Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Analysis of hematopoiesis from human pluripotent stem cells Kardel, Melanie Dawn 2011

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


24-ubc_2011_spring_kardel_melanie.pdf [ 4.21MB ]
JSON: 24-1.0071668.json
JSON-LD: 24-1.0071668-ld.json
RDF/XML (Pretty): 24-1.0071668-rdf.xml
RDF/JSON: 24-1.0071668-rdf.json
Turtle: 24-1.0071668-turtle.txt
N-Triples: 24-1.0071668-rdf-ntriples.txt
Original Record: 24-1.0071668-source.json
Full Text

Full Text

ANALYSIS OF HEMATOPOIESIS FROM HUMAN PLURIPOTENT STEM CELLS  by MELANIE DAWN KARDEL B.Sc. (Honours), The University of Alberta, 2001  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY  in THE FACULTY OF GRADUATE STUDIES (Genetics)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  April 2011 © Melanie Dawn Kardel, 2011  Abstract Human embryonic stem (ES) or induced pluripotent stem (iPS) cells have the potential ability to generate all of the cell types in the body. If their differentiation into relevant cell types of interest can be effectively controlled, they are attractive for developmental studies, disease modelling, drug testing, and advancing regenerative medicine. The generation of hematopoietic cells from human ES/iPS cells has been reported, but is highly variable and often inefficient. My specific objective in this thesis was to more fully characterize the process whereby hematopoietic cells are generated from primitive pluripotent precursors, to understand current limitations, and to design improvements that would increase the yield and reproducibility of hematopoietic cell generation. I first examined the effect of the conditions used to maintain the undifferentiated starting population of ES/iPS cells on their hematopoietic cell differentiation ability. The results showed that the initial maintenance conditions used do have a significant influence on the subsequent number and consistency of hematopoietic cells generated. In addition, I found that this process is separately influenced by optimization of sequentially manipulated (early and late) differentiation steps. Analysis of individual EBs revealed a previously unappreciated heterogeneity of hematopoietic output from single EBs in vitro. Even under the most optimal conditions studied, it was found that the majority of EBs did not generate any hematopoietic colony-forming cells (CFCs). This suggested that only a limited number of the initial ES/iPS cells were contributing to the hematopoietic progenitor cell output under these conditions. To investigate this latter phenomenon further, I developed a lentiviral system to track the subsequent hematopoietic progeny of marked undifferentiated or early differentiating ES/iPS cells. The use of this approach confirmed that few of the starting ES/iPS cells contribute to the hematopoietic output of individual EBs. Together these studies suggest that the genesis of hematopoietic progenitors from pluripotent precursors remains limited by multiple factors. Further studies to characterize cell types intermediate between fully pluripotent cells and those  ii  with hematopoietic activity are needed to define more rigorously and optimize the use of this strategy for various medical applications.  iii  Preface  The work presented in the Results Chapters of this thesis is being prepared for publication and none have been published previously. Some of the work presented in the Introduction sections has been included as contributions to manuscripts already published. Specifically, the initiation of studies to optimize the culture of undifferentiated human embryonic and induced pluripotent stem cells, and to study the role of MEDII in their differentiation was in collaboration with post-doctoral fellows in the laboratory, Dr. Michael O’Connor and Dr. Min Lu. As such, they had substantial intellectual input into my studies depicted in Figures 2.1-2.6. All experiments described in this thesis were designed, performed, and analysed by me with input from Dr. Connie Eaves. RNA purification and quantitative RT-PCR analyses in Chapter 2 were performed by Austin Hammond, a co-op undergraduate student who was working with me under my direct supervision. Viral preparations used in Chapter 4 were provided by Glenn Edin.  All studies in this thesis performed with human embryonic and induced pluripotent stem cells were approved by the University of British Columbia – British Columbia Cancer Agency Research Ethics Board, certificate number H03-60014.  iv  Table of Contents  Abstract .......................................................................................................................................ii Preface.......................................................................................................................................iv Table of Contents ....................................................................................................................... v List of Tables ............................................................................................................................ viii List of Figures.............................................................................................................................ix List of Abbreviations ................................................................................................................... x Acknowledgements ................................................................................................................... xii 1. Introduction ........................................................................................................................... 1 1.1 Stem cells ......................................................................................................................... 1 1.1.1 Human embryonic stem cells...................................................................................... 2 1.1.2 Induced pluripotent stem cells .................................................................................... 6 1.2 Hematopoietic cells in the adult ...................................................................................... 10 1.3 Generation of hematopoietic cells from pluripotent stem cells ......................................... 13 1.4 Hematopoiesis in the mammalian embryo ...................................................................... 15 1.4.1 Sites of hematopoiesis in the embryo ....................................................................... 15 1.4.2 Primitive hematopoiesis............................................................................................ 17 1.4.3 Definitive hematopoiesis........................................................................................... 18 1.4.4 Genes critical for hematopoietic development .......................................................... 20 1.5 Developmental origins of HSCs ...................................................................................... 21 1.5.1 Identification of hemogenic sites in the embryo ........................................................ 21 1.5.2 Generation of pre-hematopoietic mesoderm ............................................................. 25 1.5.3 Evidence for a hemangioblast state .......................................................................... 29 1.5.4 Evidence for hemogenic cells with mature endothelial features ................................ 30 1.6 Insights from human pluripotent cell models of hematopoietic development ................... 32 1.7 Thesis objectives ............................................................................................................ 36 2. Improved output of hematopoietic cells from human pluripotent stem cells by separate optimization of different developmental steps ........................................................................... 39 2.1 Introduction ..................................................................................................................... 39 2.2 Materials and methods.................................................................................................... 41 2.2.1 Human ES and iPS cell lines .................................................................................... 41 2.2.2 Maintenance culture of human ES and iPS cells ...................................................... 42 v  2.2.3 ES-CFC assays ........................................................................................................ 43 2.2.4 MEDII treatment ....................................................................................................... 43 2.2.5 OP9 co-cultures........................................................................................................ 44 2.2.6 EB cultures ............................................................................................................... 45 2.2.7 Hematopoietic CFC assays ...................................................................................... 46 2.2.8 Flow cytometry ......................................................................................................... 47 2.2.9 Quantitative RT-PCR analysis .................................................................................. 47 2.2.10 Statistical analyses ................................................................................................. 48 2.3 Results ........................................................................................................................... 50 2.3.1 Variation in hematopoietic CFC output using alternate methods of ES or iPS cell differentiation induction ..................................................................................................... 50 2.3.2 Highly undifferentiated human ES/iPS cells are poorly responsive to traditional hematopoietic differentiation induction protocols ............................................................... 53 2.3.3 Pretreatment of undifferentiated ES/iPS cells with MEDII conditioned medium increases the output of hematopoietic CFCs but does not alter the kinetics of their generation ......................................................................................................................... 59 2.3.4 MEDII conditioned medium primes cells for mesoderm formation upon transfer to EB conditions .......................................................................................................................... 62 2.3.5 Optimizing multiple steps of differentiation improves hematopoietic cell generation . 67 2.4 Discussion ...................................................................................................................... 70 3. Quantitative analysis of the generation and expansion of human ES and iPS cell-derived hematopoietic cells within individual EBs.................................................................................. 77 3.1 Introduction ..................................................................................................................... 77 3.2 Materials and methods.................................................................................................... 80 3.2.1 Human ES and iPS cell maintenance and differentiation cultures ............................. 80 3.2.2 Analysis of the hematopoietic CFC content of single EBs......................................... 80 3.3 Results ........................................................................................................................... 81 3.3.1 EBs show pronounced heterogeneity in their final content of hematopoietic CFCs ... 81 3.3.2 MEDII pretreatment and cytokine addition to EB cultures increases the output of hematopoietic CFCs per EB but does not alter the skewed distribution of CFCs amongst EBs ................................................................................................................................... 86 3.3.3 An increased proportion of EBs derived from MEDII-pretreated cells and cultured in the presence of cytokines generate CFU-GM.................................................................... 90 3.4 Discussion ...................................................................................................................... 96 4. Tracking human ES and iPS cell-derived hematopoietic cell development using a genetic marking strategy....................................................................................................................... 99 vi  4.1 Introduction ..................................................................................................................... 99 4.2 Materials and methods.................................................................................................. 101 4.2.1 Human pluripotent stem cell culture and differentiation........................................... 101 4.2.2 Lentiviral vectors .................................................................................................... 101 4.2.3 Production of lentivirus ........................................................................................... 103 4.2.4 Lentiviral transduction of human ES and iPS cells .................................................. 103 4.2.5 Analysis of marked EB-derived hematopoietic CFCs .............................................. 104 4.3 Results ......................................................................................................................... 104 4.3.1 Choice of lentiviral vector affects transduction efficiency and level of gene expression in human ES and iPS cells .............................................................................................. 104 4.3.2 Human ES and iPS cells can be marked by lentivirus at multiple points of their differentiation into hematopoietic CFCs ........................................................................... 107 4.3.3 Precursors of hematopoietic cells are marked by infection of undifferentiated and MEDII-pretreated ES cells ............................................................................................... 109 4.3.4 Lentiviral marking reveals that the hematopoietic output of a single EB is not necessarily from a single MEDII-generated precursor cell ............................................... 112 4.4 Discussion .................................................................................................................... 115 5. Discussion and future directions ........................................................................................ 118 5.1 Major contributions ....................................................................................................... 119 5.2 Implications and future directions .................................................................................. 121 5.2.1 Further characterization of the starting cell population ............................................ 121 5.2.2 Use of multi-step protocols to better direct differentiation ....................................... 123 5.2.3 Monitoring the clonal output of differentiating ES/iPS cells ..................................... 124 5.3 Future opportunities ...................................................................................................... 126 References............................................................................................................................. 129  vii  List of Tables Table 2.1 Primers used for Q-PCR analysis of RNA transcript levels ....................................... 49 Table 3.1 CFC outputs in individual and pooled EBs ................................................................ 85 Table 3.2 Summary of the distribution of CFC types in individual EBs ...................................... 93  viii  List of Figures Figure 1.1 Summary of the development of hematopoiesis in the mouse embryo .................... 16 Figure 1.2 Developmental steps from the morula to hematopoietic cells .................................. 26 Figure 1.3 Emergence of pre-hematopoietic mesoderm from the primitive streak..................... 27 Figure 1.4 Comparison of methods for differentiation of human ES/iPS cells ........................... 33 Figure 2.1 Comparison of hematopoietic outputs of human ES and iPS cells subjected to EB and OP9 differentiation protocols ............................................................................. 51 Figure 2.2 Characteristics of hematopoietic cells derived from human ES/iPS cells ................. 52 Figure 2.3 Optimization of passaging ES cell density and interval ............................................ 55 Figure 2.4 Changes in hematopoietic CFC output when the representation of undifferentiated cells in the starting ES/iPS cell population is increased ............................................................ 56 Figure 2.5 MEDII increases the frequency of CFCs subsequently generated in EB cultures..... 60 Figure 2.6 MEDII does not alter the kinetics of hematopoietic CFC appearance in subsequent EB cultures .............................................................................................................. 61 Figure 2.7 MEDII generates few MIXL1+ mesoderm cells ........................................................ 63 Figure 2.8 MEDII pretreatment does not significantly alter the kinetics of mesoderm formation in EBs.......................................................................................................................... 65 Figure 2.9 Mesoderm and hematopoietic gene expression in EBs derived from MEDII-pretreated and untreated starting cells ...................................................................................... 66 Figure 2.10 MEDII pretreatment with subsequent cytokine addition to the EB cultures increases the frequency of hematopoietic CFCs ...................................................................... 68 Figure 3.1 Models of CFC generation and expansion ............................................................... 78 Figure 3.2 The distribution of CFCs amongst individual EBs derived from human ES or iPS cells is highly skewed ...................................................................................................... 82 Figure 3.3 Representative analysis of CFC output from individual EBs derived from human ES or iPS cells .............................................................................................................. 84 Figure 3.4 MEDII pretreatment and cytokine addition to EB cultures increases both the proportion of positive EBs and the number of CFCs per EB..................................... 87 Figure 3.5 Relationship between EB size and hematopoietic CFC output ................................ 89 Figure 3.6 Distribution of CFC subtypes in individual EBs ........................................................ 92 Figure 3.7 Multipotent EBs have a greater output of total CFCs ............................................... 95 Figure 4.1 Schematic of lentiviral vectors ............................................................................... 102 Figure 4.2 Increased reporter gene expression in human ES and iPS cells using the PGK promoter ................................................................................................................ 106 Figure 4.3 Cells are infected with varying efficiency throughout differentiation ....................... 108 Figure 4.4 Reporter gene expression is maintained throughout differentiation into hematopoietic cells ....................................................................................................................... 110 Figure 4.5 Discrepancy between the proportion of GFP+ EB cells and GFP+ CFCs............... 111 Figure 4.6 Tracking of marked hematopoietic progenitors within single EBs........................... 114  ix  List of Abbreviations AGM α-MEM AP APC BL-CFC BMP CFC CFU CFU-E CM CMV cPPT DMEM E EB EGFP EPO ES ES-CFC FBS bFGF FACS FITC FLT3 G-CSF GEMM GFP GM GM-CSF HBSS HSC ICM IGF IL IMDM iPS IRES IVF KOSR  aorta gonad mesonephros alpha minimum essential medium alkaline phosphatase allophycocyanin blast colony-forming cell bone morphogenetic protein colony-forming cell colony-forming unit colony or burst forming unit, erythroid conditioned medium cytomegalovirus central polypurine tract Dulbecco's Modified Essential Medium embryonic day embryoid body enhanced green fluorescent protein erythropoietin embryonic stem embryonic stem colony-forming cell fetal bovine serum basic fibroblast growth factor fluoresence activated cell sorting fluorescein-isothiocyanate fms-related tyrosine kinase 3 granulocyte colony stimulating factor granulocyte/erythroid/macrophage/megakaryocyte green fluorescent protein granulocyte/macrophage granulocyte/macrophage colony stimulating factor Hanks Balanced Salt Solution hematopoietic stem cell inner cell mass insulin-like growth factor interleukin Iscove's Modified Dulbecco's Medium induced pluripotent stem internal ribosome entry site in vitro fertilization knockout serum replacement x  LIF LTC-IC M-CSF MEF MEM mEpiSC MND MPG MPSV NOD/SCID PBS PE PGK PI Q-RT-PCR RRE Rock RSV RT-PCR SCF SEM SFFV SIN SiTom SiV SSEA STAT TGF-β VE VPA VSV-G WPRE YS  leukemia inhibitory factor long-term culture-initiating cell macrophage colony stimulating factor mouse embryonic fibroblast Modified Essential Medium mouse epiblast stem cell MPSV enhancer, Negative control region deleted, dl587rev primer-binding site pCCL.PPT.MND.PGK.EGFP lentiviral vector myeloproliferative sarcoma virus non-obese diabetic severe combined immunodeficiency phosphate buffered saline phycoerythrin phosphoglycerate kinase propidium iodide quantitative reverse transcriptase polymerase chain reaction Rev responsive element Rho-associated kinase Rous Sarcoma Virus reverse transcriptase polymerase chain reaction stem cell factor standard error of the mean spleen focus forming virus self-inactivating pRRL.PPT.SF.IRES.TOM lentiviral vector pRRL.PPT.SF.IRES.VENUS lentiviral vector stage specific embryonic antigen signal transducer and activator of transcription transforming growth factor beta vascular endothelial valproic acid vesicular stomatitis virus glycoprotein woodchuck hepatitis virus post-transcriptional regulatory element yolk sac  xi  Acknowledgements  To Connie, thank you for giving me the opportunity to work in your lab. I could not have asked for a more stimulating and supportive environment. Your passion for science and dedication to research continue to amaze me. Thank you also to my supervisory committee: Keith, Pamela, Cheryl and Jamie for your many contributions over the years. To all the many members of my Eaves lab family, I couldn’t have done it without you. From Suzanne, who got me started on this path, to Mike O and Min who pioneered ES work with me in the lab, to Dianne, Darcy, and Glenn for their invaluable cell culture and technical expertise, and Dave Y, Juliya, Michael L, Sarah, Raymond, and Austin who helped maintain our cell lines, you have all been instrumental. To the “old” crowd, (from the Tim Hortons and Martinis era), you kept me going. Dave K, Mike O and Liz, Kai and Stef, you were always there when things got tough. To the “new” crowd, you got me to the finish line. Thanks to all of “Team Mouse” and Claudia, Kristen, Carla, plus my deskmates Maisam and Mike C for the final push. To my extended family of friends in TFL, Vancouver, Edmonton, and beyond, sorry if I’ve been a bit absentee lately, but I know you’re right here with me 100%. And to my family, especially my parents, thanks for all your support. I finally made it, and I guess you should have made my initials P, H and D. Financial support for my graduate training was much appreciated from scholarships received from The University of British Columbia, the Natural Sciences and Engineering Research Council of Canada, the Michael Smith Foundation for Health Research, and the Canadian Institutes of Health Research.  xii  1. Introduction  1.1 Stem cells  The general concept of a stem cell is a cell that possesses the ability to perpetuate a state that allows for subsequent activation (in response to appropriate external stimuli) of gene expression programs and cellular functions characteristic of one or more fully differentiated, specialized cell types. However, a precise molecular definition of this hypothetical state(s) is lacking. As a result, the utility of the term “stem cell” remains grounded in the specificity and validity of operational assays designed to detect the clonal demonstration of the expected differentiation and self-renewal properties. Cells that can be detected in such assays have now been discovered in many adult tissues, including the bone marrow, skin, brain, and intestine. Although not yet defined, the epigenetic state of these adult stem cells is thought to play a major role in restricting their differentiation potential under normal physiological circumstances to the generation of cells within the tissue in which they are found. The early embryo contains cells with much broader developmental abilities, and it was long hypothesized that some of these cells would exhibit stem cell properties, including the potential to differentiate into a large number of different tissues, in other words, these cells would be pluripotent. Cells with such properties were first isolated from malignant teratocarcinomas, which could be propagated as cell lines in an embryonic-like state, some of which had the capacity to generate derivatives of all three embryonic germ layers1. In 1981, after many attempts, pluripotent stem cell lines were generated from normal murine blastocysts2,3. These pluripotent stem cells are similar in concept to the stem cells found in adult tissues – they possess both the ability to differentiate into a variety of different cell types, and also the ability to self-renew, generating progeny with the same developmental potentialities.  1  What distinguishes pluripotent stem cells from other stem cell populations (except germinal stem cells) is their unlimited differentiation capacity; murine pluripotent stem cells have the ability to generate any type of cell in the body4,5, and their human equivalents are inferred to have the same capability. With this extensive differentiation capacity comes a similarly wideranging set of possible uses for these cells. It is hoped that they will eventually be useful to delineate the molecular stages of development of various cell types, to generate cells for drug and toxicity testing, for cell and tissue replacement therapies, and for modelling diseases. Another unique feature of pluripotent stem cells is that they are largely an in vitro phenomenon; pluripotent cells in the early embryo do not exhibit extensive self-renewal capacity in vivo and rapidly differentiate into cell types with more restricted potential. However, if these cells can be captured and subjected to the correct conditions in vitro, they appear to have unlimited selfrenewal ability. In contrast to adult stem cells, which are typically difficult to isolate and propagate in culture, pluripotent stem cells can be maintained as permanent lines of cells that retain their pluripotency and proliferative capacity apparently indefinitely.  1.1.1 Human embryonic stem cells  Human embryonic stem (ES) cells were first derived by Thomson et al. from cultures of the inner cell mass (ICM) of blastocyst stage embryos6. These and later human ES cell lines have been derived from embryos generated by in vitro fertilization (IVF) clinics for reproductive purposes and subsequently designated as surplus and donated for research purposes with informed consent. Most lines have been derived from blastocyst stage embryos, but morula stage embryos7 and individual blastomeres from 8-cell embryos8,9 have also been used to derive lines with similar properties. Many criteria have been used to assess the pluripotency and self-renewal potential of different putative pluripotent cell lines. These include many shared morphological features, expression of particular antigens and transcription factors, and clonal 2  growth behaviour. Bulk cultures of ES cells from most species consist of well-defined clusters of homogeneous cells that are small in size, with a high nuclear to cytoplasmic ratio. Human ES cells also express Stage Specific Embryonic Antigen (SSEA) 3, SSEA4, TRA-1-60, TRA-1-81, alkaline phosphatase (AP), and/or OCT4 (also known as POU5F1)6. A quantitative, functional measure of undifferentiated human ES cell numbers can be obtained by measuring the number of cells that can generate colonies containing at least 30 AP+ cells in an ES colony-forming cell (ES-CFC) assay10. The definition of ES cells also requires demonstrating an ability to differentiate into derivatives of all 3 germ layers. This can be achieved in vitro by culturing ES cells in non-adherent 3-dimensional aggregates called embryoid bodies (EBs) followed by immunostaining or reverse transcriptase polymerase chain reaction (RT-PCR) analysis for markers of each germ layer. However, the gold standard test for human ES cell pluripotency is functional analysis in an in vivo teratoma assay. For this procedure, large numbers of cells are injected into immunodeficient mice. If ES cells are present, tumours form several weeks later and these tumours can be shown histologically to contain cells derivative of all 3 germ layers6. More rigorous assays that can be performed to test the developmental capacity of mouse ES cells cannot be carried out using human cells. These include injection of cells into blastocysts to generate chimeric animals, or into tetraploid embryos to generate entirely ES cell-derived animals (as the tetraploid cells can give rise only to extraembryonic tissues)4,5. The self-renewal ability of ES cells is tested by their ability to propagate indefinitely in culture, while still retaining pluripotency by the criteria described above. Because it is unrealistic to test for “indefinite” maintenance, continuous passaging over a time period of several months to a year is typically accepted as an adequate substitute. In the very first report describing the generation of human ES cells, the cells were shown to be subsequently maintained in an undifferentiated state by co-culturing them with a feeder layer of irradiated primary mouse embryonic fibroblasts (MEFs) in a medium that contained a carefully selected batch of fetal bovine serum (FBS). Since then, the field has advanced rapidly, 3  with groups developing new ways to derive and culture these cells to decrease variability in growth and spontaneous differentiation, and thus increase the proportion of undifferentiated cells and the overall homogeneity of the undifferentiated population. This was accomplished by replacement of undefined biological and animal products with more defined components and also the determination of key signalling pathways to maintain the pluripotent state. The elimination of animal-derived products such as MEFs was also desired in order to generate cell lines that would be more appropriate for eventual clinical translation. To this end, FBS was rapidly replaced by a commercially available serum replacement in most labs11,12. It was also shown that the MEF feeder layer could be replaced by plates coated with extracellular matrices like Matrigel in combination with ES cell medium conditioned by MEFs13. Alternatively, the conditioned medium could be replaced by serum-free media containing a variety of growth factor supplements important for human ES cell pluripotency, most commonly including basic fibroblast growth factor (bFGF), transforming growth factor beta (TGF-β), and/or activin A14-19. In some conditions, purified extracellular matrix components such as laminin13,14 or fibronectin15,20 were found to successfully replace the more undefined composition of Matrigel. The dependence of human ES cell propagation on signalling pathways activated by FGF and activin, and the inhibition of signalling induced by bone morphogenetic proteins (BMPs)21 distinguishes human and mouse ES cells. The latter rely, instead, on signal transducer and activator of transcription 3 (STAT3) signalling activated by leukemia inhibitory factor (LIF) which is dispensable for human ES cells22,23, and signalling activated by BMPs24. Human ES cells are also distinguished from their murine counterparts by a much poorer survival as single cells and thus a low cloning efficiency, typically less than 1%10,11,25,26. In contrast, mouse ES cells can be propagated clonally with relative ease. For this reason, human ES cells are typically passaged as clumps, using either mild enzymatic treatment with collagenase or dispase, or using laborious mechanical passaging techniques in which clumps of undifferentiated cells are selected and dissected with fine needles before transfer to new culture 4  plates. This difficulty in maintaining isolated human ES cells in a viable state poses drawbacks for their analysis, e.g., using fluorescence activated cell sorting (FACS), or their manipulation, by transfection. However, by repeated passaging with trypsin, sublines of human ES cells can be isolated that are adapted to this type of enzymatic treatment, and these cells have higher cloning efficiencies25,27,28. On the other hand, it is unclear as to what allows the cells to achieve this “adapted” state or what implications there might be on the differentiation control of the cells. An associated acquisition of mutations has been reported, although it has not been demonstrated that these are responsible for the altered survival properties25,29. These cells may also contain abnormalities undetected by the standard karyotypic analysis performed, as arraycomparative genomic hybridization can more sensitively detect genomic aberrations in ES cell cultures30,31. Human ES cells can also be treated transiently with an inhibitor of Rho-associated kinase (Rock) (Y-27632) to reduce their susceptibility to dissociation induced apoptosis which thus increases their cloning efficiency10,26. These differences between human and mouse ES cells led researchers to ask why they behave so differently; possibly reflecting species-specific differences in the regulation of the same early stages of development or, alternatively, an origin from distinct developmental stages within the early embryo. Interestingly, pluripotent cell lines derived from the mouse epiblast (mEpiSCs) have now been found to share many growth characteristics, gene expression features, and growth factor responsiveness with human ES cells, and differ in these respects with mouse ES cells. mEpiSCs appear developmentally more advanced than mouse ES cells in that they do not contribute to the development of embryos when injected into blastocysts32,33. More recent findings further support this view based on the demonstration that human ES cells can be converted into a more “naïve” state associated with several characteristics of mouse ES cells34. This suggests that classical human ES cell lines may represent a cell type somewhere in between the cells present in the ICM and the epiblast. Despite this apparent difference in exact developmental stage in vivo that human and mouse ES cells most closely resemble, it appears 5  that they both rely on the same core network of transcription factors as master regulators of their pluripotent state, including OCT4, NANOG, and SOX235-37.  1.1.2 Induced pluripotent stem cells  In 2006, Takahashi and Yamanaka first demonstrated that pluripotent cells could be generated from both embryonic and adult mouse fibroblasts by the introduction of only 4 transcription factors – out of a pool of 24 candidate genes, Oct4, Sox2, Klf4, and c-Myc were sufficient when introduced by retroviral transduction38. The similarity in the core elements that regulate mouse and human pluripotency was reemphasized when a year later, the same factors were reported to reprogram human fibroblasts and mesenchymal stem cells into cells with similar pluripotent properties39,40. The lines thus derived were called “induced pluripotent stem” (iPS) cells, to reflect their remarkable similarity to ES cells in terms of their morphology and growth requirements, marker and gene expression profiles, and their ability to differentiate into cells of all 3 germ layers in vitro or in teratomas generated in vivo. The cell lines produced are thought to be capable of normal differentiation due to the induced silencing of the retroviral cDNAs used to activate their endogenous counterparts. Due to the established involvement of c-Myc and Klf4 in oncogenesis, and the demonstrated ability of the introduced c-Myc construct to cause eventual tumours to develop in mice derived from iPS cells41, there has been a particular interest in strategies to replace these factors with other genes or molecules. It was later shown that both mouse and human fibroblasts could be reprogrammed without the use of a c-MYC-encoding vector, but the frequency of iPS cells obtained was reduced by 20-fold and >100-fold, respectively42,43. The contributions of both Klf4 and c-Myc have now been shown to be replaceable, by using NANOG and LIN28 in human fibroblasts44, or the estrogen-related receptor β (Esrrb) in mouse fibroblasts45. 6  Under these conditions, the frequency of successful reprogramming events is extremely low, about 0.01% or less in both human and mouse cells. It is thought that this is limited significantly by the challenge of delivering appropriate amounts of each factor. This idea is supported by the finding that the generation of secondary iPS cell lines from differentiated derivatives of primary iPS cells created with inducible vectors is more rapid and >100-fold more efficient than the primary reprogramming46,47. It has also been shown that the introduction of 6 genes – OCT4, SOX2, KLF4, c-MYC, NANOG and LIN28, instead of 4, together with an improved delivery method, results in a >100-fold increase in reprogramming efficiency of human fibroblasts (to about 1%)48. Some of the other factors identified that appear to limit reprogramming include replicative senescence and apoptosis due to DNA damage, both of which can be at least partially overcome by inhibition of the p53 and/or Ink4a/Arf pathways49-55. Incomplete epigenetic reprogramming also seems to hinder the efficiency of iPS cell generation since this is also improved if the cells are exposed to histone deacetylase inhibitors like valproic acid (VPA) or Trichostatin A (TSA), DNA methyltransferase inhibitors such as 5-azacytidine, or one of several histone methyltransferase inhibitors (reviewed in Feng 200956). Full reprogramming also seems to be supported by the inhibition of the MEK and GSK3 pathways56. Another concern with iPS cells has been the possibility of insertional mutagenesis caused by the use of integrating viral vectors. Accordingly, methods for introducing reprogramming factors with polycistronic vectors to decrease the number of insertion sites57,58, or methods not involving DNA integration are being developed. Non-integrating methods include genetic manipulations in which the vector can subsequently be excised. One approach is to use floxed lentiviral vectors so that they can later be excised by introduction of Cre-recombinase into the cell59. Another approach is to use a polycistronic piggyBac transposon-based vector which can then be excised by introducing a transposase60,61. These methods tend to have a similar efficiency to other viral methods of reprogramming, with the added complication that lines must be screened for proper removal of the vectors. Other groups have attempted reprogramming 7  using non-integrating adenoviruses62,63, repeated plasmid transfection64, or specialized vectors such as Epstein-Barr nuclear antigen-1-based episomal vectors capable of replication in the cytoplasm48 or minicircles65. Reprogrammed human cells obtained by repeated transfection with synthetic modified mRNA molecules has also been recently reported66, and direct protein delivery using poly-arginine tagged factors and VPA have successfully reprogrammed mouse fibroblasts67. However, the efficiency of reprogramming obtained with most of these nonintegrating methods is greatly reduced when compared to those obtained using standard viral methods. Also to be considered is the type of somatic cell used for reprogramming. The most useful starting cells would be easily accessible and reprogram with high efficiency. Fibroblasts and mesenchymal cells were the first types of cells to be reprogrammed38-40,44, but now a wide variety of starting somatic cells have been used, showing that this phenomenon is not specific to the original cell type selected. In fact, it appears that human iPS cells are more efficiently generated from keratinocytes46,68, adipocyte “stem” cells69, and mesenchymal stem cells present in umbilical cord and amniotic membrane70 than from fibroblasts, which, in turn are superior to hematopoietic cells. Despite this, reprogramming of hematopoietic cells is of great interest due to the accessibility of these cells in large numbers, and has been achieved with CD133+ cord blood cells71, mobilized CD34+ peripheral blood cells72, as well as cells in unmanipulated peripheral blood which has enabled the generation of iPS cells from T and B cells with rearranged T-cell receptor and immunoglobulin genes, respectively73-76. In general, more primitive adult cell populations or cells from developmentally earlier stages seem to be more amenable to reprogramming. For example, hematopoietic stem and progenitor cells are reprogrammed at up to 100-fold greater efficiencies than their more differentiated counterparts77. Similarly, neural stem cells can be reprogrammed using only exogenous Oct478. When fibroblasts are used, the highest reprogramming efficiencies are obtained starting from ES-derived fibroblasts followed by embryonic and fetal fibroblasts, then neonatal fibroblasts, 8  and adult fibroblasts seem to reprogram most inefficiently38,39,44. The benefit of being able to generate pluripotent cell lines from any individual is already becoming apparent as the numbers of patient-derived, disease-specific human iPS cell lines available for studies of disease increases59,79-81. With the wide range of cell types, factors, and protocols now being used to generate these cell lines, there is also a growing debate as to how to best define and characterize them8284  . Much of this debate centres on the stringency of the tests that would be most informative; the  more stringent and rigorous the testing requested, the more time, resources, and expertise needed to obtain the answers. At the same time, there is evidence that many iPS cell lines can be generated that appear to be similar, but display different abilities to read out in assays typically applied to define ES cells. For example, the first mouse iPS cell lines generated were pluripotent as assessed by a teratoma assay, but failed the gold standard test of contribution to chimeric blastocysts38. Since then, murine iPS cell lines have been generated that can contribute to chimeric mice, including the germline, and even succeed in tetraploid embryo complementation assays41,85,86. This has recently been linked to the repression of the imprinted Dlk1-Dio3 gene cluster on chromosome 12q87. Similarly, there is evidence that human iPS cells may also vary in their behaviour in teratoma assays with some generating teratomas that contain >70% differentiated cells, and others that generated teratomas that contain as few as 5% differentiated cells88. It also remains to be seen how similar different human ES and iPS cell lines will prove to be, both with respect to their undifferentiated state and their differentiation capacity. Human ES cell lines have been available for over a decade, and yet only recently has there been a large scale effort to characterize how basic variables such as culture conditions, origin and inter-lab technique affect the stringent description of their undifferentiated state on a large scale89-92. At the level of global gene expression, there do appear to be some differences, both between different iPS cell lines, and between ES and iPS cell lines, which may be related to 9  reprogramming factor integration or the somatic cell type used59,93,94. Such large scale comparisons of differentiation capacity are even less advanced. There have now been several reports of differences in the capacity of human ES/iPS cell lines to differentiate towards specific functional cell types, but it is still not clear what factors influence these differences in most cases95-100. Very recently, it has been suggested that some of the difference in iPS cell differentiation capacity could be caused by epigenetic differences that reflect the somatic cell type used to generate the iPS cell line101,102. As the number and types of pluripotent cell lines continue to expand, a more sophisticated characterization of their properties will be important to establish their utility as a starting point for future investigations of how differentiation along particular lineages is regulated. In this regard, it is interesting to note that, in spite of their differences, many human ES and iPS cell lines share many similar properties. It will thus be interesting and important to define the functional consequences of variations in the maintenance and differentiation conditions as well as functional heterogeneity due to the different genetic backgrounds of the cell lines being studied.  1.2 Hematopoietic cells in the adult  Hematopoietic stem cells (HSCs) are multipotent cells, the majority of which reside in the bone marrow throughout adult life. These cells are responsible for replenishing billions of mature blood cells in our bodies every day. Their progeny include lymphoid lineage cells, such as B-cells, T-cells and natural killer (NK) cells, as well as myeloid lineages, including erythrocytes, megakaryocytes, dendritic cells, monocytes, and granulocytes. Upon demand, HSCs execute extensive self-renewal divisions, producing at least one daughter cell with the same functional capabilities as the parent cell. Hematopoiesis is organized in a hierarchical system, in which hematopoietic progenitors become gradually more restricted in their lineage 10  potential and proliferative capacity. A classical view of hematopoiesis has been that the first lineage choice determines lymphoid or myeloid cell fate, followed by restriction down these lineages103,104. However, there is also evidence that this hierarchy is not always so clear cut, for example, there are some restricted progenitors with lymphoid potential that also generate macrophages105-107. The generally described hierarchy of hematopoietic cell differentiation has been elucidated with the help of functional (developmental) assays. These assays have been developed to define hematopoietic cell types by the display of their differentiation activity and, in some cases, its persistence in vitro or in vivo. The hematopoietic colony-forming cell (CFC) assay measures both lineage-restricted and multilineage progenitors with limited proliferative potential that form colonies of mature cells in vitro when plated into semi-solid methylcellulose medium containing certain hematopoietic growth factors. The long-term culture-initiating cell (LTC-IC) assay measures very primitive cells that can proliferate for at least 6 weeks in cocultures containing a source of “stromal cells” and whose progeny at 6 weeks include CFCs108. The LTC-IC assay detects a population of cells that can overlap with HSCs defined by in vivo assays reconstitution assays109. The most definitive test for HSCs is to examine them in vivo for long-term multilineage reconstitution of the blood system of a transplant recipient, which, if performed using limiting cell doses in mice that are otherwise protected from marrow failure, can also be used to quantify the number of input HSCs in the test population. These tests rely on the use of genetically distinct donors and recipients, and the hematopoietic system of the recipient is generally compromised by myeloablative treatment to maximize the sensitivity of detecting HSCs in the input population. In vivo assays using human cells have been performed in immunocompromised mice such as the non-obese diabetic/severe combined immunodeficiency (NOD/SCID) strain, or fetal sheep recipients before the development of the immune system110-112. However, it is important to note that the extent of the immunodeficiency can influence the detection of different human cell types 11  and the sensitivity of the assay. For example, NOD/SCID-β2 microglobulin knockout mice are more severely immunocompromised, and allow the detection of additional cell populations when compared with the same assay using NOD/SCID mice113. Recently, mice deficient for the interleukin (IL)-2-receptor common-γ chain have been backcrossed onto the NOD/SCID strain by 2 groups. Both strains are highly immunodeficient and long-lived and allow very high levels of engraftment of human cells114,115. Many strategies have been used to phenotypically identify HSCs, for example, mouse HSCs can be isolated at near purity using the markers CD48-CD150+ in combination with either c-kit+Sca-1+ or E-PCR+CD45+ 116,117. These marker combinations have the advantage that they appear to be relatively stable throughout ontogeny, unlike previously identified markers which can vary with ontogeny or cell cycle status, such as CD34, MAC1, or the ability to exclude the dyes Hoescht 33342 and rhodamine-123118-122. Markers that allow equivalent purities of human HSCs have been even more difficult to identify. The highest reported purities of human HSCs have been in the CD34+CD38-CD90+CD45RA- fraction of human cord blood123,124. Human HSCs have also been identified by their high levels of aldehyde dehydrogenase (ALDH) activity125. However, the stability of these markers during human HSC activation or during ontogeny is less well understood. For example, CD34 expression seems to be more similar throughout development, but there have also been reports of rare CD34- HSC126,127. HSCs in adult human hematopoietic tissues are extremely rare and although phenotypically similar, they are less well characterized. Mouse HSCs have been shown to expand greatly in culture when engineered to overexpress HOX transcription factors, for example, HoxB4 or a fusion product of Nup98 and the homeodomain of HoxA10128,129. In contrast, unmanipulated HSCs have proven difficult to maintain in culture without loss of their cardinal stem cell properties. The best that can typically be achieved is maintenance of rigorously defined HSCs over relatively short culture periods (days) in cultures that contain stem cell factor (SCF) and/or fms-related tyrosine kinase 3 12  (FLT3)-ligand in combination with factors that act via STAT 3 (e.g. IL-11 or IL-6 or thrombopoietin)130-134.  1.3 Generation of hematopoietic cells from pluripotent stem cells  The ability to maintain and differentiate ES cells in culture has opened up a new avenue to study the development and generation of specific differentiated cell types, since the cells can be more easily visualized, accessed, analyzed and manipulated than is the case in vivo. This is particularly true for the study of human tissue development, where samples are rare and difficult to obtain during the latter two-thirds of embryonic and fetal development. Although the defined organisation and structure of the embryo may be lost in in vitro systems used to study ES differentiation, much has still been gained by monitoring cells over time under controlled conditions of external signals provided to the cells. Indeed, when combined with the use of genetic tools, analyses of ES cell differentiation processes have emerged as very powerful approaches. Many groups have shown that hematopoietic cells can be reproducibly derived from mouse or human pluripotent stem cells. The basic differentiation methods used are very similar for both species, and fall into 2 major categories: 2-D co-culture with various stromal lines, and 3-D EB cultures. The stromal cell line most often used for mouse ES cell differentiation is the OP9 line, originally derived from the bone marrow of macrophage colony stimulating factor (MCSF)-deficient mice135. This line has also been used to differentiate human ES cells136 and in both species allowed generation of both myeloid and lymphoid hematopoietic cells. A variety of additional cell lines have been used successfully to support the generation hematopoietic cells from human ES cells. These include S17 cells from mouse bone marrow136-138, C166 cells from mouse yolk sac endothelium137, immortalized human fetal liver138, AM20 from mouse AGM, UG26 from mouse embryonic urogenital ridge, and EL08 from mouse fetal liver, as well as 13  primary mouse AGM or fetal liver stromal cells139. Each of these cell types was reported to support the generation of hematopoietic CFCs at variable frequencies from human ES cells in the presence of serum and the absence of any exogenous cytokines. EB cultures allow ES cells to form 3D aggregates of cells under non-adherent conditions. The formation of EBs, in the absence of exogenously supplied factors to maintain pluripotency and/or the addition of differentiation factors, has allowed the differentiation of ES cells into many different cell types to be achieved, including hematopoietic cells. For mouse ES cells, EBs can be generated from single cells in liquid culture or in semi-solid methylcellulose medium, or by the reaggregation of small numbers of cells in hanging drops140. As human ES cells exhibit very poor survival as single cells, EB cultures of human ES cells are initiated with larger numbers of reaggregated cells141,142 or with clumps of cells143-145. For hematopoietic cell differentiation many iterations of the EB method have been developed including serum and serum-free methods, with or without the addition of various hematopoietic cytokines. As with coculture methods, many variations support the generation of hematopoietic CFCs, although the efficiencies vary. In general, the presence of serum and/or cytokines increase the CFC output in otherwise equivalent conditions141,143,146-150. Recent reports have suggested that mouse151-153 and human39,95,154,155 iPS cells can also generate hematopoietic cells in vitro using the same differentiation protocols as for ES differentiation. Analysis of hematopoietic development in the early embryo has been key to the development and validation of the ES cell model. In order for an ES differentiation protocol to be accepted as a model of development, replication of the developmental pathway seen in the embryo needs to be demonstrated. At the same time, however, ES cells have served as a useful tool to test many hypotheses and concepts that had existed for many years, but had been difficult or impossible to test in vivo. While not a perfect model system, the study of ES cell differentiation into hematopoietic cells, especially from mouse cells which have a lengthier history, has greatly helped to advance understanding of the early development of the 14  hematopoietic system. These studies will be highlighted in the following sections describing the generation of hematopoietic cells in the embryo.  1.4 Hematopoiesis in the mammalian embryo  1.4.1 Sites of hematopoiesis in the embryo  The hematopoietic system is one of the earliest to form during embryogenesis. The first hematopoietic cells appear in the yolk sac (YS) on embryonic day 7.5 (E7.5) in mouse and on E16 in humans, long before the bone marrow, the primary hematopoietic organ in the adult, has even begun to develop156,157. Hematopoietic cells are then found in spatially and temporally distinct locations (Figure 1.1); by E9 in mouse, the aorta-gonad-mesonephros (AGM) region and the placenta158-161. Later, hematopoietic cells migrate through the circulation to colonize the fetal liver (E10), the fetal thymus (E11), the fetal spleen (E12), and the fetal bone marrow (E16)162-164. Although the first hematopoietic cells colonize the bone marrow at E16, the fetal liver remains the major site of hematopoiesis until around birth165,166. The human system is less well characterized with respect to the exact timing of the initiation of hematopoiesis in different locations, but appears to generally follow a similar pattern as that seen in mice, with the YS, AGM, and placenta containing hematopoietic cells by the 3rd week of development, and the fetal liver by the 4th week which is subsequently the main hematopoietic organ from weeks 6 to 16. Other hematopoietic organs are seeded later on, including the bone marrow which, unlike in the mouse, becomes the major hematopoietic organ during the second half of gestation167-169. Also in contrast to the mouse, the human placenta continues to contribute to hematopoiesis until birth170 whereas the mouse placenta contributes very little to hematopoiesis at birth171,172. Due to this diversity of embryonic hematopoietic organs, the study of hematopoiesis in the embryo has been fraught with complexities, increased 15  Figure 1 Figure 1.1  Fetal liver  Embryonic sites of hematopoiesis  Placenta Spleen AGM Thymus Bone marrow  Yolk sac Time of development E0  E5  E10  E15  birth  Primitive erythroid Definitive CFU-E Multilineage CFCs HSCs  Appearance of hematopoietic cell types  Figure 1.1 – Summary of the development of hematopoiesis in the mouse embryo The locations of hematopoietic cells during development are shown above the timeline. This includes sites where hematopoietic cells are generated de novo (pink) or are concentrated in supportive niches (green). The timings of the first appearance of different types of hematopoietic cells are indicated below the timeline.  16  further by the movement of hematopoietic cells throughout the embryo in the circulation, which begins with the heart beating at E8.25 and reaches a steady state by about E10 in mice173.  1.4.2 Primitive hematopoiesis  There are also distinct types of hematopoietic progenitors present in the embryo and the adult which introduces additional sources of complexities to understanding the mechanisms regulating the generation of hematopoietic cells. “Embryonic” or “primitive” hematopoiesis refers specifically to a transient wave of hematopoiesis that is restricted to the YS. In this context, “primitive” refers to the similarity in appearance of the cells to the hematopoietic cells of more “primitive” vertebrate organisms such as fish and amphibians and not, as it does in reference to the adult hematopoietic hierarchy, to hematopoietic cells with a higher proliferative and selfrenewal capacity. For example, embryonic hematopoiesis has largely been characterized by the presence of “primitive” erythroid cells, beginning in the mouse embryo at E7.5. These are larger in size than their adult counterparts, and express embryonic forms of hemoglobin (zeta, betaH1, and epsilon). These erythroid cells also remain nucleated and divide even after they are released into the circulation. They continue to divide until E13.5 when they start to mature and express small amounts of adult type hemoglobin in addition to the embryonic forms, and eventually enucleate by E16.5174-176. Human embryos produce a similar early wave of primitive erythroid cells expressing embryonic hemoglobins (zeta and epsilon) that circulate in a nucleated form prior to maturation and enucleation177. Distinguishing other hematopoietic cell types that may have a distinct “primitive” form in the embryo has been more difficult, but it has been reported that primitive megakaryocyte CFCs are present concurrently and in the same locations as primitive erythroid CFCs in the mouse embryo and these produce megakaryocytes with lower ploidy that give rise to larger platelets with small granules and large canalicular systems178. Similarly, primitive macrophage progenitors have been detected179 and these early 17  macrophages are characterized by the lack of a monocytic stage180. It is as yet unclear whether these different types of “primitive” progenitors share a common origin.  1.4.3 Definitive hematopoiesis  HSCs are not found amongst these “primitive” populations and hence the continued production of embryonic type hematopoietic cells does not persist. Instead, a second wave of hematopoiesis occurs producing cells referred to as “definitive”. The production of these cells begins shortly after the initiation of primitive hematopoiesis in the embryo and generates hematopoietic progenitors of types that can also be found in the adult. For example, definitive erythrocytes are smaller than their embryonic counterparts, express adult hemoglobins (alpha and beta), and enucleate prior to entering the circulation. In human embryos, the first definitive erythrocytes express a fetal version of hemoglobin (alpha and gamma) that does not have a mouse equivalent, and then transition to adult type hemoglobins (alpha and beta) nearer birth. Definitive erythroid CFCs can be detected as early as E8-8.5 in the mouse embryo. Shortly thereafter, multi-lineage CFCs emerge and this is then followed by cells with myeloid and lymphoid progenitor activity159,179. The first HSCs with self-renewal ability are also produced shortly after that, and thus a persistent source of blood cell generation is set up (Figure 1.1). Interestingly, these HSCs, as defined by their ability to reconstitute sustained blood cell production following transplantation into an irradiated adult recipient, are not detected until E10.5 in the mouse embryo and are then found in the AGM, YS, and placenta172,181,182. The appearance of hematopoietic progenitors in the embryo in the reverse order that would be predicted from the adult hierarchy highlights the fact that the adult-type hematopoietic CFCs that appear before HSCs can be detected may have unique developmental pathways in the early embryo. Either these CFCs are derived from cells that are not self-renewing HSCs, or the HSCs in the early embryo are not detectable in traditional assays. The latter could be true if the fetal 18  cells are not yet capable of recognizing the adult environment, as is suggested by the observation that long-term engraftment can be obtained when hematopoietic cells from E9 mouse embryos are injected into the livers of neonatal mice183. Alternatively, early embryo cells could require additional signals in order to become HSCs, either to gain self-renewal potential or to mature and become detectable assays for adult cells, as is suggested by the findings that E9 cells from mouse YS can gain HSC ability when engineered to constitutively express HoxB4 184 and that E8.5 YS and AGM gain HSC ability when co-cultured on an AGM stromal line185. Mouse ES cell differentiation to hematopoietic cells in vitro mirrors these early in vivo events. Primitive erythroid progenitors are the first to be detected, and these are produced transiently and undergo a transition from embryonic to adult hemoglobin expression as seen in the embryo. Subsequently, definitive erythroid progenitors appear and these persist over a longer timeframe and express exclusively adult hemoglobins. Finally, multilineage progenitors and lymphoid cells are produced after even longer periods of time135,148,186,187. At early timepoints primitive erythroid progenitors can share a common precursor with definitive erythroid progenitors and other hematopoietic CFCs, although the ability of these cells to also generate other mesodermal cell types was not tested188. Further studies showed that the primitive and definitive lineages arise largely independently from one another152,189. It is interesting to note, however, that HSCs have not yet been obtained directly from unmanipulated ES cells or their progeny, with only rare reports of successful transplants190,191 possibly mirroring their failure to be generated in the very early embryo. On the other hand, similarly to what was observed with early YS cells, genetic manipulation to overexpress genes such as HoxB4, Stat5, or Cdx4 can promote engraftment or even HSC generation from these cells184,192-194.  19  1.4.4 Genes critical for hematopoietic development  Several genes that have critical roles in the earliest stages of hematopoietic development have been identified in genetic analysis of mice, and studies both of mutant mice and ES cell lines have been very useful in elucidating the precise effects of deficiencies in these genes. Many of these genes turned out to be transcription factors, and some affect both primitive and definitive hematopoiesis including Gata1, Gata2, and Scl. Runx1 is an example of a critical transcription factor that affects only definitive hematopoiesis. Gata1 and Gata2 are members of a larger family of transcription factors defined by a shared DNA binding site (GATA). Gata1-null mice die between E10.5 and E11.5 of severe anemia. Both primitive and definitive erythroid progenitors are formed, but they do not mature past the proerythroblast stage195,196. Gata1-null ES cells produce other types of blood cells, including macrophages, megakaryocytes and mast cells in vitro, but when used to form chimeric mice, contribute minimally to blood formation in vivo 197,198. Gata2-null mice also die between E10.5 and E11.5 of severe anemia with decreased numbers of primitive erythrocytes, and severely reduced numbers of definitive progenitors. Gata2-null ES cells are similarly deficient in their ability to generate either primitive or definitive progenitors in vitro. Chimeric mice derived from these ES cells show no contribution of the Gata2-null cells to hematopoiesis, implying that they do not allow the formation or survival of any self-renewing HSCs199,200. The effect on hematopoietic progenitor expansion appears to be dose-dependent as heterozygotes show reduced numbers of progenitors in the YS and AGM, and reduced numbers of HSCs in the AGM. Although, this deficiency in numbers is corrected by adulthood, Gata2-heterozygote HSCs compete poorly against wildtype HSCs in a co-transplant setting201. Scl, also known as Tal1, is involved in the generation of both primitive and definitive hematopoietic cells. When the gene is knocked out in mice, death occurs between E8.5 and E10.5 and the embryos lack all types of blood cells202. Further evidence that Scl is required for 20  the generation of all hematopoietic cells was provided by the study of Scl-null ES cells. Analyses of their ability to differentiate either in vitro, or in vivo in chimeras showed a failure to contribute to any of the hematopoietic lineages203,204. Thus, Scl appears to have a key role in the specification of hematopoietic cells from their precursors in the embryo. Despite this dramatic effect during development, Scl does not appear to be required for the maintenance of HSCs in the adult once they have been formed, although effects on downstream erythroid and megakaryocyte differentiation can then be seen205,206. Runx1 (also known as Aml-1 or Cbfα2) is the most studied example of a gene that does not appear to be required for primitive hematopoiesis, but is required for the formation of all definitive hematopoietic cells. Mice null for this gene die around E12.5 and have normal numbers of nucleated primitive erythroid cells in the circulation, but lack definitive blood cells of any kind. Studies of Runx1-null ES cells showed that in vitro, they also generate primitive erythroid progenitors but no other hematopoietic lineages. In addition, these null ES cells do not contribute to the adult hematopoietic compartment of chimeric mice207,208. Heterozygotes also show alterations in their development of hematopoietic cells, including decreased progenitor numbers in the early embryo and spatial and temporal effects on HSC distribution immediately after their emergence208,209. However, once the hematopoietic system has been established, Runx1 is no longer required for the maintenance of HSCs, although its absence confers alterations in the generation of megakaryocytes and certain types of lymphocytes210-212.  1.5 Developmental origins of HSCs  1.5.1 Identification of hemogenic sites in the embryo  The studies described above determined the timing of the first emergence of hematopoietic progenitors and HSCs and localized them throughout development. However, 21  these experiments did not identify the precise sites and hence cell types from which the first hematopoietic cells arise de novo. Likewise, these early studies did not distinguish sites of hemogenic cells from those that might provide a supportive niche for pre-existing cells to reach via the circulation, a distinction which could be critical in an attempt to replicate the development of HSCs in an ES model. Several strategies have been employed to distinguish hemogenic from supportive niche cells in the early embryo. First, putative hemogenic tissues have been isolated separately prior to the onset of circulation, cultured in permissive conditions in vitro, and then tested for the presence of hematopoietic cells to determine if the isolated tissue had intrinsic hematopoietic potential. Although different groups have not always obtained consistent results with respect to the different types of cells and progenitors generated by each tissue, there is now much data to show that in mice, the YS, AGM, and placenta are all capable of independently generating definitive hematopoietic cells158,185,213-216. Since the cells must then be maintained in vitro for further development, it is not clear to what extent these experiments represent the actual contribution of each of these tissues in vivo, and the different culture conditions used by different groups likely contribute to the differences in the numbers and types of cells obtained. Thus the tissues capable of independently generating HSCs remain controversial. For example, using cultures of pre-circulation whole organs, Cumano et al. obtained HSCs only from the AGM, whereas Matsuoka et al. obtained HSCs from both pre-circulation YS and AGM cells after culture with an AGM stromal cell line185,214. Recently, various knockout mouse models lacking a circulation have also been used to study this phenomenon. Vascular endothelial (VE) cadherin-null mice lack the vasculature connecting the embryo to the yolk sac with the result that hematopoietic cells cannot circulate between these locations. These mice die around E12, which is later than the time when the first hematopoietic progenitors and HSCs emerge. In these mice, the YS can still autonomously generate both embryonic and definitive types of cells217. Similarly, mice lacking the Ncx1 22  sodium-calcium exchange pump do not develop a heartbeat and consequently also lack a circulation, but can survive until about E11. In these mice, hematopoietic cells appear in the YS and placenta but are not detected in the AGM218,219. However, this approach also has limitations as neither of these mouse models has been reported to support the generation of HSCs. This is likely because the mutations have effects other than disruption of the circulation, with functional consequences for HSC generation. For instance, HSCs have been shown to be the downstream progeny of VE-cadherin+ cells220 and some early hematopoietic cells in the embryo still express VE-cadherin (see section 1.5.4)221,222. In addition, shear stress induced by circulation has been reported to influence the production of hematopoietic cells, including HSCs, via a nitric oxide dependent pathway223,224. In fact, by exposing Ncx1-null AGM cells to shear stress stimulus, Adamo et al. could show that hematopoietic cells can be generated from this tissue. Therefore, while providing additional evidence that many embryonic sites are hemogenic, these models have not yet definitively determined which or how many sites normally produce HSCs in vivo. Another approach to determining when and/or where HSCs arise has involved lineage tracing experiments using Scl and Runx1 as the drivers of the tracing marker. These studies used the permanent activation of a reporter gene by an induced recombination event at the ubiquitously expressed Rosa26 locus. In these studies, the Cre-ER dependent recombination was controlled for tissue specificity using either the Scl or Runx1 promoter to drive a Cre-ER gene, and also temporally by the injection of tamoxifen at a specific time during development to translocate the Cre-ER to the nucleus allowing its activation. In this way, Gothert et al. marked cells expressing Scl from E10.5 to E11.5. They then examined the proportion of marked cells in the adults or in recipients of transplanted E14.5 fetal liver cells obtained from the “marked” embryos. As these proportions were similar it was concluded that very little de novo generation of HSCs occurred after E14.5225. Samokhvalov et al. then showed in a similar fashion that cells marked by Runx1 at E9.5 produced nearly all hematopoietic cells subsequently found in the adult226. Interestingly, this report also showed that by marking the cells at E7.5, a time when 23  Runx1 expression appears to be confined to the YS, a portion of adult hematopoietic cells were marked (1-10%). This finding suggests that at least some HSCs originate in the YS. While intriguing, these reports are the first of their kind, and are not entirely consistent. For example, the first group achieved only a maximum 10% marking of total hematopoietic cells, whereas the second group reported up to 100%, although the extent of marking was highly dependent on the timing of induction. Furthermore, the use of one allele of a critical hematopoietic gene to generate a knock-in reporter system (as in the Runx1 study) may have functional consequences for hematopoietic development, as previously demonstrated for Runx1 heterozygotes208,209. Future studies will thus be needed to clarify these issues. Nevertheless, what has been documented is that after a slow start, HSC generation and/or expansion increases very rapidly in the midgestation mouse embryo. Although the first HSCs can be detected at E10.5, they are very rare, and present at a frequency of ≤1 HSC per embryo172,227. By about E11, the number of HSCs per embryo is still only 1 or 2 with a relatively even contribution from the YS, AGM, and placenta. However, immediately thereafter, their numbers expand rapidly to reach ~100 per embryo between E12-12.5, mostly in the fetal liver and placenta, and ~1000 by birth, the vast majority of which are still present in the fetal liver. They begin to migrate to the bone marrow after E16 and this continues in the first weeks after birth172,181,228. Based on these numbers, it seems that HSCs must be generated de novo in relatively large numbers between E11 and E12.5, however, the exact contributions of generation versus expansion of HSCs to the total numbers of present during development remains largely unanswered. For human embryos, there is much less data, but it has been shown that both YS and AGM tissues isolated before the onset of circulation are able to give rise to hematopoietic cells. Since HSCs are not present at this point, their origin remains unclear. However, under the conditions tested, the AGM, but not YS, was able to give rise to B- and T-lymphoid cells as well  24  as myeloid cells. Therefore, it was hypothesized that HSCs primarily arise from intraembryonic sites229.  1.5.2 Generation of pre-hematopoietic mesoderm  Despite the controversy surrounding the specific sites and cell types from which hematopoietic cells arise in the YS, AGM, and placenta, it is widely accepted that they all share a common origin from the mesodermal germ layer. Mesoderm cells are generated in the early embryo from cells of the ICM in a series of coordinated steps (summarized in Figure 1.2). First a cell layer called the hypoblast forms along the outer edge of the ICM, next to the blastocoele cavity, and the inner cells organize to form a second cell layer called the epiblast. The hypoblast later contributes to extraembryonic tissues, whereas the embryo proper and extraembryonic mesoderm are derived from epiblast cells. The epiblast undergoes further differentiation and reorganization in a complex and highly regulated process called gastrulation which results in the formation of the three germ layers: ectoderm (outer), mesoderm (middle), and endoderm (inner). Mesoderm and endoderm cells invaginate and migrate through a region called the primitive streak where the definitive endoderm is formed by displacement of the hypoblast, and mesoderm cells insert between the remaining epiblast or ectoderm cells and the developing endoderm. The process of gastrulation and the factors that control it have been studied extensively (reviewed in Tam and Behringer230 and Tam and Loebel231). The primitive streak forms at the posterior end of the embryo and then progresses anteriorly to the midpoint of the embryo. The time and location at which cells traverse the primitive streak is related to the type of mesoderm produced (Figure 1.3); for example, the first cells to emerge from the most posterior end of the primitive streak become extraembryonic YS mesoderm that forms the YS blood islands. Cells emerging from the same region slightly later also contribute to the extraembryonic mesoderm, 25  Figure 2 Figure 1.2  Extraembryonic tissues  Morula  Trophectoderm  Inner cell mass  Hypoblast  Epiblast  Germ cells  Embryonic ectoderm  Skin, nervous system, etc.  Primitive streak  Extraembryonic mesoderm  Embryonic endoderm  Embryonic mesoderm  Intestine, lungs, liver, etc.  Hemangioblast Hemogenic endothelium  Hemogenic endothelium  Definitive hematopoietic  Definitive hematopoietic Hematopoietic CFCs (definitive)  Primitive hematopoietic  Muscle, bone, kidney, etc.  Hematopoietic stem cells  Figure 1.2 – Developmental steps from the morula to hematopoietic cells Summary of the lineage hierarchy in the embryo from the morula stage to the emergence of hematopoietic cells. Pluripotent cells are shown in white, extraembryonic tissues in orange, ectoderm in green, mesendoderm in blue, endoderm in purple, and mesodermal derivatives in pink.  26  Figure 3 Figure 1.3  Early streak (E6.5) Extraembryonic  Mid streak (E7.0)  Late streak (E7.5)  BMP4  Embryo  Primitive Streak Extraembryonic mesoderm  nodal  Lateral plate mesoderm Figure 1.3 – Emergence of pre-hematopoietic mesoderm from the primitive streak Schematic of gastrulation stage mouse embryos showing the extent of the primitive streak (blue) at early, mid and late streak stages. The cells ingressing through the most posterior and proximal region of the primitive streak (pink) contribute to extraembryonic mesoderm throughout gastrulation. This mesoderm will go on to generate the yolk sac mesoderm (including blood islands) and the placental mesoderm. Starting at the mid-streak stage, the cells ingressing through the midsection of the primitive streak (purple) contribute to the lateral plate mesoderm from which intraembryonic hematopoietic mesoderm (AGM) is subsequently derived. BMP4 is expressed from the extraembryonic endoderm and nodal is expressed from the node at the leading edge of the primitive streak, resulting in concentration gradients as indicated.  27  including the chorionic mesoderm that generates the hematopoietic tissue of the placenta. Lateral plate mesoderm, which forms the AGM region, emerges from the midsection of the primitive streak during the mid to late stages of gastrulation232. Analysis of the cells obtained from cultures of mouse primitive streak cells maintained in hematopoietic cell-supporting conditions also shows that the highest hematopoietic potential is associated with the posterior rather than the anterior half of the streak233. Such detailed tracking studies are not possible in human embryos, but it is thought that gastrulation proceeds similarly. There is also evidence that some early extraembryonic mesoderm might be derived from adjacent YS endoderm before the formation of the primitive streak, but it is likely that after gastrulation, epiblast-derived mesoderm also contributes156,168,234. Several factors have been identified as having important roles in mesoderm formation in vertebrates. These include members of the TGF-β, WNT, and FGF families (reviewed in Smith235, and Gadue et al.236). For example, WNT3 expression localizes to the site of primitive streak formation and embryos deficient for this gene do not form a primitive streak or detectable mesoderm237. A receptor for several FGF signalling molecules, FGFR1, also becomes localized to the region of primitive streak formation during gastrulation and Fgfr1-null embryos exhibit defects in the patterning and migration of mesodermal derivatives238. The TGF-β family member nodal is also important for the formation of the primitive streak, and is subsequently present at the highest concentration at the anterior end of the primitive streak239. BMP4 is from a different branch of the TGF-β family, and Bmp4-deficient embryos either fail to form mesoderm, or show poor proliferation of derivatives of the posterior primitive streak, resulting in little to no extraembryonic and lateral plate mesoderm240. Formation of the primitive streak can be rescued by expression of BMP4 from the extraembryonic ectoderm, but later defects are not rescued as expression is still lacking in the extraembryonic and lateral plate mesoderm241 (Figure 1.3). The same factors have also been shown to have important roles in the generation of mesodermal cells from pluripotent cells in vitro. Addition of BMP4 and either bFGF or activin A 28  (which activates the same pathway as nodal) can stimulate hematopoietic cell generation from pluripotent Xenopus animal cap ectoderm242. Using mouse ES cells, activin A, WNT3A and BMP4 have all been shown to play a role in the formation of primitive streak-like cells and mesoderm147,243-246. High levels of activin A promote mesodermal derivatives of the anterior primitive streak or endoderm and BMP4 promotes mesodermal derivatives of the posterior primitive streak147,244,246. Inhibition of FGF signalling has also been shown to block activin and BMP-induced mesoderm formation in this model, demonstrating that it also has a role in vitro 246  .  1.5.3 Evidence for a hemangioblast state  Early observations of the first appearance of hematopoietic cells in the YS blood islands, in close association in location and time with endothelial cells, led to the concept that these 2 cell lineages shared a common precursor. Initially these hypothetical cells were called “angioblasts” and then more descriptively named “hemangioblasts”247,248. Further support was provided by the finding that hematopoietic and endothelial progenitors share many common markers; such as CD34, CD31, TIE1, TIE2, FLT1, and FLK1249-251. For some, such as FLK1 and TIE2, it has been shown that their loss by gene targeting results in both endothelial and hematopoietic defects252-256. Later, a cell with this bipotent precursor activity was demonstrated functionally as a derivative produced by differentiating mouse ES cells. Specifically, Choi et al. identified a population of cells within very early EBs, prior to the emergence of hematopoietic cells that had “blast colony-forming cell” (BL-CFC) activity. When these colonies of blast cells were replated, some had the ability to form endothelial cells, as well as primitive and/or definitive hematopoietic cells257. The hemangioblast, or BL-CFC, expressed the mesoderm marker brachyury as well as the hematopoietic and endothelial markers FLK1 and SCL258,259, and could be distinguished from earlier mesoderm precursors that additionally showed cardiac 29  potential by the timing of their expression of brachyury259. Using this information, Huber et al. were then able to show that Brachyury+FLK1+ BL-CFCs were generated in gastrulation stage mouse embryos and were found only for a short time, concentrated within the primitive streak260. Ueno et al. further showed using chimeric mouse embryos expressing 3 different fluorescent reporters that these hemangioblasts form each blood island in a polyclonal manner, and that many hemangioblasts may have already differentiated prior to reaching the blood islands as separate endothelial- and hematopoietic-specific precursors261. The ES model was further exploited to show that many of the genes that affect the generation of hematopoietic cells in the embryo may also be involved at this earlier step. For example, putative hemangioblasts (BLCFCs) from Scl-null ES cells are unable to proliferate and form colonies like their wild type counterparts due to defects in the subsequent formation of hematopoietic and endothelial cells262. Gata2 also appears to have a role in the survival of BL-CFCs and/or the proliferation of cells downstream of the BL-CFC263.  1.5.4 Evidence for hemogenic cells with mature endothelial features  In contrast to the YS, in the embryo proper, endothelial vessels clearly exist prior to the appearance of hematopoietic cells, and certainly before the generation of HSCs. Clusters of hematopoietic cells are found along the ventral wall of the dorsal aorta, and in the vitelline and umbilical arteries254,264 and it has been reported that these clusters include the first HSCs to be found within the developing embryo265. Therefore, instead of a hemangioblast, a slightly different model has now been proposed for the generation of hematopoietic cells and HSCs in intraembryonic sites. This model invokes their origin from a hemogenic endothelium, i.e. cells that integrate into the vasculature and turn on overt signs of endothelial differentiation prior to switching to the generation of hematopoietic cells. Such hemogenic endothelial cells thus express mature endothelial cell markers like VE-cadherin and CD31, and can take up low 30  density lipoproteins (LDL)265-267. Putative hemogenic endothelial cells have been isolated from both mouse embryos and ES cell differentiation cultures within a VE-cadherin+CD45-Ter119population and their existence demonstrated in subsequent clonal cultures showing the production of endothelial and hematopoietic progeny from single cells267. Lineage tracing experiments, using recombination activated reporter expression mediated by VE-cadherin driven Cre-recombinase, have similarly shown that most or nearly all hematopoietic cells in the adult mouse, and therefore HSCs, are derived from cells that previously expressed VEcadherin210,220. Furthermore, deletion of Runx1 specifically in VE-cadherin endothelial cells is sufficient to block hematopoietic cell generation providing further evidence that HSCs are derived from endothelium and that Runx1 is critical for the endothelial to hematopoietic transition210. Very recently, live imaging techniques have been used to directly visualise the emergence of hematopoietic cells from mouse ES or embryo-derived endothelial cells in culture268 and in situ from the endothelium in zebrafish and mouse embryos269-272. The conservation of these developmental processes would predict that the generation of human hematopoietic cells in the embryo would be similar. Direct studies of human embryos to date, although less definitive, are consistent with this prediction. Intra-aortic clusters similar to those proposed to contain the first HSCs in mice are found in the dorsal aorta of human embryos between days 27 and 40 of gestation169. These clusters also express hematoendothelial genes, including CD34, SCL, GATA2, GATA3, KDR (also known as FLK1 or VEGFR2), and c-KIT 273. Recently, CD143 was also shown to mark hematopoietic cells including HSCs, the adjacent endothelium, and select mesenchymal cells underlying the emerging hemogenic endothelium in the human embryo274,275. Potential hemogenic endothelial cells expressing CD34 or CD31 but not CD45 have been isolated from human YS, AGM, and fetal liver and shown to develop into endothelial cells in endothelial culture conditions. Although they do not form hematopoietic cells directly, they can after further culture on MS-5 stromal cells276. In agreement with the lineage tracing studies done in mouse, it has been found that 31  some HSCs retain low levels of expression of VE-cadherin protein on their surface until the early fetal liver stage221. Taken together, these studies of how hematopoietic cells are formed during development suggest that there are remarkable similarities in the processes that take place in mice and humans. It would therefore be expected that the generation of hematopoietic cells in the embryo of both species would involve the sequential differentiation of a pluripotent cell through an epiblast state, followed by a mesoderm state, and then either a hemangioblast or hemogenic endothelial cell state prior to committing to a hematopoietic fate (Figure 1.2). Similarly, these would be the predicted minimal steps required for the generation of hematopoietic cells from ES or iPS cells in vitro.  1.6 Insights from human pluripotent cell models of hematopoietic development  Section 1.3 (above) summarized the general features of the 2 major types of protocols currently used to generate hematopoietic cells from either human (or mouse) ES and iPS cells. Despite some differences between these protocols (e.g. described in Figure 1.4), some common themes emerge. First, many conditions appear able to support at least a minimal level of hematopoietic differentiation from human ES cells. Second, although the exact timing varies with the protocol used, there is generally a window of many days’ duration during which cells expressing features of hematopoietic cells are detectable136,137,139,149,150,277. Third, progenitors of both mature erythroid and myeloid cells can usually be detected, and it has been reported that over the course of differentiation the proportions of each137,139,277 and the developmental stage of the erythroid cells produced (based on their hemoglobin expression pattern) changes, in agreement with previous mouse ES results, although the production of fetal versus adult hemoglobin is not consistent138,150,278,279.  32  Figure 1.4 – Comparison of methods for differentiation of human ES/iPS cells Representative protocols and results for the generation of hematopoietic cells from human ES cells using the 2 major types of methods (co-culture with stromal cells and EB culture). Methods compared are from the following studies: 1) Kaufman DS, Hanson ET, Lewis RL, Auerbach R, Thomson JA. Hematopoietic colonyforming cells derived from human embryonic stem cells. Proc Natl Acad Sci U S A. 2001;98:10716-10721.137 2) Vodyanik MA, Bork JA, Thomson JA, Slukvin, II. Human embryonic stem cell-derived CD34+ cells: efficient production in the coculture with OP9 stromal cells and analysis of lymphohematopoietic potential. Blood. 2005;105:617-626.136 3) Ledran MH, Krassowska A, Armstrong L, Dimmick I, Renstrom J, Lang R, Yung S, Santibanez-Coref M, Dzierzak E, Stojkovic M, Oostendorp RA, Forrester L, Lako M. Efficient hematopoietic differentiation of human embryonic stem cells on stromal cells derived from hematopoietic niches. Cell Stem Cell. 2008;3:85-98.139 4) Chadwick K, Wang L, Li L, Menendez P, Murdoch B, Rouleau A, Bhatia M. Cytokines and BMP-4 promote hematopoietic differentiation of human embryonic stem cells. Blood. 2003;102:906-915.143 5) Zambidis ET, Peault B, Park TS, Bunz F, Civin CI. Hematopoietic differentiation of human embryonic stem cells progresses through sequential hematoendothelial, primitive, and definitive stages resembling human yolk sac development. Blood. 2005;106:860-870.150 6) Pick M, Azzola L, Mossman A, Stanley EG, Elefanty AG. Differentiation of human embryonic stem cells in serum-free medium reveals distinct roles for bone morphogenetic protein 4, vascular endothelial growth factor, stem cell factor, and fibroblast growth factor 2 in hematopoiesis. Stem Cells. 2007;25:2206-2214.280 7) Kennedy M, D'Souza SL, Lynch-Kattman M, Schwantz S, Keller G. Development of the hemangioblast defines the onset of hematopoiesis in human ES cell differentiation cultures. Blood. 2007;109:2679-2687.149 The left-most boxes show the conditions used to maintain the human ES cells prior to induction of differentiation (specific cell lines used). Any pre-differentiation steps performed just before the transfer to differentiation conditions are shown to the right of these first boxes. Methods 4, 5, and 7 use clumps of ES cells for EB formation, whereas method 6 forces reaggregation of single cells. The right-hand side of the figure shows the timings of exposure to the various differentiation media used, and in the case of the co-culture methods, the stromal line(s) used. Pink boxes show the range of times when hematopoietic CFCs were detected and the approximate maximal CFC frequency reported in each study. All basal media were additionally supplemented with glutamine and a reducing agent, and methods 1, 3, and 4 also included the addition of non-essential amino acids. EB culture steps in method 7 were performed in 5% O2 conditions. *SFEM in method 5 is StemSpan serum-free expansion medium **CDM in method 6 consisted of IMDM/F12 supplemented with bovine serum albumin, synthetic lipids, insulin-transferin-selenium (ITS), protein-free hybridoma mix (PFHM) and ascorbic acid (Asc). Figure 4 Figure 1.4  33  Human ES maintenance  Co-culture  1  2  3  EB culture  4  5  6  7  Prediff  Day of Differentiation 1 0  5  15  20  25  S17 or C166 DMEM+20% FBS  MEF (H1, H9)  30/105 OP9 α-MEM+10% FBS  MEF (H1, H9)  300/105 AM20, UG26, EL08, primary AGM or FL DMEM+20% FBS  MEF (H1, H9, hESNCL1)  3000/105  Matrigel+MEFCM (H1, H9) MEF (H1)  10  EB in DMEM +20% FBS IMDM+ 20%FBS +bFGF  Trypsin adapted Low on MEF density (HES3) MEF  DMEM+20% FBS+BMP4+SCF+FLT3L+IL3+IL6+G-CSF  IMDM+1% methylcell +15% FBS+EX-CYTE +Asc+PFHM  240/105 SFEM*+15% FBS+EX-CYTE+Asc+ITS+PFHM 150/105 Adherent culture step in CDM**+ TPO+SCF  CDM**+ BMP4+VEGF+SCF+bFGF 500/105  StemPro+Asc+ Trypsin adapted StemPro+ StemPro+ MEF BMP4+bFGF+ on MEF Asc+BMP4+ deplete Asc+BMP4 VEGF (H1, HES2) bFGF 100/105  34  Many of the factors implicated in mesoderm development (BMP4, activin A, and FGF) have also emerged as important candidates of mesoderm differentiation from human pluripotent stem cells145,281-283. Addition of BMP4 has been reported to be important specifically for hematopoietic differentiation from human ES cells141,143,149,280. Several groups have now also identified human bipotent cells with endothelial and hematopoietic activity in human ES differentiation cultures. The Keller and Lanza groups showed that day 3-4 human EBs, much like early mouse EBs, contain BL-CFCs that generate colonies of blast-like cells that have the ability to differentiate into both endothelial and hematopoietic cells under appropriate conditions149,284. The Civin group showed hematopoietic and endothelial cells could arise from “mesodermal hematoendothelial colonies” generated from day 7-12 EB cells although these were not proven to be clonally derived150. The Bhatia group showed that single CD45-CD31+KDR+VE-cadherin+ cells can be isolated from day 10 EBs which upon further culture could, in some cases, give rise to both hematopoietic and endothelial cells285. However, it remains unclear exactly how these observations relate to one another, although it seems likely that they all reflect an intrinsically preferred biologic process for producing hematopoietic cells from mesoderm via either a hemangioblast or hemogenic endothelial intermediate. Another similarity between human and mouse ES cells is the difficulty in producing HSCs that are capable of robust engraftment in vivo. Thus far, a few groups have reported the production from human ES cells of hematopoietic cells with some ability to reconstitute hematopoiesis in transplanted recipients, including in some cases the ability to engraft secondary recipients139,155,194,286-289. However, the levels of engraftment obtained are usually low, and the best results have proven difficult to replicate. Attempts to induce HSC generation by overexpressing HOXB4 in human ES cell derivatives, as has been effectively done with mouse ES cells, have thus far not demonstrated increased engraftment194,288,289. Such manipulations  35  still hold enormous promise, and indeed, HOXB4 overexpression increases the in vitro production of CFCs from human ES cells, albeit to variable degrees194,288-290.  1.7 Thesis objectives  The generation of hematopoietic cells and hematopoietic stem cells during development is clearly a complex process with many questions about its nature, regulation, and diversity still to be addressed. It is clear that mouse ES cells have been a particularly powerful model for addressing some of these questions over the past 25 years. At the same time, key accomplishments such as the generation of HSCs in vitro remain elusive. Compared with the mouse system, the use of human pluripotent stem cells is in its infancy, but the potential uses of these cells and benefits to be derived from them are even greater. The study of human development has always been challenging due to the scarcity of material accessible for study. Thus comparisons of mouse and human ES cell differentiation will also be useful in identifying key differences in development between the 2 species. In addition, there are great hopes that human pluripotent stem cells will provide a new paradigm for testing pre-clinical drugs in a human model system, as well as playing a central role in regenerative medicine by serving as a source of cells for cell replacement therapies. The overall goal of this thesis was to contribute to the ultimate realization of these goals with regard to the blood-forming system. To this end a program was devised to obtain information that would facilitate greater understanding of how to make the process of producing hematopoietic cells more efficient and reproducible. The first area of focus was the starting cells for differentiation, the undifferentiated human ES/iPS cells. My hypothesis was that the efficiency and reproducibility of differentiation would be increased if the homogeneity of the starting undifferentiated cell cultures was optimized and made consistent between experiments. The approach taken in Chapter 2 was to test the use of methods available at the time those experiments were initiated to achieve this 36  improvement in the maintenance of undifferentiated cells, and assess the consequent effect of these changes on the efficiency of hematopoietic differentiation. The second area of focus was to determine a differentiation protocol that would best mimic the development of hematopoietic cells. Here my hypothesis was that the efficiency of hematopoietic differentiation would be increased by using protocols that mimicked developmental signals, and that this would be best achieved by incorporating and optimizing multiple steps of differentiation to provide these diverse signals to the cells. The further studies described in Chapter 2 were thus designed to evaluate the first generation of hematopoietic differentiation protocols that had been published, and build on these to optimize both early differentiation into mesoderm and late differentiation into hematopoietic progenitors (CFCs). The final area of focus was to analyze the clonality of hematopoietic cell differentiation events. My hypothesis in these studies was that the most robust protocols would allow efficient differentiation at both early and late steps, as well as homogeneous and polyclonal differentiation within a culture, with a large number of starting ES/iPS cells contributing to the final hematopoietic output. The initial experiments on bulk cells in Chapter 2 suggested the average hematopoietic output to be between 1 and 7 CFCs per EB. To evaluate how homogeneous this output was between groups of cells in a culture, the approach taken in Chapter 3 was to assess the output of CFCs and CFC subtypes in individual EBs. In the absence of rigorously defined intermediate cell types, in order to evaluate the efficiencies of early and late steps of differentiation in various protocols, this analysis methodology was also applied. The measure of the proportion of positive EBs was taken as an output of early differentiation efficiency, and the number of CFCs per positive EB was taken as an output of later differentiation efficiency. To further evaluate the number of clones contributing to hematopoietic differentiation, the approach taken in Chapter 4 was to develop a system with which to monitor and track the hematopoietic progeny from marked ES/iPS cells. To achieve this, I determined a method for 37  marking undifferentiated or differentiating ES/iPS cells, and validated that reporter gene expression was maintained in hematopoietic progeny. Since each EB already represents different groups of ES/iPS clones, this system was then used to determine the clonality of the hematopoietic output within individual EBs.  38  2. Improved output of hematopoietic cells from human pluripotent stem cells by separate optimization of different developmental steps  2.1 Introduction  During development, hematopoietic cells arise through a coordinated series of developmental steps from the earliest totipotent cells that constitute the blastula. As delineated in Chapter 1, these steps support the sequential generation from the blastula of the inner cell mass, the epiblast, and the three major germ layers, including the mesoderm from which blood cells eventually emerge. Studies in the murine system have demonstrated that many of the same events that lead to the genesis of hematopoietic cells during the formation of the embryo in vivo can be recapitulated when murine ES or iPS cells are cultured under conditions that support their development into hematopoietic cells in vitro. More recently, additional support has been accumulating to indicate a mirroring of in vivo events is obtained during the in vitro genesis of hematopoietic cells from human ES and iPS cells. The studies described in this Chapter were initiated when this concept was first being developed. Accordingly, experiments were designed to test the assumption that optimization of each step would best be achieved by starting with as homogeneous an undifferentiated ES or iPS cell population as possible. This was compared for the two major differentiation methods that had been previously reported: co-culture with stromal cells and EB culture136,137,143. For the co-culture protocols, we concentrated on the use of OP9 cells as a feeder layer, as these had just been reported to support the generation of much higher frequencies of hematopoietic cells, over a shorter period of differentiation than other cell lines that had been tested136. Attention was then focussed on improving the output of mesoderm as measured by transcriptional changes and the use of a MIXL1 human ES reporter cell line291. In the mouse  39  embryo, MixL1 is expressed mesendodermal cells of the primitive streak and early nascent mesoderm that will go on to form multiple tissues292-294. This human reporter line was used by Davis et al. to show that MIXL1 marked most precursors of hematopoietic cells in their differentiation protocol291. A medium conditioned by the human hepatocarcinoma cell line HepG2 (called MEDII) was investigated as a candidate inducer of mesoderm based on evidence of its role in very early mouse ES cell differentiation. When exposed to MEDII, mouse ES cells are induced to form primitive ectoderm-like (epiblast) cells that can be maintained in this state, but also convert back to an ES state when transferred back into ES conditions295. When subsequently removed from MEDII and transferred into EB conditions, these cells generated mesodermal derivatives more rapidly and in increased amounts when compared to EBs produced from untreated ES cells296. Interestingly, human ES cells briefly exposed to MEDII medium reportedly show gene expression changes indicative of activation of the TGFβ/NODAL signalling pathway which is important for mesoderm formation297. Work from our laboratory then showed that MEDII conditioned medium selectively increased the expression of mesoderm markers in human ES/iPS cell populations and that BMP4 and insulin-like growth factor 2 (IGF2) were candidate factors in MEDII, based on the relatively high expression of their mRNA in HepG2 cells. Importantly, exposure of human ES and iPS cells to MEDII increased the frequency of hematopoietic cells in subsequent EBs or OP9 co-cultures, including the number of cells that could be detected in vivo when MEDII-pretreated EB cells were injected into immunodeficient mice155. Improvement in the generation, expansion, and differentiation of hematopoietic cells from mesodermal precursors was also an area of focus, where hematopoietic cytokines were used to increase the output of hematopoietic CFCs in vitro. The specific combination of SCF, FLT3 ligand, IL-3, IL-6, and granulocyte colony stimulating factor (G-CSF) was chosen for investigation as this combination of cytokines had previously been shown to be permissive for the maintenance of very primitive adult and neonatal human hematopoietic cells131,132,298. This 40  combination had also been recently reported to increase the number of CFCs produced from human ES cells during EB differentiation143. The ultimate goal of these experiments was to determine an overall set of conditions that would lead to the most effective and reproducible protocol for generating hematopoietic cells from human ES/iPS cells in our lab. Here I tested if this was best achieved by optimizing multiple, unique steps in the hypothesized developmental pathway from pluripotent to hematopoietic cells using the specific manipulations described above.  2.2 Materials and methods  2.2.1 Human ES and iPS cell lines  H1 and H9 human ES cell lines were purchased from WICELL (Madison, WI, USA). CA1 and CA2 human ES cell lines89 were provided by A. Nagy (Samuel Lunenfeld Research Institute, Toronto, ON, Canada). The MSC-iPS1 (MSC) human iPS cell line was a gift from G. Daley (Harvard University, Boston, MA, USA) and was generated from adult bone marrow mesenchymal stem cells infected with retroviruses containing OCT4, SOX2, KLF4, and c-MYC 39  . HES3-MIXL1GFP/w human ES cells were provided by A. Elefanty (Monash University,  Melbourne, Australia) and were generated by targeting GFP to one of the endogenous MIXL1 loci in HES3 cells291. Approval for the experimental use of all of these cells was obtained from the Canadian Stem Cell Oversight Committee, and the Research Ethics Board of the University of British Columbia.  41  2.2.2 Maintenance culture of human ES and iPS cells  Human ES and iPS cells were propagated as undifferentiated cells by continuous passage onto either irradiated feeders of CF1 MEFs (STEMCELL Technologies, Vancouver, BC, Canada) or Matrigel (Becton Dickson, San Jose, CA, USA)-coated tissue culture dishes. Cells were cultured on MEFs in human ES (HES) medium which was made by supplementing a mixture of 50% Dulbecco’s Modified Essential Medium (DMEM) and 50% F12 (STEMCELL Technologies) with 20% knockout serum replacement (KOSR; Invitrogen, Carlsbad, CA, USA), 1 mM L-glutamine, 0.1 mM non-essential amino acids, and 0.1 mM β-mercaptoethanol, plus 4 ng/mL of bFGF (all from STEMCELL Technologies). MEFs were cultured in gelatin-coated tissue culture flasks in high glucose DMEM with 10% FBS (cat. #6902) and 1 mM L-glutamine (all from STEMCELL Technologies), and had been maintained for no more than 5 passages when used for ES or iPS cultures. MEFs were irradiated with 30 Gy X-rays and plated onto gelatin-coated tissue culture dishes at a density of 15,000 cells/cm2 prior to addition of ES or iPS cells. For feeder-free culture on Matrigel, cells were maintained in mTeSR medium (STEMCELL Technologies), or in some experiments, HES medium conditioned by MEFs and supplemented with 8 ng/mL bFGF. Matrigel was first thawed on ice and diluted to 1 mg/12 mL with cold DMEM prior to coating dishes for 1 hour at room temperature or overnight at 4°C. Human ES and iPS cultures were split every 7 days, and morphologically differentiated cells were identified, scraped off and removed from the dish prior to enzyme treatment (selective passaging). Cultures on MEFs were treated with 1 mg/mL collagenase IV (Invitrogen) at 37°C for 20 minutes and clumps of cells removed by scraping with a pipette. Feeder-free cultures were instead treated with 1 mg/mL dispase (STEMCELL Technologies) at 37°C for 3-5 minutes, until the edges of the colonies had just begun to loosen from the dish. Dispase was removed and the plate washed twice with phosphate-buffered saline (PBS) before clumps were removed from the dish in DMEM with a cell scraper. Clumps were then centrifuged (at 300 g), 42  resuspended in fresh maintenance medium, triturated, and clumps with a minimum diameter of 60 μm were counted and seeded into new cultures at a density of 30-35 clumps/cm2. Cells were fed daily with fresh medium, except for on weekends when a double volume of medium was added for 3 days. All cultures were maintained in a 37°C incubator supplied with an atmosphere of 5% CO2 in air.  2.2.3 ES-CFC assays  Single cell suspensions were generated by incubation of cells with TrypLE (Invitrogen) for 7-10 minutes at 37°C. Cells were then removed from the dish, filtered through a 40 μm cell strainer, centrifuged (at 300 g), and resuspended in maintenance medium. Cells were plated into either MEF or feeder-free culture conditions at a density of 103-104 cells/cm2. Cultures were maintained for 7 days with no media changes, and then fixed and stained using an AP detection kit (Sigma, St. Louis, MO, USA). Colonies containing ≥30 AP+ cells were then enumerated. In some cases, cells were treated with 10 μm Rock inhibitor (Y-27632; Calbiochem, Gibbstown, NJ, USA) in maintenance medium for 1 hour prior to TrypLE treatment and then plated at a density of 102-103 cells/cm2 with 10 μm Rock inhibitor added for the first 24 hours only.  2.2.4 MEDII treatment  HepG2 human hepatocarcinoma cells were grown in high glucose DMEM with 10% FBS (cat. #6902) and 1mM L-glutamine (all from STEMCELL Technologies). Conditioned medium was prepared by seeding 8.75x106 HepG2 cells per T-175 tissue culture flask, then on day 1 washing at 37°C twice with PBS for 10 minutes, followed by DMEM/F12 for 2 hours, prior to incubation at 37°C for 3-4 days in 41 mL of HES medium/flask. The harvested conditioned medium (MEDII) was then filtered using a low protein binding 0.22 μm filter (Millipore, Billerica, 43  MA, USA), pooled and stored at -20°C until used. Prior to use, the MEDII was diluted in an equal volume of fresh HES medium. Undifferentiated ES or iPS cells exposed to this 50% MEDII prior to being seeded into OP9 co-cultures were maintained in parental cultures for 7 days then incubated with 50% MEDII for 5 days, with one media change on the 3rd day of MEDII treatment. For subsequent transfer to EB cultures, clumps of undifferentiated cells were obtained using the normal passaging protocol and were then seeded onto Matrigel-coated dishes in 50% MEDII at approximately 100 clumps/cm2. These MEDII cultures received fresh 50% MEDII medium on the same schedule as standard undifferentiated cell cultures, and were used for EB formation after 7 to 10 days.  2.2.5 OP9 co-cultures  OP9 cells (RIKEN, Tokyo, Japan) were maintained on gelatinized tissue culture dishes in Alpha Minimum Essential Medium (α-MEM) with 20% FBS (cat. #6100) and 1mM L-glutamine (all from STEMCELL Technologies), and split with trypsin-EDTA (Invitrogen) as soon as the cells became confluent. For co-culture experiments, OP9 cells were either irradiated with 80 Gy X-rays and plated onto gelatinized dishes at 25,000 cells/cm2, or allowed to grow to confluence and used 3-4 days later without irradiation. Cells to be tested in OP9 co-cultures were harvested as clumps (same protocol as for passaging undifferentiated cells), but were then suspended in a differentiation medium consisting of α-MEM or Iscove’s Modified Dulbecco’s Medium (IMDM) with 10% FBS (cat. #6100), 1 mM L-glutamine, and 0.05 mM β-mercaptoethanol (all from STEMCELL Technologies), and plated onto dishes containing a pre-formed layer of OP9 cells. After 24 hours, the differentiation medium was replaced with fresh medium of the same composition. Subsequently, half medium exchanges were performed every 3-4 days. After 7-14 days in these  44  conditions, single cells were harvested using trypsin/EDTA (Invitrogen) or TrypLE, filtered through a 40 μm cell strainer, and subjected to further analyses as described.  2.2.6 EB cultures  Cells were used for EB formation when ready for passaging or within the next 1-3 days. Cells were first incubated with 0.5 mg/mL dispase (STEMCELL Technologies) containing 0.5 mM CaCl2 (Sigma) for 20-30 minutes at 37°C until the edges of the colonies started to round up. Dispase was then carefully removed, and large clumps of cells were obtained by adding fresh DMEM and collected in a 15 mL conical tube. When the cells were difficult to remove using this method, a cell scraper was used. In some of the early experiments clumps were obtained using collagenase IV or 0.05% trypsin/EDTA with 0.5 mM CaCl2 instead of dispase. After allowing the harvested clumps of cells to settle for 1-2 minutes, they were centrifuged at low force (20 g) for 2-3 minutes. Clumps were then very gently resuspended in a medium consisting of IMDM with 20% FBS (cat. #6100), 1 mM L-glutamine, 0.1 mM non-essential amino acids, and 0.1 mM βmercaptoethanol (EB medium, all components from STEMCELL Technologies); and finally plated into ultra-low adherence 6-well plates (Corning Life Sciences, Acton, MA, USA) at high density – e.g. cells from two 35 mm wells or one 60 mm dish per single 35 mm well for EB culture. In some experiments, 10 μM Rock inhibitor was added to the media of the cells at 37°C for 1 hour prior to dispase treatment and then 10 μM Rock inhibitor was also added to the initial EB medium. To change the medium, EBs in their medium were removed from the wells in which they were being cultured and the entire volume placed into a 15 mL conical tube and the EBs allowed to settle by gravity (~5 minutes). During this time, 1 mL of fresh medium was added into each well. If EBs were very small, they were centrifuged at low force (20 g) for 2-3 minutes. The supernatant was then removed except for 0.5-1 mL, and fresh medium added to gently 45  resuspend the EBs which were then returned to their original wells in a total volume of ~4 mL. The medium was always changed within the first 2-4 days after initiation of the EB cultures, and then every 4-7 days thereafter. In some experiments, recombinant human hematopoietic cytokines were added to the EB cultures as follows: 100ng/mL SCF (R&D Systems, Minneapolis, MN, USA), 100ng/mL FLT3 ligand (Amgen, Thousand Oaks, CA, USA), 20ng/mL IL-3 (Novartis, Basel, Switzerland), 20ng/mL IL-6 (Cangene, Winnipeg, MB, Canada), 20ng/mL G-CSF (Amgen). This combination is referred to as “+GF” or “+ cytokines” throughout the text and any other cytokines that were used are indicated in the appropriate section. To analyze cells from EBs, the EBs were centrifuged, resuspended in TrypLE and incubated at 37°C for 15 minutes with vigorous trituration every 5 minutes. For colony assays, cells were additionally filtered through a 40 μm cell strainer prior to being subjected to further analyses.  2.2.7 Hematopoietic CFC assays  Cells to be tested were resuspended at an appropriate concentration in either IMDM or EB medium and 3x the final cell number per dish was added in a volume of 300 μL to 3 mL of Methocult 4230 (STEMCELL Technologies) supplemented with 50 ng/mL human SCF, 20 ng/mL human granulocyte/macrophage colony stimulating factor (GM-CSF; Novartis), 20 ng/mL IL-3, 20 ng/mL IL-6, 20 ng/mL G-CSF and 3 U/mL erythropoietin (EPO; STEMCELL Technologies), and then vortexed. The mixture was distributed in 1.1 mL volumes per dish into 2x 35 mm Greiner dishes (STEMCELL Technologies) using a 3 cc syringe and a 16 gauge blunt end needle (STEMCELL Technologies). When possible, 2 different cell concentrations were plated to improve the range of detection of the assay. For EB cells, the number of cells plated was between 1 and 15x104 cells per dish, and for cells co-cultured with OP9 cells, the range 46  was from 1-10x105 cells per dish. Dishes were incubated in a humidified atmosphere of 5% CO2 in air at 37°C for 2 weeks and then colonies were counted and scored as erythroid (from CFU-E or BFU-E, combined and labelled in analyses as CFU-E), granulocyte/macrophage (from CFUGM, CFU-G or CFU-M, combined and labelled in analyses as CFU-GM), or mixed (from CFUGEMM).  2.2.8 Flow cytometry  Single cells from EB or OP9 co-cultures were resuspended in Hanks Balanced Salt Solution (HBSS) with 10% human serum (both from STEMCELL Technologies) at a concentration of 106 cells/mL. Antibody staining for 30 minutes was performed in 100 μL volumes on ice in the dark, using the following antibodies: CD34-allophycocyanin (APC) diluted 400x; CD45-phycoerythrin (PE) or CD45-APC diluted 100x; CD31-fluorescein-isothiocyanate (FITC) diluted 100x; CD15-FITC diluted 50x; CD66b-FITC diluted 50x; CD41a-PE diluted 50x (all from BD); Glycophorin A-cyanine (Cy)5 diluted 100x (from P. Lansdorp, Terry Fox Laboratory, BC Cancer Agency, Vancouver, BC, Canada); or mouse IgG1 isotype controls for FITC, PE, and APC (BD) diluted 100x. Cells were then washed with 1mL of HBSS with 2% FBS, centrifuged (at 300 g), and resuspended in HBSS with 2% FBS and 1 μL/mL propidium iodide (PI; Sigma). Cells were analyzed using a FACSCalibur (BD) and FlowJo software (Tree Star, Ashland, OR, USA). Positive events were determined by setting gates that excluded at least 99.5% of the cells stained only with the corresponding isotype controls.  2.2.9 Quantitative RT-PCR analysis  RNA was extracted and purified using Absolutely RNA kits (Stratagene, Cedar Creek, TX, USA) and quantified using a Nanodrop ND-1000 instrument (Thermo Scientific, Waltham, 47  MA, USA). Between 0.5 and 1 μg of total RNA was reverse transcribed into cDNA using either the Superscript II or III First-Strand Synthesis System for RT-PCR (Invitrogen). Q-RT-PCR analyses were performed with either Power or Fast SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA, USA), using GAPDH as an endogenous control. Each sample was assigned a δ-ct value with respect to the level of GAPDH transcripts, and then a δδ-ct value with respect to the undifferentiated cell value in that experiment. From that, the expression relative to undifferentiated cells was calculated using the formula: expression=2-(δδ-ct). Q-PCR primers were selected using either Primer 3 or the Universal Probe Library Assay Design Guide (Roche Applied Science, Mannheim, Germany) and are listed in Table 2.1.  2.2.10 Statistical analyses  Statistical analyses were performed using GraphPad Prism Version 5.03 (GraphPad Software, La Jolla, CA, USA). Due to the >100-fold variation in CFC frequencies observed in different experiments, all CFC frequencies were log transformed prior to calculating means, errors, and assessment of differences between groups using the Students t-test or ANOVA analyses. When a particular type of CFC was not detectable within an otherwise positive experiment, the CFC frequency value assigned to that specific CFC type was that corresponding to the limit of detection (= the frequency that would have been derived if 1 colony of that type had been detected in that experiment). This strategy results in an overestimate of the average CFC frequency where many of the contributing values are at the limit of detection, as was the case for CFU-GEMMs but not other CFC types. Accordingly, further statistical analyses were not performed to compare CFU-GEMM outputs in different groups.  48  Table 2.1 – Primers used for Q-PCR analysis of RNA transcript levels Gene name GAPDH BRACHYURY MIXL1 HAND1 IGF2 GATA2 CD34 KDR CDX4 SCL CD43 HOXB4  Forward Primer CCCATCACCATCTTCCAGGAG CCCTATGCTCATCGGAACAA GGTACCCCGACATCCACTT AACTCAAGAAGGCGGATGG GCCAAGAAGGTGAGAAGCAC AACGCCTGTGGCCTCTACTA CTTGGGCATCACTGGCTATT GAACATTTGGGAAATCTCTTGC CAGTTAACCTGGGCCTTTCC GGATGCCTTCCCTATGTTCA AGTGCTGCGTCCTTATCAGC CTGGATGCGCAAAGTTCAC  Reverse Primer CTTCTCCATGGTGGTGAAGACG CAATTGTCATGGGATTGCAG CGCCTGTTCTGGAACCATAC CGGTGCGTCCTTTAATCCT CTGTTCGGTTTGCGACAC TGTTGGACATCTTCCGGTTC TCCACCGTTTTCCGTGTAAT CGGAAGAACAATGTAGTCTTTGC GGGAGATTTTCTTTTTGATCATCT AAGATACGCCGCACAACTTT ACCCCAAGGAGAAGGAGAAG AGCGGTTGTAGTGAAATTCCTT  Table 1 Table 2.1  49  2.3 Results  2.3.1 Variation in hematopoietic CFC output using alternate methods of ES or iPS cell differentiation induction  The first reports of the successful generation of hematopoietic cells from human ES cells used either EB formation or co-culture with stromal cell lines136,137,143. To determine the hematopoietic output of human ES cells in our lab, several lines (CA2, CA1, H1, and H9) and methods of differentiation induction (EB culture or co-culture with OP9 feeders in the absence of cytokines) were used. First, I assessed the robustness of the cell line or protocol used in terms of the proportion of experiments where detectable numbers of hematopoietic CFCs were obtained. Amongst those that were positive, I then compared the CFC frequencies per 105 cells in the final populations obtained. This analysis (see Figure 2.1) showed that, on average, the generation of detectable numbers of CFCs was achieved approximately half of the time (40-77%) for all cell lines regardless of the protocol used, when ≥5 experiments were performed. However, for every cell line tested, EBs gave about a 10-fold greater final CFC frequency (p<0.05) compared to the OP9 method. Some variation in final CFC frequency was also observed between different cell lines, but this variation was not as significant as the variation between the 2 protocols. Note that for one iPS line (MSC) obtained later and included in this comparison, this trend was exactly the same as for the ES lines, although the final CFC frequencies obtained in the positive cultures were, on average, slightly lower (Figure 2.1). The hematopoietic colonies generated from all lines were grossly indistinguishable from those generated from CFCs obtained directly from human hematopoietic tissues (Figure 2.2a). To further characterize the cells produced in the colonies generated from ES/iPS cell-derived hematopoietic CFCs, flow cytometry was used to examine their expression of cell surface 50  Figure 5 Figure 2.1  A  CA2  CA1  H1  H9  MSC  EB  5/8 (63)  34/44 (77)  61/80 (76)  2/5 (40)  19/30 (63)  OP9  0/3 (0)  13/17 (76)  40/63 (63)  6/15 (40)  3/6 (50)  B  iPS  ES  CFC/105 cells  1000  ** *** ***  100  *  EB OP9  10 1 0.1  nd  CA2 CA1  H1  H9  MSC  Cell line Figure 2.1 – Comparison of hematopoietic outputs of human ES and iPS cells subjected to EB and OP9 differentiation protocols A) Summary of the proportion of experiments where hematopoietic CFCs were detected when undifferentiated human ES (CA2, CA1, H1, H9) or iPS (MSC) cells were subjected directly to EB formation or OP9 co-culture and differentiated in the absence of cytokines. The number of experiments where CFCs were detected/total number of experiments (% positive) is shown for each cell line in each type of differentiation. The usual starting number of ES or iPS cells was >5x105 (OP9) or >1x106 (EB). B) Hematopoietic CFC frequency per 105 final (differentiated) cells recovered for each cell line using the indicated method of differentiation. CFC frequencies were log transformed and the values were used to calculate the geometric mean (±SEM) for the experiments where at least one CFC was detected in the final suspension. As indicated in (A), no CFC were detected in any of the experiments where CA2 cells were co-cultured with OP9 cells (nd=not detected). A significant difference between the CFC frequency obtained from EB and OP9 differentiation of the other 4 lines was determined by 2-tailed t-test of the log transformed values, * p<0.05, ** p<0.01, *** p<0.001.  51  Figure 6 Figure 2.2  i) H1  MSC  iii) CA1  MSC  iv) H1  MSC  MSC  CD41-PE  CD15/66b-FITC  B  ii) H1  GlyA-Cy5  A  CD45-APC  CD45-APC  FSC  C  E  E  M G  Figure 2.2 – Characteristics of hematopoietic cells derived from human ES/iPS cells A) Representative colonies shown from i) CFU-GEMM, ii) large CFU-GM, iii) BFU-E, and iv) small CFU-GM. For each colony shown, the starting ES and iPS cell lines from which the CFCs were derived are as indicated. B) Flow cytometric analysis of pooled human ES cell-derived hematopoietic colonies. Cells were recovered from methylcellulose dishes containing multiple colony types, and stained with antibodies recognizing the markers shown. Representative plots are shown for markers of granulocytes (CD15 and/or CD66b, CA1 cells), megakaryocytes (CD41, CA1 cells), and erythroblasts (glycophorin A, H1 cells). C) Cytospin analysis of individual human H1 ES cell-derived hematopoietic colonies. Colonies were picked from methylcellulose dishes, spun onto slides and Wright-Geimsa stained. Shown are representative images from a mixed (GM + E) colony (left) and a GM colony (right). 52  markers expected to characterize mature blood cells. For this analysis, multiple colonies were pooled, stained with appropriate antibodies, and then analyzed. Most of the cells were found to express the pan-hematopoietic marker CD45, in combination with hematopoietic lineage markers such as CD15 or CD66b (monocytes or granulocytes) or CD41a (megakaryocytes). In addition, from plates containing erythroid colonies, some smaller cells expressing glycophorin-A (erythroblasts) could also be detected (Figure 2.2b). Individual colonies of specific types were also picked, spun onto slides, and stained with Wright-Geimsa stain. This confirmed that they contained blood cells with the appropriate differentiated morphology (Figure 2.2c). These data suggest that most pluripotent cell lines are able to generate hematopoietic cells that can undergo terminal differentiation along the major myeloid lineages. However, the large variability we documented between individual experiments makes the use of this approach problematic. One major source of variation was determined to be the protocol used for differentiation, with EBs showing a greater hematopoietic CFC frequency than co-culture with OP9 cells, using the reagents and methods adopted here. Therefore, in further studies designed to analyze and optimize the steps required to ensure more consistent and higher final CFC frequencies, we made most frequent use of the EB protocol.  2.3.2 Highly undifferentiated human ES/iPS cells are poorly responsive to traditional hematopoietic differentiation induction protocols  One potential source of variation was hypothesized to be the heterogeneity of the starting “undifferentiated” cells. It is well known that under sub-optimal maintenance conditions, varying degrees of “spontaneous” differentiation occurs. To minimize this potential variable, several strategies were evaluated that increased the homogeneity of the starting cultures. These included the use of the ES-CFC assay to determine the optimal parameters for the passaging interval and the density of clumps plated during ES cell passaging by looking for a 53  combination of conditions yielding the highest ES-CFC frequencies, indicative of the purest populations of undifferentiated cells (Figure 2.3). Also included was the use of a selective passaging protocol in which areas of morphologically differentiated ES cells were removed and discarded prior to each passage, as well as the culture of the ES cells on Matrigel in defined mTeSR medium16,299. Homogeneity of undifferentiated cells in the starting cultures was assessed using the following criteria: (i) visual assessment of the proportion and size of growth areas with distinct borders also containing cells homogeneous in size and appearance with a high nuclear to cytoplasmic ratio; (ii) flow cytometric assessment of the proportion of SSEA3+ cells; and (iii) frequency of ES-CFC10. To determine the effect of these changes in the protocols used to maintain ES cells in an undifferentiated state on their subsequent ability to generate hematopoietic CFCs, the cells were subjected to the same OP9 or EB differentiation protocols. In the first set of experiments, H1 cells that were selectively passaged on MEFs were assessed for hematopoietic CFCs after co-culture with OP9 cells and compared with OP9 co-cultures where H1 cells had not been selectively passaged. Unexpectedly, we found that selectively passaging the starting (undifferentiated) ES cells resulted in both a decreased proportion of positive experiments (where some CFCs could be detected) and a decreased frequency of CFCs in the final cultures that were positive (Figure 2.4a and c). This result could have been explained by postulating either that the selective passaging removed partially differentiated cells that had an increased ability to respond to stimuli derived from OP9 cells, or that selective passaging had retained preferentially a subset of human ES cells incapable of responding to these differentiation signals. To distinguish between these 2 possibilities, H1 cells were selectively passaged, but then first exposed to FBS to try to initiate their differentiation prior to transfer to OP9 cells. This experiment showed that the lost capacity of selectively passaged ES cells to generate hematopoietic CFCs when co-cultured with OP9 feeders was restored by a preliminary period of exposure to factors present in FBS (Figure 2.4a and c). These findings favour the idea that 54  ES-CFC /103 cells  Figure 7 Figure 2.3  1.4  5  H1  4  1.0  H9  20 clumps 50 clumps 100 clumps (plated/cm2)  3  0.6  2  0.2  1 0  2  4  6  8 10  0  2  4  Days after passaging  6  8  10  Figure 2.3 – Optimization of passaging ES cell density and interval A representative experiment is shown where clumps of human ES cells (H1 and H9) were plated at the indicated densities onto MEFs and harvested at day 3, 7, or 10 after reseeding. Single cell suspensions from each of these conditions were assayed to determine the frequency of ES-CFC present to identify the best protocol parameters (number of clumps seeded per dish and passaging interval) to maintain maximum frequencies of undifferentiated cells.  55  Figure 2.4 – Changes in hematopoietic CFC output when the representation of undifferentiated cells in the starting ES/iPS cell population is increased A) CFC frequencies obtained in OP9 co-cultures initiated with H1 cells maintained under the different conditions indicated. Each dot represents a different experiment in which ≥1 CFC was detected. Log transformed values for the CFC frequencies obtained were used to calculate the geometric mean (±SEM). To determine significant differences in average CFC frequency, log transformed values were analyzed using 1-way ANOVA with Tukey’s Multiple Comparison post-test, ** p<0.01. B) CFC frequencies obtained in EB cultures initiated with ES (H1, CA1, or CA2) or iPS (MSC) cells maintained as indicated: MEF, culture on mouse embryonic fibroblasts in HES media; CM, culture on Matrigel with HES medium conditioned by MEFs; TeSR, culture on Matrigel with mTeSR. Each dot represents a different experiment in which ≥1 CFC was detected. Statistical analyses were performed as in (A), *** p<0.001. C) Summary of the proportion of experiments in which ≥1 hematopoietic CFC was detected in the above experiments. The values shown are the number of experiments in which ≥1 CFC was detected/total number of experiments (% positive). Differences in the proportions of positive and negative experiments between the groups were determined using Fisher’s exact test, * p<0.05, *** p<0.001. D) CFC frequencies for each subtype of CFC, for the experiments shown in (B). Bars represent the geometric mean (±SEM) which was calculated as for the total CFC frequency, except that the frequency representing the detection limit of the experiment was used where a specific CFC subtype was not detected within a positive experiment. Figure 8 Figure 2.4  56  A  OP9 differentiation  B  EB differentiation ***  **  1000  CFC/105 cells  CFC/105 cells  100  10  1  100 10 1 0.1 0.01  0.1 Selected  OP9 differentiation Non- Selected Selected selected +FBS 5/10 (50)  CFC/105 cells  8/8 (100)  6/7 (86)  CFU-E  EB differentiation MEF  CM  TeSR  52/55 (95)  17/33 (52)  52/90 (58)  CFU-GEMM  CFU-GM  100  100  100  10  10  10  1  1  1  0.1  MEF  CM  TeSR  ***  *  C  CM  Maintenance of undifferentiated cells  Maintenance of undifferentiated cells  D  MEF  NonSelected selected +FBS  TeSR  0.1  MEF  CM  TeSR  0.1  MEF  CM  TeSR  57  increasing the homogeneity of undifferentiated human ES cells alters the response to OP9 differentiation conditions by removing an intermediate subset of OP9-responsive cells. Interestingly, use of the more robust EB differentiation method revealed no difference in the CFC frequencies obtained when selectively passaged ES cells were used as the input population as compared to non-selectively passaged ES cells (data not shown). When the CFC frequency obtained from ES/iPS cells maintained on MEFs were compared with those for ES/iPS cells maintained without MEFs in either MEF-conditioned medium (CM) or mTeSR, a greater proportion of positive experiments (with hematopoietic CFCs) were obtained from the MEF co-cultured ES/iPS cells (Figure 2.4c). In addition, the CFC frequency in positive EB cultures derived from MEF co-cultured ES/iPS cells was also significantly higher when compared to EB cultures derived from ES/iPS cells cultured in mTeSR (p<0.05, Figure 2.4b). This effect on CFC frequencies was true for all CFC subtypes (i.e., CFU-E, CFU-GM, and CFUGEMM (Figure 2.4d)). Taken together, these results show that highly undifferentiated ES cells are poorly responsive to conditions traditionally used to induce hematopoietic cell differentiation and entry into some as yet uncharacterized intermediate state is required. This, in turn, suggested the need to develop more controlled protocols for ensuring the production of such cells as a preliminary step to initiating the production of CFCs. Since introducing a FBS pretreatment step improved the capacity of selectively passaged cells to differentiate on OP9 cells, I hypothesized that EB-induced differentiation from highly undifferentiated ES cells maintained in MEF-free conditions could also be improved by introducing a suitable pretreatment step. Treatment with MEDII conditioned medium was chosen for this, as we were concurrently doing experiments to show it is a candidate inducer of differentiation to mesoderm155 (also see Section 2.1).  58  2.3.3 Pretreatment of undifferentiated ES/iPS cells with MEDII conditioned medium increases the output of hematopoietic CFCs but does not alter the kinetics of their generation  Work by our group had shown that MEDII had the desired “pre-conditioning” effect on ES and iPS cells when compared with untreated cells co-cultured with MEFs155. I undertook additional experiments to more stringently test whether this effect would be consistent when applied and compared to highly undifferentiated ES and iPS cell populations that were continuously cultured in MEF-free conditions with mTeSR medium as described in Section 2.2.2. Clumps of cells maintained in these conditions were harvested and plated on Matrigel in either 50% MEDII or mTeSR for 7-11 days prior to EB formation. At the time of EB setup, cells cultured in the absence of MEDII maintained a mostly undifferentiated morphology as expected. In contrast, the cells cultured with MEDII appeared to contain few, if any, morphologically undifferentiated cells. At the end of the EB culture, the frequencies of CFU-E, -GM and -GEMM were determined and the results are shown in Figure 2.5 for each cell line tested in these paired experiments. In general, MEDII pretreatment increased the final CFC frequency, and this was most obvious if the CFC frequency was very low or undetectable in the untreated arm. The most pronounced effect in this regard was on the frequency of CFU-GM, where in several experiments an approximate 100-fold increase was observed in CFU-GM frequencies obtained from EBs generated from MEDII-pretreated cells (both ES and iPS). However, if the CFC frequency was already relatively high in the untreated arm, exposure to MEDII sometimes had a detrimental effect (e.g., on CFU-E from H1 cells). Since MEDII causes cells to begin differentiating, it was of interest to determine whether or not the kinetics of hematopoietic progenitor differentiation was then accelerated. To investigate this possibility, I collated the CFC frequencies measured after different periods of time in EB culture, either with or without prior MEDII treatment (Figure 2.6a). It can be seen that 59  Figure 9 Figure 2.5  CFU-E  CA1  CFC/105 cells  100  CFC/105 cells CFC/105 cells  100 10  1  1  1000  MSC  1  CFU-GM  10  0.1  H1  CFU-GEMM  0.1  10  1000 100  100 10  0.1  1  10 1  1 0.1  0.1  0.1  100  10  100 10  10 1  1  1 0.1  0.1  No pretreatment  0.1  MEDII pretreatment  Figure 2.5 – MEDII increases the frequency of CFCs subsequently generated in EB cultures CFC frequency is shown separately for each CFC subtype when the starting ES/iPS cells had been pretreated with MEDII conditioned medium (solid circles) or not (open circles) prior to being transferred to EB differentiation cultures (without the addition of cytokines). Lines connect untreated and MEDII treated points from the same experiment. CA1 n=8, H1 n=9, MSC n=18.  60  Figure 10 Figure 2.6  A  1000  No pretreatment  CFC/105 cells  100  B  No MEDII pretreatment pretreatment 40  40  30  30  20  20  10  10  10 1  0  0.1  6 11 15 22  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20 21 22  MEDII pretreatment  CFC/105 cells  100  CFC/105 cells  1000  4  4 20  3 2  10  1 0  6 11 15 22  30  10  20  1 10  0.1 0  0.01  4  5  6  7  8  6 11 15  0 330 320 310 300 50 40 30 20 10 0  4 7 11 15 20  4 7 11 15 20  5  9 10 11 12 13 14 15 16 17 18 19 20 21 22  Day of EB Culture  4 7 11 15 20  5  30  0.01  0  CFU-GM CFU-GEMM CFU-E  4 3 2 1 0  4 7 11 15  Day of EB Culture  Figure 2.6 – MEDII does not alter the kinetics of hematopoietic CFC appearance in subsequent EB cultures A) Total CFC frequencies obtained after the indicated time in EB cultures initiated with human ES cells (H1, CA1, CA2, or H9) or iPS cells (MSC) that were not pretreated with MEDII (top panel) or that were (bottom panel). Each dot represents the total CFC frequency measured in a different experiment; lines represent the geometric mean of each column. B) Frequencies and distribution of CFC subtypes in specific timecourse experiments shown in Panel A, where the starting ES/iPS cells were not pretreated with MEDII (left panels, 3 experiments, CA1) or that were MEDII-pretreated (right panels, 4 experiments, MSC). Specific CFC subtypes are shown as indicated in the legend.  61  the CFC frequencies varied markedly at all time points. However, there is a window between days 11 and 18 of EB culture where a similar average CFC frequency is obtained, regardless of prior MEDII treatment. To see if this would also hold true within individual experiments, EBs were analyzed from MEDII-pretreated cells at a series of timepoints throughout EB culture. These data were then compared to a similar set of experiments from EBs with no MEDII pretreatment. In individual experiments, the peak CFC frequency was detected at either day 11 or 15 of EB culture regardless of previous MEDII treatment (Figure 2.6b). Thus the enhancing effect of MEDII is most readily interpreted as increasing the number of cells at a very early stage that can then be induced to progress further into a hematopoietic program.  2.3.4 MEDII conditioned medium primes cells for mesoderm formation upon transfer to EB conditions  To investigate more precisely the effect of MEDII on undifferentiated ES cells, I examined the effect of MEDII treatment on HES3 ES cells that had been genetically engineered to express GFP instead of one allele of MIXL1 (HES3-MIXL1GFP/w cells291). This reporter cell line was thus used to observe the emergence of MIXL1+ cells during and after MEDII treatment. The results showed that very few cells became GFP+, even when cells were left in MEDII conditions for up to 12 days, i.e., beyond the 7-11 days of MEDII pretreatment used to demonstrate an enhancement of hematopoietic CFC output (Figure 2.7a and b). Nevertheless, the HES3MIXL1GFP/w cells showed a large increase in HAND1 (4000-fold) and IGF2 (1400-fold) transcripts after 7 days of MEDII treatment, and a slight increase (<5-fold) in BRACHYURY transcripts between 2 and 4 days of MEDII treatment (Figure 2.7c). These increases are similar to those we previously reported for other MEDII-treated ES cell lines (HAND1, >1000-fold and IGF2, ~400-fold after 7 days, and BRACHYURY, ~7-fold after 3 days)155. This suggests that the effect  62  Figure 11 Figure 2.7  A  5  % GFP  4 3 2 1 0 0  2  4  6  8  10  12  Day of MEDII treatment  B  GFP  2  4  6 8 Day of MEDII treatment  10  12  C Fold change  100  10000  BRACHYURY MIXL1  1000 10  100  HAND1 IGF2  1  10 0.1  1 0  2  4  6  8  10  12  0  2  4  6  8  10  12  Day of MEDII treatment Figure 2.7 – MEDII generates few MIXL1+ mesoderm cells A) The percentage of GFP+ cells within the population of viable (PI-) HES3-MIXL1GFP/w cells measured after different periods of incubation in 50% MEDII (3 independent experiments). B) FACS plots for the experiment indicated by the orange line in (A). C) Changes BRACHYURY, MIXL1 (left panel), HAND1, and IGF2 (right panel) transcript levels as determined by Q-RT-PCR, and averaged for the experiments indicated by the green and orange lines in (A). Each sample was assigned a δ-ct value with respect to the level of GAPDH, and then a δδ-ct value with respect to the day 0 value in that experiment (undifferentiated cells). Fold-change indicates the expression relative to undifferentiated cells calculated using the formula: fold change=2-(δδ-ct). 63  of MEDII on undifferentiated ES/iPS cells is consistent, but does not stimulate a full transition to a mesodermal state. Since Davis et al. had shown that the MIXL1-GFP reporter was rapidly turned on in EBs cultured in medium containing BMP4291, I next used the HES3-MIXL1GFP/w reporter cell line to examine the frequency and kinetics of mesoderm formation (MIXL1GFP expression) in EB cultures generated from cells pretreated with MEDII, or not. HES3-MIXL1GFP/w cells were cultured in either mTeSR or MEDII for 7 days, and then transferred to EB culture + cytokines also including 25 ng/mL BMP4 (STEMCELL Technologies). In this set of experiments, GFP was nearly undetectable in both the mTeSR and MEDII treated cells at the end of the 7-day culture. Every other day between days 1 and 15, the subsequently formed EBs were harvested, dissociated to obtain a single cell suspension, and the cells then analyzed by flow cytometry to determine the percentage of GFP+ cells. The results obtained (Figure 2.8) showed that MIXL1+ mesoderm progenitors may be generated at a slightly higher frequency but with similar kinetics in EBs derived from MEDII-pretreated cells. These findings support the observation in the previous section that MEDII does not significantly alter the kinetics of hematopoietic CFC generation in EBs. In some of these experiments, cells were also collected every 4th day between days 3 and 15 for analysis of transcript levels of genes whose expression is believed to be selectively upregulated in mesodermal and hematopoietic cells (Figure 2.9). As previously shown, transcript levels for the mesoderm markers BRACHYURY, MIXL1, HAND1 and IGF2 were higher after the 7 days of MEDII treatment. However, after another 3 days in EB culture, these were increased further, and non-pretreated EB cells already expressed similar levels of these genes. Despite the observation of similar percentages of GFP+ cells at later timepoints, transcript levels of the early mesoderm markers MIXL1 and BRACHYURY appeared to decrease more rapidly in the MEDII-pretreated EBs, whereas the more highly expressed mesoderm markers HAND1 and IGF2 remained at similarly high levels in both groups. 64  Figure 12 Figure 2.8  A  % GFP  60  TeSR EBs MEDII EBs  40  20  16  14  12  10  8  6  4  2  0  0  Day of EB culture  B  Day of EB culture 1  3  5  7  9  11  13  _ + GFP Figure 2.8 – MEDII pretreatment does not significantly alter the kinetics of mesoderm formation in EBs A) The percentage of GFP+ cells within the population of viable (PI-) HES3-MIXL1GFP/w cells cultured in either mTeSR (blue line) or MEDII (red line) for 7 days prior to transfer to EB cultures, which was considered here as day 0 of EB culture (n=4). B) FACS plots from one of the 4 experiments represented in (A) from day 1 to 13. Top row (-) and bottom row (+) show results for cells grown in mTeSR and MEDII respectively, prior to the initiation of EB cultures.  65  Figure 13 Figure 2.9  BRACHYURY  MIXL1  1000  100  100  10  10 1 1 0.1  0.1 0.01  0.01 0  3  7  11  15  0  3  HAND1  7  11  15  IGF2  GATA2  100000  100000  10000  10000  10000  1000  1000  1000  100  100  10  10  1  1  100  Fold change  10  0.1  1  0.1 0  3  7  11  15  0.1 0  3  CD34  7  11  15  0  3  KDR  1000  7  11  15  11  15  11  15  CDX4  10  10 1  100  0.1 10  1 0.01  1  0.001  0.1  0.1 0  3  7  11  15  0.0001 0  3  SCL  7  11  15  0  3  CD43  100  7  HOXB4  10  1000  100 10 1  10  1 1  0.1  0.1 0  3  7  11  15  0.1 0  3  7  11  15  0  3  7  Day of EB culture Figure 2.9 – Mesoderm and hematopoietic gene expression in EBs derived from MEDIIpretreated and untreated starting cells For each sample, every gene was assigned a δ-ct value with respect to the level of GAPDH, and then a δδ-ct value with respect to undifferentiated cells (day 0 EB culture for cells grown in mTeSR) in that experiment (HES3-MIXL1GFP/w cells, n=2). Expression relative to undifferentiated cells was then calculated by 2-(δδ-ct). Prior to initiation of EB cultures cells were pretreated with MEDII for 7 days (red line) or maintained in mTeSR (blue line).  66  Interestingly, GATA2, which has an important role in the proliferation and survival of early hematopoietic embryonic and definitive progenitors199,200,300, was expressed at relatively high levels in the MEDII-treated cells, and continued to be expressed at higher levels in MEDIIpretreated EB cells for at least the first 3-7 days of EB culture. No significant differences could be detected in transcript levels at any timepoint for several markers of hematopoietic cells or their precursors (CD34, SCL, CD43, KDR, and CDX4). However, HOXB4, a transcription factor involved in hematopoietic stem cell expansion, did appear to be increased after day 7 in EBs derived from MEDII-pretreated cells compared to EBs derived from non-pretreated cells. Taken together, this supports the data in the previous section where the frequency but not the kinetics of CFC generation is altered by MEDII treatment.  2.3.5 Optimizing multiple steps of differentiation improves hematopoietic cell generation  Another parameter that might be expected to improve hematopoietic output is the addition of hematopoietic cytokines during the EB culture, prior to and during the emergence of CFCs143,290. To test this possibility in the context of EB initiation with MEDII-treated cells, 4 groups were analyzed: EBs initiated with ES cells that had either been pretreated with MEDII or not, and within each of these two groups, EBs were cultured with or without cytokines. Only the combination of MEDII-pretreated cells cultured in EBs + cytokines led to a significantly increased CFC frequency, and this combination also decreased the variation between experiments (Figure 2.10a). When groups were compared within individual experiments, the trend was for either or both manipulations to increase the CFC frequency when compared to the untreated control. However, as seen with MEDII alone, this trend was most distinct when poorer differentiation was observed from the untreated control (Figure 2.10b). When the frequencies of different types of CFCs were analyzed, it showed that the observed increase in total CFC frequency was due to an increase in CFU-GM when MEDII, and especially MEDII + cytokines 67  Figure 2.10 – MEDII pretreatment with subsequent cytokine addition to the EB cultures increases the frequency of hematopoietic CFCs A) CFC frequencies obtained in EB cultures initiated and maintained as shown; starting cells were H1 or CA1 human ES cells or MSC human iPS cells. Each dot represents a different experiment in which ≥1 CFC was detected. Log transformed values were used to calculate the geometric mean (±SEM). To determine significant differences in average CFC frequency, log transformed values were analyzed using 1-way ANOVA with Tukey’s Multiple Comparison post-test, * p<0.05, ** p<0.01, *** p<0.001. B) Total CFC frequencies obtained in paired experiments with at least 3 of the 4 conditions tested (positive results are shown in (A)). The dot colour indicates the conditions used in each arm as in (A). Open circles indicate that no CFCs were detected and represent the limit of detection for that condition in that experiment. Results are from 7 different experiments from cell lines as indicated (shown in separate columns) in which ≥1 CFC was detected in at least 1 of the 3-4 groups studied. C) Summary of the proportion of experiments in which ≥1 CFC was detected in the experiments shown in (A). The values shown are the number of experiments in which ≥1 CFC was detected/total number of experiments performed (% positive). Differences in the proportions of positive and negative experiments between the MEDII treated and untreated groups were determined using Fisher’s exact test, ** p<0.01. D) CFC frequencies for each subtype of CFC for the experiments shown in (A). Bars represent the geometric mean (±SEM) which was calculated as for the total CFC frequency, except that the frequency representing the detection limit of the experiment was included for values where a specific CFC subtype could not be detected within a positive experiment. Differences in average CFC frequencies were also determined in the same way as in (A), **significantly different from all other groups p<0.01, ††significantly different from “No MEDII” p<0.01. Figure 14 Figure 2.10  68  ***  **  1000  B *  100  CFC/105 cells  1000 CFC/105 cells  A  100 10 1  MSC  H1  MSC  MSC  CA1  1  MSC  0.1  H1  10  Experiments  0.1  C  0.01 No MEDII  D CFC/105 cells  100  No MEDII + cytokines  CFU-E  MEDII  100  MEDII + cytokines  No MEDII  MEDII  No cytokines  41/66 (62)  50/64 (78)  With cytokines  11/24 (46)  30/39 (77)  Total  52/90** (58)  80/103** (78)  CFU-GEMM  100  10  10  10  1  1  1  0.1  0.1 No No MEDII MEDII MEDII MEDII + cytokines + cytokines  CFU-GM  †† **  0.1 No No MEDII MEDII MEDII MEDII + cytokines + cytokines  No No MEDII MEDII MEDII MEDII + cytokines + cytokines  69  were used (Figure 2.10d). Interestingly, the proportion of experiments where CFCs were detected was also increased by the pretreatment with MEDII, but this effect did not appear to be further enhanced by the addition of cytokines to the EB step (Figure 2.10c). This shows that systematically optimizing multiple steps of differentiation can improve the robustness and frequency, and decrease the variability of hematopoietic cell generation from cells maintained in feeder-free conditions.  2.4 Discussion  It is becoming widely appreciated in the field that generating hematopoietic cells from human ES/iPS cells is a viable but variable exercise95,100 (also see Figure 1.4). However, the explanations for these differences remain unknown or controversial. The purpose of the studies described in this Chapter was to address this important unresolved issue using a variety of cell lines and subjecting the cells to various different protocols using 2 endpoints: 1) presence or absence of detectable hematopoietic CFCs (robustness) starting from an estimated >5x105 (OP9 experiments) or >1x106 (EB experiments) undifferentiated ES/iPS cells, and 2) the average final CFC frequency of the positive experiments (output efficiency). Using these endpoints I found that all 4 ES and 1 iPS cell lines tested showed similar robustness (~50% positive experiments) but with considerable variability in final CFC frequencies in positive cultures. This suggests that there is not a large variability in the intrinsic hematopoietic potential of these different lines, and the variability in CFC output seen is most likely due to poor ability of undifferentiated cells to respond directly to the signals generated from OP9 cells or cells in EB aggregates and is determined by a variable presence in the starting cultures of “undifferentiated” cells of a spontaneously generated derivative that is responsive to these stimuli.  70  This finding was initially revealed by introducing changes to the protocol used to maintain the parental ES cultures to increase the representation in them of “undifferentiated” cells which led to a dramatic reduction in CFC output. Increasing the homogeneity of the starting cells might contribute to such changes in differentiation response in several ways. One possibility would be the reduction of a partially differentiated cell population that responds to a particular set of differentiation stimuli, such as those from the OP9 bone marrow stromal cell line. On the other hand, a highly homogeneous population of undifferentiated ES cells might be postulated to be producing paracrine factors reinforcing their undifferentiated state, sufficient to override the stimuli provided to induce their differentiation. It has been reported that the local density of OCT4+ undifferentiated cells can influence signalling in surrounding OCT4+ cells, creating an autoregulatory loop promoting self-renewal. Thus undifferentiated cells surrounded by higher numbers of other undifferentiated cells would be expected to suppress differentiation responses, whereas those surrounded by higher numbers of differentiated cells would be more likely to respond301. Changing the ES/iPS cell maintenance culture conditions to expand the representation of undifferentiated cells could also result in changes in the levels of many growth factors, as well as changes in the composition of extracellular matrix components to which the cells would be exposed. The majority of groups reporting results of experiments which include differentiation into hematopoietic cells appear to have mainly used MEFs or MEF conditioned medium to maintain their undifferentiated cells, suggesting that the current protocols may be optimal for starting cells maintained using these conditions. If these conditions are changed, the cells may activate different signalling pathways, alter their gene expression, or perhaps even be subject to epigenetic changes that could then change their responses to specific differentiation stimuli. Many feeder-free conditions rely on different combinations of cytokines, for example, variable levels of bFGF14,16,21. These, in turn, could predispose the starting cells to be variably responsive to differentiate along particular pathways. Notably in the case of FGF, this cytokine 71  has also been reported to be important in the maintenance of primitive neuroectodermal, mesodermal, and endoderm cells with variable effects depending on the amount and context of the signal282,302,303. Evidence that the initial cells can be intrinsically different is provided by their different morphologies when grown on MEFs versus feeder-free conditions. On MEFs, the cells form more compact colonies several cell layers thick, whereas the same cells grown in the absence of feeders form more diffuse monolayer colonies. This change in cell-cell contact could be hypothesized to influence factors that control the ability of cells to aggregate into EBs. Some changes might even be critical for the success of a particular differentiation protocol. For example, differentiation by forced reaggregation into “spin EBs” has been reported to be critically dependent on the prior adaptation of these cells to enzymatic single cell passaging304. Regardless of the exact mechanism, my observations reinforce the importance of including the method of starting cell maintenance when testing a “differentiation” protocol and in particular, when interrogating the sequence of events required to achieve the differentiated cell type of interest. The question of the extent to which iPS cells are functionally similar to ES cells is of major importance to the field. The human iPS cell line that I chose to use exhibited a similar hematopoietic differentiation potential and responded similarly to manipulations in differentiation conditions as the 4 human ES cell lines tested. However, because these results are from a single iPS cell line, further generalization is not possible. Clearly, continued rigorous characterization of additional human iPS cell lines derived in multiple ways from multiple somatic cell types will be needed to address this question in a meaningful way. Only recently have human iPS cells been derived from several somatic cell types, and it has already been reported that while they generally show a similar gene expression pattern to human ES cells, they also retain varying degrees of similarity to their cells of origin94,101. The idea that current hematopoietic differentiation protocols actually preferentially target a derivative of undifferentiated ES/iPS cells is further supported by my finding that enhancement 72  of an early mesodermal state improved the output of hematopoietic cells. This was demonstrated in 2 ways: both by pretreating the starting cells for a few days with FBS or MEDII, a medium conditioned by HepG2 cells. This extends to the human system findings first documented with mouse ES cells296. MEDII was actually removed from the cells early on, and appears to have subtle effects on immediate differentiation, as the cells do not even progress to a MIXL1+ mesodermal state, yet it enables increased CFC frequencies to be obtained at later timepoints. We previously showed that MEDII treatment drastically decreases the proportion of cells expressing the pluripotency markers SSEA3 and OCT4155. In addition to the lack of MIXL1GFP expression, I have also found that very few MEDII-treated cells have begun to downregulate cell surface expression of epithelial cell adhesion molecule (Ep-CAM; data not shown), which is decreased on mesodermal cells during gastrulation, and constitutes another early sign of human ES cell differentiation305,306. Therefore, it is not yet clear what populations of cells might be represented at the end of MEDII treatment. MIXL1 is probably not expressed in every mesoderm cell, and MixL1-null mouse embryos appear to form normal blood islands in the yolk sac before lethality at day 10292. Thus, it is likely that at least some mesodermal cells are not required to pass through a MIXL1 expressing stage in order to generate hematopoietic derivatives. One postulate could then be that MEDII acts to increase a type of MIXL1- precursor that can generate hematopoietic cells. Alternatively, MEDII could be mimicking patterning events that occur in the epiblast prior to the formation of the primitive streak. It has been reported that mesoderm cells in the chick embryo are specified in the epiblast307 and cells within the mouse epiblast are reported to show a heterogeneous pattern of gene expression308 with only particular regions containing cells that will contribute to the future mesoderm309,310. However, the contribution of these cells to specific types of mesoderm in the mouse appears to be specified during gastrulation by the timing and location of their ingression through the primitive streak232,310. Only extraembryonic and lateral plate splanchnopleuric mesoderm are thought to give rise to hematopoietic cells159 and these are derived from different regions of 73  epiblast at different times during gastrulation232. In humans, it is still not apparent to what extent these different mesodermal fates are already being programmed in the epiblast, how much of this activity is determined by positional information accrued during gastrulation, nor what signals are responsible. MEDII exposure of undifferentiated ES cells not only downregulated expression of features of undifferentiated cells (like SSEA3 and OCT4), it also caused several mesodermal and hematopoietic genes to be upregulated, although after only a short period of subsequent EB culture most of these genes were upregulated to similar levels in MEDII-pretreated and nonpretreated EBs. GATA2 and HOXB4 were identified as 2 genes that were more highly expressed in MEDII-pretreated EBs in the first or second week of culture respectively. Gata2 is induced by BMP4 and expressed in early ventral lateral mesoderm and ectoderm in Xenopus, where its overexpression augments hematopoiesis311,312. In mouse, Gata2 is similarly expressed in early lateral mesoderm and later in the paraaortic splanchnopleura and AGM regions where early hematopoietic cells are generated313,314. It is required for definitive hematopoiesis due to a role in the proliferation and maintenance of early progenitors or stem cells199,201 and Gata2-null mouse ES cell lines have a profoundly decreased ability to generate hematopoietic CFCs, as well as their precursors, BL-CFCs, in which Gata2 is highly expressed200,263. Thus, increased GATA2 expression in MEDII-pretreated EBs may reflect increased specification of prehematopoietic lateral mesoderm induced by a high concentration of BMP4 in MEDII. During the second half of the EB culture protocol, HOXB4 appeared to be increased in MEDII-pretreated EBs, although this increase was not significant when compared to EBs from untreated cells. As HOXB4 has been shown to expand primitive mouse128 and human315,316 hematopoietic cell populations, this may reflect differentiation from the cells overexpressing GATA2 and may be important for the increased frequency of hematopoietic CFCs observed from MEDII-pretreated EBs. Forced expression of high levels of HOXB4 has been reported to improve hematopoietic cell output from human ES cells289,290; but it is unclear as to whether HOXB4 can enable the 74  hematopoietic cells produced to also gain in vivo repopulating activity288,289,317 as is the case with mouse ES cells184,192. Interestingly, MEDII treatment did not appear to simply increase the number of mesodermal and hematopoietic precursors, since the different types of CFCs were not increased equally. MEDII specifically increased the generation of CFU-GM, despite the fact that these appeared after MEDII was removed from the cells. This suggests two possible explanations. One is that MEDII preferentially supports the emergence of a specific type of mesoderm or hematopoietic precursor that has increased or exclusive capacity to give rise to CFU-GM. Alternatively; it may be that MEDII enables ES cells to be epigenetically programmed in a way that predisposes subsequently generated precursors to differentiate into CFU-GM cells. The first could be achieved if, for example, extraembryonic and intraembryonic mesoderm gave rise to different types of hematopoietic cells, with MEDII favouring the formation of one type of mesoderm over the other. This is indeed plausible, as extraembryonic YS mesoderm generates primitive erythroid derivatives that the intraembryonic sites do not. Interestingly, MEDII seemed to have its greatest effect on ultimate hematopoietic CFC generation when the starting ES/iPS cells were destined to otherwise show a very poor differentiation response. This, in turn, suggests that the number of initial cells that are susceptible to the effect of MEDII may be limited. Thus if the starting population has a large proportion of cells of the type MEDII would generate, further exposure to MEDII would not have an additional effect. The cytokine combination added to the EB cultures has been shown to aid expansion of adult, neonatal, and ES-derived human hematopoietic cells131,132,143,298 and several of these cytokines have a specific role in myelopoiesis109,318,319. Addition of these cytokines to the EB cultures further increased the frequency of CFU-GM obtained, but did not appear to increase the frequency of more primitive CFU-GEMM. Interestingly, this effect was only apparent when cells had also been pretreated with MEDII, i.e., cytokines alone did not increase the final CFC frequency. This implies distinct mechanisms for these 2 manipulations, such that cytokines do 75  not promote the generation of the precursors that are generated by MEDII. In addition, cytokines alone did not improve the robustness of hematopoietic differentiation, suggesting that the success of CFC production in EBs is largely determined at a relatively early stage in this system, by the presence or absence of precursors whose generation is supported by MEDII but not the hematopoietic cytokines used. The fact that the cytokine effect was seen only when cells were pretreated with MEDII, also implies that the primary effect of the cytokines is later on in the production of CFCs from precursors generated within the EBs. Conversely, it shows that MEDII generates precursors that are more susceptible to the effects of cytokines, since EBs generated in the absence of MEDII showed no increase in CFC frequency when cytokines were added. In summary, I have shown in this Chapter that increasing the homogeneity and reproducibility of undifferentiated human ES or iPS cell populations by transfer of the cells to defined, feeder-free conditions had the unexpected downstream consequence of reducing their output of hematopoietic cells using “standard” protocols. Subsequent optimization of the differentiation protocol to provide more relevant stimuli for cells at different stages improved the hematopoietic output, and in particular was facilitated by a MEDII treatment of the starting cells, with later addition of hematopoietic cytokines known to enhance the production of primitive hematopoietic cells.  76  3. Quantitative analysis of the generation and expansion of human ES and iPS cell-derived hematopoietic cells within individual EBs  3.1 Introduction  In the previous chapter, I obtained evidence of 2 sources of variability in the output of hematopoietic CFCs from undifferentiated human ES and iPS cells. One relates to the efficiency with which undifferentiated cells initiate differentiation into a very early mesodermal cell type and which can be promoted by exposing undifferentiated cells to factors present in MEDII. The second relates to the efficiency with which these cells subsequently generate hematopoietic CFCs, as supported by the finding that this could be enhanced by the addition of hematopoietic growth factors to the EB cultures. However, even within the same conditions, much variation still remained, and CFCs were rare – about 0.01-0.1% of the total differentiated cells which was, on average, between 1-7 CFCs per EB. This then raised the question as to whether there also remains an intrinsic variability between the cells in EBs to give rise to hematopoietic cells or whether the persisting variability seen is due to heterogeneity between EBs. To begin to address this question, I used the approach of investigating the CFC content of large numbers of individual EBs. EBs derived from human ES/iPS cells are not clonal populations, so such an analysis could not be used to infer directly the output of individual cells. However, each EB is an entity generated from a discrete clump of cells, therefore allowing assessment of the variation in hematopoietic output from small groups of original ES/iPS cells. In addition, I hypothesized that this analysis method could help to identify differentiation steps that occur with variable efficiency in EB cultures. Figure 3.1 depicts 2 simplified models of CFC generation from an initially undifferentiated or early mesodermal cell where EB  77  Figure 15 Figure 3.1  “Late bottleneck”  “Early bottleneck” hPSC CFC  hPSC  precursor  CFC  CFC  hPSC  precursor  CFC  CFC  hPSC  precursor  CFC  CFC  hPSC  precursor  CFC  hPSC hPSC  precursor  hPSC  EBs: CFCs: CFC distribution highly skewed  CFCs more evenly distributed  Figure 3.1 – Models of CFC generation and expansion Two generic schemas depicting different models of hematopoietic differentiation and the predicted outcomes from analyses of single EBs. Differentiation is split into 3 stages, the starting cells – human ES or iPS cells, here collectively labelled “hPSCs” (blue), differentiating cells contained in EBs, some of which are hematopoietic precursors (purple), and cells at the end of the differentiation protocol, some of which are hematopoietic CFCs (red). The left hand model shows the generation and expansion of differentiating cells if a limiting step occurs early in the differentiation pathway, or “early bottleneck” and the right hand model shows what happens if a limiting step occurs later in differentiation, or “late bottleneck.” Underneath is a schema showing how these cell types might be expected to distribute between individual EBs in a culture; EBs shown depict aggregates of hundreds to thousands of cells each. Coloured EBs indicate the presence of one or more cells of the type indicated in the above flow chart. Grey EBs represent aggregates that have entirely differentiated into cell types outside of the hematopoietic pathway. At the end of differentiation the intensity of pink/red indicates the relative numbers of CFCs present. Below the EBs is a depiction of the expected results of multiwell CFC assays where the cells from each of the above EBs would be analyzed (one per well), with the intensity of the colour of the well representing the number of CFCs generating colonies in each well.  78  differentiation to hematopoietic cells is represented by 3 cell types: the starting cells, an intermediate state generated within EBs, and the CFCs. Between these cell types there are two transition stages: an early differentiation step that generates precursors within EBs, and a late differentiation step that generates CFCs from these precursors. The hematopoietic CFC output would then depend on the efficiencies with which both of these transitions occur. The “early bottleneck” and the “late bottleneck” depicted in the top flow charts, demonstrate how the same CFC output could be obtained with different probabilities of differentiation occurring at each step. In the “early bottleneck” model, the efficiency of the early steps of differentiation is poor and few precursors of hematopoietic cells are successfully generated. However, the later steps of differentiation are well supported and occur with higher efficiency such that once a precursor is generated, it can differentiate and generate multiple CFCs. In the “late bottleneck” model, the early steps of differentiation are relatively efficient, and many precursors can be generated from the starting cell population. However, each precursor has a limited potential and differentiates into few CFCs. Analysis of large pools of EBs precludes the ability to distinguish between these possibilities, but should be accessible to analysis by assessment of individual EBs as illustrated in the bottom portion of Figure 3.1. If the generation of precursors is extremely rare, such as in the “early bottleneck” model, then few EBs would contain any precursors. However, EBs that successfully generate a precursor would contain several CFCs as that precursor proliferates and differentiates. This would result in a highly skewed distribution of CFCs amongst individual EBs. If the generation of precursors is less limiting, then many EBs would contain precursors. However, in the “late bottleneck” model, differentiation and expansion from the precursors within each of these EBs would be limited and positive EBs would end up with relatively similar numbers of CFCs. This would result in a more even distribution of CFCs amongst EBs. Of course, these models do not represent all the possibilities, and are very simplified. For example, there is likely multiple intermediate cell types between the cells used to initiate EB formation and 79  CFCs, and it would not be possible using this system to determine the exact stage of differentiation that is affected. In general though, marked effects on early differentiation would be expected to change the proportion of EBs that generate CFCs, whereas marked effects on later stages of differentiation would be expected to change the final CFC content of positive EBs.  3.2 Materials and methods  3.2.1 Human ES and iPS cell maintenance and differentiation cultures  Maintenance of undifferentiated human ES cells (H1 or CA1) or iPS cells (MSC) in feeder-free conditions, generation of EBs from dispase-dissociated ES/iPS cells, EB cultures, the MEDII pretreatment protocol, and the cytokines added to EB cultures were all described in detail in Chapter 2.  3.2.2 Analysis of the hematopoietic CFC content of single EBs  At the end of the EB cultures, individual EBs were randomly selected and removed from the well in a volume of 5-10 μL using a P20 pipette, and were then transferred one by one, each to an individual well in a 96-well U-bottom dish. Each well was then checked under the microscope to ensure that only a single EB had been transferred, and the EB diameter was estimated using an eyepiece gradicule. Each EB was then transferred individually to a 1.5 mL microcentrifuge tube in 50 μL of TrypLE and incubated at 37°C for 20 minutes. Next, 500 μL DMEM was added to each of these tubes and the EB cells triturated vigorously. Cells were centrifuged (at 300 g) for 8 minutes and the supernatants removed. Each pellet was resuspended in 20-40 μL of DMEM and then transferred in its entirety to a single well of a 2480  well ultra low adherence plate (Corning), i.e. one EB per well. To each well, 400 μL of Methocult 4230 supplemented with 50 ng/mL SCF, 20 ng/mL GM-CSF, 20 ng/mL IL-3, 20 ng/mL IL-6, 20 ng/mL G-CSF, and 3U/mL EPO (see Chapter 2) was added on top of the cells using a 3 cc syringe and a 16 gauge blunt ended needle. Cells and Methocult were evenly distributed throughout the well, and sterile water was added to the space between the wells of the dish to maintain humidity. Plates were incubated for 2 weeks at 37°C and then colonies obtained were counted and scored as erythroid (CFU-E), granulocyte/macrophage (CFU-GM), or mixed (CFUGEMM).  3.3 Results  3.3.1 EBs show pronounced heterogeneity in their final content of hematopoietic CFCs  In a first series of experiments to examine the distribution of CFC numbers for individually analyzed EBs, I generated these under “traditional” conditions in which untreated ES cells (H1 or CA1) were used to initiate EB formation and no cytokines were added to the EB cultures. The results shown in Figure 3.2 (left panels) revealed a large variation in the output of hematopoietic CFCs from individual EBs. In fact, most EBs exhibited no hematopoietic CFC generating activity and, within the positive EBs, the number of CFCs present also showed a large variation, between 1 and 90. The very high proportion of negative EBs implies that in these conditions, there is an early bottleneck, or a limit on either the generation or the successful differentiation of early precursors resulting in a highly skewed distribution of hematopoietic CFCs. To determine if EBs derived from human iPS cells would display the same behaviour as EBs derived from ES cells, the experiment was repeated with the MSC line. As shown in Figure 3.2 (right panels) a very similar percentage of the EBs obtained gave rise to CFCs under the 81  Figure 16 Figure 3.2  ESC 100  iPSC 100  400  300  0 1 2 3-4 5-8 9-16 17-32 33-64 65-128  10  40 100  100  20  200  20  0  0  40  10  0 1 2 3-4 5-8 9-16 17-32 33-64 65-128  40  60  # EBs  200  % EBs  300  60  0  80  # EBs  % EBs  80  0  4 10  0  1 2 3-4 5-8 9-16 17-32 33-64 65-128  2  # CFC/EB  0  6  20  4 10 2 0  1 2 3-4 5-8 9-16 17-32 33-64 65-128  20  8  # EBs  6  % total EBs  30  # EBs  % total EBs  30 8  0  # CFC/EB  Figure 3.2 – The distribution of CFCs amongst individual EBs derived from human ES or iPS cells is highly skewed Distribution of CFC amongst individual EBs derived from human ES cells (left) and iPS cells (right). The number of CFCs per individual EB analyzed is shown on the x-axis, binned on a log2 scale. The percentage of all EBs that fell into each bin is indicated on the left y-axis. The absolute number of EBs that fell into each bin is indicated on the right y-axis (n=409 in 13 experiments for ES cells, n=336 in 11 experiments for iPS cells). Lower graphs show the same data with the EBs containing zero CFCs removed so that the heights of the other bars can be better visualized.  82  same conditions, and the numbers of CFCs per positive EB was similarly variable (range=1-38). Representative experiments showing the subtypes of CFCs and their distribution amongst 48 individual EBs from both sources are shown in Figure 3.3. This shows that even on this more detailed level, at least this one iPS cell line appears to have similar biological properties and differentiation activity as the 2 human ES cell lines tested. To ensure that these observations were not influenced by the modified method used to collect and dissociate single EBs, the average CFC output per total EB number was compared for all EBs analyzed singly versus the EBs pooled and harvested in bulk from the same experiments. For the single EBs, all CFCs were counted and divided by the total number of EBs analyzed, and for the pooled EBs the CFC per EB was determined by (CFC frequency) × (total cell #) ÷ (total # EBs). In 22 of 24 experiments these values were similar, showing that this analysis method gives an accurate representation of the CFC distribution and that significant numbers of CFCs are not gained or lost when compared to bulk analyses (Table 3.1). Taken together, these data show that there is a large difference in the hematopoietic CFC output of individual EBs present in the same EB differentiation cultures. The general rarity of hematopoietic CFC generation suggested that an early differentiation step might be limiting, such that most EBs failed to generate CFCs because they did not contain or generate an essential intermediate precursor cell type. However, once present, there still remains an additional source of variation in the efficiency of later steps resulting in a wide variation in CFC outputs in individual EBs.  83  Figure 17 Figure 3.3  ESC (12/48 positive)  CFC/EB  40  CFU-GM CFU-GEMM CFU-E  30 20 10 0  iPSC (9/48 positive)  CFC/EB  40 30 20 10 0  Individual EBs Figure 3.3 – Representative analysis of CFC output from individual EBs derived from human ES or iPS cells CFC output from 48 human ES cell-derived EBs (upper panel, expt. 5 in Table 3.1) and 48 human iPS cell-derived EBs (lower panel, expt. 11 in Table 3.1). One individual EB is shown in each column, and columns are arranged in order of increasing CFC/EB. Only positive EBs are shown. Colours show the number of each type of CFC as indicated.  84  Table 3.1 – CFC outputs in individual and pooled EBs  Expt #  Cell line  CFC/105 cells (pooled)  CFC/total EBs (pooled)  CFC/total EBs (individual)  1  H1  176  2.3  2.5  2  H1  64  2.3  2.8  3  H1  346  8.7  23  4  H1  459  11  1.9  5  H1  34  1.7  1.6  6  H1  nd  0.3  0.5  7  H1  72  2.0  1.9  8  H1  nd  nd  0  9  H1  19  0.5  0  10  MSC  18  1.2  1.1  11  MSC  65  2.3  1.6  12  MSC  178  2.7  4.8  13  MSC  44  1.7  3.0  14  MSC  1.7  0.14  0.29  15  MSC  4.8  0.24  0.21  16  MSC  0.4  0.02  0  17  CA1  <0.4  0  0  18  MSC  3.2  0.25  0  19  CA1  0.4  0.02  0  20  CA1  <0.25  0  0  21  MSC  <0.5  0  0  22  MSC  <0.25  0  0.04  23  MSC  <0.25  0  0  24  CA1  2  0.29  0.13  Table 2 Table 3.1  nd=not determined  85  3.3.2 MEDII pretreatment and cytokine addition to EB cultures increases the output of hematopoietic CFCs per EB but does not alter the skewed distribution of CFCs amongst EBs  Based on the finding that the output of CFCs can be improved by pretreating cells with MEDII and adding cytokines to the EB culture155 (see Chapter 2), I next designed experiments to determine whether these would also affect the distribution of CFCs at the single EB level; i.e. would a greater proportion of EBs then contain CFCs, implying improvement in an early differentiation step, and/or would a greater number of CFCs be generated per positive EB, implying an improvement in later steps. Since MEDII is applied to undifferentiated cells which are then primed for mesodermal differentiation at the time of EB formation, it was anticipated that the use of MEDII-pretreated cells to initiate EBs would increase the proportion of positive EBs. Conversely, the addition of cytokines was expected to act later, hence mostly act to increase the number of CFCs observed per positive EB. To test these possibilities, I examined single EBs cultured under each of the 4 possible conditions: ± MEDII pretreatment and ± cytokine addition to the subsequent EB cultures. Because it was not possible to harvest single EBs from all conditions in all experiments, experiments from similar time periods and using the same ES (H1 and CA1) and iPS (MSC) cell lines were combined for these comparisons. In these particular experiments, the proportion of positive EBs initiated from untreated input cells and cultured without cytokines was extremely low, but pretreatment with MEDII or the addition of cytokines increased both the percentage of positive EBs, and the number of CFCs per positive EB. Both of these increases were more pronounced when both manipulations (MEDII pretreatment and cytokine addition) were used together (Figure 3.4). Interestingly, the significant proportion of negative EBs was not eliminated and the skewed distribution of CFC per EB remained. This suggests that neither of these manipulations were sufficient to alter either the early or late proposed bottlenecks. 86  Figure 18 Figure 3.4  A 100  No MEDII No MEDII + cytokines  MEDII MEDII + cytokines  % EBs  80 60 40 20  0 1 2 3-4 5-8 9-16 17-32 33-64  0 1 2 3-4 5-8 9-16 17-32 33-64  0 1 2 3-4 5-8 9-16 17-32 33-64  0 1 2 3-4 5-8 9-16 17-32 33-64  1 2 3-4 5-8 9-16 17-32 33-64  1 2 3-4 5-8 9-16 17-32 33-64  1 2 3-4 5-8 9-16 17-32 33-64  1 2 3-4 5-8 9-16 17-32 33-64  0  % total EBs  15  10  5  0  # CFC/EB  B  No MEDII  No MEDII+ cytokines  MEDII  MEDII+ cytokines  % positive EB  2.5  16.7  6.3  32.5  Average CFC/ positive EB  1.5  5.7  7.1  8.9  Figure 3.4 – MEDII pretreatment and cytokine addition to EB cultures increases both the proportion of positive EBs and the number of CFCs per EB A) Distribution of CFC amongst individual EBs generated using different culture conditions. The number of CFCs observed in an individual EB is shown on the x-axis, binned on a log2 scale. The percentage of all EBs within each group that fell into each bin is indicated on the left y-axis (n=240 in 10 experiments for no MEDII; n=192 in 6 experiments for no MEDII + cytokines; n=334 in 14 experiments for MEDII; n=382 in 8 experiments for MEDII + cytokines). Lower graphs show the same data with the EBs containing zero CFCs removed so that the heights of the other bars can be better visualized. B) Summary of the percentage of positive EBs and the average CFC output per positive EB in the experiments in (A). 87  An additional source of heterogeneity could be EB size, as the observed diameter of EBs at the end of differentiation culture ranged from about 200-2000 μm. In these experiments, there was no difference in the average EB diameter observed between the 4 different conditions, suggesting that this was not the reason for the observed changes in CFC output per EB, however, the overall range in EB size tended to be slightly larger for EBs derived from MEDII-pretreated cells in these experiments (Figure 3.5a, left bars). Since the EBs capable of giving rise to CFCs are a small subset of the total EBs, differences in the size of this specific subset could be obscured by the larger set of negative EBs, the size of positive and negative EBs were also analyzed separately for each condition. This revealed that in each condition, the EBs positive for hematopoietic CFCs were slightly larger on average than the corresponding negative EBs. The EBs containing no CFCs were always distributed over the entire size range, implying that this is not the only factor influencing CFC generation (Figure 3.5a, right bars). In support of this hypothesis, when the positive EBs were further examined for a correlation between EB diameter and number of CFCs per EB, a very weak correlation was only observed for EBs cultured with cytokines and derived from MEDII-pretreated cells (Figure 3.5b). These data support the idea that hematopoietic cytokines can be involved in both the generation and expansion of CFCs. Also, the effect of MEDII may not be as straightforward as originally anticipated, since its use increased the proportion of subsequently obtained positive EBs (as expected), but not dramatically, and at the same time, also increased the number of CFCs subsequently generated in these EBs, despite no longer being present during the EB culture. Overall, the pretreatment of the input ES/iPS cells with MEDII and the subsequent addition of cytokines to the EB cultures increased both the likelihood of generating CFCs and later expanding their numbers. Nevertheless, these manipulations did not overcome all the current limitations for inducing the generation of hematopoietic cells as the majority of EBs, even under these conditions, still do not generate any CFCs.  88  EB diameter (μm)  A 2500  No MEDII No MEDII + cytokines MEDII MEDII + cytokines  2000 1500 1000 500 0 All EBs  B  - + - +  - + - +  No MEDII  No MEDII + cytokines  4  25  p=0.6 r2=0.07  3  20  p=0.1 r2=0.07  15  2 10  CFC/EB  1  5  0  0 0  500 1000 1500 2000 2500  0  MEDII  500 1000 1500 2000 2500  MEDII + cytokines  30  60  p=0.9 r2=0.001  20  10  p<0.05 r2=0.1  40  20  0  0 0  500 1000 1500 2000 2500  Figure 19 Figure 3.5  0  500 1000 1500 2000 2500  EB diameter (μm)  Figure 3.5 – Relationship between EB size and hematopoietic CFC output A) Diameter of individual EBs at the end of differentiation for the experiments shown in Figure 3.4. In each column, the box encloses the 25th to 75th percentile with a line at the median value and the whiskers show the range. The first 4 columns show the size distributions of all of the EBs from each condition. The right hand side of the graph shows the same data split into EBs where no CFCs were detected (-) and where CFCs were detected (+) separately for each condition. B) Scatter plots showing EB diameter versus the number of CFCs detected in individual EBs where CFCs were detected for the experiments shown in Figure 3.4. Each dot represents one EB. The linear regression line is shown with the p-value (that the slope differs from zero), indicating a correlation between the variables, and the r2 value for each condition. A weak correlation is seen only in conditions with MEDII + cytokines. 89  3.3.3 An increased proportion of EBs derived from MEDII-pretreated cells and cultured in the presence of cytokines generate CFU-GM  I showed in Chapter 2 that the addition of MEDII and cytokines to differentiation conditions increased the CFC frequency, with CFU-GMs being the most affected. Accordingly, it was of interest to determine how these changes would be manifested within single EBs; i.e. would we see a uniform effect in all EBs, or would this effect be limited to a subset of EBs, suggesting a precursor-specific effect. The distribution of CFC subtypes present in single EBs was examined in EBs generated from cells that were MEDII-pretreated and also received cytokines during EB culture, and compared with those that had neither treatment. Data showed that the average CFC number of all types, per total number of EBs was 2.6 for non-treated EBs, and 2.9 for the MEDII + cytokines group which was not significantly different in these experiments (p>0.05). However, there was a difference in the subtypes of CFCs detected overall, with the MEDII + cytokines group having relatively more CFU-GM and fewer CFU-E (Figure 3.6a). When CFC subtypes were examined within individual EBs, it became apparent that this shift was due to a greater proportion of EBs that had the ability to give rise to CFU-GM, and fewer that could give rise to CFU-E in the MEDII + cytokines condition (Figure 3.6b). These data are summarized in tabular form in Table 3.2. In both conditions, there were some EBs that could give rise to multiple CFC types within the same EB (Figure 3.6c). Due to the lower proportion of EBs that could generate CFU-E, there were actually fewer of these multipotent EBs in the MEDII + cytokines group, and no EBs that only generated CFU-E. The shift towards CFU-GM in MEDII + cytokines was also due to a much higher proportion of EBs that gave rise to only CFU-GM. These data suggested an early effect, where different types of precursors that subsequently gave rise to different CFC types were being differentially generated and supported in EBs in different conditions. To test whether this was the case, or if later effects on the 90  Figure 3.6 – Distribution of CFC subtypes in individual EBs A) The overall distribution of CFU-E (red), CFU-GEMM (yellow), and CFU-GM (blue) detected in individually analyzed EBs. The total number of CFCs was determined in all EBs either derived from non-pretreated cells with no cytokine addition (No MEDII, n=745 in 24 experiments) or EBs from MEDII-pretreated cells with cytokine addition (MEDII + cytokines, n=382 in 8 experiments) and the percentage of each CFC subtype was determined. These values are also shown in Table 3.2. B) Percentage of total EBs in each condition positive for any CFCs (black bars) or specifically for CFU-E, CFU-GM, or CFU-GEMM (coloured bars). These values are also shown in Table 3.2. C) Venn diagrams showing the overlap of specific CFC types in individual EBs. Numbers represent the percentage of EBs in each condition containing the combination of CFC types as indicated by the circle overlap. Percentages within each coloured circle add up to the total percentage indicated by the height of the equivalent bar in (B). All percentages in the coloured circles add up to the total percentage indicated by the height of the black bar in (B). D) Within the subset of EBs positive for each type of CFC, the average number per EB of CFCs of that type (±SEM) is calculated. N, No MED; M+C, MEDII + cytokines.  91  Figure 20 Figure 3.6  A  B  80  40  CFC  30  % of EBs  % CFC type  100  60 40 20  CFU-E 20  CFU-GEMM  10 0  0  CFU-GM  No MEDII + MEDII cytokines  No MEDII  MEDII+ cytokines  C  8 6 4 2 0  N M+C  2.0  CFU-GM/GM +ve EB  10  CFU-E/E +ve EB  D  MEDII+ cytokines CFU-GEMM/Mix +ve EB  No MEDII  1.5 1.0 0.5 0.0  N M+C  10 8 6 4 2 0  N M+C  92  Table 3.2 – Summary of the distribution of CFC types for individual EBs  Total EB CFC/EB Total CFC % CFU-E % CFU-GM % CFU-GEMM Total +ve EBs % +ve EBs % CFU-E +ve EBs % CFU-GM +ve EBs % CFU-GEMM +ve EBs  No MEDII 745 2.6 1943 47.1 51.7 1.2 172 23.1 14.9 19.6 2.1  MEDII + cytokines 382 2.9 1108 12.6 85.8 1.5 124 32.5 6.3 32.5 3.1  Table 3 Table 3.2  93  expansion of CFCs once formed could also be detected, the CFC numbers were examined in subsets of EBs positive for each CFC subtype. In fact, when the outputs of specific CFC types were examined just within the subset of EBs with the ability to give rise to that type of CFC, the average number was similar in both sets of conditions (Figure 3.6d). For example, despite the fact that there were far fewer EBs with erythroid potential in MEDII + cytokine conditions, these EBs still contained, on average, similar numbers of CFU-E as the greater number of erythroidpositive EBs in conditions with neither treatment. Similarly, in the presence of MEDII + cytokines, there did not appear to be additional expansion of CFU-GM within individual EBs when compared to neither treatment, therefore the increase in CFU-GM was entirely due to the increased proportion of EBs with the ability to generate CFU-GM. In other words, in these experiments, once an EB gained the ability to generate CFCs of a particular type, the number of these CFCs subsequently produced appeared to be relatively fixed. This suggests that the generation of CFC precursors was more affected by the culture conditions tested than the expansion of the CFCs thus produced, and that the differences observed are likely due to differences in the production of different types of precursors of CFCs within EBs, rather than the alteration of the differentiation of a more generic precursor. I next asked if EBs able to generate multiple types of CFCs also gave rise to more total CFCs, perhaps indicative of the initial generation of a more primitive type of hematopoietic cell. In both the no MEDII and the MEDII + cytokine conditions, such multipotent EBs were found to contain a greater number of CFCs (Figure 3.7a). Contributing to the greater total CFC output, was the combined presence of both increased numbers of CFU-E and CFU-GM (Figure 3.7b and c). Since EBs do not originate from a single cell, it is not clear whether these different types of CFCs originate from a common precursor within the EB, or whether there are multiple clones within EBs each giving rise to the CFCs. Similarly, these data cannot distinguish whether the greater number of CFCs seen arose from a single precursor with greater potential, or simply from a greater number of pre-CFCs being generated within the same EB. 94  Figure 21 Figure 3.7  MEDII+ cytokines  No MEDII 50  50  40  40  CFC/EB  Total output  CFC/EB  A  30 20 10  20 10  nd  0 nd  nd  nd  G M  E  on  o G nl y M G EM onl M y on ly E+ E+ GM G E G M MM E+ +G E G M MM +G EM M  ly on l M ix y on ly E+ G E+ M G E G M MM E+ +G E G M MM +G EM M  0 E  30  B 30  nd  ly  E+ G M  E  on  +G EM E+ G M  E  on  E+ G M  M  0 ly  0  10  M  10  20  E+ G M  20  +G EM  CFU-E/EB  Erythroid output  CFU-E/EB  30  C 25  CFU-GM/EB  20 15 10 5  20 15 10 5 +E  +M ix +E +G EM M  G M  G M  M +E +G EM M  G M  G M  +G EM  +E  G M  G M  on ly G M  on ly  0  0  G M  GM output  CFU-GM/EB  25  EBs containing Figure 3.7 – Multipotent EBs have a greater output of total CFCs Average total number (±SEM) of CFC (A), CFU-E (B), or CFU-GM (C) for the EB subset containing the indicated combination of CFC types. Results are shown for EBs with no MEDII pretreatment and no cytokine addition, or EBs derived from MEDII-pretreated cells and subsequently cultured with cytokines. nd, EB subset not detected.  95  3.4 Discussion  The experiments described in this chapter demonstrate a previously unappreciated heterogeneity in the quantitative output of hematopoietic CFCs in individual EBs that persists when they are cultured under a variety of conditions. The observation of a high proportion of negative EBs and also some EBs with very few CFCs implies that there are both early and late bottlenecks still present in the differentiation conditions tested that limit the derivation of hematopoietic CFCs. Use of more optimized differentiation conditions led to an increase in both the proportion of EBs in which hematopoietic CFCs were produced at all, and in the total average number of CFCs generated per positive EB, showing that these changes in culture conditions promote multiple steps in the production of CFCs from parental ES/iPS cells. These data are consistent with a recent report using murine ES cells where single EBs were plated into adherent cultures with or without stromal cells320. Under these conditions, the authors found that not all EBs generated hematopoietic CFCs, and a high variation in the number of CFCs per EB was also detected. Both parameters were most increased in their system when the EBs were co-cultured with the AM20.1B4 cell line, which has also been reported to support high levels of hematopoietic differentiation from human ES cells139. The proportion of positive EBs was generally higher (20-90% depending on conditions) than what I observed with human ES cells (up to 30% depending on conditions). However, this is consistent with previous reports showing that up to 85% of murine ES cell-derived EBs contain CFCs and up to 1% of the total EB cells are CFCs148, a much greater efficiency of hematopoietic CFC generation than has thus far been obtainable from human ES cells. Interestingly, Ng et al. 141 have reported that it is possible to obtain hematopoietic cells from a very high proportion (>90%) of EBs generated from human ES cells. However, there are many differences between their system and the one described here. First, the lines they used were maintained on MEFs and were adapted to single cell passaging prior to differentiation. 96  Secondly, the EBs were generated by reaggregating single cells, not from clumps of undifferentiated ES cells. Thirdly, the EBs were cultured in serum-free medium with different cytokines, and were also replated into adherent conditions with a second cytokine cocktail. Lastly, hematopoietic output was measured by microscopic visualization of mature cells, and not by CFC output. It would be very interesting to combine these two approaches to determine if in these conditions, CFCs are also present in such a high proportion of EBs, or if mature hematopoietic cells are more readily generated and supported than CFCs using their conditions. I also observed that in all conditions tested, EBs that generated hematopoietic CFCs were, on average, larger than those that did not. However, this cannot be the only determining factor, as many EBs of similar size did not develop CFCs. Due to the complex morphology of EBs, one possibility is that there is not a direct relationship between EB diameter and cell number, the latter of which could be hypothesized to have a stronger correlation with CFC generation. Nevertheless, it does suggest that EBs within a certain size range may be most optimal for hematopoietic cell development from human ES/iPS cells. This is in agreement with reports that there is an optimal input cell number per EB for hematopoietic differentiation in reaggregated EBs141 and that larger human ES cell colonies are more optimal for mesoderm formation when cultured with BMP4 and activin A and subsequently form more hematopoietic cells when cultured in EBs321. Pretreatment of human ES/iPS cells with MEDII not only increased the proportion of EBs with hematopoietic activity, it also increased the number of CFCs in the positive EBs, even in the absence of cytokines. This suggests either that MEDII is generating a different type of mesodermal precursor that subsequently forms more CFCs in EBs, or that MEDII is generating more precursors that are not randomly distributed in the culture so that when clumps are generated, these undefined precursors tend to end up in a subset of EBs. The fact that the types of CFCs produced was altered in the presence of MEDII + cytokines might be argued as supporting the first explanation. But, since these data could only be analyzed in the presence of 97  cytokines, it is difficult to determine if this could be a reason for the difference. If MEDII generates more precursors but these are grouped together in the MEDII cultures, it would be expected that the hematopoietic output from these single EBs would not be clonal. Comparison with reaggregation EB methods141,142 where the starting cells are randomly segregated into EBs could be an interesting way to address this issue. The addition of cytokines was also shown to increase both the proportion of positive EBs and the CFC output per EB which suggests that cytokines can have a role early in the EB culture, in the generation of hematopoietic cells or precursors, as well as later, in the expansion of CFC numbers per EB. The types of CFCs produced under different conditions appeared to be mostly determined at the EB level, that is, EBs with the ability to generate a particular CFC type produced a similar number of those CFCs regardless of the conditions. This suggests that the precursors were altered rather than the expansion of particular CFC types in the different conditions. Moreover, multipotent EBs were observed regardless of the differentiation condition used, and these EBs tended to contain a higher total number of CFCs, raising the possibility of their origin from a common multipotent precursor with greater proliferative potential. The limiting number of positive EBs observed under these conditions is also suggestive of a clonal hematopoietic output within these EBs. Another possibility is that multiple precursors are generated, particularly in these multipotent EBs, however, more definitive answers to all of these questions require an approach to track clonal outputs of CFCs within EBs.  98  4. Tracking human ES and iPS cell-derived hematopoietic cell development using a genetic marking strategy  4.1 Introduction  Using both standard and improved conditions to induce human ES and iPS cells to differentiate into hematopoietic CFCs, I found the hematopoietic output of individual EBs to be highly heterogeneous as described in Chapter 3. This includes the finding that a large proportion of EBs do not produce any CFCs regardless of the conditions tested. This suggested that positive EBs might contain clonally amplified CFC populations indicative of a major persisting limitation in the generation and hence distribution of cells that are precursors to CFC among individual EBs. To determine more definitively if the output of hematopoietic CFCs in individual EBs is monoclonal or polyclonal, a method to track the clonal progeny of undifferentiated or MEDII treated cells is required. The most rigorous techniques to track the output of individual cells are those where the cells are either physically isolated individually, or imparted with a unique genetic tag, followed by the subsequent tracking of their isolated or uniquely marked progeny. Retroviral and lentiviral vectors can be used to introduce reporter genes that will allow the visualisation of populations of infected cells and their progeny. In addition, the semi-random insertion of these vectors into the DNA of the infected host cell can be used as a genetic tag with which to track its clonal progeny. Since the tag is determined by the integration site alone, one benefit for the human ES cell system is that the cells to be marked do not necessarily have to be first dissociated and isolated as single cells. When mouse ES cells are infected with a retrovirus, it has been found that transgene expression tends to be lower than in other cell lines, and is silenced over time and during 99  differentiation322,323, whereas silencing does not seem to be as much of a problem when lentiviral vectors are used323,324. Several groups have reported the use of lentiviral vectors to transduce human ES cells with variable efficiencies, from 20% to >80%, presumably due to differences in vector design and methodologies324-328. For example, one group showed the addition of a central polypurine tract (cPPT) increased transduction 3-fold and the woodchuck hepatitis virus post-transcriptional regulatory element (WPRE) increased gene expression 2fold325. In these studies, use of lentiviral vectors also allowed a more prolonged maintenance of transgene expression that persisted after differentiation induction, with the use of both constitutive (EF1α, PGK) and modified viral (MND) promoters. This implies that gradual silencing of a lentiviral transgene can be minimal in transduced human ES cells. Nevertheless, Xia et al. have also reported a phenomenon they called “suppression,” where lentiviral vectors are immediately and permanently silenced when introduced into human ES cells in a promoterdependent manner, resulting in an apparent difference in transduction efficiency between lentiviruses that are identical apart from the internal promoter that they contain329. The objective of the studies described in this chapter was to develop and use a lentiviral system to determine the clonality of hematopoietic CFC outputs for individual EBs. Such a system requires that both human ES and iPS cells can be efficiently transduced and maintain high levels of transgene expression in both undifferentiated cells and their hematopoietic progeny. To achieve this, we screened available lentiviral constructs with fluorescent reporters under the control of the phosphoglycerate kinase (PGK) promoter or the spleen focus forming virus (SFFV) promoter. The former promoter was previously shown to be effective in human ES cells by others325,329, and the latter was previously reported to allow high levels of expression in hematopoietic cells330, including HSCs331, and was more recently used in mesenchymal stem cells and human ES cells332. A lentivirus with sufficiently high transduction efficiency and gene expression was then used to track the hematopoietic output within individual EBs to further investigate the data presented in Chapter 3. 100  4.2 Materials and methods  4.2.1 Human pluripotent stem cell culture and differentiation  Cells were cultured under feeder-free conditions with the exception of a few early infection experiments in section 4.3.1 which used cells cultured on MEFs. These methods, plus those used for EB culture, dissociation, and analysis, including the methods used for harvesting single EBs and determining their individual content of CFCs, were as described in detail in Chapters 2 and 3. H1 or CA1 ES cells, or MSC iPS cells were used for all experiments described in this chapter.  4.2.2 Lentiviral vectors  The pRRL.PPT.SF.IRES.YFP vector was a gift from T. Schroeder (Helmholtz Zentrum Institute of Stem Cell Research, Munich, Germany) and was used to generate the pRRL.PPT.SF.IRES.VENUS (SiV, Figure 4.1a) and pRRL.PPT.SF.IRES.TOM (SiTom) lentiviral vectors by replacement of YFP with the VENUS or tandem dimer (td)TOMATO fluorescent protein sequences, respectively. The VENUS and tdTOMATO proteins used are localized to the nuclear membrane by inclusion of the human importin a1 sequence at the 3’ end of the fluorescent protein in these vectors and are under the control of the SFFV promoter. The pCCL.PPT.MND.PGK.EGFP lentiviral vector333,334 (MPG, Figure 4.1b) was provided by D. Kohn (UCLA, Los Angeles, CA, USA) and expresses enhanced green fluorescent protein (EGFP) under the control of the human PGK promoter.  101  Figure 22 Figure 4.1  A 5’ LTR  3’ LTR-SIN  cPPT  PBS  RSV R U5 ψ ΔGAG  SFFV  RRE  IRES  VENUS  WPRE  ΔU3RU5  B 5’ LTR  CMV  PBS R U5 ψ ΔGAG  3' LTR-SIN  cPPT RRE  MND  PGK  EGFP  ΔU3RU5  Figure 4.1 – Schematic of lentiviral vectors Schematic representations of the SiV (A) and MPG (B) lentiviral vectors evaluated here. Both vectors are HIV-based and self-inactivating (SIN), and share common features including the HIV-1 primer binding site (PBS), packaging signal (ψ), minimal GAG region (ΔGAG), and Rev responsive element (RRE), as well as a cPPT to increase transduction efficiency. A) Unique features of the SiV vector include the replacement of the HIV U3 region of the 5’ LTR with the promoter and enhancer region from the U3 region of the Rous Sarcoma Virus (RSV) and the inclusion of a WPRE sequence. An internal SFFV promoter is used to drive transgene expression, and the VENUS fluorescent protein is included downstream of an internal ribosome entry site (IRES). The SiTom vector is identical except for the replacement of VENUS with the tdTOMATO fluorescent protein. B) Unique features of the MPG vector include the replacement of the HIV U3 region of the 5’ LTR with the promoter and enhancer region from the U3 region of the human cytomegalovirus (CMV) and the presence of an internal promoter to drive transgene expression, modified from the U3 region of the myeloproliferative sarcoma virus (MPSV) and called MND for MPSV enhancer, Negative control region deleted, dl587rev primer-binding site substituted. A separate internal PGK promoter drives the expression of enhanced green fluorescent protein (EGFP) and a post-transcriptional regulatory element is not included.  102  4.2.3 Production of lentivirus  Virus was produced by standard calcium phosphate transfection of 293T cells in 10 cm tissue culture dishes with 4 plasmids: lentiviral transfer vector (10 μg), ΔR (6.5 μg) and REV (2.5 μg) packaging constructs, and vesicular stomatitis virus glycoprotein (VSV-G) envelope (3.5 μg), in serum-containing medium. Packaging and envelope plasmids were gifts from P. Leboulch (Harvard University, Boston, MA, USA). Viral supernatant was collected on 2 consecutive days and filtered with a 0.45 μm low protein binding filter prior to concentration by 2 rounds of ultracentrifugation for 90 minutes at 100,000 g and storage at -80°C. Viral titres were determined by infection of Hela cells with serial dilutions of concentrated virus, and only virus with a titre of ≥1x109 U/mL was used in the experiments described here.  4.2.4 Lentiviral transduction of human ES and iPS cells  Clumps of cells were harvested as for undifferentiated or EB culture, depending on the intended use of the infected cells, and then resuspended in mTeSR or EB medium, respectively. The medium also contained 10 μM Rock inhibitor and 5 μg/mL protamine sulphate. Final viral concentrations used ranged from approximately 0.5-5x107 U/mL for SiV and SiTom and 0.51x107 U/mL for MPG. Clumps from 0.5-1 x 60 mm dish of human ES or iPS cells (~1-4x106 cells) were infected in a volume of 400 μL in a 16 mm well. Virus was mixed with the resuspended clumps of cells and incubated at 37°C for 5-6 hours. Clumps were then removed from the well and washed once with 15 mL DMEM to remove excess viral particles. Cells were then either suspended in mTeSR and plated onto Matrigel-coated dishes for propagation as undifferentiated cells, or they were suspended in EB medium with 10 μM Rock inhibitor ± cytokines and plated into ultra low adherence wells for EB culture. In these studies, the infection  103  efficiency was defined as the proportion of fluorescent cells detectable by flow cytometry, and was assessed at the earliest timepoint where cells were collected for analysis after infection.  4.2.5 Analysis of marked EB-derived hematopoietic CFCs  At the end of the EB culture, the presence of GFP+ cells in EBs was assessed using a Leica MZFLIII fluorescent microscope. Pooled or individual EBs were then dissociated into a single cell suspension and plated into hematopoietic CFC Methocult assays as previously described (see Chapters 2 and 3). Pooled EBs were also analyzed by flow cytometry to assess their content of GFP+ cells. GFP was sometimes assessed in combination with antibody staining for hematopoietic markers using the flow cytometry protocol described in Chapter 2. Two weeks after plating, CFC-derived colonies were enumerated and typed on a light microscope, and the proportion of GFP+ and GFP- colonies determined using a fluorescent microscope.  4.3 Results  4.3.1 Choice of lentiviral vector affects transduction efficiency and level of gene expression in human ES and iPS cells  Traditional infection protocols typically mix single cell suspensions or adherent monolayers of cells with viral particles at defined concentrations to allow optimal interaction between cells and virus, achieving maximal infection with minimal toxicity. Due to the poor survival of single undifferentiated human ES or iPS cells, the first strategy is not conducive to obtaining useful numbers of genetically modified human ES or iPS cells. Furthermore, single cells cannot typically be used to initiate EB differentiation experiments, limiting the types of experiments that could be carried out using this type of infection protocol. As an alternative, I 104  found that infecting adherent ES cells 1-2 days after passaging allowed efficient gene transfer. However, this required either relatively large volumes of virus, or was limited to very low cell numbers, and when cells were grown on MEFs, the MEFs were also highly infected with the virus (data not shown). In seeking other approaches, I found that the most efficient and economical protocol was to infect clumps of human ES or iPS cells at a high density for a short time (several hours), followed by their transfer into typical maintenance or differentiation conditions. This protocol may result in a somewhat decreased infection efficiency because access of the virus to cells inside the clumps is limited. However, I found I could use high concentrations of cells and virus and recover large numbers of infected cells using this protocol. Since viruses have different infection efficiencies and different promoters have varying activities in different cell types, I first sought to identify a virus-promoter combination that allowed high expression not only in undifferentiated cells, but that would retain reporter expression throughout hematopoietic differentiation. For this, I compared the MPG virus (that contains the constitutively active human PGK promoter) to the SiV and SiTom constructs (that contain the virus-based SFFV promoter). These vectors also contain several other differences as shown in Figure 4.1. When undifferentiated cells were infected and examined for expression of the encoded fluorescent reporter genes between 3 and 10 days later, cells infected with the MPG virus contained 10-fold more cells with detectable fluorescence, showing a much greater infection efficiency than was obtained using the SiV or SiTom viruses (Figure 4.2a). In addition, the PGK promoter gave a much higher level of fluorescence within the positive cells than the SFFV promoter suggesting that it was also driving a higher level of reporter gene expression (Figure 4.2b and c), since the GFP in the MPG construct is the least bright of the fluorescent proteins used in the tested constructs335. These data imply that the lentivirus construct used for the SiV and SiTom vectors either infects human ES cells much less efficiently, or it is more affected by additional factors that influence our definition of infection efficiency. These could include decreased expression from the SFFV promoter or IRES sequence, increased 105  Figure 23 Figure 4.2  % positive cells  A  ***  80 60 40 20 0  SFFV  PGK  Viral promoter  B  C  CA1  H1  H1  MSC  Figure 4.2 – Increased reporter gene expression in human ES and iPS cells using the PGK promoter A) Percentage of cells positive for fluorescent reporter expression when the reporter was under the control of the SFFV or the PGK promoter. After infection of undifferentiated cells, fluorescence was measured between 3 and 10 days later (n=8, SFFV; n=8, PGK). B) Representative FACS profiles for H1 or CA1 cells infected with SiV lentivirus and cultured for 8 days in undifferentiated human ES cell maintenance culture conditions. C) Representative FACS profiles for H1 or MSC cells infected with MPG lentivirus and cultured for 5 days (H1) or 10 days (MSC) in undifferentiated human ES/iPS cell maintenance culture conditions.  106  suppression or silencing soon after integration, or a decreased copy number per cell compared to the MPG vector. As both the proportion of cells expressing transgene and the strength of transgene expression were clearly superior, the MPG lentiviral construct was used for all further experiments.  4.3.2 Human ES and iPS cells can be marked by lentivirus at multiple points of their differentiation into hematopoietic CFCs  To track the output of precursor cell types from various points of differentiation, the virus used must efficiently infect not only undifferentiated cells but also their differentiating derivatives. I thus compared the ability of the MPG vector to infect clumps of undifferentiated cells and MEDII-pretreated cells just prior to their use to initiate EB formation. When the infection efficiency was assessed at the end of the EB culture period, the cells that were infected at the undifferentiated stage were the most efficiently marked, followed by cells infected at the end of the MEDII treatment (Figure 4.3a). When EBs were transduced even later in their course of differentiation, at day 12 of EB culture, a lower infection efficiency was obtained, likely because the virus could only access the outermost cell layers of the EB. Also, there was no longer a difference between cells in EBs derived from MEDII-pretreated vs. untreated ES cells, likely reflecting a similar overall level of differentiation and cell types in the outer layer of the EBs at this timepoint (Figure 4.3b). The large proportion of highly GFP+ cells obtained at the end of the EB culture period shows that the PGK promoter can continue to drive high levels of expression throughout the differentiation process.  107  Figure 24 Figure 4.3  A  Undifferentiated  MEDII pretreated  H1  MSC  B  MSC EBs  Figure 4.3 – Cells are infected with varying efficiency throughout differentiation A) H1 ES cells (top row) or MSC iPS cells (bottom row) were cultured in either mTeSR (undifferentiated, left column) or 50% MEDII (MEDII-pretreated, right column) for 7-10 days prior to infection of clumps with MPG lentivirus. After infection, cells were transferred to EB conditions and assessed for infection efficiency at the end of the EB differentiation culture period (15-20 days). Shown are representative paired experiments, illustrating that EBs derived from infected undifferentiated ES/iPS cells consistently show a higher proportion of GFP+ cells. Gates are set relative to untransduced controls from the same culture conditions. B) Whole EBs were infected with lentivirus at day 12 of differentiation and analyzed for the percentage of GFP+ cells at day 20 of EB differentiation. In this case, no difference in infection efficiency was observed between EBs that had been derived from MSC iPS cells that had (right) or had not (left) been pretreated with MEDII.  108  4.3.3 Precursors of hematopoietic cells are marked by infection of undifferentiated and MEDII-pretreated ES cells  EB cells were next examined for the presence of marked hematopoietic cells and CFCs. When CD34 or CD45 positive cells were detected in dissociated EBs, these populations also included GFP+ cells (Figure 4.4). CFC assays performed on these cells also demonstrated the formation of GFP+ colonies, indicating that precursors of CFCs can be targeted. Interestingly, the presence of CD45+ cells did not predict the presence of CFCs, nor did the presence of GFP+CD45+ cells predict the presence of GFP+ CFCs. In the experiment shown in Figure 4.4, for instance, a clear population of CD45+ cells was detected only in EBs derived from MEDIIpretreated iPS cells, but the CFC frequency was relatively similar for EB cells derived from untreated and MEDII-pretreated iPS-derived EBs (10.6 and 14.2 per 105 EB cells, respectively). Furthermore, all of the CFCs detected in the EBs derived from the untreated iPS cells were GFP+ despite the absence of detectable GFP+CD45+ cells, and none of the CFCs detected in the EBs derived from MEDII-pretreated iPS cells were GFP+, whereas 30% of the CD45+ cells were GFP+. Such a discrepancy between the proportion of GFP+ CFCs and CD45+ cells in EBs might be expected if only a few precursors with variable output potential contribute to the hematopoietic output, as was suggested by the data in Chapter 3. To determine if there was any correlation between the percentage of GFP+ EB cells and GFP+ CFCs, these values were compared in 10 experiments (Figure 4.5). CFC-derived GFP+ colonies were observed in most experiments, showing again that precursors of CFCs are marked by the lentivirus and also maintain expression throughout differentiation, but no significant correlation was observed with the percent GFP+ total EB cells. Unexpectedly, the number of GFP+ CFCs was almost always lower than the overall percentage of GFP+ EB cells. Variable CFC contributions by a few precursors should lead to a wide scatter in these values, but an approximately equal number of experiments with larger and smaller proportions of GFP+ 109  Figure 25 Figure 4.4  Undifferentiated  MEDII pretreated  Figure 4.4 – Reporter gene expression is maintained throughout differentiation into hematopoietic cells Differentiated cells were examined for the expression of GFP in cells expressing hematopoietic markers at the end of EB culture. MSC cells from Figure 4.3a were shown to contain some GFP+ CD34+ cells, and when CD45+ cells could be detected, as in the EBs derived from MEDIIpretreated iPS cells in this example, some GFP+ CD45+ cells were also observed.  110  % GFP+ CFCs  Figure 26 Figure 4.5  100  No pretreatment  80  MEDII pretreated  60 y=1.07x-16 p=0.16 r 2=0.23  40 20 0 0  20 40 60 80 100  % GFP+ EB cells Figure 4.5 – Discrepancy between the proportion of GFP+ EB cells and GFP+ CFCs Clumps of human ES or iPS cells with or without MEDII pretreatment were infected with lentivirus and put into EB cultures to induce hematopoietic differentiation. At the end of the EB culture period, the proportion of total EB cells that were GFP+ was determined by FACS. Cells were also plated into CFC assays. After 2 weeks, resulting colonies were examined for GFP fluorescence under a fluorescent microscope to determine the proportion of GFP+ CFCs. The dotted line indicates where equal proportions of GFP+ total cells and CFCs would be detected. The solid line shows the linear regression line, with the linear equation, p-value (that the slope differs from zero), and the r2 value indicated to the right of the graph. No significant correlation is observed (p>0.05).  111  CFCs than GFP+ EB cells. However, the fact that there are consistently fewer GFP+ CFCs implies either that precursors of CFCs are being less efficiently infected than precursors going on to form other types of cells, or that precursors of CFCs are more likely to silence the GFP lentivirus than cells undergoing other types of differentiation.  4.3.4 Lentiviral marking reveals that the hematopoietic output of a single EB is not necessarily from a single MEDII-generated precursor cell  In Chapter 3, I described substantial heterogeneity in the hematopoietic output between individual EBs and observed that most EBs showed no hematopoietic activity, implying a limit on either the generation or differentiation of a precursor of CFCs. Despite the fact that EBs are not clonal entities, this type of distribution suggested that the hematopoietic output of a single EB might in fact be clonal. To address this question, I infected clumps of MEDII-treated ES/iPS cells with MPG lentivirus immediately prior to placing the cells in an EB culture with cytokines. This produced a mixture of GFP+ and GFP- cells within each clump/EB. Following differentiation, a mixture of positive and negative cells was maintained within individual EBs, as determined by visualization of the EBs under a fluorescent microscope. The CFCs from these individual EBs were then analyzed to determine if the CFC output was clonal, as would be concluded if all colonies generated from a single EB were either GFP+ or GFP-. Alternatively, a polyclonal origin of the CFCs in individual EBs would be indicated by the generation of both GFP+ and GFP- colonies. In 4 experiments (3 with MSC, 1 with H1), 142 EBs were analyzed individually for CFCs, and pooled EBs were also analyzed to determine the infection efficiency, which ranged from 2049%. The distribution of CFCs amongst the individual EBs was similar to what had been seen in other experiments using MEDII-pretreated cells and cytokines in the EB cultures (Figure 112  4.6a). Of the single EBs analyzed, 35 contained >1 CFC and could be assessed for a mixture of GFP+ and GFP- CFCs. Of these 35 EBs; 6 contained only GFP+ CFCs, 16 contained only GFPCFCs, and 13 contained both GFP+ and GFP- CFCs (Figure 4.6b). There were also 19 EBs that contained a single CFC, which must have originated from a single precursor, of which 5 were GFP+ and 14 were GFP-, which is consistent with the estimated infection efficiency. Some of the GFP+ and GFP- only EBs contained high numbers of CFCs; however, the EBs with both GFP+ and GFP- CFCs tended to more often contain higher numbers of CFCs. If CFCs are marked at a rate equal to or less than the infection efficiency, as suggested in Figure 4.5, fewer than half of cells that give rise to CFCs should be marked. Therefore, EBs with only GFP+ CFCs are likely to be clonal, explaining why they have lower numbers of CFCs. EBs with only GFP- CFCs are more likely to contain more than one unmarked clone, as most precursors are expected to be unmarked, and therefore are also more likely to have some EBs with greater CFC numbers. In EBs with GFP+ and GFP- CFCs, it is expected that there are at least 2 clones that give rise to CFCs, and these show correspondingly greater numbers of CFCs. If there were a large number of contributing clones, this should bring the proportion of GFP+ CFCs within an EB close to equalling the proportion of total GFP+ cells, as is the case with all EBs containing a single CFC. However, within the EBs containing both GFP+ and GFP- CFCs, there was a large variation in the proportion of GFP+ CFCs. This suggests that in these multiclonal EBs, there are still relatively few clones that contribute (Figure 4.6b). In addition, such high numbers of EBs containing only GFP+ or GFP- CFCs would not be predicted if large numbers of clones were each contributing limited outputs of CFCs. It was not possible to perform integration site analysis on the colonies obtained, so the precise number of contributing precursors is unknown. Overall, these data suggest that the CFC output of an individual EB is sometimes clonal and sometimes polyclonal, but there are relatively few contributing clones within an EB, and thus each precursor can expand and generate multiple, but not very large numbers of CFCs.  113  Figure 27 Figure 4.6  100  15  40  50 25  0  0  0 1 2 3-4 5-8 9-16 17-32 33-64  20  % EBs  75  15  10  10 5  # EBs  100 60  # EBs  % EBs  80  5 0  # CFC/EB  B  20  125  1 2 3-4 5-8 9-16 17-32 33-64  A  0  # CFC/EB  100  # CFC  75  17  11 74 64 18 71 6 25 67 75 33 50  10  G FP +  on ly  G FP -o nl y G FP + an d -  1  EBs with CFCs that are:  Figure 4.6 – Tracking of marked hematopoietic progenitors within single EBs A) Distribution of CFCs amongst individual EBs derived from MEDII-pretreated cells exposed to the MPG lentivirus at the time of EB formation. The number of CFCs observed in an individual EB is shown on the x-axis, binned on a log2 scale and the number of EBs that fell into each bin is indicated on the y-axis (n=142 in 4 experiments). The right-hand graph shows the same data with the EBs containing zero CFCs removed so that the bars for the 54 positive EBs can be better visualized. B) Distribution of GFP+ and GFP- CFCs in individually analyzed EBs. Each of the 54 EBs positive for CFCs in (A) was categorized into one of 3 groups: only GFP+ CFCs, only GFPCFCs, or both GFP+ and GFP- CFCs detected within the EB. Each dot represents an individual EB and shows the number of CFCs within that EB. Of the 35 EBs containing greater than 1 CFC, 13 contained both GFP+ and GFP- CFCs. For these 13 EBs, the percentage of CFCs that were GFP+ is shown to the right of the graph in a position corresponding to the point that represents the individual EB. 114  4.4 Discussion  The high infection efficiency achieved with the MPG lentivirus (>80% of undifferentiated cells in some experiments), will make it useful for further marking or gene manipulation studies. In addition, the PGK promoter was shown to allow very strong and persistent gene expression in both undifferentiated as well as more differentiated cell types. Although the SFFV promoter has been reported to sustain high levels of expression in human hematopoietic cells331, and has also been used in human ES cells332, I did not find it gave strong reporter gene expression, demonstrating the need to validate both promoter choice and the viral construct for transducing the cell type(s) of interest. It is not clear which features of the SiV and SiTom lentiviruses are responsible for their apparent lower infection efficiency and expression when compared to MPG. Further studies to validate the correlation between transgene expression, used here as a readout of infection efficiency, and the frequency of transgene integration could help to clarify this issue. These could include qPCR analysis of viral sequences in genomic DNA or clonal PCR analyses if human ES clones could be obtained more readily after infection. The ability to mark cells efficiently at multiple points of differentiation will also be useful for further marking studies, or so that genes with detrimental or undesired effects on undifferentiated cells can be introduced directly into a more differentiated cell type. Despite possible restriction of cell-virus interaction due to the cells being in clumps, a high infection efficiency was observed, therefore, it is possible that even higher efficiencies could be consistently achieved if the protocol could be successfully modified to infect single cells, for example, in the presence of Rock inhibitor, followed by reaggregation. Robust GFP expression is retained in differentiated EB cells, including CD45+ cells generated within these EBs, and mature cells in CFC-derived colonies, showing that reporter expression can be used to track the progeny of the initial group of precursors marked during EB formation. Data collected from individual EBs showing a mix of GFP+ and GFP- CFCs highly 115  suggest that multiple clones can contribute to hematopoietic differentiation within an EB, although silencing of the lentivirus in some of the marked progeny must still be ruled out. Attempts to determine the presence or absence of viral integration in GFP- colonies were inconclusive. However, if one assumed that silencing was entirely responsible for the presence of the GFP- CFCs in EBs that also contained GFP+ CFCs, then one would conclude from the analysis of EBs with >1 CFC that 54% contained clonal GFP+ CFC precursors. This is inconsistent with both data from EBs with only 1 CFC where this value was 26% as well as the average infection efficiency of 35%. Therefore, the simplest and most compelling conclusion is that the CFCs generated in individual EBs represent the progeny of a very limited number of clones contributing to hematopoiesis, suggesting that the differentiation process is still quite restricted. However, this could be examined further by marking cells with a mixture of lentivirus vectors containing distinct fluorescent markers so that 4 populations of cells could be followed: each population of single positive, double positive, and negative cells. There is an apparent discordance between the low proportion of EBs that contain any CFCs which suggests a limited generation of precursors, and the concomitant presence of both GFP+ and GFP- precursors within these positive EBs. This could imply that the limit is not on the production of hematopoietic precursors, but rather on their continued differentiation, where some EBs provide more favourable conditions through their organisation, production of specific supportive cell types, or other unknown factors. Alternatively, since EBs are generated from clumps, there may be bias in the segregation of cell types at the start of differentiation based on their physical location near each other in the starting culture. It would therefore be interesting to see if CFCs are more evenly distributed in EBs derived from reaggregated single cells that are randomly disseminated during EB formation. Tracking the generation of different types of CFCs to determine if they share a common precursor would also be of interest. Unfortunately, this could not be analyzed in this set of experiments as there was not a sufficient variety of different colony types distributed amongst 116  the EBs (83% of positive EBs contained only CFU-GM). Only 9 EBs contained multiple types of CFCs, but of these, 6 did contain both GFP+ and GFP- CFCs. Data presented here and in Chapter 3 showed that EBs containing multiple CFC subtypes and also EBs with GFP+ and GFP- CFCs contained greater numbers of CFCs. Together, these data support the hypothesis that multipotent EBs contain higher numbers of CFCs because they contain multiple precursors that contribute to hematopoiesis. On the other hand, these data also suggest that a single precursor could produce multiple types of CFCs as in each of these 6 multipotent EBs with both GFP+ and GFP- CFCs, >1 CFC subtype was found within either the GFP+ or GFP- fraction. In addition, 2 EBs with only GFP+ CFCs and 1 EB with only GFP- CFCs contained multiple CFC types. However, without integration site analysis, it is not certain that these shared a common precursor. The data also suggest that various precursors are produced that may contribute to hematopoietic output at different times. This was evident from the finding that some EBs containing GFP+ CD45+ cells contained only GFP- CFCs, implying that a GFP+ precursor had produced hematopoietic cells that had all already fully differentiated to mature cells whereas at the same timepoint, one or more GFP- precursors produced CFCs but had not yet generated detectable numbers of mature cells. Overall in this system, few human ES or iPS cells appear to contribute significantly to the output of functional hematopoietic cells, at least not in a synchronous manner. In order to scale up these differentiation processes and produce primitive hematopoietic cells for study or other use, these conditions need to be improved by further optimizing each of the multiple differentiation steps involved. The lentiviral transduction methods described here offer a powerful approach to introducing a variety of key regulators of hematopoiesis into these cells with the goal of improving hematopoietic output and generating hematopoietic stem cells. The protocols described here also provide an important methodology to modify, monitor and prospectively track the generation of early hematopoietic cells from more primitive precursors.  117  5. Discussion and future directions  The studies described in this thesis were focussed on understanding the process by which hematopoietic cells are generated from human pluripotent stem cells. This was driven by the continuing expectation that a better understanding of the sequence of events required for the development of human hematopoietic cells from the most undifferentiated cells that appear during embryogenesis will be useful for future applications in biology and medicine. These include the production of mature blood cells and hematopoietic stem cells for use in patients, as well as the creation of disease models. Much work using mouse ES cells over the past 2 decades suggests that their manipulation in vitro can faithfully replicate many of the steps that occur during the in vivo development of the murine hematopoietic system, and this model has proven to be very useful for examining putative hematopoietic precursors, and for testing the role of many genes in the generation and later regulation of hematopoiesis. Development and validation of the equivalent system using human cells is in the early stages, and still faces many challenges. These pertain to a continuing lack of understanding of the earliest steps in mesoderm formation and differentiation as well as a persistent inability to generate hematopoietic cells as efficiently and reproducibly as occurs in vivo. The first hematopoietic differentiation protocols for human cells consisted of a single step; cells from cultures of undifferentiated ES cells were exposed to a collection of undefined factors present in FBS and/or feeder cells of various types136,137,143. The expectation was that these conditions would provide all of the cues required to initiate the multiple steps of development likely to be required. At that time, human ES cells were maintained as undifferentiated cells by culturing them on MEFs, although it was already appreciated at the time that both of these aspects of the original protocols in place had significant disadvantages for future work. Since those initial studies, many groups have been analyzing this process and have contributed to identifying particular factors or treatments required at different stages so 118  these might be separately optimized (reviewed in Chapter 1). The experiments described in this thesis add to this accumulating body of data demonstrating how the use of multi-step protocols optimized for the progression of each individual developmental stage can improve the final output of hematopoietic cells from a relatively homogeneous population of highly undifferentiated human pluripotent stem cells maintained in a relatively defined feeder-free system. Adoption of this approach then set the stage for further studies analysing the clonal variability seen in this process.  5.1 Major contributions  A first goal was to exploit emerging opportunities to start with a more biologically homogeneous population. Initially this was achieved using a selective passaging protocol299. Later, this methodology was combined with the adoption of more defined and reproducible conditions that became available in 200616. In experiments using cells maintained with these various methods, I found that the protocols used for maintaining the ES cell cultures had important downstream consequences on the yield of hematopoietic cells later obtained. This highlighted the importance of optimizing individual steps of differentiation, starting with the earliest transitions from undifferentiated cells to mesoderm. A further influence came from the demonstration from our group that that MEDII contains factors able to selectively promote ES cells to differentiate into mesoderm which, in turn, resulted in an increased output of hematopoietic cells from subsequently formed EBs155. My studies made use of these developments to show, using a reporter cell line291, that MEDII may not be directly inducing mesoderm formation. Instead they suggest that MEDII treatment of undifferentiated human ES cells can prime them for a mesodermal fate upon subsequent transfer to conditions that more strongly stimulate their differentiation. I also determined that MEDII can be used in combination with other manipulations, such as the addition of various hematopoietic cytokines in subsequent 119  EB cultures, to further increase and make more consistent the hematopoietic cell outputs obtained. The studies in Chapter 2, reinforced by those reported by others in the field underscore the significant variability seen when the hematopoietic output of ES or iPS cells in different experiments is compared, even when the same conditions are used. To investigate possible explanations for this variation, I developed a method to examine the hematopoietic output of large numbers of individual EBs and I then used it to examine EBs generated under similar conditions from different sources of pluripotent cells. The results are detailed in Chapter 3. These experiments revealed a previously unappreciated heterogeneity in the behaviour of cells between individual EBs which contributes to a large variation in the ability of individual EBs to generate hematopoietic CFCs. Specifically, the key findings were first, that at a given timepoint during the EB differentiation process, the majority of EBs contained no hematopoietic CFCs regardless of the culture conditions used; and secondly, that within those EBs that did contain at least one hematopoietic CFC, there was a large variation in the number of CFCs generated. Nevertheless, the enhancing effects of the more “optimal” conditions described for studies of bulk cultures in Chapter 2, were also evident in the studies of individual EBs. These were manifested both in terms of an increase in the proportion of EBs found to contain hematopoietic CFCs and in the number of CFCs per positive EB. However, the fact that the majority of EBs, even under these improved conditions, contained no CFCs, suggests that further optimisation is needed. These studies represent one of the first attempts to develop a lineage tracing approach to the analysis of human hematopoietic cell generation from mesodermal cells. At the same time, it was clear from the outset that human EBs as generated in these studies, are not clones. Thus, in spite of the limited number of positive EBs seen in the experiments performed in Chapter 3, it was not possible to ascribe variation in the output of positive EBs to a variability in the behaviour of a single original “hematopoiesis-generating cell”. 120  To address this issue, I established and then incorporated the use of a lentiviral marking strategy. Lentiviral marking is now well recognized as a powerful tool for tracking and manipulating precursor cells in a variety of settings because of their ability to integrate into thousands of sites in the genome in a semi-random fashion. Previous reports had shown that human ES cells were susceptible to infection using lentiviral vectors, and that this could be achieved at high efficiency in some cases324-328. As described in Chapter 4, I found that the choice of a suitable lentiviral vector and infection protocols allowed both undifferentiated and early differentiating (MEDII treated) ES/iPS cells to be transduced at high efficiency with continued expression of the introduced transgene. Using this strategy, I was then able to assess the output of individual MEDII-treated cells within individual EBs. These experiments indicated that the limited frequency of EBs containing hematopoietic CFCs was not necessarily indicative of a clonal origin of those CFCs detected within the subset of EBs that were positive. Together, this work illustrates how separation and controlled analysis of different variables affecting the output of hematopoietic cells from human ES and iPS cells can lead to new insights of the cellular steps involved and their optimization. They also serve to underscore how much information is still required to achieve the desired goals of efficiency and reproducibility suggested by the development of human hematopoietic tissues in vivo. In this regard, further characterization of the individual developmental steps identified here should facilitate these investigations.  5.2 Implications and future directions  5.2.1 Further characterization of the starting cell population  Much work has been done to show that many ES and iPS cells within a culture share similar phenotypes. Similarly, at a population level, different cultures of ES and iPS cells appear 121  reasonably comparable. Nevertheless, this may not indicate functional similarities, and some evidence of functional differences in the differentiation abilities of different lines has been reported99. It has also been suggested that the hematopoietic potential of mouse ES cells is sensitive to the maintenance conditions used336 and it has been recently shown that murine iPS cells from different somatic cell sources may have different differentiation abilities101,102. However, little is yet known about how ES/iPS cells within a culture optimized to maintain their undifferentiated state compare to each other at a molecular or functional level. If functionally distinct subtypes of cells exist within these cultures, then changes in the types of cells present and the proportions with which they are represented could influence the propensity to obtain particular types of differentiating progeny. For example, only a small fraction of human ES cells read out in the functional ES-CFC assay, even when methods such as the addition of Rock inhibitor are used to increase their viability10,11,26,337. Also, when transcript levels in individual human ES cells were examined for pluripotency and early lineage markers, large heterogeneity was observed, including the presence of pluripotency and lineage transcripts in the same cells338. The results presented in Chapter 2 support this idea, where manipulation of the maintenance conditions of the starting ES/iPS cells had profound downstream effects on the output of hematopoietic cells obtained using the same differentiation conditions. It is therefore possible that some of the variability in hematopoietic CFC output that I found to be exhibited by different EBs (Chapters 3 and 4) might be explained by heterogeneity in the initial ES/iPS starting cell populations, even though considerable effort had already been undertaken to reduce the content of spontaneously differentiating cells. Therefore, to make further progress, it may be critical to develop more rigorous ways of characterizing and separating or marking the starting undifferentiated ES/iPS cell populations to be used. Defining the characteristics of undifferentiated cells with the functional ability to generate a differentiated cell type of interest under a given set of conditions would be an important next goal. More efficient single cell differentiation protocols may facilitate such 122  advances by allowing purification of populations of interest immediately prior to differentiation. Recently, it has been suggested that the hematopoietic differentiation ability of human ES/iPS cell lines can be correlated with a gene expression profile including higher levels of transcripts involved in nodal/activin signalling100. If such characteristics could be identified, concurrent development of culture conditions that would allow tighter control over the expression of these characteristics might also prove of great benefit. Identification of specific factors or combinations of factors that predispose ES/iPS cells to a given germ layer (for example those active in MEDII) would similarly represent an additional approach to enhancing the production of mesodermal derivatives that include hematopoietic cells. Such information would also be of interest in terms of offering potential new insights into the mechanisms by which lineage choices are made in the very early human embryo.  5.2.2 Use of multi-step protocols to better direct differentiation  The results presented in Chapter 2 demonstrated that optimization of different steps of differentiation allowed both the efficiency and reproducibility of hematopoietic output to be improved. A more detailed analysis of these effects in Chapter 3 suggested that the improvements in hematopoietic CFCs obtained were indeed achieved by increases in the efficiency of induced differentiation at both early stages and later stages of differentiation (demonstrated by an increased proportion of positive EBs, and an increased number of CFCs per positive EB, respectively). Multi-step protocols have now been described for the development of 2 other commonly studied cell types that ES and iPS cells have been used to generate, spinal motor neurons339 and pancreatic β-cells (reviewed in Van Hoof340). The multi-step protocols now in place for each allow the generation of mature cells that can survive and engraft in xenotransplants and demonstrate at least some in vivo function in the case of the β-cells341,342. Another group has 123  used an initial step similar to that used for β-cell generation to produce precursors of definitive endoderm with the ultimate production of 3-dimensional intestinal epithelium organoids343. If analogous protocols could be derived for producing mesodermal and hematopoietic cells, this might further improve differentiation efficiencies. Ideally, optimization in multi-step protocols would be achieved using defined media components to replace undefined factors such as conditioned media (e.g. MEDII), feeder cells, and sera that are still used in many current protocols. An important step forward in this regard is the recent report of a defined animal product-free medium for producing hematopoietic cells from human ES/iPS cells344. An increased repertoire of human ES reporter cell lines for key transcription factors that define early developmental lineages would greatly accelerate the ability to monitor the efficiency and synchrony of differentiation and thus allow further optimization of each developmental step. These have been extremely challenging to produce, but are gradually becoming available. The MIXL1 reporter line derived by Davis et al. 291 and used in Chapter 2 is one such example. As has been demonstrated in the murine system, transcription factor-based reporter ES lines could also be valuable for the phenotypic identification and purification of different functional precursors to hematopoietic cells. If protocols for human ES/iPS cell differentiation could be modified to allow for the separation of intermediate cell types, followed by continued survival and differentiation, similar experiments could then be performed with human cells. Even in the absence of the ability to physically isolate these cells, such cell lines could aid greatly in better defining the efficiency of each step of differentiation, which should then lead to improvements in the overall efficiency as well as the purity of the final cell products.  5.2.3 Monitoring the clonal output of differentiating ES/iPS cells  As discussed in section 5.2.1, most human cells in an undifferentiated ES/iPS cell maintenance culture appear similar using currently available characterization methods. 124  Nevertheless, very little is known about their individual biological (developmental) homogeneity in terms of permanent or transient variability in their potential to generate particular types of differentiated progeny. For example, in the teratoma assay, millions of cells are usually injected thus precluding any assessment of the potentialities of individual cells in the population injected even though the teratomas as well as the specific differentiated lineages generated have been shown to be polyclonal345. Similarly, the starting cell population for formation of a single EB consists of hundreds to thousands of cells, but it is unknown how many of these cells survive, and then go on to contribute to the differentiation of a particular cell type subsequently looked for, e.g., a hematopoietic CFC. In Chapter 4, I show that from the large number of cells involved in the initial formation of an EB, none or only a very small number contribute to the hematopoietic output obtained from a single EB under the conditions studied. This suggests either that a very small number of the initial ES/iPS cells had the ability to differentiate towards the hematopoietic lineage, or that the efficiency of differentiation is still extremely limited. Further improvements to this method to combine marking of a clone at an early stage with specific integration site analysis or use of barcoded libraries of lentiviral vectors346 would allow quantitative analyses of differentiated cell outputs to address questions of the relative prevalence of “differentiation” events at varying stages of the process. For example, Stewart et al. recently asked if clones of undifferentiated human ES cells contributed similarly to differentiation protocols in vitro and in vivo. Interestingly, this group found that some clones contributed to differentiation using multiple methods, but a greater number of clones contributed to teratomas than EBs. As not all cells were marked, this study did not address the total number of contributing cells, but rather the relative contribution in multiple assays347. Within the scope of hematopoietic differentiation, it would be of great interest to identify key intermediate cell types, and assess the clonal diversity at each step of differentiation. This would allow tracking of the efficiencies of differentiation steps with increased precision compared to the single EB analysis performed in Chapter 3. 125  The methods developed in Chapter 4 to mark cells with lentiviral vectors for subsequent tracking purposes can now also be used for other applications that require the introduction of normal or mutant genes into undifferentiated or early differentiating mesodermal ES/iPS cells. Such strategies are likely to be important for evaluating their role in human hematopoietic development. For genes that increase hematopoietic output, a combination of functional and clonal analyses could then help determine the cellular mechanism of any increases obtained – whether the number of contributing clones has increased, or whether the size of the clones has increased. This would help to define genes with a role in the generation and expansion of hematopoietic cells, respectively.  5.3 Future opportunities  A major outstanding issue is the relevance of the results presented here and by others to the search for conditions required to support the in vitro generation of HSCs. Because it is not feasible to use the prolonged and labour-intensive functional assays needed to specifically identify HSCs in repeated optimization experiments, most groups have chosen to rely on CFC production as a quantitative endpoint of hematopoietic cell production. However, it is well known that assays for CFCs have little specificity for HSCs even when overlapping populations are present and, in most cases, very few if any cells detected as CFCs represent HSCs. This disparity in the spectrum of cells detected as CFCs and HSCs has repeatedly misled investigators in the past who have relied on CFC results to predict conditions anticipated to expand pre-existing human HSCs in vitro. Another issue, even more relevant to the generation of HSCs from pluripotent cells is the appearance of CFCs in the embryo prior to the generation of HSCs (as reviewed in Chapter 1). A few groups have now reported some success at generating human ES or iPS-derived hematopoietic cells that survive and engraft in vivo for a period of at least 8-16 weeks in immunodeficient mice139,155,287,289,317 or >6 months in fetal sheep 126  recipients286. However, these results typically show very low levels of human cell production, appear to be difficult to repeat, and have not always been well controlled, bringing into question their real strength. In particular, these results are not what might be expected from fetal or neonatal human HSCs. In fact, when compared to experiments using umbilical cord blood cells, ES-derived cells show decreased human cell production from similar numbers of phenotypically analogous cells and/or decreased CD45 expression on the engrafted cells286,287,317. Despite this apparent difference in in vivo function, the mature and progenitor hematopoietic cells that are produced appear to be functionally normal. For example, typical CFCs are generated and mature ES-derived erythrocytes can enucleate, express hemoglobins, and carry oxygen. However, their profiles more closely resemble embryonic or fetal erythrocytes, rather than adult erythrocytes278,348. This is not necessarily a problem for future research and clinical uses of these mature cells, but suggests that part of the difficulty in obtaining HSCs in these cultures may be due to a lack of provision of conditions required for the generation of fully functional adult HSCs. Additionally, the generation of HSCs appears to be a rare event at the earliest stages in vivo 172,181,228,270 and the signals required for the generation, expansion, and maintenance of HSCs at different stages of ontogeny are known to be different298. This complicates even further the type of experimental design that may be needed to optimize HSC production. Since methods to support a robust expansion of pre-existing human HSCs have not yet been identified131,132,349, it would be expected that this would further compromise any in vitro attempt to generate HSCs de novo. Nevertheless, new approaches for characterizing HSCs directly are rapidly emerging and hence strategies for human HSC expansion may not be far away350. Recently, there has also been much interest in producing various tissue cell types directly from fibroblasts (or other types of somatic cells) using a direct reprogramming process similar to that used to reprogram cells to iPS cells. Generally these have made use of key transcription factors already known or anticipated to be important for the development of the 127  desired end cells; e.g., Ascl1, Brn2, and Myt1l for neural cells and Gata4, Mef2c, and Tbx5 for cardiomyocytes351,352. It has also been very recently reported that the introduction of OCT4 and provision of hematopoietic cytokines enables human fibroblasts to generate mature hematopoietic cells and progenitors, as well as some cells that behave in vivo in a similar manner to ES-derived hematopoietic cells353. Interestingly, in this study, it was reported that the hemoglobin expressed in the hematopoietic cells was of the adult type unlike what has been seen in many ES-derived hematopoietic cells. It will therefore be important to understand additional features of these cells and whether their precursors include cells with better appreciated properties of human HSCs. In addition to the unique opportunities afforded for studies of normal development, the advent of iPS cell technology has introduced new access to pluripotent cells from a wide variety of congenital and acquired genetic disorders. The study and characterization of deficiencies in blood formation derived from iPS cell lines from such patients, in combination with the ability to further genetically manipulate these cells in vitro, will no doubt provide important insights into the biology of blood diseases as well as creating important therapeutic models. Cell lines generated and being studied from patients with hematopoietic disease already include lines from patients with severe combined immunodeficiency81, Fanconi anemia354, βthalassemia355,356, myeloproliferative disorders357, and chronic myeloid leukemia358. Improvements in the reliability and efficiency of differentiation from human ES/iPS cells to ultimately permit the production of functional derivatives suitable for use in translational research and regenerative medicine remain key requirements for further work with these cells. Obtaining higher yields of more pure, defined populations of cells to study would allow detailed characterization of the molecular and epigenetic states of undifferentiated cells and their differentiated derivatives in order to provide a more complete picture of the regulation of the development and generation of hematopoietic cells.  128  References 1.  Smith AG. Embryo-derived stem cells: of mice and men. Annu Rev Cell Dev Biol. 2001;17:435-462.  2.  Evans MJ, Kaufman MH. Establishment in culture of pluripotential cells from mouse embryos. Nature. 1981;292:154-156.  3.  Martin GR. Isolation of a pluripotent cell line from early mouse embryos cultured in medium conditioned by teratocarcinoma stem cells. Proc Natl Acad Sci U S A. 1981;78:7634-7638.  4.  Nagy A, Rossant J, Nagy R, Abramow-Newerly W, Roder JC. Derivation of completely cell culturederived mice from early-passage embryonic stem cells. Proc Natl Acad Sci U S A. 1993;90:84248428.  5.  Robertson EJ. Pluripotential stem cell lines as a route into the mouse germ line. Trends in Genetics. 1986;2:9-13.  6.  Thomson JA, Itskovitz-Eldor J, Shapiro SS, et al. Embryonic stem cell lines derived from human blastocysts. Science. 1998;282:1145-1147.  7.  Strelchenko N, Verlinsky O, Kukharenko V, Verlinsky Y. Morula-derived human embryonic stem cells. Reprod Biomed Online. 2004;9:623-629.  8.  Chung Y, Klimanskaya I, Becker S, et al. Human embryonic stem cell lines generated without embryo destruction. Cell Stem Cell. 2008;2:113-117.  9.  Klimanskaya I, Chung Y, Becker S, Lu SJ, Lanza R. Human embryonic stem cell lines derived from single blastomeres. Nature. 2006;444:481-485.  10. O'Connor MD, Kardel MD, Iosfina I, et al. Alkaline phosphatase-positive colony formation is a sensitive, specific, and quantitative indicator of undifferentiated human embryonic stem cells. Stem Cells. 2008;26:1109-1116. 11. Amit M, Carpenter MK, Inokuma MS, et al. Clonally derived human embryonic stem cell lines maintain pluripotency and proliferative potential for prolonged periods of culture. Dev Biol. 2000;227:271-278. 12. Price PJ, Goldsborough MD, Tilkins ML. Embryonic stem cell serum replacement: International Patent Application WO 98/30679; 1998. 13. Xu C, Inokuma MS, Denham J, et al. Feeder-free growth of undifferentiated human embryonic stem cells. Nat Biotechnol. 2001;19:971-974. 14. Li Y, Powell S, Brunette E, Lebkowski J, Mandalam R. Expansion of human embryonic stem cells in defined serum-free medium devoid of animal-derived products. Biotechnol Bioeng. 2005;91:688-698. 15. Lu J, Hou R, Booth CJ, Yang SH, Snyder M. Defined culture conditions of human embryonic stem cells. Proc Natl Acad Sci U S A. 2006;103:5688-5693. 16. Ludwig TE, Bergendahl V, Levenstein ME, Yu J, Probasco MD, Thomson JA. Feeder-independent culture of human embryonic stem cells. Nat Methods. 2006;3:637-646. 17. Vallier L, Alexander M, Pedersen RA. Activin/Nodal and FGF pathways cooperate to maintain pluripotency of human embryonic stem cells. J Cell Sci. 2005;118:4495-4509. 18. Wang L, Schulz TC, Sherrer ES, et al. Self-renewal of human embryonic stem cells requires insulinlike growth factor-1 receptor and ERBB2 receptor signaling. Blood. 2007;110:4111-4119. 19. Yao S, Chen S, Clark J, et al. Long-term self-renewal and directed differentiation of human embryonic stem cells in chemically defined conditions. Proc Natl Acad Sci U S A. 2006;103:69076912. 20. Amit M, Shariki C, Margulets V, Itskovitz-Eldor J. Feeder layer- and serum-free culture of human embryonic stem cells. Biol Reprod. 2004;70:837-845. 21. Xu R-H, Peck RM, Li DS, Feng X, Ludwig T, Thomson JA. Basic FGF and suppression of BMP signaling sustain undifferentiated proliferation of human ES cells. Nature Methods. 2005;2:185-190.  129  22. Daheron L, Opitz SL, Zaehres H, et al. LIF/STAT3 signaling fails to maintain self-renewal of human embryonic stem cells. Stem Cells. 2004;22:770-778. 23. Niwa H, Burdon T, Chambers I, Smith A. Self-renewal of pluripotent embryonic stem cells is mediated via activation of STAT3. Genes Dev. 1998;12:2048-2060. 24. Ying QL, Nichols J, Chambers I, Smith A. BMP induction of Id proteins suppresses differentiation and sustains embryonic stem cell self-renewal in collaboration with STAT3. Cell. 2003;115:281-292. 25. Chan EM, Yates F, Boyer LF, Schlaeger TM, Daley GQ. Enhanced plating efficiency of trypsinadapted human embryonic stem cells is reversible and independent of trisomy 12/17. Cloning Stem Cells. 2008;10:107-118. 26. Watanabe K, Ueno M, Kamiya D, et al. A ROCK inhibitor permits survival of dissociated human embryonic stem cells. Nat Biotechnol. 2007;25:681-686. 27. Ellerstrom C, Strehl R, Noaksson K, Hyllner J, Semb H. Facilitated expansion of human embryonic stem cells by single-cell enzymatic dissociation. Stem Cells. 2007;25:1690-1696. 28. Hasegawa K, Fujioka T, Nakamura Y, Nakatsuji N, Suemori H. A method for the selection of human embryonic stem cell sublines with high replating efficiency after single-cell dissociation. Stem Cells. 2006;24:2649-2660. 29. Buzzard JJ, Gough NM, Crook JM, Colman A. Karyotype of human ES cells during extended culture. Nat Biotechnol. 2004;22:381-382; author reply 382. 30. Elliott AM, Elliott KA, Kammesheidt A. High resolution array-CGH characterization of human stem cells using a stem cell focused microarray. Mol Biotechnol. 2010;46:234-242. 31. Werbowetski-Ogilvie TE, Bosse M, Stewart M, et al. Characterization of human embryonic stem cells with features of neoplastic progression. Nat Biotechnol. 2009;27:91-97. 32. Brons IG, Smithers LE, Trotter MW, et al. Derivation of pluripotent epiblast stem cells from mammalian embryos. Nature. 2007;448:191-195. 33. Tesar PJ, Chenoweth JG, Brook FA, et al. New cell lines from mouse epiblast share defining features with human embryonic stem cells. Nature. 2007;448:196-199. 34. Hanna J, Cheng AW, Saha K, et al. Human embryonic stem cells with biological and epigenetic characteristics similar to those of mouse ESCs. Proc Natl Acad Sci U S A. 2010;107:9222-9227. 35. Boyer LA, Lee TI, Cole MF, et al. Core transcriptional regulatory circuitry in human embryonic stem cells. Cell. 2005;122:947-956. 36. Chambers I, Tomlinson SR. The transcriptional foundation of pluripotency. Development. 2009;136:2311-2322. 37. Fong H, Hohenstein KA, Donovan PJ. Regulation of self-renewal and pluripotency by Sox2 in human embryonic stem cells. Stem Cells. 2008;26:1931-1938. 38. Takahashi K, Yamanaka S. Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell. 2006;126:663-676. 39. Park IH, Zhao R, West JA, et al. Reprogramming of human somatic cells to pluripotency with defined factors. Nature. 2008;451:141-146. 40. Takahashi K, Tanabe K, Ohnuki M, et al. Induction of pluripotent stem cells from adult human fibroblasts by defined factors. Cell. 2007;131:861-872. 41. Okita K, Ichisaka T, Yamanaka S. Generation of germline-competent induced pluripotent stem cells. Nature. 2007;448:313-317. 42. Nakagawa M, Koyanagi M, Tanabe K, et al. Generation of induced pluripotent stem cells without Myc from mouse and human fibroblasts. Nat Biotechnol. 2008;26:101-106. 43. Wernig M, Meissner A, Cassady JP, Jaenisch R. c-Myc is dispensable for direct reprogramming of mouse fibroblasts. Cell Stem Cell. 2008;2:10-12. 44. Yu J, Vodyanik MA, Smuga-Otto K, et al. Induced pluripotent stem cell lines derived from human somatic cells. Science. 2007;318:1917-1920.  130  45. Feng B, Jiang J, Kraus P, et al. Reprogramming of fibroblasts into induced pluripotent stem cells with orphan nuclear receptor Esrrb. Nat Cell Biol. 2009;11:197-203. 46. Maherali N, Ahfeldt T, Rigamonti A, Utikal J, Cowan C, Hochedlinger K. A high-efficiency system for the generation and study of human induced pluripotent stem cells. Cell Stem Cell. 2008;3:340-345. 47. Wernig M, Lengner CJ, Hanna J, et al. A drug-inducible transgenic system for direct reprogramming of multiple somatic cell types. Nat Biotechnol. 2008;26:916-924. 48. Yu J, Hu K, Smuga-Otto K, et al. Human induced pluripotent stem cells free of vector and transgene sequences. Science. 2009;324:797-801. 49. Hong H, Takahashi K, Ichisaka T, et al. Suppression of induced pluripotent stem cell generation by the p53-p21 pathway. Nature. 2009;460:1132-1135. 50. Kawamura T, Suzuki J, Wang YV, et al. Linking the p53 tumour suppressor pathway to somatic cell reprogramming. Nature. 2009;460:1140-1144. 51. Li H, Collado M, Villasante A, et al. The Ink4/Arf locus is a barrier for iPS cell reprogramming. Nature. 2009;460:1136-1139. 52. Marion RM, Strati K, Li H, et al. A p53-mediated DNA damage response limits reprogramming to ensure iPS cell genomic integrity. Nature. 2009;460:1149-1153. 53. Takenaka C, Nishishita N, Takada N, Jakt LM, Kawamata S. Effective generation of iPS cells from CD34+ cord blood cells by inhibition of p53. Exp Hematol. 2010;38:154-162. 54. Utikal J, Polo JM, Stadtfeld M, et al. Immortalization eliminates a roadblock during cellular reprogramming into iPS cells. Nature. 2009;460:1145-1148. 55. Zhao Y, Yin X, Qin H, et al. Two supporting factors greatly improve the efficiency of human iPSC generation. Cell Stem Cell. 2008;3:475-479. 56. Feng B, Ng JH, Heng JC, Ng HH. Molecules that promote or enhance reprogramming of somatic cells to induced pluripotent stem cells. Cell Stem Cell. 2009;4:301-312. 57. Carey BW, Markoulaki S, Hanna J, et al. Reprogramming of murine and human somatic cells using a single polycistronic vector. Proc Natl Acad Sci U S A. 2009;106:157-162. 58. Shao L, Feng W, Sun Y, et al. Generation of iPS cells using defined factors linked via the selfcleaving 2A sequences in a single open reading frame. Cell Res. 2009;19:296-306. 59. Soldner F, Hockemeyer D, Beard C, et al. Parkinson's disease patient-derived induced pluripotent stem cells free of viral reprogramming factors. Cell. 2009;136:964-977. 60. Kaji K, Norrby K, Paca A, Mileikovsky M, Mohseni P, Woltjen K. Virus-free induction of pluripotency and subsequent excision of reprogramming factors. Nature. 2009;458:771-775. 61. Woltjen K, Michael IP, Mohseni P, et al. piggyBac transposition reprograms fibroblasts to induced pluripotent stem cells. Nature. 2009;458:766-770. 62. Stadtfeld M, Nagaya M, Utikal J, Weir G, Hochedlinger K. Induced pluripotent stem cells generated without viral integration. Science. 2008;322:945-949. 63. Zhou W, Freed CR. Adenoviral gene delivery can reprogram human fibroblasts to induced pluripotent stem cells. Stem Cells. 2009;27:2667-2674. 64. Okita K, Nakagawa M, Hyenjong H, Ichisaka T, Yamanaka S. Generation of mouse induced pluripotent stem cells without viral vectors. Science. 2008;322:949-953. 65. Jia F, Wilson KD, Sun N, et al. A nonviral minicircle vector for deriving human iPS cells. Nat Methods. 2010;7:197-199. 66. Warren L, Manos PD, Ahfeldt T, et al. Highly efficient reprogramming to pluripotency and directed differentiation of human cells with synthetic modified mRNA. Cell Stem Cell. 2010;7:618-630. 67. Zhou H, Wu S, Joo JY, et al. Generation of induced pluripotent stem cells using recombinant proteins. Cell Stem Cell. 2009;4:381-384. 68. Aasen T, Raya A, Barrero MJ, et al. Efficient and rapid generation of induced pluripotent stem cells from human keratinocytes. Nat Biotechnol. 2008;26:1276-1284.  131  69. Sun N, Panetta NJ, Gupta DM, et al. Feeder-free derivation of induced pluripotent stem cells from adult human adipose stem cells. Proc Natl Acad Sci U S A. 2009;106:15720-15725. 70. Cai J, Li W, Su H, et al. Generation of human induced pluripotent stem cells from umbilical cord matrix and amniotic membrane mesenchymal cells. J Biol Chem. 2010;285:11227-11234. 71. Giorgetti A, Montserrat N, Aasen T, et al. Generation of induced pluripotent stem cells from human cord blood using OCT4 and SOX2. Cell Stem Cell. 2009;5:353-357. 72. Loh YH, Agarwal S, Park IH, et al. Generation of induced pluripotent stem cells from human blood. Blood. 2009;113:5476-5479. 73. Kunisato A, Wakatsuki M, Shinba H, Ota T, Ishida I, Nagao K. Direct Generation of Induced Pluripotent Stem Cells from Human Nonmobilized Blood. Stem Cells Dev. 2010. 74. Loh YH, Hartung O, Li H, et al. Reprogramming of T cells from human peripheral blood. Cell Stem Cell. 2010;7:15-19. 75. Seki T, Yuasa S, Oda M, et al. Generation of induced pluripotent stem cells from human terminally differentiated circulating T cells. Cell Stem Cell. 2010;7:11-14. 76. Staerk J, Dawlaty MM, Gao Q, et al. Reprogramming of human peripheral blood cells to induced pluripotent stem cells. Cell Stem Cell. 2010;7:20-24. 77. Eminli S, Foudi A, Stadtfeld M, et al. Differentiation stage determines potential of hematopoietic cells for reprogramming into induced pluripotent stem cells. Nat Genet. 2009;41:968-976. 78. Kim JB, Sebastiano V, Wu G, et al. Oct4-induced pluripotency in adult neural stem cells. Cell. 2009;136:411-419. 79. Hotta A, Cheung AY, Farra N, et al. Isolation of human iPS cells using EOS lentiviral vectors to select for pluripotency. Nat Methods. 2009;6:370-376. 80. Kazuki Y, Hiratsuka M, Takiguchi M, et al. Complete genetic correction of ips cells from Duchenne muscular dystrophy. Mol Ther. 2010;18:386-393. 81. Park IH, Arora N, Huo H, et al. Disease-specific induced pluripotent stem cells. Cell. 2008;134:877886. 82. Daley GQ, Lensch MW, Jaenisch R, Meissner A, Plath K, Yamanaka S. Broader implications of defining standards for the pluripotency of iPSCs. Cell Stem Cell. 2009;4:200-201; author reply 202. 83. Ellis J, Bruneau BG, Keller G, et al. Alternative induced pluripotent stem cell characterization criteria for in vitro applications. Cell Stem Cell. 2009;4:198-199; author reply 202. 84. Maherali N, Hochedlinger K. Guidelines and techniques for the generation of induced pluripotent stem cells. Cell Stem Cell. 2008;3:595-605. 85. Wernig M, Meissner A, Foreman R, et al. In vitro reprogramming of fibroblasts into a pluripotent EScell-like state. Nature. 2007;448:318-324. 86. Zhao XY, Li W, Lv Z, et al. iPS cells produce viable mice through tetraploid complementation. Nature. 2009;461:86-90. 87. Stadtfeld M, Apostolou E, Akutsu H, et al. Aberrant silencing of imprinted genes on chromosome 12qF1 in mouse induced pluripotent stem cells. Nature. 2010;465:175-181. 88. Chan EM, Ratanasirintrawoot S, Park IH, et al. Live cell imaging distinguishes bona fide human iPS cells from partially reprogrammed cells. Nat Biotechnol. 2009;27:1033-1037. 89. Adewumi O, Aflatoonian B, Ahrlund-Richter L, et al. Characterization of human embryonic stem cell lines by the International Stem Cell Initiative. Nat Biotechnol. 2007;25:803-816. 90. Akopian V, Andrews PW, Beil S, et al. Comparison of defined culture systems for feeder cell free propagation of human embryonic stem cells. In Vitro Cell Dev Biol Anim. 2010;46:247-258. 91. Hirst M, Delaney A, Rogers SA, et al. LongSAGE profiling of nine human embryonic stem cell lines. Genome Biol. 2007;8:R113.  132  92. Liu Y, Shin S, Zeng X, et al. Genome wide profiling of human embryonic stem cells (hESCs), their derivatives and embryonal carcinoma cells to develop base profiles of U.S. Federal government approved hESC lines. BMC Dev Biol. 2006;6:20. 93. Chin MH, Mason MJ, Xie W, et al. Induced pluripotent stem cells and embryonic stem cells are distinguished by gene expression signatures. Cell Stem Cell. 2009;5:111-123. 94. Ghosh Z, Wilson KD, Wu Y, Hu S, Quertermous T, Wu JC. Persistent donor cell gene expression among human induced pluripotent stem cells contributes to differences with human embryonic stem cells. PLoS One. 2010;5:e8975. 95. Choi KD, Yu J, Smuga-Otto K, et al. Hematopoietic and endothelial differentiation of human induced pluripotent stem cells. Stem Cells. 2009;27:559-567. 96. Feng Q, Lu SJ, Klimanskaya I, et al. Hemangioblastic derivatives from human induced pluripotent stem cells exhibit limited expansion and early senescence. Stem Cells. 2010;28:704-712. 97. Grigoriadis AE, Kennedy M, Bozec A, et al. Directed differentiation of hematopoietic precursors and functional osteoclasts from human ES and iPS cells. Blood. 2010;115:2769-2776. 98. Hu BY, Weick JP, Yu J, et al. Neural differentiation of human induced pluripotent stem cells follows developmental principles but with variable potency. Proc Natl Acad Sci U S A. 2010;107:4335-4340. 99. Osafune K, Caron L, Borowiak M, et al. Marked differences in differentiation propensity among human embryonic stem cell lines. Nat Biotechnol. 2008;26:313-315. 100. Ramos-Mejia V, Melen GJ, Sanchez L, et al. Nodal/Activin signaling predicts human pluripotent stem cell lines prone to differentiate toward the hematopoietic lineage. Mol Ther. 2010;18:2173-2181. 101. Kim K, Doi A, Wen B, et al. Epigenetic memory in induced pluripotent stem cells. Nature. 2010;467:285-290. 102. Polo JM, Liu S, Figueroa ME, et al. Cell type of origin influences the molecular and functional properties of mouse induced pluripotent stem cells. Nat Biotechnol. 2010;28:848-855. 103. Akashi K, Traver D, Miyamoto T, Weissman IL. A clonogenic common myeloid progenitor that gives rise to all myeloid lineages. Nature. 2000;404:193-197. 104. Kondo M, Weissman IL, Akashi K. Identification of clonogenic common lymphoid progenitors in mouse bone marrow. Cell. 1997;91:661-672. 105. Cumano A, Paige CJ, Iscove NN, Brady G. Bipotential precursors of B cells and macrophages in murine fetal liver. Nature. 1992;356:612-615. 106. Doulatov S, Notta F, Eppert K, Nguyen LT, Ohashi PS, Dick JE. Revised map of the human progenitor hierarchy shows the origin of macrophages and dendritic cells in early lymphoid development. Nat Immunol. 2010;11:585-593. 107. Montecino-Rodriguez E, Leathers H, Dorshkind K. Bipotential B-macrophage progenitors are present in adult bone marrow. Nat Immunol. 2001;2:83-88. 108. Hogge DE, Lansdorp PM, Reid D, Gerhard B, Eaves CJ. Enhanced detection, maintenance, and differentiation of primitive human hematopoietic cells in cultures containing murine fibroblasts engineered to produce human steel factor, interleukin-3, and granulocyte colony-stimulating factor. Blood. 1996;88:3765-3773. 109. Eaves CJ, Eaves AC. Anatomy and physiology of hematopoiesis. In: Pui C-H, ed. Childhood Leukemias (ed 2nd). Cambridge: Cambridge University Press; 2006:69-105. 110. Larochelle A, Vormoor J, Hanenberg H, et al. Identification of primitive human hematopoietic cells capable of repopulating NOD/SCID mouse bone marrow: implications for gene therapy. Nat Med. 1996;2:1329-1337. 111. Pflumio F, Izac B, Katz A, Shultz LD, Vainchenker W, Coulombel L. Phenotype and function of human hematopoietic cells engrafting immune-deficient CB17-severe combined immunodeficiency mice and nonobese diabetic-severe combined immunodeficiency mice after transplantation of human cord blood mononuclear cells. Blood. 1996;88:3731-3740.  133  112. Zanjani ED, Almeida-Porada G, Ascensao JL, MacKintosh FR, Flake AW. Transplantation of hematopoietic stem cells in utero. Stem Cells. 1997;15 Suppl 1:79-92; discussion 93. 113. Glimm H, Eisterer W, Lee K, et al. Previously undetected human hematopoietic cell populations with short-term repopulating activity selectively engraft NOD/SCID-beta2 microglobulin-null mice. J Clin Invest. 2001;107:199-206. 114. Ito M, Hiramatsu H, Kobayashi K, et al. NOD/SCID/gamma(c)(null) mouse: an excellent recipient mouse model for engraftment of human cells. Blood. 2002;100:3175-3182. 115. Shultz LD, Lyons BL, Burzenski LM, et al. Human lymphoid and myeloid cell development in NOD/LtSz-scid IL2R gamma null mice engrafted with mobilized human hemopoietic stem cells. J Immunol. 2005;174:6477-6489. 116. Kent DG, Copley MR, Benz C, et al. Prospective isolation and molecular characterization of hematopoietic stem cells with durable self-renewal potential. Blood. 2009;113:6342-6350. 117. Yilmaz OH, Kiel MJ, Morrison SJ. SLAM family markers are conserved among hematopoietic stem cells from old and reconstituted mice and markedly increase their purity. Blood. 2006;107:924-930. 118. Morrison SJ, Hemmati HD, Wandycz AM, Weissman IL. The purification and characterization of fetal liver hematopoietic stem cells. Proc Natl Acad Sci U S A. 1995;92:10302-10306. 119. Rebel VI, Miller CL, Thornbury GR, Dragowska WH, Eaves CJ, Lansdorp PM. A comparison of longterm repopulating hematopoietic stem cells in fetal liver and adult bone marrow from the mouse. Exp Hematol. 1996;24:638-648. 120. Sato T, Laver JH, Ogawa M. Reversible expression of CD34 by murine hematopoietic stem cells. Blood. 1999;94:2548-2554. 121. Uchida N, Dykstra B, Lyons K, Leung F, Kristiansen M, Eaves C. ABC transporter activities of murine hematopoietic stem cells vary according to their developmental and activation status. Blood. 2004;103:4487-4495. 122. Yoder MC, Hiatt K, Dutt P, Mukherjee P, Bodine DM, Orlic D. Characterization of definitive lymphohematopoietic stem cells in the day 9 murine yolk sac. Immunity. 1997;7:335-344. 123. Hogan CJ, Shpall EJ, Keller G. Differential long-term and multilineage engraftment potential from subfractions of human CD34+ cord blood cells transplanted into NOD/SCID mice. Proc Natl Acad Sci U S A. 2002;99:413-418. 124. Notta F, Doulatov S, Dick JE. Engraftment of human hematopoietic stem cells is more efficient in female NOD/SCID/IL-2Rgc-null recipients. Blood. 2010;115:3704-3707. 125. Hess DA, Meyerrose TE, Wirthlin L, et al. Functional characterization of highly purified human hematopoietic repopulating cells isolated according to aldehyde dehydrogenase activity. Blood. 2004;104:1648-1655. 126. Bhatia M, Bonnet D, Murdoch B, Gan OI, Dick JE. A newly discovered class of human hematopoietic cells with SCID-repopulating activity. Nat Med. 1998;4:1038-1045. 127. Verfaillie CM, Almeida-Porada G, Wissink S, Zanjani ED. Kinetics of engraftment of CD34(-) and CD34(+) cells from mobilized blood differs from that of CD34(-) and CD34(+) cells from bone marrow. Exp Hematol. 2000;28:1071-1079. 128. Antonchuk J, Sauvageau G, Humphries RK. HOXB4-induced expansion of adult hematopoietic stem cells ex vivo. Cell. 2002;109:39-45. 129. Ohta H, Sekulovic S, Bakovic S, et al. Near-maximal expansions of hematopoietic stem cells in culture using NUP98-HOX fusions. Exp Hematol. 2007;35:817-830. 130. Audet J, Miller CL, Eaves CJ, Piret JM. Common and distinct features of cytokine effects on hematopoietic stem and progenitor cells revealed by dose-response surface analysis. Biotechnol Bioeng. 2002;80:393-404. 131. Bhatia M, Bonnet D, Kapp U, Wang JC, Murdoch B, Dick JE. Quantitative analysis reveals expansion of human hematopoietic repopulating cells after short-term ex vivo culture. J Exp Med. 1997;186:619-624.  134  132. Conneally E, Cashman J, Petzer A, Eaves C. Expansion in vitro of transplantable human cord blood stem cells demonstrated using a quantitative assay of their lympho-myeloid repopulating activity in nonobese diabetic-scid/scid mice. Proc Natl Acad Sci U S A. 1997;94:9836-9841. 133. Ema H, Takano H, Sudo K, Nakauchi H. In vitro self-renewal division of hematopoietic stem cells. J Exp Med. 2000;192:1281-1288. 134. Yonemura Y, Ku H, Lyman SD, Ogawa M. In vitro expansion of hematopoietic progenitors and maintenance of stem cells: comparison between FLT3/FLK-2 ligand and KIT ligand. Blood. 1997;89:1915-1921. 135. Nakano T, Kodama H, Honjo T. Generation of lymphohematopoietic cells from embryonic stem cells in culture. Science. 1994;265:1098-1101. 136. Vodyanik MA, Bork JA, Thomson JA, Slukvin, II. Human embryonic stem cell-derived CD34+ cells: efficient production in the coculture with OP9 stromal cells and analysis of lymphohematopoietic potential. Blood. 2005;105:617-626. 137. Kaufman DS, Hanson ET, Lewis RL, Auerbach R, Thomson JA. Hematopoietic colony-forming cells derived from human embryonic stem cells. Proc Natl Acad Sci U S A. 2001;98:10716-10721. 138. Qiu C, Hanson E, Olivier E, et al. Differentiation of human embryonic stem cells into hematopoietic cells by coculture with human fetal liver cells recapitulates the globin switch that occurs early in development. Exp Hematol. 2005;33:1450-1458. 139. Ledran MH, Krassowska A, Armstrong L, et al. Efficient hematopoietic differentiation of human embryonic stem cells on stromal cells derived from hematopoietic niches. Cell Stem Cell. 2008;3:8598. 140. Dang SM, Kyba M, Perlingeiro R, Daley GQ, Zandstra PW. Efficiency of embryoid body formation and hematopoietic development from embryonic stem cells in different culture systems. Biotechnol Bioeng. 2002;78:442-453. 141. Ng ES, Davis RP, Azzola L, Stanley EG, Elefanty AG. Forced aggregation of defined numbers of human embryonic stem cells into embryoid bodies fosters robust, reproducible hematopoietic differentiation. Blood. 2005;106:1601-1603. 142. Ungrin MD, Joshi C, Nica A, Bauwens C, Zandstra PW. Reproducible, ultra high-throughput formation of multicellular organization from single cell suspension-derived human embryonic stem cell aggregates. PLoS One. 2008;3:e1565. 143. Chadwick K, Wang L, Li L, et al. Cytokines and BMP-4 promote hematopoietic differentiation of human embryonic stem cells. Blood. 2003;102:906-915. 144. Itskovitz-Eldor J, Schuldiner M, Karsenti D, et al. Differentiation of human embryonic stem cells into embryoid bodies compromising the three embryonic germ layers. Mol Med. 2000;6:88-95. 145. Schuldiner M, Yanuka O, Itskovitz-Eldor J, Melton DA, Benvenisty N. Effects of eight growth factors on the differentiation of cells derived from human embryonic stem cells. Proc Natl Acad Sci U S A. 2000;97:11307-11312. 146. Cerdan C, Rouleau A, Bhatia M. VEGF-A165 augments erythropoietic development from human embryonic stem cells. Blood. 2004;103:2504-2512. 147. Johansson BM, Wiles MV. Evidence for involvement of activin A and bone morphogenetic protein 4 in mammalian mesoderm and hematopoietic development. Mol Cell Biol. 1995;15:141-151. 148. Keller G, Kennedy M, Papayannopoulou T, Wiles MV. Hematopoietic commitment during embryonic stem cell differentiation in culture. Mol Cell Biol. 1993;13:473-486. 149. Kennedy M, D'Souza SL, Lynch-Kattman M, Schwantz S, Keller G. Development of the hemangioblast defines the onset of hematopoiesis in human ES cell differentiation cultures. Blood. 2007;109:2679-2687. 150. Zambidis ET, Peault B, Park TS, Bunz F, Civin CI. Hematopoietic differentiation of human embryonic stem cells progresses through sequential hematoendothelial, primitive, and definitive stages resembling human yolk sac development. Blood. 2005;106:860-870.  135  151. Hanna J, Wernig M, Markoulaki S, et al. Treatment of sickle cell anemia mouse model with iPS cells generated from autologous skin. Science. 2007;318:1920-1923. 152. Irion S, Clarke RL, Luche H, et al. Temporal specification of blood progenitors from mouse embryonic stem cells and induced pluripotent stem cells. Development. 2010;137:2829-2839. 153. Schenke-Layland K, Rhodes KE, Angelis E, et al. Reprogrammed mouse fibroblasts differentiate into cells of the cardiovascular and hematopoietic lineages. Stem Cells. 2008;26:1537-1546. 154. Lengerke C, Grauer M, Niebuhr NI, et al. Hematopoietic development from human induced pluripotent stem cells. Ann N Y Acad Sci. 2009;1176:219-227. 155. Lu M, Kardel MD, O'Connor MD, Eaves CJ. Enhanced generation of hematopoietic cells from human hepatocarcinoma cell-stimulated human embryonic and induced pluripotent stem cells. Exp Hematol. 2009;37:924-936. 156. Luckett WP. Origin and differentiation of the yolk sac and extraembryonic mesoderm in presomite human and rhesus monkey embryos. Am J Anat. 1978;152:59-97. 157. Moore MA, Metcalf D. Ontogeny of the haemopoietic system: yolk sac origin of in vivo and in vitro colony forming cells in the developing mouse embryo. Br J Haematol. 1970;18:279-296. 158. Corbel C, Salaun J, Belo-Diabangouaya P, Dieterlen-Lievre F. Hematopoietic potential of the prefusion allantois. Dev Biol. 2007;301:478-488. 159. Dzierzak E, Speck NA. Of lineage and legacy: the development of mammalian hematopoietic stem cells. Nat Immunol. 2008;9:129-136. 160. Godin IE, Garcia-Porrero JA, Coutinho A, Dieterlen-Lievre F, Marcos MA. Para-aortic splanchnopleura from early mouse embryos contains B1a cell progenitors. Nature. 1993;364:67-70. 161. Medvinsky AL, Samoylina NL, Muller AM, Dzierzak EA. An early pre-liver intraembryonic source of CFU-S in the developing mouse. Nature. 1993;364:64-67. 162. Godin I, Cumano A. The hare and the tortoise: an embryonic haematopoietic race. Nat Rev Immunol. 2002;2:593-604. 163. Houssaint E. Differentiation of the mouse hepatic primordium. II. Extrinsic origin of the haemopoietic cell line. Cell Differ. 1981;10:243-252. 164. Johnson GR, Moore MA. Role of stem cell migration in initiation of mouse foetal liver haemopoiesis. Nature. 1975;258:726-728. 165. Ema H, Nakauchi H. Expansion of hematopoietic stem cells in the developing liver of a mouse embryo. Blood. 2000;95:2284-2288. 166. Wolber FM, Leonard E, Michael S, Orschell-Traycoff CM, Yoder MC, Srour EF. Roles of spleen and liver in development of the murine hematopoietic system. Exp Hematol. 2002;30:1010-1019. 167. Huyhn A, Dommergues M, Izac B, et al. Characterization of hematopoietic progenitors from human yolk sacs and embryos. Blood. 1995;86:4474-4485. 168. Lensch MW, Daley GQ. Origins of mammalian hematopoiesis: in vivo paradigms and in vitro models. Curr Top Dev Biol. 2004;60:127-196. 169. Tavian M, Hallais MF, Peault B. Emergence of intraembryonic hematopoietic precursors in the preliver human embryo. Development. 1999;126:793-803. 170. Robin C, Bollerot K, Mendes S, et al. Human placenta is a potent hematopoietic niche containing hematopoietic stem and progenitor cells throughout development. Cell Stem Cell. 2009;5:385-395. 171. Alvarez-Silva M, Belo-Diabangouaya P, Salaun J, Dieterlen-Lievre F. Mouse placenta is a major hematopoietic organ. Development. 2003;130:5437-5444. 172. Gekas C, Dieterlen-Lievre F, Orkin SH, Mikkola HK. The placenta is a niche for hematopoietic stem cells. Dev Cell. 2005;8:365-375. 173. McGrath KE, Koniski AD, Malik J, Palis J. Circulation is established in a stepwise pattern in the mammalian embryo. Blood. 2003;101:1669-1676.  136  174. Kingsley PD, Malik J, Emerson RL, et al. "Maturational" globin switching in primary primitive erythroid cells. Blood. 2006;107:1665-1672. 175. Kingsley PD, Malik J, Fantauzzo KA, Palis J. Yolk sac-derived primitive erythroblasts enucleate during mammalian embryogenesis. Blood. 2004;104:19-25. 176. Palis J, Malik J, McGrath KE, Kingsley PD. Primitive erythropoiesis in the mammalian embryo. Int J Dev Biol. 2010;54:1011-1018. 177. Van Handel B, Prashad SL, Hassanzadeh-Kiabi N, et al. The first trimester human placenta is a site for terminal maturation of primitive erythroid cells. Blood. 2010;116:3321-3330. 178. Tober J, Koniski A, McGrath KE, et al. The megakaryocyte lineage originates from hemangioblast precursors and is an integral component both of primitive and of definitive hematopoiesis. Blood. 2007;109:1433-1441. 179. Palis J, Robertson S, Kennedy M, Wall C, Keller G. Development of erythroid and myeloid progenitors in the yolk sac and embryo proper of the mouse. Development. 1999;126:5073-5084. 180. Naito M, Yamamura F, Nishikawa S, Takahashi K. Development, differentiation, and maturation of fetal mouse yolk sac macrophages in cultures. J Leukoc Biol. 1989;46:1-10. 181. Kumaravelu P, Hook L, Morrison AM, et al. Quantitative developmental anatomy of definitive haematopoietic stem cells/long-term repopulating units (HSC/RUs): role of the aorta-gonadmesonephros (AGM) region and the yolk sac in colonisation of the mouse embryonic liver. Development. 2002;129:4891-4899. 182. Ottersbach K, Dzierzak E. The murine placenta contains hematopoietic stem cells within the vascular labyrinth region. Dev Cell. 2005;8:377-387. 183. Yoder MC, Hiatt K, Mukherjee P. In vivo repopulating hematopoietic stem cells are present in the murine yolk sac at day 9.0 postcoitus. Proc Natl Acad Sci U S A. 1997;94:6776-6780. 184. Kyba M, Perlingeiro RC, Daley GQ. HoxB4 confers definitive lymphoid-myeloid engraftment potential on embryonic stem cell and yolk sac hematopoietic progenitors. Cell. 2002;109:29-37. 185. Matsuoka S, Tsuji K, Hisakawa H, et al. Generation of definitive hematopoietic stem cells from murine early yolk sac and paraaortic splanchnopleures by aorta-gonad-mesonephros region-derived stromal cells. Blood. 2001;98:6-12. 186. Cho SK, Bourdeau A, Letarte M, Zuniga-Pflucker JC. Expression and function of CD105 during the onset of hematopoiesis from Flk1(+) precursors. Blood. 2001;98:3635-3642. 187. Nakano T, Kodama H, Honjo T. In vitro development of primitive and definitive erythrocytes from different precursors. Science. 1996;272:722-724. 188. Kennedy M, Firpo M, Choi K, et al. A common precursor for primitive erythropoiesis and definitive haematopoiesis. Nature. 1997;386:488-493. 189. Fujimoto T, Ogawa M, Minegishi N, et al. Step-wise divergence of primitive and definitive haematopoietic and endothelial cell lineages during embryonic stem cell differentiation. Genes Cells. 2001;6:1113-1127. 190. Burt RK, Verda L, Kim DA, Oyama Y, Luo K, Link C. Embryonic Stem Cells As an Alternate Marrow Donor Source: Engraftment without Graft-Versus-Host Disease. J Exp Med. 2004;199:895-904. 191. Palacios R, Golunski E, Samaridis J. In vitro generation of hematopoietic stem cells from an embryonic stem cell line. Proc Natl Acad Sci U S A. 1995;92:7530-7534. 192. Matsumoto K, Isagawa T, Nishimura T, et al. Stepwise development of hematopoietic stem cells from embryonic stem cells. PLoS One. 2009;4:e4820. 193. Schuringa JJ, Wu K, Morrone G, Moore MA. Enforced activation of STAT5A facilitates the generation of embryonic stem-derived hematopoietic stem cells that contribute to hematopoiesis in vivo. Stem Cells. 2004;22:1191-1204. 194. Wang Y, Yates F, Naveiras O, Ernst P, Daley GQ. Embryonic stem cell-derived hematopoietic stem cells. Proc Natl Acad Sci U S A. 2005;102:19081-19086.  137  195. Fujiwara Y, Browne CP, Cunniff K, Goff SC, Orkin SH. Arrested development of embryonic red cell precursors in mouse embryos lacking transcription factor GATA-1. Proc Natl Acad Sci U S A. 1996;93:12355-12358. 196. Simon MC, Pevny L, Wiles MV, Keller G, Costantini F, Orkin SH. Rescue of erythroid development in gene targeted GATA-1- mouse embryonic stem cells. Nat Genet. 1992;1:92-98. 197. Pevny L, Lin CS, D'Agati V, Simon MC, Orkin SH, Costantini F. Development of hematopoietic cells lacking transcription factor GATA-1. Development. 1995;121:163-172. 198. Pevny L, Simon MC, Robertson E, et al. Erythroid differentiation in chimaeric mice blocked by a targeted mutation in the gene for transcription factor GATA-1. Nature. 1991;349:257-260. 199. Tsai FY, Keller G, Kuo FC, et al. An early haematopoietic defect in mice lacking the transcription factor GATA-2. Nature. 1994;371:221-226. 200. Tsai FY, Orkin SH. Transcription factor GATA-2 is required for proliferation/survival of early hematopoietic cells and mast cell formation, but not for erythroid and myeloid terminal differentiation. Blood. 1997;89:3636-3643. 201. Ling KW, Ottersbach K, van Hamburg JP, et al. GATA-2 plays two functionally distinct roles during the ontogeny of hematopoietic stem cells. J Exp Med. 2004;200:871-882. 202. Shivdasani RA, Mayer EL, Orkin SH. Absence of blood formation in mice lacking the T-cell leukaemia oncoprotein tal-1/SCL. Nature. 1995;373:432-434. 203. Porcher C, Swat W, Rockwell K, Fujiwara Y, Alt FW, Orkin SH. The T cell leukemia oncoprotein SCL/tal-1 is essential for development of all hematopoietic lineages. Cell. 1996;86:47-57. 204. Robb L, Elwood NJ, Elefanty AG, et al. The scl gene product is required for the generation of all hematopoietic lineages in the adult mouse. Embo J. 1996;15:4123-4129. 205. Curtis DJ, Hall MA, Van Stekelenburg LJ, Robb L, Jane SM, Begley CG. SCL is required for normal function of short-term repopulating hematopoietic stem cells. Blood. 2004;103:3342-3348. 206. Mikkola HK, Klintman J, Yang H, et al. Haematopoietic stem cells retain long-term repopulating activity and multipotency in the absence of stem-cell leukaemia SCL/tal-1 gene. Nature. 2003;421:547-551. 207. Okuda T, van Deursen J, Hiebert SW, Grosveld G, Downing JR. AML1, the target of multiple chromosomal translocations in human leukemia, is essential for normal fetal liver hematopoiesis. Cell. 1996;84:321-330. 208. Wang Q, Stacy T, Binder M, Marin-Padilla M, Sharpe AH, Speck NA. Disruption of the Cbfa2 gene causes necrosis and hemorrhaging in the central nervous system and blocks definitive hematopoiesis. Proc Natl Acad Sci U S A. 1996;93:3444-3449. 209. Cai Z, de Bruijn M, Ma X, et al. Haploinsufficiency of AML1 affects the temporal and spatial generation of hematopoietic stem cells in the mouse embryo. Immunity. 2000;13:423-431. 210. Chen MJ, Yokomizo T, Zeigler BM, Dzierzak E, Speck NA. Runx1 is required for the endothelial to haematopoietic cell transition but not thereafter. Nature. 2009;457:887-891. 211. Growney JD, Shigematsu H, Li Z, et al. Loss of Runx1 perturbs adult hematopoiesis and is associated with a myeloproliferative phenotype. Blood. 2005;106:494-504. 212. Ichikawa M, Asai T, Saito T, et al. AML-1 is required for megakaryocytic maturation and lymphocytic differentiation, but not for maintenance of hematopoietic stem cells in adult hematopoiesis. Nat Med. 2004;10:299-304. 213. Cumano A, Dieterlen-Lievre F, Godin I. Lymphoid potential, probed before circulation in mouse, is restricted to caudal intraembryonic splanchnopleura. Cell. 1996;86:907-916. 214. Cumano A, Ferraz JC, Klaine M, Di Santo JP, Godin I. Intraembryonic, but not yolk sac hematopoietic precursors, isolated before circulation, provide long-term multilineage reconstitution. Immunity. 2001;15:477-485. 215. Palacios R, Imhof BA. At day 8-8.5 of mouse development the yolk sac, not the embryo proper, has lymphoid precursor potential in vivo and in vitro. Proc Natl Acad Sci U S A. 1993;90:6581-6585.  138  216. Zeigler BM, Sugiyama D, Chen M, Guo Y, Downs KM, Speck NA. The allantois and chorion, when isolated before circulation or chorio-allantoic fusion, have hematopoietic potential. Development. 2006;133:4183-4192. 217. Rampon C, Huber P. Multilineage hematopoietic progenitor activity generated autonomously in the mouse yolk sac: analysis using angiogenesis-defective embryos. Int J Dev Biol. 2003;47:273-280. 218. Lux CT, Yoshimoto M, McGrath K, Conway SJ, Palis J, Yoder MC. All primitive and definitive hematopoietic progenitor cells emerging before E10 in the mouse embryo are products of the yolk sac. Blood. 2008;111:3435-3438. 219. Rhodes KE, Gekas C, Wang Y, et al. The emergence of hematopoietic stem cells is initiated in the placental vasculature in the absence of circulation. Cell Stem Cell. 2008;2:252-263. 220. Zovein AC, Hofmann JJ, Lynch M, et al. Fate tracing reveals the endothelial origin of hematopoietic stem cells. Cell Stem Cell. 2008;3:625-636. 221. Oberlin E, Fleury M, Clay D, et al. VE-cadherin expression allows identification of a new class of hematopoietic stem cells within human embryonic liver. Blood. 2010;116:4444-4455. 222. Taoudi S, Gonneau C, Moore K, et al. Extensive hematopoietic stem cell generation in the AGM region via maturation of VE-cadherin+CD45+ pre-definitive HSCs. Cell Stem Cell. 2008;3:99-108. 223. Adamo L, Naveiras O, Wenzel PL, et al. Biomechanical forces promote embryonic haematopoiesis. Nature. 2009;459:1131-1135. 224. North TE, Goessling W, Peeters M, et al. Hematopoietic stem cell development is dependent on blood flow. Cell. 2009;137:736-748. 225. Gothert JR, Gustin SE, Hall MA, et al. In vivo fate-tracing studies using the Scl stem cell enhancer: embryonic hematopoietic stem cells significantly contribute to adult hematopoiesis. Blood. 2005;105:2724-2732. 226. Samokhvalov IM, Samokhvalova NI, Nishikawa S. Cell tracing shows the contribution of the yolk sac to adult haematopoiesis. Nature. 2007;446:1056-1061. 227. Muller AM, Medvinsky A, Strouboulis J, Grosveld F, Dzierzak E. Development of hematopoietic stem cell activity in the mouse embryo. Immunity. 1994;1:291-301. 228. Taylor E, Taoudi S, Medvinsky A. Hematopoietic stem cell activity in the aorta-gonad-mesonephros region enhances after mid-day 11 of mouse development. Int J Dev Biol. 2010;54:1055-1060. 229. Tavian M, Robin C, Coulombel L, Peault B. The human embryo, but not its yolk sac, generates lympho-myeloid stem cells: mapping multipotent hematopoietic cell fate in intraembryonic mesoderm. Immunity. 2001;15:487-495. 230. Tam PP, Behringer RR. Mouse gastrulation: the formation of a mammalian body plan. Mech Dev. 1997;68:3-25. 231. Tam PP, Loebel DA. Gene function in mouse embryogenesis: get set for gastrulation. Nat Rev Genet. 2007;8:368-381. 232. Kinder SJ, Tsang TE, Quinlan GA, Hadjantonakis AK, Nagy A, Tam PP. The orderly allocation of mesodermal cells to the extraembryonic structures and the anteroposterior axis during gastrulation of the mouse embryo. Development. 1999;126:4691-4701. 233. Kanatsu M, Nishikawa SI. In vitro analysis of epiblast tissue potency for hematopoietic cell differentiation. Development. 1996;122:823-830. 234. Bianchi DW, Wilkins-Haug LE, Enders AC, Hay ED. Origin of extraembryonic mesoderm in experimental animals: relevance to chorionic mosaicism in humans. Am J Med Genet. 1993;46:542550. 235. Smith JC. Mesoderm-inducing factors and mesodermal patterning. Curr Opin Cell Biol. 1995;7:856861. 236. Gadue P, Huber TL, Nostro MC, Kattman S, Keller GM. Germ layer induction from embryonic stem cells. Exp Hematol. 2005;33:955-964.  139  237. Liu P, Wakamiya M, Shea MJ, Albrecht U, Behringer RR, Bradley A. Requirement for Wnt3 in vertebrate axis formation. Nat Genet. 1999;22:361-365. 238. Yamaguchi TP, Harpal K, Henkemeyer M, Rossant J. fgfr-1 is required for embryonic growth and mesodermal patterning during mouse gastrulation. Genes Dev. 1994;8:3032-3044. 239. Zhou X, Sasaki H, Lowe L, Hogan BL, Kuehn MR. Nodal is a novel TGF-beta-like gene expressed in the mouse node during gastrulation. Nature. 1993;361:543-547. 240. Winnier G, Blessing M, Labosky PA, Hogan BL. Bone morphogenetic protein-4 is required for mesoderm formation and patterning in the mouse. Genes Dev. 1995;9:2105-2116. 241. Fujiwara T, Dehart DB, Sulik KK, Hogan BL. Distinct requirements for extra-embryonic and embryonic bone morphogenetic protein 4 in the formation of the node and primitive streak and coordination of left-right asymmetry in the mouse. Development. 2002;129:4685-4696. 242. Huber TL, Zhou Y, Mead PE, Zon LI. Cooperative effects of growth factors involved in the induction of hematopoietic mesoderm. Blood. 1998;92:4128-4137. 243. Ng ES, Azzola L, Sourris K, Robb L, Stanley EG, Elefanty AG. The primitive streak gene Mixl1 is required for efficient haematopoiesis and BMP4-induced ventral mesoderm patterning in differentiating ES cells. Development. 2005;132:873-884. 244. Nostro MC, Cheng X, Keller GM, Gadue P. Wnt, activin, and BMP signaling regulate distinct stages in the developmental pathway from embryonic stem cells to blood. Cell Stem Cell. 2008;2:60-71. 245. Park C, Afrikanova I, Chung YS, et al. A hierarchical order of factors in the generation of FLK1- and SCL-expressing hematopoietic and endothelial progenitors from embryonic stem cells. Development. 2004;131:2749-2762. 246. Willems E, Leyns L. Patterning of mouse embryonic stem cell-derived pan-mesoderm by Activin A/Nodal and Bmp4 signaling requires Fibroblast Growth Factor activity. Differentiation. 2008;76:745759. 247. Murray PDF. The Development in vitro of the Blood of the Early Chick Embryo. Proceedings of the Royal Society of London Series B, Containing Papers of a Biological Character. 1932;111:497-521. 248. Sabin FR. Studies on the origin of blood-vessels and of red blood-corpuscles as seen in the living blastoderm of chicks during the second day of incubation. Contributions to Embryology. 1920;9:213268. 249. Fong GH, Klingensmith J, Wood CR, Rossant J, Breitman ML. Regulation of flt-1 expression during mouse embryogenesis suggests a role in the establishment of vascular endothelium. Dev Dyn. 1996;207:1-10. 250. Iwama A, Hamaguchi I, Hashiyama M, Murayama Y, Yasunaga K, Suda T. Molecular cloning and characterization of mouse TIE and TEK receptor tyrosine kinase genes and their expression in hematopoietic stem cells. Biochem Biophys Res Commun. 1993;195:301-309. 251. Schmeisser A, Strasser RH. Phenotypic overlap between hematopoietic cells with suggested angioblastic potential and vascular endothelial cells. J Hematother Stem Cell Res. 2002;11:69-79. 252. Dumont DJ, Gradwohl G, Fong GH, et al. Dominant-negative and targeted null mutations in the endothelial receptor tyrosine kinase, tek, reveal a critical role in vasculogenesis of the embryo. Genes Dev. 1994;8:1897-1909. 253. Sato TN, Tozawa Y, Deutsch U, et al. Distinct roles of the receptor tyrosine kinases Tie-1 and Tie-2 in blood vessel formation. Nature. 1995;376:70-74. 254. Shalaby F, Ho J, Stanford WL, et al. A requirement for Flk1 in primitive and definitive hematopoiesis and vasculogenesis. Cell. 1997;89:981-990. 255. Shalaby F, Rossant J, Yamaguchi TP, et al. Failure of blood-island formation and vasculogenesis in Flk-1-deficient mice. Nature. 1995;376:62-66. 256. Takakura N, Huang XL, Naruse T, et al. Critical role of the TIE2 endothelial cell receptor in the development of definitive hematopoiesis. Immunity. 1998;9:677-686.  140  257. Choi K, Kennedy M, Kazarov A, Papadimitriou JC, Keller G. A common precursor for hematopoietic and endothelial cells. Development. 1998;125:725-732. 258. Chung YS, Zhang WJ, Arentson E, Kingsley PD, Palis J, Choi K. Lineage analysis of the hemangioblast as defined by FLK1 and SCL expression. Development. 2002;129:5511-5520. 259. Kouskoff V, Lacaud G, Schwantz S, Fehling HJ, Keller G. Sequential development of hematopoietic and cardiac mesoderm during embryonic stem cell differentiation. Proc Natl Acad Sci U S A. 2005;102:13170-13175. 260. Huber TL, Kouskoff V, Fehling HJ, Palis J, Keller G. Haemangioblast commitment is initiated in the primitive streak of the mouse embryo. Nature. 2004;432:625-630. 261. Ueno H, Weissman IL. Clonal analysis of mouse development reveals a polyclonal origin for yolk sac blood islands. Dev Cell. 2006;11:519-533. 262. D'Souza SL, Elefanty AG, Keller G. SCL/Tal-1 is essential for hematopoietic commitment of the hemangioblast but not for its development. Blood. 2005;105:3862-3870. 263. Lugus JJ, Chung YS, Mills JC, et al. GATA2 functions at multiple steps in hemangioblast development and differentiation. Development. 2007;134:393-405. 264. Garcia-Porrero JA, Godin IE, Dieterlen-Lievre F. Potential intraembryonic hemogenic sites at preliver stages in the mouse. Anat Embryol (Berl). 1995;192:425-435. 265. de Bruijn MF, Ma X, Robin C, Ottersbach K, Sanchez MJ, Dzierzak E. Hematopoietic stem cells localize to the endothelial cell layer in the midgestation mouse aorta. Immunity. 2002;16:673-683. 266. Jaffredo T, Gautier R, Eichmann A, Dieterlen-Lievre F. Intraaortic hemopoietic cells are derived from endothelial cells during ontogeny. Development. 1998;125:4575-4583. 267. Nishikawa SI, Nishikawa S, Hirashima M, Matsuyoshi N, Kodama H. Progressive lineage analysis by cell sorting and culture identifies FLK1+VE-cadherin+ cells at a diverging point of endothelial and hemopoietic lineages. Development. 1998;125:1747-1757. 268. Eilken HM, Nishikawa S, Schroeder T. Continuous single-cell imaging of blood generation from haemogenic endothelium. Nature. 2009;457:896-900. 269. Bertrand JY, Chi NC, Santoso B, Teng S, Stainier DY, Traver D. Haematopoietic stem cells derive directly from aortic endothelium during development. Nature. 2010;464:108-111. 270. Boisset JC, van Cappellen W, Andrieu-Soler C, Galjart N, Dzierzak E, Robin C. In vivo imaging of haematopoietic cells emerging from the mouse aortic endothelium. Nature. 2010;464:116-120. 271. Kissa K, Herbomel P. Blood stem cells emerge from aortic endothelium by a novel type of cell transition. Nature. 2010;464:112-115. 272. Lam EY, Hall CJ, Crosier PS, Crosier KE, Flores MV. Live imaging of Runx1 expression in the dorsal aorta tracks the emergence of blood progenitors from endothelial cells. Blood. 2010;116:909-914. 273. Labastie MC, Cortes F, Romeo PH, Dulac C, Peault B. Molecular identity of hematopoietic precursor cells emerging in the human embryo. Blood. 1998;92:3624-3635. 274. Jokubaitis VJ, Sinka L, Driessen R, et al. Angiotensin-converting enzyme (CD143) marks hematopoietic stem cells in human embryonic, fetal, and adult hematopoietic tissues. Blood. 2008;111:4055-4063. 275. Zambidis ET, Sinka L, Tavian M, et al. Emergence of human angiohematopoietic cells in normal development and from cultured embryonic stem cells. Ann N Y Acad Sci. 2007;1106:223-232. 276. Oberlin E, Tavian M, Blazsek I, Peault B. Blood-forming potential of vascular endothelium in the human embryo. Development. 2002;129:4147-4157. 277. Chen D, Wang P, Lewis RL, et al. A microarray analysis of the emergence of embryonic definitive hematopoiesis. Exp Hematol. 2007;35:1344-1357. 278. Ma F, Ebihara Y, Umeda K, et al. Generation of functional erythrocytes from human embryonic stem cell-derived definitive hematopoiesis. Proc Natl Acad Sci U S A. 2008;105:13087-13092.  141  279. Qiu C, Olivier EN, Velho M, Bouhassira EE. Globin switches in yolk sac-like primitive and fetal-like definitive red blood cells produced from human embryonic stem cells. Blood. 2008;111:2400-2408. 280. Pick M, Azzola L, Mossman A, Stanley EG, Elefanty AG. Differentiation of human embryonic stem cells in serum-free medium reveals distinct roles for bone morphogenetic protein 4, vascular endothelial growth factor, stem cell factor, and fibroblast growth factor 2 in hematopoiesis. Stem Cells. 2007;25:2206-2214. 281. D'Amour KA, Agulnick AD, Eliazer S, Kelly OG, Kroon E, Baetge EE. Efficient differentiation of human embryonic stem cells to definitive endoderm. Nat Biotechnol. 2005;23:1534-1541. 282. Vallier L, Touboul T, Chng Z, et al. Early cell fate decisions of human embryonic stem cells and mouse epiblast stem cells are controlled by the same signalling pathways. PLoS One. 2009;4:e6082. 283. Zhang P, Li J, Tan Z, et al. Short-term BMP-4 treatment initiates mesoderm induction in human embryonic stem cells. Blood. 2008;111:1933-1941. 284. Lu SJ, Feng Q, Caballero S, et al. Generation of functional hemangioblasts from human embryonic stem cells. Nat Methods. 2007;4:501-509. 285. Wang L, Li L, Shojaei F, et al. Endothelial and hematopoietic cell fate of human embryonic stem cells originates from primitive endothelium with hemangioblastic properties. Immunity. 2004;21:3141. 286. Narayan AD, Chase JL, Lewis RL, et al. Human embryonic stem cell-derived hematopoietic cells are capable of engrafting primary as well as secondary fetal sheep recipients. Blood. 2006;107:21802183. 287. Tian X, Woll PS, Morris JK, Linehan JL, Kaufman DS. Hematopoietic engraftment of human embryonic stem cell-derived cells is regulated by recipient innate immunity. Stem Cells. 2006;24:1370-1380. 288. Lee GS, Kim BS, Sheih JH, Moore M. Forced expression of HoxB4 enhances hematopoietic differentiation by human embryonic stem cells. Mol Cells. 2008;25:487-493. 289. Lu SJ, Feng Q, Ivanova Y, et al. Recombinant HoxB4 fusion proteins enhance hematopoietic differentiation of human embryonic stem cells. Stem Cells Dev. 2007;16:547-559. 290. Bowles KM, Vallier L, Smith JR, Alexander MR, Pedersen RA. HOXB4 overexpression promotes hematopoietic development by human embryonic stem cells. Stem Cells. 2006;24:1359-1369. 291. Davis RP, Ng ES, Costa M, et al. Targeting a GFP reporter gene to the MIXL1 locus of human embryonic stem cells identifies human primitive streak-like cells and enables isolation of primitive hematopoietic precursors. Blood. 2008;111:1876-1884. 292. Hart AH, Hartley L, Sourris K, et al. Mixl1 is required for axial mesendoderm morphogenesis and patterning in the murine embryo. Development. 2002;129:3597-3608. 293. Pearce JJ, Evans MJ. Mml, a mouse Mix-like gene expressed in the primitive streak. Mech Dev. 1999;87:189-192. 294. Robb L, Hartley L, Begley CG, et al. Cloning, expression analysis, and chromosomal localization of murine and human homologues of a Xenopus mix gene. Dev Dyn. 2000;219:497-504. 295. Rathjen J, Lake JA, Bettess MD, Washington JM, Chapman G, Rathjen PD. Formation of a primitive ectoderm like cell population, EPL cells, from ES cells in response to biologically derived factors. Journal of Cell Science. 1999;112:601-612. 296. Lake J, Rathjen J, Remiszewski J, Rathjen PD. Reversible programming of pluripotent cell differentiation. J Cell Sci. 2000;113 ( Pt 3):555-566. 297. Calhoun JD, Rao RR, Warrenfeltz S, et al. Transcriptional profiling of initial differentiation events in human embryonic stem cells. Biochem Biophys Res Commun. 2004;323:453-464. 298. Zandstra PW, Conneally E, Piret JM, Eaves CJ. Ontogeny-associated changes in the cytokine responses of primitive human haemopoietic cells. Br J Haematol. 1998;101:770-778. 299. WiCell. National Stem Cell Bank Basic hES Cell Culture Protocols.  142  300. Weiss MJ, Orkin SH. GATA transcription factors: key regulators of hematopoiesis. Exp Hematol. 1995;23:99-107. 301. Peerani R, Rao BM, Bauwens C, et al. Niche-mediated control of human embryonic stem cell selfrenewal and differentiation. Embo J. 2007;26:4744-4755. 302. Vallier L, Touboul T, Brown S, et al. Signaling pathways controlling pluripotency and early cell fate decisions of human induced pluripotent stem cells. Stem Cells. 2009;27:2655-2666. 303. Villegas SN, Canham M, Brickman JM. FGF signalling as a mediator of lineage transitions--evidence from embryonic stem cell differentiation. J Cell Biochem. 2010;110:10-20. 304. Ng ES, Davis RP, Hatzistavrou T, Stanley EG, Elefanty AG. Directed differentiation of human embryonic stem cells as spin embryoid bodies and a description of the hematopoietic blast colony forming assay. Curr Protoc Stem Cell Biol. 2008;Chapter 1:Unit 1D 3. 305. Lu TY, Lu RM, Liao MY, et al. Epithelial cell adhesion molecule regulation is associated with the maintenance of the undifferentiated phenotype of human embryonic stem cells. J Biol Chem. 2010;285:8719-8732. 306. Ng VY, Ang SN, Chan JX, Choo AB. Characterization of epithelial cell adhesion molecule as a surface marker on undifferentiated human embryonic stem cells. Stem Cells. 2010;28:29-35. 307. Stern CD, Canning DR. Origin of cells giving rise to mesoderm and endoderm in chick embryo. Nature. 1990;343:273-275. 308. Pfister S, Steiner KA, Tam PP. Gene expression pattern and progression of embryogenesis in the immediate post-implantation period of mouse development. Gene Expr Patterns. 2007;7:558-573. 309. Lawson KA, Meneses JJ, Pedersen RA. Clonal analysis of epiblast fate during germ layer formation in the mouse embryo. Development. 1991;113:891-911. 310. Parameswaran M, Tam PP. Regionalisation of cell fate and morphogenetic movement of the mesoderm during mouse gastrulation. Dev Genet. 1995;17:16-28. 311. Kelley C, Yee K, Harland R, Zon LI. Ventral expression of GATA-1 and GATA-2 in the Xenopus embryo defines induction of hematopoietic mesoderm. Dev Biol. 1994;165:193-205. 312. Maeno M, Mead PE, Kelley C, et al. The role of BMP-4 and GATA-2 in the induction and differentiation of hematopoietic mesoderm in Xenopus laevis. Blood. 1996;88:1965-1972. 313. Minegishi N, Ohta J, Yamagiwa H, et al. The mouse GATA-2 gene is expressed in the para-aortic splanchnopleura and aorta-gonads and mesonephros region. Blood. 1999;93:4196-4207. 314. Minegishi N, Suzuki N, Yokomizo T, et al. Expression and domain-specific function of GATA-2 during differentiation of the hematopoietic precursor cells in midgestation mouse embryos. Blood. 2003;102:896-905. 315. Amsellem S, Pflumio F, Bardinet D, et al. Ex vivo expansion of human hematopoietic stem cells by direct delivery of the HOXB4 homeoprotein. Nat Med. 2003;9:1423-1427. 316. Buske C, Feuring-Buske M, Abramovich C, et al. Deregulated expression of HOXB4 enhances the primitive growth activity of human hematopoietic cells. Blood. 2002;100:862-868. 317. Wang L, Menendez P, Shojaei F, et al. Generation of hematopoietic repopulating cells from human embryonic stem cells independent of ectopic HOXB4 expression. J Exp Med. 2005;201:1603-1614. 318. Alexander WS. Cytokines in hematopoiesis. Int Rev Immunol. 1998;16:651-682. 319. Metcalf D. Hematopoietic cytokines. Blood. 2008;111:485-491. 320. Gordon-Keylock SA, Jackson M, Huang C, et al. Induction of hematopoietic differentiation of mouse embryonic stem cells by an AGM-derived stromal cell line is not further enhanced by overexpression of HOXB4. Stem Cells Dev. 2010;19:1687-1698. 321. Lee LH, Peerani R, Ungrin M, Joshi C, Kumacheva E, Zandstra P. Micropatterning of human embryonic stem cells dissects the mesoderm and endoderm lineages. Stem Cell Res. 2009;2:155162.  143  322. Cherry SR, Biniszkiewicz D, van Parijs L, Baltimore D, Jaenisch R. Retroviral expression in embryonic stem cells and hematopoietic stem cells. Mol Cell Biol. 2000;20:7419-7426. 323. Hamaguchi I, Woods NB, Panagopoulos I, et al. Lentivirus vector gene expression during ES cellderived hematopoietic development in vitro. J Virol. 2000;74:10778-10784. 324. Pfeifer A, Ikawa M, Dayn Y, Verma IM. Transgenesis by lentiviral vectors: lack of gene silencing in mammalian embryonic stem cells and preimplantation embryos. Proc Natl Acad Sci U S A. 2002;99:2140-2145. 325. Gropp M, Itsykson P, Singer O, et al. Stable genetic modification of human embryonic stem cells by lentiviral vectors. Mol Ther. 2003;7:281-287. 326. Ma Y, Ramezani A, Lewis R, Hawley RG, Thomson JA. High-level sustained transgene expression in human embryonic stem cells using lentiviral vectors. Stem Cells. 2003;21:111-117. 327. Jang JE, Shaw K, Yu XJ, et al. Specific and stable gene transfer to human embryonic stem cells using pseudotyped lentiviral vectors. Stem Cells Dev. 2006;15:109-117. 328. Xiong C, Tang DQ, Xie CQ, et al. Genetic engineering of human embryonic stem cells with lentiviral vectors. Stem Cells Dev. 2005;14:367-377. 329. Xia X, Zhang Y, Zieth CR, Zhang SC. Transgenes delivered by lentiviral vector are suppressed in human embryonic stem cells in a promoter-dependent manner. Stem Cells Dev. 2007;16:167-176. 330. Baum C, Itoh K, Meyer J, Laker C, Ito Y, Ostertag W. The potent enhancer activity of the polycythemic strain of spleen focus-forming virus in hematopoietic cells is governed by a binding site for Sp1 in the upstream control region and by a unique enhancer core motif, creating an exclusive target for PEBP/CBF. J Virol. 1997;71:6323-6331. 331. Demaison C, Parsley K, Brouns G, et al. High-level transduction and gene expression in hematopoietic repopulating cells using a human immunodeficiency [correction of imunodeficiency] virus type 1-based lentiviral vector containing an internal spleen focus forming virus promoter. Hum Gene Ther. 2002;13:803-813. 332. Clements MO, Godfrey A, Crossley J, Wilson SJ, Takeuchi Y, Boshoff C. Lentiviral manipulation of gene expression in human adult and embryonic stem cells. Tissue Eng. 2006;12:1741-1751. 333. Carbonaro DA, Jin X, Petersen D, et al. In vivo transduction by intravenous injection of a lentiviral vector expressing human ADA into neonatal ADA gene knockout mice: a novel form of enzyme replacement therapy for ADA deficiency. Mol Ther. 2006;13:1110-1120. 334. Challita PM, Skelton D, el-Khoueiry A, Yu XJ, Weinberg K, Kohn DB. Multiple modifications in cis elements of the long terminal repeat of retroviral vectors lead to increased expression and decreased DNA methylation in embryonic carcinoma cells. J Virol. 1995;69:748-755. 335. Shaner NC, Steinbach PA, Tsien RY. A guide to choosing fluorescent proteins. Nat Methods. 2005;2:905-909. 336. Kennedy M, Keller GM. Hematopoietic commitment of ES cells in culture. Methods Enzymol. 2003;365:39-59. 337. Stewart MH, Bosse M, Chadwick K, Menendez P, Bendall SC, Bhatia M. Clonal isolation of hESCs reveals heterogeneity within the pluripotent stem cell compartment. Nat Methods. 2006;3:807-815. 338. Hough SR, Laslett AL, Grimmond SB, Kolle G, Pera MF. A continuum of cell states spans pluripotency and lineage commitment in human embryonic stem cells. PLoS One. 2009;4:e7708. 339. Hu BY, Zhang SC. Differentiation of spinal motor neurons from pluripotent human stem cells. Nat Protoc. 2009;4:1295-1304. 340. Van Hoof D, D'Amour KA, German MS. Derivation of insulin-producing cells from human embryonic stem cells. Stem Cell Res. 2009;3:73-87. 341. Eshpeter A, Jiang J, Au M, et al. In vivo characterization of transplanted human embryonic stem cellderived pancreatic endocrine islet cells. Cell Prolif. 2008;41:843-858. 342. Kroon E, Martinson LA, Kadoya K, et al. Pancreatic endoderm derived from human embryonic stem cells generates glucose-responsive insulin-secreting cells in vivo. Nat Biotechnol. 2008;26:443-452.  144  343. Spence JR, Mayhew CN, Rankin SA, et al. Directed differentiation of human pluripotent stem cells into intestinal tissue in vitro. Nature. 2010. 344. Ng ES, Davis R, Stanley EG, Elefanty AG. A protocol describing the use of a recombinant proteinbased, animal product-free medium (APEL) for human embryonic stem cell differentiation as spin embryoid bodies. Nat Protoc. 2008;3:768-776. 345. Blum B, Benvenisty N. Clonal analysis of human embryonic stem cell differentiation into teratomas. Stem Cells. 2007;25:1924-1930. 346. Gerrits A, Dykstra B, Kalmykowa OJ, et al. Cellular barcoding tool for clonal analysis in the hematopoietic system. Blood. 2010;115:2610-2618. 347. Stewart MH, Bendall SC, Levadoux-Martin M, Bhatia M. Clonal tracking of hESCs reveals differential contribution to functional assays. Nat Methods. 2010;7:917-922. 348. Lu SJ, Feng Q, Park JS, et al. Biologic properties and enucleation of red blood cells from human embryonic stem cells. Blood. 2008;112:4475-4484. 349. Ueda T, Tsuji K, Yoshino H, et al. Expansion of human NOD/SCID-repopulating cells by stem cell factor, Flk2/Flt3 ligand, thrombopoietin, IL-6, and soluble IL-6 receptor. J Clin Invest. 2000;105:10131021. 350. Boitano AE, Wang J, Romeo R, et al. Aryl hydrocarbon receptor antagonists promote the expansion of human hematopoietic stem cells. Science. 2010;329:1345-1348. 351. Ieda M, Fu JD, Delgado-Olguin P, et al. Direct reprogramming of fibroblasts into functional cardiomyocytes by defined factors. Cell. 2010;142:375-386. 352. Vierbuchen T, Ostermeier A, Pang ZP, Kokubu Y, Sudhof TC, Wernig M. Direct conversion of fibroblasts to functional neurons by defined factors. Nature. 2010;463:1035-1041. 353. Szabo E, Rampalli S, Risueno RM, et al. Direct conversion of human fibroblasts to multilineage blood progenitors. Nature. 2010;468:521-526. 354. Raya A, Rodriguez-Piza I, Guenechea G, et al. Disease-corrected haematopoietic progenitors from Fanconi anaemia induced pluripotent stem cells. Nature. 2009;460:53-59. 355. Wang Y, Jiang Y, Liu S, Sun X, Gao S. Generation of induced pluripotent stem cells from human beta-thalassemia fibroblast cells. Cell Res. 2009;19:1120-1123. 356. Ye L, Chang JC, Lin C, Sun X, Yu J, Kan YW. Induced pluripotent stem cells offer new approach to therapy in thalassemia and sickle cell anemia and option in prenatal diagnosis in genetic diseases. Proc Natl Acad Sci U S A. 2009;106:9826-9830. 357. Ye Z, Zhan H, Mali P, et al. Human-induced pluripotent stem cells from blood cells of healthy donors and patients with acquired blood disorders. Blood. 2009;114:5473-5480. 358. Kumano K, Arai S, Ueda K, Nakazaki K, Kamikubo Y, Kurokawa M. Generation of Induced Pluripotent Stem Cells From Primary Chronic Myelogenous Leukemia Patient Sample. Blood. 2010;116:518.  145  


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