UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Validation of two novel mouse models of conditional Meis1 deletion to study roles in adult mouse hematopoiesis Miller, Michelle Erin 2014

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata


24-ubc_2014_spring_miller_michelleerin.pdf [ 53.88MB ]
JSON: 24-1.0167309.json
JSON-LD: 24-1.0167309-ld.json
RDF/XML (Pretty): 24-1.0167309-rdf.xml
RDF/JSON: 24-1.0167309-rdf.json
Turtle: 24-1.0167309-turtle.txt
N-Triples: 24-1.0167309-rdf-ntriples.txt
Original Record: 24-1.0167309-source.json
Full Text

Full Text

VALIDATION OF TWO NOVEL MOUSE MODELS OF CONDITIONAL MEIS1 DELETION TO STUDY ROLES IN ADULT MOUSE HEMATOPOIESIS  by MICHELLE ERIN MILLER B.Sc., The University of British Columbia, 2005  A THESIS SUBMITTED IN PARTIAL FUFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPY   in   THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES  (Medical Genetics)   THE UNIVERSITY OF BRITISH COLUMBIA  (Vancouver)     April 2014  © Michelle Erin Miller, 2014 ii Abstract Meis1 is recognized as an important transcriptional regulator in hematopoietic development and is strongly implicated in the pathogenesis of leukemia, both as a Hox transcription factor co-factor and independently. Despite the emerging recognition of Meis1’s importance in the context of both normal and leukemic hematopoiesis, there is not yet a full understanding of Meis1’s functions and the relevant pathways and genes mediating its functions. In this thesis, I provide the groundwork for a novel model system to explore Meis1 function. Mice with a loxP flanked Meis1 allele were crossed with two different conditional Cre-recombinase expressing strains, MxCre mouse and the ERTCre mouse. I validated conditional deletion of the Meis1 allele and the resultant decrease in mRNA and protein expression. I additionally studied expression of Meis1 related (MEINOX family) transcription factors in highly purified hematopoietic populations to generate an atlas of gene expression and focus our future studies. The inducible Meis1-deletion mice were then used to study if Meis1 is a requirement to maintain hematopoietic homeostasis in adult mice. I provide evidence for a critical role for Meis1 in hematopoietic stem cell maintenance and megakaryocytic and erythroid progenitor expansion in vivo. I furthermore identified two novel candidate effectors of Meis1 in the adult hematopoietic system, Hlf and Msi2 using Affymetrix expression analysis. As recent studies have suggested a role for Meis1 in the regulation of hypoxia-induced reactive oxygen species (ROS), I examined the impact of the ROS-scavenger N-acetyl-L-cysteine on the Meis1-/- phenotype in vivo. The results highlighted in this thesis provide direction and an experimental platform for further dissection of the mechanisms of Meis1 function in both normal and leukemic hematopoiesis.  iii Preface  The work in this thesis was completed for the purposes of a Doctorate of Philosophy in the department of Medical Genetics at The University of British Columbia under the supervision of Dr. R. Keith Humphries (Terry Fox Laboratory, Vancouver, BC).  All experimental designs are my own, however, several people made invaluable contributions. The conditional-knock out mouse line on which the experiments are based, C57BL/6 J Meis1tmloxP/+ (Meis1fl/+), was a generous gift from Drs N. Jenkins and N. Copeland (McLaughlin Research Institute, Great Falls, MT, USA). B6;129-Gt(ROSA)26Sortm1(cre/ERT)Nat/J mice were a gift from Dr. A. Weng (Terry Fox Laboratory, BC Cancer Agency, Vancouver, BC, CAN). Dr. Weng also provided his expertise in the examination of histology slides. All animal breeding protocols and experimental procedures were approved by the UBC Research Ethics Board Animal Care Committee (Ethics certificate A13-0063, Genetic control and manipulation of normal and malignant hematopoiesis). Wenbo Xu and David Lai of the Flow Core Lab in the Terry Fox Laboratory, BC Cancer Agency were an important resource for flow cytometric analysis and sorting.  Staff at the BC Cancer Agency Centre for Translational Genomics (Vancouver, BC) prepared histology slides and ran provided samples on the Affymetrix expression array.  The Affymetrix data processing and statistical analysis was performed by Dr. Madeleine Lemieux (Bioinfo, Ottawa, ON) and independently by Dr. Lars Palmqvist (University of Gothenburg, Germany), although only Dr. M. Lemieux’s are reported here. Mature megakaryocyte RNA was provided by Dr. Susie K. Nilsson (CSIRO, Australia).   iv Patty Rosten (Terry Fox Laboratory) provided expertise and manpower in the design and execution of the various molecular methods including Southern blot, PCR, Q-PCR and Q-RT-PCR. Fellow lab members Courteney Lai, Donald Lai and Harry Chang provided occasional assistance with mouse work (bleeds, irradiation). Donald Lai additionally performed the CFC-replating experiments under my supervision as a Research Student. All Humphries lab members must be acknowledged for their contribution of ongoing discussion and generosity.Table of contents Abstract ................................................................................................................................... ii Preface .................................................................................................................................... iii Table of contents ..................................................................................................................... v List of tables ......................................................................................................................... viii List of figures .......................................................................................................................... ix List of abbreviations ............................................................................................................... x Acknowledgements ............................................................................................................... xii Dedication ............................................................................................................................. xiii Chapter 1 : Introduction ........................................................................................................ 1 Thesis Overview: Meis1 as a key regulator in both normal and leukemic hematopoiesis ........ 1 Normal and leukemic hematopoiesis as stem cell driven hierarchies ......................................... 4 Deregulation of self-renewal and differentiation leads to the generation of leukemia ................. 6 Genetic programs underpin self-renewal and differentiation in normal and leukemic hematopoiesis. .................................................................................................................................. 8 Hox genes are critical for embryonic patterning ........................................................................... 8 Hox genes have low DNA binding affinity and specificity in isolation and thus require co-factors for function ........................................................................................................................ 9 The discovery of Meis1 is entrenched in the leukemic process .................................................. 10 Meis1 confers Hox binding specificity and affinity in conjunction with other Hox co-factors .. 13 In addition to influencing Hox binding, Hox co-factors influence the cellular localization of other co-factors ............................................................................................................................ 14 The combinatorial interactions between HOX, PBX and MEIS suggest numerous possibilities for regulation of gene expression ................................................................................................ 15 Knock-out & overexpression studies of Meis, Pbx and Hox reveal critical roles in normal and leukemic hematopoiesis. ................................................................................................................ 16 Pbx, Meis and Hox knock-outs in normal hematopoiesis ........................................................... 16 Hox and Meis overexpression are potent contributors to leukemogenesis ................................. 18 Meis1 is commonly overexpressed in patient AML samples ...................................................... 18 Meis1 overexpression in patient samples is often found in conjunction with dysregulated Hox gene expression ........................................................................................................................... 20 Mouse models demonstrate overexpression of Meis1 and a Hox gene are sufficient for leukemogenesis ............................................................................................................................ 22 Meis1 and Hox gene expression may constitute a core expression program that accommodates transformation .............................................................................................................................. 24 Identifying the targets and downstream effectors of Meis1 ....................................................... 26 To date, few targets of Meis1 have been identified in normal hematopoiesis ............................ 26 Although there are several candidate targets for Meis1 in leukemogenesis, few have been validated as essential for transformation ..................................................................................... 28 Delineating functional domains of MEIS1 ................................................................................... 31 Roles for Meis1 outside of hematopoiesis .................................................................................... 34 Knowledge of regulation of Meis1 expression is relatively limited ........................................... 35  vi Meis1 is a powerful player in the context of normal and malignant hematopoiesis, however fundamental characteristics remain unknown ............................................................................ 37 Chapter 2 : Materials & Methods ....................................................................................... 39 In vivo methods ............................................................................................................................... 39 Mice ............................................................................................................................................. 39 Induction of Cre recombinase expression/localization in vivo ................................................... 40 Phenylhydrazine induced model of hemolytic anemia ................................................................ 41 Reactive oxygen species scavenging by in vivo N-acetyl-L-cysteine treatment ........................ 41 Isolation of bone marrow and peripheral blood for analysis ....................................................... 42 Long-term repopulating cell (LTRC) and competitive repopulating unit (CRU) assays ............ 42 Cell sorting and FACS analysis .................................................................................................... 43 Isolation of retrovirally transduced producer cells and 5-FU pre-treated marrow ...................... 43 Isolation and phenotypic analysis of primary marrow progenitor and mature cell populations . 44 In vitro studies ................................................................................................................................ 49 Retroviral vectors and transduction ............................................................................................. 49 Hematopoietic colony-forming cell (CFC) assays ...................................................................... 51 Long-term culture initiating cell (LTC-IC) assay ....................................................................... 53 Molecular methods ........................................................................................................................ 54 Southern blot analysis .................................................................................................................. 54 PCR detection of the 5’ LoxP site ............................................................................................... 56 RT-PCR cloning of the 5’ LoxP site and truncated transcript ..................................................... 56 Western blot analysis of MEIS1 protein following in vitro deletion in MN1-overexpressing cells ..................................................................................................................................................... 57 Histology ..................................................................................................................................... 58 Q-PCR & Q-RT-PCR .................................................................................................................. 58 Affymetrix mRNA array and analysis ......................................................................................... 62 Statistical Analysis ......................................................................................................................... 63 Chapter 3 : Validation of Conditional Meis1 Knock-Out Models ................................... 64 Introduction .................................................................................................................................... 64 Results and Discussion ................................................................................................................... 67 Identification of LoxP sites flanking Meis1 in B6-Meis1tgLoxP/+ mice ......................................... 67 Breeding of B6-Meis1fl/fl/ B6;129Gt(ROSA)26Sortm1(cre/ERT)Nat/J and B6-Meis1fl/fl/ B6.Cg-Tg(Mx1-cre)1Cgn/J mice ............................................................................................................ 72 Induction regimen of MxCre/Meis1fl/fl and ERTCre/Meis1fl/fl mice in vivo and generation of tools to quantitatively measure Meis1fl/fl deletion ....................................................................... 73 Confirmation of truncated transcript and protein expression in ERTCre/Meis1 mice ................ 77 Overexpression of a truncated protein does not interfere with CFC potential or re-plating ....... 79 Meis family member expression in sorted wild-type mouse bone marrow populations ............. 82 Summary ......................................................................................................................................... 85 Chapter 4 : Meis1 is required for hematopoietic stem cell maintenance and erythroid/megakaryocytic potential .................................................................................... 86 Introduction .................................................................................................................................... 86 Loss of Meis1 in adult mice perturbs peripheral blood composition in MxCre/Meis1-/- and ERTCre/ Meis1-/- models ............................................................................................................. 87 Generation of megakaryocyte and erythroid progenitors are impaired in MxCre/Meis1-/- mice 92 Meis1-/- results in a loss of hematopoietic stem cells, common myeloid progenitors and megakaryocyte progenitors ......................................................................................................... 97 Meis1 is required for the maintenance of primitive cell populations capable of long-term hematopoiesis in vitro and in vivo ............................................................................................. 100  vii The requirement for Meis1 for HSC is cell intrinsic ................................................................. 106 Gene expression changes following loss of Meis1 in an HSC-enriched population ................. 111 Treatment with N-acetyl-L-cysteine rescues some of the phenotypic abnormalities seen with loss of Meis1 .............................................................................................................................. 118 Discussion ..................................................................................................................................... 126 Meis1 is required for HSC maintenance and self-renewal ........................................................ 126 Meis1 is required for megakaryopoiesis and erythropoiesis in the adult .................................. 129 Hlf and Msi2 are putative effectors of Meis1 function in adult hematopoiesis ........................ 132 Regulation of ROS may play a role in Meis1 function ............................................................. 135 Summary ....................................................................................................................................... 138 Chapter 5 : Discussion ........................................................................................................ 140 Introduction .................................................................................................................................. 140 Future avenues for research using the model outside of normal adult hematopoiesis .......... 141 to delineate roles for Meis1 in leukemia ..................................................................................... 141 Power to delineate roles for Meis1 during mammalian development ....................................... 143 Insights gained and key topics for resolution highlighted for the role of Meis1 in normal hematopoiesis by the model ........................................................................................................ 145 Alternative models of the hematopoietic hierarchy ................................................................... 145 Genetic programs underpinning cell fate decisions at the megakaryocytic-erythroid junction 149 Role of Meis1 in normal cell cycle ............................................................................................ 150 Role of ROS and regulation of Hif1α regulation by Meis1 ...................................................... 151 Conclusion .................................................................................................................................... 156 References ............................................................................................................................ 158  List of tables  Table 1.1: Models of Meis1-overexpression studied to date .................................................. 24	  Table 2.1: Dilution, clone and source of antibodies for FACS phenotyping and cell sorting 45	  Table 2.2: Cell sorting and phenotyping gating strategies. .................................................... 47	  Table 2.3: References for Sorting Gates ................................................................................. 49	  Table 2.4: Probes and anticipated hybridization sizes ............................................................ 55	  Table 2.5: Primer sets for Q-PCR detection of Meis1fl/fl genomic collapse ........................... 60	  Table 2.6: Primetime Q-RT-PCR Assays ............................................................................... 61	  Table 3.1: “Walking” Intron 7 primers sequence and anticipated fragment sizes. ................. 70	  Table 3.2: QPCR Primers for Genotyping and Detection of Meis1fl Deletion ....................... 76	  Table 4.1: PCR of individual CFC colonies for Meis1 deletion. ............................................ 97	  Table 4.2: LTC-IC frequency in MxCre/Meis1-/- bone marrow ............................................ 101	  Table 4.3: Reduction in HSC in MxCre/Meis1-/- and ERTCre/Meis1-/- mice ....................... 106	  Table 4.4: HSC frequency following in vivo deletion of Meis1 in primary recipients ......... 111	  Table 4.5: Genes differentially expressed with deletion of Meis1 as determined by Affymetrix analysis within a 90% adjusted confidence interval .................................. 114	  Table 4.6: Genes with > 2-fold expression change associated with deletion of Meis1 as determined by Affymetrix analysis .............................................................................. 118	   List of figures  Figure 1.1: Simplified schematic representation of the hematopoietic hierarchy. ................... 2	  Figure 1.2: Schematic representation of quantitative and qualitative long-term repopulation assays ................................................................................................................................ 6	  Figure 1.3: Meis1 Exon structure, mRNA and MEIS1 protein. ............................................. 12	  Figure 3.1: Southern blot analysis to localize the second LoxP site 3’ or 5’ to the known site. ........................................................................................................................................ 69	  Figure 3.2: Schematic representation of Intron 7 primers to narrow the position of the second LoxP site. ........................................................................................................................ 70	  Figure 3.3: Sequence of targeted Meis1 allele. ....................................................................... 71	  Figure 3.4: Southern blot analysis to localize the second LoxP site to 3’ or 5’ to the known site ................................................................................................................................... 74	  Figure 3.5: Schematic representation of Q-PCR primers for quantitative determination of Meis1fl deletion. .............................................................................................................. 76	  Figure 3.6: Sequencing results confirming introduction of premature stop codon in exon 9 of Meis1fl with expression of Cre recombinase. .................................................................. 77	  Figure 3.7: Western blot analysis confirming loss of MEIS1 protein following in vitro induction of Cre expression in mouse bone marrow expanded with pSF91-MN1 retrovirus. ........................................................................................................................ 79	  Figure 3.9: MEIS family expression in purified hematopoietic subsets. ................................ 84	  Figure 4.1: Cre expression induction schemes and summary of experiments performed with ERTCre/Meis1 and MxCre/Meis1 mouse models ........................................................... 89	  Figure 4.2: Loss of Meis1 is associated with gross marrow and peripheral blood changes in ERTCre/Meis1 and MxCre/Meis1 mice. ......................................................................... 92	  Figure 4.5: Loss of long-term repopulation and HSC self-renewal in the absence of Meis1. ...................................................................................................................................... 105	  Figure 4.7: Affymetrix mouse Exon ST 1.0 analysis of gene expression changes as a result of loss of Meis1 in an HSC-enriched population. ............................................................. 113	  Figure 4.8: NAC treatment partially abolishes differences between MxCre/Meis1-/- and MxCre/Meis1-/+ mice. ................................................................................................... 123	  Figure 4.9: Gene expression changes as a result of NAC-treatment in sorted HSC and MkP populations .................................................................................................................... 125	  Figure 5.1: Alternative models of the hematopoietic hierarchy. .......................................... 148	  Figure 5.2: Model of NAC activity in the hematopoietic compartment. .............................. 154	       x List of abbreviations  Abbreviation Name 4-OHT 4-hydroxy-tamoxifen AML Acute myeloid leukemia APC Allophycocyanin BFU-E Burst forming unit - erythroid BM Bone marrow BrdU Bromodeoxyuridine CFC Colony forming cell CFU-GEMM Colony forming unit - granulocyte/erythrocyte/monocyte/megakaryocyte CFU-GM Colony forming unit - granulocyte/monocyte CFU-Mk Colony forming unit - megakaryocyte CLP Common lymphoid progenitor CML Chronic myeloid leukemia CMP Common myeloid progenitor CRU Competitive repopulating unit CTD C-terminal domain Cy Cyanine DAPI 4',6-diamidino-2-phenylindole DMEM Dulbecco’s Modified Eagle’s Medium  DNA deoxy-nucleic acid dpc Days post coitus EDTA Ethylenediaminetetraacetic acid EtOH Ethanol FACS Fluorescence activated cell sorting FBS Fetal bovine serum FITC Fluroescein isothiocyanate FSC Forward scatter GFP Green fluorescent protein GMP Granulocytic-monocytic progenitor HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid HM Homothorax-Meis HRP Horseradish peroxidase HSC Hematopoietic stem cell IL Interleukin IP intraperitoneal IV intravenous LSC Leukemic stem cell LTC-IC Long-term culture initiating cell  xi LTRC Long-term repopulating cell MEP Megakaryocytic-erythrocytic progenitor MES 2-(N-morpholino)ethanesulfonic acid mRNA messenger ribonucleic acid NAC N-acetyl-L-cysteine PA Phoenix Ampho PB Peripheral blood PBS Phosphate buffered saline PCR Polymerase chain reaction PE Phycoerythrin PerCP Peridinin chlorophyll PHZ phenylhydrazine PI Propidium iodide PolyI:C polyinosinic:polycytidylic acid PURO Puromycin Q-PCR Quantitative PCR Q-RT-PCR Quantitative-RT-PCR RNA ribonucleic acid ROS Reactive oxygen species RT-PCR Reverse transcriptase PCR SA Streptavidin SCF Stem cell factor SQ Subcutaneous SSC Side scatter TBS Tris-buffered saline TBS-T TBS-Tween Tris Tris(hydroxymethyl)aminomethane YFP Yellow fluorescent protein     xii Acknowledgements  Thank-you so much to my committee members Dr. Gerald Krystal, Dr. Peter Lansdorp and Dr. Fabio Rossi for their contributions to this work in the form of critical appraisal and discussion. Their patience and understanding is also greatly appreciated. Humphries Lab and Terry Fox Laboratory members past and present are to be acknowledged for their unhesitating willingness to provide expertise, resources and discussion. A heartfelt thank-you to Dr. Keith Humphries for his ongoing guidance, support and outstanding example of a researcher dedicated to the pursuit of scientific advancement, irrespective of personal gain.     xiii Dedication  To my Mother for teaching me what was possible. To Patty for making the journey fun. And to Nick for his unwavering support.     1 Chapter 1 : Introduction Thesis Overview: Meis1 as a key regulator in both normal and leukemic hematopoiesis Self-renewal, that is the ability to generate a cell with equivalent potential upon division, is a characteristic shared by a specific subset of cells across a number of tissue types and species (reviewed in Orford & Scadden, 2008). This potential defines the tissue stem cell and makes it possible to maintain a tissue for the lifetime of the organism, especially in tissues with high cell turnover, such as the gut and blood. Of these tissue stem cell systems, the hematopoietic system is perhaps the best studied. Indeed, studies focused on the hematopoietic system over the last 50 years (reviewed in Spangrude et al., 1991; Akala & Clarke, 2006; Orkin & Zon, 2008) have led to significant insights and experimental approaches relevant not only to hematopoiesis but to a wide range of stem cell based systems.  Through the study of the hematopoietic system, certain stem cell characteristics have been established. For one, progeny of the stem cell must have the ability to give rise to the differentiated, functional cell types comprising the tissue. Another essential property of HSCs is their ability to undergo self-renewal divisions in which one or both progeny cells retain essentially all of the self-renewal and differentiation functions of the parent cell.  We thus now appreciate that the hematopoietic stem cell (HSC) sits at the apex of a hierarchy from which progenitors with increasingly restricted differentiation and self-renewal potential arise (Figure 1.1; modified from Orkin & Zon, 2008). Given the central role of HSC in initiating and sustaining lifelong hematopoiesis, a major and ongoing research challenge is to identify and understand the key regulators underlying the origin and function of normal HSC.   2  Figure 1.1: Simplified schematic representation of the hematopoietic hierarchy.  The hematopoietic stem cell (HSC) sits at the apex at the hierarchy. Through the process of self-renewal, the HSC pool is maintained. To maintain homeostasis, however, the HSC undergoes commitment to multi-potent progenitors with variable long-term reconstitution capacity and narrowing lineage commitment. Lineage committed progenitors undergo successive transient amplification and differentiation to fully differentiated and function cell types.  In a similar fashion, there is a growing understanding of the processes by which key regulatory mechanisms can be eroded and ultimately lead to the emergence of  transformed hematopoietic cells and resultant hematologic malignancies. For several such malignancies, including acute and chronic myeloid leukemia, a striking feature now recognized is that the leukemic population is also hierarchical in nature with self-renewing HSCMulti-potent progenitorsCLPCMPGMPMEPErythrocytesPlatelets Granulocyte Macrophages Dendritic cellT-cell B-cell Natural killer cell 3 leukemic stem cells (LSCs) at the apex (Bonnet & Dick, 1997; reviewed in Jordan, 2004; Buzzeo, Scott & Cogle, 2007; Dick, 2008). This “stem cell model” of leukemia again draws critical attention to questions about the origin and properties of LSCs and the key underlying regulators (reviewed Gilliland et al., 2004; Jordan & Guzman, 2004; Misaghian et al., 2008).  This parallel between normal and malignant hematopoiesis has been reinforced by the elucidation of many shared regulators. Notable among the common regulators now identified are a number of transcription factors that on the one hand can be essential for the early development or maintenance of normal HSCs (e.g. Runx1 and Tel/Etv6 respectively) but on the other hand, can acquire leukemogenic potential by a variety of mechanisms including deregulated expression or being part of translocations with creation of fusion genes (eg. AML-ETO and ETV6-FLT3) (Mulloy et al., 2002; Vu, et al., 2006).  In the thesis work described here, I have focused on a transcription factor Meis1, as one such candidate regulator implicated in both normal and leukemic hematopoiesis. These studies have in large part been based on a novel conditional knockout mouse model for Meis1 that I have validated for use in vitro and in vivo. The results of this work highlight a critical requirement of Meis1 in the maintenance of adult HSC potential, cell fate determination and expansion of differentiated erythroid and megakaryocytic progenitors. Use of this model combined with RNA expression analyses has also provided an opportunity for further identification of potential effectors of MEIS1 function.   This model now provides a framework for future examination and comparative analyses of Meis1’s role in both normal and leukemia hematopoiesis.   The following sections provide further background on the stem cell models as they relate to normal and leukemic hematopoiesis, an overview of some of the key known  4 regulators of primitive normal and leukemic cells, and finally a more extensive review of emerging evidence of the central role of members of the Hox family of transcription factors and their co-factors, most notably, Meis1 in these processes.  Normal and leukemic hematopoiesis as stem cell driven hierarchies The HSC exists in a bone marrow niche, a relatively protected environment whereby cues from the surrounding environment maintain a state of relative proliferative quiescence and reduced metabolic activity (reviewed Zon, 2008; Haylock & Nilsson, 2006). Cues from the environment promote HSC division at which point it can give rise to progeny that will undergo further rounds of replication and division to generate terminally differentiated, but functional, cell types with limited replicative potential including T-cells, B-cells, granulocytes, natural killer, megakaryocytes and erythroid cells. Incompletely differentiated progenitors exist at various branch points in the hierarchy that can amplify in numbers but only give rise to a limited number of cell types. For example, once committed to the megakaryocyte-erythroid lineage, the progenitor appears to preclude alternative fates (Akashi et al., 2000). This progressive restriction of potential but gain of physiological and immunological function coincides with loss of self-renewal potential.  These characteristics have been elegantly shown in mouse experimental models whereby populations at various stages of the hierarchy are isolated by virtue of cell surface markers and transplanted into irradiated recipients (Figure 1.2). The blood system is widely accessible and advances in multi-colour fluorescence activated cell sorting (FACS) technology, along with the identification and validation of markers has made it possible to isolate enriched populations of cells with defined potential (reviewed in Fulwyler, 1980) and  5 to track their output in the peripheral blood and bone marrow of immune-compromised recipient mice (Szilvassy et al., 1990). In vitro assays have also been developed to measure the potential of multi-potent progenitors; however, the in vivo transplantation models remain the gold standard for qualitative and quantitative measurement of HSC function (reviewed in Purton & Scadden, 2007). The ability to purify these cell populations and measure their function has paved the way for genetic studies that have allowed study into what cellular processes underpin hematopoietic cell function and potential.   6  Figure 1.2: Schematic representation of quantitative and qualitative long-term repopulation assays  Deregulation of self-renewal and differentiation leads to the generation of leukemia These same techniques have also paved the way to enhancing our understanding of mechanisms underlying leukemogenesis, that is the process by which key regulatory Donor/Test mouseCD45.1 or other distinguishing cell marker (ex fluorochrome)Harvest marrowRecipent miceCD45.2Ablate endogenous HSC and progenitors with radiation or chemical treatmentTest Dose X- Monitor test cell contribution to various lineages of the peripheral blood of the recipient mouse with immunophenotyping by FACS. - Engraftment >0.1% in myeloid and lymphoid lineages for >16 weeks post-transplant supports HSC present in test populationTest Dose X/10 Test Dose X/100- Transplant test marrow into treated recipient mice with “helper” cells to ensure recipient survivalTransplantation of serial dilutions of recipient marrow (either unsorted or sorted for test popula-tion) into secondary recipeient allows quantitation of HSC self-renewal in the primary recipentNumber of positive vs negative mice for multi-linage engraftment (lymphoid & myeloid) at >16 weeks post transplantation allows quatita-tion of HSC frequency in the test population by Poisson statistics EXPERIMENT TYPESa) Quantitation of HSC numbers: Serial dilu-tions of test cells into >3 recipients per dose b) HSC quality/fitness: Competition of test cells against another cell population of known HSC frequency (CRU) 7 mechanisms are eroded and a cell population acquires malignant potential. The paths to malignancy are many, such that there is considerable heterogeneity between patients in terms of the molecular underpinnings be it fusion genes, deregulated gene expression, acquired mutations and various combinations of these. In brief, these aberrations result in enhanced proliferation, a loss of differentiation and deregulated apoptosis such that non-functional leukemic cells overwhelm the hematopoietic system.  A growing body of evidence has revealed that leukemias can mimic the normal hematopoietic system in several aspects. Notably, the leukemic population is also heterogeneous within a given individual in terms of phenotype and potential. Some of the first demonstrations of this were colony forming assays undertaken in the 70’s and 80’s where it was shown that only a small set both human and mouse tumor cells had the ability to read out as colonies in in vivo and in vitro assays (reviewed in Huntly & Gilliland, 2005). Using FACS analysis, researchers have been able to enrich for cells that, while not forming the bulk of the leukemic population, have the capacity to form transplantable leukemias in vivo. These cells have been coined leukemic initiating cells (LIC) based on their ability to recapitulate the leukemia when transplanted into mice (reviewed in Hope, Jin & Dick, 2003; Passegué et al., 2003, Warner et al., 2004; Becker & Jordan, 2011). Serial transplantation studies provide further evidence that LICs can undergo self-renewal and thus satisfy key criteria to be termed Leukemic Stem Cells (LSC). Indeed in many situations, the terms LIC and LSC have become to be used interchangeably. In this thesis the term LSC is primarily used as a way to discuss the concept of a cell at the top of a hierarchically structured leukemia and LIC is used in an operational sense when LSC are being detected by functional readouts. Importantly, a stem cell model of leukemia does not imply the origin of the LSC  8 but leaves open the possibility that it arises from a preexisting HSC or from later progenitors that reacquire crucial stem cell functions such as self-renewal.   Genetic programs underpin self-renewal and differentiation in normal and leukemic hematopoiesis. Genetic studies have shown that a large number of interconnected networks underlying different mechanisms are in play for proper function, maintenance and differentiation within the hematopoietic hierarchy. Aberrations in these networks are not well tolerated and invariably result in cell death or favor the selection of further alterations that can lead to the evolution of grossly distorted cell function that can culminate in malignancy. These mechanisms include regulation of self-renewal potential (reviewed in Reya, 2003; Zon, 2008), cell-cycle (reviewed in Matsumoto & Nakayama, 2012), epigenetic modification (reviewed in Sashida & Iwama, 2012), cell signaling (reviewed in Rizo et al., 2006), cell survival and differentiation. Transcription factors are centrally poised to influence these mechanisms by virtue of regulation of gene expression. One such family, HOX genes and their co-factors, have been established as key components of HSC function during ontogeny and in the pathogenesis of leukemia.  Hox genes are critical for embryonic patterning Hox genes were first characterized in the fruit fly Drosophila melanogaster in mutagenesis screens. Two mutations were identified, antennapedia and bithorax, that caused significant alterations to the normal body segmentation of the fly. The genes underlying this phenomemon were termed homeobox genes and were found in two clusters: the bithorax complex (BX-C: Ubx, Abd-A, Abd-B) and the antennapedia complex (Ant-C: Lab, Pb, Dfd,  9 Scr, Antp). Homologs have been found in most species, including fish, fungi and plants, indicating an ancient origin with evolutionary significance. Gene and cluster duplication over the years have left vertebrate mammals with 4 clusters of Hox genes (A through D) that are scattered through the genome. Each cluster consists of 9 to 11 members of 13 paralog groups that are assigned on the basis of sequence similarity and location to the Drosophila Ant-C cluster and Abdominal-B  (Abd-B) gene of the bithorax cluster. Thus paralog groups 1 through 8 are Ant-C-like, and paralogs 9 through 13 are Abd-B-like. During development, Hox genes are expressed in temporal and positional colinearity, that is, 3’ genes are expressed more anteriorly and earlier than those at 5’ positions. Their expression patterns define domains along the anterior-posterior axis of the embryo and specify the development of structures within these domains (Reviewed in Grier et al., 2005). Hox genes have low DNA binding affinity and specificity in isolation and thus require co-factors for function Interestingly, despite the number of Hox genes and their critical role in development, each Hox gene has remarkably low DNA binding specificity and affinity and indeed, functional specificity cannot be achieved on the basis of DNA binding alone (reviewed in Rauskolb & Wieschaus, 1994). One candidate co-factor, extradenticle (exd) was identified in Drosophila that, when mutated, caused similar homeotic transformation of body segments to loss of the bithorax or antenepedia clusters. A mutation in exd, however, did not result in changes in expression of the homeotic genes themselves, nor was expression of exd regulated by these genes, suggesting action in parallel (Peifer & Wieschaus, 1990).   Exd was found to be homologous to PBX1, a human proto-oncoprotein found in one-quarter of pediatric pre-B-cell acute lymphoblastc leukemias as a fusion with E2A in t(1;19) translocation (Rasukolb, Peifer & Wieschaus, 1993). In vitro experiments revealed that Pbx1  10 could enhance the DNA binding specificity and affinity of numerous 3’ Ant-C-like Hox genes (HoxB4, B6 and B7) but not Abd-B-like HoxA10 (Chang et al., 1995). Discovery of a novel homeodomain-containing protein, Meis1, in a mouse model of leukemia would provide further insight into the complexity of interactions required for execution of HOX programs.  The discovery of Meis1 is entrenched in the leukemic process Myeloid ecotropic virus insertion site 1 (Meis1) was first described in 1995 as a common viral integration site in the BXH-2 model of myeloid leukemogenesis (Moskow et al., 1995). BXH-2 mice develop granulocytic leukemia by 1 year of age due to activation of cellular proto-oncogenes and/or inactivation of tumor suppressor genes by retroviral insertion of two endogenous ecotropic proviruses, Emv1 and Emv2. As the insertion is relatively random, a number of genes will be disrupted during the process, many of which do not contribute to the pathogenesis of the malignancy. Disruption of a gene in multiple independently arising tumors, however, is strongly suggestive of a causative role in the generation of leukemia in these mice. In this model, disruption of Meis1 was found in 15% of tumors, in contrast to the known oncogene c-myb locus that was only disrupted in 3% of tumors. Viral integrations at the Meis1 locus did not disrupt exon coding sequences and were localized into two clusters at the 5’ and 3’ regions of the locus. Northern blotting analysis of  these tumors and subsequent studies (described below) demonstrated increased expression of Meis1 in these leukemias, supporting a role for Meis1 as an oncogene in this context.  In this seminal work on Meis1, Moskow et al. additionally localized the mouse Meis1 locus to chromosome 11 (11.11 cM, NC_000077.6: 18880428..19018969, complement) and identified the cDNA transcript from adult mouse lung and spleen libraries. The Meis1 transcript contains 12 exons and encodes a protein of 390 amino acids. Two isoforms of the  11 transcript were postulated to exist based on alternative splicing within exon 10. In comparison to the prototypic Meis1a transcript, Meis1b has a novel 93-amino acid C-terminal region. Two isoforms of Meis1 are currently validated in the mouse and humans (Figure 1.3a). Meis1a is considered the canonical sequence in mouse and has 99.7% amino acid identity to the canonical MEIS1 protein in humans, differing by only 1 amino acid. The protein is characterized by a 63 amino acid DNA binding TALE-type homeodomain (Figure 1.3b domains). MEIS1 protein is highly conserved between species, with identity at 356 of the 390 amino acids between Mus musculus, Homo sapiens, Xenopus laevis, Rattus norvegicus, Bos taurus, Danio rerio and Pan troglodytes (Uniprot alignment, August 2013).   12  Figure 1.3: Meis1 Exon structure, mRNA and MEIS1 protein.  a) Comparison of the C-terminal of MEIS1A and MEIS1B. The proteins are identical until the splice site in exon 10. The homeodomain (HD) is underlined. The designation of isoform A or B is consistent with the historical literature and UniProt database as of August 2013. b) Relative exon positions within the genomic sequence. Schematic representation of Meis1 mRNA and relative exon contributions to MEIS1 protein domains. The Homothorax-Meis (HM) domain is highly conserved within the MESI/PREP family of proteins and is considered to be composed of two subdomains (HM1 and HM2), the sequence intervening the region and 8 amino acids C-terminal to the HM2 subdomain. The subdomains are also referred to in literature as HMA(HM1) and HMB(HM2). The HM domain is required for interaction with PBX-family members (Ryoo et al., 1999).  Exon 8Exon 1Exon 2Exon 3 Exon 4Exon 5Exon 6Exon 7 Exon 9Exon 10Exon 11Exon 12EXON 1 2 3 4 5 6 7 8 9 10 11 12Meis1mRNA3346bpMEIS1 proteinIsoform A: 390aaIsoform B:465aaHM1 HM2 HD241  GDNSSEQGDGLDNSVASPSTGDDDDPDKDKKRHKKRGIFPKVATNIMRAWLFQHLTHPYP  300  MEIS1A 241  GDNSSEQGDGLDNSVASPSTGDDDDPDKDKKRHKKRGIFPKVATNIMRAWLFQHLTHPYP  300  MEIS1B      ************************************************************ 301  SEEQKKQLAQDTGLTILQVNNWFINARRRIVQPMIDQSNRAVSQGTPYNPDGQPMGGFVM  360  MEIS1A  301  SEEQKKQLAQDTGLTILQVNNWFINARRRIVQPMIDQSNRAVSQGTPYNPDGQPMGGFVM  360  MEIS1B       ************  *  *  :.  *          *                         361  DGQQHMGIRAPGPMSGMGMNMGMEGQW-------HYM-----------------------  390  MEIS1A 361  DGQQHMGIRAPGLQSMPGEYVARGGPMGVSMGQPSYTQAQMPPHPAQLRHGPPMHTYIPG  420  MEIS1B                                                    391  ---------------------------------------------  390  MEIS1A 421  HPHHPAVMMHGGQPHPGMPMSASSPSVLNTGDPTMSAQVMDIHAQ  465  MEIS1B!a) b)  13 Meis1 confers Hox binding specificity and affinity in conjunction with other Hox co-factors At the time of the Moskow et al. study, the only region of similarity of MEIS1 to known proteins was that for the homeodomain of human proteins PBX1, PBX2, PBX3 and Drosophila extradenticle (exd) and Knotted (Kn1). This homology would prove to be a key guide to understanding the role of MEIS1 in the context of leukemia and normal hematopoietic function. Hth was identified as the Drosophila homolog of murine Meis1 based on sequence similarity (88% identity in the C-terminal region including 93% identity in the homeodomain, Pai et al., 1998) and was shown to be essential for Hox patterning of Drosophila (Rieckhof et al., 1997). MEIS, PBX and their homologs are classified based on homology as TALE-type homeoproteins. TALE-type homeodomains are distinguished by the presence of three extra amino acids between the first and second alpha helices of the homeodomain as compared to the homeobox (Hox) family of proteins that defined the domain. The TALE family of genes in the mouse now includes the MEINOX family, which includes Meis1, Meis2, Meis3, Pknox1 and Pknox2, and the PBC family consisting of Pbx1, Pbx2, Pbx3 and Pbx4 (reviewed in Moens & Selleri, 2006). Both PBC and MEINOX families were found to serve as HOX co-factors and to enhance DNA-binding specificity and affinity via MEIS1 or PBX-HOX duplexes or MEIS-PBX-HOX trimeric complexes (Shen et al., 1997; Shen et al, 1999; Shanmugam et al., 1999).  In contrast to PBX proteins that were found to bind Ant-C like HOX proteins (HOX 1 through 8), MEIS1 is complementary to PBX such that MEIS1 directly binds the Abd-B-like proteins of paralogs 9-13 (Shen et al., 1997). Through in vitro binding assays, a model was developed whereby HOX proteins from paralogs 1-8 preferentially form heterodimers with  14 PBX proteins, whereas HOX paralogs 9-13 form heterodimers with MEIS1. This preferential binding is highly influenced by the presence of DNA in the assay as yeast-two hybrid interaction assays have demonstrated interaction between MEIS1 and Ant-C-like paralogs 2, 4, 5 and 8 (Williams et al., 2005). Additional binding of the alternate co-factor to both PBX-HOX and MEIS-HOX complexes confers additional binding affinity but does not alter target specificity (Shanmugam et al., 1999). These studies and additional mutagenesis studies established that HOX-PBX dimer bind to a 5’-ATGATTNATNN-3’ consensus motif (Lu et al., 1995, Chang et al., 1996) that requires a core YPWMK N-terminal to the HOX homeodomain (Phelan et al., 1995) to interact with the 3 amino acid loop extension of the PBX homeodomain (Piper et al., 1999). The N-terminal PBC-A domain of PBX1 is required for interaction with MEIS1 through the HM1 and HM2 domains at the N-terminal of MEIS1 proteins (Ryoo et al., 1999, Wang et al., 2005, Mamo et al., 2006). The C-terminal of MEIS1 (18 amino acids MEIS1a, 93 amino acids MEIS1b) is required for interaction with various domains in the N-terminus of HOX proteins, however, this was not tested in the context of DNA binding (Williams et al., 2005) and the exact domains mediating MEIS-HOX interaction remain unclear. While the HOX core binding sequence remains unchanged, the MEIS binding consensus sequence is 5’-TGACAG-3’.  In addition to influencing Hox binding, Hox co-factors influence the cellular localization of other co-factors Interaction between PBX and MEIS proteins may be important for more than just DNA binding. In several models, the presence or absence of one co-factor influences the localization of the other. For example, in a Drosophila study exploring the factors influencing HOX function, the expression of hth in cells was found to strongly correlate with  15 nuclear localization of EXD (Rieckhof et al., 1997). Enforced expression of either Meis1 or hth was found to trigger nuclear localization of EXD, even when the DNA-binding homeodomain was altered. Additional studies in Drosophila and mouse suggest that in the absence of HTH/MEIS1, EXD/PBX is exported from the nucleus by a nuclear export signal and masking of N-terminal nuclear localization sequences (Berthelsen et al., 1999; Abu-Shaar et al., 1999; Jaw et al. 2000; Saleh et al. 2000). Interestingly, in Xenopus, Xpbx1b localizes to the nucleus in the absence in the absence of ectopic Xmeis1b, and indeed Xmeis1b is cytoplasmic in the absence of Xpbx1b (Maeda et al., 2002).  In addition to organism specificity, the requirement for localization may be context specific, as in mouse fibroblasts overexpression of PBX1 results in nuclear localization in the absence of MEIS1 or other family members. In the developing Drosophila, although EXD is nuclear in all cells where hth is expressesd, hth is not expressed in all cells where EXD is nuclear (Rieckhof et al., 1997). Other molecules have been identified in the fly (Wingless and Decapentaplegic) that induce EXD nuclear translocation in endoderm of the midgut (Mann and Abu-Shaar, 1996).  The combinatorial interactions between HOX, PBX and MEIS suggest numerous possibilities for regulation of gene expression HOX-PBX-MEIS signaling is likely to be very complex and not governed by a few simple principles as the diversity of partnerships between the 47 players is immense (38 Hox, 3 Meis, 2 Pknox and 4 Pbx genes), even without accounting for the contribution of the various splice isoforms of these genes. Much of this regulation is likely accomplished by tissue specific expression of various combinations of HOX, PBX and MEIS family genes. In the hematopoietic system, Pbx1, Pbx2, Meis1 and an assortment of Hox are preferentially expressed in fetal liver and adult bone marrow populations enriched for hematopoietic stem  16 cells with long-term repopulation capacity (Sauveageau et al. 1994; Pineault et al., 2002). Expression of Meis1 is highest in the HSC-enriched bone marrow compartment and progressively down-regulated in progenitors, with the exception of increased expression in megakaryocytic progenitors (Pineault et al., 2002; Hu et al., 2009; Okada et al., 2003), suggesting a role in regulation in these progenitor cells.  Regardless of tissue limited expression, however, the functional reach of Meis1 is likely substantial as either directly or indirectly it can interact with Hox and additional non-Hox co-factors. Given the roles in normal and aberrant hematopoiesis described below, it may thus constitute a critical regulatory node in these processes.   Knock-out & overexpression studies of Meis, Pbx and Hox reveal critical roles in normal and leukemic hematopoiesis.   Manipulation of gene expression in the embryo and adult organism are powerful tools with which to study gene function. Gene deletion in the zygote allows for investigation of whether genes are essential for embryogenesis, and if so, at what stage. If a gene is embryonic lethal, conditional deletion during development or in the adult allows powerful studies into pathways regulated by the gene. Use of these models has yielded insights into key processes regulated by Hox, Meis and Pbx in both normal and leukemic hematopoiesis.  Pbx, Meis and Hox knock-outs in normal hematopoiesis Despite the apparent interconnectedness and requirement for HOX-PBX-MEIS complexes at a variety of targets, knockout phenotypes are quite unique to each of the Hox co-factors. Pbx1 or Meis1 are expressed in sites of hematopoietic development in the embryo, that is, the 11.5dpc aorta-goanad-mesonephros (AGM) and 14.5dpc fetal liver (FL) (DiMartino et al., 2001; Hisa et al., 2004; Azcoitia et al. 2005). Constitutive knock-out of  17 Pbx1, Meis1 or conditional mis-localization of MEIS1 in the developing embryo result in a decline in hematopoietic colony forming cells (CFC) and poor reconstitution in competitive transplant experiments. Although both knockouts are embryonic lethal, the physiological mechanisms underpinning this process appear to be different in important ways. Pbx1-/- embryos are anemic, edematous and show skewing of progenitor differentiation at the expense of erythroid progenitors. Meis1-/- embryos, in contrast, show massive hemorrhaging from a lack of megakaryopoiesis and vascular patterning anomalies.   Most Hox genes of the A, B, C and D clusters are expressed in hematopoietic cells, with, in general, preferential expression in HSC-enriched populations and down regulation during differentiation (Sauvageau et al., 1994; Lawrence et al., 1997; Kawagoe et al., 1999; Pineault et al., 2002). Due to high homology between Hox genes, and thus likely redundancy, knock-out of individual genes have relatively mild hematopoietic phenotypes (reviewed in Gier et al. 2005; Argipopoulos & Humphries, 2007). For example, over-expression of HoxB4 in vivo or as an exogenous protein enhances HSC self-renewal and expansion (Sauavgeau et al., 1995; Antonchuck et al. 2002; Krosl et al., 2003); however, in contrast to loss of Meis1 or Pbx1, Hoxb4-/- mice are viable and born at expected Mendelian frequencies (Brun et al., 2004). Cells have a subtle reduction in repopulating ability and frequency, although this has little impact on steady-state hematopoiesis. Greater impairment in hematopoietic potential is seen in compound Hoxb3 and Hoxb4 mutant mice, although differences again are not dramatic in the steady-state (Björnsson et al., 2003). Similarly, Hoxa9-/- bone marrow has an 8-fold reduction in repopulating activity and 30-40% reduced numbers of total leukocytes and lymphocytes, although normal red blood cell, hematocrit and platelet counts (Lawrence et al., 1997; Lawrence et al., 2005).   18 Hox and Meis overexpression are potent contributors to leukemogenesis The role of Pbx for Hox gene specification during embryogenesis had been previously well established as had an independent role for PBX in the generation of acute lymphoid leukemia as partner in the translocation arising in the fusion gene E2A-PBX1 (Kamps et al., 1990; Chang et al., 1995). From the outset of its cloning from BHX-2 mice, Meis1 had been described as Pbx-like, however, a similar role in leukemogenesis to Pbx had yet to be described. In a study published shortly after their initial publication identifying Meis1 (Moskow et al. 1995), the group headed by Drs Jenkins and Copeland highlighted the strong link between Meis1 and Hox in the generation of myeloid leukemias that would prove to be the foundation of a growing body of work investigating this phenomenon. Cloning of retroviral insertion sites in CpG islands of BHX-2 mice was undertaken in an effort to identify insertions likely to influence gene expression (Nakamura et al., 1996). 4 insertion sites were identified using this method, 2 of which clustered to Hoxa7 and Hoxa9 and were found to enhance gene expression. Although one common insertion site remained unidentified, the other was found to correspond to Meis1. Strikingly, although 9.6% of the tumors screened had retroviral insertions in Meis1 that enhanced gene expression, 95% of these also had insertions in either Hoxa9 or Hoxa7.  The preference for Hox/Meis co-expression in the development of these leukemias was further enforced by the finding that of 21 leukemias with Hox dysregulation, only 3 lacked Meis1 insertion.  Meis1 is commonly overexpressed in patient AML samples Many mouse and human studies have followed that point to a role for deregulated/overexpression of Meis1 and Hox in the generation of leukemias. In examining expression of MEIS1 in leukemia, interesting patterns have emerged. MEIS1 was found to be expressed in a selection of patient bone marrow samples from all recognized FAB AML  19 subtypes with the exception of the promyelocytic (M3) subtype (Kawagoe et al., 1999; Lawrence et al., 1999; Afonja et al., 2000; Drabkin et al., 2002; Roche et al., 2004; Camós et al., 2006; Grubach et al., 2008; Zangenberg et al., 2009). In AML samples where MEIS1 expression is present, there also exist high levels of HOX-family member expression (HOXA4, HOXA7, HOXA9, HOXA10). It is worthwhile to note that although HOX genes are not expressed in the FAB-M3 AML subtype (Lawrence et al., 1999), there are also leukemias in which high HOX-family member expression is not accompanied by a concurrent increase in MEIS1 expression (Drabkin et al., 2002, Afonja, et al., 2000).  Levels of MEIS1 expression may have a prognostic significance. In normal karyotype AML, low HOXA4 expression levels is associated with shorter overall survival (Grubach et al. 2008). When these samples are subdivided on the basis of MEIS1 expression, patients with high levels of MEIS1 expression had significantly worse outcomes than those with low MEIS1 expression (Zangenberg et al., 2009). This study, however, did not take into account expression levels of other HOX-family members in the samples. Other studies looking at broader sets of HOX and MEIS expression have found associations between high levels of HOX expression in AML with intermediate risk cytogenetics (normal karyotype and translocations not falling into the high or low risk categories). Within these samples, high levels of HOX expression were associated with high levels of FLT3 (receptor for growth factor FLT3L) expression, with the highest correlation between HOXA7 and MEIS1 expression (Roche et al., 2004).  In cell lines derived from un-selected AML patient samples, most co-express MEIS1, HOXA7 and HOXA9 (Afonja et al., 2000, Lawrence et al., 1999).   20 Meis1 overexpression in patient samples is often found in conjunction with dysregulated Hox gene expression This level of association between HOX and MEIS1 expression is unlikely to be random and indeed enforced co-expression of HOX and MEIS1 by retroviral overexpression in mouse models of leukemia supports a causative role in leukemogenesis. Nucleophosmin (NPM) mutated AML with high levels of cytoplasmic NPM accumulation (NPMc+) are also associated with high levels of Hox and Meis1 expression (Alcalay et al., 2005). Knock-down of Meis1 by shRNA impairs expansion of primary NPMc+ AML cells in vitro and triggers increased levels of apoptosis compared to NPMwt AML (Woolthuis et al., 2012).  Examination of MLL-rearranged leukemias in both human samples and mouse models provides some of the most compelling evidence for the involvement of MEIS1 in leukemogenesis. MLL is a transcriptional activator that plays an essential role in regulating HOX and other gene expression during embryogenesis through histone H3 lysine-4 methyltransferase activity. Rearrangements of MLL at 11q23 are frequent anomalies in both adult and pediatric acute leukemias whereby N-terminal of MLL is fused to a variety of C-terminal partners in 20% of ALL and 5-6% of AML samples (Somervaille & Cleary, 2010; Tamai & Inokuchi, 2010). All samples with t(4:11) translocations express high levels of MEIS1 (Rozovskaia et al., 2001; Kohlmann et al.,  2005; Trentin et al., 2009). Common t(4:11) fusion partners MLL-ENL, MLL-AF4 have shown a characteristic expression profile of HOX and MEIS1 over-expression (Rozovskaia et al., 2001; Zeisig et al., 2004). Furthermore, cells susceptible to transformation by the MLL-AF9 fusion gene express high levels of MEIS1 and Meis1 is a component of the expression profile in mouse models of MLL-AF9 induced transformation (our data, Chen et al., 2008; Kumar et al., 2010). In a study using mouse models by Wong and colleagues, the decreasing time to the onset of  21 leukemia of various MLL-fusions was shown to be linked in a dose dependent manner to the extent in which they triggered up-regulation of Meis1 (Wong et al., 2007).  These leukemias exhibited a more immature phenotype and had higher colony-forming capacity, suggestive of a primitive cell type with extensive replicative potential. In addition, Meis-/- fetal liver derived cells could not be transformed with the fusions, supporting that expression of Meis1 is an essential and rate-limiting set in leukemogenesis in the context of MLL-fusions. Other fusion genes characterized by high levels of MEIS1 expression in the context of AML include MYST3-CREBBP (Camós et al., 2006), NUP98-HOXA9 (Takeda et al., 2006) and NUP98-NSD1 (Wang et al., 2007). Although initial studies of MEIS1 expression in acute leukemia suggested MEIS1 is not expressed in lymphoblastic leukemia (Lawrence et al., 1999), expression in phenotypically lymphoid MLL-fusion samples as well as infant and adult ALL samples suggests MEIS1 also may play a role in these leukemias. Compared to normal cells, ALL cell lines and patient samples express increased levels of MEIS1 and MEIS2, and knock-down of MEIS1 expression decreases their proliferation (Rozovskaia et al., 2001; Imamura et al., 2002; Ferrando et al., 2003). Strikingly, MEIS1 and HOXA9 are co-expressed at high frequency (74%) in infant (under age 5) MLL-rearranged and non-MLL rearranged leukemias (Imamura et al., 2002).  Decrease in proliferation following down-regulation of MEIS1 may be the mechanism by which some ALL-cell lines manifest resistance to chemotherapeutics (Rosales-Aviña et al., 2011). ShRNA knock-down of MEIS1 in a B-cell leukemic line expressing MLL-AF4 resulted in impaired engraftment in mouse transplantation studies, likely due impaired proliferation and chemotaxis to the appropriate niche (Orlovsky et al., 2011).   22 Mouse models demonstrate overexpression of Meis1 and a Hox gene are sufficient for leukemogenesis Molecular manipulation of genetic sequence and expression in mouse models has proved to be invaluable in furthering our understanding of critical domains underpinning the leukemic process. Building on the initial studies demonstrating frequent retroviral co-activation of Hoxa9 and Meis1 in BHX-2 leukemias (Moskow et al., 1995), subsequent studies in mice have greatly expanded the range of Hox genes and Hox-fusions with which Meis cooperates as well as defined domains and contributory genetic pathways in this process. In studies designed to address whether all HOX genes have intrinsic leukemogenic properties, either as intact genes or as NUP98 fusions, Meis1 was found to reduce the latency to leukemia with all constructs tested (Pineault et al., 2005).  Overexpression of Meis1, 2 or 3, in isolation, however, does not have an oncogenic effect and fails to immortalize progenitors in vitro or cause leukemias when overexpressing cells are transplanted in vivo (Kroon et al., 1998; Calvo et al., 2000).  Collaboration between Hox, Nup98-Hox and Meis1 appears to be a universal property as, to date, no reports exist of Meis1 failing to cause transformation in the presence of Hox or Nup98-Hox overexpression (summarized in Table 1). Hox and Nup98-Hox partners tested to date include those found in human and mouse leukemias as well as other family members not overexpressed in leukemias or engineered fusions. Meis1 clearly contributes to the oncogenic process in these leukemias, as opposed to reinforcing programs established by the Hox or Nup98-Hox partner. This is most clearly illustrated by studies with the engineered fusion Nup98-Hoxa10-homeodomain (NA10HD), where the Hoxa10 is pared down to the DNA binding homeodomain in the fusion. Overexpression of NA10HD in HSCs results in nearly exclusive self-renewal divisions in vitro, resulting in a hugely expanded stem cell pool.  23 When transplanted in vivo, however, the cells respond appropriately to their environment and although engraftment is robust, leukemias do not form (Ohta & Sekulovic et al., 2007). Co-expression of Meis1 with NA10HD, however, results in transformation and leukemias in vivo (Pineault et al., 2004), supporting the idea that Meis1 triggers a unique program for transformation, independent of the Hox partner. Of further interest, NA10HD lacks any known motifs for direct interaction with MEIS1, thus raising questions about the basis for the powerful functional interaction seen with co-overexpression of MEIS1 and NA10HD.   Co-operation between Meis1 and Hox is extremely powerful and is synergistic, providing further support for independent programs regulated by Hox and Meis1. For example, Meis1 overexpression fails to generate transplantable leukemia cells and Hoxa9 overexpression does so at a relatively low frequency (1 in 9, Wang et al., 2005) after a long latency (280 days). The Hoxa9-leukemias that do result invariably have accumulated additional mutations that activate Meis1 expression. When co-expressed, however, the frequency of leukemia-initiating cell is minimally increased 9-fold and the latency to disease is reduced 4-fold to roughly 72 days. The decreased latency and increased frequency in Hox plus Meis1 co-expressing leukemias supports a synergistic effect between the two genes whereby the genetic programs triggered are more powerful than simply an additive effect enhanced expression of a shared program.  This synergism is also true in Hox-expressing MLL-fusion leukemias where enforced Meis1 expression reduces the latency to disease and increases leukemia-initiating cell frequency (Wong et al., 2007).       24 Table 1.1: Models of Meis1-overexpression studied to date  Mutations/Overexpression Found in Patient Samples Gene Reference HoxA9  Kroon et al., 1998; Wang et al, 2005 HoxA7  Wang et al, 2004 HoxA10 Pineault et al., 2004 HoxB3 Sauvageau et al., 1997 Nup98-HoxD13 Pineault et al., 2003 Nup98-HoxA9 Kroon et al., 2001; Iwasaki et al., 2005 Nup98-HoxA10 Pineault et al., 2004 Nup98-PMX1 Hirose et al., 2008 MLL-GAS7 Wong et al., 2007 MLL-AF10 Wong et al., 2007 MLL-LAF4 Wong et al., 2007 MLL-ENL Ferrando et al., 2003 MLL-AF5q31 Imamura et al., 2002 NPMc+ Woolthuis et al., 2012 MN1 Heuser et al., 2011 Artificial/Engineered Mutations/Overexpression  HoxB4 Pineault et al., 2004 HoxB6 Fischbach et al., 2005 Nup98-HoxA10-homeodomain Pineault et al., 2004 Nup98-HoxD13-homeodomain Pineault et al., 2004 Nup98-HoxB3 Pineault et al., 2004 Nup98-HoxB4 Pineault et al., 2004  Meis1 and Hox gene expression may constitute a core expression program that accommodates transformation For many leukemias, if not all, there is evidence to suggest that Meis1 and some Hox expression are required for a cell to be susceptible to transformation. Previous studies have demonstrated that committed progenitors, such as common myeloid and granulocyte-macrophage progenitors (CMPs and GMPs, respectively) can be transformed, provided that the oncogenic event confers or enhances self-renewal properties to allow for expansion of the leukemic clone. The self-renewal capacity of and HSC is not universally sufficient, however.  25 This is exemplified in chronic myelogenous leukemia (CML) where the BCR-ABL fusion gene is present in HSC population, but it is the outgrowth of a more committed cell that causes disease (Maguer-Satta et al., 1996). In addition, no terminally differentiated cells have been successfully transformed by leukemia-associated oncogenes to date, suggesting there is a limited population of cells susceptible to transformation with given properties.  Cells susceptible to transformation appear limited for each oncogene and are likely defined somewhat by the extent to which the oncogene can exploit naïve cellular programs. In a seminal work, B.J.P Huntly et al. tested the hypothesis that all leukemia oncogenes have the capacity to confer self-renewal to committed progenitors to generate leukemic stem cells/leukemia initiating cells. They found that while the fusion MOZ-TIF could transform committed myeloid progenitors at the CMP and GMP level, BCR-ABL could not and that transformation by this gene was limited to cells with inherent self-renewal potential, that is, the hematopoietic stem cell (Huntly et al., 2004). Recent work implicates the Meis1 program in conferring target cell susceptibility for transformation. Using the MN1 model of leukemogenesis, Heuser et al. were able to show that expression of Meis1 is a requirement for target cell transformation and that dominant negative M33-Meis1 abolishes leukemogenicity (Heuser et al., 2011). When they examined the expression profiles of cells susceptible to transformation by MN1 and MN1-leukemias, they found a high degree of correlation between Meis1 and Meis1-associated factor expression (Flt3, Mef2c). Enforced expression of Meis1 expands the range of cells that can be immortalized by MN1, however, AbdB-like Hox expression is required for full transformation. These studies and others have suggested a “code” of expression required for full transformation and that both the oncogene and target cell play a complementary role in contributing to full expression of the code. This  26 provides further support for the concept that Meis1, in conjunction with Hox gene expression, may be a cornerstone in this code relevant to normal and leukemic hematopoiesis,.  Identifying the targets and downstream effectors of Meis1   Despite its role as a transcription factor, identification of specific targets genes and pathways for Meis1 remains limited despite a number of studies looking at gene expression changes in both over-expression and knock-out models. Cell surface receptor Flt3 has been shown to be up-regulated with Hox and Meis1 in several leukemic expression studies, although it is not a requirement for transformation (Morgado et al., 2007). Several cell cycle molecules have been identified in similar over-expression and knock-down studies, including Cdk2 and Ccnd3 (Kumar et al., 2009; Argiropoulous et al., 2010). A whole genome chromatin immunoprecipitation and sequencing (ChIP-seq) approach in mouse ES-derived hematopoietic progenitors identified >8000 regions bound by MEIS1 (Wilson et al., 2010). A similar approach in and Hoxa9 and Meis1-overexpressing mouse cell lines identified 624 regions bound by MEIS1 (Huang et al., 2012). Reassuringly, previously identified targets were confirmed in these studies, however, the cell populations in which the studies were performed may differ from naïve cell populations in important respects. This highlights the need for a conditional model of gene expression at physiological levels in purified cell populations of interest. Below is a summary of our current understanding of genes regulated by Meis1 in normal and leukemic hematopoiesis.   To date, few targets of Meis1 have been identified in normal hematopoiesis  Expression of Meis1 is enriched in the most primitive HSC populations and progressively down-regulated with differentiation, with the exception of the megakaryocyte,  27 suggesting a key role in regulation of these cell types. Genome-wide ChIP-seq in the ES derived hematopoietic progenitor cell line HPC-7, suggests that Meis1 binding is enriched in promoter regions at >8000 MEIS1 sites in the genome (Wilson et al., 2010). An inducible model of Meis1 expression in the mouse ES system suggests that Meis1 expression inhibits commitment to erythroid fate via inhibition of erythroid specific genes (Hba-a1/2 and Gypa) and expression of Ptger3 (Cai et al., 2012). Ptger3 encodes for prostaglandin E2, a factor known to promote short-term reconstituting hematopoietic progenitor proliferation (Frisch et al., 2009).  Bona fide targets for Meis1 in adult HSC populations at physiological levels have not been reported to date, with the exception of recent studies implicating regulators of anaerobic metabolism, the hypoxia-inducible factors (Hifs). Studies published during the conduct and writing of this thesis work support a model whereby HSC employ anaerobic metabolism to restrict the generation of reactive oxygen species (ROS) in the low-oxygen bone marrow niche. Regulation of metabolism by restricting the generation of ROS is significant as increased levels of ROS are associated with loss of HSC potential (Ito et al., 2006; Miyamoto et al., 2007; Takubo et al., 2010). Meis1 was found to directly bind the promoter of master regulators of anaerobic metabolism Hif1α (Simsek et al., 2010) and Hif2α (Kocabas et al., 2012) and positively influence gene transcription. Conditional knock-out of Meis1 in the adult mouse results in decreased HSC function that can be rescued by ROS-scavenging compounds, supporting a causative role for Meis1 in regulating aneorobic metabolism and hence HSC function (Kocabas et al., 2012; Unnisa et al., 2012). This is a powerful demonstration of how conditional knock-out models can be exploited to identify physiologically relevant target gene expression.   28 Consistent with Meis1 up-regulation of expression in megakaryocytic progenitors (Okada et al., 2003) and requirement of Meis1for embryonic megakaryopoiesis (Hisa et al., 2004; Azcoitia et al., 2005), Platelet factor 4 (Pf4 or Cxcl4), is a direct target of MEIS1 with PBX1B and PBX2 (Okada et al., 2003). MEIS1 was recently shown to regulate a transcription of dynamin isoform DMN3 via a promoter uniquely used in megakaryopoiesis (Nürnberg et al., 2012). Inhibition of dynamin activity in megakaryocytes inhibited platelet generation in vitro, supporting a role of Meis1 regulation in megakaryopoiesis. Meis1 may also suppress erythropoiesis in the developing embryo once cells are committed to the megakaryocytic lineage. Overexpression of Meis1 in embryonic cell (ES) cultures resulted in increased levels of megakaryocyte progenitors based on CD41 expression and colony formation (Cai et al., 2012). This result is somewhat in contrast with studies in zebrafish whereby knockdown of meis1 also results in impaired definitive erythropoiesis (Cvejic et al., 2011). Regardless of species differences, collectively these results implicate Meis1 as a regulator of erythro- and megakaryopoeisis. Further studies using a conditional adult model of Meis1 deletion would help to delineate roles in maintaining normal, adult homeostasis in these linages.  Although there are several candidate targets for Meis1 in leukemogenesis, few have been validated as essential for transformation  In leukemia, Meis1 expression is invariably linked to Hox gene over-expression, and as such, the majority of studies looking a targets of Meis1 are preformed in the context of overexpression of both factors. In contrast to ChIP-seq in the ES derived progenitor population, only 5% of Meis1 and Hoxa9 DNA binding appears to localize to promoter regions (Huang et al., 2012). The majority of DNA binding that influences gene expression in this model occurred in enhancer regions, distal from transcription start sites, and was  29 associated with H3 and H4 acetylation of CBP binding, marks of active enhancer regions. Further interrogation of these regions validated enhancer and repressive activity of gene expression in ~50% of the regions tested (12/22). This study confirmed gene regulation by Meis1 and Hoxa9 of Erg1, CD34 and Flt3, targets identified by gene expression analysis in a previous overexpression study (Wang et al., 2005). Growth factor receptor Fms-related tyrosine kinase 3 (FLT3) appears to be an important, although not essential, target for MEIS1 in the transformation process, and indeed is one of the best-characterized targets to date. FLT3 is a tyrosine protein kinase receptor expressed in virtually all patients with AML and in a large portion of ALL samples. Co-expression of Meis1 in Hoxa9 immortalized mouse progenitors up-regulated Flt3 expression and chromatin immune-precipitation experiments demonstrate direct binding of Hoxa9 and Meis1 to the Flt3 promoter (Wang et al. 2005; Wang et al., 2006). Moreover, engineered overexpression could replace Meis1 for leukemic transformation by NUP98-Hox fusion genes (Palmqvist et al., 2006). However, Flt3 is not essential for transformation , as Meis1 mutants that do not trigger upregulation of Flt3 can still cooperate with NUP98-Hox fusions (Argiropoulos et al.,2008) and  Flt3 -/- cells are readily transformed by Hoxa9+Meis1 overexpression and cause leukemias with similar phenotype and latency to wild-type marrow (Morgado et al., 2007). This is likely due to an ability of MEIS1 to trigger expression of downstream Flt3 targets in response to FL signaling, that is, activation of ERK1/2, Akt and nuclear factor-κB (Argiropoulos et al., 2008).   One mechanism by which oncogenes transform cells is through enforced cell cycle progression. Evidence for a role for Meis1 in influencing cell cycle is inconsistent and may vary with cell context. In leukemia models of Hoxa9 overexpression, addition of Meis1 does  30 not increase levels of cell cycle mRNA nor does Vp16-Meis1 confer enhanced growth properties to fibroblasts, suggesting Meis1 oncogenicity is not related to enhanced proliferation. In MLL leukemia models, however, increased Meis1 expression is associated with increased cell cycle entry (Wong et al., 2007). Additionally short-hairpin RNA against Meis1 in MLL-Af9 overexpressing cell-lines reduced expression of several genes regulating DNA replication and cell-cycle entry, including Cdk2, Cdk6, Cdkn3, Ccna2, Cdc7, Cdc42, Rbl1 and Wee1 and resulted in cell cycle arrest at Go/G1 (Kumar et al., 2009). Further support for Meis1 as a modulator of cell-cycle comes from experiments with an engineered dominant repressive form of Meis1, M33-Meis1. Enforced expression of M33-Meis1 results in a partial G1 block in AML leukemia lines, mediated by a decrease in Cyclin D3 levels and reduced Rb phosphorylation (Argiropoulos et al., 2010). Outside of leukemogenesis, Meis1 enforces cyclin D1 and c-myc expression in the developing zebrafish retina (Bessa et al., 2008). Collectively this data implicates Meis1 as a regulator of cell cycle progression although perhaps only minimally in the leukemic context.  Of interest given recent studies suggesting a role for Meis1 in the regulation of ROS in the HSC, regulation of ROS in the LSC by MEIS1 may also be significant. Low levels of ROS were recently found to be associated with LSC frequency in a Hoxa9+Meis1 mouse model of leukemia (Herault et al., 2012). The expression levels of Meis1 are equivalent between the lines used in the study, however, and the authors linked low level of ROS in high frequency LSCs to expression levels of the ROS scavenger Gpx3. This does not preclude the importance of Meis1 in enforcing expression of Gpx3 or other regulators of ROS in these leukemias, and in fact, a Meis1 binding site exists in the 3’UTR of Gpx3, suggesting regulation may be possible. The sum of these studies support Meis1 as master  31 regulator of a complex genetic program that has strong oncogenic potential when dysregulated. Further investigations are required to precisely pinpoint which of the many genes with DNA binding and altered expression in the presence of Meis1 are fundamental to the transformation process.   To date, no putative Meis1 effector genes, including Flt3, can substitute for Meis1-overexpression in transformation models. For example, In a NUP98-HOXD13 overexpression model, Trib2, Dlk1, Ccl3, Ccl4 and Rgs1 were identified as up-regulated and bound by Meis1 in the transformation process. Only Trib2, however, could replace Meis1 to cause leukemia with NUP98-HOXD13 overexpression although the latency to leukemia was much longer (140 days vs 50 days with Meis1), suggesting further collaborating mutations are required for transformation (Argiropoulos et al., 2008). These studies highlight the complexity of understanding effectors of Meis1 activity, as there are likely several key effectors that have complementary and additive effects, as well as redundancy in the system.    Delineating functional domains of MEIS1  Deletion mutant studies have delineated several important domains of Meis1 required for its function in normal and leukemic hematopoiesis. While many of these studies have been conducted in the context of leukemia models, likely the findings are largely also relevant to Meis1 roles in normal hematopoiesis. For transformation, the DNA binding by MEIS1 and the C-terminal domain are critical in the context of Hoxa9 and MLL-fusion leukemia. DNA binding mutant MEIS(N51S) does not collaborate with Hoxa9 overexpression (Wang et al., 2005) in the animal models or rescue transformation capacity by MLL-AF9 on a Meis-/- background (Wong et al., 2007).  Interaction with PBX also  32 appears to be essential as mutations or deletions in the HM domains required for PBX interaction abrogate Hoxa9-overexpression transformation and fail to rescue transformation by MLL-AF9 (Mamo et al., 2006; Wang et al., 2006; Wong et al., 2007).  The C-terminal domain of MEIS1 is also essential for transformation, likely due to transactivation of expression as opposed to direct HOX-interaction. Deletion of the C-terminal domain (CTD) after the homeodomain (amino acid 334) or at the distal C-terminus (amino acid 371) results in a mutant MEIS1A protein that cannot cooperatively bind PBX-HOX to trigger gene expression in cell lines expressing reporter constructs nor collaborate with Hoxa9 overexpression in vivo (Mamo et al., 2006; Wang et al., 2006). These truncations also fail to rescue MLL-AF9 serial replating in vitro on a Meis-/- background. While it may be temping to assume this is due to a loss of HOX and DNA-binding activity, this does not appear to be the case. A more restricted truncation mutant that results in the loss of the terminal 18-amino acids responsible for HOX protein interaction, Meis370T, retains the ability to immortalize progenitors and cause AML in vivo with Hoxa9 over-expression (Wang et al. 2005). In addition, loss of the CTD does not appear to influence DNA binding in the context of HOX and PBX1, despite abrogation of reporter gene activity with the loss of the terminal 18 amino acids of MEIS1A (Huang et al, 2005).  Current theories suggest that loss of function as a result of CTD truncation may be due to a transactivation function localized in the domain that is activated following chromatin remodeling on PBX-HOX-MEIS hetero-trimer or PBX-MEIS heterodimer responsive genes (Huang et al., 2005). C-terminal mutants of Meis1a and Meis1b were examined for transactivation and DNA binding with PBX1, HOXA1 and HOXB1 on endogenous promoter elements. Researchers found that MEIS1 appeared to be recruited to  33 the promoter where PBX1 and HOX were already bound following activation by either PKA signaling or HDAC inhibition. They propose a model by which PBX exerts a repressive function on transcription from PBX-MEIS or MEIS-PBX-HOX complexes until cellular signaling triggers chromatin remodeling to a permissive state that permits recruitment of these complexes to genes or recruitment of MEIS to PBX-HOX bound promoters. Furthermore, evidence in zebrafish suggests MEIS may act in these complexes to maintain HistoneH4 acetylation and transcription via recruitment of transcriptional activator and histone acetyltransferase CBP (Choe et al., 2009). Although these studies are on developmentally regulated promoter regions, and not necessarily those involved in leukemia, the role of the C-terminal domain of MEIS1 is likely linked to transactivation of transcription as mutations abrogate function in both contexts.  The transactivation function of the MEIS1 CTD is supported by studies with Pknox1, a MEIS-family member. Pknox1 and Meis1 share high sequence similarity the conserved HM motifs and homeodomain but diverge significantly in the CTD region. PKNOX1, however, appears to have a tumor suppressive role as overexpression in mouse models prolongs the latency of Hoxa9 overexpressing cells and additionally PKnox1-/- mice are tumor prone in the context of Eµ-Myc (Thorsteinsdottir et al., 2002; Longobardi et al., 2010).  The CTD of PKNOX1 appears devoid of transactivation capacity (Huang et al., 2005), however, replacement of the CTD of Pknox1 with Meis1a converts PKNOX1 into a collaborating oncogene with HOXA9, with similar latency and phenotype to HOXA9+MEIS. Supporting a role in transactivation is similar activity of PKNOX-VP16 where the CTD of PKNOX is replaced with VP16 (Bisaillon et al. 2011). That the CTD of MEIS1 can convert PKNOX into an oncoprotein, and that lack of transactivating and  34 transformation capacity in C-terminal Meis1 mutants, argues that a key role for MEIS1 is regulation of gene expression in the context of Hox-mediated leukemogenesis.   Roles for Meis1 outside of hematopoiesis While a large bulk of study into MEIS1 function has been in the context of mammalian development and hematopoiesis, studies have implicated MEIS1 in a diverse array of other processes. In concert with PBX partners, Meis1 or homologs have been implicated in regulation of neural development of the hindbrain and eye in Drosophila, Xenopus, Zebrafish and mammalian systems (Pai et al., 1998; Waskiewicz et al., 2001; Zhang et al., 2002; Hisa et al., 2004; Azcoitia et al., 2005; Heine et al., 2008; Bessa et al., 2008). In the developing eye, regulation of cyclin D1 and c-myc by Meis1 is required to maintain proliferation of the multipotent cells of the early eye (Bessa et al., 2008). SOX3, one of the earliest neural markers in vertebrates is directly bound and regulated by Pbx1/Meis1 binding (Mojsin & Stevanovic, 2009). Pbx/Meis heterodimer-regulated expression also appears to be important in the developing heart as haploinsufficiency for Pbx1 (via reduction of Pbx2 or Pbx3) or loss of Meis1 leads to phenotypically similar abnormalities of heart development, resembling tetralogy of Fallot (Stankunas et al., 2008).  Outside the embryo, there is evidence that Meis1 expression in the maternal endometrium is required for embryo implantation via integrin expression (Xu et al., 2008). Expression of Meis1 is high in adult human endometrium, myometrium and cervical tissues as well as in ovarian cancer samples (Crijns et al., 2007).  Other tumor tissues may require Meis1 to varying extents. Down-regulation of TGF-β type II receptor occurs in a number of lung cancers, resulting in loss of function of the tumor suppressor TGF-β. Meis1 was found  35 to bind the TGF-β type II receptor promoter to repress gene transcription (Halder et al., 2011). In neuroblastoma, both expression and the chromosomal loci of Meis1 are amplified (Jones et al., 2000; Spieker et al., 2001). In conjunction with the developmental regulator OCT-1, MEIS1 and other TALE homeodomain transcription factors may regulate neuronal specific expression of gonadotropin-releasing hormone in the adult hypothalamus (Rave-Harel et al., 2004). Together, these studies suggest a far larger role for Meis1 beyond the hematopoietic system in regulating gene programs that may be subverted in oncogenesis.   Knowledge of regulation of Meis1 expression is relatively limited  Although Meis1 is a key regulator of gene expression in many processes, as of yet, little is known about regulation of its expression. Studies in zebrafish suggest there is at a minimum 13 highly conserved cis-regulator regions with enhancer activity and differing degrees of tissue specificity (Royo et al., 2012, Zhou et al.,2013). Evidence of multiple cis-regulator regions in the human MEIS1 gene are also emerging (Xiang et al., 2010; Zhou et al., 2013). In the immortalized myeloid leukemia line K562, deletion of E-twenty six (ETS) binding sites in a reporter construct for the MEIS1 promoter abrogates reporter gene expression, whereas mutation of C/EBPα, SRF or RUNX1 binding sites have minimal impact (Xiang et al., 2010). Binding of ETS family member ELF1 was observed in myeloid leukemia lines, primary samples and cord blood at the native MEIS1 promoter in these studies, supporting a role for ELF1 in transcriptional regulation of MEIS1 expression. Interestingly, ETS-1 is required for MEIS1 regulated expression of Pf4, suggesting both regulation and co-operative expression with ETS family members. The MEIS1 promoter additionally contains a partial cyclic AMP response element (CRE) that may respond to  36 increased levels of transcriptional activator CREB (cAMP responsive element binding protein 1). Elevation of CREB levels in AML samples leads to a 40-fold increase in MEIS1 expression, although direct binding to the MEIS1 promoter was not demonstrated (Esparza et al., 2008). Elevation in CREB, and hence MEIS1 expression, may be one of the oncogenic mechanisms in AML as expression is lower in patients in remission and marrow of healthy donors (Crans-Vargas et al., 2002).   A complex circuit of HOX, PBX and MEIS family member co-regulation may additionally be crucial for appropriate MEIS1 expression (Zhou et al., 2013). In AML patient samples there is a linear correlation between HOXA9 and MEIS1 mRNA expression, which is mirrored in normal hematopoietic progenitor populations (Hu et al., 2009). Hoxa9-/- mutant bone marrow shows a profound reduction in Meis1 expression that can be partially rescued by Creb1 expression, a direct target of HOXA9 and direct regulator of Meis1 expression (Hu et al., 2009; Esparza et al., 2008). Interestingly, Hoxa9 expression does not appear to be involved in the up-regulation of Meis1 in megakaryocytic progenitors, further enforcing Hox-independent roles for MEIS1 (Hu et al., 2009).  MEIS family member Pknox1 also rescues Meis1 expression in Hoxa9 deficient bone marrow and appears to be crucial for Meis1 expression during embryogenesis (Hu et al., 2009; Ferretti et al., 2006). Pknox1-/- embryos exhibit a similar phenotype to Meis1-/- mutants, that is, embryonic lethality due to anemia, impaired angiogenesis and retinal anomalies. Pknox1-deficient embryos additionally show reduction in MEIS1, PBX1 and PBX2 protein levels and concomitant DNA binding activity. The authors suggest this supports Pknox1 as a master regulator of Meis1 expression, however, this is not consistent with later embryonic lethality in Pknox1-/- embryos (17.5 dpc) compared to Meis1-/- embryos  37 (11.5 – 14.5dpc) (Hisa et al., 2004; Azcoitia et al., 2005). Epigenetic changes are also crucial for regulation of Meis1 expression. In the context of MLL leukemias, histone H3 methyltransferase DOT1 is required for MLL transformation, in part by increasing H3K79 methylation around Hoxa9 and Meis1 promoters and hence expression (Chang et al., 2010).   Meis1 is a powerful player in the context of normal and malignant hematopoiesis, however fundamental characteristics remain unknown  While Meis1 has been extensively studied in a number of contexts, many questions remain. The extent to which other members of the MEINOX family (Meis2, Meis3, Pknox1 and Pknox2) are expressed in appreciable amounts in the hematopoietic hierarchy is not well studied. The expression level of these members is of interest in light of the possibility of functional redundancy between the related genes.  In addition, while Meis1 is clearly required for embryonic hematopoiesis and viability, requirements in adult mammalian hematopoiesis remain unclear. While the maintenance of HSC in Meis1-/- mice is compromised, the extent to which the defining property of the HSC, that is self-renewal, has not been examined. In addition, little evidence exists as to how Meis1 influences megakaryocytic or erythroid differentiation in the adult mammalian system. While Meis1 is implicated as a critical regulator of leukemia and hematopoiesis, very few effectors of this role have been validated to date. As previous studies of other transcription factors, such as MLL and AML1/RUNX1, have demonstrated, knowledge in either the normal or leukemic context may be highly informative in the other.   The work presented in this thesis sought to build upon research into Meis1 function in several ways. Firstly, we validated two conditional models of Meis1 deletion that we then exploited to investigate some of the most pressing questions regarding Meis1 function. We  38 additionally examined MEINOX family expression in sorted hematopoietic fractions to compile a complete picture of MEINOX expression through the hierarchy. Once robust deletion in vitro and in vivo was established, we used the model to examine the requirement for Meis1 in adult hematopoiesis. Our results support a role for Meis1 in HSC self-renewal and additionally in maintenance of megakaryocyte and erythroid potential. HSC lacking Meis1 fail to expand in vivo while lineage differentiation remains largely intact. Erythropoiesis is impaired in adult mice lacking Meis1 as erythroid progenitors fail to expand in response to stress. Our studies also provide clarification in the adult mammalian system that Meis1 is required for both megakaryopoieis and erythropoiesis as opposed to favoring differentiation along either lineage as seen in the embryonic and zebrafish models of Meis1 deletion and overexpression. We additionally identified several genes that may be relevant to Meis1 function in adult hematopoiesis by expression analysis of HSC-enriched populations following Meis1 deletion. Building on recent studies highlighting an interplay between ROS and Meis1 function in the HSC, we used an in vivo model of ROS scavenging to examine if this mechanism extends to other hematopoietic populations requiring Meis1 expression, including megarkaryocyte and erythroid progenitors.  We additionally examined if expression of our candidate effectors of Meis1 function in the HSC are altered by ROS-scavenging in both HSC and megakaryocyte progenitor-enriched populations. In summary, the work validates a powerful tool for the study of Meis1 function and provides novel insights into roles in the adult HSC and differentiated progenitor populations.    39 Chapter 2 : Materials & Methods  In vivo methods Mice All mice were bred and maintained at the British Columbia Cancer Research Centre Animal Resource Centre with all protocols approved by the University of British Columbia Animal Care Committee (Certificate: A13-0063). Male Meis1tmloxP/+ (Meis1fl/+) mice on the CD45.2 C57BL/6 J (B6) background were a generous gift from Drs N. Jenkins and N. Copeland. Little information was available at the time with respect to the exact method of engineering the mice or targeted sequence. The CD45.2 background was confirmed by FACS analysis of peripheral blood (see below), following which Meis1fl/+ mice were bred with female B6 mice and the resulting offspring interbred. Meis1fl/fl or Meis1fl/+ mice were then bred with B6;129-Gt(ROSA)26Sortm1(Cre/ERT)Nat/J (ERTCre; Gift from Dr. A. Weng) or B6.Cg-Tg(Mx1-Cre)1Cgn/J (MxCre; Jackson Laboratory) for inducible expression of Cre recombinase and subsequent deletion of Meis1 exon 8 (MxCre/Meis1 or ERTCre/Meis1). The ERTCre strain was engineered to express Cre recombinase with a modified estrogen receptor ligand binding domain from the ubiquitous Gt(ROSA)26Sor promoter.  In the absence of the synthetic ligand tamoxifen (4-OHT), the Cre protein remains cytoplasmic, however administration of 4-OHT allows localization to the nucleus and recombinase activity (Badea et al., 2003). The MxCre system is based on controlled expression of the Cre recombinase as opposed to localization. Normally, the Mx1 promoter is expressed in response to interferon signaling triggered by double-stranded RNA virus infection. Expression of the Cre gene in MxCre mice is driven by the Mx1 promoter when the synthetic double-stranded RNA analog  40 polyinosinic:polycytidylic acid (PolyI:C) is administered, or alternatively interferon alpha or beta (Kühn et al., 1995). Progeny of the MxCre/Meis1 or ERTCre/Meis1 crosses were further interbred to yield the described genotypes. Wild-type Meis1 alleles are denoted as Meis1+, while the floxed allele prior to Cre recombinase expression is termed Meis1fl. Following expression of Cre recombinase, the allele with excised sequence intervening the LoxP sites is designated as Meis1-. For example, a mouse heterozygous for the targeted Meis1 allele on the ERTCre background following Cre expression would be denoted as ERTCre/Meis1-/+. Investigation of the targeted sequence and validation of Cre-mediated deletion is described in Chapter 3.  For transplantation assays, wild-type B6.SJL-PtprcaPeb3b/BoyJ (Peb3b, Animal Resource Centre (ARC), BC Cancer Research Centre, Vancouver, BC) were used as transplant recipients such that donor cell contribution in the peripheral blood of mice could be monitored by differential expression of CD45 antigen isoforms. Mice on the B6 background express CD45.2 on hematopoietic cells, whereas Pep3b mice express the CD45.1 isoform. Bone marrow from Peb3b mice were used to assay MEINOX family expression in sorted cell fractions as well as for cell line studies.  Induction of Cre recombinase expression/localization in vivo In vivo deletion of Meis1 exon 8 was achieved via two methods employing the Cre/LoxP system. Induction of Cre expression was achieved in MxCre/Meis1 mice by intraperitoneal (IP) injection of 300µg polyinosinic:polycytidylic acid (poly I:C; VWR/EMD Biosciences, Radnor, PA, USA), dissolved in phosphate buffered saline (PBS; STEMCELL Technologies Inc., Vancouver, BC, CAN) per mouse every 48 hours for 18 days. To induce deletion in the ERTCre/Meis1 mice, 4-hydroxytamoxifen (4-OHT; Sigma, St. Louis, MO,  41 USA) was first dissolved at 25mg/mL into filtered >99% ethanol (EtOH) and homogenized with a hand-held homogenizer. This was further diluted into autoclaved corn oil (Sigma) to achieve a 5mg/mL 4-OHT suspension. 1mg 4-OHT was injected IP into ERTCre/Meis1 mice every 48 hours for 12 days to trigger Cre expression.  Phenylhydrazine induced model of hemolytic anemia  Phenylhydrazine treatment is an experimental model for hemolytic anemia (Capron et al, 2011) and was used to examine the capacity of Meis1-/- erythroid-progenitors to expand in response to stress. At 48-hours following the last of nine PolyI:C injections, MxCre/Meis1-/+ or MxCre/Meis1-/- mice were given IV injections of phenylhydrazine hydrochloride (PHZ; Sigma) at 40mg/Kg. Four days later, mice were euthanized and analyzed for phenotype and CFC (described below) capacity.  Reactive oxygen species scavenging by in vivo N-acetyl-L-cysteine treatment  Very recent work implicates Meis1 in the regulation of hypoxia-inducible factor (Hifs) regulation and that this may be one mechanism through which reactive oxygen species (ROS) are regulated in the HSC. In order to assess the impact of ROS on the phenotypic anomalies and putative target genes we identified in MxCre/Meis1-/- mice, we used in vivo N-acetyl-L-cysteine (NAC) treatment. NAC has several functions in cells, but the ROS scavenger function in cells by the provision of substrate for H2O2 decomposition pathways (Yang et al., 2007) has been postulated to be of importance in the study of Meis1 function (Kocabas et al., 2012; Unnisa et al, 2012). Cre expression in MxCre/Meis1 mice was induced with PolyI:C according to the protocol outlined above. Following 2 PolyI:C injections (4 days), mice were then also given daily subcutaneous (SQ) injections of 100mg/Kg NAC  42 (Sigma) for 14 days. Mice were analyzed 5 to 7 days following the final PolyI:C/NAC injections.  Isolation of bone marrow and peripheral blood for analysis  Peripheral blood parameters following induction were monitored using the tail prick method and collection into heparinized capillaries. For peripheral blood complete blood count and differential, blood was transferred to ETDA-coated blood collection tubes (microtainer 365973, BD, Franklin Lakes, NJ, USA) and counts performed by a scil Vet abc automatic blood cell counter (Vet Novations, Viernheim, Germany). If mice were euthanized prior to blood collection, blood was collected from the heart of euthanized mice.  Bone marrow was isolated from euthanized mice by flushing the marrow cavities of femurs, tibias and iliac crests with 2% FBS in PBS using an 18G to 24G needle.  Long-term repopulating cell (LTRC) and competitive repopulating unit (CRU) assays Bone marrow was flushed from the tibias, femurs and iliac Crests of mice in 2% PBS with fetal bovine serum (FBS; STEMCELL Technologies Inc.) and nucleated cell counts performed using 3% v/v acetic acid with methylene blue. If no further manipulation (such as cell sorting) was required prior to transplantation, appropriately diluted Ly5.2+ cell suspensions were made and transplanted into recipient Ly5.1+ Pep3b mice irradiated at 780 cGy using an X-ray source. Whole Pep3b bone marrow was used as helper or competitor cells as indicated in the results.  Engraftment of test cells into recipients was monitored by peripheral blood collection from the tail-vein of recipient mice at various intervals. Approximately 50µL of peripheral blood was collected, lysed (Pharmlyse, BD) and incubated at 4oC with a combination of fluorochrome conjugated anti-mouse antibodies against CD45.2-FITC (anti-Ly5.2), CD4-PE,  43 CD8-PE, B220-PE, B220-APC-Cy7, Gr1-APC-Cy7 and Mac1-APC-Cy7 (see Table 2 for clones and suppliers). Samples were washed and re-suspended in 2% PBS and 1ug/mL propidium iodide (PI) prior to acquisition on a modified FACSCalibur (BD with Cytek laser and digital detector upgrades, Fremont, CA, USA). Analysis of FACS data was performed using FlowJo analysis software (TreeStar, Ashland, OR, USA). Mice were scored as positive for multi-lineage engraftment if >1% of total peripheral blood nucleated cells were donor-derived Ly5.2+ cells and within this population, the contribution of T cells (CD4CD8+), B-cells (B220+) and myeloid cells (Gr1+Mac1+) to was >1%. Long-term repopulating cell frequency and test for significant differences between groups was performed using the Extreme Limiting Dilution Analysis (ELDA) online tool (http://bioinf.wehi.edu.au/software/elda).   Cell sorting and FACS analysis  Isolation of retrovirally transduced producer cells and 5-FU pre-treated marrow  Green and yellow fluorescent proteins (GFP, YFP) are useful tools to monitor gene transfer efficiency following retroviral transduction (construct and infection details below). GFP and YFP expression is linked to that of the gene of interest through an internal ribosome entry site (IRES), thusly allowing monitoring of expression and sorting by expression for the gene of interest. We used FACS to sort both viral producer cells and retrovirally transduced 5-flurouracil (5-FU). Cells were harvested from culture, then washed with 2% PBS and prepared for sorting by staining with propidium idodide (PI, Sigma) for viability.  An Influx I (Cytopia/BD Biosciences) or FACSDIVA (BD Biosciences) machine with a 488nm argon  44 laser source, was used to sort cells on the basis of viability (PI negative) and YFP (527nm emission) and/or GFP positivity (509nm emission).  Isolation and phenotypic analysis of primary marrow progenitor and mature cell populations For phenotypic analysis and isolation of purified hematopoietic populations, the following isolation and staining protocols were used. MxCre/Meis1 mice were analyzed 5-7 days following the final PolyI:C injection as trials at earlier time-points yielded FACS profiles with indistinct separation between positive and negative populations. Sorting and analysis of ERTCre/Meis1 mice was done 2-4 days following the final 4-OHT injection.  Bone marrow was flushed from the tibias, femurs and iliac crests of mice in 2% PBS with FBS. Red blood cells were lysed with PharmLyse (BD Biosciences) reagent according to the manufacturers instructions. A viable cell count was then performed using 0.4% trypan blue in PBS. Unless the cells were destined for a myeloid progenitor sort (common myeloid progenitor (CMP), granulocyte-monocyte progenitor (GMP) or megakaryocyte-erythroid progenitor (MEP)), cells were blocked for 20 minute on ice in 5% rat sera and 1µg/1x106 cells Fc receptor (FcR, also known as CD16/32), then washed with 2% FBS in PBS. Cells were then incubated with the antibody stains outlined in Table 2.1 for 20 min on ice (BD Pharmingen; eBioScience, San Diego, CA USA; BioLegend, San Diego, CA USA, STEMCELL Technologies). Cells were then washed and re-suspended at roughly 1x107 cells/mL with 3µM 4',6-diamidino-2-phenylindole (DAPI, Life Technologies, Carlsbad, CA, USA)) for viability and 10µg/mL DNaseI solution (STEMCELL Technologies) to prevent clumping. Cells were then filtered through a 45µm filter prior sorting. An Influx II or Aria cytometer with 488nm argon, 350/405nm UV and 634nm red diode laser sources were used for cell sorting and phenotyping from primary marrow. Cells  45 were sorted/analyzed according to the gating strategies outlined in Table 2.2. Annexin V staining for apoptosis and BrDU for cell cycle staining were done according to the kit manufacturer’s instructions (BD Biosciences). Following sorting, purity of the sorted sample was assessed by running a small fraction of the sorted sample through the cytometer. Purity of >98% was deemed to be acceptable. Analysis of FACS data was performed using FlowJo analysis software (TreeStar). Unpaired two-tailed, Student’s T-tests were perfomed to determined statistical significance between Meis1-/- and control samples.  Table 2.1: Dilution, clone and source of antibodies for FACS phenotyping and cell sorting PURPOSE ANTIBODY CLONE DILUTION (FINAL) SUPPLIER Donor cell and lineage distribution of mouse peripheral blood CD45.2-APC 104 1:1000 BD Pharmingen CD4-PE L3T4 1:1600 eBioscience CD8-PE Ly-2 1:1600 eBioScience B220-PE RA3-6B2 1:2000 BD Pharmingen B220-APC-Cy7 RA3-6B2 1:2000 BD Pharmingen Gr1-APC-Cy7 RB6-8C5 1:6000 BD Pharmingen Mac1-APC-Cy7 M1/70 1:5000 BD Pharmingen  ESLAM HSC for expression profiling (Kent et al., 2009) CD45-FITC 30-F11 1:100 BioLegend EPCR-PE RMEPCR1560 1:100 STEMCELL Technologies CD48-APC HM48-1 1:100 BioLegend CD150-biotin TC15-12F12.2 1:100 BioLegend SA-PE-TxRed Streptavidin 1:400 BD Pharmingen CD150 HSC for phenotyping (Kiel et al., 2005) Lineage-PerCP-Cy5.5 See below See below See below cKit-APC 2B8 1:100 BD Pharmingen Sca-PE E13-161.7 1:100 BD Pharmingen CD48-FITC HM48-1 1:100 BioLegend CD150-PE-Cy7 SLAM 1:100 BioLegend LSK for Affymetrix expression analysis Lineage-PerCP-Cy5.5 See below See below See below cKit-APC 2B8 1:100 BD Pharmingen Sca-PE E13-161.7 1:100 BD Pharmingen Myeloid Lineage-PerCP- See below See below See below  46 PURPOSE ANTIBODY CLONE DILUTION (FINAL) SUPPLIER progenitors for expression and phenotyping (Akashi et al., 2000) Cy5.5 Sca1-PE-Cy7 E13-161.7 1:150 BD Pharmingen cKit-APC 2B8 1:100 BD Pharmingen CD34-FITC RAM34 1:400 eBioscience CD16/32-PE 2.4G2 1:300 BD Pharmingen Common lymphoid progenitors (CLP) for expression and phenotyping (Kondo et al., 1997) Lineage-FITC See below See below See below cKit-APC 2B8 1:100 BD Pharmingen Sca1-PE E13-161.7 1:100 BD Pharmingen CD127-biotin B12-1 1:100 BD Pharmingen SA-PerCP-Cy5.5 Streptavidin 1:100 BD Pharmingen Megakaryocyte progenitors for expression and phenotyping (Pronk et al., 2007) Lineage-PerCP-Cy5.5 See below See below  See below Sca1-PE E13-161.7 1:200 BD Pharmingen cKit-APC 2B8 1:100 BD Pharmingen CD150-PE-Cy7 SLAM 1:100 Biolegend CD41-FITC MWReg30 1:100 BD Pharmingen Mature megakaryocytes (Heazlewood et al., 2013) CD41 (as per Nilsson lab)    Erythroblast maturation series for expression profiling *unlysed (Socolovsky et al., 2001) Ter119-PerCP-Cy5.5 TER-119 1:150 BD Pharmingen CD71-PE C2 1:100 BD Pharmingen Lineage-FITC for expression profiling and phenotyping of CLP Gr1-FITC RB6-8C5 1:2400 BD Pharmingen Ter119-FITC TER-119 1:150 BD Pharmingen B220-FITC RA3-6B2 1:600 BD Pharmingen CD3-FITC L3T4 1:300 BD Pharmingen CD4-FITC Ly-2 1:600 BD Pharmingen CD8a-FITC 53-6.7 1:600 BD Pharmingen Lineage-PerCP-Cy5.5 for expression profiling and phenotyping of HSC and myeloid Gr1-PerCP-Cy5.5 1A8 1:2400 BD Pharmingen Ter119- PerCP-Cy5.5 TER-119 1:150 BD Pharmingen B220- PerCP-Cy5.5 RA3-6B2 1:600 BD Pharmingen CD3- PerCP- L3T4 1:300 BD Pharmingen  47 PURPOSE ANTIBODY CLONE DILUTION (FINAL) SUPPLIER lineages Cy5.5 CD4- PerCP-Cy5.5 Ly-2 1:600 BD Pharmingen CD8a- PerCP-Cy5.5 53-6.7 1:600 BD Pharmingen Mature lymphoid for expression profiling B220-APC-Cy7 RA3-6B2 1:500 BD Pharmingen CD4-PE L3T4 1:400 eBiosciences CD8-PE Ly-2 1:400 eBiosciences Mature myeloid for expression profiling (Song et al., 2005) Gr1-APC-Cy7 RB6-8C5 1:500 BD Pharmingen Mac1-FITC M1/70 1:800 BD Pharmingen Apoptosis AnnexinV-PE  kit instructions BD Pharmingen Cell Cycle BrDU-APC  kit instructions BD Pharmingen  Table 2.2: Cell sorting and phenotyping gating strategies.  Note all samples first gated on SSC-A/FSC-A for size and complexity, DAPI for viability and FSC-A/FSC-H for singlets Populations of Interest Subpopulations assessed Enrichment Gate First Purity Gate Second Purity  Gate LSK HSC for sorting and Affymetrix expression analysis n/a Lin-  cKit+Sca1+  ESLAM HSC for expression profiling n/a Generous first and second purity gates CD45+EPCR+ CD48-CD150+ CD150 HSC for phenotyping n/a Lin- cKit+Sca1+ CD48-CD150+ Myeloid Progenitors for expression and phenotyping CMP Lin- cKit+Sca1- CD16/32lowCD34+ GMP Lin- cKit+Sca1- CD16/32hiCD34+ MEP Lin- cKit+Sca1- CD16/32lowCD34low Common lymphoid progenitors n/a Lin-IL7R+ (generous) IL7R+ cKitmidSca1mid  48 Populations of Interest Subpopulations assessed Enrichment Gate First Purity Gate Second Purity  Gate for expression and phenotyping Megakaryocyte progenitors for expression and phenotyping  n/a Lin- cKit+Sca1- CD150+CD41+ Erythroblast maturation series for expression profiling Proerythroblasts & basophilic erythroblasts n/a Ter119+CD71hi  Late basophilic and chromatophilic erythroblasts n/a Ter119+CD71mid  Orthochromatophilic erythroblasts n/a Ter119+CD71lo  Lineage defined T-cells  n/a CD4+CD8+  Mature B-cells  n/a B220+  Mature Myeloid Granulocyte precursor  Gr1+Mac1+  Granulocyte  Gr1-Mac1+  Mature megakaryocytes  Lin- CD41+SSChi         49 Table 2.3: References for Sorting Gates Lineage refers to the lineage cocktail referred to as Lin- (IL7R is excluded in the CLP stain) Stain Markers Reference Lineage Gr1, Mac1, Ter119, B220, CD3, CD8 (IL7R) -  ESLAM HSC CD45modEPCR+CD48-CD150+ Kent et al., 2009 CD150 HSC Lin-cKit+Sca1+CD48-CD150+ Kiel et al., 2005 CMP Lin-cKit+Sca1-CD16/32lowCD34+ Akashi et al., 2000 GMP Lin-cKit+Sca1-CD16/32hiCD34+ Akashi et al., 2000 MEP Lin-cKit+Sca1-CD16/32lowCD34low Akashi et al., 2000 MkP Lin-cKit+Sca1-CD150+CD41+ Pronk et al., 2007 CLP Lin-IL7R+cKitmidSca1mid Kondo et al., 1997 Immature Erythroblast Ter119+CD71hi Socolovsky et al., 2001 Maturing Erythroblast Ter119+CD71mid Socolovsky et al., 2001 Mature Erythroblast Ter119+CD71low Socolovsky et al., 2001 Granulocyte Precursor Gr1+Mac1+ Song et al., 2005 Granulocyte Gr1-Mac1+ Song et al., 2005 Mature Megakaryocyte CD41+SSChi Heazlewood et al., 2013   In vitro studies Retroviral vectors and transduction  Stable ecotropic retroviral producer lines for transduction of mouse bone marrow cells, were derived from GP+E-86 packaging cells following infection with virus transiently produced from Phoenix Ampho (PA) packaging cells as previously described (protocol in Kalberer, Antonchuk & Humphries, 2002). In brief, PA cells were transfected with DNA constructs using a calcium phosphate-based system (CellPhect, GE Healthcare, Little Chalfont, Buckinghamshire, UK) for the retroviral constructs of interest  (MSCV-CRE-PURO, pSF91-MN1-GFP, MSCV-ND13-GFP, MSCV-MEIS1a-YFP, MSCV-MEIS1b-YFP or MSCV-MEIS251-YFP). MSCV-Meis1b-YFP and MSCV-Meis251-YFP were generated by cloning from cDNA using primers specific to the region with additional restriction enzyme sequence for compatibility  50 into the existing MSCV-Meis1a-YFP vector. In the case of MEIS251, a stop codon was introduced into the sequence at amino acid 251 in the MEIS1 sequence. Other viral vectors are described in Heuser et al., 2007 (pSF91-MN1-GFP), Pineault et al., 2003 (MSCV-ND13-GFP, MSCV-Meis1a-YFP). Amphotropic virus containing supernatant was then harvested 12 hours later, filtered with a 0.45µm filter to remove cellular debris and transferred to dishes containing GP+E86 cells ecotropic packaging cells and 5µg/mL protamine sulfate. The cycle of removing the media from GP+E86 cells and replacing with PA supernatant containing viral particles was repeated every 12 hours for 2 days. Following this process, virally transduced GP+E86 cells were selected by fluorescence activated cell sorting (FACS) (for GFP or YFP expressing viruses) or by antibiotic selection (1.6 – 3.2µg/mL puromycin for MSCV-CRE-PURO, Life Technologies, Carlsbad, CA, USA).  To transduce primary mouse marrow, Peb3b or C57BL/6J mice were pre-treated with 5-FU (150mg/Kg by intravenous (IV) injection; Hospira, Montral, QC, CAN) and 4 days later, bone marrow harvested (Van Zant, 1984). Red blood cells were lysed (Pharm Lyse, BD Biosciences, Franklin Lakes, NJ, USA) and stimulated for 2-days in culture in medium containing 15% fetal bovine serum (FBS), 6ng/mL mouse Interleukin 3 (mIL-3, STEMCELL Technologies, Vancouver, BC, CAN), 10ng/mL human IL-6 (STEMCELL Technologies) and 100ng/mL mouse stem cell factor (mSCF, STEMCELL Technologies) (cytokine cocktail 3/6/SCF) in Dulbecco’s Modified Eagle’s Medium (DMEM, STEMCELL Technologies). After pre-stimulation, bone marrow cells were harvested and plated onto irradiated GP+E86 viral producer cells (40Gy from an X-ray source) for 48 hours culture in growth factor containing media as above supplemented with 5µg/mL protamine sulfate (Life Technologies). Non-adherent and loosely adherent bone marrow cells were then recovered  51 transduced cells selected by FACS or antibiotic selection (MSCV-CRE-PURO). Transduced cells were then maintained in 15% FBS with the mIL-3/hIL-6/mSCF cocktail in DMEM.  Hematopoietic colony-forming cell (CFC) assays In vitro assays of colony forming cell (CFC) content are used as a surrogate to measure the capacity and frequency lineage committed hematopoietic progenitors in a test cell population, from culture or a primary cell source (reviewed in Purton & Scadden, 2007). We used CFC assays to examine the frequency of muti-potent (CFU-GEMM), myeloid (CFU-GM), erythroid (BFU-E) and megakaryocyte (CFU-Mk) progenitors from freshly euthanized experimental animals. The assay was also used to examine if over-expression of the Meis1 truncation (MEIS251) transcript had a gross impact on cell proliferation.  BM cells were flushed from the tibias, femurs and iliac crests of mice in 2% FBS/PBS while spleen cells were isolated by maceration through a 0.2µM screen. Nucleated cell counts were performed and cells further diluted in 2% FBS in Iscoves Modified Dulbecco Medium (2%FBS/IMDM, STEMCELL Technologies). Bone marrow or spleen cells were mixed with methylcellulose media containing cytokines to support CFU-GEMM, CFU-GM, and BFU-E colony growth (CAT: M3434, STEMCELL Technologies) according to manufacturer’s protocol at a concentration of 2x104 cells per dish for bone marrow or 2x105 cells per dish for spleen.  As this cocktail supports myeloid colony growth that can overwhelm the less frequent erythroid progenitor, an alternate methylcellulose media optimized for BFU-E growth (CAT: M3436, STEMCELL Technologies) was also used. For the erythroid colony cultures, 4x104 bone marrow or 4x105 spleen cells per dish was used.  Cells were cultured for 10-12 days in a humidified incubator at 37oC and 5% CO2 for  52 myeloid-biased cultures and 14 days for erythroid-biased cultures. CFC morphology and numbers were assessed using an inverted microscope. CFU-Mk are difficult to detect using in situ morphological assessment and are therefore more accurately enumerated by detection in collagen gels that have been dehydrated, fixed and treated with a cytochemical stain for acetylcholinesterase enzyme activity (Shivdasani & Schulze, 2005, manufacturer’s instructions at www.STEMCELL.com). Bone marrow was plated at a concentration 1x105 cells per culure slide into collagen-based media suuplemented with cytokines and lipids (CAT 4964, STEMCELL, Technologies) according to manufacturer’s instructions.  Cultures were incubated at 37o 5% CO2 and >95% humidity for 7 days, then fixed using acetone following dehydration with Whatman paper. Slides were then stained (acetylthiocholiniodide, sodium citrate, copper sulfate, and potassium ferricyanide - all reagents SIGMA) according to manufacturer’s protocol to identify megakaryocytes on the basis of acetylcholinesterase activity (STEMCELL Technologies). Mouse megakaryocytes and early megakaryocyte progenitors, which express acetylcholinesterase, have brown granular deposits of copper ferrocyanide in the cytoplasm resulting from the enzymatic reaction.  Granules may appear light red-brown in cells with low acetylcholinesterase content and ranges from orange-brown to dark brown/ black in cells with high acetylcholinesterase content. Harris hematoxylin solution (Sigma) was used as a counterstain for cell nuclei. CFU-Mk colonies range in size from three to approximately 50 megakaryocytes per colony, mixed CFU-Mk contain megakaryocytes and cells of granulocyte/macrophage lineages. Non-Mk colonies containing cells with stained nuclei only are also present.  53 For retroviral overexpression studies to determine if MEIS1 truncation transcript (MEIS251) had a dominant negative effect, CFCs with potential self-renewal capacity were detected using re-plating assays. 2000 retrovirally transduced (see below) bone marrow cells were plated immediately after the infection procedure dish in myeloid-biased methylcellulose media for 7 days at 37oC and 5% CO2. Following enumeration of colony number, the entire culture was then harvested by rinsing the plate gently with 4x 1 mL volumes of 2% FBS/IMDM. Pooled cells were washed twice by centrifugation at 300 g and replated (1500 cells per dish) onto secondary methylcellulose plates. Two rounds of replating were performed for each experiment. For all colony forming assays, unpaired two-tailed, Student’s T-tests were perfomed to determined statistical significance between Meis1-/- and control samples. Long-term culture initiating cell (LTC-IC) assay The in vitro long-term culture initiating cell (LTC-IC) assay, detects and quantifies a subset of primitive hematopoietic cells  (termed LTC-IC) based on their capacity to continuously produce myeloid cells for ≥ 4 weeks when cultured on a suitable feeder layer (Collins & Dorshkind, 1987). This assay consists of two steps: The first step is to co-culture test cells on a supportive feeder layer in a limiting dilution assay for 4 to 6 weeks to allow the differentiation of less primitive hematopoietic cells (present in the input cell suspension), while maintaining or expanding LTC-IC numbers. The second step is to detect LTC-IC-derived myeloid hematopoietic progenitors using the CFC assay. The frequency of LTC-IC is determined using Poisson statistics and method of maximum likelihood.  In the LTC-IC studies, S17 stromal line cells (Collins & Dorshkind, 1987) were irradiated at 2000cGy and seeded at 1.5x104 cell per well into a flat-bottom tissue culture  54 treated 96-well plate. Cultures were then seeded with un-separated BM cells at various concentrations (3x104, 1.5x104, 7.5x103, 3.75x103) from induced ERTCre/Meis+/-, ERTCre/Meis-/-, MxCre/Meis+/- or MxCre/Meis-/-mice. Cells were prepared in mouse MyeloCult™ M5300 (12.5% horse serum, 12.5% FBS, 0.2 mM i-inositol, 20 mM folic acid, 10-4 M 2-mercaptoethanol, 2 mM L-glutamine, STEMCELL Technologies) supplemented with freshly prepared 10-6 M hydrocortisone (21-hemisuccinate sodium salt)(mLTCM). Weekly one-half media changes were performed according to manufacturers protocol. Following 4 weeks in culture, adherent and non-adherent cells were harvested from individual wells and plated into methycellulose media supplemented with cytokines (CAT:3434, STEMCELL Technologies), then cultured as described for CFC assays. The number of CFCs was counted and the wells were recorded as negative if no CFC were present and positive if ≥ 1 CFC is present. LTC-IC frequency and test for significant differences between groups was performed using the Extreme Limiting Dilution Analysis (ELDA) online tool (http://bioinf.wehi.edu.au/software/elda).   Molecular methods Southern blot analysis Southern blot analysis to detect Cre-mediated deletion of Meis1fl/fl was carried out using previously described procedures (Pawliuk et al., 1994). In brief, bone marrow cells from Meis1fl/+mice were infected with MSCV-CRE-PURO retrovirus and treated with puromycin at 0, 1.6 or 3.6 µg/ml in vitro for 48-hours. Genomic DNA was extracted using DNAzol Reagent (Life Technologies). 10 µg of genomic DNA was digested with either  55 HindIII , EcoRI or BglII for 16 hrs at 37˚C and electrophoresed at 30V on a 1.0 % agarose/TAE gel.  Gels were treated with 0.1 M HCl for 8 minutes, rinsed with deionized H2O, treated with 1.5M NaCl/ 0.5N NaOH for 30 minutes and capillary transferred to ZetaProbe-GT (Bio-Rad) with 10x SSC for 16 hrs. DNA was fixed to membranes by treating @ 80˚C for 1 hour. Membranes were prehybridized 2 hours at 65˚C in buffer containing 0.8% skim milk, 8% Dextran Sulfate, 5X SSC, 8% Formamide, 0.8% SDS, 1.6 mM EDTA pH 8.0, and 400 µg/ml sheared and boiled Salmon Sperm DNA.  DNA probes were synthesized by PCR and gel purified prior to labeling. Probes and anticipated hybridization bands are outlined in Table 2.3. DNA probes were labeled by random priming and incorporation with 32P dCTP , denatured and added to the pre-hybridization buffer with the membrane, incubated for 16 hrs at 65˚C, then washed 3 times at 65˚C (0.3X SSC, 0.1%SDS, 0.1% Tetra-Sodium Pyrophosphate). The hybridized membranes were then exposed to a phosphor –imaging screen for 24 hours.  The phosphor screen was scanned by STORM Phosphorimager (Molecular Dynamics, Sunnyvale, CA, USA) and analyzed with ImageQuant 5.2 software (GE Lifesciences).   Table 2.4: Probes and anticipated hybridization sizes Probe Position Forward Reverse Internal  5’ of the 3’ LoxP site 5’-gatttgatgctcttgcgaca-3’ 5’-gaagttattaggtggatccaagct-3’ External 5’ of the 5’ LoxP site 5’-ccgtggttctccaagtttgt-3’ 5’- tccatctcaaaccccttcag-3’ Probe Digest WT hybridization MUT allele hybridization Internal HindIII 7.4 kbp 2.3 kbp Internal EcoRI 3 kbp 2.1 kpb/no band with Cre  External BglII 2.3 kbp 2.4 kbp/3.8kbp with Cre    56 PCR detection of the 5’ LoxP site As described in Chapter 3, Southern blot analysis indicated that the second LoxP site was 3’ of exon 8. Primers specific for extending regions of Intron 7 were used in concert with a reverse primer specific to the LoxP site 5’ of exon 8 in an attempt to amplify the region of the mutant allele containing both LoxP sites. Detailed description of the rationale and primers are presented in Chapter 3. PCR amplification was performed with HiFi Platinum Taq Polymerase (Life Technologies). RT-PCR cloning of the 5’ LoxP site and truncated transcript To confirm the positioning of the LoxP sites, the product of PCR amplification of with primer sets described in Chapter 3 were TOPO TA cloned into the PCR2.1 vector (Life Technologies) and sequenced by McGill University Sequencing Service. The presence and position of both LoxP sites were confirmed in the non-deleted mutant allele by sequence analysis, as well as and the deletion of genomic sequence containing exon 8 between these LoxP sites leaving a single remaining LoxP site.   In order to establish if a truncated transcript was generated following in vivo Cre-mediated Meis1 deletion, splenocytes were isolated from ERTCre/Meis1fl/fl, ERTCre/Meis1+/fl, and ERTCre/Meis1+/+ mice following in vivo Cre induction by 4-OHT. Cells were lysed in TRIZOL reagent (Life Technologies), and total RNA extracted as per recommended procedure. RNA was reverse transcribed with Superscript Vilo cDNA synthesis kit (Life Technologies) and amplified with Platinum Taq Polymerase (Life Technologies) with primers specific for Meis1 exon7 (5’-TCCACTCGTTCAGGAGGAAC-3’) and Meis1 exon 11(5’-TGCTGACCGTCCATTACAAA-3’). Predicted sizes of amplicons were obtained (428bp, intact exon 8; 282 bp deleted exon 8), gel purified and TOPO-TA cloned into PCR 2.1 Vector (Life Technologies). Individual clones were  57 sequenced by McGill University Sequencing service with M13Forward and M13Reverse sequencing primers. Western blot analysis of MEIS1 protein following in vitro deletion in MN1-overexpressing cells   pSF91-MN1-GFP transduced ERTCre/Meis1fl/fl or ERTCre/Meis1fl/+ BM cells were cultured for 48 hours in 1µM 4-OHT.  1x106 cells were harvested, spun, and rinsed twice in 2% PBS. The resulting cell pellets were re-suspended in 50µL 1x phosphate solubilization buffer (PSB – 50mM HEPES, 100nM sodium fluoride, 10mM tetrasodium phosphate, 2mM sodium vanadate, 4mM EDTA, 2mM sodium molybdate) and then diluted with 1x RIPA cell lysis buffer (2x: PSB, 1% sodium deoxycholate, 1% NP-40, 5x protease inhibitor cocktail (Roche, Penzberg, Germany)) for a total volume of 100µL. The cell pellet was sonicated for 5x 30 second intervals to homogenize the solution. Samples were prepared according to the NuPAGE Bis-Tris mini-gel kit instructions (Life Technologies) and 20µL loaded onto a 10-well 4-12% Bis-Tris Gel with SeeBlue Plus2 ladder (Life Technologies). Gels were run with MES buffer at 150V for roughly 1 hour. The gel was transferred to nitrocellulose membrane according to with the XCell II blot module according to manufacturers protocol (Life Technologies).  The nitrocellulose membrane was blocked in 5% skim milk powder in 0.1% TBS-T (150mM sodium chloride, 10mM Tris-HCl pH 8, 0.1% tween 20) for 1 hour at room temperature. Rabbit polyclonal to MEIS1 antibody (ab19867, AbCam, Cambridge, UK) was added at a 1:1000 dilution in 5mL fresh 5% skim milk powder in TBS-T (0.3µg/mL) and incubated overnight at 4oC with gentle agitation. The membrane was rinsed 3 times for 15 minutes in TBS-T to remove excess primary antibody. The membrane was then incubated for 1 hour at room temperature with biotinylated goat anti-rabbit IgG (Vector, Burlington, ON,  58 CAN) at a 1:10 000 dilution (150ng/mL). Following three 15 minute washes in TBS-T, the membrane was incubated for one hour with horseradish peroxidase conjugated streptavidin (SA-HRP, Jackson ImmunoResearch, West Grove, PA) at a 1:10 000 dilution (150ng/mL). The membrane was again washed three times for 15 minutes in TBS-T with a final 15-minute rinse in PBS. Immun-Star HRP chemiluminescence reagents (Bio-Rad, Berkeley, CA, USA) were used to visualize protein levels with light-sensitive film (Kodak, Rochester, NY, USA). The membrane was then stripped with Re-Probe (GBiosciences, St. Louis, MO, USA) to asses equivalent cell loading using GAPDH. The above protocol was repeated using mouse α-GAPDH at a 1:15 000 dilution (Jackson ImmunoResearch, 100ng/mL) and donkey α-mouse-HRP at 1:10 000 (Jackson ImmunoResearch, 150ng/mL) as a secondary antibody.  Histology The sternum, spleen and, in some cases, the tibia, of mice were isolated and placed directly into 4% paraformaldehyde solution. These specimens were fixed and mounted onto slides by the Provincial Health Services Authority Histology Laboratory (Vancouver, BC, CAN). Slides were stained with hematoxylin and eosin (H&E), anti-CD45R (RA3-62B, BD Pharmingen) or anti-Ter119 antibodies (TER-119, BD Pharmingen) using the Ventana Discovery XT (Roche) machine and protocols. Both anti-CD45R and anti-Ter119 antibodies were used at a 1:50 dilution and linked to anti-rabbit HRP using a rabbit anti-rat antibody (Jackson Immuno Research) secondary at 1:500 dilution.  Q-PCR & Q-RT-PCR Genomic DNA purification from mouse bone marrow and peripheral blood for <106 cells was performed with PureLinkTM Genomic DNA Mini Kit (Life Technologies). For bone marrow samples of >106 cells, purification was performed with DNAzol Reagent (Life  59 Technologies). DNA concentration was quantified by NanoDrop and 25 ng DNA template used per reaction. Q-PCR was performed with the 7900HT Real-Time PCR System (Applied Biosystems) with FastStart Universal SYBR Green MasterMix (Roche). Table 2.4 lists the primer sets used for Q-PCR detection of genomic deletion of the floxed Meis1 alleles. Primer set NDF (non-deleted floxed) is located between the LoxP sequences and anchored in the 3’ LoxP site and is used to detect the non-deleted floxed Meis1 allele. Primer set DF (deleted floxed) has the forward primer 5’ of the 5’LoxP site and the reverse in the targeting vector sequence 3’ of the 3’ LoxP site such that only the collapsed floxed Meis1 allele is detected. Control primers sets were used to normalize and confirm genotype. Primer set Floxed CTL uses the 3’ targeting vector sequence to detect the floxed allele regardless of collapse wheras the Exon 7 CTL set detects both wild-type and floxed Meis1 alleles. Control MxCre/Meis1-/- and MxCre/Meis1fl/fl genomic DNA samples confirmed by Southern Blot analysis to be 100% and 0% collapsed, respectively, were used as calibrator samples in Relative Quantification Analysis using RQ Manager version 2.1 software (Applied Biosystems, Foster City, CA, USA). For Q-RT-PCR, RNA from the sorted populations listed above (LSK for Affymetrix gene expression analysis, described below) and sorted progenitor/mature populations for Meis family expression) was extracted using RNaqueous-Micro Kit (Ambion/Life Technologies) as per manufacturers instructions due to the limiting cell numbers. Approximately 10ng of total RNA was reverse-trancscribed with Superscript®Vilo cDNA Synthesis kit (Life Technologies). Prior to RT-PCR, cDNA was pre-amplified 14 cycles with Taqman PreAmp Mastermix (Applied Biosytems) with the relevant primer sets (Table 2.5) diluted to 1:100 for global amplification of the genes of interest. RT-PCR was performed on  60 pre-amplified cDNA (0.125 µl per reaction) with Universal Taqman Mastermix (Applied Biosytems) and 6-FAM/ZEN/IBFQ PrimeTime® Assays (Integrated DNA Technologies, Coralville, IA, USA) on a 7900HT Fast Real-Time PCR System (Applied Biosystems).  Relative quantification analysis was performed using ABL1 as endogenous control and MxCre/Meis1-/- expression was compared to MxCre/Meis1-/+ expression for each target using RQ Manager version 1.2 Software (Applied Biosystems).  Table 2.5: Primer sets for Q-PCR detection of Meis1fl/fl genomic collapse Name Detects Forward Reverse NDF Non-deleted floxed 5’ - agcttcatttgaagttccctattg-3’                  5’- tattaggtggatccaagcttcatt-3’ DF Deleted floxed 5’- ctggactttctcctttagttggat-3’ 5’- ggaacttcatcagtcaggtacata-3’  Floxed CTL Floxed (regardless of deletion) 5’-tatgtacctgactgatgaagttcc -3’ 5’- gcgtcacttggaaaagcaat-3’ Exon 7 CTL Endogenous CTL 5’- ttggaatagagaccatgatgacac-3’  5’- gttatccccactgtgtgaagtatg-3’         61 Table 2.6: Primetime Q-RT-PCR Assays    Primetime Assays  IDT Assay ID mABL1 Mm.PT.42.14158394 mMeis1 Mm.PT.42.14235881 mMeis2 Mm.PT.45.7421202 mMeis3 Mm.PT.42.9683604 mGata1 Mm.PT.45.10444529 mGata2 Mm.PT.45.13913016.g mGata3 Mm.PT.45.11120670.g mPrep1 Mm.PT.45.16285104 mPrep2 Mm.PT.45.14033370 mNotch1 Mm.PT.45.10390781 mIL7r Mm.PT.45.14297778 mIL18r Mm.PT.45.11896127 mHbb-b1 Mm.PT.42.10942091 mHbb-b2 Mm.PT.45.10154419 mHlf Mm.PT.45.15839543 mMpo Mm.PT.45.15839547 mMsi2 Mm.PT.45.17224408 mPcgf5 Mm.PT.45.10726269 mSelp Mm.PT.45.16240531 mSenp8 Mm.PT.45.12425187.g mTyrobp Mm.PT.45.11022459 mVamp5 Mm.PT.45.5620876 mAurkb Mm.PT.47.5154422 mCcna2 Mm.PT.47.1386893 mCcnb2 Mm.PT.47.17484650 mCcnd1  Mm.PT.47.12022381 mClec12a Mm.PT.47.8638109 mDdx4 Mm.PT.47.16947842 mDgat1 Mm.PT.47.12387218 mE2f8 Mm.PT.47.7043043 mFlt3 Mm.PT.47.10501150 mGria3 Mm.PT.47.14084255 mGstm5 Mm.PT.47.11429296 mHes1 Mm.PT.47.8454373.g mHoxA6 Mm.PT.47.13073650 mPtpn5 Mm.PT.47.10380131.g mRassf4 Mm.PT.47.15971865 mSfpi1 Mm.PT.47.7508177 mSnx31 Mm.PT.47.15894188 mHIF1α Mm.PT.47.8983770 mHIF2α Mm.PT.47.7593249 mP16Ink4a Mm.PT.47.9881334 mP19Arf Mm.PT.47.5632963  62 Affymetrix mRNA array and analysis RNA was extracted from LSK sorted bone marrow from induced MxCre+/Meis1fl/fl and MxCre+/Meis1+/fl mice with Rnaqueous-Micro Kit (Ambion) as per manufacturers recommended procedure. RNA was concentrated by ultra-centrifugation, amplified and hybridized to the Affymetrix Mouse Exon ST 1.0 Array (Affymetrix, Santa Clara CA, USA) by the BC Cancer Agency Centre for Translational and Applied Genomics (CTAG) using the Ovation Pico WTA System (NuGEN) according to manufacturer’s instructions. For the quality control, 100 ng of each amplified cDNA product was tested in two separate runs by Bioanalyzer on an RNA 6000 Nano LapChip (Agilent): 1) before fragmentation and labeling (100 ng of each sample) and 2) after fragmentation and labeling. Input samples passed both the Bioanalyzer trace test for fragment size (>80% less than 200bp) and concentration and purity measurements by A320, A260 and A280 absorbance.  Affymetrix Expression Console software was used for the initial exon- and gene-based probe set signal normalization and log2 summarization. Our analysis was restricted to the well-annotated “core” exons and genes. Data were further adjusted using ComBat (Johnson, Rabinovic, & Li, 2007) to remove batch effects attributable to litter-to-litter variability rather than Meis1 status. We confirmed that the loss of Meis1 exon 8 was reflected in the exon-based signal intensity. To identify differentially regulated genes, samples were compared by two-tailed T-tests with unequal variance followed by Benjamini-Hochberg correction for multiple hypothesis testing (Benjamini & Hochberg, 1997). The R statistical computing environment was used for batch correction and expression analysis (R Development Core Team, 2009)).   63 Statistical Analysis Unpaired, unequal variance, two-tailed, Student’s T-tests were perfomed to determined statistical significance between Meis1-/- and control samples in Vet Analyzer samples, all semi-solid colony forming assays (CFC & CFU-Mk), RT-PCR, Q-RT-PCR and FACS analyses. Unpaired, unequal variance, two-tailed Student’s T-tests were also used to compare the peripheral blood engraftment between MxCre/Meis1fl/+ and MxCre/Meis1fl/fl mice prior to and following PolyI:C induction. LTRC and LTC-IC frequency and testing for significant differences between groups was performed using the Extreme Limiting Dilution Analysis (ELDA) online tool (http://bioinf.wehi.edu.au/software/elda), which utilizes Poission distribution analysis (Szilvassy et al., 1990). As described above, for the Affymetrix expression analysis, data were first normalized and summarized using the Affymetrix Expression Console software then adjusted using ComBat (Johnson, Rabinovic, & Li, 2007) to remove batch effects attributable to litter-to-litter variability rather than Meis1 status. Samples were compared by two-tailed T-tests with unequal variance followed by Benjamini-Hochberg correction for multiple hypothesis testing (Benjamini & Hochberg, 1999) 64  Chapter 3 : Validation of Conditional Meis1 Knock-Out Models  Introduction The development of knockout mouse models has revolutionized gene function studies in the mammalian system. Such models have been instrumental in understanding the role of a wide range of genes critical for normal and leukemic hematopoiesis. Previous Meis1 knockout models were limited by embryonic lethality. Hisa et al. disrupted embryonic expression of Meis1 using a βgeo gene trap cassette into exon 8 of Meis1, which allowed monitoring of embryonic stage and cell type specific expression of Meis1 through β-galactosidase expression (Hisa et al., 2004). This study showed that in the absence of Meis1, embryos do not survive past 14.5 days post coitus (dpc) and display massive hemorrhaging due to an absence of megakaryocytes. In addition, the embryos lack well-formed capillaries and have smaller lenses and partially duplicated retinas in the eye. The key results of this study were confirmed by Azcoitia et al., (Azcoitia et al., 2005) who attempted to generate an inducible model of Meis1 protein function by generating a fusion gene between Meis1 and ERTM. Theoretically, this protein should be sequestered to the cytoplasm in the absence of the synthetic estrogen-receptor ligand, tamoxifen, however, rescue experiments failed to correct embryonic lethality with in vivo tamoxifen administration to pregnant mice. Strikingly, both studies in analysis of fetal liver of Meis1-/- mice found a reduction of primitive hematopoietic cell number (KSL cells) as assessed by FACS (Azoitia et al, 2005) or loss of repopulation capacity as assessed by transplantation (Hisa et al., 2004) suggestive of a significant impairment in the number and/or function of HSC. However, these available models left  65 unanswered important questions concerning the role of Meis1 in later stages of ontogeny and in adult life.  Increasing evidence points to major differences in the properties and regulation of primitive hematopoietic cells in early ontogeny/embryonic versus post embryonic and adult stages of life. Around the time of birth, HSC activity moves from the fetal liver to bone marrow (Reviewed in Orkin & Zon, 2008). Both fetal liver and bone marrow HSC are termed to be “definitive”, that is, they can give rise to the full spectrum of mature blood cell types when transplanted into lethally irradiated recipients (Müller et al., 1994). Differences between fetal liver and bone marrow definitive HSC have been noted, however, as well as differences in bone marrow HSC in the peri-natal period compared to that of the adult. Fetal liver HSC show higher engraftment and mature cell output as well as self-renewal in vivo compared to adult bone marrow (Rebel et al., 1996). Bone marrow HSC isolated in the first three weeks after birth demonstrate similar reconstitution and expression profiles to fetal liver HSC (Bowie et al., 2007), suggesting that there are substantial differences between adult and fetal HSC.  Further support for differences in gene regulation between adult and fetal HSC comes from knock-out studies where a gene triggering embryonic lethality post-HSC development is not required in the adult. For example, loss of Aml1 (Runx1) is embryonic by 12.5dpc and results in a complete absence of definitive hematopoiesis (Okuda et al., 1996). In the adult however, absence of AML-1 results in a far subtler phenotype, with relatively normal engraftment with the exception of megakaryocyte, T -and B- cell maturation following transplantation (Ichikawa et al., 2004). Disparities in the requirement between embryonic  66 and adult hematopoiesis have also been observed with Scl(Tal), Mll and Tel/Etv6 (Mikkola et al., 2003; McMahon et al., 2007; Hock et al., 2004).  A conditional model of Meis1 deletion would also be of use in the study of leukemogenesis. Several lines of evidence highlighted in the introduction of this thesis suggest a pivotal role for Meis1 in the development of both Hox and non-Hox leukemias. For instance, it has been established that the oncogenicity of various Mll fusions is linked to the extent to which Meis1 expression is up-regulated (Wong et al., 2007) and that sh-RNA knock-down of Meis1 triggers cell cycle arrest and delays the establishment of leukemia in vivo (Kumar et al., 2009). It remains unknown, once a leukemia is established in vivo, if Meis1 is required for the survival of the LSC clone or if loss of Meis1 would have a significant impact on LSC number. The critical transcriptional networks controlled by Meis1 expression are also not yet clear.   To overcome the embryonic lethality associated with previous knockout models and enable studies in the setting of adult hematopoiesis (normal and leukemic), efforts in our lab were directed towards developing a conditional knockout model for Meis1. Fortuitously at this same time, Dr. Nancy Jenkins and Dr. Neal Copeland in personal communication revealed that they had successfully generated a conditional knockout mouse line for Meis1 based on the Cre/LoxP system. However, details of the nature of the targeting construct and resultant “floxed” region of the Meis1 gene were not available beyond an understanding that “exon 8” had been targeted. Upon receipt of breeding pairs for this mouse line, initial efforts as described in this chapter were devoted to characterizing the precise nature of the floxed allele and subsequent validation and optimization of strategies for efficient induction of Meis1 deletion both in vivo and in vitro. To these ends we first localized the position of the  67 two loxP sites flanking exon 8 and then validated the induction of Cre expression and subsequent floxed allele deletion in vivo using two inducible models of Cre expression (Mx-Cre and ERT-Cre, described below).  We confirmed the generation of a truncated RNA Meis1 transcript and loss of protein expression following Cre expression. In addition, we developed Q-PCR tools to monitor in a quantitative manner the extent of genomic collapse and related changes in Meis1 RNA expression. In order to focus our future investigations using the model we surveyed Meis family expression in the hematopoietic hierarchy in purified HSC, progenitor and differentiated cell enriched populations.  Results and Discussion Identification of LoxP sites flanking Meis1 in B6-Meis1tgLoxP/+ mice  Male B6-Meis1tgLoxP/+ (B6-Meis1fl/+) mice were a generous gift from Drs N. Jenkins and N. Copeland. The mice had been engineered with LoxP sites flanking exon 8 of Meis1, however for a variety of reasons, details on the exact position of the LoxP sites flanking exon 8 were not available when the mice were received in our lab. Based on the Meis1 RefGene sequence and the sequence of the primers provided for genotyping, it was determined that one of the LoxP sites (and accompanying targeting vector sequence) was positioned 3’ of exon 8 in intron 8 (anti-sense strand chromosome 11, 18 941 409; Dec. 2011(GRCm38/mm10 assembly)). With one site fixed, we used in vitro Cre expression by retrovirus, followed by Southern blot analysis to determine the collapsed fragment size, and thus likely location.  68  B6-Meis1+/fl bone marrow was infected with MSCV-CRE-PURO retrovirus (as described in the methods and materials) and selected for 4 days with 0, 1.6 or 3.2 µg/mL puromycin. DNA from the selected cells was digested with HindIII and EcoRI for Southern blot analysis. Probes specific to Meis1 genomic regions flanking the known LoxP site 3’ to exon 8 were generated by PCR. Compared to the size of the anticipated wild-type band, an additional hybridization band was detected using a probe 5’ to exon 8 in B6-Meis1+/fl bone marrow. This band disappeared upon expression of Cre recombinase, supporting the presence of the second LoxP site 5’ of exon 8. (Figure 3.1, Table 2.1).  PROBE A PROBE BHindIII HindIIIEcoRI EcoRI[Puro] μg/mL 0 1.6 3.2 0 1.6 3.2 0 1.6 3.2 0 1.6 3.2a) WT 7.4 Kbpfl 2.3 Kbpfl 2.1 KbpWT 3.0 KbpWT 1.9 KbpWT 7.4 Kbpfl 5.3 Kbpb) EXON 8HH HE E E E EPROBE APROBE BP1.6  Kbp1.0  KbpA 69 Figure 3.1: Southern blot analysis to localize the second LoxP site 3’ or 5’ to the known site.  a) Southern blot of Meisfl/+ 5-FU bone marrow infected with MSCV-Cre-Puro following selection with 0, 1.6 or 3.2 µg/mL puromycin and digested with HindIII or EcoRI.  Probe A was designed to detect a change in hybridization signal if the second LoxP site was 5’ to the known site (LoxP-A), whereas probe B was designed to detect a LoxP 3’ to the known site. Following selection for Cre expressing cells, there is a loss in the HindIII and EcoRI hybridization signal with probe A, indicating the unknown LoxP site is 5’ to LoxP-A. No change is seen with probe B. b) Schematic representation of the Meis1fl loci around Exon 8. The LoxP-A is represented as grey triangle “A”. For clarity the determined position of the unknown LoxP site is included as a dashed triangle. H represents HindIII sites. E represents EcoRI sites. Restriction sites in the wild-type allele are in black type while sites introduced by the targeting vector are in grey.    The location of the LoxP sites flanking exon 8 were confirmed using a series of sequence specific primers for extending regions of intron 7 (Table 3.1, Figure 3.2). These were used in concert with 2 reverse primers in an attempt to amplify the region of the mutant allele containing both LoxP sites. MeisCKO Rev1 is anchored to the known LoxP site 3’ of exon 8 and was designed to detect the targeted and uncollapsed allele as deletion of exon 8 by Cre recombinase will eliminate the site. MeisCKO Rev2 is located 3’ of the 3’ LoxP site in wild-type genomic DNA and remains intact regardless of exon 8 excision by Cre. The PCR analysis was performed on DNA from both selected and unselected B6-Meis1+/fl bone marrow infected with MSCV-CRE-PURO. The combination of Rev2/Intron 7 primer 3 revealed a significant size difference from the anticipated wild-type band, suggesting that the second LoxP site was 3’ of this primer and captured in the Rev2/Intron 7 amplicon. This material and from Intron7 Primer 3/MeisCKO Rev1 was cloned into the TOPO TA PCR2.1 vector and sent for sequencing. Sequencing confirmed the presence of the 5’ LoxP site (anti-sense strand chromosome 11 18943409; Dec. 2011(GRCm38/mm10 assembly)). The presence and position of both LoxP sites was confirmed in the non-deleted mutant allele by sequence analysis. Additionally, the deletion of genomic sequence  70 containing exon 8 between these LoxP sites leaving a single remaining LoxP site was confirmed.  Cre-mediated deletion of the Meis1fl/fl allele results in a loss of 2104bp in the targeted Meis1 locus (1998 of wild-type Meis1 locus). Collapse of the LoxP sites following Cre expression should result in a frame-shift and premature stop codon 15 bases into exon 9 in the processed RNA sequence (Figure 3.3) Table 3.1: “Walking” Intron 7 primers sequence and anticipated fragment sizes.  Sequence tethered in the LoxP site is highlighted in bold.  Name Sequence Anticipated size with Meis CKO rev1 (no Cre expression) Anticipated size with MeisCKO rev1 (Cre expression) Anticipated size with MeisCKO rev2 (no Cre expression) Anticipated size with MeisCKO rev2 (Cre expression) MeisCKO Rev1 5’- GAA GTT ATT AGG TGG ATC CAA GCT -3’ n/a n/a  n/a n/a  MeisCKO Rev2 5’- AGC GTC ACT TGG AAA AGC AAT GAT-3’ n/a n/a  n/a n/a  Intron7 Primer1 5’ – GAT TTG ATG CTC TTG CGA CA – 3’  ~1, 000 bp 0 ~1, 100 bp 0 Intron7 Primer2 5'- GCC GTG TAA CTG CCA TAG GT -3' ~1, 700 bp 0 ~1, 800 bp  0 Intron 7 Primer3 5’-TCT GAA GGG GTT TGA GAT GG -3’ ~2, 500 bp 0 ~2, 600 bp 0   Figure 3.2: Schematic representation of Intron 7 primers to narrow the position of the second LoxP site.  I7-1: Intron 7 primer1. I7-2: Intron 7 primer 2. I7-3: Intron 7 primer 3. Rev1: MeisCKO Rev1. Rev2: MeisCKO Rev2. Primer sequences found in Table 3.1. LoxP-A site is denoted    Exon 7  Exon 8 Exon 9       I7-1I7-2Rev1 Rev2I7-3 71 as a solid grey triangle. The putative locations that the Intron 7 primers would detect are represented as dashed triangles. The primers were designed such that when Cre expression was introduced, presence of an amplification product with Rev2 and the intron 7 primers would indicate the presence of the LoxP site between the primer binding sequences. For example, if the second LoxP site was located between I7-2 and I7-3, with Cre expression, there would be no amplification product with I7-1 and Rev2 as the primer binding site would be lost with LoxP site collapse. I7-2 and Rev2 would generate an amplification product whose size would help narrow the position of the second LoxP site. Sequencing of the amplification products confirmed the location of the second LoxP site (LoxP-B) beteween primers I7-3 and I7-2.    Figure 3.3: Sequence of targeted Meis1 allele.  The sequence is numbered according to the wild-type sequence on the antisense strand of chromosome 11 according to the Dec. 2011 (GRCm38/mm10) sequence from the Mouse Genome Reference Consortium. Non-wild-type sequence introduced by the targeting vector 18943465  CCATGTGACA CTGAAAATCT GGACTTTCTC CTTTAGTTGG ATGGGCTGAT GGCTTTAGGG      ATCCCCTCGA GGGACCTAAT AACTTCGTAT AGCATACATT ATACGAAGTT ATATTATTAA      GGGTTATTGA ATATGATCGG AATTGGGCTG CAGGAATTCG ATATCTGTGT TGGGGTCTTC18943395   GTAATCTTTT CTAGAAAGAA CTCAGTCTAT GAGCCCAGGA AACCTTCATG ACTTATATAG18943335    ATATTAAATG CACTAAACTT ATTATAGATG TTTTATAAAA GTTTACATTA CAAGAAGCTC18943275    TGCCAAAACT CATGGTGGTA TAGCTCTACT AACAGTCGAA ATGTTGGAAT CCATCTGCTC 18943215 CTGACTAGAA GGAGTTTATC TCTGAACACA TTAAGACAAG TGAAAGCCAG CAGGGGTCAC18943155    TTTTCACATG CCACAAGGGT GGGGCATGAG TTTCCTGATA CAATAGTTGT GAACAATTGA18943095   AGATCTGCCG TGTAACTGCC ATAGGTTACT AGAATGAATG AAATCCAACT AGCAACCCTT18943035 GTTGCACTGA GGACTTAGAG TAGCAACAAG TGATGGAAGA AGTGAAGCCA TGGACGAAAG18942975   TCTAATGGGA GTACCAATAG GTAGTAGCCA TCGGGGAAGT GTACTAGCAA AGTGTCCAGT18942915 TCTTAGGACC ATAGGTTCCC AATGACTTTA GTGGGATGAG GGTGCCCTGG GGTGTGGTTG18942855    CTACCGCACG TTGAAGCCTG GCTATTGTAG CTTTTCATTC TGGGCCCAGA TCAATCAGTC18942795   CATTTCCCTG ATAAGACGTG TCAGAGCACA GCTGCTGATC AGGATAGTGG GTTTAGTCCA18942735    TTGTTATTTG TCCTTTGACT TTGGTGCAAA GATCAAAAAT AGGTATTTCT AATGTAGGTC18942675    TTGATATATA TTACATGGGT AACTTTTTCT CATCTTATCA TATATGGCAT TATTACAGTA18942615    ATACTGTATT ATCTTTTTAC TGAGTTACAG AGATCAAGTA TTTCCTAAAA TAAGCAACTC18942555    AAGTGTTTAT AACAAATGCC AGTAGAAAAA TTGGATCCTT TAGGAATATG TTACCAGGTG18942495    ACCTCTCAGT GAGTGATCCT TGAAGAAAAA AATGTAGAAA CTGCCCAGGT TATTAGTGAG18942435    ATGCAATTCT AACATTACTT TATTTTACTT CAATTGCAAT TATTTTGTAA TTAAAAACTA18942375    TGTCAGGGGG TTTCCCTTAC TGGATTTGAT GCTCTTGCGA CACTAAGCAA AATTATATCT18942315    CTCCTTTGAA GCATTCCATG ACAAATTGCT TATGTGATGG ATTGCCTGCA GTTTCCCATG18942255    CTGCTCCCAC CCCCACCAGT CTGAAGATAG ATCACCTGTG GCAGCAGTTG CTGTATTAAA18942195    TGCATCTTCT TACCAGTTGA CGTCAGTGAC TTTATATGGA GTTCAAATAC TCAGAGACCC18942135    CCACTAACCA TGTCACCACA CTGAACTCCC CAGTAGAGGT ATTACATATA CTGAAATAAT18942075 TACCATGGCA ACAAAGCCAC TCAAGAGATG GATATTTCCC TCCAGTCCTT TTGCCTTTAT18942015    TACTGATCTT TGAAAGATTT TTTTTTATAG GCATGATTAT TTTTTAGTTG AGACCTTGTT18941955    CTTTCTCTTT GTCTTGGAAT TTAAAAAAAG TACCAAAGAA ACTACTAATC CTTTAATTGA18941895    TCTTTTGGTT TTGTTTGTAT CCAAAGGCAT TCTATTTGAA TAAAGCATGT TAAGGTAGCA18941835    TCAGTCAAAA TTTGATTCAC TATATTTGCA GGTGATGGCT TGGACAACAG TGTAGCTTCC18941775    CCCAGCACAG GTGACGATGA TGACCCTGAT AAGGACAAAA AGCGTCACAA AAAGCGTGGC18941715 ATCTTTCCCA AAGTAGCCAC CAATATCATG AGGGCGTGGC TGTTCCAGCA TCTAACAGTA18941655    AGTGGATTCT AAATGACATA TGTAAACTAA ATATGGACAA TGAACCAGTT TTGATAGAGA18941595    AAAAAACGAT CCCTCTGCGC TTCCTACATC ACTGTGTTCC CCATTTTTAA CCAAGGGTTA18941535    GCACTTATTT TCTAGACCCA TACCCTAAAA ATCAGGAGCT TCATTTGAAG TTCCCTATTG18941475 TAATTGTGTT AATTTTCAAA AGTTATAAGG CTCGGGAGTC CATGTGTACT TGTGCTTTTA18941415    TGAATGAAGC TTGGATCCAC CTAATAACTT CGTATAGCAT ACATTATACG AAGTTATATT      ATGTACCTGA CTGATGAAGT TCCTATACTT TCTAGAGAAT AGGAACTTCG GAATTCTATG18941405    TTAATATCAT TGCTTTTCCA AGTGACGCTG AAGTGCTTCC AGCTAGTTAC TAAACTTTCT18941345 AAAAGGAAAG TAAAACCCTT CCTCGAGATG GTAACTGGAG TGAGCACTGC AGTGGACCCT 72 is underlined while the LoxP sites are highlighted in grey. Exon 8 is denoted in italicized-bold type.  Breeding of B6-Meis1fl/fl/ B6;129Gt(ROSA)26Sortm1(cre/ERT)Nat/J and B6-Meis1fl/fl/ B6.Cg-Tg(Mx1-cre)1Cgn/J mice  The Mx-Cre (Kühn et al., 1995) and ERT-Cre (Hayashi and McMahon, 2002) mouse lines were used to generate an inducible model of Meis1 deletion in adult B6-Meis1fl/fl mice. The Mx-Cre mouse line allows for induction of interferon inducible Cre expression using  polyinosinic-polycytidylic acid (Poly:IC), a synthetic double stranded RNA that mimics viral infection (Kühn et al., 1995). In the case of the ERTM-Cre (ERT-Cre) line, Rosa26 driven Cre expression is controlled at the level of cellular localization of the fusion protein consisting of a mutant form of the estrogen receptor (ERTM), which cannot bind it’s natural ligand 17β-estradiol and Cre recombinase. In the absence of the synthetic ligand 4-hydroxy-tamoxifen (4-OHT), ERTM-Cre is sequestered in the cellular cytoplasm. Upon IP delivery of 4-OHT to the mouse, ERTM-Cre moves to the nucleus to promote efficient Cre-LoxP excision in many tissues, including bone marrow (Badea et al., 2003; Jo et al., 2011).   B6-Meis+/fl mice were bred onto the B6;129Gt(ROSA)26Sortm1(cre/ERT)Nat/J background (generous gift from A. Weng), for 4-OHT responsive gene deletion, and the B6.Cg-Tg(Mx1-cre)1Cgn/J background (Jackson Laboratories) for Poly:IC induced deletion. Progeny from these pairing were interbred to generate B6-Meis1fl/fl/ B6;129Gt(ROSA)26Sortm1(cre/ERT)Nat/J (ERTCre/Meis1fl/fl) and B6-Meis1fl/fl/ B6.Cg-Tg(Mx1-cre)1Cgn/J mice (MxCre/Meis1fl/fl). Mice from both strains were born in expected numbers and appropriate frequencies, suggesting that there was no selection against untreated MxCre/Meis1fl/fl or ERTCre/Meis1fl/fl germ cells or embryos in vivo.   73 Induction regimen of MxCre/Meis1fl/fl and ERTCre/Meis1fl/fl mice in vivo and generation of tools to quantitatively measure Meis1fl/fl deletion   A variety of induction schemes for both the MxCre and ERTCre models have been reported in the literature. For example, in the MxCre model, three to nine intraperitoneal injections of Poly:IC at 10µg/g every other day have been reported to be required for complete collapse of the target of interest in hematopoietic cells (Shinnick et al., 2010; Lieu & Reddy, 2009). Single monthly injections to consecutive daily injections for 5 days of 130 to 200µg/g of 4-hydroxytamoxifen (4-OHT) have been reported in the ERTCre and ERT2Cre models (Jo et al., 2011; Gan et al., 2008; Gurumurthy et al., 2010).  Starting with the MxCre/Meis1 model, we first tested three 300µg Poly:IC injections spaced 48 hours apart and measured Meis1fl/fl deletion by Southern blot analysis. In these mice the deletion was variable with as little as 46% deletion in spleen and 86% in bone marrow (Figure 3.4, panel a). An increase in dosing to 9 injections at the same 48 hour interval resulted in >98% deletion of the floxed Meis1 allele in bone marrow as measured by Southern blot (Figure 3.4, panel B). Based on these results, an induction regimen for MxCre/Meis1 of 9 injections of 300µg Poly:IC every 48 hours) was selected for future experiments.  The induction scheme for ERTCre/Meis1 (1mg 4-OHT/mouse every 48 hours for 6 injections) was initially chosen based on published reports (Gurumurthy et al., 2010) and the experience of personnel using the strain in another lab. This scheme also yielded >98% deletion of the floxed Meis1 allele in bone marrow as measured by Southern blot (Figure 3.4, panel b).   74  Figure 3.4: Southern blot analysis to localize the second LoxP site to 3’ or 5’ to the known site a) Southern blot of loxP targeted Meis1 tissues using various induction schemes for in vivo induction of Cre expression in MxCre/Meis1tg and ERTCre/Meis1tg mice. Tissue-derived DNA was digested with BamHI and probed in the region anchored in loxP-B. Percent deletion of the floxed Meis1 allele was calculated using densitometry software (ImageQuant, GE). One of each Meis1+/fl and MxCre/Meis1+/fl mice were used in the first 3 injection induction experiment whereas a total of 6 MxCre/Meis1+/+, MxCre/Meis1fl/+ and MxCre/Meis1fl/fl mice were used in the second attempt. 2 ERTCre/Meis1fl/fl mice were compared to a ERTCre/Meis1fl/+and Meis1fl/fl mouse. Mice were assessed 2 days after the last IP injection. b) Schematic representation of the Meis1fl loci around Exon 8 with BglIII restriction endonuclease sites (B) and Southern probe location.  EXON 8BPROBE AP CBB% fl allele deleted+/+ crefl/fl cre+/fl crefl/fl cre +/fl cre+/fl crefl/fl crefl/fl cre+/fl WT+/fl crea) b) WT 2.4 Kbpfl 2.5 Kbp∆ fl 2.5 Kbp+/fl 0μg/mL+/fl 3.2 μg/mL+/fl WT BM+/fl WT Spl+/fl cre BM+/fl cre Spl+/fl cre PBin vitro MSCV-Cre-Puro CTLMxCre/Meis1 in vivo IP 300μg Poly:I:C q48hrs x 9ERTCre/Meis1 in vivo 1mg 4-OHT q48hrs x 6MxCre/Meis1 in vivo IP 300μg Poly:I:C q48hrs x 313 51 - 97 99 98 97 98 - 97 97 9286 46 81- -AB 75 As key populations of interest in our studies such as HSC represent a small portion (0.0001%) of the total bone marrow, we also sought a method applicable to small cell numbers for accurately measuring the extent of Meis1 deletion at the level of DNA and the reduction in Meis1 expression at the RNA level. To accomplish this we generated Q-PCR primer sets that would allow quantitative detection of both the floxed intact and collapsed alleles using SYBR reagent with floxed and endogenous controls (Figure 3.5, Table 3.2). Primer set NDF (non-deleted floxed) amplifies the region between exon 8 and the 3’ loxP site (LoxP-A). The reverse primer is anchored in the 5’region of LoxP-A such that only intact, Meis1fl alleles are amplified. Primer set DF (deleted floxed) uses sequences 5’ of the LoxP-A site and 3’ of LoxP-B site in targeting vector sequence such that a product is only efficiently generated following Cre mediated collapse of the loxP sites. The floxed CTL primers detect the floxed allele regardless of collapse due to positioning of the forward primer in the 3’ of region of LoxP-A in the targeting vector sequence and the reverse primer in endogenous Meis1 sequence. Primer set exon 7 CTL is a control for levels of both wild-type and targeted Meis1 alleles. By comparing the CT values between the primer sets using the spleen DNA controls described below, it is possible to both confirm the genotype of the mice in the sample as well as the degree of collapse between the loxP sequences.        Exon 9       FCtl-FDF-FNDF-R DF-RE7-F E7-R FCtl-REXON 7 EXON 8LoxP-ALoxP-BNDF-F 76 Figure 3.5: Schematic representation of Q-PCR primers for quantitative determination of Meis1fl deletion.  These 4 sets of Q-PCR primers allow the quantitation of the floxed allele in a sample, regardless of deletion (FCtl), as well as non-deleted and deleted Meis1fl alleles (NDR and DF, respectively). Exon 7 (E7) primers serve as a control for the quantity of Meis1 DNA in the sample, be it wild-type or floxed.  Table 3.2: QPCR Primers for Genotyping and Detection of Meis1fl Deletion Name Detects Forward Reverse NDF Non-deleted floxed 5’-agcttcatttgaagttccctattg-3’ 5’-tattaggtggatccaagcttcatt-3’ DF Deleted floxed 5’-ctggactttctcctttagttggat-3’ 5’-ggaacttcatcagtcaggtacata-3’ Floxed CTL (FCtl) Floxed (regardless of deletion) 5’-tatagtacctgactgatgaagttcc-3’ 5’-gcgtcacttggaaaagcaat-3’ Exon 7 CTL (E7) Endogenous control 5’-ttggaatagagaccatgatgacac-3’ 5’-gttatccccactgtgtgaagtatg-3’  Spleen DNA from a MxWT/Meis1fl/fl (C0) mouse and a MxCre/Meis1-/- (C100) mouse that were 0% and >98% deleted as assessed by Southern blot were used to determine optimal primer sets in terms of efficiency and accuracy for a given percent genomic DNA collapse. Primer set efficiencies were tested by a series of five 2-fold serial dilutions from 25ng of DNA against appropriate samples for the primer sets. When plotted on a log2 scale, the slope of the CT values against the DNA dilution fell between 1.0 and 1.1, suggesting the efficiencies of all 4 primer sets are within the acceptable range and roughly equivalent. Varying proportions of C0 and C100 DNA were mixed to determine the accuracy of each primer set at a given deletion threshold. For example, when an allele is approaching 75-100% collapse, sample comparison to C0 using primer set NDF (detects the remaining intact floxed allele) is more accurate due to a larger ΔCT between the samples. We also used intron spanning RT-PCR primers to measure the amount of Meis1 transcript containing exon 8 remaining (IDT PrimeTime Mm.PT.42.14235881). When the loss of exon 8 at the level of  77 DNA is  >98% in a Meisfl/fl mouse, the loss of transcript is consistently >500-fold compared to a Meis1-/+ control.   Confirmation of truncated transcript and protein expression in ERTCre/Meis1 mice  With in vivo induction schemes established for both the ERTCre/Meis1 and MxCre/Meis1 mouse models, we wanted to determine if collapse of the floxed exon 8 resulted in the expression of a Meis1 transcript and detectable truncated protein product. We first investigated whether there was expression of an exon 8 collapsed RNA transcript in in vivo treated mice by isolating splenocytes from ERTCre/Meis1 mice treated every other day with 1mg/mouse 4-OHT as per the established induction scheme (ERTCre/Meis1-/-, ERTCre/Meis1-/+, and ERTCre/Meis1+/+ mice). Total RNA was extracted and Meis1 exon 7 and 11 primers used to amplify the segment of interest following reverse transcription. The amplified fragment was then cloned into the TOPO-TA vector and individual clones sent for sequencing. Sequencing results revealed that the expected sequence, with a frameshift mutation and premature stop codon in exon 9, was produced in treated ERTCre/Meis1-/- and ERTCre/Meis1-/+ mice (Figure 3.6).  Figure 3.6: Sequencing results confirming introduction of premature stop codon in exon 9 of Meis1fl with expression of Cre recombinase.  cDNA from ERTCre/Meis1-/+, ERTCre/Meis1-/-, and ERTCre/Meis1+/+ splenocytes was amplified using primers in exon 7 and exon 11 and cloned into the TOPO-TA vector for sequencing. Sequencing of the clones confirmed generation of the predicted transcipt with a premature stop codon in exon 9 following Cre recombinase expression. tccactcgttcaggaggaaccccgggcccttccagcggtggccatacttcacacagtggggataacagcagtgagcaagcacccttacccttctga tccactcgttcaggaggaaccccgggcccttccagcggtggccatacttcacacagtggggataacagcagtgagcaaggtgatggcttggacaac tccactcgttcaggaggaaccccgggcccttccagcggtggccatacttcacacagtggggataacagcagtgagcaaggtgatggcttggacaac tccactcgttcaggaggaaccccgggcccttccagcggtggccatacttcacacagtggggataacagcagtgagcaagcacccttacccttctga S T R S G G T P G P S S G G H T S H S G D N S S E Q G D G L D N… S T R S G G T P G P S S G G H T S H S G D N S S E Q A P L P F * EXON 8EXON 9EXON 7Predicted WT cDNART-PCR sequencePredicted proteinPredicted WT cDNART-PCR sequencePredicted protein 78   As a mutant RNA transcript was expressed in these mice, we went on to perform western blot analysis to confirm loss of full-length MEIS1 protein and investigate the possibility of expression of a truncated mutant protein (Figure 3.7a). As a large number of cells is required for western blotting and due to the limited cell populations in which Meis1 is expressed, we chose to use ERTCre/Meis1fl/fl marrow expanded by retroviral expression of the oncogene MN1 (Heuser et al., 2007). ERTCre/Meis1fl/fl marrow cells were infected, sorted for MN1 expression based on co-expression of green fluorescence protein (GFP) by FACS, expanded in culture for 20 days and then treated with either 4-OHT or the carrier control (EtOH) in vitro for 48 hours. The Western blot was probed for MEIS1 using a polyclonal rabbit αMEIS1 antibody raised against residues 200-300 of MEIS1, as well as GAPDH as a loading control. Developed blots from the EtOH treated cells showed a band at the expected size for MEIS1a at 43KDa. An additional band at roughly 28KDa was detected which may correspond to MEIS1D, an isoform of Meis1 previously reported in colorectal cancer (Crist et al., 2011). Following 4-OHT induction, the band at 43KDa is completely absent, suggesting no wild-type MEIS1 protein is produced following Meis1 exon 8 deletion. The 28KDa band present in the EtOH treated cells is also fainter and a stronger band at 27KDa is seen (Figure 3.7b). The band at 27KDa may represent expression of a truncated MEIS1 protein generated with the loss of exon 8 and premature stop in exon 9 (MEIS251) as it is at the expected molecular weight.   79  Figure 3.7: Western blot analysis confirming loss of MEIS1 protein following in vitro induction of Cre expression in mouse bone marrow expanded with pSF91-MN1 retrovirus.  a) Schematic representation of wild-type (WT) and possible truncated (MEIS251) protein due to Meis1fl/fl deletion by Cre-recombinase b) Due to the low level of MEIS1 expression in bulk marrow and the cell number required for Western blot analysis, bone marrow from 5-FU pre-treated ERTCre/Meis1fl/fl mice was infected with a retrovirus to cause overexpression of the oncogene MN1 to expand a primitive population of MEIS1 expressing cells. Following induction of Cre expression with 4-OHT, there is loss of full-length MEIS1A protein as expected. The blots were stripped and re-probed with GAPDH antibody as a loading control.  Overexpression of a truncated protein does not interfere with CFC potential or re-plating  As MEIS251 retains PBX interaction motifs and may be expressed in our knock-out model, we wanted to investigate the possibility that MEIS251 could serve as a double mutation eg. serving as a dominant negative form of Meis1 by interfering with PBX1 αMEIS1αGAPDH43 KDa27 KDa40 KDa+ 4-OHT + EtOHERTCre/Meis1fl/fl BM (pSF91-MN1)HM1 HM2 HDHM1 HM2WT Meis1aPossible truncation Meis1-251a) b)  80 availability by binding PBX protein and interfering with DNA binding. To do this, we chose to use a NUP98-HOXD13(ND13) model of overexpression in bone marrow progenitors and monitor the impact of MEIS251 on serial re-plating as compared to MEIS1A. Co-expression of Meis1 with ND13 has been shown in our lab to enhance the colony-forming frequency in semi-solid media in these cells (Pineault et al., 2005), thus we reasoned that if MEIS251 served as a dominant negative, a negative impact on serial plating would be seen in isolation or in combination with ND13.  5-FU pre-treated bone marrow from a C57Bl6/J mouse was harvested and infected by producer co-culture with retrovirus expressing ND13-GFP, MEIS1A-YFP, MEIS1B-YFP, MEIS251-YFP or a combination of ND13 + MEIS. Cells were isolated, pre-stimulated, infected and sorted as described in Chapter 2. The retrovirally-transduced cells were plated into methycellulose cultures 2 days following sorting (9 days from harvest) grown for 7 days, counted, harvested and re-plated for a total of 3 platings. This was performed in triplicate. No significant differences in re-plating were seen in MEIS251 compared to MEIS1a cultures or ND13+MEIS251 or ND13+MEIS1a cultures (Figure 3.8). There was a significant difference between ND13+MEIS1/MEIS251 cultures and the MEIS1/MEIS251 culture alone (p<0.05). This argues against a strong dominant negative effect of the truncated MEIS251 protein if it is expressed in significant amounts. There remains the possibility, however, that the truncated protein could exert an effect on a Meis1 knockout background that was not evaluated in these experiments. A lack of dominant negative effect on a Meis1 knock-out background is suggested by Hisa et al. (Hisa et al., 2004) and their Meis1-βgal gene trap knock-out. The resultant truncation protein in their model is similar to ours, however, no differences were observed between heterozygous and wild-type littermates, arguing against a  81 strong dominant negative effect of the truncation (Hisa et al., 2004). In several early experiments reported in Chapter 4, we included Cre-expressing Meis1+/+ mice on the ERT-Cre and Mx-Cre background and found no significant differences between these mice and mice heterozygous for Meis1-/+. This may not be surprising as other MEIS family members such as PREP1, MEIS2 and MEIS3 also induce nuclear accumulation of PBX1 when over-expressed (Saleh et al., 2000).  Figure 3.8: Serial re-plating of methylcellulose cultures over-expressing ND13, MEIS1A, MEIS251, ND13+MEIS1A or ND13+MEIS251.  Retroviral constructs over-expressing ND13, MEIS1A, MEIS251, ND13+MEIS1A or ND13+MEIS251 were introduced into wild-type 5-FU pre-treated mouse bone marrow. Following colony counts, the entire contents of the dish were harvested and used to re-initiate methylcellulose cultures. There was no significant difference in plating between MEIS251 or MEIS1A culture and ND13+MEIS251 or ND13+MEIS1A cultures, suggesting that MEIS251 does not exert a dominant negative effect on regulation of MEIS1 targets. ND13Meis1aMeis251ND13+Meis1aND13+Meis251100101102103104105106107Cumulative Colony TotalsPrimary PlatingSecondary PlatingTertiary Plating** 82 There was a significant difference between MEIS1A and ND13+MEIS1A cultures (p=0.05) at the third replating as well as between MEIS251 and ND13+MEIS251 cultures (p=0.04), corroborating the synergy between MEIS and HOX expression. Three replicates with duplicate methycellulose cultures were performed.  Meis family member expression in sorted wild-type mouse bone marrow populations To gain a better appreciation of hematopoietic cell types most likely to be affected by Meis1 deletion, a series of experiments were carried out to assess the level of Meis1 expression by Q-RT-PCR in highly purified sub-populations spanning very primitive, HSC enriched to late lineage positive cells.  These analyses were also extended to Meis1 family members including Meis2, Meis3, Prep1 and Prep2.  Wild-type, 8 week-old B6.SJL-Ptprca Pepcb/BoyJ (Peb3b) mice were sacrificed and stained according to Table 2.2. Individual mice were used for each of the progenitor cell populations (ESLAM HSC, LSK HSC, CMP/MEP/GMP, CLP and MkP) to obtain sufficient cell numbers whereas the mature cell populations (B-cells, T-cell, granulocytes, granulocyte progenitors, and maturing erythroid) were possible to isolate from the same mouse for sufficient cell numbers. Mature MkP RNA from individual mice were a gift from the lab of Dr. S. Nilsson. The gating strategies for each population are outlined in Table 2.3. Following pre-amplification, Meis family member expression was assessed (Figure 3.10) in addition to several key genes that would allow confirmation of the sorted populations (Table 2.3). Relative to mAbl, Meis1 is expressed at high levels in purified HSC populations (ESLAM HSC, LSK HSC) and down-regulated as cells undergo lineage commitment in common myeloid and lymphoid progenitors (Figure 3.9). Meis1 levels are similar in megakaryocyte progenitors (MkP) to HSCs, which is consistent with platelet defects seen in the embryonic knock-outs (Hisa et al., 2004; Azcoita et al., 2005). Meis1 is minimally expressed in the mature cell populations examined. Both Meis2 and Prep2 are expressed at  83 very low levels in all the populations examined, suggesting a negligible role in hematopoiesis. Meis3 seems to be expressed at a relatively consistent level across the populations but appears to be down-regulated in LSK HSCs, CMPs and GMPs as well as Mac1+ cells. Higher levels of expression in ESLAM HSCs, MkPs and maturing erythroid cells may suggest a role in these cells. To date, little is known of the function of Meis3 in the hematopoietic system, although overexpression in Icsbp-/- mice has a similar effect to Meis1 overexpression (Hara et al., 2008). Prep1 is also expressed at the highest level in granulocyte progenitors (Gr1+Mac1+) and escalating levels in subsequent steps of erythroid maturation. A role for Prep1 in these cell types have yet to be described, however a role as a tumor suppressor has been recently suggested as Prep1 deficiency predisposes mice to lymphomas and carinomas and transplanted fetal liver cells deficient in Prep1 induce lymphomas (Longobardi et al., 2010). ALL samples, however, have higher levels of both PREP1 and MEIS1, and etoposide-resistant cells up-regulate PREP1 expression, suggesting a potential oncogenic role (Rosales-Aviña et al., 2011). Overall these results support focused examination of the HSC and megakaryocyte lineages in Meis1-/- marrow as this is where the highest levels of expression are found.   84  Figure 3.9: MEIS family expression in purified hematopoietic subsets.  Populations enriched for certain hematopoietc subsets were sorted by FACS. MEIS family expression was interrogated in these subsets using Q-RT-PCR following a gene-specific pre-amplification. n=3 for each of the cell lineages studied.  Meis1ESLAM LSK CMP GMP MEP MkP CLPCD4CD8+B220+Gr1Mac1+Mac1+Immature ErythMaturing ErythMature ErythWBM0.010.1110Transcription Relative to mAblMeis2ESLAM LSK CMP GMP MEP MkP CLPCD4CD8+B220+Gr1Mac1+Mac1+Immature ErythMaturing ErythMature ErythWBM0.0000010.000010.00010.0010.010.11Transcription Relative to mAblMeis3ESLAM LSK CMP GMP MEP MkP CLPCD4CD8+B220+Gr1Mac1+Mac1+Immature ErythMaturing ErythMature ErythWBM0.010.1110100Transcription Relative to mAblPrep1ESLAM LSK CMP GMP MEP MkP CLPCD4CD8+B220+Gr1Mac1+Mac1+Immature ErythMaturing ErythMature ErythWBM0.1110Transcription Relative to mAblPrep2ESLAM LSK CMP GMP MEP MkP CLPCD4CD8+B220+Gr1Mac1+Mac1+Immature ErythMaturing ErythMature ErythWBM0.0000010.000010.00010.0010.010.11Transcription Relative to mAbl 85 Summary  In this chapter we describe results validating two conditional models for in vivo deletion of Meis1 in the adult. Following identification of the two loxP sites flanking exon 8 in the Meis1 gene, we crossed the strain onto the Poly:IC inducible MxCre and 4-OHT inducible ERTCre strains. We validated in vivo induction of Cre expression and subsequent collapse of the loxP sequences in both models and developed a robust quantitative PCR based method for measuring the efficiency of deletion applicable to even small cell numbers. We also confirmed the generation of a loxP-collapsed Meis1 transcript that results in the introduction of a premature stop codon in exon 9. Western blot analysis of MN1-expanded bone marrow cells following in vitro induction of Cre expression confirmed loss of the full length Meis1a protein in Meis1-/- mice. We additionally surveyed expression of Meis family members in a panel of purified cell populations ranging from populations highly enriched for HSCs (ESLAM HSCs) to terminally differentiated lymphoid, myeloid and erythroid populations (CD4+CD8+, B220+, Gr1-Mac1+, Gr1+Mac1+ and Ter119+CD71+/mid/lo). Overall, these studies provide confidence to the validity of the inducible Meis1-deletion models and provide clues into populations potentially impacted by loss of Meis1 expression.     86 Chapter 4 : Meis1 is required for hematopoietic stem cell maintenance and erythroid/megakaryocytic potential  Introduction  The Hox cofactor Meis1 has been identified as an important transcription factor regulator in both normal and leukemic hematopoiesis. As reviewed in detail in Chapter 1, over-expression of Meis1 is seen in a large proportion of human leukemias. In experimental models engineered overexpression of Meis1 and Hox is highly leukemogenic and Meis1 expression has been shown to be critical for leukemogenicity of a range of MLL-translocation oncogenes (Wong et al., 2007). In non-MLL models of leukemia, the levels of Meis1 appear to influence the target of transformation (Heuser et al., 2011) as well as the expansion capacity of populations enriched for LSC activity (Woolthuis et al., 2012). The critical players in initiating and maintaining transformation are likely to be oncogene and context specific and the critical genes and molecular pathways affected by Meis1 expression largely remain to be determined.  In the context of normal hematopoiesis, our recent studies along with those of others previously reported reveal that Meis1 expression is highest in populations enriched for HSC potential. Meis1 expression diminishes with differentiation, with the exception of megakaryocyte progenitors where there is a resurgence of expression. Two previous studies using embryonic Meis1 loss of expression point to a key role for Meis1 in the maintenance of fetal HSC and megakaryocyte potential (Hisa et al., 2004; Azcoitia et al., 2005). More recently, a pair of studies published at the time of writing of this thesis with the same Meis1fl/fl mouse provided new evidence of a role for Meis1 in adult hematopoiesis. Most strikingly, with induction of Meis1 deletion, there was a marked reduction in HSC as assessed at the level of phenotype and function, alterations that could, at least in part be  87 linked to the roles of Meis1 in ROS pathways and cell cycle (Kocabas et al.2012; Unnisa et al, 2012).  Conflicting lines of evidence in several different experimental models exist for a role for Meis1 in erythropoiesis and megakaryopoieis. During mouse development, in the absence of Meis1, megakaryocytes fail to form and erythroid colony numbers are reduced in the fetal liver (Hisa et al., 2004; Azcoitia et al., 2005). In zebrafish, Meis1 knock-down results in loss of erythroid progenitors in mature fish (Cvejic et al., 2010), although a recent study using ES derived hematopoietic cells suggests Meis1 represses erythroid development at the MEP stage in favor of megakaryocyte development (Cai et al. 2012). The more recent studies of Meis1 deletion in adult mice did not reveal any megakaryocytic or erythroid anomalies (Kocabas et al.2012; Unnisa et al, 2012), and thus suggested Meis1’s role in these lineages are restricted to embryonic development.   In this chapter we describe the effects of loss of Meis1 on adult hematopoiesis and additionally explore transcription programs influenced by Meis1 through expression profiling. We confirmed a role for Meis1 in maintaining HSC number and self-renewal in the adult mammalian system and obtained evidence of two novel effectors of Meis1 function in the HSC compartment that have been previously implicated in leukemia.  Additionally, our findings highlight important roles for Meis1 in adult megakaryopoesis and erythropoiesis.  Results Loss of Meis1 in adult mice perturbs peripheral blood composition in MxCre/Meis1-/- and ERTCre/ Meis1-/- models As described in Chapter 3, our studies exploited a mouse line carrying a Meis1 allele in which Exon 8 is flanked by loxP sites, thus allowing for inducible deletion upon Cre  88 expression. Two different inducible models were used: MxCre/Meis1fl/fl and ERTCre/Meis1fl/fl, allowing for induction of Cre expression upon exposure to Poly:IC or Tamoxifen respectively.  For all in vivo studies described, the induction regimens were as optimized and validated previously (see Chapter 3); these consistently yielded >95% deletion as measured 2 to 7 days following the end of the induction regimen.  A first series of experiments focused on identifying early effects of Meis1 deletion (Figure 4.1). ERTCre/Meis1 mice were bled two days after the last of six 4-OHT injections. At this time, 3 of 13 ERTCre/Meis1-/- mice were found to be moribund and euthanized along with litter matched Cre-expressing control mice. ERTCre/Meis1-/+ and ERTCre/Meis1+/+ mice are clustered in this analysis as no significant difference was found between the groups and both had significant differences compared to ERTCre/Meis1-/- mice when compared individually. The peripheral blood of all mice was analyzed on the basis of cellular composition using both an automated differential cell counter and flow cytometer for immunophenotyping. The spleen and sternum of the euthanized moribund and control mice were fixed in paraformaldehyde and sent for histological analysis on the basis of H&E staining and the cell surface markers B220 and Ter119.   89  Figure 4.1: Cre expression induction schemes and summary of experiments performed with ERTCre/Meis1 and MxCre/Meis1 mouse models  Bone marrow from ERTCreMeis1-/- moribund mice (n=3) showed a marked reduction in cellularity, with less than 1.3x106 nucleated cells per trunk (2 femurs, 2 tibias and 2 iliac crests) compared to 1.9x107 nucleated cells in controls (p<0.04), fatty replacement of bone marrow (Figure 4.2, panel a) and a marked reduction in nucleated erythroid progenitors as demonstrated by Ter119 staining on fixed tissues (Figure 4.2, panel a). Although there was no significant reduction in BM nucleated cell counts for the remaining 10 surviving ERTCreMeis1-/-, there was a significant reduction of the average number red blood cells (RBC) in the peripheral blood of ERTCre/Meis1-/- mice, even when moribund mice were excluded from the analysis (Figure 4.2, panel b). The distribution of mature nucleated cell types remained largely unchanged (Figure 4.2, panel e). Overall, there was also a significant reduction in the number of platelets in the peripheral blood of ERTCre/Meis1-/- mice compared to controls (Figure 4.2, panel d). Together these results reveal that deletion of Meis1 in an adult mouse results in a rapid and marked decrease in RBC and platelet numbers MxCre/Meis11 mg/mouse 4-OHT IPq48 hrs 300μg/mouse PolyI:C IPq48 hrs2 days post-last IP 5-7 days post-last IP2 days post-last IPFACS analysis of HSC and progenitor popuationsCFC and CFU-Mk analysisPhenylhydrazine treatment and CFC analysisLTC-IC AssayPB composition PB composition BM & Spleen histologyLTRC frequency with secondary transplantsLTRC frequency Isolate HSC-enriched BM for Affymetrix Exon arrayERTCre/Meis1 90 and in at least some mice an extreme situation of BM hypoplasia, notably in early erythroid and megakaryocytic cells.  A similar early time course study was also carried out using the MxCre/Meis1fl/fl model.  In contrast to the ERTCre/Meis1fl/fl model, no early morbidity was detected for MxCre/Meis1-/- mice and marrow and spleen cellularity and histology remained unchanged compared to controls (data not shown) as assessed 2 days after induction. There were however significant reductions in WBC and platelet numbers (Figure 4.2, panel c, d). Thus in both the MxCre/Meis1-/- and ERTCre/Meis1-/- mouse models, loss of Meis1 was associated with a rapid decrease in late erythroid and/or megakaryocytic/platelet numbers strongly suggestive of a role for Meis1 at the late progenitor stage of hematopoiesis in these lineages.   91  H&E 10xTer119 10xH&E 40xTer119 40xH&E 10xTer119 10xa) b)c)ERTCre/Meis1+/+ ERTCre/Meis1-/-d)051015*051015050010001500ERTCre/Meis1 MxCre/Meis1-/--/--/+ & +/+ -/+ & +/+RBC x 106 /mm3ERTCre/Meis1 MxCre/Meis1-/--/--/+ & +/+ -/+ & +/+WBC x 103 /mm3Platelets x 103 /mm3ERTCre/Meis1 MxCre/Meis1-/--/--/+ & +/+ -/+ & +/+** * 92  Figure 4.2: Loss of Meis1 is associated with gross marrow and peripheral blood changes in ERTCre/Meis1 and MxCre/Meis1 mice.  a) Bone marrow and spleen cross-sections of moribund ERTCre/Meis1-/- mice stained with hematoxylin & eosin (H&E) and anti-Ter119 antibody, compared to ERTCre/Meis-/+ control. b) Loss of Meis1 in vivo results in a significant decrease in red blood cells in the peripheral blood in Meis1 -/- mice. ERTCre/Meis1 -/- (n=13) compared to ERTCre/Meis1-/+ & +/+ (p=0.01; ERTCre/Meis1+/+ n=3; ERTCre/Meis1-/fl n=10). The three ERTCre/Meis1-/- mice with the lowest RBC counts were euthanized due to pallor and lethargy and had profound reductions in bone marrow cellularity, that is, less than 1.3x106 nucleated cells per trunk (2x femur, 2x iliac crest, 2x tibia). Moribund mice are represented as hollow squares on the graph. MxCre/Meis1-/- mice (n=5) show no such reduction when compared to MxCre/Meis1+/+(n=2) and MxCre/Meis1-/+ (n=3) mice. c) Loss of Meis1 results in a reduction of white blood cells (WBC) in MxCre/Meis1 -/- mice compared to MxCre/Meis1 -/+ & +/+ (p=0.004, n=5) 2 days after the final PolyI:C injection.  d) Loss of Meis1 results in a reduction of peripheral platelets (PLT) in MxCre/Meis1 -/- ERTCre+/Meis1 -/- and mice compared to MxCre+ and ERTCre+ control mice 2 days after the final PolyI:C/4-OHT injection (p=0.02, and p=0.02, respectively). e) Lineage distribution in the peripheral blood of treated mice, 2 days following the last injection. No significant differences were found.   Generation of megakaryocyte and erythroid progenitors are impaired in MxCre/Meis1-/- mice To gain further insight into the role of Meis1 at the level of erythroid and megakaryocytic progenitors, we assessed CFU-Mk numbers and proliferative potential from whole BM 7 days following the end of Poly:I:C induction in MxCre/Meis1 mice (Figure 4.3, panel a). Loss of Meis1 resulted in a reduction of total CFU-Mk (p=0.007, n=7) compared to 020406080100Percent of Total PBGr1 +Mac1 +CD4 +CD8 +B220 +MxCre/Meis1-/+ MxCre/Meis1-/-ERTCre/Meis1-/+ ERTCre/Meis1-/-e) 93 MxCre/Meis1-/+ mouse marrow. The reduction was largely due to a 6-fold reduction in CFU-Mk with high proliferative potential that form colonies composed of >10 cell clusters (p=0.01, n=7, Figure 4.3, panel a).  This loss of CFU-Mk is consistent with the drop in platelet number seen in the peripheral blood of MxCre/Meis1-/-. To examine if this reduction in progenitor number was exclusive to the megakaryocytic lineage or extended to less differentiated myeloid parent populations and their progeny, bone marrow CFC assays were carried out on MxCre/Meis1-/- BM cells 5 days after the end of induction. Compared to MxCre/Meis1-/+ controls, the total colony number was not significantly reduced in MxCre/Meis1-/- bone marrow, but there was a 4-fold reduction in the number of large BFU-E derived erythroid colonies (p=0.04, n=7) and a 7-fold reduction in the number of multi-lineage colonies derived from CFU-GEMM (p=0.05, n=7) (Figure 4.3, panel b).  The reduction in the erythroid progenitor population was even more evident when methylcellulose media optimized to support BFU-E colony formation was used (Figure 4.3, panel c). In this case, while total BFU-E colony number was unchanged compared to control, large colonies of greater than 16 cell clusters indicative of colonies derived from more primitive progenitors were reduced by 32-fold (n=7, p=0.003). The significant reduction in large BFU-E and CFU-Mk colonies in MxCre/Meis1-/- mice suggests a role for Meis1 in the proliferative potential of these progenitors, although it is unclear if it is at the level of a shared megakaryocyte-erythroid progenitor (MEP) or in the individual erythroid (EP) and megakaryocyte lineages (MkP).  To further investigate the erythroid defect seen in these mice, we used an in vivo model of hemolytic anemia induced by phenylhydrazine (PHZ) to generate a proliferative stress on erythroid progenitors. Forty-eight hours after the last of 9 Poly:I:C injections,  94 MxCre/Meis1 mice were given an IP injection of PHZ (60mg/Kg) and euthanized 4 days later.  Mice were assessed on the basis of bone marrow cellularity, spleen weight and the ability of bone marrow and spleen cells to form BFU-E in erythroid-supportive methylcellulose media. While spleen and marrow cellularity were comparable between treated control and MxCre/Meis1-/- (data not shown), erythroid colony numbers were greatly reduced in Meis1-/- mice following PHZ treatment. Again, this was particularly evident for large erythroid colonies for which there was a 73-fold reduction in number from spleens of MxCre/Meis1-/- mice compared to MxCre/Meis1-/+ (p=0.006, n=2). Moreover, there were no colonies with greater than 16 clusters from the marrow of Meis1 deficient mice, implying a greater than 2000-fold loss (p=0.002, n=2) (Figure 4.3, panel d). The BFU-E proliferation defect under normal and stress conditions demonstrates that Meis1 is required for efficient erythropoieisis. While no loss of nucleated Ter119+ cells was seen in the spleen or marrow of these mice by immunohistochemistry in our first round of experiments, in contrast to the ERTCre+/Meis1-/- model, there is clearly a large change in the potential of progenitors that cannot be distinguished by histology.   95  a) b)c)BFU-E (<16 clusters) BFU-E (>16 clusters) 10100100010000100000Colony Number/ 2xTibia, Femur, Iliac Crest MxCre+/Meis1-/+MxCre+/Meis1-/-*CFU-Mk 4-10 clustersCFU-Mk >10 clusters Total 100100010000100000Colony Number/ 2xTibia, Femur, Iliac Crest MxCre+/Meis1-/+MxCre+/Meis1--/-**BFU-ECFU-GEMMCFU-GM TOTAL1001000100001000001000000Colony Number/ 2xTibia, Femur, Iliac Crest MxCre+/Meis1-/+MxCre+/Meis1-/-* 96  Figure 4.3: Colony forming cell (CFC) capacity is selectively reduced in the absence of Meis1.  a) CFU-megakaryocytic (CFU-Mk) collagen cultures demonstrate deficiency in the ability of MxCre/Meis1-/- bone marrow to generate CFU-Mk (p=0.007, n=7), most notably a 9-fold reduction in large CFU-Mk composed of >10 clusters per colony (p=0.01, n=7). b) Bone marrow from MxCre/Meis1-/- mice (n=7) have a 4-fold reduced capacity to form burst-forming erythroid (BFU-E, p=0.04), a 7-fold reduced CFU-granulocyte-erythrocyte-monocyte-megakaryocyte capacity (CFU-GEMM, p=0.05). c) BFU-E impairment in MxCre/Meis1-/- bone marrow is further demonstrated in erythroid-specific methylcellulose media. MxCre/Meis1-/- marrow 32-fold fewer BFU-E colonies with a potential of >16 clusters per colony (p=0.003, n=7). For both CFC and BFU-E assays, mice were euthanized and cells plated 5 days after the final PolyI:C injection. d) Phenylhydrazine treated MxCre/Meis1-/- spleen cells have a 73-fold reduced ability to form large(>16 clusters/colony) BFU-E (p=0.006, n=3) compared to MxCre/Meis1-/+. Bone marrow showed a >2000-fold reduction in both small and large BFU-E colonies (p=0.002, n=3). Mice were euthanized 4 days after the phenylhydrazine and thus 6 days following the final PolyI:C injection in two replicate experiments.   To examine the possibility that erythroid progenitors, quantified by in vitro CFC assay from MxCre/Meis1-/- mice were derived from progenitors that escaped Cre-mediated Meis1 deletion, individual BFU-E, CFU-GM and CFU-GEMM derived colonies were plucked and examined for Meis1 deletion by PCR. All 40 colonies examined (Table 4.1) were collapsed around exon 8. This suggests that although the expansion of BFU-E and d)BFU-E (<16 clusters) BFU-E (>16 clusters) 1001000100001000001000000Colony Number per Spleen/Bone Marrow MxCre+/Meis1-/+MxCre+/Meis1-/-Spleen Bone Marrow Spleen Bone Marrow** 97 multi-potential progenitors is impaired in the absence of Meis1, subsequent downstream proliferation and differentiation are not overtly impaired. Together these data point to a major if not absolute requirement for Meis1 in the window of early erythroid and megakaryocytic progenitor (and possibly at the MEP stage) down to the intermediate stage of committed erythroid and late megakaryocytic development.    Table 4.1: PCR of individual CFC colonies for Meis1 deletion.  (deleted/tested; NP = not possible; ND = not done)  BFU-E CFU-GM CFU-GEMM 7-days post induction ND 18/18 2/2 From LTC-IC NP 7/15 NP From Phenylhydrazine 20/20 20/20 NP  Meis1-/- results in a loss of hematopoietic stem cells, common myeloid progenitors and megakaryocyte progenitors  As an initial approach to examine the immediate impact of Meis1 deletion on a broader spectrum of hematopoietic cells including the most primitive HSC populations, BM cells were subjected to detailed immunophenotype analyses 5 to 7 days after induction in MxCre/Meis1-/- mice and compared to MxCre/Meis1-/+ controls.  An immunophenotyping strategy as employed for isolation of subpopulatioins used for Meis expression profiling was used (see Chapters 2 and 3) with the exception that the LSKCD150+CD48- immunophenotype was used to identify primitive HSC at an estimated purity of 1 in 2 to 1 in 5 (Kiel et al., 2005) (Table 2.2). These analyses carried out in 5 experiments with a total of 7 mice per group, revealed significant decreases in absolute numbers per mouse of several of the progenitor populations (representative plots Figure 4.4, panel a; summary Figure 4.4,  98 panel b), despite normal numbers of nucleated cells and differentiated lineage distribution in the marrow. Consistent with our earlier results from progenitor assays, there was an 11-fold decrease in the number of phenotypically defined megakaryocytic progenitors (MkP) in the marrow of MxCre/Meis1-/- mice (p=0.02, n=7). Also consistent with our CFC results there was a 6-fold decrease in the number of megakaryocytic-erythroid progenitors (MEP) (p=0.05, n=7). This reduction in progenitors was also seen for the most primitive common myeloid progenitor (CMP) compartment, with a 9-fold reduction (p=0.04, n=7), which is consistent with the decrease observed in number of BFU-E and CFU-GEMM in CFC assays. No decrease was observed in the GMP subpopulaton. Together these data point to a requirement for Meis1 in the CMP, MEP and MkP compartments but limited or no requirement at the stage of GMP and later. For the most primitive subpopulations assessed by immnophenotype analysis, there was no significant difference in the number of LSK cells (HSC frequency of ~1/50). However, in the more highly purified HSC subpopulation (LSKCD150+CD48-) there was a 5-fold reduction (p=0.005, n=7) in MxCre/Meis1-/- mice compared to MxCre/Meis1-/+ mice.   99  c-Kit+ Sca-1- c-Kit+ Sca-1+c-Kit+ Sca-1+c-Kit+ Sca-1-a)0 20K 40K 60KFSC020K40K60KSSC89.30 20K 40K 60KFSC020K40K60KSSC 88.10 20K 40K 60KFSC100101102103104*PerCPCy5.513.70 20K 40K 60KFSC100101102103104*PerCPCy5.59.85100 101 102 103 104*PECy7100101102103104*APC0.4316.32100 101 102 103 104*PECy7100101102103104*APC0.3834.51100 101 102 103 104*FITC100101102103104*PE10.269.716.5100 101 102 103 104*FITC100101102103104*PE21.461.313.8100 101 102 103 104*PECy7100101102103104*FITC0.0388100 101 102 103 104*PECy7100101102103104*FITC1.57100 101 102 103 104*PECy7100101102103104*FITC0.26100 101 102 103 104*PECy7100101102103104*FITC4.04SSCFSCLin FSCCD41CD150CD48CD150CD16/32CD34CD41CD150CD48CD150CD16/32CD34SSCFSCLin FSCc-KitSca1Lin - Lin -Myeloid ProgenitorsMegakaryocyte ProgenitorsHematopoieticStem CellMxCre/Meis1-/+ MxCre/Meis1-/-c-KitSca1Myeloid ProgenitorsMegakaryocyte ProgenitorsHematopoieticStem Cell 100  Figure 4.4: Loss of Meis1 results in profound phenotypic changes in mouse bone marrow.  a) Gating strategy and representative plots for myeloid progenitors, HSCs and MkP from MxCre/Meis1-/- and MxCre/Meis1-/+ bone marrow. b) Absolute number of phenotypically defined populations in the bone marrow of MxCre/Meis1-/- and MxCre/Meis1-/+ mice. There was a 9-fold reduction in CMP (p=0.04, n=7), 6-fold reduction in MEP (p=0.05, n=7), 11-fold reduction in MkP (p=0.02, n=7) and 5-fold reduction in HSC (p=0.005, n=7) enriched populations.  Meis1 is required for the maintenance of primitive cell populations capable of long-term hematopoiesis in vitro and in vivo To assess the impact of Meis1 deletion on functionally defined primitive hematopoietic cells, in vitro LTC-IC and in vivo competitive repopulation assays were performed. The frequency and calculated absolute number of LTC-IC as assessed by limit dilution assay was reduced by 6-fold in the BM of MxCre/Meis1-/- mice compared to b)Lineagec-Kit+ Sca1 CMP GMP MEP MkP c-Kit+Sca1+CD150 HSC CLP1.0×101.0×101.0×101.0×101.0×101.0×101.0×101.0×101.0×10Number per 2x Femur, Tibia, Illiac Crest****MxCre/Meis1-/+MxCre/Meis1-/- 101 MxCre/Meis1-/+ (assays initiated 2 days post end of induction) (Table 4.2). Interestingly only ~50% of CFC derived from LTC-IC assays of MxCre/Meis1-/- BM showed complete Meis1 deletion (Table 4.1). This result is consistent with a near obligate requirement for Meis1 at the stage of LTC-IC and early downstream progeny.   Table 4.2: LTC-IC frequency in MxCre/Meis1-/- bone marrow 4 week LTC-IC 1 in X Upper 95% CI Lower 95% CI MxCre/Meis1-/+ 26 283 18 774 36 796 MxCre/Meis1-/- 157 683 78 858 315 299       To further examine the requirement for Meis1 in the most primitive hematopoietic compartment, limit dilution assays for competitive repopulating cells were carried out. Both MxCre/Meis1-/- and MxCre/Meis1+/- (Figure 4.5, panel a) and ERTCre/Meis1-/- ERTCre/Meis1+/- (Figure 4.5, panel b) bulk bone marrow cells were transplanted at different doses with competitor recipient type bone marrow cells (100,000 cells). Donor cell contributions to the peripheral blood of recipient mice were then serially assessed at 4 week intervals to 16 weeks post-transplantation.  Limit dilution analysis revealed a 10.7-fold and 19-fold reduction in the frequency of HSC in MxCre/Meis1-/- (Figure 4.5, panel c) and ERTCre/Meis1-/- (Figure 4.5, panel b) mice, respectively, when compared to MxCre/Meis1-/+ or ERTCre/Meis1-/+ mice (Table 4.3). Expressed in absolute numbers of HSC per mouse, given that there is no difference in the total bone marrow cellularity between Meis1-/- and Meis1-/+ mice in either the ERTCre or MxCre model, there is an average of 2900 HSC per MxCre/Meis1-/+ compared to only 270 in the MxCre/Meis1-/- mouse. The reduction is similar  102 in the ERTCre model whereby there are roughly 5000 HSC in ERTCre/Meis1-/+ mice compared to 230 in ERTCre/Meis1-/- mice.  No differences were seen in the lineage distribution of donor-derived peripheral blood cells. At the time of transplantation, donor cells were confirmed to be >95% deleted for Meis1 and Q-PCR analysis normalized for the Meis1fl allele showed persistence of Meis1-/- cells in the donor compartment at >80% (data not shown) at 16-weeks. Thus while Meis1 loss results in a rapid and marked reduction in HSC number, the presence of Meis1 does not appear to be an absolute requirement for at least limited repopulation capacity of downstream progenitors. As differentiated cell output at the 16 week time point is from cells with HSC potential at the time of transplantation, if Meis1 was an absolute requirement for HSC lineage restriction and differentiation, we would have expected a far greater proportion of non-deleted alleles at this late time point. In addition, recipients receiving 1 HSC equivalent from either MxCre/Meis1-/+ or MxCre/Meis1-/- mice, the relative lineage contribution of donor cells in the peripheral blood was equivalent.  To examine more closely the importance of Meis1 for HSC function, secondary transplants were performed to assess HSC self-renewal and expansion in the absence of Meis1 in the MxCre/Meis1 model. In this experiment, 3 MxCre/Meis1fl/fl and 3 McCre/Meis1fl/+ mice were treated with PolyI:C as per the established induction scheme. Limiting dilution analysis into irradiated primary recipients was then performed at 2x106, 1x106, 2x105 and 2x104 cells per replicate mouse with 100, 000 helper cells. Following 4 months in vivo, a cohort of 3 mice transplanted with 2x106 cells were sacrificed, pooled and transplanted into irradiated secondary recipient mice (Figure 4.5, panel c) based on the percent donor contribution. That is, 1x107, 1x106 or 1x105 CD45.2 cells were transplanted  103 into secondary recipients, without helper cells. Based on the percent reconstitution (8% from MxCre/Meis1-/-, 53% from MxCre/Meis1-/+), this meant that 6.6x more cells were transplanted per secondary recipient in the MxCre/Meis1-/- arm.  The highest transplant dose of 1x107 cells was expected to contain the progeny of roughly 5 HSC (minimum 2 HSC, maximum 12 HSC) in the MxCre/Meis1-/- arm, based on the limiting dilution results in primary recipients. Even at this dose, however, there was no detectable long-term repopulation in secondary recipients in the absence of Meis1 (Figure 4.5, panel d). MxCre/Meis1-/+ cells were, however, capable of minimum maintenance of the transplanted stem cell pool. This represents, at minimum, a 25-fold reduction in the ability of HSC lacking Meis1 to contribute to long-term repopulation. This deficit was evident as early as 4-weeks post-transplantation into secondary recipients, suggesting HSC lacking Meis1 are severely impaired in their ability generate progeny with even short-term repopulating potential.    104  2x1061x1062x1052x1040.1110100Percent DonorCD45.2+ Transplant Dosea)b)1x1062x1052x1040.1110100Percent Donor (CD45.2+ )Transplant Dose - CD45.2 +Primary RecipientsMxCre/Meis1-/+MxCre/Meis1-/-Primary Reciepients ERTCre/Meis1-/+ERTCre/Meis1-/- 105  Figure 4.5: Loss of long-term repopulation and HSC self-renewal in the absence of Meis1.  a) Recipient CD45.2 engraftment 16 weeks post-transplant demonstrates a 10-fold reduction in HSC capacity in MxCre/Meis1-/- (n=2 donors into 3 recipients per cell dose at 2x106 and 1x106, with 1 failed injection at 2x106 cells. 6 recipient mice were used at 1x105 and 1x104 cells per mouse) bone marrow compared to MxCre/Meis1-/+ (n=2 donors into 3 recipients per cell dose at 2x106 and 1x106. 6 recipient mice were used at 1x105 and 1x104 cells per mouse; p=0.009). Individual points represent individual mice from 2 independent experiments. b) Recipient CD45.2 engraftment 16 weeks post-transplant demonstrates a 10-fold reduction in LTRC capacity in ERTCre/Meis1-/- (n=2 donors into 6 recipients per cell dose) bone marrow compared to ERTCre/Meis1-/+ (n=2 donors into 6 recipients per cell dose; p=0.00004). Individual points represent individual mice from 2 independent experiments. For both 1x1071x1061x1050.010.1110100Percent DonorSecondary RecipientsCD45.2+ Transplant DoseMxCre/Meis1-/+MxCre/Meis1-/-c)d)300μg/mouse PolyI:C IPq48 hrs2x1061x1062x1052x104+ 100 000 CD45.1 helper cellsCD45.2 cells/mouseMxCre/Meis1fl/flMxCre/Meis1fl/+CD45.24 months1x1071x1061x105CD45.2 cells/mouse 106 ERTCre/Meis1 and MxCre/Meis1 experiments, induced mice were euthanized and transplanted into donor mice 2 days following the final IP injection. c) Experimental outline of transplants into secondary recipients. d) Secondary recipients, 16 weeks following transplantation, demonstrate MxCre/Meis1-/- HSC are not maintained in vivo.  MxCre/Meis1-/-(n=3 pooled donors into 3 recipient mice) bone marrow fails to engraft compared to MxCre/Meis1-/+ (n=3 pooled donors into 3 recipient mice) at roughly equivalent HSC transplant doses.   Table 4.3: Reduction in HSC in MxCre/Meis1-/- and ERTCre/Meis1-/- mice  Mouse HSC Frequency Upper 95% CI Lower 95% CI Primary (p=0.009) MxCre/Meis1+/- 1/28 482 1/9 473 1/85 683 MxCre/Meis1-/- 1/306 624 1/121 558 1/773 441 Primary (p=0.00004) ERTCre/Meis1+/- 1/18 199 1/6 509 1/50 887 ERTCre/Meis1-/- 1/345 174 1/151 649 1/785 665  The requirement for Meis1 for HSC is cell intrinsic  The decrease in HSC number detected following Meis1 loss as assessed at the level of immunophenotype or functional repopulation could arise from cell HSC – intrinsic requirements for HSC function and maintenance, and/or from cell-extrinsic requirements, for example in the stem cell niche.  To examine these possibilities, we transplanted 2x106 MxCre/Meis1fl/fl (n=2 donor mice into 11 recipient mice) or MxCre/Meis1+/fl (n=2 donor mice into 11 recipient mice) cells with 1x106 recipient-type cells into lethally irradiated wild-type recipients 4-weeks prior to induction of Cre expression (Figure 4.6, panel a). Both MxCre/Meis1fl/fl and MxCre/Meis1+/fl cells engrafted equivalently as read out in peripheral blood prior to Cre expression (Figure 4.6, panel b). Following PolyI:C treatment, there was a 17% decrease in MxCre/Meis1-/-  contribution to peripheral blood 2 days after the final IP injection, whereas the levels of MxCre/Meis1-/+ engraftment remained unchanged compared to control mice given PBS IP (p=0.0005, n=7 PolyI:C, n=4 PBS, Figure 4.6, panel c). The decrease in engraftment by MxCre/Meis1-/- was also evident in BM  (38% decrease) while  107 again there was no difference between the engraftment of MxCre/Meis1-/+ cells under PolyI:C treated or PBS conditions (p=6x10-6, n=7 PolyI:C, n=4 PBS Figure 4.6, panel d).  In a first experiment to determine the HSC frequency following Meis1 deletion, the 3 recipient PolyI:C-treated mice were euthanized 3 days after the final IP injection, BM pooled and transplanted into secondary recipients. In this initial experiment, BM cells were transplanted without separation such that there was a mixture of MxCre/Meis1-derived cells (CD45.2+) and recipient/competitor derived cells (CD45.1+). Cell doses were however adjusted so that equivalent numbers of MxCre/Meis1-derived cells were transplanted. In this setting, there was no detectable engraftment by MxCre/Meis1-/- cells in the peripheral blood at 16 weeks (3 recipient mice per cell dose, Figure 4.6, panel, Table 4.4) in contrast to readily detected engraftment by MxCre/Meis1-/+ cells. As no mice were positive for engraftment in the MxCre/Meis1-/- arm, the HSC frequency could be only estimated to be <1 in 18 million compared to 1 in 140,000 for MxCre/Meis1-/+ marrow and did not reach statistical significance of p<0.05.  By performing bulk transplants based on MxCre/Meis1 cell number, however, MxCre/Meis1-/- HSC had to compete against a greater number of recipient wild-type cells than MxCre/Meis1-/+ cells due to the reduction in engraftment following Poly:I:C treatment. The experiment was therefore repeated to compensate for increased competition by recipient cells by sorting for CD45.2+ bone marrow cells prior to transplantation into secondary recipients, thereby ensuring that MxCre/Meis1-/- and MxCre/Meis1-/+ HSC were competing against equivalent numbers of wild-type HSC for engraftment. 4 PolyI:C-treated primary recipient mice were euthanized, pooled and sorted by CD45.2+ by FACS and transplanted into 3 secondary recipient mice per cell dose (the injection failed in one secondary recipient  108 mouse transplanted with 1.5x106 MxCre/Meis1-/+ cells). The decreased frequency of HSC in MxCre/Meis1-/- cells was maintained in the absence of greater competition and was indeed 7-fold lower than MxCre/Meis1-/+ cells (p<1x10-125, Table 4.4). These experiments demonstrate that the major requirement for Meis1 in the maintenance of HSC number is likely cell intrinsic.   109  Engraftment 1 Month Post-Transplant, Prior to Induction020406080Percent Donor (CD45.2+ )b)c) d)Peripheral Blood Engraftment 2 days Post-Induction 020406080Percent CD45.2+-  PolyI:C+ PolyI:C*020406080Percent CD45.2+-  PolyI:C+ PolyI:CBone Marrow Engraftment 3 days Post-Induction*MxCre/Meis1+/fl MxCre/Meis1fl/flMxCre/Meis1-/+ MxCre/Meis1-/- MxCre/Meis1-/+ MxCre/Meis1-/-a)MxCre/Meis1BMCD45.1+4 weeks Check engraftmentPolyI:C (q48hrsx9)Check engraftmentIsolate BMCD45.1+Bulk or CD45.2+ transplant 110  Figure 4.6: Loss of Meis1 results in an intrinsic defect in LTRC.  a) Experimental plan to test the cell-intrinsic requirement for Meis1 in HSCs. b) Peripheral blood engraftment of primary recipients 4 weeks after transplantation and prior to induction e)6x1066x1056x1040.010.1110100Transplant Dose - CD45.2+Percent Donor (CD45.2+)f)1.5x1061.5x1051.5x1040.010.1110100Transplant Dose (sorted CD45.2+) Percent Donor (CD45.2+)MxCre/Meis1-/+MxCre/Meis1-/-MxCre/Meis1-/+MxCre/Meis1-/- Bulk Transplantation into Secondary RecipientsSorted CD45.2+ Transplantation into SecondaryRecipients 111 shows no difference in engraftment (n=11). c) Peripheral blood engraftment of primary recipients 2 days following PolyI:C administration.  MxCre+/Meis1-/- (n=7) engraftment is reduced 17% compared to MxCre/Meis1fl/fl (n=4; p=0.0005). d) Bone marrow engraftment in primary recipients 3 days following PolyI:C administration. MxCre+/Meis1-/- (n=7)_engraftment is reduced 38% compared to MxCre/Meis1fl/fl (n=4; p=6x10-6). e) Engraftment of secondary recipients transplanted with bulk CD45.2 cells from primary mice with transplant dose based on the number of CD45.2 cells 24 weeks after transplantation. f) Engraftment of secondary recipients transplanted with sorted CD45.2 cells from primary mice 24 weeks after transplantation.   Table 4.4: HSC frequency following in vivo deletion of Meis1 in primary recipients  CD45.2+ HSC Frequency Upper 95% CI Lower 95% CI MxCre/Meis1+/- BULK 1/138 474 1/31 167 1/615 000 MxCre/Meis1-/- p<0.06 1/17 857 263 1/2 532 079 1/126 000 000 MxCre/Meis1+/- SORTED 1/160 806 1/39 986 1/646 690 MxCre/Meis1-/- p<1x10-125 1/1 015 552 1/284 833 1/3 620 882  Gene expression changes following loss of Meis1 in an HSC-enriched population To search for differentially regulated genes following Meis1 deletion, Meis1 was deleted in MxCre/Meis1 mice and the HSC-enriched LSK population isolated for mRNA extraction 7 days after the final PolyI:C injection. mRNA isolated  from 3 individual MxCre/Meis1-/- and 3 MxCre/Meis1-/+ mice was amplified and subjected to transcriptome analysis using the Affymetrix Exon ST array (Figure 4.7, panel a). Following normalization for hybridization, litter batch effects and multiple hypothesis testing as described in Chapter 2, 171 differentially expressed genes were identified using a 90% confidence interval (Table 4.5). Of these differentially expressed genes, 8 had a fold change expression greater than two-fold (Table 4.6).  These 8 genes and others of interest selected from previous studies of Meis1 candidate targets were validated by RT-PCR on the original samples and an additional two  112 independent replicates (Figure 4.7, panel b). Of note, 2 genes implicated in leukemia, Hlf and Msi2 (Andres-Aguayo, et al., 2012, de Boer et al., 2011) are down-regulated in response to the loss of Meis1. Msi2 has also been reported to be up-regulated in the context of Vp16/Meis1 (Wang et al., 2006).  ChIP-Seq data for Meis1 performed in our laboratory (Eric Yung et al., unpublished) has also revealed MEIS1 binding sites in the body of Msi2 and in both the body and transcription start site of Hlf. Flt3 was also identified as direct target of MEIS1 in these experiments and was found to be differentially expressed by 1.5-fold between MxCre+/Meis1-/ - and MxCre+/Meis1-/ + LSK cells. Flt3 has previously been shown to be induced by Meis1 overexpression and is thought to be one of several pathways by which Meis1 influences leukemic activity (Argiropoulos et al. 2008).  Leading edge Gene Set Enrichment Analysis (GSEA) of all differentially expressed genes showed enrichment in cell cycle genes, consistent with the proliferation defect seen in CFC studies and previous work in leukemia models (Figure 4.7, panel c, Argiropoulos et al., 2010). We attempted to follow up with cell cycle analysis on Meis1-/- cells in both primary cells in vitro and in vivo. In the LSK subset in vitro and in vivo no differences were seen in bromodeoxyuridine (BrdU) incorporation, suggesting no major differences in the proportion of cells in Go/G1, S or G2/M in the LSK population (data not shown) of Meis1-/-.   113  Figure 4.7: Affymetrix mouse Exon ST 1.0 analysis of gene expression changes as a result of loss of Meis1 in an HSC-enriched population.  a) Experimental design. Individual, litter-matched MxCre/Meis1fl/fl and MxCre/Meis1fl/+ mice were induced for Meis1 deletion with PolyI:C every 48 hours for 18 days. 7 days following the final injection, mice were euthanized, bone marrow extracted and sorted for Lin- Sca-1+ Affymetrix WorkflowPolyI:C treatment MxCre/Meis1+/fl  & MxCre/Meis1fl/fl  Isolation of Lin-Sca1+c-Kit+ bone marrow RNA isolation and concentrationRNA amplification and hybridization to Affymetrix mouse Exon ST arrayExpression Console signal normalization and log2 summarizationComBat adjustment of litter-to-litter variabilityIdentification of differentially expressed genes by two-tailed T-tests (unequal variance) and multiple hypothesis testing correctionValidation of differentially expressed genes by RT-PCRLeading edge analysis by gene set enrichment a)b)Meis1 Il18r1Hbb-b1Hbb-b2Tyrobp Hlf Msi2mFlt3mSnx310.00010.0010.010.1110100Relative Expression to mAbl fl+/−.1fl+/−.2fl+/−.3fl−/−.1fl−/−.2fl−/−.3Cks2Smc2AurkaHat1Kif11Bub1bCcna2E2f8Mcm4Cdca8AspmPlk1Spag5Plk1Kif20aNek2Bub1Ect2AurkbMcm5Ccnb2Birc5Tk1c)  114 c-Kit+ cells by FACS. RNA was isolated from this HSC-enriched population and amplified prior to hybridization on an Affymetrix mouse Exon ST 1.0 gene expression array. Signal was normalized using Expression Console software and log2 summarization. Differentially expressed genes were identified using two-tailed t-test with multiple hypothesis correction. Changed pathways were interrogated by leading edge analysis. b) Validation of selected candidate target genes by Q-RT-PCR, using both sample RNA and two additional independent replicates. c) Leading edge gene-set enrichment analysis reveals enrichment of genes involved in cell cycle.   Table 4.5: Genes differentially expressed with deletion of Meis1 as determined by Affymetrix analysis within a 90% adjusted confidence interval Probe Name Description Adjusted t-Test Fold decrease in Meis1-/- 6836829 Dgat1 0.006761443 0.686689462 6817970 Nt5dc2 0.01133132 0.70196703 6828741 C1qtnf3 0.01148776 0.829957265 6790699 Hlf 0.01207463 3.180698125 6942692 Tmem184a 0.03904288 0.919683837 6835004 Snx31 0.03916838 1.214778913 6790621 Msi2 0.04223439 2.070406603 6871457 Incenp 0.04471138 0.867291904 6929828 Nat8l 0.04471138 0.791477701 6798334 Adam6b 0.05381157 1.562741704 6993465 Endod1 0.05650069 0.808002529 6993472 Fut4 0.05650069 0.671168355 6953607 Hoxa6 0.05650069 0.835739062 6808173 Irx2 0.05650069 0.903497659 6992172 Dusp7 0.07083209 0.709573183 6769213 Plk5 0.07433471 0.958082417 6769381 D10Wsu102e 0.07433471 0.799486601 6790046 Evi2b 0.07433471 0.754337234 6974010 Ing1 0.07433471 0.795613559 6992436 Ngp 0.07433471 0.686477168 6967109 Ptpn5 0.07433471 0.953819688 6820084 Reep4 0.07433471 0.750008593 6992178 Rrp9 0.07433471 0.876239425 6916748 Slc2a1 0.07433471 0.904081689 6883184 Slc2a10 0.07433471 0.85757217 6857885 Srbd1 0.07433471 0.75571711 6894253 Chrna4 0.07433471 1.044130276 6933084 D930016D06Rik 0.07433471 1.512825956 6782496 Dbil5 0.07433471 1.221825464 6869635 Entpd1 0.07433471 1.258515318 6900348 Gstm5 0.07433471 1.529539288 6962759 Kctd14 0.07433471 1.338515966 6964011 Scnn1g 0.07623116 0.878159125 6867632 Cabp2 0.07905712 0.92302027 6748889 Il18r1 0.07905712 0.487679618 6978369 Mmp15 0.07905712 0.946099273  115 Probe Name Description Adjusted t-Test Fold decrease in Meis1-/- 6805191 Olfr263-ps1 0.07905712 3.401042566 6789360 2810408A11Rik 0.08290535 0.859347092 6895915 Bhlhe22 0.08290535 0.907053702 6925562 Zbtb8b 0.08290535 0.91087139 6965893 Lypd4 0.08290535 1.055802188 6763295 Ralgps2 0.08290535 1.363438361 6982267 Wwc2 0.08290535 1.165781047 6846010 Cd96 0.09373904 0.655477699 6995384 Fam55b 0.09373904 0.77994938 6920609 Gja10 0.09373904 0.847973393 6969997 Hbb-b1 0.09373904 0.205045826 6852144 Lbh 0.09373904 0.742217798 6855706 Srf 0.09373904 0.796670523 6883261 Trp53rk 0.09373904 0.68789235 6815523 Naip5 0.09373904 1.459240896 6751349 Dgkd 0.09440419 1.271864279 6917813 Asap3 0.09593448 0.956210778 6767782 Lilrb4 0.09593448 0.217899404 6959584 Tyrobp 0.09593448 0.477594239 6858134 Nrxn1 0.09593448 1.52807978 6959536 Zfp30 0.09593448 1.163797528 6858910 Ttc39c 0.09641302 0.885682526 6802491 6430527G18Rik 0.09826282 0.822903471 6969016 9930013L23Rik 0.09826282 0.934908463 6775322 C030046I01Rik 0.09826282 0.914925308 6966490 C230052I12Rik 0.09826282 0.835897232 6972491 Ccnd1 0.09826282 0.78631627 6910126 Clca2 0.09826282 0.875813922 6818044 D830044D21Rik 0.09826282 0.881678297 6873363 Fgf8 0.09826282 0.839396182 6780332 Gabrb2 0.09826282 0.744622975 6927253 Gabrd 0.09826282 0.903307584 6785213 Galr2 0.09826282 0.817732162 6964600 Gpr26 0.09826282 0.804403471 6981099 Ido1 0.09826282 0.920582657 6838716 Itgb7 0.09826282 0.861326652 6789197 Ntn1 0.09826282 0.873434537 6863467 Osbpl1a 0.09826282 0.811192447 6751264 Psmd1 0.09826282 0.903584256 6917963 Rap1gap 0.09826282 0.922605325 6983163 Rfxank 0.09826282 0.922160403 6799645 Rps7 0.09826282 0.800539592 6801636 Rtn1 0.09826282 0.931952816 6865957 Slc6a7 0.09826282 0.923006603 6841739 Tomm70a 0.09826282 0.708964521 6977523 Anapc10 0.09826282 1.102988954 6775441 Atcay 0.09826282 1.093399049 6937253 Fam53a 0.09826282 1.289113157 6943142 Flt3 0.09826282 1.510209021  116 Probe Name Description Adjusted t-Test Fold decrease in Meis1-/- 6980568 Gas6 0.09826282 1.076483534 6791302 Gjd3 0.09826282 1.212372281 6866800 Katnal2 0.09826282 1.120357182 6755237 Kcnj10 0.09826282 1.080823042 6838695 Krt78 0.09826282 1.093853513 6936930 Lmbr1 0.09826282 1.268653264 6818858 Mudeng 0.09826282 1.298148084 6818523 Ptger2 0.09826282 1.153822817 6815382 Rgnef 0.09826282 1.080829635 6966041 Shkbp1 0.09826282 1.039448805 6799842 Slc26a4 0.09826282 1.065860955 6871139 Mtvr2 0.09826282 1.12901198 6864444 Stard4 0.09826282 1.188663435 6824195 Txndc16 0.09826282 1.185122457 6954615 Vamp5 0.09826282 5.77675678 6807209 Zfp346 0.09826282 1.258522896 6872206 1700028P14Rik 0.09871875 0.939795599 6867593 1810055G02Rik 0.09871875 0.858117949 6969429 4632434I11Rik 0.09871875 0.699203711 6899308 4933434E20Rik 0.09871875 0.886234308 6875832 Adamts13 0.09871875 0.949246684 6968126 Arrdc4 0.09871875 0.789476573 6784765 Axin2 0.09871875 0.909899641 6784329 BC030867 0.09871875 0.859150792 6790199 Ccl1 0.09871875 0.852827886 6773485 Cdc40 0.09871875 0.644957771 6950137 Clec12a 0.09871875 0.55226552 6748695 Cnga3 0.09871875 0.848009481 7016421 Cul4b 0.09871875 0.756217809 6861751 D18Ertd653e 0.09871875 0.859290586 6892580 D630003M21Rik 0.09871875 0.972298247 6871627 Dtx4 0.09871875 0.920747116 6959674 Gm6725 0.09871875 0.680986794 6935273 Eif3b 0.09871875 0.666257578 6953126 Fam131b 0.09871875 0.838222299 6947394 Mogs 0.09871875 0.884365136 6934506 Glt1d1 0.09871875 0.893948355 6941215 Gltp 0.09871875 0.880693101 6840400 Hes1 0.09871875 0.918382136 6762452 Igfn1 0.09871875 0.918880257 6784845 Kcnj2 0.09871875 0.869945205 6823849 Mapk8 0.09871875 0.689086426 6829297 March11 0.09871875 0.865598668 6850723 Mea1 0.09871875 0.810454696 6876010 Med27 0.09871875 0.816689095 6993272 Mmp1b 0.09871875 0.872917557 6850661 Mrpl14 0.09871875 0.782647265 6943387 N4bp2l1 0.09871875 0.789798996 6911925 Nbn 0.09871875 0.719175098  117 Probe Name Description Adjusted t-Test Fold decrease in Meis1-/- 6804893 Nid1 0.09871875 0.918265128 6832332 Nup50 0.09871875 0.795160585 6850345 Olfr130 0.09871875 0.917081849 6958995 Opa3 0.09871875 0.808539415 6919095 Pank4 0.09871875 0.869308084 6991358 Plscr1 0.09871875 0.858824362 6924882 Ptprf 0.09871875 0.954875573 6791534 Pyy 0.09871875 0.82370519 6956765 Rassf4 0.09871875 0.628405716 6789789 Rph3al 0.09871875 0.895808026 6797572 Serpina5 0.09871875 0.823835983 6762423 Shisa4 0.09871875 0.888228821 6768323 Slc25a16 0.09871875 0.888569833 6978937 Sntb2 0.09871875 0.87858046 6902192 Ssx2ip 0.09871875 0.966105003 6893057 Sulf2 0.09871875 0.927836986 6837318 Tob2 0.09871875 0.891491411 6971410 Prss53 0.09871875 0.733547615 6768898 Vpreb3 0.09871875 0.866078032 6888928 Accs 0.09871875 1.100261389 6972149 BC066028 0.09871875 1.126072063 6864520 Brd8 0.09871875 1.137338966 6968772 Cib1 0.09871875 1.279751569 6845159 Fbxo45 0.09871875 1.099376831 6934855 Gatsl2 0.09871875 1.407191182 6961012 Magel2 0.09871875 1.155860681 6907869 Mov10 0.09871875 1.303027043 6970037 Olfr649 0.09871875 1.128090383 6955066 Paip2b 0.09871875 1.161642402 6957178 Plekhg6 0.09871875 1.048243756 6857834 Prepl 0.09871875 1.156735331 6854460 Rab11fip3 0.09871875 1.075945323 6866862 Setbp1 0.09871875 1.044104365 6762355 Tmem183a 0.09871875 1.227058832 6859415 Zfp397 0.09871875 1.286213235 6862627 Zfp516 0.09871875 1.940285837 6753091 Elk4 0.09986983 1.358886916  118  Table 4.6: Genes with > 2-fold expression change associated with deletion of Meis1 as determined by Affymetrix analysis Probe Name Description Adjusted t-Test Fold decrease in Meis1-/- 6954615 VAMP5 0.0982 5.7767 6805191 OLFR4 0.0790 3.4010 6790699 HLF 0.0120 3.1807 6790621 MSI2 0.0422 2.0704 6748889 IL18R1 0.0790 0.4877 6959584 TYROBP 0.0959 0.4776 6767782 LILRB4 0.0959 0.2179 6969997 HBB-B1 0.0937 0.2050     A recent study using Meis1 -/- Lin- adult BM cells as input for the Affymetrix array (Unnisa et al., 2012) revealed down-regulation of several gene sets up-regulated in response to hypoxia, including those regulated by Hif1α. This is consistent with recent reports by another group (Simsek et al., 2010; Kocabas et al., 2012) reporting direct regulation of Hif1α by Meis1 in the HSC compartment. We were unable to find evidence of Hif1α loss of expression by RT-Q-PCR in our sorted KSL population. We also looked in the more highly HSC-enriched LSKCD150+CD48- population by Q-RT-PCR and found no difference in expression between MxCre/Meis1-/- and MxCre/Meis-/+ in terns of Hif1α or Hif2α gene expression relative to mAbl (Figure 4.9, panel a).   Treatment with N-acetyl-L-cysteine rescues some of the phenotypic abnormalities seen with loss of Meis1  Recent studies suggest that MEIS1 plays a role in regulation of Hif1α expression and resultant regulation of ROS. It is thus hypothesized that increased ROS levels and resultant damage to HSC may underlie the decrease in HSC number and function upon Meis1 deletion  119 (Simsek et al., 2010; Kocabas et al., 2012; Unnisa et al., 2012). Support for this model has been derived from studies demonstrating phenotypic rescue with N-acetyl-L-cysteine (NAC), a ROS scavenger. Despite a lack of evidence of Hif1α dysregulation at the level of mRNA in our model, we tested whether in vivo administration of NAC could rescue deficits seen in the HSC, MkP and EP populations. PolyI:C mediated Cre expression was initiated in MxCre/Meis1fl/fl and MxCre/Meis1fl/+ mice. On the third PolyI:C injection, daily subcutaneous NAC or PBS injections were initiated (Figure 4.8, panel a). Five days after the final subcutaneous injection of PBS or NAC, mice were euthanized and analyzed for level of Meis1 deletion, phenotype, CFC capacity and gene expression in sorted HSC and MkP populations. Estimated Meis1 deletion was comparable for control and NAC treated mice (97% versus 85% respectively). 3 mice per genotype and per arm were used, however one MxCre/Meis1-/- mouse treated with NAC died prior to completion of the regime. Thus the comparisons using MxCre/Meis1-/- +NAC are estimates of significance as opposed to a statistical comparison. More replicates would ideally be done. Phenotypic analysis of PBS treated mice replicated earlier experiments, that is, there was an 11-fold drop and 7-fold drop in HSC and MkP absolute numbers, respectively in Meis1 deleted mice compared to controls (p=0.001, p=0.05; Figure 4.8, panel b). CMP and GMP numbers were also reduced (14.5-fold, p=0.008; 2.5-fold, p=0.03), however the loss of the MEP population was not seen in this experiment. Strikingly, delivery of NAC abolished any phenotypic differences by FACS between MxCre/Meis-/- and MxCre/Meis1-/+ mice (Figure 4.8, panel c) with the exception of CMPs where there was still a 4.5-fold reduction in MxCre/Meis1-/- marrow (p=0.02).  CFC numbers, however, were not affected by NAC treatment as there was still a significant decrease in detectable CFU-GM (1.5x, p = 0.04) and CFU-GEMM (7.8x, p =  120 0.006) in MxCre/Meis-/- mice treated with NAC compared to MxCre/Meis1-/+ mice, also treated with NAC, in myeloid biased media (Figure 4.8, panel d). Using erythroid supportive conditions, numbers of large BFU-E (>16 colonies) also were not rescued by NAC administration in MxCre/Meis-/- mice compared to MxCre/Meis1-/+  (12.8x reduction, p = 0.03; Figure 4.8, panel e). Significant differences in these groups are based on duplicate methylcellulose cultures per mouse We additionally sorted HSC (LSKCD150+CD48-) and MkPs from MxCre/Meis-/- and MxCre/Meis1-/+ PBS and NAC treated mice 5 days after the final treatment. We wanted to confirm the targets we identified in a more highly HSC-enriched population, examine if NAC had any impact on these genes, as well as if any key expression changes may be relevant in the MkP population. These results confirmed that loss of Meis1 in HSCs results in loss of expression of Hlf and Msi2, candidate genes identified in our earlier Affymetrix analysis. In the more highly enriched LSKCD150+CD48- population, there was a 21-fold drop in Hlf (p=0.03) compared to 3-fold in the LSK population (Figure 4.9, panel a). This enrichment is roughly proportional to the increase in HSC content between the two populations, arguing for a role of Meis1 regulation of Hlf expression in the HSC. Msi2 expression dropped 4.3-fold (p=0.01) in the more enriched HSC population with the loss of Meis1. This is 2-fold more than was identified in the Affymetrix screen, suggesting this gene also plays a role in the HSC population, but likely also as well in non-HSC cells in the LSK population.  Neither Msi2 nor Hlf expression was significantly altered in the MkP population, suggesting regulation of these genes is exclusive to HSC-enriched cell populations as opposed to Meis1-expressing cells in general.  In the presence of NAC, there were still significant changes in Hlf and Msi2 expression in the absence of Meis1, although the  121 magnitude of this change was blunted. That is, there was a 3-fold lower expression of Hlf (p=0.02) and a 2-fold loss of Msi2 expression (p=0.03) in MxCre/Meis1-/- compared to MxCre/Meis1-/+ HSC.    122  a)b)300μg/mouse PolyI:C IPq48 hrsMxCre/Meis1fl/flMxCre/Meis1fl/+SubQ NAC/PBS q24 hrs- FACS for phenotype- Sort for HSC & MkP for gene expression- CFC formationHSC (CD150+ CD48- ) MkP CMP GMP MEP CLP101001000100001000001000000Absolute Number per TrunkMeis1-/- + PBSMeis1-/- + NACMeis1-/+ + PBSMeis1-/+ + NAC*** ** 123  Figure 4.8: NAC treatment partially abolishes differences between MxCre/Meis1-/- and MxCre/Meis1-/+ mice.  a) Experimental design. b) Phenotypic analysis of various cell populations by FACS between MxCre/Meis1-/- and MxCre/Meis1-/+ mice treated with PBS and NAC. The differences c)d)BFU-E (<16)BFU-E (>16)TOTAL1001000100001000001000000Colony Number/ 2xTibia, Fibia, Femur MxCre/Meis1-/- MxCre/Meis1-/+ **BFU-E CFU-GM CFU-GEMMTOTAL1001000100001000001000000Colony Number/ 2xTibia, Fibia, Femur MxCre/Meis1-/- MxCre/Meis1-/+ ** 124 between MxCre/Meis1-/- and MxCre/Meis1-/+ mice treated with PBS are consistent with those found in earlier experiments. Phenotypic differences between MxCre/Meis1-/- and MxCre/Meis1-/+ mice were abolished with NAC treatment with the exception of maintenance of a 4.5-fold reduction in CMP numbers (p = 0.02). c) CFC formation in myeloid-supportive conditions continues to be impaired in MxCre/Meis1-/- mice following in vivo NAC administration. CFU-GM are reduced 1.5-fold (p = 0.04) and CFU-GEMM are reduced 7.8-fold  (p = 0.006). d) NAC does not rescue the capacity for MxCre/Meis1-/- cells to form large BFU-E (>16 clusters/colony) in erythroid supportive media (12.8-fold reduction, p = 0.03). Both FACS data and colony numbers are expressed as absolute numbers isolated from the trunk of mice (2 femurs, 2 tibias, 2 iliac crests).  125  Figure 4.9: Gene expression changes as a result of NAC-treatment in sorted HSC and MkP populations a) Gene expression changes as a result of loss of Meis1 in the sorted HSC population treated with NAC or PBS. In PBS treated mice, there is a significant difference between MxCre/Meis1-/- mice and MxCre/Meis1-/+ mice in the expression of Meis1 (115-fold, p= a)b)Meis1Hbb-b1 Hif1α Hlf Msi2 Flt30.0010.010.1110Gene Expression Relative to mAblMeis1-/- + PBS HSCMeis1-/+ + PBS HSCMeis1-/- + NAC HSCMeis1-/+ + NAC HSC** **** **Meis1Hbb-b1 Hif1α Hlf Msi2 Flt30.00010.0010.010.1110100Gene Expression Relative to mAblMeis1-/- + PBS MkPMeis1-/+ + PBS MkPMeis1-/- + NAC MkPMeis1-/+ + NAC MkP*** 126 0.004), Hlf (21-fold, p=0.03), Msi2 (4.3-fold, p=0.01) and Hbb-b1(8-fold gain, p= 0.02). These differences are maintained with NAC treatment: Meis1 (4-fold, p= 0.04), Hlf (4-fold, p=0.02), Msi2 (2-fold, p=0.03) and Hbb-b1(12-fold gain, p= 0.0007). No differences in expression of these genes were found between PBS and NAC treated MxCre/Meis1-/- mice. b) Very few significant gene expression changes were found between MxCre/Meis1-/- and MxCre/Meis1-/+ treated MkP cells. In PBS treated mice, there was a 47-fold drop in Meis1 expression (p=0.02) and in NAC treated mice, a 8-fold loss of Hlf (p=0.01) and 140-fold gain in Hbb-b1 expression (p=0.01).   Discussion The primary goal of the studies described in this chapter was a better understanding of the roles Meis1 may play in adult hematopoiesis. To this end we have exploited mouse models in which Meis1 can be conditionally deleted using two different Cre induction strategies. Four key findings emerge from these studies. First, Meis1 is essential for the homeostatic maintenance and regenerative capacity of adult HSC. Second, Meis1 has essential roles also in the early steps in the megakaryocytic and erythroid pathways. Third, our results of gene expression analyses point to novel putative effectors of Meis1’s activity.  Fourth, the impact of Meis1 deletion can be blunted using scavengers of reactive oxygen species and thus further implicating ROS regulation as a Meis1 functional pathway.  Meis1 is required for HSC maintenance and self-renewal We founds that Meis1 is required for the HSC maintenance by two lines of evidence. In the absence of Meis1 there was a 5-fold reduction in the numbers of phenotypically defined HSC (LSKCD150+CD48-) in the MxCre model that correlates to a 10.7-fold reduction in functionally defined HSC by long-term repopulating assays. This is supported by a 19-fold reduction in HSC in long-term repopulating assays in the ERTCre model in the absence of Meis1 compared to heterozygous controls. As total bone marrow cellularity is not impacted by loss of Meis1 this corresponds to only 270 or 230 HSC per mouse in the MxCre  127 and ERTCre models, respectively. This is in contrast to Meis1-/+ mice where there are roughly 3000 or 5000 HSC per mouse in the MxCre and ERTCre models, respectively. In a serial transplantation assay we show that Meis1-/- stem cells are almost devoid of self-renewal capacity as they fail to contribute to repopulation into secondary recipients. This is in comparison to Meis1-/+ cells that are able to sustain long-term repopulation in secondary recipients. This represents, at minimum, a 25-fold reduction in self-renewal capacity between Meis1-/- and Meis1-/+ HSC. We additionally show this deficit to be largely cell intrinsic as there is a 6.3-fold drop in HSC frequency in Meis1-/- cells compared to Meis1-/+ transplanted into wild-type recipients prior to induction of allele deletion.   Overall our findings are consistent with two recently published studies using the same Meis1 conditional knock-out allele (Kocabas et al.2012; Unnisa et al, 2012). While Unnisa et al. crossed mice onto the same Rosa26ERTCre strain as some of our studies, Kocobas et al. used the SclERTCre mouse where Cre expression is driven from the Scl promoter which is expressed in primitive hematopoietic, erythroid and megakaryocytic cells as well as in endothelium and specific neural tissues (Bockamp et al. 1995; Elefanty et al., 1999).  This is of interest since although our studies have utilized different induction schemes, there are some key similarities between our studies. Although Unnisa et al. also used the ERTCre model, they used 5 daily IP injections of 1mg 4-OHT or one large 4mg bolus. Of note, they used non-quantitative PCR to document allele deletion and achieved only a 50% reduction in Meis1 mRNA as compared to our consistently >90% deletion. Unnisa et al. also confirmed loss of full-length MEIS1a by Western blotting, although only showed a narrow view of the blot, making a comparison to our study impossible. Kocabas et al. used daily IP injections of 1.2mg 4-OHT for 14 days, also used non-quantitative PCR to  128 document deletion and were able to achieve a similar 90% reduction in Meis1 mRNA to our studies.    Studies by both Unnisa et al. and Kocabas et al. documented a loss of phenotypically and functionally defined HSC similar to our studies. Unnisa et al. used the same markers to our studies and found a 2.5-fold reduction in HSC numbers. Using bulk competitive transplants of 1x106 cells from 4-OHT-treated transgenic mice, they were unable to show significant reconstitution by Meis1-/- cells in recipient mice 4 weeks following transplantation. Kocabas et al. used a similar competitive transplantation experiment using purified Meis1-/- LSKFlk2-CD34- cells enriched for long-term HSC activity and were also unable to document long-term contribution to recipient mouse engraftment. Both groups were unable to detect a significant contribution to engraftment of Meis1-/-, similar to our findings. Notably, in contrast to our studies of engraftment in primary and secondary recipients, neither group used limiting dilution assays and were thus unable to quantify HSC frequency in their studies.  Also of note is that although Kocabas used a different marker subset to define HSC-enriched populations, their studies suggest that there are increased numbers of HSC that are cycling and fail to support long-term engraftment due to loss of quiescence and exhaustion.  Together our results robustly reveal a critical and indisputable role of Meis1 in HSC function. Where our studies differ in terms of HSC function however is with respect to quantitation and interpretation of underlying mechanisms. Both Unnisa et al. and Kocabas et al. devote a number of experiments to documenting a role for ROS in the HSC compartment as well as increased cell cycling. Comparison of our ROS studies are discussed below. Both Unnisa et al. and Kocabas et al. found increased proportions of Meis1-/- cells in G1+S-G2-M  129 using Hoechst/Pyronin in their respective HSC-enriched compartments (LSKCD150+CD48- and LSKFlk2-CD34-, respectively). Unnisa et al. were also able to document a nearly 2-fold increase in bromodeoxyuridine (BrdU)-positive cells in the LSKCD150+CD48- HSC-enriched compartment using in vivo staining.  We were unable to document this using in vivo or in vitro BrdU in our studies. This may be due to our practice of using treated Meis1-/+ mice as controls in our studies. Unnisa et al used both non-Cre expressing and Cre expressing Meis1fl/+ mice as controls, whereas Kocabas et al. also included Meis1+/+ and Meis1fl/+ non-Cre expressing and Cre expressing mice as controls. Inclusion of non-Cre expressing and Meis1+/+ mice as controls does not account for cell cycle changes as a result of DNA damage and repair resultant from Cre-mediated deletion. Cre-mediated deletion of Meis1, involves DNA strand damage, Holliday quadruplex formation and DNA repair by homologous mechanisms (reviewed Craig, 1988). Even in the absence of LoxP sites, Cre expression in mammalian cells leads to reduced cell proliferation and accumulation in G2/M in a Cre-dose dependent manner (Loonstra et al., 2001). It is thus possible that differences in cell cycling in the Kocabas et al. and Unnisa et al. studies are due to a lack of Cre-mediated cell cycle depression in a significant proportion of their controls as opposed to a genuine influence of loss of Meis1. While the major finding of a loss of HSC potential remains consistent among the three studies, certain key differences highlight the need for further study.   Meis1 is required for megakaryopoiesis and erythropoiesis in the adult  While Meis1 expression had been previously implicated in fetal megakaryopoieis (Hisa et al., 2004; Azcoitia et al, 2005), our studies are the first to show a critical role in  130 adult megakaryo- and erythropoiesis. We found a loss in platelets in both our models as well as reductions in RBC numbers in the ERTCre model. Reduced numbers of terminally differentiated mature cell types was supported by a 9-fold reduction in large CFU-Mk colonies and 11-fold loss of phenotypically defined megakaryocyte progenitors. A 6-fold reduction in phenotypically defined megakaryocytic-erythroid progenitors (MEP) was mirrored by a 4-fold reduction in BFU-E colonies in the bone marrow of MxCre/Meis-/- cells compared to Meis1-/+ controls.  The loss of erythroid potential in the absence of Meis1 was exacerbated in the phenylhydrazine (PHZ) model of hemolytic anemia and stress erythropoiesis where the proliferative potential of Meis1-/- erythroid progenitors is severely blunted.  Kocabas et al. also documented a consistent reduction in RBC and platelet numbers in the peripheral blood of Meis1-/- mice and reduction in phenotypically defined CMP, GMP and MEPs. They did not examine lineage-restricted megakaryocytes either phenotypically or functionally and limited their examination of erythroid lineages to a similar finding of reduced BFU-E numbers in the bone marrow of treated mice. Their interpretation was of a pan-lineage reduction in myeloid progenitors due to a loss of primitive HSC. Our finding that Meis1-/- cells can be assayed as CFCs of myeloid, erythroid and mixed lineage may support a model whereby Meis1 is dispensable for lineage-restricted progenitors.  As GMP numbers are unchanged in our hands, however, our findings do not necessarily support this interpretation. Also at odds with this interpretation is a lack of lymphoid lineage perturbation in either of our studies. In addition, in our hands, peripheral blood platelet numbers in both the MxCre and ERTCre models normalize to normal levels 2 weeks after induction, despite persistence of the deleted Meis1 allele, that is, outgrowth of Meis1+ cells alone cannot  131 account for this phenomenon (data not shown).  That Meis1-/- bone marrow cells fail to form large BFU-E colonies following PHZ treatment (73-fold reduction in Meis1-/- spleen cells) may support an alternative model whereby Meis1 expression is required for proliferation of erythroid and megakaryocytic progenitors but is dispensable for differentiation along these lineages. Further studies are warranted to determine if the deficit in these lineages is due solely to a lack of upstream HSC or if lack of proliferation of megakaryocytic and erythroid progenitors plays a more predominant role in this phenotype.   In the ERTCre-model, we found a loss of mature megakaryocytes histologically in the bone marrow of moribund mice that was reflected in a loss of platelet numbers in the peripheral blood of these mice. Interestingly, in the MxCre model, no such loss of mature megakaryocytes was seen histologically, although there was a loss of phenotypically defined megakaryocytic progenitors, primitive colony-forming cells and mature platelets. This phenomenon of altered progenitor number without an apparent reduction of the mature cell type is reminiscent of the role of TGF-β1 in the erythroid population where expression serves to stimulate differentiation (Krystal et al., 1994). In these studies, TGF-β1 in culture lead to impaired expansion of immature progenitor cell types, likely due to early differentiation. Whereas at certain time points the mature cell output was identical between conditions, ultimately the loss of progenitors in culture lead to reductions in overall mature cell output. In other words, when examining a lineage of cells that mature along a progression of stages, where each stage has variable proliferative potential, the distribution of cell stages will be heavily influenced by the time at which you take your measurement.  A similar phenomenon could account for the lack of evidence of mature megakaryocyte loss in the marrow of MxCre/Meis1-/- mice, despite reductions in mature platelets and progenitors. Although  132 primitive cell number and mature megakaryocyte platelet output is reduced, intermediary mature megakaryocyte number may appear grossly normal. Histological evaluation at a later time point or following forced expansion in response to stress may reveal a loss of mature megakaryocytes in MxCre/Meis1-/- mice, similar to ERTCre/Meis1-/- mice.  Hlf and Msi2 are putative effectors of Meis1 function in adult hematopoiesis Overall, we identified loss of expression of 4 genes (Hlf, Msi2, Olfr4-2 and Vamp5) and gain of expression of 4 genes (Il18r1, Tyrobp, Lilrb4, and Hbb-b1) in response to loss of Meis1 that met both criteria of a >90% confidence interval and >2-fold change. Meis1 has previously been shown to be a transcriptional activator, thus, up-regulated genes in this model do not likely represent direct Meis1 targets. Up-regulated genes in our data set are primarily implicated in cell maturation and immune response, and are likely reflective of the loss of potential in the population studied and expression of Cre recombinase.  Our screen did not identify PF4 as a differentially expressed target in the LSK compartment using Affymetrix analysis (Okada et al., 2003). This is likely due to the fact that cells with megakaryocytic potential are found within the Sca1- compartment, which was not evaluated in this study. Flt3, a defined but dispensable target for Meis1 activity in leukemia, was found to be less-expressed in the absence of Meis1 by RT-Q-PCR in our study. This loss of signal was not apparent in the Affymetrix analysis however. This may be a function of the high litter-to-litter variability in sample replicates as a result of using individual mice for each sample as opposed to a pooled population. It would be of interest to perform additional individual mouse replicates to reduce the signal to noise ratio between changes as a result of individual variability and those as a result of true loss of expression of Meis1.   133 Olfr4-2 (re-designated Olfr153 May 2005), which encodes an olfactory receptor, and Vamp5, a plasma membrane protein (Entrez Gene results) were found to have lower expression with loss of Meis1 by Affymetrix analysis. Neither have been implicated in hematopoiesis to date. Olfr4-2 is not expressed at appreciable levels in HSC-enriched tissues based on information from the Immunological Genome Project (www.immgen.org; Jojic et al., 2013), whereas Vamp5 is expressed at ~1/3 the level of Meis1. The olfactory receptors share high sequence similarity, but there is also some homology (77% identity over 211 base pairs) to segments of Pdgfra cDNA, which was also down-regulated in our screen in response to loss of Meis1, however it did not reach statistical significance. This highlights a limitation of Afffymetrix gene analysis, that is, differentially expressed genes are identified by virtue of concurrent changes in hybridization of segments along a gene. Thus, transcripts with high homology are difficult to differentiate, especially transcripts with relatively low abundance.  Detection of significant changes in expression of transcripts at low abundance is one of the main weaknesses of hybridization-based gene expression arrays.  In response to this, leading-edge analysis of gene sets (Gene Set Enrichment Analysis/GSEA) was developed (Subramanian et al., 2005) in which sets of genes with associated functions/pathways are linked together based on a priori knowledge. Although individual genes in the set may not reach statistical significance, consistent changes in a gene set with related function may provide clues into differentially regulated processes between samples. In our analysis, although no cell cycle genes reached statistical and fold-change significance in isolation, GSEA was significant for cell cycle data sets, supporting a role for Meis1 in cell cycle.   134 Our Affymetrix analysis and RT-Q-PCR also identified Hlf as a possible effector of Meis1 function in an HSC-enriched population. HLF has been previously implicated in leukemia as a fusion with E2A in B-precursor ALL (de Boer et al., 2011) as well as direct target of Meis1 in Hox-mediated transformation (Argiropoulos et al., 2010). E2A-HLF translocations are also thought to mediate leukemogenesis via Hox-independent mechanisms (Ayton & Cleary, 2003), highlighting Hox-independent functions for Meis1-mediated leukemogenesis. Hlf is differentially methylated and silenced through differentiation (Ji et al, 2010) and expression is enriched in the HSC population (www.immgen.com), supporting a possible role in HSC maintenance. Evidence for direct binding and regulation of Hlf by Meis1 in leukemia (Argirpoulos et al., 2010) and our work suggests that Meis1 may mediate possible Hlf function in the HSC. Interestingly, Lmo2, a transcription factor required for hematopoietic development in the embyo, is thought to be regulated by both Hox and Hlf (de Boer et al., 2011; Calero-Nieto et al., 2013). This highlights the complex interplay between transcription factor regulation in the HSC.  Msi2 was also identified as differentially expressed in our screen of Meis1 targets. Msi2 has been gaining attention in recent years due to increasingly evident roles in the maintenance of HSC repopulation potential (Hope et al. 2010) and AML prognosis (Byers et al., 2011). Msi2 was also identified as part of a gene signature characterized by persistent Vp16-Meis transactivation in Hox models of leukemia (Wang et al., 2006). The gene was first characterized in D. melagnoster as a regulator of asymmetric cell fate and may be involved in maintaining HSC quiescence via regulation of Hes1 expression, a downstream effector of Notch signaling. Regulation of Msi2 may be partially responsible for the deficits in BFU-E expansion in our studies as Msi2 selectively expressed in cycling LT-HSC and  135 down-regulated with differentiation, with the exception of re-expression in the BFU-E (Hope et al., 2010).  Regulation of Hlf and Msi by Meis1 is also supported by ChIP-Seq experiments in our lab using a Hox-Meis1 overexpression model (Yung et al., unpublished data) that demonstrates MEIS1 binding in the body of Msi2 and in the transcription start side of Hlf. We examined changes of Hlf and Msi expression in highly purified populations of HSCs (LSKCD150+CD48-) and MkPs. These genes were consistently changed with loss of Meis1-expression in the HSC population, but not the MkP population, supporting a key role for Hlf and Msi2 as effectors of Meis1 function in the HSC compartment.  Regulation of ROS may play a role in Meis1 function  Although ROS and relative hypoxia had been previously implicated in HSC function (reviewed in Sardina et al., 2012), it was while examining the metabolic state of LT-HSC (as defined as Lin-Sca1-c-Kit+CD34-Flk2- in their studies), that Simsek et al. (2010) first drew a link between ROS and Meis1. In this work, they showed that LT-HSC could be enriched on the basis of low metabolic activity and that this fraction was enriched for Hif1α protein, a mediator of adaptation to low oxygen environments. In addition, they highlighted a conserved MEIS1 binding site in the promoter of Hif1α and demonstrated regulation of HIf1α expression by MEIS1 through luciferase and shRNA experiments. A role for Meis1 in hypoxia tolerance was a novel discovery and triggered subsequent work looking at Hif1α expression and ROS regulation by Meis1 in the HSC compartment (Kocabas et al.2012; Unnisa et al, 2012, this work).  Our studies show that treatment with the ROS scavenger N-acetyl-L-cysteine (NAC) abolished any phenotypic differences between MxCre/Meis1-/- and MxCre/Meis1-/+ mice in  136 HSC and MkP-enriched populations. Differences in CMP numbers remained unchanged, however, consistent with the lack of CFC rescue in vivo in our studies. This is in contrast to rescue of CFC formation from Lin- cells in vitro found by Unnisa et al. using in vitro NAC treatment and deletion of Meis1. Using this model of in vitro deletion, they also performed Affymetrix analysis to show down-regulation of hypoxia-related gene sets and loss of Hif1α at the mRNA and protein level. The differences in our results may be rooted in both the experimental conditions as well as interpretation of NAC activity exclusively due to alteration of ROS levels.  Hif1α dimerization to Hif1β is thought to trigger gene expression programs that allow adaptation to hypoxic environments. These programs are hypothesized to protect against the DNA damage caused by high ROS levels in the hypoxic environment that lead to HSC apoptosis. This theory has been supported by studies using NAC as an ROS scavenger. NAC provides a source of glutathione as a substrate for degradation reactions of peroxide (H2O2) into O2 and H2O and thus reduces ROS levels following cellular treatment. Several studies have used restoration of HSC function with NAC treatment to infer ROS scavenging rescues the deficient phenotype (Tothova et al. 2007; Kocabas et al.2012; Unnisa et al, 2012).  This is problematic, however, as NAC activity is not isolated to this role. NAC has been shown to modify the activity of key cell signaling and cycle cascade members Raf-1, MEK and ERK1/2, independently of its effect on ROS scavenging to promote cell survival (Zhang et al., 2011). Thus it is difficult to interpret if rescue using NAC is a result of reduction of ROS species or stimulation of the survival and cycling cascades influenced by Raf-1, MEK and ERK1/2.  137 Interplay between ROS, the niche, cell signaling and differentiation may also be of import in megakaryocyte commitment and differentiation. Recent evidence suggests that ROS may be a signal for megakaryocyte lineage commitment and proliferation from the HSC progenitor through activation of the MEK-ERK1/2 pathway (reviewed in Chen et al., 2013). In our study with in vivo NAC treatment, NAC treatment abolished HSC and MkP differences between Meis1-/+ and Meis-/- mice and perhaps increased HSC and MkP numbers in NAC Meis-/- mice compared to PBS treated controls. As both ROS and NAC treatment stimulate MEK-ERK1/2 in the MkP, it is difficult to infer whether Meis1 regulation of Hif1α plays any role in the MkP homeostasis. In our analysis of gene expression in the MkP population, no significant differences in Hif1α or Hif2α expression was found with either PBS or NAC-treatment between MxCre/Meis1-/- and MxCre/Meis1-/+mice.    It is of note that in our gene expression studies, Hif1α expression was found to be unchanged in our HSC-enriched LSK population by Affymetrix, nor by directed RT-Q-PCR with loss of Meis1. This could be due to a technical problem in our studies or low levels of bona-fide HSC in the KSL population. Most LSK cells (~49/50) are not HSCs. Thus, lack of a difference between Meis-/+ and Meis1-/- LSK cells does not preclude the possibility of cell cycle differences in the HSC compartment that we are not able to detect in this design. RT-Q-PCR in the more highly enriched LSKCD150+CD48- population, however, also did not reveal changes in Hif1α expression.   It is possible that our studies failed to show Hif1α changes in response to Meis1 levels due to sample heterogeneity. Our studies use individual mice as replicates, as opposed to pooled individuals. We additionally used Cre-expressing Meis1-/+ mice as controls, whereas other studies looking at ROS and Meis1 used a collection of Cre+/Meis+/+, Cre+/Meis1-/+ or  138 Cre-/Meis1fl/fl mice, raising the possibility of gene dosage differences between Meis1-/+ and Meis1+/+ mice that blunts detection of differentially expressed genes. This may be especially true for genes expressed at relatively low levels. For example, Hif1α is expressed at 10x lower levels than Meis1 in the LT-HSC (www.immgen.org). Unnisa et al. showed enrichment for hypoxia gene signature, however, this was in a bulk Lin- population which is ~10x enriched for HSC in bulk marrow compared to ~2000x for the LSK population used in our study.  An intriguing possibility is that while Meis1 may be able to trigger Hif1α expression in experimental models, this may be a coincident change but not a particularly physiologically relevant one.  While phenotypic rescue by NAC may apparently support a role for Meis1 in hypoxia regulation, the varied effects of NAC makes this a somewhat tenuous conclusion. Use of specific MEK and mTOR inhibitors that would isolate general ROS changes from ERK1/2 effects would be useful in this regard.  Summary  Overall, we have validated a conditional model of Meis1 deletion in the adult mouse that has lead to greater understanding of roles for Meis1 in the adult HSC and novel roles in adult megakaryopoiesis and erythropoiesis. We identified 2 novel effectors of Meis1 in adult, HSC-enriched populations, that is Msi2 and Hlf. While we demonstrate that NAC treatment in vivo is able to rescue of some of the phenotypic changes seen as a result loss of Meis1, the varied effects of NAC outside of ROS scavenging and lack of documented changes of Hif1α make it difficult to ascertain if this is due to a true influence of Meis1 on ROS levels.  Our studies demonstrate the power of the conditional model of Meis1 deletion as hypothesis  139 generating as well as testing. Further studies using the model and complementary methods will be invaluable for future study of Meis1 function and resolving remaining key questions for its role in defined hematopoietic subsets.     140 Chapter 5 : Discussion  Introduction   While up-regulation of Meis1 is a relatively common occurrence in leukemia, previous to the generation and distribution of a conditional knock-out mouse (Drs Jenkins & Copeland), a role for Meis1 in normal adult hematopoietic homeostasis remained unknown. The concurrent validation of the model by our group and others has lead to a significant advance into our understanding into Meis1 function and revealed some surprising putative mechanisms, including possible regulation of ROS (Kocabas et al.2012; Unnisa et al, 2012). For our part, the evidence presented in this work validates a conditional knock-out model to study the role of Meis1. Our subsequent studies using this model support a key role for Meis1 in HSC maintenance and revealed novel roles in the maintenance of adult megakaryopoiesis and stress erythropoiesis. We show that in the absence of Meis1, there is a dramatic loss of HSCs, CMPs, MEPs and MkP phenotypically. These losses were reflected functionally in a number of assays. The MkP defect was seen as a reduction in platelet numbers in vivo as well as a reduced numbers of CFU-Mk in the bone marrow of MxCre/Meis1-/- mice. There was a loss of RBC in the peripheral blood of ERTCre/Meis1-/- mice and in MxCre/Meis1-/- BFU-E numbers as well as proliferative potential in response to PHZ induced stress. HSC number and self-renewal capacity is all but obliterated in the absence of Meis1 based on quantitative limiting dilution CRU assays. In addition, we have used microarray expression analysis as a hypothesis-generating tool to identify Msi2 and Hlf as possible effectors of Meis1 in adult hematopoiesis, roles that have not been previously appreciated for these genes. This highlights the power of a conditional model of Meis1 deletion in order to appreciate its contribution to a number of processes including normal development, adult homeostasis and  141 leukemia. The following discussion highlights some of these unresolved mechanisms and proposes experimental models in which these questions could be studied. These are grouped into the study of leukemic processes, revision of the canonical hematopoietic hierarchy, resolution of the role of Meis1 deficiency in cell cycle regulation, and Meis1 regulation of ROS via Hif1α.   Future avenues for research using the model outside of normal adult hematopoiesis  to delineate roles for Meis1 in leukemia  Several lines of evidence suggest Meis1 plays a significant role in the generation of leukemia. Notably, it is up-regulated at high frequency in human hematologic malignancies and overexpression in immature mouse hematopoietic cells, be it in conjunction with Hox or as an MLL fusion co-expression, invariably causes leukemia. It is thus of interest to determine at which stages of leukemogenesis Meis1 expression is essential and if disruption of Meis1-regulated expression profiles has significant impact on the generation or maintenance of the leukemic population, most importantly in the LSC population. Although not reported in this work, we were able to generate leukemic cell lines on both the ERTCre/Meis1fl/fl and MxCre/Meis1flfl backgrounds using retroviral overexpression of genes of interest in leukemia including Nup98-Hox fusions, MN1 and MLL-AF9. We were able to demonstrate both in vitro and in vivo deletion of Meis1 using these lines. Other lab members have gone on to optimize generation and Meis1 deletion in these lines. Experiments that could be envisaged using this model include both qualitative and quantitative studies as well as regulation of gene expression in the leukemic context by Meis1. Depending on when the Meis1 is deleted in the target cell relative to the introduction of the oncogenes of interest,  142 it could be elucidated whether Meis1 is important for the generation or maintenance or both of leukemic potential using both in vitro assays, including serial re-plating of CFCs, and in vivo transplantation experiments. Other groups have had success enriching for LSC potential on the basis of immunophenotyping and FACS (Bonnet and Dick, 1997; Gibbs et al., 2012). Current studies into Meis1 in the leukemic setting rely on overexpression and analysis of bulk populations with leukemic potential. It would be exciting to enrich for the LSC on the basis of phenotype then delete Meis1. One could then examine the functional impact on the LSC following the loss of Meis1 as well as identify gene targets in this population. One could also monitor possible compensatory changes in gene expression following the loss of Meis1 that maintain leukemic potential.  Although not reported in this work, preliminary in vivo experiments using MN1-overexpression demonstrated that leukemia could be maintained in the absence of Meis1, however these bulk populations showed increased expression of Meis3 relative to the input population. This raises the possibility that other MEIS/PREP family members can compensate for loss of Meis1 in some contexts, a possibility that bears investigation as it may limit the effectiveness of Meis1 knock-down as a potential therapeutic target.  Erosion of regulation of cellular division is considered to be a key step in the generation of leukemia as the overgrowth of abnormal cells stifles normal cell function. Our Affymetrix expression profiling of Meis1-/- suggests upregulation of a subset of genes involved in cell cycle, although we could not measure any differences in cell cycle functionally.  In a dominant-negative model of Meis1 function in the context of Hox-overexpression, Argiropoulos et al., showed cell cycle arrest in G1 to S phase, in part due to loss of cyclin D3 (Argiropoulos et al., 2010).  In contrast, studies using a conditional model  143 of Meis1 deletion in vivo, loss of Meis1 is associated with loss of quiescence and cell cycle entry (Kocabas et al., 2012; Unnisa et al., 2012). This raises the possibility of opposite functions of Meis1 in cell cycle regulation depending on the context.  Possible explanations for this are not readily apparent as the key effectors of both HSC and LSC potential, that is, Hox genes, are expressed at high levels in both contexts. Interestingly, although loss of Meis1 is detrimental to the survival of both the LSC and the HSC, it does raise the possibility that this differential response to cell cycling could be exploited for therapeutic gain. For example, Pten deficiency leads to both the development of LSC and loss of HSC through increased cycling, however, treatment of cells with an inhibitor of the downstream pathway lead to selective preservation of the HSC population at the expense of the LSC (Yilmaz et al., 2006). Identifying such a differentially expressed downstream target in the context of Meis1-associated leukemias could lead to the development of therapeutics with improved LSC kill with fewer adverse side-effects, such as neutropenia.  Power to delineate roles for Meis1 during mammalian development One of the first studies using a conditional model for deletion of Meis1 was performed by Azcoitia et al. (Azcoitia et al., 2005) who used in-frame fusion of Meis1 to the modified estrogen receptor ERT2. Theoretically, in the absence of 4-OHT, the fusion protein would be sequestered in the cytoplasm and thus be inactive as a transcription factor. Using this model, they found similar results to previous work by Hisa et al. (Hisa et al., 2004), that is, embryos homozygous for loss of Meis1 died between day 11.5 and 14.5pc from hemorrhaging due to loss of megakaryocytes and vascular patterning deficits. Both studies also noted a reduction in HSC potential in the fetal liver of these mice, where HSC activity is localized at this developmental stage. For Azcoitia et al. administration of 4-OHT in vivo to  144 the pregnant mouse would in principle rescue the phenotype of these mice by permitting nuclear localization of the fusion MEIS1 protein. The group was, however, unable to demonstrate nuclear localization of the protein nor phenotypic changes in response to 4-OHT, making further experiments into the specific developmental time points where Meis1 is required for the various cell types unfeasible. Our model systems provide the opportunity to further examine the time points at which Meis1 is required for given cell types in further detail. ERTCre expression mice can be induced in gravid mice (Danielian et al., 1998), thus administration of 4-OHT at different d.p.c. would allow determination  at which d.p.c. is Meis1 no longer required for proper vascular endothelial cell patterning, for example, which may be different than that required for HSC function. Also of note, the fetal liver between days 10.5 p.c. and birth is where the bulk of functionally complete HSC activity is found, that is, HSC that are able to give rise to the mature lymphoid and erythroid cell types that typify the organism at birth (Reviewed in Orkin and Zon, 2008). More immature HSC potential, however, is found in the aorta/gonad/mesonephros (AGM) region days 8.5 thru 11.5 p.c. and it is thought that these cells migrate to the fetal liver and mature into full HSCs. Previous studies were unable to distinguish if Meis1 was essential for HSC development or maturation or both as Meis1 was inactivated from conception. Using ERTCre/Meis1fl/fl mice, it would be possible to administer 4-OHT at time points such as day 5, 8.5 and 10.5 p.c. and measure HSC potential in both the AGM and fetal liver at day 12 p.c. to determine the relative influence of loss of Meis1 on the populations. This information would contribute to the body of work concerning mouse development and gene expression profiles required in given cell types for the appropriate growth of the organism.   145  Insights gained and key topics for resolution highlighted for the role of Meis1 in normal hematopoiesis by the model  Alternative models of the hematopoietic hierarchy Recent studies have proposed alternative routes of differentiation of various hematopoietic cells from the HSC. These modifications suggest not all differentiated cell types arise from successively more committed progenitors, but may in fact, also or alternatively arise directly from extremely primitive progenitors (Adolfsson et al. 2005; Arinobu et al. 2007; Pronk et al., 2007; Boyer et al., 2012) (Figure 5.1b). Supporting the latter model is the maintenance of CD150 expression in the MkP compartment (Pronk, 2007) and significant overlap in concurrent hematopoietic and megakaryocytic defects in various mouse knock-out models such as c-mpl, SCL and Evi-1 (Murone et al., 1998; Mikkola et al., 2003, Goyama et al., 2008). Another model additionally suggests a minimum of two distinct progenitor subsets with myeloid potential derived from the short-term HSC – one with traditional CMP potential and one with both GMP and CLP (the GMLP or LMPP) potential and the other with MEP and GMP potential (Adolfsson et al., 2005) (Figure 5.1c).  The lack of gross lymphoid, monocytic and granulocytic perturbation in the absence of Meis1 in our studies, in the face of significant impairments of megakaryocytic/erythroid potential and specific HSC defects may support these proposed changes to the dogma of the hematopoietic hierarchy. Our studies using inducible models of Meis1 deletion hint at a minimum of 2 possible routes to generation of megakaryocytic and erythroid potential, one through the canonical CMP to GMP/MEP transition and the other directly from a HSC-like progenitor. Both CMP and MEP numbers were diminished in our studies while granulocytic/monocytic potential remained unchanged. This could be due to exclusive GMP  146 commitment from the CMP in the absence of Meis1, which supports both the canonical stepwise differentiation model and the possibility that MEPs can arise directly from the primitive HSC. Loss of Meis1 in our model could impact several independent progenitor types from which cells with megakaryocytic and erythroid potential could be derived. Induction of Cre expression in sorted progenitor populations and subsequent in vitro differentiation and/or in vivo transplantation may help to resolve some of these questions.  Experimental evidence calling for revised models of the hierarchy have been criticized for using artificial in vitro models to draw conclusions about in vivo cellular behavior.  Use of the conditional model of Meis1 deletion in conjunction with lineage tracking mice, such as those that express a fluorescent marker upon promoter expression (Boyer et al. 2012) could help elucidate if there are multiple progenitor types from which megakaryocyte/erythroid potential can be derived, in contrast to the canonical model. For example, Boyer et al. used Flk2/Flt3 driven Cre expression to demonstrate that nearly all platelet potential develops through progenitors that do not express Flk2/Flt3. This suggests that MkP are derived from a cell type closely related to the HSC and not through a CMP, which expresses Flk2/Flt3.  In our model, very few mice died in the absence of Meis1, platelet defects in the peripheral blood were transient (no difference 2 weeks after the final Cre induction dose, data not shown) and BFU-E formed in the absence of Meis1 despite reduced erythroid and megakaryocytic potential.  Given the magnitude of deficit for CFU-Mk and BFU-E in our model, one could argue that in vivo, pathways that bypass the Flk2/Flt3+ CMP may be favored and that possibly this mechanism may not require Meis1 expression.  By deleting both Meis1 and tracking progression through Flk2/Flt3 progenitor stages, one will be able to  147 monitor if megakaryocyte and/or erythroid potential that remains in our deleted mice is due to progression through the CMP progenitor. We additionally were not able to examine if the CFU-Mk were deleted for Meis1 in our assays due to the fixation method required for enumeration. Thus, although both CFU-Mk and BFU-E potential and the phenotypic MEP were reduced in Meis1-/- mice, we could not confirm that CFU-Mk were deficient for Meis1. If MkP escaped Cre-mediated deletion, this could raise the possibility that in the adult Meis1 is required for MkP generation from a HSC progenitor exclusively, while erythroid potential that depends on Meis1 arises through the CMP and not the HSC. Future studies to examine this could be done whereby MkP are isolated from Meis1-/- mice, the percent deletion in the population examined and CFU-Mk per MkP examined. Alternatively, MEPs could be sorted from Meis1fl/fl mice, deleted in vitro and examined for CFU-E and CFU-Mk potential. This could be compared to CFU-E and CFU-Mk potential from in vitro deletion at the HSC stage. Failure to form either lineage in the given populations could provide clarity for cells of origin of these progenitors in the adult mouse. These studies could provide insight into the conflicting evidence from the ES and zebrafish models for the relative contribution of Meis1 into each of these lineages as well (Cvejic et al., 2010; Cai et al., 2012).    148  Figure 5.1: Alternative models of the hematopoietic hierarchy.  a) Canonical view of the hematopoietic hierarchy as a step-wise transition through progenitors to unique differentiated cell types. b) Update to canonical model postulated by Adolfsson et al. 2005, whereby a lymphoid-primed multipotent progenitor (LMPP) exists that retains both granulocytic/monocytic and lymphoid potential. An additional modification suggested by Pronk et al., 2007 that allows for derivation of MEP potential without sequential differentiation through the CMP. c) Possible revision to the model that could be tested experimentally using a combination of conditional Meis1 mice and complementary methods, whereby MkP potential could be derived independently from the HSC or other primitive progenitor, bypassing the MEP and CMP stages.   HSCMulti-potent progenitorsCLPCMPGMPMEPHSCMulti-potent progenitorsCLPCMPGMPMEPLMPPHSCMulti-potent progenitorsCLPCMPGMPMEPLMPPEryPMkPa) b)c) 149 Genetic programs underpinning cell fate decisions at the megakaryocytic-erythroid junction Although Meis1 had previously been thought to be important, at minimum during development, in the erythroid and megakaryocyte lineages, few effectors of Meis1 function are known in these cells. Transcription and epigenetic regulation by Lsd1/Gfi1/Gfi1b complexes have been proposed as a mechanism for Meis1 regulation in erythroid and megakaryocyte progenitors. Histone demethylase LSD1 interacts with repressive transcription factors GFI1 and GFI1B to inhibit expression of Hoxa9, Pbx1 and Meis1 in hematopoietic lineages (Horman et al., 2009; Sprüssel et al., 2012; Chowdhury et al., 2013). A recent model suggests than in combination with LSD1, GF1B regulates Meis1 expression in the erythroid lineage, while Gfi1 does so at the CMP to GMP/MEP transition (Chowdhury et al., 2013). This model is problematic, however, as it suggests that Meis1 regulation does not impact the megakaryocyte lineage, in contrast to our experimental evidence. In addition, another study using LSD1 knock-down showed increased Gfi1b in both the megakaryocyte and erythroid populations and subsequent expansion of both populations (Sprüssel et al., 2012). The inconsistency with the proposed model and various lines of evidence are likely as a result of the use of fetal liver primary cells and cell lines.   Our results support a role for Meis1 in both megakaryocytic and erythroid progenitor potential. Of interest, our results do not necessarily refute the results of Chowhudry et al., which suggests Gfi1b regulation of Meis1 is not at play in the megakaryocyte, despite changes in expression of Gfi1b and megakaryocyte potential in response to knock-down of Lsd1.  Regulation of Meis1 expression in very likely specific to the cell type in question and it is increasingly thought that megakaryocytic potential can be derived from both HSC and MEP populations (discussed below). Although some common regulators may be present, it is  150 possible that differentiation to the megakaryocytic lineages from these populations employs a unique transcriptional cascade. Role of Meis1 in normal cell cycle Given the strong implication for Meis1 in cell cycle regulation in leukemic contexts as well as in recent work using the same knock-out model, it was somewhat surprising we could not identify changes in cell-cycle in vitro or in vivo, with the exception of enrichment for cell cycle gene sets by Affymetrix. Technical issues could be at the root of this, as well as some key differences in leukemic overexpression studies and conditional models of Meis1 excision. Leukemia models implicating Meis1 in cell cycle come from over-expression or shRNA studies to determine dynamic changes in cell cycle. In addition, as discussed in chapter 4, Cre mediated deletion and associated DNA repair is associated with reduced cell proliferation and accumulation in G2/M in a Cre-dose dependent manner (Loonstra et al., 2001).  Both Unnisa et al. and Kocabas et al. demonstrated increased numbers of cells in the S-G2-M compartments in Meis1-/- cells compared to controls. Their interpretation was loss of quiescence in the HSC compartments in the absence of Meis1, which is somewhat inconsistent with increased cell cycling found in Meis1-overexpressing cells. While context and cell specific regulation of cell cycling may be a play, a far more mundane explanation may account for the differences between our studies. Both groups used a combination of mouse genotypes as controls, including Cre+Meis1+/+, which would not be expected to have high levels of DNA damage and repair as there are no LoxP sites for targeted recombination. Thus, significant differences in cell cycle may be due to the relative level of DNA damage and repair between the two groups as a result of Cre excision, versus a genuine effect of loss  151 of Meis1 expression. Further in vivo replicates comparing Cre+Meis1-/+ mice to Cre+Meis1-/- mice would help resolve this, as would in vitro experiments with ERTCre/Meis1 mice.  Further work in our lab has validated in vitro Cre expression and deletion of Meis1 in a titration series. Use of varying doses of 4-OHT and hence Cre expression using ERTCre/Meis1fl/fl, ERTCre/Meis1fl/+ and ERTCre/Meis1+/+cells would allow one to investigate changes in cell cycle as a function of both Meis1 expression and DNA damage as a result of Cre expression.  Role of ROS and regulation of Hif1α regulation by Meis1  While strong evidence for regulation of Hif1a by Meis1 has been shown recently with condition models of Meis1 deletion (Simsek et al., 2010; Kocabas et al, 2012; Unnisa et al., 2012), we were unable to show changes in Hif1a expression by RT-Q-PCR nor by gene set enrichment analysis from the Affymetrix array. This disparity may be, at least in part, due to the actions of NAC beyond providing a substrate for reduction of intracellular ROS species. In our hands, CFC formation was not rescued by NAC administration in vivo, whereas Unnisa et al. used in vitro deletion and found CFC formation capacity was restored by the addition of NAC to cultures. CFC formation from progenitors in vitro would not necessarily use a Hif1α- dependent pathway to minimize ROS as these assays occur in normoxic conditions. They were also able to show, however, rescue of CFC using sh-RNA against VHL1, which would be expected to increase Hif1α levels as VHL is a principle mediator of Hif1α degradation. In our experiments NAC is postulated to mimic the downstream effects of Hif1α activation, whereas sh-VHL would be expect to increase levels of Hif1α itself. Thus, while NAC may function to activate MEK/ERK1/2 both in vitro and in vivo, sh-VHL in vitro may activate Hif1a pathways and confer a survival benefit that is not related to levels of  152 ROS. In an experiment not shown in this thesis, we also attempted to rescue CFC formation in Meis1-/- cells by induction of HIF1α stabilization via culture in hypoxic 5% O2 as opposed to normoxic 20% O2. This condition would also be expected to positively influence HIF1α activity by preventing degradation by the 26S proteasome (Reviewed in Lindsey & Papoutsakis, 2012). In our hands, CFC formation in Meis1-/- was not rescued by culture in 5% O2 conditions, regardless of the addition of NAC. This may be due to insufficient levels of hypoxia as although while 5% O2 is considered to be hypoxic in some studies (Roy et al., 2012), others use 2%, (Perry et al., 2007) or 1% O2 (Couseens et al., 2010).  To clarify the link between actual ROS, Hif1α pathways and Meis1, it would be of interest to isolate HSC from Cre+/Meis1fl/fl mice and introduce a plasmid expressing sh-VHL linked to fluorescent reporter gene. Following induction of Cre expression, the persistence of donor cells could be measured as a function of sh-VHL expression. Various donor cell fractions could then be isolated and the levels of ROS measured by dyes such as DCF-DA and correlated to CFC capacity. If shVHL and non-physiologic expression of Hif1α pathways conferred survival benefit in CFCs independently of ROS, one would expect increased CFC from committed progenitors in the absence of changes in ROS.  It is also difficult to resolve how Meis1-/- deficits could be rescued by administration of a compound following permanent deletion of a gene. For instance, Kocabas et al. completed their induction schemes before administering NAC in vivo. Both our work and that of the other groups support that a loss of Meis1 results in loss of HSC numbers. If the HSC no longer exists in the organism, upon which cell does NAC work to restore function? Possibilities include reprogramming of a downstream progenitor or perhaps rescue of an incapacitated cell with remaining HSC potential. What is additionally possible is that NAC  153 provides an advantage to transient repopulating cells without true HSC potential. Both the work by Kocabas et al. and our own show phenotypic rescue of HSC numbers by FACS, however we attempted to rescue cells with true HSC potential by administering NAC concurrently with the Cre induction scheme. While both our population and that of Kocabas et al. are enriched for HSC potential, cells in the populations are not exclusively HSCs. So while we may have a phenotypic rescue, this does not guarantee true HSC function as defined as long-term (>16 weeks) multi-lineage engraftment.  Kocabas et al. imply rescue of HSC function on the basis of donor cell engraftment at 4 weeks post transplant with no mention of lineage distribution. As discussed in the previous chapter, NAC has impacts on both ROS scavenging and cell signaling and cycle cascade members Raf-1, MEK and ERK1/2, independently of its effect on ROS scavenging to promote cell survival (Zhang et al., 2011). ERK1/2 signaling has also been shown to be important for lineage commitment, differentiation and expansion of erythroid and megakaryocytic progenitors as well as expansion and survival of myeloid progenitors (reviewed in Geest & Coffer, 2009; Chen et al., 2013). Loss of ERK1/2 in the adult mouse results in bone marrow aplasia, anemia, leukopenia, early lethality and loss of the LSK HSC-enriched compartment by FACS (Chan et al., 2013). Other groups have postulated that loss of Meis1 results in loss of Hif1α expression and thus an increase in aerobic metabolism and ROS production (Kocabas et al., 2012; Unnisa et al., 2012). It has been suggested then that ROS scavenging by NAC accounts for the phenotypic and functional rescue of some Meis1-/- deficits in their model system (Figure 5.2). As there is considerable overlap between Meis1 and ERK1/2 activity in the hematopoietic stem cell, megakaryopoiesis and erythropoiesis, it is also possible that the phenotypic rescue seen with  154 NAC in our model is a result of direct activation of ERK1/2 as opposed to elimination of ROS species. Further evaluation of the ERK1/2 knock-out mouse for the nature of the HSC deficit would be valuable, as would be experiments that examined functional and phenotypic rescue of the HSC, megakaryocyte and erythrocyte compartments using ROS scavenging compounds that do not modulate ERK1/2 activity, such as sodium pyruvate (Franco, Panayiotidis & Cidlowski, 2007).  Figure 5.2: Model of NAC activity in the hematopoietic compartment. Meis1 has been implicated in the regulation of Hif1α expression, whose transcription program modulates aerobic vs. anaerobic metabolism. Inhibition of aerobic mitochondrial metabolism by Hif1α programs is thought to result in a reduction in ROS generation. Loss of Meis1, hence Hif1α has been proposed to favor aerobic metabolism and ROS generation that results in a loss of HSC quiescence and self-renewal. NAC is theorized to eliminate these ROS and thus prevent loss of HSC quiescence and function as a result of Meis1 deletion. As NAC also stimulates Erk1/2 activity, loss of which has similar phenotypes in the HSC, megakaryocyte and erythrocyte as loss of Meis1, it is also possible that phenotypic/functional rescue by NAC in Meis1-/- cells is a result of compensatory Erk1/2 activity, versus a loss of ROS.   Confidence in true HSC rescue by NAC, be it via Hif/ROS or other pathways, would be best served by experiments that monitor long-term lineage engraftment and Meis1Hif1αROSMitochondriaNACDNA damageApoptosisLoss of quiescence Erk1/2ProliferationHSC maintenanceMegakaryopoiesisErythropoiesisProliferationLineage commitmentDifferentiation- megakaryopoiesis- erythropoiesis 155 transplantation as well as confirm Meis1 deletion in the repopulating cell compartment. One such experiment would be to perform a similar experiment to the one we performed to examine cell intrinsic/extrinsic roles for Meis1. Cre+/Meis1fl/fl or Cre+/Meis1fl/+ cells could be transplanted into recipients and Cre-expression induced concurrently with NAC/PBS control administration. Maintenance of multi-lineage engraftment as well as Meis1-deletion could be monitored over 16 weeks and/or secondary transplants could be performed to measure HSC maintenance and self-renewal. This type of experiment would irrefutably support NAC rescue of HSC numbers in the absence of Meis1.   HIF1α has been shown to positively influence TERT expression (Nishi et al., 2004; Coussens et al., 2010), a key protein component of the telomerase complex, that has activity in both that complex and independently (reviewed Li et al. 2011). In the context of the hematopoietic stem cell and embryonic stem cells, TERT as a component of the telomerase complex is proposed to maintain chromosomal stability and prevent premature senescence to allow for life-long division via maintenance of chromosomal telomere ends.  It would be of interest to explore levels of Tert mRNA as well as telomerase activity in Meis1-/- hematopoietic stem cells.   HIF pathways additionally regulate EPO synthesis in the spleen and osteoblastic niche to increase RBC number in response to hypoxia (Rankin et al., 2012; reviewed in Lee & Percy, 2011). In addition to a diminished RBC number in the peripheral blood of ERTCre/Meis1-/- mice, hematopoathologist review of H&E staining of the spleen in our moribund mice revealed abnormal iron deposition that was somewhat counterintuitive, given the drop in erythroid progenitors by Ter119-staining (Figure 4.2, panel a). Loss of erythroid progenitors could thus be due to a reduction in Hif1α –regulated Epo expression, as opposed  156 to a hematopoietic cell intrinsic effect. This may implicate Meis1 in the regulation of gene expression in previously unappreciated cell types, that is the osteoblast and splenocyte. We did not examine BFU-E numbers in our cell-extrinsic long-term transplantation experiments. In order to tease the various influences of Meis1 on erythropoiesis, both hematopoietic cell intrinsic and possible extrinsic components, one could measure serum, spleen and marrow EPO levels by Enzyme-Linked-Immunosobent Assays following deletion of Meis1 in  MxCre/Meis1fl/fl or ERTCre/Meis1fl/fl mice transplanted with wild-type bone marrow and examining BFU-E output in the transplanted wild-type cells.     In addition, congenital mutations in VHL, a negative regulator of HIF1α, leads to erythrocytosis, that is, an increase in red cell mass that may be due to increased numbers of red blood cells or iron-containing hemoglobin (Hb) molecules.  Given the accumulation of studies providing support for a network of VHL/HIF/MEIS interaction and the clinical phenotypes associated with mutations in VHL in erythrocytosis, conditional Meis1 deletion mice may provide novel models for the study of this disease.   Conclusion  Our studies have provided strong evidence that Meis1 is a requirement for adult hematopoiesis in terms of maintenance of stem cell numbers and self-renewal as well as in erythropoiesis and megakaryopoiesis. Perhaps most intriguing, however, is how the compilation and comparison of our results with that of others may suggest that further study using the conditional Meis1 knock-out mouse would be of value in elucidating alternative mechanisms of the hematopoietic hierarchy from the canonical model of stepwise differentiation. Use of this Meis1 conditional model will allow interrogation of the possibility  157 of multiple modes of derivation of megakaryocytic potential and if these are obligatorily linked to erythroid potential. In addition, differential response of the HSC and committed progenitor population to the effects of NAC will allow further study into if ROS regulation is at the root of this effect or more so a concomitant change with activation of MEK/ERK1/2 pathways. That cell cycling is apparently increased with loss of expression of Meis1 in normal hematopoietic populations as well as with overexpression in leukemic populations raises the possibility that Meis1 pathways could be differentially targeted for therapeutic benefit in the future. Overall, the work presented in this thesis demonstrates a robust model with which the function of Meis1 in an array of processes can be delineated including development, normal homeostasis, lineage commitment and leukemic processes. Future work using the model outlined in this thesis in conjunction with complementary methods will ultimately be fruitful in delineating these processes. 158 References  Abu-Shaar, M., Ryoo, H. D., & Mann, R. S. (1999). Control of the nuclear localization of Extradenticle by competing nuclear import and export signals. Genes & Development, 13, 935–945.  Adolfsson, J., Månsson, R., Buza-Vidas, N., Hultquist, A., Liuba, K., Jensen, C. T., et al. (2005). Identification of Flt3+ Lympho-Myeloid Stem Cells Lacking Erythro-Megakaryocytic Potential. Cell, 121(2), 295–306. doi:10.1016/j.cell.2005.02.013  Afonja, O., Smith, J. E. J., Cheng, D. M., Goldenberg, A. S., Amorosi, E., Shimamoto, T., et al. (2000). MEIS1 and HOXA7 genes in human acute myeloid leukemia. Leukemia Research, 24, 849–855.  Akala, O. O., & Clarke, M. F. (2006). Hematopoietic stem cell self-renewal. Current Opinion in Genetics & Development, 16(5), 496–501. doi:10.1016/j.gde.2006.08.011  Akashi, K., Traver, D., Miyamoto, T., & Weissman, I. L. (2000). A clonogenic common myeloid progenitor that gives rise to all myeloid lineages. Nature, 404, 193–197.  Alcalay, M., Alcalay, M., Tiacci, E., Bergomas, R., Bigerna, B., Venturini, E., et al. (2005). Acute myeloid leukemia bearing cytoplasmic nucleophosmin (NPMc+ AML) shows a distinct gene expression profile characterized by up-regulation of genes involved in stem-cell maintenance. Blood, 106(3), 899–902. doi:10.1182/blood-2005-02-0560  Antonchuck, J., Sauvageau, G., & Humphries, R. K. (2002). HOXB4-INDUCED EXPANSION OF ADULT HEMATOPOIETIC STEM CELLS EX VIVO. Cell, 109, 39–45.  Argiropoulos, B., & Humphries, R. K. (2007). Hox genes in hematopoiesis and leukemogenesis. Oncogene, 26(47), 6766–6776. doi:10.1038/sj.onc.1210760  Argiropoulos, B., Palmqvist, L., Imren, S., Miller, M., Rouhi, A., Mager, D. L., & Humphries, R. K. (2010a). Meis1 disrupts the genomic imprint of Dlk1 in a NUP98-HOXD13 leukemia model. Leukemia, 24(10), 1788–1791. doi:10.1038/leu.2010.161  Argiropoulos, B., Palmqvist, L., Yung, E., Kuchenbauer, F., Heuser, M., Sly, L. M., et al. (2008). Linkage of Meis1 leukemogenic activity to multiple downstream effectors including Trib2 and Ccl3. Experimental Hematology, 36(7), 845–859. doi:10.1016/j.exphem.2008.02.011  Argiropoulos, B., Yung, E., Xiang, P., Lo, C. Y., Kuchenbauer, F., Palmqvist, L., et al. (2010b). Linkage of the potent leukemogenic activity of Meis1 to cell cycle entry and  159 transcriptional regulation of cyclin D3. Blood. doi:10.1182/blood-2009-06-225573  Arinobu, Y., Mizuno, S.-I., Chong, Y., Shigematsu, H., Iino, T., Iwasaki, H., et al. (2007). Reciprocal Activation of GATA-1 and PU.1 Marks Initial Specification of Hematopoietic Stem Cells into Myeloerythroid and Myelolymphoid Lineages. Cell Stem Cell, 1(4), 416–427. doi:10.1016/j.stem.2007.07.004  Ayton, P. M., & Cleary, M. L. (2003). Transformation of myeloid progenitors by MLL oncoproteins is dependent on Hoxa7 and Hoxa9. Genes & Development, 17(18), 2298–2307. doi:10.1101/gad.1111603  Azcoitia, V., Aracil, M., Martínez-A, C., & Torres, M. (2005). The homeodomain protein Meis1 is essential for definitive hematopoiesis and vascular patterning in the mouse embryo. Developmental Biology, 280(2), 307–320. doi:10.1016/j.ydbio.2005.01.004  Badea, T. C., Wang, Y., & Nathans, J. (2003). A noninvasive genetic/pharmacologic strategy for visualizing cell morphology and clonal relationships in the mouse. The Journal of Neuroscience : the Official Journal of the Society for Neuroscience, 23(6), 2314–2322.  Becker, M. W., & Jordan, C. T. (2011). Leukemia stem cells in 2010: Current understanding and future directions. Yblre, 25(2), 75–81. doi:10.1016/j.blre.2010.11.001  Benjamini, Y., & Hochberg, Y. J. (1995). Controlling the false discovery rate: a practical and powerful approach to multiple testing (Vol. 57, pp. 289–300). R. Stat. Soc.  Berthelsen, J., Kilstrup-Nielsen, C., Blasi, F., Mavilio, F., & Zappavigna, V. (1999). The subcellular localization of PBX1 and EXD proteins depends on nuclear import and export signals and is modulated by association with PREP1 and HTH. Genes & Development, 13(8), 946–953.  Bessa, J., Tavares, M. J., Santos, J., Kikuta, H., Laplante, M., Becker, T. S., et al. (2008). meis1 regulates cyclin D1 and c-myc expression, and controls the proliferation of the multipotent cells in the early developing zebrafish eye. Development, 135(5), 799–803. doi:10.1242/dev.011932  Bisaillon, R., Wilhelm, B. T., Krosl, J., & Sauvageau, G. (2011). C-terminal domain of MEIS1 converts PKNOX1 (PREP1) into a HOXA9-collaborating oncoprotein. Blood, 118(17), 4682–4689. doi:10.1182/blood-2011-05-354076  Björnsson, J. M., Larsson, N., Brun, A. C. M., Magnusson, M., Andersson, E., Lundström, P., et al. (2003). Reduced proliferative capacity of hematopoietic stem cells deficient in Hoxb3 and Hoxb4. Mol Cell Biol, 23(11), 3872–3883.  Bockamp, E. O., McLaughlin, F., Murrell, A. M., Gottgens, B., Robb, L., Begley, C. G., & Green, A. R. (1995). Lineage-restricted regulation of the murine SCL/TAL-1 promoter.  160 Blood, 86(4), 1502–1514.  Bonnet, D., & Dick, J. E. (1997). Human acute myeloid leukemia is organized as a hierarchy that originates from a primitive hematopoietic cell. Nature Medicine, 3(7), 730–737.  Bowie, M. B., Kent, D. G., Dykstra, B., McKnight, K. D., McCaffrey, L., Hoodless, P. A., & Eaves, C. J. (2007). Identification of a new intrinsically timed developmental checkpoint that reprograms key hematopoietic stem cell properties. Proceedings of the National Academy of Sciences of the United States of America, 104(14), 5878–5882. doi:10.1073/pnas.0700460104  Boyer, S. W., Beaudin, A. E., & Forsberg, E. C. (2012). Mapping differentiation pathways from hematopoietic stem cells using Flk2/Flt3 lineage tracing. Cell Cycle (Georgetown, Tex.), 11(17), 3180–3188. doi:10.4161/cc.21279  Brun, A. C. M., Björnsson, J. M., Magnusson, M., Larsson, N., Leveén, P., Ehinger, M., et al. (2004). Hoxb4-deficient mice undergo normal hematopoietic development but exhibit a mild proliferation defect in hematopoietic stem cells. Blood, 103(11), 4126–4133. doi:10.1182/blood-2003-10-3557  Buzzeo, M. P., Scott, E. W., & Cogle, C. R. (2007). The hunt for cancer-initiating cells: a history stemming from leukemia. Leukemia, 21(8), 1619–1627. doi:10.1038/sj.leu.2404768  Byers, R. J., Currie, T., Tholouli, E., Rodig, S. J., & Kutok, J. L. (2011). MSI2 protein expression predicts unfavorable outcome in acute myeloid leukemia. Blood, 118(10), 2857–2867. doi:10.1182/blood-2011-04-346767  Cai, M., Langer, E. M., Gill, J. G., Satpathy, A. T., Albring, J. C., Kc, W., et al. (2012). Dual actions of Meis1 inhibit erythroid progenitor development and sustain general hematopoietic cell proliferation. Blood. doi:10.1182/blood-2012-01-403139  Calero-Nieto, F. J., Joshi, A., Bonadies, N., Kinston, S., Chan, W.-I., Gudgin, E., et al. (2013). HOX-mediated LMO2 expression in embryonic mesoderm is recapitulated in acute leukaemias. Oncogene. doi:10.1038/onc.2013.175  Calvo, K. R., Sykes, D. B., Pasillas, M. P., & Kamps, M. P. (2000). Hoxa9 Immortalizes a Granulocyte-Macrophage Colony-Stimulating Factor-Dependent Promyelocyte Capable of Biphenotypic Differentiation to Neutrophils or Macrophages, Independent of Enforced Meis Expression. Molecular and Cellular Biology, 20, 3274–3285.  Camós, M., Esteve, J., Jares, P., Colomer, D., Rozman, M., Villamor, N., et al. (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 Research, 66(14), 6947–6954. doi:10.1158/0008- 161 5472.CAN-05-4601  Capron, C., Lacout, C., Lécluse, Y., Wagner-Ballon, O., Kaushik, A.-L., Cramer-Bordé, E., et al. (2011). LYL-1 deficiency induces a stress erythropoiesis. Experimental Hematology, 39(6), 629–642. doi:10.1016/j.exphem.2011.02.014  Chan, G., Gu, S., & Neel, B. G. (2013). Erk1 and Erk2 are required for maintenance of hematopoietic stem cells and adult hematopoiesis. Blood, 121(18), 3594–3598. doi:10.1182/blood-2012-12-476200  Chang, C. P., Brocchieri, L., Shen, W. F., Largman, C., & Cleary, M. L. (1996). Pbx modulation of Hox homeodomain amino-terminal arms establishes different DNA-binding specificities across the Hox locus. Mol Cell Biol, 16(4), 1734–1745.  Chang, C. P., Shen, W. F., Rozenfeld, S., Lawrence, H. J., Largman, C., & Cleary, M. L. (1995). Pbx proteins display hexapeptide-dependent cooperative DNA binding with a subset of Hox proteins. Genes & Development, 9(6), 663–674.  Chang, H., Nayar, R., Li, D., & Sutherland, D. (2005). Clonality analysis of cell lineages in acute myeloid leukemia with inversion 16. Cancer Genetics and Cytogenetics, 156(2), 175–178. doi:10.1016/j.cancergencyto.2004.03.017  Chang, M. J., Wu, H., Achille, N. J., Reisenauer, M. R., Chou, C. W., Zeleznik-Le, N. J., et al. (2010). Histone H3 Lysine 79 Methyltransferase Dot1 Is Required for Immortalization by MLL Oncogenes. Cancer Research, 70(24), 10234–10242. doi:10.1158/0008-5472.CAN-10-3294  Chen, S., Su, Y., & Wang, J. (2013). ROS-mediated platelet generation: a microenvironment-dependent manner for megakaryocyte proliferation, differentiation, and maturation. Cell Death & Disease, 4, e722. doi:10.1038/cddis.2013.253  Chen, W., Kumar, A. R., Hudson, W. A., Li, Q., Wu, B., Staggs, R. A., et al. (2008). Malignant transformation initiated by Mll-AF9: Gene dosage and critical target cells. Cancer Cell, 13, 432–440.  Choe, S.-K., Lu, P., Nakamura, M., Lee, J., & SagerstrOm, C. G. (2009). Meis Cofactors Control HDAC and CBP Accessibility at Hox-Regulated Promoters during Zebrafish Embryogenesis. Developmental Cell, 17(4), 561–567. doi:10.1016/j.devcel.2009.08.007  Chowdhury, A. H., Ramroop, J. R., Upadhyay, G., Sengupta, A., Andrzejczyk, A., & Saleque, S. (2013). Differential Transcriptional Regulation of meis1 by Gfi1b and Its Co-Factors LSD1 and CoREST. PLoS ONE, 8(1), e53666. doi:10.1371/journal.pone.0053666  Christensen, J. L., & Weissman, I. L. (2001). Flk-2 is a marker in hematopoietic stem cell differentiation: a simple method to isolate long-term stem cells. Proceedings of the  162 National Academy of Sciences of the United States of America, 98(25), 14541–14546. doi:10.1073/pnas.261562798  Collins, L. S., & Dorshkind, K. (1987). A stromal cell line from myeloid long-term bone marrow cultures can support myelopoiesis and B lymphopoiesis. Journal of Immunology (Baltimore, Md. : 1950), 138(4), 1082–1087.  Coussens, M., Davy, P., Brown, L., Foster, C., Andrews, W. H., Nagata, M., & Allsopp, R. (2010). RNAi screen for telomerase reverse transcriptase transcriptional regulators identifies HIF1alpha as critical for telomerase function in murine embryonic stem cells. Proceedings of the National Academy of Sciences, 107(31), 13842–13847. doi:10.1073/pnas.0913834107  Craig, N. L. (1988). The mechanism of conservative site-specific recombination. Annu. Rev. Genet., 22, 77–105.  Crans-Vargas, H. N., Landaw, E. M., Bhatia, S., Sandusky, G., Moore, T. B., & Sakamoto, K. M. (2002). Expression of cyclic adenosine monophosphate response-element binding protein in acute leukemia. Blood, 99(7), 2617–2619.  Crijns, A. P. G., de Graeff, P., Geerts, D., Hoor, Ten, K. A., Hollema, H., van der Sluis, T., et al. (2007). MEIS and PBX homeobox proteins in ovarian cancer. European Journal of Cancer (Oxford, England : 1990), 43(17), 2495–2505. doi:10.1016/j.ejca.2007.08.025  Cvejic, A., Serbanovic-Canic, J., Stemple, D. L., & Ouwehand, W. H. (2011). The role of meis1 in primitive and definitive hematopoiesis during zebrafish development. Haematologica, 96(2), 190–198. doi:10.3324/haematol.2010.027698  Danielian, P. S., Muccino, D., Rowitch, D. H., Michael, S. K., & McMahon, A. P. (1998). Modification of gene activity in mouse embryos in utero by a tamoxifen-inducible form of Cre recombinase. Current Biology, 24, 1323–1326.  de Andrés-Aguayo, L., Varas, F., & Graf, T. (2012). Musashi 2 in hematopoiesis. Current Opinion in Hematology, 19(4), 268–272. doi:10.1097/MOH.0b013e328353c778  de Boer, J., Yeung, J., Ellu, J., Ramanujachar, R., Bornhauser, B., Solarska, O., et al. (2011). The E2A-HLF oncogenic fusion protein acts through Lmo2 and Bcl-2 to immortalize hematopoietic progenitors. Leukemia, 25(2), 321–330. doi:10.1038/leu.2010.253  Dick, J. E. (2008). Stem cell concepts renew cancer research. Blood, 112(13), 4793–4807. doi:10.1182/blood-2008-08-077941  DiMartino, J. F., Selleri, L., Traver, D., Firpo, M. T., Rhee, J., Warnke, R., et al. (2001). The Hox cofactor and proto-oncogene Pbx1 is required for maintenance of definitive hematopoiesis in the fetal liver. Blood, 98, 618–626.   163 Drabkin, H. A., Parsy, C., Ferguson, K., Guilhot, F., La. (2002). Quantitative HOX expression in chromosomally defined subsets of acute myelogenous leukemia. Leukemia, 16, 186–195. doi:10.1038/sj/leu/2402354  Elefanty, A. G., Begley, C. G., Hartley, L., Papaevangeliou, B., & Robb, L. (1999). SCL expression in the mouse embryo detected with a targeted lacZ reporter gene demonstrates its localization to hematopoietic, vascular, and neural tissues. Blood, 94(11), 3754–3763.  Esparza, S. D., Chang, J., Shankar, D. B., Zhang, B., Nelson, S. F., & Sakamoto, K. M. (2008). CREB regulates Meis1 expression in normal and malignant hematopoietic cells. Leukemia, 22(3), 665–667. doi:10.1038/sj.leu.2404933  Ferrando, A. A., Armstrong, S. A., Neuberg, D. S., Sallan, S. E., Silverman, L. B., Korsmeyer, S. J., & Look, A. T. (2003). Gene expression signatures in MLL-rearranged T-lineage and B-precursor acute leukemias: dominance of HOX dysregulation. Blood, 102(1), 262–268. doi:10.1182/blood-2002-10-3221  Ferretti, E., Villaescusa, J. C., Di Rosa, P., Fernandez-Diaz, L. C., Longobardi, E., Mazzieri, R., et al. (2006). Hypomorphic mutation of the TALE gene Prep1 (pKnox1) causes a major reduction of Pbx and Meis proteins and a pleiotropic embryonic phenotype. Mol Cell Biol, 26(15), 5650–5662. doi:10.1128/MCB.00313-06  Fischbach, N. A. (2005). HOXB6 overexpression in murine bone marrow immortalizes a myelomonocytic precursor in vitro and causes hematopoietic stem cell expansion and acute myeloid leukemia in vivo. Blood, 105(4), 1456–1466. doi:10.1182/blood-2004-04-1583  Franco, R., Panayiotidis, M. I., & Cidlowski, J. A. (2007). Glutathione depletion is necessary for apoptosis in lymphoid cells independent of reactive oxygen species formation. The Journal of Biological Chemistry, 282(42), 30452–30465. doi:10.1074/jbc.M703091200  Frisch, B. J., Porter, R. L., Gigliotti, B. J., Olm-Shipman, A. J., Weber, J. M., O'Keefe, R. J., et al. (2009). In vivo prostaglandin E2 treatment alters the bone marrow microenvironment and preferentially expands short-term hematopoietic stem cells. Blood, 114(19), 4054–4063. doi:10.1182/blood-2009-03-205823  Fulwyler, M. J. (1980). Flow cytometry and cell sorting. Blood Cells, 6(2), 173–184. Gan, B., Sahina, E., Jiang, S., Sanchez-Aguilerac, A., Scott, K. L., Chin, L., et al. (2008). mTORC1-dependent and -independent regulation of stem cell renewal, differentiation, and mobilization. Pnas, 105, 19384–19389.  Garcez, R. C., Teixeira, B. L., Schmitt, S. D. S., Alvarez-Silva, M., & Trentin, A. G. (2009). Epidermal growth factor (EGF) promotes the in vitro differentiation of neural crest cells to neurons and melanocytes. Cellular and Molecular Neurobiology, 29(8), 1087–1091.  164 doi:10.1007/s10571-009-9406-2  Geest, C. R., & Coffer, P. J. (2009). MAPK signaling pathways in the regulation of hematopoiesis. Journal of Leukocyte Biology, 86(2), 237–250. doi:10.1189/jlb.0209097  Gibbs, K. D., Jager, A., Crespo, O., Goltsev, Y., Trejo, A., Richard, C. E., & Nolan, G. P. (2012). Decoupling of tumor-initiating activity from stable immunophenotype in HoxA9-Meis1-driven AML. Cell Stem Cell, 10(2), 210–217. doi:10.1016/j.stem.2012.01.004  Gilliland, D. G., Jordan, C. T., & Felix, C. A. (2004). The Molecular Basis of Leukemia. American Society of Hematology, 1–18.  Goyama, S., Yamamoto, G., Shimabe, M., Sato, T., Ichikawa, M., Ogawa, S., et al. (2008). Evi-1 Is a Critical Regulator for Hematopoietic Stem Cells and Transformed Leukemic Cells. Cell Stem Cell, 3(2), 207–220. doi:10.1016/j.stem.2008.06.002  Grier, D., Thompson, A., Kwasniewska, A., McGonigle, G., Halliday, H., & Lappin, T. (2005). The pathophysiology of HOX genes and their role in cancer. The Journal of Pathology, 205(2), 154–171. doi:10.1002/path.1710  Grubach, L., Juhl-Christensen, C., Rethmeier, A., Olesen, L. H., Aggerholm, A., Hokland, P., & Østergaard, M. (2008). Gene expression profiling of Polycomb, Hox and Meis genes in patients with acute myeloid leukaemia. European Journal of Haematology, 81(2), 112–122. doi:10.1111/j.1600-0609.2008.01083.x  Gurumurthy, S., Xie, S. Z., Alagesan, B., Kim, J., Yusuf, R. Z., Saez, B., et al. (2010). The Lkb1 metabolic sensor maintains haematopoietic stem cell survival. Nature, 468(7324), 659–663. doi:10.1038/nature09572  Halder, S. K., Cho, Y.-J., Datta, A., Anumanthan, G., Ham, A.-J. L., Carbone, D. P., & Datta, P. K. (2011). Elucidating the mechanism of regulation of transforming growth factor β Type II receptor expression in human lung cancer cell lines. Neoplasia (New York, N.Y.), 13(10), 912–922.  Hara, T., Schwieger, M., Kazama, R., Okamoto, S., Minehata, K., Ziegler, M., et al. (2008). Acceleration of chronic myeloproliferation by enforced expression of Meis1 or Meis3 in Icsbp-deficient bone marrow cells. Oncogene, 27(27), 3865–3869. doi:10.1038/sj.onc.1211043  Hayashi, S., & McMahon, A. P. (2002). Efficient recombination in diverse tissues by a tamoxifen-inducible form of Cre: a tool for temporally regulated gene activation/inactivation in the mouse. Developmental Biology, 244(2), 305–318. doi:10.1006/dbio.2002.0597   165 Haylock, D. N., & Nilsson, S. K. (2006). Osteopontin: a bridge between bone and blood. British Journal of Haematology, 134(5), 467–474. doi:10.1111/j.1365-2141.2006.06218.x  Heazlewood, S. Y., Neaves, R. J., Williams, B., Haylock, D. N., Adams, T. E., & Nilsson, S. K. (2013). Megakaryocytes co-localise with hemopoietic stem cells and release cytokines that up-regulate stem cell proliferation. Stem Cell Research, 11(2), 782–792. doi:10.1016/j.scr.2013.05.007  Heine, P., Dohle, E., Bumsted-O'Brien, K., Engelkamp, D., & Schulte, D. (2008). Evidence for an evolutionary conserved role of homothorax/Meis1/2 during vertebrate retina development. Development, 135(5), 805–811. doi:10.1242/dev.012088  Herault, O., Hope, K. J., Deneault, E., Mayotte, N., Chagraoui, J., Wilhelm, B. T., et al. (2012). A role for GPx3 in activity of normal and leukemia stem cells. Journal of Experimental Medicine, 209(5), 895–901. doi:10.1084/jem.20102386  Heuser, M., Argiropoulos, B., Kuchenbauer, F., Yung, E., Piper, J., Fung, S., et al. (2007). MN1 overexpression induces acute myeloid leukemia in mice and predicts ATRA resistance in patients with AML. Blood, 110(5), 1639–1647. doi:10.1182/blood-2007-03-080523  Heuser, M., Yun, H., Berg, T., Yung, E., Argiropoulos, B., Kuchenbauer, F., et al. (2011). Cell of origin in AML: susceptibility to MN1-induced transformation is regulated by the MEIS1/AbdB-like HOX protein complex. Cancer Cell, 20(1), 39–52. doi:10.1016/j.ccr.2011.06.020  Hirose, K., Abramovich, C., Argiropoulos, B., & Humphries, R. K. (2008). Leukemogenic properties of NUP98-PMX1 are linked to NUP98 and homeodomain sequence functions but not to binding properties of PMX1 to serum response factor. Oncogene, 27(46), 6056–6067. doi:10.1038/onc.2008.210  Hisa, T., Spence, S. E., Rachel, R. A., Fujita, M., Nakamura, T., Ward, J. M., et al. (2004). Hematopoietic, angiogenic and eye defects in Meis1 mutant animals. The EMBO Journal, 23(2), 450–459.  Hock, H., Meade, E., Medeiros, S., Schindler, J. W., Valk, P. J. M., Fujiwara, Y., & Orkin, S. H. (2004). Tel/Etv6 is an essential and selective regulator of adult hematopoietic stem cell survival. Genes & Development, 18(19), 2336–2341. doi:10.1101/gad.1239604  Hope, K. J., & Sauvageau, G. (2011). Roles for MSI2 and PROX1 in hematopoietic stem cell activity. Current Opinion in Hematology, 18(4), 203–207. doi:10.1097/MOH.0b013e328347888a  Hope, K. J., Cellot, S., Ting, S. B., MacRae, T., Mayotte, N., Iscove, N. N., & Sauvageau, G. (2010). An RNAi Screen Identifies Msi2 and Prox1 as Having Opposite Roles in the  166 Regulation of Hematopoietic Stem Cell Activity. Stem Cell, 7(1), 101–113. doi:10.1016/j.stem.2010.06.007  Hope, K. J., Jin, L., & Dick, J. E. (2003). Human acute myeloid leukemia stem cells. Archives of Medical Research, 34(6), 507–514. doi:10.1016/j.arcmed.2003.08.007 Horman, S. R., Velu, C. S., Chaubey, A., Bourdeau, T., Zhu, J., Paul, W. E., et al. (2009). Gfi1 integrates progenitor versus granulocytic transcriptional programming. Blood, 113(22), 5466–5475. doi:10.1182/blood-2008-09-179747  Hu, Y. L., Fong, S., Ferrell, C., Largman, C., & Shen, W. F. (2009). HOXA9 Modulates Its Oncogenic Partner Meis1 To Influence Normal Hematopoiesis. Molecular and Cellular Biology, 29(18), 5181–5192. doi:10.1128/MCB.00545-09  Huang, H., Rastegar, M., Bodner, C., Goh, S. L., Rambaldi, I., & Featherstone, M. (2005). MEIS C Termini Harbor Transcriptional Activation Domains That Respond to Cell Signaling. Journal of Biological Chemistry, 280(11), 10119–10127. doi:10.1074/jbc.M413963200  Huang, Y., Sitwala, K., Bronstein, J., Sanders, D., Dandekar, M., Collins, C., et al. (2012). Identification and characterization of Hoxa9 binding sites in hematopoietic cells. Blood, 119(2), 388–398. doi:10.1182/blood-2011-03-341081  Huntly, B. J. P., & Gilliland, D. G. (2005). Leukaemia stem cells and the evolution of cancer-stem-cell research. Nature Reviews Cancer, 5, 311–321.  Huntly, B. J. P., Shigematsu, H., Deguchi, K., Lee, B. H., Mizuno, S., Duclos, N., et al. (2004). MOZ-TIF2, but not BCR-ABL, confers properties of leukemic stem cells to committed murine hematopoietic progenitors. Cancer Cell, 6(6), 587–596. doi:10.1016/j.ccr.2004.10.015  Ichikawa, M., Asai, T., Saito, T., Seo, S., Yamazaki, I., Yamagata, T., et al. (2004). AML-1 is required for megakaryocytic maturation and lymphocytic differentiation, but not for maintenance of hematopoietic stem cells in adult hematopoiesis. Nature Medicine, 10(3), 299–304. doi:10.1038/nm997  Imamura, T., Morimoto, A., Ikushima, S., Kakazu, N., Hada, S., Tabata, Y., et al. (2002a). A novel infant acute lymphoblastic leukemia cell line with MLL-AF5q31 fusion transcript. Leukemia, 16(11), 2302–2308. doi:10.1038/sj.leu.2402665  Imamura, T., Morimoto, A., Takanashi, M., Hibi, S., Sugimoto, T., Ishii, E., & Imashuku, S. (2002b). Frequent co-expression of HoxA9 and Meis1 genes in infant acute lymphoblastic leukaemia with MLL rearrangement. British Journal of Haematology, 119(1), 119–121.  Ito, K., Hirao, A., Arai, F., Takubo, K., Matsuoka, S., Miyamoto, K., et al. (2006). Reactive oxygen species act through p38 MAPK to limit the lifespan of hematopoietic stem cells.  167 Nature Medicine, 12(4), 446–451. doi:10.1038/nm1388  Iwasaki, M. (2005). Identification of cooperative genes for NUP98-HOXA9 in myeloid leukemogenesis using a mouse model. Blood, 105(2), 784–793. doi:10.1182/blood-2004-04-1508  Jaw, T. J., You, L. R., Knoepfler, P. S., Yao, L. C., Pai, C. Y., Tang, C. Y., et al. (2000). Direct interaction of two homeoproteins, homothorax and extradenticle, is essential for EXD nuclear localization and function. Mechanisms of Development, 91(1-2), 279–291.  Ji, H., Ehrlich, L. I. R., Seita, J., Murakami, P., Doi, A., Lindau, P., et al. (2010). Comprehensive methylome map of lineage commitment from haematopoietic progenitors. Nature, 467(7313), 338–342. doi:10.1038/nature09367  Jo, S. Y., Granowicz, E. M., Maillard, I., Thomas, D., & Hess, J. L. (2011). Requirement for Dot1l in murine postnatal hematopoiesis and leukemogenesis by MLL translocation. Blood, 117(18), 4759–4768. doi:10.1182/blood-2010-12-327668  Jojic, V., Shay, T., Sylvia, K., Zuk, O., Sun, X., Kang, J., et al. (2013). Identification of transcriptional regulators in the mouse immune system. Nature Immunology, 14(6), 633–643. doi:10.1038/ni.2587  Jones, T. A., Flomen, R. H., Senger, G., Nizetić, D., & Sheer, D. (2000). The homeobox gene MEIS1 is amplified in IMR-32 and highly expressed in other neuroblastoma cell lines. European Journal of Cancer (Oxford, England : 1990), 36(18), 2368–2374.  Jordan, C. T. (2004). Cancer stem cell biology: from leukemia to solid tumors. Current Opinion in Cell Biology, 16(6), 708–712. doi:10.1016/j.ceb.2004.09.002  Jordan, C. T., & Guzman, M. L. (2004). Mechanisms controlling pathogenesis and survival of leukemic stem cells. Oncogene, 23(43), 7178–7187. doi:10.1038/sj.onc.1207935  Kalberer, C. P., Antonchuck, J., & Humphries, R. K. (2002). Genetic Modification of Murine Hematopoietic Stem Cells by Retroviruses. In C. A. Klug & C. T. Jordan (Eds.), Hematopoietic Stem Cell Protocols (1st ed., pp. 231–242). Totowa, New Jersey: Humana Press.  Kamps, M. P., Murre, C., Sun, X. H., & Baltimore, D. (1990). A new homeobox gene contributes the DNA binding domain of the t(1;19) translocation protein in pre-B ALL. Cell, 60(4), 547–555.  Kawagoe, H., Humphries, R. K., Blair, A., SUTHERLAND, H. J., & Hogge, D. (1999). Expression of HOX genes, HOX cofactors, and MLL in phenotypically and functionally defines subpopulations of leukemic and normal human hematopoietic cells. Leukemia, 13, 687–698.   168 Kent, D. G., Copley, M. R., Benz, C., Wohrer, S., Dykstra, B. J., Ma, E., et al. (2009). Prospective isolation and molecular characterization of hematopoietic stem cells with durable self-renewal potential. Blood, 113(25), 6342–6350. doi:10.1182/blood-2008-12-192054  Kiel, M. J., Yilmaz, O. H., Iwashita, T., Yilmaz, O. H., Terhorst, C., & Morrison, S. J. (2005). SLAM family receptors distinguish hematopoietic stem and progenitor cells and reveal endothelial niches for stem cells. Cell, 121(7), 1109–1121. doi:10.1016/j.cell.2005.05.026  Kocabas, F., Zheng, J., Thet, S., Copeland, N. G., Jenkins, N. A., DeBerardinis, R. J., et al. (2012). Meis1 regulates the metabolic phenotype and oxidant defense of hematopoietic stem cells. Blood. doi:10.1182/blood-2012-05-432260  Kohlmann, A., Schoch, C., Dugas, M., Schnittger, S., Hiddemann, W., Kern, W., & Haferlach, T. (2005). New insights into MLL gene rearranged acute leukemias using gene expression profiling: shared pathways, lineage commitment, and partner genes. Leukemia, 19(6), 953–964. doi:10.1038/sj.leu.2403746  Kondo, M., Weissman, I. L., & Akashi, K. (1997). Identification of clonogenic common lymphoid progenitors in mouse bone marrow. Cell, 91(5), 661–672.  Kroon, E., Krosl, J., Thorsteinsdottir, U., Baban, S., Buchberg, A. M., & Sauvageau, G. (1998). Hoxa9 transforms primary bone marrow cells through specific collaboration with Meis1a but not Pbx1b. The EMBO Journal, 17(13), 3714–3725.  Krosl, J., Austin, P., Beslu, N., Kroon, E., Humphries, R. K., & Sauvageau, G. (2003a). In vitro expansion of hematopoietic stem cells by recombinant TAT-HOXB4 protein. Nature Medicine, 9(11), 1428–1432. doi:10.1038/nm951  Krosl, J., Beslu, N., Mayotte, N., Humphries, R. K., & Sauvageau, G. (2003b). The Competitive Nature of HOXB4-Transduced HSC Is Limited by PBX1: The Generation of Ultra-Competitive Stem Cells Retaining Full Differentiation Potential. Immunity, 18, 561–571.  Krystal, G., Lam, V., Dragowska, W., Takahashi, C., Appel, J., Gontier, A., et al. (1994). Transforming growth factor beta 1 is an inducer of erythroid differentiation. The Journal of Experimental Medicine, 180(3), 851–860.  Kuhn, R., Schwenk, F., Aguet, M., & Rajewsky, K. (1995). Inducible gene targeting in mice. Science, 269(5229), 1427–1429.  Kumar, A. R., Li, Q., Hudson, W. A., Chen, W., Sam, T., Yao, Q., et al. (2009). A role for MEIS1 in MLL-fusion gene leukemia. Blood, 113(8), 1756–1758. doi:10.1182/blood-2008-06-163287   169 Kumar, A. R., Sarver, A. L., Wu, B., & Kersey, J. H. (2010). Meis1 maintains stemness signature in MLL-AF9 leukemia. Blood, 114, 3642–3643.  Lawrence, H. J., Christensen, J., Fong, S., Hu, Y.-L., Weissman, I., Sauvageau, G., et al. (2005). Loss of expression of the Hoxa-9 homeobox gene impairs the proliferation and repopulating ability of hematopoietic stem cells. Blood, 106(12), 3988–3994. doi:10.1182/blood-2005-05-2003  Lawrence, H. J., Helgason, C. D., Sauvageau, G., Fong, S., Izon, D. J., Humphries, R. K., & Largman, C. (1997). Mice bearing a targeted interruption of the homeobox gene HOXA9 have defects in myeloid, erythroid, and lymphoid hematopoiesis. Blood, 89(6), 1922–1930.  Lawrence, H. J., Rozenfeld, S., Cruz, C., Matsukuma, K., KWONG, A., KOMUVES, L. G., et al. (1999). Frequent co-expression of the HOXA9 and MEIS1 homeobox genes in human myeloid leukemias. Leukemia, 13, 1993–1999.  Lee, F. S., & Percy, M. J. (2011). The HIF Pathway and Erythrocytosis. Annual Review of Pathology: Mechanisms of Disease, 6(1), 165–192. doi:10.1146/annurev-pathol-011110-130321  Li, J., Tang, B., Qu, Y., & Mu, D. (2011). International Journal of Developmental Neuroscience. International Journal of Developmental Neuroscience, 29(8), 867–872. doi:10.1016/j.ijdevneu.2011.07.010  Lieu, Y. K., & Reddy, E. P. (2009). Conditional c-myb knockout in adult hematopoietic stem cells leads to loss of self-renewal due to impaired proliferation and accelerated differentiation. Pnas, 106, 21689–21694.  Lindsey, S., & Papoutsakis, E. T. (2012). The evolving role of the aryl hydrocarbon receptor (AHR) in the normophysiology of hematopoiesis. Stem Cell Reviews and Reports, 8(4), 1223–1235. doi:10.1007/s12015-012-9384-5  Longobardi, E., Iotti, G., Di Rosa, P., Mejetta, S., Bianchi, F., Fernandez-Diaz, L. C., et al. (2010). Prep1 (pKnox1)-deficiency leads to spontaneous tumor development in mice and accelerates EmuMyc lymphomagenesis: a tumor suppressor role for Prep1. Molecular Oncology, 4(2), 126–134. doi:10.1016/j.molonc.2010.01.001  Loonstra, A., Vooijs, M., Beverloo, H. B., Allak, Al, B., van Drunen, E., Kanaar, R., et al. (2001). Growth inhibition and DNA damage induced by Cre recombinase in mammalian cells . Pnas, 98, 9209–9214.  Lu, Q., Knoepfler, P. S., Scheele, J., Wright, D. D., & Kamps, M. P. (1995). Both Pbx1 and E2A-Pbx1 bind the DNA motif ATCAATCAA cooperatively with the products of multiple murine Hox genes, some of which are themselves oncogenes. Molecular and  170 Cellular Biology, 15, 3786–3795.  Maeda, R., Ishimura, A., Mood, K., Park, E. K., Buchberg, A. M., & Daar, I. O. (2002). Xpbx1b and Xmeis1b play a collaborative role inhindbrain and neural crest gene expressionin Xenopus embryos. Pnas, 99(8), 5448–5453.  Maguer-Satta, V., Petzer, A. L., Eaves, A. C., & Eaves, C. J. (1996). BCR-ABL expression in different subpopulations of functionally characterized Ph+ CD34+ cells from patients with chronic myeloid leukemia. Blood, 88(5), 1796–1804.  Mamo, A., Krosl, J., Kroon, E., Bijl, J., Thompson, A., Mayotte, N., et al. (2006). Molecular dissection of Meis1 reveals 2 domains required for leukemia induction and a key role for Hoxa gene activation. Blood, 108(2), 622–629. doi:10.1182/blood-2005-06-2244  Mann, R. S., & Abu-Shaar, M. (1996). Nuclear import of the homeodomain protein extradenticle in response to Wg and Dpp signalling. Nature, 383(6601), 630–633. doi:10.1038/383630a0  Matsumoto, A., & Nakayama, K. I. (2012). Role of key regulators of the cell cycle in maintenance of hematopoietic stem cells. Biochimica Et Biophysica Acta. doi:10.1016/j.bbagen.2012.07.004  McMahon, K. A., Hiew, S. Y. L., Hadjur, S., Veiga-Fernandes, H., Menzel, U., Price, A. J., et al. (2007). Mll Has a Critical Role in Fetal and Adult Hematopoietic Stem Cell Self-Renewal. Cell Stem Cell, 1(3), 338–345. doi:10.1016/j.stem.2007.07.002  Mikkola, H. K. A., Klintman, J., Yang, H., Hock, H., Schlaeger, T. M., Fujiwara, Y., & Orkin, S. H. (2003). Haematopoietic stem cells retain long-term repopulating activity and multipotency in the absence of stem-cell leukaemia SCL/tal-1 gene. Nature, 421, 547–551.  Misaghian, N., Ligresti, G., Steelman, L. S., Bertrand, F. E., Bäsecke, J., Libra, M., et al. (2008). Targeting the leukemic stem cell: the Holy Grail of leukemia therapy. Leukemia, 23(1), 25–42. doi:10.1038/leu.2008.246  Miyamoto, K., Araki, K. Y., Naka, K., Arai, F., Takubo, K., Yamazaki, S., et al. (2007). Foxo3a Is Essential for Maintenance of the Hematopoietic Stem Cell Pool. Cell Stem Cell, 1(1), 101–112. doi:10.1016/j.stem.2007.02.001  Moens, C., & Selleri, L. (2006). Hox cofactors in vertebrate development. Developmental Biology, 291(2), 193–206. doi:10.1016/j.ydbio.2005.10.032  Mojsin, M., & Stevanovic, M. (2009). PBX1 and MEIS1 up-regulate SOX3gene expression by direct interaction with a consensus binding site within the basal promoter region. Biochemical Journal, 425(1), 107–116. doi:10.1042/BJ20090694   171 Morgado, E., Albouhair, S., & Lavau, C. (2007). Flt3 is dispensable to the Hoxa9/Meis1 leukemogenic cooperation. Blood, 109(9), 4020–4022. doi:10.1182/blood-2006-01-039586  Moskow, J. J. (1995). Meis1, a PBX1-Related Homeobox Gene Involved in Myeloid Leukemia in BXH-2 Mice. Mol Cell Biol, 1–10.  Mulloy, J. C., Cammenga, J., MacKenzie, K. L., Berguido, F. J., Moore, M., & Nimer, S. D. (2002). The AML1-ETO fusion protein promotes the expansion of human hematopoietic stem cells. Blood, 99(1), 15–23. doi:10.1182/blood.V99.1.15  Murone, M., Carpenter, D. A., & de Sauvage, F. J. (1998, January 1). Hematopoietic Deficiencies in c-mpl and TPO Knockout Mice. Stem Cells. Retrieved July 17, 2012, from http://onlinelibrary.wiley.com/doi/10.1002/stem.160001/abstract;jsessionid=234D6F3A83927E5F1D45CC2EEF52D498.d03t02  Müller, A. M., Medvinsky, A., Strouboulis, J., Grosveld, F., & Dzierzak, E. (1994). Development of hematopoietic stem cell activity in the mouse embryo. Immunity, 1(4), 291–301.  Nakamura, T., Largaespada, D. A., Shaughnessy, J. D., Jenkins, N. A., & Copeland, N. G. (1996). Cooperative activation of Hoxa and Pbx1-related genes in murine myeloid leukaemias. Nature Genetics, 12(2), 149–153. doi:10.1038/ng0296-149  Nishi, H., Nakada, T., Kyo, S., Inoue, M., Shay, J. W., & Isaka, K. (2004). Hypoxia-Inducible Factor 1 Mediates Upregulation of Telomerase (hTERT). Mol Cell Biol, 24(13), 6076–6083. doi:10.1128/MCB.24.13.6076-6083.2004  Nürnberg, S. T., Rendon, A., Smethurst, P. A., Paul, D. S., Voss, K., Thon, J. N., et al. (2012). A GWAS sequence variant for platelet volume marks an alternative DNM3 promoter in megakaryocytes near a MEIS1 binding site. Blood, 120(24), 4859–4868. doi:10.1182/blood-2012-01-401893  Ohta, H., Sekulovic, S., Bakovic, S., Eaves, C. J., Pineault, N., Gasparetto, M., et al. (2007). Near-maximal expansions of hematopoietic stem cells in culture using NUP98-HOX fusions. Experimental Hematology, 35(5), 817–830. doi:10.1016/j.exphem.2007.02.012  Okada, Y., Nagai, R., Sato, T., Matsuura, E., Minami, T., Morita, I., & Doi, T. (2003). Homeodomain proteins MEIS1 and PBXs regulate the lineage-specific transcription of the platelet factor 4 gene. Blood, 101(12), 4748–4756. doi:10.1182/blood-2002-02-0380  Okuda, T., van Deursen, J., Hiebert, S. W., Grosveld, G., & Downing, J. R. (1996). AML1, the target of multiple chromosomal translocations in human leukemia, is essential for normal fetal liver hematopoiesis. Cell, 84(2), 321–330.   172 Orford, K. W., & Scadden, D. T. (2008). Deconstructing stem cell self-renewal: genetic insights into cell-cycle regulation. Nature Reviews Genetics, 9(2), 115–128. doi:10.1038/nrg2269  Orkin, S. H., & Zon, L. I. (2008). Hematopoiesis: An Evolving Paradigm for Stem Cell Biology. Cell, 132(4), 631–644. doi:10.1016/j.cell.2008.01.025  Orlovsky, K., Kalinkovich, A., Rozovskaia, T., Shezen, E., Itkin, T., Alder, H., et al. (2011). Down-regulation of homeobox genes MEIS1 and HOXA in MLL-rearranged acute leukemia impairs engraftment and reduces proliferation. Proceedings of the National Academy of Sciences, 108(19), 7956–7961. doi:10.1073/pnas.1103154108  Pai, C.-Y., Kuo, T.-S., Jaw, T. J., Kurant, E., Chen, C.-T., Bessarab, D. A., et al. (1998). The Homothorax homeoprotein activates the nuclear localization of another homeoprotein, Extradenticle, and suppresses eye development in Drosophila. Genes & Development, 435–446.  Palmqvist, L., Argiropoulos, B., Pineault, N., Abramovich, C., Sly, L. M., Krystal, G., et al. (2006). The Flt3 receptor tyrosine kinase collaborates with NUP98-HOX fusions in acute myeloid leukemia. Blood, 108(3), 1030–1036. doi:10.1182/blood-2005-12-007005  Passegue, E., Jamieson, C. H. M., Ailles, L. E., & Weissman, I. L. (2003). Normal and leukemic hematopoiesis: Are leukemiasa stem cell disorder or a reacquisition of stemcell characteristics? Pnas, 100, 11842–11849.  Pawliuk, R., Kay, R., Lansdorp, P., & Humphries, R. K. (1994). Selection of retrovirally transduced hematopoietic cells using CD24 as a marker of gene transfer. Blood, 84(9), 2868–2877.  Peifer, M., & Wieschaus, E. (1990). Mutations in the Drosophila gene extradenticle affect the way specific homeo domain proteins regulate segmental identity. Genes & Development, 4(7), 1209–1223.  Perry, J. M., Harandi, O. F., & Paulson, R. F. (2007). BMP4, SCF, and hypoxia cooperatively regulate the expansion of murine stress erythroid progenitors. Blood, 109(10), 4494–4502. doi:10.1182/blood-2006-04-016154  Phelan, M. L., Rambaldi, I., & Featherstone, M. S. (1995). Cooperative interactions between HOX andPBX proteins mediated by a conserved peptide motif. Molecular and Cellular Biology, 15(8), 3989–3997.  Pineault, N., Abramovich, C., & Humphries, R. K. (2005). Transplantable cell lines generated with NUP98–Hox fusion genes undergo leukemic progression by Meis1 independent of its binding to DNA. Leukemia. doi:10.1038/sj.leu.2403696   173 Pineault, N., Abramovich, C., Ohta, H., & Humphries, R. K. (2004). Differential and Common Leukemogenic Potentials of Multiple NUP98-Hox Fusion Proteins Alone or with Meis1. Molecular and Cellular Biology, 24, 1907–1917. doi:10.1128/MCB.24.5.1907–1917.2004  Pineault, N., Buske, C., Feuring-Buske, M., Abramovich, C., Rosten, P., Hogge, D. E., et al. (2003). Induction of acute myeloid leukemia in mice by the human leukemia-specific fusion gene NUP98-HOXD13 in concert with Meis1. Blood, 101(11), 4529–4538. doi:10.1182/blood-2002-08-2484  Pineault, N., Helgason, C. D., Lawrence, H. J., & Humphries, R. K. (2002). Differential expression of Hox, Meis1 and Pbx1 genes in primitive cells throughout murine hematopoietic ontogeny. Experimental Hematology, 30, 49–57.  Piper, D. E., Batchelor, A. H., Chang, C.-P., Cleary, M. L., & Wolberger, C. (1999). Structure of a HoxB1–Pbx1 Heterodimer Bound to DNA: Role of the Hexapeptide and a Fourth Homeodomain Helix in Complex Formation. Cell, 96, 587–597.  Pronk, C. J. H., Rossi, D. J., Månsson, R., Attema, J. L., Norddahl, G. L., Chan, C. K. F., et al. (2007). Elucidation of the phenotypic, functional, and molecular topography of a myeloerythroid progenitor cell hierarchy. Cell Stem Cell, 1(4), 428–442. doi:10.1016/j.stem.2007.07.005  Purton, L. E., & Scadden, D. T. (2007). Limiting factors in murine hematopoietic stem cell assays. Cell Stem Cell, 1(3), 263–270. doi:10.1016/j.stem.2007.08.016  Rankin, E. B., Wu, C., Khatri, R., Wilson, T. L. S., Andersen, R., Araldi, E., et al. (2012). The HIF Signaling Pathway in Osteoblasts Directly Modulates Erythropoiesis through the Production of EPO. Cell, 149(1), 63–74. doi:10.1016/j.cell.2012.01.051  Rauskolb, C., & Wieschaus, E. (1994). Coordinate regulation of downstream genes by extradenticle and the homeotic selector proteins. The EMBO Journal, 13(15), 3561–3569.  Rauskolb, C., Peifer, M., & Wieschaus, E. (1993). extradenticle, a regulator of homeotic gene activity, is a homolog of the homeobox-containing human proto-oncogene pbx1. Cell, 74(6), 1101–1112.  Rave-Harel, N., Givens, M. L., Nelson, S. B., Duong, H. A., Coss, D., Clark, M. E., et al. (2004). TALE homeodomain proteins regulate gonadotropin-releasing hormone gene expression independently and via interactions with Oct-1. The Journal of Biological Chemistry, 279(29), 30287–30297. doi:10.1074/jbc.M402960200  Rebel, V. I., Miller, C. L., Eaves, C. J., & Lansdorp, P. M. (1996). The repopulation potential of fetal liver hematopoietic stem cells in mice exceeds that of their liver adult bone  174 marrow counterparts. Blood, 87, 3500–3507.  Reya, T. (2003). Regulation of Hematopoietic Stem Cell Self-Renewal. Recent Prog Horm Res, 58, 283–295.  Rieckhof, G. E., Casares, F., Ryoo, H. D., Abu-Shaar, M., & Mann, R. S. (1997). Nuclear translocation of extradenticle requires homothorax, which encodes an extradenticle-related homeodomain protein. Cell, 91(2), 171–183.  Rizo, A. (2006). Signaling pathways in self-renewing hematopoietic and leukemic stem cells: do all stem cells need a niche? Human Molecular Genetics, 15(Review Issue 2), R210–R219. doi:10.1093/hmg/ddl175  Roche, J., Zeng, C., Barón, A., Gadgil, S., Gemmill, R. M., Tigaud, I., et al. (2004). Hox expression in AML identifies a distinct subset of patients with intermediate cytogenetics. Leukemia, 18(6), 1059–1063. doi:10.1038/sj.leu.2403366  Rosales-Aviña, J. A., Torres-Flores, J., Aguilar-Lemarroy, A., Gurrola-Díaz, C., Hernández-Flores, G., Ortiz-Lazareno, P. C., et al. (2011). MEIS1, PREP1, and PBX4 are differentially expressed in acute lymphoblastic leukemia: association of MEIS1 expression with higher proliferation and chemotherapy resistance. Journal of Experimental & Clinical Cancer Research : CR, 30, 112. doi:10.1186/1756-9966-30-112  Roy, S., Tripathy, M., Mathur, N., Jain, A., & Mukhopadhyay, A. (2012). Hypoxia improves expansion potential of human cord blood-derived hematopoietic stem cells and marrow repopulation efficiency. European Journal of Haematology, 88(5), 396–405. doi:10.1111/j.1600-0609.2012.01759.x  Royo, J. L., Bessa, J., Hidalgo, C., Fernández-Miñán, A., Tena, J. J., Roncero, Y., et al. (2012). Identification and analysis of conserved cis-regulatory regions of the MEIS1 gene. PLoS ONE, 7(3), e33617. doi:10.1371/journal.pone.0033617  Rozovskaia, T., Feinstein, F., Mor, O., Foà, R., Blechman, J., Nakamura, T., et al. (2001). Upregulation of Meis1 and Hoxa9 in Acute Lymphocytic Leukemias with the T(4:11) Abnormality. Oncogene, 20, 874–878.  Ryoo, H. D., Marty, T., Casares, F., Affolter, M., & Mann, R. S. (1999). Regulation of Hox target genes by a DNA bound Homothorax/Hox/Extradenticle complex. Development, 126(22), 5137–5148.  Saleh, M., Huang, H., Green, N. C., & Featherstone, M. S. (2000). A conformational change in PBX1A is necessary for its nuclear localization. Experimental Cell Research, 260(1), 105–115. doi:10.1006/excr.2000.5010  Sardina, J. L., López-Ruano, G., Sánchez-Sánchez, B., Llanillo, M., & Hernández-Hernández, A. (2012). Reactive oxygen species: are they important for haematopoiesis?  175 Critical Reviews in Oncology/Hematology, 81(3), 257–274. doi:10.1016/j.critrevonc.2011.03.005  Sashida, G., & Iwama, A. (2012). Epigenetic regulation of hematopoiesis. International Journal of Hematology, 96(4), 405–412. doi:10.1007/s12185-012-1183-x  Sauvageau, G., Lansdorp, P. M., Eaves, C. J., Hogge, D. E., Dragowska, W., Reid, D. S., et al. (1994). Differential expression of homeobox genes in functionally distinct CD34+ subpopulations of human bone marrow cells. Proceedings of the National Academy of Sciences of the United States of America, 91(25), 12223–12227.  Sauvageau, G., Thorsteinsdottir, U., Eaves, C. J., Lawrence, H. J., Largman, C., Lansdorp, P. M., & Humphries, R. K. (1995). Overexpression of HOXB4 in hematopoietic cells causes the selective expansion of more primitive populations in vitro and in vivo. Genes & Development, 9(14), 1753–1765. doi:10.1101/gad.9.14.1753  Shanmugam, K., Green, N. C., Rambaldi, I., Saragovi, H. U., & Featherstone, M. (1999). PBX and MEIS as Non-DNA-Binding Partners in Trimeric Complexes with HOX Proteins. Molecular and Cellular Biology, 19, 7577–7588.  Shen, W., Montgomery, J., Rozenfeld, S., Moskow, J., Lawrence, H. J., Buchberg, A., & Largman, C. (1997). AbdB-Like Hox Proteins Stabilize DNA Binding by the Meis1 Homeodomain Proteins. Molecular and Cellular Biology, 17(11), 6448–6458.  Shen, W.F., Rozenfelf, S., Kwong, A., Komulves, L. G., Lawrence, H. J., & Largman, C. (1999). HOXA9 Forms Triple Complexes with PBX2 and MEIS1 in Myeloid Cells. Molecular and Cellular Biology, 19(4), 3051–3061.  Shinnick, K. M., Eklund, E. A., & McGarry, T. J. (2010). Geminin deletion from hematopoietic cells causes anemia and thrombocytosis in mice. Journal of Clinical Investigation, 120(12), 4303–4315. doi:10.1172/JCI43556DS1  Shivdasani, R. A., & Schulze, H. (2005). Culture, expansion, and differentiation of murine megakaryocytes. Current Protocols in Immunology / Edited by John E. Coligan ... [Et Al.], Chapter 22, Unit 22F.6. doi:10.1002/0471142735.im22f06s67  Simsek, T., Kocabas, F., Zheng, J., DeBerardinis, R. J., Mahmoud, A. I., Olson, E. N., et al. (2010). The Distinct Metabolic Profile of Hematopoietic Stem Cells Reflects Their Location in a Hypoxic Niche. Stem Cell, 7(3), 380–390. doi:10.1016/j.stem.2010.07.011  Socolovsky, M., Nam, H., Fleming, M. D., Haase, V. H., Brugnara, C., & Lodish, H. F. (2001). Ineffective erythropoiesis in Stat5a(-/-)5b(-/-) mice due to decreased survival of early erythroblasts. Blood, 98(12), 3261–3273.  Somervaille, T. C. P., & Cleary, M. L. (2010). Grist for the MLL: how do MLL oncogenic fusion proteins generate leukemia stem cells? International Journal of Hematology,  176 91(5), 735–741. doi:10.1007/s12185-010-0579-8  Song, X., Krelin, Y., Dvorkin, T., Bjorkdahl, O., Segal, S., Dinarello, C. A., et al. (2005). CD11b+/Gr-1+ immature myeloid cells mediate suppression of T cells in mice bearing tumors of IL-1beta-secreting cells. Journal of Immunology (Baltimore, Md. : 1950), 175(12), 8200–8208.  Spangrude, G. J., Smith, L., Uchida, N., Ikuta, K., Heimfeld, S., Friedman, J., & Weissman, I. L. (1991). Mouse hematopoietic stem cells. Blood, 78(6), 1395–1402.  Spieker, N., van Sluis, P., Beitsma, M., Boon, K., van Schaik, B. D., van Kampen, A. H., et al. (2001). The MEIS1 oncogene is highly expressed in neuroblastoma and amplified in cell line IMR32. Genomics, 71(2), 214–221. doi:10.1006/geno.2000.6408  Sprüssel, A., Schulte, J. H., Weber, S., Necke, M., Händschke, K., Thor, T., et al. (2012). Lysine-specific demethylase 1 restricts hematopoietic progenitor proliferation and is essential for terminal differentiation. Leukemia. doi:10.1038/leu.2012.157  Stankunas, K., Shang, C., Twu, K. Y., Kao, S. C., Jenkins, N. A., Copeland, N. G., et al. (2008). Pbx/Meis Deficiencies Demonstrate Multigenetic Origins of Congenital Heart Disease. Circulation Research, 103(7), 702–709. doi:10.1161/CIRCRESAHA.108.175489  Szilvassy, S. J., Humphries, R. K., Lansdorp, P. M., Eaves, A. C., & Eaves, C. J. (1990). Quantitative Assay for Totipotent Reconstituting Hematopoietic Stem Cells by a Competitive Repopulation Strategy. Pnas, 87(22), 8736–8740.  Takeda, A., Goolsby, C., & Yaseen, N. R. (2006). NUP98-HOXA9 induces long-term proliferation and blocks differentiation of primary human CD34+ hematopoietic cells. Cancer Research, 66(13), 6628–6637. doi:10.1158/0008-5472.CAN-06-0458  Takubo, K., Goda, N., Yamada, W., Iriuchishima, H., Ikeda, E., Kubota, Y., et al. (2010). Regulation of the HIF-1&alpha; Level Is Essential for Hematopoietic Stem Cells. Stem Cell, 7(3), 391–402. doi:10.1016/j.stem.2010.06.020  Tamai, H., & Inokuchi, K. (2010). 11q23/MLL acute leukemia : update of clinical aspects. Journal of Clinical and Experimental Hematopathology : JCEH, 50(2), 91–98.  Thorsteinsdottir, U., Mamo, A., Kroon, E., Jerome, L., Bijl, J., Lawrence, H. J., et al. (2002). Overexpression of the myeloid leukemia-associated Hoxa9 gene in bone marrow cells induces stem cell expansion. Blood, 99(1), 121–129. doi:10.1182/blood.V99.1.121  Tothova, Z., Kollipara, R., Huntly, B. J., Lee, B. H., Castrillon, D. H., Cullen, D. E., et al. (2007). FoxOs Are Critical Mediators of Hematopoietic Stem Cell Resistance to Physiologic Oxidative Stress. Cell, 128(2), 325–339. doi:10.1016/j.cell.2007.01.003   177 Trentin, L., Giordan, M., Dingermann, T., Basso, G., Kronnie, te, G., & Marschalek, R. (2009). Two independent gene signatures in pediatric t(4;11) acute lymphoblastic leukemia patients. European Journal of Haematology, 83(5), 406–419. doi:10.1111/j.1600-0609.2009.01305.x  Unnisa, Z., Clark, J. P., Roychoudhury, J., Thomas, E., Tessarollo, L., Copeland, N. G., et al. (2012). Meis1 preserves hematopoietic stem cells in mice by limiting oxidative stress. Blood. doi:10.1182/blood-2012-06-435800  Van Zant, G. (1984). Studies of hematopoietic stem cells spared by 5-fluorouracil. The Journal of Experimental Medicine, 159(3), 679–690.  Vu, H. A., Xinh, P. T., Masuda, M., Motoji, T., Toyoda, A., Sakaki, Y., et al. (2006). FLT3 is fused to ETV6 in a myeloproliferative disorder with hypereosinophilia and a t(12;13)(p13;q12) translocation. Leukemia, 20(8), 1414–1421. doi:10.1038/sj.leu.2404266  Wang, G. G., Cai, L., Pasillas, M. P., & Kamps, M. P. (2007). NUP98–NSD1 links H3K36 methylation to Hox-A gene activation and leukaemogenesis. Nature Cell Biology, 9(7), 804–812. doi:10.1038/ncb1608  Wang, G. G., Pasillas, M. P., & Kamps, M. P. (2005). Meis1 programs transcription of FLT3 and cancer stem cell character, using a mechanism that requires interaction with Pbx and a novel function of the Meis1 C-terminus. Blood, 106(1), 254–264. doi:10.1182/blood-2004-12-4664  Wang, G. G., Pasillas, M. P., & Kamps, M. P. (2006). Persistent Transactivation by Meis1 Replaces Hox Function in Myeloid Leukemogenesis Models: Evidence for Co-Occupancy of Meis1-Pbx and Hox-Pbx Complexes on Promoters of Leukemia-Associated Genes. Molecular and Cellular Biology, 26, 3902–3916. doi:10.1128/MCB.26.10.3902–3916.2006  Warner, J. K., Wang, J. C. Y., Hope, K. J., Jin, L., & Dick, J. E. (2004). Concepts of human leukemic development. Oncogene, 23(43), 7164–7177. doi:10.1038/sj.onc.1207933  Waskiewicz, A. J., rikhof, H. A., Hernandez, R. E., & Moens, C. B. (2001). Zebrafish Meis functions to stabilize PBX proteins and regulate hindbrain patterning. Development, 128, 4139–4151.  Williams, T. M., Williams, M. E., & Innis, J. W. (2005). Range of HOX/TALE superclass associations and protein domain requirements for HOXA13:MEIS interaction. Developmental Biology, 277(2), 457–471. doi:10.1016/j.ydbio.2004.10.004  Wilson, N. K., Foster, S. D., Wang, X., Knezevic, K., Schütte, J., Kaimakis, P., et al. (2010). Combinatorial Transcriptional Control In Blood Stem/Progenitor Cells: Genome-wide Analysis of Ten Major Transcriptional Regulators. Stem Cell, 7(4), 532–544.  178 doi:10.1016/j.stem.2010.07.016  Wong, P., Iwasaki, M., Somervaille, T. C. P., So, C. W. E., & Cleary, M. L. (2007). Meis1 is an essential and rate-limiting regulator of MLL leukemia stem cell potential. Genes & Development, 21(21), 2762–2774. doi:10.1101/gad.1602107  Woolthuis, C. M., Han, L., Verkaik-Schakel, R. N., van Gosliga, D., Kluin, P. M., Vellenga, E., et al. (2012). Downregulation of MEIS1 impairs long-term expansion of CD34+ NPM1-mutated acute myeloid leukemia cells. Leukemia, 26(4), 848–853. doi:10.1038/leu.2011.277  Xiang, P., Lo, C., Argiropoulos, B., Lai, C. B., Rouhi, A., Imren, S., et al. (2010). Identification of E74-like factor 1 (ELF1) as a transcriptional regulator of the Hox cofactor MEIS1. Experimental Hematology, 38(9), 798–808.e2. doi:10.1016/j.exphem.2010.06.006  Xu, B., Geerts, D., Qian, K., Zhang, H., & Zhu, G. (2008a). Myeloid ecotropic viral integration site 1 (MEIS) 1 involvement in embryonic implantation. Human Reproduction (Oxford, England), 23(6), 1394–1406. doi:10.1093/humrep/den082  Xu, J., Suzuki, M., Niwa, Y., Hiraga, J., Nagasaka, T., Ito, M., et al. (2008b). Clinical significance of nuclear non-phosphorylated beta-catenin in acute myeloid leukaemia and myelodysplastic syndrome. British Journal of Haematology, 140(4), 394–401. doi:10.1111/j.1365-2141.2007.06914.x  Yang, J., Su, Y., & Richmond, A. (2007). Antioxidants tiron and N-acetyl-L-cysteine differentially mediate apoptosis in melanoma cells via a reactive oxygen species-independent NF-kappaB pathway. Free Radical Biology & Medicine, 42(9), 1369–1380. doi:10.1016/j.freeradbiomed.2007.01.036  Yilmaz, O. H., Valdez, R., Theisen, B. K., Guo, W., Ferguson, D. O., Wu, H., & Morrison, S. J. (2006). Pten dependence distinguishes haematopoietic stem cells from leukaemia-initiating cells. Nature Publishing Group, 441(7092), 475–482. doi:10.1038/nature04703  Zangenberg, M., Grubach, L., Aggerholm, A., Silkjaer, T., Juhl-Christensen, C., Nyvold, C. G., et al. (2009). The combined expression of HOXA4 and MEIS1 is an independent prognostic factor in patients with AML. European Journal of Haematology, 83(5), 439–448. doi:10.1111/j.1600-0609.2009.01309.x  Zeisig, B. B., Milne, T., Garcia-Cuellar, M.-P., Schreiner, S., Martin, M.-E., Fuchs, U., et al. (2004). Hoxa9 and Meis1 Are Key Targets for MLL-ENL-Mediated Cellular Immortalization. Mol Cell Biol, 24, 617–628. doi:10.1128/MCB.24.2.617–628.2004  Zhang, F., Lau, S. S., & Monks, T. J. (2011). The cytoprotective effect of N-acetyl-L-cysteine against ROS-induced cytotoxicity is independent of its ability to enhance glutathione synthesis. Toxicological Sciences : an Official Journal of the Society of  179 Toxicology, 120(1), 87–97. doi:10.1093/toxsci/kfq364  Zhang, X., Friedman, A., Heaney, S., Purcell, P., & Maas, R. L. (2002). Meis homeoproteins directly regulate Pax6 during vertebrate lens morphogenesis. Genes & Development, 16(16), 2097–2107. doi:10.1101/gad.1007602  Zhou, J., Wu, J., Li, B., Liu, D., Yu, J., Yan, X., et al. (2013). PU.1 is essential for MLL leukemia partially via crosstalk with the MEIS/HOX pathway. Leukemia. doi:10.1038/leu.2013.384  Zon, L. I. (2008). Intrinsic and extrinsic control of haematopoietic stem-cell self-renewal. Nature, 453(7193), 306–313. doi:10.1038/nature07038  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items