a multiplicity of anti-invasive effects of farnesyl transferase inhibitor sch66336 in human head and...

A multiplicity of anti-invasive effects of farnesyl transferaseinhibitor SCH66336 in human head and neck cancer

Seung Hyun Oh1,2*, Ju-Hee Kang2*, Jong Kyu Woo3, Ok-Hee Lee1, Edward S. Kim1 and Ho-Young Lee3

1 Department of Thoracic/Head and Neck Medical Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX2 Division of Cancer Biology, National Cancer Center, Goyang-si, Republic of Korea3 College of Pharmacy, Seoul National University, Seoul, Republic of Korea

Metastasis is a critical event in the progression of head and neck squamous cell carcinoma (HNSCC) and closely correlates

with clinical outcome. We previously showed that the farnesyl transferase inhibitor SCH66336 has antitumor activities in

HNSCC by inducing the secretion of insulin-like growth factor binding protein 3 (IGFBP-3), which in turn inhibits tumor growth

and angiogenesis. In our study, we found that SCH66336 at a sublethal dose for HNSCC inhibited the migration and invasion

of HNSCC cells. The inhibitory effect of SCH66336 was associated with the blockade of the IGF-1 receptor (IGF-1R) pathway

via suppressing IGF-1R itself and Akt expression. Consistent with previous work, induction of IGFBP-3 by SCH66336 also

contributed in part to the anti-invasive effect. SCH66336 treatment also reduced the expression and activity of the urokinase-

type plasminogen activator (uPA) and matrix metalloproteinase 2 (MMP-2), both important regulators of tumor metastasis. The

effect of SCH66336 on uPA activity was inhibited partly by knockdown of IGFBP-3 using small interfering RNA. The inhibitory

effect of SCH66336 on migration or invasion was attenuated partly or completely by knockdown of IGFBP-3, Akt or IGF-1R

expression, respectively. Our results demonstrate that the IGF-1R pathway plays a major role in the proliferation, migration

and invasion of HNSCC cells, suggesting that therapeutic obstruction of the IGF-1R pathway would be a useful approach to

treating patients with HNSCC.

Despite improvements in the locoregional control of headand neck squamous cell carcinoma (HNSCC), the survivalrate of patients with this cancer has not improved substan-tially, indicating the need for more effective therapy.1 Inmost of the patients with HNSCC, the primary tumor hasmetastasized to the regional lymph nodes or distant organsby the time of diagnosis.2,3 Metastatic HNSCC tumors aremost commonly found in the aerodigestive tract, and patientswith metastatic tumors have significantly lower survival ratesthan patients without them.3,4 Consequently, inhibition ofmetastasis may be an important strategy for HNSCCtreatment.

Tumor metastasis is a complicated and dynamic processinvolving tumor cell dissociation from the primary tumor,degradation or remodeling of the extracellular matrix (ECM),infiltration of surrounding stroma, intravasation, survival ofmalignant cells in the circulation system, invasion of a targetorgan and finally growth at a secondary site.5 Tumor cellsmust interact with the ECM through a multistep process thatis tightly regulated by cytokines and growth factors andrequires the expression of tumor-associated proteases.5 Oneof the main protease systems is the plasminogen activatorsystem, which is composed of tissue-type plasminogen activa-tor and urokinase-type plasminogen activator (uPA), theirinhibitors, plasminogen activator inhibitors 1 and 2 and uPAreceptor (uPAR). uPA is expressed as a single-chain proen-zyme (pro-uPA) with little or no intrinsic enzymatic activity.Pro-uPA is converted into an enzymatically active, high-

Key words: insulin-like growth factor binding protein-3, insulin-like

growth factor-1 receptor, farnesyl transferase inhibitor, head and neck

squamous cell carcinoma, invasion, SCH66336

Abbreviations: CM: conditioned medium; ECM: extracellular

matrix; EV: parental vector; FTI: farnesyl transferase inhibitor; FBS:

fetal bovine serum; HNSCC: head and neck squamous cell

carcinoma; HMW: high molecular weight; IGF: insulin-like growth

factor; IGF-1R: IGF-1 receptor; IGFBP-3: IGF-binding protein 3;

LMW: low molecular weight; MMP: matrix metalloproteinase;

siRNA: small interfering RNA; uPA: urokinase-type plasminogen

activator; uPAR: uPA receptor

Conflict of interest: Dr. Edward S. Kim has received guest support

or honoraria from Eli Lilly, Genentech, Boehringer Ingelheim, Bayer

and Onyx.

*S.H.O. and J.-H.K. contributed equally to this work

Grant sponsor: National Research Foundation of Korea (NRF);

Grant number: 2011-0017639; Grant sponsor: The second phase of

the Brain Korea 21 program in 2011; Grant sponsor: National

Institutes of Health; Grant number: R01 CA100816; Grant sponsor:

National Cancer Center Research; Grant number: NCC0810420

DOI: 10.1002/ijc.26373

History: Received 6 Oct 2010; Accepted 28 Jun 2011; Online 16 Aug

2011

Correspondence to: Ho-Young Lee, PhD, College of Pharmacy,

Seoul National University, 599 Gwanak-ro, Gwanak-gu, Seoul

151-742, Republic of Korea, Tel.: þ82-2-880-9277, Fax:

þ82-2-872-1795, E-mail: [email protected]

Can

cerCellBiology

Int. J. Cancer: 000, 000–000 (2012) VC 2011 UICC

International Journal of Cancer

IJC

molecular-weight (HMW), two-chain form (HMW-uPA) bythe action of plasmin or other enzymes that are commonlyenriched in cancer cells, such as cathepsin B.6 HMW-uPA isfurther degraded by plasmin into the enzymatically active,low-molecular-weight (LMW), two-chain form of uPA(LMW-uPA) and the enzymatically inactive amino-terminalfragment. These uPAs can convert inactive plasminogen tothe serine protease plasmin. Plasmin can degrade mostcomponents of the ECM directly or activate other matrixmetalloproteinases (MMPs), which are important in ECMmodulation, composition, and maintenance and in cancermetastasis.7,8 uPAR binding to uPA increases plasmin activityand focuses proteolytic activity on the cell surface.7 uPAoverexpression is associated with the malignant phenotype ofthyroid cancer, non-small cell lung cancer and other solidtumors.7,9 It is also a potential biomarker of lymph node me-tastasis in oral squamous cell carcinoma.10

While searching for agents that inhibit invasive activitiesin HNSCC cells, we found that SCH66336, a nonpeptide tri-cyclic CAAX mimetic farnesyl transferase inhibitor (FTI),which has shown inhibitory effects on tumor angiogenesisand cancer cell migration,11,12 suppresses the ability ofHNSCC cells to express matrix proteinases, leading todecreased cell migration. However, the underlying molecularmechanisms by which SCH66336 inhibits cancer cell migra-tion and invasion have not been well studied. In our study,we found that SCH66336 modulated expression of insulin-like growth factor (IGF) binding protein 3 (IGFBP-3), IGF-1receptor (IGF-1R), and Akt, which in turn blocked the IGF-1R pathway and expression/activity of uPA and MMP-2.

Materials and MethodsAnimals, cells and materials

We purchased 5-week-old female nude mice from HarlanSprague Dawley (Indianapolis, IN). The human HNSCC celllines UMSCC38 and SqCC/Y1, established originally by Drs.Thomas Carey (University of Michigan, Ann Arbor, MI) andMichael Reiss (Yale University, New Haven, CT), wereobtained from Dr. Reuben Lotan (The University of TexasMD Anderson Cancer Center, Houston, TX)13; the TR146cell line was provided by Dr. Alfonse Balm (The NetherlandsCancer Institute, Amsterdam, The Netherlands). Cells werecultured in Dulbecco’s modified Eagle’s medium (DMEM)supplemented with 10% fetal bovine serum (FBS) and antibi-otics. Cell culture inserts incorporating polyester Transwellmembranes (6.4-mm diameter, 8-lm pore size) and 24-wellplates for the invasion assay were from Costar (Cambridge,MA). We obtained 3-(4,5-dimethylthazol-2-yl)-2,5-diphenyl-tetrazolium bromide (MTT) from Sigma and the AmiconUltra-4 centrifugal filter device from Millipore (Bedford,MA). Antibodies against IGF-1Rb, H-Ras and actin werepurchased from Santa Cruz Biotechnology (Santa Cruz, CA).Anti-IGFBP-3 antibody was purchased from Diagnostic Sys-tems Laboratories (Webster, TX), anti-uPA antibody from

American Diagnostica (Stamford, CT) and anti-phosphoryl-ated Akt (Ser473) and anti-Akt antibodies from Cell SignalingTechnology (Danvers, MA). SCH66336 was provided bySchering-Plough Research Institute (Kenilworth, NJ). FTI-277was purchased from Calbiochem (San Diego, CA). SCH66336and FTI-277 were dissolved in dimethyl sulfoxide at variousconcentrations to establish dose responses.

Reverse transcription-polymerase chain reaction

Total RNA was isolated from cells using Trizol reagent (Invi-trogen). cDNA was synthesized, and the reverse transcrip-tion-polymerase chain reaction (RT-PCR) was performed asdescribed elsewhere.14 The primer sequences were as follows:50-CAAGAGTGCGCTGACCATCC-30 (sense) and 50-CCGGA CTCACGCACCAAC-30 (antisense) for H-Ras; 50-GCGTGCGGCAAGACGTCTG-30 (sense) and 50-TCATAGCACCTTGCAGCAGTT-30 (antisense) for RhoB; and 50-GGTGAA GGTCGGTGTGAACGGATTT-30 (sense) and 50-AATGCCAAAGTTGTCATGGATGACC-30 (antisense) for GAPDH. The thermocycler conditions used for amplificationwere 94�C for 6 min (hot start), 94�C for 45 sec, 56–60�Cfor 45 sec and 72�C for 1 min. The control PCR was per-formed in 26 or 27 cycles with 0.5 lL of cDNA solution toallow quantitative comparisons among the cDNAs developedfrom identical reactions with primers for GAPDH. Amplifica-tion products (8 lL) were resolved in 2% agarose gel, stainedwith ethidium bromide, visualized in a transilluminator andphotographed using a Kodak Gel Logic 100 imaging system(Eastman Kodak, Rochester, NY).

Small interfering RNA transfection

H-Ras and RhoB small interfering RNAs (siRNAs) were pur-chased from Ambion (Austin, TX). IGFBP-3, Akt1, Akt2,Akt3 and nonspecific control siRNAs were synthesized atDharmacon (Chicago, IL), and IGF-1R siRNA was synthe-sized at Bioneer (Daejeon, Korea). UMSCC38 cells weretransfected with siRNA using oligofectamine (Invitrogen) andincubated in a medium with 10% FBS containing 0.1%DMSO or SCH66336 (5 lmol/L) for 2 days. Cells were thenharvested for Western blot or RT-PCR analysis. Conditionedmedium (CM) for the zymography assay was also collectedfrom cells that had been incubated in 10% FBS medium withor without SCH66336 (5 lmol/L) for 2 days and transferredto medium without FBS for 1 day. CM was concentratedusing the Amicon Ultra-4 centrifugal filter device. Proteinconcentrations were measured using the bicinchoninic acidassay (Pierce Biotechnology, Rockford, IL). UMSCC38 cellswere pretreated with SCH66336 and subjected to an in vitroinvasion assay.

Adenoviral studies

Construction and amplification of adenoviruses expressingIGFBP-3 (Ad-BP3) or uPA (Ad-uPA) have been previouslydescribed.15 UMSCC38 cells that had been infected with sev-eral doses of empty vector (Ad-EV), Ad-BP3 or Ad-uPA for

Can

cerCellBiology

2 Anti-invasive effect of SCH66336 by IGF-1R pathway regulation


2 days were used for the invasion assay. CM was also col-lected from cells that had been infected with adenoviruses,incubated in 10% FBS medium with or without SCH66336 (5lmol/L) for 2 days and transferred to medium without FBSfor 1 day. To assess the involvement of Hsp90 in theSCH66336-mediated suppression of invasion, UMSCC38 cellsthat had been infected with Ad-EV or adenovirus-expressingHsp90 (Ad-Hsp90) were incubated in 0.1% FBS mediumwith or without SCH66336 (5 lmol/L) for 1 day.

Western blot analysis

Total cell extracts were harvested from HNSCC lines aftertreatment. Whole-cell lysate preparation, protein quantifica-tion, gel electrophoresis and Western blotting were per-formed as described elsewhere.13 UMSCC38 and SqCC/Y1cells were incubated for 24 hr in the medium containing0.1% FBS with or without 5 lmol/L SCH66336. Equivalentamounts of proteins from the cell lysate or CM from eachtreatment group were resolved using sodium dodecyl sulfate-polyacrylamide gel electrophoresis and immunoblotted withprimary antibodies. Loading of equal amounts of proteins inCM samples was confirmed by Coomassie blue staining ofduplicate gels and Ponceau staining of the membrane.

Migration and invasion assays

In vitro migration and invasion assays were performed asdescribed elsewhere.16 In brief, CM obtained by culturing for18 hr in DMEM with 10% FBS was placed into the lowerchamber of each well as a chemoattractant, and 5 � 104 can-cer cells were placed in the upper chamber in DMEM with-out FBS. For the migration assay, filters were coated with a0.1 mg/mL solution of collagen type IV (Trevigen, Gaithers-burg, MD) in PBS. The invasion assay was performed in thesame manner except the Transwell units were coated withMatrigel (Becton Dickinson Labware, Bedford, MA) at a con-centration of 50 lg/mL in PBS. UMSCC38 cells were infectedwith Ad-BP3 or EV [10 and 50 plaque-forming units (pfu)per cell] for 24 hr (control cells were not infected), harvestedby trypsinization, washed and placed into the Transwell.Within 24 hr of incubation of SqCC/Y1 and TR146 cells in0.1% FBS culture medium, with or without 5 lmol/LSCH66336, cells on the membranes were fixed and stainedwith hematoxylin and eosin. The numbers of cells in four in-dependent fields were counted under a light microscope(Leica, Wetzlar, Germany). The results are expressed as thepercentage of cells that migrated or invaded compared tocontrol cells.

MTT assay

The inhibitory effect of SCH66336 on growth of HNSCCcells was determined using a MTT assay. Cells were plated in96-well plates (3 � 103/well) and cultured in medium withor without FBS. The cells then were grown for an additionaltotal incubation of 24 or 72 hr.

Gelatin and fibrinogen/plasminogen zymography

The proteolytic activity of MMP-2/MMP-9 and uPA in CMwas analyzed by substrate gel electrophoresis16 using sodiumdodecyl sulfate-polyacrylamide gel electrophoresis gels con-taining 0.2% (m/v) gelatin or 0.12% (m/v) fibrinogen and 0.01NIH unit/mL plasminogen. CM from each treatment groupwas concentrated with the Amicon Ultra-4 and loaded ontogels. After analysis, gels were washed with 2.5% Triton X-100and incubated in a zymogram incubation buffer (50 mmol/LTris-HCl, 0.15 M NaCl, 10 mmol/L CaCl2, 0.02% NaN3) over-night at 37�C. Clear bands, which are indicative of gelatino-lytic activity, were visualized by staining with Coomassie blue.

Animal experiment

All animal procedures were performed in accordance with aprotocol approved by the Institutional Animal Care andUsage Committee. Orthotopic tongue tumor model wasestablished as described elsewhere.11 Female nude mice underanesthesia were injected with UMSCC38 cells (2 � 106) intothe lateral tongue (n ¼ 5). One week later, when tumorsstarted to develop, 40 mg/kg of SCH66336 or 20% hydroxyl-propyl-betacyclodexatrin control vehicle was given orallytwice a day for 3 weeks. The tongues were removed, fixed,embedded in paraffin and sectioned for uPA staining.

Statistical analysis

Data are shown as means þ standard deviations or means þstandard error. Cell proliferation, migration and invasiondata in HNSCC cells were analyzed by Student’s t-test usingExcel 2007 software (Microsoft, Redmond, WA) to determinethe statistically significant difference between groups. All sta-tistical tests were two-sided. A p value of less than 0.05 wasconsidered statistically significant.

ResultsEffects of SCH66336 on HNSCC cell migration and invasion

Because an important process in cancer metastasis is localinvasion into the host stroma,5 we first assessed the effect ofSCH66336 on the invasiveness of UMSCC38 cells by in vitromigration and invasion assays. Pretreatment with 1–5 lmol/LSCH66336 for 24 hr, which showed a mild effect on cell pro-liferation, significantly reduced the number of migrating andinvading cells (Fig. 1a). We also observed consistent inhibi-tory effects of SCH66336 on the migration and invasion ofSqCC/Y1 and TR146 cells (Fig. 1b). Although the develop-ment of FTIs was based on the finding that oncogenic Ras ispost-translationally prenylated to transform cells,17 FTIs areknown to inhibit anchorage-independent growth in manyhuman cancer cell lines irrespective of whether they expresswild-type or mutated Ras.18,19 Hence, we first determinedwhether silencing H-Ras or RhoB expression by siRNAabolished the activity of SCH66336 in UMSCC38 cells.Transfection with siRNAs targeting H-Ras or RhoB specifi-cally inhibited H-Ras or RhoB RNA expression in UMSCC38

Can

cerCellBiology

Oh et al. 3


cells, whereas control scrambled siRNA did not affect expres-sion (Fig. 1c, left). Cell viability was not specifically altered inany cells (data not shown). An in vitro invasion assay per-formed on cells transfected with siRNAs that targeted H-Rasor RhoB revealed no attenuation of SCH66336’s inhibitoryeffects (Fig. 1c, right). This finding indicates that SCH66336suppressed the invasive activity of HNSCC cells by anH-Ras- or RhoB-independent mechanism in this cell line.

Effects of SCH66336 on uPA and MMP-2 activity

in HNSCC cells

Because uPA is considered to be the critical trigger for con-verting plasmin, which degrades various components of the

ECM (e.g., fibrin, fibronectin, laminin) under physiologic andpathologic conditions (such as invasion) through plasmin orMMP activation, we analyzed the effects of SCH66336 on theactivity of uPA. To this end, we performed Western blot andzymography analyses and analyzed expression and activity ofuPA in CM from UMSCC38 cells. As shown in Figure 2a(top), SCH66336 markedly decreased the levels of uPA pro-tein, especially LMW-uPA (Fig. 2a, top). Results of the plas-minogen/fibrinogen zymography assay on CM also showedthat SCH66336 treatment reduced the activity of uPA, withdramatic effects on LMW-uPA (Fig. 2a, middle). BecauseMMP activity is stimulated by uPA or plasmin, we also testedthe effect of SCH66336 on MMP activity. A gelatin-

Figure 1. Anti-invasive activity of SCH66336 in HNSCC. (a) Migration and invasion of UMSCC38 cells were inhibited by treatment with

SCH66336 (1 or 5 lmol/L) for 24 hr. UMSCC38 cell proliferation was marginally inhibited by treatment with SCH33663 for 24 hr.

(b) Migration and invasion of SqCC/Y1 and TR146 cells were inhibited by treatment with SCH66336 (5 lmol/L) for 24 hr. (c) Invasion of

UMSCC38 cells after H-Ras or RhoB siRNA transfection was inhibited by SCH66336. Values are means þ SEM. * p < 0.05; ** p < 0.01

compared to the control group. SCH, SCH66336. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Can

cerCellBiology



zymography assay on CM revealed that SCH66336 decreasedMMP-2 activity (Fig. 2a, bottom). MMP-9 activity was sup-pressed by SCH66336 to a much lesser extent.

We further tested the effects of SCH66336 on uPA expres-sion in HNSCC in vivo in a mouse model. An immunohisto-chemical analysis of uPA in tongue tumor tissue revealedthat treatment of mice bearing orthotopic tongue tumors

with oral SCH66336 (40 mg/kg) markedly inhibited uPAexpression in the tumors (Fig. 2b).These findings suggest thatoral administration of SCH66336 was sufficient to decreaseuPA expression in vivo.

Because uPA expression in stromal cells has been shownto contribute to tumor metastasis,5 we evaluated the effect ofSCH66336 on UMSCC38 cell invasiveness stimulated by

Figure 2. Effects of SCH66336 on the expression and activities of uPA, MMP-2, and MMP-9. (a) Effects of SCH66336 on uPA expression in

CM secreted from SCH66336-treated UMSCC38 cells (top). Effects of SCH66336 on uPA, MMP-2 and MMP-9 activities were determined by

gelatin or plasminogen/fibrinogen zymography (middle, bottom). One representative result from two independent experiments is shown.

(b) Orthotopic tongue tumor tissues from mice were stained with an anti-uPA antibody. (c) UMSCC38 cell invasion was analyzed in the CM

obtained from overnight incubation of Wi-38 lung fibroblasts with (FTI-CM) or without (Con-CM) SCH66336 (5 lmol/L) in the lower

compartment of the Transwell unit. (d) UMSCC38 cells were infected with control adenoviruses (Ad-EV) or adenoviruses expressing uPA

(Ad-uPA) at a concentration of 50 or 100 pfu per cell for 48 hr and used for the invasion assay. Shown are representative pictures of

UMSCC38 cells that invaded through the Matrigel. (e) SCH66336 inhibited invasion of UMSCC38 cells in Ad-EV- or Ad-uPA-infected cells

by 76.0% (EV), 27.4% (uPA 50) and 35.2% (uPA 100). Data shown are means + SEM of three experiments. *p < 0.05, **p < 0.01 and

***p < 0.001. SCH, SCH66336. pfu, plaque forming unit.

Can

cerCellBiology

Oh et al. 5


Wi38 fibroblasts. The plasminogen/fibrinogen zymographyassay also showed reduced activity of uPA in CM fromSCH66336-treated Wi38 cells (data not shown). Furthermore,UMSCC38 cell invasiveness was significantly decreased whenattracted with medium collected from SCH66336-treatedWi38 cells compared to medium collected from untreatedWi38 cells (Fig. 2c). These findings suggest that downregula-tion of uPA expression in HNSCC and stromal cells maydecrease the invasive potential of HNSCC cells.

To determine the specific role of uPA in the invasion ofHNSCC cells, we performed an invasion assay on UMSCC38cells infected with Ad-uPA followed by SCH66336 treatment.As shown in the representative picture of UMSCC38 cells(Fig. 2d) and by statistical analysis (Fig. 2e), Ad-uPAincreased the invasiveness of UMSCC38 cells and attenuatedthe anti-invasive effects of SCH66336 in a dose-dependentmanner. These findings suggest that uPA is a major targetfor SCH66336’s inhibitory effects on HNSCC cell invasion.

uPA downregulation by SCH66336 through

IGFBP-3 induction

We studied the mechanism by which SCH66336 regulatesuPA expression. We previously showed that the antiangio-genic effect of SCH66336 was mediated by IGFBP-3 induc-tion.11 Hence, we sought to determine whether IGFBP-3 wasimplicated in the anti-invasive effect of SCH66336 onHNSCC cells. We first tested whether expression of IGFBP-3induced by Ad-BP3 could inhibit the invasive activities ofUMSCC38 cells. As shown by Western blotting and plasmin-ogen/fibrinogen and gelatin zymography assays on CM fromIGFBP-3 adenovirus-infected cells, IGFBP-3 induced amarked decrease in the activity of MMP-2, HMW-uPA andLMW-uPA (Fig. 3a). Moreover, UMSCC38 cell invasivenesswas significantly inhibited by Ad-BP3 (Fig. 3b) at doses of10–50 pfu per cell, which showed no effects on UMSCC38cell proliferation (data not shown). To determine the impor-tance of IGFBP-3 in mediating SCH66336-induced anti-inva-sive effects in HNSCC cells, we tested the effects ofSCH66336 on UMSCC38 cells transfected with IGFBP-3-tar-geting siRNA, which effectively suppressed IGFBP-3 mRNAand protein expression (Fig. 3c, top). UMSCC38 cells exhib-ited a slight increase in basal uPA activity after IGFBP-3siRNA transfection. The effects of SCH66336 on HMW-uPAand LMW-uPA activities were partially attenuated inUMSCC38 cells transfected with IGFBP siRNA (Fig. 3c, bot-tom). The inhibitory effect of SCH66336 on the invasion ofIGFBP-3 siRNA-transfected cells was also partially attenuatedcompared to control siRNA-transfected cells (Fig. 3d). Thesefindings indicate that IGFBP-3 is an SCH66336-induced in-hibitory molecule against these ECM degraders.

SCH66336 exerts anti-invasive activities in HNSCC cells by

blocking IGF-1R and Akt expression

Because the inhibition of IGFBP-3 expression did not com-pletely attenuate the anti-invasive activity of SCH66336 in

UMSCC38 cells, we sought other molecules that acted asSCH66336-modulated regulators of invasion. Because theIGF-1R pathway is known to upregulate uPA expression,20,21

we sought to determine whether the IGF-1R signaling axis isa target of anti-invasive actions of SCH66336 in HNSCCcells. We found that SCH66336 treatment for 24 hr sup-pressed the expression of IGF-1R and Akt in UMSCC38 andSqCC/Y1 cells cultured in 0.1% FBS medium, in whichmigration and invasive activities of the cells were analyzed(Fig. 4a). We further assessed the effects of FTI-277, anotherFTI, on the expression of IGF-1R and Akt in SqCC/Y1. Anobvious decrease in IGF-1R and Akt protein levels wasobserved (Fig. 4b). No change was detected in the expressionof MAPK1/2 after FTI277 treatment. To determine the spe-cific role of the IGF-1R/Akt pathway in mediating anti-inva-sive actions of SCH66336 in HNSCC cells, we tested the in-hibitory effect of SCH66336 on the migration and invasionof UMSCC38 cells that were transfected with siRNA targetingIGF-1R. We found that IGF-1R expression was efficientlyinhibited by siRNA (Fig. 4c, left). Inhibition of IGF-1Rexpression suppressed migration (Fig. 4c, middle) and inva-sion (Fig. 4c, right) of UMSCC38 cells. The effects ofSCH66336 on migration and invasion were markedly attenu-ated when IGF-1R expression was blocked compared to con-trol siRNA-transfected cells. We further assessed the role ofAkt in anti-invasive activities of SCH66336. To completelyblock the expression of Akt, we transfected UMSCC38 cellswith siRNAs targeting specific Akt1, Akt2 and Akt3 isoforms.Transfection with these siRNAs markedly inhibited Akt pro-tein expression in UMSCC38 cells, whereas control scrambledsiRNA did not affect Akt protein expression (Fig. 4d, left).The inhibitory effect of SCH66336 on migration was com-pletely blocked in UMSCC38 cells transfected with Akt siR-NAs compared to those transfected with control siRNA (Fig.4d, right). These results indicate that SCH66336 suppressedthe invasive activity of HNSCC cells by modulating IGF-1Rsignaling components, including IGF-1R and Akt.

We investigated the mechanism that mediates downregu-lation of IGF-1R pathway by SCH66336. SCH66336 had beenoriginally developed to inhibit posttranslational activation ofRas by blocking farnesylation.22 However, none of the cellsused in our study have a ras mutation (data not shown), sug-gesting that proteins other than Ras are responsible for theeffects of SCH66336 on IGF-1R pathway. SCH66336 hasbeen shown to inhibit Hsp90 function, resulting in decreasedlevels of Hsp90 client proteins,14 such as IGF-1R and Akt.23

Hence, we studied the role of Hsp90 in SCH66336-mediatedeffects on IGF-1R pathway in UMSCC38 cells. To this end,we tested whether overexpression of Hsp90 would protectUMSCC38 cells from SCH66336-induced decrease in Aktprotein levels. To this end, we infected UMSCC38 cells withan adenoviral vector containing HA-tagged Hsp90 (Ad-Hsp90).24 Induction of Hsp90 protein expression in the cellsinfected with Ad-Hsp90 was determined by the increase inthe HA expression in Western blot analysis (Fig. 5). Indeed,

Can

cerCellBiology



cells infected with Ad-Hsp90 were completely rescued fromthe effect of SCH66336 on Akt and pAkt levels, unlike cellsinfected with Ad-EV, suggesting an impact of Hsp90 in theaction of SCH66336 against invasive activities of HNSCC.

DiscussionIn our study, we found that HNSCC cell invasion is signifi-cantly inhibited by the FTI SCH66336 at a concentrationreported to be achievable in serum.25 We were led to followthis line of research by early observations of tumor regression

in clinical trials involving SCH66336.26 Increasing evidencehas demonstrated that the Ras oncogenes are among thosemost frequently found in human cancers and that signalingpathways mediated by these oncogenes play critical roles incancer cell proliferation, angiogenesis and metastasis. Hence,the antineoplastic FTIs were developed to inhibit the catalyticfarnesylation step in Ras.17 However, it has become clear thatRas may not be the only critical target of FTIs,17–19 and othermolecules (e.g., RhoB, Akt and prostacyclin receptor) havebeen implicated in the suppression of transformed

Figure 3. Role of IGFBP-3 in SCH66336-induced inhibition of HNSCC cell invasion. (a) UMSCC38 cells were infected with control

adenoviruses (Ad-EV) or adenoviruses expressing IGFBP-3 (Ad-BP-3) at a concentration of 1, 10 or 50 pfu per cell for 48 hr. Ad-EV- or Ad-

BP-3-infected UMSCC38 cells and CM from them were used for Western blotting and gelatin and fibrinogen/plasminogen zymography. (b)

Invasion of UMSCC38 cells treated as in a through the Matrigel-coated Transwell filter was inhibited by IGFBP-3. (c) UMSCC38 cells were

transfected with control or IGFBP-3 siRNA and treated with SCH66336. CM and cells were used for fibrinogen/plasminogen zymography and

Western blotting, respectively. (d) Cells transfected as in c were used for the invasion assay in the presence or absence of SCH66336 for

24 hr. Data shown are means þ SEM of three experiments. **p < 0.01, ***p < 0.001. SCH, SCH66336. pfu, plaque forming unit.

Can

cerCellBiology

Oh et al. 7


morphologic characteristics and the induction of apoptosis byFTIs.13,27–29 We found that the blockade of H-Ras and RhoBby specific siRNAs did not influence SCH66336-induced in-hibitory effects on the invasion of UMSCC38 cells. Therefore,the anti-invasive activity of SCH66336 in HNSCC cells doesnot seem to depend on Ras and RhoB but to operate byother mechanisms of action.

We found that SCH66336 treatment reduced the amountof secreted HMW- and LMW-uPA and MMP-2 in superna-tant from UMSCC38 cells, consistent with previous find-ings.12 uPA and MMP-2 are important regulators of metasta-sis and are highly correlated with poor prognosis in patients

with HNSCC.10,30 The effects of SCH66336 on uPA expres-sion seem to have a role in its anti-invasive activities in thecells, because overexpression of uPA induced by Ad-uPAattenuated SCH66336’s actions at least in part. The pro-uPAsecreted from cells is converted to active uPA when bound touPAR. Inactive uPA–plasminogen activator inhibitor-1 com-plexes bound to uPAR are promptly internalized anddegraded in lysosomes, whereas unoccupied uPAR recycles tothe cell surface.7 The decreased levels of secreted uPAs fromSCH66336-treated cells could be the result of downregulationof uPAR expression, direct inhibition of uPA transcription ortranslation or interference with secreted uPA conversion in

Figure 4. Role of the IGF-1R/Akt pathway in SCH66336-induced inhibition of HNSCC cell invasion. (a) Western blot analysis of UMSCC38

and SqCC/Y1 cells incubated for 24 hr in the medium containing 0.1% FBS with or without 5 lmol/L SCH66336. (b) Western blot analysis

of UMSCC38 cells incubated for 24 hr in the medium containing 0.1% FBS with or without 5 lmol/L FTI277. (c) RT-PCR and Western blot

analysis using UMSCC38 cells after transfection of siRNAs targeting IGF-1R (left). Migration and invasion assay in UMSCC38 cells treated

with 5 lmol/L SCH66336 after transfection with IGF-1R siRNA (right). (d) Western blot analysis using UMSCC38 cells after transfection of

siRNAs targeting Akt1, Akt2 and Akt3 (left). Migration assay using UMSCC38 cells treated with SCH66336 after transfection with control or

Akt siRNAs (right). Data shown are means þ SEM of three experiments. **p < 0.01 compared to untreated control cells. SCH, SCH66336.

Can

cerCellBiology



UMSCC38 cells. In our study, SCH66336 did not appear tosuppress uPAR expression or to inhibit secreted uPA. Thisconclusion is based on our results showing that uPAR levelsremained unchanged (data not shown) after the SCH66336treatment, and levels of secreted HMW-uPA and LMW-uPA(as measured by plasminogen/fibrinogen zymography) werereduced in agreement with levels assessed by Western blotanalysis. SCH66336 has been shown to increase IGFBP-3expression, which mediates IGF-independent inhibition ofuPA transcription.11,16 In our study, overexpression ofIGFBP-3 markedly decreased HMW- and LMW-uPA andMMP-2 activities. Hence, the effect of SCH66336 in decreas-ing levels of secreted uPA could result from suppression ofthe expression of genes encoding uPA. However, SCH66336’sactions on secreted uPA levels and invasive activities inUMSCC38 cells were only marginally abolished by the siRNAinhibition of IGFBP-3 expression, indicating that uPA geneexpression is not the only operative factor.

The results described above suggest that SCH66336 mayhave other targets that are responsible for invasive activities inUMSCC38 cells. In our study, SCH66336 also inhibited theactivation of MMP-2 in UMSCC38 cells, which could havebeen mediated by the decreased plasminogen-to-plasmin con-version activity of uPA8 or by the direct inhibition of theexpression of MMP-2 transcription and translation. SCH66336is known to affect several proteins, such as tissue-type plasmin-ogen activator, MMP-9, tissue inhibitor of MMP 1 (TIMP-1)and TIMP-2, all of which play important roles in ECM remod-eling in prostate cancer cells.12 Akt has also been implicated inthe antitumor activity of SCH66336.13,31 SCH66336 has theability to inhibit Hsp90 function,14 which stabilizes client pro-teins, such as IGF-1R and Akt. Given that the IGF-1R/Akt

pathway has a major role in cell invasion and metastasis,32 weattempted to test whether IGF-1R and Akt regulation contrib-utes to the action of SCH66336 against invasion of HNSCCcells and whether Hsp90 function was involved in the action.We were surprised to find that SCH66336-mediated anti-inva-sive effects on UMSCC38 cells were almost completely blockedby the siRNA-mediated knockdown of IGF-1R or Akt expres-sion. These findings suggest that IGF-1R signaling is a majoroperative factor for anti-invasive activity of SCH66336 inHNSCC. In support of this suggestion, cixutumumab (the ther-apeutic antibody to IGF-1R) effectively inhibited the invasionof UMSCC38 cells in vitro (data not shown). We also notedthat overexpression of Hsp90 rescued the cells from anti-inva-sive activities of SCH66336. These findings indicate that IGF-1R pathway is regulated by SCH66336, at least in part, throughan Hsp90-dependent pathway.

In conclusion, SCH66336 inhibits uPA and MMP-2expression/activity and has anti-invasive activity by inducingIGFBP-3 expression and by decreasing IGF-1R and Akt pro-tein levels, leading to effective inhibition of the IGF-1R sig-naling pathway. SCH66336 (Lonafarnib) was well toleratedand demonstrated disease control in patients with HNSCC.33

Unfortunately, the development of the FTIs has been discon-tinued. However, our current study on the mechanism ofaction of these compounds provides rationale for the use ofthe previously discredited anticancer drug in patients withHNSCC, in which activity of IGF-1R is frequently deregu-lated through high levels of systemically circulating or tissue-derived IGFs, overexpression of IGF-1R and/or decreasedexpression of IGFBPs, loss of heterozygosity for IGF-2R andbiallelic expression of IGF-2.34–38 IGF-1R has been shown tostimulate cell proliferation and survival, protecting cells fromapoptosis. There are numerous preclinical and clinical studiesunderway to test two major anti-IGF-1R strategies, whichinvolve either monoclonal antibodies (mAbs) or small-mole-cule tyrosine kinase inhibitors (TKIs) against IGF-1R.39,40

Anti-IGF-1R monoclonal antibodies and TKIs have shownfavorable profiles in phase I clinical trials.41–43 However,recent clinical studies in patients with recurrent and/or meta-static HNSCC have revealed only very modest therapeuticbenefit from anti-IGF-1R mAbs.44–46 Previously, we andothers have demonstrated that the interplay between theEGFR and IGF-1R pathways and overlapping downstreamsignaling mediators, including Akt, induce resistance to tyro-sine kinase inhibitors of EGFR and IGF-1R.47–49 A previousstudy clearly shows that inhibition of IGF-IR stimulates acti-vation of EGFR.50 Our recent study identified activation ofAkt in HNSCC cells as a result of blockade of IGF-1R by cix-utumumab (data not shown). Given the multiplicity of effectsof SCH66336 on the IGF system, FTIs have unique featuresthat make it an effective IGF-1R-targeting antineoplastic drugin clinical trials. Further studies are warranted to identifypredictive biomarkers for selecting likely responders to FTIs-based therapy and then evaluate the efficacy of the drugs intherapy for patients with HNSCC.

Figure 5. Hsp90-mediated stabilization of Akt. Western blot

analysis in the UMSCC38 cells, which were infected with an

adenovirus carrying Hsp90 (Ad-Hsp90) or empty control (Ad-EV)

and incubated for 24 hr in the medium containing 0.1% FBS with

or without 5 lmol/L SCH66336. SCH, SCH66336; pfu, plaque

forming unit.

Can

cerCellBiology

Oh et al. 9


References

1. Marur S, Forastiere AA. Challenges ofintegrating chemotherapy and targetedtherapy with radiation in locally advancedhead and neck squamous cell cancer. CurrOpin Oncol 2010;22:206–11.

2. Day GL, Blot WJ. Second primary tumorsin patients with oral cancer. Cancer 1992;70:14–9.

3. Leon X, Hitt R, Constenla M, Rocca A,Stupp R, Kovacs AF, Amellal N, Bessa EH,Bourhis J. A retrospective analysis of theoutcome of patients with recurrent and/ormetastatic squamous cell carcinoma of thehead and neck refractory to a platinum-based chemotherapy. Clin Oncol (R CollRadiol) 2005;17:418–24.

4. Saba NF, Khuri FR. Optimizing systemictherapy in squamous cell carcinoma of thehead and neck with a focus on targetedagents. Curr Opin Oncol 2009;21:232–7.

5. Duffy MJ, McGowan PM, Gallagher WM.Cancer invasion and metastasis: changingviews. J Pathol 2008;214:283–93.

6. Goretzki L, Schmitt M, Mann K, Calvete J,Chucholowski N, Kramer M, Gunzler WA,Janicke F, Graeff H. Effective activation ofthe proenzyme form of the urokinase-typeplasminogen activator (pro-uPA) by thecysteine protease cathepsin L. FEBS Lett1992;297:112–8.

7. Smith HW, Marshall CJ. Regulation of cellsignalling by uPAR. Nat Rev Mol Cell Biol2010;11:23–36.

8. Kessenbrock K, Plaks V, Werb Z. Matrixmetalloproteinases: regulators of the tumormicroenvironment. Cell 2010;141:52–67.

9. Malinowsky K, Bollner C, Hipp S, Berg D,Schmitt M, Becker KF. UPA and PAI-1analysis from fixed tissues—newperspectives for a known set of predictivemarkers. Curr Med Chem 2010;17:4370–7.

10. Nagata M, Fujita H, Ida H, Hoshina H,Inoue T, Seki Y, Ohnishi M, Ohyama T,Shingaki S, Kaji M, Saku T, Takagi R.Identification of potential biomarkers oflymph node metastasis in oral squamouscell carcinoma by cDNA microarrayanalysis. Int J Cancer 2003;106:683–9.

11. Oh SH, Kim WY, Kim JH, Younes MN,El-Naggar AK, Myers JN, Kies M, CohenP, Khuri F, Hong WK, Lee HY.Identification of insulin-like growth factorbinding protein-3 as a farnesyl transferaseinhibitor SCH66336-induced negativeregulator of angiogenesis in head and necksquamous cell carcinoma. Clin Cancer Res2006;12:653–61.

12. Desrosiers RR, Cusson MH, Turcotte S,Beliveau R. Farnesyltransferase inhibitorSCH-66336 downregulates secretion ofmatrix proteinases and inhibits carcinomacell migration. Int J Cancer 2005;114:702–12.

13. Oh SH, Jin Q, Kim ES, Khuri FR, Lee HY.Insulin-like growth factor-I receptorsignaling pathway induces resistance to theapoptotic activities of SCH66336(lonafarnib) through Akt/mammaliantarget of rapamycin-mediated increases insurvivin expression. Clin Cancer Res 2008;14:1581–9.

14. Han JY, Oh SH, Morgillo F, Myers JN,Kim E, Hong WK, Lee HY. Hypoxia-inducible factor 1alpha and antiangiogenicactivity of farnesyltransferase inhibitorSCH66336 in human aerodigestive tractcancer. J Natl Cancer Inst 2005;97:1272–86.

15. Lee HY, Chun KH, Liu B, Wiehle SA,Cristiano RJ, Hong WK, Cohen P, KurieJM. Insulin-like growth factor bindingprotein-3 inhibits the growth of non-smallcell lung cancer. Cancer Res 2002;62:3530–7.

16. Oh SH, Lee OH, Schroeder CP, Oh YW,Ke S, Cha HJ, Park RW, Onn A, HerbstRS, Li C, Lee HY. Antimetastatic activity ofinsulin-like growth factor binding protein-3in lung cancer is mediated by insulin-likegrowth factor-independent urokinase-typeplasminogen activator inhibition. MolCancer Ther 2006;5:2685–95.

17. Sousa SF, Fernandes PA, Ramos MJ.Farnesyltransferase inhibitors: a detailedchemical view on an elusive biologicalproblem. Curr Med Chem 2008;15:1478–92.

18. Liu M, Bryant MS, Chen J, Lee S, YaremkoB, Lipari P, Malkowski M, Ferrari E,Nielsen L, Prioli N, Dell J, Sinha D, et al.Antitumor activity of SCH 66336, an orallybioavailable tricyclic inhibitor of farnesylprotein transferase, in human tumorxenograft models and wap-ras transgenicmice. Cancer Res 1998;58:4947–56.

19. Feldkamp MM, Lau N, Roncari L, Guha A.Isotype-specific Ras.GTP-levels predict theefficacy of farnesyl transferase inhibitorsagainst human astrocytomas regardless ofRas mutational status. Cancer Res 2001;61:4425–31.

20. Whitley BR, Beaulieu LM, Carter JC,Church FC. Phosphatidylinositol 3-kinase/Akt regulates the balance betweenplasminogen activator inhibitor-1 andurokinase to promote migration of SKOV-3 ovarian cancer cells. Gynecol Oncol 2007;104:470–9.

21. Dunn SE, Torres JV, Oh JS, Cykert DM,Barrett JC. Up-regulation of urokinase-typeplasminogen activator by insulin-likegrowth factor-I depends uponphosphatidylinositol-3 kinase and mitogen-activated protein kinase kinase. Cancer Res2001;61:1367–74.

22. Kohl NE, Mosser SD, deSolms SJ, GiulianiEA, Pompliano DL, Graham SL, Smith RL,Scolnick EM, Oliff A, Gibbs JB. Selective

inhibition of ras-dependent transformationby a farnesyltransferase inhibitor. Science1993;260:1934–7.

23. Xu W, Neckers L. Targeting the molecularchaperone heat shock protein 90 providesa multifaceted effect on diverse cellsignaling pathways of cancer cells. ClinCancer Res 2007;13:1625–9.

24. Oh SH, Woo JK, Yazici YD, Myers JN,Kim WY, Jin Q, Hong SS, Park HJ, SuhYG, Kim KW, Hong WK, Lee HY.Structural basis for depletion of heat shockprotein 90 client proteins by deguelin.J Natl Cancer Inst 2007;99:949–61.

25. Awada A, Eskens FA, Piccart M, CutlerDL, van der Gaast A, Bleiberg H, WandersJ, Faber MN, Statkevich P, Fumoleau P,Verweij J. Phase I and pharmacologicalstudy of the oral farnesyltransferaseinhibitor SCH 66336 given once daily topatients with advanced solid tumours. EurJ Cancer 2002;38:2272–8.

26. Morgillo F, Lee HY. Lonafarnib in cancertherapy. Expert Opin Investig Drugs 2006;15:709–19.

27. Lee HY, Moon H, Chun KH, Chang YS,Hassan K, Ji L, Lotan R, Khuri FR, HongWK. Effects of insulin-like growth factorbinding protein-3 and farnesyltransferaseinhibitor SCH66336 on Akt expression andapoptosis in non-small-cell lung cancercells. J Natl Cancer Inst 2004;96:1536–48.

28. O’Meara SJ, Kinsella BT. The effect of thefarnesyl protein transferase inhibitorSCH66336 on isoprenylation and signallingby the prostacyclin receptor. Biochem J2005;386:177–89.

29. Lebowitz PF, Davide JP, Prendergast GC.Evidence that farnesyltransferase inhibitorssuppress Ras transformation by interferingwith Rho activity. Mol Cell Biol 1995;15:6613–22.

30. Hensen EF, De Herdt MJ, Goeman JJ,Oosting J, Smit VT, Cornelisse CJ,Baatenburg de Jong RJ. Gene-expression ofmetastasized versus non-metastasizedprimary head and neck squamous cellcarcinomas: a pathway-based analysis.BMC Cancer 2008;8:168.

31. Chun KH, Lee HY, Hassan K, Khuri F,Hong WK, Lotan R. Implication of proteinkinase B/Akt and Bcl-2/Bcl-XL suppressionby the farnesyl transferase inhibitorSCH66336 in apoptosis induction insquamous carcinoma cells. Cancer Res2003;63:4796–800.

32. Brodt P, Samani A, Navab R. Inhibition ofthe type I insulin-like growth factorreceptor expression and signaling: novelstrategies for antimetastatic therapy.Biochem Pharmacol 2000;60:1101–7.

33. Hanrahan EO, Kies MS, Glisson BS, KhuriFR, Feng L, Tran HT, Ginsberg LE,

Can

cerCellBiology



Truong MT, Hong WK, Kim ES. A phaseII study of Lonafarnib (SCH66336) inpatients with chemorefractory, advancedsquamous cell carcinoma of the headand neck. Am J Clin Oncol 2009;32:274–9.

34. Wu X, Zhao H, Do KA, Johnson MM,Dong Q, Hong WK, Spitz MR. Serumlevels of insulin growth factor (IGF-I) andIGF-binding protein predict risk of secondprimary tumors in patients with head andneck cancer. Clin Cancer Res 2004;10:3988–95.

35. Papadimitrakopoulou VA, Brown EN, LiuDD, El-Naggar AK, Jack Lee J, Hong WK,Lee HY. The prognostic role of loss ofinsulin-like growth factor-binding protein-3 expression in head and neckcarcinogenesis. Cancer Lett 2006;239:136–43.

36. Ouban A, Muraca P, Yeatman T, CoppolaD. Expression and distribution of insulin-like growth factor-1 receptor in humancarcinomas. Hum Pathol 2003;34:803–8.

37. Brady G, O’Regan E, Miller I,Ogungbowale A, Kapas S, Crean SJ. Serumlevels of insulin-like growth factors (IGFs)and their binding proteins (IGFBPs), -1, -2,-3, in oral cancer. Int J Oral MaxillofacSurg 2007;36:259–62.

38. Sohda M, Kato H, Miyazaki T, NakajimaM, Fukuchi M, Manda R, Fukai Y, MasudaN, Kuwano H. The role of insulin-likegrowth factor 1 and insulin-like growthfactor binding protein 3 in humanesophageal cancer. Anticancer Res 2004;24:3029–34.

39. Bahr C, Groner B. The insulin like growthfactor-1 receptor (IGF-1R) as a drug target:

novel approaches to cancer therapy.Growth Horm IGF Res 2004;14:287–95.

40. Garcia-Echeverria C. Medicinal chemistryapproaches to target the kinase activity ofIGF-1R. IDrugs 2006;9:415–9.

41. Haluska P, Shaw HM, Batzel GN, Yin D,Molina JR, Molife LR, Yap TA, RobertsML, Sharma A, Gualberto A, Adjei AA, deBono JS. Phase I dose escalation study ofthe anti insulin-like growth factor-Ireceptor monoclonal antibody CP-751,871in patients with refractory solid tumors.Clin Cancer Res 2007;13:5834–40.

42. Tolcher AW, Sarantopoulos J, Patnaik A,Papadopoulos K, Lin CC, Rodon J,Murphy B, Roth B, McCaffery I, GorskiKS, Kaiser B, Zhu M, et al. Phase I,pharmacokinetic, and pharmacodynamicstudy of AMG 479, a fully humanmonoclonal antibody to insulin-like growthfactor receptor 1. J Clin Oncol 2009;27:5800–7.

43. Ekman S, Frodin JE, Harmenberg J,Bergman A, Hedlund A, Dahg P, AlvforsC, Stahl B, Bergstrom S, Bergqvist M.Clinical Phase I study with an Insulin-likeGrowth Factor-1 receptor inhibitor:experiences in patients with squamousnon-small cell lung carcinoma. Acta Oncol2011;50:441–7.

44. Schmitz S, Kaminsky-Forrett M, Henry S,Zanetta S, Geoffrois L, Bompas E, MoxhonA, Guigay J, Machiels JH. Phase II study offigitumumab in patients with recurrentand/or metastatic squamous cell carcinomaof the head and neck: GORTEC 2008–02. JClin Oncol 2010;28:15s (suppl; abstr 5500).

45. Patel S, Pappo A, Crowley J, Reinke D,Eid J, Ritland S, Chawla S, Staddon A,

Maki R, Vassal G, Helman L. A SARCglobal collaborative phase II trial of R1507,a recombinant human monoclonalantibody to the insulin-like growth factor-1receptor in patients with recurrent orrefractory sarcomas. J Clin Oncol 2009;2715s (suppl; abstr 10503).

46. Zha J, Lackner MR. Targeting the Insulin-like Growth Factor Receptor-1R Pathwayfor Cancer Therapy. Clin Cancer Res2010;16:2512–7.

47. Morgillo F, Woo JK, Kim ES, Hong WK,Lee HY. Heterodimerization of insulin-likegrowth factor receptor/epidermal growthfactor receptor and induction of survivinexpression counteract the antitumor actionof erlotinib. Cancer Res 2006;66:10100–11.

48. Chakravarti A, Loeffler JS, Dyson NJ.Insulin-like growth factor receptor Imediates resistance to anti-epidermalgrowth factor receptor therapy in primaryhuman glioblastoma cells throughcontinued activation of phosphoinositide 3-kinase ignaling. Cancer Res 2002;62:200–7.

49. Morgillo F, Kim WY, Kim ES, Ciardiello F,Hong WK, Lee HY. Implication of theinsulin-like growth factor-IR pathway inthe resistance of non-small cell lung cancercells to treatment with gefitinib. ClinCancer Res 2007;13:2795–803.

50. Buck E, Eyzaguirre A, Rosenfeld-FranklinM, Thomson S, Mulvihill M, Barr S,Brown E, O’Connor M, Yao Y, Pachter J,Miglarese M, Epstein D, et al. Feedbackmechanisms promote cooperativity forsmall molecule inhibitors of epidermal andinsulin-like growth factor receptors. CancerRes 2008;68:8322–32.

Can

cerCellBiology

Oh et al. 11


a multiplicity of anti-invasive effects of farnesyl transferase inhibitor sch66336 in human head and...

Documents