sbi-0640756 attenuates the growth of clinically unresponsive

Priority Report

SBI-0640756 Attenuates the Growth of ClinicallyUnresponsive Melanomas by Disrupting the eIF4FTranslation Initiation ComplexYongmei Feng1,AnthonyB. Pinkerton1, LauraHulea2,3,TongwuZhang4,MichaelA.Davies5,Stefan Grotegut1, Yann Cheli1, Hongwei Yin6, Eric Lau1, Hyungsoo Kim1, Surya K. De1,Elisa Barile1, Maurizio Pellecchia1, Marcus Bosenberg7, Jian-Liang Li1, Brian James1,Christian A. Hassig1, Kevin M. Brown4, Ivan Topisirovic2,3, and Ze'ev A. Ronai1

Abstract

Disrupting the eukaryotic translation initiation factor 4F(eIF4F) complex offers an appealing strategy to potentiate theeffectiveness of existing cancer therapies and to overcome resis-tance to drugs such as BRAF inhibitors (BRAFi). Here, we iden-tified and characterized the small molecule SBI-0640756 (SBI-756), a first-in-class inhibitor that targets eIF4G1 and disrupts theeIF4F complex. SBI-756 impaired the eIF4F complex assemblyindependently ofmTOR and attenuated growth of BRAF-resistantand BRAF-independent melanomas. SBI-756 also suppressedAKT and NF-kB signaling, but small-molecule derivatives wereidentified that only marginally affected these pathways whilestill inhibiting eIF4F complex formation and melanomagrowth, illustrating the potential for further structural and

functional manipulation of SBI-756 as a drug lead. In the geneexpression signature patterns elicited by SBI-756, DNA damage,and cell-cycle regulatory factors were prominent, with muta-tions in melanoma cells affecting these pathways conferringdrug resistance. SBI-756 inhibited the growth of NRAS, BRAF,and NF1-mutant melanomas in vitro and delayed the onset andreduced the incidence of Nras/Ink4a melanomas in vivo. Fur-thermore, combining SBI-756 and a BRAFi attenuated theformation of BRAFi-resistant human tumors. Taken together,our findings show how SBI-756 abrogates the growth of BRAF-independent and BRAFi-resistant melanomas, offering a pre-clinical rationale to evaluate its antitumor effects in othercancers. Cancer Res; 75(24); 5211–8. �2015 AACR.

IntroductionThe emergence of effective inhibitors for BRAF-mutant mela-

noma has had major impact on the clinical management ofmelanoma (1). However, the initial success of such treatmentshas been limited due to the propensity of melanomas to developresistance (2). In most cases, mechanisms underlying BRAFinhibitor (BRAFi) resistance include activation of genetic orepigenetic pathways that circumvent targeted BRAF and restoreMAPK and related signaling to levels sufficient to fuel tumori-genesis (2). This outcome has led to development of combination

therapies targeting both BRAF and associated pathways, such asMEK and PI3K (3), albeit, with limited success. Furthermore, 50%ofmelanomas, such as thoseharboringNRAS andNF1mutations,lackBRAFmutations, and are thus not amenable to BRAFi therapy(4). Thus, tumor chemoresistance and the lack of therapies forBRAF wild-type (WT) tumors remains a major clinical challenge.

We identified BI-69A11, which inhibits both AKT and NF-kBsignaling (5) and attenuates melanoma development and pro-gression inbothhuman xenografts andmouse geneticmodels (6–8). In efforts to improve the biophysical properties of BI-69A11,we identified SBI-0640756 (SBI-756), which retained the biologiceffects of the original compound while possessing superior phar-macokinetics. Extended characterization of SBI-756 identifiedeIF4G1 as its direct target. eIF4G1 is a large scaffolding proteinthat is a key component of the eukaryotic translation initiationfactor 4F (eIF4F) complex (9). Small translational repressors,eIF4E-binding proteins (4E-BP), associate with eIF4E, and impairits binding to eIF4G and the eIF4F complex assembly (10).mTORC1-mediated phosphorylation of 4E-BPs leads to theirdissociation form eIF4E, enabling eIF4E interaction with eIF4Gand the formation of the eIF4F complex (10). Although requiredfor cap-dependent translation of all nuclear-encoded mRNAs,increased eIF4F levels stimulate translation of mRNAs encodingcancer-promoting proteins while having only amarginal effect ontranslation of house-keeping mRNAs (11). Correspondingly,elevated eIF4F activity has been linked to resistance to BRAF- andMEK-targeted therapies (12). SBI-756 targeting of the eIF4G1disrupts the eIF4F complex assembly, even in BRAFi-resistant

1Cancer Center, Sanford Burnham Prebys Medical Discovery Institute,La Jolla, California. 2Lady Davis Institute for Medical Research,Sir Mortimer B. Davis-Jewish General Hospital, Montr�eal, Canada.3Department of Oncology, McGill University, Montr�eal, Canada.4Division of Cancer Epidemiology and Genetics, Laboratory of Trans-lational Genomics, NCI, Bethesda, Maryland. 5Melanoma MedicalOncology, MD Anderson Cancer Center, Houston, Texas. 6Cancer andCell Biology Division, The Translational Genomics Research Institute(TGen), Phoenix, Arizona. 7Departments of Dermatology and Pathol-ogy, Yale University, School of Medicine, New Haven, Connecticut.

Note: Supplementary data for this article are available at Cancer ResearchOnline (http://cancerres.aacrjournals.org/).

Corresponding Author: Ze'ev A. Ronai, Sanford Burnham Prebys MedicalDiscovery Institute, 10901 North Torrey Pines Road, La Jolla, CA 92037. Phone:858-646-3185; Fax: 815-366-8003; E-mail: [email protected]

doi: 10.1158/0008-5472.CAN-15-0885

�2015 American Association for Cancer Research.

CancerResearch

www.aacrjournals.org 5211

on February 14, 2018. © 2015 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 24, 2015; DOI: 10.1158/0008-5472.CAN-15-0885

http://cancerres.aacrjournals.org/

melanoma. Correspondingly, SBI-756 attenuates resistance toBRAFi and inhibits NRAS- and NF1-mutant melanomas.

Materials and MethodsWestern blot analysis and antibodies

Cells were rinsed with PBS and lysed as previously described(8). Protein concentration was determined using CoomassiePlus Protein Assay Reagent (Thermo Scientific). Equal amountsof cell lysate proteins (50 mg) were separated on SDS-PAGE andtransferred to polyvinylidene difluoride membranes (PerkinEl-mer Life Sciences). Membranes were blocked (5% BSA/TBST, 1hour) and incubated with primary antibodies (1 hour at roomtemperate or overnight at 4�C), with shaking. Following threeTBST washes, membranes were incubated for 1 hour at roomtemperature secondary antibodies (1:10,000). Detection andquantifications were made using Odyssey Infrared ImagingSystem (LiCor Biosciences), or by exposing them to X-rayfilm. Antibodies against p-AKT, p-PRAS40, p-IKK, p-IkB, p-TSC,p-mTOR, p-p70S6K, p-RPS6, p-4E-BP1, pSGK3, AKT, PRAS40,IKK, IkB, mTOR, p70S6K, RPS6, 4E-BP1, GSK3, eIF4G1, andeIF4E were purchased from Cell Signaling Technology. Anti-bodies against b-actin and a-tubulin were obtained fromSanta Cruz Biotechnology. Secondary antibodies were goatanti-rabbit Alexa-680 F(ab0)2 (Molecular Probes) and goatanti-mouse IRDye 800 F(ab0)2 (Rockland Immunochemicals).All antibodies were used according to the suppliers'recommendations.

Cell cultureMelanoma lines were obtained from the Wistar Institute, Yale

University, TGen,NCI, andATCCandmaintained in high-glucoseDulbecco modified Eagle medium (HyClone) with 5% FBS and1% penicillin–streptomycin at 37�C in 5% CO2. E1A/RAS trans-formed WT and 4E-BP double knockout (DKO) mouse embry-onic fibroblasts (MEF) were described previously (1). For cell lineauthentication, short tandem repeat (STR) analysis was per-formed on isolated genomic DNA with the GenePrint 10 Systemfrom Promega, and peaks were analyzed using GeneMarker HIDfromSoftgenetics. Allele calls were searched against STRdatabasesmaintained by ATCC (www.atcc.org), DSMZ (www.dsmz.de),Texas Tech University Children's Oncology Group (cogcell.org),and theWistar InstituteMelanomaCell STR Profiles (http://www.wistar.org/lab/meenhard-herlyn-dvm-dsc/page/melanoma-cell-str-profiles). Authentication was last performed on August 24,2015.

m7GTP pull-down assayAs previously described (13), cells growing in 100 mm plates

were washed (cold PBS), collected, and lysed in 50 mmol/LMOPS/KOH (7.4), 100 mmol/L NaCl, 50 mmol/L NaF, 2mmol/L EDTA, 2 mmol/L EGTA, 1% NP40, 1% Na-DOC, 7mmol/L b-mercaptoethanol, protease inhibitors, and phospha-tase inhibitor cocktail (Roche). Lysates were incubated with m7

-GDP-agarose beads (Jena Bioscience; 20 minutes), washed (4times) with 50mmol/LMOPS/KOH (7.4), 100mmol/L NaCl, 50mmol/L NaF, 0.5 mmol/L EDTA, 0.5 mmol/L EGTA, 7 mmol/Lb-mercaptoethanol, 0.5 mmol/L PMSF, 1 mmol/L Na3VO4 and0.1mmol/LGTP. Boundproteinswere elutedbyboiling thebeadsin loading buffer. m7-GDP-agarose pulled down material wasanalyzed by Western blot analysis.

Results and DiscussionTo improve the biophysical properties of BI-69A11, we

designed and synthesized over 60 BI-69A11 analogues (Fig.1A). Following iterative structure–activity relationship (SAR), weselected four analogies with improved pharmacokinetics (Sup-plementary Table S1; Fig. 1A). Of those, SBI-756 and SBI-726exhibited superior properties (60� improved aqueous solubilityand up to 100� improved permeability), and a favorable phar-macokinetic profile, while not exerting toxicity (SupplementaryTables S1 and S2; Supplementary Fig. S1A and S1B).

Examination of four humanmelanoma lines revealed that SBI-756 and SBI-726 were comparable with BI-69A11 in their anti-proliferative effects (Fig. 1B), and inhibition of AKT and NF-kBactivity (Fig. 1C). SBI-756 elicited comparable toxicity in mela-noma cells and melanocytes, but was less toxic against nontrans-formed fibroblasts (Supplementary Fig. S1C). SBI-756 and SBI-726 were more effective than BI-69A11 in attenuating colonyformation by BRAF- andNRAS-mutant melanoma cells (Fig. 1D).On the basis of its overall properties (in vivo and in vitro), weselected to further characterize SBI-756.

To identify proteins that interact andmay serve as direct targetsfor SBI-756 we performed gas chromatography/liquid mass spec-trometry (GC/MS-MS) using biotinylated BI-69A11. Of the 74proteins that bound specifically (outcompeted using 10� excessof soluble BI-69A11) was eIF4G1 (Supplementary Table S3). Wethus set to determine whether the eIF4F complex is indeeddisrupted by SBI-756, and further, whether the effect of SBI-756 on AKT, NFkB, and mTOR are dispensable for its effect oneIF4F as well as for suppression of melanoma growth.

We next determine whether SBI-756 dissociates eIF4G1 fromthe eIF4F complex using m7GTP-agarose pull-down, which cap-tures the eIF4F complex. SBI-756 effectively dissociated eIF4G1from the eIF4E in a dose-dependent manner, which was accom-panied by a concomitant increase in 4E-BP1:eIF4E binding(Fig. 2A), reflective of impaired eIF4F complex formation. Inhi-bition of the eIF4F complex was also confirmed for the parentcompound BI-69A11, although SBI-756 was more potent (Sup-plementary Fig. S2A).

SBI-756 also inhibits the AKT/mTORC1 signaling andmTORC1 inhibition disrupts the eIF4F complex via activation of4E-BPs (10). Todeterminewhether SBI-756 impedes eIF4F assem-bly directly or via mTORC1, we employed 4E-BP1/2 DKO MEFs,wherein mTOR inhibition does not impair the eIF4F assembly(13). Whereas torin1 induced dissociation of eIF4G1 from eIF4EinWTbut not in4E-BPDKOMEFs, SBI-756 reduced eIF4G1:eIF4Eassociation in bothWT and 4E-BP DKOMEFs (Fig. 2B). Likewise,SBI-756, but not torin1, attenuated the proliferation of E1A/RAS–transformed 4E-BP DKO MEFs (Fig. 2C). These results substan-tiate that the effect of SBI-756 on the eIF4F complex assembly islargely mTOR-independent.

As levels of the eIF4F complex inversely correlate with theeffectiveness of various cancer therapies (12, 14), we assessed theintegrity of the eIF4F complex after treatingmelanoma cells with acombination of BRAFi vemurafenib (aka PLX4032) and SBI-756.Comparing A375 melanoma cultures that are sensitive to BRAFiand resistant derivatives (A375R), BRAFi slightly reduced eIF4G1association with eIF4E, whereas SBI-756 had a more robust effect(Fig. 2D). Significantly, SBI-756, but not BRAFi, promoted a dose-dependent dissociation of eIF4G1 from eIF4E in A375R (Fig. 2D),Lu1205R and WM793R cells (Supplementary Fig. S2B). Further-

Feng et al.

Cancer Res; 75(24) December 15, 2015 Cancer Research5212




0

20

40

60

80

100

120

0

20

40

60

80

100

120

DMSO BI-69A11 SBI-726 SBI-756 SBI-970 SBI-973


0

20

40

60

80

100

120


0

20

40

60

80

100

120


0

20

40

60

80

100

120

140


−6.5 −6.0 −5.5 −5.0 −4.5−20

0

20

40

60

80

100

120 WM1346A B

C

Lu1205

WM1346

BI-69A11 SBI-756

Log(M)

% In

hibi

tion

−6.5 −6.0 −5.5 −5.0 −4.5−20

0

20

40

60

80

100

120 Lu1205

% In

hibi

tion

Log(M)

SBI-726SBI-756SBI-970SBI-973BI-69A11

SBI-726SBI-756SBI-970SBI-973BI-69A11

SBI-756 SBI-970 SBI-973

DMSO BI-69A11 SBI-726

SBI-756 SBI-970 SBI-973

DMSO BI-69A11 SBI-726

% C

FE

2 5 2 5 11 2 5 1 5 1 22

pAKT(T308)

HSP90

pPRAS40 (T246)

5

pAKT(T473)

μmol/L

AKT

PRAS40

D

DM

SO

SBI-726

SBI-756

SBI-970

SBI-973

BI-69A

11

% C

FE

NH

OO

N

Cl

NH

OO

Cl

N

F

NH

OO

Cl

NH

OO

N

NH

Cl

Cl

SBI-0640726 SBI-0640756

SBI-0645970 SBI-0645973

pIKKα/β

IKK

IκB

pIκB

pTSC2(S939)

p4E-BP1(T37/46)

pmTOR(S2448)

TSC2

A375

UACC903

WM3629

DMSO 69A11 726

DMSO 69A11 726

DMSO 69A11 726

756 970 973

756 970 973

756 970 973

Tubulin

726

756

970

973

69A11

−6.5 −6.0 −5.5 −5.0 −4.5−20

0

20

40

60

80

100

120WM793

726

756

970

973

69A11

% In

hibi

tion

Log(M)

% In

hibi

tion

Log(M)

WM1366

−20

0

20

40

60

80

100

120

−6.5 −6.0 −5.5 −5.0 -4.5

% C

FE

% C

FE

% C

FE

Figure 1.Development and characterization of SBI-756. A, more than 60 analogues were synthesized targeting numerous BI-69A11 regions including the linker (red),aryl groups (blue and green), and benzimidazole ring (gray). Top analogues are shown. B, mutant BRAF (Lu1205) or mutant NRAS (WM1346)melanoma lines were plated (triplicates 384-well plates; 1,500 cells per well) and cell viability was assessed 48 hours after treatment with indicatedcompounds. Growth inhibition is calculated as percentage of DMSO-treated controls and is plotted against the log drug concentration. C, UACC903 cellswere treated with DMSO or indicated concentrations of BI-69A11 analogues for 24 hours and whole-cell lysates were immunoblotted with indicatedantibodies. D, indicated cultures were plated at low density (500 cells/well in 6-well plates) and grown in medium containing indicated compounds. Thenumber of colonies formed after 10 days in culture was determined by crystal violet staining.

eIF4G1 Targeting in BRAF-Resistant and BRAF WT Melanoma

www.aacrjournals.org Cancer Res; 75(24) December 15, 2015 5213




more, a SBI-756/BRAFi combination decreased the amount ofeIF4G1 bound to eIF4E, which was not seen following BRAFitreatment alone (Supplementary Fig. S2C and S2D). While con-sistent with a recent study reporting that compounds that targeteIF4E (such as 4EGI-1) or eIF4A (such as rocaglate derivatives)synergize with BRAFi to inhibit proliferation of BRAFi-resistantcells (12), ourfindings demonstrate the effectiveness of SBI-756 indisrupting the eIF4F complex by targeting the eIF4G1.

To identify signaling pathways affected by SBI-756, we per-formed reverse-phase protein array analysis of BRAF- and NRAS-mutant melanomas (Supplementary Table S4). Unsupervisedclustering of proteins that significantly (P < 0.05) changed expres-sion/phosphorylation levels with SBI-756 treatment for 24 hoursdemonstrated a consistent inhibitory effect in both cell lines onmTOR signaling, and multiple translation initiation regulators,reflected by markedly decreased phosphorylation of S6, mTOR,

6 h 24 h12 h

DM

SO

DM

SO

DM

SO

β-Actin

B

A375

PLX4032

5 10 0.5 1.0

SBI-756

eIF4G1

eIF4E

4E-BP1

β-Actin

eIF4G1

eIF4E

4E-BP1

β-Actin

Input

D

SBI-756 (μmol/L)

Input

Torin1 (μmol/L)

SBI-756 (μmol/L)

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0 0.2 0.4 0.6 0.8 1.0

WTBP DKO

0.0

0.2

0.4

0.6

0.8

1.0

1.2

0 0.156 0.312 0.625 1.25 2.5

WTBP DKO

DM

SO

DM

SO

WT 4E-BP DKO

Torin

1

Torin

1

0.75 0.751.75 1.75

SBI-756 (μmol/L) SBI-756 (μmol/L)

S6K

p-S6K(T389)Input

C

A

0.75

0.50

0.25

0.75

0.50

0.25

0.75

0.50

0.25

eIF4G1

4E-BP1

eIF4E

β-Actin

eIF4G1

4E-BP1

eIF4E

m7GTPpull-down

β-Actin

eIF4G1

4E-BP1

eIF4E

β-Actin

eIF4G1

4E-BP1

eIF4E

m7GTPpull-down

DM

SO

A375R

PLX4032

5 10 0.5 1.0

SBI-756

DM

SO

(μmol/L)

m7GTPpull-down

Brd

Urd

inco

rpor

atio

n (%

ctr)

Brd

Urd

inco

rpor

atio

n (%

ctr)

Figure 2.eIF4G1 is an SBI-756 target. A, proteins prepared from UACC903 melanoma cells were treated with indicated SBI-756 concentrations and incubated withm7GTP-agarose beads to capture the eIF4F complex. Shown is dose-dependent inhibition of eIF4G1 binding to eIF4E, concomitant with increased binding ofinhibitory 4E-BP1 to eIF4E. Bottom, the total cell lysate. B, indicated cultures (WT or DKO 4E-BP) were treated with Torin1 (250 nmol/L) or indicatedSBI-756 concentrations for 6 hours.Western blots show respective protein levels in total cell lysates or following pull-down usingm7GTP agarose beads. Bottomplotshows input (5%) and top plot shows m7GTP pull-down (50%). b-Actin served as a loading control and to exclude contamination in m7GTP pull-down.C, E1A/Ras-transformed WT or 4e-bp1/4e-bp2 DKO MEFs were treated with either Torin1 or SBI-756 at indicated concentrations and cellproliferation was measured 48 hours later with the aid of BrdUrd incorporation. Values were normalized to DMSO-treated triplicates. Error bars, SD(n ¼ 3). D, A375 and A375R (PLX4032-resistant) cells were treated with vehicle (DMSO), a BRAFi (vemurafenib; PLX4032), or SBI-756 at the indicateddoses for 24 hours. Cell lysates (200 mg) were subjected to m7GTP pull-down. Amounts of the indicated proteins in input (5%) or pull-down (50%) sampleswere determined by Western blotting; b-actin served as a loading control (input) and to exclude contamination (m7GTP pull-down).

Feng et al.





P70S6K as well as TSC2 (Supplementary Fig. S3A; SupplementaryTable S4). Of interest, SBI-756 did not affect MAPK or JAK–STATsignaling pathways. Western blot analysis confirmed the dose-dependent inhibitory effects of SBI-756 on these proteins and adecrease in 4E-BP1 phosphorylation in UACC903 and A375 celllines (Supplementary Fig. S3B).

The effect of SBI-756 on mTOR, AKT, and NF-kB, led us todeterminewhethermodification of SBI-756 could reduce its effecton these signaling pathways while retaining its effect on the eIF4Fcomplex. Among SBI-756 derivatives, at least one (SBI-755199)was found to be as effective in inducing melanoma cell death(Fig. 3A) while exhibiting reduced inhibition of AKT, TSC2, PRS6,and NFkB activity (Fig. 3B). Notably, SBI-755199 retained itseffective inhibition of mRNA translation (Fig. 3C). The latter wasperformed using a bicistronic construct, which enablesmeasuringeIF4G1-dependent and eIF4G1-independent HCV-IRES–driventranslation. These data substantiate that the effects of SBI-756 onthe eIF4F complex underpin its biologic activity therebyprovidingthe basis for the SAR-based screen of SBI-756 analogues.

Dose–response analysis of SBI-756 in 21 melanoma linesidentified two groups representing respective SBI-756–sensitiveand -resistant lines (>2-fold expression difference in IC50 betweengroups; Fig. 3D). Evaluation of gene expression data from thesecell lines enabled mapping differentially expressed genes (DEG)for each group. In total, we detected 1,533 significant DEGsbetween sensitive and the more resistant cells (P < 0.05; foldchange > 1.5). Analysis of gene enrichment within canonicalpathways (Supplementary Table S5) identified higher expressionlevels of genes involved in DNA damage response (P ¼ 4.1E �10�9), ATM signaling (P ¼ 2.8 � 10�7), and CHK-mediated cell-cycle checkpoint control (P ¼ 6.0 � 10�7) in SBI-756–resistantcells, with concurrently lower expression of G2–M cell-cyclecheckpoint control genes (P ¼ 1.5 � 10�8). In addition, amonggenes exhibiting lower expression in resistant lines were key cell-cycle regulatory proteins, including, CDKN1A,CDKN2A, andRB1(P < 3.0 � 10�14).

As genes conferring drug resistance can often reveal pathwaysthat underlie the drug's response, we established multiple SBI-756–resistant clones from UACC903 and UACC3629 cultures.Exome sequencing of individual clones identified a total of 587protein-coding gene variants in 550 genes that were notdetected in the SBI-756–sensitive parental cultures (Supple-mentary Fig. S3C; Supplementary Table S6). Of these somaticmutations, 88.1% (n ¼ 517) were missense or nonsense singlenucleotide variants, with the remaining 11.9% (n ¼ 70) con-sisting of splicing or frameshift mutations that alter readingframes. IPA analysis indicated enrichment of numerous pro-teins (n ¼ 174) implicated in melanoma (P ¼ 3.8 � 10�13).Among genes with mutations identified in multiple SBI-756–resistant clone, are those implicated in fatty-acid beta oxidation(P ¼ 0.0008) and DNA damage response (P ¼ 0.004) and cell-cycle checkpoint control (P ¼ 0.01; Supplementary Table S7).These pathways were previously described to be affected uponinhibition of the eIF4G pathway at the level of translation (15–17). Consistent with these observations, a clear G2–M transi-tion was identified following SBI-756 treatment in parental(sensitive) but not in the SBI-756–resistant melanoma cells(Supplementary Fig. S3D).

To confirm the effect of SBI-756 on the eIF4F complex in BRAFi-resistant cultures on neoplastic growth, we assessed SBI-756effectiveness on their two-dimensional (2D) growth in vitro (with

BRAFi–PLX4032) and on tumorigenesis in vivo (with BRAFi–PLX4720). Growth in 2D and colony-forming efficiency (CFE)were effectively attenuated in both parental (sensitive) and resis-tant cultures, with NF1-mutant melanoma lines exhibiting equalor greater sensitivity (Fig. 4A and Supplementary Fig. S4A andS4B). Effectiveness of SBI-756 in NF1-mutant melanoma wasfurther confirmed in primary cultures (Supplementary Fig. S4C).

In vivo we first evaluated SBI-756 using an inducible NrasQ61K/Ink4a�/� genetic model in which melanoma tumors emergewithin 16 to 20 weeks (Fig. 4B). Administration of SBI-756 only,starting 11weeks after genetic inactivation of Ink4a and inductionof NRasQ61E (about 10–14 days prior to tumor appearance),delayed tumor onset (from 20–26 weeks), and reduced tumorincidence, by 50%, compared with the control nontreated group(Fig. 4B). No signs of toxicity were identified during and afteradministration of SBI-756 (21-week period), consistent withearlier studies with BI-69A11 (7). Given its effectiveness as asingle agent, one would expect that combination with MAPKinhibitors could offer novel therapeutic modalities for Nras-mutant tumors.

In vivo, we monitored growth of A375 tumors in immunode-ficient mice subjected to either BRAFi alone or BRAFi combinedwith SBI-756. Notably, SBI-756 did not elicit toxicity in mice,which was monitored by liver function and body weight (Sup-plementary Table S2, Supplementary Fig. S1B). Growth of estab-lished tumors (�250 mm3) was largely inhibited by treatmentwith either BRAFi alone or a combination of BRAFi plus SBI-756(Fig. 4C). However, as seen in human melanoma, tumors in theBRAFi-treated group (3/5) resumed growth, whereas no tumorswere seen in mice (0/4) treated with the drug combination (Fig.4C), suggesting that combining SBI-756 with BRAFi antagonizesBRAFi-resistant melanoma in vivo. When we allowed tumors toreach 500 mm3 before initiating treatment, 5 of 7 (75%) micesubjected to BRAFi treatment alone relapsed as drug-resistanttumors (4/5 within 4–6 weeks), whereas only 3 of 6 (50%)subjected to combination treatment developed resistance, albeitmore slowly (2/3 after 8 weeks; Supplementary Fig. S4D). Nota-bly, assessing SBI-756 effect on the eIF4F complex in vivo revealeda time-dependent disruption of the eIF4F complex in melanomatumors grown in animals that were subjected to treatment withboth BRAFi and SBI-756 (up to 8 hours, consistent with the halflife of SBI-756; Fig. 4D). These results demonstrate the disruptionof the eIF4F complex in vivo, consistent with the effectiveness ofSBI-756 in overcoming BRAFi-resistant phenotype.

The ability of SBI-756 to elicit inhibition of the eIF4F comp-lex independently of its effect on the mTOR/4E-BP pathwayhighlights its distinct properties compared with currently avail-able inhibitors to other components of the eIF4F complex such as4EGI-1 (14). Furthermore, identification of SBI-756 derivatives (i.e., SBI-755199) that retain the effect on the eIF4F complex and onmelanoma cells while minimizing the effect on the mTOR/AKT/NFkB substantiates eIF4F complex as the key target for the SBI-756and its derivatives.

Although expressionof the eIF4G1proteinhas been reported tobe upregulated in tumors, including melanoma (Cancer Atlas-and TCGA-based analyses) we expect that dysregulation ofeIF4G1 is not required for its support expression of cancer-promoting genes. Hence, dampening its overall effectivenessalong the translation initiation complex is expected to limitoncogene expression and corresponding addiction, justifying itstargeting. The inability of vemurafenib to affect the eIF4F complex






A

B

C

−8.0 −7.5 −7.0 −6.5 −6.0 −5.5 −5.00

20

40

60

80

120

100

Log(M)

% V

iabi

lity

SBI-0640756

SBI-0755199

SBI-0756858

SBI-0756863

D

0.0

0.4

0.8

1.2

SB

I-756

952

3291

2545

3337

2641 558

1592 25

7

903

612 91 383

2331 64

7

2496

1940

2404

1118

2512

2994

1120

UACC

Mea

n (μ

mol

/L)

Sensitive Resistant

eIF4G1dependent

eIF4G1independent

PRAS40

pPRAS40(T246)

AKT

pAKT(T308)

HSP90

DM

SO

1 42

SBI-756

1 42

SBI-858

1 42

SBI-863

1 42

SBI-199

pRPS6(S235/236)

HSP90

Tubulin

pTSC2(S939)

TSC2

RPS6

IκB

A375R

DM

SO

1 42

SBI-756

1 42

SBI-858

1 42

SBI-863

1 42

SBI-199

TNFα (20 ng/mL)

μmol/L

μmol/Lw/o

TN

Fα

100

80

60

40

20

0

100 SBI-0640756IC50 = 3.83 μmol/L

SBI-756863IC50 = 3.13 μmol/L

SBI-756858IC50 = 2.75 μmol/L

SBI-755199IC50 = 4.29 μmol/L

SBI-0640756IC50 = 4.59 μmol/L

SBI-756863IC50 = 3.68 μmol/L

SBI-756858IC50 = 3.29 μmol/L

SBI-755199IC50 = 6.11 μmol/L

80

60

40

20

0

100

80

60

40

20

0

100

80

60

40

20

0

100

80

60

40

20

0

100

80

60

40

20

0

100

80

60

40

20

0

100

80

60

40

20

0

0.10 0.32 10.10 0.100.32 0.321 13.16 3.16 0.10 0.32 1 3.16 3.16

0.10 0.32 1 3.160.10 0.32 1 3.160.10 0.32 1 3.160.10 0.32 1 3.16

Figure 3.Characterization of SBI-756 analogues. A–C, SBI-756 and indicated analogues were assessed for their effect on growth of UACC903 melanoma cells 24 hoursafter their addition (A); AKT, TSC2, PRS6, and NFkB signaling pathway 4 hours after their addition (B); inhibition of translation in vitro using the bicistronicRenilla/HCV-IRESfirefly luciferase constructs (C). D, twenty-onemelanoma lineswere treatedwith serial 2-fold dilutions of SBI-756, yielding final drug concentrationranges of 100 mmol/L to 0.2 nmol/L, and cell viability was assessed using CellTiter Glo after 72 hours. Cells were subgrouped according to IC50 values thatidentified seven most resistant and seven most sensitive cell lines.


Feng et al.




in the resistant melanoma, while SBI-756 retains its effectiveness,underlies the novelty and importance of eIF4G1 targeting.Significantly, the low dose and the lack of in vivo toxicity observedfor SBI-756 makes it an even more attractive for furtherevaluation.

Genetic support for the significance of the eIF4F complex incancer was provided through the characterization of the haploin-sufficient eif4e mice, which were found to be more resistant totumor development (18). Correspondingly, efficacy of a plethoraof inhibitors, including PI3K,mTOR,HER2, andMAPK, is limited

B

0

20

40

60

80

100

120

DMSO 0.08 0.16 0.32 0.63 1.25 2.5 5 10

A

D

0

20

40

60

80

100

120

DMSO 0.25 0.5 0.75 1.00

SBI-756 (μmol/L)

A375

A375R

% V

iabi

lity

A375

A375R

% C

FE

SBI-756 (μmol/L)

DMSODMSO

DMSO

0.25 μmol/L 0.5 μmol/L

DMSO 0.25 μmol/L 0.5 μmol/L

1.0 μmol/L

1.0 μmol/L0.75 μmol/L

0.75 μmol/L

A375

A375R

eIF4G1

β-Actin

4E-BP1

eIF4E

input

eIF4G1

β-Actin

4E-BP1

eIF4E

h0 1 3 8* 24C

6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 360

5

10

15

20

25

30

Week after doxycycline treatment

Vehicle

SBI-756

Mic

e w

ith m

elan

oma

% SBI-756

m7GTP

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 150

500

Tum

or v

olum

e (m

m3 )

1,000

1,500

2,000

2,500

3,000

3,500

Week after treatment

DMSO (#=5)

PLX4720 (#=5)

PLX4720 + SBI-756 (#=4)pull-down

Figure 4.SBI-756 inhibits growth ofNF1-mutant, NRas-mutant, andBRAFi-resistantmelanoma. A, parental andBRAFi-resistantNF1-mutantmelanomas (left)weremonitoredfor their response to SBI-756 at the indicated concentrations and for their ability to form CFE (middle and right). B, mice of the Nras(Q61K)::Ink4a�/� genotypeor WT mice (38–40 mice for each genotype) were administered vehicle or SBI-756 at 0.5 mg/kg by intraperitoneal injection twice a week starting at week 11after Nras(Q61K) activation and Ink4a inactivation. Tumor development was monitored twice per week. Latency and frequency of melanoma development overthe 21-week treatmentperiod (weeks 11–32) are shown. C, A375humanmelanomacellswere injected subcutaneously (1� 106) into theflankof nudemice andallowedto form established tumors. Once tumors reached approximately 250 mm3, mice were randomly grouped and subjected to the indicated treatments [chowcontaining PLX4720 (417 mg/kg) and/or SBI-756 (1 mg/kg, 2 times per week intraperitoneally)]. Tumor size was measured at the indicated time points. Theexperiment was repeated twice. D, proteins prepared at the indicated times from A375 tumors that were subjected in vivo to treatment with SBI-756 wereincubated with m7GTP-beads to capture the eIF4F complex. Shown is time-dependent inhibition of eIF4G1 binding to eIF4E in vivo, concomitant with increasedbinding of inhibitory 4E-BP1 to eIF4E. Bottom plot shows input, total cell lysate.






in cells that exhibit high eIF4E–4E-BP ratio and correlates withtheir inability to disrupt the eIF4F complex (14).Moreover, recentfindings show that the efficacy of 4EGI-1, which is thought todirectly target eIF4E:eIF4G association is likely to be predeter-mined by the eIF4E-4E-BP ratio. As SBI-756 decreases eIF4F levelsindependently of mTOR and cellular eIF4E–4E-BP ratio, its com-bination with currently available inhibitors is an appealing ther-apeutic modality that allows targeting the eIF4F complex inde-pendently of mTOR, eIF4E, or 4E-BP status in the cell.

Disclosure of Potential Conflicts of InterestM.A. Davies reports receiving commercial research grants from Astrazeneca,

GSK, Merck, Oncothyreon, Roche/Genentech, and Sanofi-Aventis. He is also aconsultant/advisory board member for GSK, Novartis, Roche/Genentech,Sanofi-Aventis, and Vaccinex. No potential conflicts of interest were disclosedby the other authors.

Authors' ContributionsConception and design: Y. Feng, A.B. Pinkerton, S. Grotegut, E. Barile, C.A.Hassig, K.M. Brown, I. Topisirovic, Z.A. RonaiDevelopment of methodology: Y. Feng, S. Grotegut, H. Yin, C.A. HassigAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): L. Hulea, T. Zhang, Y. Cheli, H. Yin, M. BosenbergAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): Y. Feng, A.B. Pinkerton, L. Hulea, T. Zhang, M.A.Davies, S. Grotegut, H. Yin, E. Lau, H. Kim, J.-L. Li, B. James, C.A. Hassig,K.M. Brown, I. Topisirovic, Z.A. Ronai

Writing, review, and/or revision of the manuscript: Y. Feng, A.B. Pinkerton,L. Hulea, T. Zhang, M.A. Davies, S. Grotegut, H. Yin, E. Lau, M. Bosenberg,B. James, C.A. Hassig, K.M. Brown, I. Topisirovic, Z.A. RonaiAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): T. Zhang, S. Grotegut, H. Kim, Z.A. RonaiStudy supervision: S. Grotegut, I. Topisirovic, Z.A. RonaiOther (synthesis of some compounds): S.K. De

AcknowledgmentsThe authors thankmembers of theGenomic, Proteomic, andAnimalCores at

SBP and Ronai lab members for discussions, as well as the Cancer GenomicsResearch Laboratory (CGR) at the NCI. The authors also thankDr. Jerry Pelletierfor providing luciferase constructs for in vitro translation assays. This article isdedicated in fond memory of Greg Roth, Ph.D., their colleague at SanfordBurnham Prebys Lake Nona, for his wisdom, compassion, integrity, his love ofthe sciences, and his contributions to the development of clinically meaningfulprojects.

Grant SupportThis work was supported by the Melanoma Research Alliance. Core Services

were supported by NCI Cancer Center grant P30 CA30199 (Z.A. Ronai) andCA016672 (M.A. Davies). This work was also supported by the AssistantSecretary of Defense for Health Affairs through the Peer-Reviewed CancerProgram under Award No. W81XWH-14-1-0127 (Z.A. Ronai), CIHR (MOP-115-195), CRS (01713), and CIHR new investigator salary award (IT), and bythe Intramural Research Program of the Division of Cancer Epidemiology andGenetics; National Cancer Institute (T. Zhang and K.M. Brown).

Received April 1, 2015; revised September 1, 2015; accepted September 21,2015; published OnlineFirst November 24, 2015.

References1. Flaherty KT, Fisher DE. New strategies in metastatic melanoma: oncogene-

defined taxonomy leads to therapeutic advances. Clin Cancer Res 2011;17:4922–8.

2. Lito P, Rosen N, Solit DB. Tumor adaptation and resistance to RAFinhibitors. Nat Med 2013;19:1401–9.

3. Kemper K, de Goeje PL, Peeper DS, van Amerongen R. Phenotype switch-ing: tumor cell plasticity as a resistance mechanism and target for therapy.Cancer Res 2014;74:5937–41.

4. Fedorenko IV,GibneyGT, Smalley KS.NRASmutantmelanoma: biologicalbehavior and future strategies for therapeutic management. Oncogene2013;32:3009–18.

5. Gaitonde S, De SK, TcherpakovM, Dewing A, YuanH, Riel-MehanM, et al.BI-69A11-mediated inhibition of AKT leads to effective regression ofxenograft melanoma. Pigment Cell Melanoma Res 2009;22:187–95.

6. Feng Y, Barile E, De SK, Stebbins JL, Cortez A, Aza-Blanc P, et al. Effectiveinhibition of melanoma by BI-69A11 is mediated by dual targeting of theAKT and NF-kappaB pathways. Pigment Cell Melanoma Res 2011;24:703–13.

7. Feng Y, Lau E, Scortegagna M, Ruller C, De SK, Barile E, et al. Inhibition ofmelanoma development in the Nras(Q61K): Ink4a(�/�) mouse model bythe smallmolecule BI-69A11. PigmentCellMelanomaRes 2013;26:136–42.

8. Barile E, De SK, Feng Y, Chen V, Yang L, Ronai Z, et al. Synthesis and SARstudies of dual AKT/NF-kappaB inhibitors against melanoma. Chem BiolDrug Des 2013;82:520–33.

9. Sonenberg N, Hinnebusch AG. Regulation of translation initiation ineukaryotes: mechanisms and biological targets. Cell 2009;136:731–45.

10. Roux PP, Topisirovic I. Regulation of mRNA translation by signalingpathways. Cold Spring Harb Perspect Biol 2012;a012252.

11. Graff JR, Konicek BW, Carter JH, Marcusson EG. Targeting the eukaryotictranslation initiation factor 4E for cancer therapy. Cancer Res 2008;68:631–4.

12. Boussemart L, Malka-Mahieu H, Girault I, Allard D, Hemmingsson O,Tomasic G, et al. eIF4F is a nexus of resistance to anti-BRAF and anti-MEKcancer therapies. Nature 2014;513:105–9.

13. Dowling RJ, Topisirovic I, Alain T, Bidinosti M, Fonseca BD, Petroulakis E,et al. mTORC1-mediated cell proliferation, but not cell growth, controlledby the 4E-BPs. Science 2010;328:1172–6.

14. Bhat M, Robichaud N, Hulea L, Sonenberg N, Pelletier J, Topisirovic I.Targeting the translation machinery in cancer. Nat Rev Drug Discov2015;14:261–78.

15. Badura M, Braunstein S, Zavadil J, Schneider RJ. DNA damage and eIF4G1in breast cancer cells reprogram translation for survival and DNA repairmRNAs. Proc Natl Acad Sci U S A 2012;109:18767–72.

16. Silvera D, Arju R, Darvishian F, Levine PH, Zolfaghari L, Goldberg J, et al.Essential role for eIF4GI overexpression in the pathogenesis of inflamma-tory breast cancer. Nat Cell Biol 2009;11:903–8.

17. Braunstein S, Karpisheva K, Pola C, Goldberg J, Hochman T, Yee H, et al. Ahypoxia-controlled cap-dependent to cap-independent translation switchin breast cancer. Mol Cell 2007;28:501–12.

18. TruittML,ConnCS, Shi Z, PangX, Tokuyasu T,CoadyAM, et al.Differentialrequirements for eIF4E dose in normal development and cancer. Cell2015;162:59–71.


Feng et al.




2015;75:5211-5218. Published OnlineFirst November 24, 2015.Cancer Res Yongmei Feng, Anthony B. Pinkerton, Laura Hulea, et al. Melanomas by Disrupting the eIF4F Translation Initiation ComplexSBI-0640756 Attenuates the Growth of Clinically Unresponsive

Updated version

10.1158/0008-5472.CAN-15-0885doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2015/12/23/0008-5472.CAN-15-0885.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/75/24/5211.full#ref-list-1

This article cites 17 articles, 5 of which you can access for free at:

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

Subscriptions

Reprints and

[email protected]

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

Permissions

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

.http://cancerres.aacrjournals.org/content/75/24/5211To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-15-0885

http://cancerres.aacrjournals.org/content/suppl/2015/12/23/0008-5472.CAN-15-0885.DC1

http://cancerres.aacrjournals.org/content/75/24/5211.full#ref-list-1

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

mailto:[email protected]

http://cancerres.aacrjournals.org/content/75/24/5211


sbi-0640756 attenuates the growth of clinically unresponsive

Documents