UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Site-directed integration using the Cre/lox system in hematopoietic and embryonic stem cells Faulkes, Sharlene Marie Sep 22, 2002

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

Item Metadata

Download

Media
831-ubc_2002-731596.pdf [ 7.56MB ]
Metadata
JSON: 831-1.0090532.json
JSON-LD: 831-1.0090532-ld.json
RDF/XML (Pretty): 831-1.0090532-rdf.xml
RDF/JSON: 831-1.0090532-rdf.json
Turtle: 831-1.0090532-turtle.txt
N-Triples: 831-1.0090532-rdf-ntriples.txt
Original Record: 831-1.0090532-source.json
Full Text
831-1.0090532-fulltext.txt
Citation
831-1.0090532.ris

Full Text

SITE-DIRECTED INTEGRATION USING THE CRE/LOX SYSTEM IN HEMATOPOIETIC AND EMBRYONIC STEM CELLS By Sharlene Marie Faulkes B.Sc. (Honors) Genetics, University of Manitoba, 1995 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY In THE FACULTY OF GRADUATE STUDIES Medical Genetics Program We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA January, 2002 © Sharlene Marie Faulkes, 2002 UBC Special Collections - Thesis Authorisation Form Page 1 of 1 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of /ilcJJOft < Qt^cScS The University of British Columbia Vancouver, Canada Date f<JL> *L\bT-http://www.library.ubc.ca/spcoll/thesauth.html 2/5/2002 ABSTRACT Retroviral vectors have been most commonly used for performing gene function studies in hematopoietic cells and exploited for gene therapy of hematological inherited disorders. While unquestionably powerful, they suffer from several drawbacks including size limitations and the ability to confer long-term, predictable levels of transgene expression. The goal of this thesis was to develop an alternative gene transfer strategy that may overcome some of these hurdles and further provide an improved platform for performing gene function and gene regulation studies in hematopoietic cells. This strategy is based on the Cre/lox recombination system and can be summarized in two steps: (i) a simple retroviral vector is first used to introduce lox P target sites into the genome of target cells, (ii) followed by Cre mediated integration of exogenous DNA into the chromosomally placed lox P site. Initial studies demonstrated that site directed integration was feasible and applicable to many cell types, as it was shown to occur in both non-hematopoietic and hematopoietic cell lines, pluripotent embryonic stem (ES) cells, and in hematopoietic progenitors derived from the ES in vitro differentiation system. Moreover, this strategy is rapid, efficient and can be exploited to achieve predictable expression of a transgene upon re-targeting a locus. Gene regulatory mechanisms in hematopoietic cells were also studied using the ES in vitro differentiation model. A series of human p" globin expression vectors that are of interest for gene therapy of the hemoglobinopathies were assessed. The ES system was shown to be a potent model as the development of erythroid progenitors during differentiation correlated with the profile of globin gene expression and a large number of ii mature erythroid cells for RNA/protein analysis were readily generated. The Cre/lox system was shown to, be a more critical and informative method of evaluating constructs compared to random integration, since analysis was performed at the same integration site (assess transcription levels) and at more than one integration site (assess position effects). Moreover, Cre mediated integration was achieved with human P globin constructs that have been previously reported to be unstable in retroviral vectors and the size of which go beyond the current limitations for retroviral vectors. in TABLE OF CONTENTS ABSTRACT ii TABLE OF CONTENTS iv LIST OF FIGURES viLIST OF TABLES x ABBREVIATIONS x ACKNOWLEDGMENTS xii CHAPTER 1 Introduction 1 1.1 Overview1.2 Hematopoietic stem cells 4 1.2.1 Existence of hematopoietic stem cells 4 1.2.2 Ontogeny of hematopoiesis 8 1.2.2.1 Murine hematopoiesis (in vivo model) '. 8 1.2.2.2 Embryonic stem cell hematopoiesis (in vitro model) 10 1.3 Genetic manipulation of HSCs 14 1.3.1 Gene addition strategies1.3.1.1 Existing viral approaches 6 1.3.1.1.1 Adeno-associated viral vectors 11.3.1.1.2 Retroviral vectors 17 1.3.1.1.3 Lentiviral vectors 21 1.3.1.1.3.ILimitations of retroviral vectors 23 1.3.1.1.3.2 Position effects/silencing 5 1.3.1.2 Existing non-viral approaches 32 1.3.1.2.1 Electroporation > 32 1.3.1.3 Potential approaches 34 1.3.1.3.1 Extra-chromosomal replicating vectors 31.3.1.3.1.1 Mammalian artificial chromosomes 4 1.3.1.3.1.2 Epstein bar virus based vectors 35 1.3.1.3.2 Hybrid vectors 31.3.2 Gene targeting: 6 1.3.2.1 Approaches to improve gene targeting techniques 38 1.4 Site specific recombinases 42 1.4.1 Cre/lox recombination system 43 1.4.2 Genetic engineering of the genome 5 1.4.2.1 Deletions: inducible and tissue specific 41.4.2.2 Integration of exogenous DNA 47 1.4.2.2.1 Approaches to improve integration efficiency.... 48 1.4.2.3 Cre/lox system for gene regulation studies 51 1.5 Gene therapy for inherited disorders of the hematopoietic system 52 1.5.1 Hemoglobinopathies 56 1.5.1.1 Development of globin gene vectors for the hemoglobinopathies. 61 1.6 Thesis objectives/general strategy 64 iv CHAPTER 2 Materials and Methods ? 66 2.1 Cell lines, embryonic stem cell culture and assays 62.1.1 Cell lines 62.1.2 Murine embryonic stem (ES) cell culture and assays 67 2.1.2.1 Maintenance of undifferentiated ES cells 62.1.2.2 In vitro differentiation of ES cells2.1.2.3 Generation of ES derived erythroid cells 68 2.1.2.4 Generation of ES derived mast cells 9 2.2 Gene transfer techniques 70 2.2.1 Retroviral generation and infection of cells 72.2.2 Electroporation 1 2.2.2.1 Site directed integration 72.2.2.2 Random integration 2 2.3 Molecular analysis 73 2.3.1 DNA constructs2.3.2 DNA analysis 5 2.3.2.1 Isolation of DNA2.3.2.2 Southern blot analysis and probes used 76 2.3.2.3 PCR analysis 72.3.3 RNA analysis 7 2.3.3.1 Isolation of RNA2.3.3.2 RT-PCR 8 2.3.3.3 RNASE protection : 79 2.4 FACS analysis 80 2.4.1 Staining for FACS analysis 82.4.1.1 FACS analysis and sorting 1 CHAPTER 3 Site directed integration in hematopoietic cells using the Cre/lox recombination system 82 3.1 Introduction3.2 Results 4 3.2.1 Feasibility of Cre/lox site directed integration 83.2.2 Site directed integration in the Ba/F3 hematopoietic cell line 90 3.2.3 Site directed integration in ES and ES derived hematopoietic cells 99 3.3 Discussion 10CHAPTER 4 Evaluation of gene regulatory mechanisms in hematopoietic cells using the in vitro diffrentiation of ES cells combined with the Cre/lox system 114 4.1 Introduction , 114 4.2 Results 117 4.2.1 Characterization of erythroid development and globin gene expression in differentiating ES cells 114.2.2 Comparison of expression levels of human P globin constructs integrated either randomly or at integration sites mediated by the Cre/lox system 121 v 4.3 Discussion 138 CHAPTER 5 Discussion 144 CHAPTER 6 References 151 vi LIST OF FIGURES CHAPTER 1 Figure 1-1 Schematic representation of the hematopoietic hierarchy 2 Figure 1-2 Kinetics of hematopoietic development within embryoid bodies 12 Figure 1-3 Differentiation assays used for analyzing hematopoietic cells following in vitro differentiation 13 Figure 1-4 Production of an infectious recombinant retroviral vector 20 Figure 1-5 Cre/lox recombination system from bacteriophage PI 44 Figure 1-6 Cre/lox mediated integration 49 Figure 1-7 Schematic detailed diagram of Cre/lox mediated integration using heterospecific lox P sites 50 Figure 1-8 Schematic diagram of the globin gene loci 58 CHAPTER 3 Figure 3-1 Schematic representation of Cre/lox mediated integration 86 Figure 3-2 Cre mediated integration is dependent on expression of Cre recombinase and the presence of lox P sites on the plasmid 88 Figure 3-3 Southern blot analysis confirms Cre mediated integration in a clone derived from the ES (CCE cell line) (left) and the murine J3a/F3 hematopoietic cell line (right) 89 Figure 3-4 Cre mediated integration using the promoter-less GFP-lox targeting plasmid in a Ba/F3 hematopoietic cell clone 92 Figure 3-5 Cre recombinase mediates rapid integration in a Ba/F3 hematopoietic clone 94 Figure 3-6 Predictable expression is achieved following Cre mediated re-targeting 97 Figure 3-7 Representative Southern blot analysis using the highest expressing Ba/F3 clone 98 Figure 3-8 Site directed integration occurs in undifferentiated ES cell clones 100 Figure 3-9 Schematic representation of the strategy used to test for Cre mediated integration in hematopoietic progenitors 103 Figure 3-10 FACS analysis on sorted ES derived hematopoietic cells to confirm stable integration and cells are of hematopoietic origin 105 Figure 3-11 Representative hematopoietic colonies from sorted GFP positive cells that received the promoter-less lox GFP targeting plasmid and Cre expression plasmid 106 Figure 3-12 PCR analysis confirms site directed integration in hematopoietic progenitor colonies 108 CHAPTER 4 Figure 4-1 Analysis of endogenous (5 globin gene expression during ES in vitro differentiation 118 Figure 4-2 Characterization of erythroid progenitors during ES in vitro differentiation 119 vii Figure 4-3 FACS profiles using the Terl 19 marker following expansion of ES cells in general hematopoietic liquid culture conditions (A) and specialized erythroid liquid culture conditons (B) 121 Figure 4-4 Schematic diagram of the human p globin constructs 122 Figure 4-5 Southern blot analysis to confirm random integration of the human p globin constructs lacking heterospecific lox P target sites 123 Figure 4-6 Kinetics of human P globin gene expression during ES in vitro differentiation 124 Figure 4-7 RNASE protection assay (RPA) to quantify the level of human P globin mRNA relative to mouse pmajor for human p globin constructs that were integrated at random 126 Figure 4-8 Flow cytometric analysis demonstrating minimum detecton of human p globin protein in differentiating ES cells 127 Figure 4-9 Southern blot analysis of the two randomly chosen ES clones following retroviral gene transfer of the heterospecific lox P target sites 128 Figure 4-10 Schematic diagram demonstrating the expected band sizes following Cre/lox mediated integration of the human P globin constructs 130 Figure 4-11 Southern blot anlaysis confirms Cre/lox integration of the human p globin constructs at sites A2 and A8 131 Figure 4-12 RNASE protection analysis (RPA) to quantify the level of human P mRNA expression relative to mouse P major for human p globin constructs integrated at sites A2 and A8 using the Cre/lox system 132 Figure 4-13 Graph summarizes the results obtained from the RNASE protection assays illustrated in Figure 4.12 (at sites A2 and A8) and Figure 4.7 (random) 134 Figure 4-14 Flow cytometric analysis demonstrating high levels, erythroid-specific expression of the human P globin protein in differentiating ES cells 136 Figure 4-15 RT-PCR on differentiating ES cells demonstrating erythroid-specific expression of the human p globin constructs 137 viii LIST OF TABLES CHAPTER 1 Table 1-1 Comparison of properties of various vector systems 15 Table 1-2 Evidence for position effects following retroviral gene transfer of neomycin-phosphotransferase into human hematopoietic cells 26 Table 1-3 Candidate diseases for gene therapy of the hematopoietic system 54 CHAPTER 3 Table 3-1 Accuracy of Cre mediated integration 90 Table 3-2 Average site directed integration frequency in the Ba/F3 hematopoietic cell line 95 Table 3-3 Average site directed integration frequency in the ES cell lline 101 Table 3-4 Site directed integration frequencies in hematopoietic progenitors 105 CHAPTER 4 Table 4-1 Comparison of erythroid liquid culture conditions 120 \ ix ABBREVIATIONS 5-azaC 5-AzaCytidine AAV adeno-associated virus AC artificial chromosome ADA adenosine deaminase AGM aorta-gonad mesonephros region ATCC American Type Culture Collection BM bone marrow bp base pair BSA bovine serum albumin CD cluster designation cDNA complementary deoxyribonucleic acid CAG chicken P actin CFC colony forming-cells CFU-E colony forming unit-erythroid CFU-GEMM colony forming unit-granulocyte-erythroid-monocyte-megakaryocyte CFU-GM colony forming unit-granulocyte-monocyte CFU-MEG colony forming unit-megakaryocyte cfu/ml colony formining unit/ml CFU-S colony forming unit-spleen CLP common lymphoid progenitor CMP common myeloid progenitor CMV cytomyegalovirus DMEM Dulbecco's Modified Eagles Medium DNA deoxyribonucleic acid EBs embryoid bodies EBNA-1 epstein bar virus nuclear antigen-1 EBV epstein bar virus EC embryonic carcinoma cells EDTA ethylene diaminetetraacetic acid EPO erythropoietin ES embryonic stem cells FACS fluorescence activated cell sorter FCS fetal calf serum FL fetal liver G418 geneticin GF growth factor GFP green fluorescent protein GT gene targeting Hb hemoglobin HbA adult hemoglobin FfbF fetal hemoglobin HbS sickle hemoglobin HTV-1 human immunodeficiency virus type-1 HPRT ' hypoxanthine phosphoribosyl-transferase HSC hematopoietic stem cell HS hypersensitive site HSV-tk herpes simplex virus-thymidine kinase hygrotk ^ hygromycin resistance-thymidine kinase IL interleukin IMDM Iscove's Modified Dulbecco's Medium IN integrase kb kilobase LCR locus control region LEF leukemia inhibitory factor LTC-IC long-term culture initiating cells LTR long terminal repeat LV lenti viral MAR matrix attachment region MEL mouse erythroleukemia cell line mfi mean fluorescence intensity MoMLV Moloney murine leukemia virus rnRNA messenger RNA MSCV murine stem cell virus mSF murine steel factor (stem cell factor) MTG monothioglycerol neo neomycin phosphotransferase NOD/SCID non-obese diabetic severe combined immunodeficiency PBS Phosphate Buffered Saline PCMV PCC4 embryonal carcinoma cell-passaged myeloproliferative sarcoma virus PCR polymerase chain reaction PE phycoerythrin PS . para-aortic splanchnopleure RDO RNA/DNA double stranded oligonucleotide RNA ribonucleic acid RPA RNASE protection assay RPMI Roswell Park Memorial Institute RT reverse transcriptase RV retroviral SAR scaffold attachment region Scal+ stem cell antigen-1 SCD sickle cell disease SCID severe combined immunodeficiency TAG large T antigen TFO triplex forming oligonucleotide VSV-G vesicular stomatitis virus G glycoprotein YAC yeast artificial chromosome YS yolk sac xi ACKNOWLEDGMENTS I would like to thank and express my gratitude: to my supervisor Dr. K. Humphries for his support and guidance throughout this project, to all the members of the Humphries laboratory for their help and making life during my Ph.D fun, to Drs. Dixie Mager, Rob Kay and Suzanne Lewis for serving on my graduate committee, to Patty Rosten, Cheryl Helgason and Margaret Hale for expert technical assistance, to my parents for their continuous support and providing me with the strength to keep going, And lastly to Nicolas Pineault, who always believed in me, and made me look at the positive side of life, and who got me through the most difficult times in this work. xii CHAPTER 1 INTRODUCTION 1.1 Overview Hematopoiesis is the essential, lifelong process whereby multiple types of highly specialised blood cells are generated. These cells can be functionally divided into two distinct groups termed myeloid and lymphoid. The myeloid lineage includes those cells responsible for carbon dioxide and oxygen transport (erythrocytes), blood clotting (platelets), and those involved in mounting a phagocytic response to foreign organisms (granulocytes, monocytes, macrophages). Cells that are involved in humoral (B-lymphocytes) and cellular immunity (T cells, Natural Killer cells) comprise the lymphoid lineage. Most of the above types of mature blood cells are short-lived, and therefore must be continuously replenished throughout the life of the animal. This is accomplished by a rare set of cells (estimated to comprise only 0.01% of the total marrow compartment), called pluripotent hematopoietic stem cells (HSCs). These cells have two important properties; the ability to self renew, giving rise to new pluripotent stem cells, and through extensive proliferation and division over many cell generations to give rise to all the mature and heterogeneous cells found in the blood (Figure 1.1). These critical properties have generated interest in deciphering the regulatory mechanisms that underlie HSC behaviour and in their use therapeutically (ie. bone marrow transplantation) for hematological disorders. A number of studies have suggested that both intrinsic and extrinsic regulatory mechanisms appear to control these HSC properties. The ability to genetically manipulate HSCs provides a valuable tool for elucidating the intrinsic regulatory role through gene function studies and also serves as a potential cure for 1 i I Animal repopulation (4-6 months) a (A C fe .9- 5 G bo 1 2 CMP CFU-S day 12 (4 weeks) M o c o u ten on S d o. u c J • thymus o CFU-GEMM CFU-GM / (granulocyte-monocyte) o Ql CFU-E (erythroid) CFU-MEG — (megakaryocyte) CFC assay T cells B cells monocytes 1 NK cells granulocytes erythrocytes platelets morphology Figure 1-1 Schematic representation of the hematopoietic hierarchy. Shown on the left are the different maturation stages, and on the right are the various assays used to measure the different populations of hematop'oietic cells. CFU-S day 12, colony-forming unit spleen day 12; CFC, colony-forming assay; CLP, common lymphoid progenitor; CMP, common myeloid progenitor; CFU-GEMM, colony forming unit-granulocyte-erythroid-monocyte-megakaryocyte; CFU-GM, colony forming unit-granulocyte-monocyte; CFU-E, colony forming unit erythroid; CFU-MEG, colony forming unit-megakaryocyte; NK cells, natural killer cells. genetic abnormalities of the hematopoietic system (gene therapy). Many different DNA transfer strategies have been described, but those directed towards the hematopoietic system must be able to efficiently target and stably integrate a gene into the genome of HSCs, so as to achieve sustained, high levels of transgene expression in the more mature hematopoietic lineages. The ideal approach would be to use gene targeting or homologous recombination to integrate a gene into a specific location in the genome, as clinically this approach would lead to correction of a genetic defect. To date, however, 2 this approach is not yet achievable, as the frequency of gene targeting is extremely low. Therefore, investigators have focused on gene addition strategies in which a functional/therapeutic gene is introduced at random into the genome. Retroviral vectors are the most commonly used and have proven to be a powerful tool for performing gene function studies and have also been exploited for gene therapy. However, they do suffer from several drawbacks, which include limitations on the size of the transgene that can be packaged in the viral genome and the lack of long-term, predictable transgene expression. The existence of such problems have led to defining better cis-regulatory elements that may overcome silencing and position effects. Improved model systems to systematically explore and identify optimal regulatory elements and vector design would be of considerable benefit. The overall goal of the work presented in this thesis was to develop an alternative gene transfer strategy that may overcome some of the above hurdles, and further provide an improved platform for performing gene function and gene regulation studies in hematopoietic cells. The strategy explored utilises a site specific recombination system called Cre/lox that has been used for performing site specific modifications in mammalian cells. This method presents several potentially powerful uses that include: 1) the introduction of large DNA constructs, 2) the ability to achieve more predictable expression by integration of a transgene into a site with desired expression properties (gene function studies), and 3) a more accurate analysis of cis-regulatory sequences by studying constructs at the same genomic integration site (gene regulation studies). The feasibility of this strategy was first examined in a number of cell lines, including hematopoietic, and primitive pluripotent embryonic stem cells, and subsequently in more 3 primitive cells using the ES in vitro differentiation model. The feasibility of using this to assess hematopoietic regulatory mechanisms was then explored using the ES in vitro differentiation system and p globin gene expression as a model system. As a background to these studies, the following introduction provides a brief overview of the hematopoietic system and the model systems used to study it. This is then followed by a summary of the challenges and existing and potential approaches to genetic manipulation of hematopoietic cells. 1.2 Hematopoietic stem cells Hematopoietic stem cells are undifferentiated, unspecialised cells that can renew themselves and give rise to all of the cells that comprise the blood and the immune system. To date, they have been the most thoroughly studied among all organ or tissue specific somatic stem cells (Anderson et al., 2001). 1.2.1 Existence of hematopoietic stem cells In the 1960s, thymidine labelling studies of re-infused labelled blood cells made it evident that these cells were short lived. In order to maintain normal levels, the continuous formation of new cells was required to replenish these cells, and it was established that this process occurred in the bone marrow (reviewed in Metcalf, 2001). The understanding of the cellular basis of hematopoiesis was advanced by Till and McCulloch in the early 1960's, when they injected marrow cells into irradiated mice (to remove any endogenous hematopoietic cells), and found these cells generated large macroscopic colonies in the spleens of the irradiated mice (Till and McCulloch, 1961). 4 These colonies were found to be the clonal progeny of single cells as evidenced from all the cells within a colony having the same marker (eg. a radiation induced chromosomal marker (Wu et al, 1968), or more recently, proviral marking (Dick et al., 1985; Lemischka et al, 1986)). These colonies derived from single cells are called CFU-S for colony forming unit-spleen, and were composed of multiple lineages - including mostly myeloid and erythroid. More importantly, some of these colonies also contained CFU-S, as they were capable of forming additional spleen colonies upon subsequent re-transplantation, indicating that the original cells were also capable of self-generation (Till and McCulloch, 1961). Subsequent studies showed that CFU-S could be isolated or enriched from the other marrow cells by physical separation methods, giving rise to the concept that "stem cells" are minor subsets of marrow cells that can generate multiple lineages and self renew (Curry and Trentin, 1967). The development of semi-solid medium to grow bone marrow cells provided another advance in the characterisation of the cellular basis of hematopoiesis. Cells were plated in a semi-solid medium (methylcellulose supplemented with growth factors is now used), and were found to give rise to colonies that varied in size and in the types of cells that comprised each one (Bradley and Metcalf, 1966). These cells were found to have a poor capacity for self -renewal because no or very few colonies were formed upon secondary re-plating, and were further shown to be clonal (Metcalf and Moore, 1971). These are referred to as colony forming cells (CFC) or lineage committed progenitor cells. These lineage committed progenitors were found to produce one (uni) to many (multi) lineages in which the greater the colony size and content of the different lineages, the more primitive the cell from which the colony is derived. For example, a multi-lineage progenitor, CFU-GEMM (colony forming unit - granulocyte, erythroid, macrophage, megakaryocyte) gives rise to a very large colony composed of granulocytes, macrophages, erythrocytes and megakarocytes, and is more primitive than a uni-lineage progenitor, CFU-E (colony forming unit - erythroid) which gives rise to smaller colonies of only erythroid cells (Metcalf, 1984). Cell separation techniques again showed it was possible to separate these cells, and found they were more numerous than CFU-S and could be detected in spleen colonies as the progeny of CFU-S (Metcalf, 1984). These observations gave rise to concept of the hierarchy of the hematopoietic system, in which a stem cell gives rise to more abundant progenitor cells, which in turn generates more mature cells with specialised functions (Metcalf, 2001). Subsequent studies later revealed that CFU-S are unable to repopulate irradiated recipients on a long-term basis; mice survive for only 4 weeks (Hodgson and Bradley, 1979). An even smaller subset of cells were identified as having this potential and were called pluripotent hematopoietic stem cells (HSCs) and represent the "true stem cells" responsible for hematopoiesis in an adult animal. HSCs are found at a frequency of 1/104 bone marrow cells (Szilvassy et al., 1990), are quiescent (survive treatment with chemotherapeutic drugs) and can be separated from the other hematopoietic cell types based on distinct phenotypic markers (Metcalf, 2001). Currently, HSCs cannot yet be positively identified on the basis of any unique morphological, physical or cell surface characteristics, rather a number of different phenotypic markers are used in combination to enrich for them. Monoclonal antibodies directed against specific cell surface antigens have identified several markers used for enrichment that include: Thy10, ckit+ (tyrosine kinase receptor), and Scal+ (phosphoinositol anchored protein). Moreover, HSCs do not 6 express any lineage markers found on mature hematopoietic cells (referred to as lineage" or Lin") and have been shown to efficiently efflux a mitochondrial dye called Rhodamine via a P glycoprotein pump. Therefore, mouse HSCs can be enriched and isolated from adult bone marrow cells using the phenotype: Scal+, ckit+, Thy10, Lin", Rholo/neg (reviewed in Metcalf, 2001). Due to the lack of a surface marker that is specifically expressed on HSCs, functional assays are the most reliable and rigorous method for their assessment. There are two such assays developed in the murine model for the identification and quantitation of HSCs; the competitive repopulation assay developed by Harrison (Harrison, 1980), and the limit dilution assay developed by Szlivassy (reviewed in (Metcalf, 1984)). These are based on the ability of these cells to regenerate and sustain the hematopoietic system of lethally irradiated mice. The detection of the same unique marker in both myeloid and lymphoid lineages of lethally irradiated mice is indicative of a common origin and confirms the existence of a HSC (Metcalf, 2001). Figure 1.1 summarises the current knowledge of the hematopoietic system. The rare HSCs give rise to multipotential progenitors called the common lymphoid progenitor (CLP) which gives rise to all the lymphoid cells (B, T, Natural Killer cells), and the common myeloid progenitor (CMP) which gives rise to all the myeloid cells (granulocytes, monocytes, erythrocytes and megakarocytes) through the appropriate lineage committed progenitors (Akashi et al., 2000). Thus, the hematopoietic system develops into a hierarchy of cells with decreasing proliferative and increasing differentiation capacities through successive rounds of cell division. 7 1.2.2 Ontogeny of hematopoiesis 1.2.2.1 Murine hematopoiesis (in vivo model) Interaction between endodermal and ectodermal germ layers induces the formation of the mesodermal germ layer at day 6.5 of gestation. The mesodermal germ layer can be found extraembryonically in a structure called the yolk sac (YS) and as well within the embryo, and is responsible for the formation of hematopoietic cells (Dzierzak, 2001). Murine embryonic hematopoiesis is a dynamic process in which the sequential emergence of distinct populations of blood cells occurs within different anatomical sites. In the first wave, the initial visible evidence of blood cells is observed at day 7.5 in the blood islands (structure composed of endothelial cells, primitive erythrocytes and underlying endoderm) of the YS. These primitive erythrocytes are large, nucleated erythroid cells that express embryonic globin and once circulation is established at day 8.5, they persist until late gestation. These cells are responsible for supplying oxygen and nutrients to tissues throughout the embryo prior to the development of the liver. Since these cells are the only mature hematopoietic component of the yolk sac and are only produced during this stage, this first wave is referred to as primitive or embryonic hematopoiesis (Keller et al., 1999). The second wave of hematopoiesis occurs within the embryo and is referred to as adult or definitive hematopoiesis and is responsible for generating HSCs and all other hematopoietic lineages observed in the adult system. Interestingly, the erythroid cells generated here are distinct from the primitive erythrocytes in that they are smaller, enucleated and express adult type globins and are therefore called definitive erythroid 8 cells (Keller et al, 1999). The origin of HSCs has been extensively studied and remains controversial. It is thought that HSCs originate from the trunkal/abdominal region within the embryo referred to as the PS/AGM region (para-aortic splanchnopleure - occurs between day 7-9; aorta-gonad-mesonephros - occurs between day 9-11) (Dzierzak, 2001). Evidence for this comes from the findings that HSCs that are capable of engrafting adult recipients (adult HSCs) are first detected in the AGM region at day 10, one day earlier than for the YS (Muller et al, 1994). Explant cultures of fetal liver (FL), YS or body remnants at day 10 have no potential to repopulate adult recipients, demonstrating no HSCs in these tissues (Medvinsky and Dzierzak, 1996). Thus, the current theory has been that HSCs from the AGM region seed to the FL, then to the spleen, and then bone marrow, which are responsible for seeding the adult blood system. Opposition to this theory came from studies done by Yoder et al. They isolated cells from the YS and AGM region at day 9 (before HSCs have developed), and found that both could repopulate newborn mice and then secondary adult recipients. Many more (37 times) were found in the YS than in the AGM region leading them to speculate that HSCs are produced in the YS (Yoder et al, 1997). This suggests that HSCs are present in the YS but are not functionally mature and that the AGM region does not produce HSCs, but rather provides the necessary microenvironment for their induction. The caveat with this study is by day 9, circulation has already begun making it hard to determine which tissue is the initiator and which is the one being colonized. Alternatively, the AGM region may generate a distinct population of HSCs that contributes to the fetal and then the adult hematopoietic system. Presently, it is not possible to distinguish between these two scenarios. However, 9 the ability to track the migration of cells in the embryo in situ may help to resolve this issue (Keller ef a/., 1999). 1.2.2.2 Embryonic stem cell hematopoiesis (in vitro model) Embryonic stem (ES) cells are pluripotent cells derived from the inner cell mass of mouse blastocysts and can be maintained indefinitely in an undifferentiated state in culture (Evans and Kaufman, 1981; Martin, 1981). When ES cells are injected into mouse blastocyts, these cells contribute to all tissues, including the germ cells of the developing animal (Bradley et al., 1984). ES cells are maintained in vitro by co-culture on irradiated mouse fibroblasts (Martin, 1981) or on gelatinized dishes with a differentiation inhibitory factor called leukemia inhibitory factor (LIF) (Williams et al., 1988). Upon withdrawal of cells from the feeder layer or LIF, ES cells will differentiate into complex, spherical structures called embryoid bodies (EBs) (Weiss and Orkin, 1996). EBs contain differentiated cells of all lineages which appear in a well defined temporal pattern, similar to that observed in vivo. For example, within the first few days of differentiation, EBs generate a population of cells that express genes indicative of the primitive endoderm and mesoderm. After extended periods of culture, EBs can generate cells of hematopoietic, endothelial, muscle, and neuronal lineages (Keller, 1995). Of these, hematopoiesis within the EBs has been studied most extensively. The formation of hematopoietic cells within EBs was first described by Tom Doetschman. He noted that upon differentiation of ES cells in suspension, a small percentage of EBs could develop islands of primitive erythrocytes (ie. blood islands) (Doetschman et al., 1985). Micheal Wiles and Gordon Keller further demonstrated an increase in both the efficiency of EBs that formed 10 hematopoietic cells and in the range of hematopoietic cells observed, was achieved through the addition of different growth factor combinations. As well, hematopoiesis in EBs was found to occur in distinct waves; in whole EBs expression of the embyronic p globin gene (PHI), precedes that of the adult form (P major) (Wiles and Keller, 1991). The in vitro differentiation approach was further refined by the development of a two plating assay, which allows the quantification and determination of the frequency and types of hematopoietic progenitors that are present in the EBs at different stages of differentiation. This helped to define the hematopoietic differentiation program in EBs and revealed that it strongly correlates with that observed in vivo (Figure 1.2). For example, this assay demonstrated that separate populations of erythroid precursors appear in a characteristic temporal fashion. Primitive erythrocytes arise at ~ day 3.5 of EB development and yield small colonies that synthesise embryonic globin, whereas definitive erythroid cells appear later and generate larger colonies that contain adult globins (Keller et al, 1999). 11 Lineages Detected Myeloid and multipotential progenitors Primitive erythroid (embryonic globin, BH1) 0 4 12 14 24 6 8 10 day of differentiation Figure 1-2 Kinetics of hematopoietic development within embryoid bodies. The onset and types of the various hematopoietic lineages within embryoid bodies during in vitro differentiation correlates with that observed in vivo. Compiled from Keller et al. 1993, and Keller et al. 1999. Figure 1.3 summarises the different techniques used for analysing hematopoietic cells following in vitro differentiation: 1) hematopoietic differentiated populations can be identified in intact EBs by direct morphological analysis, histological analysis or RNA analysis; 2) EBs can be dissociated with trypsin and/or collagenase into single cells and plated in methylcellulose with appropriate growth factors (GFs) for the presence of hematopoietic progenitors that can be enumerated and analysed (referred to as the two plating assay); and 3) disrupted EBs can also be placed in liquid culture with hematopoietic GFs for the generation of mature hematopoietic cells (erythroid, macrophages, granulocytes, etc). Detection of these cells is performed by staining with various antibodies for the presence of lineage specific cell surface antigens, followed by analysis using the fluorescence activated cell sorter (FACS) (Keller, 1995). 12 Analyze intact EBs (RNA, protein level) LIF <zpsp cpcp in methylcellulose for hematopoietic progenitors Dissociate EBs and plate ES cells fetal calf developing embryoid bodies (EBs) serum differentiate in methylcellulose o oee o cH mature hematopoietic cells Dissociate EBs and plate in liquid culture with hematopoietic growth factors. Assay for the presence of mature hematopoietic lineages by specific surface markers. Figure 1-3 Differentiation assays used for analyzing hematopoietic cells following in vitro differentiation. Upon withdrawl of LIF, ES cells form embryoid bodies that can be analyzed directly (1), dissociated and placed in methylcellulose for hematopoietic progenitors (2), or dissociated and plated in liquid culture for mature hematopoietic cells (3). Very few studies have demonstrated the detection of primitive HSCs within the ES system. In one report by Palacios et al, ES cells were first differentiated on a stromal cell line and in a GF cocktail that supports the growth of hematopoietic cells. These were then injected into severe combined immunodeficient (SCLD) mice that lack T and B cells, and cells of ES origin were found to generate both myeloid and lymphoid lineages in primary and secondary recipients 16-20 weeks post-transplantation, suggesting the presence of ES derived HSCs (Palacios et al, 1995). The previous reports on the lack of multilineage re-population with cells of ES origins involved direct injection of ES cells into adult irradiated mice (Keller, 1995). The fact that reconstitution was achieved in the above study with a stromal layer suggests ES cells can generate HSCs if provided the "appropriate or optimum" microenvironment. 13 In summary, ES cells can be grown and expanded to large numbers in an undifferentiated state thus making them amenable to genetic manipulation and further provides a system for testing genetic modifications in primitive hematopoietic cells. As well, ES in vitro differentiation is an accurate model of in vivo hematopoiesis that is capable of generating a large number of different hematopoietic populations and therefore provides an in vitro system for studying gene regulatory mechanisms in hematopoietic cells. 1.3 Genetic manipulation of HSCs HSCs have the capacity to maintain lifelong, blood cell production and reconstitute the entire hematopoietic system. This makes them key targets for genetic manipulations to understand how these cells are regulated (gene function studies) and as a possible long-lasting gene therapy treatment for hematological inherited disorders. Genetic modifications can be divided into two categories: 1) gene addition in which a therapeutic or functional gene is introduced randomly into the genome and 2) gene targeting in which a gene is integrated into a specific location in the genome. 1.3.1 Gene addition strategies DNA transfer systems used for gene addition strategies can be roughly divided into two classes - those that are viral mediated (transduction) and those that are non-viral mediated (transfection) (reviewed in Van Tendeloo et al., 2001). Table 1.1 summarizes the properties of various vector systems that are currently being used. This section will focus only on those vector systems that are capable of achieving stable integration of a gene into the genome of HSCs. 14 Table 1-1 Comparison of properties of various vector systems Features Retroviral Lentiviral Adenoviral AAV Non-viral Type of virus ss RNA ss RNA ds DNA ss DNA Maxium insert size 8kb 8kb 30 kb 5kb unlimited Integration Yes Yes No Yes/No very poor Infect non-and dividing cells dividing both both both both Duration of expression (in vivo) short long short long short Host range broad broad broad broad broad Immunological problems* few few extensive few none * Immune response determined upon repeated administration of vector. AAV, adeno-associated viral vector; ss, single-stranded; ds, double-stranded. Compiled from Van Tendeloo et al, 2001 and Kay et al, 2001. 15 1.3.1.1 Existing viral approaches A viral vector is a viral particle encapsulated with a modified genome containing a therapeutic/functional gene cassette in place of the viral genome. Viral vectors are the most efficient way of introducing genetic material into cells. This is because viruses have a natural ability to infect cells, avoid degradation associated with the endosomal pathway, can be translocated to the nucleus and integrated into the genome (reviewed in Kay et al, 2001; Van Tendeloo et al, 2001). 1.3.1.1.1 Adeno-associated viral vectors Adeno-associated virus (AAV) is a single stranded human DNA parvovirus that can integrate into a specific location on chromosome 19 (19ql3.3-qter called AAVS1). It appears, however, that when used as a vector, integration occurs much less efficiently and more randomly, and many show long term persistence in unintegrated forms (Russell and Kay, 1999). The main advantage of this vector is its ability to give sustained transgene expression upon in vivo administration compared with other viral vectors. For example, expression of transgenes have been detected two years after infection in mice (Snyder et al, 1999), and several months after infection in dogs (Herzog et al, 1999; Monahan et al, 1998), primates (Zhou et al, 1998) and man (Pfeifer and Verma, 2001). It appears to be extremely promising for a cure for hemophilia B; an AAV vector with clotting factor IX led to long term correction at therapeutic levels in mice and canine models of the disease after one dose in vivo administration (Herzog et al, 1999; Snyder et al, 1999). AAV vectors can transfer genes to both quiescent and proliferating cells. It has a broad 16 host range, as its host cell receptor is a membrane associated heparan sulfate proteoglycan which is ubiquitously expressed (Russell and Kay, 1999). The main disadvantage of AAV vectors concern insert size and manufacturing; the virus can only accommodate up to ~5 kb, and the manufacturing process requires the use of a helper virus (usually adenovirus) (Van Tendeloo et al., 2001). The data regarding the utility of AAV vectors for transduction of hematopoietic cells have been controversial in the literature. At the same multiplicity of infection, AAV vectors were found to significantly transduce non-hematopoietic cell lines better than hematopoietic cell lines (Hargrove et al., 1997) (Malik et al., 1997). Some groups have reported transduction of hematopoietic progenitors, (LaFace et al., 1988) (Zhou et al., 1993) whereas other reports did not (Russell and Kay, 1999) (Hirata, unpublished results). The majority of researchers investigating AAV transduction of HSCs observed a relatively low gene targeting efficiency, (Russell and Kay, 1999) (van Os et al., 1999) while others have reported efficient and stable transduction of HSCs (Chatterjee et al., 1999). These inconsistencies make this vector not a reliable tool for transduction of HSCs. 1.3.1.1.2 Retroviral vectors Among viral vectors, retroviruses are the most widely used for both gene transfer and gene therapy and therefore will be discussed in the most detail. Retroviruses contain an RNA genome that is within a protein core, surrounded by several viral enzymes (reverse transcriptase (RT), integrase (LN), proteases (PR)), which direct the reverse transcription of the viral genome into DNA, followed by integration of the DNA into the 17 host cell genome. It is also surrounded by an envelope that is studded with viral glycoproteins that determine the host cell range. The integrated DNA form of the retrovirus (provirus) consists of two long terminal repeats (or LTRs) that contain a promoter for transcription of the viral genes, as well as signals required for RNA processing, and packaging ((¥) encapsidation signal or psi sequence) of RNA into progeny viral particles. The gag gene encodes the core structural proteins, the pol sequence encodes the RT, PR and IN enzymes and the env gene encodes the glycoproteins (reviewed in Bodine, 2001). To use as a vector for gene transfer, the coding sequences (trans factors) for the viral proteins and the cis sequences required for integration are separated into distinct nucleic acid molecules in a cell line referred to as a packaging cell line (Figure 1.4). These cells have been stably transfected with the helper virus sequences; the gag, pol, env genes, but lack the signals required for packaging; RT and integration (LTR and ¥), and thus prevents the production of a replication competent retrovirus. As a further safety mechanism, the gag-pol and env genes are on separate transcriptional units, decreasing the probability for infectious helper viruses forming by recombination. Transfection of the packaging cell line with a plasmid containing the gene of interest and packaging signals (LTR and ¥), results in the generation of a recombinant vector particle (Bodine, 2001; Van Tendeloo et al., 2001). The host range and cellular tropism of the virus is determined by the env gene. Different envelope glycoproteins recognize and bind different cell surface receptors and therefore can be used to generate a recombinant vector with a desired host range, referred to as pseudotyping. Once the recombinant retroviral vector has entered the cell, it is reverse transcribed and integrated 18 at random into the host cell genome. Thereafter, it is incapable of propagating itself, due to the absence of the proteins required for viral assembly (Bodine, 2001). 19 o CN © u > o ha a s S S o o « PC O CJ 3 s CS o e u s -a o La 0. <u wo o eg O- 3 « o on a o u c ~ C 22 « '3 o 81 c « .a w ^3 (/i O 43 M ra .s -~ '5b M 43 44 =1 CJ Vi CO U 43 43 o Vi > 43 H u* O -*-» Oa <U o UJ Oa O > 0 <D <u •c Oa s Oa CJ 3 T3 O The moloney murine leukemia virus (MoMLV) is a C type retrovirus (also referred to as an oncoretrovirus) that is most commonly used for generating retroviral vectors as its life cycle and genome have been well studied. The advantage of retroviral (RV) vectors is the ability to infect murine hematopoietic cells, including HSCs with a high efficiency of up to 70-90% under ideal conditions (Bodine, 2001). Pseudotyping with a variety of different envelope gylcoproteins has further increased the host range and improved the efficiency to human HSCs. For example, the envelope G glycoprotein from Vesicular Stomatitis Virus (VSV-G) enabled the concentration of retroviral vectors to extremely high titers (from 106 to 1010 infectious units/ml), and increased the host cell range (Ory et al, 1996). As well, pseudotyping with the Feline endogenous virus envelope called RD114, was shown to allow infection of human cord blood cells with an efficiency of -90%, compared to -20% with no pseudotyping (Kelly et al., 2000). One disadvantage of using standard oncoretroviruses is their inability to infect non-dividing cells since the nucleur membrane must be disrupted (which occurs during mitosis) for the pre-integration complex to gain assess to the nucleus (Van Tendeloo et al., 2001). Some of the other limitations of using RV vectors for gene transfer will be discussed further below. 1.3.1.1.3 Lentiviral vectors Lentiviruses also belong to the retrovirus family. The best known lentivirus is the human immunodeficiency virus, in which HIV-l is most commonly used as a vector for gene delivery. The original HIV-l vectors had a limited tropism, restricted to transduction of lymphocytes. However, these vectors have now been pseudotyped with VSV-G in the 21 same manner as described for RV vectors, to further expand the tropism and therefore its potential for gene therapy (Vigna and Naldini, 2000). An attractive feature of these vectors in contrast to oncoretroviruses, is their ability to infect non-cycling quiescent cells due to the pre-integration complex being able to traverse the intact nuclear envelope (Vigna and Naldini, 2000). Numerous reports have confirmed that lentiviral (LV) vectors demonstrate superior transduction of growth-arrested mouse and human cell lines compared to MoMLV vectors. For example, Mochizuki et al, found LV vectors transduced 20-30% of contact inhibited human skin fibroblasts and rat cerebellar neurons, whereas MoMLV transduction was nearly undetectable (Mochizuki et al., 1998). Naldini et al, found that on growth arrested HeLa cells, LV vectors transduced 70-90% of the cells and MoMLV vectors only 5-8% (Naldini et al., 1996). Transduction of more primitive HSCs has been more controversial. Several studies have demonstrated better transduction with LV vectors compared to MoMLV vectors in CD34+ cells, both in vitro and in NOD/SCID repopulation assays (Evans et al., 1999; Miyoshi et al., 1999; Uchida et al., 1998). In contrast, Barrette et al, recently demonstrated that both LV and MoMLV vectors perform equally in transducing mouse HSCs. They further showed that LV vectors are NOT able to transduce mouse HSCs that are in G0, as there was NO gene transfer evident (by PCR analysis) when the infection was performed without prior stimulation of the cells with cytokines (Barrette et al., 2000). The authors speculate that lentiviral transduction may require cell cycle activation for transduction (ie, Gl or G2, the state in which most of the growth arrested cell lines were in), but mitosis is not an absolute requirement for integration (Barrette et al., 2000). This is further confirmed by 22 observed decreased lentiviral transduction in non-cycling cells like hepatocytes (Park et al, 2000). Preliminary results suggest that LV vectors may also be more resistant to silencing than MoMLV vectors. It has been speculated that this may be due to the intrinsic features of the LTRs; the oncoretroviral LTR is a more powerful enhancer compared to the LV LTR and therefore may be more prone to methylation and transcriptional shut off (Vigna and Naldini, 2000). For example, when LV vectors are injected in rodent brain, liver, muscle and eye, they give sustained expression for up to 6 months, whereas MoMLV vectors are rapidly shut down (Verma and Somia, 1997). As well, following LV transduction of ES cells, Hamaguchi et al, observed expression in day 6 EBs and day 10 hematopoietic colonies, whereas no expression was observed with the corresponding oncoretroviral vector (Hamaguchi et al., 2000). 1.3.1.1.3.1 Limitations of retroviral vectors There are several problems in using retroviruses as a vector for gene transfer. These include limitations on the size of the transcriptional unit, ability to achieve an adequate titer, generation of a stable, non-rearranged virus, and they have a tendency to be subjected to position effects and/or silencing. In the 1980's, Gelinas and Temin performed a series of experiments to define the maximum insert size for retroviral packaging. They introduced different sized DNA sequences into spleen necrosis virus-derived vectors and studied the ability of the resulting viruses to replicate in chicken embryo fibroblasts. They found that viruses larger than 9kb could not efficiently replicate, and determined that this size restriction 23 occurred during virus encapsidation (Gelinas and Temin, 1986). This limited packaging capacity can lead to problems if more than one gene is required or if a large genomic clone with appropriate regulatory elements for expression is needed. This is exemplified with the human P globin gene in which transgenic mice studies demonstrated that for high levels of expression, -8.8 kb of DNA sequence is required (Pasceri et al., 1998). The ability to achieve an adequate titer (greater than 106 infectious units/ml) is essential for efficient viral gene transfer. Several studies have demonstrated that the number of transcriptional units present in retroviral vectors can have a major influence on viral titer. For example, Correll et al, tested different retroviral vector designs to achieve long-term expression of the human glucocerebrosidase cDNA in murine hematopoietic cells in vivo. They found that retroviral vectors with the selectable marker neomycin phosphotransferase gene consistently produced low titers and therefore resulted in a low transduction efficiency to HSCs. In contrast, the same retroviral vector that contained no selectable marker was able to efficiently infect HSCs (Correll et al., 1994). This finding was also observed by Apperley et al, in which RV vectors with the human adenosine deaminase gene linked in cis to the neomycin marker gene, also failed to efficiently infect mouse bone marrow cells (Apperley et al., 1991). This also can be a major problem with very large sized inserts. As mentioned above, larger cassettes are harder to package and usually result in low viral titers (Leboulch et al., 1994). It also has been hard to achieve stable, non-rearranged virus with some transcriptional units. This has been particularly challenging when genomic DNA is used in RV vectors. The presence of introns, inappropriate cryptic splice sites and poly A addition sequences within the gene, can lead to inadvertent splicing of the viral RNA in 24 the nucleus prior to reverse transcription leading to rearrangements of the virus (Leboulch et al., 1994). To circumvent this problem, intron-containing transgenes can be placed in the opposite orientation relative to the viral LTR direction of transcription. Unfortunately, in some cases, this approach has lead to detrimental effects on gene expression due to anti-sense effects (Correll et al., 1994). Sequences within the MoMLV vector also appear to be incompatible with some regulatory elements within the transcriptional unit. For example, when the locus control region from the human CD2 locus was inserted into MoMLV vectors, it resulted in frequent rearrangements and low viral titers. Interestingly, when this same fragment was put into the context of a LV vector, they obtained high levels and stable expression of the transgene indicating HIV vector sequences may be more compatible than MoMLV sequences (Kowolik et al., 2001). One of the biggest problems associated with retroviral gene transfer is that the expression of a transgene tends to be variable and or silenced within days to weeks in vitro and even shorter in vivo (Van Tendeloo et al, 2001). The next section explores this in more detail, along with some of the approaches currently being used to overcome these problems. 1.3.1.1.3.2 Position effects/silencing Position effects refers to the variability in gene expression when a gene is placed in an ectopic location. Analysis on single clones have divided position effects into two categories: 1) pancellular - when transgene expression occurs in all cells but at a lower level than the endogenous and 2) heterocellular - when only a proportion of the cells 25 express the transgene (gene silencing). Position effects are widely observed in both retroviral and transgenic studies (Whitelaw, 2001). For example, Table 1.2 summarises studies in which the frequency of G418 resistant colonies to PCR positive colonies were compared after retroviral gene transfer of the neomycin-phosphotransferase gene into human clonogenic hematopoietic cells. The observed trend is a higher number of DNA positive colonies compared to G418 resistance colonies, which suggests a failure to express the gene in a large proportion of transduced cells (Neff et al., 1997). Table 1-2 Evidence for position effects following retroviral gene transfer of neomycin-phosphotransferase into human hematopoietic cells Cell type/culture conditions G418 resistance PCR References CFC 13% 35% Fink etal. 1990. LTC-IC-derived CFC transduced in: IL-3/G-CSF IL-3/G-CSF/SCF CFC (grown in 96 well plates) 27% 61% 54% 43% 80% 93% Hughes etal. 1992. Bregni et al. 1992. CFC from cord blood: Plated at 250 cells/ml 24% Plated at 10 cells/ml 39% Plated at 1 cell/ml 49% 83% 65% 90% Luetal. 1993. CFC 10% 90% Lu et al. 1994. CFC, colony forming cell; LTC-IC, long term culture-initiating cell; IL-3, interleukin-3; G-CSF, granulocyte-colony stimulating factor; SCF, stem cell factor. Compiled from Neff etal, 1997. 26 Position effects are thought to be the result of an equilibrium between activating transcriptional control elements (positive acting factors) and repressive chromatin (negatively acting chromatin binding proteins) at the site of integration. Sequences within the viral LTR and viral backbone have contributed to their silencing in ES and hematopoietic stem cells, thereby shifting this equilibrium towards an "inactive state" (Whitelaw, 2001). Recent modifications in these sequences have resulted in vectors that are more resistant to silencing by increasing the affinity for activation factors and decreasing the affinity for negative repressors. This was based on the observation that several viral mutants were able to express transferred genes at high levels in embryonic carcinoma (EC) cells and ES cells compared to the wild type counterparts. The PCMV (a myeloproliferative sarcoma virus mutant expressed in PCC4 EC cells) virus (Hilberg et al, 1987) and dl587 rev virus (Colicelli and Rigby, 1987) possess various deletions and base pair mutations which were found to remove several transcriptional blocks located within the LTR and primer binding sites of the original viruses (Grez et al., 1990; Grez et al, 1991; Prince and Rigby, 1991). For example PCMV LTR destroys a binding site for a repressor protein and creates a binding site for the Spl transcription factor. The dl587 rev virus removes a silencer element that overlaps the primer binding site. Hawley et al, have produced a series of vectors based upon a murine stem cell virus (MSCV) backbone, which combines the LTR from the PCMV and 5' untranslated region from dl587 virus (Hawley et al, 1994). Pawliuk et al, demonstrated that this MSCV vector conferred higher levels of expression and was more resistant to silencing in murine hematopoietic cells (Pawliuk et al, 1997). Expression from the MSCV LTR promoter has also been 27 ( shown to be resistant to silencing in human HSCs (Cheng et al, 1998). Another strategy to improve expression is to dele.te the LTR sequences that are repressive to their expression. This is performed by deleting a portion of the 3' LTR, such that following infection of the target cells, the viral enhancer and promoter regulatory regions will become non-functional. These are called self-inactivating vectors or SIN vectors and have resulted in an increased expression of ~ 5- 10 fold higher than with a functional LTR (Bodine, 2001). Many groups have correlated de novo cytosine methylation of CpG dinucleotides in the retroviral LTRs with the silenced state, and further show that expression can be reactivated using a methylation inhibitor 5-AzaCytidine (5-azaC)(Challita and Kohn, 1994; Jahner et al, 1982; Stewart et al, 1982). This methylation is thought to occur by de novo methyltransferase encoded by the dnmt 3 gene (Okano et al, 1998). Methylation can have a direct effect on silencing by preventing transcription factor binding or indirect effect in which methylation induces binding of the MeCP2 protein, which in turn recruits a repressor complex that condenses the chromatin and prevents transcription factor access (Jones et al, 1998). It should be noted that methylation cannot explain all silencing effects as expression from some viral vectors are NOT reactivated with 5-azaC (Pannell et al, 2000). Moreover, silencing of retroviral vectors in ES cells occurs within 3 days but methylation is not detected until day 10, indicating that retroviral silencing is independent of methylation in these cells (Pannell et al, 2000). Other approaches to improve transgene expression from RV vectors involve the inclusion of cis-elements that may overcome position effects. Locus control regions or LCRs are cis-regulatory elements that confer high level, tissue specific expression of 28 homologous and heterologous genes in a position independent, copy number dependent manner. Kowolik et al, recently described the use of a T cell specific LCR from the human CD2 gene in a LV vector. They demonstrated that it conferred higher levels of expression in primary human T cells. This expression was correlated with an increase in the proportion of cells that express the transgene and in the levels of expression/cell. As well, analysis of single clones demonstrated less variation in expression (Kowolik et al., 2001). These results are consistent with another study by May et al, in which the human p globin LCR in a LV vector also provided higher, long-term expression in transplanted mice (May et al, 2000). Thus, LCRs appear to overcome position effects by establishing/maintaining an active locus at the site of transgene integration. A possible mechanism for this was demonstrated by Festenstein et al, in which the LCR appears to have a dominant effect over the heterochromatin protein -M31, responsible for recruiting silencing factors (Festenstein and Kioussis, 2000). Insulators are chromatin components that serve to divide the genome into structurally and functionally discrete, topologically independent domains. They function by preventing a gene in one domain from interacting with distal enhancers and inappropriate promoters that reside in neighbouring domains, but have no inhibitory/stimulatory transcriptional effects (Neff et al., 1997). The first and best characterised vertebrate insulator is located within the chicken P globin LCR. It contains a DNase hypersensitive site called cHS4, which appears to constitute the 5' boundary of the locus and has been shown to reduce position effects in transgenic Drosophilia (Rivella et al., 2000). Previous studies demonstrated that this same insulator could confer resistance to position effects in mammalian cells (Chung et al., 1997; Pikaart et al., 29 1998). However, all of these reports used selectable markers, which may bias the analysis to integration sites that are favourable for expression. Two recent reports tested this insulator in a RV vector in vitro (in a hematopoietic cell line) and in vivo (transgenic mice), with no drug selection such that all sites are amenable to analysis. They found that the insulator increased the probability of expression of the provirus from 21% (no insulator) to 74% in a cell line and from 2% to 19% in vivo, and expression was maintained for an extended period of time (in vivo for 8 months). This was accompanied by a decrease in the level of pro viral DNA methylation, speculating that insulators may improve expression by reducing/preventing methylation of the viral LTRs. Although this insulator appears to have reduced the incidence of position effects, it did not totally eliminate it. For example, analysis of single clones in the cell line demonstrated extensive variability in both mRNA and protein levels (Emery et al, 2000; Rivella et al., 2000). Other promising cis-regulatory elements are scaffold or matrix attachment regions (SARs or MARs). These are DNA sequences that bind to the nuclear scaffold and are enriched in DNA topoisomerase LT binding sites and although there is no consensus sequence, they tend to be AT rich (Boulikas, 1993; Cockerill and Garrard, 1986). One hypothesis to their function is that they define boundaries of independent chromatin domains, similar to insulators. However, unlike insulators, they tend to have transcriptional activity that may act by recruiting transcription factors to enhancer-promoter sequences within the domains (Dang et al., 2000). Dang et al, used the SAR from the human p interferon gene in MoMLV RV vectors in an immortalised T cell line. They found that the SAR element maintained and improved long-term RV vector 30 expression. For example, in vitro experiments demonstrated -75% of the cells with the SAR element expressed the transgene and this dropped to -47%, 18 weeks later. In contrast, cells without the SAR element went from 64% to 16% during the same culture period. Moreover, analysis of the mean level of fluorescence showed no change with the SAR element and a dramatic decrease without it. That is, this element ensured that cells that contained the provirus expressed the transgene and clonal analysis showed that it reduced integration site effects (done under no selection pressure) (Dang et al, 2000). It still remains to be determined if such an element will work similarly in primitive hematopoietic cells, such as HSCs. One caveat in the above studies is that the cis-regulatory elements included in the viral vectors are analyzed at ectopic locations in the genome. This makes it difficult to distinguish if an "element" improved expression and/or suppressed position effects/silencing, OR if the element by chance integrated into a site that was permissive for expression. A strategy in which a series of constructs with different cis-acting elements are studied at the SAME integration site would therefore be extremely useful for analyzing transcriptional effects. Moreover, the influence of position effects can be further and more accurately explored by analysing such constructs at more than one integration site. This approach was performed using the Cre/lox recombination system to assess the hematopoietic regulatory mechanisms using the ES in vitro differentiation and P globin gene expression as a model system. 31 1.3.1.2 Existing non-viral approaches Non-viral gene transfer is an attractive alternative to viral vectors as it is less toxic, simpler to use, and provides flexibility with respect to the size of DNA to be transferred such that large, complex constructs can be introduced. Progress towards applying such an approach to hematopoietic cells has been hindered, because of their low gene transfer efficiency and transient expression due to lack of integration into the host genome. However, within the last year, there have been several exciting reports that have demonstrated high efficiency and stable integration (although still below that of viral vectors), in hematopoietic cells using an approach called electroporation. 1.3.1.2.1 Electroporation In this technique, cells are exposed to a high voltage electrical discharge, which results in the formation of temporary pores in the cell membrane (reviewed in Heiser, 2000). This allows the uptake of DNA into the cytoplasm and subsequently into the nucleus by a yet unknown mechanism. Electroporation can be used for transient and stable integration of DNA in a wide variety of cell lines, including hematopoietic mouse and human cell lines (Heiser, 2000). It has been demonstrated to be reproducible, can lead to integration of a construct at a low copy number, and is very useful for transfecting more than one construct into a cell at the same time (co-electroporation) (Van Tendeloo et al., 2001). Keating et al, were the first to show gene transfer using electroporation to primitive hematopoietic cells. In this study, human bone marrow cells were electroporated with a selectable marker, followed by plating in methylcellulose with selection. They obtained a stable integration frequency of 0.8% - 2.7% per 1 x 107 32 electroporated cells in hematopoietic progenitors, and verified that electroporation has no effect on their development (Toneguzzo and Keating, 1986). Another group further expanded these studies and demonstrated that electroporation also has no effect on cells responsible for murine long-term reconstitution as determined by in vivo animal repopulation studies (Bergan et al., 1996). Electroporation of more primitive cells was recently evaluated in several independent reports. Van Tendeloo et al, demonstrated an electroporation efficiency of 5.2% ± 0.4% with minimum cytotoxicity (<5%) in unstimulated human HSCs (CD34+) from bone marrow without any prior cell purification (Van Tendeloo et al., 2000). Two other reports significantly improved this efficiency by first culturing cells prior to electroporation in order to stimulate their proliferation. In these studies, human HSCs isolated from peripheral blood were cultured for 48 hours in serum free medium supplemented with Flt-3 ligand, stem cell factor and thrombopoietin. This was found to increase the number of cells in cycle from 2 - 28%. Using the green fluorescent protein (GFP) as a marker, they showed transient GFP expression by FACS 48 hours later in 21% ± 1% of the CD34+ cells. In vitro assays demonstrated -56% GFP positive hematopoietic progenitors per 1 x 105 electroporated CD34+ cells (or 1.8% absolute integration frequency), and -16% GFP positive LTC-IC cells per 1 x 105 electroporated CD34+ cells (or 0.04% absolute integration frequency) that are capable of generating hematopoietic cells for up to 5 weeks in vitro (Wu et al., 2001a; Wu et al., 2001b). Another approach was explored by Li et al. They found that electroporation induces apoptosis and therefore causes high cell mortality and this results in a low gene transfer efficiency. Here, they electroporated human CD34+ cells and introduced molecules that reduce apoptosis (caspase inhibitors). They achieved a transfection 33 efficiency of-20%, regardless of the presence of cytokines in the suspension medium (Li etal, 2001). . 1.3.1.3 Potential approaches There are several new "potential" approaches currently being explored for the purposes of overcoming some of the drawbacks associated with both viral and non-viral approaches. 1.3.1.3.1 Extra-chromosomal replicating vectors Extra-chromosomal replicating vectors refers to vectors that can be retained extra-chromosomally (rather than integrated into the genome), and have the ability to replicate with the genome upon division. 1.3.1.3.1.1 Mammalian artificial chromosomes A mammalian artificial chromosome is a DNA fragment-that possesses the properties of a chromosome such that it can replicate and segregate with the genome, and can accommodate up to 6-10 megabases of DNA. To maintain it in a linear conformation, the ends are capped with telomeres, possess an ori site for replication, and a centromere to ensure proper replication and division (Willard, 2000). The first complete construction of a human artificial chromosome (AC) was described only a few years ago (Harrington et al., 1997). The human hypoxanthine phosphoribosyl-transferase (HPRT) gene, which is mutated in the inherited neurological disorder Lesch-Nyhan disease, was recently cloned into an AC as a 90 kb genomic fragment and successfully introduced into cultured cells (Grimes, unpublished results; Mays, unpublished results). These studies 34 provide proof of principle that it may be possible to design and assemble a variety of different ACs carrying and expressing genes of biological or therapeutic interest. Although ACs do not integrate into the genome, there is still the possiblity of recombination with cellular chromosomal sequences and potential oncogenesis due to the introduced genetic instability (Van Tendeloo et al., 2000). 1.3.1.3.1.2 Epstein bar virus based vectors Epstein Bar Virus (EBV) based vectors require the EBV ori site and the Epstein Bar nuclear antigen (EBNA-1) for retention and self-replication in the nucleus (Teshigawara and Katsura, 1992). These vectors can accommodate up to 160 kb of DNA and are B cell specific, thereby making them attractive targets for some hematological disorders (Banerjee et al., 1995). However, the EBNA-1 protein has some oncogenic potential, a problem that should be addressed in more detail. As well, EBV vectors have a tendency to randomly segregate, and gradually disappears over a few months in rapidly dividing cells (Van Tendeloo et al., 2001). 1.3.1.3.2 Hybrid vectors Hybrid vectors are those transfer systems in which the positive aspects of two different types of viruses are combined to create a "hybrid" with superior properties. Zheng et al, combined the positive aspects of retroviral gene transfer (ability to integrate) with the positive aspects of adenoviral gene transfer (infect dividing and non-dividing cells with a high efficiency). Here they placed the MoMLV LTR sequence around a reporter gene and cloned this into a gutless adenoviral vector (removed for all adenoviral genes). They compared it with an adenoviral vector carrying the same reporter 35 gene, but no LTR sequences. They found that both could infect non-dividing and dividing cells in vitro and in vivo (spleen). Expression lasted up to three months in vivo with the adenoviral vector containing LTR sequences, whereas no expression was observed with the vector lacking LTR sequences. They used fluorescent in situ hybridization and probed with the reporter gene and found only the adenoviral vector with LTR sequences was integrated (-15% in the cell lines and 5% in spleen) (Zheng et al, 2000b). Another report used the positive aspects of adenoviral vectors combined with AAV vectors (ability for site specific integration). Site specific integration with AAV vectors requires the Rep protein and inverted repeats from the AAV genome. Therefore, in this study, they used two adenoviral vectors; one with the AAV Rep protein and an adenoviral vector with AAV inverted repeats that flanked the transgene. These were co-infected into hepatoma cells and found that site specific integration occurred specifically at AAVS1 in - 35% of the cells (Recchia et al, 1999). Although considerable development is needed to make these hybrid vectors suitable for therapy, these studies nonetheless illustrate their great potential. 1.3.2 Gene targeting Homologous recombination between DNA sequences that reside in the chromosome and newly introduced DNA sequences, an event termed gene targeting (GT), provides a means for introducing specific mutations in the genome. The efficiency of GT using electroporation, typically ranges from 1/105 to 1/107 treated cells and therefore requires a large number of cells in order be detected (Capecchi, 2001). This approach is 36 best known for its widespread success in mouse ES cells, as procedures have been developed for culturing and amplification of ES cells to the densities needed. Gene targeting was initially demonstrated by Smithies et al, in their classical experiment of targeted modification of the p globin locus in the mouse erythroleukemia cell line (Smithies et al, 1985). This elegant experiment illustrated that it was feasible to disrupt an endogenous gene in cultured mammalian cells. After this was established, Capecchi and colleagues extended this approach to ES cells and the generation of a so called "knock out" mouse model. In this pioneering work, a selectable marker (neomycin phosphotransferase or neo) was specifically integrated into the HPRT locus. This not only allowed expression of the neo gene (and hence its selection), but simultaneously inactivated the target gene (HPRT) in ES cells thereby generating a knock out. More importantly, these cells were then injected back into blastocysts and brought to term in a foster mother to generate chimeric mice that were capable of transmitting the mutation to their offspring (Thomas and Capecchi, 1987). This constituted a breakthrough in evaluating targeted deletions of specific genes in mice and ultimately for targeted alterations/mutations for functional studies. Homologous recombination has also been demonstrated in somatic cells including hematopoietic cell lines. Two such studies suggested that F9 embryonic carcinoma cells (Coll et al, 1995) and two pre-B cell lines (Charron et al, 1990), supported GT as efficiently as ES cells. Fibroblastoid, lymphoid, hepatic, epitheloid, myeloid, bladder and colon cell lines have also been shown to undergo targeted modifications at a ratio of homologous recombination to random integration of 1:40-l :6550 and an absolute targeting frequency of 1.2 xlO"8- 2.7 xlO"5 (Doi et al, 1992; Yanez and Porter, 1998). In 37 primary human cells, retinal epithelial cells, keratinocytes and embryonic lung fibroblasts 8 6 have been modified by GT with absolute targeting frequencies between 6x10" - 4.3 x 10" (Brown et al, 1997; Williams et al, 1994; Yanez and Porter, 1998). Can homologous recombination occur in more primitive hematopoietic cells? To date, there have been no reports of GT in HSCs owing to their low abundance in the bone marrow. However, a recent report by Smithies and colleagues demonstrated that GT can occur in clonogenic hematopoietic progenitors, which are in the same lineage as HSCs but are more abundant and mature. They used GT to replace a defective HPRT gene, and found it occurred at low but detectable frequencies in progenitors derived from both murine bone marrow cells and ES cells (4.4 ± 3.3 x 10"5; 2.3+ .4 x 10"5, respectively) (Hatada et al, 2000). Although GT is not yet feasible in HSCs, these results combined with the recent developments in GT technology (explored in the next section), suggests that site directed genetic modification of HSCs may soon be possible. 1.3.2.1 Approaches to improve gene targeting techniques One of the main barriers to GT is that random integration occurs ~ 1000 fold more frequently than homologous recombination. For this reason, strategies have been implemented to select for clones that are correctly targeted (Vasquez et al, 2001). One commonly used approach is referred to as positive/negative selection and has resulted in a 2000 fold enrichment of the "correct" clones. This involves using a targeting vector with the neomycin gene in the homologous region, and the Herpes Simplex Virus-thymidine kinase (HSV-tk) gene at the end of the same construct. If integration occurs by homologous recombination, cells will become neomycin resistant and lose the HSV-tk 38 gene. In contrast, random integration, which occurs through the ends of the DNA construct by non-homologous end joining, will result in cells becoming neomycin resistant and HSV-tk+. Selection for neomycin resistance by G418 (positive selection) followed by selection with Gancyclivor, which kills all cells with the HSV-tk gene (negative selection), allows for enrichment of those cells in which homologous recombination of the vector occurred (Mansour et al., 1988). Another approach has been to use viral vectors to introduce the targeting construct to cells because of their high gene transfer efficiency over non-viral transfection methods. Interestingly, in several of these studies, adenoviral vectors were found to decrease the frequency of random integration. For example, a replication defective adenoviral vector was used to target the fibroblast growth factor locus in mouse ES cells and yielded a ratio of targeted integration to random integration (GT:RI) of 1:2.5, whereas the same targeting construct introduced by electroporation gave a ratio of 1:20 (Mitani et al., 1995). This was similarly observed at the APRT locus in a Chinese Hamster Ovary cell line in which the ratio obtained was 1:5, a 400 fold improvement over that obtained with calcium phosphate (Wang and Taylor, 1993). Even though the absolute targeting frequency was low in both experiments (10"5 to 10"7), it appears that viruses have the ability to decrease non-homologous recombination or random integration and therefore the majority of the clones obtained will be "correct". An even more promising report by Russell et al, demonstrated a high frequency of homologous recombination using adeno-associated virus (AAV). Here, they used an AAV vector with a mutation in the HPRT gene, and introduced it into human fibrosarcoma cells and measured the frequency of disruption of the endogenous HPRT gene by selection using 6-Thioguanine resistance. They 39 demonstrated 0.02 - 0.05% homologous recombination frequency, ~ 100-1000 fold better than what occurred with the adenoviral vectors. They speculate that this dramatic increase in efficiency may be due to the single stranded nature of the AAV genome and the high multiplicity of infection obtained with this virus (Russell and Hirata, 1998). Several groups are also currently looking at ways to improve the homologous recombination frequency. Their approach has been to decipher which proteins are responsible for homologous recombination. Several of these proteins were originally identified in yeast, but their apparent human counterparts have now been identified. Overexpressing these proteins in several different murine and human cell lines has led to a 2-40 fold increase in GT. Despite these promising results, for safety purposes these must be expressed for only a short time because it could be extremely detrimental to the stability of the genome (reviewed in Vasquez et al., 2001). Templeton et al, used a GT protocol designed to increase the survival of clones (and therefore decrease the mortality) after electroporation of the targeting construct into ES cells. They demonstrated that absolute targeting efficiencies could be improved from 0.001 to 20% by this optimised GT protocol. This protocol involved plating the cells after electoporation at very high densities and then delaying the selection for at least 60 hours. This dramatic improvement would be ideal if it could be reproduced in other cell lines, like hematopoietic cells (Templeton et al., 1997). There are two other widely used techniques for performing site directed genomic modifications. One technique uses RNA/DNA double stranded oligonucleotides (RDO), referred to as chimeraplasty. Here, deoxnucleotides that contain the appropriate alteration are flanked on either side by ribonucleotides and a poly T hairpin loop. It is postulated 40 that the RNA residues and hairpin structures induce stability and enable efficient base pair exchange; however the exact mechanism of RDO-mediated exchange is still unknown (Rice et al., 2001). There have been several reports of extremely high gene correction frequencies with this technique. For example, correction of the single base pair mutation in the P globin gene that is responsible for sickle cell disease, has been shown in a lymphoblastoid cell line to occur with a frequency of-50% (Cole-Strauss et al, 1996), and in primitive human hematopoietic cells (CD34+) at a frequency of 11% (Xiang et al., 1997). There has also been a report of a high gene correction frequency of 40% in the liver of rats (in vivo), in which the hemophilia B locus was targeted (Kren et al, 1998). It should be noted that there is a lot of controversy with this technology. Investigators have tried to repeat the above results and repeatedly tried to perform similar experiments in other cells/tissues and have been unsuccessful (van der Steege et al., 2001). The reasons for the persistent failure is unknown, but indicates that broad applications of this technique may not be feasible. The other technique makes use of triplex forming oligonucleotides (TFOs), which have been shown to alter specific DNA targets on episomes and within the chromosome. TFOs can efficiently find their target within minutes and bind to polypurine tracts within the DNA. This binding forms a triple helix structure, which in turn induces DNA repair. Linking this TFO with a "repair domain" (a double stranded DNA molecule with homology to a region of interest except for one base pair change), will mediate GT or correction at the endogenous homologous site in the genome. This process seems to involve the nucleotide excision pathway, as repair is diminished in cell lines which express proteins defective in this pathway, and can be restored upon re-introduction of 41 them to cells (Rice et al., 2001). Culver et al, used this approach to correct a single base pair mutation in the adenosine deaminase (ADA) gene in a human lymphocyte cell line in vitro. They obtained a gene correction frequency of 1- 2%, with no mutations induced around the target site. Moreover, neither the repair domain or TFO alone could achieve correction (Culver et al., 1999). Vasquez et al, recently demonstrated the use of TFOs for site specific modifications in somatic cells of adult mice in vivo. They first created transgenic mice with a mutated supF reporter gene. They subsequently injected these mice with a supF-targeted TFO and ten days later analysed the tissues for site specific modifications. They found a correction frequency of ~ .05% in most tissues (Vasquez et al., 2000). Unexpectedly, however, they did find that TFOs induced mutations around the target site which were not seen in vitro, and therefore may limit its therapeutic value. In summary, even though GT frequencies are low, substantial advances have been made (and continue to be made) that may improve this technique for all cells, including HSCs. 1.4 Site specific recombinases A number of bacterial and yeast elements encode recombinases, which are enzymes that cleave and ligate DNA at specific target sequences resulting in a precisely defined recombination reaction. These "site specific recombinases" provide a novel tool for performing genomic manipulations in mammalian cells. Two of the most widely used recombinases are Cre from bacteriophage PI (Sauer and Henderson, 1989) and FLP from Saccharomyces cerevisiae (O'Gorman et al., 1991), which catalzye recombination between two 34 base pair (bp) recognition sites called lox P and FRT (FLP recognition 42 target), respectively. Initial attempts to develop the FLP/FRT system in transgenic mice resulted in either no recombination (Ludwig et al, 1996), or mosaic recombination (Dymecki, 1996)j because the wild type FLP was found to have decreased enzyme stability at 37°C (Buchholz et al, 1996). However, a thermostable variant called FLPe (Buchholz et al, 1998), was recently generated that mediates recombination in vivo at efficiencies resembling those of Cre (Rodriguez et al, 2000). Owing to the initial problems with the FLP/FRT system, the Cre/lox system has been more frequently used and will be discussed in more detail below. 1.4.1 Cre/lox recombination system Bacteriophage PI encodes a 38 kDa protein called Cre recombinase (causes recombination) and one 34 bp sequence called lox P (locus of crossing over). One lox P site contains two 13 bp inverted repeats (where cre binds) that flank an 8 bp core sequence (where recombination occurs) (Figure 1.5). The role of the Cre/lox system in the bacteriophage PI life cycle is to maintain the phage genome as a monomelic plasmid within the bacterium, Escherichia coli, in the lysogenic state. After replication, there is a tendency for the single copy bacteriophage PI plasmids to ligate and form dimers. The dimers have two lox P sites in the same direction. Cre will bind to EACH lox P site and mediate intramolecular excision to resolve the dimers into monomers, such that during cell division each cell receives one phage genome, thereby preventing any plasmid loss (Figure 1.5A). Intermolecular recombination (or integration) in which two monomers can combine to give a dimer is therefore less favoured (Kilby et al, 1993). The discovery that Cre could function in eukaryotic genomes - yeast, Droshiphila, plants, shows that host 43 specific factors were not required for its activity and opened the way to its use in mammalian systems. Dimer Cre-mediated recombination Monomers Cre binding site • Cre binding site Wild type lox P site (L2) ATAACTTCGTATA ATGTATCJC TATACGAAGTTAT lox 511 (L1) ATAACTTCGTATA ATGTATAC TATACGAAGTTAT Figure 1-5 Cre/lox recombination system from bacteriophage PI. A.Bacteriophage PI replicates as a single-copy circular plasmid in E.coli. During replication, dimers of the plasmid can form in which Cre recombinase will resolve into monomers to ensure equal partitioning into subsequent daughter cells, thereby preventing plasmid loss. Recombination can also occur between two monomers leading to the formation of a dimer, but this intermolecular recombination is less favored than intramolecular recombination. B. Diagram of one lox P target site. There are two 13 bp inverted repeats where Cre recombinase binds (underlined), and an 8 bp spacer region within which recombination occurs (designated by arrows). The wild type lox P site is also referred to as L2. A mutant lox P site (Ll or lox 511) was used in combination with L2 to increase the integration efficiency. Ll has one base pair mutation in the spacer region designated in bold. Compiled from Kilby et al, 1993. 44 1.4.2 Genetic engineering of the genome 1.4.2.1 Deletions: inducible and tissue specific The first study of using the Cre/lox system in mammalian cells was performed by Sauer et al. They first generated a mouse cell line that stably expressed Cre recombinase under an inducible promoter. This was followed by the introduction of a plasmid with a reporter gene flanked by two lox P sites (or floxed). They demonstrated that the reporter gene was removed (deleted), and that the process is dependent on expression of Cre recombinase (Sauer and Henderson, 1988). These results quickly led to using this system for generating specific alterations in vivo. For example, two initial studies cleverly demonstrated that this strategy could be used to turn off or on the expression of a gene in a tissue specific manner in mice. In the first report, they generated two transgenic mice lines: one transgenic mouse was - generated with the Cre gene under the control of a murine ccA-crystallin promoter (restricts expression to the lens of the eye). The other contained the SV40 tumor causing oncogene, large T antigen or TAG, in which a stop sequence was placed between it and the aA-crystallin promoter to prevent its expression. These two transgenic mice lines were mated, and ALL the double transgenic lines developed tumours in the lens of the eye. Tumour formation was confirmed by sequencing to be the result of TAG activation by Cre (Lakso et al, 1992). The second study was performed by Marth and colleagues. Here, they generated transgenic mice with a promoter that drives expression of a floxed P galactosidase gene, which in turn is followed by the gene for human growth hormone. In these mice, the p 45 galactosidase gene will only be expressed from the promoter. They mated these mice to transgenic mice in which Cre was under the control of a thymocyte specific promoter. This generated mice in which human growth hormone was expressed specifically in thymocytes, and P galactosidase expression in ALL OTHER tissues. The efficiency of excision was found to be extremely high ~ > 90-100% (Orban et al, 1992). In the classical knock out experiments (described in section 1.3.2), homologous recombination is used to introduce a selectable marker into a specific locus, which in turn inactivates the endogenous gene in ALL tissues and lineages of the resulting organism. The problem with this approach is that some mutations can be lethal to the embryo, and therefore it is impossible to decipher if a gene exerts its function in later stages of ontogeny and/or in different cell types. This problem can be overcome by creating a "conditional" knock out, in which an endogenous gene is deleted in a specific lineage or cell type using Cre under a tissue specific promoter (Gu et al, 1994). The Cre/lox system has now been extended to genetic modifications both in vitro and in vivo that include: the activation/inactivation of a gene (reviewed in Lewandoski, 2001), generation of inversions of genes (Feng et al, 2001; Molete et al, 2001) and large gene clusters (Alami et al, 2000), and the generation of translocations (Medberry et al, 1995; Smith et al, 1995). To further increase this potential, all of the above can also be performed in a cell/tissue specific manner using Cre specific promoters (reviewed in Lewandoski, 2001), and at a selected time in an inducible manner by fusion of Cre to ligand binding domains of steroid receptors (Danielian et al, 1998; Kellendonk et al, 1996). Cre mediated modifications have been shown to occur in all cell/tissue types, even in highly differentiated postmitotic cells like liver, neurons and resting T cells (Gorman 46 and Bullock, 2000). As mentioned earlier, most studies have reported that deletion of several kilobases of DNA is very efficient, >90-100%. Zheng et al, performed a systematic study using chromosome 11 in ES cells and found that even at distances of 2 centimorgans (~4 megabases), an -11% deletion efficiency could be obtained (Zheng et al, 2000a). 1.4.2.2 Integration of exogenous DNA Sauer and Henderson (1990) were the first to demonstrate that integration (intermolecular recombination) using the Cre/lox system could occur in yeast and in cultured mammalian cells (Sauer and Henderson, 1990). In this pioneering work, a linearized plasmid with an SV40 promoter followed by a lox P site was introduced into a mouse cell line. This was followed by co-electroporation of a plasmid with a promoterless thymidine kinase gene and a lox P site into these cells. Cre mediated integration will introduce the thymidine kinase gene under the control of the SV40 promoter, resulting in cells becoming thymidine kinase positive. Integration was shown to occur specifically at the lox P site in the genome (and no random integration occurred elsewhere in the genome) but the frequency of integration was low(10"3 to 10"4) (Sauer and Henderson, 1990). They later demonstrated, along with another group, that the efficiency of recombination is proportional to the expression of Cre recombinase (Araki et al, 1997; Baubonis and Sauer, 1993). However, only modest improvements in efficiency were achieved by using strong promoters for Cre and modifications that increase its translational proficiency (reviewed in Gorman and Bullock, 2000). 47 1.4.2.2.1 Approaches to improve integration efficiency A major hurdle in using the Cre/lox system for integration is the low frequency of the event. The reason for this is that after integration, the integrated DNA molecule has two lox P sites in the same orientation, which if any Cre remains in the cell will lead to deletion as it is favoured over integration (Figure 1.6 left). Approaches to increase the integration efficiency have been described by modifying the lox P target sites. One of these includes a single base pair mutation in the spacer region (lox 511), which is used in combination with the wild type lox P site (Figure 1.5B). Using these "heterospecific" lox P sites have previously been demonstrated to confer a 20-100 fold increase in the integration efficiency compared to the wild type lox P sites (Bethke and Sauer, 1997) (Figure 1.6 right). 48 lox P (L2) • + Cre recombinase i lox P (L2) lox P (Ll) + Cre recombinase t (unstable product) low integration efficiency (stable product) high integration efficiency Figure 1-6 Cre/lox mediated integration. Integration using the wild type lox P sites is shown on the left. After integration, gene X is flanked by two lox P sites in the same orientation. If any Cre recombinase remains in the cell, gene X will be deleted favoring the reverse reaction. This results in an overall low integration efficiency. Integration using the heterospecific lox P sites is shown on the right. Since Cre recombinase only recognizes similar lox P sites, this will result in a more stable product after integration thereby increasing the integration frequency. 49 Figure 1.7 is a schematic detailed representation of how this is achieved. A transgene (X) flanked by heterospecific lox P sites is stably introduced into the genome. The transgene to be integrated (Y) is then introduced along with a Cre recombinase expression plasmid. Cre recombinase will mediate recombination only between lox P sites with similar core sequences (ie. L2-L2, or Ll-Ll). Figure 1.7 shows recombination through L2. This results in Cre mediated integration of the entire plasmid into the genome, which is then quickly followed by deletion as the substrate is now flanked by lox sites in the same orientation. Deletion via Ll will therefore result in a "cassette exchange", replacing X transgene with Y. lox P (L2) loxP(Ll) + Cre recombinase expression plasmid Integration via L2 L2. Ll L2 gene Y -|ZZ^ Ll geneX -T~~~j)-L2 deletion via Ll Ll gene Y Integration Resolution Figure 1-7 Schematic detailed diagram of Cre/lox mediated integration using heterospecific lox P sites. The lox P sites are first integrated into the genome. A plasmid along with Cre recombinase expression plasmid is then introduced into these cells. This is followed by integration of the entire plasmid into the genome (shown is through L2), which is then quickly followed by deletion of lox sites that are in the same orientation (shown is through Ll). The overall result is a cassette exchange where gene X is replaced with gene Y. 50 Other modifications to the lox P sites have been described. These include different mutations in the Cre binding sites (Araki et al., 1997) and using inverted lox P sites (Feng et al., 1999), and will be described in more detail in Chapter 3. Cre/lox mediated integration has been exploited for carrying out a variety of genetic modifications in mammalian cell lines (Call et al., 2000) (Feng et al., 1999; Kolb et al., 1999). Since these reports were promising, this approach was extended to hematopoietic cells in this thesis work, by combining Cre/lox mediated integration with the high efficiency of retroviral gene transfer. Thus, in hematopoietic cells this strategy may allow: the introduction of large DNA constructs (that may confer more reproducible expression) since the integration step is non-viral, more predictable expression by integration of a transgene into a site with desired expression properties (gene function studies), and a more accurate analysis of cis-regulatory sequences by studying constructs at the same genomic integration site (gene regulation studies). The feasibility and characterisation of this strategy in a number of cell lines, including hematopoietic and subsequently in more primitive cells using the ES in vitro differentiation model was explored in Chapter 3. 1.4.2.3 Cre/lox system for gene regulation studies As mentioned in section 1.3.1.1.3.2, expression of a gene is strongly influenced by its site of integration into the genome. The Cre/lox system provides a novel approach for studying regulatory elements involved in gene expression since it allows a series of systematic changes to be made at a specific locus. That is, different regulatory elements of a gene can be analysed within identical chromosomal contexts, thereby eliminating the 51 influence from neighbouring chromosomal elements. As well, the effects of the site of integration on gene expression can also be studied by analysing these constructs at a variety of different sites. Chapter 4 of this thesis will describe the feasibility of this strategy in studying gene regulation in hematopoietic cells. The cis-regulatory elements of the p globin gene were evaluated in this model as a method for designing better vectors for gene therapy, since achieving high and sustained levels of expression with current viral vectors have been problematic. 1.5 Gene therapy for inherited disorders of the hematopoietic system The ability of HSCs to completely repopulate the entire hematopoietic system following transplantation makes them attractive targets for gene therapy of inherited hematological disorders. Integration of new genetic material into the genome of HSCs would ensure a continous supply of modified hematopoietic cells in the transplanted recipient. Potentially, this allows for permanent correction of the defect (Bodine, 2001). Table 1.3 summarises a number of genetic disorders of the hematopoietic system that are amenable to gene therapy. This includes the hemoglobinopathies, severe combined immunodeficiency (SCID), hemophilia A and B, gaucher's disease and chronic granulomatous disease. Gene therapy could involve either introduction of a functional/therapeutic gene to correct the defect within HSCs or these cells could be used to deliver the therapeutic gene to appropriate tissues (reviewed in Verma and Somia, 1997). In the latter case, for example, hemophilia is caused by a deficiency in blood clotting factors, which are normally produced by hepatocytes. One form of correction of 52 hemophilia is through secretion of therapeutic levels of clotting factors into the blood delivered by hematopoietic cells (Dai et al., 1992). 53 Table 1-3 Candidate diseases for gene therapy of the hematopoietic system Disease Defect Incidence Target Cells Severe Combined Immunodeficiency ADA Adenosine deaminase Rare bone-marrow cells, T-cells SCLD-X1 yc cytokine receptor subunit deficiency of LL-2,4,7,9,15 Rare bone-marrow cells, T- cells Hemophilia A Factor VIII deficiency 1:10,000 males liver, muscle, fibroblasts or Hemophilia B Factor DC deficiency 1:30,000 males bone-marrow cells Hemoglobinopathies: Thalessemias/Sickle Structural defects or 1:600 in certain bone-marrow Cell Disease deficiency in a or p ethnic groups cells globin genes Gaucher's Disease Defect in the enzyme 1:400 in bone-marrow glucocerebrosidase Ashkenazi Jews cells, or macrophages Chronic Defect in NAPDH Rare bone-marrow Granulomatous oxidase (phox): cells Disease heterogenous; p91, p47 p22, p67 Complied from information in Verma and Somia 1997, Malech et al, 1997 and Cavazzana-Calvo et al, 2000. i 54 In the 1990's, the first clinical trials for gene therapy were implemented (reviewed in Bodine, 2001). However, over a decade later, the majority of gene therapy trials have been disappointing, as only low levels of circulating gene-transduced leukocytes were observed and with no clinical benefit (Anderson, 1998) (reviewed in Brenner, 1993). For example, the autosomal dominant form of chronic granulomatous disease (CGD) is caused by a deficiency of p47phox (an NADPH oxidase), and results in defective granulocytes that are ineffective in controlling bacterial/fungal infections (Bodine, 2001). Malech et al, used retroviral transfer with p47phox to transduce mobilized peripheral CD34+ blood cells from 5 patients with CGD. Following infusion back into patients, gene transfer was only observed in less than 0.1% of the circulating cells. On the positive side, the expression of p47phox in gene corrected granulocytes were as high or higher than normal, indicating if higher levels of gene therapy could be achieved, gene therapy may be beneficial to CGD patients (Malech et al, 1997). Despite these disappointing results, there have been a number of major achievements. Several groups attempted gene therapy for patients with SCID, caused by a deficiency in the adenosine deaminase (ADA) gene. In one of these studies, CD34+cells were obtained from cord blood of 3 new-borns that were homozygous for ADA mutations. These cells were retrovirally infected with the human ADA gene and re-infused back into the patients. Four years later, patients demonstrated long-term persistence of marked peripheral blood cells (T cells with normal ADA gene frequency of 1-10%, in other lineages 0.01-0.1%), demonstrating clinical success but not correction of the disease (Kohn et al, 1995). An even more promising study was performed by Cavazzana-Calvo et al, in which they recently reported the first gene therapy trial with 55 proven clinical benefit in two SCLD-X1 patients through retroviral gene transfer of the y-common chain of the LL-2 receptor into CD34+ cells (Cavazzana-Calvo et al, 2000). Interestingly, this study used an optimized retroviral transduction protocol (including an "improved" cytokine cocktail that favors HSC cycling and the use of recombinant fibronectin fragment (CH-296) to increase contact between the virus and target cells) that may be responsible for this success (Cavazzana-Calvo et al., 2000). In summary, it is apparent from the above studies that gene therapy/gene transfer to HSCs is not a simple feat. Although, there has been significant progress in increasing gene transfer to HSCs (pseudotyping, improved cytokine cocktail conditions, and strategies to increase contact between HSCs and the virus), there is still significant problems with achieving high level, sustained expression of a transgene. This is emphasized in the next section by the laborious and extensive history of trying to obtain a suitable globin gene vector for the hemoglobinopathies. 1.5.1 Hemoglobinopathies It is estimated that ~ 7% of the world's population are carriers for inherited hemoglobinopathy disorders, making them the most common monogenic diseases (reviewed in Weatherall, 2001). The human hemoglobins are tetramers that consist of 2 a-like chains (encoded by ct-like genes on chromosome 16) and 2 P-like chains (encoded by the P-like genes on chromosome 11). The structure of hemoglobin (Hb) changes during development; adult (HbA) and fetal (HbF) hemoglobins have a chains combined with P - (HbA, ct2,P2), 5-(HbA2, 0:282) or y-(HbF, a2,y2), whereas in the embryo, a-like chains called (^-chains combine with y (Hb Portland, ^272), or s-chains (Hb Gower, t^^i) 56 and a and s chains form Hb Gower 2 (02,82) (Weatherall, 2001). As shown in Figure 1.8, both a and B like genes are arranged on the chromosome in the order that they are expressed during development, such that different hemoglobin tetramers are assembled within the red cells at different developmental stages. Embryonic Hb is confined to the yolk sac stage of development, and thereafter is replaced by HbF, which is expressed in the fetal liver until birth. After birth, HbF is replaced by HbA and HbA2, which are expressed at high and low levels, respectively in the bone marrow (normal adults still continue to produce a small amount of HbF after birth, constituting -1% of the total Hb) (Weatherall, 2001). 57 (3 globin locus LCR / 0 .Ll 5 4 3 2 1 y HS chromosome 11 50 H o '35 ° I --2 e O O OH 40—^ 30 H .2 20H oo 10H / a \V-e 6 12 18 26 30 36 1 Post-conceptual age (weeks) Birth a globin locus HS-40 o 12 18 26 30 36 Post-natal age (weeks) ct2 a 1 j±romosorne_ 16 Figure 1-8 Schematic diagram of the globin gene loci. The human P globin locus is shown on the top. It is located on chromosome 11 surrounded by olfactory receptor genes (•). The locus control region is designated by the 4 hypersensitive sites (HSs) which are erythroid specific. HS5 is constitutively expressed and thought to have insulator function, the 3'HS is erythroid specific with no known function, HS 6 and 7 have recently been discovered and are thought to also play a role in LCR function. The a globin locus is located on chromsome 16 within a cluster of ubiquitously expressed genes (O). HS-40 is a HS that plays a role in regulating the a globin locus. Complied from Weatherall, 2001 and Engel and Tanimoto, 2000. Hemoglobinopathies can be divided into two groups - the structural variants and the thalassaemias. The thalassaemias are a group of disorders due to defective and imbalanced globin production, p thalassaemia results from over 200 different mutations in the P globin gene, in which the majority are point mutations that affect gene function at the transcriptional, translational and post-translational levels, a-thalassaemias result from over 80 different mutations in the a globin gene (Weatherall, 2001). The defective P 58 globin synthesis in p thalassaemia leads to an imbalance in globin chain production resulting in an excess of a chains. These excess chains aggregate in red cell precursors, and cause abnormal cell maturation and their premature destruction in the bone marrow. This can lead to a variety of problems including anemia, splenomegaly, bone disease, cardiac and endocrine damage (Weatherall, 1998). Within the structural variants, there are over 700 mutations that have been described in which the majority are due to single amino acid substitutions. Only 3 occur at high frequency in a number of different populations, one of which includes the sickle hemoglobin (HbS) (Weatherall, 2001). Sickle cell disease (SCD) is the result of an A to T transversion in the sixth codon of the human p globin gene (ps). This mutation leads to a substitution of a non-polar valine for a polar gluatamic acid residue on the surface of the hemoglobin molecule (HbS) (McCune et al, 1994). At low oxygen tensions, ps valine interacts with a natural hydrophobic pocket of a second Hb tetramer and initiates polymerization of HbS into long fibers. The formation of fibers reduces the flexibility of erythrocytes and leads to occlusion of small capillaries. Intracellular fiber formation also results in erythroid membrane damage and increased cell lysis. The ensuing disease is characterized by hemolytic anemia with episodes of severe pain and tissue damage that can result in stroke, heart damage, kidney failure and other complications (McCune et al., 1994). There are several modifiers that can affect the clinical phenotype of the hemoglobinopathies. Both of these originally became apparent by studying patients that had severe forms of P-thalassaemia, yet exhibited milder forms of the disease. In one case, these patients also had a-thalassaemia, which is not uncommon as they occur at 59 high frequencies in many of the same populations (Knox-Macaulay et al., 1972). The fact that co-existence of ct-thalassaemia can reduce the severity of P thalassaemia, provides evidence that the main problem is imbalanced globin chain production and not under production of Hb. Therefore, although the red cells of these patients might be largely deficient in hemoglobin, the anemia is less severe and the phenotype is milder (Weatherall, 2001). The other patients with midler phenotypes were found to contain high levels of HbF in their red blood cells. It is speculated that these increased levels are due to binding of the y chains to the excess a chains to produce HbF, resulting in a selective advantage of these red blood cells (Weatherall, 2001). The red blood cells with HbF are referred to as F cells and are genetically controlled (although the number of genes involved has not been determined), and therefore may account for some of the variability in HbF production in P thalassaemic patients (Garner et al, 2000). HbF has also been found to alleviate symptoms in patients with SCD. It contains two important amino acid residues that prevent polymerization and therefore is referred to as being anti-sickling. Within these patients red blood cells, there are three types of Hb tetramers: HbS (0^2), HbF (ct2y2), and a mixed tetramer (ct2,psy). The net effect of formation of these mixed tetramers is to decrease the concentration of ps that can contribute to polymerization (Eaton and Hofrichter, 1995). The significance of this comes from kinetic studies of the polymerization reaction, in which the delay time (time before polymerization occurs in the cells) is equivalent to the inverse of the concentration of HbS to the 30th power. That is, decreasing the concentration of HbS to even a small degree (by the presence of HbF), causes a dramatic increase in the delay time; the cell has enough time to go back to the 60 lungs and become reoxygenated before polymerization occurs, preventing the pleiotropic effects from occurring (Eaton and Hofrichter, 1995). There are several potential therapies for the hemoglobinopathies. Hydroxyurea is a chemical agent that has been used to increase the HbF levels, but has resulted in clinical success in only a minority of patients (Eaton and Hofrichter, 1995). Blood transfusions have been used to treat severe anemia by replacing the red blood cells that contain the defective globin. However, this is only a short-term treatment and can lead to iron overload that can cause organ failure (Olivieri, 1999). The only cure available is bone marrow transplantation, but for success and prevention of graft versus host disease, it requires an HLA-matched sibling (Weatherall, 1998). An attractive alternative is to perform gene therapy to deliver a functional P or y globin gene (or selectively decrease the a globin production in the case of p thalaessemia) to HSC of patients with these disorders. Murine models of SCD and the thalaessemias have shown that a functional p or y globin gene expressed at levels of 10-20% of the endogenous would be required for a therapeutic benefit (Blouin et al., 2000). 1.5.1.1 Development of globin gene vectors for the hemoglobinopathies Hemoglobin disorders were among the first diseases to be considered for gene therapy. However, the development of suitable vectors that conferred high, sustained expression in mature erythroid cells proved to be extremely challenging. The first encouraging reports of successful transfer of a P globin gene with proximal cis-regulatory elements into HSCs using an oncoretroviral vector appeared in the late 1980s. These studies demonstrated stable integration into HSCs of primary and secondary recipients 61 and erythroid lineage specific expression. Despite this, the expression levels were very low (<l-5% of murine p globin), were integration site dependent and few mice showed sustained expression (Bender et al, 1989; Dzierzak et al, 1988; Karlsson et al., 1988). This suggested that appropriate regulatory elements needed for expression were missing from these vectors. This was further validated by studying p thalaessemia patients, in which the P globin gene was intact but the region upstream of the locus was deleted. Using DNase I digestion as an indicator of "open" or active chromatin, several groups found four regions upstream of the locus that were erythroid specific, referred to as hypersensitive sites (HSs) (Forrester et al., 1987; Tuan and London, 1984). Since transcription factors are known to bind to HSs and therefore designate an important regulatory region, Grosveld et al, tested these HSs in a construct with the p globin gene in transgenic mice. They found it conferred position independent, copy number dependent, erythroid specific expression at levels equivalent to the endogenous mouse p globin (Grosveld et al, 1987). Therefore, these HSs comprise a distal regulatory element which is now referred to as the locus control region or LCR (Bulger and Groudine, 1999; Engel and Tanimoto, 2000; Li et al, 1999). As soon as the LCR was discovered, it was thought that inclusion of small core sequences of 200-400 bp for each of the HSs may significantly improve globin vector expression in RV vectors. However, inclusion of such elements proved problematic, as they caused vector instability and/or low titers (103-104 colony forming unit/ml or cfu/ml) of viral producer cell clones, indicating sequences within the RV vector were incompatible with LCR type sequences (Chang et al, 1992; Novak et al, 1990; Plavec et al, 1993). Two approaches resulted in the development of stable globin gene vectors containing HS fragments. Sadelain et al, tested multiple 62 orientations of several LCR fragments and deleted an AT rich region in the second intron. This led to an unrearranged pro virus and generated vector particles of high titers (106 cfu/ml) (Sadelain et al, 1995). However, in mouse transplant experiments, although this vector initially expressed ~ 5% P globin, the level was undetectable 4 months post-transplantation, suggesting complete silencing of the vector (Rivella and Sadelain, 1998). In the second approach, Leboulch et al, performed site directed mutagenisis to remove several poly A addition sites and cryptic splice sites and also generated a stable virus (Leboulch et al., 1994). Again, testing in a murine model, they found only 2 out of the 12 mice that were positive for the vector actually expressed it, also consistent with significant globin vector silencing (Raftopoulos et al., 1997). A novel approach was performed by Kalberer et al, in which a vector encoding the green fluorescent protein (GFP) and a cis-linked p globin cassette with the HS2 fragment was used to infect bone marrow cells. Subsequent vector-expressing cells were then isolated and transplanted into mice, with the idea of pre-selecting for those vector integration sites that may favor long-term expression (Kalberer et al, 2000). Although they obtained sustained, long-term expression in a cohort of the mice, almost half of them lacked expression of the transgene, suggesting silencing/position effects still continues to be a problem. Therefore, these retroviral studies demonstrate that important regulatory elements for expression are still missing from these vectors. However, a recent exciting report by May et al, used a P globin cassette with a small promoter, AT rich deleted intron 2, and larger LCR fragments (HS 2,3,4) in a LV vector. They were able to obtain therapeutic levels of P globin mRNA and protein in transduced bone marrow in a murine model of P thalaessmia (May et al., 2000). Despite these promising results, it is still not clear if such a vector is 63 resistant to silencing and position effects, and if it could transduce human HSCs at similar efficiencies and provide therapeutic levels of expression. Moreover, recent transgenic studies have demonstrated that in order to achieve 100% expression at all integration sites, a construct with all 4 HSs, extended 1.5 kb P globin promoter, and extended 3' enhancer is needed. The size of this construct is 8.8 kb, which is beyond the current size limitations for viral vectors (Pasceri et al, 1998). Therefore, a strategy that can systematically decipher the minimum regulatory elements that can overcome position effects/silencing and confer high levels of expression, yet remain a size suitable for a viral vector would be extremely useful. Chapter 4 describes the feasibility of using the Cre/lox integration strategy combined with ES hematopoietic in vitro differentiation as a model system for evaluating p globin vectors. 1.6 Thesis objectives/general strategy The ideal gene transfer strategy to HSCs would involve integration of a gene into a specific location in the genome. As summarised in the Introduction, this is not yet achievable (although significant progress is being made), because of the low frequency of integration. On the other hand, the currently used "gene augmentation" approaches have been instrumental in both gene function studies and as a potential tool for gene therapy. However, these approaches suffer from several drawbacks, including size limitations and lack of long-term, predictable levels of expression. The overall goal of this thesis was to investigate an alternative gene transfer strategy that may overcome these hurdles and further provides an improved platform for performing gene function and gene regulation studies in hematopoietic cells. The Cre/lox recombination system was chosen for its 64 ability to perform site specific modifications in the genome. To apply such an approach to hematopoietic cells, the efficient and stable integration properties of a RV vector was used to randomly introduce modified or heterospecific lox P target sites into the genome. This was then followed by using the Cre/lox system to integrate a gene flanked by compatible lox P sites into the chromosomally placed lox P site. Since the integration step is non-viral, there are no constraints on the size of DNA that can be integrated. The first objective of this thesis work was to characterise this approach in a number of cell lines, including hematopoietic. These studies focused on determing the feasibility of this strategy to enable the efficient, stable, integration of exogenous DNA into a hematopoietic cell line and primitive pluripotent ES cells. The ability to achieve predictable/reproducible expression upon re-targeting a locus was also explored. Using the ES in vitro differentiation model, I also investigated the feasibility of this strategy in more primitive hematopoietic cells (Chapter 3). My second objective was to investigate the use of this approach for studying gene regulatory mechanisms in hematopoietic cells using the ES in vitro differentiation model. The initial studies performed using this model involved characterising and improving conditions for the generation of hematopoietic cells. The cis-regulatory elements of the (3 globin gene were then evaluated as a method for designing better vectors, as the elements are well characterised, the constructs are extremely large and complex, and achieving high and sustained levels of expression with current viral vectors has been problematic. The utility of this strategy was demonstrated by analysing these constructs at the same integration site, and at 2 different sites, and comparing their expression properties to those that were randomly integrated (Chapter 4). 65 CHAPTER 2 MATERIALS AND METHODS 2.1 Cell lines, embryonic stem cell culture and assays 2.1.1 Cell lines The cell lines used in this study were obtained from the American Type Culture Collection (ATCC) unless otherwise specified. The human erythroleukemia cell line, K562, was originally derived from the pleural effusion of a patient in blast phase of chronic myelogenous leukemia (Lozzio, 1975) and was maintained in RPMI1640 with 10% fetal calf serum (FCS) (Gibco/BRL, Life Technologies, Burlington, Canada). The murine IL-3-dependent pro-B cell line Ba/F3 (Palacios and Steinmetz, 1985) was maintained in RPMI 1640 supplemented with 5% FCS and 10 ng/ml of murine IL -3 (supplied as a COS cell derived supernatant prepared at the Terry Fox Laboratory (TFL)). The murine NTH 3T3 fibroblast cell line (Jainchill et al, 1969) was cultured in Dulbecco's modified essential medium (DMEM) and 10% calf serum (Gibco). Helper free recombinant retrovirus was generated by using either the ecotropic GP+E-86 (Markowitz et al, 1988b) or the amphotropic GP+AM12 (Markowitz et al, 1988a) stable packaging cell lines and were maintained in DMEM supplemented with 10% heat inactivated (55°C for 30 minutes) newborn calf serum (Gibco). All cells were cultured at 37°C in a humidified atmosphere of 5% CO2 in air. 66 2.1.2 Murine embryonic stem (ES) cell culture and assays 2.1.2.1 Maintenance of undifferentiated ES cells Unless otherwise stated, all reagents for ES cell culture and in vitro differentiation of ES cells were purchased from StemCell Technologies Inc. (STI) (Vancouver, Canada) and all growth factors were used as a diluted supernatant from transfected COS cells prepared at the TFL. CCE ES cells (kindly provided by Dr. G. Keller, National Jewish Center, Denver, CO) were maintained on gelatinized dishes in DMEM supplemented with 15% FCS (STI #6901), 2mM L-glutamine, 0.1 mM nonessential amino acids, lOOuM monothioglycerol (MTG) (Sigma Chemical Co, St. Louis MO), and 10 ng/ml leukemia inhibitory factor (LIF). 2.1.2.2 In vitro differentiation of ES cells The method used for differentiation of ES cells involved initial embryoid body (EB) formation in methylcellulose followed by dissociation of EBs and hematopoietic assays based on the procedure developed by Keller et al. (Keller et al., 1993) and described previously (Helgason et al., 1996). In brief, forty-eight hours prior to differentiation, cells were passaged at a low density (5 x 104 cells per 25 cm2 flask) in Iscove's modified dulbecco's medium (LMDM) supplemented as described above for the maintenance of ES cells. Cells were then resuspended in primary differentiation methylcellulose (containing 0.9% methylcellulose (STI, #MC-3120) supplemented with 15% FCS (STI, #6900), 150 uM MTG (Sigma), 2mM L-glutamine, 50 ng/ml murine steel factor (mSF) and LMDM to volume) to yield a final density of 300 cells per ml and 1.0 ml aliquots were distributed into 35 mm petri-style dishes. Cultures were fed at day 7 of 67 differentiation by layering 0.5 ml of 0.5% primary methylcellulose containing 15% FCS (STI # 6900), 150 uM MTG, recombinant human erythropoietin (rhEpo; STI 3U/ml), mSF (160 ng/ml), and 30 ng/ml of both murine IL-3 (mIL-3) and human IL-6 (hIL-6) onto each dish. At various time points of primary differentiation, EBs were harvested and cells were plated in secondary methylcellulose cultures to detect hematopoietic progenitors, plated in liquid culture conditions for the generation of mature hematopoietic cells and also used for RNA analysis. EBs were disrupted either by treatment with Trypsin-EDTA (Gibco) for 2 minutes (day 6 EBs) or incubation at 37°C for one hour in collagenase (day 7 or later EBs), followed by passage through a 21 gauge needle to obtain a single cell suspension. Cells were plated at a density of 3 x 104 cells/dish in secondary methylcellulose (0.9% methylcellulose (STI #3120) with 15% FCS (STI # 6900), 2 mM L-glutamine, 150 uM MTG, 1% bovine serum albumin, 10 ug/ml insulin, 200 ug/ml transferrin (STI BIT # 9500) supplemented with 3U/ml rhEpo, 160 ng/ml mSF, 30 ng/ml mIL-3 and 30 ng/ml of hIL-6) and hematopoietic colonies were scored microscopically after seven days using standard criteria. To obtain mature hematopoietic cells, 5 x 105 cells/ml were plated in liquid culture conditions which are identical to that described above for secondary methylcellulose conditions but replacing the methylcellulose with IMDM. 2.1.2.3 Generation of ES derived erythroid cells The generation of erythroid cells from differentiating ES cells was evaluated in Chapter 4 using the above liquid culture conditions and those previously described for the growth of erythroid cells from human cord blood (Panzenbock et al, 1998). Evidence for 68 mature erythroid cells was assessed by expression of the erythroid specific lineage marker Terl 19, p globin expression and morphological analysis. Cells were disrupted from day 10 EBs and plated at density of 5 x 105 cells/ml in the above liquid culture conditions for six days and then analyzed based on these criteria. The protocol previously described for human erythroid cells consists of a proliferation phase where erythroid progenitors are first selectively amplified, followed by a differentiation phase where these progenitors then undergo terminal differentiation (Panzenbock et al., 1998). Modifications to this protocol were made for adaptation to murine ES cells. In brief, cells were disrupted from day 10 EBs and 1 x 106 cells/ml were plated in proliferation media for 6 days, which consisted of 15% FCS (STI # 6900), 2mM L-glutamine, 150 uM MTG, 1% BSA (Intergen), 200 ug/ml human transferrin (Bayer) supplemented with 3U/ml rhEpo, 160 ng/ml mSF, 40 ng/ml of insulin growth factor-1 (Sigma), 10"6 mol/L dexamethasone (Sigma), and 10"6 mol/L p-estradiol (Sigma F4389). After six days of proliferation, cells were induced to differentiate for two days by washing twice with serum-free medium and seeding cells at a density of 2 x 10° cells/ml in culture medium containing 3 U/ml rhEpo, 80 ng/ml mSF and lug/ml bovine insulin (Sigma). 2.1.2.4 Generation of ES derived mast cells Mast cells were generated from differentiating ES cells in Chapter 4 to create a non-erthyroid cell line to examine specificity of the human p globin constructs. Conditions for mast cell development from ES cells have been described previously (Tsai et al., 2000). In brief, cells from disrupted day 10 EBs were plated at a density of 5 x 105 cells/ml in mast cell medium containing 15% FCS (STI # 6900), 150 uM MTG, 2mM L-69 glutamine, supplemented with 50 ng/ml mSF and 30 ng/ml mIL-3. Half media changes were done every other day. After 2 weeks, > 95% of the cells were identifiable as mast cells as determined by morphology and FACS analysis for lineage specific markers (the high affinity IgE receptor and the stem cell receptor - ckit). 2.2 Gene transfer techniques 2.2.1 Retroviral generation and infection of cells Retroviral gene transfer was used as the initial step to introduce the lox P target sites into the genome of cells. The production of high titer helper free retrovirus was carried out by standard procedures (Pawliuk et al, 1994), using calcium phosphate transfection into the amphotropic GP+AM-12 packaging cell line and also using the supernatant to infect the ecotropic packaging cell line GP+E-86. Both viruses were used interchangeably throughout this study to infect a number of different murine and human cell lines. The viral titers were determined by transfer of hygromycin resistance to NTH 3T3 cells to be 3 - 5 x 105 cfu/ml. Infections were carried out by collecting the viral supernatants, followed by filtering through a 0.45uM filter (Millipore; Bedford, MA) and supplemented with polybrene (Sigma) (2ug/ml for ES, 7ug/ml for Ba/F3, K562, and NTH 3T3 cells) before addition to the cells. Freshly passaged ES cells were resuspended at 5xl05 cells in a 100 cm dish containing 3 ml of viral supernatant plus 5 ml ES maintenance media, whereas 1 x 105 NTH 3T3 cells were seeded in a 60 cm dish with 4 ml viral supernatant. Logarithmically growing Ba/F3 and K562 cells were resuspended at 2x10 cells in 5 ml viral supernatant per 21 cm flask and all cells were incubated with two changes of viral supernatant over 24 hours. At 48 hours post-infection, media was 70 supplemented with 250 ng/ml Hygromycin B (BOEHPJNGER) for the ES cells and 350 ug/ml Hygromycin B for the NIH 3T3, Ba/F3 and K562 cells. For experiments where a single integration site of the initial retroviral vector was desired, single clones from each of the cell lines were isolated. Following seven days of selection, ~ 25 individual colonies from the NTH 3T3 and ES cells were picked into a 96 well plate and expanded. Approximately 50 -100 Ba/F3 and K562 cells were plated in methylcellulose (STI #H4100) supplemented with 50% EVIDM, 10% FCS, 350 ug/ml Hygromycin B and with 10 ng/ml mIL-3 for the Ba/F3 cells. Southern blot analysis was performed on expanded colonies to detect clones containing a single integration site and with the provirus intact. 2.2.2 Electroporation 2.2.2.1 Site directed integration Electroporation was utilized throughout this study as the second step for site directed integration. The electroporation conditions for each of the cell lines was first optimized using the Biorad gene pulser (Biorad, Hercules, CA) and 10 ug of pCMV-GFP-C1 plasmid (Clontech Palo Alto, CA) whereby the efficiency ranged between 30-50%. Site directed integration was performed by electroporating 100 ug of the promoter-less integrating vector with 20 ug of a Cre expression plasmid into cells harboring the initial lox target site. For each experiment, a control for the gene transfer efficiency (100 ug of pCMV-GFP-Cl or pMCI-neo (Thomas and Capecchi, 1987)) and a control for Cre mediated integration (100 ug of the integration vector alone) were used. Approximately 71 Ix 107 undifferentiated ES cells were resuspended in 400 p.1 of Phosphate buffered saline (PBS) mixed with DNA and incubated on ice for ten minutes before and after electroporation with optimized settings set at 240 volts and 500 u,F. Using final conditions set at 270 volts and 960 pF, ~ 5 x 106 Ba/F3 and K562 cells were resuspended in 500 pi of RPMI mixed with DNA and incubated at room temperature for 5 minutes before and after electroporation. The conditions used for NLH 3T3 cells are similar to those described above for the Ba/F3 and K562 cells but with the exception of using 250 volts. In Chapter 4, electroporation conditions were also optimized for differentiated ES cells such that site directed integration could be executed in primitive hematopoietic cells. Approximately 1 x 107 cells were resuspended in 1 ml of DMEM and DNA, incubated at room temperature for 5 minutes before and after electroporation with final settings set at 500 uF and 400 volts. For Cre mediated integration where the MSCV LTR drives the expression of the neomycin gene, media was changed 24 hours later and 1 mg/ml (approximately 0.92 mg/ml active compound) of the neomycin analogue G418 (Gibco/BRL) was added to the media 48 hours later. Studies performed using cell surface or intracellular markers such as THY-1 or the green fluorescent protein (GFP) respectively, were analyzed three days post-electroporation using flow cytometry. 2.2.2.2 Random integration In Chapter 4, each of the human p globin constructs were randomly integrated into undifferentiated ES cells. This was achieved by co-electroporating 8 u.g of the linearized human P globin constructs with the linearized pMCI neomycin plasmid at a 10:lmolar 72 ratio. The media was changed 24 hours later to allow cells to recover, and then replaced with media containing 800 ug/ml G418. After 7 days, visible neomycin resistant colonies were pooled, expanded for differentiation and Southern blot analysis performed for the presence of the globin transgene. In Chapter 3, the original lox P target sites were introduced by electroporating 10 ug of the linearized RAVtkneo plasmid into undifferentiated ES cells and selection applied as described above. 2.3 Molecular analysis 2.3.1 DNA constructs Retroviral gene transfer was used as the initial step to introduce the lox P target sites into the genome of cells. The proviral plasmid DNA used to create the recombinant retroviral vector consists of a hygromycin resistance-thymidine kinase fusion gene flanked by lox sites that differ by one base pair in their spacer region (wild type lox P site; L2 and mutant lox P site; Ll) and whose expression is driven by the MSCV LTR (Fint 52, kindly provided by Dr. P. Leboulch, Harvard University, Cambridge, MA). A series of integration vectors were used throughout this study for the second step of site directed integration. All the vectors contain a promoter-less cassette (either a gene or selectable marker) flanked by the heterospecific lox P sites and include: the neomycin resistant gene (Fint 1908), the GFP gene (Fint 81), (both gifts from Dr. P. Leboulch) and a human cDNA THY-1 construct. The floxed human cDNA THY-1 construct was generated by removing the 700 bp Neo I/Not I fragment encompassing the GFP gene from the Fint 81 plasmid and ligating the empty vector with a 650 bp Sail fragment containing the THY-1 cDNA from the pCTV78 plasmid (gift from Dr. R. Kay, TFL, Vancouver, Canada). As a 73 control for gene transfer, several plasmids that contain a functional promoter were used throughout this study: the red shifted variant of the wild type GFP (pCMV-GFP-Cl), the yellow-green variant of the wild type GFP (pCMV-YFP-Nl) (Clontech, Palo Alto, CA) both expressed from the human cytomegalovirus (CMV) promoter and a plasmid containing the neomycin resistant gene expressed from the MCI promoter (pMCI neo). Two Cre expression plasmids (pBS185, and pCAG-CRE) were used interchangeably at various times during this study. The Cre coding sequence is driven the CMV promoter in pBS185 (Sauer and Henderson, 1990) (Life Technologies, Gaithersburg, MD) and by the chicken (3 actin promoter in pCAG-CRE (a generous gift from Dr. P. Vassilli, Geneva, Switzerland (Araki et al, 1995)). In Chapter 4, ES cell clones were created that contained a series of randomly integrated human p globin constructs. The B33 and BGT9 human p globin constructs were gifts from Dr. J. Ellis (Hospital for Sick Children, Toronto, ON) and described in detail elsewhere (Pasceri et al, 1998). In brief, B33 (referred to in this thesis as HS 1-4 p globin containing construct), contains a 4.0 kb DNA fragment consisting of ~ 1 kb each of the 4 hypersensitive sites (HSs), 1.5 kb P globin promoter, the entire P globin gene and 1.65 kb 3'enhancer. BGT9 (referred to as HS3 P globin construct), contains the 850 bp DNA fragment of HS3 (230 bp HS3 core element and -600 bp flanking region), 815 bp p globin promoter, the entire P globin gene and 1.65 kb 3' enhancer. The basic human P globin construct was created by removing the HS3 site from the BGT9 construct. Cre mediated integration was performed with each of the above globin constructs. A parent vector (kindly provided by Dr. P. Leboulch) that consists of a promoter-less neomycin resistant gene, flanked by heterospecific lox P sites was used to clone each of 74 the above constructs, such that site directed integration could be executed. Each of the vectors were digested to remove the P globin cassette as follows: an 8.8 kb EcoRV fragment from HS 1-4, a 5.0 kb EcoRV/Sallfragment from HS3, and a 4.2 kb Kpn I/EcoRVfragment from the basic construct, and then cloned into the multiple cloning site, down stream of the neomycin resistant gene. A plasmid was used in Chapter 3 that would allow testing of site directed integration in differentiating ES cells. RAVTkneo (generated by P. M. Rosten) consists of a Herpes Simplex virus thymidine kinase (tk) gene promoter which drives the expression of a neomycin resistant gene, which in turn is flanked by inverted heterospecific lox P target sites (Ll and 2L) to further increase the integration frequency. The human polypeptide chain elongation factor 1 alpha (EFl-a) promoter, which has been previously shown to stay on during ES differentiation (Hanaoka et al., 1991), was then cloned upstream of this entire cassette. The integrating exchange vector (gift from Dr. R. Kay) contains a promoter-less GFP construct similar to the one described above but with compatible lox P sites (Ll and 2L). 2.3.2 DNA Analysis 2.3.2.1 Isolation of DNA The genomic DNA from all cell lines used in this study were isolated using the reagent DNAzol and the recommendation provided by the manufacturer (Gibco/BRL). An alternative procedure was used for isolating DNA in Chapter 4 due to the small number of cells available. In brief, these cells were washed twice with PBS and lOOpl of IX lysis buffer (0.32M Sucrose, lOmM Tris HCL pH 7.5, 5mM MgCL2, 1% Triton X-100) and 10 75 mg of protienase K (Gibco/ BRL) was added and incubated for one hour at 55°C, followed by heat inactivation of the enzyme at 95°C for 10 minutes. The lysate was then directly used for PCR analysis. 2.3.2.2 Southern blot analysis and probes used Genomic DNA (10 u.g) was digested using the appropriate restriction enzyme. The DNA fragments were fractionated on a 1% agarose gel, followed by treatment for 8 minutes with a 0.1M HCL solution to break large fragments of DNA into smaller more transferable ones, and then denatured using a 0.5 M/1.5M NaCL solution for 30 minutes. The DNA was then transferred by using 10X SSC and fixed by vacuum heat at 80°C for one hour onto a nylon membrane (Zetaprobe, Bio-Rad) as previously described (Pawliuk et al., 1994). Hybridization and washes were performed as previously described in Sambrook and Maniatis (1989). Probes used in this study included: an MluI/BamHI 860 bp fragment of the neomycin resistant gene derived from the pMCI neo plasmid, a BamHI/EcoRI 900 bp fragment encompassing intron 2 of the human P globin gene from the HS3 p globin containing plasmid, a Narl/EcoRI 111 bp fragment of the GFP gene from the pCMV-GFP-Cl plasmid (Clontech), a Clal/Sstll 1 .1 bp fragment from the hygromycin gene from the CT3VB plasmid (gift from Dr. R. Kay) and a Narl/Ncol 800 bp fragment containing the entire human THY-1 cDNA from the cloned THY-1 plasmid. 2.3.2.3 PCR analysis Unless otherwise stated, all enzymes and nucleotides were obtained from Gibco/BRL, Life Technologies, Burlington, Canada. PCR analysis was used for the kinetic studies in Chapter 3 and was performed as recommended by the manufacturer. 76 The same protocol and conditions were used for all PCR reactions and primer sets. In brief, the PCR reaction consisted of 1 X PCR buffer, 0.2 mM of each dNTP, 1.5 mM MgCb, 0.2 uM of each primer and 2.5 units of platinum Taq polymerase. A 35 cycle PCR was performed using the following parameters: denaturation at 94°C for 30 seconds, annealing at 68°C for 30 seconds and extension at 72°C for 1 minute. The sequence of the primers used were: a sense primer in the 5' leader sequence of the MSCV LTR (5' TGTTTGCGCCTGCGTCTGTA 3') and an antisense primer in the GFP sequence (5' TGTTGTAGCTCCTGCCGTCG 3') which gives a distinctive band of 1.2 kb. Also in Chapter 3, PCR was also used to detect site directed integration in differentiated ES cells. A primer was designed in the eLF-a promoter (5' CCTCGATTAGTTCTCGAGCT 3') and the same antisense GFP primer described above were used together to give a characteristic band size of 750 bp. As a control for DNA amplification, an additional primer set was constructed that detects the original construct. The same EFl-a promoter was used with an antisense primer made to the neomycin gene (5' TTGAGCCTGGCGAACAGTTC 3') which results in a band at 1.2 kb. 2.3.3 RNA analysis 2.3.3.1 Isolation of RNA Genomic free RNA was isolated from undifferentiated ES cells, EBs, and differentiating ES cells using the TRIzol reagent (Gibco/BRL) and manufacturer's instructions and used to perform RT-PCR or RNASE protection. 77 2.3.3.2 RT-PCR RT-PCR was used in Chapter 4 for the detection of murine P major and human p globin in EBs and differentiating ES cells. For RT-PCR analysis, reverse transcription was carried out utilizing 0.5ug of RNA, Superscript II reverse transcriptase, .01M DTT, 200 ng/ul random hexamers, and 2.5 mM each of dNTPs in a final volume of 20 ul. Reverse transcription was performed at 70°C for ten minutes, followed by annealing and extension at 42°C for 30 minutes, and then heat inactivation at 70°C for 15 minutes. One fourth of the resulting cDNA was subjected to a 26 cycle PCR using the following parameters: 94°C for 1 minute, 60° or 55°C (see below) for 1 minute, and 72°C for 1 minute with each gene specific primer pair. All the primers used were designed such that they spanned at least one intron. The sequences of the primers and expected sizes of the cDNA specific transcripts are as follows: murine P major 5' primer (in exon 1) 5'CACAACCCCAGAAACAGACA 3' and 3' primer (in exon 3) 5'CTGACAGATGCTCTCTTGGG 3' results in a 578 bp product at 60°C annealing temperature. Human p globin 5'primer (in exon 1) 5 'GCAAGGTGAACGTGGATGAAG 3' and 3' primer (in exon 3) 5 'TTCTGATAGGCAGCCTGCACT 3' results in a 349 bp product at 60°C annealing temperature. Murine pHl 5' primer (in exon 2) 5'CTCAAGGAGACCTTTGCTCA 3' and 3' primer (in exon 3) 5'AGTCCCCATGGAGTCAAAGA 3' results in a 265 bp product at 55°C annealing temperature. As a control for cDNA amplification, primers were made to the murine HPRT gene as follows: 5' primer (in exon 7) 5' GCTGGTGAAAAGGACCTCT 3' and 3' primer (in exon 9) 5'CACAGGACTAGAACACCTGC 3' which confers a 249 bp product at 55°C annealing 78 temperature. Reaction products were analyzed on a 1% agarose gel and visualized by ethidium bromide staining. 2.3.3.3 RNASE protection The RNASE protection assay (RPA) system was used in Chapter 4 to quantify the amount of human P globin mRNA. Approximately 5 u.g of total RNA was isolated from differentiating ES cells and RPA was performed as recommended in the RiboQuant instruction manual (PharMingen, San Diego, CA). RPA probes were uniformly labeled by in vitro transcription with SP6 polymerase and in the presence of [oc-33P] UTP under the same experimental conditions as described in (Leboulch et al., 1994). A human specific probe was kindly provided by Tom Maniatis [Harvard University, Cambridge]: the specific protected fragment is 350 bp long and corresponds to the first and second exons of the P globin mRNA up to the exonic BamHI site. A murine specific probe was constructed so that a 145 bp fragment corresponding to the first exon of P major globin mRNA is protected [kindly provided by D. Galson, MIT, Cambridge]. RPA products were assessed using a denaturing polyacrylamide gel electrophoresis followed by autoradiography. The intensity of the radioactive bands corresponding to the specific protected fragments were measured using the phoshoimager STORM 860 and the Image Quant Software from Molecular Dynamics (Sunnyvale, CA, USA). The human P globin mRNA expression was normalized relative to the mouse P globin mRNA as descibed in (Leboulch, 1994). Human P globin/murine P globin mRNA ratios were corrected for the number of uridine residues in each probe (factor 2.5) and on the volume loaded per well (loaded 1/10th of the volume for murine Pmajor): 79 human B globin murine P major [human p globin band] x 100 [murine P major band] x 10 x 2.5 2.4 FACS analysis 2.4.1 Staining for FACS analysis Fluorescent activated cell sorting (FACS) analysis of cells that express GFP were first rinsed once with Hanks Balanced Salt Solution containing 2% FCS (HF) and then resuspended in HF with lug/ml of propidium iodide (HF/PI) (Sigma Chemicals, St. Louis, USA.). HF/PI staining was performed on all cells prior to flow cytometry to distinguish between live and dead cells. Analysis of cells that express human THY- 1 were rinsed once with HF, stained using a 1:20 final ratio of 5E10-PE antibody (gift from Dr. P. Lansdorp, TFL, Vancouver, Canada) for 30 minutes and then rinsed with HF/PI. Differentiating ES cells were analyzed for the percentage of erythroid cells (using the erythroid specific lineage marker Terl 19), formation of mast cells (for presence of the IgE receptor and cKit receptor expression) and the proportion of cells that express the human p globin protein. Unless otherwise specified, all cells were labeled with phycoerythrin-conjucated (PE) antibodies (PharMingen). Approximately, 1-3 x 106 cells were resuspended in a 100 pi of HF and incubated at a final concentration of 5 ug/ml with an anti-mouse IgG Fc receptor antibody (2.4G2, produced from a hybridoma provided by Dr. S. Szilvassy) to prevent nonspecific binding for 15 minutes on ice. For erythroid detection, cells were stained at a final concentration of 1:500 with the Terl 19 antibody for 30 minutes on ice. Mast cells were identified by staining with the ckit antibody at a final concentration of 1:1000 and the anti-IgE receptor antibody (prepared in 80 Dr. G. Krystal's lab) at a final concentration of 1:25 for 30 minutes on ice. As a negative control, cells were also stained with the strepdavidin antibody at a final concentration of 1:2000 for the same time period. After the 30 minutes of staining, all cells were washed with HF/PI prior to flow cytometry. The proportion of cells expressing human p globin protein was evaluated by an intracellular P globin staining protocol previously described in Kalberer et al, 2000. In brief, ~ 3 x 106 cells were first fixed in a 1% Formaldehyde/0.05% Glutaraldehyde solution and then washed twice with PBS. The cells were then permeabilized with a 5% non-fat dry milk/PBS solution and then stained with a biotinylated anti-human p globin antibody at a final concentration of 1:500 (Biolabs, Northbrook, IL), followed by strepdavidin staining at a final concentration of 1:2000. 2.4.1.1 FACS analysis and sorting Flow cytometric analysis was performed using the FACsort or FACscalibur machine (Becton Dickinson, CA). Cells were sorted using the FACStar+ (Becton Dickinson) equipped with a 5-W argon and a 30-mW helium neon laser. 81 CHAPTER 3 SITE DIRECTED INTEGRATION IN HEMATOPOIETIC CELLS USING THE CRE/LOX RECOMBINATION SYSTEM 3.1 Introduction Gene transfer to hematopoietic cells has many important potential and current applications ranging from gene function studies to gene therapy (Bodine, 2001). As summarized in the Introduction chapter, retroviral vectors are the most commonly used method to efficiently transfer a gene to HSCs and achieve stable integration. Although proven to be a powerful tool in this regard, they suffer from several drawbacks including size limitations and the ability to confer long-term, predictable levels of transgene expression (Verma and Somia, 1997; Vigna and Naldini, 2000). Site specific recombination systems are an attractive alternative integration strategy that have been exploited for carrying out genetic modifications in undifferentiated ES cells and in mammalian cell lines (Bouhassira et al, 1997; Feng et al, 2001; Soukharev et al., 1999). Site-specific recombinases such as Cre and Flp mediate recombination at their specific target sequences (lox P or FRT, respectively) and in principle can be applied to excise as well to insert DNA into the genome depending on the orientation of the target sites (Schubeler et al, 1998). Cre/lox recombination has been shown to occur more efficiently than the FLP/FRT system in mammalian cells and for this reason has been most widely used (Dymecki, 1996; Ludwig et al, 1996). This method presents several powerful uses that include: (1) a more accurate analysis of cis-82 acting sequences of a gene since constructs containing different regulatory elements are analyzed within identical chromosomal contexts (Bouhassira et al, 1997; Feng et al., 2001), (2) the introduction of large DNA constructs since there is no size restrictions (Alami et al, 2000) and (3) the ability to achieve more predictable and reproducible transgene expression levels by directing a transgene into a locus with desired expression properties (Feng et al, 1999; Kolb et al, 1999; Schubeler et al, 1998). There has been some effort to implement the Cre/lox recombination system to hematopoietic cells. Studies by Bouhassira et al, have demonstrated the feasibility of using this approach for novel analysis of regulatory elements of the P globin gene in the mouse erythroleukemia cell line (Bouhassira et al, 1997; Feng et al, 2001). These studies used a non-viral delivery method to introduce the initial target sites into the genome of cells. This method is feasible for studies in cell lines, but is inefficient for primary hematopoietic cells. Several groups have used retroviral gene transfer to efficiently integrate the initial lox P sites, followed by using the Cre/lox system to delete genes from the integrated provirus in primary cells, including primary hepatocytes and myogenic cells (Berghella et al, 1999; Kobayashi et al, 2000). However, the feasibility of such an approach for integration of exogenous DNA in primitive hematopoietic cells has not yet been investigated. The focus of this work was to investigate an alternative gene transfer strategy that uses a retroviral vector to introduce two lox P sites that differ by a single base pair in their spacer region (heterospecific lox P sites), followed by using Cre/lox mediated integration of a DNA cassette with compatible lox P sites introduced by electroporation into hematopoietic cells. This strategy may overcome some of the hurdles associated with 83 existing viral approaches and provide a useful tool for genetic manipulation of hematopoietic cells. In the work presented in this Chapter, I demonstrate the feasibility and efficiency of this approach in a number of cell lines, including those of hematopoietic origin, and illustrate its use in achieving predictable transgene expression levels. I further document the feasibility of this strategy in more primitive cells, modeled here by ES cells and use this model to demonstrate for the first time that Cre/lox site directed integration can occur in hematopoietic progenitors. 3.2 Results 3.2.1 Feasibility of Cre/lox site directed integration Retroviral gene transfer was used as the initial step to introduce lox P target sites into the genome of NTH 3T3 cells, undifferentiated ES cells (CCE cell line), the K562 human erythroleukmia cell line and the murine Ba/F3 hematopoietic cell line. The retrovirus contains the hygromycin resistance-fhymidine kinase fusion (hygrotk) gene flanked by lox sites that differ by one base pair in their spacer region (wild type lox P site; L2 and mutant lox P site; Ll). The expression of the hygrotk fusion gene is under the control of the MSCV LTR promoter and therefore cells that have been transduced will become hygromycin resistant (Figure 3.1). Clones for each of the cell lines that contained a single integration site and with the provirus intact, as assessed by Southern blot analysis (data not shown) were chosen for further studies. A promoter-less plasmid that contains the neomycin resistant gene flanked by heterospecific lox P sites (L2 and Ll in Figure 3.1), was used as a reporter to test for Cre mediated integration. A clone for each of the cell lines that contained a single hygrotk 84 target sequence was electroporated with this plasmid, either alone or in combination with a Cre expression plasmid under the control of the cytomegalovirus (CMV) promoter. Only cells that have simultaneously taken up both the Cre and the neomycin targeting plasmids should result in site directed integration, which can be detected by expression of the neomycin resistant gene. This is illustrated in Figure 3.1, in which Cre mediates integration via compatible lox sites on the plasmid and in the genome, followed by deletion of intervening sequences to result in an exchange of neomcyin for the hygrotk fusion gene, such that the neomcyin gene will now be expressed from the viral LTR promoter. 85 v/\ fr LTR Ll hygrotk LTR Kpnl 2.5 kb 1 Kpnl Electroporation + Promoter-less targeting construct + Cre Cre-mediated integration/ Cre expression plasmid Retroviral transfer of lox P sites Introduction of targeting construct and Cre expression plasmid Cre .Random integration LTR Ll neo L2 LTR Kpnl i LlH HL2 Kpnl l> 2.2 kb GFP Kpnl 1.2 kb ] > 2.2 kb [GFP] Figure 3-1 Schematic representation of Cre/lox mediated integration. Cre mediated integration will result in cassette exchange and expression of the neomycin gene for the hygrotk fusion gene. Targeted integration is shown by digestion with Kpnl and probing with neomycin to give a 2.2kb and loss of the 2.5kb band when probing with hygromycin. A promoter-less GFP plasmid was used in subsequent experiments (refer to section 3.2.2.) in place of the neomycin plasmid. Primers were made to the MSCV LTR promoter and within the GFP gene (depicted as arrows) in which a 1.2kb band is indicative of Cre mediated integration. Three days after electroporation, cells were selected with the neomycin analogue, G418, and the number of subsequent neomycin resistant colonies were assessed (Figure 3.2A). As predicted, only those clones that received both neomycin and Cre recombinase plasmids resulted in a significant number of neomycin resistant colonies. To confirm that 86 these neomycin resistant subclones were the result of Cre mediated integration, Southern blot analysis was performed. As shown in Figure 3.1, Cre mediated integration will result in a 2.2kb band following digestion with Kpnl and probing with the neomycin resistant gene. Since individual clones were chosen that contained a single hygrotk target sequence, targeted integration is further confirmed by the loss of a 2.5kb band following probing with the hygromycin resistant gene. All of the subclones from the NTH 3T3 cell line that received the promoter-less neomycin targeting and Cre expression plasmids were correctly targeted (Figure 3.2B). In contrast, subclones that received only the promoter-less neomycin plasmid or a neomycin plasmid with a functional promoter and no lox target sites (pMCI-neo) were randomly integrated, confirming site directed integration is dependent on Cre recombinase and the presence of lox P sites on the plasmid. 87 A. NIH 3T3 Ba/F3 Clones B. hi T3 •a Probe used neomycin hygromycin T3 <u o CP c a O O o c X o J ? S o 3 ^ S T3 promoter-less neo-lox ^3 o 'g plasmid +Cresubclones ii! IS I 8 | S 3 fi « a ft 2.2 kb <-2.5 kb Figure 3-2 Cre mediated integration is dependent on expression of Cre recombinase and the presence of lox P sites on the plasmid. A. A single clone for each of the cell lines was electroporated with the promoter-less neomycin targeting plasmid either alone or in combination with the Cre expression plasmid. Only clones that have taken up both Cre and the neomycin targeting plasmids resulted in neomycin resistant colonies. B. Southern blot anlaysis confirms targeted integration in subclones from the NIH 3T3 clone. Subclones that received only the promoter-less neomycin plasmid or neomycin plasmid with no lox sites (pMCI-neo) were randomly integrated as shown by the larger 2.2kb band when probing with neomycin and the presence of the 2.5kb band when probing with hygromycin. 88 Subsequent Southern blot analysis on individual subclones for the ES cell line and the Ba/F3 hematopoietic cell line further confirmed that targeted integration occurred, demonstrating that this approach is a generally applicable phenomenon independent of the particular cell type used (Figure 3.3). neomycin plasmid; no lox sites (pMCI-neo) rh Probe used: neomycin hygromycin a o u u •a = 1 OJ 3 3 « sr <r o oj o ,fl a a 5 o o g c c c i e promoter-less neo-lox plasmid +Cre subclones Probe used: neomycin 2.2 kb-* M-2.2kb 2.5 kb-*\ K2.5 kb 1 I + . g promoter-less neo-lox § J? plasmid +Cre subclones JJ 2 I 1 1 S _& I I &_ ES (CCE cell line) Ba/F3 cell line Figure 3-3 Southern blot analysis confirms Cre mediated integration in a clone derived from the ES (CCE cell line) (left) and the murine Ua/I 3 hematopoietic cell line (right). Subclones from both cell lines that received both the promoter-less neomycin targeting and Cre expression plasmids were correctly targeted. The arrow in the ES blot demonstrates that in addition to the correctly targeted band, the plasmid also integrated at random in the genome. Random integration was observed in subclones with or without the initial lox target sequence (+/- hygrotk) that received the neomycin plasmid with no lox target sites (pMCI-neo). 89 In addition, as illustrated in Table 3.1, the majority of the integrants are predominantly (>95%) single copy insertions that occurred at the chromosomal lox P target site, with no random integration of the plasmid elsewhere in the genome. Table 3-1 Accuracy of Cre mediated integration Clones Number of Number of Accuracy of targeted subclones subclones correctly integration (%) screened targeted* Ba/F3 8 8 100 NIH3T3 3 3 100 CCE 20 19 95 * Targeted integration with no random integration occuring elsewhere in the genome as shown by Southern blot analysis. 3.2.2 Site Directed Integration in the Ba/F3 hematopoietic cell line The green fluorescent protein (GFP) was used as a reporter (in place of the neomycin resistant gene as shown in Figure 3.1) to determine the efficiency of integration. Initial experiments were performed to ascertain if the promoter-less GFP-lox targeting plasmid could be used to accurately assess the stable integration frequency. A Ba/F3 clone containing a single hygrotk target sequence was electroporated with the promoter-less GFP-lox targeting plasmid either alone or in combination with the Cre recombinase expression plasmid, and three days later GFP expression was assessed by fluorescent flow cytometry. Representative results of one experiment are shown in Figure 3.4. A GFP vector with a functional promoter but lacking lox sites (pCMV-GFP) resulted in -30% GFP positive cells 72 hours after electroporation, demonstrating high level initial 90 transfection efficiency. However, expression from this vector was only transient, with the proportion of GFP positive cells dropping rapidly to -0.1% by day 14 and is consistent with high initial expression due to episomal DNA but with a low integration frequency. As expected, the promoter-less GFP- lox targeting plasmid transfected in the absence of the Cre expression plasmid yielded no GFP expression at any time. In contrast, the same targeting plasmid transfected with the Cre expression plasmid yielded ~1.5% GFP positive cells 72 hours post-electroporation, consistent with Cre recombinase mediated integration. In support of this, as seen in Figure 3.4, the proportion of GFP positive cells remained stable with culture, demonstrating that the GFP targeting plasmid provides an accurate measure of the stable integration frequency. These results also indicate that the efficiency of integration mediated by Cre recombinase was ~10 fold higher as compared to random integration (as assessed by the pCMV-GFP plasmid). 91 non-transfected - positive control (A) ^promoter-less lox- (•) -„ promoter-less lox- (•) FSC-H\FSC-Heighl—> 100 3 0.01 50 100 isb 200 240 FSC-H\FSC-Height—> ~5o" TM f5o 200 250 FSC-H\FSC-HeigM—> FSC 5 10 15 20 25 Number of days post-electroporation 50 100 1?0 200 250 FSC-H\FSC-Heighl—> 30 Figure 3-4 Cre mediated integration using the promoter-less GFP-lox targeting plasmid in a Ba/F3 hematopoietic cell clone. Analysis of GFP expression by fluorescent flow cytometry three days post-electroporation (top) and over a period of 30 days of the following constructs: non-transfected or negative control; positive control using a GFP plasmid with a functional promoter lacking lox sites (pCMV-GFP) (A); promoter-less GFP-lox targeting plasmid only (•); and promoter-less GFP-lox targeting plasmid with the CMV driven Cre expression plasmid (•). . The above results indicated that stable integration had occurred by 72 hours suggesting rapid integration after transfection. To assess this in more detail, the same Ba/F3 clone with the hygrotk target sequence was co-electroporated with the promoter-less GFP-lox and Cre plasmids and GFP positive cells were isolated by FACS at 24, 48, and 72 hours post-electroporation. Sorted cells were then plated back in culture for 92 continued monitoring of the proportion of GFP positive cells. As shown in Figure 3.5, when GFP positive cells were isolated at 24 hours, -60% stably expressed GFP. This suggests that integration has already occurred at this time point, but is not totally completed since GFP expression was less than 100%. As expected from the above experiment, when cells were sorted at 72 hours the proportion of GFP positive cells remained at 100%, consistent with the notion that Cre mediated integration has been completed by this time point. To confirm that these results represent site directed integration and examine time points even earlier than 24 hours, a PCR approach was used. To distinguish between site directed and random or non-integration, primers were designed to a sequence within the GFP gene and the leader sequence in the viral MSCV LTR (Figure 3.1). A 1.2kb band indicative of Cre mediated integration was shown to occur as early as 8 hours post-electroporation and likely is completed by 72 hours (Figure 3.5). Thus, cre mediated integration can occur very rapidly after electroporation. Moreover, these results also demonstrate that GFP expression can be used to isolate targeted cells that have undergone site directed integration. 93 1000 5 0 1H 1 1 1 1 1 1 1 0 3 6 9 12 15 18 21 Number of days post-electroporation 1.2 kb-U T3 03 O d o o o is 8 CO U pre-sort post-sort 1 s 1 5 00 w S 2 00 00 3 o 00 O 4= 00 o 43 00 O IT) Tf OO ^ OO (S o u i : i i j ' s *"* SZ. MM kMf |j§f * 1 pi« mm >«g r—-r •" • Figure 3-5 Cre recombinase mediates rapid integration in a Ba/F3 hematopoietic clone. Shown is a representative PCR from three independent experiments. Top. Analysis of GFP expression of cells that were electroporated with the promoter-less GFP-lox targeting plasmid and Cre expression plasmid, followed by isolation at 24 hours (•), 48 hours (•), 72 hours (O) and unsorted cells (A). Bottom. PCR analysis was performed on cells that received both GFP-lox targeting and Cre expression plasmids at indicated time points post-electroporation. Integration was shown to occur as early as 8 hours, demonstrated by the presence of the 1.2kb band. 94 Five independent Ba/F3 clones with different initial lox P integration sites were then used to determine the site directed integration frequency. Based on the above experiments, each were co-electroporated with the promoter-less GFP-lox targeting plasmid and Cre expression plasmid and analyzed three days later. As shown in Table 3.2, an average of -0.7% was obtained with only minimal variation between clones. This overall frequency will be affected by the proportion of cells successfully co-electroporated with both the GFP targeting and Cre expression plasmids. This frequency was estimated based on co-electroporation of the GFP and the yellow fluorescent protein co-expression using plasmids in which both were driven by the CMV promoter. The co-electroporation frequency was estimated to be -17% in the Ba/F3 cell line, indicating that the real Cre mediated integration frequency might be as high as - 7%. Table 3-2 Average site directed integration frequency in the Ba/F3 hematopoietic cell line Cell line n Observed % of Calculated site stable GFP positive directed integration cells frequency (%) Ba/F3 5 0.67 4.1 (range 0.14-1.1) (range 0.8 - 6.7) Results show the average percentage of stable GFP positive cells measured by FACS analysis after each clone was co-electroporated with the promoter-less GFP-lox and CMV-driven Cre expression plasmids. The calculated site directed integration frequency was determined by the co-electroporation frequency using pCMV-GFP and pCMV-YFP in similar ratios used for the targeting and Cre plasmids. This was determined to be 16.5%. «, number of independent clones tested. 95 The analysis was extended to determine if Cre/lox mediated integration yields predictable expression levels following re-targeting of an expression cassette to selected integration sites that exhibit distinct expression properties. To test this, a retrovirus with the GFP gene flanked by heterospecific lox P sites was used to infect the Ba/F3 hematopoietic cell line. Three clones were identified in which GFP expression levels spanned a 200 fold range and each were confirmed to have a single integration site. Using their mean levels of green fluorescence, they were designated as low (10.4 mean fluorescence intensity (mfi)), medium (386.7 mfi) and high (1741.5 mfi) GFP expressers (Figure 3.6 A). Each of these clones were then co-electroporated with a plasmid containing the human THY-1 cDNA expression cassette flanked by heterospecific lox P sites and the Cre expression plasmid. FACS was used to sort for THY-1 expression three days later, followed by analysis to see if the expression properties were maintained. As shown in Figure 3.6B, the mean levels of THY-1 expression after exchange were predicted by the initial GFP expression yielding: low (45.7 mfi), medium (568.3 mfi), and high (1018.3 mfi) expression. Although different methods were used to analyze the two genes, its is evident that the differential expression was maintained following Cre mediated replacement. For example, the mean fluorescence for the "low" expresser never reached levels of either the "medium or high " expressers after the exchange or vice versa. 96 A. GFP expression • THY-1 expression • Figure 3-6 Predictable expression is achieved following Cre mediated re-targeting. Top. Ba/F3 cells were infected with a retrovirus containing a GFP gene flanked by heterospecific lox sites. Individual clones containing a single copy of the retrovirus were isolated using the FACStar"1" based on their level of GFP expression. Bottom. Cre mediated integraton was performed by co-electroporating each of these clones with the promoter-less lox THY-1 plasmid and Cre expresson plasmid. FACStar+was used to isolate cells that express THY-1 and subsequent FACS analsysis confirms that the expression properties were maintained after re-targeting. Southern blot analysis confirms targeted integration occurred in all three clones. Figure 3.7 is a representative Southern blot using the highest expressing Ba/F3 clone. These results confirm the concept of reusing a characterized genomic site for predictable expression of a gene in hematopoietic cells. In summary, using a hematopoietic model, the Cre/lox recombination system mediated efficient integration of ~ 7%. It occurred very 97 rapidly following electroporation and could be used to achieve predictable levels of transgene expression. Before integration: After integration: claI CM \f\yLTR -Ll - GFP-L21 LTR\f\y \f\y LTR -Ll - THY-L21 \XK\f\y genomic DNA | 1 genomic DNA | 1 EcoRV 12 kb EcoRV EcoRV 23 kb EcoRV Before After Single —» Integration 2.2 kb Single —1 Integration 2.3 kb—*\ Before > -2 8 After > 3 8 GFP THY-1 Figure 3-7 Representative Southern blot analysis using the highest expressing Ba/F3 clone. Digestion with EcoRV, which cuts once within each LTR, generates a 2.2kb band when probing with GFP and a 2.3kb band when probing with THY-1. Confirmation that the Ba/F3 clone contains a single proviral integration site is achieved by digestion with Clal, which cuts only once in the retroviral vector. 98 3.2.3 Site directed integration in ES and ES derived hematopoietic cells As a means to extend this to more primitive cells, ES cells were chosen as a model. These cells are pluripotent and also have the potential to give rise to hematopoietic cells in vitro. As an initial step, the frequency of site directed integration in undifferentiated ES cells was first assessed. Similar experiments to those described in the Ba/F3 hematopoietic cell line were performed on cells containing a single lox flanked hygrotk target sequence in three individual ES clones. Each of the clones were co-electroporated with the promoter-less GFP-lox plasmid and the Cre expression plasmid and the site directed integration frequencies were determined. Figure 3.8 represents the FACS analysis performed on these clones three days later. 99 non-transfected 50 ioo lib 26b 250 FSC-H\FSC-Haight—> FSC positive control 38% d 50 100 150 21)0 250 FSC-H\FSC-Height—> • clone C6 clone C4 clone B8 . GFP-lox plasmid +Cre -GFP-lox plasmid +Cre -GFP-lox plasmid +Cre FSC-H\FSC-Height—> FSC-H\FSC-Height—> FSC p 5%: 3.5% 50 i bo 1So' 2O0 250 FSC-H\FSC-Height—> Figure 3-8 Site directed integration occurs in undifferentiated ES cell clones. Flow cytometric analysis for GFP expression three days post-electroporation in three ES clones (C6, C4, B8) harboring the initial hygrotk provirus. As summarized in Table 3.3, the average site directed integration frequency was ~ 2.3%. The co-electroporation frequency was determined to be -16% in the ES cell line, which indicates that of the cells that received both the Cre and integrating vector, a high frequency of integration on average of - 15% was obtained. 100 Table 3-3 Average site directed integration frequency in the ES cell line Cell line n Observed % of Calculated site stable GFP positive directed integration cells frequency (%) ES(CCE) 3 2.3 14.6 (range 1.2 - 3.5) (range 7.8 - 22.2) Results show the average percentage of stable GFP positive cells measured by FACS analysis after each clone was co-electroporated with the promoter-less GFP-lox and CMV-driven Cre expression plasmids. The calculated site directed integration frequency was determined by the co-electroporation frequency using pCMV-GFP and pCMV-YFP in similar ratios used for the targeting and Cre plasmids. This was determined to be 15.8%. n, number of independent clones tested. To see if Cre/lox mediated integration could be achieved in primary hematopoietic cells, this approach was tested in hematopoietic progenitors. The in vitro differentiation of ES cells has been well characterized for their hematopoietic development. The kinetics and types of hematopoietic cells formed during this process have been shown to accurately recapitulate what occurs in vivo and thus serve as a reliable system to test the feasibility of this approach (Keller et al., 1999). The approach used to obtain hematopoietic cells from ES cells was originally described by Keller et al, and involves differentiation of ES cells in methylcellulose to allow the formation of embryoid bodies (EBs). After dissociating EBs, cells can either be plated in methylcellulose for the development of hematopoietic progenitors or in liquid culture for the generation of more mature hematopoietic cells (Keller et al., 1993). The experiment to test for Cre/lox mediated integration in hematopoietic progenitors is outlined in Figure 3.9. Electroporation was used to introduce a tk-neo (thymidine kinase promoter -neomycin resistant gene) cassette flanked by lox P sites and 101 expressed from the human polypeptide chain elongation factor 1 alpha (EFl-rx) promoter, followed by selection with G418 for the generation of neomycin resistant clones. The polyclonal G418 resistant ES cells were differentiated to form EBs, disrupted and put into liquid culture conditions (which include murine Stem Cell Factor, murine Interleukin-3, human Interlekin-6, and human Erythropoietin) to specifically expand hematopoietic cells for electroporation. Previous work in our laboratory have shown that after six days in these culture conditions, >50% of the cells are hematopoietic as evidenced by the presence of hematopoietic specific cell surface antigens (ie. Ly5.1, ckit+, Macl+, Grl+). After six days of expansion, electroporation was again used to introduce a promoter-less GFP targeting vector containing compatible lox sites (Ll and 2L) either alone or in combination with a Cre expression plasmid driven by the chicken p actin promoter (CAG). Three days later, cells were analyzed by FACS to detect GFP expression. As expected, cells electroporated with the GFP targeting vector alone were essentially negative for GFP expression (-.08%). In contrast, a proportion (-0.24%) of the cells electroporated with both the GFP and Cre expression vectors demonstrated GFP expression. 102 undifferentiated ES cells differentiate eLFa-Ll-tk- neoR-2L -> <-1.2 kb embryoid body formation expansion i mIL-3, hIL-6, mSF, Epo o0o. hematopoietic cells ^ expansio  for 6 days in ° °°o ° electroporation with targeting plasmid +/-Cre plasmid sort GFP+ cells methylcellulose assay for hematopoietic progenitors expand cells for FACS analysis eLFa - Ll- GFP - 2L 780 bp Figure 3-9 Schematic representation of the strategy used to test for Cre mediated integration in hematopoietic progenitors. Electroporation of the tk-neo cassette was used to introduce the initial lox P target sites into undifferentiated ES cells. Hematopoietic cells were generated following differentiation and expansion of these cells from disrupted day 10 embryoid bodies. Cells were then electroporated with appropriate DNA, sorted for GFP expression and plated in both methylcellulose medium and expanded for FACS analysis. To confirm that GFP expression was due to stable integration and occurred in primitive hematopoietic cells, GFP positive cells were isolated and plated in either liquid culture for expansion or in a clonogenic assay, respectively. Based on the kinetic experiments performed in the Ba/F3 cell line, GFP positive cells were sorted using the 103 FACStar+ three days later. Figure 3.10 represents the FACS profiles obtained after expansion of these sorted cells. There was no GFP expression from cells that had been transfected with the GFP targeting vector alone, whereas a frequency of -97% was obtained in cells that received both plasmids, confirming that GFP expression is due to stable integration. This reinforces the previous results obtained with the Ba/F3 cell line, demonstrating that integration has been completed by 72 hours and this can be used to isolate or enrich for correctly targeted cells. These results were further confirmed in a second independent experiment (Table 3.4). From this, the frequency of site directed integration was determined to occur with a mean and standard deviation of-0.18% ± .022. As shown in Figure 3.10, these cells were also hematopoietic as evidenced by staining with an antibody against the IgE Fc receptor, which is indicative of mast cells. 104 A. promoter-less GFP-lox promoter-less GFP-lox non-transfected plasmid -Cre .a plasmid +Cre 1023 0 0% FSC 1023 0 97% B. t <u o a. o W -• so non-transfected ft promoter-less GFP-lox promoter-less GFP-lox plasmid -Cre 27% 10° GFP • Figure 3-10 FACS analysis on sorted ES derived hematopoietic cells to confirm stable integration and cells are of hematopoietic origin. ES derived hematopoietic cells were electroporated with the promoter-less lox-GFP targeting plasmid with or without the Cre expression plasmid and isolated using the FACStar+. A. FACS profiles on these expanded sorted cells from one experiment demonstrtaing stable integration. B. FACS profiles on expanded sorted cells from the second experiment using the IgE receptor antibody to confirm that they are hematopoietic. Table 3-4 Site directed integration frequencies in hematopoietic progenitors Experiment Constructs Percent of GFP Percent of GFP Estimated number on day of sort after expansion Frequency (%) (%) (%) 1 GFP alone 0.08 0 GFP and Cre 0.2 97 0.19 2 GFP alone 0.10 0 GFP and Cre 0.23 70 0.16 0.18± .022 105 Analysis of the colonies formed in methylcellulose demonstrated the presence of hematopoietic progenitors. More importantly, the majority (>80%) of these clonogenic cells were GFP positive as assessed under the fluorescent microscope. Representative hematopoietic progenitors that were designated as "positive" (express GFP) and "negative" (no GFP expression) are shown in Figure 3.11. Figure 3-11 Representative hematopoietic colonies from sorted GFP positive cells that received the promoter-less lox GFP targeting plasmid and Cre expression plasmid. Left, positive GFP colony (top, dark field; bottom, fluorescent microscope). Right, negative colony (top, dark field; bottom, fluorescent microscope). 106 To confirm that GFP expression in the hematopoietic progenitors was the result of Cre mediated integration, individual colonies were plucked and PCR was performed. To detect site directed integration, primers were designed to a sequence within the GFP gene and the EFl-a. promoter (as shown in Figure 3.9). A 780bp band is indicative of Cre mediated integration and was shown to occur in all of the positive clones (Figure 3.12). Moreover, none of the negative clones exhibited this band except for clone #2, which appeared as a mixed colony under the fluorescent microscope, exhibiting some positive and negative GFP cells. In the case of the negative clones, absence of a band may also be the result of the lack of cells used in the PCR. For this reason, PCR was repeated using a primer pair to detect the original construct. Primers were made within the neomycin resistant gene and was used in combination with the above primer for the EFl-a promoter to produce a 1.2kb band (as designated in Figure 3.9). As shown in Figure 3.12, all of the negative clones exhibited the 1.2kb band. Therefore, PCR analysis confirms that Cre mediated integration occurs in hematopoietic progenitors. 107 Positive Colonies Negative o I I I <N X [M - + 1 2 34 5 6 7 8 910 1 2 3 1 Figure 3-12 PCR analysis confirms site directed integration in hematopoietic progenitor colonies. Shown is a representative PCR from two independent experiments. PCR was performed on each colony with a primer set to detect integration (780 bp, see arrows in figure 3.9) and a primer set to detect the original construct (1.2kb, see arrows in figure 3.9). M, lkb ladder; -, non-transfected control; +, positive control for each primer set; individual positive colonies (1-10); individual negative colonies (1-3). 108 3.3 Discussion A modified gene transfer strategy based on Cre/lox mediated site directed integration has been characterized in these studies. Initial experiments in the NTH 3T3 cell line, the K562 human erythroleukemia cell line, the murine Ba/F3 hematopoietic cell line and subsequently in primitive pluripotent ES cells demonstrated the feasibility of this two step process using the neomycin resistant gene as a reporter. Use of a GFP reporter further illustrated that this process is rapid, efficient, and can be exploited to achieve predictable expression of a transgene. Furthermore, using the ES in vitro differentiation model, the feasibility of achieving Cre/lox integration in primary hematopoietic progenitors was demonstrated. Initial feasibility studies using the neomycin resistant gene demonstrated that site directed integration requires both Cre recombinase and the presence of lox P target sites on the incoming plasmid. Southern blot analysis of a large number of individual clones further illustrated that integration occurs specifically at the initial target site, without additional integrants elsewhere in the genome. This provides an advantage over other site directed integration strategies, such as homologous recombination, in which random integration occurs -1000 fold higher than targeted integration (Vasquez et al., 2001). This approach also appears to be applicable to many cell types as integration was shown to occur in both non-hematopoietic and hematopoietic (murine and human) cell lines, primitive pluripotent ES cells and even primary hematopoietic progenitors. Using the GFP as a reporter, the site directed frequency was determined to be as high as 7% in one of the Ba/F3 clones and 22% in an ES clone. In this work, heterospecific lox P sites with 109 a single base pair change in the spacer region were used,.as previous results have demonstrated 20-100 fold increase in the integration efficiency compared to the wild type lox P sites. Modifications in the portion of the lox sites in which Cre recombinase binds (in the inverted repeat), have also been constructed but appear to confer similar efficiencies to that obtained in these studies. For example, Yamamura and colleagues, reported an integration efficiency of 2-16% in several independent ES clones using these lox sites (Araki et al., 1997). On the other hand, Feng et al, have recently reported a significant increase in Cre/lox mediated integration by using inverted lox P target sites. For example, they show an efficiency of 100% in three independent clones from the mouse erythroleukemia cell line and as high as 50% in one clone from undifferentiated ES cells (Feng et al., 1999). Thus, retroviral gene transfer combined with the Cre/lox system permits efficient integration of exogenous DNA into the genome and these frequencies might be further improved by re-engineering the lox target sites. There has been only one previous report on the kinetics of site directed integration using the Cre/lox system in a mouse cell line (Sauer and Henderson, 1990). Sauer et al, showed that by applying selection to cells at various time points after electroporation of the targeting vector and the Cre recombinase, site directed integration could occur as early as 24 hours. Using the GFP reporter gene, the kinetics of integration were further examined in hematopoietic cells. FACS analysis demonstrated that site directed integration occurred as early as 24 hours and was completed by 72 hours post-electroporation, confirming the results by Sauer et al. PCR was used to further show that integration occurs even earlier than previously reported, at 8 hours post-electroporation. The fact that integration occurs rapidly after electroporation opens the possibilities to 110 either pre-enrich for cells or isolate cells that have undergone site directed integration at early times post gene transfer. This could prove to be useful in applications to primary hematopoietic cells where cells can be isolated at the earliest time point to avoid excessive manipulation. At the time this study was initiated, electroporation conditions were not optimal for murine bone marrow cells, and initial attempts demonstrated very low gene transfer (<1%) to hematopoietic progenitors. For this reason, I turned to the ES cell in vitro differentiation model to test the feasibility of this approach in more primitive hematopoietic cells. By exploiting this system, I demonstrated and provided the first evidence that Cre mediated integration can be achieved in hematopoietic progenitor cells that form colonies in methylcellulose culture. Recently, Hatada et al, reported on homologous recombination in primary hematopoietic progenitor cells. Here, homologous recombination was used to correct a defective HPRT gene at a frequency of 4.4 ± 3.3 x 10"5 in bone marrow cells and 2.3 ± 0.4 x 10"5 in ES cells (Hatada et al., 2000). In this work, a significantly higher frequency (-0.18% + .022) of site directed integration using the Cre/lox system was demonstrated. This is clearly a usable frequency and provides encouragement that site directed integration in HSCs may be an achievable goal. However, it is apparent that some form of enrichment for stem cells would be required. For example, functional assays have estimated that there is -1 HSC per 104 bone marrow cells, and assuming that HSCs are corrected at the same frequency as hematopoietic progenitors, (one event per 5x10 cells) an average of 5 x 10 bone marrow cells would be required to detect site directed integration in one HSC. Strategies to expand HSCs in vitro and/or those aimed at amplifying only "targeted" HSCs after site directed integration 111 may prove beneficial towards this goal, and will be discussed in more detail in Chapter 5. The effectiveness of this procedure could also be further improved by using re-engineered lox target sites (as mentioned above) and improved electroporation conditions designed specifically to target HSCs. For example, Wu et al, cultured human HSCs in a growth factor cocktail (to stimulate their proliferation) prior to electroporation and increased the number of cells in cycle from 2 - 28%. In vitro assays further show -56% GFP positive hematopoietic progenitors per 1 x 105 electroporated CD34+ cells, and -16% GFP positive LTC-IC cells per 1 x 105 electroporated CD34+ cells, that are capable of generating hematopoietic cells for up to 5 weeks in vitro (Wu et al., 2001a; Wu et al, 2001b). A non-viral delivery approach was used to introduce the initial lox target sites into ES cells because of the ease in genetically manipulating these cells. However, to apply this approach to HSCs, it is apparent that retroviral gene transfer would be needed to introduce the lox target sites, since the efficiency of using non-viral delivery is extremely low. Furthermore, recent studies using lentiviral vectors may prove to be more practical than retroviral vectors as they have the ability to transduce non-dividing cells (which include HSCs). The transduction efficiency is similar to retroviral vectors and some studies also report a decreased propensity for silencing in hematopoietic cells (Vigna and Naldini, 2000). The site of integration can have a major influence on the expression characteristics of a transgene (Wallace et al., 2000). Investigators have used this to their advantage by using site directed integration strategies to achieve predictable and reproducible expression of a transgene. For example, one report used homologous recombination to place lox P target sites into the P-casein locus, which is highly 112 expressed in milk. They subsequently used the Cre/lox system to insert the luciferase reporter gene so it was under the control of the P-casein promoter and demonstrated that its expression/tissue specificity was similar to the endogenous gene (Kolb et al., 1999). Bode et al, also demonstrated using the retrovirus and FLP/FRT system, that the observed expression of a re-targeted luciferase gene followed the expression level of P-galactasidase in the initial provirus in a hamster kidney cell line (Schubeler et al., 1998). The studies reported here extend this approach to a hematopoietic cell line where the initial lox P target sites were introduced by an efficient retroviral method. Here, GFP expression was correlated to the expression of another readily detected marker gene, THY-1 cell surface antigen, and confirm signature levels of relative expression that is dependent on the initial lox P integration site. These results further confirm the critical importance that the site of integration has on the level of transgene expression. This strategy should prove useful in hematopoietic cells for gene function studies since it allows the expression of a gene at a desired level, and for more accurate gene regulation studies where different regulatory elements of a gene can be re-targeted to the same site, without the confounding influence of the surrounding chromatin. 113 CHAPTER 4 EVALUATION OF GENE REGULATORY MECHANISMS IN HEMATOPOIETIC CELLS USING THE IN VITRO DIFFRENTIATION OF ES CELLS COMBINED WITH THE CRE/LOX SYSTEM 4.1 Introduction Hematopoiesis is the process by which blood cells of multiple and distinct lineages are generated from multi-potential stem cells. This complex process is regulated through the expression of specific genes brought about through the co-ordinated interaction of cis-regulatory elements and trans-acting factors (reviewed in Metcalf, 2001). Hematopoietic cell lines have been widely used and are instrumental in deciphering which regulatory elements are important for gene expression. Although extremely convenient, hematopoietic cell lines do suffer from several limitations. For example, the majority of cell lines only recapitulate a single state of ontogeny during hematopoiesis, making it difficult to study gene regulation in multiple lineages or within a specific lineage as it progresses through development (Chiba et al., 1991; Tani et al, 1996). There have also been reports in which the same construct analyzed in a cell line and in transgenic mice gave conflicting results (Lemarchandel et al., 1993; Prandini et al., 1992). For example, a number of human p globin constructs that were not subject to position effects in transgenic mice, were sensitive in the mouse erythroleukemia (MEL) cell line resulting in expression levels that were highly variable (Skarpidi et al., 1998). On the other hand, the transgenic mouse model has proven particularly useful to explore developmental and 114 lineage specific regulatory elements in vivo (Pasceri et al, 1998). While unquestionably powerful, the transgenic model is labour intensive and interpretation can be complicated by integration of test vectors at various, unknown sites and usually at more than one copy in the genome. An attractive alternative model system for studying gene regulatory mechanisms is the murine embryonic stem (ES) cell system. ES cells are pluripotent cells that are amenable to genetic manipulation and can differentiate into a variety of cell types, including hematopoietic, in a process that accurately reflects in vivo hematopoietic development (Keller et al., 1993). It is well established that confounding variables such as copy number and site of integration play an important role in gene expression (Whitelaw, 2001). These problems can now be controlled by using site specific recombinases, such as the Cre/lox recombination system, which allows a series of systematic changes to be performed at a specific locus. As discussed earlier, this approach permits a more accurate analysis of cis-acting sequences of a gene since constructs containing different regulatory elements can be analyzed at the same integration site. Several reports have used this Cre/lox approach for studying gene regulation in hematopoietic cell lines (Bouhassira et al., 1997; Molete et al, 2001). Bouhassira et al, have demonstrated the benefits of using this approach for studying the regulatory elements of the human p globin gene in the MEL cell line (Bouhassira et al., 1997). This approach has also been used to obtain predictable and reproducible expression of transgenes in undifferentiated ES cells and in several mammalian cell lines (Feng et al, 1999; Kolb et al, 1999). Moreover, there have been reports of other recombination systems being similarly used for evaluating the design of retroviral vectors (Verhoeyen et al., 2001). Since these reports were promising, the 115 Cre/lox recombination system combined with the in vitro differentiation of murine ES cells was investigated in this Chapter for studying gene regulation in hematopoietic cells. An intriguing study performed by Grosveld et al, demonstrated the use of ES in vitro differentiation as a dynamic model system to study human P globin gene switching (Lindenbaum and Grosveld, 1990). The attractiveness of this approach led to implementing the ES in vitro differentiation system described in this Chapter to assess human p globin gene expression vectors that are of interest for gene therapy. The hemoglobinopathies, which are characterised by either a deficiency or structural alteration of globin proteins, were the first genetic disorders considered for gene therapy. However, the elucidation of a vector that is capable of conferring high levels and sustained expression in mature erythroid cells has been problematic (reviewed in Weatherall, 2001). Improved model systems to systematically explore and identify optimal regulatory elements and vector design would therefore be beneficial. In this study, the feasibility of using Cre/lox recombination combined with ES in vitro differentiation was assessed by using human p globin gene constructs containing a range of regulatory elements from a minimal promoter to an extensive locus control region (LCR), and evaluating their expression levels when integrated at random or at specific integration sites by the Cre/lox system. 116 4.2 Results 4.2.1 Characterization of erythroid development and globin gene expression in differentiating ES cells An evaluation of erythroid development and globin gene expression in differentiating ES cells was first performed. ES cells (CCE line) were differentiated based on the two step procedure developed by Keller et al. (Keller et al, 1993). In the first step, ES cells are suspended in a methylcellulose-based medium which promotes their differentiation into a complex structure called an embryoid body (EB), that includes hematotopoietic cells. In the second step, EBs are disrupted into single cells that can be analyzed either directly, plated in methycellulose for hematopoietic progenitor development or plated in liquid culture for the induction of more mature cells. Initial experiments ascertained the kinetics of expression of the endogenous embryonic (pHl) and adult (Pmajor) globin genes during in vitro differentiation by RT-PCR analysis of RNA from EBs at various times post-differentiation. As shown in Figure 4.1, ES cells at day 0 showed no evidence of any globin gene expression (Lane 1). EBs harvested at day 6 of differentiation showed expression of the embryonic PHI gene, with no or minimal Pmajor expression evident (Lane 2). Expression of Pmajor increased dramatically by day 10, and persisted through to day 14 (Lanes 3-5). The embryonic globin gene was also found to be continuously expressed during these later stages of differentiation. 117 coil c 53 V, ° <N t ^ o ^ », g V© _< FH _ g,g »J W T3 -o ffl g h E S (embryonic) Murine pHl 265 bp (adult) Murine (3 major 578 bp HPRT 249 bp Figure 4-1 Analysis of endogenous p globin gene expression during ES in vitro differentiation. Shown is a representative RT-PCR from 5 independent experiments. RNA was isolated from ES cells at day 0 (Lane 1) or from embryoid bodies (EBs) harvested at day 6 (Lane 2), 10 (Lane 3), 12 (Lane 4) and day 14 (Lane 5) of differentiation. RT-PCR was performed using a panel of primer pairs specific for embryonic pHl, adult Pmajor and an HPRT primer as a control for cDNA amplification. The sizes of the specific products in basepairs (bp) are indicated. A positive control for the adult pmajor included RNA isolated from the spleen of a C57/B6 mouse (Lane 6). A negative control for the adult Pmajor included RNA isolated from the spleen of a transgenic mouse that expresses only human P globin (Lane 7). Analysis of erythroid progenitors during in vitro differentiation demonstrated that this expression profile correlated with the initial appearance of predominantly primitive type erythroid progenitors (giving rise to small clusters of nucleated erythroblasts, see figure 4.2i left), followed by an increased number of definitive erythroid progenitors (giving rise to larger multi-clustered erythroblasts, see figure 4.2ii left). Moreover, the frequency of erythroid progenitors during differentiation was determined to be highest at day 10 118 (corresponds to -62%), followed by a slight decline by day 12 (-48%), and a more dramatic decrease by day 14 (-22%) (Figure 4.2 right). k 1 1001 GM Mix Erythroid Total Types of Hematpoietic Progenitors Figure 4-2 Characterization of erythroid progenitors during ES in vitro differentiation. Shown are representatve results from 3 independent experiments. Morphology of erythroid progenitors formed during in vitro differentiation, i; primitive erythroid colony present at day 6, ii; definitive erythroid colony present at days 10-14 (Left). Graph representing the various hematopoietic progenitors formed at different stages during in vitro differentiation (Right), with the frequency of erythroid progenitors (number of erythroid progenitors/total number of progenitors) shown in the inset. Based on the above experiments, this system was then evaluated at day 10 of differentiation for its efficacy in generating mature erythroid cells, as assessed by the expression of the erythroid specific lineage marker, Terl 19. Approximately, 300 ES cells were plated in primary differentiation media for the formation of EBs, followed by the 119 disruption of EBs into single cells at day 10 of differentiation. These cells were then subsequently plated in liquid culture conditions supplemented with murine stem cell factor (mSF), murine IL-3, and human IL-6, to support general expansion of hematopoietic cells, and also plated in liquid culture conditions to promote selective growth/expansion of erythroid cells (Panzenbock et al., 1998). The general conditions gave an ~ 9 fold expansion in cell number after six days of expansion in liquid culture, and a high percentage of erythroid cells as determined by FACS analysis for Terl 19. As shown in Table 4.1 and Figure 4.3, -21% were Terl 19 positive which corresponds to 6.2 x 105 erythroid cells. The specialized conditions resulted in approximately the same fold expansion in cell number, and a slight increase of -2 fold in the number of erythroid cells (59% Terl 19) (Figure 4.3). Table 4-1 Comparison of erythroid liquid culture conditions Experiment Number of ES Cell number Fold Percentage Number of Number cells/dish from after 6 days of expansion of Terl 19 erythroid disrupted day expansion* (%) cells 10 EBs 1 8.6 x 105 A 1.7 xlO7 8.5 22.1 7.3 x 105 B 2.0 x 107 9.9 59.0 1.2 x 106 2 7x 105 A 1.6 xlO7 8.0 21.1 5.0 x 105 B 9.8 x 106 9.8 60.0 1.1 x 106 * ES cells were expanded in two different conditions. A; supplemented with mSF, mIL-3 and hIL-6, B; and conditions previously described in Panzenbock et al, 1998. 120 Condition A Condition B 59% FCS FCS Figure 4-3 FACS profiles using the Terll9 marker following expansion of ES cells in general hematopoietic liquid culture conditions (A) and specialized erythroid liquid culture conditons (B). 4.2.2 Comparison of expression levels of human p globin constructs integrated either randomly or at integration sites mediated by the Cre/lox system The expression characteristics of various human P globin constructs were next examined in this system. As summarized in Figure 4.4, three constructs were chosen that were expected to confer very poor to high p globin expression levels, as indicated by the types of regulatory elements included within each one. The simplest or basic p globin construct included ~0.8 kb 5'flanking minimal promoter region, no LCR sequence, and the entire human p globin gene. The HS3 p globin construct included in addition to this, - 0.85 kb DNA fragment of HS3 (consists of 230 bp HS3 core element and -600 bp flanking region) from the LCR. The most complex construct, HS1-4 p globin, included an extended 1.5 kb promoter region and - 4.0 kb DNA fragment which consists of - 1 kb each of all 4 HSs (1-4) of the LCR. 121 Parent vector Ll neor multiple cloning site L2 Basic pLl -i . . r^~i p gi0bin [O—I neo^J^j |fi globin gene| |j>J u > s O N H 'in 4.2 00 w ° i 8 0.8 OH U HJ 0> N p globin [> rWHSAlJ BUHP globin gene| \j>j 5.0 0.8 HS3 (0.85) HS1-4 L1 2 p globin [> l_M^[JA]l-|HSfflHS2j{H^ globin gene| ^D>] 8.8. 1.5 HS 1-4 (4.0) Figure 4-4 Schematic diagram of the human P globin constructs. For random integration studies, each construct (with no lox sites) was linearized into undifferentiated ES cells. For Cre mediated integration each vector was cloned into the parent vector to result in constructs that were flanked by heterospecific lox P sites (Ll and L2), and contained an upstream promoter-less neomycin resistant gene to select for cells that underwent site directed integration. The HS3 and HS 1-4 P globin containing constructs are described in detail in Pasceri et al, 1998. neoR, neomycin resistant gene; pA, poly A addition signal; LCR, locus control region; HS, hypersensitive site. Each of these constructs was linearized and co-transfected with a linearized plasmid containing the neomycin resistant gene in a 10:1 molar ratio into undifferentiated ES cells. After G418 selection, at least 20 clones for each of the constructs were pooled and Southern blot analysis was performed to confirm integration (Figure 4.5). As shown in Figure 4.5, the ES cell pools with the HS3 and HS 1-4 containing P globin constructs were polyclonal, as demonstrated by the large number (>5) of distinct integration bands. In contrast, it appears that a clone harbouring a single integration site is predominant in the ES cell pool containing the basic p globin construct. 122 EcoRV P globin constructs intron 2 EcoRV exonl exon2t genomic DNA exon3 BamHI genomic DNA BamHI 1.53 kb BamHI 12.0 kb. 9.0 kb 6.0 kb bands represent random integrations (polyclonal) minimum size for clonal integration Figure 4-5 Southern blot analysis to confirm random integration of the human p globin constructs lacking heterospecific lox P target sites. Representative Southern blot from two independent experiments. Undifferentiated ES cells were co-transfected with human p globin constructs linearized with EcoRV and a linearized plasmid containing the neomycin resistant gene in a 10:1 molar ratio, followed by selection with G418. Clonal integration was determined by digestion with BamHI, which cuts once in the construct and at various sites in the genome, followed by probing with intron 2 of the human P globin gene. The minimum size for clonal integration (if BamHI site in the genome occurred immediately after the EcoR V site) was determined to be 1.5 kb. Therefore, bands >1.5kb represent true random integrations. 123 Each of these ES cell pools was differentiated and RT-PCR analysis performed on the RNA isolated from EBs at various times post-differentiation. A representative RT-PCR using the HS1-4 containing p globin vector is shown in Figure 4.6. RT-PCR analysis revealed that for all of the constructs, human P globin expression was observed at day 10, similar to the mouse Pmajor and interestingly, later than the mouse embryonic PHI. (embryonic) Murine pHl 265 bp (adult) Murine P major 578 bp (adult) Human p globin 349 bp HPRT 249 bp Figure 4-6 Kinetics of human p globin gene expression during ES /// vitro differentiation. RT-PCR was performed on EBs derived from the ES cell pools on all three constructs in 5 independent experiments. Shown is a representative RT-PCR using the HS1-4 containing p globin construct. RNA was isolated from ES cells at day 0 (Lane 1), or from embryoid bodies at day 6 (Lane 2), 10 (Lane 3), 12 (Lane 4), and day 14 (Lane 5) of differentiation. RT-PCR was performed using a panel of primers specific for embryonic PHI, adult pmajor, human p and an HPRT primer pair as a control for cDNA amplification. The size of the specific products in basepairs (bp) are indicated. A positive control for the adult pmajor included RNA isolated from the spleen of a C57/B6 mouse (Lane 6). Also, a positive control for the human p globin included RNA isolated from the spleen of a transgenic mouse that expresses only human p globin (Lane 7). 124 The levels of human P globin expression relative to mouse Pmajor were next examined for each of the constructs. To perform this, ES liquid culture conditions were exploited to obtain an enriched hematopoietic cell population with significant erythroid component and in sufficient numbers for RNASE protection assay (RPA). The minimal promoter construct and the HS3 containing p globin construct yielded barely detectable levels of human P mRNA relative to mouse pmaj0r as assessed by RPA (<0.1%, see figure 4.7) or FACS analysis to detect the human p globin protein (Figure 4.8). The HS1-4 containing p globin construct, in contrast, yielded detectable levels of human P globin mRNA relative to mouse pmajor(9.5%, see figure 4.7), and detectable levels of human p globin protein (-5% of the cells express human P globin) (Figure 4.8). In summary, it appears that the HS1-4 containing p globin vector performed better than the other two constructs. However, comparisons of the different regulatory elements within each construct could not accurately be assessed, since two of the vectors were below the detection level. 125 % human p globin/ 0.009 0.02 0.07 9.5 mouse P major Figure 4-7 RNASE protection assay (RPA) to quantify the level of human p globin mRNA relative to mouse Pmajor for human p globin constructs that were integrated at random. RNA was isolated after in vitro differentiation of ES cells until day 10 and six days of expansion in general hematopoietic liquid culture conditions. Each sample was tested with a probe for the human p globin gene (350bp protected fragment, HP) and a probe for the mouse Pmajor (145bp protected fragment). The percentage of human mRNA relative to mouse mRNA for each sample is shown on the bottom of the figure. The rank order of these constructs was reproduced in a second independent experiment. 126 JO O 00 X jo o co cS JO 00 O CN 2 4—• co C O o .01 IOL iD O CN mm constn .: :.r$ .01 IOI .0 •01 w u u 0s-v.'.i .S3 iOI tOt ,01 tfl + jo o CO m •am-2 CO ' ' o o • JO o CO JO a u U (01 Ol' nOl true CS CO C o o l|| ,0V [01 iOI .01 oO. mqopl cj t/5 H a 5? o o [01 tOl .01 U fa a cu 1-'33 •M o u a 15 _© "3D CQ. a es E a _© CJ CU CU T3 s *a DX) eS I-co a o S CU •a cOL IOI .01 .01 •4 uiqoTg^ CO >, a a CS u *c CU E o CJ o 00 I Tf CU L, s 6X1 c <u C <u 1-1 U s-cu CO U CO (U C o '•5 cu .2 xt 2 CD M CL 00 H c . <u s s o o I» C '53 >> ^ JO S •o °-<u c vs S co o w coo en. s o u w u u c o u T3 <U o C*H co J3 c3 •2 § U J? co -^J 1^ ^ ^ <U >< J=l c o c co O CO C O o CU I -J-» s co U CO CJ OH co c2 CO C O » C o o <u l-l 3 i-~ "S •<§ 1 •s a as 2 n. 6 M CU .S O •a | — s 2 JS cj <u CO CCJ s eg 0) C jo o CU l-l CU £ CU co CO p 00 O e ^ _ cu ^ •S H O a u ° o > cc] co O T) 3 ft Cre/lox mediated exchange was then used to compare these vectors at the same integration site. Retroviral gene transfer was used to introduce heterospecific lox P target sites (Ll and L2) into undifferentiated ES cells, followed by the isolation of individual clones. Approximately 22/56 of these contained a single copy of the provirus, two of which were chosen at random for the following studies (referred to as A2 and A8) (Figure 4.9). LTR—Ll-genomic DNA Xbal hygro-tk Hindlll -L L2 LTR/\/)/^ I genomic DNA 3.5 kb Xbal Xbal Hindlll u a g o £ CJ u oo < c o U w g U £ u u oo < u c o U 3.5 kb • Figure 4-9 Southern blot analysis of the two randomly chosen ES clones following retroviral gene transfer of the heterospecific lox P target sites. To confirm that the provirus was intact, DNA was digested with Xbal, which cuts once within each LTR, and then probed with the hygromycin gene (designated as a black bar) to give a 3.5kb band. Digestion with Hindlll, which cuts only once within the vector, followed by probing with hygromycin confirms a single integraton site, hygrotk, hygromycin-thymidine kinase fusion gene; L2, wild type lox P site; Ll, lox P site with a single base pair mutation in the spacer region. 128 Cre mediated integration was then performed with each of the human (3 globin vectors at sites A2 and A8, followed by Southern blot analysis on the resulting G418 resistant clones. Site directed integration was verified by achieving the correct DNA fragment sizes after digestion with Kpnl or KpnI/BamHI, followed by probing with the neomycin resistant gene and intron 2 of the human P globin gene (Figure 4.10 and 4.11). This confirms that site directed integration mediated by the Cre/lox system is feasible with human p globin constructs that have previously been reported to be unstable in retroviral vectors (Leboulch et al, 1994), and the size of which go beyond the current limitations for retroviral vectors (Pasceri et al., 1998). 129 ^robe^ probe Basic p globin LTR - Ll - neoR- f^-^^—. L2 - LTR KpnI/BamHI I "I 1 II 3.5 kb I Kpnl neomycin SaJ|// 8lobin Kpnl probe Pr°be HS3 P globin LTR -Ll - neoR- |HS3]-B--H—H - L2 - LTR KPnI I 2.6 kb || llkb [ Kpnl ne°myCm 8lobin AJ,«/ HS1-4 p globin pi^e LTR -Ll - neoR 1 4.7 kb I Kpnl neomycin/globin BamHI | : 11 0 kh HS3 HS4 L2 - LTR KpnI/BamHI Kpnl neomycin/globin Kpnl Kpnl Figure 4-10 Schematic diagram demonstrating the expected band sizes following Cre/lox mediated integration of the human P globin constructs. Digestion with Kpnl and/or KpnI/BamHI followed by probing with the neomycin resistant gene and intron 2 of the human P globin gene (designated by black boxes above the constructs) confers distinctive band sizes indicative of Cre mediated integration. 130 Clone A2 Clone A8 Clone A2 Clone A8 ll.Okb— 4.7 kb — 3.1 kb 2.6 kb —I 1.6 kb probed with neomycin ll.Okb 1.6 kb probed with globin Figure 4-11 Southern blot anlaysis confirms Cre/lox integration of the human p globin constructs at sites A2 and A8. DNA was isolated and digested with Kpnl and/or KpnI/BamHI, followed by probing with the neomycin resistant gene (left) and stripping the blot and re-probing with intron 2 of the human P globin gene (right). All of the band sizes expected from Cre/lox mediated integration (detailed in Figure 4.10) occurred, confirming site directed integration of all three human p globin constructs. The ES cells containing each of the human p globin constructs at sites A2 and A8 were differentiated until day 10, expanded in liquid culture for six days, whereupon RNA was isolated from erythroid cells. Figure 4.12 shows the RPA for the human P globin constructs obtained at sites A2 and A8. 131 CD CD x> X O 0 »i cu u a •2, 3 £ "5b ca ca CD C 1 I A2 site A8 site bin d obin bin c obin lob obin elo O obin &o ca 60 ca ca ca ca o 1 o bas OO bas 00 bas X bas X o "Hb ca % human P globin/ mouse P major Hp •(3 major 0.8 0.9 22 Figure 4-12 RNASE protection analysis (RPA) to quantify the level of human P mRNA expression relative to mouse p major for human p globin constructs integrated at sites A2 and A8 using the Cre/lox system. RNA was isolated after in vitro differentiation of ES cells until day 10, and six days of expansion in general hematopoietic liquid culture conditions. Each sample was tested with a probe for the human p globin gene (350bp protected fragment, Hp) and a probe for the mouse P major (145 bp protected fragment). The percentage of human mRNA relative to mouse mRNA for each sample is shown on the bottom of the figure. These results are summarised along with the RPA obtained from constructs that were randomly integrated (figure 4.7) in Figure 4.13 and are instructive in several ways. At site A2, all three human p globin constructs could be detected. As a result, the 132 constructs could be ordered according to their expression levels; the HS 1-4 containing P globin construct was superior to the minimal promoter construct and the HS3 containing P globin construct, both of which were observed to perform similarly. At the second site (A8), all three constructs were again detectable. Moreover, the rank order of each of the constructs was observed to be the same as the first site (A2), (HS 1-4 > HS3 = minimal promoter), but at significantly lower levels of expression. In contrast, at the random integration site (polyclonal), only the HS 1-4 containing p globin construct was detected, although at very low levels. Thus, from these studies, the HS1-4 containing p globin construct performed better relative to the HS3 and minimal promoter containing P globin constructs at both A2 and A8 sites, an observation that was only apparent by using the Cre/lox system. In addition, comparison of the vectors at two different integration sites illustrated the influence of the site of integration on expression (or position effects), as levels were significantly higher at the A2 site than at the A8 site, again demonstrating that this strategy is much more informative than random integration. 133 Site 1 (A2) u Site 2 (A8) CCE Basic HS3 HS1-4 * ° CCE Basic HS3 HS1-4 Random (polyclonal) I 80 -, O | C 60 • ' S? o -I . . .—^m— CCE Basic HS3 HS1-4 Figure 4-13 Graph summarizes the results obtained from the RNASE protection assays illustrated in Figure 4.12 (at sites A2 and A8) and Figure 4.7 (random). The level of expression is represented as the percent of human p globin mRNA relative to mouse p major. FACS analysis to detect the human P globin protein further confirms these results. For example, Figure 4.14 shows that the A2 site is "permissive" for expression (at the protein level) as the percentage of cells that express human p globin is significantly increased compared to the barely detectable protein levels obtained following random integration (Figure 4.8), and further correlates with the rank order observed with RPA (HS1-4 >HS3=minimal promoter). Lastly, each of these constructs also exhibited erythroid-specific expression as demonstrated by the absence of globin mRNA and protein following expansion for two weeks in mast cell specific conditions (Tsai et al, 2000). A representative FACS profile using the HS1-4 containing P globin construct at the A2 site is shown in Figure 4.14. The generation of mast cells was confirmed by the > 95% 134 expression of the high-affinity IgE receptor and the lack of Terl 19 expression. Moreover, no human p globin protein was observed with any of the constructs. 135 IT) cn C o c o U c o 'cn C ct CX X W 2 .Ok .01 .01 .01 d5l mqoiS cj -«— JOjd3D3J HSI U fa 00 a on u X 3 61I»X' a o o u > CN uiqoiS cj utqoig (j a 'S3 o a a IS JO "©JD ca a E CU CO tn u. a x o s '5 cu a o cu .2f la o E CU T3 Vi *5 73 B 03 CU °c V s _^ o — 1- 4> CJ o 2 i en <u u s WD S3 cu u <u fe -3 CH ^ £ ca .S = o 2 H cd a3 This was further confirmed by RT-PCR analysis. Figure 4.15 is a representative RT-PCR using the HS 1-4 p globin construct at the A2 site after expansion in mast cell specific liquid culture conditions. Erythroid specific expression was demonstrated by the absence of human globin mRNA and expression of the HPRT gene. Murine P major 578 bp Human P globin 349 bp HPRT 249 bp <U cd U CU CU CO VO <u CO ;mou mouse ntrol ;hum nstrui ntrol: mouse ntrol ;hum ca o o CO W o 00 o + H + o CN X Figure 4-15 RT-PCR on differentiating ES cells demonstrating erythroid-specific expression of the human fi globin constructs. RT-PCR was performed on all three constructs from three independent experiments. Shown is a representative RT-PCR analysis using the HS 1-4 containing p globin construct at the A2 site after expansion in mast cell specific liquid culture conditions to generate non-erythroid cells. Eyrythoid specific expression was demonstrated by the absence of human globin mRNA and expression of the HPRT gene. 137 4.3 Discussion A strategy that combines the Cre/lox recombination system and ES in vitro differentiation was explored in this Chapter for studying gene regulation in hematopoietic cells. ES cells are an attractive model system to use as these cells are amenable to genetic manipulation and can differentiate into a variety of cell types, including hematopoietic. Moreover, the kinetics of in vitro differentiation is considered to accurately reflect in vivo hematopoietic development, thereby providing a potent model to study gene regulatory elements in hematopoietic cells. Including the Cre/lox system in this model allows a more critical evaluation of regulatory elements compared to random integration as constructs are analyzed at the same integration site. Furthermore, with this strategy the influence of a vector at different integration sites can also be compared for a more thorough analysis of position effects. This strategy allowed vector comparisons to be made, such that they could be ordered according to their expression levels. This proved to be very effective as it confirms and extends the results previously reported with these similar constructs in transgenic mice studies. For example, the HS 1-4 containing P globin construct ranked the highest in transgenic mice, as it is expressed at levels higher than the lowest ranked minimal promoter containing p globin construct (Pasceri et al., 1998). However, these transgenic mice studies are much more labour intensive, as they require the analysis of many transgenic lines to ensure that the expression observed is vector specific. It should be noted that there was a discrepancy in the results obtained with the HS3 containing P globin construct in the transgenic studies and in the results obtained here. In transgenic mice studies, the HS3 containing p globin construct was found to perform significantly 138 better (express -50% of human p globin mRNA relative to the mouse mRNA, (Rubin et al, 2000)) than the minimal promoter containing construct (HSl-4> HS3>minimal promoter), whereas in this study they appeared to perform similarly (HSl-4> HS3=minimal promoter). This may be attributed to the repressive influence of the MSCV LTR promoter during in vitro differentiation, which in turn may lead to lower levels of expression of the P globin construct. In this case, the HS3 containing p globin construct may be more sensitive to this repressive influence than the HS1-4 containing P globin construct. On the positive side, such an approach may be ideal to evaluate potential gene therapy vectors as a means to test their feasibility in the context of a retroviral vector. This strategy is not limited to using retroviral vectors to introduce the initial lox target sites. Lentiviral vectors could also be used, as recent studies have demonstrated that they have a decreased propensity for silencing in both differentiating EBs and in hematopoietic colonies (Hamaguchi et al., 2000). Furthermore, a recent report used a non-viral approach to obtain integration sites that would permit high levels of expression of a transgene (Wallace et al, 2000). Here, they randomly integrated the Oct 4 promoter/lacZ transgene into undifferentiated ES cells, followed by screening for those clones in which the promoter was properly regulated during in vitro differentiation. They demonstrated that this site could then be reused following homologous recombination to achieve high, appropriately regulated expression of lacZ from different tissue specific promoters both in vitro and in vivo following the development of transgenic mice (Wallace et al., 2000). Integration of very large, complex, constructs was also shown to be feasible in this study. Previous reports of using a p globin gene in a MoMLV retroviral vector demonstrated that stable transmission of the cassette was only achieved by deleting an AT 139 rich region in intron 2 and using only small core regions (-200-300 bp each) of the individual HSs (Leboulch et al, 1994). In contrast, Cre mediated integration was achieved with all of the human P globin constructs, despite the presence of this AT rich element. Moreover, larger segments of each of the HSs could be used (including the core region and flanking region which is - lkb each in size), that have recently been shown to confer higher levels of P globin gene expression relative to the core regions in an MEL cell line (Molete et al., 2001). As well, the largest construct tested was ~9kb, which is beyond the size capacity for retroviral vectors, was shown to be stably integrated using this approach (Pasceri et al, 1998). These studies also revealed the importance of the LCR in conferring high levels of P globin expression. For example, the HS1-4 containing p globin construct demonstrated -6 fold to a 30 fold increase in expression levels at the A2, A8 site respectively, compared to the minimal promoter containing P globin construct. These results, however, further demonstrated that this extensive LCR containing construct is still subjected to position effects, by the analysis of variable (lower) levels of expression at a second integration site (A8). Although only two integration sites were tested in this Chapter, these results have also been confirmed by transgenic mice studies involving a yeast artificial chromosome (YAC) containing the entire 150 kb human p globin locus. Alami et al, showed that despite using such a large construct, position effects were still observed at 3A integration sites analyzed (Alami et al., 2000). Interestingly, cytological analysis demonstrated that all of the YAC integrations occurred at non-centromeric locations, suggesting that such non-permissive sites for expression may occur frequently in the genome than previously thought (Alami et al., 2000). This suggests that cis-regulatory 140 elements that may prevent silencing and position effects, like insulators, may prove to be extremely useful in future vector designs for gene therapy. The expression levels at both of the targeted integration sites (A2 and A8) appeared to be significantly improved relative to random integration. For example, the expression levels of the minimum promoter containing construct were increased to ~ 40 fold (at the A8 site) and as high as 600 fold (at the A2 site), whereas barely detectable levels were observed when it was randomly integrated. It is interesting to speculate that retroviral vectors used in the first step may seek out more "permissive" integration sites that confer higher expression levels relative to random integration of a plasmid delivered by electroporation. It is thought that retroviruses select for integration sites with a high and stable transcriptional potential to ensure expression from the LTR promoter and are reported to be near DNase I HSs and/or sequences with an affinity for the nuclear matrix (reviewed in Bodine, 2001). However, as mentioned earlier only two integration sites were tested in this Chapter and the ES cell pool harboring the basic P globin construct contained a predominant clone with a single integration site. Analysis of more clones at both targeted and random integration sites are needed to confirm these results. ES cells have been used as an alternative in vitro model to hematopoietic cell lines as they more accurately reflect hematopoiesis in vivo and can be induced to differentiate into several different lineages. For example, a recent report using D3 ES cells on the OP9 stromal cell line demonstrated that large numbers of erythrocytes, megakaryoctes, and macrophages could be selectively amplified by the addition of appropriate growth factors (Era et al., 2000). This was also illustrated in this study using CCE ES cells, in which two different growth factor cocktails were compared for their 141 ability to generate erythroid cells. Even though there was no significant difference in cell number, the erythroid specific conditions were able to induce a more homogenous population of erythroid cells relative to the general hematopoietic growth factor conditions. In addition, lineage specific expression of the human P globin constructs was analyzed by changing the growth factor cocktail to induce the selective proliferation of mast cells. Thus, a construct can be integrated into the genome of undifferentiated ES cells and studied within one or more hematopoietic lineages, through the addition of different growth factor combinations during differentiation. This has the advantage over hematopoietic cell lines, in which expression in multiple lineages is commonly examined by using several different cell lines. This study further documents and provides the first detailed comparison of P globin gene regulation in ES cells. One other report performed by Grosveld et al, used the ES system to study globin gene regulation by analyzing RNA from EBs at different times post-differentiation (Lindenbaum and Grosveld, 1990). In this study, a more extensive evaluation of erythroid development and globin gene expression during ES in vitro differentiation was performed. Using this model system, a large number of mature erythroid cells were generated from a relatively small number of undifferentiated ES cells, such that studies (RPA and protein analysis) could be performed with erythroid cells derived from differentiating ES cells, rather than EBs. The development of primitive and definitive erythroid progenitors were also shown to correlate with the profile of globin gene expression during in vitro differentiation, which mirrors that observed in the developing embryo. For example, the embryonic PHI expression was illustrated to precede the expression of both the adult pmajorand human P globin. The embryonic globin, however, was not down regulated at later time points in 142 this study, but rather continued to be expressed at all stages. These results cannot be attributed to lack of specificity of the primers since expression was never observed in the adult mouse spleen control or in undifferentiated ES cells. Grosveld et al, also reported a similar expression pattern and speculated that this could be due to heterogeneity in the synchrony of differentiation due to the growing or differentiation of new EBs during the differentiation process (Lindenbaum and Grosveld, 1990). Even though only the PHI gene showed promiscuous expression in this work, all of the comparison studies between the constructs were performed using RNA from erythroid cells rather than EBs in order to avoid this problem. 143 CHAPTER 5 DISCUSSION Retroviral gene transfer to HSCs has proven to be a powerful tool in performing gene function studies and has also been exploited for gene therapy. Although instrumental in this regard, they do suffer from several drawbacks which include limitations on the size of the transgene that can be packaged in the viral genome and the lack of long-term, predictable transgene expression. The overall goal of the work presented in this thesis was to develop an alternative integration strategy that may overcome some of these hurdles, and further provides an improved platform for performing gene function and gene regulation studies in hematopoietic cells. This site directed integration strategy conferred several advantages over existing viral delivery approaches. Cre mediated integration was feasible on very large, complex contructs that have previously been reported to be unstable in retroviral vectors and the size of which go beyond the current limitations for retroviral vectors. The ability to integrate large DNA fragments from an endogenous locus of a transgene may allow for more reproducible expression, as it is likely to include all of the sequences required to establish its native epigenetic organization. This also provides a powerful way to create mouse models of human diseases. For example, Call et al, demonstrated the feasibility of using the Cre/lox system to integrate a YAC containing defined genomic regions of chromosome 21 into ES cells to correlate the imbalance in gene dosage with specific Down Syndrome features in transgenic mice derived thereof (Call et al., 2000). This strategy also provided an advantage over viral vectors by achieving predictable levels of transgene expression upon re-targeting a locus with desired expression properties. This 144 could prove to be particularly important for those studies in which the appropriate regulatory elements needed for expression have not been defined. As well, such an approach may facilitate the dissection of protein function by generating mutated versions (in the DNA binding domain or protein binding domain) of endogenous gene products. Moreover, it may also aid in the production of proteins that are of biological value by insertion of a transgene into loci in which expression is high or appropriately regulated. For example, Schubeler et al, used the FLP/FRT system to produce therapeutic levels of the human clotting factor Vll gene following re-targeting of the gene into proviral integration sites that were expressed at high levels (Schubeler et al., 1998). The studies described above may be performed in a wide range of cell types, as the Cre/lox integration strategy was shown to be applicable and occurred at a high efficiency in a number of non-hematopoietic and hematopoietic (both human and mouse) cell lines. Attainment of this approach in pluripotent ES cells also provides an avenue for performing similar studies in germline competent ES cells that are capable of generating transgenic mice. In Chapter 3, this strategy was shown to occur in hematopoietic progenitors at a useable frequency. However, it still remains to be determined if such an approach could occur in HSCs. As discussed in Chapter 3, assuming site directed integration occurs at the same frequency in hematopoietic progenitors as HSCs, then ~ 5 x 106 bone marrow cells would be required to detect one targeted HSC. This frequency may be improved by using phenotypic markers to separate HSCs from the other hematopoietic cell types, followed by selection strategies that can enhance their proliferation. One of these includes the 145 transcription factor Hox B4, which has been demonstrated to be a potent enhancer of HSC growth and therefore may provide a powerful way to enhance proliferation of transduced cells in vitro prior to Cre mediated integration. For example, Antonchuck et al, showed that overexpression of Hox B4 in murine bone marrow cells conferred a 40 fold net expansion of HSCs compared to transduced bone marrow control cells (Antonchuk et al., 2001). Another intriguing strategy would be to amplify only "targeted" HSCs after site directed integration. This could be accomplished by introducing a selectable marker along with the transgene of interest into the target locus, followed by a selection strategy to enrich for these cells, while simultaneoulsy killing unmodified stem cells. For example, recent work by Allay et al, in a murine transplant model demonstrated amplification of vector expressing cells in vivo of multiple lineages from less than 10% to more than 70% using a retroviral vector containing a mutant dihydrofolate reductase antifolate resistance gene and an antifolate treatment regimen (Allay et al., 1998). Since there is difficulty in expanding HSCs ex vivo, increasing evidence within the last few years have suggested that this may be overcome by substituting stem cells from other sources. The long-standing concept has been that organ specific stem cells are restricted to making differentiated cell types of the tissue in which they reside. That is, they have lost their capacity to generate other cell types in the body (Anderson et al., 2001). Recent experiments challenge this and provide support that adult stem cells from one tissue or organ can be induced to differentiate into cells of other organs in vitro and in vivo. One of the first provocative experiments demonstrated that bone marrow derived cells could target and differentiate into muscle cells (Ferrari et al., 1998). This was quickly followed by a report that adult neural cell cultures containing neural stem cells injected into 146 irradiated mice (to destroy HSCs), could differentiate into blood cells (Bjornson et al, 1999). A flurry of studies followed that reported: bone marrow-to brain, bone marrow derived stroma-to brain, bone marrow-to liver, skin to brain, brain to heart and other such stem cell differentiations (reviewed in (Morrison, 2001)). It also appears that adult somatic stem cells may be similar to ES cells, if given the right cues. For example, Clarke et al, injected adult neural stem cells into the inner cell mass of a blastocyst, and found it contributed to a wide variety of tissues, including liver, brain, intestine and heart. Interestingly, they did not find higher rates of incorporation into the brain, which suggests that adult stem cells do not retain their bias to the tissues in which they were originally derived from (Clarke et al,, 2000). Hence, it may not be unrealistic to envision in the near future of obtaining stem cells from tissues that are easier to grow (ie. skin), genetically manipulating them, followed by their induction to become hematopoietic. The current limitation of this strategy is the requirement that the genome be first modified with a lox P target site. Recent studies have revealed a number of intriguing approaches that may overcome this problem. Hartung et al, constructed a number of Cre mutants with amino acid substitutions in different positions of the DNA binding region. Interestingly, in vivo recombination experiments demonstrated that one of these modifications changed the target site specificity, such that it recognized a different sequence from the wild type lox P site (Hartung and Kisters-Woike, 1998). Moreover, the eukaryotic type IB topoisomerases have been shown to contain regions that are highly homologous to the catalytic regions of site specific recombinases (Cao and Hayes, 1999). These studies are appealing as they suggest it may be possible to utilize a "modified" Cre recombinase or even a eukaryotic enzyme to target a gene to a specfic site (and/or at more 147 than one site) in the genome. In other studies, Cre recombinase and (f>C31 integrase from a Streptomyces phage were found to mediate integration of a plasmid in the absence of lox or att target sites, respectively. Sequencing of the flanking regions of these integration sites illustrated preferential integration into specific loci in both human and mouse cells. These results suggest the potential existence of "psuedo lox/att sites" in the eukaryotic genome (Thyagarajan et al, 2000; Thyagarajan et al, 2001). Although the frequency of integration at these sites was low (0.07% in human cells, 1.2% in mouse cells), it is interesting to speculate the use of these endogenous target sites for site directed integration of a gene in the near future. In Chapter 4, Cre/lox site directed integration was used for studying regulatory mechanisms in hematopoietic cells and for its use in designing better vectors for gene therapy. In transgenic mice studies, transgenes are studied in ectopic genomic locations and deciphering if a particular cis element in a transgene stimulates transcription, suppresses position effects or does both is difficult as transcription levels are affected by position effects (Bender et al, 2001). The strategy developed here allows such a distinction to be made by analyzing constructs at the same integration site (assess transcription levels) and at more than one site (assess effects on the site of integration). This approach would be useful for evaluating the design of a vector for gene therapy, as a cassette that can be expressed not only at appropriate levels, but also at ectopic locations in the genome is required. The studies performed in Chapter 4, revealed the importance of the LCR in conferring high level of p globin gene expression but also demonstrated that this extensive LCR containing construct is still subjeted to position effects. This suggests that cis-regulatory elements that may prevent silencing and position effects, like 148 insulators or MARs/SARs, may prove to be useful in future vector designs for gene therapy. The majority of reports on insulators in retroviral vectors have focused on their placement in both the 5' and 3' LTRs to prevent position effects from surronding host chromatin (Chung et al, 1997; Emery et al, 2000; Pikaart et al, 1998). These studies have shown that insulators can improve expression from a variety of integration sites, but do not have the ability to overcome all position effects (Emery et al, 2000; Rivella et al, 2000). A more effective method may be to place the insulator around the gene itself, thereby also preventing repressive effects from the viral LTR. Therefore, in addition to studying cis regulatory elements, this strategy may also provide a valuable model for testing insulator function and its most effective placement within a retroviral context. Even though the work presented in Chapter 4 involved hematopoietic cells, the ES model could potentially be used for gene regulation studies in any cell lineage. For example, given the proper combination of growth factors, EBs can develop into such diverse cell types as adipocytes, muscular smooth muscle cells, oligodendrocytes or astrocytes (Fuchs and Segre, 2000). A further extension of using the ES model system was recently demonstrated by Wallace et al. They screened a number of ES clones for sites in the mouse genome that allowed proper regulated expression of a reporter gene during in vitro differentiation. Two clones were chosen which displayed appropriate and inappropriate gene regulation. Interestingly, homologous recombination of a second transgene into these sites recapitulated the expression pattern at each of the corresponding sites, both in vitro and in vivo (Wallace et al, 2000). Thus, it would be interesting to obtain a number of predefined and reusable reference ES cells that contain lox target sequences integrated at sites in the 149 genome that exhibit a wide spectrum of expression properties. For example, clones that are expressed at high levels, low levels, subjected to position effects, exhibit lineage specific expression or ubiquitously expressed. Re-targeting of constructs into these sites followed by analysis in vitro and subsequently in vivo, would further increase our understanding of gene regulatory mechanisms. In summary, the gene transfer strategy developed in this work may prove to be a powerful tool for performing gene function and gene regulation studies in a wide variety of cell lineages both in vitro and in vivo, and with the continued progress on stem cell expansion and alternative sources for stem cells, may eventually pave the way to its use in HSCs. 150 CHAPTER 6 REFERENCES Akashi, K., Traver, D., Miyamoto, T. and Weissman, LL. (2000) A clonogenic common myeloid progenitor that gives rise to all myeloid lineages. Nature, 404, 193-7. Alami, R., Greally, J.M., Tanimoto, K., Hwang, S., Feng, Y.Q., Engel, J.D., Fiering, S. and Bouhassira, E.E. (2000) Beta-globin YAC transgenes exhibit uniform expression levels but position effect variegation in mice. Hum Mol Genet, 9, 631-6. Allay, J.A., Persons, DA., Galipeau, J., Riberdy, J.M., Ashmun, R.A., Blakley, R.L. and Sorrentino, B.P. (1998) In vivo selection of retrovirally transduced hematopoietic stem cells. Nat Med, 4, 1136-43. Anderson, D.J., Gage, F.H. and Weissman, LL. (2001) Can stem cells cross lineage boundaries? Nat Med, 7, 393-5. Anderson, W.F. (1998) Human gene therapy. Nature, 392, 25-30. Antonchuk, J., Sauvageau, G. and Humphries, R.K. (2001) HOXB4 overexpression mediates very rapid stem cell regeneration and competitive hematopoietic repopulation. Exp Hematol, 29, 1125-34. Apperley, J.F., Luskey, B.D. and Williams, D.A. (1991) Retroviral gene transfer of human adenosine deaminase in murine hematopoietic cells: effect of selectable marker sequences on long-term expression. Blood, 78, 310-7. Araki, K., Araki, M., Miyazaki, J. and Vassalli, P. (1995) Site-specific recombination of a transgene in fertilized eggs by transient expression of Cre recombinase. Proc Natl Acad Sci US A, 92,160-4. Araki, K., Araki, M. and Yamamura, K. (1997) Targeted integration of DNA using mutant lox sites in embryonic stem cells. Nucleic Acids Res, 25, 868-72. Banerjee, S., Livanos, E. and Vos, J.M. (1995) Therapeutic gene delivery in human B-lymphoblastoid cells by engineered non-transforming infectious Epstein-Barr virus. Nat Med, 1, 1303-8. Barrette, S., Douglas, J.L., Seidel, N.E. and Bodine, D.M. (2000) Lentivirus-based vectors transduce mouse hematopoietic stem cells with similar efficiency to moloney murine leukemia virus-based vectors. Blood, 96, 3385-91. Baubonis, W. and Sauer, B. (1993) Genomic targeting with purified Cre recombinase. Nucleic Acids Res, 21, 2025-9. 151 Bender, M.A., Gelinas, R.E. and Miller, A.D. (1989) A majority of mice show long-term expression of a human beta-globin gene after retrovirus transfer into hematopoietic stem cells. Mol Cell Biol, 9, 1426-34. Bender, M.A., Roach, J.N., Halow, J., Close, J., Alami, R., Bouhassira, E.E., Groudine, M. and Fiering, S.N. (2001) Targeted deletion of 5'HS1 and 5'HS4 of the beta-globin locus control region reveals additive activity of the DNasel hypersensitive sites. Blood, 98, 2022-7. Bergan, R., Hakim, F., Schwartz, G.N., Kyle, E., Cepada, R., Szabo, J.M., Fowler, D., Gress, R. and Neckers, L. (1996) Electroporation of synthetic oligodeoxynucleotides: a novel technique for ex vivo bone marrow purging. Blood, 88, 731-41. Berghella, L., De Angelis, L., Coletta, M., Berarducci, B., Sonnino, C, Salvatori, G., Anthonissen, C, Cooper, R., Butler-Browne, G.S., Mouly, V., Ferrari, G., Mavilio, F. and Cossu, G. (1999) Reversible immortalization of human myogenic cells by site-specific excision of a retrovirally transferred oncogene. Hum Gene Ther, 10, 1607-17. Bethke, B. and Sauer, B. (1997) Segmental genomic replacement by Cre-mediated recombination: genotoxic stress activation of the p53 promoter in single-copy transformants. Nucleic Acids Res, 25, 2828-34. Bjornson, C.R., Rietze, R.L., Reynolds, B.A., Magli, M.C. and Vescovi, A.L. (1999) Turning brain into blood: a hematopoietic fate adopted by adult neural stem cells in vivo. Science, 283, 534-7. Blouin, M.J., Beauchemin, H., Wright, A., De Paepe, M., Sorette, M., Bleau, A.M., Nakamoto, B., Ou, C.N., Stamatoyannopoulos, G. and Trudel, M. (2000) Genetic correction of sickle cell disease: insights using transgenic mouse models. Nat Med, 6, 177-82. Bodine, D.M. (2001) Hematopoietic Stem Cell Gene Therapy: Progress and Prospects. In Zon, G. (ed.) Hematopoiesis: A Development Approach. Oxford University Press, New York, Vol. 36. Bouhassira, E.E., Westerman, K. and Leboulch, P. (1997) Transcriptional behavior of LCR enhancer elements integrated at the same chromosomal locus by recombinase-mediated cassette exchange. Blood, 90, 3332-44. Boulikas, T. (1993) Nature of DNA sequences at the attachment regions of genes to the nuclear matrix. J Cell Biochem, 52, 14-22. Bradley, A., Evans, M., Kaufman, M.H. and Robertson, E. (1984) Formation of germ-line chimaeras from embryo-derived teratocarcinoma cell lines. Nature, 309, 255-6. 152 Bradley, T.R. and Metcalf, D. (1966) The growth of mouse bone marrow cells in vitro. AustJExp Biol Med Sci, 44, 287-99. Brenner, M.K. (1993) Gene transfer into human hematopoietic progenitor cells: a review of current clinical protocols. JHematother, 2, 7-17. Brown, J.P., Wei, W. and Sedivy, J.M. (1997) Bypass of senescence after disruption of p21CIPT/WAFl gene in normal diploid human fibroblasts. Science, 277, 831-4. Buchholz, F., Angrand, P.O. and Stewart, A.F. (1998) Improved properties of FLP recombinase evolved by cycling mutagenesis. Nat Biotechnol, 16, 657-62. Buchholz, F., Ringrose, L., Angrand, P.O., Rossi, F. and Stewart, A.F. (1996) Different thermostabilities of FLP and Cre recombinases: implications for applied site-specific recombination. Nucleic Acids Res, 24, 4256-62. Bulger, M. and Groudine, M. (1999) Looping versus linking: toward a model for long distance gene activation. Genes Dev, 13, 2465-77. Call, L.M., Moore, C.S., Stetten, G. and Gearhart, J.D. (2000) A cre-lox recombination system for the targeted integration of circular yeast artificial chromosomes into embryonic stem cells. Hum Mol Genet, 9, 1745-51. Cao, Y. and Hayes, F. (1999) A newly identified, essential catalytic residue in a critical secondary structure element in the integrase family of site-specific recombinases is conserved in a similar element in eucaryotic type IB topoisomerases. J Mol Biol, 289, 517-27. Capecchi, M.R. (2001) Generating mice with targeted mutations. Nat Med, 7, 1086-90. Cavazzana-Calvo, M., Hacein-Bey, S., de Saint Basile, G., Gross, F., Yvon, E., Nusbaum, P., Selz, F., Hue, C, Certain, S., Casanova, J.L., Bousso, P., Deist, F.L. and Fischer, A. (2000) Gene therapy of human severe combined immunodeficiency (SCID)-Xl disease. Science, 288, 669-72. Challita, P.M. and Kohn, D.B. (1994) Lack of expression from a retroviral vector after transduction of murine hematopoietic stem cells is associated with methylation in vivo. Proc Natl Acad Sci USA, 91, 2567-71. Chang, J.C., Liu, D. and Kan, Y.W. (1992) A 36-base-pair core sequence of locus control region enhances retrovirally transferred human beta-globin gene expression. Proc Natl Acad Sci USA, 89, 3107-10. 153 Charron, J., Malynn, B.A., Robertson, E.J., Goff, S.P. and Alt, F.W. (1990) High-frequency disruption of the N-myc gene in embryonic stem and pre-B cell lines by homologous recombination. Mol Cell Biol, 10, 1799-804. Chatterjee, S., Li, W., Wong, C.A., Fisher-Adams, G., Lu, D., Guha, M., Macer, J.A., Forman, S.J. and Wong, K.K., Jr. (1999) Transduction of primitive human marrow and cord blood-derived hematopoietic progenitor cells with adeno-associated virus vectors. Blood, 93, 1882-94. Cheng, L., Du, C, Lavau, C, Chen, S., Tong, J., Chen, B.P., Scollay, R., Hawley, R.G. and Hill, B. (1998) Sustained gene expression in retrovirally transduced, engrafting human hematopoietic stem cells and their lympho-myeloid progeny. Blood, 92, 83-92. Chiba, S., Takaku, F., Tange, T., Shibuya, K., Misawa, C, Sasaki, K., Miyagawa, K., Yazaki, Y. and Hirai, H. (1991) Establishment and erythroid differentiation of a cytokine-dependent human leukemic cell line F-36: a parental line requiring granulocyte-macrophage colony-stimulating factor or interleukin-3, and a subline requiring erythropoietin. Blood, 78, 2261-8. Chung, J.H., Bell, A.C. and Felsenfeld, G. (1997) Characterization of the chicken beta-globin insulator. Proc Natl Acad Sci USA, 94, 575-80. Clarke, D.L., Johansson, C.B., Wilbertz, J., Veress, B., Nilsson, E., Karlstrom, H., Lendahl, U. and Frisen, J. (2000) Generalized potential of adult neural stem cells. Science, 288, 1660-3. Cockerill, P.N. and Garrard, W.T. (1986) Chromosomal loop anchorage of the kappa immunoglobulin gene occurs next to the enhancer in a region containing topoisomerase U. sites. Cell, 44, 273-82. Cole-Strauss, A., Yoon, K., Xiang, Y., Byrne, B.C., Rice, M.C., Gryn, J., Holloman, W.K. and Kmiec, E.B. (1996) Correction of the mutation responsible for sickle cell anemia by an RNA-DNA oligonucleotide. Science, 273, 1386-9. Colicelli, J. and Rigby, P.W.J. (1987) Isolation of a recombinant murine leukemia virus utilizing a new primer tRNA. J. Virol, 57, 37. Coll, J.L., Ben-Ze'ev, A., Ezzell, R.M., Rodriguez Fernandez, J.L.; Baribault, H., Oshima, R.G. and Adamson, E.D. (1995) Targeted disruption of vinculin genes in F9 and embryonic stem cells changes cell morphology, adhesion, and locomotion. Proc Natl Acad Sci US A, 92, 9161-5. Correll, P.H., Colilla, S. and Karlsson, S. (1994) Retroviral vector design for long-term expression in murine hematopoietic cells in vivo. Blood, 84, 1812-22. 154 Culver, K.W., Hsieh, W.T., Huyen, Y., Chen, V., Liu, J., Khripine, Y. and Khorlin, A. (1999) Correction of chromosomal point mutations in human cells with bifunctional oligonucleotides. Nat Biotechnol, 17, 989-93. Curry, J.L. and Trentin, J.J. (1967) Hemopoietic spleen colony studies. I. Growth and differentiation. Dev Biol, 15, 395-413. Dai, Y., Roman, M., Naviaux, R.K. and Verma, I.M. (1992) Gene therapy via primary myoblasts: long-term expression of factor IX protein following transplantation in vivo. Proc Natl Acad Sci USA, 89, 10892-5. Dang, Q., Auten, J. and Plavec, I. (2000) Human beta interferon scaffold attachment region inhibits de novo methylation and confers long-term, copy number-dependent expression to a retroviral vector. J Virol, 74, 2671-8. Danielian, P.S., Muccino, D., Rowitch, D.H., Michael, S.K. and McMahon, A.P. (1998) Modification of gene activity in mouse embryos in utero by a tamoxifen-inducible form of Cre recombinase. Curr Biol, 8, 1323-6. Dick, J.E., Magli, M.C., Huszar, D., Phillips, RA. and Bernstein, A. (1985) Introduction of a selectable gene into primitive stem cells capable of long-term reconstitution of the hemopoietic system of W/Wv mice. Cell, 42, 71-9. Doetschman, T.C., Eistetter, H., Katz, M., Schmidt, W. and Kemler, R. (1985) The in vitro development of blastocyst-derived embryonic stem cell lines: formation of visceral yolk sac, blood islands and myocardium. JEmbryol Exp Morphol, 87, 27-45. Doi, S., Campbell, C. and Kucherlapati, R. (1992) Directed modification of genes by homologous recombination in mammalian cells. Academic press, London. Dymecki, S.M. (1996) Flp recombinase promotes site-specific DNA recombination in embryonic stem cells and transgenic mice. Proc Natl Acad Sci USA, 93, 6191-6. Dzierzak, E.A., Papayannopoulou, T. and Mulligan, R.C. (1988) Lineage-specific expression of a human beta-globin gene in murine bone marrow transplant recipients reconstituted with retro virus-transduced stem cells. Nature, 331, 35-41. Dzierzak, E.a.O., R. (2001) Hematopoietic Stem Cell Development in Mammals. In Zon, G. (ed.) Hematopoiesis: A Development Approach. Oxford University Press, New York, Vol. 36. Eaton, W.A. and Hofrichter, J. (1995) The biophysics of sickle cell hydroxyurea therapy. Science, 268, 1142-3. 155 Emery, D.W., Yannaki, E., Tubb, J. and Stamatoyannopoulos, G. (2000) A chromatin insulator protects retrovirus vectors from chromosomal position effects. Proc Natl Acad Sci US A, 97, 9150-5. Engel, J.D. and Tanimoto, K. (2000) Looping, linking, and chromatin activity: new insights into beta-globin locus regulation. Cell, 100, 499-502. Era, T., Takagi, T., Takahashi, T., Bories, J.C. and Nakano, T. (2000) Characterization of hematopoietic lineage-specific gene expression by ES cell in vitro differentiation induction system. Blood, 95, 870-8. Evans, J.T., Kelly, P.F., O'Neill, E. and Garcia, J.V. (1999) Human cord blood CD34+CD38- cell transduction via lentivirus-based gene transfer vectors. Hum Gene Ther, 10, 1479-89. Evans, M.J. and Kaufman, M.H. (1981) Establishment in culture of pluripotential cells from mouse embryos. Nature, 292, 154-6. Feng, Y.Q., Lorincz, M.C., Fiering, S., Greally, J.M. and Bouhassira, E.E. (2001) Position effects are influenced by the orientation of a transgene with respect to flanking chromatin. Mol Cell Biol, 21, 298-309. Feng, Y.Q., Seibler, J., Alami, R., Eisen, A., Westerman, K.A., Leboulch, P., Fiering, S. and Bouhassira, E.E. (1999) Site-specific chromosomal integration in mammalian cells: highly efficient CRE recombinase-mediated cassette exchange. J Mol Biol, 292, 779-85. Ferrari, G., Cusella-De Angelis, G., Coletta, M., Paolucci, E., Stornaiuolo, A., Cossu, G. and Mavilio, F. (1998) Muscle regeneration by bone marrow-derived myogenic progenitors. Science, 219, 1528-30. Festenstein, R. and Kioussis, D. (2000) Locus control regions and epigenetic chromatin modifiers. CurrOpin Genet Dev, 10, 199-203. Forrester, W.C., Takegawa, S., Papayannopoulou, T., Stamatoyannopoulos, G. and Groudine, M. (1987) Evidence for a locus activation region: the formation of developmentally stable hypersensitive sites in globin-expressing hybrids. Nucleic Acids Res, 15, 10159-77. Fuchs, E. and Segre, J.A. (2000) Stem cells: a new lease on life. Cell, 100, 143-55. Garner, C, Tatu, T., Reittie, J.E., Littlewood, T., Darley, J., Cervino, S., Farrall, M., Kelly, P., Spector, T.D. and Thein, S.L. (2000) Genetic influences on F cells and other hematologic variables: a twin heritability study. Blood, 95, 342-6. 156 Gelinas, C. and Temin, H.M. (1986) Nondefective spleen necrosis virus-derived vectors define the upper size limit for packaging reticuloendotheliosis viruses. Proc Natl Acad Sci USA, 83, 9211-5. Gorman, C. and Bullock, C. (2000) Site-specific gene targeting for gene expression in eukaryotes. Curr Opin Biotechnol, 11, 455-60. Grez, M., Akgun, E., Hilberg, F. and Ostertag, W. (1990) Embryonic stem cell virus, a recombinant murine retrovirus with expression in embryonic stem cells. Proc Natl Acad Sci US A, 87,9202-6. Grez, M., Zornig, M., Nowock, J. and Ziegler, M. (1991) A single point mutation activates the Moloney murine leukemia virus long terminal repeat in embryonal stem cells. J Virol, 65,4691-8. Grimes, B.e.a. (unpublished results). Grosveld, F., van Assendelft, G.B., Greaves, D.R. and Kollias, G. (1987) Position-independent, high-level expression of the human beta-globin gene in transgenic mice. Cell, 51,975-85. Gu, H., Marth, J.D., Orban, P.C., Mossmann, H. and Rajewsky, K. (1994) Deletion of a DNA polymerase beta gene segment in T cells using cell type-specific gene targeting. Science, 265, 103-6. Hamaguchi, I., Woods, N.B., Panagopoulos, I., Andersson, E., Mikkola, H., Fahlman, C, Zufferey, R., Carlsson, L., Trono, D. and Karlsson, S. (2000) Lentivirus vector gene expression during ES cell-derived hematopoietic development in vitro. J Virol, 74, 10778-84. Hanaoka, K., Hayasaka, M., Uetsuki, T., Fujisawa-Sehara, A. and Nabeshima, Y. (1991) A stable cellular marker for the analysis of mouse chimeras: the bacterial chloramphenicol acetyltransferase gene driven by the human elongation factor 1 alpha promoter. Differentiation, 48, 183-9. Hargrove, P.W., Vanin, E.F., Kurtzman, G.J. and Nienhuis, A.W. (1997) High-level globin gene expression mediated by a recombinant adeno-associated virus genome that contains the 3' gamma globin gene regulatory element and integrates as tandem copies in erythroid cells. Blood, 89,2167-75. Harrington, J.J., Van Bokkelen, G., Mays, R.W., Gustashaw, K. and Willard, H.F. (1997) Formation of de novo centromeres and construction of first-generation human artificial microchromosomes. Nat Genet, 15, 345-55. 157 Harrison, D.E. (1980) Competitive repopulation: a new assay for long-term stem cell functional capacity. Blood, 55, 77-81. Hartung, M. and Kisters-Woike, B. (1998) Cre mutants with altered DNA binding properties. J Biol Chem, 273, 22884-91. Hatada, S., Nikkuni, K., Bentley, S.A., Kirby, S. and Smithies, O. (2000) Gene correction in hematopoietic progenitor cells by homologous recombination. Proc Natl Acad Sci U S y*. 97, 13807-11. Hawley, R.G., Lieu, F.H., Fong, A.Z. and Hawley, T.S. (1994) Versatile retroviral vectors for potential use in gene therapy. Gene Ther, 1, 136-8. Heiser, W. (2000) Optimizing electroporation conditions for the transformation of mammalian cells. Humana Press Inc., Totowa, NJ. Helgason, CD., Sauvageau, G., Lawrence, H.J., Largman, C. and Humphries, R.K. (1996) Overexpression of HOXB4 enhances the hematopoietic potential of embryonic stem cells differentiated in vitro. Blood, 87, 2740-9. Herzog, R.W., Yang, E.Y., Couto, L.B., Hagstrom, J.N., Elwell, D., Fields, P.A., Burton, M., Bellinger, D.A., Read, M.S., Brinkhous, K.M., Podsakoff, G.M., Nichols, T.C, Kurtzman, G.J. and High, K.A. (1999) Long-term correction of canine hemophilia B by gene transfer of blood coagulation factor IX mediated by adeno-associated viral vector. Nat Med, 5, 56-63. Hilberg, F., Stocking, C, Ostertag, W. and Grez, M. (1987) Functional analysis of a retroviral host-range mutant: altered long terminal repeat sequences allow expression in embryonal carcinoma cells. Proc Natl Acad Sci USA, 84, 5232-6. Hirata, R.K.a.H., D.W.R. (unpublished results). Hodgson, G.S. and Bradley, T.R. (1979) Properties of haematopoietic stem cells surviving 5-fluorouracil treatment: evidence for a pre-CFU-S cell? Nature, 281, 381-2. Jahner, D., Stuhlmann, H., Stewart, C.L., Harbers, K., Lohler, J., Simon, I. and Jaenisch, R. (1982) De novo methylation and expression of retroviral genomes during mouse embryogenesis. Nature, 298, 623-8. Jainchill, J.L., Aaronson, S.A. and Todaro, G.J. (1969) Murine sarcoma and leukemia viruses: assay using clonal lines of contact-inhibited mouse cells. J Virol, 4, 549-53. Jones, P.L., Veenstra, G.J., Wade, P.A., Vermaak, D., Kass, S.U., Landsberger, N., Strouboulis, J. and Wolffe, A.P. (1998) Methylated DNA and MeCP2 recruit histone deacetylase to repress transcription. Nat Genet, 19, 187-91. 158 Kalberer, CP., Pawliuk, R., Imren, S., Bachelot, T., Takekoshi, K.J., Fabry, M., Eaves, C.J., London, I.M., Humphries, R.K. and Leboulch, P. (2000) Preselection of retro virally transduced bone marrow avoids subsequent stem cell gene silencing and age-dependent extinction of expression of human beta-globin in engrafted mice. Proc Natl Acad Sci U S A, 97, 5411-5. Karlsson, S., Bodine, D.M., Perry, L., Papayannopoulou, T. and Nienhuis, A.W. (1988) Expression of the human beta-globin gene following retroviral-mediated transfer into multipotential hematopoietic progenitors of mice. Proc Natl Acad Sci USA, 85, 6062-6. Kay, M.A., Glorioso, J.C. and Naldini, L. (2001) Viral vectors for gene therapy: the art of turning infectious agents into vehicles of therapeutics. Nat Med, 1, 33-40. Kellendonk, C, Tranche, F., Monaghan, A.P., Angrand, P.O., Stewart, F. and Schutz, G. (1996) Regulation of Cre recombinase activity by the synthetic steroid RU 486. Nucleic Acids Res, 24,1404-11. Keller, G., Kennedy, M., Papayannopoulou, T. and Wiles, M.V. (1993) Hematopoietic commitment during embryonic stem cell differentiation in culture. Mol Cell Biol, 13, 473-86. Keller, G., Lacaud, G. and Robertson, S. (1999) Development of the hematopoietic system in the mouse. Exp Hematol, 27, 777-87. Keller, G.M. (1995) In vitro differentiation of embryonic stem cells. Curr Opin Cell Biol, 7, 862-9. Kelly, P.F., Vandergriff, J., Nathwani, A., Nienhuis, A.W. and Vanin, E.F. (2000) Highly efficient gene transfer into cord blood nonobese diabetic/severe combined immunodeficiency repopulating cells by oncoretroviral vector particles pseudotyped with the feline endogenous retrovirus (RD114) envelope protein. Blood, 96, 1206-14. Kilby, N.J., Snaith, M.R. and Murray, JA. (1993) Site-specific recombinases: tools for genome engineering. Trends Genet, 9, 413-21. Knox-Macaulay, H.H., Weatherall, D.J., Clegg, J.B., Bradley, J. and Brown, M.J. (1972) The clinical and biosynthetic characterization of -thalassaemia. Br J Haematol, 22, 497-512. Kobayashi, N., Fujiwara, T., Westerman, K.A., Inoue, Y., Sakaguchi, M., Noguchi, H., Miyazaki, M., Cai, J., Tanaka, N., Fox, I.J. and Leboulch, P. (2000) Prevention of acute liver failure in rats with reversibly immortalized human hepatocytes. Science, 287, 1258-62. 159 Kohn, D.B., Weinberg, K.I., Nolta, J.A., Heiss, L.N., Lenarsky, C, Crooks, G.M., Hanley, M.E., Annett, G., Brooks, J.S., el-Khoureiy, A. and et al. (1995) Engraftment of gene-modified umbilical cord blood cells in neonates with adenosine deaminase deficiency. Nat Med, 1,1017-23. Kolb, A.F., Ansell, R., McWhir, J. and Siddell, S.G. (1999) Insertion of a foreign gene into the beta-casein locus by Cre-mediated siteyspecific recombination. Gene, 227, 21-31. Kowolik, CM., Hu, J. and Yee, J.K. (2001) Locus control region of the human CD2 gene in a lentivirus vector confers position-independent transgene expression. J Virol, 75, 4641-8. Kren, B.T., Bandyopadhyay, P. and Steer, CJ. (1998) In vivo site-directed mutagenesis of the factor IX gene by chimeric RNA/DNA oligonucleotides. Nat Med, 4, 285-90. LaFace, D., Hermonat, P., Wakeland, E. and Peck, A. (1988) Gene transfer into hematopoietic progenitor cells mediated by an adeno-associated virus vector. Virology, 162, 483-6. Lakso, M., Sauer, B., Mosinger, B., Jr., Lee, E.J., Manning, R.W., Yu, S.H., Mulder, K.L. and Westphal, H. (1992) Targeted oncogene activation by site-specific recombination in transgenic mice. Proc Natl Acad Sci USA, 89, 6232-6. Leboulch, P., Huang, G.M., Humphries, R.K., Oh, Y.H., Eaves, C.J., Tuan, D.Y. and London, I.M. (1994) Mutagenesis of retroviral vectors transducing human beta-globin gene and beta-globin locus control region derivatives results in stable transmission of an active transcriptional structure. Embo J, 13, 3065-76. Lemarchandel, V., Ghysdael, J., Mignotte, V., Rahuel, C. and Romeo, P.H. (1993) GATA and Ets cis-acting sequences mediate megakaryocyte-specific expression. Mol Cell Biol, 13, 668-76. Lemischka, I.R., Raulet, D.H. and Mulligan, R.C. (1986) Developmental potential and dynamic behavior of hematopoietic stem cells. Cell, 45, 917-27. Lewandoski, M. (2001) Mouse genomic technologies: conditional control of gene expression in the mouse. Nat Rev Genet, 2, 743-55. Li, L.H., McCarthy, P. and Hui, S.W. (2001) High-efficiency electrotransfection of human primary hematopoietic stem cells. Faseb J, 15, 586-8. Li, Q., Harju, S. and Peterson, K.R. (1999) Locus control regions: coming of age at a decade plus. Trends Genet, 15, 403-8. 160 Lindenbaum, M.H. and Grosveld, F. (1990) An in vitro globin gene switching model based on differentiated embryonic stem cells. Genes Dev, 4, 2075-85. Lozzio, C.B.a.L., B.B. (1975) Human chronic myelogenous leukemia cell line with positive Philadelphia chromosome, blood, 45, 312-317. Ludwig, D.L., Stringer, J.R., Wight, D.C., Doetschman, H.C. and Duffy, J.J. (1996) FLP-mediated site-specific recombination in microinjected murine zygotes. Transgenic Res, 5, 385-95. Malech, H.L., Maples, P.B., Whiting-Theobald, N., Linton, G.F., Sekhsaria, S., Vowells, S.J., Li, F., Miller, J.A., DeCarlo, E., Holland, S.M., Leitman, S.F., Carter, C.S., Butz, R.E., Read, E.J., Fleisher, T.A., Schneiderman, R.D., Van Epps, D.E., Spratt, S.K., Maack, C.A., Rokovich, J.A., Cohen, L.K. and Gallin, J.I. (1997) Prolonged production of NADPH oxidase-corrected granulocytes after gene therapy of chronic granulomatous disease. Proc Natl Acad Sci U S A, 94, 12133-8. Malik, P., McQuiston, S.A., Yu, X.J., Pepper, K.A., Krall, W.J., Podsakoff, G.M., Kurtzman, G.J. and Kohn, D.B. (1997) Recombinant adeno-associated virus mediates a high level of gene transfer but less efficient integration in the K562 human hematopoietic cell line. / Virol, 71, 1776-83. Mansour, S.L., Thomas, K.R. and Capecchi, M.R. (1988) Disruption of the proto-oncogene int-2 in mouse embryo-derived stem cells: a general strategy for targeting mutations to non-selectable genes. Nature, 336, 348-52. Markowitz, D., Goff, S. and Bank, A. (1988a) Construction and use of a safe and efficient amphotropic packaging cell line. Virology, 167,400-6. Markowitz, D., Goff, S. and Bank, A. (1988b) A safe packaging line for gene transfer: separating viral genes on two different plasmids. J Virol, 62, 1120-4. Martin, G.R. (1981) Isolation of a pluripotent cell line from early mouse embryos cultured in medium conditioned by teratocarcinoma stem cells. Proc Natl Acad Sci USA, 78,7634-8. May, C, Rivella, S., Callegari, J., Heller, G., Gaensler, K.M., Luzzatto, L. and Sadelain, M. (2000) Therapeutic haemoglobin synthesis in beta-thalassaemic mice expressing lentivirus-encoded human beta-globin. Nature, 406, 82-6. Mays, R.W., et al. (unpublished results). McCune, S.L., Reilly, M.P., Chomo, M.J., Asakura, T. and Townes, T.M. (1994) Recombinant human hemoglobins designed for gene therapy of sickle cell disease. Proc Natl Acad Sci US A, 91, 9852-6. 161 Medberry, S.L., Dale, E., Qin, M. and Ow, D.W. (1995) Intra-chromosomal rearrangements generated by Cre-lox site-specific recombination. Nucleic Acids Res, 23, 485-90. Medvinsky, A. and Dzierzak, E. (1996) Definitive hematopoiesis is autonomously initiated by the AGM region. Cell, 86, 897-906. Metcalf, D. (1984) The Hemopoietic Colony Stimulating Factors. Metcalf, D. (2001) Some General Aspects of Hematopoietic Cell Development. In Zon, G. (ed.) Hematopoiesis: A Development Approach. Oxford University Press, New York, Vol.36. Metcalf, D. and Moore, M.A.S. (1971) Hematopoietic cells, Amsterdam, North Holland. Mitani, K., Wakamiya, M., Hasty, P., Graham, F.L., Bradley, A. and Caskey, CT. (1995) Gene targeting in mouse embryonic stem cells with an adenoviral vector. Somat Cell Mol Genet, 21, 221-31. Miyoshi, H., Smith, K.A., Mosier, D.E., Verma, I.M. and Torbett, B.E. (1999) Transduction of human CD34+ cells that mediate long-term engraftment of NOD/SCID mice by HIV vectors. Science, 283, 682-6. Mochizuki, H., Schwartz, J.P., Tanaka, K., Brady, R.O. and Reiser, J. (1998) High-titer human immunodeficiency virus type 1-based vector systems for gene delivery into nondividing cells. J Virol, 72, 8873-83. Molete, J.M., Petrykowska, H., Bouhassira, E.E., Feng, Y.Q., Miller, W. and Hardison, R.C. (2001) Sequences flanking hypersensitive sites of the beta-globin locus control region are required for synergistic enhancement. Mol Cell Biol, 21, 2969-80. Monahan, P.E., Samulski, R.J., Tazelaar, J., Xiao, X., Nichols, T.C, Bellinger, D.A., Read, M.S. and Walsh, CE. (1998) Direct intramuscular injection with recombinant AAV vectors results in sustained expression in a dog model of hemophilia. Gene Ther, 5, 40-9. Morrison, S.J. (2001) Stem cell potential: can anything make anything? Curr Biol, 11, R7-9. Muller, A.M., Medvinsky, A., Strouboulis, J., Grosveld, F. and Dzierzak, E. (1994) Development of hematopoietic stem cell activity in the mouse embryo. Immunity, 1, 291-301. 162 Naldini, L., Blomer, U., Gallay, P., Ory, D., Mulligan, R., Gage, F.H., Verma, I.M. and Trono, D. (1996) In vivo gene delivery and stable transduction of nondividing cells by a lentiviral vector. Science, 272, 263-7. Neff, T., Shotkoski, F. and Stamatoyannopoulos, G. (1997) Stem cell gene therapy, position effects and chromatin insulators. Stem Cells, 15 Suppl l', 265-71. Novak, U., Harris, E.A., Forrester, W., Groudine, M. and Gelinas, R. (1990) High-level beta-globin expression after retroviral transfer of locus activation region-containing human beta-globin gene derivatives into murine erythroleukemia cells. Proc Natl Acad Sdt/S^, 87, 3386-90. O'Gorman, S., Fox, D.T. and Wahl, G.M. (1991) Recombinase-mediated gene activation and site-specific integration in mammalian cells. Science, 251, 1351-5. Okano, M., Xie, S. and Li, E. (1998) Cloning and characterization of a family of novel mammalian DNA (cytosine-5) methyltransferases. Nat Genet, 19, 219-20. Olivieri, N.F. (1999) The beta-thalassemias. N EnglJ Med, 341, 99-109. Orban, P.C., Chui, D. and Marth, J.D. (1992) Tissue- and site-specific DNA recombination in transgenic mice. Proc Natl Acad Sci USA, 89, 6861-5. Ory, D.S., Neugeboren, B.A. and Mulligan, R.C. (1996) A stable human-derived packaging cell line for production of high titer retrovirus/vesicular stomatitis virus G pseudotypes. Proc Natl Acad Sci USA, 93, 11400-6. Palacios, R., Golunski, E. and Samaridis, J. (1995) In vitro generation of hematopoietic stem cells from an embryonic stem cell line. Proc Natl Acad Sci USA, 92, 7530-4. Palacios, R. and Steinmetz, M. (1985) 11-3-dependent mouse clones that express B-220 surface antigen, contain Ig genes in germ-line configuration, and generate B lymphocytes in vivo. Cell, 41, 727-34. Pannell, D., Osborne, C.S., Yao, S., Sukonnik, T., Pasceri, P., Karaiskakis, A., Okano, M., Li, E., Lipshitz, H.D. and Ellis, J. (2000) Retrovirus vector silencing is de novo methylase independent and marked by a repressive histone code. Embo J, 19, 5884-94. Panzenbock, B., Bartunek, P., Mapara, M.Y. and Zenke, M. (1998) Growth and differentiation of human stem cell factor/erythropoietin-dependent erythroid progenitor cells in vitro. Blood, 92, 3658-68. Park, F., Ohashi, K., Chiu, W., Naldini, L. and Kay, M.A. (2000) Efficient lentiviral transduction of liver requires cell cycling in vivo. Nat Genet, 24, 49-52. 163 Pasceri, P., Pannell, D., Wu, X. and Ellis, J. (1998) Full activity from human beta-globin locus control region transgenes requires 5'HS1, distal beta-globin promoter, and 3' beta-globin sequences. Blood, 92, 653-63. Pawliuk, R., Eaves, CJ. and Humphries, R.K. (1997) Sustained high-level reconstitution of the hematopoietic system by preselected hematopoietic cells expressing a transduced cell-surface antigen. Hum Gene Ther, 8, 1595-604. Pawliuk, R., Kay, R., Lansdorp, P. and Humphries, R.K. (1994) Selection of retrovirally transduced hematopoietic cells using CD24 as a marker of gene transfer. Blood, 84, 2868-77. Pfeifer, A. and Verma, I.M. (2001) GENE THERAPY: Promises and Problems. Annu Rev Genomics Hum Genet, 2, 177-211. Pikaart, M.J., Recillas-Targa, F. and Felsenfeld, G. (1998) Loss of transcriptional activity of a transgene is accompanied by DNA methylation and histone deacetylation and is prevented by insulators. Genes Dev, 12, 2852-62. Plavec, I., Papayannopoulou, T., Maury, C. and Meyer, F. (1993) A human beta-globin gene fused to the human beta-globin locus control region is expressed at high levels in erythroid cells of mice engrafted with retro virus-transduced hematopoietic stem cells. Blood, 81, 1384-92. Prandini, M.H., Uzan, G., Martin, F., Thevenon, D. and Marguerie, G. (1992) Characterization of a specific erythromegakaryocytic enhancer within the glycoprotein lib promoter. JBiol Chem, 267, 10370-4. Prince, V.E. and Rigby, P.W. (1991) Derivatives of Moloney murine sarcoma virus capable of being transcribed in embryonal carcinoma stem cells have gained a functional Spl binding site. J Virol, 65, 1803-11. Raftopoulos, H., Ward, M., Leboulch, P. and Bank, A. (1997) Long-term transfer and expression of the human beta-globin gene in a mouse transplant model. Blood, 90, 3414-22. Recchia, A., Parks, R.J., Lamartina, S., Toniatti, C, Pieroni, L., Palombo, F., Ciliberto, G., Graham, F.L., Cortese, R., La Monica, N. and Colloca, S. (1999) Site-specific integration mediated by a hybrid adenovirus/adeno-associated virus vector. Proc Natl Acad Sci US A, 96, 2615-20. Rice, M.C., Czymmek, K. and Kmiec, E.B. (2001) The potential of nucleic acid repair in functional genomics. Nat Biotechnol, 19, 321-6. 164 Rivella, S., Callegari, J.A., May, C, Tan, C.W. and Sadelain, M. (2000) The cHS4 insulator increases the probability of retroviral expression at random chromosomal integration sites. J Virol, 74, 4679-87. Rivella, S. and Sadelain, M. (1998) Genetic treatment of severe hemoglobinopathies: the combat against transgene variegation and transgene silencing. Semin Hematol, 35, 112-25. Rodriguez, C.I., Buchholz, F., Galloway, J., Sequerra, R., Kasper, J., Ayala, R., Stewart, A.F. and Dymecki, S.M. (2000) High-efficiency deleter mice show that FLPe is an alternative to Cre-loxP. Nat Genet, 25, 139-40. Rubin, J.E., Pasceri, P., Wu, X., Leboulch, P. and Ellis, J. (2000) Locus control region activity by 5'HS3 requires a functional interaction with beta-globin gene regulatory elements: expression of novel beta/gamma-globin hybrid transgenes. Blood, 95, 3242-9. Russell, D.W. and Hirata, R.K. (1998) Human gene targeting by viral vectors. Nat Genet, 18,325-30. Russell, D.W. and Kay, MA. (1999) Adeno-associated virus vectors and hematology. Blood, 94, 864-74. Sadelain, M., Wang, C.H., Antoniou, M., Grosveld, F. and Mulligan, R.C. (1995) Generation of a high-titer retroviral vector capable of expressing high levels of the human beta-globin gene. Proc Natl Acad Sci US A, 92, 6728-32. Sambrook, J.F. and Maniatis, T. (1989) Molecular Cloning: a laboratory manual. Cold Spring Harbour Laboratory Press, Cold Spring Harbour, New York. Sauer, B. and Henderson, N. (1989) Cre-stimulated recombination at loxP-containing DNA sequences placed into the mammalian genome. Nucleic Acids Res, 17, 147-61. Sauer, B. and Henderson, N. (1990) Targeted insertion of exogenous DNA into the eukaryotic genome by the Cre recombinase. New Biol, 2, 441-9. Schubeler, D., Maass, K. and Bode, J. (1998) Retargeting of retroviral integration sites for the predictable expression of transgenes and the analysis of cis-acting sequences. Biochemistry, 37, 11907-14. Skarpidi, E., Vassilopoulos, G., Stamatoyannopoulos, G. and Li, Q. (1998) Comparison of expression of human globin genes transferred into mouse erythroleukemia cells and in transgenic mice. Blood, 92, 3416-21. 165 Smith, A.J., De Sousa, M.A., Kwabi-Addo, B., Heppell-Parton, A., Impey, H. and Rabbitts, P. (1995) A site-directed chromosomal translocation induced in embryonic stem cells by Cre-loxP recombination. Nat Genet, 9, 376-85. Smithies, O., Gregg, R.G., Boggs, S.S., Koralewski, M.A. and Kucherlapati, R.S. (1985) Insertion of DNA sequences into the human chromosomal beta-globin locus by homologous recombination. Nature, 317, 230-4. Snyder, R.O., Miao, C., Meuse, L., Tubb, J., Donahue, B.A., Lin, H.F., Stafford, D.W., Patel, S., Thompson, A.R., Nichols, T., Read, M.S., Bellinger, D.A., Brinkhous, K.M. and Kay, M.A. (1999) Correction of hemophilia B in canine and murine models using recombinant adeno-associated viral vectors. Nat Med, 5, 64-70. Soukharev, S., Miller, J.L. and Sauer, B. (1999) Segmental genomic replacement in embryonic stem cells by double lox targeting. Nucleic Acids Res, 27, e21. Stewart, C.L., Stuhlmann, H., Jahner, D. and Jaenisch, R. (1982) De novo methylation, expression, and infectivity of retroviral genomes introduced into embryonal carcinoma cells. Proc Natl Acad Sci USA, 79, 4098-102. Szilvassy, S.J., Humphries, R.K., Lansdorp, P.M., Eaves, A.C. and Eaves, C.J. (1990) Quantitative assay for totipotent reconstituting hematopoietic stem cells by a competitive repopulation strategy. Proc Natl Acad Sci USA, 87, 8736-40. Tani, T., Ylanne, J. and Virtanen, I. (1996) Expression of megakaryocyte and erythroid properties in human leukemic cells. Exp Hematol, 24, 158-68. Templeton, N.S., Roberts, D.D. and Safer, B. (1997) Efficient gene targeting in mouse embryonic stem cells. Gene Ther, 4, 700-9. Teshigawara, K. and Katsura, Y. (1992) A simple and efficient mammalian gene expression system using an EBV-based vector transfected by electroporation in G2/M phase. Nucleic Acids Res, 20, 2607. Thomas, K.R. and Capecchi, M.R. (1987) Site-directed mutagenesis by gene targeting in mouse embryo-derived stem cells. Cell, 51, 503-12. Thyagarajan, B., Guimaraes, M.J., Groth, A.C. and Calos, M.P. (2000) Mammalian genomes contain active recombinase recognition sites. Gene, 244, 47-54. Thyagarajan, B., Olivares, E.C., Hollis, R.P:, Ginsburg, D.S. and Calos, M.P. (2001) Site-specific genomic integration in mammalian cells mediated by phage phiC31 integrase. Mol Cell Biol, 21,3926-34. 166 Till, J.E. and McCulloch, E.A. (1961) A direct measurement of the radiation sensitivity of normal mouse bone marrow cells. Radiat. Res., 14, 213-222. Toneguzzo, F. and Keating, A. (1986) Stable expression of selectable genes introduced into human hematopoietic stem cells by electric field-mediated DNA transfer. Proc Natl Acad Sci US A, 83, 3496-9. Tsai, M., Wedemeyer, J., Ganiatsas, S., Tarn, S.Y., Zon, L.I. and Galli, S.J. (2000) In vivo immunological function of mast cells derived from embryonic stem cells: an approach for the rapid analysis of even embryonic lethal mutations in adult mice in vivo. Proc Natl Acad Sci USA, 97, 9186-90. Tuan, D. and London, I.M. (1984) Mapping of DNase I-hypersensitive sites in the upstream DNA of human embryonic epsilon-globin gene in K562 leukemia cells. Proc Natl Acad Sci USA, 81,2718-22. Uchida, N., Sutton, R.E., Friera, A.M., He, D., Reitsma, M.J., Chang, W.C, Veres, G., Scollay, R. and Weissman, LL. (1998) HLV, but not murine leukemia virus, vectors mediate high efficiency gene transfer into freshly isolated G0/G1 human hematopoietic stem cells. Proc Natl Acad Sci USA, 95, 11939-44. van der Steege, G., Schuilenga-Hut, P.H., Buys, C.H., Scheffer, H., Pas, H.H. and Jonkman, M.F. (2001) Persistent failures in gene repair. Nat Biotechnol, 19, 305-6. van Os, R., Avraham, H., Banu, N., Mauch, P.M., Whater, J., Yang, Y. and Du, B. (1999) Recombinant adeno-associated virus-based vectors provide short-term rather than long-term transduction of primitive hematopoietic stem cells. Stem Cells, 17, 117-20. Van Tendeloo, V.F., Van Broeckhoven, C. and Berneman, Z.N. (2001) Gene therapy: principles and applications to hematopoietic cells. Leukemia, 15, 523-44. Van Tendeloo, V.F., Willems, R., Ponsaerts, P., Lenjou, M., Nijs, G., Vanhove, M., Muylaert, P., Van Cauwelaert, P., Van Broeckhoven, C, Van Bockstaele, D.R. and Berneman, Z.N. (2000) High-level transgene expression in primary human T lymphocytes and adult bone marrow CD34+ cells via electroporation-mediated gene delivery. Gene Ther,l, 1431-7. Vasquez, K.M., Marburger, K., Intody, Z. and Wilson, J.H. (2001) Manipulating the mammalian genome by homologous recombination. Proc Natl Acad Sci USA, 98, 8403-10. Vasquez, K.M., Narayanan, L. and Glazer, P.M. (2000) Specific mutations induced by triplex-forming oligonucleotides in mice. Science, 290, 530-3. 167 Verhoeyen, E., Hauser, H. and Wirth, D. (2001) Evaluation of retroviral vector design in defined chromosomal loci by ftp-mediated cassette replacement. Hum Gene Ther, 12, 933-44. Verma, I.M. and Somia, N. (1997) Gene therapy ~ promises, problems and prospects. Nature, 389, 239-42. Vigna, E. and Naldini, L. (2000) Lentiviral vectors: excellent tools for experimental gene transfer and promising candidates for gene therapy. J Gene Med, 2, 308-16. Wallace, H., Ansell, R., Clark, J. and McWhir, J. (2000) Pre-selection of integration sites imparts repeatable transgene expression. Nucleic Acids Res, 28, 1455-64. Wang, Q. and Taylor, M.W. (1993) Correction of a deletion mutant by gene targeting with an adenovirus vector. Mol Cell Biol, 13, 918-27. Weatherall, D.J. (1998) Pathophysiology of thalassaemia. Baillieres Clin Haematol, 11, 127-46. Weatherall, D.J. (2001) Phenotype-genotype relationships in monogenic disease: lessons from the thalassaemias. Nat Rev Genet, 2, 245-55. Weiss, M.J. and Orkin, S.H. (1996) In vitro differentiation of murine embryonic stem cells. New approaches to old problems. J Clin Invest, 97, 591-5. Whitelaw, E., et al. (2001) Epigenetic Effects on Transgene Expression. In Kola, M.J.T.a.I. (ed.) Methods in Molecular Biology. Humana Press Inc., Totowa, NJ, Vol. 158. Wiles, M.V. and Keller, G. (1991) Multiple hematopoietic lineages develop from embryonic stem (ES) cells in culture. Development, 111, 259-67. Willard, H.F. (2000) Genomics and gene therapy. Artificial chromosomes coming to life. Science, 290, 1308-9. Williams, R.L., Hilton, D.J., Pease, S., Willson, T.A., Stewart, C.L., Gearing, D.P., Wagner, E.F., Metcalf, D., Nicola, NA. and Gough, N.M. (1988) Myeloid leukaemia inhibitory factor maintains the developmental potential of embryonic stem cells. Nature, 336, 684-7. Williams, S.R., Ousley, F.C., Vitez, L.J. and DuBridge, R.B. (1994) Rapid detection of homologous recombinants in nontransformed human cells. Proc Natl Acad Sci US A, 91, 11943-7. 168 Wu, A.M., Siminovitch, L., Till, J.E. and McCulloch, E.A. (1968) Evidence for a relationship between mouse hemopoietic stem cells and cells forming colonies in culture. Proc Natl Acad Sci U S A, 59, 1209-15. Wu, M.H., Liebowitz, D.N., Smith, S.L., Williams, S.F. and Dolan, M.E. (2001a) Efficient expression of foreign genes in human CD34(+) hematopoietic precursor cells using electroporation. Gene Ther, 8, 384-90. Wu, M.H., Smith, S.L., Danet, G.H., Lin, A.M., Williams, S.F., Liebowitz, D.N. and Dolan, M.E. (2001b) Optimization of culture conditions to enhance transfection of human CD34+ cells by electroporation. Bone Marrow Transplant, 21, 1201-9. Xiang, Y., Cole-Strauss, A., Yoon, K., Gryn, J. and Kmiec, E.B. (1997) Targeted gene conversion in a mammalian CD34+-enriched cell population using a chimeric RNA/DNA oligonucleotide. J Mol Med, 75, 829-35. Yanez, R.J. and Porter, A.C. (1998) Therapeutic gene targeting. Gene Ther, 5, 149-59. Yoder, M.C., Hiatt, K., Dutt, P., Mukherjee, P., Bodine, D.M. and Orlic, D. (1997) Characterization of definitive lymphohematopoietic stem cells in the day 9 murine yolk sac. Immunity, 1, 335-44. Zheng, B., Sage, M., Sheppeard, E.A., Jurecic, V. and Bradley, A. (2000a) Engineering mouse chromosomes with Cre-loxP: range, efficiency, and somatic applications. Mol Cell Biol, 20, 648-55. Zheng, C, Baum, B.J., Iadarola, M.J. and O'Connell, B.C. (2000b) Genomic integration and gene expression by a modified adenoviral vector. Nat Biotechnol, 18, 176-80. Zhou, S., Murphy, J.E., Escobedo, J.A. and Dwarki, V.J. (1998) Adeno-associated virus-mediated delivery of erythropoietin leads to sustained elevation of hematocrit in nonhuman primates. Gene Ther, 5, 665-70. Zhou, S.Z., Broxmeyer, H.E., Cooper, S., Harrington, M.A. and Srivastava, A. (1993) Adeno-associated virus 2-mediated gene transfer in murine hematopoietic progenitor cells. Exp Hematol, 21, 928-33. 169 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

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"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0090532/manifest

Comment

Related Items