A Combinatorial Strategy for Targeting › content › ...Darmood Wei4, Sudheer Kumar Gara4, Lisa Zhang5, Ya-qin Zhang6, Min Shen6, Sanjit Mukherjee7, and Electron Kebebew1 ABSTRACT

CLINICAL CANCER RESEARCH | TRANSLATIONAL CANCER MECHANISMS AND THERAPY

A Combinatorial Strategy for TargetingBRAFV600E-Mutant Cancers with BRAFV600E Inhibitor(PLX4720) and Tyrosine Kinase Inhibitor (Ponatinib) A C

Chandrayee Ghosh1, Suresh Kumar2, Yevgeniya Kushchayeva3, Kelli Gaskins4, Myriem Boufraqech4,Darmood Wei4, Sudheer Kumar Gara4, Lisa Zhang5, Ya-qin Zhang6, Min Shen6, Sanjit Mukherjee7, andElectron Kebebew1

ABSTRACT◥

Purpose: Most aggressive thyroid cancers are commonlyassociated with a BRAFV600E mutation. Preclinical and clinicaldata in BRAFV600E cancers suggest that combined BRAF andMEK inhibitor treatment results in a response, but resistance iscommon. One mechanism of acquired resistance is throughpersistent activation of tyrosine kinase (TK) signaling by alter-nate pathways. We hypothesized that combination therapy withBRAF and multitargeting TK inhibitors (MTKI) might be moreeffective in BRAFV600E thyroid cancer than in single-agent orBRAF and MEK inhibitors.

Experimental Design: The combined drug activity wasanalyzed to predict any synergistic effect using high-throughput screening (HTS) of active drugs. Weperformed follow-up in vitro and in vivo studies to validateand determine the mechanism of action of synergisticdrugs.

Results: The MTKI ponatinib and the BRAF inhibitor PLX4720showed synergistic activity by HTS. This combination significantlyinhibited proliferation, colony formation, invasion, and migration inBRAFV600E thyroid cancer cell lines and downregulated pERK/MEKand c-JUN signaling pathways, and increased apoptosis. PLX4720-resistant BRAFV600E cells became sensitized to the combinationtreatment, with decreased proliferation at lower PLX4720 concentra-tions. In an orthotopic thyroid cancer mouse model, combinationtherapy significantly reduced tumor growth (P < 0.05), decreased thenumber of metastases (P < 0.05), and increased survival (P < 0.05)compared with monotherapy and vehicle control.

Conclusions: Combination treatment with ponatinib andPLX4720 exhibited significant synergistic anticancer activity in pre-clinical models of BRAFV600E thyroid cancer, in addition to over-coming PLX4720 resistance. Our results suggest this combinationshould be tested in clinical trials.

IntroductionApproximately 10% to 15% of thyroid cancers have aggressive

disease behavior and are associated with a high disease-specificmortality (1). Anaplastic thyroid cancer (ATC) is one of the mostaggressive and fatal malignancies in humans (2, 3). Treatmentapproaches, including surgery, chemotherapy, radiotherapy, andmul-timodal therapy, rarely result in a durable response or prolongedsurvival in patients with ATC (4–6). Multiple genetic aberrations arecommon in ATC, and a better understanding of the genetic alterationshas led to important and effective treatment advances (6).

Among the different genetic mutations identified in differenthistologic subtypes of thyroid cancer, BRAFV600E has been studiedextensively because it is the most common driver and actionablemutation in thyroid cancer, with additional secondary mutations thatpromote progressive dedifferentiation into ATC (7). Monotherapywith a BRAF inhibitor may result in initial clinical benefits, but this isfor a short duration as acquired resistance is common and leads to asignificant relapse in almost all cases (8). Several mechanisms ofacquired resistance to BRAF inhibitor therapy have been reported,including tyrosine kinase (TK) upregulation, NRASmutation, mutantBRAF amplification or alternative splicing, andMEKmutation (9, 10).

Combination therapy using both a BRAF inhibitor and a MEKinhibitor was initially investigated in metastatic melanoma and wasshown to be more effective compared with BRAF inhibition therapyalone (8). Subsequently, clinical trials combining inhibitors of bothMEK and BRAFwere conducted, as this combination was predicted todelay MAPK-driven acquired resistance, resulting in longer responseduration and a higher tumor response rate (11). As a result of thosestudies, combination therapy with a BRAF and MEK inhibitors forBRAFV600E-mutant ATCwas recently approved by the FDA.However,both BRAF and MEK act on the same downstream target and actlargely on the same pathway; hence, they may not target any alterna-tively activated pathways, such as upregulation of the upstream TKsignaling pathways, which, in turn, may result in acquired resistancefor this combination therapy as well (12). Furthermore, studies haveshown that other cancers treated with combined BRAF and MEKinhibitor treatment, such as melanoma, also develop resistance (13).Thus, approaches using a combination of BRAF andmultitargeted TKinhibitorsmay prove evenmore effective inBRAFV600E thyroid cancer.

1Department of Surgery, StanfordUniversity, Stanford, California. 2Laboratory ofGenetics and Genomics, National Institute of Aging, Bethesda, Maryland.3National Institute of Digestive and Kidney Diseases, NIH, Bethesda, Maryland.4NCI, NIH, Bethesda, Maryland. 5National Institute of Child Health and Devel-opment, NIH, Bethesda, Maryland. 6National Center for Advancing TranslationalSciences, NIH, Bethesda, Maryland. 7Genetics Branch, NCI, Bethesda, Maryland.

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

C. Ghosh and S. Kumar contributed equally to this article.

Corresponding Author: Electron Kebebew, Stanford University School of Med-icine, 300 Pasteur Drive, H3642, Stanford, CA 94305. Phone: 650-725-7830;E-mail: [email protected]

Clin Cancer Res 2020;26:2022–36

doi: 10.1158/1078-0432.CCR-19-1606

�2020 American Association for Cancer Research.

AACRJournals.org | 2022

In this study, we hypothesized that combination therapy with aBRAF inhibitor and a multitargeting TK inhibitor would be moreeffective in BRAFV600E-mutant thyroid cancer than BRAF inhibitionalone. We performed high-throughput screening (HTS) onBRAFV600E-mutant cells with BRAFV600E inhibitors and TK inhibitors.The drug combination with synergistic activity (ponatinib and vemur-afenib/PLX4720) was then studied in vitro and in vivo to determine itsanticancer activity and the mechanism of action for synergism.

Materials and MethodsTheAnimalCare andUseCommittee of theNCI,NIH approved the

animal study protocol. No human sample data were used in this study.

Cell linesHuman thyroid cancer cell lines 8505C and BCPAP cell harboring

the BRAFV600E mutation were purchased from the European Collec-tion of Cell Culture, BRAFWild Type (WT) cell lines, THJ-16T derivedfrom a patient with ATC was a kind gift from Dr. John A. Copland(Mayo Clinic, Jacksonville, FL), and C643 was obtained fromCell LineService. All cell lines were maintained in Dulbecco's Eagle medium(DMEM) containing 4,500 mg/L of D-glucose, 2 mmol/L of L-gluta-mine, and 110mg/L of sodium pyruvate, supplemented with 10% fetalbovine serum (FBS), thyroid stimulating hormone (TSH, 10 mU/M;Sigma), penicillin (10,000 U/mL), streptomycin (10,000 U/mL), fun-gizone (250 ng/mL), and insulin (10 mg/mL) in a standard humidifiedincubator at 37�C in 5% CO2 and 95% O2 atmosphere. All cell lineswere authenticated by short-tandem repeat profiling. Cell lines weretested for Mycoplasma from Idexx BioAnalytics and were foundnegative for any contamination.

Cell proliferation assayCell proliferation assays were performed in 96-well plates in trip-

licate. Cells were plated in 96-well black bottom well plates (GreinerBio-one) at 1.5� 103 cells/well in 100 mL of culture medium. The nextday, cells were supplemented with 100 mL of fresh medium containingdrug(s) PLX4720 and/or ponatinib (Sellekchem), or vehicle dimethylsulphoxide (DMSO).Mediumwith drug/vehiclewas replenished every48 hours. Cell proliferation data were collected for 0, 24, 48, 72, and96 hours after treatment. CyQUANT (Thermo Fisher Scientific) cell

proliferation assays were performed according to the manufacturer'sinstructions. The cell numbers were determined using a fluorescencemicroplate reader (Molecular Devices) at 485 nm/538 nm.

Plasmids and siRNA transfectionspCLXSN-c-JUN retroviral vector was purchased from Addgene. c-

JUN (2 mg) plasmid was transfected into Amp-phoenix cells, aretroviral packing cell line (a gift from Dr. George Xu, University ofPennsylvania), using Fugene6 transfection reagent (Promega). At24 hours after transfection, fresh culture media were added, and thecells were incubated for an additional 48 hours before harvesting theretrovirus-containing media. The cell lines 8505C, BCPAP, C643, andTHJ16T at approximately 50% to 70% confluence were then infectedwith c-JUN containing retrovirus in cell culture media containing 10%FBS for 48 hours at 37�C. The cells were treated with 1 mg/mLpuromycin. The puromycin resistant cell clones were selected forfurther studies.

Combination drug activity calculation methodCombination drug effect on cell proliferation was determined

using CompuSyn software (Version 1.0, http://www.compusyn.com), which is based on the median-effect principle (Chou) andthe combination index-isobologram theorem (Chou–Talalay). Thecombination index (CI) values were calculated, where CI < 1.0indicates a synergistic effect, CI ¼ 1.0 indicates an additive effect,and CI > 1 indicates an antagonistic effect (14). Drug combinationsat nonconstant ratios were used to calculate CI. For calculating thefold change in sensitivity to PLX4720, the IC50 was determined for apanel of BRAFV600E-mutant and BRAFWT cell lines grown in thepresence and absence of ponatinib. Experiments were performedindependently three times.

Apoptosis assayA Caspase-Glo 3/7 assay (Promega) was used to measure caspase

activity. Cells were plated in 96-well plates at a density of 2� 103 cells/well in 200 mL of media in triplicate. After 24 hours, fresh cell culturemedium with the drug(s) or vehicle control was added to each well.After 48 hours, cells were analyzed for caspase-3/7 activity using theCaspase-Glo3/7 assay kit according to the manufacturer's instruction.The relative luminescence was calculated and normalized to the totalnumber of cells.

To detect the drug-induced apoptosis, thyroid cancer cells weretreated with drug(s) or vehicle (48 hours), then the fraction of bothfloating and attached cells were pooled in 1� annexin V binding bufferand resuspended. Next, 400 mL of cell suspension containing 1 � 105

cells were transferred to a 5-mL culture tube. The BioLegend FITCannexin V Apoptosis Detection Kit was used for staining the cells.Further, 5 mL of FITC annexin V and 10 mL of FxCycle Violet(Invitrogen) were added to the cell suspension, vortexed gently, andincubated for 15 minutes at room temperature in the dark beforemeasurement by flow cytometry (FACSCanto II, BDBiosciences) withexcitation blue laser (488 nm). FlowJo software was used for analyzingthe data.

Western blot and antibodiesTotal cell lysates were prepared from the cells after treatment with

the drug(s) or vehicle with 10 mmol/L Tris buffer (pH 7.4) and 1%sodium dodecyl sulfate (SDS). The protein concentration was deter-mined using the Pierce BCA assay kit solution (Thermo FisherScientific). An equal amount of proteins was resolved by electropho-resis on 4%–15% SDS-PAGE gradient gels and transferred to PVDF

Translational Relevance

Most aggressive thyroid cancers are commonly associated with aBRAFV600E mutation. Recent preclinical and clinical findings onBRAF-mutant cancers suggest that combination therapy withBRAF andMEK inhibitors may result in a response, but resistanceis common, leading to disease progression. We used quantitativehigh-throughput screening on BRAFV600E-mutant cells withBRAFV600E inhibitors and tyrosine kinase inhibitors for identifyingnew therapeutic combinations and found PLX4720 showed goodsynergism with ponatinib. The synergistic activity of this combi-nation was validated on cellular proliferation, colony formation,invasion, and migration assays and was seen to induce apoptosis.In vivo study also showed the effectiveness of this drug combinationin regressing tumor and lung metastasis in an orthotopic model.Our preclinical evaluation of PLX4720 and ponatinib shows thatthis novel combination is a potential candidate for a clinical trial inBRAFV600E thyroid cancer.

Targeting BRAFV600E Cancers with Novel Combination Agents

AACRJournals.org Clin Cancer Res; 26(8) April 15, 2020 2023

membrane, then immunostained overnight using specific antibodies(Supplementary Method). All primary (1:1,000) and secondary anti-bodies (1:5,000) were purchased from Cell Signaling Technologies.Band densitometry analysis was performed using ImageJ software(NCI, Bethesda, MD).

Gene-expression analysisTotal mRNA was extracted from the cells after treatment with the

drug(s) or vehicle using an RNeasy Kit (Qiagen) and was used forreverse transcription to make cDNA (Applied Biosystems). Quan-titative real-time PCR was performed using reagents from Applied

Biosystem and TaqMan probes (JUN, FOS, Thermo FisherScientific).

Cell migration and invasion assayThe cell migration and invasion assays were performed according to

the manufacture's instruction (BD Biosciences). Briefly, the cells wereseeded in triplicate into 6-well plates (5� 103 cells/well); after 24 hours,cells were treated with the drug(s) and vehicle for 24 hours. The cellswere trypsinized, and then 5 � 104 cells were seeded in 500 mL totalvolume of serum-free medium in the upper chamber of the transwellplate. In the bottom well, 750 mL of DMEM supplemented with 10%

Figure 1.

HTS matrix drug screen identifies synergy between BRAF inhibitor and TK inhibitor in ATC 8505c cell line. A, a 10 � 10 combination response illustration plotrepresenting drug synergy screen. DrugA is added to each column from left to right, concentration from high to low. Drug B is added to each row from top to bottom,concentration from high to low. B, Complete dose–response curves for vemurafenib and ponatinib in the 8505c cell line. Single-agent vemurafenib curve is shown inred, single-agent ponatinib curve is shown in blue, other curves represent dose–response from different combinations. C, 10� 10 matrix plot for the combination ofvemurafenib (0.0007–15 mmol/L) and ponatinib (0.0007–15 mmol/L) in cell viability assay after data normalization in 0–100 scale. 100 means strong cell killing.D,Data are shown in excess HSA (highest single-agent) format. Greenmeans strong synergistic effect with excess HSA≥ 20, cyanmeansweak synergistic effect withexcess HSA < 20, yellow and red indicate possible antagonistic effect. The mean excess HSA and BLISS score for 8505C was 21.143 and 20.355, respectively.(Continued on the following page.)

Ghosh et al.

Clin Cancer Res; 26(8) April 15, 2020 CLINICAL CANCER RESEARCH2024

FBS was added to act as a chemoattractant. Cells were incubated at37�C for 22 hours, at which point the cells that migrated and invadedwere fixed and stained using the DiffiQuick Staining Kit (VWR). Cellswere then photographed and analyzed using ImageJ software (NCI,Bethesda, MD).

Clonogenic assayCells were seeded in triplicate in collagen precoated 12-well

plates (5 � 102 cells/well) and allowed to adhere overnight inculture media. The cells were then cultured with the drug(s) aloneor in combination, or with the vehicle in complete media for 12 to

Figure 1.

(Continued.) E–H, Effect of PLX4720 and ponatinib on cellular proliferation. PLX4720 (5, 10, 15 mmol/L), ponatinib (0.062, 0.185, 0.556 mmol/L), or combinations ofthe two for up to 96hourswere analyzed for cell viability. TheX axis represents the elapsed time in hours and theY axis represents the relativefluorescence unit. I,Thesynergistic effect of combination PLX4720 and ponatinib treatment onBRAFV600E thyroid cancer cell is depicted as an isobologram of drug combinations. The X axisrepresents the drug concentration, and the Y axis represents their combination effect at FA of 0.5 (50% reduction of cell growth). The red and blue plots are theresponse of the single agents, and the combination plots (green) indicate synergistic effect when cell growth inhibition was greater than �50% (for fractionalinhibition of Fa¼ 0.50–0.90). Synergistic responses are between the two lines and above Fa¼ 0.5 on the Y axis. PLX¼ PLX4720; PTB¼ ponatinib in all the figures.

Targeting BRAFV600E Cancers with Novel Combination Agents

AACRJournals.org Clin Cancer Res; 26(8) April 15, 2020 2025

14 days. Growth media with vehicle or drug(s) were replaced every48 hours. The cells were fixed with 0.4% buffered paraformaldehydeand then stained with 0.5% crystal violet in methanol for 10 minutes.The colonies were counted and photographed using a ChemiDocsystem (Bio-Rad).

Screening of phosphoprotein kinase proteinsAfter treatment of the 8505C cell line, the relative phosphorylation

level of 43 proteins was determined using the Proteome ProfilerHuman-Phospho-Kinase Array Kit (R&D Systems), according to themanufacturer's instruction. The expression level of the phosphorylatedproteins was quantified using the Biorad ChemiDoc system, and thespots were quantified by ImageJ software.

Virtual docking of drug bindingDocking analysis was performed using Autodock Vina and PyMol

software (15).

Xenograft ATC modelsFor orthotopic ATC cell implantation, 5 � 105 8505C-Luc2 cells

(Luc2 denotes cells with stable expression of a luciferase reporter) wereimplanted into the region of the thyroid gland of NOD Cg-Prkdcscid

Il2rgtm1WjI/SzJ mice. Tumor luminescence was measured using theXenogen IVIS in vivo imaging system after intraperitoneal luciferininjection (10 mg/body weight of animals). After 1 week of orthotopicimplantation, mice were randomized into four treatment groups. Themice were given a PLX4720 drug containing diet or vehicle control diet(Plexxikon). Group Imice were treated with the PLX4720 diet (417mgdrug per kg of chow, ppm), plus ponatinib (20 mg/day) by oralgavage. Group II mice were treated with the PLX4720 diet (417ppm), plus 0.9% NaCl daily by oral gavage. Group III mice weretreated with the vehicle control normal diet and ponatinib(20 mg/day) by oral gavage. Group IV control mice received thevehicle control normal diet and 0.9% NaCl daily by oral gavage.Mice were imaged and weighed weekly. They were treated for21 days, after which all mice were euthanized using CO2 inhalation.Tumors, regional lymph nodes, lungs, and liver tissue were collectedafter euthanasia for histologic analysis.

We also used an ATC metastasis mouse model, in which 8505C-Luc2 cells (3 � 104 cells/200 mL) were injected into the tail vein of6-month-old NOD Cg-PrkdcscidIl2rgtm1Wjl/SzJ mice. A week later,in vivo imaging was performed to confirm lung metastases andtreatment was started. Mice were randomized into four groups. Mice

were treated with oral gavage with 0.9% NaCl (control group),PLX4720 (30 mg/kg/day), ponatinib (20 mg/kg/daily), or a combina-tion PLX4720 (30 mg/kg/day) and ponatinib (20 mg/kg/day). Micewere imaged and weighed weekly. Treatment was continued until thefirst mouse reached humane endpoint criteria, upon which all micewere euthanized using CO2 inhalation. Tumors, lungs, and liver tissuewere collected after euthanasia for histologic examination. The sur-vival time was calculated from the day that the 8505C-Luc2 cells wereinoculated (day 0) to the day the animal died. Kaplan–Meier survivalcurves were drawn for each group.

Statistical analysesData are presented as mean � SEM or mean � SD values. The

effect of treatment and differences among experimental group wereassessed using analysis of variance (ANOVA) and appropriate posthoc test. Kaplan–Meier survival analysis, using Mantel–Cox andGehen–Breslow–Wilcoxon tests, was performed for the in vivostudy. A two-tailed P value of ≤ 0.05 was considered statisticallysignificant. GraphPad Prism software (version 8: GraphPad Soft-ware Inc.) was used to perform all statistical analyses. The followingsymbols are used to denote the statistical significance of data:nonsignificant ¼ ns; �, P < 0.05; ��, P < 0.01; ���, P < 0.001; ����,P < 0.0001.

ResultsPLX4720 and ponatinib combination therapy has a synergisticactivity in BRAFV600E thyroid cancer cells

Based on our previous studies, 32 drugs and drug candidates wereselected for pair-wise HTS (16–21). Combination TK inhibitor pona-tinib and vemurafenib showed synergistic activity (Fig. 1A–D). Thecombination of vemurafenib andponatinib had good synergistic effect,based on their effective HSA and BLISS score (22, 23). We selectedPLX4720, also a BRAF inhibitor, for further studies as it had a lowerIC50 inBRAF

V600E-mutant 8505C cells when used in combinationwithponatinib (Supplementary Table S1). For follow-up studies, we usedtwo BRAFV600E-mutant cell lines (8505C and BCPAP) and twoBRAFWT cell lines (THJ-16T and C643).

The proliferation assay showed the drugs had effect on all the celllines tested, both single agents and in combination (Fig. 1E–H). Chou-Talaley algorithm was used to identify drug interaction. PLX4720 andponatinib had a synergistic effect on BRAFV600E cell lines (8505C andBCPAP), but the combination in BRAFWT cell lines had mostly

Table 1. PLX4720 and ponatinib combination shows synergistic activity on BRAFV600E thyroid cancer cells.a

Celllines

PLX4720 5mmol/Lþponatinib0.062mmol/L

PLX4720 5mmol/Lþponatinib0.185mmol/L

PLX4720 5mmol/Lþponatinib0.556mmol/L

PLX4720 10mmol/Lþponatinib0.062mmol/L

PLX4720 10mmol/Lþponatinib0.185mmol/L

PLX4720 10mmol/Lþponatinib0.556mmol/L

PLX4720 15mmol/Lþponatinib0.062mmol/L

PLX4720 15mmol/Lþponatinib0.185mmol/L

PLX4720 15mmol/Lþponatinib0.556mmol/L

8505C 4.22416 0.89812 0.45152 1.97866 9.36626 0.4395 1.11639 0.97061 0.6195C6343 3.21843 2.86244 1.50454 3.39379 0.99065 1.66398 2.74089 2.61221 2.40038BCPAP 0.56469 0.64187 0.97572 0.46722 0.9291 1.15374 0.45521 1.61142 1.01058THJ-16T 1.37974 1.81091 1.56236 1.32762 2.22532 1.36109 1.74805 8.74873 1.72797

Synergistic: CI < 1.0 Additive: CI ¼ 1 Antagonist: CI >1

aThe combination index (CI) was calculated using the Chow–Talalay method with the PLX4720 and ponatinib combination treatment at 72 hours, which had asynergistic effect on BRAFV600E cell lines (8505C and BCPAP). This was compared with BRAFWT cell lines (THJ-16T and C643), which mostly had an antagonisticeffect with the combination treatment. The CI is determined by the following range: CI < 1, synergist; CI ¼ 1, additive; CI > 1, antagonist.

Ghosh et al.

Clin Cancer Res; 26(8) April 15, 2020 CLINICAL CANCER RESEARCH2026

Figure 2.

PLX4720 and ponatinib combination treatment inhibits colony formation, invasion, and migration in BRAFV600E-mutant thyroid cancer cells. A, PLX4720 andponatinib combination inhibits colony formation in 8505C and BCPAP cells. BRAFV600E and BRAFWT cells were incubated with increasing doses of PLX4720 (5, 10,15 mmol/L), ponatinib (0.062, 0.185, 0.556 mmol/L), and their combinations for 12 to 14 days. Images are representative of three independent experiments. B,Histogram representing the mean colony counts of 8505C. PLX4720 (15 mmol/L) and ponatinib (0.556 mmol/L) reduced colony formation more than PLX4720 andponatinib alone in 8505C and BCPAP compared with control (��, P < 0.01) as well as compared with PLX4720 (15 mmol/L) alone (P < 0.05). C and D, Combinationtreatmentwith PLX4720 (15mmol/L) andponatinib (0.556mmol/L) reduced colony formationmore thanPLX4720andponatinib alone inBCPAP cells comparedwithcontrol (��� ,P<0.001).E,PLX4720 (15mmol/L) andponatinib (0.556mmol/L) significantly inhibited cellular invasion andmigration ofBRAFV600E thyroid cancer cells.The combination of PLX4720 (15 mmol/L) and ponatinib (0.556 mmol/L) showed the most significant inhibition of cellular invasion and migration compared withcontrol (� ,P <0.05; �� , P <0.01; ��� , P <0.001) aswell as comparedwith PLX4720 (15mmol/L) alone (P <0.05, P <0.01). PLX¼PLX4720 and PTB¼ ponatinib in all thefigures.

Targeting BRAFV600E Cancers with Novel Combination Agents

AACRJournals.org Clin Cancer Res; 26(8) April 15, 2020 2027

Figure 3.

PLX4720 andponatinib combination treatment induces apoptosis inBRAFV600E-mutant thyroid cancer cells.A andB,FACSanalysis usingAnnexinV–FITCandPIwasperformedafter the cellswere incubatedwith PLX4720, ponatinib, and their combination for 48hours. Therewas an increase (>3.42-fold) in double-positive annexin/PI-labeled cells (late apoptosis, right upper square) and annexin V single-positive cells (early apoptosis, left upper square) with combination treatment (PLX4720 15mmol/L andponatinib 0.556mmol/L) than the single agents PLX4720 (>1.88-fold), ponatinib (0.690-fold), and vehicle control (DMSO) after 48 hours in 8505C cells.Cand D, Quantification of apoptosis of BRAFV600E cells with treatment (percentage). The X axis represents the treatment groups, and the Y axis represents thepercentage of cells in each quadrant (% in quadrant fromA andB). E and F, Effect on caspase-3/7 activity with treatment. Therewas a significant increase in caspase-3/7 activity in BRAFV600E cells after 48 hours of treatment with the combination of drugs (>4.6-fold) than the PLX4720 (>2.1-fold), ponatinib (>0.82-fold) comparedwith vehicle control (����, P < 0.0001). Also, the combination treatment showed significantly higher caspase-3/7 activity compared with individual drug treatments.The result is represented in the relative light unit (RLU, y axis). PLX ¼ PLX4720, PTB ¼ ponatinib. The results are mean � SD.

Ghosh et al.

Clin Cancer Res; 26(8) April 15, 2020 CLINICAL CANCER RESEARCH2028

antagonistic effect (Fig. 1I; Table 1). These results suggested that thetwo drugs may have a complementary mechanism of action inBRAFV600E-mutant cell lines.

PLX4720 and ponatinib combination therapy inhibits colonyformation, cellular invasion, and migration in BRAFV600E-mutant thyroid cancer cells

We found that PLX4720 (15 mmol/L) and ponatinib (0.556 mmol/L)combination treatment reduced colony formation more than

PLX4720 or ponatinib alone in 8505C and BCPAP cells comparedwith control (P < 0.05, P < 0.01, and P < 0.001, respectively;Fig. 2A–D). The reduced colony formation with combinationtherapy was also seen in BRAFWT cells but at higher doses (Sup-plementary Fig. S1A). Additionally, we found that PLX4720and ponatinib combination most significantly inhibited cellularinvasion and migration in BRAFV600E cells compared with control(P < 0.05, P < 0.01, and P < 0.001, respectively) and compared withPLX4720 (15 mmol/L; Fig. 2E).

Figure 4.

PLX4720 and ponatinib combination reduces phosphorylation of ERK (P44/42) and MEK in BRAFV600E-mutant thyroid cancer cells. Both BRAFV600E and BRAFWT

thyroid cancer cells were treated with PLX4720 (10–20 mmol/L), ponatinib (0.185–0.556 mmol/L), or their combination, and phosphorylation of ERK and MEK wasmeasured byWestern blot.A,No significant reduction of ERK andMEK phosphorylationwas detected after up to 48 hours of incubationwith single agents PLX4720or ponatinib in BRAFV600E cells; however, significant reduction in the phosphorylation of ERK and MEK was observed after treatment with the combination ofPLX4720 and ponatinib in BRAFV600E cells (8505C and BCPAP). B, No synergistic effect and no changes in the phosphorylation of ERK and MEK were observed inBRAFWT cells (C643 and THJ-16T) with single or combination drug treatments.

Targeting BRAFV600E Cancers with Novel Combination Agents

AACRJournals.org Clin Cancer Res; 26(8) April 15, 2020 2029

PLX4720 and ponatinib combination therapy induces apoptosisin BRAFV600E-mutant thyroid cancer cells

FACS analysis was performed after the cells were treated withPLX4720, ponatinib, or their combination, and an increase inapoptosis was found with combination treatment in 8505C andBCPAP cell lines (Fig. 3A–D). A significant increase in caspase-3/7activity was also observed in BRAFV600E cells after 24- and 48-hourtreatment with the drug combination (>4.6-fold), PLX4720 (>2.1-fold), and ponatinib (>0.82-fold) as compared with vehicle control

(P < 0.001, P < 0.0001). The combination treatment group alsoshowed significantly higher caspase-3/7 activity when comparedwith single-drug treatment groups (Fig. 3E and F). This suggeststhat the combination treatment results in increased caspase-3/7activity. In the BRAFWT cell lines, the result was heterogeneous withno consistent increase of caspase activity after treatment. In THJ-16T, the single-agent treatment resulted in higher caspase activitywhen compared with control and compared with combinationtreatment (P < 0.05, P < 0.01), and combination treatment had

Figure 5.

PLX4720 and ponatinib combinationinhibits c-JUN and other kinases inBRAFV600E-mutant thyroid cancercells, and c-JUN enhances the antipro-liferative activity of combination treat-ment in vitro. A, A proteome humanphosphokinase array was used toinvestigate the additional mechanismof action of PLX4720 and ponatinibcombination. Combination drug treat-ments decreased phosphorylation ofp38a (T180/Y182), c-Jun (S63), JNK1/3(T183/Y185), and HSP27 (S78/S82)and upregulated phosphorylation ofSTAT3 (S727), AKT1/2/3 (T308),EGF-R (Y1080), GDK3(a/b), b-cate-nin, p70S Kinase (S78/S82), CREB(S100), and p27 (T198) in 8505C. Thered box in the figure is around c-JUN(S63), and the band densitometryrepresentation of these data is shownat the bottom. B, Validation ofdecreased c-JUN phosphorylationwith drug combination treatment byWestern blot. C–F, Western blotshowing c-JUN knockdown in8505C, BCPAP, THJ16T, and C643cells. Bottom, the significant relativeknockdown by band densitometryin 8505C, BCPAP, THJ16T, andC643 cells, respectively (� , P < 0.05;��, P < 0.01; ��� , P < 0.001). (Continuedon the following page.)

Ghosh et al.

Clin Cancer Res; 26(8) April 15, 2020 CLINICAL CANCER RESEARCH2030

significantly lower activity compared with PLX4720 alone (P <0.0001). Whereas in C643 cells, the highest caspase activity wasobserved after PLX4720 treatment and no difference betweenponatinib and combination treatment was observed (P < 0.05,P < 0.01; Supplementary Fig. S1B). Combination treatmentshowed significantly lower activity compared with PLX4720 alone(P < 0.01).

PLX4720 and ponatinib combination therapy inhibitsphosphorylation of ERK and MEK in BRAFV600E-mutant thyroidcancer cells

Combination treatment with PLX4720 and ponatinib significantlydownregulated the phosphorylation of ERK and MEK in BRAFV600E-mutant cell compared with single agents and vehicle control (Fig. 4A;Supplementary Fig. S1C) at 48 hours of treatment. We also evaluatedthis effect over time (24 and 48 hours) and found greater reduction inphosphorylated-ERK (pERK) and phosphorylated-MEK (pMEK) at48 hours as compared with 24 hours with combination treatment(Supplementary Fig. S4C and S4D).

There was no synergistic effect or changes in the phosphorylationof ERK and MEK observed in BRAFWT cell lines (Fig. 4B; Sup-

plementary Fig. S1C). Collectively, these data suggest the enhancedanticancer activity of combination drug treatment in BRAFV600E-mutant cell lines may be due to the more effective reduction inpERK and pMEK levels.

PLX4720 and ponatinib combination therapy reduces c-JUNlevels in BRAFV600E-mutant thyroid cancer cells, and c-JUNmediates the antiproliferative activity of combination drugtreatment

Because PLX4720 and ponatinib combination treatment had asynergistic effect on pMEK/pERK levels in BRAFV600E-mutant cells,we wanted to further explore the possible mechanism(s) that mightcause this enhanced effect on pMEK/pERK in BRAFV600E cells. Theproteome profile human phospho-kinase array was used to explore theeffect of combination PLX4720 and ponatinib treatment. Combina-tion treatment decreased phosphorylation of p38a (T180/Y182), c-JUN (S63), JNK1/3 (T183/Y185), and HSP27 (S78/S82) and increasedphosphorylation of STAT3 (S727), AKT1/2/3 (T308), EGF-R (Y1080),GDK3(a/b), b-catenin, p70S kinase (S78/S82), CREB (S100), and p27(T198; Fig. 5A; Supplementary Table S2). c-JUN expression andphosphorylation was investigated as a possible mechanism of the

Figure 5.

(Continued.) G, c-JUN knockdown affects cellular proliferation. c-JUN knockdown (siRNA1) reduced proliferation significantly compared with control siRNA(���� , P < 0.0001) in 8505C. H, Effect of drug treatment with c-JUN knockdown. Combination drug treatment and single agent reduced proliferation inc-JUN-deficient 8505C cells significantly. Treatment with drug combination as compared with single agents decreased cellular proliferation 1.6-foldcompared with controls (PLX4720 vs. combination; � , P < 0.01). I, Cell number at 96 hours of c-JUN knockdown cells and treatment groups compared withcontrol (�� , P < 0.002; ���� , P < 0.0001) and PLX4720 (15 mmol/L) alone (P < 0.0001). @, PLX vs. other groups.

Targeting BRAFV600E Cancers with Novel Combination Agents

AACRJournals.org Clin Cancer Res; 26(8) April 15, 2020 2031

Figure 6.

PLX4720 and ponatinib combination treatment reduces tumor growth in an orthotopic model and decreases metastasis in a tail-vein metastasis model.A, Ex vivo orthotopic tumor luciferase activity in control and treatment groups showing tumor burden. Representative image of three mice from eachgroup. Mice in the combination group (PLX4720 and ponatinib) had lower tumor burden and showed no signs of sickness. (Continued on the following page.)

Ghosh et al.

Clin Cancer Res; 26(8) April 15, 2020 CLINICAL CANCER RESEARCH2032

synergistic drug action, because c-JUN and ERK cross-talk has beenreported (24), and ponatinib has also been shown to act on c-JUN inbreast cancer (25).

A molecular docking technique was used to evaluate the potentialbinding of ponatinib and PLX4720 in the protein cavity of JNK (PDBid: 2ELJ). The JNK protein is crystallized and the bis-anilino-pyrolopyrimidine inhibitor binds with amino acids (GLU109,MET111, and ASN114), providing a binding site for virtual screening.Ponatinib binds to the ligand binding site of JNKwith a binding energyof�9.5 Kcal/mol and stabilized itself by forming hydrogen bonds withMET111 (3.1 Å) and ASP112 (3.4 Å). PLX4720 binds with a bindingenergy of �7.8 Kcal/mol and stabilized itself by forming hydrogenbonds with ILE32 (2.6 Å). This suggests the potential binding ofponatinib with the ligand binding site of JNK and the amino acidresidue MET111 is more stable than the binding with PLX4720 andmight be involved in inhibition of JNK activity, leading to lowering inphosphorylation of c-JUN. Also, the different binding sites of the twodrugsmight indicate that the drugswhen used in combination have thepotential to bind at different sites reducing the chance of competitiveinhibition and possibly increasing the activity of their combined use(Supplementary Fig. S1D).

Combination treatment had an effect on the phosphorylation ofJNK/c-JUN in BRAFV600E cells as seen in the phospho-kinase array;hence, we performed further experiments to validate the inhibitionof JNK/c-JUN activity by combination treatment. The combinationtreatment downregulated c-JUN phosphorylation in BRAFV600E

cells as shown by Western blot (Fig. 5B). As c-JUN, in combinationwith c-Fos, forms the AP-1 early response transcription factor, themRNA expression levels of both c-JUN and c-FOS were analyzed,and a significant downregulation by the single agents and combi-nation treatment was seen (P < 0.01, P < 0.001; SupplementaryFig. S2A).

Next, we wanted to determine whether the activity of the combi-nation of PLX4720 and ponatinib was dependent on c-JUN. Therefore,c-JUNwas ectopically overexpressed in BRAFV600E-mutant cells (Sup-plementary Fig. S2B). No significant change was observed in cellularproliferation with c-JUN overexpression with single-agent treatmentor combination treatment (Supplementary Fig. S2C and S2D). Knock-down of c-JUN was done in 8505C, BCPAP, C643, and THJ16T(Fig. 5C–F). Knockdown of c-JUN in 8505C (Fig. 5C) cells resulted inreduced cellular proliferation. c-JUN siRNA1 showed significantreduction in proliferation compared with control siRNA, which wasconsistent with the level of knockdown as siRNA1 had a greaterknockdown of c-JUN than siRNA2 (P < 0.0001; Fig. 5G). Moreimportantly, c-JUN knockdown in BRAFV600E-mutant cells was asso-ciated with greater sensitivity to combination treatment compared

with control (P < 0.0001) and PLX4720 (15 mmol/L) alone (P <0.0001; Fig. 5H and I). Combination treatment significantly reducedcolony formation compared with single-agent and vehicle control(Supplementary Fig. S2E and S2F). Cellular proliferation after c-JUNknockdown was studied on BRAFWT cells as well (SupplementaryFig. S5A–S5D).

PLX4720 and ponatinib combination treatment reduces tumorgrowth and metastasis in vivo

Combination PLX4720 and ponatinib treatment significantlyreduced tumor growth (70.3%,P< 0.05) comparedwith vehicle controlgroup and PLX4720 alone (P < 0.05), whereas in the single-agentgroups (PLX4720, 49.4%; ponatinib, 19.95%), there was no significantdifference compared with vehicle control (Fig. 6A and B). Mice in thecontrol group (25� 1.7 g), PLX4720 group (25� 1.9 g), and ponatinibgroup (20.4� 4.0 g) had significantly lower weights as compared withthe combination treatment group (29.4 � 3.2 g; P < 0.05), whomaintained their body weight and showed no signs of illness orcachexia (Fig. 6C). Hematoxylin and eosin staining demonstratedthat pulmonary metastases were lower in mice treated with PLX4720(single) and combination treatment (PLX4720 þ ponatinib) whencompared with control group (P < 0.01; P < 0.001). No significantdifference was seen in the ponatinib group (Fig. 6D).

As most cancer deaths are due to metastatic disease, an ATCmetastasis mouse model was used to test the efficacy of combinationtreatment. Survival analysis showed a significantly longer survival timein mice treated with combination PLX4720 and ponatinib, with 100%of the mice surviving, as compared with control and single-agenttreated groups (P < 0.05; Fig. 6E). Mice in the control and ponatinibgroups had significantly lower body weight and signs of cachexia(≤0.8 g), while mice treated with PLX4720 showed no weight loss orsigns of cachexia (≥1.5 g) when comparedwith baseline bodyweight (P< 0.01). Interestingly, mice in the combination treatment group hadhigher body weight from baseline (≥3.4 g) and no signs of cachexia(Fig. 6F). The mice in the combination group also showed reducedtumor progression (not statistically significant, but this may be due tothe different route of PLX4720 administration) compared with thesingle-agent and control groups based on whole-body luminescencemeasurements (Supplementary Fig. S3A). Tumor burden as measuredby body luciferase signal indicated the lowest tumor burden in thecombination treatment group (Supplementary Fig. S3B). Hematoxylinand eosin staining demonstrated that lung metastasis was not signif-icantly different with treatment (Fig. 6G; Supplementary Fig. S3D),but liver metastasis was significantly lower in mice treated with thecombination of ponatinib and PLX4720 (P < 0.01; SupplementaryFig. S3C and S3E).

(Continued.) B, There was a significant reduction of whole-body luciferase signal in the mice treated with the drug combination (70.3%, P < 0.05) than the single-agent groups (PLX, 49.4%; ponatinib, 19.95%, respectively) comparedwith the control and PLX4720 alone (P <0.05).C, Themice in control (25� 1.67), PLX4720 (25� 1.92), and ponatinib (20.40 � 4.03) group had lower body weight and cachexia compared with combination treatment. The mice in the combination groupmaintained their body weight and showed no sign of cachexia (� , P < 0.05). All the values are in mean � SD. D, Metastasis analysis in lungs from H&E staining.Metastasis is significantly low inPLX420-treatedmice and combination treatedmice comparedwith control (�� ,P<0.01; ��� ,P<0.001). The scoringwasperformedat4�magnification. E, Tail-vein metastasis model study. Survival curve analysis showed the mice in the combination group (PLX4720 and ponatinib, 100%) survivedlonger compared with the control group (50%). The X axis represents the elapsed time in weeks and the Y axis the percentage surviving. Kaplan–Meier survivalanalysis usingMantel–Cox andGehen–Breslow–Wilcoxon testwas performed todetermine statistical significance in survival. Therewas a significantly longer survivalin the combination group (� , P < 0.05). F, Tumor burden measured by body luciferase signal indicated the lowest tumor burden in the combination treatment[nonsignificant (n.s.)]. G, Representative image of H&E staining. There was lower pulmonary metastasis in mice treated with the combination of PLX4720 andponatinib, but it was not statistically significant based on randomly selected three fields (4�magnification) and counting the number of tumor foci. In the orthotopicmodel, 20 skidmice (NODCg-Prkdcscid Il2rgtm1WjI/SzJ), 8505C-Luc2 cells taggedwith a luciferase reporter genewere implanted orthotopically into the right thyroidlobe, and aweek after tumor implantation, luciferase activity was detected by in vivo bioluminescence. In themetastasismodel, 8505C-Luc2 cellswere injected in thetail vein of NOD Cg-PrkdcscidIl2rgtm1Wjl/SzJ mice (n ¼ 40) to see the effect of treatment. @, PLX vs. other groups.

Targeting BRAFV600E Cancers with Novel Combination Agents

AACRJournals.org Clin Cancer Res; 26(8) April 15, 2020 2033

PLX4720 and ponatinib combination treatment overcomesBRAF inhibitor (PLX4720 > 60 mmol/L) resistance in 8505Ccells

Resistance to BRAF inhibition occurs in BRAFV600E-mutant can-cers. Therefore, 8505C cells that were resistant to PLX4720 treatment(not sensitive to 30mmol/Lwith an IC50 of 60mmol/L)were used to testif the combination treatment could be effective (SupplementaryFig. S2G). Combination PLX4720 and ponatinib treatment signifi-cantly inhibited cellular proliferation (PLX4720 15 mmol/L and pona-tinib 0.185 mmol/L, PLX 4720 30 mmol/L and ponatinib 0.185 mmol/L,PLX4720 30mmol/L and ponatinib 0.556mmol/L, PLX4720 60mmol/Land ponatinib 0.185, PLX4720 60 mmol/L and ponatinib 0.556 mmol/L) as compared with control (P < 0.0001) and PLX4720 alone (P <0.001; Supplementary Figs. S2H and S3F). We also observed reducednumber and size of colonies when treated with the combination ofPLX4720 and ponatinib (Supplementary Fig. S3G). Chou–Talalayanalysis of PLX4720 and ponatinib in 8505C-resistant cells demon-strated that the synergismof the drugs at the respective doses thatmakethe combination effective, although the single drugs at those doseswere not effective (Supplementary Fig. S4A). Western blot analysisdemonstrated lower pERK levels with combination treatment in8505C-resistant cells (Supplementary Fig. S4B).

DiscussionIn this study, we report the combination of PLX4720 and ponatinib

has synergistic anticancer activity in BRAFV600E-mutant cells based onin vitro and in vivo studies. The synergisticmechanism of action resultsin more effective inhibition of the MAPK pathway with lower pERKand c-JUN levels, and induces caspase-dependent apoptosis. More-over, the combination of PLX4720 and ponatinib is able to overcomeresistance in PLX4720-resistant BRAFV600E-mutant cells.

Approximately 20% of patients harboring activating mutations inBRAF develop intrinsic resistance and do not respond to BRAFinhibitors (26), which limits the effectiveness of the treatment. Thus,combination therapy with BRAF andMEK inhibitors has been studiedand may achieve a synergistic therapeutic effect, which could reducetreatment dose and toxicity. Dabrafenib and trametinib (BRAFV600E

and MEK inhibitors) were approved by the FDA to treat BRAFV600E-mutant ATC, and it is the only combination treatment being used forthis type of cancer. The combination has also been approved forBRAF-mutated melanoma and squamous cell carcinoma of the lungs.Although recent reports suggest that BRAF and MEK inhibitors usedin combination in BRAFV600E-mutant melanoma have initial goodresponse rates, most patients develop resistance to the combinationtherapy (27–30). Our preclinical data suggest a novel combination ofPLX4720 and ponatinib might be effective as there is clearly room forimprovements in the current therapies available for ATC.

In this study, we report for the first time that a novel combination ofdrugs, the BRAF inhibitor PLX4720 and the FDA-approved multi-target TK inhibitor ponatinib, synergistically reduces p-MEK and p-ERK compared with single-drug treatments in BRAFV600E cells. Thechemoresistance of ponatinib has been frequently raised as criticalissues in the preclinical and clinical settings; this might result in lowereffect of ponatinib in vivo compared with in vitro study (31).

Pharmacologic inhibitors of BRAFV600E are usually the drug ofchoice for first-line therapy in BRAFV600E tumors as this is a targetablemutation that drives cancer initiation/progression (32). The effect ofvemurafenib in combination with ponatinib showed a synergisticeffect in BRAFV600E-mutant cells. The BRAFV600E inhibitor vemur-afenib has provided a major advance for the treatment of patients with

BRAFV600E-mutant metastatic melanoma. However, in BRAFV600E-mutant thyroid cancer, it has been found to be relatively resistant tovemurafenib treatment (33, 34) and cell lines have a higher IC50 thanmelanoma cells, and the reason for this difference in response remainsunclear (35). Furthermore, PLX4720 treatment alone had a similareffect on cellular proliferation and colony formation in BRAFWT andBRAFV600E cell lines, results similar to previous studies (36). Thepresence of additional mutations (excluding BRAF) and altered sig-naling pathways present in the cell lines tested may account for this.Furthermore, although PLX4720 showed a similar effect on cellularproliferation inBRAFV600E-negative cell lines, the effect on pERK/ERKand pMEK/MEK was variable. For our follow-up studies, PLX4720, asecond-generation BRAFV600E inhibitor, was used because of its muchstronger affinity and favorable pharmacokinetic properties. Nuceraand colleagues (36) reported that PLX4720 (10 mmol/L) reduced theMEK/ERK phosphorylation in BRAFV600E-mutant 8505C cells.PLX4720 (5, 10, 15 mmol/L) in combination with ponatinib (0.185mmol/L, 0.556 mmol/L) inhibited cell migration, invasion, and colonyformation in BRAFV600E-mutant 8505C and BCPAP cell lines. There-fore, the combination of BRAF inhibition and the multitarget TKinhibitor ponatinib can be useful in targeting thyroid cancer and othercancers harboring a BRAFV600E mutation.

Previous studies have shown that cancer cells are able to adaptsignaling pathway circuits in response to drug treatment by establish-ing alternate signaling routes. Hence, one critical aspect to improvecancer treatment is not only to inhibit the primary oncogenic pathwaythat reduces cell proliferation, but simultaneously to prevent func-tional redundancies and pathway cross-talk that facilitates survival ofcancer cell populations, rendering tumors resistant to therapy (37). Assuch, studies based on network pharmacology principles have tried toidentify synergistic multitarget intervention strategies to improveclinical efficacies. Our studywas designed to target themost commonlyactivated pathway BRAF/MEK/ERK, with a BRAFV600E inhibitor(PLX4720) and a multitarget TK inhibitor ponatinib.

A link between ERK and JNK signaling in melanoma, a cancer withhigh rates of BRAFV600E mutations, has been reported (38). Consti-tutively active ERK affects the c-JUN oncogene, its upstream kinaseJNK, and its downstream targets RAC1 and cyclin D1. Understandinghow these signaling pathways are rewired offers new targets fortherapy. Cross-talk between these two pathways can contribute to arobust and integrated signaling transduction network involved in cellproliferation (24). In breast cancer metastasis, ponatinib represses theexpression of BCLM-associated genes,mainly through theERK/c-JUNsignaling pathway by inhibiting the transcription of JUN and accel-erating the degradation of the c-JUN protein (25). These investigatorsidentified ponatinib as a new drug to inhibit BCLM, and c-JUN wasfound to be a crucial factor and a potential drug target.We investigatedthe effects of PLX4720 and ponatinib on c-JUN as a possible mech-anism for their synergistic action. Our data showed a lower level of c-JUN and JNK kinase activity after combination treatment inBRAFV600E-mutant cells, suggesting a functional interaction of ERKand JNK pathway, and the inhibitory effect of PLX4720 and ponatinibon this cross-talkmight cause the synergistic effect.We also discoveredthat c-JUN knockdown along with the combination of PLX4720 andponatinib significantly inhibited cell proliferation and colony forma-tion in BRAFV600E-mutant cells, demonstrating the important role c-JUN plays in BRAFV600E thyroid cancer cells. Interestingly, JNK is aserine/threonine kinase and not a TK, but inhibition of its effectorc-JUN by combination treatment suggests that it might not be thedirect target but gets regulated from upstream inhibition. These resultssuggest that c-JUN mediates some of the synergistic effect of

Ghosh et al.

Clin Cancer Res; 26(8) April 15, 2020 CLINICAL CANCER RESEARCH2034

combination PLX4720 and ponatinib treatment but is not the solemediator as illustrated by our knockdown studies showing c-JUN isnot the sole regulator of cellular proliferation.

Developing resistance to BRAFV600E inhibition alone is a significantproblem; thus, we performed experiments on PLX4720-resistant8505C cells generated by established protocol (39, 40). Thyroid cancercell lines are known to have complex mutational profiles and harborconcurrent alteration in many genes (41). The 8505C cell line hasmutations in TP53, EGFR, PIK3R1, PIK3R2, NF2, SMARCA4,SMARCD1, and TERT, in addition to BRAFV600E (19, 40, 41). Thepresence of an activating BRAFV600E mutation generally predictsresponse to BRAF inhibitors, but resistance to this treatment devel-ops (39). Interestingly, we found, in the resistant cells, reducedphosphorylation of the downstream MEK1/2 (on Ser217/221) andERK1/2 (on T202/Y204) signaling pathway targeted by PLX4720 at itsIC50 dose, unlike the parental cells. Ponatinib as a single agent waseffective in the PLX4720-resistant 8505C cells. Although the concen-tration we used might not be achievable in vivo, this observation in thein vitro system may be due to the additional mutations present in thiscell line. Resistant cell lines have also been reported to acquire a moreinvasive phenotype characterized by increased cell mobility andmetastatic capacity (40).

Upregulation of a distinct receptor TK has been shown tosustain signaling through a signaling pathway, despite continuedinhibition of the primary oncoprotein with the targeted drug (42).Thus, drug combinations to block proliferation pathways are indevelopment, but the fundamental combinatorial principle is stillelusive. Here, we propose the combination of a BRAF-mutantinhibitor and a multitargeted TK inhibitor might prove to beeffective in overcoming BRAF inhibitor resistance, as it can targetmultiple pathways.

In summary, the current study demonstrates that the combina-tion of PLX4720 and ponatinib has significant and synergisticanticancer activity in BRAFV600E-mutant thyroid cancer cellsin vitro and in vivo. Combination treatment with PLX4720 and

ponatinib is a highly promising new combination targeted therapyfor BRAFV600E-mutant thyroid cancer, as well as for BRAF inhib-itor–resistant BRAF V600E-mutant cell. Our preclinical studies sug-gest that the combination of PLX4720 and ponatinib treatmentshould be tested in clinical trials.

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

Authors’ ContributionsConception and design: C. Ghosh, S. Kumar, Y. Kushchayeva, L. Zhang, E. KebebewDevelopment of methodology: C. Ghosh, S. Kumar, D. Wei, S.K. Gara, Y. Zhang,E. KebebewAcquisition of data (provided animals, acquired and managed patients, providedfacilities, etc.): C. Ghosh, K. Gaskins, D. Wei, M. Shen, S. MukherjeeAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): C. Ghosh, S. Kumar, M. Boufraqech, D. Wei, S.K. Gara,L. Zhang, M. Shen, E. KebebewWriting, review, and/or revision of the manuscript: C. Ghosh, S. Kumar,M. Boufraqech, D. Wei, S.K. Gara, L. Zhang, M. Shen, S. Mukherjee, E. KebebewAdministrative, technical, or material support (i.e., reporting or organizing data,constructing databases):C. Ghosh, Y. Kushchayeva, K. Gaskins, S.K. Gara, Y. Zhang,E. KebebewStudy supervision: S. Kumar, E. KebebewOther (acquisition of experimental data): S. Mukherjee

AcknowledgmentsThe authors are thankful to Dr. Gideon Bollag from Plexxikon Inc., Berkley, CA,

for his intellectual contribution and critical assessment of the work. This work wassupported by the intramural research program of the Center for Cancer Research,NCI, NIH (1ZIABC011286-09).

The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby marked advertisement in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.

Received May 15, 2019; revised September 3, 2019; accepted January 10, 2020;published first January 14, 2020.

References1. Nilubol N, Kebebew E. Should small papillary thyroid cancer be observed? A

population-based study. Cancer 2015;121:1017–24.2. Landa I, Ibrahimpasic T, Boucai L, Sinha R, Knauf JA, Shah RH, et al. Genomic

and transcriptomic hallmarks of poorly differentiated and anaplastic thyroidcancers. J Clin Invest 2016;126:1052–66.

3. Shell J. Combination therapy with sorafenib and vemurafenib is effective inanaplastic thyroid cancer. J Am Coll Surg 2016;223:S45.

4. Ito K, Hanamura T, Murayama K, Okada T, Watanabe T, Harada M, et al.Multimodality therapeutic outcomes in anaplastic thyroid carcinoma: improvedsurvival in subgroups of patients with localized primary tumors. Head Neck2012;34:230–7.

5. Liu TR, Xiao ZW, Xu HN, Long Z, Wei FQ, Zhuang SM, et al. Treatment andprognosis of anaplastic thyroid carcinoma: a clinical study of 50 cases. PLoS One2016;11:e0164840.

6. Cabanillas ME. Recent advances and emerging therapies in anaplastic thyroidcarcinoma. F1000Research 2018;7. pii: F1000 Faculty Rev-87.

7. Kunstman JW, Juhlin CC, Goh G, Brown TC, Stenman A, Healy JM, et al.Characterization of the mutational landscape of anaplastic thyroid cancer viawhole-exome sequencing. Hum Mol Genet 2015;24:2318–29.

8. Lim AM, Taylor GR, Fellowes A, Cameron L, Lee B, Hicks RJ, et al. BRAFinhibition in BRAFV600E-positive anaplastic thyroid carcinoma. J Natl ComprCanc Netw 2016;14:249–54.

9. Long GV, Stroyakovskiy D, Gogas H, Levchenko E, de Braud F, Larkin J, et al.Combined BRAF and MEK inhibition versus BRAF inhibition alone in mela-noma. N Engl J Med 2014;371:1877–88.

10. Arozarena I, Wellbrock C. Overcoming resistance to BRAF inhibitors.Ann Transl Med 2017;5:387.

11. Subbiah V, Kreitman RJ, Wainberg ZA, Cho JY, Schellens JHM, Soria JC, et al.Dabrafenib and trametinib treatment in patients with locally advanced or meta-static BRAF V600-mutant anaplastic thyroid cancer. J Clin Oncol 2018;36:7–13.

12. Nazarian R, Shi H, Wang Q, Kong X, Koya RC, Lee H, et al. Melanomas acquireresistance to B-RAF(V600E) inhibition by RTK or N-RAS upregulation. Nature2010;468:973–7.

13. Lu H, Liu S, Zhang G, Bin W, Zhu Y, Frederick DT, et al. PAK signalling drivesacquired drug resistance to MAPK inhibitors in BRAF-mutant melanomas.Nature 2017;550:133–6.

14. Chou TC. Drug combination studies and their synergy quantification using theChou-Talalay method. Cancer Res 2010;70:440–6.

15. TrottO,OlsonAJ. AutoDockVina: improving the speed and accuracy of dockingwith a new scoring function, efficient optimization, and multithreading.J Comput Chem 2010;31:455–61.

16. Mehta A, Zhang L, Boufraqech M, Liu-Chittenden Y, Zhang Y, Patel D, et al.Inhibition of survivin with YM155 induces durable tumor response in anaplasticthyroid cancer. Clin Cancer Res 2015;21:4123–32.

17. Mehta A, Zhang L, BoufraqechM, Zhang Y, Patel D, ShenM, et al. Carfilzomib isan effective anticancer agent in anaplastic thyroid cancer. Endocr Relat Cancer2015;22:319–29.

18. Sadowski SM, BoufraqechM, Zhang L,Mehta A, Kapur P, Zhang Y, et al. Torin2targets dysregulated pathways in anaplastic thyroid cancer and inhibits tumorgrowth and metastasis. Oncotarget 2015;6:18038–49.

AACRJournals.org Clin Cancer Res; 26(8) April 15, 2020 2035

Targeting BRAFV600E Cancers with Novel Combination Agents

19. ZhangL, BoufraqechM, LakeR,KebebewE.CarfilzomibpotentiatesCUDC-101-induced apoptosis in anaplastic thyroid cancer. Oncotarget 2016;7:16517–28.

20. Zhang L, He M, Zhang Y, Nilubol N, Shen M, Kebebew E. Quantitative high-throughput drug screening identifies novel classes of drugs with anticanceractivity in thyroid cancer cells: opportunities for repurposing. J Clin EndocrinolMetab 2012;97:E319–28.

21. Zhang L, Zhang Y, Mehta A, Boufraqech M, Davis S, Wang J, et al. Dualinhibition of HDAC and EGFR signaling with CUDC-101 induces potentsuppression of tumor growth and metastasis in anaplastic thyroid cancer.Oncotarget 2015;6:9073–85.

22. Berenbaum MC. What is synergy? Pharmacol Rev 1989;41:93–141.23. Yadav B, Wennerberg K, Aittokallio T, Tang J. Searching for drug synergy in

complex dose-response landscapes using an interaction potency model.Comput Struct Biotechnol J 2015;13:504–13.

24. Jiang T, Pan CQ, Low BC. BPGAP1 spatially integrates JNK/ERK signalingcrosstalk in oncogenesis. Oncogene 2017;36:3178–92.

25. Shao W, Li S, Li L, Lin K, Liu X, Wang H, et al. Chemical genomics revealsinhibition of breast cancer lung metastasis by Ponatinib via c-Jun. Protein Cell2019;10:161–77.

26. Turajlic S, Furney SJ, StampG, Rana S, Ricken G, Oduko Y, et al.Whole-genomesequencing reveals complex mechanisms of intrinsic resistance to BRAF inhi-bition. Ann Oncol 2014;25:959–67.

27. Wagle N, Van Allen EM, Treacy DJ, Frederick DT, Cooper ZA, Taylor-Weiner A,et al. MAP kinase pathway alterations in BRAF-mutant melanoma patients withacquired resistance to combinedRAF/MEK inhibition.CancerDiscov2014;4:61–8.

28. Welsh SJ, Rizos H, Scolyer RA, Long GV. Resistance to combination BRAF andMEK inhibition in metastatic melanoma: where to next? Eur J Cancer 2016;62:76–85.

29. Johnson DB, Childress MA, Chalmers ZR, Frampton GM, Ali SM, RubinsteinSM, et al. BRAF internal deletions and resistance to BRAF/MEK inhibitortherapy. Pigment Cell Melanoma Res 2018;31:432–6.

30. MoriceauG,HugoW,HongA, ShiH,KongX,YuCC, et al. Tunable-combinatorialmechanisms of acquired resistance limit the efficacy of BRAF/MEKcotargeting butresult in melanoma drug addiction. Cancer Cell 2015;27:240–56.

31. Li H, Wang Y, Chen Z, Lu J, Pan J, Yu Y, et al. Novel multiple tyrosine kinaseinhibitor ponatinib inhibits bFGF-activated signaling in neuroblastoma

cells and suppresses neuroblastoma growth in vivo. Oncotarget 2017;8:5874–84.

32. Charles RP, Silva J, Iezza G, Phillips WA, McMahon M. Activating BRAF andPIK3CA mutations cooperate to promote anaplastic thyroid carcinogenesis.Mol Cancer Res 2014;12:979–86.

33. Montero-Conde C, Ruiz-Llorente S, Dominguez JM, Knauf JA,VialeA, ShermanEJ, et al. Relief of feedback inhibition of HER3 transcription by RAF and MEKinhibitors attenuates their antitumor effects in BRAF-mutant thyroid carcino-mas. Cancer Discov 2013;3:520–33.

34. Hyman DM, Puzanov I, Subbiah V, Faris JE, Chau I, Blay JY, et al. Vemurafenibin multiple nonmelanoma cancers with BRAF V600 mutations. N Engl J Med2015;373:726–36.

35. Wang W, Kang H, Zhao Y, Min I, Wyrwas B, Moore M, et al. Targetingautophagy sensitizes BRAF-mutant thyroid cancer to vemurafenib. J ClinEndocrinol Metab 2017;102:634–43.

36. Nucera C, Nehs MA, Nagarkatti SS, Sadow PM, Mekel M, Fischer AH, et al.Targeting BRAFV600E with PLX4720 displays potent antimigratory and anti-invasive activity in preclinicalmodels of human thyroid cancer. Oncologist 2011;16:296–309.

37. Jaeger S, Igea A, Arroyo R, Alcalde V, Canovas B, OrozcoM, et al. Quantificationof pathway cross-talk reveals novel synergistic drug combinations for breastcancer. Cancer Res 2017;77:459–69.

38. Lopez-Bergami P, Huang C, Goydos JS, Yip D, Bar-Eli M, Herlyn M, et al.Rewired ERK-JNK signaling pathways in melanoma. Cancer Cell 2007;11:447–60.

39. Varmeh S, Vanden Borre P, Gunda V, Brauner E, Holm T, Wang Y, et al.Genome-wide analysis of differentially expressed miRNA in PLX4720-resistantand parental human thyroid cancer cell lines. Surgery 2016;159:152–62.

40. BoufraqechM, Patel D,NilubolN, PowersA, King T, Shell J, et al. Lysyl oxidase isa key player in BRAF/MAPK pathway-driven thyroid cancer aggressiveness.Thyroid 2019;29:79–92.

41. Landa I, Pozdeyev N, Korch C, Marlow LA, Smallridge RC, Copland JA, et al.Comprehensive genetic characterization of human thyroid cancer cell lines: avalidated panel for preclinical studies. Clin Cancer Res 2019;25:3141–51.

42. Neel DS, Bivona TG. Resistance is futile: overcoming resistance to targetedtherapies in lung adenocarcinoma. NPJ Precis Oncol 2017;1:pii: 3.

Clin Cancer Res; 26(8) April 15, 2020 CLINICAL CANCER RESEARCH2036

Ghosh et al.