Epigenetic Regulation of Acute Myeloid Leukaemic Stem Cells

Epigenetic Regulation of Acute Myeloid

Leukaemic Stem Cells

Halina Hoi Laam Leung

Bachelor of Medicinal Chemistry Advanced (Honours Class 1)

A thesis submitted in fulfilment of the requirements for the degree of

Doctor of Philosophy

Children‘s Cancer Institute Australia for Medical Research

School of Women‘s and Children‘s Health

Faculty of Medicine

University of New South Wales

November 2015

PLEASE TYPE THE UNIVERSITY OF NEW SOUTH WALES Thesis/Dissertation Sheet

Surname or Family name: Leung

First name: Halina Other name/s: Hoi Laam

Abbreviation for degree as given in the University calendar: PhD

School: Women’s and Children’s Health Faculty: Medicine

Title: Epigenetic Regulation of Acute Myeloid Leukaemic Stem Cells

Abstract 350 words maximum: (PLEASE TYPE)

Acute myeloid leukaemia (AML) is a heterogeneous blood disease with high relapse rates. A small population of highly drug resistant leukaemic stem cells (LSC) persists after chemotherapy, causing disease relapse. This emphasises the need to develop novel LSC-targeted therapies. While increasing evidence has emerged in recent years highlighting the role of epigenetic regulators in cancer development and maintenance, knowledge in epigenetic regulation of LSC is limited. This study investigated the role of two novel histone demethylases, Jmjd1c and Jmjd5, in regulating LSC, as well as the effectiveness of currently available epigenetic agents on leukaemogenesis.

This study has shown that Jmjd1c is required for AML development and maintenance. Experiments involving enforced expression of Jmjd1c in haematopoietic stem cells (HSC) suggested that Jmjd1c has the capacity to enhance stem cell proliferation. Jmjd1c overexpression also promoted proliferation of HSC transduced with oncogenes, Hoxa9/Meis1a, and accelerated leukaemia development in mice. Gene expression profiling identified Jmjd1c to be a critical regulator of metabolic pathways such as glycolysis. These novel findings support the importance of targeting Jmjd1c.

This study also provided the first evidence of Jmjd5 as a potential tumour suppressor in AML. Overexpression of Jmjd5 significantly impaired LSC proliferation and prolonged mouse survival. Gene expression profiling and functional studies suggested that Jmjd5 negatively regulated a LSC self-renewal pathway driven by G protein-coupled receptor 84 (Gpr84). Given the partially shared downstream signalling of Jmjd5 and Gpr84, it is likely that Jmjd5 negatively regulates LSC function, at least in part, by inhibiting Gpr84 signalling.

Jmjd1c and Jmjd5 have previously been shown to regulate histone methylation. Although global H3K9me2 levels were not altered by Jmjd1c, Jmjd5 reduced global H3K36me2 and H3K27me3 levels. Reduction in H3K27me3 could also be achieved by DZNep-induced Ezh2 inhibition at low doses, which altered the gene expression profile of DZNep-treated AML cells.

In summary, both Jmjd1c and Jmjd5 were found to be crucial in the proliferation and maintenance of AML cells, suggesting that targeting aberrant methylation mediated by these epigenetic regulators may provide a promising approach in AML therapy.

Declaration relating to disposition of project thesis/dissertation

I hereby grant to the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or in part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all property rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation.

I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstracts International (this is applicable to doctoral theses only).

…………………………………………………………… ……………………………………..……………… ……….…....11/11/2015..…….… Signature Witness Date The University recognises that there may be exceptional circumstances requiring restrictions on copying or conditions on use. Requests for restriction for a period of up to 2 years must be made in writing. Requests for a longer period of restriction may be considered in exceptional circumstances and require the approval of the Dean of Graduate Research.

FOR OFFICE USE ONLY Date of completion of requirements for Award:

THIS SHEET IS TO BE GLUED TO THE INSIDE FRONT COVER OF THE THESIS

ORIGINALITY STATEMENT

‗I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, or substantial proportions of material which have been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis. I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project's design and conception or in style, presentation and linguistic expression is acknowledged.‘

Signed ………………………………………

Date ……11/11/2015..……………………

‗I hereby grant the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all proprietary rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation

I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstract International (this is applicable to doctoral theses only).

I have either used no substantial portions of copyright material in my thesis or I have obtained permission to use copyright material; where permission has not been granted I have applied/will apply for a partial restriction of the digital copy of my thesis or dissertation.'

Signed ………………………………………

Date ………11/11/2015…………………

AUTHENTICITY STATEMENT

‗I certify that the Library deposit digital copy is a direct equivalent of the final officially approved version of my thesis. No emendation of content has occurred and if there are any minor variations in formatting, they are the result of the conversion to digital format.‘

Signed ………………………………………

Date ………11/11/2015..…………………

iii

ABSTRACT

This study has shown that Jmjd1c is required for AML development and maintenance. Experiments involving enforced expression of Jmjd1c in haematopoietic stem cells (HSC) suggested that Jmjd1c has the capacity to enhance stem cell proliferation. Jmjd1c overexpression also promoted proliferation of HSC transduced with oncogenes, Hoxa9/Meis1a, and accelerated leukaemia development in mice. Gene expression profiling also identified Jmjd1c to be a critical regulator of metabolic pathways such as glycolysis. These novel findings support the importance of targeting Jmjd1c.

ACKNOWLEDGEMENTS

I would like to thank a number of people for their support and patience during the last four challenging years. I would not have made it without you all.

Firstly, I would like to thank my supervisor, Dr Jenny Wang, for giving me the opportunity to join her research group. I have learnt so much and appreciate the supervision and advice she has given me throughout the time spent in her lab.

I also would like to thank my co-supervisor, Professor Murray Norris, for the time taken out of his busy schedule for his constructive feedback and support in getting me through my PhD. I also appreciate the support and guidance given to me by my panel members, Professor Michelle Haber and Professor Maria Kavallaris. Thank you to Dr Amanda Philp for our catch-ups and the support she has given me during my PhD.

Thank you to Dr Bing Liu for his help with the expression profiling and data analysis, and to Dr Aliaksei Holik for his help with the chromatin immunoprecipitation assays. Thank you to Chris Brownlee and David Snowden for their assistance with flow sorting and always being so accommodating when I give late/no notice. It has been great having someone to chat to downstairs during our flow sessions. Thank you to Janelle McPhee for giving me the opportunity to work in the animal facility and to Anick Standley, Alison Richards, Gabrial Gomes and Aliza Yong for looking after my animals. Thank you to Kathryn Evans for our always amusing training sessions.

Thank you to Dr Owen Sprod for helping me settle into the lab. Thank you to Ashley Yang, Kimberley Anderson and Kathryn Mathews for your friendship and help with my projects. Good luck with your PhDs. Thank you to Florida Voli, you will make a great PhD student; Brendon Martinez, your Canada trip awaits; Andrea Naim, you work too hard for an honours student; and Hannah McCalmont, I‘m sure you‘ll find someone to flick tips at when you ‗unintentionally‘ miss the bin, and may they continue our tradition of leaving messages on your ethanol/vodka bottle.

Thank you to the amazing postdoctorate researchers, Dr Katerina Bendak, Dr Jennifer Lynch, Dr Elizabeth Roundhill and Dr Denise Yu. I would not have made it through my PhD without your help. All of you were there for me when I was at my lowest and have vi all helped me become a stronger person. Katschi and Jen, good luck with your upcoming weddings; Liz, I will visit you soon in the UK; and Denise, no one deserved that promotion more than you. To the recently accepted and soon to be Dr‘s: Tony Huynh, Philipp Dietrich, Hangyu Yi, Estrella Gonzales, Amelia Parker, Rebekka Williams and Luke Jones. Thank you all for your help in and out of the lab, the coffee breaks, night outs, hugs and putting up with all the tears. Tony, thanks for all the hugs when I needed it; Phil, thanks for all the help in the lab and the chats to pass the time, I will make sure steak is on the menu when you come to visit; Hangyu, you‘re so close to finishing, good luck; Star, Cassie is bound to become a great scientist like you with her frequent visits to CCI; Amelia, thanks for your support, good luck with finishing off your PhD; Bex, thanks for the late night chats at work; and Luke, thanks for being there when I needed to chat. Thank you all for your friendship and support.

I would like to thank everyone at the Children‘s Cancer Institute. It has been a pleasure to meet you all and have made working in CCI so enjoyable.

To my family, thank you for all your patience and support over the last few years. Mum, thanks for all the midnight snacks and the trips to work at the most inconvenient hours of the day. Dr Helen Leung, Tina Leung and Vivian Tse, we have all been studying and working so hard over the last four years and I really appreciate all your constant love and support. Thanks for getting me through the rough patches with your emergency visits and dinners. I have eaten so much good food these last few years thanks to our nights out. I can‘t wait to start the next chapter of my life.

vii

TABLE OF CONTENTS

Originality Statement...... i Copyright Statement ...... ii Authenticity Statement ...... iii Abstract ...... iv Acknowledgements ...... vi Table of Contents ...... viii List of Figures ...... xii List of Tables ...... xv Abbreviations ...... xvi Publications, Conferences and Awards ...... xix

Chapter 1: Introduction ...... 1

1.1. Acute myeloid leukaemia ...... 1 1.1.1. Incidence and symptoms ...... 1 1.1.2. Subtypes of AML ...... 2 1.1.3. Current therapy ...... 9 1.2. Leukaemic stem cells...... 11 1.2.1. Normal haematopoietic stem cell properties ...... 11 1.2.2. Origin and identification of LSC ...... 13 1.2.3. Disease relapse and drug resistance ...... 17 1.2.4. Hypoxia and glucose metabolism ...... 19 1.2.5. Self-renewal pathways...... 22 1.3. Epigenetics ...... 25 1.3.1. Epigenetic regulation ...... 25 1.3.1.1. DNA methylation ...... 26 1.3.1.2. Histone modifications ...... 29 1.3.1.2.1. Histone acetylation ...... 33 1.3.1.2.2. Histone methylation ...... 35 1.3.2. Recent advances in epigenetic cancer therapy ...... 42 1.4. Summary and thesis perspectives ...... 47

viii

Chapter 2: Materials and Methods ...... 50

2.1 Reagents ...... 50 2.2 Cell biology techniques ...... 50 2.2.1 Murine leukaemic cells...... 50 2.2.2 Cell culture ...... 52 2.2.3 Cryopreservation ...... 53 2.2.4 Thawing cell lines...... 53 2.2.5 Colony forming assay ...... 53 2.2.6 Alamar blue cytotoxicity assay ...... 53 2.2.7 Wright-Giemsa staining...... 54 2.3 Molecular biology techniques ...... 55 2.3.1 Bacterial transformation ...... 55 2.3.2 Plasmid preparation ...... 56 2.3.3 shRNA/cDNA transfection ...... 56 2.3.4 Viral transduction ...... 57 2.4 RNA analysis ...... 57 2.4.1 RNA extraction ...... 57 2.4.2 cDNA synthesis ...... 58 2.4.3 Real-time quantitative PCR protocol...... 58 2.4.4 Microarray expression profiling ...... 62 2.5 DNA analysis ...... 63 2.5.1 Chromatin immunoprecipitation sample preparation ...... 63 2.6 Protein analysis ...... 63 2.6.1 Protein extraction...... 63 2.6.2 Western blotting ...... 64 2.6.2.1 Protein separatioin by SDS-PAGE ...... 64 2.6.2.2 Protein transfer ...... 64 2.6.2.3 Immunoblotting ...... 65 2.7 Flow cytometry ...... 66 2.7.1 Cell cycle analysis ...... 67 2.7.2 Apoptosis assay ...... 68 2.7.3 BrdU staining...... 68 2.7.4 Fluorescence activated cell sorting ...... 69 ix

2.7.5 Cell surface staining ...... 69 2.8 Metabolic assays ...... 69 2.8.1 ATP assay ...... 69 2.8.2 Glycolysis assay ...... 70 2.8.3 Oxidative phosphorylation assay ...... 70 2.9 Animal work ...... 71 2.9.1 Mouse injections ...... 71 2.9.2 Organ harvesting ...... 71 2.10 Limiting dilution analysis ...... 72 2.11 Statistical analysis ...... 72

Chapter 3: The Role of JMJD1C in Regulating Acute Myeloid Leukaemic Stem Cells ...... 73

3.1. Introduction ...... 73 3.2. Jmjd1c is upregulated in mouse and human MLL AML models ...... 75 3.3. Suppression of Jmjd1c impairs leukaemic properties of KLSMLLAF9 in vitro ...... 78 3.4. Jmjd1c is required for the maintenance of established KLSMLLAF9 leukaemia in vivo ...... 81 3.5. Jmjd1c plays a more critical role in stem cell-derived MLLAF9 leukaemia . 86 3.6. Overexpression of Jmjd1c confers a growth advantage to KLSA9M pre-LSC in vitro ...... 86 3.7. Jmjd1c enhances leukaemogenesis ...... 89 3.8. Jmjd1c enhances proliferation of HSC ...... 96 3.9. Jmjd1c target genes are involved in regulation of metabolic pathways ...... 97 3.10. Discussion ...... 116

Chapter 4: The Role of JMJD5 in Regulating MLL-Rearranged Acute Myeloid Leukaemia ...... 123

4.1. Introduction ...... 123 4.2. Jmjd5 expression is suppressed in KLSMLLAF9 AML cells ...... 124 4.3. Overexpression of Jmjd5 impairs KLSMLLAF9 colony formation in vitro...... 128

4.4. Jmjd5 impairs leukaemogenesis ...... 132 4.5. KLSMLLAF9 cells are more sensitive to doxorubicin treatment upon enforcced Jmjd5 expression...... 136 4.6. Downstream pathways regulated by Jmjd5 ...... 141 4.7. Discussion ...... 151

Chapter 5: Epigenetic Regulation of Acute Myeloid Leukaemic Stem Cells ...... 158

5.1. Introduction ...... 158 5.2. Global histone 3 lysine 9 methylation levels are not altered by Jmjd1c ...... 161 5.3. Jmjd5 preferentially demethylates histone 3 lysine 36 dimethylation ...... 163 5.4. Interplay between H3K36 and H3K27 methylation ...... 166 5.5. DZNep reduces H3K27 methylation levels at low concentrations without inducing apoptosis ...... 168 5.6. The origin and developmental stage of malignant stem cells determine downstream regulatory pathways in response to DZNep treatment ...... 181 5.7. Discussion ...... 185

Chapter 6: Concluding Remarks ...... 191

6.1. Jmjd1c is required for the maintenance and proliferation of stem-cell derived leukaemia ...... 192 6.2. Jmjd5 plays a tumour suppressive role in leukaemogenesis ...... 193 6.3. DZNep as an epigenetic agent in AML ...... 197 6.4. Potential therapeutic agents ...... 197 6.5. Conclusion ...... 203

Supplementary Figures ...... 204 References ...... 210

LIST OF FIGURES

Figure 1.1: Categories of structural genomic alterations in paediatric AML ...... 6 Figure 1.2: Distribution of major MLL fusion partner genes in de novo childhood AML ...... 7 Figure 1.3: Classical model of haematopoiesis ...... 12 Figure 1.4: Model of haematopoiesis in AML ...... 16 Figure 1.5: Conventional chemotherapy and LSC-targeted therapy ...... 18 Figure 1.6: Cellular metabolism of normal and tumour cells ...... 20 Figure 1.7: Schematic diagram of the Wnt/β-catenin signalling pathway ...... 23 Figure 1.8: Epigenetic modifications to DNA and histones dictate nucleosome spacing and transcription ...... 27 Figure 1.9: Covalent modifications of the N-terminal tail of core histones ...... 32 Figure 1.10: Mechanism of lysine demethylation ...... 41 Figure 3.1: Jmjd1c expression is upregulated in mouse and human MLL AML, and associated with poor clinical prognosis ...... 76-77 Figure 3.2: Knockdown of Jmjd1c impairs the proliferative and colony forming abilities of KLSMLLAF9 pre-LSC in vitro ...... 79 Figure 3.3: Suppression of Jmjd1c blocks cell cycle progression but does not induce apoptosis ...... 80 Figure 3.4: Suppression of Jmjd1c does not alter MLL leukaemia initiation ...... 82 Figure 3.5: Jmjd1c suppression retains an AML immunophenotype in vivo ...... 83 Figure 3.6: Jmjd1c is required for leukaemia maintenance ...... 85 Figure 3.7: Jmjd1c is not required for GMPMLLAF9 maintenance ...... 87 Figure 3.8: Jmjd1c confers a growth advantage to KLSA9M pre-LSC in vitro ...... 88 Figure 3.9: Jmjd1c enhances leukaemogenesis in KLSA9M ...... 90-91 Figure 3.10: Overexpression of Jmjd1c in KLSA9M enhances engraftment in the bone marrow ...... 93 Figure 3.11: Jmjd1c increases proliferation of KLSA9M in vivo ...... 94-95 Figure 3.12: Transformation of HSC with Jmjd1c ...... 98 Figure 3.13: Gene set enrichment analysis of microarray data identifies Jmjd1c downstream pathways ...... 100-101 Figure 3.14: Validation of microarray data by qRT-PCR ...... 103 xii

Figure 3.15: Validation of Jmjd1c targets by flow cytometry ...... 108 Figure 3.16: Valdiation of Jmjd1c targets by western blotting ...... 110 Figure 3.17: Pkm2 inhibition by shikonin impairs colony formation ...... 112 Figure 3.18: Inhibition of Pdk1 by GSK2334470 impairs colony formation ...... 114 Figure 3.19: Enhanced Jmjd1c expression is associated with increased ATP production ...... 115 Figure 3.20: Enforced Jmjd1c expression enhances glycolytic activity ...... 117 Figure 4.1: Jmjd5 is downregulated in aggressive AML where low expression predicts poor clinical outcome ...... 125-126 Figure 4.2: Jmjd5 is highly expressed in DOT1L inhibitor-treated KLSMLLAF9 ... 127 Figure 4.3: Jmjd5 impairs colony forming ability of KLSMLLAF9 in vitro ...... 129 Figure 4.4: Jmjd5 induces apoptosis and differentiation in KLSMLLAF9 ...... 130-131 Figure 4.5: Overexpression of Jmjd5 impairs leukaemia initiation ...... 133 Figure 4.6: Overexpression of Jmjd5 impairs leukaemia maintenance ...... 134-135 Figure 4.7: Heatmap depicting Jmjd5 target genes ...... 137 Figure 4.8: Validation of Jmjd5 targets by western blotting ...... 140 Figure 4.9: Jmjd5 sensitises KLSMLLAF9 to doxorubicin ...... 142 Figure 4.10: Suppression of Gpr84 correlates with reduced β-catenin and increased Jmjd5 expression ...... 143 Figure 4.11: Jmjd5 downregulates the expression of β-catenin targets ...... 144 Figure 4.12: Heatmap of genes inversely co-regulated by Jmjd5 and Gpr84 ...... 146 Figure 4.13: Validation of Gpr84 and Jmjd5 common target genes by qRT-PCR in Jmjd5 overexpressing cells ...... 149 Figure 4.14: Validation of Gpr84 and Jmjd5 common target genes by qRT-PCR in Gpr84 overexpressing cells ...... 150 Figure 4.15: Gpr84 and β-catenin rescue the Jmjd5-overexpressing phenotype in KLSMLLAF9 pre-LSC ...... 152-153 Figure 5.1: Jmjd1c does not alter global H3K9 methylation levels ...... 162 Figure 5.2: Jmjd5 preferentially demethylates H3K36me2 ...... 164 Figure 5.3: H3K36 dimethylation levels are altered by the ectopic expression of Gpr84 or β-catenin ...... 165 Figure 5.4: Enrichment of H3K36me2 in the coding region of Tcf7l2 in KLSMLLAF9 ...... 167

xiii

Figure 5.5: Jmjd5 indirectly reduces H3K27me3 methylation ...... 169 Figure 5.6: Suppression of Ezh2 inhibits colony formation without inducing apoptosis ...... 170 Figure 5.7: DZNep does not induce apoptosis at low concentrations in MLL pre-LSC ...... 172 Figure 5.8: DZNep impairs clonogenic growth of pre-LSC by inducing differentiation ...... 173-174 Figure 5.9: DZNep impairs clonogenic growth of LSC by inducing differentiation ...... 176-177 Figure 5.10: DZNep does not induce apoptosis at low concentrations in MLL LSC .. 178 Figure 5.11: DZNep preferentially targets HSC-derived malignant stem cells at low doses ...... 179 Figure 5.12: Low doses of DZNep are sufficient to reduce global levels of H3K27me3 ...... 180 Figure 5.13: DZNep restores expression of different downstream targets in pre-LSC generated from distinct cellular origins ...... 182 Figure 5.14: DZNep restores expression of different downstream targets in LSC generated from distinct cellular origins ...... 183 Figure 6.1: Therapeutic targeting of leukaemic stem cell niche interactions ...... 196 Figure 6.2: Chemical structures of iron-dependent enzyme inhibitors ...... 199 Figure 6.3: Chemical structures of Jmjd inhibitors ...... 200 Figure S.1: Generation of an inducible Jmjd1c knockout MLLAF9 AML model ..... 204 Figure S.2: Confirming overexpression of Jmjd1c-S in KLSA9M by qRT-PCR ...... 205 Figure S.3: H3K36me2 enrichment at Jmjd5 target genes using ChIP-qPCR ...... 206

xiv

LIST OF TABLES

Table 1.1: French-American-British (FAB) morphology classification of AML subtypes ...... 3 Table 1.2: World Health Organisation (WHO) classification of AML subtypes ...... 4 Table 1.3: Function and frequency of MLL fusion partner genes in leukaemia ...... 8 Table 1.4: Classes of epigenetic writers, erasers, and readers of hisone marks and their regulatory function ...... 30-31 Table 1.5: Summary of histone acetylase specificity on histone 3 (H3) and histone 4 (H4) lysine residues ...... 34 Table 1.6: Summary of histone methyltransferase and demethylase specificity on histone 3 (H3) and histone 4 (H4) lysine residues ...... 37 Table 1.7: Family of histone demethylases and their methylated substrate ...... 38 Table 1.8: Epigenetic-based therapeutics in preclinical and clinical development ...... 43-44 Table 2.1: Summary of cell types ...... 51 Table 2.2: List of plasmids ...... 55 Table 2.3: List of primer sequences ...... 59-62 Table 2.4: Primary antibody dilutions used for western blotting ...... 65-66 Table 2.5: List of flow cytometry antibodies ...... 67 Table 3.1: Summary of GSEA results following microarray expression profiling of Jmjd1c overexpression in KLSA9M pre-LSC ...... 99 Table 3.2: Brief summary of the function of selected Jmjd1c target genes ...... 104-107 Table 4.1: Brief summary of the function of selected Jmjd5 target genes ...... 138-139 Table 4.2: Brief summary of the function of selected common target genes of Jmjd5 and Gpr84 ...... 147-148 Table 5.1: Brief summary of the known function of DZNep-related target genes ... 184 Table S.1: Limiting dilution analysis of KLSMLLAF9 or KLSA9M leukaemia upon gain or loss of Jmjd1c ...... 207 Table S.2: Transcripts of Jmjd1c in Mus musculus ...... 208 Table S.3: Limiting dilution analysis of KLSMLLAF9 leukaemia upon gain of Jmjd5 ...... 209 xv

ABBREVIATIONS

°C Degrees Celsius

7-AAD 7-Aminoactinomycin D aa Amino acids

A9M Hoxa9/Meis1a

ALL Acute lymphoblastic leukaemia

AML Acute myeloid leukaemia bp Base pairs

BrdU Bromodeoxyuridine c-Kit Stem cell factor receptor

ChIP Chromatin immunoprecipitation

CLP Common lymphoid progenitor

CML Chronic myeloid leukaemia

CSC Cancer stem cell

DMEM Dulbecco‘s Modified Eagle‘s media

DMSO Dimethyl sulphoxide

DNA Deoxyribonucleic acid

DNMT DNA methyltransferase dNTP Deoxynucleoside triphosphate

DTT Dithiothreitol

EFS Event free survival

FAB French-American-British

FACS Fluorescence activated cell sorting

FCS Fetal calf serum

FSB First strand buffer g Gram(s), mass; gravitational force, centrifugation steps xvi

GMP Granulocyte macrophage progenitor

GPR G protein-coupled receptor

GSK3 Glycogen synthase kinase-3-β

GVHD Graft-versus-host disease h Hour(s)

H3 Histone 3

H4 Histone 4

HAT Histone acetylase

HDAC Histone deacetylase

HEK293 Human embryonic kidney 293

HEPES 4-(2-Hydroxyethyl)-1-piperazineethanesulfonic acid

HMT Histone methyltransferase

HSC Haematopoietic stem cell

IC50 Half maximal inhibitory concentration

IMDM Iscove‘s Modified Dulbecco‘s Medium

KDM Histone demethylase

JMJD JmjC domain containing demethylase

KLS c-Kithigh Lin Sca-1

L Litre(s)

Lin Haematopoietic lineage markers

LSC Leukaemic stem cells

LSD Lysine specific demethylase

LT-HSC Long-term HSC mA MilliAmps me1 Monomethylation me2 Dimethylation me3 Trimethylation xvii min Minute(s)

MLL Mixed lineage leukaemia

MLLAF9 MLL containing the chromosome translocation t(9;11)(p22;q23)

MPP Multipotent progenitor

NOD/SCID Non-obese diabetic/severe combined immunodeficiency mouse model

PBS Phosphate buffered saline

PI Propidium iodide

Pre-LSC Early developmental stage of LSC qRT-PCR Quantitative real-time polymerase chain reaction s seconds

Sca-1 Stem cell antigen-1

SEM Standard error of the mean

ST-HSC Short-term HSC

Tcf T cell-specific transcription factor

V Volts

WBC White blood cell

WHO World Health Organization

Wnt Wingless-related integration site

xviii

PUBLICATIONS, CONFERENCES AND AWARDS

Publications

1. Dietrich PA, Yang C, Leung HHL, Lynch JR, Gonzales E, Liu B, Haber M, Norris MD, Wang J, Wang JY. GPR84 sustains aberrant beta-catenin signaling in leukemic stem cells for maintenance of MLL leukemogenesis. Blood (2014) 124(22):3284-3294 2. Leung, HHL, Wang, J, Wang JY (2014). Jmjd1c, a histone demethylase, is required for the maintenance of MLL-AF9 AML leukemic stem cells. Blood, 124(21):3530*

Conferences

Poster presentations:

1. Leung, HHL, Gonzales, E, Anderson, KJ, Wang JY (2015). Cell of origin and developmental stage influence the sensitivity of acute myeloid leukaemic stem cells to treatment with 3-deazaneplanocin A. 23rd Australian Society for Medical Research NSW Scientific Meeting. Powerhouse Museum, Sydney, Australia 2. Leung, HHL, Wang, J, Wang JY (2014). Jmjd1c, a histone demethylase, is required for the maintenance of MLL-AF9 AML leukemic stem cells. 56th American Society of Hematology Annual Meeting and Exposition. Moscone Center, San Francisco, USA* 3. Leung, HHL, Norris MD, Wang JY (2014). Targeting a novel epigenetic oncogene required for the maintenance of acute myeloid leukaemic stem cells. 26th Lorne Cancer Conference. Mantra Lorne, Lorne, Victoria, Australia

Oral presentations:

1. Leung, HHL, Norris MD, Wang JY (2013). Defining the tumourigenic function of a novel epigenetic oncogene required or the development of acute myeloid leukaemic stem cells. Genetics Society of AustralAsia Conference. University of New South Wales, Sydney, New South Wales, Australia xix

2. Leung, HHL, Norris MD, Yang JY (2013). Unravelling the crucial role of a novel epigenetic oncogene in regulation of acute myeloid leukaemic stem cells. 21st Australian Society for Medical Research NSW Scientific Meeting. Australian Technology Park, Eveleigh, Australia 3. Leung, HHL, Norris MD, Wang JY (2013). Characterising a novel oncogene required for the development of acute myeloid leukaemic stem cells. Stem Cells and Cancer Symposium. Melbourne Brain Centre, University of Melbourne, Parkville, Australia

Awards and Achievements

October 2014 ASH Abstract Excellence Award

February 2014 Lorne Cancer Travel Bursary

January 2014 Postgraduate Research Support Scheme Travel Scholarship to attend ASH in Dec 2014

April 2013 Breakthrough Speaker at Stem Cells and Cancer Conference

CHAPTER 1

Introduction

1.1 Acute myeloid leukaemia

Leukaemia is a malignant disease of the blood and bone marrow and is the most common type of childhood cancer, accounting for 30% of cancers (Linet et al., 1999). Acute lymphoblastic leukaemia (ALL) and acute myeloid leukaemia (AML) are the two main types of paediatric haematopoietic malignancies, while chronic leukaemias are more common in adults than in children (Downing and Shannon, 2002). ALL is characterised by the accumulation of lymphoblasts in the bone marrow or blood, while AML is characterised by the accumulation of malignant myeloid blasts. In AML, self- renewing stem or progenitor cells demonstrate limited differentiation capacity or fail to differentiate into functional granulocytes or monocytes (Fialkow et al., 1987).

1.1.1 Incidence and symptoms

AML accounts for approximately 20% of acute leukaemias in children (Rubnitz et al., 2010a, Lehmbecher et al., 2004), but accounts for 30% of acute leukaemia deaths (Rubnitz et al., 2010a). Children younger than 15 years of age have a five-year event- free survival (EFS) rate of approximately 58%, while adolescents aged between 15 and 19 years have a five-year EFS rate of approximately 40% (Smith et al., 2010).

Fatigue, pallor and fever are the most common symptoms of AML (Miller and Daoust, 2005, American Cancer Society, 2015). Pallor and weakness can manifest from anaemia, whilst fever can result from infection or the leukaemia itself. Leukaemic infiltration can occur in various organs or tissues termed granulocytic sarcoma or chloroma (deposits of myeloid cells) (Bonilla and Ribeiro, 2014). In children, leukaemic cells may spread to the gums, causing swelling, pain and bleeding (Miller and Daoust, 2005). Easy bleeding or bruising may occur due to the lack of blood platelets, as well as bone or joint pain from the accumulation of leukaemic cells near the

1 surface of the bone or inside the joint. Hepatosplenomegaly and adenopathy may also occur, but are less common in AML than in ALL (Bonilla and Ribeiro, 2014).

1.1.2 Subtypes of AML

AML is a heterogeneous malignancy due to dysregulation of genes and pathways in haematopoietic stem or progenitor cells that may be sequentially acquired and cooperate to display the leukaemic phenotype. There are two classifications of AML subtypes which yield information critical to prognosis and relevant therapy. The French- American-British (FAB) criteria categorises leukaemia by morphological and cytochemical criteria (Besa, 1992), dividing AML into eight major subtypes ranging from M0 to M7 (Table 1.1). The World Health Organization (WHO) classification divides AML subtypes into categories based on cytomorphological, cytogenetic and immunophenotypic information with an attempt to be more clinically useful and provides more prognostic information than the FAB criteria (Vardiman et al., 2009) (Table 1.2).

The diagnosis of acute leukaemia patients involves differentiating between AML and ALL, followed by the classification of AML or ALL into categories that define treatment and prognostic groups. The majority of cases can be appropriately designated by morphological and cytochemical identification. The FAB classification is based on the degree of maturation, lineage differentiation and blast cell percentage. Although the main advantage of the FAB classification is the ease of use and quick preliminary diagnosis, difficulties arise in cases with negative cytochemical staining or when attempting to distinguish between M1 (AML FAB classification) from L2 (ALL FAB classification), or M1 from M2, or M2 from M4. In addition, there is also a substantial subset of cases that show no karyotypic abnormalities. The WHO classification, however, combines features of both cytogenetic and molecular criteria, and correlates this with specific cytogenetic findings associated with myelodysplasia, thus enhancing clinical and prognostic utility. The most significant differences between these two classifications include: a lower blast threshold for the diagnosis of AML by the WHO (20%) compared to FAB (30%); patients with recurring cytogenetic abnormalities

Table 1.1: French-American-British (FAB) morphology classification of AML subtypes Adapted from (Canadian Cancer Society, 2015, Bennett et al., 1985). Subtype Leukaemia Description Approx % of Cases M0 Acute myelogenous Minimal evidence of myeloid 1-6 leukaemia differentiation M1 Acute myeloblastic Myeloblasts are the dominant 11-19 leukaemia with minimal leukaemic cells at time of maturation diagnosis M2 Acute myeloblastic Myeloblasts are present, however, 25-30 leukaemia with some cells have matured to fully maturation formed blood cells M3 Acute promyelocytic Hypergranular, abnormal 3-12 leukaemia (APL) promyelocytes with Auer rods M4 Acute myelomonocytic Both granulocytic and monocytic 15-23 leukaemia differentiation are present M5 Acute Leukaemic cells have features of 13-29 monocytic/monoblastic developing monocytes leukaemia M6 Acute erythroleukaemia Leukaemic cells have features of 1-5 developing erythroid cells M7 Acute megakaryoblastic Leukaemic cells have features of 4-14 leukaemia developing platelets

Table 1.2: World Health Organization (WHO) classification of AML subtypes Adapted from (Vardiman et al., 2009). Leukaemia Subtypes AML with  AML with t(8;21) (AML1-ETO) characteristic genetic  AML with abnormal bone marrow eosinophils and abnormalities inv(16) or t(16;16)  APL with t(15;17) (PML-RAR) and variants  AML with mixed-lineage leukaemia (MLL) abnormalities such as t(9;11) (MLLAF9)  Megakaryoblastic AML with t(1;22) (OTT-MAL) Therapy-related  Related to alkylating agents myeloid neoplasms  Related to topoisomerase type II inhibitors AML with multi-  AML following myelodysplastic syndrome (MDS) or lineage dysplasia myeloproliferative disorder (MPS) o Chronic myelomonocytic leukaemia o Atypical chronic myelogenous leukaemia o Juvenile myelomonocytic leukaemia o Myelodysplatic/myeloproliferative disease AML, not otherwise  Acute myelomonocytic leukaemia specified  Acute monoblastic/monocytic leukaemia  Acute erythroid leukaemia  Acute megakaryoblastic leukaemia Myeloid sarcoma  Granulocytic sarcoma or chloroma MDS  Refractory anaemia  Refractory cytopenia

4 should be considered as AML regardless of blast percentage; and immunology marker expression is heterogeneous in AML making WHO subtype classification difficult, as well as the lack of expression of commonly investigated myeloid antigens (Angelescu et al., 2012). These methods of classification should therefore be used in complement to each other to offer a more accurate approach to diagnosis of AML in both children and adults.

Genomic alterations (duplications, deletions, and translocations) can be grouped into several categories which are valuable as diagnostic and prognostic tools for allocation of clinical therapy. The three main categories of cytogenetic profiles of paediatric AML cases contain either core binding factors (21%; AML1-ETO or t(8;21), and CBF-MYH1 or inv(16)), rearrangements involving the mixed-lineage leukaemia (MLL) gene (19%), or do not have identifiable karyotypic abnormalities (20%; normal karyotype) (Figure 1.1) (Horan et al., 2014). The MLL gene is highly promiscuous and forms fusion partners with a large number of translocation partners where MLL rearrangements are found in greater than 70% of infant leukaemias (ALL and AML) (Biondi et al., 2000). The function and frequency of leukaemia containing MLL fusion partners are described in Table 1.3. The clinical outcome of patients with MLL gene alterations therefore depends on the specific translocation partner (Balgobind et al., 2011).

The most frequent MLL rearrangements accounting for approximately 80% of all leukaemias include: MLLAF4 or t(4;11)(q21;q23); MLLAF9 or t(9;11)(p22;q23); MLLENL or t(11;19)(q23;p13.3); MLLAF10 or t(10;11)(p12;q23); and MLLAF6 or t(6;11)(q27;q23) (Huret et al., 2001, Meyer et al., 2006). The most prevalent MLL- translocation in childhood AML are those that contain the MLLAF9 translocation accounting for 33%, followed by MLLAF10 with 17% (Figure 1.2). MLL-fusions are efficient in transforming haematopoietic stem cells (HSC), granulocyte-macrophage progenitors (GMP) and other committed progenitors into leukaemic stem cells (LSC) (Krivtsov et al., 2006, Somervaille and Cleary, 2006, Cozzio et al., 2003). AML containing the MLLAF9 translocation has been associated with treatment failure (Meyer et al., 2013) and AML derived from stem cells with MLLAF9 develop a particularly aggressive and drug resistant subtype of AML (Krivtsov et al., 2013), however, the molecular mechanisms underlying MLLAF9 AML LSC survival remain obscure.

+8 Rare recurrent, 6 2% Del 5q/-5, 1 -7, 2 Del 9q, 2

Normal karyotype, 20 Other, 16

MLL, 19 Inv(16), 9

t(15;17), 11 t(8;21), 12

Figure 1.1: Categories of structural genomic alterations in paediatric AML Cytogenetic profiles of paediatric AML containing either core binding factors, rearrangements involving the mixed-lineage leukaemia (MLL) gene, or do not have identifiable karyotypic abnormalities. Adapted from (Horan et al., 2014).

Other, 22 AF9, 33

EPS15, 3

ELL, 11

AF10, 17 AF6, 14

Figure 1.2: Distribution of major MLL fusion partner genes in de novo childhood AML Mixed lineage leukaemia (MLL) rearrangements are found in AML, with MLLAF9 being the most predominant MLL translocation. Adapted from (Krivtsov and Armstrong, 2007).

Table 1.3: Function and frequency of MLL fusion partner genes in leukaemia Adapted from (Meyer et al., 2006, Huret et al., 2001, Krivtsov and Armstrong, 2007). Group Putative function Chromosome Fusion Frequency partner 1 Nuclear, putative DNA- 4q21 AF4 > 80% of binding proteins 9q23 AF9 MLL 19p13.3 ENL rearranged 10p12 AF10 leukaemias 19p13.1 ELL 2 Cytoplasm, presence of 1q32 EPS15 > 10% coiled-coil 17p13 GAS7 oligomerisation domain 19p13 EEN 6q27 AF6 Xq13 AFX 3 Cytoplasm, septin Xq22 SEPT2 > 1% family, interact with 22q11 SEPT5 cytoskeletal filaments, Xq24 SEPT6 have a role in mitosis 17q25 SEPT9 4q21 SEPT11 4 Nuclear, histone 16q13 CBP 1% acetyltransferases 22q13 P300 5 MLL partial tandem 11q23 N/A 4-7% of all duplication of exons 5-11 AML with (MLL-PTD) normal karyotype

1.1.3 Current therapy

AML therapy has changed little over the past twenty-five years (O‘Donnell et al., 2011) and consists of induction chemotherapy and consolidation therapy. Induction therapy attempts to induce remission by eliminating leukaemic bone marrow cells and allow repopulation of the bone marrow with normal cells. Morphological remission has been defined as the presence of less than 5% blast cells in bone marrow with evidence of myelopoiesis recovery, while complete remission is achieved when the leukaemia burden is less than 0.1% (Coustan-Smith et al., 2003, Rubnitz et al., 2010b).

The current mainstay of induction therapy for AML involves a combination of cytarabine (Ara-C) and daunorubicin. Cytarabine is a pyrmidine nucleoside analogue that inhibits DNA synthesis by incorporating into DNA, exhibiting cell phase specificity, primarily killing cells undergoing DNA synthesis (S-phase) (Perry, 2008). Daunorubicin (Daunomycin®) was the first anthracycline discovered, followed by doxorubicin (Adriamycin®) (Weiss, 1992). Anthracyclines exhibit antimitotic activity by intercalating between DNA base pairs and inhibiting topoisomerase II activity. Doxorubicin is more widely used in solid tumours while daunorubicin is exclusively used for leukaemia therapy (Doroshow, 2010). Differences in pharmacokinetics exist between these two anthracyclines, which may explain the difference in clinical applications (Paul et al., 1989). Enhanced selectivity of doxorubicin shown by increased intracellular accumulation was observed when administered as a DNA-conjugate for the treatment of acute leukaemia (Paul et al., 1989), suggesting its applicability as a therapeutic agent for AML.

The standard regimen for AML involves three days of an anthracycline such as daunorubicin and seven days of cytarabine, which results in improved complete remission and overall survival rates (Stone et al., 2004, Vormoor et al., 1996, Rees et al., 1986, Wells et al., 1993). Modifications to intensify the induction treatment include the use of diverse anthracycline drugs, shortening the intervals between initial cycles of chemotherapy, increasing cytarabine dosage (Gibson et al., 2011b, Creutzig et al., 2005, Rai et al., 1981) or the addition of DNA topoisomerase II inhibitor, etoposide, or purine analogues (Stevens et al., 1998) to improve complete remission and overall survival rates, and overcome resistance mechanisms in leukaemic cells (Hubeek et al., 2005, Gale, 1979). A combination of the standard chemotherapy with etoposide or guanine

9 analogue, 6-thioguanine, induced remission rates of up to 90% in paediatric patients (Creutzig et al., 2005, Gibson et al., 2011b).

Consolidation therapy is delivered post-remission to eradicate residual disease to prevent disease relapse. Treatment options include: stem cell transplant (autologous, syngeneic or allogeneic), high-dose chemotherapy, and investigational therapy (within a clinical trial). The most effective anti-leukaemic approach is allogeneic stem cell transplantation; however, it is associated with a high degree of initial mortality and a significant degree of long-term morbidity in the form of chronic graft-versus-host disease (GVHD), tending to offset the low likelihood of disease relapse (Stone et al., 2004). Chemotherapy-based approaches can be performed relatively safely, but a high chance for disease relapse remains, with many groups reporting relapse rates of 30% to 40%, EFS rates of 50%, and overall survival rates of 60% (Armendariz et al., 2005, Creutzig et al., 2005, Dluzniewska et al., 2005, Entz-Werle et al., 2005, Gibson et al., 2005, Kardos et al., 2005, Lie et al., 2005, Perel et al., 2005, Pession et al., 2005, Quintana et al., 2005, Ravindranath et al., 2005). The standard consolidation treatment for most paediatric AML trials is a total of two to five courses of high-dose cytarabine, combined with non-cross-resistant agents similar to what was given during induction (Lie et al., 2005, Perel et al., 2005, Pession et al., 2005, Ravindranath et al., 2005). A number of trials demonstrated that consolidation therapy with high-dose cytarabine ( 1 g/m2/dose) played a key role in improving patient outcome (Roboz, 2011). Haematopoietic stem cell transplants, however, yielded the best outcome for children with relapsed or refractory AML (Gibson et al., 2011a, Gorman et al., 2010). Improvements in paediatric AML outcome have also been attributed to improved supportive care, risk-stratification and management of infections (Jastaniah et al., 2012).

Due to the heterogeneity of AML, not all patients in each AML subset have the same response to treatment, as the chromosomal and genetic abnormalities vary within each population, making the treatment of AML challenging. Various therapeutic strategies are being investigated for the treatment of AML, including differentiation-inducing therapeutics, inhibitors of histone deacetylases, angiogenesis, signalling pathways (cell cycle, mTOR, PARP, Bcl-2, aminopeptidase, tyrosine kinase), and immunotherapy (Kumar, 2011). Disease relapse continues to contribute to poor AML patient outcome

(Stoiser et al., 2000) attributable to the persistence of chemotherapy-resistant LSC (Huntly and Gilliland, 2005), emphasising the need for LSC-targeted therapies.

1.2 Leukaemic stem cells

Disease relapse in AML patients can arise from residual LSC in the patient‘s bone marrow following chemotherapy. The cancer stem cell concept has been controversial and received a great deal of attention in recent years. LSC share similar properties with normal HSC; however, we have a limited understanding of the key genes and pathways which control LSC development and maintenance. Thus, identifying LSC-specific pathways is vital in order to develop therapies to selectively target LSC.

1.2.1 Normal haematopoietic stem cell properties

Tissue-specific stem cells, like embryonic stem cells, have the ability to self-renew as well as being capable of giving rise to mature effector cells in a sustained manner through differentiation. HSC are in routine clinical use in the context of grafts and bone marrow transplants for the treatment of a variety of blood cell diseases such as leukaemia and autoimmune disorders (Weissman, 2000). HSC account for about 0.01% of total nucleated cell in bone marrow in mice and approximately 5000 can be isolated from an individual mouse depending on the age, sex and strain of mice as well as purification method (Challen et al., 2009).

HSC reside within and utilise the normal bone marrow microenvironment known as the niche. The bidirectional signals between HSC and the niche allow for the regulation of normal HSC numbers (Calvi et al., 2003, Stier et al., 2005) and maintenance of quiescent long-term HSC (Fleming et al., 2008). Haematopoiesis takes place in a hierarchical order with multipotent progenitor cells being the immediate progeny of HSC which retain full lineage potential but have a limited capacity for self-renewal. Multipotent progenitors in turn give rise to oligopotent progenitors, which in turn give rise to more linage-restricted progenitors from which all the mature terminally differentiated, functional hematopoietic cells arise (Figure 1.3). The hierarchical model assists HSC in minimising the proliferative pressure on itself or being subjected to 11

HSC LT-HSC

ST-HSC

Multipotent MPP Progenitors

CMP CLP Oligopotent Progenitors MEP GMP

Pro-B Pro-T Pro-NK Lineage Restricted Progenitors

Mature effector cells Erythrocytes Macrophages B-cells NK-cells Platelets Granulocytes Dendritic cells T-cells

Figure 1.3: Classical model of haematopoiesis Haematopoietic stem cells (HSC) have the potential to self-renew, where long term-HSC (LT- HSC) give rise to short-term HSC (ST-HSC) with limited self-renewal capacity. ST-HSC give rise to multipotent progenitors (MPP), which differentiate into common myeloid progenitors (CMP) or common lymphoid progenitors (CLP). CMP differentiate into granulocyte- macrophage progenitors (GMP) or megakaryocyte-erythrocyte progenitors (MEP), which in turn give rise to the mature myeloid lineage cells. CLP differentiate into NK cell (Pro-NK), B- cell (Pro-B) and T-cell progenitors (Pro-T), which in turn differentiate into lymphoid haematopoietic lineages. Adapted from (Seita and Weissman, 2010).

12 potentially mutagenic hazards of DNA replication and cell division, which may contribute to the integrity and longevity of these cells by cycling very infrequently and primarily residing in the G0 phase of the cell cycle (Bradford et al., 1997). Stem cells express high levels of numerous ABC/multi-drug resistant transporter genes (Rossi et al., 2005, Zhou et al., 2001), which have physiological roles in cytoprotection.

Like all haematopoietic cells, HSC express distinct cell surface markers that allow them to be isolated and characterised using fluorescence activated cell sorting (FACS) (Figure 1.3). HSC express stem cell antigen (Sca-1), low levels of thymocyte differentiation antigen (Thy-1) expression, and do not express cell surface markers which characterise differentiated haematopoietic cell lineages (Lin) (Spangrude et al., 1988). HSC can also be identified from additional cell surface markers that enrich for HSC activity which include CD34 (Osawa et al., 1996), Flk2 (Adolfsson et al., 2001, Christensen and Weissman, 2001), CD105 (endoglin) (Chen et al., 2002a), and CD150 (Slamf1) (Kiel et al., 2005). In mice, HSC activity is enriched within the c-Kit (receptor for stem cell factor) positive, lineage negative, and Sca-1 positive fraction (KLS; c- Kithigh, Lin-, Sca-1+) of the bone marrow (Bryder et al., 2006).

1.2.2 Origin and identification of LSC

HSC possess a number of properties such as multipotency, ability to differentiate, self- renew, and proliferate, which have parallels with oncogenesis and malignancy. The concept of a rare population of cells in which tumours are derived from was first discussed in the late nineteenth century (Virchow, 1881), however, it was not until recently that putative cancer stem cells have been identified in cancers of the brain (Singh et al., 2003), breast (Al-Hajj et al., 2003), lung (Eramo et al., 2008, Ho et al., 2007), and haematopoietic system (Bonnet and Dick, 1997). It is also evident that many cancer stem cells can arise not only from normal stem cells (Kim et al., 2005, Barker et al., 2010, Barker et al., 2009, Fialkow et al., 1978, Wang et al., 2009, Ooi et al., 2010), but also from differentiated progenitor cells (Jamieson et al., 2004, Goldstein et al., 2010, Lim et al., 2009, Molyneux et al., 2010). The current definition of cancer stem cells is described as ―a small subset of cancerous population responsible for tumour initiation and growth, which also possess the characteristic properties of quiescence,

13 indefinite self-renewal, intrinsic resistance to chemo- and radiotherapy and capability to give rise to differentiated progeny‖ (O'Flaherty et al., 2012).

The strongest evidence for the existence of cancer stem cells was first demonstrated in AML using a xenogeneic transplant models (Lapidot et al., 1994, Bonnet and Dick, 1997). Early studies demonstrated that only the most primitive Lin- CD34+ CD38- fraction of human AML cells and not the more mature Lin- CD34+ CD38+ or CD34- populations were capable of transferring disease to non-obese diabetic/severe combined immunodeficient (NOD/SCID) mice (Bonnet and Dick, 1997). In the recipient mice, the cells differentiated into leukaemic blasts and recapitulated the phenotype of the disease observed in the patient and were able to give rise to AML in secondary recipients, indicating that these cells acquired self-renewal properties of LSC (Bonnet and Dick, 1997).

The controversy surrounding the cancer stem cell theory involves the limitations of current experimental models used to obtain the experimental evidence supporting the cancer stem cell model. The majority of the cancer stem cell experimental data involve transplanting putative cancer stem cells into immunocompromised mice (Hill, 2006, Kelly et al., 2007). One objection was that these cells are simply more adept at forming tumours in the xenogeneic microenvironment of an immunocompromised mouse. Another report using a genetically engineered lymphoma mouse model isolated a small subpopulation of stem-like cells based on murine stem cell markers Sca-1 and CD93, and found that tumour cells with Sca-1+ CD93high and Sca-1+ CD93low were equally capable of forming tumours in immunocompetent mice hence LSC may not be as rare as previously believed (Kelly et al., 2007). These studies infer that LSC may not be reliably identified by surface marker expression since the phenotype of LSC is heterogenous. Furthermore, the functional heterogeneity within the LSC compartment possesses differing self-renewal and engraftment capacity, similar to HSC (Hope et al., 2004). Without the cancer stem cell theory, it is unclear why this cancer stem cell-like population persists following chemotherapy and possesses high tumourigenicity if it is not necessary for continued propagation of the disease. The LSC-specific phenotype distinguishing LSC from the bulk of leukaemic blasts will be described in this chapter.

AML haematopoiesis according to the cancer stem cell model is similar to normal haematopoiesis, where a small population of LSC that possesses extensive self-renewal and undergoes limited differentiation, exists at the apex of an AML hierarchy, which gives rise to a large population of more mature leukaemic blasts which lack self-renewal capacity (Figure 1.4). The hierarchical organisation of the cancer stem cell model proposes that the most abundant cancer cells derive from a much smaller and often elusive pool of stem-like cells. This model was commonly accepted only in the field of leukaemia, however, it now appears to also apply to many solid tumours (Foreman et al., 2009). The transformation of HSC or committed progenitors by fusion oncogenes (Figure 1.4) capable of engendering limitless self-renewal may lead to LSC that are more immunophenotypically mature than HSC (Lane and Gilliland, 2010). The MLLAF9 rearrangement immortalises HSC or committed progenitors to form pre-LSC (Pandolfi et al., 2013), which represents LSC at an early developmental stage. In pre- LSC, MLLAF9 hijacks the self-renewal pathway to facilitate the development of fully transformed LSC through accumulation of further mutations (Krivtsov et al., 2006). LSC generated by retroviral transformation of committed granulocyte-macrophage progenitors (GMP) with oncogenes such as MLLT3-MLL, MLLT1-MLL or MYST3- NCOA2 (also known as MLLAF9, MLLENL and MOZ-TIF2, respectively) gave rise to a transplantable AML in vivo and had an immunophenotype similar to normal GMP, that is Linlow Sca-1− c-Kit+ FcγRII+ CD34+ (Cozzio et al., 2003, Huntly et al., 2004, Krivtsov et al., 2006, Scholl et al., 2007, Somervaille and Cleary, 2006). In addition, the transformation of GMP involving leukaemogenic fusion proteins has been shown to activate genes such as Hoxa9 and Meis1 (Cozzio et al., 2003, Huntly et al., 2004, Krivtsov et al., 2006). As enforced expression of these oncogenes in mouse bone marrow leads to rapid development of AML (Kroon et al., 1998), LSC in mice models of AML can also be induced by the coexpression of the Hoxa9 and Meis1a oncogenes (Wang et al., 2010).

The frequency of LSC can vary greatly between different AML samples, ranging from 1 in 104 to 107 cells (Bonnet and Dick, 1997). Other groups located LSC within both the CD34 and CD34 cell compartments and showed that frequencies were 0.2 to 200 per 106 mononuclear cells (Hope et al., 2003, Hope et al., 2004, Terpstra et al., 1996, Warner et al., 2004). Putative LSC in a MLLAF9 transgenic model of AML is found

HSC Pre-LSC

Multi-potent progenitors Mutations LSC that confer self-renewal capacity Progenitors

Leukaemic Mature blasts effector cells

Figure 1.4: Model of haematopoiesis in AML In AML, HSC or progenitor cells acquire a hit such as the fusion oncogene MLLAF9, which transforms these cells into pre-LSC. Pre-LSC can then develop into LSC when further mutations are acquired. LSC can self-renew or differentiate into leukaemic blasts which populate the bone marrow. Adapted from (Horton and Huntly, 2012).

16 within the Linlow c-Kit+ Sca-1+ compartment, enriched for HSC in normal haematopoiesis (Chen et al., 2008a) and this LSC immunophenotype is also observed in patient samples. The differentiated cells that form the bulk of the tumour population express low c-Kit levels whereas LSC are predominantly contained within the c-Kithigh population (Krivtsov et al., 2006, Somervaille and Cleary, 2006).

1.2.3 Disease relapse and drug resistance

LSC produce differentiated progeny and have the capacity for self-renewal (Reya et al., 2001), which allow these cells to possess the ability to instigate, maintain and serially propagate leukaemia in vivo. This has been observed in their ability to form tumours after xenotransplantation in immunodeficient mice. While most leukaemia cells are initially sensitive to chemo- and radiotherapy, current AML chemotherapy regimens predominantly target the highly proliferative leukaemic blasts (Huntly and Gilliland, 2005). LSC are frequently resistant to therapy and are therefore considered to be the basis for disease relapse (Perona and Sanchez-Perez, 2004), emphasising the need for targeted LSC therapy (Figure 1.5).

Similar to HSC, LSC home to and engraft to the bone marrow niche. The high resistance of LSC to conventional chemotherapy has been attributed to LSC taking refuge in the endothelial bone marrow niche during chemotherapy, where they are protected from chemotherapy-induced apoptosis, potentially through niche-induced LSC quiescence (Blair et al., 1997). LSC receive vital cues from the bone marrow microenvironment that dictate their behaviour and eventual disease phenotype. For instance, the leukaemia phenotype in a human xenograft model of MLLAF9 leukaemia could be altered between lymphoid, biphenotypic or myeloid leukaemia by the expression of human cytokines stem cell factor/Kit ligand (SCF/KITLG), colony stimulating factor 2 (CSF-2) and interleukin 3 (IL-3) in the microenvironment (Wei et al., 2008). The relationship between LSC and the niche is not unidirectional, as normal HSC can be altered by signals within a pathological niche to cause haematopoietic disorders or dyscrasias (Walkley et al., 2007a, Walkley et al., 2007b) and consequently re-emerging to initiate disease relapse (Lane et al., 2009). Since LSC remain sheltered in the bone marrow microenvironment, exhibit resistance to chemotherapy, and

Leukaemic blasts lose ability to generate new cells

LSC targeted Tumour therapy degenerates resulting in remission

LSC

LSC LSC Disease Conventional relapse therapy LSC persist and repopulate bone marrow with leukaemic blasts

Figure 1.5: Conventional chemotherapy and LSC-targeted therapy Conventional chemotherapies predominantly target leukaemic blasts but the more chemoresistant LSC remain after therapy and re-establish the tumour, causing disease relapse. Therapies which target LSC specifically may effectively kill the stem cells resulting in tumour degeneration. Adapted from (Reya et al., 2001).

18 serve as the origin of relapse capable of propagating leukaemia, it is crucial to target LSC in order to improve AML patient outcome. Examining the LSC microenvironment and dysregulated signalling pathways may also reveal potential LSC targets and further our understanding of LSC biology, enabling the potential development of LSC-specific anti-cancer therapy.

1.2.4 Hypoxia and glucose metabolism

It has been demonstrated that the bone marrow microenvironment can aid the growth and spreading of leukaemic cells, conferring chemoresistance and survival to LSC. The microenvironment of solid tumours have been shown to be hypoxic (low levels of O2) (Milosevic et al., 2004). Similarly, the bone marrow microenvironment becomes altered in AML, where leukaemia progression is associated with marked expansion of hypoxia compared to normal bone marrow (Mortensen et al., 1998). This has also been observed in the bone marrow environment involving multiple myeloma (Colla et al., 2010).

Key regulators of oxygen homeostasis are the hypoxia-inducible factor 1 (Hif1) transcription factors (Iyer et al., 1998, Loboda et al., 2010). One member of this family, Hif1a, whose expression is directly influenced by hypoxia (Drolle et al., 2015), has been observed to be highly expressed in multiple myeloma cells (Colla et al., 2010) and primary childhood ALL bone marrow specimens (Wellmann et al., 2004). The activation of Hif1a results in two major effects on cellular metabolism. Hif1a stimulates glycolytic production by promoting the transcription of genes involved in extracellular glucose import (such as Slc2a1/Glut1) and glycolytic breakdown of intracellular glucose (such as aldolase C). Hif1a also downregulates oxidative phosphorylation within the mitochondria by activating genes such as pyruvate dehydrogenase kinase 1 (Pdk1) (Kim et al., 2006, Papandreou et al., 2006). The result of these two effects is a promotion in the switch from mitochondrial oxidative phosphorylation to glycolytic metabolism, thereby favouring the conversion of glucose to lactate, where tumour cells have sufficient energy in a hypoxic environment (Figure 1.6) (Semenza, 2007). This is known as the Warburg effect, based on Otto Warburg and his co-workers pioneering research into cancer cell metabolism, showing a 10-fold increase in glucose consumption rate and 7-fold increase in lactate production in carcinoma tissue

A Quiescent normal cell

B Proliferating tumour cell

Figure 1.6: Cellular metabolism of normal and tumour cells The cellular metabolism of normal cells (a) compared to tumourigenic cells (b). A shift to aerobic glycolysis in tumour cells activates mTOR, which in turn enhances hypoxia-inducible factor 1 (HIF1) activity, further promoting glycolysis. Hif1 increases the expression of glucose transporters (GLUT), glycolytic enzymes and pyruvate dehydrogenase kinase, isozyme 1 (PDK1), which blocks pyruvate from entering the TCA cycle. Oct1 (also known as POU2F1) activates the transcription of glycolysis genes and suppresses oxidative phosphorylation. Pyruvate kinase M2 (PKM2) diverts substrates from the glycolysis pathway into alternative biosynthetic and reduced nicotinamide adenine dinucleotide phosphate (NADPH)-generating pathways. Adapted from (Cairns et al., 2011).

20 compared to normal liver tissue (Warburg et al., 1927). ATP production from glycolysis (approximately 4 mol ATP/mol glucose for aerobic glycolysis; approximately 4 mol ATP/mol glucose for anaerobic glycolysis) is less efficient than oxidative phosphorylation (approximately 36 mol ATP/mol of glucose) (Vander Heiden et al., 2009), therefore indicating that cancer cells implement an abnormally high rate of glucose uptake to meet the demands of the proliferating tumour cell (Warburg, 1956). Clinical data have also emerged supporting the importance of glucose in a number of malignancies (Jadvar et al., 2009, Gambhir, 2002).

Since Hif1a reduces mitochondrial function in tumour cells, it has been suggested that this should reduce the amount of metabolically produced reactive oxygen species (ROS) from the mitochondria (Chandel and Schumacker, 2000, Droge, 2002). High ROS levels can be toxic to cells causing oxidative stress by damaging cellular macromolecules such as lipids, proteins and DNA (Collins et al., 2005, Miyamoto et al., 2003). A recent study observed AML LSC to have a ROS-low profile (Lagadinou et al., 2013), thus minimising ROS-induced damage.

Although there is evidence of enhanced glycolysis under hypoxic conditions in AML cell lines (Lodi et al., 2011), and resistance to chemotherapy have been associated with high levels of cellular glycolysis in solid cancers (Gatenby and Gillies, 2004), limited number of studies have been conducted investigating the metabolic dependence and changes occurring in AML LSC. Pyruvate kinase M2 (Pkm2) is a key regulator of glycolysis as it catalyses the rate-limiting step of glycolysis and thus ATP production, and its expression is abundant in AML cell lines and primary AML patient samples (Sturgill and Guzman, 2013). The oncoprotein MYC has also been shown to promote Pkm2 expression (David et al., 2009). Pkm2 acts by inhibiting glycolysis by shuttling substrates through the pentose phosphate pathway and other pathways involved in macromolecule production which are required for redox maintenance, cell proliferation, and reducing equivalents such as nicotinamide adenine dinucleotide phosphate (NADPH) (Vander Heiden et al., 2009, Marshall et al., 1991, Christofk et al., 2008). Although this opposes the Warburg effect, Pkm2 can exist in either active tetramers or inactive dimers, where the inactive form is predominant in tumour cells (Hathurusinghe et al., 2007, Christofk et al., 2008). Phosphorylation at the tyrosine residue 105 (Y105) inhibits the formation of active Pkm2, thus promoting the Warburg effect by providing

21 a metabolic advantage to tumour cells and promoting tumour growth (Hitosugi et al., 2009). A number of haematopoietic cancer cell lines have been shown to express phosphorylated Y105, including HEL (contains the Janus kinase 2 (JAK2) Val617Phe mutant), KG-1a (contains FOP2-FGFR1 fusion tyrosine kinase), Mo91 (contains the est variant 6 (ETV6)-neurotrophic tyrosine kinase, receptor, type 3 (NTRK3) fusion tyrosine kinase), Molm14 (contains fms-related tyrosine kinase 3 (FLT3)-internal tandem duplication (ITD) mutant), and K562 (contains breakpoint cluster region (Bcr)- Abl fusion tyrosine kinase) (Hitosugi et al., 2009). It was also recently reported that Pkm2 deficient mice retarded the progression of established leukaemia induced by Bcr- Abl or MLLAF9 (Israelsen et al., 2013, Wang et al., 2014c), suggesting that the glycolytic pathway is important for leukaemia maintenance and progression.

1.2.5 Self-renewal pathways

Studies have shown that pathways governing cellular proliferation, differentiation and apoptosis in normal stem cells are also prominent in cancer stem cells. As LSC have the unique capacity to self-renew, targeting pathways specifically governing LSC self- renewal has the potential to enable these cells to be selectively killed, while sparing HSC. The limitless self-renewal characteristic of LSC has been implicated to be governed by pathways such as Wnt/-catenin (Reya and Clevers, 2005), Notch (Maillard et al., 2003) and Hedgehog (Bhardwaj et al., 2001). The importance of different self-renewal pathway appears to be dependent on the subtype of AML and oncogene involved as evidenced by the importance of Hedgehog pathway signalling in chronic myeloid leukaemia (CML) LSC (Sengupta et al., 2007, Irvine and Copland, 2012), but not in MLLAF9-induced AML (Hofmann et al., 2009, Gao et al., 2009).

The Wnt signalling pathway is a highly conserved signalling system that has been associated with proliferation, differentiation, self-renewal and motility in embryonic development and regeneration of adult tissues, which are key properties in carcinogenesis (Hanahan and Weinberg, 2000, Zhao et al., 2007). Dysregulation of Wnt signalling has also been associated with numerous malignancies, such as breast, colon, gastric, head and neck, hepatocellular, ovarian, prostate and non small cell lung cancer

Figure 1.7: Schematic diagram of the Wnt/-catenin signalling pathway Components of the Wnt signalling pathway. In the absence of Wnt (a), β-catenin is degraded and target genes are repressed. In the active state (b), Wnt binds to the Frizzled receptors and LRP5/6 co-receptors and β-catenin translocates to the nucleus to bind to T-cell factor (TCF) and lymphoid enhance-binding protein (LEF)-family transcription factors, activating transcription. Adapted from (Moon et al., 2004).

(Clevers and Nusse, 2012, Yao et al., 2011) and facilitates gene expression of intracellular -catenin. In the Wnt pathway, the Wnt-ligand is a secreted glycoprotein that binds to Frizzled receptors, which triggers a signalling cascade resulting in displacement of the multifunctional kinase GSK-3β from the APC/Axin/GSK-3β- complex (Figure 1.7). In the absence of Wnt, the transcriptional co-regulator, β-catenin, is targeted for degradation by the APC/Axin/GSK-3β-complex. Phosphorylation of β- catenin by the coordinated action of CK1 and GSK-3β leads to its ubiquitination and proteasomal degradation through the β-TrCP/SKP complex. In the presence of Wnt binding, Dishevelled (Dvl) is activated, which in turn recruits GSK-3β away from the degradation complex. This allows for stabilization of β-catenin levels, Rac1-dependent nuclear translocation and recruitment to the LEF/TCF DNA-binding factors, where β- catenin acts as an activator for transcription.

Wnt/β-catenin signalling is perturbed in a leukaemic setting (Wang et al., 2010) and studies have shown that dysfunctional β-catenin signalling imparts self-renewal properties to haematopoietic progenitors in MLLAF9-AML (Meyer et al., 2013). Microarray analysis also found β-catenin to be the only Wnt pathway component expressed at significant levels in AML LSC (Krivtsov et al., 2006). β-catenin is a prognostic marker and driver of disease progression in AML patients (Ysebaert et al., 2006) and required for the development of LSC in AML (Wang et al., 2010). β-catenin knockdown experiments using shRNA showed impaired leukaemogenesis in MLLENL transformed AML LSC in vivo (Yeung et al., 2010) and delayed disease onset in recipient mice. Furthermore, knockdown of β-catenin in AML patient samples significantly impaired AML cell proliferation and blast cell count in vitro (Yeung et al., 2010). The Wnt/-catenin signalling pathway was also found to be dispensable in normal adult HSC (Cobas et al., 2004, Jeannet et al., 2008, Koch et al., 2008), suggesting β-catenin may be an ideal therapeutic target. However, targeting β-catenin has proven difficult due to its structure and lack of discernible enzymatic activity (Polakis, 2012, Curtin and Lorenzi, 2010). Since the Wnt/β-catenin pathway plays an important role in AML LSC, upstream or downstream targets of β-catenin need to be identified in order to develop therapeutics against this aberrantly activated pathway.

1.3 Epigenetics

Studies into the epigenetic dysregulation of Wnt/β-catenin signalling pathway have shown that abnormal DNA methylation of Wnt antagonists was associated with decreased 4-year relapse-free survival in AML (P = 0.03) (Valencia et al., 2009). Similar results were also observed in primary multiple myeloma samples, where Wnt antagonists were hypermethylated, resulting in transcriptional repression (Chim et al., 2007). Furthermore, several studies have shown the involvement of MLL fusion partners in multiprotein complexes mediating transcriptional control and elongation (Slany, 2005). ELL acts as an elongation factor that associates with RNA polymerase II (Shilatifard et al., 1996), while AF4, AF9 and ENL contain C-terminal transcriptional activation domains (Krivtsov and Armstrong, 2007) and AF10 associates with DOT1L, a histone methyltransferase that methylates histone lysine 79 residue (H3K79) (Okada et al., 2005). In recent years, numerous reports have emerged regarding the role of epigenetic regulation in cancer and studying this regulation is a promising approach in targeting β-catenin.

1.3.1 Epigenetic regulation

In addition to genetic alterations, it is now well established that pathogenesis can arise from epigenetic dysfunctions. Epigenetics can be defined as the heritable changes in gene expression that are not due to alterations in the DNA sequence (Holliday, 1987). Epigenetics control the organisation of the genome and gene activity. DNA is packaged around histones to form nucleosomes, the repeating core unit of chromatin. The nucleosome consists of 146 bp of DNA wrapped around an octamer of four core histones (H2A, H2B, H3 and H4) (Luger et al., 1997). Within chromatin, a vast number of posttranslational modifications can occur to the DNA and to the histones around which DNA is bound (Sarma and Reinberg, 2005, Klose and Bird, 2006, Kouzarides, 2007, Zlatanova et al., 2009, Allfrey et al., 1964). These epigenetic modifications not only regulate normal growth and development but epigenetic malfunction underpins many complex diseases including neurological diseases (Molfese, 2011, Jakovcevski and Akbarian, 2012), cancer (Chim et al., 2007, Sharma et al., 2010a, Esteller, 2008, Jones and Baylin, 2002) and diabetes (Reddy et al., 2015, Ling and Groop, 2009).

The complex interplay of posttranslational modification of histone tails on specific residues coupled with the DNA methylation status at particular loci determines chromatin folding and its switch from heterochromatin, transcriptionally silent, to euchromatin, transcriptionally active, and vice versa (Figure 1.8). Transcriptional activation and repression are highly regulated, dynamic events and controlled by enzymes that often exist in macromolecular complexes. The epigenetic mechanisms that modify chromatin structure can be divided into two main categories: DNA methylation and histone modifications.

1.3.1.1 DNA methylation

Methylation of cytosine bases in DNA is the most extensively studied epigenetic modification. DNA methylation provides a stable gene silencing mechanism likely due to steric hindrance of transcription complexes binding to regulatory DNA. DNA methylation plays an important role in gene expression and chromatin architecture, in association with histone modifications and other chromatin associated proteins (Bernstein et al., 2007). CpG-rich DNA termed ‗CpG islands‘ are concentrated in regions of the human genome (Suzuki and Bird, 2008, Bird, 2002) and are preferentially located at the 5‘ end of genes and occupy approximately 60% of human gene promoters (Wang and Leung, 2004). Most CpG sites in the genome are methylated, but the majority of CpG islands remain unmethylated during development and in differentiated tissues (Shen et al., 2007).

DNA methylation by DNA methyltransferases can lead to gene silencing by preventing recruitment of regulatory proteins to DNA such as the blocking of transcriptional factors from accessing target-binding sites (Figure 1.8). Alternatively, methylation can provide binding sites for methyl-binding domain proteins, which can mediate gene expression through interactions with histone deacetylases and histone methyltransferases prompting coordinated epigenetic modifications of the surrounding chromatin (Zhang and Reinberg, 2001, Bird and Wolffe, 1999). Aberrant DNA methylation patterns were the first examples of epigenetic deregulation to be characterised in cancer. Examples of deregulated DNA methylation in cancer include solid tumours (Hansen et al., 2011, Irizarry et al., 2009) and AML, where AML

Figure 1.8: Epigenetic modifications to DNA and histones dictate nucleosome spacing and transcription Modifications such as methylation of DNA and histones at specific residues cause nucleosomes to pack tightly together (heterochromatin), preventing transcription factors from binding to the DNA, inactivation gene transcription (top). Modifications such as histone acetylation at specific residues results in loose packing of nucleosomes (euchromatin) (bottom). This enables transcription factors to bind to the DNA and genes are expressed. Adapted from (Choi, 2013).

27 containing PML-RARa and AML1-ETO aberrations showed highly distinct methylation profiles (Figueroa et al., 2010b, Cancer Genome Atlas Research Network, 2013). Transcriptional silencing of CDKN2a and CDKN1a via DNA hypermethylation have been associated with poor clinical outcome in acute leukaemias (Bhalla, 2005).

The most frequently mutated DNA methyltransferase in myeloid malignancies is DNA methyltransferase 3a (DNMT3A), where mutations are found in up to 22% of AML, 8% of myelodysplastic syndrome and 15% of myeloproliferative neoplasm cases (Shih et al., 2012). Patients with DNMT3A mutations were associated with an increased risk of relapse and decreased overall survival (Thol et al., 2011). Missense mutations at R882 occur in approximately 60% of DNMT3A mutant samples and cause a reduction in DNA binding affinity and catalytic activity (Gowher et al., 2006, Yamashita et al., 2010). AML patients with DNMT3A mutations are enriched in classes M4 or M5 (FAB classification) (Yan et al., 2011) and frequently harbour nucleophosmin (NPM1), FMS- related tyrosine kinase 3 (FLT3) and isocitrate dehydrogenase 1 (IDH1) mutations (Ley et al., 2010).

Like DNMT3A, mutations in tet methylcytosine dioxygenase 2 (TET2), IDH1 and IDH2 have been found to persist from diagnosis to relapse in AML (Wakita et al., 2012). TET family of proteins (TET1-3) are 2-oxoglutarate-Fe(II)- and α-ketoglutarate-dependent enzymes that catalyse the conversion of 5-methylcytosine to 5-hydroxymethylcytosine (Tahiliani et al., 2009). TET2 mutations occur in up to 25% of myelodysplastic syndrome, 13% of myeloproliferative neoplasm and 23% of AML cases (Shih et al., 2012) and associated with poorer prognosis in cytogenetically normal AML (Metzeler et al., 2011). Normal DNA-binding-proteins that confer transcriptional silencing are blocked by 5-hydroxymethylcytosine (Valinluck et al., 2004) and is therefore associated with increased gene expression when the 5-hydroxymethycytosine enrichment in CpG dinucleotides occur near transcriptional start sites and intragenic regions (Pastor et al., 2011).

TET2 and IDH mutations are mutually exclusive in AML and share similar methylation profiles, characterised by global promoter hypermethylation and reduced 5- hydroxymethylcytosine (Figueroa et al., 2010a). IDH1 and IDH2 catalyse the conversion of isocitrate to α-ketoglutarate in the Krebs cycle and would normally produce 2-oxoglutarate needed for TET2 activity. Mutations in IDH convert α-

28 ketoglutarate to 2-hydroxygutarate (Ward et al., 2010, Dang et al., 2009), which inhibits TET family members (Figueroa et al., 2010a, Xu et al., 2011) and jumonji-domain- containing (JmjC) family of histone lysine demethylases (Chowdhury et al., 2010) as they are both α-ketoglutarate-dependent enzymes. Mutations in IDH1 and IDH2 are present in up to 20% of de novo normal karyotype AML and 20% of secondary AML cases (Kosmider et al., 2010, Pardanani et al., 2010, Patnaik et al., 2010, Zhou et al., 2012) where AML patients with this mutation are associated with poorer overall survival and lower complete remission rates (Feng et al., 2012, Im et al., 2014). Alterations in these epigenetic regulators have led to greater risk stratifications for AML and investigations into the epigenome are in progress to determine its contribution to myeloid leukaemogenesis. To determine the transcriptional status of genes, DNA methylation acts in coordination with histone modifications (Guil and Esteller, 2009), which can be regulated by various epigenetic enzymes.

1.3.1.2 Histone modifications

Histones contain NH2-terminal tails which protrude from nucleosomal cores and are targeted for posttranslational modifications by different classes of enzymes (Allfrey et al., 1964, Cheung et al., 2000, Kouzarides, 2007, Strahl and Allis, 2000). These enzymes can be divided into three categories based on their epigenetic activity and function: writers (enzymes that add chemical marks to histone substrates); erasers (enzymes that remove chemical covalent modification from histone substrates) and readers (proteins that recognise specific histone marks and initiate downstream target gene regulation) (Table 1.4) (Strahl and Allis, 2000, Jenuwein and Allis, 2001, Kouzarides, 2007, Mai and Altucci, 2009). Covalent modifications include acetylation of lysines, methylation of lysines and arginines, phosphorylation of serines and threonines, ADP-ribosylation of glutamic acids and ubiquitination and sumoylation of lysine resides (Figure 1.9) (Kouzarides, 2007). Through these histone modifications, intra- and interneucleosomal contacts are altered via changes in steric or charge interactions. These histone marks can also regulate the modification status of another mark when the modifications occur within the same histone tail or among tails involving one or multiple nucleosomes (Ahn et al., 2006), adding another level of complexity.

Table 1.4: Classes of epigenetic writers, erasers, and readers of histone marks and their regulatory function Adapted from (Arrowsmith et al., 2012). Class of Epigenetic Regulator No. of Proteins Major Classes and Functions Writers Histone 18 -MYST family proteins: involved in DNA damage and oncogenic translocation methyltransferases -GNAT: involved in EGF signalling and cell cycle progression -EP300: promiscuous (range of cellular events) Protein 60 -SET domain: methylates both histone and non-histone lysines methyltransferases -PRMT: methylates both histone and non-histone arginines -PRDM: SET domain-like tissue-specific factors Erasers Histone deacetylases 17 -Classes I, IIb and IV: histone and non-histone substrates, involved in gene silencing -Class IIa: scaffolding proteins -Sirtuins (Class III): NAD-dependent, deacetylation and ADP-ribosylation activity Lysine demethylases 25 -Lysine-specific demethylases: flavin-dependent enzymes that regulate transcription during development -Jumonji domain: 2-oxoglutarate-dependent Readers Bromodomain- 61 -Targeting of chromatin-modifying enzymes to specific sites, often linked to containing proteins PHD fingers and the catalytic domain of histone acetyltransferases

Methyl-lysine- and/or 95 -Tudor domains: bind dimethylated lysine, trimethylated lysine and methyl-arginine- dimethylated arginine binding domain- MBT domains: bind monomethylated and dimethylated lysine with low containing proteins sequence specificity -Chromodomains: bind trimethylated lysine with sequence specificity -PWWP domains: bind to both trimethylated lysine and DNA PHD-containing 104 -Large and diverse family that act on multiple substrates proteins

P A A H2A SGRGKQBBKARAKAK-

A A P A A H2B PEPSKSAPAPKKGSKKAITKAQKK-

P A P P A A A P H3 ARTKQTARKSTGGKAPRKQLATKAARKSAPATGGVKKP- M M M M MM M

P A A A A P H4 SGRGKGGKGLGKGGARKRHRKVLR- M M

Figure 1.9: Covalent modifications of the N-terminal tail of core histones Each nucleosome contains four core histone proteins, H2A, H2B, H3 and H4. The histone tails which protrude from each core are subject to posttranslational modifications at specific residues. Phosphorylation is indicated as P in yellow, acetylation indicated as A in red, and methylation indicated as M in blue. Adapted from (Lund and van Lohuizen, 2004).

1.3.1.2.1 Histone acetylation

Histone acetylation and deacetylation are governed by the activity of histone acetylases and histone deacetylases (Table 1.5), where the acetylation of lysine residues on histones 3 and 4 are strongly associated with transcriptional activation. It has also been associated with DNA replication and repair, and chromatin assembly (Roth and Allis, 1996, Gong and Miller, 2013). Changes in the balance of histone acetylation and deacetylation can lead to cancer (Kouzarides, 2007).

The three main families of histone acetyltransferases are the MOZ/YBF2/SAS2/TIP60 (MYST) family, the Gen5-related N-acetyltransferase (GNAT) family and the CBP/p300 family (Yang, 2004). Histone acetyltransferases are recruited as co-activators of transcription by transcription factors usually in the context of large chromatin remodelling complexes. Histone acetyltransferases such as PCAF, p300, and CBP can also acetylate multiple non-histone proteins, which alter protein stability, cellular localisation and protein-nucleotide/protein-protein interactions. These include p53, nuclear factor-B, (NF-B), p65, CBP, p300, c-Myc, Hif1a, heat shock protein (Hsp)- 90, E2F, Bcr-Abl, FLT3 kinase and c-Raf kinase (Yang and Seto, 2008, Glozak et al., 2005, Yoo and Jones, 2006). Missense mutations of p300 have been identified in a variety of solid tumours including colorectal, gastric, breast, and pancreatic cancers (Iyer et al., 2004). Tip60 and its involvement in the regulation of the transcriptional activation of p53 and Myc play an important role in regulating tumourigenesis (Tang et al., 2006, Frank et al., 2003). Furthermore, translocations involving histone acetyltransferases have been identified in AML and ALL, in which t(11;16)(q23;p13) results in MLLCBP (Ayton and Cleary, 2001), and p300 can aberrantly fuse to MLL via the t(11;22)(q23;p13) translocation (Ida et al., 1997).

The role of histone deacetylases is to remove acetyl groups from lysine residues of histone tails and non-histone substrates. There are eighteen histone deacetylase members which can be categorised into four classes based on their homology (de Ruijter et al., 2003): class 1 (HDAC 1, 2, 3 and 8) is localised to the nucleus; class II (HDAC 4, 5, 6, 7, 9 and 10) found in both nucleus and cytoplasm; class III (sirtuins SIRT 1-7); and class IV (HDAC 11) which display features of both class I and II histone deacetylases. Class I, II and IV histone deacetylases share homology in both

Table 1.5: Summary of histone acetylase specificity on histone 3 (H3) and histone 4 (H4) lysine residues Adapted from (Kouzarides, 2007). Residue Histone Acetylases H3K9 PCAF/GCN5 H3K14 PCAF/GCN5 CBP/p300, TIP60, ScSAS3 H3K18 PCAF/GCN5 CBP/p300 H3K23 ScSAS3 H3K56 ScRTT108 H4K5 SBP/p300, HAT1, TIP60, HB01 H4K8 SBP/p300, TIP60, HB01 H4K12 HAT1, TIP60, HB01 H4K16 TIP60, ScSAS2 H2AK5 SBP/p300 H2BK12 SBP/p300 H2BK15 SBP/p300

34 sequence and structure and all require a zinc ion for catalytic activity (de Ruijter et al., 2003). In contrast, the sirtuin family shares no similarities in sequence or structure with class I, II or IV histone deacetylases and requires nicotinamide adenine dinucleotide (NAD+) for catalytic activity, with SIRT 1, 6 and 7 localised in the nucleus, SIRT 3-5 in the mitochondria and SIRT2 in the cytoplasm. (Blander and Guarente, 2004). Histone deacetylases show no specificity for a particular acetyl group, except SIRT2 which targets H4K16 (Kouzarides, 2007) The acetylation of histones neutralises the positive charge of histones and loosens their interaction with negatively-charged DNA, resulting in an open chromatin structure and gene transcription (Grunstein, 1997). Therefore, removal of an acetyl group from lysine residues by histone deacetylases condenses chromatin structure and inhibits gene transcription.

Histone deacetylases can also function in deacetylating non-histone proteins such as transcription factors and signalling molecules (Glozak et al., 2005). Deacetylation of p53 and RUNX3 results in degradation via the ubiquitin-proteasome pathway (Ito et al., 2002, Jin et al., 2004), while deacetylation of HSP90 by HDAC6 is required for stability and function of Bcr-Abl, c-RAF and AKT (Bali et al., 2005). The regulatory roles of histone deacetylases are cell-type and context-dependent. Knockdown of HDAC1 and 3 in HeLa inhibited cell proliferation, whereas knockdown of HDAC4 and 7 had no effect (Glaser et al., 2003). Silencing of HDAC1 resulted in cell cycle arrest, cell growth inhibition and induction of apoptosis in osteosarcoma and breast cancer cells (Senese et al., 2007), while silencing of HDAC1 and 2 suppressed cell growth in colon cancer cells (Weichert et al., 2008). Similarly knockdown of HDAC1 and 2 but not 3, 6 and 7 sensitises CML cells for TRAIL-induced apoptosis (Inoue et al., 2006). In acute promyelocytic leukaemia, HDAC3 acting within a repressor complex targets PML- RARa promoters to repress transcription, restoring retinoic acid dependent genes (Atsumi et al., 2006). AML1-ETO fusion protein recruits HDAC1, 2 and 3 to repress transcription of target genes (Amann et al., 2001).

1.3.1.2.2 Histone methylation

Histone methylation can activate or repress gene transcription depending on the site and degree of methylation. Histone lysine residues can occur in four states: unmethylated,

35 monomethylated, dimethylated, and trimethylated. Histone methyltransferases add methyl groups to lysine and arginine residues and are removed by two families of histone lysine demethylases, lysine-specific demethylases (LSDs) and the Jumonji- domain (JMJD)-containing protein family. Unlike histone acetyltransferases and deacetylases, histone methyltransferases and demethylases show high specificity for not only lysine residues but also methylation status (Tables 1.6 and 1.7).

The action of histone methyltransferases was first recognised in genetic studies in the fruit fly, where Su(Var)3-9 (mammalian homologues Suv39h1 and Suv39h2 (Peterson and Laniel, 2004)) was required for gene silencing (Lachner et al., 2003, Sinclair et al., 1992). These histone methyltransferases contain the SET (Su(var)3-9, Enhancer of Zeste, Trithorax) domains. There are three families of histone methyltransferases identified to date: SET-domain containing proteins (Rea et al., 2000), DOT1-like (DOT1L) proteins (Feng et al., 2002), and arginine N-methyltransferase (PRMT) family (Bannister and Kouzarides, 2011). The polycomb group (PcG) proteins which were also originally defined in Drosophila, govern H3K27 methylation, where polycomb repressor complexes (PRC) 1 and 2 mediate gene silencing through H3K27 trimethylation (Sparmann and van Lohuizen, 2006). PRC2 is comprised of enhancer of zeste homologue 2 (Ezh2), suppressor of zeste 12 (SUV12) and embryonic ectoderm development (Eed). Ezh2, a SET domain methyltransferase for H3K27, is highly expressed in prostate, breast, colorectal, endometrial and bladder cancer, lymphoma, myeloma and melanoma (Moss and Wallrath, 2007), and has been shown to regulate embryonic development and stem cell renewal via association with histone deacetylases and DNA methyltransferases (Bracken et al., 2006, Sparmann and van Lohuizen, 2006). Another protein that also promotes transcriptional repression via association with DNA methyltransferases and histone deacetylases is heterochromatin protein (HP) 1, which binds to methylated H3K9 (Fuks et al., 2003, Wang et al., 2007b).

Chromosomal translocations involving histone methyltransferases appear in AML, where 5% of childhood AML contain the translocation between nucleoporin 98 (NUP98) and the H3K36 methyltransferase nuclear receptor binding SET domain protein, NSD1 (Cerveira et al., 2003). The fusion of NUP98 to NSD1 (Wang et al., 2007c, Jaju et al., 2001) and NSD3 (Rosati et al., 2002) gives rise to AML whilst

Table 1.6: Summary of histone methyltransferase and demethylase specificity on histone 3 (H3) and histone 4 (H4) lysine residues Adapted from (Kouzarides, 2007, Greer and Shi, 2012).

Residue Histone Histone Transcriptional Role methyltransferases demetehylases H3K4 MLL1, MLL2, MLL3, LSD1, H3K4me2-Activation MLL4, SET1A, SET1B, JARID1A-D H3K4me3-Activation ALL-1, SET7, ALR-1/2, Activation ALR, ASH1, SMYD3, PRMD9 H3K9 Suv39h1, Suv39h2, G9a, LSD1/AR, H3K9me1-Activation EHMT1/GLP, JHDM2A-B, H3K9me2-Repression ESET/SETDB1, CLL8, JMJD2A-D H3K9me3-Repression SpClr4, RIZ1, PRMD2 Repression, Activation H3K27 EZH2, EZH1 UTX, JMJD3 H3K27me1-Activation H3K27me3-Repression Repression H3K36 HYPB, SMYD2, NSD1, JHDM1A-B, H3K36me2-Activation NSD2, NSD3, SET2 JMJD2A, H3K36me3-Activation, JMJD2C, Elongation FBXL10 H3K79 DOT1L H3K79me1-Activation H3K79me3- Activation Activation H4K20 PR-Set7/8, Suv420h1, PHF8 H4K20me1-Activation Suv420h2, SET9 H4K20me3-Repression Repression

Table 1.7: Family of histone demethylases and their methylated substrate Adapted from (Kouzarides, 2007, Greer and Shi, 2012).

Name Synonyms Targets Transcriptional role KDM1A LSD1, AOF2 H3K4me2/me1 Repression H3K9me2/me1 Activation KDM1B LSD2, AOF1 H3K4me2/me1 Repression KDM2A FBXL11A, JHDM1A H3K36me2/me1 Repression KDM2B FBXL10B, JHDM1B H3K36me2/me1 Repression H3K4me3 Repression KDM2C JMJD1C, TRIP8, JHDM2c H3K9me2/me1 Activation KDM3A JMJD1A, JHDMA H3K9me2/me1 Activation KDM3B JMJD1B, JHDM2B H3K9me2/me1 Activation KDM4A JMJD2A, JHDM3A H3K9me3/me2 Activation H3K36me3/me2 Repression KDM4B JMJD2B H3K9me3/me2 Activation H3K36me3/me2 Repression KDM4C JMJD2C, H3K9me3/me2 Activation GASC1/JMJD2A H3K36me3/me2 Repression KDM4D JMJD2D H3K9me3/me2/me1 Activation H3K36me3/me2 Repression KDM4E JMJD2E H3K9me3/me2 Activation KDM5A JARID1A, RBP2 H3K4me3/me2 Repression KDM5B JARID1B, PLU1 H3K4me3/me2 Repression KDM5C JARID1C, SMCX H3K4me3/me2 Repression KDM5D JARID1D, SMCY H3K4me3/me2 Repression KDM6A UTX, MGC141941 H3K27me3/me2 Activation KDM6B JMJD3, KIAA0346, H3K27me3/me2 Activation PHF8, KIAA1111, H3K9me2/me1 Activation ZNF422 H4K20me1 Repression KDM7 KIAA1718 H3K9me2/me1 Repression H3K27me2/me1 Activation KDM8 JMJD5, FLJ13798 H3K36me2 Repression

38 translocations containing NSD2 lead to multiple myeloma (Chesi et al., 1998, Stec et al., 1998).

The complex interaction of repressive and active marks has been reported, where the expression of EZH2 and NSD2, responsible for H3K27me3 and H3K36me2, respectively, are tightly correlated in cancers (Asangani et al., 2013). EZH2 functions upstream and initiates transcriptional repression through methylation at H3K27me3. A complex network of microRNAs regulated by H3K27me3 cooperates to upregulate NSD2 which results in increased H3K36me2, associated with transcriptional activation.

Reports on the involvement of MLL fusions in epigenetic regulation include the H3K4 methyltransferase, MLL1, a frequent target of chromosomal translocations in AML and ALL, which has been implicated in regulating proliferation in normal HSC (McMahon et al., 2007). The H3K79 methyltransferase, DOT1L, has been reported to bind to MLL- fusion partners such as AF9, ENL and AF10 (Mohan et al., 2010, Okada et al., 2005,

Zhang et al., 2006). DOT1L has also been implicated in the development of MLLAF9 AML, where the inactivation of DOT1L led to the downregulation of MLLAF9 targets without affecting global gene expression (Bernt et al., 2011). Although the transcriptionally active H3K79 methylation mark is enhanced on several MLL-fusion target loci, the epigenetic landscape of MLL LSC has also been shown to have genome- wide low H3K79me2 levels and high H3K4me3 levels (Wong et al., 2015). The regulation of the H3K4me3 methylome by histone demethylase, Jarid1b, was shown to negatively regulate LSC oncogenic potential, where the reversal of these epigenetic states induced LSC differentiation and downregulation of crucial MLL target genes (Wong et al., 2015).

Histone demethylases are the most recently discovered family of erasers, largely due to the widely held belief that histone methylation was permanent. Unlike acetylation and phosphorylation, methylation was suggested to be biochemically stable due to the high thermodynamic stability of N-CH3 bond resulting in an irreversible methyl mark. Furthermore, the half-life of histones and methyl-lysine residues within them were the same (Duerre and Lee, 1974, Byvoet et al., 1972) and histone demethylases had not been identified. Evidence then appeared showing the methylation turnover was slow but active (Annunziato et al., 1995, Borun et al., 1972), which supported the role of histone methylation in regulating gene expression, where gene expression status should be

39 concomitant with a change in its methylation status. To date, the family of histone demethylases comprises twenty-eight members which can be categorised based on their phylogeny (Table 1.7) (Arrowsmith et al., 2012, Mosammaparast and Shi, 2010). Each histone demethylase category contains a number of different domains for DNA-binding, substrate recognition, stabilisation and also uncharacterized domains (Lohse et al., 2011b).

Histone demethylase activity was first reported in 2004, where a flavin-containing amino oxidase (AO) named lysine-specific demethylase 1 (LSD1/KDM1A) specifically demethylates mono- or dimethylated lysine 4 at histone 3 (H3K4me1 and H3K4me2) (Shi et al., 2004). LSD1 has also been reported to demethylate p53 at lysine 370 to repress p53 activity (Lan and Shi, 2009). More recently, a second flavin-dependent H3K4me1/2 demethylase, LSD2/AOF1/KDM1B, has been identified, which forms active complexes with histone methyltransferases G9a and NSD3 (Fang et al., 2010, Huang et al., 2007a). LSDs demethylate substrates via a flavin adenine dinucleotide (FAD)-dependant amine oxidase reaction, which requires a protonated methyl - ammonium group to catalyse oxidation and thus prevents LSDs from demethylating trimethylated lysine residues (Figure 1.10A).

A structurally different family of histone demethylases with a JmjC-domain was discovered in 2006, and hence named JmjC-domain containing histone demethylase (JMJD). They demethylate through a Fe2/2-oxoglurate (2-OG) mechanism for catalysis where the last two steps of the reaction involve the generation of a carbinolamine intermediate on the substrate, which releases formaldehyde. This leaves one less methyl group attached to the amine (Figure 1.10B) (Hou and Yu, 2010).

The aberrant expression or activity of JMJD has been linked to regulating self-renewal, hypoxic stress, and drug resistance, which all play key roles in cancer. JMJD2A, a H3K9me2/me3 demethylase was originally named gene amplified in squamous cell carcinoma 1 (GASC1) due to its amplification in oesophageal cell lines (Yang et al., 2001, Cloos et al., 2006). JMJD2 proteins are overexpressed in prostate cancer (Wissmann et al., 2007), and have been shown to promote genomic instability, suppress cellular senescence and activate transcription (Cloos et al., 2008). JMJD1 and JMJD2 family have been linked to the regulation of self-renewal in embryonic stem cells, where Oct4 controls the expression of JMJD1A and JMJD2C, and knockdown of these genes 40

Figure 1.10: Mechanism of lysine demethylation (A) LSD1 demethylates via an amine oxidation reaction using FAD as a cofactor. The imine intermediate is hydrolysed to an unstable carbinolamine that degrades to release formaldehyde. (B) JMJD uses α-ketoglutarate and Fe as cofactors to hydroxylate the methylated histone substrate. Fe(II) in the active site activates dioxygen to form a highly reactive oxoferryl [Fe(IV) = O] species to react with the methyl group. The resulting carbinolamine intermediate degrades to release formaldehyde. The wavy line indicates attachment to the peptide backbone. Adapted from (Cloos et al., 2008).

41 induced differentiation (Loh et al., 2007). JMJD1A (Beyer et al., 2008, Krieg et al., 2010), JMJD2B (Beyer et al., 2008, Krieg et al., 2010), JMJD3 (Lee et al., 2014) and JARID1B (Krieg et al., 2010) are transcriptional targets of Hif where hypoxic conditions facilitate hypoxic gene expression, demethylation of target genes and enhanced tumour growth. JARID1A has also been shown to elevate drug resistance in prostate carcinoma cells (Sharma et al., 2010b). Interestingly, Jmjd1c, a putative H3K9me2 demethylase, was identified as a MLLAF9 target gene along with HoxA cluster genes, Meis1a and Runx2 (Bernt et al., 2011). Some histone demethylases have been identified as candidate tumour suppressors in leukaemia, such as JMJD5/KDM8 and FBXL10 (Suzuki et al., 2006). JMJD5 has also been shown to contribute to genomic stability by protecting the genome against mutations in C. elegans (Pothof et al., 2003).

1.3.2 Recent advances in epigenetic cancer therapy

Aberrant expression of epigenetic regulators and hence improper maintenance of epigenetic marks can result in the inappropriate activation or inhibition of various signalling pathways and lead to cancer. Due to the reversible nature of epigenetic modifications, drugs can be designed to reverse the aberrant epigenetic state in neoplastic cells. The United States Food and Drug Administration (FDA) has approved a small number of epigenetic drugs for cancer treatment (Table 1.8), including the DNA methyltransferase inhibitors azacytidine (Vidaza) in 2004 and decitabine (Dacogen) in 2006, both of which are for the treatment of myelodysplastic syndrome. Suberoylanilide hydroxamic acid or SAHA (Zolinza) has also been approved by the FDA for the treatment of cutaneous T-cell lymphoma. Promising results have emerged from a phase 3 trial involving multiple myeloma patients, where the histone deacetylase inhibitor, panobinostat (LBH589), was combined with the proteosome inhibitor, bortezomib (Velcade®), receiving a priority review by the FDA (San-Miguel et al., 2014). Due to the relatively recent discovery of epigenetic regulators such as histone demethylases, the focus has shifted in the last few years to develop agents targeting lysine methyltransferases, lysine demethylases, lysine acetyltransferases and bromodomain inhibitors.

Table 1.8: Epigenetic-based therapeutics in preclinical and clinical development HDAC, histone deacetylase; DNMT, DNA methyltransferases; HMT, lysine methyltransferases; LSD, lysine demethylase. Adapted from (Ellis et al., 2009, Mack, 2006, Arrowsmith et al., 2012).

Epigenetic Target Compound Developmental Stage HDAC (Class I, IIa, IIb, Suberoylanilide hydroxamic FDA approved for cutaneous IV) acid (SAHA)/Vorinostat T-cell lymphoma (Zolinza®) Panobinostat (LBH589) Phase II Belinostat (PXD101) Phase II ITF2357 Phase I PCI-24781 Phase I HDAC (Class I, IIa) Phenylbutyrate (Buphenyl) FDA approved for urea cycle disorders Valproic acid (Depakote®) FDA approved for anti- seizure HDAC 1 and 2 Romidepsin (Istodax®) FDA approved for cutaneous T-cell lymphoma SK-7401 Experimental SK-7068 Experimental PCI-24781 Phase I/II Mocetinostat Phase II HDAC 1 and 3 MS-275 Phase I/II CI-994 Phase I MGCD0103 Phase I/II AR-42 Phase I HDAC 2 and 3 Apicidin Experimental HDAC6 Tubacin Experimental ACY-1215 Phase I/II HDAC8 SB-379872A Experimental PCI-34051 Experimental DNMT 5-Azacytidine (Vidaza®) FDA approved for myelodysplastic syndrome Decitabine/5-aza-2‘- FDA approved for

deoxycytidine (Dacogen®) myelodysplastic syndrome Zebularine Experimental RG108 Experimental Procaine Experimental Hydralazine Phase I (-)-Epigallocatechin-3-gallate Experimental noncovalent enzyme (EGCG) SIRT1-7 Nicotinamide Experimental SIRT 1, 2 Tenovin-1-3, 5-6 Experimental Sirtinol Experimental Splitomycin Experimental Cambinol Experimental SIRT 1 SIRT1720 Experimental EX-527 Experimental NF657 Experimental SEN196 Phase II (Huntington‘s disease) SIRT 2 AGK2 Experimental SIRT 5 Sumarin Experimental HMT (G9a) BIX-01294 Experimental HMT (SU(VAR)3-9, Chaetocin Experimental G9a) HMT (DOT1L) EPZ-5676 Phase 1 Polycomb group E7438 Phase 1 proteins DZNep Experimental (Ezh2) LSD1 Polyamine analogues Experimental

With regards to histone methylation, specific inhibitors for well-characterised histone methyltransferases such as EZH2 and DOT1L have been identified (Copeland, 2013). EZH2 inhibitors (GSK126 and EI1) decreased proliferation and caused cell cycle arrest in lymphoma cells (Qi et al., 2012, McCabe et al., 2012), while treatments with inhibitors of DOT1L (EPZ004777) have also been effective in killing leukaemic cells with MLL translocations (Daigle et al., 2011). These inhibitors have entered phase 1 trials in leukaemia and lymphoma patients (Copeland, 2013).

Not only are epigenetic drugs currently being evaluated in combination therapies, but the matter of dosing regimens for currently licensed epigenetic agents has also become a focus for a number of studies. An innovative approach to cancer therapy is to use drugs that target well-known epigenetic pathways already on the market in combination or in different doses to increase their efficacy. The concept is to treat cancer by changing their gene expression profile rather than cause toxicity. New schedules and doses of decitabine and SAHA for MDS and AML are currently in phase I trial following the completion of a 3 year study where decitabine was used as a single agent in the treatment of MDS and CML (Mack, 2006). The study used three different doses and observed that the intravenous low dose showed the highest response rate and was 25% lower than the decitabine label dose. This response rate was higher than the original phase III trial that led to the approval of decitabine. Furthermore, the low dose induced the greatest DNA demethylation suggesting that the effect observed was due to epigenetic changes. This concept of using epigenetic-related doses was also applied in a phase I/II trial where pre-treated patients with recurrent metastatic non-small cell lung cancer (NSCLC) were treated with the DNA methyltransferase inhibitor, azacitidine, and the histone deacetylase inhibitor, entinostat. Previous studies showed that both DNA methyltransferase inhibitors, azacitidine and decitabine, were associated with minimal efficacy and extensive toxicity when maximal tolerated doses were used to treat patients with solid tumours (Glover et al., 1987) but had improved clinical efficacy and long-term tolerance when MDS patients were treated at doses well below maximally tolerated doses (Silverman et al., 2002). The results of the NSCLC trial demonstrated that demethylation of a set of four epigenetically silenced genes known to be associated with lung cancer was associated with improved progression-free and overall survival when the concentration required to reverse tumour-specific DNA methylation was used (Juergens et al., 2011). Furthermore, 25% of patients in the trial 45 responded to subsequent chemotherapies (Juergens et al., 2011). As this concentration was much lower than that shown to produce maximal cytotoxicity, the effects observed in this trial are likely due to the epigenetic drugs altering the epigenetic landscape of NSCLC cells.

Difficulties in developing epigenetic therapeutics arise from the heterogeneity of AML, which affects the epigenetic profile of the disease (Swanton and Beck, 2014, Landau et al., 2014, Easwaran et al., 2014). Another emerging concern involves the cancer cell‘s ability to repair DNA damage. Recent reports have identified a role for the histone demethylase, Jmjd1c, in regulating targets to elicit chromatin responses to repair DNA breaks (Watanabe et al., 2013, Lu and Matunis, 2013). Since many standard chemotherapy regimens involve DNA damaging agents, drug resistance to these chemotherapeutic agents may involve epigenetic regulation of DNA repair pathways. Epigenetic regulators may also influence genomic instability due to the regulation of chromatin remodelling. This highlights the importance of identifying the involvement of epigenetic events in drug resistance and genome stability for effective cancer treatment.

Personalised epigenetic therapy is also an evolving concept in cancer treatment, where chemosensitivity testing will assist in providing more effective treatments to patients. Strategies such as predictive modelling of anticancer drug sensitivity using ex vivo patient samples will become crucial in the future for tailoring cancer treatment for each patient (Majumder et al., 2015). Another potential strategy for treating patients will be to analyse epigenetic profiles of patients to identify aberrant epigenetic regulation based on epigenetic biomarkers for diagnosis, prognosis and prediction of drug response.

The field of epigenetics has grown enormously, leading to several large consortium- based epigenome mapping programs (National Institutes of Health (NIH) Roadmap Epigenomics Program or the International Human Epigenetics Consortium) (Roadmap Epigenomics Consortium et al., 2015). With the assistance of these consortiums, epigenetic regulation of human diseases will assist in identifying aberrant epigenetic marks underpinning cancer and therefore support the development of novel cancer therapeutics.

In summary, investigations into epigenetic regulators as potential drug targets have increased the number of epigenetic drugs being developed and tested experimentally, as

46 well as in clinical trials either as single agents or in combination with other chemotherapeutic agents. Studies are also being conducted on FDA approved epigenetic drugs at a dose capable of altering the epigenetic phenotype of specific cancers in order to elicit an epigenetic change rather than induce toxicity. This concept of focusing on low doses and the level of epigenetic change should be factors to be considered in epigenetic drug development. With the clinical success of histone deacetylase and DNA methyltransferase inhibitors and emerging data on epigenetic regulators that govern the balance of histone modifications and gene regulation in cancer, epigenetic regulators hold considerable promise as drug targets for AML treatment.

1.4 Summary and thesis perspectives

Recent findings have yielded important insights into the dynamic nature of epigenetic regulation, presenting it as a means to manipulate the expression of downstream targets critical for AML development and maintenance of LSC. Current AML therapies target the tumour bulk, namely leukaemic blasts, however, LSC are highly resistant and persist after chemotherapy. The main cause of patient relapse has therefore been attributed to AML LSC, emphasising the need to develop LSC-targeted therapies. Signalling pathways and epigenetic regulation have been shown to be perturbed in LSC, hence the need for elucidating key players in regulating LSC and understanding the biology of these cells in order to design agents to target these cells.

This thesis is directed towards investigating aberrantly expressed histone demethylases in MLLAF9 AML and elucidating the mechanism(s) by which they regulate AML LSC. Since leukaemia is a heterogeneous disease, previous AML patient gene expression studies have involved comparing differentially regulated genes between subtypes of leukaemia, however, none of these leukaemic groups can be considered a control group. Analysis of global gene expression changes in murine HSC with MLLAF9 and Hoxa9/Meis1a (A9M) LSC identified Jmjd1c and Jmjd5 to be the most highly upregulated and downregulated histone demethylases in MLLAF9 LSC, respectively.

The mechanisms of leukaemic transformation involved proof-of-principle studies and mouse models to demonstrate the important role epigenetic regulators such as histone demethylases play in regulating LSC. Few studies have been conducted on the function 47 and regulation of Jmjd1c in AML, allowing this research to provide a novel insight into the mechanism of this epigenetic regulator. Ectopically upregulating Jmjd1c was found to enhance the leukaemogenicity of A9M AML through the regulation of metabolic pathways such as glycolysis. Recent reports suggest that the level of glycolysis may be therapeutically explored for treating leukaemia while preserving HSC function. MLLAF9 AML cells were impaired upon Jmjd1c suppression, suggesting a requirement of Jmjd1c for MLLAF9 leukaemia maintenance. At present, there is a lack of Jmjd1c inhibitors providing a challenge in studying the pharmacological inhibition of this epigenetic regulator. Inhibitors of key glycolytic regulators have been previously identified and were used to investigate the ability to target AML cells with glycolysis inhibitors. Jmjd1c has emerged as a critical regulator in AML LSC and represents a novel therapeutic target.

This study provides the first evidence demonstrating a tumour suppressor role of Jmjd5 in AML, as exemplified by the upregulated expression of Jmjd5 in HSC and DOT1L inhibitor-treated KLSMLLAF9, induction of apoptosis and differentiation, increased sensitivity to doxorubicin, reduction in leukaemic burden and delayed leukaemia onset. Gene expression profiling also revealed a number of Jmjd5 targets that are co-regulated by G protein-coupled receptor 84 (Gpr84). Gpr84 is known to regulate β-catenin expression, which in turn is essential for AML LSC. This receptor has also been shown to regulate a small set of known MLL-fusion target genes and Wnt-associated genes. The disruption of the Wnt/β-catenin pathway is crucial in retarding the progression of established leukaemia induced by MLLAF9.

Histone methylation changes were also analysed, where Jmjd1c did not alter global H3K9 methylation levels, however, Jmjd5 overexpression resulted in marked suppression in H3K36me2 methylation. Lack of global methylation changes suggests that Jmjd1c acts on a small subset of genes. On the other hand, Jmjd5 overexpression reduced H3K36 methylation levels, in addition to global H3K27me3 levels. These methylated lysine residues have previously been shown to cooperate in the same regulatory axis, where Ezh2 and NSD2 cooperated in methylating H3K27me3 and H3K36me2. A common Ezh2 inhibitor was also characterised in the MLLAF9 and A9M AML model system where doses required to elicit epigenetic changes were used for drug treatments in both early stage and fully established AML LSC. Treatments at

48 these low doses were found to alter different target genes compared to those identified in previous reports where doses were used to cause cytotoxicity. The concept of epigenetic therapies capable of restoring aberrant epigenetic marks in cancer supports the data obtained in this study.

Since LSC are highly resistant to current chemotherapeutic regimens and can cause disease relapse and metastasis, eradication of these cells is necessary to achieve long- lasting remission, thus LSC-targeted therapies must be developed. The data presented in this thesis provide significant insights into the molecular mechanism of two novel histone demethylases in AML development, which are clinically relevant as biomarkers of disease prognosis and drug response. Elucidating the mechanism of histone demethylases in AML LSC provides great insights into the role of histone demethylases in leukaemogenesis and presents the possibility of developing new diagnostic tools as well as therapeutic targets in AML. Compounds regulating histone demethylases should be of considerable interest as novel anticancer agents with the potential to lead to improved treatment and prognosis for AML patients.

CHAPTER 2 Materials and Methods

2.1 Reagents

General laboratory reagents for cell biology and molecular biology experiments were obtained from BD Biosciences, Sigma, Merck, Invitrogen, Millipore, Australian Bioresearch, Bio-Rad, Thermo Fisher Scientific, Therma Trace, Amresco and Stem Cell Technologies. All reagents and solvents for molecular biology were of analytical grade.

2.2 Cell biology techniques

2.2.1 Murine leukaemic cells

Bone marrow cells were harvested from 6- to 8-week-old C57BL/6 mice (Australian BioResources, Moss Vale, NSW, Australia), subsequently stained and sorted for c-Kit+ Sca-1+ Lin- (CD3- CD4- CD8a- CD19- B220- Gr-1- Ter119- IL-7R-, BioLegend) for haematopoietic stem cell (HSC)-enriched KLS cells, while granulocyte-macrophage progenitors (GMP) were isolated by sorting for c-Kithigh Sca-1- Lin- IL-7R- CD34+ FcγRII/IIIhigh (BioLegend) using a BD Influx high-speed cell sorter (BD Biosciences). The panel of pre-LSC used in this study include KLSA9M, GMPMLLAF9, and KLSMLLAF9 (Table 2.1). These cells were generated by retroviral transduction of HSC and GMP cells with concentrated virus encoding for MSCV-Hoxa9-GFP/MSCV- Meis1a-Puromycin or MSCV-MLLAF9-GFP, as described previously (Krivtsov et al., 2006, Wang et al., 2010).

Pre-LSC were then transplanted into C57BL/6 syngeneic sublethally irradiated (600 rad) recipients at 1 × 106 cells per mouse for leukaemia development. GFP LSC (also known as L-GMPs; c-Kit+ Sca-1- Lin- IL-7R- CD34+ FcγRII/III+) (Wang et al., 2010) were sorted from the bone marrow of mice with primary AML. GMPMLLAF9 LSC and KLSMLLAF9 LSC were cell types used in this study. Animal experiments were

approved by the Animal Care and Ethics Committee of the University of New South Wales, Australia.

Table 2.1: Summary of cell types Cell Type Description Exogenous Gene Growth Media HSC c-Kit Sca-1 Lin (CD3 M3234 CD4 CD8a CD19 supplemented B220 Gr-1 Ter119 IL- with IL-3, IL-6 7R) mouse bone marrow and SCF cells LSC c-Kit Sca-1 Lin IL-7R M3234 CD34+ FcγRII/III+ mouse supplemented bone marrow cells with IL-3 KLSA9M HSC mouse bone marrow Hoxa9, Meis1a M3234 cells supplemented with IL-3 GMPMLLAF9 GMP mouse bone marrow MLLAF9 M3234 cells supplemented with IL-3 KLSMLLAF9 HSC mouse bone marrow MLLAF9 M3234 cells supplemented with IL-3 HEK293T Human embryonal kidney SV40 large T 10% FCS/DMEM cells antigen

2.2.2 Cell culture

Murine pre-LSC/LSC were grown and maintained in methylcellulose (MethoCult M3234, Stem Cell Technologies) enriched with 19% Iscove‘s Modified Dulbecco‘s Medium (IMDM, Invitrogen) and 1% of penicillin/streptomycin (PSG, 10 mg/mL penicillin, 10 mg/mL streptomycin, 29.2 mg/mL L-glutamine, Invitrogen) supplemented with recombinant mouse interleukin-3 (IL-3, 50 ng/mL, Australian Biosearch) in 35 mm culture dishes (Stem Cell Technologies). Cells were typically replated every 5 days and involved harvesting the cells by dissolving the methylcellulose in phosphate buffered saline (PBS, Sigma) and transferring it into 15 mL conical tubes. The tubes were centrifuged at 1500 rpm for 5 min, supernatant was removed and the cell pellet was resuspended in PBS and centrifuged again to wash the cells. The cells were resuspended in PBS and counted using the trypan blue exclusion assay (0.4% Trypan blue, Invitrogen) and a haemaocytometer, where trypan blue negative cells were considered viable. The cells were then replated in fresh methylcellulose supplemented with IL-3 at 5000 cells/dish. HSC were grown in a similar manner, but the methylcellulose was supplemented with IL-3 (50 ng/mL), IL-6 (50 ng/mL, Australian Biosearch) and SCF (100 ng/mL, Australian Biosearch).

Human embryonic kidney (HEK) cells previously transfected with 293 expressing simian virus T-antigen (293T) were grown in Dulbecco‘s Modified Eagle Medium (DMEM, Invitrogen) supplemented with 10% fetal calf serum (FCS, Therma Trace) and 1% PSG. The cells were grown in T75 culture flasks and were subcultured, typically once a week, before reaching confluence. Culturing HEK293T cells involved aspirating the culture medium from the flask and adding 3 mL of trypsin ethylenediaminetetraacetic acid solution (0.25% trypsin, 0.02% EDTA v/v in PBS, pH 7.4, Invitrogen) for 2 min, followed by the addition of fresh media to inactivate the trypsin. The cells were collected into 15 mL conical tubes, centrifuged for 5 min at 1500 rpm. The supernatant was removed and the pellet was resuspended in fresh tissue culture media, where a fraction of the cells (1:10) were transferred to a new T75 flask.

All cells were maintained in a 37C/5% CO2 incubator.

2.2.3 Cryopreservation

Cells were washed in PBS and centrifuged at 1500 rpm for 5 min. The media was aspirated and the pellet was resuspended in cryomedia containing 90% FCS and 10% dimethyl sulphoxide (DMSO, Sigma) with a concentration of 1  106 cells/mL. This was aliquoted into cryovials and frozen at -80C for 24 h before being transferred to liquid nitrogen for long-term storage.

2.2.4 Thawing cell lines

A vial of 1 × 106 cells from -80°C or liquid nitrogen storage was thawed in a 37°C water bath and transferred to a centrifuge tube. Fresh tissue culture media (80% DMEM, 20% FCS and PSG for HEK293T; IMDM for pre-LSC/LSC) was added to the centrifuge tube and the cells were centrifuged at 1500 rpm for 5 min. The supernatant was removed and cells were washed again in fresh medium. The cells were then plated into 35 mm dishes or T75 flasks.

2.2.5 Colony forming assay

This in vitro assay is based on the ability of a single cell to divide and form colonies, and is commonly used for research into haematopoietic cells (Pereira et al., 2007, Franken et al., 2006). Cells were harvested and seeded in a 35 mm dish at 1 × 103 cells/dish containing 1 mL methylcellulose supplemented with 50 ng/ml IL-3. Plates were incubated (37°C/5% CO2) for 5 days. The colonies were counted directly over the entire plate by inverted light microscopy. Cells were subsequently harvested and the cell number was counted using trypan blue assay.

2.2.6 Alamar blue cytotoxicity assay

Cells were harvested, washed in PBS and seeded into a 96-well plate at 1 × 103 per well containing 100 µL IMDM supplemented with 10% FCS and 50 ng/mL IL-3. The cytotoxic drug was diluted with IMDM-supplemented media to make up the desired

53 concentrations and 100 µL of the drug was added to wells containing cells. Blank controls contained 200 µL media without cells and untreated cells were included as a control. The plate was incubated at 37C/5% CO2 for 24 h. Alamar blue (20 µL,

Invitrogen) was added to each well and incubated for 4 h at 37C/5% CO2. The absorbance was then read on a Benchmark Plus microplate spectrophotometer (Bio- Rad) at 570 nm with a 575 nm reference wavelength (Bio-Rad). The principle of this assay is based on resazurin, a non-toxic, active ingredient of alamar blue, which permeates through the cell membrane where live cells reduce this compound in its nonfluorescent oxidised state to resorufin, its fluorescent reduced state, causing a change in the colour of the culture media (Shiloh et al., 1997, Larson et al., 1997, O'Brien et al., 2000). From this cell viability assay, the half maximal inhibitory concentration (IC50) can be determined, which is the concentration of a drug to inhibit the viability of half of the cells.

2.2.7 Wright-Giemsa staining

Cells were harvested and washed twice in PBS. The cells were resuspended in PBS and loaded into the cytofunnel of a spin column (5 × 104 cells, 150 µL) with a microscope slide attached. The column was centrifuged for 10 min at 1000 rpm, the microscope slide was detached and then air dried. The cells were fixed in ice-cold pure methanol for 5 min, and then allowed to dry completely. Cytospin slides were submerged in filtered Wright stain (Sigma Aldrich; 3 g/L Wright stain in methanol) for 30 minutes, followed by 1:10 Wright stain diluted in Phosphate buffer (Sigma Aldrich) for 1 h. The slides were then stained with Giemsa stain (Sigma Aldrich; 1 g Giemsa stain in 66 mL glycerol and 66 mL methanol) diluted 1:10 in phosphate buffer for 30 min. The slides were washed with phosphate buffer by submerging the slide in the buffer solution for 5 min and repeated three times. The slides were air dried then covered with by a coverslip using Eukitt quick-hardening mounting medium (Sigma-Aldrich). The stained samples were left to dry overnight. The slides were visualised using a light microscope.

2.3 Molecular biology techniques

2.3.1 Bacterial transformation

Several expression plasmids were used in this study to examine the effect of suppressing or overexpressing the gene of interest. The plasmids used for stable transfections are summarised in Table 2.2. Plasmids to be amplified (2 µL) were added to DH5α bacteria (500 µL, Invitrogen) and incubated on ice for 30 min, then 42C for 40 s, and then left on ice for 2 min. Pre-warmed Luria broth (LB, 500 µL, Sigma) was added to each sample and incubated at 37C for 1 h with agitation. The bacterial samples were poured onto LB agar plates (Sigma) containing 100 µg/mL ampicillin (Sigma) and incubated overnight at 37C.

Table 2.2: List of plasmids Plasmid Reporter Gene Supplier Scrambled shRNA mCherry/puromycin GeneCopoeia Jmjd1c shRNA 1 mCherry/puromycin GeneCopoeia Jmjd1c shRNA 2 mCherry/puromycin GeneCopoeia Jmjd1c shRNA 3 mCherry/puromycin GeneCopoeia Jmjd1c shRNA 4 mCherry/puromycin GeneCopoeia Neo empty vector Neomycin GenScript Jmjd1c-S cDNA Neomycin GenScript Jmjd1c-L cDNA Neomycin GenScript LeGO-iT2 dTomato Addgene Jmjd5 cDNA dTomato Open Biosystems Gpr84 shRNA mCherry/puromycin GeneCopoeia Gpr84 cDNA Neomycin Genscript β-catenin-Flag cDNA Neomycin ClonTech Ezh2 shRNA 1 Puromycin Sigma Ezh2 shRNA 2 Puromycin Sigma Ezh2 shRNA 3 Puromycin Sigma Ezh2 shRNA 4 Puromycin Sigma

2.3.2 Plasmid preparation

Bacterial colonies from LB agar plate containing the plasmid of interest were picked into microcentrifuge tubes containing PBS. The bacterial samples were pelleted and the plasmids were extracted and purified using the QIAprep Spin Miniprep Kit (Qiagen). Buffer P1 (250 µL) was used to resuspend the pellet of bacterial cells and transferred into a microcentrifuge tube. Buffer P2 (250 µL) was added and the contents were mixed gently by inverting the tube. Buffer N3 (350 µL) was added, mixed by inversion, then centrifuged for at 13000 rpm for 10 min. The supernatant was transferred into a QIAprep spin column and centrifuged for 60 s. The column was washed by adding 0.5 mL Buffer PB and then centrifuged for 60 s. This washing step was repeated using 0.75 mL Buffer PE. The column was placed in a clean 1.5 mL microcentrifuge tube and 50 µL of water was added to elute the DNA. The concentration of DNA was quantified using the NanoDrop (Thermo Fisher Scientific) and stored at -20C.

2.3.3 shRNA/cDNA transfection

Retroviral and lentiviral stocks were produced by seeding HEK293T cells (5 × 105 cells/well with 10% FCS/DMEM) in a 6-well plate for 24 h at 37C/5% CO2. The media was then replaced with 200 µL DMEM/well. Lipofectamine 2000 (6 µL, Invitrogen) was diluted in Opti-MEM I reduced serum media (244 µL, Invitrogen). In a separate microcentrifuge tube, plasmid DNA (1 µg) was added to either psPAX2 packaging plasmid (0.75 µg) and psMD2.G envelope plasmid (0.25 µg) for lentivirus, or psi-eco packaging plasmid (1 µg) for retrovirus. The final volume of 250 µL was reached by adding Opti-MEM. The lipofectamine and plasmid solutions were combined, incubated for 30 min at room temperature, and then added dropwise to each well containing HEK293T cells. The virus-containing media was replaced with 3 mL of fresh media (10% FCS/DMEM) following an overnight incubation at 37C/5% CO2.

After another incubation for 24 h at 37C/5% CO2, the media was harvested, filtered through a 0.45 µm low-protein binding filter (Millex-HP syringe filter, Millipore) and concentrated by centrifugation at 3000 g for 24 h at 4C. The virus-containing media was aspirated and resuspended in 500 µL 15% FCS/IMDM and stored at -80C. Fresh media (10% FCS/DMEM) was added to each well of the 6-well plate and incubated for

56 another 24 h. The media was harvested, filtered and concentrated as before. Transfection efficiency was determined by harvesting the cells for flow cytometry to measure the expression of the fluorescent reporter gene.

2.3.4 Viral transduction

Cells (5 × 104) were seeded into a 96-well plate containing 250 µL concentrated virus supplemented with 0.7 µg/mL hexadimethrine bromide (polybrene, Sigma) and 50 ng/mL IL-3 for pre-LSC/LSC or 50 ng/mL IL-3, 50 ng/mL IL-6 and 100 ng/mL SCF for HSC. The plate was centrifuged for 2 h at 1500 rpm at 30C. Following a 4 h incubation at 37C/5% CO2, the plate was centrifuged for 5 min and the media was replaced with fresh virus supplemented with polybrene and cytokine(s). This plate was centrifuged for 1 h at 1500 rpm at 30C then incubated overnight at 37C/5% CO2. The media was then aspirated, washed twice with IMDM and then seeded into 35 mm dish with M3234 supplemented with cytokine(s). Two to 4 days after transduction, the infected cells were selected by fluorescence activated sorting (GFP, dTomato or mCherry cells) or by antibiotic selection (1 µg/mL puromycin or 1 mg/mL G418) depending on the vector.

2.4 RNA analysis

2.4.1 RNA extraction

RNA was extracted and purified using the Rneasy mini kit (Qiagen) according to the manufacturer‘s protocol. Cells were harvested, washed with PBS and pelleted. Buffer RLT (350 µL) with β-mercaptoethanol (10 µL/mL) was added to lyse the cells. Ethanol (70%, 350 µL) was added to the cells, mixed thoroughly and transferred to an Rneasy mini spin column. After centrifugation ( 10000 rpm, 15 s), Buffer RW1 (700 µL) was added to the column then centrifuged again ( 10000 rpm, 15 s). Buffer RPE (500 µL) was added twice to the column, first for 15 s to wash, then for 2 min to dry the membrane. The column was transferred to a new microcentrifuge tube and the RNA

was eluted with 30 µL Rnase-free water (Invitrogen) and spinning for 1 min at  10000 rpm. The concentration of RNA was measured using a NanoDrop and stored at -80C.

2.4.2 cDNA synthesis

To construct cDNA, 1 µg total cellular RNA was treated with Dnase 1 (0.5 µL, Invitrogen) with 1 × First Strand Buffer (FSB, Invitrogen) made up to 10 µL with nuclease-free water (Invitrogen), heated at 37C for 30 min, followed by 5 min at 75C to inactivate Dnase. A mastermix (10 µL/sample; 2 μL FSB, 1 μL deoxynucleoside triphosphate (dNTP, Invitrogen), 1 μL 0.1M dithiothreitol (DTT, Invitrogen), 1 μL random primers (Invitrogen), 1 μL Moloney Murine Leukaemia Virus Reverse Transcriptase (M-MLV RT, Invitrogen), 0.5 μL RNAsin (Invitrogen) and 3.5 μL nuclease-free water), was added to each sample and heated for 1 hr at 37C. Nuclease- free water (30 µL) was then added to the cDNA samples and stored at -20C.

2.4.3 Real-time quantitative PCR protocol

Real-time PCR mastermix (22.5 µL) containing 12.5 μL SYBR green (Life Technologies), 1 μL forward primer (5 μM), 1 μL reverse primer (5 μM) and 8 μL nuclease-free water, was added to each well of a MicroAmp Optical 96-well reaction plate (Applied Biosystems) for each primer pair. In addition to the wells containing samples (2.5 µL cDNA), a no template control containing 22.5 µL master mix and 2.5 µL nuclease-free water was included. The PCR plate was sealed with MicroAmp Optical Adhesive Film (Applied Biosystems) and centrifuged for 1 min at 1500 rpm. Primers used in this study are described in Table 2.3. Cycling was performed on an ABI7900HT Fast Real-Time PCR system (Applied Biosystems) with the following profile:

Activation: 50C for 2 min, 95C for 10 minutes Cycle: 95C for 15 s, 60C for 60 s, 40 cycles Dissociation: 95C for 15 min, 60C for 15 min, 95C for 15 min Melt: Ramp from 65C to 95C, 1C/5 s

Data were analysed using SDSv 2.4 software (Applied Biosystems), where differences in threshold cycles CT between target genes and housekeeping gene (59lyceraldehydes-3-phosphate dehydrogenase, GAPDH) were calculated.

Table 2.3: List of primer sequences Gene Sequence Aldh3b1 Forward: 5‘- CCA AGA ATC TGG CTA CAC AGC TGG -3‘ Reverse: 5‘- CCA CAC AAT TCC CTG CAG CTA TGG -3‘ AldoC Forward: 5‘- GGC ATT CTC GTA GGC ATC AAG GT -3‘ Reverse: 5‘- CGG CGT GCG ATC ACT GAT T -3‘ Alox5 Forward: 5‘- GGA TGG ACG TGC AAA ATT GGC C-3‘ Reverse: 5‘- GGG TTC CAC TCC ATC CAT CGA TAC -3‘ Bhlhb2 Forward: 5‘- AAC GGA GCG AAG ACA GCA AG -3‘ Reverse: 5‘- GCT TTT TCC AAG TGA CCC AAA GTA GT -3‘ Cbr3 Forward: 5‘- AAA TTC TCC GGG GAC GTG GTG -3‘ Reverse: 5‘- ATG CTC TGCGGG TCG TCG -3‘ Ccnd1 Forward: 5‘- CAA CAA CTT CCT CTC CTG CTA CC -3‘ Reverse: 5‘- AGA CCA GCC TCT TCC TCC AC -3‘ Ccnd2 Forward: 5‘- CTG GAT GCT AGA GGT CTG TGA GGA AC -3‘ Reverse: 5‘- ATG CAC ACT GCA CCC AGG A -3‘ Ccne1 Forward: 5‘-GGT CTG AGT TCC AAG CCC AAG-3‘ Reverse: 5‘-AGC GGA CTG AAA GGT CGG AG-3‘ CD177 Forward: 5‘- TAC GAT GGA GTC CTC AGG CTC -3‘ Reverse: 5‘- GTT TCC AAG CTC CCA CTA TTG CAA -3‘ Chi3l1 Forward: 5‘- ACT GAA TGC GGA ATT CAC AAA GGA GG -3‘ Reverse: 5‘- GAG ATT GAT AAA ATC CAG GTG TTG GGC TAT C -3‘ Cux2 Forward: 5‘- GGC AGC GAG GCA GCA T -3‘ Reverse: 5‘- CTT GGA CTC ACC CCC CTT CTG -3‘ Fbxo32 Forward: 5‘-CCA TTC TAC ACT GGC AGC AGC A-3‘ Reverse: 5‘-AGT TGT AAG CAC ACA GGC AGG TC-3‘ GAPDH Forward: 5‘- CAT GGC CTT CCG TGT TCC TA -3‘

Reward: 5‘- CCT GCT TCA CC ACCT TCT TGA T -3‘ Hif1a Forward: 5‘- GAG AAA TGC TTA CAC ACA GAA ATG GCC -3‘ Reverse: 5‘- CCG TGC AGT GAA GCA CCT TC -3‘ Hoxa9 Forward: 5‘-ACA AT GCC GAG AAT GAG AGC G-3‘ Reverse: 5‘-CTT CTT CCG AGT GGA GCG AG-3‘ Irx3 Forward: 5‘- GGC ACA CAG ACT CGT GTC G -3‘ Reverse: 5‘- AGA GCA GCG TCC AGA TGG TTC -3‘ Jmjd1c Forward: 5‘- CTT CTG TGT CCC TGC CTG AGT CT -3‘ Reverse: 5‘- GAT GCT GGA GGT GAT GTG CTG C -3‘ Jmjd5 Forward: 5‘- AGC ATT TTC TAG TTC CTG GGA G -3‘ Reward: 5‘- TTC ATC TGT GTA CCT TGA GCC -3‘ Ldha Forward: 5-‗ GTG CCA TCA GTA TCT TAA TGA AGG ACT TG -3‘ Reverse: 5‘- GAG TTC GCA GTT ACA CAG TAG TCT TTG -3‘ Lgals3 Forward: 5‘- CCG CTG GAC CAC TGA CG -3‘ Reverse: 5‘- TCC CTC TCC TGA AAT CTA GAA CAA TCC T -3‘ Mgmt Forward: 5‘- CAC CCT GTG TTC CAG CAA GAT TC -3‘ Reverse: 5‘- ATG AGG ATG GGG ACC GGA -3‘ Mmp9 Forward: 5‘- TCC AAC CTC ACG GAC ACC CA -3‘ Reverse: 5‘- GGT ACA AGT ATG CCT CTG CCA GC -3‘ Mpo Forward: 5‘- GCA ACA ACA GAC GAA GCC CC -3‘ Reverse: 5‘- GGG GCA CCT TGA AGC CAT -3‘ Ms4a3 Forward: 5‘- CCT GAG GAG GAA GAG ACT GCT G -3‘ Reverse: 5‘- CGT GAC TCA TCC AAG GGC TG -3‘ p16 Forward: 5‘-AAC TCT TTC GGT CGT ACC CCG ATT CA-3‘ Reverse: 5‘-CGA ATC TGC ACC GTA GTT GAG CA-3 p21 Forward: 5‘-TTG CAC TCT GGT GTC TGA GC-3‘ Reverse: 5‘-TGC GCT TGG AGT GAT AGA AA-3‘ Pdk1 Forward: 5‘- CGT ACA GCT GGT GCA AAG TTG GTA T -3‘ Reverse: 5‘- CGG TTT CTG ATC CTT ATC ACT GTG TCT GTG -3‘ Pgk1 Forward: 5‘- CTC AAA TCT CTG CTG GGC AAG GAT -3‘ Reverse: 5‘- CAG TGA GGC TCG GAA AGC ATC -3‘ Pkm2 Forward: 5‘- TGG AAC CCA TGA GTA CCA TGC AGA G -3‘

Reverse: 5‘- TGC CGC TGC CCT TGA TGA -3‘ Pitx2 Forward: 5‘- CCT CAC CCT TCT GTC ACC ATC C -3‘ Reverse: 5‘- CCC ACA TCC TCA TTC TTT CCT TGC -3‘ Plaur Forward: 5‘- TGC TGC TGC TGC TGT TGC -3‘ Reverse: 5‘- GCA CAC TCC TCT ACC AGG CA -3‘ Rhou Forward: 5‘- CCC ACC GAG TAC ATC CCT ACG -3‘ Reverse: 5‘- TCC CTG AGG TCC GAC TGT GTC -3‘ S100a8 Forward: 5‘- ATG CCG TCT GAA CTG GAG AAG G-3‘ Reverse: 5‘- CGA TAT TTA TAT TCT GCA CAA ACT GAG G -3‘ S100a9 Forward: 5‘- GCA GCA TAA CCA CCA TCA TCG ACA -3‘ Reverse: 5‘- CAT TTC TCT TCT CTT TCT TCA TAA AGG TTC -3‘ Slc2a1 Forward: 5‘- TGG CAT CCT TAT TGC CCA GGT G -3‘ Reverse: 5‘- GGC AAC AGG ATA CAC TGT AGC AGG -3‘ Slc6a3 Forward: 5‘- GCT ATG CTC TAT GGC ACA GGA C -3‘ Reverse: 5‘- GCC AGA CCC AAG CCA GT -3‘ Stra6 Forward: 5‘- CCA CTG CCT TTC CTG AAC CTC -3‘ Reverse: 5‘- GCC CTC TTG ATG TCT CCA TCT C -3‘ Tcf7l2 Forward: 5‘- TTC CTC CGA TCA CAG ACC TGA G -3‘ Reverse: 5‘- GCT GCC TTC ACC TTG TAT GTA GC -3‘ Txnip Forward: 5‘-ACA TCC CAG ATA CCC CAG AAG C-3‘ Reverse: 5‘-CGT CCA CAT CGT CCA GCA GAG-3‘ β-catenin Forward: 5‘- GCA GGA TAC ACG GTG CCG-3‘ Promoter Reverse: 5‘- GGC TTC ACA GGA CAC GAG CT-3‘ β-catenin Forward: 5‘- CTG ACC TGA TGG AGT TGG ACA TGG-3‘ exon 3 Reverse: 5- AGG GTT GCC CTT GCC ACT-3‘ (14-156) β-catenin Forward: 5‘- GGA CTC TAG TGC AGC TTC TGG GTT C-3‘ exon 9 Reverse: 5‘- TGC GTA CAA GAG CCT CTA TGC CA-3‘ (1208- 1348) Gpr84 Forward: 5‘- AGT TCC ATG GCA GAC TGG GAT AAT GAA TC-3‘ Promoter Reverse: 5‘- CCA AAT GTT ATT TCC TTC CTT TCT CTG AGG C-3‘

Gpr84 Forward: 5‘- ATG TGG AAC AGC TCA GAT GCC A-3‘ exon 2.1 Reverse: 5‘- GAA TGG CCA AGG CCA GCA GA-3‘ (1-142) Gpr84 Forward: 5‘- GAC CAA TAC GGG CTG CAT CAG G-3‘ exon 2.2 Reverse: 5‘- GCA GCA CTG ACT GGC TCA GAT GA-3‘ (637-785) Tcf7l2 Forward: 5‘- CGC CTT TGA ACT GAA AAG CTC AGT CTA AC-3‘ Promoter Reverse: 5‘- CAA CAA CAA CAA CAA GAA GAA GTC AGA TAT AAG AGC-3‘ Tcf7l2 Forward: 5- TGC AAC ACC CCC ACC ATG T-3‘ exon 5 Reverse: 5‘- CGT CAG CTG GTA AGT GCG GAG-3‘ (503-601) Tcf7l2 Forward: 5‘- CCA TCA CAC TCT GCA CAC GAC-3‘ exon 8 Reverse: 5‘- AGC TGT GGA GTG AGC CGA-3‘ (831-931)

2.4.4 Microarray expression profiling

RNA was extracted and evaluated using a NanoDrop spectrophotometer. RNA samples with 260/280 and 260/230 ratios of 2.00 or greater and RNA integrity number (RIN) exceeding 9.00 were selected for microarray analysis. Three samples from each cohort were submitted to the Ramaciotti Centre (UNSW) for gene expression analysis carried out on the Illumina platform (Illumina Mouse WG-6 v2.0 Expression BeadChip, UNSW). Microarray data were analysed by Dr Bing Liu (Children‘s Cancer Institute). Gene set enrichment analysis was also performed by Dr Bing Liu, where the statistical criteria included normalised enrichment scores based on a false discovery rate (FDR) ≤ 0.25, normalised enrichment score (NES) and nominal P value ≤ 0.01.

2.5 DNA Analysis

2.5.1 Chromatin immunoprecipitation sample preparation

Cells (1 × 106 cells) were washed in PBS and pelleted by centrifugation. Formaldehyde (2 mL, 1% final concentration, diluted with PBS) was used to resuspend the pellet and samples were incubated for 10 min at room temperature. Glycine (200 µL, 1.25 M) was added, mixed by inversion and incubated for 5 min at room temperature. Cells were centrifuged at 1000 g for 5 min at 4C and then washed with ice cold PBS (3 mL). Cell lysis buffer (500 µL; 10 mM Tric/HCL pH 7.2, 2 mM MgCl2, 0.5% Triton X100) containing 2.5 µL protease inhibitor cocktail was added to resuspend the pellet before a 15 min incubation on ice. Cell suspensions were vortexed every 5 min for 10 s during incubation. The nuclei were pelleted by centrifuging at 5000 g for 5 min at 4C, the supernatant was removed and the fixed nuclei was snap frozen using dry ice, then stored at -80C. Nuclei pellets were transported on dry ice to collaborators for chromatin immunoprecipitation (ChIP) assays.

2.6 Protein analysis

2.6.1 Protein extraction

RIPA buffer (100 mL) was prepared with 3 mL 5 M NaCl, 1 mL Nonidet P40, 5 mL 10% Na deoxycholate, 1 mL 10% SDS, 5 mL 1 M Tris pH 7.5, 85 mL water and used for whole cell protein extraction. Cells were prepared by harvesting and washing the cells with PBS. The cells were then counted and pelleted. For each 2 × 106 cells, 30 µL of RIPA buffer was added, supplemented with 1 × proteinase inhibitor cocktail, 1 µL phenylmethanesulfonyl fluoride (1 mM PMSF, Sigma) and 1 µL sodium orthovanadate (1 mM, Sigma). Samples were then incubated on ice for 30 min then centrifuged at 4000 rpm for 15 min at 4C. The supernatant was transferred into new microcentrifuge tubes, aliquoted and stored at -80C.

Histone extracts were prepared using the Episeeker Histone Extraction kit (Abcam) according to the manufacturer‘s protocol. For cell pellets containing 1 × 106 cells, 1 × pre-lysis buffer (100 µL) was added and incubated on ice for 10 min. Samples were

63 then centrifuged at 10000 rpm for 1 min at 4C. Pellets were resuspended with 20 µL lysis buffer and incubated on ice for 30 min before being centrifuged at 12000 rpm for 5 min at 4C. The supernatant was transferred to a new microcentrifuge and 0.3 mL of Balance-DTT buffer (1 µL DTT/500 µL balance buffer) was added per 1mL of supernatant. The lysates were aliquoted and stored at -80C.

Protein lysates were quantified using the Pierce Bicinchoninic acid (BCA) protein assay (Thermo Fisher Scientific). A protein standard using bovine serum albumin (BSA) was included in the assay. Lysates (2.5 µL) was added to each well of a 96-well plate containing RIPA buffer (22.5 µL), where duplicate wells were used for each sample and standard. Bicinchonic acid (BCA) solution was added to each well, and the plate was incubated at 37C for 30 min in the dark. Absorbance was measured at 570 nm using a Benchmark Plus microplate spectrophotometer.

2.6.2 Western blotting

2.6.2.1 Protein separation by SDS-PAGE

Protein samples (40 µg) were prepared for electrophoresis by adding 4 × sample buffer, 10 × reducing agent and made up with milliQ water to 15 µL. Samples were then heated for 5 min at 95C, then loaded on a precast 4 – 20% Mini-Protean TGX gel (Bio-Rad). Precision Plus Protein Dual Colour standard (Bio-Rad) was also loaded onto the polyacrylamide gel. Gels were electrophoresed in 1 × Running Buffer (2.9 g Tris, 14.4 g glycine dissolved in 1 L milliQ water) for 80 min at 120 V.

2.6.2.2 Protein transfer

Polyvinylidene difluoride (PVDF) Immobilon-P membrane (Millipore) was prepared by soaking in methanol for 15 s, then in water for 2 min and in transfer buffer for at least 5 min. The gel was removed from the SDS-PAGE apparatus, assembled into the western blotting apparatus, and protein transferred onto a PVDF membrane in 1 × transfer buffer (3.03 g Tris, 14.4 g glycine, 20 mL methanol and 80 mL water) for 120 min at 200 mA. Following transfer, the PVDF membrane was stained with Ponceau S solution for 5 min

64 to confirm uniform protein transfer. The membrane was then blocked in 10% non-fat dietary milk (NFDM) dissolved in tris buffered saline (TBS (Amresco)) containing 1% Tween-20 (Sigma) (TBST), for 1 h at room temperature.

2.6.2.3 Immunoblotting

After blocking, the membrane was incubated with primary antibody (see Table 2.4 for antibody dilutions) in 5% NFDM overnight at 4C with agitation. The blot was then washed three times for 10 min in TBST. The membrane was then incubated in secondary antibody for 1 h with agitation and then washed three times for 10 min in TBST. The secondary antibodies used in this study include anti-rabbit horseradish peroxidase (HRP) (1:5000, 2% NFDM, Sigma), anti-mouse HRP (1:2000, 2% NFDM, Thermo Scientific) and anti-goat HRP (1:10000, 2% NFDM, SantaCruz). Pierce ECL Western Blotting Substrate (Thermo Fisher Scientific) used for histone blots, and SuperSignal West Dura Extended Duration Substrate (Thermo Fisher Scientific) used for visualising all other proteins, are luminol-based enhanced chemiluminescent that detects HRP substrates. The substrate was applied on the membrane and protein bands were visualised using Super HR-G 30 autoradiography film (Fujifilm). To determine the approximate intensity of each band, the exposed films were scanned with an Epson Perfection V200 and the intensity of the bands as determined by the scientific image analysis software Image J (Abramoff et al., 2004, Hartig, 2013, Gassmann et al., 2009). For analysis of additional antibodies, the blot was incubated in Restore Western Blot Stripping Buffer (Thermo Fisher Scientific) for 30 min. Membranes were then washed four times for 10 min in TBST with agitation, and then placed in blocking solution for 1 h as above.

Table 2.4: Primary antibody dilutions used for western blotting Antibody Host/Conjugate Dilution Supplier Actin Rabbit 1:1000 Sigma AldoC Rabbit 1:1000 Abcam Bcl2a1 Rabbit 1:500 Abcam Cbr1 Mouse 1:500 Abcam 65

Ccnd2 Rabbit 1:500 Abcam Cux2 Rabbit 1:1000 Abcam Ezh2 Mouse 1:1000 Cell Signaling Flag Mouse 1:1000 Sigma Gpr84 Goat 1:500 SantaCruz H3K9me1 Rabbit 1:1000 Abcam H3K9me2 Mouse 1:1000 Abcam H3K9me3 Rabbit 1:1000 Abcam H3K27me3 Rabbit 1:2000 Millipore H3K36me1 Rabbit 1:2000 Millipore H3K36me2 Rabbit 1:1000 Millipore H3K36me3 Rabbit 1:2000 Millipore Hlx1 Rabbit 1:1000 Millipore Jmjd1c Rabbit 1:1000 Millipore Jmjd5 Rabbit 1:1000 Abcam Mct4 Rabbit 1:2000 SantaCruz Ms4a3 Rabbit 1:500 LifeSpan BioSciences Pdk1 Mouse 1:500 Abcam Pgk1 Rabbit 1:500 Abcam Phospho-Pkm2 Rabbit 1:1000 Cell Signaling Pkm2 Rabbit 1:1000 Cell Signaling Tcf7l2 Rabbit 1:1000 Sapphire Bioscience Total H3 Rabbit 1:2000 Abcam β-catenin Mouse 1:500 BD Biosciences

2.7 Flow cytometry

Samples were analysed by flow cytometry using a FACSCanto II or LSRFortessa (BD Biosciences). Flow cytometry data were analysed using FlowJo software (FlowJo, TreeStar). A list of antibodies used for flow cytometry analysis are summarised in Table 2.5.

Table 2.5: List of flow cytometry antibodies

Antibody Fluorochrome Conjugate Supplier 7-Amino-actinomycin (7AAD) Percp-Cy5.5 BD Biosciences Annexin V APC BD Biosciences Anti-BrdU APC BD Biosciences CD3 Biotin BD Biosciences CD4 Biotin BD Biosciences CD8a Biotin BD Biosciences CD11b (Mac-1) APC BD Biosciences CD19 Biotin BD Biosciences CD45 (B220) PE BD Biosciences CD117 (c-Kit) APC-Cy7 BD Biosciences IL-7R Biotin BD Biosciences Ly-6G (Gr-1) PE BD Biosciences Propidium iodide (PI) PE Invitrogen Sca-1 FITC BD Biosciences Streptavidin PerCP-Cy5.5 BD Biosciences Ter119 Biotin BD Biosciences

2.7.1 Cell cycle analysis

To detect changes in cell cycle, propidium iodide (PI, Invitrogen) staining was performed. This staining allows for the distribution of cells in different phases of cell cycle (G1, S and G2/M) to be analysed by flow cytometry due to the stoichiometric binding of PI to the amount of DNA present within the cell (Pozarowski and Darzynkiewicz, 2004). The assay involved harvesting cells (1 × 106) and washing with ice cold PBS. Cells were then resuspended by the drop-wise addition of 70% ethanol while vortexing. Samples were then stored at -20C for at least 24 h. Fixed samples were then centrifuged at 500 g for 15 min at 4C, washed with PBS then resuspended in PI (100 µL; 50 µg/mL PI and 100 µg/mL DNase-free RNAse (Invitrogen) in PBS).

Cells were incubated on ice for 20 min protected from light, centrifuged at 500 g for 10 min at 4C, and then resuspended in 1 mL PBS for analysis by flow cytometry.

2.7.2 Apoptosis assay

Apoptosis is a mechanism of programmed cell death which can be detected using Annexin V (BD Biosciences) with the addition of 7-amino-actinomycin (7AAD) (BD Biosciences) in Annexin binding buffer (10 mM HEPES, pH 7.4; 140 mM NaCl; 2.5 mM CaCl2). Annexin V stains phosphatidylserines which are expressed on the plasma membrane in early apoptotic cells, whilst 7AAD is a DNA intercalator which can identify late apoptotic and necrotic cells (Vermes et al., 2000, Zimmerman and Meyer, 2011). The assay involved harvesting cells (1 × 106), washing in PBS and resuspending in 100 µL diluted annexin antibody (99 µL annexin binding buffer with 1 µL Annexin V). The samples were incubated for 10 min at room temperature protected from light. Diluted 7AAD antibody (400 µL; 395 µL annexin binding buffer with 5 µL 7AAD) was added to the sample and incubated on ice for 15 min protected from light and then analysed by flow cytometry. The percentage of total apoptotic cells, including Annexin V-positive/7AAD-negative cells in the early stage of apoptosis and Annexin V/7AAD double positive cells in the end stage of apoptosis, was determined by flow cytometry.

2.7.3 BrdU staining

5-Bromo-2′-deoxyuridine (BrdU) immunostaining was used to detect proliferating cells in vivo. BrdU is a thymidine analogue which incorporates into the DNA during DNA synthesis and can be detected by flow cytometry using anti-BrdU antibodies (Holm et al., 1998, Dolbeare et al., 1983). Mice were injected with BrdU solution (100 µL 10 mg/mL, BD Biosciences) in the intraperitoneal region and then sacrificed after 2 h. The bone marrow was harvested, centrifuged at 500 g for 8 min at 4C and resuspended in 100 µL BD Cytofix/Cytoperm buffer. Following incubation on ice for 20 min, samples were washed with 1 mL BD Perm/Wash Buffer, centrifuged for 8 min, then incubated with 100 µL BD Cytoperm Permeabilization Buffer Plus for 10 min on ice, then washed again with BD Perm/Wash Buffer. Cells were then refixed in 100 µL BD

Cytofix/Cytoperm Buffer for 5 min, washed with BD Perm/Wash Buffer, then incubated with 100 µL of DNase (300 µg/mL) for 1 h at 37C. Samples were then washed with BD Perm/Wash Buffer and stained with diluted BrdU antibody (50 µL BD Perm/Wash Buffer containing 1 µL BrdU antibody) for 20 min at 20C before analysis by flow cytometry.

2.7.4 Fluorescence activated cell sorting

Cells were harvested, washed and centrifuged at 1500 rpm for 5 min. Samples were then resuspended in 1 mL PBS and strained through a 35 µm cell strainer cap into 5 mL round bottom polystyrene tubes (BD Biosciences). Cells were sorted using an Influx fluorescence activated cell sorter (BD Biosciences) under sterile conditions at 4C. The purity of sorted cell populations was > 95%.

2.7.5 Cell surface staining

Diluted cell surface antibody (200 µL; 1 µL antibody in 1 mL of PBS) was added to cells (2 × 105) and incubated on ice for 20 min protected from the light. Following incubation, the samples were analysed by flow cytometry.

2.8 Metabolic assays

2.8.1 ATP assay

A colorimetric adenosine-5‘-triphosphate (ATP) assay kit (Abcam) was employed to measure ATP, where the phosphorylation of glycerol generates a colorimetric product. Cell pellets (1 × 106) were washed in ice cold PBS, resuspended in ATP assay buffer (100 µL) then homogenised by pippeting rapidly. Samples were then centrifuged for 2 min at top speed at 4C. The supernatant was transferred to a new microcentrifuge tube for deproteinisation, where 30 µL of ice cold perchloric acid (4 M, PCA) was added and then incubated on ice for 5 min. Samples were centrifuged at 13000 g for 2 min at 4C, the supernatant was removed and 34 µL of ice cold 2 M KOH was added to neutralise

69 the sample and precipitate excess PCA. The microcentrifuge tube was vortexed, pH was adjusted with 0.1 M KOH to ensure pH was between 6.5 and 8, and then centrifuged at 13000 g for 15 min at 4C. The ATP assay was conducted in a 96-well plate, where sample wells and background control sample wells contained 50 µL of sample, and standard wells contained 50 µL of ATP using standard dilutions between 0 nM and 0.2 mM. The ATP reaction mix (50 µL; 44 µL ATP assay buffer, 2 µL ATP Probe, 2 µL ATP Converter and 2 µL Developer Mix) was added to each sample and standard well, while a background reaction mix (50 µL; 46 µL ATP assay buffer, 2 µL ATP Probe and 2 µL Developer Mix) was added to the background wells. The plate was incubated at room temperature for 30 min protected from light, and then the absorbance was measured at 570 nm using a Benchmark Plus microplate spectrophotometer.

2.8.2 Glycolysis assay

Glycolysis Assay Kit (pH-Xtra, Luxcel Biosciences) was used to measure glycolysis activity. Cells (5 × 105 cells/well) were seeded into a 96-well black wall clear bottom plate (Costar) with 150 µL respiration buffer. Blank wells (160 µL respiration buffer), signal control (150 µL respiration buffer) and positive control (150 µL media with 10 µL glucose oxidase (1mg/mL)) were used in this assay. Reconstituted pH-Xtra reagent (10 µL) was added to each well, except for blank control wells. Fluorescence was measured (excitation 380 nm, emission 615 nm) using the SpectraMax plate reader (Molecular Devices) for 180 min with plate temperature set at 37C. Triplicate wells were used per sample. Glycolytic flux was determined by measuring extracellular acidification, where lactate production is the main contributor to this acidification (Hynes et al., 2012, Hynes et al., 2009). To calculate the extracellular acidification rate, linear regression is applied to the linear portion of the signal profiles to determine the slope of intensity versus time.

2.8.3 Oxidative phosphorylation assay

Oxygen Consumption Rate Assay Kit (MitoXpress, Luxcel Biosciences) was used to measure oxygen phosphorylation activity. This measurement is based on mitochondrial

70 function, where the amount of oxygen-sensing fluorophore is inversely proportional to the amount of extracellular oxygen in the sample (Hynes et al., 2006, Hynes et al., 2009). Cells (8 × 104 cells/well) were seeded into a 96-well black wall clear bottom plate (Costar) with 150 µL media (15% FCS/IMDM). Blank wells (170 µL media), negative control (150 µL media with 10 µL antimycin A (2 µM)) and positive control (150 µL media with 10 µL glucose oxidase (10 mg/mL)) were included in this assay. MitoXpress (10 µL) was added to all wells except for blank wells, and then 100 µL of HS Mineral Oil was used to cover all wells before measuring fluorescence (excitation 380 nm, emission 650 nm) on a SpectraMax M5 microplate reader (Molecular Devices) for 120 min with plate temperature set at 37C. Triplicate wells were used per sample. To calculate the oxygen consumption rate, linear regression was applied to the linear portion of the signal profiles to determine the slope of intensity versus time.

2.9 Animal work

2.9.1 Mouse injections

Murine pre-LSC or LSC were injected into the lateral tail vein of sublethally irradiated (680 cGy, using X-RAD 320 biological irradiator (Precision X-Ray)) female C57BL/6 mice and monitored for symptoms of leukaemia. When the mice displayed any leukaemic symptoms, they were sacrificed, the spleen was weighed and cells from the spleen and bone marrow were harvested for flow cytometric analysis. All animal work was conducted using protocols approved by the Animal Care and Ethics Committee (ACEC# 11/124B; NLRD 14-06).

2.9.2 Organ harvesting

To harvest bone marrow cells, muscle and connective tissue were removed from the femur and tibia, the bones were flushed with PBS using a syringe into 50 mL conical tubes, and then centrifuged at 1500 rpm for 5 min. For spleen cells, connective tissue was removed and the spleen was crushed using a mortar and pestle. The cells were filtered through a 40 µM strainer (BD Biosciences) into a 50 mL falcon tube and centrifuged at 1500 rpm for 5 min. The bone marrow and spleen cells were resuspended 71 in 3 mL BD PharmLyse red blood cell lysis buffer (diluted 1:10 in milliQ water, BD Biosciences) then incubated at room temperature for 10 min protected from light. PBS (30 mL) was added to the conical tube, the cells were centrifuged at 1500 rpm for 5 min and then resuspended in 10 mL PBS to strain the cells through a 40 µM strainer (BD Biosciences) into a new 50 mL conical tube. The tubes were then centrifuged, cells were counted using trypan blue assay. The cells were analysed by flow cytometry and then cryopreserved using 90% FCS/10% DMSO.

2.10 Limiting dilution analysis

Limiting dilution analysis was performed using the extreme limiting dilution analysis (ELDA) software designed to predict the frequency of active cells in a given population following transplantations such as the depletion or enrichment of stem cells (Hu and Smyth, 2009).

2.11 Statistical analysis

Statistical analysis was performed using the Student t test. Results are given as mean ± standard error mean (SEM). P-values less than 0.05 are considered statistically significant.

CHAPTER 3

The Role of JMJD1C in Regulating Acute Myeloid Leukaemic Stem Cells

3.1 Introduction

As reviewed in Chapter 1, MLL rearranged-AML is one of the more common and aggressive AML subtypes (Krivtsov et al., 2006, Raimondi et al., 1999, Meyer et al., 2013). Experiments involving chromatin immunoprecipitation followed by next- generation sequencing identified a number of direct MLLAF9 target genes, which included the extensively studied HoxA cluster genes and Meis1, and less well-defined targets such as the histone demethylase, Jmjd1c (Bernt et al., 2011). Although histone demethylases are the most recent family of epigenetic regulators to be discovered, members of this family have been shown to play important roles in cancer initiation and metastasis (Greer and Shi, 2012) in both solid cancers and leukaemias (Hoffman et al., 2012). JMJD1C, also known as JHDM2C or thyroid receptor interacting protein 8 (TRIP8), was originally identified as a thyroid receptor protein (Lee et al., 1995) and an androgen receptor coactivator (Wolf et al., 2007), and has been the subject of limited studies. Existing publications on JMJD1C include its involvement in spermatogenesis (Kuroki et al., 2013), autism (Saez et al., 2015), embryonic stem cells (Shakya et al., 2015), gastric cancer (Katoh and Katoh, 2007) and leukaemia (Sroczynska et al., 2014).

Studies in embryonic stem cells have identified Oct4 as one of a small group of pluripotency regulators, including Nanog and Sox2 (Boyer et al., 2005). This transcription factor, which is indispensable in controlling the establishment and maintenance of pluripotent, undifferentiated cells (Soltanian and Matin, 2011, Tantin, 2013), has been shown to play a role in maintaining self-renewal capacity in lung, pancreatic and liver cancer cells (Tai et al., 2005, Chen et al., 2008b). Oct4 was previously shown to positively regulate Jmjd1a and Jmjd2c, both histone 3 lysine 9 demethylases (H3K9me2 and H3K9me3, respectively) (Loh et al., 2007). JMJD1A was found to demethylate H3K9me2 at the promoter regions of a number of pluripotency- associated genes activating their transcription, whilst JMJD2C positively regulates the

73 transcription factor for self-renewal, Nanog. Interestingly, JMJD1C, closely related to JMJD1A and JMJD1B, was recently identified to cooperate with Oct4 to mediate H3K9me2 demethylation of known Oct4 targets in embryonic stem cells (Shakya et al., 2015), where comparative genomics analyses revealed binding sites for Pou5f1 (Oct3/Oct4), AP-1 and bHLH transcription factors (Katoh and Katoh, 2007). JMJD1C was also recently found to inhibit differentiation of embryonic stem cells (Wang et al., 2014a), which would provide cells a survival advantage when applied to a cancer context suggesting a possible oncogenic role in cancer.

In leukaemia, JMJD1C is highly expressed in MLL rearranged AML and ALL cell lines (Sroczynska et al., 2014) and MLL AML paediatric samples (Balgobind et al., 2010, Ross et al., 2014) compared with other mutations. It is, however, crucial to only target genes that are upregulated in leukaemic cells compared to healthy cells in order to reduce side effects when targeting the gene of interest and this has not yet been established for JMJD1C. The persistence of LSC has been attributed to be the cause of disease relapse due to their high resistance to conventional chemotherapy (Ishikawa et al., 2007), enhanced β-catenin signalling (Wang et al., 2010), and aberrant regulation of metabolic pathways (Lagadinou et al., 2013). It is therefore important to also examine the expression of JMJD1C in AML LSC to assess the potential of targeting this gene to selectively impair LSC, the main contributor to disease relapse in AML patients. Furthermore, it is important to uncover the role of JMJD1C in the various stages of leukaemia development, from leukaemia initiation to the maintenance of established leukaemia in order to understand how targeting JMJD1C at various stages of the disease may affect AML development and survival.

This chapter aims to examine the role of JMJD1C in regulating AML LSC and to determine the potential of targeting this oncogene as a novel therapeutic strategy for AML. To address this aim, the expression of Jmjd1c in HSC and AML LSC derived from KLSA9M or KLSMLLAF9 cells was examined in an AML mouse model. In addition, the effect of aberrantly expressed Jmjd1c on AML leukaemogenesis was investigated, where downstream pathways regulated by Jmjd1c were identified.

3.2 Jmjd1c is upregulated in mouse and human MLL AML models

To determine whether JMJD1C plays a crucial role in regulating LSC and has the potential to be targeted without affecting haematopoietic stem cells (HSC), microarray data generated previously from our laboratory was analysed which compared HSC (c- Kithigh Lin− Sca-1+, KLS) with LSC (c-Kithigh Lin− Sca-1− CD34+ FcγRII/III+) transformed with different oncogenes, Hoxa9/Meis1a (KLSA9M) or MLLAF9 (KLSMLLAF9) (Figure 3.1A). Jmjd1c was identified as the most significantly upregulated gene in the highly aggressive KLSMLLAF9 subtype of AML compared to KLSA9M LSC and HSC (P < 0.05, fold change > 2). Since Jmjd1c was shown to be highly expressed in early-stage and established MLLAF9 LSC compared to HSC, targeting this demethylase may have the potential to selectively eliminate KLSMLLAF9 LSC without harming HSC.

The microarray data were confirmed by western blotting in primary mouse cells from different developmental stages of leukaemia, namely early stage of LSC development termed pre-LSC and fully established LSC (Figures 3.1B-E). Jmjd1c protein expression was consistent with the microarray data, where Jmjd1c expression was higher in KLSMLLAF9 LSC compared to HSC (Figure 3.1B). When comparing KLSA9M and KLSMLLAF9 LSC, Jmjd1c expression was also consistent with the microarray data, showing an upregulation of Jmjd1c in KLSMLLAF9 LSC, the more aggressive cell type (Figure 3.1C). This was also observed in pre-LSC, where Jmjd1c was upregulated in KLSMLLAF9 pre-LSC compared to HSC and KLSA9M pre-LSC (Figures 3.1D and E). Furthermore, survival data from AML patients showed that patients with high JMJD1C expression have a significantly poorer prognosis compared to patients with low JMJD1C expression (P = 0.0029, Figure 3.1F) implicating JMJD1C to be a potential determinant of clinical outcome in AML patients. Collectively, these data suggest that JMJD1C may play a role in the aggressive phenotype observed in KLSMLLAF9 AML and may function as an oncogene.

A B LSC LSC HSC A9M MLL HSC KLSMLLAF9 1 2 3 1 2 3 1 2 3 Jmjd1c

Actin

C LSC KLSA9M KLSMLLAF9

Jmjd1c

Actin

low high

D E HSC KLSMLLAF9 KLSA9M KLSMLLAF9

Jmjd1c Jmjd1c

Actin Actin

F 0.8

Low (N = 114)

0.4

Survival probability

P = 0.0029 High (N = 49) 0.0

0 200 400 600 800 1000 1200 Days

Figure 3.1: Jmjd1c expression is upregulated in mouse and human MLL AML, and associated with poor clinical prognosis (A) Microarray was conducted comparing HSC-enriched c-Kithigh Lin− Sca-1+ (KLS) cells to the LSC-enriched c-Kithigh Lin− Sca-1− CD34+ FcγRII/III+ population isolated from KLSA9M (A9M) and KLSMLLAF9 (MLL). The top most differential genes are shown on the heatmap (P < 0.05, fold change > 2). Western blotting confirms that Jmjd1c is highly expressed in KLSMLLAF9 LSC compared to (B) HSC and (C) KLSA9M LSC. This trend was also observed in pre-LSC, where KLSMLLAF9 had high Jmjd1c expression compared to (D) HSC and (E) KLSA9M pre- LSC. (F) Survival data of AML patients show that subjects with high JMJD1C expression have a significantly poorer clinical prognosis compared to patients with low JMJD1C expression (P = 0.0029). Survival data obtained from PrognoScan, a publicly available database for evaluating patient prognosis.

3.3 Suppression of Jmjd1c impairs leukaemic properties of KLSMLLAF9 in vitro

By profiling genes differentially expressed in HSC and LSC, Jmjd1c was shown to be significantly upregulated in KLSMLLAF9 AML compared to KLSA9M pre-LSC and LSC. To investigate whether suppression of Jmjd1c expression could impair the leukaemic properties of KLSMLLAF9-AML, shRNA-mediated inhibition of Jmjd1c (shJmjd1c) was used in KLSMLLAF9 pre-LSC with a non-targeting shRNA (Scramble/Scr) as a control (detailed in Chapter 2). Reduced expression of Jmjd1c at the mRNA level was confirmed by qRT-PCR and showed that shRNA1 had 15% knockdown, whereas shRNAs 2, 3 and 4 had significant knockdown of 42%, 39% and 44% respectively (P < 0.0005) (Figure 3.2A). This correlated to the knockdown observed at the protein level, where western blotting showed knockdown in shRNAs 2, 3 and 4 (Figure 3.2B). The reduction in Jmjd1c expression was quantified using densitometric analysis, which showed a 72%, 55% and 67% decrease in Jmjd1c expression in shRNAs 2, 3 and 4, respectively, compared to control cells.

To assess whether Jmjd1c knockdown would influence the clonogenic ability of KLSMLLAF9-AML cells, colony-forming cell assays were performed. ShRNA 1 had no significant change in colony forming ability (Figure 3.2C) or cell number (Figure 3.2D) from colony assays compared to control, which could be explained by the poor knockdown efficiency of this shRNA. Effective suppression of Jmjd1c expression by shRNAs 2, 3 and 4 resulted in a significant reduction in colony forming capacity (2 to 3.3-fold, P = 0.018; Figure 3.2C) and proliferative potential (1.8 to 2.6-fold, P < 0.0001; Figure 3.2D). Additionally, the colony morphology was analysed and the reduction in colony size and density supports the role of Jmjd1c in cell proliferation (Figure 3.2E).

To determine whether the reduction in proliferation seen in Jmjd1c knockdown cells could be attributed to changes in cell cycle, propidium iodide staining was performed. The distribution of cells in each cell cycle phase showed a significant accumulation of cells in G1 phase (12% increase, P = 0.01) along with a significant reduction of cells in S phase (8% decrease, P = 0.0008) for shRNA 4 cells compared to control (Figure 3.3A).

A 2 a 1 c s h e x p B

1 .2 KLSMLLAF9

1 .0

n Scr sh1 sh2 sh3 sh4

)

s e

s 0 .8 g

e *** n ***

r *** a

p Jmjd1c h

x 0 .6

l o

N 0 .4

f (

R Actin

m 0 .2

0 .0 1.00 0.56 0.28 0.45 0.33 S c r s h 1 s h 2 s h 3 s h 4

C 2 c 1 c s h c fu D 2 d 1 c s h c e ll n o 5 0 0 1 .2

) 1 .0

6 l

l 4 0 0

 0 .8

(

0 3 0 0 * * * * e

0 0 .6 *

b 1

y 2 0 0

u n

n 0 .4

l e

o 1 0 0 C C 0 .2

0 0 .0 S c r s h 1 s h 2 s h 3 s h 4 S c r s h 1 s h 2 s h 3 s h 4

E Scr sh1 sh2 sh3 sh4

200 µm

Figure 3.2: Knockdown of Jmjd1c impairs the proliferative and colony forming abilities of KLSMLLAF9 pre-LSC in vitro Suppression of Jmjd1c expression in lentivirally transduced KLSMLLAF9 pre-LSC with four Jmjd1c shRNAs (sh1, sh2, sh3 and sh4) compared to scramble control (Scr) was analysed by (A) qRT-PCR and (B) western blotting. The intensity of the bands comparing shJMJD1c relative to Scr was quantified by Image J to measure the reduction in Jmjd1c levels normalised to actin. Colony forming assays showed a significant reduction in (C) colony number and (D) cell number for shRNAs 2, 3 and 4 compared to control. (E) Representative microscope images of colony morphology following colony forming assays with Scr or shJmjd1c. Data represents mean values ± SEM (Standard Error of the Mean) of three independent experiments. Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (* P < 0.05, *** P < 0.0005). Images captured using 10× magnification lens.

3 b 1 c s h c e ll c y c le A 6 0 * *** S c r

) s h 4

% (

4 0

u m

u 2 0

c A

0 G 1 S G 2 /M

B Scr sh4

V Annexin

7AAD

Figure 3.3: Suppression of Jmjd1c blocks cell cycle progression but does not induce apoptosis (A) Representative histograms of cell cycle assay showing an accumulation of cells in G1 phase in shRNA 4 (sh4) compared to control (Scr). (B) Levels of apoptosis remained relatively unchanged upon Jmjd1c suppression. Data represent mean values ± SEM of three independent experiments. Asterisks indicate statistical significance obtained using an unpaired Student‘s t- test (* P < 0.05, *** P < 0.0005).

Since colony morphology changes were observed upon suppression of Jmjd1c, an apoptosis assay was implemented to assess the presence of apoptotic cells, discriminating between cells in early or late apoptosis/necrosis. There was no significant change detected between shRNA 4 and the control cells following AnnexinV/7AAD staining (Figure 3.3B). Hence, the reduction in colony forming capacity and proliferative potential observed in shJmjd1c appear to be due to the block in G1/S phase cell cycle transition.

3.4 Jmjd1c is required for the maintenance of established KLSMLLAF9 leukaemia in vivo

Due to the promising results observed in vitro supporting the hypothesis that Jmjd1c may function as an oncogene in AML, it was important to determine whether the suppression of Jmjd1c could impair leukaemogenesis in vivo. KLSMLLAF9 pre-LSC transduced with Scr or shRNAs 2 and 4 were injected into the lateral tail vein of sublethally irradiated C57BL/6 mice and monitored for symptoms of leukaemia. Although the in vitro data showed impaired LSC properties upon Jmjd1c suppression, the survival curve of shRNAs 2 and 4 did not alter the rate of leukaemia initiation compared to control, as there was no delay in disease onset (Figure 3.4A). The flow cytometric analysis also showed no significant change in the percentage of GFP- positive leukaemic cells in the bone marrow between the sick mice injected with Scr or Jmjd1c knockdown cells suggesting that the leukaemic cells were able to infiltrate and proliferate in the bone marrow to a similar degree (Figure 3.4B). There was, however, a significant reduction in the percentage of GFP-positive cells infiltrating the spleen in mice injected with both shRNA 2 (2.2-fold, P = 0.04) and shRNA 4 cells (3.4-fold, P = 0.02) (Figure 3.4C). This reduction in infiltration into the spleen may contribute to the significant reduction in spleen weight observed in mice inoculated with shJmjd1c (2.7- fold for shRNA 2, P = 0.0028; 1.7-fold for shRNA 4, P = 0.026) compared to control (Figure 3.4D).

To confirm that the mice which displayed leukaemic symptoms had succumbed to AML, the leukaemic cells harvested from sick mice were analysed by flow cytometry

B A 4 b in filta tio n b o n e m a rro w 1 2 0 1 0 0 S c r

s h 2

l )

a 7 5 s h 4 1 0 0

(

s e

8 0 t

5 0 c

F e

2 5 G 6 0 P

0 4 0 0 5 0 1 0 0 1 5 0 2 0 0 S c r s h 2 s h 4 D a y s p o s t tra n s p la n ta tio n

* * * ** D C 4 c in filtra tio n s p l 4 d s p l w e ig h t 1 0 0 0 .6

8 0

) )

% 0 .4

(

s 6 0

i +

4 0 e P

W 0 .2

F G 2 0

0 0 .0 S c r s h 2 s h 4 S c r s h 2 s h 4

Figure 3.4: Suppression of Jmjd1c does not alter MLL leukaemia initiation (A) Survival curve of primary recipient mice inoculated with 1 × 106 KLSMLLAF9 pre-LSC with Scr or Jmjd1c shRNAs 2 and 4 (N = 12) showed no difference in survival. Infiltration of GFP cells in the (B) bone marrow and (C) spleen was determined at time of sacrifice. The significantly lower infiltration of leukaemic cells into the spleen is reflected in the significantly reduced spleen weights (D). Horizontal line represents the median. Each data point represents one mouse. Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (* P < 0.05, ** P < 0.005).

Scr Scr

Gr B220

CD3 Mac-1

sh4 sh4

B220 Gr

CD3 Mac-1

Figure 3.5: Jmjd1c suppression retains an AML immunophenotype in vivo Representative scatter plots from flow cytometric analysis of leukaemic cells harvested from primary recipients display a CD3− B220− Gr-1 Mac-1 immunophenotype consistent with AML in both (A) Scr and (B) shRNA 4.

83 for an AML immunophenotype, which is negative for both B-cell and T cell markers, B220 (CD45) and CD3, respectively, and positive for myeloid maturation markers Gr-1 (Ly-6G) and Mac-1 (CD11b) (Stubbs et al., 2008, Wang et al., 2010). Flow cytometric analysis showed that mice inoculated with shJmjd1c cells (Figure 3.5B) retained the AML immunophenotype as seen in mice injected with Scr cells (Figure 3.5A), confirming that the sick mice in all cohorts developed AML.

The lack of improvement in survival observed following knockdown of Jmjd1c by shRNA in KLSMLLAF9 pre-LSC cells suggests that this demethylase does not regulate leukaemia initiation. Rather, Jmjd1c may be involved in leukaemia maintenance, and to investigate this, leukaemia burden in secondary recipient mice was assessed. The leukaemic cells harvested from primary leukaemic mice (1 × 105 KLSMLLAF9 LSC with Scr or shJmjd1c) were transplanted into secondary mice. Survival curves show a significant delay in survival for shJmjd1c compared to the Scr control (P < 0.0001; Figure 3.6A). To determine the LSC frequency within a leukaemic population, limiting dilution assays were performed, where mice were injected with 1 × 104 KLSMLLAF9 LSC or 1 × 103 KLSMLLAF9 LSC. Based on the frequency of disease onset for each dilution, the frequency of repopulating stem cells can be estimated (Illa-Bochaca et al., 2010, Hu and Smyth, 2009). Survival curves showed that when secondary mice were injected with 1 × 104 KLSMLLAF9 LSC, mice with shJmjd1c had a significant delay in leukaemia onset compared to control mice (Figure 3.6B). When mice were inoculated with 1 × 103 cells, only one mouse from the shRNA 2 cohort developed leukaemia, and none of the shRNA 4 cohort developed leukaemia, whereas all of the control mice developed aggressive leukaemia (Figure 3.6B), suggesting a higher frequency of LSC in the latter population of leukaemic cells compared to shRNAs 2 and 4 (Table S.1). Furthermore, tertiary transplantation showed that dysregulation of leukaemic maintenance persisted in mice with shJmjd1c as shown by the significant delay in leukaemia onset (Figure 3.6C). These data suggest that Jmjd1c is required for leukaemia maintenance of MLLAF9-AML, but not leukaemia initiation.

A 1 0 0 S c r

s h 2 l

a 7 5 s h 4

t 5 0

c r

e 2 5 P

P < 0.0001 0 0 2 0 4 0 6 0 8 0 1 0 0 D a y s p o s t tra n s p la n ta tio n

1 103 B 1 0 0 1 0 k s c r

1 103 1 0 k s h 2 l

a 7 5 1 0 k s h 4

i v

r 1 k s c r 4

u 1x10 s

1 k s h 2

t 5 0

n 4

e 1 10 1 k s h 4

c r

e 2 5 P

1 104 1 103 0 0 2 0 4 0 6 0 8 0 1 0 0 1 2 0 1 4 0 1 6 0 1 8 0 2 0 0 2 2 0 D a y s p o s t tra n s p la n ta tio n S u rv iv a l p ro p o rtio n s : S u rv iv a l o f 6 c te r tia ry s u r v iv a l

C 1 0 0 1 0 0 k s c r

1 0 0 k s h 2 l

a 7 5

v 1 0 0 k s h 4

t 5 0

c r

e 2 5 P P = 0.001 0 0 2 0 4 0 6 0 D a y s p o s t tra n s p la n ta tio n

Figure 3.6: Jmjd1c is required for leukaemia maintenance (A) Survival curves of secondary recipient mice inoculated with 1 × 105 KLSMLLAF9 LSC shRNA 2 (N = 9) or shRNA 4 (N = 13) showed a significant delay in leukaemia onset upon Jmjd1c suppression compared to Scr (N = 12). (B) Limiting dilution assay in secondary recipient mice injected with 1 × 103 or 1 × 104 KLSMLLAF9 LSC illustrates a reduction in LSC frequency in Jmjd1c-deficient cells (N = 12) compared to Scr (N = 12). (C) Tertiary transplantation of 1 × 105 KLSMLLAF9 LSC resulted in prolonged survival in mice with shJmjd1c (N = 5), suggesting impaired leukaemic maintenance compared to Scr (N = 5). Each data point in survival curves represents one mouse. 85

3.5 Jmjd1c plays a more critical role in stem cell-derived MLLAF9 leukaemia

Thus far, the results have shown that Jmjd1c plays a critical role in leukaemic maintenance in HSC-derived MLLAF9 (KLSMLLAF9) AML cells. To determine whether the cell of origin influences the role of Jmjd1c, MLLAF9 LSC derived from GMP (c-Kithigh Sca-1 Lin IL-7R CD34 FcγRII/IIIhigh) were transduced with shRNA 4 or the scramble control. Validation of efficient gene silencing by shRNA-mediated knockdown was assessed by western blotting (Figure 3.7A). Colony forming assays showed that there was a significant reduction in colony formation (1.6-fold change, P = 0.0089; Figure 3.7B) and clonogenic potential (1.7-fold change, P = 0.0235; Figure 3.7C), however, this was not translated in vivo, as there was no delay observed in survival between mice inoculated with shRNA 4 or control (Figure 3.7D). This suggests therefore, that Jmjd1c is only required for MLLAF9 leukaemia maintenance when derived from stem cells (KLSMLLAF9) and not the more differentiated progenitor cells (GMPMLLAF9).

3.6 Overexpression of Jmjd1c confers a growth advantage to KLSA9M pre-LSC in vitro

Previous studies have shown that KLSA9M cells develop a less aggressive form of AML compared to KLSMLLAF9 cells in mice, while maintaining a similar disease pathology and phenotype (Krivtsov et al., 2013, Krivtsov et al., 2006, Wang et al., 2010). To determine whether Jmjd1c has the potential to augment KLSA9M leukaemogenicity, Jmjd1c was overexpressed in KLSA9M pre-LSC that otherwise have low Jmjd1c expression (Figure 3.1E).

Jmjd1c cDNA or empty vector (EV) was retrovirally introduced into KLSA9M pre- LSC. The expression of Jmjd1c cDNA in KLSA9M was confirmed using western blotting, where there was a marked increase in Jmjd1c expression (Figure 3.8A). The overexpression of Jmjd1c significantly increased the colony forming potential by 2-fold (P = 0.0093, Figure 3.8B). Similarly, KLSA9M with Jmjd1c overexpressing construct

A GMPMLLAF9 LSC Scr sh4

Jmjd1c

Actin

B c fu g m ls c C

4 0 0 2 .0

)

l l

0 1 .5 e

3 0 0 1

** 

(

0 e

0 1 .0 b

1 2 0 0

l o

l 0 .5

1 0 0 e

C C

0 .0 0 S c r s h 4 S c r s h 4

D 1 0 0 S c r

s h 4 l

a 7 5

t 5 0

c r

e 2 5 P

0 0 2 0 4 0 6 0 8 0 D a y s p o s t tra n s p la n ta tio n

Figure 3.7: Jmjd1c is not required for GMPMLLAF9 maintenance (A) Jmjd1c expression in GMPMLLAF9 LSC lentivirally transduced with scramble (Scr) or shRNA 4 (sh4) was confirmed by western blotting. Colony forming assays showed a significant reduction in (B) colony number and (C) cell number for shRNA 4 compared to control. (D) Survival curves of mice inoculated with 1 × 105 GMPMLLAF9 LSC resulted in a lack of delay in survival for shRNA 4 (N = 6) compared to Scr (N = 6), which suggests that Jmjd1c does not impair leukaemic maintenance for GMPMLLAF9 LSC. Data represent mean values ± SEM of three independent experiments. Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (* P < 0.05, ** P < 0.005). Each data point in survival curves represents one mouse.

KLSA9M

A EV Jmjd1c Jmjd1c

Actin

9 c 1 c c e ll n o B 9 b 1 c c fu C 2 .5 8 0 **

) ****

6 2 .0

1 e

6 0



(

0 1 .5

0 b

1 4 0

/ m

1 .0

l l

2 0 l o

e 0 .5

C C

0 0 .0 E V Jm jd 1 c E V Jm jd 1 c

D E 9 e 1 c c d a c e ll c y c le 6 0 E V EV Jmjd1c *

) J m jd 1 c

% (

4 0

o i

2D t

u m

u 2 0

c c 200 µm A

0 G 1 S G 2 /M

Figure 3.8: Jmjd1c confers a growth advantage to KLSA9M pre-LSC in vitro KLSA9M pre-LSC were retrovirally transduced with Jmjd1c cDNA or empty vector (EV). (A) Jmjd1c expression was confirmed using western blotting. Colony forming assays with KLSA9M pre-LSC showed a significant increase in (B) colony number and (C) cell number for cells with Jmjd1c cDNA compared to control. (D) Representative microscope images of colony morphology illustrate a marked increase in colony size and density upon Jmjd1c overexpression. (E) Cell cycle analysis found that high expression of Jmjd1c significantly increases the accumulation of cells in S phase. Data represent mean values ± SEM of three independent experiments. Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (* P < 0.05, ** P < 0.005, **** P < 0.0001). Images captured using 10× magnification lens.

88 had a 2.4-fold increased rate of proliferation (P < 0.0001; Figure 3.8C), a finding also supported by the colony morphology observed in the representative microscope images (Figures 3.8D). This suggests that Jmjd1c confers a growth advantage to KLSA9M cells by significantly increasing the clonogenic and proliferative ability in vitro.

To confirm whether the increased proliferation observed in vitro could be explained by changes in cell cycle, KLSA9M with EV or Jmjd1c cDNA were fixed and stained with PI. Following flow cytometric analysis, Jmjd1c overexpressing cells were found to have a 4.1% reduction in G1 phase and an 8.4% increase in accumulation of cells in the S phase compared to control cells (P = 0.006; Figure 3.8E). This result is consistent with the Jmjd1c shRNA data, where Jmjd1c knockdown caused the cells to accumulate in G1 phase blocking them from proceeding onto S phase (Figure 3.3B). Overall, these findings suggest that Jmjd1c is able to enhance the leukaemic properties of KLSA9M pre-LSC through regulation of cell cycle progression, allowing the cells to proliferate at a higher capacity.

3.7 Jmjd1c enhances leukaemogenesis

To determine whether Jmjd1c could enhance leukaemogenesis of KLSA9M in vivo, KLSA9M cells transduced with EV or Jmjd1c cDNA were transplanted into sublethally irradiated mice and monitored for disease onset. The survival curve of primary recipient mice showed that mice inoculated with KLSA9M overexpressing Jmjd1c developed AML at a significantly accelerated rate compared to the control group (P < 0.0001; Figure 3.9A). The infiltration of leukaemic cells into the bone marrow was similar in both the control and Jmjd1c overexpressing cell types, however, there was a 14% increase in infiltration into the spleen (P = 0.033; Figure 3.9B), indicated by the GFP- positive population. This increase in migration of leukaemic cells to the spleen may contribute to the 11% increase in spleen weight observed in mice injected with Jmjd1c (P = 0.0288; Figure 3.9C).

To examine the maintenance of leukaemia burden in mice, GFP-positive cells harvested from leukaemic mice from primary transplantation were inoculated into secondary recipient mice. The acceleration in leukaemic onset observed from primary transplantation of KLSA9M overexpressing Jmjd1c was also seen in the secondary 89

AS u r v iv a l p ro p o rtio n s : S u r v iv a l o f 1 0 a c d n a s u rv iv a l 1 0 0 E V

J m jd 1 c l

a 7 5

t 5 0

c r

e 2 5 P

P < 0.0001 0 0 3 0 6 0 9 0 1 2 0 1 5 0 1 8 0 2 1 0 D a y s p o s t tra n s p la n ta tio n

B 1 0 b in filtra tio n C 1 0 c s p l w e ig h t * 1 0 0 * 0 .5

8 0 ) 0 .4

)

(

6 0 g 0 .3

4 0 n 0 .2

p G

2 0 S 0 .1

0 0 .0 E V Jm jd 1 c E V Jm jd 1 c E V Jm jd 1 c

BMS u rv iv a l p ro pSoplr tio n s : S u r v iv a l o f 3 .9 d s e c o n d a r y

1 0 0 D E V

J m jd 1 c l

a 7 5

t 5 0

c r

e 2 5 P P < 0.0001

0 0 2 0 4 0 6 0 S u rv iv aSl uprrvoivpaol rptDrioa ypnsos pr:ot iSsotu ntrrsav:n iSsvpualarlvn oitvafat i1ol n0off l1i0mai tcidnnga d siulurvtiiov anl

1 0 0 E 1 0 0 E V 1 103 1 0 k E V

J m jd 1 c l

a 7 5 1 0 k J M J D 1 c

i a

7 5 v

r v

i 1 k E V

t 5 0

u 1 k J M J D 1 c n

s 4

e 1 10

t 5 0

n e

e 2 5

c r

e 2 5 P 0 1 103 0 3 0 1 6100 4 9 0 1 2 0 1 5 0 1 8 0 2 1 0 0 D a y s p o s t tra n s p la n ta tio n 0 2 5 5 0 7 5 1 0 0 D a y s p o s t tra n s p la n ta tio n

Figure 3.9: Jmjd1c enhances leukaemogenesis in KLSA9M (A) Primary recipient mice inoculated with 1 × 106 KLSA9M pre-LSC showed a significant acceleration in disease onset in mice with Jmjd1c cDNA (N = 23) compared to EV (N = 14). (B) Infiltration of GFP cells in the bone marrow and spleen was determined at time of sacrifice. (C) The significantly higher infiltration of leukaemic cells into the spleen is reflected in the significantly larger spleens for mice with KLSA9M overexpressing Jmjd1c compared to control. (D) Secondary recipient mice (N = 11) were injected with 1 × 105 KLSA9M LSC harvested from sick primary recipient mice. (E) Limiting dilution assay in secondary recipient mice injected with 1 × 103 (N = 5) or 1 × 104 (N = 9) KLSA9M LSC illustrates an increase in LSC frequency in Jmjd1c-overexpressing cells compared to control. Horizontal line represents the median. Each data point represents one mouse. Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (* P < 0.05).

91 transplantations (P < 0.0001; Figure 3.9D). These data therefore support the hypothesis that Jmjd1c functions as an oncogene and enhances leukaemogenesis.

A limiting dilution assay was performed in vivo to assess the frequency of self-renewing leukaemic cells. The accelerated disease latency observed in primary survival curves was maintained when secondary recipient mice were inoculated with 1 × 104 cells or 1 × 103 cells compared to control; however, only two thirds of the 1 × 104 control cohort succumbed to leukaemia, whilst none of the 1 × 103 control cohort developed leukaemia (Figure 3.9E). This suggests that Jmjd1c overexpressing leukaemic cells have a higher frequency of LSC enabling these cells to reconstitute leukaemia in subsequent serial transplantations at low cell numbers (Table S.1).

To further assess the oncogenic potential of Jmjd1c, homing experiments were performed in vivo to examine the ability of leukaemic cells to home to the bone marrow post transplantation. Representative scatter plots following flow cytometric analysis showed that KLSA9M cells overexpressing Jmjd1c had an increased capacity to home to the bone marrow (Figure 3.10A). Quantification of the percentage of GFP-positive cells showed that Jmjd1c increased the homing ability of KLSA9M LSC by 22-fold (P = 0.0013) compared to the control (Figure 3.10B). This is a favourable property for leukaemic cells, as homing more easily to the bone marrow allows KLSA9M cells to proliferate and reconstitute the bone marrow with leukaemic cells to initiate leukaemia.

Proliferation assays were also performed in vivo to confirm the enhanced proliferative potential of Jmjd1c in KLSA9M observed in vitro. Secondary recipient mice were injected with LSC and sacrificed 10 days post inoculation. The representative scatter plots show an increase in GFP-positive bone marrow cells in mice injected with Jmjd1c overexpressing KLSA9M cells (Figure 3.11A). Normal bone marrow from healthy C57/Bl6 mice was used as a control for gating the GFP-positive population. The improved proliferative potential of Jmjd1c observed in vitro (Figure 3.8C) is supported in vivo, where analysis of flow cytometry data revealed a 9.3-fold increase in the GFP- positive population (P < 0.0001; Figures 3.11A and B). Jmjd1c is therefore able to enhance proliferation of KLSA9M-AML cells. To further support this finding, 5- bromo-2′-deoxyuridine (BrdU) immunostaining was used to detect proliferating cells in vivo. Representative histograms and dot plots illustrate a significant increase in BrdU

A Normal BM EV Jmjd1c SSC

GFP

B 1 1 b ***h o m in g 0 .4

) 0 .3

(

l l

e 0 .2

P F

G 0 .1

0 .0 E V Jm jd 1 c

Figure 3.10: Overexpression of Jmjd1c in KLSA9M enhances engraftment in the bone marrow (A) Representative scatter plots and (B) dot plot of KLSA9M LSC with Jmjd1c cDNA or EV in mice 24 h after transplantation. The significant increase in GFP-positive percentage in the bone marrow illustrates the ability for Jmjd1c overexpressing cells to home to the bone marrow. Horizontal line represents the median. Each data point represents one mouse. Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (*** P < 0.0005).

A Normal BM EV Jmjd1c SSC

GFP 1 2 b g fp B

1 5 ****

) %

( 1 0

c +

P 5

F G

0 E V Jm jd 1 c

EV Jmjd1c

C % Max %

BrdU

** D 1 2 d b r d u 5 0

4 0

l l

e 3 0

+ U

d 2 0

r B

1 0

0 E V Jm jd 1 c

Figure 3.11: Jmjd1c increases proliferation of KLSA9M in vivo (A) Representative scatter plots and (B) dot plot of KLSA9M LSC with Jmjd1c cDNA or empty vector (EV) in mice 10 days after transplantation. The significant increase in GFP-positive percentage in the bone marrow illustrates the enhanced proliferation of Jmjd1c overexpressing cells compared to control. (C) Representative histograms and (D) dot plot illustrates a significant increase in BrdU incorporation of secondary recipient mice. Each data point represents one mouse. Horizontal line represents the median. Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (** P < 0.005, **** P < 0.0001).

95 incorporation in mice with Jmjd1c cDNA compared to control cohort, suggesting that there is a greater percentage of proliferating cells within the leukaemic population (16% increase, P = 0.0072; Figures 3.11C and D). Jmjd1c is therefore involved in the positive regulation of cell proliferation in vivo.

The investigations into the enforced expression of Jmjd1c in KLSA9M demonstrate that Jmjd1c enhances cell proliferation both in vitro and in vivo likely through the accumulation of cells in S phase. These factors could contribute to the development of AML where mice with high Jmjd1c expression had shorter disease latency in both primary and secondary transplantations with increased infiltration into the spleen at time of sacrifice upon primary transplantation, and increased LSC homing, which consequently provide KLSA9M cells a leukaemic advantage.

3.8 Jmjd1c enhances proliferation of HSC

HSC are a rare population in the bone marrow that present an opportunity to better understand stem cell traits (Ema et al., 2000, Goodell et al., 1996, Morrison et al., 1996, Osawa et al., 1996, Wagers and Weissman, 2006). Maintenance of these cells in vitro is difficult even when cultured with a mixture of currently known haematopoietic cytokines, therefore HSC lose their repopulating ability and cannot be maintained long- term (Krosl et al., 2003). Co-cultures of HSC with primary bone marrow stromal cells can only extend the repopulating period up to several weeks in vitro (Dexter et al., 1977, Schofield, 1978, Frimberger et al., 2001). Oncogenes such as MLLAF9 have the ability to transform HSC into cancerous cells, allowing the self-renewal properties of HSC to be retained and therefore be maintained long-term in vitro.

Jmjd1c has been shown to enhance the colony forming capacity, proliferation and leukaemogenicity of KLSA9M pre-LSC, so in order to determine whether Jmjd1c has the potential to transform HSC to promote proliferation and prolong survival, HSC were flow sorted from naive mice and retrovirally transduced with Jmjd1c cDNA or EV. The cells were replated weekly following antibiotic selection for colony forming assays where cell proliferation was measured over 8 weeks. Overexpression of Jmjd1c significantly increased the proliferation potential of HSC and extended the cell viability of HSC for a further 4 weeks compared to control, where HSC transduced with EV 96 ultimately lost the ability to proliferate after week 4 (Figure 3.12A). The results indicate that Jmjd1c-transduced HSC showed significant expansion in vitro, where weeks 2, 3 and 4 had a 10-, 11.9-, and 16.2-fold increase in cell number, respectively, compared to control (P < 0.01). Representative microscope images illustrate the colony morphology at week 4, where HSC transduced with EV were single cells, however, HSC overexpressing Jmjd1c were still able to form colonies (Figure 3.12B). The number of HSC colonies initially observed in this assay was high, but the gradual reduction each week suggests that Jmjd1c may not fully transform HSC (Figure 3.12A) but rather possesses oncogenic properties such as enhanced proliferation.

3.9 Jmjd1c target genes are involved in regulation of metabolic pathways

In order to identify downstream targets of Jmjd1c and determine the pathways regulated by Jmjd1c, KLSA9M cells with EV or Jmjd1c cDNA were submitted for microarray analysis. The RNA expression profile obtained from the microarray was used for gene set enrichment analysis (GSEA), which assesses microarray data at the level of gene sets. The gene sets are defined based on existing published biological knowledge and represent groups of genes that share common biological function or regulation of pathways.

GSEA provides biologically meaningful insights, and previously published cancer and disease datasets have revealed a relationship between Jmjd1c and a number of metabolic pathways. In this analysis, normalised enrichment scores identified a number of pathways associated with Jmjd1c overexpression such as: o-glycan biosynthesis, extracellular matrix interaction, glycolysis gluconeogenesis pathway, the mitochondrial tricarboxylic acid (TCA) cycle or Krebs cycle, pyruvate metabolism, mTor signalling pathway, pentose phosphate pathway, glycosaminoglycan biosynthesis heparin sulphate pathway, prion disease, and complement and coagulation cascade (Figure 3.13 and Table 3.1). Of particular interest were the metabolic pathways given their well established roles in cancer development and regulation (Schulze and Downward, 2011, Warburg et al., 1927).

A 1 2 0 0 EV ****

) 1 0 0 0 J m jd 1 c

 8 0 0

(

e 6 0 0 b

m ***

u **

n 4 0 0

l l

e ***

C 2 0 0 **** **** * 0 2 3 4 5 6 7 8

W e e k

B EV Jmjd1c

100 µm

Figure 3.12: Transformation of HSC with Jmjd1c (A) Colony forming assays over 8 weeks show an increase in cell number when HSC were retrovirally transduced with Jmjd1c cDNA compared to EV. (B) Representative microscope images of colony morphology illustrate a lack of colonies in EV compared to colonies of Jmjd1c cDNA at week 4. Data represent mean values ± SEM of three independent experiments. Images captured using 20× magnification lens.

Table 3.1: Summary of GSEA results following microarray expression profiling of Jmjd1c overexpression in KLSA9M pre-LSC GSEA target genes from KEGG molecular signature database (Liberzon et al., 2011). False discovery rate (FDR) ≤ 0.25, normalised enrichment score (NES), and nominal P value ≤ 0.01 cut-offs were used. GSEA DETAILS FDR NES KEGG_O_Glycan_Biosynthesis 0.159 1.34 KEGG _ECM_Receptor_Interaction 0.165 1.34 KEGG _Glycolysis_Gluconeogenesis 0.173 1.36 KEGG _Citrate_Cycle_TCA_Cycle 0.167 1.39 KEGG _Pyruvate_Metabolism 0.176 1.41 KEGG _mTor_Signaling_Pathway 0.166 1.45 KEGG _Complement_and_Coagulation_Cascades 0.179 1.45 KEGG _Glycosaminoglycan_Biosynthesis_Heparan_Sulfate 0.177 1.47 KEGG _Pentose_Phosphate_Pathway 0.202 1.47 KEGG _Prion_Diseases 0.202 1.57

A B Glycolysis gluconeogenesis TCA cycle

FDR = 0.166 FDR = 0.173 P = 0.01 P = 0.01 NES = 1.39 NES = 1.36

C D Pyruvate metabolism mTOR signaling pathway

FDR = 0.166 FDR = 0.176 P = 0.01 P = 0.01 NES = 1.45 NES = 1.41

Glycosaminoglycan biosynthesis E Pentose phosphate pathway F Heparansulfate

FDR = 0.202 P = 0.01 FDR = 0.177 NES = 1.47 P = 0.01 NES = 1.47

H G Prion diseases Complement and coagulation cascades

FDR = 0.202 P = 0.01 FDR = 0.179 NES = 1.57 P = 0.01 NES = 1.45

Low High

100

Figure 3.13: Gene set enrichment analysis of microarray data identifies Jmjd1c downstream pathways GSEA in KLSA9M pre-LSC overexpressing Jmjd1c for target genes identified in (A) glycolysis gluconeogenesis, (B) tricarboxylic acid (TCA) cycle, (C) pyruvate metabolism, (D) mTOR signalling pathway, (E) pentose phosphate pathway, (F) glycosaminoglycan biosynthesis and (G) prion diseases. False discovery rate (FDR) ≤ 0.25, normalised enrichment score (NES), and nominal P value ≤ 0.01 cut-offs were used. KEGG molecular signature database was used (Liberzon et al., 2011).

101

The expression of a number of genes involved in cancer-related processes was validated by qRT-PCR. As described in Table 3.2, these genes are highly expressed in AML patient samples, are biomarkers of disease prognosis and regulate metabolic pathways. Phosphoglycerate kinase, isozyme 1 (Pgk1), basic helix-loop-helix domain containing class-B2 (Bhlhb2), pyruvate kinase isoforms M2 (Pkm2), glucose transporter 1 (Glut1/Slc2a1) and hypoxia inducible factor-1a (Hif1a) were significantly upregulated with fold changes greater than 2, whilst aldolase C (AldoC), lectin, galactoside-binding, soluble 3 (Lgals3), solute carrier family 6, member 3 (Slc6a3), neutrophil specific antigen 1 (CD177) and chitinase-3-like protein 1 (Chi3l1) had a greater than 10-fold increase in mRNA levels in Jmjd1c overexpressing cells compared to control cells (Figure 14). The upregulated expression of these genes thus implicates Jmjd1c as an enhancer of leukaemogenesis since Jmjd1c may promote metabolic pathways and chemoresistance, which are considered hallmarks of cancer.

It has previously been shown that Hoxa9 and Meis1, known downstream targets of MLLAF9 (Bernt et al., 2011), are strongly expressed in most AML patients (Golub et al., 1999, Kawagoe et al., 1999), and regulate the haematopoietic progenitor cell antigen, CD34 (Wang et al., 2007c). Flow cytometric analysis showed CD34 expression to increase by 52% in Jmjd1c cDNA and decrease by 38% in shRNA cells (Figures 3.15A and B, respectively), suggesting that CD34 expression is enhanced by Jmjd1c. This finding is consistent with the microarray data described previously.

As discussed in Chapter 1, cancer cell metabolism involving glycolysis and respiration (i.e. the Warburg effect), and the hypoxic environment of the bone marrow niche, play key roles in regulating cancer. The association of Jmjd1c with genes involved in the metabolic pathway and hypoxia was investigated. Protein expression of hypoxia inducible factor-1a and -1b (Hif1a and Hif1b) were analysed by flow cytometry in KLSA9M pre-LSC transduced with EV or Jmjd1c cDNA (Figure 3.15A) as well as KLSMLLAF9 pre-LSC with Scr or Jmjd1c shRNA 4 (Figure 3.15B). Flow cytometric analysis showed an increase in Hif1a and Hif1b protein expression by 24% and 50%, respectively, in Jmjd1c overexpressing KLSA9M pre-LSC (Figure 3.15A). In KLSMLLAF9 pre-LSC, Hif1a decreased by 12% while Hif1b decreased by 10% in Jmjd1c shRNA cells compared to control (Figure 3.15B). Although downregulation of

102

**** 6 0 EV ** 4 0 J m jd 1 c ***

n 2 0

)

i s

e 1 5

s **

g e

n 1 0

r **

p h

x 5

3 ** ** ** ****

d A l ***

o **

N f

( 2

R ** * m 1

0 t a 1 1 2 2 1 a C 3 3 7 l1 m h k k b a f1 o ls a 7 3 g d d g h m 2 i d 6 1 i L l k c l a c h M P P h P l H A g l D B S L S C C

Figure 3.14: Validation of microarray data by qRT-PCR The expression of target genes identified from the microarray of KLSA9M pre-LSC transduced with Jmjd1c cDNA or EV was validated by qRT-PCR. Data represent mean values ± SEM of three independent experiments. Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (* P < 0.05, ** P < 0.005, *** P < 0.0005 **** P < 0.0001).

103

Table 3.2: Brief summary of the function of selected Jmjd1c target genes Function of selected downstream targets identified from the microarray of in KLSA9M pre-LSC with Jmjd1c cDNA or control. Gene Function Reference AldoC -glycogenolysis regulator (Colla et al., 2010, (Aldolase C) -increased expression under hypoxic conditions Walczak-Drzewiecka et -binds to Hif1a al., 2008, Wang et al., -suppression of Hif1a downregulates AldoC in multiple myeloma 2007a) Blhb2/Sharp2 -transcription factor that represses transcription of target genes (Ismail et al., 2008, Jia et (Basic helix-loop-helix -activates mitogen-activated protein kinase (MAPK) pathway and promotes al., 2011, Yamada and domain containing, class proliferation in SW-13 adrenal carcinoma Miyamoto, 2005) B2) -interacts with aryl hydrocarbon receptor nuclear translocator (ARNT) -transcriptionally regulated by PML-RARa (APL) -expression induced by hypoxia CD177 -member of the leukocyte antigen 6 (Ly-6) gene superfamily (Stroncek et al., 2004) (Neutrophil specific -involved in neutrophil proliferation and polycythemia vera antigen 1) -mRNA levels used as a biomarker for diagnosing myeloproliferative disorders Chi3l1 -elevated levels correlate with poor prognosis in patients with cancer and (Libreros et al., 2013, (Chitinase-3-like protein inflammatory diseases Rehli et al., 2003) 1) -induces pro-inflammatory, pro-tumourigenic and angiogenic factors, promoting tumour growth and metastasis -encodes the glycoprotein HC-gp39, which stimulates proliferation

104

Hif1a -associated with chemoresistance in AML (Wang et al., 2011, (Hypoxia inducible -maintains lymphoma cancer stem cells by repressing a negative feedback loop in Zhang et al., 2012, Song factor-1a) Notch pathway et al., 3025) -regulates tumourigenic capacity of glioma stem cells under hypoxic conditions -Hif1a signalling activated in stem cells of mouse lymphoma and human AML under normoxia -required for survival maintenance of CML LSC Ldha -also known as cell proliferation-inducing gene 19 (Ferrara and Mirto, 2003, (Lactate dehydrogenase -serum levels used as an independent prognostic marker in MDS and AML Kornberg and Polliack, A) -LDH levels correlated with WBC in childhood ALL patients and elderly AML 1980, Mani et al., 2006, patients Suarez et al., 1984, Wimazal et al., 2007) Lgals3/Galectin-3 -B-cell precursor ALL and stromal cells communicate through Lgals3 (Cheng et al., 2013, Fei (Lectin, Galactoside- -leukaemia microenvironment induces Lgals3 et al., 2015, Yamamoto- binding, Soluble 3) -promotes drug resistance in CML Sugitani et al., 2011) -high expression is an independent unfavourable prognostic factor in AML patients Mct4/Slc16a3 -mediates lactic acid efflux from tissues dependent on glycolysis for ATP (Ullah et al., 2006, Baek (Monocarboxylate production et al., 2014) transporter 4) -expression upregulated under hypoxic conditions -not activated in cells lacking Hif1a -inhibition suppresses tumour growth

105

-inhibition causes cancer cells to be vulnerable to metabolic stress Mgmt -co-methylated p15/MGMT exhibit shorter overall survival and lower frequency of (Kraguljac Kurtovic et (O(6)-methylguanine- complete remission in AML patients al., 2012, Brandwein et DNA methyltransferase) -high expression attenuate effects of triazenes al., 2014, Bonmassar et -low expression associated with higher response to temozolamide in AML or MDS al., 2013, Su et al., 2012) patients -Azacytidine inhibits proliferation of AML by hypomethylation of MGMT Pdk1 -highly expressed in AML (Pearn et al., 2007, (Pyruvate -promotes PKC-mediated survival of leukaemic blasts Zabkiewicz et al., 2014, dehydrogenase kinase, -expression promoted by Ras in AML Cairns et al., 2011, Hu et isozyme 1) -regulates LSC maintenance in MLLAF9 AML in mice al., 2015) -blocks pyruvate in the TCA cycle which drives the Warburg effect -upregulated by Myc and Hif1 Pgk1 -predictor of poor survival and prognostic biomarker for paclitaxel chemoresistant (Sun et al., 2015, (Phosphoglycerate breast cancer Heddleston et al., 2012, kinase-1) -reduced expression in MLL1 knockdown and Hif1a knockdown Holmquist-Mengelbier et al., 2006) Pkm2 -role in maintenance of aerobic glycolysis in tumour cells (Warburg effect) (Chritofk et al., 2008, (Pyruvate kinase isoform -abundant in AML cell lines and primary AML patient samples Gao et al., 2012, Hitosugi M2) et al., 2009, Sturgill and Guzman, 2013)

106

Slc2a1/Glut1 -facilitates glucose transport (Song et al., 3025) (Glucose transporter 1) -associated with chemoresistance in AML Slc6a3 -dopamine transporter (Paulsson et al., 2006) (Solute carrier family 6, -highly expressed in MLL-rearranged infant ALL resistant to prednisolone member 3) -upregulated in MDS and AML with trisomy 8

107

A B CD34 CD34c d 3 4 s h

1 .8 * 1 .2

1 .5 n 1 .0

s s

s 1 .2 0 .8

e **

p x

x 0 .9 0 .6

i e

e 0 .6 0 .4

r P 0 .3 P 0 .2

0 .0 0 .0 E V Jm jd 1 c S c r s h 4

c o m b o s h + 2 0 1 5 0 6 2 4

D 1 .2 C 1 .8 ** E V

n * J m jd 1 c

1 .5 o

s i

s S c r

s 0 .8 e

s 1 .2 r

e s h 4

p x

0 .9 e

n i

e 0 .4 t

e 0 .6

r P P 0 .3

0 .0 0 .0 H if1 a H if1 b H if1 a H if1 b

Figure 3.15: Validation of Jmjd1c targets by flow cytometry Cell surface staining with CD34 antibody was used to determine protein expression of CD34 in (A) KLSA9M pre-LSC transduced with Jmjd1c cDNA or EV and (B) KLSMLLAF9 pre-LSC transduced with shRNA 4 (sh4) or control (Scr). Protein expression of hypoxia inducible factors, Hif1a and Hifb, was analysed in (C) KLSA9M pre-LSC transduced with Jmjd1c cDNA or EV and (D) KLSMLLAF9 pre-LSC transduced with shRNA 4 or Scr. Data represent mean values ± SEM of three independent experiments. Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (* P < 0.05, ** P < 0.005).

108 protein expression of both Hif1a and Hif1b in KLSMLLAF9 are only slight, the trend observed showed that both Hif proteins tended to increase upon Jmjd1c overexpression, hence the expression of other glycolysis regulators and Hif-regulated targets was examined.

Pkm2, one of the target genes identified from the GSEA enrichment sets (Table 3.1), is highly expressed in AML patient samples (Sturgill and Guzman, 2013) and maintains aerobic glycolysis in tumour cells (Hitosugi et al., 2009, Gao et al., 2012, Christofk et al., 2008). Total Pkm2 expression was shown to be upregulated by Jmjd1c in KLSA9M pre-LSC, which is also consistent with the microarray data (Figure 3.14). The balance between the highly active tetrameric form and nearly inactive dimeric form determines whether glucose is used for biosynthetic processes or glycolytic ATP production. Hence, glycolysis and tumour cell proliferation and survival are controlled by the different isoforms of Pkm2. Phosphorylated-Pkm2 (phospho-Pkm2, Tyr105) has been shown to inhibit active Pkm2 and has been observed in lung cancer and leukaemia cell lines (Hitosugi et al., 2009). Western blot analysis showed that the majority of Pkm2 present in Jmjd1c overexpressing cells is in the phosphorylated form (Figure 3.16A). This observation was consistent in KLSA9M LSC transduced with Jmjd1c cDNA (Figure 3.16B), while Jmjd1c-deficient cells showed a slight decrease in total Pkm2 levels but a larger reduction in phosho-Pkm2 protein expression (Figure 3.16C). The enhanced phospho-Pkm2 expression may explain the proliferative potential and tumour expansion phenotype observed in vitro and in vivo when Jmjd1c expression is altered.

Other members of the glycolysis pathway which are upregulated in proliferating tumour cells and regulated by Hif1 include: AldoC, a glycogenolysis regulator whose expression is upregulated under hypoxic conditions (Colla et al., 2010); pyruvate dehydrogenase kinase, isozyme 1 (Pdk1), whose mechanism of action drives the Warburg effect (Semenza, 2010b); Pgk1, a glycolytic enzyme that catalyses the conversion of 1,3-diphosphoglycerate to 3-phosphoglycerate during glycolysis (Semenza, 2010b); and monocarboxylate transporter 4 (Mct4), a transporter of lactate during enhanced glycolysis (Ullah et al., 2006, Zheng et al., 2015). Western blotting showed that AldoC, Mct4, Pgk1, and Pdk1 expression were all upregulated by Jmjd1c, validating the GSEA data (Figures 16D, E and F). Pdk1 has been previously shown to regulate KLSMLLAF9 LSC maintenance, thus its expression in KLSMLLAF9 pre-LSC

109

A KLSA9M B KLSA9M LSC C KLSMLLAF9 Scr sh4 EV Jmjd1c EV Jmjd1c Phospho-Pkm2 Phospho-Pkm2 Phospho-Pkm2

Pkm2 Pkm2 Pkm2 Actin Actin Actin

D E F KLSA9M KLSA9M KLSA9M EV Jmjd1c EV Jmjd1c EV Jmjd1c

Pgk1 AldoC Pdk1

Actin Mct4 Actin

Actin

G KLSMLLAF9 Scr sh4

Pdk1

Actin

Figure 3.16: Validation of Jmjd1c targets by western blotting Western blot analysis was performed to validate microarray data. Pkm2 is upregulated in (A) KLSA9M pre-LSC and (B) KLSA9M LSC with Jmjd1c cDNA compared to EV, and downregulated in (C) Jmjd1c knockdown in KLSMLLAF9 pre-LSC. (A-C) Phospho-Pkm2 expression was also analysed to determine relative expression compared to total Pkm2. (D) AdoC, (D) Mct4, (E) Pgk1, and (F) Pdk1 expression was also upregulated upon Jmjd1c overexpression in KLSA9M pre-LSC. (G) Pdk1 expression was also shown to be downregulated in shRNA4 KLSMLLAF9 pre-LSC compared to Scr.

110 with Jmjd1c shRNA was examined. Pdk1 expression was confirmed to be downregulated in Jmjd1c shRNA 4-treated KLSMLLAF9 pre-LSC (Figure 3.16G). Pdk1 expression may therefore influence the phenotype observed in vivo, where suppression of Jmjd1c impairs KLSMLLAF9 LSC maintenance.

Since a number of downstream targets of Jmjd1c have critical roles in maintenance or regulation of AML, commercially available inhibitors were used to see whether high Jmjd1c expression would influence the phenotypic effects of the inhibitors by increasing resistance to these drugs. Small molecule inhibitors, shikonin (Cayman Chemicals) and GSK2334470 (Tocris) are specific inhibitors of Pkm2 (Chen et al., 2002b) and Pdk1 (Najafov et al., 2011), respectively. Pharmacological actions attributed to shikonin range from induction of necroptosis, elevation in reactive oxygen species (ROS) and inhibition of adipogenesis by modulation of the Wnt/β-catenin pathway (Han et al., 2007, Lee et al., 2011). The Wnt/β-catenin pathway has been shown to be crucial for AML where MLLAF9 cells have high β-catenin expression (Wang et al., 2010). To determine an optimal range of concentrations for treating AML pre-LSC, alamar blue assays were performed. The results indicated that higher concentrations of shikonin were required to effectively inhibit KLSMLLAF9 growth, which was expected since KLSMLLAF9 AML are more drug resistant, malignant and have enhanced Wnt/β- catenin activity by comparison with KLSA9M AML (Krivtsov et al., 2006). The IC50 values calculated for KLSA9M and KLSMLLAF9 cells were 65 nM and 470 nM, respectively, when these cells were treated with shikonin (Figure 3.17A). These concentrations were used as the basis for colony forming assays, however, the IC50 concentrations were too high to determine its effect on colony formation as the cells were unable to form colonies. This indicates that although these concentrations were required to inhibit cell viability by 50% following treatment for 24 h, lower concentrations of shikonin, starting from 1 nM, were needed to observe impairment in the clonogenicity of these cells when treated for 5 days. Consistent with data from alamar blue assays, colony forming assays showed a 2.1-fold increase in IC50 values when Jmjd1c was overexpressed in KLSA9M cells (IC50 = 58 nM) compared to control cells (IC50 = 27 nM) (Figure 3.17B). Suppression of Jmjd1c also showed a marked sensitisation of KLSMLLAF9 cells (IC50 = 16 nM), where the concentration required to inhibit 50% of colony formation was 2.5-fold lower compared to Scr control cells (IC50 = 40 nM) (Figure 3.17C). Representative microscope images of KLSMLLAF9 pre-LSC 111

T ra n s fo r m o f a la m a r s h ik K + M

A 1 2 5 K S L A 9 M K L S M L L A F 9

) 1 0 0

(

t 7 5

a i

v 5 0

l e

C 2 5

0 0 .0 0 .5 1 .0 1 .5 2 .0 2 .5 3 .0 3 .5 4 .0 L o g [S h ik o n in ] (n M )

T ra n s fo rm o f c d n a s h ik T r a n s fo rm o f s c r s h ik B C 1 2 5 1 2 5

E V S c r

s l

J m jd 1 c l s h 4 l

1 0 0 l 1 0 0

)

t 0

7 5 0 7 5

o y

5 0 y 5 0

(

( o

2 5 o 2 5

C C

0 0 -0 .5 0 .0 0 .5 1 .0 1 .5 2 .0 2 .5 3 .0 -0 .5 0 .0 0 .5 1 .0 1 .5 2 .0 2 .5 3 .0 L o g [S h ik o n in ] (n M ) L o g [S h ik o n in ] (n M )

D Shikonin

Control 20 nM

Scr

sh4 200 µm

Figure 3.17: Pkm2 inhibition by shikonin impairs colony formation (A) Alamar blue cytotoxicity assay showed that KLSA9M pre-LSC were more sensitive to shikonin than KLSMLLAF9 pre-LSC. Colony forming assays using a range of concentrations showed that high expression of Jmjd1c increased chemoresistance to shikonin in (B) KLSA9M and (C) KLSMLLAF9 pre-LSC. (D) Representative microscope images of colony morphology following treatment of KLSMLLAF9 pre-LSC with shRNA 4 (sh4) or scramble (Scr) with 20 nM shikonin or DMSO (control). Data represent mean values ± SEM of three independent experiments. Images captured using 10× magnification lens.

112 treated with shikonin 20 nM showed a reduction in colony size compared to DMSO control, in both Scr and shRNA 4 cells, suggesting that shikonin was able to impair proliferation of KLSMLLAF9 cells within a nanomolar concentration range (Figure 3.17D).

As previously mentioned, GSK2334470 is a novel and highly specific inhibitor of Pdk1, where treatment with GSK2334470 suppressed the activation of SGK, S6K1, RSK and Akt (kinases activated by Pdk1) (Najafov et al., 2011). Treatment with GSK233470 was also shown to delay melanomagenesis and metastasis in BrafV600E::Pten-/- mice (Scortegagna et al., 2014). The assays performed with shikonin were therefore applied to GSK2334470. Alamar blue assays showed that GSK2334470 was not as potent against AML pre-LSC compared to shikonin, however, KLSMLLAF9 were still more drug resistant to treatment (IC50 = 4711 nM), which was 8-fold higher than KLSA9M

(IC50 = 593 nM) (Figure 3.18A). A similar trend of higher IC50 values was observed for cells with high Jmjd1c expression, where KLSA9M overexpressing Jmjd1c were 2.8- fold more resistant to GSK2334470 (Figures 3.18B), while Jmjd1c knockdown cells were 2.2-fold more sensitive to GSK2334470 than control cells (Figures 3.18C). The reduction in colony size and cell density observed in shJmjd1c and Scr KLSMLLAF9 cells following treatment with GSK2334470 are shown in Figure 3.18D. Collectively, the pharmacological inhibition of Pkm2 and Pdk1 by shikonin and GSK2334470, respectively, suggests that they can effectively impair cell viability and colony formation of both KLSA9M and KLSMLLAF9 pre-LSC in the nanomolar concentration range.

From the GSEA data, a number of metabolic pathways were shown to be regulated by Jmjd1c, including glycolysis, TCA cycle, pyruvate metabolism and pentose phosphate pathway (Table 3.2). The dependence of AML cells on various metabolic pathways for energy supply was recently published and showed that quiescent cells differed from rapidly proliferating leukaemic cells (Lagadinou et al., 2013). To examine the metabolic pathways regulating KLSA9M and KLSMLLAF9 AML, ATP production, glycolytic activity and cellular respiration were analysed.

A colorimetric ATP assay used to quantify ATP production demonstrated significantly higher ATP levels in cells with high Jmjd1c expression, where Jmjd1c overexpression increased ATP production by 1.7-fold in KLSA9M (P = 0.016; Figure 3.19A), while

113

T ra n s fo rm o f a la m a r g s k K + M

A 1 2 5 K L S A 9 M

1 0 0 K L S M L L A F 9

)

(

t 7 5

a i

v 5 0

l e

C 2 5

0 0 .5 1 .0 1 .5 2 .0 2 .5 3 .0 3 .5 4 .0 4 .5 L o g [G S K 2 3 3 4 4 7 0 ] (n M )

T r a n s fo rm o f c d n a g s k T r a n s fo rm o f s c r g s k B C 1 2 5 1 2 5

E V S c r

s l

l J m jd 1 c s h 4 l

l 1 0 0 1 0 0

)

t 0

0 7 5 7 5

o o

y 5 0

(

( o

o 2 5 2 5

C C

0 0 -0 .5 0 .0 0 .5 1 .0 1 .5 2 .0 2 .5 3 .0 3 .5 0 .0 0 .5 1 .0 1 .5 2 .0 2 .5 3 .0 3 .5 4 .0 L o g [G S K 2 3 3 4 4 7 0 ] (n M ) L o g [G S K 2 3 3 4 4 7 0 ] (n M )

D GSK2334470

Control 200 nM

Scr

sh4 200 µm

Figure 3.18: Inhibition of Pdk1 by GSK2334470 impairs colony formation (A) Alamar blue cytotoxicity assay showed that KLSA9M pre-LSC were more sensitive to GSK2334470 than KLSMLLAF9 pre-LSC. Colony forming assays using a range of concentrations showed that high expression of Jmjd1c increased chemoresistance to GSK2334470 in (B) KLSA9M and (C) KLSMLLAF9 pre-LSC. (D) Representative microscope images of colony morphology following treatment of KLSMLLAF9 pre-LSC with shRNA 4 (sh4) or scramble (Scr) with 200 nM GSK2334470 or DMSO (control). Data represent mean values ± SEM of three independent experiments. Images captured using 10× magnification lens.

114

A B 2 6 b a tp s h

1 0 * 2 0

8 1 5

)

M M

6 u

( u

1 0 (

****

T T

4 A A 5 2

0 0 S c r s h 4 E V Jm jd 1 c

Figure 3.19: Enhanced Jmjd1c expression is associated with increased ATP production Total cellular ATP production was measured in (A) KLSA9M pre-LSC stably expressing Jmjd1c cDNA or EV, and (B) KLSMLLAF9 pre-LSC stably expressing shRNA control (Scr) or

Jmjd1c shRNA 4 (sh4). Data represent mean values ± SEM of three independent experiments. Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (* P < 0.05, **** P < 0.0001).

115

Jmjd1c knockdown significantly reduced ATP production by 2-fold in KLSMLLAF9 cells (P < 0.0001; Figure 3.19B). This would suggest that AML cells with high Jmjd1c expression, which have increased proliferation potential, generate more ATP to accommodate the metabolic processes required for these proliferating cells.

Luxcel‘s fluorescent assays for determining extracellular acidification rates and oxygen consumption rates were employed to measure cellular glycolytic flux and cellular respiration, respectively. Extracellular acidification rates are illustrated as changes in fluorescence in response to pH over time (Figure 3.20A), where lactate production from glycolytic activity is the main contributor to this acidification. Interestingly, there was a 1.5-fold increase in basal extracellular acidification rates for KLSA9M cells with Jmjd1c cDNA compared to control (P = 0.005; Figure 3.20B). This observation is consistent with previous data which demonstrated that tumour cells implement can implement an abnormally high glucose uptake and lactate production compared to normal cells due to enhanced glycolytic activity (Schulze and Downward, 2011, Vander Heiden et al., 2009, Warburg et al., 1924, Warburg et al., 1927). Mitochondrial function based on oxygen consumption rates, however, showed no significant change when Jmjd1c was overexpressed in KLSA9M (Figures 3.20 C and D). Since basal oxidative phosphorylation rates remained unchanged, these data suggest that enforced Jmjd1c expression depends on glycolysis rather than oxidative phosphorylation to produce ATP required for these cells.

3.10 Discussion

Previous reports have demonstrated that MLLAF9 targets the promoter of Jmjd1c (Bernt et al., 2011) and that JMJD1C expression is aberrantly upregulated in human primary MLL AML (Balgobind et al., 2010, Ross et al., 2014). In this study, Jmjd1c was identified to be the most differentially expressed gene in LSC derived from MLLAF9 compared to the less aggressive KLSA9M LSC and healthy HSC, highlighting the potential clinical relevance for targeting JMJD1C in MLL AML. Clinical data also support the importance of targeting JMJD1C, as AML patients with high JMJD1C expression have an unfavourable prognosis.

116

c d n a

A 6 5 0 B 6 0

E V n o

i ** t

6 0 0 J m jd 1 c a

)

i )

r 4 0



c (

5 5 0 s

(

l l

5 0 0 a

e 2 0

a r

4 5 0 t

x E 0 4 0 0 E V Jm jd 1 c 0 3 0 6 0 9 0 1 2 0 1 5 0 1 8 0 T im e (m in ) C D 1 2 0 E V 8

J m jd 1 c

i )

t 6 s

1 1 0 p

)



(

i o

( 4

t i

1 0 0 n

r g

y 2

x O 9 0 0 0 3 0 6 0 9 0 1 2 0 E V Jm jd 1 c T im e (m in )

Figure 3.20: Enforced Jmjd1c expression enhances glycolytic activity The extracellular acidification rate and oxygen consumption rate of KLSA9M transduced with EV or Jmjd1c cDNA were measured to determine glycolytic and oxidative phosphorylation activity, respectively. (A) Glycolysis was measured with a phosphorescent pH-sensitive probe over 180 min. (B) Data presented as extracellular cellular acidification over time. (C) Oxygen consumption was measured using the phosphorescent oxygen-sensing probe MitoXpress over 120 min. (D) Data presented as oxygen consumption over time. Data represent mean values ± SEM obtained for each time point in three independent experiments. Asterisk indicates statistical significance obtained using an unpaired Student‘s t-test (** P < 0.01).

117

The suppression of Jmjd1c in KLSMLLAF9 resulted in cell cycle progression defects where the block in G1/S transition could explain the impaired colony-forming ability and proliferation rate in vitro and in vivo. Furthermore, shRNA-mediated suppression of Jmjd1c in KLSMLLAF9 achieved a 67% to 72% reduction in protein expression for shRNA 2 and 4, and in vivo studies showed that this reduction was sufficient to impair leukaemia maintenance but not delay leukaemia onset. To further support the role of Jmjd1c as a leukaemia maintenance gene in MLLAF9-AML, the biological effects of silencing Jmjd1c on leukaemogenesis should be examined.

A conditional Jmjd1c knockout AML model was established using mice purchased from the Wellcome Trust Sanger Institute (UK) as it was previously shown that straight Jmjd1c knockout male mice exhibit an age-dependent sterility from the reduction or elimination of germ cells through the regulation of spermatogonia in mouse testis (Kuroki et al., 2013). There are a number transcripts of Jmjd1c (Table S.2) and qRT- PCR primers were designed to detect these protein-coding variants. Significant knockdown in total Jmjd1c expression was observed in both Jmjd1c-/+ and Jmjd1c-/- Cre recombinase cells (Cre), however, a full knockout was not observed in Jmjd1c-/- Cre cells (Figure S.1). A possibility for not observing complete knockout in these cells could be due to low expression of Cre recombinase. Cre was virally transduced into the ex vivo knockout cells due to the time constraints involved in waiting for the litters from crossbreeding Jmjd1c knockout mice with Mx1-Cre mice. Although the Cre-infected cells were selected with antibiotics for puromycin resistance, the antibiotics would select all cells containing the vector but the expression level may not be high. Jmjd1c knockout could be more accurately confirmed using genomic DNA to detect the deletion of the specific exons flanked by the loxP sites. Crossbreeding the Mx1-Cre and Jmjd1c knockout mice would be the method of choice to ensure complete knockout in Jmjd1c-/- cells. Since a full knockout was not observed, these cells were not continued for further experiments.

One approach to achieve effective gene silencing would be to use the prokaryotic type II clustered regularly interspaced short palindromic repeat associated (CRISPR-Cas) protein gene-editing technologies (Jinek et al., 2012, Shalem et al., 2014, Wang et al., 2014b). The CRISPR-Cas9 technology utilises single guide RNA (sgRNA) to bring the Cas9 nuclease to the target locus and cut the target DNA strand. Non-homologous end

118 joining or homologous recombination results in mutations, insertions or deletions that disrupt or knockout a gene. CRISPR-Cas9 genome editing has previously been used in models of myeloid malignancies such as the loss-of-function studies in Flt3-ITD- expressing KLS cells of commonly mutated genes in AML (Tet2, Runx1, Dnmt3a, Ezh2, Nf1 and p53) (Heckl et al., 2014), and drug resistance studies through complete gene knockout of the rate-limiting enzyme involved in the metabolic activation of cytarabine, deoxycytidine kinase (Dck), in cytarabine-resistant AML cell lines (Rathe et al., 2014). The complete loss of Jmjd1c that could be achieved through CRISPR-Cas genome editing in KLSMLLAF9 AML cells may be required to impair leukaemia initiation, which was not able to be achieved by a partial loss of Jmjd1c achieved through shRNA-mediated knockdown, and would clarify the role of Jmjd1c as a leukaemia initiating and/or maintenance gene in KLSMLLAF9 AML.

The oncogenic role of Jmjd1c in AML was also revealed by the enhanced leukaemogenesis of KLSA9M in stably transfected Jmjd1c overexpressing cells where increased infiltration into the spleen and higher abundance of LSC residing within the leukaemic population was observed. These phenotypic effects therefore promote leukaemia development and propagation. Like HSC, LSC engage in bidirectional signals with the microenvironment and localisation of LSC to the niche has been attributed to the protection of LSC from the effects of chemotherapy in part mediated through niche-induced LSC quiescence (Ishikawa et al., 2007). The bone marrow niche has also been proposed to preserve the reconstituting ability of stem cells (Schofield, 1978). Overexpression of Jmjd1c showed an enhanced ability for KLSA9M cells to home to the bone marrow following transplantation in mice, where they were also shown to proliferate at a higher rate within the bone marrow, reflecting the enhanced leukaemogenic properties of these cells. Whether these cells home to and are protected by the niche was not confirmed in this study, but it would be interesting to determine the location in which these cells may engraft and home to the bone marrow.

It has previously been shown that overexpression of Meis1a alone fails to transform haematopoietic cells, however, coexpression of both Hoxa9 and Meis1a could transform HSC to induce AML in vivo (Kroon et al., 1998). Although Jmjd1c alone is not sufficient to fully transform HSC into pre-LSC, the enforced expression of Jmjd1c is able to confer a proliferative advantage to HSC, and extend the longevity of these cells

119 in vitro, displaying a phenotype that is similar to that caused by Meis1a (Calvo et al., 2001, Cai et al., 2012).

Evidence is accumulating showing aberrant upregulation of glycolysis as a cancer phenotype observed in primary and metastatic cancers, where it confers growth advantages by promoting unconstrained proliferation and invasion (Gatenby and Gillies, 2004). In support of these findings, gene expression profiling identified a number of metabolic pathways regulated by Jmjd1c, including glycolysis. The expression levels of key players of glycolysis were highly upregulated in association with Jmjd1c overexpression. Pkm2 is one of the main contributors to cancer metabolism and the replacement of Tyr105 (phospho-Pkm2) by phenylalanine results in reduced cell proliferation during hypoxia as well as reduced tumour growth in xenografts (Hitosugi et al., 2009). Phospho-Pkm2 is therefore critical for the metabolism of cancer cells to promote the Warburg effect, which describes a pro-oncogenic metabolic switch. The ability for Jmjd1c to induce the expression of phospho-Pkm2 is integral in determining the metabolic switch of AML LSC and conferring growth advantages to these cells.

The role of Jmjd1c in regulating cellular metabolism was further examined by measuring ATP production, glycolysis and oxidative phosphorylation, where the increased basal ATP levels in Jmjd1c overexpressed cells was attributed to an enhanced rate of glycolysis. These results suggest that Jmjd1c promotes tumour proliferation through the promotion of the glycolytic pathway. In support of these findings, a recent study found that metabolic processes and Myb depletion gene sets were enriched in Jmjd1c knockdown targets (Sroczynska et al., 2014). Although they did not examine the role of Jmjd1c as a metabolic regulator, they found that overexpression of Myb and Myc, the transcriptional target of Myb, could partially rescue the Jmjd1c shRNA- mediated phenotype. The interaction of Myc and Hif has been shown to confer metabolic advantages to tumour cells (Dang et al., 2008), in addition to the fact that they are both involved in glycolysis, further supporting the role of Jmjd1c as a glycolysis regulator. Furthermore, treatment with commercially available inhibitors targeting key players in glycolysis, Pkm2 and Pdk1, showed that Jmjd1c overexpressing cells were highly resistant to these inhibitors, suggesting a protective role Jmjd1c in leukaemic cells. The inhibitor concentration sufficient to impair the proliferation and clonogenicity of control KLSA9M cells only partially impairs Jmjd1c overexpressing

120 cells as there is an abnormally high activity of glycolysis in these cells. These findings highlight a novel role for Jmjd1c in regulating the glycolytic pathway in KLSA9M AML. One mechanism of action of shikonin is the modulation of the Wnt/β-catenin pathway (Han et al., 2007, Lee et al., 2011). Higher concentrations of shikonin are therefore required to block the glycolysis pathway and impair the growth of KLSMLLAF9 cells as they have high levels of Jmjd1c and β-catenin, where Jmjd1c enhances glycolysis. Since the suppression of Jmjd1c did not result in changes in Myb or Myc expression, and the overexpression of Myb or Myc only partially rescues the Jmjd1c-deficient phenotype (Sroczynska et al., 2014), ectopic expression of genes such as Hif1a, Pdk1, Pkm2 or Pgk1, which are key players in glycolysis may have more of an effect in fully rescuing the phenotype of Jmjd1c-depleted cells. Efforts to determine the phenotype of overexpressing these glycolytic genes in shJmjd1c KLSMLLAF9 cells are currently in progress.

Hif1 not only plays a critical role in glycolysis, but is also the master regulator of hypoxia where Hif function is primary regulated by levels of Hif1a (Keith and Simon, 2007). Studies have shown areas within the bone marrow niche of AML patients are hypoxic, varying from 1% to 6% (Fiegl et al., 2009). It was also shown that under hypoxic conditions (1% O2), Hif1 expression in primary AML cells was increased compared to atmospheric (21% O2) conditions (Hatfield et al., 2010). Since AML cells with high Jmjd1c expression adopt high glycolysis pathway activity and therefore produce more lactate, the location and microenvironment in which these cells engraft to the bone marrow may be examined to determine whether these areas are hypoxic in a similar manner to leukaemic bone marrow niche.

Hif1 signalling was also selectively activated in human AML LSC (Wang et al., 2011) and shown to be required for the maintenance of T-ALL LSC (Giambra et al., 2015). The suppression of Jmjd1c was found to alter the maintenance of AML LSC and this may, in part, be due to the regulation of Hif1 signalling by JMJD1C. There are several mechanisms by which Hif signalling and hypoxia may alter histone methylation and vary depending on oxygen availability, since all 2-oxoglutarate-dependent oxygenases depend on dioxygen as a co-substrate. Studies have identified several 2-oxoglutarate- dependent oxygenases as hypoxia-inducible gene products (Metzen et al., 2005, Takahashi et al., 2000, Hofbauer et al., 2003, Pescador et al., 2005) and microarray data

121 of MCF7 cells following exposure to hypoxia or treatment with dimethyloxalylglycine (DMOG), a prolyl-4-hydroxylase inhibitor, shown to competitively inhibit a number of 2-oxoglutarate oxygenases that regulate HIF (Epstein et al., 2001), identified JMJD1A as the most upregulated transcript (Pollard et al., 2008). In this study, JMJD2B (3-fold) and JMJD1C (2-fold) were the second and third most highly regulated JMJD by HIF signalling (Pollard et al., 2008). This finding raises an interesting question of whether the JMJD upregulated by HIF identified in this study would show a similar response in hypoxia- or DMOG-treated leukaemic cells.

The potential therapeutic value of targeting Jmjd1c is exemplified by the aberrant activation of glycolysis in AML LSC, inherited proliferative advantages in KLSA9M and HSC upon enforced expression of Jmjd1c, as well as the enhanced leukaemogenesis, where these cells adopted an increased ability to engraft and proliferate in the bone marrow. It would also be of interest to investigate whether these findings are cell type specific or whether they can be translated to other tissues or tumours with abnormal JMJD1C expression, such as diffuse-type gastric cancer (Katoh and Katoh, 2007) and autism (Saez et al., 2015). These novel findings highlight the potential importance of targeting this demethylase as a form of LSC-targeted therapy to inhibit its ability to facilitate efficient leukaemia development and progression. The combination of JMJD1C inhibitors with conventional chemotherapy or glycolysis inhibitors also may have the potential to improve the survival of patients with AML.

122

CHAPTER 4

The Role of JMJD5 in Regulating MLL-Rearranged Acute Myeloid Leukaemia

4.1 Introduction

Dysregulation of the epigenome plays a pivotal role in facilitating the transformation of leukaemic cells and regulating stem cell function. In Chapter 3, Jmjd1c was demonstrated to regulate specific metabolic pathways and enhance the leukaemogenicity of KLSA9M LSC, as well as being required for the maintenance of KLSMLLAF9-derived AML. As histone demethylases can display either oncogenic or tumour suppressive functions, gene expression profiling of HSC and LSC would assist in identifying a histone demethylase which can function as a tumour suppressor by targeting signalling pathways required for stem cell function.

This chapter aims to explore the role of JMJD5, a novel histone demethylase that has not been characterised in depth in cancer. Jmjd5 was identified to be highly expressed in murine HSC compared to KLSMLLAF9 LSC in this study. Previous studies found JMJD5 to be highly expressed in breast cancer (Hsia et al., 2010) and colon cancer cell lines (Zhang et al., 2015). These findings are a departure from other studies which considered Jmjd5 as a tumour suppressor. For example, Jmjd5 was reported to be required for normal embryonic growth (Ishimura et al., 2012, Oh and Janknecht, 2012), while studies based on retrovirus insertional mutagenesis in Blm-deficient B-cell lymphoma also identified JMJD5 as a tumour suppressor (Suzuki et al., 2006). The biological role and molecular characterisation of JMJD5 have yet to be examined in the context of AML. Analysis of clinical data would assist in determining whether JMJD5 may act as a tumour suppressor or oncogene. The effect of reversing the aberrant expression of JMJD5 in KLSMLLAF9 LSC on leukaemogenesis was investigated.

123

4.2 Jmjd5 expression is suppressed in KLSMLLAF9 AML cells

The histone demethylase family includes a number of attractive candidate target genes that regulate LSC properties via transcriptional effects. Concentrating only on histone demethylases, microarray analysis comparing HSC and KLSMLLAF9 LSC identified Jmjd1c as the most highly expressed histone demethylase in KLSMLLAF9 LSC, while Jmjd5 was the most highly expressed histone demethylase in HSC (fold change > 1.5 and P < 0.05; Figure 4.1A). The microarray data were validated using qRT-PCR and western blotting, where Jmjd5 expression was significantly decreased at both the mRNA (9.2-fold, P = 0.0014; Figure 4.1B) and protein level (Figure 4.1C) in KLSMLLAF9 LSC compared to HSC. AML patients with high JMJD5 expression had a significantly better prognosis and survival rate than those with low JMJD5 expression (P = 0.0034; Figure 4.1D) supporting the hypothesis that JMJD5 may function as a tumour suppressor in AML.

A recent report showed that DOT1L, the H3K79 methyltransferase, is required for MLLAF9 induced AML, and DOT1L-mediated H3K79me2 modifications regulate the expression of MLLAF9 target genes (Bernt et al., 2011). Interestingly, Ezh2, an epigenetic regulator that plays a crucial role in maintenance of MLL AML, and Jmjd5 have been identified as potential targets of H3K79me2. Given the important role of DOT1L in MLL leukaemogenesis and to determine if Jmjd5 and Ezh2 are regulated by H3K79me2, KLSA9M and KLSMLLAF9 pre-LSC were treated with SGC0946, a specific DOT1L-inhibitor. SGC0946 is a brominated analogue of EPZ004777, a potent DOT1L inhibitor currently under investigation to enter human clinical trials (Daigle et al., 2013). Colony forming assays revealed that KLSMLLAF9 cells were more sensitive to SGC0946 than KLSA9M (Figure 4.2A) which is consistent with previous reports showing the selectivity of SGC0946 towards MLL leukaemic cells (Yu et al., 2013). Following treatment with SGC0946, western blotting showed that Ezh2 expression was reduced, while Jmjd5 expression was upregulated compared to the control in KLSMLLAF9 pre-LSC (Figure 4.2B). Collectively, these data implicate Jmjd5 as having a potential tumour suppressor function, whose expression is suppressed in AML cells which develop aggressive AML in vivo. The role of Ezh2 is discussed further in the conclusion of this chapter, as well as in Chapter 5.

124

LSC HSC KLSMLLAF9 A B 1 2 3 1 2 3 4 1 2 Top 5: n

o H S C

i s

s L S C e

Jmjd1c r 9

x e Phf8

A 6 N

Jmjd2c R

Jmjd1b v

i 3

t **

a l

Jmjd6 e R 0 H S C L S C Top 3: Jmjd5 Jarid1c Jarid1b C LSC HSC KLSMLLAF9

Jmjd5

Actin

High Low P < 0.05; Fold > 1.5

D P = 0.0034

D 0.8

High Survival

0.4 (N = 13) probability Low

(N = 21) 0.0

0 200 400 600 800 1000 Days

Figure 4.1: Jmjd5 is downregulated in aggressive AML where low expression predicts poor clinical outcome (A) Microarray heatmap showing the top upregulated and downregulated histone demethylase genes in HSC-enriched population compared to LSC-enriched population isolated from KLSMLLAF9 (P < 0.05, fold change > 1.5). Jmjd5 was identified as the most downregulated gene in KLSMLLAF9 LSC. Jmjd5 expression was confirmed by (B) qRT-PCR and (C) Western blotting. Data represent mean values ± SEM of three independent experiments. Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (** P < 0.005). (D) Survival data from AML patients show that patients with low Jmjd5 expression have a 125 significantly poorer survival compared to patients with high Jmjd5 expression (P = 0.0034. Survival data obtained from PrognoScan, a publicly available database for evaluating patient prognosis).

126

A 1 2 5

K L S A 9M s

l K L S M L L A F 9

l 1 0 0

)

0 t

0 7 5

y 5 0

l (

o 2 5 C

0 -1 0 1 2 3 4 L o g [S G C 0 9 4 6 ] n M

B KLSMLLAF9

Control 1 µm

Jmjd5

Ezh2

Actin

Figure 4.2: Jmjd5 is highly expressed in DOT1L inhibitor-treated KLSMLLAF9 (A) Pharmacological inhibition of DOT1L with SGC0946 selectively reduces the colony forming potential of KLSMLLAF9 and not KLSA9M pre-LSC. (B) Western blot analysis depicting a reduction in Ezh2 and an increase Jmjd5 expression upon DOT1L inhibition in KLSMLLAF9 pre-LSC. Data represent mean values ± SEM of three independent experiments.

127

4.3 Overexpression of Jmjd5 impairs KLSMLLAF9 colony formation in vitro

To determine whether Jmjd5 may impair the leukaemic properties of AML cells, KLSMLLAF9 pre-LSC, which have low Jmjd5 expression, were transduced with Jmjd5 cDNA or control vector. Transduction efficiency was confirmed using qRT-PCR and western blotting, where Jmjd5 expression was shown to be significantly upregulated at both the mRNA (7.1-fold increase, P = 0.0158; Figure 4.3A) and protein level (Figure 4.3B). To examine the phenotypic properties of Jmjd5 in KLSMLLAF9, colony forming assays were performed. A significant reduction in both colony (2.5-fold, P = 0.002; Figure 4.3C) and cell (2-fold, P = 0.0126; Figure 4.3D) numbers was observed in cells overexpressing Jmjd5 compared to control. This is supported by the representative colony images in Figure 4.3E, which show a marked decrease in colony size in Jmjd5 overexpressed cells compared to control.

To explain the mechanism in which Jmjd5 may impair colony forming capacity and proliferative potential of KLSMLLAF9 cells in vitro, cell cycle and apoptosis analyses were performed. PI staining showed no significant change in G1, S or G2/M phases between cells overexpressing Jmjd5 and control cells (Figure 4.4C), however, staining with Annexin V and 7AAD showed a significant increase in an early apoptosis population (Annexin V positive, 7AAD negative; 1.7-fold, P = 0.02; Figures 4.4A and B). Overexpression of Jmjd5 in KLSMLLAF9 also generated a number of differentiated myeloid cells, as illustrated in the representative Wright-Giemsa staining panels (Figure 4.4D). In contrast, control cells consisted of only blast cells. To further confirm Jmjd5 as a regulator of differentiation, the expression of lineage markers for macrophages (Mac-1/CD11b) and granulocytes (Gr-1/Ly-6G) were examined by flow cytometry. The expression of Mac-1 and Gr-1 was significantly upregulated by 2-fold (P = 0.0004) and 1.7-fold (P = 0.0268), respectively, in Jmjd5 overexpressing cells compared to control (Figures 4.4E and F), supporting the result observed following staining of intracellular content with Wright-Giemsa dyes. Collectively, these results suggest that Jmjd5 may be involved in the regulation of apoptosis and induction of differentiation in KLSMLLAF9, thereby causing a reduction in proliferation and colony forming potential.

128

A B KLSMLLAF9 1 0 n C o n tro l

o * i

s Control Jmjd5

s 8 J m jd 5

x e

6 Jmjd5

A N

R 4

e Actin

v i

t 2

e R 0 C o n tro l Jm jd 5

C D 1 5 0 2 .5

s C o n tro l

) e

6 2 .0 J m jd 5

1 0

1 0 0



( 0

1 .5 **

/ **

s m

e 1 .0 i

5 0 u

l e

o 0 .5

C C

0 0 .0 C o n tro l Jm jd 5 C o n tro l Jm jd 5

H S C M L L H S C M L L

E Control Jmjd5

200 µm

Figure 4.3: Jmjd5 impairs colony forming ability of KLSMLLAF9 in vitro KLSMLLAF9 pre-LSC were transduced with Jmjd5 cDNA or control. Following FACS, Jmjd5 overexpression was validated by (A) qRT-PCR and (B) western blotting. Colony forming assays showed a significant decrease in (C) colony number and (D) cell number in KLSMLLAF9 transduced with Jmjd5 cDNA compared to control. (E) Representative microscope images of colony morphology from colony forming assays illustrate a reduction in colony size in KLSMLLAF9 with Jmjd5 cDNA or control. Data represent mean values ± SEM of three independent experiments. Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (* P < 0.05, ** P < 0.005). Images captured using 10× magnification lens.

129

A Control Jmjd5 7AAD

Annexin V 4 e p i p r e B 4 d p r e a p o p C 6 0 3 0 C o n tro l

J m jd 5 )

* )

(

s 4 0 n

l 2 0

t u

p 1 0 2 0

A A

0 0 C o n tro l Jm jd 5 G 1 S G 2 /M

Jmjd5 D Control Jmjd5

20 µm

Control Jmjd5 Control Jmjd5

Max (%) Max Max (%) Max

Mac-1 Gr-1

F 2 .5 C o n tro l

n 2 .0 *** J m jd 5

o i

s *

s e

r 1 .5

n 1 .0

t o

r 0 .5 P

0 .0 M a c -1 G r-1 130

Figure 4.4: Jmjd5 induces apoptosis and differentiation in KLSMLLAF9 (A) Representative scatter plots and (B) bar graphs following Annexin V and 7-aminoactinomycin (7AAD) staining show that KLSMLLAF9 overexpressing Jmjd5 are induced to enter early apoptosis compared to control. (C) Flow cytometric analysis of propidium iodide (PI) staining showed no difference in the distribution of cells in different phases of cell cycle following enforced Jmjd5 expression. (D) Wright-Giemsa staining illustrates a mixture of blast and differentiated cells in Jmjd5 overexpressing KLSMLLAF9 pre-LSC. (E) Representative histograms and (F) bar graphs show an increase in protein expression of lineage markers for macrophages (Mac-1/CD11b) and granulocytes (Gr-1/Ly-6G) determined by flow cytometry. (F) Increased Jmjd5 expression correlated with a significant increase in Mac-1 and Gr-1 expression. Data represent mean values ± SEM of three independent experiments. Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (* P < 0.05, *** P < 0.0005). Images captured using 100× magnification lens.

131

4.4 Jmjd5 impairs leukaemogenesis

Since the overexpression of Jmjd5 impaired the proliferation and clonogenic growth of KLSMLLAF9, Jmjd5 was hypothesised to have a role in leukaemia initiation and maintenance. To assess this, KLSMLLAF9 pre-LSC transduced with either Jmjd5 or control cDNA were injected into sublethally irradiated mice (Figure 4.5A). The survival outcome of these mice demonstrated the tumour suppressor effect of Jmjd5, where overexpression of this demethylase significantly extended the latency of KLSMLLAF9- leukaemia compared to control (P < 0.0001; Figure 4.5A). Assessment of leukaemic infiltration at time of sacrifice showed a 2.4-fold reduction in bone marrow infiltration (P = 0.0051) and an 18-fold reduction in spleen infiltration (P = 0.1049) (Figure 4.5B). Although the reduction of leukaemic cell infiltration in the spleen was not statistically significant due to the large spread and few data points for the Jmjd5 cohort, the reduction in leukaemic burden was confirmed by the significant reduction in spleen weight (2.5-fold, P = 0.03; Figure 4.5C). Furthermore, the lower disease burden was also reflected in the peripheral blood, where there was a 1.7-fold decrease in white blood cells in mice with Jmjd5 cDNA compared to control (P = 0.04; Figure 4.5D), further supporting the hypothesis that Jmjd5 acts as a tumour suppressor in MLLAF9- AML.

To determine whether the suppression of Jmjd5 was required to maintain fully established MLL AML, 1×105 cells from the bone marrow of leukaemic mice were inoculated into secondary recipient mice and monitored for disease onset. Mice injected with KLSMLLAF9 LSC overexpressing Jmjd5 were shown to have a significant delay in disease latency compared to control, suggesting a regulatory role in leukaemia maintenance (P = 0.00333; Figure 4.6A). This observed delay in leukaemia development could be in part due to the 10% reduction in proliferation rate in Jmjd5 overexpressing cells compared to control cells as demonstrated by BrdU staining (P = 0.047; Figure 4.7B). There was also a reduction in LSC frequency in the Jmjd5 leukaemic population (Figure 4.6C and Table S.3), which was determined by limiting dilution assays, where mice injected with 1 × 104 and 1 × 103 cells all developed leukaemia in the control cohort, whereas only two mice injected with 1 × 104 cells and none of the mice injected with 1 × 103 cells from the Jmjd5 cohort succumbed to AML.

132

A 1 0 0 C o n tro l

J m jd 5 l

a 7 5

t 5 0

c r

e 2 5 P

0 0 5 0 1 0 0 1 5 0 2 0 0 2 5 0 3 0 0 D a y s p o s t tra n s p la n ta tio n

B 1 2 5 **

) l

a 1 0 0

% 7 5

(

l e

c 5 0

+ P

F 2 5 G

0 C o n tro l Jm jd 5 C o n tro l Jm jd 5

BM Spleen C * D *

0 .6 0 .6

) )

g g

( (

0 .4 t 0 .4

h h

g g

i i

e e

w w

n n

e e e

0 .2 e 0 .2

l l

p p

S S

0 .0 0 .0 C o n tro l Jm jd 5 C o n tro l Jm jd 5

Figure 4.5: Overexpression of Jmjd5 impairs leukaemia initiation (A) Survival curves of primary recipient mice following transplantation of KLSMLLAF9 pre- LSC transduced with control (N = 14) or Jmjd5 cDNA (N = 12). (B) Leukaemic infiltration was determined by assessing GFP-positive population in the bone marrow and spleen. (C) Spleen weight and (D) white blood cell (WBC) counts were reduced in mice transplanted with Jmjd5 cDNA compared to control. Horizontal line represents the median. Each data point represents one mouse. Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (* P < 0.05, ** P < 0.005). 133

1 0 0 1 0 0 k C o n tro l

1 0 0 k J m jd 5 l

a 7 5

t 5 0

c r

e 2 5 P

0 0 3 0 6 0 9 0 1 2 0 1 5 0 D a y s p o s t tra n s p la n ta tio n

B * 4 0

3 0

+ 2 0

r B 1 0

0 C o n tro l Jm jd 5

C 1 0 0 l

a 7 5

t 5 0

n 1 0 k C o n tr o l

e c r 1 k C o n tr o l e 2 5 P 1 0 k J m jd 5

1 k J m jd 5 0 0 3 0 6 0 9 0 1 2 0 1 5 0 D a y s p o s t tra n s p la n ta tio n

Figure 4.6: Overexpression of Jmjd5 impairs leukaemia maintenance (A) Survival curves of secondary recipient mice show Jmjd5 overexpression (N = 7) causes a delay in leukaemia onset following transplantation of 1 × 105 KLSMLLAF9 LSC compared to control (N = 7). (B) Dot plots illustrate a significant decrease in BrdU incorporation in Jmjd5 134 overexpressed KLSMLLAF9 LSC. (C) Limiting dilution assays in secondary recipient mice reveal a reduction in LSC frequency upon enforced Jmjd5 expression (N = 6) compared to control (N = 6). Horizontal line represents the median. Each data point represents one mouse. Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (* P < 0.05).

135

These in vivo studies highlight the tumour suppressor potential of Jmjd5, where restoring the expression of the aberrantly downregulated Jmjd5 in KLSMLLAF9 impaired leukaemia initiation and maintenance, with a reduction in leukaemic infiltration and cell proliferation in vivo.

4.5 KLSMLLAF9 cells are more sensitive to doxorubicin treatment upon enforced Jmjd5 expression

To identify downstream targets of Jmjd5, microarray analysis was performed on KLSMLLAF9 pre-LSC transduced with Jmjd5 or control cDNA. Gene expression analysis found 107 genes upregulated and 22 genes downregulated by Jmjd5 in KLSMLLAF9 (P < 0.05, fold change > 2; Figure 4.7). A brief summary of the known function of the top downregulated and upregulated candidate target genes identified by mRNA expression profiling is described in Table 4.1. These top differentially expressed genes were selected due to their function or regulation of pathways which may be critical for AML LSC phenotype. Western blotting validated the microarray data, where H2.0-like homeobox 1 (Hlx1), G-protein coupled receptor 84 (Gpr84), Cut-like homeobox 2 (Cux2), carbonyl reductase 1 (Cbr1), transcription factor 7-like 2 (Tcf7l2) and B-cell lymphoma 2-related protein A1 (Bcl2a1) were downregulated following Jmjd5 overexpression, while cyclin D2 (Ccnd2) and membrane-spanning 4-domain, subfamily A, member 3 (Ms4a3) expression were upregulated by comparison with control cells (Figure 4.8). Hlx1, Gpr84, Cux2 have all been previously shown to be highly expressed in myeloid disorders, while Cbr1, the predominant regulator of daunorubicin metabolism, also contributes to chemotherapy-resistance in AML (Table 4.1).

Since Cbr1 is responsible for converting anthracyclines such as doxorubicin, a common drug used for curative treatment in AML (Burnett, 2012), to a less active compound (Kassner et al., 2008, Varatharajan et al., 2012), and Cbr1 expression was reduced upon enforced Jmjd5 expression in KLSMLLAF9 pre-LSC (Figure 4.7), alamar blue assays were conducted. IC50 values determined from these cytotoxicity assays were 2.2-fold

136

Control Jmjd5

1 2 3 1 2 3 Downregulated: Hlx1 Gpr84 Cutl2/Cux2 Cbr1 Tcf7l2 Bcl2a1c

107 genes

Upregulated: Ccnd2 22 genes Ms4a3

High Low

P < 0.05; Fold > 2

Figure 4.7: Heatmap depicting Jmjd5 target genes Microarray heatmap showing the most differentially expressed genes in KLSMLLAF9 pre-LSC with Jmjd5 cDNA or control using an Illumina platform. 107 genes were downregulated and 22 genes were upregulated by Jmjd5 (P < 0.05, fold change > 2).

137

Table 4.1: Brief summary of the function of selected Jmjd5 target genes Function of selected downstream targets identified from the microarray comparing KLSMLLAF9 pre-LSC transduced with Jmjd5 cDNA or control. Gene Function Reference Hlx1 -transcription factor essential for haematopoietic development (Jawad et al., 2006, (H2.0-like homeobox 1) -high levels in CD34 bone marrow cells but not granulocytes or macrophages Kennedy et al., 1994, -involved in proliferation Deguchi et al., 1992) -increases the risk of therapy-related AML (t-AML)

Gpr84 -upregulated in human MLL-rearranged AML compared to normal HSC (Saito et al., 2010, (G-protein coupled -regulates β-catenin signalling Kohlmann et al., 2008, receptor 84) -required for MLLAF9-AML maintenance Dietrich et al., 2014) Cux2/Cutl2 -transcription factor affecting proliferation of neuronal progenitors of the (Stegelmann et al., 2010, (Cut-like homeobox 2) subventricular zone Cubelos et al., 2008, -expression regulated by Notch signalling Iulianella et al., 2009) -highly expressed in myeloproliferative neoplasms Cbr1 -predominant doxorubicin reductase (Lal et al., 2008, Kassner (Carbonyl reductase 1) -highly expressed in prostate, lung and endometrial cancer et al., 2008, Varatharajan -AML patients with high expression are at risk of developing resistance to et al., 2012, Jang et al., daunorubicin 2012) -overexpression protects APL cells against arsenic trioxide via generation of ROS

138

Tcf7l2/ Tcf4 -forms a complex with β-catenin in the nucleus, which inhibits differentiation (de Wetering et al., 2002, (Transcription factor 7- through its control over c-MYC and p21CIP1/WAF1 in colorectal cancer Kendziorra et al., 2010) like 2) -main downstream effector of the Wnt signalling pathway -overexpressed in resistant rectal adenocarcinoma tumours Bcl2a1c -member of the anti-apoptotic Bcl2 family, frequently overexpressed during (Jenal et al., 2010, (B-cell lymphoma 2- tumourigenesis, including AML, CML and ALL patient samples Lagadinou et al., 2013, related protein A1) -transcriptional target of PU.1 (key player in myeloid and Β-cell development) Kirkin et al., 2004, Taylor -high expression causes resistance to etoposide, staurosporine, doxorubicin and et al., 2000, Sartorius and cisplatin Krammer, 2002) -inhibition selectively eradicates quiescent human LSC

Ccnd2 -controls G1/S transition (Camos et al., 2006, (Cyclin D2) -decreased expression in t(8;16) and t(11q23)/MLL AML Haferlach et al., 2009, -repressed in quiescent cells Bouchard et al., 1999) Ms4a3/Htm4 -cell surface signalling molecule (Donato et al., 2002, (Membrane-spanning 4- -regulates G1/S phase Chinami et al., 2005) domains, subfamily A, -expression restricted to haematopoietic lineage Member 3 -transcriptional repression by Ecotropic Virus Integration site 1 (EVI1) promotes tumour growth

139

KLSMLLAF9

Control Jmjd5

Jmjd5

Hlx1

Gpr84

Cux2

Cbr1

Tcf7l2

Bcl2a1

Ccnd2

Ms4a3

Actin

Figure 4.8: Validation of Jmjd5 targets by western blotting Protein expression of Jmjd5, Hlx1, Gpr84, Cux2, Cbr1, Tcf7l2, Bcl2a1, Ccnd2 and Ms4a3 was examined by western blotting, where actin was used as a loading control.

140 higher in control (IC50 = 11.5 ng/mL) compared to Jmjd5 overexpressing cells (IC50 = 5.2 ng/mL) (Figure 4.9A). In vitro proliferation assays were also performed where the cell number was monitored following 48 h treatment with doxorubicin. The hypothesis that Jmjd5 sensitises KLSMLLAF9 cells to doxorubicin was supported by these proliferation assays where doxorubicin at 0.5 ng/mL was efficient in significantly impairing the proliferation of Jmjd5 overexpressing cells after 48 h (P = 0.0028), whereas control cells required double the concentration to achieve a similar reduction in cell number (P = 0.0272; Figure 4.9B). These assays demonstrate that enforced Jmjd5 expression can sensitise KLSMLLAF9 pre-LSC to doxorubicin.

4.6 Downstream pathways regulated by Jmjd5

To further investigate the pathways regulated by Jmjd5, the relationship between Jmjd5 and Gpr84 was assessed. KLSMLLAF9 pre-LSC were transduced with Gpr84 shRNA or a non-targeting shRNA (Scr), where the shRNA used has been described previously (Dietrich et al., 2014). Western blotting confirmed suppression of Gpr84 expression in Gpr84 shRNA cells, which was associated with a reduction in β-catenin expression (Figure 4.10). This supports recently published data by our laboratory, that Gpr84 regulates the β-catenin signalling pathway (Dietrich et al., 2014). Interestingly, the shRNA-mediated suppression of Gpr84 resulted in the upregulation of Jmjd5 expression (Figure 4.10). These findings suggest that Jmjd5 and Gpr84 may regulate each other in a negative feedback loop.

The role of Jmjd5 in negatively regulating Gpr84 signalling was examined, where qRT- PCR was used to analyse the expression of common downstream targets of Gpr84 and β-catenin identified previously (Dietrich et al., 2014). A comparison of the microarray data from Gpr84 overexpressing cells with that of the published Wnt/β-catenin datasets, demonstrated that the most highly enriched genes identified by GSEA were also significantly downregulated by Jmjd5 in KLSMLLAF9 (P ≤ 0.02; Figure 4.11). These genes included the Rho family member, Ras homolog family member U (Rhou), transmembrane protein, stimulated by retinoic acid 6 (Stra6), paired-like homeodomain 2 (Pitx2), plasminogen activator, urokinase receptor (Plaur), Iroquois homeobox gene (Irx3), matrix metallopeptidase 9 (Mmp9) and cyclin D1 (Ccnd1) (Dietrich et al., 2014).

141

T ra n s fo rm o f 5 a a la m a r d o x A 1 2 5 C o n tr o l J m jd 5

1 0 0

)

(

t 7 5

a i

v 5 0

l e

C 2 5

0 -4 -3 -2 -1 0 1 L o g [D o x o ru b ic in ] ( g /m L )

1 2 5

) 0 n g /m L d o x o ru b ic in

% (

0 .1 n g /m L d o x o ru b ic in

r 1 0 0 e

b 0 .5 n g /m L d o x o ru b ic in

m 7 5 *

u ** 1 n g /m L d o x o ru b ic in

l **

l e

c 5 0

i t

a 2 5

e R 0 C o n tro l Jm jd 5

K L S M L L A F 9

Figure 4.9: Jmjd5 sensitises KLSMLLAF9 to doxorubicin (A) Alamar blue cytotoxicity assays showed that Jmjd5 overexpression sensitises KLSMLLAF9 pre-LSC to doxorubicin compared to control. (B) Proliferation assays measuring cell number at 48 h showed increased sensitisation in Jmjd5 overexpressing KLSMLLAF9 cells compared to control when treated with doxorubicin. Data represent mean values ± SEM of three independent experiments relative to control treatment. Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (* P < 0.05, ** P < 0.005, *** P < 0.0005).

142

KLSMLLAF9

Scr Gpr84 shRNA

Gpr84

β-catenin

Jmjd5

Actin

Figure 4.10: Suppression of Gpr84 correlates with reduced β-catenin and increased Jmjd5 expression KLSMLLAF9 pre-LSC were transduced with Gpr84 shRNA or shRNA-control (scramble/Scr). Western blotting confirmed a reduction in Gpr84 expression in Gpr84 shRNA cells, which was accompanied by a decrease in β-catenin and increase in Jmjd5 expression. Actin was used as a loading control.

143

C o n tr o l

1 .5 J m jd 5

n 1 .2

)

g e

n 0 .9

r a

p *

d 0 .6

A ***

l ** ** o

N **

( R

m 0 .3 ****

0 .0 R h o u S tra 6 P itx2 P la u r Irx3 M m p 9 C c n d 1

Figure 4.11: Jmjd5 downregulates the expression of β-catenin targets GSEA revealed β-catenin target genes were highly enriched by Gpr84 in KLSA9M pre-LSC. Expression of these target genes analysed by qRT-PCR confirmed a significant reduction in KLSMLLAF9 pre-LSC overexpressing Jmjd5 compared to control. Data represent mean values ± SEM of three independent experiments. Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (* P < 0.05, ** P < 0.005, *** P < 0.0005, **** P < 0.0001).

144

These data thus support the hypothesis that Jmjd5 negatively regulates the expression of Gpr84 and its downstream targets, which include β-catenin co-effector genes.

To determine the extent to which Jmjd5 target genes overlap with Gpr84, data from gene expression profiling of Jmjd5 transduced KLSMLLAF9 cells were analysed against the data of Gpr84 transduced KLSA9M cells (P < 0.05, fold change > 2; Figure 4.12). From the 256 genes differentially expressed by Gpr84, 92 of these genes were inversely regulated by Jmjd5. A summary of the function of some of the top downregulated and upregulated genes potentially co-regulated by Jmjd5 and Gpr84 are described in Tables 4.1 and 4.2. The combined microarray data were validated using qRT-PCR. Myeloperoxidase (Mpo), Ms4a3 and Ccnd2 (two of which were identified in Figure 4.7) were the most significantly upregulated genes in Jmjd5 overexpressing cells compared to control cells, by 4.5-, 4.8- and 1.5-fold, respectively (P < 0.0001; Figure 4.13A). Consistent with the microarray data, expression of aldehyde dehydrogenase (Aldh3b1), S100 calcium binding protein A8 and A9 (S100a8 and S100a9), arachidonate 5-lipoxygenase (Alox5), Tcf7l2, carbonyl reductase 3 (Cbr3) and Cux2 were all significantly downregulated by 1.8 to 10-fold in Jmjd5 overexpressed cells compared to control (P < 0.0005; Figure 4.13B). In support of the hypothesis that a number of specific genes are co-regulated by Jmjd5 and Gpr84, overexpression of Gpr84 in KLSA9M pre-LSC was associated with a significant suppression Mpo, Ms4a3 and Ccnd2 expression by at least 4.6-fold (P < 0.005; Figure 4.14A). This was accompanied by a 2.8 to 5-fold increase in Alox5, Tcf7l2, Cbr3 and Cux2, and a 69 to 470-fold increase in Aldh3b1, S100a8, S100a9 expression (P < 0.05; Figure 4.14B). These data suggest that although there is only a 36% overlap in differentially expressed genes following Jmjd5 and Gpr84 overexpression, the common targets identified appear to play crucial roles in regulating AML and CML LSC where the expression of these genes is associated with patient survival outcome and chemoresistance.

Since the data implicate Jmjd5 as a negative regulator of Gpr84 expression, where downstream targets of Gpr84 involved in Wnt regulation are also altered, Jmjd5 overexpressing cells were transduced with Gpr84 or a constitutively active β-catenin cDNA to examine whether they could reverse the previously described Jmjd5 phenotype in KLSMLLAF9 (Figures 4.3 and 4.5). A common empty vector (EV) was

145

Control Jmjd5 EV Gpr84

1 2 3 1 2 3 1 2 3 1 2 3 Top 13: Aldh3b1 S100a8 S100a9 Cbr3 Alox5 Cux2 Cd33 Hopx Tns4 Ltb4r1 Camp Timp2 Zfp36l1

Top 3: Mpo Ms4a3 Ccnd2

P ≤ 0.05; Fold change > 2

Figure 4.12: Heatmap of genes inversely co-regulated by Jmjd5 and Gpr84 From the 256 differentially expressed genes regulated by Gpr84, 92 of those genes were inversely regulated by Jmjd5 (P < 0.05, fold change > 2). The top co-regulated genes by Jmjd5 and Gpr84 are indicated.

146

Table 4.2: Brief summary of the function of selected common target genes of Jmjd5 and Gpr84 Function of selected inversely co-regulated targets of Jmjd5 and Gpr84 identified from microarray analysis comparing genes regulated by Jmjd5 in KLSMLLAF9 and Gpr84 in KLSA9M pre-LSC. Gene Function Reference Aldh3b1 -cytosolic enzyme involved in metabolism of retinoic acid and detoxification of (Cheung et al., 2007, Corti et (Aldehyde alkylating drugs such as cyclophosphamide al., 2006, Hess et al., 2004, dehydrogenase -LSC enriched among ALDH subsets Pearce et al., 2005, Ran et -increased expression of ALDH in primary AML patient samples and associated al., 2009) with adverse clinical outcome S100a8 and S100a9 -high expression predicts poor survival in de novo AML patients (Nicolas et al., 2011, Yang et (S100 Calcium -regulator of chemoresistance in AML al., 2012, Spijkers- Binding Protein A8 -pro-autophagic protein that enhances cell survival by directly interacting with Hagelstein et al., 2012) and A9) Beclin1 displacing Bcl2 -elevated expression causes glucocorticoid resistance in MLL-rearranged infant ALL Cbr3 -NAPDH-dependent short-chain dehydrogenases (Bains et al., 2010, Lal et al., (Carbonyl reductase -along with Cbr1, are key anthracycline metabolisers 2008, Varatharajan et al., 3) -doxorubicin is a better substrate that daunorubicin for wild-type Cbr3 and 2012) nonsynonymous single-nucleotide polymorphisms (ns-SNPs) encoding Cbr3 Alox5/ 5-LO -critical regulator of LSC in Bcr-Abl-induced CML (Chen et al., 2009, Catalano (Arachidonate 5- -suppression caused impaired differentiation and survival of long-term CML LSC, et al., 2005, Radmark et al.,

147

Lipoxygenase) preventing the initiation of Bcr-Abl-induced CML 2007, Zhao et al., 2004, -involved in oxidative stress response, inflammation, and cancer Chen et al., 1994) Mpo -associated with a myeloid lineage commitment (Catovsky et al., 1972, (myeloperoxidase) -high expression associated with a favourable prognosis in AML patients Itonaga et al., 2014, Davey et -polymorphonucleic neutrophils lacking Mpo were found in 43% of AML cases al., 1988) -abnormal neutrophils in AML and MDS were myeloperoxidase-deficient

148

A 6 C o n tr o l ****

**** J m jd 5

)

s e

s 4

l o

N 2 ****

(

R m

o 3 2 p a d 4 n M s c M C

1 .2 C o n tr o l

1 .0 J m jd 5

)

s e

s 0 .8

r a

p *** h

x 0 .6

c e

****

d A l ****

o **** ***

N 0 .4

( R

m **** 0 .2 ****

0 .0

1 8 9 5 l2 r3 2 b a a x 7 x 3 0 0 lo f b u h 0 0 A c C C ld 1 1 T A S S

Figure 4.13: Validation of Gpr84 and Jmjd5 common target genes by qRT-PCR in Jmjd5 overexpressing cells (A) Upregulated and (B) downregulated common Jmjd5 and Gpr84 targets identified by microarray analysis were validated by qRT-PCR in KLSMLLAF9 pre-LSC overexpressing Jmjd5. Data represent mean values ± SEM of three independent experiments. Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (* P < 0.05, ** P < 0.005, *** P < 0.0005, **** P < 0.0001).

149

A 2 .0 E V

G p r 8 4

o )

i 1 .5

p h

x 1 .0

f (

R 0 .5 * m ** ** 0 .0

o 3 2 p a d 4 n M s c M C

1 3 d 8 4 u p c o m m o n B 5 0 0 ****

4 5 0 EV

n 1 5 0 G p r8 4

o ****

)

s e

s *

g 1 0 0

p h

x 5 0

6 **

A l

o **** N

f **** (

R 4 **** m 2

1 8 9 5 l2 r3 2 b a a x 7 x 3 0 0 lo f b u h 0 0 A c C C ld 1 1 T A S S

Figure 4.14: Validation of Gpr84 and Jmjd5 common target genes by qRT-PCR in Gpr84 overexpressing cells (A) Downregulated and (B) upregulated common Jmjd5 and Gpr84 targets identified by microarray analysis were validated by qRT-PCR in KLSA9M pre-LSC overexpressing Gpr84 compared to control. Data represent mean values ± SEM of three independent experiments. Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (* P < 0.05, ** P < 0.005, *** P < 0.0005, **** P < 0.0001).

150 used as a control. Overexpression of Gpr84 or β-catenin was confirmed using western blotting, where Gpr84 protein levels were increased in Gpr84 cDNA transduced Jmjd5 cells (Jmjd5-Gpr84) compared to EV cells (Jmjd5-EV) (Figure 4.15A). β-catenin overexpression was confirmed using a flag antibody, as the β-catenin construct contained a flag/DYKDDDDK epitope tag. As expected, only the samples containing β- catenin cDNA showed a band when probed with the flag-antibody (Figure 4.15B). Colony forming assays showed a consistent increase in colony formation following overexpression of either Gpr84 (1.5-fold, P < 0.036) or β-catenin (1.7-fold, P < 0.022) compared to EV control in Jmjd5 overexpressing KLSMLLAF9 cells (Figure 4.15C). Furthermore, this increase in colony forming potential rescued the Jmjd5 phenotype to the extent in which there was no significant difference when Jmjd5-Gpr84 (P = 0.052) or Jmjd5-β-catenin (P = 0.47) was compared to control-EV. To determine whether restoring the expression of Gpr84 or β-catenin could enhance the leukaemogenicity of Jmjd5 overexpressing KLSMLLAF9 pre-LSC, the cells were injected into sublethally irradiated primary recipient mice. Similar to the phenotype observed in vitro, overexpression of Gpr84 or β-catenin accelerated leukaemia onset in Jmjd5- overexpressing cells compared to Jmjd5-EV (P < 0.05, Figure 4.15D), suggesting that both Gpr84 and β-catenin are key components of the Jmjd5 regulatory pathway.

4.7 Discussion

Several histone demethylases have been revealed to be dysregulated in cancers such as AML. Previous studies in various normal and cancer tissues suggest that Jmjd5 expression is cell-type specific and could act as a proto-oncogene (Xiao-nan and Zhi- peng, 2013, Zhang et al., 2015, Hsia et al., 2010) or tumour suppressor (Ishimura et al., 2012, Oh and Janknecht, 2012, Suzuki et al., 2006) depending on the cancer type. In addition, Jmjd5 was also found to be regulated by the Sonic hedgehog pathway in head and neck squamous cell carcinoma (Xiao-nan and Zhi-peng, 2013). Since Jmjd5 is associated with self-renewal pathways in this cancer setting, it would be of interest to examine whether Jmjd5 may be involved in other self-renewal pathways crucial in AML such as the Wnt/β-catenin signalling pathway.

151

A Control Jmjd5 B Control Jmjd5

EV Gpr84 β-cat EV Gpr84 β-cat EV Gpr84 β-cat EV Gpr84 β-cat

Flag Gpr84

Actin Actin

1 4 d c fu re s c u e C * 4 0 0

l l

e 3 0 0

0 0

1 2 0 0

n o

l 1 0 0

o C

0 EVE V GGpr84p r8 4  -cβa-tcate n in EVE V GGpr84p r8 4  -cβa-tcate n in

Control Jmjd5

D 1 0 0 C tr+ E V C tr + G p r 8 4

l C tr +  - c a te n in

a 7 5

v i

v J m jd 5 + E V

r u

s J m jd 5 + G p r8 4

t 5 0 n

e J m jd 5 +  - c a te n in

c r

e 2 5 P

0 0 3 0 6 0 9 0 1 2 0 1 5 0 1 8 0 2 1 0 D a y s p o s t tra n s p la n ta tio n

Figure 4.15: Gpr84 and β-catenin rescue the Jmjd5-overexpressing phenotype in KLSMLLAF9 pre-LSC KLSMLLAF9 pre-LSC with Jmjd5 cDNA or control were transduced with Gpr84 cDNA, constitutively active β-catenin cDNA, or empty-vector control (EV). (A) Western blotting confirmed an increase in Gpr84 expression in Jmjd5 with Gpr84 cDNA (Jmjd5-Gpr84) compared to Jmjd5 with EV (Jmjd5-EV) in KLSMLLAF9 pre-LSC. (B) As the β-catenin cDNA construct contains a flag-tag, flag antibody was used to confirm transduction efficiency of β- catenin cDNA (β-cat) in KLSMLLAF9 pre-LSC overexpressing Jmjd5 (Jmjd5-β-catenin) and 152 control (control-β-catenin). (C) Colony forming assays showed that the Jmjd5 phenotype could be reversed upon enforced expression of Gpr84 (Jmjd5-Gpr84) or β-catenin (Jmjd5-β-catenin). (D) Survival outcome of primary recipient mice injected with KLSMLLAF9 pre-LSC (N = 6) shows that the introduction of Gpr84 or β-catenin cDNA can enhance leukaemogenicity of Jmjd5 overexpressing. Data represent mean values ± SEM of three independent experiments. Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (* P < 0.05).

153

This study is the first to reveal Jmjd5 as a critical demethylase aberrantly downregulated in KLSMLLSAF9 AML. A methyltransferase known to form a complex with MLLAF9 to regulate downstream MLLAF9 target genes is DOT1L (Bernt et al., 2011). The pharmacological inhibition of DOT1L with SGC0946 reduced the clonogenic property of KLSMLLAF9 cells, which was consistent with previous findings where a DOT1L inhibitor was shown to selectively target MLLAF9 cells (Daigle et al., 2011). The downregulation of Ezh2 expression would assist in impairing KLSMLLAF9 colony formation, as inhibition of Ezh2 blocks proliferation (Qi et al., 2012). Furthermore, Ezh2 is known to augment leukaemogenicity in AML by blocking differentiation (Tanaka et al., 2012). The upregulation of Jmjd5 expression upon DOT1L inhibition implicates the tumour suppressor role of Jmjd5 in AML. This hypothesis is also supported by clinical data where AML patients with high JMJD5 expression have a more favourable prognosis (Figure 4.1D).

Functional studies following enforced expression of Jmjd5 in KLSMLLAF9 cells demonstrated that Jmjd5 might regulate a number of pathways including apoptosis and differentiation, as well as impairing colony forming potential. In this context, the downregulation in the expression of the anti-apoptotic gene, Bcl2a1, and differentiation inhibiting gene, Tcf7l2, may explain the increase in apoptosis and induction of differentiation observed in these cells. This finding has potential therapeutic implications since AML is characterised by an accumulation of immature myeloid cells unable to differentiate into mature effector cells (Bruserud et al., 2000) and quiescent LSC are able to evade apoptosis as a mechanism of resistance (Ketley et al., 2000). This is most relevant to AML subtypes showing a block in differentiation including M3 (APL, accumulation of promyelocytes), M4 and M5 (myelomonocytic differentiation), M6 (erythroid predominance) and M7 (acute megakaryoblastic leukaemia) according to the FAB classification. All trans-retinoic acid (ATRA) used in acute promyelocytic leukaemia combination therapy is regarded as the standard treatment for this disease where arsenic trioxide can also be used as an effective treatment for refractory acute promyelocytic leukaemia patients (Lo-Coco et al., 2013). Arsenic trioxide was found to have a dual effect on ATRA-resistant acute promyelocytic leukaemic cells in vivo, both inducing apoptosis and differentiation of leukaemic cells (Kinjo et al., 2000), which is a similar phenotype observed when Jmjd5 is upregulated in KLSMLLAF9. Furthermore, the phenotype of Jmjd5 observed in vitro is similar to that of HOXB1, whose expression 154 is downregulated in primary paediatric AML (Yan-Fang et al., 2012) and the enforced expression of this homeobox gene in AML cells leads to a reduction in cell proliferation, induction of apoptosis and cell differentiation, suggesting a tumour suppressor role in AML (Petrini et al., 2013). Although both apoptosis and differentiation are observed upon Jmjd5 overexpression, these two events are likely to occur via different regulatory pathways (Bruserud et al., 2000). The tumour suppressive effect of Jmjd5 was also observed in vivo, where KLSMLLAF9 cells proliferated at a slower rate and there was a significant delay in AML development due to the overexpression of Jmjd5. Collectively, these results highlight the crucial role Jmjd5 plays in reducing the leukaemogenicity of MLL-rearranged AML, which is one of the most aggressive subtypes of AML.

As mentioned in Chapter 1, current challenges in AML therapeutics are that LSC exhibit high chemoresistance (Huntly and Gilliland, 2005) where the persistence of LSC serves as the origin of relapse (Burnett et al., 2011). Cbr1, which was identified as a potential downstream target of Jmjd5, is the predominant reductase responsible for converting doxorubicin to doxorubinol, a less antineoplastic compound (Kassner et al., 2008). In vitro cytotoxicity of daunorubicin, another anthracycline antibiotic used in induction regimens for AML patients (Burnett, 2012), reduced with increased Cbr1 expression resulting in increased intracellular daunorubicinol levels (Varatharajan et al., 2012). Furthermore, Bcl2a1 expression was also downregulated by Jmjd5, where high Bcl2 expression in AML has previously been reported to cause resistance to etoposide and doxorubicin (Sartorius and Krammer, 2002, Taylor et al., 2000, Kirkin et al., 2004). This is of particular clinical relevance as etoposide has been trialled in combination with standard chemotherapy and shown to increase remission rates in paediatric AML patients (Creutzig et al., 2005, Gibson et al., 2011b). This would suggest that AML patients with higher Jmjd5 expression, and hence lower Cbr1 and Bcl2a1 expression, may be more responsive to anthracycline and etoposide treatment.

Gpr84, a proto-oncogene required for the maintenance of MLL-rearranged AML, has been shown to positively regulate MLL-fusion and Wnt/β-catenin target genes (Dietrich et al., 2014). It is proposed here that Jmjd5 suppresses the expression of Wnt target genes via the downregulation of Gpr84. The reduced expression of Wnt regulators by Jmjd5 such as Rhou, Stra6, Pitx2, Mmp9 and Ccnd1 may contribute to the impaired

155 leukaemogenic potential observed by Jmjd5 overexpression, as these targets have been reported to regulate proliferation, migration and β-catenin activity (Tao et al., 2001, Szeto et al., 2001, Kioussi et al., 2002, Wu et al., 2007, Sansom et al., 2005). The shRNA-mediated suppression of Gpr84 perturbs the β-catenin signalling pathway, which surprisingly enhances Jmjd5 expression, suggesting that a negative feedback loop, comprised of Gpr84, β-catenin and Jmjd5, maintain Wnt/β-catenin signalling in AML LSC. Overexpression of Gpr84 or β-catenin in Jmjd5 overexpressing KLSMLLAF9 cells was sufficient to restore the leukaemogenic potential of KLSMLLAF9 in vitro and in vivo, likely from the rescued Wnt pathway which was impaired by Jmjd5 alone.

A number of genes including Aldh3b1, S100a8 and S100a9, found to be downregulated upon Jmjd5 overexpression, have all been previously shown to be predictors of poor clinical outcome in AML patients (Nicolas et al., 2011, Cheung et al., 2007). In addition, enhanced expression of Mpo, which was also achieved by upregulating Jmjd5, is associated with a favourable prognosis (Davey et al., 1988).

This chapter created a foundation from which the biological function of Jmjd5 was determined, where enforced Jmjd5 expression in LSC impaired MLL AML leukaemogenesis. Epigenetics is dynamic and changes in epigenetic regulators such as the upregulation of JMJD5 in LSC is proposed to cause a shift towards the active chromatin state to remedy the abnormal epigenetic balance, however, it is also important to determine whether this overexpression is comparable to that of HSC to ensure that this level of upregulation is similar to normal physiological levels.

Experiments involving human AML patient samples are also needed to determine whether the tumour suppressor phenotype observed in the AML mouse model used in this study is also observed in human AML. Next generation deep sequencing is an approach that can be utilised to analyse patient samples to detect a mutation or abundance level of a malignant clone in AML (Ding et al., 2012, Ley et al., 2010, Ley et al., 2008). Identifying genetic mutations is indispensable for targeted therapy and prognosis (Burnett et al., 2011) and data from this experiment could provide insights into the tightly regulated network of histone regulation in leukaemogenesis. Since there have been no studies to date that have searched for mutations of JMJD5 in AML patient samples, the lower expression or loss of function of JMJD5 in patients with aggressive

156

AML would assist in determining whether JMJD5 is a crucial determinant of clinical outcome in AML. By screening AML patient samples or AML xenografts for those with higher JMJD5 expression and comparing the expression of proposed target genes and functional activity with samples with lower JMJD5 expression, the findings would allow the discovery gained from this murine model to be translated into a clinical setting and support the potential of JMJD5 to be used as a biomarker for responsiveness to cancer treatment and survival outcome.

Collectively, these data reveal Jmjd5 as a putative novel tumour suppressor in MLLAF9 AML and as such this demethylase may have significant therapeutic potential given its role in lowering resistance to chemotherapeutic agents, dysregulating AML development and progression, and perturbing the Wnt pathway. The findings therefore raise the possibility of using Jmjd5 as a predictive biomarker for diagnosis and survival outcome, and as a determinant of the risk of developing resistance to common chemotherapeutic drugs currently used for AML treatment. This study has established Jmjd5 as a key regulator of MLLAF9-associated leukaemogenesis in AML that should be explored as a novel biomarker and therapeutic strategy.

157

CHAPTER 5 Epigenetic Regulation of Acute Myeloid Leukaemic Stem Cells

5.1 Introduction

Epigenetic alterations maintain proper control of gene expression, however, dysfunction in the writing, reading or erasing of epigenetic marks can be involved in the pathogenesis and development of cancer. Targeting epigenetic machinery can restore normal chromatin modification states and gene expression levels in cancer cells (Popovic and Licht, 2012). Since histone demethylases are the most recently discovered family of erasers, identification of dysregulated members in this family of enzymes could represent new potential diagnostic tools and therapeutic options in targeting chemotherapy-resistant LSC.

As described in Chapter 1, chromatin modifications are abundant, complex and dynamic, and their involvement in gene regulation has been studied extensively in recent years. Several chromatin-associated proteins have been reported to be required for MLL-rearranged leukaemia, including H3K79 methyltransferase, DOT1L (Bernt et al., 2011, Chang et al., 2010), histone demethylase, LSD1 (Schenk et al., 2012, Harris et al., 2012), bromodomain-containing 4 (BRD4) (Zuber et al., 2011, Dawson et al., 2011), PRC2 complex components Ezh1, Ezh2, Eed, and Suz12 (Neff et al., 2012, Shi et al., 2013), PRC1 complex member, CBX8 (Tan et al., 2011), H2B ubiquitin ligase, RNF20 (Wang et al., 2013) and methylcytosine dioxygenase, TET1 (Huang et al., 2013). Results from Chapters 3 and 4 have extended the list of epigenetic regulators required for MLL-rearranged leukaemia, where the upregulation of Jmjd1c and downregulation of Jmjd5 are required for the development and maintenance of MLL AML.

The JmjC histone lysine demethylase family demethylates specific substrates, with particular reference to certain methylated lysine residues within histones. Jmjd1c belongs in the KDM3 histone demethylase family, which also includes Jmjd1a and Jmjd1b. The selective demethylation activity and specificity of Jmjd1a and Jmjd1b have been published previously, having been shown to target mono-or dimethylated histone 3

158 lysine 9 (H3K9me1 or H3K9me2) (Yamane et al., 2006, Hu et al., 2001). However, contrasting reports have arisen on the histone demethylase activity of Jmjd1c. Studies in human embryonic stem cells and mouse testis Leydig cells reported histone demethylase activity (Wang et al., 2014a, Kim et al., 2010), whilst studies in MLLAF9- AML, HEK293 cells and male-germ cells did not detect any activity following cellular and biochemical assays (Sroczynska et al., 2014, Kuroki et al., 2013, Brauchle et al., 2013).

There has also been some controversy regarding the histone demethylase activity of Jmjd5. Crystallography and sequence analysis show that Jmjd5 contains a JmjC domain with all the residues required for histone demethylation activity, however, the exact substrate specificity of Jmjd5 remains unclear. Recent studies in embryonic cells (Ishimura et al., 2012) and breast cancer (Hsia et al., 2010) have shown that Jmjd5 demethylates H3K36me2 and is capable of regulating epigenetic modification in the coding region of genes, however, no global changes in H3K36me2 were observed by Oh and Janknecht in mouse embryonic stem cells (Oh and Janknecht, 2012). Furthermore, investigations into the mechanism of Jmjd5 have demonstrated a possible hydroxylase function towards histone and non-histone proteins (Del Rizzo et al., 2012, Youn et al., 2012), suggesting a context-dependent function.

As reviewed in Chapter 1, H3K9 methylation is generally identified as a transcriptional silencing mark, with enrichment at the promoter region (Kouzarides, 2007, Shilatifard, 2006), however, H3K9me1 has also been associated with active gene transcription (Shilatifard, 2006). Similar complexities in gene regulation also occur for H3K36. H3K36me3 correlates with transcriptional activation which is commonly accepted, however, H3K36me1 has been reported to be an active enhancer (Armstrong, 2014, Carlberg and Molnar, 2014), whilst others showed that H3K36 di- and tri- but not monomethylation correlates with transcriptional activation (Xu et al., 2008). Although H3K36me2 is generally accepted to be associated with transcriptional activation, few studies have identified the H3K36me2 mark to be associated with transcriptional repression when enriched at the transcription start sites in yeast (Strahl et al., 2002, Landry et al., 2003, Biswas et al., 2006). The spatial distribution of di- and trimethylation adds another complexity to H3K36 with regards to gene regulation, however, it is generally accepted that H3K36me2 and H3K36me3 modifications mainly

159 occur within the coding region of genes to enhance the elongation of active transcription (Guenther et al., 2007, Bannister et al., 2005).

A histone methyltransferase that controls H3K36 dimethylation is the nuclear SET domain-containing 2 (NSD2) (Kuo et al., 2011, Wagner and Carpenter, 2012, Li et al., 2009). Found to be overexpressed in up to 20% of multiple myeloma cases (Chesi et al., 1998, Stec et al., 1998), NSD2 has been shown to be linked to Ezh2 through a network of microRNAs (Asangani et al., 2013). Ezh2 is also overexpressed in a diverse range of cancers including AML, and associated with poor prognosis and metastatic progression (Fiskus et al., 2009, Varambally et al., 2002, Bachmann et al., 2006, McCabe et al., 2012, Kleer et al., 2003). Through the regulation of global H3K27 methylation levels associated with transcriptional repression (Cao et al., 2002, Czermin et al., 2002, Cao and Zhang, 2004), Ezh2 mediates the overexpression of NSD2 and can therefore indirectly regulate H3K36me2 levels, which is associated with transcriptional activation (Asangani et al., 2013). The interplay between H3K36 and H3K27 methylation states is an example of the complexity involved in gene regulation.

Studies with Ezh2 have shown that its knockdown blocks cell proliferation, cell invasion, tumour growth, and metastasis (Richter et al., 2009, Takeshita et al., 2005, Varambally et al., 2008). Furthermore, suppression of Ezh2 impairs LSC by reducing stem cell frequency and inducing cell differentiation, and consequently perturbing the progression of AML without affecting HSC (Tanaka et al., 2012, Neff et al., 2012). Thus, the Ezh2-mediated pathway could be an attractive target for epigenetic therapy. 3- Deazaneplanocin A (DZNep), an inhibitor of S-adenosylmethionine-dependent methyltransferase that targets the degradation of Ezh2 (Fujiwara et al., 2014), depletes H3K27me3 levels and induces robust apoptosis in various cancer cells including AML (Zhou et al., 2011, Tan et al., 2007). Unlike chemotherapeutic agents that kill both cancer and normal cells through their cytotoxic capacity, recent clinical studies suggest that the use of epigenetic agents at doses lower than the threshold for cytotoxicity can effectively restore the epigenetic landscape of a malignant cell to a normal state in patients with lung cancer and myelodysplastic syndrome (Itzykson and Fenaux, 2014, Juergens et al., 2011). As genetic depletion of Ezh2 does not appear to substantially increase apoptosis in MLL-rearranged AML (Neff et al., 2012), it was crucial to determine the appropriate concentrations for DZNep treatment that would be sufficient 160 to reverse aberrant epigenetic changes and restore gene expression without leading to any significant increase in apoptotic cell death in MLLAF9 transduced haematopoietic stem/progenitor cells, since these cells are capable of developing aggressive AML in patients (Wang et al., 2010, Zeisig et al., 2012).

This chapter aims to investigate the epigenetic landscape of AML to examine whether Jmjd1c or Jmjd5 regulate histone methylation in order to control gene transcription of downstream targets. In addition, the principle of using epigenetic drugs at concentrations which restore aberrant epigenetic marks rather than causing cytotoxicity was examined with DZNep.

5.2 Global histone 3 lysine 9 methylation levels are not altered by Jmjd1c

Jmjd1c is a putative H3K9 demethylase, targeting H3K9me2/H3K9me1 to generate unmethylated H3K9 (Peeken et al., 2013, Kim et al., 2010). Furthermore, the shortest isoform of Jmjd1c (Jmjd1c-S) has been shown to be functional and may also possess demethylase activity (Wolf et al., 2007). To determine whether global changes in H3K9 methylation could be observed by overexpression of either the long or short isoforms of Jmjd1c (Jmjd1c-L and Jmjd1c-S) in KLSA9M, histone extracts were used for western blotting and probed for all three H3K9 methylation states. The overexpression of Jmjd1c-S in KLSA9M was confirmed by qRT-PCR (P < 0.0002; Figure S.2) and Jmjd1c-L was the same isoform used previously in Chapter 3.

When both the Jmjd1c-S and Jmjd1c-L were overexpressed in KLSA9M pre-LSC (Figure 5.1A) and LSC (Figure 5.1B), there was no marked difference in methylation levels for H3K9me1, H3K9me2 or H3K9me3. Although global methylation changes were not observed, this does not preclude Jmjd1c as a H3K9 demethylase in specific instances.

161

KLSA9M KLSA9M LSC A B Jmjd1c Jmjd1c EV S L EV S L

H3K9me1 H3K9me1

H3K9me2 H3K9me2

H3K9me3 H3K9me3

Total H3 Total H3

Figure 5.1: Jmjd1c does not alter global H3K9 methylation levels Western blot analysis of histone extracts were used to identify changes in global H3K9 mono-, di- and trimethylation levels (H3K9me1, H3K9me2 and H3K9me3, respectively). (A) KLSA9M pre-LSC and (B) LSC transduced with Jmjd1c cDNA (short (S) and long (L) isoforms) or empty vector (EV) showed no observable decrease in H3K9 methylation levels in Jmjd1c overexpressed cells.

162

5.3 Jmjd5 preferentially demethylates histone 3 lysine 36 dimethylation

A recent report suggested that Jmjd5 demethylated H3K36me2 in breast cancer (Hsia et al., 2010). To determine whether Jmjd5 also targets this methylation mark in AML, histone extracts of Jmjd5-transdcued KLSMLLAF9 were analysed. Western blotting detected a substantial decrease in all three H3K36 methylation states by comparison with control cells (Figure 5.2A), suggesting that Jmjd5 can demethylate all three methylation marks. By quantifying the intensity of the bands from replicate blots, it was revealed that the reduction in H3K36me2 levels was greater (83% reduction, P < 0.0001) than H3K36me1 (55% reduction, P = 0.0099) and H3K36me3 (65% reduction, P = 0.01) (Figure 5.2B). This suggests that Jmjd5 preferentially demethylates H3K36me2, although it is not specific for only this methylation state as the overall levels of H3K36me1 and H3K36me3 were also altered. As a proof of principle study, H3K36me2 levels were examined in Gpr84 shRNA cells. As shown in Chapter 4, the expression of Jmjd5 was shown to be upregulated in response to the suppression of Gpr84 (Figure 4.10), suggesting a negative regulation of Jmjd5. As anticipated, Gpr84- suppressed KLSMLLAF9 pre-LSC demonstrated a marked decrease in H3K36me2 levels compared to control (Figure 5.3A). It was also shown in Chapter 4 that the overexpression of Gpr84 or β-catenin was able to reverse the Jmjd5 overexpressed phenotype in vitro and in vivo. To examine whether this may be due to the dysregulation of H3K36me2 levels, the histone extracts of Gpr84- and β-catenin- transduced KLSMLLAF9 pre-LSC were analysed. As expected, low levels of H3K36me2 were observed in Jmjd5-EV cells, however, it was interesting to find an increase in H3K36me2 levels upon enforced expression of Gpr84 or β-catenin in Jmjd5 overexpressed cells (Figure 5.3B). These data suggest that the tumour suppressive phenotype observed in Jmjd5 overexpressing cells may be due to the dysregulation of H3K36me2 and thus the gene expression of downstream Jmjd5 targets.

163

A KLSMLLAF9

Control Jmjd5

H3K36me1

H3K36me2

H3K36me3

Total H3

l e

v 1 .2 e

l C o n tro l

3 l

H J m jd 5

t n

o 0 .8

o **

t *

t y

a 0 .4

t e

e ****

6 3

K 0 .0 3

H 1 2 3 e e e m m m 6 6 6 3 3 3 K K K 3 3 3 H H H

Figure 5.2: Jmjd5 preferentially demethylates H3K36me2 (A) Western blot analysis of histone extracts of KLSMLLAF9 pre-LSC transduced with Jmjd5 cDNA or control was used to identify changes in global H3K36 mono-, di- and trimethylation (H3K36me1, H3K36me2 and H3K36me3). Total H3 was used as a loading control. (B) Replicate western blots were analysed by Image J to measure the intensity of the bands, where H3K36 methylation was quantified relative to total H3. The ratios of H3K36 methylation in KLSMLLAF9 Jmjd5 overexpressing cells were then measured relative to control KLSMLLAF9 pre-LSC. These data showed H3K36me2 to have the greatest reduction in global methylation levels compared to H3K36me1 and H3K36me3. Data represent mean values ± SEM of three independent experiments. Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (* P < 0.05, ** P < 0.01, **** P < 0.0001).

164

A KLSMLLAF9

Scr Gpr84 shRNA

H3K36me2

Total H3

B KLSMLLAF9

Control Jmjd5 EV Gpr84 -cat EV Gpr84 -cat

H3K36me2

Total H3

Figure 5.3: H3K36 dimethylation levels are altered by the ectopic expression of Gpr84 or β-catenin Western blot analysis of histone extracts were used to identify changes in global H3K36 dimethylation (H3K36me2). (A) Knockdown of Gpr84 in KLSMLLAF9 pre-LSC resulted in a reduction in global H3K36me2 levels. (B) Enforced expression of Gpr84 or β-catenin in Jmjd5 overexpressing KLSMLLAF9 pre-LSC resulted in an upregulation in global H3K36me2 levels compared to Jmjd5-EV KLSMLLAF9 pre-LSC.

165

To assess whether Jmjd5 transcriptionally regulates downstream targets through H3K36me2, chromatin immunoprecipitation coupled to detection by quantitative real- time PCR (ChIP-qPCR) analysis was conducted with the assistance of collaborators at WEHI. Since a number of studies have reported that H3K36me2 is enriched within the coding region of genes (Hsia et al., 2010, Bannister et al., 2005), primers were designed to detect the promoter and multiple coding regions of Gpr84 (promoter, beginning and middle of exon 2 (exon 2.1 and 2.2), β-catenin (promoter, exon 3 and exon 9) and Tcf7l2 (promoter, exon 5 and exon 8). Preliminary results showed that H3K36me2 was enriched in control KLSMLLAF9 pre-LSC at exon 2.1 of Gpr84 (Figure S.3A), a modest enrichment at exon 3 of β-catenin (Figure S.3B), and the greatest enrichment at exons 5 and 8 of Tcf7l2 (Figure S.3C), compared to Jmjd5 overexpressing cells. It was difficult to determine which exon of a gene would show the most H3K36me2 enrichment as it varies between genes. Since the qPCR primers for Tcf7l2 showed the most promising results, biological replicates of control and Jmjd5 overexpressing KLSMLLAF9 pre-LSC were used to validate the ChIP-qPCR data. As expected, there was no significant difference in H3K36me2 enrichment at the promoter (Figure 5.4A), however, there was a 2.5-fold decrease at exon 5 (P = 0.016; Figure 5.4B) and 2.1-fold decrease at exon 8 of Tcf7l2 in Jmjd5 overexpressing samples (P = 0.00075; Figure 5.4C). Collectively, these data suggest that enforced expression of Jmjd5 results in a reduction in H3K36me2 enrichment in the coding region of Tcf7l2 and potentially other Jmjd5 target genes, suggesting that this demethylase may regulate the transcriptional silencing of downstream oncogenes through H3K36me2.

5.4 Interplay between H3K36 and H3K27 methylation

As described above, epigenetic regulation is complex due to the interplay between different posttranslational modifications on specific residues within and between histone tails in addition to DNA methylation, which cooperate to regulate gene transcription. The treatment of KLSMLLAF9 pre-LSC with SGC0946, a specific DOT1L inhibitor shown to specifically target and kill MLLAF9 cells, revealed that both Jmjd5 and Ezh2 expression were inversely altered (Figure 4.2). Furthermore, a recent study found that NSD2, a histone methyltransferase that catalyses H3K36me2 associated with active gene transcription, is involved in a regulatory axis with Ezh2, a histone 166

A Tcf7l2 Promoter B Tcf7l2 Exon 5 C Tcf7l2 Exon 8

Percent ofPercent input

Percent ofPercent input Percent ofPercent input

Control Jmjd5 Control Jmjd5 Control Jmjd5

Figure 5.4: Enrichment of H3K36me2 in the coding region of Tcf7l2 in KLSMLLAF9 Chromatin immunoprecipitation coupled to detection by qRT-PCR revealed that H3K36me2 is not enriched at the (A) promoter region of Tcf7l2, but was significantly enriched at (B) exon 5 and (C) exon 8 in control KLSMLLAF9 compared to Jmjd5 overexpressing cells. Horizontal line represents mean values ± SEM of five biological replicates. Statistical significance obtained using an unpaired Student‘s t-test

167 methyltransferase that deposits the H3K27me3 repressive mark (Asangani et al., 2013). Since epigenetic regulation relies on the balance between demethylases and methyltransferases, Jmjd5 overexpressing cells were examined to determine whether there may be an association between H3K36me2 and H3K27me3 marks. Western blotting revealed that H3K27me3 was reduced in response to increased Jmjd5 expression (Figure 5.5A). This result was also observed when Gpr84 knockdown cells were examined, where there was a reduction in global H3K27me3 levels (Figure 5.5B). The link between H3K36me2 and H3K27me3 was also supported by western blotting of Jmjd5 rescue cells, where restoring Gpr84 or β-catenin expression in Jmjd5 overexpressing cells reversed the downregulation of the H3K27me3 mark observed in Jmjd5-EV cells (Figure 5.5C). Since the substrate specificity of Jmjd5 has previously been shown to be towards residue 36 on histone 3 (Hsia et al., 2010), these data suggest that Jmjd5 indirectly reduces H3K27me3, potentially through Ezh2.

5.5 DZNep reduces H3K27 methylation levels at low concentrations without inducing apoptosis

3-Deazaneplanocin A (DZNep), a chromatin-remodelling compound, depletes Ezh2 and the associated H3K27me3, and induces robust apoptosis in various cancer cells including AML (Zhou et al., 2011, Tan et al., 2007). Asangani and co-workers, who identified the regulatory axis of Ezh2 and NSD2, also reported that DZNep treatment reduced NSD2 expression as well as its substrate, H3K36me2 (Asangani et al., 2013). Although an increase in Jmjd5 expression promotes the downregulation of H3K36me2, there are currently no epigenetic drugs available that can pharmaceutically enhance the regulation of Jmjd5. Since treatment with DZNep can indirectly reduce H3K36me2 levels, DZNep was examined as an epigenetic agent to target AML LSC.

Recent studies suggest that Ezh2 knockout does not induce considerable apoptosis in cancer cells (Neff et al., 2012, Tanaka et al., 2012). To confirm whether this was also observed in MLL-rearranged AML, KLSMLLAF9 pre-LSC were transduced with Ezh2-shRNA or a scrambled control (Scr). The knockdown efficiency was determined by western blotting, which showed shRNA 3 to have the largest reduction in Ezh2 expression (Figure 5.6A). Consistent with previous observations, inhibition of Ezh2

168

A KLSMLLAF9 B KLSMLLAF9

Control Jmjd5 Scr Gpr84 shRNA

H3K27me3 H3K27me3

Total H3 Total H3

C KLSMLLAF9

Control Jmjd5 EV Gpr84 -cat EV Gpr84 -cat

H3K27me3

Total H3

Figure 5.5: Jmjd5 indirectly reduces H3K27me3 methylation Western blot analysis probed for H3K27 trimethylation (H3K27me3) in (A) KLSMLLAF9 transduced with control or Jmjd5 cDNA, (B) KLSMLLAF9 transduced with Scramble (Scr) or Gpr84 shRNA, and (C) KLSMLLAF9 with control or Jmjd5 overexpressing Gpr84, β-catenin or EV.

169

1 b e z h 2 c fu A KLSMLLAF9 B 4 0 0

Scr sh1 sh2 sh3 sh4 s

l l

e 3 0 0

Ezh2 0 **

0 0

1 2 0 0

Actin

n o

l 1 0 0

o C

0 S c r s h 3 C 1 c e z h 2 c e ll n o D 1 d e z h 2 a n n e x in

3 1 5

)

(

s l

 2 1 0 l

( *

u p

n 1 5

A C

0 0 S c r s h 3 S c r s h 3

Figure 5.6: Suppression of Ezh2 inhibits colony formation without inducing apoptosis (A) Western blotting of KLSMLLAF9 pre-LSC transduced with Ezh2 shRNA or scramble (Scr) revealed shRNA 3 (sh3) to have the most reduction in Ezh2 expression compared to Scr. Upon Ezh2 suppression, there was a reduction in (B) colony number and (C) cell number. (D) Annexin V and 7AAD staining revealed no significant difference in apoptosis between sh3 and Scr. Data represent mean values ± SEM of three independent experiments. Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (* P < 0.05, ** P < 0.01).

170 resulted in reduced colony formation (30% reduction, P = 0.008; Figure 5.6B) and proliferative capacity (35% decrease, P = 0.035; Figure 5.6C), but did not induce a significant increase in apoptosis (P = 0.79; Figure 5.6D).

Previous studies using DZNep reported a substantial increase in apoptosis (Tan et al., 2007, Fiskus et al., 2009), however, no significant increase in apoptosis in Ezh2 deficient cells was observed. To imitate the Ezh2-deficient phenotype, KLSMLLAF9 pre-LSC were treated with a range of concentrations of DZNep to determine an appropriate concentration that would not lead to a significant increase in apoptotic death. Since DZNep was used at concentrations greater than 500 nM in previous studies showing significant apoptotic cell death (Fiskus et al., 2009), 100 nM was used as the maximum dose. The assay results revealed DZNep concentrations of 50 nM or lower did not result in any significant increase in apoptotic cells, whilst treatment with 100 nM (P = 0.02) resulted in a 2.1-fold increase in apoptotic cells (Figure 5.7).

Since DZNep treatment at concentrations above 50 nM induced significant apoptosis, it was hypothesised that lower concentrations of DZNep could be used more effectively to exert epigenetic rather than cytotoxic effects in AML cells and therefore provide a greater therapeutic potential. To test this hypothesis, GMPMLLAF9 and KLSMLLAF9 pre-LSC were treated with 10 nM or 25 nM DZNep. Treatment of GMPMLLAF9 with DZNep showed a 1.8- and 2.7-fold reduction in colony formation (P < 0.002; Figure 5.8A) and 2.2- and 9.3-fold reduction in cell number (P < 0.007; Figure 5.8C) at 10 nM and 25 nM, respectively. DZNep was also effective in significantly reducing both colony number by 1.9- and 5.9-fold (P < 0.03; Figure 5.8B) and cell number by 2.9- and 11.1-fold (P < 0.0007; Figure 5.8D) in KLSMLLAF9 pre-LSC when treated with 10 nM and 25 nM, respectively. When comparing between GMPMLLAF9 and KLSMLLAF9 pre-LSC, treatment with 10 nM DZNep displayed similar inhibitory effects regardless of cellular origin, however, DZNep was more potent towards KLSMLLAF9 than GMPMLLAF9 when the concentration was increased to 25 nM. Interestingly, DZNep treatment at the lowest concentration, 10 nM, was sufficient to induce differentiation in both GMPMLLAF9 (Figure 5.8E) and KLSMLLAF9 pre-LSC (Figure 5.8F), as revealed by the presence of neutrophils following Wright-Giemsa staining, whereas the controls consisted only of blast cells. The significant impairment in colony formation and proliferation of DZNep-treated pre-LSC is illustrated by the representative colony

171

2 b a n n e x in m ll

5 0 *

) 4 0

(

l l

e 3 0

i t

o 2 0

p o

p 1 0 A

0 C tr 1 0 2 5 5 0 1 0 0

[D Z N e p ] n M

Figure 5.7: DZNep does not induce apoptosis at low concentrations in MLL pre-LSC KLSMLLAF9 pre-LSC treated with various concentrations of DZNep then stained with Annexin V and 7AAD for flow cytometry to measure apoptosis. Data represent mean values ± SEM of three independent experiments. Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (* P < 0.05).

172

A B GMPMLLAF9g m c fu KLSMLLAF9m c fu

3 0 0 3 2 0

l e

e 2 4 0 c

c *

2 0 0

0 0

** 0

0 1

1 1 6 0

***

y n

1 0 0 n

o l

l 8 0 **

C C

0 0 C o n tro l 1 0 2 5 C o n tro l 1 0 2 5

[D Z N e p ] n M [D Z N e p ] n M C g m c e ll n o D m c e ll n o

0 .6 1 .0

)

) 6

6 0 .8



 0 .4

( (

0 .6

** r

b m

m 0 .4 ***

n n

0 .2

l l

*** e

e 0 .2 ***

C C

0 .0 0 .0 C o n tro l 1 0 2 5 C o n tro l 1 0 2 5

[D Z N e p ] n M [D Z N e p ] n M

E F

20 µm

0 nM 0 nM 10 nM 10 nM G Concentration of DZNep Control 10 nM 25 nM

GMPMLLAF9

KLSMLLAF9

100 µm

173

Figure 5.8: DZNep impairs clonogenic growth of pre-LSC by inducing differentiation Colony forming assays following DZNep treatment with control (DMSO), 10 nM or 25 nM DZNep showed a reduction in (A-B) colony number and (C-D) cell number in GMPMLLAF9 and KLSMLLAF9 pre-LSC. Representative images of Wright-Giemsa stained DZNep-treated (E) GMPMLLAF9 and (F) KLSMLLAF9 pre-LSC revealed differentiated cells following treatment. (G) Representative microscope images of colony morphology following colony forming assays with DZNep or control. Data represent mean values ± SEM of three independent experiments. Asterisks indicate statistical significance obtained using an unpaired Student‘s t- test (* P < 0.05, ** P < 0.01, *** P < 0.0005). Images captured using 10× magnification lens.

174 morphology images (Figure 5.8G). Taken together, these results suggest that DZNep at concentrations as low as 10 nM, is potent in impairing colony formation and proliferation in both GMPMLLAF9 and KLSMLLAF9 pre-LSC by inducing differentiation, where DZNep preferentially targets stem cell-derived pre-LSC at the early stage of LSC development.

To determine whether similar inhibitory effects could be achieved in LSC, GMPMLLAF9 or KLSMLLAF9 LSC were treated with DZNep at 10, 25 or 50 nM. Unlike pre-LSC, LSC required higher concentrations of DZNep to significantly impair their clonogenic growth, where 25 and 50 nM DZNep inhibited LSC function as revealed by the 1.6- and 2.9-fold reduction in colony number (P < 0.01; Figure 5.9A) and 2.5- and 8.3-fold reduction in cell proliferation (P < 0.02 Figure 5.9C), respectively. Similarly, KLSMLLAF9 LSC demonstrated significantly reduced colony formation (2.1 to 3.6-fold, P < 0.01; Figure 5.9B) and proliferation (5.1 to 51-fold, P = 0.006; Figure 5.9D) at both 25 nM and 50 nM, which is also reflected by the smaller colony size and reduced density (Figure 5.9G). Consistent with its effects on pre-LSC, 25 nM DZNep appeared to be slightly more potent at suppressing stem cell-derived LSC than progenitor-derived LSC, supporting the potential role of DZNep in targeting KLSMLLAF9 LSC. Furthermore, low doses of DZNep induced differentiation in both GMPMLLAF9 LSC (Figure 5.9E) and KLSMLLAF9 LSC (Figure 5.9F). It was also confirmed that both stem cell and progenitor-derived LSC showed no significant increase in apoptosis when treated with 25 nM DZNep compared to control (Figure 3.10). DZNep was also approximately 2- and 3-fold more effective in blocking the growth of pre-LSC than LSC for GMPMLLAF9 (P = 0.02) and KLSMLLAF9 (P = 0.008), respectively (Figure 3.11), suggesting that the inhibitory effects of DZNep depend not only on the cell of origin but also on the stage of LSC development.

As DZNep is considered an epigenetic drug that reduces H3K27me3 (Cheng et al., 2012, Miranda et al., 2009, Zhou et al., 2011, Fiskus et al., 2009), global H3K27me3 levels were assessed following treatment with this drug. In agreement with previous studies (Fiskus et al., 2009), western blot analysis revealed a reduction in H3K27me3 levels in both GMPMLLAF9 (Figure 5.12A) and KLSMLLAF9 pre-LSC (Figure 5.12C) after treatment with 10 nM, 25 nM and 100 nM DZNep. Further quantitative analysis of independent results showed that H3K27me3 levels were reduced by

175

A GMPMLLAF9g m ls c c fu B KLSMLLAF9m ls c c fu

1 5 0 2 0 0

l e

e 1 5 0

1 0 0 *

0 1

1 1 0 0

y n

n 5 0 **

o l

l 5 0

C C

0 0 C o n tro l 2 5 5 0 C o n tro l 2 5 5 0

[D Z N e p ] n M [D Z N e p ] n M C g m ls c c e ll n o D m ls c c e ll n o

1 .0 1 .0

)

) 6

6 0 .8 0 .8



(

0 .6 0 .6

e b

b * m

m 0 .4 0 .4

l l

l ** * e

e 0 .2 0 .2 C C ** 0 .0 0 .0 C o n tro l 2 5 5 0 C o n tro l 2 5 5 0

[D Z N e p ] n M [D Z N e p ] n M E F

20 µm

0 nM 25 nM 0 nM 25 nM Concentration of DZNep G Control 25 nM 50 nM

GMPMLLAF9

KLSMLLAF9

100 µm

176

Figure 5.9: DZNep impairs clonogenic growth of LSC by inducing differentiation Colony forming assays following DZNep treatment with control (DMSO), 25 nM or 50 nM DZNep showed a reduction in (A-B) colony number and (C-D) cell number in GMPMLLAF9 and KLSMLLAF9 LSC. Representative images of Wright-Giemsa stained DZNep-treated (E) GMPMLLAF9 and (F) KLSMLLAF9 LSC revealed differentiated cells following treatment. (G) Representative microscope images of colony morphology following colony forming assays with DZNep or control. Data represent mean values ± SEM of three independent experiments. Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (* P < 0.05, ** P < 0.01). Images captured using 10× magnification lens.

177

A 4 a g m a p o p B 4 b m a p o p

4 0 4 0

)

) %

% 3 0 (

( 3 0

2 0

2 0 c

p o o 1 0

1 0 p

A A

0 0 C o n tro l 2 5 n M C o n tro l 2 5 n M

[D Z N e p ] n M [D Z N e p ] n M

Figure 5.10: DZNep does not induce apoptosis at low concentrations in MLL LSC GMPMLLAF9 and KLSMLLAF9 LSC treated with 25 nM DZNep were stained with Annexin V and 7AAD for flow cytometry to measure apoptosis. Data represent mean values ± SEM of three independent experiments. Statistical significance obtained using an unpaired Student‘s t- test.

178

1 .2 K L S M L L A F 9 p r e -L S C

) s

l G M P M L L A F 9 p re - L S C

o e

r 0 .9 t

c K L S M L L A F 9 L S C

n 0

o ** 0

c G M P M L L A F 9 L S C

o 1

t 0 .6

l l

e 0 .3

( C

0 .0 C o n tro l 2 5

[D Z N e p ] n M

Figure 5.11: DZNep preferentially targets HSC-derived malignant stem cells at low doses The cell of origin (GMP and KLS) and stage of LSC development (pre-LSC and LSC) are factors which determine sensitivity of AML cells to DZNep treatment. Data represent mean values ± SEM of three independent experiments. Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (* P < 0.05, ** P < 0.01).

179

6 b g m h is m o d

A Concentration of DZNep (nM) B 1 .2

) 3

Control 10 25 100 s

l v ** *

a 0 .8

o ***

H3K27me3 e

GMPMLLAF9 m

2 i

t 0 .4

a l

Total H3 3

r (

0 .0 C o n tro l 1 0 2 5 1 0 0

[D Z N e p ] n M

C D 6 d m h is m o d 1 .2

Concentration of DZNep (nM)

)

l Control 10 25 100 v

a 0 .8 *

t 3

** ***

H3K27me3 m e

KLSMLLAF9 7

2 i

t 0 .4

3 e

Total H3 H

r (

0 .0 C o n tro l 1 0 2 5 1 0 0

[D Z N e p ] n M

Figure 5.12: Low doses of DZNep are sufficient to reduce global levels of H3K27me3 (A) Western blot analysis of total H3 (loading control) and repressive mark H3K27me3 following treatment of GMPMLLAF9 and KLSMLLAF9 pre-LSC with control (DMSO) or indicated concentration of DZNep. (B) The intensity of the bands from replicate blots were quantified by Image J, relative to total H3 and normalised to control treatment. Data represent mean values ± SEM of three independent experiments. Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (* P < 0.05, ** P < 0.01, *** P < 0.005).

180 approximately 40% in GMPMLLAF9 (P < 0.01; Figure 5.12B) and 56% in KLSMLLAF9 pre-LSC (P < 0.04; Figure 5.12D). These data suggest that low-dose DZNep is effective in reversing aberrant epigenetic marks.

5.6 The origin and developmental stage of malignant stem cells determine downstream regulatory pathways in response to DZNep treatment

To examine whether the downstream targets regulated by DZNep vary between LSC derived from a different cellular origin and developmental stage, the expression of genes (i.e. p16, Fbxo32, p21, Txnip, Hoxa9 and Ccne1) that have been implicated as DZNep targets in several studies were examined using real-time qRT-PCR (Fiskus et al., 2009, Zhou et al., 2011). A brief description of the known function of these genes is summarised in Table 5.1. Since concentrations of at least 500 nM were used in previous studies, it was anticipated that differential expression levels and pathways may be regulated from those previously published.

Consistent with our functional results, DZNep treatment at 25 nM compared to 10 nM showed more profound effects on target gene expression in pre-LSC/LSC (Figures 5.12 and 5.13). Using the criteria of differential expression of at least 1.8-fold and P < 0.05, qRT-PCR results showed that p16 and Fbxo32 were upregulated and Ccne1 was downregulated in GMPMLLAF9 pre-LSC (Figure 5.12A); p21 was upregulated and Hoxa9 was downregulated in KLSMLLAF9 pre-LSC (Figure 5.12B); p16 was upregulated and Ccne1 was downregulated in GMPMLLAF9 LSC (Figure 5.13A); and p16, Fbxo32 and p21 were upregulated in KLSMLLAF9 LSC (Figure 5.13B) in response to DZNep treatment at 25 nM. These results suggest that appropriate doses of DZNep can exert epigenetic effects through different downstream regulatory pathways that largely depend on the origin and stage of malignant stem cells.

181

A GMPMLLAF9 pre-LSC

3 *

n o

i CtrC tr s

s 101 0 nM n M e

r ** ** p 252 5 nM n M

x 2 e

R m

e v

i *

e R 0 p 1 6 F b x o 3 2 p 2 1 T x n ip H o x a 9 C c n e 1

KLSMLLAF9 pre-LSC

n o i 2 .5 * s *

s CtrC tr e r 101 0 nM n M

p 2 .0 * x

e 252 5 nM n M

A 1 .5 *

N R

m 1 .0

e v

i **

t 0 .5

e R 0 .0 p 1 6 F b x o 3 2 p 2 1 T x n ip H o x a 9 C c n e 1

Figure 5.13: DZNep restores expression of different downstream targets in pre-LSC generated from distinct cellular origins The mRNA expression levels of known DZNep target genes (i.e. p16, Fbxo32, p21, Txnip, Hoxa9 and Ccne1) determined by qRT-PCR following treatment of (A) GMPMLLAF9 pre-LSC and (B) KLSMLLAF9 pre-LSC with DMSO control (Ctr) or DZNep at the indicated concentrations. Red arrows indicate genes that were differentially expressed by at least 1.8-fold when compared to control treatment with P < 0.05. Data represent mean values ± SEM of four independent experiments. Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (* P < 0.05, ** P < 0.01).

182

GMPMLLAF9 LSC

n o

i 3 .0

s *

CtrC tr

s e r 2 .5 101 0nM n M

p * x 25 nM

e 2 5 n M

2 .0 A

N 1 .5

e 1 .0 **

v * i

t ** a

l 0 .5

e R 0 .0 p 1 6 F b x o 3 2 p 2 1 T x n ip H o x a 9 C c n e 1

KLSMLLAF9 LSC B

n 3 .0 o

i CtrC tr s

s 2 .5

e ** 101 0 nM n M

r * p

x 252 5 nM n M 2 .0

e **

N 1 .5

R m 1 .0

e * *

t a

l 0 .5

e R 0 .0 p 1 6 F b x o 3 2 p 2 1 T x n ip H o x a 9 C c n e 1

Figure 5.14: DZNep restores expression of different downstream targets in LSC generated from distinct cellular origins The mRNA expression levels of known DZNep target genes (i.e. p16, Fbxo32, p21, Txnip, Hoxa9 and Ccne1) determined by qRT-PCR following treatment of (A) GMPMLLAF9 LSC and (B) KLSMLLAF9 LSC with DMSO control (Ctr) or DZNep at the indicated concentrations. Red arrows indicate genes that were differentially expressed by at least 1.8-fold when compared to control treatment with P < 0.05. Data represent mean values ± SEM of four independent experiments. Asterisks indicate statistical significance obtained using an unpaired Student‘s t- test (* P < 0.05, ** P < 0.01).

183

Table 5.1: Brief summary of the known function of DZNep-related target genes Gene Function Reference p16 -tumour suppressor, cell-cycle regulator (G1/S phase), inhibitor of cdk4 and cdk6 (Kotake et al., 2007, Bracken et al., (INK4A/ -associated with oncogene-induced senescence 2007, Ueda et al., 2014, Kaneda et al., CDKN2A) -epigenetically regulated by PRC2 and PRC1 2011, Liggett and Sidransky, 1998) -knockdown exacerbates MLL/ENL leukaemia Fbxo32 -component of the stem cell factor ubiquitin protein E3 ligase complex (Thorsteinsdottir et al., 2001, Schisler (F-box protein 32) -induces apoptosis in breast and colorectal cancer cells et al., 2008, Fiskus et al., 2009) -repressed by Ezh2-mediated H3K27me3 p21 -cell-cycle regulator (G1/S phase), inhibitor of cdk2, cdk1 and cdk4/6 complexes (Fiskus et al., 2009, Brakensiek et al., (CIP1/WAF1) -associated with growth arrest and differentiation in AML cells 2005) Txnip -inhibits thioredoxin activity and controls redox (Junn et al., 2000, Zhou and Chng, (Thioredoxin- -increases ROS production and induces endoplasmic reticulum stress 2013, Shah et al., 2013) binding protein 2) -induces apoptosis Hoxa9 -homeobox transcription factor (Ayton and Cleary, 2003, Kumar et (Homeobox A9) -essential for maintaining the malignant MLL phenotype in AML al., 2004, Milne et al., 2005, Tan et -repressed by Ezh2 al., 2007) Ccne1 -target of Fbxo32 (Abdel Alim et al., 2010, Iida et al., (Cyclin e1) -overexpressed in AML patient samples and several solid tumours 1997) -potential prognostic marker for AML

184

5.7 Discussion

This chapter examined the epigenetic regulation of LSC by modifying the expression of histone demethylases, Jmjd1c and Jmjd5. The data demonstrated that Jmjd1c was unable to alter global H3K9 in any of the three methylation states, which was consistent with recent reports in AML (Sroczynska et al., 2014), however, other studies have reported demethylase activity of Jmjd1c using ChIP experiments (Wang et al., 2014a, Kim et al., 2010, Shakya et al., 2015). Further investigations are necessary to confirm whether Jmjd1c possesses demethylase activity in AML, and whether the downstream target genes of Jmjd1c identified in Chapter 3 are epigenetically regulated by Jmjd1c.

On the other hand, Jmjd5 was found to alter all three H3K36 methylation states, preferentially targeting H3K36me2. As mentioned previously, studies using primary human AML samples are required to support the evidence presented in this thesis using a murine AML model. Following the screening of primary samples to determine the relative levels of JMJD5 expression, the H3K36 methylation levels should be assessed to determine whether there is a correlation between expression and methylation levels. This would more closely predict the function of JMJD5 in altering the epigenetic balance under physiological conditions.

To address the discrepancy in the demethylase function of JMJD1C and JMJD5, structure function analyses would reveal the importance of the JmjC domain for demethylase activity. JMJD1A and JMJD1B have both been shown to possess demethylase activity (Yamane et al., 2006) and are structurally similar to JMJD1C, with a zinc-finger-like domain, JmjC domain, and conserved Fe(II)-binding and α- ketoglutarate-binding sites. JMJD5 only contains a JmjC domain along with Fe(II)- binding and α-ketoglutarate-binding sites, however, the serine residues are in place of threonine in the first α-ketoglutarate-binding site (Klose et al., 2006). To evaluate which domains are important for its enzymatic activity, a series of expression constructs with deletions or mutations of the various domains would be required followed by demethylase activity assays to demonstrate which domains are critical for enzymatic activity and how this may alter the substrate specificity.

Furthermore, shRNA-mediated suppression of Gpr84, which has been shown to upregulate the expression of Jmjd5, also showed a reduction in global H3K36me2

185 levels. The reversal of H3K36me2 levels brought upon by the overexpression of Gpr84 and β-catenin in Jmjd5 overexpressed cells suggests that the restoration of this methylation mark may play a role in rescuing the leukaemogenic phenotype observed in vitro and in vivo. Although the epigenetic regulation of Gpr84 or β-catenin was not confirmed due to the difficulties in determining the ideal spatial distribution of H3K36me2 enrichment in the coding region of target genes, preliminary ChIP-qPCR data involving Tcf7l2 look promising as it demonstrates that Jmjd5 overexpressed cells have less H3K36me2 enrichment in the coding region of the gene which suggests that the gene is epigenetically silenced. Further analysis involving ChIP-sequencing would be beneficial and necessary to confirm that differences in enrichment are observed in other Jmjd5 target genes. The interplay between H3K36me2 and H3K27me3 proposed by the NSD2/Ezh2 regulatory axis, was also observed in this study, where a reduction in H3K36me2 was associated with a reduction in H3K27me3. The restoration of proper epigenetic balance may require this association between H3K36 and H3K27, which could also be analysed through ChIP-sequencing.

DZNep is an epigenetic drug shown to reduce Ezh2 expression and its associated H3K27me3 mark. Previous studies have revealed that DZNep induces apoptosis following treatment at high concentrations of DZNep (Xie et al., 2011, Zhou et al., 2011, Myssina et al., 2009), however, knockdown of Ezh2 does not appear to substantially increase apoptosis in MLL-rearranged AML (Neff et al., 2012). Recent clinical studies have also shown that epigenetic agents at doses lower than the cytotoxicity threshold can effectively reverse aberrant epigenetic marks in cancer (Juergens et al., 2011, Itzykson and Fenaux, 2014). The data obtained in this chapter showed that global reductions in H3K27me3 levels were induced following treatment with DZNep at concentrations as low as 10 nM. The effect of using various doses was examined further to provide evidence that the phenotype observed following DZNep treatment is due to the epigenetic activity of the drug and not cytotoxicity.

In agreement with previous studies in AML cell lines (Zhou et al., 2011, Fiskus et al., 2009) the data in this study indicate that DZNep is highly effective at suppressing the aggressive MLL phenotype, reducing proliferation of MLL-transformed HSC and progenitors, as well as inducing differentiation in all cell types. The origin of LSC has been reported to influence disease outcomes in MLL-rearranged AML, where stem cell-

186 derived leukaemias are more resistant to chemotherapy than progenitor-derived leukaemias (Krivtsov et al., 2012). The results here showed that DZNep exhibited profound inhibitory effects, and stem cell-derived pre-LSC/LSC displayed higher sensitivity to treatment compared to their respective progenitor-derived cells when treated at 25 nM. This finding is of particular interest since stem cell-derived malignant cells are considered to be one of the main contributors for relatively poor outcome in various cancers (Shlush et al., 2014, Krivtsov et al., 2012). Hence, targeting this specific cell type has received special attention in cancer research. Given the growth inhibition effects observed for DZNep on stem cell-derived pre-LSC/LSC in this study, and minimal effects on DZNep-treated normal human hematopoietic stem cells, where concentrations of 0.5, 1 and 2 µM DZNep were used (< 20% reduction in colony formation) (Zhou et al., 2011), these findings provide evidence supporting a role for DZNep as a promising therapeutic agent for cancer treatment.

To investigate whether the phenotypic effects induced by DZNep in pre-LSC/LSC are due to epigenetic-mediated changes, the expression of several genes previously reported to be DZNep targets in a diverse range of cancer cells including primary AML were examined (Fiskus et al., 2009). Consistent with the observation that sensitivity to DZNep treatment is determined by the origin and developmental stage, the expression levels and patterns of DZNep target genes were distinct in each cell type. Following DZNep treatment, KLSMLLAF9 pre-LSC showed the most growth inhibition which was associated with a significant reduction in Hoxa9 expression, a well-known target of MLL oncoproteins (Zeisig et al., 2004). Overexpression of Hoxa9 is associated with poor prognosis in AML patients (Golub et al., 1999), whilst suppression of Hoxa9 results in differentiation of human MLLAF9 leukaemic cells (Faber et al., 2009). Given the important functional role of Hoxa9 in MLL-rearranged AML, it is not surprising that DZNep induces the greatest inhibitory effect on KLSMLLAF9 by targeting Hoxa9. GMPMLLAF9 pre-LSC and KLSMLLAF9 LSC showed similar sensitivity to DZNep treatment. Consistent with this finding, these two cell types upregulated the expression of the common DZNep target genes, p16 and Fbxo32, which are well-characterized Ezh2 targets, often silenced in cancer cells (Merlo et al., 1995, Chou et al., 2010, Wilson et al., 2010), which themselves possess tumour suppressive properties by reducing proliferation (Serrano et al., 1993). Although the apoptosis-related gene, Txnip, has been reported to be up-regulated by DZNep and to induce apoptosis in 187 human AML LSC (Zhou et al., 2011), no significant alteration in its expression was detected regardless of cellular origins and developmental stages. This result is consistent with the observed phenotype in the current study, where the concentrations of DZNep used effectively triggered differentiation rather than apoptosis in AML pre- LSC/LSC. The expression of p21, which has been associated with growth arrest and differentiation in AML cells (Fiskus et al., 2009), was significantly upregulated by stem-cell derived pre-LSC/LSC, which may also contribute to the greater phenotypic changes observed in these cells compared to their respective progenitor-derived cells.

Both stem cell- and progenitor-derived LSC were found to be less sensitive to DZNep treatment than pre-LSC, suggesting a developmental stage-dependent effect, which may be due to the different target genes regulated by DZNep treatment. These results are in line with recent studies indicating that the cell of origin and developmental stage influence gene expression programs in AML (Krivtsov et al., 2012, Shlush et al., 2014), and suggest that DZNep targets different functional programs and downstream regulatory pathways largely depending on the origin and stage of LSCs. Future studies into the identification of pathways mediated by DZNep treatment on a genome-wide scale in mouse and human pre-LSC and LSC will expand the understanding on how these pathways are regulated in a context-dependent manner.

DZNep was the first drug proposed to inhibit EZH2 (Tan et al., 2007), however, it was first described as an S-adenosylhomocysteine hydrolase inhibitor when it was isolated in 1986 (Glazer et al., 1986). The inhibition of S-adenosylhomocysteine hydrolase results in the accumulation of adenosylhomocysteine, which leads to product inhibition of methyltransferases dependent on S-adenosylmethionine (Chiang and Cantoni, 1979) which indirectly inhibits methyltransferases activity due to limited methyl donor groups (Chiang et al., 1992). The specificity of DZNep in targeting Ezh2 has been questioned and proposed to be through indirect mechanisms (Miranda et al., 2009), which has led to the identification of selective small molecule inhibitors for EZH2. GSK296 and GSK343 were identified as S-adenosylmethionine inhibitors with high selectivity towards EZH2 (GSK296: IC50 = 0.020 µM; GSK343 IC50 = 0.004 µM) of greater than

1000-fold over most other methyltransferases, except EZH1 (GSK296: IC50 = 2.5 µM;

GSK343 IC50 = 0.24 µM), which has 76% homology with EZH2 (Verma et al., 2012). Another EZH2-selective small molecule inhibitor, El1, was shown to competitively

188 bind to the S-adenosylethionine pocket of EZH2 SET domain (Qi et al., 2012). EPZ005687 (Knutson et al., 2012) and GSK126 (McCabe et al., 2012), which inhibit EZH2 in a similar manner to El1, were identified through high-throughput screening. GSK126 has been used to assess lymphoma xenograft models and showed complete inhibition of tumour growth and increased survival in mice (McCabe et al., 2012), however, the use of EPZ005687 have been limited to in vitro studies due to pharmacokinetic properties (Konze et al., 2013). EPZ-6438, based on EPZ005687, was therefore designed with improved pharmacological properties such as increased potency and oral bioavailability, and demonstrated antitumour activity in malignant rhabdoid tumour xenograft (Knutson et al., 2013) and lymphoma xenograft models (Knutson et al., 2014). This has led to the current study of EPZ-6438 (labelled as E7438) in a phase 1 clinical trial for patients with B-cell lymphomas or advanced solid tumours (Clinical Trials.gov Identifier: NCT01897571).

Unlike these EZH2-specific inhibitors, UNC1999 is a dual EZH2 and EZH1 small molecule inhibitor recently shown to be more effective than EZH2-specific inhibitors in MLLAF9-transformed leukaemia cells (Xu et al., 2015). Since MLL-rearranged leukaemia coexpress EZH1 and EZH2 (Xu et al., 2015) and requires both EZH2 and EZH1 to be disrupted in order to inhibit leukaemia growth (Neff et al., 2012, Shi et al., 2013), UNC1999 and not GSK126 was shown to be effective against MLLAF9- leukaemia in vitro and in vivo (Xu et al., 2015). Although the specificity of DZNep has been called into question by Miranda and co-workers (Miranda et al., 2009), there have been no studies to analyse the specificity and IC50 of DZNep against a panel of S- adenosylmethionine-dependent methyltransferases. Furthermore, a number of histone methyltransferases, not only EZH2 and EZH1, may play oncogenic roles critical for the regulation of MLL-rearranged leukaemia and could also possibly be targeted by DZNep. Genome-wide expression analyses would therefore provide more insight into the genes that are regulated by DZNep-treatment and account for the functional effects observed. Nevertheless, studies with DZNep have continued to emerge, extending to many types of cancer models including lung (Konishi et al., 2012), breast, prostate, colorectal, hepatoma (Tan et al., 2007), gastric cancer (Cheng et al., 2012), myeloma (Xie et al., 2011), AML (Fiskus et al., 2009) and lymphoma (Fiskus et al., 2012). DZNep appears to preferentially target cancer cells not only in AML compared to normal HSC (Zhou et al., 2011), but also noncancerous breast epithelial cells, lung 189 epithelial cells, primary human lung fibroblast cells and skin fibroblast cells (Tan et al., 2007), further showcasing DZNep as a promising epigenetic therapeutic agent.

Due to the complexity of the epigenetic landscape in cancer, the use of combination epigenetic therapy is of interest as it has been shown to enhance the anti-cancer effects of particular epigenetic inhibitors (Gore et al., 2006, Juergens et al., 2011). A study using ChIP assay examined the treatment of gastric and liver cancer cells with a combination of histone deacetylase inhibitor, SAHA, and DZNep which revealed that EZH2 was inhibited from binding to the promoter regions of a number of tumour suppressor microRNA and exerted its affects through activation of these microRNAs (Hibino et al., 2014). Reports have also emerged demonstrating enhanced efficacy against AML when DZNep is combined with the histone deacetylase inhibitor, panobinostat (Fiskus et al., 2009), or the combination of DZNep with histone deacetylase inhibitor, trichostatin-A, and DNA methyltransferase inhibitor, azacitidine (Momparler et al., 2014). This concept may also apply to histone demethylase inhibitors, however, it has not yet been studied in the context of AML. Furthermore, histone demethylase inhibitors with high specificity are currently limited.

Taken together, these results provide the first evidence of a potential role for DZNep in targeting stem cell-derived pre-LSC/LSC, which are often less responsive to chemotherapeutic agents compared to progenitor-derived cells (Krivtsov et al., 2012). This study also supports the notion that low doses of epigenetic agents are required to specifically target malignant stem cells through its ability to reverse aberrant cancer- specific methylation without causing significant cytotoxicity. Agents disrupting epigenetic programming of gene expression could potentially avoid chemotherapy- induced side effects, and indeed offer a more reliable and more effective therapeutic option for patients with poor prognosis cancers.

190

CHAPTER 6

Concluding Remarks

In addition to genetic alterations, epigenetic dysregulation is well established to be involved in the pathogenesis and development of cancer. This aberrant regulation has been reported to promote many of the hallmarks of cancer such as impaired growth inhibitory signals, evasion of apoptosis, increased proliferation potential, sustained angiogenesis, and the capability of metastasis and invasion (Esteller, 2008, Guil and Esteller, 2009). Since epigenetic regulation is reversible and dynamic, altering cancer- specific aberrant epigenetic modifications has emerged as a potential strategy for the treatment of cancer.

Increasing evidence suggests that epigenetic regulatory proteins such as histone demethylases are among those dysregulated in cancer, and that histone methylation can be altered globally or locally in different cancers (Jones and Baylin, 2007). The discovery of the first histone demethylase in 2004, H3K4 demethylase, KDM1a/LSD1, revealed that histone methylation, which was originally thought to be irreversible (Byvoet et al., 1972), was in fact reversible (Shi et al., 2004). Then in 2006, the first JmjC-domain-containing histone demethylase, JHDM1a/FBXL11/KDM2a, was discovered which converted di- and monomethylated H3K36 to the unmethylated form (Tsukada et al., 2006), followed by the discovery of JHDM2a/ Jmjd1a/KDM3a, found to promote the demethylation of H3K9me2 (Yamane et al., 2006). This class of epigenetic regulators that mediate the removal of methyl groups from different lysine residues on histones now comprise of twenty-eight members. Since the overexpression or mutation of a number of histone demethylases has been found in many types of cancer, the role of critical histone demethylases in the regulation of LSC was investigated in this thesis since it could provide potential diagnostic tools as well as epigenetic targets to be modulated by pharmaceutical agents as novel AML therapies.

191

6.1 Jmjd1c is required for the maintenance and proliferation of stem-cell derived leukaemia

At the commencement of this study the oncogenic role of Jmjd1c in AML had not been investigated, however, the identification of Jmjd1c as a downstream target of MLLAF9 and MLLAF4 (Bernt et al., 2011, Wilkinson et al., 2013, Guenther et al., 2008) suggested that Jmjd1c could be critical in the regulation of MLL-rearranged AML. In 2014, a paper was published where shRNA-mediated suppression of Jmjd1c impaired growth and colony formation of MLLAF9 cells in vitro, and also impaired leukaemia maintenance (Sroczynska et al., 2014). Although the data described in Chapter 3 are consistent with those in this recent publication, the cells used in this published paper were selected for a c-Kit-enriched population, which does not distinguish between stem cells and the more differentiated progenitors such as GMP. There was also no description in the paper of the survival data for the primary transplantation of these c- Kit-enriched MLLAF9 cells with Jmjd1c shRNA, which would reveal whether Jmjd1c is essential for the establishment of MLLAF9 AML. This thesis was able to show that Jmjd1c is required for the maintenance of stem cell-derived MLLAF9 AML rather than progenitor-derived MLLAF9 AML, as well as having a role in leukaemia maintenance but not initiation. This finding is supported by GSEA analysis by Sroczynska and co- workers, who identified enrichment of Jmjd1c target genes in LSC maintenance programs (Somervaille et al., 2009). Since Jmjd1c has been shown not to be crucial for the survival of HSC (Sroczynska et al., 2014) and it is a target of MLLAF4 (Wilkinson et al., 2013), a frequent MLL rearrangement found in leukaemias(Meyer et al., 2006, Krivtsov and Armstrong, 2007), studies into the Jmjd1c-deficient phenotype in MLLAF4 leukaemia would strengthen the data supporting the oncogenic role of Jmjd1c presented here in MLLAF9 and KLSA9M AML, and give impetus to develop Jmjd1c inhibitors for leukaemia therapy.

In extension to the results illustrated in the published paper, Chapter 3 provides evidence for the ability of Jmjd1c to enhance the proliferation of HSC, and accelerate disease latency of KLSA9M-derived leukaemia when ectopically expressed in these cell types. Furthermore, overexpression of Jmjd1c was associated with an increase in ATP production, likely through the promotion of metabolic pathways, such as glycolysis. Investigations into the glycolysis pathway are growing as this metabolic pathway has

192 been found to be induced by hypoxic microenvironments and enhanced in AML cells (Lodi et al., 2011). In this regard, future work should investigate the mechanism in which HIF signalling and Jmjd1c may be regulated, as hypoxia is a complex network involving feed-forward and feed-back activation. This would be essential in providing insight into the influence that Jmjd1c on the hypoxic microenvironment of leukaemic bone marrow (Mortensen et al., 1998, Colla et al., 2010) and therefore LSC growth and survival.

Although a demethylase activity of Jmjd1c was unable to be confirmed by western blotting, along with inconsistent reports of whether Jmjd1c possesses this ability, an alternative approach was to use a demethylase kit using K9me, K9me2 and K9me3 peptide as substrates. It was not feasible, however, to use this assay for primary AML cells due to the large cell number required to perform this assay. ChIP experiments would be recommended in future to determine whether the phenotype observed in both KLSMLLAF9 and KLSA9M was due to the direct transcriptional regulation of downstream target genes. Furthermore, many epigenetic regulators are known to cooperate and form complexes in order to recruit and bind to chromatin and regulate gene transcription (Atsumi et al., 2006, Brauchle et al., 2013, Kotake et al., 2007, Mohan et al., 2010, Neff et al., 2012, Shi et al., 2013, Tan et al., 2007, Zhang et al., 2006). Although Oct4 was identified as a partner of Jmjd1c to cooperate in the demethylation of target genes in embryonic stem cells (Shakya et al., 2015), it will be necessary to determine if this transcription factor or others is required for the demethyation of target genes in AML. The experiments undertaken in this study have expanded what is known about Jmjd1c and its role in regulating of AML LSC.

6.2 Jmjd5 plays a tumour suppressive role in leukaemogenesis

In addition to the identification of Jmjd1c, the results of this study suggest a tumour suppressor function for Jmjd5 in AML cells. Such a finding may need to be interpreted in a context-specific manner, as previous reports have described the tumourigenic function of Jmjd5 in breast cancer (Hsia et al., 2010). Nevertheless, other groups (Oh and Janknecht, 2012, Suzuki et al., 2006) along with this study support the tumour suppressor role of Jmjd5 in embryonic cells and Blm-deficient mice.

193

Conflicting reports have also emerged on the demethylase activity of Jmjd5, where some reports conclude that Jmjd5 functions as a histone demethylase (Ishimura et al., 2012, Hsia et al., 2010), while others show a lack of demethylase activity (Oh and Janknecht, 2012) or propose that it functions as a hydoxylase (Del Rizzo et al., 2012, Youn et al., 2012). This study observed a global reduction in H3K36 methylation, with a preference for dimethylation. A number of histone demethylases have been reported to function by hydroxylating the methyl groups of methyl lysines in histones and non- histone proteins, including JHDM1 (KDM2), JHDM2 (KDM3), Jmjd2 (KDM4), Jarid1, (KDM5), UTX/Jmjd3 (KDM6), and PHF8/KIAA1718 (KDM7) (Horton et al., 2010, Mosammaparast and Shi, 2010, Shi and Whetstine, 2007). Other JmjC enzymes which hydoxylate stereo- and site-specific residues in non-histone proteins include Jmjd6, a lysyl-5S-hydroxylase that oxidizes the splicing factor, U2AF65 (Webby et al., 2009, Mosammaparast and Shi, 2010), and factor inhibiting HIF-1 (FIH-1), an asparaginyl, aspartyl, and histidinyl hydroxylase whose substrates include HIF and ankyrin repeat (ANK)-containing proteins (Cockman et al., 2009, Schofield and Ratcliffe, 2004, Yang et al., 2011a, Yang et al., 2011b). Crystallography studies illustrate that Jmjd5 shares similar sequence and structural homology with Jmjd6 and FIH-1 but displays limited homology to JmjC lysine demethylases (Del Rizzo et al., 2012). Since JMJDs are 2- oxoglutarate-dependent dioxygenases that use Fe(II) in their active site to form a highly reactive oxoferryl (Fe(IV)=O) species, the substrates are hydroxylated by the enzyme which leads to demethylation (Trewick et al., 2005). Therefore, the consequence of hydroxylating a methyl group can also result in demethylation of mono-, di- or trimethylated histones. The results in this study did not investigate the mechanism in which Jmjd5 removes methyl groups from histones i.e. by demethylation or hydroxylation, however, it was observed that global H3K36 methylation levels were reduced upon enforced Jmjd5 expression and ChIP-qPCR data revealed the coding region of Tcf7l2 had lower H3K36me2 enrichment compared to control cells. Tcf7l2 was identified as the most frequently dysregulatd Wnt signalling component in AML (Daud et al., 2010) whose expression is found to be prognostic for reduced complete remission rates (Daud, 2014). Efforts to utilise ChIP-sequencing are in progress to confirm target genes identified in Chapter 4 and pathways such as Wnt signalling to be dysregulated by Jmjd5. Due to the observed association of H3K36me2 and H3K27me3, as well as the inverse expression of Jmjd5 and Ezh2 upon DOT1L inhibitor treatment in

194

KLSMLLAF9, it would be interesting to investigate this further and determine whether there is a functional link between these two epigenetic regulators and how they may contribute to disease development, as Ezh2 is known to play a role in enhancing proliferation and metastasis, blocking differentiation and is frequently mutated in haematopoietic malignancies (Bachmann et al., 2006, Tanaka et al., 2012, Ueda et al., 2014, Ernst et al., 2010, Nikoloski et al., 2010) where AML patients harbouring EZH2 mutations have an inferior survival (Ernst et al., 2012).

Another approach recommended for future experiments would be to investigate the leukaemia microenvironment. The endosteal niche, composed of mesenchymal stem cells, stromal cells, osteoprogenitors, osteoblasts and osteoclasts, supports the survival and maintenance of HSC, however, the identity of the cells that protects AML cells is unclear (Mihara et al., 2003, Lévesque et al., 2010, Ehninger and Trumpp, 2011, Singbrant et al., 2011). In addition, the main contributor to disease relapse in AML patients is the persistence of LSC as they are protected from chemotherapy by their interactions with the niche (Huntly and Gilliland, 2005). Researchers have also shown that AML can influence normal bone marrow cells to create a niche for leukaemia cells (Frisch et al., 2012). It is therefore of interest to investigate the involvement of Jmjd5 as it was implicated as a negative regulator of osteoclast differentiation (Youn et al., 2012). The nuclear factor of activated T-cells, cytoplasmic, calcineurin-dependent 1 (NFATc1), the primary osteoclastogenic transcription factor that attenuates osteoclastogenesis has been shown to promote cancer cell invasion (Oikawa et al., 2013). Jmjd5 posttranslationally corepresses NFATc1 by hydroxylation, facilitating its degradation (Youn et al., 2012). A number of approaches can be developed to target the microenvironment: blocking pro-survival and self-renewal pathways in LSC promoted by the niche; disrupting homing and adhesion of LSC to the niche through interference with chemokines and adhesion molecules; targeting the hypoxic leukaemic microenvironment; and inhibiting abnormally activated pathways within the cells of the niche (Figure 6.1) (Konopleva and Jordan, 2011).

195

CD44 Leukaemic Stem Cell

CXCR4 Blocking anti-apoptosis signalling MAPK

Receptor Affecting homing and adhesion antagonists PI3K PTEN Targeting leukaemic hypoxic milieu Integrin

ILK Akt HIF Integrin antagonists GSK3 Β-catenin Affecting self-renewal pathways

Wnt/ STAT NFB Frizzled/LRP 3 Blocking survival signalling PI3K Bcl2 Cytokine receptor MAPK Mcl1 Kinase inhibitors

Figure 6.1: Therapeutic targeting of leukaemic stem cell niche interactions Targeting interactions between LSC and the niche as an approach for AML therapy. Pro- survival signalling pathways (phosphatidylinositol-3 kinase (PI3K)/Akt, mitogen-activated protein kinase (MAPK), signal transducer and activator of transcription 3 (STAT3), and nuclear factor kappa B (NF-B)) activated by cytokines, chemokines and the extracellular matrix can be targeted to block the proliferation and survival of LSC. PTEN, phosphatase and tensin homolog; ILK, integrin-linked kinase; HIF-1, hypoxia-inducible transcription factor-1; GSK3, glycogen synthase kinase 3; LRP, leukocyte common antigen-related phosphatase; Bcl2, B-cell lymphoma 2; Mcl-1, myeloid cell leukemia-1. Adapted from (Konopleva and Jordan, 2011).

196

6.3 DZNep as an epigenetic agent in AML

Prior to this study, previous publications investigating DZNep used concentrations in the micromolar range and reported on its apoptotic activity. The data in this study suggests that apoptosis is only induced at high concentrations, while concentrations as low as 10 nM were sufficient to cause a significant reduction in H3K27me3. This was accompanied by the induction of different expression profiles in AML cells of various cellular origins and developmental stages. Furthermore, stem cell-derived AML cells were shown to be more sensitive to DZNep treatment than their progenitor counterpart. This was a favourable result as the stem cell-derived cells are often more drug resistant and difficult to target using traditional chemotherapeutics (Krivtsov et al., 2013). The characterisation of DZNep provides data supporting the therapeutic value of this drug as an AML therapeutic agent. Further studies into the interplay between different epigenetic marks should provide an insight into the associations between both active and repressive marks that in turn control transcriptional gene regulation and facilitate leukaemia transformation.

6.4 Potential therapeutic agents

Following the success of DNA methyltranferase and histone deacetylase inhibitors being tested in clinical trials (Bolden et al., 2006, Minucci and Pelicci, 2006) and the growing evidence showing the importance of histone demethylases in the maintenance of aggressive cancer cells, the development of histone demethylase inhibitors are emerging as promising candidates for epigenetic cancer therapies. Since LSD1 is an amino oxidase, several known inhibitors of monoamine and polyamine oxidases have been found to efficiently inhibit the histone demethylase activity of LSD1 (Huang et al., 2007b, Lee et al., 2006, Schmidt and McCafferty, 2007). The anti-cancer potential of these agents has not yet been evaluated. JMJDs, however, are structurally different from LSD1 and structural information is utilised to guide inhibitor design. This JmjC class of enzymes can be categorised into seven different subfamilies (KDM2, KDM3, KDM4, KDM5, JARID2, KDM6, KDM7 and KDM8) (Arrowsmith et al., 2012, Mosammaparast and Shi, 2010, Lohse et al., 2011a), based on a number of different domains for DNA-binding, substrate recognition, stabilisation and uncharacterised

197 domains (Lohse et al., 2011a). Due to the limited structure activity relationship data and limited biological information of these enzymes, as well as the identification of the first JMJD being less than a decade ago, potent and specific inhibitors of JMJD have not yet been identified (Rose et al., 2008, Rose et al., 2010, Thalhammer et al., 2011).

Potential JMJD inhibitors have been developed as analogues of JMJD substrates or by targeting the cofactors essential for JMJD activity, thereby inhibiting the demethylation activity (Lohse et al., 2011a, Suzuki and Miyata, 2011), however, this results in a lack of specificity and selectivity. Inhibitors of iron (II)-dependent and α-ketoglutarate- dependent enzymes have been identified to potentially inhibit JMJD (Schofield et al., 2010, Rose et al., 2008) (Figures 6.2 and 6.3). A mimetic of 2-oxoglutarate, N- oxalylglycine (NOG), was found to specifically inhibit the Zn(II) binding site of JMJD2A since other histone demethylases do not contain Zn-binding sites near the catalytic domain (Couture et al., 2007, Horton et al., 2010, Horton et al., 2011, Sengoku and Yokoyama, 2011, Sekirnik et al., 2009). 8-Hydroxy-5-carboxyquinoline was also identified as a potent inhibitor of JMJD2 via the chelation of the Fe ion in the active site (Oh and Janknecht, 2012, King et al., 2010). Due to the similar sequence identity and zinc binding capabilities of the JMJD2 enzymes and histone deacetylases, the histone deacetylase inhibitor, SAHA, was found to be an inhibitor of JMJD2E with activity in the micromolar range (Rose et al., 2008). Structural studies have shown that the JmjC catalytic domain of KDM2A, KDM4A, KDM6A and KDM7 share a high similarity (Hoffman et al., 2012). Designing specific inhibitors of JMJD1C (KDM2C) and JMJD5 (KDM8) may therefore be easier as the residues of the cofactor binding site are more distinct from the other subfamilies and therefore offer potential for a selective structure based drug design. A combination of structure-guided design, virtual screening and high-throughput screening, will undoubtedly assist in the development of novel Jumonji-specific inhibitor compounds.

198

Figure 6.2: Chemical structures of iron-dependent enzyme inhibitors Inhibitor scaffolds based on other iron-dependent enzymes and their optimised derivatives. Adapted from (Hoffman et al., 2012).

199

Figure 6.3: Chemical structures of JMJD inhibitors (A) JMJD inhibitors based on disulfiram derivative and hydroxyglutarate identified from screening natural products libraries and a diverse chemical library. (B) Other chemical structures of more recently discovered small-molecule inhibitors of JMJD. Figure adapted from (Hoffman et al., 2012). 200

Another advantage of targeting epigenetic processes is that it can potentiate the effects of traditional chemotherapy and radiation (Ellis et al., 2009). A number of epigenetic therapies have been shown to lower the apoptotic threshold of tumour cells, sensitising the cells towards agents that target microtubule stability (docetaxel, paclitaxel), proteosomal degradation (bortezomib), Bcr-Abl tyrosine kinase (imatinib/Gleevec®), and molecular protein chaperones (geldanomycin) (Bolden et al., 2006, Fiskus et al., 2006, Xu et al., 2007, Verheul et al., 2008, Drummond et al., 2005). Clinical investigations into the combination of epigenetic therapies with targeted anticancer treatments are currently underway (Drummond et al., 2005).

Data from Chapter 3 implicated the glycolysis pathway in playing a critical role in AML. Understanding the biological differences between normal and cancer cells has led to the suggestion that the glycolysis pathway may be inhibited to preferentially target malignant cells, as ATP generation via glycolysis is less efficient than through oxidative phosphorylation, and cancer cells have greater consumption of glucose than normal cells (Munoz-Pinedo et al., 2003, Izyumov et al., 2004, Xu et al., 2005). A number of inhibitors targeting various molecules in the glycolysis pathway are being tested in clinical trials. 2-Deoxyglucose acts as a competitive inhibitor of glucose metabolism, due to it being a glucose analogue that does not get metabolised by phosphohexose isomerase and is therefore accumulated in the cell, leading to inhibition of glycolysis and depletion in cellular ATP (Maher et al., 2004, Weindruch et al., 2001, Brown, 1962). This inhibitor, however, only partially reduces the availability of glucose as the presence of glucose is still evident in the cells (Pelicano et al., 2006). A derivative of indazole-3-carboxylic acid, lonidamine, has been shown to suppress glycolysis in cancer through inhibition of hexokinase (Floridi et al., 1981). Treatment of B-cell chronic leukaemia with lonidamine resulted in decreased lactate production (Natali et al., 1984) and enhancement in the activity of other anticancer agents, resulting in this drug being tested in clinical trials (Phase II/III) for the treatment of glioblastoma, breast, ovarian and lung cancer as well as prostatic hyperplasia (De Lena et al., 2001, Di Cosimo et al., 2003, Oudard et al., 2003).

Imatinib is a compound that targets Bcr-Abl, an aberrantly expressed tyrosine kinase responsible for the development of CML. These cells express high levels of Glut-1, the molecule responsible for glucose uptake. Treatment with imatinib suppresses aerobic

201 glycolysis by decreasing the activity of hexokinase and glucose-6-phosphate dehydrogenase (Boren et al., 2001, Gottschalk et al., 2004, Serkova and Boros 2005). ATP production is reduced not only from the suppressed glycolysis but also by the reduction in metabolic intermediates required for the pentose pathway because of the decreased activity of glucose-6-phosphate dehydrogenase. Another inhibitor of hexokinase is 3-bromopyruvate, which has been shown to reduce ATP production (Xu et al., 2005, Geschwind et al., 2004, Ko et al., 2001) and possesses therapeutic activity against liver cancer (Geschwind et al., 2004, Ko et al., 2001). A thiamine antagonist and inhibitor of transketolase and pyruvate dehydrogenase is oxythiamine (Strumilo et al., 1984). These enzymes are crucial for the pentose phosphate pathway and require thiamine pyrophosphate as a cofactor for their activity. Oxythiamine acts as a competitive inhibitor leading to suppression of the pentose phosphate pathway and a reduction in the metabolic intermediates required for ATP generation. This compound has shown anticancer activity against Ehrlich ascite tumours in vitro and in vivo (Rais et al., 1999, Comin-Anduix et al., 2001). In order to prevent potential side effects arising from the use of glycolysis inhibitors, animal toxicity studies need to be considered since the brain, retina and testis use glucose as their main energy source. Preliminary studies have shown glycolysis inhibitors to be effective and highly cytotoxic to cancer cells under hypoxia (Pelicano et al., 2006), providing evidence that these inhibitors have the potential to be used as anti-cancer agents against AML.

Another approach to treating AML is by examining the hypoxic and acidic microenvironment. Upregulation of glycolysis leads to hypoxia and then acidosis, resulting in the acceleration of metastasis and resistance to therapeutic strategies, such as radiotherapy and anthracyclines (Secomb et al., 1998). Inhibitors of Hif1 and Pdk1, whose expression is mediated by Hif1, have the potential to reverse these tumourigenic effects. The Pdk1 inhibitor, dichloroacetic acid (DCA), and the Hif1 inhibitor, echinomycin, are under investigation to evaluate their therapeutic potential in haematological malignancies (Semenza, 2010a, Michelakis et al., 2008, Cairns et al., 2009, Onnis et al., 2009, Fujiwara et al., 2013). The inhibition of Hif1 is proposed to have minimal side effects as it has low activity in normal well-oxygenated tissues.

An alternative treatment of AML would be to focus on downstream targets of Jmjd5 such as those identified in Chapter 4. By targeting LSC-specific genes, these therapeutic

202 agents should produce a wider therapeutic index and hence demonstrate lower toxicity to normal cells at physiologic dose levels. Although there are currently no available inhibitors for the majority of targets identified, an antagonist of Gpr84, GLPG1205 (Galapagos®), is currently being investigated in Phase II clinical trials for inflammatory bowel disease, where Phase I trials revealed a favourable tolerability to the drug (Dupont et al., 2014, Vanhouttee et al., 2014). Hence, targeting the Gpr84 signalling axis may also be a viable approach for targeting AML LSC.

6.5 Conclusion

Prior to the commencement of this study, the role of JMJD1C and JMJD5 in AML, as well as the phenotype and gene expression profile of AML cells following treatment with DZNep at concentrations in the nanomolar range, was not known. The experiments undertaken in this thesis have expanded what is known about epigenetic regulators of AML LSC. This study investigated the role of Jmjd1c and Jmjd5 in AML LSC, and the importance in considering dosing when using epigenetic drugs as cancer therapeutics to ensure the effects from the drug are due to epigenetic mechanisms. The data presented in this study support the development of inhibitors against Jmjd1c for the treatment of AML in an effort to control its activity, as it acts as a positive effector of glycolysis and enhances the leukaemogenicity of KLSA9M LSC. Moreover, this study provides data on the tumour suppressor role of Jmjd5 in AML and the mechanism in which Jmjd5 mediates the suppression of leukaemia. These two novel histone demethylases are therefore likely to selectively eliminate AML LSC whilst sparing HSC, preventing potential toxic side effects. Data from these experiments has helped elucidate the biological targets of Jmjd1c and Jmjd5, and has provided a platform from which their molecular mechanisms of action can be determined with relevance to the regulation of LSC and AML development. Future work on these two histone demethylases will build upon the data reported in this thesis, and is likely to provide insight into the tightly regulated epigenetic program in which these epigenetic regulators repress or activate the transcription of downstream target genes.

203

SUPPLEMENTARY FIGURES

8 a to ta l o n ly A B 1 .2

Jmjd1c-/- Jmjd1c-/+

n o

) 0 .9

e s

EV Cre EV Cre g

e *

p h

x 0 .6

Jmjd1c

o N

f *** ( R 0 .3 Actin m

0 .0 E V C re E V C re

-/+ -/- Jmjd1c Jmjd1c

Figure S.1: Generation of an inducible Jmjd1c knockout MLLAF9 AML model Suppression of Jmjd1c expression in Jmjd1c heterozygous (Jmjd1c-/+) and homozygous (Jmjd1c-/-) knockouts transduced with Cre recominbase (Cre) or control (EV) was analysed by (A) western blotting and (B) qRT-PCR. Data represent mean values ± SEM of three independent experiments. Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (* P < 0.05, *** P < 0.0005).

204

s u p p l s .2 1 c -s c d n a e x p ***

6 0

)

s n

s 4 0

i t

N 2 0

m (

0 E V Jm jd 1 c -S

Figure S.2: Confirming overexpression of Jmjd1c-S in KLSA9M by qRT-PCR KLSA9M pre-LSC were transduced with Jmjd1c-S or EV as a control and the mRNA levels confirm the overexpression of Jmjd1c-S by qRT-PCR. Data represent mean values ± SEM of three independent experiments Asterisks indicate statistical significance obtained using an unpaired Student‘s t-test (*** P < 0.0005).

205

Gpr84 Gpr84 Gpr84 A Promoter Exon 2.1 Exon 2.2 8

2 Percent ofPercent input 0 Ctr Jmjd5 Ctr Jmjd5 Ctr Jmjd5

β-catenin β-catenin β-catenin B Promoter Exon 3 Exon 9 8

2 Percent ofPercent input 0 Ctr Jmjd5 Ctr Jmjd5 Ctr Jmjd5

Tcf7l2 Tcf7l2 Tcf7l2 C Promoter Exon 5 Exon 8

2 Percent ofPercent input 0 Ctr Jmjd5 Ctr Jmjd5 Ctr Jmjd5

Figure S.3: H3K36me2 enrichment at Jmjd5 target genes using ChIP-qPCR Quantifying H3K36me2 enrichment in various regions of (A) Gpr84 (promoter, beginning of exon 2 and middle of exon 2), (B) β-catenin (promoter, exon 3 and exon 9) and (C) Tcf7l2 (promoter, exon 5 and exon 8) by ChIP-qPCR in control versus Jmjd5 overexpressing KLSMLLAF9 pre-LSC.

206

Table S.1: Limiting dilution assay of KLSMLLAF9 and KLSA9M leukaemia upon loss or gain of Jmjd1c Limiting dilution analysis for estimating the frequency of leukaemic stem cells.

Dose Response/Tested P-value Scramble Jmjd1c shRNA Jmjd1c shRNA 2 4 100000 12/12 9/9 13/13 < 0.0001 10000 12/12 6/10 8/12 < 0.0001 1000 12/12 1/11 0/12 < 0.0001 LSC frequency 1 in 1 1 in 10986 1 in 10719 Confidence intervals 1/1 – 1/5154 – 1/5347 – 1/1071 1/23418 1/21488 Dose Response/Tested P-value EV Jmjd1c cDNA 100000 7/11 14/14 < 0.0001 10000 6/9 9/9 < 0.0001 1000 0/7 5/5 0.0003 LSC frequency 1 in 55040 1 in 1 Confidence intervals 1/28281 – 1/107116 1/1 – 1/1254

207

Table S.2: Transcripts of Jmjd1c in Mus musculus The gene and protein size of the eleven transcripts of Jmjd1c in mice. Five of these transcripts encode for a protein, where four of these containing the JmjC demethylases functional domain. Transcripts Gene (bp) Protein (aa) Contains JmjC domain 1 8609 2350 Yes 2 8382 2530 Yes 3 8377 2531 Yes 4 3076 744 No, 3‘ truncated 5 2015 578 Yes, 5‘ truncated 6 2772 No protein 7 2574 No protein 8 2365 No protein 9 1579 No protein 10 651 No protein 11 522 No protein

208

Table S.3: Limiting dilution assay of KLSMLLAF9 leukaemia upon gain of Jmjd5 Limiting dilution analysis for estimating the frequency of leukaemic stem cells.

Dose Response/Tested P-value Control Jmjd5 cDNA 100000 7/7 5/7 0.0033 10000 6/6 2/6 0.0005 1000 6/6 0/6 0.0005 LSC frequency 1 in 1 1 in 63201 Confidence intervals 1/1 – 1/1071 1/27409 – 1/145733

209

REFERENCES

ABDEL ALIM, A., YOUSSEF, S. & FAROUK, H. 2010. The prognostic potential of bone marrow cyclin E and P27, in Egyptian patients with acute myeloid leukemia. Egypt J Immunol, 17, 9-18. ABRAMOFF, M., MAGALHAES, P. & RAM, S. 2004. Image processing with ImageJ. Biophotonics International, 11, 36-42. ADOLFSSON, J., BORGE, O., BRYDER, D., THEILGAARD-MONCH, K., ASTRAND-GRUNDSTROM, I., SITNICKA, E., SASAKI, Y. & JACOBSEN, S. 2001. Upregulation of Flt3 expression within the bone marrow Lin- Sca1+ c- kit+ stem cell compartment is accompanied by loss of self-renewal capacity. Immunity, 15, 659-669. AHN, S., DIAZ, R., GRUNSTEIN, M. & ALLIS, C. 2006. Histone H2B deacetylation at lysine 11 is required for yeast apoptosis induced by phosphorylation of H2B at serine 10. Mol Cell, 24, 211-220. AL-HAJJ, M., WICHA, M., BENITO-HERNANDEZ, A., MORRISON, S. & CLARKE, M. 2003. Prospective identification of tumorigenic breast cancer cells. Proc Natl Acad Sci USA, 100, 3983-3988. ALLFREY, V., FAULKNER, R. & MIRSKY, A. 1964. Acetylation and methylation of histones and their possible role it the regulation of RNA synthesis. Proc Natl Acad Sci USA, 51, 786-794. AMANN, J., NIP, J., STROM, D., LUTTERBACH, B., HARADA, H., LENNY, N., DOWNING, J., MEYERS, S. & HIEBERT, S. 2001. ETO, a target of t(8;21) in acute leukemia, makes distinct contacts with multiple histone deacetylases and binds mSin3A through its oligomerization domain. Mol Cell Biol, 21, 6470-83. AMERICAN CANCER SOCIETY 2015. Cancer Facts and Figures 2015. ANGELESCU, S., BERBEC, N., COLITA, A., BARBU, D. & LUPU, A. 2012. Value of Multifaced Approach Diagnosis and Classification of Acute Leukemias. Maedica (Buchar), 7, 254-260. ANNUNZIATO, A., EASON, M. & PERRY, C. 1995. Relationship between methylation and acetylation of arginine-rich histones in cycling and arrested HeLa cells. Biochem, 34, 2916-24.

210

ARMENDARIZ, H., BARBIERI, M., FREIGEIRO, D., LASTIRI, F., FELICE, M. & DIBAR, E. 2005. Treatment strategy and long-term results in pediatric patients treated in two consecutive AML-GATLA trials. Leukemia, 19, 2139-2142. ARMSTRONG, L. 2014. Epigenetic control of the mitotic cycle. Epigenetics. Abingdon: Garland Science, Taylor & Frnacis group. ARROWSMITH, C., BOUNTRA, C., FISH, P., LEE, K. & SCHAPIRA, M. 2012. Epigenetic protein families: a new frontier for drug discovery. Nat Rev Drug Discov, 11, 384-400. ASANGANI, I., ATEEQ, B., CO, Q., DODSON, L., PANDHI, M., KUNJU, L., MEHRA, R., LONIGRO, R., SIDDIQUI, J., PALANISAMY, N., WU, Y., CAO, X., KIM, J., ZHAO, M., QIN, Z., IYER, M., MAYER, C., KUMAR- SINHA, C., VARAMBALLY, S. & CHINNAIYAN, A. 2013. Characterization of the EZH2-MMSET Histone Methyltransferase Regulatory Axis in Cancer. Mol Cell, 49, 80-93. ATSUMI, A., TOMIA, A., KIYOI, H. & NAOE, T. 2006. Histone deacetylase 3 (HDAC3) is recruited to target promoters by PML-RARalpha as a component of the N-CoR co-repressor complex to repress transcription in vivo. Biochem Biophys Res Commun, 345, 1471-80. AYTON, P. & CLEARY, M. 2001. Molecular mechanisms of leukemogenesis mediated by MLL fusion proteins. Oncogene, 20, 5695-707. AYTON, P. & CLEARY, M. 2003. Transformation of myeloid progenitors by MLL oncoproteins is dependent on Hoxa7 and Hoxa9. Genes Dev, 17. BACHMANN, I. M., HALVORSEN, O. J., COLLETT, K., STEFANSSON, I. M., STRAUME, O., HAUKAAS, S. A., SALVESEN, H. B., OTTE, A. P. & AKSLEN, L. A. 2006. EZH2 expression is associated with high proliferation rate and aggressive tumor subgroups in cutaneous melanoma and cancers of the endometrium, prostate, and breast. J Clin Oncol, 24, 268-273. BAEK, G., TSE, Y., HU, Z., COX, D., BUBOLTZ, N., MCCUE, P., YEO, C., WHITE, M., DEBARARDINIS, R., KNUDSEN, E. & WITKIEWICZ, A. 2014. MCT4 Defines a Glycolytic Subtype of Pancreatic Cancer with Poor Prognosis and Unique Metabolic Dependencies. Cell Rep, 9, 2233-49.

211

BAINS, O., KARKLING, M., LUBIENIECKA, J., GRIGLATTI, T., REID, R. & RIGGS, K. 2010. Naturally occurring variants of human CBR3 alter anthracycline in vitro metabolism. J Pharmacol Exp Ther, 332, 755-63. BALGOBIND, B., ZWAAN, C., PIETERS, R. & VAN DEN HEUVEL-EIBRINK, M. 2011. The heterogeneity of pediatric MLL-rearranged acute myeloid leukemia. Leukemia, 25, 1239-1248. BALGOBIND, B., ZWAAN, C., REINHARDT, D., ARENTSEN-PETERS, T., HOLLINK, I., DE HAAS, V., KASPERS, G., DE BONT, E., BARUCHEL, A., STARY, J., MEYER, C., MARSCHALEK, R., CREUTZIG, U., DEN BOER, M., PIETERS, R. & VAN DEN HEUVEL-EIBRINK, M. 2010. High BRE expression in pediatric MLL-rearranged AML is associated with favorable outcome. Leukemia, 24, 2048-2065. BALI, P., PRANPAT, M., BRADNER, J., BALASIS, M., FISKUS, W., GUO, F., ROCHA, K., KUMARASWAMY, S., BOYAPALLE, S. A., P, SETO, E. & BHALLA, K. 2005. Inhibition of histone deacetylase 6 acetylates and disrupts the chaperone function of heat shock protein 90: a novel basis for antileukemia activity of histone deacetylase inhibitors. J Biol Chem, 280, 26729-34. BANNISTER, A., SCHNEIDER, R., MYERS, F., THORNE, A., CRANE- ROBINSON, C. & KOUZARIDES, T. 2005. Spatial Distribution of Di- and Tri- methyl Lysine 36 of Histone H3 at Active Genes. J Biol Chem, 280, 17732- 17736. BANNISTER, A. J. & KOUZARIDES, T. 2011. Regulation of chromatin by histone modifications. Cell Res, 21, 381-95. BARKER, N., HUCH, M., KUJALA, P., VAN DE WETERING, M., SNIPPERT, H., VAN ES, J., SATO, T., STRANGE, D., BEGTHEL, H., VAN DEN BORN, M., DANENBERG, E., VAN DEN BRINK, S., KORVING, J., ABO, A., PETERS, P., WRIGHT, N., POULSOM, R. & CLEVERS, H. 2010. Lgr5(+ve) stem cells drive self-renewal in the stomach and build long-lived gastric units in vitro. Cell Stem Cell, 6, 25-36. BARKER, N., RIDGWAY, R., VAN ES, J., VAN DE WETERING, M., BEGTHEL, H., VAN DEN BORN, M., DANENBERG, E., CLARKE, A., SANSOM, O. & CLEVERS, H. 2009. Crypt stem cells as the cells-oforigin of intestinal cancer. Nature, 457, 608-11.

212

BENNETT, J., CATOVSKY, D., DANIEL, M., FLANDRIN, G., GALTON, D., GRALNICK, H. & SULTAN, C. 1985. Proposed revised criteria for the classification of acute myeloid leukemia. A report of the French-American- British Cooperative Group. Ann Intern Med, 103, 620-625. BERNSTEIN, B., MEISSNER, A. & LANDER, E. 2007. The mammalian epigenome. Cell, 128, 669-81. BERNT, K., ZHU, N., SINHA, A., VEMPATI, S., FABER, J., KRIVTSOV, A., FENG, Z., PUNT, N., DAIGLE, A., BULLINGER, L., POLLOCK, R., RICHON, V., KUNG, A. & ARMSTRONG, S. 2011. MLL-rearranged leukemia is dependent on aberrant H3K79 methylation by DOT1L. Cancer Cell, 20, 66-78. BESA, E. 1992. Neoplastic Disorders of Granulocytes, Monocytes and Lymphocytes. In: BESA, E., CATALANO, P., KANT, J. & JEFFERIES, L. (eds.) Hematology. Lippincott Williams And Wilkins. BEYER, S., KRISTENSEN, M., JENSEN, K., JOHANSEN, J. & STALLER, P. 2008. The Histone Demethylases JMJD1A and JMJD2B Are Transcriptional Targets of Hypoxia-inducible Factor HIF. J Biol Chem, 283, 36542-36552. BHALLA, K. 2005. Epigenetic and chromatin modifiers as targeted therapy of hematologic malignancies. J Clin Oncol, 23, 3971-93. BHARDWAJ, G., MURDOCH, B., WU, D., BAKER, D., WILLIAMS, K., CHADWICK, K., LING, L., KARANU, F. & BHATIA, M. 2001. Sonic hedgehog induces the proliferation of primitive human hematopoietic cells via BMP regulation. Nat Immunol, 2, 172-80. BIONDI, A., CIMINO, G., PIETERS, R. & PUI, C. H. 2000. Biological and therapeutic aspects of infant leukemia. Blood, 96, 24-33. BIRD, A. 2002. DNA methylation patterns and epigenetic memory. Genes Dev, 16, 6- 21. BIRD, A. P. & WOLFFE, A. P. 1999. Methylation-induced repressio—belts, braces, and chromatin. Cell, 99, 451-4. BISWAS, D., DUTTA-BISWAS, R., MITRA, D., SHIBATA, Y., STRAHL, B., FORMOSA, T. & STILLMAN, D. 2006. Opposing roles for Set2 and yFACT in regulating TBP binding at promoters. EMBO J, 25, 4479-89.

213

BLAIR, A., HOGGE, D., AILLES, L., LANSDORP, P. & SUTHERLAND, H. 1997. Lack of expression of Thy-1 (CD90) on acute myeloid leukemia cells with long- term proliferative ability in vitro and in vivo. Blood, 89, 3104-3112. BLANDER, G. & GUARENTE, L. 2004. The Sir2 family of protein deacetylases. Annu Rev Biochem, 73, 417-35. BOLDEN, J., PEART, M. & JOHNSTONE, R. 2006. Anticancer activities of histone deacetylase inhibitors. Nat Rev Drug Discov, 5, 769-784. BONILLA, M. & RIBEIRO, R. 2014. Acute Myeloid Leukemia. In: STEFAN, D. & RODRIGUEZ-GALINDO, C. (eds.) Pediatric Hematology-Oncology in Countries with Limited Resources New York: Springer. BONMASSAR, L., MARCHESI, F., PASCALE, E., FRANZESE, O., MARGISON, G., BIANCHI, A., D'ATRI, S., BERNARDINI, S., LATUADA, D., BONMASSAR, E. & AGUINO, A. 2013. Triazene compounds in the treatment of acutee myeloid leukemia: a short review and a case report. Curr Med Chem, 20, 2389-401. BONNET, D. & DICK, J. 1997. Human acute myeloid leukaemia is organised as a hierarchy that originates from a primitive haematopoietic cell. Nat Med, 3, 730- 737. BOREN, J., CASCANTE, M., MARIN, S., COMIN-ANDUIX, B., CENTELLES, J., LIM, S., BASSILIAN, S., AHMED, S., LEE, W. & BOROS, L. 2001. Gleevec (STI571) Influences Metabolic Enzyme Activities and Glucose Carbon Flow toward Nucleic Acid and Fatty Acid Synthesis in Myeloid Tumor Cells. J Biol Chem, 276, 37747-53. BORUN, T., PEARSON, D. & PAIK, W. 1972. Studies of histone methylation during the HeLa S-3 cell cycle. J Biol Chem, 247, 4288-4298. BOUCHARD, C., THIEKE, K., MAIER, A., SAFFRICH, R., HANLEY-HYDE, J., ANSORGE, W., REED, S., SICINSKI, P., J, B. & EILERS, M. 1999. Direct induction of cyclin D2 by Myc contributes to cell cycle progression and sequestration of p27. EMBO J, 18, 5321-5333. BOYER, L., LEE, T., COLE, M., JOHNSTONE, S., LEVINE, S., ZUCKER, J., GUENTHER, M., KUMAR, R., MURRAY, H., JENNER, R., GIFFORD, D., MELTON, D., JAENISCH, R. & YOUNG, R. 2005. Core transcriptional regulatory circuitry in human embryonic stem cells. Cell, 122, 947-956.

214

BRACKEN, A., DIETRICH, N., PASINI, D., HANSEN, K. & HELIN, K. 2006. Genome-wide mapping of Polycomb target genes unravels their roles in cell fate transitions. Genes Dev, 20, 1123-36. BRACKEN, A., KLEINE-KOHLBRECHER, D., DIETRICH, N., PASINI, D., GARGIULO, G., BEEKMAN, C., THEILGAARD-MONCH, K., MINUCCI, S., PORSE, B., MARINE, J., HANSEN, K. & HELIN, K. 2007. The Polycomb group proteins bind throughout the INK4A-ARF locus and are disassociated in senescent cells. Genes Dev, 21, 525-30. BRADFORD, G., WILLIAMS, B., ROSSI, R. & BERTONCELLO, I. 1997. Quiescence, cycling, and turnover in the primitive hematopoietic stem cell compartment. Exp Hematol, 25, 445-453. BRAKENSIEK, K., LANGER, F., KREIPE, H. & LEHMANN, U. 2005. Absence of p21(CIP 1), p27(KIP 1) and p 57(KIP 2) methylation in MDS and AML. Leuk Res, 29, 1357-60. BRANDWEIN, J., KASSIS, J., LEBER, B., HOGG, E. D., HOWSON-JAN, K., MINDEN, M., GALARNEAU, A. & POULIOT, J. 2014. Phase II study of targeted therapy with temozolomide in acute myeloid leukaemia and high-risk myelodysplastic syndrome patients pre-screened for low O(6) -methylguanine DNA methyltransferase expression. Br J Haematol, 167, 664-70. BRAUCHLE, M., YAO, Z., ARORA, R., THIGALE, S., CLAY, I., INVERARDI, B., FLETCHER, J., TASLIMI, P., ACKER, M., GERRITS, B., VOSHOL, J., BAUER, A., SCHÜBELER, D., BOUWMEESTER, T. & RUFFNER, H. 2013. Protein complex interactor analysis and differential activity of KDM3 subfamily members towards H3K9 methylation. PLoS One, 8, e60549. BROWN, J. 1962. Effects of 2-deoxyglucose on carbohydrate metablism: review of the literature and studies in the rat. Metabolism, 11, 1098-1112. BRUSERUD, O., GJERTSEN, B. & HUANG, T. 2000. Induction of Differentiation and Apoptosis— A Possible Strategy in the Treatment of Adult Acute Myelogenous Leukemia. Oncologist, 5, 454-62. BRYDER, D., ROSSI, D. & WEISSMAN, I. 2006. Hematopoietic stem cells: The paradigmatic tissue-specific stem cell. Am J Pathol, 169, 338-346. BURNETT, A. 2012. New induction and postinduction strategies in acute myeloid leukemia. Curr Opin Hematol, 19, 76-81.

215

BURNETT, A., WETZLER, M. & LÖWENBERG, B. 2011. Therapeutic advances in acute myeloid leukemia. J Clin Oncol, 29, 487-494. BYVOET, P., SHEPHARD, G., HARDIN, J. & NOLAND, B. 1972. The distribution and turnover of labelled methyl groups in hsitone fractions of cultured mammalian cells. Arch Biochem Biophys, 148, 558-67. CAI, M., LANGER, E., GILL, J., SATPATHY, A., ALBRING, J., KC, W., MURPHY, T. & MURPHY, K. 2012. Dual actions of Meis1 inhibit erythroid progenitor development and sustain general hematopoietic cell proliferation. Blood, 120, 335-46. CAIRNS, R., HARRIS, I. & MAK, T. 2011. Regulation of cancer cell metabolism. Nat Rev Cancer, 11, 85-95. CAIRNS, R. A., BENNEWITH, K., GRAVES, E., GIACCAI, A., CHANG, D. & DENKO, N. C. 2009. Pharmacologically increased tumor hypoxia can be measured by 18F‑ Fluoroazomycin arabinoside positron emission tomography and enhances tumor response to hypoxic cytotoxin. Clin Cancer Res, 15, 7170- 4. CALVI, L., ADAMS, G., WEIBRECHT, K., WEBER, J., OLSON, D., KNIGHT, M., MARTIN, R., SCHIPANI, E., DIVIETI, P., BRINGHURST, F., MILNER, L., KRONENBERG, H. & SCADDEN, D. 2003. Osteoblastic cells regulate the haematopoietic stem cell niche. Nature, 425, 841-6. CALVO, K., KNOEPFLER, P., SYKES, D., PASILLAS, M. & KAMPS, M. 2001. Meis1a suppresses differentiation by G-CSF and promotes proliferation by SCF: potential mechanisms of cooperativity with Hoxa9 in myeloid leukemia. Proc Natl Acad Sci USA, 98, 13120-5. CAMOS, M., ESTEVE, J., JARES, P., COLOMER, D., ROZMAN, M., VILLAMOR, N., COSTA, D., CARRIO, A., NOMDEDEU, J., MONTSERRAT, E. & CAMPO, E. 2006. Gene expression profiling of acute myeloid leukemia with translocation t(8;16)(p11;p13) and MYST3-CREBBP rearrangement reveals a distinctive signature with a specific pattern of HOX gene expression. Cancer Res, 66, 6947-6954. CANADIAN CANCER SOCIETY 2015. Subtypes of childhood AML.

216

CANCER GENOME ATLAS RESEARCH NETWORK 2013. Genomic and epigenomic landscapes of adult de novo acute myeloid leukemia. N Eng J Med, 368, 2059-74. CAO, R., WANG, L., WANG, H., XIA, L., ERDJUMENT-BROMAGE, H., TEMPST, P., JONES, R. & ZHANG, Y. 2002. Role of histone H3 lysine 27 methylation in Polycomb-group silencing. Science, 298, 1039-43. CAO, R. & ZHANG, Y. 2004. The functions of E(Z)/EZH2-mediated methyation of lysine 27 in histone H3. Curr Opin Genet Dev, 14, 155-164. CARLBERG, C. & MOLNAR, F. 2014. Chromatin Modifiers. Mechanisms of Gene Regulation. Finland: Springer Science. CATALANO, A., RODILOSSI, S., CAPRARI, P., COPPOLA, V. & PROCOPIO, A. 2005. 5-Lipoxygenase regulates senescence like growth arrest by promoting ROS-dependent p53 activation. EMBO J, 24, 170-9. CATOVSKY, D., GALTON, D. & ROBINSON, J. 1972. Myeloperoxidase-Deficient Neutrophils in Acute Myeloid Leukaemia. Scand J Haematol, 9, 142-8. CERVEIRA, N., CORREIA, C., DORIA, S., BIZARRO, S., ROCHA, P., GOMES, P., TORRES, L., NOORTON, L., BORGES, B., CASTEDO, S. & TEIXEIRA, M. 2003. Frequency of NUP98‑ NSD1 fusion transcript in childhood acute myeloid leukaemia. Leukemia, 17. CHALLEN, G., BOLES, N., LIN, K. & GOODELL, M. 2009. Mouse hematopoietic stem cell identification and analysis. Cytometry Part A, 75, 14-24. CHANDEL, N. S. & SCHUMACKER, P. T. 2000. Cellular oxygen sensing by mitochondria: old questions, new insight. J Appl Physiol, 88. CHANG, M., WU, H., ACHILLE, N., REISENAUER, M. R., CHOU, C., ZELEZNIK- LE, N., HEMENWAY, C. & ZHANG, W. 2010. Histone H3 lysine 79 methyltransferase Dot1 is required for immortalization by MLL oncogenes. Cancer Res, 70, 10234-10242. CHEN, C., LI, M., DE GRAAF, D., MONTI, S., GOTTGENS, B., SANCHEZ, M., LANDER, E., GOLUB, T., GREEN, A. & LODISH, H. 2002a. Identification of endoglin as a functional marker that defines long-term repopulating hematopoietic stem cells. Proc Natl Acad Sci USA, 99, 15468-15473.

217

CHEN, W., KUMAR, A., HUDSON, W., LI, Q., WU, B., STAGGS, R., LUND, E., SAM, T. & KERSEY, J. 2008a. Malignant transformation initiated by Mll-AF9: gene dosage and critical target cells. Cancer Cell, 13, 432-440. CHEN, X., SHELLER, J., JOHNSON, E. & FUNK, C. 1994. Role of leukotrienes revealed by targeted disruption of the 5-lipoxygenase gene. Nature, 372, 179- 182. CHEN, X., YANG, L., OPPENHEIM, J. & HOWARD, M. 2002b. Cellular pharmacology studies of shikonin derivatives. Phytother Res, 16, 199-209. CHEN, Y., HSU, H., CHEN, Y., TSAI, T., HOW, C., WANG, C., HUNG, S., CHANG, Y., TSAI, M., LEE, Y., KU, H. & CHIOU, S. 2008b. Oct-4 expression maintained cancer stem-like properties in lung cancer-derived CD133-positive cells. PLoS One, 3, 2637. CHEN, Y., HU, Y., ZHANG, H., PENG, C. & LI, S. 2009. Loss of the Alox5 gene impairs leukemia stem cells and prevents chronic myeloid leukemia. Nat Genet, 41, 783-792. CHENG, C., HOU, H., LEE, M., LIU, C., JHUANG, J., LAI, Y., LIN, C., CHEN, H., LIU, F., CHOU, W., CHEN, C., TANG, J., YAO, M., HUANG, S., KO, B., WU, S., TSAY, W. & TIEN, H. 2013. Higher bone marrow LGALS3 expression is an independent unfavorable prognostic factor for survival in patients with acute myeloid leukemia. Blood, 121, 3172-3180. CHENG, L., ITAHANA, Y., DLEI, Z. & AL, E. 2012. TP53 genomic status regulates sensitivity of gastric cancer cells to the histone methylation inhibitor 3- deazaneplanocin A (DZNep). Clin Cancer Res, 18, 4201-4212. CHESI, M., NARDINI, E., LIM, R. S., SMITH, K. D., KUEHL, W. M. & BERGSAGEL, P. L. 1998. The t(4;14) Translocation in Myeloma Dysregulates Both FGFR3and a Novel Gene, MMSET, Resulting in IgH/MMSET Hybrid Transcripts. Blood, 92, 3025–3034. CHEUNG, A., WAN, T., LEUNG, J., CHAN, L., HUANG, H., KWONG, Y., LIANG, R. & LEUNG, A. 2007. Aldehyde dehydrogenase activity in leukemic blasts defines a subgroup of acute myeloid leukemia with adverse prognosis and superior NOD/SCID engrafting potential. Leukemia, 21, 1423-30. CHEUNG, P., ALLIS, C. D. & SASSONE-CORIS, P. 2000. Signaling to chromatin through histone modifications. Cell, 103, 263-271.

218

CHIANG, P., BURBELO, P., BRUGH, S., GORDON, R., FUKUDA, K. & YAMADA, Y. 1992. Activation of collagen IV gene expression in F9 teratocarcinoma cells by 3-deazaadenosine analogs. Indirect inhibitors of methylation. J Biol Chem, 267, 4988-91. CHIANG, P. & CANTONI, G. 1979. Perturbation of biochemical transmethylation by 3-deazaadenosine in vivo. biochem Pharmacol, 28, 1897-1902. CHIM, C., PANG, R., FUNG, T., CHOI, C. & LIANG, R. 2007. Epigenetic dysregulation of Wnt signaling pathway in multiple myeloma. Leukemia, 21, 2527-2536. CHINAMI, M., YANO, Y., YANG, X., SALAHUDDIN, S., MORIYAMA, K., SHIROISHI, M., TURNER, H., SHIRAKAWA, T. & ADRA, C. 2005. Binding of HTm4 to Cyclin-dependent Kinase (Cdk)-associated Phosphatase (KAP)·Cdk2·Cyclin A Complex Enhances the Phosphatase Activity of KAP, Dissociates Cyclin A, and Facilitates KAP Dephosphorylation of Cdk2. J Biol Chem, 280, 17235-42. CHOI, J. 2013. Eukaryotic Epigenetic Gene Regulation. In: AVISSAR, Y., CHOI, J., DESAIX, J., JURUKOVSKI, V., WISE, R. & RYE, C. (eds.) Biology. Wiley PLUS. CHOU, J. L., SU, H. Y., CHEN, L. Y., LIAO, Y. P., HARTMAN-FREY, C., LAI, Y. H., YANG, H. W., DEATHERAGE, D. E., KUO, C. T., HUANG, Y. W., YAN, P. S., HSIAO, S. H., TAI, C. K., LIN, H. J., DAVULURI, R. V., CHAO, T. K., NEPHEW, K. P., HUANG, T. H., LAI, H. C. & CHAN, M. W. 2010. Promoter hypermethylation of FBXO32, a novel TGF-beta/SMAD4 target gene and tumor suppressor, is associated with poor prognosis in human ovarian cancer. Lab Invest, 90, 414-25. CHOWDHURY, R., YEOH, K., TIAN, Y., HILLRINGHAUS, L., BAGG, E., ROSE, N. R., LEUNG, I., LI, X., WOON, E., YANG, M., MCDONOUGH, M. A., KING, O., CLIFTON, I., CLARIDGE, T., RATCLIFFE, P., SCHOFIELD, C. J. & KAWAMURA, A. 2010. The oncometabolite 2-hydroxyglutarate inhibits histone lysine demethylases. EMBO Rep, 12, 463-9. CHRISTENSEN, J. & WEISSMAN, I. 2001. Flk-2 is a marker in hematopoietic stem cell differentiation: a simple method to isolate long-term stem cells. Proc Natl Acad Sci USA, 98, 14541-14546.

219

CHRISTOFK, H., VANDER HEIDEN, M., HARRIS, M., RAMANATHAN, A., GERSZTEN, R., WEI, R., FLEMING, M., SCHREIBER, S. & CANTLEY, L. 2008. The M2 splice isoform of pyruvate kinase is important for cancer metabolism and tumour growth. Nature, 452, 230-3. CHRITOFK, H., VANDER HEIDEN, M., HARRIS, M., RAMANATHAN, A., GERSZTEN, R., WEI, R., FLEMING, M., SCHREIBER, S. & CANTLEY, L. 2008. The M2 splice isoform of pyruvate kinase is important for cancer metabolism and tumour growth. Nature, 452, 230-3. CLEVERS, H. & NUSSE, R. 2012. Wnt/beta-catenin signaling and disease. Cell, 149, 1192-1205. CLOOS, P., CHRISTENSEN, J., AGGER, K. & HELIN, K. 2008. Erasing the methyl mark: histone demethylases at the center of cellular differentiation and disease. Genes Dev, 22, 1115-40. CLOOS, P. A., CHRISTENSEN, J., AGGER, K., MAIOLICA, A., RAPPSILBER, J., ANTAL, T., HANSEN, K. H. & HELIN, K. 2006. The putative oncogene GASC1 demethylates tri- and dimethylated lysine 9 on histone H3. Nature, 442, 307-11. COBAS, M., WILSON, A., ERNST, B., MANCINI, S., MACDONALD, H., KEMLER, R. & RADTKE, F. 2004. b-catenin is dispensable for hematopoiesis and lymphopoiesis. JEM, 199, 221-229. COCKMAN, M., WEBB, J. & RATCLIFFE, P. 2009. FIH-dependent asparaginyl hydroxylation of ankyrin repeat domain-containing proteins. Ann N Y Acad Sci, 1177, 9-18. COLLA, S., STORTI, P., DONOFRIO, G., TODOERTI, K., BOLZONI, M., LAZZARETTI, M., ABELTINO, M., IPPOLITO, L., NERI, A., RIBATTI, D., RIZZOLI, V., MARTELLA, E. & GIULIANI, N. 2010. Low bone marrow oxygen tension and hypoxia-inducible factor-1α overexpression characterize patients with multiple myeloma: role on the transcriptional and proangiogenic profiles of CD138+ cells. Leukemia, 24, 1967-1970. COLLINS, P., JONES, C., CHOUDHURY, S., DAMELIN, L. & HODGSON, H. 2005. Increased expression of uncoupling protein 2 in HepG2 cells attenuates oxidative damage and apoptosis. Liver Int, 25, 880-7.

220

COMIN-ANDUIX, B., BOREN, J., MARTINEZ, S., MORO, C., CENTELLES, J., TREBUKHINA, R., PETUSHOK, N., LEE, W., BOROS , L. & CASCANTE, M. 2001. The effect of thiamine supplementation on tumour proliferation. A metabolic control analysis study. Eur J Biochem, 268, 4177-82. COPELAND, R. 2013. Molecular pathways: protein methyltransferases in cancer. Clin Cancer Res, 19, 6344-50. CORTI, S., LOCATELLI, F., PAPADIMITRIOU, D., DONADONI, C., SALANI, S., DEL BO, R., STRAZZER, S., BRESOLIN, N. & COMI, G. 2006. Identification of a primitive brain-derived neural stem cell population based on aldehyde dehydrogenase activity. Stem Cells, 24, 975-85. COUSTAN-SMITH, E., RIBEIRO, R., RUBNITZ, J., RAZZOUK, B., PUI, C., POUNDS, S., ANDREANSKY, M., BEHM, F., RAIMONDI, S., SHURTLEFF, S., DOWNING, J. & CAMPANA, D. 2003. Clinical significance of residual disease during treatment in childhood acute myeloid leukaemia. Br J Haematol, 123, 243-52. COUTURE, J. F., COLLAZO, E., ORTIZ-TELLO, P. A., BRUNZELLE, J. S. & TRIEVEL, R. C. 2007. Specificity and mechanism of JMJD2A, a trimethyllysine-specific histone demethylase. Nat Struct Mol Biol, 14, 689-695. COZZIO, A., PASSEGUE, E., AYTON, P., KARSUNKY, H., CLEARY, M. & WEISSMAN, I. 2003. Similar MLL-associated leukemias arising from self- renewing stem cells and shortlived myeloid progenitors. Genes Dev, 17, 3029- 3035. CREUTZIG, U., ZIMMERMANN, M., RITTER, J., REINHARDT, D., HERMANN, J., HENZE, G., JURGENS, H., KABISCH, H., REITER, A., RIEHM, H., GADNER, H. & SCHELLONG, G. 2005. Treatment strategies and long-term results in paediatric patients treated in four consecutive AML-BFM trials. 19, 12, 2030-2042. CUBELOS, B., SEBASTIAN-SERRANO, A., KIM, S., MORENO-ORTIZ, C., REDONDO, J., WALSH, C. & NIETO, M. 2008. Cux-2 controls the proliferation of neuronal intermediate precursors of the cortical subventricular zone. Cereb Cortex, 18, 1758-70. CURTIN, J. & LORENZI, M. 2010. Drug discovery approaches to target Wnt signalling in cancer stem cells. Oncotarget, 1, 552-566.

221

CZERMIN, B., MELFI, R., MCCABE, D., SEITZ, V., IMHOF, A. & PIRROTTA, V. 2002. Drosophila enhancer of Zeste/ESC complexes have a histone H3 methyltransferase activity that marks chromosomal Polycomb sites. Cell, 111, 185-96. DAIGLE, A., OLHAVA, E., THERKELSEN, C., MAJER, C., SNEERINGER, C., SONG, J., JOHNSTON, L., SCOTT, M., SMITH, J., XIAO, Y., JIN, L., KUNTZ, K., CHESWORTH, R., MOYER, M., BERNT, K., TSENG, J., KUNG, A., ARMSTRONG, S., COPELAND, R., RICHON, V. & POLLUCK, R. 2011. Selective killing of mixed lineage leukemia cells by a potent small- molecule DOT1L inhibitor. Cancer Cell, 20, 53-65. DAIGLE, S., OLHAVA, E., THERKELSEN, C., BASAVAPATHRUNI, A., JIN, L., BORJACK-SJODIN, P., ALLAIN, C., KLAUS, C., RAIMONDI, A., SCOTT, M., WATERS, N., CHESWORTH, R., MOYER, M., COPELAND, R., RICHON, V. & POLLOCK, R. 2013. Potent inhibition of DOT1L as treatment of MLL-fusion leukemia. Blood, 122, 1017-1025. DANG, C., KIM, J. W., GAO, P. & YUSTEIN, J. 2008. The interplay between MYC and HIF in cancer. Nat Rev Cancer, 8, 51-56. DANG, L., WHITE, D., GROSS, S., BENNETT, B., BITTINGER, M., DRIGGERS, E., FANTIN, V., JANG, H., JIN, S., KEENAN, M., MARKS, K., PRINS, R., WARD, P., YEN, K., LIAU, L., RABINOWITZ, J., CANTLEY, L., THOMPSON, C. B., VANDER HEIDEN, M. & SU, S. 2009. Cancer-associated IDH1 mutations produce 2-hydroxyglutarate. Nature, 462, 739-744. DAUD, S. 2014. The Role of WNT Transcription Factor TCF7L2 in Acute Myeloid Leukaemia. PhD, Cardiff University. DAUD, S., BURNETT, A., DARLEY, R. & TONKS, A. 2010. Large-Scale Inegration of Gene Profiling Identifies TCF7L2/TCF4 as the Most Frequently Dysregulated Wnt Signalling Component in AML. Blood, 116, 1029. DAVEY, F., ERBER, W., GATTER, K. & MASON, D. 1988. Abnormal neutrophils in acute myeloid leukemia and myelodysplastic syndrome. Hum Pathol, 19, 454-9. DAVID, C. J., CHEN, M., ASSANAH, M., CANOLL, P. & MANLEY, J. L. 2009. HnRNP proteins controlled by c‑ Myc deregulate pyruvate kinase mRNA splicing in cancer. Nature, 463, 364-8.

222

DAWSON, M., PRINJHA, R., DITTMANN, A., GIOTOPOULOS, G., BANTSCHEFF, M., CHAN, W., ROBSON, S., CHUNG, C., HOPF, C., SAVITSKI, M., HUTHMACHER, C., GUDGIN, E., LUGO, D., BEINKE, S., CHAPMAN, T., ROBERTS, E., SODEN, P., AUGER, K., MIRGUET, O., DOEHNER, K., DELWEL, R., BURNETT, A., JEFFREY, P., DREWES, G., LEE, K., HUNTLY, B. & KOUZARIDES, T. 2011. Inhibition of BET recruitment to chromatin as an effective treatment for MLL-fusion leukaemia. Nature, 478, 529-533. DE LENA, M., LORUSSO, V., LATORRE, A., FANIZZA, G., GARGANO, G., CAPORUSSO, L., GUIDA, M., CATINO, A., CRUCITTA, E., SAMBIASI, D. & MAZZEI, A. 2001. Paclitaxel, cisplatin and lonidamine in advanced ovarian cancer. A phase II study. Eur J Cancer, 37, 364-8. DE RUIJTER, A., VAN GENNIP, A., CARON, H., KEMP, S. & VAN KUILENBURG, A. 2003. Histone deacetylases (HDACs): characterization of the classical HDAC family. Biochem J, 370, 737-749. DE WETERING, M., SANCHO, E., VERWEIJ, C., DE LAU, W., OVING, I., HURLSTONE, A., DER HORN, K., BATLLE, E., COUDREUSE, D., HARAMIS, A., TJON-PON-FONG, M., MOERER, P., VAN DEN BORN, M., SOETE, G., PALS, S., EILER, S. M., MEDEMA, R. & CLEVERS, H. 2002. The β-Catenin/TCF-4 Complex Imposes a Crypt Progenitor Phenotype on Colorectal Cancer Cells. Cell, 111, 241-250. DEGUCHI, Y., KIRSCHENBAUM, A. & KEHRL, J. 1992. A diverged homeobox gene is involved in the proliferation and linage commitment of human hematopoietic progenitors and highly expressed in acute myelogenous leukemia. Blood, 79, 2841-2848. DEL RIZZO, P., KRISHNAN, S., TRIEVEL, R. & 4044–4052. 2012. Crystal structure and functional analysis of JMJD5 indicate an alternate specificity and function. 32. DEXTER, T., MOORE, M. & SHERIDAN, A. 1977. Maintenance of hemopoitic stem cells and production of differentiated progeny in allogeneic and semiallogeneic bone marrow chimeras in vitro. J Exp Med, 145, 1612-1616.

223

DI COSIMO, S., FERRETTI, G., PAPALDO, P., CARLINI, P., FABI, A. & COGENETTI, F. 2003. Lonidamine: efficacy and safety in clinical trials for the treatment of solid tumors. Drugs Today (Barc), 39, 157-74. DIETRICH, P., YANG, C., LEUNG, H., LYNCH, J., GONZALES, E., LIU, B., HABER, M., NORRIS, M., WANG, J. & WANG, J. 2014. GPR84 sustains aberrant beta-catenin signaling in leukemic stem cells for maintenance of MLL leukemogenesis. Blood, 124, 3284-3294. DING, L., LEY, T. J., LARSON, D., MILLER, C., KOBOLDT, D., WELCH, J., RITCHEY, J., YOUNG, M., LAMPRECHT, T., MCLELLAN, M., MCMICHAEL, J., WALLIS, J., LU, C., SHEN, D., HARRIS, C., DOOLING, D., FULTON, R., FULTON, L., CHEN, K., SCHMIDT, H., KALICKI- VEIZER, J., MAGRINI, V., COOK, L., MCGRATH, S., VICKERY, T., WENDL, M., HEATH, S., WATSON, M., LINK, D., TOMASSON, M., SHANNON, W., PAYTON, J., KULKARNI, S., WESTERVELT, P., WALTER, M., GRAUBERT, T., MARDIS, E., WILSON, A. & DIPERSIO, J. 2012. Clonal evolution in relapsed acute myeloid leukaemia revealed by whole- genome sequencing. Nature, 481, 506-10. DLUZNIEWSKA, A., BALWIERZ, W., ARMATA, J., BALCERSKA, A., CHYBICKA, A., KOWALCZYK, J., MATYSIAK, M., OCHOCKA, M., RADWANSKA, U., ROKICKA-MILEWSKA, R., SONTA-JAKIMCZYK, D., WACHOWIAK, J. & WYSOCKI, M. 2005. Twenty years of Polish experience with three consecutive protocols for treatment of childhood acute myelogenous leukemia. Leukemia, 19, 2117-2124. DOLBEARE, F., GRATZNER, H., PALLAVICINI, M. & GRAY, J. 1983. Flow cytometric measurement of total DNA content and incorporated bromodeoxyuridine. Proc Natl Acad Sci USA, 80, 5573-5577. DONATO, J., KO, J., KUTOK, J., CHENG, T., SHIRAKAWA, T., MAO, X., BEACH, D., SCADDEN, D., SAYEGH, M. & ADRA, C. 2002. Human HTm4 is a hematopoietic cell cycle regulator. J Clin Invest, 109, 51-58. DOROSHOW, J. 2010. Topoisomerase II inhibitors: Anthracyclines. In: CHABNER, B. & LONGO, D. (eds.) Cancer Chemotherapy and Biotherapy: Principles and Practice. 5th ed. Philadelphia: Lippincott Williams & Wilkins.

224

DOWNING, J. & SHANNON, K. 2002. Acute leukemia: A pediatric perspective. Cancer Cell, 2, 437-445. DROGE, W. 2002. Free radicals in the physiological control of cell function. Physiol Rev, 71, 47-95. DROLLE, H., WAGNER, M., VASOLD, J., KUTT, A., DENIFFEL, C., SOTLAR, K., SIRONI, S., HEROLD, T., RIEGER, C. & FIEGL, M. 2015. Hypoxia regulates proliferation of acute myeloid leukemia and sensitivity against chemotherapy. Leuk Res, 39, 779-785. DRUMMOND, D., NOBLE, C., KIRPOTIN, D., GUO, Z., SCOTT, G. & BENZ, C. 2005. Clinical development of histone deacetylase inhibitors as anticancer agents. Annu Rev Pharmacol Toxicol, 45, 495-528. DUERRE, J. & LEE, C. 1974. In vivo methylation and turnover of rat brain histones. J Neurochem, 23, 541-7. DUPONT, S., LABEGUERE, F., BLANQUE, R., DE VOS, S., CELEMENT- LACROIX, P., PARENT, I., SACCOMANI, C., NELLES, L., HAGERS, A., COTTEREAUX, C., MERCIRIS, D., CECCOTTI, M., DA COSTA, C., FLETCHER, S. & BRYS, R. 2014. GPR84, a novel target for the development of therapies for IBD. United European Gastroenterology, Abstract #OP183. EASWARAN, H., TSAI, H. & BAYLIN, S. 2014. Cancer epigenetics: tumor heterogeneity, plasticity of stem-like states and drug resistance. Mol Cell, 54, 716-27. EHNINGER, A. & TRUMPP, A. 2011. The bone marrow stem cell niche grows up: mesenchymal stem cells and macrophages move in. J Exp Med, 208, 421-8. ELLIS, L., ATADJA, P. & JOHNSTONE, R. 2009. Epigenetics in cancer: Targeting chromatin modifications. Mol Cancer Ther, 8, 1409-1420. EMA, H., TAKANO, H., SUDO, K. & NAKAUCHI, H. 2000. In vitro self-renewal division of hematopoietic stem cells. J Exp Med, 192, 1281-1288. ENTZ-WERLE, N., SUCIU, S., VAN DER WERFF TEN BOSCH, J., VILMER, E., BERTRAND, Y., BENOIT, Y., MARGUERITTE, G., PLOUVIER, E., BOUTARD, P., VANDECRUYS, E., FERSTER, A., LUTZ, P., UYTTEBROECK, A., HOYOUX, C., THYSS, A., RIALLAND, X., NORTON, L., PAGES, M., PHILIPPE, N., OTTEN, J. & BEHAR, C. 2005. Results of 58872 and 58921 trials in acute myeloblastic leukemia and relative value of

225

chemotherapy vs allogeneic bone marrow transplantation in first complete remission: the EORTC Children Leukemia Group report. Leukemia, 19, 2072- 2081. EPSTEIN, A., GLEADLE, J., MCNEILL, L., HEWITSON, K., O'ROURKE, J., MOLE, D., MUKHERJI, M., METZEN, E., WILSON, M., DHANDA, A., TIAN, Y., MASSON, N., HAMILTON, D., JAAKKOLA, P., BARSTEAD, R., HODGKIN, J., MAXWELL, O., PIUGH, C., SCHOFIELD, C. J. & RATCLIFFE, P. 2001. C. elegans EGL-9 and mammalian homologs define a fmily of dioxygenases that regulate HIF by prolyl hydroxylation. Cell, 107, 43- 54. ERAMO, A., LOTTI, F., SETTE, G., PILOZZI, E., BIFFONI, M., DI VIRGILIO, A., CONTICELLO, C., RUCO, L., PESCHLE, C. & DE MARIA, R. 2008. Identification and expansion of the tumorigenic lung cancer stem cell population. Cell Death Differ, 15, 504-14. ERNST, T., CHASE, A. J., SCORE, J., HIDALGO-CURTIS, C. E., BRYANT, C., JONES, A. V., WAGHORN, K., ZOI, K., ROSS, F. M., REITER, A., HOCHHAUS, A., DREXLER, H., DUNCOMBE, A., CERVANTES, F., OSCIER, D., BOULTWOOD, J., GRAND, F. & CROSS, N. C. 2010. Inactivating mutations of the histone methyltransferase gene EZH2 in myeloid disorders. Nat Genet, 42, 722-726. ERNST, T., PFLUG, A., RINKE, J., ERNST, J., BIERBACH, U. & BECK, J. 2012. A somatic EZH2 mutation in chidlhood acute myeloid leukemia. Leukemia, 26, 1701-3. ESTELLER, M. 2008. Epigenetics in Cancer. N Eng J Med, 358, 1148-1159. FABER, J., KRIVTSOV, A. V., STUBBS, M. C., WRIGHT, R., DAVIS, T. N., VAN DEN HEUVEL-EIBRINK, M., ZWAAN, C. M., KUNG, A. L. & ARMSTRONG, S. A. 2009. HOXA9 is required for survival in human MLL- rearranged acute leukemias. Blood, 113, 2375-85. FANG, R., BARBERA, A., XU, Y., RUTENBERG, M., LEONOR, T., BI, Q., LAN, F., MEI, P., YUAN, G., LIAN, C., PENG, J., CHENG, D., SUI, G., KAISER, U., SHI, Y. & SHI, Y. 2010. Human LSD2/KDM1b/AOF1 regulates gene transcription by modulating intragenic H3K4me2 methylation. Mol Cell, 39, 222-33.

226

FEI, F., JOO, E., TARIGHAT, S., SCHIFFER, I., PAZ, H., FABBRI, M., ABDEL- AZIM, H., GROFFEN, J. & HEISTERKAMP, N. 2015. B-cell precursor acute lymphoblastic leukemia and stromal cells communicate through Galectin-3. Oncotarget, 6, 11378-94. FENG, J., GUO, X., CHEN, Y., WANG, Z., CHENG, Y. & TANG, Y. 2012. Prognostic significance of IDH1 mutations in acute myeloid leukemia: a meta- analysis. Am J Blood Res, 2, 254-64. FENG, Q., WANG, H., NG, H., ERDJUMENT-BROMAGE, H., TEMPST, P., STRUHL, K. & ZHANG, Y. 2002. Methylation of H3‑ lysine 79 is mediated by a new family of HMTases without a SET domain. Curr Biol, 12, 1052-8. FERRARA, F. & MIRTO, S. 2003. Serum LDH value as a predictor of clinical outcome in acute myelogenous leukaemia of the elderly. Br J Haematol, 92, 627-31. FIALKOW, P., DENMAN, A., JACOBSON, R. & LOWENTHAL, M. 1978. Chronic myelocytic leukemia. Origin of some lymphocytes from leukemic stem cells. J Clin Invest, 62, 815-23. FIALKOW, P., SINGER, J., RASKIND, W., ADAMSON, J., JACOBSON, R., BERNSTEIN, I., DOW, L., NAJFELD, V. & VEITH, R. 1987. Clonal development, stem-cell differentiation, and clinical remission in acute nonlymphocytic leukemia. N Engl J Med, 317, 468-473. FIEGL, M., SAMUDIO, I., CLISE-DWYER, K., BURKS, J., MNJOYAN, Z. & ANDREEFF, M. 2009. CXCR4 expression and biologic activity in acute myeloid leukemia are dependent on oxygen partial pressure. Blood, 113, 1504- 12. FIGUEROA, M., ABDEL-WAHAB, O., LU, C., WARD, P., PATEL, J., SHIH, A., LI, Y., BHAGWAT, N., VASANTHAKUMAR, A., FERNANDEZ, H., TALLMAN, M., SUN, Z., WOLNIAK, K., PEETERS, J., LIU, W., CHOE, S., FANTIN, V., PAIETTA, E., LOWENBERG, B., LICHT, J., GODLEY, L., DELWEL, R., VALK, P., THOMPSON, C. B., LEVINE, R. L. & MELNICK, A. 2010a. Leukemic IDH1 and IDH2 mutations result in a hypermethylation phenotype, disrupt TET2 function, and impair hematopoietic differentiation. 18, 6. FIGUEROA, M., LUGTHART, S., LI, Y., ERPELINCK-VERSCHUEREN, C., DENG, X., CHRISTOS, P., SCHIFANO, E., BOOTH, J., VAN PUTTEN, W.,

227

SKRABANEK, L., CAMPAGNE, F., MAZUMDAR, M., GREALLY, J., VALK, P., LOWENBERG, B., DELWEL, R. & MELNICK, A. 2010b. DNA methylation signatures identify biologically distinct subtypes in acute myeloid leukemia. Cancer Cell, 17, 13-27. FISKUS, W., PRANPAT, M., BALI, P., BALASIS, M., KUMARASWAMY, S., BOYAPALLE, S., ROCHA, K., WU, J., GILES, F., MANLEY, P., ATADJA, P. & BHALLA, K. 2006. Combined effects of novel tyrosine kinase inhibitor AMN107 and histone deacetylase inhibitor LBH589 against Bcr-Abl-expressing human leukemia cells. Blood, 108, 645-52. FISKUS, W., RAO, R., BALUSU, R., GANGULY, S., TAO, J., SOTOMAYOR, E., MUDUNURU, U., SMITH, J., HEMBRUFF, S., ATADJA, P., MARQUEZ, V. & BHALLA, K. 2012. Superior efficacy of a combined epigenetic therapy against human mantle cell lymphoma cells. Clin Cancer Res, 18, 6227-38. FISKUS, W., WANG, Y. C., SREEKUMAR, A., BUCKLEY, K. M., SHI, H. D., JILLELLA, A., USTUN, C., RAO, R., FERNANDEZ, P., CHEN, J. G., BALUSU, R., KOUL, S., ATADJA, P., MARQUEZ, V. E. & BHALLA, K. N. 2009. Combined epigenetic therapy with the histone methyltransferase EZH2 inhibitor 3-deazaneplanocin A and the histone deacetylase inhibitor panobinostat against human AML cells. Blood, 114, 2733-2743. FLEMING, H., JANZEN, V., LO, C. C., GUO, J., LEAHY, K., KRONENBERG, H. & SCADDEN, D. 2008. Wnt signaling in the niche enforces hematopoietic stem cell quiescence and is necessary to preserve self-renewal in vivo. Cell Stem Cell, 2, 274-83. FLORIDI, A., PAGGI, M., MARCANTE, M., SILVESTRINI, B., CAPUTO, A. & DE MARTINO, C. 1981. Lonidamine, a selective inhibitor of aerobic glycolysis of murine tumor cells. J Natl Cancer Inst, 66, 497-499. FOREMAN, K., RIZZO, P., OSIPO, C. & MIELE, L. 2009. The Cancer Stem Cell Hypothesis. In: BAGLEY, R. & TEICHER, B. (eds.) Stem Cells and Cancer. New York: Humana Press. FRANK, S., PARISI, T., TAUBERT, S., FERNANDEZ, P., FUCHS, M., CHAN, H., LIVINGSTON, D. & AMATI, B. 2003. MYC recruits the TIP60 histone acetyltransferase complex to chromatin. EMBO Rep, 4, 575-80.

228

FRANKEN, N., RODERMOND, H., STAP, J., HAVEMAN, J. & VAN BREE, C. 2006. Clonogenic assays of cells in vitro. Nat Protoc, 1, 2315-9. FRIMBERGER, A., STERING, A. & QUESENBERRY, P. 2001. Characterization of engraftable hematopoietic stem cells in murine long-term bone marrow cultures. Exp Hematol, 29, 643-652. FRISCH, B., ASHTON, J., XING, L., BECKER, M., JORDAN, C. & CALVI, L. 2012. Functional inhibition of osteoblastic cells in an in vivo mouse model of myeloid leukemia. Blood, 119, 540-550. FUJIWARA, S., KAWANO, Y., YUKI, H., OKUNO, Y., NOSAKA, K., MITSUYA, H. & HATA, H. 2013. PDK1 inhibition is a novel therapeutic target in multiple myeloma. Br J Cancer, 108, 170-8. FUJIWARA, T., SAITOH, H., INOUE, A., KOBAYASHI, M., OKITSU, Y., KATSUOKA, Y., FUKUHARA, N., ONISHI, Y., ISHIZAWA, K., ICHINOHASAMA, R. & HARIGAE, H. 2014. 3-Deazaneplanocin A (DZNep), an inhibitor of adenosylmethionine-dependent methyltrnasferase promotes erythroid differentiation. J Biol Chem, 289, 8121-34. FUKS, F., HURD, P., R, D. & KOUZARIDES, T. 2003. The DNA methyltransferase associate with HP1 and SUV39H1 histone methyltransferases. Nucleic Acids Res, 31, 2305-12. GALE, R. 1979. Advances in treatment of acute myelogenous leukemia. N Eng J Med, 300, 1189-99. GAMBHIR, S. 2002. Molecular imaging of cancer with positron emission tomography. Nat Rev Cancer, 2, 683-693. GAO, J., GRAVES, S., KOCH, U., LIU, S., JANKOVIC, V., BUONAMICI, S., EL ANDALOUSSI, A., NIMER, S., KEE, B., TAICHMAN, R., RADTKE, F. & AIFANTIS, I. 2009. Hedgehog signaling is dispensable for adult hematopoietic stem cell function. Cell Stem Cell, 4, 548-558. GAO, X., WANG, H., YANG, J., LIU, X. & LIU, Z. 2012. Pyruvate kinase M2 regulates gene transcription by acting as a protein kinase. Mol Cell, 945, 598- 609. GASSMANN, M., GRENACHER, B., ROHDE, B. & VOGEL, J. 2009. Quantifying Western blots: pitfalls of densitometry. Electrophoresis, 30.

229

GATENBY, R. & GILLIES, R. 2004. Why do cancers have high aerobic glycolysis? Nat Rev Cancer, 4, 891-9. GESCHWIND, J., GEORGIADES, C., KO, Y. & PEDERSON, P. 2004. Recently elucidated energy catabolism pathways provide opportunities for novel treatments in hepatocellular carcinoma. Expert Rev Anticancer Ther, 4, 449-57. GIAMBRA, V., JENKINS, C., LAM, S., HOOFD, C., BELMONTE, M., WANG, X., GUSSCOTT, S., GRACIAS, D. & WENG, A. 2015. Leukemia stem cells in T- ALL require active Hif1a and Wnt signaling. Blood, 125, 3917-3927. GIBSON, B., PERENTESIS, J., ALONZO, T. & KASPERS, G. 2011a. Treatment of acute myeloid leukaemia. In: REAMAN, G. & SMITH, F. (eds.) Childhood leukemia. Berlin: Springer. GIBSON, B., WEBB, D., HOWMAN, A., DE GRAAF, S., HARRISON, C., WHEATLEY, K., UNITED KINGDOM CHILDHOOD LEUKAEMIA WORKING GROUP & DUTCH CHILDHOOD ONCOLOGY GROUP 2011b. Results of a randomized trial in children with acute myeloid leukaemia: Medical research council AML12 trial. Br J Haematol, 155, 366-76. GIBSON, B., WHEATLEY, K., HANN, I., STEVENS, R., WEBB, D., HILLS, R., GRAAF, S. & HARRISON, C. 2005. Treatment strategy and long-term results in paediatric patients treated in consecutive UK AML trials. Leukemia 19, 2130- 2138. GLASER, K., LI, J., STAVER, M., WEI, R., ALBERT, D. & DAVIDSEN, S. 2003. Role of class I and class II histone deacetylases in carcinoma cells using siRNA. Biochem Biophys Res Commun, 2003. GLAZER, R., KNODE, M., TSENG, C., HAINES, D. & MARQUEZ, V. 1986. 3- Deazaneplanocin A: a new inhibitor of S-adenosylhomocysteine synthesis and its effects in human colon carcinoma cells. Biochem Pharmacol, 35, 4523-7. GLOVER, A., LEYLAND-JONES, B., CHUN, H., DAVIES, B. & HOTH, D. 1987. Azacitidine: 10 years later. Cancer Treat Rep, 71, 737-46. GLOZAK, M., SENGUPTA, N., ZHANG, X. & SETO, E. 2005. Acetylation and deacetylation of non-histone proteins. Gene, 363, 15-23. GOLDSTEIN, A., HUANG, J., GUO, C., GARRAWAY, I. & WITTE, O. 2010. Identification of a cell of origin for human prostate cancer. Science, 329, 568-71.

230

GOLUB, T., SLONIM, D., TAMAYO, P., HUARD, C., GAASENBEEK, M., MESIROV, J., COLLER, H., LOH, M., DOWNING, J., CALIGIURI, M., BLOOMFIELD, C. & LANDER, E. 1999. Molecular Classification of Cancer: Class Discovery and Class Prediction by Gene Expression Monitoring. Science, 286, 531-537. GONG, F. & MILLER, K. 2013. Mammalian DNA repair: HATs and HDACs make their mark through histone acetylation. Mutat Res, 750, 23-30. GOODELL, M., BROSE, K., PARADIS, G., CONNER, A. & MULLIGAN, R. 1996. Isolation and functional properties of murine hematopoietic stem cells that are replicating in vivo. J Exp Med, 183, 1797-1806. GORE, S. D., BAYLIN, S., SUGAR, E., CARRAWAY, H., MILLER, C. B., CARDUCCI, M., GREVER, M., GALM, O., DAUSES, T., KARP, J. E., RUDEK, M. A., ZHAO, M., SMITH, B. D., MANNING, J., JIEMJIT, A., DOVER, G., MAYS, A., ZWIEBEL, J., MURGO, A., WENG, L. J. & HERMAN, J. G. 2006. Combined DNA methyltransferase and histone deacetylase inhibition in the treatment of myeloid neoplasms. Cancer Research, 66, 6361-6369. GORMAN, M., JI, L., KO, R., BARNETTE, P., BOSTROM , B., HUTCHINSON, R. & AL, E. 2010. Outcome for children treated for relapsed or refractory acute myelogenous leukaemia (rAML): A therapeutic advances in Childhood leukaemia (TACL) Consortium study. Pediatr Blood Cancer, 55, 421-9. GOTTSCHALK, S., ANDERSON, N., HAINZ, C., ECKHARDT, S. & SERKOVA, N. 2004. Imatinib (STI571)-mediated changes in glucose metabolism in human leukemia BCR-ABL-positive cells. Clin Cancer Res, 10, 6661-8. GOWHER, H., LOUTCHANWOOT, P., VOROBJECA, O., HANDA, V., JURKOWSKA, R., JURKOWSKI, T. & JELTSCH, A. 2006. Mutational analysis of the catalytic domain of the murine Dnmt3a DNA-(cytosine C5)- methyltransferase. J Mol Biol, 357, 928-41. GREER, E. & SHI, Y. 2012. Histone methylation: a dynamic mark in health, disease and inheritance. Nat Rev Genet, 13, 343-57. GRUNSTEIN, M. 1997. Histone acetylation in chromatin structure and transcription. Nature, 389, 349-52.

231

GUENTHER, M., LAWTON, L., ROZOVSKAIA, T., FRAMPTON, G. M., LEVINE, S. S., VOLKERT, T. L., CROCE, C. M., NAKAMURA, T., CANAANI, E. & YOUNG, R. A. 2008. Aberrant chromatin at genes encoding stem cell regulators in human mixed-lineage leukemia. Genes Dev, 22, 3403-8. GUENTHER, M., LEVINE, S., BOYER, L., JAENISCH, R. & YOUNG, R. 2007. A chromatin landmark and transcription initiation at most promoters in human cells. Cell, 130, 77-88. GUIL, S. & ESTELLER, M. 2009. DNA methylomes, histone codes and miRNAs: Tying it all together. Int J Biochem Cell Biol, 41, 87-95. HAFERLACH, T., KOHLMANN, A., KLEIN, H., RUCKERT, C., DUGAS, M., WILLIAMS, P., KERN, W., SCHNITTGER, S., BACHER, U., LOFFLER, H. & HAFERLACH, C. 2009. AML with translocation t(8;16)(p11;p13) demonstrates unique cytomorphological, cytogenetic, molecular and prognostic features. Leukemia, 23, 934-943. HAN, W., LI, L., QIU, S., LU, Q., PAN, Q., GU, Y., LUO, J. & HU, X. 2007. Shikonin circumvents cancer drug resistance by induction of a necroptotic death. Mol Cancer Res, 6, 1641-9. HANAHAN, D. & WEINBERG, R. 2000. Hallmarks of Cancer: The Next Generaton. Cell, 144, 646-674. HANSEN, K., TIMP, W., BRAVO, H., SABUNCIYAN, S., LANGMEAD, B., MCDONALD, O., WEN, B., WU, H., LIU, Y., DIEP, D., BRIEM, E., ZHANG, K., IRIZARRY, R. & FEINBERG, A. 2011. Increased methylation variation in epigenetic domains across cancer types. Nat Genet, 43, 768-75. HARRIS, W., HUANG, X., LYNCH, J., SPENCER, G. J., HITCHIN, J., LI, Y., CICERI, F., BLASER, J., GREYSTOKE, B., JORDAN, A., MILLER, C., OGILVIE, D. & SOMERVAILLE, T. 2012. The histone demethylase KDM1A sustains the oncogenic potential of MLL-AF9 leukemia stem cells. Cancer Cell, 21, 473-487. HARTIG, S. 2013. Basic image analysis and manipulation in ImageJ. Curr Protoc Mol Biol, 14. HATFIELD, K., BEDRINGSAAS, S., RYNINGEN, A., GJERTSEN, B. & BRUSERUD, O. 2010. Hypoxia increases HIF-1α expression and constitutive cytokine. Eur Cytokine Netw, 21, 154-64.

232

HATHURUSINGHE, H., GOONETILLEKE, K. & SIRIWARDENA, K. 2007. Current status of tumor M2 pyruvate kinase (tumor M2-PK) as a biomarker of gastrointestinal malignancy. Ann Surg Oncol, 14, 2714-20. HECKL, D., KOWALCZWK, M., YUDOVICH, D., BELIZAIRE, R., PURAM, R., MCCONKEY, M., THIELKE, A., ASTER, J., REGEV, A. & EBERT, B. 2014. Generation of mouse models of myeloid malignancy with combinatorial genetic lesions sing CRISPR-Cas9 genome editing. Nat Biotechnol, 32, 941-6. HEDDLESTON, J., WU, Q., RIVERA, M., MINHAS, S., LATHIA, J., SLOAN, A., ILIOPOULOS, O., HJELMELAND, A. & RICH, J. 2012. Hypoxia-induced mixed-lineage leukemia 1 regulates glioma stem cell tumorigenic potential. Cell Death Differ, 19, 428-439. HESS, D., MEYERROSE, T., WIRTHLIN, L., CRAFT, T., HERRBRICH, P., CREER, M. & NOLTA, J. 2004. Functional characterization of highly purified human hematopoietic repopulating cells isolated according to aldehyde dehydrogenase activity. Blood, 104, 1648-55. HIBINO, S., SAITO, Y., MURAMATSU, T., OTANI, A., KASAI, Y., KIMURA, M. & SAITO, H. 2014. Inhibitors of enhancer of zeste homolog 2 (EZH2) activate tunmor-suppressor microRNAs in human cancer cells. Oncogenesis, 3, 1-10. HILL, R. 2006. Identifying cancer stem cells in solid tumors: case not proven. Cancer Res, 66, 1891-5. HITOSUGI, T., KANG, S., VANDER HEIDEN, M., CHUNG, T., ELF, S., LYTHGOE, K., DONG, S., LONIAL, S., WANG, X., CHEN, G., XIE, J., GU, T., POLAKIEWICZ, R., ROESEL, J., BOGGON, T., KHURI, F., GILLILAND, D., CANTLEY, L., KAUFMAN, J. & CHEN, J. 2009. Tyrosine phosphorylation inhibits PKM2 to promote the Warburg effect and tumor growth. Sci Signal, 2, ra73. HO, M., NG, A., LAM, S. & HUNG, J. 2007. Side population in human lung cancer cell lines and tumors is enriched with stem-like cancer cells. Cancer Res, 67, 4827-33. HOFBAUER, K., GESS, B., LOHAUS, C., MEYER, H., KATSCHINSKI, D. & KIRTZ, A. 2003. Oxygen tension regulates the expression of a group of procollagen hydroxylases. Eur J Biochem, 270, 4515-22.

233

HOFFMAN, J., ROATSCH, M., SCHMITT, M., CARLINO, L., PIPPEL, M., SIPPL, W. & JUNGI, M. 2012. The role of histone demthylases in cancer therapy. Mol Oncol, 6, 683-703. HOFMANN, I., STOVER, E., CULLEN, D., MAO, J., MORGAN, K., LEE, B., KHARAS, M., MILLER, P., CORNEJO, M., OKABE, R., ARMSTRONG, S., GHILARDI, N., GOULD, S., DE SAUVAGE, F., MCMAHON, A. & GILLILAND, D. 2009. Hedgehog signaling is dispensable for adult murine hematopoietic stem cell function and hematopoiesis. Cell Stem Cell, 4, 559-67. HOLLIDAY, R. 1987. The inheritance of epigenetic defects. Science, 238, 163-70. HOLM, M., THOMSEN, M., HOYER, M. & HOKLAND, P. 1998. Optimization of a flow cytometric method for the simultaneous measurement of cell surface antigen, DNA content, and in vitro BrdUrd incorporation into normal and malignant hematopoietic cells. Cytometry, 32, 28-36. HOLMQUIST-MENGELBIER, L., FREDLUND, E., LOFSTEDT, T., NOGUERA, R., NAVARRO, S., NILSSON, H., PIETRAS, A., VALLON-CHRISTERSSON, J., BORG, A., GRADIN, K., POELLINGER, L. & PAHLMAN, S. 2006. Recruitment of HIF-1α and HIF-2α to common target genes is differentially regulated in neuroblastoma: HIF-2α promotes an aggressive phenotype. Cancer Cell, 10, 413-423. HOPE, K., JIN, L. & DICK, J. 2003. Human acute myeloid leukemia stem cells. Arch Med Res, 34, 507-514. HOPE, K., JIN, L. & DICK, J. 2004. Acute myeloid leukemia originates from a hierarchy of leukemic stem cell classes that differ in self-renewal capacity. Nat Immunol, 5, 738-743. HORAN, J., HASLE, H. & MASHINCHI, S. 2014. Acute Myeloid Leukemia. In: SMITH, F., REAMAN, G. & RACADIO, J. (eds.) Hematopoietic Cell Transplantation in Children with Cancer. New York: Springer-Verlag Berlin Heidelberg. HORTON, J., UPADHYAY, A. K., QI, H. H., ZHANG, X., SHI, Y. & CHENG, X. 2010. Enzymatic and structural insights for substrate specificity of a family of Jumonji histone lysine demethylases. Nat Struct Mol Biol, 17, 38-43.

234

HORTON, J. R., UPADHYAY, A. K., HASHIMOTO, H., ZHANG, X. & CHENG, X. 2011. Structural basis for human PHF2 Jumonji domain interaction with metal ions. J Mol Biol, 406, 1-8. HORTON, S. & HUNTLY, B. 2012. Recent Advances in Acute Myeloid Leukemia Stem Cell Biology. Haematologica, 97, 966-974. HOU, H. & YU, H. 2010. Structural insights into histone lysine demethylation. Curr Opin Struct Biol, 20, 739-48. HSIA, D., TEPPER, C., POCHAMPALLI, M., HSIA, E., IZUMIYA, C., HUERTA, S., WRIGHT, M., CHEN, H., KUNG, H. & IZUMIYA, Y. 2010. KDM8, a H3K36me2 histone demethylase that acts in the cyclin A1 coding region to regulate cancer cell proliferation. Proc Natl Acad Sci USA, 107, 9671-9676. HU, T., LI, C., ZHANG, Y., WANG, L., PENG, L., CHENG, H., WANG, W., CHU, Y., XU, M., CHENG, T. & YUAN, W. 2015. Phosphoinositide-dependent kinase 1 regulates leukemia stem cell maintenance in MLL-AF9-induced murine acute myeloid leukemia. Biochem Biophys Res Commun, 459, 692-8. HU, Y. & SMYTH, G. K. 2009. ELDA: Extreme limiting dilution analysis for comparing depleted and enriched populations in stme cell and other assays. J Immunol Methods, 347, 70-78. HU, Z., GOMES, I., HORRIGAN, S., KRAVARUSIC, J., MAR, B., ARBIEVA, Z., CHYNA, B., FULTON, N., EDASSERY, S., RAZA, A. & WESTBROOK, C. 2001. A novel nuclear protein, 5qNCA (LOC51780) is a candidate for the myeloid leukemia tumor suppressor gene on chromosome 5 band q31. Oncogene, 20, 6946-54. HUANG, H., JIANG, X., LI, Z., SONG, C., HE, C., SUN, M., CHEN, P., GURBUXANI, S., WANG, J., HONG, G., ELKAHLOUN, A., ARNOVITZ, S., WANG, J., SZULWACH, K., LIN, L., STREET, C., WUNDERLICH, M., DAWLATY, M., NEILLY, M., JAENISCH, R., YANG, F., MULLOY, J., JIN, P., LIU, P., ROWLEY, J., XU, M., HE, C. & CHEN, J. 2013. TET1 plays an essential oncogenic role in MLL-rearranged leukemia. Proc Natl Acad Sci USA, 110, 11994-11999. HUANG, J., SENGUPTA, R., ESPEJO, A., LEE, M., DORSEY, J., RICHTER, M., OPRAVIL, S., SHIEKHATTAR, R., BEDFORD, M., JENUWEIN, T. &

235

BERGER, S. 2007a. p53 is regulated by the lysine demethylase LSD1. Nature, 449, 105-8. HUANG, Y., GREENE, E., MURRAY STEWART, T., GOODWIN, A., BAYLIN, S., WOSTER, P. & CASERO, R. J. 2007b. Inhibition of lysine-specific demethylase 1 by polyamine analogues results in reexpression of aberrantly silenced genes. Proc Natl Acad Sci USA, 104, 8023-8028. HUBEEK, I., STAM, R., PETERS, G., BROEKHUIZEN, R., MEIJERINK, J., VAN WERING, E., GIBSON, B., CREUTZIG, U., ZWAAN, C., CLOOS, J., KUIK, D., PIETERS, R. & KASPERS, G. 2005. The human equilibrative nucleoside transporter 1 mediates in vitro cytarabine sensitivity in childhood acute myeloid leukaemia. Br J Cancer, 93, 1388-94. HUNTLY, B. & GILLILAND, D. 2005. Leukaemia stem cells and the evolution of cancer-stem-cell research. Nat Rev Cancer, 5, 311-321. HUNTLY, B., SHIGEMATSU, H., DEGUCHI, K., LEE, B., MIZUNO, S., DUCLOS, N., ROWAN, R., AMARAL, S., CURLEY, D., WILLIAMS, I., AKASHI, K. & GILLILAND, G. 2004. MOZTIF2, but not BCR-ABL, confers properties of leukemic stem cells to committed murine hematopoietic progenitors. Cancer Cell, 6, 587-596. HURET, J., DESSEN, P. & BERNHEIM, A. 2001. An atlas of chromosomes in hematological malignancies. Example: 11q23 and MLL partners. Leukemia, 15, 987-989. HYNES, J., MARROQUIN, L., OGURTSOV, V., CHRISTIANSEN, K., STEVENS, G., PAPKOVSKY, D. & WILL, Y. 2006. Investigation of drug-induced mitochondrial toxicity using fluorescence-based oxygen-sensitive probes. Toxicol Sci, 92, 186-200. HYNES, J., NADANACIVA, S., SWISS, R., CAREY, C., KIRWAN, S. & WILL, Y. 2012. A high-throughput dual parameter assay for assessing drug-induced mitochondrial dysfunction provides additional predictivity over two established mitochondrial toxicity assays. Toxicol In Vitro, 27, 560-569. HYNES, J., NATOLI , E. & WILL, Y. 2009. Fluorescent pH and oxygen probes of the assessment of mitochondrial toxicity in isolated mitochondria and whole cells. Curr Protoc Toxicol, Chapter 2, 16.

236

IDA, K., KITABAYASHI, I., TAKI, T., TANIWAKI, M., NORO, Y., YAMAMOTO, M., OHKI, M. & HAYASHI, Y. 1997. Adenoviral E1A-associated protein p300 is involved in acute myeloid leukemia with t(11;22)(q23;q13). Blood, 90, 4699- 704. IIDA, H., TOWATARI, M., TANIMOTO, M., MORISHITA, Y., KODERA, Y. & SAITO, H. 1997. Overexpression of cyclin E in acute myelogenous leukemia. Blood, 90, 3707-13. ILLA-BOCHACA, I., FERNANDEZ-GONZALEZ, R., SHELTON, D., WELM, B., ORTIZ-DE-SLORZANO, C. & BARCELLOS-HOFF, M. 2010. Limiting- dilution transplantation assays in mammary stem cell studies. Methods Mol Biol, 621, 29-47. IM, A., SEHGAL, A., CARROLL, M., SMITH, B., TEFFERI, A., JOHNSON, D. & BOYIADZIS, M. 2014. DNMT3A and IDH mutations ina cute myeloid leukemia and other myeloid malignancies: associations with prognosis and potential treatment strategies. Leukemia, 28, 1774-83. INOUE, S., MAI, M., DYER, M. & COHEN, G. 2006. Inhibition of histone deacetylase class I but not class II is critical for the sensitization of leukemic cells to tumor necrosis factor-related apoptosis-inducing ligand-induced apoptosis. Cancer Res, 66, 6785-92. IRIZARRY, R., LADD-ACOSTA, C., WEN, B., WU, Z., MONTANO, C., ONYANGO, P., CUI, H., GABO, K., RONGIONE, M., WEBSTER, M., JI, H., POTASH, J., SABUNCIYAN, S. & FEINBERG, A. 2009. The human colon cancer methylome shows similar hypo- and hypermethylation at conserved tissue-specific CpG island shores. Nat Genet, 41, 178-86. IRVINE, D. & COPLAND, M. 2012. Targeting hedgehog in hematologic malignancy. Blood, 119, 2196-2204. ISHIKAWA, F., YOSHIDA, S., SAITO, Y., HIJIKATA, A., KITAMURA, H., TANAKA, S., NAKAMURA, R., TANAKA, T., TOMIYAMI, H., SAITO, N., FUKATA, M., MIYAMOTO, T., LYONS, B., OHSHIMA, K., UCHIDA, N., TANIGUCHI, S., OHARA, O., AKASHI, K., HARADA, M. & SHULTZ, L. 2007. Chemotherapy-resistant human AML stem cells home to and engraft within the bone-marrow endosteal region. Nat Biotechnol, 25, 1315-1321.

237

ISHIMURA, A., MINEHATA, K., TERASHIMA, M., KONDOH, G., HARA, T. & SUZUKI, T. 2012. Jmjd5, an H3K36me2 histone demethylase, modulates embryonic cell proliferation through the regulation of Cdkn1a expression. Development, 139, 749-759. ISMAIL, A., TOH, H., TREMBLAY, M., HADDAD, L. & BATEMAN, A. 2008. BHLHB2 and BHLHB3 are targets of FGF-4 and promotes proliferation and anchorage independent cell growth in SW-13 carcinoma. Cancer Res, 68, LB- 16. ISRAELSEN, W., DAYTON, T., DAVIDSON, S., FISKE, B., HOSIOS, A., BELLINGER, G., LI, J., YU , Y., SASAKI, M., HORNER, J., BURGA, L., XIE, J., JURCZAK, M., DEPINHO, R., CLISH, C., JACKS, T., KIBBEY, R., WULF, G., DI VIZIO, D., MILLS, G., CANTLEY, L. & VANDER HEIDEN, M. 2013. PKM2 isoform-specific deletion reveals a differential requirement for pyruvate kinase in tumor cells. Cell, 155, 397-409. ITO, A., KAWAGUCHI, Y., LAI, C., KOVACS, J., HIGASHIMOTO, Y., APPELLA, E. & YAO, T. 2002. MDM2-HDAC1-mediated deacetylation of p53 is required for its degradation. EMBO J, 21, 6236-45. ITONAGA, H., IMANISHI, D., WONG, Y., SATO, S., ANDO, K., SAWAYAMA, Y., SASAKI, D., TSURUDA, K., HASEGAWA, H., IMAIZUMI, Y., TAGUCHI, J., TSUSHIMA, H., YOSHIDA, S., FUKUSHIMA, T., HATA, T., MORIUCHI, Y., YANAGIHARA, K. & MIYAZAKI, Y. 2014. Expression of myeloperoxidase in acute myeloid leukemia blasts mirrors the distinct DNA methylation pattern involving the downregulation of DNA methyltransferase DNMT3B. Leukemia, 28, 1459-66. ITZYKSON, R. & FENAUX, P. 2014. Epigenetics of myelodysplastic syndromes. Leukemia, 28, 497-506. IULIANELLA, A., SHARMA, M., VANDEN HEUVEL, G. & TRAINOR, P. 2009. Cux2 functions downstream of Notch signaling to regulate dorsal interneuron formation in the spinal cord. Development, 136, 2329-2334. IYER, N., KOTCH, L., AGANI, F., LEUNG, S., LAUGHNER, E., WENGER, R., GASSMANN, M., GEARHART, J., LAWLER, A., YU, A. & SEMENZA, G. L. 1998. Cellular and developmental control of O2 homeostasis by hypoxia- inducible factor 1α. Genes Dev, 149-62.

238

IYER, N., OZDAG, H. & CALDAS, C. 2004. p300/CBP and cancer. Oncogene, 23, 4225-31. IZYUMOV, D., AVETISYAN, A., PLETJUSHKINA, O., SAKHAROV, D., WIRTZ, K., CHERNYAK, B. & SKULACHEV, V. 2004. "Wages of fear": transient threefold decrease in intracellular ATP level imposes apoptosis. Biochim Biophys Acta, 1658, 141-7. JADVAR, H., ALAVI, A. & GAMBHIR, S. 2009. 18F-FDG uptake in lung, breast, and colon cancers: molecuar biology correlates and disease characterisation. J Nucl Med, 50, 1820-1827. JAJU, R. J., FIDLER, C., HAAS, O. A., STRICKSON, A. J., WATKINS, F., CLARK, K., CROSS, N. C., CHENG, J. F., APLAN, P. D., KEARNEY, L., BOULTWOOD, J. & WAINSCOAT, J. S. 2001. A novel gene, NSD1, is fused to NUP98 in the t(5;11)(q35;p15.5) in de novo childhood acute myeloid leukemia. Blood, 98, 1264–1267. JAKOVCEVSKI, M. & AKBARIAN, S. 2012. Epigenetic mechniams in neurological disease. Nat Med, 18, 1194-204. JAMIESON, C., ALILLES, L., DYLLA, S., MUIJTJENS, M., JONES, C., ZEHNDER, J., GOTLIB, J., LI, K., MANZ, M., KEATING, A., SAWYERS, C. & WEISSMAN, I. 2004. Granulocyte-macrophage progenitors as candidate leukemic stem cells in blast-crisis CML. N Eng J Med, 351, 657-67. JANG, M., KI, Y., WON, H., LIM, S., JYOTHI, K., DASHDORJ, A., MIN, Y., KIM, S., SHOKAT, K., HA, J. & KIM, S. 2012. Carbonyl Reductase 1 Offers a Novel Therapeutic Target to Enhance Leukemia Treatment by Arsenic Trioxide. Cancer Res, 72, 1-11. JASTANIAH, W., ABRAR, M. & KHATTAB, T. 2012. Improved outcome in pediatric AML due to augmented supportive care. Pediatr Blood Cancer, 59, 919-21. JAWAD, M., SEEDHOUSE, C., RUSSELL, N. & PLUMB, M. 2006. Polymorphisms in human homeobox HLX1 and DNA repair RAD51 genes increase the risk of therapy-related acute myeloid leukemia. Blood, 108, 3916-8. JEANNET, G., SCHELLER, M., SCARPELLINO, L., DUBOUX, S., GARDIOL, N., BACK, J., KUTTLER, F., MALANCHI, I., BIRCHMEIER, W., LEUTZ, A. & AL, E. 2008. Long-term, multilineage hematopoiesis occurs in the combined absence of {beta}-catenin and {gamma}-catenin. Blood, 111, 142-149.

239

JENAL, M., BATLINER, J., REDDY, V., HAFERLACH, T., TOBLER, A., FEY, M., TORBETT, B. & TSCHAN, M. 2010. The anti-apoptotic gene BCL2A1 is a novel transcriptional target of PU.1. Leukemia, 24, 1073-1076. JENUWEIN, T. & ALLIS, C. 2001. Translating the histone code. Science, 293, 1074- 80. JIA, X., YANG, W., ZHU, X., YANG, X., ZHANG, J. & WANG, K. 2011. Transcriptional regulation of BHLHB2 induced by PML-RARa. Zhongguo Shi Yan Xue Ye Xue Za Zhi, 19, 1005-9. JIN, Y., JEON, E., LI, Q., LEE, Y., CHOI, J., KIM, W., LEE, K. & BAE, S. 2004. Transforming growth factor-beta stimulates p300-dependent RUNX3 acetylation, which inhibits ubiquitination-mediated degradation. J Biol Chem, 279, 29409-17. JINEK, M., CHYLINSKI, K., FONFARA, I., HAUER, M., DOUDNA, J. & CHARPENTIER, E. 2012. A programmable dual-RNA-guided DNA endonuclease in adapive bacterial immunity. Science, 337, 816-21. JONES, P. & BAYLIN, S. 2002. The fundamental role of epigenetic events in cancer. Nat Rev Genet, 3, 415-428. JONES, P. & BAYLIN, S. 2007. The epigenomics of cancer. Cell, 128, 683-692. JUERGENS, R., WRANGLE, J., VENDETTI, F., MURPHY, S., ZHAO, M., COLEMAN, B., SEBREE, R., RODGERS, K., HOOKER, C., FRANCO, N., LEE, B., TSAI, S., DELGADO, I., RUDEK, M., BELINSKY, S., HERMAN, J., BAYLIN, S., BROCK, M. & RUDIN, C. 2011. Combination Epigenetic Therapy Has Efficacy in Patients with Refractory Advanced Non–Small Cell Lung Cancer. Cancer Discov, 1, 598-607. JUNN, E., HAN, S., IM, J., YANG, Y., CHO, E., UM, H., KIM, D., LEE, K., HAN, P., RHEE, S. & CHOI, I. 2000. Vitamin D3 up-regulated protein 1 mediates oxidative stress via suppressing the thioredoxin function. J Immunol, 164, 6287- 6295. KANEDA, A., FUJITA, T., ANAI, M., YAMAMOTO, S., NAGAE, G., MORIKAWA, M., TSUJI, S., OSHIMA, M., MIYAZONO, K. & ABURATANI, H. 2011. Activation of Bmp2-Smad1 signal and its regulation by coordinated alteration of H3K27 trimethylation in Ras induced senescence. PloS Genet, 7, e1002359.

240

KARDOS, G., ZWAAN, C., KASPERS, G., DE-GRAAF, S., DE BONT, E., POSTMA, A., BOKKERINK, J., WEENING, R., VAN DER DOES-VAN DEN BERG, A., WERING, E., KORBIJN, C. & HAHLEN, K. 2005. Treatment strategy and results in children treated on three Dutch Childhood Oncology Group acute myeloid leukemia trials. Leukemia, 19, 2063-2071. KASSNER, N., HUSE, K., MARTIN, H., GODTEL-ARMBRUST, U., METZGER, A., MEINEKE, I., BROCKMOLLER, J., KLEIN, Z., ZANGER, U., MASER, E. & WOJNOWSKI, L. 2008. Carbonyl reductase 1 is a predominant doxorubicin reductase in the human liver. Drug Metab Dispos, 36, 2113-20. KATOH, M. & KATOH, M. 2007. Comparative integromics on JMJD1C gene encoding histone demethylase: Conserved POU5F1 binding site elucidating mechanism of JMJD1C expression in undifferentiated ES cells and diffuse-type gastric cancer. Int J Oncol, 31, 219-223. KAWAGOE, H., HUMPHRIES, R., BLAIR, A., SUTHERLAND, H. & HOGGE, D. 1999. Expression of HOX genes, HOX cofactors, and MLL in phenotypically and functionally defined subpopulations of leukemic and normal human hematopoietic cells. Leukemia, 13, 687-698. KEITH, B. & SIMON, M. 2007. Hypoxia-inducible factors, stem cells, and cancer. Cell, 129, 465-472. KELLY, P., DAKIE, A., ADAMS, J., NUTT, S. & STRASSER, A. 2007. Tumor growth need not be driven by rare cancer stem cells. Science, 317, 337. KENDZIORRA, E., SPITZNER, M., EMONS, G., AHLBORN, K., GAEDCKE, J., RAVE-FRANK, M., BECKER, H., BEIßBARTH, T., GHADIMI, T., RIED, T. & GRADE, M. 2010. Inhibition von TCF7L2 mittels RNA-Interferenz sensibilisiert kolorektale Krebszellen für Bestrahlung. Chirurgisches Forum und DGAV Forum 2010, Deutsche Gesellschaft für Chirurgie. Springer Berlin Heidelberg. KENNEDY, M., RAYNER, J. & MORRIS, C. 1994. Genomic structure, promoter sequence, and revised translation of human homeobox gene HLX1. Genomics, 22, 348-55. KETLEY, N., ALLEN, P., KELSEY, S. & NEWLAND, A. 2000. Mechanisms of resistance to apoptosis in human AML blasts: the role of differentiation-induced perturbations of cell-cycle checkpoints. Leukemia, 14, 620-8.

241

KIEL, M., YILMAZ, O., IWASHITA, T., YILMAZ, O., TERHORST, C. & MORRISON, S. 2005. SLAM family receptors distinguish hematopoietic stem and progenitor cells and reveal endothelial niches for stem cells. Cell, 121, 1109-1121. KIM, C., JACKSON, E., WOOLFENDEN, A., LAWRENCE, S., BABAR, I., VOGEL, S., CROWLEY, D., BRONSON, R. & JACKS, T. 2005. Identification of bronchioalveolar stem cells in normal lung and lung cancer. Cell, 121, 823-5. KIM, J. W., TCHERNYSHYOV, I., SEMENZA, G. & DANG, C. 2006. HIF-1- mediated expression of pyruvate dehydrogenase kinase: a metabolic switch required for cellular adaptation to hypoxia. Cell Metab, 3, 177-185. KIM, S., KIM, J., CHOE, N., CHO, I., KIM, J., KIM, D., SEOL, J., LEE, S., KOOK, H., NAM, K., KOOK, H., BHAK, Y. & SEO, S. 2010. Regulation of mouse steroidogenesis by WHISTLE and JMJD1C through histone methylation balance. Nucleic Acids Res, 38, 6389-6403. KING, O., LI, X., SAKURAI, M., KAWAMURA, A., ROSE, N., NG, S., QUINN, A., RAI, G., MOTT, B., BESWICK, P., KLSOE, R., OPPERMANN, U., JADHAV, A., HEIGHTMAN, T., MALONEY, D., SCHOFIELD, C. J. & SIMEONOV, A. 2010. Quantitative high-throughput screening identifies 8-hydroxyquinolines as cell-active histone demethylase inhibitors. PLoS One, 5, 15535. KINJO, K., KIZAKI, M., MUTO, A., FUKUCHI, Y., UMEZAWA, A., YAMATO, K., NICHIHARA, T., HATA, J., ITO, M., UEYAMA, Y. & IKEDA, Y. 2000. Arsenic trioxide (As2O3)-induced apoptosis and differentiation in retinoic acid- resistant acute promyelocytic leukemia model in hGM-CSF-producing transgenic SCID mice. Leukemia, 14, 431-8. KIOUSSI, C., BRIATA, P., BAEK, S., ROSE, D., HAMBLE, T. N., HERMAN, T., OHGI, K., LIN, C., GLEIBERMAN, A., WANG, J., BRAULT, V., RUIZ- LOZANO, P., NGUYEN, H., KEMLER, R., GLASS, C., WYNSHAW-BORIS, A. & ROSENFELD, M. 2002. Identification of a Wnt/Dvl/beta-Catenin-> Pitx2 pathway mediating cell-type-specific proliferation during development. Cell, 111, 673-85. KIRKIN, V., JOOS, S. & ZORNIG, M. 2004. The role of Bcl-2 family members in tumorigenesis. Biochim Biophys Acta, 1644, 229-49.

242

KLEER, C., CAO, Q., VARAMBALLY, S., SHEN, R., OTA, I., TOMLINS, S., GHOSH, D., SEWALT, R., OTTE, A. P., HAYES, D., SABEL, M., LIVANT, D., WEISS, S., RUBIN, M. & CHINNAIYAN, A. 2003. EZH2 is a marker of aggressive breast cancer and promotes neoplastic transformation of breast epithelial cells. Proc Natl Acad Sci USA, 100, 11606-11. KLOSE, R. & BIRD, A. 2006. Genomic DNA methylation: The marks and its mediators. Trends Biochem Sci, 31, 89-97. KLOSE, R., KALLIN, E. & ZHANG, Y. 2006. JmjC-domain-containing proteins and histone demethylation. Nat Rev Genet, 7, 715-727. KNUTSON, S., KAWANO, S., MINOSHIMA, Y., WARHOLIC, N., HUANG, K., XIAO, Y., KADOWAKI, T., UESUGI, M., KUZNETSOV, G., KUMAR, N., WIGLE, T., KLAUS, C., ALLAIN, C., RAIMONDI, A., WATERS, N., SMITH, J., PORTER-SCOTT, M., CHESWORTH, R., MOYER, M., COPELAND, R., RICHON, V., UENAKA, T., POLLOCK, R., KUNTZ, K., YOKOI, A. & KEILHACK, H. 2014. Selective inhibition of EZH2 by EPZ- 6438 leads to potent antitumor activity in EZH2-mutant non-hodgkin lymphoma. Mol Cancer Ther, 13, 824-54. KNUTSON, S., WARHOLIC, N., WIGLE, T., KLAUS, C., ALLAIN, C., RAIMONDI, A., PORTER-SCOTT, M., CHESWORTH, R., MOYER, M., COPELAND, N., RICHON, V., POLLOCK, R., KUNTZ, K. & KEILHACK, H. 2013. Durable tumor regression in genetically altered malignant rhabdoid tumors by inhibition of methyltrnasferase EZH2. Proc Natl Acad Sci USA, 110, 7922-7. KNUTSON, S., WIGLE, T., WARHOLIC, N., SNEERINGER, C., ALLAIN, C., KLAUS, C., SACKS, J., RAIMONDI, A., MAJER, C., SONG, J., SCOTT, M., JIN, L., SMITH, J., OLHAVA, E., CHESWORTH, R., MOYER, M., RICHON, V., COPELAND, R., KEILHACK, H., POLLOCK, R. & KUNTZ, K. 2012. A selective inhibitor of EZH2 blocks H3K27 methylation and kills mutant lymphoma cells. Nat Chem Biol, 8, 890-6. KO, Y., PEDERSON, P. & GESCHWIND, J. 2001. Glucose catabolism in the rabbit VX2 tumor model for liver cancer: characterization and targeting hexokinase. Cancer Lett, 173, 83-91.

243

KOCH, U., WILSON, A., COBAS, M., KEMLER, R., MACDONALD, H. & RADTKE, F. 2008. Simultaneous loss of {beta}- and {gamma}-catenin does not perturb hematopoiesis or lymphopoiesis. Blood, 111, 160-164. KOHLMANN, A., KIPPS, T., RASSENTI, L., DOWNING, J., SHURTLEFF, S., MILLS, K., GILKES, A., HOFMANN, W., BASSO, G., DELL'ORTO, M., FOA, R., CHIARETTI, S., VOS, J., RAUHUT, S., PAPENHAUSEN, P., HERNANDEZ, J., LUMBRERAS, E., YEOH, A., KOAY, E., LI, R., LIU, W., WILLIAMS, P., WIERCZOREK, L. & HAFERLACH, T. 2008. An international standardization programme towards the application of gene expression profiling in routine leukaemia diagnostics: the Microarray Innovations in LEukemia study prephase. Br J Haematol, 142, 802-7. KONISHI, J., OIZUMI, S., MARQUEZ, V., NISHIMURA, M. & DOSAKA-AKITA, H. 2012. Epigenetic therapy with 3-deazaneplanocin A, an inhibitor of the histone methyltransferase EZH2, inhibits growth of non-small cell lung cancer cells. Clin Cancer Res, 18, 4201-12. KONOPLEVA, M. Y. & JORDAN, C. T. 2011. Leukemia Stem Cells and Microenvironment: Biology and Therapeutic Targeting. J Clin Oncol, 29, 591- 9. KONZE, K., MA, A., LI, F., BARSYTE-LOVEJOY, D., PARTON, T., MACNEVIN, C., LIU, F., GAO, C., HUANG, X., KUZNETSOVA, E., ROUGIE, M., JIANG, A., PATTENDEN, S., NORRIS, J., JAMES, L., ROTH, B., BROWN, P., FRYE, S., ARROWSMITH, C., HAHN, K., WANG, G., VEDADI, M. & JIN, J. 2013. An orally bioavailable chemical probe of the lysine methyltrnasferases EZH2 and EZH1. ACS Chem Biol, 8, 1324-34. KORNBERG, A. & POLLIACK, A. 1980. Serum lactic dehydrogenase (LDH) levels in acute leukemia: marked elevations in lymphoblastic leukemia. Blood, 56, 351-5. KOSMIDER, O., GELSI-BOYER, V., SLAMA, L., DREYFUS, F., BEYNE-RAUZY, O., QUESNEL, B., HUNAULT-BERGER, M., SLAMA, B., VEY, N., LACOMBE, C., SOLARY, E., BIRNBAUM, D., BERNARD, O. & FONTENAY, M. 2010. Mutations of IDH1 and IDH2 genes in early and accelerated phases of myelodysplastic syndromes and MDS/myeloproliferative neoplasms. Leukemia, 24, 1094-6.

244

KOTAKE, Y., CAO, R., VIATOUR, P., SAGE, J., ZHANG, Y. & XIONG, Y. 2007. pRB family proteins are required for H3K27 trimethylation and Polycomb repression complexes binding to and silencing p16INK4alpha tumor suppressor gene. Genes Dev, 21, 49-54. KOUZARIDES, T. 2007. Chromatin modifications and their function. Cell, 128, 693- 705. KRAGULJAC KURTOVIC, N., KRAJNOVIC, M., BOGDANOVIC, A., SUVAJDZIC, N., DIMITRIJEVIC, B., COLOVIC, M. & KRTOLICA, K. 2012. Concomitant aberrant methylation of p15 and MGMT genes in acute myeloid leukemia: association with a particular immunophenotype of blast cells. Med Oncol, 29, 3547-56. KRIEG, A., RANKIN, E., CHAN, D., RAZORENOVA, O., FERNANDEZ, S. & GIACCIA, A. 2010. Regulation of the Histone Demethylase JMJD1A by Hypoxia-Inducible Factor 1α Enhances Hypoxic Gene Expression and Tumor Growth. Mol Cell Biol, 30, 344-353. KRIVTSOV, A. & ARMSTRONG, S. 2007. MLL translocations histone modifications and leukaemia stem-cell development. Nat Rev Cancer, 7, 828-833. KRIVTSOV, A., FIGUEROA, M., SINHA, A., STUBBS, M., FENG, Z., VALK, P., DELWEL, R., DOHNER, K., BULLINGER, L., KUNG, A., MELNICK, A. & ARMSTRONG, S. 2013. Cell of origin determines clinically relevant subtypes of MLL-rearranged AML. Leukemia, 27, 852-60. KRIVTSOV, A., TWOMEY, D., FENG, Z., STUBBS, M., WANG, Y., FABER, J., LEVINE, J., WANG, J., HAHN, W., GILLILAND, D., GOLUB, T. & ARMSTRONG, S. 2006. Transformation from committed progenitor to leukemia stem cell initiated by MLL-AF9. Nature, 442, 818-822. KRIVTSOV, A. V., FIGUEROA, M. E., SINHA, A. U., STUBBS, M. C., FENG, Z., VALK, P. J., DELWEL, R., DOHNER, K., BULLINGER, L., KUNG, A. L., MELNICK, A. M. & ARMSTRONG, S. A. 2012. Cell of origin determines clinically relevant subtypes of MLL-rearranged AML. Leukemia, 27, 852-60. KROON, E., KROSL, J., THORSTEINSDOTTIR, U., BABAN, S., BUCHBERG, A. & SAUVAGEAU, G. 1998. Hoxa9 transforms primary bone marrow cells through specific collaboration with Meis1a but not Pbx1b. EMBO J, 17, 3714-25.

245

KUMAR, A., HUDSON, W., CHEN, W., NISHIUCHI, R., YAO, Q. & KERSEY, J. 2004. Hoxa9 influences the phenotype but not the incidence of MLL-AF9 fusion gene leukemia. Blood, 103, 1823-8. KUMAR, C. 2011. Genetic Abnormalities and Challenges in the Treatment of Acute Myeloid Leukemia. Genes and Cancer, 20, 1-13. KUO, A. J., CHEUNG, P., CHEN, K., ZEE, B. M., KIOI, M., LAURING, J., XI, Y., PARK, B. H., SHI, X., GARCIA, B. A., LI, W. & GOZANI, O. 2011. NSD2 links dimethylation of histone H3 at lysine 36 to oncogenic programming. Mol Cell, 44, 609-620. KUROKI, S., AKIYOSHI, M., TOKURA, M., MIYACHI, H., NAKAI, Y., KIMURA, H., SHINKAI, Y. & TACHIBANA, M. 2013. JMJD1c, a JmjC domain- containing protein, is required for long-term maintenance of male germ cells in mice. Biol Reprod, 89, 93. LACHNER, M., O‘SULLIVAN, R. & JENUWEIN, T. 2003. An epigenetic road map for histone lysine methylation. J Cell Sci, 116, 2117–2124. LAGADINOU, E., SACH, A., CALLAHAN, K., ROSSI, R., NEERING, S., MINHAJUDDIN, M., ASHON, J., PEI, S., GROSE, V., O'DWYER, K., LIESVELD, J., BROOKES, P., BECKER, M. & JORDAN, C. 2013. BCL-2 inhibition targets oxidative phosporyation and selectively eradicates quiescent human leukemia stem cells. Cell Stem Cell, 12, 329-341. LAL, S., SANDANARAJ, E., WONG, Z., ANG, P., WONG, N., LEE, E. & CHOWBAY, B. 2008. CBR1 and CBR3 pharmacogenetics and their influence on doxorubicin disposition in Asian breast cancer patients. Cancer Sci, 99, 2045-54. LAN, F. & SHI, Y. 2009. Epigenetic regulation: methylation of histone and non-histone proteins. Sci China C Life Sci, 52, 311-22. LANDAU, D., CLEMENT, K., ZILLER, M., BOYLE, P., FAN, J., GU, H., STEVENSON, K., SOUGNEZ, C., WANG, L., LI, S., KOTLIAR, D., ZHANG, W., GHANDI, M., GARRAWAY, L., FERNANDES, S., LIVAK, K., GABRIEL, S., GNIRKE, A., LANDER, E., BROWN, J., NEUBERG, D., KHARCHENKO, P., HACOHEN, N., GETZ, G., MEISSNER, A. & WU, C. 2014. Locally disordered methylation forms the basis of intratumor methylome variation in chronic lymphocytic leukemia. Cancer Cell, 26, 813-25.

246

LANDRY, J., SUTTON, A., HESMAN, T., MIN, J., XU, R. M., JOHNSTON, M. & STERNGLANZ, R. 2003. Set2-catalyzed methylation of histone H3 represses basal expression of GAL4 in Saccharomyces cerevisiae. Mol Cell, 23, 5972-8. LANE, S. & GILLILAND, G. 2010. Leukemia stem cells. Seminars in Cancer Biology, 20, 71-76. LANE, S., SCADDEN, D. & GILLILAND, D. 2009. The leukemic stem cell niche- current concepts and therapeutic opportunities. Blood, 114, 1150-1157. LAPIDOT, T., SIRARD, C., VORMOOR, J., MURDOCH, B., HOANG, T., CACERES-CORTES, J., MINDEN, M., PATERSON, B., CALIGIURI, M. & DICK, J. 1994. A cell initiating human acute myeloid leukaemia after transplantation into SCID mice. Nature, 367, 645-648. LARSON, E., DOUGHMAN, D., GREGERSON, D. & OBRITSCH, W. 1997. A new, simmple, nonradioactive, nontoxic in vitro assay to monitor corneal endothelial cell viability. Invest Opthalmol Vis Sci, 38, 1929-33. LEE, H., BAE, S., KIM, K., KIM, W., CHUNG, S., YANG, Y. & YOON, Y. 2011. Shikonin inhibits adipogeneseis by modulation of the WNT/B-catenin pathway. Life Sci, 88, 294-301. LEE, H., CHOI, K., OH, H., PARK, Y. & PARK, Y. 2014. HIF-1 Dependent induction of Jumunji Domain-Containing Protein (JMJD) 3 under hypoxic conditions. Mol Cells, 37, 43-50. LEE, J., CHOI, H.-S., GYURIS, J., BRENT, R. & MOORE, D. 1995. Two classes of proteins dependent on either the presence or absence of thyroid hormone for interaction with the thyroid hormone receptor. Mol Endocrinol, 9, 243-254. LEE, M., WYNDER, C., SCHMIDT, D., MCCAFFERTY, D. & SHIEKHATTAR, R. 2006. Histone H3 lysine 4 demethylation is a target of nonselective antidepressive medications. Chem Biol, 13, 563-567. LEHMBECHER, T., VARWIG, D., KAISER, J., REINHARDT, D., KLINGEBIEL, T. & CREUTZIG, U. 2004. Infectious complications in paediatric acute myeloid leukaemia: analysis of the prospective multi-institutional clinical trial AML- BFM 93. Leukemia, 18, 72-77. LÉVESQUE, J., HELWANI, F. & WINKLER, I. 2010. The endosteal 'osteoblastic' niche and its role in hematopoietic stem cell homing and mobilization. Leukemia, 24, 1979-92.

247

LEY, T. J., DING, L., WALTER, M., MCLELLAN, M., LAMPRECHT, T., LARSON, D., KANDOTH, C., PAYTON, J., BATY, J., WELCH, J., HARRIS, C., LICHTI, C., TOWNSEND, R., FULTON, R., DOOLING, D., KOBOLDT, D., SCHMIDT, H., ZHANG, Q., OSBORNE, J., LIN, L., O'LAUGHLIN, M., MCMICHAEL, J., DELEHAUNTY, K., MCGRATH, S., FULTON, L., MAGRINI, V., VICKERY, T., HUNDAL, J., COOK, L., CONYERS, J., SIFT, G., REED, J., ALLDREDGE, P., WYLIE, T., WALKER, J., KALICKI, J., WATSON, M., HEATH, S., SHANNON, W., VARGHESE, N., NAGARAJAN, R., WESTERVELT, P., TOMASSON, M., LINK, D., GRAUBERT, T., DIPERSIO, J., MARDIS, E. & WILSON, R. 2010. DNMT3A mutations in acute myeloid leukemia. N Eng J Med, 363, 2424-33. LEY, T. J., MARDIS, E., DING, L., FULTON, B., MCLELLAN, M., CHEN, K., DOOLING, D., DUNFORD-SHORE, B., MCGRATH, S., HICKENBOTHAM, M., COOK, L., ABBOTT, R., LARSON, D., KOBOLDT, D., POHL, C., SMITH, S., HAWKINS, A., ABBOTT, S., LOCKE, D., HILLIER, L., T, M., FULTON, L., MAGRINI, V., WYLIE, T., GLASSCOCK, J., CONYERS, J., SANDER, N., SHI, X., OSBORNE, J., MINX, P., GORDON, D., CHINWALLA, A., ZHAO, Y., RIES, R., PAYTON, J., WESTERVELT, P., TOMASSON, M., WATSON, M., BATY, J., IVANOVICH, J., HEATH, S., SHANNON, W., NAGARAJAH, R., WALTER, M., LINK, D., GRAUBERT, T., DIPERSIO, J. & WILSON, R. 2008. DNA sequencing of a cytogenetically normal acute myeloid leukaemia genome. Nat Biotechnol, 456, 66-72. LI, Y., TROJER, P., XU, C. F., CHEUNG, P., KUO, A., DRURY, W. J., 3RD, QIAO, Q., NEUBERT, T. A., XU, R. M., GOZANI, O. & REINBERG, D. 2009. The target of the NSD family of histone lysine methyltransferases depends on the nature of the substrate. J Biol Chem, 284, 34283-34295. LIBERZON, A., SUBRAMANIAN, A., PINCHBACK, R., THORVALDSDOTTIR, H., TAMAYO, P. & MESIROV, J. 2011. Molecular signatures database (MSigDB)3.0. Bioinformatics, 27, 1739-1740. LIBREROS, S., GARCIA-AREAS, R. & IRAGAVARAPU-CHARYULU, V. 2013. CHI3L1 plays a role in cancer through enhanced production of pro- inflammatory/pro-tumorigenic and angiogenic factors. Immunol Res, 57, 99-105.

248

LIE, S., ABRAHAMSSON, J., CLAUSEN, N., FORESTIER, E., HASLE, H., HOVI, L., JONMUNDSSON, G., MELLANDER, L., SIIMES, M., YSSING, M., ZELLER, B. & GUSTAFSSON, G. 2005. Long-term results in children with AML: NOPHO-AML Study Group-report of three consecutive trials. Leukemia, 19, 2090-100. LIGGETT, W. J. & SIDRANSKY, D. 1998. Role of the p16 tumor suppressor gene in cancer. J Clin Oncol, 16, 1197-206. LIM, E., VAILLANT, F., WU, D., FORREST, N., PAL, B., HART, A., ASSELIN- LABAT, M., GYORKI, D., WARD, T., PARTANEN, A., FELEPPA, F., HUSCHTSCHA, L., THORNE, H., KCONFAB, FOX, S., YAN, M., FRENCH, J., BROWN, M., SMYT, G., VISVADER, J. & LINDEMAN, G. 2009. Aberrant luminal progenitors as the candidate target population for basal tumor development in BRCA1 mutation carriers. Nat Med, 15, 907-13. LINET, M., RIES, L., SMITH, M., TARONE, R. & DEVESA, S. 1999. Cancer surveillance series: recent trends in childhood cancer incidence and mortality in the United States. J Natl Cancer Inst, 91, 1051-1058. LING, C. & GROOP, L. 2009. Epigenetics: A Molecular Link Between Environmental Factors and Type 2 Diabetes. Diabetes, 58, 2718-25. LO-COCO, F., AVVISATI, G., BIGNETTI, M., THIEDE, C., ORLANDO, S., IACOBELLI, S., FERRARA, F., FAZI, P., CICCONI, L., DI BONA, E., SPECCHIA, G., SICA, S., DIVONA, M., LEVIS, A., FIEDLER, W., CERQUI, E., BRECCIA, FIORITONI, G., SALIH, H., CAZZOLA, M., MELILLO, L., CARELLA, A., BRANDTS, C., MORRA, E., VON LILLIENFELD-TOAL, M., HERTENSSTEIN, B., WATTAD, M., LUBBERT, M., FABBIANO, F., DOHNER, K., SAUER, M., GANSER, A., AMADORI, S., MANDELLI, F., DOHNER, H., EHNINGER, G., SCHLENK, R., PLATZBECKER, U., GRUPPO ITALIANO MALATTIE EMATOLOGICHE DELL'ADULTO, GERMAN-AUSTRIAN ACUTE MYELOID LEUKEMIA STUDY GROUP & STUDY ALLIANCE LEUKEMIA 2013. Retinoic acid and arsenic trioxide for acute promyelocytic leukemia. N Eng J Med, 369, 111-21. LOBODA, A., JOZKOWICZ, A. & DULAK, J. 2010. HIF-1 and HIF-2 transcription factors- similar but not identical. Mol Cells, 29, 435-42.

249

LODI, A., TIZIANI, S., KHANIM, F., DRAYSON, M., GUNTHER, U., BUNCE, C. & VIANT, M. 2011. Hypoxia triggers major metabolic changes in AML cells without altering indomethacin-induced TCA cycle deregulation. ACS Chem Biol, 6, 169-75. LOH, Y., ZHANG, W., CHEN, X., GEORGE, J. & NG, H. 2007. Jmjd1a and Jmjd2c histone H3 Lys 9 demethylases regulate self-renewal in embryonic stem cells. Genes Dev, 21, 2545-2557. LOHSE, B., KRISTENSEN, J., KRISTENSEN, L., AGGER, K., HELIN, K., GAJHEDE, M. & CLAUSEN, R. 2011a. Inhibitors of histone demethylases. Bioorg Med Chem, 19, 3625-3636. LOHSE, B., NIELSEN, A., KRISTENSEN, J., HELGSTRAND, C., CLOOS, P., OLSEN, L., GAJHEDE, M., CLAUSEN, R. & KRISTENSEN, J. 2011b. Targeting histone lysine demethylases by truncating the histone 3 tail to obtain selective substrate-based inhibitors. Angew Chem Int Ed, 50, 9100-9103. LU, J. & MATUNIS, M. 2013. A mediator methylation mystery: JMJD1c demethylates MDC1 to regulate DNA repair. Nat Struct Mol Biol, 20, 1346-8. LUGER, K., MADER, A., RICHMOND, R., SARGENT, D. & RICHMOND, T. 1997. Crystal structure of the nucleosome core particle at 2.8 A° resolution. Nature, 389, 251-260. LUND, A. & VAN LOHUIZEN, M. 2004. Epigenetics and Cancer. Genes Dev, 18, 2315-2335. MACK, G. 2006. Epigenetic Cancer Therapy Makes Headway J Natl Cancer Inst, 98, 1443-4. MAHER, J., KRISHAN, A. & LAMPIDIS, T. 2004. Greater cell cycle inhibition and cytotoxicity induced by 2-deoxy-D-glucose in tumor cells treated under hypoxic vs aerobic conditions. Cancer Chemother Pharmacol, 53, 116-22. MAI, A. & ALTUCCI, L. 2009. Epi-drugs to fight cancer: From chemistry to cancer treatment, the road ahead. Int J Biochem Cell Biol, 41, 199-213. MAILLARD, I., ADLER, S. & PEAR, W. 2003. Notch and the immune system. Immunity, 19, 781-91. MAJUMDER, B., BARANEEDHARAN, U., THIYAGARAJAN, S., RADHARKRISHNAN, P., NARASIMHAN, H., DHANDAPANI, M., BRIJWANI, N., PINTO, D., PRASATH, A., SHANTHAPPA, B.,

250

THAYAKUMAR, A., SURENDRAN, R., BABU, G., SHENOY, A., KURIAKOSE, M., BERGTHOLD, G., HOROWITZ, P., LODA, M., BEROUKHIM, R., AGARWAL, S., SENGUPTA, S., SUNDARAM, M. & MAJUMDER, P. 2015. Predicting clinical response to anticancer drugs using an ex vivo platform that captures tumour heterogeneity. Nat Commun, 6. MANI, R., MURTHY, S. & JAMIL, K. 2006. Role of serum lactate dehydrogenase as a bio-marker in therpay related hematological malignancies. Int J Can Res, 2. MARSHALL, S., BACOTE, V. & TRAXINGER, R. R. 1991. Discovery of a metabolic pathway mediating glucose‑ induced desensitization of the glucose transport system. Role of hexosamine biosynthesis in the induction of insulin resistance. J Biol Chem, 266, 4706-4712. MCCABE, M., OTT, H., GANJI, G., KORENCHUK, S., THOMPSON, C. B., VAN ALLER, G., LIU, Y., GRAVES, A., DELLA PETRA, A., DIAZ, E., LAFRANCE, L., MELLINGER, M., DUGUENNE, C., TIAN, X., KRUGER, R., MCHUGH, C., BRANDT, M., MILLER, W., DHANAK, D., VERMA, S., TUMMINO, P. & CREASY, C. 2012. EZH2 inhibition as a therapeutic strategy for lymphoma with EZH2-activating mutations. Nature, 492, 108-12. MCMAHON, K. A., HIEW, S., HADJUR, S., VEIGA-FERNANDES, H., MENZEL, U., PRICE, A., KIOUSSIS, D., WILLIAMS, O. & BRADY, H. 2007. Mll has a critical role in fetal and adult haematopoietic stem cell self-renewal. Cell Stem Cell, 1, 338-345. MERLO, A., HERMAN, J. G., MAO, L., LEE, D. J., GABRIELSON, E., BURGER, P. C., BAYLIN, S. B. & SIDRANSKY, D. 1995. 5' CpG island methylation is associated with transcriptional silencing of the tumour suppressor p16/CDKN2/MTS1 in human cancers. Nat Med, 1, 686-92. METZELER, K., MAHARRY, K., RADMACHER, M., MROZEK, K., MARGESON, D., BECKER, H., CURFMAN, J., HOLLAND, K., SCHWIND, S., WHITMAN, S., WU, Y., BLUM, W., POWELL, B., CARTER, T., WETZLER, M., MOORE, J., KOLITZ, J., BAER, M., CARROLL, A., LARSON, R., CALIGIURI, M., MARCUCCI, G. & BLOOMFIELD, C. 2011. TET2 mutations improve the new European LeukemiaNet risk classification of acute myeloid leukemia: a Cancer and Leukemia Group B study. J Clin Oncol, 29, 1373-81.

251

METZEN, E., STIEHL, D., DOEGE, K., MARXSEN, J., HELLWIG-BURGEL, T. & JELKMANN, W. 2005. Regulation of the prolyl hydroxylase domain protein 2 (phd2/egln-1) gene: identification of a functional hypoxia-responsive element. Biochem J, 387, 711-7. MEYER, C., HOFFMAN, J., BURMEISTER, T., GROGER, D., PARK, T., EMERENCIANO, M., POMBO DE OLIVEIRA, M., RENNEVILLE, A., VILLARESE, P., MACINTYRE, E., CAVE, H. & CLAPPIER, E. 2013. The MLL recombinome of acute leukemias in 2013. Leukemia, 27, 2165-76. MEYER, C., SCHNEIDER, B., JAKOB, S., STERHL, S., ATTARBASCHI, A., SCHNITTGER, S., SCHOCH, C., JANSEN, M., VAN DONGEN, J., DEN BOER, M., PIETERS, R., ENNAS, M., ANGELUCCI, E., KOEHL, U., GREIL, J., GRIESINGER, F., ZUR STADT, U., ECKERT, C., SZCZEPARISKI, T., NIGGLI, F., SCHAFER, B., KEMPSKI, H., BRADY, H., ZUNA, J., TRKA, J., NIGRO, L., BIONDI, A., DELABESSE, E., MACINTYRE, E., STANULLA, M., SCHRAPPER, M., HAAS, O., BURMEISTER, T., DINGERMANN, T., KLINGEBIEL, T. & MARSCHALEK, R. 2006. The MLL recombinome of acute leukemias. Leukemia, 20, 777-84. MICHELAKIS, E. D., WEBSTER, L. & MACKEY, J. R. 2008. Dichloroacetate (DCA) as a potential metabolictargeting therapy for cancer. Br J Cancer, 99, 989-94. MIHARA, K., IMAI, C., COUSTAN-SMITH, E., DOME, J., DOMINICI, M., VANIN, E. & CAMPANA, D. 2003. Development and functional characterization of human bone marrow mesenchymal cells immortalized by enforced expression of telomerase. Br J Haematol, 120, 846-9. MILLER, K. & DAOUST, P. 2005. Clinical manifestations of acute myeloid leukemia. In: HOFFMAN, R., BENZ, E., SHATTIL, S., FURIE, B., COHEN, H., SILBERSTEIN, L. & MCGLAVE, P. (eds.) Hematology: Basic Principles and Practice. 4th ed. Philadelphia: Elsevier. MILNE, T., MARTIN, M., BROCK, H., SLANY, R. & HESS, J. 2005. Leukemogenic MLL fusion proteins bind across a broad region of the Hox a9 locus, promoting transcription and multiple histone modifications. Cancer Res, 65, 11367-11374. MILOSEVIC, M., FYLES, A., HEDLEY, D. & HILL, R. 2004. The human tumor microenvironment: invasive (needle) measurement of oxygen and interstitial fluid pressure. Semin Radiat Oncol, 14, 249-58.

252

MINUCCI, S. & PELICCI, P. 2006. Histone deacetylase inhibitors and the promise of epigenetic (and more) treatments for cancer. Nat Rev Cancer, 6, 38-51. MIRANDA, T., CORTEZ, C., YOO, C., LIANG, G., ABE, M., KELLY, T., MARQUEZ, V. & JONES, P. 2009. DZNep is a global histone methylation inhibitor that reactivates developmental genes not silenced by DNA methylation. Mol Cancer Ther, 8, 1579-88. MIYAMOTO, Y., KOH, Y., PARK, Y., FUJIWARA, N., SAKIYAMA, H., MISONOU, Y., OOKAWARA, T., SUZUKI, K., HONKE, K. & TANIQUCHI, N. 2003. Oxidative stress caused by inactivation of glutathione peroxidase and adaptive responses. Biol Chem, 384, 567-74. MOHAN, M., HERZ, H. M., TAKAHASHI, Y., LIN, C., LAI, K. C., ZHANG, Y., WASHBURN, M. P., FLORENS, L. & SHILATIFARD, A. 2010. Linking H3K79 trimethylation to Wnt signaling through a novel Dot1-containing complex (DotCom). Genes Dev, 24, 574-589. MOLFESE, D. 2011. Advancing neuroscience through epigenetics: molecular mechanisms of learning and memory. Dev Neuropsychol, 36, 810-27. MOLYNEUX, G., GEYER, F., MAGNAY, F., MCCARTHY, A., KENDRICK, H., NATRAJAN, R., MACKAY, A., GRIGORIADIS, A., TUTT, A., ASHWORTH, A., REIS-FILHO, J. & SMALLEY, M. 2010. BRCA1 basal-like breast cancers originate from luminal epithelial progenitors and not from basal stem cells. Cell Stem Cell, 7, 403-17. MOMPARLER, R., COTE, S., MOMPARLER, L. & IDAGHDOUR, Y. 2014. Epigenetic therapy of acute myeloid leukemia using 5-aza-2'-deoxycytidine (decitabine) in combination with inhibitors of histone methylation and deacetylation. Clin Epigenetics, 6, doi:10.1186/1868-7083-6-19. MOON, R., KOHN, A., DE FERRARI, G. & KAYKAS, A. 2004. Wnt and b-catenin signalling: diseases and therapies. Nat Rev Genet, 5, 691-701. MORRISON, S., WANDYCZ, A., AKASHI, K., GLOBERSON, A. & WEISSMAN, I. 1996. The aging of hematopoietic stem cells. Nat Med, 2, 1011-1016. MORTENSEN, B., JENSEN, P., HELLEDIE, N., IVERSEN, P., RALFKIAER, E., LARSEN, J. & MADSEN, M. 1998. Changing bone marrow micro-environment during development of acute myeloid leukaemia in rats. Br J Haematol, 102.

253

MOSAMMAPARAST, N. & SHI, Y. 2010. Reversal of histone methylation: biochemical and molecular mechanisms of histone demethylases. Annu Rev Biochem, 79, 155-79. MOSS, T. & WALLRATH, L. 2007. Connections between epigenetic gene silencing and human disease. Mutat Res, 618, 163-74. MUNOZ-PINEDO, C., RUIZ-RUIZ, C., RUIZ DE ALMODOVAR, C., PALACIOS, C. & LOPEZ-RIVAS, A. 2003. Inhibition of glucose metabolism sensitizes tumor cells to death receptor-triggered apoptosis through enhancement of death- inducing signaling complex formation and apical procaspase-8 processing. J Biol Chem, 278, 12759-68. MYSSINA, S., JORGENSEN, G., YU, Y. & HOLYOAKE, T. 2009. 3-deazaneplanocin A (DZNep) activity against CD34+ leukaemia stem cells. Haematologica, 94, 129 abs. 0327. NAJAFOV, A., SOMMER, E., AXTEN, J., DEYOUNG, M. & ALESSI, D. 2011. Characterization of GSK2334470, a novel and highly specific inhibior of PDK1. Biochem J, 433, 357-69. NATALI, P., SALSANO, F., VIORA, M., NISTA, A., MALORNI, W., MAROLLA, A. & DE MARTINO, C. 1984. Inhibition of aerobic glycolysis in normal and neoplastic lymphoid cells induced by Lonidamine [1-(2,4-dichlorobenzyl)-I-H- indazol-3-carboxylic acid. Oncology, 41, Suppl 7-14. NEFF, T., SINHA, A., KLUK, M., ZHU, N., KHATTAB, M., STEIN, L., XIE, H., ORKIN, S. & ARMSTRONG, S. 2012. Polycomb repressive complex 2 is required for MLL-AF9 leukemia. Proc Natl Acad Sci USA, 109, 5028-33. NICOLAS, E., RAMUS, C., BERTHIER, S., ARLOTTO, M., BOUAMRANI, A., LEFEBVRE, C., MOREL, F., GARIN, J., IFRAH, N., BERGER, F., CAHN, J. & MOSSUZ, P. 2011. Expression of S100A8 in leukemic cells predicts poor survival in de novo AML patients. Leukemia, 25, 57-65. NIKOLOSKI, G., LANGEMEIJER, S. M., KUIPER, R. P., KNOPS, R., MASSOP, M., TONNISSEN, E. R., VAN DER HEIJDEN, A., SCHEELE, T. N., VANDENBERGHE, P., DE WITTE, T., VAN DER REIJDEN, B. & JANSEN, J. 2010. Somatic mutations of the histone methyltransferase gene EZH2 in myelodysplastic syndromes. Nat Genet, 42, 665-667.

254

O'BRIEN, J., WILSON, I., ORTON, T. & POGNAN, F. 2000. Investigation of the Alamar Blue (resazurin) fluorescent dye for the assessment of mammalian cell cytotoxicity. Eur J Biochem, 267, 5421-6. O'FLAHERTY, J., BARR, M., FENNELL, D., RICHARD, D., O'LEARY, J. & O'BYRNE, K. 2012. The cancer stem-cell hypothesis: its emerging role in lung cancer biology and its relevance for future therapy. J Thorac Oncol, 7, 1880-90. O‘DONNELL, M., ABBOUD, C., ALTMAN, J., APPELBAUM, F., COUTRE, S., DAMON, L., FORAN, J., GOORHA, S., MANESS, L., MARCUCCI, G., MASLAK, P., MILLENSON, M., MOORE, J., RAVANDI, F., SHAMI, P., SMITH, B., STONE, R., STRICKLAND, S., TALLMAN, M. & WANG, E. 2011. Acute myeloid leukemia. J Natl Compr Canc Netw, 9, 280-317. OH, S. & JANKNECHT, R. 2012. Histone demethylase JMJD5 is essential for embryonic development. Biochem and Biophys Res Com, 420, 61-65. OIKAWA, T., NAKAMURA, A., ONISHI, N., YAMADA, T., MATSUO, K. & SAYA, H. 2013. Acquired expression of NFATc1 downregulates E-cadherin and promotes cancer cell invasion. Cancer Res. OKADA, Y., FENG, Q., LIN, Y., JIANG, Q., LI, Y., COFFIELD, V., SU, L., XU, G. & ZHANG, Y. 2005. hDOT1L links histone methylation to leukemogenesis. Cell, 121, 167-178. ONNIS, B., RAPISARDA, A. & MELILLO, G. 2009. Development of HIF‑ 1 inhibitors for cancer therapy. J Cell Mol Med, 13, 2780-2786. OOI, A., MAH, V., NICKERSON, D., GILBERT, J., HA, V., HEGAB, A., HORVATH, S., ALAVI, M., MARSESH, E., CHIA, D., GOWER, A., LENBURG, M., SPIRA, A., SOLIS, L., WISTUBA, I., WALSER, T., WALLACE, W., DUBINETT, S., GOOGLICK, L. & GOMPERTS, B. 2010. Presence of a putative tumor-initiating progenitor cell population predicts poor prognosis in smokers with non-small cell lung cancer. Cancer Res, 70, 6639-48. OSAWA, M., HANADA, K., HAMADA, H. & NAKAUCHI, H. 1996. Long-term lymphohematopoietic reconstition by a single CD34-low/negative hematopoietic stem cell. Science, 273, 242-5. OUDARD, S., CARPENTIER, A., BANU, E., FAUCHON, F., CELERIER, D., POUPON, M., DUTRILLAUX, B., ANDRIEU, J. & DELATTRE, J. 2003.

255

Phase II study of lonidamine and diazepam in the treatment of recurrent glioblastoma multiforme. J Neurooncol, 63, 81-6. PANDOLFI, A., BARREYRO, L. & STEIDL, U. 2013. Concise Review: Preleukemic Stem Cells: Molecular Biology and Clinical Implications of the Precursors to Leukemia Stem Cells. Stem Cells Transl Med, 2, 143-150. PAPANDREOU, I., CAIRNS, R., FONTANA, L., LIM, A. & DENKO, N. C. 2006. HIF-1 mediates adaptation to hypoxia by actively downregulating mitochondrial oxygen consumption. Cell Metab, 3, 187-197. PARDANANI, A., LASHO, T., FINKE, C., MAI, M., MCCLURE, R. & TEFFERI, A. 2010. IDH1 and IDH2 mutation analysis in chronic- and blast-phase myeloproliferative neoplasms. Leukemia, 24, 1146-51. PASTOR, W., U, P., HUANG, Y., HENDERSON, H., LISTER, R., KO, M., MCLOUGHLIN, E., BRUDNO, Y., MAHAPATRA, S., KAPRANOV, P., TAHILIANI, M., DALEY, G., LIU, X., ECKER, J., MILOS, P., AGARWAL, S. & RAO, A. 2011. Genome-wide mapping of 5-hydroxymethylcytosine in embryonic stem cells. Nature, 473, 394-7. PATNAIK, M., HANSON, C., HODNEFIELD, J., LASHO, T., FINKE, C., KNUDSON, R., KETTERLING, R., PARDANANI, A. & TEFFERI, A. 2010. Differential prognostic effect of IDH1 versus IDH2 mutations in myelodysplastic syndromes: a Mayo Clinic study of 277 patients. Leukemia, 26, 101-5. PAUL, C., LILIEMARK, J., TIDEFELT, U., GAHRTON, G. & PETERSON, C. 1989. Pharmacokinetics of daunorubicin and doxorubicin in plasma and leukemic cells from patients with acute nonlymphoblastic leukemia. Ther Drug Monit, 11, 140- 8. PAULSSON, K., HEIDENBLAD, M., STROMBECK, B., STAAF, J., JONSSON, G., BORG, A., FIORETOS, T. & JOHANSSON, B. 2006. High-resolution genome- wide array-based comparative genome hybridization reveals cryptic chromosome changes in AML and MDS cases with trisomy 8 as the sole cytogenetic aberration. Leukemia, 20, 840-846. PEARCE, D., TAUSSIG, D., SIMPSON, C., ALLEN, K., ROHATINER, A., LISTER, T. & BONNET, D. 2005. Characterization of cells with a high aldehyde

256

dehydrogenase activity from cord blood and acute myeloid leukemia samples. Stem Cells, 23, 752-60. PEARN, L., FISHER, J., BURNETT, A. & DARLEY, R. 2007. The role of PKC and PDK1 in monocyte lineage specification by Ras. Blood, 109, 4461-9. PEEKEN, J., SEEGER, T., WEHRLE, J., SCHANNE, D., GOTHWAL, M., GRUNDER, A. & PAHL, H. 2013. Overexpression Of The Histone Demethylase JMJD1C In Polycythemia Vera Contributes To NF-E2 Overexpression Via Epigenetic Dysregulation and An Auto-Regulatory Loop. Blood, 122, 1602-1602. PELICANO, H., MARTIN, D., XU, R.-H. & HUANG, P. 2006. Glycolysis inhibition for anticancer treatment. Oncogene, 25, 4633-4646. PEREIRA, C., CLARKE, E. & DAMEN, J. 2007. Hematopoietic colony-forming cell assays. Methods Mol Biol, 407, 177-208. PEREL, Y., AUVRIGNON, A., LEBLANC, T., MICHEL, G., REGUERRE, Y., VANNIER, J., DALLE, J., GANDEMER, V., SCHMITT, C., MECHINAUD, F., LEJARS, O., PIGUET, C., COUILLAUD, G., PAUTARD, B., LANDMAN- PARKER, J., THURET, I., ALADJIDI, N., BARUCHEL, A. & LEVERGER, G. 2005. Treatment of childhood acute myeloblastic leukemia: dose intensification improves outcome and maintenance therapy is of no benefit- multicenter studies of the French LAME (Leucemie Aigue Myeloblastique Enfant) Cooperative Group. Leukemia, 19, 2082-2089. PERONA, R. & SANCHEZ-PEREZ, I. 2004. Control of oncogenesis and cancer therapy resistance. Br J Cancer, 90, 573-577. PERRY, M. 2008. The Chemotherapy source book, Philadelphia, Lippincott Williams & Wilkins. PESCADOR, N., CUEVAS, Y., NARANJO, S., ALCAIDE, M., VILLAR, D., LANDAZURI, M. & DEL PESO, L. 2005. Identification of a functional hypoxia-responsive element that regulates the expression of the egl nine homologue 3 (egln3/phd3) gene. Biochem, 390, 189-97. PESSION, A., RONDELLI, R., BASSO, G., RIZZARI, C., TESTI, M., FAGIOLI, F., STEFANO, P. & LOCATELLI, F. 2005. Treatment and long-term results in children with acute myeloid leukaemia treated according to the AIEOP AML protocols. Leukemia, 19, 2043-2053.

257

PETERSON, C. & LANIEL, M. 2004. Histones and histone modifications. Curr Biol, 14, R546-551. PETRINI, M., FELICETTI, F., BOTTERO, L., ERRICO, M., MORSILLI, O., BOE, A., DE FEO, A. & CARE, A. 2013. HOXB1 restored expression promotes apoptosis and differentiation in the HL60 leukemic cell line. Cancer Cell Int, 13, 101. POLAKIS, P. 2012. Drugging Wnt signalling in cancer. EMBO J, 31, 2737-2746. POLLARD, P., LOENARZ, C., MOLE, D., MCDONOUGH, M. A., GLEADLE, J., SCHOFIELD, C. J. & RATCLIFFE, P. 2008. Regulation of Jumonji-domain- containing histone demethylases by hypoxia-induble factor (HIF)-1a. Biochem J, 416, 387-94. POPOVIC, R. & LICHT, J. D. 2012. Emerging epigenetic targets and therapies in cancer medicine. Cancer Discov, 2, 405-13. POTHOF, J., HAAFTEN, G., THIJSSEN, K., KAMATH, R., FRASER, A., AHRINGER, J., PLASTERK, R. & TIJSTERMAN, M. 2003. Identification of genes that protect the C. elegans genome against mutations by genome-wide RNAi. Genes Dev, 17, 443-8. POZAROWSKI, P. & DARZYNKIEWICZ, Z. 2004. Analysis of cell cycle by flow cytometry. Methods Mol Biol, 281, 301-11. QI, W., CHAN, H., TENG, L., LI, L., CHUAI, S., ZHANG, R., ZENG, J., LI, M., FAN, H., LIN, Y., GU, J., ARDAYFIO, O., ZHANG, J., YAN, X., FANG, J., MI, Y., ZHANG, M., ZHOU, T., FENG, G., CHEN, Z., LI, G., YANG, T., ZHAO, K., LIU, X., YU, Z., LU, C., ATADJA, P. & LI, E. 2012. Selective inhibition of Ezh2 by a small molecule inhibitor blocks tumor cell proliferation. Proc Natl Acad Sci USA, 109, 21360-5. QUINTANA, J., ADVIS, P., BECKER, A., BERESI, V., CAMPBELL, M., VINES, E., GARCIA, H., SLAGADO, C., VARGAS, L., ROJAS, J., OBANDO, M., NEIRA, L., PAEZ, E. & ZOLEZZI, P. 2005. Acute myelogenous leukemia in Chile PINDA protocols 87 and 92 results. Leukemia, 19, 2143-2146. RADMARK, O., WERZ, O., STEINHILBER, D. & SAMUELSSON, B. 2007. 5- Lipoxygenase: regulation of expression and enzyme activity. Trends Biochem Sci, 32, 332-341.

258

RAI, K., HOLLAND, J., GLIDWELL, O., WEINBERG, V., BRUNNER, K., OBRECHT, J., PREISLER, H., NAWABI, I., PRAGER, D., CAREY, R., COOPER, M., HAURANI, F., HUTCHINSON, J., SILVER, R., FALKSON, G., WIERNIK, P., HOAGLAND, H., BLOOMFIELD, C., JAMES, G., GOTTLIEB, A., RAMANAN, S., BLOM, J., NISSEN, N., BANK, A., ELLISON, R., KUNG, F., HENRY, P., MCINTYRE, O. & KAAN, S. 1981. Treatment of acute myelocytic leukaemia: a study by cancer and leukaemia group B. Blood, 58, 1203-12. RAIMONDI, S., CHANG, M., RAVINDRANATH, Y., BEHM, F., GRESIK, M., STEUBER, C., WEINSTEIN, H. & CARROLL, A. 1999. Chromosomal Abnormalities in 478 Children With Acute Myeloid Leukemia: Clinical Characteristics and Treatment Outcome in a Cooperative Pediatric Oncology Group Study - POG 8821. Blood, 94, 3707-16. RAIS, B., COMIN, B., PUIGJANER, J., BRANDES, J., CREPPY, E., SABOUREAU, D., ENNAMANY, R., LEE, W., BOROS , L. & CASCANTE, M. 1999. Oxythiamine and dehydroepiandrosterone induce a G1 phase cycle arrest in Ehrlich's tumor cells through inhibition of the pentose cycle. FEBS Lett, 456, 113-8. RAN, D., SCHUBERT, M., PIETSCH, L., TAUBERT, I., WUCHTER, P., ECKSTEIN, V., BRUCKNER, T., ZOELLER, M. & HO, A. 2009. Aldehyde dehydrogenase activity among primary leukemia cells is associated with stem cell features and correlates with adverse clinical outcomes. Exp Hematol, 37, 1423-34. RATHE, S., MORIARITY, B., STOLTENBERG, C., KURATA, M., AUMANN, N., RAHRMANN, E., BAILEY, N., MELROSE, E., BAECKMANN, D., LISKA, C. & LARGAESPADA, D. 2014. Using RNA-seq and targeted nucleases to identify mechanisms of drug resistance in acute myeloid leukaemia. Sci Rep, 4, 1-9. RAVINDRANATH, Y., CHANG, M., STEUBER, C., BECTON, D., DAHL, G., CIVIN, C., CAMITTA, B., CARROLL, A., RAIMONDI, S. & WEINSTEIN, H. 2005. Pediatric Oncology Group (POG) studies of acute myeloid leukemia (AML): a review of four consecutive childhood AML trials conducted between 1981 and 2000. Leukemia, 19, 2101-2116.

259

REA, S., EISENHABER, F., O'CARROLL, D., STRAHL, B., SUN, Z., SCHMID, M., OPRAVIL, S., MECHTLER, K., PONTING, C., ALLIS, C. & JENUWEIN, T. 2000. Regulation of chromatin structure by site-specific histone H3 methyltransferases. Nature, 406, 593-9. REDDY, M., ZHANG, E. & NATARAJAN, R. 2015. Epigenetic mechanisms in diabetic complications and metabolic memory. Diabetologia, 58, 443-55. REES, J., GRAY, R., SWIRSKY, D. & HAYHOE, F. 1986. Principal results of the medical research council‘s 8th actue myeloid leukaemia trial. Lancet, 2, 1236- 41. REHLI, M., NILLER, H., AMMON, C., LANGMANN, S., SCHQARZFISCHER, L., ANDREESEN, R. & KRAUSE, S. 2003. Transcriptional Regulation of CHI3L1, a Marker Gene for Late Stages of Macrophage Differentiation. J Biol Chem, 2778, 44058-67. REYA, T. & CLEVERS, H. 2005. Wnt signalling in stem cells and cancer. Nature, 434, 843-50. REYA, T., MORRISON, S., CLARKE, M. & WEISSMAN, I. 2001. Stem cells, cancer, and cancer stem cells. Nature, 414, 105-111. RICHTER, G. H., PLEHM, S., FASAN, A., ROSSLER, S., UNLAND, R., BENNANI- BAITI, I. M., HOTFILDER, M., LOWEL, D., VON LUETTICHAU, I., MOSSBRUGGER, I., QUINTANALLI-MARTINEZ, L., KOVAR, H., STAEGE, M., MULLER-TIDOW, C. & BURDACH, S. 2009. EZH2 is a mediator of EWS/FLI1 driven tumor growth and metastasis blocking endothelial and neuro-ectodermal differentiation. Proc Natl Acad Sci USA, 106, 5324-5329. ROADMAP EPIGENOMICS CONSORTIUM, KUNDAJE, A., MEULEMAN, W., ERNST, J., BILENKY, M., YEN, A., HERAVI-MOUSSAVI, A., KHERADPOUR, P., ZHANG, Z., WANG, J., ZILLER, M., AMIN, V., WHITAKER, J., SCHULTZ, M., WAR, D. L., SARKAR, A., QUON, G., SANDSTROM, R., EATON, M., WU, Y., PFENNING, A., WANG, X., CLAUSSNITZER, M., LIU, Y., COARFA, C., HARRIS, R., SHORESH, N., EPSTEIN, C., GJONESKA, E., LEUNG, D., XIE, W., HAWKINS, R., LISTER, R., HONG, C., GASCARD, P., MUNGALL, A., MOORE, R., CHIAH, E., TAM, A., CANFIELD, T., HANSEN, R., KAUL, R., SABO, P., BANSAL, M., CARLES, A., DIXON, J., FARH, K., FEIZI, S., KARLIC, R.,

260

KIM, A., KULKARNI, A., LI, D., LOWDON, R., ELLIOT, G., MERCER, T., NEPH, S., ONUCHIC, V., POLAK, P., RAJAGOPAL, N., RAY, P., SALLARI, R., SIEBENTHALL, K., SINNOTT-ARMSTRONG, N., STEVENS, M., THURMAN, R., WU, J., ZHANG, B., ZHOU, X., BEAUDET, A., BOYER, L., DE JAGER, P., FARNHAM, P., FISHER, S., HAUSSLER, D., JONES, S., LI, W., MARRA, M., MCMANUS, M., SUNYAEV, S., THOMSON, J., TISTY, T., TSAI, L., WANG, W., WATERLAND, R., ZHANG, M., CHADWICK, L., BERNSTEIN, B., COSTELLO, J., ECKER, J., HIRST, M., MEISSNER, A., MILOSAVLJEVIC, A., REN, B., STAMATOYANNOPOULOS, J., WANG, T. & KELLIS, M. 2015. Integrative analysis of 111 reference human epigenomes. Nature, 5188, 317-330. ROBOZ, G. 2011. Novel approaches to the treatment of acute myeloid leukemia. ASH Educ Program Book, 2011, 43-50. ROSATI, R., LA STARZA, R., VERONESE, A., AVENTIN, A., SCHWIENBACHER, C., VALLESPI, T., NEGRINI, M., MARTELLI, M. F. & MECUCCI, C. 2002. NUP98 is fused to the NSD3 gene in acute myeloid leukemia associated with t(8;11)(p11.2;p15). Blood, 99, 3857-3860. ROSE, N., WOON, E., KINGHAM, G., KING, O., MECINOVIC, J., CLIFTON, I., NG, S., TALIB-HARDY, J., OPPERMANN, U., MCDONOUGH, M. A. & SCHOFIELD, C. J. 2010. Selective inhibitors of the JMJD2 histone demethylases: combined nondenaturing mass spectrometric screening and crystallographic approaches. J Med Chem, 53, 1810-8. ROSE, N. R., NG, S. S., MECINOVIC, J., LIENARD, B. M., BELLO, S. H., SUN, Z., MCDONOUGH, M. A., OPPERMANN, U. & SCHOFIELD, C. J. 2008. Inhibitor scaffolds for 2-oxoglutarate-dependent histone lysine demethylases. J Med Chem, 51, 7053-7056. ROSS, M., MAHFOUZ, R., ONCIO, M., LIU, H., ZHOU, X., SONG, G., SSHURTLEFF, S., POUNDS, S., CHENG, C., MA, J., RIBEIRO, R., RUBNITZ, J., GIRTMAN, K., WILLIAMS, W., RAIMONDI, S., LIANG, D., SHIH, L., PUI, C. & DOWNING, J. 2014. Gene expression profiling of pediatric acute myelogenous leukemia. Blood, 104, 3679-3687.

261

ROSSI, D., BRYDER, D., ZAHN, J., AHLENIUS, H., SONU, R., WAGERS, A. & WEISSMAN, I. 2005. Cell intrinsic alterations underlie hematopoietic stem cell aging. Proc Natl Acad Sci USA, 102, 9194-9199. ROTH, S. & ALLIS, C. 1996. Histone acetylation and chromatin assembly: a single escort, multiple dances? Cell, 87, 5-8. RUBNITZ, J., GIBSON, B. & SMITH, F. 2010a. Acute myeloid leukaemia. Hematol Oncol Clin N Am, 24, 35-63. RUBNITZ, J., INABA, H., DAHL, G., RIBEIRO, R., BOWMAN, W., TAUB, J., POUNDS, S., RAZZOUK, B., LACAYO, N., CAO, X., MESHINCHI, S., DEGAR, B., AIREWELE, G., RAIMONDI, S., ONCIU, M., COUSTAN- SMITH, E., DOWNING, J., LEUNG, W., PUI, C. & CAMPANA, D. 2010b. Minimal residual disease-directed therapy for childhood acute myeloid leukaemia:results fo the AML02 multicentre trial. Lancet Oncol, 11, 543-52. SAEZ, M., FERNANDEZ-RODRIGUEZ, J., MOUTINHO, C., SANCHEZ-MUT, J., GOMEZ, A., VIDAL, E., PETAZZI, P., SZCZESNA, K., LOPEZ-SERRA, P., LUCARIELLO, M., LORDEN, O., DELGADO-MORALES, R., DE LA CARIDAD, O., HUERTAS, D., GELPI, J., OROZCO, M., LOPEZ-DORIGA, A., MILA, M., PEREZ-JURADO, L., PINEDA, M., ARMSTRONG, J., LAZARO, C. & ESTELLER, M. 2015. Mutations in JMJD1c are involved in Rett syndrome and intellectual disability. Genet Med. SAITO, Y., KITAMURA, H., HIJIKATA, A., TOMIZAWA-MURASAWA, M., TANAKA, S., TAKAGI, S., UCHIDA, N., SUZUKI, N., SONE, A., NAJIMA, Y., OZAWA, H., WAKE, A., TANIGUCHI, S., SHULTZ, L., OHARA, O. & ISHIKAWA, F. 2010. Identification of therapeutic targets for quiescent, chemotherapy-resistant human leukemia stem cells. Sci Transl Med, 2, 17ra9. SAN-MIGUEL, J., HUNGRIA, V., YOON, S., BEKSAC, M., DIMOPOULOS, M., ELGHANDOUR, A., JEDRZEJCZK, W., GUNTHER, A., NAKORN, T., SIRITANARATKUL, N., CORRADINI, P., CHUNCHARUNEE, S., LEE, J., SCHOLSSMAN, R., SHELEKHOVA, T., YONG, K., TAN, D., NUMBENJAPON, T., CAVENAGH, J., HHOU, J., LEBLANC, R., NAHI, H., QIU, L., SALWENDER, H., PULINI, S., MOREAU, P., WARZOCHA, K., WHITE, D., BLADE, J., CHEN, W., DE LA RUBIA, J., GIMSLING, P., LONIAL, S., KAUFMAN, J., EINSELE, H., SOPALA, M., CORRADO, C.,

262

RBENGOUDIFA, B., BINLICH, F. & RICHARDSON, P. 2014. Panobinostat plus bortezomib and dexamethasone versus placebo plus bortezomib and dexamethasone in patients with relapsed or relapsed and refractory multiple myeloma: a multicentre,randomised, double blind phase 3 trial. Lancet Oncol, 15, 1195-206. SANSOM, O., REED, K., VAN DE WETERING, M., MUNCAN, V., WINTON, D., CLEVERS, H. & CLARKE, A. 2005. Cyclin D1 is not an immediate target of beta-catenin following Apc loss in the intestine. J Biol Chem, 280, 28463-28467. SARMA, K. & REINBERG, D. 2005. Histone variants meet their match. Nat Rev Mol Cell Biol, 6, 139-149. SARTORIUS, U. & KRAMMER, P. 2002. Upregulation of Bcl-2 is involved in the mediation of chemotherapy resistance in human small cell lung cancer cell lines. Int J Cancer, 97, 584-592. SCHENK, T., CHEN, W., GOLLNER, S., HOWELL, L., JIN, L., HEBESTREIT, K., KLEIN, H., POPESCU, A., BURNETT, A., MILLS, K., CASERO, R. J., MARTON, L., WOSTER, P., MINDEN, M., DUGAS, M., WANG, J., DICK, J., MULLER-TIDOW, C., PETRIE, K. & ZELENT, A. 2012. Inhibition of the LSD1 (KDM1A) demethylase reactivates the all-trans-retinoic acid differentiation pathway in acute myeloid leukemia. Nat Med, 18, 605-611. SCHISLER, J., WILLIS, M. & PATTERSON, C. 2008. You spin me round: MaFBx/Atrogin-1 feeds forward on FOXO transcription factors (like a record). Cell Cycle, 7, 440-443. SCHMIDT, D. & MCCAFFERTY, D. 2007. Trans-2-Phenylcyclopropylamine is a mechanism-based inactivator of the histone demethylase LSD1. Biochem, 46, 4408-4416. SCHOFIELD, C. & RATCLIFFE, P. 2004. Oxygen sensing by HIF hydroxylases. Nat Rev Mol Cell Biol, 5, 343-354. SCHOFIELD, C. J., MCDONOUGH, M. A., ROSE, N. R. & THALHAMMER, A. 2010. Histone lysine demethylase inhibitors. Patent No. WO 2010/043866 A2, GB. SCHOFIELD, R. 1978. The relationship between the spleen colony-forming cell and the haemopoietic stem cell. Blood Cells, 4, 7-24.

263

SCHOLL, C., BANSAL, D., DOHNER, K., EIWEN, K., HUNTLY, B., LEE, B., RUCKER, F., SCHLENK, R., BULLINGER, L., DOHNER, H., GILLILAND, G. & FROHLING, S. 2007. The homeobox gene CDX2 is aberrantly expressed in most cases of acute myeloid leukemia and promotes leukemogenesis. J Clin Invest, 117, 1037-1048. SCHULZE, A. & DOWNWARD, J. 2011. Flicking the Warburg Switch—Tyrosine Phosphorylation of Pyruvate Dehydrogenase Kinase Regulates Mitochondrial Activity in Cancer Cells. Mol Cell, 44, 846-848. SCORTEGAGNA, M., RULLER, C., FENG, Y., LAZOVA, R., KLUGER, H., JLI, J.- L., KDE, S., RICKERT, R., PELLECCHIA, M., BOSENBERG, M. & RONAI, Z. 2014. Genetic inactivation or pharmacological inhibition of Pdk1 delays development and inhibits metastasis of Braf(V600E)::Pten-/- melanoma. Oncogene, 33, 4330-4339. SECOMB, T., HSU, R., BRAUN, R., ROSS, J., GROSS, J. & DEWHIRST, M. 1998. Theoretical simulation of oxygen transport to tumors by three-dimensional networks of microvessels. Adv Exp Med Biol, 454, 629-634. SEITA, J. & WEISSMAN, I. 2010. Hematopoietic Stem Cell: Self-renewal versus Differentiation. Wiley Interdiscip Rev Syst Biol Med, 2, 640-653. SEKIRNIK, R., ROSE, N., THALHAMMER, A., SEDEN, P., MECINOVIC, J. & SCHOFIELD, C. 2009. Inhibition of the histone lysine demethylase JMJD2A by ejection of structural Zn(II). Chem Commun (Camb), 42, 6376-8. SEMENZA, G. 2007. HIF-1 mediates the Warburg effect in clear cell renal carcinoma. J Bioenerg Biomembr, 39, 231-234. SEMENZA, G. L. 2010a. Defining the role of hypoxia‑ inducible factor 1 in cancer biology and therapeutics. Oncogene, 29, 625-634. SEMENZA, G. L. 2010b. HIF-1: upstream and downstream of cancer metabolism. Curr Opin Genet Dev, 20, 51-56. SENESE, S., ZARAGOZA, K., MINARDI, S., MURADORE, I., RONZONI, S., PASSAFARO, A., BERNARD, L., DRAETTA, G., ALCALAY, M., SEISER, C. & CHIOCCA, S. 2007. Role for histone deacetylase 1 in human tumor cell proliferation. Mol Cell Biol, 27, 4784-95. SENGOKU, T. & YOKOYAMA, S. 2011. Structural basis for histone H3 Lys 27 demethylation by UTX/KDM6A. Genes Dev, 25, 2266-2277.

264

SENGUPTA, A., BANERJEE, D., CHANDRA, S., BANERJI, S., GHOSH, R., ROY, R. & BANERJEE, S. 2007. Deregulation and cross talk among Sonic hedgehog, Wnt, Hox and Notch signaling in chronic myeloid leukemia progression. Leukemia, 21, 949-55. SERKOVA, N. & BOROS , L. 2005. Detection of resistance to imatinib by metabolic profiling: clinical and drug development implications. Am J Pharmacogenomics, 5, 293-302. SERRANO, M., HANNON, G. J. & BEACH, D. 1993. A New Regulatory Motif in Cell-Cycle Control Causing Specific-Inhibition of Cyclin-D/Cdk4. Nature, 366, 704-707. SHAH, A., XIA, L., GOLDBERG, H., LEE, K., QUAGGIN, S. & FANTUS, I. 2013. Thioredoxin-interacting protein mediates high glucose-induced reactive oxygen species generation by mitochondria and the NADPH oxidase, Nox4, in mesangial cells. J Biol Chem, 288, 6835-48. SHAKYA, A., CALLISTER, C., GOREN, A., YOSEF, N., GARG, N., KHODDAMI, V., NIX, D., REGEV, A. & TANTIN, D. 2015. Pluripotency Transcription Factor Oct4 Mediates Stepwise Nucleosome Demethylation and Depletion. Mol Cell Biol, 35, 1014-1025. SHALEM, O., SANJANA, N., HARTENIAN, E., SHI, X., SCOTT, D., MIKKELSEN, T., HECKL, D., EBERT, B., ROOT, D., DOENCH, J. & ZHANG, F. 2014. Genome-scale CRISPR-Cas9 knockout screening in human cells. Science, 343, 84-7. SHARMA, S., KELLY, T. & JONES, P. 2010a. Epigenetics in cancer. Carcinogenesis, 31, 27-36. SHARMA, S., LEE, D., LI, B., QUINLAN, M., TAKAHASHI, F., MAHESWARAN, S., MCDERMOTT, U., AZIZIAN, N., ZOU, L., FISCHBACH, M., WONG, K., BRANDSTETTER, K., WITTNER, B., RAMASWAMY, S., CLASSON, M. & SETTLEMAN, J. 2010b. A chromatin-mediated reversible drug-tolerant state in cancer cell subpopulations. Cell, 141, 69-80. SHEN, L., KONDO, Y., GUO, Y., ZHANG, J., ZHANG, L., AHMED, S., SHU, J., CHEN, X., WATERLAND, R. & ISSA, J. 2007. Genome-wide profiling of DNA methylation reveals a class of normally methylated CpG island promoters. PloS Genet, 3, 2023-36.

265

SHI, J., WANG, E., ZUBER, J., RAPPAPORT, A., TAYLOR, M., JOHNS, C., LOWE, S. & VAKOC, C. 2013. The Polycomb complex PRC2 supports aberrant self- renewal in a mouse model of MLL-AF9; Nras(G12D) acute myeloid leukemia. Oncogene, 32, 930-938. SHI, Y., LAN, F., MATSON, C., MULLIGAN, P., WHETSTINE, J. R., COLE, P. A. & CASERO, R. A. 2004. Histone demethylation mediated by the nuclear amine oxidase homolog LSD1. Cell, 119, 941–953. SHI, Y. & WHETSTINE, J. 2007. Dynamic regulation of histone lysine methylation by demethylases. Mol Cell, 25, 1-14. SHIH, A. H., ABDEL-WAHAB, O., PATEL, J. P. & LEVINE, R. L. 2012. The role of mutations in epigenetic regulators in myeloid malignancies. Nat Rev Cancer, 12, 599-612. SHILATIFARD, A. 2006. Chromatin modifications by methylation and ubiquitination: implications in the regulation of gene expression. Annu Rev Biochem, 75, 243- 69. SHILATIFARD, A., LANE, W. S., JACKSON, K. W., CONAWAY, R. C. & CONAWAY, J. W. 1996. An RNA polymerase II elongation factor encoded by the human ELL gene. Science, 271, 1873-6. SHILOH, M., RUAN, J. & NATHAN, C. 1997. Evaluation of bacterial survival and phagocyte function with a fluorescence-based microplate assay. Infect Immun, 65, 3193-3198. SHLUSH, L. I., ZANDI, S., MITCHELL, A., CHEN, W. C., BRANDWEIN, J. M., GUPTA, V., KENNEDY, J. A., SCHIMMER, A. D., SCHUH, A. C., YEE, K. W., MCLEOD, J. L., DOEDENS, M., MEDEIROS, J. J., MARKE, R., KIM, H. J., LEE, K., MCPHERSON, J. D., HUDSON, T. J., BROWN, A. M., YOUSIF, F., TRINH, Q. M., STEIN, L. D., MINDEN, M. D., WANG, J. C. & DICK, J. E. 2014. Identification of pre-leukaemic haematopoietic stem cells in acute leukaemia. Nature, 506, 328-33. SILVERMAN, L., DEMAKOS, E., PETERSON, B., KORNBLITH, A., HOLLAND, J., ODCHIMAR-REISSIG, R., STONE, R., NELSON, D., POWELL, B., DECASTRO, C., ELLERTON, J., LARSON, R., SCHIFFER, C. & HOLLAND, J. 2002. Randomized controlled trial of azacitidine in patients with the

266

myelodysplastic syndrome: a study of the cancer and leukemia group B. J Clin Oncol, 20. SINCLAIR, D., RUDDELL, A., BROCK, J., CLEGG, N., LLOYD, V. & GRIGLIATTI, T. 1992. A cytogenetic and genetic characterization of a group of closely linked second chromosome mutations that suppress position-effect variegation in Drosophila melanogaster. Genetics, 130, 333-344. SINGBRANT, S., ASKMYR, M., PURTON, L. & WALKLEY, C. 2011. Defining the hematopoietic stem cell niche: the chicken and the egg conundrum. J Cell Biochem, 112, 1486-90. SINGH, S., CLARKE, I., TERASAKI, M., BONN, V., HAWKINS, C., SQUIRE, J. & DIRKS, P. 2003. Identification of a cancer stem cell in human brain tumors. Cancer Res, 63, 5821-8. SLANY, R. K. 2005. When epigenetics kills: MLL fusion proteins in leukemia. Hematol Oncol, 23, 1-9. SMITH, M., SEIBEL, N., ALTEKRUSE, S., RIES, L., MELBERT, D., O'LEARY, M., SMITH, F. & REAMAN, G. 2010. Outcomes for children and adolescents with cancer: challenges for the twenty-first century. J Clin Oncol, 28, 2625-34. SOLTANIAN, S. & MATIN, M. 2011. Cancer stem cells and cancer therapy. Tumor Biol, 32, 425-440. SOMERVAILLE, T. & CLEARY, M. 2006. Identification and characterization of leukemia stem cells in murine MLL-AF9 acute myeloid leukemia. Cancer Cell, 10, 257-268. SOMERVAILLE, T., MATHENY, C. J., SPENCER, G. J., IWASAKI, M., RINN, J. L., WITTEN, D., CHANG, H., SHURTLEFF, S., DOWNING, J. & CLEARY, M. 2009. Hierarchical maintenance of MLL myeloid leukaemia stem cells employs a transcriptional program shared with embryonic rather than adult stem cells. Cell Stem Cell, 4, 129-40. SONG, K., LI, M., XU, X., XUAN, L., HUANG, G., SONG, X. & LIU, Q. 3025. HIF- 1α and GLUT1 gene expression is associated with chemoresistance of acute myeloid leukemia. Asian Pac J Cancer Prev, 15, 1823-9. SPANGRUDE, G., HEIMFELD, S. & WEISSMAN, I. 1988. Purification and characterisation of mouse hematopoietic stem cells. Science, 241, 58-62.

267

SPARMANN, A. & VAN LOHUIZEN, M. 2006. Polycomb silencers control cell fate, development and cancer. Nat Rev Cancer, 6, 846-856. SPIJKERS-HAGELSTEIN, J., SCHNEIDER, P., HULLEMAN, E., DE BOER, J., WILLIAMS, O., PIETERS, R. & STAM, R. 2012. Elevated S100A8/S100A9 expression causes glucocorticoid resistance in MLL-rearranged infant acute lymphoblastic leukemia. Leukemia, 26, 1255-1265. SROCZYNSKA, P., CRUICKSHANK, V., BUKOWSKI, J., MIYAGI, S., BAGGER, F., WALFRIDSSON, J., SCHUSTER, M., PORSE, B. & HELIN, K. 2014. shRNA screening identifies JMJD1c as being required for leukemia maintenance. Blood. STEC, I., WRIGHT, T. J., VAN OMMEN, G. J., DE BOER, P. A., VAN HAERINGEN, A., MOORMAN, A. F., ALTHERR, M. R. & DEN DUNNEN, J. T. 1998. WHSC1, a 90 kb SET Domain-Containing Gene, Expressed in Early Development and Homologous to a Drosophila Dysmorphy Gene Maps in the Wolf-Hirschhorn Syndrome Critical Region and is Fused to IgH in t(1;14) Multiple Myeloma. Hum Mol Genet, 7, 1078-1082. STEGELMANN, F., BULLINGER, L., GRIESSHAMMER, M., HOLZMANN, K., HABDANK, M., KUHN, S., MAILE, C., SCHAUR, S., DOHNER, H. & DOHNER, K. 2010. High-resolution single-nucleotide polymorphism array- profiling in myeloproliferative neoplasms identifies novel genomic aberrations. Haematologica, 95, 666-9. STEVENS, R., HANN, I., WHEATLEY, K. & GRAY, R. 1998. on behalf of the MRC Childhood Leukaemia working party. Marked improvements in outcome and chemotherapy alone in paediatric acute myeloid leukaemia: results of the united kingdom medical research council‘s 10th AML trial. Br J Haematol, 101, 130- 40. STIER, S., KO, Y., FORKERT, R., LUTZ, C., NEUHAUS, T., GRÜNEWALD, E., CHENG, T., DOMBKOWSKI, D., CALVI, L., RITTLING, S. & SCADDEN, D. 2005. Osteopontin is a hematopoietic stem cell niche component that negatively regulates stem cell pool size. J Exp Med, 201, 1781-91. STOISER, B., KNOBL, P., FONATSCH, C., HAAS, O., MITTERBAUER, G., WELTERMANN, A., GEISSLER, K., VALENT, P., SPERR, W., PAGINBER,

268

I., LECHNER, K. & JAEGER, U. 2000. Prognosis of patients with a second relapse of acute myeloid leukemia. Leukemia, 14, 2059-63. STONE, R., O'DONNELL, M. & SEKERES, M. 2004. Acute myeloid leukaemia. Hematology Am Soc Hematol Educ Program, 98, 98-117. STRAHL, B. D. & ALLIS, C. D. 2000. The language of covalent histone modifications. Nature, 403, 41-45. STRAHL, B. D., GRANT, P. A., BRIGGS, S. D., SUN, Z., BONE, J., CALDWELL, J., MOLLAH, S., COOK, R., SHABANOWITZ, J., HUNT, D. & ALLIS, C. 2002. Set2 is a nucleosomal histone H3-selective methyltransferase that mediates transcriptional repression. Mol Cell Biol, 22, 1298–1306. STRONCEK, D., CARUCCIO, L. & BETTINOTTI, M. 2004. CD177: A member of the Ly-6 gene superfamily involved with neutrophil proliferation and polycythemia vera. J Transl Med, 2, 8. STRUMILO, S., SENKEVICH, S., SELEVICH, M. & VINOGRADOV, V. 1984. Inhibition of kinase activity of the adrenal pyruvate dehydrogenase complex by the mitochondrial component. Biokhimiia, 49, 155-9. STUBBS, M., KIM, Y., KRIVTSOV, A., WRIGHT, R., FENG, Z., AGARWAL, J., KUNG, A. & ARMSTRONG, S. 2008. MLL-AF9 and FLT3 cooperation in acute myelogenous leukemia: development of a model for rapid therapeutic assessment. Leukemia, 22, 66-77. STURGILL, E. & GUZMAN, M. 2013. Cytokine induced nuclear localisation of pyruvate kinase M2 in Acute Myeloid Leukemima. Blood, 122, 5406-5406. SU, Y., XU, H., XU, Y., YU, J., XIAN, Y. & LUO, Q. 2012. Azacytidine inhibits the proliferation of human promyelocytic leukemia cells (HL60) by demethylation of MGMT, DAPK and p16 genes. Hematology, 17, 41-6. SUAREZ, C., ANDREEFF, M. & MILLER, D. 1984. Serum LDH values in childhood acute leukemias and non-Hodgkin's lymphoma. Med Pediatr Oncol, 12, 89-92. SUN, S., LIANG, X., ZHANG, X., LIU, T., SHI, Q., SONG, Y., JIANG, Y., WU, H., JIANG, Y., LU, X. & PANG, D. 2015. Phosphoglycerate kinase-1 is a predictor of poor survival and a novel prognostic biomarker of chemoresistance to paclitaxel treatment in breast cancer. Br J Cancer, 112, 1332-9. SUZUKI, M. & BIRD, A. 2008. DNA methylation landscapes: provocative insights from epigenomics. Nat Rev Genet, 9, 465-76.

269

SUZUKI, T., MINEHATA, K., AKAGI, K., JENKINS, N. & COPELAND, N. 2006. Tumor suppressor gene identification using retroviral insertional mutagenesis in Blm-deficient mice. EMBO Journal, 25, 3422-3431. SUZUKI, T. & MIYATA, N. 2011. Lysine demethylases inhibitors. J Med Chem, 54, 8236-8250. SWANTON, C. & BECK, S. 2014. Epigenetic noise fuels cancer evolution. Cancer Cell, 26, 775-6. SZETO, W., JIANG, W., TICE, D., RUBINFELD, B., HOLLINGSHEAD, P., FONG, S., DUGGER, D., PHAM, T., YANSURA, D., WONG, T., GRIMALDI, J., CORPUZ, R., SINGH, J., FRANTZ, G., DEVAUS, B., CROWLEY , C., SCHWALL, R., EBERHARD, D., RASTELLI, L., POLAKIS, P. & PENNICA, D. 2001. Overexpression of the retinoic acid-responsive gene Stra6 in human cancers and its synergistic induction by Wnt-1 and retinoic acid. Cancer Res, 61, 4197-4205. TAHILIANI, M., KOH, K., SHEN, Y., PASTOR, W., BANDUKWALA, H., BRUDNO, Y., AGARWAL, S., IYER, L., LIU, D., ARAVIND, L. & RAO, A. 2009. Conversion of 5-methylcytosine to 5-hydroxymethylcytosine in mammalian DNA by MLL partner TET1. Science, 324, 930-5. TAI, M., CHANG, C., OLSON, L. & TOSKO, J. 2005. Oct-4 expression in adult human stem cells: evidence in support of the stem cell theory of carcinogenesis. Carcinogenesis, 26, 495-502. TAKAHASHI, Y., TAKAHASHI, S., SHIGA, Y., YOSHIMI, T. & MIURA, T. 2000. Hypoxic induction of prolyl 4-hydroxylas a (1) in cultured cells. J Biol Chem, 275, 14139-46. TAKESHITA, F., MINAKUCHI, Y., NAGAHARA, S., HONMA, K., SASAKI, H., HIRAI, K., TERATANI, T., NAMATAME, N., YAMAMOTO, Y., HANAI, K., KATO, T., SANO, A. & OCHIYA, T. 2005. Efficient delivery of small interfering RNA to bone-metastatic tumors by using atelocollagen in vivo. Proc Natl Acad Sci USA, 102, 12177-12182. TAN, J., JONES, M., KOSEKI, H., NAKAYAMA, M., MUNTEAN, A., MAILLARD, I. & HESS, J. 2011. CBX8, a polycomb group protein, is essential for MLLAF9- induced leukemogenesis. Cancer Cell, 20, 563-575.

270

TAN, J., YANG, X., ZHUANG, L., JIANG, X., CHEN, W., LEE, P., KARUTURI, R., TAN, P., LIU, E. & YU, Q. 2007. Pharmacologic disruption of Polycomb- repressive complex 2-mediated gene repression selectively induces apoptosis in cancer cells. Genes Dev, 21, 1050-63. TANAKA, S., MIYAGI, S., SASHIDA, G., CHIBA, T., YUAN, J., MOCHIZUKI- KASHIO, M., SUZUKI, Y., SUGANO, S., NAKASEKO, C., YOKOTE, K., KOSEKI, H. & IWAMA, A. 2012. Ezh2 augments leukemogenicity by reinforcing differentiation blockage in acute myeloid leukemia. Blood, 120, 1107-17. TANG, Y., LUO, J., ZHANG, W. & GU, W. 2006. Tip60-dependent acetylation of p53 modulates the decision between cell-cycle arrest and apoptosis. Mol Cell, 24, 827-39. TANTIN, D. 2013. Oct transcription factors in development and stem cells: insights and mechanisms. Development, 140, 2857–2866. TAO, W., PENNICA, D., XU, L., KALEJTA, R. & LEVINE, A. 2001. Wrch-1, a novel member of the Rho gene family that is regulated by Wnt-1. Genes Dev, 15, 1796-1807. TAYLOR, S., HICKMAN, J. & DIVE, C. 2000. Epigenetic Determinants of Resistance to Etoposide Regulation of Bcl-xL and Bax by Tumor Microenvironmental Factors. J Natl Cancer Inst, 92, 18-23. TERPSTRA, W., PRINS, A., PLOEMACHER, R., WOGNUM, B., WAGEMAKER, G., LOWENBERG, B. & WIELENGA, J. 1996. Long-term leukemia-initiating capacity of a CD34-subpopulation of acute myeloid leukemia. Blood, 87, 2187- 2194. THALHAMMER, A., MECINOVIC, J., LOENARZ, C., TUMBER, A., ROSE, N., HEIGHTMAN, T. & SCHOFIELD, C. J. 2011. Inhibition of the histone demethylase JMJD2E by 3-substituted pyridine 2,4-dicarboxylates. Org Biomol Chem, 9, 127-35. THOL, F., DAMM, F., LUDEKING, A., WINSCHEL, C., WAGNER, K., MORGAN, M., YUN, H., GOHRING, G., SCHLEGELBERGER, B., HOELZER, D., LUBBERT, M., KANZ, L., FIEDLER, W., KIRCHNER, H., HEIL, G., KRAUTER, J., GANSER, A. & HEUSER, M. 2011. Incidence and prognostic

271

influence of DNMT3A mutations in acute myeloid leukemia. J Clin Oncol, 29, 2889-96. THORSTEINSDOTTIR, U., KROON, E., JEROME, L., BLASI, F. & SAUVAGEAU, G. 2001. Defining roles for HOX and MEIS1 genes in induction of acute myeloid leukemia. Mol Cell Biol, 21, 224-234. TREWICK, S., MCLAUGHLIN, P. & ALLSHIRE, R. 2005. Methylation: Lost in hydroxylation? EMBO J, 6, 315-320. UEDA, K., YOSHIMI, A., KAGOYA, Y., NISHIKAWA, S., MARQUEZ, V., NAKAGAWA, M. & KUROKAWA, M. 2014. Inhibition of histone methyltransferase EZH2 depletes leukemia stem cell of mixed lineage leukemia fusion leukemia through upregulation of p16. Cancer Sci, 105, 512-519. ULLAH, M., DAVIES, A. & HALESTRAP, A. 2006. The plasma membrane lactate transporter, MCT4, but not MCT1, is upregulated by hypoxia through a HIF- 1alpha-dependent mechanism. J Biol Chem, 2006, 14. VALENCIA, A., ROMAN-GOMEZ, J., CERVERA, J., SUCH, E., BARRAGAN, E., BOLUFER, P., MOSCARDO, F., SANZ, G. & SANZ, M. 2009. Wng signaling pathway is epigenetically regulated by methylation of Wnt antagonists in acute myeloid leukemia. Leukemia, 23, 1658-1666. VALINLUCK, V., TSAI, H., ROGSTAD, D., BURDZY, A., BIRD, A. & SOWERS, L. 2004. Oxidaive damage to methyl-CpG sequences inhibits the binding of the methyl-CpG binding domain (MBD) of methyl-CpG binding protein 2 (MeCP2). Nucleic Acids Res, 32, 4100-8. VANDER HEIDEN, M. G., CANTLEY, L. C. & THOMPSON, C. B. 2009. Understanding the Warburg Effect: The Metabolic Requirements of Cell Proliferation. Science, 324, 1029–1033. VANHOUTTEE, F., DUPONT, S., VAN KAEM, T., GOUY, M., GHEYLE, L., BLANQUE, R., BRYS, R., N, V., HAAZEN, W., VAN KLOOSTER, G. & BEETENS, J. 2014. Human safety, pharmacokinetics and pharmacodynamics of the GPR84 antagonist GLPG1205, a potential new approach to treat IBD. United European Gastroenterology, Abstract #PO341. VARAMBALLY, S., CAO, Q., MANI, R. S., SHANKAR, S., WANG, X., ATEEQ, B., LAXMAN, B., CAO, X., JING, X., RAMNARAYANAN, K., BRENNER, J., YU, J., KIM, J., HAN, B., TAN, P., KUMAR-SINHA, C., LONIGRO, R.,

272

PALANISAMY, N., MAHER, C. & CHINNAIYAN, A. 2008. Genomic loss of microRNA-101 leads to overexpression of histone methyltransferase EZH2 in cancer. Science, 322, 1695-9. VARAMBALLY, S., DHANASEKARAN, S., ZHOU, M., BARRETTE, T., KUMAR- SINHA, C., SANDA, M., GHOSH, D., PIENTA, K., SEWALT, R., OTTE, A. P., RUBIN, M. & CHINNAIYAN, A. 2002. The polycomb group protein EZH2 is involved in progression of prostate cancer. Nature, 419, 624-629. VARATHARAJAN, S., ABRAHAM, A., ZHANG, W., SHAJI, R., AHMEND, R., GEORGE, B., SRIVASTAVA, A., CHANDY, M., MATHEWS, V. & BALASUBRAMANIAN, P. 2012. Carbonyl reductase 1 expression influences daunorubicin metabolism in acute myeloid leukemia. Eur J Clin Pharmacol, 68, 1577-86. VARDIMAN, J., THIELE, J., ARBER, D., BRUNNING, R., BOROWITZ, M., PORWIT, A., HARRIS, N., LE BEAU, M., HELLSTROM-LINDBERG, E., TEFFERI, A., BLOOMFIELD, C., PORWIT, A., HARRIS, N., LE BEAU, M., HELLSTROM-LINDBERG, E., TEFFERI, A. & BLOOMFIELD, C. 2009. The 2008 revision of the World Health Organization (WHO) classification of myeloidneoplasms and acute leukemia: rationale and important changes. Blood, 114, 937-51. VERHEUL, H., SALUMBIDES, B., VAN ERP, K., HAMMERS, H., QIAN, D., SANNI, T., ATADJA, P. & PILI, R. 2008. Combination strategy targeting the hypoxia inducible factor-1 α with mammalian target of rapamycin and histone deacetylase inhibitors. Clin Cancer Res, 14, 3589-97. VERMA, S., TIAN, X., LAFRANCE, L., DUQUENNE, C., SUAREZ, D., NEWLANDER, K., ROMERIL, S., BURGESS, J., GRANT, S., BRACKLEY, J., GRAVES, A., SCHERZER, D., SHU, A., THOMPSON, C. B., OTT, H., VAN ALLER, G., MACHUTTA, C., DIAZ, E., JIANG, Y., JOHNSON, N., KNIGHT, S., KRUGER, R., MCCABE, M., DHANAK, D., TUMMINO, P., CREASY, C. & MILLER, W. 2012. Identification of potent, selective, cell- active inhibitors of the histone lysine methyltransferase EZH2. ACS Med Chem Lett, 3, 1091-6. VERMES, I., HAANEN, C. & REUTELINGSPERGER, C. 2000. Flow cytometry of apoptotic cell death. J Immunol Methods, 243, 1-2.

273

VIRCHOW, R. 1881. An address on the value of pathological experiments. Br Med J, 2, 198-203. VORMOOR, J., BOOS, J., STAHNK, E., JURGENS, H., RITTER, J. & CREUTZIG, U. 1996. Therapy of childhood acute myelogenous leukaemias. Ann Hematol, 73, 11-24. WAGERS, A. & WEISSMAN, I. 2006. Differential expression of alpha2 integrin separates long-terma nd short-term reconstituting Lin-/loThy1.1(lo)c-kit+ Sca- 1+ hematopoietc stem cells. Stem Cells, 24, 1087-1094. WAGNER, E. J. & CARPENTER, P. B. 2012. Understanding the language of Lys36 methylation at histone H3. Nat Rev Mol Cell Biol, 13, 115-126. WAKITA, S., YAMAGUCHI, H., OMORI, I., TERADA, K., UEDA, T., MANABE, E., KUROSAWA, S., IIDA, S., IBARAKI, T., SATO, Y., TOKDOROKI, T., HIRAKAWA, T., RYOTOKUJI, T., ARAI, K., KITANO, T., MITAMURA, Y., KOSAKA, F., DAN, K. & INOKUCHI, K. 2012. Mutations of the epigenetics- modifying gene (DNMT3a, TET2, IDH1/2) at diagnosis may induce FLT3-ITD at relapse in de novo acute myeloid leukemia. Leukemia, 27, 1044-52. WALCZAK-DRZEWIECKA, A., RATAJEWSKI, M., WAGNER, W. & DASTYCH, J. 2008. HIF-1alpha is up-regulated in activated mast cells by a process that involves calcineurin and NFAT. J Immunol, 181, 1665-72. WALKLEY, C., OLSEN, G., DWORKIN, S., FABB, S., SWANN, J., MCARTHUR, G., WESTMORELAND, S., CHAMBON, P., SCADDEN, D. & PURTON, L. 2007a. A microenvironment-induced myeloproliferative syndrome caused by retinoic acid receptor gamma deficiency. Cell, 129, 1097-1110. WALKLEY, C., SHEA, J., SIMS, N., PURTON, L. & ORKIN, S. 2007b. Rb regulates interactions between hematopoietic stem cells and their bone marrow microenvironment. Cell, 129, 1081-1095. WANG, C., YUAN, C., WANG, S., ZHANG, H., HU, X., ZHANG, L., WU, C. & LI, N. 2007a. Differential gene expression of aldolase C (ALDOC) and hypoxic adaptation in chickens. Anim Genet, 38, 203-10. WANG, E., KAWAOKA, S., YU, M., SHI, J., NI, T., YANG, W., ZHU, J., ROEDER, R. & VAKOC, C. 2013. Histone H2B ubiquitin ligase RNF20 is required for MLLrearranged leukemia. Proc Natl Acad Sci USA, 110, 3901-3906.

274

WANG, G., ALLIS, C. & CHI, P. 2007b. Chromatin remodeling and cancer, Part 1: Covalent histone modifications. Trends Mol Med, 13, 363-72. WANG, G., CAI, L., PASILLAS, M. & KAMPS, M. 2007c. NUP98–NSD1 links H3K36 methylation to Hox-A gene activation and leukaemogenesis. Nat Cell Biol, 9, 804-12. WANG, J., PARK, J., DRISSI, H., WANG, X. & XU, R. 2014a. Epigenetic Regulation of miR-302 by JMJD1C Inhibits Neural Differentiation of Human Embryonic Stem Cells. J Biol Chem, 289, 2384-95. WANG, T., WEI, J., SABATINI, D. & LANDER, E. 2014b. Genetic screens in human cells using the CRISPR-Cas9 system. Science, 343, 80-4. WANG, X., KRUITHOF-DE JULIO, M., ECONOMIDES, K., WALKER, D., YU, H., ALILI, M., HU, Y., PRICE, S., ABATE-SHEN, C. & SHEN, M. 2009. A luminal epithelial stem cell that is a cell of origin for prostate cancer. Nature, 461, 495-500. WANG, Y., ISRAELSEN, W., LEE, D., YU, V., JEANSON, N., CLISH, C., CANTLEY, L., VANDER HEIDEN, M. & SCADDEN, D. 2014c. Cell-state- specific metabolic dependency in hematopoiesis and leukemogenesis. Cell, 158, 1309-23. WANG, Y., KRIVTSOV, A., SINHA, A., NORTH, T., GOESSLING, W., FENG, Z., ZON, L. & ARMSTRONG, S. 2010. The Wnt/b-catenin pathway is required for the development of leukemia stem cells in AML. Science, 327, 1650-1653. WANG, Y. & LEUNG, F. 2004. An evaluation of new criteria for CpG islands in the human genome as gene markers. Bioinformatics, 20, 1170-7. WANG, Y., LIU, Y., MALEK, S., ZHENG, P. & LIU, Y. 2011. Targeting HIF1a eliminates cancer stem cells in hematological malignances. Cell Stem Cell, 8, 399-411. WARBURG, O. 1956. On the origin of cancer cells. Science, 123, 309-314. WARBURG, O., POSENER, K. & NEGELEIN, E. 1924. Über den Stoffwechsel der Carcinomzelle. Biochem. Zeitschr, 152, 309–344. WARBURG, O., WIND, F. & NEGELEIN, E. 1927. The metabolism of tumors in the body. J Gen Physiol, 8, 519-530. WARD, P., PATEL, J., WISE, D., ABDEL-WAHAB, O., BENNETT, B., COLLER, H., CROSS, J., FANTIN, V., HEDVAT, C., PERL, A., RABINOWITZ, J.,

275

CARROLL, M., SU, S., SHARP, K., LEVINE, R. L. & THOMPSON, C. B. 2010. The common feature of leukemia-associated IDH1 and IDH2 mutations is a neomorphic enzyme activity converting a-ketoglutarate to 2-hydroxyglutarate. Cancer Cell, 17, 225-234. WARNER, J., WANG, J., HOPE, K., JIN, L. & DICK, J. 2004. Concepts of human leukemic development. Oncogene, 23, 7164-7177. WATANABE, S., WATANABE, K., AKIMOV, V., BARTKOVA, J., BLAGOEV, B., LUKAS, J. & BARTEK, J. 2013. JMJD1C demethylates MDC1 to regulate the RNF8 and BRCA1–mediated chromatin response to DNA breaks. Nat Struct Mol Biol, 20, 1425-1433. WEBBY, C., WOLF, A., GROMAK, N., DREGER, M., KRRAMER, H., KESSLER, B., NIELSEN, M., SCHMITZ, C., BUTLER, D., YATES, J., DELAHUNTY, C., HAHN, P., LENGELING, A., MANN, M., PROUDFOOT, N., SCHOFIELD, C. J. & BOTTGER, A. 2009. Jmjd6 catalyses lysyl-hydroxylation of U2AF65, a protein associated with RNA splicing. Science, 325, 90-93. WEI, J., WUNDERLICH, M., FOX, C., ALVAREZ, S., CIGUDOSA, J., WILHELM, J., ZHENG, Y., CANCELAS, J., GU, Y., JANSEN, M., MDIMARTINO, J. & MULLOY, J. 2008. Microenvironment determines lineage fate in a human model of MLL-AF9 leukemia. Cancer Cell, 13, 483-495. WEICHERT, W., ROSKE, A., NIESPOREK, S., NOSKE, A., BUCKENDAHL, A., DIETEL, M., GEKELER, V., BOEHM, M., BECKERS, T. & DENKERT, C. 2008. Class I histone deacetylase expression has independent prognostic impact in human colorectal cancer: specific role of class I histone deacetylases in vitro and in vivo. Clin Cancer Res, 14, 1669-77. WEINDRUCH, R., KEENAN, K., CARNEY, J., FERNANDES, G., FEUERS, R., FLOYD, R., HALTER, J., RAMSEY, J., RICHARDSON, A., ROTH, G. & SPINDLER, S. 2001. Caloric restriction mimetics: metabolic interventions. J Gerontol A Biol Sci Med Sci, 56, 20-33. WEISS, R. 1992. The anthracyclines: will we ever find a better doxorubicin? Semin Oncol, 19, 670-86. WEISSMAN, I. 2000. Translating stem and progenitor cell biology to the clinic: barriers and opportunities. Science, 287, 1442-1446.

276

WELLMANN, S., GUSCHMANN, M., GRIETHE, W., ECKERT, C., VON STACKELBERG, A., LOTTAZ, C., MODEREGGER, E., EINSIEDEL, H., ECKARDT, K., HENZE, G. & SEEGER, K. 2004. Activation of the HIF pathway in childhood ALL, prognostic implications of VEGF. Leukemia, 18, 926-933. WELLS, R., WOODS, W., LAMPKIN, B., NESBIT, M., LEE, J., BUCKLEY, J., VERSTEEG, C. & HAMMOND, G. 1993. Impact of high-dose cytarabine and asparaginase intensification on childhood acute myeloid leukaemia: a report from the Childrens Cancer Group. J Clin Oncol, 11, 538-45. WILKINSON, A. C., BALLABIO, E., GENG, H., NORTH, P., TAPIA, M., KERRY, J., BISWAS, D., ROEDER, R., ALLIS, C., MELLNICK, A., DE BRUJIN, M. & MILNE, T. 2013. RUNX1 is a key target in t(4;11) leukemias that contributes to gene activation through an AF4-MLL complex interaction. Cell Rep, 3, 116- 27. WILSON, B. G., WANG, X., SHEN, X., MCKENNA, E. S., LEMIEUX, M. E., CHO, Y. J., KOELLHOFFER, E. C., POMEROY, S. L., ORKIN, S. H. & ROBERTS, C. W. 2010. Epigenetic antagonism between polycomb and SWI/SNF complexes during oncogenic transformation. Cancer Cell, 18, 316-28. WIMAZAL, F., SPERR, W., KUNDI, M., VALES, A., FONATSCH, C., THALHAMMER-SCHERRER, R., SCHWARZINGER, I. & VALENT, P. 2007. Prognostic significance of serial determinations of lactate dehydrogenase (LDH) in the follow-up of patients with myelodysplastic syndromes. Ann Oncol, 19, 970-6. WISSMANN, M., YIN, N., MULLER, J. M., GRESCHIK, H., FODOR, B. D., JENUWEIN, T., VOGLER, C., SCHNEIDER, R., GUNTHER, T., BUETTNER, R., METZGER, E. & SCHULE, R. 2007. Cooperative demethylation by JMJD2C and LSD1 promotes androgen receptor-dependent gene expression. Nat Cell Biol, 9, 347-353. WOLF, S., PATCHEV, V. & OBENDORF, M. 2007. A novel variant of the putative demethylase gene, s-JMJD1c, is a coactivator of the AR. Arch Biochem Biophys, 460, 56-66. WONG, S., GOODE, D., IAWASAKI, M., WEI, M., KUO, H., SCHNEIDAWIND, D., DUQUE-AFONSO, J., WENG, Z. & CLEARY, M. 2015. The H3K4Methyl

277

Epigenome Regulates Leukemia Stem Cell Oncogenic Potential. Cancer Cell, 28, 198-209. WU, B., CRAMPTON, S. & CC, H. 2007. Wnt signalling induces matrix metalloproteinase expression and regulates T cell transmigration. Immunity, 26, 227-239. XIAO-NAN, Y. & ZHI-PENG, F. 2013. Sonic hedge hog signaling enhanced the expression of histone demethylase JMJD5 in the head and neck squamous cell carcinoma. Oral Oncol, S93. XIE, Z., CHONGLEI, B., CHEONG, L., LIU, S., GAOFENG, H., ZHOU, J., YU, Q., CHEN, C. & CHNG, W. 2011. Determinants of sensitivity to DZNep induced apoptosis in multiple myeloma cells. Plos One, 6, e21583. XU, B., ON, D., MA, A., PARTON, T., KONZE, K., PATTENDEN, S., ALLISON, D., CAI, L., ROCKOWITZ, S., LIU, S., LIU, Y., LI, F., VEDADI, M., FRYE, S., GARCIA, B. A., ZHENG, D., JIN, J. & WANG, G. 2015. Selective inhibition of EZH2 and EZH1 enzymatic activity by a small molecule suppresses MLL- rearranged leukemia. Blood, 125, 346-57. XU, L., ZHAO, Z., DONG, A., SOUBIGOU-TACANNAT, L., RENOU, J., STEINMETZ, A. & SHEN, W. 2008. Di- and tri- but not monomethylation on histone H3 lysine 36 marks active transcription of genes involved in flowering time regulation and other processes in Arabidopsis thaliana. Mol Cell Biol, 28, 1348-60. XU, R., PELICANO, H., ZHOU, Y., CAREW, J., FENG, L., BHALLA, K., KEATING, M. & HUANG, P. 2005. Inhibition of glycolysis in cancer cells: a novel strategy to overcome drug resistance associated with mitochondrial respiratory defect and hypoxia. Cancer Res, 65, 613-21. XU, W., PARMIGIANI, R. & MARKS, P. 2007. Histone deacetylase inhibitors: molecular mechanisms of action. Oncogene, 26, 5541-52. XU, W., YANG, H., LIU, Y., YANG, Y., WANG, P., KIM, S., ITO, S., YANG, C., WANG, P., XIAO, M., LIU, X., JIANG, W., LIU, J., ZHANG, J., WANG, B., FRYE, S., ZHANG, Q., GUAN, K., ZHAO, S. & XIONG, Y. 2011. Oncometabolite 2-hydroxyglutarate is a competitive inhibitor of a-ketoglutarate- dependent dioxygenases. Cancer Cell, 19, 17-30.

278

YAMADA, K. & MIYAMOTO, K. 2005. Basic helix-loop-helix transcription factors, BHLHB2 and BHLHB3; their gene expressions are regulated by multiple extracellular stimuli. Front Biosci, 2005. YAMAMOTO-SUGITANI, M., KURODA, J., ASHIHARA, E., NAGOSHI, H., KOBAYASHI, T., MATSUMOTO, Y., SASKI, N., SHIMURA, Y., KIYOTA, M., NAKAYAMA, R., AKAJI, K., TAKI, T., UOSHIMA, N., KOBAYASHI, Y., HORIIKE, S., MAEKAWA, T. & TANIWAKI, M. 2011. Galectin-3 (Gal-3) induced by leukemia microenvironment promotes drug resistance and bone marrow lodgement in chronic myelogenous leukemia. Proc Natl Acad Sci USA, 108, 17468-17473. YAMANE, K., TOUMAZOU, C., TSUKADA, Y., ERDJUMENT-BROMAGE, H., TEMPST, P., WONG, J. & ZHANG, Y. 2006. JHDM2A, a JmjC-containing H3K9 demethylase, facilitates transcription activation by androgen receptor. Cell, 125, 483-495. YAMASHITA, Y., YUAN, J., SUETAKE, I., SUZUKI, H., ISHIKAWA, Y., CHOI, Y. U., T, SODA, M., HAMADA, T., HARUTA, H., TAKADA, S., MIYAZAKI, Y., KIYOI, H., ITO, E., NAOE, T., TOMONAGA, M., TOYOTA, M., TAJIMA, S., IAWAMA, A. & MANO, H. 2010. Array-based genomic resequencing of human leukemia. Oncogene, 29, 3723-31. YAN-FANG, T., DONG, W., LI, P., WEN-LI, Z., JUN, L., NA, W., JIAN, W., XING, F., YAN-HONG, L., JIAN, N. & JIAN, P. 2012. Analyzing the gene expression profile of pediatric acute myeloid leukemia with real-time PCR arrays. Cancer Cell Int, 12, 40-51. YAN, X. J., XU, J., GU, Z. H., PAN, C. M., LU, G., SHEN, Y., SHI, J. Y., ZHU, Y. M., TANG, L., ZHANG, X. W., LIANG, W. X., MI, J. Q., SONG, H. D., LI, K. Q., CHEN, Z. & CHEN, S. J. 2011. Exome sequencing identifies somatic mutations of DNA methyltransferase gene DNMT3A in acute monocytic leukemia. Nat Genet, 43, 309-15. YANG, L., YANG, M., ZHANG, H., WANG, Z., YU, Y., XIE, M., ZHAO, M., LIU, L. & CAO, L. 2012. S100A8-targeting siRNA enhances arsenic trioxide-induced myeloid leukemia cell death by down-regulating autophagy. Int J Mol Med, 29, 65-72.

279

YANG, M., CHOWDHURY, R., GE, W., HAMED, R., MCDONOUGH, M. A., CLARIDGE, T., KESSLER, B., COCKMAN, M., RATCLIFFE, P. & SCHOFIELD, C. J. 2011a. Factor-inhibiting hypoxia-inducible factor (FIH) catalyses the post-translational hydroxylation of histidinyl residues within ankyrin repeat domains. FEBS J, 278, 1086-1097. YANG, M., GE, W., CHOWDHURY, R., CLARIDGE, T., KRAMER, H., SCHMIERER, B., MCDONOUGH, M. A., GONG, L., KESSLER, B., RATCLIFFE, P., COLEMAN, M. & SCHOFIELD, C. J. 2011b. Asparagine and aspartate hydroxylation of the cytoskeletal ankyrin family is catalyzed by factor- inhibiting hypoxia-inducible factor. J Biol Chem, 286, 7648-7660. YANG, X. 2004. The diverse superfamily of lysine acetyltransferases and their role in leukemia and other diseases. Nucleic Acids Res, 32, 959-76. YANG, X. & SETO, E. 2008. Lysine acetylation: codified crosstalk with other posttranslational modifications. Mol Cell, 31, 449-61. YANG, Z. Q., IMOTO, I., PIMKHAOKHAM, A., SHIMADA, Y., SASAKI, K., OKA, M. & INAZAWA, J. 2001. A novel amplicon at 9p23-24 in squamous cell carcinoma of the esophagus that lies proximal to GASC1 and harbors NFIB. Jpn J Cancer Res, 92, 423-8. YAO, H., ASHIHARA, E. & MAEKAWA, T. 2011. Targeting the Wnt/b-catenin signaling pathway in human cancers. Expert Opin Ther Targets 15, 873-87. YEUNG, J., ESPOSITO, M., GANDILLET, A., ZEISIG, B., GRIESSLINGER, E., BONNET, D. & SO, C. 2010. B-catenin mediates the establishment and drug resistance of MLL leukemic stem cells Cancer Cell, 18, 606-618. YOO, C. & JONES, P. 2006. Epigenetic therapy of cancer: past, present and future. Nat Rev Drug Discov, 5, 37-50. YOUN, M., YOKOYAMA, A., FUJIYAMA-NAKAMURA, S., OHTAKE, F., MINEHATA, K., YASUDA, H., SUZUKI, T., KATO, S. & IMAI, Y. 2012. JMJD5, a Jumonji C (JmjC) domain-containing protein, negatively regulates osteoclastogenesis by facilitating NFATc1 protein degradation. J Biol Chem, 287, 12994-13004. YSEBAERT, L., CHICANNE, G., DEMUR, C., DE TONI, F., PRADE-HOUDLLIER, N., RUIDAVETS, MANSAT-DE MAS, V., RIGAL-HUGUET, F., LAURENT, G., PAYRASTRE, B., ANENTI, S. & C, R.-S. 2006. Expression of b-catenin by

280

acute myelod leukemia cells predicts enhaned clonogenic capacities and poor prognosis. Leukemia, 20, 1211-1216. YU, W., CHORY, E., WERNIMONT, A., TEMPEL, W., SCOPTON, A., FEDERATION, A., MARINEAU, J., QI, J., BARSYTE-LOVEJOY, D., YI, J., MARCELLUS, R., IACOB, R., ENGEN, J., GRIFFIN, C., AMAN, A., WIENHOLDS, E., LI, F., PINEDA, J., ESTIU, G., SHATSEVA, T., HAJIAN, T., AL-AWAR, R., DICK, J., VEDADI, M., BROWN, P., ARROWSMITH, C., BRADNER, J. & SCHAPIRA, M. 2013. Catalytic site remodelling of the DOT1L methyltransferase by selective inhibitors. Nat Commun, 3, 1288. ZABKIEWICZ, J., PEARN, L., HILLS, R., MORGAN, R., TONKS, A., BURNETT, A. & DARLEY, R. 2014. The PDK1 master kinase is over-expressed in acute myeloid leukemia and promotes PKC-mediated survival of leukemic blasts. Haematologica, 99, 858-64. ZEISIG, B. B., KULASEKARARAJ, A. G., MUFTI, G. J. & SO, C. W. 2012. SnapShot: Acute myeloid leukemia. Cancer Cell, 22, 698-698 e1. ZEISIG, B. B., MILNE, T., GARCIA-CUELLAR, M. P., SCHREINER, S., MARTIN, M. E., FUCHS, U., BORKHARDT, A., CHANDA, S. K., WALKER, J., SODEN, R., HESS, J. L. & SLANY, R. K. 2004. Hoxa9 and Meis1 are key targets for MLL-ENL-mediated cellular immortalization. Mol Cell Biol, 24, 617- 28. ZHANG, H., LI, H., XI, H. & LI, S. 2012. HIF1a is required for survival maintenance of chronic myeloid leukemia stem cells. Blood, 119, 2595-607. ZHANG, R., HUANG, W., LI, Y., SONG, Y. & LI, Y. 2015. JMJD5 is a potential oncogene for colon carcinogenesis. Int J Clin Exp Pathol, 8, 6482-6489. ZHANG, W., XIA, X., REISENAUER, M. R., HEMENWAY, C. S. & KONE, B. C. 2006. Dot1a-AF9 complex mediates histone H3 Lys-79 hypermethylation and repression of ENaCalpha in an aldosterone-sensitive manner. J Biol Chem, 281, 18059-18068. ZHANG, Y. & REINBERG, D. 2001. Transcription regulation by histone methylation: interplay between different covalent modifications of the core histone tails. Genes Dev, 15, 2343-60.

281

ZHAO, C., BLUM, J., CHEN, A., KWON, H., JUNG, S., COOK, J., LAGOO, A. & REYA, T. 2007. Loss of b-catenin impairs the renewal of normal and CML stem cells in vivo. Cancer Cell, 12, 528-541. ZHAO, L., MOOS, M., GRABNER, R., PEDRONO, F., FAN, J., KAISER, B., JOHN, N., SCHMIDT, S., SPANBROEK, R., LOTZER, K., HUANG, L., CUI, J., RADAR, D., EVANS, J., HABENICHT, A. & FUNK, C. 2004. The 5- lipoxygenase pathway promotes pathogenesis of hyperlipidemia-dependent aortic aneurysm. Nat Med, 10, 966-973. ZHENG, T., XIE, N., BANERJEE, S., CUI, H., FU, M., THANNICKAL, V. & LIU, G. 2015. The monocarboxylate transporter 4 is required for glycolytic reprogramming and inflammatory response in macrophages. J Biol Chem, 290, 46-55. ZHOU, J., BI, C., CHEONG, L. L., MAHARA, S., LIU, S. C., TAY, K. G., KOH, T. L., YU, Q. & CHNG, W. J. 2011. The histone methyltransferase inhibitor, DZNep, up-regulates TXNIP, increases ROS production, and targets leukemia cells in AML. Blood, 118, 2830-9. ZHOU, J. & CHNG, W. 2013. Roles of thioredoxin binding protein (TXNIP) in oxidative stress, apoptosis and cancer. Mitochondrion, 13, 163-9. ZHOU, K., JIANG, L., SHANG, Z., WANG, J., HUANG, L. & ZHOU, J. 2012. Potential application of IDH1 and IDH2 mutations as prognostic indicators in non-promyelocytic acute myeloid leukemia: A meta-analysis. Leuk Lymphoma, 53, 2423-9. ZHOU, S., SCHUETZ, J., BUNTING, K., COLAPIETRO, A., SAMPATH, J., MORRIS, J., LAGUTINA, I., GROSVELD, G., OSAWA, M., NAKAUCHI, H. & SORRENTINO, B. 2001. The ABC transporter Bcrp1/ABCG2 is expressed in a wide variety of stem cells and is a molecular determinant of the side- population phenotype. Nat Med, 7, 1028-1034. ZIMMERMAN, M. & MEYER, N. 2011. Annexin V/7-AAd staining in keratinocytes. Methods Mol Biol, 740, 57-63. ZLATANOVA, J., BISHOP, T., VICTOR, J., JACKSON, V. & VAN HOLDE, K. 2009. The nucleosome family: Dynamic and growing. Structure, 17, 160-171. ZUBER, J., SHI, J., WANG, E., RAPPAPORT, A., HERRMANN, H., SISON, E., MAGOON, D., QI, J., BLATT, K., WUNDERLICH, M., TAYLOR, M.,

282

JOHNS, C., CHICAS, A., MULLOY, J., KOGAN, S., BROWN, P., VALENT, P., BRADNER, J., LOWE, S. & VAKOC, C. 2011. RNAi screen identifies Brd4 as a therapeutic target in acute myeloid leukaemia. Nature, 478, 524-528.

283