Synergistic Antileukemic Therapies in NOTCH1-Induced T-ALL

Synergistic antileukemic therapies in NOTCH1-induced T-ALL

Marta Sanchez-Martina, Alberto Ambesi-Impiombatoa, Yue Qina, Daniel Herranza, Mukesh Bansalb, Tiziana Girardic,d, Elisabeth Paiettae, Martin S. Tallmanf, Jacob M. Roweg, Kim De Keersmaeckerc,d, Andrea Califanob, and Adolfo A. Ferrandoa,h,i,1

aInstitute for Cancer Genetics, Columbia University, New York, NY 10032; bDepartment of Systems Biology, Columbia University, New York, NY 10032; cKU Leuven, University of Leuven, 3000 Leuven, Belgium; dDepartment of Oncology, Leuven Cancer Institute, 3000 Leuven, Belgium; eMontefiore Medical Center, New York, NY 10467; fDepartment of Hematologic Oncology, Memorial Sloan Kettering Cancer Center, New York, NY 10065; gTechnion, Israel Institute of Technology, Haifa 3200003, Israel; hDepartment of Pediatrics, Columbia University, New York, NY 10032; and iDepartment of Pathology, Columbia University, New York, NY 10032

Edited by Bruce A. Chabner, Harvard Medical School, Boston, MA, and accepted by Editorial Board Member Rakesh K. Jain January 9, 2017 (received for review July 18, 2016) The Notch1 gene is a major oncogenic driver and therapeutic tar- Results get in T-cell acute lymphoblastic leukemia (T-ALL). However, inhi- Expression-Based Screen of T-ALL Antileukemic Drugs. Transcriptomic bition of NOTCH signaling with γ-secretase inhibitors (GSIs) has studies have linked inhibition of NOTCH1 signaling with gene shown limited antileukemic activity in clinical trials. Here we per- expression signatures related to down-regulation of anabolic formed an expression-based virtual screening to identify highly pathway genes and up-regulation of genes associated with catabolic active antileukemic drugs that synergize with NOTCH1 inhibition functions (9, 10). Significantly, these metabolic effects are antag- in T-ALL. Among these, withaferin A demonstrated the strongest onized by activation of the PI3K-protein kinase B (AKT)–signaling cytotoxic and GSI-synergistic antileukemic effects in vitro and pathway upon either Pten deletion or via expression of a consti- in vivo. Mechanistically, network perturbation analyses showed tutively active form of AKT (10, 11). Here we hypothesize that eIF2A-phosphorylation–mediated inhibition of protein translation pharmacologic perturbations converging on this core transcriptional as a critical mediator of the antileukemic effects of withaferin A response could yield drugs and drug targets with synergistic anti- and its interaction with NOTCH1 inhibition. Overall, these results MEDICAL SCIENCES leukemic effects in T-ALL when combined with NOTCH1 in- support a role for anti-NOTCH1 therapies and protein translation inhibitor combinations in the treatment of T-ALL. hibition. Toward this goal we searched for positive associations between gene sets generated by drug treatments in the Connectivity leukemia | T-ALL | NOTCH1 | protein translation | synergy Map (cMAP) (12) and the gene expression signatures induced by NOTCH1 inhibition and reversed byPI3K-AKTactivationinT-ALL (SI Appendix,Fig.S1). To generate a NOTCH1 inhibition signature, -cell acute lymphoblastic leukemias (T-ALL) are immature we profiled mouse Notch1-induced T-ALL cells treated with vehicle lymphoid tumors characterized by the diffuse infiltration of T only or a GSI [((S)-2-(2-(3,5-difluorophenyl)acetamido)-N-((S)-5- the bone marrow by malignant lymphoblasts expressing imma- methyl-6-oxo-6,7-dihydro-5H-dibenzo[b,d]azepin-7-yl)propanamide), ture T-cell markers (1). Clinically, T-ALL patients typically present with elevated white cell counts in peripheral blood and frequently show mediastinal thymic masses and meningeal in- Significance filtration of the central nervous system at diagnosis (1). In the early days of combination chemotherapy, T-ALL was recognized The clinical development of targeted therapies has been ham- as a high-risk leukemia group; however, current cure rates with pered by their limited intrinsic antitumor activity and the rapid intensified therapy have improved to about 80% in children (2) emergence of resistance, highlighting the need to identify highly and 60% in adults (3). Despite this progress, the prognosis of active and synergistic drug combinations. However, empirical primary resistant and relapsed T-ALL remains very poor (4). In synergistic drug-screening approaches are challenging, and elu- cidating the mechanisms that underlie such drug interactions this context, the identification of activating mutations in the is typically complex. Here, we performed an expression-based NOTCH1 gene has created major interest in the development of screen and network analyses to identify drugs amplifying the γ-secretase inhibitors (GSI), which block a proteolytic cleavage antitumor effects of NOTCH inhibition in T-cell acute lympho- of NOTCH1 receptor at the membrane required for the activa- blastic leukemia (T-ALL). These studies uncovered a druggable tion of NOTCH1 signaling, as potential targeted therapy in synthetic lethal interaction between suppression of protein T-ALL (5). However, the clinical development of GSIs as anti- translation and NOTCH inhibition in T-ALL. Our results illustrate NOTCH1 therapy has been hampered by a paucity of thera- the power of expression-based analyses toward the identifica- – peutic responses in early trials (6 8). Thus, the identification of tion and functional characterization of antitumor drug combi- highly effective and synergistic GSI drug combinations capable nations for the treatment of human cancer. of eliciting strong and synergistic cytotoxic antileukemic effects has become a major priority toward the development of effective Author contributions: M.S.-M., K.D.K., A.C., and A.A.F. designed research; M.S.-M., D.H., and anti-NOTCH1 therapies in the clinic. T.G. performed research; E.P., M.S.T., and J.M.R. contributed new reagents/analytic tools; Here, we implemented and integrated a systems biology ap- M.S.-M., A.A.-I., Y.Q., D.H., M.B., T.G., K.D.K., and A.A.F. analyzed data; and M.S.-M. and A.A.F. wrote the paper. proach toward the identification of active drugs synergistic with The authors declare no conflict of interest. GSIs for the treatment of NOTCH1-driven T-ALL. These analy- This article is a PNAS Direct Submission. B.A.C. is a Guest Editor invited by the Editorial ses identified eIF2A-mediated translation inhibition as therapeutic Board. target for the development of synergistic drug combinations. Data deposition: Expression data are accessible at Gene Expression Omnibus using acces- Our results uncover highly active drug combinations for the sion codes: GSE71087, GSE71089, GSE78189, and GSE5827. treatment of T-ALL and identify a targetable synthetic lethality 1To whom correspondence should be addressed. Email: [email protected]. interaction between anti-NOTCH1 therapies and eIF2A-mediated This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. translation inhibition. 1073/pnas.1611831114/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1611831114 PNAS Early Edition | 1of6 Downloaded by guest on September 24, 2021 DBZ] in vivo (SI Appendix,Fig.S1). Genes differentially expressed candidate drugs potentially enhancing the effects of NOTCH1 in- in vehicle-only versus GSI-treated Notch1-induced mouse T-ALL hibition in T-ALL (Fig. 1A). These included drugs with known cells identified 16 positive cMAP associations (P < 0.01) indicative of mechanisms of action and redundant activities, such as histone deacetylase inhibitors (vorinostat, valproic acid, and trichostatin A); phenothiazine-derivative antipsychotic compounds (trifluoperazine, thioridazine, and prochlorperazine); and antimalarial drugs (aste- A NOTCH on vs. NOTCH off mizole and mefloquine); together with rapamycin, an mTOR protein complex 1 (mTORC1) inhibitor; geldanamycin, a heat shock Drug name Enrichment P value Functional class score protein 90 (HSP90) inhibitor; resveratrol, an antioxidant sirtuin vorinostat 0.93 0.00E+00 HDAC inhibitor trichostatin A 0.84 0.00E+00 HDAC inhibitor agonist; parthenolide, an nuclear factor KB (NFKB) inhibitor with trifluoperazine 0.75 0.00E+00 antipsychotic leukemia stem cell suppressor activity; withaferin A, a steroidal thioridazine 0.74 0.00E+00 antipsychotic lactone natural compound with antiinflamatory and antiangiogenic rapamycin 0.63 0.00E+00 mTOR inhibitor prochlorperazine 0.62 0.00E+00 anti-malarial activities; phenoxybenzamine, an antiadrenergic alpha receptor astemizole 0.85 1.80E-04 antipsychotic antagonist; pyrvinium pamoate, an antihelmintic compound with geldanamycin 0.54 2.00E-04 HSP90 inhibitor resveratrol 0.66 2.60E-04 antioxidant, Sirt1 agonist antitumor activity and preferential cytotoxicity following glucose parthenolide 0.83 1.23E-03 NFΚB inhibitor starvation; and lanatoside C, a cardiac glycoside ion channel in- withaferin A 0.82 1.87E-03 angiogenesis inhibitor hibitor used in the treatment of congestive heart failure and cardiac phenoxybenzamine 0.82 2.19E-03 adrenergic receptor antagonist pyrvinum pamoate 0.67 3.69E-03 antihelmintic arrhythmias. Notably, and most reassuringly of this approach, rapa- valproic acid 0.23 3.81E-03 HDAC inhibitor mycin (13), vorinostat (14), and different antipsychotic phenothia- lanatoside C 0.65 5.28E-03 cardiac glycoside zine drugs (15) have been recently described to have antileukemic mefloquine 0.70 5.95E-03 anti-malarial effects in T-ALL and to increase the activity of GSIs. Loss of Pten Pten on vs. Pten off rescues the metabolic and antileukemic effects of NOTCH inhibition Drug name Enrichment P value Functional class with GSI (10). Thus, we also investigated negative associations score thioridazine -0.58 0.00E+00 antipsychotic between cMAP gene sets and the expression signatures driven by f/f ERT2 rapamycin -0.57 0.00E+00 mTOR inhibitor Pten loss following tamoxifen treatment of Pten Cre Notch1- trichostatin A -0.37 0.00E+00 HDAC inhibitor induced T-ALL cells (SI Appendix,Fig.S1). Here, genes differen- trifluoperazine -0.44 2.51E-03 antipsychotic tially expressed in Pten-positive vs. Pten-knockout Notch1-induced wortmannin -0.41 3.02E-03 PI3K inhibitor T-ALL cells identified five negative cMAP associations (P < 0.01) Pten Pten on, NOTCH off vs. Pten off, NOTCH off indicative of candidate drugs antagonizing the effects of loss. These included two PI3K-mTOR inhibitor drugs (rapamycin and Drug name Enrichment P value Functional class score wortmannin), the trichostatin A histone deacetylase inhibitor, and trifluoperazine -0.64 0.00E+00 antipsychotic two antipsychotic drugs (trifluoperazine and thioridazine) (Fig. 1A). rapamycin -0.36 0.00E+00 mTOR inhibitor trichostatin A -0.34 0.00E+00 HDAC inhibitor Notably, and consistent with the antagonistic effects of NOTCH1 parthenolide -0.84 1.90E-03 NFΚB inhibitor inhibition and Pten inactivation in T-ALL, our cMAP analyses of drugs potentially enhancing the effects of NOTCH1 inhibition and compounds antagonizing the effects of Pten loss identified rapamy- B cin, thioridazine, trifluoperazine, and trichostatin A as redundant hits in both categories (Fig. 1A). Finally, and consistent with these p roch results, cMAP analysis of the gene expression signatures induced by annin resveratrolwithaferinpe A lor Pten rtm razi - NOTCH1 inhibition in WT cells, but no longer present upon wo ne Pten Notch1 SI d astemizolel GSI treatment of -deleted -induced leukemias ( ci an a a Appendix,Fig.S1), identified three of these drugs (rapamycin, ic to o s r i trifluoperazine, and trichostatin A), as well as parthenolide, as lp d geldana- a e v mycin candidate agents to abrogate the prosurvival effects of Pten loss in

A the context of NOTCH1 inhibition (Fig. 1 ).

Following on these results we next explored the functional e n mefloquine i relationships between these cMAP hits in CUTLL1, a NOTCH1- z

a r vorinostat dependent T-cell lymphoblastic cell line (16), analyzing their e

p pyrvinium

o pamoate gene expression profiles and measuring the pairwise relationship

i between their gene expression signatures using gene set enrich- r

t ment analysis (GSEA) (17). Circos plot (Fig. 1B) and MANTRA p

SI Appendix

plot (Mode of Action by NeTwoRk Analysis) ( , Fig.

n S1) representations of the similarities between the transcrip-

i n

l tional signatures induced by each of these compounds identified i

d s

e o

h drugs with highly interrelated transcriptional programs (wort-

c i

t mannin, valproic acid, trifluoperazine, thrichostatin A, thioridazine,

e rapamycin, phenoxybenzamine, and parthenolide), suggestive of a

n e

o n

y common downstream effector mechanism. In contrast, pyrvinium a

i e

i z

a h

m pamoate, resveratrol, withaferin A, astemizole, mefloquine, and

e n i

c r

a y p m a vorinostat showed the least number of associations with other drugs, suggesting more distinct mechanisms of action. Fig. 1. Gene expression-based identification and characterization of candidate drugs antagonizing NOTCH and PI3K-AKT oncogenic programs in Analysis of Antileukemic Effects and Interaction with NOTCH1 T-ALL. (A) cMAP top-scoring drugs positively associated with NOTCH in- Inhibition. activation signatures and negatively associated with Pten deletion signa- To evaluate the antileukemic effects of these drugs tures in Notch1-induced leukemias and with Pten deletion signatures in and their potential interaction with NOTCH1 inhibition, we first Notch1-induced tumors treated with the DBZ GSI. (B) Circos plot represen- analyzed the response of CUTLL1 cells to each of our cMAP hits tation of pairwise relationships between the gene expression signatures alone and in combination with the DBZ GSI. These analyses induced by drug treatments in CUTLL1 cells. identified 11 compounds with strong synergism with GSI

2of6 | www.pnas.org/cgi/doi/10.1073/pnas.1611831114 Sanchez-Martin et al. Downloaded by guest on September 24, 2021 treatment (Combination Index < 0.4), 8 of which showed high Figs. S6–S8). These analyses demonstrated a major proapototic (IC50 < 0.5 μM) (withaferin A, rapamycin, vorinostat, parthe- activity for withaferin A and parthenolide as single agents and nolide, and wortmannin) or moderate (IC50 0.5–5 μM) (aste- markedly increased apoptotic responses for withaferin A, rapamycin, mizole, trifluoperazine, and trichostatin A) intrinsic antileukemic and astemizole when used in combination with DBZ (Fig. 3 and SI Appendix activities based on reduced cell viability (Fig. 2 and SI Appendix, ,Figs.S6andS7). In contrast, vorinostat, wortmannin, Figs. S2–S4). These results were verified in a broader panel of trichostatin A, and trifluoperazine showed a strong and synergistic cytostatic activity in combination with GSI treatment, but with lim- T-ALL lines, including PTEN WT (DND41 and KOPTK1) and ited apoptosis (SI Appendix,Figs.S6andS7). Analyses of primary PTEN mutant (RPMI8402 and CCRF-CEM) leukemias (SI Appendix SI Appendix – T-ALL cells from three independent NOTCH1-mutant leukemia , Table S1and , Figs. S2 S5). samples demonstrated strong antileukemic effects for the combina- To elucidate the antileukemic effects of these drugs, we next tion of withaferin A and DBZ (Fig. 4A). Consistently, and despite its analyzed their effects on cell cycle progression and apoptosis short in vivo half-life (18), withaferin A induced a marked en- alone and in combination with DBZ (Fig. 3 and SI Appendix, hancement of the antitumor effects of Notch inhibition with DBZ in vivo in a mouse model of Notch1-induced T-ALL (10) (Fig. 4 B and C). In this experiment, we allografted mice with NOTCH1 HD- ΔPEST mouse-induced ALL cells and upon full leukemia devel- ABC DWithaferin A 1.5 1.5 DBZ+ opment treated them with vehicle only, withaferin A, DBZ, or the 0.20 CI=0.18 Withaferin A 1.0 0.15 1.0 combination of withaferin A plus DBZ, observing a marked and 0.5 0.10 0.5 viability viability 0.05 significant reduction in tumor burden by in vivo bioimaging after 6 d 0.0

0.0 [DBZ] M 0.0 0 0.11 10 0 0.1 110 20 0 0.10 0.20 0.30 0.01 of treatment in animals treated with withaferin A plus DBZ (Fig. Withaferin A 0.01 0.001 [Withaferin A], μM [Withaferin A], μM [Withaferin A], μM 4B), which translated in significant extension in survival from 20 d in Rapamycin 1.5 0.20 CI=0.17 1.5 control mice to 60 d in the combination treatment group (Fig. 4C). 1.0 0.15 DBZ+ 0.10 1.0 Rapamycin Finally, we evaluated the efficacy of this treatment against two in- 0.5 0.05 0.5 viability

viability Notch1 SI

[DBZ] M 0.0 dependent -mutant human primary T-ALL xenografts ( 0.0 0.0 0 1 10 20 0 0.01 0.11 10 20 0.01 0.1 0 0.1 0.2 0.3 0.4 Appendix,TableS2) in vivo. These analyses demonstrated variable [Rapamycin], μM [Rapamycin], μM [Rapamycin], μM Rapamycin Parthenolide response to withaferin A alone, but robust responses to withaferin A 1.5 1.2 CI=0.001 1.5 D–G SI Appendix 1.0 0.9 1.0 DBZ+ plus DBZ in combination (Fig. 4 , ,Fig.S9). 0.6 Parthenolide 0.5 0.3 0.5 Of note, therapy with withaferin A plus DBZ in combination viability viability MEDICAL SCIENCES 0.0 [DBZ] M 0.0 0.0 1 0 1 0.1 was well tolerated clinically. Analysis of toxicity showed increased 0 0 e-4 00 01 1 2e-4 3e-44e-4 0005 0. Parthenolide 1e-41e-31e-21e- 0. 0. [Parthenolide], μM [Parthenolide], μM [Parthenolide], μM goblet cell numbers in the intestine of mice treated with DBZ, a Wortmannin phenotype linked with systemic inhibition of NOTCH signaling, 1.5 0.20 CI=0.03 1.0 1.0 0.15 DBZ+ 0.10 0.5 Wortmannin and similar changes with no signs of increased toxicity were noted 0.5 0.05 in animals treated with withaferin A plus DBZ (SI Appendix, Fig. viability viability 0.0 [DBZ] M 0.0 0.0 0 0.30.1 1310 4 -4 010.3 S10). We observed no changes in the weight of C57BL/6 mice 0 - e-4 0.1 1e-4 2e 3 4e Wortmannin [Wortmannin], μM [Wortmannin], μM [Wortmannin], μM treated for 6 d with vehicle, DBZ, withaferin A, or both drugs in Vorinostat SI Appendix 1.5 0.20 CI=0.27 1.5 DBZ+ combination ( ,Fig.S11). Moreover, hematologic 1.0 0.15 1.0 Vorinostat 0.10 analyses revealed only a mild decrease of white blood cells at the 0.5 0.5

viability 0.05 viability expense of lymphocytes and monocytes in animals treated with the 0.0 [DBZ] M 0 1 10 20 0.0 0 0.2 0.4 0.6 0.8 0.00 0.1 1 10 20 0.01 0.1 combination of DBZ and withaferin A (SI Appendix,Fig.S12). Vorinostat [Vorinostat], μM [Vorinostat], μM [Vorinostat], μM 1.5 0.25 1.5 Astemizole 0.20 CI=0.13 DBZ+ eIF2A Translation Inhibition Mediates the Antileukemic Effects of 1.0 0.15 1.0 Astemizole 0.5 0.10 0.5 Withaferin A. Withaferin A, a bioactive steroidal lactone origi- viability viability 0.05 1 0.0 [DBZ] M 0.0 0.0 Withania Somnifera 0 0.31 5 10 051020 05100.3 1 nally isolated from , has shown antitumor [Astemizole], μM [Astemizole], μM [Astemizole], μM effects against colorectal and breast cancer cell lines (19, 20). Astemizole Trifluoperazine However, the mechanisms of action of this natural compound 1.5 1.5 0.20 CI=0.08 DBZ+ 1.0 0.15 1.0 Trifluoperazine remain incompletely understood. To explore the potential ef- 0.5 0.10 0.5 fector mechanisms mediating the antileukemic activities of viability viability 0.05

0.0 [DBZ] M 0.0 0 0.3 1 5 10 0.0 02.04.06.0 05100.3 1 withaferin A in T-ALL and its interaction with NOTCH1 in- [Trifluoperazine], μM [Trifluoperazine], μM [Trifluoperazine], μM hibition we further analyzed the transcriptional signatures in- Trifluoperazine Trichostatin A duced by this drug. In CUTLL1 cells, withaferin A treatment 1.5 0.20 CI=0.19 1.5 DBZ+ induced broad changes in the gene expression profile with 433 1.0 0.15 1.0 Trichostatin A 0.10 0.5 0.5 up-regulated and 424 down-regulated genes (fold change 1.3, viability

0.05 viability

0.0 [DBZ] P < A 10 3.5 2010 0.0 0.0 05100.3 1 0.001) (Fig. 5 ). Notably, this signature was markedly en- 0 01 03 0. 0.020. 0.040.05 [Trichostatin A], M μ [Trichostatin A], μM riched in genes and pathways implicated in protein translation, Trichostatin A [Trichostatin A], M CUTLL1 CEM including translation (GO: 0006412), translational elongation DND41 RPMI (GO: 0006414), ribosome (GO: 0005840), and protein bio- KOPTK1 synthesis (SP_PIR_KEYWORD: protein biosynthesis). More- Fig. 2. Antileukemic effects of cMAP drugs that antagonize NOTCH1 and over, gene set enrichment analysis (GSEA) revealed significant PI3K-AKT expression programs in T-ALL. (A) Withaferin A, rapamycin, par- down-regulation of translation-related pathways upon withaferin thenolide, wortmannin, vorinostat, astemizole, trifluoperazine, and tri- A treatment, including aminoacyl transfer RNA (t-RNA) bio- chostatin A structures. (B) Dose–response cell-viability curves relative to synthesis, 3′UTR translation regulation, peptide elongation, and vehicle-treated controls in T-ALL (72 h treatment). (C) Isobologram analysis ribosome (Fig. 5B). Consistently with our cMAP analysis results, of the effects of fixed molar ratio combinations of each cMAP drug with the genes down-regulated by GSI in T-ALL, including NOTCH1 DBZ GSI in CUTLL1 cells during 6 d of treatment. (D) Cell-viability curves DTX1 MYC relative to vehicle-treated controls of cMAP drugs alone and in combination direct targets (e.g., and ), are also significantly down- regulated by GSEA in CUTLL1 cells treated with withaferin A with a fully inhibitory concentration (250 nM) of the DBZ GSI (72 h treat- SI Appendix ment). All treatments were performed in triplicate and were repeated at ( , Fig. S13). However, the transcriptional effects of least twice. CUTLL1 cells were used in C and D. Data in B and D represent DBZ and withaferin A are not completely overlapping, as evi- mean ± SD. denced by the lack of HES1 down-regulation upon withaferin A

Sanchez-Martin et al. PNAS Early Edition | 3of6 Downloaded by guest on September 24, 2021 P P<0.001 <0.001 as a critical mediator in the antileukemic effects of withaferin ABDMSO withaferin A DMSO withaferin A P P =0.06 =0.004 A in T-ALL and support the role of therapies inhibiting pro- 100 P P<0.001 60 <0.001 80 50 tein translation in combination with NOTCH inhibition for 60 40 the treatment of T-ALL. Consistent with this model, inhibi- DBZ + 40 DBZ + 30 20 DBZ withaferin A 20 DBZ withaferin A tion of protein translation with silvestrol, an inhibitor of eIF4A-

Apoptosis (%) 10 % cells in G1 0 cap–mediated translation, induced synergistic antileukemic 0 SI Appendix DBZ effects with NOTCH inhibition by GSI in T-ALL ( , DMSO DBZ DMSO Counts 7-AAD Fig. S16). PI withaferin A Annexin V withaferin A DBZ+withaferin A P CDDMSO withaferin A P=0.026 DMSO withaferin A DBZ+withaferin <0.001 A Discussion P<0.001 100 P =0.05 60 P<0.001 There is an urgent need to identify drugs that synergistically 80 P=0.004 50 60 40 enhance the antileukemic effects of anti-NOTCH1 therapies in 30 DBZ + 40 DBZ + T-ALL. However, empirical screening approaches to the iden- withaferin A 20 DBZ withaferin A 20 DBZ

Apoptosis (%) tification of synergistic drug combinations are cumbersome and % cells in G1 10 0 0 often do not directly inform on the mechanisms mediating drug DBZ DMSO DBZ interactions. To overcome these obstacles we implemented an 7-AAD DMSO

Counts AnnexinA V PI withaferin A withaferin A expression-based discovery strategy that capitalizes on accurately

DBZ+withaferin A DBZ+withaferin A

Fig. 3. Antitumor effects of withaferin A in combination with the DBZ GSI in P <0.001 P <0.001 P <0.001 T-ALL cells in vitro. (A)Cellcycleand(B) apoptosis analysis of CUTLL1 cells P A P <0.001 P=0.001 =0.005 treated with DMSO, withaferin A (200 nM), the DBZ GSI (250 nM), and 120 P 120 P <0.001 P <0.001 120 =0.001 withaferin A (200 nM) plus DBZ (250 nM) in combination. (C)Cellcycleand(D) 80 80 80 apoptosis analysis of CCRF-CEM cells treated with DMSO, withaferin A (200 40 40 T-ALL #80 T-ALL T-ALL #38 T-ALL T-ALL #22 T-ALL 40 cell viability (%) cell viability (%) nM), the DBZ GSI (250 nM), and withaferin A (200 nM) plus DBZ (250 nM) in 0 0 cell viability (%) 0 combination. Cells were treated with drug alone or in combination per trip- Vehicle Withaferin A DBZ DBZ+ Withaferin A licate during 72 h. Bar graphs in A, B, C, and D indicate mean ± SD. BCP <0.001 NOTCH1 L1601P ΔPEST P <0.001 VehicleWithaferinDBZ ADBZ+ Withaferin A P =0.003 100 P =0.001 SI Appendix 270 90 22 1.8 mean treatment ( ,Fig.S13). Further analysis of withaferin Day 0 A-induced gene expression programs using DeMAND, a regulatory 1000 50 100 Day 3 network algorithm for the identification of potential drug effector 10 Survival (%) 0 1 0204060 Fold change

mechanisms as deregulated nodes induced by a drug treatment, tumor burden Time (days) Day 6 identified 21 components of the translation machinery among the 402010 60 (luciferase counts) Vehicle DBZ vehicle DBZ top 50 nodes perturbed by withaferin A (SI Appendix,TableS3). Luciferase counts (x1000) Withaferin A DBZ+ Withaferin A Withaferin A DBZ+ Withaferin A vehicle withaferin A P =0.005 Notably, these included two subunits of eIF2A (eIF2S1 and P D Day 0 E <0.001 eIF2S2), a protein complex that mediates the recruitment of the P =0.01 first Met-coupled t-RNA to the 40S ribosome subunit (21) and the Day 6 100 18 14 4 1.8 mean

inhibition of protein synthesis under conditions of cellular stress DBZ DBZ+withaferin A (22) (SI Appendix,Fig.S14). In all, these results suggest translation 10

PD-TALL 10 PD-TALL Day 0

inhibition and the eIF2A complex as effectors of the antileukemic 10 PD-TALL 1

Day 6 (luciferase counts) activities of withaferin A. Vehicle DBZ

To test this hypothesis, we analyzed the effects of withaferin A Fold change tumor burden Withaferin A DBZ+ Withaferin A vehicle withaferin A in the control of protein synthesis in polysome profiling and P <0.01 F Day 0 G nascent protein synthesis assays. These studies revealed re- P <0.01 P <0.01 duction of polysome numbers and a significant abrogation of Day 6 100 17 5 11 1 mean nascent protein synthesis in withaferin A-treated T-ALL cells DBZ DBZ+withaferin A compared with vehicle-only–treated controls (Fig. 5 C and D). PD-TALL 19 PD-TALL Day 0 10 Moreover, and consistently with our network analyses prediction,

withaferin A treatment of T-ALL cells induced dose-dependent Day 6 19 PD-TALL 1 A SI Ap- (luciferase counts) phosphorylation of eIF2S1 at residue S51 (Fig. 6 and 2010 40 Vehicle DBZ pendix Luminescence Fold change tumor burden Withaferin A DBZ+ Withaferin A , Fig. S15), a posttranslational modification responsible for luciferase counts (x1000) blocking the formation of eIF2A Met–t-RNA complexes in conditions of amino acid starvation and in response to oxidative Fig. 4. Characterization of the antileukemic effects of withaferin A in and endoplasmic reticulum stress (21–23). Moreover, and con- combination with the DBZ GSI in vivo. (A) In vitro analyses of the antileu- comitant with eIF2S1 S51 phosphorylation, we observed in- kemic responses of three independent primary T-ALL samples treated with DMSO control, DBZ (250 nM), withaferin A (200 nM), or with the same doses creased expression of ATF4, a transcription factor specifically of DBZ plus withaferin A in combination during 72 h. (B) Representative activated by alternative translation in the context of eIF2A-me- images and quantitative analyses of changes in tumor burden in isogenic diated translation inhibition (Fig. 6B). Notably, expression of a Notch1-positive, Pten-positive murine leukemia, upon treatment with vehi- phosphomimic mutant form of eIF2S1 (eIF2S1-S51D) in cle only, the DBZ GSI, withaferin A ,and withaferin A plus DBZ in combina- JURKAT and CUTLL1 cells impaired leukemia cell viability tion (n = 5 C57BL6 mice per group). (C) Survival analysis of mice harboring C D SI Appendix NOTCH1 L1601P Δ-PEST induced T-ALL upon treatment with vehicle only, and proliferation (Fig. 6 and and , Fig. S15), − − DBZ (10 mg kg 1), withaferin A (10 mg kg 1), or withaferin A + DBZ (with- whereas expression of a nonphosphorylatable form of eIF2S1 + ≤ – – (eIF2S1-S51A) abrogated the antileukemic effects of withaferin aferin A DBZ vs. all other groups P 0.001 Gehan Breslow Wilconson test; n = 10 per group). (D, E, F, and G) Luciferase images and quantitative A (Fig. 6 E and F and SI Appendix, Fig. S15) and of DBZ plus G H SI Appendix analyses of changes in tumor burden in two independent human primary withaferin A in combination (Fig. 6 and and , T-ALL xenografts upon treatment with vehicle only, the DBZ GSI, withaferin Fig. S15). These results demonstrate a direct role for eIF2S1 A, and withaferin A plus DBZ in combination (n = 5 per group). Horizontal phosphorylation and inhibition of eIF2A-dependent translation bars in B, E, and G indicate mean luciferase levels.

4of6 | www.pnas.org/cgi/doi/10.1073/pnas.1611831114 Sanchez-Martin et al. Downloaded by guest on September 24, 2021 A role of withaferin A in eIF2A translation inhibition. Consistently, withaferin A treatment induced decreased protein synthesis in T-ALL cells, which was mediated by eIF2S1 phosphorylation. AKR1C3 AKR1C4 ATP5F1 ABHD3 ARPC1B ALDH3B2 OSGIN1 HLA−E KLF6 TALDO1 TCEAL8 MED27 SMPD3 TGS1 LAS1L OGFOD1 GCSAM FBLN2 PPP1R3E PFKFB3 NAT10 PFKFB4 MTA2 GPR56 FTH1 SRGN H2AFY PSMC4 FOS TCF7 RFFL CD82 CD69 CACYBP SMN1 PLCG2 NMT1 EPHX2 URGCP UBA52 PHF19 IP6K1 FANCD2 COX15 GBE1 RPS23 RPS29 SRP54 CLEC2D IDS Key oncogenes and signaling pathways involved in the patho- genesisofT-ALL,includingNOTCH1,MYC,andthePI3K-AKT-

withaferin A mTOR pathway, participate in the regulation of ribosome biogenesis and translation (9, 28, 29). Consistently, gene expression signatures DMSO -5 0 5 B Aminoacyl 3’ UTR mediated C P =0.005 t-RNA biosynthesis translational regulation 120 P P 0.1 <0.0001 0.1 <0.0001 A B 0 0 Polysomes 80 withaferin A (μM) vehiclewithaferin A ES

ES 40 Relative ATF4 DMSO 0.1 0.25 0.5 1 2 −0.7 −0.7 Sedimentation 0 Absorbance 254 nm p-eIF2S1 p-eIF2S1 5 5 DMSO polysome fraction (%) DMSO withaferin A withaferin A −5 up down −5 up down eIF2S1 eIF2S1 t−score t−score P <0.001 GAPDH GAPDH Peptide elongation Ribosome D P P P<0.0001 120 =0.001 0.1 <0.0001 0.1 P 0 0 CD <0.001 80 8 * P 100 <0.001 ES ES 40 6 80 −0.7 −0.7 60 0 4 * 5 5 40 synthesis rate (%) μM μM (millions) −5 up down −5 up down 1 2 20 t−score t−score DMSO 0.5 Cell number Live cells (%)

Relative nascent protein 0 withaferin A 0 0h 24h 48h 72h Empty Fig. 5. Inhibition of protein translation by withaferin A in T-ALL. (A) Heat map Empty vector eIF2S1 eIF2S1 S51D vectoreIF2S1 representation of top differentially expressed genes in CUTLL1 cells treated with P <0.001 vehicle only and withaferin A (24 h). Up-regulated transcripts are shown in red, E 1.5 F eIF2S1 S51D 100 P <0.001 and down-regulated transcripts are shown in blue. Bar at bottom indicates * * 1.0 * 80 MEDICAL SCIENCES differential expression levels in SD units. (B) Representative examples of gene set 60 enrichment plots corresponding to GSEA analysis of MSigDB C2 data sets 0.5 40 enriched in the expression signature associated with withaferin A treatment Cell viability 20 samples compared with vehicle-treated controls. (C) Polysome profiling analysis 0.0 Live cells (%) (relative to DMSO) 0 0.12 0.25 1 0 - ++- - + of JURKAT cells treated with withaferin A or vehicle only (DMSO). (D)Quanti- [Withaferin A], μM withaferin A tative analysis of nascent protein synthesis measured by ClickITonJURKATcells (500nM) Empty eIF2S1 eIF2S1 Empty vector eIF2S1 eIF2S1 S51A vector S51A treated with DMSO or withaferin A. Bars in C and D represent mean ± SD. G P =0.01 P =0.006 P <0.001 H n.s P <0.001 P <0.001 n.s P P n.s P =0.02 P =0.003 <0.001 <0.001 120 140 engineered mouse models of NOTCH1-induced T-ALL with 100 120 80 100 Pten 80 conditional loss of the tumor suppressor gene. Reassuringly, 60 60 this approach recovered inhibitors of the mTOR/PI3K/AKT 40 40 pathway (rapamycin and wortmannin), histone deacetylase in- Nascent protein synthesis rate 20 20 0 0 DBZ Withaferin A DBZ+ hibitors (vorinostat, trichostatin A, and valproic acid), an NFKB vehicle only controls DBZ Withaferin A DBZ+ Cell viability relative to inhibitor (parthenolide), and several phenothiazines (trifluoperazine, Withaferin A Withaferin A prochlorperazine, and thioridazine). However, the most active cy- empty vector eIF2S1 wt eIF2S1 S51A totoxic and synergistic compound identified in this screen was Fig. 6. eIF2A-dependent inhibition of protein translation mediates the withaferin A, a natural compound with antitumor, antiinflammatory, antitumor effects of withaferin A in T-ALL. (A) Western blot analysis of antibacterial, and immunomodulary properties. Mice treated with eIF2S1-S51 phosphorylation in JURKAT cells treated with vehicle only or DBZ plus withaferin A showed markedly reduced tumor loads, low withaferin A. (B) Western blot analyses of ATF4 expression in JURKAT cells μ levels of minimal residual disease, and increased survival in vivo. treated with vehicle only or withaferin A (0.5 M) (C) Cell growth analysis of JURKAT cells after infection with empty-vector control lentiviruses and len- Moreover, the broad response of multiple T-ALL cell models tiviral constructs expressing eIF2S1 WT or the phosphomimic form of eIF2S1 tested here, including GSI-sensitive (CUTLL1, DND41, KOPT- (eIF2S1-S51D). (D) Cell-viability analysis (annexin-V APC exclusion) of JURKAT K1) and GSI-resistant (CCRF-CEM, RPMI 8402, JURKAT) cells after infection with empty-vector control lentiviruses and lentiviral lines, to withaferin A plus GSI in combination supports that this constructs expressing either WT or the phosphomimic form of eIF2S1 combination can overcome resistance to anti-NOTCH1 therapies. (eIF2S1-S51D) (72 h). (E) Differential cell growth of JURKAT cells treated with It is unclear at this point if relapsed tumors following GSI plus withaferin A relative to vehicle-treated controls after infection with empty- withaferin A combination therapy would respond to retreatment. vector control lentiviruses and lentiviral constructs expressing eIF2S1 WT or However, we have not observed development of resistance in vitro the nonphosphorylatable form of eIF2S1 (eIF2S1-S51A) (72 h). (F) Cell- viability analysis (annexin-V APC exclusion) of JURKAT cells after infection after sustained treatment with DBZ plus withaferin A in any of with empty-vector control lentiviruses and lentiviral constructs expressing our models, suggesting that T-ALL cells need to overcome sig- eIF2S1 WT or the nonphosphorylatable form of eIF2S1 (eIF2S1-S51A) treated nificant barriers to develop resistance to this combination. with vehicle only or withaferin A (72 h). (G) Nascent protein synthesis in Mechanisms proposed for the antitumor effects of withaferin A JURKAT cells infected with empty-vector control lentiviruses and lentiviral include proteasome (24), HSP90 (25), NFKB (26) inhibition, and constructs expressing WT eIF2S1 or the nonphosphorylatable form of eIF2S1 induction of reactive oxygen species (ROS) (27). However, these (eIF2S1-S51A) upon treatment with withaferin A or withaferin A plus DBZ mechanisms seem to be cell-type specific, as functional assays revealed (250 nM) relative to vehicle-treated controls. (H) Cell-viability analysis of JURKAT cells infected with empty-vector control lentiviruses and lentiviral decreased ROS levels and no or minimal impact of withaferin A in SI Appendix constructs expressing WT eIF2S1 or the nonphosphorylatable form of eIF2S1 HSP90, proteasome, or NFKB activity in T-ALL cells ( , (eIF2S1-S51A) upon treatment with withaferin A or withaferin A plus DBZ Fig. S17). Functional annotation of transcriptomic data coupled with (250 nM) relative to vehicle-treated controls (72 h). Graphs in C, D, E, F, G, regulatory network analysis (DeMAND) pointed toward a prominent and H indicate mean ± SD. Asterisks in C and E indicate P < 0.001.

Sanchez-Martin et al. PNAS Early Edition | 5of6 Downloaded by guest on September 24, 2021 induced by NOTCH inhibition in T-ALL show marked down- inhibition with DBZ as before after tamoxifen-induced deletion of Pten. We regulation on translation-related genes (SI Appendix,Fig.S18). applied Drug Mode of Action through Network Dysreguation (DeMAND) to Moreover, forced expression of the eIF4A translation initiation withaferin A signature to investigate potential effectors of its antileukemic factor has been shown to accelerate NOTCH1-induced T-ALL in activity. A detailed description of analytical parameters is given in SI Appendix, Materials and Methods. mice, and inhibition of eIF4A cap-mediated translation induces apoptosis in T-ALL (30). Our results shown here further high- Mice and Animal Procedures. We maintained animals in the animal facility at light a central role of protein translation in T-ALL homeostasis the Irving Cancer Center at Columbia University Medical Campus. All animal and suggest a therapeutic role for targeting eIF2A-mediated procedures were approved by the Columbia University Institute for Animal translation in combination with GSIs. Care and Use Committee (IACUC). A detailed description of experimental therapeutic procedures is provided in SI Appendix, Materials and Methods. Materials and Methods T-ALL Human Cell Lines. CCRF-CEM, RPMI-8402, KOPT-K1, and JURKAT T-ALL Statistical Analyses. We performed statistical analysis by Student’s t test. We cell lines were obtained from ATCC, and HPBALL and DND41 cell lines were considered results with P < 0.05 as statistically significant. We analyzed drug obtained from DSMZ (The Leibniz Institute). The CUTLL1 NOTCH-dependent synergism using the median-effect method developed by Chou and Talalay T-cell lymphoblastic cell line has been previously described (16). (31, 32) and used the CalcuSyn software (Biosoft) to calculate the combination index (CI) and perform isobologram analysis of drug interactions. We Human Primary Leukemia Samples. T-ALL primary samples were obtained used GraphPad Prism 4.0 software (GraphPad Software) for determination from the ECOG-ACRIN Cancer Research Group (ECOG-ACRIN) tumor bank. of drug IC50s using nonlinear regression analysis of dose–response curves. Clinical leukemia samples were obtained with informed consent at local Statistical comparison of survival curves in pairs was performed using institutions and used under the supervision of the Columbia University Gehan–Breslow–Wilconson test (GraphPad Software). Medical Center Institutional Review Board Committee. ACKNOWLEDGMENTS. We thank the ECOG-ACRIN Cancer Research Group for Δ In Vitro Studies. We analyzed cell viability and proliferation with the Cell clinical specimens; J. Aster for the MigR1-NOTCH1 L1601P PEST vector; D. Ron Proliferation Kit I (Roche) and cell cycle analysis by flow cytometry after for the pCDNA3 eIF2S1, pCDNA3 eIF2S1-S51A and pCDNA3 eIF2S1-S51D plas- mids; P. P. Pandolfi for the Ptenfl conditional knockout mouse; T. Ludwig for Propidium Iodide (Sigma Aldrich) DNA staining. We quantified cell viability + the ROSA26Cre-ERT2/ mouse; S. Indraccolo for xenograft T-ALL cells, R. Baer for and apoptosis by flow cytometry after Annexin V-allophycocyanin (APC) and helpful discussions and revision of the manuscript; and L. Xu for technical 7AAD markers (AnnexinV BD Pharmigen). Western blot was performed using assistance in mouse experiments. This work was supported by the National standard methods. Antibodies against NOTCH1 (Val1744) (#4147), GAPDH Institute of Health (Grants R01CA129382 and CA120196 to A.A.F.; Grants (#5174S), eIF2S1 (#9722), phospho-eIF2S1 S51 (#9721S), and ATF4 (#11815) CA180827 and CA196172 to E.P.; and CA180820, CA189859, CA14958, were purchased from Cell Signaling Technologies; CDK4 (sc-260) and FYN CA180791, and CA17145 to ECOG-ACRIN), the Stand Up to Cancer Innovative (sc-271294) antibodies were obtained from Santa Cruz Biotechnology. Research Award (to A.A.F.), the Swim Across America Foundation (to A.A.F.), the William Lawrence & Blanche Hughes Foundation (to A.A.F.), Fonds Weten- schappelijk Onderzoek Vlandereen (G084013) and European Research Council Microarray Gene Expression Profile Analysis. We performed cMAP analyses on Starting Grant 334946 (to K.D.K.). D.H. is supported by the US National Insti- gene expression signatures derived from Pten conditional-inducible knock- tutes of Health Grant K99/R00 CA197869 and an Alex Lemonade Stand Foun- out NOTCH1-induced mouse T-ALLs upon treatment with the DBZ GSI; upon dation Young Investigator grant. M.S.-M. is a postdoctoral fellow funded by Pten deletion induced by tamoxifen treatment; and upon NOTCH1 the Rally Foundation.

1. Pui CH, et al. (1990) Heterogeneity of presenting features and their relation to treatment 17. Subramanian A, et al. (2005) Gene set enrichment analysis: A knowledge-based ap- outcome in 120 children with T-cell acute lymphoblastic leukemia. Blood 75(1):174–179. proach for interpreting genome-wide expression profiles. Proc Natl Acad Sci USA 2. Möricke A, et al.; German-Austrian-Swiss ALL-BFM Study Group (2008) Risk-adjusted 102(43):15545–15550. therapy of acute lymphoblastic leukemia can decrease treatment burden and improve 18. Patil D, et al. (2013) Determination of withaferin A and withanolide A in mice plasma survival: treatment results of 2169 unselected pediatric and adolescent patients en- using high-performance liquid chromatography-tandem mass spectrometry: Appli- rolled in the trial ALL-BFM 95. Blood 111(9):4477–4489. cation to pharmacokinetics after oral administration of Withania somnifera aqueous 3. Stock W, et al. (2013) Dose intensification of daunorubicin and cytarabine during extract. J Pharm Biomed Anal 80:203–212. treatment of adult acute lymphoblastic leukemia: Results of Cancer and Leukemia 19. Das T, Roy KS, Chakrabarti T, Mukhopadhyay S, Roychoudhury S (2014) Withaferin A Group B Study 19802. Cancer 119(1):90–98. modulates the Spindle assembly checkpoint by degradation of Mad2-Cdc20 complex 4. Sutton R, et al. (2015) Persistent MRD before and after allogeneic BMT predicts re- in colorectal cancer cell lines. Biochem Pharmacol 91(1):31–39. lapse in children with acute lymphoblastic leukaemia. Br J Haematol 168(3):395–404. 20. Nagalingam A, Kuppusamy P, Singh SV, Sharma D, Saxena NK (2014) Mechanistic 5. Weng AP, et al. (2004) Activating mutations of NOTCH1 in human T cell acute lym- elucidation of the antitumor properties of withaferin a in breast cancer. Cancer Res phoblastic leukemia. Science 306(5694):269–271. 74(9):2617–2629. 6. Deangelo D, et al. (2006) A phase I clinical trial of the notch inhibitor MK-0752 in 21. Holcik M, Sonenberg N (2005) Translational control in stress and apoptosis. Nat Rev patients with T-cell acute lymphoblastic leukemia/lymphoma (T-ALL) and other leu- Mol Cell Biol 6(4):318–327. kemias. Journal of Clinical Oncology, 2006 ASCO Annual Meeting Proceedings Part I. 22. Silvera D, Formenti SC, Schneider RJ (2010) Translational control in cancer. Nat Rev 24(18S):6585. Cancer 10(4):254–266. 7. Wei P, et al. (2010) Evaluation of selective gamma-secretase inhibitor PF-03084014 for 23. Donnelly N, Gorman AM, Gupta S, Samali A (2013) The eIF2α kinases: Their structures its antitumor efficacy and gastrointestinal safety to guide optimal clinical trial design. and functions. Cell Mol Life Sci 70(19):3493–3511. Mol Cancer Ther 9(6):1618–1628. 24. Yang H, et al. (2012) Withaferin A inhibits the proteasome activity in mesothelioma 8. Takebe N, Nguyen D, Yang SX (2014) Targeting notch signaling pathway in cancer: in vitro and in vivo. PLoS One 7(8):e41214. Clinical development advances and challenges. Pharmacol Ther 141(2):140–149. 25. Gu M, et al. (2014) Structure-activity relationship (SAR) of withanolides to inhibit 9. Palomero T, et al. (2006) NOTCH1 directly regulates c-MYC and activates a feed-forward- Hsp90 for its activity in pancreatic cancer cells. Invest New Drugs 32(1):68–74. loop transcriptional network promoting leukemic cell growth. Proc Natl Acad Sci USA 26. Heyninck K, Lahtela-Kakkonen M, Van der Veken P, Haegeman G, Vanden Berghe W 103(48):18261–18266. (2014) Withaferin A inhibits NF-kappaB activation by targeting cysteine 179 in IKKβ. 10. Herranz D, et al. (2015) Metabolic reprogramming induces resistance to anti-NOTCH1 Biochem Pharmacol 91(4):501–509. therapies in T cell acute lymphoblastic leukemia. Nat Med 21(10):1182–1189. 27. Hahm ER, et al. (2011) Withaferin A-induced apoptosis in human breast cancer cells is 11. Palomero T, et al. (2007) Mutational loss of PTEN induces resistance to NOTCH1 in- mediated by reactive oxygen species. PLoS One 6(8):e23354. hibition in T-cell leukemia. Nat Med 13(10):1203–1210. 28. Campbell KJ, White RJ (2014) MYC regulation of cell growth through control of 12. Lamb J, et al. (2006) The Connectivity Map: Using gene-expression signatures to transcription by RNA polymerases I and III. Cold Spring Harb Perspect Med 4(5): connect small molecules, genes, and disease. Science 313(5795):1929–1935. a018408. 13. Cullion K, et al. (2009) Targeting the Notch1 and mTOR pathways in a mouse T-ALL 29. Iadevaia V, Liu R, Proud CG (2014) mTORC1 signaling controls multiple steps in ri- model. Blood 113(24):6172–6181. bosome biogenesis. Semin Cell Dev Biol 36:113–120. 14. Sanda T, et al. (2010) Interconnecting molecular pathways in the pathogenesis and 30. Wolfe AL, et al. (2014) RNA G-quadruplexes cause eIF4A-dependent oncogene drug sensitivity of T-cell acute lymphoblastic leukemia. Blood 115(9):1735–1745. translation in cancer. Nature 513(7516):65–70. 15. Gutierrez A, et al. (2014) Phenothiazines induce PP2A-mediated apoptosis in T cell 31. Chou TC (2010) Drug combination studies and their synergy quantification using the acute lymphoblastic leukemia. J Clin Invest 124(2):644–655. Chou-Talalay method. Cancer Res 70(2):440–446. 16. Palomero T, et al. (2006) CUTLL1, a novel human T-cell lymphoma cell line with t(7;9) 32. Chou TC (2006) Theoretical basis, experimental design, and computerized simulation rearrangement, aberrant NOTCH1 activation and high sensitivity to gamma-secretase of synergism and antagonism in drug combination studies. Pharmacol Rev 58(3): inhibitors. Leukemia 20(7):1279–1287. 621–681.

6of6 | www.pnas.org/cgi/doi/10.1073/pnas.1611831114 Sanchez-Martin et al. Downloaded by guest on September 24, 2021