UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The role of SHIP in hematopoiesis Hughes, Michael R. 2005

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

Item Metadata


831-ubc_2005-10404X.pdf [ 17.52MB ]
JSON: 831-1.0092241.json
JSON-LD: 831-1.0092241-ld.json
RDF/XML (Pretty): 831-1.0092241-rdf.xml
RDF/JSON: 831-1.0092241-rdf.json
Turtle: 831-1.0092241-turtle.txt
N-Triples: 831-1.0092241-rdf-ntriples.txt
Original Record: 831-1.0092241-source.json
Full Text

Full Text

THE ROLE OF SHIP IN HEMATOPOIESIS by MICHAEL R. HUGHES B.Sc, The University of Victoria, 1995 A THESIS SUBMITTED TN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY In THE FACULTY OF GRADUATE STUDIES (Experimental Medicine) THE UNIVERSITY OF BRITISH COLUMBIA March 2005 © Michael R. Hughes, 2005 ABSTRACT The SH2-containing inositol-5'-phosphatase, SHIP, has been shown, by comparing SHTP+/+ and -/- cells, to be. a critical negative regulator of the phosphatidylinositol (PI)-3-kinase (PI3K) pathway in many hematopoietic cell lineages. However, SHIP-/- mice are neither polycythemic nor profoundly anemic and we wanted to know why. Our studies revealed that SHIP-/- mice suffer from reticulocytosis and show an enhanced recovery from phenylhydrazine (PHz)-induced anemia accompanied by a more rapid accumulation of CFU-E and late erythroid progenitors in the spleens of SHIP-/- mice. In addition, following PHz treatment, SHIP-/- plasma possesses higher erythroid growth-stimulating activity than SHIP+/+ plasma. Interestingly, however, erythropoietin (Epo) dose-response studies revealed little difference between SHIP+/-I- and -/- progenitors and this may be because perturbed erythropoiesis in SHIP-/- mice is caused by extrinsic mechanisms. Although SHIP becomes tyrosine phosphorylated in response to multiple stimuli, the kinase(s) responsible has not been identified. Using phosphospecific antibodies we developed, we found that the two NPXYs within SHIP are the major sites of tyrosine phosphorylation in response to cytokines, growth factors, G protein-coupled receptor ligands, immunoreceptor ligands and osmotic stress. Moreover, using the Src family inhibitor, PP2 as well as cells from Lyn-/- mice, we found that the Src family is primarily responsible for SHIP'S tyrosine phosphorylation. Moreover, consistent with this being a direct effect, we found SHIP and Lyn associate in B-cells and this association increases with BCR + FcyRITB aggregation. Interestingly, Lyn-SHIP association appears to be mediated via Lyn's SH3 and a previously unidentified PXXP motif proximal to SHIP'S SH2 domain. Lastly, we investigated the role of SHIP in human hematopoietic cells using small interfering RNA to silence SHIP expression in the erythroleukemic cell line, TF-1. We demonstrated complete and long-lasting knockdown of SHIP protein in these cells and this resulted in a more rapid phosphorylation of Akt, GSK3|3 and ERK1/2 in response to GM-CSF. Furthermore, SHIP-deficient TF-1 cells survived better with low or no cytokines but displayed reduced proliferation at high cytokine concentrations. Further investigation into the enhanced survival of SHIP-knockdown TF-1 cells suggested that this was mediated in part by the PI3K-induced maintenance of the pro-survival Bcl-2 member, Mcl-1. ii TABLE OF CONTENTS ABSTRACT ii TABLE OF CONTENTS iv LIST OF FIGURES vii LIST OF TABLES x LIST OF ABBREVIATIONS xi CONTRIBUTIONS OF OTHERS xix ACKNOWLEDGEMENTS xx CHAPTER 1: INTRODUCTION 1 1.1 HEMATOPOBESIS 1 1.1.1 The Hematopoietic stem cell 2 1.1.2 Hematopoietic development 3 1.1.3 Regulation of hematopoiesis 3 1.1.4 Hierarchy of blood cells and progenitors 6 1.1.5 Functions of mature blood cells 7 1.2 ERYTHROPOIESIS 11 1.2.1 Erythroid developmental stages 11 1.2.2 Regulation of erythropoiesis 14 1.2.3 Erythropoietin 16 1.2.4 Erythropoietin receptor structure and function 17 1.2.5 Transcription factors involved in erythropoiesis 23 1.3 PROTEIN BINDING DOMAINS 25 1.4 REGULATION OF SRC-FAMILY KINASE ACTIVHY 28 1.5 THE PI3K PATHWAY 33 1.5.1 SHIP 36 1.5.2 SHIP2 39 1.5.3 PTEN 39 1.6 THE PHENOTYPE OF SHIP KNOCKOUT MICE 40 1.7 SHIP AND HUMAN DISEASE 42 1.8 AIMS OF STUDY 43 CHAPTER 2: MATERIALS AND METHODS 44 2.1 TISSUE CULTURE 44 2.1.1 Immortal cell lines 44 iii 2.1.2 Bone marrow-derived mast cells (BMMCs) 44 2.1.3 Bone marrow-derived macrophages (BMM<ps) 45 2.2 siRNA TRANSFECTION 45 2.3 ASSESSMENT OF ERYTHROPOIESIS WHHIN SHIP+/+ AND -/- MICE 46 2.3.1 Mice 46 2.3.2 Phenylhydrazine-induced anemia 46 2.2.3 Determination of hematocrit and reticulocyte index 46 2.3.4 , Plasma collection 47 2.3.5 Cell isolation and nucleated cell counts 47 2.3.6 Progenitor analysis 47 2.3.7 CD71+purification 49 2.3.8 Murine hematopoietic enrichment (Lin" selection) 49 2.4 PROTEIN ANALYSIS 50 2.4.1 Cell stimulation, immunoprecipitations, and immunoblotting 50 2.4.2 Inhibitors and blocking antibodies 51 2.4.3 Antibodies 52 2.5 FLOW CYTOMETRY AND FACS 54 2.6 BIOLOGICAL ANALYSIS OF CELLS 55 2.6.1 Survival and differentiation studies 55 2.6.2 3H-thymidine assays 56 2.7 SHIP MUTANTS AND PHOSPHO-SHIP ANTIBODIES 56 2.8 BINDING OF SHIP TO PHOSPHORYLATED TTIM BEADS 57 2.9 GLUTATHIONE S-TRANSFERASE (GST) FUSION PROTEINS. 57 2.9.1 GST Fusion Protein Constructs 57 2.9.2 In vitro Protein Binding Assay 58 CHAPTER 3: CHARACTERIZATION OF ERYTHROPOIESIS IN SHIP -/- MICE 59 3.1 INTRODUCTION 59 3.2 RESULTS 60 3.2.1 SHIP-/- mice are only mildly anemic but display a marked reticulocytosis 60 3.2.2 SHIP-/- mice recover more rapidly from PHz-induced anemia... 64 3.2.3 Enhanced recruitment of erythroid progenitors to the spleen of SHIP-/- mice following PHz treatment 66 3.2.4 CFU-E accumulate more rapidly in PHz-treated SHIP-/- mice.... 72 3.2.5 Epo-responsiveness of erythroid progenitors from the spleens of SHIP+/+ and -/- mice following PHz-treatment 78 i v 3.2.6 Epo-responsiveness of SHIP+/+ and -/- BFU-E and CFU-E 86 3.2.7 Epo-induced tyrosine phosphorylations in Lin" cells from Day 3 spleens of PHz-treated mice 89 3.2.8 CD71 h ' s TER1 i 9 m e d / l o w erythroblasts from SHJP+/+ and -/-mice express similar EpoR, STAT5a and GATA-1 levels 92 3.2.9 SHIP expression is normally lost during erythropoiesis 96 3.2.10 Enhanced induction of erythropoietic growth promoting activity in the plasma of SHIP-/- mice 98 3.3 DISCUSSION 102 CHAPTER 4: SHIP IS TYROSINE PHOSPHORYLATED BY SRC FAMILY KINASES IN RESPONSE TO MULTIPLE STIMULI 110 4.1 INTRODUCTION 110 4.2 RESULTS I l l 4.2.1 The two NPXY motifs in SHIP are tyrosine phosphorylated in response to multiple stimuli I l l 4.2.2 The Src family inhibitor PP2 blocks the tyrosine phosphorylation of SHIP, regardless of the extracellular stimulus 115 4.2.3 SHIP phosphorylation is reduced in Lyn-/- BMMCs in response to multiple stimuli 121 4.2.4 Lyn associates with SHIP in BJAB cells and this is not blocked by PP2 124 4.2.5 SHIP associates, via a previously unrecognized proline-rich motif, with the SH3 domain of Lyn 129 4.2.6 PP2 does not block BCR internalization 131 4.2.7 The tyrosine phosphorylation of SHIP reduces its affinity for the phosphorylated ITIM of the negative co-receptor, FcyRIIB 135 4.3 DISCUSSION 137 CHAPTER 5: SHIP-DEFICIENT TF-1 CELLS DISPLAY CYTOKINE INDEPENDENT GROWTH/SURVIVAL BUT REDUCED RESPONSIVENESS TO HIGH CYTOKINE L E V E L 147 5.1 INTRODUCTION 147 5.2 RESULTS 148 5.2.1 Silencing human SHIP expression in TF-1 cells 148 5.2.2 Akt, GSK3 and ERK phosphorylation are enhanced in SHIP deficient TF-1 cells 150 5.2.3 SHIP-deficient TF-1 cells proliferate more at low but less at high cytokine concentrations 152 5.2.4 SHIP deficient TF-1 cells survive longer in the absence of cytokines 154 5.2.5 The survival of SHIP-deficient TF-1 cells is dependent on PI3K.... 156 5.2.6 SHIP-deficient TF-1 cells display elevated Mcl-1 levels upon v starvation 159 5.3 DISCUSSION 161 CHAPTER 6: SUMMARY 163 REFERENCES 167 vi LIST OF FIGURES CHAPTER 1 Figure 1.1 The hierarchical model of hematopoiesis 9 Figure 1.2 Stages of erythroid differentiation 13 Figure 1.3 Features of the cytokine receptor superfamily 18 Figure 1.4 Regulation of Src-family kinases by intra- and inter-molecular interactions 32 Figure 1.5 The structure of PIP3, its precursors and products 34 Figure 1.6 Functional domains of SHIP, SHIP2 and PTEN 38 CHAPTER 2 No figures CHAPTER 3 Figure 3.1 SHIP-/- mice have a higher spleen index, are slightly anemic and display reticulocytosis 62 Figure 3.2 SHIP-/- mice display a more rapid recovery after PHz treatment 65 Figure 3.3 The spleens of SHIP-/- mice rapidly increase in size and cellularity in response to PHz-induced anemia 67 Figure 3.4 CD71TER119 expression profiles of SHIP+/+ and SHIP-/- following PHz treatment 70 Figure 3.5 SHIP-/- mice have higher total CFU-E and BFU-E than SHIP+/-I- mice. PHz-treated SHIP-/- mice rapidly accumulated CFU-E in the spleen 74 Figure 3.6 Ammonium chloride lysis of spleen cells prepared from Day 3 PHz treated mice enriches for Epo-responsive cells... 80 Figure 3.7 SHIP-I-/+ and SHIP-/- lineage-depleted (Lin) Day 3 PHz splenocytes are equally responsive to increasing concentrations of Epo 82 Figure 3.8 SHIP+/+ and SHIP-/- CFU-E are equally responsive to Epo 88 Figure 3.9 Epo-induced signaling events in SHrP+/+ and SHIP-/- Lin" Day 3 PHz spleen cells 90 vii Figure 3.10 SHIP-/- and SHIP+/+ erythroblasts isolated by FACS express similar levels of EpoR, GATA-1 and STAT5a 94 Figure 3.11 Late TER119+ erythroblasts do not express SHIP 97 Figure 3.12 Enhanced induction of erythropoietic growth-promoting activity in the plasma of SHIP-/- mice 99 C H A P T E R 4 Figure 4.1 Generation and characterization of phospho-specific antibodies to the two NPXpY motifs of SHIP 113 Figure 4.2 PP2 inhibits SHIP tyrosine phosphorylation in response to a variety of stimuli 117 Figure 4.3 SHIP tyrosine phosphorylation in Lyn-/- BMMC is reduced compared to Lyn+/+ BMMCs in response to a variety of stimuli 122 Figure 4.4 SHIP phosphorylation is not required for Lyn association 126 Figure 4.5 PP2 does not inhibit internalization of the BCR in BJAB cells stimulated with intact a-IgM 130 Figure 4.6 The N-terminal PxxP sequence of SHIP binds Lyn 132 Figure 4.7 Tyrosine phosphorylated SHIP has a reduced affinity for pITIM beads 136 Figure 4.8 SHIP associates with the SH3 domain of Lyn 138 Figure 4.9 Model for interactions of SHIP, Lyn, She andFcyRUB 143 C H A P T E R 5 Figure 5.1 TF-1 cells transfected with siRNA to human SHIP (siSHIP) or with a non-silencing control siRNA (siNS) 149 Figure 5.2 GM-CSF stimulation of TF-1 cells transfected with siSHIP display altered signaling 151 Figure 5.3 TF-1 cells transfected with siSHIP display enhanced proliferation in the absence of cytokines 153 Figure 5.4 TF-1 cell survival is enhanced by SHIP knockdown 155 viii Figure 5.5 Cytokine-independent survival of SHIP-deficient TF-1 cells requires active PI3K 157 Figure 5.6 Pro-apoptotic proteins are reduced in siSHIP transfected TF-1 cells 160 CHAPTER 6 No figures ix CHAPTER 2 Table 2.1 LIST OF TABLES List of antibodies used in this thesis LIST OF ABBREVIATIONS -/- null genotype +/+ wild-type genotype 2NPXF double Y to F point mutation in SHIP amino acid sequence 3H-Tdr tritiated thymidine 4G10 pan-specific anti-phosphotyrosine antibody A alanine AA arachidonic acid Abs antibodies Ag antigen A G M aorta-gonad-mesonephros Akt also known as PKB A L L acute lymphoblastic leukemia A M L acute myelogenous leukemia ATP adenosine triphosphate BaER Ba/F3 cell line with ectopically expressed EpoR Bcl-XL B cell lymphoma X (large) BCR B cell receptor BFU-E burst forming unit-erythroid BJAB mature human B cell line B M bone marrow BMMC bone marrow-derived mast cell SFFV spleen focus-forming virus BMmO bone marrow-derived macrophage BSA bovine serum albumin Btk Bruton's tyrosine kinase C cysteine C- carboxy c.d.f. cumulative distribution function CI phorbol ester/diacylglycerol binding domain xi C2 calcium dependent/independent phospholipids binding domains Ca 2 + calcium CBP CREB-binding protein CD150 also known as SLAM or CDw 150 CD16 low affinity Fc receptor (FcyRIIIa), a major activating receptor on NK cells CD22 FTIM containing B cell immunoreceptor CD28 T cell co-receptor CD71 transferrin receptor CD79a/b Signaling subunits of BCR complex (Iga (CD79a) and Igp (CD79b)) CDK cyclin dependent kinase CDKI cyclin dependent kinase inhibitor cDNA complementary deoxyribonucleic acid CFU-E colony forming unit-erythroid CFU-GM colony forming unit-granulocyte/macrophage CFU-Mk colony forming unit-megakaryocyte CFU-S colony forming unit-spleen CLP(s) common lymphoid progenitors CML chronic myelogenous leukemia CMP(s) common myeloid progenitors cPKC classical protein kinase C cpm counts per minute CREB cAMP-response element-binding protein CSF colony stimulating factor Csk C-terminal Src kinase D aspartic acid DAG diacylglycerol DMSO dimethylsulfoxide DNA deoxyribonucleic acid DNP-HSA dinitrophenyl-human serum albumin Dok docking protein xii E. coli Eschericia coli EDTA ethylenediaminetetraacetic acid EGF epidermal growth factor EGTA ethyleneglycol-bis ((3-aminoethyl ether) N,N,N,N'tetraacetate E K L F erythroid Kruppel-like factor ELISA enzyme-linked immunosorbant assay Epo erythropoietin EpoR erythropoietin receptor ER endoplasmic reticulum ERK1/2 extracellular regulated kinase (also called p44 (ERK1) or p42 (ERK2) MAPK) ES embryonic stem F phenylalanine F(ab')2 fragment antigen binding FACS fluorescent activated cell sorting FasL Fas ligand Fc fragment crystalline FceRI Fc epsilon receptor I FcyRIIB Fc gamma receptor IIB FcR receptor for Fc portion of immunoglobulins FCS fetal calf serum FITC fluorescein isothiocyanate FKHRL-1 forkhead (Drosophila) homolog (rhabdomyosarcoma) like 1 FN fibronectin FOG-1 Friend of GATA FynT Fyn isoform expressed in T cells G glycine Gab 1/2 Grb2 associated binding protein 1/2 GAP GTPase activating protein GAPDH glyceraldehyde-3-phosphate dehydrogenase GAS INF-gamma activated sequence xiii GC glucocorticoid G-CSF granulocyte-colony stimulating factor GFP green fluorescent protein GH growth hormone G M granulocyte and macrophage GM-CSF granulocyte-macrophage-colony stimulating factor GPCR G protein-coupled receptor (also called serpentine receptor) GSK-3 glycogen synthase kinase-3 GST ^ glutathione s-transferase HA hemagglutinin Hb hemoglobin HBSS Hank's balanced salt solution Hct hematocrit HFN HBSS (with fetal calf serum and sodium azide) HMBA hexamethylene bisacetamide Hr hour HRP horseradish peroxidase HSCs hematopoietic stem cells hSHIP human SHIP IFNy interferon-gamma Ig immunoglobulin IgE immunoglobulin E IgG immunoglobulin G IL- interleukin- (eg. IL-3 or interleukin-3) IL-3RP P subunit of IL-3 receptor IMDM Iscove's modified Dulbecco's medium IP immunoprecipitation IP4 inositol-1,3,4,5-tetrakisphosphate IRS insulin receptor substrate ITAMs immunoreceptor tyrosine based activation motifs ITIMs immunoreceptor tyrosine based inhibitory motifs xiv JAK JNK Kb kDa LAT Lin Lin" Lin* L M 0 2 LPS MAPK Mcl-1 M - C S F M E L cells M g 2 + MHC Min mRNA MSP M T G NaN 3 NFKB NH4CI NK NP-40 NPXY NPXY1 Janus kinase c-Jun N-terminal kinase kilobase kilodalton linker for activation of T cells hematopoietic lineage cell-surface markers CD5 (Ly-1), Erythroid Cells (TER119), CD45R (B220), Ly-6G (Gr-1), CDl lb (Mac-1), Neutrophils (7-4) lineage-depleted (cells) (depleted by immunomagnetic negative selection) hematopoietic lineage cell-surface markers without TER119 LEVI domain only 2 lipopolysaccharide mitogen activated protein kinase myeloid cell leukemia sequence 1 macrophage-colony stimulating factor murine erythroleukemic cells magnesium major histocompatibility complex minute messenger ribonucleic acid macrophage-stimulating protein monothioglycerol sodium azide nuclear factor KB ammonium chloride natural killer nonidet P-40 PTB-binding domain core sequence PTB binding consensus of SHIP conserved in mouse and human (at Y915 in hSHIP) xv NPXY2 P P.I. P/S P1C1 PAGE PARP-1 PBS Pet PDGF PDK1 PDZ PE PG PGE 2 PH PHz PI PI3K PIP3 pITIM PKB PKC PLC PP2 (or P2) PP3 (or P3) pS PSB PTB binding consensus of SHIP (at Y1022 in hSHIP) proline propidium iodide penicillin (100 U/ml) and streptomycin (100 pg/ml) anti-SHIP monoclonal antibody polyacrylamide gel electrophoresis poly (ADP-ribose) polymerase phosphate buffered saline piceatannol (pharmacological inhibitor of Syk) (3,4,3',5'-Tetrahydroxy-trans-stilbene) platelet-derived growth factor phosphoinositide-dependent protein kinase 1 protein binding domain (binds C-terminal isoleucine or valine) phycoerythrin prostaglandin prostaglandin E2 pleckstrin homology phenylhydrazine phosphatidylinositol phosphatidylinositol 3-kinase PI-3,4,5-trisphosphate tyrosine phosphorylated ITLVI motif of FcyRIIB protein kinase B protein kinase C phospholipase C pharmacological inhibitor of Src family kinases (4-Amino-5-(4-chlorophenyl)-7-(r-butyl)pyrazolo[3,4-d]pyrimidine) inactive chemical analog of PP2 (4-Amino-7-phenylpyrazol[3,4-d]pyrimidine) phosphoserine phosphorylation solubilization buffer xvi pT phosphothreonine PTB phosphotyrosine binding PTEN phosphatase and tensin homolog deleted on chromosome 10 PTK protein tyrosine kinase PTP protein tyrosine phosphatase PTP protein tyrosine phosphatase PVDF polyvinylidene fluoride PXXP proline-X-X-proline (where X= any amino acid) pY (or pTyr) phosphotyrosine R arginine RBCs red blood cells RON recepteur d-origine nantais (human) macrophage stimulating protein 1 receptor (c-met-related tyrosine kinase) RTK receptor tyrosine kinase SA streptavidin SAP SLAM-associated protein SCF stem cell factor (or c-kit ligand) SCL stem cell leukemia hematopoietic transcription factor SCN severe congenital neutropenia SD standard deviation SDF-la stromal cell-derived factor-1 alpha SDS sodium dodecyl sulphate SEM standard error measure SFK Src family kinase SH2 Src-homology 2 SH3 Src-homology 3 She Src homology and collagen SHIP SH2-containing inositol 5'-phosphatase shRNA short hairpin RNA siNS non-silencing siRNA xvii siRNA small interfering RNA siSHIP SHIP expression silencing siRNA SLAM signaling lymphocyte activation molecule sSHIP stem cell SHIP STAT signal transducer and activator of transcription STK stem cell-derived tyrosine kinase, mouse form of RON Syk spleen tyrosine kinase T threonine TBST tris-buffered saline (with Tween-20) T C L total cell lysate TCR T cell receptor T c T L cytotoxic T cell TER119 erythroid lineage-specific surface marker (also Ly-76) TF-1 tri-factor dependent cell line (human erythroleukemic cell line) TFs transcription factors TGFp transforming growth factor P TH1 cell type I helper T cell TH2 cell type II helper T cell TLR toll like receptor TNFa tumor necrosis factor-alpha Tpo thrombopoietin TX-100 Triton X-100 WCL whole cell lysate WEHI/WEHI231 immature mouse B cell line wsxws amino acid sequence motif conserved in hematopoietin receptoi superfamily (W=tryptophan, S=serine, X=any amino acid) X any amino acid Y tyrosine AFCS (or dFCS) heat-inactivated fetal calf serum u€i microCurie (3-ME p-mercaptoethanol (also called 2-mercaptoethanol or 2-ME) xviii CONTRIBUTIONS OF OTHERS The completion of this thesis would not be possible without the contributions of reagents and data from many friends and colleagues. I would like to acknowledge the assistance of Frann Antignano in Chapter 3 including the breeding of SHIP+/- mice and her help in many, many 20 hour days required to complete these experiments. Thanks also to Dr. Jackie Damen (StemCell Technologies, Inc.) for her assistance in the design and interpretation of the CFC assays presented in Chapter 3. I would like to acknowledge the contributions of Dr. Mark Ware for the initiation of the studies described in Chapter 4 and the generation of data presented in Figure 4.6 therein. In addition, thanks also to Dr. Jackie Damen and Dr. Mark Ware for the construction of SHIP wild-type and 2NPXF SHIP mutants (and the retroviral infection of SHIP-/- bone marrow mast cells) used to characterize the phospho-SHD? antibodies described in Figure 4.1. Thanks also to Dr. Li-Ping Cao and Vivian Lam for the preparation and affinity purification of the phospho-SHIP antibodies. In addition, I would like to thank Dr. Cao for the generation of pITIM binding study data presented in Figure 4.7. The bone marrow mast cells derived from Lyn+/+ and Lyn-/- mice used in Chapter 4 were kindly provided by Dr. Janet M . Oliver of the University of New Mexico (Albuquerque, NM). Vivian Lam was instrumental in optimizing the siRNA transfection studies in Chapter 5 (and many attempts with anti-sense RNA techniques that are not presented in this thesis). Frann Antignano was also of great assistance in the generation of data presented in Chapter 5. Finally, I have had the pleasure working with several dedicated co-op and summer students during the completion of this thesis. I would like to acknowledge the assistance of Matthew James, Keith Tsang, Ambrose Wong, Wency Ip and Stephanie Omeis for many hours of enthusiastic work. x i x ACKNOWLEDGEMENTS Dr. Gerry Krystal is a research supervisor that is wholly dedicated to the well-being and development of his students. He has provided to me not only a thriving work environment but has guided my work through many rough patches. I would like to thank him for his patience, support and mentorship over many years. Also, I would like to thank the support and encouragement of my family, and particularly my brother James for getting me out of the lab to do those important things in life - hockey! I have also had the pleasure of working with many fine friends that understand the stresses of research and provide an outlet for these frustrations. Drs Mark Ware, Jackie Damen and Janet Kalesnikoff in the early years provided an example of dedication that was useful in the final years. Dr. Laura Sly, Michael Rauh, Frann Antignano and Jens Ruschmann continue to set those high standards in our lab today. In particular, the self-less dedication of Frann Antignano and her assistance in the completion of this thesis were invaluable. xx Chapter 1 INTRODUCTION 1.1 HEMATOPOIESIS The modern era of hematopoiesis, the process of blood cell formation (Smith 2003), has its roots in the Manhattan Project at the University of Chicago. As an outgrowth of war-time projects studying the effects of chemical agents and ionizing radiation on blood and bone marrow, L.O. Jacobson and colleagues discovered that shielding the mouse spleen from radiation while administering a lethal dose of X-rays prevented death due to bone marrow aplasia (Jacobson, Marks et al. 1949). Curiously, survival was achieved even if the shielded spleen was removed shortly after irradiation or if spleen cells from a non-irradiated mouse were injected into a non-shielded lethally irradiated recipient. Jacobson initially credited recovery from radiation injury to unknown humoral factors (perhaps produced by the protected spleen), largely based on the observation that rabbit spleen cells were also an effective saviour of an irradiated mouse. He argued that the rabbit spleen cells would be destroyed by the immune system of the mouse and thus humoral factors were a more likely source for radioprotection than a cell-mediated mechanism - a conclusion made at a time when the nature of immunity was poorly understood. Subsequent studies by Jacobson (Jacobson, Simmons et al. 1951) and others (Lorenz, Uphoff et al. 1951) revealed that populations of spleen or marrow cells are responsible for marrow repopulation and the rescue of mice from radiation injury. Further animal experiments showed that radioprotection was achieved by a small population of transplantable cells (Main and Prehn 1955; Ford, Hamerton et al. 1956; Nowell, Cole et al. 1956). These studies laid the groundwork for subsequent experimentation that resulted in our current understanding of hematopoiesis and hematopoietic stem cells (HSCs). This knowledge has yielded innovative treatments for human diseases such as leukemias (Thomas, Storb et al. 1975) and immunological disorders, the development of therapeutic bone marrow transplantation in humans and, most recently, the emerging promise of regenerative medicine (Eaves 2003). 1 1.1.1 The Hematopoietic Stem Cell In 1961, Till and McCulloch demonstrated the presence of cells in mouse bone marrow (BM) that were capable of forming macroscopic colonies in the spleen of an irradiated mouse (Till and McCulloch 1961). These spleen nodules contained cells from both myeloid and lymphoid lineages (see Fig. 1.1). Subsequent transplantation of cells derived from splenic nodules into secondary irradiated recipients yielded new splenic colonies and also reconstituted hematopoiesis - demonstrating both self-renewal capacity and differentiation potential (Wu, Till et al. 1967; Till and McCulloch 1961). Cytological and chromosomal marker studies showed that splenic nodules contained a mass of cells (105 to 10 cells) derived from a single progenitor cell (termed colony forming unit-spleen (CFU-S)) (Wu, Till et al. 1967). CFU-S cells display the two hallmark properties of a 'stem cell': self-renewal capacity and the ability to differentiate into at least one cell type. It is now known that most CFU-S are not bona fide HSCs (i.e., capable of long term engraftment) but a more mature population of cells that give short term engraftment (Na Nakorn, Traver et al. 2002) . Strictly speaking, a HSC is defined as a primitive hematopoietic cell with the potential to reconstitute both myeloid and lymphoid hematopoiesis in vivo for the life of the recipient (Kondo, Wagers et al. 2003). In mice, a single HSC can reconstitute the entire hematopoietic system for the lifetime of the mouse (Osawa, Hanada et al. 1996). HSCs are primarily found in BM but also in peripheral blood, umbilical cord blood and in low numbers in the liver and spleen (Holyoake, Nicolini et al. 1999). B M HSCs are capable of self-renewal and differentiation into all the mature blood cell types and some other cell types such as bone osteoclasts and dendritic cells. In the absence of injury, the size of the total pool of HSCs remains roughly constant. Thus, at the population level, HSC divisions must, on average, produce one daughter cell that returns to quiescence and one daughter cell that proceeds to yield the various hematopoietic lineages (Kondo, Wagers et al. 2003) . Whereas most HSCs and early progenitors are quiescent in the G 0 phase of the cell cycle (non-proliferating), most mature progenitors are proliferating and producing mature effector cells (Hao, Thiemann et al. 1996). In the absence of stresses, proliferation within the hematopoietic compartment is balanced by apoptosis in progenitors and mature cells (Domen, Cheshier et al. 2000; Domen 2001). Stresses such as bleeding, hypoxia, and infection trigger stored pools of cells (mostly in the BM) to be released into the circulation 2 and recruited to sites of injury and inflammation (Rogowski, Sasson et al. 1998). Under such conditions, fewer progenitors and mature cells undergo apoptosis (Koury, Sawyer et al. 2002; Endo, Odb et al. 2001) and quiescent progenitors and HSCs are stimulated by a variety of growth factors and cytokines to proliferate and differentiate into mature white cells, red blood cells (RBCs) and platelets (Smith 2003). When the demand for blood returns to normal, the anti-apoptotic and proliferative pathways are turned off, the progenitors return to storage sites and hematopoiesis settles back to maintenance levels (Smith 2003). 1.1.2 Hematopoietic Development During the development of an embryo, hematopoiesis occurs in a variety of anatomical locations throughout gestation. Primitive, or embryonic, hematopoiesis first manifests in yolk sac blood islands (Haar and Ackerman 1971) and is relocated to other organs as the embryo matures. Primitive hematopoietic cells consist mainly of large nucleated erythroid cells designed to carry oxygen from the placenta to the embryo. Definitive, or adult, hematopoiesis first arises in the yolk sac and the aorta-gonad-mesonephros (AGM) region of the embryo prior to the onset of a mature circulatory system. In the human embryo, definitive hematopoiesis is established in the liver at 6 weeks of gestation. Fetal liver HSCs migrate to the spleen at 12 weeks of gestation and then finally to the B M at 20 weeks (Ikuta 1993; Morrison, Hemmati et al. 1995). The B M becomes the predominant site of hematopoietic cell development soon after birth and throughout adult life (Clapp, Freie et al. 1995). 1.1.3 Regulation of Hematopoiesis Homeostasis of blood is maintained by a complex fusion of genetic and environmental mechanisms that balance cellular decisions such as quiescence (survival without growth), proliferation, differentiation, self-renewal and apoptosis (Domen and Weissman 1999; Domen, Cheshier et al. 2000; Orkin and Zon 2002). The mechanisms involved in the production of different types of blood cells as well as the regulation of differentiation and cellular maturation are still not fully understood (Rane and Reddy 2002). Furthermore, the relative importance of the exogenous or intrinsic signals that regulate these processes remains controversial (Kondo, Wagers et al. 2003). 3 Preprogrammed genetic factors and changing environmental factors regulate hematopoietic development. However, the various proposed models for the regulation of hematopoiesis put different weights on the importance of genetic and environmental influences. Intrinsic genetic events are critical to hematopoiesis. Current evidence suggests that many genetic events are preprogrammed to occur in a certain sequence and timing (Smith 2003). Thus, the developmental fate of cells may be predetermined by intrinsic genetic processes and the environmental signals then act upon these cells to amplify or modulate the genetic effects (Smith 2003). For example, in a stochastic-selective model, the commitment decision of hematopoietic precursors is essentially random with success determined by the availability of essential growth and survival factors (Fisher 2002). Alternatively, a sequential lineage-determination model proposed by Brown et al (Brown, Bunce et al. 1985) suggests a predetermined order of developmental choices (Fisher 2002). The Brown model proposes that HSCs undergo an intrinsic program of decisions to generate cells that can differentiate along one or two discrete pathways (Fisher 2002). In these hematopoietic models, the presence or absence of environmental factors influences lineage choices and survival. However, in these models, environmental factors do not drive lineage decisions. Environmental factors are critical for hematopoiesis. The instructional hematopoietic model hypothesizes that the environment plays a primary role in determining the fate of HSCs and progenitors and can direct cells toward any of the various lineages and developmental outcomes (Lemischka 2001). Environmental regulators of hematopoiesis include cytokines, chemokines and extracellular structural components (Zhu and Emerson 2002; Metcalf 1989; Kishimoto, Taga et al. 1994). Cytokines are soluble factors that act by binding to their cognate receptors and mediate intracellular signal transduction events that result in the modulation of gene expression (Rane and Reddy 2002); they can induce positive or negative effects on cellular quiescence, apoptosis, proliferation and differentiation (Smith 2003) and facilitate the interactions between progenitors and the extracellular matrix (Kinashi and Springer 1994). Some cytokines function in a redundant manner and several different cytokines can exert similar and overlapping effects (Rane and Reddy 2002). The more primitive hematopoietic progenitors require some of the same factors such as stem cell factor (SCF, also known as Steel Factor or c-kit ligand), interleukin-3 (TL-3) and 4 granulocyte-macrophage-colony stimulating factor (GM-CSF) whereas more lineage-restricted cells require one or more of the lineage-restricted factors such as erythropoietin (Epo), macrophage-colony stimulating factor (M-CSF), granulocyte-colony stimulating factor (G-CSF) and thrombopoietin (Tpo) (Metcalf 1993; Lok, Kaushansky et al. 1994; Kaushansky, Lok et al. 1994). Chemokines are also important regulators of hematopoiesis (Christopherson and Hromas 2001; Broxmeyer 2001; Majka, Janowska-Wieczorek et al. 2001) since they regulate blood cell trafficking and homing (e.g., to sites of injury) and, like cytokines, may also be negative and positive-growth regulators (Wright, Bowman et al. 2002) . Whether through exogenous factors present in the HSC environment or via hardwired genetic programs, the gene expression profile of a cell ultimately determines its phenotype (Kondo, Wagers et al. 2003). In turn, the expression and activity of transcription factors (TFs), regulators of gene expression, ultimately determine the fate of a cell (Cantor and Orkin 2001). Lineage-specific TFs have essential roles in lineage decisions (Shivdasani and Orkin 1996); they may act in a positive manner to specify or reinforce lineage choices and/or exert inhibitory effects on alternate lineage gene programs (Cantor and Orkin 2001). Multilineage gene expression is common in progenitor cells and may be a prerequisite of a multipotent cell (Kondo, Wagers et al. 2003). At each decision point along the differentiation hierarchy, genes associated with an appropriate pathway are upregulated, while genes not necessary for the chosen lineage are silenced. Figure 1.1 shows a selection of TFs that are important in hematopoietic development and lineage choice. A more detailed discussion regarding the role of TFs in erythroid development is provided in Section 1.2.5. Current dogma professes that HSC progeny are restricted to the hematopoietic system and that "hematopoiesis is a one-way hierarchical process" (Smith 2003). However, new studies suggest that cells previously believed to be developmentally restricted to a particular lineage could be put into an environment where they could recover their ability to commit to other lineages (King, Kondo et al. 2002). This reversible nature of hematopoiesis (and differentiation in general) is termed plasticity. However, much of the evidence for plasticity in hematopoietic development has come from experiments using genetically modified cells or transformed cell lines (reviewed in (Graf 2002)). Whether this process is common and significant in vivo is unclear (Kondo, Wagers et al. 2003). 5 1.1.4 Hierarchy of blood cells and progenitors The hematopoietic system of adult humans produces about one trillion blood cells each day (reviewed in (Cantor and Orkin 2001)). As multipotential HSCs differentiate there is a regulated commitment to more-restricted progenitor cells, and finally to functionally specialized mature cells (Cantor and Orkin 2001; Kondo, Wagers et al. 2003). Terminally differentiated cells are produced that, in general, cannot divide and undergo apoptosis after a period ranging from hours (neutrophils) to decades (for some lymphocytes). Only mast cells, macrophages and the lymphocytes (T and B cells) retain proliferation capacity once matured. All mature blood cells are generated from relatively small numbers of HSCs and progenitors (Weissman 2000; Lemischka 2001). Hematopoietic development is classically illustrated as a hierarchical relationship between HSCs and common myeloid and common lymphoid progenitors (Fisher 2002) (Fig. 1.1). Evidence for this model comes from experiments that showed chromosome-marked BM can repopulate the myeloid or lymphoid compartments of transplant recipients (Becker, McCulloch et al. 1963; Abramson 1978). Current evidence supports the linear model of hematopoiesis in which a cell that loses the potential to develop into a specific lineage never regains that potential (Kondo, Wagers et al. 2003). Thus the first 'decision' of hematopoietic development is whether to become a lymphoid (CLP) or myeloid (CMP) cell type (Kondo, Wagers et al. 2003). Subsequent identification of progenitors that have the functional and phenotypic characteristics of common myeloid progenitors (CMPs) (Akashi, Traver et al. 2000) and common lymphoid progenitors (CLPs) (Kondo, Weissman et al. 1997) and a comparison of their gene transcription profiles (Miyamoto, Iwasaki et al. 2002) have added weight to the argument that the myeloid versus lymphoid choice is a primary decision (Fisher 2002). Downstream of CMPs and CLPs lie more mature progenitors that are further restricted in the number and type of lineages that they can generate (Akashi, Traver et al. 2000). Figure 1.1 depicts the hierarchical model of hematopoiesis: HSCs generate multiple hematopoietic lineages through a series of increasingly lineage-restricted intermediate progenitors. For example, a CLP may give rise to B, T or natural killer (NK) cells and a CMP can generate RBCs, megakaryocytes/platelets, granulocytes or monocytes (Akashi, Traver et al. 2000). 6 1.1.5 Functions of mature blood cells Mature cells usually die after several days or weeks, creating a demand for a constant supply of hematopoietic cells. Historically, the mature cells that circulate in the blood stream have been classified into 3 categories: RBCs, platelets and white blood cells (WBCs). WBCs (also called leukocytes) include the granulocytes, monocytes and lymphocytes. However, a more useful classification of blood cells is based on the hierarchical hematopoietic model depicted in Figure 1.1 where mature blood cells can be categorized based on common precursors. Using this system of classification, two major classes of blood cells can be defined as the lymphoid and the myeloid compartments. Cells in the myeloid compartment arise from a CMP and include RBCs, platelets, monocytes/macrophages, the granulocytes (eosinophils, neutrophils and basophils), microglial cells and dendritic cells (Akashi, Traver et al. 1999; Cantor and Orkin 2001). RBCs (also called erythrocytes) deliver oxygen to and remove carbon dioxide from body tissues (Migliaccio, Vannucchi et al. 1996). These cells remain within the arteries, veins and capillaries. Platelets, which are cell fragments derived from megakaryocytes, circulate in the blood and help repair damaged blood vessels and aid in blood clotting. Granulocytes harbour numerous secretory vesicles or granules; they are further classified based on granule content. The most abundant granulocyte is the neutrophil (also called polymorphonuclear leukocyte) which has a multilobed nucleus and primarily functions to phagocytose and destroy invading bacteria. Basophils and mast cells secrete histamine and serotonin from their granules to regulate inflammatory reactions. Eosinophils destroy parasites and modulate allergic inflammatory responses. Monocytes leave blood vessels and mature into macrophages, which function as phagocytes along with neutrophils. However, macrophages are much larger and_ longer lived than neutrophils and also have a role in clearing dead or damaged cells in many tissues. Similarly, cells of the lymphoid compartment are derived from a CLP. Important lymphoid cells included natural killer (NK) cells, T cells and B cells (Akashi, Traver et al. 1999; Cantor and Orkin 2001). B cells differentiate into plasma cells upon encountering a recognized antigen and are responsible for the production of antibodies. T cells can further be divided into helper T ( T H ) or cytotoxic T ( T C T L ) cells. T H and TCTL cells regulate the 7 activities of other leukocytes and kill virus-infected cells, respectively. NK cells are lymphocyte-like cells that can kill some types of tumour cells and virus-infected cells. 8 Figure 1.1: The hierarchical model of hematopoiesis. The hematopoietic stem cell (HSC) is normally quiescent and divides infrequently to regenerate stem cells (self-renewal) or produce the more restricted hematopoietic progenitors, the common lymphoid (CLP) and common myeloid progenitor (CMP). Progenitor cells are stimulated to proliferate and differentiate by a variety of growth factors but progressively lose their capacity for division as they become increasingly lineage-restricted and eventually yield terminally differentiated blood cells. Mature T cells, mature B cells, macrophages and mast cells are among the end-stage cells that retain proliferative potential. Some of the developmental pathways, such as the pathway to mast cells, remain uncertain (dotted arrow). Lineage important TFs are indicated in red below the arrows. Lineage important cytokines and growth factors are indicated above the arrows and include the ligand for flk-2/flt-3 (flk-2/flt-3 L), SCF, various interleukins, GM-CSF, G-CSF, M-CSF, Tpo and Epo. (Alberts 1994; Zhu and Emerson 2002). 9 Pre B cell EZA Q CE|| CFU-Bas BasoDhil CFU-Eos SCF IL-3 ||flk-2/flt-3 ligand CFU-GM TPO SCF IL-3 CFU-Meg SCF IL-3 BFU-E IL-3 GM-CSF GATA-1 C/EBP Eosinophil Osteoclast SCF IL-3 3M-CSF IL-11 IL-6 TPO Monocyte Megakaryocyte GATA-1 FOG NF-E2 Macrophage Platelets SCF IL-3 GM-CSF EPO GATA-1 IL-3 GM-CSF EPO CFU-E GATA-1 NF-E2 EKLF Erythrocyte FOG 10 1.2 ERYTHROPOIESIS 1.2.1 Erythroid developmental stages RBCs are highly specialized cells that require the expression of specialized proteins for their unique shape and oxygen carrying function (Migliaccio, Vannucchi et al. 1996). Mutations in one of more of these proteins can result in a deleterious loss of function and is the root of several RBC disorders including sickle cell anemia and thalassemias (Bull B. S. and Breton-Gorius 1995). Since RBCs have a limited life span of about 120 days (in humans), they must be continually replaced such that adult humans produce over 200 billion cells daily (Lappin 2003). This large volume of cells is derived from a pool of undifferentiated cells in the marrow by a regulated process of proliferation and differentiation called erythropoiesis (Migliaccio, Vannucchi et al. 1996). Erythropoiesis maintains the red cell volume by replacing cells lost by senescence, bleeding or destruction (Erslev and Besarab 1995). The erythroid lineage, which includes only cells that are 'irreversibly' committed to RBC maturation, is one of several differentiation pathways arising from the CMP (Fig. 1.1). Embryonic development of erythropoiesis involves both primitive and definitive steps (Migliaccio and Migliaccio 1998; Palis and Segel 1998; Dame and Juul 2000). In humans, blood cells are first formed outside the embryo in numerous blood islands within the yolk sac (reviewed in (Erslev 1995)). Cells formed in the blood islands are large, nucleated erythroblasts that synthesize embryonic globin forms (Palis, Robertson et al. 1999; Orkin 1996). Definitive erythropoiesis, established in the fetal liver, yields enucleated erythrocytes that synthesize adult forms of globin (Dzierzak and Medvinsky 1995). Near the fifth fetal month, the B M begins to support definitive erythropoiesis. At birth and through adult life, the B M assumes the bulk of erythropoietic activity except in cases of severe anemia (Erslev 1995). In normal adult BM, erythropoiesis occurs within multicellular structures known as erythroblastic islands (Bull B. S. and Breton-Gorius 1995). These structures consist of "nursing" macrophages surrounded by closely associated erythroid cells undergoing maturation (Bull B. S. and Breton-Gorius 1995). As the erythroid progenitors mature, they move from a position next to the body of the macrophage to the periphery of the blood islands (Bull B. S. and Breton-Gorius 1995). When the nucleus of the mature reticulocyte is 11 expelled, the erythrocyte then leaves the marrow to enter the circulation (Bull B. S. and Breton-Gorius 1995). Following commitment, erythroid progenitors progress through several stages, becoming more functionally specialized with maturation (Bull B. S. and Breton-Gorius 1995). Major committed progenitors include the burst forming unit-erythroid (BFU-E) and the more mature colony forming unit-erythroid (CFU-E), so named for their colony morphology within in vitro assays (Migliaccio, Vannucchi et al. 1996). The BFU-E, defined by its ability to create a burst of cell clusters in semisolid media (Bull B. S. and Breton-Gorius 1995), is the earliest known committed erythroid progenitor. BFU-E are derived from a common erythroid/megakaryocyte (Migliaccio, Vannucchi et al. 1996) progenitor, which, in turn, is derived from a CMP. After several divisions, the BFU-E becomes a CFU-E that, in turn, gives rise to differentiated erythroid cells, the different stages of which have distinct morphology and staining histology (Migliaccio, Vannucchi et al. 1996). Like HSCs, most BFU-Es are not actively proliferating (i.e., most are in the Go/G| phase of the cell cycle); whereas most CFU-E are actively proliferating (i.e., most are in the S-phase of the cycle). As CFU-Es differentiate to late-stage erythroblasts, they stop dividing and accumulate in the Go phase before enucleation (Koury, Sawyer et al. 2002). Since some BFU-Es yield single colonies containing both erythroblasts and megakaryocytes, the commitment to erythroid differentiation apparently occurs between the BFU-E and the CFU-E stages (Koury, Sawyer et al. 2002). Erythroid progenitors at the CFU-E stage and later can express proteins found in megakaryocytes (Goldfarb, Wong et al. 2001) but differentiation of CFU-E or erythroblasts into megakaryocytes has not been demonstrated (Koury, Sawyer et al. 2002). Proerythroblasts (also called pronormoblasts) result from cell division of CFU-E and are the first morphologically identifiable RBC progenitor. Figure 1.2 depicts the course of maturation of proerythroblasts in the BM through the various stages of development to mature RBCs. 12 Pronormoblast Figure 1.2 : Stages of erythroid differentiation (adapted from (Riley, Ben-Ezra et al. 2002) and (Koury, Sawyer et al. 2002)). Proerythroblast - The first morphologically recognizable erythroid precursor is the largest progenitor cell with a nucleus comprising the bulk of the cell area (80%). Basophilic normoblast - These cells are smaller than proerythroblast with a nucleus comprising 75% of the cell area and stain with distinct staining properties (i.e., basophilic). Polychromatophilic normoblast - Smaller than basophilic normoblasts and the basophilic (pink) stain is diluted by the presence of hemoglobin (Hb). After the polychromatophilic stage, erythroid cells no longer divide (Koury, Sawyer et al. 2002). Orthochromatic normoblast - Hb levels are high enough such that this cell stains like a mature erythrocyte. About the same size as an early reticulocyte but with a nucleus comprising 25% of the cell area. The orthochromatic erythroblast enucleates to become a reticulocyte. Reticulocyte - immature RBCs that arise after enucleation of the orthrochromatic normoblast in the B M . They contain mitochondria, RNA, ribosomes, a centriole and remnants of the golgi. The loss of transferrin receptors and the breakdown of RNA signals the last stages of maturation of reticulocytes into mature RBCs. Erythrocyte (RBC) - highly specialized cell with no organelles and has a biconcave disc structure because it lacks a nucleus. 13 Enumeration of immature RBCs, or reticulocytes, is a diagnostic tool often used in the clinic because the number of reticulocytes in peripheral blood reflects the erythropoietic activity of BM, the rate of reticulocyte delivery from hematopoietic organs to the blood and the rate of reticulocyte maturation (reviewed in (Riley, Ben-Ezra et al. 2002)). Reticulocytosis is an increase in the number of peripheral blood reticulocytes that occurs in anemic patients with functional BM (i.e., an appropriate recovery response to anemia). On the other hand, reticulocytopenia occurs in anemia patients with dysfunctional bone marrow (loss of erythropoietic function). Therefore, reticulocyte enumeration is a valuable measure of the regenerative activity of bone marrow after chemotherapy for cancer or AIDS, or after B M transplants, or other illnesses that may cause anemia. 1.2.2 Regulation of erythropoiesis The growth factor requirements for proliferation and differentiation of erythroid cells have been identified using cultures of purified populations of HSCs and pure growth factors (Migliaccio, Migliaccio et al. 1990). From these studies it has been shown that purified mouse and human HSCs have a requirement for SCF in order to survive and proliferate in culture (Migliaccio, Migliaccio et al. 1991; Hirayama, Shih et al. 1992). Also, SCF is involved in the early steps of erythroid commitment since spontaneous mutations occurring in the SCF or c-kit (surface receptor for SCF) loci are associated with anemic phenotypes (Russell 1979). Specifically, the generation of BFU-E from more primitive progenitors is largely dependent on SCF while their subsequent differentiation occurs through a SCF-independent pathway (Migliaccio, Vannucchi et al. 1996). SCF is capable of promoting the expansion of erythroid progenitors by enhancing proliferation while at the same time delaying maturation (Nocka, Majumder et al. 1989; Krantz 1991; Ogawa, Nishikawa et al. 1993). Several studies have shown that the optimal proliferation of BFU-E in culture also requires the presence of IL-3 (and in humans GM-CSF), Epo (Migliaccio, Migliaccio et al. 1988; Sonoda, Yang et al. 1988) and one of the insulin family of growth factors (e.g., IGF-1 or insulin) (Dainiak and Kreczko 1985; Akahane, Tojo et al. 1987; Migliaccio, Bruno et al. 14 1987; Merchav, Tatarsky et al. 1988). However, more mature progenitors (CFU-E) require only Epo and insulin (Iscove, Guilbert et al. 1980; Migliaccio and Migliaccio 1988). Glucocorticoids (GC) also enhance erythropoiesis via binding to their nuclear receptors. Mice deficient in glucocorticoid receptors exhibit a complete loss of stress erythropoiesis (i.e., the ability to increase erythropoietic production under stress conditions such as hypoxia or hemolysis (Bauer, Tranche et al. 1999)). Stress erythropoiesis is discussed in more detail in Chapter 3. As well, low concentrations of the chemokine stromal cell-derived factor-1 alpha (SDF-la) promote hematopoietic cell growth (Lataillade, Clay et al. 2000) but high levels decrease erythroid progenitor growth through upregulation of Fas ligand (FasL) production and subsequent erythroid apoptosis by the FasL/Fas pathway (Gibellini, Bassini et al. 2000). During normal erythropoiesis many erythroid progenitors die via apoptosis (Koury, Sawyer et al. 2002) and this can be prevented by high levels of 3 known factors: SCF, insulin-like growth factor (IGF-1) or Epo (Koury, Sawyer et al. 2002). They appear to prevent programmed cell death by activating STAT5 and, as a result, upregulating anti-apoptotic factors such as BC1-XL (Testa 2004). The production of Epo, the most specific trophic factor for apoptosis prevention in erythroid cells, is regulated by tissue oxygenation via oxygen sensing transcription factors like HIF-la (Koury, Sawyer et al. 2002). The mechanisms responsible for apoptosis during the Epo-dependent stages of erythropoiesis have not been completely vetted, but multiple factors in the erythropoietic environment may be involved (Koury, Sawyer et al. 2002). For example, inflammatory cytokines such as tumor necrosis factor-alpha (TNFa) and interferon-gamma (IFNy) can induce apoptosis through upregulation of the FasL/Fas pathway in erythroid progenitor cells (Maciejewski, Selleri et al. 1995; Dai, Price et al. 1998). Interestingly, in this regard, FasL is expressed on the plasma membranes of mature erythroblasts and, if there is an overabundance of these cells, the FasL on their surface can bind Fas on the surface of more immature erythroblasts to induce apoptosis in a negative feedback loop (Testa 2004). Death receptors also have a physiological role as regulators of erythropoieisis. For example, the activation of death receptors in the presence of Epo has been shown to block erythroid maturation via a caspase-dependent mechanism that involves the cleavage of SCL/Tal-1 and 15 GATA-1 (both transcription factors important for erythroid differentiation) (Zeuner, Eramo et al. 2003; De Maria, Zeuner et al. 1999; reviewed in Testa 2004). Interestingly, normal maturation of late erythroblasts requires 'mild caspase activation' (Testa 2004). Inhibition of G|-phase cyclins through their associated cyclin-dependent kinases (CDKs), specifically cdk2, is important for the cessation of cell division as CFU-Es mature to late erythroblasts (Koury, Sawyer et al. 2002). Cessation of cell division and Gi cell cycle arrest at this stage is crucial to the final stages of differentiation (Koury, Sawyer et al. 2002). 1.2.3 Erythropoietin (Epo) The biological effects directly associated with Epo stimulation of their target cells include suppression of apoptosis (Kelley, Green et al. 1994; Koury and Bondurant 1990) and induction of proliferation (Miura, D'Andrea et al. 1991). Erythropoietin is a glycoprotein hormone that circulates at 1/100 the concentration of most other hormones in the body (Lappin, Maxwell et al. 2002; Maxwell 2002). The first evidence that RJ3C production was hormonally regulated was from experiments by Carnot and Deflandre in 1906 who demonstrated increased RBCs in the circulation of normal rabbits injected with plasma from rabbits made anemic by bleeding (Carnot P 1906). Two seminal works that provided clear evidence that a humoral factor regulated RBC production were published in the 1950's by Kurt Reissmann and Allan Erslev, both pioneers in Epo research. Kurt Reissmann showed that when one partner of a set of parabiotic rats (animals in which the circulatory systems have been surgically joined) was subjected to hypoxic conditions, both animals developed reticulocytosis, increased hemoglobin concentration and developed bone marrow hyperplasia (presumably owing to increased erythropoietic activity in the BM) (Reissman 1950). Similarly, Allan Erslev injected large volumes of plasma taken from donor rats (after bleeding-induced anemia) into normal recipient rats and discovered that the 'anemic' plasma caused a reticulocytosis in the receipients (Erslev 1953). After a span of more than two decades, the active humoral factor, Epo was purified from the urine of patients with aplastic anemia (10 mg of pure Epo from 2500 L of aplastic anemic urine) (Miyake, Kung et al. 1977) and it was cloned in 1985 by scientists at the Genetics Institute (Jacobs, Shoemaker et al. 1985) and Amgen (Lin, Suggs et al. 1985). Through recombinant DNA technologies, Epo is available in pharmacological quantities and 16 is used to treat patients with anemia due to chemotherapy or kidney failure (Shaw 1967; Eschbach, Egrie et al. 1987). In adults, the major site of Epo synthesis is the kidney (Jacobson, Goldwasser et al. 1957). Therefore patients with renal failure require supplemental Epo to maintain normal levels of RBCs (Winearls, Oliver et al. 1986; Eschbach, Egrie et al. 1987; Eschbach, Kelly et al. 1989). B M suppressive therapies used to treat cancer or AIDS can also reduce RBC levels and lead to anemia. Erythropoiesis is regulated by a negative feedback loop in which Epo production is inversely correlated with oxygen tension in arterial blood (Zanjani and Ascensao 1989). Hypoxic or chemical stimuli increase the de novo synthesis of Epo, rather than stimulate the release of pre-formed Epo from cellular stores (Beru, McDonald et al. 1986). Epo circulates in the plasma and induces RBC production in the bone marrow where it binds to specific receptors expressed on the surface of erythroid progenitor cells (reviewed in (Lappin 2003)). Deletion of the Epo gene or the Epo receptor (EpoR) is lethal in mice (and presumably in humans) and death occurs by embryonic day 13-15 (when definitive erythropoiesis begins) as a result of profound anemia, despite normal BFU-E and CFU-E numbers in the fetuses (Wu, Liu et al. 1995; Lin, Lim et al. 1996; Shivdasani and Orkin 1996; Zon 1995). This indicates that the Epo/EpoR interaction is essential for normal erythropoiesis past the CFU-E stage of development but perhaps dispensable at earlier stages. 1.2.4 The erythropoietin receptor (EpoR) and its signalling properties The mouse (D'Andrea, Lodish et al. 1989) and human EpoRs (Jones, D'Andrea et al. 1990) were cloned in 1989 and 1990, respectively, and share 82% amino acid identity (Migliaccio, Vannucchi et al. 1996). As a member of the class I family of cytokine receptors, the EpoR shares structural characteristics with some of the interleukin receptors (IL-2, 3, 6, 7, and 9) and receptors for G-CSF, leukemia inhibitory factor, oncostatin M , ciliary neurotrophic factor, growth hormone, prolactin and Tpo (D'Andrea and Zon 1990; Youssoufian, Longmore et al. 1993; reviewed in Constantinescu, Ghaffari et al. 1999). Members of this class of cytokine receptors share high homology in their ligand binding domains (Bazan 1990). Features of this family are shown in Figure 1.3: 17 1. A single transmembrane domain oriented in the plasma membrane such that its amino-terminus is exposed to the extracellular side of the membrane (Migliaccio, Vannucchi et al. 1996); 2. Four conserved cysteine (C) residues in amino terminal region and a conserved membrane proximal (but in the extracellular region) 20 amino acid motif containing a tryptophan-serine-x-tryptophan-serine (WSXWS) repeat essential for ligand binding (Yoshimura, Zimmers et al. 1992; Migliaccio, Vannucchi et al. 1996; Constantinescu, Ghaffari et al. 1999). 3. An intracellular region that lacks a tyrosine kinase domain but contains two domains (designated Box 1 and 2) capable of interacting with receptor-associated tyrosine kinases (e.g., JAKs). These receptor-associated tyrosine kinases are essential for the propagation of signals upon receptor-ligand binding (Migliaccio, Vannucchi et al. 1996). Figure 1.3: Features of the cytokine receptor superfamily. 18 The EpoR forms homodimers (formerly thought to be ligand-induced), like the receptors for growth hormone (GH), prolactin, G-CSF and Tpo whereas other members of the family form heterodimers (Constantinescu, Ghaffari et al. 1999) (e.g., IL-3, IL-5 and GM-CSF). EpoRs are expressed on the surface of cells of erythroid (Sawada, Krantz et al. 1988; Landschulz, Noyes et al. 1989; Broudy, Lin et al. 1991) and megakaryocyte (Fraser, Tan et al. 1989; Ishibashi, Koziol et al. 1987) progenitors where they have well-studied roles in growth and differentiation; however, the EpoR is ubiquitous and its expression has been detected on diverse cell types (e.g., neuronal cells (Brines, Grasso et al. 2004; Coleman and Brines 2004)). The surface expression of EpoRs on different stages of development correlates with the cell's in vitro responsiveness to Epo (Sawada, Krantz et al. 1988; Landschulz, Noyes et al. 1989; Broudy, Lin et al. 1991). BFU-E, which respond to Epo only in combination with other hematopoietic growth factors (like SCF), express approximately 25-50 high affinity Epo-binding sites per cell (Migliaccio, Vannucchi et al. 1996). CFU-E and proerythroblasts, on the other hand, have 300-400 high affinity binding sites per cell (Migliaccio, Vannucchi et al. 1996) and are much more responsive to Epo. EpoRs are low or absent on late normoblasts and reticulocytes (Migliaccio, Vannucchi et al. 1996). Thus, the CFU-E and proerythroblast stages are the most sensitive to Epo (Lappin 2003). Crystallographic analysis of extracellular domains of the EpoR suggests that EpoR dimers exist in the absence of Epo (Koury, Sawyer et al. 2002) and it is therefore thought that Epo brings pre-existing dimers into the correct orientation for Janus kinase 2 (JAK2)-induced tyrosine phosphorylation and alteration of EpoR conformation (Fig. 1.3) (Miura, Nakamura et al. 1994). JAK2 first tyrosine phosphorylates and activates itself and then phosphorylates some or all of the 8 tyrosine residues within the EpoR cytoplasmic domain (Koury, Sawyer et al. 2002). Phosphorylation of these residues, in turn, attracts Src-homology 2 (SH2) containing proteins and phosphotyrosine binding (PTB) proteins and, as a result, these proteins become tyrosine phosphorylated themselves (Miura, D'Andrea et al. 1991; Komatsu, Adamson et al. 1992; Gobert, Porteu et al. 1995). Alternatively, by virtue of translocation to the plasma membrane, proteins can act on plasma membrane-associated substrates and trigger multiple cascades culminating in the activation of specific sets of genes responsible for mediating the survival, and perhaps differentiation, of erythroid progenitors. 19 However, it is worthy of note that JAK2 is not the only tyrosine kinase activated in response to Epo. The Src family member, Lyn, has also been implicated (Arai, Kanda et al. 2001) and appears to play an important positive role in erythropoiesis (Tilbrook, Palmer et al. 2001), perhaps, in part, by binding to and activating phospholipase C (PLCy2) and phosphatidylinositol 3-kinase (PI3K) (Boudot, Dasse et al. 2003). Finally, proteins responsible for attenuating or turning off these Epo-induced signals (i.e., tyrosine and lipid phosphatases) are activated as well, but with slightly slower kinetics (Klingmuller, Lorenz et al. 1995; Tauchi, Damen et al. 1996). Signalling proteins that are attracted to the tyrosine phosphorylated residues in the EpoR include: the transcription factor signal transducer and activator of transcription (STAT5) which then dimerizes and binds to IFN-gamma activated sites (GAS) sequences on DNA and upregulates various proteins including the pro-survival Bcl-2 family member, BC1-XL; Ras and Grb2/Sos complexes which activate the ERK/MAPK pathway (Koury, Sawyer et al. 2002); phosphatidylinositol 3-kinase (PI3K) which activates the PKB/Akt pathway (a pro-survival pathway) and; the SH2-containing protein tyrosine phosphatases SHP-1 and SHP-2 (Constantinescu, Ghaffari et al. 1999). Several reports suggest that membrane proximal Y 3 4 3 of EpoR (see Fig. 1.3) is required for STAT5 activation whereas the distal end of the EpoR is required for activation of the PI3K pathway and ERK cascades and activation of the negative regulator SHP-1 (Koury, Sawyer et al. 2002). However, because mutant and severely truncated EpoRs (when expressed in EpoR-/- progenitors) support normal erythroid differentiation, some of these signalling pathways might be redundant (Constantinescu, Ghaffari et al. 1999). However, it must be borne in mind that in vivo, other factors like IGF-1 may compensate for the lack of certain intracellular signals by mutant EpoRs (Damen, Krosl et al. 1998). As well, a poorly functioning mutant EpoR may still result in a near normal hematocrit in vivo with only a small compensatory increase in Epo production (Koury, Sawyer et al. 2002) - an example of homeostatic compensation. With regard to the distal end of the EpoR, it appears, on balance, to act primarily as a negative regulatory domain by binding SHP-1, which dephosphorylates JAK2 (Daigle, Yousefi et al. 2002) and thus attenuates EpoR signalling (Koury, Sawyer et al. 2002). In keeping with this, mice expressing a truncated EpoR have slightly elevated hematocrits 20 (Zhang, Johnson et al. 2001). Truncated EpoRs have also been found in several families and these people display a benign autosomal dominant erythrocytosis (de la Chapelle, Traskelin et al. 1993). Althought these findings demonstrate the importance of SHP-1 in downmodulating EpoR signalling (Constantinescu, Ghaffari et al. 1999), the negative regulatory role of the distal cytoplasmic domain of the EpoR has been challenged experimentally since SHP-1 has been shown to interact with JAK2 directly (Jiao, Berrada et al. 1996). This, and more recent studies of EpoR truncation mutants (in the absence of marked erythrocytosis) have down-played the regulatory role of the EpoR distal cytoplasmic domain (Zang, Sato et al. 2001; Divoky, Liu et al. 2001). Negative regulators of EpoR signaling implicated in the regulation of RBC production included members of the suppressors of cytokine signaling (SOCS) family. Of the 8 members of the SOCS family, CIS, SOCS-1 and SOCS-3 appear to have roles in erythroid cells. Each of the SOCS family members contain an SH2 domain and a 'SOCS box', a domain that functions to target SOCS-bound proteins for proteasomal degradation (Starr, Willson et al. 1997; Hilton, Richardson et al. 1998; Kamura, Sato et al. 1998; Zhang, Farley et al. 1999). SOCS-1 and SOCS-3, but not CIS, also have domains that can directly inhibity JAK2 tyrosine kinase activity (Ilangumaran and Rottapel 2003). CIS (cytokine-inducible SH2 domain containing protein), the founding member of the SOCS family, was first identified as an Epo-inducible protein (Yoshimura, Ohkubo et al. 1995). CIS expression is, in part, regulated by STAT5 as CIS expression increases upon activation of JAK2/STAT5 and the CIS promoter contains STAT-binding elements (Minamoto, Ikegame et al. 1997; Verdier, Rabionet et al. 1998). CIS may negatively regulated EpoR signaling by targeting the receptor for proteasome-mediated degradation or via the CIS SH2 domain outcompeting STAT5 for EpoR binding (both CIS and STAT5 preferentially bind the same EpoR phosphotyrosine motif) (Verdier, Chretien et al. 1998; Sasaki, Yasukawa et al. 2000). Forced expression of CIS has been shown to inhibit cell growth in response to Epo and IL-3 and this correlates with reduced STAT5 activation in response to these stimuli (Jegalian and Wu 2002; Matsumoto, Seki et al. 1999). SOCS-1 and SOCS-3 inhibit cytokine receptor phosphorylation by directly suppressing JAK2 activity (Endo, Masuhara et al. 1997; Naka, Narazaki et al. 1997; Starr, Willson et al. 1997; Cohney, Sanden et al. 1999; Sasaki, Yasukawa et al. 2000; Cacalano, Sanden et al. 2001). 21 It has been shown that, in the terminal stages of erythropoiesis, the expression of B c l - X L (which is upregulated by STAT5) is greatly increased (Gregoli and Bondurant 1997; Motoyama, Kimura et al. 1999) and mice with a conditional deletion of B c l - X L (deletion occurs only after birth) are anemic because of major losses of erythroid cells in the late erythroblast and reticulocyte stages (Wagner, Claudio et al. 2000). Interestingly, erythroid cells from the BM of patients with polycythemia vera have increased levels of BC1-XL compared to those from normal erythroblasts (Silva, Richard et al. 1998). It is likely that, in situations where STAT5 signalling does not appear to be necessary for Epo-induced erythropoiesis, compensatory factors (e.g., IGF-1) are present and its role can only be clearly demonstrated during stress erythropoiesis (Socolovsky, Nam et al. 2001). The EpoR is thought to exist in a 'signalsome complex' with several receptor tyrosine kinases such as c-kit (Wu, Klingmuller et al. 1997) and RON (van den Akker, van Dijk et al. 2004) and other non-receptor tyrosine kinases such as JAK2 (Witthuhn, Quelle et al. 1993), Lyn (Tilbrook, Ingley et al. 1997) and Tec (Machide, Mano et al. 1995). RON, also known as Stk (mouse form), is as receptor tyrosine kinase expressed in erythroblasts as a full length (flRON) and a short form (sfRON). The short form lacks the extracellular domain of the full length variety. RON's ligand, macrophage-stimulating protein (MSP), which binds the extracellular domain of the long form, has a role in the induction of macrophage spreading, migration and phagocytosis but also inhibits the lipopolysaccharide (LPS)-induced production of pro-inflammatory mediators (Wang, Zhou et al. 2002). However, RON can also be activated independent of ligand binding through heterodimerization with various receptors (e.g., c-Met, integrins, the common P subunit of IL-3, IL-5 and GM-CSF and the EpoR) (Liu and Rohrschneider 2002; Ney and D'Andrea 2000; Follenzi, Bakovic et al. 2000; Mera, Suga et al. 1999; Danilkovitch-Miagkova and Leonard 2001). In erythroblasts, MSP does not induce flRON phosphorylation or downstream signalling (van den Akker, van Dijk et al. 2004). However, Epo-induced activation of the EpoR and JAK2 efficiently induces RON tyrosine phosphorylation and activation (van den Akker, van Dijk et al. 2004). These events subsequently lead to the tyrosine phosphorylation of Gabl (by RON directly) and Gab2 (through Src kinases) which, in turn, results in the recruitment of signaling intermediates, independent of EpoR tyrosine phosphorylation, to the EpoR signalsome complex (van den Akker, van Dijk et al. 2004). The signaling pathways initiated in this way 22 may have redundant or distinct cell-fate consequences than those mediated by signaling intermediates recruited to the tyrosine residues of the EpoR. RON's role in erythroid progenitors is illustrated by the transformation of murine erythroid progenitors (defined by Epo-independent expansion) by the Friend spleen focus-forming virus (SFFV). The Epo-independent expansion of (SFFV)-infected erythroblasts requires the expression of sfRON (Liu and Rohrschneider 2002). The mechanism of transformation involves the binding of a viral protein (gp55) to EpoR and sfRON (Ney and D'Andrea 2000; Ruscetti 1999). This coupling results in the constitutive phosphorylation and activation of RON and downstream signalling pathways (e.g., Gabl mediated activation of the PI3K and ERK pathways) (Agazie, Ischenko et al. 2002; Finkelstein, Ney et al. 2002). The presence of the EpoR, JAK2 and RON in signaling complex in erythroblasts could explain the viability (without major defects in erythropoiesis) of mice expressing a tyrosine-null EpoR (Zang, Sato et al. 2001). However, the tyrosine residues of the EpoR may have a distinct role in erythroblasts since, while RON signaling contributes to erythroblast expansion, it fails to support survival and differentiation, a process which has been shown to rely on STAT5 (Dolznig, Habermann et al. 2002) - RON does not activate STAT5 (van den Akker, van Dijk et al. 2004). 1.2.5 Transcription factors involved in erythropoiesis Erythroid differentiation is regulated by a complex network of transcription factors (TFs) that influence the expression of target genes and this, in turn, determines whether progenitors proliferate, differentiate, apoptose or express specialized proteins such as globin. Both embryonic and adult erythropoiesis require broad spectrum TFs as well as erythroid-specific TFs (Perry and Soreq 2002; Shivdasani and Orkin 1996) and targeted gene disruption studies in mice have revealed blocks in hematopoietic maturation that occur in the absence of specific TFs (see Fig. 1.1) (Cantor and Orkin 2002). Many of these TFs are also associated with various human and murine leukemias (Cantor and Orkin 2002). TFs that have been shown to markedly affect erythropoiesis include GATA-1, FOG-1 (Friend of GATA), SCL, LM02 , E K L F (erythroid Kruppel-like factor), CREB-binding protein (CBP)and PU.l . 23 GATA-1 was identified by its ability to regulate the expression of globin genes, and subsequently, GATA-binding motifs have been found in the promoters or enhancers of all erythroid and megakaryocytic-specific genes examined to date (Orkin 1992; Weiss and Orkin 1995); reviewed in (Cantor and Orkin 2002)). GATA-1 is expressed in multipotential hematopoietic cells as well as these other lineages: erythroid, megakaryocytic, eosinophilic and mast cells (Evans and Felsenfeld 1989; Tsai, Martin et al. 1989). Gene knockout studies in mice have shown that GATA-1 is essential for normal erythropoiesis (Pevny, Simon et al. 1991) and its expression accumulates during erythroid differentiation, with the highest levels at the CFU-E/proerythroblast stage of development (Suzuki, Suwabe et al. 2003). GATA-1" mice (GATA-1 is on the X-chromosome) die during embryonic gestation from severe anemia as erythroid maturation is blocked at the proerythroblast stage (Fujiwata et al, 1996). GATA-1~ embryonic stem (ES) cells also fail to mature past the proerythroblast stage and apoptose, indicating a role for GATA-1 in survival as well as maturation (Weiss and Orkin 1995). Several other TFs important in erythroid development have been reported to interact physically with GATA-1 including Friend of GATA (FOG-1) (Tsang, Visvader et al. 1997). FOG-1, discovered in a yeast two-hybrid screen for GATA-1 binding proteins (Tsang, Visvader et al. 1997), is highly expressed in erythroid and megakaryocytic cells, and is co-expressed with GATA-1 during development (Tsang, Visvader et al. 1997). FOG-1-/- mice die during mid-embryonic gestation due to severe anemia with arrest at a stage of development similar to that observed in GATA-1- mice (Tsang, Fujiwara et al. 1998). Direct physical interaction between GATA-1 and FOG-1 is required for normal erythropoiesis (Crispino, Lodish et al. 1999). Interestingly, unlike the GATA-1 knockout, FOG-1-/- mice have a complete failure of megakaryopoiesis, suggesting that FOG-1 has a GATA-1-independent role early in the development of megakaryocytes (Cantor and Orkin 2002). SCL a was first identified by the study of frequent chromosomal translocations in patients with T-cell acute lymphoblastic leukemia (T-ALL) (Begley, Apian et al. 1989; Finger, Kagan et al. 1989; Chen, Cheng et al. 1990). SCL-/- mice die during embryogenesis with a complete absence of yolk sac blood (Porcher, Swat et al. 1996). A loss of function of LM02 is identical to that of SCL and is consistent with studies showing that SCL and LM02 physically interact (Warren, Colledge et al. 1994; Osada, Grutz et al. 1995). Both GATA-1 24 (Vyas, McDevitt et al. 1999) and c-kit (Krosl, He et al. 1998; Lecuyer, Herblot et al. 2002) gene expression are downstream targets of SCL. EKLF participates in the switch from embryonic to adult p-globin expression in humans (Cantor and Orkin 2002). EKLF-deficient mice die from severe anemia during embryogenesis from a block in adult |3-globin gene activation (Cantor and Orkin 2002). DNA binding sites for EKLF and GATA-1, which can also physically associate, are found in close proximity in erythroid specific genes (Cantor and Orkin 2002). CBP, a histone acetyl transferase, is a ubiquitously expressed TF that interacts with a large variety of proteins including GATA-1 (Blobel, Nakajima et al. 1998). Acetylated GATA-1 has enhanced transcriptional activity and thus, in this context, CBP is an activator of erythroid gene transcription (Cantor and Orkin 2002). PU.l is essential in granulocytic, monocytic and lymphoid development (Scott, Simon et al. 1994; McKercher, Torbett et al. 1996; Hromas, Orazi et al. 1993). Transgenic mice that overexpress PU.l in erythroid cells develop erythroleukemias at higher rates than non-transgenic animals. Additionally, overexpressed PU.l efficiently immortalizes B M -derived erythroblasts (Moreau-Gachelin, Wendling et al. 1996; Schuetze, Stenberg et al. 1993). Like many of the other important erythroid transcription factors, PU.l and GATA-1 interact directly: in this case the interaction is functionally inhibitory (reviewed in (Cantor and Orkin 2002)). 1.3 PROTEIN BINDING DOMAINS Protein binding domains allow for the interaction of signalling proteins with each other as well as with certain phospholipids (Sudol 1998). Control of these interactions is a central theme of signal transduction. A signalling protein can express a variety of different protein interaction domains and thus can interact with a complex network of signalling molecules (reviewed in (Pawson and Nash 2000)). These domains are true protein modules in that they are independently functioning domains that do not require the surrounding protein structure. Protein interaction domains can be loosely divided into two types: phosphorylation-dependent and phosphorylation-independent. Some protein interaction domains have features of both types. 25 The majority of protein-protein interactions are facilitated by phosphorylation-independent domains. These include a very large (and growing) group of domains such as the PDZ, EH, SAM, WD40, GYF, WW and EVH1 and Src-homology 3 (SH3) domains, each with unique expression patterns, specificities and binding properties. Several of these modules (SH3, GYF, WW and EVH1) bind to proline-rich sequences on target proteins (Sudol 1998). A discussion of the features of the SH3 domain is particularly relevant for Chapter 4 of this thesis. SH3 domains are expressed in many proteins and often found in molecules also expressing an SH2 domain or other interaction domains and motifs. SH3's are expressed in all Src family kinases, the regulatory subunit of Class IA PI3Ks and the adaptor proteins Grb2, Nek, and Crk, to name only a few. SH3 domains bind to the minimum consensus sequence Proline-X-X-Proline (PXXP) to form signalling complexes and regulate protein localization and enzymatic activity (Ren, Mayer et al. 1993; Sudol 1998). The specificity of SH3 ligands are determined by the sequence surrounding the core structure. Class I SH3 ligands have a positively charged residue N-terminal to the PXXP core (R/KxxPxxP) whereas Class II ligands have a 'mirror image' consensus with the positive charged residue C-terminal to the core sequence (PxxPxxR/K). In other words, both classes interact with the SH3 domain in a similar way but with opposite symmetry. SH3 domains are comprised of 3 shallow binding pockets, 2 of which interact with the core prolines of the binding consensus and a third that binds the positively charged residue either N-terminal (Class I) or C-terminal (Class II) to the core sequence (Kay, Williamson et al. 2000). Specificity of the SH3-ligand interaction is conferred by both the symmetry of interactions and by other residues surrounding the binding regions (Kay, Williamson et al. 2000). Phosphorylation-dependent domains bind to sites of serine/threonine or tyrosine phosphorylation as well as to specific phospholipids (e.g., PIP3). These protein-protein or protein-lipid interactions are critical in receptor-mediated signalling events (Sudol 1998; Yaffe 2002) and are particularly important in the temporal regulation of signalling events because the interactions can be regulated in a reversible way. Phosphorylation-dependent domains include the SH2 and PTB domains that bind phosphotyrosine (pY) residues of proteins, the PH domains that bind phosphatidylinositides (Pis) and domains that bind phosphorylated serine or threonine residues. 26 The SH2 domain was the first phosphorylation-dependent signalling domain identified (Sadowski, Stone et al. 1986). The major functions of SH2 domains include the recruitment of signalling molecules to sites of tyrosine phosphorylation (such as in activated transmembrane receptors), the recruitment of adaptor molecules or substrates (Pawson and Nash 2000; Yaffe 2002), and the direct regulation of enzymatic function (e.g., intramolecular regulation of the activity of Src family kinases) (Yaffe 2002). An SH2 domain consists of 2 conserved pockets. One pocket has basic residues, including an invariant arginine (R) that forms hydrogen bonds with pY residues. The second pocket confers binding specificity by recognizing 3 to 6 residues immediately C-terminal to the pY site (Yaffe 2002; Pawson and Scott 1997). PTB domains are similar to SH2 domains in that they bind to specific pY residues. However, their binding specificity is conferred by amino acids N-terminal rather than C-terminal to the pY, with a core sequence NPXY (Pawson and Scott 1997; Yaffe 2002). The first PTB domain identified was in the adaptor protein She (which also contains an SH2 domain). Many of the proteins that express PTB domains are adaptor or docking proteins and facilitate recruitment of other proteins to signalling complexes (Pawson and Scott 1997; Yaffe 2002). Domains that have a high affinity for phosphoserine (pS) and phosphothreonine (pT) residues have also been identified and, like PTB and SH2 domains, they mediate complex assembly in a stimulus-regulated, reversible way (Pawson and Nash 2000; Tzivion, Shen et al. 2001). 14-3-3 was the first protein discovered that has a pS binding domain that mediates localization and activation of binding partners such as Raf-1, cdc25, Bad, and the forkhead TFs (Muslin, Tanner et al. 1996; Tzivion, Shen et al. 2001). In particular, the S/T phosphorylation of the forkhead (Drosophila) homolog (rhabdomyosarcoma) like 1 (FKHRL-1) or FOX03A by Akt allows 14-3-3 to bind and sequester FKHRL-1 in the cytoplasm and prevent the transcription of apoptotic genes. Several domains that mediate protein-lipid interactions have also been identified and include CI , C2, F Y V E , PX, F E R M and PH domains (DiNitto, Cronin et al. 2003). The classical and novel isoforms of protein kinase C (PKC), for example, are serine/threonine kinases that contain CI domains (DiNitto, Cronin et al. 2003) and/or C2 domains. These domains enable these PKCs to translocate to the plasma membrane to bind diacylglycerol 27 (DAG) when the latter is transiently generated following growth factor- or cytokine-stimulated activation of the enzyme PLCy (Feng, Becker et al. 2000; Vallentin, Prevostel et al. 2000). Binding to D A G leads to release of a pseudo-substrate PKC domain and an increase in the catalytic activity of these PKCs (Jaken 1996). C2 domains (expressed in the regulatory domain of Class II PI3Ks), regulate binding of proteins to acidic phospholipids in a Ca -dependent or independent manner (Pawson and Scott 1997). The pleckstrin homology (PH) domain regulates the recruitment of various signalling proteins to the plasma membrane. This domain specifically binds Pis (including PI-4,5-P2, PI-3,4-5-P3 (PIP3) or PI-3,4-P2) (Pawson and Scott 1997; Toker 2002). PH domain-containing proteins are attracted to these modified Pis (which reside in the plasma membrane) and their binding can influence the enzymatic activity of the PH-containing proteins. PH domains have variable preferences for PI species; for example, the PH domain of PLCy preferentially binds to its substrate PI-4,5-P2, whereas the PH domain of Bruton's tyrosine kinase (Btk) binds to PIP3 in order to regulate downstream effects such as extracellular calcium [Ca ] entry (Sato, Overduin et al. 2001; Turner and Kinet 1999). Most of the PH domain-containing proteins localize to membranes in a PI3K-dependent manner. Since PI3K generates PIP3, which is rapidly metabolized to PI-3,4-P2, most PH domains display a preference for either PIP3 (predominantly) and/or PI-3,4-P2. Section 1.5 discusses in greater detail the regulation of PIP3 and PI-3,4-P2 and their role as signaling intermediates. 1.4 REGULATION OF SRC-FAMILY KINASE ACTIVITY The Src-family kinases (SFKs) are a large family of non-receptor tyrosine kinases comprised of 9 members: Lyn, Src, Yes, Fgr, Fyn, Lck, Hck, Blk and Yrk (Brown and Cooper 1996). SFKs are widely expressed in hematopoietic and non-hematopoietic cells and multiple members are often expressed together in a variety of cell-type specific expression patterns (Lowell 2004). For example, Lyn is the predominant SFK in B cells (which also express Fyn and Blk) but is not expressed in T cells where Fyn and Lck predominant (Lowell 2004). Src-family members have a common structure consisting of a unique N-terminal domain (U), followed by three Src-homology domains: SH3, SH2 and SHI (the tyrosine kinase domain) (Fig. 1.4) (Lowell 2004). The N-terminus of SFKs contains acylation sites for myristate and palmitate (except for Src and Blk, which have only myristoylation sites) 28 (Lowell 2004). These N-terminal modifications anchor SFKs to membranes, specifically to the cytoplasmic side of the plasma membrane within lipid rafts, and are required for the in vivo activity of SFKs (Resh 1999; Janes, Ley et al. 2000). Figure 1.4 illustrates the mechanism by which the SH3 and SH2 domains of SFKs participate in the intramolecular regulation of SFK activity (Sicheri and Kuriyan 1997; Sicheri, M o a r e f i et al. 1997). A n important (and conserved amoung SFK members) tyrosine motif within the C-terminal domain (Y507 in Lyn) negatively regulates SFKs, when phosphorylated, by binding its own SH2 domain and folding the SFK into an inactive conformation. This inactive/closed conformation is further stabilized by an SH3 interaction with a SFK motif that lies between the SH2 and kinase domain. The C-terminal tyrosine residue is phosphorylated (favouring SFK inactivation) mainly by the C-terminal Src kinase (Csk) and is dephosphorylated (favouring SFK activation) in immune cells by CD45, a transmembrane protein tyrosine phosphatase (PTP). Csk is recruited to the membrane by the Csk-binding protein (Cbp) (also called phosphoprotein associated with glycosphingolipid-enriched domains (PAG)), an integral membrane protein that is localized to membrane rafts, where recruitment of Csk can act directly on SFKs (Davidson, Bakinowski et al. 2003). Csk binds Cbp/PAG through an SH3 and SH2 mediated association: the tyrosine phosphorylation of Cbp's SH2 binding motif is, in turn, regulated by SFKs, an example of negative feedback (reviewed in (Lindquist, Simeoni et al. 2003)). A conserved autophosphorylation site (Y396 in Lyn) within the SFK's kinase domain is required for full activation. In some cells, the autophosphorylation site can be dephosphorylated by CD45 giving this PTP a potential role in the negative regulation of SFKs (Irie-Sasaki, Sasaki et al. 2003). Finally, the interaction of SFK's SH2 or SH3 domain with other proteins such as membrane receptors can serve to stabilize the active/open conformation. For example, the HIV-1 N e f protein, which binds the Hck SH3 domain with high affinity, binds and activates the C-terminally phosphorylated form o f Hck (Moarefi, LaFevre-Bernt et al. 1997). And, Lck is activiated by binding, via its SH3 domain, to a proline motif in the CD28 T cell co-receptor following recruitment of Lck via CD4 (Holdorf, Lee et al. 2002). Sustained activation of Lck in activated T cells requires the presence of CD28 (Holdorf, Lee et al. 2002). Cell signalling often takes place in specialized cell compartments or microdomains that facilitate the concentration and assembly of signalling complexes. Lipid rafts and the F-29 actin submembranous cytoskeleton are two such microdomains important for signalling at the plasma membrane. Lipid rafts, also called low-density detergent-resistant membrane domains (LD-DRM) (Luna and Hitt 1992), consist of glycosphingolipid- and cholesterol-rich lipid domains organized within the plasma membrane (Brown and London 2000; Horejsi 2003). These membrane microdomains are proposed to function as signalling platforms by concentrating some signalling complexes and excluding others. The submembranous F-actin skeleton consists of F-actin polymers connected to the plasma membrane via a series of linkages to integral membrane proteins (Luna and Hitt 1992). The F-actin skeleton compartment has an important role in cell morphological changes (e.g., phagocytosis, endocytosis and exocytosis), chemotaxis and cell division. This compartment is involved in cell signalling via direct recruitment of signalling molecules by actin-binding proteins such as filamin-1 (Lesourne, Fridman et al. 2005). Importantly, lipid rafts and the submembranous F-actin skeleton may interact with each other and thus may act in concert to coordinate signalling events (Kwik, Boyle et al. 2003). The BCR, T cell receptor (TCR) and IgE receptor (FcsRI) all have ability to associate with lipid rafts after being engaged by the appropriate ligand and 'to further coalesce with pre-existing rafts' that contain signalling proteins such as Lyn (Weintraub, Jun et al. 2000; Petrie, Schnetkamp et al. 2000). Accumulation of BCR complexes in lipid rafts does not require Src-kinase activation (Weintraub, Jun et al. 2000) or tyrosine phosphorylation of the Iga/p ITAMs (reviewed in (Matsuuchi and Gold 2001)). However, lipid rafts are particularly important for BCR signalling as it may promote the recruitment of signalling components while at the same time excluding negative co-receptors, such as CD22 (Weintraub, Jun et al. 2000) and the subsequent recruitment of negative regulators like SHP-1. The B cell receptor (BCR) complex consists of antigen-binding subunit (mlg -membrane bound immunoglobulin heavy and light chains) and a signalling subunit composed of CD79a (Iga) and CD79b (IgP) that translates the binding of antigen complexes into intracellular signalling cascades (recently reviewed in (Dal Porto, Gauld et al. 2004)). The cytoplasmic domains of CD79a (Iga) and CD79b (IgP) contain immunoreceptor tyrosine-based activation motifs (ITAMs) that, when phosphorylated on tyrosine residues, recruit and further activate SFKs (Fyn, Lyn and Blk) and Syk (Lam, Kuhn et al. 1997; 30 Nagata, Nakamura et al. 1997) and many other signalling intermediates (Dal Porto, Gauld et al. 2004). The outcome of BCR signalling (activation, apoptosis or antibody production) depends on maturation state of B cell, the context of the signal (i.e., whether inhibitory co-receptors are also engaged) and the strength and duration of the signal (Gold 2002). Lyn has both positive and negative roles in BCR-induced signal transduction; it appears to have a redundant role as positive regulator but has an irreplaceable function in attenuation of BCR-initiated signals (Nishizumi, Horikawa et al. 1998; Lowell 2004; Xu, Harder et al. 2005). As a negative regulator of BCR activated signalling, Lyn has a critical role in the tyrosine phosphorylation of the ITIM containing negative co-receptors such as FcyRIIB and CD22 (Nishizumi, Horikawa et al. 1998; Xu, Harder et al. 2005). As a case in point, Lyn-deficient B cells are resistant to the inhibitory effects of co-aggregation of FcyRIIB and BCR (Gold 2002; Xu, Harder et al. 2005) in large part because the ITIM of the FcyRIIB fails to become tyrosine phosphorylated in the absence of Lyn and SHIP recruitment to the FcyRIIB is phosphorylation dependent (Coggeshall 1998). Importantly, the phenotypes for Lyn-/- and SHIP-/- mice (see Section 1.6) reveals some intriguing similarities (Hibbs, Tarlinton et al. 1995; Chan, Meng et al. 1997; Helgason, Kalberer et al. 2000) and both Lyn and SHIP have been found to be important negative regulators in B cells (Satterthwaite and Witte 2000) and mast cells (Huber, Helgason et al. 1999; Hernandez-Hansen, Mackay et al. 2004) and thus may work in tandem to attenuate immunoreceptor signalling. 31 CD45 Figure 1.4: Regulation of Src-family kinases by intra- and intermolecular interactions. Regulation of SFK activity (Lyn is used as an example). SFKs consist of a unique N-terminal domain (U) followed by and SH3, SH2 and kinase domains. SFKs are anchored to lipid raft microdomains via N-terminal acylation sites. The active and inactive structural conformations of SFKs are regulated by the phosphorylation state of a C-terminal regulatory tyrosine (Y507) via a C-terminal Src kinase (Csk) and a protein tyrosine phosphatase (PTP). The kinase activities of SFKs are further activated by an autophosphorylation site within the kinase domain (Y396). Proteins that interact with the SH2 and SH3 domains of SFKs (e.g., SHIP, the regulatory subunit of PI3K (p85) and surface receptors (CD28)) may stabilize the active/open conformation of SFKs. Dashed arrows indicated protein interactions, solid arrows are tyrosine phosphorylation or dephosphorylation events (modified from (Xu, Harder et al. 2005)). 3 2 1.5 THE PI3K PATHWAY The PI3K pathway is a critical regulator of many biological processes through the generation of PIP3 (reviewed in Krystal 2000; Sly, Rauh et al. 2003). Although PIP3 is present at low levels in the plasma membrane of unstimulated cells, it is rapidly synthesized from PI-4,5-P2 by PI3K in response to a diverse stimuli (reviewed in (Rameh and Cantley 1999). The structure of PIP3, a potent second messenger, and its precursors are shown in Figure 1.5. Transiently generated PIP3 attracts PH domain-containing proteins such as the serine/threonine kinase Akt (also known as protein kinase B (PKB)) to the plasma membrane, a critical step in its activation and propagation of its survival and growth-promoting effects (reviewed in (Sly, Rauh et al. 2003; Maxwell, Yuan et al. 2004)). There are at least 4 different Epo-induced pathways that lead to PI3K activation. (1) Direct interaction of PI3K with the EpoR via the SH2-containing p85 subunit of PI3K and the p Y 4 7 9 in the activated EpoR. (2) Tyrosine phosphorylation of the Grb2-associated binding protein (Gabl) - perhaps by RON (see below). Following its binding to the activated EpoR complex, the phosphotyrosyl residues in Gabl act as docking sites for PI3K (Lecoq-Lafon, Verdier et al. 1999). (3) Through tyrosine phosphorylation of the scaffolding protein, insulin-regulated substrate 2 (IRS2), and subsequent binding to the SH2-containing p85 subunit of PI3K (Bouscary, Pene et al. 2003). (4) Association and activation by Lyn (Boudot, Dasse et al. 2003). Activation of the PI3K pathway has been shown to be critical for erythroid survival, proliferation and differentiation by many investigators, using pharmacological inhibitors of PI3K (e.g., wortmannin and LY294002) (Sui, Krantz et al. 1998; Haseyama, Sawada et al. 1999; Myklebust, Blomhoff et al. 2002). Interestingly, in this regard, LY294002 has been shown to block the Epo-stimulated decrease in the cyclin-dependent kinase inhibitor (CDKI), p27kip. This CDKI blocks entry into S phase and Epo stimulates proliferation in part ci/ by activating, via the PI3K pathway, the ubiquitin E3 ligase SCF complex to proteolytically cleave p27 k i p l (Sutterluty, Chatelain et al. 1999; Carrano, Eytan et al. 1999; Mamillapalli, Gavrilova et al. 2001). 33 Phosphatidylinositol o H 2 c - o x R 2 ^ 0 - C H 0 A R1 1,2 -Diacylglycerol 0 II o-p-o. OH I _ 0 HO Y^OH Phosphate Inositol PH b i n d i n g SHIP SHIP2 Synaptojanin 1,2 3-ptase PI-5P PI-3,5 P 2 Figure 1.5: The structure of PI-3,4,5-P3 (PIP3), its precursors and products. Phosphatidylinositol (PI) is the precursor for all phosphoinositides. Arrows pointing to the right indicate kinase reactions while arrows pointing to the left represent reactions regulated by phosphatases. Enzymes involved in the synthesis of the phosphoinositides are indicated above the arrows. Dotted arrows represent possible reactions that have yet to be confirmed in vivo. The position of the lipid side chains are indicated (Ri and R2). K = kinase, P = phosphate, ptase = phosphatase. Adapted from (Rameh and Cantley 1999). 34 Many of the biological effects induced by PI3K appear to be mediated by activation of the serine/threonine (S/T) kinase, Akt. Akt, for example, has been shown to phosphorylate the pro-apoptotic protein, Bad. This phosphorylation allows the association of Bad with 14-3-3 and this association prevents the interaction of Bad with pro-survival members of the Bcl-2 family, thereby preventing apoptosis (Nunez and del Peso 1998). Akt also phosphorylates the Forkhead TF, FKHRL-1, and this reduces the ability of this TF to enter the nucleus and upregulate apoptotic genes like FasL and Bim. Activation of Akt also leads to an increase in the transcription of survival genes via activation of the TFs, N F K B (nuclear factor KB) and cAMP-response element-binding protein (CREB). As well, Akt activation has been shown to inactivate, via phosphorylation, the pro-apoptotic enzyme caspase-9 (Cardone, Roy et al. 1998; Fujita, Jinbo et al. 1999) and the S/T kinase, glycogen synthase kinase 3 (GSK-3). GSK-3 inactivation results in the promotion of cell cycle entry (reviewed in (Scalia, Heart et al. 2001)). An overly active PI3K pathway is associated with many different cancers (Ayala, Thompson et al. 2004; Thompson and Thompson 2004) and, as such, this pathway must normally be tightly regulated. The ubiquitously expressed, tumour suppressor PTEN (phosphatase and tensin homologue deleted on chromosome ten), a 3'-phosphatase, hydrolyzes PIP3 to Pl-4,5-P2 (removing the phosphate of the 3' position of the inositol ring), while the hematopoietic-restricted SH2-containing inositol 5'-phosphatase (SHIP or SHIP1), the stem cell-restricted SHIP (sSHIP) isoform, and the ubiquitously expressed SHIP2 remove the 5'-phosphate of PIP3 to yield PI-3,4-P2. The importance of damping the PI3K pathway via negative regulation is exemplified by the fact that nearly half of human cancers contain biallelic deletions or inactivating mutations of PTEN (Sansal and Sellers 2004). For non-hematopoietic cells, PTEN and SHIP2 are the primary enzymes that keep PIP3 levels under control while, in hematopoietic cells, SHIP appears to be key (Sly, Rauh et al. 2003). 35 1.5.1 SHIP SHIP (Inpp5D) was the first bone fide 'signalling' lipid phosphatase identified (i.e., it was not simply a housekeeping lipid phosphatase) (Damen, Liu et al. 1996; Lioubin, Algate et al. 1996; Osborne, Zenner et al. 1996). It was originally identified as a 145 kDa protein that became both tyrosine phosphorylated and associated with She upon activation of several growth factor and cytokine receptors (Damen, Liu et al. 1993; Damen, Liu et al. 1996; Liu, Damen et al. 1997). SHIP'S structure, as shown in Figure 1.6, consists of an N-terminal SH2 domain that preferentially binds phosphotyrosyl motifs with the consensus pY(Y/D)X(L/I/V). SHIP'S SH2 is known to bind the tyrosine phosphorylated forms of She, SHP-2 (formerly known as Syp), Dok, Gab, CD 150 (SLAM), and certain immunoreceptor tyrosine-based inhibitory and immunoreceptor tyrosine-based activation motifs (ITIMs and ITAMs, respectively) (Sly, Rauh et al. 2003). SHIP'S centrally located inositol phosphatase domain selectively hydrolyzes the 5'-phosphate from the lipid membrane component PIP3 and the cytosolic inositol-l,3,4,5-tetrakisphosphate (IP4). Although SHIP'S role in the hydrolysis of PIP3 has been extensively described, the functional consequences of IP4 hydrolysis in vivo are not currently known. Finally, the C-terminus of SHIP possess two NPXY sequences (PTB domain-binding motifs) that, when phosphorylated, bind the PTB domains of She and Dok2, and a critical proline-rich region that binds SH3-containing proteins including Grb2 and members of the Tec family (Sly, Rauh et al. 2003; Tomlinson, Heath et al. 2004). The C-terminal region of SHIP is also involved in the binding of cytoskeletal components such as filamin (Lesourne, Fridman et al. 2005). Proline-rich motifs near the N-terminus of SHIP may have significance in mediating protein-protein interactions and this is discussed in Chapter 4. SHIP is expressed in all hematopoietic cells including lymphocytes, NK cells, mature granulocytes, monocyte/macrophages, mast cells, platelets and erythroid progenitors (reviewed in Sly, Rauh et al. 2003). There is some evidence that SHIP is not expressed during the later stages of erythroid development (Kalesnikoff, Sly et al. 2003) and this is discussd further in Chapter 3. Related to this fact, the expression of SHIP protein varies considerably within the hematopoietic compartment; for example, SHIP expression increases during T cell maturation (Liu, Shalaby et al. 1998), shows a bimodal expression pattern during B cell development (Geier, Algate et al. 1997) and increases dramatically in activated B cells 36 (Brauweiler, Tamir et al. 2000). As a further complication, there are two alternatively spliced forms of SHIP (note that this is distinct from the sSHIP isoform) whose expression(s) vary during hematopoiesis (Lucas and Rohrschneider 1999; Wolf, Lucas et al. 2000). In addition, sSHIP is the only form of SHIP expressed in embryonic stem (ES) cells (Tu, Ninos et al. 2001) and may be co-expressed (albeit briefly) with full length SHIP in HSCs (Tu, Ninos et al. 2001). Current evidence indicates that the expression of sSHIP and full length SHIP may be mutually exclusive with sSHIP expression fading soon after HSC differentiation. sSHIP mRNA is transcribed from an alternative promoter between SHIP'S exons 5 and 6. The protein product is essentially an N-terminally truncated SHIP isoform that lacks a SH2 domain and, perhaps subsequently, is neither tyrosine phosphorylated nor associated with She following stimulation (Tu, Ninos et al. 2001). However, sSHIP does have an intact 5'-phosphatse catalytic domain and C-terminal tail that allows constitutive association with Grb2 (Tu, Ninos et al. 2001). Therefore, sSHIP may be recruited via Grb2's SH2 domain to the plasma membrane and thus may function to regulate PIP3 levels in stem cells (Tu, Ninos et al. 2001). However, the functional role of sSHIP in stem cells has yet to be elucidated. It is still not known how (or if) the tyrosine phosphorylation of SHIP affects its function. However, we and others have shown that tyrosine phosphorylation does not alter SHIP'S phosphatase activity in vitro (Damen, Liu et al. 1996). It is therefore likely, SHIP is regulated by compartmentilzation and the tyrosine phosphorylation of SHIP, via intermolecular binding interactions, may have a role in this regard. In fact, by using green fluorescent protein (GFP)-tagged SHIP or other immunofluorescent labeling techniques, SHIP has been shown to translocate to the plasma membrane, to membrane microdomains (lipid rafts) or receptor aggregates (e.g., BCR caps) in response to stimulation (Cox, Dale et al. 2001; Phee, Rodgers et al. 2001). The mechanisms of translocation appear to depend on the type of extracellular stimulus (reviewed in (Rauh, Sly et al. 2004); for example, SHIP may use its own SH2 domain to take to the tyrosine phosphorylated ITIM motif (pITIM) of FcyRIIB. Alternatively, SHIP may use an adaptor protein like She to take it to the p-subunit of the IL-3 receptor (IL-3RP) in mast cells, via She's PTB domain (Bone and Welham 2000; Velazquez, Gish et al. 2000) or to CD16 (FcyRIIIa), the low affinity receptor for the Fc fragment of immunoglobin G (IgG) on NK cells, via She's SH2 domain (Galandrini, Tassi et al. 2001; Galandrini, Tassi et al. 2002). 37 Many of the studies of the biological role of SHIP in hematopoietic cells use, as a model, a comparison of the immune cell responses from SHIP+/+ and -/- mice. This work has revealed a role for SHIP in the negative regulation of mast cell activation and adhesion to fibronectin (Kalesnikoff, Lam et al. 2002; Lam, Kalesnikoff et al. 2003); proliferation, chemotaxis and activation of B cells (reviewed in Sly, Rauh et al. 2003); neutrophil survival (Gardai, Whitlock et al. 2002), monocyte/macrophage phagocytosis (Cox, Dale et al. 2001) and early erythroid progenitor (BFU-E) colony formation (Mason 2002). PI3Ks PTEN 4 • 4 t -Ptase M 1 • II I 1 TENSIN DOMAIN 54 kD SHIP sSHIP SHIP2 - • P I - 3 , 4 - P 2 N P N Y N P L Y 1 S H 2 5'- Ptase Pro-rich N P N Y N P L Y V V 5'- Ptase Pro- rich N P A Y V I S H 2 5'- Ptase Pro- rich 145 kD 104 kD 150 kD Figure 1.6: Functional domains of SHIP, SHIP2 and PTEN. The structure of SHIP consists of an N-terminal SH2 domain, a centrally located catalytic domain and a proline-rich C-terminus with 2 NPXY PTB-binding motifs. SHIP also has several PXXP motifs capable of binding SH3 domains (not shown). SHIP and stem cell SHIP (sSHIP) are hematopoietic cell specfic inositol 5'-phosphatases. While sSHIP expression is limited to embryonic stem and hematopoietic stem cells, SHIP is expressed in more mature hematopoietic cells. sSHIP arises from an alternative start site within the full length SHIP gene and results in a protein product without an SH2 domain. SHIP2 and PTEN are ubiqutously expressed. The SHIP family and PTEN share the same phosphotidylinositol substrate (PI-3,4,5-P3) but their activities yield separate products with PTEN, a 3'-phosphatase, regenerating the pool of the PI3K substrate PI-4,5P2 whereas the product of SHIP is PI-3,4-P2. 38 1.5.2 SHIP2 SHIP2 (Inppll) is widely expressed in non-hematopoietic and hematopoietic tissues and is even co-expressed with SHIP in some hematopoietic cells. However, SHIP2 does not appear to play a dominant enzymatic role in hematopoietic cells based on studies in platelets (Giuriato, Pesesse et al. 2003) and the lack of hematopoietic defects in SHIP2-deficient mice (Clement, Krause et al. 2001; Sleeman, Wortley et al. 2005). SHIP2 is slightly larger than SHIP at 150 kDa, with a similar structure (Fig. 1.6) and, like SHIP and sSHIP, it hydrolyzes the 5'-phosphate from PIP3 and IP4. Like SHIP, SHIP2 is tyrosine phosphorylated and associates with She in response to extracellular stimuli. The greatest divergence in the amino acid sequence identity between SHIP and SHIP2 is within the proline-rich C-terminus such that SHIP2 does not bind the SH3 domain of Grb2 (Wisniewski, Strife et al. 1999). Since SHIP2's C-terminus is very different from that of SHIP, the compartmental regulation of SHIP and SHIP2 (Damen, Ware et al. 2001; Aman, Walk et al. 2000), and therefore the functions of these proteins, may not be redundant. 1.5.3 PTEN PTEN (MMAC1) is widely expressed in non-hematopoietic and hematopoietic cells (Gimm, Attie-Bitach et al. 2000; Luukko, Ylikorkala et al. 1999). Loss of expression or inactivating mutations of PTEN are very common in a variety of human cancers, highlighting its role as a tumour suppressor (as does its synonym, MMAC1 - for mutated in multiple advanced cancers) (Cantley and Neel 1999; Maehama and Dixon 1999; Parsons and Simpson 2003). The structure of PTEN (Fig. 1.6) includes a putative PI-4,5-P2 binding motif at its N -terminus, a 3'-phosphatase domain with specificity towards PI-3P's, particularly PIP3, and a C-terminal C2 domain that is thought to be involved in targeting PTEN to negatively charged phospholipids in the plasma membrane (Leslie and Downes 2002; Lee, Yang et al. 1999). PTEN also possesses a PDZ-binding motif that may facilitate its interaction with other proteins such as membrane-associated scaffold protiens (Wu, Hepner et al. 2000). PTEN has been shown to reside mainly in the cytosol but is also found in other cell compartments including the plasma membrane and, in some cells, the nucleus. However, there is no evidence to date for the regulation of PTEN function by compartmentalization (reviewed in (Leslie and Downes 2002)). The tumor suppressor properties of PTEN have 39 been attributed to its lipid phosphatase activity which, by primarly acting to reduce PIP3 levels, suppresses the activation of Akt, a known oncogene (Maehama and Dixon 1999; Myers, Pass et al. 1998; Leslie and Downes 2002). Although, PTEN-deficient mice are not viable (embryonic-lethal phenotype), studies using PTEN-/- ES cells reveal a significant increases in PIP3 levels in response to stimulation, compared to wild-type ES cells, and a constitutively active Akt (reviewed in (Krystal 2000)). PTEN heterozygote mice (PTEN+/-) are viable but frequently develop cancers and autoimmune disease (Parsons and Simpson 2003). Since PTEN-/- ES cells exhibit high basal levels of Akt activity (Parsons and Simpson 2003; Downward 2004), while basal levels of Akt in resting SHIP-/- BMMCs are only slightly elevated (Huber, Helgason et al. 1998), PTEN may act as a basal regulator of PIP3, while SHIP/SHIP2 are important extracellular-activated regulators of PIP3 and PI-3,4-P2 Importantly, PI-3,4-P2 is an important second messenger in its own right (Scheid, Huber et al. 2002). For example, PI-3,4-P2 may act attracts PH-containing proteins, such as Bam32 and TAPP2, that have a binding preference for PI-3,4-P2 to the plasma membrane (Marshall, Krahn et al. 2002; Krahn, Ma et al. 2004). In addition, both PIP3 and PI-3,4-P2 may be important for full activation of Akt (Scheid, Huber et al. 2002). 1.6 THE PHENOTYPE OF SHIP KNOCKOUT MICE The SHIP-/- mouse provides a valuable tool to study the role of SHIP in vivo. SHIP-/- mice are viable and fertile but have a shortened life span marked by several hematopoietic and physiological abnormalities. The phenotype of these mice is similar to PTEN +/- mice (Fox, Ung et al. 2002; Moody, Xu et al. 2004) and Lyn-/- mice (Hernandez-Hansen et al. 2004 ). SHIP-/- mice have a distinctly 'white' bone marrow accompanied by extramedullary hematopoieis that causes splenomegaly: spleen weights and total cellularity are increased 5-7 fold in SHIP-/- mice (Helgason, Damen et al. 1998). Furthermore, the lungs of SHIP-/-mice are enlarged due to infiltration by myeloid cells (macrophages and neutrophils). They also have increased granulocyte-macrophage (GM) progenitors in the bone marrow and the spleen and these progenitors are hyperresponsive to M-CSF, GM-CSF, IL-3 and SCF (Helgason, Damen et al. 1998). Like Lyn-/- B cells, SHIP-/- B cells have enhanced function and consequently SHIP-/- mice have elevated serum IgG and IgM levels (Helgason, Kalberer 40 et al. 2000). This immune hyperactivity predisposes SHIP-/- mice to a lupus-like autoimmune disease (Rauh and Krystal 2002). SHIP-/- mice are also osteoporotic (Takeshita, Namba et al. 2002) and have perturbed NK cell development (Wang, Howson et al. 2002). And finally, recent work by Rauh et al (unpublished) has shown that the chronic activation of the PI3K pathway in SHIP-/- mice skews SHIP-/- macrophage development away from an M l killer to an M2 'healing' phenotype. The first studies of a SHIP2-/- mice revealed a more severe phenotype than that observed with SHIP-/- mice - perinatal death apparently from insulin hypersensitivity-induced hypoglycemia (Clement, Krause et al. 2001). This difference in severity of the SHIP2-/- and SHIP-/- phenotypes was attributed to the fact that SHIP2 is expressed (albeit at low levels) in hematopoietic cells (Sly, Rauh et al. 2003), and therefore may partially compensate for the loss of SHIP in these cells, whereas SHIP is not expressed in non-hematopoietic tissues. Corroborating evidence for the observed phenotype of SHIP2 came from the discovery of a deletion in the 3' untranslated region (UTR) of SHIP2 in some type II diabetic patients (Marion, Kaisaki et al. 2002). The deletion may lead to overexpression of SHIP2 in these patients and could account for observed insulin hyposensitivity. However, these first studies of the SHIP2-/- phenotype in mice used a gene knock-out strategy that left the first 18 exons of SHIP2 intact (resulting in a severely truncated but not ablated gene product) and furthermore resulted in an unintended deletion of a second gene (Phox2a) (Clement, Krause et al. 2004). A more recent description of the phenotype of the SHIP2-/-mouse that is truly SHIP2 null (and without the secondary gene defect) has been reported (Sleeman, Wortley et al. 2005). In contrast to the original SHIP2 ablation studies, these mice are viable and fertile with normal glucose and insulin levels and tolerances. However, these mice resist weight gain when placed on a high fat diet. This latest study questions the previously reported role of SHIP2 in modulating glucose homeostasis and highlights the potential of SHIP2 as a specific target for the treatment of obesity. 41 1.7 SHIP AND HUMAN DISEASE Unlike the prevalent role of PTEN in human disease, the role of SHIP in human disease has proved to be less obvious. However, a subgroup of hyper-allergic people express substantially lower than normal levels of SHIP protein in their basophils (MacDonald and Vonakis 2002); this is interesting in light of SHIP'S role in regulating the threshold of mast cell activation (Huber, Kalesnikoff et al. 2002). Also, given that SHIP-/- mice are osteoporotic (Takeshita, Namba et al. 2002), the human SHIP gene maps to q36/37 of chromosome 2 (Ware, Rosten et al. 1996), the same locus associated with a group of patients with Paget's disease (also called osteitis deformans), an osteoporotic-like disorder. Most importantly, an inactivating mutation in the catalytic domain of human SHIP has recently been found within the blast cells of a patient with acute myelogenous leukemia (AML) (Luo, Yoshida et al. 2003). This mutation replaces a valine with a glycine in motif 2 of the 5-phosphatase domain and in vitro analysis has shown that this mutant results in reduced 5'-phosphatase activity. Leukemic cells with this mutation display enhanced Akt phosphorylation following IL-3 stimulation. Furthermore, exogenous expression of this putative dominant-negative SHIP in K562 cells (a human erythroleukemic cell line lacking SHIP) allowed for increased survival of these cells in the absence of serum and a growth advantage under low serum conditions. The presence of such a mutant in AML patients suggests that SHIP may act as a tumor suppressor in hematopoietic progenitors. The putative role of SHIP in human disease processes was the impetus for our studies described in Chapter 5. 42 1.8 AIMS OF STUDY As mentioned earlier, SHIP is a potent negative regulator of the PI3K pathway in hematopoietic cells and in its absence there is a marked increase in myeloid progenitors and mature granulocytes and monocyte/macrophages. However, even though activation of the PI3K pathway has been shown to prevent apoptosis of normal erythroid progenitors (Bouscary, Pene et al. 2003), SHIP-/- mice are neither polycythemic nor profoundly anemic. One of our aims, therefore, was to look into why erythropoiesis was not significantly affected in SHIP-/- mice and the results of this investigation are described in Chapter 3. Although our lab and others have shown that SHIP becomes tyrosine phosphorylated in response to a diverse array of extracellular stimuli, the kinase(s) responsible for this phosphorylation has not been identified. Furthermore, the functional role of the tyrosine phosphorylation of SHIP is not well understood. The results of these studies are described in Chapter 4. Lastly, our studies with SHIP-/- mice have provided substantial insights into the signalling and biological functions of murine SHIP. However, it has been difficult to assess the function of human SHIP within human hematopoietic cells and so little work has been carried out in this area. However, with the recent advent of siRNA technology this has now become possible and our third aim, using siRNA to human SHIP, was to explore the role of SHIP in the human erythroleukemic cell line, TF-1. The results of these studies are described in Chapter 5. 43 Chapter 2 MATERIALS AND METHODS 2.1 TISSUE CULTURE 2.1.1 Immortalized cell lines The murine cell lines Ba/F3 (from Dr. A. Miyajima, University of Tokyo) and Ba/F3 with exogenously expressed EpoRs (BaER) (Damen, Liu et al. 1993) were maintained in RPMI 1640 medium (StemCell Technologies, Vancouver, BC) with 10% (v/v) fetal calf serum (FCS), 5 ng/ml murine IL-3, 5 xlO"5 M P-mercaptoethanol (P-ME) with penicillin (100 U/ml) and streptomycin (100 ug/ml) (P/S) (StemCell Technologies). The murine B cell line WEHI 231 (from Dr. Mike Gold, University of British Columbia, Vancouver, BC) and the human B cell line BJAB (from Dr. Vincent Duronio, Jack Bell Research Centre, Vancouver, BC) were grown in RPMI 1640 with 10% (v/v) bovine growth serum and P/S. The human leukemic cell line TF-1 (from Dr. Trang Hoang, McGill University) was grown in Iscove's modified Dulbecco's medium (IMDM) (StemCell Technologies) containing 10% (v/v) FCS, 5 ng/ml human GM-CSF and P/S (TF-1 growth medium). 2.1.2 Bone marrow-derived mast cells (BMMCs) B M cells were aspirated from the femurs and tibias of 4-8 week old SHIP+/+ and -/-and Lyn+/+ and -/- (gift from Dr. Janet Oliver, University of New Mexico) C57BL/6 littermates using IMDM with P/S as described previously (Huber et al., 1998; Leitges et al., 2001). B M cells were plated at lxlO 6 cells/ml in IMDM containing 10 ng/ml murine IL-3, 10 ng/ml human IL-6, 50 ng/ml murine SCF, and 3 units/ml of human Epo for 7 days. Non-adherent cells were then harvested and resuspended in IMDM containing 15% FCS, P/S, 150uM monothioglycerol (MTG) (Sigma, St. Louis MO) and 30 ng/ml murine IL-3 (StemCell Technologies Inc). Cell cultures were maintained between 2 x 105 and 8 x 105 cells/ml with complete medium replacement every 2 weeks. After 6 weeks in culture, the surface expression of Fc £ Rl (IgE receptor) and c-kit were assessed by flow cytometry using fluorescein isothiocyanate (FITC)-labelled anti-c-kit (BD Pharmingen, Mississauga, ON) and FITC-labelled IgE (anti-Epo 26; StemCell Technologies) antibodies. 44 2.1.3 Bone marrow-derived macrophages (BMm<Ds) B M cells were aspirated from the femurs and tibias of 4-8 week old SHIP+/+ and -/-mice with IMDM containing 10% FCS (v/v) and P/S. Cells were plated in a 175cm flask for 3 hr to allow mature cells to adhere at 37°C and 5%CC>2. The supernatants were removed and resuspended in IMDM containing 10% FCS (v/v), 5% C127 conditioned medium (as a source of M-CSF) and P/S. These cells were allowed to mature for 10-12 days with medium changes at days 5 and 10. 2.2 siRNA TRANSFECTION TF-1 cells were maintained at 2-4 x 105 cells/ml in TF-1 growth medium and resuspended in fresh growth medium at 2 x 105 cells/ml 24 hrs before transfection. On the day of transfection, cells were seeded in 12 well plates at 2.5 - 3 x 104 cells/well in 600 ul TF-1 growth medium. The siRNA transfection reagent was freshly prepared for each transfection in microfuge tubes by diluting 1 ug of the appropriate siRNA (Qiagen) in 90 pi EC-R buffer (Qiagen) (for 3x104 cells). RNAifect™ transfection reagent was added and to the diluted siRNA vortexed for 10 sec and then left to stand at 23°C for 15 min to allow siRNA:RNAifect™ complex formation. Complexes were added drop-wise onto cells, swirled gently and then incubated for 48 hr at 37°C. Before each experiment, a sample of the transfection reaction was assayed for gene silencing either by Western blotting or by intracellular labeling with anti-SHIP antibodies and flow cytometry. The target sequences of the siRNA were generated by Qiagen and are as follows: siSHIP r(GAGUCAGGAAGGAGAAAAU)d(TT) r(AUUUUCUCCUUCCUGACUC)d(TT) siShc r(CUACUUGGUUCGGUACAUGUU)d(TT) r(CAUGUACCGAACCAAGUAGGA)d(TT) siNS r(UUCUCCGAACGUGUCACGU)d(TT) 45 2.3 ASSESSMENT OF ERYTHROPOIESIS WITHIN SHIP+/+ AND -/- MICE 2.3.1 Mice SHIP+/+ and -/- mice (on a mixed 129JxC57BL/6 background, generation N=2) were housed in the Joint Animal Facility (JAF) of the BC Cancer Research Centre. Femurs and tibias extracted from Lyn+/+ and Lyn-/- mice (on a similar mixed background) or BMMCs from these mice were provide by Dr. Janet Oliver (University of New Mexico, USA). At the completion of experiments, mice were humanely euthanized by C 0 2 asphyxiation. 2.3.2 Phenylhydrazine-induced anemia Stock phenylhydrazine (PHz) was prepared at a concentration of 6 mg/ml in RPMI and filter sterilized (0.2pm). Age and sex matched SHIP+/+ and -/- mice from the same colony (or littermates where indicated) were injected intraperitoneally on 2 consecutive days with PHz to give a final dose of 60 mg/kg mouse. The first day of treatment was designated Day 0. 2.3.3 Determination of hematocrit and reticulocyte index Blood for hematocrit measurements was collected from tail veins into heparinized capillary tubes (Fisher, Pittsburg, PA) and centrifuged for 5 min in a Readacrit® centrifuge (Becton Dickenson). The packed RBC/blood volume ratio was determined manually. Reticulocyte analysis was performed from tail vein bleeds or from blood obtained from cardiac punctures (see below) by flow cytometry using thiazol orange. The thiazole orange working solution was freshly prepared for each analysis from a stock solution (1 mg/ml in methanol stored at -20°C). A small aliquot (3 pi) of blood was incubated in 2 ml of thiazole solution or PBS alone for 30 min at 23°C and then analyzed immediately or refrigerated and analyzed within 6 hr. The reticulocyte index was calculated based on the hematocrit and the number of reticulocytes assuming a desired hematocrit of 0.45. Reticulocyte index = % reticulocytes X Hct/0.45 (Riley, Ben-Ezra et al, 2002). 46 2.3.4 Plasma collection Immediately after sacrifice, whole blood was aseptically obtained by cardiac puncture using a 22 gauge needle. Blood was transferred into a heparinized collection tube, mixed several times by inversion, and then centrifuged at 3500 rpm for 5 min in a Heraeus biofuge. The plasma was transferred into a new tube and frozen. 2.3.5 Cell isolation and nucleated cell counts Femurs from mice were harvested and cells from individual femurs extracted. To obtain nucleated counts, an aliquot of the cell preparation was diluted (1:100) in 3% acetic acid (v/v) in PBS and loaded onto a Neubauer counting chamber or a Reichert hemocytometer (Buffalo, NY). The spleen from each animal was weighed and the cells extracted by homogenization in either in IMDM with 2% FCS or in PBS. A nucleated cell count was determined as above. To remove excess RBCs, spleen cell suspensions were incubated with ammonium chloride (NH4C1) (StemCell Technologies) buffer solution (0.8%, O.lmM EDTA) (1 vol cells: 4 vol NH4CI) for 10 min on ice. This suspension was centrifuged and the supernatant containing the lysed RBCs decanted. The remaining spleen cells were passed through a 100pm nylon cell strainer and treated with deoxyribonuclease I (DNase I). A final cell count was performed as above to calculate the total nucleated cells/spleen. 2.3.6 Progenitor Analysis Clonogenic progenitors of the erythroid lineage were detected in methylcellulose-based cultures. For CFU-E and the more primitive BFU-E, MethoCult™ 3334 was used (a methylcellulose based medium containing human insulin, human transferin and recombinant human Epo (StemCell Technologies)). Cells were added to tubes containing MethoCult™ 3334 at 2 x 10s - 4 x 105 cells/dish for both B M and spleen cells. On Days 2 and 3 following PHz treatment, cells were plated at 0.5 x 105 -2 x 105 cells/dish for both B M and spleen cells. Clonogenic progenitors of the erythroid, myeloid (CFU-granulocyte-monocyte (GM)) and multi-potential (CFU-GEMM) lineages were detected in methylcellulose-based cultures using MethoCult™ 3434 (medium containing recombinant murine SCF (50 ng/ml), human IL-6 (10 ng/ml), murine IL-3 (10 ng/ml) and human Epo (3 U/ml) (StemCell Technologies, 47 Inc.). All cultures were set up in triplicate and were incubated in a humidified 5% C 0 2 incubator at 37°C, for 2 days to enumerate mature CFU-E (in MethoCult™ 3334) and for 8-10 days to enumerate primitive BFU-E (in MethoCult™ 3434). Myeloid and multi-potential progenitors were assessed in MethoCult™ 3434 after 8-10 days. Clonogenic progenitors of the megakaryocytic lineage (CFU-Mk) were detected in serum-free collagen-based cultures containing MegaCult-C™ (a medium containing recombinant human Tpo (50 ng/ml), human IL-11 (50 ng/ml), mouse IL-3 (10 ng/ml) and human IL-6 (20 ng/ml) (StemCell Technologies). B M cells from each mouse were added to tubes containing the MegaCult-C™ to give 1 x 105 cells/slide for each time point tested. Spleen cells from each mouse were added to tubes containing MegaCult-C1M to give 5xl0 3 cells/slide for all the time points tested. Collagen was then added and the tubes vortexed and the contents dispensed into double chamber slides. All cultures were set up in triplicate and were incubated for 8-9 days at 37°C in a 5% C O 2 incubator. Following incubation, the chambers were assessed microscopically for colony formation prior to dehydration and fixation of the slides. Using a staining protocol to detect acetylcholinesterase activity, the colonies on the slide were stained with acetylthiocholiniodide as described in the StemCell Technical Manual, "Assays for the Quantitation of Human and Murine Megakaryocytic Progenitors", section 17. The colonies were divided into the following categories, based on size and morphology; CFU-Mk (3-30), CFU-Mk (> 30), CFU-Mk Mixed and Non-CFU-Mk. To assess the total progenitor content/femur, the following equation was employed: Mean CFC per dish X Total nucleated cell count/ femur Cells plated in each culture To assess the total progenitor content/spleen, the following equation was employed: Mean CFC/slide X Total nucleated cell count/spleen before lysis/ total nucleated count after lysis xlOO") Cells plated in each culture Standard t-tests were performed to assess if there was a difference in the number of colonies generated between the normal and knockout genotypes. Due to the potential subjectivity of colony enumeration, only a P value of < 0.01 was deemed significant. 48 2.3.7 CD71+ purification CD71+ (transferring receptor positive) cells were isolated from NEUCl-treated spleen cells (prepared as above from Day 3 or Day 4 PHz-treated mice) by immunomagnetic positive selection (EasySep™ biotin selection kit, StemCell Technologies) based on the manufacturer's instructions. Briefly, the Fc receptors were blocked on the spleen cells with anti-FcR antibody (2.4G2) and the cells incubated for 15 min at 23°C with 2.5pg of anti-CD71 -biotin/1 x 10s cells in 5ml polystyrene round bottom tubes (Falcon®, Beckton Dickinson). Anti-biotin tetrameric antibody complexes (100 pi) were then incubated for 15 min at 23°C. Next, magnetic nanoparticles (50pl/l x 10s cells) were added and incubated for 15 min at 23°C. Lastly, cells were placed in an EasySep™ magnet for positive selection, where CD71+ cells remained in the tube and CD7T cells were washed away. Cell yields were routinely between 1.5-3.0 x 107 cells/SHIP+/+ spleen and 4.0-5.5 x 107 cells/SHIP-/-spleen (with Day 3 PHz-treated mice). 2.3.8 Murine hematopoietic enrichment (Lin" selection) Lineage depletion of MLCl-treated spleens was performed using StemSep™ negative selection. Specifically, cells were labelled with a murine hematopoietic enrichment cocktail (containing biotinylated antibodies to lineage markers: CD5 (Ly-1), Erythroid Cells (TER119), CD45R (B220), Ly-6G (Gr-1), CD1 lb (Mac-1), Neutrophils (7-4) using 16pl/5 x 107cells) or a custom cocktail where TER119 was omitted. Anti-biotin tetrameric antibody complexes were added (100pl/5 x 107 cells) and incubated at 4°C for 15 min. Next, magnetic colloid was added (60pl/5 x 107 cells) and incubated for 15 min at 4°C. Cells were then loaded onto a magnetic column for separation. Typical cell yields were 2-4 x 106cells/spleen. Enriched cells were assayed for lineage markers by flow cytometry using the biotinylated antibodies in the custom cocktail described above (or a biotin-labelled rat IgGi isotype control) and PE-TER119 (or a PE-labelled isotype control) as primary antibodies. Streptavidin-AlexaFluor488 (SA-488) (Molecular Probes, Eugene, OR) was then added for detection of the biotinylated primary antibody. 49 2.4 PROTEIN ANALYSIS 2.4.1 Cell Stimulations, immunoprecipitations, and immunoblotting All cells were maintained in liquid suspension culture as specified in Section 2.1.1, 2.1.2 and 2.1.3 and were then growth factor-deprived for up to 18 hr at 37°C (with the exception of WEHI 231 and BJAB cells which were washed into starvation medium and stimulated immediately) in their normal culture medium with 0.1% (v/v) bovine serum albumin (BSA) or 10% heat-inactivated FCS (AFCS), in the case of TF-1 cells. Al l cells were resuspended in fresh culture medium with 0.1% (v/v) BSA (10-50 x 106 cells/ml) and stimulated as indicated. Cells were then either washed once with ice-cold lx PBS (StemCell Technologies) and solubilized at 1-5 x 10 cells/ml in phosphorylation solubilization buffer (PSB; 50 mM HEPES, pH 7.4, 100 mM sodium fluoride, 10 mM sodium pyrophosphate, 2 mM sodium orthovanadate, 4 mM EGTA, 2 mM phenylmethyl sulfonyl fluoride, 10 pg/ml leupeptin, 2 ug/ml aprotinin) with 1% (v/v) Triton-X 100 (TX-100) at 4°C for 1 hr for immunoprecipitation studies or pelleted and lysed with lx sample buffer (4 x SDS sample buffer: 34% (v/v) glycerol, 2% (w/v) SDS, 2.84M 2-mercaptoethanol) for total cell lysate (TLC) analysis. Insoluble material was pelleted by centrifugation (13000 rpm for 5 min at 4°C in a Heraeus biofuge and the supernatants transferred to new tubes. Immunoprecipitations "were performed as follows: Antibodies (0.5 to 2 ug/reaction) were preincubated for 2 hr at 4°C with protein G-conjugated agarose beads (Pierce, Rockford, IL) in 500 pi PSB containing 1% BSA (v/v) and 0.1% TX-100, using a Nutator™. The antibody/bead complexes were then washed twice and the cell lysates, pre-cleared with protein G agarose beads where indicated, were added and incubated with rotation for 3 hr to overnight at 4°C. The agarose beads were then washed 3 times with 1 ml of PSB + 0.1% (v/v) TX-100, resuspended in 120 pi lx SDS sample buffer and heated for 3 min at 95°C. Samples were then fractionated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) using the methods and equipment included with the Protean II™ electrophoresis system (BioRad) and the proteins from these gels transferred to Immobilon™ polyvinylidene fluoride (PVDF) membranes using a TransBlot Cell™ transfer system (BioRad) as per the manufacturer's instructions. Immunoblot analysis was carried out by first blocking the PVDF membranes with 5% (w/v) BSA dissolved in PSB for 1 hr at 23°C or overnight at 4 °C. The blot was then washed 50 twice with TBST (lOx TBS: lOOmM Tris-HCl, pH 7.4, 1.5 M sodium chloride, 214 mM potassium chloride with 0.05% (v/v) Tween-20), then incubated with the appropriate concentration of primary antibody diluted in TBST with 10% (v/v) BSA for 1 hr at 23°C. Membranes were then thoroughly washed and incubated with secondary antibodies in TBST containing donkey anti-mouse IgG, donkey anti-rabbit IgG or donkey anti-rat IgG antibodies (Jackson Immuno Research Laboratories, Inc., West Grove, PA) directly conjugated to horseradish peroxidase (HRP) for 45 min at 23°C. Blots were then washed 5 times (5 min each), treated with Western Lightning chemiluminescence reagent (Perkin Elmer, Boston, MA) for 1 min, and exposed to Kodak X-Omat Blue film. 2.4.2 Inhibitors and Blocking antibodies Pharmacological inhibitors For inhibitor studies, inhibitors were added at the indicated concentrations 20-30 min before the addition of the stimulus or just prior to adding the cells for survival studies. The inhibitors used were the PI3K inhibitor LY294002, the inactive analogue LY303511, the Src kinase inhibitor PP2 and its inactive analogue PP3, and the spleen tyrosine kinase (Syk) inhibitor piceatannol (Calbiochem, San Diego, CA). The working concentrations for each experiment are indicated in the figure legends. All of these inhibitors were dissolved in dimethylsulfoxide (DMSO) and in cases were an inactive analogue was not used as a control, DMSO was used as a vehicle control. Neutralizing Epo in mouse plasma The blocking antibody a-Epo 16 (StemCell Technologies, Vancouver, BC) was added at 200 pg/ml with either Epo or mouse plasma for 30 min before addition of the cells. 51 2.4.3 Antibodies Antibody Type Cone./ Dilution Application Source Actin mAb (Ms) 1:1000 W Santa Cruz Biotechnology (Santa Cruz, CA) Akt pAb (Rb) 1:1000 W Cell Signaling Technologies (Beverly, MA) CD71-biotin rat IgG i 2.5ug/ml Flow Cytometry/FACS BD Pharmingen (Mississauga,ON) EpoR (M-20) pAb (Rb) 1:500 W Santa Cruz Biotechnology (Santa Cruz, CA) FITC-a-human IgM F(ab')2 pAb (goat) 20 ug/ml Stimulation/IF Jackson ImmunoResearch (West Grove, PA) FITC-a-human IgM, intact (whole molecule) pAb (goat) 20ug/ml Stimulation/IF Jackson ImmunoResearch (West Grove, PA) FITC-mouse-IgG mAb (Ms) 30ug/ml Flow Cytometry Dr. Peter Lansdorp (UBC, Vancouver, BC) GAPDH mAb (Ms) 1:15000 W Research Diagnostic Inc. (Flanders, NJ) GATA-1 mAb (rat) 1:1000 W Santa Cruz Biotechnology (Santa Cruz, CA) Grb-2 mAb (Ms) 1:1000 W Chemicon International (Temecula, CA) HA mAb (Ms) 1:1000 W Santa Cruz Biotechnology (Santa Cruz, CA) Lyn (44) pAb (Rb) 1:1000 W Santa Cruz Biotechnology (Santa Cruz, CA) Lyn (H6) mAb (Ms) 1:50(IP) IP Santa Cruz Biotechnology (Santa Cruz, CA) Mcl-1 pAb (Rb) 1:1000 W Santa Cruz Biotechnology (Santa Cruz, CA) PARP pAb (Rb) 1:1000 w Cell Signaling Technologies (Beverly, MA) PE-a-rat-IgG mAb (Ms) 30ug/ml Flow Cytometry Jackson Laboratories (West Grove, PA) 52 PTEN mAb (Ms) 1:500 W Santa Cruz Biotechnology (Santa Cruz, CA) SHIP (C+S) pAb (Rb) 1:2000 (W) W Krystal Lab SHIP(PICI) mAb (Ms) 1:50(IP); 1:1000 (W) IP, W Santa Cruz Biotechnology (Santa Cruz, CA) SHIP2 pAb (Rb) 1:2000 W Clarkson Lab STAT5 pAb (Rb) 1:1000 W Santa Cruz Biotechnology (Santa Cruz, CA) STAT5a pAb (Rb) 1:1000 W Santa Cruz Biotechnology (Santa Cruz, CA) STAT5b mAb (Ms) 1:1000 W Santa Cruz Biotechnology (Santa Cruz, CA) TER119-biotin rat IgG2a 0.5ug/ml Flow Cytometry/FACS BD Pharmingen (Mississauga,ON) TER 119-PE rat IgG2a 0.5ug/ml Flow Cytometry/FACS StemCell Technologies Inc. (Vancouver, BC) P-tubulin mAb (IgG2b) 1:1000 W Chemicon International 2.4G2 mAb (Ms) lOug/ml Flow Cytometry Dr. S. Szilvassy, (Vancouver, BC) Phospho-Specific Antibodies Antibody Type Cone./ Dilution Procedure Source 4G10 pan-specific anti-phosphotyrosine antibody mAb (IgG,) 0.5ug/ml W Cell Signaling Technologies (Beverly, MA) pAkt (S473) pAb (Rb) 1:1000 W Cell Signaling Technologies (Beverly, MA) pAkt (T308) pAb (Rb) 1:1000 w Cell Signaling Technologies (Beverly, MA) pFcyRIIb(Y292) pAb (Rb) 1:1000 w Cell Signaling Technologies (Beverly, MA) pGSK3p (S9) pAb (Rb) 1:1000 w Cell Signaling Technologies (Beverly, MA) pLyn (Y507) pAb (Rb) 1:1000 w Cell Signaling Technologies (Beverly, MA) 53 pPLCy! (Y783) pAb (Rb) 1:1000 W Cell Signaling Technologies (Beverly, MA) pPLCy2 (Y1217) pAb (Rb) 1:1000 W Cell Signaling Technologies (Beverly, MA) pp44/42 MAPK (pERKl/2) (T202/Y204) mAb (Ms) 1:1000 W New England BioLabs (Beverlay, MA) pShc (Y239/240) pAb (Rb) 1:2000 W Cell Signaling Technologies (Beverly, MA) pSHIP(Y1022) (NPXY2) pAb (Rb) 1:5000 W, Flow Cytometry, IF Krystal lab (Vancouver, BC) pSHIP(Y915)(NPXYl) pAb (Rb) 1:5000 W Krystal lab pSHIP (Y867) pAb (Rb) 1:2000 W Krystal lab pSTAT5 (Y694) pAb (Rb) 1:1000 W Cell Signaling Technologies (Beverly, MA) Table 2.1 List of antibodies used in this thesis W= Western, IP= Immunoprecipitation, IF= Immunofluorescence, p= phospho, pAb= polyclonal antibody, mAb= monoclonal antibody, Rb= rabbit, Ms= mouse 2.5 FLOW CYTOMETRY AND FLUORESCENCE ACTIVATED C E L L SORTING (FACS). BMMCs were resuspended in HFN (Hank's balanced salt solution (HBSS) containing 2% (v/v) FBS and 0.08% (v/v) sodium azide) at 1 x 107cells/ml in Falcon 2058 test tubes (Becton Dickinson, Franklin Lakes, NJ). Directly conjugated FITC- or phycoerythrin (PE)-labelled antibodies were added at the indicated concentration to 100 pi of cells and incubated for 30 min on ice. Cells were then washed twice with 4 ml of HFN and resuspended in 500 pi of HFN + 1 ul/ml propidium iodide (P.I.). Cells were analyzed on a FACScalibur™ (Becton Dickinson, Franklin Lakes, NJ) and analyzed using FlowJo™ software (TreeStar, Ashland, OR). For analysis of GFP-expressing cells, cells were washed twice with HFN and resuspended in 500 pi of HFN + 1 pi /ml propidium iodide. Spleen cells were resuspended in HFN at 2.5 x 106cells/ml containing 5% rat serum and incubated for 15 min at 4°C before the addition PE-labelled TER119 and CD71 -biotin or 54 biotinylated lineage markers (Lin) for 30 min at 4°C. Cells were then washed twice with 4 ml of HFN, resuspended in 200 ul of HFN containing SA-488, and then incubated at 4°C for 30 min. Cells were then washed twice with HFN and resuspended in 500 ul of HFN for analysis on a FACScalibur™. Alternatively, CD71-SA488 and TER119-PE-labelled cells were sorted based on receptor expression on a FACSAria™. For intracellular staining of BJABs, TF-1 and BMMCs, cells were washed once with PBS and resuspended in 50ul PBS and 50ul of the Caltag™ Reagent A was added for 15 min at 23°C. The cells were then washed once with HFN and resuspended in 50 ul of Caltag™ Reagent B with 5% goat serum for 15 min at 4°C. Primary antibodies were then added at the appropriate concentration (see Table 2.1) for 20 min at 4°C. Cells were then washed twice with HFN, resuspended in 100 pi of HFN containing the secondary antibody, and incubated at 23°C for 15 min in the dark. Lastly, cells were washed twice and resuspended in 500 pi HFN for analysis on a FACScalibur™. To measure internalized BCR, BJABs were washed in RPMI, resuspended in RPMI + 0.1%BSA (v/v) at 2xl0 5 cells/ml ± PP2 or PP3 (Calbiochem, Cat#s 529573 & 529574) and then incubated for 30min at 37°C. The cells were then preloaded with FITC-algM at 4°C for lhour followed by warming at 37°C for the indicated times. Cells were washed once in ice cold PBS with 0.05% azide followed by a wash in 1 ml 4°C 150 mM NaCl and by an acid wash in 1 ml 4°C 0.29% acetic acid in 150 mM NaCl (pH 3.3). The cells were then immediately resuspended in 1 ml PBS, spun down, washed once more with PBS and resuspended in 500pl HFN for analysis by flow cytometry. A no acid wash control was also performed to provide a control of maximal fluorescence. 2.6 BIOLOGICAL ANALYSIS OF CELLS 2.6.1 Survival and Differentiation Studies TF-1 cells were set up at 3- 4 x 104 cells/well in 0.1% BSA in flat bottom 96 well plates (total volume 200 ul/well) containing the test substance, (e.g., GM-CSF) as indicated. Viability was assessed by trypan blue exclusion. The human cytokines, IL-3, SCF, GM-CSF, IL-5, T N F a and Epo were from StemCell Technologies. 55 To assess the differentiation of TF-1 cells in response to Epo, cells were washed free of TF-1 cell growth medium and starved for 24 hrs in 10% heat-inactivated FCS. Differentiation was assessed by benzidine staining. 2.6.2 3H-thymidine assays Two days after siRNA transfection, TF-1 cells were washed twice and resuspended in IMDM and 10% heat inactivated FCS and seeded at 2 x 103 cells/ml in a 96 well round bottom plate (total volume 100 ul/well) containing the indicated cytokines/growth factors. After 42 hr incubation, the TF-1 cells were labelled with 20 ul of a 25 Ci/ml solution of tritiated thymidine (3H-Tdr) in IMDM to give a final concentration of 0.5 uCi/well. After another 6 hr at 37°C in a 5% C O 2 incubator the contents of each well were harvested onto filtermats and counted using an LKB Betaplate Harvester and Liquid Scintillation Counter (LKB Wallac, Turku, Finland). NH4Cl-treated spleen cells were washed and resupended in RPMI at l x l 0 6 cells/ml with 20% heat-inactivated FCS and 2 nM p-ME and seeded into a 96 well round bottom plate (total volume 100 pl/well) for Epo dose response studies. Proliferation assays using H-Tdr uptake into spleen cells were also performed in the presence of plasma collected from SHIP+/+ and -/- mice pre- and post-PHz injection. Spleen cells were incubated at 37°C in a 5% C 0 2 incubator for 18 hr followed by 6 hr incubation with 0.5 uCi/well of 3H-Tdr and harvesting onto filtermats. BaER cells were also used to assess the proliferative potential of the mouse plasma. Cells were seeded at 2.5 x 104 cells/well in the presence of increasing concentrations of mouse plasma obtained from untreated SHIP+/+ and -/- mice or mice treated with PHz for the times indicated in the figure legends. Plates were incubated at 37°C in a 5% C O 2 incubator for 18 hr followed by 6 hr incubation with 0.5 pCi/well. 3H-Tdr-labelled plates were harvested and counted as described above. 2.7 SHIP MUTANTS AND PHOSPHO-SHIP ANTIBODIES SHIP point mutations were generated by Dr. Jackie Damen and Dr. Mark Ware (Damen et al., 2001). Constructs consisted of a hemagglutinin (HA) tag at the N-terminus and a green fluorescent protein (GFP) tag at the C-terminus of SHIP. B M from 5-56 fluorouracil treated SHIP-/- mice were infected with MSCV containing wild-type or SHIP mutants (as described in Damen et al, 2001) and the cells were maintained in 2 ug/ml puromycin and checked for GFP expression prior to use for stimulations and immunoprecipitations. Both wild-type (WT) SHIP and a mutant in which the tyrosines (Ys) in the two NPXY motifs of SHIP were converted to phenylalanines (F) (2NPXF SHIP) were used in these experiments. Rabbits were immunized with phosphopeptides corresponding to the regions encompassing the two NPXpY peptides (Y 9 1 5 and Y 1 0 2 2 in human SHIP). The amino acid residues used were identical in human and murine cells and the resulting sera were immunogen-affinity purified. Rabbits were also immunized with a phosphopeptide corresponding to the region encompassing Y 8 6 7 in murine SHIP and purified in the same way. The NPXF mutants, as well as BMMC and WT mast cells, were used to characterize the antibodies generated, as the NPXF mutant cannot be phosphorylated at Y 9 1 5 and Y 1 0 2 2 but can be phosphorylated at Y . 2.8 BINDING OF SHIP TO PHOSPHORYLATED ITIM BEADS BMMCs were starved and left unstimulated or stimulated with 100 ng/ml SCF for 2 min at 37°C. Cell lysates were precleared with streptavidin (SA)-agarose beads and then incubated with limiting amounts of pITIM-biotin (biotinylated phosphopeptide 9 8 9 corresponding to ITIM consensus of Fey RUB (biotin-EAENTITpY SLLKH) pre-bound to SA-agarose beads) pre-bound to SA-agarose beads. pITIM beads were allowed to interact with cell lysates at 4°C for 1 hr. By using limiting amounts of beads, the unphosphorylated and phosphorylated forms of SHIP have to compete for the limited pITIM sites. The beads were then washed thoroughly, boiled in 1 x SB and fractionated on SDS-PAGE for Western analysis. The pSHIP/SHIP ratio was calculated based on densitometry values. 2.9 GLUTATHIONE S-TRANSFERASE (GST) FUSION PROTEINS 2.9.1 GST Fusion Protein Constructs GST fusion proteins containing regions fused in-frame behind the 27 kDa coding region GST were used in the studies described in chapter 4 of this thesis. Many of these proteins were created by others and were generously donated for these studies: 57 Protein Source Lyn SH3 Fyn SH3 Dr. F. Jirik (University of Calgary) Dr. F. Jirik Yes SH3 Dr. F. Jirik SHIP SH2+PXXP Dr. L. Liu (Liu et al., 1997) The remainder of the GST-fusion proteins described in this thesis were produced by Dr. Mark Ware by subcloning PCR products corresponding to the coding regions of the protein fragments of interest into the multi-cloning site 3' to the GST open reading frame of the pGEX-5X3 plasmid (Pharmacia/LKB). 2.9.2 In vitro Protein Binding Assay Cells were stimulated and solubilized as described in Section 2.4.1 at 1-3x10 cells/ml. Cell lysates were pre-incubated with 2-5pg of GST protein bound to GSH beads for 4-16 hrs at 4°C to remove non-specific protein binding partners, then removed from the GST beads, and incubated with 2-5pg of individual GST fusion proteins immobilized on reduced glutathione beads for 20 min. at 4°C. Finally, beads were washed 3x with PSB containing 0.1% (v/v) NP-40 or Triton X-100 and proteins size separated and visualized by SDS-PAGE and immunoblotting. 58 Chapter 3 CHARACTERIZATION OF ERYTHROPOIESIS IN THE SHIP KNOCKOUT MOUSE 3.1 INTRODUCTION In 1998 our laboratory generated a SHIP-/- mouse, in collaboration with Dr. Keith Humphries, by deleting SHIP's first exon (Helgason, Damen et al. 1998). Although these mice are viable they have a shortened lifespan, likely due to a massive infiltration of neutrophils and macrophages into their lungs (Helgason, Damen et al. 1998). Related to this, SHIP-/- mice overproduce granulocytes and macrophages, apparently forcing erythropoiesis out of the B M (which causes the B M to appear white) and into the spleen and elsewhere (Helgason, Damen et al. 1998). As a result, these mice suffer from progressive splenomegaly, due to both extramedullary erythropoiesis and myelopoiesis. Interestingly, much of the phenotype of the SHIP-/- mouse appears to result from the enhanced ability of their myeloid progenitors to survive and proliferate (Helgason, Damen et al. 1998; Liu, Sasaki et al. 1999). For example, the large pool of Macl + myeloid cells in these mice secrete high levels of IL-6 into the plasma and this not only reduces the level of B cells as the mice age but further enhances myeloid cell development (Nakamura, Kouro et al. 2004). However, even, though activation of the PI3K pathway has been shown to prevent apoptosis of normal erythroid progenitors (Bouscary, Pene et al. 2003), SHIP-/- mice are neither polycythemic nor significantly anemic (Helgason, Damen et al. 1998). As a result, investigators have not paid much attention to the role that SHIP plays in erythropoiesis. Ironically, Dr. Jackie Damen in our laboratory first showed SHIP was tyrosine phosphorylated and associated with She in response to Epo stimulation of DA-3 and M07-E cells expressing ectopic EpoRs (Damen, Liu et al. 1993). We were therefore very interested in knowing why the absence of SHIP did not appear to have much impact on this pathway and so embarked on a comprehensive analysis of the role of SHIP in murine RBC generation. 59 3.2 RESULTS 3.2.1 SHIP-/- mice are only mildly anemic but display a marked reticulocytosis Although the BMs of SHIP+/+ and -/- mice have been reported to contain approximately equal numbers (Helgason, Damen et al. 1998); Moody, Xu et al. 2004) of early erythroid progenitors (i.e., BFU-E), the more mature progenitors (CFU-E), along with developing erythroblasts, appear to be depleted in SHIP-/- BM, possibly because, as mentioned above, they are "pushed out" of the B M by the dramatically increased pool of myeloid progenitors (Helgason, Damen et al. 1998). Excess myeloid progenitors are also forced out of the BM, perhaps because of a limiting number of "niches" there. Together, this leads to extramedullary hematopoiesis and results in the spleen weight and cellularity being dramatically increased in SHIP-/- mice (Helgason, Damen et al. 1998). To confirm this, we first calculated both the spleen index (i.e., spleen weight/mouse weight X 1000) and the hematocrit of several SHIP-/- and littermate control mice and found, as reported previously, that SHIP-/- mice indeed have a higher spleen index (Fig. 3.1A) and are not profoundly anemic (Fig. 3.IB). We could not detect a correlation of Hct and spleen index in the SHIP-/-mice. This is likely because the profound splenomegaly in the SHIP-/- mouse is a consequence of myeloid infiltration in addition to the extramedullary erythropoiesis (Helgason, Damen et al. 1998). Furthermore, we did not detect significant differences in the mean cell volume (MCV) or hemoglobin (Hb) content of SHIP-/- red cells (not shown). Importantly, the SHIP-/- mice and the littermate control wild-type mice used in this chapter have a mixed 129JxC57BL/6 background. These mice survive, on average, up to 14 weeks (Helgason, Damen et al. 1998) but display variable severity of phenotype and some mice die considerably sooner. However, subsequent backcrossing on to the C57BL/6 background yields SHIP-/- mice that display a more severe phenotype and substantially shorter life span (Moody, Xu et al. 2004 and our unpublished observations). Along with other abnormalities, the anemia of these severe phenotypes is exacerbated (Moody, Xu et al. 2004). Interestingly, we discovered that some of our mixed background SHIP-/- mice display a marked reticulocytosis (Fig. 3.1C) that has not been reported previously. Typically, in humans, reticulocytosis occurs in anemic patients with functional B M (i.e., an appropriate 60 recovery response to anemia) and so the high reticulocyte index in SHIP-/- mice may be consistent with a more active erythropoiesis (Erslev 1995) and may reflect a compensatory response to anemic pressure. Although the reticulocyte index in SHIP-/- mice was highly variable (and may reflect severity of phenotype), some of the SHIP-/- mice had profound reticulocytosis. Interestingly, although we found all SHIP-/- to have increased TER119+ erythroblasts in their spleens compared to wild-type littermate controls (see Section 3.2.2), the SHIP-/- mice with the highest reticulocyte indices also had the highest numbers of late erythroid progenitors (i.e., TER119+ erythroblasts) in their spleens. 61 Figure 3.1: SHIP-/- mice have a higher spleen index, are slightly anemic and display reticulocytosis. Hematological parameters of untreated SHIP-/- (red triangles) and SHIP+/+ (blue diamonds) were compared. Individual mice were plotted as separate points and the horizontal lines represent the mean of the populations (A) Spleen index P < .001 (B) Hematocrit (packed cell volume) P < .1 (C) Reticulocyte index (reticulocyte fraction of RBC x Hct/0.45 (ideal Hct)) P < .01. Age (6-9 weeks) and sex-matched mice (as a group) were taken from the same colony. P values were calculated using two-tailed t-test, unequal variance. 62 retic. index © p p • • 4 • • • • • • • • • • ON Hct. p 01 00 spleen Index _ i • • • • • • • • • • • • • • • • • • • • • • • • • o _ i • • • • • • • • • • 3.2.2 SHIP-/- mice recover more rapidly from PHz-induced anemia Since a high reticulocyte index is often indicative of increased erythropoietic activity, we were interested in determining how SHIP-/- mice respond during chemically induced stress erythropoiesis. Chemically induced anemia has previously been used to highlight intrinsic abnormalities in adult mouse erythropoiesis (Vannucchi, Bianchi et al. 2001; Socolovsky, Nam et al. 2001). We therefore induced anemia in SHIP-/- and +/+ mice with phenylhydrazine (PHz). Mice treated with PHz develop anemia via a mechanism involving membrane peroxidation and eventual hemolysis (reviewed in (Hodges, Winter et al. 1999). This hemolysis results in a rapid reduction in the hematocrit (Spivak, Marmor et al. 1972) and a subsequent recovery phase. During the recovery phase, the spleens of PHz-treated mice become an important site for RBC production that is associated with a synchronous wave of erythropoiesis (Hara and Ogawa 1978; Krystal 1983). In these experiments, mice are typically injected (intraperitoneal) with PHz on two consecutive days (designated Days 0 and 1 in this thesis), and by Day 4 the spleen becomes engulfed with nucleated erythroblasts such that approximately 90% of the cells are erythroid (Spivak, Marmor et al. 1972; Krystal 1983). Interestingly, we found that while both SHIP+/+ and -/- mice experienced an equivalent drop in hematocrit following PHz treatment, the SHIP-/- mice were able'to recover at a faster rate (Fig. 3.2A). Specifically, SHIP-/- mice experienced a dramatic recovery of hematocrit on Days 3 and 4 compared to SHIP+/+ mice. On Day 5 and afterwards the hematocrits of the SHIP-/- and wild-type mice were not significantly different. The faster rate of recovery of the SHIP-/- mice was preceded by a much more rapid and robust increase in reticulocyte index (Fig. 3.2B). Worthy of note, both SHIP-/- and SHIP+/+ mice fully recovered their hematocrits by Day 8 (Fig. 3.2A). As well, recticulocyte indices were similar by Day 8 and, by Day 10, reticulocytes had returned to baseline levels. Since reticulocytes are macrocytic (greater MCV than mature red cells), the rapid recovery in Hct observed in the SHIP-/- mice could be attributed to both the more rapid release of reticulocytes in the SHIP-/- mice and/or a more rapid production of RBCs in these mice. Although absolute RBC and reticulocyte counts would help to differentiate the contribution of these two possiblites, absolute counts after PHz-treatment proved to be unreliable due to broad differences in the morphologies of RBCs in PHz-treated mice. 64 A Figure 3.2: SHIP-/- mice display a more rapid recovery after PHz treatment The (A) hematocrit (Hct) and (B) reticulocyte index of sex-matched SHIP-/- (red triangles, dashed line) and SHIP+/+ (blue squares, solid line) littermates were determined immediately before (Day 0, control) and daily (starting on Day 2) after PHz-induced anemia. Each point represents the mean of 4-5 mice (± standard deviation (SD)); (*P < .005). 6 5 3.2.3 Enhanced accumulation of erythroid progenitors to the spleen of SHIP-/- mice following PHz treatment Splenic erythropoiesis is particularly important under conditions of erythropoietic stress in the mouse whereas the B M is not predominantly involved (Ou, Kim et al. 1980; Broudy, Lin et al. 1996). Recovery from PHz-induced anemia is accompanied by the mobilization of erythropoietic progenitors into cell cycle and accumulation of erythroid progenitors, especially in the spleen (Hodges, Winter et al, 1999). This may be in part because the spleen is capable of a substantial increase in both size and cellularity during recovery from PHz-induced anemia while the B M is restricted by the available volume within the sinus cavity. Given that extramedullary myelopoiesis and erythropoiesis are already present in resting SHIP-/- mice, we wanted to determine the capacity of SHIP-/- B M and spleen for further recruitment of erythroid progenitors during stress erythropoiesis. We found that even though SHIP-/- spleens are already substantially larger than wild-type spleens at rest (Fig. 3.1A), they were still capable of a dramatic expansion in size (Fig. 3.3A) and cellularity (Fig. 3.3B) following PHz treatment. Moreover, although it is well know that the spleens of wild-type mice increase in size in response to PHz (Hodges, Winter et al. 1999), the spleens in the SHIP-/- mice reach a maximum size much faster (Fig. 3.3A). We also looked at the cellularity of the B M and found that it declined (rather than expanded as in the spleen, although the decreases did not reach statistical significance) in both SHIP+/+ and -/- mice following PHz treatment (Fig. 3.3C). This limited capacity of the B M to accommodate erythroid cell expansion highlights the importance in the mouse of the spleen in both SHIP+/+ and -/- mice for recovery from stress-induced anemia. 66 Figure 3.3: The spleens of SHIP-/- mice rapidly increase in size and cellularity in response to PHz-induced anemia. SHIP-/- (red columns) and SHIP+/+ (blue columns) mice were treated or not with PHz to induce anemia. On the days indicated, the mice were sacrificed and their (A) spleen index (n = 6-12 mice/genotype/day), (B) spleen cellularity (TNC= total nucleated cells) (n = 4-10 mice/genotype/day) and (C) bone marrow cellularity (3-6 mice/genotype/day) were determined. For (A), (B) and (C) the error bars are ± SEM. 67 TNC per spleen 00 We then monitored the accumulation of erythroblasts in the spleens of PHz-treated mice at several time points during their recovery using CD71TER119 surface expression and flow cytometry. As pronormoblasts differentiate to orthochromatic erythroblasts, the cell surface levels of CD71 decline (Lok and Ponka 2000) and those of TER119 increase (Kina, Ikuta et al. 2000). Socolovsky et al (Socolovsky, Nam et al. 2001) have shown the utility of these surface markers in monitoring erythropoeisis in hematopoietic organs by flow cytometry. The CD71TER119 expression profile roughly correlates with the morphological characteristics of erythroblasts as follows: proerythroblasts (CD71 h l g hTERl 19med), basophilic erythroblasts (CD71 h l g hTERl 19hlgh), late basophilic and polychromatophilic erythroblasts (CD71m e dTER119 h i 8 h) and orthochromatophilic erythroblasts (CD71 l o wTERl 19high) (Socolovsky, Nam et al. 2001). Using this technique, we found that SHIP-/- mice accumulated erythroblasts in their spleens in response to PHz substantially faster than their wild-type counterparts (Fig. 3.4A). In particular, both CD71 + TER l o w / m e d as well as CD71+TER119h i g h cells accumulated much more rapidly in the spleens of SHIP-/- mice following PHz treatment (Fig. 3.4A). Since the spleens of SHIP-/- mice were already enriched for CFU-E and erythroid progenitors before PHz treatment (Fig. 3.4A, control), we also followed the accumulation of erythroblasts in the BM of SHIP-/- and +/+ mice to document the contribution of this compartment to anemic recovery. We found that although the SHIP-/- B M was comparatively deficient in erythroid progenitors before PHz treatment, as described previously (Helgason, Damen et al. 1998), the accumulation of erythroid progenitors in the B M in response to PHz was similar in SHIP+/+ and -/- mice on Day 2 and 3 (Fig. 3.4B). However, based on the relative contributions of the cellularity of the spleen and the B M during recovery from PHz (see Fig. 3.3B & C), the spleen is likely a more important site for stress-induced erythropoiesis in both SHIP+/+ and SHIP-/- mice. 69 Figure 3.4: CD71TER119 expression profiles of SHIP+/+ and SHIP-/- following PHz treatment. (A) Spleen cells or (B) bone marrow of untreated and PHz-treated mice were monitored by flow cytometry at the times indicated. B M or spleen cell preparations were blocked in 5% normal rat serum and then labelled with antibodies against CD71 and TER119 (upper'panel array) as described in the Materials and Methods. The FACS profiles are representative examples of more than 3 trials. The experiments shown compare sex-matched littermates as indicated. 70 A s p l e e n control Day 2 Day 3 Day 4 Day 5 71 3.2.4 CFU-E accumulate more rapidly in PHz-treated SHIP-/- mice The more rapid accumulation of late erythroid progenitors in the spleens of SHIP-/-mice in response to PHz-induced anemia could either be due to the a priori presence of erythroid progenitors (particularly CFU-E) in the SHIP-/- spleens, an increased BFU-E or CFU-E pool in SHIP-/- mice, an increase in the number or rate of cell divisions from the BFU-E or CFU-E compartments or enhanced survival of erythroblast populations. To address some of these possibilities, we first determined the total number of BFU-E and CFU-E in the spleens and B M of SHIP+/+ and SHIP-/- mice at rest and under conditions of PHz-induced anemia. By taking into account the contribution from both the B M and the spleen we hoped to compare the total pool of BFU-E and CFU-E available to SHIP-/- and SHIP+/+ mice during recovery. Using methylcellulose based colony assays we counted the number of BFU-E and CFU-E present in the B M and spleens of untreated mice and Day 2 and Day 3 PHz-treated mice (Fig. 3.5). In untreated mice we found a significantly larger number of CFU-E per spleen in SHIP-/- mice than in the SHIP+/+ littermate controls (Fig. 3.5B, control). On the other hand, the SHIP+/+ mice had more CFU-E per femur than the SHIP-/-counterparts (Fig. 3.5A, control). However, the estimated total CFU-E per mouse in the SHIP+/+ and SHIP-/- mice are approximately equal when the cellularity of the entire BM is considered (previous work has shown that the B M of one femur is about 6% of the total B M compartment in the mouse (Boggs 1984)) (see Fig. 3.5E). This means that untreated SHIP-/-mice had nearly equivalent pools of CFU-E compared to SHIP+/+ mice when the spleen and B M hematopoietic compartments were considered together. And, the diminution of CFU-E in the B M of SHIP-/- mice appears to be fully compensated by the CFU-E contribution in the spleen. Although we did not detect a significant difference in the absolute numbers of BFU-E in the B M of untreated SHIP+/+ and -/- mice (Fig. 3.5C), the SHIP -/- mice had substantially higher numbers of BFU-E in the spleen (Fig. 3.5D) and thus the total BFU-E pool in SHIP-/- mice may be larger than in SHIP+/+ mice. Because these are only estimated total mouse values, we did not determine the statistical significance of this difference. However, we can say that the BFU-E pool in SHIP-/- mice is at least as large, if not larger that the wild-type genotype. We then assayed the absolute numbers of CFU-E and BFU-E during anemic recovery at Day 2 and 3 following PHz-treatment since these time points preceded the recovery of the hematocrit. Figure 3.5A shows that in the BM, the CFU-E 72 increased steadily at Day 2 and 3 for both genotypes with the SHIP+/+ mice maintaining nearly double the CFU-E present in SHIP-/- BM. However, as noted above, since the spleen contributed substantially more CFU-E to the total pool of these progenitors (Fig. 3.5B), the PHz-treated SHIP-/- mice at Day 2 had significantly more CFU-E than the wild-type control mice (both spleen alone and in the total mouse estimate). By Day 3, however, the SHIP+/+ CFU-E had increased (whereas the total number of CFU-E in the SHIP-/- mouse had declined) such that the CFU-E content of the spleens were now not significantly different. Taken together, these data corroborate an enhanced recovery of SHIP-/- mice from PHz-induced anemia and show that the rapid accumulation of CFU-E in the SHIP-/- mouse precedes the recovery of the hematocrit. The BFU-E content of the spleen (Fig. 3.5D) and the B M (Fig. 3.5C) was relatively constant (compared to the increase in CFU-E) in both SHIP+/+ and SHIP-/- mice at Day 2 and Day 3 following PHz treatment. However, because of the greater contribution of the spleen to the total pool of BFU-E, the SHIP-/- mice contained higher absolute numbers of BFU-E at all times examined (Fig. 3.5D and 3.5F). 73 Figure 3.5: SHIP-/- mice have higher total CFU-E and BFU-E than SHIP+/+ mice. PHz-treated SHIP-/- mice rapidly accumulated CFU-E in the spleen. (A) SHIP-/- (red bars) and SHIP+/+ (blue bars) mice were left untreated or treated with PHz. Bone marrow aspirates and ammonium chloride treated splenocytes were plated in triplicate (3 mice per genotype per time point) to assay (A) B M and (B) spleen CFU-E and '(C) B M and (D) spleen BFU-E as described in the Materials and Methods. Total mouse (E) CFU-E and (F) BFU-E were estimated based on the results of A, B, C and D and the formula: total mouse progenitors = spleen progenitors + (femur progenitors x 16.7) ((Kean, Brown et al. 2002; Boggs 1984)). The error bars are ± SEM. P values were calculated using the counts from each replicate for the three mice; * P <0.05. 74 A BM CFU-E i - 1.5x1005-i Days 75 c Sp leen B F U - E a» 2.0*10 0) Days D B M B F U - E 1.5x10°S Days 76 Total mouse CFU-E o 7.5x1006n • 5.0x10 Q. 06J Z 2.5x10 o ,06. control I E S S 2 Days I ™total BM +/+ total BM -/-• spleen +/+ i spleen -/-Total mouse BFU-E • 4.0x1005 \ 3.0x10°H a 8 2.0x10°H I 2 Days I D M total BM +/+ total BM -/-S i spleen +/+ spleen -/-77 3.2.5 Epo-responsiveness of erythroid progenitors from the spleens of SHIP+/+ and -/-mice following PHz-treatment We then asked if there was any difference in the Epo-responsiveness of SHIP+/+ and -/- erythroid progenitors. Interestingly, in this regard, we found that treatment of both SHIP+/+ and -/- PHz-treated spleen cells with ammonium chloride (NH4C1) enriched for Epo-responsive cells (Fig. 3.6A). Subsequent CD71TER119 flow cytometry profiles revealed that the reason for this was because the standard NH4CI treatment used to eliminate mature RBCs from spleen cell preparations also lysed late nucleated erythroblasts, which express low levels of CD71 (compare the non-NH4Cl-treated spleen cells in Fig. 3.4A to the NH4Cl-treated spleen cells in Fig. 3.6B). Although these lysed late erythroblasts are known to express EpoRs, they have little capacity for further cell divisions. Thus, their elimination 3 3 by N H 4 C I lysis prior to H-Thymidine ( H-Tdr) assay effectively enriched the Epo-induced proliferative potential, on a per cell basis, of the remaining spleen population. In addition, these relatively mature erythroblasts are known to feedback enhance apoptosis in less mature erythroblasts via a FasL/Fas interaction and so their removal may also have contributed to the increase in the Epo-induced proliferative potential we observed (Testa 2004). One major concern we had with comparing the Epo-responsiveness of SHIP+/+ and -/- spleen cells from PHz-treated mice was that we could be comparing "apples and oranges", (i.e., the differentiation states of the two cell populations could be very different). To address this we tried to compare the Epo-induced proliferation of the various CD71TER119 compartments but, unfortunately, we found that sorting of these cells markedly reduced their biological responses. We therefore tried non-FACS approaches. Specifically, we used lineage depletion, using a cocktail of antibodies to deplete lineage-committed cells (including TER119) from NH4Cl-treated Day 3 PHz splenocytes. As shown in Figure 3.7A (top set of panels) this dramatically enriched for lineage negative cells in both SHIP+/+ and -/- spleen cells and yielded highly similar preparations of C D71 h i e h TERl 19 m e d / l o w cells (bottom set of panels) that contained CFU-E (see section 3.2.6). However, the majority of these Lin" cells appeared to be proerythroblasts and basophilic erythroblasts from cytospin analyses (not shown). Using this preparation of cells, we performed H-Tdr incorporation assays to measure the Epo-induced proliferation of Lin" cells from SHIP-/- and SHIP+/+ Day 3 PHz spleens (Fig. 3.7B). The absolute magnitude of H-Tdr incorporation of Lin" Day 3 PHz 78 splenocytes was lower for the SHIP +/+ mouse (left panel). However, when plotted as a percentage of maximum incorporation, the dose response curves were similar for the SHIP+/+ and -/- Lin" cells (Fig. 3.7C). This was consistent with the SHIP-/- and +/+ cells having a similar capacity to respond to Epo. The observed differences in the absolute level of H-Tdr incorporation may reflect, despite our efforts, non-equivalent populations of cells. 79 Figure 3.6: Ammonium chloride lysis of spleen cells prepared from Day 3 PHz treated mice enriches for Epo-responsive cells. (A) Untreated spleen cells (solid diamonds, dashed line) and NH4CI treated spleen cells (up solid triangles, solid line) from Day 3 PHz spleens (SHIP+/+) were washed twice in RPMI containing 20% AFCS and reseeded with increasing doses of Epo. 3H-Tdr assays were performed as described in the Materials and Methods. The assay was performed in duplicate and the error bars are ± SEM. The results shown are representative of several experiments. (B) CD71TER119 FACS profiles of untreated and NH4Cl-treated spleens where Region II (CD71 h ' 8 hTERl 19low) ~ proerythroblasts/basophilic erythroblasts, Region III (CD71m e dTERl 19 h i g h / m e d) ~ basophilic/polychromatophilic erythroblasts and Region IV (CD71 l o wTERl 19low) ~ orthrochromatophilic erythroblasts. Mature red cells and dead cells (propidium iodide positive) were gated out of the profiles 80 A 400000-1 c o 5000 Epo (mU/mL) B untreated spleen NH4Cl-treated spleen 10° 101 102 103 104 10° 101 102 103 104 TER119 • SI Figure 3.7: SHIP+/+ and SHIP-/- lineage-depleted (Lin) Day 3 PHz splenocytes are equally responsive to increasing concentrations of Epo. (A) Lineage-depleted (Lin: Ly-1), Erythroid Cells (TER119), CD45R (B220), Ly-6G (Gr-1), CDllb (Mac-1), Neutrophils (7-4)) of spleen cells prepared from Day 3 PHz-treated SHIP +/+ and SHIP-/-mice. (A) Flow cytometry analysis showing the Lin*TERl 19 profile of (left) unfractionated spleen and (right) lineage-depleted splenocytes. (B) Shows CD71TER119 profile of either (left) unfractionated or (right) lineage-depleted Day 3 PHz spleen cells. (C and D) 3H-Tdr incorporation assays of Lin" splenocytes from Day 3 PHz mice in response to increasing doses of Epo. Graphs are plotted based on (C) 3H-Tdr incorporation in counts per minute (cpm) in logio scale or (D) as a % of the maximum 3H-Tdr incorporation with Epo dose on the x-axis. The maximum Epo dose was 3 U/ml. Error bars are ± SEM of duplicate determinations. Lin* = CD5 (Ly-1), Erythroid Cells (TER119), CD45R (B220), Ly-6G (Gr-1), CD1 lb (Mac-1), Neutrophils (7-4)). Lin" = lineage-depleted cell population. 82 A D a y 3 P H z S p l e e n unfractionated spleen -MR • * ;'. 'ijr* '"I I 1 1 ' ' 'i"" Lineage-depleted 10" i f SHIP+/+ r y r m , 10" 10' 10' 10° 10" 10" 10' 10' 10° 10" 10* tar 10" SHIP- / -1 * ""1 1 1 1 10" 101 102 103 104 Lin*: CD5 (Ly-1), CD45R (B220), Ly-6G (Gr-1), C D l l b (Mac-1), Neutrophils (7-4) 83 c 85 3.2.6 Epo-responsiveness of SHIP +/+ and -/- BFU-E and CFU-E Since Helgason et al. (Helgason, Damen et al. 1998) had used methylcellulose assays to show hyper-responsiveness of SHIP-/- myeloid progenitors in response to SCF, IL-3, G-CSF and M-CSF we wanted to compare the Epo-responsiveness of BFU-E and CFU-E in methylcellulose cultures of SHIP-/- and SHIP +/+ mice. For these colony assays we used the same Lin" Day 3 spleen cells that we used for our suspension culture Epo-responsiveness studies, in part because our CFU-E assay results above suggested a similar CFU-E content in the SHIP+/+ and the -/- spleens on that day. We not only assayed for CFU-E and BFU-E but for CFU-GM (granulocyte/macrophage progenitors) and CFU-Mk (megakaryocyte progenitors) as well. These studies revealed that the frequency of CFU-Mk colonies was less than 1/10,000 and the CFU-GM frequency was less than 1/1500 cells plated in both SHIP+/+ and -/- Day 3 PHz spleen Lin" populations. Interestingly, the frequency of BFU-E in the Lin~ cell populations was not significantly different for the SHIP+/+ (0.37 ± 0.13 % (SD)) and -/- (0.27 ±0.10 % (SD)) cells when assayed with saturating levels of cytokines. However, we did detect a significant difference in the frequency of CFU-E in the Lin"" population of 23 ±3 % (SD) and 9.4 ± 1 % (SD) for the SHIP+/+ and SHIP-/-, respectively. This calculation is based on the frequencies determined in triplicate from 6 SHIP+/+ and 3 -/-Day 3 PHz mice using saturating cytokine conditions. Related to this, although there was also a trend to more CFU-E in the unfractionated, NHiCl-treated spleen cells from Day 3 PHz-treated SHIP+/+ mice, this was not statistically significant when the data when several experiments were combined (see Fig. 3.5A, Day 3). Since the frequency of CFU-E in unfractionated, NH4Cl-treated Day 3 PHz spleen was (0.5 - 1.5%) for the SHIP -/- mice and (1 - 2.6%) for SHIP +/+ mice, the Lin" purification of Day 3 PHz spleen cells allowed for a 10-20 fold enrichment of CFU-E. In order to test the Epo-responsiveness of these progenitors we plated the Lin" selected cells under saturating conditions (3 U/ml Epo) as well as decreasing concentrations of Epo and scored the number and size of the resulting colonies. As shown in Figure 3.8A, the frequency of CFU-E in the Lin" populations at saturating and limiting concentrations of Epo was reduced in the SHIP-/- compared to +/+ cells. However, when the number of colonies obtained at limiting Epo concentrations was plotted as a percentage of the 86 maximum, there was no significant difference in the Epo-responsiveness of SHIP+/+ and SHIP-/- CFU-E (Fig. 3.8B). 87 A Ji 1 cr x £ CO SHIP+/+ SHIP-/-Epo U/mL B % max plating efficiency 30 100 Epo (mU/mL) 300 Figure 3.8: SHIP+/+ and SHIP-/- CFU-E are equally responsive to Epo. Lin" spleen cells from Day 3 PHz mice were plated in methylcellulose under conditions of saturating concentrations of Epo (3 U/mL) or limiting concentrations of Epo (0.01, 0.03, 0.1, 0.3 U/mL). Colonies were counted in triplicate (A) The frequency of SHIP+/+ (squares) and SHIP-/- (diamonds) CFU-E colonies per 3xl0 3 Lin" cells observed at each Epo dose were determined. Error bars are ± SD of triplicate determinations (B) The % maximum plating efficiency (based on the number of colonies detected under saturating cytokine conditions (3U/mL)) at limiting doses of Epo. SHIP-/- (red open bars), SHIP+/+ (blue solid bars). 88 3.2.7 Epo-induced tyrosine phosphorylation in Lin" cells from Day 3 spleens of PHz-treated mice Since SHIP is an important negative regulator of the PI3K/Akt pathway in many cells and also regulates the activation of the ERK pathway in some cell types (Kalesnikoff, Baur et al. 2002), we next wanted to determine if the phosphorylation of Akt or ERK were enhanced in Lin" cells purified from Day 3 spleens of SHIP-/- PHz-treated mice. As can be seen in Figure 3.9A, SHIP itself was rapidly phosphorylated at the NPXY2 site following stimulation of SHIP+/+ cells with Epo. Interestingly, however, the phosphorylation of Akt was not enhanced in the SHIP-/- cells in response to Epo, at least not at the two time points examined. However, ERK phosphorylation was enhanced at 10 min after Epo stimulation in the SHIP-/- cells. The only caveat in this study is that, as mentioned before, the cell populations of the SHIP+/+ and -/- samples may not be identical. Since we found that Src family members are upstream of SHIP tyrosine phosphorylation in response to a variety of stimuli (see Chapter 4), we tested whether this was also true of the Lin" population (consisting largely of CFU-E and proerythroblasts). As shown in Figure 3.9B, PP2, but not its inactive analogue, PP3, reduced the level of SHIP tyrosine phosphorylation in response to Epo. 89 Figure 3.9: Epo-induced signalling events in SHIP+/+ and SHIP-/- Lin- Day 3 PHz spleen cells. Lin- spleen cells from Day 3 PHz spleens were starved in serum-free medium for 4 hours immediately after immunomagnetic negative selection as described in the Materials and Methods. Cells washed and stimulated or not (-) in SFM with 100 U/mL Epo for time points indicated (mins) and then immediately lysed to prepare total cell lysates, fractionated by SDS-PAGE and transferred to PVDF membranes. Each lane contains 5X105 cell equivalents of cell lysate. (A) Comparison of Epo stimulations in SHIP+/+ and SHIP-/-Lin" cells (B) SHIP+/+ cells were preincubated for 30 min with 30 uM PP2 or PP3 and then simulated with Epo as above. 90 SHIP+/+ 2 10 • • SHIP-/-2 10 min pSHIP - T 7 T — - — SHIP pSTAT5 pAkt (S473) pERK1/2 ERK1/2 GAPDH B PP3 PP2 + Epo pSHIP (NPXY2) P-tubulin 9 1 3.2.8 CD71 h i g h TER119 m c d / l o w erythroblasts from SHIP+/+ and -/- mice express similar EpoR, STAT5a and GATA-1 levels There are a number of possible reasons why SHIP-/- erythroid progenitors, unlike SHIP-/- myeloid progenitors, are not hyperresponsive to cytokine stimulation. One possibility is that they have fewer EpoRs or reduced expression of downstream signalling components than their wild-type counterparts. Although our signalling studies above suggested that Epo was capable of signalling to a similar extent in SHIP+/+ and -/- Lin" cells, this could have been in spite of a reduced number of EpoRs. Since the Lyn-/- and SHIP-/-share considerable similarity in phenotype (particularly aged Lyn-/- mice) including a myeloproliferative disorder, immune abnormalities and mild anemia with reticulocytosis (Harder, Quilici et al. 2004; Ingley, McCarthy et al. 2004) we were mindful of very recent reports evaluating the erythroid compartment of Lyn-deficient mice. One such report (Ingley, McCarthy et al. 2004) had found decreased expression of the erythroid-specific transcription factors GATA-1, STAT5a/b and EKLF in the spleens of PHz-treated mice (equivalent to Day 4 PHz unfractionated spleen in our study). These findings were significant because a Lyn-deficient murine erythroleukemic cell line (J2E-NR), which has lost the ability to differentiate in response to Epo, also displayed reduced expression of these transcription factors (Ingeley, McCarthy et al. 2004). Furthermore, the genetic mouse mutants GATA-1" (null mutant) (Pevny, Simon et al. 1991), G A T A - l l o v v (McDevitt, Shivdasani et al. 1997; Vannucchi, Bianchi et al. 2001) and STAT5a-/-STAT5b-/-(Socolovsky, Nam et al. 2001) all have mild to severe deficiencies in erythropoiesis. The G A T A - l l o w phenotype (erythroid compartment) in particular resembled the observed erythropoietic abnormalities observed in Lyn-deficient mice, and to a lesser extent the SHIP-/- mice (Ingley, McCarthy et al. 2004; Vannucchi, Bianchi et al 2001). Surprisingly, we found that SHIP-/- spleen cells (unfractionated) from Day 4 PHz mice had reduced levels of GATA-1, STAT5a/b and the EpoR (not shown). However, the results were highly variable amoung littermate pairs of SHIP+/+ and SHIP-/- mice. This was consistent with the observations of Ingley et al (Ingley, McCarthy et al. 2004) who had reported highly variable expression (20-80% reduction based on western analysis) of GATA-1, STAT5a/b and EKLF in Lyn-deficient PHz spleen cells compared to wild-type controls. Since both the Lyn-/-(Harder et al 2004; Ingley, McCarthy et al. 2004) and SHIP-/- mice display a more rapid 92 accumulation of CFU/erythroblasts in the spleen in response to PHz we were concerned that the variability reported in the Lyn-/- results and our own results were due to unequal populations of cells in the spleens of PHz treated mice. We therefore used an alternative approach that would allow us to directly compare SHIP+/+ and SHIP-/- erythroblasts at equivalent stages of differentiation. We thus FACS-sorted the NH4Cl-treated spleen cells from SHIP+/+ and -/- mice on Day 3 (and Day 4, not shown) after PHz treatment into 3 subpopulations based on their CD71TER119 levels. As can be seen in Figure 3.10A, Western blots of the starting population and the 3 sorted fractions, loaded with equal levels of total protein, revealed, as expected, that very low numbers of EpoRs were present on the primitive CD71TER119" subset (Region I) from both SHIP+/+ and -/- preparations (Fig. 3.10B). On the other hand, there were substantial EpoR levels in the CD71 h i g hTERl 19m e d (Region II) and more mature CD71 m e dTERl 19h i g h subsets (Region III). Importantly, there were no obvious differences in EpoR levels between the SHIP+/+ and -/- samples at equivalent stage of erythroid differentiation. Also of interest, STAT5a, which was also expressed at equal levels in SHIP+/+ and -/- samples, was expressed primarily in subset II and disappeared with subsequent differentiation. The important erythroid specific TF, GATA1 (Orkin 1992), on the other hand, was first expressed on CD71 h i g hTERl 19m e d cells and remained at high levels in the more mature CD71 m e d TERl 19h l g h cell population. Interestingly, we were surprised to find that (3-tubulin was very specifically restricted in its expression to CD71 h i g hTERl 19m e d cells and could therefore serve in the future as a good marker of this stage of erythroid differentiation. Also of interest, although we did not detect actin in the CD71 m e d TERl 19 h i g h population, GATA-1 and EpoR levels were still high and this high level of GATA-1 and the EpoR in Region II is consistent with these cells being predominantly early erythroblasts (Orkin 1992). 93 Figure 3.10: SHIP-/- and SHIP+/+ erythroblasts isolated by F A C S express similar levels of EpoR, GATA-1 and S T A T 5 a . (A) CD71TER119 FACS profile of SHIP+/+ and SHIP-/- N H 4 C I treated Day 3 PHz spleen. Populations were sorted (FACSAria™ ) as follows: Region I (CD71TER119); Region II (CD71 l l i g hTERl 19med); Region III (CD71 m e d / h i g h TERl 19high). (B) Sorted cells were lysed in SDS sample buffer and fractionated by SDS-PAGE for immunoblotting. Equal amounts of protein were loaded in each lane (based on parallel protein determination assay). Regions are as indicated in (A). S represents pre-sort spleen cells. Similar results were obtained in 2 independent experiments. 94 SHIP+/+ SHIP-/-TER119 B SHIP+/+ S I II II SHIP-/-S I II II , , . | | , | || SHIP Stat5a EpoR GATA-1 p-tubulin actin 95 3.2.9 SHIP expression is lost during erythropoiesis To gain some insight into why the absence of SHIP did not have a more marked effect on erythropoiesis we looked at the expression of SHIP in wild-type mice as erythroid progenitors matured. To our surprise, we found that SHIP expression was significantly reduced or absent in more mature erythroid progenitors. Using the CD71TER119 profile we were able to sort Day 3 PHz splenocytes from wild-type mice and probe for SHIP protein expression by Western blotting and found that the CD71 h i g hTERl 19m e d cells (Region II, Fig. 3.10B) expressed SHIP at a level similar to the non-erythroid population (Region I, Fig. 3.10B), but more mature erythroblasts (CD71m e dTERl 19high) (Region III, Fig. 3.10B) expressed little or no SHIP. To confirm our findings we isolated the Lin" fraction from Day 5 PHz spleens using a custom murine hematopoietic antibody enrichment cocktail (Lin*: CD5 (Ly-1), CD45R (B220), Ly-6G (Gr-1), CDl lb (Mac-1), Neutrophils (7-4) as described in the Materials & Methods) in which the antibody for TER119 was omitted. This allowed us to enrich greatly our spleen cell population for the erythroid lineage and include the TER119 erythroblasts. Using intracellular labelling and flow cytometry we were able to detect SHIP expression in the TER119" population but not the TER119+ population in SHIP+/+ spleen cells. SHIP-/-spleen cells were used as a negative control (Fig. 3.11). Since PI3K activity is important in the differentiation of murine and human erythroid progenitors, the regulated loss of SHIP expression during erythroid differentiation may have functional significance. 96 U3iS6AUQei3 U3:26AUGet5 FL1-HNFL1-H*ighl > < C 3 o ? C M O 3- FLl-WsR.1-Height > F L l - m F L l - H « i « h t > U3I26AUG816 X to i 6 ° i 6 * F L I - H v F L l - H » i g h t > SHIP+/+ SHIP-/-TER119 Figure 3.11: Late TER119 erythroblasts do not express SHIP. Day 5 PHz spleen from SHIP+/+ (upper panels) and SHIP-/- (lower panels) mice were purified by negative immunomagnetic selection using a custom lineage-depletion cocktail of antibodies (Lin*: CD5 (Ly-1), CD45R (B220), Ly-6G (Gr-1), CDl lb (Mac-1), Neutrophils (7-4)). This method enriched the Day 5 PHz spleen population for the erythroid lineage and included both early and late TER119+ erythroblasts. Purified cells were first labelled anti-TERl 19 (biotin) followed by SA-FITC and then fixed, permeablized and labelled with either anti-SHIP plus anti-rabbit PE (right panels) or anti-rabbit PE alone (2° Ab control, left panels). 97 3.2.10 Enhanced induction of erythropoietic growth-promoting activity in the plasma of SHIP-/- mice Lastly, we wanted to look at the level of erythroid growth promoting activity in the plasma of SHIP+/+ and -/- mice during recovery from PHz. As mentioned earlier, the dramatically expanded myeloid compartment in SHIP-/- mice leads to an increased production from these expanded Macl+ cells of IL-6 into the plasma of SHIP-/- mice and this contributes both to the decrease in B lymphocytes and to even more pronounced myelopoiesis seen in older SHIP-/- mice. We therefore wanted to know if the plasma from PHz-treated SHIP-/- mice might possess higher or lower erythropoietic-stimulating activity than plasma from SHIP+/+ mice. Specifically, we collected plasma from SHIP+/+ and -/-mice on different days before and after PHz treatment and assayed their ability to promote the proliferation of erythroid progenitors from the spleens of SHIP+/+ PHz-treated mice. Using this method, we discovered that there was greater growth-promoting activity in the plasma of SHIP-/- mice at time-points very soon after PHz-treatment (Fig. 3.12A). However, since even PHz spleen contains a mixed population of cells, we used an alternative approach to in an attempt to identify the activity in the plasma of the PHz-treated mice. To gain some insight into the nature of this activity we tested the various plasma samples for their ability to stimulate proliferation of BaER cells (Ba/F3 cells expressing ectopic EpoRs) in the presence and absence of a neutralizing antibody to murine Epo (anti-Epol6) (StemCell Technologies, Inc.). Once again, we found that SHIP-/- mouse plasma possessed more growth promoting activity and that much of this activity could be blocked with Epo-neutralizing antibody (anti-Epol6) but not an isotype control antibody (Fig. 3.12C). Although one could estimate the level of plasma Epo using a standard Epo dose response (Fig. 3.12B) with pure Epo, it is likely that even though we could block most of the growth-promoting activity in the plasma samples with anti-Epol6, the remaining Epo likely synergizes with other factors (e.g., IGF-1 (Damen, Krosl et al 1998) or glucocorticoids (GC) (Bauer, Tronche et al 1999)) in plasma and would likely overestimate the Epo levels in these samples. It is worthy of note in this regard that there are no commercially available Epo ELISAs for murine Epo and so we could not directly determine the level of Epo in our mouse plasma. Attempts to measure the Epo levels with an ELISA reagent designed for use in the clinic (human Epo detection) were unsuccessful. 98 Figure 3.12: Enhanced induction of erythropoietic growth-promoting activity in the plasma of SHIP-/- mice. (A) Increasing concentrations of mouse plasma obtained from SHIP+/+ and SHIP-/- mice following PHz-treatment (on days indicated) were assayed for growth-promoting activity using 3H-Tdr incorporation assays of Day 4 NHtCl-treated spleen (SHIP+/+ in all experiments). (B) BaER cells were subjected' to 24 hr with increasing concentrations of Epo (solid triangle, solid line) in the presence of neutralizing Epo-antibody (open triangles, dashed line) or an isotype control antibody (solid diamonds, dashed line). (C) Increasing concentrations of plasma obtained from SHIP+/+ and SHIP-/- mice following PHz-treatment (as indicated) were assayed for growth-promoting activity in 3H-Tdr assays using BaER cells in the presence of neutralizing anti-Epo (open symbols, dashed line) or an isotype control antibody (solid lines). Each line represents the growth-promoting plasma in separate mice. All assays were performed in duplicate and the error bars are ± SEM. This experiment is representative of 2 independent experiments with similar results. 99 B 100000 E p o d o s e r e s p o n s e 0 1000 2000 3000 4000 5000 Epo (mU/mL) SHIP+/+ plasma SHIP-/- plasma 20000 ? i 115000 5 c l | £ 2 IOOOOH " 5000H 0.0 2.5 5.0 7.5 10.0 12.5 15.0 % plasma 20000 | = 115000-) B c f l ^ 5 10000 « 5000H control 0.0 2.5 5.0 7.5 10.0 % plasma 12.5 15.0 20000-] E £ j i 15000' i i £ S 10000-1 t— o o ft S 5000 0.0 2.5 5.0 7.5 10.0 % plasma 12.5 15.0 Day 1.5 15.0 20000 ? c i i 15000 12.5 15.0 % plasma Day 2 0.0 2.5 5.0 7.5 10.0 12.5 15.0 % plasma 101 3.3 DISCUSSION In the preceding studies, we identified a perturbation in the erythropoiesis in SHIP-/-mice that results in a mild anemia coupled with a reticulocytosis indicative of enhanced erythropoietic rate in resting SHIP-/- mice. Furthermore, SHIP-/- mice displayed enhanced recovery from PHz-induced anemia. A high reticulocyte index is typically associated with an increased rate of erythropoiesis and, in the clinic, is used as a measure of the erythropoietic state of the BM. Moreover, the number of reticulocytes in peripheral blood usually reflects the erythropoietic activity of B M (or extramedullary organs), the rate of reticulocyte delivery from hematopoietic organs to the blood and the rate of reticulocyte maturation (reviewed in (Riley, Ben-Ezra et al. 2 0 0 2 ) . It is conceivable that this reticulocytosis and the elevated erythropoietic activity that we observed in SHIP-/- mouse plasma under stress is a result of perturbed erythropoiesis. One could hypothesize that an intrinsic deficiency in erythropoiesis leading to inefficient erythropoiesis may result in a higher turnover of erythroid progenitors at rest (and thus the reticulocytosis) as a method of homeostatic compensation. However, we were not able to define an intrinsic defect in the erythropoiesis of SHIP-/- erythroid progenitors. Evidences to the contrary include the fact that we did not detect a difference in the Epo-responsiveness of late erythroid progenitors in 3H-Tdr incorporation assays using a highly pure population of CFU-E/erythroblasts (Lin" Day 3 PHz spleen), we did not detect a difference in the plating efficiency of CFU-E in response to Epo in cells derived from this same population, nor did we detect a difference in the expression of erythroid specific transcription factors (GATA-1 and STAT5a) or EpoR when erythroblasts at equivalent stages of differentiation were compared directly. These results suggest that the presumed increased erythropoietic rate of SHIP-/- mice at rest and enhanced response to stress erythropoiesis is not simply a result of increased sensitivity of SHIP-/- to Epo. Note that this result differs from similar studies in SHIP-/- granulocyte/macrophage (GM) progenitors that demonstrated that SHIP-/- GM progenitors are hyperresponsive to various cytokines (Helgason, Damen et al. 1998). However, we did detect differences in the phosphorylation/activation of Akt, ERK and STAT5 in SHIP-/- compared to SHIP+/+ Lin" Day 3 PHz spleen cells in response to Epo stimulation. Surprisingly, the predicated increase in phosphorylation of Akt (S473), a PI3K-mediated event, in the SHIP-/- spleen cells was not apparent, but this may have been due to the differences in the Epo-sensitivity of unequal 102 populations and may not truly reflect intrinsic signalling differences in these cells. Interestingly, in the Lyn-/- erythroblasts, Ingely et al (Ingley, McCarthy et al. 2004) reported reduced STAT5 tyrosine phosphorylation in response to EpoR activation using unfractionated spleen cells. However, another group (Harder, Quilici et al. 2004), using CD71-immunomagnetic purified spleen populations, did not detect differences in the dose or time-dependent phosphorylation of STAT5, SHIP-1, SHP-1 or ERK 1/2 in response to Epo. Many of the contradictions in these reports may be explained by differences in the techniques and methods used to isolate erythroblasts from phenylhydrazine spleens since, because of the non-equivalent stress response of Lyn-/- (Harder, Quilici et al. 2004; Ingley, McCarthy et al. 2004) and SHIP-/- mice, the populations of erythroblasts derived from the PHz spleens of these mice are not identical. Consequently, even in our studies, which used a highly pure population of Lin" Day 3 PHz spleen cells (CD71 h i g hTERl 19"/med) that predominantly resembled proerythroblasts/basophilic erythroblasts, further analysis revealed that, despite our best efforts, these populations of cells were not equivalent and this may account for the differences in phosphorylation levels of STAT5, ERK1/2 and Akt (S473) we observed. Note also that, although the Epo-responsiveness of SHIP+/+ and -/- Lin" Day 3 PHz spleens are equivalent, we typically see differences in the absolute magnitude of H-Tdr incorporation in SHIP-/- Lin" cells isolated from Day 3 PHz-spleens. This result may also be explained by differences in the composition of the Lin" populations from the different genotypes since the SHIP+/+ cells contained a higher proportion of CFU-E. Part of the impetus to look for differences in the expression of GATA-1 and STAT5 in SHIP-/- erythroblasts came from studies of others that have shown a reduced expression of these TFs in Lyn-/- erythroblasts (also derived from PHz spleen). We were further encouraged by the fact that the phenotype of the Lyn-/- mice (especially aged Lyn-/- mice) (Harder, Quilici et al. 2004; Ingley, McCarthy et al. 2004) shared considerable similarities with the SHIP-/- mice. Recent studies also reveal similarities in the erythroid compartments of the Lyn-/- and SHIP-/- mice and to a lesser extent the STAT5a-/-5b-/- (Socolovsky, Nam et al 2001) and G A T A - l l o w (Vannucchi, Bianchi et al. 2001) mouse models. In fact, the perturbed erythropoiesis in the Lyn-/- mouse was attributed in one study (Ingley, McCarthy et al 2004), in part, to the decrease in expression of STAT5a/b and GATA-1. Both the STAT5a-/-5b-/- and GATA-1 l o w mouse models have been documented to contain intrinsic 103 defects in erythropoiesis (Socolovsky, Nam et al 2001; Vannucchi, Bianchi et al. 2001). Although we were initially able to replicate these results showing decreased expression of STAT5a/b and GATA-1 (not shown) in the SHIP-/- mouse in unfractionated PHz spleen (Day 4), using an alternative approach that allowed us to compared erythroblasts at the same stage of development proved otherwise. Namely, GATA-1, STAT5a (and STAT5b (not shown)) and the EpoR are approximately equally and appropriately expressed in erythroblasts in both SHIP-/- and +/+ mice. Although we did not examine the expression of these TFs in Lyn-/- mice ourselves, it is possible that the differences in expression observed by these authors is due to unequal populations of cells used in the comparison (Day 4 PHz unfractionated spleen) as this may account for the variability in the results reported by these authors. Like SHIP-/- mice, the spleens of SHIP-/- mice accumulate erythroblasts at a faster rate that wild-type controls. If Lyn-/- mice indeed have low STAT5a/b and GATA-1 expression, it is likely a problem that arises late in the development of the mouse since both the STAT5a-/-5b-/- and the G A T A - l l o w mice have severe fetal or neonatal defects in erythropoiesis due to reduced survival of erythroid progenitors and erythroblasts (Socolovsky, Nam et al. 2001; Vannucchi, Bianchi et al. 2001). In contrast, the relatively mild anemia in the Lyn-/- (and SHIP-/-) develop in parallel with the myeloproliferative disorder. In the Lyn-/- mouse in particular, profound splenomegaly does not occur until the mice are of relatively advanced age. Also, in the Lyn-/- mouse, anemia at the fetal and neonatal stages was not detected nor was there a difference in the survival of erythroblasts in Lyn-/- mice; whereas, the profound anemia of the STAT5a-/-5b-/- and G A T A - l l o v v are resolved as the mice reach adulthood due to homeostatic compensation (Vannucchi, Bianchi et al 2001; Ihle 2000). Interestingly, both the SHIP-/- and Lyn-/- have resting reticulocytosis, whereas no such condition is reported for the STAT5a-/-5b-/- and G A T A - l l o w models. We plan to explore the survival of erythroblasts in the SHIP-/- mice but face two obstacles: (1) In the resting mice we may be required to compare survival of B M SHIP+/+ erythroblasts to splenic SHIP-/- erythroblasts since erythropoiesis beyond the CFU-E stage is predominantly in the spleen of SHIP -/- mice and (2) SHIP expression may be functionally lost in late erythroblasts such that the study of the survival of relatively early erythroblasts (CD71 h i g hTERl 19med) is of prime importance. 104 Intrinsic defects in erythropoiesis are normally apparent at the prenatal or neonatal stage of development (Socolovsky, Fallon et al. 1999) and no such abnormalities are observed in Lyn-/- (Harder, Quilici et al. 2004) or SHIP-/- mice. Given this, and the fact that the mild anemia in SHIP-/- and Lyn-/- are exacerbated by age and severity of the myeloproliferative disorder, we must consider that, in the SHIP-/- mouse, (and perhaps the Lyn-/- mouse) there is no intrinsic defect in erythropoiesis, but rather the anemia and reticulocytosis is a symptom of extrinsic mechanisms. For example, SHIP-/- mice are likely chronically hypoxic due to infiltration of neutrophils and macrophages into the lungs. Such hypoxic conditions may result in an increased erythropoietic rate in these mice. Splenic erythropoiesis is consistent with a stress erythropoiesis mechanism (and increase circulating reticulocytes) (Bauer, Tronche et al. 1999). However, it is obvious that the stress erythropoiesis is never resolved since anemia persists despite the apparent chronic hypoxia. Stress erythropoiesis (hypoxia or hemolysis) requires both Epo and glucocorticoids (GC) (Bauer, Tronche et al. 1999). The latter allows for greater proliferation over differentiation of erythroblast populations (Bauer, Tronche et al. 1999). We may therefore expect higher Epo or GC levels in resting SHIP-/- mice compared to wild-type controls. In this regard, Koury has shown that only a small increase in Epo is sufficient to compensate for a poorly functioning mutant EpoR (Koury, Sawyer et al. 2002) and it is possible that slightly elevated Epo levels exist in the plasma of unstressed SHIP-/- mice. Of note, the BFU-E numbers are apparently normal in B M of SHIP -/- mice, but a pool of BFU-E in the spleen may explain the greater CFU-E levels in spleen (rather than migration of late BFU-E or early CFU-E) and the presence of at least some erythroblasts in the SHIP-/- resting B M suggest that there is not a particular erythropoiesis defect that dissuades BFU-E development in this compartment. Rather, the splenic erythropoiesis may be a preferred response to the chronic hypoxia or other extrinsic pressures. Alternatively, extramedullary hematopoiesis/erythropoiesis could be caused simply because late progenitors are pushed out of the B M by the hyper-proliferative myeloid progenitors in the B M or erythroid progenitors could be stressed, perhaps more likely, because of inhibitory plasma factors produced by the overly abundant population of myeloid cells. In Lyn-deficient, and perhaps SHIP-deficient mice, the anemia gets worse with age and this coincides with an increase in myeloproliferative disorder in the Lyn-/- mouse 105 (Harder, Quilici et al. 2004). On the other hand, the myeloproliferative disorder in SHIP-/-mice is comparatively worse (particularly at a younger age) than in the Lyn-deficient model (Harder, Quilici et al. 2004). Interestingly, the BFU-E and CFU-E numbers in the Lyn-/-mouse are normal in the B M at all ages suggesting that the suppressive factors produced by the myeloid cells rather than overcrowding of the myeloid progenitors alone is responsible for the relocation of erythropoiesis from the B M to the spleen (Harder, Quilici et al. 2004; Ingley, McCarth et al. 2004). Although we had ppreviously reported a hyporesponsiveness of SHIP-/- to erythropoietin in colony forming assays and 3H-Tdr assays (Kalesnikoff, Sly et al. 2003), those experiments were performed using unfractionated spleens from SHIP-/- and SHIP +/+ mice. We suspect that the high myeloid background in these studies may have been the reason for these previous findings. Further experiments investigating the potential erythroid suppressive role of myeloid cells in SHIP-/- cells will be required to verify this hypothesis. Of interest, Ingley et al. reported morphological abnormalities of erythroblasts in the spleen and BM of Lyn-/- (megaloblasts) as well as instances of erythrophagocytosis in the BM. Thus, the presence of numerous monocytes and macrophages lead to increased destruction of erythroblasts in the Lyn-/- mouse (Ingley, McCarthy et al 2004). This mechanism is highly probable in the SHIP-/- mice owing to the similarity in the myeloproliferative disorder of these mice. Further investigation will reveal if this mechanism can explain the resting erythropoietic activity of SHIP-/- mice. Related to this, autoimmune hemolytic anemia is marked by reticulocytosis, erythrophagocytosis and extramedullary erythropoiesis (Ingley, McCarthy et al. 2004). Since the phenotype of SHIP-deficient mice include immune abnormalities (Moody, Xu et al. 2004), this is yet another extrinsic mechanism that may contribute to an increased erythropoietic rate in these mice. In the Lyn-/- mouse, one group ruled out an autoantibody mechanism as the cause of the erythroid abnormalities (Harder, Qulici et al 2002). One of the surprising findings in our studies was the apparent loss of SHIP expression in late (CD71m e dTERl 19high) erythroblasts. Although SHIP expression was co-expressed with EpoR, STAT5a and GATA-1 in erythroblasts (in CD71 h i g hTERl 19 l o w / m e d population), it was lost soon after further differentiation into CD71 m e d TERl 19hlgl1 erythroblasts. Also, as observed in EpoR-expressing cell lines (Damen, Liu et al. 1993), SHIP is tyrosine 106 phosphorylated in response to Epo-stimulation. This phosphorylation was at the NPXY2 site (a PTB binding consensus) and was partially blocked by the Src-family kinase inhibitor PP2. Although Harder et al did not detect a difference in Epo-induced SHIP tyrosine phosphorylation in Lyn-/- erythroblasts, our studies show that at least one Src-family member is required for maximal tyrosine phosphorylation of SHIP (the regulation of SHIP tyrosine phosphorylation by Src-family members is explored in detail in Chapter 4). Vivian Lam in our lab has also noted loss of SHIP expression during normal human erythropoiesis (unpublished observations). It is conceivable that this turnoff is required for the late stages of erythropoiesis. Put another way, it is possible that elevated PIP3 levels are required for terminal erythroid differentiation. In keeping with this possibility, we (unpublished, not shown) and others (Bavelloni, Faenza et al. 2000) have found that the PI3K inhibitors, wortmannin and LY294002, block hexamethylene bisacetamide (HMBA)-induced or dimethyl sulfoxide (DMSO)-induced differentiation of murine erythroleukemic (MEL) cells without affecting their proliferation. This MEL cell line was generated via transformation with Friend Virus and is thought to be arrested at a CFU-E/proerythroblast stage of development (Ney and D'Andrea 2000). As well, it has been shown that ectopic expression of catalytically active, but not inactive, SHIP in the human erythroleukemic cell line, K562 (which normally does not express SHIP) does not affect its growth but blocks hemin-induced erythroid differentiation (Siegel, Li et al. 1999). It has also been shown that LY294002 blocks normal human CD34+ progenitors from expressing glycophorin A, a marker for the late stages of erythropoiesis (Myklebust, Blomhoff et al. 2002). Interestingly, our studies with SHIP+/+ and -/- myeloid progenitors also suggest that elevated PIP3 levels accelerate mast cell and macrophage differentiation (Rauh, Sly et al. 2004). Although we did not detect an intrinsic abnormality in the erythropoiesis of SHIP-/-mice, the enhanced recovery of SHIP-/- mice to PHz-induced anemia (a stress erythropoiesis mechanism) is similar to the enhanced response of GATA-1 l o w mice and contrasts sharply with the sluggish recovery of STAT5a-/-5b-/- mice. In addition, Lyn-/- mice also have been reported to recover faster from PHz-induced anemia (Ingley, McCarthy et al. 2004) although one report detected no difference in the recovery hematocrit but did detect increased erythroblast accumulation in the spleens of PHz-treated Lyn-deficient mice (Harder, Quilici et al. 2004). Since the mechanism of stress erythropoiesis caused by PHz-induced anemia is 107 primarily a GC-mediated erythroid expansion in the spleen, the a priori presence of CFU-E and erythroblasts in the SHIP-/- mice may be sufficient to explain the enhanced recovery. Thus, the anemic pressures in SHIP-/- mice caused by chronic hypoxia and high numbers of myeloid cells (leading to erythrophagocytosis or erythroid suppression by the production of suppressive factors) may cause the reticulocytosis and splenic erythropoiesis. This erythropoiesis in the spleen consequently may allow for a more rapid response to stress erythropoiesis a process known to occur predominantly in the spleen (Bauer, Tranche et al. 1999). Although we were not successful in out attempts to directly measure the resting and stress-induced levels of Epo in the plasma of SHIP-/- mice and wild-type mice, we were able to show an increase in the Epo-dependent the growth-promoting activity in SHIP-/- plasma soon after PHz-treatment. However, although the Epo levels in the SHIP-/- plasma may be elevated, the activity of the plasma is likely due to Epo acting in concert with other factors (perhaps including glucocorticoids). Interestingly, there was no difference in the growth-promoting activity of plasma obtained from resting SHIP-/- and SHIP+/+ mice, suggesting that the chronic hypoxia and therefore chronic stress erythropoiesis we suspect occurs in SHIP-/- mice has not saturated the pathways involved in the stress-mediated mechanisms (as noted for the G A T A - l l o w mice by others (Vannucchi, Nam et al. 2001)). A more sensitve assay for Epo or other erythropoietic-promoting factors may reveal differences in the plasma activating of resting SHIP+/+ and SHIP-/- mice in future endeavours. In particular, the bioassay we used to detect Epo-dependent activity in the plasma of PHz-treated mice may not be sensitive enough to measure Epo concentrations in non-anemic mice. In order to determine whether there is any case for an intrinsic defect in SHIP-/-erythroid development several studies may prove fruitful. Our lab is currently engaged in the study of the in vitro differentiation of SHIP-/- mouse embryonic stem (ES) cells into pure erythroid progenitors (Carotta, Pilat et al. 2004). Such a system of would allow direct determination of the proliferative and differentiation capabilities of SHIP-deficient erythroid progenitors at a various stages of development without the interference of large populations of myeloid cells. Furthermore, we are exploring methods of competitive reconstitution in order to determine if the erythropoietic abnormalities in the SHIP-/- mice are transplantable. Finally, related to Chapter 5 of this thesis, we are working to silence SHIP expression in highly purified human BFU-E (Krystal, Lam et al. 1994) using siRNA technology. This 108 study would allow the direct observation of the proliferation and differentiation of SHIP-deficient human erythroid progenitors in response to cytokines and inducers of differentiation (e.g., transforming growth factor beta (TGFP)) . 109 Chapter 4 SHIP IS TYROSINE PHOSPHORYLATED IN RESPONSE TO MULTIPLE STIMULI BY SRC FAMILY KINASES 4.1 INTRODUCTION The PI3K pathway plays a central role in regulating survival and proliferation, as well as many other biological responses, in eukaryotic cells via the generation of the key second messenger, PIP3. This phospholipid, which is transiently generated in the plasma membrane, attracts PH domain-containing proteins to mediate its effects (Huber, Helgason et al. 1999; Rameh and Cantley 1999). This pathway is kept in check, in hematopoietic cells, primarily by the phospholipid 5'-phosphatase, SHIP (Huber, Helgason et al. 1999). This 145 kDa protein is restricted in expression to hematopoietic cells and becomes both tyrosine phosphorylated and associated with She after cytokine, growth factor, immunoreceptor, integrin receptor or serpentine receptor stimulation of hematopoietic cells (Liu, Damen et al. 1994; Giuriato, Payrastre et al. 1997). SHIP contains two NPXY sequences that, when phosphorylated, bind the PTB domains of She (Huber, Helgason et al. 1999), Dokl (Sattler, Verma et al. 2001) and Dok2 (Tamir, Stolpa et al. 2000). However, although SHIP is a major negative regulator of the PI3K pathway in hematopoietic cells, our understanding of how SHIP activity itself is regulated has not been fully elucidated. For example, the identity of the tyrosine kinase(s) that phosphorylates SHIP is not yet known nor has the role of this phosphorylation in the function(s) of SHIP been completely elucidated. In this study, we have used the Src kinase family inhibitor, PP2, as well as Lyn-/- BMMCs to show that the kinase responsible for tyrosine phosphorylating SHIP is a Src family member, regardless of the extracellular stimulus used. Moreover, our data suggest that members of the Src family likely phosphorylate SHIP directly since we can demonstrate a physical interaction between Lyn and SHIP in B cells that involves the SH3 domain of Lyn and a previously unsuspected proline-rich region in SHIP. Lastly, to gain some insight into the repercussions of this in vivo phosphorylation, we have compared the affinity of tyrosine phosphorylated and unphosphorylated SHIP for the phosphorylated ITIM of the FcyRIIB receptor and found that phosphorylation reduces this affinity. 110 4.2 RESULTS 4.2.1 The two NPXY motifs in SHIP are tyrosine phosphorylated in response to multiple stimuli Since previous studies had established that SHIP was tyrosine phosphorylated at its two NPXY motifs following TCR activation in murine BYDP cells (Lamkin, Walk et al. 1997), we raised immunogen-purified antibodies to phosphopeptides corresponding to these two known PTB binding domains of SHIP (NPNpY 9 1 5 and NPLpY 1 0 2 2 in human SHIP) to facilitate our identification of the kinase(s) responsible for SHIP phosphorylation. Western analysis of whole cell lysates revealed that immunization with either phosphopeptide was effective at raising antibodies in rabbits that reacted far more strongly with SHIP from stimulated than unstimulated murine and human hematopoietic cells (Fig. 4.1A, left panel= NPXpYl, right panel =NPXpY2). The NPXY2 antibody was also found to be highly effective at selectively staining phosphorylated SHIP in flow cytometry (Fig. 4.1B) and immunofluorescent studies (data not shown). To further confirm the specificity of the two anti-phospho-SHIP antibodies, SHIP was immunoprecipitated from IgE + antigen (Ag)-stimulated SHIP+/+ BMMCs as well as from SHIP-/- BMMCs expressing either wild-type (WT) SHIP or a mutant SHIP in which the two NPXY motifs were replaced with phenylalanines (2NPXF). As can be seen in the top panel, the total level of tyrosine phosphorylation (Fig. 4.1C, 1st panel) was markedly reduced in IgE + Ag-treated BMMCs expressing the 2NPXF although there was a slight increase over unstimulated levels. Both the NPXY1 and the NPXY2 antibodies detected phosphorylated SHIP in the parental SHIP+/+ BMMCs and in SHIP-/- BMMCs expressing WT SHIP add-backs but not in cells expressing the 2NPXF mutant (Fig. 4.1C, 2n d and 3 r d panels). Of interest, there appeared to be a slightly higher level of constitutive phosphorylation at the NPXY1 site (compare the 2n d and 3 r d panels). As far as the slight increase over unstimulated levels in BMMCs expressing the 2NPXF SHIP was concerned, we found, serendipitously, that a commercially available anti-phospho-Btk antibody (Cell Signalling, Cat# 353IS) cross-reacted with SHIP and did so substantially more after stimulation with various ligands. This antibody was made against a 223 phosphopeptide corresponding to a region in Btk containing the sequence LpY DY and a similar sequence (LYDF) was found in both murine and human SHIP. We therefore 111 immunized rabbits with a phosphopeptide corresponding to this region (i.e., Y in murine SHIP) and found that immunogen-affinity purified antibodies to this sequence reacted more strongly to IgE + Ag-stimulated than unstimulated BMMCs expressing the 2NPXF mutant (Fig. 4.1C, 4th panel). The biological role of this phosphorylated tyrosine is currently being explored. As shown in the 5 t h panel (anti-SHIP IP) of Figure 4.1 C, similar levels of SHIP were loaded in the stimulated and unstimulated lanes. The total cell lysates from the same experiment were anlayzed for downstream phosphorylation events (right set of panels, Fig. 4.1C) to show that each of the SHIP-mutant expressing cells responded equivalently to IgE + Ag stimulation. Each of the transfected cell lines expressed the same number of surface FcsRI as measured by flow cytometry (not shown). 112 Figure 4.1: Generation and characterization of phospho-specific antibodies to the two NPXpY motifs of SHIP. (A) TF-1, WEHI 231 and BMMCs were stimulated (+) or not (-) with GM-CSF, algM (intact) and SCF, respectively, for 2 min and the TCLs subjected to Western analysis with rabbit antisera raised against the NPXpY sites and affinity-purified using the immunizing phosphopeptides corresponding to the NPXpY 1 (left panel) and NPXpY2 (right panel) sites of SHIP. (B) SHIP+/+ and SHIP-/- BMMCs were starved, preloaded with 0.1 ng/ml IgE for 16hr, stimulated with 100 ng/ml DNP-HSA for 2 min (shaded peaks) or left unstimulated (control, empty peaks), fixed, permeabilized and either stained for pSHIP (NPXpY2) followed by an anti-rabbit AlexaFluor488 secondary antibody or secondary antibody alone (not shown). (C) SHIP+/+ and transfected SHIP-/- BMMCs expressing WT SHIP or 2NPXF SHIP were starved and preloaded as in (B) for 16 hr, then stimulated ± 100 ng/ml DNP-HSA for 2 min. Cells were either solubilized for anti-SHIP IP (left panel) or TCLs. SHIP IPs were split into equal aliquots and subjected to Western analysis with the phospho-specific SHIP antibodies as indicated or with a pan-specific anti-phosphotyrosine antibody (4G10). Total SHIP protein was assessed to show equal loading of the IPs and approximately equal expression of SHIP in the transfected cells. The TCLs were analyzed for the phosphorylation of downstream effectors of IgE receptor stimulation (e.g., phosphorylation of PLCy 1 and ERK1/2) to illustrate similar levels of stimulation. The figure shown and subsequent figures in this chapter are representative of at least 2 (usually more) independent experiments with similar results. 113 TF-1 WEHI BMMC TF-1 WEHI B M M C + + - + - + mm pSHIP (NPXY1) SHIP pSHIP (NPXY2) SHIP B 3 a '8 cd — O control eE/DNP pSHIP (NPXY2) 1 11 SHIP"'" 10° to1 • a-SHIP IP(PICI) SHIP + + WT 2NPXF - + - + - + pSHIP(4GlO) pSHIP (NPXY1) pSHIP (NPXY2) pSHIP (Y 8 6 7 ) SHIP(PICI) whole cell lysate SHIP + / + WT 2NPXF + - + + *-SHIP(C+S) pPLCyl J pErk1/2 ^ - G A P D H 114 4.2.2 The Src family inhibitor PP2 blocks the tyrosine phosphorylation of SHIP, regardless of the extracellular stimulus Since it had recently been reported that Lyn tyrosine phosphorylated SHIP in the human monocytic cell line, THP-1, following M-CSF stimulation (Baran, Tridandapani et al. 2003) we tested i f this was also true with primary bone marrow-derived macrophages (BMm<j)s). Specifically, we used the Src family inhibitor PP2 (Zhu and Emerson 2002), as well as its inactive analogue, PP3, to see i f it could block SHIP'S tyrosine phosphorylation in response to M-CSF. As shown in Figure 4.2A, whole cell lysates probed with the anti-NPXpY2 antibody revealed that PP2, but not PP3, inhibited both resting and M-CSF-induced phosphorylation of SHIP. We then asked if PP2 would inhibit intact a-IgM (to co-aggregate the B C R and FcyRIIB receptors) or F(ab')2 a-IgM (to activate B C R alone) induced SHIP tyrosine phosphorylation in the human B cell line, BJAB and indeed this was the case (Fig. 4.2B, upper panels). However, since Lyn activation is a very early and critical event in BCR signalling, this was not surprising. We therefore asked if SHIP was immediately downstream of Lyn or the spleen tyrosine kinase, Syk. This was especially of interest since SHIP had previously been reported to associate with Syk, in Ag-stimulated B cells and LPS activated macrophages (Crowley, Harmer et al. 1996). We therefore tested the effect of the Syk inhibitor, piceatannol (3,4,3',5'-tetrahydroxy-trans-stilbene), and found it had only a small effect on SHIP tyrosine phosphorylation, even at a concentration approximately 5-fold higher than its IC50 (i.e., 10 pM) (Oliver, Burg et al. 1994). Even though SHIP phosphorylation was barely affected, the tyrosine phosphorylation of PLCy2, a Syk dependent event (Kurosaki 1999), was markedly inhibited, confirming the activity of the piceatannol in our studies (Fig. 4.2C, lower panels). Previous studies in our lab established that SHIP is tyrosine phosphorylated in response to several stimuli in primary B M M C s , including IL-3, SCF and IgE + Ag. In the current study, we also tested other stimuli on B M M C s and found that prostaglandin E 2 ( P G E 2 ) (which utilizes G protein-coupled receptors) (Nguyen, Solle et al. 2002) and osmotic stress (e.g., hypertonic NaCl) stimulated the tyrosine phosphorylation of SHIP. We therefore carried out time course studies in which we simulated B M M C s with these various inducers in the presence of PP2 or PP3 and, in each case, PP2 inhibited tyrosine phosphorylation of 115 SHIP at the NPXY2 site (Fig. 4.2D). In an effort to assess the specificity of PP2 for Src family members we carried out a PP2 dose response study with IL-3 stimulated BaF3 cells. As shown in Figure 4.2E, although SHIP tyrosine phosphorylation was dose-dependently inhibited by PP2, there was no effect of STAT5 phosphorylation, consistent with no effect of PP2 on JAK2 (the principal kinase involved in IL-3R signalling (Ihle 1994)). However, similar studies with SCF-induced BMMCs suggested that PP2 was equally effective at inhibiting c-kit autophosphorylation as it was at inhibiting Src family members (Fig. 4.2F), in keeping with a recent report (Tatton, Morley et al. 2003). 116 Figure 4.2: PP2 inhibits SHIP tyrosine phosphorylation in response to a variety of stimuli. Where indicated, cells were preincubated with PP2, PP3 or piceatannol (Pet) for 30 min at 37°C prior to stimulation for the times indicated (in minutes). Final concentrations of PP2 or PP3 were 20 pM (BMmfs and BMMCs) or 30 pM (BJAB) and 12.5 ng/ml Pet. (A) BMM(|>s were growth factor deprived for 16 hr and then stimulated with 100 ng/ml M-CSF. (B) BJAB cells stimulated with either 20 pg/ml goat a-human F(ab')2 IgM (left panel) or goat a-human intact a-IgM. (C) BJAB cells were stimulated for 5 min with intact a-IgM (after preincubation with PP2, PP3) or Pet and the cell lysates either solubilized with 1 x SB and used as TCLs (left panel) or immunoprecipitated (IP) with a-Lyn (H6) (2 x 107 cells) (right panel). (D) BMMCs were growth factor deprived for 16 hr, or starved and preloaded with IgE, and stimulated as indicated. (E) BaF3 cells were deprived of IL-3 for 16 hr and stimulated with 400 ng/ml IL-3 for 5 min, solubilized, immunoprecipitated (5 x 107cells) with either a-SHIP (top panel) or a-STAT5 (bottom panel). (F) BMMCs were growth factor deprived for 16 hr, then pretreated with DMSO, PP3 or PP2 as indicated prior to 5 min stimulation with 400 ng/ml SCF. Cells were solubilized and immunoprecipitated with either a-SHIP (top panel) or a-c-kit (bottom panel). 117 A M - C S F PP3 PP2 - 1 3 5 20 • fi • 5 min pSHIP (NPXY2) SHIP B F(ab') 2 a-IgM PP3 PP2 intact a-IgM PP3 PP2 2 5 2060 - 2 5 2 5 2060 - 2 5 1 1 % pSHIP ( N P X Y 2 ) •SHIP Total Cell Lysate PP3 PP2 Pet - + p S H I P ( N P X Y 2 ) p P L C Y 2 ( Y 7 8 3 ) pErkl /2 G A P D H a-Lyn IP PP3 PP2 Pet - + - + - + pSrc ( Y 4 1 6 ) 118 D IL-3 SCF PP3 PP2 PP3 PP2 1 3 5 20 1 5 ***m « M M § mmm mmm mm# 1 3 5 20 1 5 min pSHIP (NPXY2) SHIP IgE/DNP PP3 PP2 PGE, PP3 PP2 3 5 20 - 1 5 - 1 3 5 20 - 1 5 min pSHIP (NPXY2) JMMB i^ Mi' itttM' iiMiik liisiis I i w" jflv--v«v ...'.(....jip B B f SHRI SHIP 0.3 M NaCl PP3 PP2 1 3 5 20 1 5 min H j^l fi^ fc WPP pSHIP (NPXY2) SHIP 119 a - S H I P ( P l C l ) IP PP2 PP3 1 10 100 100 100 100 ^iM + + + - + - IL-3 pSHIP (a -pTyr ) SHIP a S T A T 5 IP p S T A T 5 (a -pTyr ) S T A T 5 a S H I P ( P l C l ) I P DMSO PP3 J>P2__ 10 0.1 1 10 nM - + + + + + S C F dBMtO^ ^ ^ ^ ^ ^ pSHIP (a-pTyr) SHIP a c -K i t IP p c -Ki t (a-pTyr) c -Ki t 120 4.2.3 SHIP phosphorylation is reduced in Lyn-/- BMMCs in response to multiple stimuli Because of the inherent specificity problems associated with PP2 we wanted to confirm that Src family kinases were responsible for the tyrosine phosphorylation of SHIP using a genetic approach and therefore generated BMMCs from Lyn+/+ and -/- mice. As can be seen in Figure 4.3A & B, we found that, Lyn-/- BMMCs had reduced SHIP tyrosine phosphorylation at the NPXY2 site compared to Lyn+/+ littermates in response to all stimuli tested (i.e., SCF, IL-3, hypertonic NaCl, PGE2 and IgE + antigen (Ag)). Interestingly, Lyn appeared to be particularly important for SHIP'S tyrosine phosphorylation in response to FcsRl crosslinking and for IL-3R stimulation. While Lyn is the most prevalent Src kinase in BMMCs (Hernandez-Hansen, Mackay et al. 2004), it is by no means the only one in these cells. Src, Fyn, and Yes are ubiquitously expressed and Hck and Fgr are found in most myeloid cell lineages (Thomas and Brugge 1997). Related to this, Rivera's group recently showed that Fyn carries out a distinct set of signalling events in response to IgE + Ag than that elicited by Lyn and these events are important for degranulation (Parravicini, Gadina et al. 2002). Because of the potential for redundancy in the functioning of Src-family members, we treated Lyn-/- BMMCs with PP2 to determine if the residual tyrosine phosphorylation of SHIP observed in these cells could be attributed to other Src kinases. As shown in Figure 4.3A & B, the residual SHIP phosphorylation in the Lyn -/- BMMCs was significantly reduced when preincubated with PP2 suggesting that other Src kinases may also be involved in the tyrosine phosphorylation of SHIP in BMMCs and can at least partially compensate for the loss of Lyn expression. One caveat in these studies, however, particularly those involving SCF, is that PP2 could be reducing SHIP phosphorylation by directly inhibiting c-kit's intrinsic tyrosine kinase. To more carefully explore the ramifications of the loss of Lyn on all of SHIP'S tyrosine phosphorylation sites we carried out a time course study with SCF-stimulated BMMCs in Lyn+/+ and -/- BMMCs, immunoprecipitated SHIP from these cells and Western analysis using a pan-specific anti-phosphotyrosine antibody (4G10). As can be seen in Figure 4.3C, we found that the tyrosine phosphorylation of SHIP was both delayed and short-lived in Lyn-/- BMMCs. A reprobing with anti-SHIP antibodies established equal loading. 121 Figure 4.3: SHIP tyrosine phosphorylation in Lyn-/- BMMC is reduced compared to Lyn+/+ BMMCs in response to a variety of stimuli. (A) BMMCs derived from Lyn+/+ and Lyn-/- mice were growth factor deprived for 16 hr and subjected to 5 min stimulation, as indicated, after incubation with 30 pM PP3 or PP2. (B) BMMCs were growth factor deprived and preloaded with 0.1 ng/ml IgE for 16hr, incubated with 30 pM PP3 or PP2 and stimulated with DNP-HSA (Ag) for 5 min. (C) BMMCs were growth factor deprived for 16 hr, then stimulated with 400 ng/ml SCF for the indicated times. Cells were lysed and a-SHIP IP performed. For all, samples were fractionated by SDS-PAGE and analyzed by immunoblotting with the indicated antibodies. 122 no stimulation +/+ -/-SCF +/+ -/-IL-3 +/+ -/-PP3 PP2 PP3 PP2 PP3 PP2 PP3 PP2 PP3 PP2 PP3 PP2 pSHIP (NPXY2) SHIP GAPDH 0.3M NaCl +/+ -/-PGE 2 +/+ -/-PP3 PP2 PP3 PP2 PP3 PP2 PP3 PP2 pSHIP (NPXY2) SHIP GAPDH B IgE preload +/+ -/-IgE + Ag +/+ -/-PP3 PP2 PP3 PP2 PP3 PP2 PP3 PP2 mm* WMWH iiWMiii^ in'iiiii dijiiidiiiHfe .ajNMb^u-OT Wm W£t pSHIP (NPXY2) SHIP GAPDH a S H I P IP Lyn +/+ Lyn -/-0 0.5 2 5 20 0 0 5 2 5 20 min *— pSHIP (a pTyr) 123 4.2.4 Lyn associates with SHIP in BJAB cells and this is not blocked by PP2 To gain some insight into whether Lyn phosphorylates SHIP directly or indirectly via a downstream kinase we asked if Lyn could be found physically associated with SHIP at any time. Specifically, SHIP was immunoprecipitated from the human B cell line, BJAB, before and after stimulation with intact anti-IgM and the precipitates probed for Lyn. As shown in Figure 4.4A, Lyn co-precipitated with SHIP in and this association increased dramatically upon stimulation. Of note, while the F(ab')2 fragment of anti-IgM also stimulated SHIP'S tyrosine phosphorylation, it did not increase the association between SHIP and Lyn (data not shown). Intriguingly, Lyn association continued to increase with intact anti-IgM during the 120 min time course of this study even though the tyrosine phosphorylation of SHIP returned to baseline levels by this time (Fig. 4.4A). More importantly, while PP2 substantially reduced the tyrosine phosphorylaton of both SHIP and Lyn it had no effect on the increase in their anti-IgM stimulated association. Thus, SHIP'S association with Lyn in BJAB cells did not appear to be dependent on either the tyrosine phosphorylation of SHIP or Lyn. Since we could detect a small amount of Lyn in anti-SHIP IPs in unstimulated BJAB cells we wanted to gain some insight into whether SHIP and Lyn might be weakly pre-associated (and therefore sensitive to 1% Triton X-100 solubilization). Active signalling events at 4°C are consistent with closely associated proteins. We therefore examined whether we could detect Lyn in SHIP IPs at 4°C. As can be seen in Figure 4.4B, SHIP, Btk and the ITIM-containing inhibitory receptor, FcyRIIB, are rapidly tyrosine phosphorylated at 4°C and this level of tyrosine phosphorylation was maintained for up to 2 hr. The rapid and sustained tyrosine phosphorylation of these proteins likely reflect the fact that BCR induced protein kinases, but not protein tyrosine phosphatases are activated at 4°C (Kholodenko, Demin et al. 1999; Moehren, Markevich et al. 2002). Notably, Cheng et al (Cheng, Brown et al. 2001) reported that at 4°C the BCR is maximally tyrosine phosphorylated and associated with lipid rafts and that warming to 37°C reduces both events after only 2 min. Note that cold exposure alone (unstimulated lanes) did not induce tyrosine phosphorylation of these proteins. As shown in Figure 4.4C warming of BJAB cells (stimulated with anti-IgM for 1 hr at 4°C) to 37°C results in the rapid dephosphorylation of SHIP and FcyRIIB. Next, we wanted to test if Lyn could become associated with SHIP in BJAB cells stimulated with anti-IgM at 4°C and, if so, whether the association was maintained after warming. Figure 4.4E 124 shows that stimulation of BJAB cells with anti-IgM at 4°C resulted in increased Lyn association with SHIP. Furthermore, although warming of the stimulation reactions resulted in dephosphorylation of SHIP (Fig. 4.4C), the association of SHIP with Lyn was not affected by warming nor by the addition of PP2 (Fig. 4.4D). Figure 4.4D shows that the activity of PP2 was active at 4°C since PP2, but not PP3, blocked tyrosine phosphorylation of SHIP and FcyRIIB in both cold stimulated (anti-IgM); and cold stimulated, and then warmed BJAB cells. Although these results do not prove pre-association, the rapid phosphorylation of SHIP at 4°C is certainly consistent with this possibility. 125 Figure 4.4: SHIP phosphorylation is not required for Lyn association. (A) BJAB cells (2 x 107) were preincubated with 30 pM PP3 or PP2 for 30 min followed by stimulation with 20 pg/ml intact a-IgM for the indicated times, the cells solubilized and a-SHIP IPs performed. Western analysis was then carried out with the indicated antibodies. (B) BJAB cells (5 x 105) were stimulated at 4°C with 20pg/ml intact a-IgM for the indicated time. (C) BJABs were either stimulated (+) or not (-) with 20 pg/ml intact a-IgM at 4°C for 1 hr followed by warming to 37°C for the indicated times. (D) cold BJAB cells were stimulated as in (C) followed by solubilization and IP with a-SHIP antibodies. (E) BJAB cells were preincubated at 37°C with 30 pM PP3 or PP2 for 30 min followed by stimulation with 20 pg/ml intact a-IgM at 4°C for 1 hr, then warmed to 37°C for 5 min or lysed while still cold as indicated. 126 a-SHIP IP 30 u M PP3 30 u M PP2 0 2 5 20 60 120 0 2 5 20 60 120 min I- S f f l P ( P l C l ) pSHIP (NPXY2) She pSrc ( Y 4 1 6 ) pLyn ( Y 5 0 7 ) 56kDa _ 53kDa Grb2 127 B 5 15 30 60 120 180 min **•*»' « • • * •*•«<• pSHIP(NPXY2) pBtk(Y 2 2 3 ) pFcyRIIb(Y292 ) GAPDH D 0 2 5 min pSHIP(NPXY2) pBtk(Y 2 2 3 ) pFcyRIIb(Y292 ) GAPDH PP3 PP2 0 5 0 5 ., -:5:: n • pSHIP(NPXY2) pFcyRIIb(Y292 ) GAPDH F a-SHIP IP PP3 PP2 0 5 - 0 5 min -SHIP 56/53kDa Lyn 128 4.2.5 PP2 does not block BCR internalization BCR activation involving co-aggregation of the inhibitory receptor, FcyRIIB, results in the recruitment of SHIP to the phosphorylated ITIM of this inhibitory receptor. This recruitment can be observed in the co-capping of SHIP with the BCR when stimulated with intact-IgM. This association does not occur with BCR crosslinking in the absence of FcyRIIB participation (e.g., with the F(ab')2 fragment of anti-IgM) (Phee, Rodgers et al. 2001). We were interested in knowing if PP2 would still permit the internalization of the BCR when stimulated with intact anti-IgM. Others have reported that PP2 does not inhibit the formation of lipid rafts nor the translocation of BCR to lipid rafts in activated B cells (Cheng, Brown et al. 2001). We therefore stimulated BJAB cells with intact anti-IgM in the presence and absence of PP2 and found that, PP2 delayed, but did not block BCR internalization (Fig. 4.5). One caveat here is that we are equating acid resistance to internalization and although this is an assumption that is accepted in scientific circles, it may not be correct. Interestingly this "internalization" occurred even though PP2 appeared to abolish most tyrosine phosphorylations in BJAB cells (see Fig. 4.2C). 129 PP3 PP2 Figure 4.5: PP2 does not inhibit internalization of the BCR in BJAB cells stimulated with intact a-IgM. BJAB cells were preincubated with 50 pM PP3, 25 pM PP2 or 50 pM PP2 for 30 min at 37°C. FITC-a-IgM was then added to cells at 4°C for 1 hr with mixing. Cells were warmed to 37°C for the indicated times, quenched with ice cold PBS containing 0.05% (w/v) NaN3 followed by an acid wash treatment to remove surface bound FITC-a-IgM as described in the Materials & Methods. The intensity of fluorescence remaining after acid wash was monitored by flow cytometry. The cumulative distribution functions (c.d.f) of the resulting histograms are shown. Internalization was defined as resistance to acid wash. 130 4.2.6 SHIP associates, via a previously unrecognized proline-rich motif, with the SH3 domain of Lyn While the increase in SHIP and Lyn association in BJAB cells following intact anti-IgM stimulation suggested that increased tyrosine phosphorylation of SHIP and/or Lyn might be facilitating an SH2-mediated interaction between the two, our PP2 results suggested otherwise. To test this further we examined whether the SH2 domain of Lyn could recognize and bind tyrosine phosphorylated SHIP from BCR-activated WEHI 231 cells. Conversely, SHIP also possesses an SH2 domain that could bind tyrosine phosphorylated Lyn and so we assessed the ability of GST-SHIP SH2 and GST-Lyn SH2 domains to bind to phosphorylated Lyn and SHIP, respectively. As shown in Figure 4.6A, our in vitro protein binding assays revealed that the Lyn SH2 did not interact with SHIP (upper panel) and the SHIP SH2 domain (aa 1-111) did not bind Lyn (lower panel), Interestingly, however, a GST-SHIP SH2 domain fusion protein containing the juxtaposed PXXP motif (aa 7-133), referred to as SHIP SH2+PXXP, strongly interacted with Lyn and did so equally well in resting and stimulated cells (Figure 4.6A, lower panel). This suggested that the PXXP motif in close proximity to SHIP's SH2 domain was a potential binding site for the SH3 domain of Lyn. To test this further we overexpressed an HA-tagged wild-type SHIP and a C-terminally truncated SHIP lacking all four of its C-terminal proline-rich (PXXP) motifs (Tl SHIP (aa 1-912)) (Damen, Ware et al. 2001) (shown diagrammatically in Figure 4.6B). We then carried out in vitro protein binding assays using GST-SH3 fusion proteins of the Src members Lyn, Fyn, and Yes and all three were found to bind to both the full-length wild-type SHIP and the C-terminally truncated Tl SHIP mutant. To determine whether the other 5 PXXP motifs in SHIP could participate in binding to Src family SH3 domains, several GST-SHIP fusion proteins containing proline-rich motifs (see Fig. 4.6B) were incubated with WEHI 231 cellular lysates and analyzed for Lyn association. The results, shown in Figure 4.6C, revealed that the SH2+PXXP fragment (aa 7-133), but not GST by itself nor the PXXP-containing fragments B (aa 232-415) or G (aa 994-1190), could associate with Lyn. 131 Figure 4.6: The N-terminal PXXP sequence of SHIP binds Lyn. (A) Resting WEHI 231 cells were stimulated, or not, with 20 ug/ml intact a-IgM, followed by solubilization. Cell lysates were incubated for 30 min at 4°C with 2-5 pg of immobilized GST-Lyn SH2 domain, GST-SHIP SH2 domain, or GST-SHIP SH2+PXXP. Samples were fractionated by SDS-P A G E and probed with a-SHIP and a-Lyn. (B, upper panel) The structures of H A and GFP-tagged WT and T l mutants of SHIP. (B, lower panel) Cell lysates from resting WEHI 231 cells expressing retrovirally introduced, HA-tagged WT or C-terminally truncated SHIP (Tl) were incubated with 2-5 pg immobilized GST-Lyn, -Fyn or -Yes SH3 domains for 30 min at 4°C. To determine the ability of the different SH3 domains to associate with WT and T l SHIP in vitro SDS-PAGE separated samples were probed with anti-HA antibodies. (C, upper panel) The structures of the GST-SHIP fragments. (C, lower panel) WEHI 231 cell lysates were incubated with 2-5 pg immobilized GST-SHIP SH2+PXXP, -SHIP G or -SHIP B fragments for 30 min at 4°C. SDS-PAGE separated samples were immunoblotted with a-Lyn. Immobilized GST protein was also incubated with stimulated (where appropriate) and unstimulated cell lysate as a negative control and a TCL was included as a positive control. 132 SHIP Lyn SHIP SH2 + GST SH2 SH2 PxxP TCL -/+ - + - + - + a IgM SHIP 56kDa T 53kDa L ^ B WT HA | S H 2 5'-Ptase 1 1 1 A A (1-1190 aa) 2 5 3 4 T1 HA | S H 2 5'-Ptase GFP • P X X P motif WT Tl Lyn Fyn Yes Lyn Fyn Yes GST SH3 SH3 SH3SH3 SH3SH3 H A - S H i P - O T ( W i ) H^-SHIP-CFP(T) 133 c A A A A A A Proline-rich GST SH2+ PXXP (aa 7-133) (994-1 190) GST B (232-415) SHIP Fragnents SH2+ TCL GST PxxP G B 56kDa 53kDa Lyn 134 4.2.7 The tyrosine phosphorylation of SHIP reduces its affinity for the phosphorylated ITIM of the negative co-receptor, FcyRIIB Previous studies in our lab and others have established that the tyrosine phosphorylation of SHIP at its two NPXY motifs may have at least a limited role in its ability to hydrolyze PIP3 in vivo, perhaps in part because of its reduced ability to associate with She (Damen, Ware et al. 2001; Lamkin, Walk et al. 1997). Interestingly, She may play different roles with respect to SHIP depending on the stimulus and the cell type. In response to activation of the IL-3R, for example, there is good evidence that She brings SHIP to the plasma membrane (Bone and Welham 2000) while in the activation of the negative co-receptor, FcyRIIB, it has been proposed that She actually aids in detaching SHIP from the plasma membrane (Tridandapani, Pradhan et al. 1999). To explore further the role of SHIP's tyrosine phosphorylation in the second model, we asked if the tyrosine phosphorylation of SHIP affected its ability to bind to the activated FcyRIIB. Specifically, we compared the ability of SHIP from unstimulated and SCF-stimulated BMMCs to bind to beads bearing the phosphorylated ITIM of the FcyRIIB receptor. As can be seen in Figure 4.7, equal amounts of SHIP from unstimulated BMMCs bound to the pITIM more readily than SHIP from stimulated cells. Although we chose SCF-stimulated BMMCs as a source of tyrosine phosphorylated SHIP because it triggered a pronounced tyrosine phosphorylation of SHIP, as determined by Western analysis, we cannot say at this time what proportion of total cellular SHIP is phosphorylated at the times analyzed because of the inherent limitations with determining phosphoSHIP/total SHIP in Western blots. However, if anything, our results are an underestimate of the difference between phosphorylated SHIP and unphosphorylated SHIP because of the presence of unphosphorylated SHIP in SCF-stimulated BMMCs. 135 TCL pITIM beads beads 30 30 60 + + S/N 30 30 60 mwm ......... bead volume (pi) pSHIP(NPXY2) total SHIP .15 .23 1.1 0.9 pSHIP total SHIP ratio Figure 4.7: Tyrosine phosphorylated SHIP has a reduced affinity for pITIM beads. B M M C s were starved and stimulated (+) or not (-) with 100 ng/ml SCF for 2 min at 37°C. Cell lysates were precleared with streptavidin (SA)-agarose beads and then incubated with limiting amounts of pITIM-biotin bound to SA-agarose beads at 4°C for 1 hr. The beads were washed thoroughly, boiled in 1 x SB and fractionated on SDS-PAGE for Western analysis. The pSHIP/SHIP ratio was calculated based on densitometry values. S/N is the total cell lysate supernatant (unbound material) following incubation with pITIM beads. 136 4.3 DISCUSSION We demonstrate herein, using phosphospecific antibodies to the 2 N P X Y motifs of SHIP, that these two sites are tyrosine phosphorylated following hematopoietic cell activation via cytokine receptors (IL-3R), tyrosine kinase receptors (M-CSF or SCF receptors), serpentine receptors (PGE2 receptors (EP1.4)), immunoreceptors (BCR ± FcyRIIB, FcsRI) or osmotic stress inducers (hypertonic NaCl). Interestingly, we have also identified a putative third site of phosphorylation at Y and we are currently generating a Y to F mutant SHIP at this site to evaluate its biological role. We also found, from studies utilizing both the Src kinase family inhibitor, PP2 and Lyn-/- B M M C s that SHIP's tyrosine phosphorylation in response to multiple stimuli is likely mediated via a Src family member. In retrospect, this is not surprising, since the Src family is implicated in signal propagation from every major class of cell surface receptor (Thomas and Brugge 1997) and they often interact directly with these receptors to facilitate access to their substrates (Torigoe, R et al. 1992; Rao and Mufson 1995; Adachi, Pazdrak et al. 1999; Yamanashi, Kakiuchi et al. 1991; Yamamoto, Yamanashi et al. 1993; Linnekin, DeBerry et al. 1997; Timokhina, Kissel et al. 1998; Lennartsson, Blume-Jensen et al. 1999). Additionally, we have established that Lyn and SHIP are associated in response to BCR+FcyRIIB stimulation in the human BJAB B cell line. Moreover, this interaction was not blocked by the Src family inhibitor PP2, suggesting a binding mechanism independent of tyrosine phosphorylation. Related to this, we have established that the SH3 domains of the Src family members are capable of directly binding to an N-terminal P X X P motif of SHIP. The relevant P X X P motif (PELPPR) fits the class II core sequence (PxxPxR/K) perfectly (Sparks, Rider et al. 1996; Sparks, Rider et al. 1998; Tong, Drees et al. 2002) and lies just C-terminal of SHIP'S SH2 domain and is not part of the SH2 domain consensus (Marchler-Bauer, Anderson et al. 2005). Although other Src-family SH3 domains have been shown to prefer class I ligands, a similar P X X P motif of this class was identified as a Src SH3 target (Alexandropoulos, Cheng et al. 1995). A model of the interaction of the SHIP P X X P motif (designated PXXPi ) with the SH3 domain of Lyn is depicted in Figure 4.8. 137 A P X X P motif Figure 4.8: SHIP associates with the SH3 domain of Lyn. In vitro and in vivo evidence suggests an interaction (indicated by the dashed arrow) between the SH3 domain of Lyn with the SHIP'S PXXP motif just C-terminal of SHIP's SH2 domain (PXXP motifs are depicted as open triangles). The SFK inhibitor, PP2, blocks Lyn kinase activity (solid arrows) and the tyrosine phosphorylation of the NPXY motifs near SHIP's C-terminus. PP2 does not block the interaction of SHIP and Lyn in B cells following BCR + FcyRIIB co-aggregation. In support of our findings in this thesis, Lyn has been shown to be required for SHIP tyrosine phosphorylation in B cells (Hibbs, Harder et al. 2002). Furthermore, SHIP has previously been shown to be a substrate for both Lyn and Lck in vitro (Osborne, Zenner et al. 1996; Phee, Jacob et al. 2000; Sarmay, Koncz et al. 1999; Lamkin, Walk et al. 1997), and others have reported that c-Src is constitutively bound to SHIP in human platelets (Giuriato, Bodin et al. 2000). In addition, both Lyn and Fgr have been found to associate with SHIP in unstimulated (Lyn) and following CDwl50 (SLAM) activation in B cells (Lyn and Fgr) - a stimulus that results in transient tyrosine dephosphorylation of SHIP (Mikhalap, Shlapatska et al. 1999). Although we did not detect a similar constitutive interaction in BJAB B cells, our studies at 4°C are suggestive of a close association of SHIP and Lyn in resting B cells. 138 A similar inference has been used previously to predict a close association of the IL-3R with members of the JAK family (Okuda, Druker et al. 1991; Miyajima, Mui et al. 1993). Finally, in the mouse B cell line (WEHI 231), we have detected Lyn in anti-SHIP immunoprecipitates before BCR activation with increasing association following activation (not shown). Although Syk had previously been reported to associate with SHIP in antigen-stimulated B cells and LPS-activated macrophages (Crowley, Harmer et al. 1996), our results using the Syk inhibitor piceatannol demonstrated a limited role for Syk in SHIP tyrosine phosphorylation following BCR activation. This is consistent with the observation that SHIP tyrosine phosphorylation was not reduced in Syk-/- mast cell lines stimulated by FceRI activation (Kimura, Sakamoto et al. 1997). In the THP-1 human monocyte cell line, Lyn was detected in SHIP immunoprecipitates from lysates prepared form M-CSF stimulated but not unstimulated cells (Baran, Tridandapani et al. 2003). Furthermore, this same group has suggested a possible model of Lyn-SHIP interaction involving the SHIP SH2 domain and unspecified phospho tyro sine residues of Lyn. However, they also showed that PP2 did not block this association. Importantly, this group does not report the full sequence of the SH2 domain used in their studies, and it would be interesting to know if the SH2-proximal PXXP domain that we have identified as a putative Src-family SH3 interaction site in this study is included in their SH2 amino acid sequence (Baran, Tridandapani et al. 2003). A recent report illustrates that the SH3 domain of at least one SFK can associate with another protein in a non-proline and non-phosphotyrosine dependent manner. Chan et al (Cahn, Lanyi et al. 2003) have shown that the SH2 domain of the SLAM-associated protein (SAP) binds to the SH3 domain of FynT (Fyn isoform expressed in T cells) and couples FynT to SLAM (CD 150). "SAP binds the FynT SH3 through a surface-surface interaction that does not involve canonical SH3 or SH2 binding interactions" (Chan, Lanyi et al. 2003). This SAP-FynT association promotes the activation of the FynT kinase domain (Chan, Lanyi et al. 2003). Since Baran et al (Baran, Tridandapani et al. 2003) have reported a stimulation-dependent (MCSF receptor activation) but not phosphotyrosine-dependent interaction between SHIP and Lyn in a macrophage cell line, and have implicated the SHIP SH2 in the binding interaction, it is possible that, like the FynT SH3 domain, the Lyn SH3 domain may have the ability to associate with SH2 domain surfaces (Cahn, Lanyi et al. 2003). Curiously, 139 the SAP SH2 domain amino acid sequence shares high identity/homology with the SHIP SH2. However, the SHIP SH2 appears to lack the critical residues that are required for interaction with the FynT SH3 domain. It remains to be determined whether an analogous interaction between SFK SH3 domains and the SHIP SH2 is possible. Recently, in mast cells, the submembranous F-actin skeleton has been shown to be essential for FcyRIIB-dependent negative regulation of the IgE receptor (FccRI). In addition, SHIP was shown to be constitutively associated with filamin-1 and the submembranous skeleton in mast cells. Upon co-aggregation, FcsRI and FcyRIIB interact with the F-actin skeleton and engage SHIP and filamin-1. Following co-aggregation, filamin-l and F-actin dissociate from Fc receptor (FcR) complexes, whereas SHIP remains associated with FcyRIIB (via SHIP'S interaction with the tyrosine phosphorylated ITIM motif). Thus, at least in mast cells, the submembranous F-actin skeleton may act as a 'donor' of SHIP to inhibitory signalling complexes (Lesourne, Fridman et al. 2005). Although association of SHIP with the ITIM containing receptor requires tyrosine phosphorylation of the ITIM motif, the initial stages of complex formation (i.e., the redistribution of FcR's and SHIP into the microdomains following stimulation) did not require the cytoplasmic domain of the FcyRIIB (and thus ITIM phosphorylation). This suggests that the physical concentration of signalling components within microdomains may permit phosphorylation-independent interactions (such as those between Lyn and SHIP proposed in this thesis) in the absence of enzyme-dependent signalling events such as tyrosine phosphorylation. Although Lyn is predominantly associated with lipid raft microdomains and the FcyRIIB and SHIP are associated with the submembranous cytoskeletal domains, the SFK-dependent phosphorylation of the FcR's and SHIP suggest that these microdomains may interact. The results of our studies raise the intriguing possibility that SHIP may be recruited into lipid raft microdomains in the absence of tyrosine phosphorylation events. In particular, the presence of PP2, which effectively blocks the tyrosine phosphorylation of the FcyRIIB ITIM, SHIP and, predominantly, the activation of the Lyn, does not appear to block association of SHIP and Lyn. However, stable Lyn-SHIP association in our studies was only detected in the context of BCR+FcyRIIB co-aggregation. This was despite our demonstration that SFKs, especially Lyn, are responsible for the tyrosine phosphorylation of SHIP in response to other stimuli in these cells, and multiple stimuli in many other cell types. 140 Of note, and as mentioned above, other groups have detected SHIP binding various SFKs, including Lyn. Our in vitro binding data and in vivo experiments suggest a mechanism whereby, in BCR+FcyRIIB stimulated B cells, SHIP is colocalized with Lyn (and presumably the BCR complex following co-aggregation of the B C R and FcyRIIB (see Figure 4.9A and 4.9B). SHIP has been detected in lipid raft microdomains following B C R activation in the context of co-aggregation of the FcyRIIB co-receptor (Aman, Tosello-Trampont et al. 2001). It remains to be determined if the F-actin submembranous cytoskeleton associates and 'donates' SHIP to Lyn-containing lipid raft domains in B cells (as reported in mast cells) or if ITIM tyrosine phosphorylation is a prerequisite for SHIP localization to lipid rafts. The concentration of SHIP, Lyn and the FcyRIIB in a signalling compartment may then allow the interaction of SHIP and Lyn via Lyn's SH3 domain and the SH2-proximal P X X P motif we have identified. This association may stabilize the active/open conformation of Lyn and promote the tyrosine phosphorylation of the ITIM motif within the FcyRIIB. The phosphorylated ITIM motifs may then attract SHIP, which may or may not remain in a complex with Lyn. Phosphorylation of the 2 N P X Y motifs within SHIP's C-terminus, by Lyn, would then allow binding of She's PTB domain followed by tyrosine phosphorylation of She by Lyn. Our data and reports by others support the contention that following tyrosine phosphorylation of SHIP and its association with She, SHIP disengages the FcyRIIB. This dissociation may be promoted by either SHIP's association with She and/or binding to Lyn's SH3 domain. The SHIP-Lyn complex appears to be long-lived following BCR+FcyRIIB activation and is present in BJAB B cells long after other proteins in the signalling complex (FcyRIIB, She and SHIP) have returned to baseline-phosphorylated levels. The presence of PP2, which inhibits the tyrosine phosphorylation of the inhibitory receptor, SHIP, She and Lyn, does not block the association of SHIP with Lyn in BCR+FcyRIIB activated BJAB cells. Thus, our model requires that the co-aggregation of the ITIM-containing inhibitory receptor with the B C R complex and other signalling intermediates is independent of SFK-mediated tyrosine phoshorylation events and, furthermore, the lack of SFK activity does not preclude the binding of SHIP to Lyn within the signalling complex (Figure 4.9C). The biological importance of SHIP phosphorylation on tyrosine residues is not well understood. One possibility is that this post-translational modification regulates SHIP's 141 enzymatic activity directly. However, our work and that of others have not been able to show a difference in the phosphatase activity of SHIP immurioprecipitates from unstimulated and stimulated cells (Damen, Liu et al. 1996; Phee, Jacob et al. 2000). Recent studies have shown that SHIP tyrosine phosphorylation and association with the FcyRIIB within the BCR complex is critical for the recruitment of several proteins to the membrane. Furthermore the recruitment of She to phosphorylated SHIP appears to regulate the tyrosine phosphorylation of She in B cells (Ingham, Okada et al. 1999) and may down-regulate SHIP activity by causing dissociation of FcyRIIB/SHIP complexes (Tridandapani, Pradhan et al. 1999). Such a mechanism would distance SHIP from its substrate, PIP3. Our findings in vitro support the idea that tyrosine phosphorylated SHIP as a lower affinity for the pITIM of FcyRIIB when competing with unphosphorylated SHIP for limited binding sites. 142 Figure 4.9: Models for interactions of SHIP, Lyn, She and FcyRIIB. (A) In resting mature B cells Src family kinases (e.g., Lyn) are located primarily in membrane lipid rafts but have also been detected in the F-actin submembranous skeleton in some cells (light coloured areas represent signalling microdomains). The B cell receptor (BCR) complex and the ITIM-containing inhibitory co-receptor (FcyRIIB) are primarily outside of lipid rafts within the plasma membrane before activation. SHIP has peri-plasma membrane localization before stimulation and may be associated with the F-actin submembranous skeleton. SHIP and Lyn (or other SFKs) may be loosely associated in some resting cells. (B) Association of SHIP, Lyn and SHIP in mature B cells after co-clustering of BCR and FcyRIIB. (1) The co-aggregation of the BCR with the FcyRIIB results in clustering of the BCR+FcyRIIB with activated Lyn within lipid rafts or other signalling microdomains (depicted as light coloured membrane regions). (2) Active Lyn then contributes to the phosphorylation of the BCR complex ITAMs (not depicted) and is the primary kinase responsible for phosphorylation of the ITIM motif in FcyRIIB (sites of tyrosine phosphorylation indicated by encircled 'P'). SHIP is initially concentrate with the FcR's and Lyn within signalling microdomains following co-aggregation and is further stabilized within these domains by both the active conformation of Lyn (via an interaction with Lyn's SH3 and the SH2 proximal PxxP motif of SHIP) and/or the association of SHIP's SH2 domain with the phosphorylated ITIM motif. (3) Active Lyn then phosphorylates SHIP at its NPXY motifs which (4) allows SHIP to bind the PTB domain of She. (5) SHIP's association with She and/or association of SHIP with the active conformation of the Lyn SH3 results in dissociation of SHIP from the FcyRIIB. (6) The protein complexes are subsequently dephosphorylated by a variety of protein tyrosine phosphatases that have not been fully described but include SHP-1 and SHP-2. Following dephosphorylation of the NPXY domains of SHIP, She dissociates from the complex. SHIP and Lyn may remain associated in a phosphorylation independent manner and may serve to retain SFIIP within signalling platforms. (C) Association of SHIP, Lyn and SHIP in mature B cells in the presence of PP2. (1) The co-clustering of the BCR+FcyRIIB and Lyn in signalling microdomains does not require Lyn kinase activity. (2) Lyn in an open conformation (but not fully active) allows binding of SHIP via Lyn's SH3. This interaction may stabilize the open conformation of Lyn and, in turn, the association of SHIP with membrane microdomains in the absence of tyrosine phosphorylated FcyRIIB. (3) The presence of PP2 blocks the Lyn-dependent tyrosine phosphorylation of SHIP and therefore blocks SHIP-Shc association. (4) SHIP and Lyn remain bound via a phosphorylation independent mechanism. Protein-protein interactions are depicted as dotted arrows. Tyrosine phosphorylation or dephosphorylation events are depicted with solid arrows. 143 144 145 c B C R + F c y R I I B 146 Chapter 5 SHIP-DEFICIENT TF-1 CELLS DISPLAY CYTOKINE-INDEPENDENT GROWTH/SURVIVAL BUT REDUCED RESPONSIVENESS TO HIGH CYTOKINE LEVELS 5.1 INTRODUCTION The generation of SHIP-/- mice in our lab in 1998 has allowed our group and others to make rapid progress in our understanding of the role that SHIP plays in murine hematopoietic progenitors and mature blood cells. However, progress with human SHIP has been much slower because of the absence of naturally occurring mutations or deletions in SHIP in the human population. However, it is well known that the PI3K pathway plays a similar, positive role in murine and human cells in stimulating proliferation and differentiation. Moreover, the presence of biallelic inactivating mutations of PTEN in a large number of advanced human cancers has shown that tightly regulating the PI3K pathway in human cells is critical in order to avoid uncontrolled cell growth. In addition, as mentioned in the Introduction, a number of recent reports suggest that SHIP may play a role in repressing mast cell activation since a subgroup of hyper-allergic people appear to express substantially lower than normal levels of SHIP protein in their basophils (MacDonald and Vonakis 2002). As well, an inactivating mutation in human SHIP has very recently been found within the blast cells of a patient with AML (Luo, Yoshida et al. 2003), suggesting that SHIP may serve as a tumour suppressor in human hematopoietic cells. Lastly, it has been shown that the presence of the tyrosine kinase fusion protein, BCR-ABL, in CML patients not only leads to the constitutive phosphorylation of SHIP and its association with She but markedly reduces the levels of SHIP in late CML progenitors and it is possible that reduced SHIP activity might be a prerequisite for the proliferative advantage of some CML clones. Interestingly, however, in early (i.e., CD34+) CML progenitors SHIP is elevated compared to that in normal CD34+ cells (Jiang, Stuible et al. 2003) and this elevation in SHIP may be an attempt by the abnormal primitive progenitors to counter the BCR-ABL-induced over-activation of the PI3K pathway. 147 To investigate further the role of SHIP in human hematopoietic cells we used the newly developed small interfering RNA (siRNA) technology to deplete specifically the levels of SHIP in the human erythroleukemic cell line, TF-1. By careful selection of an appropriate targeting sequence, we were able to markedly reduce SHIP protein levels in these cells and, as shown below, dramatically affect the biological properties of these cells. 5.2 RESULTS 5.2.1 Silencing human SHIP expression in TF-1 cells To study the role of human SHIP (hSHIP) we transfected the human erythroleukemic cell line, TF-1, with several small interfering RNAs directed against different regions of hSHIP mRNA in an effort to silence SHIP protein expression. With one of these constructs in particular, we were able to achieve efficient transfection of the siRNA and, after 48 hr, we could detect only small amounts hSHIP protein by Western blotting and this knockdown persisted for at least 5 days (Fig. 5.1A). This is consistent with previous reports using siRNA to knock down other gene products (i.e., gene silencing with siRNA constructs is transient but typically capable of effective gene silencing over many days and through many cell divisions (Agrawal, Dasaradhi et al. 2003) even after washing the cells out of the transfection reagent). In our experiments, the silencing of hSHIP protein expression was maintained for at least 5 days in GM-CSF following transfection and often longer (Fig. 5.1 A). SHIP knockdown also lasted several days in the absence of growth factor and will be discussed below. Further analysis by intracellular flow cytometry revealed two populations of cells in transfected cultures (Fig. 5.1B). Typically, we were able to achieve almost complete silencing of hSHIP protein expression in about 80 - 90% of the population with more limited knockdown in the remainder of cells (i.e., 10-20%). For the remainder of the experiments described in this chapter, the efficiency of hSHIP silencing was assayed in parallel either by Western blotting or by intracellular flow cytometry. 148 B siNS s iSHIP SHIP expression Figure 5.1: TF-1 ceils transfected with siRNA to human SHIP (siSHIP) or with a non-silencing control siRNA (siNS). (A) Western analysis of hSHIP protein expression in TF-1 cells 2 and 5 days after siRNA transfection. GAPDH is used as a loading control. (B) Intracellular flow cytometry of TF-1 cells 2 days after transfection with siNS or siSHIP. Fixed and permeablized cells were labelled with either mouse a-hSHIP (shaded peaks) or isotype control antibody (empty peaks), followed by cc-mouse IgG AlexaFluor647. This figure and subsequent figures in this chapter are representative of multiple experiments with similar results. 149 5.2.2 Akt, GSK3 and ERK phosphorylations are enhanced in SHIP-deficient TF-1 cells Since SHIP is known to be an important negative regulator of the PI3K/Akt pathway through its ability to hydrolyze PIP3, a key secondary messenger upstream of Akt phosphorylation/activation, we first wanted to determine if SHIP knockdown would affect the phosphorylation levels of Akt in response to cell stimulation. We therefore transfected TF-1 cells with si SHIP, starved the transfected cells overnight and then restimulated with GM-CSF for various times. As shown in Figure 5.2, stimulation of siSHIP-transfected cells with recombinant human GM-CSF revealed enhanced phosphorylation of Akt at serine 473, compared to TF-1 cells similarly transfected with a non-silencing siRNA (siNS) control. In addition, the phosphorylation of GSK-3 P, a direct downstream target of Akt, was also enhanced in GM-CSF-stimulated SHIP-deficient TF-1 cells. Although indirect, the enhanced Akt phosphorylation at a key activation site and the subsequent phosphorylation of a downstream regulator of the PI3K/Akt pathway is evidence of increased concentration of PIP3 in SHIP-deficient TF-1 cells in response to GM-CSF stimulation. As well, as can be seen Figure 5.2, the phosphorylation of ERK1/2 was enhanced in siSHIP transfected cells compared to siNS controls. This is consistent with previous reports suggesting that in hematopoietic cells, the PI3K pathway activates ERK1/2 by virtue of the production of PIP3 and the subsequent recruitment of PKC isoforms to the plasma membrane (Klingmuller, Wu et al. 1997). These PKCs, in turn, phosphorylate/activate Raf-1 and this phosphorylates/activates ERK 1/2. Alternatively, a lack of SHIP could lead to enhanced ERK 1/2 phosphorylation because SHIP becomes associated with She upon the activation of a variety of membrane receptors (Liu, Damen et al. 1994; Damen, Liu et al. 1996), including cytokine receptors like the GM-CSF receptor and it has been postulated by us (Liu, Damen et al. 1994) that SHIP may sequester She and, by this mechanism, attenuate Ras activation of the subsequent phosphorylation of ERK1/2. Interestingly, this phosphorylation of ERK1/2, as well as GSK3, appeared to peak earlier in siSHIP transfected cells, (i.e., at 2 min) and then declined rapidly. In comparison, the phosphorylation of ERK1/2 and GSK3 in siNS control transected TF-1 cells was less rapid but perhaps more prolonged. 150 siNS siSHIP 0 2 5 10 30 0 2 5 10 30 min 4 -fj-ii iriMiiiiiMirt sHflHBI pSHIP (NPXY2) SHIP <«- pAkt(S 4 7 3) *- Akt pGSK3p 1= pErkl/2 4- Erkl/2 Figure 5.2: GM-CSF stimulation of TF-1 cells transfected with siSHIP display altered signalling. siNS and siSHIP transfected TF-1 cells were stimulated with 40 ng/ml hGM-CSF for the indicated times. The cells were then lysed with 1 x SB and TCLs separated by SDS-PAGE and subjected to immunoblot analysis with the indicated phospho-specific antibodies. The blots were then reprobed for total protein. 151 5.2.3 SHIP-deficient TF-1 cells proliferate more at low but less at high cytokine concentrations The loss of SHIP's enzymatic function, either by a mutation in its phosphatase domain or loss of expression is associated with a growth advantage in response to many cytokines and growth factors (Rauh et al 2004). For example, myeloid progenitors derived from SHIP-/- mice are hyperresponsive to M-CSF, SCF, IL-3 and G-CSF in CFC assays (Helgason, Damen et al. 1998). In addition, a dominant-negative hSHIP mutant (V684E) found in the blood of a patient with AML resulted in a growth advantage in low serum concentrations and enhanced survival in the absence of serum (Luo, Yoshida et al. 2003). We therefore wanted to test the proliferative potential of SHIP-deficient TF-1 cells in response to cytokines. Specifically, TF-1 cells were transfected with siSHIP or sins, as described above, and after 48 hr were washed out of media containing GM-CSF and reseeded in media containing increasing doses of GM-CSF, IL-3 or Epo. We then monitored their rate of DNA synthesis using H-Tdr and found, as shown in Figure 5.3, that siSHIP transfected cells were capable of proliferating to some extent in only heat-inactivated fetal calf serum (AFCS) as well as with very low concentrations of cytokine whereas the control cells did not. However, at high concentrations of GM-CSF, IL-3 and Epo the control cells proliferated better than the SHIP knockdown cells. This very reproducible result (i.e., 3 separate experiments) suggested that the presence of SHIP might offer a growth advantage to hematopoietic progenitors, but only under what are usually non-physiologically elevated levels of cytokines. Interestingly, the proliferation in response to IL-5, which was less than with the other cytokines (probably because of the reported low levels of the IL-5Ra subunit on the surface of TF-1 cells (Yen, Hsieh et al. 1995), did not show this inhibition at high IL-5 levels. This could be because the cells were not proliferating as rapidly with IL-5 and therefore a key signalling intermediate (e.g., perhaps SHIP-generated PI-3,4-P2) was not becoming limiting. 152 hGM-CSF hlL-3 100000 £ o)^ 75000 — c - . 2 1,2 50000-1 I— Q. § 25000 10° I M I H ! | ! I 111111} I I l l l M f I 101 102 103 I I 1 104 10s pg/mL hll_-5 30000 E c (cp T3 c 20000-iymi o "5 o X 110000-o c , - 0 - 0 - 0 — 0 - 0 I I I M l i q I I—I I I Min I I I i i n q i i i 10° 101 102 103 10" 105 pg/mL 75000-1 50000 25000H 75000-1 50000 25000-m | — i i inrai—i 11imq—i 11miq 10° 101 102 103 pg/mL hEpo 104 105 •T I'minj— 10° 101 102 103 mU/mL •PI i 11 nm 104 105 Figure 5.3: TF-1 cells transfected with siSHIP display enhanced proliferation in the absence of cytokines. 3H-Tdr incorporation assays of TF-1 cells 2 days after siRNA transfection. Proliferation of siSHIP (red open diamonds, dashed lines) TF-1 cells in response to increasing doses of hGM-CSF, rhIL-3, rhIL-5 or hEpo compared to siNS-transfected (blue solid diamonds, solid line) TF-1 control cells. Each point is the mean of duplicates ± SEM. 153 5.2.4 SHIP deficient TF-1 cells survive longer in the absence of cytokines TF-1 cells deprived of cytokines typically start to die in one or two days (Kitamura, Tange et al. 1989; Bradbury, Zhu et al. 1994). Cell death has all the hallmarks of apoptosis and involves the activation of caspase cascades (Kitamura, Tange et al. 1989; Bradbury, Zhu et al. 1994). Since in 10% AFCS alone we detected a small amount of 3H-Tdr incorporation in SHIP-deficient TF-1 cells (Fig. 5.3), we surmised that silencing SHIP expression would allow for limited cytokine-independent proliferation or at least enhanced survival of TF-1 cells. To determine if this was true, we monitored the cytokine-deprived survival of TF-1 cells by trypan blue exclusion following siRNA transfection with either siSHIP or siNS. As shown in Figure 5.4A, SHIP-deficient TF-1 cells were capable of prolonged survival compared to SHIP-expressing TF-1 cells. Whereas the siNS TF-1 cells survived for about 1-2 days, by the third day most of the cells were dead. However, TF-1 cells depleted of SHIP protein survived well past the third day and in fact experienced a slight increase in cell numbers over the first 1-2 days. Since we had previously starved siSHIP- and siNS-transfected cells in very low serum conditions (0.5 % FCS) and had found that both died rapidly, it appeared that the survival of SHIP-deficient cells, while not cytokine dependent per se, did have a nutrient or factor requirement provide by the sera. We therefore repeated the survival experiments using 30% FCS (Fig. 5.4B) in place of the 10% AFCS and found that, although the survival of both the non-silencing and the SHIP-deficient TF-1 cells were enhanced in 30% FCS, SHIP-deficient cells had a decided survival advantage. Although we had previously found that the presence of bovine IGF-1 in FCS is an important co-stimulatory factor for the Epo-mediated growth of BAER cells, neither the addition of IGF-1, nor neutralizing antibody against IGF-1, had any effect on the survival of siSHIP- or siNS-transfected TF-1 cells maintained in serum. 154 A 1 0 % A F C S lOOOOOn B 3 0 % F C S IOOOOOT 0 1 2 3 4 5 6 7 Days Figure 5.4: TF-1 cell survival is enhanced by SHIP knockdown. (A) TF-1 cells transfected with siNS (blue solid squares, solid line) or siSHIP (red solid triangles, dashed line) siRNA were grown in 10% AFCS and the total number of viable cells counted daily using trypan blue exclusion. (B) TF-1 cells transfected with siNS (blue solid squares, solid line) or siSHIP (red solid triangles, dashed line) were grown in 30% FCS and the total number of viable cells counted daily by trypan blue exclusion. Each point is the mean of duplicates ± SEM. 155 5.2.5 The survival of SHIP-deficient TF-1 cells is dependent on PI3K Since we established that SHIP-deficient TF-1 cells had enhanced Akt phosphorylation, and by inference, enhanced PI3K activity, we wanted to test whether a pharmacological inhibitor of PI3K would block the survival of SHIP-deficient cells. We therefore monitored the survival of siNS- and si SHIP-transfected TF-1 cells (in 10% AFCS) in the presence of LY294002, a potent inhibitor of PI3K or its inactive analog, LY303511. As can be seen in Figure 5.5A & B, the addition of LY294002 (Fig. 5.5B, right panel), but not its inactive chemical analog, LY303511 (Fig. 5.5A, right panel), blocked the survival of SHIP-deficient TF-1 cells in a dose-dependent manner. At a concentration as low as 1 pM, LY294002 significantly reduced survival and at 5 pM, the survival was as poor as that seen with siNS-transfected TF-1 cells. These concentrations of LY294002 did not have significant effect on the GM-CSF mediated growth of SHIP-deficient or siNS transfected TF-1 cells (not shown). Interestingly, LY303511 at the higher dose of 5 pM did have a mild deleterious effect on the survival of the SHIP-deficient cells compared to the cultures with only DMSO as a vehicle control. Also worthy of note is that the survival of the control siNS-transfected TF-1 cells was not substantially diminished by LY294002 (Fig. 5.5B, left panel) indicating that the PI3K activity in these cells, in the absence of this inhibitor, was not contributing significantly to it's, albeit low, survival. 156 Figure 5.5: Cytokine-independent survival of SHIP-deficient TF-1 cells requires active PI3K. (A) TF-1 cells transfected with either non-silencing siRNA (siNS) (left panel, in blue) or siSHIP (right panel, in red) were grown in IMDM containing 30% FCS with either DMSO (• solid diamonds, solid lines), 30% FCS with 1 pM LY303511 (V open down triangles, dashed line) or 5 pM LY303511 (0 open diamonds, dashed line) and viable cells counted on the indicated days. (B) TF-1 cells transfected with either siNS (left panel, blue) or siSHIP (right panel, red) were grown in IMDM containing 30% FCS with DMSO (• solid diamonds, solid line), 30% FCS with 1 pM LY294002 (V open down triangles, dashed line) or 5 pM LY294002 (0 open diamonds, dashed line). Each point is the mean of duplicates ± SEM. 157 A siNS (±LY303511) siSHIP (± LY303511) 80000-1 "g 60000-| a | 40000 TS H 20000 B siNS (± LY294002) siSHIP (± LY294002) 80000-1 Days 158 5.2.6 SHIP-deficient TF-1 cells display elevated levels Mcl-1 levels upon starvation In a previous study, Huang et al (Huang, Huang et al. 2000) demonstrated that TF-1 cell survival correlated with the maintenance of the anti-apoptotic Bcl-2 family member, myeloid cell leukemia sequence 1 (Mcl-1), and that this maintenance of high Mcl-1 levels was dependent on both the ERK and PI3K pathways (Huang, Huang et al. 2000). Subsequently, Schubert and Duronio showed that the ERK pathway was critical for maintaining Mcl-1 mRNA levels while the PI3K pathway upregulated Mcl-1 protein translation (Schubert and Duronio 2001). To determine if SHIP-deficient TF-1 cells were more resistant to cell death than siNS-transfected cells because they were capable of maintaining higher levels of Mcl-1, we starved these two cell types in 10% AFCS for increasing lengths of time and monitored Mcl-1 levels by Western analysis. As can be seen in Figure 5.6, we not only compared the levels of Mcl-1 with increasing times in starvation medium but also that of poly (ADP-ribose) polymerase (PARP-1). This DNA repair enzyme is proteolytically cleaved by both caspases and calpains and thus serves as an early indicator of apoptosis in many cell types (Duriez and Shah 1997). Interestingly, comparing the SHIP-deficient and control TF-1 cells, in starvation medium versus optimal GM-CSF (last lanes of the left and right panels) it appears that PARP levels increase in the starved control cells, perhaps to try and deal with the DNA fragmentation occurring in these apoptosing cells. This increase is not occurring in the SHIP-deficient cells. Moreover, there is clear evidence of PARP cleavage in the control TF-1 cells, especially at 48 hr when there is a marked loss in viability of these cells (Fig. 5.4), but much less in the SHIP-deficient cells. More importantly, there appears to be a marked elevation in Mcl-1 protein levels in the SHIP-deficient TF-1 cells, both in the presence and absence of GM-CSF. As well, in the control cells there is clear evidence of a shorter form of this protein. Although the molecular mass of this protein is consistent with the alternate splice form of Mcl-1, Mcl-1 s (Bae, Leo et al. 2000) (which is pro-apoptotic unlike its full length counterpart) it is also consistent with a caspase-cleaved form of this protein that has very recently been shown to be formed during apoptosis (Herrant, Jacquel et al. 2004). In either case, this shorter Mcl-1 form appears with starvation of control but not SHIP-deficient TF-1 cells. 159 siNS siSHIP 10%dFCS 10%dFCS 0 12 24 36 48 GM 0 12 24 36 48 GM SHIP «esa -jjsaoi. •at. ~~***4ttmW PARP McML Mcl-1s jflagfe iM&f&*mmm::: wmam mmm< ^ * W F ^*"p' ™ <W lilBIL HS^^ (3-actin Figure 5.6: Pro-apoptotic proteins are reduced in siSHIP transfected TF-1 cells. TF-1 cells were transfected with either siNS or siSHIP and starved in 10% AFCS. Samples (5 x 105 viable cells) were taken every 12 hr and TCLs subjected to Western analysis with the indicated antibodies. Cells grown in GM-CSF (GM) were used as a control. 160 5.3 DISCUSSION The TF-1 cells used in this study are CD34+ cells derived from a patient with erythroleukemia. Their name refers to the fact that they are tri-factor dependent, (i.e., they can grow continuously in either GM-CSF or IL-3 and for a short term in Epo) (Kitamura, Tange et al. 1989). Several versions that have slightly different cytokine responses and biological activities have been derived and long term growth in IL-5 results in the upregulation (or selection) of TF-1 cells with higher IL-5Ra expression (Ettinger, Fong et al. 1997). Some sub-lines can also differentiate into macrophage-like cells following growth arrest or into erythroid-like cells with Epo (Kitamura et al. 1989) ( Relevant to this, the EpoR in TF-1 cells has been shown to be truncated (Winkelmann, Ward et al. 1995) and this truncation eliminates all intracellular tyrosines save the membrane proximal tyrosine (Y 3 4 3 ) and the motifs required for JAK2 association. Because of this truncation, signalling proteins are poorly recruited and Epo stimulation of TF-1 cells results in relatively weak tyrosine phosphorylations compared to those induced by IL-3 and GM-CSF. Also relevant to this is that the membrane receptors for IL-3, GM-CSF and IL-5 are all heterodimers and share a common Pc subunit and a unique a subunit, the latter conferring specificity of binding (Hayashida, Kitamura et al. 1990; Tavernier, Devos et al. 1991; Miyajima, Mui et al. 1993). The Pc subunit is capable of binding JAK2, and undergoes multiple tyrosine phosphorylations upon stimulation with GM-CSF, IL-3 or IL-5 (Silvennoinen, Witthuhn et al. 1993; Witthuhn, Quelle et al. 1993; Mui, Wakao et al. 1995). In this study, we demonstrate for the first time that a siRNA to SHIP can markedly reduce SHIP protein levels in human cells. Interestingly, we found that reducing SHIP levels did not result in a compensatory increase in the protein levels of either PTEN or SHIP2, both of which are expressed in TF-1 cells. Also of interest, we found that in the SHIP-deficient TF-1 cells the phosphorylation of Akt, GSK3P and ERK1/2 was not only more pronounced but peaked earlier. This was also true for STAT5 (data not shown). At least with GSK2 and ERK1/2 this also seemed to result in a more rapid deposphorylation, in keeping with other studies suggesting that a stronger stimulation results in a more acute response while a weaker one gives a lower but more prolonged response (Murphy, Smith et al. 2002). Intriguingly, while SHIP-deficient TF-1 cells survive and proliferate more readily that control TF-1 cells in the absence of cytokines or in the presence of very low cytokines, this is 161 not the case at high cytokine concentrations. Although the high levels tested may have no physiological significance, these results are reminiscent of Cheryl Helgason's results in Keith Humphries lab, when she monitored myeloid colonies from SHIP-/- and +/+ bone marrow in methylcellulose assays. At limiting cytokines, she saw far more colonies with the SHIP-/-bone marrow cells but at high cytokine concentrations, she saw significantly fewer. However, since this was somewhat unsettling at the time she plotted her results as a percent of maximal colonies and this effect was obscured (Helgason, Damen et al. 1998). One possibility to explain this inhibition of proliferation in SHIP-deficient cells at high cytokine concentrations is that the stimulatory signals are more short-lived when they are very potent or that a compensatory a negative feedback signal is stronger with higher stimulatory signals. Alternatively, in the absence of SHIP, less PI-3,4-P2 is generated and this has been shown to be a second messenger in its own right (Jones, Klinghoffer et al. 1999; Rameh and Cantley 1999; Scheid, Huber et al. 2002) and may limit long term Akt activation under some circumstances (Scheid, Huber et al. 2002). Our finding that SHIP-deficient TF-1 cells express higher Mcl-1 protein levels is interesting but perhaps not too surprising given that it has been shown to be positively regulated at the translational level by the PI3K pathway (Schubert and Duronio 2001). This has some ramifications as far as human disease is concerned since it has been shown that patients with AML express higher levels of Mcl-1 at relapse than at first presentation for the disease (Kaufmann, Karp et al. 1998). Like Bcl-2, Mcl-1 may play a role in resistance to chemotherapy. Related to this is has also been shown that patients with severe congenital neutropenia (SCN) have a mutant G-CSF receptor that is postulated to contribute to transformation into AML (Hunter and Avalos 2000). This mutant receptor lacks Y 7 6 4 which normally binds SHIP (and CIS, a SOCS member that acts as a negative regulator by blocking JAK2 signalling) (Hansen, Lindberg et al. 1999) and cells expressing this mutant are hypersensitive to G-CSF and give sustained cellular activation (Hunter, Jacob et al. 2004). Taken together, we have demonstrated, using siRNA technology, that SHIP plays an important role in human hematopoietic cells, most likely by restraining the PI3K pathway, as it does in murine hematopoietic cells. 162 Chapter 6 SUMMARY In Chapter 3 of this thesis, we explored the role that SHIP plays in erythropoiesis. Interestingly, although SHIP-/- mice are not overtly polycythemic (with erythrocytosis) as one might expect from the loss of this negative regulator of the PI3K pathway, we can estimate that the total number (/. e., spleen + bone marrow) of BFU-E and CFU-E are higher or at least similar to wild-type mice in the unstressed state. In spite of this, SHIP-/- mice are slightly anemic and display reticulocytosis, suggesting a higher erythropoietic rate and perhaps a shorter half-life for their late progenitors or RBCs. Related to this but not reported in Chapter 3, we also found that the while the total number of CFU-Mk (megakaryocyte progenitors) was reduced in the BM, the number of CFU-Mk progenitors in the spleen compensated for this loss. Despite this fact, others (Moody, Xu et al. 2004) and we have detected a mild thrombocytopenia in SHIP-/- compared to wild-type mice. As was the case for anemia, the severity of the thrombocytopenia is exacerbated in animals with a more severe (myeloproliferative) phenotype (Moody, Xu et al. 2004). This similarity in megakaryopoiesis and erythropoiesis is not too surprising, given that the megakaryocytic and erythroid progenitors are derived from a common progenitor, but it again suggests that something acting downstream of these progenitors is having a negative impact on the survival of later end cell precursors or the end cells themselves. This may be related to the large number of granulocytes or monocyte/macrophages in SHIP-/- mice, since, as mentioned earlier, the overproduction of IL-6 by the elevated numbers of macrophages in SHIP-/- mice negatively influences B cell proliferation and differentiation in these mice. Unlike platelets however, SHIP is not expressed in very late erythroblasts or mature RBCs. This fact means that if there is an intrinsic defect in SHIP-/- erythropoiesis, it must occur before or at the basophilic stage of erythroid differentiation. In addition, the more enhanced and more rapid secretion of growth-promoting factors (which act in concert with Epo) into the plasma of PHz-treated may also be attributed to the greater numbers of granulocytes and macrophages in the SHIP-deficient mice. In future studies it would be interesting to see the effect of IL-6 or of co-culturing peripheral blood white blood cells from SHIP+/+ and -/-163 mice with lineage-depleted (Lin") erythroid progenitors from SHIP+/+ and -/- mice and monitoring both the number of emerging benzidine positive cells and the number of CFU-E and smaller clusters in methylcellulose. However, our CFU-E and BFU-E studies described in Chapter 3 yielded colonies of approximately the same size and morphology, suggesting that, at least under defined culture conditions, SHIP+/+ and SHIP-/- erythroid progenitors undergo the same number of cell divisions during differentiation. As well it would be interesting to compare the life span of the RBCs in the SHIP+/+ and -/- mice. Our findings would predict a shorter half-life for SHIP-/- RBCs and serve to confirm an increased erythropoietic rate. Examination of the marrow and spleen for evidence of erythrophagocytosis as observed in Lyn-/- mice would further advance our understanding of the mechanisms of anemia in these mice. Finally, we are embarking on studies using in vitro differentiation of SHIP-/- ES cells and gene-silencing of SHIP in purified human erythroid progenitors to better understand the role of SHIP (if any) in the proliferation and differentiation of the erythroid lineage outside of the influence of a myeloproliferative environment. As also mentioned in Chapter 3, we found that SHIP-/- mice show an enhanced recovery from PHz-induced anemia. This is not surprising given the elevated numbers of BFU-E and CFU-E in the spleen of resting SHIP-/- mice. In addition, we found that, following PHz treatment, SHIP-/- plasma possesses higher erythroid growth-stimulating activity than SHIP+/+ plasma and this elevated activity may be due to an increased level of Epo. In future studies we would like to develop an Epo ELISA for murine Epo that would allow us to measure the Epo level in both resting and PHz-treated mice to determine if the higher growth-promoting activity in SHIP-/- plasma is due to higher levels of Epo itself or factors that synergize with Epo such as glucocorticoids. We have some preliminary evidence of enhanced growth of SHIP-/- erythroblasts in serum-free cultures containing Epo, SCF and dexamethasone (a glucocorticoid receptor agonist). These findings would confirm our model explaining enhanced stress erythropoiesis in the SHIP-/- mice. Of note, in our studies in Chapter 4, we did detect tyrosine phosphorylation of SHIP in response to stimulation with dexamethasone and other steroids of various classes (unpublished observations). Like the inducers described in Chapter 4, this steroid mediated phosphorylation of SHIP was dependent on Src-family member activity and steroid receptors are known to interact with 164 Src-family members (reviewed in (Shupnik 2004). Thus, a study of SHIP'S role in the regulation of steroid receptor/Src-family mediated signalling may be warranted. In Chapter 3, we found that the Epo responsiveness was similar in SHIP+/+ and -/-progenitors. Given this, one might expect the survival of the late erythroid progenitors to be very similar in SHIP+/+ and -/- mice (unless the presence of SHIP at earlier stages irreversibly impacts on the differentiation potential) and this supports the notion of extrinsic forces, perhaps from the excess numbers of monocytes/macrophages, being responsible for the reduced ability of SHIP-/- erythroblasts or RBCs to survive. In Chapter 4, using phosphospecific antibodies we developed for this study, we found that the two NPXYs within SHIP are the major sites of tyrosine phosphorylation in response to cytokines, growth factors, serpentine receptor ligands, immunoreceptor ligands and osmotic stress. However, we also discovered a previously unknown tyrosine phosphorylation site in SHIP (i.e., Y 8 6 7 in murine SHIP) and future studies will examine the biological role of this tyrosine by mutating it to a phenylalanine. We also found, using PP2 as well as cells from Lyn-/- mice, that the Src family is primarily responsible for SHIP's tyrosine phosphorylation. Moreover, consistent with this being a direct effect, we found SHIP and Lyn associate in BJAB B-cells and this association increases with BCR stimulation. Interestingly, this association appears to be mediated via a previously unidentified proline-rich motif just C-terminal of the SH2 domain of SHIP. How SHIP can increase its association (or the strength of its association) with Lyn after anti-IgM stimulation of BJAB cells in the presence of PP2 is intriguing and may be related to our finding that ERK 1/2 phosphorylation is relatively unaffected by inhibition of the Src family. Related to this, it is conceivable that the tyrosine kinase Syk can be activated in the absence of an upstream Src family member in these cells and this may allow several tyrosine phosphorylations, including that of a protein that is complexed with or becomes complexed with SHIP after stimulation and enhances the association of SHIP with Lyn. It would be interesting, in future studies, to explore the proteins complexed with SHIP before and after stimulation of BJAB cells in the presence and absence of PP2. Lastly, in Chapter 5, we investigated the role of SHIP in human hematopoietic cells using siRNA to knockdown SHIP in the erythroleukemic cell line, TF-1. We demonstrated nearly complete knockdown of SHIP protein in these cells and this resulted in a more rapid 165 phosphorylation of Akt, GSK3(3 and ERK1/2 in response to GM-CSF. As well, SHIP-deficient TF-1 cells survived better with low or no cytokines but, interestingly, displayed reduced proliferation at high cytokine concentrations. This reduced proliferation at high cytokine levels warrants further study and might be due to either the activation or upregulation of negative regulators or limiting amounts of PI-3,4-P2 (SHIP's product). With regard to the latter, we have previously reported that optimal Akt activation requires both PIP3 and PI-3,4-P2 (Scheid, Huber et al, 2002) and it would be interesting to examine the activation of Akt (via the phosphorylation of a downstream target like GSK3) at different concentrations of GM-CSF in SHIP knockdown and control TF-1 cells. We also found in Chapter 5 that the enhanced survival of SHIP-knockdown TF-1 cells was mediated, at least in part, by maintenance of the pro-survival Bcl-2 member, Mcl-1. We would now like to knockdown SHIP in normal human progenitors. However, preliminary studies using CD34+ cells from G-CSF mobilized peripheral blood and square wave electroporation resulted in only partial knockdown (approximately 50% by Western analysis) and variable biological effects. The lab, therefore, has embarked on a lentiviral approach using short hairpin RNA (shRNA) to human SHIP. 166 REFERENCES Abramson, S., Miler, RG., Phillips, RA (1978). "The identification in adult bone marrow of pluripotent and restricted stem cells of the myeloid and lymphoid systems." J Exp Med 145: 1567. Adachi, T., K. Pazdrak, et al. (1999). "The mapping of the Lyn kinase binding site of the common beta subunit of IL-3/granulocyte-macrophage colony-stimulating factor/IL-5 receptor." J Immunol 162(3): 1496-501. Agazie, Y., I. Ischenko, et al. (2002). "Concomitant activation of the PI3K-Akt and the Ras-ERK signaling pathways is essential for transformation by the V-SEA tyrosine kinase oncogene." Oncogene 21(5): 697-707. Agrawal, N. , P. V. Dasaradhi, et al. (2003). "RNA interference: biology, mechanism, and applications." Microbiol Mol Biol Rev 67(4): 657-85. Akahane, K., A. Tojo, et al. (1987). "Pure erythropoietic colony and burst formations in serum-free culture and their enhancement by insulin-like growth factor I." Exp Hematol 15(7): 797-802. Akashi, K., D. Traver, et al. (1999). "Lymphoid development from hematopoietic stem cells." Int J Hematol 69(4): 217-26. , Akashi, K., D. Traver, et al. (2000). "A clonogenic common myeloid progenitor that gives rise to all myeloid lineages." Nature 404(6774): 193-7. Alberts, B., Bray, D., Lewis, J., Raff, M. , Roberts, K. and Watson, JD. (1994). Molecular Biology of the Cell. New York and London, Garland Publishing. Alexandropoulos, K., G. Cheng, et al. (1995). "Proline-rich sequences that bind to Src homology 3 domains with individual specificities." Proc Natl Acad Sci U S A 92(8): 3110-4. Aman, M. J., A. C. Tosello-Trampont, et al. (2001). "Fc gamma RIIBl/SHIP-mediated inhibitory signaling in B cells involves lipid rafts." J Biol Chem 276(49): 46371-8. Aman, M. J., S. F. Walk, et al. (2000). "Essential role for the C-terminal noncatalytic region of SHIP in FcgammaRIIBl -mediated inhibitory signaling." Mol Cell Biol 20(10): 3576-89. Arai, A., E. Kanda, et al. (2001). "CrkL is recruited through its SH2 domain to the erythropoietin receptor and plays a role in Lyn-mediated receptor signaling." J Biol Chem 276(35): 33282-90. Ayala, G , T. Thompson, et al. (2004). High levels of phosphorylated form of Akt-1 in prostate cancer and non-neoplastic prostate tissues are strong predictors of biochemical recurrence. Clin Cancer Res. 10: 6572-8. Bae, J., C. P. Leo, et al. (2000). "MCL-1S, a splicing variant of the antiapoptotic BCL-2 family member MCL-1, encodes a proapoptotic protein possessing only the BH3 domain." J Biol Chem 275(33): 25255-61. Baran, C. P., S. Tridandapani, et al. (2003). "The inositol 5'-phosphatase SHIP-1 and the Src kinase Lyn negatively regulate macrophage colony-stimulating factor-induced Akt activity." J Biol Chem 278(40): 38628-36. Bauer, A., F. Tronche, et al. (1999). "The glucocorticoid receptor is required for stress erythropoiesis." Genes Dev 13(22): 2996-3002. 167 Bavelloni, A., I. Faenza, et al. (2000). "Inhibition of phosphoinositide 3-kinase impairs pre-commitment cell cycle traverse and prevents differentiation in erythroleukaemia cells." Cell Death Differ 7(1): 112-7. Bazan, J. F. (1990). "Structural design and molecular evolution of a cytokine receptor superfamily." Proc Natl Acad Sci U S A 87(18): 6934-8. Becker, A. J., E. A. McCulloch, et al. (1963). "Cytological demonstration of the clonal nature of spleen colonies derived from transplanted mouse marrow cells." Nature 197: 452-4. Begley, C. G., P. D. Apian, et al. (1989). "The gene SCL is expressed during early hematopoiesis and encodes a differentiation-related DNA-binding motif." Proc Natl Acad Sci U S A 86(24): 10128-32. Beru, N , J. McDonald, et al. (1986). "Expression of the erythropoietin gene." Mol Cell Biol 6(7): 2571-5. Blobel, G. A., T. Nakajima, et al. (1998). "CREB-binding protein cooperates with transcription factor GATA-1 and is required for erythroid differentiation." Proc Natl Acad Sci U S A 95(5): 2061-6. Boggs, D. R. (1984). "The total marrow mass of the mouse: a simplified method of measurement." Am J Hematol 16(3): 277-86. Bone, H. and M. J. Welham (2000). "She associates with the IL-3 receptor beta subunit, SHIP and Gab2 following IL-3 stimulation. Contribution of She PTB and SH2 domains." Cell Signal 12(3): 183-94. Boudot, C , E. Dasse, et al. (2003). "Involvement of the Src kinase Lyn in phospholipase C-gamma 2 phosphorylation and phosphatidylinositol 3-kinase activation in Epo signalling." Biochem Biophys Res Commun 300(2): 437-42. Bouscary, D., F. Pene, et al. (2003). "Critical role for PI 3-kinase in the control of erythropoietin-induced erythroid progenitor proliferation." Blood 101(9): 3436-43. Bradbury, D., Y. M . Zhu, et al. (1994). "Regulation of Bcl-2 expression and apoptosis in acute myeloblasts leukaemia cells by granulocyte-macrophage colony-stimulating factor." Leukemia 8(5): 786-91. Brauweiler, A., I. Tamir, et al. (2000). "Differential regulation of B cell development, activation, and death by the src homology 2 domain-containing 5' inositol phosphatase (SHIP)." J Exp Med 191(9): 1545-54. Brines, M. , G. Grasso, et al. (2004). Erythropoietin mediates tissue protection through an erythropoietin and common beta-subunit heteroreceptor. Proc Natl Acad Sci U S A . 101: 14907-12. Broudy, V. C , N . Lin, et al. (1991). "Erythropoietin receptor characteristics on primary human erythroid cells." Blood 77(12): 2583-90. Brown, D. A. and E. London (2000). "Structure and function of sphingolipid- and cholesterol-rich membrane rafts." J Biol Chem 275(23): 17221-4. Brown, G., C. M . Bunce, et al. (1985). "Sequential determination of lineage potentials during haemopoiesis." Br J Cancer 52(5): 681-6. Brown, M . T. and J. A. Cooper (1996). "Regulation, substrates and functions of src." Biochim Biophys Acta 1287(2-3): 121-49. Broxmeyer, H. E. (2001). "Regulation of hematopoiesis by chemokine family members." Int J Hematol 74(1): 9-17. 168 Bull B. S. and Breton-Gorius, J. (1995). Chapter 32: Morphology of the erythron. New York, McGraw-Hill, Inc. Cacalano, N . A., D. Sanden, et al. (2001). "Tyrosine-phosphorylated SOCS-3 inhibits STAT activation but binds to pi20 RasGAP and activates Ras." Nat Cell Biol 3(5): 460-5. Cantley, L. C. and B. G. Neel (1999). "New insights into tumor suppression: PTEN suppresses tumor formation by restraining the phosphoinositide 3-kinase/AKT pathway." Proc Natl Acad Sci U S A 96(8): 4240-5. Cantor, A. B. and S. H. Orkin (2001). "Hematopoietic development: a balancing act." Curr Opin Genet Dev 11(5): 513-9. Cantor, A. B. and S. H. Orkin (2002). "Transcriptional regulation of erythropoiesis: an affair involving multiple partners." Oncogene 21(21): 3368-76. Cardone, M . H., N . Roy, et al. (1998). "Regulation of cell death protease caspase-9 by phosphorylation." Science 282(5392): 1318-21. Carnot P, D. C. (1906). " Sur l'activite hemopoietique du serum au cours de la regeneration du sang." Comte Rendus Hepdomadaires des Seances de l'academie de Sciences 143(384). Carotta, S., S. Pilat, et al. (2004). "Directed differentiation and mass cultivation of pure erythroid progenitors from mouse embryonic stem cells." Blood 104(6): 1873-80. Carrano, A. C , E. Eytan, et al. (1999). "SKP2 is required for ubiquitin-mediated degradation of the CDK inhibitor p27." Nat Cell Biol 1(4): 193-9. Chan, B., A. Lanyi, et al. (2003). "SAP couples Fyn to SLAM immune receptors." Nat Cell B i d 5(2): 155-60. Chan, V. W., F. Meng, et al. (1997). "Characterization of the B lymphocyte populations in Lyn-deficient mice and the role of Lyn in signal initiation and down-regulation." Immunity 7(1): 69-81. Chen, Q., J. T. Cheng, et al. (1990). "The tal gene undergoes chromosome translocation in T cell leukemia and potentially encodes a helix-loop-helix protein." Embo J 9(2): 415-24. Cheng, P. C , B. K. Brown, et al. (2001). "Translocation of the B cell antigen receptor into lipid rafts reveals a novel step in signaling." J Immunol 166(6): 3693-701. Christopherson, K., 2nd and R. Hromas (2001). "Chemokine regulation of normal and pathologic immune responses." Stem Cells 19(5): 388-96. Clapp, D. W., B. Freie, et al. (1995). "Molecular evidence that in situ-transduced fetal liver hematopoietic stem/progenitor cells give rise to medullary hematopoiesis in adult rats." Blood 86(6): 2113-22. Clement, S., U. Krause, et al. (2001). "The lipid phosphatase SHIP2 controls insulin sensitivity." Nature 409(6816): 92-7. Clement, S., U. Krause, et al. (2004). "corrigendum: The lipid phosphatase SHIP2 controls insulin sensitivity." Nature 431(7010): 878. Coggeshall, K. M . (1998). "Inhibitory signaling by B cell Fc gamma Rllb." Curr Opin Immunol 10(3): 306-12. Cohney, S. J., D. Sanden, et al. (1999). "SOCS-3 is tyrosine phosphorylated in response to interleukin-2 and suppresses STAT5 phosphorylation and lymphocyte proliferation." Mol Cell Biol 19(7): 4980-8. Coleman, T. and M . Brines (2004). Science review: Recombinant human erythropoietin in critical illness: a role beyond anemia? 169 Constantinescu, S. N. , S. Ghaffari, et al. (1999). "The Erythropoietin Receptor: Structure, Activation and Intracellular Signal Transduction." Trends Endocrinol Metab 10(1): 18-23. Cox, D., B. M . Dale, et al. (2001). "A regulatory role for Src homology 2 domain-containing inositol 5'-phosphatase (SHIP) in phagocytosis mediated by Fc gamma receptors and complement receptor 3 (alpha(M)beta(2); CD1 lb/CD18)." J Exp Med 193(1): 61-71. Crispino, J. D., M . B. Lodish, et al. (1999). "Use of altered specificity mutants to probe a specific protein-protein interaction in differentiation: the GATA-1 :FOG complex." Mol Cell 3(2): 219-28. Crowley, M . T., S. L. Harmer, et al. (1996). "Activation-induced association of a 145-kDa tyrosine-phosphorylated protein with She and Syk in B lymphocytes and macrophages." J Biol Chem 271(2): 1145-52. Dai, C. H., J. O. Price, et al. (1998). "Fas ligand is present in human erythroid colony-forming cells and interacts with Fas induced by interferon gamma to produce erythroid cell apoptosis." Blood 91(4): 1235-42. Daigle, I., S. Yousefi, et al. (2002). "Death receptors bind SHP-1 and block cytokine-induced anti-apoptotic signaling in neutrophils." Nat Med 8(1): 61-7. Dainiak, N . and S. Kreczko (1985). "Interactions of insulin, insulinlike growth factor II, and platelet-derived growth factor in erythropoietic culture." J Clin Invest 76(3): 1237-42. Dal Porto, J. M. , S. B. Gauld, et al. (2004). "B cell antigen receptor signaling 101." Mol Immunol 41(6-7): 599-613. Dame, C. and S. E. Juul (2000). "The switch from fetal to adult erythropoiesis." Clin Perinatol 27(3): 507-26. Damen, J. E., J. Krosl, et al. (1998). "The hyperresponsiveness of cells expressing truncated erythropoietin receptors is contingent on insulin-like growth factor-1 in fetal calf serum." Blood 92(2): 425-33. Damen, J. E., L. Liu, et al. (1993). "Erythropoietin stimulates the tyrosine phosphorylation of She and its association with Grb2 and a 145-Kd tyrosine phosphorylated protein." Blood 82(8): 2296-303. Damen, J. E., L. Liu, et al. (1996). "The 145-kDa protein induced to associate with She by multiple cytokines is an inositol tetraphosphate and phosphatidylinositol 3,4,5-triphosphate 5-phosphatase." Proc Natl Acad Sci U S A 93(4): 1689-93. Damen, J. E., M . D. Ware, et al. (2001). "SHIP's C-terminus is essential for its hydrolysis of PIP3 and inhibition of mast cell degranulation." Blood 97(5): 1343-51. D'Andrea, A. D., H. F. Lodish, et al. (1989). "Expression cloning of the murine erythropoietin receptor." Cell 57(2): 277-85. D'Andrea, A. D. and L. I. Zon (1990). "Erythropoietin receptor. Subunit structure and activation." J Clin Invest 86(3): 681-7. Danilkovitch-Miagkova, A. and E. J. Leonard (2001). "Cross-talk between RON receptor tyrosine kinase and other transmembrane receptors." Histol Histopathol 16(2): 623-31. Davidson, D., M . Bakinowski, et al. (2003). "Phosphorylation-dependent regulation of T-cell activation by PAG/Cbp, a lipid raft-associated transmembrane adaptor." Mol Cell Biol 23(6): 2017-28. 170 de la Chapelle, A., A. L. Traskelin, et al. (1993). "Truncated erythropoietin receptor causes dominantly inherited benign human erythrocytosis." Proc Natl Acad Sci U S A 90(10): 4495-9. De Maria, R., A. Zeuner, et al. (1999). "Negative regulation of erythropoiesis by caspase-mediated cleavage of GATA-1." Nature 401(6752): 489-93. DiNitto, J. P., T. C. Cronin, et al. (2003). Membrane recognition and targeting by lipid-binding domains. Sci STKE. Divoky, V., Z. Liu, et al. (2001). "Mouse model of congenital polycythemia: Homologous replacement of murine gene by mutant human erythropoietin receptor gene." Proc Natl Acad Sci U S A 98(3): 986-91. Dolznig, H., B. Habermann, et al. (2002). "Apoptosis protection by the Epo target Bcl-X(L) allows factor-independent differentiation of primary erythroblasts." Curr Biol 12(13): 1076-85. Domen, J. (2001). "The role of apoptosis in regulating hematopoietic stem cell numbers." Apoptosis 6(4): 239-52. Domen, J., S. H. Cheshier, et al. (2000). "The role of apoptosis in the regulation of hematopoietic stem cells: Overexpression of Bcl-2 increases both their number and repopulation potential." J Exp Med 191(2): 253-64. Domen, J. and I. L. Weissman (1999). "Self-renewal, differentiation or death: regulation and manipulation of hematopoietic stem cell fate." Mol Med Today 5(5): 201-8. Downward, J. (2004). "PI 3-kinase, Akt and cell survival." Semin Cell Dev Biol 15(2): 177-82. Duriez, P. J. and G. M. Shah (1997). "Cleavage of poly(ADP-ribose) polymerase: a sensitive parameter to study cell death." Biochem Cell Biol 75(4): 337-49. Dzierzak, E. and A. Medvinsky (1995). "Mouse embryonic hematopoiesis." Trends Genet 11(9): 359-66. Eaves, C. J. (2003). "Manipulating hematopoietic stem cell amplification with Wnt." Nat Immunol 4(6): 511-2. Endo, T., A. Odb, et al. (2001). "Stem cell factor protects c-kit+ human primary erythroid cells from apoptosis." Exp Hematol 29(7): 833-41. Endo, T. A., M . Masuhara, et al. (1997). "A new protein containing an SH2 domain that inhibits JAK kinases." Nature 387(6636): 921-4. Erslev, A. J. and A. Besarab (1995). "The rate and control of baseline red cell production in hematologically stable patients with uremia." J Lab Clin Med 126(3): 283-6. Erslev, A. J. a. B., E. (1995). Chapter 39: Production and destruction of erythrocytes. New York, McGraw-Hill, Inc. Eschbach, J. W., J. C. Egrie, et al. (1987). "Correction of the anemia of end-stage renal disease with recombinant human erythropoietin. Results of a combined phase I and II clinical trial." N Engl J Med 316(2): 73-8. Eschbach, J. W., M . R. Kelly, et al. (1989). "Treatment of the anemia of progressive renal failure with recombinant human erythropoietin." N Engl J Med 321(3): 158-63. Ettinger, S., D. Fong, et al. (1997). "Lack of correlation between growth of TF-1 cells and tyrosine phosphorylation signals in response to IL-3, IL-5 and GM-CSF." Cytokine 9(9): 650-9. Evans, T. and G. Felsenfeld (1989). "The erythroid-specific transcription factor Eryfl: a new finger protein." CeU 58(5): 877-85. 171 Feng, X., K. P. Becker, et al. (2000). "Regulation of receptor-mediated protein kinase C membrane trafficking by autophosphorylation." J Biol Chem 275(22): 17024-34. Finger, L. R., J. Kagan, et al. (1989). "Involvement of the TCL5 gene on human chromosome 1 in T-cell leukemia and melanoma." Proc Natl Acad Sci U S A 86(13): 5039-43. Finkelstein, L. D., P. A. Ney, et al. (2002). "Sf-Stk kinase activity and the Grb2 binding site are required for Epo-independent growth of primary erythroblasts infected with Friend virus." Oncogene 21(22): 3562-70. Fisher, A. G. (2002). "Cellular identity and lineage choice." Nat Rev Immunol 2(12): 977-82. Follenzi, A., S. Bakovic, et al. (2000). "Cross-talk between the proto-oncogenes Met and Ron." Oncogene 19(27): 3041-9. Ford, C. E., J. L. Hamerton, et al. (1956). "Cytological identification of radiation-chimaeras." Nature 177(4506): 452-4. Fox, J. A., K. Ung, et al. (2002). "Disruption of a single Pten allele augments the chemotactic response of B lymphocytes to stromal cell-derived factor-1." J Immunol 169(1): 49-54. Fraser, J. K., A. S. Tan, et al. (1989). "Expression of specific high-affinity binding sites for erythropoietin on rat and mouse megakaryocytes." Exp Hematol 17(1): 10-6. Fujita, E., A. Jinbo, et al. (1999). "Akt phosphorylation site found in human caspase-9 is absent in mouse caspase-9." Biochem Biophys Res Commun 264(2): 550-5. Galandrini, R., I. Tassi, et al. (2002). "SH2-containing inositol phosphatase (SHIP-1) transiently translocates to raft domains and modulates CD16-mediated cytotoxicity in human NK cells." Blood 100(13): 4581-9. Galandrini, R., I. Tassi, et al. (2001). "The adaptor protein she is involved in the negative regulation ofNK cell-mediated cytotoxicity." Eur J Immunol 31(7): 2016-25. Gardai, S., B. B. Whitlock, et al. (2002). "Activation of SHIP by NADPH oxidase-stimulated Lyn leads to enhanced apoptosis in neutrophils." J Biol Chem 277(7): 5236-46. Geier, S. J., P. A. Algate, et al. (1997). "The human SHIP gene is differentially expressed in cell lineages of the bone marrow and blood." Blood 89(6): 1876-85. Gibellini, D., A. Bassini, et al. (2000). "Stroma-derived factor 1 alpha induces a selective inhibition of human erythroid development via the functional upregulation of Fas/CD95 ligand." Br J Haematol 111(2): 432-40. Gimm, O., T. Attie-Bitach, et al. (2000). "Expression of the PTEN tumour suppressor protein during human development." Hum Mol Genet 9(11): 1633-9. Giuriato, S., S. Bodin, et al. (2000). "pp60c-src associates with the SH2-containing inositol-5-phosphatase SHIP1 and is involved in its tyrosine phosphorylation downstream of alphallbbeta3 integrin in human platelets." Biochem J 348 Pt 1: 107-12. Giuriato, S., B. Payrastre, et al. (1997). "Tyrosine phosphorylation and relocation of SHIP are integrin-mediated in thrombin-stimulated human blood platelets." J Biol Chem 272(43): 26857-63. Giuriato, S., X. Pesesse, et al. (2003). "SH2-containing inositol 5-phosphatases 1 and 2 in blood platelets: their interactions and roles in the control of phosphatidylinositol 3,4,5-trisphosphate levels." Biochem J 376(Pt 1): 199-207. Gobert, S., F. Porteu, et al. (1995). "Tyrosine phosphorylation of the erythropoietin receptor: role for differentiation and mitogenic signal transduction." Blood 86(2): 598-606. Gold, M. R. (2002). "To make antibodies or not: signaling by the B-cell antigen receptor." Trends Pharmacol Sci 23(7): 316-24. 172 Goldfarb, A. N. , D. Wong, et al. (2001). "Induction of megakaryocytic differentiation in primary human erythroblasts: a physiological basis for leukemic lineage plasticity." Am J Pathol 158(4): 1191-8. Graf, T. (2002). "Differentiation plasticity of hematopoietic cells." Blood 99(9): 3089-101. Gregoli, P. A. and M. C. Bondurant (1997). "The roles of Bcl-X(L) and apopain in the control of erythropoiesis by erythropoietin." Blood 90(2): 630-40. Haar, J. L. and G. A. Ackerman (1971). "A phase and electron microscopic study of vasculogenesis and erythropoiesis in the yolk sac of the mouse." Anat Rec 170(2): 199-223. Hansen, J. A., K. Lindberg, et al. (1999). "Mechanism of inhibition of growth hormone receptor signaling by suppressor of cytokine signaling proteins." Mol Endocrinol 13(11): 1832-43. Hao, Q. L., F. T. Thiemann, et al. (1996). "Extended long-term culture reveals a highly quiescent and primitive human hematopoietic progenitor population." Blood 88(9): 3306-13. Hara, H. and M. Ogawa (1978). "Murine hemopoietic colonies in culture containing normoblasts, macrophages, and megakaryocytes." Am J Hematol 4(1): 23-34. Harder, K. W., C. Quilici, et al. (2004). Perturbed myelo/erythropoiesis in Lyn-deficient mice is similar to that in mice lacking the inhibitory phosphatases SHP-1 and SHIP-1. Blood. Harder, K. W., C. Quilici, et al. (2004). "Perturbed myelo/erythropoiesis in Lyn-deficient mice is similar to that in mice lacking the inhibitory phosphatases SHP-1 and SHIP-1." Blood 104(13): 3901-10. Haseyama, Y., K. Sawada, et al. (1999). "Phosphatidylinositol 3-kinase is involved in the protection of primary cultured human erythroid precursor cells from apoptosis." Blood 94(5): 1568-77. Hayashida, K., T. Kitamura, et al. (1990). "Molecular cloning of a second subunit of the receptor for human granulocyte-macrophage colony-stimulating factor (GM-CSF): reconstitution of a high-affinity GM-CSF receptor." Proc Natl Acad Sci U S A 87(24): 9655-9. Helgason, C. D., J. E. Damen, et al. (1998). "Targeted disruption of SHIP leads to hemopoietic perturbations, lung pathology, and a shortened life span." Genes Dev 12(11): 1610-20. Helgason, C. D., C. P. Kalberer, et al. (2000). "A dual role for Src homology 2 domain-containing inositol-5-phosphatase (SHIP) in immunity: aberrant development and enhanced function of b lymphocytes in ship -/- mice." J Exp Med 191(5): 781-94. Hernandez-Hansen, V., G. A. Mackay, et al. (2004). "The Src kinase Lyn is a negative regulator of mast cell proliferation." J Leukoc Biol 75(1): 143-51. Herrant, M. , A. Jacquel, et al. (2004). "Cleavage of Mcl-1 by caspases impaired its ability to counteract Bim-induced apoptosis." Oncogene 23(47): 7863-73. Hibbs, M. L., K. W. Harder, et al. (2002). "Sustained activation of Lyn tyrosine kinase in vivo leads to autoimmunity." J Exp Med 196(12): 1593-604. Hibbs, M . L., D. M. Tarlinton, et al. (1995). "Multiple defects in the immune system of Lyn-deficient mice, culminating in autoimmune disease." Cell 83(2): 301-11. Hilton, D. J., R. T. Richardson, et al. (1998). "Twenty proteins containing a C-terminal SOCS box form five structural classes." Proc Natl Acad Sci U S A 95(1): 114-9. 173 Hirayama, F., J. P. Shih, et al. (1992). "Clonal proliferation of murine lymphohemopoietic progenitors in culture." Proc Natl Acad Sci U S A 89(13): 5907-11. Hodges, V. M. , P. C. Winter, et al. (1999). "Erythroblasts from friend virus infected- and phenylhydrazine-treated mice accurately model erythroid differentiation." Br J Haematol 106(2): 325-34. Holdorf, A. D., K. H. Lee, et al. (2002). "Regulation of Lck activity by CD4 and CD28 in the immunological synapse." Nat Immunol 3(3): 259-64. Holyoake, T. L., F. E. Nicolini, et al. (1999). "Functional differences between transplantable human hematopoietic stem cells from fetal liver, cord blood, and adult marrow." Exp Hematol 27(9): 1418-27. Horejsi, V. (2003). "The roles of membrane microdomains (rafts) in T cell activation." Immunol Rev 191: 148-64. Hromas, R., A. Orazi, et al. (1993). "Hematopoietic lineage- and stage-restricted expression of the ETS oncogene family member PU.l ." Blood 82(10): 2998-3004. Huang, H. M. , C. J. Huang, et al. (2000). "Mcl-1 is a common target of stem cell factor and interleukin-5 for apoptosis prevention activity via MEK/MAPK and PI-3K/Akt pathways." Blood 96(5): 1764-71. Huber, M. , C. D. Helgason, et al. (1999). "The role of SHIP in growth factor induced signalling." Prog Biophys Mol Biol 71(3-4): 423-34. Huber, M. , C. D. Helgason, et al. (1998). "Targeted disruption of SHIP leads to Steel factor-induced degranulation of mast cells." Embo J 17(24): 7311-9. Huber, M. , J. Kalesnikoff, et al. (2002). "The role of SHIP in mast cell degranulation and IgE-induced mast cell survival." Immunol Lett 82(1-2): 17-21. Hunter, M. G. and B. R. Avalos (2000). "Granulocyte colony-stimulating factor receptor mutations in severe congenital neutropenia transforming to acute myelogenous leukemia confer resistance to apoptosis and enhance cell survival." Blood 95(6): 2132-7. Hunter, M . G., A. Jacob, et al. (2004). "Loss of SHIP and CIS recruitment to the granulocyte colony-stimulating factor receptor contribute to hyperproliferative responses in severe congenital neutropenia/acute myelogenous leukemia." J Immunol 173(8): 5036-45. Ihle, J. N . (1994). "The Janus kinase family and signaling through members of the cytokine receptor superfamily." Proc Soc Exp Biol Med 206(3): 268-72. Ihle, J. N . (2000). "The challenges of translating knockout phenotypes into gene function." Cell 102(2): 131-4. Ikuta, K., Weissman, IL (1993). "T lymphocyte development from fetal hematopoietic stem cells." Sem. Dev. Biol 4: 371-378. Ilangumaran, S. and R. Rottapel (2003). "Regulation of cytokine receptor signaling by SOCS1." Immunol Rev 192: 196-211. Ingham, R. J., H. Okada, et al. (1999). "Tyrosine phosphorylation of she in response to B cell antigen receptor engagement depends on the SHIP inositol phosphatase." J Immunol 163(11): 5891-5. Ingley, E., D. J. McCarthy, et al. (2004). "Lyn deficiency reduces GATA-1, EKLF and STAT5, and induces extramedullary stress erythropoiesis." Oncogene. Irie-Sasaki, J., T. Sasaki, et al. (2003). "CD45 regulated signaling pathways." Curr Top Med Chem 3(7): 783-96. 174 Iscove, N . N. , L. J. Guilbert, et al. (1980). "Complete replacement of serum in primary cultures of erythropoietin-dependent red cell precursors (CFU-E) by albumin, transferrin, iron, unsaturated fatty acid, lecithin and cholesterol." Exp Cell Res , 126(1): 121-6. Ishibashi, T., J. A. Koziol, et al. (1987). "Human recombinant erythropoietin promotes differentiation of murine megakaryocytes in vitro." J Clin Invest 79(1): 286-9. Jacobs, K., C. Shoemaker, et al. (1985). "Isolation and characterization of genomic and cDNA clones of human erythropoietin." Nature 313(6005): 806-10. Jacobson, L. O., E. Goldwasser, et al. (1957). "Role of the kidney in erythropoiesis." Nature 179(4560): 633-4. Jacobson, L. O., E. K. Marks, et al. (1949). Effect of spleen protection on mortality following x-irradiation. J Lab Clin Med. 34: 1538-1543. Jacobson, L. O., E. L. Simmons, et al. (1951). "Recovery from radiation injury." Science 113(2940): 510-11. Jaken, S. (1996). "Protein kinase C isozymes and substrates." Curr Opin Cell Biol 8(2): 168-73. Janes, P. W., S. C. Ley, et al. (2000). "The role of lipid rafts in T cell antigen receptor (TCR) signalling." Semin Immunol 12(1): 23-34. Jegalian, A. G. and H. Wu (2002). "Differential roles of SOCS family members in EpoR signal transduction." J Interferon Cytokine Res 22(8): 853-60. Jiang, X., M . Stuible, et al. (2003). "Evidence for a positive role of SHIP in the BCR-ABL-mediated transformation of primitive murine hematopoietic cells and in human chronic myeloid leukemia." Blood 102(8): 2976-84. Jiao, H., K. Berrada, et al. (1996). "Direct association with and dephosphorylation of Jak2 kinase by the SH2-domain-containing protein tyrosine phosphatase SHP-1." Mol Cell Biol 16(12): 6985-92. Jones, S. M. , R. Klinghoffer, et al. (1999). "PDGF induces an early and a late wave of PI 3-kinase activity, and only the late wave is required for progression through G l . " Curr Biol 9(10): 512-21. Jones, S. S., A. D. D'Andrea, et al. (1990). "Human erythropoietin receptor: cloning, expression, and biologic characterization." Blood 76(1): 31-5. Kalesnikoff, J., N . Baur, et al. (2002). "SHIP negatively regulates IgE + antigen-induced IL-6 production in mast cells by inhibiting NF-kappa B activity." J Immunol 168(9): 4737-46. Kalesnikoff, J., V. Lam, et al. (2002). "SHIP represses mast cell activation and reveals that IgE alone triggers signaling pathways which enhance normal mast cell survival." Mol Immunol 38(16-18): 1201-6. Kalesnikoff, J., L. M . Sly, et al. (2003). "The role of SHIP in cytokine-induced signaling." Rev Physiol Biochem Pharmacol 149: 87-103. Kamura, T., S. Sato, et al. (1998). "The Elongin BC complex interacts with the conserved SOCS-box motif present in members of the SOCS, ras, WD-40 repeat, and ankyrin repeat families." Genes Dev 12(24): 3872-81. Kaufmann, S. H., J. E. Karp, et al. (1998). "Elevated expression of the apoptotic regulator Mcl-1 at the time of leukemic relapse." Blood 91(3): 991-1000. Kaushansky, K., S. Lok, et al. (1994). "Promotion of megakaryocyte progenitor expansion and differentiation by the c-Mpl ligand thrombopoietin." Nature 369(6481): 568-71. 175 Kay, B. K., M . P. Williamson, et al. (2000). "The importance of being proline: the interaction of proline-rich motifs in signaling proteins with their cognate domains." Faseb J 14(2): 231-41. Kean, L. S., L. E. Brown, et al. (2002). "Comparison of mechanisms of anemia in mice with sickle cell disease and beta-thalassemia: peripheral destruction, ineffective erythropoiesis, and phospholipid scramblase-mediated phosphatidylserine exposure." Exp Hematol 30(5): 394-402. Kelley, L. L., W. F. Green, et al. (1994). "Apoptosis in erythroid progenitors deprived of erythropoietin occurs during the G l and S phases of the cell cycle without growth arrest or stabilization of wild-type p53." Mol Cell Biol 14(6): 4183-92. Kholodenko, B. N. , O. V. Demin, et al. (1999). Quantification of short term signaling by the epidermal growth factor receptor. J Biol Chem. 274: 30169-81. Kimura, T., H. Sakamoto, et al. (1997). "The negative signaling molecule SH2 domain-containing inositol-polyphosphate 5-phosphatase (SHIP) binds to the tyrosine-phosphorylated beta subunit of the high affinity IgE receptor." J Biol Chem 272(21): 13991-6. Kina, T., K. Ikuta, et al. (2000). "The monoclonal antibody TER-119 recognizes a molecule associated with glycophorin A and specifically marks the late stages of murine erythroid lineage." Br J Haematol 109(2): 280-7. Kinashi, T. and T. A. Springer (1994). "Steel factor and c-kit regulate cell-matrix adhesion." Blood 83(4): 1033-8. King, A. G., M . Kondo, et al. (2002). "Lineage infidelity in myeloid cells with TCR gene rearrangement: a latent developmental potential of proT cells revealed by ectopic cytokine receptor signaling." Proc Natl Acad Sci U S A 99(7): 4508-13. Kishimoto, T., T. Taga, et al. (1994). "Cytokine signal transduction." Cell 76(2): 253-62. Kitamura, T., T. Tange, et al. (1989). "Establishment and characterization of a unique human cell line that proliferates dependently on GM-CSF, IL-3, or erythropoietin." J Cell Physiol 140(2): 323-34. Klingmuller, U., U. Lorenz, et al. (1995). "Specific recruitment of SH-PTP1 to the erythropoietin receptor causes inactivation of JAK2 and termination of proliferative signals." Cell 80(5): 729-38. Klingmuller, U., H. Wu, et al. (1997). "Identification of a novel pathway important for proliferation and differentiation of primary erythroid progenitors." Proc Natl Acad Sci U S A 94(7): 3016-21. Komatsu, N. , J. W. Adamson, et al. (1992). "Erythropoietin rapidly induces tyrosine phosphorylation in the human erythropoietin-dependent cell line, UT-7." Blood 80(1): 53-9. Kondo, M. , A. J. Wagers, et al. (2003). "Biology of hematopoietic stem cells and progenitors: implications for clinical application." Annu Rev Immunol 21: 759-806. Kondo, M. , I. L. Weissman, et al. (1997). "Identification of clonogenic common lymphoid progenitors in mouse bone marrow." Cell 91(5): 661-72. Koury, M . J. and M . C. Bondurant (1990). "Erythropoietin retards DNA breakdown and prevents programmed death in erythroid progenitor cells." Science 248(4953): 378-81. Koury, M. J., S. T. Sawyer, et al. (2002). "New insights into erythropoiesis." Curr Opin Hematol 9(2): 93-100. 176 Krahn, A. K., K. Ma, et al. (2004). "Two distinct waves of membrane-proximal B cell antigen receptor signaling differentially regulated by Src homology 2-containing inositol polyphosphate 5-phosphatase." J Immunol 172(1): 331-9. Krantz, S. B. (1991). "Erythropoietin." Blood 77(3): 419-34. Krosl, G., G. He, et al. (1998). "Transcription factor SCL is required for c-kit expression and c-Kit function in hemopoietic cells." J Exp Med 188(3): 439-50. Krystal, G. (1983). "A simple microassay for erythropoietin based on 3H-thymidine incorporation into spleen cells from phenylhydrazine treated mice." Exp Hematol 11(7): 649-60. Krystal, G. (2000). "Lipid phosphatases in the immune system." Semin Immunol 12(4): 397-403. Kurosaki, T. (1999). "Genetic analysis of B cell antigen receptor signaling." Annu Rev Immunol 17: 555-92. Kwik, J., S. Boyle, et al. (2003). "Membrane cholesterol, lateral mobility, and the phosphatidylinositol 4,5-bisphosphate-dependent organization of cell actin." Proc Natl Acad Sci U S A 100(24): 13964-9. Lam, K. P., R. Kuhn, et al. (1997). "In vivo ablation of surface immunoglobulin on mature B cells by inducible gene targeting results in rapid cell death." Cell 90(6): 1073-83. Lam, V., J. Kalesnikoff, et al. (2003). "IgE alone stimulates mast cell adhesion to fibronectin via pathways similar to those used by IgE + antigen but distinct from those used by Steel factor." Blood 102(4): 1405-13. Lamkin, T. D., S. F. Walk, et al. (1997). "She interaction with Src homology 2 domain containing inositol phosphatase (SHIP) in vivo requires the She-phosphotyrosine binding domain and two specific phosphotyrosines on SHIP." J Biol Chem 272(16): 10396-401. Landschulz, K. T., A. N . Noyes, et al. (1989). "Erythropoietin receptors on murine erythroid colony-forming units: natural history." Blood 73(6): 1476-86. Lappin, T. (2003). "The cellular biology of erythropoietin receptors." Oncologist 8 Suppl 1: 15-8. Lappin, T. R., A. P. Maxwell, et al. (2002). "EPO's alter ego: erythropoietin has multiple actions." Stem Cells 20(6): 485-92. Lataillade, J. J., D. Clay, et al. (2000). "Chemokine SDF-1 enhances circulating CD34(+) cell proliferation in synergy with cytokines: possible role in progenitor survival." Blood 95(3): 756-68. Lecoq-Lafon, C , F. Verdier, et al. (1999). "Erythropoietin induces the tyrosine phosphorylation of GAB1 and its association with SHC, SHP2, SHIP, and phosphatidylinositol 3-kinase." Blood 93(8): 2578-85. Lecuyer, E., S. Herblot, et al. (2002). "The SCL complex regulates c-kit expression in hematopoietic cells through functional interaction with Spl." Blood 100(7): 2430-40. Lee, J. O., H. Yang, et al. (1999). "Crystal structure of the PTEN tumor suppressor: implications for its phosphoinositide phosphatase activity and membrane association." CeU 99(3): 323-34. Lemischka, I. (2001). "Stem cell dogmas in the genomics era." Rev Clin Exp Hematol 5(1): 15-25. 177 Lennartsson, J., P. Blume-Jensen, et al. (1999). "Phosphorylation of She by Src family kinases is necessary for stem cell factor receptor/c-kit mediated activation of the Ras/MAP kinase pathway and c-fos induction." Oncogene 18(40): 5546-53. Leslie, N . R. and C. P. Downes (2002). "PTEN: The down side of PI 3-kinase signalling." Cell Signal 14(4): 285-95. Lesourne, R., W. H. Fridman, et al. (2005). "Dynamic interactions of Fc{gamma} receptor IIB with filamin-bound SHIP1 amplify filamentous actin-dependent negative regulation of Fc{epsilon} receptor I signaling." J Immunol 174(3): 1365-73. Lin, C. S., S. K. Lim, et al. (1996). "Differential effects of an erythropoietin receptor gene disruption on primitive and definitive erythropoiesis." Genes Dev 10(2): 154-64. Lin, F. K., S. Suggs, et al. (1985). "Cloning and expression of the human erythropoietin gene." Proc Natl Acad Sci U S A 82(22): 7580-4. Lindquist, J. A., L. Simeoni, et al. (2003). "Transmembrane adapters: attractants for cytoplasmic effectors." Immunol Rev 191: 165-82. Linnekin, D., C. S. DeBerry, et al. (1997). "Lyn associates with the juxtamembrane region of c-Kit and is activated by stem cell factor in hematopoietic cell lines and normal progenitor cells." J Biol Chem 272(43): 27450-5. Lioubin, M . N. , P. A. Algate, et al. (1996). "pl50Ship, a signal transduction molecule with inositol polyphosphate-5-phosphatase activity." Genes Dev 10(9): 1084-95. Liu, L., J. E. Damen, et al. (1994). "Multiple cytokines stimulate the binding of a common 145-kilodalton protein to She at the Grb2 recognition site of She." Mol Cell Biol 14(10): 6926-35. Liu, L., J. E. Damen, et al. (1997). "SHIP, a new player in cytokine-induced signalling." Leukemia 11(2): 181-4. Liu, Q., T. Sasaki, et al. (1999). "SHIP is a negative regulator of growth factor receptor-mediated PKB/Akt activation and myeloid cell survival." Genes Dev 13(7): 786-91. Liu, Q., F. Shalaby, et al. (1998). "The SH2-containing inositol polyphosphate 5-phosphatase, ship, is expressed during hematopoiesis and spermatogenesis." Blood 91(8): 2753-9. Liu, Y. and L. R. Rohrschneider (2002). "The gift of Gab." FEBS Lett 515(1-3): 1-7. Lok, C. N . and P. Ponka (2000). "Identification of an erythroid active element in the transferrin receptor gene." J Biol Chem 275(31): 24185-90. Lok, S., K. Kaushansky, et al. (1994). "Cloning and expression of murine thrombopoietin cDNA and stimulation of platelet production in vivo." Nature 369(6481): 565-8. Lorenz, E., D. Uphoff, et al. (1951). "Modification of irradiation injury in mice and guinea pigs by bone marrow injections." J Natl Cancer Inst 12(1): 197-201. Lowell, C. A. (2004). "Src-family kinases: rheostats of immune cell signaling." Mol Immunol 41(6-7): 631-43. Lucas, D. M . and L. R. Rohrschneider (1999). "A novel spliced form of SH2-containing inositol phosphatase is expressed during myeloid development." Blood 93(6): 1922-33. Luna, E. J. and A. L. Hitt (1992). "Cytoskeleton—plasma membrane interactions." Science 258(5084): 955-64. Luo, J. M. , H. Yoshida, et al. (2003). "Possible dominant-negative mutation of the SHIP gene in acute myeloid leukemia." Leukemia 17(1): 1-8. 178 Luukko, K., A. Ylikorkala, et al. (1999). "Expression of LKB 1 and PTEN tumor suppressor genes during mouse embryonic development." Mech Dev 83(1-2): 187-90. MacDonald, S. M . and B. M . Vonakis (2002). "Association of the Src homology 2 domain-containing inositol 5' phosphatase (SHIP) to releasability in human basophils." Mol Immunol 38(16-18): 1323-7. Machide, M. , H. Mano, et al. (1995). "Interleukin 3 and erythropoietin induce association of Vav with Tec kinase through Tec homology domain." Oncogene 11(4): 619-25. Maciejewski, J., C. Selleri, et al. (1995). "Fas antigen expression on CD34+ human marrow cells is induced by interferon gamma and tumor necrosis factor alpha and potentiates cytokine-mediated hematopoietic suppression in vitro." Blood 85(11): 3183-90. Maehama, T. and J. E. Dixon (1999). "PTEN: a tumour suppressor that functions as a phospholipid phosphatase." Trends Cell Biol 9(4): 125-8. Main, J. M . and R. T. Prehn (1955). "Successful skin homografts after the administration of high dosage X radiation and homologous bone marrow." J Natl Cancer Inst 15(4): 1023-9. Majka, M. , A. Janowska-Wieczorek, et al. (2001). "Numerous growth factors, cytokines, and chemokines are secreted by human CD34(+) cells, myeloblasts, erythroblasts, and megakaryoblasts and regulate normal hematopoiesis in an autocrine/paracrine manner." Blood 97(10): 3075-85. Mamillapalli, R., N . Gavrilova, et al. (2001). "PTEN regulates the ubiquitin-dependent degradation of the CDK inhibitor p27(KIPl) through the ubiquitin E3 ligase SCF(SKP2)." Curr Biol 11(4): 263-7. Marchler-Bauer, A., J. B. Anderson, et al. (2005). "CDD: a Conserved Domain Database for protein classification." Nucleic Acids Res 33 Database Issue: D192-6. Marion, E., P. J. Kaisaki, et al. (2002). "The gene INPPL1, encoding the lipid phosphatase SHIP2, is a candidate for type 2 diabetes in rat and man." Diabetes 51(7): 2012-7. Marshall, A. J., A. K. Krahn, et al. (2002). "TAPP1 and TAPP2 are targets of phosphatidylinositol 3-kinase signaling in B cells: sustained plasma membrane recruitment triggered by the B-cell antigen receptor." Mol Cell Biol 22(15): 5479-91. Mason, J. M., Halupa, A., Hyam, D., Iscove, N. , Dumont, D., Barber, D. (2002). "SHIP-1 regulates the proliferation and mobilization of the erythroid lineage." Blood 100: 519a (abstract). Matsumoto, A., Y. Seki, et al. (1999). "Suppression of STAT5 functions in liver, mammary glands, and T cells in cytokine-inducible SH2-containing protein 1 transgenic mice." Mol Cell Biol 19(9): 6396-407. Matsuuchi, L. and M . R. Gold (2001). "New views of BCR structure and organization." Curr Opin Immunol 13(3): 270-7. Maxwell, A. P. (2002). "Novel erythropoiesis-stimulating protein in the management of the anemia of chronic renal failure." Kidney Int 62(2): 720-9. Maxwell, M. J., Y. Yuan, et al. (2004). "SHIP1 and Lyn Kinase Negatively Regulate Integrin alpha lib beta 3 signaling in platelets." J Biol Chem 279(31): 32196-204. McKercher, S. R., B. E. Torbett, et al. (1996). "Targeted disruption of the PU.l gene results in multiple hematopoietic abnormalities." Embo J 15(20): 5647-58. Mera, A., M. Suga, et al. (1999). "Induction of cell shape changes through activation of the interleukin-3 common beta chain receptor by the RON receptor-type tyrosine kinase." J Biol Chem 274(22): 15766-74. 179 Merchav, S., I. Tatarsky, et al. (1988). "Enhancement of human granulopoiesis in vitro by biosynthetic insulin-like growth factor I/somatomedin C and human growth hormone." J Clin Invest 81(3): 791-7. Metcalf, D. (1989). "The molecular control of cell division, differentiation commitment and maturation in haemopoietic cells." Nature 339(6219): 27-30. Metcalf, D. (1993). "Hematopoietic regulators: redundancy or subtlety?" Blood 82(12): 3515-23. Migliaccio, A. R., M . Bruno, et al. (1987). "Evidence for direct action of human biosynthetic (recombinant) GM-CSF on erythroid progenitors in serum-free culture." Blood 70(6): 1867-71. Migliaccio, A. R. and G. Migliaccio (1988). "Human embryonic hemopoiesis: control mechanisms underlying progenitor differentiation in vitro." Dev Biol 125(1): 127-34. Migliaccio, A. R. and G. Migliaccio (1998). "The making of an erythroid cell. Molecular control of hematopoiesis." Biotherapy 10(4): 251-68. Migliaccio, A. R., A. M. Vannucchi, et al. (1996). Molecular control of erythroid differentiation. Int J Hematol. 64: 1-29. Migliaccio, A. R., A. M. Vannucchi, et al. (1996). "Molecular control of erythroid differentiation." Int J Hematol 64(1): 1-29. Migliaccio, G., A. R. Migliaccio, et al. (1988). "In vitro differentiation of human granulocyte/macrophage and erythroid progenitors: comparative analysis of the influence of recombinant human erythropoietin, G-CSF, GM-CSF, and IL-3 in serum-supplemented and serum-deprived cultures." Blood 72(1): 248-56. Migliaccio, G., A. R. Migliaccio, et al. (1990). "The biology of hematopoietic growth factors: studies in vitro under serum-deprived conditions." Exp Hematol 18(9): 1049-55. Migliaccio, G., A. R. Migliaccio, et al. (1991). "Stem cell factor induces proliferation and differentiation of highly enriched murine hematopoietic cells." Proc Natl Acad Sci U S A 88(16): 7420-4. Mikhalap, S. V., L. M . Shlapatska, et al. (1999). "CDwl50 associates with src-homology 2-containing inositol phosphatase and modulates CD95-mediated apoptosis." J Immunol 162(10): 5719-27. Minamoto, S., K. Ikegame, et al. (1997). "Cloning and functional analysis of new members of STAT induced STAT inhibitor (SSI) family: SSI-2 and SSI-3." Biochem Biophys Res Commun 237(1): 79-83. Miura, O., A. D'Andrea, et al. (1991). "Induction of tyrosine phosphorylation by the erythropoietin receptor correlates with mitogenesis." Mol Cell Biol 11(10): 4895-902. Miura, O., N. Nakamura, et al. (1994). "Erythropoietin induces association of the JAK2 protein tyrosine kinase with the erythropoietin receptor in vivo." Blood 84(5): 1501-7. Miyajima, A., A. L. Mui, et al. (1993). "Receptors for granulocyte-macrophage colony-stimulating factor, interleukin-3, and interleukin-5." Blood 82(7): 1960-74. Miyake, T., C. K. Kung, et al. (1977). "Purification of human erythropoietin." J Biol Chem 252(15): 5558-64. Miyamoto, T., H. Iwasaki, et al. (2002). "Myeloid or lymphoid promiscuity as a critical step in hematopoietic lineage commitment." Dev Cell 3(1): 137-47. Moarefi, I., M . LaFevre-Bernt, et al. (1997). "Activation of the Src-family tyrosine kinase Hck by SH3 domain displacement." Nature 385(6617): 650-3. 180 Moehren, G., N . Markevich, et al. (2002). Temperature dependence of the epidermal growth factor receptor signaling network can be accounted for by a kinetic model. Biochemistry. 41: 306-20. Moody, J. L., L. Xu, et al. (2004). "Anemia, thrombocytopenia, leukocytosis, extramedullary hematopoiesis, and impaired progenitor function in Pten+/-SHIP-/- mice: a novel model of myelodysplasia." Blood 103(12): 4503-10. Moreau-Gachelin, F., F. Wendling, et al. (1996). "Spi-l /PU.l transgenic mice develop multistep erythroleukemias." Mol Cell Biol 16(5): 2453-63. Morrison, S. J., H. D. Hemmati, et al. (1995). "The purification and characterization of fetal liver hematopoietic stem cells." Proc Natl Acad Sci U S A 92(22): 10302-6. Motoyama, N . , T. Kimura, et al. (1999). "bcl-x prevents apoptotic cell death of both primitive and definitive erythrocytes at the end of maturation." J Exp Med 189(11): 1691-8. Mui , A . L. , H . Wakao, et al. (1995). "Interleukin-3, granulocyte-macrophage colony stimulating factor and interleukin-5 transduce signals through two STAT5 homologs." EmboJ 14(6): 1166-75. Murphy, L. O., S. Smith, et al. (2002). "Molecular interpretation of E R K signal duration by immediate early gene products." Nat Cell Biol 4(8): 556-64. Muslin, A. J., J. W. Tanner, et al. (1996). "Interaction of 14-3-3 with signaling proteins is mediated by the recognition of phosphoserine." Cell 84(6): 889-97. Myers, M . P., I. Pass, et al. (1998). "The lipid phosphatase activity of PTEN is critical for its tumor supressor function." Proc Natl Acad Sci U S A 95(23): 13513-8. Myklebust, J. H. , H. K. Blomhoff, et al. (2002). "Activation of phosphatidylinositol 3-kinase is important for erythropoietin-induced erythropoiesis from CD34(+) hematopoietic progenitor cells." Exp Hematol 30(9): 990-1000. NaNakorn, T., D. Traver, et al. (2002). "Myeloerythroid-restricted progenitors are sufficient to confer radioprotection and provide the majority of day 8 CFU-S." J Clin Invest 109(12): 1579-85. Nagata, K. , T. Nakamura, et al. (1997). "The Ig alpha/lgbeta heterodimer on mu-negative proB cells is competent for transducing signals to induce early B cell differentiation." Immunity 7(4): 559-70. Naka, T., M . Narazaki, et al. (1997). "Structure and function of a new STAT-induced STAT inhibitor." Nature 387(6636): 924-9. Nakamura, K., T. Kouro, et al. (2004). "Src homology 2-containing 5-inositol phosphatase (SHIP) suppresses an early stage of lymphoid cell development through elevated interleukin-6 production by myeloid cells in bone marrow." J Exp Med 199(2): 243-54. Ney, P. A. and A. D. D'Andrea (2000). "Friend erythroleukemia revisited." Blood 96(12): 3675-80. Nguyen, M . , M . Solle, et al. (2002). "Receptors and signaling mechanisms required for prostaglandin E2-mediated regulation of mast cell degranulation and IL-6 production." J Immunol 169(8): 4586-93. Nishizumi, H. , K. Horikawa, et al. (1998). "A double-edged kinase Lyn: a positive and negative regulator for antigen receptor-mediated signals." J Exp Med 187(8): 1343-8. 181 Nocka, K., S. Majumder, et al. (1989). "Expression of c-kit gene products in known cellular targets of W mutations in normal and W mutant mice—evidence for an impaired c-kit kinase in mutant mice." Genes Dev 3(6): 816-26. Nowell, P. C , L. J. Cole, et al. (1956). "Growth and continued function of rat marrow cells in x-radiated mice." Cancer Res 16(3): 258-61. Nunez, G. and L. del Peso (1998). "Linking extracellular survival signals and the apoptotic machinery." Curr Opin Neurobiol 8(5): 613-8. Ogawa, M. , S. Nishikawa, et al. (1993). "Expression and function of c-Kit in fetal hemopoietic progenitor cells: transition from the early c-Kit-independent to the late c-Kit-dependent wave of hemopoiesis in the murine embryo." Development 117(3): 1089-98. Oliver, J. M. , D. L. Burg, et al. (1994). "Inhibition of mast cell Fc epsilon Rl-mediated signaling and effector function by the Syk-selective inhibitor, piceatannol." J Biol Chem 269(47): 29697-703. Orkin, S. H. (1992). "GATA-binding transcription factors in hematopoietic cells." Blood 80(3): 575-81. Orkin, S. H. (1996). "Development of the hematopoietic system." Curr Opin Genet Dev 6(5): 597-602. Orkin, S. H. and L. I. Zon (2002). "Hematopoiesis and stem cells: plasticity versus developmental heterogeneity." Nat Immunol 3(4): 323-8. Osada, H., G. Grutz, et al. (1995). "Association of erythroid transcription factors: complexes involving the LIM protein RBTN2 and the zinc-finger protein GATA1." Proc Natl Acad Sci U S A 92(21): 9585-9. Osawa, M. , K. Hanada, et al. (1996). "Long-term lymphohematopoietic reconstitution by a single CD34-low/negative hematopoietic stem cell." Science 273(5272): 242-5. Osborne, M . A., G. Zenner, et al. (1996). "The inositol 5'-phosphatase SHIP binds to immunoreceptor signaling motifs and responds to high affinity IgE receptor aggregation." J Biol Chem 271(46): 29271-8. Palis, J. , S. Robertson, et al. (1999). "Development of erythroid and myeloid progenitors in the yolk sac and embryo proper of the mouse." Development 126(22): 5073-84. Palis, J. and G. B. Segel (1998). "Developmental biology of erythropoiesis." Blood Rev 12(2): 106-14. Parravicini, V., M . Gadina, et al. (2002). "Fyn kinase initiates complementary signals required for IgE-dependent mast cell degranulation." Nat Immunol 3(8): 741-8. Parsons, R. and L. Simpson (2003). "PTEN and cancer." Methods Mol Biol 222: 147-66. Pawson, T. and P. Nash (2000). "Protein-protein interactions define specificity in signal transduction." Genes Dev 14(9): 1027-47. Pawson, T. and J. D. Scott (1997). "Signaling through scaffold, anchoring, and adaptor proteins." Science 278(5346): 2075-80. Perry, C. and H. Soreq (2002). "Transcriptional regulation of erythropoiesis. Fine tuning of combinatorial multi-domain elements." Eur J Biochem 269(15): 3607-18. Petrie, R. J., P. P. Schnetkamp, et al. (2000). "Transient translocation of the B cell receptor and Src homology 2 domain-containing inositol phosphatase to lipid rafts: evidence toward a role in calcium regulation." J Immunol 165(3): 1220-7. 182 Pevny, L., M. C. Simon, et al. (1991). "Erythroid differentiation in chimaeric mice blocked by a targeted mutation in the gene for transcription factor GATA-1." Nature 349(6306): 257-60. Phee, H., A. Jacob, et al. (2000). "Enzymatic activity of the Src homology 2 domain-containing inositol phosphatase is regulated by a plasma membrane location." J Biol Chem 275(25): 19090-7. Phee, H., W. Rodgers, et al. (2001). "Visualization of negative signaling in B cells by quantitative confocal microscopy." Mol Cell Biol 21(24): 8615-25. Porcher, C , W. Swat, et al. (1996). "The T cell leukemia oncoprotein SCL/tal-1 is essential for development of all hematopoietic lineages." Cell 86(1): 47-57. Rameh, L. E. and L. C. Cantley (1999). "The role of phosphoinositide 3-kinase lipid products in cell function." J Biol Chem 274(13): 8347-50. Rane, S. G. and E. P. Reddy (2002). "JAKs, STATs and Src kinases in hematopoiesis." Oncogene 21(21): 3334-58. Rao, P. and R. A. Mufson (1995). "A membrane proximal domain of the human interleukin-3 receptor beta c subunit that signals DNA synthesis in NIH 3T3 cells specifically binds a complex of Src and Janus family tyrosine kinases and phosphatidylinositol 3-kinase." J Biol Chem 270(12): 6886-93. Rauh, M . J. and G. Krystal (2002). "Of mice and men: elucidating the role of SH2-containing inositol 5-phosphatase (SHIP) in human disease." Clin Invest Med 25(3): 68-70. Rauh, M. J., L. M . Sly, et al. (2004). "The role of SHIP1 in macrophage programming and activation." Biochem Soc Trans 32(Pt 5): 785-8. Ren, R., B. J. Mayer, et al. (1993). "Identification of a ten-amino acid proline-rich SH3 binding site." Science 259(5098): 1157-61. Resh, M. D. (1999). "Fatty acylation of proteins: new insights into membrane targeting of myristoylated and palmitoylated proteins." Biochim Biophys Acta 1451(1): 1-16. Riley, R. S., J. M . Ben-Ezra, et al. (2002). "The virtual blood film." Clin Lab Med 22(1): 317-45. Rogowski, O., Y. Sasson, et al. (1998). "Down-regulation of the CD62L antigen as a possible mechanism for neutrophilia during inflammation." Br J Haematol 101(4): 666-9. Ruscetti, S. K. (1999). "Deregulation of erythropoiesis by the Friend spleen focus-forming virus." Int J Biochem Cell Biol 31(10): 1089-109. Russell, E. S. (1979). "Hereditary anemias of the mouse: a review for geneticists." Adv Genet 20: 357-459. Sadowski, I., J. C. Stone, et al. (1986). "A noncatalytic domain conserved among cytoplasmic protein-tyrosine kinases modifies the kinase function and transforming activity of Fujinami sarcoma virus P130gag-fps." Mol Cell Biol 6(12): 4396-408. Sansal, I. and W. R. Sellers (2004). "The biology and clinical relevance of the PTEN tumor suppressor pathway." J Clin Oncol 22(14): 2954-63. Sarmay, G , G. Koncz, et al. (1999). "Cooperation between SHP-2, phosphatidyl inositol 3-kinase and phosphoinositol 5-phosphatase in the Fc gamma Rllb mediated B cell regulation." Immunol Lett 68(1): 25-34. Sasaki, A., H. Yasukawa, et al. (2000). "CIS3/SOCS-3 suppresses erythropoietin (EPO) signaling by binding the EPO receptor and JAK2." J Biol Chem 275(38): 29338-47. Sato, T. K., M . Overduin, et al. (2001). "Location, location, location: membrane targeting directed by PX domains." Science 294(5548): 1881-5. 183 Satterthwaite, A. B. and O. N . Witte (2000). "The role of Bruton's tyrosine kinase in B-cell development and function: a genetic perspective." Immunol Rev 175: 120-7. Sattler, M. , S. Verma, et al. (2001). "SHIP1, an SH2 domain containing polyinositol-5-phosphatase, regulates migration through two critical tyrosine residues and forms a novel signaling complex with DOK1 and CRKL." J Biol Chem 276(4): 2451-8. Sawada, K., S. B. Krantz, et al. (1988). "Quantitation of specific binding of erythropoietin to human erythroid colony-forming cells." J Cell Physiol 137(2): 337-45. Scalia, P., E. Heart, et al. (2001). "Regulation of the Akt/Glycogen synthase kinase-3 axis by insulin-like growth factor-II via activation of the human insulin receptor isoform-A." J Cell Biochem 82(4): 610-8. Scheid, M. P., M. Huber, et al. (2002). "Phosphatidylinositol (3,4,5)P3 is essential but not sufficient for protein kinase B (PKB) activation; phosphatidylinositol (3,4)P2 is required for PKB phosphorylation at Ser-473: studies using cells from SH2-containing inositol-5-phosphatase knockout mice." J Biol Chem 277(11): 9027-35. Schubert, K. M . and V. Duronio (2001). "Distinct roles for extracellular-signal-regulated protein kinase (ERK) mitogen-activated protein kinases and phosphatidylinositol 3-kinase in the regulation of Mcl-1 synthesis." Biochem J 356(Pt 2): 473-80. Schuetze, S., P. E. Stenberg, et al. (1993). "The Ets-related transcription factor PU.l immortalizes erythroblasts." Mol Cell Biol 13(9): 5670-8. Scott, E. W., M . C. Simon, et al. (1994). "Requirement of transcription factor PU.l in the development of multiple hematopoietic lineages." Science 265(5178): 1573-7. Shaw, A. B. (1967). "Haemolysis in chronic renal failure." Br Med J 2(546): 213-6. Shivdasani, R. A. and S. H. Orkin (1996). "The transcriptional control of hematopoiesis." Blood 87(10): 4025-39. Shupnik, M. A. (2004). "Crosstalk between steroid receptors and the c-Src-receptor tyrosine kinase pathways: implications for cell proliferation." Oncogene 23(48): 7979-89. Sicheri, F. and J. Kuriyan (1997). "Structures of Src-family tyrosine kinases." Curr Opin Struct Biol 7(6): 777-85. Sicheri, F., I. Moarefi, et al. (1997). "Crystal structure of the Src family tyrosine kinase Hck." Nature 385(6617): 602-9. Siegel, J., Y. Li , et al. (1999). "SHIP-mediated inhibition of K562 erythroid differentiation requires an intact catalytic domain and She binding site." Oncogene 18(50): 7135-48. Silva, M. , C. Richard, et al. (1998). "Expression of Bcl-x in erythroid precursors from patients with polycythemia vera." N Engl J Med 338(9): 564-71. Silvennoinen, O., B. A. Witthuhn, et al. (1993). "Structure of the murine Jak2 protein-tyrosine kinase and its role in interleukin 3 signal transduction." Proc Natl Acad Sci U S A 90(18): 8429-33. Sleeman, M . W., K. E. Wortley, et al. (2005). "Absence of the lipid phosphatase SHIP2 confers resistance to dietary obesity." Nat Med 11(2): 199-205. Sly, L. M. , M. J. Rauh, et al. (2003). "SHIP, SHIP2, and PTEN activities are regulated in vivo by modulation of their protein levels: SHIP is up-regulated in macrophages and mast cells by lipopolysaccharide." Exp Hematol 31(12): 1170-81. Smith, C. (2003). "Hematopoietic stem cells and hematopoiesis." Cancer Control 10(1): 9-16. Socolovsky, M. , H. Nam, et al. (2001). "Ineffective erythropoiesis in Stat5a(-/-)5b(-/-) mice due to decreased survival of early erythroblasts." Blood 98(12): 3261-73. 184 Sonoda, Y., Y. C. Yang, et al. (1988). "Analysis in serum-free culture of the targets of recombinant human hemopoietic growth factors: interleukin 3 and granulocyte/macrophage-colony-stimulating factor are specific for early developmental stages." Proc Natl Acad Sci U S A 85(12): 4360-4. Sparks, A. B., J. E. Rider, et al. (1996). "Distinct ligand preferences of Src homology 3 domains from Src, Yes, Abl, Cortactin, p53bp2, PLCgamma, Crk, and Grb2." Proc Natl Acad Sci U S A 93(4): 1540-4. Sparks, A. B., J. E. Rider, et al. (1998). "Mapping the specificity of SH3 domains with phage-displayed random-peptide libraries." Methods Mol Biol 84: 87-103. Spivak, J. L., J. Marmor, et al. (1972). "Studies on splenic erythropoiesis in the mouse. I. Ribosomal ribonucleic acid metabolism." J Lab Clin Med 79(4): 526-40. Starr, R., T. A. Willson, et al. (1997). "A family of cytokine-inducible inhibitors of signalling." Nature 387(6636): 917-21. Sudol, M. (1998). "From Src Homology domains to other signaling modules: proposal of the 'protein recognition code'." Oncogene 17(11 Reviews): 1469-74. Sui, X. , S. B. Krantz, et al. (1998). "Synergistic activation of MAP kinase (ERK1/2) by erythropoietin and stem cell factor is essential for expanded erythropoiesis." Blood 92(4): 1142-9. Sutterluty, H., E. Chatelain, et al. (1999). "p45SKP2 promotes p27Kipl degradation and induces S phase in quiescent cells." Nat Cell Biol 1(4): 207-14. Suzuki, N. , N . Suwabe, et al. (2003). "Identification and characterization of 2 types of erythroid progenitors that express GATA-1 at distinct levels." Blood 102(10): 3575-83. Takeshita, S., N . Namba, et al. (2002). "SHIP-deficient mice are severely osteoporotic due to increased numbers of hyper-resorptive osteoclasts." Nat Med 8(9): 943-9. Tamir, I., J. C. Stolpa, et al. (2000). "The RasGAP-binding protein p62dok is a mediator of inhibitory FcgammaRIIB signals in B cells." Immunity 12(3): 347-58. Tatton, L., G. M. Morley, et al. (2003). "The Src-selective kinase inhibitor PP1 also inhibits Kit and Bcr-Abl tyrosine kinases." J Biol Chem 278(7): 4847-53. Tauchi, T., J. E. Damen, et al. (1996). "Tyrosine 425 within the activated erythropoietin receptor binds Syp, reduces the erythropoietin required for Syp tyrosine phosphorylation, and promotes mitogenesis." Blood 87(11): 4495-501. Tavernier, J., R. Devos, et al. (1991). "A human high affinity interleukin-5 receptor (IL5R) is composed of an IL5-specific alpha chain and a beta chain shared with the receptor for GM-CSF." Cell 66(6): 1175-84. Testa, U. (2004). "Apoptotic mechanisms in the control of erythropoiesis." Leukemia 18(7): 1176-99. Thomas, E. D., R. Storb, et al. (1975). "Bone-marrow transplantation (second of two parts)." N Engl J Med 292(17): 895-902. Thomas, S. M . and J. S. Brugge (1997). "Cellular functions regulated by Src family kinases." Annu Rev Cell Dev Biol 13: 513-609. Thompson, J. E. and C. B. Thompson (2004). Putting the rap on Akt. J Clin Oncol. 22: 4217-26. Tilbrook, P. A., E. Ingley, et al. (1997). "Lyn tyrosine kinase is essential for erythropoietin-induced differentiation of J2E erythroid cells." Embo J 16(7): 1610-9. 185 Tilbrook, P. A., G. A. Palmer, et al: (2001). "Maturation of erythroid cells and erythroleukemia development are affected by the kinase activity of Lyn." Cancer Res 61(6): 2453-8. Till, J. E. and E. A. McCulloch (1961). A direct measurement of the radiation sensitivity of normal mouse bone marrow cells. Radiat. Res. 14: 213-22. Timokhina, I., H. Kissel, et al. (1998). "Kit signaling through PI 3-kinase and Src kinase pathways: an essential role for Racl and JNK activation in mast cell proliferation." Embo J 17(21): 6250-62. Toker, A. (2002). "Phosphoinositides and signal transduction." Cell Mol Life Sci 59(5): 761-79. Tomlinson, M. G., V. L. Heath, et al. (2004). "SHIP family inositol phosphatases interact with and negatively regulate the Tec tyrosine kinase." J Biol Chem 279(53): 55089-96. Tong, A. H., B. Drees, et al. (2002). "A combined experimental and computational strategy to define protein interaction networks for peptide recognition modules." Science 295(5553): 321-4. Torigoe, T., O. C. R, et al. (1992). "Regulation of SRC-family protein tyrosine kinases by interleukins, IL-2, and IL-3." Leukemia 6 Suppl 3: 94S-97S. Tridandapani, S., M . Pradhan, et al. (1999). "Protein interactions of Src homology 2 (SH2) domain-containing inositol phosphatase (SHIP): association with She displaces SHIP from FcgammaRIIb in B cells." J Immunol 162(3): 1408-14. Tsai, S. F., D. I. Martin, et al. (1989). "Cloning of cDNA for the major DNA-binding protein of the erythroid lineage through expression in mammalian cells." Nature 339(6224): 446-51. Tsang, A. P., Y. Fujiwara, et al. (1998). "Failure of megakaryopoiesis and arrested erythropoiesis in mice lacking the GATA-1 transcriptional cofactor FOG." Genes Dev 12(8): 1176-88. Tsang, A. P., J. E. Visvader, et al. (1997). "FOG, a multitype zinc finger protein, acts as a cofactor for transcription factor GATA-1 in erythroid and megakaryocytic differentiation." Cell 90(1): 109-19. Tu, Z., J. M . Ninos, et al. (2001). "Embryonic and hematopoietic stem cells express a novel SH2-containing inositol 5'-phosphatase isoform that partners with the Grb2 adapter protein." Blood 98(7): 2028-38. Turner, H. and J. P. Kinet (1999). "Signalling through the high-affinity IgE receptor Fc epsilonRI." Nature 402(6760 Suppl): B24-30. Tzivion, G., Y. H. Shen, et al. (2001). "14-3-3 proteins; bringing new definitions to scaffolding." Oncogene 20(44): 6331-8. Vallentin, A., C. Prevostel, et al. (2000). "Membrane targeting and cytoplasmic sequestration in the spatiotemporal localization of human protein kinase C alpha." J Biol Chem 275(8): 6014-21. van den Akker, E., T. van Dijk, et al. (2004). "Tyrosine kinase receptor RON functions downstream of the erythropoietin receptor to induce expansion of erythroid progenitors." Blood 103(12): 4457-65. Vannucchi, A. M. , L. Bianchi, et al. (2001). "Accentuated response to phenylhydrazine and erythropoietin in mice genetically impaired for their GATA-1 expression (GATA-l(low) mice)." Blood 97(10): 3040-50. 186 Velazquez, L., G. D. Gish, et al. (2000). "The she adaptor protein forms interdependent phosphotyrosine-mediated protein complexes in mast cells stimulated with interleukin 3." Blood 96(1): 132-8. Verdier, F., S. Chretien, et al. (1998). "Proteasomes regulate erythropoietin receptor and signal transducer and activator of transcription 5 (STAT5) activation. Possible involvement of the ubiquitinated Cis protein." J Biol Chem 273(43): 28185-90. Verdier, F., R. Rabionet, et al. (1998). "A sequence of the CIS gene promoter interacts preferentially with two associated STAT5A dimers: a distinct biochemical difference between STAT5A and STAT5B." Mol Cell Biol 18(10): 5852-60. Vyas, P., M. A. McDevitt, et al. (1999). "Different sequence requirements for expression in erythroid and megakaryocytic cells within a regulatory element upstream of the GATA-1 gene." Development 126(12): 2799-811. Wagner, K. U., E. Claudio, et al. (2000). "Conditional deletion of the Bcl-x gene from erythroid cells results in hemolytic anemia and profound splenomegaly." Development 127(22): 4949-58. Wang, J. W., J. M. Howson, et al. (2002). "Influence of SHIP on the NK repertoire and allogeneic bone marrow transplantation." Science 295(5562): 2094-7. Wang, M . H., Y. Q. Zhou, et al. (2002). "Macrophage-stimulating protein and RON receptor tyrosine kinase: potential regulators of macrophage inflammatory activities." Scand J Immunol 56(6): 545-53. Ware, M . D., P. Rosten, et al. (1996). "Cloning and characterization of human SHIP, the 145-kD inositol 5-phosphatase that associates with SHC after cytokine stimulation." Blood 88(8): 2833-40. Warren, A. J., W. H. Colledge, et al. (1994). "The oncogenic cysteine-rich LIM domain protein rbtn2 is essential for erythroid development." Cell 78(1): 45-57. Weintraub, B. C , J. E. Jun, et al. (2000). "Entry of B cell receptor into signaling domains is inhibited in tolerant B cells." J Exp Med 191(8): 1443-8. Weiss, M. J. and S. H. Orkin (1995). "Transcription factor GATA-1 permits survival and maturation of erythroid precursors by preventing apoptosis." Proc Natl Acad Sci U S A 92(21): 9623-7. Weissman, I. L. (2000). "Stem cells: units of development, units of regeneration, and units in evolution." Cell 100(1): 157-68. Winearls, C. G., D. O. Oliver, et al. (1986). "Effect of human erythropoietin derived from recombinant DNA on the anaemia of patients maintained by chronic haemodialysis." Lancet 2(8517): 1175-8. Winkelmann, J. C , J. Ward, et al. (1995). "A translocated erythropoietin receptor gene in a human erythroleukemia cell line (TF-1) expresses an abnormal transcript and a truncated protein." Blood 85(1): 179-85. Wisniewski, D., A. Strife, et al. (1999). "A novel SH2-containing phosphatidylinositol 3,4,5-trisphosphate 5-phosphatase (SHIP2) is constitutively tyrosine phosphorylated and associated with src homologous and collagen gene (SHC) in chronic myelogenous leukemia progenitor cells." Blood 93(8): 2707-20. Witthuhn, B. A., F. W. Quelle, et al. (1993). "JAK2 associates with the erythropoietin receptor and is tyrosine phosphorylated and activated following stimulation with erythropoietin." CeH 74(2): 227-36. 187 Wolf, I., D. M. Lucas, et al. (2000). "Cloning of the genomic locus of mouse SH2 containing inositol 5-phosphatase (SHIP) and a novel 110-kDa splice isoform, SHIPdelta." Genomics 69(1): 104-12. Wright, D. E., E. P. Bowman, et al. (2002). "Hematopoietic stem cells are uniquely selective in their migratory response to chemokines." J Exp Med 195(9): 1145-54. Wu, A. M. , J. E. Till, et al. (1967). "A cytological study of the capacity for differentiation of normal hemopoietic colony-forming cells." J Cell Physiol 69(2): 177-84. Wu, H., U. Klingmuller, et al. (1997). "Functional interaction of erythropoietin and stem cell factor receptors is essential for erythroid colony formation." Proc Natl Acad Sci U S A 94(5): 1806-10. Wu, H , X. Liu, et al. (1995). "Generation of committed erythroid BFU-E and CFU-E progenitors does not require erythropoietin or the erythropoietin receptor." Cell 83(1): 59-67. Wu, X. , K. Hepner, et al. (2000). "Evidence for regulation of the PTEN tumor suppressor by a membrane-localized multi-PDZ domain containing scaffold protein MAGI-2." Proc Natl Acad Sci U S A 97(8): 4233-8. Xu, Y., K. W. Harder, et al. (2005). "Lyn tyrosine kinase: accentuating the positive and the negative." Immunity 22(1): 9-18. Yaffe, M . B. (2002). "Phosphotyrosine-binding domains in signal transduction." Nat Rev Mol Cell Biol 3(3): 177-86. Yamamoto, T., Y. Yamanashi, et al. (1993). "Association of Src-family kinase Lyn with B-cell antigen receptor." Immunol Rev 132: 187-206. Yamanashi, Y., T. Kakiuchi, et al. (1991). "Association of B cell antigen receptor with protein tyrosine kinase Lyn." Science 251(4990): 192-4. Yen, J. J., Y. C. Hsieh, et al. (1995). "Restoring the apoptosis suppression response to IL-5 confers on erythroleukemic cells a phenotype of IL-5-dependent growth." J Immunol 154(5): 2144-52. Yoshimura, A., T. Ohkubo, et al. (1995). "A novel cytokine-inducible gene CIS encodes an SH2-containing protein that binds to tyrosine-phosphorylated interleukin 3 and erythropoietin receptors." Embo J 14(12): 2816-26. Yoshimura, A., T. Zimmers, et al. (1992). "Mutations in the Trp-Ser-X-Trp-Ser motif of the erythropoietin receptor abolish processing, ligand binding, and activation of the receptor." J Biol Chem 267(16): 11619-25. Youssoufian, H., G. Longmore, et al. (1993). "Structure, function, and activation of the erythropoietin receptor." Blood 81(9): 2223-36. Zang, H., K. Sato, et al. (2001). "The distal region and receptor tyrosines of the Epo receptor are non-essential for in vivo erythropoiesis." Embo J 20(12): 3156-66. Zanjani, E. D. and J. L. Ascensao (1989). "Erythropoietin." Transfusion 29(1): 46-57. Zeuner, A., A. Eramo, et al. (2003). "Control of erythroid cell production via caspase-mediated cleavage of transcription factor SCL/Tal-1." Cell Death Differ 10(8): 905-13. Zhang, D., M . M . Johnson, et al. (2001). "An optimized system for studies of EPO-dependent murine pro-erythroblast development." Exp Hematol 29(11): 1278-88. Zhang, J. G., A. Farley, et al. (1999). "The conserved SOCS box motif in suppressors of cytokine signaling binds to elongins B and C and may couple bound proteins to proteasomal degradation." Proc Natl Acad Sci U S A 96(5): 2071-6. 188 Zhu, J. and S. G. Emerson (2002). "Hematopoietic cytokines, transcription factors and lineage commitment." Oncogene 21(21): 3295-313. Zon, L. I. (1995). "Developmental biology of hematopoiesis." Blood 86(8): 2876-91. 189 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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


Related Items