Cancer Biology and Translational Studies

EZH2 Inhibition by Tazemetostat Results inAltered Dependency on B-cell ActivationSignaling in DLBCLDorothy Brach, Danielle Johnston-Blackwell, Allison Drew, Trupti Lingaraj,Vinny Motwani, Natalie M.Warholic, Igor Feldman, Christopher Plescia,Jesse J. Smith, Robert A. Copeland, Heike Keilhack, Elayne Chan-Penebre,Sarah K. Knutson, Scott A. Ribich, Alejandra Raimondi, and Michael J. Thomenius

Abstract

The EZH2 small-molecule inhibitor tazemetostat (EPZ-6438) is currently being evaluated in phase II clinical trialsfor the treatment of non-Hodgkin lymphoma (NHL). Wehave previously shown that EZH2 inhibitors display anantiproliferative effect in multiple preclinical models of NHL,and that models bearing gain-of-function mutations in EZH2were consistently more sensitive to EZH2 inhibition thanlymphomas with wild-type (WT) EZH2. Here, we demon-strate that cell lines bearing EZH2 mutations show a cytotoxicresponse, while cell lines with WT-EZH2 show a cytostaticresponse and only tumor growth inhibition without regres-sion in a xenograft model. Previous work has demonstratedthat cotreatment with tazemetostat and glucocorticoid recep-tor agonists lead to a synergistic antiproliferative effect inboth mutant and wild-type backgrounds, which may provide

clues to the mechanism of action of EZH2 inhibition inWT-EZH2 models. Multiple agents that inhibit the B-cellreceptor pathway (e.g., ibrutinib) were found to have syner-gistic benefit when combined with tazemetostat in bothmutant and WT-EZH2 backgrounds of diffuse large B-celllymphomas (DLBCL). The relationship between B-cell acti-vation and EZH2 inhibition is consistent with the proposedrole of EZH2 in B-cell maturation. To further support this, weobserve that cell lines treated with tazemetostat show anincrease in the B-cell maturation regulator, PRDM1/BLIMP1,and gene signatures corresponding to more advanced stagesof maturation. These findings suggest that EZH2 inhibition inboth mutant and wild-type backgrounds leads to increased B-cell maturation and a greater dependence on B-cell activationsignaling. Mol Cancer Ther; 16(11); 2586–97. �2017 AACR.

IntroductionDisruption of chromatin modulation is emerging as an

important, if not requisite, step in the process of oncogenesis.Mutations in chromatin modifiers are frequent occurrencesin a number of cancers and are often associated with aberrantcell fate decisions (1–3). These lesions are particularly frequentevents in non-Hodgkin B-cell Lymphoma. Loss-of-functionmutations in EP300, CREBBP, or KMT2D (MLL2) or gain-of-function mutations in EZH2 occur frequently in B-cell lym-phomas (4–6). Recent reports have demonstrated that loss ofKMT2D or EZH2 gain-of-function leads to disruption of cellfate decisions (e.g., differentiation or apoptosis) and responsesto extracellular signaling (7–10). Loss of cell fate control allows

lymphomas to remain in a highly proliferative state and notproceed through the normal B-cell maturation process. Forinstance, germinal center lymphomas (follicular, GCB-DLBCL)maintain a transcriptional program that supports autonomousproliferation (11). For normal B-cells, the germinal centerreaction is maintained by a gene expression program that islargely governed by the transcriptional repressor, BCL6 (12).During this phase, cells undergo a period of rapid proliferation,somatic hypermutation, reduced DNA damage response, andclass switch recombination. B-Cells that successfully produce ahigh affinity B-cell receptor–antigen interaction will ultimatelyundergo activation and differentiation; those that fail to pro-duce a B-cell receptor that recognizes antigen or that recognizesa self-antigen will undergo apoptosis via tightly regulatedsignaling pathways (11, 12). In addition to B-cell receptoractivity, many other pathways such as Toll-like receptor (13),cytokine (14), and CD40 signaling contribute to B-cell activa-tion, differentiation, and apoptosis regulation (11) and theseresponses are frequently altered in DLBCL (15). While follicularand GC lymphomas often have many characteristics of germi-nal center B-cells, they typically lose the ability to undergoapoptosis and maturation (15).

BCL6, a known regulator of B-cell maturation is misregu-lated by translocations with the 3q27 locus, resulting in itsoverexpression, an enhancement of the germinal center pro-gram, and a subsequent deficiency in B-cell differentiation.

Epizyme Inc., Cambridge, Massachusetts.

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

D. Brach and D. Johnston-Blackwell contributed equally to this article.

Corresponding Authors: Alejandra Raimondi, Epizyme Inc., 400 TechnologySquare, 4th Floor, Cambridge, MA 02139. Phone: 617-500-0708; Fax: 617-349-0707; E-mail: araimondi@epizyme.com; and Michael Thomenius,mthomenius@epizyme.com

doi: 10.1158/1535-7163.MCT-16-0840

�2017 American Association for Cancer Research.

MolecularCancerTherapeutics

Mol Cancer Ther; 16(11) November 20172586

on September 20, 2020. © 2017 American Association for Cancer Research. mct.aacrjournals.org Downloaded from

Published OnlineFirst August 23, 2017; DOI: 10.1158/1535-7163.MCT-16-0840

Additional genetic lesions, including mutations in BCL6,MEF2B, and FBX011 have been reported to enhance the abilityof BCL6 to maintain the germinal center phenotype. Thesemutations lead to enhanced BCL6 stability or altered activity(12, 16, 17). Defective apoptosis and differentiation allowB-cells to remain in the germinal center reaction indefinitely,which results in a germinal center lymphoma (16). In additionto GCB lymphomas, it has been shown that gain-of-function inBCL6 frequently occurs in ABC lymphomas (6), which display amore mature, post-germinal center phenotype. The role ofBCL6 in ABC lymphomas remains largely unknown and it ispossible that its alteration is required only for initiation of thislymphoma subtype (18), although recent reports indicate thatactivated B-cell lymphomas may require sustained BCL6 activ-ity (19). As above, KMT2D and CREBBP are frequently mutatedin follicular lymphoma and DLBCL and have also been impli-cated in B-cell maturation specifically via a reduction in CD40Land cytokine responsive transcription (8, 9, 20). The loss ofchromatin modifiers in multiple lymphoma subtypes indicatesthat chromatin regulation may be critical at multiple pointsalong the B-cell lineage. In contrast, EZH2 gain-of-functionmutation is found almost exclusively in follicular lymphomaand GCB-DLBCL (6), suggesting that increased EZH2 activitydoes not allow cells to progress past the germinal center. Inaddition, ABC lymphomas also have unique lesions, such asloss of PRDM1, that prevent lymphomas from progressingbeyond the plasmacytic phenotype (21).

Consistent with the importance of chromatin modificationsin B-cell maturation, recent reports have demonstrated thatEZH2 activity is required for cells to maintain the germinalcenter reaction. These experiments have shown that loss ofEZH2 prevents the formation of germinal centers in mouselymph nodes. Moreover, knock-in of gain-of-function muta-tions of EZH2 occurring primarily at the Y646 residue havebeen shown to induce or contribute to lymphomagenesis inmouse models (7, 10). It was recently discovered that treatmentof normal B-cells or DLBCL cell lines with small-moleculeinhibitors of EZH2 can induce several hallmarks of B-cellmaturation (immunoglobulin and PRDM1 induction; refs. 7,10, 22) and that B-cell maturation in the lymph node iscontrolled by the transcriptional repressive effects of BCL6 andEZH2 (23). These reports are consistent with many studies thathave explored the effects of EZH2 inhibitors in cell lines derivedfrom B-cell lymphomas. Many B-cell lymphoma cell lines areunaffected or show only a modest decrease in proliferation orcell viability in response to EZH2 inhibition, while cell lineswith mutant EZH2 generally show a more robust response thantheir wild-type counterparts (24, 25).

In this report, we explore a variety of drug combinations withtazemetostat (EPZ-6438) todeterminewhether its antilymphomaactivity can be enhanced in cancers with wild-type EZH2. Taze-metostat was previously shown to have a synergistic relationshipwith glucocorticoid receptor agonists (GRAG) in cotreatmentmodels of DLBCL (26). This work showed that cell lines withwild-type EZH2 were sensitive to the tazemetostat/GRAG com-bination, while only showing modest effects with single-agenttazemetostat. To expand on this effort, here we explore combi-nation effects with clinically and biologically relevant agents thattarget B-cell biology. We find that agents that inhibit B-cellactivation (BTKi, SYKi, GRAGs) show a strong synergistic rela-tionship with tazemetostat. In contrast, we find that tazemetostat

is less cytotoxic when combined with agents that activate B-cells(LPS, CD40L). Consistent with previous studies demonstrating arole for EZH2 in B-cell differentiation, we observe that cotreat-ment of cell lines with CD40L and tazemetostat leads to anupregulation of the B-cell maturation regulator PRDM1, whileno effect was observed for either agent individually. Taken togeth-er, these results suggest that EZH2 inhibition leads to a cell fatedecision governed by B-cell activation.

Materials and MethodsCompounds

Dexamethasone (S1322), everolimus (S1120), ibrutinib(S2680), idelalisib (S2226), MK-2206 (S1078; ref. 27), prednis-olone (S1737), R406 (tamatinib, S2194; ref. 28), and trametinib(S2673), were all obtained from Selleck Chemicals. APRIL (5860-AP), CD40 ligand (6420-CL), and BAFF/TNFSF13B (2149-BF)were all obtained from R&D Systems. LPS (L4391) was obtainedfrom Sigma-Aldrich. F(ab')2 Anti-human IgMþIgG (16-5099-85)was obtained from eBioscience. Tazemetostat was synthesized asdescribed previously (29).

Cell cultureAll cell lines used in this article were obtained from the

following sources between 2011 and 2016. No cell line waspassaged for more than 6months prior to an experiment. Cultureof diffuse large B-cell lymphoma cells WSU-DLCL2, SU-DHL-10,Toledo, and OCI-LY19 has been described previously (24, 30).KARPAS-422 (ACC 32), OCI-LY7 (ACC 688), SU-DHL-5 (ACC571), and OCI-LY3 (ACC 761) were obtained from DSMZ(German Collection of Microorganisms and Cell Cultures).SU-DHL-6 (CRL-2959), HT (CRL-2260), Farage (CRL-2630), andSU-DHL-2 (CRL-2956) were obtained from ATCC. TMD8 wasobtained from Tokyo Medical and Dental University (Tokyo,Japan). No authentication information for TMD8 is available.HT, Farage, SU-DHL-5, TMD8, KARPAS-422, SU-DHL-6, OCI-LY3, and SU-DHL-2 were maintained in RPMI supplementedwith 10%FBS and cultured in a humidified atmosphere including5% CO2. OCI-LY7 was maintained in Iscove's modified Dulbec-co's medium supplemented with 20% FBS and cultured in ahumidified atmosphere including 5% CO2. Mutations in EZH2,MYD88, and CD79B in cell lines were evaluated using the CCLE(https://portals.broadinstitute.org/ccle/home) and COSMIC(http://cancer.sanger.ac.uk/cosmic) databases.

RNA-SeqRNA-seq was executed by Expression Analysis (http://www.

q2labsolutions.com/genomicslaboratories). RNA was convertedinto cDNA libraries using the Illumina TruSeq Stranded mRNAsample preparation kit (Illumina # RS-122-2103). RNA-seq wasperformed on the Illumina HiSeq platform using 50 basepair, paired-end sequencing. Approximately 30 million readswere collected.Data qualitywas evaluated using spike-in controls.ERCC controls were from Ambion via Life Technologies. Ninety-two oligos were added to cover a 106-fold concentration range.Genome alignments of readswere performedusing STAR softwareversion2.4 andwerequantifiedusingRSEMversion1.2.14.Upperquartile normalized values were used for subsequent analysis.RNA-seq datasets were analyzed using GSEA software (BroadInstitute, Cambridge, MA). Normalized RNA-seq values fromsamples with and without tazemetostat treatment were run

Tazemetostat Creates Dependency on B-cell Activation

www.aacrjournals.org Mol Cancer Ther; 16(11) November 2017 2587

on September 20, 2020. © 2017 American Association for Cancer Research. mct.aacrjournals.org Downloaded from

Published OnlineFirst August 23, 2017; DOI: 10.1158/1535-7163.MCT-16-0840

through the algorithm. Altered genes were compared against genesets found to be induced in the transition from germinal centerB-cells to memory B-cells (31).

CHIP SeqCells were fixed with 1% formaldehyde for 15 minutes and

quenched with 0.125 mol/L glycine. Chromatin was isolated bythe addition of lysis buffer, followed by disruption with aDounce homogenizer. Lysates were sonicated and the DNAsheared to an average length of 300–500 bp. Genomic DNAwas prepared by treating aliquots of chromatin with RNase,proteinase K, and heat for decrosslinking, followed by ethanolprecipitation. Pellets were resuspended and the resulting DNAwas quantified on a NanoDrop spectrophotometer. Extrapola-tion to the original chromatin volume allowed quantitation ofthe total chromatin yield.

An aliquot of chromatin (25 or 30 mg) was precleared withprotein A agarose beads (Invitrogen). Genomic DNA regions ofinterest were isolated using antibodies against H3K4me3 (ActiveMotif, 39159), K3K27me3 (Cell Signaling Technology, 9733),and EZH2 (ActiveMotif, 39901). Complexes were washed, elutedfrom the beads with SDS buffer, and subjected to RNase andproteinase K treatment. Crosslinks were reversed by incubationovernight at 65�C, and ChIP DNA was purified by phenol–chloroform extraction and ethanol precipitation.

Quantitative PCR (qPCR) reactions were carried out intriplicate on specific genomic regions using SYBR Green Super-mix (Bio-Rad). The resulting signals were normalized for prim-er efficiency by carrying out qPCR for each primer pair usinginput DNA.

Illumina sequencing libraries were prepared from the ChIPand input DNAs by the standard consecutive enzymatic steps ofend-polishing, dA-addition, and adaptor ligation. After a finalPCR amplification step, the resulting DNA libraries were quan-tified and sequenced on Illumina NextSeq 500 (75 nt reads,single end). Reads were aligned to the human genome (hg19)using the BWA algorithm (default settings). Duplicate readswere removed and only uniquely mapped reads (mappingquality �25) were used for further analysis. Alignments wereextended in silico at their 30 ends to a length of 200 bp, which isthe average genomic fragment length in the size-selected library,and assigned to 32-nt bins along the genome. The resultinghistograms (genomic "signal maps") were stored in bigWig files.Data were visualized using UCSC genome browser (https://genome.ucsc.edu).

In vitro cellular assaysFor long-term proliferation assays, seeding densities for each

cell line are provided in Supplementary Table S1. On day 0,cells were seeded in T75 flasks and either treated with 1 mmol/Ltazemetostat, 0.25 mmol/L tazemetostat, or DMSO. Cells weresplit and reseeded every 3–4 days. Eleven-day IC50 determina-tion was performed as described previously (32). To determinethe combinatorial effects of tazemetostat with other com-pounds, lymphoma cells were treated according to a 4þ3treatment model (4-day pretreatment with tazemetostat fol-lowed by a 3-day cotreatment with tazemetostat and com-pound of interest) or in the case of Toledo cells only, to a6þ5 treatment model (6-day pretreatment with tazemetostatfollowed by 5-day cotreatment with tazemetostat and com-pound of interest). For combination studies of tazemetostat

and biologics in the SU-DHL-10 cell line, a 4-day cotreatmentmodel was established due to rapid apoptosis following single-agent tazemetostat treatment.

Lymphoma cells were seeded into flasks and pretreated withseven concentrations of tazemetostat or DMSO for 4 days (withthe exception of Toledo cells, which were pretreated for 6 daysand Farage cells, which were pretreated for 7 or 8 days). Cellswere then split and cotreated with tazemetostat and a secondagent of interest using the HP D300 digital dispenser (TecanGroup) in 96-well plates for additional 3 days, with theexception of Toledo cells, that were cotreated for 5 days.SU-DHL-10 cells were plated and treated directly in solid white96-well plates and cotreated with tazemetostat and compoundsof interest for 4 days. Doses of drugs used in these assays werenear the IC50 for each agent (Supplementary Table S2). Bothagents were serially diluted 2-fold and combined in a matrixwith constant ratios diagonally across the plate with a finalDMSO content of 0.11% (v/v). Three concentrations with a4-fold dilution were chosen to combine with tazemetostat.After 3-day cotreatment (5 days for Toledo assays), cell viabilitywas measured via ATP content using CellTiter-Glo (Promega)and luminescence was detected using a SpectraMax M5 micro-plate reader (Molecular Devices).

In vivo modelsAll of the procedures related to animal handling, care, and their

treatment in this study were performed according to the guide-lines approved by the Institutional Animal Care and Use Com-mittee (IACUC) of Shanghai Chempartner following the guid-ance of the Association for Assessment and Accreditation ofLaboratory Animal Care (AAALAC). Female CB17 SCID mice(6–8weeks old) were obtained from Shanghai Lingchang BiotechCo. A total of 1 � 106 OCI-LY19 and 5 � 106 Toledo cells weremixed with Matrigel and inoculated subcutaneously in the rightflank. Because of high animal-to-animal variability, SU-DHL-5cells were subcultured from a previously established tumor. Atotal of 5 � 106 of the explanted SU-DHL-5 cells were alsoinoculated in 50% Matrigel into the right flank. Tazemetostatwas suspended in 0.5% NaCMC with 0.1% Tween-80. Eighttumor-bearing animals were used per dose group. Animals weretreated with 125 mg/kg or 500 mg/kg twice daily. Tumor volumewas measured every three or four days for the duration of thestudy. Mice were taken off study and euthanized once tumormeasurements exceeded a volume of 2,000 mm3. Mean tumorvolumes were calculated with SEM values (n ¼ 8).

Apoptosis analysisApoptosis analysis was performed as described previously

(32). All cells were seeded at densities that ensured logarithmicgrowth for each segment of treatment. Seeding densities areshown in Supplementary Table S1. For combinations withibrutinib, KARPAS-422, Farage, SU-DHL-5, and SU-DHL-2 cellswere treated with DMSO or 1 mmol/L of tazemetostat for4 days. Cells were split and retreated with DMSO or 1 mmol/Ltazemetostat, 3 mmol/L ibrutinib, or in combination withtazemetostat and ibrutinib at those concentrations for 3 addi-tional days. Cells were harvested on days 4 and 7 to allow foranalysis of Annexin V staining. For the CD40 ligand combina-tions, KARPAS-422 cells were treated with DMSO or 1 mmol/Lof tazemetostat for 4 days. Cells were split and retreated withDMSO, 1 mmol/L tazemetostat, 100 ng/mL CD40 ligand, or

Brach et al.

Mol Cancer Ther; 16(11) November 2017 Molecular Cancer Therapeutics2588

on September 20, 2020. © 2017 American Association for Cancer Research. mct.aacrjournals.org Downloaded from

Published OnlineFirst August 23, 2017; DOI: 10.1158/1535-7163.MCT-16-0840

with the combination of both agents for 3 additional days.Cells were split one additional time in the same mannerabove for 3 more days, for a total of 10 days in treatment.Cells were harvested on days 4, 7, and 10 to allow for analysisof Annexin V staining. In apoptosis experiments, statisticalanalysis was performed using one-way ANOVA plus Bonferroniposttest where ����, P < 0.0001; ���, P < 0.001; ��, P < 0.01;�, P < 0.05; not significant P > 0.05.

Combination synergy quantitationConcentration response plots were generated in GraphPad

Prism version 7.0 for Windows, GraphPad Software and curvesfitted to a four-parameter logistic model with variable slope(two biological replicates). Fraction affected (Fa), a measure ofcompound inhibition, was calculated at each treatment con-centration by the following equation: Fa ¼ 1–((RLU cmpd –

RLU min)/(RLU max – RLU min)), where RLU ¼ relativeluminescence units, RLU cmpd ¼ value for test treatment, RLUmax is the maximum signal control (DMSO treatment), andRLUmin is the minimum signal control (positive control or fullinhibition control). The 50% inhibitory concentrations wereinterpolated from the four-parameter logistic model as theconcentration that achieved Fa ¼ 0.5. Analysis of combinatorialeffects and synergy quantification were performed by differentmethods according to the single-agent activities of the combi-nation partners: when both compounds had a calculable IC50

value then the combination index (CI) was derived by theChou–Talalay method (33) using the CalcuSyn for Windowsversion 2.1 software (Biosoft). Combinations with CI < 1demonstrated synergy (SYN), those with CI > 1, antagonism(ANT), and those with CI ¼ 1 additivity (ADD). When one ofthe combined agents did not reach an IC50, the extent of thecombination benefit was determined by maximal fold shift inthe IC50 of the other agent as a function of the concentration ofthe second agent compared with DMSO alone. Combinationbenefit or enhancement was demonstrated when the presence ofthe second agent produced a decrease of the IC50 value greaterthan 2-fold. In contrast, antagonism was demonstrated by arightward shift of the IC50 greater than 2-fold. If no fold shiftwas observed, then it was denoted as no effect (N/E). If neithercombined agent had a calculable IC50, CHALICE software(Horizon Discovery) was used to calculate the Loewe volume(VLoewe; ref. 34). A positive Loewe volume denoted synergy,and a negative Loewe volume, antagonism. A volume between�1 and 1 was denoted as additivity.

Western blottingWhole-cell lysates were prepared using RIPA buffer [10� RIPA

lysis buffer, Millipore; 20–188], supplemented with 0.1% SDS,and protease inhibitor tablet [Sigma Aldrich; 4693124001]. Cellswere pelleted, washed with ice-cold PBS, resuspended in ice-coldRIPA buffer, and incubated on ice for 5 minutes. Lysates werethen sonicated three times for 5 seconds and incubated on icefor 10 minutes. Lysates were centrifuged at maximum speed for10 minutes at 4�C. Protein concentrations were determined byBCA (Pierce; 23225). Twenty micrograms were loaded onto gelsfor PRDM1/BLIMP1/PRD1-BF1 (CST; 9115) analysis, 5 mg forH3K27me3 (CST; 9733), and 25 mg for IkBa (CST; 4814). b-Actin(Cell Signaling Technology; 3700) and GAPDH (Millipore;MAB374) were used as protein loading controls. Lysates werefractionated on 4%–12% Bis-Tris gel (Life Technologies;

WG1402BX10), transferred using iBlot (7 minutes on program3, using nitrocellulose transfer stacks), and probed antibodies inOdyssey blocking buffer (Odyssey; 927-40000).

Quantitative PCRAll cells were seeded at densities that ensured logarithmic

growth for each segment of treatment. Seeding densities can beseen in Supplementary Table S1. All cells were treated withDMSO or 1 mmol/L tazemetostat for 4 days. Cells were splitand retreated with DMSO or 1 mmol/L tazemetostat for 3 addi-tional days. Cells were harvested and total mRNA was extractedfrom cell pellets using the RNeasy Plus Mini Kit (Qiagen;74134). cDNA was made using the RT2 First Strand Kit (Qia-gen; 330401). RT-PCR was performed using ViiA 7 Real-TimePCR Systems (Applied Biosystems) TaqMan probe-basedqPCR was carried out using TaqMan Fast Advanced MasterMix (AB; 4444964) and TaqMan primer/probe sets for PRDM1(Invitrogen; Hs00153357_m1). Gene expression was normal-ized to housekeeping gene, 18S (Thermo, 4319413E), andfold change compared with DMSO was calculated using theDDCt method.

ResultsEffects of tazemetostat in wild-type versus mutant EZH2 celllines

Treatment of select B-cell lymphoma cell lines in culture withEZH2 inhibitors has been shown to decrease their proliferationand in some cases induce apoptosis. We and others havedemonstrated that cell lines bearing gain-of-function muta-tions in EZH2 are significantly more sensitive to EZH2 inhibi-tion (24, 25). As we have previously published, lymphoma celllines with mutant EZH2 are several orders of magnitude moresensitive to EZH2 inhibitors, as measured by IC50 in an 11-dayproliferation assay (Fig. 1B). However, we observe decreases inproliferative activity in the majority of B-cell cancer cell linestested, regardless of EZH2 mutation status, cell of origin, orlymphoma subtype. As shown in Fig. 1B, EZH2 wild-type celllines with canonical ABC mutations like OCI-LY3 (MYD88),TMD8 (CD79B) display similar sensitivity to tazemetostat withEZH2wild-type cell lines of GCB origin like Farage, OCI-LY7, orSU-DHL-5 (35). When comparing mutant and wild-type celllines at identical concentrations of inhibitor, we observe thatthe EZH2-mutant cell line, KARPAS-422, shows a strong cyto-toxic response at 250 nmol/L tazemetostat while cell lines withwild-type EZH2, like SU-DHL-5 (GCB), Farage (GCB), andTMD8 (ABC) show dose-dependent decreases in proliferation(Fig. 1C). Moreover, the KARPAS-422 cell line shows anincrease in apoptosis in response to tazemetostat treatment,while the three wild-type cell lines show little or none (Fig. 1D).These findings suggest that mutant EZH2 cell lines have astrong dependency on EZH2 activity for their viability, whilevery little if any EZH2 activity is required for survival of wild-type EZH2 cell lines. These findings are consistent with previouswork demonstrating that cytotoxicity occurs at very low dosesof EZH2 inhibitor in mutant cell lines, but is not observed withany relevant concentration in wild-type cells (24, 25).

While previous xenograft studies of mutant EZH2 lym-phoma models have demonstrated sensitivity to tazemetostat(30), few studies have been carried out to evaluate responsefrom models of wild-type EZH2. Three wild-type cell line

Tazemetostat Creates Dependency on B-cell Activation

www.aacrjournals.org Mol Cancer Ther; 16(11) November 2017 2589

on September 20, 2020. © 2017 American Association for Cancer Research. mct.aacrjournals.org Downloaded from

Published OnlineFirst August 23, 2017; DOI: 10.1158/1535-7163.MCT-16-0840

Pfei

ffer

KA

RPA

S 42

2SU

-DH

L-6

SU-D

HL-

10W

SU-D

LCL2 DB

OC

I-LY1

SU-D

HL-

4R

LFa

rage H

TSU

-DH

L-2

A3/

KA

WA

4/Fu

kU

-294

0O

CI-L

Y-3

MC

116

OC

I-LY7

RC

-K8

DO

HH

-2U

-293

2SU

-DH

L-5

TMD

8R

i-1O

CI-L

Y-19

Tole

do SRU

-290

4

0.0001

0.001

0.01

0.1

1

10

100

Cell linePr

olife

ratio

n IC

50 (m

icro

m o

lar)

Mutant EZH2 Wild-Type EZH2

Perc

ent o

f gat

ed

DMSODMSO

TAZTAZ

DMSODMSO

TAZTAZ

0

20

40

60

80

100

KARPAS-422 Farage SU-DHL-5 TMD8

** *

0 7 14 21 2810-2

100

102

104

106

108

Time (days) Time (days)4 10

Farage: WT EZH2

0 7 14 21 2810-2

100

102

104

106

108

Time (days) Time (days)

Viab

le c

ells

per

mL

Viab

le c

ells

per

mL

Viab

le c

ells

per

mL

Viab

le c

ells

per

mL

4 10

KARPAS-422:Mutant EZH2

0 7 14 21 2810-5

10-3

10-1

101

103

105

107

109

4 10

SU-DHL-5: WT EZH2

0 7 14 21 28104

106

108

1010

1012

1014TMD8: WT EZH2ABC

4 10

OCI-LY19

Day Day Days0 5 10 15 20 25

0

500

1,000

1,500

2,000

2,500

SU-DHL-5

0 5 10 15 20

Toledo

0 10 20 30 400

1,000

2,000

3,000

4,000

A B

C D

E

***

Tum

or v

olum

e (m

m3 )

0

1,000

2,000

3,000

4,000

Tum

or v

olum

e (m

m3 )

Tum

or v

olum

e (m

m3 )

DMSO 0.06 µmol/L TAZ 0.25 µmol/L TAZ 1 µmol/L TAZ

Vehicle TAZ 125 mg/kg PO BID TAZ 500 mg/kg PO BID

Nuclear debris annexin(–) 7-AAD(+)Late stage apoptotic annexin(+) 7-AAD(+)Early apoptotic annexin(+) 7-AAD(–)Nonapoptotic annexin(–) 7-AAD(–)

O

O O

ONN

HN

HN

Figure 1.

Lymphoma cell lines with EZH2 mutation have a cytotoxic response to EZH2 inhibition with tazemetostat, while wild-type cell lines show a cytostatic response.A, Chemical structure of tazemetostat (EPZ-6438). B, EZH2 gain-of-function mutant cell lines are more sensitive to tazemetostat (TAZ) than EZH2 wild-typecell lines in vitro. Eleven-day proliferation IC50 values for 19 wild-type DLBCL cell lines (left) and 9 mutant DLBCL cell lines (right) shown on a log plot.Proliferation was assessed by viable cell counts. C, Mutant EZH2 KARPAS-422 cell line, wild-type EZH2 GCB cell lines Farage and SU-DHL-5 and the ABCcell line, TMD8 cell line were treated with tazemetostat over a 42-day time course show varying degrees of antiproliferative responses. Cell number was measured atthe indicated time points. The viable cell count (y-axis) in each panel is presented on a logarithmic scale. D, Early-stage apoptotic cell death is observed inmutant EZH2 KARPAS-422 cells but not in wild-type cell lines treated for 10 days with 1 mmol/L tazemetostat or DMSO as measured by Annexin-V. Green stacksrepresent percentages of cells in early-stage apoptosis (means � SD, n ¼ 2). �� , P < 0.01 (ANOVA plus Bonferroni posttest) KARPAS-422 cells treated withtazemetostat compared with DMSO, � , P < 0.1 (ANOVA plus Bonferroni posttest) Farage cells treated with tazemetostat compared with DMSO. E, OCI-LY19,SU-DHL-5, and Toledo cell lines xenograft-bearing mice were treated with vehicle, 125 or 500 mg/kg twice daily tazemetostat by oral gavage for theindicated times. Eight tumor-bearing SCID mice per group were treated with the indicated doses of tazemetostat once tumors reached a minimum volume of100 mm2. Mean tumor volumes � SEM (n ¼ 8) are plotted. Endpoint was reached when tumor volume exceeded 2,000 mm2. � , P < 0.01; ��, P < 0.05.

Brach et al.

Mol Cancer Ther; 16(11) November 2017 Molecular Cancer Therapeutics2590

on September 20, 2020. © 2017 American Association for Cancer Research. mct.aacrjournals.org Downloaded from

Published OnlineFirst August 23, 2017; DOI: 10.1158/1535-7163.MCT-16-0840

xenograft models were developed to study this in greater depth.Animals with Toledo, SU-DHL-5, and OCI-LY19 cell line xeno-grafts were treated with vehicle, or 125 or 500 mg/kg twice dailytazemetostat for their respective tumor growth windows. Asshown in Fig. 1E, significant dose-dependent tumor growthinhibition was observed in the OCI-LY19 model. However, theToledo and SU-DHL-5 models showed only modest tumorgrowth inhibition that could not be readily differentiated fromanimal-to-animal variability from the vehicle control. Inhibi-tion of H3K27me3 levels in SU-DHL-5 tumors and animalbody weight changes are shown in Supplementary Fig. S1,demonstrating that dosing is sufficient and tolerated for thismouse strain. In contrast, previous work has shown that taze-metostat induces robust tumor regressions in multiple tumormodels bearing EZH2 gain-of-function mutations (30).

Tazemetostat creates a dependency on B-cell activationsignaling

The phenotypic differences in response to tazemetostatbetween wild-type and mutant EZH2 suggest that wild-typecell lines are able to compensate for loss of EZH2 activity andmaintain viability. We further explored this observation byevaluating various agents for combination benefit with taze-metostat. Consistent with our previously published findings(26), we observed synergistic interactions between glucocorti-coids and tazemetostat treatment in the majority of cell linestested. While glucocorticoids have a number of effects onlymphoma cells, one of these effects is a reduction in NFkBsignaling (36) which is critical for B-cell activation. On the basisof this finding, we treated lymphoma cells with kinase inhibi-tors that target the B-cell activation pathway at various pointsalong the B-cell receptor signaling cascade. Figure 2 shows theeffects of combinations of tazemetostat with ibrutinib (BTKinhibitor) or tamatinib (SYK inhibitor). Both of these inhibi-tors target signal transduction through the B-cell receptorpathway. We observed synergy between these agents and taze-metostat suggesting a relationship between EZH2 and B-cellactivation. Moreover, agents targeting additional nodes of theB-cell receptor pathway, including PI3K, AKT, mTOR, andMEK1 (Table 1) were also found to be synergistic with taze-metostat. The synergistic relationship between the B-cell recep-tor pathway and EZH2 inhibition suggests that loss of EZH2activity makes cells more dependent on B-cell activation signalsfor survival. To test this more directly, we sought to determine ifby artificially stimulating B-cell activation, the effects of taze-metostat would be antagonized. We found that addition of LPS,BAFF, APRIL, or CD40L (all activators of B-cells) was able toantagonize the effects of EZH2 inhibition in multiple cell lines(Fig. 2A; Table 1). It is important to note that extendedtreatment times were used for Toledo and Farage due to theirgrowth rates and drug combinations in the SU-DHL-10 cell linewere added simultaneously for only 4 days because of a rapidonset of apoptosis in response to single-agent tazemetostat,providing a caveat to direct comparisons between all cell lines.CD40L showed the most consistent effect across cell lines, andimportantly treatment with CD40L alone had no effect onproliferation of the majority of the cell lines (SupplementaryFig. S2). These results suggest that much of the cytotoxic andgrowth-inhibitory effects of EZH2 inhibitor are due to insuf-ficient compensatory B-cell activation signaling. To furtherexplore this observation, RNA-seq was performed on Farage,

SU-DHL-5, and KARPAS-422 cell lines treated with 1 mmol/Ltazemetostat for 4 days. Gene set enrichment analysis (GSEA)comparison of cell lines with and without tazemetostat treat-ment shows a decrease in expression of a BCR-responsive geneset (37) following EZH2 inhibition (Fig. 2B) suggesting thatEZH2 alters how cells respond to this signaling pathway.

Inhibition of both EZH2 and the B-cell receptor pathwayinduces apoptosis in wild-type cell lines

The strong synergy observed between B-cell receptor pathwayinhibitors and tazemetostat in cell viability assays suggested thatcombination of these agents may be inducing a cytotoxic effectrather than the cytostatic effect observed with single agent. Totest this hypothesis, KARPAS-422, SU-DHL-5, and Farage cells,cotreated with tazemetostat and ibrutinib, were stained withAnnexin V and 7-AAD to evaluate apoptosis. Combination oftazemetostat and ibrutinib in all three cell lines showed asignificant increase in apoptosis than what was observed foreither single-agent treatment (Fig. 3A–C). The most significantincrease in apoptosis was observed in SU-DHL-5 cells that arenot sensitive to either agent individually. To determine whetherB-cell activation can suppress the proapoptotic effect of taze-metostat in a mutant cell line, we measured apoptosis inKARPAS-422 cells cotreated with tazemetostat and CD40L.Addition of CD40L greatly reduced tazemetostat-induced apo-ptosis (Fig. 3D), while having no effect on the ability oftazemetostat to reduce the levels of H3K27me3 (Fig. 3E). Wealso examined whether upstream signaling through CD40 isaffected by tazemetostat treatment; in Supplementary Fig. S3,we show that CD40L-induced IkBa degradation is not affectedby tazemetostat treatment suggesting that EZH2 does notimpact the CD40 signal transduction machinery.

Tazemetostat induces molecular events of B-celldifferentiation

To better understand the effects of tazemetostat in wild-type EZH2 lymphomas, RNA-seq studies were performed onmutant EZH2 GCB, wild-type EZH2 GCB, and wild-type EZH2ABC cell lines with and without EZH2 inhibitor treatment.Consistent with previous reports (7, 10), gene set enrichmentanalysis showed that genes associated with B-cell maturationare induced upon EZH2 inhibition. We find that enrichment inmemory cell gene sets (31) can be observed regardless ofEZH2 mutation status, cell of origin, or in vitro IC50. As shownin Fig. 4A, KARPAS-422 (GCB EZH2 Mut), Farage (GCB EZH2WT), SU-DHL-5 (GCB EZH2 WT), and TMD8 (ABC EZH2 WT)show increases in expression of this gene set. In addition, wealso observed increases in the B-cell maturation regulator andtumor suppressor, PRDM1/BLIMP1. PRDM1 message (Fig. 4B)and protein (Fig. 4C) were increased most strongly in theEZH2-mutant cell line KARPAS-422, but increases were alsoobserved in SU-DHL-5 and Farage. No increase in PRDM1 wasobserved in the TMD8 cell line, as PRDM1 is already expressed(Fig. 4B and C), which is common in ABC lymphomas.However, our results suggest that the role of EZH2 in B-cellmaturation is not limited to PRDM1 regulation, as the upre-gulation of the maturation gene expression signature isobserved in TMD8s despite no observable change in PRDM1(Fig. 4A–C). Consistent with the upregulation of PRDM1, agene set corresponding to ABC-DLBCL (38) was induced in thethree GCB cell lines tested Supplementary Fig. S4, indicating

Tazemetostat Creates Dependency on B-cell Activation

www.aacrjournals.org Mol Cancer Ther; 16(11) November 2017 2591

on September 20, 2020. © 2017 American Association for Cancer Research. mct.aacrjournals.org Downloaded from

Published OnlineFirst August 23, 2017; DOI: 10.1158/1535-7163.MCT-16-0840

molecular events of maturation were occurring. Not surpris-ingly, this gene set was not increased in the ABC-DLBCL cellline TMD8 as these genes were likely already elevated, as TMD8is an ABC-derived cell line.

Similar to a previous report (7), we observe that like manyother genes involved in B-cell maturation, the PRDM1 locus isbivalent. In Fig. 4D, we show CHIP-Seq data from the PRDM1

locus in the Farage cell line that shows that EZH2, H3K27me3,and H3K4me3 are all present at the PRDM1 promoter. Tofurther evaluate the repression of PRDM1 by wild-type EZH2,we explored the effect of tazemetostat on known regulatorsof PRDM1 expression in the SU-DHL-5 cell line. One suchinducer, CD40L, is an important mechanism for T follicularhelper cells to stimulate B-cell maturation and PRDM1

Figure 2.

DLBCL cell lines become dependent on activation signals upon treatment with tazemetostat. A, SU-DHL-5 (EZH2 WT, GCB), KARPAS-422 (EZH2 mutant,GCB), and Farage (EZH2 WT, GCB) cell lines were pretreated with tazemetostat (TAZ) for 4 days followed by cotreatment with the indicated doses ofmodulators of B-cell signaling for additional 3 days. All dose–response curves were plotted as fraction affected versus the logarithmic concentration oftazemetostat. Synergy was quantified for each combination by the Loewe method using the Chalice software. The calculated Loewe volume (VLoewe) foreach combination and the 3-day single-agent IC50 value of the combined drug are indicated on the respective graphs. All analysis methodologies aredescribed in Materials and Methods. Synergistic antiproliferative activity was observed when tazemetostat was combined with ibrutinib or tamatinib in all celllines. CD40L in combination with tazemetostat showed antagonism in KARPAS-422 and SU-DHL-5 but had no effect in Farage. B, BCR ligation–responsivegene sets decrease in response to tazemetostat treatment. GSEA was performed on RNA-seq data from the indicated cell lines with and without tazemetostattreatment for a gene set corresponding to genes upregulated following anti-BCR ligation in normal B-Cells (37). Anti-BCR–induced genes weredecreased in all three cell lines as indicated by normalized enrichment scores. False discovery rate (FDR) < 0.001 for each comparison.

Brach et al.

Mol Cancer Ther; 16(11) November 2017 Molecular Cancer Therapeutics2592

on September 20, 2020. © 2017 American Association for Cancer Research. mct.aacrjournals.org Downloaded from

Published OnlineFirst August 23, 2017; DOI: 10.1158/1535-7163.MCT-16-0840

expression via NFkB activation. We therefore assessed recom-binant CD40L-mediated PRDM1 induction in the SU-DHL-5cell line in the presence or absence of tazemetostat. As seenin Fig. 4E, tazemetostat has only a modest effect on PRDM1levels in the SU-DHL-5 line as a single agent and neither CD40Lnor IL21 affect PRDM1 protein levels in SU-DHL-5 cells(Supplementary Fig. S5), while they have been shown to inducePRDM1 expression in normal B-cells (39). However, weobserve that treatment with combinations of CD40L andtazemetostat strongly upregulate PRDM1 in SU-DHL-5s (Fig.4E). While its importance in normal B-cell differentiation hasbeen established, addition of IL21 to tazemetostat treatment inthe SU-DHL-5 cell line had no additional effect on PRDM1protein levels (Supplementary Fig. S5). These findings suggestthat EZH2 suppresses PRDM1 levels in response to B-cellactivation, which may be important for the antilymphomaeffects of tazemetostat. In addition, gene set enrichment anal-ysis showed an increase in CD40L-response genes with taze-metostat treatment alone in KARPAS-422, Farage, SU-DHL-5,and TMD8 cell lines (Supplementary Fig. S6). This is consistentwith previous reports demonstrating that EZH2 suppressesCD40L and other B-cell maturation–related pathways at thelevel of transcription (7).

DiscussionAs described in the work presented here and in other studies

(24, 25), B-cell lymphoma cell line models bearing EZH2 muta-tions show a strong cytotoxic response to EZH2 inhibitors. Incontrast, wild-type cell lines show much less dependence onEZH2 activity for survival and only a cytostatic response totreatment. These findings suggest that wild-type EZH2 lympho-mas have additional active pathways that support their growthand survival. A previous study has demonstrated a synergisticrelationship between tazemetostat and glucocorticoids in bothwild-type and mutant EZH2 DLBCL cell lines (26). We presenthere that in addition to glucocorticoids, agents that selectivelytarget the B-cell receptor pathway, ibrutinib (BTKi) and tamatinib(SYKi) also have a synergistic relationship with tazemetostat.Moreover, several other inhibitors that target kinases involved inB-cell receptor signaling (AKT, PI3K, mTOR, and MEK1) alsoshow this effect. These findings suggest that EZH2 inhibitioncreates a dependency on B-cell activation signaling for cell via-bility. This concept is further illustrated by the finding that theeffects of single-agent tazemetostat can be diminished by artifi-cially activating B-cell lymphoma lines with CD40L, BAFF, LPS, orBCR ligation.

On the basis of previous studies describing the role of EZH2in B-cell differentiation (7, 22, 25), we speculated that thealtered dependence on B-cell activation signaling observed inlymphoma cell lines is due to progression through the B-cellmaturation pathway. Previous studies have demonstrated thatEZH2 is essential for the maintenance of germinal center B-cellsand that the key oncogenic effect of EZH2 gain-of-functionmutations is to delay or block progression through this step ofB-cell maturation. We demonstrate that while no significantcytotoxicity is observed in cell lines with wild-type EZH2, genesets associated with memory B-cells and the B-cell differenti-ation transcription factor PRDM1/BLIMP1 are upregulatedfollowing tazemetostat treatment. In addition, we demonstratethat in the EZH2 wild-type cell line Farage, the PRDM1Ta

ble

1.In

vitroco

mbinatorial

effectsoftazemetostat

withinhibitors

oftheB-cellreceptorpathw

ayan

dmodulators

ofB-cellactivation

Mutan

tEZH2GCB

WTEZH2GCB

WTEZH2ABC

Modality

Drugna

me

WSU

-DLC

L2KARPAS-422

^SU-D

HL-10

SU-D

HL-6

OCI-LY

19HT

#Farag

eOCI-LY

7SU

-DHL-5

Toledo

TMD8

OCI-LY

3SU

-DHL-2

GRAG

Prednisolone

SYN

SYN

SYN

SYN

SYN

N/E

SYN

SYN

SYN

N/E

SYN

SYN

SYN

Dexam

etha

sone

SYN

SYN

SYN

SYN

SYN

N/E

SYN

SYN

SYN

N/E

SYN

SYN

SYN

B-cell

Recep

tor

pathw

ay

Eve

rolim

usSYN

ADD

SYN

SYN

N/E

SYN

SYN

SYN

SYN

N/E

SYN

SYN

SYN

Trametinib

SYN

SYN

SYN

SYN

N/E

N/E

SYN

SYN

SYN

N/E

SYN

SYN

N/E

MK-220

6SYN

SYN

SYN

SYN

SYN

SYN

SYN

SYN

SYN

N/E

SYN

SYN

N/E

Ibrutinib

SYN

SYN

SYN

SYN

N/E

SYN

SYN

SYN

SYN

N/E

SYN

ADD

ADD

Idelalisib

SYN

SYN

SYN

SYN

N/E

SYN

SYN

SYN

SYN

N/E

SYN

SYN

N/E

Tam

atinib

SYN

SYN

SYN

SYN

N/E

SYN

SYN

SYN

SYN

N/E

SYN

SYN

SYN

B-cell

Activators

CD40L

ANT

ANT

ANT

ANT

N/E

ANT

N/E

ANT

ANT

SYN

SYN

N/E

ANT

LPS

ANT

N/E

N/E

N/E

N/E

N/E

N/E

ANT

ANT

SYN

ANT

N/E

N/E

BAFF

ANT

N/E

N/E

ANT

N/E

ANT

N/E

N/E

ANT

N/E

ANT

N/E

N/E

APRIL

ANT

N/E

N/E

ANT

N/E

N/E

N/E

N/E

ANT

N/E

ANT

N/E

N/E

Anti-IgM/IgG

ANT

ANT

N/E

SYN

N/E

N/E

N/E

N/E

N/E

N/E

SYN

N/E

N/E

IC50Tazem

etostat

7day

100nm

ol/L

100nm

ol/L

50nm

ol/L

100nm

ol/L

1mmol/L

500nm

ol/L

1mmol/L

1.5mmol/L

3.8mmol/L

>10mmol/L

5.5mmol/L

5.1mmol/L

1.3mmol/L

IC50Tazem

etostat

14day

8.6

nmol/L

1.8nm

ol/L

5.8nm

ol/L

4.7

nmol/L

6.2

mmol/L

140nm

ol/L

99nm

ol/L

1.4mmol/L

2.1mmol/L

7.6mmol/L

2.6mmol/L

820

nmol/L

230nm

ol/L

NOTE:Invitroassayswere

perform

edintheindicated

modelswith4-day

pretrea

tmen

twithtaze

metostat,follo

wed

by3day

softheindicated

combinationun

less

otherwiseindicated

.Syn

ergy(SYN),an

tagonism

(ANT),ad

ditivity(A

DD)an

dnoco

mbinationeffect

(N/E

)were

assigned

bydifferent

methodsacco

rdingto

thesingle-agen

tactivity

oftazemetostat

andco

mbinationpartners

asdescribedin

theMaterialsan

dMethodssection.

^ SU-D

HL-10

and#Farag

ece

lllin

esused

different

dosingsche

dulesforco

mbinations

(see

Materialsan

dMethodsSection).

Tazemetostat Creates Dependency on B-cell Activation

www.aacrjournals.org Mol Cancer Ther; 16(11) November 2017 2593

on September 20, 2020. © 2017 American Association for Cancer Research. mct.aacrjournals.org Downloaded from

Published OnlineFirst August 23, 2017; DOI: 10.1158/1535-7163.MCT-16-0840

promoter is enriched for H3K27me3 and H3K4me3 suggestingthat it is bivalent and poised for expression. We also note thatin the SU-DHL-5 cell line, combination of CD40L and taze-metostat induces PRDM1 expression while neither treatmentaffects PRDM1 levels as a single agent. These findings suggestthat EZH2 suppresses PRDM1 expression and that with theappropriate activation stimulus and decrease in EZH2 activity;its expression can be induced. As B-cells mature, their require-ments for activation signals change (15). This concept is con-sistent with the observation that addition of CD40L or other B-cell –activating agents can support viability in cell lines thathave an otherwise cytotoxic response to tazemetostat. Thesynergistic cytotoxic effect of tazemetostat with inhibitors ofB-cell activation signaling and the synergistic induction of

PRDM1 with CD40L suggest that EZH2 inhibition "opens thedoor" for a cell fate decision for B-cell lymphomas regardlessof EZH2 mutation status. This model is illustrated in Supple-mentary Fig. S7.

The molecular events of differentiation observed in vitrofollowing tazemetostat treatment in wild-type EZH2 lympho-ma cell lines and the effects of EZH2 inhibition on normalB-cells in mice suggest that tazemetostat may have beneficialeffects in lymphomas with either wild-type or mutant EZH2.Our findings show examples of molecular events of differen-tiation in cell lines with GCB characteristics but many of thosechanges also occur in an ABC cell line following tazemetostattreatment. This suggests that the effects of EZH2 on B-cellmaturation may not be limited to a single cell of origin. But

SU-DHL-5: Ibrutinib

0

20

40

60

80

100

120****

KARPAS-422: Ibrutinib

Perc

ent o

f gat

edPe

rcen

t of g

ated

Perc

ent o

f gat

edPe

rcen

t of g

ated

DMSODMSO

TAZ 1 mm

ol/L

TAZ 1 mm

ol/L

TAZ + Ibru

t

TAZ + Ibru

t

Ibrutin

ib 3 mm

ol/L

DMSODMSO

TAZ 1 mm

ol/L

TAZ 1 mm

ol/L

TAZ + Ibru

t

TAZ + CD40

L

CD40L 10

0 ng/m

L

Ibrutin

ib 3 mm

ol/L

Ibrutin

ib 0.7 m

mol/L0

20

40

60

80

100

120**

Farage: Ibrutinib

0

20

40

60

80

100

120**

KARPAS-422: CD40L

0

20

40

60

80

100

120*

A B

C D

Day 4 Day 7

H3K27me3

Total H3

DMSO TAZ CD40L Combo DMSO TAZ CD40L Combo

ENuclear debris annexin(–) 7-AAD(+)Late stage apoptotic annexin(+) 7-AAD(+)Early apoptotic annexin(+) 7-AAD(–)Nonapoptotic annexin(–) 7-AAD(–)

Figure 3.

Inhibitors of B-cell activation cooperate with tazemetostat (TAZ) to induce apoptotic cell death in GCB cells. A–C, Cells were treated for 7 days with singleagents (1 mmol/L tazemetostat or 3 mmol/L ibrutinib) or in combination. A significant increase of apoptosis was observed with the combination oftazemetostat and ibrutinib in GCB cells lines compared with the single agents (��, P < 0.01; ��� , P < 0.001; ���� , P < 0.0001 (ANOVA plus Bonferroni posttest.D, KARPAS-422 cells were treated for 10 days with single agents (1 mmol/L tazemetostat or 100 ng/mL CD40L) or in combination. A significantincrease of apoptosis is seen with single-agent tazemetostat compared with DMSO and a significant decrease of apoptosis is seen with the combinationof TAZ and CD40L compared with tazemetostat single-agent treatment [� , P < 0.1 (ANOVA plus Bonferroni posttest)]. E, KARPAS-422 cells weretreated with tazemetostat and CD40L for 4 or 7 days (1 mmol/L tazemetostat, 100 ng/mL CD40L, and DMSO control) or in combination. CD40 liganddid not affect H3K27 trimethylation as a single agent or in combination with tazemetostat.

Brach et al.

Mol Cancer Ther; 16(11) November 2017 Molecular Cancer Therapeutics2594

on September 20, 2020. © 2017 American Association for Cancer Research. mct.aacrjournals.org Downloaded from

Published OnlineFirst August 23, 2017; DOI: 10.1158/1535-7163.MCT-16-0840

PRDM1

b-Ac�n

TAZ Alone TAZ + CD40L LP1

0.04

2.5

100.16

0.6

0.01

DMSO

0.04

2.5

100.16

0.6

0.01

DMSO

(mmol/L) Tazemetostat

Unt

Farage SU-DHL-5 TMD8

DMSO

TAZ

DMSO

TAZ

DMSO

TAZ

DMSO

TAZ

Farage, WT GCB0.099 µmol/L IC50NES = 1.57 FDR <0.001

SU-DHL-5, WT GCB

1 mmol/L TAZ 1 mmol/L TAZ 1 mmol/L TAZ 1 mmol/L TAZ

2 µmol/L IC50NES = 1.4FDR <0.01

KARPAS-422, mut0.0018 µmol/L IC50NES = 1.4 FDR < 0.01

DMSO – EZH2

DMSO – H3K4me3

DMSO – H3K27me3

TAZ – H3K27me3

KARPAS-422

TMD8 WT ABC 2.6 µmol/L IC50NES = 1.67 FDR<0.001

BA

C

D

E

Fold

cha

nge

over

DM

SO

25020015020

15

10

5

0

PRDM1

KARPAS-422 Farage SU-DHL-5 TMD8

Figure 4.

Tazemetostat induces memory cell gene sets and PRDM1 (BLIMP1) expression independently of EZH2 mutation status or tazemetostat sensitivity. A, EZH2inhibition induces a gene set that is heavily enriched for memory cell genes in DLBCL cell lines. RNA-seq data from KARPAS-422 (EZH2-mutant GCB),Farage (EZH2 WT GCB), SU-DHL-5 (EZH2 WT GCB), and TMD8 (EZH2 WT ABC) treated with 1 mmol/L tazemetostat (TAZ) for 4 days was analyzed usingGSEA (Broad Institute) for enrichment of the GSE11961 memory B-cell versus germinal center B-cell gene set (31). NES and FDR scores are indicated. B and C,KARPAS-422 (EZH2-mutant GCB), Farage (EZH2 WT GCB), SU-DHL-5 (EZH2 WT GCB), and TMD8 (EZH2 WT ABC) cell lines were treated for 7 dayswith 1 mmol/L tazemetostat or DMSO. PRDM1 (BLIMP1) expression was evaluated by qPCR (B) or Western blot (C) relative to DMSO control. D, CHIP-Seq wasperformed on cell pellets from Farage cells with or without a 1 mmol/L tazemetostat 4-day treatment. Antibodies recognizing EZH2, H3K4me3, andH3K27me3 were used. CHIP-Seq traces from the PRDM1 locus are shown. E, SU-DHL-5 (WT EZH2) cells treated for 14 days with tazemetostat (6 concentrationsstarting at 10 mmol/L, 4-fold dilution) in combination with 100 ng/mL of CD40L leads to enhanced upregulation of PRDM1 compared with tazemetostat as asingle agent as measured by Western blot analysis. Na€�ve LP1 (multiple myeloma derived) cells were used as positive control for PRDM1 expression.

Tazemetostat Creates Dependency on B-cell Activation

www.aacrjournals.org Mol Cancer Ther; 16(11) November 2017 2595

on September 20, 2020. © 2017 American Association for Cancer Research. mct.aacrjournals.org Downloaded from

Published OnlineFirst August 23, 2017; DOI: 10.1158/1535-7163.MCT-16-0840

more importantly, the effects of EZH2 inhibitors on cell fateand the synergistic effects observed with inhibitors of B-cellactivation in vitro provide multiple avenues for drug combina-tions in the future. To that end, tazemetostat is currentlybeing evaluated as monotherapy in clinical studies that haveincluded patients with either wild-type or activating EZH2mutations. Patients enrolled in these studies have also includedthose with follicular lymphoma, transformed follicularlymphoma, GCB-DLBCL, ABC-DLBCL, marginal zone lympho-ma and primary mediastinal B-cell lymphoma. In addition,there are also ongoing combination studies with R-CHOP andwith prednisolone.

Disclosure of Potential Conflicts of InterestD. Brach has ownership interest (including patents) in Epizyme.

D. Johnston-Blackwell has ownership interest (including patents)in Epizyme, Inc. C. Plescia is a principal research associate and hasownership interest (including patents) in Epizyme. J.J. Smith has own-ership interest (including patents) in EPZM. R.A. Copeland has owner-ship interest (including patents) in Epizyme, Inc. H. Keilhack has own-ership interest (including patents) in Epizyme Inc. S.K. Knutson hasownership interest (including patents) in patents and shares. M.J. Tho-menius is a principal scientist and has ownership interest (includingpatents) in Epizyme, Inc. No potential conflicts of interest were disclosedby the other authors.

Authors' ContributionsConception and design:D. Brach, D. Johnston-Blackwell, C. Plescia, J.J. Smith,R.A. Copeland, H. Keilhack, E. Chan-Penebre, S.A. Ribich, A. Raimondi,M.J. ThomeniusDevelopment of methodology: D. Brach, D. Johnston-Blackwell, N.M.Warholic, I. Feldman, C. Plescia, S.K. Knutson, A. Raimondi, M.J. ThomeniusAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.):D. Brach, D. Johnston-Blackwell, A. Drew, T. Lingaraj,V. Motwani, N.M. Warholic, S.K. Knutson, A. RaimondiAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): D. Brach, D. Johnston-Blackwell, T. Lingaraj,V. Motwani, I. Feldman, C. Plescia, J.J. Smith, R.A. Copeland, S.A. Ribich,A. Raimondi, M.J. ThomeniusWriting, review, and/or revision of the manuscript: D. Brach, A. Drew,J.J. Smith, R.A. Copeland, S.A. Ribich, A. Raimondi, M.J. ThomeniusAdministrative, technical, or material support (i.e., reporting or organiz-ing data, constructing databases): D. Johnston-Blackwell, M.J. ThomeniusStudy supervision: J.J. Smith, H. Keilhack, S.K. Knutson, S.A. Ribich,A. Raimondi, M.J. Thomenius

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

Received December 5, 2016; revised May 4, 2017; accepted August 16, 2017;published OnlineFirst August 23, 2017.

References1. Gonzalez-Perez A, Jene-Sanz A, Lopez-Bigas N. The mutational landscape

of chromatin regulatory factors across 4,623 tumor samples. Genome Biol2013;14:r106.

2. Chen T, Dent SY. Chromatin modifiers and remodellers: regulators ofcellular differentiation. Nat Rev Genet 2014;15:93–106.

3. Comet I, Riising EM, Leblanc B, Helin K. Maintaining cell identity: PRC2-mediated regulation of transcription and cancer. Nat Rev Cancer 2016;16:803–10.

4. Lunning MA, Green MR. Mutation of chromatin modifiers; an emerg-ing hallmark of germinal center B-cell lymphomas. Blood Cancer J2015;5:e361.

5. Lohr JG, Stojanov P, Lawrence MS, Auclair D, Chapuy B, Sougnez C, et al.Discovery and prioritization of somatic mutations in diffuse large B-celllymphoma (DLBCL) bywhole-exome sequencing. ProcNatl Acad Sci U S A2012;109:3879–84.

6. Morin RD, Mendez-Lago M, Mungall AJ, Goya R, Mungall KL, Corbett RD,et al. Frequent mutation of histone-modifying genes in non-Hodgkinlymphoma. Nature 2011;476:298–303.

7. Beguelin W, Popovic R, Teater M, Jiang Y, Bunting KL, Rosen M, et al.EZH2 is required for germinal center formation and somatic EZH2mutations promote lymphoid transformation. Cancer Cell 2013;23:677–92.

8. Ortega-Molina A, Boss IW, Canela A, Pan H, Jiang Y, Zhao C, et al. Thehistone lysine methyltransferase KMT2D sustains a gene expression pro-gram that represses B cell lymphoma development. Nat Med 2015;21:1199–208.

9. Zhang J, Dominguez-Sola D, Hussein S, Lee JE, Holmes AB, BansalM, et al.Disruption of KMT2D perturbs germinal center B cell development andpromotes lymphomagenesis. Nat Med 2015;21:1190–8.

10. CaganovaM, Carrisi C, Varano G,Mainoldi F, Zanardi F, Germain PL, et al.Germinal center dysregulation by histone methyltransferase EZH2 pro-motes lymphomagenesis. J Clin Invest 2013;123:5009–22.

11. Klein U, Dalla-Favera R. Germinal centres: role in B-cell physiology andmalignancy. Nat Rev Immunol 2008;8:22–33.

12. Basso K, Dalla-Favera R. BCL6: master regulator of the germinal centerreaction and key oncogene in B cell lymphomagenesis. Adv Immunol2010;105:193–210.

13. Peng SL. Signaling in B cells via Toll-like receptors. Curr Opin Immunol2005;17:230–6.

14. DavisMR, Zhu Z,HansenDM, BaiQ, Fang Y. The role of IL-21 in immunityand cancer. Cancer Lett 2015;358:107–14.

15. Basso K, Dalla-Favera R. Germinal centres and B cell lymphomagenesis.Nat Rev Immunol 2015;15:172–84.

16. Hatzi K, Melnick A. Breaking bad in the germinal center: how dereg-ulation of BCL6 contributes to lymphomagenesis. Trends Mol Med2014;20:343–52.

17. Duan S, Cermak L, Pagan JK, Rossi M, Martinengo C, di Celle PF, et al.FBXO11 targets BCL6 for degradation and is inactivated in diffuse largeB-cell lymphomas. Nature 2012;481:90–3.

18. Green MR, Vicente-Duenas C, Romero-Camarero I, Long Liu C, Dai B,Gonzalez-Herrero I, et al. Transient expression of Bcl6 is sufficientfor oncogenic function and induction of mature B-cell lymphoma.Nat Commun 2014;5:3904.

19. Cardenas MG, Yu W, Beguelin W, Teater MR, Geng H, Goldstein RL, et al.Rationally designed BCL6 inhibitors target activated B cell diffuse largeB cell lymphoma. J Clin Invest 2016;126:3351–62.

20. Zhang J, Vlasevska S, Wells VA, Nataraj S, Holmes AB, Duval R, et al.The Crebbp Acetyltransferase is a haploinsufficient tumor suppressorin B cell lymphoma. Cancer Discov 2017;7:322–37.

21. Mandelbaum J, Bhagat G, Tang H, Mo T, Brahmachary M, Shen Q,et al. BLIMP1 is a tumor suppressor gene frequently disrupted inactivated B cell-like diffuse large B cell lymphoma. Cancer Cell 2010;18:568–79.

22. Qi W, Chan H, Teng L, Li L, Chuai S, Zhang R, et al. Selective inhibitionof Ezh2 by a small molecule inhibitor blocks tumor cells proliferation.Proc Natl Acad Sci U S A 2012;109:21360–5.

23. Beguelin W, Teater M, Gearhart MD, Calvo Fernandez MT, Goldstein RL,Cardenas MG, et al. EZH2 and BCL6 cooperate to assemble CBX8-BCORcomplex to repress bivalent promoters,mediate germinal center formationand lymphomagenesis. Cancer Cell 2016;30:197–213.

24. Knutson SK, Wigle TJ, Warholic NM, Sneeringer CJ, Allain CJ, Klaus CR,et al. A selective inhibitor of EZH2 blocks H3K27 methylation and killsmutant lymphoma cells. Nat Chem Biol 2012;8:890–6.

25. McCabe MT, Ott HM, Ganji G, Korenchuk S, Thompson C, Van Aller GS,et al. EZH2 inhibition as a therapeutic strategy for lymphoma with EZH2-activating mutations. Nature 2012;492:108–12.

26. Knutson SK, Warholic NM, Johnston LD, Klaus CR, Wigle TJ, IwanowiczD, et al. Synergistic anti-tumor activity of EZH2 inhibitors and

Brach et al.

Mol Cancer Ther; 16(11) November 2017 Molecular Cancer Therapeutics2596

on September 20, 2020. © 2017 American Association for Cancer Research. mct.aacrjournals.org Downloaded from

Published OnlineFirst August 23, 2017; DOI: 10.1158/1535-7163.MCT-16-0840

glucocorticoid receptor agonists in models of germinal center non-Hodgkin lymphomas. PLoS One 2014;9:e111840.

27. Hirai H, Sootome H, Nakatsuru Y, Miyama K, Taguchi S, Tsujioka K, et al.MK-2206, an allosteric Akt inhibitor, enhances antitumor efficacy bystandard chemotherapeutic agents or molecular targeted drugs in vitroand in vivo. Mol Cancer Ther 2010;9:1956–67.

28. Cha HS, Boyle DL, Inoue T, Schoot R, Tak PP, Pine P, et al. A novel spleentyrosine kinase inhibitor blocks c-Jun N-terminal kinase-mediated geneexpression in synoviocytes. J Pharmacol Exp Ther 2006;317:571–8.

29. Knutson SK, Warholic NM, Wigle TJ, Klaus CR, Allain CJ, Raimondi A,et al. Durable tumor regression in genetically altered malignantrhabdoid tumors by inhibition of methyltransferase EZH2. Proc NatlAcad Sci U S A 2013;110:7922–7.

30. Knutson SK, Kawano S, Minoshima Y, Warholic NM, Huang KC, Xiao Y,et al. Selective inhibition of EZH2 by EPZ-6438 leads to potent antitumoractivity in EZH2-mutant non-Hodgkin lymphoma. Mol Cancer Ther2014;13:842–54.

31. Kaji T, Ishige A, HikidaM, Taka J, Hijikata A, KuboM, et al. Distinct cellularpathways select germline-encoded and somatically mutated antibodiesinto immunological memory. J Exp Med 2012;209:2079–97.

32. Daigle SR, Olhava EJ, Therkelsen CA, Majer CR, Sneeringer CJ, Song J, et al.Selective killing ofmixed lineage leukemia cells by apotent small-moleculeDOT1L inhibitor. Cancer Cell 2011;20:53–65.

33. Chou TC. Theoretical basis, experimental design, and computerized sim-ulation of synergism and antagonism in drug combination studies. Phar-macol Rev 2006;58:621–81.

34. Lehar J, Zimmermann GR, Krueger AS, Molnar RA, Ledell JT, Heilbut AM,et al. Chemical combination effects predict connectivity in biologicalsystems. Mol Syst Biol 2007;3:80.

35. Laursen MB, Falgreen S, Bodker JS, Schmitz A, Kjeldsen MK, Sorensen S,et al. Human B-cell cancer cell lines as a preclinical model for studies ofdrug effect in diffuse large B-cell lymphoma and multiple myeloma. ExpHematol 2014;42:927–38.

36. Auphan N, DiDonato JA, Rosette C, Helmberg A, Karin M. Immunosup-pression by glucocorticoids: inhibition of NF-kappa B activity throughinduction of I kappa B synthesis. Science 1995;270:286–90.

37. Arenzana TL, Smith-Raska MR, Reizis B. Transcription factor Zfx controlsBCR-induced proliferation and survival of B lymphocytes. Blood2009;113:5857–67.

38. Wright G, Tan B, Rosenwald A, Hurt EH, Wiestner A, Staudt LM. A geneexpression-based method to diagnose clinically distinct subgroups ofdiffuse large B cell lymphoma. Proc Natl Acad Sci U S A 2003;100:9991–6.

39. Ding BB, Bi E, Chen H, Yu JJ, Ye BH. IL-21 and CD40L synergisticallypromote plasma cell differentiation through upregulation of Blimp-1 inhuman B cells. J Immunol 2013;190:1827–36.

www.aacrjournals.org Mol Cancer Ther; 16(11) November 2017 2597

Tazemetostat Creates Dependency on B-cell Activation

on September 20, 2020. © 2017 American Association for Cancer Research. mct.aacrjournals.org Downloaded from

Published OnlineFirst August 23, 2017; DOI: 10.1158/1535-7163.MCT-16-0840

2017;16:2586-2597. Published OnlineFirst August 23, 2017.Mol Cancer Ther   Dorothy Brach, Danielle Johnston-Blackwell, Allison Drew, et al.   B-cell Activation Signaling in DLBCLEZH2 Inhibition by Tazemetostat Results in Altered Dependency on

  Updated version

  10.1158/1535-7163.MCT-16-0840doi:

Access the most recent version of this article at:

  Material

Supplementary

  http://mct.aacrjournals.org/content/suppl/2017/08/23/1535-7163.MCT-16-0840.DC1

Access the most recent supplemental material at:

   

   

  Cited articles

  http://mct.aacrjournals.org/content/16/11/2586.full#ref-list-1

This article cites 39 articles, 14 of which you can access for free at:

  Citing articles

  http://mct.aacrjournals.org/content/16/11/2586.full#related-urls

This article has been cited by 2 HighWire-hosted articles. Access the articles at:

   

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

  Subscriptions

Reprints and

  .pubs@aacr.org

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://mct.aacrjournals.org/content/16/11/2586To request permission to re-use all or part of this article, use this link

on September 20, 2020. © 2017 American Association for Cancer Research. mct.aacrjournals.org Downloaded from

Published OnlineFirst August 23, 2017; DOI: 10.1158/1535-7163.MCT-16-0840