Scott A, Silinski M, Huang K, et al.SNX2112, a synthetic heat shock protein 90 inhibitor, has potentantitumor activity against HER kinase-dependent cancers. Clin Can-cer Res 2008;14:240–8.

48. Caldas-Lopes E, Cerchietti L, Ahn JH, Clement CC, Robles AI, RodinaA, et al. Hsp90 inhibitor PU-H71, a multimodal inhibitor of malignancy,induces complete responses in triple-negative breast cancer models.Proc Natl Acad Sci U S A 2009;106:8368–73.

49. Cerchietti LC, Lopes EC, Yang SN, Hatzi K, Bunting KL, Tsikitas LA,et al. A purine scaffold Hsp90 inhibitor destabilizes BCL-6 and hasspecific antitumor activity in BCL-6-dependent B cell lymphomas. NatMed 2009;15:1369–76.






2011;17:2237-2249. Published OnlineFirst March 3, 2011.Clin Cancer Res Marco Breinig, Philipp Mayer, Andreas Harjung, et al. Epithelial TumorsGrowth Factor 1 Receptor Contributes to Malignancy of Thymic Heat Shock Protein 90-Sheltered Overexpression of Insulin-Like

Updated version

10.1158/1078-0432.CCR-10-1689doi:

Access the most recent version of this article at:

Material

Supplementary

http://clincancerres.aacrjournals.org/content/suppl/2011/04/14/1078-0432.CCR-10-1689.DC1

Access the most recent supplemental material at:

Cited articles

http://clincancerres.aacrjournals.org/content/17/8/2237.full#ref-list-1

This article cites 49 articles, 17 of which you can access for free at:

Citing articles

http://clincancerres.aacrjournals.org/content/17/8/2237.full#related-urls

This article has been cited by 1 HighWire-hosted articles. Access the articles at:

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

Subscriptions

Reprints and

[email protected]

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

Permissions

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

.http://clincancerres.aacrjournals.org/content/17/8/2237To request permission to re-use all or part of this article, use this link



http://clincancerres.aacrjournals.org/lookup/doi/10.1158/1078-0432.CCR-10-1689

http://clincancerres.aacrjournals.org/content/suppl/2011/04/14/1078-0432.CCR-10-1689.DC1

http://clincancerres.aacrjournals.org/content/17/8/2237.full#ref-list-1

http://clincancerres.aacrjournals.org/content/17/8/2237.full#related-urls

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

mailto:[email protected]

http://clincancerres.aacrjournals.org/content/17/8/2237


heat shock protein 90-sheltered overexpression of insulin-like … · the hitherto disappointing...

Documents