crizotinib-resistant npm-alk mutants confer differential sensitivity to unrelated alk inhibitors

2013;11:122-132. Published OnlineFirst December 13, 2012.Mol Cancer Res Monica Ceccon, Luca Mologni, William Bisson, et al. Sensitivity to Unrelated Alk InhibitorsCrizotinib-Resistant NPM-ALK Mutants Confer Differential

Updated Version 10.1158/1541-7786.MCR-12-0569doi:

Access the most recent version of this article at:

MaterialSupplementary

http://mcr.aacrjournals.org/content/suppl/2012/12/13/1541-7786.MCR-12-0569.DC1.htmlAccess the most recent supplemental material at:

Cited Articles http://mcr.aacrjournals.org/content/11/2/122.full.html#ref-list-1

This article cites 41 articles, 14 of which you can access for free at:

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

SubscriptionsReprints and

[email protected] atTo order reprints of this article or to subscribe to the journal, contact the AACR Publications

[email protected]

To request permission to re-use all or part of this article, contact the AACR Publications Department at

American Association for Cancer Research Copyright © 2013 on February 28, 2013mcr.aacrjournals.orgDownloaded from

Published OnlineFirst December 13, 2012; DOI:10.1158/1541-7786.MCR-12-0569

http://mcr.aacrjournals.org/lookup/doi/10.1158/1541-7786.MCR-12-0569

http://mcr.aacrjournals.org/content/suppl/2012/12/13/1541-7786.MCR-12-0569.DC1.html

http://mcr.aacrjournals.org/content/11/2/122.full.html#ref-list-1

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

mailto:[email protected]

mailto:[email protected]

http://mcr.aacrjournals.org/

http://www.aacr.org/

Cell Death and Survival

Crizotinib-Resistant NPM-ALK Mutants Confer DifferentialSensitivity to Unrelated Alk Inhibitors

Monica Ceccon1, Luca Mologni1, William Bisson3, Leonardo Scapozza3, and Carlo Gambacorti-Passerini1,2

AbstractThe dual ALK/MET inhibitor crizotinib was recently approved for the treatment of metastatic and late-stage

ALKþ NSCLC, and is currently in clinical trial for other ALK-related diseases. As predicted after other tyrosinekinase inhibitors' clinical experience, the first mutations that confer resistance to crizotinib have been described inpatients with non–small cell lung cancer (NSCLC) and in one patient inflammatorymyofibroblastic tumor (IMT).Here, we focused our attention on the anaplastic large cell lymphoma (ALCL), where the oncogenic fusion proteinNPM-ALK, responsible for 70% to 80% of cases, represents an ideal crizotinib target. We selected andcharacterized 2 human NPM-ALKþ ALCL cell lines, KARPAS-299 and SUP-M2, able to survive and proliferateat different crizotinib concentrations. Sequencing of ALK kinase domain revealed that a single mutation becamepredominant at high crizotinib doses in each cell line, namely L1196Q and I1171N in Karpas-299 and SUP-M2cells, respectively. These mutations also conferred resistance to crizotinib in Ba/F3 cells expressing human NPM-ALK. The resistant cell populations, as well as mutated Ba/F3 cells, were characterized for sensitivity to twoadditional ALK inhibitors: the dual ALK/EGFR inhibitor AP26113 and NVP-TAE684. While L1196Q-positivecell lines were sensitive to both inhibitors, cells carrying I1171N substitution showed cross-resistance to all ALKinhibitors tested. This study provides potentially relevant information for the management of patients with ALCLthat may relapse after crizotinib treatment. Mol Cancer Res; 11(2); 122–32. �2012 AACR.

IntroductionThe interest in anaplastic lymphoma kinase (ALK) as an

effective oncogenic target is rapidly growing. ALK over-expression (1, 2) or mutations (3–6) were shown to beoncogenic events in neuroblastoma, whereas several onco-genic fusion proteins involving ALK kinase domain werefound in a broad range of solid and hematologic tumors.Among these, the fusion protein NPM-ALK is present inabout 70% to 80% of ALK-positive anaplastic large celllymphomas (ALCL), an aggressive nonHodgkin T-cellLymphoma currently treated with polychemotherapy and,sometimes, bone marrow transplantation. Moreover,

fusions between TPM3-ALK or TPM4-ALK are foundin approximately 50% of inflammatory myofibroblastictumor (IMT), whereas 5% of patients with non–small celllung cancer (NSCLC) express the EML4-ALK fusion (7).Recently, ALK was discovered to be involved at low fre-quencies in several diseases, such as extramedullary plasma-cytoma (8), renal cell carcinoma (9), thyroid cancer (10),breast cancer, and colorectal cancer (11, 12) enlarging thepopulation of patients that will possibly benefit from ALKinhibition.NPM-ALK has been originally recognized as the leading

cause of ALCL by Morris and colleagues in 1994 (13). Thisfusion protein originates from the t(2;5)(p23;q35) translo-cation, and is composed of the first 117 amino acids ofNPM1 protein, followed by a completely functional C-terminal part of ALK (residues 1,058–1,620), that compriseits catalytic domain. The NPM1 part maintains its dimer-ization domain and shuttling activity (14), and its function iscrucial for homodimerization and consequent ALK-deregu-lated activation (15). After dimerization, similar to otherkinases belonging to the insulin receptor family, all 3tyrosines from the well conserved YXXXYY motifs in theactivation loop are activated by transautophosphorylation.In particular, tyrosine 1278 plays an important role inactivation and signal transduction (16). The main ALKdownstream effectors are JAK3-STAT3 (17, 18) andPI3K-AKT (19) pathways, involved in cell survival andphenotypic changes, and ERK pathway (20), leading tocellular proliferation. Also, the JAK2-STAT5 pathway

Authors' Affiliations: 1Department of Health Sciences, University ofMilano-Bicocca; 2Section of Haematology, San Gerardo Hospital, Monza,Italy; and 3Pharmaceutical Biochemistry, Section des sciences pharma-ceutiques, Universit�e de Gen�eve, Geneva, Switzerland

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

Current address for W. Bisson: Environmental Health Sciences Center,Department of Environmental and Molecular Toxicology, Oregon StateUniversity, Portland, Oregon.

M. Ceccon and L. Mologni contributed equally to the work.

Corresponding Author:Monica Ceccon, Department of Health Sciences,University of Milano-Bicocca, Via Cadore 48, Monza 20900, Italy. Phone:3902-6448-8362; Fax: 3902-6448-8363; E-mail:[email protected]

doi: 10.1158/1541-7786.MCR-12-0569

�2012 American Association for Cancer Research.

MolecularCancer

Research

Mol Cancer Res; 11(2) February 2013122





seems to be involved in NPM-ALK–mediated cellular pro-liferation, although the role of STAT5A and STAT5B is stillcontroversial (21, 22)Crizotinib (PF-02341066, Xalkori) was recently approved

for advancedALKþNSCLC and is nowundergoing phase IIclinical trials for ALK-positive diseases different fromNSCLC. However, despite encouraging results (23–25),several ALK kinase domain mutations able to confer crizo-tinib resistance were found in patients withNSCLC (26–29)and IMT (30). This is not surprising, as already shown byclinical experiencewith other tyrosine kinase inhibitors, suchas imatinib in chronic myeloid leukaemia (31). Secondgeneration ALK inhibitors are currently under clinical inves-tigation, for example, AP26113 developed by ARIAD (32),currently in phase I/II clinical trial (NCT01449461),LDK378 by NOVARTIS (33) now in phase I clinical trial(NCT01283516), and ASP3026 by ASTELLAS PHARMA(34), structurally related to NVP-TAE684, which is also inphase I trials (NCT01284192 and NCT01401504). Asexpected, interest in new drugs that may be able to overcomecrizotinib resistance is growing fast.In the present work, we focused our attention on ALK

kinase domain mutations that may confer crizotinib resis-tance in ALCL, a rare nonHodgkin lymphoma that affectsboth adult and pediatric patients, currently treated withstandard polychemotherapy or bone marrow transplanta-tion. Two human ALCL ALKþ cell lines, Karpas-299(K299) and SUP-M2, were grown in the presence ofincreasing doses of crizotinib, thus obtaining different poly-clonal populations able to survive and to proliferate at well-defined drug doses. We found 2 mutations that becamepredominant at high crizotinib doses and, after their intro-duction in Ba/F3 model, we studied their biologic effects,exploring also sensitivity to 2 unrelated ALK inhibitors.Computational studies were run to provide a possibleexplanation for the experimental results.

Materials and MethodsCell culture and selection of resistant cellsALCL T-cell lymphoma cell lines Karpas-299 and SUP-

M2 carrying the t(2;5) translocation and the pro-B murineBa/F3 cell line were purchased from DSMZ, where they areroutinely verified using genotypic and phenotypic testing toconfirm their identity.Cells were cultured in RPMI-1640 supplemented with

9% FBS U.S. origin (Euroclone EUS0080966), 100 U/mLpenicillin, 100 mg/mL streptomycin, and 2 mmol/LL-glutamine (Invitrogen) and incubated at 37�C with 5%CO2. Ba/F3 cells' medium was supplemented with Chinesehamster ovary cells' supernatant (1:2,000) as a source ofinterleukin (IL)-3.Crizotinib was kindly provided by Pfizer and added to the

medium starting from an initial dose of 50 nmol/L.Mediumwas replaced with fresh RPMI-1640 supplemented withcrizotinib every 48 or 72 hours, and cell number and viabilitywere assessed by Trypan Blue count. Once the new cell linewas established, medium was replaced and a higher crizo-tinib dose was added.

AP26113 was kindly provided by ARIAD and NVP-TAE684 was purchased from Selleck Chemicals.

Site-directed mutagenesis and Ba/F3 cell transfectionThe pcDNA3.0 vector containing wild-type NPM/ALK

(pcDNA3-NA) was kindly provided by Dr. S. W. Morris (StJude Research Hospital, Memphis, TN). Site-directed muta-genesis was conducted usingQuikChange IIXLSite-DirectedMutagenesis Kit (Stratagene) and pcDNA3-NA as a template,according to manufacturer instructions, with the followingoligonucleotide sequences: CAATCCCTGCCCCGGTT-CATCCTGCAGGAGCTCATGGCGGG to insert the4539T!A mutation and GGAAGCCCTGATCAAT-AGCAAATTCAACCACCAGAAC to reproduce 4464-65TC!AT double mutation. Ba/F3 cells were transfectedby electroporation (270V; 0.975 mF) and selected first withG418 (Euroclone) 1 mg/mL, followed by IL-3 withdrawal.

TopoCloning and direct sequencingNPM-ALK sequence was amplified by quantitative real-

time (qRT)-PCR and reaction efficiency was checked byagarose gel. NPM-ALK clonal sequencing was conductedusing the TOPO TA Cloning Kit for Sequencing (Invitro-gen), according to manufacturer's instructions. Plasmid wasrecovered from 30 clones per cell line using Zyppy plasmidMiniprep Kit (Zymo Research). Correct insertion was ver-ified by EcoRI (Roche) enzymatic digestion, and only clonesthat displayed the expected restriction map were sequenced.Sequence numbering is related toGenBank IDNM004304.

PCR and quantitative RT-PCRThe cells were lysed in EUROGOLD TRIFAST solution

(Euroclone), and total RNA was extracted according tomanufacturer's instructions. RNA was then retrotranscribedusing TaqMan Reverse Transcription reagents (Roche).Quantitative RT-PCR (qRT-PCR) was conducted usingBrilliant III Sybr Green Mastermix (Agilent Technologies),according to the manufacturer's instructions. The followingprimers were used: NPM-FW, TGCATATTAGTGGA-CAGCAC and ALK-REV, CAGCTCCATCTGCATG-GCTTG. Standard PCRwas conducted usingHigh FidelityTaq Polymerase (Roche), according to the manufacturer'sinstructions, with forward TGCATATTAGTGGACAG-CAC and reverse CAGAGGTCTCTGTTCGAGTC pri-mers. Normalization was carried out on the housekeepinggene glyceraldehyde-3-phosphate dehydrogenase (GAPDH)and amplified using the following primers: GAPDH_SybrGFW, tgcaccaccacctgcttagc; GAPDH_SybrG_REV, GGCA-TGGACTGTGGTCATGAG.

Western blotting and antibodiesCells were seeded in 12-well plate and compounds were

added at different concentrations. After 4-hour treatmentwith the indicated drug, cells were harvested, washed once inPBS at 4�C, and resuspended in lysis buffer [50 mmol/LTris-HCl (pH 7.4), 1% Triton X-100, 5 mmol/L EDTA,150 mmol/L NaCl, 1 mmol/L Na3VO4, 1 mmol/L NaF, 1mmol/L phenylmethylsulfonyl fluoride, and protease

Prediction of Crizotinib-Resistant Mutations

www.aacrjournals.org Mol Cancer Res; 11(2) February 2013 123





inhibitor cocktail (10 mmol/L benzamidine-HCl and 10 mg/mL each of aprotinin, leupeptin, and pepstatin A)] followedby incubation on ice for 30 minutes. After centrifugation toremove cell debris, Laemmli buffer supplemented with 10%b-mercaptoethanol was added and lysates were denatured at97�C for 20 minutes and used for electrophoresis. Equalamounts of total proteins were loaded on 10% SDS-PAGE,transferred to nitrocellulose membrane Hybond ECL(Amersham), and incubated overnight at 4�C with primaryantibody. Secondary horseradish peroxidase–conjugatedanti-mouse or anti-rabbit antibodies (Amersham) were incu-bated for 1 hour and then visualized by chemiluminescenceas recommended by the manufacturer.Anti-ACTIN antibody was purchased from Sigma; poly-

clonal anti-STAT3 is from Calbiochem; monoclonal anti-phospho-ALK (Y-1604), monoclonal anti-ALK (31F12)andmonoclonal anti-phospho-STAT3 (Tyr 705) antibodieswere from Cell Signaling Technology

Proliferation assayCells (5,000/well) were seeded in 96-well plates and

exposed to serial dilutions (1:3) of the compounds, startingfrom 10 or 1mmol/L concentration.Medium alone was usedas a control. After 72 hours of incubation, 3H-methylthymidine (Perkin Elmer) was added to the medium (1mCi/well) and incubated for 8 hours at 37�C. The cells werethen harvested onto glass fiber filters using a Tomtecautomated cell harvester and [3H]-thymidine incorporationwas measured using a filter scintillation counter (1430MicroBeta).

Apoptosis and MTS assayCells were treated for 72 hours at indicated doses and then

were harvested, washed with PBS, and resuspended intoAnnexin V binding buffer. Approximately, 106 cells werestained with Annexin V and propidium iodide (PI) andanalyzed by Flow Cytometry (Bender Medsystems). A smallaliquot was plated for MTS assay using the CellTiter 96AQueous One Solution Cell Proliferation Assay kit (Pro-mega), according to instructions.

Molecular modelingThe three-dimensional coordinates of wild-type (WT)

human ALK catalytic domain bound to crizotinib andNVP-TAE684 and of the L1196M human ALK catalyticdomain bound to crizotinib were retrieved from the ProteinData Bank (PDB codes: 2XP2, 2XB7, and 2YFX; refs. 35–37). The human L1196Q-ALK models were prepared bymutation of the desired amino acid followed by localminimization using Tripos force field (SYBYL8.0, Tripos).For this operation, the L/M1196 side chain orientationcoming from WT and L1196M crystal structures wereused to include in the study both possible Q1196 rotamers(here referred to as Q-R1 and Q-R2). Molecular dockingwith rigid protein was run with the program Surflex-Dockimplemented in SYBYL8.0 (Tripos). Docking scores werecalculated with the consensus-scoring program CScoreimplemented in SYBYL8.0 (Tripos; ref. 38).

Statistical analysesDose–response curves were analyzed using GraphPad

Prism4 software. IC50 indicates the concentration ofinhibitor that gives half-maximal inhibition. Densitome-try values are (ALK treated/ALK untreated) divided by(ACTIN treated/ACTIN untreated) or (P-ALK treated/P-ALK untreated) divided by (ALK treated/ALK untreated).Normalized IC50 was calculated as the IC50 of the targetcell line divided by the IC50 of the parental cell line.Therapeutic index was the ratio between IC50 of Ba/F3parental cell line and IC50 of WT or mutant NPM-ALK–expressing Ba/F3. MTS assay results were analyzed byStudent unpaired t test.

ResultsEstablishment of crizotinib-resistant cell linesTo establish human ALCL cell lines resistant to cri-

zotinib, we cultured SUP-M2 and K299 in the presenceof increasing doses of crizotinib. The selection startedusing a concentration close to the IC90 value (50 nmol/Lfor both cell lines) and gradually reached a final concen-tration of 1 mmol/L for K299 and 0.3 mmol/L for SUP-M2. At each step increase, after an initial drop in cellviability and absolute number (Supplementary Fig S1), astable population able to live and proliferate in thepresence of crizotinib was selected and named as CR(crizotinib resistant) followed by the crizotinib concen-tration [in mmol/L units] in which cells were growing. Ateach step, we collected and analyzed cells. To checkwhether resistance was due to NPM-ALK amplificationor stabilization, NPM-ALK expression levels were inves-tigated both at the transcriptional level (Table 1), usingqRT-PCR, and at the protein level (Fig. 1A). While amoderate increase in NPM-ALK mRNA expression levelwas noted, we could not detect any significant increase ofNPM-ALK protein, except for the K299CR03 cell line,in which ALK expression is quantified by densitometry as4-fold higher than in K299 parental cell line. However,this increase in protein level was no more detectable incell lines selected at higher crizotinib doses. Next, we

Table 1. ALK mRNA expression was measuredby qRT-PCRand normalized onparental cell lineexpression level

Normalized ALKexpression

K299 1K299CR01 1.7K299CR03 2.68K299CR06 4.19K299CR1 4.58SUP-M2 1SUPM2CR02 3.76SUPM2CR03 1.52

Ceccon et al.

Mol Cancer Res; 11(2) February 2013 Molecular Cancer Research124





evaluated cell proliferation rates in the presence of cri-zotinib for each cell line (Fig. 1B). As expected, IC50values increased in proportion to the amount of crizotinibin which cell lines were selected, except for K299CR1(Table 2). We also checked for NPM-ALK activity in thepresence of increasing doses of crizotinib, by analyzing itsphosphorylation status. We found that in all resistant celllines, NPM-ALK phosphorylation was detectable at drugconcentrations higher than in the parental cell line, andthe increase was proportional to the resistance index(Supplementary Fig. S2A). To assess if ALK intrinsicactivity in our resistant cell lines is higher than the oneobserved in parental cells, we measured ALK phosphor-ylation status of untreated cells by densitometry and thenwe normalized this value on ALK expression levels (Sup-plementary Fig. S2B). In two cell lines, K299CR1 andSUPM2CR03, we observed a slight increase in ALKphosphorylation signal (2.17- and 1.84-fold, respective-

ly). To investigate the presence of point mutations and toassess their frequency within the resistant populations, wesequenced the NPM-ALK kinase domain. Interestingly,we found that in both cell lines, a single NPM-ALKkinase domain mutation became predominant at highdoses of crizotinib. In K299 cells, the 4539T!A trans-version, leading to L1196Q substitution, reached afrequency of 86% at 0.6 mmol/L concentration(K299CR06), whereas in SUPM2CR03 cell line, thedouble 4464-65TC!AT mutation, responsible for theI1171N substitution, was present in 100% of the ana-lyzed clones (Table 3 and Supplementary Fig. S3)

Human cell lines carrying L1196Q and I1171Nmutations show different sensitivity to other ALKinhibitorsTo further investigate the role of mutations found at

higher crizotinib doses, we focused our attention on thosecell lines expressing the relevant mutation at higher frequen-cy, that is, K299CR06 and SUPM2CR03. We confirmedboth cell lines to be resistant to crizotinib and then wechallenged them with 2 additional ALK inhibitors: the dualALK/EGFR inhibitor AP26113 and the diaminopyrimidineNVP-TAE684. We evaluated cellular proliferation in thepresence of increasing drug concentrations (Fig. 2A). IC50values obtained are summarized in Table 4. In K299CR06cells, bearing at high frequency the L1196Q substitution,the IC50 values observed with AP26113 and NVP-TAE684were 4- and 13-fold higher than in parental cells, respec-tively, but the absolute values were still low nanomolar (<20nmol/L). In SUPM2CR03 cells, bearing the I1171N sub-stitution, the IC50 values for AP26113 and NVP-TAE684were 112 nmol/L and 52 nmol/L, respectively, that is 8- and10.4-fold higher than in the parental cell line (Table 4). Wetested NPM-ALK autophosphorylation by Western blotanalysis in the presence of increasing doses of all inhibitors,confirming cell proliferation data (Fig. 2B). We also assessed

Figure 1. Human cell linessurviving at different crizotinibconcentrations were characterized.A, ALK expression at protein levelwas assessed by Western blotanalysis and quantified bydensitometry, after normalization onactin expression level. B, cell lineswere cultured for 72 hours in thepresence of different crizotinibconcentrations. Then, proliferationrate was measured as 3H-thymidineincorporation level. Each curvederives from 2 independentexperiments. Corresponding IC50

values are summarized in Table 2.

K299A

B

K299

CR01

K299

CR1

SUP-M2K299

CR03

SUP-M2

CR02

SUP-M2

CR03

1.52.0911

ALK

Actin

Crizotinib Crizotinib

Log (concentration)

1.5

1.0

0.5

0.0–5 –4 –3 –2 –1 0 1 2

% Inhib

itio

n

Log (concentration)

1.5

1.0

0.5

0.0–5 –4 –3 –2 –1 0 1

SUP-M2SUPM2CR02SUPM2CR03

K299

K299CR01K299CR03K299CR06

K299CR1

2

% Inhib

itio

n

1.02 3.95 1.74 1.6

K299

CR06

Table 2. IC50 values obtained by 3H-thymidineincorporation test for each crizotinib-resistanthuman cell line are summarized

Cell line IC50 (mmol/L) Normalized

K299 0.023 1.0K299CR01 0.27 11.7K299CR03 0.45 19.6K299CR06 0.948 41.2K299CR1 0.726 31.6SUP-M2 0.056 1.0SUPM2CR01 0.26 4.6SUPM2CR02 0.25 4.5SUPM2CR03 0.388 6.9

NOTE: Each value derives from 2 independent experiments.







the inhibition of STAT3, an ALK downstream target, aftereach drug administration at different doses. In NPM-ALK–positive cells, STAT3 phosphorylation is ALK-dependent

and recapitulates ALK activation. Moreover, ALK phos-phorylation status was further analyzed in K299CR06 cellsin the presence of lower doses of AP26113 and NVP-

Table 3. All mutations identified, corresponding substitutions and relative frequencies are presented

Cell line Dose (mmol/L) Mutation Substitution Frequency (%)

K299 0.1 mmol/L 4539T!A L1196Q 12.54485A!G N1178S 34596C!T P1215L 34936G!A M1328I 3

0.3 mmol/L 4539T!A L1196Q 83.84521T!C L1190P 2.7

0.6 mmol/L 4539T!A L1196Q 85.71 mmol/L 4539T!A L1196Q 58

SUP-M2 0.2 mmol/L 4464–65TC!AT I1171N 500.3 mmol/L 4464–65TC!AT I1171N 100

Log (concentration)

1.5

1.0

0.5

0.0

µmol/L 0 0.1

P-ALK

ALK

Actin

P-STAT3

STAT3

P-ALK

ALK

Actin

P-STAT3

STAT3

P-ALK

ALK

Actin

6

5

4

3

2

1

0

1201101009080706050403020100

CTRL

Crizotinib 0.3 µmol/L

NPV-TAE684 0.02 µmol/L

AP26113 0.02 µmol/L

P-STAT3

STAT3

Fo

ld in

du

ctio

n

% M

TS

0.3 1 3 0 0.1 0.3 1 3 0 0.1 0.3 1 3 0 0.1 0.3 1 3

–5 –4 –3 –2 –1 0 1 2

1.5

1.0

0.5

0.0–5 –4 –3 –2 –1 0 1 2

1.5

1.0

0.5

0.0–5 –4 –3 –2 –1 0 1 2

% in

hib

itio

n

K299K299CR06SUP-M2SUPM2CR03

K299

Crizotinib AP26113 NPV-TAE684

Cri

zoti

nib

AP

2611

3N

PV-

TAE

684

K299CR06 SUP-M2 SUPM2CR03

K299 K299CR06 SUP-M2 SUPM2CR03 K299 K299CR06 SUP-M2 SUPM2CR03

A

B

C D

Figure 2. Human cell lines carryingdominant NPM-ALK mutationswere characterized. A, dose–response curves obtained by 3H-thymidine incorporation assayafter 72-hour treatment withcrizotinib, AP26113, and NVP-TAE684. IC50 values related aresummarized in Table 4. B, ALK andSTAT3 phosphorylation status wasassessed by Western blot analysisafter 4-hour incubation withdifferent concentrations ofcrizotinib, AP26113, or NVP-TAE684. C, apoptosis wasquantified after 72-hour treatmentwith crizotinib, AP26113, andNVP-TAE684 using Annexin V–PIstaining and analyzed by flowcytometry. D, after 72 hours ofincubation with crizotinib,AP26113, and NVP-TAE684, cellviability was analyzed by MTSassay. K299CR06 viability wassignificantly reduced afterAP26113 (P < 0.0001) and NVP-TAE684 (P ¼ 0.0008) exposure butnot in response to crizotinib,whereas SUPM2CR03 were notaffected by any kind of treatment.Values obtained derive from atleast 2 independent experiments.

Ceccon et al.






TAE684, and we observed a complete loss of NPM-ALKphosphorylation at 10 nmol/L AP26113 and NVP-TAE684, in agreement with values obtained by thymidineincorporation rate (Supplementary Fig. S4). We then stud-ied induction of apoptosis in the presence of 300 nmol/Lcrizotinib, 20 nmol/L AP26113, or 20 nmol/L NVP-TAE684, using ANNEXIN V–PI staining and extendedthe analysis to cell viability at the same drug doses usingMTSassay. We found that in K299CR06 cells treated withcrizotinib apoptosis was comparable with the untreatedcontrol, whereas significant apoptosis was induced in the

K299 parental line (P < 0.05). In contrast, K299CR06 cellstreated with AP26113 or NVP-TAE684 were as sensitive asthe parental cells. SUPM2CR03 survived in the presence ofall drugs (Fig 2C). Interestingly, no apoptosis was noted inSUPM2CR03 cells even at higher concentrations ofAP26113 and NVP-TAE684 (100 nmol/L; SupplementaryFig. S5).MTS assay confirmed Annexin data, (Fig. 2D), andthese results are consistent with the resistance profiles pre-viously found.Together, these data indicate that the L1196Qmutation is

highly resistant to crizotinib but sensitive to AP26113 and

Table 4. IC50 values, expressed in mmol/L, obtained for K299CR06 and SUPM2CR03 cell lines and thesame values normalized on parental cell line, are summarized

Crizotinib AP26113 NVP-TAE684

IC50 Norm IC50 Norm IC50 Norm

K299 0.014 1.00 0.004 1.00 0.001 1.00K299CR06 0.651 46.50 0.016 4.00 0.019 12.88SUP-M2 0.057 1.00 0.014 1.00 0.005 1.00SUPM2CR03 0.388 6.81 0.112 8.00 0.052 10.40

NOTE: All values shown derive from at least 2 independent experiments.

Figure 3. Characterization of Ba/F3cell lines stably transfected with WTor mutated human NPM-ALK. A, celllines were cultured for 72 hours inthe presence of crizotinib, AP26113,and NVP-TAE684 at differentconcentrations; 3H-thymidineincorporation was then measured.Curves derive from at least 2independent experiments.Corresponding IC50 values aresummarized in Table 5. B, ALKphosphorylation status in thepresence of increasing doses ofcrizotinib, AP26113, and NVP-TAE684. Activation of the main ALKdownstream effector STAT3 is alsoshown. ACTINwas used as a loadingcontrol.

1.25

1.00

0.75

0.50

0.25

0.00–5–6 –4 –3 –2 –1 0 1 2

1.25

1.00

0.75

0.50

0.25

0.00–5–6 –4 –3 –2 –1 0 1 2

1.25

1.00

0.75

0.50

0.25

0.00–5 –4 –3 –2 –1 0 1 2

% I

nh

ibitio

n

Crizotinib AP26113 NPV-TAE684

Ba/F3 NA WTBa/F3 NA L1196QBa/F3 NA I1171NL1196MBa/F3

µmol/L 0 0.1 0.3 1 3 0 0.1 0.3 1 3 0 0.1 0.3 1 3 0 0.1 0.3 1 3

WT L1196Q I1171N L1196M

Log (concentration)

A

BP-ALK

ALK

Actin

P-STAT3

STAT3

P-ALK

ALK

Actin

P-STAT3

STAT3

P-ALK

ALK

Actin

P-STAT3

STAT3

Cri

zoti

nib

AP

2611

3N

PV-

TAE

684







NVP-TAE684, whereas the I1171N mutant is resistant tothese drugs.

Effects of crizotinib, AP26113, and NVP-TAE684 in Ba/F3 cells carrying NPM-ALKL1196Q and NPM-ALKI1171NmutantsTo evaluate the actual biologic role of L1196Q and

I1171N mutations out of the context where they wereidentified, we studied their effects in the Ba/F3 cellularmodel transfected with human NPM-ALK. We introducedboth mutations in Ba/F3 NPM-ALK cell lines (Supplemen-tary Fig. S6) and we selected 2 new cell lines, named Ba/F3NA L1196Q and Ba/F3 NA I1171N. Both new cell lineswere able to grow in the absence of IL-3 in the medium andshowed comparable NPM-ALK expression levels (data notshown). We evaluated their proliferation when exposed tocrizotinib, AP26113, and NVP-TAE684 (Fig. 3A). All IC50values and relative therapeutic indexes are summarizedin Table 5. Both cell lines are resistant to crizotinib, butshow different sensitivity to the 2 other compounds. WhileBa/F3 NA L1196Q are sensitive to both AP26113 andNVP-TAE684, with an IC50 of 9 nmol/L, Ba/F3 NAI1171N are resistant to both drugs, with an IC50 of 329and 759 nmol/L, respectively. We also checked the phos-phorylation status of ALK and its downstream effectorSTAT3 after treatment with several drug doses (Fig 3B).Resistance to crizotinib was confirmed for both cell lines,whereas sensitivity to AP26113 andNVP-TAE684 differed:in Ba/F3 NA L1196Q, phosphorylated (p)-ALK andpSTAT3 signals disappear after 4 hours at 0.1 mmol/LAP26113 and NVP-TAE684, whereas in Ba/F3 NAI1171N cells, the band is still present at 0.3 mmol/L ofboth drugs. This finding is consistent with IC50 values

previously shown. As a comparison, the well-known mutantL1196M was analyzed and showed a similar pattern toL1196Q (Fig 3B). These data are in line with results fromthe original human crizotinib-resistant cell lines, K299CR06and SUPM2CR03.

Computational studiesThe results obtained with crizotinib (PF-02341066) and

NVP-TAE684 onWT-, L1196Q-, and I1171N-ALK inBa/F3 cells were investigated computationally. As AP26113structure is not available, we could not extend our analysis tothis drug.The inhibitory trend of crizotinib and NVP-TAE684 in

L1196Q-ALK cells is similar to the one observed with theL1196M mutant. The consistent difference in activitybetween the 2 inhibitors was then studied throughmoleculardocking into the L1196Q-mutant model.The protocols were validated by docking the ligand back

to its binding pocket in the crystal structure. Crizotinibestablished hydrogen bonds with backbone atoms from 2hinge region residues, M1199-NH and E1197-CO, withlow all-atoms root-mean-square deviation (RMSD) com-pared with the crystallographic orientation and with ascore of 7.16 (Table 6). Crizotinib was then docked intothe L1196Q-ALK model using both Q-R1 and Q-R2rotamers (see Material and Methods section). Crizotinibdid not dock into the Q-R1 model, whereas it docked intoQ-R2 model in BM2 mode, with a score of 6.73 (Table 6).Thus, in the case of crizotinib, the possibility of dockinginto the ATP site is clearly affected by the type of Q-rotamer used. In cells, the inhibitor is less active towardsthe L1196Q mutant compared with WT. This might becaused by the flexibility and the size of the gatekeeper

Table 5. IC50 values, expressed in mmol/L, obtained for each Ba/F3 cell line and relative therapeutic index(TI) calculated as IC50 parental line/IC50 mutant, are summarized

CRIZOTINIB AP26113 NVP-TAE684

IC50 Norm T.I. IC50 Norm T.I. IC50 Norm T.I.

Ba/F3 1.261 36.090 1.000 0.820 78.021 1.000 0.760 127.389 1.000Ba/F3 NA WT 0.035 1.000 36.090 0.011 1.000 78.021 0.006 1.000 127.389Ba/F3 L1196Q 0.212 6.076 5.940 0.009 0.856 91.111 0.009 1.509 84.444Ba/F3 I1171N 0.215 6.156 5.862 0.329 31.256 2.496 0.759 127.254 1.001Ba/F3 L1196M 0.459 13.125 2.750 0.042 3.983 19.589 0.018 3.022 42.152

NOTE: IC50 values obtained derive from at least 2 independent experiments.

Table 6. Docking scores of crizotinib and NVP-TAE684 on human WT and L1196Q ALK

WT-ALK (PDB 2XP2) WT-ALK (PDB 2XB7) L1196Q-ALK (Q-R1) L1196Q-ALK (Q-R2)

Crizotinib 7.16 - ND 6.73NVP-TAE684 - 5.67 4.47 5.63

Abbreviation: ND, not defined.

Ceccon et al.






glutamine side chain not favoring energetically the bind-ing of the inhibitor. (Fig. 4B).The same procedure was used with NVP-TAE684 start-

ing with the validation of the protocol. The inhibitor dockedinto WT-ALK model establishing 2 hydrogen bonds withonly one hinge region amino acid, M1199-NH and –CO,with low all-atoms RMSD compared with the crystallo-graphic orientation and with a score of 5.67 (Table 6). Then,NVP-TAE684 was docked in L1196Q-ALK model usingagain Q-R1 and Q-R2 rotamers. In contrast to crizotinib,NVP-TAE684 docked into both models in a BM1 modewith a score of 4.47 (Q-R1) and 5.63 (Q-R2; Table 6). In thelatter case, the ligand establishes a second hydrogen bondbetween its sulphonic group and the side chain of Q1196.(Fig. 4C). In cells, the inhibitor is active against bothWTandmutant forms. This can be explained by thermodynamicallyfavorable binding poses into the ATP site independently ofthe orientation of the gatekeeper glutamine side chain.Residue I1171 is distant from the ATP binding site and

thus, mutations were not studied with molecular dockingbecause they were unable to affect WT docking outputs. Inaddition, for both inhibitors the activity in cells against this

mutant is similar. Hence, we tried to find out a commoncause, not related to the structure of the ligand, explainingthe observed decrease in the biologic response. The hydro-phobic residue I1171, together with C1182, F1271 (DGFmotif), H1247 (HRDmotif), and the polar D1311 (F-helix)define the hydrophobic R-spine (regulatory spine) connect-ing the 2 lobes of the kinase (Fig. 4D; refs. 39, 40). Allstructural motifs important in the catalytic cycle are con-nected to the F-helix, which in this case, is represented by theaspartic acid D1311. (39, 40). The R-spine, together withthe salt bridge E1167-K1150, stabilizes the active confor-mation of the ALK (35, 36). According to our model,mutations that alter this hydrophobic assembly, likeI1171N, decrease the stability of the inhibitor-bound DFGin conformation. In addition, the mutation of I1171 to Asnallows the formation of a hydrogen bond between the sidechain ofN1171 and the backbone carbonylCOofV1180, aspreviously reported in literature (36). The novel interactionmaintains the R-spine compact, balancing the disruptiveeffect of the polar Asn residue on the hydrophobic assembly.This will ultimately stabilize the activated conformationwhile destabilizing inhibitors binding. The described

Figure 4. Molecular modeling studies show the interaction between ALK domain bearing L1196Q and I1171N substitutions in combination with crizotinib andNVP-TAE684. A, the position of residues L1196 and I1171 is displayed in the catalytic domain of the human WT-ALK kinase bound to NVP-TAE684(PDB 2XB7). The ligand is displayed as capped sticks and colored by atom type (carbon atoms in green). The protein is displayed as secondary structure.The residues are colored by atom type (carbon atoms in orange) and displayed as capped sticks. B, docking pose of crizotinib in the humanWT-ALK kinase(PDB 2XP2). The ligand is displayed as capped sticks and colored by atom type (carbon atoms in green). The human L1196Q-ALK mutant model isdisplayed as secondary structure. The residue Q1196-R1 is colored by atom type (carbon atoms in orange) and displayed as space fill. C, docking pose ofNVP-TAE684 in the human L1196Q-ALK mutant model. The ligand is displayed as capped sticks and colored by atom type. The protein is displayed assecondary structure. The residue Q1196-R2 is colored by atom type (carbon atoms in orange) and displayed as capped sticks. H-bonds are representedby black dashed lines, respectively, between the donor and the acceptor. D, the R-spine of the catalytic domain of the human WT-ALK kinasebound toNVP-TAE684 (PDB2XB7) is shown. Theprotein is displayed as secondary structure. TheR-spine residues are colored by atom type anddisplayed asspace fill.







destabilizing event takes place independently of the structureand binding mode of the inhibitor. This explains bothcrizotinib (BM2 mode) and NVP-TAE684 (BM1 mode)decreased activity in I1171N-ALK cells.

DiscussionIn the presentwork, we selected humanALKþ lymphoma

cell lines resistant to different concentrations of crizotinib, adual ALK and MET inhibitor developed by Pfizer. Crizo-tinib has recently been approved for the treatment ofadvanced lung cancer and is in clinical trial for otherALK-related diseases. We preferred human cell lines overthe broadly used Ba/F3 cells for our screening, because theyrepresent a more relevant biologic model of the disease.Analysis of resistant cell lines and following cloning anddirect sequencing of NPM-ALK kinase domain showed aninteresting scenario, where in both cell lines, a single mutantclone becomes predominant at higher crizotinib doses. Aftersequencing, we found L1196Q in 85.7% of screenedK299CR06 clones and I1171N in 100% of SUPM2CR03clones. The frequency of the L1196Q clone diminished inthe K299CR1 population (58%), maybe because othermechanisms of resistance involving different effectors wererandomly selected while subculturing. In fact, detailedanalysis of K299CR1population revealed that 33%of clonescarry WT NPM-ALK kinase domain, compared with only14% of K299CR06 clones (data not shown). However, wecan exclude gene amplification and mRNA or proteinstabilization as causes of resistance (Table 1 and Fig. 1A).We can also exclude that observed resistance was related toenhanced ALK intrinsic activity, despite we observed a slightincrease in basal ALK phosphorylation status (Supplemen-tary Fig. S2B). We could not select a SUP-M2 cell line atcrizotinib concentrations higher than 0.3 mmol/L. Wehypothesize that the selected I1171N substitution wassufficient to confer resistance at this dose, but not at higherdoses. This hypothesis is consistent with the work of Zhangand colleagues who, in a different cellular model (Ba/F3EML4-ALK), could observe I1171N appearance, but thissubstitution disappeared at doses more than 720 nmol/L(41). Whether such mutation will have a clinical signifi-cance, remains to be verified. However, if we refer toimatinib experience in CML context, some mutations witha low in vitro resistance profile are indeed clinically relevant(42, 43), so we cannot exclude that, despite I1171N muta-tion was found at a relatively low crizotinib dose, it may havea significant role in the clinic. Of note, the reported plasmaconcentration of crizotinib at the recommended dose (250mg twice daily) is 0.57 mmol/L (44). From a patient'sviewpoint, I1171N-mutant cells may respond to a crizotinibdose increase. Leucine 1196 is the so called "gatekeeper"residue (45); it is a part of the N-lobe binding to ATP, and itis located in the vicinity of the inhibitor (Fig. 4A). In ALK,mutation of L1196 to methionine (L1196M) leads toclinical forms of cancer drug resistance (26). As reported inliterature, crizotinib is less active against the L1196Mmutant. In contrast, NVP-TAE684 maintains the sameactivity as on the WT kinase (41). In our screening, a

different mutation of the gatekeeper was selected by crizo-tinib, namely a L1196Q substitution. The same mutant waspredicted in a previous mutagenesis screen using Ba/F3 cellscarrying human EML4-ALK oncogene at 500 nmol/L doseof crizotinib (41). We found that the L1196Q mutant wasresistant to crizotinib but was inhibited by both AP26113and NVP-TAE684. Although K299CR06 showed 1-logshift in IC50 upon NVP-TAE684 treatment compared withparental cells, they were still sensitive to low nanomolarconcentrations of the compound. On the other hand, Ba/F3NA L1196Q cells were as sensitive as WT cells, even inproliferation assays. Therefore, we conclude that theL1196Q mutant is inhibited both by AP26113 andNVP-TAE684, as described for L1196M (41, 46). In silicodocking analysis revealed that crizotinib, but not NVP-TAE684, binding was profoundly influenced by the spatialorientation adopted by the glutamine side chain. Therefore,binding of crizotinib may be thermodynamically disfavored.The second mutation identified in our resistant cells was anIle to Asn change at position 1171. The same I1171Nmutation has been found as an activating event in patientswith neuroblastoma (6). Isoleucine in position 1171 is partof the hydrophobic spine of the ALK kinase domain,together with C1182, F1271, H1247, and D1311 (Fig4D; refs. 36, 40). This hydrophobic structural motif hasbeen recognized as an important feature of tyrosine kinases.Molecular studies led to the hypothesis that the high resis-tance observed toward all inhibitors in our NPM-ALKI1171N cell lines may be due to an alteration of the kineticsof ALK structural plasticity, independently from the struc-ture of the inhibitor and its binding mode. Mutation ofhydrophobic Ile to a polar Asn residue is likely to bedestabilizing for the R-spine and for the conformation thatis bound by the inhibitors. However, the additional hydro-gen bond with V1180 will possibly compensate and con-tribute to maintain an activated state of the kinase. Alter-natively, as proposed by Bossi and colleagues (36), themutation may have no effect on inhibitor binding, butsimply increase catalytic efficiency: according to this model,the N1171-V1180 hydrogen bond stabilizes the R-spineand, adds to the E1167-K1150 salt bridge, and shifts thethermodynamic equilibrium toward a fully active confor-mation. In neuroblastoma, I1171N is an activating muta-tion and thus, the plasticity of the ALK-KD plays a majorrole in favoring a particular conformation of the mutant. Inany case, I1171N is predicted to modify the structure of thekinase-inhibitor complex. More detailed computationalstudies need to be conducted to increase our understandingof I1171N and other resistant ALK-mutant forms distantfrom the ATP site.AP26113 is a dual ALK and EGFR inhibitor developed by

ARIAD now in phase I/II clinical trial (NCT01449461).Preclinical data presented at AACR 2010 by Zhang andcolleagues (32) and Rivera and colleagues (47) were prom-ising, showing a potent, selective, and orally active com-pound. AP26113 may soon become a second-line therapyfor ALK-related diseases. While cells carrying L1196Qsubstitution are slightly less sensitive to AP26113 compared

Ceccon et al.






with the parental cell lines, the IC50 values are still lownanomolar (Tables 4 and 5) so we consider these cell lines assensitive to the drug, as was also shown by cell death assaysin Fig. 2C and 2D. In contrast, cells carrying the I1171Nmutation are resistant to AP26113.Moreover, this mutationshowed persistent ALK phosphorylation inWestern blottingand protected from cell death. Unfortunately, AP26113structure is not available. So, we could not conduct compu-tational studies to explain the molecular mechanisms under-lying our biologic results. However, these data provideimportant information to guide the oncologist towards themost effective therapeutic option for those patients carryingI1171N mutation.In conclusion, we selected 2 mutations within the ALK

kinase domain of NPM-ALK that confer resistance tocrizotinib, using human ALKþ lymphoma cell lines as amodel. While L1196Q was sensitive to crizotinib-unrelatedcompounds AP26113 and NVP-TAE684, I1171N wasresistant to both drugs, encouraging or excluding all relatedcompounds as possible therapeutic options in case of relapseafter treatment with crizotinib in patients with ALCL.Future clinical correlates will be necessary to assess theclinical relevance of these 2 ALK mutations.

Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.

Authors' ContributionsConception and design: M. Ceccon, L. Mologni, W. Bisson, L. Scapozza, C.Gambacorti-PasseriniDevelopment of methodology: M. Ceccon, W. BissonAcquisition of data (provided animals, acquired and managed patients, providedfacilities, etc.): M. CecconAnalysis and interpretation of data (e.g., statistical analysis, biostatistics, compu-tational analysis): M. Ceccon, L. Mologni, W. Bisson, L. ScapozzaWriting, review, and/or revision of the manuscript: M. Ceccon, L. Mologni, W.Bisson, L. Scapozza, C. Gambacorti-PasseriniAdministrative, technical, or material support (i.e., reporting or organizing data,constructing databases): L. ScapozzaStudy supervision: L. Mologni, C. Gambacorti-Passerini

AcknowledgmentsThis work was supported by the Italian Association for Cancer Research

(AIRC), Fondazione Cariplo, the Italian Ministry of Health (Programma Integratodi Oncologia), and Lombardy Region: NEDD (Network Enabled Drug Design)and ASTIL.

The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be herebymarked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.

Received September 28, 2012; revisedNovember 14, 2012; acceptedNovember 20,2012; published OnlineFirst December 13, 2012.

References1. Lamant L, Pulford K, Bischof D, Morris SW, Mason DY, Delsol G, et al.

Expression of the ALK tyrosine kinase gene in neuroblastoma. AmJ Pathol 2000;156:1711–21.

2. Passoni L, Longo L, Collini P, Coluccia AM, Bozzi F, Podda M, et al.Mutation-independent anaplastic lymphoma kinase overexpression inpoor prognosis neuroblastoma patients. Cancer Res 2009;69:7338–46.

3. George RE, Sanda T, Hanna M, Fr€ohling S, Luther W, Zhang J, et al.Activating mutations in ALK provide a therapeutic target in neuroblas-toma. Nature 2008;455:975–8.

4. Janoueix-Lerosey I, Lequin D, Brugi�eres L, Ribeiro A, de Pontual L,Combaret V, et al. Somatic and germline activating mutations of theALK kinase receptor in neuroblastoma. Nature 2008;455:967–70.

5. ChenY, Takita J,Choi YL,KatoM,OhiraM,SanadaM,et al.Oncogenicmutations of ALK kinase in neuroblastoma. Nature 2008;455:971–4.

6. Mosse YP, Laudenslager M, Longo L, Cole KA, Wood A, Attiyeh EF,et al. Identification of ALK as a major familial neuroblastoma predis-position gene. Nature 2008;455:930–5.

7. Chiarle R, Voena C, Ambrogio C, Piva R, Inghirami G. The anaplasticlymphoma kinase in the pathogenesis of cancer. Nat Rev Cancer2008;8:11–23.

8. Wang WY, Gu L, Liu WP, Li GD, Liu HJ, Ma ZG. ALK-positive extra-medullary plasmacytoma with expression of the CLTC-ALK fusiontranscript. Pathol Res Pract 2011;207:587–91.

9. DebelenkoLV,RaimondiSC,DawN,ShivakumarBR,HuangD,NelsonM, et al. Renal cell carcinoma with novel VCL-ALK fusion: newrepresentative of ALK-associated tumor spectrum. Mod Pathol2011;24:430–42.

10. Murugan AK, Xing M. Anaplastic thyroid cancers harbor novel onco-genic mutations of the ALK gene. Cancer Res 2011;71:4403–11.

11. Lipson D, Capelletti M, Yelensky R, Otto G, Parker A, Jarosz M, et al.Identification of new ALK and RET gene fusions from colorectal andlung cancer biopsies. Nat Med 2012;18:382–4.

12. Lin E, Li L, Guan Y, Soriano R, Rivers CS, Mohan S, et al. Exon arrayprofiling detects EML4-ALK fusion in breast, colorectal, and non-smallcell lung cancers. Mol Cancer Res 2009;7:1466–76.

13. Morris SW, Kirstein MN, Valentine MB, Dittmer KG, Shapiro DN,Saltman DL, et al. Fusion of a kinase gene, ALK, to a nucleolar proteingene, NPM, in non-Hodgkin's lymphoma. Science 1994;263:1281–4.

14. Borer RA, Lehner CF, Eppenberger HM, Nigg EA. Major nucleolarproteins shuttle between nucleus and cytoplasm. Cell 1989;56:379–90.

15. Bischof D, Pulford K, Mason DY, Morris SW. Role of the nucleophos-min (NPM) portion of the non-Hodgkin's lymphoma-associated NPM-anaplastic lymphoma kinase fusion protein in oncogenesis. Mol CellBiol 1997;17:2312–25.

16. Donella-Deana A,Marin O, Cesaro L, GunbyRH, Ferrarese A, ColucciaAM, et al. Unique substrate specificity of anaplastic lymphoma kinase(ALK): development of phosphoacceptor peptides for the assay of ALKactivity. Biochemistry 2005;44:8533–42.

17. Zamo A, Chiarle R, Piva R, Howes J, Fan Y, Chilosi M, et al. Anaplasticlymphoma kinase (ALK) activates Stat3 and protects hematopoieticcells from cell death. Oncogene 2002;21:1038–47.

18. Chiarle R, Simmons WJ, Cai H, Dhall G, Zamo A, Raz R, et al. Stat3 isrequired for ALK-mediated lymphomagenesis and provides a possibletherapeutic target. Nat Med 2005;11:623–9.

19. Slupianek A, Nieborowska-SkorskaM, Hoser G,Morrione A,MajewskiM, Xue L, et al. Role of phosphatidylinositol 3-kinase-Akt pathway innucleophosmin/anaplastic lymphoma kinase-mediated lymphoma-genesis. Cancer Res 2001;61:2194–9.

20. Marzec M, Kasprzycka M, Liu X, Raghunath PN, Wlodarski P, WasikMA. Oncogenic tyrosine kinase NPM/ALK induces activation of theMEK/ERK signaling pathway independently of c-Raf. Oncogene2007;26:813–21.

21. Ruchatz H, Coluccia AM, Stano P, Marchesi E, Gambacorti-PasseriniC. Constitutive activation of Jak2 contributes to proliferation andresistance to apoptosis in NPM/ALK-transformed cells. Exp Hematol2003;31:309–15.

22. Zhang Q, Wang HY, Liu X, Wasik MA. STAT5A is epigeneticallysilenced by the tyrosine kinase NPM1-ALK and acts as a tumorsuppressor by reciprocally inhibiting NPM1-ALK expression. Nat Med2007;13:1341–8.

23. Gambacorti-Passerini C, Messa C, Pogliani EM. Crizotinib in anaplas-tic large-cell lymphoma. N Engl J Med 2011;364:775–6.

24. Kwak EL, Bang YJ, Camidge DR, ShawAT, Solomon B,Maki RG, et al.Anaplastic lymphoma kinase inhibition in non-small-cell lung cancer.N Engl J Med 2010;363:1693–703.







25. Shaw AT, Yeap BY, Solomon BJ, Riely GJ, Gainor J, Engelman JA,et al. Effect of crizotinib on overall survival in patients with advancednon-small-cell lung cancer harbouring ALK gene rearrangement: aretrospective analysis. Lancet Oncol 2011;12:1004–12.

26. Choi YL, SodaM,Yamashita Y,Ueno T, Takashima J,Nakajima T, et al.EML4-ALK mutations in lung cancer that confer resistance to ALKinhibitors. N Engl J Med 2010;363:1734–9.

27. Sasaki T, Koivunen J,OginoA,YanagitaM,NikiforowS, ZhengW, et al.A novel ALK secondarymutation andEGFR signaling cause resistanceto ALK kinase inhibitors. Cancer Res 2011;71:6051–60.

28. Doebele RC, Pilling AB, Aisner DL, Kutateladze TG, Le AT, WeickhardtAJ, et al. Mechanisms of resistance to crizotinib in patients with ALKgene rearranged non-small cell lung cancer. Clin Cancer Res 2012;18:1472–82.

29. Katayama R, Shaw AT, Khan TM, Mino-Kenudson M, Solomon BJ,Halmos B, et al. Mechanisms of acquired crizotinib resistance in ALK-rearranged lung cancers. Sci Transl Med 2012;4:120ra17.

30. Sasaki T, Okuda K, Zheng W, Butrynski J, Capelletti M, Wang L, et al.The neuroblastoma-associated F1174L ALK mutation causes resis-tance to an ALK kinase inhibitor in ALK-translocated cancers. CancerRes 2010;70:10038–43.

31. Gambacorti-Passerini CB, Gunby RH, PiazzaR, Galietta A, Rostagno R,Scapozza L. Molecular mechanisms of resistance to imatinib in Phila-delphia-chromosome-positive leukaemias. LancetOncol 2003;4:75–85.

32. Zhang S, Wang F, Keats J, Ning Y, Wardwell SD, Moran L, et al.AP26113, a potent ALK inhibitor, overcomes mutations in EML4-ALKthat confer resistance to PF-02341066. In: Proceedings of the 101stAnnual Meeting of the American Association for Cancer Research;2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; CancerRes 2010;70(8 Suppl):Abstract nr LB-298.

33. Li N, Michellys PY, Kim S, Culazzo Pferdekamper A, Li J, Kasibhatla S,et al. Activity of a potent and selective phase I ALK inhibitor LDK378 innaive and crizotinib-resistant preclinical tumor models. In: Proceed-ings of the AACR-NCI-EORTC International Conference: MolecularTargets and Cancer Therapeutics; 2011 Nov 12-16; San Francisco,CA. Philadelphia (PA): AACR; Mol Cancer Ther 2011;10(11 Suppl):Abstract nr B232.

34. Kuromitsu S, Mori M, Shimada I, Kondoh Y, Shindoh N, Soga T, et al.Antitumor activities of ASP3026 against EML4-ALK-dependent tumormodels. In: Proceedings of the AACR-NCI-EORTC International Con-ference:Molecular Targets andCancer Therapeutics; 2011Nov12–16;San Francisco, CA. Philadelphia (PA): AACR;Mol Cancer Ther 2011;10(11 Suppl):Abstract nr A227.

35. Cui JJ, Tran-Dub�e M, Shen H, Nambu M, Kung PP, Pairish M, et al.Structure based drug design of crizotinib (PF-02341066), a potentand selective dual inhibitor of mesenchymal-epithelial transition factor

(c-MET) kinase and anaplastic lymphoma kinase (ALK). J Med Chem2011;54:6342–63.

36. Bossi RT, Saccardo MB, Ardini E, Menichincheri M, Rusconi L, Mag-naghi P, et al. Crystal structures of anaplastic lymphoma kinase incomplex with ATP competitive inhibitors. Biochemistry 2010;49:6813–25.

37. Mctigue M, Deng Y, Liu W, Brooun A. Structure of L1196M mutantanaplastic lymphoma kinase in complex with crizotinib. [Availablefrom: http://www.rcsb.org/pdb/explore.do?structureId=2YFX (PDBCODE: 2YFX)].

38. Kellenberger E, Rodrigo J, Muller P, Rognan D. Comparative evalua-tion of eight docking tools for docking and virtual screening accuracy.Proteins 2004;57:225–42.

39. Kornev AP, Haste NM, Taylor SS, Eyck LF. Surface comparison ofactive and inactive protein kinases identifies a conserved activationmechanism. Proc Natl Acad Sci U S A 2006;103:17783–8.

40. Kornev AP, Taylor SS, Ten Eyck LF. A helix scaffold for the assemb-ly of active protein kinases. Proc Natl Acad Sci U S A 2008;105:14377–82.

41. Zhang S, Wang F, Keats J, Zhu X, Ning Y, Wardwell SD, et al.Crizotinib-resistant mutants of EML4-ALK identified through an accel-erated mutagenesis screen. Chem Biol Drug Des 2011;78:999–1005.

42. Redaelli S, Piazza R, Rostagno R, Magistroni V, Perini P, Marega M,et al. Activity of bosutinib, dasatinib, and nilotinib against 18 imatinib-resistant BCR/ABL mutants. J Clin Oncol 2009;27:469–71.

43. Ernst T, Hochhaus A. Chronic myeloid leukemia: clinical impact ofBCR-ABL1 mutations and other lesions associated with diseaseprogression. Semin Oncol 2012;39:58–66.

44. CostaDB,Kobayashi S, PandyaSS, YeoWL,ShenZ, TanW, et al. CSFconcentration of the anaplastic lymphoma kinase inhibitor crizotinib.J Clin Oncol 2011;29:e443–5.

45. Gunby RH, Ahmed S, Sottocornola R, Gasser M, Redaelli S, MologniL, et al. Structural insights into the ATP binding pocket of theanaplastic lymphoma kinase by site-directed mutagenesis, inhibitorbinding analysis, and homology modeling. J Med Chem 2006;49:5759–68.

46. Rivera VM. AP26113, a potent ALK inhibitor, is also active againstEGFR T790M in mouse models of NCSLC. 14thWorld Conference onLung Cancer. Amsterdam RAI Convention Centre, Amsterdam, TheNetherlands 2011. Presentation #MO11.12

47. Rivera VM, Anjum R, Wang F, Zhang S, Keats J, Ning Y, et al. Efficacyand pharmacodynamic analysis of AP26113, a potent and selectiveorally active inhibitor of Anaplastic Lymphoma Kinase (ALK). In: Pro-ceedings of the 101st Annual Meeting of the American Association forCancerResearch; 2010Apr 17–21;Washington, DC.Philadelphia (PA):AACR; Cancer Res 2010;70(8 Suppl):Abstract nr 3623.

Ceccon et al.






crizotinib-resistant npm-alk mutants confer differential sensitivity to unrelated alk inhibitors

Documents