Treatment Ef
ficacy and Resistance Mechanisms
Using the Second-Generation ALK Inhibitor
AP26113 in Human NPM-ALK
–Positive Anaplastic
Large Cell Lymphoma
M. Ceccon
1, L. Mologni
1, G. Giudici
2, R. Piazza
1, A. Pirola
1, D. Fontana
1, and
C. Gambacorti-Passerini
1,3Abstract
ALK is a tyrosine kinase receptor involved in a broad range of solid and hematologic tumors. Among 70% to 80% of ALKþ anaplastic large cell lymphomas (ALCL) are caused by the aberrant oncogenic fusion protein NPM-ALK. Crizotinib was thefirst clinically relevant ALK inhibitor, now approved for the treatment of late-stage and metastatic cases of lung cancer. However, patients frequently develop drug resistance to Crizo-tinib, mainly due to the appearance of point mutations located in the ALK kinase domain. Fortunately, other inhibitors are available and in clinical trial, suggesting the potential for second-line therapies to overcome Crizotinib resistance. This study focuses on the ongoing phase I/II trial small-molecule tyrosine kinase inhibitor (TKI) AP26113, by Ariad Pharmaceu-ticals, which targets both ALK and EGFR. Two NPM-ALKþ human cell lines, KARPAS-299 and SUP-M2, were grown in the presence of increasing concentrations of AP26113, and
eight lines were selected that demonstrated resistance. All lines show IC50values higher (130 to 1,000-fold) than the parental line. Mechanistically, KARPAS-299 populations resistant to AP26113 show NPM-ALK overexpression, whereas SUP-M2– resistant cells harbor several point mutations spanning the entire ALK kinase domain. In particular, amino acid substitu-tions: L1196M, S1206C, the double F1174VþL1198F and L1122VþL1196M mutations were identified. The knowledge of the possible appearance of new clinically relevant mechan-isms of drug resistance is a useful tool for the management of new TKI-resistant cases.
Implications: This work defines reliable ALCL model systems of AP26113 resistance and provides a valuable tool in the manage-ment of all cases of relapse upon NPM-ALK–targeted therapy.Mol Cancer Res; 13(4); 775–83. 2014 AACR.
Introduction
The pharmacologic inhibition of the ALK (anaplastic lym-phoma kinase) tyrosine kinase has become in the last years an issue of great interest, given its involvement in different malig-nancies and the recent development of several ALK inhibitors. ALK oncogenic role was recognized for thefirst time in the mid-1990s (1), when the NPM-ALK fusion protein was identified as the main cause of a particular subset of anaplastic large cell lymphoma (ALCL). After that, other oncogenic fusion proteins involving the functional ALK kinase domain have been described in different kinds of both solid and hematologic tumors, such as 50% of inflammatory myofibroblastic tumors (IMT; ref. 2), about 5% cases of non–small cell lung cancer
(NSCLC; ref. 3), and at low frequency in other types of tumor, such as rhabdomyosarcoma (RMS; ref. 4), extramedullary plas-macytoma (5), renal cell carcinoma (6), thyroid cancer (7), basal cell carcinoma (8), breast cancer, and colorectal cancer (9, 10). Moreover, aberrant activation or overexpression of full-length ALK has an oncogenic role in neuroblastoma (11, 12) and glioblastoma multiforme (13, 14). The oncogenic properties of aberrantly activated ALK are mainly due to the deregulated activation of different downstream pathways such as RAS/ RAF/MEK/ERK1/2 and PCLg, involved in cellular proliferation, and PI3K/AKT and JAK3/STAT3, that promote cell survival (15– 17). Thefirst clinically available drug able to efficiently target ALK was the dual ALK and MET inhibitor Crizotinib, now approved for the treatment of late-stage and metastatic ALKþ NSCLC and in clinical trial for other ALK-related diseases. The latest clinical data about ALKþ NSCLC patients treated with Crizotinib reveal an increase of median progression-free survival (PFS) and response rate compared with chemotherapy. No significant advantage in overall survival (OS) was observed; however, Crizotinib-treated patients had a better quality of life (18). Very limited data are available on ALCL patients (19). The first long-term follow-up of ALCL patients who received Crizo-tinib reported, after 2 years, a PFS of 63.7% and an OS of 72.7%, and the drug was in general well tolerated (20). However, despite great success, as expected from previous experience with tyrosine kinase inhibitors (TKI; refs. 21, 22), thefirst cases of relapse due
1
Department of Health Science, University of Milano-Bicocca, Monza, Italy. 2Tettamanti Research Centre, Pediatric Clinic, University of Milano-Bicocca, Monza, Italy.3Section of Haematology, San Gerardo Hospital, Monza, Italy.
Note: Supplementary data for this article are available at Molecular Cancer Research Online (http://mcr.aacrjournals.org/).
Corresponding Author: Monica Ceccon, University of Milano-Bicocca, Via Cadore 48, 20900 Monza, Italy. Phone: 39-0264488362; Fax: 39-0264488363; E-mail: mceccon.manuscript@gmail.com
doi: 10.1158/1541-7786.MCR-14-0157
2014 American Association for Cancer Research.
to the positive selection of mutant clones have already been detected in several Crizotinib-treated NSCLC and ALCL patients (20, 23–26). To effectively overcome Crizotinib resistance, the development of second-generation ALK inhibitors is exponen-tially growing (27). Currently, several ALK inhibitors are already in clinical trial: the phase II/III ceritinib (LDK-378, Novartis), the phase I/II alectinib (CH5424802, Roche), and the phase I ASP3026 (Astellas). In this work, we decided to focus our attention on the dual ALK/EGFR inhibitor AP26113 because, after Crizotinib, it was thefirst inhibitor that entered in clinical trials. Preclinical data showed that this inhibitor is 10-fold more potent than Crizotinib and active against the gatekeeper mutant L1196M. In xenograft mouse model, lower AP26113 doses compared with Crizotinib lead to complete tumor disappear-ance (28). Unfortunately, the chemical structure remains undis-closed. In this article, we selected 8 new human NPM-ALKþcell lines, 4 derived from KARPAS-299 and 4 from SUP-M2, able to live and proliferate at high AP26113 doses (K299AR300A, K299AR300B, K299AR300C, K299AR300D, SUP-M2AR500A, SUP-M2AR500B, SUP-M2AR500C, and SUP-M2AR500D). For KARPAS-299–derived cell lines, we observed oncogene over-expression as the main resistance mechanism, whereas in SUP-M2–derived cell lines, we identified several point mutations located within the NPM-ALK kinase domain, which could explain drug resistance. We also found an L1196M mutation in two of four SUP-M2–derived cell lines, but it had no role in conferring resistance at high drug doses. To find a way to overcome AP26113 resistance, we explored the cross-resistance of our KARPAS-299–derived cell lines and mutated Ba/F3 NPM-ALK against other clinically relevant NPM-ALK inhibitors nowadays available, namely Crizotinib, ceritinib, alectinib, ASP3026. KARPAS-299–derived cell lines are highly resistant to all inhi-bitors tested, whereas all mutants studied were targetable with at least a compound, except S1206C. Collectively, our data predict in a human cell-based model the appearance of different mechanisms of resistance to AP26113, and we explored different ways to overcome resistance using a set of clinically relevant ALK inhibitors. This kind of knowledge is a powerful tool to manage clinical cases of Crizotinib and AP26113 relapse.
Materials and Methods
Cell lines, inhibitors, and selection of AP26113-resistant cell lines
The human NPM-ALKþ KARPAS-299 and SUP-M2 cell lines bearing the t(2;5) translocation and the pro-B murine cell line Ba/F3 were purchased from DSMZ, where they are routinely verified using genotypic and phenotypic testing to confirm their identity. Cell lines were subcultured as previously described (29). AP26113 was kindly provided by Ariad Pharmaceutical and added to the medium initially at 20 nmol/L. Medium was replaced with fresh RPMI-1640 supplemented with AP26113 every 48 or 72 hours, and cell number and viability were assessed by Trypan Blue count. Ba/F3 cells were maintained in RPMI medium supplemented with 10% FBS and CHO cells supernatant (1:2,000) as a source of IL3. Ba/F3 cells were transduced with mutagenized pcDNA3.0-NPM/ALK plasmid (see below) by electroporation, as previously described (29) and selectedfirst with G418 (Euroclone) 2 mg/mL, followed by IL3 withdrawal. Crizotinib was kindly provided by Pfizer, ceritinib (LDK-378) by Novartis, ASP3026 by Astellas, whereas alectinib (CH5424802) was purchased from Selleck Chemicals.
Site-directed mutagenesis
pcDNA3.0 bearing human WT NPM-ALK (pcDNA3.0 NA) was kindly provided by Dr. S.W. Morris (St Jude Research Hospital, Memphis, TN). Site-directed mutagenesis on pcDNA3-NA was performed using the QuikChange II XL Site-Directed Mutagenesis Kit (Stratagene), according to the manufacturer's instructions. Primers used for mutagenesis were as follows: NPM-ALK L1122V FW: 50-ATCACCCTCATTCGGGGTGTGGGCCATGGCGC-30; NPM-ALK S1206C FW: 30 -GAGACCTCAAGTGCTTCCTCCGAGAGA-CCCGCC-50; NPM-ALK L1196M FW: 50 -CTGCCCCGGTTCATCCT-GATGGAGCTCATGGCG-30; NPM-ALK F1174V FW: 30 -GAA-GCCCTGATCATCAGCAAAGTCAACCACCAGAACATTG-50; NPM-ALK L1198F FW: 30 -GTTCATCCTGCTGGAGTTCATGGCGGGGG-GAGAC-50. Reverse primers are reverse complements of FW primers. Sequence numbering is related to GenBank ID NM004304. FISH
Interphase FISH was performed using ALK Dual Color Break Apart Rearrangement Probe (Cytocell). This probe consists of a green 420-kb probe, which spans the majority of the ALK gene and a red 486-kb probe, which is telomeric to the ALK gene. Slides were prepared following standard cytogenetic procedures. Codenaturation/hybridization of the specimen slides and probes were performed at 72C for 2 minutes and at 37C overnight, followed by washing in saline-sodium citrate (SSC) buffer pH 5.3 (Vysis/ABBOTMolecular -Abbott Laboratories. Abbott Park, Illi-nois, U.S.A.) and counter staining with anti-fade solution contain-ing DAPI.
PCR, quantitative RT-PCR, and sequencing
An NPM-ALK fragment encompassing the breakpoint and comprising the whole kinase domain was amplified by PCR using high-fidelity Taq polymerase (Roche), according to instructions. Primers used were FW: 50-TGCATATTAGTGGACAGCAC-30; REV: 50-CTGTAAACCAGGAGCCGTAC-30. PCR products were sent to Eurofins MWG Operon for sequencing. Quantitative real-time PCR (qRT-PCR) was performed as previously described (29). Housekeeping genes used for normalization were murine HPRT FW: 50–TCAGTCAACGGGGGACATAAA-30; REV: 50 -GGGGC-TGTACTGCTTAACCAG-30 or human GAPDH FW: 50 -Tgcaccac-cacctgcttagc-30; REV: 50–GGCATGGACTGTGGTCATGAG–30. Deep sequencing was performed as previously described (20). Exome sequencing
Exome libraries were generated from 1 mg of genomic DNA extracted with a PureLink Genomic DNA kit (Life Technology). Genomic DNA was fragmented to a size of 200 to 500 bp and processed according to the standard protocol for the Illumina TruSeq DNA Sample Preparation Kit. Genomic libraries were then enriched with the Illumina TruSeq Exome Enrichment Kit. Librar-ies were sequenced on an Illumina Genome Analyzer IIx with 76-bp paired-end reads using Illumina TruSeq SBS kit v5.
Bioinformatics
were initially filtered by proper pair, then converted into the binary BAM alignment format, sorted, and indexed. Removal of duplicates was performed using the SAMtools rmdup command. Unique BAMfiles were then converted into the mPileup format. mPileup data generated from paired cancer and control samples were cross-matched using a dedicated in-house software tool. Copy number and allelic imbalance/loss of heterozygosity anal-yses from whole-exome data were performed using CEQer (31). Western blotting and antibodies
Cells were seeded in complete medium in 12-well plate, and compounds were added at different concentrations. After 4 hours, cells were harvested, washed once in PBS at 4C, and resuspended in Laemmli buffer 1 supplemented with 10% b-mercaptoetha-nol (100 mL/106 cells) and denatured at 97C for 20 minutes before electrophoresis. Equal volumes were loaded on 10% SDS-PAGE, transferred to nitrocellulose membrane Hybond ECL (Amersham), and incubated overnight at 4C with primary anti-body (1:1,000 dilution in BSA 2.5%). Secondary horseradish peroxidase–conjugated anti-mouse or anti-rabbit antibodies (Amersham) were incubated for 1 to 2 hours and then visualized by chemiluminescence as recommended by the manufacturer. Monoclonal anti–phospho-ALK (Y-1604), monoclonal anti-ALK (31F12), and monoclonal anti–phospho-STAT3 (Tyr 705) bodies were from Cell Signaling Technology. Anti-ACTIN anti-body was purchased from Sigma; polyclonal anti-STAT3 is from Calbiochem.
Proliferation assay
Five thousand cells per well were seeded in 96-well plates in the presence of serial dilutions (1:2 or 1:3) of each compound, starting from a concentration of 10 mmol/L or 1 mmol/L, based on drug potency and specific cell line sensitivity. Incubation with radioactive-labeled thymidine and radioactivity detection was performed as previously described (29).
Software and statistical analysis
Dose–response curves were analyzed using GraphPad Prism 5 software. IC50indicates the concentration of inhibitor that gives half-maximal inhibition. Densitometry values are calculated as follows: total ALK level for each well is normalized on its own loading control (ALK/ACTIN). Values are the average of at least three independent wells. P-ALK value is normalized both on the untreated sample and on the total ALK level: (P-ALK treated/ P-ALK untreated) divided by (ALK treated/ALK untreated). Rela-tive resistance (RR) index was calculated as the ratio between mutant and WT IC50values (32). qPCR data were analyzed using the DDCt method, normalized on the proper housekeeping gene. ALK kinase domain was drawn using PDB viewer software (PDB code:3LCS).
Results
Establishment and characterization of human AP26113-resistant cell lines
To obtain a reliable resistance model, we grew two different human NPM-ALK–positive cell lines, KARPAS-299 and SUP-M2, in the presence of increasing AP26113 doses (29). We divided each cell line into four differentflasks, assuming that a stochastic selection would occur in each independent population. Cell selection was started at a drug concentration close to the IC90
value, calculated with a preliminary proliferation assay as 13 nmol/L for KARPAS-299 and 19.4 nmol/L for SUP-M2. Each cell line was challenged withfive or six sequential drug doses (Sup-plementary Fig. S1). An independent cell line was established before each step increase of drug concentration. This process was stopped at 300 nmol/L AP26113 for KARPAS-299 and 500 nmol/L for SUP-M2 cells, based on the comparison between IC50values for resistant and parental cells. In our experience, for a highly drug-resistant population, an IC50increase of at least 10-fold compared with parental cells was expected. In fact, all AP26113-resistant cell lines showed a RR index higher than 100 (Table 1). Among all cell lines established, we decided to focus our attention on the highest AP26113 dose–resistant populations, referred to as K299AR300A, K299AR300B, K299AR300C, K299AR300D (AP26113-resistant at 300 nmol/L) and M2AR500A, SUP-M2AR500B, SUP-M2AR500C, SUP-M2AR500D (AP26113-resis-tant at 500 nmol/L). For each cell line, NPM-ALK expression and activation were assessed by Western blot (Fig. 1; Supplementary Fig. S7A) using specific antibodies for total ALK and phosphory-lated ALK (Tyr 1604). Although in parental cells 100 nmol/L AP26113 completely abrogates ALK activation, in all resistant populations, Tyr 1604 phosphorylation is still detectable at 100 or even 300 nmol/L, indicating that drug resistance is due to a mechanism that directly impairs ALK inactivation.
Identification of resistance mechanism
Western blot analysis revealed an increase in NPM-ALK expres-sion, quantified by densitometry as 9.4 2.1, 6.0 1.3, 7.1 1.8, and 6.5 0.5-fold compared with parental cells for K299AR300A, K299AR300B, K299AR300C, and K299AR300D cells, respectively (Supplementary Fig. S2). Notably, in thefirst three cell lines (A, B, and C), the basal phospho-ALK signal was higher than in parental cells (Fig. 1A). This hyperactivation could be simply due to the increased ALK expression rather than to an intrinsic NPM-ALK hyperactivation, as highlighted in Supplementary Fig. S2. To further confirm that in AP26113-resistant cell lines, NPM-ALK targeting was ineffective, the phosphorylation status in Tyr705 of the NPM-ALK downstream effector STAT-3 was also assessed. In all AP26113-resistant KARPAS-derived cell lines, STAT-3 P-Tyr705 was present at higher drug doses than the one observed for parental cells and correlated with the persistent NPM-ALK phos-phorylation (Fig. 1A). qRT-PCR confirmed that oncogene over-expression was present also at transcriptional level, because a 23.7-, 16-, 25.5-, and 5.1-fold increase of NPM-ALK transcript was detected in K299AR300A, B, C, and D, respectively, compared with parental (Fig. 2A; Table 2). Values are obtained upon normalization on the proper housekeeping gene. Of note, protein and mRNA levels correlated. Moreover, a FISH experiment
Table 1. IC50values obtained for AP26113-resistant KARPAS-derived cell lines
and SUP-M2–derived cell lines
Cell line IC50(mmol/L) RR index
revealed that the increase in NPM-ALK overexpression is due to gene amplification (Fig. 2B). Because the IC50value observed for all resistant cell lines is extremely high, we explored the presence of low-frequency point mutations in ALK kinase domain as an additional resistance mechanism by deep sequencing. The results excluded this hypothesis (data not shown). NPM-ALK overexpres-sion in K299AR300D was less evident than in the other cell lines. Moreover, the band corresponding to P-Tyr1604 ALK disappears at 300 nmol/L, suggesting low molecular resistance to AP26113, despite the fact that STAT-3 phosphorylation remains also at high drug doses and the cells RR index was 180. Whole-exome sequenc-ing and copy-number analysis of this cell line highlighted that the
increase in ALK expression was due to NPM-ALK amplification. In addition, some huge copy-number alterations were present. Also missense mutations were detected, but none of them involved proteins clearly related to tumorigenesis or drug resistance (Sup-plementary Fig. S3).
Also in SUP-M2–derived resistant cells, we could observe a great increase in IC50value compared with SUP-M2 parental cell line, and, as expected, this corresponded to a persistent P-Tyr1604
Figure 1.
Human cell lines characterization. Parental or resistant KARPAS-299 (A) or SUP-M2 (B) cell lines were incubated for 4 hours in the presence of increasing AP26113 concentrations: 0, 100, 300, and 1000 nmol/L. P-ALK (Tyr1604), ALK, P-STAT3 (Tyr705), STAT3, andbACTIN expression levels were assessed by Western blot.
Figure 2.
ALK signal (Fig. 1; Supplementary Fig. S7A). No increase in total ALK or in basal ALK phosphorylation was evident in SUP-M2– derived cell lines (Table 2; Figs. 1B and 2A). We speculated that SUP-M2 resistance could be due to the positive selection of point mutations in the NPM-ALK kinase domain, so a fragment com-prising the whole kinase domain from total RNA was amplified after retrotranscription and analyzed by direct sequencing. In all resistant SUP-M2–derived cell lines, we found at least one point mutation as a possible resistance mechanism (Table 3; Fig. 2C). Moreover, STAT-3 was still phosphorylated at the highest AP26113 dose (Fig. 1B), confirming an aberrant upregu-lation of NPM-ALK–driven signaling. To confirm the role of these point mutations in resistance to AP26113, all of them were introduced into the pro-B murine Ba/F3 cell line expressing human NPM-ALK model (Supplementary Fig. S4). The AP26113 IC50value was also calculated for the Ba/F3 cell line together with the assessment of NPM-ALK P-Tyr1604 and STAT3 P-Tyr 705 phosphorylation status in the presence of increasing drug doses (Tables 4 and 5; Fig. 3; Supplementary Fig. S7B). All mutants, except L1196M, showed intermediate to high resistance to AP26113, paralleled by persistent NPM-ALK phosphorylation. These data confirmed an effective role for these mutations in AP26113 resistance, except for the gatekeeper substitution L1196M. Clonal sequencing of SUP-M2AR500A showed the simultaneous presence of two mutations: F1174V and L1198F (Table 6), whereas in SUP-M2AR500B cell line revealed the simultaneous presence of L1196M and at least another substitu-tion, mainly L1122V, found in 80% of clones analyzed, but also D1203N (6.7%; Tables 7). Notably, F1174V alone was unable to confer AP26113 resistance, whereas the double-mutant F1174VþL1198F was highly resistant to the same compound. Similarly, L1122V alone was sufficient to induce AP26113 resis-tance (Table 4 and 5), thus supporting the idea that this was the effective driver mutation, as well as the L1198F for the previous double mutant. Moreover, we could not select, after two
experi-ments, a Ba/F3 cell line expressing the single NPM-ALK D1203N able to survive in the absence of IL3. Moreover, the double L1196MþD1203N mutant is highly AP26113 resistant, so it is possible that both mutations are required for an advantageous selection and drug resistance. We detected also the presence of the P1139S mutation in one clone of 15, but it is not able to confer high AP26113 resistance (IC50¼ 0.015 mmol/L; RR index ¼ 2.14). Interestingly, F1174V and L1198F mutations were not able to induce resistance to AP26113 singularly, but their cooperation was necessary.
Cross-resistance to other ALK inhibitors
To explore a possible way to overcome AP26113 resistance, we tested the sensitivity of KARPAS-derived cell lines and mutated NPM-ALK–expressing Ba/F3 cells to other clinically relevant ALK inhibitors: Crizotinib, ceritinib, alectinib, and ASP3026. As expected, all KARPAS-299–derived cell lines are highly resistant to all inhibitors, according to the proliferation assay (Tables 8 and 9). These data are consistent with the observation that in K299AR300A, B, and C cell lines, a general mechanism of drug resistance, oncogene amplification, has been selected.
We also challenged Ba/F3 NPM-ALK WT and mutated cell lines with all the ALK inhibitors (Tables 4 and 5). Cells carrying the NPM-ALK L1122V mutation are moderately resistant to Crizoti-nib and resistant to alectiCrizoti-nib, ceritiCrizoti-nib, and ASP3026 (RR index is 2.6, 5.3, 9.1, and 4.9, respectively); our data about the gate-keeper-mutant L1196M suggest a moderate resistance to AP26113, Crizotinib, and alectinib (RR¼ 2.1, 3.4, and 2.9), confirming our previous data (29), resistance against ASP3026 and sensitivity to ceritinib. Combination of L1122V with the L1196M substitution increases the resistance values of all drugs, especially of AP26113, thus giving an impressive advantage upon treatment with all compounds that directly inhibit the target. Ba/ F3 NPM-ALK bearing the S1206C substitution are resistant to crizotinib, alectinib, and ceritinib (RR¼ 4.3, 5.7, and 4.1) and highly resistant to ASP3026. The double F1174VþL1198F mutant is resistant to all drugs except crizotinib, with an RR index of 12.2 and 10 for AP26113 and ceritinib and 9.0 and 5.8 for alectinib and ASP3026, respectively. Interestingly, the single F1174V muta-tion is completely sensitive to all drugs, AP26113 included, but
Table 2. Values obtained for NPM-ALK expression by QRT-PCR, in terms of absolute number and fold change compared with parental cells
Cell line mRNA expression value Normalization
K299 0.0222 1.0 KAR300A 0.5261 23.7 KAR300B 0.3546 16.0 KAR300C 0.5653 25.5 KAR300D 0.1129 5.1 SUPM2 0.0134 1.0 SUP-M2AR500A 0.0182 1.4 SUP-M2AR500B 0.0314 2.3 SUP-M2AR500C 0.0405 3.0 SUP-M2AR500D 0.0368 2.7
Table 3. Mutations reported by direct sequencing
Cell line Mutation Substitution SUP-M2AR500A 4472T>Gþ4544C>T F1174VþL1198F SUP-M2AR500B 4316C>Gþ4538C>A L1122VþL1196M SUP-M2AR500C 4538C>A L1196M SUP-M2AR500D 4569C>G S1206C
Table 4. IC50values obtained by proliferation assay for each Ba/F3 NPM-ALK WT or mutagenized cell line are summarized
IC50(mmol/L) AP26113 CRIZOTINIB CH5424802 LDK-378 ASP3026
resistant to alectinib and ASP3026, whereas the single L1198F is per se resistant to all drugs except Crizotinib. Together, these two mutations cooperate in conferring higher resistance to AP26113. Moreover, we explored the cross-resistance of other two muta-tions found at lower frequencies in SUP-M2AR500B cell line, namely P1139S and D1203N. Although P1139S mutant is sen-sitive to all drugs except ceritinib (RR index¼ 3.5), we could not establish an IL3-independent Ba/F3 cell line carrying the single D1203N substitution. On the other hand, the double L1196Mþ D1203N mutant is highly resistant to AP26113 (RR index¼ 33.2), confirming that the latter may be the driver mutation for this compound. Cell proliferation data are confirmed by Western blot analysis (Supplementary Fig. S5).
In conclusion, according to these data, we can foresee that for each mutation, alone or in combination, except S1206C, there is a clinically available ALK inhibitor able to overcome acquired resistance to AP26113.
Discussion
In the last 4 years, the management of ALK-related diseases has successfully changed, thanks to the availability of a new ALK inhibitor, Crizotinib. However, as expected, the problem arising from drug resistance soon became reality, in fact several patients relapsed after Crizotinib treatment, mainly because of the positive selection of point mutations. So, other different ALK inhibitors were developed with the purpose of overcoming Crizotinib resistance. The first second-generation ALK inhibitor was AP26113, a dual ALK and EGFR inhibitor by Ariad Pharmaceu-tical. In this work, we took advantage of a reliable cellular model, based on human NPM-ALKþcell lines, to predict the possible resistance mechanism to AP26113. We could establish eight AP26113-resistant cell lines, 4 derived by KARPAS and 4 from SUP-M2 cell lines (Supplementary Fig. S1). All cell lines were less sensitive to drug targeting than the parental one (Fig. 1; Table 1; Supplementary Fig. S7A), although for K299AR300D this obser-vation was less evident. KARPAS-derived cell lines A, B, and C clearly showed oncogene amplification as the main cause of resistance (Fig. 2A); moreover, we could exclude by deep sequenc-ing of the NPM-ALK fragment comprissequenc-ing the whole ALK kinase domain the presence of point mutations as an additional mechanism. NPM-ALK overexpression was less evident in K299AR300D cells, both at protein and at transcriptional levels, and is clearly due to NPM-ALK amplification. The presence of copy-number alterations spread throughout all the genome may explain the drug resistance. However, further studies will be necessary to clarify this issue (Supplementary Fig. S3). For all
KARPAS-derived cell lines, we could not detect any point muta-tion in NPM-ALK kinase domain. On the other hand, in SUP-M2 cell lines, we detected the presence of point mutations in the NPM-ALK kinase domain that could explain AP26113 resistance: L1122VþL1196M, L1196M, F1174VþL1198F, S1206C. Clonal sequencing of SUP-M2AR500B revealed the presence of other point mutations at lower frequency, namely P1139S and the double L1196MþD1203N (Table 7). We introduced in the broad-ly used murine pro-B cell line Ba/F3 all these mutations, and
Table 5. RR index corresponding to each IC50value is reported
AP26113 CRIZOTINIB CH5424802 LDK-378 ASP3026
WT 1 1 1 1 1 L1122V 8.4 2.6 5.3 9.1 4.9 P1139S 1.5 1.1 0.6 3.5 1.9 F1174V 1.5 1.0 3.7 0.9 4.0 L1196M 2.1 3.4 2.9 1.1 5.0 L1198F 5.8 0.1 12.0 22.2 5.4 S1206C 14.2 4.3 5.7 4.1 17.1 L1122VþL1196M 65.1 7.6 20.5 8.7 25.8 F1174VþL1198F 12.2 0.0 9.0 10.0 5.8 L1196MþD1203N 33.2 6.0 4.2 3.0 1.7 NOTE: Green: RR< 2; yellow: RR ¼ 2.1–4; orange: RR ¼ 4.1–10; red: RR > 10. Data are the average from at least 2 independent experiments.
Figure 3.
NPM-ALK targeting by AP26113 in Ba/F3-mutant cell lines. All cell lines were incubated for 4 hours in the presence of increasing doses of the compound, then ALK P-Tyr 1604, total ALK, STAT3 P-Tyr705, and total STAT3 level are assessed by Western blot.b-Actin was used as loading control.
Table 6. Clonal sequencing of SUPM2AR500A SUP-M2AR500A
further confirmed their effective role in drug resistance. Unfortu-nately, AP26113 structure is not available, so we could not perform any molecular modeling analysis. Residue L1122 is located at the P-loop and extends into the drug binding pocket (Supplementary Fig. S6). To our knowledge, mutations involving this residue have never been observed thus far, neither for Cri-zotinib nor for AP26113 resistance. It is interesting to note that L1122 corresponds to Abl L248, whose substitution with valine or with arginine causes resistance to several TKIs (33). The gatekeeper L1196M was one of thefirst causes of Crizotinib resistance found in patients. A moderate AP26113 resistance was predicted in vitro (29, 34, 35); however, the IC50value is low enough to consider this mutation sensitive to AP26113. Because in SUP-M2AR500B cells both L1122V and L1196M were found, we hypothesize that thefirst is the mutation driving resistance, while the latter was selected at low doses and never counterselected, remaining as a passenger in the high-dose–resistant clone. Topo-cloning per-formed on SUPM2AR500B cell line supported this hypothesis, in fact both mutations are simultaneously present in 12 of 15 clones (80%); moreover, where L1122V is not detected, the D1203N is present (13% of cases), highlighting the weakness of L1196M in conferring AP26113 resistance (Table 7). The fact that we could not establish an IL3-independent Ba/F3 NPM-ALK D1203N cell line means that this mutation alone is disadvanta-geous. However, the double L1196MþD1203N mutant was not only able to grow upon IL3 withdrawal but also showed high AP26113 resistance, thus highlighting the fact that a cooperation between the two single mutations is favorable for drug resistance. Curiously, the double mutant is targetable by ASP3026, whereas the single L1196M is not, but it is moderately resistant or resistant to all other inhibitors. P1139S alone, unique in SUP-M2AR300B clone #9, is sensitive to AP26113 (RR index¼ 1.8); thus, we can speculate that, in this clone, other unknown mechanisms may cooperate in its positive selection. F1174 is located at the end of
the aC helix (Supplementary Fig. S6) and lies in a hydrophobic cluster composed by F1098, F1271, and F1245. Mutations involv-ing F1174 were recognized as activatinvolv-ing in neuroblastoma and phenylalanine substitution with a leucine was found in an IMT patient that relapsed after Crizotinib treatment (24). Clones carrying cysteine, valine, or isoleucine in residue 1174 instead of phenylalanine were selected at AP26113 100 nmol/L and 200 nmol/L in a previous Ba/F3 screen (34). In our screening, Ba/F3 cells bearing the single mutation are completely sensitive to all drugs studied, including AP26113. L1198F alone in our Ba/F3 cells was sufficient to confer resistance to AP26113, alectinib, ceritinib, and ASP3026; however, it was completely sensitive to Crizotinib. A methionine substitution in this position was pre-dicted to confer resistance at AP26113 at 100 nmol/L (34), whereas a proline was predicted to confer crizotinib resistance in an in vitro screening (36). This residue corresponds to the Abl F317, a site that, if mutated, induces resistance to several TKIs. Notably, in our model, L1198F and F1174V cooperate in confer-ring resistance against AP26113. Finally, S1206 is located into the aD helix. A tyrosine substitution was found in an NSCLC patient that relapsed after Crizotinib treatment, whereas substitutions with an arginine, an isoleucine, or a glycine were predicted in vitro as resistant to AP26113. Notably, the S1206C was the only mutation detected at AP26113 500 nmol/L, indicating that this residue has a key role in conferring high AP26113 resistance (34). All data obtained by3H thymidine incorporation test were val-idated by Western blot (Figs. 1 and 3), evaluating both NPM-ALK activation by phosphorylation status of its tyrosine 1605 and its downstream target STAT3 phosphorylation in Tyrosine 705. The pattern of STAT3 phosphorylation recapitulates the one found for NPM-ALK; moreover, in some cases it appears even stronger, likely because NPM-ALK–driven signaling is amplified while trans-duced. Targeting the molecular chaperon HSP90 has been pro-posed as an alternative way to hit NPM-ALK and overcome TKI's resistance, because NPM-ALK is a well-known HSP90 client. For this reason, we tested all our NPM-ALK–overexpressing KARPAS-derived cell lines and Ba/F3 cells bearing all single and double mutations against the HSP90 inhibitor 17-AAG (Supplementary Table S1). AP26113-resistant KARPAS cells seemed to be more sensitive to 17-AAG than to other TKIs, whereas all mutations except the S1206C were sensitive to the inhibitor, and this could be due to the vast heterogeneity of HSP90 clients, because other molecules impaired by HSP90 inhibition may cooperate in cell survival and proliferation. Overall, our cross-resistance experi-ments revealed that, except for S1206C, all point mutations detected may be targeted simply switching to another inhibitor, already available in clinic. In the light of these data, we could speculate that most of the efforts should be directed infinding a new inhibitor, able to target mutations involving the S1206 residue. This knowledge, together with all data nowadays avail-able on Crizotinib resistance, is a useful tool to manage cases of
Table 9. RR index corresponding to each IC50value is reported
CRIZOTINIB CH5424802 LDK-378 ASP3026 K299 1 1 1 1 K299AR300A >10 >10 >10 >10 K299AR300B >10 >10 >10 >10 K299AR300C >10 >10 >10 >10 K299AR300D >10 >10 >10 >10 NOTE: Green: RR< 2; yellow: RR ¼ 2.1–4; orange: RR ¼ 4.1–10; red: RR > 10. Table 7. Clonal sequencing of SUPM2AR500B
SUP-M2AR500B
Clone Mutations Substitutions #1 4316C>G 4538C>A L1122VþL1196M #2 4316C>G 4538C>A 4575T>C L1122VþL1196MþL1208P #3 4316C>G 4538C>A L1122VþL1196M #4 4316C>G 4538C>A L1122VþL1196M #5 4316C>G 4538C>A L1122VþL1196M #6 4316C>G 4538C>A L1122VþL1196M #7 4316C>G 4538C>A 4556G>A L1122VþL1196MþG1202R #8 4316C>G 4538C>A L1122VþL1196M #9 4367C>T P1139S #10 4316C>G 4538C>A L1122VþL1196M #11 4316C>G 4538C>A L1122VþL1196M #12 4316C>G 4538C>A L1122VþL1196M #13 4559G>A 4538C>A L1196MþD1203N #14 4316C>G 4538C>A L1122VþL1196M #15 4559G>A 4538C>A 4593G>A L1196MþD1203NþR1214H
Table 8. IC50values obtained by proliferation assay for each AP26113-resistant
KARPAS-299–derived cell line are summarized
IC50(mmol/L) CRIZOTINIB CH5424802 LDK-378 ASP3026
AP26113 resistance, both for the oncologist and for the drug designer.
Disclosure of Potential Conflicts of Interest
Professor C. Gambacorti-Passerini is principal investigator of the trial A8081013 sponsored by Pfizer. No potential conflicts of interest were disclosed by the other authors.
Authors' Contributions
Conception and design: M. Ceccon, L. Mologni, C. Gambacorti-Passerini Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): M. Ceccon, A. Pirola, D. Fontana
Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): M. Ceccon, L. Mologni, R. Piazza, C. Gambacorti-Passerini
Writing, review, and/or revision of the manuscript: M. Ceccon, L. Mologni, C. Gambacorti-Passerini
Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): G. Giudici
Study supervision: M. Ceccon
Acknowledgments
The authors thank Ariad Pharmaceutical, Pfizer, Astellas, and Novartis that kindly provided all drugs used for this work.
Grant Support
This work was supported by the Lombardy Region (ID14546A). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Received March 21, 2014; revised October 3, 2014; accepted November 12, 2014; published OnlineFirst November 24, 2014.
References
1. Morris SW, Kirstein MN, Valentine MB, Dittmer KG, Shapiro DN, Saltman DL, et al. Fusion of a kinase gene, ALK, to a nucleolar protein gene, NPM, in non-Hodgkin's lymphoma. Science 1994;263:1281–4.
2. Coffin CM, Patel A, Perkins S, Elenitoba-Johnson KS, Perlman E, Griffin CA. ALK1 and p80 expression and chromosomal rearrangements involving 2p23 in inflammatory myofibroblastic tumor. Mod Pathol 2001;14: 569–76.
3. Soda M, Choi YL, Enomoto M, Takada S, Yamashita Y, Ishikawa S, et al. Identification of the transforming EML4-ALK fusion gene in non-small-cell lung cancer. Nature 2007;448:561–6.
4. van Gaal JC, Flucke UE, Roeffen MH, de Bont ES, Sleijfer S, Mavinkurve-Groothuis AM, et al. Anaplastic lymphoma kinase aberrations in rhabdo-myosarcoma: clinical and prognostic implications. J Clin Oncol 2012;30: 308–15.
5. Wang WY, Gu L, Liu WP, Li GD, Liu HJ, Ma ZG. ALK-positive extrame-dullary plasmacytoma with expression of the CLTC-ALK fusion transcript. Pathol Res Pract 2011;207:587–91.
6. Debelenko LV, Raimondi SC, Daw N, Shivakumar BR, Huang D, Nelson M, et al. Renal cell carcinoma with novel VCL-ALK fusion: new representative of ALK-associated tumor spectrum. Mod Pathol 2011;24:430–42. 7. Murugan AK, Xing M. Anaplastic thyroid cancers harbor novel oncogenic
mutations of the ALK gene. Cancer Res 2011;71:4403–11.
8. Ning H, Mitsui H, Wang CQ, Suarez-Farinas M, Gonzalez J, Shah KR, et al. Identification of anaplastic lymphoma kinase as a potential therapeutic target in Basal Cell Carcinoma. Oncotarget 2013;4:2237–48.
9. Lin E, Li L, Guan Y, Soriano R, Rivers CS, Mohan S, et al. Exon array profiling detects EML4-ALK fusion in breast, colorectal, and non-small cell lung cancers. Mol Cancer Res 2009;7:1466–76.
10. Lipson D, Capelletti M, Yelensky R, Otto G, Parker A, Jarosz M, et al. Identification of new ALK and RET gene fusions from colorectal and lung cancer biopsies. Nat Med 2012;18:382–4.
11. Mosse YP, Laudenslager M, Longo L, Cole KA, Wood A, Attiyeh EF, et al. Identification of ALK as a major familial neuroblastoma predisposition gene. Nature 2008;455:930–5.
12. Azarova AM, Gautam G, George RE. Emerging importance of ALK in neuroblastoma. Semin Cancer Biol 2011;21:267–75.
13. Powers C, Aigner A, Stoica GE, McDonnell K, Wellstein A. Pleiotrophin signaling through anaplastic lymphoma kinase is rate-limiting for glio-blastoma growth. J Biol Chem 2002;277:14153–8.
14. Hudson L, Kulig K, Young D, McLendon R, Abernethy A. Abstract "A42: ALK and cMET expression in glioblastoma multiforme: Implications for therapeutic targeting." Abstracts: AACR-NCI-EORTC International Con-ference: Molecular Targets and Cancer Therapeutics—Nov 12–16, 2011; San Francisco, CA.
15. Roskoski R Jr. Anaplastic lymphoma kinase (ALK): structure, oncogenic activation, and pharmacological inhibition. Pharmacol Res 2013;68:68–94. 16. Chiarle R, Simmons WJ, Cai H, Dhall G, Zamo A, Raz R, et al. Stat3 is required for ALK-mediated lymphomagenesis and provides a possible therapeutic target. Nat Med 2005;11:623–9.
17. Chiarle R, Voena C, Ambrogio C, Piva R, Inghirami G. The anaplastic lymphoma kinase in the pathogenesis of cancer. Nat Rev Cancer 2008; 8:11–23.
18. Shaw AT, Kim DW, Nakagawa K, Seto T, Crino L, Ahn MJ, et al. Crizotinib versus chemotherapy in advanced ALK-positive lung cancer. N Engl J Med 2013;368:2385–94.
19. Gambacorti-Passerini C, Messa C, Pogliani EM. Crizotinib in anaplastic large-cell lymphoma. N Engl J Med 2011;364:775–6.
20. Gambacorti Passerini C, Farina F, Stasia A, Redaelli S, Ceccon M, Mologni L, et al. Crizotinib in advanced, chemoresistant anaplastic lymphoma kinase-positive lymphoma patients. J Natl Cancer Inst 2014;106:djt378.
21. Deininger MW, Holyoake TL. Can we afford to let sleeping dogs lie? Blood 2005;105:1840–1.
22. Gambacorti-Passerini CB, Gunby RH, Piazza R, Galietta A, Rostagno R, Scapozza L. Molecular mechanisms of resistance to imatinib in Philadel-phia-chromosome-positive leukaemias. Lancet Oncol 2003;4:75–85. 23. Choi YL, Soda M, Yamashita Y, Ueno T, Takashima J, Nakajima T, et al.
EML4-ALK mutations in lung cancer that confer resistance to ALK inhibi-tors. N Engl J Med 2010;363:1734–9.
24. Sasaki T, Okuda K, Zheng W, Butrynski J, Capelletti M, Wang L, et al. The neuroblastoma-associated F1174L ALK mutation causes resistance to an ALK kinase inhibitor in ALK-translocated cancers. Cancer Res 2010;70: 10038–43.
25. Doebele RC, Pilling AB, Aisner DL, Kutateladze TG, Le AT, Weickhardt AJ, et al. Mechanisms of resistance to crizotinib in patients with ALK gene rearranged non-small cell lung cancer. Clin Cancer Res 2012;18:1472–82.
26. 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.
27. Mologni L. Inhibitors of the anaplastic lymphoma kinase. Expert Opin Investig Drugs 2012;21:985–94.
28. Rivera VM, Anjum R, Wang F, Zhang S, Keats J, Ning Y, et al. Efficacy and pharmacodynamic analysis of AP26113, a potent and selective orally active inhibitor of Anaplastic Lymphoma Kinase (ALK). Washington, D.C.: AACR; 2010.
29. Ceccon M, Mologni L, Bisson W, Scapozza L, Gambacorti-Passerini C. Crizotinib-resistant NPM-ALK mutants confer differential sensitivity to unrelated Alk inhibitors. Mol Cancer Res 2013;11:122–32.
30. Li H, Handsaker B, Wysoker A, Fennell T, Ruan J, Homer N, et al. The Sequence alignment/map format and SAMtools. Bioinformatics 2009;25: 2078–9.
31. Piazza R, Magistroni V, Pirola A, Redaelli S, Spinelli R, Galbiati M, et al. CEQer: a graphical tool for copy number and allelic imbalance detection from whole-exome sequencing data. PLoS One 2013;8:e74825. 32. Redaelli S, Piazza R, Rostagno R, Magistroni V, Perini P, Marega M, et al.
33. Redaelli S, Mologni L, Rostagno R, Piazza R, Magistroni V, Ceccon M, et al. Three novel patient-derived BCR/ABL mutants show different sensitivity to second and third generation tyrosine kinase inhibitors. Am J Hematol 2014;87:E125–8.
34. Zhang S, Wang F, Keats J, Ning Y, Wardwell SD, Moran Lauren, et al. AP26113, a potent ALK inhibitor, overcomes mutations in EML4-ALK that confer resistance to PF-02341066. AACR annual meeting, Apr 17–21, 2010; Washington, DC.
35. Katayama R, Khan TM, Benes C, Lifshits E, Ebi H, Rivera VM, et al. Therapeutic strategies to overcome crizotinib resistance in non-small cell lung cancers harboring the fusion oncogene EML4-ALK. Proc Natl Acad Sci U S A 2011;108:7535–40.
NPM-ALK
–Positive Anaplastic Large Cell
Lymphoma
In this article (Mol Cancer Res 2015;13:775–83), which appeared in the April 2015 issue of Molecular Cancer Research (1), the grant support section was missing a funding source. The corrected grant support section is below. The authors regret the error. This work was supported by the Lombardy Region (ID14546A) and the Italian Association for Cancer Research (AIRC 2013 IG-14249).
Reference
1. Ceccon M, Mologni L, Giudici G, Piazza R, Pirola A, Fontana D, et al. Treatment efficacy and resistance mechanisms using the second-generation ALK inhibitor AP26113 in human NPM-ALK– positive anaplastic large cell lymphoma. Mol Cancer Res 2015;13:775–83.
Published OnlineFirst September 29, 2015. doi: 10.1158/1541-7786.MCR-15-0277
