MEF2C Phosphorylation Is Required for Chemotherapy Resistance in Acute Myeloid Leukemia

Published OnlineFirst February 5, 2018; DOI: 10.1158/2159-8290.CD-17-1271

Research Article

Fiona C. Brown1, Eric Still1, Richard P. Koche2, Christina Y. Yim1, Sumiko Takao1, Paolo Cifani1, Casie Reed1, Shehana Gunasekera1, Scott B. Ficarro3, Peter Romanienko4, Willie Mark4, Craig McCarthy1, Elisa de Stanchina1, Mithat Gonen5, Venkatraman Seshan5, Patrick Bhola6, Conor O’Donnell1, Barbara Spitzer7, Crystal Stutzke8, Vincent-Philippe Lavallée9,10, Josée Hébert9,10,11,12, Andrei V. Krivtsov2,13, Ari Melnick14, Elisabeth M. Paietta15, Martin S. Tallman16, Anthony Letai6,17, Guy Sauvageau9,10,11,12, Gayle Pouliot6, Ross Levine2,7,16,18, Jarrod A. Marto3, Scott A. Armstrong2,13, and Alex Kentsis1,7,14

abstract In acute myeloid leukemia (AML), chemotherapy resistance remains prevalent and poorly understood. Using functional proteomics of patient AML specimens, we identified MEF2C S222 phosphorylation as a specific marker of primary chemoresistance. We found that Mef2cS222A/S222A knock-in mutant mice engineered to block MEF2C phosphorylation exhibited normal hematopoiesis, but were resistant to leukemogenesis induced by MLL–AF9. MEF2C phosphorylation was required for leukemia stem cell maintenance and induced by MARK kinases in cells. Treatment with the selective MARK/SIK inhibitor MRT199665 caused apoptosis and conferred chemosensitivity in MEF2C-activated human AML cell lines and primary patient specimens, but not those lacking MEF2C phosphorylation. These findings identify kinase-dependent dysregulation of transcription factor control as a determinant of therapy response in AML, with immediate potential for improved diagnosis and therapy for this disease.

SIGNIFICANCE: Functional proteomics identifies phosphorylation of MEF2C in the majority of primary chemotherapy-resistant AML. Kinase-dependent dysregulation of this transcription factor confers susceptibility to MARK/SIK kinase inhibition in preclinical models, substantiating its clinical investiga- tion for improved diagnosis and therapy of AML. Cancer Discov; 8(4); 1–20. ©2018 AACR.

INTRODUCTION specific mutational classes of cytogenetically normal and chromosomally rearranged leukemias (1–3). Overall, AML is Acute myeloid leukemias (AML) are cancers of the blood characterized by a prevalence of mutations of genes encoding that originate in the hematopoietic progenitor cells as a regulators of gene expression, such as the MLL–AF9 (KMT2A– result of the accumulation of genetic mutations that lead MLLT3) gene fusion that dysregulates the expression of genes to cell transformation. Recent advances in genomic profil- controlling self-renewal, differentiation, and cell survival ing have revealed distinct genetic subsets of AML, including (4). Recent studies have begun to reveal specific molecular dependencies that can be used for improved targeted 1Molecular Pharmacology Program, Sloan Kettering Institute, Memorial therapies in AML (5). In spite of this knowledge, intensive 2 Sloan Kettering Cancer Center, New York, New York. Center for Epige- chemotherapy and stem cell transplantation continue to be netics Research, Sloan Kettering Institute, Memorial Sloan Kettering Can- cer Center, New York, New York. 3Department of Biological Chemistry and essential means to achieve cure in the treatment of AML. Molecular Pharmacology, Harvard Medical School, Dana-Farber Cancer However, current chemotherapy regimens remain inadequate Institute, Boston, Massachusetts. 4Mouse Genetics Core Facility, Sloan Ket- and fail to induce or sustain remissions in more than 50% of tering Institute, Memorial Sloan Kettering Cancer Center, New York, New adults and 30% of children with AML (6–8). Thus, improved York. 5Department of Epidemiology and Biostatistics, Memorial Sloan Ket- tering Cancer Center, New York, New York. 6Department of Medical Oncol- therapeutic strategies to overcome or block chemotherapy ogy, Dana-Farber Cancer Institute, Boston, Massachusetts. 7Department of resistance are needed. Pediatrics, Memorial Sloan Kettering Cancer Center, New York, New York. Patients with distinct molecularly identifiable types of leuke- 8PhosphoSolutions, Aurora, Colorado. 9The Leucegene Project at Institute mia have been found to exhibit varying degrees of response to for Research in Immunology and Cancer, University of Montreal, Montreal, chemotherapy, leading to therapy intensification for patients Quebec, Canada. 10Division of Hematology-Oncology, Maisonneuve-Rose- mont Hospital, Montreal, Quebec, Canada. 11Quebec Leukemia Cell Bank, with high-risk disease (9–12). However, human leukemias also Maisonneuve-Rosemont Hospital, Montreal, Quebec, Canada. 12Department exhibit large heterogeneity of chemotherapy response, and its of Medicine, University of Montreal, Montreal, Quebec, Canada. 13Depart- molecular determinants continue to remain poorly under- ment of Pediatric Oncology, Dana-Farber Cancer Institute, Boston, Mas- stood. For example, mutations of DNMT3A, as well as gene sachusetts. 14Departments of Pediatrics, Pharmacology, and Physiology and Biophysics, Weill Cornell Medical College, Cornell University, New York, New fusions such as MLL–AF9 and NUP98–NSD1, have been found York. 15Montefiore Medical Center-North Division, Albert Einstein College to contribute to chemotherapy resistance (13–15). However, of Medicine, Bronx, New York, New York. 16Department of Medicine, Leuke- patients with leukemias with these and other mutations can mia Service, Memorial Sloan Kettering Cancer Center, New York, New York. nonetheless achieve durable remissions (3). Likewise, leuke- 17 18 Harvard Medical School, Boston, Massachusetts. Human Oncology and mias that exhibit primary chemotherapy resistance that is Pathogenesis Program, Memorial Sloan Kettering Cancer Center and Weill Medical College of Cornell University, New York, New York. refractory to induction chemotherapy have not been found Note: Supplementary data for this article are available at Cancer Discovery to show significant enrichment for these or other single-gene Online (http://cancerdiscovery.aacrjournals.org/). mutations (9, 11, 16). Thus, additional genetic or molecular Current affiliation for F.C. Brown: Australian Centre for Blood Diseases, mechanisms must cause chemotherapy resistance in AML. Monash University and Alfred Health, Melbourne, Australia. In support of this idea, human leukemias and their geneti- Corresponding Author: Alex Kentsis, Memorial Sloan Kettering Cancer Center, cally engineered mouse models are characterized by distinct 1275 York Avenue, Box 223, New York, NY 10065. Phone: 646-888-3557; cell populations, comprising defined functional compart- E-mail: [email protected] ments such as leukemia stem or initiating cells that exhibit doi: 10.1158/2159-8290.CD-17-1271 unique phenotypic properties, including self-renewal and ©2018 American Association for Cancer Research. enhanced cell survival (17, 18). In part, these behaviors are

April 2018 CANCER DISCOVERY | OF2

RESEARCH ARTICLE Brown et al. caused by co-option of developmental programs that regulate Fig. S1A and S1B). We identified phosphorylation of serine 222 normal hematopoiesis, such as for example self-renewal and (pS222) in MEF2C among the top 20 most highly abundant therapy resistance of MLL–AF9 leukemias (4, 19). In particu- phosphoproteins in induction failure specimens as compared lar, MEF2C, a member of the MADS family of transcription with age, therapy, and disease-matched remission specimens factors, normally regulates hematopoietic self-renewal and (P = 5.0 × 10−3, t test, Fig. 1A and B; Supplementary Fig. S1B). differentiation, supports the proliferation of MLL–AF9 leu- MEF2C was the preferred candidate to study given its kemias, and is associated with the increased risk of relapse known function as an oncogene in lymphoid malignancies when highly expressed in multiple subtypes of AML in (35) and cooperation with MLL-rearranged AML (19). The patients (19–25). However, the precise molecular mechanisms observed MEF2C pS222 peptide was distinct from the related by which MEF2C is dysregulated in AML are not known. MEF2A, MEF2B, and MEF2D peptides (Supplementary Fig. Recurrent mutations of MEF2 family members, including S1C). We also observed phosphorylation of S396 MEF2C MEF2B, MEF2C, and MEF2D in refractory lymphoblastic leu- (Supplementary Data S1), but this modification was present kemias and lymphomas (26–28), suggest that MEF2C and in substoichiometric amounts, as measured using quanti- its orthologs may regulate a general mechanism of therapy tative targeted mass spectrometry (36). In addition, this resistance. analysis revealed phosphoproteins previously implicated in Here, using recently developed high-accuracy mass spec- therapy resistance, such as HSBP1 pS15 (37), as well as those trometry techniques, we determined phospho-signaling pro- not previously observed but likely functional, such as HGF files of human AML specimens collected at diagnosis from pT503 (29). patients with primary chemotherapy resistance and failure of To investigate the diagnostic significance of MEF2C induction chemotherapy. Analysis of these profiles revealed pS222 in primary AML chemotherapy resistance, we high levels of phosphorylation of S222 of MEF2C, which was assessed the prevalence of MEF2C pS222 in an independent found to be significantly associated with primary chemo- cohort of 47 pediatric and adult primary AML specimens therapy resistance in an independent cohort of cytogeneti- matched for age, disease biology, and therapy (Supplemen- cally normal and MLL-rearranged leukemias. By integrating tary Table S2; ref. 16). As previously observed, this cohort genome editing and biochemical and cell biological approaches, included specimens with gene mutations associated with we tested the hypothesis that MEF2C phosphorylation pro- poor prognosis such as those with cryptic rearrangements motes chemotherapy resistance and that its blockade can of MLL/KMT2A and combined DNMT3A and NPM1 muta- be leveraged for improved AML therapy. These studies have tions (Supplementary Table S2). We developed an affinity- revealed an unexpected dependence on kinase-dependent dys- purified antibody against MEF2C pS222 and validated its regulation of transcription factor control as a determinant of specificity using synthetic peptide competition and phos- therapy response in AML, with immediate potential for trans- phatase assays (Supplementary Fig. S2A–S2C). We found lation into improved diagnosis and therapy for this disease. that expression of MEF2C pS222 and total MEF2C were significantly associated with induction failure and primary −4 −3 RESULTS chemotherapy resistance (P = 6.5 × 10 and 6.0 × 10 , respectively; Fig. 1C and D; Supplementary Fig. S2D–S2G), Phosphorylation of S222 in MEF2C Is a Specific remaining statistically significant in multivariate analyses Marker of AML Chemotherapy Resistance (Supplementary Table S3). Overall, pS222 and total MEF2C Previously, we assembled a cohort of primary AML speci- expression levels were high in the induction failure as com- mens matched for AML subtypes and therapy and collected pared with complete remission specimens (Supplementary at diagnosis from patients with failure of induction chemo- Fig. S2H and S2I) and significantly associated with induc- therapy and those who achieved remission after two cycles of tion failure as compared with disease relapse (P = 2.7 × cytarabine and daunorubicin-based induction chemotherapy 10−2; Fig. 1E). In addition, presence of MEF2C pS222 above (16). In this analysis, we found that defined gene mutations its median expression level correlated with reduced overall were associated with primary chemotherapy resistance only event-free survival (EFS; P = 3.8 × 10−2; Fig. 1F) and was a in a minority of cases. Thus, we sought to investigate alterna- significant binary predictor of poor outcome as assessed tive molecular mechanisms that may explain primary chemo- by the ROC analysis (P = 3.2 × 10−2; Fig. 1G). In addition, therapy resistance in AML. we analyzed MEF2C expression data from the ECOG E1900 We focused on phospho-signaling because kinase activa- cohort of young adult AML and found that high MEF2C tion is one of the hallmarks of AML pathogenesis (29, 30). expression correlated with reduced EFS (P = 3.8 × 10−2; Sup- Recent advances in quantitative proteomics, particularly in plementary Fig. S2K). In contrast, the highly homologous high-efficiency, multidimensional fractionation platforms MEF2D did not exhibit this association (P = 0.98; Fig. 1D; (31), enable in-depth analysis of signaling molecules from Supplementary Fig. S2J). Post hoc review of the top 10 speci- rare cell populations (32). Leukemia cells purified from a mens with moderate MEF2C pS222 expression from remis- discovery cohort of eight diagnostic adult AML bone marrow sion cases at diagnosis revealed that 6 of these cases had in aspirate specimens with normal karyotypes (Supplementary fact developed early relapse after initial induction therapy Table S1) were analyzed by metal affinity chromatography treatment (Fig. 1C), corroborating the specific association (IMAC; ref. 33) and isobaric tagging (iTRAQ) mass spec- of MEF2C pS222 with AML chemotherapy resistance and trometry (34). This yielded 2,553 unique phosphopeptides, suggesting that MEF2C phosphorylation is necessary but 34 of which were significantly enriched in induction fail- not sufficient to cause chemotherapy resistance. Overall, ure specimens (Supplementary Data S1; Supplementary these findings indicate that MEF2C pS222 is a specific

OF3 | CANCER DISCOVERY April 2018 www.aacrjournals.org

MEF2C Phosphorylation in Chemoresistant AML RESEARCH ARTICLE

A B 3 4 pITGA4 pSRRM2 pVIM 3 pITGB7 pMAST3 pRRAS2

pMEF2C ) 2 2 pTOP2B pMEF2C pHSPB1 ( P

1 10 pAP4B1 0

− Log 1 −1 −2

(moderated t statistic) −3 0 −2 −10 123

Extent of differential expression Extent of differential −4 Induction failure/remission Proteins (log2 ratio iTRAC ion intensity) C Remission Failure OCI-AML2U937 MEF2C pS222 MEF2C ######

−88 Log2 relative abundance (normalized to actin) E D Remission Failure 10 8 6 MW (kDa) OCI-AML2U937 MSK103MSK109MSK110MSK113MSK106MSK159MSK104MSK105MSK107MSK108 * 4 53 MEF2C 2 0 53 pS222 MEF2C −2 normalized to actin 2 −4 MEF2D log 53 pS222 MEF2C expression −6 Actin Failure RelapseComplete 41 remission

FG1.0 100 0.8 Low pS222 MEF2C 75 High pS222 MEF2C 0.6 50 0.4

Sensitivity High pS222 25 0.2 MEF2C

Event-free survival (%) 0 0.0 0246810 0.00.2 0.40.6 0.81.0 Years 1-Specificity

Figure 1. Phosphorylation of MEF2C at serine 222 is associated with primary AML chemoresistance. A, Phosphoproteomic screen for differentially abundant protein phosphorylation sites detected in diagnostic AML specimens in patients with primary chemotherapy resistance and induction failure, as compared with patients who achieved complete induction remission, with pS222 is marked in red (Supplementary Data S1, Supplementary Fig. S1A and S1B). B, Volcano plot of protein phosphorylation sites detected in induction failure versus complete remission specimens, with candidate phosphoproteins marked, including pMEF2C (red). C, Heat map of MEF2C expression and S222 phosphorylation in a matched cohort of 47 specimens, as measured using quantitative fluorescence immunoblotting, and normalized to actin. #, Specimens from patients with high pS222 expression who achieved complete remission but experienced AML relapse. ^, P = 6.0 × 10−3 and ^^, 6.5 × 10−4 for remission versus failure for MEF2C and pS222 MEF2C, respectively (t test). D, Representative western immunoblot analysis for MEF2C, pS222 MEF2C, and MEF2D in a cohort of age, disease, and therapy-matched AML patient specimens with induction failure and complete remission. The human AML cell lines OCI-AML2 and U937 serve as positive and negative controls for

MEF2C expression and S222 phosphorylation, respectively. E, Normalized log2 expression of pS222 MEF2C compared with actin in induction failure, relapse and complete remission AML patient specimens. *, P = 2.7 × 10−2 and ^, 3.5 × 10−3 for induction failure versus relapse and remission, respectively (t test). F, Event-free survival analysis of 47 AML patient specimens assessed in C–E, separated above or below median pS222 MEF2C expression levels. P = 3.8 × 10−2 (log-rank test). G, ROC curve analysis for pS222 MEF2C in this cohort. P = 3.2 × 10−2 (Wilcoxon test).

April 2018 CANCER DISCOVERY | OF4

RESEARCH ARTICLE Brown et al. marker of primary chemoresistance and failure of induc- recipient mice as compared with wild-type controls over 16 tion therapy in AML. weeks after transplantation (Fig. 2E and F; Supplementary Fig. S3I–S3J). Thus, MEF2C S222 phosphorylation is dis- MEF2C pS222 Is Dispensable for pensable for normal steady-state and stress mouse hemat- Normal Hematopoiesis opoiesis. The activity of MEF2C is known to be regulated by posttranslational modifications, including acetylation, sumoyla- MEF2C Phosphorylation Is Required for tion, and phosphorylation, that can affect the recruitment of MLL–AF9 Leukemogenesis transcriptional corepressors and coactivators (38–40). Analy- Overexpression of Mef2c itself is not sufficient to cause sis of existing phospho-signaling data revealed the presence of leukemia, but potently cooperates with other leukemia onco- MEF2C S222 phosphorylation in human K562 leukemia cells genes (45). In particular, Mef2c is required for the main- and cytokine-stimulated hematopoietic progenitor cells (41, tenance of MLL-rearranged mouse leukemias (19, 23) and 42). The specific association of MEF2C pS222 with failure of high MEF2C expression is associated with MLL-rearranged induction chemotherapy raises the possibility that MEF2C primary AML (23), as supported by the finding of three S222 phosphorylation promotes chemotherapy resistance in cryptic MLL rearrangements, including MLL–AF9 fusions, AML. To test this hypothesis, we first sought to investigate in our cohort (Supplementary Table S2). Thus, we reasoned the potential function of MEF2C S222 phosphorylation in that MEF2C phosphorylation may be required for MLL–AF9- normal hematopoiesis. Thus, we engineered knock-in mice induced leukemogenesis. harboring a loss-of-function Mef2c S222A allele that can- To test this hypothesis, we assessed leukemia induction by not be phosphorylated, and a gain-of-function S222D allele retroviral MLL–AF9 in bone marrow GMP cells of Mef2cS222A/S222A, that mimics phospho-serine by using CRISPR/Cas9 genome Mef2cS222D/S222D, or wild-type littermates. Bone marrow GMP editing (Fig. 2A; Supplementary Fig. S3A). Genotyping of cells were transduced with the MSCV-IRES-GFP retrovirus founder animals identifiedMef2c S222A and Mef2cS222D mutant encoding MLL–AF9, and GFP-expressing cells were purified alleles in 10% and 7% of the mice born, respectively. We con- by FACS and plated in methylcellulose for clonogenic assays firmed the absence of apparent off-target mutations of the (Fig. 2G). We observed a significant reduction of clonogenic Mef2c locus by genomic DNA sequencing of each founder efficiency in serial replating of theMef2c S222A/S222A progenitor animal. To control for possible off-target effects, we obtained two cells transformed with MLL–AF9, as compared to wild-type independent founder strains for both Mef2cS222A and Mef2cS222D controls (P = 3.3 × 10−3, t test, Fig. 2G). We observed the same alleles, and back-crossed them to wild-type C57BL/6J mice. phenotype in clonogenic assays of bone marrow GMP cells Subsequently, Mef2cS222A/S222A and Mef2cS222D/S222D mice from two independent Mef2cS222A/S222A founder strains (Sup- obtained from heterozygous intercrosses were detected at plementary Fig. S3K), consistent with the specific effects of expected Mendelian ratios (Supplementary Fig. S3B) and Mef2c S222A. exhibited approximately equal MEF2C protein expression, as To test the effects of MEF2C S222 phosphorylation on leu- measured by western immunoblotting of B220+ and Gr-1+/ kemia initiation in vivo, MLL–AF9 transduced cells were trans- CD11b+ bone marrow cells, which are known to have rela- planted into lethally irradiated C57BL/6J mice by intravenous tively high and low MEF2C expression, respectively (Fig. 2B). tail-vein injection (Supplementary Fig. S3L). Recipient mice Mef2cS222A/S222A and Mef2cS222D/S222D animals exhibited normal transplanted with MLL–AF9-transduced cells derived from growth as compared with wild-type littermates (Supplemen- wild-type littermate control and Mef2cS222D/S222D mice became tary Fig. S3C and S3D) and had no discernible defects on moribund at approximately 80 days after transplant with histologic analysis of tissues where physiologic activities of more than 90% of the bone marrow replaced with GFP+ leu- MEF2C are required for normal development (43, 44), includ- kemia cells (Fig. 2H; Supplementary Fig. S3M). In contrast, ing brain, heart, skeletal muscle, and spleen. mice transplanted with MLL–AF9-transduced Mef2cS222A/S222A Consistent with the dispensability of MEF2C S222 phos- cells survived for more than 170 days after transplant with- phorylation for normal hematopoiesis, Mef2cS222A/S222A and out physical signs of leukemia development (P = 6.8 × 10−9, Mef2cS222D/S222D mice exhibited normal peripheral blood cell log-rank test; Fig. 2H). We observed no GFP-expressing cells counts and morphologies (Supplementary Fig. S3E and in the peripheral blood and bone marrow of these mice (Sup- S3F). Likewise, we observed no significant differences in the plementary Fig. S3M), consistent with the essential function absolute numbers of mature blood cells from myeloid and for MEF2C pS222 in leukemia stem cell homing or engraft- lymphoid lineages in the bone marrow of Mef2cS222A/S222A ment, as supported by prior studies (23). Thus, MEF2C S222 and Mef2cS222D/S222D mice, as compared with wild-type litter- phosphorylation is required for leukemia initiation by the mates (Fig. 2C and D; Supplementary Fig. S3G). Similarly, MLL–AF9 oncogene. Mef2cS222A/S222A and Mef2cS222D/S222D bone marrow granulo- cyte–macrophage progenitor (GMP) cells exhibited preserved MEF2C Phosphorylation Cooperates with Distinct clonogenic capacities as compared with wild-type Mef2c GMP Leukemia Oncogenes and Is Required for Leukemia cells from littermate control mice (Supplementary Fig. S3H). Stem Cell Maintenance Likewise, CD45.2+ Mef2cS222A/S222A and Mef2cS222D/S222D hemat- Protection of Mef2cS222A/S222A hematopoietic progeni- opoietic stem cells displayed preserved repopulation capaci- tor cells from MLL–AF9-induced leukemogenesis suggests ties in competitive bone marrow transplants with wild-type that MEF2C S222 phosphorylation may be necessary for CD45.1+ cells in lethally irradiated mice, as monitored by the survival of MLL–AF9-transformed leukemia cells. To CD45.2+/CD45.1+ chimerism of peripheral blood cells in investigate the function of MEF2C S222 phosphorylation in

OF5 | CANCER DISCOVERY April 2018 www.aacrjournals.org

MEF2C Phosphorylation in Chemoresistant AML RESEARCH ARTICLE

A WT WT B N S P NSP

MW (kDa) WT S222A/S222AS222D/S222DWT S222A/S222AS222D/S222D AACCT AACC AACCT AACC 53 MEF2C S222A/S222A S222D/S222D 53 N A PND P pS222 MEF2C

41 Actin B220+ Gr-1/CD11b+ AACCG GCCA AACG A T CCA

C D ) 7 1.6 WT Myeloid cells B cells T cells 5 5 5 S222A/S222A 10 2.28 55.0 10 10 1.93 0.096 104 4 1.2 S222D/S222D 104 10 3 43.3 10 103

Gr-1 3 CD8 102 B220 10 102 0.8 101 0 101 0 40.5 2.23 103 94.4 3.58 10 − 100 Bone marrow 0.4 100 101 102 103 104 105 0K 50K 100K 150K 200K 250K 100 101 102 103 104 105 CD11b FSC-A CD4

absolute cell counts (1 × 10 0.0 Gr-1/ B220+ CD4+/ CD11b+ CD8+

E F 80 80 4 Weeks 8 Weeks

sm 16 Weeks 60 sm 60

40 40

20 20 CD45.2 chime ri CD45.2 chime ri 0 (% of mononuclear PB cells) 0 WT S222A/ S222D/ (% of mononuclear PB cells) WT S222A/ S222D/ S222A S222D S222A S222D GH 1,400 WT; MLL-AF9 S222A/S222A; MLL-AF9 125 1,200 S222D/S222D; MLL-AF9 S222A/S222A 100

1,000 l

800 75 * S222D/ 600 50 S222D

400 survi va Percent

Colonies/1,000 cells 25 200 WT 0 0 01080 120 160 123456 Time (days) Replating

Figure 2. A therapeutic window for targeting MEF2C phosphorylation in AML. A, Sequencing electropherograms of tail genomic DNA from Mef2cS222A/S222A and Mef2cS222D/S222D mice, demonstrating specific CRISPR/Cas9-induced c.TCA>GCG and c.TCA>GAT mutations, as underlined red and green, respectively. WT, wild-type. B, Western immunoblot of bone marrow B220+ and Gr-1+/CD11b+ cells from Mef2cS222A/S222A and Mef2cS222D/S222D mice. C, Total numbers of myeloid (Gr-1+/CD11b+), B-cell (B220+), and T-cell (CD4+/CD8+) bone marrow cells from Mef2cS222A/S222A and Mef2cS222D/S222D mice as assessed by (D) FACS analysis. Error bars, SD of the mean from 3 mice. E, Peripheral blood chimerism of CD45.2+ cells at 4 weeks following competitive transplantation. Error bars, SD of the mean of 10 animals per group. F, Peripheral blood engraftment of CD45.2+ cells as a function of time posttransplant as in E. Bars, mean. G, Serial replating clonogenic efficiencies of bone marrow GMP cells fromMef2c S222A/S222A, Mef2cS222D/S222D, and wild-type littermates transduced with MLL–AF9. Error bars, SD of the mean of 3 biological replicates (additional data in Supplementary Fig. S6). *, P = 3.3 × 10−3 of WT;MLL–AF9 versus S222A/S222A;MLL–AF9 (t test). H, Kaplan–Meier survival curves of mice transplanted with MLL–AF9-transformed bone marrow GMP cells from Mef2cS222A/S222A, Mef2cS222D/S222D and wild-type littermate controls. P = 6.8 × 10−9 for Mef2cS222A/S222A vs. wild-type littermates, log-rank test for 10 animals per group. Solid and dashed black lines denote wild-type littermates for Mef2cS222A/S222A and Mef2cS222D/S222D, respectively.

April 2018 CANCER DISCOVERY | OF6

RESEARCH ARTICLE Brown et al.

MLL–AF9-induced leukemia maintenance, we transformed that MEF2C phosphorylation is necessary but insufficient to wild-type bone marrow GMP cells by retroviral transduction of cause enhanced leukemia cell survival. Notably, when allowed MSCV-IRES-GFP MLL–AF9 and coexpressed wild-type MEF2C, to develop leukemia (Supplementary Fig. S4E) and secondar- S222A, or S222D mutants using the MSCV-IRES-tdTomato ily transplanted into sublethally irradiated mice to assess (MIT) retrovirus (Supplementary Fig. S4A). This allowed us leukemia maintenance in vivo, MLL–AF9;MEF2C S222A leuke- to purify GFP/tdTomato double-positive MLL–AF9-trans- mias were significantly impaired in leukemia development as formed cells using FACS where MEF2C transgenes can func- compared with MLL–AF9;MEF2C wild-type leukemias (P = 2.2 × tion as dominant mutants because of their dimerization with 10−10, log-rank test, Fig. 3A). Using limiting dilution analysis, endogenous MEF2C (46, 47). We confirmed that all MEF2C we found that the expression of wild-type phosphorylated transgenes were expressed at approximately equal levels by MEF2C increased the leukemia initiating cell frequency as western immunoblotting, with MEF2C S222A–transduced compared with MIT control (1/16 vs. 1/85) and consistent cells exhibiting substantially reduced but not completely with previous reports (48), whereas the expression of MEF2C eliminated levels of pS222 as compared with wild-type con- S222A significantly blocked this effect (1/139,P = 6.7 × 10−8, trols (Supplementary Fig. S4B). We found that MLL–AF9 χ2 test, Fig. 3B; Supplementary Fig. S4F–S4H). These results leukemia cells expressing MEF2C S222A exhibited transient indicate that MEF2C S222 phosphorylation is required for reduction in clonogenic efficiencyin vitro (Supplementary the enhanced maintenance of leukemia stem cells and its Fig. S4C) and displayed increased apoptosis (Supplemen- cooperation with MLL–AF9 leukemogenesis in vivo. tary Fig. S4D) as compared with cells expressing wild-type To investigate MEF2C S222 phosphorylation in non– MEF2C. MEF2C S222D–expressing cells behaved similarly to MLL-rearranged leukemias, we assessed its function in leu- MEF2C wild-type–expressing cells, similar to the phenotype of kemias with inactivating mutations of RUNX1 and internal MLL–AF9-transduced Mef2cS222D/S222D leukemias, indicating tandem duplication (ITD) mutations of FLT3, a leukemia

A B

MLL-AF9 0 MIT = 1/85 100 MEF2C = 1/16 S222A = 1/139 75 MIT − 1 S222A

50 − 2

25 Percent survival Percent MEF2C − 3 0 0104080 120 160 200 Log fraction nonresponding Time (days) 0 200 400 600 800 1,000 Dose (cells transplanted) CD Runx1−/−;Flt3 ITD 600 100 MIT MIT MEF2C S222A 75 400 S222D 50 MEF2C S222A

200 25 Percent survival Percent

Colonies/5,000 cells 0 * 0 0515 20 25 30 123 Time (days) Replating

MIT MEF2C S222A S222D

Figure 3. MEF2C phosphorylation is required for leukemia stem cell survival and maintenance. A, Kaplan–Meier survival curves of secondary recipient mice transplanted with 100 cells of wild-type MLL–AF9;MEF2C, dominant-negative MLL–AF9;MEF2C S222A, or control MLL–AF9;MIT-transformed leukemias. P = 2.2 × 10−10 for MLL-AF9;MEF2C S222A versus MLL-AF9;MEF2C, log-rank test for 20 animals per group. B, Limiting dilution analysis of frequency of leukemia-initiating cells in secondary MLL–AF9 transplants. Solid and dashed lines represent the calculated stem cell frequencies and their 95% confidence intervals, respectively.P = 6.7 × 10−8 for S222A versus MEF2C (χ2 test). C, Colony formation of primary Runx1−/−;Flt3ITD;MEF2C leukemia cells. Below, representative micrographs of colonies at day 7. *, P = 1.9 × 10−5 S222A vs. MEF2C, t test. Error bars, SD of the mean of 3 biological replicates. Scale bar, 100 μm. D, Kaplan–Meier survival curves of tertiary recipient mice transplanted with 50,000 cells of wild-type Runx1−/−;Flt3ITD;MEF2C, dominant-negative Runx1−/−;Flt3ITD;MEF2C S222A and Runx1−/−;Flt3ITD;MEF2C S222D or control Runx1−/−;Flt3ITD;MIT-transduced leukemias. P = 3.1 × 10−3 for MLL–AF9;MEF2C S222A versus MLL–AF9;MEF2C, log-rank test for 5 animals per group. (continued on following page)

OF7 | CANCER DISCOVERY April 2018 www.aacrjournals.org

MEF2C Phosphorylation in Chemoresistant AML RESEARCH ARTICLE subtype associated with high rates of therapy failure that MEF2C Phosphorylation Is Required for Leukemia also exhibits high levels of pS222 MEF2C. First, we obtained Maintenance in MLL-Rearranged Human AML Cells genetically engineered mouse Runx1−/−;Flt3ITD leukemias (49), which we transduced with wild-type MEF2C and transgenes To assess the function of MEF2C S222 phosphoryla- modeling loss and gain of S222 phosphorylation using the tion in human AML, we analyzed its expression in a panel MIT retrovirus (Supplementary Fig. S5A–S5C). We found of human AML cell lines and identified bothMLL -rear- that Runx1−/−;Flt3ITD leukemias expressing MEF2C S222A dis- ranged and nonrearranged leukemias with activated levels of played significantly reduced clonogenic efficiency in serial MEF2C pS222, including OCI-AML2, MOLM-13, K562, and replating as compared with cells transduced with wild-type Kasumi-1 (Supplementary Fig. S6A and S6G). Thus, we used or S222D MEF2C (P = 1.9 × 10−5, t test, Fig. 3C). Similar to a doxycycline-inducible lentivirus vector to express wild-type the MLL–AF9 leukemia model, we observed that mice trans- and dominant-negative MEF2C S222A and S222D mutants planted with secondary Runx1−/−;Flt3ITD;MEF2C wild-type leu- and confirmed near-physiologic transgene expression and kemia cells exhibited accelerated leukemia development as pS222 levels by western immunoblotting (Fig. 3E and F; compared with Runx1−/−;Flt3ITD MIT control leukemias, which Supplementary Fig. S6B). We observed that OCI-AML2 and was blocked in Runx1−/−;Flt3ITD;MEF2C S222A–expressing leu- MOLM-13 AML cell lines with endogenous MEF2C S222 kemias (P = 3.1 × 10−3, log-rank test; Fig. 3D). In contrast, we phosphorylation had significantly reduced viability upon the found that MEF2C S222 phosphorylation was dispensable doxycycline-induced expression of MEF2C S222A as com- for leukemias induced by ectopic retroviral expression of pared with wild-type MEF2C (P = 8.4 × 10−3 and 3.4 × 10−3, t Hoxa9 and Meis1 (Supplementary Fig. S5D–S5H), consistent test, respectively; Fig. 3G) and S222D MEF2C, at least in part with the epistatic function of aberrant Hox gene expression due to the induction of apoptosis (Supplementary Fig. S6C in AML (50, 51). Thus, MEF2C S222 phosphorylation is and S6D) and reduction of cells in S phase (Supplementary required for maintenance of Runx1−/−;Flt3ITD leukemias. Fig. S6E and S6F), in agreement with prior reports of cell

EF 6 MEF2C pS222 MEF2C MW (kDa) * OCI-AML2MEF2C S222A 5

53 β -actin) MEF2C 4

53 pS222 MEF2C 3 2 Actin ** 41 1

(normalized to 0 Relative protein expression Relative OCI- MEF2C S222A AML2

G Wild-type MEF2C S222A S222D H 120 100 S222A 100 75 80 OCI-AML2 50 ** MEF2C 60

Percent survival Percent 25 40 * 0 20 0103040 50 60 Normalized luminescence (% no doxycycline control) (% no doxycycline Time (days) 0 MOLM-13 OCI- HL-60 U937 AML2 MEF2C+ MEF2C−

Figure 3. (Continued) E, Western immunoblot analysis for MEF2C and pS222 MEF2C in OCI-AML2 cells lentivirally transduced with wild-type MEF2C or dominant-negative MEF2C S222A transgenes and treated for 48 hours with 600 ng/mL of doxycycline to induce transgene expression. F, Quantita- tive analysis of MEF2C and pS222 MEF2C, as measured using quantitative fluorescence immunoblotting, and normalized to actin, demonstrating equal expression of MEF2C and MEF2C S222A protein (*, P = 0.96, t test) and significantly reduced abundance of pS222 MEF2C (**, P = 1.1 × 10−3 for MEF2C S222A vs. MEF2C, t test). Error bars, SD of the mean for 3 biological replicates. G, Growth of human AML cell lines lentivirally transduced with wild-type MEF2C or dominant-negative MEF2C S222A and MEF2C S222D transgenes and treated for 72 hours with 600 ng/mL doxycycline to induce transgene expression. Error bars, SD of the mean for 3 biological replicates. *, P = 3.4 × 10−3 and **, 8.4 × 10−3 for MEF2C versus MEF2C S222A, respectively (t test). H, Kaplan–Meier survival curves of NSG mice transplanted with OCI-AML2 cells transduced with wild-type MEF2C and dominant-negative MEF2C S222A transgenes, and treated with doxycycline in chow 3 days following transplantation continuously in vivo. P = 3.5 × 10−6 MEF2C versus MEF2C S222A, log- rank test for 10 animals per group.

April 2018 CANCER DISCOVERY | OF8

RESEARCH ARTICLE Brown et al.

A 25 C

20 2 Downregulated MS4A3 1 IFT57 15 0 BMP8B * MARC1 10 −1 CTSG Upregulated (fold control) (fold −2 5 NOG anscriptional activity SMIM14 Tr 0 BTG2 HEK Vector WT S222A HIST1H2AC 293T control CALCOCO2

B MW (kDa) 53 – MEF2C 53 – pS222 MEF2C

41 – Actin 123 123 123123 MEF2C MEF2C S222A S222A Doxycycline − + − + DE

Hallmark_mTorc1_Signaling ODonnell_DN_Targets_MYC Hallmark_apoptosis Hallmark_E2F_Targets HPC_RUNX1_RUNX1T1_Fusion_UP 0.5 Reactome_S_Phase CD34_NUP98_HOXA9_Fusion_UP 0.4 0.3 Schuhmacher_MYC_Targets_UP mTor_Inhibitor_DN 0.2 Zhao_Cell_Cycle_Genes Bcell_vs_CD8_Tcell_UP 0.1 NK_DN_Unstim_vs_IL2_Stim Mature_Bcell_UP −0.0 chment score (ES) 0.1 Hallmark_MYC_Targets Notch_Targets_UP − Manalo_Hypoxia_DN Hallmark_Apoptosis En ri −32−2 −1 0 10 3 Normalized enrichment score S222A+ S222A−

Figure 4. MEF2C phosphorylation is required for MEF2C-mediated gene expression program. A, Activity of luciferase transcriptional MEF2 reporter in HEK293T cells lentivirally transduced with wild-type MEF2C or mutant MEF2C S222A, as compared with vector control. Error bars, SD of the mean for 3 biological replicates. *, P = 4.0 × 10−2 for MEF2C S222A versus MEF2C, t test. B, Western immunoblot analysis for MEF2C and pS222 MEF2C in transcriptional reporter cells, demonstrating equal protein expression of MEF2C transgenes, and reduced pS222 in MEF2C S222A transduced cells. C, Hierarchical clustering of gene expression of the most differentially expressed genes in OCI-AML2 cells transduced with wild-type MEF2C or dominant- negative MEF2C S222A transgenes (−), and treated with 600 ng/mL doxycycline for 48 hours to induce transgene expression (+). Three biological replicates are shown, as indicated. Blue-to-red color gradient indicates relative gene expression. D, Gene set enrichment analysis (GSEA) of significantly upregulated and downregulated gene sets. E, GSEA illustrates enrichment of apoptotic genes in MEF2C S222A–expressing cells. Normalized enrichment score = 1.76 and false discovery rate q = 5.2 × 10−3. (continued on following page) cycle–dependent regulation of MEF2C (52, 53). In contrast, MEF2C Phosphorylation Is Required for HL-60 and U937 AML cells that lack MEF2C expression MEF2C-Mediated Gene Repression Program exhibited no significant effects upon expression of wild-type or mutant MEF2C S222A (Fig. 3G). Overall, these results MEF2C functions as a transcription factor by sequence- indicate that MEF2C S222 phosphorylation is required spe- specific recognition of MEF2 response elements (54), through cifically for the survival of MEF2C-activated AML cells but regulated recruitment of transcriptional coactivators and core- not those lacking MEF2C expression. pressors via its transactivation (residues 118–473) domain To investigate the function of MEF2C phosphorylation in (38, 55). To test the hypothesis that MEF2C S222 phospho- human leukemias in vivo, we next assessed the effects of inhib- rylation regulates MEF2C transcriptional activity, we assessed iting MEF2C S222 phosphorylation by expression of MEF2C the transactivation of MEF2 response element–driven firefly S222A in OCI-AML2 cells transplanted orthotopically by luciferase as compared with a cytomegalovirus promoter– tail-vein injection in immunodeficient mice (Fig. 3H). Cells driven Renilla luciferase control introduced by electropora- were transplanted into NOD-SCID-IL2Rcγnull (NSG) mice, tion in K562 cells. To model the effects of MEF2C S222 and transgene expression was induced 3 days after trans- phosphorylation, we used lentivirus transduction to generate plant using doxycycline chow. Mice transplanted with OCI- K562 cells expressing wild-type or dominant-negative S222A AML2 cells expressing dominant-negative MEF2C S222A MEF2C upon doxycycline treatment (Supplementary Fig. had significantly prolonged survival and impaired leukemia S6G–S6H). As expected, expression of wild-type MEF2C led development as compared with those transplanted with cells to an 8-fold increase in MEF2C transcriptional activity as expressing wild-type MEF2C or nontransduced control OCI- compared with unmodified control K562 cells, which was AML2 cells (P = 3.5 × 10−6, log-rank test; Fig. 3H). These find- rescued by mutant S222A protein expression (P = 3.1 × 10−3, ings indicate that MEF2C S222 phosphorylation is required t test; Supplementary Fig. S6I). We confirmed these findings for the survival of human AML cells in vitro and in mouse in HEK 293T cells stably expressing wild-type and mutant xenografts in vivo. MEF2C S222A (Fig. 4A). Expression of the wild-type MEF2C

OF9 | CANCER DISCOVERY April 2018 www.aacrjournals.org

MEF2C Phosphorylation in Chemoresistant AML RESEARCH ARTICLE

FH S222A-activated genes 80 *** OCI-AML2 3 NCOA3 CASP8 70 MEF2C BCOR FOXO3 S222A 2 BMF BCL6 60 50 ** * 1 fold change) fold 40 2 30 0 20

Gene expression −1 10 LYL1 MYC Cytochrome c release (%) 0 (RNA-seq, log −2 BIM BID PUMA BAD HRK S222A-repressed genes −202468 Chromatin accessiblity of MEF2 response element-containing loci I (ATAC-seq, log 2 fold change) DSS 0004 mmol/L MW (kDa) 235 – 170 – 130 – 93 – Dimer G 70 – OCI-AML2MEF2CS222A MW (kDa) 53 – MEF2C LYL1 30 – 41 – Actin 41 – Actin Relative LYL1 1 1 0.6 Expression +/– +/– +/– OCI-AML2 MEF2C S222A 0.01 0.1 0.1

Figure 4. (Continued) F, Combined analysis of differentially expressed genes identified in RNA sequencing (RNA-seq) shown inC versus differentially accessible genes identified in the assay for transposase accessible chromatin with sequencing (ATAC-seq) containing canonical MEF2 sequence motifs. S222A-induced and S222A-repressed genes are highlighted in red and blue, respectively. G, Western immunoblot analysis for LYL1 in OCI-AML2 cells lentivirally transduced with wild-type MEF2C or dominant-negative MEF2C S222A transgenes and treated for 48 hours with 600 ng/mL of doxycycline to induce transgene expression. Below, quantitative analysis of LYL1 expression normalized to actin. H, BH3 profiling of OCI-AML2 cells transduced with wild-type MEF2C or dominant-negative MEF2C S222A transgenes, and treated with 600 ng/mL doxycycline for 48 hours to induce transgene expression. Error bars, SD of the mean for 3 biological replicates. *, P = 1.3 × 10−2; **, P = 1.2 × 10−2; ***, P = 2.9 × 10−2 for MEF2C S222A versus MEF2C respectively, t test. I, Western immunoblot analysis for MEF2C in DSS-treated OCI-AML2 cells lentivirally transduced with wild-type MEF2C or dominant-negative MEF2C S222A transgenes and treated for 48 hours with 600 ng/mL of doxycycline to induce transgene expression. in HEK293T cells that lack endogenous MEF2C led to an 276 genes that were significantly altered in expression in 18-fold increase in MEF2C transcriptional activity as com- cells expressing MEF2C S222A as compared with uninduced pared with the vector control (P = 6.3 × 103, t test, Fig. 4A). cells (Fig. 4C; Supplementary Data S2). Using qRT-PCR, we Despite equal levels of wild-type MEF2C and mutant S222A found that none of the previously reported canonical MEF2C protein expression (Fig. 4B), MEF2C transcriptional activity target genes, such as Nr4a1/Nur77, Hdac7, Jun, and Cebpa, was significantly reduced due to the effects of MEF2C S222 were significantly changed in expression upon expression of phosphorylation in the S222A mutant (P = 4.0 × 10−2, t test; MEF2C S222A (Supplementary Fig. S4I–S4L). Instead, gene Fig. 4A and B). This suggests that MEF2C S222 phosphoryla- set enrichment analyses (GSEA) revealed that loss of MEF2C tion is required for maximal activation of MEF2C-dependent S222 phosphorylation led to downregulation of distinct gene gene transcription. expression programs, including E2F and MYC target genes To elucidate the gene expression program controlled by (Fig. 4D; Supplementary Data S2). In addition, we observed MEF2C S222 phosphorylation in AML cell survival, we ana- altered expression of genes regulating apoptosis and cell lyzed gene expression profiles of OCI-AML2 cells express- cycle (Fig. 4D and E; Supplementary Data S2), consistent ing wild-type as compared with dominant-negative MEF2C with the phenotypic induction of apoptosis and cell-cycle S222A using RNA sequencing (RNA-seq; Fig. 4C). Consistent defects upon the blockade of MEF2C S222 phosphorylation with the near-physiologic expression of MEF2C transgenes in mouse and human leukemia cells (Supplementary Fig. (Fig. 3E), we found essentially no significant differences in S4D and S6C–S6F). gene expression profiles of cells expressing wild-type MEF2C To identify genes specifically regulated by S222 phosphoryl- upon doxycycline-induced transgene expression (Fig. 4C). In ated MEF2C, we profiled chromatin accessibility using the contrast, we observed a significant change in gene expres- assay for transposase accessible chromatin with sequencing sion of cells upon doxycycline-induced expression of MEF2C (ATAC-seq), annotated for the presence of canonical MEF2 S222A, as compared with both uninduced cells and cells binding sites, in OCI-AML2 cells expressing wild-type and expressing wild-type MEF2C (Fig. 4C). In particular, we found S222A-mutant MEF2C upon 48 hours of doxycycline-induced

April 2018 CANCER DISCOVERY | OF10

RESEARCH ARTICLE Brown et al. transgene expression. This approach was necessary, because for antileukemic therapy. Given that direct inhibition of mapping MEF2C genome localization using chroma- transcription factors is pharmacologically challenging, we tin immunoprecipitation with sequencing (ChIP-seq) was reasoned that inhibition of the upstream kinases phosphoryl- not possible due to the inadequate chromatin enrichment ating MEF2C at S222 could downregulate MEF2C signaling achieved with currently available MEF2C-specific antibodies. to impair aberrant AML cell survival. To identify candidate Using ATAC-seq analysis annotated for the presence of MEF2 kinases that can phosphorylate S222 of MEF2C, we screened binding sites as a surrogate of MEF2C binding, we identi- a library of 172 recombinant human serine kinases in their fied genes that exhibited significant changes in chromatin ability to phosphorylate a model substrate in the presence of accessibility of MEF2 binding sites and concomitantly had synthetic pS222-containing peptide as a putative competi- significant changes in their expression, as measured using tive product inhibitor (Fig. 5A; Supplementary Data S3). We RNA-seq (Fig. 4F; Supplementary Fig. S6J and Data S2). This found that only five protein kinases scored significantly in analysis identified a total of 262 genes, including 18 genes this screen, of which four belonged to a closely related micro- repressed by the expression of MEF2C S222A, consistent with tubule-associated protein/microtubule affinity regulating their aberrant induction by S222 phosphorylated MEF2C, kinase (MARK) 1, 2, 3, 4 family (Fig. 5A). Analysis of primary such as MYC and LYL1 (marked in blue, Fig. 4F). Observed AML specimens revealed that MARK3 and MARK2 genes are gene expression changes in LYL1 repression were verified by highly expressed in most subtypes of AML (Supplementary protein immunoblotting (Fig. 4G), potentially explaining Fig. S7A–S7B), with high MARK3 expression in particular previous findings of MEF2C-dependentLYL1 expression and significantly associated with inferior survival among adults its aberrant expression in AML (56–58). Similarly, we also with AML (P = 1.7 × 10−3, log-rank test; Supplementary Fig. observed 159 genes that were upregulated by MEF2C S222A S7C). We validated the specific kinase activity of recombinant expression, consistent with their repression by phosphoryl- MARK3 with synthetic MEF2C pS222 peptide, as compared ated MEF2C in AML cells. These genes included BCOR, BCL6, with the closely related CAMK1α kinase as specificity control and NCOA3, as well as apoptotic regulators BMF, CASP8, (Fig. 5B). These results indicate that MARK3 and related and FOXO3 (marked in red, Fig. 4F). In agreement with the MARK kinase family members can specifically phosphorylate MEF2C S222 phosphorylation–induced suppression of these S222 MEF2C in vitro. proapoptotic factors, we found that OCI-AML2 cells express- To evaluate whether MARK3 could phosphorylate MEF2C ing MEF2C S222A exhibited enhanced apoptotic priming S222 and regulate its transcriptional activity in cells, we compared to MEF2C wild-type cells as assessed by BH3 pro- coexpressed MEF2C and MARK3 in HEK293T cells and filing flow cytometry, with particular sensitivity to BIM and assessed MEF2C phosphorylation and transcriptional activ- BID (Fig. 4H), consistent with their reported interactions ity by western immunoblotting and transcriptional reporter with BMF, CASP3, and FOXO3 (59–62). assays, respectively (Fig. 5C). We found that MEF2C tran- Homodimerization or heterodimerization of the MEF2 scriptional activity was significantly increased upon coex- family of transcription factors is a key mechanism of their pression of MEF2C and MARK3 as compared with cells regulation, which can be modulated by posttranslational expressing MEF2C alone (P = 3.4 × 10−5, t test; Fig. 5C). This modifications (46, 63). To investigate whether MEF2C S222 increased transcriptional reporter activity was associated with phosphorylation can regulate its oligomerization, we treated increased levels of pS222, which was blocked by the coexpres- OCI-AML2 cells induced to express MEF2C wild-type or sion of the enzymatically impaired MARK3 T211A activation S222A with disuccinimidyl suberate (DSS), a cell-permeable loop mutant (refs. 64, 65; P = 5.0 × 10−5, Fig. 5C). Importantly, amine-reactive cross-linking agent, to label and capture oli- expression of MARK3 failed to induce S222 phosphorylation gomeric MEF2C complexes. Analysis of DSS-treated cells or activate the transcriptional activity of the MEF2C S222A revealed the formation of a reduced mobility complex, con- mutant that cannot be phosphorylated (Fig. 5C). Taken sistent with dimerization of MEF2C, which was significantly together, these data indicate that MARK kinase signaling can upregulated upon the loss of MEF2C S222 phosphorylation regulate MEF2C transcriptional activity specifically in an in MEF2C S222A–expressing cells as compared with wild-type S222 phosphorylation–dependent manner. controls (P = 7.9 × 10−4, t test; Fig. 4I; Supplementary Fig. Next, we tested the hypothesis that inhibition of MARK- S6K), consistent with previous studies of dominant-negative mediated phosphorylation of MEF2C may have antileuke- MEF2 mutants (46). Thus, MEF2C S222 phosphorylation mic efficacy. Genetic depletion of MARK3 using shRNA controls the assembly of its transcriptionally active complex, interference was not sufficient to reduce MEF2C pS222 which is associated with induction of both activating and phosphorylation (Supplementary Fig. S7D), consistent with repressive gene expression programs, leading to the dys- potentially redundant functions of MARK3, MARK2, and regulation of genes controlling leukemia cell survival such as MARK4. Therefore, we identified the ATP-competitive kinase LYL1, MYC, BMF, CASP8, and FOXO3. inhibitor MRT199665, which exhibits high selectivity and potency against MARK and the structurally similar SIK and MARK-Mediated Phosphorylation of MEF2C S222 MELK kinases as compared with other AMP kinase fam- Confers Susceptibility to MARK/SIK Inhibitors ily members (66). We found that SIK, NUAK, and MELK in AML kinases, which can be inhibited by MRT199665 at nanomolar The functional requirements of MEF2C S222 phospho- concentrations, scored in the bottom 45% of the recom- rylation for leukemia initiation and maintenance, combined binant kinase screen for MEF2C phosphorylation (Sup- with its dispensability for normal hematopoiesis, suggest plementary Data S3). Thus, we reasoned that MRT199665 that blockade of MEF2C phosphorylation may be exploited can be used as a selective inhibitor to block MEF2C S222

OF11 | CANCER DISCOVERY April 2018 www.aacrjournals.org

MEF2C Phosphorylation in Chemoresistant AML RESEARCH ARTICLE phosphorylation. To test this directly, we first investigated with untreated cells, as assessed by quantitative fluorescent whether MRT199665 treatment can phenocopy the suppres- western immunoblotting (Fig. 5G; Supplementary Fig. S7G). sion of MEF2C phosphorylation–induced gene expression MRT199665 treatment also caused a decrease of total MEF2C program by the dominant-negative MEF2C S222A mutant. protein expression, suggesting that its phosphorylation can Transcriptome profiling of OCI-AML2 cells treated with affect MEF2C stability or degradation, similar to previous 100 nmol/L MRT199665 for 48 hours revealed significant reports (53). Notably, we observed that MRT199665 treat- gene expression changes as compared with vehicle control- ment led to dose-dependent loss of MEF2C S222 phospho- treated cells, where 191 and 128 genes were upregulated and rylation in OCI-AML2 cells expressing wild-type MEF2C, downregulated, respectively (Supplementary Fig. S7E and but not MEF2C S222A, consistent with its specific effects on Data S4). GSEA revealed that MRT199665-treated cells and MEF2C S222 phosphorylation (Supplementary Fig. S7H–S7I). MEF2C S222A–expressing cells exhibit similar changes in Thus, MRT199665 treatment can block MEF2C S222 phos- gene expression (r = 0.64, Fig. 5D), including downregula- phorylation in AML cells. tion of cell-cycle and MYC target genes and upregulation Consistent with this mechanism, we found that human of genes mediating apoptosis (Fig. 5E; Supplementary Data AML cell lines with endogenous MEF2C phosphorylation S4). Furthermore, treatment of OCI-AML2 and MOLM-13 (OCI-AML2, MV4-11, MOLM-13, and Kasumi-1) were more cells with increasing concentrations of MRT199665 led to a sensitive to MRT199665 as compared with cell lines lack- dose-dependent reduction in total and pS222 MEF2C (Fig. ing MEF2C (NB-4, HEL, HL-60, and U937), with mean 50%

5F; Supplementary Fig. S7F), causing more than 40% reduc- inhibitory concentrations (IC50) of 26 ± 13 versus 990 ± tion in MEF2C phosphorylation at 10 nmol/L as compared 29 nmol/L, respectively (P = 5.6 × 10−5, t test, Fig. 5H),

A BC 25 ** 60 MARK3 0.4 20 MARK4 15 40 MARK1 0.3 MARK2 *** RSK1 0.2 10

20 * control) (fold * (MSA)

0.1 anscriptional activity 5 **** 0 Tr 0 0.0 Vector control

Inhibition (% control) ++

Phosphorylation activity ++ −20 MARK3 ++ + MEF2C + + + + 040 80 120 160 Staurosporine MEF2C S222A + + MEF2C + + + + + Kinase + MARK3 CAMK1a + MARK3 T211A + MW (kDa) 70 – MARK3 53 – MEF2C D 53 – pS222 MEF2C 41 – Actin r = 0.64 E 12 Chang_Cycling_Genes GNF2_CASP1_Apoptosis Yu_Myc_Targets_UP GNF2_TNFRSF1B_Apoptosis Reactome_DNA_Replication GNF2_CARD15_Caspase GNF2_CDC2_Cell Cycle HOXA9_DN.V1_UP Hallmark_E2F_Targets Hematopoietic_Stem Cell_DN MORF_RFC4_Replication Mo_RUNX1_RUNX1T1_Fusion_DN

− 10 Lee_Early_T_Lymphocyte_UP GSK3_Inhibitor_SB216763_UP GNF2_CCNA2_Cell Cycle Ramalho_Stemness_DN

− 2 −3 −2 −100123 Normalized enrichment score

MRT199556-treated cells (gene set NES) MRT199556-treated −3 −2 −1012 S222A-expressing cells (gene set NES)

Figure 5. Chemical inhibition of MARK-induced MEF2C phosphorylation exhibits selective toxicity against MEF2C-activated human AML cells. A, Recombinant screen for serine kinases that phosphorylate MEF2C S222, as assayed by significant pS222 MEF2C product inhibition marked in red. B, Phosphorylation activity of recombinant MARK3 (red) as compared with control CAMK1α (black) on model substrate as product inhibited by synthetic pS222 MEF2C peptide. Staurosporine serves as positive control. Error bars, SD of the mean for 3 biological replicates. *, P = 4.1 × 10−5 for MARK3 activity with and without MEF2C, t test. MSA, mobility shift assay. C, Activity of luciferase transcriptional MEF2 reporter in HEK293T cells lentivirally transduced with MEF2C, MARK3 or their mutants, as indicated. Error bars, SD of the mean for 3 biological replicates. *, P = 2.3 × 10−4; **, P = 3.4 × 10−5; ***, P = 5.0 × 10−5; ****, P = 1.5 × 10−7 for MEF2C S222A;vector versus MEF2C;vector, MEF2C;MARK3 versus MEF2C;vector, MEF2C;MARK3 T211A versus MEF2C;MARK3, and MEF2C S222A;MARK3 versus MEF2C;MARK3, respectively (t test). Below, western immunoblot analysis demonstrating equal protein expression of MEF2C and MARK3 transgenes, with reduced S222 phosphorylation by expression of MEF2C S222A and MARK3 T211A mutants. D, Correlation analysis of differentially expressed gene sets between S222A-expressing OCI-AML2 cells and MRT199665-treated OCI-AML2 cells. r = 0.64, Pearson correlation coefficient. E, GSEA of significantly upregulated and downregulated gene sets. continued( on next page)

April 2018 CANCER DISCOVERY | OF12

RESEARCH ARTICLE Brown et al.

F G 1.2 MEF2C pS222 MEF2C MRT199665 0 1,000 nmol/L 1.0 MW (kDa) MARK3 70 – β -actin) 0.8 53 – MEF2C 0.6 53 – pS222 MEF2C 0.4

Protein expression 0.2 41 – Actin (normalized to 0.0 DMSO 10 100 500 1,000 MRT199665 (nmol/L)

HI100 J NS 125 MOLM-13 80 100 MV4-11 * 103 MEF2C+ 60 (nmol/L) 75 OCI-AML2 50 Kasumi-1 50 40 NB-4 102 ** 25 MEF2C– HEL (% DMSO control) 20 (% DMSO control) 0 U937 Normalized luminescence 1 Normalized luminescence HL-60 0 IC MRT199665 10 0 100 101 102 103 104 105 OCI- MEF2C S222A S222D Primary MEF2C+ MEF2C– AML2 MRT199665 (nmol/L) AML AML AML

Figure 5. (Continued) F, Western immunoblot for MEF2C, pS222 MEF2C, and MARK3 in OCI-AML2 cells treated with increasing concentrations of MRT199665 for 12 hours. G, Quantitative analysis of MEF2C and pS222 MEF2C abundance, as measured using quantitative fluorescence immunoassays, and normalized to actin, demonstrating significant reduction of MEF2C phosphorylation by MRT199665 as compared with DMSO control P( = 5.5 × 10−3 and 3.4 × 10−4 for 500 and 1,000 nmol/L, respectively; t test). Error bars, SD of the mean for 2 biological replicates. H, Growth of human AML cell lines at 48 hours as a function of MRT199665 concentration for MEF2C-expressing cells (red) as compared to those that lack MEF2C (black), as indicated. Error bars, SD of the mean for 3 biological replicates. I, Growth of OCI-AML2 cells lentivirally transduced with wild-type MEF2C or dominant-negative MEF2C S222A or MEF2C S222D transgenes and treated with 600 ng/mL doxycycline to induce transgene expression and 100 nmol/L MRT199665 for 72 hours. (*, P = 3.7 × 10−2; **, P = 1.3 × 10−3 for MEF2C and S222A versus OCI-AML2, respectively; t test). Error bars, SD of the mean for 3 biological replicates. J, MRT199665 IC50 values for primary patient AML specimens and human AML cell lines with (red) and without (black) MEF2C activation upon 48 hours of drug treatment in vitro. Box plots represent mean and quartile, with whiskers denoting maximum and minimum values. and displayed reduced leukemia growth in clonogenic assays as compared with cells expressing wild-type MEF2C, MEF2C (Supplementary Fig. S7J). Importantly, expression of MEF2C S222D, or untransduced control OCI-AML2 cells (P = 8.8 × in OCI-AML2 cells conferred enhanced susceptibility to 10−8 and P = 3.4 × 10−9, respectively, χ2 test; Fig. 6A and B). MRT199665, which was partially rescued by the expression To investigate this therapeutic effect in vivo, we transplanted of MEF2C S222D that models constitutive MEF2C S222 OCI-AML2 cells induced to express MEF2C S222A as com- phosphorylation (Fig. 5I). Likewise, we found that primary pared with wild-type MEF2C into NSG mice, induced expres- patient AML specimens with MEF2C activation including sion of transgenes by doxycycline chow, and treated leukemic MLL-rearranged and non–MLL-rearranged leukemias (Sup- mice with cytarabine or PBS control (Fig. 6C and D). We plementary Table S4 and Supplementary Fig. S8A) treated in found that mice with leukemias expressing MEF2C S222A short-term cultures ex vivo exhibited enhanced susceptibility were significantly sensitized to cytarabine treatmentin vivo as and apoptosis in response to MRT199665 treatment as com- compared with MEF2C wild-type expressing controls (P = 6.8 × −5 pared with AML cell lines lacking MEF2C expression (IC50 = 10 , log-rank test; Fig. 6D). 280 ± 136 vs. 1300 ± 360 nmol/L, respectively, P = 1.8 × 10−5, Because MEF2C phosphorylation is required for AML t test, Fig. 5J; Supplementary Fig. S8B–S8D). Thus, activation chemotherapy resistance, we investigated the antileukemic of MEF2C in human AML confers susceptibility to MARK/ efficacy of MARK/SIK kinase inhibition in combination with SIK kinase inhibition. chemotherapy. Because MRT199665 exhibits rapid metabo- lism and elimination in mice and consequently is not suitable MEF2C Phosphorylation Is Required for for animal studies (67), its effects on chemotherapy resist- Chemotherapy Resistance in AML ance could only be assessed in vitro. We found that human Because MEF2C phosphorylation was specifically observed AML cell lines with endogenous MEF2C phosphorylation in diagnostic AML specimens in patients with failure of (OCI-AML2, MV4-11, MOLM-13, and Kasumi-1) were sig- induction therapy (Fig. 1), we hypothesized that MEF2C nificantly sensitized to cytarabine in the presence of 100 S222 phosphorylation induces chemotherapy resistance. In nmol/L MRT199665, as compared with cell lines lacking agreement with this prediction, OCI-AML2 cells induced to MEF2C (NB-4, HEL, HL-60, and U937) that remained unaf- express dominant-negative MEF2C S222A were significantly fected by MRT199665 (P = 0.024, paired t test; Fig. 6E). Simi- more sensitive to both cytarabine and doxorubicin treatment larly, patient-derived primary AML specimens with MEF2C

OF13 | CANCER DISCOVERY April 2018 www.aacrjournals.org

MEF2C Phosphorylation in Chemoresistant AML RESEARCH ARTICLE

A B

125 125 OCI-AML2 OCI-AML2 100 MEF2C 100 MEF2C S222A S222A 75 S222D 75 S222D

50 50

(% PBS control) 25 25 (% PBS control) Normalized luminescence

0 Normalized luminescence 0

0100 101 102 103 104 0100 101 102 103 104 Cytarabine (nmol/L) Doxorubicin (nmol/L)

CDDoxycycline 100 100

75 75 S222A

50 OCI-AML2 50 rcent survival rcent survival MEF2C Pe 25 Pe 25 OCI-AML2 S222A MEF2C 0 0 02550 75 100 02550 75 100 125 Time (days) Time (days) Vehicle Cytarabine EF 1,000 1,000 100

10 Normal CD34

(nmol/L) MSK21 (nmol/L)

50 1

50 100 * 0.1

0.01 MSK17 MSK14 Cytarabine IC

0.001 Cytarabine IC 10 MSK54 0.0001 −MRT +MRT −MRT +MRT −MRT +MRT MEF2C+ MEF2C−

Figure 6. MARK kinase inhibition overcomes chemotherapy resistance of MEF2C-activated AML cell lines and patient cells. Growth of OCI-AML2 cells lentivirally transduced with wild-type MEF2C or dominant-negative MEF2C S222A or MEF2C S222D transgenes and treated with 600 ng/mL doxycycline to induce transgene expression and increasing concentrations of cytarabine (A) and doxorubicin (B) for 72 hours. Error bars, SD of the mean for 3 biological replicates. P = 8.8 × 10−8 and 3.4 × 10−9 for MEF2C versus MEF2C S222A by nonlinear regression for cytarabine and doxorubicin, respectively. C, Kaplan–Meier survival curves of NSG mice transplanted with OCI-AML2 cells transduced with wild-type MEF2C and dominant-negative MEF2C S222A transgenes, and (D) treated with doxycycline in chow 3 days following transplantation continuously in vivo. One week following transplant, animals were treated with vehicle or cytarabine (blue arrow) intraperitoneally for 5 days. P = 6.8 × 10−5 MEF2C versus S222A cytarabine treated, + log-rank test for 10 animals per group. E, Cytarabine IC50 values for human AML cell lines with MEF2C activation (MEF2C ) as compared with those lacking MEF2C (MEF2C−) after 48 hours of drug treatment in the absence (−MRT) or presence (+MRT) of 100 nmol/L MRT199665. Each data point represents the mean of biological triplicates of an individual sample. *, P = 0.024 for +MRT versus −MRT for MEF2C-activated cells by paired t test. F, Cytarabine IC50 values for primary patient AML specimens or normal CD34 cells with MEF2C after 48 hours of drug treatment in the absence (−MRT) or presence (+MRT) of 100 nmol/L MRT199665 in vitro. Each data point represents the mean of biological triplicates of an individual sample.

April 2018 CANCER DISCOVERY | OF14

RESEARCH ARTICLE Brown et al. activation (2 MLL-rearranged and 2 non–MLL-rearranged) mechanisms may be engaged by distinct genetic and molecu- exhibited enhanced susceptibility to cytarabine treatment in lar classes of AML, such as for example distinct groups the presence of MRT199665 in vitro (IC50 = 58 ± 76 vs. 230 ± of mutations and/or their induction in specific leukemia- 280 nmol/L in the presence vs. absence of MRT199665, initiating cell populations. Likewise, recurrent mutations of respectively, P = 0.036, t test, Fig. 6F), in contrast to healthy MEF2B, MEF2C, and MEF2D in refractory lymphoid cancers CD34+ human blood umbilical cord progenitor cells, which (26–28) suggest that MEF2 family members may regulate showed no sensitization (Fig. 6F). Taken together, these essential survival or homeostatic mechanisms in hematopoi- results indicate that MEF2C phosphorylation is required etic cells, which cause therapy resistance in myeloid and lym- for AML chemotherapy resistance, which can be blocked by phoid malignancies. However, whereas mutational activation MARK/SIK kinase inhibition. of MEF2C in refractory T-cell acute lymphoblastic leukemias appears to inhibit apoptosis via NR4A1/NUR77-mediated effects on BCL2 (35), apoptosis resistance induced by MEF2C DISCUSSION phosphorylation in AML appears to involve BMF, CASP8, Our current findings indicate that MEF2C S222 phos- and FOXO3 dysregulation of BIM and BID (Fig. 4). Addition- phorylation is a specific marker of primary chemotherapy ally, MEF2C phosphorylation–dependent regulation of LYL1 resistance and failure of induction chemotherapy in patients expression suggests that its oncogenic functions in AML may with both cytogenetically normal and chromosomally rear- contribute to the co-option of hematopoietic stem cell pro- ranged AMLs (Fig. 1). Activity of S222-phosphorylated grams (71), at least in part mediated by aberrant transcription MEF2C appears to be aberrant, insofar as mice genetically factor complexes (72), which remains an important question engineered to block its phosphorylation exhibit normal for future studies. steady-state and stress hematopoiesis, but are resistant to We observe that MEF2C is both phosphorylated and leukemogenesis induced by the MLL–AF9 oncogene in vivo highly expressed in AML, and its leukemogenic activities (Fig. 2). At least in part, this effect is due to the requirement may therefore be due to both its high abundance and S222 for MEF2C phosphorylation for the survival of leukemia phosphorylation. Changes in MEF2C abundance and S222 stem or initiating cells (Fig. 3) and enhanced transcriptional phosphorylation may affect its activity by (i) recruitment of activity and induction of gene expression programs regu- distinct coactivators or corepressors, (ii) altered dimeriza- lating cell survival and apoptosis (Fig. 4). Finally, MARK tion with endogenous MEF2 family members, and/or (iii) kinases can specifically phosphorylate MEF2C, potentiating altered target gene localization due to secondary changes its transcriptional activity (Fig. 5), and inhibition of MEF2C in the posttranslational modifications and sequence bind- phosphorylation can overcome chemotherapy resistance of ing preferences of its DNA binding domain (47, 73–75). We MEF2C-activated human AML cell lines and patient leuke- found that MEF2C S222 phosphorylation can regulate the mias (Fig. 6). assembly of its transcriptionally active complex, which is Resistance to chemotherapy remains the major barrier to associated with its induction of both activating and repres- improving clinical outcomes for patients with AML, yet our sive gene expression programs. This leads to the dysregula- current concepts of chemoresistance lack sufficient explana- tion of genes controlling cell survival such as LYL1, MYC, tory power. For example, inactivation of TP53 and overex- BMF, CASP8, and FOXO3 (Fig. 4), consistent with the previ- pression of xenobiotic transporters have been found to cause ously reported activities of MEF2C as both transcriptional chemoresistance, but the majority of patients with primary activator and repressor (76). This dysregulation is likely refractory or relapsed AML lack identifiable mutations of important for both AML pathogenesis and chemotherapy TP53 or known components of TP53-dependent DNA dam- resistance, and it will be important to determine whether age response, and do not exhibit xenobiotic transporter over- shared mechanisms regulate leukemia-initiating cell popu- expression (1–3). Likewise, we do not yet understand how lations and those resistant to chemotherapy. Because our recently identified groups of genetic mutations associated functional studies were largely based on MLL–AF9-rear- with inferior clinical outcomes cause chemotherapy resist- ranged and RUNX1-mutant leukemias, it is possible that ance (3). Recently, impaired susceptibility to mitochondrial MEF2C phosphorylation may involve distinct molecular apoptosis has been associated with chemoresistance in AML mechanisms in different leukemia subtypes. It is also pos- (68, 69). However, the molecular mechanisms responsible for sible that the marked phenotype of nonphosphorylatable these differences also remain poorly understood. Our find- MEF2C mutants may be due to dominant-negative effects ings identify kinase-dependent dysregulation of transcription on other MEF2 family members, given their apparent dimer- factor control as a determinant of therapy response in AML, ization. Similarly, although MARK3 can specifically phos- at least in part mediated by aberrant survival and apoptosis phorylate MEF2C S222, additional serine kinases inhibited resistance of leukemia stem cells (Fig. 7). This suggests that by MRT199665, or those which were not included in our varied mechanisms of chemotherapy resistance and disease screen, such as mTOR, may induce MEF2C phosphorylation relapse in patients may be ultimately caused by aberrant gene in distinct types of AML. expression programming of privileged leukemia cell subsets The MARK family was originally discovered based on that provide a chemoresistant disease reservoir (70). its functions in controlling cell polarity and microtubule The observation of MEF2C S222 phosphorylation in dynamics as part of the basolateral polarity complex (77, 78), genetically diverse AML subtypes, including cytogenetically but their functions in hematopoietic cells are not yet defined normal and MLL-rearranged leukemias, raises the possibil- (79). Our finding that MARKs can regulate MEF2C sug- ity that common gene expression programs and regulatory gests that the basolateral polarity complex may have distinct

OF15 | CANCER DISCOVERY April 2018 www.aacrjournals.org

MEF2C Phosphorylation in Chemoresistant AML RESEARCH ARTICLE

A Normal hematopoiesis

MARK kinases P P MEF2C Normal hematopoiesis

MARK S222A kinases

MEF2C Normal hematopoiesis

B Leukemia

MARK kinases P P MEF2C BMF, CASP8, FOXO3 Cell survival

LYL1, MYC LSC maintenance

MARK MRT199665 S222A kinases

MEF2C BMF, CASP8, FOXO3 Apoptosis

LYL1, MYC LSC defect

Figure 7. Model of kinase-dependent dysregulation of MEF2C in AML and chemotherapy resistance. A, MEF2C S222 phosphorylation is regulated by MARK kinases and is dispensable for normal hematopoiesis, but (B) is required for enhanced leukemia stem cell (LSC) survival and chemotherapy resistance via control of genes such as LYL1, MYC, BMF, CASP8, and FOXO3. This differential functional dependency can be exploited for therapy and blockade of chemotherapy resistance using selective MARK kinase inhibition.

functions in normal hematopoiesis, which may be dys- Finally, the lack of apparent phenotypes in knock-in regulated in leukemia cells. Similarly, the mechanisms of mutant mice homozygous for Mef2cS222A/S222A or Mef2cS222D/S222D MARK3-mediated phosphorylation of MEF2C in chemother- suggests that phosphorylation of MEF2C S222 is dispensable apy-resistant AML remain to be defined. For example, MARK for normal development, establishing a compelling thera- kinases can be phosphorylated and activated by GSK-3β (80), peutic window and substantiating its therapeutic targeting. which can in turn by activated by integrin signaling, includ- However, additional studies of MEF2C phosphorylation in ing reports of its activation in AML (81). Thus, activation of specific physiologic states (24, 25), and further preclinical MARK signaling and MEF2C phosphorylation may be linked development of selective non-pyrrolopyrimidinone MARK to autocrine or paracrine signaling by AML cells and the bone inhibitors, will be needed to advance their use as MEF2C- marrow niche (29, 82, 83). targeted therapeutics. Future functional profiling studies of

April 2018 CANCER DISCOVERY | OF16

RESEARCH ARTICLE Brown et al. large AML cohorts should establish the prevalence of MEF2C generated using CRISPR/Cas9 genome editing and homologous phosphorylation and other mechanisms of primary chemo- recombination. A CRISPR gRNA targeting exon 5 of the Mef2c gene therapy resistance. was designed (GGAATGGATACGGCAACCCC) and expressed in vitro using the approach previously described (85). The donor templates for recombination were designed according to Wefers and colleagues METHODS (86), which produced in-frame substitutions and the addition of restriction sites for genotyping. Mouse zygotes from C57BL/6J:CBA Collection of Patient Samples F2 hybrids were injected with 100 ng/μL of gRNA, 50 ng/μL of Cas9 Written informed consent and approval by the Institutional mRNA, and 50 ng/μL of donor oligonucleotide into the pronucleus. Review Boards of participating institutions was obtained in accord- Founder mice were first screened using restriction endonuclease ance with the Declaration of Helsinki for all subjects. Primary digestion of genomic DNA isolated from tail tissues and positive chemotherapy resistance was defined based on the presence of at mice analyzed by PCR. For additional details, see the Supplementary least 5% of abnormal blasts by morphologic and immunophenotypic Methods. For all experiments using Mef2c knock-in mutant animals, assessment of bone marrow aspirates obtained after two cycles of wild-type littermates were used as controls. induction chemotherapy, as assessed by the respective institutional or central pathologic reviews. Specimens were collected and leukemia Murine Bone Marrow Transplantation Models cells purified as previously described (16). Mononuclear cells were For primary MLL–AF9 and Hoxa9/Meis1 mouse leukemia trans- purified using Ficoll gradient centrifugation, and leukemia cells were plants, 150,000 and 500,000 transduced cells, respectively, were trans- purified by negative immunomagnetic selection against CD3, CD14, planted by intravenous injection into lethally irradiated (900 rad) CD19, and CD235a, based on the absence of their expression by the C57BL/6J animals with 125,000 support wild-type bone marrow cells majority of AML specimens. and monitored for the development of leukemia. For Runx1−/−;Flt3ITD mouse leukemia transplants, 750,000 and 50,000 secondary leuke- Production and Purification of pS222 MEF2C Antibody mia bone marrow cells harvested from moribund primary recipients A phospho-specific antibody against MEF2C S222 (RefSeq ID: were transplanted by intravenous injection into sublethally irradiated NM_002397.4) was generated by PhosphoSolutions (Catalog p1208- (600 rad) C57BL/6J mice. For competitive bone marrow transplanta- 222, RRID:AB_2572427; PhosphoSolutions). Rabbits were immu- tion studies, 1 million CD45.2 mononuclear bone marrow cells from nized with a synthetic phosphopeptide corresponding to amino Mef2C transgenic mice were mixed with 1 million CD45.1 mononu- acids surrounding S222 of human MEF2C, conjugated with Key- clear bone marrow cells from B6.SJL-Ptprca Pepcb/BoyJ mice and hole limpet hemocyanin. Serum isolated from peripheral blood transplanted via tail-vein injection in lethally irradiated C57BL/6J of immunized rabbits was screened for phospho-specificity using recipient animals. Transplanted mice were monitored by CD45.1/ enzyme-linked immunosorbent assays (ELISA). ELISA-positive sera CD45.2 peripheral blood chimerism using FACS with the surface were pooled and sequentially affinity purified using both the phos- markers CD45.2 and CD45.1. Antibodies used in flow cytometry are phopeptide and non-phosphopeptide columns to isolate the affinity- detailed in Supplementary Table S5. purified pS222 MEF2C antibody. Human AML Xenograft Model Phosphoproteomics Screen NSG mice (8–10 weeks old) were sublethally irradiated (200 rad) Phosphoproteomic profiling was performed as described previ- and transplanted with 500,000 OCI-AML2 cells via tail-vein injection. ously (31). Briefly, purified leukemia cells were lysed using guani- Doxycycline-inducible transgene expression was induced using doxy- dine hydrochloride, and proteins were reduced with dithiothreitol, cycline chow 3 days after transplantation (Harlan). In xenotransplants alkylated with iodoacetamide, and digested using trypsin. Tryptic involving cytarabine treatment, 1.2 million OCI-AML2 cells were trans- peptides were purified by solid phase extraction, and purified pep- planted into nonirradiated NSG mice to account for the toxicity effects tides were labeled using iTRAQ reagents. Phosphopeptides were of cytarabine. Cytarabine in PBS (50 mg/kg/day) was administered enriched using iron affinity chromatography, and separated using 7 days after transplantation intraperitoneally for a total of 5 doses. three-dimensional RP-SAX-RP chromatography coupled to nanoelec- trospray ion source. Spectra were recorded using Orbitrap Velos mass Recombinant Kinase Screen spectrometer (ThermoFisher Scientific) in data-dependent mode. The protein kinase screen was performed using a recombinant ser- Data files were analyzed using multiplierz (84). Analysis of phospho- ine kinase library as previously described (87). Briefly, 172 serine/thre- proteomics data is detailed in Supplementary Methods. onine kinases (Supplementary Data S3) were individually expressed as N-terminal GST-fusion proteins in insect cells, and purified using Cell Lines glutathione sepharose chromatography. A synthetic peptide cor- Human AML cell lines were obtained from DSMZ and HEK293T responding to phosphoserine 222 for human MEF2C (RefSeq ID: cells from ATCC in 2013. Cell line authentication testing by STR NM_002397.4) Ac-GNPRN[pS]PGLLVC-NH2 was synthesized (Tufts genotyping (Genetica DNA Laboratories) for all cell lines was per- University), purified by HPLC, and confirmed by mass spectrometry. formed in May 2014 and February 2017. The absence of Mycoplasma For the screen, the MEF2C peptide was dissolved in DMSO at 10, 1, sp. contamination was verified most recently in August 2017 using and 0.1 μmol/L, and specific kinase activity on respective substrates Lonza MycoAlert (Lonza Walkersville, Inc.). Cell lines were used for was determined by the off-chip mobility shift assay using the Lab- experiments in this article within 20 passages from thawing. ChipTM3000 instrument (Caliper Life Sciences). Methods for analysis are detailed in Supplementary Methods. Mouse Studies All mouse studies were conducted with approval from the Memo- Data Accession rial Sloan Kettering Cancer Center Institutional Animal Care and RNA-seq and ATAC-seq are available at Gene Expression Omni- Use Committee. C57BL/6J, NOD-SCID IL2Rγ-null (NSG), and bus (GEO series number GSE94453, https://www.ncbi.nlm.nih. B6.SJL-Ptprca Pepcb/BoyJ mice were all obtained from The Jackson gov/geo/), and mass spectrometry proteomics data are available at Laboratories. Knock-in Mef2cS222A and Mef2cS222D mutant mice were MassIVE (MassIVE ID MSV000080646, https://massive.ucsd.edu/).

OF17 | CANCER DISCOVERY April 2018 www.aacrjournals.org

MEF2C Phosphorylation in Chemoresistant AML RESEARCH ARTICLE

Statistical Analysis 2. Cancer Genome Atlas Research N. Genomic and epigenomic land- scapes of adult de novo acute myeloid leukemia. N Engl J Med All in vitro experiments were performed a minimum of three inde- 2013;368:2059–74. pendent times with a minimum of three biological replicates. Statis- 3. Papaemmanuil E, Gerstung M, Bullinger L, Gaidzik VI, Paschka P, tical significance was determined using a two-tailed Studentt test for Roberts ND, et al. Genomic classification and prognosis in acute 2 continuous variables, Pearson χ test for limiting dilution assays, and myeloid leukemia. N Engl J Med 2016;374:2209–21. log-rank test for survival analysis. For ROC curve analysis, MEF2C 4. Zuber J, Rappaport AR, Luo W, Wang E, Chen C, Vaseva AV, et al. S222 phosphorylation expression above the median was considered An integrated approach to dissecting oncogene addiction implicates to be positive. Data were plotted as mean values with error bars rep- a Myb-coordinated self-renewal program as essential for leukemia resenting standard deviation. maintenance. Genes Dev 2011;25:1628–40. 5. Coombs CC, Tallman MS, Levine RL. Molecular therapy for acute Disclosure of Potential Conflicts of Interest myeloid leukaemia. Nat Rev Clin Oncol 2016;13:305–18. No potential conflicts of interest were disclosed. 6. de Rooij JD, Zwaan CM, van den Heuvel-Eibrink M. Pediatric AML: from biology to clinical management. J Clin Med 2015;4:127–49. Authors’ Contributions 7. Breems DA, Van Putten WL, Huijgens PC, Ossenkoppele GJ, Verhoef Conception and design: F.C. Brown, W. Mark, M. Gonen, G. Pouliot, GE, Verdonck LF, et al. Prognostic index for adult patients with acute S.A. Armstrong, A. Kentsis myeloid leukemia in first relapse. J Clin Oncol 2005;23:1969–78. 8. Burnett A, Wetzler M, Lowenberg B. Therapeutic advances in acute Development of methodology: F.C. Brown, E. Still, P. Cifani, myeloid leukemia. J Clin Oncol 2011;29:487–94. P. Romanienko, C. McCarthy, C. O’Donnell, A.V. Krivtsov, J.A. Marto, 9. Klco JM, Miller CA, Griffith M, Petti A, Spencer DH, Ketkar-Kulkarni A. Kentsis S, et al. Association between mutation clearance after induction Acquisition of data (provided animals, acquired and managed therapy and outcomes in acute myeloid leukemia. JAMA 2015;314: patients, provided facilities, etc.): F.C. Brown, E. Still, C.Y. Yim, 811–22. S. Takao, P. Cifani, C. Reed, S. Gunasekera, S.B. Ficarro, P. Romanienko, 10. Ding L, Ley TJ, Larson DE, Miller CA, Koboldt DC, Welch JS, et al. W. Mark, C. McCarthy, E. de Stanchina, C. O’Donnell, C. Stutzke, Clonal evolution in relapsed acute myeloid leukaemia revealed by J. Hébert, A. Melnick, E.M. Paietta, M.S. Tallman, G. Pouliot whole-genome sequencing. Nature 2012;481:506–10. Analysis and interpretation of data (e.g., statistical analysis, 11. Farrar JE, Schuback HL, Ries RE, Wai D, Hampton OA, Trevino LR, biostatistics, computational analysis): F.C. Brown, R.P. Koche, C.Y. et al. Genomic profiling of pediatric acute myeloid leukemia reveals Yim, S. Takao, P. Cifani, S.B. Ficarro, P. Romanienko, C. McCarthy, a changing mutational landscape from disease diagnosis to relapse. M. Gonen, V. Seshan, B. Spitzer, V.-P. Lavallée, A. Letai, G. Sauvageau, Cancer Res 2016;76:2197–205. R. Levine, S.A. Armstrong, A. Kentsis 12. Kihara R, Nagata Y, Kiyoi H, Kato T, Yamamoto E, Suzuki K, et al. Writing, review, and/or revision of the manuscript: F.C. Brown, Comprehensive analysis of genetic alterations and their prognos- C.Y. Yim, C. Reed, S.B. Ficarro, M. Gonen, V. Seshan, E.M. Paietta, M.S. tic impacts in adult acute myeloid leukemia patients. Leukemia Tallman, G. Pouliot, R. Levine, J.A. Marto, S.A. Armstrong, A. Kentsis 2014;28:1586–95. Administrative, technical, or material support (i.e., reporting 13. Guryanova OA, Shank K, Spitzer B, Luciani L, Koche RP, Garrett- or organizing data, constructing databases): F.C. Brown, E. Still, Bakelman FE, et al. DNMT3A mutations promote anthracycline P. Cifani, S. Gunasekera, P. Bhola, A.V. Krivtsov, A. Melnick, A. Kentsis resistance in acute myeloid leukemia via impaired nucleosome remodeling. Nat Med 2016;22:1488–95. Study supervision: F.C. Brown, A. Kentsis 14. Hollink IH, van den Heuvel-Eibrink MM, Arentsen-Peters ST, Prat- Other (created the model organism for the study): W. Mark corona M, Abbas S, Kuipers JE, et al. NUP98/NSD1 characterizes a Acknowledgments novel poor prognostic group in acute myeloid leukemia with a distinct HOX gene expression pattern. Blood 2011;118:3645–56. We are grateful to Nathanael Gray, Alejandro Gutierrez, Marc 15. Zuber J, Radtke I, Pardee TS, Zhao Z, Rappaport AR, Luo W, et al. Mansour, Leo Wang, and Michael Kharas for critical discussions. We Mouse models of human AML accurately predict chemotherapy thank John Schwarz and Hanna Mikkola for Mef2c conditional mice, response. Genes Dev 2009;23:877–89. Jeffrey Magee and Shaina Porter for Runx1−/−;Flt3ITD mice, Philip 16. Brown FC, Cifani P, Drill E, He J, Still E, Zhong S, et al. Genomics Cohen for MRT199665, and Ralph Garippa, Antoine Gruet, Matthew of primary chemoresistance and remission induction failure in pae- Witkin, and Yang Li for technical assistance. This work was sup- diatric and adult acute myeloid leukaemia. Br J Haematol 2017;176: ported by NIH R21 CA188881, R01 CA204396, P30 CA008748, U10 86–91. CA180827, and U24 CA196172, Burroughs Wellcome Fund, Josie Rob- 17. Hope KJ, Jin L, Dick JE. Acute myeloid leukemia originates from ertson Investigator Program, Rita Allen Foundation, Alex’s Lemonade a hierarchy of leukemic stem cell classes that differ in self-renewal Stand Foundation, American Society of Hematology, and Gabrielle’s capacity. Nat Immunol 2004;5:738–43. Angel Foundation. A. Kentsis is the Damon Runyon-Richard Lumsden 18. Somervaille TC, Cleary ML. Identification and characterization of Foundation Clinical Investigator. leukemia stem cells in murine MLL-AF9 acute myeloid leukemia. Cancer Cell 2006;10:257–68. 19. Krivtsov AV, Twomey D, Feng Z, Stubbs MC, Wang Y, Faber J, et al. The costs of publication of this article were defrayed in part by Transformation from committed progenitor to leukaemia stem cell the payment of page charges. This article must therefore be hereby initiated by MLL-AF9. Nature 2006;442:818–22. marked advertisement in accordance with 18 U.S.C. Section 1734 20. Laszlo GS, Alonzo TA, Gudgeon CJ, Harrington KH, Kentsis A, solely to indicate this fact. Gerbing RB, et al. High expression of myocyte enhancer factor 2C (MEF2C) is associated with adverse-risk features and poor outcome Received November 13, 2017; revised January 22, 2018; accepted in pediatric acute myeloid leukemia: a report from the Children’s January 30, 2018; published first February 5, 2018. Oncology Group. J Hematol Oncol 2015;8:115. 21. Schuler A, Schwieger M, Engelmann A, Weber K, Horn S, Muller U, et al. The MADS transcription factor Mef2c is a pivotal modulator of REFERENCES myeloid cell fate. Blood 2008;111:4532–41. 1. Schuback HL, Arceci RJ, Meshinchi S. Somatic characterization of 22. Stehling-Sun S, Dade J, Nutt SL, DeKoter RP, Camargo FD. Regula- pediatric acute myeloid leukemia using next-generation sequencing. tion of lymphoid versus myeloid fate ‘choice’ by the transcription Semin Hematol 2013;50:325–32. factor Mef2c. Nat Immunol 2009;10:289–96.

April 2018 CANCER DISCOVERY | OF18

RESEARCH ARTICLE Brown et al.

23. Schwieger M, Schuler A, Forster M, Engelmann A, Arnold MA, Delwel nominal mass and sub-Da isotopologue quantitative proteomic rea- R, et al. Homing and invasiveness of MLL/ENL leukemic cells is regu- gents. J Am Soc Mass Spectrom 2014;25:636–50. lated by MEF2C. Blood 2009;114:2476–88. 43. Lin Q, Schwarz J, Bucana C, Olson EN. Control of mouse cardiac 24. Wang W, Org T, Montel-Hagen A, Pioli PD, Duan D, Israely E, et al. morphogenesis and myogenesis by transcription factor MEF2C. MEF2C protects bone marrow B-lymphoid progenitors during stress Science 1997;276:1404–7. haematopoiesis. Nat Commun 2016;7:12376. 44. Li H, Radford JC, Ragusa MJ, Shea KL, McKercher SR, Zaremba JD, 25. Herglotz J, Unrau L, Hauschildt F, Fischer M, Kriebitzsch N, Alawi M, et al. Transcription factor MEF2C influences neural stem/progenitor et al. Essential control of early B-cell development by Mef2 transcrip- cell differentiation and maturation in vivo. PNAS 2008;105: tion factors. Blood 2016;127:572–81. 9397–402. 26. Ying CY, Dominguez-Sola D, Fabi M, Lorenz IC, Hussein S, Bansal 45. Du Y, Spence SE, Jenkins NA, Copeland NG. Cooperating cancer- M, et al. MEF2B mutations lead to deregulated expression of the gene identification through oncogenic-retrovirus-induced insertional oncogene BCL6 in diffuse large B cell lymphoma. Nat Immunol mutagenesis. Blood 2005;106:2498–505. 2013;14:1084–92. 46. Molkentin JD, Black BL, Martin JF, Olson EN. Mutational analysis 27. Gu Z, Churchman M, Roberts K, Li Y, Liu Y, Harvey RC, et al. of the DNA binding, dimerization, and transcriptional activation Genomic analyses identify recurrent MEF2D fusions in acute lymph- domains of MEF2C. Mol Cell Biol 1996;16:2627–36. oblastic leukaemia. Nat Commun 2016;7:13331. 47. Mao Z, Bonni A, Xia F, Nadal-Vicens M, Greenberg ME. Neuronal 28. Pon JR, Wong J, Saberi S, Alder O, Moksa M, Grace Cheng SW, activity-dependent cell survival mediated by transcription factor et al. MEF2B mutations in non-Hodgkin lymphoma dysregulate cell MEF2. Science 1999;286:785–90. migration by decreasing MEF2B target gene activation. Nat Commun 48. Xiu Y, Dong Q, Li Q, Li F, Borcherding N, Zhang W, et al. Stabiliza- 2015;6:7953. tion of NF-kappaB-inducing kinase suppresses MLL-AF9-induced 29. Kentsis A, Reed C, Rice KL, Sanda T, Rodig SJ, Tholouli E, et al. Auto- acute myeloid leukemia. Cell reports 2018;22:350–8. crine activation of the MET receptor tyrosine kinase in acute myeloid 49. Porter SN, Cluster AS, Yang W, Busken KA, Patel RM, Ryoo J, et al. leukemia. Nat Med 2012;18:1118–22. Fetal and neonatal hematopoietic progenitors are functionally and 30. Zheng R, Klang K, Gorin NC, Small D. Lack of KIT or FMS internal transcriptionally resistant to Flt3-ITD mutations. Elife 2016;5. tandem duplications but co-expression with ligands in AML. Leuk 50. Kroon E, Krosl J, Thorsteinsdottir U, Baban S, Buchberg AM, Sau- Res 2004;28:121–6. vageau G. Hoxa9 transforms primary bone marrow cells through 31. Ficarro SB, Zhang Y, Carrasco-Alfonso MJ, Garg B, Adelmant G, specific collaboration with Meis1a but not Pbx1b. EMBO J Webber JT, et al. Online nanoflow multidimensional fractionation 1998;17:3714–25. for high efficiency phosphopeptide analysis. Mol Cell Proteomics 51. Bach C, Buhl S, Mueller D, Garcia-Cuellar MP, Maethner E, Slany 2011;10:O111 011064. RK. Leukemogenic transformation by HOXA cluster genes. Blood 32. Wang LD, Ficarro SB, Hutchinson JN, Csepanyi-Komi R, Nguyen 2010;115:2910–8. PT, Wisniewski E, et al. Phosphoproteomic profiling of mouse pri- 52. Di Giorgio E, Gagliostro E, Clocchiatti A, Brancolini C. The control mary HSPCs reveals new regulators of HSPC mobilization. Blood operated by the cell cycle machinery on MEF2 stability contributes 2016;128:1465–74. to the downregulation of CDKN1A and entry into S phase. Mol Cell 33. Ficarro SB, Adelmant G, Tomar MN, Zhang Y, Cheng VJ, Marto Biol 2015;35:1633–47. JA. Magnetic bead processor for rapid evaluation and optimiza- 53. Badodi S, Baruffaldi F, Ganassi M, Battini R, Molinari S. Phospho- tion of parameters for phosphopeptide enrichment. Anal Chem rylation-dependent degradation of MEF2C contributes to regulate 2009;81:4566–75. G2/M transition. Cell Cycle 2015;14:1517–28. 34. Choe L, D’Ascenzo M, Relkin NR, Pappin D, Ross P, Williamson 54. Yu YT, Breitbart RE, Smoot LB, Lee Y, Mahdavi V, Nadal-Ginard B, et al. 8-plex quantitation of changes in cerebrospinal fluid pro- B. Human myocyte-specific enhancer factor 2 comprises a group tein expression in subjects undergoing intravenous immunoglobulin of tissue-restricted MADS box transcription factors. Genes Dev treatment for Alzheimer’s disease. Proteomics 2007;7:3651–60. 1992;6:1783–98. 35. Nagel S, Meyer C, Quentmeier H, Kaufmann M, Drexler HG, MacLeod 55. Zhu B, Gulick T. Phosphorylation and alternative pre-mRNA splicing RA. MEF2C is activated by multiple mechanisms in a subset of converge to regulate myocyte enhancer factor 2C activity. Mol Cell T-acute lymphoblastic leukemia cell lines. Leukemia 2008;22:600–7. Biol 2004;24:8264–75. 36. Cifani P, Kentsis A. High sensitivity quantitative proteomics using 56. Meng YS, Khoury H, Dick JE, Minden MD. Oncogenic potential automated multidimensional nano-flow chromatography and accu- of the transcription factor LYL1 in acute myeloblastic leukemia. mulated ion monitoring on quadrupole-Orbitrap-linear ion trap Leukemia 2005;19:1941–7. mass spectrometer. Mol Cell Proteomics 2017;16:2006–16. 57. Homminga I, Vuerhard MJ, Langerak AW, Buijs-Gladdines J, Pieters 37. Kasimir-Bauer S, Ottinger H, Meusers P, Beelen DW, Brittinger G, R, Meijerink JP. Characterization of a pediatric T-cell acute lympho- Seeber S, et al. In acute myeloid leukemia, coexpression of at least blastic leukemia patient with simultaneous LYL1 and LMO2 rear- two proteins, including P-glycoprotein, the multidrug resistance- rangements. Haematologica 2012;97:258–61. related protein, bcl-2, mutant p53, and heat-shock protein 27, is 58. Gerby B, Tremblay CS, Tremblay M, Rojas-Sutterlin S, Herblot S, predictive of the response to induction chemotherapy. Exp Hematol Hebert J, et al. SCL, LMO1 and Notch1 reprogram thymocytes into 1998;26:1111–7. self-renewing cells. PLos Genet 2014;10:e1004768. 38. Kang J, Gocke CB, Yu H. Phosphorylation-facilitated sumoylation 59. Pinon JD, Labi V, Egle A, Villunger A. Bim and Bmf in tissue homeo- of MEF2C negatively regulates its transcriptional activity. BMC Bio- stasis and malignant disease. Oncogene 2008;27:S41–52. chem 2006;7:5. 60. Hornsveld M, Tenhagen M, van de Ven RA, Smits AM, van Triest 39. McKinsey TA, Zhang CL, Olson EN. MEF2: a calcium-dependent MH, van Amersfoort M, et al. Restraining FOXO3-dependent tran- regulator of cell division, differentiation and death. Trends Biochem scriptional BMF activation underpins tumour growth and metastasis Sci 2002;27:40–7. of E-cadherin-negative breast cancer. Cell Death Differ 2016;23: 40. Ma K, Chan JK, Zhu G, Wu Z. Myocyte enhancer factor 2 acetylation 1483–92. by p300 enhances its DNA binding activity, transcriptional activity, 61. Essafi A, ernandezF de Mattos S, Hassen YA, Soeiro I, Mufti GJ, and myogenic differentiation. Mol Cell Biol 2005;25:3575–82. Thomas NS, et al. Direct transcriptional regulation of Bim by FoxO3a 41. Zhou H, Di Palma S, Preisinger C, Peng M, Polat AN, Heck AJ, et al. mediates STI571-induced apoptosis in Bcr-Abl-expressing cells. Toward a comprehensive characterization of a human cancer cell Oncogene 2005;24:2317–29. phosphoproteome. J Proteome Res 2013;12:260–71. 62. Li H, Zhu H, Xu CJ, Yuan J. Cleavage of BID by caspase 8 mediates 42. Ficarro SB, Biagi JM, Wang J, Scotcher J, Koleva RI, Card JD, et al. the mitochondrial damage in the Fas pathway of apoptosis. Cell Protected amine labels: a versatile molecular scaffold for multiplexed 1998;94:491–501.

OF19 | CANCER DISCOVERY April 2018 www.aacrjournals.org

MEF2C Phosphorylation in Chemoresistant AML RESEARCH ARTICLE

63. Zhao M, New L, Kravchenko VV, Kato Y, Gram H, di Padova F, et al. 75. Kim H, Lee JE, Kim BY, Cho EJ, Kim ST, Youn HD. Menin represses Regulation of the MEF2 family of transcription factors by p38. Mol JunD transcriptional activity in protein kinase C theta-mediated Cell Biol 1999;19:21–30. Nur77 expression. Exp Mol Med 2005;37:466–75. 64. Timm T, Balusamy K, Li X, Biernat J, Mandelkow E, Mandelkow 76. Di Giorgio E, Franforte E, Cefalu S, Rossi S, Dei Tos AP, Brenca M, EM. Glycogen synthase kinase (GSK) 3beta directly phosphorylates et al. The co-existence of transcriptional activator and transcriptional Serine 212 in the regulatory loop and inhibits microtubule affinity- repressor MEF2 complexes influences tumor aggressiveness. PLos regulating kinase (MARK) 2. J Biol Chem 2008;283:18873–82. Genet 2017;13:e1006752. 65. Nesic D, Miller MC, Quinkert ZT, Stein M, Chait BT, Stebbins CE. 77. Guo S, Kemphues KJ. par-1, a gene required for establishing polarity Helicobacter pylori CagA inhibits PAR1-MARK family kinases by in C. elegans embryos, encodes a putative Ser/Thr kinase that is asym- mimicking host substrates. Nat Struct Mol Biol 2010;17:130–2. metrically distributed. Cell 1995;81:611–20. 66. Clark K, MacKenzie KF, Petkevicius K, Kristariyanto Y, Zhang J, Choi 78. Shulman JM, Benton R, St Johnston D. The Drosophila homolog of HG, et al. Phosphorylation of CRTC3 by the salt-inducible kinases C. elegans PAR-1 organizes the oocyte cytoskeleton and directs oskar controls the interconversion of classically activated and regulatory mRNA localization to the posterior pole. Cell 2000;101:377–88. macrophages. PNAS 2012;109:16986–91. 79. Hurov JB, Huang M, White LS, Lennerz J, Choi CS, Cho YR, et al. Loss 67. Haidle AM, Childers KK, Zabierek AA, Katz JD, Jewell JP, Hou Y, et al. of the Par-1b/MARK2 polarity kinase leads to increased metabolic MARK inhibitors: declaring a No-Go decision on a chemical series rate, decreased adiposity, and insulin hypersensitivity in vivo. PNAS based on extensive DMPK experimentation. Bioorg Med Chem Lett 2007;104:5680–5. 2017;27:109–13. 80. Kosuga S, Tashiro E, Kajioka T, Ueki M, Shimizu Y, Imoto M. GSK- 68. Pan R, Hogdal LJ, Benito JM, Bucci D, Han L, Borthakur G, et al. 3beta directly phosphorylates and activates MARK2/PAR-1. J Biol Selective BCL-2 inhibition by ABT-199 causes on-target cell death in Chem 2005;280:42715–22. acute myeloid leukemia. Cancer Discov 2014;4:362–75. 81. Miller PG, Al-Shahrour F, Hartwell KA, Chu LP, Jaras M, Puram RV, 69. Pierceall WE, Kornblau SM, Carlson NE, Huang X, Blake N, Lena et al. In Vivo RNAi screening identifies a leukemia-specific depend- R, et al. BH3 profiling discriminates response to cytarabine-based ence on integrin beta 3 signaling. Cancer Cell 2013;24:45–58. treatment of acute myelogenous leukemia. Mol Cancer Ther 2013; 82. Zahreddine HA, Culjkovic-Kraljacic B, Assouline S, Gendron 12:2940–9. P, Romeo AA, Morris SJ, et al. The sonic hedgehog factor GLI1 70. Goardon N, Marchi E, Atzberger A, Quek L, Schuh A, Soneji S, et al. imparts drug resistance through inducible glucuronidation. Nature Coexistence of LMPP-like and GMP-like leukemia stem cells in acute 2014;511:90–3. myeloid leukemia. Cancer Cell 2011;19:138–52. 83. Zhou HS, Carter BZ, Andreeff M. Bone marrow niche-mediated sur- 71. Capron C, Lecluse Y, Kaushik AL, Foudi A, Lacout C, Sekkai D, et al. vival of leukemia stem cells in acute myeloid leukemia: Yin and Yang. The SCL relative LYL-1 is required for fetal and adult hematopoi- Cancer Biol Med 2016;13:248–59. etic stem cell function and B-cell differentiation. Blood 2006;107: 84. Parikh JR, Askenazi M, Ficarro SB, Cashorali T, Webber JT, Blank NC, 4678–86. et al. multiplierz: an extensible API based desktop environment for 72. Chen M, Zhu N, Liu X, Laurent B, Tang Z, Eng R, et al. JMJD1C is proteomics data analysis. BMC Bioinformatics 2009;10:364. required for the survival of acute myeloid leukemia by functioning 85. Romanienko PJ, Giacalone J, Ingenito J, Wang Y, Isaka M, Johnson as a coactivator for key transcription factors. Genes Dev 2015;29: T, et al. A vector with a single promoter for in vitro transcrip- 2123–39. tion and mammalian cell expression of CRISPR gRNAs. PLoS One 73. Lu J, McKinsey TA, Nicol RL, Olson EN. Signal-dependent activation 2016;11:e0148362. of the MEF2 transcription factor by dissociation from histone dea- 86. Wefers B, Ortiz O, Wurst W, Kuhn R. Generation of targeted mouse cetylases. PNAS 2000;97:4070–5. mutants by embryo microinjection of TALENs. Methods 2014;69:94–101. 74. Barneda-Zahonero B, Roman-Gonzalez L, Collazo O, Rafati H, Islam 87. Kitagawa D, Gouda M, Kirii Y, Sugiyama N, Ishihama Y, Fujii I, AB, Bussmann LH, et al. HDAC7 is a repressor of myeloid genes et al. Characterization of kinase inhibitors using different phospho- whose downregulation is required for transdifferentiation of pre-B rylation states of colony stimulating factor-1 receptor tyrosine kinase. cells into macrophages. PLos Genet 2013;9:e1003503. J Biochem 2012;151:47–55.

April 2018 CANCER DISCOVERY | OF20

MEF2C Phosphorylation Is Required for Chemotherapy Resistance in Acute Myeloid Leukemia

Fiona C. Brown, Eric Still, Richard P. Koche, et al.

Cancer Discov Published OnlineFirst February 5, 2018.

Updated version Access the most recent version of this article at: doi:10.1158/2159-8290.CD-17-1271

Supplementary Access the most recent supplemental material at: Material http://cancerdiscovery.aacrjournals.org/content/suppl/2018/02/03/2159-8290.CD-17-1271.DC1

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

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerdiscovery.aacrjournals.org/content/early/2018/03/15/2159-8290.CD-17-1271. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.