A Gain-Of-Function P53-Mutant Oncogene Promotes Cell Fate Plasticity and Myeloid Leukemia Through the Pluripotency Factor FOXH1

Published OnlineFirst May 8, 2019; DOI: 10.1158/2159-8290.CD-18-1391

RESEARCH ARTICLE

A Gain-of-Function p53-Mutant Oncogene Promotes Cell Fate Plasticity and Myeloid Leukemia through the Pluripotency Factor FOXH1

Evangelia Loizou1,2, Ana Banito1, Geulah Livshits1, Yu-Jui Ho1, Richard P. Koche3, Francisco J. Sánchez-Rivera1, Allison Mayle1, Chi-Chao Chen1, Savvas Kinalis4, Frederik O. Bagger4,5, Edward R. Kastenhuber1,6, Benjamin H. Durham7, and Scott W. Lowe1,8

ABSTRACT Mutations in the TP53 tumor suppressor gene are common in many cancer types, including the acute myeloid leukemia (AML) subtype known as complex karyotype AML (CK-AML). Here, we identify a gain-of-function (GOF) Trp53 mutation that accelerates CK-AML initiation beyond p53 loss and, surprisingly, is required for disease maintenance. The Trp53 R172H mutation (TP53 R175H in humans) exhibits a neomorphic function by promoting aberrant self-renewal in leukemic cells, a phenotype that is present in hematopoietic stem and progenitor cells (HSPC) even prior to their transformation. We identify FOXH1 as a critical mediator of mutant p53 function that binds to and regulates stem cell–associated genes and transcriptional programs. Our results identify a context where mutant p53 acts as a bona fi de oncogene that contributes to the pathogenesis of CK-AML and suggests a common biological theme for TP53 GOF in cancer.

SIGNIFICANCE: Our study demonstrates how a GOF p53 mutant can hijack an embryonic transcription factor to promote aberrant self-renewal. In this context, mutant Trp53 functions as an oncogene to both initiate and sustain myeloid leukemia and suggests a potential convergent activity of mutant Trp53 across cancer types.

INTRODUCTION capacity to transactivate wild-type target genes, despite being frequently stabilized owing to reduced interaction with neg- TP53 mutations occur in the majority of human cancers and ative regulators ( 10 ). These mutant proteins can instill neo- are often associated with poor outcomes (1, 2). TP53 encodes morphic gain-of-function (GOF) activities that contribute to a sequence-specifi c transcription factor (3) that is normally cancer phenotypes beyond p53 loss (11 ). At the organismal maintained at low levels through strict post-translational con- level, mice harboring certain germline missense mutations trol ( 4 ). In response to DNA damage, activated oncogenes, or in Trp53 develop an altered tumor spectrum compared with other forms of cellular stress, p53 is stabilized and promotes Trp53 -null mice, including a larger fraction of epithelial can- cell-cycle arrest, apoptosis, senescence, or other antiprolifera- cers with increased metastatic potential ( 12, 13 ). At the cellu- tive programs depending on cellular context ( 5–8 ). Most TP53 lar level, some GOF p53 mutants promote chemoresistance, mutations occur in the DNA binding domain and disrupt invasiveness, and/or an epithelial-to-mesenchymal transi- its transcriptional activity, thereby preventing these stress tion (EMT) through diverse mechanisms ( 12–14 ). Another responses and enabling aberrant proliferation and survival of neomorphic function of mutant p53 involves its ability to mutated cells ( 9 ). facilitate the formation of induced pluripotent stem cells Cancer-associated mutations typically inactivate p53 (iPSC) more so than p53 loss (15, 16), although the extent through a two-hit mechanism, whereby one allele acquires to which this GOF activity is relevant to cancer is poorly a missense mutation and the other undergoes “loss-of- understood. heterozygosity” (LOH) via chromosomal deletion ( 8 ). Mis- In contrast to their high prevalence in most solid tumors, sense TP53 mutations encode proteins that have attenuated TP53 mutations occur in around 10% of blood cancers, although when they occur they are associated with poor prognosis (17, 18). In acute myeloid leukemia (AML), TP53 1 Cancer Biology and Genetics Program, Sloan Kettering Institute, Memo- mutations are associated with a subtype known as complex rial Sloan Kettering Cancer Center, New York, New York. 2 Weill Cornell karyotype AML (CK-AML), which is defi ned by the presence Graduate School of Medical Sciences, New York, New York. 3 Center for of three or more cytogenetic abnormalities and a dismal Epigenetics Research, Memorial Sloan Kettering Cancer Center, New York, New York. 4 Center for Genomic Medicine, Rigshopitalet, University of 5-year survival rate of less than 2% ( 17, 18 ). Functional stud- Copenhagen, Copenhagen, Denmark. 5 Department of Biomedicine and ies in mice indicate that Trp53 inactivation in the hematopoi- Swiss Institute of Bioinformatics, UKBB Universitats-Kinderspital, Basel, etic compartment can produce chemoresistant malignancies Switzerland. 6 Louis V. Gerstner Jr. Graduate School of Biomedical Sci- with increased leukemia-initiating potential, mirroring key ences, Sloan Kettering Institute, Memorial Sloan Kettering Cancer Center, features linked to TP53 mutations in patients with AML New York, New York. 7 Human Oncology and Pathogenesis Program, Memo- rial Sloan Kettering Cancer Center, New York, New York. 8 Howard Hughes (19–21 ). Still, whether and how TP53 missense mutations Medical Institute, New York, New York. confer GOF activities to p53 in AML is not known. Note: Supplementary data for this article are available at Cancer Discovery In this study, we set out to test whether mutant p53 has Online (http://cancerdiscovery.aacrjournals.org/). GOF activity in AML and, if so, to determine the underlying Corresponding Author: Scott W. Lowe, Memorial Sloan Kettering Cancer mechanisms behind this effect. We chose to study Trp53 R172H Center, 417 E 68th Street, Box 373, New York, NY 10065. Phone: 646- (R175H in humans), a mutant form of Trp53 that has been 888-3342; Fax: 646-888-3347; E-mail: [email protected] shown to confer GOF activity in solid tumors and is the Cancer Discov 2019;9:962–79 most common allele in patients with AML (Dr. Elli Papaem- doi: 10.1158/2159-8290.CD-18-1391 manuil, personal communication). Several complementary ©2019 American Association for Cancer Research. in vitro and in vivo systems were used to compare the biological

JULY 2019 CANCER DISCOVERY | 963

RESEARCH ARTICLE Loizou et al. features of wild-type, Trp53-null, or Trp53-mutant alleles, functional evidence that a Trp53-mutant allele can contribute leading us to identify a neomorphic function of mutant to AML beyond the effects of p53 loss. p53 in hematopoietic stem and progenitor cells (HSPC) that exerts its effect by enhancing cellular self-renewal beyond that Sustained Expression of Mutant p53 Is Required R172H produced by p53 inactivation. We also identify a novel medi- for the Maintenance of p53 Leukemic Cells ator of mutant p53 function, FOXH1, which contributes to We next examined whether the p53R172H-mutant protein the aberrant self-renewal phenotype. As such, suppression of was required for survival of p53-mutant leukemic cells. We either mutant p53 or FOXH1 ablates this stemness capacity utilized previously characterized Trp53Δ/Δ or Trp53R172H/Δ and by triggering differentiation. These observations illustrate Trp53R172H/11B3 CK-AML murine cell lines that were created how mutant p53 can acquire a pro-oncogenic activity that by cotransduction of short hairpin RNAs (shRNA) target- magnifies loss of its tumor-suppressive functions and creates ing Nf1 (GFP-labeled) and Mll3 (mCherry-labeled), hereafter a previously unappreciated molecular dependency in AML. referred to as PNM (20, 26). These cell lines were generated to reflect the most common genetic configurations found in Δ/Δ R172H RESULTS patients, including Trp53 loss alone (Trp53 ), Trp53 over a focal Trp53 deletion (Trp53R172H/Δ), or Trp53R172H over a larger R172H p53 Accelerates the Onset of Hematologic genomic deletion that encompasses Trp53 (Trp53R172H/11B3; ref. Malignancies beyond Effects of p53 Deficiency 25). Control shRNAs against Renilla luciferase (sh.Ren) or We first compared the ability of a mutant or nullTrp53 allele Trp53 (sh.p53) were coexpressed with a blue fluorescent protein to promote leukemogenesis in a well-defined genetic model. (BFP) reporter (Supplementary Fig. S1A; ref. 27). The percent- Because of its previously defined GOF activity in other set- age of BFP+ cells was monitored over time as a measurement of tings, we used a conditional mutant Trp53 allele harboring fitness under the two shRNA-containing conditions. an R172H mutation downstream of a “lox-stop-lox” cassette We observed a significant reduction in the relative number (LSL-Trp53R172H; ref. 12), which corresponds to TP53R175H, the of sh.p53 BFP+ cells compared with sh.Ren BFP+ cells over time most frequent missense mutation in clinical cases of AML (18). (Fig. 2A, light blue and dark blue lines), indicating negative Mouse cohorts were produced to harbor one Trp53R172H allele selection against cells in which the mutant protein is depleted. and a Trp53 floxed allele (Trp53LSL-R172H/F), or were homozygous In contrast, sh.p53 had no effect on Trp53Δ/Δ AML cells, rul- for the floxedTrp53 allele (Trp53F/F), such that all hematopoietic ing out potential off-target effects of the p53 shRNA (Fig. cells would become Trp53R172H/Δ and Trp53Δ/Δ, respectively, in the 2A, black line). p53 depletion in cells harboring Trp53R172H presence of the hematopoietic-specificVav1-Cre transgene (22). promoted differentiation and triggered apoptosis as assessed Although both the Trp53R172H/Δ and Trp53Δ/Δ cohorts eventu- by flow cytometry for CD11b and Annexin V, respectively ally succumbed to disease with full penetrance, mice harboring (Fig. 2B and C). Thus, AML cells expressing p53R172H acquire the Trp53R172H allele became moribund significantly faster (Fig. a molecular dependency on mutant p53 that contributes to a 1A). As commonly observed in Trp53−/− mice (23), most animals differentiation block that sustains leukemogenesis. develop thymic lymphoma, although some had leukemias and, To determine whether human AML can also depend on rarely, a mixed disease (leukemia and lymphoma; Fig. 1B). sustained expression of mutant p53, we first examined pub- The above observations are consistent with those previously licly available CRISPR/Cas9 genome-wide screening data (28). obtained using mice harboring a humanized TP53R248Q allele In these studies, single guide RNAs (sgRNA) were introduced expressed in the whole body of mice, which also develop T-cell into human AML lines expressing Cas9 followed by assess- lymphoma faster than Trp53-null mice (24). However, neither ment of their enrichment or depletion over time using next- model addresses the biology of mutant p53 action in the clini- generation sequencing. Interestingly, sgRNAs targeting p53 cally relevant context of AML, likely due to the high penetrance were invariably enriched in AML lines harboring wild-type of lymphomas that arise in the mouse. Therefore, to bias the TP53 but depleted to varying degrees in those expressing model against T-cell lymphoma development, we transplanted mutant TP53 genes (e.g., R248Q, another hotspot TP53 muta- bone marrow (BM) cells from Mx1-Cre;Trp53R172H/F and Mx1- tion), suggesting that a subset of the latter group might Cre;Trp53F/F mice into thymectomized recipients. Three weeks depend on sustained expression of the mutant protein for post-transplantation, intraperitoneal injection of PolyI:C was their proliferation and/or survival (Supplementary Fig. S1B). used to activate the IFN-inducible Mx1-Cre, triggering the None of the AML cell lines used in the aforemen- recombination of floxed alleles in transplanted cells (25). Of tioned CRISPR/Cas9 screen harbored an R175H mutation note, the use of the inducible Mx1-Cre recombinase enabled (corresponding to the mutation we studied in mice). How- transplantation of functionally wild-type whole bone mar- ever, we identified a human AML cell line (KY821) that row (WBM) cells, avoiding biases in engraftment efficiency harbored this mutation in the absence of a wild-type that might occur upon p53 alteration. As in the autochtho- (WT) TP53 allele (Supplementary Fig. S1C) and used nous model, Trp53R172H/Δ mice succumbed to disease faster it to functionally test the requirement for this mutant than their Trp53-null counterparts (Fig. 1C), although, in form of TP53 in vitro and in vivo. Cells were transduced this instance, the disease was immunophenotypically and with a validated shRNA targeting human p53 (sh.TP53; histologically characterized as AML based on the expression ref. 29) or empty-vector control (Ctrl) and examined for of CD11b and Gr1 on the surface of infiltrating leukemic cells their ability to form colonies in methylcellulose culture (Fig. 1D and E). Thus, p53R172H gives rise to an aggressive leu- in vitro or give rise to leukemia upon transplantation into kemia that significantly shortens overall survival as compared sublethally irradiated NOD-SCID IL2R−/− (NSG) immuno- with the disease arising from p53-deficient BM cells, providing deficient mice. Similar to observations from murine AML,

964 | CANCER DISCOVERY JULY 2019 www.aacrjournals.org

A Native mice B Trp53∆ /∆ Trp53 R172H/∆ (Vav1-Cre) Trp53∆ /∆ 100 Lymphoma Trp53 R172H/∆ Leukemia Leukemia/ lymphoma 50

**** Percent survival 0 n = 15 n = 11 0 100 200 300 Time (days)

Transplant in thymectomized mice R172H /∆ C D Trp53∆ /∆ Trp53 (Mx1-Cre) 100 105 47.9 5.52 105 37.6 8.92 Trp53WT ** Trp53∆ /∆ 104 104 Trp53 R172H/∆ 3 3 50 10 10

CD11b-PE 0 0 3.53 3.28

Percent survival 43 50.2 0 0 103 104 105 0 103 104 105 0 100 200 300 400 Gr1-APC Time (days)

E BM Spleen Liver WT Trp53

50 µm / ∆ ∆ Trp53

50 µm ∆ R172 H/ Trp53

50 µm

Figure 1. Trp53R172H leads to accelerated onset of hematologic malignancies. A, Kaplan–Meier survival curves of native mice with Trp53Δ/Δ and Trp53R172H/Δ hematopoietic cells recombined using Vav1-Cre. B, Spectrum of hematologic malignancies arising in Trp53Δ/Δ (n = 15) and Trp53R172H/Δ (n = 11) native mice. C, Kaplan–Meier survival curves of thymectomized mice transplanted with Trp53f/f, Trp53f/f;Mx1-Cre and Trp53R172H/F;Mx1-Cre BM. D, Expression of CD11b and Gr1 in the peripheral blood of moribund mice from C. E, Representative hematoxylin and eosin image of mice with AML arising after transplant in thymectomized recipients. Black arrows show myeloid blasts.**, P < 0.005; ****, P < 0.0001, log-rank (Mantel–Cox) test.

JULY 2019 CANCER DISCOVERY | 965

RESEARCH ARTICLE Loizou et al.

AB1.5 CD11b C Annexin V Trp53∆ /∆ Trp53 R172H/∆ Trp53 R172H/∆

+ 40 8 Trp53 R172H/11B3 *** ** 1.0 R172H/∆ Trp53 6 30

4 20 0.5 Fold change 2 Annexin V (%) 10 sh.p53/sh.Ren BFP

0.0 0 0 2518 22 sh.Ren sh.p53 sh.Ren sh.p53 Time (days)

D TP53 EFKY821 (TP53 R175H) 100 Ctrl (5.94%) 1.5 100 *** 80 sh.TP53 (21.60%) *** 80 60 1.0 60 40

40 Count 0.5 CFU (#) 20 20

Relative expression 0 0 0 Ctrl sh.TP53 Ctrl sh.TP53 hCD16 - PeCy7

G HIMoribund mice 100 100 Ctrl 80 sh.TP53 60 Ctrl sh.TP53 sh.TP53_1sh.TP53_2sh.TP53_3 50 ** 40 p53 % hCD45 20 Percent survival ACTB 0 0 Ctrl/ sh.TP53 sh.TP53 010203040 50 Ctrl moribund mice Time (days) moribund mice

Figure 2. Depletion of mutant p53 impairs the growth of p53-mutant leukemias. A, The ratio of sh.p53 BFP+ cells compared with sh.Ren BFP+ cells over time in leukemic lines. CD11b expression in the shRNA BFP+ containing cells (n = 3; B) and Annexin V (C) shown by FACS 7 days after infection. Data presented as mean ± SEM. D, Relative expression of TP53 7 days postinfection with pRS sh.TP53 in human KY821 cells. E, Colony-forming units (CFU) in human AML cells with sh.Ren or sh.TP53. Data presented as mean ± SEM (n = 3). F, hCD16 expression by FACS in pRS Ctrl or pRS sh.TP53 cells from growing in CFU assay. G, Percentage of hCD45 in the peripheral blood of mice 5 weeks post-transplant in shCtrl and shCtrl moribund mice (black), sh.TP53 (blue), and sh.TP53 moribund (orange) mice. Data are represented as mean ± SEM (n = 5 and n = 5). H, Kaplan–Meier curve from NSG mice transplanted with equal numbers of KY821 cells transduced with either shCtrl or sh.TP53 (n = 5 and n = 5). I, Western blot analysis for human p53 in cells before injection and in sorted hCD45 cells from the peripheral blood of three independent moribund sh.TP53 mice. B, Kruskal–Wallis one-way ANOVA; C, D, E, H, unpaired t test. *, P < 0.05; **, P < 0.005; ***, P < 0.0005. suppression of mutant TP53 in KY821 cells significantly the survival of transplanted mice compared with controls (Fig. reduced their colony-forming ability (Fig. 2D and E) and led 2H). At the time of death, the resulting leukemias invariably to the upregulation of the myeloid differentiation marker selected against expression of the p53-targeting shRNA and CD16 (Fig. 2F), again suggesting that this p53 mutant con- restored the expression of mutant p53, consistent with our tributes to a differentiation block. observations from murine models (Fig. 2I; Supplementary Fig. To test whether TP53R175H-mutant cells transduced with S1D–S1F). Therefore, KY821 cells require mutant p53 protein sh.TP53 exhibited reduced fitnessin vivo, we analyzed periph- to sustain leukemia. Collectively, these results reveal an onco- eral blood for the presence of human CD45-positive leu- genic function for mutant p53 in murine and human AML. kocytes in recipient mice 5 weeks post-transplantation. R172H Remarkably, this analysis revealed a significant reduction p53 -Derived Leukemia Is Associated with of cells containing sh.TP53 versus control cells (Fig. 2G), Stem Cell Transcriptional Signatures confirming the suppressive effect of mutant p53 knock- Most studies suggest that the GOF activities of mutant down in p53 mutant–expressing human leukemias. In addi- p53 proteins arise from their ability to alter transcription in tion, knockdown of mutant p53 significantly increased a manner that is distinct from effects due to p53 loss (30).

966 | CANCER DISCOVERY JULY 2019 www.aacrjournals.org

A p53R172H/∆ vs. p53∆/∆ B 20 VERRECCHIA_EARLY_RESPONSE_TO_TGFB1 ROSS_AML_OF_FAB_M7_TYPE BOQUEST_STEM_CELL_UP 15 ANASTASSIOU_MULTICANCER_INVASIVENESS_SIGNATURE LIM_MAMMARY_STEM_CELL_UP Trp53 GRAHAM_CML_QUIESCENT_VS_NORMAL_QUIESCENT_UP 10 VALK_AML_CLUSTER_8 VERRECCHIA_RESPONSE_TO_TGFB1_C2 MCBRYAN_PUBERTAL_TGFB1_TARGETS_UP

−Log10 (padj) Foxh1 5 VALK_AML_CLUSTER_7 LIANG_HEM_STEM_CELL_NUMBER_LARGE_VS_TINY_DN REACTOME_EXTRACELLULAR_MATRIX_ORGANIZATION

0 0 1234

−10 −5 0510 –Log10 (FDR)

Log2 fold change C BIOCARTA STEM PATHWAY BIOCARTA TGFB PATHWAY *** *** ** *** ** 7.6 *** 4.6 * * 7.4 4.4 7.2 4.2 7.0 4.0 6.8 3.8 6.6 6.4

3.6 expression 2 expression

2 6.2 3.4

Log 6.0 Log 3.2 5.8 3.0 5.6 2.8 5.4

MY BC BC HSCMPPCMPGMPMEP MM PMNMono HSCMPPCMPGMPMEP MY MM PMNMono Early_PMLate_PM Early_PMLate_PM AML t(8;21) AML t(8;21) AML t(15;17) AML complex AML t(15;17) AML complex AML t(11q23)MLL AML t(11q23)MLL AML inv(16)/t(16;16) Cell types AML inv(16)/t(16;16) Cell types

D WONG__EMBRYONIC STEM_ BOQUEST_STEM_CELL_UP CELL_CORE LT-HSC HSC PASSEGUE

FDR = 0 FDR = 0.005 FDR = 0 FDR = 0 NES = 1.83 NES = 1.65 NES = 2.28 NES = 1.61

MUT NULL MUT NULL MUT NULL MUT NULL

Figure 3. Gene expression profile ofTrp53 R172H/Δ leukemias. A, Volcano plot from RNA-seq comparing Trp53R172H/Δ leukemias to Trp53Δ/Δ showing Trp53 as the most highly expressed gene between the two groups. B, Gene ontology (GO) analysis of pathways associated with genes upregulated in Trp53R172H/Δ leukemias. C, Expression of GO-associated pathways in human AML samples and normal human hematopoietic cells (www.bloodspot.eu). D, Gene set enrichment analysis comparing gene expression between Trp53R172H/Δ and Trp53Δ/Δ leukemias to other known signatures related to stem cells and hematopoietic stem cells (HSC). Two-tailed t test, *, P < 0.05; **, P < 0.005; ***, P < 0.0005. MPP, multipotential progenitors; early_PM, early promyelocyte; late_PM, late promyelocyte; MY, myelocyte; MM, metamyelocytes; BC, band cell; PMN, polymorphonuclear cells; mono, monocytes; LT-HSC, long-term HSC.

Therefore, to examine how mutant p53 altered the tran- the AML subtype in which missense TP53 mutations most scriptome of CK-AML cells, we performed gene expression commonly occur, compared with other leukemia subtypes profiling of PNMTrp53 -null and Trp53-mutant leukemias via or normal hematopoietic cells (www.bloodspot.eu; ref. 32; RNA sequencing (RNA-seq) and identified 252 significantly Fig. 3C). Beyond general stem cell signatures, gene set enrich- differentially expressed genes (Fig. 3A; Supplementary Table ment analysis (GSEA) indicated that p53-mutant AML was S1). A systematic and unbiased analysis of the data using gene enriched for genes linked to hematopoietic stem cells (HSC) ontology (GO) pathway analysis revealed TGFβ and stem cell and long term-HSC (LT-HSC; ref. 31; Fig. 3D). Therefore, pathways to be positively correlated with the top upregulated both murine and human mutant p53-expressing (but not genes in our Trp53R172H murine leukemias (Fig. 3B). Notably, p53-deficient) leukemias converge on similar transcriptional these same signatures were enriched in the transcriptional pathways, some of which are linked to stemness and inhibi- profiles of leukemias obtained from patients with CK-AML, tion of myeloid differentiation.

JULY 2019 CANCER DISCOVERY | 967

RESEARCH ARTICLE Loizou et al. p53R172H Enhances Self-Renewal in Hematopoietic exacerbated in secondary transplants (Fig. 4E). Analysis of Cells In Vitro and In Vivo peripheral blood (Fig. 4F) and BM (Fig. 4G) demonstrated that Trp53R172H/Δ cells were significantly enriched with respect AML arises from alterations (e.g., mutations) that pro- to Trp53WT cells. Furthermore, within the BM of Trp53R172H/Δ mote proliferation or enhance the self-renewal of myeloid mice, HSCs were expanded in relative and absolute number progenitor cells (33, 34). To directly examine the functional compared with Trp53Δ/Δ mice (Supplementary Fig. S3H–S3K). consequences of mutant p53 expression in the normal hemat- All downstream progenitors, including megakaryocyte/eryth- opoietic compartment, we isolated adult WBM cells from roid progenitors (MEP), granulocyte/macrophage progeni- Trp53WT/WT, Trp53Δ/Δ, and Trp53R172H/Δ mice and performed tors (GMP) and common myeloid progenitors (CMP), were limiting dilution cultures in methylcellulose medium fol- significantly increased inTrp53 R172H/Δ mice (Fig. 4H–J). lowed by serial replating. Under these cytokine-rich culture For further validation, we examined the consequences of conditions, wild-type cells lose their replating capacity after inhibiting mutant p53 in normal mouse HSCs. For these three sequential passages (35). As described previously (36), experiments, we used the BM of Trp53R172H/Δ mice to isolate loss of p53 modestly increased the replating capacity of hemat- lineage-negative (Lin−)/Sca1+/c-KIT+ cells (LSK), a cell popu- opoietic cells as compared with those expressing wild-type p53 lation capable of maintaining long-term hematopoiesis in (Fig. 4A). In stark contrast, hematopoietic cells expressing mice (40). To suppress p53, these cells were infected with Trp53R172H/Δ not only replated more frequently than Trp53Δ/Δ retroviruses encoding an shRNA (targeting mutant p53 or and WT controls, but also could replate indefinitely (Fig. 4A). Renilla control) linked to BFP (Fig. 4K). Equal numbers of This enhanced capability could not be explained by dif- either sh.Ren or sh.p53 cells were mixed with non–shRNA- ferences in the preexisting frequency of stem and progenitor expressing p53R172H BFP− cells and transplanted into CD45.1 cells in the BM, because immunophenotyping of 7-week-old lethally irradiated mice in an in vivo competition format. The mice (the time of BM collection) did not show an increase fraction of BFP+ cells in the peripheral blood was monitored in any type of stem or progenitor cell (Supplementary Fig. for 4 months. Trp53R172H/Δ BM cells transduced with the p53 S2A–S2K). This increase in self-renewal was observed in shRNA were progressively depleted from peripheral blood, Trp53R172H/Δ but not Trp53R172H/WT cells, suggesting that loss becoming nearly undetectable by the end of the experiment of the residual WT allele is a prerequisite for this GOF effect (Fig. 4L). (Supplementary Fig. S3A). Furthermore, Trp53R172H/Δ cells Analysis of BM at the final time point revealed that fewer exhibited higher nuclear-to-cytoplasmic ratio when com- cells expressing sh.p53 remained as compared with those pared with Trp53Δ/Δ and WT cells at the same passage (P3; expressing control sh.Ren (Fig. 4M). Moreover, the propor- Supplementary Fig. S3B) and expressed the highest levels tion of LSKs within the sh.p53 population was lower than of CD34 (Supplementary Fig. S3C), an important marker in the control sh.Ren population (Fig. 4N). Altogether, these of long-term repopulation of HSCs (37). The replating phe- observations establish a role for p53R172H in promoting aber- notype produced by mutant Trp53 was similar to that pro- rant self-renewal in adult hematopoietic cells in a manner duced by disruption of Tet2, a well-characterized leukemia that renders these cells dependent on its action for their tumor suppressor gene whose loss promotes enhanced self- maintenance in a premalignant setting. Hence, phenotypes renewal (Supplementary Fig. S3C; refs. 38, 39). Nevertheless, linked to p53-mutant GOF can emerge in the hematopoietic the Trp53R172H/Δ cells were different than Tet2Δ/Δ cells at the system prior to neoplastic transformation. immunophenotypic and morphologic levels (Supplementary Fig. S3C and S3D), suggesting subtle differences between FOXH1 Is Required for Mutant p53–Induced how these alterations act. Self-Renewal in Both Normal and Leukemic Cells To further explore the relationship between mutant p53 To explore the mechanism by which mutant p53 and self-renewal potential, we isolated c-KIT+ BM cells from enhances cellular self-renewal, we surveyed the genes that the Trp53R172H/Δ mice and transduced them with viral vectors were expressed at higher levels in Trp53R172H/Δ leukemias coexpressing an shRNA-targeting mutant p53 (sh.p53) or a compared with Trp53Δ/Δ leukemias. One of the most signifi- Renilla control (sh.Ren). These vectors encoded puromycin cantly upregulated genes in our analysis was Foxh1 (Supple- resistance, allowing us to select and maintain pure shRNA- mentary Table S1), which encodes a key transcription factor expressing cells in culture. This isogenic comparison validated that is known to mediate Nodal/TGFβ signaling during our previous observation: p53 knockdown in Trp53-mutant embryogenesis, but is not expressed at appreciable levels in cells halted replating capacity after the fourth passage (com- the adult hematopoietic compartment or any other mouse parable with Trp53Δ/Δ WBM), whereas sh.Ren cells (which tissue (Supplementary Fig. S4A–S4C). Like mutant p53 (41), retain expression of mutant p53) were able to replate indefi- FOXH1 has been linked to EMT (42, 43) and can facilitate nitely (Fig. 4B and C). We confirmed that the normalization of the reprogramming of fibroblasts into iPSCs (44). Further- self-renewal potential by sh.p53 was not due to an off-target more, many of the top upregulated genes in Trp53R172H/Δ effect of RNAi, as the same shRNA had no effect on the fitness leukemias were known FOXH1 targets (ref. 45; Fig. 5A). of Trp53Δ/Δ hematopoietic cells (Supplementary Fig. S2L). Taken together, these observations led us to hypothesize Consistent with these findings, competitive transplanta- that FOXH1 might be a key mediator of the p53R172H GOF tion studies using Trp53WT/WT, Trp53Δ/Δ, and Trp53R172H/Δ BM activity in leukemia. (Fig. 4D) demonstrated that Trp53R172H/Δ cells outcompeted In agreement, analysis of publicly available expression wild-type cells to a greater extent than Trp53Δ/Δ cells (Sup- data from normal human hematopoietic cells and differ- plementary Fig. S3E–S3G), a phenotype that was further ent leukemia subtypes revealed that FOXH1 expression was

968 | CANCER DISCOVERY JULY 2019 www.aacrjournals.org

300 AB80 Trp53 R172H /∆ Trp53 R172H /∆; sh.p53 C ∆ /∆ Trp53 250 Trp53 R172H /∆; sh.Ren Trp53 R172H /∆ Trp53 WT/WT 60 200 150 40 80 * sh.Ren sh.p53 60 p53 CFU (average) CFU (average) 20 40 * 20 ACTB 0 0 1234567 1 2345678910 P4 Platings Platings

D E F F/F Peripheral blood Trp53 80 100 F/F * Trp53 ; Mx1Cre 80 60 ** Trp53 R172H/F; Mx1Cre 60 40 500,000 40 CD45.2 1:1 CD45.1 20 CD45.2 (%) CD45.2 20

WT (%) PB in CD45.2 0 0 WT ∆ /∆ R172H /∆ CD45.1 500,000 01234 Trp53 Trp53 Trp53 Time (months)

GHBM MEP IJGMP CMP 80 1.0 0.8 0.25 * * * ** * 60 0.8 0.6 0.20 0.6 0.15 40 0.4

0.4 0.10

MEP (%) MEP CMP (%) CMP 20 (%) GMP 0.2 CD45.2 (%) CD45.2 0.2 0.05 0 0.0 0.0 0.00 Trp53 WT Trp53 ∆ /∆ Trp53 R172H/∆ Trp53 WT Trp53 ∆ /∆ Trp53 R172H/∆ Trp53 WT Trp53 ∆ /∆ Trp53 R172H/∆ Trp53 WT Trp53 ∆ /∆ Trp53 R172H/∆

K L M BFP+ in BM N BFP+ LSK n.s. sh.Ren 1.4 40 n.s. 0.6 1.2 Trp53 R172H /∆ 30

1.0 0.4 (%)

0.8 (%) + 20 + 0.6 *

sh.p53 0.4 *** 0.2 BFP * 10 BFP 0.2

Fold change of BFP of change Fold 0 0.0 0 1234 sh.Ren sh.p53 sh.Ren sh.p53 Time (months)

Figure 4. p53R172H induces aberrant self-renewal in vitro and in vivo. A, Total number of colony-forming units (CFU) generated by Trp53WT (black), Trp53Δ/Δ (red) and Trp53R172H/Δ (blue) cells. Data are representative of four independent experiments. Error bars correspond to mean ± SEM (n = 3). B, Total number of CFUs generated by Trp53R172H/Δ c-KIT+ cells containing sh.Ren or sh.p53. Data are representative of three independent experiments. Error bars correspond to mean ± SEM (n = 3). C, Western blotting for p53 and β-Actin at passage 4 (P4). D, Schematic representation of the competitive transplantation protocol. E, Percentage of CD45.2 cells during the secondary competitive transplant in the peripheral blood (PB) of mice monitored monthly by bleeding (n = 5 per group). F, Percentage of CD45.2 cells in the PB four months after transplant (n = 4–7). G, Percentage of CD45.2 cells in the BM 4 months after transplant (n = 4–7). H, Percentage of megakaryocyte erythrocyte progenitors (MEP), granulocyte monocyte progenitors (GMP; I) and common myeloid progenitors (CMP; J) in the BM of mice four months after transplant (n = 4–7). K, Schematic representation of the transplant layout. LSKs were sorted from Trp53R172H/Δ mice and were infected with an sh.Ren or sh.p53 conjugated to BFP fluorescence. Two days after infection equal numbers of cells were transplanted in lethally irradiated CD45.1 mice. L, BFP fluorescence over time as assessed by monthly bleeds in recipient mice n( = 5). M, Percentage of BFP+ cells in the BM of mice with sh.Ren or sh.p53 4 months after transplant (n = 3–5). N, Percentage of BFP+ LSKs in the BM of mice with sh.Ren or sh.p53 4 months after transplant (n = 3–5). B, C, and D, two-tailed t test, *, P < 0.05; **, P < 0.005; ***, P < 0.0005. Error bars correspond to mean ± SEM.

JULY 2019 CANCER DISCOVERY | 969

RESEARCH ARTICLE Loizou et al.

ACB FOXH1 expression 100 Enrichment plot 6 FOXH1 WT *** 90 * FOXH1_BOUND_GENES FOXH1 AMP or high mRNA 80 0.25 5.5 0.20 FDR = 0 70 NES = 1.38 0.15 60 0.10

ES 5 0.05 expression 50 0.00 2 −0.10 40 −0.15 4.5 30 Overall survival (%) Mean log 20

4 10 MUT NULL 0 BC MC

MM 0102030405060708090 100 110 120 HSC MPP MEP CMP PMN MNC GMP L-PM E-PM Time (months) AML-NK AML-CK AML t(8;21) AML t(15;17) Inv(16)/t(16;16) AML t(11q23)/MLL Normal cells AML

D Bone marrow EFBone marrow Murine AML G Human AML

Trp53 R172H/∆ Trp53 R172H/∆ Trp53 R172H/11B3 TP53 R172H/∆ /∆ ∆ R172H /∆ sh.Ren sh.p53 Trp53 sh.Rensh.p53 sh.Ren sh.p53 sh.Ren sh.TP53 Trp53 p53 p53 p53 p53 FOXH1 FOXH1 FOXH1 FOXH1 FOXH1

ACTB ACTB ACTB ACTB ACTB

Figure 5. Identification of FOXH1 as a p53-mutant mediator. A, GSEA comparing the expression of genes associated with genomic regions occupied by FOXH1 when compared with the genes differentially expressed in Trp53R172H and Trp53Δ/Δ leukemias. B, Microarray data showing the expression of FOXH1 across normal human hematopoietic cells and different types of leukemias. C, Overall survival in patients with AML who have FOXH1 mRNA upregulation or genetic amplification compared with patients without (cBioPortal –TCGA). D, Western blotting for FOXH1 in HSPCs from Trp53R172H and Trp53Δ/Δ cells at Passage 4. E, Western blotting for p53 and FOXH1 in Trp53R172H HSPCs transduced with sh.Ren or sh.p53. F, Western blotting for TP53 and FOXH1 in two independent Trp53R172H murine leukemic cells transduced with sh.Ren or sh.p53. G, Western blotting for p53 and FOXH1 in Trp53R175H human leukemic cells transduced with sh.Ren or sh.TP53. *, P < 0.05; ***, P < 0.0005. MPP, multipotential progenitors; CMP, common myeloid progenitor cell; GMP, granulocyte monocyte progenitors; MEP, megakaryocyte-erythroid progenitor cell; early_PM, early promyelocyte; late_PM, late promyelocyte; MY, myelocyte; MM, metamyelocytes; BC, band cell; PMN, polymorphonuclear cells; mono, monocytes. significantly higher in all AMLs (particularly in patients (ChIP-seq) to identify the genome-wide FOXH1 binding with CK-AML, the majority of whom harbor TP53 muta- sites in Trp53-mutant murine leukemic cells (Fig. 6A). Sites tions) compared with any type of normal cell (Fig. 5B; bound by FOXH1 correlated with the presence of canonical Supplementary Fig. S5A and S5B). In addition, like mutant FOXH1 binding motifs (P < 1e-1915) as well as with other TP53, FOXH1 upregulation correlated with poor survival in TGFβ mediators (refs. 48, 49; Fig. 6B). GO pathway analy- patients with AML (refs. 46, 47; Fig. 5C). Consistent with sis indicated that genes associated with FOXH1 peaks are the RNA-seq results, FOXH1 expression was also higher in involved in BMP and TGFβ signaling pathways, as well as Trp53R172H/Δ HSPCs compared with Trp53Δ/Δ cells growing in myeloid leukemia (Fig. 6C). colonies (Fig. 5D), and knockdown of mutant p53 reduced To investigate how many of the genes bound by FOXH1 are FOXH1 levels in murine HSPCs (Fig. 5E) as well as murine dynamically regulated, we suppressed FOXH1 and compared and human leukemic cells (Fig. 5F and G, respectively). the downregulated genes (green circle) to those physically Thus, mutant p53 expression led to higher levels of FOXH1 bound by FOXH1 (orange circle; Fig. 6D). Consistent with in both normal and leukemic cells from mouse and human a molecular relationship between mutant p53 and FOXH1 leukemias. expression, there was a significant overlap between the genes that were decreased upon knockdown of mutant p53 (blue R172H FOXH1 Is a Critical Effector of p53 via Its circle) or FOXH1 (green circle), with more than 50% of genes Ability to Bind and Regulate Stem Cell Genes showing evidence of coregulation (Fisher exact test P < 2.2e- Because FOXH1 is a transcription factor and its role in 16). GSEA confirmed that genes that were downregulated cells of hematopoietic origin is not known, we performed upon FOXH1 knockdown significantly overlapped with chromatin immunoprecipitation (ChIP) with an anti- those downregulated upon mutant p53 knockdown (NES = FOXH1 antibody followed by high-throughput sequencing −14.5, FDR = 0; Fig. 6E). In line with these results, many of

970 | CANCER DISCOVERY JULY 2019 www.aacrjournals.org

AB C Motif enriched in FOXH1 Annotation of all leukemia FOXH1 sites Best match TF P-value bound sites BMP signaling pathway FOXH1 <1e−1915 Developmental biology Intergenic 57.7% Signaling by SCF−KIT KLF10 <1e−1810 Sema4D in semaphorin signaling TGFβ receptor complex NRF1 <1e−1728 Exon 0.3% Pluripotency of stem cells 5 _UTR 0% PB0130.1_Gm397 <1e−933 Acute myeloid leukemia ′ Chronic myeloid leukemia 3′_UTR 0.3% RUNX2 <1e−790 Serine/threonine kinase activity TSS 1% Growth hormone signaling pathway PB0170.1_SOX17 <1e−72 Promoter−TSS 1% p38 MAPK signaling pathway Noncoding 0.1% EKLF(Zf)/Klf1 <1e−61 MAPK signaling pathway TFAP2A <1e−37 0102030 Intron 39.7% TEAD1 <1e−3 Pathway GO_Biological_Process_2017b GO_Molecular_Function_2017b KEGG_2016 Reactome_2016 BioCarta_2016 sh.p53_DN sh.FOXH1_DN D E sh.p53 downregulated genes HSC Passegue LSC signature_Somerveille 153 608 139 FDR = 0.0 FDR = 0.0 FDR = 0.0 91 NES = −14.5 NES = −6.58 NES = −5.48 61 187

5599 sh.FOXH1 sh.Ren sh.FOXH1 sh.Ren sh.FOXH1 sh.Ren FOXH1 ChIP F 9.2 9.5 Input Input 9.2 9.5 FOXH1 ChIP FOXH1 ChIP

0.13 0.4 sh.Ren sh.Ren 0.13 0.4 sh.Ren sh.Ren 0.13 0.4 sh.FOXH1 sh.FOXH1 0.13 0.4 sh.FOXH1 RNA-seq sh.FOXH1 0.13 0.4 RNA-seq sh.p53 sh.p53 0.13 sh.p53 sh.p53

Mef2c

G H Trp53 R172H_upregulated genes FOXH1_regulated genes

**** **** **** **** * **** 5.2 **** 6.0 *

5.0

expression 5.5 2

expression 4.8 2 Mean log 4.6 Mean log 5.0

4.6 BC BC MY MY MM MM HSC HSC MEP MEP MPP MPP PMN PMN CMP CMP GMP GMP L-PM L-PM E-PM E-PM Mono Mono AML-CK AML-CK AML t(8;21) AML t(8;21) AML t(15;17) AML t(15;17) Inv(16)/t(16;16) Inv(16)/t(16;16) AML t(11q23)/MLL AML t(11q23)/MLL

Figure 6. FOXH1 directly binds genes involved in leukemia and stem cell properties. A, Pie charts depicting the genome-wide distribution of FOXH1 binding in murine leukemic cells. TSS, transcription start site. B, Motif enrichment analysis for the FOXH1-bound genomic regions in leukemic cells. C, GO analysis of genes bound by FOXH1 in murine leukemic cells. D, Venn diagram depicting the overlap between the genes downregulated upon p53 KD (blue circle), FOXH1 KD (green circle), and FOXH1 ChIP (orange circle). E, GSEA from transcriptional profile of cells upon FOXH1 KD. F, Gene tracks for FOXH1 ChIP-seq at the Runx2 and Mef2c loci together with RNA-seq tracks from either sh.Ren (red), sh.FOXH1 (green), or sh.p53 (blue). Mean log2 expression of transcriptional signature composed of significantly upregulated genes inTrp53 R172H/Δ versus Trp53Δ/Δ murine leukemias across AML and normal cells (G), as well as same comparison for genes downregulated following FOXH1 KD (H). Student t test between CK-AML and other AML types, *, P < 0.05; ****, P < 0.00005.

JULY 2019 CANCER DISCOVERY | 971

RESEARCH ARTICLE Loizou et al. the transcriptional signatures related to HSCs and leukemia to overexpress either a Foxh1 cDNA or control vector and (previously shown to be upregulated in Trp53R172H/Δ cells) examined for self-renewal capacity in serial replating assays. were conversely enriched upon suppression of either FOXH1 Enforced FOXH1 expression partially reversed the loss of (Fig. 6E) or p53 (Supplementary Fig. S6A). As examples, replating potential produced by suppressing mutant p53 two genes involved in leukemia and hematopoietic cell dif- (Fig. 7E), suggesting that FOXH1 functions downstream of ferentiation, Runx2 (50) and Mef2c (51), are both bound by mutant p53 to enforce aberrant self-renewal. Collectively, FOXH1 and their expression is reduced by either FOXH1 or these data demonstrate that FOXH1 is necessary and suffi- p53 knockdown (Fig. 6F). Hence, our data show that mutant cient to support an aberrant self-renewal phenotype in both p53 induces FOXH1, which can directly bind to and control premalignant and malignant p53-mutant cells. genes linked to the differentiation status and self-renewal properties of AML. To confirm that these FOXH1-dependent transcripts were DISCUSSION also observed in human AML, we derived a transcriptional Although mutations in the TP53 tumor suppressor gene signature composed of differentially upregulated genes in ablate its antiproliferative functions, some mutant TP53 Trp53R172H/Δ versus Trp53Δ/Δ murine leukemias and compared alleles also appear to produce “GOF” phenotypes that display it with expression data from different karyotypes of human more profound phenotypes than those produced by complete AML as well as normal hematopoietic cells (32). We observed a p53 loss. Nonetheless, studies investigating the basis for these significant upregulation of theTrp53 R172H/Δ murine leukemia– effects have failed to coalesce around a common biological derived transcriptional signature in patients with CK-AML or molecular mechanism and, indeed, a recent comprehen- (dark red) compared with all other leukemia types (light red) sive analysis of DNA sequence and functional genomics data or normal hematopoietic cells (black; Fig. 6G), suggesting our suggested that the primary selective pressure for the appear- model transcriptionally resembles CK-AML harboring TP53 ance of p53 mutation in cancer arises from their purely mutations. A similar increase in FOXH1-dependent transcripts inactivating and/or dominant-negative effects (52). Here we (sh.FOXH1 vs. sh.Ren) was also observed in patients with studied the action of the most common TP53-mutant allele CK-AML (Fig. 6H). These results reveal a FOXH1 signature in in the initiation and maintenance of AML, a well-character- human AML harboring TP53 mutations, further supporting ized liquid tumor where p53 loss is linked to aggressive and the existence of a functional mutant p53–FOXH1 axis in this chemoresistant disease and one in which the potential GOF disease. effects of mutant p53 have not been previously examined. We show that this p53 mutant accelerates leukemia onset FOXH1 Is Necessary and Sufficient for the and, strikingly, is required for leukemia maintenance, Enhanced Self-Renewal Phenotype Produced such that suppression of mutant p53 eliminates leukemia- by Mutant p53 initiating potential. To further explore the links between mutant p53, FOXH1, The biological basis of AML involves the emergence of and aberrant self-renewal, we knocked down mutant p53 or leukemic blasts that have acquired mutations that drive FOXH1 in Trp53R172H/Δ HSPCs and performed colony-form- increased proliferation of stem and progenitor cells, block ing unit (CFU) and immunophenotyping assays. Knockdown their maturation, and promote their aberrant self-renewal of FOXH1 ablated the replating capacity of these cells (Fig. 7A), (53, 54). In our models, the murine Trp53R172H mutant con- phenocopying, in large part, the effects observed upon tributes to the latter activity, such that its elimination in leu- mutant p53 knockdown. Furthermore, similar to mutant kemic cells releases the differentiation block. Similar results p53 ablation, FOXH1 suppression leads to the downregula- are observed with human AML cells harboring the equivalent tion of the stem cell factor receptor c-KIT and the stem cell R175H allele and likely explain the apparent requirement antigen 1 (Sca-1) in Trp53R172H/Δ HSPCs (but not in p53- of certain other leukemia cells to additional mutant TP53 deficient cells), which was functionally accompanied by alleles (e.g., R248Q; see Supplementary Fig. S1B; ref. 28). coarsening of chromatin and nuclear segmentation, mor- Taking advantage of well-validated assays of self-renewal phologic changes indicative of differentiation (Fig. 7B; Sup- in the hematopoietic system, we show that p53R172H acts in plementary Fig. S7A and S7B). Conversely, overexpression of nontransformed hematopoietic cells to promote self-renewal FOXH1 in both p53WT/WT and p53Δ/Δ HSPCs endowed these to a substantially greater degree than p53 loss, and that sup- cells with increased colony-forming ability (Fig. 7C) and pression of mutant p53 reverses these effects. Thus, in AML, also led to the upregulation of c-KIT and Sca-1 along with mutant p53 can produce phenotypes analogous to bona concomitant downregulation of Gr1/CD11b (Fig. 7D; Sup- fide oncogenes, such as the MLL–AF9 fusion oncoprotein, plementary Fig. S7C). Analogous results were also observed which also contributes to leukemia by enforcing aberrant in leukemia cells, where suppression of either mutant p53 self-renewal and exerts a continuous dependency for disease or FOXH1 in two Trp53-mutant AML lines reduced c-KIT maintenance (55–57). expression (Supplementary Fig. S8A–S8D) and led to cellular Although most TP53 mutations disrupt its DNA-binding differentiation determined by morphologic criteria (Sup- capabilities, the resulting proteins are frequently stable and plementary Fig. S8E and S8F) and transcriptional profiling retain an intact transcriptional activation domain (4, 30). (Supplementary Fig. S8G). Finally, to determine whether Some studies suggest that mutant p53 can be recruited to FOXH1 could phenocopy mutant p53 effects, we performed chromatin to induce gene expression via interaction with an epistasis experiment using Trp53R172H/Δ HSPCs where other transcriptional regulators, such as NF-Y and ETS2 mutant p53 was suppressed and cells were cotransduced (30, 58–60). Other studies support a model where mutant

972 | CANCER DISCOVERY JULY 2019 www.aacrjournals.org

A B sh.Ren sh.p53 sh.FOXH1 R172H/∆ 25 Trp53 (P3) 77.4 86.5 73.8 Trp53 WT 20

10104 10 75.9 87.2 94.2

CFU (average) 5 Trp53 ∆ /∆ 0 FOXH1

85.60.00 25.0 ACTB

Trp53 R172H/∆

sh.p53 sh.Ren

sh.FOXH1_1sh.FOXH1_2 c-KIT Sca-1

C WT ∆/∆ D 15 Trp53 Trp53 Ctrl 80 **** Ctrl Foxh1 Foxh1 10 60 **** (%) +

5 40 CFU (average) Sca-1

20 0 c-KIT FOXH1 0 ACTB Trp53 WT Trp53 ∆ /∆

E Trp53 R172H/∆ 40 sh.Ren + pLVX control n.s. sh.Ren + pLVX Foxh1 * 30 ** n.s. sh.p53 + pLVX control *

20 sh.p53 + pLVX Foxh1 CFU (average) 10

Figure 7. FOXH1 is necessary and sufficient for the enhanced self-renewal phenotype produced by mutant p53. A, Total number of CFUs generated by Trp53R172H/Δ HSPCs containing sh.p53 (blue), sh.Ren (black), shFOXH1.1072 (1; plum), shFOXH1.1116 (2; maroon). Data are representative of three experiments. Error bars correspond to mean ± SEM (n = 3). Bottom, Western blotting intensities for FOXH1 for the different conditions. B, c-KIT+ and Sca-1+ expression analyzed by FACS in sh.Ren, sh.p53 and sh.FOXH1 containing HSPCs from Trp53WT, Trp53Δ/Δ, and Trp53R172H/Δ mice. Data are representative of three experiments. C, Total number of CFUs generated by Trp53Δ/Δ and Trp53WT HSPCs containing pLVX Control (black) or pLVX Foxh1 cDNA (gray). Data are representative of three experiments. Error bars correspond to mean ± SEM (n = 3). Bottom, Western blotting for FOXH1 for the different conditions. D, Percentage of c-KIT+Sca-1+ HSPCs analyzed by FACS in Trp53Δ/Δ and Trp53WT cells containing pLVX Control or pLVX Foxh1 cDNA. Data are representative of three experiments. Error bars correspond to mean ± SEM (n = 3). E, Total number of CFUs generated by Trp53R172H/Δ HSPCs containing sh.Ren + pLVX control (black), sh.Ren + pLVX Foxh1 cDNA (gray), sh.p53 + pLVX control (light blue), sh.p53 + pLVX Foxh1 cDNA (dark blue). Data are representative of three experiments. Error bars correspond to mean ± SEM (n = 3; *, P < 0.05; **, P < 0.005; ***, P < 0.0005).

JULY 2019 CANCER DISCOVERY | 973

RESEARCH ARTICLE Loizou et al. p53 proteins can directly bind to and perturb the func- TP53 mutations are often observed in clonal hematopoiesis, a tion of other transcription factors, including p63, p73, and phenomenon where stem cells harboring leukemia-associated SREBP (13, 14, 58–60). In the AML system studied herein, mutations expand in the peripheral blood of individuals in mutant p53 increases the expression of the FOXH1 tran- more than 0.02 VAF without overt disease (65–67). We identi- scription factor, leading to its significant upregulation com- fied 9 cases of clonal hematopoiesis that harboredTP53 R175H pared with Trp53Δ/Δ leukemias. Functional studies revealed mutations with a high VAF that ranged from 2% to 9%. Strik- that p53R172H is necessary and sufficient to induce FOXH1 ingly, 3 of 9 cases (∼33%) harbored only a single mutation, which, in turn, binds stem cell–associated genes and con- TP53 R175H, strongly suggesting that this is an early event. tributes to enhanced self-renewal. The precise details of how Of the remaining 6 cases, 2 individuals harbored clonal mutant p53 induces FOXH1 remain to be determined and, TP53R175H mutations whereas the remaining 4 contained to date, we have been unable to detect a robust ChIP signal subclonal mutations. Furthermore, individuals presenting for mutant p53 on DNA or validate an interaction with with clonal hematopoiesis involving a TP53 mutation have other p53 family members. Nevertheless, the importance of the highest risk for progressing to AML (66). Taken together, this molecular axis is supported by the conserved ability of these observations strongly suggest that TP53R175H mutation mutant p53 to increase FOXH1 expression across species is an early, “first-hit” event in the pathogenesis of human and by the significant upregulation of FOXH1 expression AML. in CK-AML, a leukemia subtype characterized by a high Our data predict that some cancers might depend on con- frequency of TP53 mutations. tinuous expression of mutant TP53 and are consistent with FOXH1 is expressed in embryonic stem cells and during studies on the p53R248Q-mutant protein in mouse models of early embryonic development. There, it is induced by the lymphoma, colon cancer, and others (68, 69). Nevertheless, it TGFβ family member Nodal (61) and acts to promote an seems unlikely that all mutant TP53 alleles have GOF activity, EMT that takes place during gastrulation (43, 62). FOXH1 and even those that do may impinge on distinct molecular is not normally expressed in the adult hematopoietic system mechanisms depending on context. For example, although and, as such, mutant p53 appears to hijack FOXH1-driven p53R172H can display GOF effects by promoting invasion and programs to promote leukemia by enforcing the expression metastasis in a pancreatic cancer model, it is not required of genes normally involved in promoting stemness and cell for the sustained proliferation of pancreatic cancer cells, at plasticity during hematopoiesis. Like the Trp53R172H allele, least in vitro (13). Moreover, a recent report analyzing broad p53 deficiency, as produced by two homozygous nullTrp53 functional genomic datasets concluded that mutant TP53 alleles, can also accelerate AML development by promot- was generally dispensable for the proliferation of human ing aberrant self-renewal and enhancing leukemia-initiating cell lines in culture (52). However, this study assessed only potential (19). However, the aberrant program produced the effects of gene disruption on proliferation under in vitro by the p53-null state is not as potent as that produced by conditions and did not examine signals associated with allele p53R172H and does not depend on elevated levels of FOXH1. specificity or cellular context that are highly prevalent in the Presumably, this distinct rewiring of self-renewal programs biology of p53 (8), whereas studies in vivo have confirmed a explains why cells differentiate and lose leukemia-initiating greater advantage of some of the hotspot mutations (70). capacity upon mutant p53 elimination to produce a p53- Indeed, our functional experiments show a requirement for null state, the latter event being known to promote leuke- p53R175H in the leukemia-initiating potential of a human mogenesis if present as an initiating event. Thus, in this AML line, which could have been missed by performing a dif- context, mutant p53 GOF produces a more profound phe- ferent type of analysis. In addition, upon detailed inspection notype than p53 loss while concomitantly creating unique of publicly available CRISPR screening data, we note that molecular dependencies. some TP53-mutant lines appear to depend on mutant TP53 Our results may also have relevance to the pathogenesis of for their survival (Supplementary Fig. S1B; ref. 30). Dissect- human AML. Specifically, beyond the GOF effect of p53R172H ing the underlying mechanism of this context specificity will in leukemogenesis, this work reveals that mutant p53 can be important to determine when and where mutant p53 pro- exert its phenotypes even prior to leukemia onset, leading teins and their associated molecular programs may represent to the continuous self-renewal of HSPCs in vitro and in vivo. therapeutic targets. Most previous GOF studies of mutant p53 have been in the Still, despite the perplexing range of molecular and bio- context of established cancer and, to our knowledge, this logical mechanisms that have been attributed to GOF TP53 is the first study to reveal that a GOF activity can produce mutations, our studies reveal convergent features between molecular dependencies prior to neoplastic transformation. mutant p53 GOF action in solid and liquid tumors, as well as The effect is such that elimination of the mutant protein in in experimental models of induced pluripotency (44). In solid premalignant cells promotes differentiation, impairs colony- tumors, the best characterized phenotype associated with forming ability, and reduces competitive fitness during long- TP53 GOF mutations involves their ability to drive invasion term hematopoiesis in vivo. Although TP53 mutations are and metastasis (12, 13, 69, 71, 72), which is often associated considered a later event during the progression of many solid with an EMT and cellular plasticity that enables the acquisi- tumors (63), they are present as “truncal” mutations with a tion of a more stem cell–like state. Consistent with this, the high variant allele frequency in human AML such that we transcriptional profiles of p53-mutant tumors are enriched identified 35 AML cases with the R175H mutation and a for gene signatures expressed in embryonic stem cells (73–75). median variant allele frequency (VAF) of 36% that showed In AML, Trp53R172H promotes self-renewal, as well as a tran- evidence of clonal or subclonal presentation (64). Indeed, scriptional landscape enriched for gene signatures linked to

974 | CANCER DISCOVERY JULY 2019 www.aacrjournals.org

Downloaded from cancerdiscovery.aacrjournals.org on September 27, 2021. © 2019 American Association for Cancer Research. Published OnlineFirst May 8, 2019; DOI: 10.1158/2159-8290.CD-18-1391 p53R172H has Gain-of-Function Effects in AML RESEARCH ARTICLE stemness and, intriguingly, EMT. One such gene is Foxh1, Histology and Microscopy which is known to promote EMT during embryonic develop- Tissues were dissected from mice for fixation overnight in 10% ment, yet we demonstrate here that it can drive aberrant self- formalin (Thermo Fisher Scientific). Bones were decalcified as per the renewal in AML. Remarkably, these same players contribute to manufacturer’s instructions (IDEXX BioResearch), and fixed tissues the generation of iPS cells by the Yamanaka factors: Trp53R172H were dehydrated and embedded in paraffin for sectioning. Paraffin enhances iPS cell formation beyond p53 loss (16), and FOXH1 sections (5 μm) were prepared and stained with hematoxylin and cooperates with p53 deficiency to achieve a similar effect (44). eosin (Leica Autostainer XL). Collectively, these observations point to a convergent action of p53 GOF on cellular plasticity and stemness, which can influ- Flow Cytometry ence distinct cancer processes depending on context. Single-cell suspensions were prepared from BM, spleen, or peripheral blood. Red blood cells were lysed with ammonium–chloride– potassium (ACK) buffer, and the remaining cells were resuspended in METHODS PBS with 3% FBS. Nonspecific antibody binding was blocked by incubation with 20 μg/mL rat IgG (Sigma-Aldrich) for 15 minutes, and cells Animal Models were then incubated with the indicated primary antibodies for 30 min- All mouse experiments were approved by the Institutional Animal utes on ice. Stained cells were quantified using a Fortessa analyzer (BD Care and Use Committee at Memorial Sloan Kettering Cancer Center Biosciences) or isolated with a FACSAria II (BD Biosciences). FlowJo (MSKCC). Trp53LSLR172H and Trp53LoxP mice were described previously software (TreeStar) was used to generate flow cytometry plots. More (12). C57BL/6 (B6.SJL-Ptprca Pepcb/BoyJ (CD45.1) mice, thymecto- details on FACS antibodies can be found in Supplementary Table S2. mized mice, and Vav1-Cre and Mx1-Cre mice were purchased from Jackson Laboratory and bred in a mouse facility at MSKCC. BM Competitive Transplantation F/F F/F Primary Cell Culture Femurs and tibiae were isolated from donor Trp53 (WT), Trp53 ; Mx1-Cre or Trp53R172H/F;Mx1-Cre (CD45.2), as well as WT (CD45.1) R172H/F F/F Total BM cells from Trp53 ;Vav1Cre, Trp53 ;Vav1Cre and mice. Bones were flushed with PBS and the single-cell suspension Trp53F/F (WT) mice (6–8 weeks) or purified BM c-KIT+ cells were was centrifuged (5 minutes, 0.5 × g, 4°C) and treated with red cell transduced as described below with sh.Ren or sh.Trp53 retrovirus ACK lysis buffer, as described above. Total nucleated BM cells were and plated in methylcellulose medium (Methocult M3434, Stem- resuspended in PBS, passed through a 40-μm cell strainer, and Cell Technologies). Cells were seeded at 10,000 total BM cells or counted. Donor cells (1 106 per genotype, per mouse) were mixed + × 4,000 purified c-KIT cells per replicate. Colony-forming units were 1:1 with support BM cells (CD45.1) and transplanted via retro-orbital enumerated using a Zeiss Axio Observer microscope, and cells were injection into lethally irradiated (950 Rad) CD45.1 recipient mice. replated (4,000 cells/replicate) every 7 to 10 days. Primary mouse Chimerism was monitored by flow cytometry (anti-CD45.1 and anti- R172H/F F/F PMN leukemias derived from Trp53 ;Vav1-Cre and Trp53 ; CD45.2; BD Biosciences) of peripheral blood at 4-week intervals post- Vav1-Cre mice were cultured in 1:1 Dulbecco’s Modified Eagle transplant three weeks after last pIpC injection and for 16 weeks, at Medium (DMEM) to Iscove’s Modified Dulbecco’s Medium with which time mice were sacrificed and chimerism was assessed in other 10% FBS (Gibco), 100 U/mL penicillin, 100 μg/mL streptomycin, hematopoietic compartments (BM). 50 μmol/L 2-mercaptoethanol, and recombinant mouse stem cell factor (SCF; 50 ng/mL), IL3 (10 ng/mL), and IL6 (10 ng/mL) at 37°C Retroviral Transduction in 5% CO2. Puromycin and doxycycline were used at a final concen- − + + tration of 1 μg/mL for all cell culture assays. HSPCs were isolated by sorting Lin c-KIT Sca1 cells or enriched using an autoMACS Pro separator (Miltenyi Biotec) using CD117 Cell Lines Magnetic Microbeads (Miltenyi Biotec, 130-091-224), then cultured in the presence of 50 ng/mL SCF, 10 ng/mL IL3, and 10 ng/mL The human leukemia line KY821 (JCRB cell bank) was grown IL6, and infected with concentrated supernatants of LMNe-sh.Ren/ in RPMI-1640 and 20% FBS (Gibco). KL922 ( Δ/Δ , KL973 Trp53 ) sh.p53-BFP or MLPe sh.Ren/sh.p53/sh.FoxH1 retrovirus after 24 and R172H/11B3 R172H/Δ (Trp53 ), and KL974 (Trp53 ) primary mouse cell lines 48 hours (more details about the shRNAs in Supplementary (19, 25) were cultured with RPMI, 10% FBS (Gibco), and recombi- Table S2). Transduction efficiency was determined by reporter fluo- nant mouse SCF (50 ng/mL), IL3 (10 ng/mL), and IL6 (10 ng/mL). rescence at 48 hours, and 5,000 sorted cells were injected into CD45.1 PlatinumE (76) cells were a gift from Craig Thomson’s laboratory mice as described above. and HEK293T were purchased from ATCC (CRL-3216) and were used for retroviral production while maintained in DMEM with 10% FBS. Media supplemented with 100 U/mL penicillin, 100 μg/mL Cell Viability and DNA Synthesis streptomycin, and cultures were incubated at 37°C, 5% CO2. All cells Apoptosis assays were performed by staining with Annexin V (BD tested negative for Mycoplasma (MP0025-1KT) and were not main- Pharmingen) according to the manufacturer’s instructions, in combi- tained for longer than 10 passages in culture. For cell line authenti- nation with DAPI. For cell-cycle studies, BrdU (BD Pharmingen) was cation, we confirmed p53 expression by Western blot analysis, and used per manufacturer’s instructions and stained with 2 μg/mL DAPI Sanger sequencing confirmed the specific mutation. prior to flow cytometric analysis.

Xenotransplantation Quantitative RNA Expression Assays Experiments were carried out in accordance with institutional guide- Total RNA was extracted from cells using the RNeasy Plus Mini lines approved by MSKCC. Immunodeficient NSG mice (6–10 weeks Kit (Qiagen). RNA quantity and quality were determined using an old) were bred under pathogen-free conditions and were sublethally Agilent 2100 Bioanalyzer. RNA-seq libraries were prepared from total irradiated (200–250 cGy) 3 to 24 hours before transplantation. KY821 RNA by polyA selection using oligo-dT beads (Life Technologies) cells were injected by tail-vein injection. Complete blood count analysis according to the manufacturer’s instructions. The resulting poly-A+ was performed on peripheral blood collected from submandibular RNA served as input for library construction using standard Illumina bleeding using a Drew Hemavet Analyzer (Veterinary Diagnostics). protocols. Sequencing was performed on an Illumina HiSeq 2500

JULY 2019 CANCER DISCOVERY | 975

RESEARCH ARTICLE Loizou et al. sequencer using 50 bp pair-end reads. For mRNA quantification, Immunoblotting total RNA was used for cDNA synthesis (Agilent). Real-time PCR For protein lysates, cells were incubated with RIPA buffer sup- reactions were carried out using SYBR Green Master Mix (Quanta- plemented with protease inhibitors (Protease inhibitor tablets, Bio) and a Viia7 (Applied Biosystems). Roche) for 30 minutes and cleared by centrifugation (15 minutes 14.000 rpms 4C). Protein was quantified using the DC protein assay RNA-seq Analysis (Bio-Rad). The following antibodies were used for immunoblot- RNA-seq library construction and sequencing were performed at ting: β-Actin (ac-15, Sigma), mp53 (CM5, Leica Microsystems), hp53 the integrated genomics operation (IGO) Core at MSKCC. Poly-A (DO-1, Santa Cruz Biotechnology), mFOXH1 (Abcam 49133), and selection was performed. For sequencing, approximately 30 million hFOXH1 (Abcam 102590). 50-bp paired-end reads were acquired per replicate condition. Result- ing RNA-seq data were analyzed by removing adaptor sequences using ChIP Trimmomatic (77). RNA-seq reads were then aligned to GRCh37.75 ChIP was performed as described previously (82). Briefly, KL974 (hg19) with STAR (78), and genome-wide transcript counting was (Trp53R172H/Δ) cells were fixed with 1% formaldehyde for 15 minutes, performed by HTSeq (79) to generate a matrix of fragments per and the cross-linking reaction was stopped by adding 125 mmol/L kilobase of exon per million fragments mapped. Gene expressions of glycine. Cells were washed twice with cold PBS and lysed in buffer RNA-seq data were managed by hierarchical clustering based on one (150 mmol/L NaCl, 1%v/v Nonidet P-40, 0.5% w/v deoxycholate, 0.1% minus Pearson correlation test using Morpheus (https://software. w/v SDS, 50 mmol/L Tris pH 8, 5 mmol/L EDTA) supplemented broadinstitute.org/morpheus/). All results from the RNA-seq experi- with protease inhibitors. Cell lysates were sonicated using a Covaris ments can be found under GSE125097. E220 Sonicator to generate fragments less than 400 bp. Sonicated lysates were centrifuged and incubated overnight at 4°C with specific GSEA antibodies (FOXH1 Abcam 49133). Immunocomplexes were recov- GSEA was performed using gene set as permutation type, 1,000 ered by incubation with 30 μL protein A/G magnetic beads (Thermo permutations, and log2 ratio of classes, or with gene set and Signal- Fisher Scientific) for 2 hours at °4 C. Beads were sequentially washed 2Noise as metrics for ranking genes. Gene sets used in this study twice with RIPA buffer, increasing stringency ChIP wash buffers were identified from the Molecular Signatures Database (MSigDB (150 mmol/L NaCl, 250 mmol/L NaCl, 250 mmol/L LiCl), and finally Curated v4.0). Gene pathways and functions were assessed using TE buffer. Immunocomplexes were eluted using elution buffer (1%

Ingenuity Pathway Analysis (Qiagen Bioinformatics). SDS, 100 mmol/L NaHCO3), and cross-linking was reverted by the addition of 300 mmol/L NaCl and incubation at 65°C overnight. Expression of Gene Signatures in Human AML DNA was purified using a PCR Purification Kit (Qiagen). Input chro- and Normal Cells matin was used for estimation of relative enrichment. The results for the ChIP-seq experiments can be found under GSE125097. Visualization of gene signatures were derived from transcriptional data as log2 expression values for different karyotypes of patients with AML described previously in GSE42519 (80). For normal hemat- ChIP-seq Library Preparation, Illumina Data Analysis, opoietic cells, values were averaged over all genes in each signature and Peak Detection and visualized by using Sinaplot. Student t test was performed using ChIP-seq libraries were prepared at the Center for Epigenetic the R programming language (https://www.R-project.org/). Research (MSKCC) using the NEBNext ChIP-Seq Library Prep Master Mix Set for Illumina (New England BioLabs) following the manu- Quantification and Statistical Analysis facturer’s instructions. Reads were trimmed for quality and Illumina adapter sequences using “trim_galore” before aligning to mouse All P values were calculated using unpaired two-tailed Student assembly mm9 with bowtie2 using the default parameters. Aligned t test with GraphPad Prism software, unless otherwise described in reads with the same start site and orientation were removed using the methods or figure legends. No specific randomization or blinding the Picard tool MarkDuplicates (http://broadinstitute.github.io/ protocol was used for these analyses. Statistically significant differ- picard/). Density profiles were created by extending each read to the ences are indicated with asterisks in figures with the accompanying average library fragment size and then computing density using the P values in the legend. Error bars in figures indicate SD or SEM for BEDTools suite (http://bedtools.readthedocs.io). Enriched regions the number of replicates, as indicated in the figure legend. were discovered using MACS2 and scored against matched input libraries (fold change > 2 and P < 0.005). Dynamic regions between Plasmids and Viral Production two conditions were discovered using a similar method, with the All vectors were derived from the murine stem cell virus (MSCV; second ChIP library replacing input. Peaks were then filtered against Clontech) retroviral vector backbone. miRE-based shRNAs were genomic “blacklisted” regions (http://mitra.stanford.edu/kundaje/ designed and cloned as described previously (27) into the LT3GEPIR akundaje/release/blacklists/mm9-mouse/mm9blacklist.bed.gz) and (TRE3G-GFP-miRE-PGK-PuroR-IRES-rtTA3) vector (81). The those within 500 bp were merged. All genome browser tracks and mouse Foxh1 cDNA was cloned in a pVXL-hygromycin-mFoxh1 vec- read density tables were normalized to a sequencing depth of ten tor (provided by Yilong Zou, Massagué lab, MSKCC). All constructs million mapped reads. Dynamic peaks were annotated using linear were verified by sequencing. Lentiviruses were produced by cotrans- genomic distance, and motif signatures were obtained using the fection of 293T cells with 10 μg LT3GEPIR construct and helper “de novo” approach with Homer v4.5 (http://homer.ucsd.edu/). vectors (3.75 μg psPAX2 and 1.25 μgVSV-G). For retroviral infection, PlatinumE cells were plated in 15-cm diameter dishes each Disclosure of Potential Conflicts of Interest transfected with 54 μg of MSCV vectors and 2.7 μg of VSV-G and 8.1 μg of pCl-eco. Transfection of packaging cells was performed using S.W. Lowe has ownership interest (including stock, patents, etc.) in PMV Pharmaceuticals. No potential conflicts of interest were dis- HEPES-buffered saline and CaCl2 or Mirus transfection reagent (MIR 2304). Viral supernatants were passed through a 0.45-μm filter (Mil- closed by the other authors. lipore) and concentrated with Centrifugal Filter units (Millipore) to obtain high virus titer. Concentrated virus was supplemented with 8 Authors’ Contributions μg/mL of polybrene (Sigma) before adding to target cells. Conception and design: E. Loizou, S.W. Lowe

976 | CANCER DISCOVERY JULY 2019 www.aacrjournals.org

Acquisition of data (provided animals, acquired and managed 11. Muller PA, Vousden KH. Mutant p53 in cancer: new functions and patients, provided facilities, etc.): E. Loizou, A. Banito, A. Mayle therapeutic opportunities. Cancer Cell 2014;25:304–17. Analysis and interpretation of data (e.g., statistical analysis, 12. Olive KP, Tuveson DA, Ruhe ZC, Yin B, Willis NA, Bronson RT, et al. biostatistics, computational analysis): E. Loizou, A. Banito, Mutant p53 gain of function in two mouse models of Li-Fraumeni G. Livshits, Y.-J. Ho, R.P. Koche, C.-C. Chen, S. Kinalis, F.O. Bagger, syndrome. Cell 2004;119:847–60. E.R. Kastenhuber, B.H. Durham 13. Weissmueller S, Manchado E, Saborowski M, Morris JPt, Wagenblast Writing, review, and/or revision of the manuscript: E. Loizou, E, Davis CA, et al. Mutant p53 drives pancreatic cancer metastasis A. Banito, G. Livshits, F.J. Sanchez-Rivera, A. Mayle, E.R. Kastenhuber, through cell-autonomous PDGF receptor beta signaling. Cell 2014; 157:382–94. B.H. Durham, S.W. Lowe 14. Freed-Pastor WA, Mizuno H, Zhao X, Langerod A, Moon SH, Administrative, technical, or material support (i.e., reporting or Rodriguez-Barrueco R, et al. Mutant p53 disrupts mammary tissue organizing data, constructing databases): Y.-J. Ho architecture via the mevalonate pathway. Cell 2012;148:244–58. Study supervision: S.W. Lowe 15. Sarig R, Rivlin N, Brosh R, Bornstein C, Kamer I, Ezra O, et al. Mutant Acknowledgments p53 facilitates somatic cell reprogramming and augments the malignant potential of reprogrammed cells. J Exp Med 2010;207:2127–40. The authors are grateful to Charles J. Sherr for extensive scientific 16. Yi L, Lu C, Hu W, Sun Y, Levine AJ. Multiple roles of p53-related path- discussion, suggestions for experiments, and help in formulating ways in somatic cell reprogramming and stem cell differentiation. and editing the manuscript. We thank Omar Abdel-Wahab, Ross L. Cancer Res 2012;72:5635–45. Levine, and Scott Armstrong for their key input and reagents, and 17. Rucker FG, Schlenk RF, Bullinger L, Kayser S, Teleanu V, Kett H, et al. Elli Papaemmanuil for providing analysis of human AML datasets. TP53 alterations in acute myeloid leukemia with complex karyotype We also thank members of the Lowe laboratory for comments and correlate with specific copy number alterations, monosomal karyo- discussions, Yilong Zou and Qiong Wang for providing us with the type, and dismal outcome. Blood 2012;119:2114–21. mFOXH1-expressing vector, and Jiajun Zhu for helping to optimize 18. Papaemmanuil E, Gerstung M, Bullinger L, Gaidzik VI, Paschka P, ChIP-seq conditions. This work was supported by a grant from the Roberts ND, et al. Genomic classification and prognosis in acute NCI (R01 CA190261) and the Memorial Sloan Kettering Cancer myeloid leukemia. N Engl J Med 2016;374:2209–21. Center Support Grant (P30 CA008748). E. Loizou was supported 19. Zhao Z, Zuber J, Diaz-Flores E, Lintault L, Kogan SC, Shannon K, et al. p53 loss promotes acute myeloid leukemia by enabling aberrant by the Dorris J. Hutchison Student Fellowship (SKD). A. Banito was self-renewal. Genes Dev 2010;24:1389–402. supported by a European Molecular Biology Organization long-term 20. Chen C, Liu Y, Rappaport AR, Kitzing T, Schultz N, Zhao Z, et al. fellowship. E.R. Kastenhuber was supported by an F31 National MLL3 is a haploinsufficient 7q tumor suppressor in acute myeloid Research Service Award predoctoral fellowship from the NCI/NIH leukemia. Cancer Cell 2014;25:652–65. under award number F31CA192835. F.J. Sanchez-Rivera is a Howard 21. Zuber J, Radtke I, Pardee TS, Zhao Z, Rappaport AR, Luo W, et al. Hughes Medical Institute Hanna Gray Fellow and was partially sup- Mouse models of human AML accurately predict chemotherapy ported by an MSKCC Translational Research Oncology Training response. Genes Dev 2009;23:877–89. Fellowship (NIH T32-CA160001). S.W. Lowe is an investigator at the 22. Ogilvy S, Elefanty AG, Visvader J, Bath ML, Harris AW, Adams JM. Howard Hughes Medical Institute and is the Geoffrey Beene Chair of Transcriptional regulation of vav, a gene expressed throughout the Cancer Biology. hematopoietic compartment. Blood 1998;91:419–30. 23. Donehower LA, Harvey M, Vogel H, McArthur MJ, Montgomery CA Received November 27, 2018; revised March 20, 2019; accepted Jr, Park SH, et al. Effects of genetic background on tumorigenesis in May 2, 2019; published first May 8, 2019. p53-deficient mice. Mol Carcinog 1995;14:16–22. 24. Hanel W, Marchenko N, Xu S, Yu SX, Weng W, Moll U. Two hot spot mutant p53 mouse models display differential gain of function in REFERENCES tumorigenesis. Cell Death Differ 2013;20:898–909. 25. Kuhn R, Schwenk F, Aguet M, Rajewsky K. Inducible gene targeting 1. Kandoth C, McLellan MD, Vandin F, Ye K, Niu B, Lu C, et al. in mice. Science 1995;269:1427–9. Mutational landscape and significance across 12 major cancer types. 26. Liu Y, Chen C, Xu Z, Scuoppo C, Rillahan CD, Gao J, et al. Deletions Nature 2013;502:333–9. linked to TP53 loss drive cancer through p53-independent mecha- 2. Olivier M, Hollstein M, Hainaut P. TP53 mutations in human can- nisms. Nature 2016;531:471–5. cers: origins, consequences, and clinical use. Cold Spring Harb Per- 27. Zuber J, McJunkin K, Fellmann C, Dow LE, Taylor MJ, Hannon GJ, spect Biol 2010;2:a001008. et al. Toolkit for evaluating genes required for proliferation and 3. Kern SE, Kinzler KW, Bruskin A, Jarosz D, Friedman P, Prives C, et al. survival using tetracycline-regulated RNAi. Nat Biotechnol 2011;29: Identification of p53 as a sequence-specific DNA-binding protein. 79–83. Science 1991;252:1708–11. 28. Wang T, Yu H, Hughes NW, Liu B, Kendirli A, Klein K, et al. Gene 4. Laptenko O, Prives C. Transcriptional regulation by p53: one protein, essentiality profiling reveals gene networks and synthetic lethal inter- many possibilities. Cell Death Differ 2006;13:951–61. actions with oncogenic ras. Cell 2017;168:890–903. 5. Serrano M, Lin AW, McCurrach ME, Beach D, Lowe SW. Oncogenic 29. Brummelkamp TR, Bernards R, Agami R. A system for stable ras provokes premature cell senescence associated with accumulation expression of short interfering RNAs in mammalian cells. Science of p53 and p16INK4a. Cell 1997;88:593–602. 2002;296:550–3. 6. Lowe SW, Ruley HE, Jacks T, Housman DE. p53-dependent apoptosis 30. Weisz L, Oren M, Rotter V. Transcription regulation by mutant p53. modulates the cytotoxicity of anticancer agents. Cell 1993;74:957–67. Oncogene 2007;26:2202–11. 7. Kastan MB, Onyekwere O, Sidransky D, Vogelstein B, Craig RW. 31. Pietras EM, Reynaud D, Kang YA, Carlin D, Calero-Nieto FJ, Leavitt Participation of p53 protein in the cellular response to DNA damage. AD, et al. Functionally distinct subsets of lineage-biased multipotent Cancer Res 1991;51:6304–11. progenitors control blood production in normal and regenerative 8. Kastenhuber ER, Lowe SW. Putting p53 in context. Cell 2017;170: conditions. Cell Stem Cell 2015;17:35–46. 1062–78. 32. Bagger FO, Kinalis S, Rapin N. BloodSpot: a database of healthy 9. Bouaoun L, Sonkin D, Ardin M, Hollstein M, Byrnes G, Zavadil J, and malignant haematopoiesis updated with purified and single cell et al. TP53 variations in human cancers: new lessons from the IARC mRNA sequencing profiles. Nucleic Acids Res 2019;47(D1):D881–D5. TP53 Database and Genomics Data. Hum Mutat 2016;37:865–76. 33. Passegue E, Jamieson CH, Ailles LE, Weissman IL. Normal and 10. Prives C, Hall PA. The p53 pathway. J Pathol 1999;187:112–26. leukemic hematopoiesis: are leukemias a stem cell disorder or a

JULY 2019 CANCER DISCOVERY | 977

RESEARCH ARTICLE Loizou et al.

reacquisition of stem cell characteristics? Proc Natl Acad Sci U S A expression profile that distinguishes a unique leukemia. Nat Genet 2003;100:11842–9. 2002;30:41–7. 34. Nowak D, Stewart D, Koeffler HP. Differentiation therapy of leuke- 56. Somervaille TC, Cleary ML. Identification and characterization of mia: 3 decades of development. Blood 2009;113:3655–65. leukemia stem cells in murine MLL-AF9 acute myeloid leukemia. 35. Nakahata T, Ogawa M. Identification in culture of a class of hemopoi- Cancer Cell 2006;10:257–68. etic colony-forming units with extensive capability to self-renew and 57. Zuber J, Rappaport AR, Luo W, Wang E, Chen C, Vaseva AV, et al. generate multipotential hemopoietic colonies. Proc Natl Acad Sci An integrated approach to dissecting oncogene addiction implicates U S A 1982;79:3843–7. a Myb-coordinated self-renewal program as essential for leukemia 36. Liu Y, Elf SE, Miyata Y, Sashida G, Liu Y, Huang G, et al. p53 regulates maintenance. Genes Dev 2011;25:1628–40. hematopoietic stem cell quiescence. Cell Stem Cell 2009;4:37–48. 58. Zhu J, Sammons MA, Donahue G, Dou Z, Vedadi M, Getlik M, et al. 37. Matsuoka S, Ebihara Y, Xu M, Ishii T, Sugiyama D, Yoshino H, Gain-of-function p53 mutants co-opt chromatin pathways to drive et al. CD34 expression on long-term repopulating hematopoietic cancer growth. Nature 2015;525:206–11. stem cells changes during developmental stages. Blood 2001;97: 59. Di Agostino S, Strano S, Emiliozzi V, Zerbini V, Mottolese M, Sacchi 419–25. A, et al. Gain of function of mutant p53: the mutant p53/NF-Y pro- 38. Moran-Crusio K, Reavie L, Shih A, Abdel-Wahab O, Ndiaye-Lobry tein complex reveals an aberrant transcriptional mechanism of cell D, Lobry C, et al. Tet2 loss leads to increased hematopoietic stem cycle regulation. Cancer Cell 2006;10:191–202. cell self-renewal and myeloid transformation. Cancer Cell 2011;20: 60. Do PM, Varanasi L, Fan S, Li C, Kubacka I, Newman V, et al. Mutant 11–24. p53 cooperates with ETS2 to promote etoposide resistance. Genes 39. Challen GA, Sun D, Jeong M, Luo M, Jelinek J, Berg JS, et al. Dnmt3a Dev 2012;26:830–45. is essential for hematopoietic stem cell differentiation. Nat Genet 61. Brennan J, Lu CC, Norris DP, Rodriguez TA, Beddington RS, 2011;44:23–31. Robertson EJ. Nodal signalling in the epiblast patterns the early 40. Seita J, Weissman IL. Hematopoietic stem cell: self-renewal versus dif- mouse embryo. Nature 2001;411:965–9. ferentiation. Wiley Interdiscip Rev Syst Biol Med 2010;2:640–53. 62. Yamamoto M, Meno C, Sakai Y, Shiratori H, Mochida K, Ikawa Y, 41. Muller PA, Vousden KH. p53 mutations in cancer. Nat Cell Biol et al. The transcription factor FoxH1 (FAST) mediates Nodal signal- 2013;15:2–8. ing during anterior-posterior patterning and node formation in the 42. Chiu WT, Charney Le R, Blitz IL, Fish MB, Li Y, Biesinger J, et al. mouse. Genes Dev 2001;15:1242–56. Genome-wide view of TGFbeta/Foxh1 regulation of the early mesen- 63. Vogelstein B, Papadopoulos N, Velculescu VE, Zhou S, Diaz LA Jr, doderm program. Development 2014;141:4537–47. Kinzler KW. Cancer genome landscapes. Science 2013;339:1546–58. 43. Attisano L, Silvestri C, Izzi L, Labbe E. The transcriptional role of 64. Wong TN, Ramsingh G, Young AL, Miller CA, Touma W, Welch JS, Smads and FAST (FoxH1) in TGFbeta and activin signalling. Mol Cell et al. Role of TP53 mutations in the origin and evolution of therapy- Endocrinol 2001;180:3–11. related acute myeloid leukaemia. Nature 2015;518:552–5. 44. Takahashi K, Tanabe K, Ohnuki M, Narita M, Sasaki A, Yamamoto 65. Xie M, Lu C, Wang J, McLellan MD, Johnson KJ, Wendl MC, et al. Age- M, et al. Induction of pluripotency in human somatic cells via a related mutations associated with clonal hematopoietic expansion transient state resembling primitive streak-like mesendoderm. Nat and malignancies. Nat Med 2014;20:1472–8. Commun 2014;5:3678. 66. Abelson S, Collord G, Ng SWK, Weissbrod O, Mendelson Cohen 45. Charney RM, Forouzmand E, Cho JS, Cheung J, Paraiso KD, Yasuoka N, Niemeyer E, et al. Prediction of acute myeloid leukaemia risk in Y, et al. Foxh1 occupies cis-regulatory modules prior to dynamic healthy individuals. Nature 2018;559:400–4. transcription factor interactions controlling the mesendoderm gene 67. Steensma DP, Bejar R, Jaiswal S, Lindsley RC, Sekeres MA, Hasserjian program. Dev Cell 2017;40:595–607. RP, et al. Clonal hematopoiesis of indeterminate potential and its 46. Cerami E, Gao J, Dogrusoz U, Gross BE, Sumer SO, Aksoy BA, et al. distinction from myelodysplastic syndromes. Blood 2015;126:9–16. The cBio cancer genomics portal: an open platform for explor- 68. Alexandrova EM, Yallowitz AR, Li D, Xu S, Schulz R, Proia DA, et al. ing multidimensional cancer genomics data. Cancer Discov 2012;2: Improving survival by exploiting tumour dependence on stabilized 401–4. mutant p53 for treatment. Nature 2015;523:352–6. 47. Gao J, Aksoy BA, Dogrusoz U, Dresdner G, Gross B, Sumer SO, et al. 69. Schulz-Heddergott R, Stark N, Edmunds SJ, Li J, Conradi LC, Integrative analysis of complex cancer genomics and clinical profiles Bohnenberger H, et al. Therapeutic ablation of gain-of-function using the cBioPortal. Sci Signal 2013;6:pl1. mutant p53 in colorectal cancer inhibits Stat3-mediated tumor 48. Spittau B, Krieglstein K. Klf10 and Klf11 as mediators of TGF-beta growth and invasion. Cancer Cell 2018;34:298–314. superfamily signaling. Cell Tissue Res 2012;347:65–72. 70. Kotler E, Shani O, Goldfeld G, Lotan-Pompan M, Tarcic O, Gershoni 49. Lee KS, Kim HJ, Li QL, Chi XZ, Ueta C, Komori T, et al. Runx2 is a A, et al. A systematic p53 mutation library links differential func- common target of transforming growth factor beta1 and bone mor- tional impact to cancer mutation pattern and evolutionary conserva- phogenetic protein 2, and cooperation between Runx2 and Smad5 tion. Mol Cell 2018;71:178–90. induces osteoblast-specific gene expression in the pluripotent mesen- 71. Morton JP, Timpson P, Karim SA, Ridgway RA, Athineos D, Doyle chymal precursor cell line C2C12. Mol Cell Biol 2000;20:8783–92. B, et al. Mutant p53 drives metastasis and overcomes growth 50. Kuo YH, Zaidi SK, Gornostaeva S, Komori T, Stein GS, Castilla LH. arrest/senescence in pancreatic cancer. Proc Natl Acad Sci U S A Runx2 induces acute myeloid leukemia in cooperation with Cbfbeta- 2010;107:246–51. SMMHC in mice. Blood 2009;113:3323–32. 72. Lang GA, Iwakuma T, Suh YA, Liu G, Rao VA, Parant JM, et al. 51. Schuler A, Schwieger M, Engelmann A, Weber K, Horn S, Muller U, Gain of function of a p53 hot spot mutation in a mouse model of et al. The MADS transcription factor Mef2c is a pivotal modulator of Li-Fraumeni syndrome. Cell 2004;119:861–72. myeloid cell fate. Blood 2008;111:4532–41. 73. Shibue T, Weinberg RA. EMT, CSCs, and drug resistance: the mechanis- 52. Giacomelli AO, Yang X, Lintner RE, McFarland JM, Duby M, Kim J, tic link and clinical implications. Nat Rev Clin Oncol 2017;14:611–29. et al. Mutational processes shape the landscape of TP53 mutations in 74. Mizuno H, Spike BT, Wahl GM, Levine AJ. Inactivation of p53 in human cancer. Nat Genet 2018;50:1381–7. breast cancers correlates with stem cell transcriptional signatures. 53. Kelly LM, Gilliland DG. Genetics of myeloid leukemias. Annu Rev Proc Natl Acad Sci U S A 2010;107:22745–50. Genomics Hum Genet 2002;3:179–98. 75. Ben-Porath I, Thomson MW, Carey VJ, Ge R, Bell GW, Regev A, et al. 54. Takahashi S. Current findings for recurring mutations in acute An embryonic stem cell-like gene expression signature in poorly dif- myeloid leukemia. J Hematol Oncol 2011;4:36. ferentiated aggressive human tumors. Nat Genet 2008;40:499–507. 55. Armstrong SA, Staunton JE, Silverman LB, Pieters R, den Boer 76. Morita S, Kojima T, Kitamura T. Plat-E: an efficient and stable system ML, Minden MD, et al. MLL translocations specify a distinct gene for transient packaging of retroviruses. Gene Ther 2000;7:1063–6.

978 | CANCER DISCOVERY JULY 2019 www.aacrjournals.org

77. Bolger AM, Lohse M, Usadel B. Trimmomatic: a flexible trimmer for identifies new disease entities and common transcriptional programs Illumina sequence data. Bioinformatics 2014;30:2114–20. in AML patients. Blood 2014;123:894–904. 78. Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, 81. Fellmann C, Hoffmann T, Sridhar V, Hopfgartner B, Muhar M, Roth et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics M, et al. An optimized microRNA backbone for effective single-copy 2013;29:15–21. RNAi. Cell Rep 2013;5:1704–13. 79. Anders S, Pyl PT, Huber W. HTSeq–a Python framework to work with 82. Banito A, Li X, Laporte AN, Roe JS, Sanchez-Vega F, Huang CH, et al. high-throughput sequencing data. Bioinformatics 2015;31:166–9. The SS18-SSX oncoprotein hijacks KDM2B-PRC1.1 to drive synovial 80. Rapin N, Bagger FO, Jendholm J, Mora-Jensen H, Krogh A, Kohlmann sarcoma. Cancer Cell 2018;33:527–41. A, et al. Comparing cancer vs normal gene expression profiles

JULY 2019 CANCER DISCOVERY | 979

A Gain-of-Function p53-Mutant Oncogene Promotes Cell Fate Plasticity and Myeloid Leukemia through the Pluripotency Factor FOXH1

Evangelia Loizou, Ana Banito, Geulah Livshits, et al.

Cancer Discov 2019;9:962-979. Published OnlineFirst May 8, 2019.

Updated version Access the most recent version of this article at: doi:10.1158/2159-8290.CD-18-1391

Supplementary Access the most recent supplemental material at: Material http://cancerdiscovery.aacrjournals.org/content/suppl/2019/05/08/2159-8290.CD-18-1391.DC1

Cited articles This article cites 82 articles, 29 of which you can access for free at: http://cancerdiscovery.aacrjournals.org/content/9/7/962.full#ref-list-1

Citing articles This article has been cited by 6 HighWire-hosted articles. Access the articles at: http://cancerdiscovery.aacrjournals.org/content/9/7/962.full#related-urls

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/9/7/962. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.