UNIVERSITY of CALIFORNIA, SAN DIEGO Characterization Of

UNIVERSITY OF CALIFORNIA, SAN DIEGO

Characterization of granulocyte-macrophage colony-stimulating factor signaling in

RUNX1-ETO leukemogenesis

A dissertation submitted in partial satisfaction of the requirements for the degree

Doctor of Philosophy

Biomedical Sciences

Stephanie Weng

Committee in charge:

Professor Dong-Er Zhang, Chair Professor Steven Dowdy Professor Gen-Sheng Feng Professor Mark Kamps Professor David Traver

2016

Stephanie Weng, 2016

The Dissertation of Stephanie Weng is approved, and it is acceptable in quality and form for publication on microfilm and electronically:

Chair

University of California, San Diego

2016

iii

DEDICATION

This dissertation is dedicated in loving memory to my mother, Linda Chien Weng.

Thank you for being a continual source of inspiration in every aspect of my life.

TABLE OF CONTENTS

Signature page ...... iii Dedication ...... iv Table of Contents ...... v List of Abbreviations ...... ix List of Figures ...... xi List of Tables ...... xiii Acknowledgements ...... xiv Vita ...... xviii Abstract of the Dissertation ...... xx Chapter 1: Introduction ...... 1 1.1. Loss of a sex chromosome in hematological malignancies ...... 2 1.2. Pseudoautosomal regions ...... 4 1.3. Pseudoautosomal region genes in the hematopoietic system ...... 6 1.3.1. Tumor suppressors ...... 7 1.3.2. Proto-oncogenes ...... 13 1.3.3. Context-dependent pseudoautosomal region genes ...... 16 1.3.4. Inactive pseudoautosomal region genes ...... 20 1.3.5. Other less characterized pseudoautosomal region genes and regions ...... 21 1.4. Sex chromosome amplifications in cancer ...... 24 1.5. Discussion ...... 24 1.6. Objectives of the dissertation ...... 25 1.7. Acknowledgements ...... 25 Chapter 2: Multiple GM-CSF-induced genes in RUNX1-ETO hematopoietic stem/progenitor cells function to inhibit self-renewal potential ...... 26 2.1. Introduction ...... 27 2.2. Results ...... 29 2.2.1. GM-CSF induces a human myelopoiesis gene expression profile in RUNX1- ETO hematopoietic stem/progenitor cells ...... 29 2.2.2. GM-CSF promotes differentiation and reduces the long-term culture initiating cell frequency of primary human RUNX1-ETO hematopoietic stem/progenitor cells ...... 36

2.2.3. Characterization of the GM-CSF-induced gene expression profile in inhibiting self-renewal potential of RUNX1-ETO hematopoietic stem/progenitor cells ...... 40 2.2.4. Validation and characterization of candidates from screen to identify GM- CSF-induced genes that reduce self-renewal potential of RUNX1-ETO hematopoietic stem/progenitor cells ...... 45 2.3. Discussion ...... 50 2.4. Materials and Methods...... 51 2.4.1. Plasmids ...... 51 2.4.2. Gene expression profiling...... 52 2.4.3. Gene expression profiling data analysis ...... 52 2.4.4. Quantitative reverse-transcription PCR (qRT-PCR)...... 53 2.4.5. CD34 + cell isolation, transduction, and culture ...... 56 2.4.6. Long-Term Culture-Initiating Cell (LTC-IC) assays ...... 56 2.4.7. Flow cytometry ...... 57 2.4.8. Barcoded cDNA screen ...... 57 2.4.9. Cell culture ...... 60 2.4.10. Proliferation assays ...... 60 2.4.11. Replating assays ...... 60 2.4.12. Microscopy ...... 60 2.4.13. Statistical analysis ...... 61 2.5. Acknowledgements ...... 61 Chapter 3: Attenuation of MYC reduces the leukemic potential of RUNX1-ETO cells ....63 3.1. Introduction ...... 64 3.2. Results ...... 66 3.2.1 MYC-associated gene signatures are attenuated in GM-CSF-treated RUNX1- ETO hematopoietic stem/progenitor cells ...... 66 3.2.2 MXI1 expression in RUNX1-ETO hematopoietic stem/progenitor cells and t(8;21) cell lines reduces leukemic potential ...... 67 3.2.3 MYC knockdown reduces the leukemic potential of t(8;21) cell lines ...... 70 3.2.4 JQ1 treatment reduces the leukemic potential of RUNX1-ETO hematopoietic stem/progenitor cells and t(8;21) cell lines ...... 73 3.2.5 MXI1 relieves the RUNX1-ETO-induced repression of RUNX1 target sites ...75

3.3. Discussion ...... 79 3.4. Materials and Methods...... 81 3.4.1 Plasmids ...... 81 3.4.2 Cell culture ...... 81 3.4.3 Proliferation assays ...... 81 3.4.4 Flow Cytometry ...... 82 3.4.5 Colony forming assays ...... 82 3.4.6 Western blot ...... 82 3.4.7 Quantitative reverse-transcription PCR (qRT-PCR) ...... 82 3.4.8 Luciferase assay ...... 84 3.4.9. Co-immunoprecipitation ...... 84 3.4.10. Statistical analysis ...... 85 3.5. Acknowledgements ...... 85 Chapter 4: CSF2RA is post-transcriptionally downregulated in t(8;21) cell lines ...... 86 4.1. Introduction ...... 87 4.2. Results ...... 89 4.2.1. CSF2RA is downregulated in t(8;21) cell lines ...... 89 4.2.2 CSF2RA is downregulated through the 3’ UTR in t(8;21) cell lines ...... 89 4.2.3. The CSF2RA coding sequence and 3’ UTR are required for maximal downregulation ...... 94 4.2.4. The strongest AU-rich element in the 3’ UTR of CSF2RA does not entirely mediate the downregulation of CSF2RA ...... 99 4.3. Discussion ...... 100 4.4. Materials and Methods...... 104 4.4.1. Plasmids ...... 104 4.4.2. Flow cytometry ...... 104 4.4.3. Quantitative reverse-transcription PCR (qRT-PCR)...... 104 4.5. Acknowledgements ...... 105 Chapter 5: Conclusion and Future Directions ...... 106 5.1. Conventional t(8;21) AML treatment and alternative approaches ...... 107 5.2. Caveats to the use of GM-CSF in the treatment of AML ...... 108 5.3. Therapeutic implications of studying GM-CSF signaling in t(8;21) cells ...... 108

vii

5.4. Additional implications for GM-CSF signaling in the pathogenesis of t(8;21) AML ...... 110 5.5. Impact of the dissertation ...... 111 5.6. Acknowledgments ...... 111 References ...... 113

viii

LIST OF ABBREVIATIONS

ADP Adenine diphosphate ALL Acute lymphoid leukemia AML Acute myeloid leukemia ARE-BP Adenylate-uridylate-rich elements binding protein ATP Adenine triphosphate ATRA All-trans retinoic acid AU-rich Adenylate-uridylate-rich BET Bromodomain and extraterminal domain Bmp2 Bone morphogenic protein 2 bp basepair BSA Bovine serum albumin C/EBPα CCAAT/enhancer-binding protein alpha CAR Chimeric antigen receptor Cdkn2a Cyclin-dependent kinase inhibitor 2a Cdkn2b Cyclin-dependent kinase inhibitor 2b cDNA Complementary deoxyribonucleic acid CEBP CCAAT/enhancer-binding protein CLL Chronic lymphoid leukemia CML Chronic myeloid leukemia CMML Chronic myelomonocytic leukemia CNL Chronic neutrophilic leukemia CRISPR Clustered regularly interspaced short palindromic repeats Cxcl1 C-X-C motif ligand 1 DC Dendritic cell DMSO Dimethyl sulfoxide DNA Deoxyribonucleic acid EGFP Enhanced green fluorescent protein ES Enrichment score FACS Fluorescent activated cell sorting FBS Fetal bovine serum FDR False discovery rate FLT3L FMS-like tyrosine kinase 3 ligand GADD45A Growth arrest and DNA damage inducible alpha GM-CSF or GM Granulocyte-macrophage colony-stimulating factor GSEA Gene Set Enrichment Analysis HA Hemagluttinin HDAC Histone deacetylase HSC Hematopoietic stem cell HSPC Hematopoietic stem/progenitor cell IL-3 Interleukin-3 IL-6 Interleukin-6 IPA Ingenuity Pathway Analysis IRES Internal ribosomal entry site JAK Janus kinase JMML Juvenile myelomonocytic leukemia

Kb Kilobase pair Lin- Lineage negative LOH Loss of heterozygosity LOS Loss of a sex chromosome LTB4 Leukotriene B4 Ltb4r1 Leukotriene B4 receptor LTC-IC Long-term culture-initiating cell MAPK Mitogen-activated protein kinase MAX MYC-associated factor Mb Megabase pair MDS Myelodysplastic syndrome MEK Mitogen-activated protein kinase kinase MIG MSCV-IRES-GFP MIP MSCV-IRES-puromycin R miRNA MicroRNA mRNA Messenger ribonucleic acid MSCV Murine stem cell virus MXI1 Max interacting protein-1 MYC V-myc myelocytomatosis viral oncogene homolog NES Nominal enrichment score NK Natural killer PABP Poly(A) binding protein PAR Pseudoautosomal region PBS Phosphate buffered saline PCR Polymerase chain reaction pDC Plasmacytoid dendritic cell PI3K Phosphoinositide 3-kinase qPCR Quantitative Polymerase Chain Reaction qRT-PCR Quantitative reverse transcriptase-polymerase chain reaction RAF Raf oncogene serine/threonine kinase RAS Ras viral oncogene homolog RE RUNX1-ETO, AML1-ETO, RUNX1-RUNX1T1, AML1-MTG8 RNA Ribonucleic acid SCF Stem cell factor SCL Sex chromosome loss SD Standard deviation SEM Standard error of the mean SFEM Serum free expansion medium shRNA Small hairpin ribonucleic acid STAT Signal transducer and activator of transcription TK Thymidine kinase TNF Tumor necrosis factor TPA 12-O-Tetradecanoylphorbol-13-acetate UTR Untranslated region

LIST OF FIGURES

Figure 1.1. Schematic of the X and Y sex chromosomes with the distal pseudoautosomal regions ...... 5 Figure 1.2. Protein phosphatase 2A subunits and respective encoding genes ...... 10 Figure 1.3. Overview of melatonin synthesis ...... 12 Figure 1.4. Schematic of the P2RY8-CRLF2 rearrangement ...... 14 Figure 1.5. The GM-CSF receptor complex and its downstream signaling pathways ... 17 Figure 2.1. Diagram of experimental methods for gene expression profiling of murine RUNX1-ETO hematopoietic stem/progenitor cells treated with GM-CSF ... 30 Figure 2.2. Analysis of differentially expressed genes after GM-CSF-treatment of RUNX1-ETO hematopoietic stem/progenitor cells ...... 32 Figure 2.3. Gene Set Enrichment Analysis of GM-treated RUNX1-ETO hematopoietic stem/progenitor cells ...... 33 Figure 2.4. Validation of microarray data ...... 34 Figure 2.5. Gene Set Enrichment Analysis of GM-CSF-treated RUNX1-ETO hematopoietic stem/progenitor cells with a human myelopoiesis dataset .... 35 Figure 2.6. Validation of RUNX1-ETO expression in CD34 + hematopoietic stem/progenitor cells ...... 38 Figure 2.7. GM-CSF promotes differentiation and reduces the long-term culture initiating cell frequency of primary human RUNX1-ETO hematopoietic stem/progenitor cells ...... 39 Figure 2.8. CD34 + percentages and proliferation of primary human hematopoietic stem/progenitor cells cultured with GM-CSF ...... 41 Figure 2.9. Bioinformatic analysis of differentially expressed genes in GM-CSF-treated RUNX1-ETO hematopoietic stem/progenitor cells ...... 43 Figure 2.10. Barcoded cDNA screen to identify GM-CSF-induced genes that inhibit self- renewal potential of RUNX1-ETO hematopoietic stem/progenitor cells ...... 44 Figure 2.11. Barcode screen reveals genes that reduce RUNX1-ETO hematopoietic stem/progenitor cells self-renewal capacity ...... 46 Figure 2.12. Ltb4r1 is a GM-CSF-induced gene that mediates its inhibitory effects on RUNX1-ETO hematopoietic stem/progenitor cells ...... 49 Figure 3.1. GM-CSF treatment of RUNX1-ETO hematopoietic stem/progenitor cells attenuates MYC-associated gene signatures and MXI1 expression reverses RUNX1-ETO-associated phenotypes ...... 68 Figure 3.2. MXI1 expression inhibits RUNX1-ETO-associated phenotypes in t(8;21) AML cell lines ...... 69 Figure 3.3. MYC knockdown induces apoptosis in t(8;21) human cell lines ...... 71 Figure 3.4. MYC knockdown reduces proliferation of t(8;21) cell lines and restores MYC- repressed targets in Kasumi-1 cells ...... 72

Figure 3.5. JQ1 treatment reduces proliferation and induces apoptosis in t(8;21) cell lines ...... 74 Figure 3.6. Proposed mechanism of the inhibitory effects of GM-CSF on the leukemic potential of RUNX1-ETO hematopoietic stem/progenitor cells ...... 76 Figure 3.7. MXI1 alleviates the RUNX1-ETO-induced repression of RUNX1 transcriptional target sites independently of inhibiting Sin3A recruitment to ETO ...... 78 Figure 4.1. CSF2RA expression decreases in t(8;21) cell lines ...... 90 Figure 4.2. EGFP reporter expression decreases over time in t(8;21) cell lines ...... 91 Figure 4.3. The 3’ UTR mediates downregulation of CSF2RA in t(8;21) cell lines ...... 93 Figure 4.4. Candidate CSF2RA 3’ UTR-targeting miRNAs do not promote downregulation of CSF2RA in K562 cells ...... 95 Figure 4.5. The CSF2RA coding sequence and 3’ UTR are targeted for downregulation in Kasumi-1 cells ...... 97 Figure 4.6. The coding sequence of CSF2RA is required for 3’ UTR-mediated downregulation of CSF2RA in Kasumi-1 cells ...... 98 Figure 4.7. Mutation of the strongest AU-rich element does not fully rescue the downregulation of CSF2RA ...... 101

xii

LIST OF TABLES

Table 1.1. Overview of the pseudoautosomal region genes and their functions...... 8 Table 2.1. Ingenuity Pathway Analysis Upstream Regulator Analysis...... 37 Table 2.2. Percentage cDNA Barcode Reads Normalized to T0...... 47 Table 2.3. Mouse qPCR primers ...... 54 Table 2.4. Human qPCR primers ...... 55 Table 3.1. Human qPCR primers ...... 83

xiii

ACKNOWLEDGEMENTS

The completion of this dissertation would not have been possible without the

support of many people in my life. First, I would like to acknowledge my thesis advisor,

Dr. Dong-Er Zhang, for her mentorship and guidance during my PhD studies. Dong-Er,

thank you for constantly pushing me to surpass my own boundaries and limitations to

show me that I am capable of more than I think I am. I am also extremely grateful to the

past and present members of the Zhang lab for their support and friendship over the

years. I would like to acknowledge Miao-Chia Lo, Dan Ran, Ming Yan, Christoph

Burkart, Yukiko Komeno, Russell DeKelver, and Kentson Lam, who were especially kind

and helpful when I first began my studies in the lab. Additionally, I thank my

undergraduate student Cody Mowery for all of his hard work and dedication to helping

me with my projects. Current lab members Junbao Fan, Kei-ichiro Arimoto, Sam Stoner,

Amanda Davis, Yi-Jou Huang, Sayuri Ishida, Takahiro Shima, and Daniel Johnson have also been extremely generous with their support throughout the years. I thank my committee, Steven Dowdy, Gen-Sheng Feng, Mark Kamps, and David Traver, for their insightful discussions of my thesis work and advice during our committee meetings. I thank the Biomedical Sciences Program and its wonderful administrators, Leanne

Nordeman and Gina Butcher, for always offering a helping hand. I am grateful to have received support from the National Cancer Institute Cancer Cell Biology Training Grant

(T32CA067754) and the Achievement Rewards for College Scientists (ARCS)

Foundation.

I am fortunate to have had Huilin Zhou as a former mentor and who was extremely supportive of me and taught me to “keep it simple”. Additionally, I am appreciative of my former CytomX collegaues, Nancy Stagliano, who took a chance and

xiv

gave me an amazing opportunity, as well as Jason Sagert and Vanessa Huels for all of

their encouragement. My former labmates and colleagues, Claudio Ponte De

Albuquerque, Jason Liang, and Gabriel Pineda have also been so supportive to me.

In addition to those who have supported me in the academic setting, numerous

family members and friends have also provided me with invaluable forms of support and

have been overwhelmingly understanding when I have failed to be physically or mentally

present because, as my advisor says, “there are no holidays in science”. First, I would

like to thank my dad, Ken Weng, for raising me to share his passion for seeking answers

to life’s questions and instilling his innovative problem-solving tactics in me. I would also

like to thank my stepmom, Leona Fung, for teaching me to standing up for myself and for

what I believe in. I thank my little my brother, Andrew Weng, for all of his support. I am

grateful to my grandparents, Grace and Charles Chien and Sarah Weigh, and my cousin

Adam Wu, for always being there for me and believing in me.

I am infinitely grateful for all of my incredible friends who have supported me

through the ups and downs. I am especially grateful to Jacqueline Benthuysen for

always listening to what I have to say, for all of her advice, and for being by my side

throughout this journey. Thank you, Hiruy Meharena, for continually reminding me to find

the positive in any situation. I am extremely appreciative of Erica Voelz for always going

out of her way to show her support in everything that I do. I have been extremely

blessed to have Starr and Quetzalsol Lopez as my San Diego family; Starr, thank you for

always joining me in escaping through music. I am grateful to have had the opportunity

to meet great friends like Jerome Karpiak and Danjuma Quarless from graduate school.

Thank you Matthew Ruppel, for providing me with much-needed advice and support

over the years. Also, I thank Jenelle Shapiro for her friendship over the past 19 years. I

would also like to acknowledge Ellen Hames, Justin and Andrea Hickman, Jenny Holl,

Mary Hull, Ashley Philpot-Mattos, Elona Rista, and Julie Sheline, who have been great friends to me. I am so grateful to my San Diego music family who have welcomed me with open arms and provided me with much-needed therapy throughout the course of my graduate studies. Finally, I would like to thank music, which has been there for me when I needed it most.

Chapter 1 includes material currently under revision for publication as a review article. Weng S, Stoner SA, Zhang DE “Sex chromosome loss and the pseudoautosomal regions in hematological malignancies.” The dissertation author was the primary investigator and author of this manuscript. S.W., S.S., and D.Z. wrote the manuscript.

S.W. prepared figures.

Chapters 2, 3, and 5 include published material from Weng S, Matsuura S,

Mowery CT, Stoner SA, Lam K, Ran D, Davis AG, Lo MC, Zhang DE. “Restoration of

MYC-repressed targets mediates the negative effects of GM-CSF on RUNX1-ETO leukemogenicity.” Leukemia, 2016. The dissertation author was the primary investigator and author of this study. S.W. designed and performed the experiments. S.M. performed microarray gene expression profiling at the UC San Diego IGM Genomics Center, La

Jolla, CA. Barcode sequencing was performed at the UC San Diego IGM Genomics

Center, La Jolla, CA. C.M., S.S., K.L., D.R., A.D., and M.L. assisted with experiments.

S.W. analyzed the data and prepared figures. S.W. and D.Z. wrote the manuscript.

Chapter 3 also includes work performed by Russell DeKelver. The authors would like to thank D. Neuberg and K. Stevenson for advising with microarray data analysis, O.

Harismendy and K. Jepsen for their assistance with the barcode sequencing and data

xvi

analysis, and D.J. Young for flow cytometry expertise. This work was funded by the

National Institutes of Health (NIH R01CA192924).

Chapter 4 includes material currently being prepared for submission as a manuscript. Weng S, Johnson D, Zhang DE. “ CSF2RA is post-transcriptionally downregulated in t(8;21) acute myeloid leukemia cell lines.” The dissertation author was the primary investigator and author of this study. S.W. designed and performed the experiments. D.J. assisted in cloning constructs for this study. S.W. analyzed the data and prepared figures. S.W. and DE.Z. are writing the manuscript. This work was funded

by the National Institutes of Health (5R01CA192924).

xvii

VITA

2005-2007 Undergraduate Researcher, University of California, Santa Barbara

2007-2008 Lab Assistant/Research, CytomX Therapeutics

2008 Bachelor of Science, Microbiology, University of California, Santa Barbara

2008-2010 Research Associate, Ludwig Institute for Cancer Research, University of California, San Diego Branch

2010-2014 Graduate Student Researcher, University of California, San Diego

2014-2016 Ph.D. Candidate, University of California, San Diego

2016 Doctor of Philosophy, Biomedical Sciences, University of California, San Diego

PUBLICATIONS

Weng S , Zhang DE. Sex chromosome loss and the pseudoautosomal regions in hematological malignancies. Oncotarget , 2016. (Accepted with minor revisions)

Weng S , Matsuura S, Mowery CT, Stoner SA, Lam K, Ran D, Davis AG, Lo MC, Zhang DE. Restoration of MYC-repressed targets mediates the negative effects of GM-CSF on RUNX1-ETO leukemogenicity. Leukemia , 2016.

Arimoto K, Weng S , Zhang DE. Plakophilin-2 induced EGFR phosphorylation: a focus on the intracellular activators of EGFR. Receptors & Clinical Investigation , 2015.

Lam K, Muselman A, Du R, Harada Y, Yan M, Matsuura S, Weng S , Harada H, Zhang DE. Hmga2 is a direct target gene of RUNX1 and regulates expansion of myeloid progenitors. Blood , 2014.

Arimoto K, Burkart C, Yan M, Ran D, Weng S , Zhang DE. Plakophilin-2 Promotes Tumor Development by Enhancing Ligand Dependent and -Independent Epidermal Growth Factor Receptor Dimerization and Activation. Molecular and Cellular Biology , 2014.

Dekelver RC, Lewin B, Weng S , Yan M, Biggs J, Zhang DE. (2013) RUNX1-ETO induces a type I interferon response which negatively effects t(8;21)-induced increased self-renewal and leukemia development. Leukemia & Lymphoma , 2014.

Matsuura S, Yan M, Lo MC, Ahn EY, Weng S , Dangoor D, Matin M, Higashi T, Feng GS, Zhang DE. Negative effects of GM-CSF signaling in a murine model of t(8;21)- induced leukemia. Blood , 2012.

xviii

Wang G, Tong X, Weng S , Zhou H. Multiple phosphorylation of Rad9 by CDK is required for DNA damage checkpoint activation. Cell Cycle , 2012.

Zhou H, Albuquerque CP, Liang J, Suhandynata R, Weng S . Quantitative phosphoproteomics: New technologies and applications in the DNA damage response. Cell Cycle , 2010.

FIELDS OF STUDY

Major Field: Cancer Biology

Studies in Leukemia and Hematopoiesis Professor Dong-Er Zhang

xix

ABSTRACT OF THE DISSERTATION

Characterization of granulocyte-macrophage colony-stimulating factor signaling in

RUNX1-ETO leukemogenesis

Stephanie Weng

Doctor of Philosophy in Biomedical Sciences

University of California, San Diego, 2016

Professor Dong-Er Zhang, Chair

The t(8;21)(q22;q22) generates the oncofusion protein RUNX1-ETO (AML1-ETO,

RUNX1-RUNX1T1, AML1-MTG8) and is one of the most common chromosomal abnormalities associated with acute myeloid leukemia (AML). Loss of a sex chromosome (LOS) is detected in 32-59% of t(8;21) AML patients, which suggests haploinsufficiency of key tumor suppressors on the sex chromosomes may be involved in t(8;21) leukemogenesis. Though LOS is frequently observed in t(8;21) patients, there have been no mechanistic studies examining its role in the pathogenesis of t(8;21) AML.

The gene encoding the granulocyte-macrophage colony-stimulating factor (GM-CSF) receptor, CSF2RA , is located on the sex chromosomes. With LOS, t(8;21) cells suffer

haploinsufficiency of CSF2RA , resulting in reduced GM-CSF signaling. It was recently

discovered that GM-CSF signaling, which is compromised upon LOS, inhibits t(8;21)

leukemogenesis. Therefore, this dissertation explores the mechanisms by which GM-

CSF signaling prevents t(8;21) leukemogenesis and aims to identify mediating

mechanisms that are therapeutically targetable. Gene expression profiling of GM-CSF-

treated RUNX1-ETO cells revealed that GM-CSF signaling elicits a unique gene

expression profile in RUNX1-ETO cells. This gene expression profile correlates with

attenuated MYC-associated gene signatures and human myelopoiesis, which together

reduce the self-renewal potential of RUNX1-ETO cells and inhibit leukemogenesis. To

identify GM-CSF-induced genes mediating the inhibition of RUNX1-ETO

leukemogenesis, a functional screen was conducted to identify several genes, including

Mxi1 , which reduce the self-renewal potential of RUNX1-ETO cells. MXI1 is a well-

established inhibitor of MYC; therefore, the effects of MXI1 and MYC inhibition were

examined in t(8;21) cells. It was found that MYC inhibition promotes apoptosis and

reduces the proliferation and self-renewal potential of t(8;21) cells, demonstrating that

MYC is critical for maintaining leukemic potential and may serve as an effective target in

t(8;21) AML treatment. Furthermore, it was discovered that in addition to being

downregulated upon LOS-driven haploinsufficiency, CSF2RA is also downregulated in

t(8;21) cell lines through its 3’ UTR. This reveals an additional mechanism present in

t(8;21) cells to reduce GM-CSF signaling and promote leukemogenesis. Together, these

studies highlight the importance of GM-CSF signaling in attenuating MYC in t(8;21) cells

and elucidate the mechanisms by which t(8;21) cells downregulate CSF2RA to reduce

GM-CSF signaling.

xxi

CHAPTER 1:

Introduction

1 2

Chromosomal abnormalities are frequently detected in cancers. These

abnormalities include events such as aneuploidy, chromosomal duplications, deletions,

inversions, and translocations. Genomic regions containing critical proto-oncogenes and

tumor suppressors are often targeted by these events. The subsequent alteration of the

copy number, expression, or sequence of these proto-oncogenes and tumor

suppressors ultimately contributes to or promotes tumorigenesis. Numerous studies

have largely expanded our understanding of the mechanisms underlying various

chromosomal abnormalities and their impact on cancer development and progression 1.

However, there remain chromosomal aberrations that have not been investigated as

thoroughly, such as sex chromosome abnormalities.

Sex chromosome aberrations have been reported in a variety of malignancies.

One of the most commonly observed sex chromosome aberrations is aneuploidy or loss,

which has been reported in a variety of cancers, including lung 2, pancreatic 3, breast 4, 5,

colorectal 6, and hematological cancers 7. These observations suggest that tumor

suppressor genes may reside on the sex chromosomes, and their loss may contribute to

cancer development or progression.

1.1. Loss of a sex chromosome in hematological malignancies

Loss of a sex chromosome (LOS) is frequently observed in hematological

malignances. In a study of 868 patients with hematologic diseases, 5.1% of patients

exhibited LOS 8. Acute myeloid leukemia (AML) patients displayed the highest frequency

of LOS, 9.5%, compared to a range of 0-6% for acute lymphoid leukemia (ALL), chronic

lymphoid leukemia (CLL), or chronic myeloid leukemia (CML) patients 8. Loss of the Y chromosome has also been observed in myelodysplastic syndrome (MDS) 9.

Cytogenetic studies of normal non-malignant bone marrow and peripheral blood

cells indicate that loss of a sex chromosome or sex chromosome loss (SCL) is a

phenomenon associated with aging 10-12 . Although aneuploidy generally increases with age 13, 14 , LOS has been reported to occur at a rate 10-fold greater than autosomal

chromosome loss 15 . Additionally, LOS is observed at higher frequencies and at younger

ages in cancers. Several studies of childhood ALL have reported LOS in patient

leukemia cells 16, 17 . Another larger cytogenetic study of 270 AML patient samples

revealed that LOS was observed in 22 patients. Of these 22 patients, 50% were under

the age of 14, 40% were aged 14 to 50, and 10% were aged above 50 18 , which further corroborates that LOS occurs at younger ages in cancers. Additionally, in a search for cytogenetic aberrations that may confer adverse risk in childhood ALL, LOS was correlated with increased risk 17 . Altogether, these observations find a positive correlation between LOS and hematological malignancies, implicating LOS as a risk factor for cancer development and prognosis.

Although LOS is common in hematological malignancies, only 1.8% of patients exhibit LOS as a sole genetic abnormality 8, and LOS is frequently observed with

additional mutations 7. However, LOS is typically observed in clonal populations, which further suggests that this cytogenetic aberration is favored for oncogenesis 19 . These

findings indicate that LOS may not be sufficient for malignant transformation, but is likely

to serve as a cooperating event 7. In support of this is the frequent co-occurrence of LOS

in around 32-59% of t(8;21) AML patients, which is much higher compared to other types

of AMLs and leukemias 20, 21 . Unlike other oncogenic translocations, t(8;21) is insufficient for leukemogenesis and requires additional cooperating mutations, which LOS may serve as 22 .

Altogether, these findings support that the sex chromosomes may harbor critical

tumor suppressor genes, which when lost upon LOS, may contribute to disease

development. However, the mechanistic contributions of LOS to disease development

are not currently understood and have yet to be thoroughly investigated.

1.2. Pseudoautosomal regions

The X chromosome is composed of just over 156 million base pairs, which

contain 1,805 genes. The Y chromosome is roughly a third of the size and is composed

of around 57 million base pairs, which contain 458 genes. Females possess both a

maternal and paternal X chromosome, whereas males possess a maternal X

chromosome and a paternal Y chromosome 23 . In females, one of the X chromosomes is subjected to high levels of DNA methylation to generate repressive heterochromatin, resulting in X-inactivation and gene silencing 23 .

Although the X and Y sex chromosomes are highly divergent in sequence, the

pseudoautosomal regions (PARs) are homologous between the two chromosomes,

escape X-inactivation, and facilitate proper pairing and segregation of the sex

chromosomes during meiosis 24 . There are two main pseudoautosomal regions (PAR1 and PAR2) on the sex chromosomes, and they are localized to the distal ends of the p and q chromosomal arms, respectively 25 (Figure 1.1.). PAR1 is 2.6Mb and contains 16 genes, whereas PAR2 is 360Kb and contains 3 genes and 1 pseudogene, CXYorf1 or

WASH6P . In 2013, a third pseudoautosomal region (PAR3) was identified and reported to have arisen from duplication and transposition of the q21.3 region of the X chromosome, which contains contains the genes PCDH11X/Y and TGIF2LX/Y, to the p11.2 region of the Y chromosome 26 . Because PAR3 is only observed in

Figure 1.1. Schematic of the X and Y sex chromosomes with the distal pseudoautosomal regions. PAR1 contains 16 genes, with PLCXD1 as the furthermost PAR1 gene at the distal telomeric end and XG at the boundary of PAR1 at the centromeric end. PAR2 contains 3 genes, with SPRY3 at the centromeric boundary and IL9R at the distal telomeric end.

approximately2% of the general population, we focus on PAR1 and 2 here.

Because the PARs escape X-inactivation, PAR genes are biallelically expressed

in both females and in males. In the case of LOS, PAR genes suffer from

haploinsufficiency and are the only genes commonly affected in both males and females

with LOS. Consequently, the PAR genes are the most likely candidates as common

tumor suppressors amongst the X and Y chromosomes.

Further substantiating that PAR genes may possess tumor suppressive functions

is the observation that pseudoautosomal regions are frequent targets for loss of

heterozygosity (LOH) in cancers. LOH in a region of the PAR, which results in deletion of

SHOX, CRLF2, and CSF2RA , has been observed in mantle cell lymphoma 27 .

While LOS is frequently observed in hematological malignancies, especially

AMLs, there is a shortage of studies examining PAR genes in hematological disease

development and progression. Moreover, many of the PAR genes are not well

characterized. Further mechanistic studies of the role of PAR genes in normal and

malignant hematological conditions will contribute immensely to our understanding of

how LOS may be involved in oncogenesis. Additionally, systematic knockout or enforced

expression of individual or a combination of PAR genes in the hematopoietic system will

greatly elucidate the function of these genes in hematological diseases.

Here, we overview what is currently known about the PAR genes in the

hematopoietic system and discuss the implications for LOS and haploinsufficiency of

PAR genes in the development of hematological malignancies.

1.3. Pseudoautosomal region genes in the hematopoietic system

The nineteen PAR genes have a broad range of diverse functions. Some of

these genes act as tumor suppressors or proto-oncogenes. Others have been reported

to function as both, depending on the cellular context, which highlights the need for

examining the roles of PAR genes in individual hematological conditions (Table 1.1).

1.3.1. Tumor suppressors

PPP2R3B

There are several PAR genes that have been implicated in possessing tumor

suppressive functions. One of these genes is PPP2R3B (NYREN8, PPP2R3L,

PPP2R3LY, PR48 ) which encodes one of the subunits of protein phosphatase 2A

(PP2A). PP2A is a holoenzyme composed of A, B, and C subunits, each of which is

encoded by several genes 28 (Figure 1.2.). The A and C subunits form a catalytic core, and the B subunit is involved in determining substrate specificity. PP2A functions as a serine-threonine phosphatase and regulates various intracellular signaling pathways, and is generally regarded as a tumor suppressor 28 . Deletion, mutation, or inactivation of

PP2A is frequently observed in various cancer types28 . The oncofusion protein BCR-

ABL1, which is generated from t(9;22), has been reported to inhibit PP2A activity 29 .

Interestingly, inhibition of PP2A was found to be required for proliferative and pro-

survival signals in CML 29 . Furthermore, the sphingosine analog FTY720, which has been

demonstrated to activate PP2A, was found to reduce cell proliferation and induce

apoptosis in models of CML, as well as on the Kasumi-1 t(8;21) AML cell line 30, 31 .

. enes and their functions their and enes

Table 1.1. Overview of the pseudoautosomal region g region pseudoautosomal the of Overview Table 1.1.

Because PP2A is composed of multiple subunits encoded by 17 different genes, there

are multiple avenues for altering the expression or activity of this phosphatase. Analysis

of the genes encoding the different subunits of PP2A revealed that one of the genes

encoding the A subunit, PPP2R1B , is downregulated in AML cells, and two genes encoding the B subunits, PPP2R3B and PPP2R2C are not expressed in AML cells 32 .

Altogether, reduction in PPP2R3B gene dosage via LOS may result in reduced overall

activity of PP2A, which could promote increased proliferation and confer survival

advantage for leukemic cells.

SLC25A6

In a screen for genes required for TNF-induced apoptosis in the MCF-7 breast

cancer cell line, Solute carrier family 25 member 6 ( SLC25A6, AAC3, ANT3 ) was

identified 33 . In this study, SLC25A6 was found to be necessary for apoptosis induced by

TNF and oxidative stress, but not other types of apoptosis. SLC25A6 is a member of a family of genes that encode a component of the mitochondrial permeability transition pore (MPTP), which is critical in regulating cellular energy metabolism and apoptosis 33 . It

functions as a transporter of ADP from the cytosol into the mitochondrial matrix and ATP

from the mitochondrial matrix to the cytosol. Exogenous expression of SLC25A6 in HeLa

cells increased sensitivity to ATRA-induced apoptosis 34 . SLC25A6 has also been reported to facilitate apoptosis induced by camptothecin, a cytotoxic agent, and knockdown of SLC25A6 impairs apoptosis. Interestingly, SLC25A6 was downregulated in several cancer cell lines upon treatment with camptothecin, including the leukemia cell lines HL-60 and K562, suggesting that reduced levels of SLC25A6 may aid in conferring

resistance to camptothecin-induced apoptosis 35 . Although the functions of SLC25A6 in hematopoiesis or leukemogenesis are not well characterized, these data indicate that it

Figure 1.2. Protein phosphatase 2A subunits and respective encoding genes. Schematic of the serine-threonine protein phosphatase (PP2A) complex. The PP2A complex is composed of three subunits: A, B, and C. Each respective subunit is encoded by the listed genes. The A and C subunits form a catalytic core, and the B subunit is involved in determining substrate specificity.

may serve as a tumor suppressor through its role in activating apoptosis.

SHOX

Short stature homeobox ( SHOX, GCFX, PHOGY, SS ) regulates bone growth and

maturation. Several groups have reported gender concordance in siblings with Hodgkin’s

lymphoma 36, 37 . Additionally, Hodgkin’s lymphoma has been found to occur in association

with Leri-Weill dyschondrosteosis, a bone growth disorder that results in shortened arm

and leg bones, which is caused by deletions or mutations of the PAR gene SHOX .

These findings led to the hypothesis that disruption to SHOX may be linked to the

development of Hodgkin’s lymphoma 38 . Additionally, deletion of SHOX has also been

reported in mantle cell lymphoma 27 .

ASMT

Melatonin is an antioxidant that easily passes through cell membranes.

Treatment of several human leukemia cell lines, including CMK, Jurkat, MOLT-3 and

MOLT-4 with melatonin resulted in cytotoxic effects39, 40 . ASMT (ASMTY , HIOMT ,

HIOMT ) encodes acetylseratonin O-methyltransferase, an enzyme that catalyzes the

final step in melatonin synthesis (Figure 1.3.). Melatonin is predominantly synthesized in

the pineal gland; however, melatonin has also been found to be synthesized by human

bone marrow cells and several human hematopoietic cell lines 41 . ASMT expression is

highest in T lymphocytes 42 , and ASMT enzyme activity was found to be decreased in B

lymphoblastoid cell lines derived from patients with ASMT mutations. Additionally, a

study of clofarabine-resistant lymphoblastic leukemia cell lines revealed that melatonin

treatment induced cytotoxic effects, and resulted in increased total H3 and H4

acetylation in these cells 43 . These findings indicate that melatonin can induce cytotoxicity

Figure 1.3. Overview of melatonin synthesis. Melatonin is synthesized from tryptophan. The enzymes catalyzing each synthesis step are shown to the left of the arrows. The PAR1 gene ASMT catalyzes the final step in melatonin synthesis to convert N-Acetyl-Serotonin to melatonin.

of leukemia cells, and loss of function mutations or haploinsufficiency of ASMT may

reduce melatonin levels and enhance cell survival.

DHRSX

In normal cells, autophagy is a critical regulator of cellular homeostasis.

Autophagy is a lysosome-mediated mechanism to degrade intracellular components and it can act as a tumor-suppressive mechanism, but also a chemo-resistance mechanism 44 . It has been reported that during early stages of cellular transformation, autophagy is disrupted or suppressed, which results in increased genomic instability 44 .

Dehydrogenase/reductase (SDR family) X-linked ( DHRSX, DHRSY, DHRSXY,

DHRS5X, DHRS5Y, SDR7C6, CXORF11, SDR46C1 ) is a member of the

oxidoreductase enzyme family. Ectopic expression of DHRSX increased starvation-

induced autophagy, and knockdown of DHRSX reduced autophagy in HeLa and U2OS

cells 45 . Therefore, deletion or loss of DHRSX due to LOS could reduce autophagy and

leave cells vulnerable to genomic insults that promote cellular transformation.

1.3.2. Proto-oncogenes

Some PAR genes have also been reported to possess proto-oncogenic properties in the hematopoietic system. Interestingly, these genes encode cytokine receptors.

CRLF2

CRLF2 (TSLPR ) encodes the cytokine receptor-like factor 2, a type I cytokine receptor. It forms a heterodimer with IL7R (Interleukin 7 Receptor) to bind to its ligand, thymic stromal lymphopoietin (TSLP), and activates JAK-STAT signaling 46 . Deletions in the PAR1 region have been detected in various types of ALL and results in the fusion of

Figure 1.4. Schematic of the P2RY8-CRLF2 rearrangement. Schematic of the affected genomic region upon the P2RY8-CRLF2 rearrangement. This genomic deletion results in the fusion of the P2RY8 promoter to the CRLF2 gene, which results in upregulation of CRLF2 . Additionally, the coding region of P2RY8 is deleted, along with ASMTL , SLC25A6 , IL3RA , and CSF2RA .

the noncoding region of exon 1 of P2RY8 to the CRLF2 gene, which causes CRLF2 expression to be driven by the P2RY8 promoter 47-49 (Figure 1.4.). This fusion has been reported to increase CRLF2 expression by up to 10-fold, and results in increased activation of downstream signaling pathways, such as the JAK-STAT pathway, that promote proliferation of B cell precursors 50 . B-ALL patients with JAK2 mutations also frequently overexpress CRLF2 , which leads to cytokine-independent growth 47 .

Additionally, patients with high expression of CRLF2 have poor prognoses. However, the

deletion leading to the fusion of P2RY8 and CRLF2 also results in the deletion of several

genes ( P2RY8 , CSF2RA , IL3RA , SLC25A6 , and ASMTL ) whose potential roles as tumor

suppressors should not be discounted.

IL3RA

The other proto-oncogenic receptor is the Interleukin 3 receptor alpha ( IL3RA or

CD123), which is highly expressed on monocytes, macrophages, alveolar macrophages,

and dendritic cells 42 . The IL-3 receptor is composed of the CD123 alpha receptor and the CD131 common GM-CSF beta receptor. Although IL-3 signaling promotes differentiation of hematopoietic stem/progenitor cell, it also stimulates proliferation of myeloid cells and is often used as a growth factor for myeloid cell lines 51 . CD123 has

been found to be highly expressed in a variety of hematological malignancies, including

AML, myelodysplastic syndrome, Hodgkin’s lymphoma, hairy cell leukemia, systemic

mastocytosis, plasmacytoid dendritic cell (pDC) leukemia, chronic myeloid leukemia, and

B-cell acute lymphoblastic leukemia 52-54 . Therefore, CD123 has become an attractive hematopoietic cancer cell biomarker to target. Anti-CD123 therapies, such as monoclonal antibodies and chimeric antigen receptor (CAR) T-cells, have shown

promise in mouse models, and are now being investigated clinically for therapeutic

efficacy in various hematological malignancies 54-56 .

1.3.3. Context-dependent pseudoautosomal region genes

Several of the PAR genes cannot be as readily classified as tumor suppressors or proto-oncogenes, due to conflicting discoveries in varying cellular contexts.

CSF2RA

Colony stimulating factor 2 receptor alpha ( CSF2RA, GMRα, or CD116 ) is one of the receptors for granulocyte macrophage-colony stimulating factor (GM-CSF or GM).

The GM receptor is an oligomer of the alpha and beta ( CSF2RB , GMRβ, or CD131) receptors (Figure 1.5.). The alpha subunit binds with high affinity to GM and confers specificity of signaling. The beta subunit of GM also serves as the beta receptor for the

IL-3, as aforementioned, and IL-5 (Interleukin 5) receptors, which both have unique alpha receptors that provide signaling specificity 57 . GM signaling regulates various

cellular functions including cell proliferation, survival, and differentiation. Increased levels

of CSF2RA and GM signaling have been observed in hematologic malignancies such as

CMML (chronic myelomonocytic leukemia) and JMML (juvenile myelomonocytic leukemia) 58, 59 . Leukemia cells from JMML and CMML also frequently exhibit

hypersensitivity to GM and harbor oncogenic RAS, which results incontinuous activation

of the RAS-RAF-MEK-ERK signaling pathway 58 . Studies with an Nras G12D/+ driven mouse

model for CMML demonstrated that eliminating GM signaling via knockout of the GM

beta receptor was insufficient to inhibit leukemia development 60 . However, eliminating

GM signaling prolonged the survival of mice and reduced splenomegaly of NRas G12D/+

mice, and reduced colony formation of their hematopoietic cells. These

Figure 1.5. The GM-CSF receptor complex and its downstream signaling pathways The GM-CSF receptor is a dodecamer complex composed of 4 GMRα receptor subunits (green), 4 GMRβ receptor subunits (purple), and 4 GM-CSF molecules (orange). Upon GM-CSF ligand binding and receptor oligomerization, the juxtaposed GMRβ subunits bound to JAK2 promotes receptor transphosphorylation to initiate activation of various signaling pathways (grey) that regulate cellular proliferation, survival, and differentiation.

findings suggest that GM signaling provides mitogenic signals in CMML and JMML, but

indicates there are other oncogenic drivers, such as mutant RAS, responsible for

leukemic transformation.

Conversely, LOH of CSF2RA , along with SHOX and CRLF2 , due to deletion in a

portion of the PAR has been reported in mantle cell lymphoma 27 . As mentioned

previously, the t(8;21) requires additional cooperating mutations for leukemogenesis,

and roughly 32-59% of t(8;21) patients also display LOS 20, 21 , suggesting that genes on

the sex chromosomes, such as CSF2RA , may act as tumor suppressors. In fact, our

group previously reported that loss of GM signaling cooperated with RUNX1-ETO, the

oncofusion protein generated from t(8;21), to induce AML in a murine model 61 . Together,

these data support that GM signaling has negative effects in t(8;21) AML development,

and substantiates that reduced CSF2RA levels due to LOS may cooperate with t(8;21) in

leukemogenesis.

P2RY8

Another PAR gene with context-dependent functions is P2RY8 . This gene

encodes the Purinergic receptor P2Y8 ( P2RY8 ), an orphan receptor that is a member of the G-protein coupled receptor family. This family of receptors is activated upon binding of adenosine and uracil nucleotides. P2RY8 is highly expressed in hematopoietic cells, including granulocyte-macrophage progenitor cells, B cells, and T cells 42 . Additionally,

P2RY8 has been found to be upregulated in leukemias and expression of P2RY8 in 3T3

cells resulted in their ability to form tumors in vivo , suggesting it may serve as a proto-

oncogene 62 .

Contrarily, loss of function mutations of P2RY8 are frequently detected in

Burkitt’s lymphoma and lymphomas derived from germinal center B cells. Five of the six

mutations identified in one study resulted in loss of P2RY8 expression at the cell

surface 63 . Additionally, ectopic expression of P2RY8 was demonstrated to suppress germinal center B cell proliferation. Therefore, P2RY8 may aid in suppressing B cell proliferation. In the case of the aforementioned P2RY8 -CRLF2 rearrangement, where the coding sequence of P2RY8 is deleted, it may consequently result in deletion of a potential tumor suppressor 47 (Figure 1.4.). Given that the regulatory element of P2RY8 is highly active in lymphoid cells, it is fitting that the P2RY8-CRLF2 rearrangement, which places the proto-oncogenic CRLF2 under control of the P2RY8 regulatory element, is

correlated with lymphoid leukemia.

IL9R

Interleukin 9 Receptor ( IL9R or CD129) heterodimerizes with interleukin 2

receptor gamma (IL2R) to form the receptor for its ligand interleukin 9 (IL9). IL9 is a

cytokine that stimulates proliferation and inhibits apoptosis. It is also a growth factor for

several hematological cell types, including T cells, mast cells, and leukemia cell lines 64 .

IL9R is most highly expressed on T cells, and has been found to be upregulated in various hematological malignancies, including B and T cell leukemias, B cell lymphomas, and AML 42 . Although IL9R generally forms heterodimers with IL2R, IL9R has also been found to form homodimers that interact with and activate mutant JAK1, which is common in ALL, to further activate its downstream STAT targets 65 . Contrarily,

IL9 has been reported to activate tumor immunity. In a pulmonary melanoma model, it

was found that neutralization of IL9 resulted in increased tumor growth and reduced

leukocyte infiltration in the tumors 66 . In another study of mice bearing B16F10

melanoma, homozygous knockout of IL9RA resulted in increased tumor growth, and administration of IL9 to wildtype tumor-bearing mice reduced tumor growth 67 . Altogether,

these findings suggest that IL9 signaling promotes activation of immune cells to

recognize cancer cells and has tumor suppressive functions. In hematopoietic cells with

LOS, reduced IL9R levels could hinder activation of an antitumor response, thus allowing cancer cells to escape from immunosurveillance.

CD99

CD99 (MIC2 , HBA71 , MSK5X ), also known as the E2 antigen, is a glycoprotein

that is expressed across the hematopoietic system and its expression decreases with

differentiation. CD99 is most highly expressed on thymocytes where it regulates T-cell

adhesion and apoptosis 68 . CD99 has been found to be upregulated in acute

lymphoblastic leukemias (ALL) 68, 69 . Incubation with CD99 antibody has been reported to

trigger apoptosis of the Jurkat cell line, thymocytes, TEL/AML1-positive ALL cells, and

normal B cell precusors 69, 70 . Additionally, binding of specific epitopes of CD99 has been

found to promote cell death of transformed T cells via a caspase-independent pathway 71 .

Although CD99 overexpression is observed in various hematologic malignancies, CD99

haploinsufficiency could also perturb normal T cell apoptosis and cell adhesion and

result in increased T cell survival and mobility.

The conflicting findings for these PAR genes stress the importance of cellular

context when studying the functions of PAR genes in hematological malignancies.

1.3.4. Inactive pseudoautosomal region genes

Although most PAR genes escape inactivation, two PAR genes, SPRY3 and

VAMP7 , are located in the PAR but are silenced on the inactive X chromosome in

females and the Y chromosome in males, which results in monoallelic expression. The

mechanisms of inactivation of the two genes vary. Treatment of cells with DNA

methylation inhibitors enhanced VAMP7 , but not SPRY3 expression, suggesting that

promoter methylation is responsible for inactivating VAMP7 , but not SPRY3 72, 73 . Instead,

histone modifications appear to be responsible for the maintenance of SPRY3

repression, as treatment with histone deacetylase (HDAC) inhibitors reactivated its

expression 73 . SPRY3 (sprouty RTK signaling antagonist 3) encodes one of the four mammalian Sprouty proteins (SPRY1-4), which were originally identified in Drosophila to

function as inhibitors of MAPK signaling 74 . However, very little is known about SPRY3.

Because SPRY3 is a monoallelically expressed PAR gene, its expression should not be affected by LOS.

Vesicle associated membrane protein 7 ( VAMP7, SYBL1, TIVAMP ) is found in

the membranes of late endosomes and lysosomes. VAMP7 is a member of the soluble

NSF attachment protein receptor (SNARE) family. VAMP7 aids in the fusion of vesicles

to their target membranes and is required for the formation of mature

autophagosomes 75 . Knockdown of VAMP7 reduced autophagosome formation in starvation conditions and resulted in an accumulation of autophagosomes due to impaired ability to fuse to lysosomes 76 . VAMP7 has also been reported to be required for the release of cytotoxic granules in natural killer (NK) cells 77 , and controls T cell activation by regulating the transport and docking of vesicles containing critical adaptor proteins for T cell activation to T cell receptor-activation sites 78 . VAMP7 has various functions that could be regarded as tumor suppressive; therefore, aberrant epigenetic silencing of the active allele could be involved in cellular transformation.

1.3.5. Other less characterized pseudoautosomal region genes and regions

As the remaining PAR genes are not as well characterized, there exists great potential for novel functional studies of these genes in hematopoietic development and diseases. In a recent report attempting to identify tumor suppressors in malignant

melanoma, PLCXD1 (Phosphatidylinositol-specific phospholipase C, X domain

containing 1) was one of 7 genes to be methylated or deleted in all 14 melanoma tumor

samples that were examined 79 . Additionally, when PLCXD1 was ectopic expressed, it

reduced proliferation in 2 of the 4 melanoma cell lines that were assessed 79 .

Unfortunately, PLCXD1 remains a relatively uncharacterized gene.

ASMTL is expressed highly in T lymphocytes 42 and codes for the N-

acetylserotonin O-methyltransferase-like protein. ASMTL has two domains with

homology to two different proteins. The N-terminal domain shares roughly 60%

sequence homology to two bacterial proteins: maf of B. subtilis and orfE of E. Coli . The

C-terminal domain is highly homologous to ASMT (acetylserotonin O-methyltransferase),

which is encoded by another PAR gene 80 . Very little has been reported on the function of

ASMTL. However, in a genomic study of high risk childhood B-cell precursor ALL, recurrent point mutations of ASMTL were discovered 81 . One of the identified mutations results in a frameshift mutation at V171, and the other results in an R395H missense mutation.

A-kinase anchoring protein 17A ( AKAP17A, XE7, CCDC133, PRKA17A,

SFRS17A, AKAP-17A, DXYS155E, CXYorf3 ) is highly expressed in T cells, B cells, and dendritic cells 42 . AKAP17A is a protein kinase A anchoring protein (AKAP) and binds to

both type I and II PKA to localize it to subcellular regions 82 . In addition to binding PKA,

AKAP17A has also been found to interact with the splicing factors ZNF265, ASF/SF2,

and SC35, and therefore acts as a regulator of alternative splicing, which is often

perturbed in cancer cells 82, 83 .

Zinc finger BED-containing 1 ( ZBED1 , ALTE , DREF , TRAMP ) is a transcription factor that regulates DNA replication and cell proliferation genes 84, 85 . There have been

no studies of ZBED1 the hematopoietic system. However, knockdown of ZBED1/DREF

in HeLa cells reduced cell proliferation and reduced expression of its target gene FNC16

(histone H1) 85 .

XG or PBDX encodes a blood group antigen that shares 48% sequence

homology with CD99 86 . XG expression is believed to be restricted to red blood cells, and

cells are either Xg(a+) or Xg(a-). In red blood cells of those who are Xg(a-), no CD99

was detected, suggesting that expression of XG is co-regulated with CD99 87 . Further studies confirmed that the expression of these two genes are in fact co-regulated, but only at the transcriptional level 88, 89 . Aside from encoding a blood cell antigen, there are

no other reported functions of this gene.

GTPases have been demonstrated to regulate hematopoietic stem cell functions

and differentiation, and both activating and inactivating mutations have been reported in

hematologic diseases 90 . GTPBP6 (PGPL ) encodes a protein that is 442 amino acids long. Because it contains 4 characteristic GTP-binding sites, it has been characterized as a putative GTP-binding protein 91 . GTPases have been demonstrated to regulate

hematopoietic stem cell functions and differentiation, and both activating and inactivating

mutations have been reported in hematologic diseases 90 . Although GTPBP6 is highly expressed in CD4 T lymphocytes, myoblasts, and erythroid cells, there have been no studies of GTPBP6 in the hematopoietic system 42 . Given the importance of GTP-binding

proteins in regulating cellular functions, additional mechanistic studies on the functions

of GTPBP6 in normal and malignant hematopoietic contexts would be of great value.

In addition to protein coding genes, there are several uncharacterized long intergenic non-protein coding RNAs, including LINC00106, LINC00685, and LINC00102,

and the pseudogene DDX11L16 (DEAD/H box polypeptide 11 like 16), in the PARs 92, 93 .

There are no validated microRNAs (miRNAs) in the PARs.

1.4. Sex chromosome amplifications in cancer

Although we have focused on LOS, sex chromosome amplifications have also been reported in hematological malignancies. Extra copies of the X and Y chromosomes have been detected in chronic neutrophilic leukemia (CNL) that progressed to myeloid blast crisis 94 and AML 95 . Given that some PAR genes do possess potentially oncogenic functions, such as promoting cell proliferation, it is not surprising that sex chromosome polysomies are also observed in hematological malignancies and their roles in oncogenesis should not be overlooked.

1.5. Discussion

Given the high frequency of LOS in hematological malignancies and the shortage of systematic studies of PAR genes in the hematopoietic system, there remains the potential to uncover novel functions for PAR genes, which could be targeted therapeutically. Importantly, elucidating the role of haploinsufficiency of PAR genes in the development of hematological malignancies and chemoresistance could reveal a novel role for LOS as a predictive biomarker for patient prognoses and therapeutic response. Studies clarifying the mechanisms of action of PAR genes may also provide valuable insight into viable combination therapies that would be especially beneficial for patients exhibiting LOS compared to those who do not. Altogether, mechanistic studies on the effects of LOS and haploinsufficiency of PAR genes are currently lacking and may greatly assist in opening the door to alternative therapeutic strategies for patients suffering from hematological malignancies.

1.6. Objectives of the dissertation

Due to the shortage of studies on the involvement of LOS and PAR genes in the development of hematological malignancies, the objective of this dissertation was to mechanistically examine the consequence of haploinsufficiency of the PAR gene

CSF2RA , and subsequent reduction of its downstream GM signaling, in the pathogenesis of t(8;21) AML. The following works center on elucidating the mechanisms mediating the inhibitory effects of GM signaling on t(8;21) leukemogenesis through (1) gene expression profiling GM-treated RUNX1-ETO hematopoietic stem/progenitor cells,

(2) a functional screen of GM-induced genes capable of reducing the self-renewal potential of RUNX1-ETO hematopoietic stem/progenitor cells, (3) the identification of a therapeutically targetable pathway in t(8;21) cells, and (4) elucidation of the mechanisms involved in the downregulation of CSF2RA in t(8;21) cells.

1.7. Acknowledgements

S.W. prepared figures.

CHAPTER 2:

Multiple GM-CSF-induced genes in RUNX1-ETO hematopoietic stem/progenitor

cells function to inhibit self-renewal potential

26 27

2.1. Introduction

The t(8;21)(q22;q22) is one of the most commonly observed cytogenetic

aberrations in adult AML patients 96, 97 . Although t(8;21) AML is generally associated with favorable prognoses, the 5-year survival rate is around 50% and patients frequently relapse, implying a need for alterative therapeutic strategies for targeting t(8;21) AML 98-

100 . The t(8;21) generates the oncofusion protein RUNX1-ETO (also known as RUNX1-

RUNX1T1, AML1-ETO, AML1-MTG8, hereinafter referred to as RE), which contains the

N-terminal DNA binding domain of RUNX1 (AML1), a transcription factor that functions

as a master regulator of hematopoiesis, and the majority of the transcriptional co-

repressor ETO 101 . As a result, RE is able to bind to RUNX1 target sites, but recruits

transcriptional repressors through the ETO domain to suppress expression of RUNX1

target genes that are critical for normal hematopoietic cell development 102, 103 . Although

RE alone is insufficient for leukemogenesis and requires additional secondary mutations,

RE induces an early myeloid differentiation block, enhances self-renewal, and promotes

expansion of hematopoietic stem/progenitor cells (HSPCs), which together increase

leukemic potential 22, 96, 104-106 .

Interestingly, one of the most frequently observed cytogenetic abnormalities in

t(8;21) patients is LOS, which is detected in 32-59% of patients 107 . The high incidence of

LOS in t(8;21) AML suggests that potential tumor suppressor genes may reside on the

sex chromosomes. The similar frequency of X and Y chromosome loss further suggests

they may be located in regions homologous between and expressed by both sex

chromosomes, such as the pseudoautosomal regions (PARs). Upon examining the

expression of PAR genes across AML patient samples, CSF2RA was found to be the most significantly downregulated PAR gene in t(8;21) patients compared to non-t(8;21)

M2 AML patients 108, 109 . Altogether, this suggests that LOS-driven haploinsufficiency of

CSF2RA may aid in promoting pathogenesis of t(8;21) AML.

The granulocyte macrophage colony-stimulating factor (GM-CSF or GM) receptor

is a heterodimeric receptor composed of alpha (GM-CSFRα, GMRα, and CD116) and

beta (GM-CSFRβ, GMRβ, β c, and CD131) subunits, encoded by CSF2RA and CSF2RB ,

respectively 110 . GM-CSFRβ serves as a common beta subunit for the GM-CSF, IL-3, and

IL-5 receptors 110 . Therefore, cytokine signaling specificity is determined by the alpha subunits. Upon GM-CSF binding to the alpha receptor, the ligand-bound alpha receptor binds to the beta receptor 57 . This complex then heterodimerizes to form a hexamer, which binds with high-affinity to ligand, and further oligomerizes to form a dodecamer complex, which activates receptor signaling 57 . GM binding and receptor oligomerization activates a multitude of signal transduction pathways, such as JAK-STAT, MAPK, PI3K, and NFκB pathways, that regulate proliferation, survival, hematopoietic differentiation, and activation of immune responses 110, 111 . Expression of GM and its receptor has been

found to be dysregulated in various leukemias 112, 113 , which suggests GM signaling also

plays a role in regulating leukemia cells.

In t(8;21) cells with LOS, LOS-driven haploinsufficiency of CSF2RA decreases

functional GM receptors, and renders the cells hyporesponsive to GM 114, 115 . We previously reported that GM signaling inhibits leukemogenesis in a t(8;21) murine model by reducing the self-renewal potential of RE hematopoietic stem/progenitor cells

(HSPCs) 61 . Interestingly, RE has also been found to negatively regulate GM

expression 116 . Thus, hyporesponsiveness to GM due to CSF2RA haploinsufficiency and

RE-mediated repression of GM in t(8;21) cells cooperate to allow cells to escape the inhibitory effects of GM signaling, which ultimately promotes pathogenesis. In accordance with this, leukemia cells from t(8;21) patients are hyporesponsive to GM and

29 have reduced GM binding 114, 115 . These observations suggest that in RE HSPCs, certain

GM-induced molecular events mediate the negative effects of GM signaling in preventing leukemic cell transformation. Direct reactivation or restoration of these mechanisms may serve as an alternative therapeutic strategy for treating t(8;21) AML patients, including those who are hyporesponsive to GM.

In this study, we validated that the negative effects of GM on RE HSPCs observed previously in the murine system were also applicable to primary human RE

HSPCs. Additionally, we performed gene expression profiling of primary RE HSPCs in response to GM to identify candidate genes mediating the negative effects of GM on t(8;21) cells. A functional screen of these candidate GM-stimulated genes revealed multiple genes that diminish the self-renewal potential of RE HSPCs. Additionally, when these genes were individually co-expressed with RE, they reduced proliferation of RE

HSPCs and aided them in overcoming their differentiation block.

2.2. Results

2.2.1. GM-CSF induces a human myelopoiesis gene expression profile in RUNX1-

ETO hematopoietic stem/progenitor cells

To gain insight into the molecular mechanisms mediating the inhibitory effects of

GM on leukemic transformation of RE cells, we examined the gene expression profile of control (MIG) and RE-expressing (MIG-RE) HSPCs (Lin-/c-Kit +/GFP +) after 10 ng/mL GM treatment (Figure 2.1.).

Microarray data analysis revealed that only 3 genes were differentially expressed after GM treatment of control MIG HSPCs (Figure 2.2.A.). In contrast, 122 genes were

Figure 2.1. Diagram of experimental methods for gene expression profiling of murine RUNX1-ETO hematopoietic stem/progenitor cells treated with GM. (A) Schematic diagram of the control MIG and MIG-RE retroviral vector constructs. (B) Lineage negative (Lin -) cells were transduced with empty vector control (MIG) or MIG- RE (RE) retrovirus. The following day, cells were washed and cultured in StemSpan SFEM media with or without 10 ng/mL GM for 24 hours. Lin -/c-Kit +/GFP + cells were isolated by FACS and used for microarray analysis.

differentially expressed after GM treatment of RE HSPCs compared to untreated RE

HSPCs, none of which overlapped with the MIG+GM differentially expressed genes. This

indicates that RE HSPCs are hypersensitive to GM. RE expression alone induced

differential expression of 1111 genes (Figure 2.2.B); however, 35 of these 1111 genes

were further differentially expressed upon GM treatment (Figure 2.2.B. and Figure

2.2.C.). Furthermore, GM treatment of RE cells (RE+GM) uniquely affected 87 genes

(Figure 2.2.B. and Figure 2.2.D.), which were unchanged by RE expression alone or by

GM treatment of control cells. Interestingly, the majority of these genes were

upregulated.

We utilized Gene Set Enrichment Analysis (GSEA) to confirm that our RE gene

expression signature significantly correlates with previously published datasets of RE

expression in HSPCs (Figure 2.3.A)117, 118 . Furthermore, GSEA also identified significant correlation of RE+GM differentially expressed genes with the gene expression signatures from a murine myeloid cell line expressing two distinct constitutively active forms of the GM receptor 118 , which confirms these were GM-induced genes (Figure

2.3.B.). The microarray data was validated by qRT-PCR (Figure 2.4.).

Although murine HSPCs were used for gene expression profiling, we also conducted GSEA with our dataset using human myelopoiesis gene expression data published by Ferrari et al 119 . This analysis revealed that the murine RE+GM gene

expression signature displayed significant correlation with that of primary human HSPCs

undergoing myelopoiesis and with monocytes and neutrophils (Figure 2.5.). Importantly,

these human myelopoiesis gene signatures were not observed in control cells treated

with GM. These results reveal the importance of and requirement for RE expression to

enable GM to induce a unique gene expression profile that correlates with human

myeloid differentiation programs. This is particularly noteworthy because RE enforces an

Figure 2.2. Analysis of differentially expressed genes after GM-CSF-treatment of RUNX1-ETO hematopoietic stem/progenitor cells. (A) GM-treated control MIG HSPCs were compared to untreated MIG HSPCs (MIG+GM). GM-treated RE HSPCs were compared to untreated RE HSPCs (RE+GM). (B) RE cells were compared to control MIG cells (RE). GM-treated RE cells were compared to untreated RE cells (RE+GM). A 2-fold differential gene expression cutoff was applied. (C) Unsupervised hierarchical clustering of the 35 common RE and RE+GM genes using the average clustering of Pearson correlation coefficient, depicted as a heat map. (D) Unsupervised hierarchical clustering of the 87 unique RE+GM genes using the average clustering of Pearson correlation coefficient, depicted as a heat map.

Figure 2.3. Gene Set Enrichment Analysis of GM-CSF-treated RUNX1-ETO hematopoietic stem/progenitor cells. (A) Gene Set Enrichment Analysis (GSEA) of the gene expression profile after RE expression with gene sets generated from publicly available data from Liu et al of murine HSPCs expressing RE. (B) GSEA of RE+GM differentially expressed genes using gene sets generated from microarray data of FDB1 cells, a factor-dependent bipotent cell line, expressing constitutively active FIΔ and V449E mutants of the GM beta receptor, which promote differentiation and proliferation, The enrichment score (ES), nominal enrichment score (NES), nominal p-value (NOM p-val), and false discovery rate (FDR) for each gene set is shown.

Figure 2.4. Validation of microarray data Quantitative reverse transcription PCR (qRT-PCR) validation of microarray data. MYC- repressed targets genes, as identified by GSEA and Ingenuity Pathway Analysis (IPA), were also validated. Error bars represent standard deviation (SD) of independent experiments.

Figure 2.5. Gene Set Enrichment Analysis of GM-CSF-treated RUNX1-ETO hematopoietic stem/progenitor cells with a human myelopoiesis dataset. (A) Figure adapted from Ferrari et al . Gene sets for GSEA were generated from class comparisons between CD34 + cells and Monocytes or Neutrophils. The Myelopoiesis gene set was generated from a class comparison of CD34 + cells and the rest of the samples combined. (B) Gene set enrichment analysis (GSEA) of the RE+GM gene expression signature with human myelopoiesis, monocyte, and neutrophil gene sets generated from publicly available GEO dataset GSE12803 . The enrichment score (ES), nominal enrichment score (NES), nominal p-value (NOM p-val), and false discovery rate (FDR) for each gene set are shown.

early myeloid differentiation block in HSPCs and our findings indicate that the presence

of RE sensitizes cells to myeloid differentiation induced by GM.

GSEA also found correlation of RE+GM upregulated genes with C/EBP family

target genes (data not shown), although none of the CEBP s themselves were

upregulated. IPA upstream regulator analysis also predicted CEBPB activation (Table

2.1.). The C/EBP family of transcription factors regulates differentiation of various cell

types, including myeloid cells 121 . These findings further support that the GM signaling in

RE HSPCs induces the expression of genes that aid cells in overcoming their RE-

induced differentiation block.

2.2.2. GM-CSF promotes differentiation and reduces the long-term culture

initiating cell frequency of primary human RUNX1-ETO hematopoietic

stem/progenitor cells

Our previous work revealed that GM signaling inhibits RE-induced

leukemogenesis in the murine system 61 . Therefore, we sought to validate this negative

effect of GM on primary human RE HSPCs. CD34 + cells transduced with MIG control or

RE retrovirus were sorted and confirmed for RE expression (Figure 2.6.). As was

previously reported, we observed that RE expression results in a higher percentage of

immature cells expressing CD34, compared to control (Figure 2.7.A. and 2.8.A.) 122 . In the presence of 10 ng/mL (10GM) or 100 ng/mL (100GM) GM, we detected an accelerated reduction in the percentage of CD34 + RE cells. Meanwhile, GM also

increased the total number of cells in culture as was reported previously (Figure

2.8.B.) 123 . Morphological analysis of cells cultured in methylcellulose showed immature

blast-like cells with RE expression, which were reduced upon GM treatment, indicating

GM relieves the RE-induced early myeloid differentiation block (Figure 2.7.B.). This

lator Analysis. lator Target molecules in dataset dataset in molecules Target CCL3L1/CCL3L3,CCL4,CDKN2A,CXCL2,EDN1,FST, GADD45A,HBEGF,IER3,IFNGR2,IL1A,IL2RA,IL6,IRF 4,LTB4R,MMP10,MMP13,NFKBIZ,P2RY6,PLAU,PLA UR,RND3,SDC4,SERPINB2,THBS1,TNF,TNFRSF11 B,UBD CCL3L1/CCL3L3,CCL4,CDKN2B,HBEGF,IER3,IL2RA ,IL6,IRF4,TNF,UBD GADD45A,GLIPR2,IER3,IL6,MMP10,MMP13,MSH5, NFKBIZ,PLAUR,TNF CDKN2B,CSRP2,EDN1,EMP1,EXOSC7,GADD45A, GATA6,GDI2,GPC1,ID1,IER3,IL1RAP,IRX3,LYZ,PFK M,PLAU,PLAUR,PLS3,S100A10,S100A6,THBS1, TIMP1,TNF,TPM1 AKR1B10,APOC4,APOE,AQP9,BMP2,CAV1,CCL17, AKR1B10,CCL3L1/CCL3L3,CCL4,CDKN2B,CXCL1, ACSL4,ACTN1,ALCAM,C9orf3,CAV1,CDKN2A, of p-value p-value overlap overlap 8.03E-16 8.03E-16 4.01E-05 1.45E-09 3.58E-08 z-score z-score Activation Activation 4.717 4.717 3.101 2.577 -1.323 State State Predicted Predicted Activation Inhibited Inhibited Activated Activated Activated Activated Upstream Upstream Regulator Regulator Table 2.1. Ingenuity Pathway Analysis Upstream Regu Upstream Analysis Pathway Ingenuity Table 2.1. TNF TNF CSF2 CEBPB MYC

Figure 2.6. Validation of RUNX1-ETO expression in CD34 + hematopoietic stem/progenitor cells. (A) Immunofluorescence staining of sorted primary human CD34 +/GFP + control MIG and RE cells for nuclear HA-tagged RUNX1-ETO expression (original magnification, 400X). (B) qRT-PCR of relative RUNX1-ETO mRNA levels after 3 weeks of liquid culture. RUNX1-ETO levels were normalized to RE cells with no GM treatment.

Figure 2.7. GM-CSF promotes differentiation and reduces the long-term culture initiating cell frequency of primary human RUNX1-ETO hematopoietic stem/progenitor cells. (A) Percentage of CD34 + cells, as determined by flow cytometric analysis, of sorted primary human CD34 +/GFP + control and RE cells in liquid culture with 0 ng/mL (Control), 10 ng/mL (10GM), and 100ng/mL (100GM) GM. Error bars represent standard deviation (SD) of technical replicates of a representative experiment. Two additional independent experiments are shown in Figure S4. (B) Wright-Giemsa staining and differential counts of sorted CD34 +/GFP + control and RE cells seeded in methycellulose with 0 ng/mL (Control), 10 ng/mL (10GM), and 100 ng/mL (100GM) GM for 17 days of culture. Error bars represent SD of technical replicates of a representative experiment. (C) Relative limiting dilution LTC-IC frequencies of sorted CD34+/GFP + control MIG or RE cells with increasing concentrations of GM. Frequencies were normalized to control cells with no GM treatment. Error bars represent standard error of the mean (SEM) of 3 experiments (* indicates p < 0.05).

reflects the GSEA finding, which correlated the RE+GM gene expression signature with

human myelopoiesis.

To further investigate the effects of GM on long-term culture-initiating cell (LTC-

IC) frequency, we performed limiting dilution LTC-IC assays, which serves as a

surrogate in vitro assay for quantifying primitive hematopoietic cells that share phenotypic and functional properties with in vivo repopulating stem cells 124 . Sorted primary human control and RE CD34 + cells were seeded for limiting dilution LTC-IC assays and cultured with GM. GM did not affect LTC-IC frequencies of control cells.

However, the LTC-IC frequencies of RE cells were significantly reduced in the presence of GM (Figure 2.7.C.). These results reflect the gene expression profiling data and demonstrate that GM treatment in the presence of RE also elicits a unique effect on primary human RE cells, which ultimately aids them in overcoming their RE-induced myeloid differentiation block and reduces their LTC-IC frequency.

2.2.3. Characterization of the GM-CSF-induced gene expression profile in inhibiting self-renewal potential of RUNX1-ETO hematopoietic stem/progenitor cells

We sought to identify GM-induced mechanisms of reducing leukemic potential of

RE HSPCs with aims to activate or restore them as an alternative method of inducing a

GM-like response in t(8;21) cells. RE expression enhances self-renewal potential and confers serial replating ability to HSPCs in vitro 104, 105 . Our previous studies demonstrated that the serial replating ability of RE HSPCs is inhibited by GM 61 .

Therefore, we conducted a functional screen to identify GM-induced genes capable of

inhibiting the serial replating ability and reducing the self-renewal potential of RE cells.

Many of the GM-induced genes in RE HSPCs did not exhibit dramatic upregulation,

Figure 2.8. CD34 + percentages and proliferation of primary human hematopoietic stem/progenitor cells cultured with GM-CSF. (A) CD34 + percentages of sorted primary human CD34 +/GFP + control MIG and RE cells derived from two additional independent cord blood samples cultured with 0ng/mL, 10ng/mL (10GM), and 100ng/mL (100GM) GM. Error bars represent SD of representative experiments. (B) Proliferation of sorted primary human CD34 + /GFP + control MIG and RE cells cultured with 0 ng/mL, 10 ng/mL (10GM), and 100 ng/mL (100GM) GM. Cross indicates cell death. Error bars represent SD from one representative experiment of three independent experiments.

42 suggesting that the modest but concerted upregulation of a group of genes may be cooperatively functioning to mediate the negative effects of GM on RE HSPCs.

However, we aimed to identify individual genes capable of reducing the self-renewal potential of RE HSPCs. Pathway analysis assisted in the selection of 10 genes of interest (See Materials and Methods and Figure 2.9. for details), and a barcoded complementary DNA (cDNA) mini-library was generated to screen multiple genes simultaneously.

Each cDNA was cloned into the MIG vector, along with a common primer sequence and a cDNA-specific barcode (Figure 2.10.A.). Murine HSPCs were co- transduced with puromycin resistance MIP-RE retrovirus and control MIG or a pool of barcoded MIG-cDNA retroviruses. After selection of co-transduced cells with puromycin drug selection and sorting for GFP expression, cells were serially replated for 8 weeks. A subset of cells was saved just after sorting (T0), midway at 4 weeks (T4), and at the final timepoint of 8 weeks (T8) (Figure 2.10.B.). Retroviral integration of the vector into genomic DNA allows for PCR amplification of the cDNA-specific barcode region from purified genomic DNA using common primers, followed by next-generation sequencing of the resulting PCR products to identify and quantify barcodes present at each timepoint.

As expected, control cells lost replating ability, whereas cells transduced with RE or RE + cDNAs continued replating (Figure 2.10.C.). Morphological analysis of cells after the first replating indicated that co-expression of RE and the pool of cDNAs resulted in enhanced myeloid differentiation when compared to RE alone. However, because the

RE + cDNAs and control RE colony numbers were relatively similar, this suggests there existed cDNAs in the pool that do not elicit inhibitory effects on the self-renewal capacity of RE cells. RE cells expressing a cDNA that reduces self-renewal capacity and/or

Figure 2.9. Bioinformatic analysis of differentially expressed genes in GM-CSF- treated RUNX1-ETO hematopoietic stem/progenitor cells. (A) Heat map of the 48 genes IPA identified to be involved in regulating hematopoiesis, hematological system development and function, cell growth and proliferation, and cell death. The IPA functional categorization of each gene is also shown. (B) GSEA of RE+GM differentially expressed genes significantly correlated with (A) NFKB signaling 14 , (B) upregulated genes during B lymphocyte late differentiation 15 , (C) upregulated genes upon LPS stimulation 16 , (D) upregulated genes after TNF treatment 17 . The enrichment score (ES), nominal enrichment score (NES), nominal p-value (NOM p-val), and false discovery rate (FDR) for each gene set is shown.

Figure 2.10. Barcoded cDNA screen to identify GM-CSF-induced genes that inhibit self-renewal potential of RUNX1-ETO hematopoietic stem/progenitor cells. (A) Schematic of genomic DNA from cells transduced with barcoded cDNA retrovirus. A common forward universal primer and IRES reverse primer are utilized to PCR amplify the barcode region from transduced cells. (B) Diagram outlining the serial replating cDNA screen. Mouse Lin - HSPCs were transduced with control MIP or MIP-RE, selected with puromycin, and co-transduced with control MIG or a pool of barcoded cDNAs (MIG- cDNAs). Lin -/GFP + cells were sorted and seeded for serial replating. Cells were saved after sorting as the initial timepoint (T0), and GFP + cells were sorted and saved after 4 weeks (T4) and 8 weeks (T8) of replating for quantitative barcode analysis. (C) Colony numbers of control, RE, and RE + cDNAs cells over the course of 8 weeks. Wright- Giemsa staining of cells after the first plating (original magnification, 400X).

induces differentiation or apoptosis would be expected to have a disadvantage in serial replating. Consequently, these cells would be less abundant or absent at later timepoints and fewer of those cDNA barcodes would be detected over the course of the experiment. Quantification of the barcodes present at each timepoint revealed that 6 of the 10 cDNAs displayed a statistically significant dropout by T8 (Figure 2.11.A. and

Table 2.2.). Cdkn2a and Cdkn2b , two well-established tumor suppressors that inhibit cell cycle progression, demonstrated significant dropout125 . Bmp2 , Cxcl1 , Ltb4r1 , and Mxi1 , also significantly dropped out.

2.2.4. Validation and characterization of candidates from screen to identify GM-

CSF-induced genes that reduce self-renewal potential of RUNX1-ETO hematopoietic stem/progenitor cells

We selected a few of the cDNAs that displayed significant dropout in our screen for validation. Murine bone marrow was co-transduced with pMSCV-Neo (Control) or pMSCV-RE-Neo (RE) and MIP (Control) or individual MIP-cDNA retroviruses and subjected to serial replating (Figure 2.11.B.). RE cells co-expressing CXCL1, LTB4R1, and MXI1 completely lost serial replating ability by 8 weeks, indicating that expression of these cDNAs reduces the self-renewal potential of RE cells. Expression of BMP2 did not fully inhibit the serial replating ability of RE cells, but did significantly reduce the number of colonies formed with each replating.

Ltb4r1 encodes the leukotriene B4 receptor 1 (LTB4R1 or BLT1), which is expressed highly on myeloid cells, including monocytes and granulocytes 126 . Leukotriene

B4 (LTB4), the ligand for LTB4R1, is a pro-inflammatory chemoattractant for neutrophils.

LTB4 synthesis and LTB4 receptor expression have been found to increase during 12-

O-Tetradecanoylphorbol-13-acetate (TPA) and 1 alpha,25-dihydroxyvitamin D3-induced

Figure 2.11. Barcode screen reveals genes that reduce RUNX1-ETO hematopoietic stem/progenitor cells self-renewal capacity. (A) Percentage of barcode reads identified by sequencing of GFP + cells saved from T0, T4, and T8, normalized to T0. Percentages are the average of three independent experiments. (B) Serial replating assay of control or RE cells co-transduced with individual cDNAs that displayed significant dropout by T8 from the barcoded screen. Error bars represent SD of technical replicates of a representative experiment.

Table 2.2. Percentage cDNA Barcode Reads Normalized to T0.

cDNA p-value T0 Percentage T4 Percentage T8 Percentage Name (T0 vs. T8) BATF 100 11.90 127.03 0.7720 BMP2 100 48.90 3.95 8.3188E-06 CDH1 100 303.55 127.19 0.7409 CDKN2A 100 0.67 0.21 1.1767E-11 CDKN2B 100 3.40 3.67 1.5793E-06 CXCL1 100 38.16 4.60 2.0129E-05 IRX3 100 965.81 1159.52 0.2024 LTA 100 157.74 229.65 0.4176 LTB4R1 100 39.23 0.14 3.2532E-13 MXI1 100 5.15 8.39 7.7261E-05

differentiation of the human promyelocytic leukemia cell line HL-60, respectively 127, 128 .

This suggests that the LTB4 pathway may be involved in regulating myeloid cell

differentiation 127, 128 . Along these lines, the LTB4 pathway has also been reported to regulate the stemness of hematopoietic stem cells. Culture of primary human CD34 +

cells with leukotriene B4 (LTB4) reduced CD34 expression and promoted differentiation,

which was inhibited upon addition of an LTB4R1 antagonist CP105696 129 . These findings support that activation of the LTB4 pathway aids in reducing the self-renewal potential of hematopoietic stem cells and promotes their differentiation. LTB4R1 expression is also significantly downregulated in AML patient samples compared to normal hematopoietic cell controls and in t(8;21) AML patients compared to non-t(8;21)

AML patients 108, 130 , suggesting that downregulation of signaling through LTB4R1 may advantageous for leukemogenesis (Figure 2.12.A.). Furthermore, from our barcode screen and subsequent validation experiments, we discovered that ectopic expression of

LTB4R1 inhibits the self-renewal potential of RE HSPCs as well.

In addition to promoting differentiation of CD34 + cells, LTB4 has been reported to

promote their proliferation 129 . To address whether ectopic LTB4R1 expression also elicits

these effects on RE cells, we assessed proliferation and differentiation of co-transduced

murine bone marrow cells. Conversely to the increased proliferation of CD34 + cells upon

LTB4 treatment 129 , LTB4R1 expression inhibited proliferation of RE cells (Figure

2.12.B.). As expected, flow cytometric analysis of CD11b, which is expressed on monocytes and macrophages, and Gr-1, which is expressed on granulocytes, indicates that LTB4R1 expression enhances granulocytic differentiation of RE cells, which was also confirmed by morphological analysis (Figure 2.12.C. and D.). Although LTB4R1 expression is effective at reversing various RE-associated phenotypes, the LTB4 pathway is well-known for its role in promoting inflammation 131 . Therefore, the LTB4

Figure 2.12. Ltb4r1 is a GM-CSF-induced gene that mediates its inhibitory effects on RUNX1-ETO hematopoietic stem/progenitor cells. (A) Comparison of the gene expression of LTB4R1 in AML samples from two datasets. (B) Relative cell numbers of primary mouse bone marrow cells co-transduced with empty vector control or RE and empty vector control or LTB4R1 in liquid culture. Cell numbers for each condition were normalized to their respective Day 0 cell counts. Error bars represent SD of one experiment. (C) Flow cytometric analysis of Gr-1 and CD11b expression of cells from (B). (D) Wright-Giemsa staining of cells from (B).

pathway is not an ideal pathway to activate therapeutically. Moreover, though there is an

antagonist for LTB4R1, there is currently no agonist for this receptor.

However, our findings demonstrate that our cDNA barcode screen was effective

at identifying single genes induced by GM-treatment of RE HSPCs that can mediate the

negative effects of GM, through the reduction of self-renewal potential and promotion of

differentiation of RE HSPCs but without its mitogenic effects.

2.3. Discussion

In this study, we validated that GM signaling also elicits negative effects on the

self-renewal potential of primary human RE HSPCs. Although GM enhanced the

proliferation of primary human RE HSPCs, this was paralleled with a reduction in the

percentage of CD34 + RE cells and of RE LTC-IC frequencies. Additionally, GM aided RE

cells in overcoming the RE-induced early myeloid differentiation block and promoted

their differentiation.

Gene expression profiling revealed that RE HSPCs are hypersensitive to GM

treatment. This finding confirms previous reports that RE expression sensitizes cells to

GM, which has been attributed to RE-induced repression of NF1 (Neurofibromin 1), a negative regulator of RAS 61, 132 . Gene expression profiling also revealed that RE

expression induced a 2.5-fold increase in CSF2RB , which may further contribute to the

enhanced GM response. Interestingly, the hypersensitive gene expression profile

displayed by GM-treated RE HSPCs resembles human myelopoiesis and correlates with

the activation of CEBP targets. Importantly, this human myelopoiesis gene expression

profile was not observed in control cells. Therefore, these findings stress the importance

of GM signaling in aiding RE HSPCs in overcoming their differentiation block and

51 reducing their self-renewal potential. Moreover, downregulating GM signaling in t(8;21) cells, through CSF2RA haploinsufficiency driven by LOS for example, is critical for evading negative effects of GM during the leukemic transformation process.

Because GM is known to function as a mitogen, which was observed when primary human RE HSPCs were cultured in the presence of GM, identifying factors mediating the negative effects of GM on RE HSPCs without inducing a mitogenic effect is critical for elucidating novel therapeutic strategies. From the barcoded cDNA screen, we have identified several GM-induced genes that reduce the self-renewal potential of

RE HSPCs without inducing a mitogenic response. Individual validation of a candidate identified by our screen, Ltb4r1 , reveals that this GM-induced gene significantly reduces

RE HSPC proliferation and promotes their differentiation.

GM treatment of RE HSPCs resulted in moderate, but concerted, upregulation of

85 genes. Although in this screen, we only selected 10 of these genes to screen, the other GM-induced genes may also possess tumor suppressive functions, which warrants further investigation and may aid in the identification of novel pathways for therapeutic intervention in t(8;21) AML patients, including those with LOS.

2.4. Materials and Methods

2.4.1. Plasmids

MSCV-IRES-EGFP (MIG) and MSCV-HA-RUNX1-ETO-IRES-EGFP (RE) were previously described 133 . MSCV-IRES-puromycin R (MIP) and MSCV-HA-RUNX1-ETO-

IRES-puromycin R (MIP-RE) are identical to MIG and RE, except the EGFP has been replaced with the puromycin R gene. For co-expression of RE with other cDNAs, pMSCV-

Neo (Control) (Clontech, Mountain View, CA) or pMSCV-RE-Neo (RE) and MIP (Control)

or MIP-cDNA retroviruses were used.

2.4.2. Gene expression profiling

Lin - cells isolated from bone marrow of C57BL/6 mice were transduced with

control (MIG) or RE retrovirus. The subsequent day, cells were washed and treated with

10 ng/mL recombinant murine GM (Peprotech, Rocky Hill, NJ) for 24 hours in StemSpan

serum-free expansion medium (SFEM) (StemCell Technologies, Vancouver, BC). Lin -/c-

Kit +/GFP + cells were sorted using a BD FACS Aria II (BD Biosciences, San Jose, CA) and RNA was isolated with the RNeasy Micro Kit (Qiagen, Hilden, Germany). Total

RNAs from three independent experiments were labeled and hybridized on Mouse Ref-8 v2.0 Expression BeadChips following manufacturer’s protocol (Illumina, San Diego, CA).

RNA quality control and sample preparation for BeadChips were performed at the

UCSD, Biomedical Genomics Core Facility. The microarray data have been deposited in the Gene Expression Omnibus database and are accessible through GEO series

number GSE72567.

2.4.3. Gene expression profiling data analysis

Microarray data normalization was performed using dCHIP 134 invariant set

normalization and filtered for a coefficient of variation defined by standard

deviation/mean values between 0.5 and 1000. Class comparisons were performed with

dCHIP and BRB-Arraytools 135 using a univariate 2-sided t- test requiring a nominal p <

0.05 and p < 0.001, respectively. Class comparisons were performed to identify significantly differentially expressed genes: 1) after GM treatment of control MIG HSPCs

(MIG vs. MIG+GM), 2) after GM treatment of RE HSPCs (RE vs. RE+GM), 3) after RE expression (MIG vs. RE), and 4) after GM treatment of RE HSPCs compared to MIG

HSPCs (MIG+GM vs. RE+GM). Each of the class comparisons generated a list of

differentially expressed genes, which were then compared and overlapped to generate a

list of uniquely differentially expressed genes upon each perturbation, and a 2-fold gene

expression cutoff was applied. Gene set enrichment analysis (Broad Institute,

Cambridge, MA) 136 was performed on raw data using the c2.all.v4.0.symbols gene set

database with 1000 gene set permutations, and required p < 0.01 and FDR < 0.01 to be

considered significant. GSEA was also conducted with gene sets generated from

publicly available datasets from Ferrari et al 119 , Liu et al 117 , and Brown et al 137 . and required p < 0.01 and FDR < 0.01 to be considered significant. Pathway analysis and upstream regulator analysis was performed with Ingenuity Pathway Analysis (Qiagen,

Hilden, Germany). Ingenuity Pathway Analysis (IPA) (Ingenuity, Qiagen Redwood City,

CA) was run with confidence settings including both experimentally observed and highly predicted, on mouse and human species only, and with default settings for the remaining parameters.

2.4.4. Quantitative reverse-transcription PCR (qRT-PCR)

RNA was isolated using the RNeasy Micro Kit (Qiagen) and cDNA was generated using the Superscript III (Life Technologies). Quantitative PCR was done and analyzed with the CFX Connect Real-Time PCR Detection System (Bio-Rad) using the

KAPA SYBR FAST qPCR Kit (Kapa Biosystems, Wilmington, MA) with 0.25 μM of each primer performing 40-50 cycles of 95°C for 5 seconds and 60°C for 30 seconds. Primers

for qPCR are provided in Tables 2.3. and 2.4. The cycle threshold (Ct) values for the no

reverse transcriptase control samples were more than 10 Ct values greater than the

samples with reverse transcriptase (data not shown). The qPCR products were run out

on 1.8% agarose gels to confirm there was no product in the no reverse transcriptase

Table 2.3. Mouse qPCR primers

Mouse Primers Sequence (5’ to 3’) mBATF-For CAGTCAAGAAGGGGAGCCAG mBATF-Rev AGATGAGTCCTGTTTGCCAGG mCAV1-For TGGTCAAGATTGACTTTGAAGATGT mCAV1-Rev CCTTCCAGATGCCGTCGAAA mCDKN2A-For GAGCGGGGACATCAAGACAT mCDKN2A-Rev AGCTCTGCTCTTGGGATTGG mCDKN2B-For AGGACCATTTCTGCCACAGAC mCDKN2B-Rev CTGCCCATCATCATGACCTGGAT mCXCL1-For ACCGAAGTCATAGCCACACTC mCXCL1-Rev CTCCGTTACTTGGGGACACC mGADD45A-For CTGTGTGCTGGTGACGAACC mGADD45A-Rev TCCATGTAGCGACTTTCCCG mGAPDH-For GGTGCTGAGTATGTCGTGGAGTCTA mGAPDH-For AAAGTTGTCATGGATGACCTTGG mGATA6-For AGTCAAAAGCTTGCTCCGGT mGATA6-Rev TGATGCCCCTACCCCTGAG mGRP-For ACTGGGCTGTGGGACACTTA mGRP-Rev ATCCCTTGCAGCTTCTTCCC mIER3-For CACACCATGACTGGCCTGAG mIER3-Rev TGGCGCCGGACCACT mIRX3-For TCTGGGTCCCTATCCAATGTG mIRX3-Rev GGTCCCCGAACTGGTACTG mLTA-For TTCCTGCCTTCGACTGAAACA mLTA-Rev TGTCATGTGGAGAACCTGCTG mLTB4R1-For CCA ACAACTAGACAAGACATTACCTTCTGA mLTB4R1-Rev AGGATGCTCCACACACAAAGCTATTG mMMP13-For TCCCTGCCCCTTCCCTATGG mMMP13-Rev CTCGGAGCCTGTCAACTGTGG mMXI1-For TTAAAGCGGCGACTGGAACA mMXI1-Rev CCACTTCAATCTCCTCTCGCT mNTS-For ACATCCAAGATCAGCAAAGCAA mNTS-Rev CATGTCTCCTGCTTCCTCGG mPLAUR-For ACAGGCTTAGATGTGCTGGG mPLAUR-Rev CAGAGACGTTGAGGTGGGTC mPPBP-For CTCAGACCTACATCGTCCTGC mPPBP-Rev GTGGCTATCACTTCCACATCAG Universal-Fwd-Primer CGAAATGTCCGTTCGGTTGG IRES-Rev-Primer GCAAAGGGTCGCTACAGACGTTG

Table 2.4. Human qPCR primers

Human Primers Sequence (5’ to 3’)

BCR/ABL-For CATTCCGCTGACCATCAATAAG BCR/ABL-Rev GATGCTACTGGCCGCTGAAG hGAPDH-For TCGCTCAGACACCATGGGGAAG hGAPDH-Rev GCCTTGACGGTGCCATGGAATTTG RUNX1/ETO-For TGGAAGAGGGAAAAGCTTCA RUNX1/ETO-Rev ATTGCGTCTTCACATCCACA

control samples and that products were of the correct size in samples with reverse

transcriptase (data not shown).

2.4.5. CD34 + cell isolation, transduction, and culture

The CD34 Microbead Kit (Miltenyi Biotech, San Diego, CA) was used to purify

CD34 + cells from human umbilical cord blood (NDRI, Philadelphia, PA). CD34 + cells were pre-stimulated in StemSpan serum-free expansion medium (SFEM) (StemCell

Technologies, Vancouver, BC) supplemented with 100 U/mL penicillin/streptomycin, 100 ng/mL SCF (Stem Cell Factor), 100 ng/mL FLT3L (Fms-related tyrosine kinase 3 ligand)

, and 100 ng/mL TPO (Thrombopoietin) (Peprotech, Rocky Hill, NJ) for 2 days before undergoing spinoculation at 3000 rpm for 3 hours aset 32°C with retrovirus added at

50% v/v and 4 μg/mL polybrene for 2 consecutive days. CD34 +/GFP + cells were sorted

at the UCSD Cancer Center Shared Resource on a BD FACS Aria II after transduction

with control MIG and MIG-RE retrovirus, and used for subsequent experiments. Cells

were cultured in StemSpan SFEM (StemCell Technologies) supplemented with 100U/mL

penicillin/streptomycin, 50 ng/mL SCF, 50 ng/mL FLT3L, and 50 ng/mL TPO (Peprotech)

and weekly medium changes were administered. When applicable, recombinant human

granulocyte-macrophage colony-stimulating factor (rhGM-CSF or GM) (Peprotech) was

added. Cells were seeded in H4534 (StemCell Technologies) for colony forming assays.

2.4.6. Long-Term Culture-Initiating Cell (LTC-IC) assays

Sorted CD34+/GFP+ cells were seeded at limiting dilutions of 1, 3, 5, 16, 50, and

100 cells per well on MS-5 cells irradiated with 8000 cGy and cultured in MyeloCult

H5100 (StemCell Technologies) with 1μM hydrocortisone (Sigma), 1x Anti-Anti

(ThermoFisher, Grand Island, NY) and the indicated concentrations of GM 124 . Cells were fed with weekly one-half media exchanges for 5 weeks and seeded in methylcellulose

(Myelocult H4534 or H4434, StemCell Technologies). The frequency of LTC-ICs was

scored after 17 days of methylcellulose culture. Stem cell frequencies were calculated

using L-Calc (StemCell Technologies).

2.4.7. Flow cytometry

Primary human hematopoietic cells were incubated with Human F c Receptor

Binding Inhibitor (eBioscience, 14-9161, San Diego, CA) and stained with

allophycocyanin (APC)-conjugated antibody against human CD34 (eBioscience, 17-

0349) for flow cytometric analysis. Antibodies used for mouse lineage staining are listed

in Table 2.5. Cells were analyzed with a BD FACS Canto cytometer (BD Biosciences,

San Jose, CA). For differentiation studies, murine bone marrow cells were incubated

with Mouse F c Receptor Binding Inhibitor (eBioscience, 14-0161-86) and stained with

PE-conjugated anti-mouse CD11b (eBioscience, 14-0112-82) and APC-conjugated anti- mouse Gr-1 (eBioscience, 17-5931-72) for flow cytometric analysis.

2.4.8. Barcoded cDNA screen

IPA was used to identify genes that were upregulated by 2-fold or greater in RE

HSPCs upon GM treatment (RE+GM) that are involved in regulating hematopoiesis, hematological system development and function, cell growth and proliferation, and cell death. These 48 genes were also grouped into functional categories by IPA. Of particular interest were genes involved in pathways with potential to be targeted therapeutically. GSEA revealed enrichment of RE+GM genes with tumor necrosis factor

(TNF) targets, and IPA found TNF to be the most significantly activated upstream regulators of the RE+GM upregulated genes (Table 2.1.). It has been previously reported that TNF induces apoptosis of the t(8;21) Kasumi-1 cell line and promotes differentiation of leukemic blasts 138-140 . Therefore, cytokines and growth factors

Table 2.5. Mouse Lineage Antibodies

Antigen Conjugate Source

CD3 PerCPCy5.5 BD Bioscience, 560527 CD4 PerCPCy5.5 eBioscience, 45-0042 CD8a PerCPCy5.5 eBioscience, 45-0081 CD19 PerCPCy5.5 eBioscience, 45-0193 B220 PerCPCy5.5 eBioscience, 45-0452 CD11b PerCPCy5.5 eBioscience, 45-0112 Gr-1 PerCPCy5.5 eBioscience, 35-5931 Ter119 PerCPCy5.5 eBioscience, 45-5921 IL-7Ra PerCPCy5.5 eBioscience, 45-1271

that are members of the TNF superfamily or targets of TNF were also included.

Additionally, cytokines and growth factors that have synthetic analogs or that are ligands

for receptors with chemical agonists were chosen. Transcriptional regulators, which are

targetable or have not been previously studied in the hematopoietic system, were also

included in the screen. Of the remaining genes, many have been implicated in promoting

leukemia development and may mediate the mitogenic effects of GM, are not readily

targetable, or are not expressed in the hematopoietic system and were therefore not

included. With the 10 candidate genes that were selected, a common primer sequence

and a cDNA-specific 6-nucleotide barcode sequence were added 3’ to each

corresponding cDNA and cloned into the retroviral MIG construct. Individual barcoded

MIG-cDNAs were pooled at an equimolar ratio for transfection. Lin - cells were transduced with MIP or MIP-RE and selected with 2 μg/mL puromycin for 48 hours.

Selected cells were then co-transduced with control MIG or a pool of barcoded MIG- cDNA retrovirus at an MOI ranging from 0.375-0.5. GFP +/Lin - cells (Lineage antibodies

are listed in Table 2.5.) were isolated by fluorescence activated cell sorting (FACS) and

seeded in M3434 (StemCell Technologies) with 0.5 μg/mL puromycin (Sigma, St. Louis,

MO) and 100 U/mL penicillin/streptomycin for serial replating. Based on the

quantification of barcode reads at T0 and the number of cells initially seeded, each

barcoded cDNA was present at an average frequency ranging from 89 to 594 cells

between 3 independent experiments. On average, 15% of initially seeded cells

possessed colony-forming potential, which yielded a range of 13 to 89 colony-forming

clones for each barcoded cDNA. Each subsequent week, 1 x 10 4 cells were replated under the same conditions. GFP + cells from the 4th and 8th replating were sorted and

genomic DNA was isolated using the AllPrep DNA/RNA/Protein Mini Kit (Qiagen).

Barcode regions were PCR amplified using the common forward primer and an IRES

reverse primer. The Nextera XT Index Kit (Illumina, San Diego, CA) was used to

generate a sequencing library. PCR products were gel purified, quantified, and pooled at

an equimolar ratio for MiSeq sequencing (UCSD IGM Genomics Center).

2.4.9. Cell culture

Primary C57BL/6 murine bone marrow was cultured in IMDM supplemented with

10% FBS, 100U/mL penicillin/streptomycin, 2% SCF conditioned media from BHK/MKL

cells stably expressing the secretory form of mouse SCF, and 2% IL-3 conditioned

media prepared from X63Ag-653 myeloma cells. Cells were selected with 1 μg/mL

puromycin and 500 μg/mL G418, when applicable.

2.4.10. Proliferation assays

Viable cells were counted via trypan blue exclusion.

2.4.11. Replating assays

Initially after transduction, 1 x 10 5 transduced primary murine bone marrow cells

were seeded for one week of drug selection in M3134 (StemCell Technologies)

supplemented with 20% bovine serum albumin, insulin, and transferrin (BIT 9500,

StemCell Tech), 15% fetal bovine serum (FBS) 100 U/mL penicillin/streptomycin,

10ng/mL rmIL-3 (Peprotech), 50ng/mL rmSCF (Peprotech), and 10ng/mL rhIL-6

(Peprotech). For selection, 1μg/mL puromycin (Sigma) and 500μg/mL G418 (Sigma)

were used, when applicable. Each subsequent week, cells were resuspended and 1 x

10 4 cells were replated with half the aforementioned drug concentrations.

2.4.12. Microscopy

Cytospin slides were fixed in methanol and stained with Wright-Giemsa (Sigma) for cytological analysis. For immunofluorescence staining, slides were fixed in methanol and

61 permeabilized with phosphate-buffered saline (PBS) containing 0.1 % Triton X-100. After blocking with PBS containing 0.1 % Triton X-100 and 2% BSA, and anti-HA antibody

(Sigma, H9658) were added. Slides were washed 3 times with PBS containing 0.2%

Tween20 and incubated with PBS containing 0.1 % Triton X-100, 2% BSA, secondary antibody conjugated to Alexa-568 (ThermoFisher , A-11004) and Hoescht 33342 stain

(ThermoFisher, H3570) then washed with PBS containing 0.2% Tween-20. All images were acquired with an Olympus BX51 microscope, a DP71 digital camera, and the DP-

BSW acquisition software (Olympus, Center Valley, PA).

2.4.13. Statistical analysis

Statistical significance was determined from adequately powered sample sizes of similar variation using two-tailed unpaired Student’s t-tests and was defined as P < 0.05.

Sample sizes are given in figure legends.

2.5. Acknowledgements

Chapter 2 includes, in part, published material. Weng S, Matsuura S, Mowery

CT, Stoner SA, Lam K, Ran D, Davis AG, Lo MC, Zhang DE. “Restoration of MYC- repressed targets mediates the negative effects of GM-CSF on RUNX1-ETO leukemogenicity.” Leukemia, 2016. The dissertation author was the primary investigator and author of this study. S.W. designed and performed the experiments. S.M. performed microarray gene expression profiling at the UC San Diego IGM Genomics Center, La

Jolla, CA. Barcode sequencing was performed at the UC San Diego IGM Genomics

Center, La Jolla, CA. C.M., S.S., K.L., D.R., A.D., and M.L. assisted with experiments.

S.W. analyzed the data and prepared figures. S.W. and D.Z. wrote the manuscript. The

62 authors would like to thank D. Neuberg and K. Stevenson for advising with microarray data analysis, O. Harismendy and K. Jepsen for their assistance with the barcode sequencing and data analysis, and D.J. Young for flow cytometry expertise.

CHAPTER 3:

Attenuation of MYC reduces the leukemic potential of RUNX1-ETO cells

63 64

3.1. Introduction

The MYC proto-oncogene is a frequent target of activating mutations and is

commonly upregulated in cancers 141 . MYC encodes a helix-loop-helix transcription factor

that requires its obligatory partner MYC-associated factor X (MAX) to form the MYC-

MAX heterodimer to bind DNA at specific sequences. Genome-wide analyses of MYC

occupancy, coupled with gene expression profiling, have revealed that MYC functions

both as a transcriptional activator and repressor. Its target genes are involved in

regulating a multitude of cellular functions such as cell cycle, self-renewal, metabolism,

DNA damage response, and differentiation. Therefore, increased MYC levels and

activity, and consequent dysregulation of its critical target genes, drive cellular

transformation.

Expression of MYC is tightly regulated during hematopoiesis. Its expression is highest in hematopoietic stem and progenitor cells (HSPCs) where it functions to maintain self-renewal capacity and promote proliferation 142 , and decreases during myeloid differentiation 143 . Downregulation of MYC to alleviate its repression on key target

genes, such as CEBPA and GADD45A , is critical for initiating hematopoietic

differentiation and apoptosis, respectively 144, 145 . Previous studies have demonstrated that RE expression enhances MYC expression and knockdown of RE in the t(8;21)

Kasumi-1 cell line results in reduced MYC levels 146, 147 , which suggest that MYC

upregulation may be involved in promoting and maintaining t(8;21) leukemogenesis.

MAX, in addition to heterodimerizing with MYC, also heterodimerizes with MAD

family proteins, which include MAD1, MAD3, MAD4, and MXI1. MAX-MAD heterodimers

oppose MYC activity by displacing MYC from its target sites and recruiting transcriptional

co-repressors, such as nuclear receptor co-repressor 1 (N-CoR), and repressors, such

65 as transcription regulator family member A (Sin3A), and histone deacetylase 1 (HDAC1), to repress MYC-activated target genes 148, 149 . Therefore, the ratio of MYC-MAX and

MAX-MAD complexes, which is determined by competitive binding of MAX with either

MYC or MAD proteins, is contingent on the abundance of each of these proteins and largely influences the cellular transcriptional profile. MAX is ubiquitously and stably expressed in cells. However, during differentiation MYC expression decreases and the expression of MAD family genes, such as MXI1 ( MAX interacting protein-1), increases to favor MAX-MAD complex formation and inhibit MYC.

Interestingly, gene expression profiling of RE HSPCs and GSEA revealed that

MYC activity is attenuated upon GM treatment. This finding provides greater insight into the molecular mechanisms that may drive reduced self-renewal potential and increased differentiation of GM-treated RE HSPCs. Moreover, Mxi1 expression was also found to be induced by GM in RE HSPCs, and Mxi1 was identified as a candidate gene that reduces self-renewal potential of RE HSPCs from our screen. Altogether these data suggest that Mxi1 may mediate the negative effects of GM on RE-associated phenotypes and MYC inhibition may serve as an effective strategy for reducing the leukemogenicity of RE HSPCs. Given the well-established role of MXI1 in inhibiting MYC activity, and the previously reported activation of MYC by RE, we explored whether MXI1 expression and MYC knockdown are sufficient to reduce the leukemic potential of RE

HSPCs. Our data demonstrates that RE HSPCs and t(8;21) cell lines are highly sensitive to MYC inhibition, which supports that MYC inhibition may provide an effective therapeutic strategy for treating t(8;21) AML patients. To explore the feasibility of using small molecules to inhibit MYC in t(8;21) cells, we utilized JQ1, an inhibitor of BET

(bromodomain and extraterminal domain)-containing proteins including BRD1-4, which has been reported to reduce MYC expression in various cell types. Therefore, we

examined whether JQ1 can effectively inhibit MYC expression in t(8;21) cell lines and assessed the consequences of JQ1 treatment.

In addition to examining the effects of Mxi1 on the MYC pathway, we also investigated the effects of MXI1 expression on the transcriptional activity of RE. RE is generally believed to repress the expression of critical RUNX1 target genes via the recruitment of transcriptional co-repressor, N-CoR, and repressors, Sin3A and histone deacetylases (HDACs), to RUNX1 target genes through its ETO domain 150 . As aforementioned, MXI1 also binds these transcriptional co-repressors and repressors to inhibit gene expression. Therefore, we investigated whether MXI1 is capable of sequestering transcriptional repressors away from RE and relieve the RE-induced repression of RUNX1 target genes.

Overall, we discovered that RE cells are highly dependent on MYC for survival and proliferation, which highlights the importance of targeting the MYC pathway during t(8;21) AML treatment. We also found that MXI1 expression relieves RE-induced repression of RUNX1 targets; however this does not appear to be mediated by MXI1 sequestering Sin3A from RE.

3.2. Results

3.2.1 MYC-associated gene signatures are attenuated in GM-CSF-treated RUNX1-

ETO hematopoietic stem/progenitor cells

We utilized GSEA and Ingenuity Pathway Analysis (IPA) to identify significantly altered pathways in GM-treated RE HSPCs. One significantly altered pathway in these cells was the MYC pathway. GSEA revealed that GM treatment of RE HSPCs restores

the expression of MYC-downregulated targets from two independent data sets, which

suggests that GM partially attenuates MYC-associated gene signatures in RE cells

(Figure 3.1.A.) 144, 151 . Additionally, IPA upstream regulator analysis predicted MYC to be

inhibited in RE HSPCs treated with GM (Table 2.1.). Interestingly, the microarray data

revealed that Myc expression was unaffected after GM treatment of RE HSPCs.

Therefore, the upregulation of MYC-repressed targets to reduce leukemic potential of

GM-treated RE HSPCs appears to occur independently of MYC downregulation.

3.2.2 MXI1 expression in RUNX1-ETO hematopoietic stem/progenitor cells and

t(8;21) cell lines reduces leukemic potential

The aforementioned pathway analyses converged on attenuated MYC gene

signatures, suggesting MYC as a critical regulator of RE leukemic potential. Therefore,

we performed additional studies with Mxi1 , a GM-induced gene in RE HSPCs that also displayed significant dropout in our screen. MXI1 functions as a MYC inhibitor through its competitive binding to the obligatory MYC binding partner MAX to interfere with the

MAX-MYC heterodimerization necessary for MYC transcriptional activity 152 .

Co-expression of MXI1 with RE in primary murine BM cells demonstrated that

MXI1 significantly inhibited RE cell proliferation and enhanced monocyte and macrophage differentiation (Figure 3.1.B. and C.) and induced apoptosis (Figure 3.1.D. and E.). Kasumi-1 and SKNO-1, the only established t(8;21) cell lines, are chemoresistant and suffer from LOS 153, 154 . Although Kasumi-1 cells have also been found to be hyporesponsive to GM 155 , expression of MXI1 in these cells reduced

proliferation and induced apoptosis (Figure 3.2.A and B.), which indicates they remain

sensitive to MYC inhibition. MXI1 had similar effects in SKNO-1 cells, though to a lesser

extent. These effects were not observed in the non-t(8;21) myeloid cell lines U937 and

Figure 3.1. GM-CSF treatment of RUNX1-ETO hematopoietic stem/progenitor cells attenuates MYC-associated gene signatures and MXI1 expression reverses RUNX1-ETO-associated phenotypes. (A) GSEA of RE+GM differentially expressed genes significantly correlated with two MYC downregulated targets gene sets. The enrichment score (ES), nominal enrichment score (NES), nominal p-value (NOM p-val), and false discovery rate (FDR) for each gene set are shown. (B) Relative cell numbers of primary mouse bone marrow cells co- transduced with empty vector control or RE and empty vector control or MXI1 in liquid culture. Cell numbers for each condition were normalized to their respective Day 0 cell counts. Error bars represent SEM of 3 independent experiments. (C) Wright-Giemsa staining of Day 4 cells. Montage images were made for some due to low cell numbers and density on cytospin slides. (Original magnification, 400X). (D) Flow cytometric analysis of the percentage of Annexin V + primary mouse bone marrow cells co- transduced control or RE and control and MXI1 during liquid culture. Annexin V + percentages were normalized to control cells. Error bars represent SEM of 3 independent experiments (* indicates p<0.05, # indicates p<0.01). (E) Representative flow cytometric analysis of Annexin V + and 7AAD + populations of cells from (B) at Day 2.

Figure 3.2. MXI1 expression inhibits RUNX1-ETO-associated phenotypes in t(8;21) AML cell lines. (A) Relative cell numbers of and (B) flow cytometric analysis of the percentage of Annexin V + Kasumi-1 and SKNO-1 cells transduced with control or MXI1 in liquid culture (bottom). Cell numbers for each condition were normalized to their respective Day 0 cell counts. Error bars indicate SEM of 3 independent experiments (* indicates p < 0.05, # indicates p < 0.01). (C) Relative cell numbers of and flow cytometric analysis of the percentage of Annexin V + U937 and K562 cells transduced with control or MXI1 in liquid culture. Cell numbers for each condition were normalized to their respective Day 0 cell counts. Error bars indicate SEM of 3 independent experiments (* indicates p < 0.05, # indicates p < 0.01). (E) Relative colony forming units (CFUs) and (F) Wright-Giemsa staining of sorted primary human CD34 + cells transduced with RE and MXI1 after 17 days in methycellulose. CFUs were normalized to control cells with no RE or MXI1 expression. Error bars represent SEM of 3 independent experiments (* indicates p < 0.05).

K562 (Figure 3.2.C. and D.). MXI1 expression also specifically reduced the colony

forming ability and promoted myeloid differentiation of primary human CD34 + RE HSPCs

(Figure 3.2.E. and F.).

3.2.3 MYC knockdown reduces the leukemic potential of t(8;21) cell lines

In order to verify that the effects of MXI1 are due to its role in regulating MYC

activity, we knocked down MYC using three independent shRNA sequences in Kasumi-

1, SKNO-1, U937, and K562 cells. Although the t(8;21) cell lines Kasumi-1 and SKNO-1

displayed less MYC knockdown compared to the non-t(8;21) cell lines U937 and K562

(Figure 3.3.A. and B.), they displayed greater increases in apoptosis upon MYC

knockdown (Figure Figure 3.3.C.) and greater reductions in cell proliferation (Figure

3.4.A.), indicating they are highly dependent on MYC for survival and proliferation.

Next, the expression of two critical MYC-repressed targets, CEBPA and

GADD45A , upon MYC knockdown was also examined. C/EBPα is a critical regulator of granulopoiesis and its expression enables hematopoietic progenitors to differentiate.

Importantly, RE also represses CEBPA expression, which results in the early myeloid differentiation block that is characteristic of RE expression 102, 156 . GADD45A has been

reported to function as a tumor suppressor by inhibiting cell cycle progression and

promoting apoptosis 157 , and its methylation has been implicated in poor prognostic outcome in AML 158 . Interestingly, although the t(8;21) Kasumi-1 cells displayed less

MYC knockdown compared to the non-t(8;21) U937 and K562 cells, the expression of

the critical MYC-repressed targets CEBPA and GADD45A was restored in these cells

(Figure 3.4.B.).

Figure 3.3. MYC knockdown induces apoptosis in t(8;21) human cell lines. (A) Western blot analysis of MYC levels in cells expressing non-targeting control shRNAs or 3 unique shRNAs targeting MYC . (B) Quantification of MYC levels by Western Blot upon shRNA-mediated knockdown. Error bars represent SEM of 3 independent experiments. MYC levels are normalized to shCtrl samples. (C) Flow cytometric analysis of the percentage of Annexin V + apoptotic cells upon MYC knockdown (bottom). Error bars represent SEM of 3 independent experiments (* indicates p < 0.05, # indicates p < 0.01).

Figure 3.4. MYC knockdown reduces proliferation of t(8;21) cell lines and and restores MYC-repressed targets in Kasumi-1 cells. (A) Relative cell numbers of Kasumi-1 (top, left), SKNO-1 (bottom, left), U937 (top, right), and K562 (bottom, right) cells upon MYC knockdown. Error bars represent SEM of 3 independent experiments. (B) Quantitative reverse transcription PCR (qRT-PCR) of CEBPA mRNA levels (top) and GADD45A mRNA levels (bottom) upon MYC knockdown at Day 4. The mRNA levels were normalized to the respective controls for each cell line. Error bars represent SEM of 3 independent experiments (* indicates p < 0.05, # indicates p < 0.01).

3.2.4 JQ1 treatment reduces the leukemic potential of RUNX1-ETO hematopoietic

stem/progenitor cells and t(8;21) cell lines

JQ1 is a potent small molecule inhibitor of BET-containing proteins, which

include the BRD1-4 chromatin adaptors. JQ1 has been demonstrated to most strongly

inhibit BRD4. BRD4 has been reported to activate MYC expression, and inhibition of

BRD4 with JQ1 has been demonstrated to downregulate MYC 159, 160 . Additionally, gene

expression profiling of cells treated with JQ1 revealed MYC to be one of the most

downregulated genes 159 . However, tumor cells display varying sensitivity to JQ1 in MYC

downregulation 161 . Therefore, we investigated whether JQ1 could effectively downregulate MYC in chemoresistant t(8;21) cell lines and elicit similar phenotypes as shRNA-mediated knockdown of MYC. In fact, the t(8;21) cell lines exhibited significant reduction in MYC upon JQ1 treatment (Figure 3.5.A. and B.). Additionally, the t(8;21) cell lines exhibited reduced cell viability, increased apoptosis (Figure 3.5.C. and D.), and reduced colony forming ability at much lower concentrations of JQ1 compared to the

U937 and K562 cell lines (data not shown). The effect of JQ1 on the colony forming ability of primary human CD34 + RE HSPCs was also found to be more significantly

reduced at lower concentrations of JQ1 compared to control CD34 + HSPCs (Figure

3.5.E.). These results reinforce that JQ1 treatment reduces leukemia cell proliferation in

RE9a/NRas G12D /p53 -/- AML, a mouse model for RE leukemia 162 , and indicates that JQ1

also reduces the leukemic potential of primary human RE cells. Expression of the critical

MYC-repressed targets CEBPA and GADD45A were also restored upon JQ1 treatment

in the t(8;21) cell lines, and to a lesser extent in U937 cells (Figure 3.5.F. and G.).

These findings reveal the importance of GM signaling in RE cells to counteract

MYC-associated gene signatures, and activate the expression of critical MYC-repressed

tumor suppressor genes and other differentiation-related genes to reduce leukemic

Figure 3.5. JQ1 treatment reduces proliferation and induces apoptosis in t(8;21) cell lines. (A) Western blot analysis and (B) quantification of MYC levels in Kasumi-1, SKNO-1, U937, and K562 cells treated with indicated JQ1 concentrations. Error bars represent SEM of 3 independent experiments. (C) Viability and (D) flow cytometric analysis of the percentage of Annexin V + cells of Kasumi-1, SKNO-1, U937, and K562 cells treated with indicated JQ1 concentrations. (E) Relative colony forming units (CFUs) of sorted primary human CD34 + cells transduced with control or RE and seeded with indicated concentrations of JQ1. CFUs were normalized to untreated control cells. Error bars indicate SEM of 3 independent experiments (* indicates p < 0.05, # indicates p < 0.01). (F) Quantitative reverse transcription PCR (qRT-PCR) of CEBPA mRNA levels and (G) GADD45A mRNA levels (right) in indicated cells after a 48-hour treatment with indicated JQ1 concentrations. The mRNA levels were normalized to the respective untreated controls for each cell line. Error bars represent SEM of 3 independent experiments (* indicates p < 0.05, # indicates p < 0.01).

potential (Figure 3.6.).

3.2.5 MXI1 relieves the RUNX1-ETO-induced repression of RUNX1 target sites

The RE oncofusion protein contains the RUNX1 DNA-binding domain, which

localizes it to RUNX1 target sites, and the ETO regions, which recruits transcriptional co-

repressors, such as N-CoR, and repressors, such as Sin3A and HDACs, to facilitate

transcriptional repression 163-166 . This transcriptional repression of RUNX1 target genes

by RUNX1-ETO is generally believed to initiate the molecular events involved in

promoting leukemogenesis 102, 103 . Because MXI1 also binds and recruits N-CoR, Sin3A, and HDACs to suppress the expression of MYC target genes, we hypothesized that

MXI1 expression may compete with RE for the binding of transcriptional repressors.

Therefore, overexpression of MXI1 may sequester these transcriptional repressors away from RE, which could potentially alleviate the RE-induced suppression of RUNX1 targets and reverse RE-associated phenotypes. To test this, HEK293T cells were transfected with linearized pBAAE#35-TK-luc vector, which contains two validated RUNX1-binding sites and a minimal thymidine kinase (TK) promoter upstream of a Firefly luciferase reporter and the puromycin resistance gene 167 . HEK293T cells were cultured with puromycin to select for cells with stable genomic integration of the vector (HEK293T- pBAAE#35-TK-luc). Independently transfected and puromycin-selected HEK293T cells were assessed for luciferase activity and two independent cell lines with high basal luciferase activity were used for subsequent experiments.

HEK293T-pBAAE#35-TK-luc cell lines were co-transfected with a Renilla luciferase vector and pcDNA3.1-HA (Control) or pcDNA3.1-HA-RE (RE) or pcDNA3.1-

HA-MXI1 (MXI1) or pcDNA3.1-HA-AE and pcDNA-HA-MXI1 (RE + MXI1). Luciferase activity was measured 48 hours and 72 hours after transfection (Figure 3.7.A.). RE has

Figure 3.6. Proposed mechanism of the inhibitory effects of GM-CSF on the leukemic potential of RUNX1-ETO hematopoietic stem/progenitor cells. Hematopoietic stem/progenitor cells (HSPCs) rely on high levels of MYC to maintain self-renewal potential and prevent differentiation. MYC represses the expression of critical target genes that function as positive regulators of differentiation, such as the transcription factor CEBPA. The transition from HSPCs to differentiated cells requires repression of MYC and MYC activity. Reduced MYC activity allow upregulation of MYC- repressed target genes, such as CEBPA , which aid in enforcing terminal differentiation. In RE HSPCs, RE expression upregulates MYC and enhances self-renewal potential. Increased MYC levels in RE HSPCs also results in the downregulation of MYC- repressed targets, such as CEBPA and other positive regulators of differentiation. Additionally, RE has also been reported to repress CEBPA expression, which further contributes to the myeloid differentiation block. However, upon GM treatment of RE HSPCs, upregulation of MYC-repressed targets is observed. Although GM does not affect MYC levels directly, it restores the expression of MYC-repressed targets which aid RE HSPCs in overcoming their differentiation block and undergo myelopoiesis. MYC inhibition also elicits similar effects. Altogether, both GM and MYC inhibition promote upregulation of MYC-repressed target genes, which results in the differentiation of RE HSPC to ultimately reduce leukemic potential.

77 previously been demonstrated to suppress pBAAE#35-TK-luc promoter activity 167 . As expected, HEK293T-pBAAE#35-TK-luc cells transfected with RE exhibited less than half the promoter activity relative to control-transfected cells. Transfection with MXI1 alone did not significantly alter promoter activity. Interestingly, co-transfection with RE and

MXI1 alleviated the RE-induced promoter repression and resulted in restoration of promoter activity to levels similar to those observed in control-transfected cells.

To delve deeper into the mechanism underlying the alleviation of RE-induced promoter repression in the presence of MXI1, we performed co-immunoprecipitation experiments. Because RE and MXI1 bind to similar co-repressors to facilitate transcriptional repression, we assessed whether ectopic expression of MXI1 sequesters the transcriptional repressor Sin3A from RE. If so, one would expect a reduction in RE and Sin3A binding upon ectopic MXI1 expression. High levels of RE expression are toxic to cells, which generally results in cells being forced to express low levels of RE.

Therefore, since the RE oncofusion protein includes the majority of the ETO protein, we substituted RE with FLAG-tagged ETO to avoid cellular toxicity and to enhance protein expression. Additionally, we used an HA-tagged MXI1. HEK293T cells were transiently transfected with FLAG-ETO (ETO) or with FLAG-ETO and HA-MXI1 (ETO + MXI1).

Expression of tagged proteins and endogenous Sin3A expression was confirmed by western blot of whole cell lysate (Figure 3.7.B.). We performed co-immunoprecipitation of endogenous Sin3A and transiently transfected ETO or ETO + MXI1. As a negative control, co-immunoprecipitation of GST was performed in parallel. ETO was successfully co-immunoprecipitated with Sin3A, although there was also ETO pulled down in the GST control (Figure 3.7.C.). In lysate from cells transfected with ETO + MXI1, MXI1 co- immunoprecipitated with Sin3A and there was no reduction in ETO co- immunoprecipitation. Altogether our data indicates that MXI1 expression does not

Figure 3.7. MXI1 alleviates the RUNX1-ETO-induced repression of RUNX1 transcriptional target sites independently of inhibiting Sin3A recruitment to ETO. (A) Relative luciferase activity of HEK293T-pBAAE#35-TK-luc cells transiently transfected with Renilla luciferase vector and pcDNA3.1-HA (Control) or pcDNA3.1-HA- RE (RE) or pcDNA3.1-HA-MXI1 (MXI1) or pcDNA3.1-HA-AE and pcDNA-HA-MXI1 (RE + MXI1). Luciferase activity was measured 48 hours and 72 hours after transfection. Error bars represent SEM of 4 independent experiments (* indicates p < 0.05, # indicates p < 0.01) (B) Western blot of HEK293T cells transfected with FLAG-tagged ETO and HA-tagged MXI1 for endogenous Sin3A and ectopic ETO and MXI1 expression. (C) Co- immunoprecipitation of FLAG-ETO and HA-MXI1 with control GST or Sin3A from HEK293T cells transfected with FLAG-tagged ETO and HA-tagged MXI1.

sequester Sin3A from or reduce Sin3A binding to ETO, and presumably RE. Therefore,

how MXI1 alleviates the RE-induced transcriptional repression of RUNX1 target sites

remains unclear and should be investigated more thoroughly.

3.3. Discussion

By identifying mechanisms mediating the inhibitory effects of GM on RE

leukemogenesis, and activating or restoring them directly in t(8;21) cells, there exists

potential to uncover alternative therapeutic strategies that could be broadly applicable to

t(8;21) AML. In this study, we discovered that attenuation of MYC-associated gene

signatures is a crucial mechanism mediating the inhibitory effects of GM on RE

leukemogenesis. Inhibition of MYC, either by ectopic MXI1 expression or shRNA-

mediated MYC knockdown, in t(8;21) cell lines efficiently reduced cell proliferation and induced apoptosis. Additionally, we discovered that MYC inhibition resulted in the reactivation of critical MYC-repressed target genes, which were also activated by GM treatment of RE HSPCs. Altogether, our data indicate that MYC inhibition or restoration of critical MYC-repressed targets may serve as a promising therapeutic strategy for treating t(8;21) AML patients, including those who are hyporesponsive to GM.

Importantly, the t(8;21) Kasumi-1 cell line, which is chemoresistant and hyporesponsive to GM, was highly sensitive to MYC inhibition. This suggests that although GM-responsiveness has been compromised in these cells, they remain sensitive to the downstream mechanisms responsible for mediating the negative effects of GM. However, the chemoresistant t(8;21) SKNO-1 cell line was less sensitive to MYC- inhibition compared to Kasumi-1 cells. We hypothesize that these differences are due to the fact that the SKNO-1 cell line was initially established as a GM-dependent cell line

and are therefore more resistant to the negative effects of GM, as well as the

downstream mechanisms mediating its effects. These findings reinforce that MYC

inhibition is effective in reducing the leukemic potential of t(8;21) cells, regardless of their

degree of sensitivity to GM, which is especially pertinent to patients also exhibiting LOS.

Furthermore, we made the novel discovery that MXI1 expression aids in

alleviating the RE-induced transcriptional repression of RUNX1 target sites. This

provides an additional mechanism, aside from MYC inhibition, by which MXI1 promotes

differentiation and inhibits proliferation and self-renewal in RUNX1-ETO cells. Co-

immunoprecipitation experiments found that ectopic MXI1 expression does not result in

reduced Sin3A and ETO binding, suggesting that this alleviation occurs independently of

MXI1 sequestering Sin3A away from RE. Nevertheless, N-CoR and HDACs are also

critical transcriptional co-repressors and repressors that are commonly bound by RE and

MXI and may exhibit differential binding to ETO and RE upon MXI1 expression;

therefore, they should be investigated further.

In summary, here we elucidate a novel mechanism for the negative effects of GM

on t(8;21) cells. Our findings indicate that attenuation or inhibition of MYC, either from

GM treatment or shRNA-mediated MYC knockdown, restores the expression of MYC-

repressed genes that are critical in promoting differentiation and apoptosis in t(8;21)

cells to reduce leukemic potential. We also found that the small molecule inhibitor JQ1

effectively reduced MYC expression in t(8;21) cells. Altogether, we provide further experimental support for MYC inhibition as an effective therapeutic strategy for t(8;21)

AML patients, which should be investigated clinically.

3.4. Materials and Methods

3.4.1 Plasmids

For co-expression of RE with MXI1 in primary human CD34 + cells, RE was cloned into MSCV-IRES-tdTomato, which is identical to MIG, except the EGFP has been replaced with tdTomato and MXI1 was cloned into the MIG construct. For shRNA knockdown, control scrambled shRNA (gift from David Sabatini, Addgene plasmid

#1864) and human MYC shRNA constructs TRCN0000039639, TRCN0000039640, and

TRCN0000039641 (GE Lifesciences, Pittsburgh, PA) were used.

3.4.2 Cell culture

Cell lines were cultured in RPMI supplemented with 10% fetal bovine serum

(FBS) and 100U/mL penicillin/streptomycin. Although SKNO-1 cells were initially

established as a GM-CSF-dependent cell line 154 , over time in culture, they have lost their

cytokine dependence. Primary C57BL/6 murine bone marrow was cultured in IMDM

supplemented with 10% FBS, 100U/mL penicillin/streptomycin, 2% SCF conditioned

media from BHK/MKL cells stably expressing the secretory form of mouse SCF, and 2%

IL-3 conditioned media prepared from X63Ag-653 myeloma cells. Cell lines were

validated by reverse transcriptase-polymerase chain reaction (RT-PCR) of oncofusion

proteins (data not shown).

3.4.3 Proliferation assays

Viable cells were counted via trypan blue exclusion or subjected to the CellTiter

96 Aqueous One Solution Cell Proliferation Assay (MTS) according to manufacturer’s

protocol (Promega, Madison, WI).

3.4.4 Flow Cytometry

For apoptosis studies, APC-conjugated Annexin V and 7AAD (BD Pharmingen,

550475 and 559925, San Jose, CA) were used.

3.4.5 Colony forming assays

For colony forming assays, human cells were seeded in H4534 (StemCell Tech)

with 100 U/mL penicillin/streptomycin.

3.4.6 Western blot

Primary antibodies included rabbit anti-c-MYC (1:5000) (Abcam, ab32072.

Cambridge, UK), rabbit anti-Sin3A (Santa Cruz Biotechnology, sc-994), mouse anti-

HA.11 (Covance, MMS101P, Princeton, NJ), mouse anti-FLAG M2 (Sigma, F1804, St.

Louis, MO), mouse anti-α-tubulin (Sigma, T9026), and mouse anti-β-actin (Sigma,

A2228) antibodies. Licor (Lincoln, NE) IRDye 680 anti-rabbit (926-32221) and IRDye 800

anti-mouse (926-32210) secondary antibodies (1:10000) were used for visualization on a

LI-COR Odyssey Classic imager.

3.4.7 Quantitative reverse-transcription PCR (qRT-PCR)

RNA was isolated using the TRIzol method (Life Technologies, Carlsbad, CA).

cDNA was generated using QuantiTect Reverse Transcription (Qiagen) kits. Quantitative

PCR was done and analyzed with the CFX Connect Real-Time PCR Detection System

(Bio-Rad) using the KAPA SYBR FAST qPCR Kit (Kapa Biosystems, Wilmington, MA)

with 0.25 μM of each primer performing 30-40 cycles of 95°C for 5 seconds and 60°C for

30 seconds. Primers for qPCR are provided in Table 3.1. The cycle threshold (Ct) values

for the no reverse transcriptase control samples were more than 10 Ct values greater

than the samples with reverse transcriptase (data not shown). The qPCR products were

Table 3.1. Human qPCR primers

Human Primers Sequence (5’ to 3’) hCEBPA-For AACCTTGTGCCTTGGAAATG hCEBPA-Rev CCCTATGTTTCCACCCCTTT hGADD45A-For CCCTGATCCAGGCGTTTTG hGADD45A-Rev GATCCATGTAGCGACTTTCCC hGAPDH-For TCGCTCAGACACCATGGGGAAG hGAPDH-Rev GCCTTGACGGTGCCATGGAATTTG hMXI1-For GCGCCTTTGTTTAGAACGCTT hMXI1-Rev AATGCTGTCCATTCGTATTCGT hRNAPolII-For GCACCACGTCCAATGACAT hRNAPolII-Rev GTGCGGCTGCTTCCATAA

run out on 1.8% agarose gels to confirm there was no product in the no reverse

transcriptase control samples and that products were of the correct size in samples with

reverse transcriptase (data not shown).

3.4.8 Luciferase assay

HEK293T cells were transfected with 5 μg of gel purified pBAAE#35-TK-luc

linearized with BglII. 24 hours after transfection, cells were selected and passaged with 2

μg/mL puromycin for 6 days, after which puromycin was reduced to 1 μg/mL. Single

wells of a 6-well plate of HEK293T-pBAAE#35-TK-luc cells were co-transfected with 10

ng of Renilla luciferase vector and a total of 800 ng of pcDNA3.1 vector. Cells from a

single well of a 6-well plate were used to generate lysate, and 8% of the lysate was used

for the Dual-Luciferase Reporter Assay System (Promega) according to manufacturer’s

protocol.

3.4.9. Co-immunoprecipitation

HEK293T cells were transfected with 1 μg of pCMV3xFLAG7.1-ETO and 9 μg of

pcDNA3.1(+) or pcDNA3.1(+)-HA-MXI1 in a 10 cm plate. Cells were harvested 48 hours

post transfection and lysed in 900 μL of 50mM HEPES, 150mM NaCl, 1% glycerol, 1%

Triton-X 100, 1.5mM MgCl 2, 5mM EGTA, cOmplete Protease Inhibitor (Roche), and

PhosStop (Roche). 10 μL of Protein A beads (Thermo Fisher Scientific, 15918014) were pre-cleared and incubated with 250 μL of lysate, 250 μL of lysis buffer, and 1 μg of rabbit anti-Sin3A antibody (Santa Cruz, sc-994) or 2ug of anti-GST antibody (A455) with rotation for 5 hours at 4ºC. Beads were washed 5 times with 500 μL of lysis buffer and eluted in sample buffer for western blotting.

3.4.10. Statistical analysis

Statistical significance was determined from adequately powered sample sizes of similar variation using two-tailed unpaired Student’s t-tests and was defined as P < 0.05.

Sample sizes are given in figure legends.

3.5. Acknowledgements

Chapter 3 includes unpublished and published material. Unpublished material includes co-immunoprecipitation experiments performed by Russell DeKelver. Published material presented from Weng S, Matsuura S, Mowery CT, Stoner SA, Lam K, Ran D,

Davis AG, Lo MC, Zhang DE. “Restoration of MYC-repressed targets mediates the negative effects of GM-CSF on RUNX1-ETO leukemogenicity.” Leukemia, 2016. The dissertation author was the primary investigator and author of this study. S.W. designed and performed the experiments. S.M., C.M., S.S., K.L., D.R., A.D., and M.L. assisted with experiments. S.W. analyzed the data and prepared figures. S.W. and D.Z. wrote the manuscript.

CHAPTER 4:

CSF2RA is post-transcriptionally downregulated in t(8;21) cell lines

86 87

4.1. Introduction

Given the importance of GM-CSF signaling in negatively regulating t(8;21)

leukemogenesis, proper regulation of the expression of CSF2RA and CSF2RB , the genes encoding the GM-CSF receptor, is critical to prevent leukemic transformation.

LOS, which is frequently observed in t(8;21) patients, results in haploinsufficiency of the

PAR gene CSF2RA and aids in reducing GM signaling in these cells to promote leukemogenesis. In support of this is the observation that CSF2RA is one of the most significantly downregulated PAR genes in t(8;21) patients 108, 109 .

In addition to gene dosage alteration, which occurs with LOS-driven haploinsufficiency, the expression of CSF2RA can be regulated through alternative mechanisms. The PU.1 and C/EBPα transcription factors have been reported to bind the

CSF2RA promoter and activate its expression 168 . Interestingly, expression of SPI1 , the

gene encoding PU.1, and CEBPA have been found to be repressed by RE 102, 103 , which

may further contribute to reduced CSF2RA expression in RE-expressing cells.

Gene expression can also be regulated post-transcriptionally through the 5’ and

3’ untranslated regions (UTRs) of the messenger RNA (mRNA) 169 . The 5’ UTR is

upstream of the start codon and contains ribosomal binding sites that allow for

translation of the protein coding sequence 169 . The 5’ UTR can also contain cis-acting

regulatory elements that can impact translation 169 . The 3’ UTR is downstream of the translational stop codon and often contain binding sites for microRNAs (miRNAs), adenylate-uridylate-rich (AU-rich) elements binding proteins (ARE-BP), poly(A) binding proteins (PABP), and other regulatory elements 169 . The UTRs provide an additional

complex layer of post-transcriptional gene regulation that are often targeted and

perturbed during cellular transformation 170, 171 .

The GM receptor, encoded by CSF2RA and CSF2RB , is expressed on

hematopoietic cells of the myeloid lineage. Its expression increases with differentiation,

is highest on monocytes, macrophages, dendritic cells, and granulocytes, and has been

found to be dysregulated in AML 112, 113, 172-174 . As mentioned previously, CSF2RA is

significantly downregulated in t(8;21) AML patients compared to non-t(8;21) AML

patients of the M2 subtype 108, 109 . Additionally, the t(8;21) Kasumi-1 cell line has lower levels of CSF2RA expression compared to other myeloid leukemia cell lines and leukemia cells from t(8;21) patients have been reported to be hyporesponsive to GM 114,

115, 175 . Although these observations could be attributed to the high frequency of LOS and

subsequent haploinsufficiency of CSF2RA in t(8;21) patients, we found that when the

cDNA of CSF2RA was ectopically expressed, CSF2RA expression in t(8;21) cell lines was much lower compared to in non-t(8;21) cell lines. In this study, we sought to address the differences in CSF2RA expression in t(8;21) cells versus non-t(8;21) cells and discovered that CSF2RA is downregulated via the 3’ UTR in t(8;21) cells. This mechanism of post-transcriptional downregulation of CSF2RA through its 3’ UTR may also aid in explaining the hyporesponsiveness of t(8;21) leukemia cells to GM. With the newfound role of GM signaling in negatively regulating t(8;21) leukemogenesis, downregulation of CSF2RA through its 3’ UTR serves as an additional mechanism to

reduce GM signaling in t(8;21) cells and promote leukemic transformation. Therefore, it

is imperative to clarify the mechanisms by which CSF2RA is downregulated via it’s 3’

UTR in t(8;21) cells.

4.2. Results

4.2.1. CSF2RA is downregulated in t(8;21) cell lines

To assess the effects of CSF2RA cDNA expression, we transduced the t(8;21) cell lines, Kasumi-1 and SKNO-1, and the U937 and K562 cell lines, with MIP-

5’+CSF2RA+3’, which includes the ~170bp CSF2RA 5’ UTR, the 1.2Kb coding sequence, and the ~460bp 3’ UTR (Figure 4.1.). Unexpectedly, flow cytometric analysis of CSF2RA (CD116) expression revealed that the percentage of CSF2RA + cells

decreased over time in the t(8;21) Kasumi-1 and SKNO-1 cells. Next, we transduced the

same cell lines with MIG and MIG-5’+CSF2RA+3’, which is identical to MIP with the

exception of containing an EGFP reporter downstream of the internal ribosomal entry

site (IRES) instead of the puromycin resistance gene. The IRES allows for GFP to be

independently translated from the same mRNA transcript as CSF2RA and serves as a

reporter for retroviral transcript expression. In cells transduced with control MIG, the

percentage of GFP + cells remained constant across all cell lines over time (Figure

4.2.A.). Conversely, in cells transduced with MIG-5’+CSF2RA+3’, the percentage of

GFP + cells decreased in the t(8;21) Kasumi-1 and SKNO-1 cell lines over time (Figure

4.2.B.). Since GFP acts as a reporter for transcript expression and its expression also decreased in the t(8;21) cell lines, this implies that the reduction in both CSF2RA and

GFP expression is likely due to the entire retroviral transcript being downregulated in t(8;21) cell lines.

4.2.2 CSF2RA is downregulated through the 3’ UTR in t(8;21) cell lines

As aforementioned, MIP-5’+CSF2RA+3’ contains the entire CSF2RA cDNA,

including the 5’ and 3’ UTRs. Therefore, we hypothesized that the UTRs were subjecting

the retroviral transcript to post-transcriptional downregulation. To determine whether the

Figure 4.1. CSF2RA expression decreases in t(8;21) cell lines. Cell lines were transduced with MIP-5’+CSF2RA+3’, which includes the CSF2RA ~170bp 5’ UTR (purple), the 1.2Kb coding sequence (grey), and the ~460bp 3’ UTR (blue). Flow cytometric analysis of CSF2RA (CD116) expression over time. Error bars represent SEM of 3 independent experiments.

Figure 4.2. EGFP reporter expression decreases over time in t(8;21) cell lines. Cell lines were transduced with control MIG and MIG-5’+CSF2RA+3’, which includes the CSF2RA ~170bp 5’ UTR, the 1.2Kb coding sequence, and the ~460bp 3’ UTR, followed by an internal ribosomal entry site (RES) and an EGFP reporter to allow EGFP expression from the same transcript as CSF2RA. Flow cytometric analysis of GFP expression over time. Error bars represent SEM of 3 independent experiments.

UTRs mediate the observed downregulation of CSF2RA expression, several constructs

with various UTR combinations were generated (Figure 4.3.A.). The coding sequence of

CSF2RA , with no UTRs, was cloned into MIP (MIP-CSF2RA), as well as the 5’ UTR and

CSF2RA coding sequence (MIP-5’+CSF2RA), and the CSF2RA coding sequence and 3’

UTR (MIP-CSF2RA+3’) (Figure 4.3.A.). These constructs were then transduced into the

Kasumi-1, SKNO-1, U937, and K562 cell lines (Figure 4.3.B.) All cells transduced with

MIP-CSF2RA and MIP-5’+CSF2RA exhibited a constant percentage of CSF2RA + cells

over time. However, when cells were transduced with MIP-5’+CSF2RA+3’ and MIP-

CSF2RA+3’, the percentage of CSF2RA + cells decreased over time in the Kasumi-1 cell line, and slightly decreased in the SKNO-1 cell line. Furthermore, to assess whether

CSF2RA expression is being downregulated at the mRNA or protein level, we compared the expression of CSF2RA mRNA in Kasumi-1 cells transduced with the various UTR combinations at Days 0 and 6 (Figure 4.3.C.). Kasumi-1 cells transduced with constructs containing the 3’ UTR exhibited downregulation of the CSF2RA mRNA at Day 6 compared to Day 0, which was not observed when the 3’ UTR was absent. These data indicate that the CSF2RA transcript is subjected to downregulation via the 3’ UTR at the mRNA level in t(8;21) cell lines.

Because the downregulation of CSF2RA was only observed in the t(8;21) cell lines Kasumi-1 and SKNO-1, we sought to address whether RUNX1-ETO influences

CSF2RA expression. The non-t(8;21) K562 cell line was transduced with MIP control or

MIP-RUNX1-ETO and selected with puromycin. After drug selection, these cells were co-transduced with MIP-CSF2RA+3’. The expression of CSF2RA in K562 cells was unaffected with the expression of RUNX1-ETO, suggesting that RUNX1-ETO is not responsible for the downregulation of CSF2RA (Figure 4.3.D.).

Figure 4.3. The 3’ UTR mediates downregulation of CSF2RA in t(8;21) cell lines. (A) Schematic of the various MIP constructs containing combinations of the CSF2RA 5’ UTR (purple), the coding sequence (grey), and the 3’ UTR (blue). (B) Flow cytometric analysis of CSF2RA (CD116) expression over time in cell lines transduced with constructs from (A). Error bars represent SEM of 3 independent experiments. (C) qRT- PCR of CSF2RA expression of Kasumi-1 cells from (B) at Day 0 and Day 6. CSF2RA expression was normalized to Day 0 values. Error bars represent SD of 1 experiment. (D) K562 cells were transduced with MIP control or MIP-RUNX1-ETO. After drug selection, cells were co-transduced with MIP-CSF2RA+3’ and expression of CSF2RA (CD116) was measured by flow cytometry. Error bars represent SD of 1 experiment.

Next we sought to identify factors targeting the 3’ UTR for the post-transcriptional

downregulation. The 3’ UTR can house binding sites for miRNAs. Binding of a miRNA to

its target initiates recruitment of the RNA-induced silencing complex (RISC) and results

in either cleavage of the transcript or inhibition of translation. Using miRWalk 2.0, the 3’

UTR of CSF2RA was searched for putative miRNA binding sites (Figure 4.4.A.).

Additionally, global miRNA expression profiling datasets of the Kasumi-1 and K562 cell lines were used to identify miRNAs predicted by miRWalk 2.0 to bind the CSF2RA 3’

UTR that are expressed in Kasumi-1 cells. Together, this resulted in the identification of

9 candidate miRNAs. Each miRNA, along with 300 nucleotides upstream and downstream of the miRNA sequence, was cloned into the MIP vector. The cell line K562, which does not exhibit downregulation of CSF2RA when the 3’ UTR is present, was utilized to screen for miRNAs capable of downregulating the 3’ UTR. K562 cells were transduced with individual miRNAs. Following puromycin drug selection for transduced cells, cells were co-transduced with MIP-CSF2RA+3’ to screen for downregulation of

CSF2RA in the presence of the various miRNAs. Expression of the 9 miRNAs in K562

cells did not result in the downregulation of CSF2RA, suggesting these miRNAs are not

responsible for the downregulation of CSF2RA through its 3’ UTR (Figure 4.4.B.).

4.2.3. The CSF2RA coding sequence and 3’ UTR are required for maximal

downregulation

In order to further confirm that the 3’ UTR is solely responsible for the

downregulation of the retroviral transcript, we generated two constructs to eliminate the

possibility that IRES or coding sequence is involved in this process. To test whether the

IRES is involved, the IRES and puromycin resistance gene were removed from the MIP-

CSF2RA and MIP-CSF2RA+3’ construct to generate MSCV-CSF2RA and MSCV-

Figure 4.4. Candidate CSF2RA 3’ UTR-targeting miRNAs do not promote downregulation of CSF2RA in K562 cells. (A) miRWalk 2.0 was used to search various miRNA targets databases for putative miRNA binding sites and miRNAs predicted to target the 3’ UTR of CSF2RA . For each database, a score of 1 is given when a miRNA is predicted to target the 3’ UTR of CSF2RA . A total score is generated by adding up individual scores from each database, and a total score minimum threshold of 4 was set for this study. Two publicly available GEO datasets of global miRNA expression in the Kasumi-1 (GSM506744) and K562 (GSE57679) cell lines was used to generate an expression heatmap of the candidate miRNAs (highlighted in red). (B) K562 cells were transduced with MIP control or with the indicated MIP-miRNA. After drug selection, cells were co-transduced with MIP- CSF2RA+3’ and the expression of CSF2RA (CD116) was measured by flow cytometry. Error bars represent SD of 1 experiment.

CSF2RA+3’, respectively (Figure 4.5.A.). The Kasumi-1 and K562 cell lines were

transduced with MSCV-CSF2RA and MSCV-CSF2RA+3’. Surprisingly, Kasumi-1 cells

transduced with MSCV-CSF2RA displayed reduced CSF2RA expression over time;

however, the reduction was not as severe as in cells transduced with MSCV-

CSF2RA+3’. This result indicates that the IRES is not involved in the downregulation

process, and removal of the IRES enhanced CSF2RA downregulation even in the

absence of the 3’ UTR. Quantification of CSF2RA mRNA levels in these cells also

revealed that CSF2RA mRNA was downregulated in Kasumi-1 cells transduced with

MSCV-CSF2RA, but was more severely downregulated when the 3’ UTR was present

(Figure 4.5.B.). As expected, downregulation of CSF2RA mRNA was not observed in

K562 cells. Altogether, this suggests that the coding sequence of CSF2RA is also targeted for downregulation in Kasumi-1 cells. This result was unexpected because

Kasumi-1 cells transduced with the MIP-CSF2RA did not exhibit the same degree of

CSF2RA downregulation as those transduced with MSCV-CSF2RA. The observed differences may be attributed to a steric block initiated by the recruitment of translational machinery to the IRES, which could hinder factors from binding to the coding sequence to promote its degradation, and is rescued upon its removal.

With this newfound discovery that the CSF2RA coding sequence is also targeted for downregulation, we sought to test whether the coding sequence of CSF2RA was required for downregulation of CSF2RA via its 3’ UTR. The CSF2RA coding sequence was removed from MSCV-CSF2RA and MSCV-CSF2RA+3’ and replaced with EGFP to generate MSCV-EGFP and MSCV-EGFP+3’, respectively (Figure 4.6.). The Kasumi-1 and K562 cell lines were transduced with MSCV-EGFP and MSCV-EGFP+3’ and the percentage of GFP + cells was measured over time. Unexpectedly, Kasumi-1 cells

transduced with MSCV-EGFP 3’ did not exhibit any decrease in the percentage of GFP +

Figure 4.5. The CSF2RA coding sequence and 3’ UTR are targeted for downregulation in Kasumi-1 cells. (A) Kasumi-1 and K562 cells were transduced with MSCV-CSF2RA and MSCV- CSF2RA+3’ and CSF2RA (CD116) expression was measured by flow cytometry. Error bars represent SD of 2 independent experiments. (B) qRT-PCR of CSF2RA expression of Kasumi-1 cells from (A) at Day 0 and Day 6. CSF2RA expression was normalized to Day 0 values. Error bars represent SD of 1 experiment.

Figure 4.6. The coding sequence of CSF2RA is required for 3’ UTR-mediated downregulation of CSF2RA in Kasumi-1 cells. Kasumi-1 and K562 cells were transduced with MSCV-EGFP and MSCV-EGFP+3’ and GFP expression was measured by flow cytometry over time. Error bars represent SD of 2 independent experiments.

cells over time.

Altogether, these data suggest that CSF2RA is maximally downregulated in the

t(8;21) Kasumi-1 cell line when the 3’ UTR is present; however, the 3’ UTR is not

sufficient to promote downregulation and the coding sequence of CSF2RA is required.

Additionally, we discovered that the coding sequence itself can also be targeted for slight

downregulation in t(8;21) cell lines.

4.2.4. The strongest AU-rich element in the 3’ UTR of CSF2RA does not entirely

mediate the downregulation of CSF2RA

Due to the discovery that multiple regions of the transcript are required for

maximal post-transcriptional downregulation of CSF2RA in the Kasumi-1 cell line, we

hypothesized that the transcript is unlikely to be targeted by miRNAs, which generally

bind to short RNA sequences complementary to its seed sequence, and instead recruits

RNA-binding proteins that regulate its stability. One class of RNA-binding proteins that

are known to bind to regulatory sequences in the 3’ UTRs are AU-rich elements binding

proteins (ARE-BPs). ARE-BPs bind to AU-rich elements that are commonly found in 3’

UTRs. The core of an AU-rich element consist of a conserved AUUUA motif 176 . This core

AUUUA motif is often flanked by several repeated A’s and U’s to generate a more

optimal UUAUUUA(U/A)(U/A) motif 177, 178 .

Examination of the CSF2RA 3’ UTR sequence for AU-rich elements revealed that

it contains two core AUUUA motifs, one that lies within a more optimal

UUAUUUA(U/A)(U/A) motif. Therefore, to test whether this optimal AU-rich element is

recruiting ARE-BPs that are responsible for the downregulation of CSF2RA , we mutated

the AUUUA motif in the MSCV-CSF2RA+3’ construct to AGGGA to generate the MSCV-

CSF2RA+3’ Mut (x1). Mutation of the AUUUA motif in the AU-rich element did not

100

abolish the downregulation of CSF2RA in Kasumi-1 cells, suggesting that this AU-rich

element is not responsible for its downregulation (Figure 4.7.).

4.3. Discussion

Gene expression profiling of AML patients has revealed that t(8;21) AML patients

exhibit significantly lower levels of CSF2RA expression compared to non-t(8;21) AML

patients 108, 109 . The hyporesponsiveness to GM that is observed in leukemic blasts from

t(8;21) AML patients also functionally reflects the generally reduced GM receptor

expression in t(8;21) patients 114, 115 . Although LOS, and subsequent CSF2RA

haploinsufficiency, may explain the reduced expression levels in t(8;21) patients, the

t(8;21) Kasumi-1 cell line expresses the lowest levels of CSF2RA compared to other myeloid leukemia cell lines, many of which also exhibit LOS. These findings collectively suggest that CSF2RA expression in t(8;21) Kasumi-1 cells is lower than expected with

CSF2RA haploinsufficiency, and there may exist alternative mechanisms responsible for repressing CSF2RA expression. Our discovery that CSF2RA transcript is downregulated in the t(8;21) Kasumi-1 and SKNO-1 cell lines provides an explanation for the lower level of CSF2RA expression in these cells. Importantly, our discovery also uncovers an additional mechanism employed by t(8;21) cells to downregulate CSF2RA expression and further escape the inhibitory effects of GM signaling to promote leukemogenesis.

Unexpectedly, we found that although the 3’ UTR promotes downregulation of the CSF2RA transcript, it is not sufficient for downregulation and the gene’s coding region is also necessary. Truncations of the coding sequence would aid in the identification of the region required for the downregulation in the presence of the 3’ UTR.

Additionally, although we found that the strongest AU-rich element is not responsible for

101

Figure 4.7. Mutation of the strongest AU-rich element does not fully rescue the downregulation of CSF2RA. (A) The DNA sequence of the CSF2RA 3’ UTR with the canonical AU-rich element AUUUA core sequences (red) and the optimal UUAUUUA(U/A)(U/A) motif (pink) highlighted. (B) The core sequence (red) located within the larger optimal AU-rich element motif (pink) was mutated from AUUUA to AGGGA in the MSCV-CSF2RA+3’ construct. Kasumi-1 cells were transduced with the indicated constructs and CSF2RA (CD116) expression was measure by flow cytometry.

102

promoting downregulation of CSF2RA , there remains another AUUUA motif that could

also be mutated to fully eliminate AU-rich elements from the 3’ UTR. Due to the well-

established role of 3’ UTRs in post-transcriptionally regulating translation and the

stability of mRNA transcripts, we have presumed that the downregulation of CSF2RA is

occurring post-transcipritonally through the 3’ UTR. However, all of our studies involved

the expression of CSF2RA from a retroviral long terminal repeat (LTR), which are known to contain regulatory elements, including promoters, enhancers, and polyadenylation signals, that could potentially be influencing our results 179 . Therefore, the effects of the

CSF2RA coding sequence and 3’ UTR should also be confirmed in an LTR-free expression system, as is the case with many luciferase reporter constructs.

Due to the fact that the coding sequence and 3’ UTR are required for maximal downregulation of CSF2RA , we hypothesize that an RNA-binding protein may be

responsible to promoting degradation of the transcript. A screen could be conducted to

identify direct and indirect factors that are necessary for downregulation of the CSF2RA

transcript. The genome-wide CRISPR-Cas9 knock-out (GeCKO) library, which targets

19,050 genes and 1,864 miRNAs, has become a preferred tool for genome-wide

screens 180 . To pinpoint direct and indirect factors required to dowregulate CSF2RA , the

t(8;21) Kasumi-1 cell line could be co-transduced with the GeCKo library and with a

vector containing the CSF2RA coding sequence and its 3’ UTR. Kasumi-1 cells with

CRISPR-Cas9 mediated knockout of a gene essential to the process of post-

transcriptionally downregulating CSF2RA should consequently express high levels of

CSF2RA. Therefore, sorting Kasumi-1 cells with high levels of CSF2RA expression after

a set period of culture would isolate Kasumi-1 cell clones with knockout of a critical

factor for the post-transcriptional downregulation of CSF2RA . Due to genomic integration

of the CRISPR-Cas9 GeCko provirus upon transduction, these factors can be identified

103

via amplification and sequencing of the sgRNA sequence from genomic DNA of sorted

cells.

To directly identify proteins binding to and regulating the CSF2RA transcript, the

portion of the coding sequence and the 3’ UTR that is required for the downregulation

can be screened for factors that target it. One method to identify binding proteins to the

region would be to in vitro transcribe and biotinylate the transcript. The biotinylated transcript can then be incubated with Kasumi-1 cell lysate and immunoprecipitated to isolate and identify binding proteins via western blotting or mass spectrometry.

More recently, it has been reported that epigenetic marks on mRNA can regulate mRNA stability and translation, thus providing an additional layer of gene regulation 181 .

N6-methyladenosine (m 6A), a dynamic epigenetic mark that is written by METTL3 and

METTL14 and erased by FTO and ALKBH5, has been found on nearly half of human

gene transcripts and is enriched near stop codons and in 3’ UTRs 182, 183 . Genome-wide

analysis of sequences surrounding m6A marks have revealed an enriched RRACU core

motif 182 , which the CSF2RA coding sequence and 3’ UTR has 19 of. Therefore,

assessing the ability for CSF2RA to be downregulated after mutating these motifs or

knocking down the METTL3 and METTL14 m 6A writers would elucidate whether this

epigenetic mark is involved.

Together, these experiments would reveal critical proteins, miRNAs, epigenetic

marks involved in the post-transcriptional downregulation of CSF2RA and have the

potential to expose a cohort of novel targets that could be inhibited to restore CSF2RA

expression and re-sensitize t(8;21) AML cells to GM signaling.

104

4.4. Materials and Methods

4.4.1. Plasmids

The CSF2RA protein coding sequence was cloned into the MIP vector to generate MIP-CSF2RA. The 5’ UTR and 3’ UTR of CSF2RA were included to generate the MIP-5’+CSF2RA+3’ vector. The MIP-5’+CSF2RA vector includes the CSF2RA 5’

UTR and protein coding sequence, and the MIP-CSF2RA+3’ vector contains the

CSF2RA protein coding sequence and 3’ UTR sequence. To generate MSCV vectors, the IRES and puromycin resistance gene were removed from the MIP vector.

Additionally, the CSF2RA coding sequence was replaced with EGFP to generate the

MSCV-EGFP and MSCV-EGFP+3’ vectors. MicroRNA sequences, in addition to 300

nucleotides upstream and downstream of the microRNA sequence, were PCR amplified

from genomic DNA isolated from Kasumi-1 cells and cloned into the MIP vector. Site

directed mutagenesis of the MSCV-CSF2RA+3’ vector was performed with forward

primer: 5’-GAACCTTTATATCATTTTCTATGTTTTTAGGGAAAAACATGACATTTG and

reverse primer: 5’-ACAGCATCATCAAGAAAACAGTTC to generate the MSCV-

CSF2RA+3’ Mut (1x) vector.

4.4.2. Flow cytometry

For CSF2RA expression analysis, cells were stained with PE-conjugated anti-

human CD116 antibody (eBioscience, 12-1169) and analyzed with a BD FACS Canto

cytometer (BD Biosciences).

4.4.3. Quantitative reverse-transcription PCR (qRT-PCR)

RNA was isolated using the TRIzol method (Life Technologies), treated with RQ1

RNase-Free DNase (Promega), and re-purified using the TRIzol method. cDNA was

105 generated using MCLAB First Strand cDNA synthesis kit (MCLAB, San Francisco, CA).

Primers to determine CSF2RA expression included CSF2RA forward primer: 5’-

GCTCCTGGAGTGAAGCCATT and reverse primer: 5’-

CCCAGATGATCTCGTCTTCCA. Quantitative PCR was done and analyzed with the

CFX Connect Real-Time PCR Detection System (Bio-Rad) using the KAPA SYBR FAST qPCR Kit (Kapa Biosystems, Wilmington, MA) with 1 μM of each primer performing 50 cycles of 95°C for 5 seconds and 60°C for 30 seconds. The cycle threshold (Ct) values for the no reverse transcriptase control samples were more than 10 Ct values greater than the samples with reverse transcriptase. PCR products were run out on a 1.8% agarose gel (data not shown).

4.5. Acknowledgements

CHAPTER 5:

Conclusion and Future Directions

106 107

5.1. Conventional t(8;21) AML treatment and alternative approaches

The t(8;21)(q22;q22), which was discovered by Janet Rowley in 1982, is one of

the most commonly observed cytogenetic aberrations in adult AML patients 96, 97, 184 .

Studies have found that t(8;21) is insufficient and additional mutations are required to drive leukemia development 22 . LOS is observed in 35-59% of t(8;21) patients, which

suggests this event is favored during and may provide sufficiency for disease

development. The conventional treatment for t(8;21) patients has relied on induction and

consolidation chemotherapy for over 30 years 99, 185-187 . Although prognoses for t(8;21)

AML patients are generally favorable, roughly 50% of patients relapse 98-100 . This signifies

that conventional therapy fails to fully eradicate the disease and completely eliminate

leukemia stem cell populations, which are responsible for initiating relapse. By studying

the mechanisms of disease development and identifying essential pathways for the

survival of leukemia stem cells, there is potential to uncover novel strategies to more

effectively eliminate leukemia stem cell populations.

The focus of this dissertation was to examine how LOS contributes to t(8;21)

leukemia development. Although there are numerous potential mechanisms by which

LOS may cooperate in t(8;21) AML development, I concentrated on the role of

haploinsufficiency of the PAR gene CSF2RA , and therefore reduced GM signaling, in the

pathogenesis of t(8;21) AML. By clarifying the mechanisms by which GM signaling

prevents t(8;21) leukemogenesis, there lies hope in restoring or activating these

mechanisms, which would provide novel targeted approaches to treating t(8;21) patients.

Therefore, the objective of this dissertation was to identify novel therapeutically

targetable mechanisms mediating the inhibitory effects of GM signaling on t(8;21)

leukemogenesis. To this end, I (1) analyzed the gene expression profile of GM-treated

108

RE HSPCs to broadly assess the effects of GM on these cells, (2) performed a screen to

identify GM-induced genes that effectively reduce the self-renewal potential of RE

HSPCs, (3) identified that the MYC pathway is critical for maintaining the self-renewal

potential of RE HSPCs, which may provide an attractive target in t(8;21) treatment, and

(4) discovered that CSF2RA , in addition to being downregulated by LOS-induced haploinsufficiency, is also downregulated through its 3’ UTR in t(8;21) cell lines.

5.2. Caveats to the use of GM-CSF in the treatment of AML

GM signaling regulates numerous cellular processes including differentiation, proliferation, and survival of hematopoietic cells 110 . GM is used clinically to enhance white blood cell production and facilitate recovery from chemotherapy-induced myelosuppression 188, 189 . Additionally, it has been investigated for efficacy in inducing proliferation and sensitizing leukemia blasts to chemotherapy, although these studies yielded inconclusive results 190, 191 . Notably, GM has also been implicated in leukemogenesis, and has been found to be involved in CMML and JMML 59, 192-194 .

Altogether, these findings suggest that the effects of GM are diverse and highly dependent on cellular context and that administration of GM to t(8;21) patients could result in the undesired cellular consequence of increased leukemia cell proliferation, which was observed when primary human RE HSPCs were culture with GM.

5.3. Therapeutic implications of studying GM-CSF signaling in t(8;21) cells

Although GM signaling was also found to enhance proliferation of RE HSPCs,

GM reduced their self-renewal potential and promoted their differentiation, which further

109

emphasizes the complexity of GM signaling in differing cellular contexts. Regardless, the

increased proliferation stimulated by GM would be unfavorable in the context of

leukemia treatment. Therefore, the principal goal of my dissertation was to identify GM-

induced mechanisms that reduce self-renewal potential of RE HSPCs without the

induction of its mitogenic effects. I discovered that GM treatment of RE HSPCs results in

the attenuation of MYC-associated gene signatures and the reactivation of MYC-

repressed targets aid in promoting RE HSPC differentiation and apoptosis. Additionally, I

found that t(8;21) cells are highly sensitive to MYC inhibition. These discoveries were

especially timely with the identification of a small molecule BET inhibitor, JQ1, which has

been demonstrated to effectively inhibit MYC in various cancer cell types 160 . Although

JQ1 treatment effectively inhibited MYC , reduced proliferation, and induced apoptosis in the t(8;21) cell lines tested, this molecule is relatively insoluble and suffers a short half life. However, newer generations of BET inhibitors have been developed and are currently being tested clinically. The findings in this dissertation suggest that these new generation BET inhibitors may hold great promise in effectively eradicating t(8;21)

HSPCs.

As is observed with most therapeutic agents, the initial response is typically robust and patients enter remission. However, resistant leukemia stem cell clones will often persist through treatment, generally due to the acquisition of additional mutations to escape drug toxicity, and eventually initiate relapse. Therefore, the greatest challenge in cancer treatment is the comprehensive eradication of leukemia stem cells within the patient. Combination therapies, which target multiple pathways simultaneously, are proving to be significantly more effective than single therapies. Thus, it becomes increasingly imperative to identify additional essential pathways that could be targeted concurrently to ensure complete elimination of leukemic stem cell populations. Gene

110

expression profiling revealed that GM induced slight but concerted upregulation of 85

genes in RE HSPCs. Of the 10 genes screened, 6 successfully inhibited the self-renewal

potential of RE HSPCs. This leaves an open door for exploration of 75 additional genes,

which may elucidate additional pathways that hold therapeutic promise and warrants

further investigation.

5.4. Additional implications for GM-CSF signaling in the pathogenesis of t(8;21)

AML

GM has recently risen to the forefront in immunotherapy due to its effectiveness

in promoting dendritic cell (DC) differentiation and activating immune cells 195, 196 . It is also being investigated clinically as an adjuvant in numerous cancer vaccines 197 . These current applications of GM are of relevance to t(8;21) AML patients because their leukemic blasts are generally hyporesponsive to GM due to reduced CSF2RA expression from LOS 114, 198 . Although LOS and t(8;21) are frequently observed together, it is unclear whether they are sufficient for leukemogenesis. Therefore, cells harboring both t(8;21) and LOS may remain dormant in these patients for extended periods of time. It has been reported that peripheral blood AML cells differentiate into DCs in the

presence of cytokine cocktails, which include GM, and activate T cells to have cytotoxic

activity against AML blasts 199 . Thus, t(8;21) cells with LOS, which secrete less GM due

to the RE-induced repression of GM and are hyporesponsive to GM due to LOS, would

have impaired differentiation into myeloid DCs. Consequently, their ability to activate T

cells and participate in cancer immunosurveillance would also be compromised 199, 200 .

This may further contribute to leukemia development if these t(8;21) cells evade

immunosurveillance and remain in the bone marrow until additional mutations

111

accumulate for disease initiation. Altogether, this implies a greater immunological role for

GM signaling in preventing RE leukemia development that should be explored.

5.5. Impact of the dissertation

The high incidence of LOS in cancers, together with the observation that

pseudoautosomal regions are frequent targets for LOH in cancers, has led to the

postulation that tumor suppressor genes reside in the PARs. Although LOS and LOH of

sex chromosome PARs are frequently observed in cancers 6, 201 , there are few studies

examining the function of PAR genes in cancer development, progression, and

maintenance. This dissertation aimed to provide an in-depth characterization of the

function of the PAR gene CSF2RA and consequences of reduced GM signaling in t(8;21) leukemia. Though GM, its receptor, and GM signaling have been heavily studied since the discovery of GM in 1984 202, 203 , we have been the first to suggest that GM

signaling can prevent leukemogenesis. Importantly, the studies in this dissertation are

the first to clarify the mechanisms mediating the inhibitory role of GM on t(8;21) AML

pathogenesis and support targeting of the MYC pathway in the treatment of t(8;21) AML.

5.6. Acknowledgments

Chapter 5 contains, in part, published material. Weng S, Matsuura S, Mowery

CT, Stoner SA, Lam K, Ran D, Davis AG, Lo MC, Zhang DE. “Restoration of MYC-

repressed targets mediates the negative effects of GM-CSF on RUNX1-ETO

leukemogenicity.” Leukemia, 2016. The dissertation author was the primary investigator

and author of this study. S.M., C.M. S.S., K.L., D.R., A.D., and M.L. assisted with

112 experiments. S.W. analyzed the data and prepared figures. S.W. and D.Z. wrote the manuscript.

References

113 114

1. Fröhling S, Döhner H. Chromosomal abnormalities in cancer. N Engl J Med 2008 Aug; 359 (7) : 722-734.

2. Korkmaz DT, Demirhan O, Abat D, Demirberk B, Tunç E, Kuleci S. Microchimeric Cells, Sex Chromosome Aneuploidies and Cancer. Pathol Oncol Res 2015 Sep; 21 (4) : 1157-1165.

3. Missiaglia E, Moore PS, Williamson J, Lemoine NR, Falconi M, Zamboni G, Scarpa A. Sex chromosome anomalies in pancreatic endocrine tumors. Int J Cancer 2002 Apr; 98 (4) : 532-538.

4. Jacobs PA, Maloney V, Cooke R, Crolla JA, Ashworth A, Swerdlow AJ. Male breast cancer, age and sex chromosome aneuploidy. Br J Cancer 2013 Mar; 108 (4) : 959-963.

5. Borah V, Shah PN, Ghosh SN, Sampat MB, Jussawalla DJ. Further studies on the prognostic importance of Barr body frequency in human breast cancer: with discussion on its probable mechanism. J Surg Oncol 1980; 13 (1) : 1-7.

6. Bottarelli L, Azzoni C, Necchi F, Lagrasta C, Tamburini E, D'Adda T, Pizzi S, Sarli L, Rindi G, Bordi C. Sex chromosome alterations associate with tumor progression in sporadic colorectal carcinomas. Clin Cancer Res 2007 Aug; 13 (15 Pt 1) : 4365-4370.

7. Cantú ES, Moses MD, Nemana LJ, Pierre RV. Sex chromosome loss in adults with haematological neoplasms. Br J Haematol 2015 Jun; 169 (6) : 899-901.

8. Huh J, Moon H, Chung WS. [Incidence and clinical significance of sex chromosome losses in bone marrow of patients with hematologic diseases]. Korean J Lab Med 2007 Feb; 27 (1) : 56-61.

9. Wong AK, Fang B, Zhang L, Guo X, Lee S, Schreck R. Loss of the Y chromosome: an age-related or clonal phenomenon in acute myelogenous leukemia/myelodysplastic syndrome? Arch Pathol Lab Med 2008 Aug; 132 (8) : 1329-1332.

10. Stone JF, Sandberg AA. Sex chromosome aneuploidy and aging. Mutat Res 1995 Oct; 338 (1-6) : 107-113.

115

11. Guttenbach M, Koschorz B, Bernthaler U, Grimm T, Schmid M. Sex chromosome loss and aging: in situ hybridization studies on human interphase nuclei. Am J Hum Genet 1995 Nov; 57 (5) : 1143-1150.

12. Pierre RV, Hoagland HC. Age-associated aneuploidy: loss of Y chromosome from human bone marrow cells with aging. Cancer 1972 Oct; 30 (4) : 889-894.

13. JACOBS PA, COURT BROWN WM, DOLL R. Distribution of human chromosome counts in relation to age. Nature 1961 Sep; 191: 1178-1180.

14. Sandberg AA, Cohen MM, Rimm AA, Levin ML. Aneuploidy and age in a population survey. Am J Hum Genet 1967 Sep; 19 (5) : 633-643.

15. Hando JC, Tucker JD, Davenport M, Tepperberg J, Nath J. X chromosome inactivation and micronuclei in normal and Turner individuals. Hum Genet 1997 Oct; 100 (5-6) : 624-628.

16. Riesch M, Niggli FK, Leibundgut K, Caflisch U, Betts DR. Loss of X chromosome in childhood acute lymphoblastic leukemia. Cancer Genet Cytogenet 2001 Feb; 125 (1) : 27-29.

17. Heerema NA, Nachman JB, Sather HN, Sensel MG, Lee MK, Hutchinson R, Lange BJ, Steinherz PG, Bostrom B, Gaynon PS, Uckun F. Hypodiploidy with less than 45 chromosomes confers adverse risk in childhood acute lymphoblastic leukemia: a report from the children's cancer group. Blood 1999 Dec; 94 (12) : 4036-4045.

18. Bakshi S, Kakadia P, Brahmbhatt M, Trivedi P, Rawal S, Bhatt S, Parikh B, Patel K, Shukla S, Shah P. Loss of sex chromosome in acute myeloid leukemia. Indian Journal of Human Genetics: Indian Journal of Human Genetics; 2004. p. 22-25.

19. Chapiro E, Antony-Debre I, Marchay N, Parizot C, Lesty C, Cung HA, Mathis S, Grelier A, Maloum K, Choquet S, Azgui Z, Uzunov M, Leblond V, Merle-Beral H, Sutton L, Davi F, Nguyen-Khac F. Sex chromosome loss may represent a disease-associated clonal population in chronic lymphocytic leukemia. Genes Chromosomes Cancer 2014 Mar; 53 (3) : 240-247.

20. Kuchenbauer F, Schnittger S, Look T, Gilliland G, Tenen D, Haferlach T, Hiddemann W, Buske C, Schoch C. Identification of additional cytogenetic and molecular genetic abnormalities in acute myeloid leukaemia with t(8;21)/AML1- ETO. Br J Haematol 2006 Sep; 134 (6) : 616-619.

116

21. Marcucci G, Mrozek K, Ruppert AS, Maharry K, Kolitz JE, Moore JO, Mayer RJ, Pettenati MJ, Powell BL, Edwards CG, Sterling LJ, Vardiman JW, Schiffer CA, Carroll AJ, Larson RA, Bloomfield CD. Prognostic factors and outcome of core binding factor acute myeloid leukemia patients with t(8;21) differ from those of patients with inv(16): a Cancer and Leukemia Group B study. J Clin Oncol 2005 Aug 20; 23 (24) : 5705-5717.

22. Yuan Y, Zhou L, Miyamoto T, Iwasaki H, Harakawa N, Hetherington CJ, Burel SA, Lagasse E, Weissman IL, Akashi K, Zhang DE. AML1-ETO expression is directly involved in the development of acute myeloid leukemia in the presence of additional mutations. Proc Natl Acad Sci U S A 2001 Aug 28; 98 (18) : 10398- 10403.

23. Briggs SF, Reijo Pera RA. X chromosome inactivation: recent advances and a look forward. Curr Opin Genet Dev 2014 Oct; 28: 78-82.

24. Rappold GA. The pseudoautosomal regions of the human sex chromosomes. Hum Genet 1993 Oct; 92 (4) : 315-324.

25. Helena Mangs A, Morris BJ. The Human Pseudoautosomal Region (PAR): Origin, Function and Future. Curr Genomics 2007 Apr; 8(2) : 129-136.

26. Veerappa AM, Padakannaya P, Ramachandra NB. Copy number variation-based polymorphism in a new pseudoautosomal region 3 (PAR3) of a human X- chromosome-transposed region (XTR) in the Y chromosome. Funct Integr Genomics 2013 Aug; 13 (3) : 285-293.

27. Nieländer I, Martín-Subero JI, Wagner F, Baudis M, Gesk S, Harder L, Hasenclever D, Klapper W, Kreuz M, Pott C, Martinez-Climent JA, Dreyling M, Arnold N, Siebert R. Recurrent loss of the Y chromosome and homozygous deletions within the pseudoautosomal region 1: association with male predominance in mantle cell lymphoma. Haematologica 2008 Jun; 93 (6) : 949- 950.

28. Mumby M. PP2A: unveiling a reluctant tumor suppressor. Cell 2007 Jul; 130 (1) : 21-24.

29. Perrotti D, Jamieson C, Goldman J, Skorski T. Chronic myeloid leukemia: mechanisms of blastic transformation. J Clin Invest 2010 Jul; 120 (7) : 2254-2264.

117

30. Neviani P, Harb JG, Oaks JJ, Santhanam R, Walker CJ, Ellis JJ, Ferenchak G, Dorrance AM, Paisie CA, Eiring AM, Ma Y, Mao HC, Zhang B, Wunderlich M, May PC, Sun C, Saddoughi SA, Bielawski J, Blum W, Klisovic RB, Solt JA, Byrd JC, Volinia S, Cortes J, Huettner CS, Koschmieder S, Holyoake TL, Devine S, Caligiuri MA, Croce CM, Garzon R, Ogretmen B, Arlinghaus RB, Chen CS, Bittman R, Hokland P, Roy DC, Milojkovic D, Apperley J, Goldman JM, Reid A, Mulloy JC, Bhatia R, Marcucci G, Perrotti D. PP2A-activating drugs selectively eradicate TKI-resistant chronic myeloid leukemic stem cells. J Clin Invest 2013 Oct; 123 (10) : 4144-4157.

31. Yang Y, Huang Q, Lu Y, Li X, Huang S. Reactivating PP2A by FTY720 as a novel therapy for AML with C-KIT tyrosine kinase domain mutation. J Cell Biochem 2012 Apr; 113 (4) : 1314-1322.

32. Ramaswamy K, Spitzer B, Kentsis A. Therapeutic Re-Activation of Protein Phosphatase 2A in Acute Myeloid Leukemia. Front Oncol 2015; 5: 16.

33. Yang Z, Cheng W, Hong L, Chen W, Wang Y, Lin S, Han J, Zhou H, Gu J. Adenine nucleotide (ADP/ATP) translocase 3 participates in the tumor necrosis factor induced apoptosis of MCF-7 cells. Mol Biol Cell 2007 Nov; 18 (11) : 4681- 4689.

34. Zamora M, Ortega JA, Alaña L, Viñas O, Mampel T. Apoptotic and anti- proliferative effects of all-trans retinoic acid. Adenine nucleotide translocase sensitizes HeLa cells to all-trans retinoic acid. Exp Cell Res 2006 Jun; 312 (10) : 1813-1819.

35. Hu Z, Guo X, Yu Q, Qiu L, Li J, Ying K, Guo C, Zhang J. Down-regulation of adenine nucleotide translocase 3 and its role in camptothecin-induced apoptosis in human hepatoma QGY7703 cells. FEBS Lett 2009 Jan; 583 (2) : 383-388.

36. Altieri A, Hemminki K. The familial risk of Hodgkin's lymphoma ranks among the highest in the Swedish Family-Cancer Database. Leukemia 2006 Nov; 20 (11) : 2062-2063.

37. Grufferman S, Cole P, Smith PG, Lukes RJ. Hodgkin's disease in siblings. N Engl J Med 1977 Feb; 296 (5) : 248-250.

38. Horwitz MS, Mealiffe ME. Further evidence for a pseudoautosomal gene for Hodgkin's lymphoma: Reply to 'The familial risk of Hodgkin's lymphoma ranks among the highest in the Swedish Family-Cancer Database' by Altieri A and Hemminki K. Leukemia 2007 Feb; 21 (2) : 351.

118

39. Büyükavci M, Ozdemir O, Buck S, Stout M, Ravindranath Y, Savaşan S. Melatonin cytotoxicity in human leukemia cells: relation with its pro-oxidant effect. Fundam Clin Pharmacol 2006 Feb; 20 (1) : 73-79.

40. Perdomo J, Cabrera J, Estévez F, Loro J, Reiter RJ, Quintana J. Melatonin induces apoptosis through a caspase-dependent but reactive oxygen species- independent mechanism in human leukemia Molt-3 cells. J Pineal Res 2013 Sep; 55 (2) : 195-206.

41. Conti A, Conconi S, Hertens E, Skwarlo-Sonta K, Markowska M, Maestroni JM. Evidence for melatonin synthesis in mouse and human bone marrow cells. J Pineal Res 2000 May; 28 (4) : 193-202.

42. Hruz T, Laule O, Szabo G, Wessendorp F, Bleuler S, Oertle L, Widmayer P, Gruissem W, Zimmermann P. Genevestigator v3: a reference expression database for the meta-analysis of transcriptomes. Adv Bioinformatics 2008; 2008: 420747.

43. Yamanishi M, Narazaki H, Asano T. Melatonin overcomes resistance to clofarabine in two leukemic cell lines by increased expression of deoxycytidine kinase. Exp Hematol 2015 Mar; 43 (3) : 207-214.

44. Sehgal AR, Konig H, Johnson DE, Tang D, Amaravadi RK, Boyiadzis M, Lotze MT. You eat what you are: autophagy inhibition as a therapeutic strategy in leukemia. Leukemia 2015 Mar; 29 (3) : 517-525.

45. Zhang G, Luo Y, Li G, Wang L, Na D, Wu X, Zhang Y, Mo X. DHRSX, a novel non-classical secretory protein associated with starvation induced autophagy. Int J Med Sci 2014; 11 (9) : 962-970.

46. Rochman Y, Kashyap M, Robinson GW, Sakamoto K, Gomez-Rodriguez J, Wagner KU, Leonard WJ. Thymic stromal lymphopoietin-mediated STAT5 phosphorylation via kinases JAK1 and JAK2 reveals a key difference from IL-7- induced signaling. Proc Natl Acad Sci U S A 2010 Nov; 107 (45) : 19455-19460.

47. Hertzberg L, Vendramini E, Ganmore I, Cazzaniga G, Schmitz M, Chalker J, Shiloh R, Iacobucci I, Shochat C, Zeligson S, Cario G, Stanulla M, Strehl S, Russell LJ, Harrison CJ, Bornhauser B, Yoda A, Rechavi G, Bercovich D, Borkhardt A, Kempski H, te Kronnie G, Bourquin JP, Domany E, Izraeli S. Down syndrome acute lymphoblastic leukemia, a highly heterogeneous disease in

119

which aberrant expression of CRLF2 is associated with mutated JAK2: a report from the International BFM Study Group. Blood 2010 Feb; 115 (5) : 1006-1017.

48. Mullighan CG, Collins-Underwood JR, Phillips LA, Loudin MG, Liu W, Zhang J, Ma J, Coustan-Smith E, Harvey RC, Willman CL, Mikhail FM, Meyer J, Carroll AJ, Williams RT, Cheng J, Heerema NA, Basso G, Pession A, Pui CH, Raimondi SC, Hunger SP, Downing JR, Carroll WL, Rabin KR. Rearrangement of CRLF2 in B-progenitor- and Down syndrome-associated acute lymphoblastic leukemia. Nat Genet 2009 Nov; 41 (11) : 1243-1246.

49. Russell LJ, Capasso M, Vater I, Akasaka T, Bernard OA, Calasanz MJ, Chandrasekaran T, Chapiro E, Gesk S, Griffiths M, Guttery DS, Haferlach C, Harder L, Heidenreich O, Irving J, Kearney L, Nguyen-Khac F, Machado L, Minto L, Majid A, Moorman AV, Morrison H, Rand V, Strefford JC, Schwab C, Tönnies H, Dyer MJ, Siebert R, Harrison CJ. Deregulated expression of cytokine receptor gene, CRLF2, is involved in lymphoid transformation in B-cell precursor acute lymphoblastic leukemia. Blood 2009 Sep; 114 (13) : 2688-2698.

50. Tasian SK, Loh ML. Understanding the biology of CRLF2-overexpressing acute lymphoblastic leukemia. Crit Rev Oncog 2011; 16 (1-2) : 13-24.

51. Testa U, Riccioni R, Diverio D, Rossini A, Lo Coco F, Peschle C. Interleukin-3 receptor in acute leukemia. Leukemia 2004 Feb; 18 (2) : 219-226.

52. Muñoz L, Nomdedéu JF, López O, Carnicer MJ, Bellido M, Aventín A, Brunet S, Sierra J. Interleukin-3 receptor alpha chain (CD123) is widely expressed in hematologic malignancies. Haematologica 2001 Dec; 86 (12) : 1261-1269.

53. Testa U, Riccioni R, Militi S, Coccia E, Stellacci E, Samoggia P, Latagliata R, Mariani G, Rossini A, Battistini A, Lo-Coco F, Peschle C. Elevated expression of IL-3Ralpha in acute myelogenous leukemia is associated with enhanced blast proliferation, increased cellularity, and poor prognosis. Blood 2002 Oct; 100 (8) : 2980-2988.

54. Busfield SJ, Biondo M, Wong M, Ramshaw HS, Lee EM, Ghosh S, Braley H, Panousis C, Roberts AW, He SZ, Thomas D, Fabri L, Vairo G, Lock RB, Lopez AF, Nash AD. Targeting of acute myeloid leukemia in vitro and in vivo with an anti-CD123 mAb engineered for optimal ADCC. Leukemia 2014 Nov; 28 (11) : 2213-2221.

55. Pizzitola I, Anjos-Afonso F, Rouault-Pierre K, Lassailly F, Tettamanti S, Spinelli O, Biondi A, Biagi E, Bonnet D. Chimeric antigen receptors against CD33/CD123

120

antigens efficiently target primary acute myeloid leukemia cells in vivo. Leukemia 2014 Aug; 28 (8) : 1596-1605.

56. Mardiros A, Dos Santos C, McDonald T, Brown CE, Wang X, Budde LE, Hoffman L, Aguilar B, Chang WC, Bretzlaff W, Chang B, Jonnalagadda M, Starr R, Ostberg JR, Jensen MC, Bhatia R, Forman SJ. T cells expressing CD123- specific chimeric antigen receptors exhibit specific cytolytic effector functions and antitumor effects against human acute myeloid leukemia. Blood 2013 Oct; 122 (18) : 3138-3148.

57. Hansen G, Hercus TR, McClure BJ, Stomski FC, Dottore M, Powell J, Ramshaw H, Woodcock JM, Xu Y, Guthridge M, McKinstry WJ, Lopez AF, Parker MW. The structure of the GM-CSF receptor complex reveals a distinct mode of cytokine receptor activation. Cell 2008; 134 (3) : 496-507.

58. de Vries AC, Zwaan CM, van den Heuvel-Eibrink MM. Molecular basis of juvenile myelomonocytic leukemia. Haematologica 2010 Feb; 95 (2) : 179-182.

59. Padron E, Painter JS, Kunigal S, Mailloux AW, McGraw K, McDaniel JM, Kim E, Bebbington C, Baer M, Yarranton G, Lancet J, Komrokji RS, Abdel-Wahab O, List AF, Epling-Burnette PK. GM-CSF-dependent pSTAT5 sensitivity is a feature with therapeutic potential in chronic myelomonocytic leukemia. Blood 2013 Jun; 121 (25) : 5068-5077.

60. Zhang J, Ranheim EA, Du J, Liu Y, Wang J, Kong G. Deficiency of beta common receptor moderately attenuates the progression of myeloproliferative neoplasm in Nras G12D/+ mice. J Biol Chem 2015 Jun.

61. Matsuura S, Yan M, Lo MC, Ahn EY, Weng S, Dangoor D, Matin M, Higashi T, Feng GS, Zhang DE. Negative effects of GM-CSF signaling in a murine model of t(8;21)-induced leukemia. Blood 2012 Mar 29; 119 (13) : 3155-3163.

62. Fujiwara S, Yamashita Y, Choi YL, Watanabe H, Kurashina K, Soda M, Enomoto M, Hatanaka H, Takada S, Ozawa K, Mano H. Transforming activity of purinergic receptor P2Y, G protein coupled, 8 revealed by retroviral expression screening. Leuk Lymphoma 2007 May; 48 (5) : 978-986.

63. Muppidi JR, Schmitz R, Green JA, Xiao W, Larsen AB, Braun SE, An J, Xu Y, Rosenwald A, Ott G, Gascoyne RD, Rimsza LM, Campo E, Jaffe ES, Delabie J, Smeland EB, Braziel RM, Tubbs RR, Cook JR, Weisenburger DD, Chan WC, Vaidehi N, Staudt LM, Cyster JG. Loss of signalling via Gα13 in germinal centre B-cell-derived lymphoma. Nature 2014 Dec; 516 (7530) : 254-258.

121

64. Goswami R, Kaplan MH. A brief history of IL-9. J Immunol 2011 Mar; 186 (6) : 3283-3288.

65. Hornakova T, Staerk J, Royer Y, Flex E, Tartaglia M, Constantinescu SN, Knoops L, Renauld JC. Acute lymphoblastic leukemia-associated JAK1 mutants activate the Janus kinase/STAT pathway via interleukin-9 receptor alpha homodimers. J Biol Chem 2009 Mar; 284 (11) : 6773-6781.

66. Lu Y, Hong S, Li H, Park J, Hong B, Wang L, Zheng Y, Liu Z, Xu J, He J, Yang J, Qian J, Yi Q. Th9 cells promote antitumor immune responses in vivo. J Clin Invest 2012 Nov; 122 (11) : 4160-4171.

67. Purwar R, Schlapbach C, Xiao S, Kang HS, Elyaman W, Jiang X, Jetten AM, Khoury SJ, Fuhlbrigge RC, Kuchroo VK, Clark RA, Kupper TS. Robust tumor immunity to melanoma mediated by interleukin-9-producing T cells. Nat Med 2012 Aug; 18 (8) : 1248-1253.

68. Dworzak MN, Fröschl G, Printz D, Zen LD, Gaipa G, Ratei R, Basso G, Biondi A, Ludwig WD, Gadner H. CD99 expression in T-lineage ALL: implications for flow cytometric detection of minimal residual disease. Leukemia 2004 Apr; 18 (4) : 703- 708.

69. Husak Z, Printz D, Schumich A, Pötschger U, Dworzak MN. Death induction by CD99 ligation in TEL/AML1-positive acute lymphoblastic leukemia and normal B cell precursors. J Leukoc Biol 2010 Aug; 88 (2) : 405-412.

70. Bernard G, Breittmayer JP, de Matteis M, Trampont P, Hofman P, Senik A, Bernard A. Apoptosis of immature thymocytes mediated by E2/CD99. J Immunol 1997 Mar; 158 (6) : 2543-2550.

71. Pettersen RD, Bernard G, Olafsen MK, Pourtein M, Lie SO. CD99 signals caspase-independent T cell death. J Immunol 2001 Apr; 166 (8) : 4931-4942.

72. Matarazzo MR, De Bonis ML, Gregory RI, Vacca M, Hansen RS, Mercadante G, D'Urso M, Feil R, D'Esposito M. Allelic inactivation of the pseudoautosomal gene SYBL1 is controlled by epigenetic mechanisms common to the X and Y chromosomes. Hum Mol Genet 2002 Dec; 11 (25) : 3191-3198.

122

73. De Bonis ML, Cerase A, Matarazzo MR, Ferraro M, Strazzullo M, Hansen RS, Chiurazzi P, Neri G, D'Esposito M. Maintenance of X- and Y-inactivation of the pseudoautosomal (PAR2) gene SPRY3 is independent from DNA methylation and associated to multiple layers of epigenetic modifications. Hum Mol Genet 2006 Apr; 15 (7) : 1123-1132.

74. Hacohen N, Kramer S, Sutherland D, Hiromi Y, Krasnow MA. sprouty encodes a novel antagonist of FGF signaling that patterns apical branching of the Drosophila airways. Cell 1998 Jan; 92 (2) : 253-263.

75. Moreau K, Rubinsztein DC. The plasma membrane as a control center for autophagy. Autophagy 2012 May; 8(5) : 861-863.

76. Moreau K, Ravikumar B, Renna M, Puri C, Rubinsztein DC. Autophagosome precursor maturation requires homotypic fusion. Cell 2011 Jul; 146 (2) : 303-317.

77. Krzewski K, Gil-Krzewska A, Watts J, Stern JN, Strominger JL. VAMP4- and VAMP7-expressing vesicles are both required for cytotoxic granule exocytosis in NK cells. Eur J Immunol 2011 Nov; 41 (11) : 3323-3329.

78. Larghi P, Williamson DJ, Carpier JM, Dogniaux S, Chemin K, Bohineust A, Danglot L, Gaus K, Galli T, Hivroz C. VAMP7 controls T cell activation by regulating the recruitment and phosphorylation of vesicular Lat at TCR-activation sites. Nat Immunol 2013 Jul; 14 (7) : 723-731.

79. Mithani SK, Smith IM, Califano JA. Use of integrative epigenetic and cytogenetic analyses to identify novel tumor-suppressor genes in malignant melanoma. Melanoma Res 2011 Aug; 21 (4) : 298-307.

80. Ried K, Rao E, Schiebel K, Rappold GA. Gene duplications as a recurrent theme in the evolution of the human pseudoautosomal region 1: isolation of the gene ASMTL. Hum Mol Genet 1998 Oct; 7(11) : 1771-1778.

81. Zhang J, Mullighan CG, Harvey RC, Wu G, Chen X, Edmonson M, Buetow KH, Carroll WL, Chen IM, Devidas M, Gerhard DS, Loh ML, Reaman GH, Relling MV, Camitta BM, Bowman WP, Smith MA, Willman CL, Downing JR, Hunger SP. Key pathways are frequently mutated in high-risk childhood acute lymphoblastic leukemia: a report from the Children's Oncology Group. Blood 2011 Sep; 118 (11) : 3080-3087.

123

82. Jarnaess E, Stokka AJ, Kvissel AK, Skålhegg BS, Torgersen KM, Scott JD, Carlson CR, Taskén K. Splicing factor arginine/serine-rich 17A (SFRS17A) is an A-kinase anchoring protein that targets protein kinase A to splicing factor compartments. J Biol Chem 2009 Dec; 284 (50) : 35154-35164.

83. Mangs AH, Speirs HJ, Goy C, Adams DJ, Markus MA, Morris BJ. XE7: a novel splicing factor that interacts with ASF/SF2 and ZNF265. Nucleic Acids Res 2006; 34 (17) : 4976-4986.

84. Matsukage A, Hirose F, Yoo MA, Yamaguchi M. The DRE/DREF transcriptional regulatory system: a master key for cell proliferation. Biochim Biophys Acta 2008 Feb; 1779 (2) : 81-89.

85. Ohshima N, Takahashi M, Hirose F. Identification of a human homologue of the DREF transcription factor with a potential role in regulation of the histone H1 gene. J Biol Chem 2003 Jun; 278 (25) : 22928-22938.

86. Ellis NA, Tippett P, Petty A, Reid M, Weller PA, Ye TZ, German J, Goodfellow PN, Thomas S, Banting G. PBDX is the XG blood group gene. Nat Genet 1994 Nov; 8(3) : 285-290.

87. Gelin C, Aubrit F, Phalipon A, Raynal B, Cole S, Kaczorek M, Bernard A. The E2 antigen, a 32 kd glycoprotein involved in T-cell adhesion processes, is the MIC2 gene product. EMBO J 1989 Nov; 8(11) : 3253-3259.

88. Fouchet C, Gane P, Cartron JP, Lopez C. Quantitative analysis of XG blood group and CD99 antigens on human red cells. Immunogenetics 2000 Jul; 51 (8- 9) : 688-694.

89. Fouchet C, Gane P, Huet M, Fellous M, Rouger P, Banting G, Cartron JP, Lopez C. A study of the coregulation and tissue specificity of XG and MIC2 gene expression in eukaryotic cells. Blood 2000 Mar; 95 (5) : 1819-1826.

90. Mulloy JC, Cancelas JA, Filippi MD, Kalfa TA, Guo F, Zheng Y. Rho GTPases in hematopoiesis and hemopathies. Blood 2010 Feb; 115 (5) : 936-947.

91. Gianfrancesco F, Esposito T, Montanini L, Ciccodicola A, Mumm S, Mazzarella R, Rao E, Giglio S, Rappold G, Forabosco A. A novel pseudoautosomal gene encoding a putative GTP-binding protein resides in the vicinity of the Xp/Yp telomere. Hum Mol Genet 1998 Mar; 7(3) : 407-414.

124

92. Konishi K, Morishima Y, Ueda E, Kibe Y, Nonomura K, Yamanishi K, Yasuno H. Cataloging of the genes expressed in human keratinocytes: analysis of 607 randomly isolated cDNA sequences. Biochem Biophys Res Commun 1994 Jul; 202 (2): 976-983.

93. Strausberg RL, Feingold EA, Grouse LH, Derge JG, Klausner RD, Collins FS, Wagner L, Shenmen CM, Schuler GD, Altschul SF, Zeeberg B, Buetow KH, Schaefer CF, Bhat NK, Hopkins RF, Jordan H, Moore T, Max SI, Wang J, Hsieh F, Diatchenko L, Marusina K, Farmer AA, Rubin GM, Hong L, Stapleton M, Soares MB, Bonaldo MF, Casavant TL, Scheetz TE, Brownstein MJ, Usdin TB, Toshiyuki S, Carninci P, Prange C, Raha SS, Loquellano NA, Peters GJ, Abramson RD, Mullahy SJ, Bosak SA, McEwan PJ, McKernan KJ, Malek JA, Gunaratne PH, Richards S, Worley KC, Hale S, Garcia AM, Gay LJ, Hulyk SW, Villalon DK, Muzny DM, Sodergren EJ, Lu X, Gibbs RA, Fahey J, Helton E, Ketteman M, Madan A, Rodrigues S, Sanchez A, Whiting M, Young AC, Shevchenko Y, Bouffard GG, Blakesley RW, Touchman JW, Green ED, Dickson MC, Rodriguez AC, Grimwood J, Schmutz J, Myers RM, Butterfield YS, Krzywinski MI, Skalska U, Smailus DE, Schnerch A, Schein JE, Jones SJ, Marra MA, Team MGCP. Generation and initial analysis of more than 15,000 full-length human and mouse cDNA sequences. Proc Natl Acad Sci U S A 2002 Dec; 99 (26) : 16899-16903.

94. Yamamoto K, Nagata K, Kida A, Hamaguchi H. Acquired gain of an X chromosome as the sole abnormality in the blast crisis of chronic neutrophilic leukemia. Cancer Genet Cytogenet 2002 Apr; 134 (1) : 84-87.

95. Ozkaynak MF, Ying KL, Laug WE. Triple X chromosome constitution and acute nonlymphocytic leukemia. Cancer Genet Cytogenet 1988 Oct; 35 (1) : 1-3.

96. Peterson LF, Boyapati A, Ahn E-Y, Biggs JR, Okumura AJ, Lo M-C, Yan M, Zhang D-E. Acute myeloid leukemia with the 8q22;21q22 translocation: secondary mutational events and alternative t(8;21) transcripts. Blood 2007; 110 (3) : 799-805.

97. Erickson P, Gao J, Chang KS, Look T, Whisenant E, Raimondi S, Lasher R, Trujillo J, Rowley J, Drabkin H. Identification of breakpoints in t(8;21) acute myelogenous leukemia and isolation of a fusion transcript, AML1/ETO, with similarity to Drosophila segmentation gene, runt. Blood 1992 Oct 1; 80 (7) : 1825- 1831.

98. Nishii K, Usui E, Katayama N, Lorenzo F, Nakase K, Kobayashi T, Miwa H, Mizutani M, Tanaka I, Nasu K, Dohy H, Kyo T, Taniwaki M, Ueda T, Kita K, Shiku H. Characteristics of t(8;21) acute myeloid leukemia (AML) with additional

125

chromosomal abnormality: concomitant trisomy 4 may constitute a distinctive subtype of t(8;21) AML. Leukemia 2003 Apr; 17 (4) : 731-737.

99. Byrd JC, Mrózek K, Dodge RK, Carroll AJ, Edwards CG, Arthur DC, Pettenati MJ, Patil SR, Rao KW, Watson MS, Koduru PR, Moore JO, Stone RM, Mayer RJ, Feldman EJ, Davey FR, Schiffer CA, Larson RA, Bloomfield CD, 8461) CaLGBC. Pretreatment cytogenetic abnormalities are predictive of induction success, cumulative incidence of relapse, and overall survival in adult patients with de novo acute myeloid leukemia: results from Cancer and Leukemia Group B (CALGB 8461). Blood 2002 Dec; 100 (13) : 4325-4336.

100. Roboz GJ. Current treatment of acute myeloid leukemia. Curr Opin Oncol 2012 Nov; 24 (6) : 711-719.

101. Lam K, Zhang DE. RUNX1 and RUNX1-ETO: roles in hematopoiesis and leukemogenesis. Front Biosci (Landmark Ed) 2012; 17: 1120-1139.

102. Pabst T, Mueller BU, Harakawa N, Schoch C, Haferlach T, Behre G, Hiddemann W, Zhang DE, Tenen DG. AML1-ETO downregulates the granulocytic differentiation factor C/EBPalpha in t(8;21) myeloid leukemia. Nat Med 2001 Apr; 7(4) : 444-451.

103. Vangala RK, Heiss-Neumann MS, Rangatia JS, Singh SM, Schoch C, Tenen DG, Hiddemann W, Behre G. The myeloid master regulator transcription factor PU.1 is inactivated by AML1-ETO in t(8;21) myeloid leukemia. Blood 2003 Jan; 101 (1) : 270-277.

104. Higuchi M, O'Brien D, Kumaravelu P, Lenny N, Yeoh EJ, Downing JR. Expression of a conditional AML1-ETO oncogene bypasses embryonic lethality and establishes a murine model of human t(8;21) acute myeloid leukemia. Cancer Cell 2002 Feb; 1(1) : 63-74.

105. Okuda T, Cai Z, Yang S, Lenny N, Lyu CJ, van Deursen JM, Harada H, Downing JR. Expression of a knocked-in AML1-ETO leukemia gene inhibits the establishment of normal definitive hematopoiesis and directly generates dysplastic hematopoietic progenitors. Blood 1998 May 1; 91 (9) : 3134-3143.

106. Mulloy JC, Cammenga J, MacKenzie KL, Berguido FJ, Moore MA, Nimer SD. The AML1-ETO fusion protein promotes the expansion of human hematopoietic stem cells. Blood 2002 Jan; 99 (1) : 15-23.

126

107. Appelbaum FR, Kopecky KJ, Tallman MS, Slovak ML, Gundacker HM, Kim HT, Dewald GW, Kantarjian HM, Pierce SR, Estey EH. The clinical spectrum of adult acute myeloid leukaemia associated with core binding factor translocations. Br J Haematol 2006 Oct; 135 (2) : 165-173.

108. Valk PJM, Verhaak RGW, Beijen MA, Erpelinck CAJ, Barjesteh van Waalwijk van Doorn-Khosrovani S, Boer JM, Beverloo HB, Moorhouse MJ, van der Spek PJ, Löwenberg B, Delwel R. Prognostically useful gene-expression profiles in acute myeloid leukemia. The New England Journal of Medicine 2004; 350 (16) : 1617- 1628.

109. Wouters BJ, Löwenberg B, Erpelinck-Verschueren CA, van Putten WL, Valk PJ, Delwel R. Double CEBPA mutations, but not single CEBPA mutations, define a subgroup of acute myeloid leukemia with a distinctive gene expression profile that is uniquely associated with a favorable outcome. Blood 2009 Mar; 113 (13) : 3088-3091.

110. de Groot RP, Coffer PJ, Koenderman L. Regulation of proliferation, differentiation and survival by the IL-3/IL-5/GM-CSF receptor family. Cellular Signalling 1998; 10 (9) : 619-628.

111. Perugini M, Brown AL, Salerno DG, Booker GW, Stojkoski C, Hercus TR, Lopez AF, Hibbs ML, Gonda TJ, D'Andrea RJ. Alternative modes of GM-CSF receptor activation revealed using activated mutants of the common beta-subunit. Blood 2010 Apr 22; 115 (16) : 3346-3353.

112. Young DC, Griffin JD. Autocrine secretion of GM-CSF in acute myeloblastic leukemia. Blood 1986 Nov; 68 (5) : 1178-1181.

113. Lanza F, Castagnari B, Rigolin G, Moretti S, Latorraca A, Ferrari L, Bardi A, Castoldi G. Flow cytometry measurement of GM-CSF receptors in acute leukemic blasts, and normal hemopoietic cells. Leukemia: Official Journal of the Leukemia Society of America, Leukemia Research Fund, UK 1997; 11 (10) : 1700-1710.

114. Kita K, Shirakawa S, Kamada N. Cellular characteristics of acute myeloblastic leukemia associated with t(8;21)(q22;q22). The Japanese Cooperative Group of Leukemia/Lymphoma. Leuk Lymphoma 1994 Apr; 13 (3-4) : 229-234.

115. Jahns-Streubel G, Braess J, Schoch C, Fonatsch C, Haase D, Binder C, Wörmann B, Büchner T, Hiddemann W. Cytogenetic subgroups in acute myeloid leukemia differ in proliferative activity and response to GM-CSF. Leukemia:

127

Official Journal of the Leukemia Society of America, Leukemia Research Fund, UK 2001; 15 (3) : 377-384.

116. Frank R, Zhang J, Uchida H, Meyers S, Hiebert SW, Nimer SD. The AML1/ETO fusion protein blocks transactivation of the GM-CSF promoter by AML1B. Oncogene 1995 Dec; 11 (12) : 2667-2674.

117. Liu Y, Chen W, Gaudet J, Cheney MD, Roudaia L, Cierpicki T, Klet RC, Hartman K, Laue TM, Speck NA, Bushweller JH. Structural basis for recognition of SMRT/N-CoR by the MYND domain and its contribution to AML1/ETO's activity. Cancer Cell 2007; 11 (6) : 483-497.

118. Brown AL, Peters M, D'Andrea RJ, Gonda TJ. Constitutive mutants of the GM- CSF receptor reveal multiple pathways leading to myeloid cell survival, proliferation, and granulocyte-macrophage differentiation. Blood 2004 Jan 15; 103 (2) : 507-516.

119. Ferrari F, Bortoluzzi S, Coppe A, Basso D, Bicciato S, Zini R, Gemelli C, Danieli GA, Ferrari S. Genomic expression during human myelopoiesis. BMC Genomics 2007; 8: 264.

120. Gery S, Gombart AF, Yi WS, Koeffler C, Hofmann WK, Koeffler HP. Transcription profiling of C/EBP targets identifies Per2 as a gene implicated in myeloid leukemia. Blood 2005 Oct; 106 (8) : 2827-2836.

121. Ma O, Hong S, Guo H, Ghiaur G, Friedman AD. Granulopoiesis requires increased C/EBPα compared to monopoiesis, correlated with elevated Cebpa in immature G-CSF receptor versus M-CSF receptor expressing cells. PLoS One 2014; 9(4) : e95784.

122. Mulloy JC, Cammenga J, Berguido FJ, Wu K, Zhou P, Comenzo RL, Jhanwar S, Moore MA, Nimer SD. Maintaining the self-renewal and differentiation potential of human CD34+ hematopoietic cells using a single genetic element. Blood 2003 Dec; 102 (13) : 4369-4376.

123. Gu L, Chiang KY, Zhu N, Findley HW, Zhou M. Contribution of STAT3 to the activation of survivin by GM-CSF in CD34+ cell lines. Exp Hematol 2007 Jun; 35 (6) : 957-966.

124. Sutherland HJ, Lansdorp PM, Henkelman DH, Eaves AC, Eaves CJ. Functional characterization of individual human hematopoietic stem cells cultured at limiting

128

dilution on supportive marrow stromal layers. Proc Natl Acad Sci U S A 1990 May; 87 (9) : 3584-3588.

125. Ragione FD, Iolascon A. Inactivation of cyclin-dependent kinase inhibitor genes and development of human acute leukemias. Leuk Lymphoma 1997 Mar; 25 (1- 2) : 23-35.

126. Wu C, Orozco C, Boyer J, Leglise M, Goodale J, Batalov S, Hodge CL, Haase J, Janes J, Huss JW, Su AI. BioGPS: an extensible and customizable portal for querying and organizing gene annotation resources. Genome Biol 2009; 10 (11) : R130.

127. Ziboh VA, Wong T, Wu MC, Yunis AA. Lipoxygenation of arachidonic acid by differentiated and undifferentiated human promyelocytic HL-60 cells. J Lab Clin Med 1986 Aug; 108 (2) : 161-166.

128. Goldman DW, Olson DM, Payan DG, Gifford LA, Goetzl EJ. Development of receptors for leukotriene B4 on HL-60 cells induced to differentiate by 1 alpha,25- dihydroxyvitamin D3. J Immunol 1986 Jun; 136 (12) : 4631-4636.

129. Chung JW, Kim GY, Mun YC, Ahn JY, Seong CM, Kim JH. Leukotriene B4 pathway regulates the fate of the hematopoietic stem cells. Exp Mol Med 2005 Feb; 37 (1) : 45-50.

130. Haferlach T, Kohlmann A, Wieczorek L, Basso G, Kronnie GT, Béné MC, De Vos J, Hernández JM, Hofmann WK, Mills KI, Gilkes A, Chiaretti S, Shurtleff SA, Kipps TJ, Rassenti LZ, Yeoh AE, Papenhausen PR, Liu WM, Williams PM, Foà R. Clinical utility of microarray-based gene expression profiling in the diagnosis and subclassification of leukemia: report from the International Microarray Innovations in Leukemia Study Group. J Clin Oncol 2010 May; 28 (15) : 2529- 2537.

131. Sharma JN, Mohammed LA. The role of leukotrienes in the pathophysiology of inflammatory disorders: is there a case for revisiting leukotrienes as therapeutic targets? Inflammopharmacology 2006 Mar; 14 (1-2) : 10-16.

132. Yang G, Khalaf W, van de Locht L, Jansen JH, Gao M, Thompson MA, van der Reijden BA, Gutmann DH, Delwel R, Clapp DW, Hiebert SW. Transcriptional repression of the Neurofibromatosis-1 tumor suppressor by the t(8;21) fusion protein. Mol Cell Biol 2005 Jul; 25 (14) : 5869-5879.

129

133. Yan M, Burel SA, Peterson LF, Kanbe E, Iwasaki H, Boyapati A, Hines R, Akashi K, Zhang DE. Deletion of an AML1-ETO C-terminal NcoR/SMRT-interacting region strongly induces leukemia development. Proc Natl Acad Sci U S A 2004 Dec; 101 (49) : 17186-17191.

134. Li C, Hung Wong W. Model-based analysis of oligonucleotide arrays: model validation, design issues and standard error application. Genome Biol 2001; 2(8) : RESEARCH0032.

135. Simon R, Lam A, Li M-C, Ngan M, Menenzes S, Zhao Y. Analysis of gene expression data using BRB-array tools. Cancer Informatics; 2007. p. 11-17.

136. Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, Paulovich A, Pomeroy SL, Golub TR, Lander ES, Mesirov JP. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A 2005 Oct; 102 (43) : 15545-15550.

137. Brown AL, Wilkinson CR, Waterman SR, Kok CH, Salerno DG, Diakiw SM, Reynolds B, Scott HS, Tsykin A, Glonek GF, Goodall GJ, Solomon PJ, Gonda TJ, D'Andrea RJ. Genetic regulators of myelopoiesis and leukemic signaling identified by gene profiling and linear modeling. J Leukoc Biol 2006 Aug; 80 (2) : 433-447.

138. Ido M, Hayashi K, Kato S, Ogawa H, Komada Y, Zhau YW, Zhang XL, Sakurai M, Suzuki K. Isolation and characterisation of Kasumi-1 human myeloid leukaemia cell line resistant to tumour necrosis factor alpha-induced apoptosis. Br J Cancer 1996 Feb; 73 (3) : 360-365.

139. Oehler L, Berer A, Kollars M, Keil F, König M, Waclavicek M, Haas O, Knapp W, Lechner K, Geissler K. Culture requirements for induction of dendritic cell differentiation in acute myeloid leukemia. Ann Hematol 2000 Jul; 79 (7) : 355-362.

140. Tsuchiya T, Hagihara M, Shimakura Y, Ueda Y, Gansuvd B, Munkhbat B, Inoue H, Tazume K, Kato S, Hotta T. The generation of immunocompetent dendritic cells from CD34+ acute myeloid or lymphoid leukemia cells. Int J Hematol 2002 Jan; 75 (1) : 55-62.

141. Meyer N, Penn LZ. Reflecting on 25 years with MYC. Nat Rev Cancer 2008 Dec; 8(12) : 976-990.

130

142. Delgado MD, León J. Myc roles in hematopoiesis and leukemia. Genes Cancer 2010 Jun; 1(6) : 605-616.

143. Gowda SD, Koler RD, Bagby GC. Regulation of C-myc expression during growth and differentiation of normal and leukemic human myeloid progenitor cells. J Clin Invest 1986 Jan; 77 (1) : 271-278.

144. Zeller KI, Jegga AG, Aronow BJ, O'Donnell KA, Dang CV. An integrated database of genes responsive to the Myc oncogenic transcription factor: identification of direct genomic targets. Genome Biol 2003; 4(10) : R69.

145. Dang CV. c-Myc target genes involved in cell growth, apoptosis, and metabolism. Mol Cell Biol 1999 Jan; 19 (1) : 1-11.

146. Müller-Tidow C, Steffen B, Cauvet T, Tickenbrock L, Ji P, Diederichs S, Sargin B, Köhler G, Stelljes M, Puccetti E, Ruthardt M, deVos S, Hiebert SW, Koeffler HP, Berdel WE, Serve H. Translocation products in acute myeloid leukemia activate the Wnt signaling pathway in hematopoietic cells. Mol Cell Biol 2004 Apr; 24 (7) : 2890-2904.

147. Salat D, Liefke R, Wiedenmann J, Borggrefe T, Oswald F. ETO, but not leukemogenic fusion protein AML1/ETO, augments RBP-Jkappa/SHARP- mediated repression of notch target genes. Mol Cell Biol 2008 May; 28 (10) : 3502-3512.

148. Schreiber-Agus N, Chin L, Chen K, Torres R, Rao G, Guida P, Skoultchi AI, DePinho RA. An amino-terminal domain of Mxi1 mediates anti-Myc oncogenic activity and interacts with a homolog of the yeast transcriptional repressor SIN3. Cell 1995 Mar; 80 (5) : 777-786.

149. Alland L, Muhle R, Hou H, Potes J, Chin L, Schreiber-Agus N, DePinho RA. Role for N-CoR and histone deacetylase in Sin3-mediated transcriptional repression. Nature 1997 May; 387 (6628) : 49-55.

150. Licht JD. AML1 and the AML1-ETO fusion protein in the pathogenesis of t(8;21) AML. Oncogene 2001 Sep; 20 (40) : 5660-5679.

151. Ellwood-Yen K, Graeber TG, Wongvipat J, Iruela-Arispe ML, Zhang J, Matusik R, Thomas GV, Sawyers CL. Myc-driven murine prostate cancer shares molecular features with human prostate tumors. Cancer Cell 2003 Sep; 4(3) : 223-238.

131

152. Zervos AS, Gyuris J, Brent R. Mxi1, a protein that specifically interacts with Max to bind Myc-Max recognition sites. Cell 1994 Oct; 79 (2) : following 388.

153. Asou H, Tashiro S, Hamamoto K, Otsuji A, Kita K, Kamada N. Establishment of a human acute myeloid leukemia cell line (Kasumi-1) with 8; 21 chromosome translocation. Blood 1991; 77 (9) : 2031-2031.

154. Matozaki S, Nakagawa T, Kawaguchi R, Aozaki R, Tsutsumi M, Murayama T, Koizumi T, Nishimura R, Isobe T, Chihara K. Establishment of a myeloid leukaemic cell line (SKNO-1) from a patient with t(8;21) who acquired monosomy 17 during disease progression. Br J Haematol 1995 Apr; 89 (4) : 805-811.

155. Brown MA, Harrison-Smith M, DeLuca E, Begley CG, Gough NM. No evidence for GM-CSF receptor alpha chain gene mutation in AML-M2 leukemias which have lost a sex chromosome. Leukemia 1994 Oct; 8(10) : 1774-1779.

156. Westendorf JJ, Yamamoto CM, Lenny N, Downing JR, Selsted ME, Hiebert SW. The t(8;21) fusion product, AML-1-ETO, associates with C/EBP-alpha, inhibits C/EBP-alpha-dependent transcription, and blocks granulocytic differentiation. Mol Cell Biol 1998 Jan; 18 (1): 322-333.

157. Liebermann DA, Tront JS, Sha X, Mukherjee K, Mohamed-Hadley A, Hoffman B. Gadd45 stress sensors in malignancy and leukemia. Crit Rev Oncog 2011; 16 (1- 2) : 129-140.

158. Perugini M, Iarossi DG, Kok CH, Cummings N, Diakiw SM, Brown AL, Danner S, Bardy P, Bik To L, Wei AH, Lewis ID, D'Andrea RJ. GADD45A methylation predicts poor overall survival in acute myeloid leukemia and is associated with IDH1/2 and DNMT3A mutations. Leukemia 2013 Jul; 27 (7) : 1588-1592.

159. Mertz JA, Conery AR, Bryant BM, Sandy P, Balasubramanian S, Mele DA, Bergeron L, Sims RJ. Targeting MYC dependence in cancer by inhibiting BET bromodomains. Proc Natl Acad Sci U S A 2011 Oct; 108 (40) : 16669-16674.

160. Delmore JE, Issa GC, Lemieux ME, Rahl PB, Shi J, Jacobs HM, Kastritis E, Gilpatrick T, Paranal RM, Qi J, Chesi M, Schinzel AC, McKeown MR, Heffernan TP, Vakoc CR, Bergsagel PL, Ghobrial IM, Richardson PG, Young RA, Hahn WC, Anderson KC, Kung AL, Bradner JE, Mitsiades CS. BET bromodomain inhibition as a therapeutic strategy to target c-Myc. Cell 2011 Sep; 146 (6) : 904- 917.

132

161. Fowler T, Ghatak P, Price DH, Conaway R, Conaway J, Chiang CM, Bradner JE, Shilatifard A, Roy AL. Regulation of MYC expression and differential JQ1 sensitivity in cancer cells. PLoS One 2014; 9(1) : e87003.

162. Zuber J, Shi J, Wang E, Rappaport AR, Herrmann H, Sison EA, Magoon D, Qi J, Blatt K, Wunderlich M, Taylor MJ, Johns C, Chicas A, Mulloy JC, Kogan SC, Brown P, Valent P, Bradner JE, Lowe SW, Vakoc CR. RNAi screen identifies Brd4 as a therapeutic target in acute myeloid leukaemia. Nature 2011 Oct; 478 (7370) : 524-528.

163. Peterson LF, Zhang DE. The 8;21 translocation in leukemogenesis. Oncogene 2004 May; 23 (24) : 4255-4262.

164. Heinzel T, Lavinsky RM, Mullen TM, Söderstrom M, Laherty CD, Torchia J, Yang WM, Brard G, Ngo SD, Davie JR, Seto E, Eisenman RN, Rose DW, Glass CK, Rosenfeld MG. A complex containing N-CoR, mSin3 and histone deacetylase mediates transcriptional repression. Nature 1997 May; 387 (6628) : 43-48.

165. Wang J, Hoshino T, Redner RL, Kajigaya S, Liu JM. ETO, fusion partner in t(8;21) acute myeloid leukemia, represses transcription by interaction with the human N-CoR/mSin3/HDAC1 complex. Proceedings of the National Academy of Sciences of the United States of America 1998; 95 (18) : 10860-10865.

166. Lutterbach B, Westendorf JJ, Linggi B, Patten A, Moniwa M, Davie JR, Huynh KD, Bardwell VJ, Lavinsky RM, Rosenfeld MG, Glass C, Seto E, Hiebert SW. ETO, a target of t(8;21) in acute leukemia, interacts with the N-CoR and mSin3 corepressors. Mol Cell Biol 1998 Dec; 18 (12) : 7176-7184.

167. Okumura AJ, Peterson LF, Okumura F, Boyapati A, Zhang DE. t(8;21)(q22;q22) Fusion proteins preferentially bind to duplicated AML1/RUNX1 DNA-binding sequences to differentially regulate gene expression. Blood 2008 Aug; 112 (4) : 1392-1401.

168. Hohaus S, Petrovick MS, Voso MT, Sun Z, Zhang DE, Tenen DG. PU.1 (Spi-1) and C/EBP alpha regulate expression of the granulocyte-macrophage colony- stimulating factor receptor alpha gene. Mol Cell Biol 1995 Oct; 15 (10) : 5830- 5845.

169. Mignone F, Gissi C, Liuni S, Pesole G. Untranslated regions of mRNAs. Genome Biol 2002; 3(3) : REVIEWS0004.

133

170. Croce CM. Causes and consequences of microRNA dysregulation in cancer. Nat Rev Genet 2009 Oct; 10 (10) : 704-714.

171. Mayr C, Bartel DP. Widespread shortening of 3'UTRs by alternative cleavage and polyadenylation activates oncogenes in cancer cells. Cell 2009 Aug; 138 (4) : 673-684.

172. Gasson JC, Kaufman SE, Weisbart RH, Tomonaga M, Golde DW. High-affinity binding of granulocyte-macrophage colony-stimulating factor to normal and leukemic human myeloid cells. Proc Natl Acad Sci U S A 1986 Feb; 83 (3) : 669- 673.

173. DiPersio J, Billing P, Kaufman S, Eghtesady P, Williams RE, Gasson JC. Characterization of the human granulocyte-macrophage colony-stimulating factor receptor. J Biol Chem 1988 Feb; 263 (4) : 1834-1841.

174. Wognum AW, Westerman Y, Visser TP, Wagemaker G. Distribution of receptors for granulocyte-macrophage colony-stimulating factor on immature CD34+ bone marrow cells, differentiating monomyeloid progenitors, and mature blood cell subsets. Blood 1994 Aug; 84 (3) : 764-774.

175. Rücker FG, Sander S, Döhner K, Döhner H, Pollack JR, Bullinger L. Molecular profiling reveals myeloid leukemia cell lines to be faithful model systems characterized by distinct genomic aberrations. Leukemia: Official Journal of the Leukemia Society of America, Leukemia Research Fund, UK 2006; 20 (6) : 994- 1001.

176. Savant-Bhonsale S, Cleveland DW. Evidence for instability of mRNAs containing AUUUA motifs mediated through translation-dependent assembly of a > 20S degradation complex. Genes Dev 1992 Oct; 6(10) : 1927-1939.

177. Lagnado CA, Brown CY, Goodall GJ. AUUUA is not sufficient to promote poly(A) shortening and degradation of an mRNA: the functional sequence within AU-rich elements may be UUAUUUA(U/A)(U/A). Mol Cell Biol 1994 Dec; 14 (12) : 7984- 7995.

178. Barreau C, Paillard L, Osborne HB. AU-rich elements and associated factors: are there unifying principles? Nucleic Acids Res 2005; 33 (22) : 7138-7150.

179. Temin HM. Function of the retrovirus long terminal repeat. Cell 1982 Jan; 28 (1) : 3-5.

134

180. Sanjana NE, Shalem O, Zhang F. Improved vectors and genome-wide libraries for CRISPR screening. Nat Methods 2014 Aug; 11 (8) : 783-784.

181. Liu N, Pan T. N6-methyladenosine–encoded epitranscriptomics. Nat Struct Mol Biol 2016 Feb; 23 (2) : 98-102.

182. Dominissini D, Moshitch-Moshkovitz S, Schwartz S, Salmon-Divon M, Ungar L, Osenberg S, Cesarkas K, Jacob-Hirsch J, Amariglio N, Kupiec M, Sorek R, Rechavi G. Topology of the human and mouse m6A RNA methylomes revealed by m6A-seq. Nature 2012 May; 485 (7397) : 201-206.

183. Meyer KD, Saletore Y, Zumbo P, Elemento O, Mason CE, Jaffrey SR. Comprehensive analysis of mRNA methylation reveals enrichment in 3' UTRs and near stop codons. Cell 2012 Jun; 149 (7) : 1635-1646.

184. Rowley JD. Identification of the constant chromosome regions involved in human hematologic malignant disease. Science 1982 May; 216 (4547) : 749-751.

185. Mayer RJ, Davis RB, Schiffer CA, Berg DT, Powell BL, Schulman P, Omura GA, Moore JO, McIntyre OR, Frei E. Intensive postremission chemotherapy in adults with acute myeloid leukemia. Cancer and Leukemia Group B. N Engl J Med 1994 Oct; 331 (14) : 896-903.

186. Löwenberg B, Suciu S, Archimbaud E, Ossenkoppele G, Verhoef GE, Vellenga E, Wijermans P, Berneman Z, Dekker AW, Stryckmans P, Schouten H, Jehn U, Muus P, Sonneveld P, Dardenne M, Zittoun R. Use of recombinant GM-CSF during and after remission induction chemotherapy in patients aged 61 years and older with acute myeloid leukemia: final report of AML-11, a phase III randomized study of the Leukemia Cooperative Group of European Organisation for the Research and Treatment of Cancer and the Dutch Belgian Hemato-Oncology Cooperative Group. Blood 1997 Oct; 90 (8) : 2952-2961.

187. Byrd JC, Dodge RK, Carroll A, Baer MR, Edwards C, Stamberg J, Qumsiyeh M, Moore JO, Mayer RJ, Davey F, Schiffer CA, Bloomfield CD. Patients with t(8;21)(q22;q22) and acute myeloid leukemia have superior failure-free and overall survival when repetitive cycles of high-dose cytarabine are administered. J Clin Oncol 1999 Dec; 17 (12) : 3767-3775.

188. Peters WP, Rosner G, Ross M, Vredenburgh J, Meisenberg B, Gilbert C, Kurtzberg J. Comparative effects of granulocyte-macrophage colony-stimulating

135

factor (GM-CSF) and granulocyte colony-stimulating factor (G-CSF) on priming peripheral blood progenitor cells for use with autologous bone marrow after high- dose chemotherapy. Blood 1993 Apr; 81 (7) : 1709-1719.

189. Socinski MA, Cannistra SA, Elias A, Antman KH, Schnipper L, Griffin JD. Granulocyte-macrophage colony stimulating factor expands the circulating haemopoietic progenitor cell compartment in man. Lancet 1988 May; 1(8596) : 1194-1198.

190. Hast R, Hellström-Lindberg E, Ohm L, Björkholm M, Celsing F, Dahl IM, Dybedal I, Gahrton G, Lindberg G, Lerner R, Linder O, Löfvenberg E, Nilsson-Ehle H, Paul C, Samuelsson J, Tangen JM, Tidefelt U, Turesson I, Wahlin A, Wallvik J, Winquist I, Oberg G, Bernell P. No benefit from adding GM-CSF to induction chemotherapy in transforming myelodysplastic syndromes: better outcome in patients with less proliferative disease. Leukemia 2003 Sep; 17 (9) : 1827-1833.

191. Ravandi F. Role of cytokines in the treatment of acute leukemias: a review. Leukemia 2006 Apr; 20 (4) : 563-571.

192. Gaipa G, Bugarin C, Longoni D, Cesana S, Molteni C, Faini A, Timeus F, Zecca M, Biondi A. Aberrant GM-CSF signal transduction pathway in juvenile myelomonocytic leukemia assayed by flow cytometric intracellular STAT5 phosphorylation measurement. Leukemia 2009 Apr; 23 (4) : 791-793.

193. Gasson JC. Molecular physiology of granulocyte-macrophage colony-stimulating factor. Blood 1991 Mar; 77 (6) : 1131-1145.

194. Biethahn S, Alves F, Wilde S, Hiddemann W, Spiekermann K. Expression of granulocyte colony-stimulating factor- and granulocyte-macrophage colony- stimulating factor-associated signal transduction proteins of the JAK/STAT pathway in normal granulopoiesis and in blast cells of acute myelogenous leukemia. Experimental Hematology 1999; 27 (5) : 885-894.

195. van de Laar L, Coffer PJ, Woltman AM. Regulation of dendritic cell development by GM-CSF: molecular control and implications for immune homeostasis and therapy. Blood 2012 Apr; 119 (15) : 3383-3393.

196. Waller EK. The role of sargramostim (rhGM-CSF) as immunotherapy. Oncologist 2007; 12 Suppl 2: 22-26.

136

197. Quezada SA, Peggs KS, Curran MA, Allison JP. CTLA4 blockade and GM-CSF combination immunotherapy alters the intratumor balance of effector and regulatory T cells. J Clin Invest 2006 Jul; 116 (7) : 1935-1945.

198. Touw I, Donath J, Pouwels K, van Buitenen C, Schipper P, Santini V, Hagemeijer A, Lowenberg B, Delwel R. Acute myeloid leukemias with chromosomal abnormalities involving the 21q22 region identified by their in vitro responsiveness to interleukin-5. Leukemia 1991 Aug; 5(8) : 687-692.

199. Woiciechowsky A, Regn S, Kolb HJ, Roskrow M. Leukemic dendritic cells generated in the presence of FLT3 ligand have the capacity to stimulate an autologous leukemia-specific cytotoxic T cell response from patients with acute myeloid leukemia. Leukemia 2001 Feb; 15 (2) : 246-255.

200. Tarte K, Fiol G, Rossi JF, Klein B. Extensive characterization of dendritic cells generated in serum-free conditions: regulation of soluble antigen uptake, apoptotic tumor cell phagocytosis, chemotaxis and T cell activation during maturation in vitro. Leukemia 2000 Dec; 14 (12) : 2182-2192.

201. Roncuzzi L, Brognara I, Cocchi S, Zoli W, Gasperi-Campani A. Loss of heterozygosity at pseudoautosomal regions in human breast cancer and association with negative hormonal phenotype. Cancer Genet Cytogenet 2002 Jun; 135 (2) : 173-176.

202. Gough NM, Gough J, Metcalf D, Kelso A, Grail D, Nicola NA, Burgess AW, Dunn AR. Molecular cloning of cDNA encoding a murine haematopoietic growth regulator, granulocyte-macrophage colony stimulating factor. Nature 1984 1984 Jun 28-Jul 4; 309 (5971) : 763-767.

203. Wong GG, Witek JS, Temple PA, Wilkens KM, Leary AC, Luxenberg DP, Jones SS, Brown EL, Kay RM, Orr EC. Human GM-CSF: molecular cloning of the complementary DNA and purification of the natural and recombinant proteins. Science 1985 May; 228 (4701) : 810-815.