UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Interleukin-3 signal transduction : purification and characterization of the murine interleukin-3 receptor Mui, Alice Low-Fung 1992

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

Item Metadata


831-ubc_1992_spring_mui_alice_low_fung.pdf [ 5.2MB ]
JSON: 831-1.0098911.json
JSON-LD: 831-1.0098911-ld.json
RDF/XML (Pretty): 831-1.0098911-rdf.xml
RDF/JSON: 831-1.0098911-rdf.json
Turtle: 831-1.0098911-turtle.txt
N-Triples: 831-1.0098911-rdf-ntriples.txt
Original Record: 831-1.0098911-source.json
Full Text

Full Text

INTERLEUKIN-3 SIGNAL TRANSDUCTION:PURIFICATION AND CHARACTERIZATION OF THE MURINE INTERLEUKIN-3 RECEPTORbyALICE LOW-FLING MUIB.Sc.(honours), The University of British Columbia, 1986A THESIS SUBM1i iLD IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIES(Department of Pathology)We accept this thesis as ctoe required stt____.77F.,--UNrff^IVERSITY OF BRITISH Cal. :, JMBIAFebruary 1992©Alice L-F. Mui, 1992In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.Department of ^PGA University of British ColumbiaVancouver, CanadaDate ^Mat/ , I qg 2- DE-6 (2/88)ABSTRACTMurine interleukin-3 (mIL-3) is a potent hemopoietic growth factor that is producedprimarily by activated T lymphocytes and stimulates the proliferation and differentiation ofpluripotent stem cells and committed myeloid and early lymphoid progenitors. In order to gainsome insight into its mechanism of action, we set out to purify and characterize its cell surfacereceptor. To monitor this purification, an assay capable of detecting detergent solubilizedmIL-3R's was devised. With this assay, a simple two step purification protocol, involvinganti-phosphotyrosine Sepharose and biotinylated mIL-3/streptavidin agarose, was developed topurify the mIL-3R This protocol was based, in part, on a previous observation in ourlaboratory that the mIL-3R becomes tyrosine phosphorylated upon ligand binding. Two formsof the receptor were obtained using this procedure; a tyrosine/serine phosphorylated 140 kDform that was greater that 98% pure, and a less pure, serine phosphorylated 120 kD form.Alkaline phosphatase treatment, chymotrypsin digestion and Western analysis usingantibodies to the N-terminus of the 120 kD form established that, apart from phosphorylationdifferences, these two proteins were identical. Intriguingly, the 140 kD tyrosine/serinephosphorylated mIL-3R was exquisitely sensitive to proteolysis whereas the 120 kD receptorwas not. This proteolysis was also shown to occur in intact cells in response to mIL-3 and tookplace rapidly at 37°C in the presence of lysosomal inhibitors. These observations suggest thatthis mIL-3 stimulated proteolysis occurs at the cell surface and could play an important role inthe mechanism of action of mIL-3.N-terminal amino acid sequencing and amino acid composition analysis of the purifiedmIL-3R showed it to be identical to an mIL-3 binding protein subsequently cloned byexpression in COS cells. Examination of this sequence revealed no consensus kinase domains,indicating the tyrosine phosphorylation of the mIL-3R must be mediated by an associatedkinase. Studies directed towards the identification and purification of this mIL-3R associatedtyrosine kinase, as well as the characterization of other mIL-3R associated proteins that wehave identified, are presented. Our results to date have led us to propose a model of mIL-3Rinduced signal transduction that may also apply, to some extent, to other members of thehemopoietic receptor superfamily.iiiivTABLE OF CONTENTSABSTRACTTABLE OF CONTENTS^ ivLIST OF TABLES viiLIST OF FIGURES viiiLIST OF ABBREVIATIONSACKNOWLEDGEMENTSChapter 1A.157INTRODUCTIONHemopoiesis1. Cells of the hemopoietic system2. Control of hemopoiesis^2.1^Stromal control of hemopoiesis2.2^Humoral control of hemopoiesisB. Hemopoietic Growth Factor Receptors1. Intrinsic Tyrosine kinase receptors 101.1^Structure of tyrosine kinase receptors 101.2^Signalling from tyrosine kinase receptors 132. The Hemopoietin receptor superfamily 162.1^Structure of the hemopoietin receptors 162.1.1 Hemopoietin receptors with solved structures 192.1.2 Hemopoietin receptors with unsolved structures 202.1.3 The receptors for IL-3, GM-CSF and IL-5 212.2^Signalling from the hemopoietin receptors 23C. Proteins Involved in Signal Transduction 241. GTP Binding proteins 251.1^G proteins 251.2^The ras related GTP binding proteins 271.3^Regulators of ras 281.3.1 GTPase activating protein 291.3.2 Nucleotide diphosphate kinase 311.3.3 Nucleotide exchange proteins 312. Protein kinase C and inositol phosphate metabolism 322.1^Protein kinase C 332.1.1 Regulators of PKC 332.1.2 Targets of PKC 352.2^Phospholipase C 362.3^PI3-kinase and other PIP2 metabolites 373. Tyrosine kinases 383.1^Structure of the src family of tyrosine kinase 393.2^Targets of tyrosine kinases 423.3^Association of tyrosine kinases with receptors 424. Serine/threonine specific kinases 434.1^Raf kinase 444.2^MAP kinase 465. Phosphatases 475.1^Tyrosine specific phosphatases 475.2^Serine/threonine specific phosphatases 486. Signalling pathways implicated in mIL-3 action 49D. Cell cycle control 501.^Integration of growth factor with cell cycle control 54E. Thesis Objectives 57VF.^References^ 58Chapter II MATERIALS AND METHODS^86A.^Materials1. Materials^ 862. Growth factors, antibodies and cDNA's^ 863. Cells 874. Preparation of a-mIL-3R peptide antibodies 87B.^General biochemical techniques^ 881. Protein determination 882. Radio-iodinating proteins for SDS-PAGE^ 883. One and two dimensional gel electrophoresis 884. Western blotting^ 885.^Phosphoamino acid analysis^ 89C.^Purification and derivitization of growth factors^ 891. B6SUtA1 cell proliferation assay for mIL-3 and mGM-CSF^892. Production and purification of COS mIL-3 and mGM-CSF 893. Biotinylation of COS cell derived mIL-3^ 904. Iodination of mIL-3^ 90D.^In vivo labelling of B6SUtAsis cells^ 911. Labelling cells with "P-orthophosphate^ 912. Labelling cells with 35S-methionine 91E.^Purification of the mIL-3R^ 911. mIL-3R solubilization 912. Soluble receptor assay 9I3. Purification of the mIL-3R from intact B6SUtA1 cells^924. Amino acid sequence & composition analysis of the mIL-3R^925. Displacement analysis with the 120 kD mIL-3R 936. Scatchard analysis of the mIL-3R on intact cells^ 93F.^Analysis of the tyrosine phosphorylated mIL-3R 931. Purification of tyr and non-tyr phosphorylated mIL-3R's^942. Phosphatase treatment of the tyr phosphorylated mIL-3R 94G. Analysis of mIL-3 induced tyrosine phosphoproteins^ 951. Preparation of detergent lysates from factor treated cells^952. Immunoprecipitation and SH2 precipitation of proteins 953. mIL-3R tyrosine kinase assay^ 964. Sephadex G 150 fractionation of tyrosylphosphoproteins^965.^Preparation of SH2 Sepharose 97H. References^ 98Chapter III PURIFICATION OF THE mIL-3R^ 99A. Introduction^ 99B. Results 100C^Discussion 105D.^References^ 124Chapter IV TYROSINE PHOSPHORYLATION OF THE mIL-3R INCREASES ITS^126SUSCEPTIBILITY TO PROTEOLYSISA. Introduction^ 126B. Results 127C.^Discussion 130viD.^References^ 141Chapter V^IDENTIFICATION OF PROTEINS POTENTIALLY ASSOCIATED^143WITH THE mIL-3RA. Introduction^ 143B. Results 144C. Discussion 149D. References^ 162Chapter VI DEVELOPMENT OF STRATEGIES TO PURIFY mIL-3R^ 164ASSOCIATED PROTEINSA. Introduction^ 164B. Results 165C. Discussion 167D. References^ 174Chapter VII SUMMARY AND CONCLUSIONS^ 175A. Purification of the mIL-3R 175B. Tyrosine phosphorylation of the mIL-3R increases itssusceptibility to cleavage^ 176C. Identification of mIL-3R associated proteins^ 177D. A model for mIL-3R signal transduction 178E.^References^ 180LIST OF TABLESTable 1. Hemopoietic growth factors^ 9Table 2. Hemopoietic inhibitory factors 9Table 3. Molecules associated with intrinsic tyrosine kinase receptors^14Table 4. Tyrosine kinases associated with cell surface receptors 42Table 5. Purification of the 140 kD mIL-3R^ 108Table 6. Comparison of amino acid compositions 109VII114Figure 1.Figure 2.Figure 3.Figure 4A.Figure 4BLIST OF FIGURESA schematic of the hemopoietic hierarchyThe three classes of tyrosine kinase containing receptorsStructure of the hemopoietin receptorsStructure of the src family of tyrosine kinasesProteins with SH2 and SH3 domains212174040Figure 5.^Specificity and sensitivity of the lectin assay for thesolubilized mIL-3R^110Figure 6.^Scatchard analyses 111Streptavidin-agarose elution profile of the mIL-3RTwo dimensional O'Farrell gels of the Streptavidin purified mIL-3RSDS-PAGE of the purified mIL-3R from 35S-labelledB6SUtA1Flow diagram showing the purification of the mIL-3RN-glycanase digestion of 32P-labelled mIL-3R'sEffect of phosphorylation on apparent molecular massChymotryptic maps of the tyrosine phosphorylated andtyrosine unphosphorylated mIL-3R proteinsWestern analysis.of plasma membranes from B6SUtA1 cells.Scatchard analysis with various purified mIL-3R preparationsChymotryptic map of the purified mIL-3R and Aic 2A proteinWestern analysis of B6SUtA 1 cells plasma membranes aftertunicamycin treatmentScatchard analysis of B6SUtA1 cells following tunicamycintreatmentDisplacement kinetics of 1251-mIL-3 from the purified 120 kDmIL-3RTyrosine phosphorylation of the mIL-3R increases its apparentmolecular massTyrosine phosphorylation of the mIL-3R increases itssusceptibility to proteolysisIncreased proteolytic susceptibility of the tyrosinephosphorylated mIL-3R is an intrinsic propertyEffect of phosphatase treatment on receptor integrityFigure 7A.Figure 7B.Figure 8.Figure 9.Figure 10.Figure 11.Figure 12.Figure 13.Figure 14.Figure 15.Figure 16A.Figure 16B.Figure 17.Figure 18.Figure 19.Figure 20.Figure 21.112113118119120121122122123134135137138ixFigure 22.^Degradation of the mIL-3R in intact cells^ 139Figure 23A.^Effect of various inhibitors on mIL-3R cleavage 140Figure 23B.^Effect of kinase inhibitors of mIL-3R tyr phosphorylation^140Figure 24.^Comparison of tyrosine phosphorylations induced by mIL-3,mGM-CSF and SF^ 152Figure 25.^Time course of mIL-3 stimulated phosphorylations at 4°C^153Figure 26A.^Comparison of pp56 and pp32 with cyclin A and cdc2 kinase^154Figure 26B.^a-fps and a-GAP immunoprecipitation from mIL-3 treatedB6SUtA1 cells^ 155Figure 26C.^Effect of growth factor stimulation on the tyrosinephosphorylation of vavFigure 26D. A 10 X longer exposure of the Western blot from Figure 26c^156Figure 27.^Effects of various inhibitors on mIL-3 stimulated tyrosinephosphorylations^ 157Figure 28.^Scheme of mIL-3R tyrosine kinase assay^ 158Figure 29.^Specificity of the mIL-3R tyrosine kinase assay 159Figure 30.^mIL-3R tyrosine kinase activity is present in thea-phosphotyrosine bound protein fraction from mIL-3stimulated B6SUtA 1 cells^ 160Figure 31.^Sephadex G150 fractionation of mIL-3 inducedtyrosine phosphoproteins^ 161Figure 32^B-mIL-3/streptavidin agarose precipitation of tyrosinephosphorylated proteins fromB6SUtA 1 cells^ 169Figure 33^Binding of B6SUtA1 cell proteins to the SH2 domain of GAP^170Figure 34^Binding of B6SUtA 1 cell proteins to various SH2 domains 172Figure 35.^Binding of mIL-3, mGM-CSF and SF induced tyrosinephosphoproteins to GAP SH2 domains^ 173Figure 36.^Model of mIL-3 receptor mediated signal transduction^179LIST OF ABBREVIATIONSATPp-MEB-mIL-3BSAc/mLCNBrCon ADMSDSSDTrEpFCSGAPgstIL-xIL-xRKLHmGM-CSFmIL-3NP40OGPBSPI3-K(inase)PLCyPMSFPSBSASDSSFSH2TBSTBSTadenosine triphosphate3-mecaptoethanolBiotinylated murine interleukin-3bovine serum albumincells/mLcyanogen bromideConcanavalin Adimethyl suberimidatedisuccinimidyl suberatedithiothreitolerythropoietinfetal calf serumGTPase activating proteinglutathione-S-transferaseinterleukin-x (x= 1-12)interleukin-x receptor (x=1-12)keyhole limpet hemocyaninmurine granulocyte-macrophage colony stimulating factormurine interleukin-3Nonidet P40octylglucosidephosphate buffered salinephosphatidylinositol 3-kinasephospholipase Cyphenylmethylsufonyl flouridephosphorylation solubilization bufferstreptavidin agarosesodium dodecyl sulfatesteel factorsrc homology domain 220 mM Tris-C1, 0.15M NaC1, pH 7.420 mM Tris-C1, 0.15M NaC1, pH 7.4, 0.05% Tween 20xiACKNOWLEDGEMENTSI have been priviledged to have studied under the kind and patient supervision of Dr. GeraldKrystal. I owe Gerry my deepest thanks and eternal gratitude for allowing me to work and growin his laboratory.I would also like to thank my supervisory committee members: Dr. Don Brooks, Dr. RogerBrownsey, Dr. Peter Lansdorp and Dr. Katarina Zis for their support and guidance through theyears.I would also like to express my sincere appreciatiation to the staff and students of the TerryFox Laboratory for providing an exciting scientific environment in which to study. Inparticular, I would like to thank Dr. Keith Humphries and Dr. Fumio Takei for their criticalinterest and helpful suggestions. Also, I would like to acknowledge my "big brothers", Dr. PoulSorensen and Dr. Sid Murthy, for intiating me to the lab and Vivian Lam and BarbaraDelaplace for helping me survive Drs. Sorensen and Murthy.Finally, I would like to thank my mother and father, without whose love and support, none ofthis would have been possible.1CHAPTER IINTRODUCTIONA. HEMOPOIESISHemopoiesis is the process by which a pluripotent, hemopoietic stem cell proliferates anddifferentiates to generate all the functional end cells circulating in the blood stream. Sinceblood cells have a limited life span they must be continuously replaced throughout life. Inaddition to maintaining this steady state, the hemopoietic system must also be capable ofresponding to hematologic or immunologic insults by rapid production of appropriate cells.Control of this process is achieved by a complex regulatory system consisting of direct cell-cellinteractions as well as soluble regulatory molecules.1. Cells of the Hemapoietic SystemThe hemopoietic system can be functionally divided into a hierarchy of four cellularcompartments characterized by decreasing self-renewal capability and increasing differentiativestate.' At the top of this hierarchy (Figure 1) are the pluripotent hemopoietic stem cells whichare characterized by extensive self-renewal capacity and unrestricted differentiation potential.These cells differentiate to give the more committed progenitors, of both myeloid and lymphoidlineages, which are more limited in their differentiative capacity. Both stem cells andcommitted progenitors lack distinct morphological characteristics and can only be detected bytheir functional properties. These progenitors divide and differentiate further to give rise to themorphologically identifiable, immediate precursors of the mature blood cells.Stem cells, in the adult, are primarily located in the bone marrow, having migrated therefrom developmentally earlier sites of hemopoiesis in the embryonic yolk sac and fetal liver. 2Totipotent stem cells are characterized functionally by pluripotentiality, self-renewal capacityOG.IL-6 (J0^IL-1GM-CSPLYMPHOIDIL-3GIM -CS,IL(4IL 9 i (IL 4)ai^I \spl? 1^1aw^0 li I*gp i^4ai ('‘G-CSPOfeCil•• IL-3.„.4GM-CSP(11;11CSF-1THYMUS^HURSAL EQUIVALENTI (IL-1,6)IL-2,4,7 5.b.,(igPIIL - 1)•2 ° LYMPHATIC TISSUE1 IL-1,2,4,9(IL-7)(IL-4)O G-CSP I/IL-4GM-CSFNEUTROPHILIL-1,8G-CSPGM-CSFMOMOCYTE^BASOPHILMACROPHAGE MAST CELLIL-1, 11=3^IL-1,3.4CSP -1GM -CSP00; ^.5.•VO O,PLATELETSIL-I,2,6.7IL-8IL-1,4.'x .IL-1,2,3,4,5,6•RED BLOODCELLSEOSIMOPHILIL-1,3,5G14-CSI,Figure 1. A schematic of the hemopoietic hierarchy. Also shown are some of thehemopoietic growth factors and where they are thought to act (fromreference 32)23and high proliferative potential.3 In the murine system, the existence of totipotent,lymphomyelopoietic stem cells was shown in experiments examining the long term hemopoieticreconstitution of lethally irradiated mice by genetically marked bone marrow cells:4,6 Inhumans, evidence for a similar cell, with both lymphoid and myeloid potential, derivesprincipally from studies of patients with hematologic disorders. For example, the Philadelphiachromosome (Ph I), a diagnostic, cytogenetic abnormality associated with chronic myelogenousleukemia (CML), is observed in both myeloid and B-lymphoid tissues of most CML patients. 6 '7In addition, examination of the restriction fragment length polymorphisms within the X-linkedhypoxanthine phosphoribosyl transferase gene of a patient after allogeneic bone marrowtransplantation revealed long term monoclonal hemopoiesis of donor origin. 8Since stem cells are morphologically unremarkable, the study of stem cells has dependedon assays based on their biological characteristics. The first technique for measuring murinepluripotent stem cells, the spleen colony assay, 9 was based on the finding that intravenousinjection of bone marrow cells into lethally irradiated, histocompatible mice, gave rise tomacroscopic splenic nodules of hemopoietic tissue. Cytological l° and chromosomal 1 1markers showed that each nodule was derived from a single pluripotent cell. However,although some of the cells (referred to as CFU-S or Colony Forming Unit-Spleen) detected bythis assay appear to be capable of hemopoietic reconstitution of lethally irradiated mice, severallines of evidence suggest that most cells detected by spleen colony assays are not capable oflong term hemopoietic repopulation and thus are not equivalent to the most primitive toti-potent stem cell. 12 A more recent assay for this more primitive cell has been developed basedon competitive long term reconstitution of lethally irradiated mice. 13 Similar in vivo assaysfor human stem cells are obviously not available. However, two in vitro assays have beenreported that detects very primitive human hemopoietic cells with stem cell properties. 14,16With these assays, various questions have been examined. Of particular interest is whatinfluences stem cell self-renewal versus differentiation. Two theories regarding how thisdecision is made have been suggested. 16 In the instructive model, cells become committed to4a particular lineage as a consequence of instructive, microenvironmental signals. In contrast,the stochastic model proposes that commitment to differentiation is a random processdetermined by probabilities intrinsic to the stem cell. According to this second model, externalregulatory factors merely permit the growth and amplification of a committed cell. Indeed,evidence for this random commitment of progenitor cells has come from experiments in whichpaired daughter cells are separated by micromanipulation. 17 Despite being placed intoidentical culture conditions, the daughter cells develop independently down different lineagesin a stochastic manner. However, aspects of the first, deterministic, model may also be correctsince extrinsic factors may change the stem cell's intrinsic probability of commitment to aspecific lineage.'In contrast to stem cells, the committed progenitors within the next compartment are inactive cell cycle. In addition, these more developed progenitors have a diminished capacity forself-renewal and proliferation and are committed to specific lineages. However, like stem cells,progenitors do not possess morphologically distinct characteristics. They can only be identifiedby the differentiated progeny they produce when cultured in vitro. To assay for progenitors,3cells are suspended in semi-solid medium (such as agar, methylcellulose or plasma clot)supplemented with serum and hemopoietic growth factors (see below). The lineage specificityof a progenitor can then be recognized by the colonies of morphologically recognizable progenythey produce. 3 A progenitor that produces a colony consisting of granulocytes andmacrophages is termed a colony forming unit-granulocyte/macrophage (CFU-GM). Similarly,progenitors giving rise to granulocytes, macrophages or megakaryocytes are called CFU-G,CFU-M or CFU-Meg respectively. Three types of erythroid progenitors have also been defined 18based on colony size, composition and time of maturation. The most primitive progenitorsproduce multi-clustered erythroid colonies, or 'bursts" and are called BFU-E (burst formingunit erythroid). These are further classified as either primitive of mature depending on thenumber of colonies (>8 or 3-8 respectively) in the burst. Progenitors which give rise to single orpaired clusters of erythroblasts are called CFU-erythroid (CFU-E). In addition to these5unipotent or bipotent myeloid progenitors, cells have been detected which produce mixedcolonies of granulocytes, erythrocytes, macrophages and megakaryocytes (CFU-GEMM). 18Lymphoid progenitors have also been described. 19The committed progenitors differentiate further to produce cells of the quantitativelylargest hemopoietic compartment. These cells are morphologically recognizable and consist oftwo classes. The first contains precursor cells, such as myeloblasts and normoblasts, whichhave restricted proliferative potential. These terminally differentiate to produce the secondclass of fully differentiated, mature cells which are then released into the circulation to carryout their designated functions.2. Control of HemopoiesisDuring normal hemopoiesis, the levels of each cell type is maintained within strict limits.However, during stresses such as blood loss or infection, the hemopoietic system rapidlyresponds by producing the necessary cells. This exquisite regulation is mediated by twointeracting systems of stromal and humoral control.2.1 Stromal control of hemopoiesisHemopoietic cells exist in intimate contact with a specialized microenvironmentconsisting of stromal cells and extracellular matrix (ECM) molecules. The stromal cellpopulation includes fibroblasts, endothelial cells, adipocytes, osteoblasts and osteoclasts. 16Fixed marrow macrophages, although they are derived from hemopoietic stem cells, are alsoconsidered part of the stroma. These stromal cells produce and maintain a lattice-work ofextracellular proteins which include collagen types I and IV, glycosaminoglycans, fibronectin,laminin16 and hemonectin. 2° Hemopoietic cells nestle in this network and bind to theextracellular matrix through specific adhesion receptors. 21The contribution of the stroma to hemopoietic regulation is suggested by a number ofobservations. Firstly, the cellular composition of spleen cell colonies varies with their locationin the spleen. Granulocytic colonies tend to form along splenic trabeculae while erythroid ormixed lineage colonies localize to the splenic capsule. 22 Secondly, examinationof bone marrow ultrastructure, similarly revealed specific localizations for certain cells such asCFU-GEMM.23 Evidence for the supportive nature of the stoma is also observed in vitro.When bone marrow cells are cultured for extended periods of time, stromal cells establish anadherent layer of cells and ECM at the bottom of the long term culture dish. Hemopoietic cellscan be maintained on this stromal layer and differential localization of primitive progenitors onthis layer has been observed. 24 In addition, the mIL-3 dependent cell line, FDCP-mix, willproliferate and differentiate when grown on an irradiated stromal layer in the absence of anyexogenously added growth factor. 25One of the ways by which stromal cells might influence hemopoiesis is through localproduction of humoral factors. In fact, as discussed in the next section, stromal cells such asfibroblasts, endothelial cells and macrophages can produce soluble growth regulatorymolecules. Intriguingly, the manner in which the stroma presents these factors may be animportant aspect of stromal regulation of hemopoiesis. Some humoral factors such asinterleukin-3 (IL-3), 26 granulocyte/macrophage colony stimulating factor (GM-CSF) 27 andleukemia inhibitory factor (LIF)28 are bound and sequestered by glycosaminoglycans in theECM. The binding of factors by the ECM may serve to store, localize or regulate the supply offactors that act on hemapoietic cells. In addition, at least two growth factors which wereoriginally characterized as soluble molecules, macrophage colony stimulating factor (M-CSF) 29and Steel factor (SF) 3° are also synthesized as bioactive, membrane bound forms. Importantly,at least in the case for SF, the membrane bound form is absolutely essential for normalhemopoietic development, ie., mice with a Steel mutation (Sid) that results in the sequencesencoding the transmembrane and intracellular regions to be deleted suffer from macrocyticanemia even though these mice produce soluble SF. 3° Melanogenesis and gametogenesis arealso affected since SF is involved in these pathways during development as well. Thesefindings indicate that one of the roles of the stroma is to present SF, and perhaps othermolecules, as a cell bound protein. However, it is important to keep in mind that the stomamay also serve other purposes which are not yet as well characterized. For example, the7binding of hemopoietic cells by the stroma may limit access to soluble factors, or regulate therelease of cells into the circulation.2.2 Humoral control of hemopoiesisThe initial identification of humoral factors that could modulate hemopoiesis came fromstudies of the in vitro culture requirements of hemopoietic cells. Hemopoietic cells will notgrow, or form colonies of recognizable cells as described above, unless the culture issupplemented with conditioned media from a variety of tissues or cell lines. 31 Biochemicalpurification of these colony stimulating activities (CSF's, Table 1) led to the identification ofglycoproteins defined by their ability to support the growth of certain colony types.Subsequently, many more cytokines (Table 1) which act on the hemopoietic system, but do notnecessarily support colony formation, were discovered using different types of assays.Importantly, a number of negative regulatory proteins have also been identified (Table 2).As can be seen from tables 1 and 2, a large number of factors act on hemopoietic cells.This number continues to increase as new sources are tested for activities or new assays foranalyzing hemopoietic cells are developed. None of these cytokines, except for 1L-6, G-CSF andLIF, share significant primary sequence homology. However, the cytokines that act on thehemopoietic system can be summarized as follows:1. They are all glycoproteins. The carbohydrate is typically not required for bioactivity,but glycosylation does contribute to in vivo stability.2. They are extremely potent, with activities in the pM range. The notable exception isSF, which as discussed above, probably exists physiologically as a membrane boundform.3. Many factors, especially IL-3, are not lineage restricted and act on many cell types.As a general rule of thumb, the earlier discovered factors (the lower numberedinterleukins) have the widest range of activities.4. Many cytokines have overlapping activities. However, their in vivo functions may bequite distinct. For example, IL-3 and GM-CSF both support the growth of8granulocyte/macrophage precursors, but whereas IL-3 is produced during animmune response by activated T cells, GM-CSF may be produced constitutively bystromal cells.5. Most cells respond to more than one cytokine. The ability of most hemopoietic cellsto respond to more than one factor is the basis of the intricate cytokine networksobserved in vivo43 which confers great complexity and specificity to the regulation ofhemopoiesis.Not listed in Tables 1 and 2 are indirect acting factors, such as platelet derived growth factor(PDGF), which stimulate stromal cells to produce factors, such as IL-1, which then act directlyon hemopoietic cells.Murine interleukin-3 (mIL-3), the subject of this thesis, is a 14-30 IUD glycoprotein, thatis produced by activated T cells" or IgE cross-linked mast cells. 45 Recently the mRNA formIL-3 was detected in neural tissue, 46 however, whether protein is actually produced is notclear. Murine IL-3 supports the proliferation and differentiation of pluripotent stem cells aswell as a variety of committed myeloid and lymphoid progenitors. The action of mIL-3 onprimitive cells, however, is restricted to those cells which are already in cycle; thus the abilityof IL-6 and SF to bring quiescent cells out of Go is believed to be the basis of the synergyobserved between mIL-3 and IL-647 or SF48,49 in many systems. Murine IL-3 also exertsactions on more mature cells. For example, it has been shown to induce thy-1 antigenexpression on committed myeloid cells 51 and stimulate mature macrophage cytoxicity. 50Since the physiological sources of mIL-3 are likely restricted to activated mast and Tcells, this factor is probably not involved in the regulation of normal, steady state hemopoiesis.In fact, of the growth factors listed in Table 1, only GM-CSF, SF, IL-7 and CSF-1 have beenshown to be constitutively produced. As noted above, GM-CSF, SF and CSF-1 are intimatelyassociated with the stoma as membrane or ECM bound forms. These observations suggest thatsteady state hemopoiesis is controlled predominantly by stromal elements while Immoralfactors become important during times of hematologic or immunologic stress.Table 1. Hemopoietic Growth FactorsFactor^ Abbreviation^Cellular Source^Cellular Targetserythropoietin^ Ep^renal peritubular cells^erythroklgranulocyte CSF G-CSF^macrophages,^ neutrophilsendothelial cellsfibroblastsgranulocyte/^ GM-CSF^activated T cells^granulocytesmacrophage CSF endothelial cells macrophagesfibroblasts early progenitorsmacrophage CSF^M-CSF^fibroblasts^ monocyte/macrophages(CSF-1) endothelial cellsleukemia inhibitory^LIF^stromal cells early progenitorsfactorsteel factor^ SF^fibroblasts^ mast cellsearly progenitorsinterleukin-1 IL-1^macrophages co-stimulator of T cells and progeni .interleukin-2^ IL-2 activated T cells^B and T cellsinterleukin-3 IL-3^activated T cells most myeloid lineagesactivated mast cellsinterleukin-4^ IL-4^activated T cells^B,T and mast cellsinterleukin-5 IL-5^activated T cells B and T cellseosinophilsinterleukin-6^ IL-6^activated T cells^B and T cellsearly progenitorsinterleukin-7 IL-7^stromal cells T and pro B cellsinterleukin-9^ IL-9^activated T cells^myeloid cellsprogenitorsinterleukin-10 IL-10^activated T cells mast, B cellsinterleukin-11^ IL-11^stromal cells^ granulocytes, macrophagesmegakaryocytesinterleukin-12 IL-12^activated B cells T cellsreferences: 3 .32-38Table 2. Hemopoietic Inhibitory FactorsFactor^ Abbreviation^Cellular Source^Cellular Targetsinterleukin-1receptor antagonistmacrophage inflammatoryprotein-1 atumour growth factor Pinterferon yIL-lra^macrophages^IL-1 responsive cellsMIP-la^macrophages^early progenitorsTGFI3^platelets^myeloid cellsearly progenitorsIFNy T cells most mature cell typesReferences: 39-4210B. HEMOPOIETIC GROWTH FACTOR RECEPTORSMurine IL-3, as described above, possesses a wide range of activities depending on theresponding cell. The activity that we chose to investigate was the ability of mIL-3 to supportproliferation. Since the first step in growth factor action is the ligand induced activation of itscell surface receptor, the first step in studying mIL-3 signal transduction is the study of itsreceptor.At the time the studies described in this thesis were initiated, the colony stimulatingfactor-1 receptor(CSF-1R) and c-kit were the only hemopoietic receptors which had beencharacterized at the molecular level. Both of these are receptors with intrinsic tyrosine kinaseactivity. During the past two years, however, many hemopoietic growth factor receptors havebeen cloned and it has become apparent that although the hemopoietic growth factors are notrelated in primary structure, many of their receptors are. Indeed, many of the cytokinereceptors are now classed together in the hemopoietin receptor superfamily. Thus, the twomajor classes of receptors present on hemopoietic cells are the intrinsic tyrosine kinasereceptors and the members of the hemopoietin receptor superfamily.1. Intrinsic Tyrosine Kinase ReceptorsMany cell surface receptors resemble allosteric enzymes with an external regulatory,ligand binding domain and an intracellular, catalytic domain. Examples of these receptorsinclude the activin52 and the Drosophila, daf-1 53 receptors which contain an intrinsic serinekinase and the atrial natriuretic receptor54 which harbours an intrinsic guanylate cyclaseactivity. However, the most common and best studied receptors with intrinsic enzymic activityare those with tyrosine kinase activity. These include the receptors for EGF, PDGF andinsulin. Among the hemopoietic growth factors, at least three utilize tyrosine kinase receptors.1.1 Structure of tyrosine kinase receptorsTyrosine kinase receptors have been well characterized because many of the earliestdiscovered polypeptide growth factors, such as epidermal growth factor (EGF), PDGF andinsulin, possess receptors of this type. Based on structural features, all tyrosine kinase11receptors can be clarified into three groups (Figure 2). 55 The first group consists of receptorswith only a single polypeptide chain. The extracellular, ligand binding domain of thesereceptors contain two cysteine rich repeats. The EGF receptor (EGFR), the neu proto-oncogeneand the Drosophila sevenless protein are examples of this first group. In addition, the productof the erb-B2 oncogene encodes a truncated version of the EGFR The second class of receptorsinclude those for insulin and insulin-like growth factor 1 (IGF-1). These receptors areheterotetrameric with two a and 13 disulfide bonded subunits. The a subunits compose theligand binding site, and each contain one of the cysteine rich repeats also found in the type Ireceptors, while the kinase domain is located on the f chains. The PDGF receptor (PDGFR)typifies the third class of receptors. These receptors are single chain polypeptides withimmunoglobulin (Ig) like repeats in the external domain. Comparison of the catalytic domain ofthis class of receptors with those of the other two, revealed that an insert region separates theATP binding site from the phosphotransferase domain. The sequence of this kinase insert (KI)domain is quite heterogeneous for differ ent growth factor receptors in this class. However, forthe same receptor, the KI region is very well conserved among different species. 56 Othermembers of this last class of tyrosine kinase containing receptors include the hemopoieticreceptors CSF-1R, 57 c-kit58 andflk-2. 59The CSF-1R, expressed on cells of the mononuclear phagocyte lineage, 60 is the cellularhomologue of the transforming gene of the McDonough strain of feline sarcoma virus, v-fms. 61The activated v-fins differs from the cellular gene, c-fins, by having a point mutation in theextracellular domain and a short cytoplasmic truncation. 62 Both of these alterationscontribute to the constitutively activated kinase activity, and thus the transforming potential ofv-fms. The c-kit protein is the receptor for SF. 63 Approximately 10% of all bone marrow cellsand most hemopoietic progenitors express c-kit. 58 A transforming version of c-kit, v-kit, isfound in the Hardy-Zuckerman-4 strain of feline sarcoma virus and differs from c-kit bydeletions within both the ligand binding and transmembrane region. The c-kit gene is allelicwith the murine W (white spotting) locus. Mice with mutations at the W locus exhibit12I6EGF-R^I-R ^PDGF - Rneu/HER2^IGF - 1 - R^c-fms/CSF-14.):c.k:tFigure 2. The three classes of tyrosine kinase containing receptors. Shaded boxesrepresent the cysteine rich domains; clear boxes designate the kinasedomains, which are split by an insert region in receptor class III. Thecircles represent the conserved cysteines in the Ig domains of the thirdreceptor. (from reference 55)13defects in hemopoiesis, melanogenesis and gametogenesis; these defects resemble thoseobserved in mice with mutations at the Steel locus. Complementation analysis withreconstitution experiments and finally with the purification of the Steel gene product63 showedthat the c-kit protein is the receptor for the Steel ligand. The most recently discovered tyrosinekinase receptor in hemopoietic cells isflk-2. Unlike, c-frns or c-kit, j7k-2 was not discovered byits homology to any viral oncogenes. Instead,flk-2 was identified during a deliberate search fortyrosine kinases that are expressed in hemopoietic cells. 59 Although, ilk-2 has only beencharacterized to date at the DNA level, the fact that its expression is restricted to stem cell andprogenitor populations makes it and its, as yet, unknown ligand potentially important inhemopoietic regulation.1.2 Signalling from receptors with intrinsic tyrosine kinase activityFrom extensive studies with the classical tyrosine kinase receptors, ie., the EGFR,insulinR and PDGFR, a model by which these receptors transduce signals across themembrane has been proposed. Specifically, the binding of a ligand to the external domainresults in the activation of the cytoplasmic tyrosine kinase domain. 64 This activation, whichreflects an increase in the Vmax of the phosphorylation reaction, 65 is an essential step in thereceptor signalling since receptors with defective kinase domains can not signal. 55,66Accompanying receptor activation is autophosphorylation of the receptor itself.In an early model of receptor action, the binding of a ligand to the external domain wasthought to induce a conformational change that is transmitted through the transmembraneregion to activate the kinase. However studies with chimeric receptors 67,68 or receptors withaltered transmembrane domains 69 suggested that the transmembrane segment does notfunction in this manner. A large body of evidence now supports a mechanism in whichreceptor activation occurs as a result of ligand induced oligomerization. For example, certainanti-insulinR antibodies mimic the effects of insulin and this insulinometic property dependson antibody divalency. 7° Also, the isolated ectodomains of the EGFR have been observed toundergo dimerization upon the addition of EGF.71,72 In the case of the insulin and EGF14receptors, dimerization may be the result of ligand induced conformational changes. However,the ligands for PDGFR, CSF-1R and c-kit are dimeric so one ligand molecule may bridge andthus bring together two receptor molecules. 73-76 The involvement of receptor clustering hasbeen proposed for the basis of the constitutive kinase activity of the oncogenic version of neu.Oncogenic neu differs from the normal cellular version by a point mutation in thetransmembrane domain and this mutation apparently causes oncogenic neu to exist asaggregates in the absence of ligand stimulation. 76One result of receptor clustering is the transphosphorylation of other receptormolecules.77 Indeed, the demonstration that receptor autophosphorylation is primarily anintermolecular,78 .79 rather than an intramolecular, reaction further strengthens thedimerization model of receptor activation. Recent studies suggest that receptorautophosphorylation is a key event in receptor signalling. Several proteins implicated in signaltransduction, including the p74 ra, phospholipase Cy (PLCy), phosphatidylinositol 3-kinase(PI3-K) and ras GTPase activating protein (GAP), have been shown to physic-ally associate withtyrosine phosphorylated receptors. 80-82 The function of these proteins in signal transductionwill be discussed in later sections. Since the initial demonstration that p74raf binds the PDGFreceptor,83 the number of reports showing association of receptor and signalling proteins hasgrown exponentially (Table 3).Table 3. Molecules Associated with Intrinsic Tyrosine Kinase ReceptorsPI3-K GAP PLCy p74rafEGFR + + +PDGFR + + + +InsulinR + ?CSF- 1R + + _c-kit + ? + ?References in text. Association is designated with a + or - sign. The ? refers to caseswhere association is not known.15Intriguingly, with the exception of p74 ra, the proteins which bind to intrinsic tyrosinekinase receptors all contain SH2 domains (Src Homology region 2). 84 SH2 domains areapproximately 100 amino acids in length and were first described as regions of homology incytoplasmic tyrosine kinases such as src and were subsequently found in other, non-kinaseproteins involved in signal transduction. The SH2 domains of PLCy, PI3-K and GAP bind tospecific phosphotyrosines in the appropriate receptor. 80,84,85 PLCy and GAP each have twoadjacent SH2 domains and these act synergistically in binding. 86 The raf polypeptide does nothave an SH2 motif, so it either associates with a protein that does contain an SH2 domain orbinds receptors through an alternate mechanism. It is important to note, however, thatalthough many early intermolecular interactions associated with signal transduction aremediated through SH2 domains, not all may be. For example, GAP has been shown toassociate with 190 and 62 Id) tyrosine phosphorylated proteins in mitogen activated ortransformed cells. 84 Although pp62 binds to the GAP SH2 motif, pp 190 binds some otherregion in the molecule. 85 In addition, although PLC), binds tightly to tyrosine phosphorylatedPDGFR's through its SH2 domains, PLCy is loosely associated, presumably through an SH2independent mechanism, with the receptor prior to activation (see below).The structural requirements in the receptor for the binding of signalling proteins has alsobeen studied. Phosphorylation of tyr-706 in the kinase insert region of the CSF-1R isimportant for PI3-K binding. Similarly, tyr-857 in the PDGFR appears to be specific for GAPassociation. 87 Interestingly, deletion of tyr-809 in the human CSF-1R specifically uncouplesthe CSF-1R to the pathway leading to the elevation of c-myc gene expression and impairs themitogenic response to CSF-1. This suggests that a protein specific to the signalling pathwayleading to c-myc elevation associates with the CSF-1R through tyr-809 88 since all other earlycellular signalling pathways appear to be intact. This finding also illustrates the importance ofthe discovery that receptor tyrosine phosphorylation followed by association with other proteins16is a key event in signal transduction. Namely, mutagenesis of specific tyrosines in the receptormay allow identification of proteins which link receptors with intracellular signalling pathways.2. The Hemopoietin Receptor SuperfamilyUnlike the receptors with intrinsic tyrosine kinases, the other receptors on hemopoieticcells are normally present only in low numbers. Because of this low abundance, most of themembers of the hemopoietin superfamily, with the exception of the mIL-3R (this work) and theIL-411.89,90 were isolated using cDNA cloning and COS cell protein expression systems. 91 Inthis method, COS cells are transfected with pools of cDNA's from an appropriate source andthe COS cells allowed to express the proteins encoded by the cDNA. Pools that contain thecDNA coding for the receptor of interest are identified by screening the transfected COS cellswith either an anti-receptor antibody, if available, or with labelled ligands. The human IL-6R asubunit was the first to be cloned in this manner. The EpR and IL-2R (3 chain were isolated acouple of years later. In the past year, however, more than 10 hemopoietin superfamilyreceptors have been cloned (Figure 3C). Thus considerable information regarding the structureof this family of receptors has recently been accumulated. However, knowledge of thesignalling mechanisms utilized by these receptors is still at a nascent stage.2.1 Structure of the hemopoietin receptorsSome of these receptors consist of two subunits, one or both of which may belong to thehemopoietin superfamily, while others only have one. The common feature of all members ofthe superfamily is a 210 amino acid module in the extracellular domain. 92 This module(Figure 3A) is characterized by a W-S-X-S-W (single letter codes, X=any amino acid) sequencenear the C-terminal and a motif of four spatially conserved cysteinesCX9-20-CXWX22_36CX8_25C near the N-terminal. A fibronectin type III domain is also foundin the C-terminal region which has led to speculation that these receptors are evolutionarilyderived from primitive adhesion molecules. 93 From an analysis of the primary amino acidsequence, Bazan94 has suggested that the extracellular hemopoietin domains fold into a17A.^ B. cysteine motif—wsxwsLIFC.Structure of the hemopoietin receptors. The hemopoietin consensusdomain is shown in 3A. Figure 3B shows the barrel model for receptorfolding; the black box represents a ligand.94 The members of thehemopoietin receptor superfamily are displayed in 3C. In Fig. 3C, thestippled segment represents a contactin domain.Figure 3.18double barrel configuration, with each barrel consisting of seven (3 strands (Figure 38). Theligand binding site is predicted to be located in the crevice between the two barrels. This modelhas been substantiated, to some extent, by the finding that replacement of amino acid residuesin the putative hinge region between the two barrels, greatly diminishes ligand binding. 95 TheWSXWS motif has also been shown to be important in ligand binding. 96Analysis of the ligand binding characteristics of the hemopoietin receptors in native,hemopoietic cells revealed that while some, ie. the G-CSF receptor, exhibit only one highaffinity class, others, such as the Ep receptor, exhibit both high and low affinity binding.However, not all of the cloned receptors, when expressed in an ectopic cell such as thefibroblast COS cell, demonstrate the same ligand binding characteristics as the intact receptoron hemopoietic cells. This anomaly suggests that either, the receptor protein is modified inhemopoietic cells in a manner that can not occur in non-hemopoietic cells or that the receptorcontains additional protein subunits present in hemopoietic cells which are absent in COScells. Thus the hemopoietin receptors that have been cloned, thus far, can be divided into twogroups based on whether the cloned molecules display native ligand binding characteristics.Examples of receptors for which all of the components contributing to ligand binding have beenisolated include the receptors for G-CSF, GM-CSF, IL-4, IL-5, IL-6, IL-7 and human IL-3. Incontrast, the receptor structures for erythropoietin (Ep), leukemia inhibitory factor (LIF) IL-2and murine IL-3 are still under investigation. The following section summarizes ourknowledge, to date, of the hemopoietin receptors. The receptors for which the structure hasbeen "solved", including the ones for G-CSF, IL-4, IL-6 and IL-7, will be discussed first.Following these will be the "unsolved" receptors for erythropoietin, leukemia inhibitory factorand IL-2. Finally, the murine interleukin-3 receptor will be discussed together along with themurine IL-5, GM-CSF and human IL-3, GM-CSF and IL-5 receptors. These last receptors aretreated together, even though the murine IL-3R is the only one of the group for which thenative structure is not yet known, because they are inextricably linked by the way they werecloned.192.1.1 Hemopoietin receptors with solved structuresGranulocyte-colony stimulating factor receptor (G-CSFR). Although normalhemopoietic cells and cell lines express a single high affinity class (120 pM) of G-CSFR's,solubilization of this receptor generates an additional low (3 nM) affinity class. However,sucrose density gradient centrifugation has shown that the high affinity receptors are dimerswhich dissociate to give low affinity binding monomers. 97 The mouse 130 kD G-CSFR wascloned by expression with radiolabelled ligand and expresses high affinity binding whenexpressed in COS cells. 98 In addition to the typical hemopoietin receptor domain, the G-CSFRalso has an N-terminal, Ig like domain which is not involved in G-CSF binding or biologicalactivity99 and a contactin domain near the transmembrane region. A related human proteinwith 63% amino acid identity has been isolated. Interestingly, in human cells. two mRNAsplice variants of the G-CSFR exist. One contains a 27 amino acid insertion in the cytoplasmicdomain which may alter its signalling characteristics, while the second encodes a truncatedform of the receptor that retains only the N-terminal ligand binding domain. In fact, the G-CSFR is only one of many hemopoietin receptors that can exist as soluble molecules.Interleukin-4 receptor (IL-4R). Only one high (20-80 pM) affinity class of IL-4R hasbeen observed on hemopoietic cells. 113° Initial cross-linking experiments suggested that 1251labelled IL-4 bound to two proteins with apparent molecular weights of 80 and 150 kD.However, the lower molecular weight form was shown to be a proteolytic fragment of the largermurine 102,103 and humanlyspecies. 1° 1^IL-4R's have been isolated which possess 50%homology at the amino acid level. Analysis of the requirements for signalling demonstratedthat the terminal 400 amino acids of the cytoplasmic domain can be deleted with no effect onactivity.92 Like the G-CSFR, a mRNA species that encodes a soluble IL-4 binding domain hasbeen described. 102Interleukin-6 receptor (IL-6R). Both subunits of the IL-6R belong to the hemopoietinreceptor superfamily. The 80 kD a chain has an N-terminal Ig domain in addition to thehemopoietin receptor module and binds IL-6 with low affinity. 104 The 130 kD 13 subunit20resembles the G-CSFR, with one copy of the hemopoietin domain and a membrane proximalcontactin domain. 105 Reconstitution of the a with the 0 chain produces high affinitybinding. m6 Interestingly, the a and f3 chains do not exist as preformed complexes and onlyassociate, through interactions of their extracellular domains, in the presence of ligand. 1°6Interleukin-7 receptor (IL-7R). Unlike the G-CSF and IL-4 receptors, the IL-7R exhibitstwo affinity classes on hemopoietic cells. Since the cloned 80 kD murine protein 90 alsoproduces two affinity classes when expressed in COS cells, the high affinity class is thought tobe a result of dimerization. The human and mouse receptors share 64% amino acid homology;a soluble form of the IL-7R has also been detected. 9°2.1.2 Hemopoietin receptors with unsolved structuresErythropoietin receptor (EpR). The cDNA encoding a 65 kD murine EpR was clonedfrom an erythroleukemia cell line 107 and a human homologue was isolated which is 82%identical to the murine protein at the amino acid leve1. 108 Although the EpR is the beststudied of the hemopoietin receptors, probably because it was one of the earliest to becloned, 107 there is some disagreement in the literature as to whether there are one or twoaffinity classes of receptors. 109 When expressed in fibroblasts, the EpR exhibits one highaffinity (200 pM) class. However, chemical cross-linking studies with 125I-Ep have revealedtwo bands, one corresponding to the 65 kD cloned EpR polypeptide, and another with anapparent weight of 100 kD. Cross-linking experiments with normal EpR bearing cells alsoyields similar, though not identical, bands. These experiments suggests that the EpR complexmay contain additional proteins.Mutagenesis of the cytoplasmic domain of the EpR has mapped a region likely to beimportant for mitogenesis to a membrane proximal region. 111 Provocatively, this region has ahigh degree of homology with the domain of the IL-2R [3 subunit 112 that has been shown to beimportant in signalling.Leukemia inhibitory factor receptor (LIFR). The native LIFR displays low (1 nM) andhigh affinity (400 pM) binding. Since the 80 kD, LIF binding protein that was cloned binds LIF21with low affinity when expressed in COS cells, a second subunit has been postulated that isrequired for high affinity binding. 113 The 8010 protein has two copies of the hemopoietindomain, and like the G-CSFR and IL-6R 0 chain, has a contactin domain near thetransmembrane region. In fact the G-CSFR the 8010 LIFR and the IL-6 0 chain are the mostclosely related among the members of the hemopoietin receptor superfamily. This relationshipis even more intriguing given the fact that the ligands for these three receptors also share somesequence similarity and exhibit similar spectra of activity on hemopoietic cells. 114Interleukin-2 receptor (IL-2R). From a comparison of the IL-2 binding characteristicsof activated and resting T cells, it has long been postulated that the IL-2R consists of twosubunits.92 The 55 kd a subunit was cloned early in the history of cytokine receptors becauseof the availability of the Tac antibody and was shown to bind IL-2 with low affinity (10 nM)when expressed in COS cells. This 55 kD subunit does not belong to the hemopoietin receptorfamily and is not present on resting T cells. The other subunit, called the p subunit, isconstitutively produced on T cells and binds TL-2 with intermediate affinity. Activation of Tcells induces expression of the a subunit and together the two form a complex that binds IL-2with high affinity (10 pM). Subsequently, a 75 kd 13 subunit was isolated. Unexpectedly, whenthis 0 chain was expressed in COS cells, it bound IL-2 with very low affinity (100 nM) and notwith the expected intermediate affinity. However, when the 0 subunit was expressed in anoligodendroglioma cell line, it exhibited the expected intermediate affinity (2 nM) bindingcharacteristics. These results have led to the speculation that the IL-2R actually consists ofthree subunits, ie., in resting T cells, the 0 exists as a complex with the putative y subunit andtogether the two are responsible for intermediate affinity binding. This y chain is absent fromCOS cells but is present in oligodendroglioma cells. Biochemical evidence for a 100 kD y chainhas been reported. 115,1162.1.3 The receptors for interleukin-3, GM-CSF and interleukin-5 Maine interleukin-3 receptor (mIL-3R). The mIL-3R is qualitatively very differentfrom the human IL-3R (hIL-3R). For example, although mIL-3 is able to downmodulate surface22expression of the murine GM-CSF and G-CSF receptors, neither mGM-CSF, mG-CSF or otherligands can compete with mIL-3 for binding to the mIL-3R 117 In contrast, in the humansystem, IL-3, IL-5 and GM-CSF compete with each other for binding to their receptors. 92,115An mIL-3 binding protein was cloned by COS cell expression 119 using an antibody thatwas believed to recognize the mIL-3R because of its ability to partially inhibit mIL-3 binding tomIL-3R expressing cells. 12° This 120 kD mIL-3 binding protein, called Aic 2A, has two copiesof the hemopoietin receptor domain and when expressed in COS cells, binds mIL-3 with lowunity 119 Because of the low affinity binding, a second subunit has been hypothesized toexist.92 Interestingly, a second protein, called Aic 2B, with 91% sequence identity to Aic 2Awas isolated during the cloning of Aic 2A. 121 This second protein is expressed in similaramounts to Aic 2A in all mIL-3 dependent cells tested. However, when expressed by itself inCOS cells, Aic 2B does not bind mIL-3 or any other known ligand.Murine interleukin-5 and GM-CSF receptors (mIL-5R and mGM-CSFR). These tworeceptors are discussed together since they share a common subunit responsible for highaffinity binding. Both mIL-5 and mGM-CSF bind to normal hemopoietic cells with high andlow affinities.92 In 1990, a 60 Kd mIL-5 binding protein was cloned 122 which binds mIL-5with low affinity when expressed in COS cells. This protein is a member of the hemopoietinsuperfamily. However, when this 60 kD protein was expressed in an mIL-3 dependent cellwhich normally does not bind mIL-5, high and low affinity mIL-5 affinity binding wasobserved. This suggested that a protein present in the mIL-3 dependent line could be thesecond subunit of the mIL-5R complex. This theory was confirmed when high and low affinitymIL-5 binding was reconstituted by co-expression, of the 60 10 mIL-5 binding protein and Aic2B in COS cells. 106,123 Several lines of evidence suggest that Aic 2B also serves as the f3subunit for the mGM-CSFR. 124Human interleukin-3, interleukin-5 and GM-CSF receptors. In an attempt to clonethe hIL-3R, a human cDNA library was screened with probes based on the mIL-3 bindingprotein, Me 2A. Only one protein was isolated, 125 but it did not bind any known factor when23expressed in COS cells. However, when co-expressed with the previously described low affinityreceptors for hIL-5 126 or hGM-CSF, 127 high affinity binding of the appropriate ligand was126,127observed.^Thus the human IL-5 and GM-CSF receptors share a common subunit thatconfers high affinity binding to the low affinity, ligand specific receptor subunits. Solubleforms of the hIL-5 126 and hGM-CSF 128 a subunits have also been described.It had long been speculated that the receptors for human IL-3 and GM-CSF shared somecommon receptor component since their ligands showed a complex pattern of cross-competition. 118,129 Specifically, hIL-3 displayed only one class of high affinity binding sitesand hGM-CSF was able to effectively compete with hIL-3 for binding to those sites. In contrast,hGM-CSF had both high and low affinity sites and hIL-3 was able to compete with GM-CSF forbinding to these high affinity sites. From these observations, Kitamura suggested that the hIL-3R complex shares a 13 subunit with the hGM-CSFR and both hGM-CSF and hIL-3 receptorcomplexes have unique ligand specific a subunits. 13° However, although the hGM-CSF asubunit alone can bind to hGM-CSF with low affinity, the hIL-3 a subunit by itself can notbind hIL-3 at all. This model was confirmed with the expression cloning of the hIL-3R asubunit in COS cells. 130 The expression cloning was successful only because, the COS cellswere co-transfected with the cDNA encoding the hGM-CSFR 13 subunit. The hIL-3R a subunit,like the hIL-5 and hGM-CSF a subunits are members of the hemopoietin receptor superfamily.Other members of the hemopoietin superfamily. The u-mpl oncogene of themyeloproliferative murine leukemia virus, is capable of immortalizing hemopoietic progenitorsand resembles a hemopoietin receptor. 131 The ligand for this putative receptor is not known.The prolactin and growth hormone receptors are also considered members of the hemopoietinsuperfamily although they lack the usual W-S-X-W-S motif. 922.2 Signalling from the hemopoietin receptorsStudies of the mechanism by which the hemopoietin receptor family transmits signalsacross the membrane are only just beginning since most of the receptors have only been24recently cloned. However, from the work reported so far, it appears that the hemopoietinreceptors may utilize mechanisms similar to the intrinsic tyrosine kinase receptors.Dimerization, which is important in tyrosine kinase receptor signalling, may also beinvolved in hemopoietin receptor activation. For example, intact anti-prolactin receptorantibodies can mimic the effects of prolactin, but monovalent fragments can not. 132 Secondly,it has recently been shown that one molecule of growth hormone binds to two receptormolecules. 133 Thirdly, an activating point mutation in the EpR extra-cellular domain wasshown to be a replacement of an arginine by a cysteine ll° and analysis of this mutationshowed that replacement of this arginine by any amino acid except for cysteine did not activatethe receptor. Furthermore, non-reducing SDS-PAGE suggested that dimerization throughdisulfide bonding was involved in receptor activation.After dimerization, tyrosine kinase receptors autophosphorylate and become associatedwith signalling proteins. Although members of the hemopoietin receptor family do not haveintrinsic tyrosine kinase domains, many become phosphorylated on tyrosine residues upongand binding. This phosphorylation must be mediated by an associated tyrosine kinase andboth the growth hormone iM and IL-2 receptors have been shown to be associated withtyrosine kinases. 135 Phosphorylation of the IL-2R on tyrosines following ligand binding maybe the signal for the observed association of PI3-K, 136 a molecule better known for itsassociation with intrinsic tyrosine kinase receptors.Therefore, although structurally very unlike the better characterized tyrosine kinasereceptors, the hemopoietin receptors may use many of the same signalling pathways. Thefollowing section discusses some of the proteins described in many systems to be involved insignal transduction.C. PROTEINS INVOLVED IN SIGNAL TRANSDUCTIONThe binding of a growth factor to its cell surface receptor induces many intracellularchanges including: changes in ion fluxes, 137,138 cytoskeletal reorganization,56 and proteinphosphorylations. 139 These early signals lead to changes in gene expression 14° and25eventually culminate with the initiation of DNA synthesis and mitosis. 141 Some of theproteins involved in these changes have been characterized and they are discussed below.1. GTP binding proteinsThe members of the GTP binding protein superfamily, all of which bind and hydrolyzeGTP, are involved in regulating information flow and ensuring fidelity of specificmacromolecular interactions in diverse processes ranging from photoreception and hormoneaction to intracellular vesicle transport and ribosomal function. 142 Two members of thisfamily that are implicated in signal transduction are the heterotrimeric G proteins and the lowmolecular weight, ras related GTP binding proteins. The mechanism of action of theheterotrimeric G proteins are well understood. In contrast, the regulators and function of theras like proteins are not well characterized and our knowledge of ras protein function derivesmostly from comparison of these proteins with the more familiar members of the GTP bindingprotein superfamily.1.1 G proteinsThe term G protein is reserved for the signal transducing GTP binding proteins composedof a 40 - 45 ItD a, a 35 - 36 kD Os, and 8 - 10 kD y subunit. 143 Both the a and y subunits arepost-translationally modified, the a with a myristate and the y with a geranylgeranyl moiety,which appears to mediate the observed association of the trimeric complex with the plasmamembrane or plasma membrane proteins. 144 Almost 20 G proteins have now beendescribed, 145 and they differ from each other primarily in their a subunits. The a subunit,which contains the GTP binding site and GTPase activity, is subject to ADP ribosylation, withdifferent a's being differentially susceptible to ribosylation by Vibrio Cholera and BordetellaPertussis toxins. Originally, the few G proteins that were known, were classified by their toxinsensitivity and by their action on the downstream effector molecules, adenylate cyclase andGMP phosphodiesterase. 146 However, in the wake of the discovery of multiple a's and themultiplicity of effector molecules, 145 G proteins are now categorized by sequence homologieswithin the a subunit.26In the basal state, G proteins exist as a GDP bound, trimeric complex in associationwith a receptor. 145 Upon activation, the receptor stimulates exchange of a bound GDP forGTP and this reduces the affinity of the G-protein for the receptor. In the absence of receptorstimulation, the nucleotide exchange rate is very slow. The f3 and y subunits, which form atightly associated heterodimer, dissociate from the GTP bound, activated a subunit to allowinteraction of the activated a with downstream effector molecules. These effector moleculesinclude diverse proteins such as adenylate cyclases, phospholipases and ion channelcomponents. 147 GTP hydrolysis, catalyzed by an intrinsic GTPase activity, deactivates the asubunit and it reassociates with the (3y subunits. Reassembly is required for reassociation withthe receptor. 145 The fay subunits, important as regulators of a subunit activity, were once alsopostulated to have a direct mediator role. For example, adenylate cyclase inhibitory G proteinswere thought to exert their effects by releasing13ys which would bind to and inactivateadenylate cyclase stimulatory G proteins (G s). 146 However, this model has fallen intodisrepute since inhibition has been shown to occur in G s deficient cells and because inhibitiondoes not exhibit competitive kinetics. 145G proteins have been shown to couple receptors, including receptors for light,chemotactic factors and neurotransmitters, to downstream events in over 100 systems. 145Most of these receptors are opsin proteins which span the membrane seven times. G proteinsinteract with the third cytoplasmic loop of these integral membrane proteins; 148 however,single membrane spanning receptors may also couple to G proteins. 149 The importance of Gproteins in signal transduction is evidenced not only by their wide ranging expression andusage, but also by the discovery that oncogenesis and viral subversion of host cell metabolismcan proceed through G protein pathways. For example, the gsp and gip2 oncogenes implicatedin neuroendocrine tumours are GTPase deficient G protein a subunits. 15° Also, the CMVgenome encodes three seven-times-membrane-spanning, G protein coupled receptors thatmight, when expressed, modify host cell signalling to suppress anti-viral effects.151271.2 The ras related GTP binding proteinsThe mechanism of action of the ras family of GTP binding proteins is less wellunderstood. Ras (rat sarcoma) was first identified as the transforming gene of the Harvey andKirsten rat sarcoma viruses. 152 The ras oncogenes, Ha-ras, K-ras and N-ras, are the mostfrequently encountered oncogenes in human tumours with an overall incidence ofapproximately 30%. The discovery that this viral oncogene was a mutated version of a normalcellular gene suggested that the normal homologue might be important in controllingmitogenesis.Ras encodes a 21 kD GTP binding protein with an intrinsic GTPase activity in its Nterminal functional domain. 153 Transforming activity, and thus presumably normal activity ofthe non-oncogenic form of p21ras requires addition of both a farnesyl and palmitic acid to theC terminal end of the protein. 153 By analogy with the heterotrimeric G proteins, the GTPbound molecule is the active form. Oncogenic ras proteins typically have point mutations atcodons 12, 13 and 59 and these result in both decreased intrinsic GTPase activity andrefractoriness to regulatory proteins which normally enhance the GTPase activity of rasproteins. 154 Ras belongs to a family of approximately 30 members 155457 which includes:rap/ (K-rev-1) which appears to antagonize ras activity; the rho subfamily, which is involved incytoskeletal regulation; the rab's, which may control intracellular vesicle transport and ran,which links completion of DNA synthesis to mitosis. The yeast RAS1 and RAS2 proteins arealso members of this family. Ras itself is implicated in cell growth and signal transductionIL-2,158 IL-3158^159,160 serum ^and insulinsince ^GM-CSF,158 PDGF,^ increase theratio of GTP to GDP bound p2l ras in responsive cells. Ras expression will also inducedifferentiation in rat pheochromocytoma cells 163 and meiotic maturation of Xenopusoocytes. 164In mammalian cells, neither the upstream regulators or the downstream effectormolecules by which p2lras mediates these actions have been well defined. Tyrosine kinasesare apparently upstream of p21ras since tyrosine kinase oncogenes can not transform28p2lrasdeficient cells. 168 Moreover, studies with antibodies to phospholipase Cy (PLCy) suggestp2lras acts upstream of this molecule. 165 Ras expression also been shown to activate akinase which phosphorylates and activates c-jun, a component of the AP-1 transcriptionfactor.' However, no direct connection has been demonstrated for p21ras with any of thesesignalling proteins. In S. cerevisiae, the ras signalling pathway is, perhaps, slightly moredefined. The yeast RAS1 and RAS2 proteins activate adenylate cyclase through a cyclaseassociated protein (CAP) in response to external signals. 167 However, the coupling of p21 rasto adenylate cyclase in yeast has not been conserved in mammalian cells. On the other hand,recent data which link CAP to an actin and phosphatidylinositol 4,5-bisphosphate (PIP 2)binding protein called profilin, suggest certain aspects of ras function have been conserved inevolution. It has long been suspected that regulation of adenylate cyclase is not the solefunction of the yeast ras proteins since RAS deletions are lethal whereas adenylate cyclase nullmutations are not; 167 the discovery that CAP is a bifunctional protein which couples p21ras toprofilin and PIP2 as well as to adenylate cyclase confirmed this suspicion) 68 Importantly, thissecond p21ras effector pathway may have been conserved in higher organisms. As mentionedearlier, PLCy is a downstream molecule in the mammalian168 p2lras pathway, and as will bediscussed below, PLCy is regulated by profilin sequestration of PIP2 in mammalian cells. 169The finding that p2lras action in yeast involves profilin, along with the observation that p2lrasaction in mammalian cells involves PLCy, 165 suggests the intriguing possibility that profilinmay be an intermediate mediator if not a direct effector of p2lras in eucaryotes. Regardless ofthe role of profilin, the discovery that the yeast and mammalian p2lras pathways are moresimilar than initially thought, may allow use of powerful yeast genetics to dissect p21ras actionin mammalian cells.1.3 Regulators of rasBy analogy with better known GTP binding proteins, p2l ras can be regulated byswitching between the GDP and GTP bound state. One report showed that ras becomestyrosine phosphorylated in response to GM-CSF in normal murine hemopoietic progenitor29cells. 170 However the effect of this phosphorylation on ras activation is not known. Bettercharacterized are the factors which affect p2l ras intrinsic GTPase activity and factors thatcatalyze exchange of bound GDP for GTP.1.3.1 GTPase activating protein as a regulator of rasNormal and oncogenic forms of p21ras differ in their ability to hydrolyze GTP. However,the magnitude of this difference can not be accounted for by the slight differences in theirintrinsic GTPase activities. 171 This discrepancy was resolved by the discovery of a GTPaseactivating protein (GAP) that binds to both normal and oncogenic p2lras but is able tostimulate the GTPase activity only of the normal ras protein. 172 GAP, a 120 kD protein, iswidely expressed in all tissues173 and is homologous to the yeast IRA1 and IRA2 (Inhibitoryregulator of ras/cAMP pathway) proteins. 174 GAP's for other members of the ras family havealso been isolated, including GAP's for rapi,175 rh0, 176 ra/ 177and rab 3A178 proteins. Theneurofibromatosis type 1 179,180 and bcr genes 181 also encode functioning GAP activity.Interestingly, a GTPase inhibitory protein (GIP) for ras has also been described. l82 Consistentwith an upstream regulator model of GAP action is the observation that GAP can suppress thetransformation induced by c-ras over-expression. 183 In addition, phorbol ester stimulation ofT cells apparently elevates p2lras-GTP levels by inhibiting GAP activity. 184Interest in GAP was heightened recently by the finding that GAP may provide themolecular basis for the observed connection between tyrosine kinases and p2l ras. GAPbecomes tyrosine phosphorylated in cells stimulated with EGF,185,186 pDGF187,188 ortransformed with fps, 189 src189 or the erbB-2 19° tyrosine kinases and is physicallyassociated with activated EGF,82 PDGF87 and CSF-1 191 receptors. Association of GAP withtyrosine kinases occurs via SH2 interactions as discussed earlier. Over-expression of GAPinhibits src transformation, 192 presumably by acting on p2l ras, since none of the p56sreinduced tyrosine phosphorylations are inhibited. Interestingly, the catalytic C-terminal regionof GAP is more efficient in suppressing src transformation than the full length protein,suggesting that p56sre phosphorylation of the GAP N-terminal domain may regulate GAP's30activity on p2lras,192 However, no in vitro correlation of tyrosine phosphorylation and GAPactivity has been demonstrated. Moreover, in some systems, the stoichiometry of tyrosinephosphorylation is quite low with most of the phosphorylation occurring on serinesresidues. 186 An alternate model of GAP regulation involves sequestering GAP into an inactivecomplex. 186 In mitogen stimulated or tyrosine kinase transformed fibroblasts, GAP becomesassociated with two proteins with molecular weights of 62 and 190 kD. Comparison of freeGAP to the form bound to the 190 kD protein, showed that the complex had diminished rasGTPase stimulating activity. 186 An attractive hypothesis for the function of GAP receptorassociation is a similar sequestration of GAP, thus resulting in accumulation of active p21ras-GTP complexes. This may not be the whole story however. Certain lipids, the levels of whichchange during mitogenic stimulation, like arachidonic acid and phosphatidylinositolphosphate 193 are also capable of inhibiting GAP activity in vitro. The same lipids which inhibitGAP activate GIP. 182 Perhaps these multiple layers of regulation allow fine tuning of p2lrasregulation.Because the function of p2lras is not known, GAP may be actually a target of p2lrasaction instead of, or in addition to, being an upstream p21ras regulator. 171 Genetic andbiochemical evidence have shown that GAP interacts with the effector domain of p2lras; 194regions of p2l ras required for oncogenicity are also those which interact with GAP. 195 Thebest evidence that GAP is a target for p2l ras comes from experiments examining the control ofK+ channels by muscarinic receptors. 196 Addition of recombinant GAP, p21 ras or GAP-p2lrascomplexes to isolated atrial membranes inhibited K+ flux through activated channels. Theeffect of free GAP was dependent on the presence of endogenous p21ras since neutralizingantibodies to p21 ras abrogated this effect. Intriguingly, the C-terminal domain of GAP, whichcontains the GTPase activating domain and is sufficient to suppress ras mediated cellulartransformation 192 and also stimulate ras GTPase activity 197 but was not sufficient forchannel blocking. These observations suggest that GAP-p2lras complexes uncouple the K+31channels from muscarinic receptors and that the GAP polypeptide contains important activitiesoutside of its GTPase activating domain.1.3.2 Nucleotide diphosphate kinases as regulators of rasA downstream effector role for GAP is attractive since an upstream regulator modelwould predict a futile cycle in which GAP, unless deactivated by a mitogenic stimulus, keepsp2lras in a GDP bound state by catalyzing hydrolyis of bound GTP. A more conservativemodel of p2 gas regulation involves control of the rate with which GDP is replaced by GTP. Inthis scenario, p2lras is in a GDP bound state until agonist stimulation signals exchange ofGTP for GDP. Two mechanisms exist for this replacement: direct phosphorylation of p21rasbound GDP by a nucleotide diphosphate kinase, or stimulation of the rate at which GDP isexchanged for GTP. The action of a nucleotide diphosphate kinase on protein bound GDP hasalways been a theoretical possibility. 198 The simplest interpretation of the site of action ofthese kinases suggest that they act on free nucleoside diphosphates which would result ingreater pools of nucleoside triphosphates. Exchange of bound nucleotide diphosphates fortriphosphates would then give the appearance of direct phosphorylation of the original proteinbound nucleotide diphosphate. Although direct phosphorylation has been difficult to proveexperimentally because nucleotide exchange has been difficult to control for, a recent reportsuggests that purified nucleotide diphosphate kinase catalyzes the phosphorylation of GDPbound to ARF (adenosine diphosphate ribosylation factor). 199 It is too early, at this point, topredict the relevance of this mechanism to the control of p2l ras guanine nucleotidephosphorylation state.1.3.3 Nucleotide exchange proteins as regulators of rasFar better characterized are the proteins which catalyze the exchange of ras bound GDPfor GTP. A number of nucleotide exchange proteins have been described for various membersof the ras family including ones for ras",201 rho,202 and rar-L203 The ras family proteinshave very low rates of nucleotide dissociation in the absence of these catalysts," an essentialfeature if p2lras is controlled by the regulated release of GDP. Support for this model has32been accumulating. It has been recently proven that the yeast CDC25 gene product, whichacts upstream of the ras proteins, is a nucleotide exchange protein. 204 In Drosophila, anucleotide exchange protein has been recently implicated in the ras signalling pathway fromthe sevenless tyrosine kinase receptor. 206 These provocative results suggest that p2l ras maybe controlled by agonist stimulation of nucleotide exchange, in a manner very similar toreceptor regulation of the heterotrimeric G proteins.Regulation of p2lras at the level of nucleotide exchange does not preclude regulation atthe level of GTP hydrolysis by GAP, ie. both mechanisms may be operative. However, it isbecoming increasingly apparent that GAP has functions other than stimulating GTPase activitysince two of the known GAP's contain activities in addition to their GTPase activating domain.Ras GAP has SH2 domains important to its interaction with tyrosine kinases and otherproteins; the interaction, directly or indirectly, uncoupling the muscarinic receptor from K+charmels. 196 The bcr protein contains, in addition to its GAP domain, a region which bindsSH2 sequences and a serine/threonine kinase activity.410 Perhaps the GTPase stimulatingactivities of these molecules are important as downstream signal terminators of these otherfunctions rather than as upstream regulators of p2lras.2.^Protein Kinase C and Inositol Phosphate MetabolismProtein kinase C (PKC) is involved in the regulation of many cellular metabolic processesincluding endocrine and exocrine tissue secretion, smooth muscle contraction, steroidogenesis,ion channel fluxes, cAMP dependent phosphorylations, receptor desensitization, and geneexpression.206 The importance of PKC in growth control is suggested by the fact that PKC isthe site of action of tumour promoting phorbol esters 207 and by the discovery that theretroviral akt oncogene codes for a PKC analogue. 208 PKC was originally described as a proteinkinase activity that could be activated in vitro by limited proteolysis with calpain. 206 Furthercharacterization of this activity showed that the physiological activator of PKC, diacylglycerol(DAG), increases the enzyme's affinity for its cofactors, Ca2+ and phosphatidylserine. 209,210In the classically described PKC activation pathway, agonist induced receptor activation33stimulates hydrolysis of membrane PIP2 which generates diacylglycerol (DAG) and inositol1,4,5 trisphosphate (InP3). DAG remains in the membrane to stimulate PKC while InP3 bindsto cognate receptors on the ER membrane211,212 to liberate Ca2+ stores. This picture hasbeen complicated by the discovery of new regulators of PKC activr --2Ly- 13,214 and by the cloningof multiple PKC isozymes. 215 In addition, although agonist controlled phosphatidylinositol (PI)hydrolysis is intimately involved in PKC activation because of the generation of DAG and InP3 ,other metabolites of PI turnover have been implicated in signal transduction.2162.1 Protein kinase C Several PKC isozymes have been isolated which have been categorized into fivegroups.215 Although there is much inter-group variation, the members within each group arevery homologous and highly conserved between species. The various isozymes differ in theirdistribution, the a group being widely expressed, while the 13 and y isoforms are highest inendocrine and brain tissues, respectively. The 8 and e forms have not as yet been wellcharacterized. The a,f3 and y isozymes are similar in structure with five variable and fourconstant domains. 217,218 The C-terminal half of the enzyme contains the catalytic site, theN-terminal portion has the regulatory domain with its cysteine rich, zinc finger motif essentialfor DAG binding. 210 The 8 and e isozymes which are independent of Ca 2+ , lack the C2 regionin the N-terminal domain which may be important in Ca 2+ binding. In all PKC's, a regionbetween the catalytic and regulatory domains is a unique region with high affinity for thecatalytic site of the enzyme. 219 This pseudosubstrate region appears to be important inregulating PKC activity. 220 PKC activators such as DAG may induce a conformational changethat shifts the pseudosubstrate domain out of the catalytic site to allow access of substrates.2.1.1 Regulators of PKC The variable regions determine co-factor requirements and substrate specificity of thedifferent isozymes. For example, the a isozyme is the most dependent on DAG for activation,the 13 form exhibits considerable activity in the absence of Ca2+ , and the y can be activated invitro by arachidonic acid without Ca2+ or phospholipid. Although co-factor requirements for34the 8,e and are not as well characterized, these isozymes, unlike the a, 13 and y forms, do notrecognize histone as a substrate. 206 Moreover, phorbol esters bind to and activate the variousisozymes to varying degrees.21° The different isozymes of PKC also differ in their sensitivity toproteolysis.21° Proteolysis may be important in PKC regulation. Specifically, calpainpreferentially cleaves the activated form of PKC221 in a region between its N-terminalregulatory and catalytic domain to generate a potentially active fragment referred to as PKMwhich might be important in persistent PKC activation. However, this PKM fragment may bebroken down further since it is often difficult to detect, either immunologically or by activitymeasurements. The generation of PKM may therefore be involved in PKC downregulationrather than persistent PKC activation. The distinct differences among the isozymes, however,suggests that a cell's response to PKC activating signals depends on the particular combinationof isozymes expressed.Other regulators of PKC include unsaturated fatty acids, such as arachidonic, oleic orlinoleic acid, which synergize with DAG to allow activation of PKC when Ca 2+ levels arelow.213 These fatty acids can be produced by receptor mediated activation of phospholipaseA2.214 Sphingolipids, on the other hand, are inhibitory and may be responsible formaintaining PKC in a basal, inactivated state in the presence of cellular levels of DAG. 222However, PKC may never actually be exposed to DAG generated during normal housekeepingprocesses because of compartmentalization. For example, a cardinal sign of PKC activation isthe translocation of this enzyme from the cytosol to the plasma membrane. Recent studieshave also shown that some PKC molecules become associated with the detergent insolublecytoskeleton. Activated PKC binds to cytoskeleton associated proteins called RACKs (Receptorsfor Activated C-Kinase); this association alters the subcellular localization of PKC and thus maybe involved in its regulation. 223 The different PKC isozymes display differing RACK bindingefficiencies. 223 PKC may also translocate to the nucleus224,225 in response to nuclear DAGproduction226 and act on nuclear targets.352.1.2 Targets of PKCVarious targets of PKC action have been described and receptor desensitization is oneimportant consequence of PCK activation. Specifically, the intrinsic tyrosine kinase activities ofthe EGF and insulin receptors are inhibited, while the beta-adrenergic and glucagon receptorsbecome uncoupled from G-proteins upon PKC phosphorylation. 227 PKC phosphorylation ofthe EGFR was also once believed to be responsible for decreasing its affinity for EGF. However,site-directed mutagenesis of the residues involved showed that these phosphorylations wereimportant only in receptor internalization. 227 The effect of PKC on EGFR binding affinity isprobably mediated indirectly, for example, through phosphorylation of other proteins whichassociate with the receptor. Another prominent PKC substrate is a cytosolic "80 kD" proteinwhich is phosphorylated in a PKC dependent manner in a wide range of cells in response tomany factors. 228 This almost ubiquitous marker of PKC activation was cloned and the cDNAfound to encode a 32 kD polypeptide. Since this protein is myristylated, which gives it anaberrant mobility on SDS-PAGE, and contains a high alanine content, it was named MARCKSfor Myristylated, Alanine Rich C-kinase Substrate. MARCKS contains a binding site forcalmodulin, 229 an ubiquitous Ca2+ dependent regulatory protein that binds to and regulatescellular proteins. PKC phosphorylation of MARCKS releases MARCKS sequestered calmodulinand may allow activation of calmodulin dependent processes. MARCKS also has an actinbinding site which suggests that it may be involved in the cytoskeletal rearrangementsobserved upon PKC activation. 230 Transcription factors are another important target of PKCaction. PKC activates NF-KB 225 via releasing free, active NF-KB from an inactive NF-KB/ IkBcomplex by phosphorylation of the IKB inhibitory protein. 231 PKC also phosphorylates andactivates AP-1 and AP-2. 231 In the case of NF-KB, PKC acts on a cytosolic, inactive complex togenerate an active factor that enters the nucleus. 231 However, PKC itself may translocate intothe nucleus to phosphorylate factors in situ.224,225 Regardless of the subcellular location, theaction of kinases such as PKC on transcription factors is an important, intermediate link fromearly receptor to later nuclear events.362.2 Phospholipase CPKC is a downstream effector, rather than a receptor associated signal transductionmolecule since its activation depends on upstream production of regulators such as DAG andCa2+ . DAG production is controlled by agonist sensitive phospholipase C's and much hasbeen learned recently about the role of phospholipases in signal transduction, including theexciting discovery that one isozyme of PLC may directly couple tyrosine kinase receptors todownstream effectors such as PKC.Phosphatidylinositol specific phospholipases are classified into five groups based onprimary structure. 232,233 They share two regions of significant homology which probablycontain the catalytic domain. However, outside of these domains, the sequences are highlyvariable. Of the different types of PLC's, only PLC13 1 , PLCe and PLCy appear to be involved insignal transduction. PLC13 1 and PLCe233 but not PLCy234 are regulated through Gproteins233 of the pertussis toxin insensitive, 235 Gq class. 234 In the case of PLCI3 1 , the Gprotein activates by reducing PLC13 1 's Ca2+ requirement. 236 Interestingly, a specific growthfactor can activate different types of PLC in different cell types. EGF, for example, stimulatesinositol phosphate metabolism through a pertussis toxin sensitive PLC in hepatocytes butactivates a G protein independent PLC in A431 cells. 237PLCy, the only isoform with SH2 domains, has been the subject of considerable attentionbecause of its physical association with intrinsic tyrosine kinase receptors. As with other SH2containing proteins, these domains are dispensable for in vitro enzymic activity. 238 Thesedomains are required, however, for binding to ligand activated, tyrosine phosphorylatedreceptors.239,24° Activation of PLCy in intact cells is accompanied by tyrosinephosphorylation of the enzyme itself, although this phosphorylation does not increase catalyticactivity, since replacement of the relevant tyrosine with phenylalanine has no effect on in vitroactivity.241 Tyrosine phosphorylation is essential, however, both for interaction with profilin, aprotein which sequesters PIP2 in vivo, and subsequent cleavage of profilin bound PIP 2 . 169These observations are consistent with a model in which EGF,239 PDGF,24° nerve growth37factor (NGF)242 and fibroblast growth factor (FGF)242 stimulate the receptor intrinsic tyrosinekinase that phosphorylates receptor associated PLCy. Tyrosine phosphorylated PLCy thencatalyses DAG and InP3 production from profilin bound PIP2. That PLCy is associated with thereceptor prior to receptor autophosphorylation, in a presumably SH2 independent manner,comes both from the observation that PLCy phosphorylation occurs at 4° C243-246 and fromimmunoprecipitation experiments. 245DAG, in some systems, can also be generated from phosphatidylcholine (PC)hydrolysis247 via an agonist sensitive PC specific PL.248 Significantly, DAG from this sourcedoes not result in proteolysis and down-regulation of PKC, either because the signal thatactivates PKC is different or because the fatty acid composition of the PIP2 and PC derivedDAG's differ.249 As a consequence, PKC activation can be long term, and this sustained PKCactivation is important for stimulation of AP-1 enhancer activity in macrophages 25° and foractivation of resting human T cells. 251It is not clear how Ca2+ mobilization occurs or whether it is required in systems whereDAG is derived from PC. However, in the more common PIP 2 pathway, Ca2+ is released fromintracellular endoplasmic reticulum (ER) stores. Two ER InP 3 receptors have been cloned andshown to consist of a calcium channel domain joined to a cytoplasmic ligand binding domainthrough a coupling domain.252,253 The two receptors are the least homologous in the bindingdomain and have different affinities for InP3. 253 The coupling domain is subject to cAMPdependent kinase phosphorylation, suggesting that it may have a regulatory role in controllingthe function of the ion channe1. 254 Ca2+ release may therefore dependent on InP3 receptorexpression in a particular cell type as well as the activation state of other signalling pathways.2.3 Phosphatidylinositol 3-kinase and other PIP2 metabolites Although the best studied product of PIP 2 metabolism is InP3, other metabolites mayalso be important in signalling. Inositol 1,3,4,5-tetrakisphosphate, for example, may beresponsible for promoting the influx of external Ca2+255 and inositol 1,4-bisphosphate mayactivate a low specific activity form of DNA polymerase a. 216 However, much interest has38focussed on inositol polyphosphates that are phosphorylated on the D-3 position of the inositolring because of the discovery that the enzyme responsible for this modification,phosphatidylinositol 3-kinase (PI3-K), which associates with tyrosine phosphorylated growthfactor receptors. 139PI3-K was first described as an enzyme activity associated with cytoplasmic tyrosinekinases. 256 Intriguingly, PI3-K associates only with transforming versions of src and notnormal src.257 This association of PI3-K with these tyrosine kinases is thought to recruit PI3-K to the plasma membrane and thus allow access substrates. 258 Plasma membranerecruitment may also be the purpose of PI3-K association with the ligand activated, intrinsictyrosine kinase receptors for insulin, 259 CSF-1,26° PDGF261 and EGF.262 PI3-K consists ofsubunits with molecular weights of 110 and a 85 kl3 263 and it is the 85 kD subunit whichcontains two SH2 domains264,265 and mediates the association of PI3-K with tyrosinephosphorylated receptors. 266Of interest are the recent observations that PI3-K also associates with activated IL-2Rs 136 and P 2 lras,267  As discussed earlier, the first observation is important since itdemonstrates that some of the signalling proteins that bind to intrinsic tyrosine kinasereceptor may also associate with receptors that lack tyrosine kinase domains. The way inwhich the second observation, although potentially very exciting, fits into the known scheme ofsignal transduction is not clear.3. Tyrosine KinasesThe importance of tyrosine kinases in signal transduction is demonstrated by their manymanifestations in the form of viral oncogenes268 and growth factor receptors. Tyrosinekinases can be classified as those which are transmembrane receptors, which were discussedearlier, and the cytoplasmic tyrosine kinases. The cytoplasmic tyrosine kinases can further bedivided into src,268 ab/269 and fps270 families based on sequence homologies. Of these, thesrc family of tyrosine kinases are the best characterized.393.1 Structure of the src family of tyrosine kinaseThe overall structure of a src kinase is shown in figure 4A. The N-terminus ismyristylated and the recognition sequence for this modification is found in the first 7-10 aminoresidues.271 p605 rc must be membrane associated in order to exerts its action andmyristylation is necessary, although not sufficient, for this to occur. 272 The association withthe membrane may be necessary to b ring p6Osre in contact with certain substrates andappears to be mediated by specific p60sre receptors. 273,274 The next 75 residues are highlydivergent among different members of the src family275 and may determine substratespecificity and subcellular localization. The C terminus, on the other hand, contains a negativeregulatory domain. Phosphorylation of a C-terminal tyrosine, tyr-527 in p6Os re, inhibitstyrosine kinase activity as shown by phosphatase276 and site directed mutagenesisexperiments.277,278 Members of the src family can also be activated in vivo275 by theassociation with middle T antigen in polyoma virus infected cells. Since the region required formiddle T binding is also the region required for in vivo phosphorylation at tyr-527, the bindingof middle T is believed to activate p60sre by preventing phosphorylation of tyrosine 527 by aregulatory kinase. The receptor intrinsic tyrosine kinases may have analogous C terminalregulatory domains.62,279The SH 1 region contains the tyrosine kinase domain which , is conserved in all tyrosinekinases and homologous to serine/threonine (Ser/Thr) catalytic domains. 280 The major site ofautophosphorylatlon, tyrosine 416 in p6osre is located in the catalytic domain, andphosphorylation at this residue activates kinase activity. 277,281 In contrast, SH2 and SH3domains are found in proteins other than tyrosine kinases. The SH3 domain is thought to beinvolved in cytoskeletal interactions. 84 However, attention has focused on the SH2 domainsbecause of their role in mediating interactions between tyrosine kinase receptors and signallingproteins.A GENERIC SRC FAMILY MEMBER1%-spirl,...___SpecifIc Protein Interactions SH3^SH2,...„Common Protein InteractionsATP^AUTOPHOSPHORYLATION100^ 200^300^ 400^500Amino Acids1:4442- -Mk-s 14,2-1MEM=irSTI rs-Er2.1p85 PI3KtensinvavFigure 4.^Panel 4A shows the structure of a src family tyrosine kinase (from reference280). Proteins with SH2 and SH3 domains are shown in Panel 4B.40B.srcGAPPLCcrkn ck41As mentioned earlier, the SH2 domains have been found in a number of proteins (Figure4B). Many of these proteins: PLCy, PI3-K and ras GAP, have been shown to be involved insignal transduction by associating with tyrosine phosphorylated proteins through SH2 domainmediated interactions. The presence of SH2 domains in other proteins 84 : nck, tensin and vav,may therefore suggest that these proteins are also important in signalling. However, regardlessof whether these three do participate in signal transduction, the SH2 domains do providepotential sites for protein/protein interactions.The role of SH2 domains in p60sre have been extensively studied and they are thought toserve either in regulation of kinase activity or in substrate recognition. Both of these modelsare based on the tyrosylphosphoprotein binding function of the SH2 domains. In the regulatormodel, the SH2 domain is postulated to bind to the C terminal, negative regulatoryphosphotyrosine and thus mask the tyrosine kinase domain. In support of this model, limitedproteolysis of p6Osre suggests that p6Osrc changes its conformation upon activation. 282 Inaddition, the observation that point mutations in the SH2 domain activate the transformingpotential of p60sre, without any detectable change in the phosphorylation state of the C-terminal tyrosine, is also consistent with this model. This model also suggests a mechanism bywhich the crk oncogene, which does not encode a tyrosine kinase, transforms cells. The crkprotein,a 47 kD polypeptide consisting almost entirely of SH2 and SH3 domains, 283 associateswith p60sre in an SH2 dependent manner. 284 The binding of p47erk SH2 to the terminalphosphotyrosine, in the place of the intramolecular p60s re SH2, would open up the tyrosinekinase domain. Release of endogenous tyrosine kinases from negative regulatory control wouldaccount for the increases in tyrosine phosphorylation level in crk transformed cells. 283 Notmutually exclusive with its role in regulating tyrosine kinase activity, is a model in which theSH2 domain mediates interaction with regulatory proteins or substrates. Various mutationshave been described in the SH2 domain that decrease the transforming potential of variouskinases285,286 or change the cellular range of the transforming protein287 without altering its42in vitro kinase activity. These observations suggest that interaction with cellular substrates isimpaired in these SH2 mutants.3.2 Targets of tyrosine kinase actionActivation of a tyrosine kinase results in phosphorylation of many cellular proteins.Many of these substrates have not been characterized further than by molecular weight.However, several have been identified. For example, p6Os re activation correlates withphosphorylation of vinculin, talin, erzin, calpactin, and lipocortin 11. 28° The physiologicaleffects of these phosphorylation are not known. Despite the difficulty in proving directconsequences of tyrosine phosphorylation, the involvement of cytoplasmic tyrosine kinases insignalling cascades is not doubted. For example, p6Osre is thought to be involved in EGFsignalling because over expression of c-src in fibroblasts potentiates the mitogenic response toEGF.288 The discovery of SH2 and SH3 mediated interactions has opened up the possibilitythat tyrosine phosphorylation serves mainly to direct protein traffic.3.3 Association of tyrosine kinases with receptorsMany cell surface receptors, including the hemopoietin receptors, lack intrinsic tyrosinekinase domains yet they become tyrosine phosphorylated upon ligand binding. Recently, veryexciting work has shown that several of these receptors are physically associated with srcfamily tyrosine kinases (Table 4).TABLE 4: Tyrosine Kinases Associated with Cell Surface ReceptorsReceptor^ tyrosine kinase^ReferenceIL-2R13 chain^ lckT cell antigen receptor (TcR)^fynB cell antigen receptor lynCD4/CD8^ lckbasophil/mast cell IgE Fe receptor^lyn, yesplatelet glycoprotein CD36^fyn, yes, lyn135289290291,29229329343These kinases become activated upon Nand stimulation of the receptor, resulting inphosphorylation of the receptor on tyrosines residues.The structural features important in these interactions has been investigated forCD4/CD8 and the IL-2R A membrane proximal +-+-X-C-X-C-(P) (where + is any basic aminoacid) sequence on both the CD4 and CD8 molecules appear to be important in the binding ofp661ck 294 A similar cysteine sequence in the bovine metallothionein I protein is known tointeract with zinc. Since 1,10-ortho-phenanthroline, a zinc chelator, disrupts the associationof p561ck with CD4 or CD8, formation of a zinc coordination complex may be involved in thebinding of p561ck to CD4/CD8. 294 In contrast, the region in the IL-2R f chain important inp561ck association is characterized by an abundance of acidic amino acids. 135 This regioncontains the two tyrosines (residues 355 and 358), phosphorylated by p561ck, butphosphorylation of these residues is not required for p56 1ck binding. Interestingly, the regionin p 561C1C important for association with CD4/CD8 differs from the region involved in bindingto the IL-2R The N-terminal unique region, which distinguishes different members of the srcfamily, is involved in p561ck association with CD4/CD8, 294 whereas a region in the N terminalhalf of the SH1 domain mediates interaction with the IL-2R 5 chain. 135In addition to being associated with receptors lacking intrinsic tyrosine kinase domains,members of the src family of tyrosine kinases have also been shown to associate with the PDGFreceptor.295 Three src family kinases, p60src.p6Ofn and p603'es become activated in PDGFstimulated cells and co-precipitate with the PDGFR. 295 These kinases may serve tophosphorylate substrates that the PDGFR is not capable of phosphorylating.4. Serine/threonine Specific KinasesSerine/threonine kinases are far more common than tyrosine specific kinases and many,including protein kinase C and the cyclic nucleotide dependent kinases are thought to play keyroles in signal transduction. However, of all the serine/threonine kinases in the cell, only threehave been shown to be controlled by both tyrosine and serine/threonine phosphorylation,suggesting perhaps these kinases play a pivotal role in the signal transduction cascade by44integrating information from both tyrosine and serine/threonine kinase pathways. One ofthese is the cyclin dependent kinases which will be discussed in the following section on cellcycle control: the other two are the Raf and MAP kinases.4.1 Raf-1 kinase Raf-1 the normal cellular homologue of v-raf, the transforming gene of murine moloneysarcoma virus,296 belongs to a family of closely related cytosolic 74 kD serine/threoninekinases. Interestingly, a transmembrane, receptor-like raf homologue has been detected in C.elegans. 53 Other members of this family include A-raf, B-raf, the raf-2 pseudogene, and theavian v-mil. While Raf-1 mRNA is ubiquitously expressed in all proliferating cells thedistribution of A- and B-raf is mainly restricted to urogenital and brain tissues,respectively.298 The structure of p74ral--1 is reminiscent of other serine/threonine kinasessince it has an N-terminal regulatory domain and a C-terminal catalytic domain. 297Interestingly, the N-terminal regulatory region contains a cysteine rich region similar to PKCsuggesting, perhaps, similar small ligand regulation. Truncation of the N-terminal region, as inv-raf, results in activation of the transforming potential of raf-1.p74 -1 has been studied intensely because its activity is upregulated in mitogenstimulated cells. IL-3, 299,30° GM-CSF,299,300CSF-1,301 PDGF,302 insulin,3°3 andEGF304 all stimulate p74raf-1 kinase activity. In all cases, p74raf-1 kinase activation isaccompanied by serine phosphorylation of the p74raf-1 protein itself. The kinase responsiblefor this modification has not as yet been identified but it is not believed to be PKC sincep74raf." 1 activation can occur in PKC downregulated cells. Tyrosine phosphorylation alsoaccompanies activation in some cases and this phosphorylation is most convincinglydemonstrated in the mouse IL-3, GM-CSF299 and PDGF302 systems where considerablephosphotyrosine can be detected. However, in the remaining systems, tyrosinephosphorylation is of very low stoichiometry or not detectable. This has raised the question ofwhether tyrosine phosphorylation is essential for p74 ra1-1 activation. In support, in vitrophosphorylation of p74raf-1 with recombinant, baculovirus PDGFR tyrosine kinase has been45shown to increase p74 -1 kinase activity six fold.83 It is possible that tyrosinephosphorylation of p74 -1 is very transient in some cells, but nonetheless is necessary forserine phosphorylation which then is itself sufficient for activation. One group 301 however,suggests that in vivo activation of p74raf" 1 does not require tyrosine phosphorylation at all.This group points out possible deficiencies in antibody specificity which makes interpretation ofcertain data difficult in studies302 which do show p74ral" 1 tyrosine phosphorylation. Thequestion of whether tyrosine phosphorylation is indeed important in regulating the kinaseactivity of p74 -1 waits to be resolved by site directed mutagenesis of the tyrosine residue(s)thought to be involved.p74' -1is also notable among signal transduction molecules because it is one of thefour proteins which have been shown to associate with ligand activated, tyrosinephosphorylated receptors. p74raf-1 associates with the PDGFR83 •302 and the EGFR304 andthis association may allow it to be phosphorylated by the tyrosine kinase intrinsic to thesereceptors. However, p74raf-1 does not associate with either the CSF-1 301 or the insulin303receptors even though both of these contain intrinsic tyrosine kinases and p74raf -1 doesbecome stimulated in cells incubated with ligands for these receptors.The placement of p74 -1 downstream of tyrosine kinase receptors in the signaltransduction cascade, whether p74 -1 physically associates with the receptor or not, isapparently conserved evolutionarily, ie. the Drosophila homologue, D-raf, acts downstream oftorso, a transmembrane tyrosine kinase similar to the PDGFR. 305 p74raf-1 may also actdownstream of p2lras, or alternatively, use a pathway independent of ras since u-raf is able toraf -1transform ras deficient cell lines. 296,298 p74may be an upstream regulator oftranscription factors such as those responsible for regulating c-fos.3°8 In this regard it isinteresting to note that PDGF or TPA stimulation of quiescent cells results in rapidredistribution of raf protein from a uniform cytoplasmic distribution to a perinuclearlocation.298464.2 MAP kinaseMAP-2 kinase is a mitogen activated 42 kD serine/threonine kinase originally assayed invitro with, and thus named for, microtubule associated protein-2 as a substrate. 307 However,with the recent discovery that MAP-2 kinase was actually one member of a farnily308 andbecause of its increasingly apparent importance and wide range of activities in many signallingsystems, "MAP" kinase is now considered an acronym for Mitogen Activated Protein kinase.However, several synonyms, such as "ERK" (Extra-cellular signal Regulated Kinase), extst. 308MAP kinases are stimulated in various cells in response to many factors, including IL-3,309 EGF,31° NGF,311 FGF,311 insulin312 and TcR ligation. 313 In contrast to raf kinase,many substrates for MAP kinases have been described. One of the major in vivo substrates isthe microtubular network, the phosphorylation of which may lead to disassembly. 314 Othersubstrates include other kinases important in signal transduction, namely, the 90 kD S6kinase311 (or pp90rsk.315) and the EGFR MAP kinase mediated phosphorylation of S6kinase has been shown to increase S6 kinase activity in response to insulin, indicating thatMAP kinase acts as a downstream effector in a kinase cascade initiated by the insulin receptortyrosine kinase. 312 MAP phosphorylation of the EGFR, on the other hand, apparentlydecreases EGFR kinase activity316 and thus may be involved in a negative regulatory feedbackloop responsible for EGFR desensitization. MAP kinase has also been shown to act late in thesignal transduction cascade by phosphorylating and activating the transcription factor c-jun.317Activation of MAP kinase is accompanied by phosphorylation on both tyrosine andthreonine residues and both are required for MAP kinase activity. 318 Because most kinasesdescribed to date phosphorylate substrates on either tyrosine or serine/threonine residues, butnot both, the requirement for both phosphorylations suggests MAP kinases may integratesignals from two upstream kinases. Efforts directed towards purifying and identifying theupstream kinases have led to some controversy.319,320 One group for example believes thatMAP kinase is phosphorylated and activated by a MAP kinase kinase that is, surprisingly,47capable of phosphorylating both tyrosine and threonine residues.32° Krebs' group, on theother hand, describes the upstream molecule as a MAP kinase activator319 that does not haveintrinsic kinase activity. This second interpretation has been strengthened by the unexpectedobservation that, although phosphorylation of exogenous substrates appears to be stillrestricted to serine/threonine residues, MAP kinases can autophosphorylate on both tyrosineand threonine residues.321The possibility that MAP kinase is capable of autophosphorylating on both tyrosine andthreonine residues casts in doubt any theory that the MAP, raf and cyclin dependent kinasesoccupy pivotal, integrative positions in signal transduction which is based only on theobservation of both serine/threonine and tyrosine phosphorylations. These kinases areinteresting, nevertheless, since they have many actions and functions in many signallingsystems. The possibly shared ability 3 21 for autophosphorylation on both tyrosine andserine/threonine residues may indicate a unique method of regulation, the significance ofwhich remains to be discovered.5. PhosphatasesProtein phosphorylation is regulated not only by kinases, but by phosphatases as well.As with protein kinases, the phosphatases involved in intracellular signal transduction can bedivided into tyrosine322 and serine/threonine specific323 classes. However, unlike thekinases, the two classes are not evolutionarily related, as evidenced by differing catalyticdomain structure.3225.1 Tyrosine specific phosphatases Tyrosine specific phosphatases are a fairly recent discovery 3 24 and have been implicatedin the mechanism of action of at least one of the interleukins (IL-4).325 A role in signaltransduction for tyrosine phosphatases is also suggested by the discovery of a tyrosine specificphosphatase containing two copies of SH2 domains.326 Like the tyrosine specific kinases,tyrosine phosphatases are found in both cytoplasmic and transmembrane forms. 322 Thetransmembrane tyrosine phosphatases are especially intriguing because they resemble48receptors322 and one of these transmembrane tyrosine phosphatases, CD45, appears to beessential in the action of the src family of tyrosine kinases.327-33°CD45 is a molecule originally identified as an antigenic marker of leukocytes. Differentsplice variants of CD45, which are expressed on different subsets of leukocytes, generatedifferent N-terminal, external, putative ligand binding domains which may suggest each variantmay respond differently to an external stimulus. Like all transmembrane tyrosinephosphatases, the cytoplasmic region of CD45 contains two phosphatase catalytic domainswhich may possess different specificities. 324 The action of CD45 has been implicated inTcR327-33° and B cell antigen receptor331 activation of receptor associated src family kinases.As discussed above, TcR ligation results in activation of a receptor associated p 561ek kinaseHowever, like other members of the src family, p56" is inhibited by the phosphorylation of itscarboxy terminal tyrosine (tyr-505 p561ck, 275  Activation of p56" requiresdephosphorylation of this residue and CD45 may be the phosphatase responsible since T cellsdeficient in CD45 are impaired in TcR signalling. 329 Furthermore, CD45 phosphatasetreatment of p56" in vitro activates p56" kinase activity328 and overexpression of CD45 invivo results in decreased phosphorylation of p56 1ck .327 The CD45 tyrosine phosphatase,which may be regulated by serine phosphorylation, 33° has also been similarly implicated in Bcell antigen receptor activation of lyn. 3315.2 Serin.e/threonine specific phosphatasesNumerous Ser/Thr specific phosphatases have been described323 and some of thesemay be involved in hemopoietic signal transduction pathways. For example, CD2 mediated Tcell activation is associated with the dephosphorylation of a 19 Id) protein 332 and aphosphatase may be involved in antagonizing the PKC mediated phosphorylation of CD3. 333Also, unexpectedly, the inununosuppressants cyclosporin and FK506 were recently shown toact by suppressing the activity of a Ca2+/calmodulin dependent phosphatase calledcalcineurin.334 Thus, like tyrosine phosphatases, Ser/Thr specific phosphatases may also beimportant in signalling.496. Signalling Pathways Implicated in the Mechanism of Action of IL-3Murine IL-3, like all other hemopoietic growth factors, is absolutely required for cellgrowth and survival. Cells deprived of mIL-3 quickly lose viability and die by a process ofprogrammed cell death referred to as apoptosis. 335,336 Early studies investigating thebiochemical basis of mIL-3 action showed that mIL-3, like many growth factors, stimulatedglucose uptake337,338 and ATP elevation339 in target cells. Eventually, similar to thatobserved in many other growth factor systems, 14° these initial events lead to elevation of c-mycand c-fos expression.340,341 Studies directed towards further dissecting the mIL-3 signallingpathway have thus far concentrated on protein kinase C activation and protein tyrosinephosphorylation events.With respect to protein kinase C, the literature contains many apparently conflictingresults. In the majority of studies, mIL-3 was observed to activate PKC activity. 342-345However, whereas activation of PKC was accompanied by its translocation to the plasmamembrane in some instances,344-346 no change in localization was detected in others. 342Similarly, while some studies have implied that diacylglycerol is generated through theclassical inositol phosphate pathway, 346 others have suggested that DAG may be derived froman alternate source343,347 such as PC.348 This variability may, to some extent, be attributedto differences intrinsic to the cells studied. Different cells, as discussed above, may possess adifferent complement of PKC and PLC isozymes. These differences may also account for thefact that phorbol esters can support the growth of some mIL-3 dependent cells 349,35° buthave only modest effects on others. 351There is less variation amongst cell lines when it comes to mIL-3 stimulated tyrosinephosphorylation events. A role for tyrosine phosphorylation in mIL-3 action was first suggestedby the fact that v-abl could confer growth factor independence on a mIL-3 dependent cell linevia a non-autocrine mechanism. 352 Subsequently, mIL-3 was shown to rapidly inducetyrosine phosphorylations in a number of different cell lines.351,353-356 In all cases, a 140k.13 tyrosine phosphorylated protein, shown to be the mIL-3R, 357,358 was observed. The50identity of other notable tyrosine phosphorylated proteins, with apparent molecular masses of90-100 ki3,351,354 70 kD353,355,356 and 56 kD354,356 are not yet known.A logical next step in furthering our knowledge of mIL-3 signal transduction, is tocharacterize these phosphoproteins, especially with regard to their relationship to the mIL-3RThe studies described in this thesis focuses on these phosphoproteins and the earliestbiochemical events that occur following growth factor stimulation. However, it is important tokeep in mind that the biochemistry set into motion by growth factor stimulation musteventually interact with intrinsic cell cycle controls.D. CELL CYCLE CONTROLThere are three points in the cell cycle at which progression can be controlled: at theentrance into G1, the G1 /S transition or at the G2/M boundary. The first two of thesecheckpoints are subject to control by external growth stimulatory or inhibitory signals, whilethe third is apparently subject to an intrinsic cellular program.Much of our understanding of mammalian cell cycle control comes from studies insimpler systems such as that encountered in the Xenopus oocyte during maturation or yeastcell division. In xenopus, the G2/M transition is controlled by an activity called M-phasePromoting Factor (MPF), 359 originally assayed by its ability to induce maturation uponinjection into oocytes. Purification and biochemical characterization of this activity, which canbe isolated from any mitotic, eucaryotic cell, showed the catalytic component of the multi-subunit complex to be identical to a 34 kd serine/threonine kinase originally identified inyeast. In yeast, study of the cell cycle is facilitated by the ease of isolation of conditional celldivision cycle (cdc) mutants which are blocked, when grown in non-permissive conditions, atvarious points in the cell cycle. The 34 kd kinase is the gene product of the fission yeast cdc2gene; a mutation at cdc2 results in M phase arrest. 360,361 The cdc2 kinase is essential forcell cycle progression through mitosis and its importance is suggested by its ubiquitousexpression and high degree of conservation in all eucaryotic cells. 362 In fusion yeast, theCDC28 gene encodes an homologous kinase. In humans, a cdc2 homologue was isolated by51complementation of the cdc2 mutation in fission yeast, 363 and shown to be essential formitosis with the use of either temperature sensitive mutants364 or microinjection ofneutralizing antibodies.365The activity of cdc2 kinase fluctuates within the cell cycle, being highest at the onset ofmitosis362,366 and undetectable outside of M phase. However, cdc2 protein levels remainconstant throughout the cycle and its activity is regulated by phosphorylation and associationwith cyclin proteins. 367 Towards the beginning of G2 phase, unphosphorylated cdc2 kinasebecomes phosphorylated on threonine residues and associated with cyclins.366-368 Thisassociation targets cdc2 for tyrosine phosphorylation369 and nuclear localization. 368 Theinactive cdc2/cyclin complex is then activated at the onset of mitosis by cdc25 phosphatasemediated37° dephosphorylation of phosphotyrosine residues. 371 The activated cdc2 kinasemay then phosphorylate histones, nuclear larnins and other substrates responsible forchromosome condensation, nuclear envelope breakdown and mitotic spindle formation. 362Cyclin proteolysis366,367 at metaphase/anaphase via an ubiquitin dependent pathway372then inactivates the cdc2 kinase.Cyclins were originally described in marine invertebrates as proteins that cyclicallyaccumulate to high levels in interphase and undergo abrupt destruction at the end ofmitosis.373 In addition to these mitotic cyclins, 0 1 specific cyclins (CLN1, CLN2, CLN3, andCLN4) which mediate the G i /S transition have since been isolated in fusion yeast. 374,375The levels of two of these, CLN1 and CLN2, undergoe cell cycle fluctuation as expected, withhighest levels at late G1. CLN4 has only just been discovered and little is known about its levelof expression.375 However, the level of CLN3 remains constant through the cell cycle.Intriguingly, 0 1 cyclin expression is regulated through a positive feedback loop. 376,377Transcription of CLN1 and CLN2 requires the action of CLN3/CDC28 kinase activity, possiblythrough the phosphorylation and activation of transcription factors SWI4 and SWI6377 whichcontrol cyclin gene expression. CLN3, the only G 1 cyclin not to undergo cycle dependentchanges in abundance, is necessary as an upstream activator but is not sufficient in itself for52cell cycle progression since G 1 /S transition requires the presence of CLN1, CLN2 or CLN4. Thefeedback loop is closed by the ability of any of the synthesized cyclins to positively regulate itsown transcription. This positive feedback mechanism is an effective way to convert an initialsignal to a large increase in cyclin protein and may also be the basis for the irreversibility of Sphase commitment.In higher eucaryotes the cycle is complicated by the presence of many more cyclins.Several human cyclins have been isolated378-381 which have been classified into five typesaccording to sequence homology or function.380 Cyclins A and B are considered mitoticcyclins since highest levels exist at M phase. However, the abundance of cyclin A rises earlierin the cell cycle than cyclin B,382 implying it possesses a function outside of M phase. Indeed,differences exist between the two with respect to the kinase they prefer. In humans, it hasrecently been shown that there are at least two different cdc2-like kinases. 382,383 Cyclin Bcomplexes with the original cdc2, which can also be referred to as Cyclin Dependent Kinase-1(CDK-1), while cyclin A but not cyclin B associates with the newly cloned CDK-2.384,385 Thehistone H-1 kinase activity of cyclin A/CDK-2 is not as efficient as cyclin B/cdc2, 382,384suggesting different substrates and functions. As well, cyclin A differs from cyclin B in that itis sequestered by E1A in adenovirus infected cells whereas cyclin B is not. 382,384 Cyclins C, Dand E, on the other hand, are G 1 cyclins. Cyclin C mRNA is highest during early G 1 whilecyclin E is highest at late G 1 . Expression of the D type cyclins is more complex since theyexhibit both tissue specific expression386,387 and cell specific regulation. Cyclin D1 (alsocalled CYL1) is, at present, of great interest because it is the first cyclin whose expression hasbeen shown to be under growth factor regulation. 386 Specifically, in monocytes, CSF-1 isrequired for cell cycle progression; in its absence cells arrest at G 1 and die. Cyclin D1 mRNAand protein levels in G 1 arrested monocytes are low, but the addition of CSF-1 induces cyclinD1 mRNA during late G i . Continued presence of CSF-1 through S phase is required for mRNAstability. Cyclin D1 was independently isolated as a bcll linked gene from HeLa cells. 387 Inthese cells, cyclin D1 mRNA is induced following S phase and degraded at the G 1 /M boundary.53Perhaps the differences in cyclin expression reflect cell cycle control differences between CSF-1dependent monocytes and factor independent HeLa cells.Targets for the cyclin dependent kinases are not yet well defined. 226 Nuclear lamins388and histone H1 are two potential substrates since they undergo cell cycle specificphosphorylation coincident with the activation of mitotic cdc2 kinase. Phosphorylation of theseproteins may initiate the nuclear envelope disassembly and chromosome condensationobserved at mitosis. Other proteins have also been implicated as substrates since they alsoexhibit cycle dependent phosphorylations mediated by cdc2 kinase. These include p6Osrc389,390 p150abl 391 and the transcription factor encoded by oct 1.392 Thesephosphorylations may regulate their activity or alter substrate specificity in a cycle specificmanner.Recently, however, considerable attention has focussed on the cycle specific regulation ofthe retinoblastoma (Rb) gene product. The Rb susceptibility gene encodes a 100 kd nuclearprotein that functions as a negative regulator of cell growth. 393 Loss of Rb function correlateswith development of certain human tumours. The Rb protein is phosphorylated in a cell cycledependent manner with maximal phosphorylation during S-phase, reduced phosphorylationafter and no phosphorylation during G 1 .394 This phosphorylation is believed to be mediatedby cdc2 kinase.395 Elegant microinjection studies have shown that it is the unphosphorylatedform which suppresses cell proliferation by restricting cell cycle progression at a specific pointin G 1396Active, unphosphorylated, Rb protein inhibits transcription of several cellular earlyresponse genes including c-fos and c-myc. Initially, it was believed that Rb exerts thissuppressive action by direct interaction with DNA promoter regions. However, evidence nowsuggests that Rb does not bind directly to DNA, but interacts with and regulates transcriptionfactors such E2F397,398 and DRTF1. 399,40° The association of E2F and Rb has beenrigorously studied. Free, or Rb uncomplexed E2F has higher transcriptional activity than Rbbound E2F. In quiescent, serum starved cells or cells synchronized at G1, E2F is sequestered54by unphosphorylated Rb.397 Cell cycle progression is associated with dissociation of thiscomplex. This complex can also be disrupted by the transforming proteins of several DNAtumour viruses including adenovirus ElA, SV40 and polyoma virus large T and humanpapilloma virus E6. The conserved regions in these viral transforming proteins essential fortransformation are also those which are required for interaction with the Rb protein,suggesting that these oncogenes stimulate cell proliferation by sequestering Rb and thusinactivating its growth suppressive activity. Efforts have now turned towards identifying thenormal cellular homologue of the DNA viral oncogenes which mediate the normal transitionfrom Rb bound E2F to active E2F. 395 These predicted cellular Rb binding proteins must thenbe integrated into a model in which disruption of Rb/E2F complexes occurs with cycledependent changes in phosphorylation of the Rb protein.Another piece to add to this puzzle is the observation that E2F complexes with cyclin Aduring S phase. It is therefore tempting to speculate that during the G 1 /S transition, as morecyclin A accumulates,a cyclin A dependent kinase associates with the G 1 Rb/E2F complex andthe resulting phosphorylation of Rb by the CDK then releases Rb. Indeed, in studies withDRTF1,400 a transcription factor probably identical to E2F, Rb and cyclin A have beendetected in the same DRTF1 complex. 399 Adenovirus E1A will also disrupt cyclin A/E2Fcomplexes. 398,399,401 However, the role of cyclin A in the E2F complexes is open tospeculation. Perhaps the association with E2F targets the CDK to S-phase specific targets, orperhaps the association with cyclin A changes the specificity of E2F. The effect of thecombination is, however, not a simple global activation of E2F since cyclin A/E2F complexeshave lower transcriptional activity than free E2F.398 Considerable evidence points to the theimportance of Rb and its interaction with transcription factors, cyclins and CDK's, not only incell cycle control but, as will discussed in the next section, in growth factor responses as well.1. Integration of Growth Factor Control With Cell Cycle ControlUntil very recently, work on growth factor control and cell cycle analysis have progressedindependently. Progress in our understanding of the cell cycle has been achieved primarily via55deletion mutants in simple organisms such as yeast where genetic manipulations are relativelysimple. Growth factor studies have, on the other hand, concentrated on accessible earlybiochemical changes in mammalian cells. Thus, control of the cell cycle was essentially dividedinto two solitudes. Lately this dichotomy has been disappearing as rapid progress is beingmade from both ends of the cycle. We are beginning to to able to integrate growth factorcontrol mechanisms with cell cycle events.An example of this kind of integration is the inhibition by TGF13 of c-myctranscription. 402 TGFI3 also inhibits G i/S cdc2 kinase activity 403 and Rbphosphorylation.4°3 These observations suggest a model of TGFI3 action in which TGFI3inhibition of G i/S cdc2 kinase results in failure to phosphorylate directly, or indirectly, the Rbprotein, a phosphowlation which ordinarily undergoes cell cycle oscillation. Failure toinactivate Rb allows persistence, perhaps, of an inactive Rb/E2F complex so that c-myc is nottranscribed. Consistent with this model is the abrogation of TGFP mediated inhibition of c-myc(which contains an E2F binding motif in its promoter region) transcription when cells aretransfected with Rb binding proteins such as ElA. 4°4A more concrete instance of a growth factor pathway feeding into cell cycle control is therecent, direct demonstration that certain growth factors activate cyclins or CDK's directly.CSF-1, as mentioned earlier, induces the synthesis of cyclin D protein. The continuedpresence of CSF-1 during G1 is required for persistence of cyclin D. Removal of CSF-1 andthus cyclin D at any time before G 1 /S commitment results in cycle arrest. More intriguing isthe involvement of CDK's very early in the signal transduction cascade. Both EGF 405 andNGF406 stimulate activation of a proline directed protein kinase (PDPK). Biochemicalcharacterization of PDPK has shown it to be a cyclin A/cdc2 complex. 405 This surprisingobservation suggests that certain CDK's act much earlier than the G 1 restriction point and apossible role for this factor responsive complex in cell division control will be discussed below.In fact, a role outside of cell division altogether has been hypothesized for p34cdc2 since56platelet activating factor apparently activates cdc2 histone H1 kinase activity in platelets, a cellincapable of undergoing division.407Work in growth factor and cell cycle control has progressed sufficiently such that ourknowledge of the two fields may soon become integrated completely. An important conceptthat has emerged from studies so far, is the high degree of conservation of fundamental growthand cycle control mechanisms from yeast to man. Because of this conservation, there arecertain observations made in yeast that might by integrated with certain growth factorresponses in mammalian cells. For example, in fusion yeast the G 1 transition is mediated byexpression of four CLN genes. Upregulation of the three oscillating cyclins is mediated by theactivation of CLN3. However, since the protein level of CLN3 does not oscillate through thecycle, it must be controlled post-translationally. In mammalian cells, an analogous systemmay exist; G i/S commitment may also be initiated by the activation of a constitutivelyexpressed cyclin. Indeed, although cyclin D1 induction in the first round through the cellcycle requires CSF-1, in subsequent cycles there is a basal, constant level throughout thecycle. Genetic complementation analyses demonstrate that CLN3 is regulated in yeast by theproduct of the FUS3 gene.408 •4°9 Very intriguingly, FUS3 encodes a Ser/Thr kinasehomologous to the mammalian MAP kinases. 307 It is very tempting to speculate that growthfactor stimulation of a MAP kinase activates the basally expressed cyclin which then initiatesthe cyclin transcription positive feedback loop. Alternatively, instead of cyclin regulation byMAP kinase activation similar to the FUS3 kinase signalling system in yeast, mammalian cellscould utilize a cyclin/CDK complex which is more directly coupled to growth factor stimulation.PDPK is exactly such a complex. One can envisage a signal transduction cascade in whichgrowth factor stimulation activates either MAP kinase or PDPK, which then, in a manneranalogous to the events described in yeast, elevates cyclin levels by phosphorylatingtranscription factors which regulate cyclin gene expression. These models are readily testableand the results could be very interesting and may potentially provide a direct path fromreceptor activation to the initialization of a cell's intrinsic cell cycle machinery.57E. THESIS OBJECTIVEAt the time the work described in this thesis was initiated, relatively little was knownabout the receptors for the hemopoietic growth factors. The best characterized receptor, anintrinsic tyrosine kinase receptor, was the one for CSF-1. No member of the hemopoietinreceptor superfamily had yet been cloned. Indeed, studies into the mechanism of action of theinterleukins ( and there were only four at this time), had only just begun with the recentavailability of recombinant growth factors. Thus a logical first step in understanding themechanism of action of mIL-3 was the purification and characterization of its cell surfacereceptor. Once this was accomplished, the mIL-3R could then be studied to see if it utilizedsignalling mechanisms similar to those used by better characterized receptors, such as thosefor CSF-1 and EGF. Specifically, the thesis objectives were to:1. Develop an assay for detecting the detergent solubilized mIL-3R2. Purify the mIL-3R in sufficient amounts for amino acid sequencing.3. Conduct biochemical characterization studies of the purified receptor.4. Investigate the signalling pathways used by this receptor, especially with respect tothe kinase responsible for tyrosine phosphorylation of the mIL-3R58F. REFERENCES1. Metcalf D. The molecular control of cell division, differentiation commitment andmaturation in haemopoietic cells. Nature 1989;339:27-30.2. Moore MAS. Embryologic and phylogenetic development of the hematopoieticsystem. In: Burkhardt R, Conley CL, Lennert K, Adler SS, Pincus T, Till JE.Advances in the Biosciences 16. Dahlem Workshop on Myelofibrosis-OsteosclerosisSyndrome. Vieweg; Pergamon Press, 1975:87-101.3. Eaves CJ, Eaves AC. Fundamental control of hematopoiesis. In: Fisher JW.Biochemical Pharmacology of Blood and Blood Forming Organs. Heidelberg;Springer Verlag, (in press).4. Lemischka IR, Raulet DH, Mulligan RC. Developmental potential and dynamicbehavior of hematopoietic stem cells. Cell 1986;45:917-27.5. Abramson S, Miller RG, Phillips RA. The identification in adult bone marrow ofpluripotent and restricted stem cells of the myeloid and lymphoid systems. J ExpMed 1977;145:1567-79.6. Raskind WH, Fialkow PJ. The use of cell markers in the study of humanhematopoietic neoplasia. Adv Cancer Res 1987;49:127-67.7. Nitta M, Kato Y, Strife A et al. Incidence of involvement of the B and T lymphocytelineages in chronic myelogenous leukemia. Blood 1985;66:1053.8. Turhan AG, Humphries RK, Phillips GL, Eaves AC, Eaves CJ. Clonal hematopoiesisdemonstrated by X-linked DNA polymorphisms after allogeneic bone marrowtransplantation. N Engl J Med 1989;320:1655-61.9. Till JE, McCulloch EA. A direct measurement of the radiation sensitivity of normalmouse bone marrow cells. Radiat Res 1961;14:213-22.10. Kaplan SS, Basford RE, Wing EJ, Shadduck RK. The effect of recombinant humangranulocyte macrophage colony-stimulating factor on neutrophil activation inpatients with refractory carcinoma. Blood 1989;73:636-8.11. Becker AJ, McCulloch EA, Till JE. Cytological demonstration of the clonal nature ofspleen colonies derived from transplanted mouse marrow cells. Nature1963;197:452-4.12. Szilvassy SJ, Lansdorp PM, Humphries RK, Eaves AC, Eaves CJ. Isolation in asingle step of a highly enriched murine hematopoietic stem cell population withcompetitive long-term repopulating ability. Blood 1989;74:930-9.13. Szilvassy SJ, Humphries RK, Lansdorp PM, Eaves AC, Eaves CJ. Quantitativeassay for totipotent reconstituting hematopoietic stem cells by a competitiverepopulation strategy. Proc Natl Acad Sci U S A 1990;87:8736-40.14. Nakahata T, Ogawa M. Identification in culture of a class of hemopoieticcolony-forming units with extensive capability to self-renew and generatemultipotential hemopoietic colonies. Proc Natl Acad Sci U S A 1982;79:3843-7.5915. Sutherland HJ, Lansdorp PM, Henkelman DH, Eaves AC, Eaves CJ. Functionalcharacterization of individual human hematopoietic stem cells cultured at limitingdilution on supportive marrow stromal layers. Proc Natl Mad Sci U S A1990;87:3584-8.16. Cannistra SA, Griffin JD. Regulation of the production and function of granulocytesand monocytes. Semin Hematol 1988;25:173-88.17. Leary AG, Straus LC, Civin CI. Disparate differentiation in hemopoietic coloniesderived from human paired progenitors. Blood 1985;66:327-32.18. Eaves CJ, Humphries RK, Eaves AC. In vitro characterization of erythroid precursorcells and the erythropoietic differentiation process. In: Stamatoyannopoulos G,Nienhuis AW. Cellular and Molecular Regulation of Hemoglobin Switching. NewYork; Grune and Stratton, 1979:251-73.19. Keller G, Paige C, Gilboa E, Wagner EF. Expression of a foreign gene in myeloid andlymphoid cells derived from multipotent haematopoietic precursors. Nature1985;318:149-54.20. Campbell AD, Long MW, Wicha MS. Haemonectin, a bone marrow adhesion proteinspecific for cells of granulocyte lineage. Nature 1987;329:744-6.21. Tsai S, Patel V, Beaumont E, Lodish HF, Nathan DG, Sieff CA. Differential bindingof erythroid and myeloid progenitors to fibroblasts and fibronectin. Blood1987;69:1587-94.22. Curry JL, Trentin JJ. Hemopoietic spleen colony studies: I. Growth anddifferentiation. Dev Biol 1967;15:395-413.23. Lorimore SA, Wright EG. Stem cell maintenance and commitment to differentiationin long-term cultures of murine marrow obtained from different spatial locations inthe femur. Cell Tissue Kinet 1987;20:427-34.24. Coulombel L, Eaves AC, Eaves CJ. Enzymatic treatment of long-term humanmarrow cultures reveals the preferential location of primitive hemopoieticprogenitors in the adherent layer. Blood 1983;62:291-7.25. Spooncer E, Heyworth CM, Dunn A, Dexter TM. Self-renewal and differentiation ofinterleukin-3-dependent multipotent stem cells are modulated by stromal cells andserum factors. Differentiation 1986;31:111-8.26. Roberts R, Gallagher J, Spooncer E, Allen TD, Bloomfield F, Dexter TM. Heparansulphate bound growth factors: A mechanism for stromal cell mediatedhaemopoiesis. Nature 1988;332:376-8.27. Gordon MY, Riley GP, Watt SM, Greaves MF. Compartmentalization of ahaematopoietic growth factor (GM-CSF) by glycosaminoglycans in the bone marrowmicroenvironment. Nature 1987;326:403-5.28. Rathjen PD, Toth S, Willis A, Heath JK, Smith AG. Differentiation inhibiting activityis produced in matrix-associated and diffusible. Cell 1990;62:1105-14.6029. Rettenmier CW, Roussel MF, Ashmun RA, Ralph P, Price K, Sherr CJ. Synthesis ofmembrane-bound colony-stimulating factor 1 (CSF-1) and downmodulation ofCSF-1 receptors in NIH 3T3 cells transformed by cotransfection of the human CSF-1and c-fms (CSF-1 receptor) genes. Mol Cell Biol 1987;7:2378-87.30. Flanagan JG, Chan DC, Leder P. Transmembrane form of the kit li and growthfactor is determined by alternative splicing and is missing in the SIc7 mutant. Cell1991;64:1025-35.31. Metcalf D. The hemopoietic colony stimulating factors. Amsterdam; Elsevier, 1984.32. Krystal G, Alai M, Cutler RL, Dickeson H, Mui ALF, Wognum AW. Hematopoieticgrowth factor receptors. Hematol Pathol 1991;5:141-62,33. Holbrook ST, Ohis RK, Schibler KR, Yang YC, Christensen RD. Effect of interleukin9 on clonogenic maturation and cell cycle status of fetal and adult hematopoieticprogenitors. Blood 1991;77:2129-34.34. Thompson-Snipes L, Dhar V, Moore K, Bond M, Rennick D. A T cell derivedlymphokine termed cytokine synthesis inhibitory factor (CSIF) (interleukin-10) is apotent costimulant of mast cell growth [abstract). Exp Hematol 1990;18:612.35. Go NF, Castle BE, Kastelein R et al. Interleukin 10, a novel B cell stimulatoryfactor: unresponsiveness of X chromosome-linked immunodeficiency. J Exp Med1990;172:1625-31.36. Gubler U, Chua AO, Schoenhaut DS et al. Coexpression of two distinct genes isrequired to generate secreted bioacive cytotoxic lymphocyte maturation factor. ProcNatl Acad Sci U S A 1991;88:4143-7.37. Bruno E, Briddell RA, Cooper RJ, Hoffman R Effects of recombinant interleukin 11on human megakaryocytic progenitor cells. Exp Hematol 1991;19:378-81.38. Musashi M, Yang Y-C, Paul SR, Clark SC, Sudo T, Ogawa M. Direct and synergisticeffects of interleukin 11 on murine hemopoiesis in culture. Proc Natl Acad Sci U S A1991;88:765-9.39. Heath JK, Smith AG, Hsu LW, Rathjen PD. Growth and differentiation factors ofpluripotent stem cells. J Cell Sci 1990;13:75-85.40. Gough NM, Williams RL. The pleiotropic actions of leukemia inhibitory factor.Cancer Cells 1989;1:77-80.41. Graham GJ, Wright EG, Hewick R et al. Identification and characterization of aninhibitor of haemopoietic stem cell proliferation. Nature 1990;344:442-4.42. Axeirad AA. Some hemopoietic negative regulators. Exp Hematol 1990;18:143-50.43. Arai K, Lee F, Miyajima A, Miyatake S, Arai N, Yokota T. Cytokines: Coordinators ofimmune and inflammatory responses. Annu Rev Biochem 1990;59:783-836.44. Schrader JW. The panspecific hemopoietin of activated T lymphocytes(interleukin-3). Annu Rev Immunol 1986;4:205-30.6145. Wodnar-Filipowicz A, Heusser CH, Moroni C. Production of the haemopoieticgrowth factors GM-CSF and interleukin-3 by mast cells in response to IgEreceptor-mediated activation. Nature 1989;339:150-2.46. Farrar WL, Vinocour M, Hill JM. In situ hybridization histochemistry localization ofinterleukin-3 mRNA in mouse brain. Blood 1989;73:137-40.47. Ikebuchi K, Wong GG, Clark SC, Ihle JN, Hirai Y, Ogawa M. Interleukin 6enhancement of interleukin 3-dependent proliferation of multipotential hemopoieticprogenitors. Proc Natl Acad Sci U S A 1987;84:9035-9.48. Anderson DM, Lyman SD, Baird A et al. Molecular cloning of mast cell growthfactor, a hematopoietin that is active in both membrane bound and soluble forms.Cell 1990;63:235-43.49. Leary AG, Zeng HO, Clark SC, Ogawa M. Growth factor requirements for survival inGo and entry into the cell cycle of primitve human hemopoeiticprogentors [abstract]. Blood 1991;78 Suppl 1:159a.50. Metcalf D, Begley CG, Johnson GR Effects of purified baterially synthesized murinemulti-CSF(IL-3) on hematopoiesis in normal adult mice. Blood 1986;68:46-57.51. Keller JR, Ihle JN. Unique pathway of IL-3 driven hemopoietic differentiation. JImmunol 1989;143:4025-33.52. Mathews IS, Vale WW. Expression cloning of an activin receptor, a predictedtransmembrane serine kinase. Cell 1991;65:973-82.53. Georgi LL, Albert PS, Riddle DL. daf-1, a C. elegans gene controlling dauer larvadevelopment, encodes a novel receptor protein kinase. Cell 1990;61:635-45.54. Schulz S, Singh S, Belle RA et al. The primary structure of a plasma membraneguanylate cyclase demonstrates diversity within this new receptor family. Cell1989;58:1155-62.55. Yarden Y, Ullrich A. Growth factor receptor tyrosine kinases. Annu Rev Biochem1988;57:443-78.56. Williams LT. Signal transduction by the platelet-derived growth factor receptor.Science 1989;243:1564-70.57. Sherr CJ, Rettenmier CW, Sacca R, Roussel MF, Look AT, Stanley ER The c-fmsproto-oncogene product is related to the receptor for the mononuclear growth factor,CSF-1. Cell 1985;41:665-76.58. Ogawa M, Matsuzaki Y, Nishikawa S et al. Expression and function of c-kit inhemopoietic progenitor cells. J Exp Med 1991;174:63-71.59. Matthews W, Jordan CT, Wiegand GW, Pardoll D, Lemischka IR A receptor tyrosinekinase specific to hematopoietic stem and progenitor cell-enriched populations. Cell1991;65:1143-52.60. Sherr CJ. Colony-stimulating factor-1 receptor. Blood 1990;75:1-12.6261. Bortin MM, Gale RP, Rimm AA. Allogeneic bone marrow transplantation for 144patients with severe aplastic anemia. JAMA 1981;245:1132-9.62. Roussel MF, Downing JR, Rettenmier CW, Sherr CJ. A point mutation in theextracellular domain of the human CSF-1 receptor (c-fms proto-oncogene product)activates its transforming potential. Cell 1988;55:979-88.63. Zsebo KM, Williams DA, Geissler EN et al. Stem cell factor is encoded at the Sllocus of the mouse and is the ligand for the c-kit tyrosine kinase receptor. Cell1990;63:213-24.64. Downing JR, Rettenmier CW, Sherr CJ. Ligand induced tyrosine kinase activity ofthe colony stimulating factor 1 recptor in a murine macrophage cell line. Mol CellBiol 1988;8:1795-9.65. Henegger A, Dull TJ, Szapary D et al. Kinetic parameters of the protein tyrosinekinase activity of EGF receptor mutants with individually alteredautophosphorylation sites. EMBO J 1988;7:3053-60.66. Felder S, Miller K, Moehren G, Ullrich A, Schlessinger J, Hopkins CR Kinaseactivity controls the sorting of the epidermal growth factor receptor within themultivesicular body. Cell 1990;61:623-34.67. Lammers R, Gray A, Sschlessinger J, Ullrich A. Differential signalling potential ofinsulin and IGF-1 receptor cytoplasmic domains. EMBO J 1989;8:1369-75.68. Lee J, Dull TJ, Laz I, Schlessinger J, Ullrich A. HER2 cytoplasmic domain generatesnormal mitogenic and transforming signals in a chimeric receptor. EMBO J1989;8:167-73.69. Kashles 0, Szapary D, Bellot F, Ullrich A, Schlessinger J, Schmidt A.Ligand-induced stimulation of epidermal growth factor receptor mutants withaltered transmembrane regions. Proc Natl Acad Sci U S A 1988;85:9567-71.70. O'Brien RM, Soos MA, Siddle K. Monoclonal antibodies to the insulin receptorstimulate the intrinsic tyrosine kinase activity by cross-linking recptor molecules.EMBO J 1987;6:4003-10.71. Hurwitz DR, Emanuel SL, Nathan MH et al. EGF induces increased ligand bindingaffinity and dimerization of soluble epidermal growth factor (EGF) receptorextracellular domain. J Biol Chem 1991;266:22035-43.72. Greenfiel C, Hiles I, Waterfield MD et al. Epidermal growth factor binding induces aconformational change in the external domain of its receptor. EMBO J1989;8:4115-23.73. Bishayee S, Majumdar S, Khire J, Das M. Ligand induced dimerization of theplatelet derived growth factor receptor. J Biol Chem 1989;264:11699-705.74. Heldin C-H, Ernlund A, Rorsman C, Ronnstrand L. Dimerization of B-type platletderived growth factor receptors occurs after ligand binding and is closely associatedwith receptor kinase activation. J Biol Chem 1989;264 Suppl 8905:8912.6375. Hammacher A, Mellstrom K, Heldin C-H, Westermark B. Isoform-specific inductionof actin reorganization by platelet-derived growth factor suggests that thefunctionally active receptor is a dimer. EMBO J 1989;8:2489-95.76. Weiner DB, Liu J, Cohen JA, Williams WV, Greene MI. A point mutation in the neuoncogene mimics ligand induction of receptor aggregation. Nature 1989;339:230-1.77. Ballotti R, Lammers R, Scimeca J-C et al. Intermolecular transphosphorylationbetween insulin receptors and EGF-insulin receptor chimerae. EMBO J1989;8:3303-9.78. Cobb MH, Sang B-C, Gonzalez R, Goldsmith E, Ellis L. Autophosphorylationactivates the soluble cytoplasmic domain of the insulin receptor in anintermolecular reaction. J Biol Chem 1989; Suppl 264:18701-6.79. Honegger AM, Schmidt A, Ullrich A, Schlessinger J. Evidence for epidermal growthfactor (EGF) induced intermolecular autophosphorylation of the EGF receptors inliving cells. Mol Cell Biol 1990;10:4035-44.80. Mayer BJ, Jackson PK, Baltimore D. The noncatalytic src homology region 2segment of abl tyrosine kinase binds to tyrosine-phosphorylated cellular proteinswith high affinity. Proc Natl Acad Sci USA  1991;88:627-31.81. Kazlauskas A, Cooper JA. Autophosphorylation of the PDGF receptor in the kinaseinsert region regulates interactions with cell proteins. Cell 1989;58:1121-33.82. Margolis B, Koch A, Mohammadi M et al. The tyrosine phosphorylatedcarboxyterminus of the EGF receptor is a binding site for GAP and PLC gamma.EMBO J 1990;9:4375-80.83. Morrison DK, Kaplan DR, Escobedo JA, Rapp UR, Roberts TM, Williams LT. Directactivation of the serine/threonine kinase activity of raf-1 through tyrosinephosphorylation by the PDGF receptor. Cell 1989;58:649-57.84. Koch CA, Anderson D, Moran MF, Ellis C, Pawson T. SH2 and SH3 domains:Elements that control interaction of cytoplasmic signaling proteins. Science1991;252:668-74.85. Moran MF, Koch CA, Anderson D et al. Src homology region 2 domains directprotein-protein interactions in signal transduction. Proc Natl Acad Sci U S A1990;87:8622-6.86. Anderson D, Koch CA, Grey L, Ellis C, Moran MF, Pawson T. Binding of SH2domains of phospholipase C gamma 1, GAP and src to activated growth factorreceptors. Science 1990;250:979-82.87. Kazlauskas A, Ellis C, Pawson T, Cooper JA. Binding of GAP to activated PDGFreceptors. Science 1990;247:1578-81.88. Sherr CJ. Mitogenic response to colony-stimulating factor 1. Trends Biochem Sci1991;7:398-402.89.^Galizzi J-P, Castle B, Djossou 0 et al. Purification of a 130 kDa T cell glycoproteinthat binds human interleukin 4 with high affinity. J Biol Chem 1990;265:439-44.90. Goodwin RG, Friend D, Ziegler SF et al. Cloning of the human and murineinterleukin-7 receptors: Demonstration of a soluble form and homology to a newreceptor superfamily. Cell 1990;60:941-51.91. Seed B, Aruffo A. AAA. Proc Natl Acad Sci U S A 1987;84:8573-7.92. Miyajima A, Kitamura T, Harada N, Yokota T, Arai K. Cytokine receptors and signaltransduction. Annu Rev Immunol 1991;10:295-331.93. Patthy L. Homology of a domain of the growth hormone/prolactin receptor familywith Type III modules of fibronectin. Cell 1990;61:13-4.94. Bazan JF. Structural design and molecular evolution of a cytokine receptorsuperfamily Proc Natl Acad Sci U S A 1990;87:6934-8.95. Wang H-M, Ogorochi T, Arai K-I, Miyajima A. Sturcture of mouse interleukin 3(IL-3) binding protein (AIC2A). J Biol Chem (in press).96. Miyazaki T, Maruyama M, Yamada G, Hatakeyama M. The integrity of theconserved WS motif common to IL-2 and other cytokine receptors is essential forligand binding and signal transduction. EMBO J 1991;10:3191-7.97. Fukunaga R, Ishizaka-Ikelda E, Nagata S. Purification and characterisation of thereceptor for murine granulocyte colony-stimulating factor. J Biol Chem1990;265:14008-15.98. Fukunaga R, Ishizaka-Ikeda E, Seto Y, Nagata S. Expression cloning of a receptorfor murine granulocyte colony-stimulating factor. Cell 1990;61:341.99. Fukunaga R, Ishizaka-Ikeda E, Pan C, Seto Y, Nagata S. Functional domains of thegranulocyte colony-stimulating factor receptor. EMBO J 1991;10 Suppl 10:2855-65.100. Idzerda RL, March CJ, Mosley B et al. Human interleukin 4 receptor confersbiological responsiveness and defines a novel receptor superfamily. J Exp Med1990;171:861-73.101. Park LS, Friend D, Sassenfeld HM, Urdal DL. Characterization of the human B-cellstimulatory factor 1 receptor. J Exp Med 1987;166:476-88.102. Mosley B, Beckmann MP, March CJ et al. The murine interleukin-4 receptor:Molecular cloning and characterization of secreted and membrane bound forms.Cell 1989;59:335-48.103. Harada N, Castle BE, Gorman DM et al. Expression cloning of a cDNA encoding themurine interleukin 4 receptor based on ligand binding. Proc Natl Acad Sci U S A1990;87:857-61.104. Yamasaki K, Taga T, Hirata Y et al. Cloning and expression of the humaninterleukin-6 (BSF-2/IFN(3 2) receptor. Science 1988;241:825-8.105. Hibi M, Murakami M, Saito M, Hirano T, Taga T, Kishimoto T. Molecular cloningand expression of an IL-6 signal transducer, gp 130. Cell 1990;63:1149-57.65106. Takaki S, Mita S, Kitamura T et al. Identification of the second subunit of themurine interleukin-5 receptor: interleukin-3 receptor-like protein, AIC2B is acomponent of the high affinity interleukin-5 receptor. EMBO J 1991;10:2833-8.107. D'Andrea AD, Lodish HF, Wong GG. Expression cloning of the murineerythropoietin receptor. Cell 1989;57:277.108. Winkelmann J. Penny L, Deavan L, Forget B, Jenkins R. The gene for the humanerythropoietin receptor: Analysis of the coding sequence and assignment tochromosome 19p. Blood 1990;76:24-30.109. D'Andrea AD, Zon LI. Erythropoietin receptor subunit structure and activation. JClin Invest 1990;86:681-7.110. Yoshimura A, Longmore G, Lodish HF. Point mutation in the exoplasmic domain ofthe erythropoietin receptor resulting in hormone-independent activation andtumorigenicity. Nature 1990;348:647-9.111. D'Andrea Ad, Yoshimura A, Youssoufian H, Zon LI, Koo J-W, Lodish H. Thecytoplasmic region of the erythropoietin receptor contains nonoverlapping positiveand negative growth-regulatory domains. Mol Cell Biol 1991;11:1980-7.112. guelle DE, Wojchowski DM. Localized cytosolic domains of the erythropoietinreceptor regulate growth signaling and down-modulate responsiveness togranulocyte-macrophage colony-stimulating factor. Proc Natl Acad Sci U S A1991;88:4801-5.113. Gearing DP, Thut CJ, VandenBos T et al. Leukemia inhibitory factor receptor isstructurally related to the IL-6 signal transducer, gp 130. EMBO J1991;10:2839-48.114. Lord KA, Abdollahi A, Thomas SM et al. Leukemia inhibitory factor andinterleukin-6 trigger the same immediate early response, including tyrosinephosphorylation, upon induction of myeloid leukemia differentiation. Mol Cell Biol1991;11:4371-9.115. Colamonici OR, Neckers LM, Rosolen A. Putative gamma subunit of the IL-2receptor is detected in low, intermediate, and high affinity IL-2 receptor bearingcells. J Immunol 1990;145:155-60.116. Sharon M, Gnarra JR, Leonard WJ. A 100 kilodalton protein is associated with themurine interleukin 2 receptor: Biochemical evidence that p100 is distinct from thealpha and beta chains. Proc Natl Acad Sci U S A 1990;87:4869-73.117. Walker F, Nicola NA, Metcalf D, Burgess AW. Hierarchical down-modulation ofhemopoietic growth factor receptors. Cell 1985;43:269-76.118. Park LS, Friend D, Price V et al. Heterogeneity in human interleukin-3 receptors. JBiol Chem 1989;264:5420-7.119. Itoh N, Ybnehara S, Schreurs J et al. Cloning of an interleukin-3 receptor gene: Amember of a distinct receptor gene family. Science 1990;247:324-7.120. Yonehara S, Ishii A, Yonehara M et al. Identification of a cell surface 105 kd protein(Aic-2 antigen) which binds interleukin-3. Int Immunol 1990;2:143-50.121. Gorman DM, Itoh N, Kitamura T et al. Cloning and expression of a gene encodingan interleukin 3 receptor-like protein: Identification of another member of thecytokine receptor gene family. Proc Natl Acad Sci U S A 1990;87:5459-63.122. Takaki S, Tominaga A, Hitoshi Y et al. Molecular cloning and expression of themurine interleukin-5 receptor. EMBO J 1990;9:4367-74.123. Devos R, Plaetinck G, Van der Heyden J et al. Molecular basis of a high affinitymurine interleulcin-5 receptor. EMBO J 1991;10:2133-7.124. Kitamura T, Hayashida K, Sakamaki K, Yokota T, Arai K-I, Miyajima A.Reconstitution of functional receptors for human granulocyte/macrophagecolony-stimulating factor (GM-CSF): Evidence that the protein encoded by theAIC2B cDNA is a subunit of the murine GM-CSF receptor. Proc Natl Arad Sci U S A1991;88:5082-6.125. Hayashida K, Kitamura T, Gorman DM, Arai K-I, Yokota T, Miyajima A. Molecularcloning of a second subunit of the receptor for human granulocyte-macrophagecolony-stimulating factor (GM-CSF): Reconstitution of a high-affinity GM-CSFreceptor. Proc Natl Acad Sci U S A 1990;87:9655-9.126. Tavenier J, Devos R, Cornelis S et al. A human high affinty interleukin-5 receptor(IL5R) is composed of an III specific alpha chain and a beta chain shared with thereceptor for GM-CSF. Cell 1991;66:1175-84.127. Metcalf D, Nicola N, Gearing D, Gough N. Low-affinity placenta-derived receptorsfor human granulocyte-macrophage colony-stimulating factor can deliver aproliferative signal to murine hempoeitic cells. Proc Natl Acad Sci U S A1990;87:4670-4.128. Raines M, Liu L, Quan S, Joe V, Dipersio J, Golde D. Identification and molecularcloning of a soluble human granulocyte-macrophage colony stimulating factorreceptor. Proc Natl Acad Sci U S A 1991;88:8203-7.129. Park IS, Waldron PE, Friend D et al. Interleukin-3, GM-CSF, and G-CSF receptorexpression on cell lines and primary leukemia cells: Receptor heterogeneity andrelationship to growth factor responsiveness. Blood 1989;74:56-65.130. Kitamura T, Sato N, Arai K-I, Miyajima A. Expression cloning of the human IL-3receptor cDNA reveals a shared R subunit for the human IL-3 and GM-CSFreceptors. Cell 1991;66:1165-74.131. Souyri M, Vigon I, Penciolelli J-F, Heard J-M, Tambourin P, Wendling F. A putativetruncated cytokine receptor gene transduced by the myeloproliferative leukemiavirus immortalizes hematopoietic progenitors. Cell 1990;63:1137-47.132.^Djiane J, Dusanter-Fourt I, Katoh M, Kelly PA. Biological activities of binding sitespecific monoclonal antibodies to prolactin receptors of rabbit mammary gland. JBiol Chem 1985;260:11430-5.67133. Cunningham BC, Ultsch M, de Vos AM, Mulkerrrin MG, Clauser KR, Wells JA.Dimerization of the extracellular domain of the human growth hormone receptor bya single hormone molecule. Science 1991;254:821-5.134. Carter-Su c, Stubbart JR, Wang X, Stred SE, Argetsinger LS, Shafer JA.Phosphorylation of highly purified growth hormone receptors by a growth hormonereceptor associated tyrosine kinase. J Biol Chem 1989;264:18654-61.135. Hatakeyama M, Kono T, Kobayashi N et al. Interaction of the IL-2 receptor with thesrc-family kinase p56 lck: identification of novel intermolecular association. Science1991;252:1523-8.136. Remillard B, Petrillo R, Malinski W et al. Interleukin-2 receptor regulates activationof phosphatidylinositol 3-kinase. J Biol Chem 1991;266:14167-70.137. Vairo G, Hamilton JA. Activation and proliferation signals in murine macrophages:Stimulation of Na and K-ATPase activity by hemopoietic growth factors and otheragents. J Cell Physiol 1988;134:13-24.138. Grinstein S, Rotin D, Mason MJ. Na/H exchange and growth factor-inducedcytosolic pH changes. Role in cellular proliferation. Biochim Biophys Acta1989;988:73-97.139. Cantley LC, Auger KR, Carpenter C et al. Oncogenes and signal transduction. Cell1991;64:281-302.140. Bravo R Growth factor-responsive genes in fibroblasts. Cell Growth Differ1990;1:305-9.141. Pardee AB. G1 events and regulation of cell proliferation. Science 1989;246:603-8.142. Bourne HR, Sanders DA, McCormick F. The GTPase superfamily: a conservedswitch for diverse cell functions. Nature 1990;348:125-32.143. Casey PJ, Gilman AG. G protein involvement in receptor-effector coupling. J BiolChem 1988;263:2577-80.144. Spiegel AM, Backlund PS, Butrynski JE, Jones TLZ, Simonds WF. The G proteinconnection: molecular basis of membrane association. Trends Biochem Sci1991;16:338-41.145. Birnbaumer L. G proteins in signal transduction. Ann Rev Biochem1990;30:675-705.146. Gilman AG. G proteins and dual control of adenylate cyclase. Cell 1984;36:577-9.147. Neer EJ, Clapham DE. Roles of G protein subunits in transmembrane signalling.Nature 1988;333:129-34.148. Lechleiter J, Hellmiss R, Duerson K et al. Distinct sequence elements control thespecificity of G protein activation by muscarinic acetylcholine receptor subtypes.EMBO J 1990;9:4381-90.68149. Okamoto T, Nishimoto I. Analysis of stimulation-G protein subunit coupling byusing active insulin-like growth factor II receptor peptide. Proc Natl Acad Sci U S A1991;88:8020-3.150. Lowndes JM, Guptua SK, Osawa S, Johnson GL. GTPase-deficient Gait oncogenegip2 inhibits adenylylcyclase and atttenuates receptor-stimulated phospholipase A2activity. J Biol Chem 1991;266:14193-7.151. Chee MS, Satchwell SC, Preddie E, Weston KM, Barren BG. Humancytomegalovirus encodes three G protein-coupled receptor homologues. Nature1990;344:774-7.152. Barbacid M. ras genes. Ann Rev Biochem 987;56:779-827.153. Der CJ, Cox AD. Isoprenoid modification and plasma membrane association:critical factors for ras oncogenicity. Cancer Cells 1991;3:331-40.154. Milburn MV, Tong L, DeVos AM et al. Molecular switch for signal transduction:structural differences between active and inactive forms of protooncogenic rasproteins. Science 1990;247:939-45.155. Madaule P, Axel R A novel ras-related gene family. Cell 1985;41:31-40.156. Lowe DG, Capon DJ, Delwart E, Sakaguchi AY, Naylor SL, Goeddel DV. Structureof the human and murine R-ras genes, novel genes closely related to rasproto-oncogenes. Cell 1987;48:137-46_157. Hall A. The cellular functions of small GTP-binding proteins. Science1990;249:635-40.158. Satoh T, Nakafuku M, Miyajima A, Kaziro Y. Involvement of ras p21 protein insignal-transduction pathways from interleukin 2, interleukin 3, andgranulocyte/macrophage colony-stimulating factor, but not from interleukin 4. ProcNatl Acad Sci U S A 1991;88:3314-8.159. Gibbs JB, Marshall MS, Scolnick EM, Dixon RAF, Vogel US. Modulation of guaninenucleotides bound to ras in NIH3T3 cells by oncogenes, growth factors, and theGTPase activating protein (GAP). J Biol Chem 1990;265:20437-42.160. Satoh T, Endo M, Nakafuku M, Nakamura S, Kazira Y. Platelet-derived growthfactor stimulates formation of active p2lrasGTP complex in swiss mouse 3T3 cells.Proc Natl Acad Sci U S A 1990;87:5993-7.161. Mulcahy LS, Smith MR, Stacey DW. Requirement for ras proto-oncogene functionduring serum-stimulated growth of NIH 3T3 cells. Nature 1985;313:241-3.162. Benito M, Porras A, Nebreda AR, Santos E. Differentiation of 3T3-L1 fibroblasts toadipocytes induced by transfection of ras oncogenes. Science 1991;253:565.163.^Smith MR, DeGudicibus SJ, Stacey DW. Requirement for c-ras proteins during viraloncogene transformation. Nature 1986;320:540-3.69164. Akin JG, Wetss R, Bradshaw RA. Microinjection of a p2 lras antibody into PC12cells inhibits neurite outgrowth induced by nerve growth factor and basic fibroblastgrowth factor. UNLISTED 1991;4:145-55.165. Smith MR, Liu Y-L, Kim H, Rhee SG, Kung H-F. Inhibition of serum- and rasstimulated DNA synthesis by antibodies to phospholipase C. Science1990;247:1074-7.166. Binetruy B, Smeal T, Karin M. Ha-ms augments c-jun activity and stimulatesphosphorylation of its activation domain. Nature 1991;351:122-7.167. Goldschmidt-Clermont PJ, Janmey PA. Profilin, a weak CAP for actin and ras. Cell1991;66:419-21.168. Vojtek A, Haarer B, Field J et al. Evidence for a functional link between profilin andCAP in the yeast S. cerevisiae. Cell 1991;66:497-505.169. Goldschmidt-Clermont PJ, Kim JW, Machesky LM, Rhee SG, Pollard TD. Regulationof phospholipase C-y-1 profilin and tyrosine phosphorylation. Science1991;251:1231-3.170. Telfer JC, Rudd CE. A 32 kD GTP-binding protein associated with the CD4-p561ckand CD8-p561ck T cell receptor complexes. Science 1991;254:439-42.171. McCormick F. ras GTPase activating protein: Signal transmitter and signalterminator. Cell 1989;56:5-8.172. Trahey M, McCormick F. A cytoplasmic protein stimulates normal N-ms p21GTPase, but does not affect oncogenic mutants. Science 1987;238:542-5.173. Trahey M, Wong G, Halenbeck R et al. Molecular cloning of two types of GAPcomplementary DNA form human placenta. Science 1988;242:1697-700.174. Tanaka K, Nakafuku M, Satoh T, Matsumoto K, Kaziro Y, Toh-e A. S. cerevisiaegenes IRA1 and IRA2 encode proteins that may be functionally equivalent tomammalian ras GTPase activating protein. Cell 1990;60:803-7.175. Rubinfeld B, Munemltsu S, Clark R et al. Molecular cloning of a GTPase activatingprotein specific for the Krev-1 protein p21 rapl. Cell 1991;65:1033-42.176. Garrett MD, Self AF, van Oers C, Hall A. Identification of distinct cytoplasmictargets for ras/R-ras and rho regulatory proteins. J Biol Chem 1989;264:10-3.177. Burstein ES, Linko-Stenz K, Lu Z, Macara IG. Regulation of the GTPase activity ofthe ras-like protein p25rab3A. J Biol Chem 1991;266:2689-92.178. Emkey R, Freedman S, Feig LA. Characterization of a GTPase-activating protein forthe ms-related ral protein. J Biol Chem 1991;266:9703-6.179. Ballester R, Marchuk D, Boguski M et al. The NF1 locus encodes a proteinfunctionally related to mammalian GAP and yeast IRA proteins. Cell 1990;63:851-9.180.^Martin GA, Viskochil D, Bollag G et al. The GAP-related domain of theneurofibromatosis tyep 1 gene product interacts with ras p21. Cell 1990;63:843-9.70181. Diekmann D, Brill S, Garrett MD et al. Bcr encodes a GTPase-activating protein forp2 lrac. Nature 1991;351 Suppl 400:402.182. Tsai M-H, Yu C-L, Stacey DW. A cytoplasmic protein inhibits the GTPase activity ofH-Ras in a phospholipid-dependent manner. Science 1990;250:982-5.183. Zhang K, DeClue JE, Vass WC, Papageorge AG, McCormick F, Lowy DRSuppression of c-ras transformation by GTPase-activating protein. Nature1990;346:754-6.184. Downward J, Graves JD, Warne PH, Rayter S, Cantrell DA. Stimulation of p2lrasupon T-cell activation. Nature 1990;346:719-23.185. Liu X, Pawson T. The epidermal growth factor receptor phosphorylatesGTPase-activating protein (GAP) at tyr-460, adjacent to the GAP SHS domains. MolCell Biol 1991;11:2511-6.186. Moran MF, Polakis P, McCormick F, Pawson T, Ellis C. Protein-tyrosine kinasesregulate the phosphorylation, protein interactions, subcellular distribution, andactivity of p21 ras GTPase-activating protein. Mol Cell Biol 1991;11:1804-12.187. Molloy CJ, Bottaro DP, Fleming TP, Marshall MS, Gibbs JB, Aaronson SA. PDGFinduction of tyrosine phosphorylation of STPase activating protein. Nature1989;342:711-4.188. Kaplan DR, Morrison DK, Wong G, McCormick F, Williams LT. PDGF beta-receptorstimulates tyrosine phosphorylation of GAP and association of GAP with a signallingcomplex. Cell 1990;61:125-33.189. Ellis C, Moran M, McCormick F, Pawson T. Phosphorylation of GAP andGAP-associated proteins by transforming and mitogenic tyrosine kinases. Nature1990;343:377-81.190. Fazioli F, Kim U-H, Rhee SG, Molloy CJ, Segatto 0, Di Fiore PP. The erbB-2mitogenic signaling pathway: Tyrosine phosphorylation of phospholipase C-gammaand GTPase-activating protein does not correlate with erbB-2 mitogenic potency.Mol Cell Biol 1991;11:2040-8.191. Reedijk M, Liu X, Pawson T. Interactions of phosphatidylinositol kinase, GTPaseactivating protein (GAP) and GAP associated proteins with the colony stimulatingfactor 1 receptor. Mol Cell Biol 1990;10:5601-8.192. Nori M, Vogel US, Gibbs JB, Weber MJ. Inhibition of v-src induced transformationby a GTPase activating protein. Mol Cell Biol 1991;11:2812-8.193. Yu C-L, Tsai M-H, Stacey DW. Serum stimulation of NIH 3T3 cells induces theproduction of lipids able to inhibit GTPase-activating protein activity. Mol Cell Biol1990;10:6683-9.194. Adarai H, Lowy DR, Willumsen BM, Der CJ, McCormick F. Guanosinetriphosphatase activating protein (GAP) interacts with the p21 ras effector bindingdomain. Science 1988;240:518-21.71195. Maruta H, Holden J, Sizeland A, D'Abaco G. The residues of ras and rap proteinsthat determine their GAP specificities. J Biol Chem 1991;266:11661-8.196. Yatani A, Okabe K, Po'aids P. Halenbeck R. McCormick F, Brown AM. ras p21 andGAP inhibit coupling of muscarinic receptors to atrial K+ channels. Cell1990;61:769-76.197. Marshall MS, Hill WS, Ng AS et al. A c-terminal domain of GAP is sufficient tostimulate ras p21 GTPase activity. EMBO J 1989;8:1105-1110x.198. Ruggieri R, McCormick F. Ras and the awd couple. Nature 1991;353:390-1.199. Randazzo PA, Northup JK, Kahn RA. Activation of a small GTP-binding protein bynucleoside diphosphate kinase. Science 1991;254:850-3.200. Wolfman A, Macara IG. A cytosolic protein catalyzes the release of GDP fromp2lras. Science 1990;248:67-9.201. Huang YK, Kung H-F, Kamata T. Purification of a factor capable of stimulating theguanine nucleotide exchange reaction of ras proteins and its effect on ras-relatedsmall molecular mass G proteins. Proc Natl Mad Sci U S A 1990;87:8008-12.202. Mizuno T, Kaibuchi K, Yamamoto T et al. A stimulatory GDP/GTP exchange proteinfor smg p21 is active on the post-translationally processed form of c-Ki-ras p21 andshoA p21. Proc Natl Acad Sci U S A 1991;88:6442-6.203. Bischoff FR, Ponstingl H. Catalysis of guanine nucleotide exhange on ran by themitotic exchange on ran by the mitotic regulator RCC1. Nature 1991;354:80-3.204. Jones S, Vignais M-L, Borach JR. The CDC25 protein of saccharomyces cerevisiaepromotes exchange of guanine nucleotides bound to ras. Mol Cell Biol1991;11:2641-6.205. Simon MA, Bowtell DLL, Dodson GS, Laverty TR, Rubin GM. Rasl and a putativeguanine nucleotide exchange factor perform crucial steps in signalling by thesevenless protein tyrosine kinase. Cell 1991;67:701-16.206. Farago A, Nishizuka Y. Protein kinase C in transmembrane signalling. FEBS Lett1990;268:350-4.207. Castagna M. Takai Y. Kaibuchi K, Sano K, Kikkawa U, Nishizuka Y. Directactivation of calcium-activated, phospholipid-dependent protein kinase bytumor-promoting phorbol esters. J Biol Chem 1982;257:7847-51.208. Bellacosa A, Testa JR, Staal SP, Tsichlis PN. A retroviral oncogene, aid, encoding aserine threonine kinase containing an SH2-like domain. Science 1991;254:274-7.209. Berridge MJ, Irvine RF. Inositol trisphosphate, a novel second messenger in cellularsignal transduction. Nature 1984;312:315-21.210. Huang K-P. Role of protein kinase C in cellular regulation. BioFactors1990;2:171-8.72211. Mignery GA, Sudhof TC, Takei K, De Camill P. Putative receptor for inositol1,4,5-trisphosphate similar to wanodine receptor. Nature 1989;342:192-5.212. Maeda N, Niinove M, Mikoshiba K. A cerebellar purkinje cell marker p400 protein isan inositol 1.4,5-trisphosphate (Insp3) receptor protein. Purification andcharacterization of InsP3 receptor complex. EMBO J 1990;9:61-7.213. Shinomura T, Asaoka Y, Oka M, Toshida K, Nishizuk Y. Synergistic action ofdiacylglycerol and unsaturated fatty acid for protein kinase C activation: Itspossible implications. Proc Natl Acad Sci U S A 1991;88:5149-53.214. Clark JD, Lin L-L, Kriz RW et al. A novel arachidonic acid-selective cytosolic PLA2contains a Ca2+-dependent translocation domain with homology to PKC and GAP.Cell 1991;65:1043-51.215. Iiikkawa U, Ogita K, Shearman MS et al. The family of protein kinase C: Itsmolecular heterogeneity and differential expression. Cold Spring Harbor SympQuant Biol 1988;53:97-102.216. Sylvia V, Curtin G, Norman J, Stec J, Busbee C. Activation of a low specific activityform of DNA polymerase a by inositol-1,4-bisphosphate. Cell 1988;54:651-8.217. Nishizuka Y. The molecular heterogeneity of protein kinase C and its implication forcellular regulation. Nature 1988;334:661-5.218. Osada S, Mizuno K, Saida TC et al. A phorbol ester receptor/protein kinase,nPKCn, a new member of the protein kinase C family predominantly expressed inlung and skin. J Biol Chem 1990;265:22434-40.219. House C, Kemp BE. Protein kinase C contains a pseudo substrate prototope in itsregulatory domain. Science 1987;238:1726-9.220. Makowske M, Rosen OM. Complete activation of protein kinase C by an antipeptideantibody directed against the pseudosubstrate prototope. J Biol Chem1989;264:16155-9.221. Kishlmoto A, Mikawa K, Hashimoto K et al. Limited proteolysis of protein kinase Csubspecies by calcium dependent neutral protease (calpain). J Biol Chem1989;264:4088-92.222. Bell RM, Loomis CR, Hannun YA. Protein kinase C regulation bysphingosine/lysosphingolipids. Cold Spring Harbor Symp Quant Biol1988;53:103-10.223. Mochly-Rosen D, Khaner H, Lopez J. Identification of intracellular receptor proteinsfor activated protein kinase C. Proc Natl Acad Sci U S A 1991;88:3997-4000.224. Masmoudi A, Labourdette G, Mersel M et al. Protein kinase C located in rat livernuclei. J Biol Chem 1989;264:1172-9.225. Nelsen B, Hellman L, Sen R The NF-kB binding site mediates phorbolester-inducible transcription in nonlymphoid cells. Mol Cell Biol 1988;8:3526-31.226. Moreno S, Nurse P. Substrates for p34cdc2: in vivo veritas. Cell 1990;61:549-51.73227. Houslay MD. 'Crosstalk': a pivotal role for protein kinase C in modulatingrelationships between signal transduction pathways. Eur J Biochem 1991;195:9-27.228. Stumpo DJ, Graff JM, Albert KA, Greengard P, Blackshear PJ. Molecular cloning,characterization, and expression of a cDNA encoding the "80 to 87 kDa"myristoylated alanine-rich C kinase substrate: A major cellular substrate forprotein kinase C. Proc Natl Acad Sci U S A 1989;86:4012-6.229. Mclroy BK, Walters JD, Blackshear PJ, Johnson JD. Phosphorylation-dependentbinding of a synthetic MARCKS peptide to calmodulin. J Biol Chem1991;266:4959-64.230. Thelen M, Rosen A, Nairn AC, Aderem A. Regulation by phosphorylation ofreversible association of a myristoylated protein kinase C substrate with the plasmamembrane. Nature 1991;351:320-2.231. Imagawa M, Chiu R, Karin M. Transcription factor AP-2 mediates induction by twodifferent signal-transduction pathways: Protein kinase C and cAMP. Cell1987;51:251-60.232. Rhee SG, Suh P-G, Ryu S-H, Lee SY. Studies of Inositol phospholipid-specificphospholipase C. Science 1989;244:546-50.233. Thomas GMH, Geny B, Cockcroft S. Identification of a novel cytosolicpoly-phosphoinositide-specific phospholipase C (PLC-86) as the majorG-protein-regulated enzyme. EMBO J 1991;10:2507-12.234. Taylor SJ, Chae HZ, Rhee SG, Exton JH. Activation of the bl isozyme ofphospholipase C by a subunits of the Gq class of G proteins. Nature1991;350:516-8.235. Blank JL, Ross AH, Exton JH. Purification and characterization of two G-proteinsthat activate the b 1 isozyme of phosphoinositide-specific phospholipase C. J BiolChem 1991;266:18206-16.236. Smith CD, Cox C, Snyderman R Receptor-coupled activation ofphosphoinositide-specific phospholipase C by an N protein. Science1986;232:97-100.237. Liang M, Garrison JC. The epidermal growth factor receptor is coupled to apertussis toxin-sensitive guanine nucleotide regulatory protein in rat hepatocyes. JBiol Chem 1991;266:13342-9.238. Rhee SG. Inositol phospholipid-specific phospholipase C: interaction of the gisoform with tyrosine kinase. Trends Biochem Sci 1991;16:297-301.239. Nishibe S, Wahl MI, Wedegaertner PB, Kim JJ, Rhee SG, Carpenter G. Selectivity ofphospholipase C phosphorylation by the epidermal growth factor receptor, theinsulin receptor, and their cytoplasmic domains. Proc Natl Acad Sci U S A1990;87:424-8.240. Downing JR, Margolis BL, Ziberstein A et al. Phospholipase C-g, a substrate ofPDGF receptor kinase, is not phosphorylated on tyrosine during the mitogenicresponse to CSF-1. EMBO J 1989;8:3345-50.74241. Kim HK, Kim JW, Zilberstein A et al. PDGF stimulation of inositol phospholipidhydrolysis requires PLC-gamma 1 phosphorylation on tyrosine residues 783 and1254. Cell 1991;65:435-41.242. Vetter ML, Martin-Zanca D, Parada LF, Bishop JM, Kaplan DR Nerve growth factorrapidly stimulates tyrosine phosphorylation of phospholipase C-gl by a kinaseactivity associated with the product of the trk protooncogene. Biochemistry1991;88:5650-4.243. Wahl MI, Olashaw NE, Nishibe S, Rhee SG, Pledger WJ, Carpenter G.Platelet-derived growth factor induces rapid and sustained tyrosine phosphorylationof phospholipase C-gamma in quiescent BALB/c #T# cells. Mol Cell Biol1989;9:2934-43.244. Toderud G, Wahl MI, Rhee SG, Carpenter G. Stimulation of phospholipaseC-gammal membrane association by epidermal growth factor. Science1990;249:296-8.245. Margolis B, Bellot F, Honegger AM, Ullrich A, Schlessinger J, Zilberstein A. Tyrosinekinase activity is essential for the association of phospholipase C-gammma with theepidermal growth factor receptor. Mol Cell Biol 1990;10:435-41.246. Kumjian DA, Wahl MI, Rhee SG, Daniel TO. Platelet-derived growth factor (PDGF)binding promotes physical association of PDGF receptor with phospholipase C. ProcNatl Acad Sci U S A 1989;86:8232-6.247. Besterman JM, Duronio V, Cuatrecasas P. Rapid formation of diacylglycerol fromphosphatidylcholine: a pathway for generation of a second messenger. Proc NatlAcad Sci U S A 1986;83:6785-9.248. Larrodera P, cornet ME, Diaz-Meco MT et al. Phospholipase C mediated hydrolysisof phosphatidylcholine is an important step in PDGF stimulated DNA synthesis. Cell1990;61:1113-20.249. Diaz-Laviada I, Larrodera P, Diaz-Meco MT et al. Evidence for a role ofphosphatidylcholine-hydrolysing phospholipase C in the regulation of protein kinaseC by ras and src oncogenes. EMBO J 1990;9:3907-12.250. William F, Wagner F, Karin M, Kraft AS. Multiple doses of diacylglycerol andcalcium ionophore are necessary to activate AP-1 enhancer activity and inducemarkers of macrophage differentiation. J Biol Chem 1990;265:18166-71.251. Berry N, Ase K, Kishimoto A, Nishizuka Y. Activation of resting human T cellsrequires prolonged stimulation of protein kinase C. Proc Natl Acad Sci U S A1990;87:2294-8.252. Mignery GA, Newton CL, Archer BT, Sudhof TC. Structure and expression of the ratinositol 1,4,5-trisphosphate receptor. J Biol Chem 1990;21:12679-85.253.^Sudhof TC, Newton CL, Archer BT, Ushkaryov YA, Mignery GA. Structure of a novelInsP3 receptor. EMBO J 1991;10:3199-206.75254. Ferris CD, Huganir RL, Bredt DS, Cameron AM, Snyder SH. Inositol trisphosphatereceptor: Phosphorylation by protein kinase C and calcium calmodulin-dependentprotein kinase in reconstituted lipid vesicles. Proc Natl Acad Sci U S A1991;88:2232-5.255. Berridge MJ, Irvine RF. Inositol phosphates and cell signalling. Nature1989;341:197-205.256. Kaplan DR, Pallas DC, Morgan W, Schaffhausen B, Roberts TM. Mechanisms oftransformation by polyoma virus middle T antigen. Biochim Biophys Acta1988;948:345-64.257. Chan T-0, Tanaka A, Bjorge JD, Fujita DJ. Association of type Iphosphatidylinositol kinase activity with mutationally activated forms of humanpp60 c-src. Mol Cell Biol 1990;10:3280-3.258. Varticovski L, Daley GC), Jackson P, Baltimore D, Cantley LC. Activation ofphosphatidylinositol 3-kinase in cells expressing abl oncogene variants. Mol CellBiol 1991;11:1107-13.259. Ruderman NB, Kapeller R, White MF, Cantley LC. Activation of phsohatidylinositol3-kinase by insulin. Proc Natl Acad Sci U S A 1990;87:1411-5.260. Varticovski L, Druker B, Morrison D, Cantley L, Roberts T. The colony stimulatingfactor-1 receptor associates with and activates phsophatidylinosito1-3 kinase.Nature 1989;342:699-701.261. Kaplan DR, Whitman M, Schaffhausen B et al. Common elements in growth factorstimulation and oncogenic transformation: 85 kD phosphoprotein andphsophatidylinositol kinase activity. Cell 1987;50:1021-9.262. Bjorge JD, Chan T-0, Antczak M, Kung H-J, Fujita DJ. Activated type Iphosphatidylinositol kinase is associated with the epidermal growth factor (EGF)receptor following EGF stimulation. Proc Natl Acad Sci U S A 1990;87:3816-20.263. Carpenter CL, Duckworth BC, Auger KR, Cohen B, Schaffhausen BS, Cantley LC.Purification and characterization of phosphoinositide 3-kinase from rat liver. J BiolChem 1990;265:19704-8.264. Escobedo JA, Navankasattusas S, Kavanaugh WM, Milfay D, Fired VA, Williams LT.cDNA cloning of a novel 85 kD protein that has SH2 domains and regulates bindingof PI3-kinase to the PDGF beta-receptor. Cell 1991;65:75-82.265. Skolnik EY, Margolis B, Mohammadi M et al. Cloning of PI3 kinase-associated p85utilizing a novel method for expression/cloning of target proteins for receptortyrosine kinases. Cell 1991;65:83-90.266. Escobedo JA, Kaplan DR, Kavanaugh WM, Turch CW, Williams LT. Aphosphatidylinositol-3 kinase binds to platelet derived growth factor receptorsthrough a specific receptor sequence containing phosphotyrosine. Mol Cell Biol1991;11:1125-32.267. Sjolander A, Yamamtoto K, Huber BE, Lapetina EG. Association of p21 ras withphosphatidylinositol 3-kinase. Proc Natl Acad Sci U S A 1991;88:7908-12.76268. Bishop JM. Viral Oncogenes. Cell 1985;42:23-38.269. Van Etten RA, Jackson P, Baltimore D. The type IV c-abl gene product is a nuclearprotein, and activation of transforming ability is associated with cytoplasmiclocalization. Cell 1989;58:669-78.270. Hao g-L, Ferris DK, White G, Heisterkamp N, Groffen J. Nuclear and cytoplasmiclocation of the FER tyrosine kinase. Mol Cell Biol 1991;11:1180-3.271. Resh MD. Membrane interactions of pp60 v-src: a model for myristylated tyrosineprotein kinase oncogene. Oncogene 1990;5:1437-44.272. Kaplan JM, Varmus HE, Bishop JM. The src protein contains multiple domains forspecific attachment to membranes. Mol Cell Biol 1990;10:1000-9.273. Resh MD. Specific and saturable binding of pp60 v-src to plasma membranes:Evidence for a myristyl-src receptor. Cell 1989;58:281-6.274. Resh MD, Ling H-P. Identification of a 32K plasma membrane protein that binds tothe myristylated amino-terminal sequence of p60v-src. Nature 1990;346:84-6.275. Cooper JA. The src family of protein-tyrosine kinases. In: Kemp BE. Peptides andPhosphorylation. Boca Raton; CRC Press Inc., 1990:85-113.276. Cooper JA, King CS. Dephosphorylation or antibody binding to the carboxyterminus stimulates pp60 c-src. Mol Cell Biol 1986;6:4467-77.277. Kmiecik TE, Shalloway D. Activation and suppression of pp60 c-src transformingability by mutation of its primary sites of tyrosine phosphorylation. Cell1987;49:65-73.278. Cartwright CA, Eckhart W, Simon S, Kaplan PL. Cell transfomation by pp60 c-srcmutated in the carboxy-terminal regulatory domain. Cell 1987;49:83-91.279. Akiyama T, Matsuda S, namba Y, Saito T, Toyoshima K, Yamamoto T. Thetransforming potential of the c-erb-2 protein is regulated by its autophosphorylationat the carboxyl-terminal domain. Mol Cell Biol 1991;11:833-42.280. Bolen JB, Thompson PA, Eiseman E, Horak ID. Expression and interactions of thesrc family of tyrosine kinases in T lymphocytes. Biochimica Biophysica Acta1989;1:103-49.281. Cheng SH, Piwnica-Worms H, Harvey RW, Roberts TM, Smith AE. The carboxyterminus of pp60 c-src is a regulatory domain and is involved in complex formationwith the middle-T antigen of polyomavirus. Mol Cell Biol 1988;8:1736-47.282. MacAuley A, Cooper JA. Structural differences between repressed and derepresseedforms of p60c-src. Mol Cell Biol 1989;9:2648-56.283. Mayer BJ, Hamaguchi M, Hanafusa H. A novel viral oncogene with structuralsimilarity to phospholipase C. Nature 1988;332:272-5.77284. Matsuda M, Mayer BJ, Fukui Y, Hanafusa H. Binding of transforming protein, p47gag-crk, to a broad range of phosphotyrosine-containing proteins. Science1990;248:1537-9.285. Hirai H, Varmus HE. Site-directed mutagenesis of the SH2 and SH3 codingdomains for c-src produces varied phenotypes, including oncogenic activation ofp60 c-src. Mol Cell Biol 1990;10:1307-18.286. Koch CA, Moran M, Sadowski I, Pawson T. The common src homology region 2domain of cytoplasmic signaling proteins is a positive effector of v-fps tyrosinekinase function. Mol Cell Biol 1989;9:4131-40.287. Declue JE, Sadowski I, Martin GS, Pawson T. A conserved domain regulatesinteractions of the v-fps protein tyrosine kinase with the host cell. Proc Natl MadSci U S A 1987;84:9064-8.288. Luttrel DK, Luttrell LM, Parsons SJ. Augmented mitogenic responsiveness toepidermal growth factor in murine fibroblasts that overexpress pp60 c-src. Mol CellBiol 1988;8:497-501.289. Samelson LE, Philips AF, Loung ET, Klausner RD. Association of thefin proteintyrosine kinase with the T cell antigen receptor. Proc Natl Acad Sci U S A1990;87:4358-62.290. Yamanashi Y, Kakiuchi T, Mizuguchi J, Yamamoto T, Toyoshima K. Association ofB cell antigen receptor with protein tyrosine kinase lyn. Science 1991;251:192-84.291. Veillette A, Bookman MA, Horak EM, Bolen JB. The CD4 and CD8 T cell surfaceantigens are associated with the internal membrane tyrosine protein kinase p56 1ck .Cell 1988;55:301-8.292. Shaw AS, Chlaupny J, Whitney A et al. Short related sequences in the cytoplasmicdomains of CD4 and CD8 mediate biding to the amino-terminal domain of the p56lck protein kinase. Mol Cell Biol 1990;10:1853-62.293. Bolen J. Signal transduction by the src family of tyrosine protein kinases inhemopoietic cells. Cell Growth and Differentiation. 1991;2:409-14.294. Turner JM, Brodsky MH, Irving BA, Levin SD, Perlmutter RM, Littman DRInteraction of the unique amino-terminal region of the tyrosine kinase p56 ick withthe cytoplasmic domains of CD4 and CD8 is mediated by cysteine motifs. Cell1990;60:755-65.295. Kypta RM, Boldberg Y, Ulug ET, Coutneidge SA. Association between the PDGFreceptor and members of the src family of tyrosine kinases. Cell 1990;62:481-92.296. Rapp UR, Cleveland JL, Bonner TI, Storm SM. The raf oncogenes. In: Reddy EP,Skalka AM, Curran T. The Oncogene Handbook. ; Elsevier Science Publishers,1988:213-53.297. Zmuidzinas A, Mamon HJ, Roberts TM, Smith KA. Interleukin-2 triggered raf-1expression, phosphorylation, and associated kinase activity increase through G1and S in CD3-stimulated primary human T cells. Mol Cell Biol 1991;11:2794-803.78298. Rapp UR, Heidecker G, Huleihel M et al. Raf family serine/threonine proteinkinases in mitogen signal transduction. Cold Spring Harbor Quant Biol1988;53:173-84.299. Kanakura Y, Druker B, Wood KW et al. Granulocyte-macrophagecolony-stimulating factor and interleukin.-3 induce rapid phosphorylation andactivation of the proto-oncogene Raf-1 in a human factor-dependent myeloid cellline. Blood 1991;77:243-8.300. Carroll MP, Clark-Lewis I, Rapp UR, May WS. Interleukin-3 andgranulocyte-macrophage colony-stimulating factor mediate rapid phosphorylationand activation of cytosolic c-raf. J Biol Chem 1990;265:19812-7.301. Siegel JN, Klausner RD, Rapp UR, Samelson LE. T cell antigen receptorengagement stimulates C-raf phosphorylation and induces c-raf associated kinaseactivity via a protein ldanse C dependent pathway. J Biol Chem1990;265:18472-80.302. Morrison DK, Kaplan DR, Rapp U, Roberts TM. Signal transduction frommembrane to cytoplasm: growth factors and membrane-bound oncogene productsincrease Raf-1 phosphorylation and associated protein kinase activity. Proc NatlAcad Sci U S A 1988;85:8855-9.303. Lee R-M, Rapp UR, Blackshear PJ. Evidence for one or more raf-1 kinase kinase(s)activated by insulin and polypeptide growth factors. J Biol Chem1991;266:10351=7.304. App H, Hazan R, Zilberstein A, Ullrich A, Schlessinger J. Rapp U. Epidermal growthfactor (EGF) stimulates association and kinase activity of raf-1 with the EGFreceptor. Mol Cell Biol 1991;11:913-9.305. Ambrosio L, Mahowald AP, Perrimon N. Requirement of the drosophila rafhomologue for torso function. Nature 1989;342:288-91.306. Siegfried Z, Ziff EB. Altered transcriptional activity of c-fos promoter plasmids inv-raf transformed NIH 3T3 cells. Mol Cell Biol 1990;10:6073-8.307. Boulton TG, Yancopoulous GD, Gregory JS et al. An insulin-stimulated proteinkinase similar to yeast kinases involved in cell cycle control. Science 1990;249:64-7.308. Boulton TG, Nye SH, Robbins DJ et al. ERKs: A family of protein-serine/threoninekinases that are activated and tyrosine phosphorylated in response to insulin andNGF. Cell 1991;65:663-75.309. Okuda K. Kanakura Y, Hallek M, Druker B, Griffin JD. GM-CSF, IL-3 and steelfactor induce rapid tyrosine phosphorylation of p42 MAP kinase (abstract). Blood1991;78:370a.310.^Ahn N, Krebs E. Evidence for epidermal growth facor-stimulated protein kinasecascade in 3T3 cells. Activation of serine peptide kinase activity by myelin basicprotein kinases in vitro. J Biol Chem 1990;265:11495-591.79311. Hoshi M, Nishida E, Sakai H. Activation of a Ca2+ inhibitable protein kinase thatphosphorylates microtubule-associated protein 2 in vitro by growth factors, phorbolesters, and serum in quiescent cultured human fibroblasts. J Biol Chem1988;263:5396-401.312. Ray LB, Sturgill TW. Rapid stimulation by insulin of a serine/threonine kinase in3T3-L1 adipocytes that phosphorylates a microtubule-associated protein 2 in vitro.Proc Natl Acad Sci U S A 1987;84:1502-6.313. Nel A, Hanekom c, Rheeder A et al. Stimulation of MAP-2 kinase activity in Tlymphocytes by anti-CD3 or anti-Ti monoclonal antibody is partially dependent onprotein kinase C. J Immunol 1990;144:2683-9.314. Gotoh Y, Nishida E, Matsuda S et al. In vitro effects on microtubule dynamics ofpurified Xenopus M phase-activated MAP kinase. Nature 1991;349:251-4.315. Alcorta DA, Crews CM, Sweet LI, Bankston L, Jones SW, Erikson RL. Sequenceand expression of chicken and mouse rsk: homologs of Xenopus laevis ribosomalS6 kinase. Mol Cell Biol 1989;9:3850-9.316. Takishima K, Griswold-Prenner I, Ingebritsen T, Rosner MR Epidermal growthfactor (EGF) receptor T669 peptide kinase from 3T3-L1 cells is an EGF-stimulated"MAP" kinase. Proc Natl Acad Sci U S A 1991;88:2520-4.317. Pulverer BJ, Kyriakis JM, Avruch J, Nikolakaki E, Woodgett JR Phosphorylation ofc-jun mediated by MAP kinases. Nature 1991;353:670-5.318. Anderson NG, Mailer JL, Tonics NK, Sturgill TW. Requirement for integration ofsignals from two distinct phosphorylation pathways for activation of MAP kinase.Nature 1990;343:651-3.319. Ahn NG, Seger R, Bratlien RL, Diltz DC, Tonks NK. Multiple components in anEGF-stimulated protein kinase cascade. In vitro activation of a MBP/MAP2 kinase.J Biol Chem 1991;266:4220-7.320. Gomez N, Cohen P. Dissection of the protein kinase cascade by which nerve growthfactor activates MAP kinases. Nature 1991;353:170-3.321. Seger R, Aim NG, Boulton TG et al. Microtubule-associated protein 2 kinases,ERK1 and ERK2, undergo autophosphorylation on both tyrosine and threonineresidues: Implication for their mechanism of activation. Proc Natl Acad Sci U S A1991;88:6142-6.322. Fischer EH, Charbonneau H, Tonks NK. Protein tyrosine phsophatases: a diversefamily of intracellular and transmembrane enzymes. Science 1991;253:401-6.323. Cohen PTW, Brewis ND, Hughes V, Mann DJ. Protein serine/threoninephosphatases; an expanding family. FEBS Lett. 1990;268:355-9.324. Hunter T. Protein-tyrosine phosphatasees: the other side of the coin. Cell1989;58:1013-6.80325. Mire-Sluis AR, Thorpe R Inerleukin-4 proliferative signal transduction involves theactivation of a tyrosine specific phosphatase and the dephosphorylation of a 80 kDaprotein. J Biol Chem 1991;266:18113-8.326. Shen S-H, Bastien L, Posner BI, Chretien P. A protein-tyrosine phosphatase withsequence similarity to the SH2 domain of the protein-tyrosine kinases. Nature1991;352:736-9.327. Ostergaard HL, Shackelford DA, Hurley TR et al. Expression of CD45 altersphosphorylation of the lck-encoded tyrosine protein kinase in murine lymphomaT-cell lines. Proc Natl Acad Sci U S A 1989;86:8959-63.328. Mustelin T, Cogeshall KM, Altman A. Rapid activation of the T cell tyrosine proteinkinase pp56 lck by the CD45 phosphotyrosine phosphatase. Proc Natl Acad Sci U SA 1989;86:6302-6.329. Weaver CT, Pingerl JT, Nelson JO, Thomas ML. CD8+ T cell clones deficient in theexpression of the CD45 protein tyrosine phosphatase have impaired responses toT-cell receptor stimuli. Mol Cell Biol 1991;11:4415-22.330. Ostergaard HL, Trowbridge IS. Negative regulation of CD45 protein tyrosinephosphatase activity by ionomycin in T cells. Science 1991;253:1423-6.331. Justement LB, Campbell KS, Chien NC, Cambier JC. Regulation of B cell antigenreceptor signal transduction and phosphorylation by CD45. Science1991;252:1839-42.332. Samstag Y, Bader A, Meuer SC. A serine phosphatase is involved in CD2 mediatedactivation of human T lymphocytes and natural killer cells. J Immunol1991;147:788-94.333. Alexander DR, Cantrell DA. Kinases and phosphatases in T-cell activation.Immunol Today 1989;10:200-5.334. McKeon F. When worlds collide: Immunosuppressants meet protein phosphatases.Cell 1991;66:823-6.335. Rodriguez-Tarduchy G, Collins M, Lopez-Rivas A. Regulation of apoptosis ininterleukin-3 dependent hemopoieitic cells by inteleukin-3 and calcium ionophores.EMBO J 1990;9:2997-3002.336. Cromptom T. IL3 dependent cells die by apoptosis on removal of their growthfactors. UNLISTED 1991;4:109-16.337. Whetton AD, Bazill GW, Dexter TM. Haemopoeitic growth factor mediates cellsurvival via its action on glucose transport. EMBO J 1984;3:409-13.338. Whetton AD, Bazill GW, Dexter TM. Stimulation of hexose uptake by haemopoieticcell growth factor occurs in WEHI-3B myelomonocytic leukaemia cells: A possiblemechanism for loss of growth control. J Cell Physiol 1985;123:73-8.339.^Whetton AD, Dexter TM. Effect of haematopoietic cell growth factor on intracellularATP levels. Nature 1983;303:629-31.81340. Cleveland JL, Dean M, Rosenberg N, Wang JYJ, Rapp UR Tyrosine kinaseoncogenes abrogate interleukin-3 dependence of murine myeloid cells throughsignaling pathways involving c-myc: Conditional regulation of c-myc transcriptionby temperature-sensitive v-abl. Mol Cell Biol 1989;9:5685-95.341. Isfort RJ, Ihle JN. Multiple hemotopoietic growth factors signal through tyrosinephosphorylation. Growth Factors. 1990;2:213-20.342. Pelech SL, Paddon HB, Charest DL, Federsppiel BS. IL-3 induced activation ofprotein kinases in the mast cell/megakaryocyte R6-XE.4 line. J Immunol1990;144:1759-66.343. Whetton AD, Valiance SJ, Monk PN, Cragoe EJ, Dexter TM, Heyworth CM.Interleukin-3-stimulated haemopoietic stem cell proliferation: Evidence foractivation of protein kinase C and Na +/H+ exchange without inositol lipidhydrolysis. Biochem J 1988;256:585-92.344. Farrar WL, Thomas TP, Anderson WB. Altered cytosol/membrane enzymeredistribution on interleukin-3 activation of protein kinase C. Nature1985;315:235-7.345. Fields AP, Pincus SM, Kraft AS, May WS. Interleukin-3 and bryostatin 1 mediaterapid nuclear envelope protein phosphorylation in growth factor-dependent FDC-P1hematopoietic cells. J Biol Chem 1989;264:21896-901.346. Chaikin E, Ziltener HJ, Razin E. Protein kinase C plays an inhibitory role ininterleukin-3 and interleukin-4 mediated mast cell proliferation. J Biol Chem1990;265:22109-16.347. Whetton AD, Monk PN, Consalvey SD, Huang SJ, Dexter TM, Downes CP.Interleukin 3 stimulates proliferation via protein kinase C activation withoutincreasing inositol lipid turnover. Proc Nati Acad Sci U S A 1988;85:3284-8.348. Duronio V, Nip L, Pelech SL. Interleukin 3 stimulates phosphatidylcholine turnoverin a mast/meakaryocyte cell line. Biochem Biophys Res Comm. 1989;164:804-8.349. Hogan BB, Spivak JL. Tumor promoting phorbol esters stimulate the proliferationof interleukin-3 dependent cells. J Cell Physiol 1988;137:346-52.350. Whetton AD, Heyworth CM, Dexter TM. Phorbol esters activate protein kinase Cand glucose transport and can replace the requirement for growth factor ininterleukin-3-dependent multipotent stem cells. J Cell Sci 1986;84:93-104.351. Sorensen PHB, Mui ALF, Murthy SC, Krystal G. Interleukin-3, GM-CSF and TPAinduce distinct phosphorylation events in an interleukin 3-dependent multipotentialcell line. Blood 1989;73:406-18.352. Pierce JH, Di Fiore PP, Aaronson SA et al. Neoplastic transformation of mast cellsby abelson-MuLV: aborgation of IL-3 dependence by a nonautocrine mechanism.Cell 1985;41:685-93.353. Evans SW, Rennick D. Farrar WL. Multilineage hematopoietic growth factorinterleukin 3 and direct activators of protein kinase C stimulate phosphorylation ofcommon substrates. Blood 1986;68:906-13.82354. Marla AO, Schreurs J, Miyajima A, Wang JYJ. Hematopoietic growth factorsactivate the tyrosine phosphorylation of distinct sets of proteins ininterleukin-3-dependent murine cell lines. Mot Cell Biol 1988;8:2214-8.355. Evans SW, Rennick D, Farrar WL. Identification of a signal-transduction pathwayshared by haematopoietic growth factors with diverse biological specificity. BiochemJ 1987;244:683-91.356. Isfort K Huhn RD. Frackelton AR Jr., Ihle JN. Stimulation of factor-dependentmyeloid cell lines with Interleukin 3 induces tyrosine phosphorylation of severalcellular substrates. J Biol Chem 1988;263:19203-9.357. Isfort RJ, Stevens D. May WS, Ihle JN. Interleukin 3 binds to a 140-kDaphosphotyrosine-containing cell surface protein. Proc Natl Mad Sci U S A1988;85:7982-6.358. Koyasu S. Tojo A, Miyajima A et al. Interleukin 3-specific tyrosine phosphorylationof a membrane glycoprotein of Mwt 150000 in multi-factor-dependent myeloid celllines. EMBO J 1987;6:3979-84.359. Dunphy WG, Brizuela L, Beach D, Newport J. The Xenopus cdc2 protein is acomponent of MPF, a cytoplasmic regulator of mitosis. Cell 1988;54:423-31.360. Nurse P, Bisset Y. Gene required in G1 for commitment to cell cycle and G2 forcontrol of mitosis in fission yeast. Nature 1981;292:558-60.361. Simanis V. Nurse P. The cell cycle control gene cdc2 of fission yeast encodes aprotein kinase potentially regulated by phosphorylation. Cell 1986;45:261-8.362. Nurse P. Universal control mechanism regulating onset of M-phase. Cell1990;344:503-8.363. Lee MG, Nurse P. Complementation used to clone a human homologue of thefission yeast cell cycle control gene cdc2. Nature 1987;327:31-3.364. Th'ng JPH, Wright PS, Hamaguchi J et al. The FT210 cell line is a mouse G2 phasemutant with a temperature-sensitive CDC2 gene product. Cell 1990;63:313-24.365. Riabowol K, Draetta G, Brizuela L, Vandre D. Beach D. The cdc2 kinase is anuclear protein that is essential for mitosis in mamalian cells. Cell1989;57:393-401.366. Draetta G, Luca F, Westendorf J. Brizuela L, Ruderman J, Beach D. cdc2 proteinkinase is complexed with both cyclin and B: evidence for proteolytic inactivation ofMPF. Cell 1989;56:829-38.367. Muray AW, Solomon MJ, Kirschner MW. The role of cyclin synthesis anddegradation in the control of maturation promoting factor activity. Nature1989;339:280-6.368.^Booher RN, Alfa CE, Hymans JS, Beach DH. The fission yeast cdc2/cdc 13/suc 1protein kinase: Regulation of catalytic activity and nuclear localization. Cell1989;58:485-97.83369. Meijer L, Azzi L, Wang J. Cyclin B targets p34 cdc2 for tyrosine phosphorylation.EMBO J 1991;10:1545-54.370. Strausfeld U, Labbe JC, Fesquet D et al. Dephosphorylation and activation of a p34cdc2/cyclin B complex in vitro by human CDC25 protein. Nature 1991;351:242-5.371. Morla AO, Draetta G, Beach D, Wang JYJ. Reversible tyrosine phosphorylation ofcdc2: dephosphorylation accompanies activation during entry into mitosis. Cell1989;58:193-203.372. Glotzer M, Marray AW, Kirschner MW. Cyclin is degraded by the ubiquitin pathway.Nature 1991;349:132-8.373. Evans T, Rosental ET, Youngblom J, Distel D, Hunt T. Cyclin: a protein specifiedby maternal mRNA in sea urchin eggs that is destroyed at each cleavage division.Cell 1983;33:389-96.374. Wittenberg C, Sugimoto K, Reed SI. G1 specific cyclins of S. cervisiae: cell cycleperiodicity, regulation by mating pheromone, and association with the p34 CDC28protein kinase. Cell 1990;62:225-37.375. Ogas J, Andrews BJ, Herskowitz I. Transcriptional activation of CLN1, CLN2, and aputative new G1 cyclin (HCS26) by SWI4, a positive regulator of G1 specifictranscription. Cell 1991;66:1015-26.376. Cross FR, Tinkelenberg AH. A potential positive feedback loop controlling CLN1 andCLN2 gene expression at the start of the yeast cell cycle. Cell 1991;65:875-83.377. Dirick L, Nasmyth K. Positive feedback in the activation of G1 cyclins in yeast.Nature 1991;351:754-7.378. Pines J, Hunter T. Isolation of a human cyclin cDNA: evidence for cyclin mRNAand protein regulation in the cell cycle and for interaction with p34 cdc2. Cell1989;58:833-46.379. Lew DJ, Dulic V, Reed SI. Isolation of three novel human cyclins by rescue of G1cyclin (cln) function in yeast. Cell 1991;66:1197-206.380. xiong Y, Connolly T, Futcher B, Beach D. Human D-type cyclin. Cell1991;65:691-9.381. Koff A, Cross F, Fisher A et al. Human Cyclin E, a new cyclin that interacts withtwo members of the CDC2 gene family. Cell 1991;66:1217-28.382. Pines J, Hunter T. Human cyclin A is adenovirus E1A-associated protein p60 andbehaves differently from cyclin B. Nature 1990;346:760-3.383. Fang F, Newport JW. Evidence that the Gl-S and G2-M transitions are controlledby different cdc2 proteins in higher eucaryotes. Cell 1991;66:731-42.384. Tsai L-H, Harlow E, Meyerson M. Isolation of the human cdk2 gene that encodesthe cyclin A and adenovirus ElA associated p33 kinase. Nature 1991;353:174-7.84385. Elledge SJ, Spottswood MR. A new human p34 protein kinase, CDK2, identified bycomplementation of a cdc28 mutation in Saccharomyces cerevisiae, is a homolog ofXenopus Egl. EMBO J 1991;10:2653-9.386. Matsushime H, Roussel MF, Ashmun RA, Sherr CJ. Colony-stimulating factor 1regulates novel cyclins during the 01 phase of the cell cycle. Cell 1991;65:701-13.387. Lewin. B. Driving the Cell Cycle: M Phase Kinase, Its Partners, and Substrates. Cell1990;61:743-52.388. Luscher B. Brizuela L, Beach D, Eisenman RN. A role for the p34cdc2 kinase andphosphatases in the regulation of phosphorylation and disassembly of laminin B2during the cell cycle. EMBO J 1991;10:865-75.389. Morgan DO, Kaplan JM, Bishop JM, Varmus HE. Mitosis-specific phosphorylationof p60 c-src by p34cdc2-associated protein kinase. Cell 1989;57:775-86.390. Shenoy S. Choi J-K, Bagrodia S, Copeland TD, Mailer JL, Shalloway D. Purifiedmaturation promoting factor phosphorylates pp60 c-src at the sites phosphorylatedduring fibroblast mitosis. Cell 1989;57:763-74.391. Kipreos El', Wang JYJ. Differential phosphorylation of c-abl in cell cycle determinedby cdc2 kinse and phosphatase activity. Science 1990;248:217-20.392. Roberts SN, Segil N, Heintz N. Differential phosphorylation of the transcriptionfactor Octi during the cell cycle. Science 1991;253:1022-6.393. Marshall CJ. Tumor suppressor genes. Cell 1991;64:313-26.394. Ludlow JW, Shon J. Pipas JM, Livingston DM, DeCaprio JA. The retinoblastomasusceptibility gene product undergoes cell cycle-dependent dephosphorylation andbinding to and release from SV40 large T. Cell 1990;60:387-96.395. Lin BTY, Gruenwald S, Morla AO, Lee W-H, Wang JYJ. Retinoblastoma cancersuppressor gene product is a substrate of the cell cycle regulator cdc2 kinase.EMBO J 1991;58:857-64.396. Goodrich DW, Wang NP, Qian Y-W, Lee EYH, Lee W-H. The retinoblastoma geneproduct regulates progression through the G1 phase of the cell cycle. Cell1991;67:293-302.397. Chellappan SP, Hiebert S, Mudryj M. Horowitz JM, Nevins JR The E2Ftranscription factor is a cellular target for the RB protein. Cell 1991;65:1053-61.398. Mudryj M. Devoto SH, Hiebert SW, Hunter T, Pines J. Nevins JR Cell cycleregulation of the E2F transcription factor involves an interaction with cyclin A. Cell1991;65:1243-53.399.^Bandara L, Adamczewski JP, Hunt T, La Thangue NB. Cyclin A and theretinoblastoma gene product complex with a common transcription factor. Nature1991;352:249-54.85400. Bandara LR, La Thangue NB. Adenovirus E la prevents the retinoblastoma geneproduct from complexing with a cellular transcription factor. Nature1991;351:494-7.401. Giordano A, Whyte P, Harlow E, Franza BR, Beach D, Draetta G. A 60 10cdc2-associated polypeptide complexes with the E1A proteins in adenovirus-infectedcells. Cell 1989;58:981-90.402. Pietenpol JA, Holt Jr. Stein RW, Moses HL. Transforming growth factor B1suppression of c-myc gene transcription: Role in inhibition of keratinocyteproliferation. Proc Natl Acad Sci U S A 1990;87:3758-2.403. Howe PH, Draetta G, Leof EB. Transforming growth factor B1 inhibition of p34cdc2phosphorylation and histone H1 kinase activity is associated with Gl/S phasegrowth arrest. Mol Cell Biol 1991;11:1185-94.404. Pietenpol JA, Stein RW, Moran E et al. TGF B1 inhibiton of c-myc transcription andgrowth in keratinocytes is abrogated by viral transforming proteins with pRBbinding domains. Cell 1990;61:777-85.405. Hall FL, Braun RIC, Mihara K et al. Characterization of the cytoplasmicproline-directed protein kinase in proliferative cells and tissues as a heterodimercomprised of p34 cdc2 and p58 cyclin A. J Biol Chem 1991;266:17430-40.406. Hall FL, Mitchell JP, Vulliet PR. Phosphorylation of synapsin I at a novel site byproline-directed protein kinase. J Biol Chem 1991;265:6944-8.407. Samiei M, Daya-Makin M, Clark-Lewis I, Pelech SL. Platelet-activating factor andthrombin-induced stimulation of p34 cdc2 cyclin histone H1 kinase activity inplatelets. J Biol Chem 1991;266:14889-92.408. Elion EA, Grisafi PL, Fink GR FUS3 encodes a cdc2+/CDC28 related kinaserequired for the transistion from mitosis into conjugation. Cell 1990;60:649-64.409. Chang F, Herskowitz I. Identification of a gene necessary for cell cycle arrest by anegative growth factor of yeast: FAR1 is an inhibitor of a 01 cyclin, CLN2. Cell1990;63:999-1011.410.^Maru Y, Witte ON. The BCR gene encodes a novel serine/threonine kinase activitywithin a single exon. Cell 1991; Suppl 67:459-68.86CHAPTER IIMATERIALS AND METHODSA. MATERIALS1. MaterialsStreptavidin-agarose beads and protein grade Nonidet P-40 (NP40) were purchased fromCalbiochem. (32 131-orthophosphate (50 mCi/mL; carrier free) in acid free solution, Na1 125 I1(100 mCi/mL; carrier free) and 3H-Thymidine (2 Ci/mmole) were from ICN Biochemicals.Expre35S35S protein labeling mix (11.5 nCi/mL) was from NEN. Co-bind plates were fromMicromembranes, Inc (Newark, NJ). All other reagents were purchased from the SigmaChemical Company unless otherwise indicated.2. Growth Factors, Antibodies and cDNA'sPurified E. coli derived recombinant mIL-3 1 was generously supplied by Biogen, Geneva.Switzerland. Purified recombinant murine SF was provided by Immunex, Seattle, Washington.Polyclonal rabbit serum raised against GAP, p9OfPs, p56hek and p97vav were kindly providedby Drs. Tony Pawson (Samuel Lunenberg Research Centre, Toronto), Peter Greer (SamuelLunenberg Research Centre), Roger Perlmutter (University of Washington, Seattle) and MarianoBarbacid (Bristol-Meyers Squibb, Princeton, NJ), respectively. Affinity purified rabbitpolyclonal antibodies against cyclin A and PSTAIRE were the kind gifts of Drs. Fred Hall(University of Southern California, Los Angeles) and Steven Pelech (University of BritishColumbia, Vancouver), respectively. Affinity purified rabbit a-phosphotyrosine antibodies wereobtained from Upstate Biotechnology, Inc.(UBI, Lake Placid, NY) and purified mousemonoclonal a-phosphotyrosine antibody 4G 10 was generously provided by Dr. Brian Druker(Dana Farber Institute, Boston, Ma) and both antibodies were used at 2 pg/mL in Westernanalyses. Monoclonal a-phosphotyrosine 1G2 was purified using protein A agarose fromhybridoma supernatants and coupled to CNBr Sepharose (Pharmacia) according to standard87procedures. Antibodies against PLCy and MAP kinase were from UBI. The AIC 2A cDNA waskindly provided by Dr. Atsushi Miyajima of DNAX, California. E. colt transformed with pGEXvectors containing the SH2 domains of GAP, PLCy and PI3-K were generously supplied Dr.Tony Pawson.3. CellsThe mIL-3 dependent cell line, B6SUtA, generously provided by Dr. J. Greenberger,(University of Massachusetts, Worchester, MA) was derived from a Friend virus infected cultureof C57B1.S bone marrow cells. 2 The clone, B6SUtA1, was obtained by plucking an individualB6SUtA colony growing in methylcellulose. These cells were propagated in RPMI 1640supplemented with 10% fetal calf serum (FCS), 100 u/mL penicillin G. 100 p.g/mLstreptomycin and either 5% pokeweed mitogen stimulated spleen cell conditioned medium or 3ng/mL COS cell derived murine granulocyte-macrophage colony stimulating factor (mGM-CSF).The mIL-3 dependent cell line FDC-P1 was also kindly supplied by Dr. Joel Greenberger andpropagated in the same medium as the B6SUtA1 cells. COS-7 cells were grown in DMEMcontaining 5% FCS. Normal mouse bone marrow cells were isolated from the femurs and tibiaof B6xC3H (F1) mice, treated with ammonium chloride to lyse the mature red blood cells andwashed 2x with Hank's Balanced Salt Solution (HBSS).4. Preparation of Polyclonal Rabbit Serum Against mIL-3R PeptidesRabbit anti-serum to the amino terminal 15 amino acids of the AIC 2A product3 (whichcorresponds exactly to the amino terminus of the mIL-3R that we purified 4) and the terminal15 amino acids of AIC 2A were produced by immunizing rabbits with peptide conjugated tokeyhole limpet hemocyanin (KLH) in complete Freund's adjuvant followed by monthly boostswith peptide-KLH conjugate in incomplete Freund's adjuvant. This antiserum was purified onan amino terminal peptide affinity column and used at 1.0 Kg/mL in Western analyses.Specificity for the mIL-3R was established by comparing its reactivity with AIC 2A transfectedand non-transfected COS cells.88B. GENERAL BIOCHEMICAL TECHNIQUES 1. Protein DeterminationProtein concentrations were determined using the Coomassie Brilliant Blue G-250 dye-binding technique, 5 the soluble silver binding assay, 6 and densitometric analysis of silverstained gels.2. Radio-iodinating Proteins for SDS-PAGE AnalysisProtein samples, usually 50 pL, were buffered to pH 7.5 with 0.1 M Tris-Cl. 100 pei ofNa 125I was added and the sample made 100 µg/mL with chloramine T. After incubating at23° for 20 min, the reaction was terminated with 200 gig/mL Na bisulfite and 10 mM Na!.Unincorporated 1251 was removed by centrifugation through a 1 mL Sephadex G-25(Pharmacia) column equilibrated with 1% SDS in PBS.3. One and Two Dimensional Gel ElectrophoresisSamples for one-dimensional sodium dodecyl-polyacrylamide gel electrophoresis (SDS-PAGE) were adjusted to 2% SDS, 5% 2-mercaptoethanol, 10% glycerol and boiled for 2 minprior to electrophoresis on one-dimensional 7.5% or 5-15% gradient SDS-polyacrylamide gels,as described by Laemmli. 7 Two dimensional gel analysis, isoelectrofocusing in the first andSDS-PAGE in the second dimension, was performed with the Bio-Rad Protean 2D cellaccording to the manufacturer's instructions. 14C-labelled molecular weight markers werefrom Amersham and consisted of myosin (200 kD), phosphorylase-b (97.4 kD) , bovine serumalbumin (69 kD), ovalbumin (46 kD), carbonic anhydrase (30 kD) and lysozyrne (14.3 kD).4. Western BlottingProteins were electrophoresed on 7.5% polyacrylamide gels at 100 V for 12 h and thenelectrotransferred to Imrnobilon-P membranes at 100 V for 2 h at 23°C. Blots were blocked for8 h at 23°C with 20 mM Tris-CI, 0.15M NaCl, pH 7.4 (TBS) containing 5% BSA, and 0.02%NaN3, and then incubated for 2 h with the appropriate first antibody in TBS with 0.05% Tween20 (TBST). Blots were then washed 3 X 5 min with TBST, incubated with 200,000 cpm/mL of1251_ goat anti-rabbit IgG or 125I-goat anti-mouse IgG (Jackson Immunoresearch Labs, Inc..89West Grove, PA) in TBST for 1 h at 23°C, washed 3 times with TBST, air dried and subjected toautoradiography.5. Phosphoamino Acid AnalysisThe analysis of phosphoamino acids was based on the method of Hunter and Sefton.8Phosphoproteins of interest were localized by autoradiography of unfixed gels dried ontocellulose acetate sheets. Phosphoproteins were electro-eluted from gel slices into 0.1 MNaHCO3, 0.05% SDS, dialyzed against 0.01% SDS in distilled water and lyophilized. Residueswere dissolved in 300 of 6 M HC1, boiled for 1.5 h at 100°C under N2 in sealed tubes,diluted with 1 mL distilled water and lyophilized a further 2 times. The dried residues werethen dissolved in 50 % EtOH containing 1 mg/mL phosphoserine, phosphothreonine andphosphotyrosine standards and applied to cellulose thin- layer plates. The plates were wettedevenly with a solution consisting of 1:10:189 pyridine/acetic acid/water and electrophoresed at100 V for 40 min at 23°C. The 32P-labelled phosphoamino acids and the standards weredetected by autoradiography and ninhydrin staining, respectively.C. PURIFICATION AND DERIVITIZATION OF GROWTH FACTORS1. B6SUtA 1  Cell Proliferation Assay for mIL-3 and mGM-CSFTo assay mIL-3, mGM-CSF or SF, B6SUtA 1 cells were washed twice with RPMI 1640 andthen incubated at 2 X 105 cells/mL with 10% FCS in RPMI in the presence of the test samplein a total volume of 0.1 mL in Linbro U-shaped microtitre plate wells. After 18-24 h at 37°C ina humidified atmosphere of 5% CO2 and 95% air, 20 pL of 3H-thymidine stock containing 50pCi/mL (2 Ci/mmole) in RPMI 1640 was added to each well to give a final 3H-thymidineconcentration of 1 pCi/well. After a further 2 h incubation at 37°C, the well contents wereharvested onto glass fiber filters using an LKB 1295 cell harvester and the filters counted on anLKB Betaplate scintillation counter.2. Production and Purification of COS cell Derived mIL-3 and mGM-CSFThe coding region of mIL-3 was assembled from cDNA and genomic mIL-3 clonesobtained from Biogen and expressed in COS cells using a pAX expression vector. 9 An mGM-90CSF vector was similarly constructed using a cDNA clone obtained from Dr. N. Gough. 1°Murine IL-3 and mGM-CSF were purified from serum free supernatants by sequential phenylSepharose and Sephadex G-75 gel filtration chromatography. Preparations were >95% pure asassessed by SDS-PAGE and autoradiography of 125I-labeled material.3. Biotinylation of COS Cell Derived mIL-3In preliminary experiments biotinylated mIL-3 (B-mIL-3) was prepared as describedpreviously 1 1 with NHS-LC-Biotin (Pierce). In later experiments, to avoid the batch to batchvariability in biological activity of NHS-LC-biotinylated mIL-3, B-mIL-3 was prepared from COScell derived mIL-3 using biotin-X-hydrazide (Calbiochem). Purified mIL-3 was made pH 5.5 with0.5 M sodium acetate and oxidized with 10 mM sodium periodate. After 1 h at 23°C, theperiodate was removed by gel filtration through a Sephadex G-25 (Pharmacia) columnequilibrated with 0.1 M sodium acetate, 0.05% Tween 20, pH 5.5. The oxidized mIL-3 was thenmade 2 mM in biotin-X-hydrazide and allowed to react for 5 h at 23°C. B-mIL-3 was separatedfrom free biotin by gel filtration through a Sephadex G-25 Sephadex column equilibrated withPBS containing 0.05% Tween 20.4.^Iodination of mIL-3 Aliquots of recombinant, E. colt derived mIL-3 containing 4 pg of growth factor wereresuspended in a total volume of 100 4, of 0.1 M Na phosphate, pH 7.2, containing 10%dimethyl sulfoxide and 1001.fg/mL polyethylene glycol 4000. To this was added 1 mCi ofNa 1251, followed by 20 µL of freshly prepared 1 mg/mL chloramine T. After incubating for 20min at 23°C, 20 4, each of freshly prepared 3 mg/mL Na metabisulfite and 0.1 M NaI wereadded. Free 1251 was then removed by gel filtration on a 10 mL G-25 Sephadex columnequilibrated with 0.05% Tween 20 in PBS. The void volume was collected, made 0.1% in BSAand 0.02% and NaN3 and stored at 4°C. Following iodination, mIL-3 retained >95% of itsbiological activity as assessed by the B6SUtA 1 3H-thymidine incorporation assay. Specificactivity was determined by self-displacement analysis according to the method of Calvo.1291D. IN VIVO ISOTOPIC LABELLING OF B6SUtA1  CELLS1. Labelling Cells with 32P-Orthophosphate To prepare 32P-labeled B6SUtA1 cells, the cells were washed twice with phosphate freeRPMI 1640 and incubated at 2 x 106 c/mL in phosphate-free RPMI 1640 containing 3 ng/mLmGM-CSF and 0.25 mCi/mL of carrier-free [32P]-orthophosphate for 2 h at 37°C. The cellswere then washed twice with PBS, resuspended at 7.5 x 10 7 c/mL in PBS containing 0.1%ovalbumin, 501.1.M sodium orthovanadate (Na3VO4), 0.02% sodium azide (NaN3) for growthfactor stimulation.2. Labelling Cells with 35S-MethionineTo prepare 355-labelled B6SUtA 1 cells, the cells were first grown in RPMI 1640containing 10% FCS and 3 ng/mL mIL-3 to downregulate mIL-3R levels. 13 The cells were thenwashed twice with 37°C methionine-free DMEM and incubated at 1 X 10 6 c/mL in methionine-free DMEM supplemented with 100 µM methionine, 50 p.Ci/rnL Expre 355355 protein labellingmix, 10% FCS (dialyzed against PBS) and 3 ng/mL mGM-CSF (to upregulate mIL-3R levels 13),for 6 h at 37°C. The cells were then washed twice in PBS and resuspended at 7.5 x 107 c/mLin PBS containing 0.1% ovalbumin, 501.1.M Na3VO4, 0.02% NaN3 for growth factor stimulation.E. PURIFICATION OF THE mIL-3R1. mIL-3R SolubilizationB6SUtAi cells or B6SUtA 1 cell membranes were suspended at 2 X 10 7 c/mL or 2mg/mL respectively in 0.5% NP40 in PBS, or other buffers as indicated, containing 0.2 mMphenylmethyl sulfonylfluoride (PMSF), 100 KIU/mL aprotinin and 2 µg/mL leupeptin. Sampleswere then agitated for 1 h at 4°C before being centrifuged at 100 000 X g for 1 h to removeinsoluble material.2. Soluble Receptor AssayDetergent solubilized samples, usually 50 III., were incubated with 30 nM 125I-mIL-3 for2 to 6 h at 4°C and then transferred to 5 mL polystyrene tubes containing 4 mL of 0.1 % NP40in TBS and 15 ill (packed volume) of Concanavalin A (Con-A) Sepharose beads. The beads were92rocked at 4°C for 6 h, washed twice with TBS containing 0.1% NP40 and counted in aBeckman 5500 gamma counter.3. Purification of the mIL3-R From Intact B6SUtA1  CellsB6SUtA 1 cells, propagated in 3 ng/mL mGM-CSF to upregulate mIL-3 receptors, 13 werewashed twice in PBS and resuspended at 7.5 x 10 7 c/mL in PBS containing 0.1% ovalbumin,501.1.M sodium orthovanadate, 0.02% sodium azide and 200 nM B-mIL-3. After incubating for4 h at 4°C, the cells were washed twice with ice-cold PBS and resuspended at 2 X 10 7 c/mLwith 0.5% NP40 in 50 mM Hepes, 100 mM sodium fluoride, 10 mM sodium pyrophosphate, 2mM Na3VO4, 4 mM EDTA, 2 mM PMSF, 100 KIU aprotinin, 24/mL leupeptin (pH 7.4)(phosphorylation solubilization buffer, PSB) and solubilized by shaking for 1 h at 4°C.Insoluble material was removed by centrifuging for 1 h at 100,000 x g. The supernatant wasincubated batchwise with a-phosphotyrosine Sepharose overnight at 4°C. The beads werethen poured into a column and washed with phosphorylation buffer containing 0.1% NP40 toelute the 120 kDa mIL-3R and other non-tyrosine phosphorylated proteins. The tyrosinephosphoproteins were then eluted with PSB containing 40 mM phenylphosphate and 0.1%NP40. Both the a-phosphotyrosine bound fraction, containing the 140 kDa mIL-3R and theunbound fraction containing the 120 kD mIL-3R were applied separately to streptavidin-agarose (SA) columns and washed sequentially with PSB containing 0.1% NP40, with 20 mMsodium citrate, 0.15 M NaC1 (pH 6.0) containing 0.05% NP40, and then with 0.05% NP40. 20mM sodium citrate, 0.15 M NaC1 (pH 3.5). The low pH fractions were neutralized with 2 MTris-C1 (pH 7.5).4. Determination of the Amino Acid Sequence and Composition of the 120 kD mIL-3RThe unbound fraction from the a-phosphotyrosine Sepharose column waschromatographed through a SA column to recover non-tyrosine phosphorylated. B-mIL-3bound mIL-3Rs. These 120 kD mIL-3Rs were then eluted as above, made 5% with SDS and (3-ME, heated at 60°C for 15 min and applied to an SDS gel consisting of a 2% agarose stackinggel and an acrylamide separating gel (7.5% T, 0.4% C) that had been dialyzed for 48 h against9350^reduced glutathione, 375 mM Tris-C1, 0.1% SDS (pH 8.8) in order to reduce aminoterminal blocking. Following electrophoresis for 5 h at 40 mA with 100 p.M thioglycolate in thecathodic buffer, the gel was equilibrated for 15 min in electrotransfer buffer (0.5 mM DTT, 10mM Caps (pH 10), 10% Me0H) before electrotransfer to an Immobilon-P membrane (Millipore)at 90 V for 2 h at 23°C. The blot was stained for 5 min with 0.1% Coomassie R-250, destainedin 10% acetic acid, 50% methanol, rinsed well with distilled water and air-dried. The 120 kDaband was excised and sent to the University of Victoria Protein Microchemistry Centre forsequencing on an Applied Biosystems 470A gas phase sequencer. A portion of the same bandwas also acid hydrolyzed and analyzed using an Applied Biosystems Model 420A Derivatizer-Analyzer System to obtain amino acid composition data.5. Displacement Analysis With the 120 If.D mIL-3RThe purified 120 kDa mIL-3R (prepared in 0.1 M glycine buffer, pH 3.0, instead of citratebuffer, since the latter inhibits N-glycanase) was supplemented with 0.2 M Tris-C1, pH 8.0, and10 mM phenanthroline and incubated with or without 30 U/mL N-glycanase (Genzyme). Thesamples were then incubated with 30 nM 1251-mIL-3 at 4°C. After 2 h, 1 jiM unlabelled mIL-3was added and, at various times, duplicate aliquots were applied to 1 mL A1.5m (Bio-Rad)columns equilibrated with PBS containing 0.1% NP-40. The void volume containing 1251-m1L-3/mIL-3R complexes was collected and counted in a Beckman 5500 gamma counter.6. Scatchard Analysis of the mIL-3R on Intact CellsCells were washed twice in PBS and resuspended at either 2 X 10 5 cells/80 pl., (cell lines)or 3 X 106 cells/80 p.L (normal mouse bone marrow) in PBS containing 1% BSA, 0.1% gelatin,0.02% NaN3. The cells were then added to 0.5 mL microfuge tubes containing 20 pL of 2 folddilutions of 1251-mIL-3, starting at 10 nM, in the presence or absence of a 20 fold excess ofunlabelled mIL-3. After incubating 2 h at 4°C, the sample was layered over 250 pL of 2:3dioctyl:dibutyl phthalate and microfuged for 2 min at 14 000 rpm. The tubes were then frozenon dry ice and the cell pellets clipped into gamma tubes for counting in a Beckman 5500counter. To examine the effect of carbohydrate on receptor binding affinity, B6SUtA1 cells,94growing exponentially in 3 ng/mL mIL-3, were washed twice in RPMI and grown for 6 h with 10gg/rxiL tunicamycin in 3 ng/mL mGM-CSF, 10% FCS in RPMI 1640. These cells were thensubjected to Scatchard analysis as described.F. ANALYSIS OF THE PROPERTIES OF THE TYROSINE PHOSPHORYLATED mIL-3R1. Purification of Tyrosine Phosphorylated and Tyrosine Unphosphorylated  mIL-3R'sB6SUtA1 cells propagated in 3 ng/mL mGM-CSF to upregulate mIL-3Rs, 13 were washedtwice in PBS and resuspended at 7.5 x 10 7 c/mL in PBS containing 0.1% ovalbumin, 50 I.LMNa3VO4, 0.02% NaN3 and 200 nM B-mIL-3. After incubating for 4 h at 4°C, the cells werewashed twice with ice-cold PBS and resuspended at 2 x 10 7 c/mL with 0.5% NP40 in PSB andsolubilized by shaking for 1 h at 4°C. Insoluble material was removed by centrifuging for 1 hat 100,000 x g. The supernatant was incubated batchwise with a-phosphotyrosine Sepharoseovernight at 4°C. The beads were then poured into a column and washed with PSB containing0.1% NP40 to recover the 120 kDa mIL-3R and other non-tyrosine phosphorylated proteins.The tyrosine phosphoproteins were then eluted with PSB containing 40 mM phenylphosphateand 0.1% NP40. Both the a-phosphotyrosine bound fraction, containing the 140 kDa mIL-3Rand the unbound fraction containing the 120 kDa mIL-3R were applied separately tostreptavidin agarose columns and washed sequentially with PSB containing 0.1% NP40, with20 mM sodium citrate, 0.15 M NaC1 (pH 6.0) containing 0.05% NP40, and then with 0.05%NP40, 20 mM sodium citrate, 0.15 M NaCI (pH 3.5). The low pH fractions were neutralizedwith 2 M Tris-C1 (pH 7.5). Typically, 10 - 20% of the total mIL-3R preparation was tyrosinephosphorylated.2. Phosphatase Treatment of the Tyrosine Phosphorylated mIL-3RTyrosine phosphorylated mIL-3R's were purified as described on page 92 except thatunderivatized mIL-3 was used instead of B-mIL-3 and the streptavidin step was omitted. Thea-phosphotyrosine bound, phenyl phosphate eluate was chromatographed through a SephadexG-25 column equilibrated with 20 mM imidazole, pH 7.2, 0.15 M NaCI, 0.1 mM EDTA and0.1% P-mercaptoethanol to remove the phosphate and protease inhibitors. Receptor95preparations were then incubated 16 h at 37°C with or without each of the followingphosphatases: 100 U/mL T cell specific phosphatase (truncated form, kindly provided by Dr.Deborah Cool, University of Washington, Seattle, WA), 1 U/mL each of the catalytic subunits ofthe serine phosphatase 1 and 2A (kindly provided by Dr. Nick Tonics, Cold Spring HarbourLaboratory, NY), or 1 U/mL calf intestinal alkaline phosphatase (Boehringer Mannheim). Theincubation mixtures were supplemented to 2 mM EDTA for the tyrosine phosphatase samplesand to 1 mM MgC12 and 2 mM ZnC12 for the alkaline phosphatase samples. Receptor integritywas then analyzed by SDS-PAGE and Western blotting with affinity purified anti-N terminalmIL-3R antibody as described above.G. ANALYSIS OF mIL-3R INDUCED TYROSINE PHOSPHOPROTEINS1. Preparation of NP40 Detergent Lysates from Factor Stimulated B6SUtA1  CellsB6SUtA 1 cells were washed twice with RPMI 1640, resuspended at 2 X 10 5 cells/mL in0.1% BSA in RPMI 1640 and incubated at 37°C for 3 h. The cells were then pelleted andresuspended at 2 X 107 cells/mL in 0.1% BSA, RPMI 1640 equilibrated to 4°C or 37°C.Growth factors were then added to give a final concentration of 30 ng/mL mIL-3, 5 ng/mLmGM-CSF or 20 ng/mL SF. For the 4°C growth factor stimulations, the cells were pelletedafter the appropriate time by centrifuging for 4 min at 2000 rpm in an Eppendorfmicrocentrifuge 5415. For the 37°C stimulations, 20 volumes of ice cold PBS were added tostop the reaction before pelleting the cells by centrifuging for 5 min at 600 X g in a BeckmanTJ-6 centrifuge. The cells were then resuspended at 2 X 10 7 cells/mL in 0.5% NP40 in PSBand solubilized for 1 h at 4°C. Insoluble material was pelleted by a 20 min centrifugation at 12000 X g and the supernatants taken for study. If the supernatants were to be analyzed directlyby SDS-PAGE, the samples were made 4% in SDS and 5 % in 2-ME prior to boiling for 2 mM at100°C.2. Immunoprecipitation and SH2 Precipitation of Cellular ProteinsDetergent lysates, prepared as described above, were incubated with first antibody, a-phosphotyrosine-Sepharose (100 pg IG2/20111, packed beads/mL of lysate) or SH2 Sepharose96(20111, packed beads/mL lysate, prepared as described below). After 2 h at 4°C, protein Abeads were added to the soluble antibody samples and incubated an additional 2 h. All beadswere then washed three times with 0.1% NP40 in PSB and either boiled in SDS-PAGE samplebuffer (protein A and SH2 beads) or eluted in 90 p.L of 5 mM phenyl phosphate, 0.1% NP40 inPSB (a-phosphotyrosine Sepharose).3. mIL-3R Tyrosine Kinase AssayCo-bind plates were incubated for 2 h at 23°C with 10 pg/mL BSA in PBS, washed fourtimes with PBS and then incubated with 0.1 mg/mL NHS-LC-Biotin (Pierce). After 1 h at 23°C,the plates were blocked with 1% BSA in PBS. The BSA blocked plates were then incubatedwith 10 .tg/mL SA in PBS for 1 hr at 23°C. B-mIL-3R complexes were prepared by incubatingB6SUtA1 plasma membranes (4 mg/mL) with 200 nM B-mIL-3 for 2 h at 4°C. The membraneswere washed once with PBS and solubilized at 4 mg/mL with 0.5% NP40 in PBS containing 0.2mM PMSF, 100 KIU/mL aprotinin and 2 gg/mL leupeptin for 1 h at 4°C. Insoluble materialwas then pelieted by microfuging 20 min at 12 000 X g, and the solubilized B-mIL-3/mIL-3Rcomplexes were incubated with the biotin/BSA treated Co-bind plates. After 2 h at 4°C, theplates were washed 4 times with 0.1% NP40 in PBS.Samples to be assayed, usually 100111,, were added to the wells in duplicate andsupplemented with 10 mM MgC12 , 5 mM ATP, 2 mM NaVO4, 0.2 mM PMSF, 100 KIU/mLaprotinin and 2 gg/mL leupeptin. The plates were incubated overnight at 4°C, washed 3 timeswith 0.1% NP40 in PSB and incubated a further 2 h at 4°C with 1 pg/mL 1251-1G2 a-phosphotyrosine. After 2 h, the plates were washed 3 X with 0.1% NP40 in PSB and elutedwith 1% SDS. The SDS eluates were transferred to gamma tubes for counting as above.4. Sephadex G 150 Fractionation of mIL-3 Induced Tyrosine PhosphoproteinsTyrosine phosphorylated proteins, from B6SUtA 1 cells stimulated for 10 min at 4°C withmIL-3, were purified from 8 X 10 7 cells using a-phosphotyrosine Sepharose as describedabove. The 500 !IL phenyl phosphate eluate, was applied to a 30 cm X 0.3 cm Sephadex G 15097column equilibrated with 0.1% NP40, 2 mM NaVO4, 50 mM Hepes, pH 7.5. The column wasdeveloped at 2 mL/h and 0.5 mL fractions were collected.5. Preparation of SH2 SepharoseE. coli, transformed with pGEX (Pharmacia) alone or pGEX containing inserts coding forthe SH2 domains of GAP, PLCy or PI3-K, were cultured overnight at 37°C in LB brothsupplemented with 100 µg/mL ampicillin. The stationary phase cultures were then diluted 10fold with fresh 100 µg/mL ampicillin in LB broth and grown for a further 2 h at 37°C. Isothio-p-D-thiogalactopyranoside was then added to induce production of the glutathione-S-transferase (gst) fusion proteins. After 4 h, the cells were harvested by centrifugation andwashed once with PBS. The cell pellets were resuspended in 20 volumes of 1% NP40 in PSBcontaining 10 mM DTT, 0.2 mM PMSF, 2 pg/mL leupeptin and 100 KIU/mL aprotinin, andthen sonicated with 3 X 5 sec bursts at 30% power on a Biosonik III sonicator (BronwellScientific, Rochester, NY). Insoluble material was pelleted with a 20 min, 12 000 x gcentrifugation and the supernatants incubated with glutathione-agarose. After 15 min, thebeads were washed extensively with PSB containing 0.1% NP40, 1 mM DTI' followed by PBScontaining 1 mM MT The proteins were then eluted with 15 mM reduced glutathione, 5 mMUM', 50 mM Tris Cl (pH 7.5), desalted by gel filtration on a Sephadex G-25 column equilibratedwith 1 mM DTT, PBS and coupled to CNBr Sepharose at 1 mg protein/mL swollen beads. Inpreliminary experiments, GAP SH2-gst fusion proteins were also coupled directly toglutathione-agarose with 10 mM dimethyl suberimidate (DMS) according to standardprocedures.98H. REFERENCES1. Kindler V, Thorens B, de Kossodo S et al. Stimulation of hematopoiesis in vivo byrecombinant bacterial murine interleukin 3. Proc Natl Acad Sci U S A1986;83:1001-5.2. Greenberger JS, Sakakeeny MA, Humphries RK, Eaves CJ, Eckner RJ.Demonstration of permanent factor-dependent multipotential(erythroid/neutrophil/basophil) hematopoietic progenitor cell lines. Proc Natl AcadSci U S A 1983;80:2931-5.3. Itoh N, Yonehara S, Schreurs J et al. Cloning of an interleukin-3 receptor gene: Amember of a distinct receptor gene family. Science 1990;247:324-7.4. Mui ALF, Kay RJ, Humphries RK, Krystal G. Purification of the murineinterleukin-3 receptor. J Biol Chem (submitted).5. Bradford M. Anal Biochem 1983;72:248-54.6. Krystal G. A silver-binding assay for measuring nanogram amounts of protein insolution. Anal Biochem 1987;167:86-96.7. Laemmli UK. Cleavage of structural proteins during the assembly of the head ofbacteriophage, T4. Nature 1970;227:680-5.8. Hunter T, Sefton BM. Transforming gene product of rous sarcoma virusphosphorylates tyrosine. Proc Natl Acad Sci U S A 1980;77:1311-5.9. Kay R, Humphries RK. New vectors and procedures for isolating cDNAs encodingcell surface proteins by expression cloning in COS cells. Methods Mol Cell Biol1991;2:254-65.10. Gough NM, Metcalf D, Gough J, Grail D, Dunn AR. Structure and expression of themRNA for murine granulocyte-macrophage colony stimulating factor. EMBO J1985;4:645.11. Sorensen P, Mui ALF, Krystal G. Interleukin-3 stimulates the tyrosinephosphorylation of the 140-kilodalton interleukin-3 receptor. J Biol Chem1989;264:19253-8.12. Calvo JC, Radicella JP, Charreau EH. Measurements of specific radioactivites inlabelled hormones by self displacement analysis. Biochem J 1983;212:259-64.^13.^Murthy SC, Sorensen PHB, Mui ALF, Krystal G. Interleukin-3 down-regulates itsown receptor. Blood 1989:73:1180-7.99CHAPTER IIIPURIFICATION OF THE mIL-3 RECEPTORA. INTRODUCTIONMurine IL-3 1 is a potent hemopoietic growth factor that is produced primarily byactivated T lymphocytes and stimulates the proliferation and differentiation of pluripotent stemcells and committed myeloid and early lymphoid progenitors.' Characterization studies of itscell surface receptor using 125I-mIL-3, various crosslinking agents and intact mIL-3 responsivecells have suggested the presence of two mIL-3 binding proteins with apparent molecularmasses of approximately 140 kD and 70 kD, 2 the former apparently being cleaved to the latterupon mIL-3 binding by a receptor associated protease. 2,3 Studies in our laboratory, andothers, have shown that the higher molecular weight form contains two N-linked carbohydratemoieties,2 and becomes tyrosine phosphorylated following mIL-3 binding (3,4). As a steptowards purifying the receptor for more detailed characterization, we have developed an assayto monitor detergent solubilized receptors that exploits the differential ability of unglycosylated125I-mIL-3 and receptor bound 125I-mIL-3 complexes to bind to Con A-Sepharose beads.With this assay, we have optimized solubilization conditions for the receptor and this, in turn,has made possible its purification from intact cells using a simple two step procedure involvingB-mIL-3, SA and a-phosphotyrosine-Sepharose beads. Two forms of the receptor wereobtained; a tyrosine phosphorylated 140 kD form which was purified to apparent homogeneityand a non-tyrosine phosphorylated 120 kD form. Alkaline phosphatase treatment,chymotrypsin digestion and Western analysis of the 2 forms (using antibodies that weresubsequently developed against the N- and C- termini of the AIC 2A cDNA produce')established that, apart from phosphorylation differences, these two proteins were identical.Sequence and amino acid composition analysis of the 120 kD receptor indicated that it wasvery similar, if not identical to the 120 kD product5 of the AIC 2A cDNA recently cloned by Itoh100et al.4 Western analysis using antibodies against the amino- and carboxy-termini of thepredicted AIC 2A product and chymotryptic maps comparing our purified mIL-3R with the AIC2A protein also suggested identity. However, a comparison of the 1251-rnIL-3 binding propertiesof our 12010 and 140 kD purified mIL-3R preparations with the binding properties of the AIC2A cDNA product expressed on 3T3 and COS cells suggested that the purified mIL-3Rpreparations had the same high affinity as the mIL-3R in intact cells whereas the AIC 2Aproduct displayed a substantially lower affinity for 1251-mIL-3.To investigate this discrepancy in affinity, we examined the contribution that mIL-3Rglycosylation makes to mIL-3 binding. Inhibition of normal mIL-3R glycosylation in B6SUtA 1cells with tunicamycin resulted in loss of high affinity binding. Furthermore in vitro removal ofcarbohydrate from purified mIL-3Rs with N-glycanase also produced low affinity binding.These results suggest that proper glycosylation of the mIL-3R may be required for high affinitybinding to mIL-3.B. RESULTS1. Development of an Assay for Solubilized mIL-3RsAs a first step in purifying the mIL-3R, it was necessary to develop an assay to monitorits purification. The final form of this assay was based in part on the results of preliminaryexperiments involving DSS crosslinking of 1251-mIL-3 to mIL-3 dependent cells before andafter solubilization with different detergents. These results demonstrated that receptorssolubilized with 0.5% NP40 or 2% octylglucoside retained their ability to bind 1251-mIL-3 whilereceptors solubilized with various concentrations of CHAPS or deoxycholate did not. Then,based on our previous finding that the mIL-3R is a glycoprotein, 2 we tested various Sepharosebound lectins for their ability to bind DSS crosslinked complexes containing unglycosylated Ecoli derived 1251-mIL-3 and mIL-3Rs. Our results suggested that the lectins Con-A and wheatgerm agglutinin were equally capable of binding to these 1251-mIL-3R complexes. The finalform of this "lectin assay", described in chapter II, was found to be mIL-3R specific and gave alinear dose response with increasing solubilized mIL-3Rs (Figure 5). With this assay we were101able to further optimize solubilization conditions in order to obtain solubilized mIL-3Rs frommIL-3 dependent cells in high yield and with a similar KD , ie., 1-5 nM, to that observed withintact mIL-3 dependent cells (Figure 6).2. Purification of the mIL-3RFrom previous studies, we found that a subline of B6SUtA cells, which we havedesignated B6SUtA 1 , was capable of expressing 100,000 mIL-3Rs/cell when propagated inmGM-CSF.6 This level of expression is approximately 10-20 times higher than that found ontypical mIL-3 dependant cell lines. 1 These cells were therefore used as starting material for thepurification of the receptor. Standard chromatographic techniques involving ion exchange, gelfiltration, hydrophobic interaction, isoelectric focussing, reverse phase high performance liquidchromatography failed to give reasonable yields or significant purification of the mIL-3RMoreover, affinity chromatography techniques in which mIL-3 was covalently linked to variousactivated matrixes, such as Affi-gel 10, Affi-gel 15, Affi-gel Hz, CNBr activated Sepharose andcarbodimide-aminohexyl Sepharose all gave poor yields. However, an initial purification of themIL-3R was made possible by incubating mIL-3R bearing cells or plasma membranes fromthese cells with B-mIL-3, solubilizing the membrane proteins with NP40 or octyiglucoside andthen chromatographing the solubilized material through a SA column. After washing atneutral pH, mIL-3R activity could be eluted at pH 3.5 (Figure 7A). Although greatly enrichedfor the receptor, these preparations still contained several other proteins as evidenced by 2DO'Farrell gels of pH 3.5 eluates from cells prepared in the presence and absence of excessunlabeled mIL-3 (Figure 7B).To further purify the mIL-3R we took advantage of our previous finding that the mIL-3Rbecomes tyrosine phosphorylated upon binding mIL-3. 7 Biotinylated mIL-3 was first allowed tobind to B6SUtA 1 cells in the presence of sodium orthovanadate and sodium azide to inhibittyrosine phosphatases and receptor internalization, respectively. The cells were thensolubilized in NP40 and the phosphorylated B-mIL-3R complexes allowed to bind to a-phosphotyrosine Sepharose. The bound complexes were eluted with phenylphosphate and put102directly onto a SA column. The pH 3.5 eluate from this SA column yielded a 140 kD band onSDS-polyacrylamide gels. This protein represented more than 98% of the total protein present,as assessed by SDS-PAGE of either silver stained, iodinated, or 35S-labelled preparations(Figure 8). In addition, the a-phosphotyrosine-Sepharose unbound fraction, which containedfrom 2 to 10 times the mIL-3 binding activity of the bound fraction, was chromatographedthrough SA as well and this less pure mIL-3R fraction displayed a prominent 120 kD bandupon SDS-PAGE (Figure 8). A flow chart depicting the purification of these two mIL-3 bindingspecies is shown in Figure 9 and a summary of the purification of the 140 kD species is givenin Table 5.3. Characterization of the Purified mIL-3RPreliminary characterization studies revealed that the highly purified 140 kD mIL-3Rpossessed an isoelectric point of approximately 5.4 and migrated identically on SDS-PAGEunder reducing and non-reducing conditions suggesting that disulfide bridges, if present, didnot significantly restrain this receptor from assuming a random coil conformation in thepresence of SDS (data not shown). The less pure 120 kD mIL-3R also migrated identically onSDS-PAGE under reducing and non-reducing conditions and had a slightly more basicisoelectric point (i.e. approximately 5.6). N-glycanase studies suggested that both speciescontained 10 kD of N-linked carbohydrate (shown for the 120 kD mIL-3 in Figure 10)confirming earlier studies using 125I-mIL-3, DSS crosslinkers and intact cells. 2 The 20 kDdifference in apparent molecular mass between the two receptor preparations appeared to bedue solely to differences in the level of phosphorylation, since alkaline phosphatase digestionreduced both bands to the same molecular mass of 115 kD (Figure 11). The relationship ofthese two proteins was studied further by comparing their chymotrypsin digestion patternsbefore and after alkaline phosphatase treatment. Following alkaline phosphatase digestion,complete identity was observed on SDS polyacrylamide gels (Figure 12). In addition, affinitypurified rabbit anti-sera which was raised against the N- and C- termini of the AIC 2A product(see below), not only recognized both receptor preparations on Western blots, but, more103importantly, only detected the 120 kD protein in unstimulated B6SUtA1 cells. On addition ofmIL-3, the antibodies detected a new band at 140 kD and a reduced amount of the 120 kDspecies (Figure 13). This provided strong evidence that the 140 kD mIL-3R was a tyrosine-phosphorylated derivative of the 120 kD mIL-3R Scatchard analysis with the two purifiedmIL-3R preparations indicated that their affinity for mIL-3 was 2-5 nM, similar to that observedin intact B6SUtA1, normal mouse bone marrow and mIL-3 dependent FDC-P1 cells (Figure 14).Further studies comparing the two receptor preparations revealed a marked difference intheir stability. The tyrosine phosphorylated receptor rapidly degraded on storage, even at 4°C(Chapter IV). We therefore used the 120 kD species to obtain sequencing and amino acidcomposition data. Although this receptor preparation contained a few contaminating proteins,2-D gel analysis yielded, as expected, a very similar pattern to that obtained with SA purifiedtotal cell extracts (Figure 7B) and demonstrated that the 120 k13 receptor was the only proteinat this molecular mass. This made possible the sequencing of the amino terminus of thisprotein from a preparative one dimensional SDS polyacrylamide gel and yielded x-glu-val-thr-glu-glu-glu-x-thr-val-pro-leu-lys-thr-leu-glu-x-tyr-asn-asp. This sequence was confirmed by asecond purification and sequence analysis. To address the possibility that the mIL-3R in the12010 band might have been N-blocked and that a contaminant with an identical isoelectricpoint and molecular mass had been sequenced, we determined the amino acid composition ofthe 120 kD band (Table 6). The calculated amount of protein present suggested very littleamino terminal blockage.A search of the NBRF protein sequence data bank using the FASTA program of theUniversity of Wisconsin Genetics Computer Group 8 revealed this to be a previously unreportedsequence. However, during the preparation of this manuscript, Itoh et a14 reported theexpression cloning of a gene encoding a protein reactive with a monoclonal antibody thatpartially inhibited mIL-3 binding to mIL-3R bearing cells. The amino terminal amino acidsequence we report here corresponds exactly with their predicted amino terminal sequence.104The proline and serine rich content of the mIL-3R we have purified also closely resembles thepredicted amino acid composition of Itoh et al's protein encoded by the AIC 2A cDNA (Table 6).4. Comparison of the Purified mIL-3R with the Product of the AIC 2A cDNATo determine whether our purified mIL-3R was identical to the AIC 2A cDNA productwhich has a predicted molecular mass of only 95 kD, 4 we compared the apparent molecularmasses of the receptors isolated from B6SUtA 1 cells with that of 3T3 cells transfected with theAIC 2A cDNA. Side by side comparisons using SDS-PAGE and Western analysis with the N-terminal antibody to the predicted AIC 2A sequence revealed that the AIC 2A product had anidentical apparent molecular mass to the 120 kD mIL-3R (data not shown) Addition of mIL-3 tothe cells resulted in the appearance of the 140 kD mIL-3R species in B6SUtA 1 cells but not inthe transfected 3T3 cells, consistent with Itoh et al's finding that mIL-3 does not elicit anytyrosine phosphorylation events in these transfected cells. 4 We subsequently prepared affinitypurified rabbit antiserum to the predicted C-terminal 15 amino acids of the AIC 2A cDNA andobtained identical results, further confirming the relatedness of our 140 kD and 120 kD mIL-3R species and their identity with the AIC 2A product (data not shown). Chymotryptic peptidemap comparisons of the 120 kD mIL-3R with the AIC 2A protein expressed in COS cells, alsosuggested identity (Figure 15).However, a major difference between the AIC 2A product and our purified mIL-3Rpreparations was in their affinity for mIL-3. Scatchard analyses with the purified 140 kD and120 kD species as well as with intact normal mouse bone marrow and mIL-3 dependent FDC-P1 cells all yielded KD's of 1-5 nM (Figure 14A). On the other hand, the AIC 2A productdisplayed a substantially lower affinity (-200 nM) when expressed on 3T3 cells or COS cells(Figure 14D). A low affinity for the AIC 2A product has also been reported by Itoh et a14 whoshowed that this was due to a more rapid off rate than that for mIL-3Rs on mIL-3 dependantcell lines (t1/2 or 4 min vs 2 h). This could suggest that another subunit is required for highaffinity binding. However, the fact that purified preparations of our mIL-3Rs exhibit high105affinity binding suggests only one chain is sufficient and perhaps the difference in affinitymight be due to post-translational modifications.5. Effect of mIL-3R Glycosylation on mIL-3 BindingTo examine the role of carbohydrate in mIL-3R function, B6SUtA1 cells were treated withtunicamycin for 6 h and the receptor number and affinity determined by Scatchard analysis.As shown in Figure 16A, a 6 h tunicamycin treatment decreased the molecular mass of thebulk of the mIL-3R population by 10 kD. The total amount of receptor protein was alsoreduced, reflecting the inhibitory effect of tunicamycin on protein synthesis. Scatchardanalysis indicated two affinity classes on tunicamycin treated cells, the high affinity classcorresponding to that observed on control untreated B6SUtA1 cells while the low affinity classhad a KD similar to that of COS cells transfected with the AIC 2A cDNA (Figure 16B)The generation of low affinity binding with tunicamycin treatment suggests thatcarbohydrate may be important for mIL-3 binding. Alternatively, tunicamycin could beaffecting the synthesis of a putative second subunit or the loss of carbohydrate may disrupt theassembly of the mIL-3R complex. To rule out these two possibilities, the carbohydrate wasremoved from the purified 120 kD mIL-3R in vitro with N-glycanase. As the removal ofcarbohydrate made use of the lectin assay impossible, 125I-mIL-3 binding had to be assessedusing gel filtration columns to separate receptor bound from free 125I-mIL-3. In our hands,this technique proved too insensitive for Scatchard analysis but cold displacement studiesprovided an estimate of the mIL-3 dissociation rate. As Figure 17 shows, N-glycanasetreatment of the mIL-3R dramatically increased the rate of dissociation.C. DISCUSSIONBased on our previous finding that mIL-3 induces the tyrosine phosphorylation of themIL-3R,7 we have devised a simple two step procedure to purify this cell surface proteinutilizing B-mIL-3, streptavidin-agarose and a-phosphotyrosine-Sepharose beads. With thispurification strategy, both a highly purified 140 kD tyrosine phosphorylated and a less pure,more stable 120 kD non-tyrosine phosphorylated preparation were obtained. The relatedness106of these 2 proteins was investigated using alkaline phosphatase, chymotrypsin and twodifferent antisera and our results suggest that the 140 kD protein is the tyrosinephosphorylated form of the 12010 protein. The physical properties of both the tyrosinephosphorylated receptor and its non-tyrosine phosphorylated precursor are consistent with theproperties of the bonafide mIL-3R, established using crosslinking studies with 125I-mIL-3 andvarious intact mIL-3 dependent cell lines. 2 .9 Scatchard analyses carried out with the partiallypurified 12010 and highly purified 14010 forms of the receptor reveal a single affinity class of1-5 nM, identical to the value obtained with intact B6SUtA1 cells.From amino terminal amino acid sequence and composition analysis, molecular massdata as deduced from SDS-PAGE (under both reducing and non-reducing conditions) andWestern analysis results, we also conclude that the receptor we have purified is very similar, ifnot identical, to the protein encoded by the AIC 2A cDNA recently cloned by Itoh et al. 4However, according to these authors, their AIC 2A cDNA clone yields a low affinity mIL-3binding protein with a KD of 17.9 ± 3.6 nM at 4°C and 5.7 ± 1.0 nM at 37°C when stablytransfected into L cell fibroblasts. The low affinity of the AIC 2A protein for mIL-3 may indicatethat more than one subunit is required for high affinity binding as has been shown for thehuman IL-3, GM-CSF and IL-5 receptors. In the human system, the IL-3, GM-CSF and IL-5receptors consist of unique a subunits but share a common f3 subunit. The shared subunitforms the basis of the cross-competition observed between the three ligands. The mIL-3R, likethe human IL-3R, could be composed of two subunits as well. However, the mIL-3R appears tobe qualitatively different from its human counterpart in several respects: There is no cross-competition between mIL-3, mGM-CSF and mIL-5; the AIC 2B 1 ° product does not confer highaffinity binding for mIL-3 to AIC 2A5 and; the mIL-3R is a 14010 protein, not 60-80 kD as hasbeen reported for the hIL-3R 11 and the a subunits of the hGM-CSF 12 and mIL-5R 13Moreover, our Scatchard plots indicate that the mIL-3R on intact cells, plasma membranes,solubilized membranes or as a purified molecule exhibits the same ligand affinity. Thus themIL-3R may be more like the IL-4R, which is a 140 kD protein closely related in structure to107the mIL-3R and capable of high affinity binding in the absence of a second subunit. 14,15 Tohelp resolve the difference in affinity between our purified mIL-3R and the AIC 2A product, weinvestigated whether glycosylation of the receptor could play a role. Consistent with thishypothesis we found that tunicamycin treatment of B6SUtA, cells resulted in the generation oftwo mIL-3R affinity classes. Furthermore, in vitro N-glycanase treatment of purified mIL-3Rsincreased the dissociation rate of mIL-3. These observations suggest, as has been described forthe insulin receptor, 16 and the basic fibroblast growth factor receptor 17 that carbohydratestructure may be important for high affinity binding. To further study the role of glycosylationit would be of interest to express the extracellular domain of the AIC 2A cDNA in bothfibroblasts and mIL-3 dependent cell lines in order to compare the binding characteristics oftheir soluble products.108Table 5. Purification of the 140 kD mIL-3RTotal Total Binding Specific OverallProtein Activity Activity Yield(mg) (pmol) (pmol/mg) (%)375 57 0.13 10030 ^ **<0.15 8 >9100 14Step1. Solubilized B6SUtA1 cells2. a-phosphotyrosine-Sepharosephenylphosphate eluate3. Streptavidin agarosepH 3.5 eluateThis typical purification began with 3 X 109 mIL-3 deprived B6SUtA 1 cells and the numberin brackets refers to total protein.** The phenylphosphate eluate could not be assayed since the mIL-3R's at this stage. beingalready bound with B-mIL-3, were not available for 125I-mIL-3 binding.Similar results were obtained in two separate purifications.Table 6. Comparison of Amino Acid CompositionsAmino Acid Number of ResiduesPredicted*^Determin.ed**ala 35 47arg 30 41asx 66 62cys 17glx 104 87gly 54 73his 20 18ile 29 31leu 85 101lys 37 37met 12 14phe 24 26pro 102 104ser 101 109thr 40 39trp 18tyr 31 29val 51 62* Calculated from the AIC 2A cDNA sequence reported by Itoh et al. 41094000cn3000 -02000 -0Q. 1000110f`TVIT-IT^10^100^200^300^400^500 600[plasma membrane protein] (ug/mt)Figure 5. Specificity and sensitivity of the lectin assay for the solubilized mIL-3R NP40solubilized B6SUtA 1 (0) or P815 (mIL-3R negative8 (*) plasma membranes wereincubated with 1251-mIL-3 in the presence (a) or absence (a M of a 20 fold excessof unlabelled mIL-3 for 2 h at 4°C. Con-A Sepharose beads (15 packed volume)were then added and the mixture rocked at 4°C for 6 h before the beads werewashed twice with 0.1% NP40 in TBS and counted in a Beckman 5500 gammacounter. Similar results were obtained in three separate experiments.A.B.1110.20.1 -0.000.0140.0120.010 -0.008 -LL0.006 -00300C.0.004 -0.002 -0.000 •^•^•^► •^• I^I0 10 20 30 40 50 60 70 80 90 100[Bound] (pM)0.0700.06 -0.05 - 0•0.04 -0.03 -0.02 -0.01 -0.000^100 ^200^300[Bound] (pM)Scatchard analyses were performed with (A) intact cells as described earlier 2 orwith (B) B6SUtA1 plasma membrane proteins solubilized in 2% octylglucoside inPBS or (C) 0.5% NP40 in PBS. All 3 samples gave KD's between 1 and 5 nM.Similar results were obtained in three separate experiments.Figure 6.100008000.20>. 6000-cr 4000-MEEey)Ca)O20001120^ 10^ 20fraction numberFigure 7A. Streptavidin-agarose elution profile of the mIL-3R. B-mIL-3 was incubated withB6SUtA 1 cells for 2 h at 4°C and the membranes solubilized with 0.5% NP40 andapplied to SA columns. mIL-3R activity was monitored using the lectin assay andprotein concentrations were determined with the Coomassie Dye binding assay 18and the silver-binding assay. 19 Similar results were obtained in ten separateexperiments.2Acidic^Basic^ Acidic^Basic-200^ -200-97.5^ -97.5-69^ -694fillabra.^411,or .IOW^ -46^ -46Figure 7B. Two dimensional O'Farrell gels of the SA purified mIL-3R. B-mIL-3 wasincubated with B6SUtA1 cells in the presence (2) and absence (1) of a 20 foldexcess of unbiotinylated mIL-3 for 2 h at 4°C. The cells were then solubilizedwith 0.5% NP40 in PBS and incubated with SA. After washing the beads, boundproteins were eluted at pH 3.5. An mIL-3 specific protein at approximately 140kD (see arrow) was consistently observed when eluants were radioiodinated andsubjected to 2-D SDS-PAGEli and autoradiography. Similar results wereobtained in three separate experiments.113114200-97.5-69-46-30-Figure 8. SDS-PAGE of the purified mIL-3R from 35S labelled B6SUtA 1 cells. The SAeluates of the a-phosphotyrosine Sepharose bound (lane 2) and unboundfractions (lane 1) were subjected to SDS-PAGE and fluorography. Similar resultswere obtained in five separate experiments.a-phosphotyrosineagarosestreptavidinagarosetyrosine phosphorylated 140 kD formstreptavidinagarosenon-tyrosine phosphorylated 120 kD form.B6SUtA 1 cell115Figure 9.^Flow diagram showing the purification of the mIL-3R Both highly purified 140kD tyrosine phosphorylated and less pure 120 kD tyrosine-unphosphorylatedmIL-3R species were obtained.1^2200-69-46-116Figure 10. N-glycanase digestion of 32P-labelled mIL-3R's. B6SUtA i cells were equilibratedwith 32P-orthophosphate as described in chapter II. 32P-labeled 120 kD mIL-3Rs were purified from the a-phosphotyrosine unbound fraction as described inChapter II, denatured by boiling 2 min in 0.5% SDS, 0.1 MI3-mercaptoethanoland diluted into 0.1 M sodium phosphate buffer (pH 8.6) containing 10 mM 1.10phenanthroline and a 7 fold excess of NP40. The sample was then incubated inthe presence (lane 1) and absence (lane 2) of 5 U/ml of N-Glycanase (Genzyme)for 16 h at 37°C. Both samples were then made 4% in SDS and 5% inf3-mercaptoethanol, separated by SDS-PAGE and subjected to autoradiography.Similar results were obtained in two separate experiments.1172^32 0 0-—115 9 7 . 5 -6 9 -Figure 11. Effect of phosphorylation on apparent molecular mass. The mIL-3R, SA purifiedfrom either the a-phosphotyrosine unbound or bound fraction, wasradioiodinated subjected to SDS-PAGE and electroeluted from the gel. Theelectroeluates from the unbound (lanes 1 and 3) and the bound (lanes 2 and 4)fractions were incubated with (lanes 1 and 2) or without (lanes 3 and 4) 0.07U/m1 alkaline phosphatase (Boehringer Mannheim) in 100 mM Tris-C1, pH 8.0,10 mM MgC12 , 0.1 mM ZnC12, for 2 h at 37°C and analyzed by SDS-PAGE anddetected by autoradiography. Similar results were obtained in two separateexperiments.200-69-46-30-14-1 2 3 4118Figure 12. Chymotryptic maps of the tyrosine phosphorylated 140 kD (lanes 1 and 3) andthe tyrosine unphosphorylated 120 kD (lanes 2 and 4) proteins before (lanes 1and 2) and after (lanes 3 and 4) alkaline phosphatase digestion. Receptorpreparations were radiolabelled and treated, or not, with alkaline phosphatase &subjected to SDS-PAGE as in figure 11. The receptor bands were excised, placedin the wells of a 15% SDS-polyacrylamide gel & digested in situ with 10 µg/m1chymotrypsin according to the method of Cleveland.20 Fragments werevisualized by autoradiography. Similar results were obtained in three separateexperiments.1191^2200-97.5-69-46-30-Figure 13. Western analysis of plasma membranes from B6SUtA 1 cells exposed to controlbuffer (lane 1) or 30 nM mIL-3 for 15 min at 37°C. The Immobilon blot wasprobed with affinity purified rabbit antibody to the amino-terminus of the AIC 2Aproduct as described in Chapter II. Similar experiments were obtained in fiveseparate experiments.0.030.02 -0.01 -0.00 .^•^•0^20 • 40^60^80 100 120 140[Bound] (pM)B.A. 0.030.02 -LL LL0.01 -I1 0^20^30^40^50[Bound] (pM)0.006 0C•^0.030.02D. 0.010 0.009 -0.008 -0.007 -0.006 -0.005 -0.01 -0.00Figure 14.•^► 0.004^.^,^, ^,0^10^20^3 0^40^200 400 600 800 1 000 1200 1 400 1 600 1 800[Bound] (pM)[Bound] (pM)Scatchard analyses of the purified 140 ItD (A) and 120 Id) (B) miL-3R species as well as intact normalmouse bone marrow cells (C,r1) and FDC-P1 cells (C,4 ) and COS cells transfected with the AIC 2A cDNA(D).1 2200-69-46-30-14-Figure 15. Cymotryptic map of the purified 120 kD mIL-3R and the AIC 2A protein. The Aic2A protein was expressed in COS-7 cells using the DEAE-Dextran method andpurified using B-IL-3/SA. Peptide maps were then prepared of the 120 kD mIL-3R (lane 1) and the purified AIC 2A protein (lane 2) as described in figure 12.12197.5-69-u.46-^0.010^0.0090.008 -0.007-0.006  -0.005 -0.004 -0.003  -0.0020.001^0.000 •^' 1'1'1'1.10CI0 0 05CI8A.1^2200- B.30-^ Bound (pM)_Figure 16A. Western analysis of plasma membranes from B6SUtA 1 cells following tunicamycin treatment. Cells were grownfor 6 h in RPMI containing 3 ng/mL GM-CSF, & 10% FCS with (lane 2) or without (lane 1) 10 µg/mltunicamycin. Plasma membranes were then prepared and subjected to Western analyses with affinity purifiedrabbit antibody to the amino-terminus of the AIC 2A product, as described in Chapter II.Figure 16B. Scatchard analysis of B6SUtA 1 cells following tunicamycin treatment. Tunicamycin treated cells were washedand used for Scatchard analysis as described in Chapter II. Similar results were obtained in three separateexperiments.Displacement Analysis123- N-g I yc an ase111 + N-glycanase0^45 min^120 mintimeFigure 17. Displacement kinetics of 1251-m1L-3 from the purified 120 kD mIL-3R Thereceptor preparation was incubated in the presence and absence of N-glycanase.1251-mIL-3 was then added and, after 2 h, excess unlabeled mIL-3 was includedto displace the receptor bound 1251-mIL-3, as described in Chapter II. Datapoints represent means (n=2) +/- SEM. The significance of the difference betweenthe 0 time point and the 45 and 120 min time points, in the N-glycanase treatedsample, has a p value < 0.05.124D. REFERENCES1. Schrader JW. The panspecific hemopoietin of activated T lymphocytes(interleukin-3). Annu Rev Immunol 1986;4:205-30.2. Murthy SC, Mui ALF, Krystal G. Characterization of the interleukin 3 receptor. ExpHematol 1990;18:11-7.3. Isfort RJ, Stevens D, May WS, Ihle JN. Interleukin 3 binds to a 140-kDaphosphotyrosine-containing cell surface protein. Proc Natl Acad Sci U S A1988;85:7982-6.4. Itoh N, Yonehara S, Schreurs J et al. Cloning of an interleukin-3 receptor gene: Amember of a distinct receptor gene family. Science 1990;247:324-7.5. Miyajima A, Kitamura T, Hayashida K et al. Molecular analysis of the IL-3 andGM-CSF receptors [abstract]. J Cell Biochem 1991; Suppl 15F:37.6. Murthy SC, Sorensen PHB, Mui ALF, Krystal G. Interleukin-3 down-regulates itsown receptor. Blood 1989;73:1180-7.7. Sorensen P, Mui ALF, Krystal G. Interleukin-3 stimulates the tyrosinephosphorylation of the 140-kilodalton interleukin-3 receptor. J Biol Chem1989;264:19253-8.8. Devereux J, Haeberli P, Smithies 0. A comprehensive set of sequence analysisprograms for the VAX. Nucleic Acids Res 1984;12:387-95.9. Schreurs J, Miyajima A, Arai K. Murine interleukin-3 (IL-3) receptorstructure [abstract]. FASEB J 1988;2:A1652.10. Gorman DM, Itoh N, Kitamura T et al. Cloning and expression of a gene encodingan interleukin 3 receptor-like protein: Identification of another member of thecytokine receptor gene family. Proc Natl Acad Sci U S A 1990;87:5459-63.11. Miyajima A, Hayashida K, Sato N et al. Molecular structure of the receptors forIL-3, GM-CSF and IL-5 [abstract]. Int J Cell Cloning 1991;9:371.12. Chiba S, Shibuya K, Miyazono K et al. Affinity purification of human granulocytemacrophage colony-stimulating factor receptor a-chain. J Biol Chem1990;265:19777-81.13. Takaki S, Tominaga A, Hitoshi Y et al. Molecular cloning and expression of themurine interleukin-5 receptor. EMBO J 1990;9:4367-74.14. Fernandez-Botran R, Vitetta ES. A soluble, high-affinity, interleukin-4-bindingprotein is present in the biological fluids of mice. Proc Natl Acad Sci U S A1990;87:4202-6.^15.^Keegan AD, Beckmann MP, Park LS, Paul WE. The IL-4 receptor: Biochemicalcharacterization of IL-4-binding molecules in a T cell line expressing large numbersof receptors. J Immunol 1991;146:2272-9.12516. Podskalny JM, Rouiller DG, Grunberger G, Baxter RC, McElduff A, Gorden P.Glycosylation defects alter insulin but not insulin-like growth factor I binding toChinese hamster ovary cells. J Biol Chem 1986;261:14076-81.17. O'Farrell PH. High resolution two-dimensional electrophoresis of proteins. J BiolChem 1975;250:4007-21.18. Bradford M. A rapid and sensitive method for the quantitation of microgramquantities of protein utilizing the principle of protein-dye binding. Anal Biochem1976;72:248-54.19. 'Crystal G. A silver-binding assay for measuring nanogram amounts of protein insolution. Anal Biochem 1987;167:86-96.20.^Cleveland DW, Fischer SG, Kirschner MW, Laemmli UK. Peptide mapping by limitedproteolysis in sodium dodecyl sulfate polyacrylamide gels. J Biol Chem1977;252:1102-6.126CHAPTER WTYROSINE PHOSPHORYLATION OF THE mIL-3 RECEPTORINCREASES ITS SUSCEPTIBILITY TO CLEAVAGEA. INTRODUCTIONAs described in chapter III, we have purified the 140 kD protein to apparent homogeneityon the basis of its ability to bind mIL-3. 1 In addition, we have established the likely identity ofthe 140 kD protein with the mIL-3 binding protein, Aic 2A, recently cloned by Itoh et a1. 2through amino acid sequence and composition analysis, Cleveland maps and its ability to reactwith antisera to the amino and carboxy termini of the Aic 2A protein.The AIC 2A cDNA product belongs to a unique family of growth factor receptors thatincludes the 13 chain of the IL-2 receptor (IL-2R) and the receptors for IL-4, IL-5, IL-6, IL-7, G-CSF, GM-CSF, LIF and erythropoietin.3,4 The members of this hemopoietic growth factorreceptor superfamily are characterized by conserved Lys^eines and Trp-Ser-X-Trp-Ser motifs intheir extracellular domains, and an exceptionally high proline and serine content in theircytoplasmic domains.3 None of these receptors possess intrinsic tyrosine kinase domains andtheir mechanisms of action are, as yet, unknown. However, although these receptors do notappear to possess tyrosine kinase activities, at least three of them, the mIL-3R, 5,6 the EpR7 '8and the f3 chain of the IL-2R, 9 become tyrosine phosphorylated upon ligand binding. We haveshown previously that mIL-3 induces tyrosine phosphorylation of its receptor 5 and have thusbeen able to use immobilized a-phosphotyrosine coupled with biotinylated mIL-3 andstreptavidin agarose to purify the tyrosine phosphorylated form of the mIL-3R to homogeneityas described in Chapter III. During the course of these studies we noted that different mIL-3Rpreparations differed in their susceptibility to proteolysis and that this differentialsusceptibility appeared to correlate with the tyrosine phosphorylation state of the mIL-3R beingisolated. We now show that tyrosine phosphorylation of the mIL-3R is the signal for its own127cleavage. This generation of a receptor fragment may be important to the mechanism of actionof mIL-3.B. RESULTS1. Purification of Tyrosine Phosphorylated and Tyrosine Unphosphorylated mIL-3RsThe 140 kD tyrosine phosphorylated and the 120 kD tyrosine unphosphorylated mIL-3R's were purified from B6SUtA 1 cells that were grown under conditions which allowedexpression of approximately 100,000 mIL-3Rs/ce11. 10 The two step purification procedureinvolved B-mIL-3, streptavidin agarose and a-phosphotyrosine Sepharose, as previouslydescribed in Chapter III. As shown in Figure 18A when 32P-labeled B6SUtA 1 cells weresubjected to this purification procedure only 1 band was observed in both the tyrosinephosphorylated and the non-tyrosine phosphorylated preparations, as assessed by SDS-PAGE.This is consistent with our previous data in which we observed only a 140/120 kD doubletfollowing B-mIL-3 stimulation of 32P-labeled B6SUtA 1 cells and subsequent streptavidinagarose purification. 5 The tyrosine phosphorylated mIL-3R had an apparent molecular mass20 kD higher than the tyrosine unphosphorylated form. Phosphoamino acid analysisconfirmed that the two forms of the receptor differed in phosphotyrosine content (Figure. 18B).Previous studies using alkaline phosphatase, Cleveland maps and antisera to the aminoterminal of the mIL-3R demonstrated that these two proteins were otherwise identical.'2. Effect of Tyrosine Phosphorylation on the Stability of the mIL-3RTo compare the stability of the two receptor preparations, they were stored at 4°C and, atvarious times, aliquots were either assayed for 1251-mIL-3 binding activity or directlyradioiodinated and subjected to SDS-PAGE. The results clearly demonstrated that the non-tyrosine phosphorylated receptor was unaffected by storage at 4°C, since it retained its fullability to bind 1251-mIL-3 (Figure. 19A) and remained intact as a 120 kD protein (Figure. 19B).In contrast, the tyrosine phosphorylated receptor rapidly lost mIL-3 binding activity and thiscorrelated kinetically with the disappearance of the 140 kD mIL-3R moiety. Since addition ofprotease inhibitors could prevent this (Figure. 19B), it seemed likely that this instability was128due to proteolysis. To determine whether this sensitivity to proteolysis was an intrinsicproperty of the tyrosine phosphorylated mIL-3R (due to tyrosine phosphorylation inducedconformational alterations which allow protease binding and cleavage) or a secondary effectdue to the possible presence of proteases or protease inhibitors in one but not the otherreceptor preparation, mixing experiments were performed. Tyrosine phosphorylated andtyrosine unphosphorylated m1L-3Rs were combined and their ability to bind 1251-mIL-3examined (Figure. 19B). The resulting drop in activity corresponded to the level of tyrosinephosphorylated receptor present in the mixture, indicating that if a protease co-purified withthe tyrosine phosphorylated mIL-3R, it was not capable of degrading the tyrosineunphosphorylated receptor. A more rigorous mixing experiment was then performed with 35S-labelled and unlabelled receptor preparations. 35S-labelled tyrosine phosphorylated andtyrosine unphosphorylated mIL-3Rs were purified from B6SUtA 1 cells equilibrated with 35S-methionine and mixed with either tyrosine phosphorylated or tyrosine unphosphorylated,unlabelled mIL-3R preparations. As shown in Figure. 20, the 35S-labelled tyrosinephosphorylated mIL-3R degraded to 70 kD fragments and this degradation was not inhibitedby the presence of unlabelled, non-tyrosine phosphorylated mIL-3R. Conversely, the 35S-labelled tyrosine unphosphorylated mIL-3R did not undergo proteolysis even in the presence ofunlabelled tyrosine phosphorylated mIL-3R These results strongly support the conclusionthat tyrosine phosphorylation of the receptor causes an intrinsic change in the receptor whichgreatly increases its susceptibility to proteolysis. These results further suggest the intriguingpossibility of a specific protease that becomes associated with (or is already associated with)the mIL-3R, since it apparently co-purifies with the tyrosine phosphorylated receptor.3. Receptor Degradation Requires Both Tyrosine and Serine PhosphorylationTo investigate the phosphorylation requirements for receptor cleavage, mIL-3Rs werepurified from mIL-3 stimulated B6SUtA 1 cells with a-phosphotyrosine Sepharose. Receptorsisolated using this procedure contain both phosphotyrosine and phosphoserine residues (seeFigure 18B). This receptor preparation was incubated at 37°C for 16 h in the presence or129absence of a tyrosine specific phosphatase, a mixture of serine phosphatases or alkalinephosphatase, as described in Chapter II. As shown in Figure 21, receptor degradation wasmarkedly inhibited by all three phosphatases, suggesting that both serine and tyrosinephosphorylation contribute to the proteolytic sensitivity of the 140 kD tyrosine phosphorylatedmIL-3R Alternatively, it is conceivable that the protease responsible for the cleavage is itselfactivated by tyrosine and serine phosphorylation and that dephosphorylation inactivates it.However, this is not consistent with our previous mixing studies. The p70 cleavage productwas not observed in the phosphatase experiments shown in Figure 21 because of the lengthyincubation period and the less pure receptor preparation used (i.e. streptavidin agarose wasnot employed). This resulted in the further fragmentation of the initial cleavage product.4. Degradation of the mIL-3R in Intact Cells. To examine whether mIL-3R cleavage also occurs in intact cells, B6SUtA 1 cells wereincubated at 37°C with mIL-3 for various times and plasma membranes were prepared andelectrophoresed in SDS-polyacrylamide gels. Western analysis was then carried out using anaffinity purified antibody raised against a peptide corresponding to the N-terminal 15 aminoacids of the mIL-3R 1 .2 In the absence of mIL-3 (Figure. 22, time 0) only the 120 kDunphosphorylated mIL-3R (lower arrow) was detected. However, in the presence of mIL-3, the140 kD phosphorylated species (upper arrow) appeared. Consistent with earlier studies," the140 kD band was maximal within 10 min of incubation with mIL-3 and then began to declineby 30 min. As can also be seen in Figure. 22, this decline was accompanied by the appearanceof three bands at approximately 70 kD. These three bands might reflect some secondaryprocessing, or heterogeneity in the 140 kD mIL-3R (perhaps due to differential glycosylation) orin the cleavage site. By 60 min, the 140 kD species was barely visible and the 70 kD bandsmore prominent. Identical results were obtained when whole cell lysates were analyzed insteadof plasma membranes, although the blots contained additional bands because of cross-reactivity with cellular proteins (data not shown). Interestingly, treatment of these cells withthe phorbol ester TPA, instead of mIL-3, stimulated, exclusively, the serine and/or threonine130phosphorylation of the mIL-3R (Figure. 23A, lane 6) (i.e. no tyrosine phosphorylation wasdetected by Western analysis using a-phosphotyrosine antibodies (Figure. 23B, lane 6). Thisphosphorylation was suggested by the increase in apparent molecular mass of the receptor(Figure. 23A) No breakdown to 70 kD fragments was observed under these conditions,consistent with our in vitro findings that both tyrosine and serine phosphorylations arerequired for cleavage. Also shown in Figure 23A is the effect of various inhibitors on mIL-3Rstimulated mIL-3R cleavage. Chloroquine did not prevent mIL-3R cleavage (lane 3), furtherestablishing that this cleavage reaction occurs at the cell surface and not in lysosomes.Genistein, an inhibitor of certain tyrosine kinases, also did not prevent mIL-3R cleavage (Figure23B) but it also did not inhibit mIL-3R receptor tyrosine phosphorylation, suggesting that itdoes not affect the tyrosine kinase responsible for phosphorylating the mIL-3R This inhibitordid, however, inhibit the tyrosine phosphorylation of several other, constitutivelyphosphorylated proteins (Figure. 23B, lane 3). Interestingly, staurosporine, originallydescribed as a protein kinase C inhibitor, 12 did prevent the mIL-3 stimulated mIL-3R cleavage(Figure. 23A, lane 5) but, in our hands, also inhibited mIL-3R tyrosine phosphorylation (Figure.23B, lane 4). These observations are consistent with our hypothesis that tyrosinephosphorylation of the mIL-3R is essential for cleavage.C. DISCUSSIONOur previous studies 13 and those by others,6 using 126I-labeled mIL-3, intact cells andvarious crosslinking agents, established that mIL-3 binds to both a 140 kD and a 70 kDspecies. Upon incubation at 37°C, we found that the intensity of the 140 kD crosslinked banddecreased and the 70 kD band increased, suggesting conversion of the p140 to the p70. Basedon these and other data showing that the p70 could be generated with both plasma membraneand solubilized membrane preparations, 13 we proposed that a protease closely associated withthe mIL-3R was responsible for this cleavage. This interpretation was strengthened by carryingout crosslinking studies with our purified 140 kD mIL-3R preparations. 14 Specifically, asreported in the present study, a slow conversion of the 140 kD receptor to a 70 kD fragment131was noted in the absence of cross-linkers. However, this conversion was markedly acceleratedwhen the crosslinker DSS and 125I-m1L-3 were added, suggesting that crosslinking altered theconformation of the mIL-3R so as to make it more susceptible to cleavage. We confirmed that itwas DSS and not mIL-3 binding that accelerated the cleavage by using unlabeled mIL-3 and32P-labeled 140 kD mIL-3Rs incubated in the presence and absence of DSS. Only when DSSwas present did the 32P-140 kD mIL-3R rapidly disappear (data not shown).In the present study, both our in vitro and in vivo data are consistent with a model ofmIL-3 action in which the mIL-3 induced tyrosine and serine phosphorylation of its cell surfacereceptor leads to receptor cleavage. Very recent studies in our laboratory indicate that theseevents also occur readily at 4°C in intact cells, suggesting, perhaps, that the tyrosine andserine kinases and the protease are all closely associated with the mIL-3R (data not presented).This would be consistent with our observation that the highly purified tyrosine phosphorylated140 kD mIL-3R undergoes cleavage, and suggests that the protease remains physicallyassociated during the purification procedure. The amino-terminal 70 kD fragment (i.e.identified using an anti N-terminal mIL-3R antibody) likely contains the transmembrane regionof the mIL-3R since it is detected in Western blots of plasma membrane preparations andbecause of its likely identity with the 70 kD cross-linked species seen on intact cells andplasma membranes. 13Studies to date to examine the fate of the tyrosine and serine phosphorylated C-terminalfragment have been hindered by the lack of an anti-C terminal mIL-3R antibody that recognizesthe mIL-3R well in Western blots and by the inability to observe 32P-labelled fragmentsfollowing 4°C incubation of 32P-labelled 140 kD mIL-3Rs. The latter could be due either todephosphorylation or to further in vitro hydrolysis of this C-terminal fragment to peptides thatelectrophorese with the dye front on our 7.5% polyacrylamide gels. We are currentlyinvestigating the possible biological function(s) of the generated C-terminal fragment(s) byexpressing cDNAs corresponding to the intracellular domain in mIL-3 dependent 32D cells.132The identity of the protease responsible for receptor cleavage is unknown. Onepossibility is that the receptor itself contains an intrinsic protease, as has been hypothesizedfor the insulin receptor. 15 The protease activity of the insulin receptor is postulated to bepresent as an inactive zymogen until activated by the binding of insulin and this may explainthe insulinometic effect of trypsin. We have, however, found batch to batch variation inprotease sensitivity with our purified 140 kD mIL-3R preparations, consistent with the notionthat the receptor is acted upon by an exogenous protease that copurifies to varying extents. Apotential candidate for this mIL-3R specific protease may be calpain since it has been shownrecently to be involved in signal transduction. Specifically, in neutrophils stimulated with f-met-leu-phe it cleaves the Ca2+ and lipid dependent protein kinase C to the Ca2+ and lipidindependent protein kinase M. 16The function of the proteolytic cleavage of the mIL-3R is, of course, unknown at present.Receptor proteolysis could release a cytoplasmic receptor fragment that serves as part of amitogenic signal transducing pathway. Alternatively, the proteolysis may result in receptordown-regulation, analogous to phorbol ester stimulated cleavage and down-regulation of thecolony stimulating factor-1 receptor. 17 One possible mechanism, in this regard, is that mIL-3binding induces a conformational change in its receptor such that the associated tyrosinekinase becomes activated and phosphorylates the receptor and other receptor associatedproteins (which then act as second messengers to stimulate mitogenesis). The tyrosinephosphorylated mIL-3R then becomes a substrate for the associated serine kinase and thissubsequently leads to receptor cleavage and shutting down of the mitogenic signal. A thirdalternative is that generation of mIL-3R fragments may not play any role in mIL-3 inducedsignal transduction. Instead, receptor cleavage may be secondary and incidental to theactivation of the associated protease. The activated protease might then play an important rolein the propagation or regulation of the mitogenic signal by acting on mIL-3R associatedproteins. To gain further insights into the role of mIL-3R cleavage it would be of interest toexplore the biological consequences of preventing tyrosine phosphorylation and/or cleavageusing site directed mutagenesis of the AIC 2A product.1331 2134B-200-14-30-46-69-97.5- P-Ser- P-Thr- P-Tyrunbound^boundFigure 18. Tyrosine phosphorylation of the mIL-3R increases its apparent molecular mass.3 P-labeled tyrosine phosphorylated (panel A, lane 2) and tyrosineunphosphorylated (panel A, lane 1) mIL-3Rs were purified from 1 x 10 8 32P-labeled B6SUtA1 cells, as described in Chapter II. To compare thephosphoamino acid content of the two preparations, they were electrophoresedon SDS-polyacrylamide gels as shown in (A) and, the bands excised andsubjected to phosphoamino acid analysis (B). In panel B, three times moretyrosine unphosphorylated (unbound) than tyrosine phosphorylated (bound)mIL-3R's were analyzed. Similar results were obtained in two separateexperiments.4000 -2 3000 -a:5t 2000 -CCE 100000^10^20^30^40^50Incubation Time (hours)Figure 19. Tyrosine phosphorylation of the mIL-3R increases its susceptibility toproteolysis. Both tyrosine phosphorylated (4) and tyrosine unphosphorylated(C) mIL-3R preparations and a mixture of the two (a), were incubated for theindicated times at 4°C and either assayed for mIL-3 binding activity (A) in thelectin mIL-3R assayl or directly radioiodinated using the chloramine Tprocedure previously described 13 and analyzed by SDS-PAGE andautoradiography (B , on following page). For the lectin mIL-3R assay results,non-specific binding (which was determined using a 20 fold excess of unlabeledmIL-3) represented approximately 10 % of total binding and has beensubtracted. In B a mixture of protease inhibitors (0.5 mM PMSF, 1µg/mlleupeptin and 100 Kill/nil aprotinin) was also added to an aliquot of thetyrosine phosphorylated mIL-3R at the beginning of the incubation. The levelof protein in the non-tyrosine phosphorylated preparation was approximately10 fold that in the tyrosine phosphorylated preparation. Similar results wereobtained in two separate experiments.1350^16^25i) Non-tyrosine phosphorylated mIL-3RIncubation Time (hours)ii)Tyrosine phosphorylated mIL-3R0^16^25^25 +protease inhibitorsIncubation Time (hours)Figure 19B.^Tyrosine phosphorylation of the mIL-3R increases its susceptibility toproteolysis. Legend on previous page.1361371^2^3^4^5^6^7^8200- 97.5-69-46-30- jFigure 20. Increased proteolytic susceptibility of the tyrosine phospho rylated mIL-3R is anintrinsic property of the tyrosine phosphorylated receptor. 35S-labelled,tyrosine phosphorylated (lanes 5-6) and tyrosine unphosphorylated (lanes 1-4)mIL-3R's were purified as described in Chapter II from B6SUtA1 cellsequilibrated with 35S-methionine and stored at -20°C (lanes 1 and 5) orincubated at 4°C for 24 h with either no additions (lanes 2,6), or withunlabelled tyrosine phosphorylated mIL-3R (lanes 3,7) or unlabelled tyrosineunphosphorylated mIL-3R (lanes 4,8). The level of protein in the 35S-labeledand unlabeled preparations were identical so that the total amount of proteinwas approximately equal in lanes 3 and 8. The 35S-labeled bands, withapparent molecular masses ranging from 40 to 90 kD in lanes 1 through 4simply reflect the fact that the 120 kD mIL-3R preparation is less pure than the140 kD mIL-3R preparation. Proteins were visualized by autoradiography.Similar results were obtained in three separate experiments.1381^2 3 4^5 6200-97.5-69-46-30-Figure 21. Effect of phosphatase treatment on receptor integrity. Aliquots of a 140 kDmIL-3R preparation were incubated with three different phosphatases (orphosphatase control buffers) for 16 h at 37°C and than subjected to Westernanalysis using affinity purified anti-N terminal mIL-3R antibody. Lane 1, noaddition; 2, 100 U/ml T cell specific tyrosine phosphatase; 3, 1 U/ml each ofthe catalytic subunits of serine phosphatase 1 and 2A; 4, 1 U/ml of calfintestinal alkaline phosphatase; 5, tyrosine phosphatase buffer containing2mM EDTA; and 6, alkaline phosphatase buffer containing 1 mM mga2 and 2mM ZnC12. Similar results were obtained in three separate experiments.1390^5^10 15 30 60200-97.5-69-46-Figure 22. Degradation of the mIL-3R in intact cells. B6SUtA 1 cells were incubated at37°C with 30 nM mIL-3. At the indicated times (in minutes) the incubationwas stopped, plasma membranes were prepared and processed for Westernanalysis with anti-mIL-3R antibody as described in Chapter II. Similar resultswere obatined in four separate experiments.A 1^2 4 5 6140200-97.5-69-46-30-B^1 2 3 4 5 6200-97.5-69-Figure 23A.^Effect of various inhibitors on mIL-3R cleavage. B6SUtA 1 cells were incubatedfor 30 min in 10% FCS/RPMI with no addition (lane 1,2,6), 1 nM chloroquine(lane 3), 2 mM genistein (lane 4), or 0.3 gIVI staurosporine (lane 5). The cellswere then stimulated for 15 min with 30 nM mIL-3 (lanes 2,3,4,5) or 100 nMTPA (lane 6). Plasma membranes were prepared, solubilized in SDS-PAGEsample buffer, separated by SDS-PAGE and subjected to Western analysis withpolyclonal rabbit a-phosphotyrosine antibody. Similar results were obtained inthree separate experiments.Figure 23B.^Effect of kinase inhibitors on mIL-3R tyrosine phosphorylation. B6SUtA 1 cellswere treated as described in Figure 23A. The cells were then pelleted,solubilized with NP40 and irnmunoprecipitated with 1G2 a-phosphotyrosineSepharose. The immunoprecipitates were separated by SDS-PAGE andsubjected to Western analysis with polyclonal rabbit a-N terminal mIL-3Rantibody. Similar results were obtained in two separate experiments.141D. REFERENCES1. Mui ALF, Kay RJ, Humphries RK, Krystal G. Purification of the murineinterleukin-3 receptor. J Biol Chem (submitted).2. Itoh N, Yonehara S, Schreurs J et al. Cloning of an interleukin-3 receptor gene: Amember of a distinct receptor gene family. Science 1990;247:324-7.3. Krystal G, Alai M, Cutler RL, Dickeson H, Mui ALF, Wognum AW. Hematopoieticgrowth factor receptors. Hematol Pathol 1991;5:141-62.4. Park LS, Gillis S. Characterization of hematopoietic growth factor receptors. In:Dainiak V, Cronkite EP, McCaffrey R, Shadduck RK. The Biology of Hematopoiesis.Vol 352. New York; Wiley-Liss, 1990:189-96.5. Sorensen P, Mui ALF, "Crystal G. Interleukin-3 stimulates the tyrosinephosphorylation of the 140-kilodalton interleukin-3 receptor. J Biol Chem1989;264:19253-8.6. Isfort RJ, Stevens D, May WS, Ihle JN. Interleukin 3 binds to a 140-kDaphosphotyrosine-containing cell surface protein. Proc Natl Acad Sci U S A1988;85:7982-6.7. Miura 0, D'Andrea A, Kabat D, Ihle JN. Induction of Tyrosine Phosphorylation bythe Erythropoeitin Receptor Correlates with Mitogenesis. Mol Cell Biol1991;11:4895-902.8. Damen J, Hughes P, Humphries RK, Krystal G. Erythropoietin induces the tyrosinephosphorylation of a 70 KD erythropoietin receptor associated protein [abstract].Blood 1991;78:45.9. Mills GB, May C, McGill M et al. Interleukin 2-induced tyrosine phosphorylation. JBiol Chem 1990;265:3561-7.10. Murthy SC, Sorensen PHB, Mui ALF, Krystal G. Interleukin-3 down-regulates itsown receptor. Blood 1989;73:1180-7.11. Sorensen PHB, Mui ALF, Murthy SC, Krystal G. Interleukin-3, GM-CSF and TPAinduce distinct phosphorylation events in an interleukin 3-dependent multipotentialcell line. Blood 1989;73:406-18.12. Tamaoki K, Nomoto H, Takahashi I, Kato Y, Morimoto M, Tomita F. Staurosporine,a potent inhibitor of phospholipid-calcium dependent protein kinase. BiochemBiophys Res Commun 1986;135:397-402.13. Murthy SC, Mui ALF, Krystal G. Characterization of the interleukin 3 receptor. ExpHematol 1990;18:11-7.14. Mui ALF, Sorensen PHB, Murthy SC, Krystal G. Properties of the murineinterleukin-3 receptor. Exp Hematol Today 1989.^15.^Vega-SaenzdeMiera EC, Rubalcava B. Proteolytic activity in the insulin receptor.Biochem Biophys Res Commun 1988;156:30-7.14216. Pontremoli S, Michetti M, Mellon! E, Sparatore B, Salamino F, Horecker BL.Identification of the proteolytically activated form of protein kinease C in stimulatedhuman neutrophils. Proc Nall Acad Sci U S A 1990;87:3705-7.17. Downing JR, Roussel MF, Sherr CJ. Ligand and protein kinase C downmodulatethe colony-stimulating factor 1 receptor by independent mechanisms. Mol Cell Biol1989;9:2890-6.143CHAPTER VIDENTIFICATION OF PROTEINS POTENTIALLY ASSOCIATED WITH THE mIL-3RA. INTRODUCTIONThe binding of many hemopoietic growth factors to their receptors induces the tyrosinephosphorylation of several cellular substrates 1,2 including the receptor itself.3 '4 In the case ofCSF-1 and SF, the cellular phosphorylations are initiated by the activation of a receptorintrinsic tyrosine kinase. However, although at least three members of the hemopoietinreceptor superfamily, ie. the mIL-3R,4,8 EpR6 and IL-2R,3 become tyrosine phosphorylatedupon ligand binding, none of the hemopoietin receptor superfamily members possess a kinasedomain similar to any known tyrosine kinases. It is possible that these receptors possess anovel tyrosine kinase domain that is not homologous to previously characterized kinases. Aprecedent for this has recently been reported, namely, the bcr protein which is aserine/threonine kinase with no obvious homology to other known kinases. Alternatively,receptor tyrosine phosphorylation could be mediated by a receptor associated tyrosine kinase.Several members of the src family of tyrosine kinases have been shown recently to associatewith and phosphorylate cell surface molecules Examples include the p56 1c1c kinase whichbinds to the cytoplasmic tails of CD4 or CD8 and becomes activated upon crosslinking of thesemolecules:7,8 and the p60f3rn kinase, which is associated with the TcR and becomes activatedupon TcR ligation. 9 Recently, p561ck has also been shown to associate with the IL-2R, and beresponsible for tyrosine phosphorylation of the IL-2R 5 chain following IL-2 stimulation.'°Since we and others 11 have been unable to detect tyrosine kinase activity in purifiedpreparations of the mIL-3R, it is likely that tyrosine phosphorylation of the receptor is mediatedby an associated kinase. We therefore undertook studies directed at identifying andcharacterizing this putative kinase, as well as any other receptor associated proteins that maybe involved in mIL-3 induced signal transduction.144B. RESULTS 1. Comparison of Tyrosine Phosphorylations Induced by mIL-3, mGM-CSF and SFMost tyrosine kinases autophosphorylate upon activation, so as a first step in identifyingthe mIL-3R tyrosine kinase, we examined the tyrosine phosphorylations induced by mIL-3. Asshown in Figure 24, at 37°C , mIL-3 (lane 6) stimulated the tyrosine phosphorylation of severalproteins in B6SUtA 1 cells including ones at 140, 130, 95, 70, 56, and 32 kD. The 140 kDband we had previously shown to be the mIL-3R4 Interestingly, mGM-CSF (lane 7) and SF(lane 8) also induced the tyrosine phosphorylation of 56 and 32 kD proteins. Although furtherwork (ie. 2D gels, tryptic maps) is needed to establish that these two tyrosine phosphoproteinsare identical, this result does suggest that these two proteins may be components of a commonsignalling pathway utilized by all three growth factors. The 95 kD protein was tyrosinephosphorylated in response to both mIL-3 and mGM-CSF (lanes 6 and 7) as observedpreviously2 and may represent a common signal transduction intermediate. A slowermigrating 97 kD protein was induced by SF (lane 8). In order to determine whether thesephosphorylations were receptor proximal in the signalling cascade, we also looked at thephosphorylations induced at 4°C. At this temperature, membrane and cytoskeletal movementsare minimal so interactions should occur only between molecules which are already closelyassociated and events which are further down in the signalling cascade should not take placeat 4°C. 12 As shown in Figure 24, the mIL-3 induced phosphorylations after 10 min at 4°C(lane 2) were identical to those observed following 5 min at 37°C (lane 6). The rapid inductionkinetics at 4°C suggests that the mIL-3 activated tyrosine kinase may be associated with themIL-3R before ligand binding. In addition to the receptor, the other substrates of this kinase,i.e. the other proteins tyrosine phosphorylated in response to mIL-3 also appear to be receptorassociated. Interestingly, the SF induced tyrosine phosphorylation of the 150 kD protein,perhaps the c-kit tyrosine kinase, was more intense at 4°C (lane 4) than at 37°C (lane 8). Thismay reflect dephosphorylation of labile tyrosine phosphates at 37°C.145The kinetics of phosphorylation were then examined in more detail in order to try todistinguish the first autophosphorylation event of the activated tyrosine kinase fromsubsequent phosphorylations of exogenous substrates by this activated enzyme. As shown inFigure 25, phosphorylation of all proteins as well as the mIL-3R occurred, remarkably, within 2min at 4°C. However, the kinetics of phosphorylation and dephosphorylation of each proteindiffered, with the phosphorylation of pp56 and pp32 continuing to increase with time at 4°Cwhereas the level of phosphorylation of the 140, 90 and 70 kD proteins peaked at 2 min anddeclined rapidly thereafter. By 15 min, approximately half of the original tyrosinephosphorylation of the 140, 90 and 70 kD proteins remained. Although these results were veryinteresting, they did not allow us to clearly distinguish a defined sequence of phosphorylationevents and so did not allow assignment of any one of these proteins as a possible tyrosinekinase.2. Identity of the mIL-3 Stimulated Tyrosine Phosphoproteins with Known  ProteinsTo examine whether any of the mIL-3 stimulated tyrosine phosphorylated proteins wereidentical to tyrosine phosphorylated signal transduction intermediates previously described inthe literature, antibodies to potential candidates were obtained for analysis. Given that themolecular weight of a major protein tyrosine phosphorylated in response to mIL-3, mGM-CSFand SF was 56 kD, it was tempting to speculate that this protein might be a member of the srcfamily. However, pp56 did not react with antiserum to p5611ck (data not shown), a src familymember present predominantly in hemopoietic cells. 13 Moreover, other investigators haveshown that other src family members, ie. lck, src and lyn , are not involved in mIL-3 stimulatedsignal transduction. 6 However, this does not rule out the possibility that pp56 is a src familymember since a number of new src family members have recently been identified using lessstringent oligonucleotide screening techniques (Dr. J. Ihle, personal communication).Because of a recent report2° showing that the activation of a cdc2 kinase/cyclin Acomplex is an early event in in the response to PDGF, immunoblotting was carried out withantibodies to cdc2 (aPSTAIRE) and cyclin A (aCHLA-4) following immunoprecipitation of146tyrosine phosphoproteins from mIL-3 stimulated cells (Figure 26A). As Figure 26A shows, theaCHLA-4 reactive band did not line up with the mIL-3 stimulated pp56. In contrast, theaPSTAIRE reactive band aligned exactly with pp32. However, the aPSTAIRE reactive proteindid not appear to be pp32 since its tyrosine phosphorylation level appeared to drop rather thanincrease upon mIL-3 stimulation.A potential candidate for the identity of the pp95 was the fps kinase. However, althoughimmunoprecipitation with afps precipitated a 95 kD band, it did not undergo any changes inits phosphorylation state with mIL-3 (Figure 26B). In the same experiment, proteins wereimmunoprecipitated from mIL-3 treated and untreated cells with aGAP anti-serum, both as acontrol for the afps immunoprecipitation and to see if GAP became tyrosine phosphorylated inresponse to mIL-3. A 120 kD band, probably GAP, was precipitated from both control andmIL-3 stimulated cells, suggesting the tyrosine phosphorylation state of GAP does not changewith stimulation. Of interest, however, is the co-immunoprecipitation of a 95 kD protein frommIL-3 stimulated cells. This is especially interesting since GAP has been shown recently toassociate with tyrosine kinases 14 and so this associated 95 kD protein, which is most likelyidentical to the mIL-3R associated 95 kD phosphoprotein, may be a tyrosine kinase.GAP associates with tyrosine phosphorylated proteins through its two SH2 domains.Many proteins, in addition to GAP have now been described which contain SH2 domains, l4including the 97 kD proto-oncogene^. Since this protein has recently been shown to betyrosine phosphorylated under certain conditions, we investigated whether it could be the 95kD protein tyrosine phosphorylated in response to mIL-3 and mGM-CSF or the 97 kD proteintyrosine phosphorylated in response to SF. As Figure 26C shows, p 97vav appeared to beconstitutively tyrosine phosphorylated in B6SUtAi cells (lanes 1,5), although the addition ofeither mIL-3 or SF increased the level of tyrosine phosphorylation slightly. Thus the mIL-3stimulated pp95 is not p97vay. However, we did find that it was the 97 kD protein tyrosinephosphorylated in response to SF, both in B6SUtAi cells and in the human myeloblastic cellline, MOTE (15 and data not shown). Interestingly, upon longer exposure of the autoradiogram147in Figure 26C, a smaller band just below p97 vav could be seen in the mIL-3 and mGM-CSFlanes which might represent either a protein co-precipitating with activated p97vav or a p97vavdegradation product (Figure 26D).3. Effect of Various Kinase Inhibitors on mIL-3 Stimulated Tyrosine PhosphorylationsWithin the limitations of analysis, it did not appear that the mIL-3R associated tyrosinekinase nor any of the mIL-3 stimulated tyrosine phosphoproteins were previously isolatedprotein. Thus to further characterize the mIL-3R tyrosine kinase, the effect of genistein andstaurosporine on mIL-3 induced tyrosine phosphorylations were examined. Genistein is aspecific tyrosine kinase inhibitor while staurosporine is a serine/threonine kinase specificinhibitor at low concentrations and an inhibitor of tyrosine kinases at high concentrations. Asshown in Figure 27, genistein (lane 4) did not affect mIL-3 stimulated tyrosine phosphorylation.Thus the mIL-3R associated kinase is genistein insensitive. In contrast, the mIL-3R tyrosinekinase was profoundly inhibited by high concentrations of staurosporine (lane 5).Interestingly, TPA (lane 6), which is slightly mitogenic for B6SUtA 1 cells in the absence of mIL-3,2 did not induce tyrosine phosphorylations similar to mIL-3. However, in separateexperiments in which TPA was added to B6SUtA 1 cells and immunoblotting carried out withantibodies against the mIL-3R, it appeared that this phorbol ester caused the mIL-3R tomigrate more slowly in SDS polyacrylamide gels (Chapter III, Figure 23A, Lane 6). Thissuggests that protein kinase C may be the serine/threonine kinase responsible for serinephosphorylation of the mIL-3R Lastly, chloroquine, a lysosomal inhibitor was tested for itseffect on mIL-3 stimulated tyrosine phosphorylation events (lane 3). Interestingly, there was noobvious effect suggesting that internalization of the mIL-3R may not play a role in regulating,ie. limiting, the levels of tyrosine phosphorylation (at least not within 10 min at 37°C).4. Development of an Assay for the mIL-3R Associated Tyrosine Kinase To facilitate identification and purification of the mIL-3R associated tyrosine kinase, anassay was developed to detect mIL-3R specific tyrosine kinase activity. In this assay, B-mIL-3/mIL-3R complexes were isolated from B6SUtA 1 cell plasma membranes, as described in148Chapter II, in a buffer not containing phosphatase inhibitors. Preliminary experiments showedthat when no precautions were taken to inhibit tyrosine phosphatases, the tyrosinephosphorylation of the mIL-3R quickly declines to very low levels. These receptor complexeswere immobilized on Co-bind plates as described in Chapter II. Samples were then incubatedwith the immobilized receptors in the presence of ATP and Mg2+. After this incubation, thewells were washed and tyrosine phosphorylation detected with 125I-labelled a-phosphotyrosineantibodies. A summary of the assay procedure is shown in Figure 28. The specificity of thisassay is shown in Figure 29. As a source of mIL-3R tyrosine kinase, tyrosine phosphorylatedproteins were purified with a-phosphotyrosine Sepharose from B6SUtA1 cells stimulated withmIL-3 for 10 min at 4°C. As Figure 29 shows, when all components of the assay were present(+ATP), approximately 500 cpm were associated with the well in this particular experiment.Omission of ATP (-ATP) or tyrosine kinase sample (buffer) decreased the amount of countsbound. Omission of the mIL-3R, by adding a 20 fold excess of unbiotinylated mIL-3 during theB-mIL-3 isolation of mIL-3R's from plasma membranes, also decreased the level of tyrosinephosphorylation detected (+ cold mIL-3). Thus this assay appears to be capable of detectingtyrosine specific phosphorylation of the mIL-3RWith this assay we confirmed that the mIL-3R specific tyrosine kinase activity was in thea-phosphotyrosine bound fraction of B6SUtA 1 cells stimulated with mIL-3. Figure 30compares the mIL-3R tyrosine kinase activity isolated with a-phosphotyrosine Sepharose fromstimulated and unstimulated B6SUtA 1 cells. Considerably more activity was present in thesample from mIL-3 stimulated cells, suggesting that one of the tyrosine phosphorylatedproteins, or a non-tyrosine phosphorylated protein physically associated with one of theseproteins, was the mIL-3R kinase. In an attempt to determine which of these proteins was thetyrosine kinase, samples were fractionated on a Sephadex G150 column. The eluted fractionswere analyzed with this tyrosine kinase assay and the results plotted in Figure 31. Four peaksof activity were consistently observed, suggesting that either there are four distinct tyrosine149kinases, or that one kinase can associate to varying degrees with potential regulatorymolecules.C. DISCUSSIONAs was previously reported from our laboratory, several proteins in the mIL-3 dependentmurine hemopoietic cell line, B6SUtAi, become tyrosine phosphorylated upon mIL-3stimulation.2 However, these earlier studies were extended in the present work by thedemonstration that almost all the tyrosine phosphorylations observed at 37°C are also seen at4°C, the only exception being the minor tyrosine phosphorylated 42 IcD protein which we (datanot shown) and others 16 have identified as MAP kinase. The 4°C studies were conducted in anattempt to identify the earliest tyrosine phosphorylation events, and thus potentially identifythe mIL-3R specific tyrosine kinase. However, similar to results obtained from studies in thehuman GM-CSF system reported by Okuda et a1, 17 most of the phosphorylations observed at37°C were also seen at 4°C. This suggests either that our working hypothesis (that at 4°C,phosphorylations can only occur if proteins are already associated before ligand binding) isincorrect, or that all the observed phosphorylated proteins are indeed already receptorassociated. We favour the latter interpretation since phosphorylation occurs by 2 min at 4°C;the rapid kinetics makes it unlikely that these phosphorylated substrates diffuse together. Inaddition, many receptor systems, notably the intrinsic tyrosine kinase receptors and the srcfamily associated receptors such as the IL-2R and the TcR have been shown to exist inpreformed complexes with signalling proteins. 9,18 The physiological significance of thephosphorylation events we observe at 4°C is further supported by the fact that not allphosphorylations take place at 4°C. Phosphorylation and activation of p74raf-1,17 and MAP 16kinases do not occur at 4°C. An interesting corollary to this last observation is that p74raf -1and MAP may act further downstream in the signalling cascade than the processes regulatedby the phosphorylation events observed at 4°C.The five major mIL-3 stimulated phosphoproteins do not appear, within the limits of ouranalysis, to be p9OfPs, p97vav, cyclin A, p 5 6 hck or p 3 CdC2. However, the tyrosine150phosphorylation state of p34cdc2 decreased following 10 min of mIL-3 stimulation at 4°C. Thisobservation suggests that p34cdc 2 may have a role early in mIL-3 signal transduction sincetyrosine dephosphorylation of p34cdc2 occurs during activation of the kinase in M-phase. 19Furthermore, a proline directed protein kinase activity (PDPK) has been shown to be activatedearly in PDGF stimulated signal transduction. 20 This PDPK was subsequently shown toconsist of a cdc2/cyclin A complex. 21 Perhaps a cdc2 kinase activity has a similar role in themIL-3 system.Another possible identity for pp32 is the ras guanine nucleotide exchange protein(GEF). 22,23 The levels of ras GTP increase in cells in response to mIL-3 24 and this may be aconsequence of GAP inhibition. However, there is increasing evidence that acceleratednucleotide exchange may be important in mediating cellular responses to growth factors. 25Nucleotide exchange proteins may potentially be regulated by phosphorylation. 22 Thus it isinteresting to speculate that mIL-3 may regulate ras GTP levels by regulation, through tyrosinephosphorylation, of a guanine nucleotide exchange protein.One of the mIL-3 stimulated tyrosine phosphoproteins is likely to be the mIL-3R tyrosinekinase itself since most tyrosine kinases autophosphorylate upon activation. To identify thekinase, we developed an assay specific for tyrosine phosphorylation of the mIL-3R In order tomake the assay specific for tyrosine phosphorylation, we monitored phosphate incorporationusing labelled a-phosphotyrosine antibodies instead of 32P-ATP. A similar tyrosine kinaseassay was recently reported which utilized synthetic tyrosine containing peptides as asubstrate.26 However, since our assay uses purified mIL-R's as the substrate, it is morespecific for the mIL-3R tyrosine kinase. The major limitation of this assay is that it does notdistinguish between the action of an mIL-3R tyrosine kinase and the presence of an ATPdependent association of tyrosine phosphorylated proteins with the mIL-3R However, bothpossibilities are of interest to the study of mIL-3 induced signal transduction so this lack ofspecificity does not disqualify the utility of this assay. The assay is simple and allows the rapid151screening of multiple samples, e.g., column fractions. However, once an activity is isolated, itmust be analyzed for its ability to phosphorylate the mIL-3R on tyrosine residues.Four peaks of mIL-3R kinase activity were detected following Sephadex G150 gelfiltration chromatography. This suggests that there are either four kinases (or mIL-3associating proteins) or that the putative kinase is associated with other proteins which mayserve as regulatory subunits. The resolution of these possibilities awaits further purification ofthese mIL-3 induced tyrosine phosphoproteins.1521^2^3^5^6 7 8200-9 7 . 5 -6 9 -4 6 -3 0 -Figure 24. Comparison of tyrosine phosphorylations induced by mIL-3, mGM-CSF and SFin B6SUtA 1 cells. Cells were incubated for 10 min at 4°C (lanes 1-4) or 5 minat 37°C (lanes 5-8) with control buffer (1,5), mIL-3 (2,6), mGM-CSF (3,7) ormSF (4,8). Cells (1 X 106 per lane) were then solubilized with 0.5% NP40 asdescribed in Chapter II and subjected to Western analysis with a-phosphotyrosine Ab 4G10. Similar results were observed in three separateexperiments.Time course of tyrosine phosphorylation^4'C^ 37'CI—I0 2^5 10 15^15 30 60 120 15 mins200-15397 . 5 -6 9 -46-30-Figure 25. Time course of mIL-3 stimulated phosphorylations at 4°C. B6SUtA 1 cellswere stimulated for the indicated times at 4°C and solubilized in 0.5% NP40 asdescribed in Chapter II. The cell lysates (2 X 10 6 cells/sample) wereimmunoprecipitated with a-phosphotyrosine Sepharose and the phenylphosphate eluates analyzed by Western blotting with rabbit polyclonal a-phosphotyrosine Ab. For comparison, one sample (far right) was stimulated for10 min at 37°C. Similar results were obtained in two separate experiments.69-1541G2^CHLA-4^PSTAIRE200-97.5-46-30-Figure 26A. Comparison of pp56 and pp32 with cyclin A and cdc2 kinase. B6SUtA1 cellswere stimulated (+) or not (-) with mIL-3 for 10 min at 4°C. Tyrosinephosphorylated proteins were purified as described in Chapter II and separatedby SDS-PAGE and electrotransferred onto Immobilon. The membrane was cutand immunoblotted with a-phosphoWrosine Ab 4G10, a-CHLA4 or a-PSTAIREand developed with the appropriate 1251 labelled second antibody. Similarresults were obtained in three separate experiments.1 2 3 4Figure 26B. a-fps and a-GAP immunoprecipitation from mIL-3 treated B6SUtAl cells. Cellswere stimulated (lanes 2 and 4) or not (lanes 1 and 3) with mIL-3 andsolubilized in 0.5% NP40 as described in Chapter II. The cell lysates wereincubated with a-fps (lanes 1 and 2) or a-GAP (lanes 3 and 4) (both at 1/200dilution) rabbit anti-serum for 2 h and Protein A Sepharose was added for 2 h.The beads were washed and boiled in SDS-PAGE sample buffer and theproteins subjected to Western analysis with rabbit polyclonal a-phosphotyrosine Ab. Similar results were obtained in three separateexperiments.200-97.5-69-46-30-14-1551561^2^3 4^5^6^7 8 1 2^3^4^5^6^7^8200- 200-97.5.69-46-30-wryEffect of growth factor stimulation on the tyrosine phosphorylation of vay. B6SUtA1 cells were incubatedfor 10 min at 4°C (lanes 1-4) or 5 min at 37°C (lanes 5-8) with control buffer (1,5), mIL-3 (2,6), mGM-CSF(3,7) or mSF (4,8). Cells (2 X 106 cells/sample) were solubilized in 0.5% NP40 as described in Chapter ILThe lysates were incubated with a-vav rabbit serum (1/500 dilution) for 2 h, protein A Sepharose was thenadded and incubated for a further 2 h. After washing the beads, proteins were eluted by boiling in SDS-PAGE sample buffer and subjected to Western analysis with a-phosphotyrosine (4G10). Similar resultswere observed in three separate experiments.A 10 X longer exposure of the Western blot from Figure 26C.Figure 26C.Figure 26D.1571^2^3 4 5^6200-9 7 . 5 -6 9 -4 6 -3 0 -Figure 27. Effect of various inhibitors on mIL-3 stimulated tyrosine phosphorylations. B6SUtA 1cells (2 X 107 cells/sample) were incubated with 100 JIM chloroquine diphosphate(lane 3), 2 mM genistein (lane 4), 1 1.a4 staurosporine (lane 5) or control buffer (lanes1,2 and 6) for 30 min at 37°C. mIL-3 (lanes 2-5) or TPA (100 4M, lane 6) was thenadded, and the cells incubated a further 10 min. Samples were then processed as inFigure 24 and analyzed by immunoblotting with a-phosphotyrosine (4G10). Similarresults were obtained in three separate experiments.Figure 28. Scheme of mIL-3R tyrosine kinase assay. B-mIL-3/mIL-3R complexes were preparedin 0.5% NP40/PBS as described in Chapter II and immobilized onto Co-bind platesmodified with biotinylated gelatin. Samples to be assayed are added to the wells, intriplicate, and made 10 mM ATP and 10 mM MgC1 2 . The plates were incubated 12-16 h at 4°C, washed with 0.1 % NP40/PSB and incubated 2 h with 10 5 cpm/mL (50ng/mL) 123I labelled goat anti-mouse lg. After washing away unbound antibody,the contents of the well were eluted with 1% SDS and counted.158+ ATP^- ATP^BUFFER + COLD 1L3Figure 29. Specificity of the mIL-3R tyrosine kinase assay. B-mIL-3/mIL-3R complexes wereimmobilized to Co-bind plates as described in Chapter II. In one set of wells (+ coldmIL-3), a 20 fold excess of unbiotinylated mIL-3 was included with B-mIL-3 in thepreparation of mIL-3R complexes in order to control for non-specific binding of non-mIL-3R proteins to B-mIL-3. The wells were then incubated with the kinase sampleand ATP/MgC12 . ATP was omitted from some wells (-ATP); control buffer wassubstituted for the kinase sample in others (buffer). Tyrosine phosphateincorporation was measured as described above. Data points represent means (n=3)+/- SEM. Similar results were observed in two separate experiments.159200-97.5-69-46-40003000200010000A. B.1601^2- IL-3 + IL-330-laMMINIONgFigure 30. mIL-3R tyrosine kinase activity is present in the a-phosphotyrosine bound proteinfraction from mIL-3 stimulated B6SUtA1 cells. Tyrosine phosphorylated proteinswere purified from rnIL-3 unstimulated (panel A, lane 1) and stimulated (panel A lane2) cells with a-phosphotyrosine as described above and analyzed for activity in themIL-3R tyrosine kinase assay. The protein sample was visualized by a-phosphotyrosine Western (Figure 30A, also shown in Figure 32) and the tyrosinekinase activity compared in Figure 30B. In panel B, data points represent means(n=2) +/- SEM. Similar results were seen in two separate experiments.6000500040003000a.2000100016110^20^30fraction numberFigure 31. Sephadex G 150 fractionation of mIL-3 induced tyrosine phosphoproteins. a-phosphotyrosine Sepharose purified proteins from mIL-3 stimulated B6SUtA 1 cellswere separated on a 10 mL G150 sephadex size exclusion column equilibrated with0.1% NP40/2 mM Na3VO4/50 mM Hepes, pH7.5. Fractions, 0.5 mL each, werecollected and assayed for tyrosine kinase activity. Arrows mark the elution positionsof (from left to right) thyroglobulin (670 kD), Ig (158 kD), ovalbumin (44 kD) andmyoglobin (17 kD). Similar results were obtained in two separate experiments.162D. REFERENCES1. Uckun FM, Dibirkik I, Smith R et al. Interleukin 7 receptor ligation stimulatestyrosine phosphorylation, inositol phospholipid turnover, and clonal proliferation ofhuman B-cell precursors. Proc Natl Mad Sci U S A 1991;88:3589-93.2. Sorensen PHB, Mui ALF, Murthy SC, Krystal G. Interleukin-3, GM-CSF and TPAinduce distinct phosphorylation events in an interleukin 3-dependent multipotentialcell line. Blood 1989;73:406-18.3. Mills GB, May C, McGill M et al. Interleukin 2-induced tyrosine phosphorylation. JBiol Chem 1990;265:3561-7.4. Sorensen P, Mui ALF, Krystal G. Interleukin-3 stimulates the tyrosinephosphorylation of the 140-kilodalton interleukin-3 receptor. J Biol Chem1989;264:19253-8.5. Isfort RJ, Stevens D, May WS, Ihle JN. Interleukin 3 binds to a 140-kDaphosphotyrosine-containing cell surface protein. Proc Natl Acad Sci U S A1988;85:7982-6.6. Miura 0, D'Andrea A, Kabat D, Ihle JN. Induction of tyrosine phosphorylation by theerythropoeitin receptor correlates with mitogenesis. Mol Cell Biol 1991;11:4895-902.7. Veillette A, Bookman MA, Horak EM, Bolen JB. The CD4 and CD8 T cell surfaceantigens are associated with the internal membrane tyrosine protein kinase p56".Cell 1988;55:301-8.8. Shaw AS, Chlaupny J, Whitney A et al. Short related sequences in the cytoplasmicdomains of CD4 and CD8 mediate binding to the amino-terminal domain of the p56p 56kk protein kinase. Mol Cell Biol 1990;10:1853-62.9. Samelson LE, Philips AF, Loung ET, Klausner RD. Association of thefin proteintyrosine kinase with the T cell antigen receptor. Proc Natl Acad Sci U S A1990;87:4358-62.10. Yamanashi Y, Kakiuchi T, Mizuguchi J, Yamamoto T, Toyoshima K. Association of Bcell antigen receptor with protein tyrosine kinase lyn. Science 1991;251:192-84.11. Schreurs J, Hung P, May WS, Arai K, Miyajima A. AIC 2A is a component of thepurified high affinity mouse IL-3 receptor: temperature-dependent modulation ofAIC2A structure. Int Immunol (in press).12. Siegel JN, Klausner RD, Rapp UR, Samelson LE. T cell antigen receptor engagementstimulates C-raf phosphorylation and induces c-raf associated kinase activity via aprotein Manse C dependent pathway. J Biol Chem 1990;265:18472-80.13. Bolen JB, Thompson PA, Eiseman E, Horak ID. Expression and interactions of thesrc family of tyrosine kinases in T lymphocytes. Biochimica Biophysica Acta1989;1:103-49.16314. Koch CA, Anderson D, Moran MF, Ellis C, Pawson T. SH2 and SH3 domains:Elements that control interaction of cytoplasmic signaling proteins. Science1991;252:668-74.15. Mui ALF, Alai M, Cutler R, Bustelo X, Barbacid M, Krystal G. Steel factor stimulatesthe tyrosine phosphorylation of vav in both murine and human hemopoietic celllines (abstract). Blood 1991;78:161a.16. Okuda K, Kanakura Y, Hallek M, Druker B, Griffin JD. GM-CSF, IL-3 and steelfactor induce rapid tyrosine phosphorylation of p42 MAP kinase (abstract). Blood1991;78:370a.17. Okuda K, Druker B, Kanakura Y, Koenigsman M, Griffin JD. Internalization of thegranulocyte-macrophage colony stimulating factor receptor is not required forinduction of protein tyrosine phosphorylation. Blood 1991;78:1928-35.18. Hatakeyama M, Kono T, Kobayashi N et al. Interaction of the IL-2 receptor with thesrc-family kinase p56 p561ck: identification of novel intermolecular association.Science 1991;252:1523-8.19. Mm-la AO, Draetta G, Beach D, Wang JYJ. Reversible tyrosine phosphorylation ofcdc2: dephosphorylation accompanies activation during entry into mitosis. Cell1989;58:193-203.20. Hall FL, Mitchell JP, Vulliet PR Phosphorylation of synapsin I at a novel site byproline-directed protein kinase. J Biol Chem 1991;265:6944-8.21. Hall FL, Braun RK, Mihara K et at Characterization of the cytoplasmicproline-directed protein kinase in proliferative cells and tissues as a heterodimercomprised of p34 cdc2 and p58 cyclin A. J Biol Chem 1991;266:17430-40.22. Wolfrnan A, Macara IG. A cytosolic protein catalyzes the release of GDP from p2 lras.Science 1990;248:67-9.23. Huang YK, Kung H-F, Kamata T. Purification of a factor capable of stimulating theguanine nucleotide exchange reaction of ras proteins and its effect on ras-relatedsmall molecular mass G proteins. Proc Natl Acad Sci U S A 1990;87:8008-12.24. Satoh T, Nakafuku M, Miyajima A, Kaziro Y. Involvement of ras p21 protein insignal-transduction pathways from interleukin 2, interleukin 3, andgranulocyte/macrophage colony-stimulating factor, but not from interleukin 4. ProcNatl Acad Sci U S A 1991;88:3314-8.25. Jones S, Vignais M-L, Borach JR The CDC25 protein of saccharomyces cerevisiaepromotes exchange of guanine nucleotides bound to ras. Mol Cell Biol1991;11:2641-6.26. Babcock J, Watts J, Aebersold R, Ziltener HJ. Automated non-isotopic assay forprotein-tyrosine kinase and protein-tyrosine phosphatase activities. Anal Biochem(in press).164CHAPTER VIDEVELOPMENT OF STRATEGIES TO PURIFY mIL-3R ASSOCIATED PROTEINSA. INTRODUCTIONAs discussed in Chapter V, we have identified several proteins which could be potentiallyassociated with and mediate mIL-3R signal transduction. To further characterize theseproteins and determine their possible roles in mIL-3 stimulated signal transduction, weinvestigated various protocols to purify them to homogeneity. From our initial attempts atfractionating the mIL-3 induced phosphoproteins, and from our experience with thepurification of the mIL-3R itself, we concluded that standard chromatographic techniqueswould probably be inefficient and result in yields too low to obtain amino acid sequence data.We therefore examined various affinity purification protocols. For pragmatic reasons, wefocused our attention on the purification of the 95 IUD tyrosine phosphoprotein which co-precipitated with GAP. Since GAP often binds proteins through its SH2 domains, we reasonedthat an affinity column consisting of the SH2 domain of GAP might allow for the rapidpurification of pp95. We also decided to focus on pp95 because, as mentioned in Chapter V, itmay be a signal transduction intermediate common to mIL-3, mGM-CSF and SF. Sorensen etall for example, first described a cytosolic 95 Ic13 protein that became tyrosine phosphorylatedin response to both mIL-3 and mGM-CSF stimulated B6SUtA 1 cells. Others have also reporteda similar phosphoprotein induced by mIL-2, mIL-3, mGM-CSF and Ep in cells responsive tothese factors. 2 •3 In the human system, a protein with a very similar molecular weight wasreported to be the major tyrosine phosphorylated protein observed in response to hIL-3 andhGM-CSF. 4,5 These observations, along with the fact that pp95 undergoes phosphorylationwithin 2 minutes at 4°C, suggested that pp95 might be an important molecule in signaltransduction.165B. RESULTS1. The use of B-mIL-3 to purify mIL-3R associated proteins. Biotinylated mIL-3 was successfully used in the affinity purification of the mIL-3R asdiscussed in Chapter III. To determine if B-mIL-3/SA could also be used to purify mIL-3Rassociated proteins, B6SUtA 1 cells were stimulated with B-mIL-3 at 4°C for 5-15 min (Figure32), and were then solubilized and incubated with SA. For comparison, tyrosinephosphoproteins were immunoprecipitated with a-phosphotyrosine Sepharose. As Figure 32shows, the only tyrosine phosphorylated protein isolated with B-mIL-3 was the receptor itself.Thus, the other tyrosine phosphoproteins induced by mIL-3, which can be seen in the anti-PYlanes, did not co-purify with the mIL-3R There are several possible explanations for thisobservation. For example, the NP40 solubilization conditions might have been too harsh toallow survival of an intact complex. However, when we tried lower NP40 concentrations anddifferent detergents ( ie., OG and Chaps) they did not yield results different from those shownin Figure 32 (data not shown). It is also conceivable that, because of steric hindrance, the SAmight only be able to bind B-mIL-3/mIL-3R complexes from which receptor associated proteinshave dissociated. Lastly, it is possible that the tyrosine phosphorylated proteins, putativelyassociated with the mIL-3R, might have remained associated with the mIL-3R upon cell lysisbut became dissociated, because of high kD's, during the SA column washing steps.2. Association of mIL-3 Induced Tyrosine Phosphoproteins With SH2  domainsRecently, GAP and other SH2 domain containing proteins have been shown to bindtightly, via their SH2 domains, to various tyrosine phosphorylated proteins. 6 We thereforeexplored the possibility of using the SH2 domain of GAP to purify pp95. Specifically the SH2domain of GAP was expressed in E. co/i as a glutathione S-transferase (gst) fusion protein (seeChapter II), purified on a glutathione agarose column, and immobilized to Sepharose beads.These beads were then tested for their ability to bind pp95. As shown in Figure 33A, controlbeads containing only the gst portion of the fusion protein did not bind any tyrosinephosphorylated proteins from mIL-3 stimulated B6SUtA 1 cells. The GAP SH2-gst fusion166protein beads, however, bound several tyrosine phosphorylated B6SUtA1 proteins, includingthe pp140, pp95, pp70, and pp56 tyrosine phosphorylated proteins induced by mIL-3. The140 kD protein was shown to be the mIL-3R by immunoblotting duplicate samples with an a-Nterminal mIL-3R antibody (Figure 33B). The presence of all four mIL-3 induced tyrosinephosphoproteins could suggest that they all independently bind GAP SH2. This, however, isprobably unlikely. Alternatively, this apparent co-precipitation could simply be due toinsufficient washing of the beads and that the beads are not specifically binding proteins at all.However, this is not likely, since the GAP SH2 beads bound mIL-3R only from mIL-3 stimulatedcells (Figure 33B). A third possibility is that only one of these proteins actually binds to theGAP SH2 and the other proteins are simply associated with it. In this case, more stringentwashing should remove all but one phosphoprotein A fourth possibility is that the fourphosphoproteins specifically associate in vivo with SH2 domains of different proteins, eg.domains of PLCy or the PI3-kinase p85 subunit, but they can also bind with lower affinity tothe SH2 domain of GAP.To investigate these possibilities, gst fusion proteins of the SH2 domains of PLCy and thep85 subunit of PI3-kinase were immobilized to beads. The binding of mIL-3 induced tyrosinephosphoproteins was then analyzed as before, except with more extensive washes. Figure 34shows that, as before, no proteins bound to the control gst beads. However this time, aftersubjecting the beads to more washes, only the 140 kD mIL-3R bound to GAP SH2.Gratifyingly, the other SH2 domains had different specificities: the C-terminal SH2 domain ofthe p85 subunit of PI3-kinase specifically bound pp95. This dramatic finding not onlysuggests a powerful new technique for purifying proteins that are tyrosine phosphorylated inresponse to mIL-3 stimulation but also hints at specific protein/protein interactions that mayoccur in vivo as the result of mIL-3R activation.3. Hemopoietic Growth Factor Receptors and GAP SH2 DomainsOur finding that the mIL-3R bound to the SH2 domain of GAP begged the question ofwhether other hemopoietic growth factor receptors could do the same. As a preliminary test,167B6SUtA1 cells were stimulated with mIL-3, mGM-CSF or SF, solubilized as usual andincubated with GAP SH2 beads. The bound proteins were then analyzed by immunoblottingwith a-phosphotyrosine antibodies. As shown in Figure 35, a tyrosine phosphorylated band(lane 3) slightly higher in molecular weight than the phosphorylated mIL-3R (lane 2) wasspecifically precipitated from mGM-CSF stimulated cells. This protein had a molecular weighthighly reminiscent of the I3-subunit of the murine GM-CSF. 7 An even larger protein, with thesame molecular weight as the c-kit protein,8 was purified from SF treated cells. Further workis needed to establish that these are indeed the receptors for mGM-CSF and SF and todetermine the generality of this phenomenon within the hemopoietic system.C. DISCUSSIONAs discussed in Chapter V, our working hypotheisis is that the proteins which becometyrosine phosphorylated in response to mIL-3 at 4°C are mIL-3R associated proteins. In asurvey of purification techniques that might be used to purify these proteins, the mIL-3R wasfound to bind the SH2 domain of GAP while pp95 associated with the SH2 domain of the PI3-kinase p85 subunit. Thus these SH2 affinity matrices should prove highly useful in purifyingboth the mIL-3R and pp95. More importantly, the binding of the mIL-3R and pp95 to theseSH2 domains suggests that GAP and PI3-kinase might be involved in mIL-3 induced signaltransduction. Although it is well established that proteins bearing SH2 domains can bind togrowth factor receptors containing intrinsic tyrosine kinase domains, it has only recently beenshown that the IL-2R is associated with both p56k" and PI3-kinase. 1 ° This is not entirelysurprising since the IL-2R R chain has been shown to become tyrosine phosphorylatedfollowing IL-2 binding. 11From our observations, a model of mIL-3 signal transduction can be proposed in whichthe inactive mIL-3R is associated with a 95 kD protein and both of these becomephosphorylated on tyrosines upon mIL-3 binding. GAP may then bind to both thephosphorylated mIL-3R and pp95 may through SH2 and non-SH2 dependent interactions,respectively. pp95 also binds to PI3-K in an SH2 dependent manner. If this model is correct168then one might expect to see mIL-3R's as well as the pp95 in a-GAP immunoprecipitates.However, only pp95 was observed. To explain this discrepancy, one could postulate that thebinding of tyrosine phosphorylated mIL-3R's to GAP could make the GAP epitopes inaccessibleto a-GAP. In support of this, perhaps, is the fact that we observe binding of a 150 kD tyrosinephosphorylated protein, most likely p150c -kit to GAP SH2 domains (Figure 35). However,neither Rottapel et al8 nor Miyazawa 12 could detect a 150 kD tyrosine phosphorylated proteinin a-GAP immunoprecipitates. If the 150 kD protein is p150c -kit, then the failure of theseinvestigators to co-precipitate pp 150 might be due, as mentioned above, to the blocking ofepitope(s) recognized by a-GAP antiserum. For example, it is possible that the association ofthis large, detergent solubilized, SF receptor molecule with GAP prevents precipitation with a-GAP because of steric hindrance or epitope masking. While it may not be possible at this timeto fit the data we have gathered using the SH2 domains of GAP and PI3-K with our a-GAPresults, our preliminary findings suggest models that can be readily tested.169B-mIL-3/SA^anti-PYI _ 5' 10' 15' 1177200-97.5-69-46-30-Figure 32. B-mIL-3/streptavidin agarose precipitation of tyrosine phosphorylatedproteins from B6SUtA1 cells. For the B-mIL-3/SA lanes, B6SUtA 1 cells wereincubated with B-mIL-3 for the indicated times, or with a 20 fold excess ofunbiotinylated mIL-3 (-) for 10 minutes and solubilized in 0.5% NP40/PSB asdescribed in chapter II. Streptavidin-agarose was then added to precipitate theB-mIL-3 bound proteins. For the anti-PY lanes, B6SUtA1 cells were stimulated(+) or not (-), with mIL-3, solubilized and the phosphoproteins precipitated witha-phosphotyrosine as described in the legend to Figure 25. Similarexperiments were obtained in three separate experiments.170SH2 GAP SH2 GAP^gst(DMS)^(CNBr)^(DMS)200-97 . 5 -6 9 -46- ;3 0 -Figure 33A. Binding of B6SUtA 1 cell proteins to the SH2 domain of GAP. Gst fusionproteins were immobilized to Sepharose using the DMS or CNBr methods asdescribed in chapter II. The beads were then incubated with cell lysates,prepared as described in the legend to figure 25, from mIL-3 stimulated (+) andunstimulated (-) B6SUtA 1 . After washing, the beads were boiled in SDS-PAGEsample buffer and subjected to Western analysis with a-phosphotyrosine Ab(4G 10) (Figure 33A) or a-N terminal mIL-3R Ab (Figure 33B, on following page).171SH2 GAP SH2 GAP^gst(DMS)^(CNBr)^(DMS)200-97.5-69-46-30- sFigure 33B.^Binding of B6SUtA1 cell proteins to the SH2 domain of GAP. Legend onprevious page.N-GAP N-p85 C-p85 N-PLC C-PLC200-97 . 5-69 -46-Figure 34. Binding of B6SUtA1 cell proteins to various SH2 domains Gst fusion proteinsof the N-terminal GAP (N-GAP), N-terminal and C-terminal PI3-kinase p85subunit (N-p85 and C-p85), N-terminal and C-terminal PLCy (N-PLC and C-PLC) or the underivitized gst protein (-) were coupled to CNBr activatedSepharose as described in chapter II. The beads were then incubated withmIL-3 stimulated (+) and unstimulated (-) B6SUtA1 cell lysates as described inFigure 25. Bound proteins were then analyzed by immunoblotting with a-phosphotyrosine Ab (4G10). Similar results were observed in two separateexperiments.1721731^2^3^4200-97.5-69-46-30-Figure 35. Binding of mIL-3, mGM-CSF and SF induced tyrosine phosphoproteins to GAPSH2 domains. B6SUtA 1 cells were stimulated with control buffer (lane 1), mIL-3 (2), mGM-CSF (lane 3) or SF (lane 4) for 10 min at 4°C. Cell lysates wereprepared and incubated with GAP SH2 beads and bound proteins analyzed asdescribed in the legend to Figure 33A.174D. REFERENCES1. Sorensen PHB, Mui ALF, Murthy SC, Krystal G. Interleukin-3, GM-CSF andTPA induce distinct phosphorylation events in an interleukin 3-dependentmultipotential cell line. Blood 1989;73:406-18.2. Mona AO, Schreurs J, Miyajima A, Wang JYJ. Hematopoietic growth factorsactivate the tyrosine phosphorylation of distinct sets of proteins ininterleukin-3-dependent murine cell lines. Mol Cell Biol 1988;8:2214-8.3. Miura 0, D'Andrea A, Kabat D, Ihle JN. Induction of tyrosine phosphorylationby the erythropoietin receptor correlates with mitogenesis. Mol Cell Biol1991;11:4895-902.4. Kanakura Y, Druker B, Cannistra SA, Furukawa Y, Torimoto Y, Griffin JD.Signal transduction of the human granulocyte-macrophage colony-stimulatingfactor and interleukIn-3 receptors involves tyrosine phosphorylation of acommon set of cytoplasmic proteins. Blood 1990;76:706-15.5. Okuda K, Kanakura Y, Hallek M, Druker B, Griffin JD. GM-CSF, IL-3 and steelfactor induce rapid tyrosine phosphorylation of p42 MAP kinase (abstract).Blood 1991;78:370a.6. Koch CA, Anderson D, Moran MF, Ellis C, Pawson T. SH2 and SH3 domains:Elements that control interaction of cytoplasmic signaling proteins. Science1991;252:668-74.7. Kitamura T, Hayashida K, Sakamaki K, Yokota T, Arai K-I, Miyajima A.Reconstitution of functional receptors for human granulocyte/macrophagecolony-stimulating factor (GM-CSF): Evidence that the protein encoded by theAIC2B cDNA is a subunit of the murine GM-CSF receptor. Proc Natl Acad Sci US A 1991;88:5082-6.8. Rottapel R, Reedijk M, Williams DE et al. The steel/W transduction pathway:Kit autophosphorylation and its association with a unique subset ofcytoplasmic signaling proteins is induced by the Steel factor. Mol Cell Biol1991;11:3043-51.9. Hatakeyama M, Kono T, Kobayashi N et al. Interaction of the IL-2 receptorwith the src-family kinase p56 lck: identification of novel intermolecularassociation. Science 1991;252:1523-8.10. Remillard B, Petrillo R, Malinski W et al. Interleukin-2 receptor regulatesactivation of phsophatidylinositol 3-kinase. J Biol Chem 1991;266:14167-70.11. Yamanashi Y, Kakiuchi T, Mizuguchi J, Yamamoto T, Toyoshima K.Association of B cell antigen receptor with protein tyrosine kinase lyn. Science1991;251:192-84.12. Miyazawa K, Hendrie PC, Mantel C, Wood K, Ashman LK, Braxmeyer HE.Comparative analyisis of signaling pathways between mast cell growth factor(c-kit ligand) and granulocyte-macrophage colony stimulating factor in ahuman factor dependent myeloid cell line involves phosphorylation of raf-1.GTPase-activating protein and MAP kinase. Exp Hematol 1991;19:1110-23.175CHAPTER VIISUMMARY AND CONCLUSIONSA.^Purification of the mIL-3RIn order to purify the mIL-3R, a lectin assay based on Con A beads and unglycosylated1251_mm_3, was developed that could detect detergent solubilized mIL-3R's. With this assay,solubilization conditions that optimized retention of mIL-3R binding ability were determined.None of a range of standard chromatographic techniques which were tested gave satisfactoryincreases in purity or substantial yields of mIL-3R Similarly, none of the standard methods ofpreparing immobilized ligand affinity columns yielded an mIL-3 column that bound mIL-3R'swith good efficiency. However, we found that mIL-3 could be biotinylated without loss ofbioactivity and that this B-mIL-3 could be used in a simple two step procedure to the purify themIL-3R to apparent homogeneity. This two step protocol involved B-mIL-3, a-phosphotyrosineSepharose and streptavidin agarose and took advantage of the fact that the mIL-3R becomesphosphorylated on tyrosine residues upon mIL-3 binding to intact B6SUtA 1 cells. N-terminalamino acid sequencing of the purified mIL-3R revealed it to be identical to the Aic 2A proteinsubsequently cloned by expression in COS cells using an antibody which partially inhibits mIL-3 binding to mIL-3R positive cells.'As assessed by the lectin assay, both the NP40 and OG solubilized mIL-3R's displayedhigh affinity binding equal to that detected on intact cells. Scatchard analysis of semi-purifiedand pure mIL-3R preparations suggested that high affinity binding was not lost uponpurification of the 120 - 140 kD mIL-3R This result is not readily reconciled with theobservation that the Aic 2A protein exhibits low affinity binding when expressed in fibroblasts.Examination of the contribution of carbohydrate to mIL-3 binding affinity showed that thedeglycosylated mIL-3R possessed lower binding affinity that the properly glycosylated molecule.Thus it is possible that glycosylation differences between hemopoietic cells and fibroblasts may176partially explain the differences in affinity. However, many other members of the hemopoieticreceptor family consist of two subunits, and both are required for high affinity binding. 2 Aswell, in chemical cross-linking experiments with 125I-mIL-3, two proteins with molecularweights of 120 and 70 kD are consistently seen. Although we have shown previously that thehigher molecular weight protein undergoes proteolysis to generate a 70 kD fragment, 3 it ispossible that a portion of the 70 kD labelled band seen on cross-linking gels is due to anothersubunit that associates with the 120 kD molecule to produce high affinity binding. In thiscase, our high affinity estimate of the purified 140 kD molecule might be due to an artefact ofthe lectin assay or to aggregation of solubilized mIL-3R's. Aggregation of receptors in othersystems has been reported to increase their affinity for their ligands. 4 ' 5 Further study,including, the isolation of the putative second mIL-3R subunit, is needed to resolve the issue ofwhether the mIL-3R is composed of more than one ligand binding subunit.B.^Tyrosine Phosphorylation of the mIL-3R Increases its Susceptibility to ProteolysisDuring the course of the purification of the mIL-3R, we noticed that, the serinephosphorylated mIL-3R was very stable while the tyrosine/serine phosphorylated 140 kD formrapidly degraded. Closer examination of this phenomenon using specific phosphatasesrevealed that both tyrosine and serine phosphorylations were required for degradation to occur.Moreover, this proteolysis of the mIL-3R was also found to occur in intact cells in response tomIL-3. Agents that inhibited the tyrosine phosphorylation of the mIL-3R inhibited mIL-3Rcleavage, whereas agents that did not affect tyrosine phosphorylation were unable to altercleavage. These observations are consistent with a model in which mIL-3 binding to itsreceptor results in both the tyrosine and serine phosphorylation of the mIL-3R This convertsthe mIL-3R into a substrate for a receptor associated protease which cleaves the TnIL-3R togenerate a 70 kD membrane associated molecule and release a cytoplasmic fragment. Thegeneration of these fragments may be involved in receptor downregulation and/or propagatingthe mitogenic signal initiated by mIL-3 binding. Determination of the proteolytic site, followed177by site directed mutagenesis of this region to prevent proteolysis will allow testing of thesepossibilities.C.^Identification of mIL-3R Associated ProteinsThe mIL-3R does not contain an intrinsic tyrosine kinase domain, although it doesbecome phosphorylated on tyrosine residues upon mIL-3 stimulation. Thus at least a tyrosinekinase, and also an mIL-3R protease, become associated with the mIL-3R. A kineticexamination of mIL-3 induced tyrosine phosphorylations showed, remarkably, thatphosphorylation of the mIL-3R and four other proteins, with apparent molecular weights of 95,70, 56 and 32 Id), occur by 2 min at 4°C. This observation suggests that the kinase isassociated with the mIL-3R and its other substrates prior to mIL-3 binding. One of these mIL-3 induced tyrosine phosphoproteins, pp95, co-immunoprecipitates with GAP. Since GAPassociates with tyrosine kinases, this association makes pp95 a good candidate for the mIL-3Rspecific tyrosine kinase.Since tyrosine kinases associate with GAP through GAP SH2 domains, an affinity matrixconsisting of GAP SH2 domains was tested for its suitability in purifying pp95. Unexpectedly,we discovered that the mIL-3R bound tightly to GAP SH2 domains whereas pp95 did not. Thisobservation suggests that the association of pp95 with GAP might occur through an SH2independent mechanism. However, pp95 did bind to an affinity matrix constructed of the SH2domains of the p85 subunit of PI3-kinase. Further work is needed to determine whether thepp95 binding in vitro to the p85 SH2 domains reflects an in vivo association with PI3-kinase.However, these preliminary results are intriguing in the light of the recent report of theassociation of PI3-kinase with p2lras. 8 One could envisage a model in which ras co-precipitates with PI3-kinase because of the association of pp95 with GAP and PI3-kinase,through SH2 independent and dependent mechanisms, respectively. Further experiments arerequired to test this model, and may yield insights into the mechanism of action of mIL-3.178D. A Model for mIL-3R Signal TransductionBased on the studies reported in this thesis and on previous work carried out in ourlaboratory, the following model of mIL-3 induced signal transduction events is proposed (Figure36). The ligand unoccupied, inactive receptor (lower left hand corner of Figure 36) is a 120 kDglycoprotein that is associated via its cytosolic domain with at least four other signallingmolecules with apparent molecular weights of 95, 70, 56 and 32 kD. Upon mIL-3 binding, thereceptor becomes phosphorylated on tyrosines and serines and the tyrosine/serinephosphorylated receptor becomes cleaved to release a cytoplasmic fragment. The mIL-3Rassociated proteins also become phosphorylated on tyrosines and serines and are releasedfrom the mIL-3R One of these proteins, pp95, becomes associated with both GAP and PI3-kinase. The association of pp95 with GAP may attenuate GAP's ras GTPase stimulatory activityand thus account for the accumulation of ras GTP that is observed in cells stimulated withmIL-3. 6 The association of pp95 with PI3-kinase may similarly modulate the activity of thiskinase. Also shown in Figure 36, is a 42 kD protein, MAP kinase, which we and others 7 haveshown to become tyrosine phosphorylated and activated in response to mIL-3. However, thisreaction occurs only at 37°C, suggesting perhaps that this is a late event in the signaltransduction cascade. Although many aspects of this model are still somewhat speculative, itdoes provide an interesting basis for further experimentation.179Figure 36. Model of mIL-3R mediated signal transduction180E. REFERENCES1. Yonehara S, Ishii A, Yonehara M et al. Identification of a cell surface 105 kD protein(Aic-2 antigen) which binds interleukin-3.•Int Immunol 1990;2:143-50.2. Miyajima A, Kitamura T, Harada N, Yokota T, Arai K. Cytokine receptors and signaltransduction. Annu Rev Immunol 1991;10:295-331.Murthy SC, Mui ALF, Krystal G. Characterization of the interleukin 3 receptor. ExpHematol 1990;18:11-7.4. Fukunaga R, Ishizaka-Ikelda E, Nagata S. Purification and characterisation of thereceptor for murine granulocyte colony-stimulating factor. J Biol Chem1990;265:14008-15.5. Goodwin RG, Friend D, Ziegler SF et al. Cloning of the human and murineinterleukin-7 receptors: Demonstration of a soluble form and homology to a newreceptor superfamily. Cell 1990; 60:941-51.6. Satoh T, Nakafuku M, Miyajima A, Kaziro Y. Involvement of ras p21 protein insignal-transduction pathways from interleukin 2, interleukin 3, andgranulocyte/macrophage colony-stimulating factor, but not from interleukin 4. ProcNatl Acad Sci U S A 1991;88:3314-8.7. Okuda K, Kanakura Y, Hallek M, Druker B, Griffin JD. GM-CSF, IL-3 and steelfactor induce rapid tyrosine phosphorylation of p42 MAP kinase [abstract]. Blood1991;78:370a.8.^Sjolander A, Yamamtoto K, Huber BE, Lapetina EG. Association of p21 ras withphosphatidylinositol 3-kinase. Proc Natl Acad Sci U S A 1991;88:7908-12.


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