UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The role of SRC family protein tyrosine kinases and the GAB1 adapter protein in BCR-mediated activation… Santos, Lorna 2004

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

Item Metadata

Download

Media
831-ubc_2004-902625.pdf [ 31.97MB ]
Metadata
JSON: 831-1.0092322.json
JSON-LD: 831-1.0092322-ld.json
RDF/XML (Pretty): 831-1.0092322-rdf.xml
RDF/JSON: 831-1.0092322-rdf.json
Turtle: 831-1.0092322-turtle.txt
N-Triples: 831-1.0092322-rdf-ntriples.txt
Original Record: 831-1.0092322-source.json
Full Text
831-1.0092322-fulltext.txt
Citation
831-1.0092322.ris

Full Text

THE ROLE OF SRC FAMILY PROTEIN TYROSINE KINASES AND THE GAB 1 ADAPTER PROTEIN IN BCR-MEDIATED ACTIVATION OF AKT by LORNA SANTOS Bachelor of Science, University of British Columbia, 1997  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENT FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Department of Zoology)  We accept this thesis as-eenforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA, © Lorna Santos, 2003  ABSTRACT Antibody production by B lymphocytes is important for the eradication of foreign pathogens and requires signaling by the BCR. Crosslinking of the BCR results in the activation of three main signaling pathways — the PI 3-kinase, MAP kinase and PLC-y2 pathways. It is proposed that various combinations of protein tyrosine kinases and adapter proteins differentially regulate the three main signaling pathways following BCR engagement. A non-lymphoid cell system that stably expresses functional BCRs at the cell surface and expresses only one Src kinase (Fyn) was used to examine the role of the Gabl adapter protein and Src family members, Lyn, Blk and Lck, in the BCR-induced activation of the Akt pathway. It was determined that Syk kinase activity is required for amplifying and sustaining Akt pathway activation through the BCR.  Lck has no effect on the  activation of this pathway while Lyn and Blk both inhibit the activation of this pathway in response to BCR crosslinking. This inhibition correlated with the ability of Lyn and Blk to interact with the SHP-2 protein tyrosine phosphatase. Altering the plasma membrane localization of Lyn and Blk in lipid raft and non-rafts domains has no effect on the BCRinduced Akt phosphorylation. However, expression of a cytosolic mutant form of Blk in the AtT20 cells alleviates the inhibitory effect on Akt phosphorylation. Using the AtT20 cells, it was established that Gabl could inducibly translocate to the plasma membrane and become tyrosine phosphorylated in response to BCR crosslinking. Akt phosphorylation was 2-fold higher following BCR crosslinking in cell transfected with Gabl.  The  recruitment of Gabl to the plasma membrane, its inducible tyrosine phosphorylation, association with signaling molecules and increased Akt phosphorylation all required the PH domain of Gabl as well as the Syk kinase. Together, both parts of this thesis demonstrated that various BCR-activated PTKs and Gabl are involved in the regulation of Akt pathway downstream of the BCR. The Akt pathway could be modulated by the use of bona fide adapter proteins or PTKs acting as adapter proteins that recruit important signaling components to the plasma membrane where PI 3-kinase/Akt signaling takes place.  TABLE OF CONTENTS Abstract  ii  Table of Contents  iii  List of Figures  ix  List of Tables  xiv  List of Abbreviations  xv  Acknowledgements  xix  Chapter 1:  Introduction  1  1.1  The importance and role of lymphocytes in the immune response  1  1.2  B lymphocyte development  4  1.3  The B cell antigen receptor  8  1.4  Membrane proximal events in BCR-mediated signal transduction  13  1.5  Signaling pathways regulated by the BCR  16  1.5.1  PLC-y pathway  16  1.5.2  MAP kinase and SAP kinase pathways  21  1.5.3  PI3K/Akt pathway  25  1.6  Protein tyrosine kinases involved in proximal signaling events by the BCR . . 34 1.6.1  SykPTK  34  1.6.2  Src kinase family tyrosine kinases  35  1.6.3  Src kinase structure and activation  41  1.6.4  Src kinase function in B lymphocytes  46  1.6.5  Src kinase knockout studies  49  1.7  The Gabl adaptor protein  51  1.8  Thesis Goals  59  1.9  Thesis Summary  60  Chapter 2:  2.1  2.2  2.3  2.4  Materials and Methods  62  Reagents  62  2.1.1  Antibodies  62  2.1.2  Plasmids  63  2.1.3  GST fusion proteins  64  Tissue culture and cell stimulations  66  2.2.1  Tissue culture cell lines  66  2.2.2  Culture of cell lines  66  2.2.3  Stimulation and lysis of cells  67  DNA transfection of cells  69  2.3.1  Transient transfection of BOSC 23 cells  69  2.3.2  Calcium phosphate transfection of AtT20 cells  70  2.3.3  Drug selection of transfected cells and isolation of individual clones . .71  Retroviral infection of cells . . .  72  2.4.1  Production of retroviruses using BOSC 23 packaging cell line  72  2.4.2  Retroviral infection of AtT20 cells  72  2.5  SDS-PAGE and Western immunoblotting  73  2.6  Precipitation experiments  74  2.6.1  Immunoprecipitations  74  2.6.2  Pull-down experiments with GST fusion proteins  75  2.6.3  In vitro kinase assay  76  2.7  2.8  Isolation of lipid rafts by sucrose density gradient ultracentrifugation  76  2.7.1  Cell stimulation for sucrose density gradient ultracentrifugation  76  2.7.2  Discontinuous sucrose density gradient ultracentrifugation  77  Membrane enrichment  78  2.9  2.10  2.11  Confocal microscopy  79  2.9.1  Preparation of poly-D-lysine coated coverslips  79  2.9.2  Culture and stimulation of cells on coverslips  79  2.9.3  Preparation of cells for confocal microscopy  80  Molecular biology methods  80  2.10.1 Restriction endonuclease reactions  80  2.10.2 Alkaline phosphatase reactions  81  2.10.3 Agarose gel electrophoresis  81  2.10.4 Gel purification of digested DNA  81  2.10.5 Ligation of purified DNA fragments  82  2.10.6 Transformation of competent Escherichia coli bacteria  82  2.10.7 Site-directed mutagenesis  82  2.10.8 Small scale preparation of DNA  83  2.10.9 Large scale preparation of DNA  83  Plasmids generated  84  Chapter 3:  Syk and Src family protein tyrosine kinases differentially regulate the PI 3-kinase/Akt pathway  92  3.1  Introduction  92  3.2  Results  95  3.2.1  Fyn is sufficient to activate Akt and Syk is required for sustained Akt activation  3.2.2  Syk kinase activity is required for promoting and sustaining Akt phosphorylation and activation  3.2.3  110  Lyn and Blk, in conjunction with Syk inhibit BCR-induced Akt phosphorylation and activation  3.3  104  The Src kinases, Lyn, Lck and Blk, differentially regulate BCR -mediated Akt phosphorylation and activation  3.2.4  95  Discussion  124 137  Chapter 4:  Lyn and Blk inhibit Akt phosphorylation and activation by associating with the SHP-2 protein tyrosine phosphatase  145  4.1  Introduction  145  4.2  Results  147  4.2.1  Lyn and Blk associate with the SHP-2 protein tyrosine phosphatase in AtT20 transfected cells  4.2.2  BCR-induced SHP-2 tyrosine phosphorylation is greater in AtT20 cells co-expressing Lyn and Syk  4.2.3  Discussion  Chapter 5:  152  Lyn and Blk associate with the SH2 domains of SHP-2 protein tyrosine phosphatase in vitro  4.3  147  153 161  The kinase activity of Lyn is required for BCR-induced inhibition of Akt phosphorylation and activation  170  5.1  Introduction  170  5.2  Results  171  5.2.1  The kinase activity of Lyn is required for BCR- mediated suppression of Akt phosphorylation and activation  5.2.2  Wild type and kinase dead forms of Lyn associate with SHP-2 protein tyrosine phosphatase in transfected AtT20 cells  5.2.3  179  The kinase activity of Lyn and Syk are required for BCR -induced phosphorylation of cellular SHP-2  5.3  171  Discussion  182 183  Chapter 6:  Altering the membrane localization of Lyn and Blk does not affect BCR-mediated inhibition of Akt phosphorylation and activation. .189  6.1  Introduction  189  6.2  Results  193  6.2.1  Chains of the BCR translocate into lipid rafts following receptor cross-linking  6.2.2  The Src kinases, Lyn, Lck and Blk, are differentially distributed in lipid rafts  6.2.3  193  201  Altering the membrane localization of Lyn and Blk does not alleviate the inhibition of Akt phosphorylation following BCR cross-linking  6.2.4  Altering the membrane localization of Lyn and Blk does not affect BCR-mediated SHP-2 phosphorylation  6.2.5  Discussion  Chapter 7:  221  Targeting of Blk to the cytoplasm alleviates the inhibition of Akt phosphorylation following BCR cross-linking  6.3  210  221 231  The Gabl adaptor protein enhances PI3K/Akt pathway activation through the B cell antigen receptor  234  7.1  Introduction  234  7.2  Results  236  7.2.1  Gabl is recruited to the plasma membrane in response to BCR cross-linking  7.2.2  Gabl is tyrosine phosphorylated in response to BCR engagement . . .245  7.2.3  Gabl interacts with various tyrosine phosphorylated proteins  7.2.4 7.3  236  following BCR cross-linking  248  BCR-induced Akt phosphorylation is enhanced by Gabl  251  Discussion  252  Chapter 8  Discussion  261  8.1  Summary of results and future directions  261  8.2  Discussion  272  Reference List  280  Appendix  326  LIST OF FIGURES  Figure 1.1  Schematic representation of B lymphocyte development  Figure 1.2  Schematic representation of the BCR complex  Figure 1.3  BCR cross-linking results in the activation of the PLC-y pathway . . . 19  Figure 1.4  The MAP kinase and SAP kinase pathways are activated in response to BCR cross-linking  Figure 1.5  23  27  BCR cross-linking results in the activation of the PI 3-kinase pathway  Figure 1.7  11  Schematic representation of the PI 3-kinase subunits, PDK1 and Akt structures  Figure 1.6  6  32  Schematic representation of the primary structures of tyrosine kinases involved in the proximal signaling downstream of the BCR  37  Figure 1.8  Phylogenetic tree of the Src kinase family members  40  Figure 1.9  Mechanism of Src kinase activation  45  Figure 1.10  Schematic representation of Gabl structure  54  Figure 1.11  Model of pathways regulated by the Gabl adapter protein  58  Figure 2.1  Strategy for generating the pMSCV-Lyn expression vector  86  Figure 2.2  Strategy for generating the pMSCV-Blk (cytosolic mutant) expression vector  Figure 3.1  BCR cross-linking in the AtT20 BCR+Fyn+cells can promote phosphorylation of. Akt  Figure 3.2  101  BCR-induced Akt phosphorylation in the AtT20 cells is PI 3-kinase-dependent  Figure 3.4  98  Syk PTK is required for amplifying and sustaining BCR-induced Akt phosphorylation in the AtT20 cell system  Figure 3.3  88  103  Syk kinase activity is required for BCR-induced Akt phosphorylation in AtT20 cells  107  Figure 3.5  Hyper-expression of the SHP-1 protein tyrosine phosphatase in Sykexpressing AtT20 cells inhibits BCR-induced Akt phosphorylation . . 109  Figure 3.6  Multiple Src kinase family members are expressed in various B lymphoma cell lines and mouse splenic B cells  Figure 3.7  Expression  levels of  Src kinases and BCR-induced  112 tyrosine  phosphorylation of total cellular protein in transfected AtT20 cells . .115 Figure 3.8  Lck has no effect BCR-induced Akt phosphorylation of the serine 473 residue in transfected AtT20 cells  Figure 3.9  Lyn increases the kinetics of BCR-induced Akt phosphorylation of the serine 473 residue in transfected AtT20 cells  Figure 3.10  121  Blk decreases the kinetics of BCR-induced Akt phosphorylation of the serine 473 residue in transfected AtT20 cells  Figure 3.11  119  123  Expression levels of Src kinases and BCR-induced tyrosine phosphorylation of total cellular protein in transfected AtT20 cells co-expressing Syk PTK  Figure 3.12  Lck in the presence of Syk does not affect BCR-induced Akt phosphorylation in transfected AtT20 cells  Figure 3.13  132  Blk has an inhibitory effect on BCR-induced Akt phosphorylation of the serine 473 residue in the presence of Syk PTK  Figure 3.15  130  Lyn has an inhibitory effect on BCR-induced Akt phosphorylation of the serine 473 residue in the presence of Syk PTK  Figure 3.14  126  134  The extent of BCR-mediated inhibition of Akt phosphorylation in the AtT20 BCR+ Fyn+ Syk+ Lyn+ and AtT20 BCR+ Fyn+ Syk+ Blk+ cell lines is similar  Figure 3.16  Proposed mechanism of BCR-induced phosphorylation of Akt in Syk-transfected AtT20 cells  Figure 4.1  140  The Src kinases, Lyn and Blk, constitutively associate with the SHP-2 protein tyrosine phosphatase in the transfected AtT20 cells  Figure 4.2  136  151  Phosphorylation of SHP-2 on tyrosine 580 was slightly higher in AtT20 cells co-expressing Syk and Lyn PTKs  155  Figure 4.3  The SH2 domain(s) of SHP-2 mediate its interaction with Lyn  and Blk in vitro Figure 4.4  SHP-2 interaction with Lyn and Blk is inhibited by SHP-2 GST fusion protein  Figure 4.5  164  Amino acid sequence of putative ITIM motifs within the kinase domain of the Src family members  Figure 5.1  160  Schematic representation of SHP-2 structure and mechanism of activation  Figure 4.6  158  169  Expression levels of wild type and kinase dead forms of Lyn and Syk and BCR-induced tyrosine phosphorylation of total cellular protein in transfected AtT20 cells  Figure 5.2  Lyn kinase activity is required for BCR-induced suppression of Akt phosphorylation in the AtT20 cells  Figure 5.3  174  178  The wild type and kinase dead forms of Lyn, constitutively associate with the SHP-2 protein tyrosine phosphatase in the transfected AtT20 cells  Figure 5.4  181  The kinase activity of both Lyn and Syk are important for the phosphorylation of SHP-2 on the tyrosine 580 residue  185  Figure 6.1  Fractions collected from the sucrose gradient  195  Figure 6.2  Inducible translocation of the BCR chains into lipid rafts following receptor cross-linking in B lymphoma cell lines  Figure 6.3  Inducible translocation of the BCR chains into lipid rafts following receptor cross-linking in transfected AtT20 cells  Figure 6.4  203  Src kinases are differentially localized in the different sucrose gradient fractions from transfected AtT20 cells  Figure 6.6  200  Distribution of Lyn, Lck, Blk and Syk in B lymphoma cell lines in the sucrose gradient  Figure 6.5  197  207  Generation of mutants of Lyn and Blk with alterations in their acylation sites  212  Figure 6.7  Expression levels and distribution of non-palmitylated Lyn and palmitylated Blk in the sucrose gradient  Figure 6.8  Distribution of mutant Lyn and Blk proteins in the different sucrose gradient fractions is unaffected by BCR cross-linking  Figure 6.9  215  217  Altering the distribution of Lyn and Blk in lipid rafts has no effect on Akt phosphorylation following BCR cross-linking  220  Figure 6.10  SHP-2 interacts with the acyl mutants of Lyn and Blk in vitro  223  Figure 6.11  Altering the distribution of Lyn and Blk in lipid rafts has no effect on SHP-2 phosphorylation following BCR cross-linking  Figure 6.12  Expression levels and distribution of cytosolic Blk mutant in the sucrose gradient  Figure 6.13  244  Gabl-EFGP is inducibly tyrosine phosphorylated following BCR engagement in AtT20 cells co-transfected with Syk PTK  Figure 7.5  242  Recruitment of EGFP-tagged Gabl to the plasma membrane following BCR engagement is PI 3-kinase dependent  Figure 7.4  238  Gabl-EGFP translocates to the plasma membrane following BCR cross-linking  Figure 7.3  230  EGFP-tagged wild type and APH Gabl structure and expression in transfected AtT20 cells  Figure 7.2  228  BCR-induced Akt phosphorylation was not inhibited in AtT20 cells expressing the soluble form of Blk  Figure 7.1  225  247  Gabl-EFGP interacts with several tyrosine phosphorylated proteins following BCR cross-linking in AtT20 cells cotransfected with Syk  Figure 7.6  250  Wild type Gabl-EFGP, but not GablAPH-EGFP, potentiates BCRinduced Akt phosphorylation in AtT20 BCR+ Fyn+ cells co-  Figure 7.7  expressing Syk PTK  254  Proposed model of BCR-mediated regulation of Gabl  260  Figure 8.1  Figure 8.2  Proposed model of Lyn- and Blk-mediated inhibition of the PI 3kinase/Akt pathway in AtT20 cells  265  Proposed location of cellular SHP-2 molecules in AtT20 cells  267  Appendix Figure 1  Sequence alignments of Src kinase family members  326  LIST OF TABLES Table 1.1  Expression patterns of the Src family members expressed in cells of the lymphocyte lineage  Table 2.1  47  Oligonucleotide primers used for site-directed mutagenesis reactions  Appendix Table 1  Components of AtT20 transfected cell lines  91  328  LIST OF ABBREVIATIONS APC  antigen-presenting cell  ASK1  apoptosis-signal regulating kinase 1  ATP  adenosine triphosphate  AtT20  mouse pituitary gland tumor cell line  Bam32  B cell adaptor molecule of 32 kDa  BCA  bicinchoninic acid  BCAP  B cell adapter for PI 3-kinase  BCR  B cell antigen receptor  bFGF  basic fibroblast growth factor  BH  break point cluster region homology domain  BLNK  B cell linker adapter protein  BME  2-P-mercaptoethanol  BSA  bovine serum albumin  Btk  Bruton's tyrosine kinase  CAMs  cell adhesion molecules  CD  cluster of differentiation  CD16  FcyRIII (low affinity receptor for IgG)  cm  centimeter  CMV  cytomegalovirus  CSF-1  macrophage colony stimulating factor-1  Csk  C-terminal Src kinase  DABCO  1,4-diazabicyclo(2.2.2)octane  DAG  diacylglycerol  DIGs  detergent-insoluble glycolipid-enriched membranes  DMEM  Dulbecco's modified Eagle medium  DMSO  dimethylsulfoxide  DNA  deoxyribonucleic acid  dNTPs  deoxynucleotide triphosphates  DOC  deoxycholate  D-PBS  Dulbecco's phosphate-buffered saline  DRMs  detergent-resistant membranes  DTT  dithiothreitol  ECL  enhanced chemiluminescence  ;  EDTA  (ethylenedinitrilo)tetraacetic acid  EGF  epidermal growth factor  EGFP  enhanced green fluorescent protein  ER  endoplasmic reticulum  ERK  extracellular regulated kinases  FasL  Fas ligand  FceRI  high affinity receptor for IgE  FcyRIIB  low affinity receptor for IgG  FCS  fetal calf serum  FGF  fibroblast growth factor  FH  Forkhead  g G-CSF  grams granulocyte colony stimulating factor  Gab  Grb2-associated binding protein  GDP  guanosine diphosphate  GEMs  glycolipid-enriched membranes  GFP  green fluorescent protein  GSK-3p  glycogen synthase kinase-3p  GST  glutathione S-transferase  GTP  guanosine triphosphate  HIV  human immunodeficiency virus  HPK1  hematopoietic progenitor kinase 1  HRP  horseradish peroxidase  IB  immunoblot  IFN  interferon  Ig IgG  immunoglobulin immunoglobulin G  IgM  immunoglobulin M  IKK  IK kinase  IL  interleukin  ILK  integrin-linked kinase  IPs IPTG  inositol-1,4,5-triphosphate  IRS-1  insulin receptor substrate-1  ITAM  immunoreceptor tyrosine-based activation motif  ITIM  immunoreceptor tyrosine-based inhibitory motif  isopropy 1-1 -thio- P-galactopyranoside  JIP  JNK-interacting protein  JNK  c-Jun NH2-terminal kinase  KD  kinase dead  kDa  kilodaltons  L  liters  LAB  linker for activation of B cells  LAT  linker for activation of T cells  LB  Luria-Bertani  LPS  lipopolysaccharide  LY294002  Eli Lilly (Company) 294002  pCi  microCurie  Hg pM  micrograms  MAP kinase  mitogen-activated protein kinase  MES  2-(N-morpholino)ethanesulfonic acid  mlg  membrane immunoglobulin  mg  milligrams  MHC  major histocompatibility  mlg  membrane immunoglobulin  ml  milliliter  MLK  mixed lineage kinase  mM  millimolar  MMAC1  mutations in multiple cancers 1  MSCV  murine stem cell virus  NF-ATC  nuclear factor of activated T cells  NGF  nerve growth factor  OD  optical density  PAGE  polyacrylamide gel electrophoresis  PAMPs  pathogen-associated molecular patterns  PBS  phosphate-buffered saline  PCR  polymerase chain reaction  PDGF  platelet-derived growth factor  PDK  phosphoinositide-dependent kinase  PH  Pleckstrin homology  APH  deleted Pleckstrin homology  PI  phosphatidylinositol  micromolar  PIR  paired immunoglobulin-like receptor  PKA  protein kinase A  PKC  protein kinase C  PLC  phospholipase C  PMSF  phenylmethylsulfonyl fluoride  PRK-2  PKC-related kinase-2  PRS  proline-rich sequence  PTB  phosphotyrosine-binding  PTK(s)  protein tyrosine kinase(s)  Raftlin  Raft linking protein  RNA  ribonucleic acid  rpm  revolutions per minute  RPMI  Roswell Park Memorial Institute  RSV  Rous sarcoma virus  SAP kinase  stress-activated protein kinase  SDS  sodium dodecyl sulfate  SDS-PAGE  sodium dodecyl sulfate-polyacrylamide gel electrophoresis  SH2  Src homology 2 domain  SH3  Src homology 3 domain  SHIP  SH2 domain-containing inositol 5-phosphatase  SHP  SH2 domain-containing protein tyrosine phosphatase  SOS  son-of-sevenless  SR  sarcoplasmic reticulum  TBS  Tris-buffered saline  TCA  trichloroacetic acid  TCR  T cell antigen receptor  TEP1  TGF-P-regulating and epithelial-cell enriched phosphatase 1  TLRs  Toll-like receptors  TNF  tumor necrosis factor  WEHI  Walter and Eliza Hall Institute  ACKNOWLEDGMENTS I would like to thank several people who have made working in the lab more enjoyable. I thank May Dang-Lawson for being an great technician. Thank you for all the help with experiments, for sharing your secrets to being more efficient in the lab and mostly for your friendship throughout the years. I thank Elisa Vicencio for all the hard work with the Lck project.  I thank Teresa Jackson for answering all my molecular biology  questions and for all the helpful scientific as well as non-scientific discussions. I thank Dr. Mike Gold for his help throughout the years and for the numerous reference letters. Most of my thanks go to my supervisor, Dr. Linda Matsuuchi. I would not have made it through graduate school without her support, guidance and patience from the very beginning. Mostly, I thank her for believing in my abilities as a researcher. I thank my parents and sister for their love, understanding and support. Finally, I would like to thank my husband, Jared Lopes, for all the love and encouragement, especially at the end of this journey through graduate school.  CHAPTER 1  Introduction  1.1  The importance and role of lymphocytes in the immune response  The immune systems of organisms are constantly faced with the challenge of preventing the occurrence of diseases due to continual exposure to bacteria, fungi and viruses in the environment. The immune system plays a critical role in fighting off these various pathogens. In the absence of a properly functioning immune system, organisms would be unable to destroy and eliminate invading microorganisms from their system. Without the help of the immune system's cells, organisms would undoubtedly be unable to survive infectious attacks by these foreign pathogens.  The immune system is divided into two main categories: the innate immune system and the adaptive or acquired immune system. The innate immune system is a non-specific defense mechanism against infection that utilizes immune cells such as macrophages, dendritic cells and natural killer cells. The macrophages and dendritic cells engulf the foreign microorganisms resulting in their elimination from the organism. Alternatively, the foreign microorganisms can specifically interact with pattern recognition receptors present on macrophages and dendritic cells (reviewed by Akira et al., 2001; Gordon, 2002; Underhill and^ Ozinsky, 2002; Heine and Lien, 2003).  The main pattern  recognition receptors in mammals are the Toll-like receptors (TLRs) which recognize highly conserved molecular patterns called p_athogen-associated molecular patterns (PAMPs) found in many microorganisms (Medzhitov and Janeway, 1997; Hoffmann et al., 1999; reviewed by Aderem and Ulevitch, 2000'; Akira et al., 2001; Underhill and Ozinsky, 2002; Heine and Lien, 2003). TLRs can recognize PAMPs found in bacterial components including lipopolysaccharide (LPS), lipoproteins, glycolipids, flagellin and  bacterial deoxyribonucleic acid (DNA) containing unmethylated CpG dinucleotides (reviewed by Akira et al., 2001).  The interaction with these components allows  macrophages and dendritic cells to phagocytose the microorganism. However, the TLRs can also be activated at the cell surface, leading to the activation of signaling cascades within the macrophages and dendritic cells. TLR activation results in the production of interferon (IFN)-y, which promotes the activation of more macrophages and dendritic cells that can eliminate the pathogens by phagocytosis. TLR activation can also initiate the activation of a caspase cascade that leads to the proteolytic destruction of the foreign microorganisms. Finally, activation of TLRs can results in the production of cytokines, which promote the activation and expansion of T lymphocytes.  These activated T  lymphocytes can aid in the activation of B lymphocytes (reviewed by Akira et al., 2001; Heine and Lien, 2003). In this latter case, the TLRs act as a link between the innate and adaptive immune responses.  The adaptive immune response, on the other hand, is a highly specific mechanism that involves the production of antibodies and cell-mediated immunity. The adaptive immune response is specific and has memory, which enables animals to develop long-lasting immunity to re-infection with the exact same pathogen (Arpin et al., 1995; reviewed by Sprent, 1994).  For instance, if a person can recover from an illness caused by an  infectious microorganism, then any subsequent exposures to the same foreign pathogen would not result in the same intensity of the disease. This is due to the production of memory immune cells following the initial encounter (reviewed by Sprent, 1994). These memory cells are activated when their receptors recognize the pathogen. The activation leads to their subsequent activation and differentiation into effector cells that help eradicate the foreign pathogen.  The main cell types involved in the adaptive immune response belong to a class of white blood cells referred to as lymphocytes, in particular, B and T lymphocytes.  These  specialized lymphoid cells are found predominantly in the blood, lymph and in lymphoid organs such as the thymus, lymph nodes, spleen and appendix. In mammals, as opposed to chickens, B lymphocytes mature in the bone marrow and are responsible for the  production of antibodies. T lymphocytes, however, mature in the thymus and can take on one of two fates. They can develop into effector T cells and aid in the destruction of other cells which have been infected with intracellular pathogens or they can develop into helper T cells which help in the activation of other immune systems cells such as B lymphocytes, macrophages or dendritic cells (Clark and Ledbetter, 1994).  The B and T lymphocytes work in collaboration to produce the appropriate immune response against specific disease-causing agents. When an animal encounters a foreign agent, the initial event of the immune response involves the engulfment or phagocytosis of the agent by various antigen presenting cells (APC).  Such APC's include  macrophages, dendritic cells and B lymphocytes. Macrophages and dendritic cells engulf the foreign agent via phagocytosis, while B lymphocytes utilize receptor-mediated endocytosis.  Once the foreign agent is inside the APC, it undergoes proteolytic  degradation in a specialized intracellular compartment and peptides derived from it are eventually presented at the cell surface. These peptide fragments are presented on major histocompatibility (MHC) molecules to two subsets of T lymphocytes. The receptors on the T lymphocytes only recognize the foreign peptides in the context of MHC. There are two outcomes from this interaction between T lymphocytes and APC's depending on which subsets of T lymphocytes recognize the MHC-peptide complex on specific types of APC's.  If, for example, a cytotoxic killer T cell interacts with an infected  macrophage, the T cell can become activated and eventually kills the macrophage. On the other hand, if a helper T lymphocyte recognizes the MHC-peptide complex on the surface of a B lymphocyte, the helper T lymphocyte would release interleukins, which are growth-promoting factors (reviewed by Parker, 1993). Interleukins promote the proliferation and differentiation of B cells into antibody-secreting plasma cells that secrete large amounts of soluble antibodies. Thus, it is evident that dramatically different outcomes can be produced depending on the different cell types involved. In all cases, however, these outcomes protect the organism from severe illness and death.  1.2  B lymphocyte development  As mentioned earlier, one of the main cell types involved in the specific adaptive immune response are B lymphocytes. Like all other blood cells including erythrocytes, platelets and T lymphocytes, B lymphocytes are derived from one common pluripotent hematopoietic stem cell which is located in the bone marrow. This hematopoietic stem cell initially gives rise to two different specialized cell types, a common lymphoid progenitor and a myeloid progenitor. The myeloid progenitor eventually gives rise to the polymorphonuclear leukocytes, erythrocytes and platelets. The lymphoid progenitor, however, differentiates to produce the B and T lymphocytes.  In order to produce differentiated mature B lymphocytes, the hematopoietic stem cell must undergo a series of changes. The production of mature B lymphocytes from the hematopoietic stem cell generally involves four stages of development (Figure 1.1) (reviewed by Rolink et al., 1999; Meffre et al, 2000; Melchers et al., 2000). The first stage of development involves the differentiation of the hematopoietic stem cell into pro-B cells which do not express a complete B cell antigen receptor (BCR) on its cell surface. Some examples of these cells, however, were found to express Ig-P monomers and Ig-a/Ig-p heterodimers at the cell surface without the membrane immunoglobulins (mlg) (Koyama et al., 1997). This surface expression is facilitated by the association of the Ig-a and Ig-P chains with calnexin (Nagata et al., 1997). Calnexin is a resident endoplasmic reticulum (ER) chaperone protein that is responsible for ensuring that proteins passing through the ER are properly folded and assembled before traveling through the rest of the secretory pathway. Therefore, it was surprising to discover that calnexin was associated with the Ig-a and Ig-P chains at the plasma membrane. The function of the Ig-P monomers and Ig-a/Ig-P heterodimers at the cell surface remains unclear. It is hypothesized that these 'receptors' play a role in signal transduction since several signaling molecules were found to be tyrosine phosphorylated following Ig-p cross-linking (Nagata et al., 1997).  Figure 1.1: Schematic representation of B lymphocyte development. The various BCR(s) expressed in the different stages of development are indicated. The p heavy chains (green), 8 heavy chains (orange), light chains (gray), Ig-a (blue) and Ig-P (purple) accessory chains are shown.  Hematopoietic stem cell  pro-B cell lg-a/lg-p  pre-B cell  calnexin  heterodimers^^  lg-p monomer  pre-BCR  Secreted Igs  (mlgM)  Plasma cell Figure 1.1  Mature B lymphocyte  Immature B lymphocyte  The next stage immediately following the pro-B cell stage, the pre-B cell stage, is characterized by the expression of the pre-BCR which consists of a p heavy chain and two immunoglobulin (Ig) like proteins  (VpreB  and X 5 ) that make up the surrogate light  chain (Figure 1.1) (Tsubata and Reth, 1990; Minegishi et al., 1999a). This surrogate light chain is linked to the p heavy chain by disulfide bonds. Also non-covalently associated with the p heavy and the surrogate light chain complex are the Ig-a and Ig-P accessory proteins (Kashiwamura et al., 1990; van Noesel et al., 1991; Clark et al., 1992a). The pre-BCR complex is capable of the induction of a signaling cascade following receptor cross-linking (Kuwahara et al., 1996; Guo et al., 2000; Kato et al., 2000; Hess et al., 2001). These signals initiated by the pre-BCR inform the cell to stop p heavy chain gene rearrangement and initiate light chain gene rearrangements (Miyazaki et al., 1999; reviewed by Neuberger, 1997; Benschop and Cambier, 1999; Rolink et al, 1999; Rolink et al., 2001).  The production of light chains characterizes the next step in the  development of B lymphocytes, the immature B lymphocyte stage. lymphocytes express functional mlgM on their cell surface.  Immature B  The main difference  between the pre-BCR and the immature BCR is the presence of the light chain as opposed to the surrogate VpreB and X$ light chains.  All of these developmental stages from hematopoietic stem cell to the immature B lymphocyte occur in the bone marrow. The immature B lymphocytes produced then circulate throughout the bone marrow where they encounter self peptides. If the BCR on the immature B lymphocyte specifically recognizes a self-antigen, one of two outcomes may occur. The first possibility is that the immature B lymphocytes will undergo a type of programmed cell death referred to as apoptosis, resulting in clonal deletion. The other possible outcome is that the cell may become anergic. Anergy is defined as a state of unresponsiveness. In other words, the immature B lymphocytes will not respond even if they encounter the same self antigen at a later time. These two processes, clonal deletion and anergy, result in self tolerance by preventing immature B lymphocytes whose receptors recognize self antigens from differentiating into mature B lymphocytes (reviewed by Rolink et al, 2001). If the self-reactive immature B lymphocytes are not deleted in the bone marrow, then severe destruction of host tissues and organs will occur,  leading to the occurrence of autoimmune diseases (reviewed by LeBien, 1998; LeBien, 2000; Cariappa and Pillai, 2002)  If the BCR on an immature B lymphocyte does not recognize a self-antigen, then the cell can proceed through the developmental pathway and differentiate into a mature B lymphocyte (reviewed by Pillai, 1999; Cariappa and Piliiai, 2002).  In such cases,  expression of mlgD along with mlgM is up-regulated on the cell surface of these cells. These mature cells then migrate from the bone marrow to the periphery where they may encounter foreign antigens that would enable them to differentiate into antibody-secreting plasma cells with the help of T helper lymphocytes. These differentiated plasma cells are no longer responsive to activation by antigens or by T helper lymphocytes.  These  terminally-differentiated B lymphocytes live for approximately two weeks during which time they produce a large amount of soluble antibodies that can circulate throughout the entire body where they may bind to and inactivate foreign pathogens (Slifka et al., 1998). This results in the inability of the foreign pathogens to bind to host cells, thereby blocking their ability to enter and infect the cells. The bound antibodies on the foreign pathogens can also target them for destruction by phagocytosis or complement-mediated lysis. Therefore, the production of antibodies by differentiated B lymphocytes results in the elimination of the foreign pathogen from the organism.  1.3  The B cell antigen receptor  The development of hematopoietic stem cells into antibody-secreting plasma cells are triggered by the signals which emanate from the BCR located at the cell surface of pre-B, immature and mature B lymphocytes (reviewed by Desiderio, 1994; Gauld et al., 2002). The BCR expressed on the surface of immature and mature B lymphocytes has been the focus of much research for several years. This receptor is an oligomeric complex consisting of four different proteins: the heavy chain, light chain, Ig-a and Ig-P (reviewed by DeFranco, 1993b; Reth, 1994). The mlg consists of two heavy' and two light chains that are linked by disulfide bonds (Figure 1.2). Light chains are 25 kDa proteins that do  not contain any sequences specifying hydrophobic transmembrane domains. Only one of two different light chains, A, or K, can be expressed on the surface of any given cell. The P heavy chain, however, is a 78 kDa transmembrane glycoprotein.  One light chain  associates with one heavy chain by a disulfide linkage. This heavy-light chain complex then associates with another one, through disulfide bonds, to form the mlg (Figure 1.2). There are five different mlg classes or isotypes, defined by which heavy chain gene they express (in brackets): IgA (a), IgD (5), IgE (e), IgG (y) and IgM (p). All of these isotypes can exist as soluble or membrane-bound forms. The IgM isotype is found on the surface of immature B lymphocytes whereas the IgM and IgD isotypes are expressed on mature B lymphocytes. The mlgE, mlgG and mlgA isotypes are found on memory B lymphocytes.  Two accessory proteins are associated with the mlg at the cell surface (Hombach et al., 1988; Campbell and Cambier, 1990; Hombach et al., 1990a). These accessory proteins, Ig-cc and Ig-P, are transmembrane N-glycosylated polypeptides encoded by the mb-1 and B29 genes, respectively (Sakaguchi et al., 1988; Hombach et al., 1990b; Campbell et al., 1991; Matsuo et al., 1991; Clark et al., 1992a; Flaswinkel and Reth, 1992; Ha et al., 1992; Ishihara et al., 1992; Lankester et al., 1994). Both these genes contain sequences encoding one hydrophobic transmembrane domain, one extracellular Ig-like domain and a long cytoplasmic domain (Yu and Chang, 1992). The cytoplasmic domain of murine Ig-a is 61 amino acids long while that of murine Ig-P is 48 amino acids in length. The Ig-a and Ig-P molecules found at the cell surface of B lymphocytes are non-covalently associated with the mlg complex. These two accessory proteins form heterodimers using a disulfide bond (Figure 1.2). Originally it was thought that two Ig-a/Ig-P heterodimers associated with one mlg. However, Schamel and Reth (2000) have shown that only one Ig-a/Ig-P heterodimer is associated with the mlg at the cell surface. The mlgM and mlgD expressed on the surface associate with the same Ig-a and Ig-p molecules, but they are differentially glycosylated (Campbell et al., 1991 and Gold et al., 1991).  Figure 1.2: Schematic representation of the BCR complex. The p heavy chains (green), light chains {yellow), Ig-a (blue) and Ig-(3 (purple) accessory chains are shown. Each BCR is composed of an antigen-binding membrane Ig (mlg) which is bound to one Ig-a/Ig-P heterodimer that is involved in signal transduction. The IT AM motifs (red) on the Ig-a and Ig-P accessory chains are also indicated.  Membrane Igantigen recognition  plasma membrane  — ITAM  lg-a/lg-[3 heterodimersignal transduction  Figure 1.2  The Ig-a and Ig-P molecules are involved in many cellular processes that occur during the lifetime of a B lymphocyte.  These two proteins are important for cell surface  expression of a functional BCR. The Ig-a and Ig-P proteins were sufficient to promote cell surface expression of a transfected BCR in a non-lymphoid cell line (Venkitaraman et al., 1991; Matsuuchi et al., 1992). Transfection of the mb-1 gene alone was unable to promote surface expression of the receptor, however, transfection of both the mb-1 and B29 genes were able to do so. In the absence of Ig-a and Ig-P, the heavy and light chains of the mlg complex are retained in the ER, associated with ER retention proteins. The ER retention proteins are responsible for preventing incomplete or improperly assembled complexes and improperly folded proteins from escaping the ER and traveling through the secretory pathway to the plasma membrane. Association of the Ig-a/Ig-P heterodimer displaces the ER retention protein bound to the heavy-light chain complex, thereby forming a functional BCR which can travel to the cell surface (Melnick et al., 1992; Melnick et al., 1994; reviewed by Melnick and Argon, 1995).  Another function of Ig-a, but not Ig-p chain involves BCR internalization (Cassard et al., 1998). When the Ig-a chain lacks certain portions of their cytoplasmic domain, the BCR expressed at the cell surface could not be internalized. Examination of many chimeric molecules containing different segments of Ig-a and Ig-P revealed that only Ig-a contained the 'internalization signal,' which allowed the BCR complex to become internalized by endocytosis. However, Ig-a and Ig-p are both required for subsequent antigen presentation (Siemasko et al., 1999).  Finally, the Ig-a and Ig-P molecules play a major role in BCR-mediated signal transduction. The cytoplasmic carboxy terminal domain of the murine p heavy chain is relatively short, it consists of only three amino acids, K-V-K (single letter amino acid code).  This short sequence is insufficient to mediate the binding of the different  signaling molecules. For a long time, it remained a mystery as to how the mlg could signal since its cytoplasmic tail was so short. It was later discovered that the Ig-a and Ig-P proteins were responsible for relaying the signals received by the BCR on the  outside of the cell to the inside of the cell (Clark et al, 1992b; Sanchez et al., 1993; reviewed by DeFranco, 1993b; Reth, 1995).  Within the cytoplasmic domains of Ig-a and Ig-P, there is a conserved 26 amino acid sequence referred to as the immunoreceptor tyrosine-based activation motif (ITAM) (Reth, 1989). The consensus sequence of the ITAM of Ig-a and Ig-P is D/E-X 7 -D/E-YX3-L-X7-Y-X2-L/I, where X refers to any amino acid. This ITAM motif is also found in the cytoplasmic tails of other receptors such as the high affinity FceRI receptor and the T cell antigen receptor (Wegener et al., 1992; Samelson and Klausner, 1992). This domain plays a key role in the signal transduction mediated by the BCR. It has been shown by many independent research groups that the cytoplasmic tails of the Ig-a and Ig-P proteins are capable of interacting with different effector molecules involved in signaling including the Syk protein tyrosine kinase (PTK) (Hutchcroft et al., 1991; Hutchcroft et al., 1992; Yamada et al., 1993; Rowley et al., 1995) and several Src family members (Lin and Justement 1992; Clark et al, 1994; Pleiman et al., 1994a; reviewed by Justement,  2000).  1.4  Membrane proximal events in BCR-mediated signal transduction  As mentioned earlier, the Ig-a and Ig-P accessory proteins play a crucial role in the activation of the signaling pathways downstream of the BCR.  These two proteins  actually relay the information received at the receptor outside the cell to the cell's interior. How does activation of the BCR complex at the cell surface result in altered gene expression in the nucleus? The answer to this has been derived from many years of research conducted by several different groups.  Signal transduction in B lymphocytes is initiated when the BCRs at the cell surface are cross-linked by foreign antigens or by anti-Ig antibodies under experimental conditions. This receptor aggregation or clustering at the cell surface results in a protein tyrosine cascade within the cells (Gold et al., 1990; Campbell and Sefton 1990; Law et al., 1992).  PTKs play a major role in the initiation of the signaling cascades from the BCR. Protein kinases can be grouped into three categories, tyrosine kinases, serine/threonine kinases and dual specificity kinases.  PTKs are a class of enzymes that catalyze the  phosphorylation of tyrosine residues while serine/threonine kinases catalyze the phosphorylation of serine or threonine residues on proteins. Dual specificity kinases phosphorylate tyrosine and serine/threonine residues. The source of the high energy phosphate group that is transferred by the kinase is derived from adenosine triphosphate (ATP).  Approximately 10% of all total cellular proteins are capable of becoming  phosphorylated. Of this 10%, approximately 90% are phosphorylated on serine while about 9% are phosphorylated on threonine. The amount of tyrosine phosphorylation that occurs within cells is less than the amount of serine or threonine phosphorylation. Less than 0.1% of total cellular protein becomes tyrosine phosphorylated. This small amount, however, is biologically important. Tyrosine phosphorylation plays a critical role in the initial steps of B cell signal transduction and in signaling by many receptors including growth factor receptor, and immune receptors.  BCR cross-linking results in the  activation of PTKs. Serine/threonine kinases are also activated by the BCR and are also important in B lymphocyte signaling, but these kinases are not involved in the membrane proximal signaling following BCR cross-linking.  B lymphocyte activation involves cross-linking of many BCR complexes at the cell surface. There are two proposed models of how the BCRs aggregate upon encountering the specific antigen. Originally, it was thought that BCRs existed as individual receptors at the cell surface. These individual receptors could form aggregates upon encountering the appropriate antigen recognized by the BCR. However, Reth's group has proposed another model (reviewed by Reth et al., 2000; Matsuuchi and Gold, 2001). This newer model proposes that BCRs exist as pre-formed aggregates at the cell surface that are capable of maintaining low levels of basal signaling (Schamel and Reth, 2000). However, cross-linking of the BCRs results in the formation of even larger receptor complexes through the clustering of the pre-formed complexes. The aggregation of the BCRs at the cell surface results in a tyrosine phosphorylation cascade within the cell and activation of several signaling pathways. The activation of the signaling pathways  ultimately leads to changes in gene transcription that results in phenotypic changes within the cells. The purpose of receptor aggregation is to bring the components, including PTKs, in close proximity to each other, which results in their activation.  Some of these components that are activated by BCR cross-linking are the Syk and Src PTKs, which are very important for signaling through the BCR (Takata and Kurosaki, 1995). In resting, unactivated B lymphocytes, some Src kinases constitutively associate with the unphosphorylated Ig-a/Ig-P heterodimer in a phospho-tyrosine-independent manner (Pleiman et al., 1994a). Upon receptor aggregation, the Src kinases become activated by autophosphorylation and phosphorylate tyrosine residues within the ITAM motifs of Ig-a and Ig-P creating docking sites for the Src homology 2 (SH2) domaincontaining proteins (Burkhardt et al, 1991; Gold et al, 1991; Campbell and Sefton, 1992; Li et al., 1992).  SH2 domains interact with specific sequences on proteins  containing phosphorylated tyrosine residues (Songyang et al., 1993; Songyang et al., 1994; reviewed by Pawson, 1995; Pawson et al., 2001). The phosphorylation of these ITAM motifs allows for the recruitment and activation of more Src family kinases (reviewed by Yamamoto et al., 1993). In addition, Syk molecules are also recruited to the phosphorylated ITAM sequences of Ig-a and Ig-P (Hutchcroft et al., 1991; Rowley et al,  1995; Poa and Cambier, 1997).  Syk is an SH2 domain-containing PTK found  predominantly in B lymphocytes that is crucial for B lymphocyte development (Cheng et al., 1995; reviewed by Turner et al., 2000).  The recruitment of Syk to the BCR  complexes results in its tyrosine phosphorylation and activation by Src family kinases (Kurosaki et al., 1994; Zoller et al., 1997). Association of Syk with the phosphorylated ITAM sequences is also required for Syk activation (Pao et al., 1998). Perhaps the binding of the SH2 domains of Syk with the phosphorylated ITAM sequences results in a conformational change that is necessary for Syk activation to occur. These activated Syk molecules can further enhance the tyrosine phosphorylation of the tyrosine residues with in the ITAM motifs of the Ig-a and Ig-p accessory chains resulting in the generation of a positive feedback loop, which can amplify signaling through the BCR (Rolli et al., 2002). In addition, the activated Syk molecules can phosphorylate numerous downstream signaling effector molecules involved in the activation of the phosphatidylmositol 3-  kinase (PI 3-kinase), mitogen-activated protein kinase (MAP kinase) and phospho lipase C (PLC)-y pathways (reviewed by Weiss and Littman, 1994; Bolen, 1995; Chan and Shaw, 1995; Gold and Matsuuchi, 1995; Reth and Wienands, 1997; Gold, 2000; Gold et  al., 2000).  1.5  Signaling pathways regulated by the B cell antigen receptor  The tyrosine cascade induced by BCR cross-linking results in the activation of four major signaling pathways mediated by the following enzymes: PI 3-kinase/Akt, PLC-y, MAP kinase and stress-activated protein kinase (SAP kinase). While this study focuses mainly on the activation of the PI 3-kinase/Akt pathway, brief overviews of MAP kinase and PLC-y pathway activation through the BCR will be provided  1.5.1  PLC-y pathway  BCR signaling leads to the tyrosine phosphorylation and activation of both PLC-y isoforms, PLC-yl and PLC-y2 (Bijsterbosch et al., 1985; Fahey and DeFranco, 1987; Carter et al., 1991; Coggeshall et al., 1992; Hempel et al., 1992; Roifman and Wang, 1992; Siderenko et al., 1995; DeBell et al., 1999). PLC-y2 was shown to be essential for B lymphocyte development (Hashimoto et al., 2000; Wang et al., 2000).  PLC-y  signaling through the BCR is a multi-step process that requires tyrosine phosphorylation of PLC-y and recruitment to the plasma membrane by adapter proteins where it is in close proximity to other effector molecules (reviewed by Rudd, 1999; Tsukada and Kurosaki, 2000). The BCR-associated PTKs, Syk and the Src family kinases, are important for PLC-y pathway activation through the BCR (Takata et al., 1994). Following BCR crosslinking and activation of the Src family and Syk PTKs, an adapter protein for PLC-y, B cell linker (BLNK; also referred to as BASH and SLP-65) becomes inducibly tyrosine phosphorylated (Fu et al., 1998; Goitsuka et al., 1998; Wienands et al., 1998). BLNK is a cytosolic adapter protein that is expressed only in B lymphocytes throughout all stages of B lymphocyte development (reviewed by Campbell, 1999; Leo and Schraven, 2001).  BLNK, along with the recently discovered transmembrane adapter protein, linker for activation of B cells (LAB), are thought to serve the same function as the SLP-76 and linker for activation of T cells (LAT) PLC-y adapter proteins found in T lymphocytes (Pivniouk et al, 1998; Boerth et al., 2000; Yoder et al., 2001; Janssen et al., 2003; reviewed by Jordan et al., 2003).  Like SLP-76, BLNK is required for normal  development and activation of B lymphocytes (Jumaa et al., 1999; Minegishi et al., 1999b; Pappu et al., 1999; Hayashi et al., 2000; Xu et al., 2000; Flemming et al., 2003). The phosphorylation of BLNK by Syk on multiple tyrosine residues results in the generation of docking sites for PLC-y as well as other signaling molecules including the Grb2 adapter protein, the Vav guanine nucleotide exchange factor, the Nek adapter protein, the Bruton's tyrosine kinase (Btk), Syk and the hematopoietic progenitor kinase I (HPK1) (Figure 1.3) (Fu et al, 1998, Hashimoto et al., 1999). Once in close proximity, Btk and Syk phosphorylate PLC-y on tyrosine residues that results in an increase in its enzyme activity (Takata and Kurosaki, 1996; Fluckiger et al., 1998; Ishiai et al., 1999). Since the substrates for PLC-y are located at the plasma membrane, its translocation to the plasma membrane is essential for its function within the cell. BLNK performs this function in B lymphocytes, however, the mechanism by which BLNK is recruited to the plasma membrane is not completely understood.  It does not contain any acyl  modifications that could anchor it to the plasma membrane and it does not have a Pleckstrin homology (PH) domain that binds to phospholipids within the plasma membrane (reviewed by Lemmon et al., 1997; Fruman et al., 1999a; Vanhaesebroeck et al., 2001). It has been recently proposed that LAB is responsible for the recruitment of BLNK to the plasma membrane (Janssen et al., 2003).  LAB is a transmembrane  molecule that contains nine tyrosine residues within its cytoplasmic domain, five of which are within the Grb2 binding motif. BLNK has not been shown to directly interact with LAB. It is thought that BLNK associates with LAB through the Grb2 adapter protein (Figure 1.3). Alternatively, the tyrosine kinase Btk may be responsible for BLNK plasma membrane localization. The PH domain of Btk is responsible for its recruitment to the plasma membrane since this domain can mediate the interaction with membraneassociated inositol phospholipids (Li et al., 1997; Varnai et al, 1999). Therefore, Btk may not only phosphorylate and activate PLC-y when co-localized on BLNK, but it could  Figure 1.3:  BCR cross-linking results in the activation of the PLC-y pathway.  Briefly, BLNK, LAB and Bam32 recruit PLC-y to the plasma membrane resulting in the breakdown of PI-3,4-P2 to IP3 and DAG.  Generation of IP3 leads to an increase in  intracellular Ca 2 + from intracellular stores, which results in the activation of the calcineurin phosphatase. Activated calcineurin molecules de-phosphorylate NF-AT C allowing it to enter the nucleus and initiate transcription. The DAG aids in the activation of some PKC isoforms.  PKC's are responsible for the activation of MAP kinases and  SAP kinases, JNK and p38, which results in changes in the transcription of certain genes in the nucleus. The red circles represent tyrosine phosphorylated residues. Several other proteins including Btk, Nek, Vav, Grb2 and HPK1 also associate with BLNK.  Antigen  Antigen  [Antigen PLASMA MEMBRANE  BLNK  Bam32  Other effectors (i.e. Vav. Nek, Btk, Syk, HPK1) ERK1/2 mitochondria vo  calcinuerin  JNK1/2  calmodulin  Transcription  nucleus  also aid in the translocation of the whole complex to the plasma membrane. Another possibility is that BLNK recruitment to the plasma membrane may be mediated by its interaction with the Ig-a accessory chain of the BCR. BLNK can associate with Ig-a, however, whether this association involves the ITAM sequence or sequences outside the ITAM remains controversial (Engels et al., 2001; Kabak et al., 2002).  The recruitment of PLC-y to the plasma membrane following BCR activation can also be performed by the B cell adaptor molecule of 32 kDa (Bam32) adapter protein (Marshall et al., 2000). The SH2 domain of Bam32 interacts with PLC-y2, and to a lesser extent PLC-yl in B lymphocytes. Unlike BLNK, Bam32 contains a PH domain that can bind inositol phospholipids present within the plasma membrane and thus recruit PLC-y. The recruitment of Bam32 to the plasma membrane requires its PH domain and the lipid products generated by PI 3-kinase (Marshall et al., 2000).  Bam32-deficient B  lymphocytes show decreased Ca 2+ flux following BCR activation compared to wild type cells, indicating that Bam32 plays a role in PLC-y activation in B lymphocytes (Niiro et al., 2002).  The recruitment of PLC-y to the plasma membrane by either BLNK/LAB or Bam32 results in the hydrolysis of PI-4,5-P2 leading to the production of diacylglycerol (DAG) and the inositol triphosphate (IP3) (Bihsterbosch et al., 1985; Fahey and DeFranco, 1987; Rhee et al., 1989; reviewed by Berridge, 1993). IP3 binds to IP3 receptors, which are present on organelles within the cell including the ER, mitochondria and sarcoplasmic reticulum (SR) (Figure 1.3). The binding of IP3 to the receptors allows them to open and Ca 2 + is then released from these intracellular stores resulting in the generation of a Ca 2+ flux within the cell (Sugawara et al., 1997). The Ca 2+ ions associate with, for example, calmodulin, a calcium-dependent kinase, resulting in its activation.  The activated  calmodulin then phosphorylates and activates the serine/threonine phosphatase calcineurin.  Once activated, calcineurin de-phosphorylates the nuclear factor for  activated T cells (NF-ATC) transcription factor, allowing it to translocate into the nucleus and regulate gene expression.  DAG, on the other hand, activates several of the  conventional and novel protein kinase C (PKC) isoforms (Figure 1.3) (reviewed by  Jaken, 1996; Newton, 1997; Mellor and Parker, 1998). PKC can also act as an upstream activator of the MAP kinases extracellular-regulated kinase (ERK) and SAP kinases cJun NFL-terminal kinases (JNK) and p38 that also results in increased transcription (Figure 1.4) (Hashimoto et al., 1998; Krappmann et al., 2001; reviewed by Musashi et al., 2000). Thus, activation of the various signaling enzymes by both PLC-y products, IP3 and DAG, leads to changes in gene transcription within the nucleus.  1.5.2  MAP kinase and SAP kinase pathways  Another pathway activated by BCR cross-linking is the ERK kinase pathway (briefly summarized in Figure 1.4, right side of diagram) (Casillas et al., 1991; Gold et al., 1992b; Sutherland et al., 1996). BCR engagement results in the tyrosine phosphorylation of the She adapter protein that then allows it to associate with the Grb2 adapter protein (Saxton et al., 1994). The guanine nucleotide exchange factor son-of-sevenless (SOS) then interacts with Grb2. This complex is recruited to the plasma membrane through the association of She with the Ig-a/Ig-P subunit of the BCR. Once at the plasma membrane, SOS activates p21Ras by exchanging the guanosine diphosphate (GDP) for guanosine triphosphate (GTP) (reviewed by Boguski and McCormick, 1993; Marshall, 1996). p21Ras is a low molecular weight GTPase that acts as a "molecular switch." In the inactive state, it is associated with GDP while in the activated state, it is bound to GTP. Several groups have shown that p21Ras can be activated by BCR engagement (Harwood and Cambier, 1993; Lazarus et al., 1993; Saxton et al., 1994; Tordai et al., 1994). The GTP-bound activated p21Ras can phosphorylate Rafl (Cook and McCormick, 1993; reviewed by Schaeffer and Weber, 1999). Rafl is a MAP kinase kinase kinase that can activate the MEK1 and MEK2 dual specificity MAP kinase kinases (Figure 1.4, right side of the diagram). Activated MEK1 and MEK2 then specifically phosphorylate the ERK1 and ERK2 MAP kinases resulting in their activation (reviewed by Cobb and Goldsmith, 1995; Su and Karin, 1996). Once activated, ERK1 and ERK2 can translocate into the nucleus and regulate gene expression by phosphorylating multiple transcription factors.  Figure 1.4: The MAP kinase and SAP kinase pathways are activated in response to BCR cross-linking. Activation of the MAP kinases, ERK1/ERK2, is through the BCR is initiated by the formation of the Shc/Grb2/SOS complex (right side of diagram). This complex recruits SOS to the plasma membrane where it can activate the GTPase p21Ras. The activated p21Ras phosphorylates and activates Rafl, which ultimately leads to the activation of ERK1/ERK2. Some PKC isoforms can also activate the ERK1/2 through MEK. Activation of the SAP kinase, JNK1/JNK2 and p38, is thought to involve the BLNK adapter protein (left side of diagram). BLNK can associate with HPK1 and the Vav guanine nucleotide exchange factor that are both involved in SAP kinase activation. Vav can activate Racl by exchanging its bound GDP for GTP.  This GTP-bound  activated Racl then activates the SAP kinase kinase, MEKK, which then activates JNK1/JNK2 and p38 kinases. Alternatively, HPK1 can phosphorylate and activate the SAP kinase kinase kinase, MLK, which can also activate MEKK leading to the activation of the JNK1/JNK2 and p38 kinases. Activation of the JNK1/JNK2 and p38 kinases results in altered gene transcription within the nucleus.  The SAP kinases are also activated in response to BCR cross-linking (Sutherland et al., 1996; Hashimoto et al., 1998; Jiang et al., 1998). Unlike ERK MAP kinases, the SAP kinases p38 and JNK1/2, are not activated through the p21Ras pathway downstream of the BCR. There are two ways to activate the JNK and p38 kinases through the BCR. One of these involves the BLNK adapter protein, described earlier. This adapter protein appears to be important for activation of both the PLC-y and SAP kinase pathways following BCR cross-linking. BLNK can interact with HPK1 (Fu et al., 1998). HPK1 can phosphorylate and activate the mixed lineage kinase (MLK), which is a MAP kinase kinase kinase (Hu et al., 1996; Kiefer et al., 1996; Wang et al., 1997). MLK then activates the MAP kinase kinase MEKK, leading the activation of the JNK and p38 (Figure 1.4, left side of the diagram). The other way in which the SAP kinase pathway can be activated is through the Vav guanine nucleotide exchange factor. Vav is a protooncogene which is required for signaling in B lymphocytes and is activated by the BCR (Gulbins et al., 1994; Tarakhovsky et al., 1995; Zhang et al., 1995; Bachmann et al., 1999; Doody et al., 2001; Tedford et al., 2001). Although Vav can also interact with BLNK, it not known if this association is required for its recruitment to the plasma membrane or Vav-mediated activation of the JNK and p38 kinases. Vav can be recruited to the plasma membrane through the interaction of its PH domain with membraneassociated phospholipids. Once at the plasma membrane, Vav activates the Racl GTPase in the same manner that SOS activates p21Ras (Crespo et al., 1997; Ishiai et al., 1999; Sauer et al., 2001). The activated Racl then phosphorylates and activates the MAP kinase kinase, MEKK, ultimately resulting in the activation of JNK and p38 kinases (Figure 1.4, left side of diagram). Activation of the SAP kinases results in increased gene transcription by the c-Jun and c-fos transcription factors (reviewed by Davis, 2000).  1.5.3  PI3K/Akt pathway  Another pathway activated by the BCR that is required for B lymphocyte development and function is the PI 3-kinase pathway (Fruman et al., 1999b; Suzuki et al., 1999; Clayton et al., 2002; reviewed by Gold et al., 1999; Koyasu, 2003; Okkenhaug and Vanhaesebroeck, 2003). PI 3-kinase is a lipid kinase that is composed of two subunits, the catalytic subunit and the non-catalytic regulatory adapter subunit (reviewed by Hunter, 1995; Vanhaesebroek et al., 1997; Vanhaesebroek et al., 2001). Three classes of PI 3-kinase exist but signaling through the BCR activates class Ia PI 3-kinases. There are three class Ia mammalian catalytic subunits (pi 10a, pi 10(3 and pi 105) and five mammalian regulatory subunits (p85a, p85P, p55a, p55y and p50a) (reviewed by Hunter, 1995; Vanhaesebroek et al., 1997; Wymann and Pirola, 1998; Funaki et al., 2000; Okkenhaug and Vanhaesebroeck, 2003). The pi 10 catalytic subunit consists of a carboxy terminal lipid kinase domain and a binding site for the p85 subunit (Figure 1.5 A). The p85 regulatory subunit contains many protein interaction domains including an SH3, a break point cluster region homology domain (BH), two SH2 domains and two proline-rich sequences (PRS) (Figure 1.5 A). The pi 10 catalytic subunit can interact with the p85, p55 or p50 regulatory subunits forming a complete PI 3-kinase molecule, however, signaling through the BCR activates the class Ia pi 10/p85 PI 3-kinases (Gold et al., 1992a). Class I A PI 3-kinases can phosphorylate PI-4-P and PI-4,5-P2 at the 3' position of the inositol ring resulting in the production of PI-3,4-P2 and PI-3,4,5-P3 (reviewed by Carpenter and Cantley, 1996). The levels of these phospholipids were found to increase following BCR cross-linking (Gold and Aebersold, 1994).  Like PLC-y, PI 3-kinase activation also requires the aid of adapter proteins. PI 3-kinase adapter proteins in B lymphocytes include the transmembrane molecule CD 19 (Tuveson et al., 1993; Buhl et al., 1997; Buhl and Cambier, 1999), the B cell adapter for PI 3kinase (BCAP) (Okada et al., 2000; Inabe and Kurosaki, 2002; Yamazaki et al., 2002) and the Grb2-associated binding protein 1 (Gabl) (Ingham et al., 1998; Ingham et al., 2001). The docking proteins allow for the recruitment of PI 3-kinase to the plasma  Figure 1.5: Schematic representation of the PI 3-kinase subunits, PDK1 and Akt structures. (A) Structure of the pi 10 catalytic subunit and p85 regulatory subunits of PI 3-kinase. The pi 10 catalytic subunit consists of the carboxy terminal lipid kinase domain and amino terminal regulatory binding. The p85 regulatory subunit consists of an SH3 domain, two SH2 domains, BCR homology domain (BH), two proline-rich sequences and a pi 10 binding domain. (B) Structure of PDK1 and Akt. PDK1 contains a carboxy terminal PH domain and a kinase domain. Akt consists of an amino terminal PH domain and kinase domain. The threonine 308 residue is located within the kinase domain and the serine 473 residue is located within the regulatory tail. The approximate molecular weight of the proteins (in kDa) is indicated.  (A) Class l A PI 3-kinase subunits p110a/|3 catalytic subunit Regulatory domain  Kinase domain  p85a/p  p110  PRS  SH3  binding site  PRS  BH  •  l  SH2  SH2  (B) PDK1 and Akt PDK1 (Molecular weight = 68 kDa) PH  Kinase domain  Regulatory tail  Akta/Aktp/Akty (Molecular weight = 60 kDa) PH  Kinase domain  7308  S473  Figure 1.5  membrane where its substrates are PI-4-P and PI-3,4-P2 are located. BCR cross-linking leads to the phosphorylation of these adapter proteins on several tyrosine residues, creating binding sites for signaling molecules including PI 3-kinase. The SH2 domains of the p85 regulatory subunit interact with the tyrosine phosphorylated residues on the adapter proteins. Once at the plasma membrane, PI 3-kinase phosphorylates its substrates leading to the generation of PI-3,4-P2 and PI-3,4,5-P3, which bind to PH domaincontaining molecules and allow them to be recruited to the plasma membrane (Figure 1.6, top part of diagram, associated with the plasma membrane) (reviewed by Corvera and Czech, 1998; Leevers et al., 1999; Rameh and Cantlyey, 1999). The PH domains directly interact with the inositol phospholipids within the plasma membrane (Rameh et al., 1997; reviewed by Lemmon et al, 1997; Fruman et al, 1999a). Some well-known examples of PH domain-containing PI 3-kinase target molecules are phosphoinositidedependent kinases 1 and 2 (PDK1 and 2) (Alessi et al., 1997a; Alessi et al., 1997b) and Akt, which is the main focus of this study (Bugering and Coffer, 1995; Stokoe et al., 1997). Akt was the first target of PI 3-kinase identified (Franke et al., 1997; reviewed by Downward, 1995; Hemmings, 1997; Marte and Downward, 1997; Aoki et al, 1998; Coffer et al, 1998; Brazil and Hemmings, 2001).  Once recruited to the plasma membrane and brought into close proximity with one another, PDK1 and PDK2 regulate Akt activation by phosphorylating it on two residues (reviewed in Cohen et al., 1997; Downward, 1998a; Downward, 1998b; Vanhaesebroek and Alessi, 2000). PDK1 phosphorylates the threonine 308 residue of Akt and the ability of PDK1 to phosphorylate this residue requires its carboxy terminal PH domain (Figure 1.5 B) (Alessi et al., 1997a; Stokoe et al, 1997; Anderson et al, 1998; Filippa et al, 2000; Wick et al., 2000). Phosphorylation of another residue is also required for full activation of Akt, the serine 473 residue. The serine/threonine kinase responsible for this phosphorylation has remained elusive. One potential candidate for this kinase, though controversial, is the integrin-Unked kinase (ILK) (reviewed by Dedhar, 2000). It is possible that ILK may be able to phosphorylate Akt on the serine 473 residue since inhibition of ILK has been shown to result in the suppression of Akt activation (Delcommenne et al, 1998; Persad et al, 2000). However, it is also thought that ILK acts as adapter protein that facilitates the phosphorylation of Akt on the serine 473  residue. Some investigators have postulated that PDK1 is also PDK2. PDK1 has been shown to phosphorylate the serine 473 residue when it was associated with the PKCrelated kinase-2 (PRK-2) (Balendran et al., 1999). Other groups, however, believe that PDK2 is a novel kinase that is sensitive to the immunosuppressive antibiotic, rapamycin (Ziegler et al., 1999). Akt (also referred to as protein kinase B and RAC-PK) was identified by three independent groups as a retroviral oncogenic serine/threonine kinase and as a kinase that was homologous to both protein kinase A (PKA) and PKC (Bellacosa et al., 1991; Coffer and Woodgett, 1991; Jones et al., 1991). There are three human and mouse isoforms of Akt (a, (3 and y) which share a high degree of sequence homology (approximately 80%) and are encoded by three separate genes (reviewed by Coffer et al., 1998; Downward, 1998b). All three isoforms have the same general structure, which consists of an amino terminal PH domain, a glycine-rich region, a serine/threonine kinase domain and a regulatory tail (Figure 1.5 B). Akt can be activated by several growth factor receptors (Alessi et al., 1996; Dudek et al., 1997) as well as immune receptors including the T cell antigen receptor (TCR) (Genot et al., 2000) and the BCR (Astoul et al., 1999; Craxton et al., 1999; Gold et al., 1999; Li et al., 1999; reviewed by Gold et al., 2000). Akt has been shown to be very important in regulating the cell survival of many cell types including B lymphocytes (Pogue et al., 2000). Many cellular targets of Akt have been identified (Figure 1.6, lower part of the figure) (reviewed by Vanhaesebroek and Alessi, 2000). Below is a discussion of the important targets of Akt (as diagrammed in Figure 1.6). One of these targets is glycogen synthase kinase-3(3 (GSK-3(3) (Figure 1.6, pathway 1) (Hajduch et al., 1998; van Weeren et al., 1998). Phosphorylation of GSK-3(3 by Akt inactivates it thereby preventing it from associating with and phosphorylating the (3-catenin transcription factor (Cross et al., 1995; Pap and Cooper, 1998).  The  phosphorylation of (3-catenin targets it for ubiquitin-mediated degradation (Figure 1.6, pathway 1). Therefore, phosphorylation and inactivation of GSK-3(3 prevents (3-catenin degradation and allows it to translocate into the nucleus and induce transcription of genes that promote cell survival (Rubinfeld et al., 1996; Monick et al., 2001). Akt can also phosphorylate BAD on two serine residues (serine 112 and 136) and inhibit its activity.  BAD is a pro-apoptotic protein that is a member of the Bcl-2 gene family. When BAD is not phosphorylated, it can interact with Bcl-2 and Bcl-XL which are anti-apoptotic members of the Bcl-2 gene family (Figure 1.6, pathway 2). When complexed with BAD, Bcl-2 and Bcl-XL cannot perform their anti-apoptotic functions resulting in apoptosis (Zha et al, 1996; Datta et al, 1997; del Peso et al, 1997). Therefore, inhibition of BAD through phosphorylation by Akt allows Bcl-2 and Bcl-XL to prevent apoptosis and promote cell survival. The transcription factor, N F - K B is also regulated by Akt (Scott et al, 1998; Kane et al, 1999; Ozes et al, 1999; Romashkova and Makarov, 1999). Akt can phosphorylate and activate IK kinase (IKK) which can subsequently phosphorylate IKB  molecules that are complexed with N F - K B (Figure 1.6, pathway 3).  The  phosphorylated IKB molecules are targeted for degradation resulting in the release of NFKB. NF-KB  molecules can then enter the nucleus and up-regulate transcription of several  pro-survival genes.  Another target of Akt is caspase 9 (Figure 1.6, pathway 4).  Phosphorylation of this caspase family member prevents it from participating in the proteolytic cascade that leads to apoptosis (Cardone et al., 1998). Akt is also involved in the regulation of the SAP kinase, JNK (Figure 1.6, pathway 5). An increase in JNK activity is observed when Akt signaling is suppressed, indicating that Akt is a negative regulator of this SAP kinase pathway (Suhara et al., 2002). Akt phosphorylates and inhibits the activity of apoptosis-signal regulating kinase 1 (ASK1) and MLK3, both of which are SAP kinase kinase kinases (Kim et al, 2001; Barthwal et al., 2003). ASK1 and MLK3 are upstream regulators of JNK, which when activated leads to apoptosis. Therefore, negative regulation of the JNK pathway by Akt prevents apoptosis and promotes cell survival. Finally, other transcription factors regulated by Akt that promote cell survival are the Forkhead family members (FKH) (Brunet et al., 1999; Kops et al, 1999). Phosphorylation of Forkhead transcription factors causes them to be retained in the cytoplasm where they cannot initiate transcription of pro-apoptotic proteins (Figure 1.6, pathway 6) (Biggs et al, 1999; Guo et al, 1999; Tang et al, 1999). One known target of the Forkhead transcription factors is the Fas ligand (FasL). Binding of FasL to its receptor, Fas, results in activation of the apoptotic pathway culminating in cell death (reviewed by Zimmermann and Green, 2001; Kim, 2002). Therefore, inhibition of the Forkhead transcription factors by Akt prevents the production of FasL protein and apoptosis (Suhara et al, 2002).  Figure 1.6: BCR cross-linking results in the activation of the PI 3-kinase pathway. The tyrosine phosphorylation of an adapter protein following BCR cross-linking leads to the recruitment of PI 3-kinase to the plasma membrane. Once recruited to the plasma membrane, PI 3-kinase phosphorylates PI-4,5-P2 to produce PI-3,4,5-P3 which results in the recruitment of PDK1, the putative PDK2 and Akt, allowing for the activation of Akt by phosphorylation. Akt regulates the activity of multiple proteins that control cell survival (refer to the text for more details). These proteins include GSK-3P, BAD, IKK, caspase 9, ASK1, and Forkhead transcription factors (FH). Other PI 3-kinase targets include a few novel and atypical PKC isoforms which are regulated by the PI-3,4,5-P3 lipid product.  Besides activating Akt, PDK1 can also regulate the atypical and  conventional PKC isoforms. The activation of some of these PKC molecules leads to the activation of the SAP kinases, JNK1/JNK2 and p38. The red circles represent tyrosine phosphorylated residues.  PLASMA MEMBRANE  > PIP  PIP3 PIP3  PIP  PDK1  PDK2  PIP, • •• 3  PIP3 PKC£  7  ~PKC  o/pl/pll/y  I  ©  5/e/n  u> K>  ©  ASK1  ©  MLK3  erki/2 [ JNK1/2]  I JNK1/2  [fkh]  FKH  Apoptosis  \ — > \  nucleus  f  T FasL transcription  Transcription  Besides regulating the activation of the Akt pathway, PI 3-kinase has also been implicated in the regulation of PKC isoforms including novel isoforms, PKC8, PKCe, PKCr) and the atypical PKC£ (Parekh et al, 1999; Popoff and Deans, 1999; Parekh et al, 2000; Ting et al, 2002). These PKC isoforms are activated by the PI 3-kinase products PI-3,4-P2 and PI-3,4,5-P3 (Nakanishi et al, 1993; Toker et al, 1994). However, PKC£ can also be activated by PDK1 (Chou et al, 1998; Le Good et al, 1998). In addition, PDK1 can also phosphorylate the conventional PKC isoforms, PKCa, PKCpI, PKCpII and PKCy (Good et al, 1998). Although phosphorylation by PDK1 does not lead to the activation of these PKC isoforms, it is a necessary step for their activation which also requires the PLC-y product, DAG (Dutil et al, 1998). Once activated, the PKC enzymes regulate the activation of MAP kinases and SAP kinases (JNK and p38), which ultimately leads to increased transcription within the nucleus (Figure 1.6, pathway 7).  The negative regulation of the PI 3-kinase pathway involves the lipid phosphatases PTEN (also referred to as MMAC1 (mutations in multiple cancers 1) and TEP1 (TGF-Pregulating and epithelial-cell enriched phosphatase I)) and the SH2-domain-containing inositol phosphatases (SHIPs) (Plutzky et al, 1992; Li et al, 1997; Li and Sun, 1997; Myers et al, 1998; Brauweiler et al, 2000a). These lipid phosphatases have both been shown to inhibit Akt activation (Aman et al, 1998; Stambolic et al, 1998; Wu et al, 1998; Sun et al, 1999; reviewed by Cantley and Neel, 1999). PTEN hydrolyses the 3phosphate from inositol phospholipids converting PI-3,4-P2 to PI-4-P and PI-3,4,5-P3 to PI-4,5-P2 (Maehuma and Dixon, 1998; reviewed by Cantley and Neel, 1999). Akt activation is decreased due to the decrease in amount of PI-3,4,5-P3 and number of binding sites within the plasma membrane (Haas-Kogan et al, 1998). Other negative regulator of the lipid signaling pathway in B lymphocytes are the SHIP family of lipid phosphatases (SHIP1 and SHIP2) (reviewed by Brauweiler et al, 2000b; Krystal, 2000; March and Ravichandran, 2002).  These lipid phosphatases also decrease the  phospholipids required for the recruitment of PH domain-containing proteins to the plasma membrane. However, the SHIP phosphatases de-phosphorylate the 5-phosphate of PI-3,4,5-P3 and produce PI-3,4-P2, the substrate for PLC-y (Pesesse et al, Muraille et al., 1999).  1998;  1.6  Protein tyrosine kinases involved in proximal signaling events through the B cell antigen receptor  PTKs are responsible for initiating the signaling that occurs downstream of the BCR immediately after receptor aggregation at the cell surface. These PTKs are at the top of a chain of many biochemical reactions that culminate in the alteration of cellular phenotype. However, it is unclear as to how different combinations of kinases are able to specifically activate certain signal transduction pathways while failing to activate others. One hypothesis to explain this differential regulation is that certain kinases may be able to interact specifically with certain signaling components through their special proteinprotein interaction domains.  Alternatively, the kinases may be compartmentalized  differentially within the cell, only allowing access to specific subsets of enzymes. As mentioned earlier, two main classes of PTKs are involved in membrane proximal signaling from the BCR. These PTKs are the Syk and the Src family kinases (reviewed by Courtneidge et al, 1993; Chan and Shaw, 1995; Justement, 2000; Turner et al., 2000).  1.6.1  Syk PTK  Syk is a 72 kDa non-receptor PTK that contains two tandem SH2 domains (Figure 1.7 A) (Zioncheck et al., 1986; Zioncheck et al., 1988; Taniguchi et al., 1991). Syk does not have any acylation modification and is therefore unable to directly interact with the plasma membrane. This kinase, which is predominantly expressed in B lymphocytes (Hutchcroft et al, 1991; Hutchcroft et al, 1992), is a member of the ZAP-70/Syk family of PTKs. The other member of this family, ZAP-70, is expressed in T lymphocytes and is one of the key PTKs involved in signaling by the TCR. Syk associates with the Ig-ot/Ig-P heterodimer of the BCR in B lymphocytes following receptor cross-linking (Hutchcroft et al., 1992; Rowley et al., 1995). Specifically, the SH2 domains of Syk interact with tyrosine phosphorylated ITAM sequences of the Ig-a and Ig-P accessory proteins (Law et al,  1993; Johnson et al, 1995; Jugloff and Jongstra-Bilen, 1997).  Experiments involving the Syk knockout cells indicate that this PTK is essential for B lymphocyte development since it is required for signaling through pre-BCR (Cheng et al., 1995). The lack of Syk molecules prevents pre-B cells from maturing into mature B lymphocytes.  1.6.2  Src kinase family tyrosine kinases  Members of the Src family of PTKs participate in a variety of signal transduction pathways involved in signaling from many different receptors including growth factor receptors, cytokine receptors and adhesion molecules (reviewed by Parsons and Parsons, 1997; Thomas and Brugge, 1997; Tatosyan and Mizenina, 2000; Schlessinger, 2000). For example, some Src family members have been shown to take part in signaling through receptor protein tyrosine kinases including the platelet-derived growth factor (PDGF) receptor, epidermal growth factor (EGF) receptor, basic fibroblast growth factor (bFGF) receptor, insulin receptor, nerve growth factor (NGF) receptor, fibroblast urowth factor (FGF) receptor and macrophage colony stimulating factor-1 (CSF-1) receptor (reviewed by Erpel and Courtneidge, 1995; Abram and Courtneidge, 2000; Abram and Courtneidge, 2000). Src kinase family members also play a role in signaling pathways activated by immune receptors including the BCR, TCR and Fc receptors for Ig's (reviewed by Corey and Anderson, 1999; Korade-Mirnics and Corely, 2000). A few Src kinases have been implicated in integrin signaling (reviewed by Erpel and Courtneidge, 1995). Integrins are a family of transmembrane receptors involved in cell to matrix and cell to cell contact. Other cell adhesion molecules that also utilize Src kinases in their signaling pathways are cadherins, selectins and cell adhesion molecules (CAMs). Gprotein coupled seven transmembrane receptors have also been shown to use Src family members in signaling.  These include the receptors for thrombin, angiotensin II,  bradykinin and vasopressin. Finally, many of the Src family members are involved in signaling through cytokine receptors including those for granulocyte colony stimulating factor (G-CSF), tumor necrosis factor (TNF), erythropoietin, prolactin and many  Figure 1.7: Schematic representation of the primary structures of tyrosine kinases involved in the proximal signaling downstream of the BCR. (A) Structure of Syk and (B) Src kinases are shown. The SH2 and SH3 protein interaction domains and the kinases domains are indicated. The two regulatory tyrosine residues on the Src kinases are also indicated. The family members Src and Blk do not contain the cysteine residue at position 3. "Myr" refers to the myristylation site; "palm" refers to the palmitylation site on the Src kinases.  (A) S y k (Molecular weight = 72 kDa)  SH2  SH2  Kinase domain  u>  (B) Src kinases  (Molecular weight = 53-59 kDa)  myr palm Unique  SH3  SH2  Kinase domain  (?)  (?)  Y  Y  interleukin receptors (IL-2, IL-3, IL-4, IL-5, IL-6, IL-7, IL-11, IL-12 and IL-15) (reviewed by Thomas and Brugge, 1997). Src family kinases play major roles in cell death, survival, proliferation, differentiation, migration and adhesion since they are involved in signaling through a large repertoire of cell surface receptors.  At the present time, there are eleven known Src family members: Src, Fyn, Yes, Fgr, Lyn, Hck, Blk, Lck, Yrk, Frk/Rak and Iyk/Bsk (Figure 1.7 and 1.8) (reviewed by Thomas and Brugge, 1997; Corey and Anderson, 1999; Korade-Mirnics and Corey, 2000; Tatosyan and Mizenina, 2000).  Three of the Src members, Src, Fyn and Yes, are  expressed in a wide range of cell types and tissues. In contrast, the five other family members (Lyn, Hck, Blk, Lck and Fgr) are quite limited in their cellular expression patterns. These five family members are found primarily in cells of hematopoietic lineage. However, Lck and Lyn are also found in neuronal cells. The Frk/Rak and Iyk/Bsk belong to a subgroup of Src kinases and share homology in all regions with the other Src kinases (Cance et al., 1994; Lee et al., 1994; Oberg-Welsh and Welsh, 1995; Thuveson et al., 1995; Chandrasekharan et al., 2002). These two kinases are expressed primarily in epithelial cells. The final member of this family, Yrk, is found only in chickens where it is expressed in nearly all cell types.  Figure 1.8: Phylogenetic tree of the Src kinase family members. The tree of the murine  Src  kinases  was  (http://www.ebi.ac.uk/clustalw/).  constructed  using  the  ClustalW  program  Note: The family member Yrk is only expressed  chicken cells and was not included in the phylogenetic tree. The sequences for Frk/Rak and Iyk/Bsk family members were not available and could not be included in the tree.  Fyn  Lyn Hck Lck Blk Src Yes Fgr  1.6.3  Src kinase structure and activation  The general structure of all Src family members is highly conserved (Figure 1.7 B). Src family members all contain one SH2 and one Src homology 3 (SH3) domain. These domains were first identified in Src kinases. The SH2 domains are approximately 100 amino acids in length and mediate protein-protein interactions with phospho-tyrosine containing motifs present in a variety of other signaling molecules.  Although the  phosphotyrosine residue on the SH2 ligand is critical for mediating this protein interaction, the three to five amino acids following the phosphorylated residue are important for the binding specificity for the individual SH2 domains to the target protein. The SH2 domains of different Src family kinases have been shown to, have different binding specificities to their target proteins. These domains on Src kinases preferentially bind to distinct sets of proteins that can be mutated to change the binding specificity (Malek and Desiderio, 1993; reviewed by Pawson, 1995).  The differences in the  specificities of the SH2 domains of the various Src kinases may explain the differential activation of the downstream signal transduction pathways. SH3 domains of proteins are responsible for the binding to proteins containing proline-rich sequences that have a P-XX-P consensus sequence (Ren et al., 1993; Yu et al., 1994; Rickles et al., 1995). Aside from forming intermolecular interactions, the SH2 and SH3 domains of the Src kinases have also been implicated in the maintenance of the inactive form of the kinases (reviewed by Williams et al., 1998).  The catalytic or kinase domain is located at the carboxy end of the Src kinases (Figure 1.7 B). This catalytic domain contains an important tyrosine residue that is a positive regulatory autophosphorylation site. This tyrosine residue must become phosphorylated in order to activate the Src kinases. Like most tyrosine kinase domains, the catalytic domain of Src kinases contain the motif required for ATP-binding (G-X-G-X2-G). Also contained within the catalytic domain is a lysine residue which, when altered, renders the kinase constitutively inactive (Hanks et al., 1988). A negative regulatory tyrosine residue is located outside the kinase domain at the carboxy terminal end of the kinase (Figure 1.7 B). This tyrosine residue, unlike the one within the catalytic domain, is dephosphorylated in the active state and must be phosphorylated to inactivate the kinase. The final domain  is the unique region that is located at the very amino terminal end of the Src kinases. This region is distinct for each member of the family. There is no homology within this region between all family members. It is thought that this domain is important for mediating interactions between the individual kinases and the different receptors or signaling components (Shaw et al., 1990; Turner et al., 1990; Pleiman et al., 1994; Gervais and Veillette, 1995). This unique region also contains two important amino acid residues that are required for lipid modifications. The glycine residue at position 2 is important for myristylation. The consequence of this particular lipid modification is association with the inner layer of the plasma membrane (van't Hof and Resh, 1997). Therefore, since all Src family members possess this glycine residue at position 2, they are all found anchored to the plasma membrane (reviewed by Resh, 1994; Resh, 1999). Another lipid modification site exists in the unique region at position 3. The cysteine residue at this site is required for palmitylation. The palmitylation modification may allow the Src kinases to be localized to specialized domains within the plasma membrane called lipid rafts, which may have an effect on their ability to differentially regulate downstream pathways (Alland et al., 1994; Koegl et al., 1994; Sigal et al., 1994). However, palmitylation is not sufficient to promote targeting of proteins to lipid rafts. Lipid raft localization is dependent on other factors such as interactions with other proteins and additional lipid modifications.  Not all Src family members have this  cysteine residue at position 3 and thus are not subject to this type of lipid modification. Src and Blk are the only two Src family members that do not have cysteine residues at position 3 and therefore are not palmitylated. All the other family members are subject to both myristylation and palmitylation modifications (Alland et al, 1994; Koegl et al., 1994; Sigal et al, 1994; reviewed by Resh, 1994; Resh, 1999).  All Src kinase family members are regulated in a similar manner and the mechanism of this activation has been inferred from previous investigations on the Src and Hck family members (Sicheri et al, 1997; reviewed by Thomas and Brugge, 1997; Cooper and Howell, 1993; Mayer, 1997). Since Src kinases have the potential to cause uncontrolled cell growth when activated, their regulation from the inactive to the active states must be under stringent control within the cells. The inactive form is maintained by specific intramolecular interactions involving several regions of the kinase.  The postulated inactive form of Src kinases is inferred from the crystal structure of the Hck family member (Sicheri et al, 1997; Moarefi et al., 1997). In the inactive state, the SH2 domain of the Src kinases interacts with the negative regulatory tyrosine residue at its carboxy-terminus which is phosphorylated (Figure 1.9 A). This interaction enables the kinases to bend.  The SH3 domain also provides an additional intramolecular  interaction. The linker region between the SH2 and catalytic domains contains two proline residues which form a left-handed polyproline type II helix allowing it to interact with the SH3 domain (Arold et al., 2001). The intramolecular interactions mediated by the SH2 and SH3 domains are, by themselves, both weak interactions. If only one of these interactions were solely responsible for maintaining an inactivated kinase, spontaneous activation would occur frequently.  However, the two intramolecular  interactions together provide enough extra stability to maintain the kinase in the inactive conformation. Since the positive regulatory tyrosine residue within the kinase domain is still exposed, Src kinases can be rapidly activated by autophosphorylation at this site.  The phosphorylation state of the negative regulatory tyrosine residue at the carboxy terminal end of all Src family members is known to be regulated in part by the cytoplasmic tyrosine C-terminal Src kinase (Csk) (reviewed by Chow and Veillette, 1995). Csk phosphorylation of this residue results in the inhibition of Src kinase activity (Hata et al., 1994; Cooper et al, 1986; Okada and Nakagawa, 1989; Nada et al, 1991). The other regulator of this tyrosine residue in B lymphocytes is the receptor tyrosine phosphatase CD45 (Clark and Ledbetter, 1989; Justement et al, 1991; Brown et al, 1994; Yanagi et al, 1996; Pao and Cambier 1997; reviewed by Justement et al, 1994). In T lymphocytes, CD45 regulation of the Src family member Fyn and Lck has been well-defined.  The negative regulatory tyrosine residues of Fyn and Lck are de-  phosphorylated by CD45 in vitro (Mustelin et al, 1992; Cahir et al, 1993; Sieh et al, 1993).  In T lymphocytes, CD45 greatly affects the activity of both Fyn and Lck  (Ostergaard et al, 1989; Mustelin et al, 1989; Mustelin et al, 1990; Ostergaard et al, 1990).  Figure 1.9: Mechanism of Src kinase activation. (A) The inactive state of Src kinases is maintained by intramolecular interactions involving the SH2 and SH3 domains (refer to the text for details and to Sicheri et al. (1997) and Moarefi et al. (1997). (B) In the activated state, the negative regulatory tyrosine residue is unphosphorylated while the positive regulatory tyrosine residue is phosphorylated.  (A) Inactive state  (B) Active state  SH3  SH2 Phosphorylation by Csk Autophosphorylation kinase  Dephosphorylation by CD45  >  Y-  1.6.4  Src kinase function in B lymphocytes  Three of the eleven Src family members are found in normal B lymphocytes, Lyn, Fyn and Blk. The Src family kinase predominantly expressed in B lymphocytes is Lyn. Two isoforms of Lyn exist, p53 and p56, which are produced by alternative splicing as indicated by the presence of two different Lyn mRNA products (Yi et al,  1991;  Yamanashi et al., 1991b). The two Lyn products differ by 21 amino acids contained within the unique region of the kinase. Since this region is thought to play a role in mediating the interactions between the kinase and other proteins, then it is possible that the different isoforms of Lyn may interact differentially with certain receptors or signaling components. The ability of the two isoforms of Lyn to interact with the BCR were previously shown to differ (Yamanashi et al., 1991b). Although the two forms of the kinase may associate differently with certain receptors, the expression patterns of the p53 and p56 isoforms are identical in the different cells and cell lines examined. Lyn is predominantly expressed in B lymphocytes, but it is also found in myeloid and neuronal cells (Table 1.1). The Src family members Lck and Fyn play important roles in the initiation of signaling through the TCR, but are also expressed in some B lymphoma tissue culture cell lines (Table 1.1). Fyn is expressed in a wide variety of cells and tissues while the Fyn(T) isoform is found only in T lymphocytes. Lck, on the other hand, is expressed in T lymphocytes and neuronal cells, but not in normal splenic B lymphocytes. Lck is only expressed in B lymphoma tissue culture cell lines and in B lymphocytes from patients with leukemia or lymphomas (Von Knethen et al., 1997; Majolini et al., 1998). Normal splenic B lymphocytes do not express Lck, suggesting that Lck may contribute to the formation and maintenance of transformed B cell lymphomas and leukemias. The Src kinase Blk, is the only family member with very restricted expression. Blk (which stands for B lymphocyte kinase) is expressed only in B lymphocytes (Dymecki et al., 1990; Debrin et al., 1995; Lin et al, 1995).  Table 1.1: Expression patterns of the Src family members expressed in cells of the lymphocyte lineage.  Src family kinase  Cell types  Lyn  B lymphocytes, myeloid cells, neuronal cells  Blk  B lymphocytes  Lck  T lymphocytes, natural killer cells, B lymphoma cell lines, neuronal cells  Fyn  monocytes, granulocytes, low levels in some B lymphoma cell lines  Fyn(T)  T lymphocytes  The role of Src family members in BCR-mediated signal transduction has been wellstudied.  One of the initial events which occurs following BCR engagement is the  activation of Src family kinases including Lyn, Blk and Fyn (Burkhardt et al., 1991; Li et al., 1992; Law et al., 1992; Saouaf et al., 1994). These family members have been shown to interact with the Ig-a and Ig-P accessory chains of the BCR complex (Yamanashi et al., 1991a; Lin and Justement, 1992; Campbell and Sefton, 1992; Law et al., 1993; Burg et al., 1994; Clark et al., 1994). In vitro, it was observed through binding studies that Lyn and Fyn were able to associate with phosphorylated as well as unphosphorylated forms of the Ig-a protein (Pleiman et al., 1994a). This interaction between Lyn and Fyn with the unphosphorylated Ig-a chain indicates that low levels of Src kinases associate with the receptor complex at the plasma membrane in resting, unstimulated cells. Once the receptor is activated, however, the phosphorylation of the ITAM motifs of the Ig-a and Ig-P proteins results in the recruitment of more Src kinases to the receptor complex (Jugloff and Jongstra-Bilen, 1997). The recruitment of more Src kinase molecules to the BCR complexes subsequently results in their activation through autophosphorylation (Johnson et al., 1995; Burg et al., 1994). The phosphorylation of the ITAM sequences also enables Syk to be recruited to the BCR complexes and become activated (Johnson et al., 1995).  The association between Syk and the Src kinase Lyn has been observed in B lymphocytes (Siderenko et al., 1995). The interaction between these two kinases suggests that Lyn may be responsible for the activation of Syk.  Experiments performed using Lyn-  deficient and Syk-deficient DT40 chicken B lymphoma cell lines indicated that Syk can be activated by Lyn.  In Syk-deficient cells, tyrosine phosphorylation of Lyn was  unaffected by the lack of Syk (Takata and Kurosaki, 1995). Meanwhile, Lyn-negative cells showed reduced BCR-induced tyrosine phosphorylation and activation of Syk PTK (Kurosaki et al., 1994). These two pieces of information indicate that Lyn is responsible for the tyrosine phosphorylation and activation of Syk. Results from co-transfection experiments have also provided evidence to support this model. In co-transfected COS cells, Syk was phosphorylated and activated only when Lyn was also present in the cells (Zotter et al., 1997).  According to one model of BCR signaling proposed by Kurosaki (1997), once the Src kinase and Syk are activated, they subsequently activate the Tec family PTK, Btk. One study conducted by Rawlings et al. (1996), in which Lyn and Btk were co-transfected into a non-lymphoid cell line, indicated that only Lyn is required for Btk activation. The presence of Lyn resulted in a five to ten fold increase in the enzymatic activity of Btk. In contrast, no increase in Btk tyrosine phosphorylation or enzymatic activity was observed in cells co-transfected with Syk PTK. These results suggest that Lyn plays a role in the activation of Btk. Lyn can activate Btk by phosphorylating a tyrosine residue at position 551 within the catalytic domain. This initial phosphorylation event up-regulates Btk activity. This increase in activity then stimulates the autophosphorylation of a tyrosine residue at position 223 within the SH3 domain of the kinase (Wahl et al., 1997). This autophosphorylation induces a conformational change in Btk leading to its activation (Rawlings et al., 1996; Kurosaki, 1997). It is possible that other Src kinases can regulate Btk activation (Afar et al., 1996).  Some evidence also suggests that Syk can also  phosphorylate and activate Btk (Baba et al. , 2001).  1.6.5  Src kinase knockout studies  It has proved difficult to determine the precise function of each individual Src family kinases because of the redundancy and expression of multiple family members within the same cell. However, knockout mouse studies have provided some insight into Src kinase function (reviewed in Brandon et al., 1995a; Brandon et al., 1995b; Brandon et al., 1995c; Lowell and Soriano, 1996). Although Lyn, Fyn and Blk participate in multiple cellular functions in immune cells, single Src knockout mice are still viable and lymphocytes are still produced in these animals. B lymphocyte population numbers and function are normal in Fyn and Lck knockout mice indicating that these two Src kinases do not play unique roles in and are not required for B lymphocyte development and activation (Grant et al., 1992; Molina et al., 1992; Stein et al., 1992; Yagi et al., 1993). T lymphocyte function and proliferation, however, are impaired in the Fyn knockout mice. Interestingly, the Fyn Src double knockout mice are lethal suggesting that these two family members may have overlapping functions that cannot be performed by other family members (Stein et al., 1994). The Lck knockout mouse showed more severe phenotype compared to the Fyn knockout mice. No mature thymocytes were detected in the Lck knockout mice indicating that Lck is crucial for T lymphocyte development (Molina et al., 1992). Although Lyn is important for B lymphocyte function, B lymphocytes were still found in Lyn knockout mice. However, the populations of B lymphocytes at various stages of development and signaling through the BCR in these mice were different from those in wild type animals (Hibbs et al., 1995; Wang et al., 1996). Circulating autoreactive antibodies were also found in these knockout mice, indicating that Lyn is important for B lymphocyte development during the clonal deletion stage where self-reactive cells are deleted from the repertoire (Nishizumi et al., 1995). The mice also developed a form of systemic lupus due to the accumulation of antibody complexes in the kidneys. Since Lyn plays a role in signaling through the high affinity receptor for IgE (FceRI), it was not surprising that IgE cross-linking of the FceRI receptor in the Lyn-deficient mice failed to induce an allergic response (Vonakis et al., 1997). Finally, the Lyn knockout mice also showed defective mast cell function since Lyn is also expressed in cells of the myeloid  lineage. Surprisingly, Hck Fgr Lyn triple knockout mice have no additional phenotypic differences compared to single Lyn knockout mice (Meng and Lowell, 1997). Only minor problems with macrophage activation were observed in these mice. The knockout mouse studies have also provided evidence for the role of Lyn in signaling through the BCR (reviewed by DeFranco et al., 1998). One group reported that B lymphocytes from Lyn knockout mice showed enhanced and prolonged MAP kinase and PKC activation (Chan et al, 1997). This inhibitory effect of Lyn on BCR signaling may be due to the fact that Lyn plays a role in signaling through the CD22 BCR co-receptor (Schulte et al, 1992; Cornall et al, 1998; Smith et al, 1998; Fujimoto et al, 1999; Otipoby et al, 2001).  Alternatively, the Lyn-mediated negative regulation in B  lymphocytes may be due to Lyn's ability to participate in signaling through Fc receptors including the low affinity receptor for IgG (FcyRIIB), which attenuates signaling through the BCR (Takai et al, 1996; Nishizumi et al, 1998). Indeed, it was observed that FcyRIIB-mediated signaling was defective in the Lyn-deficient cells from the knockout mice (Chan et al, 1997; Wang et al, 1996).  The formation of B lymphomas was observed in transgenic mice expressing a constitutively activated form of Blk (Malek et al,  1998).  T lymphomas were also  generated when this constitutively active form of Blk was introduced into T lymphocytes. These results indicate that Blk may control proliferative responses in B lymphocytes. However, unlike the phenotype observed in the Lyn knockout mice, Blk knockout had no abnormalities compared to wild type mice (Texido et al, 2000). B and T lymphocyte development was normal in the Blk single knockout mice. Furthermore, both in vitro and in vivo B lymphocyte responses were also normal in Blk knockout cells. These results suggest that Blk does not have a unique function in B lymphocytes and that its function is compensated by the other Src family members present in these cells.  Mice with double and triple knockouts of the three major Src kinases expressed in B lymphocytes have been generated in an effort to understand the roles of these Src family kinases in B lymphocyte development and signaling. In the Blk Fyn double knockout mice, B lymphocyte development and responses were normal compared to wild type  animals (Texido et al., 2000).  The Fyn Lyn double knockout mice have the same  phenotypic abnormalities as the Lyn single knockout mice (Horikawa et al., 1999). However, proliferation in response to BCR cross-linking in B lymphocytes from these double knockout mice was significantly lower compared to the single Lyn knockout mice. Therefore, Fyn may play a role in the regulation of B lymphocyte proliferation through the BCR. Surprisingly, B lymphocytes are still produced in the Lyn Fyn Blk triple knockout mice. The combination of these three Src kinases is required early in B lymphocyte development as reduced numbers of pre-B cells were present in these animals (Saijo et al., 2003). The developmental block from the pro-B to pre-B cell stage of development was due to the lack of activation of the N F - K B transcription factor through the pre-BCR. Despite the large body of information that has been accumulated by studying Src kinase knockout mice, a few questions still remain unanswered. For example, the specific or unique role(s) of the individual Src kinases in B lymphocyte signaling is not known. In addition, the extent of redundancy and overlapping functions of the Src kinases has also remained elusive due to the presence of multiple family members within cells.  1.7  The Gabl adaptor protein  Some cytosolic signaling molecules require recruitment to specific cellular locations for their activation, as described in the activation of PLC-y and PI 3-kinase earlier. Adapter proteins serve the function of recruiting the components of these signaling pathways to the plasma membrane for activation. Adapter proteins are also important for bringing together different signaling components to form macromolecular complexes, which facilitates the activation of the associated molecules. Adapter and docking proteins that do not possess intrinsic catalytic activity are mainly composed of protein interaction domains or motifs. Some domains and motifs found on adapter proteins are SH2 domains, SH3 domains, PH domains, phosphotyrosine-binding (PTB) domains, WW domains and proline-rich sequences (Lu et al., 1999; reviewed by Cohen et al., 1995; Pawson, 1995). Adapter proteins are required for translocation of proteins to specific cellular locations, generally to  the plasma membrane, where their substrates are located or where they can become activated by other signaling components. Furthermore, adapter proteins have the ability to co-localize multiple signaling components enabling them to regulate each other's activity. This co-localization also allows for cross-talk between signaling molecules from different pathways within the cells. Signaling in B lymphocytes uses multiple adapter proteins for PI 3-kinase and PLC-y activation, as mentioned earlier (reviewed by Niiro and Clark, 2002; Gu and Neel, 2003). One adapter protein used in BCR signaling is Gabl. Gabl was first identified in fibroblasts and was found to associate with multiple SH2 domain-containing proteins (HolgadoMadruga et al, 1996). Gabl is a member of a family of PH domain-containing proteins that include the insulin receptor substrate-! (IRS-1), p62 d o k , Gab2 and Gab3 (Gu et al., 1998; Wolf et al., 2002; reviewed by Liu and Rohrschneider, 2002). Gabl has been shown to play a role in signaling downstream of several receptors (reviewed in Lui and Rohrschneider, 2002; Gu and Neel, 2003). These include receptors for growth factors including the insulin receptor; the EGF receptor (Holgado-Madruga, 1998; Yart et al., 2001); the NGF receptor (Holgado-Madruga et al., 1997); the cytokine receptors, IL-3, IL6, IFN-a and IFN-y (Takahashi-Tezuka et al., 1998); FGF receptor (Ong et al., 1998); the erythropoietin receptor (Lecoq-Lafon et al., 1999); and the c-Met receptor (Maroun et al., 1999a; Maroun et al., 1999b; Lock et al, 2002). Immune receptors also use Gab proteins for signaling. Gabl plays a role in BCR signaling (Ingham et al, 1999; Ingham et al., 2001) while Gab2 has been shown to participate in signaling through the TCR (Pratt et al, 2000; Yamasaki et al., 2001).  Gab proteins have an amino terminal PH domain, multiple proline-rich motifs and twelve potential tyrosine phosphorylation sites (Figure 1.10).  Receptor-induced tyrosine  phosphorylation of Gabl, results in the creation of docking sites for SH2 domaincontaining molecules including the She adapter protein, the p85 regulatory subunit of PI 3kinase and the SHP-2 protein tyrosine phosphatase (Holgado-Madruga et al., 1996). Although the tyrosine phosphorylation motifs in Gabl match the consensus sequences for the SH2 domains of PLC-y and the Grb2 adapter protein, the interaction between these two proteins with Gabl has not been shown.  Figure 1.10: Schematic representation of Gabl structure. Gabl contains an amino terminal PH domain, four proline-rich sequences (PRS) and multiple tyrosine residues (Y). The Y* indicates the SHP-2 binding site; the Y** indicate the potential binding sites for the p85 subunit of PI 3-kinase.  Gabl  (Molecular weight = 115 kDa)  Y  Y Y Y Y  Y  Y  Y  PH domain PRS PRS  PRS  Immune receptors such as the BCR, TCR and FceRI use Gab proteins to activate PI 3kinase since these receptors lack binding sites for this lipid kinase (Ingham et al, 1998; Pratt et al., 2000; Gu et al., 2001; Ingham et al., 2001). The phosphorylation of Gabl on its Y-X-X-M sequences provides docking sites for the p85 regulatory subunit of PI 3kinase, which can then be recruited to the plasma membrane where its substrates (PI-4P and PI-4,5-P2) are located. The PI 3-kinase-Gabl association is critical for PI 3-kinase activation as deletion of the PI 3-kinase binding sites on Gabl abrogate signaling through this pathway in B lymphocytes (Ingham et al., 2001).  Gabl can also regulate the activity of the SHP-2 phosphatase.  Its association with  phosphotyrosine-containing sequences on Gabl increases SHP-2 phosphatase activity (Cunnick et al., 2000). SHP-2 appears to play both a positive and negative role in signaling through Gabl. The interaction between SHP-2 and Gabl has been shown to be essential for ERK activation (Cunnick et al., 2000; Maroun et al., 2000; Rodrigues et al., 2000; Schaeper et al., 2000). Indeed growth factor-induced ERK activation is impaired in Gabldeficient fibroblast cells (Itoh et al., 2000; Yart et al., 2000). The ability of Gabl to activate ERK is due to the fact that the SHP-2 molecules associated with Gabl are upstream activators of the Ras GTPase (Cunnick et al., 2002). The activation of p21Ras is due to the ability of SHP-2 to associate with the Grb2/SOS complex (Figure 1.11) (Takahashi-Tezuka et al., 1998).  Gabl recruits the SHP-2/Grb2/SOS complex to the  plasma membrane where SOS can activate p21Ras. The activated p21Ras molecules can activate the MAP kinase pathway that culminates in ERK activation (Figure 1.4). The activation of the MAP kinase pathway through Gabl-associated SHP-2 is consistent with the findings that SHP-2 and PI 3-kinase are required for Ras-ERK activation (King et al., 1997; Sutor et al., 1999; Zhang et al., 2002). Gab2 has also been shown to play a role in ERK activation (Liu et al, 2001; Nishida et al, 2002).  Besides playing a role in ERK activation, SHP-2 may also be responsible for the negative regulation of Gabl signaling. SHP-2 has the ability to de-phosphorylate the tyrosine residues on the Gab proteins, Gabl (Cunnick et al., 2001), Gab2 (Gu et al, 1997) and Dos,  the Drosophila Gabl homolog (Herbst et al., 1996).  The de-phosphorylation of the  tyrosine residues on Gabl results in the dissociation of the signaling complexes and termination of signaling. Fibroblast cells containing an inactive mutant of SHP-2 showed increased PI 3-kinase-Gabl association and increased PI 3-kinase activation (Zhang et al., 2002).  Gabl is involved in the activation of the MAP kinase pathway through its association with the She adapter protein (Itoh et al., 2000). She, like SHP-2, can interact with Grb2-SOS complexes that can be recruited to the plasma membrane by Gabl and initiate the activation of the MAP kinase pathway through p21Ras. Interestingly, the Grb2 molecules associated with She can also interact with the SHIP lipid phosphatase (Ingham et al., 1998). SHIP phosphatases, as mentioned earlier, can de-phosphorylate PI 3-kinase products thereby decreasing the PI-3,4,5-P3 available at the plasma membrane (Figure 1.11). This results in decreased number of binding sites for PH domain-containing proteins including Gabl. Thus, the recruitment of SHIP to the plasma membrane provides another mechanism for down-regulating signaling through Gabl by promoting its dissociation from the plasma membrane. Therefore, She, in addition to SHP-2, can play a role in the negative regulation of Gabl-mediated signaling.  The ability on Gabl to form complexes with various signaling molecules allows it to function in multiple signaling pathways downstream of many receptors. The ability of Gabl to interact with negative regulators also provides a mechanism for controlling its own signaling.  Figure 1.11: Model of pathways regulated by the Gabl adapter protein. Briefly, Gabl can interact with PI 3-kinase, leading to its activation and the production of PI-3,4,5P3 within the plasma membrane. Gabl can also activate the MAP kinase pathway through its association with both SHP-2 and She (refer to text for details). However, Gabl's interaction with SHP-2 and She could also negatively regulate signaling through Gabl. SHP-2 could inhibit Gabl activity by de-phosphorylating its tyrosine residues. This dephosphorylation could result in the dissociation of the complex at the membrane and attenuation of signaling.  Alternatively, She could interact with SHIP lipid phosphatase  through Grb2. The recruitment of SHIP to the plasma membrane by Gabl could lead to a decrease in the amount of PI 3-kinase products within the plasma membrane. SHIP can de-phosphorylate PI-3,4,5-P3 and produce PI-3,4-P2 which could result in a decrease in binding sites for PH domain-containing signaling molecules.  1.8 Thesis Goals  Hypothesis: Lymphoid-expressed Src kinase family members and adapter proteins play different roles in regulating BCR-induced signaling by the PI 3-kinase/Akt pathway.  Specific Objectives: (1)To determine the roles of Syk and Src kinase family members in the regulation of signaling by the PI 3-kinase/Akt pathway following BCR cross-linking. (2) To determine the mechanism of BCR-mediated PI 3-kinase/Akt pathway inhibition by specific Src family kinases. (3) To determine if membrane localization of Src kinases affects BCR-mediated PI 3kinase/Akt pathway activation. (4) To determine the role of the Gabl adapter protein in the regulation of the PI 3kinase/Akt pathway in response to BCR cross-linking.  Expermiental System: It would be difficult to detemine the individual roles of the Src kinases in the regulation of the Akt pathway using B lymphoma cell lines since they generally express multiple family members. Therefore, the non-lymphoid AtT20 murine endocrine cell line was used for this study of BCR-mediated PI 3-kinase/Akt regulation. Since AtT20 cells do not express Lyn, Lck or Blk, this system was ideal for comparing the effects of the lymphoid-expressed Src family members on the activation of downstream signaling pathways. Thus, the kinases can be reconstituted individually or in various combinations by DNA transfection or retroviral infection. Another unique characteristic of this system is that signaling is not constitutively active, as is common in most heterologous systems used previously to study Src kinase function and BCR signaling (Kuroaski et al., 1994; Rawlings et al., 1996; El-Hillal et al., 1997; Zoller et al., 1997). Instead, activation of the various signaling pathways was dependent on the cross-linking of the BCR, which was previously introduced into the AtT20 cells by DNA transfection (Matsuuchi et al., 1992). The BCR in the AtT20 cells has been shown to be functional since downstream signaling  pathways can be activated by cross-linking the receptor with anti-IgM antibodies (Matsuuchi et al., 1992; Richards et al., 1996). However, one main disadvantage of this cell system is the lack of other lymphoid-specific proteins including adapter proteins, which may be essential for the proper functioning of the different signaling pathways. Using this gain-of-function approach, different signaling components of interest, including various PTKs such as Syk and Src kinases, were stably introduced into the cells and their individual functions assessed following BCR cross-linking.  1.9  Thesis Summary  In order to test the hypothesis, the non-lymphoid AtT20 cell system was used to express functional BCRs at the cell surface along with different combinations of the Src PTKs, Syk and Gabl. It was determined that Syk kinase activity is required for amplifying and sustaining PI 3-kinase/Akt pathway activation through the BCR [Gold MR, Scheid MP, Santos L. Dang-Lawson M. Roth RA. Matsuuchi L. Duronio V, Krebs PL. 1999. The B cell antigen receptor activates the Akt (protein kinase B)/glycogen synthase kinase-3 signaling pathway via phosphatidylinositol 3-kinase. Journal of Immunology  163: 1894-  1905]. Previously, an in vitro system was used to detemine that the lymphoid-expressed Src kinases, Lyn, Blk and Fyn, differentially associate with the p85 regulatory subunit of PI 3-kinase (Pleiman et al., 1993). If these Src kinases could differentially associate with PI 3-kinase, then they may be able to regulate its activity in some way. In this project, it was established that the various Src kinase expressed in lymphoid cells (Lyn, Blk and Lck) played different roles in the regulation of the PI 3-kinase/Akt pathway in response to BCR cross-linking. Lck has no effect on the activation of this pathway while Lyn and Blk both inhibited the activation of this pathway in response to BCR cross-linking in the presence of Syk. This inhibition correlated with the ability of Lyn and Blk to interact with the SHP-2 protein tyrosine phosphatase. The Lyn-SHP-2 association resulted in a 2fold increase in SHP-2 tyrosine phosphorylation on the tyrosine 580 residue. This increase in phosphorylation, however, was not observed with Blk. The kinase activity of Lyn was also required for the inhibition of Akt phosphorylation.  To test if the  localization within specialized domains at the plasma membrane also contributed to the inhibition of Akt phosphorylation, AtT20 cells expressing acylation mutants of Lyn and Blk were generated. Altering the membrane localization of both these Src kinases had a marginal effect on the BCR-induced Akt phosphorylation. However, expression of a cytosolic mutant form of Blk the AtT20 cells did not have an inhibitory effect on BCRmediated Akt phosphorylation. This is presumably due to the fact that this mutant Blk could not recruit SHP-2 to the plasma membrane. It is hypothesized that recruitment of SHP-2 to the plasma membrane by Lyn and Blk enabled SHP-2 to de-phosphorylate the adapter protein(s) required for PI 3-kinase recruitment and activation. If PI 3-kinase is not recruited to the plasma membrane, it cannot gain access to its substrates and generate the phospholipids required for Akt activation.  The role of the Gabl adapter protein in PI 3-kinase/Akt pathway activation was also examined using the AtT20 cell system [Ingham RJ, Santos L. Dang-Lawson M, HolgadoMadruga M, Dudek P, Maroun CR, Wong AJ, Matsuuchi L and Gold MR. 2001. The Gabl docking protein links the B cell antigen receptor to the phosphatidylinositol 3kinase/Akt signaling pathway and to the SHP2 tyrosine phosphatase.  Journal  of  Biological Chemistry 276:12257-12265]. Gabl can translocate to the plasma membrane and can be inducibly tyrosine phosphorylated in response to BCR cross-linking. The tyrosine phosphorylation of Gab 1 results in its association with the She adapter protein and SHP-2 phosphatase. Akt phosphorylation was 2-3 fold higher following BCR crosslinking in cell transfected with Gabl. The recruitment of Gabl to the plasma membrane, its inducible tyrosine phosphorylation, association with signaling molecules and increased Akt phosphorylation all required the PH domain of Gabl as well as Syk kinase activity. This study demonstrated that various BCR-activated PTKs and the Gabl adapter protein are involved in the regulation of PI 3-kinase/Akt pathway activation downstream of the BCR. The ability of these signaling components to form complexes with other proteins greatly affected the outcome of pathway activation. Furthermore, the cellular localization of the various signaling molecules was shown to be important for the activation of the PI 3-kinase/Akt pathway.  CHAPTER 2 Materials and Methods  2.1  Reagents  2.1.1  Antibodies  The goat anti-mouse IgM (p chain specific), goat anti-human IgM (p chain specific) and goat anti-mouse IgG (Fab')2 fragment specific antibodies used for the stimulation of B cells and BCR-expressing AtT20 cells were purchased from Jackson ImmunoResearch Laboratories, Inc. (West Grove, Pennsylvania; distributed by Bio/Can Scientific, Mississauga, Ontario). The anti-Lyn, anti-Blk, anti-SHP-1, anti-SHP-2, anti-caveolin-1 and anti-Cbl rabbit polyclonal antibodies; anti-Lck, anti-Fyn, anti-CD45 and antiglutathione S-transferase (GST) mouse monoclonal antibodies; and the anti-Btk goat polyclonal antibody were all purchased from Santa Cruz Biotechnology (Santa Cruz, California). The anti-phospho-Akt (serine 473), anti-phospho-Akt (threonine 308) and anti-SHP-2 (tyrosine 580), anti-SHP-2 rabbit polyclonal antibodies were purchased from Cell Signaling Technology (Beverly, Massachusetts). The anti-K light chain rabbit polyclonal and anti-actin mouse monoclonal antibodies were acquired from ICN Biomedicals (Irvine, California). The anti-tubulin mouse monoclonal was acquired from Sigma-Aldrich Canada Limited (Oakville, Ontario). The rabbit polyclonal antibody that recognizes the p85 subunit of PI 3-kinase was obtained from Transduction Laboratories (Lexington, Kentucky). The rabbit polyclonal anti-Gabl antibodies were obtained from Transduction Laboratories. The 4G10 anti-phosphotyrosine mouse monoclonal, anti-She rabbit polyclonal and anti-SHIP-1' rabbit polyclonal antibodies were purchased from Upstate Biotechnology. The rabbit anti-mouse IgM (p chain specific) antibody was obtained from Jackson ImmunoResearch Laboratories and the rabbit polyclonal anti-A. light chain antibody was purchased from Bethyl Laboratories (Montgomery, Texas; distributed by Cedarlane Laboratories, Hornby, Ontario).  Anti-Lck rabbit antiserum was a gift from Dr. Pauline Johnson (University of British Columbia (UBC), Vancouver, British Columbia). The rabbit anti-Ig-P antibody that recognizes the cytoplasmic tail of Ig-P was a gift from Dr. Marcus Clark (University of Chicago, Chicago, Illinois).  The rabbit anti-mouse Ig-a antibody was previously  described (Gold et al., 1991). The rabbit anti-mouse Syk antibody was generated in the Matsuuchi Laboratory with assistance from the UBC Animal Care Facility and has been previously described (Richards et al., 1996).  Normal rabbit IgG and normal mouse IgG used as a negative controls for immunoprecipitations were purchased from Santa Cruz Biotechnology. The horseradish peroxidase (HRP)-conjugated secondary antibodies used for Western immunoblotting were purchased from the following sources: Protein A-HRP and Protein G-HRP were from Amersham Biosciences Canada (Baie d'Urfe, Quebec); goat anti-rabbit IgG-HRP was from Jackson ImmunoResearch Laboratories; and goat anti-mouse IgG-HRP was from Invitrogen Life Technologies Canada (Burlington, Ontario).  2.1.2  Plasmids  The pRC/CMV  Syk and pRC/CMV  kinase dead Syk plasmids (constructs containing genes  encoding for wild type murine and catalytically inactive Syk, respectively) have been previously described (Richards et al., 1996). The plasmids containing the murine Lyn (pBS-Lyn) and the murine Blk genes (pBS-Blk) in the pBlueScript vector were gifts from Dr. Anthony DeFranco (University of California San Francisco, San Francisco, California). The RSV pLpA mammalian expression vector was created by Dr. Linda Matsuuchi (Ph.D. thesis supervisor) using a plasmid containing a Rx)us sarcoma virus (RSV) promoter obtained from Dr. William Rutter (University of California San Francisco). The murine Lyn gene (p56 isoform) was cloned into the RSV pLpA vector by Dr. Linda Matsuuchi, thus creating the RSVpLpA-Lyn EGFP-Gabl  and pMX-EGFP-APH  expression vector. The pMX-  Gabl were gifts from Dr. Christiane Moroun (McGill  University, Montreal, Canada) and obtained from Dr. Michael Gold (UBC) and have been previously described (Ingham et al., 2001). The pMSCV-CD16/CD17  plasmid was  also from Dr. Michael Gold. The pRc/CMV-Lck  was a gift from Dr. Jamey Marth  (University of California San Diego, La Jolla, California). The pWZL Blast 1 andpWZL Blast 2 vectors containing the gene that confers resistance to blasticidin were gifts from Dr. Stephen Robbins (University of Calgary, Calgary, Alberta). The LPCsrf plasmid, which contains the gene that confers resistance to puromycin, was a gift from Dr. Anthony DeFranco (University of California San Francisco). The pMSCV-puro retroviral expression vector andpSVneo were purchased from BD Biosciences Clontech (Palo Alto, California).  The pBlueScript  II KS vector was purchased from Stratagene (La Jolla,  California).  The RSVpLpA-Blk  (cytosolic mutant) was created by Teresa Jackson (a former research  assistant in the laboratory). Briefly, polymerase chain reaction (PCR) was performed to introduce a start codon in the murine Blk gene within pBS-Blk construct. PCR was also used to create Xbal sites that flanked the Blk gene. These sites were used to insert the Blk gene into the Xbal site of the RSV pLpA mammalian expression vector, resulting in the creation of RSV pLpA-Blk,  However, it was later found that the Blk gene within the  construct contained mutations at the amino terminal (the amino acids at position 2 and 3 were not similar to the wild type amino acids mutated from glycine2-leucine3 to aspartic acid2-proline3) and was therefore renamed RSV pLpA-Blk (cytosolic mutant).  2.1.3  GST fusion proteins  The construct containing GST fused to the amino-terminus of the tandem SH2 domains of SHP-2 (GST-SHP-2 (SH2-SH2)) was a gift from Dr. Frank Jirik (University of Calgary). The construct encoding for the GST protein alone was a gift from Dr. Stephen Robbins.  The preparation and purification of the GST proteins was performed with the help of Gabe Woollam, a former undergraduate student in the laboratory. Briefly, a single bacterial colony was used to inoculate 20 ml of Luria-Bertani (LB) liquid culture (85 mM NaCl, 5 mg/ml yeast extract, 10 mg/ml tryptone) containing 100 pg/ml ampicillin. Following an overnight incubation at 37°C with shaking, the 20 ml of bacterial culture  was used to inoculate 1 liter (L) of LB liquid culture containing 100 pg/ml ampicillin. The entire culture was then incubated at 37°C until an OD600 between 0.8 and 1.0 was reached (this usually takes a few hours).  Isopropylthio-P-galactopyranoside (IPTG)  (Invitrogen Life Technologies Canada) was then added to a final concentration of 100 pM and the culture was incubated overnight at 26°C with shaking. The following day, the bacteria was collected by centrifugation and lysed in lysis buffer (50 mM Tris-HCl pH 8.0, 150 mM NaCl, 1% Triton X-100 (Sigma-Aldrich Canada), 1 mg/ml lysozyme (Sigma-Aldrich Canada), 0.1 mg/ml DNAse I, 10 pg/ml soybean trypsin inhibitor (Sigma-Aldrich Canada), 10 pg/ml leupeptin (Roche Diagnostics, Laval, Quebec), 1 pg/ml aprotinin (Roche Diagnostics), 1 mM phenylmethylsulfonyl fluoride (PMSF) (Roche Diagnostics)) for 30 minutes on ice. For further lysis, the cells were sonicated for 2 minutes while on ice to prevent overheating. The lysed bacterial cells were centrifuged at 30,000 rpm for 45 minutes at 4°C in a Beckman L8-70 Ultracentrifuge in a SW70 Ti rotor (Beckman Coulter, Fullerton, California) to separate the insoluble material from the supernatant. The supernatant containing the GST-SHP-2 (SH2-SH2) fusion protein was used for the pull-down precipitation experiments without further purification. The GST protein used for pre-clearing cell lysates in the pull-down experiments, however, was purified using glutathione-Sepharose 4B beads (Amersham Biosciences Canada). The bacterial lysates were incubated with the washed glutathione-Sepharose 4B beads for 1 hour in the cold while rocking. The beads were then collected by centrifugation and washed three times with wash buffer (25 mM Tris-HCl pH 8.0, 150 mM NaCl, 0.1 % Triton X-100, 10 pg/ml soybean trypsin inhibitor, 10 pg/ml leupeptin, 1 pg/ml aprotinin, 1 mM PMSF. The bound fusion protein on the glutathione-Sepharose 4B beads were eluted multiple times under mild non-denaturing conditions with elution buffer (50 mM Tris base, 20 mM glutathione (Amersham Biosciences), pH 8.0). The eluted GST was dialyzed against 10 mM Tris-HCl pH 8.0 at 4°C to remove the free glutathione. The purity of the GST protein was assessed by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) and Coomasie staining. The concentration of the protein was determined by examining the OD280 with the Ultrospec 2100 pro UV/visible spectrophotometer (Biochrom, Cambridge, United Kingdom; distributed by Fisher Scientific Canada, Nepean, Ontario).  2.2  Tissue culture and cell stimulations  2.2.1  Tissue culture cell lines  The AtT20 murine endocrine cell line was from Dr. Regis Kelly (University of California San Francisco) and have been previously described (Matsuuchi and Kelly, 1991). The WEHI 231 murine B lymphoma, BJAB human B lymphoma, Ramos human B lymphoma, Daudi human B lymphoma, A20 mature murine B lymphoma and 2PK3 mature murine B lymphoma cell lines were obtained from the American Type Culture Collection (Manassas, Virginia).  The Ball7 murine B lymphoma, J558 murine  plasmacytoma, CHI2 murine B lymphoma, CH31 murine B lymphoma, K40-B1 pro-B and K40-B2 pre-B cell lines were generous gifts from Dr. Anthony DeFranco. The BOSC 23 human retrovirus packaging cell line was a gift from Dr. Warren Pear (University of Pennsylvania, Philadelphia, Pennsylvania).  2.2.2  Culture of cell lines  The human and murine B lymphoma cell lines were maintained in complete Roswell Park Memorial Institute (RPMI)-1640 media (Invitrogen Life Technologies Canada) containing 10% heat-inactivated fetal calf serum (FCS) (Invitrogen Life Technologies Canada), 2 mM L-glutamine (Invitrogen Life Technologies Canada), 1 mM sodium pyruvate (Invitrogen Life Technologies Canada), 50 pM 2-(3-mercaptoethanol (SigmaAldrich Canada), 50 units/ml penicillin and 50 pg/ml streptomycin sulfate (Invitrogen Life Technologies Canada). The B lymphoma cell lines were grown in tissue culture flasks of varying sizes (Falcon, Franklin Lakes, New Jersey; distributed by VWR International, Edmonton, Alberta) at 37°C in an atmosphere of 5% C 0 2 in a waterjacketed incubator (Forma Scientific, Marietta, Ohio). The B lymphoma cell lines were spilt into new flasks every 3-4 days.  AtT20-derived and BOSC 23 cells were maintained in complete Dulbecco's modified Eagle's medium (DMEM) (Invitrogen Life Technologies Canada) containing 4.5 g/L glucose, 2 mM L-glutamine, 110 mg/L sodium pyruvate and supplemented with 10% FCS, 50 units/ml penicillin and 50 pg/ml streptomycin sulfate. The AtT20 and BOSC 23  cells were grown at 37°C in a 10% CO2 atmosphere water-jacketed incubator. The adherent cells were cultured in 10 cm polystyrene tissue culture dishes (Falcon). The cells were generally grown to approximately 90%-100% confluency prior to being split into a new tissue culture dish. The adherent cells were removed from the tissue culture dishes by aspirating off the old DMEM media and adding 1 ml of a 0.25% trypsin/1 mM (ethylenedinitrilo)tetraacetic acid (EDTA) solution (Invitrogen Life Technologies Canada) for 2 minutes. Once the cells were loosened and lifted off the tissue culture plate, 9 ml of complete DMEM media was added to neutralize the trypsin. The cells were then pipetted up and down with a 10 ml sterile plastic pipette (Corning, Corning, New York) to break up any cell clumps. Once the cell clumps have been broken up, 0.51.0 ml of the cell suspension was placed into a new 10 cm tissue culture dish containing 10 ml of fresh complete DMEM media and grown as described above.  Generally, the B lymphoma and AtT20 cell lines were kept in cell culture for no more than three months since some cells may change in culture and these changes may interfere with the signaling pathways being studied. BOSC 23 cells are kept in culture for a maximum of three passages since they lose their ability to produce a high titer of retroviruses if kept in culture for a longer period of time. Multiple copies of the different cell lines are stored in liquid nitrogen and individual vials of cells are thawed when required. The various cell lines are frozen in freezing media (FCS containing 10% dimethyl sulfoxide (DMSO) (Sigma-Aldrich Canada) for lymphoid cells; complete DMEM containing 10% DMSO for AtT20-derived and BOSC 23 cells).  2.2.3  Stimulation and lysis of cells  AtT20-transfected cells were generally grown to approximately 90% confluency on 5 cm or 10 cm tissue culture dishes (Falcon) for cell stimulation experiments. To reduce signaling from serum-derived growth factors in the media, the media in which the AtT20 cells were cultured was removed and the cells were washed twice with Dulbecco's phosphate-buffered saline (D-PBS) (Invitrogen Life Technologies Canada). Following the washes, the cells were cultured for an additional 13-16 hours at 37°C in low serum media (DMEM supplemented with 0.2% FCS, 50 units/ml penicillin and 50 pg/ml  streptomycin sulfate). Before being stimulated, the cells were washed twice with D-PBS and incubated for 15 minutes at 37°C in modified HEPES-buffered saline (25 mM HEPES pH 7.2, 125 mM NaCl, 5 mM KC1, 1 mM CaCl2, 1 mM Na 2 HP0 4 , 500 pM MgS0 4 , 1 mg/ml glucose, 2 mM L-glutamine, 1 mM sodium pyruvate, 2% bovine serum albumin (BSA) (ICN Biomedicals). To cross-link the BCRs on the cell surface, affinity purified goat anti-mouse IgM (p chain specific) antibodies were added to the cells to a final concentration of 20 pg/ml. The reaction was terminated by aspirating the modified HEPES-buffered saline containing the stimulating antibodies and washing the cells twice with ice cold phosphate-buffered saline (PBS) (1.5 mM NaCl, 1.9 mM N a H 2 P 0 4 H 2 0 , 8.4 mM Na 2 HP0 4 , pH 7.2) containing 1 mM Na 3 V0 4 (Sigma-Aldrich Canada). After the final wash, the cells were solubilized with ice cold 0.5-1.0 ml of Triton X-100 lysis buffer (20 mM Tris-HCl pH 8.0, 137 mM NaCl, 1% Triton X-100, 2 mm EDTA (Fisher Scientific Canada), 10% glycerol, 10 pg/ml leupeptin, 1 pg/ml aprotinin, 1 mM pepstatin A (Sigma-Aldrich Canada), 1 mM Na 3 V0 4 , 1 mM PMSF).  Following a 20 minute  incubation on ice, the lysed cellular material was transferred to a 1.5 ml eppendorf tube (Eppendorf/Brinkmann Instruments, Mississauga, Ontario). The detergent insoluble material was separated from the soluble material by centrifugation in a microcentrifuge at 14 000 rpm for 15 minutes in the cold. After the soluble cell lysates were transferred to a new 1.5 ml tube, SDS (Bio-Rad Laboratories, Hercules, California) and sodium deoxycholate (DOC) (Fisher Scientific Canada) were added to a final concentration of 0.3% and 0.4%, respectively. The lysates were then quickly frozen in liquid nitrogen for approximately 5-10 seconds and subsequently stored at -20°C. The protein concentration of each sample was determined using the bicinchoninic acid (BCA) protein assay kit (Pierce Biotechnology, Rockford, Illinois).  For stimulation of B lymphoma cell lines, the cells were pelleted by low speed centrifugation (1500 rpm) in a IEC Centra-8R tabletop centrifuge (International Equipment Company, Needham Heights, Massachusetts) for 3 minutes. Once collected, the cells were washed once with modified HEPES-buffered saline and then re-suspended in the same buffer at a concentration of 2.5 x 107 cells per ml. The cells were warmed to 37°C for 15 minutes in a water bath and stimulated with anti-BCR antibodies at a final  concentration of 100 pg/ml.  Goat anti-mouse IgG F(ab')2 antibodies were used to  stimulate the more mature A20 and 2PK3 B cell lines, while the less mature B lymphoma lines were stimulated with goat anti-mouse IgM (p chain specific) antibodies. Goat antihuman IgM (p chain specific) antibodies were used for stimulation of the human B lymphoma cell lines. The reactions were terminated by adding ice cold PBS containing 1 mM Na3V04 and pelleting the cells by low speed centrifugation for 2 minutes at 4°C. The cell pellets were washed an additional time with ice cold PBS containing 1 mM Na 3 V04 and then lysed in Triton X-100 lysis buffer for 20 minutes on ice. Following lysis, the detergent soluble and insoluble material was separated by centrifugation for 15 minutes at 14 000 rpm in a microcentrifuge in the cold. The cell extracts were then transferred to new tubes and detergents SDS and DOC were added to a final concentration of 0.3% and 0.4%, respectively. The lysates were quickly frozen in liquid nitrogen for approximately 5-10 seconds and subsequently stored at -20°C. The protein concentration of each sample was determined using the BCA protein assay kit.  For experiments in which cell stimulation was not required, the cells were simply washed twice with cold PBS and lysed as described above.  2.3  DNA transfection of cells  2.3.1  Transient transfection of BOSC 23 cells  Transient transfections of BOSC 23 cells were routinely conducted to determine if a new plasmid construct was expressible in mammalian cells. Calcium phosphate-mediated transfection of BOSC 23 cells was performed basically as described by Pear et al. (1993) and Krebs et al. (1999) with a few minor modifications. BOSC 23 cells were grown to approximately 80% confluency in 10 cm tissue culture dishes (as described in detail in Chapter 2.2.2 Culture of cell lines). The day before the transfection, the DNA to be used was sterilized by adding one drop of chloroform and incubating the DNA overnight in the cold. On the day of the transfection, the DNA was prepared by mixing 8 pg of the desired plasmid DNA with 800 pi of 250 mM CaCh in a 12 x 75 mm clear polystyrene  tube (Falcon) and vortexing for 5 seconds at maximum speed. 800 pi of 2X HEPESbuffered saline (50 mM HEPES pH 7.2, 10 mM KC1, 12 mM glucose, 280 mM NaCl, 1.5 mM Na2HPC>4) was then added dropwise to the DNA/CaCl 2 mixture while vortexing at maximum speed for 10-20 seconds.  The whole mixture was then vortexed for an  additional 20 seconds. Once the DNA mixture was complete, the media was aspirated from he tissue culture dishes containing the BOSC 23 cells and replaced with 5 ml of room temperature complete DMEM containing 25 pM chloroquinone (Sigma-Aldrich Canada). The DNA mixture was then added dropwise to the cells and swirled gently. The cells were incubated with the DNA mixture for a minimum of 8 hours in a 37°C 10% CO2 incubator.  Following the incubation, the DMEM media containing the DNA  mixture was aspirated off the cells and replaced with complete DMEM media. The cells were lysed 48 hours post-transfection (from the time the DNA was added to the cells) and the whole cell lysates analyzed for the presence of the specific protein encoded by the transfected plasmid. 2.3.2  Calcium phosphate transfection of AtT20 cells  Transfection of AtT20 cells was performed essentially as described previously in Matsuuchi and Kelly (1991) with some modifications. The appropriate AtT20 cells were grown to 80% confluency on 10 cm tissue culture dishes (as described in Chapter 2.2.2 Culture of cell lines) for at least 4 days to ensure that the cells are well-adhered to the dish. The DNA for the transfection was prepared by mixing 50 pg plasmid DNA, 20 pg selectable drug resistance marker plasmid DNA, 94 pi 2M CaCh and sterile distilled water up to a final volume of 750 pi. While vortexing at maximum speed, 750 pi of 2X HEPES-buffered saline was added dropwise using cotton-plugged pipette tips (VWR International).  Once mixed thoroughly, the DNA mixture was left to sit at room  temperature for 40 minutes to form a calcium phosphate precipitate with the DNA. After the 40 minute incubation, the dishes containing the cells to be transfected were washed twice with IX HEPES-buffered saline (25 mM HEPES pH 7.2, 5 mM KC1, 6 mM glucose, 140 mM NaCl, 750 pM Na 2 HP0 4 ). The DNA calcium precipitate was then carefully added to the washed cells and incubated at 37°C for 20 minutes. 10 ml of complete DMEM media was then added to dilute the DNA calcium precipitate and the  cells were once again incubated in a 37°C 10% CO2 incubator for 6-7 hours. Following the incubation period, the media containing the DNA calcium precipitate was aspirated from the cells and discarded. The cells were then glycerol shocked by adding 1 ml of 25% glycerol solution (complete DMEM containing 25% glycerol) dropwise onto the cells and swirling the plates for exactly 1 minute. The glycerol was subsequently diluted with 10 ml of D-PBS and then aspirated. The cells were rinsed two more times with 10 ml of D-PBS to ensure that no glycerol was left on the cells as it is toxic to them. Following the final wash, 10 ml of complete DMEM media was added back to the cells and then incubated for three days in a 37°C 10% CO2 incubator.  2.3.3  Drug selection of transfected cells and isolation of individual clones  Three days following the transfection, the cells were split into five to ten 10 cm tissue culture dishes with complete DMEM media containing 2 pg/ml blasticidin S (Invitrogen Life Technologies Canada), 0.4 mg/ml G418 neomycin (Invitrogen Life Technologies Canada) or 0.4 pg/ml puromycin (Calbiochem, La Jolla, California). Over the course of 4-6 weeks, the transfected cells were washed with D-PBS every 4 days to remove any dead cells that have not taken up the drug resistance plasmid DNA. Once stable drug resistant clones were visible to the naked eye on the tissue culture dishes (approximately the 64-128 cell stage), individual clones were isolated using sterilized 9 mm Teflon cloning rings (Fisher Scientific Canada).  The cloning rings were sterilized by  autoclaving on a layer of high vacuum grease (Dow Corning, Midland, Michigan), which facilitated the formation of a seal around the individual clones on the tissue culture dishes. Once placed over single, well-separated clones, 1 drop of trypsin/EDTA solution was added into the ring to loosen the clone from the dish. Once lifted from the tissue culture dish, the cloning rings were filled with complete DMEM media containing the appropriate drug following which the media was pipetted several times to break up the cluster of cells and transferred into 24-well tissue culture plates (Falcon). The clones were successively transferred into plates with larger wells (from 24- to 12- to 6- cell plates). Once in 12- or 6-well plates, the clones were screened to determine if they expressed the particular protein encoded for by the transfected plasmid by Western immunoblotting and in some cases, by immunofluorescence.  In this thesis, the  components that have been transfected during the course of this thesis will be in bold letters.  2.4  Retroviral infection of cells  2.4.1  Production of retroviruses using the BOSC 23 packaging cell line  BOSC 23 cells were transfected with the desired retroviral plasmid as described earlier (Chapter 2.3.1 Transient transfection of BOSC 23 cells). However, instead of lysing the cells following the transfection, the culture supernatant that the cells were grown in was collected 45-46 hours post-transfection.  This culture supernatant contains the virus  particles that the BOSC 23 cells release. This supernatant was syringe-filtered using a 0.22 pm filter (Millipore, Billerica, Massachusetts). This sterile solution of culture supernatant containing the virus particles was either frozen and -80°C or used immediately for the infection of AtT20 cells.  2.4.2  Retroviral infection of AtT20 cells  The AtT20 cells were grown to 80% confluency on 10 cm tissue culture dishes (as described in Chapter 2.2.2 Culture of cell lines) for at least 4 days to ensure that the cells are well-adhered to the dish. On the day of the infection, the media in which the AtT20 cells were grown was aspirated and replaced with 5-6 ml of the sterilized media (from the BOSC 23 cells) containing the virus particles and 10 pg/ml polybrene (hexadimethrine bromide, Sigma-Aldrich Canada).  The AtT20 cells were incubated with the virus  particles at 37°C for 6-7 hours. Following the incubation, the media containing the virus particles was aspirated and the cells were washed once with D-PBS. 10 ml of complete DMEM media was added to the cells and grown for an additional 48 hours at 37°C in a 10% CO2 incubator. The cells were split 48 hours post-transfection into five to ten 10 cm tissue culture dishes containing 0.4 pg/ml puromycin.  Dead cells, which were not  successfully infected, were washed off with D-PBS every 4 days over the course of 2-3 weeks. Once individual clones were visible (this usually took 2-3 weeks), they were  isolated as described above. The components that have been added through transfection ore retroviral infection during the course of this thesis will be in bold.  2.5  SDS-PAGE and Western immunoblotting  Samples of whole cell lysates or immune complexes containing SDS-PAGE reducing sample buffer (to a final concentration of 62.5 mM Tris-HCl pH 6.8, 4% glycerol, 2.5% SDS, 0.02% bromophenol blue, 100 mM dithiothreitol (DTT) (Sigma-Aldrich Canada)) were heated in a boiling water bath for 5 minutes. The samples along with BenchMark pre-stained protein molecular weight standards (ranging in size from 6-190 kilodaltons (kDa); Invitrogen Life Technologies Canada) were separated on 1.0 and 1.5 mm thick SDS-PAGE mini-gels at a constant current of 15 milliamps per gel for approximately 2-3 hours in a dual vertical mini-gel apparatus unit (CBS Scientific, Del Mar, California). The mini-gels used ranged from 5-15% acrylamide (Bio-Rad Laboratories) depending on the size of the particular protein being examined. The separated proteins within the minigel were then transferred onto Protran nitrocellulose (VWR International) in a Transblotter. transfer apparatus (Bio-Rad Laboratories) at a constant voltage of 110 volts for 2 hours in transfer buffer (20 mM Tris-HCl pH 8.0, 150 mM glycine, 20% methanol). Following the transfer, the separated proteins were visualized by Ponceau staining (0.2% Ponceau powder (Sigma-Aldrich Canada), 3% trichloroacetic acid (TCA)) for 30 seconds to ensure that the proteins had transferred properly onto the nitrocellulose filter. The Ponceau was subsequently washed off with water prior to incubating the nitrocellulose filters in antibody solution.  When the 4G10 anti-phosphotyrosine mouse monoclonal antibody was used for Western immunoblotting, the nitrocellulose filters were blocked in Tris-buffered saline (TBS) (10 mM Tris-HCl pH 8.0, 150 mM NaCl) containing 5% BSA for a minimum of 2 hours at room temperature. For all the other Western blots, the filters were blocked with TBS containing 5% skim milk powder for 1 hour at room temperature. Once blocked, the filters were rinsed quickly with TBS containing 0.1% Tween 20 (Sigma-Aldrich Canada) to remove the excess BSA or skim milk powder. The filters were subsequently incubated  with primary antibody that was diluted in TBS containing 0.1% Tween 20 and 10% BSA (for less sensitive antibodies and the 4G10 anti-phosphotyrosine monoclonal) or 2.5% skim milk powder (for strong antibodies or ones that have a lot of background) rocking in the cold overnight. The following day, the filters were washed for 1 hour (changing the TBS containing 0.1% Tween 20 every 15 minutes) and then incubated with HRPconjugated secondary antibodies for 1 hour at room temperature.  The secondary  antibodies were all diluted 1:10 000 in TBS containing 0.1% Tween 20 and 5% skim milk powder or 5% BSA (for the 4G10 anti-phosphotyrosine monoclonal). Following a second set of four 15 minute washes to rinse off the unbound secondary antibodies, the bands on the nitrocellulose filters were visualized by enhanced chemiluminescence (ECL) detection (Amersham Biosciences Canada) and exposure to Kodak autoradiography film (Mandel Scientific, Guelph, Ontario) for varying lengths of time depending on the strength of the antibody signal.  The nitrocellulose filters were re-probed by first vigorously washing off the bound antibodies with low pH TBS (pH 1.5-2.0) for 20 minutes at room temperature. The filters were then washed twice for 10 minute increments with TBS pH 8.0 to neutralize the low pH solution. Following removal of the bound antibodies, the filters were subsequently blocked and re-probed with antibodies, as described above.  2.6  Precipitation experiments  2.6.1  Immunoprecipitations  For immunoprecipitations, the cells were lysed in Triton X-100 lysis buffer as described earlier (Chapter 2.2.3 Stimulation and lysis of cell lines). No extra detergents (SDS or DOC) were added to any of the cell lysates used for immunoprecipitations. One to two mg of cell lysate was first pre-cleared with 10-20 pi of packed washed Protein A- or Protein G-Sepharose 4B beads (Sigma-Aldrich Canada) for 1 hour in the cold in 1.5 ml eppendorf tubes. The samples were then centrifuged for a few seconds to collect the beads. The pre-cleared lysates were then transferred to new tubes containing 1-5 pg of  the desired antibody and 20-30 pi of packed washed Protein A- or Protein G-Sepharose 4B beads and subsequently rocked for 1 hour in the cold. Following the incubation the immune complexes were collected by centrifugation and washed twice with Triton X-100 lysis buffer. The bound proteins were then eluted by adding SDS-PAGE reducing sample buffer and heating the samples for 5 minutes in a boiling water bath. The precipitates were then transferred to new tubes and loaded onto a SDS-PAGE mini-gel.  2.6.2  Pull-down experiments with GST fusion proteins  To pre-clear cell extracts, 4 pg of purified GST protein was first bound to 5 pi of packed washed glutathione-Sepharose 4B beads (Amersham Biosciences Canada) by incubating them together with Triton X-100 lysis buffer for 30 minutes in the cold. The GSTglutathione-Sepharose 4B bead complexes were then washed twice with Triton X-100 lysis buffer following the incubation. One to two mg of cell lysate was then added to the washed GST-glutathione-Sepharose 4B bead complexes and rocked for 30 minutes in the cold. This pre-clearing step eliminates proteins from the cell lysates that may associate with either the GST protein or the glutathione-Sepharose 4B beads and allows for the isolation of proteins which only interact with the GST-fusion protein. Following the incubation step, the glutathione-Sepharose 4B beads were collected by centrifugation and the pre-cleared lysates were transferred to a new tube containing GST-SHP-2(SH2-SH2) fusion protein bound to glutathione-Sepharose 4B beads (created in the same manner as described above, using 20 pi bacterial cell lysates containing the GST fusion protein and 20 pi packed washed glutathione-Sepharose 4B beads). Following a 1 hour incubation in the cold, the complexes were washed twice with Triton X-100 lysis buffer and the bound proteins were eluted by boiling the samples with SDS-PAGE reducing sample buffer for 5 minutes.  2.6.3  In vitro kinase assay  One confluent 10 cm plate of cells were lysed in Triton X-100 lysis buffer as described earlier (Chapter 2.2.3 Stimulation and lysis of cell lines). All of the cell lysates obtained were first pre-cleared with 10-20 pi of packed washed Protein A- or Protein G-Sepharose 4B beads for 1 hour in the cold. The samples were then centrifuged for a few seconds to collect the beads. The pre-cleared lysates were then transferred to new tubes containing 5 pg of the desired antibody along with 20-30 pi of packed washed Protein A- or Protein G-Sepharose 4B beads and subsequently rocked for 2-3 hours in the cold. Following the incubation the immune complexes were collected by centrifugation and washed twice with Triton X-100 lysis buffer containing 500 mM NaCl and once with in vitro kinase assay buffer (20 mM Tris-HCl pH 7.2, 10 mM MgCl 2 , 0.1% NP-40, 10 mM MnCl2). Following the washes, the immunoprecipitates were incubated with 100 pi in vitro kinase assay buffer containing 10 pCi radiolabeled [y- 32 P]-ATP (Amersham Biosciences Canada) for 20 minutes at room temperature. To terminate the reaction, 500 pi of cold wash buffer (10 mM Tris-HCl pH 7.2, 100 mM NaCl, 1 mM EDTA, 1% NP-40, 0.3% SDS) was added to the samples.  The immunoprecipitates were collected by  centrifugation and the supernatant was discarded. The bound proteins were then eluted by adding SDS-PAGE reducing sample buffer and heating the samples for 5 minutes in a boiling water bath. The precipitates were then transferred to new tubes and loaded onto a SDS-PAGE mini-gel. The mini-gels were subsequently dried using a gel dryer.  2.7  Isolation of lipid rafts by sucrose density gradient ultracentrifugation  2.7.1  Cell stimulations for sucrose density gradient ultracentrifugation  For the B lymphoma cell lines, at least 109 cells per treatment (i.e. unstimulated and BCR-stimulated) were used for the sucrose density gradient ultracentrifugation. The cells were stimulated with goat anti-mouse IgM (p chain specific) or goat anti-mouse IgG F(ab')2 antibodies as described in a previous section (in Chapter 2.2.3 Stimulation and lysis of cell lines).  For transfected AtT20 cells, at least ten confluent 10 cm dishes of cells (approximately 109 cells) were used per treatment. The cells were removed from the tissue culture dishes by chelating them with 1-2 ml of D-PBS containing 10 mM EDTA for approximately 3 minutes while rocking. The cells were loosened by vigorously shaking the dishes and pipetting.  Once loosened, the cells were collected in 15 ml polypropylene conical  centrifuge tubes (Falcon) and centrifuged at low speed (1500 rpm) in a tabletop centrifuge. The cells were washed once with D-PBS and once with modified HEPESbuffered saline to rinse off excess EDTA before being re-suspended in 5 ml of modified HEPES-buffered saline. The cells were warmed in a 37°C water bath for 20 minutes following which they were stimulated with goat anti-mouse IgM (p chain specific) antibodies for 10 minutes at 37°C. The reactions were terminated by adding 10 ml of ice cold PBS containing 1 mM Na3V04 and pelleting the cells by low speed centrifugation for 2 minutes at 4°C. The cell pellets were washed an additional time with ice cold PBS containing 1 mM Na3V04 and then transferred to 1.5 ml polypropylene tubes (Kimble/Kontes, Vineland, New Jersey)/  2.7.2  Discontinuous sucrose gradient density ultracentrifugation  Lipid rafts were isolated by sucrose gradient density ultracentrifugation essentially as previously described (Deans et al., 1998; Petrie and Deans, 2002) with a few modifications.  Once in the 1.5 ml polypropylene tubes, the cells were pelleted by  centrifugation, all of the PBS containing 1 mM Na3V04 was aspirated and the cells were lysed with 2-(N-morpholino)ethanesulfonic acid (MES) (Fisher Scientific Canada)-lysis buffer (25 mM MES pH 6.5, 150 mM NaCl, 1% Triton X-100, 10 pg/ml leupeptin, 1 pg/ml aprotinin, 1 mM pepstatin A, 1 mM Na 3 V0 4 , 1 mM PMSF, 2 mM EDTA) for 20 minutes on ice. Following lysis, the samples were homogenized (15 strokes) using a pellet pestle (Kimble/Kontes; product number 749520-0090). The homogenized lysates were then mixed with and equal volume of 80% sucrose solution (made in MES-buffered saline (25 mM MES pH 6.5, 150 mM NaCl) containing 10 pg/ml leupeptin, 1 pg/ml aprotinin, 1 mM pepstatin A, 1 mM Na 3 V0 4 , 1 mM PMSF, 2 mM EDTA), making the concentration of the homogenized lysates 40% sucrose. The lysates were transferred to 9/16"  x 3 V2" ultra-clear centrifuge tubes (Beckman Coulter) and carefully overlaid with 4  ml 30% sucrose solution and then with 6 ml 5% sucrose solution. The samples were centrifuged at 37 000 rpm, no brake, for 3 hours at 4°C in a SW41 Ti swinging bucket rotor in a Beckman L8-70 Ultracentrifuge. Following centrifugation, a band of material visible at the 5%/30% sucrose interface, which was previously shown to contain the lipid rafts (Petrie et al, 2000), was collected, mixed with cold MES-buffered saline and recentrifuged at 37 000 rpm for 1 hour at 4°C. Following the second centrifugation, the lipid rafts were re-suspended in a small volume (generally less than 200 pi) of cold MESbuffered saline. Additional samples containing non-lipid raft or soluble fractions were collected at 12 ml (at the 30%/40% sucrose interface, Fraction 12) and 14 ml (Fraction 14) from the top of the tube. The pelleted material at the bottom of the tube was washed twice and then re-suspended by vortexing in ice cold MES-buffered saline. Alternative methods in which equal volumes of each fraction is analyzed have been used by other investigators (Deans et al., 1998; Cheng et al., 1999). However, due to the variation in the amount of lipid raft samples collected from experiment to experiment, I chose to determine the protein concentration of each sample using a BCA assay kit. Thus, for the purpose of analysis and to allow for consistent interpretation of the data, equal amounts of protein from each fraction collected were separated by SDS-PAGE. The variation may have been due to the membrane composition of the cells, how well the cells were lysed or how well the lysates were homogenized.  2.8  Membrane-enrichment  The AtT20 cells were removed from the tissue culture dishes by chelating them with 1 -2 ml of D-PBS containing 10 mM EDTA for approximately 3 minutes while rocking. The cells were loosened by vigorously shaking the dishes and pipetting. Once loosened, the cells were collected in 15 ml polypropylene conical centrifuge tubes (Falcon) and centrifuged at low speed (1500 rpm) in a tabletop centrifuge. The cells were washed once with D-PBS and transferred to 1.5 ml eppendorf tubes. The tube of cells was then frozen in liquid nitrogen for 15 seconds. Two hundred pi of "no detergent lysis buffer" (20 mM Tris-HCl pH 8.0, 137 mM NaCl, 10% glycerol, 2 mM EDTA) was then added to the frozen pellet of cells and incubated on ice for 5 minutes. Following the incubation,  the cell pellet was re-suspended in the "no detergent lysis buffer" and then centrifuged at 14 000 rpm for 5 minutes in the cold. The supernatant (cytosolic fraction) was then transferred to a new 1.5 eppendorf tube, 20 pi of 10% Triton X-100 was subsequently added to this fraction. The remaining pellet was washed twice with "no detergent lysis buffer." Following the washes, the pellet was re-suspended in 220 pi 1% Triton X-100 lysis buffer and centrifuged at 14 000 rpm for 5 minutes in the cold. The supernatant (membrane fraction) was subsequently remove and the remaining pellet discarded. The protein concentration of each sample was determined using the BCA assay kit.  2.9  Confocal microscopy  2.9.1  Preparation of poly-D-lysine coated coverslips  Circular glass coverslips (12mm diameter x 0.13-0.17 mm thick; Fisher Scientific Canada) were submerged in solution containing poly-D-lysine (1 mg/ml poly-D-lysine (Sigma-Aldrich Canada), 50 mM sodium borate pH 8.5) for 2 hours at room temperature. The coverslips were then rinsed well three times with PBS and three times with distilled water to get rid of excess poly-D-lysine since it is toxic to cells. Following the multiple washes, the coverslips were subsequently autoclaved in a covered glass dish and air dried in a sterile biosafety cabinet until thoroughly dry.  2.9.2  Culture and stimulation of cells on coverslips  Using sterile techniques, single poly-D-lysine coated coverslips were placed into individual wells of 24 well cluster dishes (Falcon) containing 1 ml of complete DMEM media. A small number of AtT20 transected cells expressing green fluorescent protein (GFP)-tagged proteins were initially seeded into the wells containing the coverslips to ensure that the cells did not overgrow into a confluent monolayer. The cells were grown to 75% confluency at 37 °C in an atmosphere of 10% C02 for a few days to ensure that the cells adhered well to the coated coverslips.  One day prior to stimulation, the  complete media in which the cells were grown was aspirated and the cells were rinsed twice with D-PBS and cultured for a maximum of 16 hours under low serum conditions  (in DMEM supplemented with 0.2% FCS). On the day of the stimulation, the cells were rinsed twice with D-PBS and incubated for 20 minutes at 37°C in 1 ml of modified HEPES-buffered saline without any inhibitors, with 25 pM LY294002 (BIOMOL Research Laboratories, Plymoth Meeting, Pennsylvania) or with 30 nM wortmannin (BIOMOL Research Laboratories). Following the incubation, cells were then washed once with D-PBS. To cross-link the BCRs on the cell surface, the cells were incubated in modified HEPES-buffered saline containing the appropriate inhibitors and 20 pg/ml goat anti-mouse IgM (p chain specific) antibodies for varying periods of time. The reactions were terminated by aspirating the modified HEPES-buffered saline and rinsing the cells twice with PBS containing 1 mM Na3V04.  2.9.3  Preparation of cells for confocal microscopy  Following the final wash, the cells attached to the coverslips were fixed by incubating them in 3% paraformaldehyde solution, pH 7.4 (3% paraformaldehyde (Sigma-Aldrich Canada), 100 pM CaCl2, 100 pM MgCl 2 in D-PBS) for 20 minutes at room temperature. The cell-coated coverslips were then washed twice with D-PBS and once with distilled water. The coverslips were then mounted cell-side down onto glass coverslides (Fisher Scientific Canada) using mounting media (10% PBS, 87.5% glycerol, 2.5% 1,4diazabicyclo (2.2.2) octane (DABCO) (Sigma-Aldrich Canada)). The coverslips were secured onto the coverslides by sealing the edges with clear nail polish. Images of the cells expressing GFP-tagged proteins were collected using a Bio-Rad Radiance Plus confocal microscope with settings optimized for GFP.  The collected images were  analyzed using the NIH Image version 1.62 program.  2.10  Molecular biology methods  2.10.1 Restriction endonuclease reactions The restriction enzymes used for the experiments were purchased from Roche Diagnostics, Invitrogen Life Technologies Canada, New England Biolabs Canada (Pickering, Ontario) and Promega (Madison, Wisconsin). Restriction enzyme reactions  were performed as suggested by the manufacturers.  The total amount of restriction  enzyme(s) used per reaction was always 10% or less than the total volume of the reaction. The reactions were incubated in a 37°C incubator for a minimum of 2 hours, unless otherwise recommended by the manufacturer). For enzymes that function optimally at temperatures other than 37°C, the reactions were incubated in a water bath at the appropriate temperature. The eppendorf tube lids containing the reactions were sealed with parafilm and then completely submerged in the water for a minimum of 2 hours.  2.10.2 Alkaline phosphatase reactions In instances where vector DNA was to be used for ligation reactions, the restriction enzyme-digested vector was dephosphorylated by treatment with alkaline phosphatase (Roche Diagnostics). 1 pi of alkaline phosphates was used to treat 1 pi of digested vector DNA for 45 minutes at 37°C. The reactions were terminated by adding EDTA to a final concentration of 1 mM and heating the samples to 65°C for 5 minutes. The DNA was then cleaned using the QIAquick Gel Extraction kit (Qiagen, Mississauga, Ontario) according to the manufacturer's instructions.  2.10.3 Agarose gel electrophoresis To separate digested DNA fragments, agarose gels (ranging from 0.5-2.0% agarose (Fisher Scientific Canada)) containing 25 pg/ml ethidium bromide (Invitrogen Life Technologies Canada) were electrophoresed at 100 volts for varying lengths of time depending on the size of the gel and percentage of agarose. The DNA fragments were visualized by UV illumination.  2.10.4 Gel purification of digested DNA Following restriction enzyme digestion, and separation of the DNA fragments by agarose gel electrophoresis, the appropriate DNA fragment was quickly and carefully excise from the gel using a clean razor blade. The gel pieces containing the DNA were heated to 55°60°C to melt the agarose.  The DNA was then purified using the QIAquick Gel  Extraction kit according to the manufacturer's instructions.  2.10.5 Ligation of purified DNA fragments The Rapid DNA Ligation kit (Roche Diagnostics) was used to perform ligations. The ligations were performed according to the manufacturer's instructions. Briefly, 2 pi of 5X DNA dilution buffer (included in the kit) was mixed with 0.5 pg of purified vector DNA and 3-10 pg of insert DNA. The amount of insert DNA added was dependent on the relative sizes of the insert and vector. Generally, the molar ratio of inset DNA to vector DNA used was 3 or greater. Sterile distilled deionized water was added to the DNA mixture to a final volume of 10 pi. 10 pi of 2X DNA ligation buffer and 1 pi (5 units) of T4 DNA ligase was then added to the DNA mixture and incubated for 30 minutes at room temperature. One-third to one-half of the ligation mix was used to transform competent bacteria.  2.10.6 Transformation of competent Escherichia coli bacteria The competent DH5a and XL 1-Blue E. coli strains were prepared by May Dang-Lawson, the laboratory technician. One hundred to three hundred ng of plasmid DNA or ligation mixture was added to 150 pi of competent bacteria and incubated on ice for 20 minutes. The bacteria were then heat-shocked for exactly 2 minutes at 42°C and subsequently incubated on ice for at least 5 minutes. The entire mixture was then plated onto LB agar plates containing 100 pg/ml ampicillin. Once dry, the plates were incubated upside down at 37°C overnight or until individual bacterial colonies were visible.  2.10.7 Site-directed mutagenesis Site-directed mutagenesis was performed using the QickChange Site-Directed Mutagenesis kit (Stratagene).  The oligonucleotide primers for all the site-directed  mutagenesis and DNA sequencing reactions were created by and purchased from the Nucleic Acid Protein Service Unit (UBC). The site-directed mutagenesis reactions were performed according to the manufacturer's instructions.  Briefly, the 5' and 3'  oligonucleotide primers and double-stranded DNA template (purified from DH5a or XL 1-Blue dam+ E. coli strains) were mixed with 10X reaction buffer, a mixture of deoxynucleotide triphosphates (dNTPs) and distilled, deionized water. 1 pi (2.5 units) of Pfu Turbo DNA polymerase was added to the mixture and combined thoroughly. The  entire mixture was overlaid with sterilized light mineral oil (Fisher Scientific Canada). These initial steps were performed in a sterilized biosafety cabinet to prevent contamination. The samples were then subjected to polymerase chain reaction (PCR) in a thermocylcer (PerkinElmer Canada, Woodbridge, Ontario) under conditions recommended by the manufacturer. Following the PCR reaction, 1 pi (10 units) of Dpn I restriction enzyme was added to the reaction mixture and incubated for 1 hour at 37°C. Dpn I specifically cleaves methylated DNA produced in most dam+ E. coli strains. Since the parental, unmutated DNA is dam methylated, it will be susceptible to Dpn I digestion while the PCR-derived mutated plasmids remain intact. 1 pi of the digested reaction mixture was subsequently used to transform XL 1-Blue Supercompetent bacteria, as described previously.  Once DNA was isolated from single bacterial colonies,  confirmation of the mutation was verified by restriction enzyme digestion and DNA sequencing (performed by Lone Star Laboratories, Houston, Texas).  2.10.8 Small scale preparation of DNA The QIAprep Spin Miniprep kit (Qiagen) was used to isolate smaller amounts of plasmid DNA from bacteria. Briefly, individual bacteria colonies were inoculated into 3 ml of LB liquid culture containing 100 pg/ml ampicillin and incubated overnight at 37°C with shaking.  After 16 hours, the bacteria were collected in an eppendorf tube by  centrifugation at 8000 rpm for 15 seconds. Following the removal of the supernatant by aspiration, the plasmid DNA was isolated using the kit according to the manufacturer's instructions. The purified DNA was quantified using the Ultrospec 2100 pro UV/visible spectrophotometer and stored at -20°C until ready for use.  2.10.9 Large scale preparation of DNA Larger amounts of plasmid DNA was isolated using either the Qiagen Plasmid Maxi kit (Qiagen) or the Concert High Purity. Maxiprep kit (Invitrogen Life Technologies Canada). Briefly, an individual bacterial colony was inoculated into 50 ml of LB liquid culture containing 100 pg/ml ampicillin and incubated at 37°C for 9-10 hours while shaking. The entire culture was then added to 600 ml of LB liquid culture containing 100 pg/ml ampicillin and incubated for at least 20 hours at 37°C while shaking. Following the 20 hour incubation, the bacteria was collected by centrifugation at 8000 rpm for 5  minutes. The supernatant was removed and the plasmid DNA was isolated according to the manufacturer's instructions with one additional step to ensure the purity of the DNA. Following bacterial lysis and centrifugation of the bacterial membranes, the lysate containing the plasmid DNA was filtered through filter or Whatman paper to ensure that the DNA was not contaminated by proteins. The purified DNA was quantified using a spectrophotometer and subsequently stored at -20°C. Generally, 500 pg DNA was obtained from the Qiagen Plasmid Maxi kit and 1 mg of DNA was isolated using the Concert High Purity Maxiprep kit. 2.11  Plasmids generated  pMSCV-Lyn The DNA encoding the murine Lyn gene was excised from the RSVpLpA-Lyn with BamH I and ligated into the Bgl II site of the pMSCV-puro  plasmid  retroviral mammalian  expression vector, since BamH I and Bgl II have compatible cohesive ends (Figure 2.1). The orientation of the Lyn insert was determined by restriction enzyme digests and DNA sequencing.  pMSCV-Blk (cytosolic mutant) The DNA encoding the murine Blk gene containing mutations within the amino-terminal (the second and third amino acids were mutated) was cut out of the RSV pLpA-Blk (cytosolic mutant) plasmid with Xba I and cloned into the Xba I site of pBluescript to create pBluescript-Blk  (cytosolic mutant).  The correct orientation of the mutant Blk  gene was determined by restriction enzyme digestion. subsequently excised from pBluescript-Blk  KSII  The mutant Blk gene was  (cytosolic mutant) by digestion with Xho I  (which is located upstream of the Xba I site at the 5' end of the gene) and with BssH II (which is located downstream of the Xba I site at the 3' end of the gene), refer to Figure 2.2. The pMSCV-CD 16/CD7 was digested with Xho I and Mlu I to excise the DNA encoding CD16/CD7, leaving the pMSCV vector. The excised Blk (cytosolic mutant) gene fragment was then cloned into the Xho I and Mlu I (BssH II and Mlu I have compatible cohesive ends) of the digested pMSCV wector, thereby producing pMSCV-Blk (cytosolic mutant)..  Figure 2.1 Strategy for generating the pMSCV-Lyn text for details.  expression vector. Refer to the  Digest with Bam HI  Alw l/Bst Yl/Dpn II  Digest with Bgl II Bgl II  Ligation V  Alw l/Bst Yl/Dpn II  Figure 2.2 Strategy for generating the pMSCV-Blk vector. Refer to the text for details.  (cytosolic mutant) expression  Xba  -Xbal >  Digest with Xba I  Xho I  oc  oo Digested with Xho I and Mlu I  Xho!_  Mlu I  Xba U  Xho /.  Xba I J  XbaI |  BssHII  ^ut^al-BJ^en^s. Xbal  Ligation Digested with Xba I Xba I  f  Xho I —77  _\  Xba I  V r - BssHII  Digested with Xho I and BssHII  Ligation Xbal ~Bst Ul/Hha I  RSV pLpA-Blk (wild type) To correct the amino acids at positions 2 and 3 of the RSV pLpA-Blk  (cytosolic  mutant)  and generate a wild type Blk construct, site-directed mutagenesis was performed on the RSV pLpA-Blk  (cytosolic  mutant)  construct.  The WT BLK 5' and WT BLK 3'  oligonucleotide primers (the sequences of which are indicated in Table 2.1) were used for the site-directed mutagenesis. This resulted in the creation of the RSV pLpA-Blk  (wild  type) construct, in which the 2 n d and 3 r d amino acids mutated from aspartic acid and proline to glycine and leucine, respectively. The correct sequence of the construct was verified by DNA sequencing.  pMSCV-Blk (wild type) To correct the amino acids at positions 2 and 3 of the pMSCV-Blk  (cytosolic mutant) and  create a wild type Blk construct, site-directed mutagenesis was performed on the pMSCVBlk (cytosolic mutant) plasmid. The WT BLK 5' and WT BLK 3' oligonucleotide primers (the sequences of which are indicated in Table 2.1) were used for the PCR reaction, resulting in the generation of the pMSCV-Blk  (wild type) construct. The 2 n d and  3 r d amino acids in this construct were mutated from aspartic acid and proline to glycine and leucine, respectively. The correct sequence of the construct was verified by DNA sequencing.  RSV pLpA-palm Blk A murine Blk construct was created in which the 3 r d amino acid (leucine) was mutated into cysteine to create a palmitylation modification site. To generate this construct, sitedirected mutagenesis was performed on the RSV pLpA-Blk (wild type) plasmid. The PM BLK 5' and PM BLK 3' oligonucleotide primers (the sequences of which are indicated in Table 2.1) were used for the PCR reaction. The site-directed mutagenesis resulted in the generation of a construct containing DNA encoding for palmitylated Blk (RSVpLpApalm Blk), as verified by restriction enzyme digests and DNA sequencing.  RSV pLpA-non-palm Lyn A murine construct was created in which the 3 r d amino acid (cysteine) was mutated into leucine to eliminate the palmitylation site. This construct was created by performing sitedirected mutagenesis on the RSVpLpA-Lyn  plasmid.  The NP LYN5' and NP LYN 3'  oligonucleotide primers, the sequences of which are indicated in Table 2.1, were used for the site-directed mutagenesis reaction. The site-directed mutagenesis resulted in the generation of the RSV pLpA-non-palm  Lyn construct which contained the gene encoding  for Lyn in which the 3 r d amino acid was concerted into leucine, as- verified by DNA sequencing.  RSVpLpA-kinase dead Lyn A construct containing a catalytically inactive kinase dead form of murine Lyn was generated in which the lysine at position 275 was mutated into alanine. This lysine residue has been shown to be essential for Src kinase activity. This plasmid was created by performing site-directed mutagenesis on the RSV pLpA-Lyn vector. The KD LYN 5' and KD LYN 3' oligonucleotide primers were used for the PCR reaction (Table 2.1). The site-directed mutagenesis resulted in the generation of the RSV pLpA-kinase Lyn construct.  dead  Table 2.1: Oligonucleotide primers used for site-directed mutagenesis reactions. The underlined nucleotides were targeted for mutagenesis and the original nucleotides are shown below the primer sequences. The resulting amino acid changes are also indicated, number indicate the position of the amino acids. Primer name WT BLK 5'  Primer sequence (5'—>3') CTA GAG CAA ATG GGG CTG CTG AGC AGC AAG AGG  AT WT BLK 3'  C  CCT CTT GCT GCT CAG CAG CCC CAT TTG CTC TAG  G PM BLK 5'  Aspartic acid2—>Glycine2 Proline3—>Leucine3  AT  CTA GAG CAA ATG GGG TGT CTG AGC AGC AAG AGG Leucine3—>Cysteine3  CTG PM BLK 3'  Resulting amino acid change(s)  CCT CTT GCT GCT CAG ACA CCC CAT TTG CTC TAG  CAG NP LYN 5'  CCA CCA CGA GCG AGA AAT ATG GGA CTT ATT AAA TCA AAA AGG  TG NP LYN 3'  CCT TTT TGA TTT AAT AAG TCC CAT ATT TCT CGC TCG TGG TGG  KD LYN 5'  GCA CAA AGG TGG CTG TGG CGA CCC TCA AGC TCG GC AA GCC GAG CTT GAG GGT CGC CAC AGC CAC CTT TGT GC  Cysteine3—>Leucine3  CA  KD LYN 3'  IT  Lysine275 —> Alanine275  CHAPTER 3 Syk and Src family protein tyrosine kinases differentially regulate the PI 3-kinase/Akt pathway  3.1  Introduction  Cross-linking of the BCRs at the cell surface by pathogen-derived antigens results in the activation of multiple downstream signaling pathways regulated by four key enzymes — PLCy-2, MAP kinase, SAP kinases and PI 3-kinase. The activation of these signaling enzymes through the BCR requires multiple steps that are initiated by several PTKs. These PTKs include Syk and various Src family members.  Although there is considerable  information available regarding the signaling enzymes activated by the BCR, less is known about the specific mechanisms by which the BCR activates or regulates these pathways. In particular, it is not clearly understood how the expression of multiple Src kinases regulates the activation of the downstream signaling substrates or if multiple Src kinases are required for the activation of individual pathways. In addition, it is unclear if there is a specific relationship between an individual Src kinase family member and the regulation of an individual signaling pathway. Single Src kinase_knockout mice have not provided further information since Src kinase knockout mice are generally viable with intact immune systems, indicating that Src family members have some redundant functions (reviewed by Thomas and Brugge, 1997; Korade-Mirnics and Corey, 2000). Interestingly, viable mice in which two or three Src kinases are deleted show fascinating phenotypes (refer to Chapter 1.6.5).  However, there are still other Src kinases present in these animals that are  compensating for the lack of the deleted family members. Although the function of some Src kinases appears to be replaceable by other family members, reasons must exist as to why there are so many different Src kinases within individual organisms and within different cell types. The fact that multiple family members are expressed in the same cells raises an important question, do the Src kinases have overlapping functions or does each Src kinase play a unique role that cannot be performed by any other family member?  While a great deal of information is available on the various BCR-activated signaling pathways, less is known about how these signaling pathways are regulated by the BCRassociated Syk and Src family PTKs. This project is aimed at comparing the roles of the lymphoid-expressed Src kinases, Lyn, Lck and Blk in BCR signaling, focusing specifically on their influence on the PI 3-kinase/Akt pathway. To this end, I have adapted a nonlymphoid cell expression system using AtT20 endocrine cells that have been previously transfected with the BCR (Matsuuchi et al., 1992) and previously used to study the role of Syk in BCR signaling (Richards et al., 1996). In particular, the BCR-mediated regulation of the PI3-kinase/Akt pathway by the lymphoid-expressed Src kinases was of interest. It was hypothesized that the various Src family kinases would play different roles in linking the BCR to PI3-kinase/Akt pathway activation.  Although very similar in structure, there are some differences between the various Src kinases involved in this study (Lyn, Lck, Blk and Fyn) that suggests that each family member plays a unique role that cannot be performed by other family members. These differences include their ability to associate with different signaling components through their various protein interaction domains (reviewed by Thomas and Brugge, 1997). These include the amino-terminal unique region, SH2 and SH3 domains. Another difference between the Src kinases, which may influence their distinct roles, is their ability to become localized to different domains within the plasma membrane called lipid rafts.  Their  localization inside or outside of lipid rafts may dictate which signaling molecules they can interact with and regulate.  Experiments to determine if Src kinases have different functions in BCR-mediated signaling would be difficult to perform in lymphoid cell lines since B cells generally express several Src kinase family members (Figure 3.6). In addition, it would be difficult for our laboratory to perform these experiments in knockout mice that only express a single lymphoid-expressed Src kinase. Only a few viable double and triple knockout mice lacking different combinations of Src kinases have been created. Therefore, due to the overlapping functions of Src kinases, creating double and triple knockout mice is possible. However, since there are generally multiple Src family kinases in B cells (other than Lyn, Fyn and Blk), generation of viable mice with multiple knockouts would be necessary to  study the individual roles of the different family members. Knocking out several Src kinase genes that are expressed in the whole animal is a daunting and probably impossible task. The AtT20 cell system was used to test the hypothesis that Src kinases play different roles in regulating BCR-mediated Akt activation. The individual Src kinases were transfected into AtT20 cells and their function assessed following BCR cross-linking.  It was  previously established that the endogenously expressed Src family kinase, Fyn, was sufficient to promote the signaling by the PI3-kinase pathway (Gold et al., 1991). Later, as part of this thesis project (refer to Chapter 3.2.1; Gold et al., 1999), it was shown that BCRmediated activation of Akt, as monitored by changes in the levels of Akt phosphorylation was supported by Fyn alone. Syk kinase activity was found to be required for sustained and enhanced Akt phosphorylation. The three lymphoid-expressed Src kinases (Lyn, Lck and Blk) appeared to play different roles in the regulation of Akt following BCR engagement. Interestingly, addition of these Src kinases to AtT20 cells without Syk was not able to enhance Akt phosphorylation, indicating that Lyn, Lck and Blk could not substitute for Syk function. Since Syk interacts with Src kinases in B cells, their joint ability to regulate BCR-mediated Akt activation was also examined in the AtT20 cells. In the presence of Syk, two of the Src kinases, Lyn and Blk, had inhibitory effects on BCRinduced Akt phosphorylation and activation while Lck did not change the extent of this phosphorylation in response to BCR cross-linking. These results supply evidence that Lyn, Blk and Lck play different roles in regulating the PI 3-kinase/Akt pathway and that this regulation is further modified by the presence of Syk.  3.2  Results  3.2.1  Fyn is sufficient to activate Akt and Syk is required for sustained Akt activation  BCR engagement has been shown to lead to activation of the lipid kinase, PI 3-kinase (Gold et al., 1992a; Gold and Aebersold, 1994). PI 3-kinase plays multiple roles in BCR signaling by producing PI-3,4,5-P3 within the plasma membrane. The production of PI3,4,5-P3 regulates the activity and subcellular localization of several important PH domaincontaining signaling components including Btk, SOS, vav and Akt. Although it is wellknown that BCR cross-linking results in PI 3-kinase activation, it had not been established if BCR engagement also leads to Akt activation. Since PI 3-kinase is activated by BCR cross-linking, then Akt should also become activated downstream of the BCR. This was tested in lymphocytes and by using the BCR-expressing AtT20 cell system (Astoul et al., 1999; Craxton et al., 1999; Gold et al., 1999; Li et al., 1999). The study using BCRexpressing AtT20 cell system is presented below as part of this thesis.  As mentioned earlier, Akt phosphorylation on two amino acid residues, threonine 308 and serine 473, could be used indirectly to assay for PI 3-kinase. In addition, the levels of Akt phosphorylation on these two residues could also be used as an indication of the extent of Akt activation. The phosphorylation state of Akt in the various cell lines was determined first by cross-linking the BCRs on the cell surface for varying periods of time. The cell lysates were then analyzed by SDS-PAGE and Western immunoblotted for Akt that is specifically phosphorylated on the serine 473 or threonine 308 residues, using commercially available phospho-specific antibodies.  BCR-expressing AtT20 cells (AtT20 BCR+ Fyn+) cells were examined for the ability of the BCR working with the endogenous Fyn to activate Akt. The BCRs on these cells were cross-linked for varying periods of time and the levels of serine 473 and threonine 308 phosphorylated Akt was examined. A transient increase in Akt phosphorylation on the serine 473 residue was observed (Figure 3.1 top panel). This phosphorylation began by 3 minutes, peaked at 15-30 minutes before it decreased following 60 minutes of receptor  cross-linking. The pattern of phosphorylation on the threonine 308 residue showed a similar rise and fall of phosphorylation, though the antibody did not work as well (Figure 3.1 middle panel).  Thus, BCR cross-linking can promote Akt phosphorylation and  activation in the AtT20 BCR+ Fyn+ cells (Gold et al., 1999). This data is consistent with the idea that Fyn (the only Src kinase family member present in AtT20 cells) is able to perform these functions. These results are consistent with the previous observation that Fyn alone is sufficient for signaling by the PI 3-kinase pathway, as monitored by the production of PI-3,4,5-P3 in the AtT20 BCR+ Fyn+ cell line (Matsuuchi et al., 1992). Akt phosphorylation was also examined in AtT20 Fyn+ cells, which did not express the BCR, as a negative control. Low levels of Akt phosphorylation, which did not incresae in response to BCR cross-linking, was observed in these cells (data not shown).  To examine the role of Syk in BCR-induced signaling, murine Syk PTK was cloned and transfected into the AtT20 BCR+ Fyn+ cells, creating the cell lines referred to as AtT20 BCR+ Fyn+ Syk+ (Richards et al., 1996). This cell line was used in my study to examine the role of Syk in BCR-induced Akt phosphorylation and activation. One specific clone (#13), which will hereafter be referred to as AtT20 BCR+ Fyn+ Syk+, from the transfection was used for all the subsequent experiments, unless otherwise stated. Akt activation in response to BCR cross-linking was enhanced in this clone compared to cells  Figure 3.1 BCR cross-iinking in the AtT20 BCR+ Fyn+ cells can promote phosphorylation of Akt. AtT20 BCR+ Fyn+ cells were serum-starved for 16 hours in media containing 0.2% FCS before being stimulated with 20 pg/ml of anti-mouse IgM antibodies for the indicated time periods. Following cell lysis, total cell extracts were separated by SDS-PAGE and analyzed by Western blotting. Filters were probed with anti-phospho-serine 473 Akt (top), anti-phospho-threonine 308 Akt (middle) and anti-Akt (bottom) antibodies. The bands were visualized by enhanced chemiluminescence. The molecular weight standards (in kDa) are indicated to the left.  These results are  representative of a minimum of four independent experiments with similar results. These results were published in Gold et al. (1999).  AtT20 BCR+ Fyn+ Anti-IgM (min)  IB: anti-phosphoserine 473 Akt  65-  IB: anti-phosphothreonine 308 Akt  65-  IB: anti-Akt  65-  0  3  5  -  15 30 60 <  p-serine 473 Akt  •<  p-threonine 308 Akt  <  Akt  Figure 3.1  not expressing Syk. In contrast to the untransfected AtT20 BCR+ Fyn+ parental cell line, phosphorylation of Akt on the serine 473 residue in AtT20 BCR+ Fyn+ Syk+ cells was faster, reached almost maximum levels after a shorter period of BCR cross-linking and was more sustained (Figure 3.2 A).  The extent of phosphorylation on this residue  remained high for at least 60 minutes of receptor activation.  The pattern of  phosphorylation of the threonine 308 residue was similar to that of the serine 473 residue in that the phosphorylation remained at the maximum levels for 60 minutes of BCR activation. Phosphorylation of the serine 473 residue remained at these high levels even after 90 minutes of receptor activation (personal observation, data not shown). BCRinduced Akt phosphorylation was also examined in another clone, AtT20 BCR+ Fyn+ Syk+ [clone #41] which expressed slightly lower levels of Syk compared to clone #13 (Richards et al., 1996). Akt phosphorylation in this clone mirrored what was observed for clone #13 (Figure 3.2 B).  To determine if the observed induction of Akt phosphorylation following BCR crosslinking in the AtT20 transfectants is dependent on PI 3-kinase activity, the cells were treated with the PI 3-kinase-inhibitors wortmannin and LY294002, prior to and during receptor stimulation. No Akt phosphorylation was observed in the AtT20 BCR+ Fyn+ or AtT20 BCR+ Fyn+ Syk+ cell lines prior to and following BCR stimulation (Figure 3.3). Thus, BCR-induced Akt phosphorylation in the AtT20 transfected cells is dependent on PI 3-kinase activity.  Taken together, these results indicate that one Src family kinase, in this case the endogenous Fyn homolog, is sufficient to promote BCR-induced Akt phosphorylation and hence activation. However, Syk PTK is required to further amplify and sustain this response. This is consistent with the idea that Syk is an upstream regulator of the PI 3kinase pathway downstream of the BCR. In addition, the patterns of Akt phosphorylation on the serine 473 and threonine 308 residues was similar in the AtT20 BCR+ Fyn+ and AtT20 BCR+ Fyn+ Syk+ cell lines examined, suggesting that PDK1 and putative PDK2 phosphorylate and regulate Akt to the same extent in response to BCR activation.  Figure 3.2 Syk PTK is required for amplifying and sustaining BCR-induced Akt phosphorylation in the AtT20 cell system. AtT20 BCR+ Fyn+, AtT20 BCR+ Fyn+ Syk+ [clone #13] (A) and AtT20 BCR+ Fyn+ Syk+ [clone #41] (B) cells were serumstarved for 16 hours in media containing 0.2% FCS before being stimulated with 20 pg/ml of anti-mouse IgM antibodies for the indicated time periods. Following cell lysis, the total cell extracts were separated by SDS-PAGE and analyzed by Western blotting. Filters were probed with anti-phospho-serine 473 Akt, anti-phospho-threonine 308 and/or anti-Akt  (bottom)  antibodies.  The bands were visualized  by  enhanced  chemiluminescence. The molecular weight standards (in kDa) are indicated to the left. These results are representative of a minimum of four independent experiments with similar results. These results were published in Gold et al. (1999).  (A)  AtT20 B C R + Fyn+ Syk+[#13]  AtT20 B C R + Fyn+  IB: anti-phosphoserine 473 Akt  (min) 55.  0  3  5 15 30 601 0  3 5 15 30 60 -p-serine 473 Akt  IB: anti-phosphothreonine 308 Akt 65-  IB: anti-Akt  -p-threonine 308 Akt  -Akt  65 -  (B)  AtT20 B C R + Fyn+ S y k + [#41] Anti-IgM  (min)  IB: anti-phosphoserine 473 Akt  IB: anti-Akt  6 5  65  0  3  5  15 30 60  p-serine 473 Akt  •Akt  Figure 3.2  Figure 3.3 BCR-induced Akt phosphorylation in the AtT20 cells is PI 3-kinasedependent. AtT20 BCR+ Fyn+ and AtT20 BCR+ Fyn+ Syk+ were serum-starved for 16 hours in media containing 0.2% FCS before incubating them in buffer alone, 25 pM LY294002 (top panel) or 30 nM wortmannin (bottom panel) for 20 minutes prior to cell stimulation. The cells were then stimulated with 20 pg/ml of anti-mouse IgM antibodies for 15 minutes. Following cell lysis, 30-35 pg of total cell extract was separated by SDSPAGE and analyzed by Western blotting. Filters were probed with anti-phospho-serine 473 Akt. The bands were visualized by enhanced chemiluminescence. The molecular weight standards (in kDa) are indicated to the left. These results are representative of two experiments.  LY294002 (25 i M ) Anti-IgM (15 min)  IB: anti-phosphoserine 473 Akt  AtT20 BCR+ Fyn+  AtT20 B C R + F y n +  T +T  T I -  +  l -  \  65 -  AtT20 BCR+ Fyn+ Wortmannin (30 nM) Anti-IgM (15 min)  IB: anti-phosphoserine 473 Akt  65 -  AtT20 B C R + F y n +  3.2.2  Syk kinase activity is required for promoting and sustaining Akt phosphorylation and activation  Since Syk was important for amplifying  and sustaining BCR-induced Akt  phosphorylation and activation, the necessity of its tyrosine kinase activity in mediating this response in transfected AtT20 cells was further analyzed. Two transfected AtT20 cell lines in which Syk kinase activity was ablated were examined for their ability to promote and sustain BCR-induced response.  The AtT20 BCR+ Fyn+ KD Syk+ [clone #17] cell line was created by transfecting a plasmid encoding for kinase dead (KD) catalytically inactive mutant form of Syk into AtT20 BCR+ Fyn+ cells, as previously described (Richards et al., 1996). This particular clone expressed as much kinase dead Syk as wild type Syk in the AtT20 BCR+ Fyn+ Syk+ cell line (Richards et al., 1996). An in vitro kinase assay confirmed that the mutant Syk in these cells was catalytically inactive (Richards et al., 1996). Indeed, the induction of tyrosine phosphorylation of the total cellular protein following BCR stimulation was attenuated in these cells due to the lack of Syk kinase activity (Figure 3.4 A). There was also a marked decrease in BCR-induced Akt phosphorylation in these cells compared to cells expressing catalytically active Syk (Figure 3.4 B). There was a small increase in Akt phosphorylation in the AtT20 BCR+ Fyn+ KD Syk+ cells observed at 15 minutes of BCR cross-linking which decreased following longer periods of receptor activation. This phosphorylation was most likely due to the presence of the endogenous Fyn homolog within the cells.  This data suggests that the kinase activity of Syk is required for  amplifying and sustaining BCR-induced Akt phosphorylation.  BCR-induced Akt phosphorylation was examined in another AtT20 cell line that was cotransfected with Syk and hyper-expressed the SHP-1 protein tyrosine phosphatase (generated by Tally Vertinsky, a former undergraduate student). The AtT20 BCR+ Fyn+ Syk+ SHP-1+ cell line [clone # 10] expressed slightly less Syk compared to the AtT20 BCR+ Fyn+ Syk+ cell line (Figure 3.5 A, left panel).  Since AtT20 cells express  endogenous SHP-1, clone #10 was chosen for further study due to the fact that it hyper-  expressed SHP-1. Clone #10 expressed approximately twice as much SHP-1 as the parental cell line that was originally transfected (Figure 3.5 A, right panel). It was of interest to examine BCR-induced Akt phosphorylation and activation in these cells because SHP-1 has been shown to be a negative regulator of BCR signaling. Specifically, SHP-1 downregulates the activation of several PTKs including Syk and Src family kinases (Dustin et al., 1999; Somani et al., 2001).  Since SHP-1 was hyper-  expressed in the AtT20 BCR+ Fyn+ Syk+ SHP-1+ cells, it was predicted that BCR signaling would be diminished even in the presence of Syk. Indeed, the induction of tyrosine phosphorylation of total cellular protein following various lengths of BCR stimulation was noticeably lower compared to cells only transfected with Syk (AtT20 BCR+ Fyn+ Syk+ cells) (Figure 3.5 B). This attenuation of BCR-induced tyrosine phosphorylation may be due to the inhibition of Syk kinase activity by the hyperexpressed SHP-1 within the cells. Since Syk kinase activity is probably inhibited in the AtT20 BCR+ Fyn+ Syk+ SHP-1+ cells, it was hypothesized that Akt activation following BCR cross-linking would also be decreased in these cells. As predicted, BCRinduced Akt phosphorylation on the serine 473 residue was considerably lower in the AtT20 BCR+ Fyn+ Syk+ SHP-1+ cells compared to the AtT20 BCR+ Fyn+ Syk+ cells, which contained half as much SHP-1 (Figure 3.5 C). Thus, hyper-expressing SHP-1 in AtT20 cells that co-express Syk results in the inhibition of Akt phosphorylation in response to BCR engagement, as was observed in cells that expressed the catalytically inactive mutant form of Syk (Figure 3.4 B).  Interestingly, Akt phosphorylation in the AtT20 BCR+ Fyn+ KD Syk+ and AtT20 BCR+ Fyn+ Syk+ SHP-1+ cells appeared to be lower than in the AtT20 BCR+ Fyn+ cells (compared Figures 3.4 B and 3.5 C with 3.2 A). It is possible that the kinase dead Syk has a dominant negative effect within the cells. It could interact with its downstream targets and prevent them from becoming activated by other kinases such as the endogenous Fyn. Additionally, the hyper-expressed SHP-1 also had a similar effect on Akt phosphorylation. In these cells, a SHP-1 could be inhibiting Akt signaling by dephosphorylating substrates that lead to decreased PI 3-kinase signaling.  Figure 3.4 Syk kinase activity is required for BCR-induced Akt phosphorylation in AtT20 cells. AtT20 BCR+ Fyn+ Syk+ and AtT20 BCR+ Fyn+ KD Syk+ [clone #17] cells were serum-starved for 16 hours in media containing 0.2% FCS before being stimulated with 20 pg/ml of anti-mouse IgM antibodies for the indicated time periods. Following cell lysis, the total cell extract was separated by SDS-PAGE and analyzed by Western blotting. The filter were probed with (A) anti-phosphotyrosine monoclonal antibody (4G10) and (B) anti-phospho-serine 473 Akt. The bands were visualized by enhanced chemiluminescence.  The location of Syk and Fyn are indicated.  molecular weight standards (in kDa) are indicated to the left. representative of two experiments.  The  These results are  (A)  AtT20 BCR+ Fyn+ Syk+ Anti-IgM (min; 1  0 3  85-  605040-  25-  20-  3 5 15 30 60 I  mi  185 115-  5 15 30 60 lo  AtT20 BCR+ Fyn+ KD Syk+  « ii | I II 1  WW  *  H '*"*' ISt "tff  igvf fsf  w m  m  •<  -Wild type/kinase dead Syk  -<  -Fyn  ft  -  m  m  *  HI MM  IB: anti-phosphotyrosine (4G10)  (B) AtT20 BCR+ Fyn+ Syk+ Anti-IgM (min)  IB: anti-phosphoserine 473 Akt  65-  1 0  3  5 15 30 60  AtT20 BCR+ Fyn+ KD Syk+ 0 3 5 15 30 60 —  p-Akt  Figure 3.4  Figure 3.5 Hyper-expression of the SHP-1 protein tyrosine phosphatase in Sykexpressing AtT20 cells inhibits BCR-induced Akt phosphorylation. (A) Left panel: Relative levels of Syk PTK and SHP-1 in AtT20 BCR+ Fyn+ Syk+ and AtT20 BCR+ Fyn+ Syk+ SHP-1+. Following cell lysis, 20 pg of total cell extract was separated by SDS-PAGE and analyzed by Western blotting with anti-Syk antibodies. Right panel: To determine how much more SHP-1 was present in the transfected cells, varying amounts of total cell extract from AtT20 BCR+ Fyn+ Syk+ SHP-1+ cells was compared to 30 pg from AtT20 BCR+ Fyn+ Syk+ cells. AtT20 BCR+ Fyn+ Syk+ and AtT20 BCR+ Fyn+ Syk+ SHP-1+ cells were serum-starved for 16 hours in media containing 0.2% FCS before being stimulated with 20 pg/ml of anti-mouse IgM antibodies for the indicated time periods. Following cell lysis, the total cell extract was separated by SDS-PAGE and analyzed by Western blotting with (B) anti- phosphotyrosine monoclonal antibody (4G10) and (C) anti-phospho-serine 473 Akt. The bands were visualized by enhanced chemiluminescence. The location of Syk and Fyn are indicated. The molecular weight standards (in kDa) are indicated to the left. These results are representative of a two independent experiments with similar results.  W 9? 5 D) 3.  o  CD T' -si G O ^O > T 3 7T zr ~ o cn Cn  if  \  o  1 1 g1  [  (I  o  1  V  o CO < ro U1 > > o + CD > / O cn C X T> T J + CO A TI a oo> +  ro 0) • "a oo Tczr 3 •o <ri — o cn  Z3 CD ^ O  o  1  »•  t  t ip  9  t  OI  o  -t  s i K -  •v  1  00 cn  m  I  r t  1  t  1  X  1  m  Syk  CD  •a • > 5  I S I I  Fyn  CD C -5  m  -<  1 i  ro CO o OI (/> CD •< o U1 n CO < + o cn o  •P-  to to O OI  O) o  0 0 cn AtT20 BCR+ Fyn+ AtT20 BCR+ Fyn-i- Syk+  o  AtT20 BCR+ Fyn+ Syk+ SHP-1 +  CO  o c n K (/> 00 ° e r cn -n g r co < 0 + 0 o1 co cn cn co o CD  o  << N> > O + ro  </> o X J3 TJ + I  -k 11 +  con  co cn AtT20 BCR+ Fyn+ (30 /vg)  3.2.3  The Src kinases, Lyn, Lck and Blk, differentially regulate BCR-mediated Akt phosphorylation and activation  It was previously shown that multiple Src family kinases are expressed at the mRNA level in B lymphoma cells lines (Law et al., 1992). The presence and expression levels of the four main lymphoid-expressed Src kinase family members that are the focus of this study were examined in several B lymphoma cell lines used to study BCR signal transduction (Figure 3.6). The p56 and p53 isoforms of Lyn were expressed in mouse splenic B cells and in all of the B lymphoma cell lines examined (Figure 3.6 A). This was expected since this family member plays an important role in the initiation of signaling through the BCR. Lyn was most abundant in the K40-B1 pro-B and WEHI 231 immature B lymphoma lines. Lower levels of Lyn were found in the more mature B lymphoma lines, Ball 7, A20 and 2PK3. Like Lyn, Fyn was also expressed in all of the B lymphoma cell lines examined (Figure 3.6 B). This p59 isoform of Fyn expressed in the B lymphoma cell lines and in the AtT20 cells (Richards et al., 1996) is distinct from the isoform found only in T cells, Fyn(T), which plays an important role in signaling by the TCR. Another Src kinase that is important for the initial stages of TCR signaling, Lck, was also expressed in the B lymphoma cell lines examined. Lck was expressed in most of the B lymphoma lines except in the mature A20, K40-B1 pro-B and K40-B2 pre-B cells (Figure 3.6 C). It was difficult to determine if Lck was expressed in the mature 2PK3 B lymphoma lines since the y heavy chain was the exact molecular weight as Lck. It was previously shown that Lck was expressed in multiple transformed B lymphoma cell lines (Von Knethen et al., 1997). Interestingly, Lck was also found in B cells from patients with various leukemias and lymphomas (Majolini et al., 1998). However, it was not expressed in normal splenic B cells suggesting that Lck may contribute to the formation and maintenance of transformed B cell lymphomas and leukemias (Majolini et al., 1998; Figure 3.6 C). Like the other Src kinases examined, the B cell-specific Src kinase Blk was also expressed in mouse splenic B cells and in the majority of the B lymphoma cell lines (Figure 3.6 D). The only cell lines that did not express detectable levels of Blk were the K40-B1 pro-B and the Ball7 cells. High levels of Blk were found  Figure 3.6  Multiple Src kinase family members are expressed in various B  lymphoma cell lines and mouse splenic B cells. Whole cell lysates from the various B lymphoma cells lines indicated (40-50 pg) and mouse splenic B cells (15 pg) lysate was separated by SDS-PAGE and Western immunoblotted with (A) anti-Lyn, (B) anti-Fyn, (C) anti-Lck, (D) anti-Blk or (E) anti-Syk antibodies. The bands were visualized by enhanced chemiluminescence. The molecular weight standards (in kDa) are indicated to the left.  The expression of the Src kinases in AtT20 cells has been previously  characterized by Richards et al. (1996).  Zl\  m  D  CD  O  •  '  65 hi 0)  65  65 0)  1—>•  1—H r0 7T  <•7T oo o  cn o  cd o  cn o  CD o  '  65 oi D <—tTI D  cn o  > 65 03 <—tr-  <  13  cn o  K40-B1 K40-B2 WEHI 231  J it  CH31 CH12 Ball 7  I  t cn  •XT  T  TOIJ 3 c7T n  (  £  A20  M  2PK3  ((  BJAB  ffi (ii  t "O  splenic B cells  CO  TI  3  Ramos AtT20  01  ><  Daudi  A  1OI3 T3 OI W o>  in the immature B lymphoma cells, WEHI 23.1 and CH31. Since Blk is the only B cellspecific Src kinase, it is not surprising that it ws expressed in most of the B lymphoma cell lines tested. Blk expression in the human B lymphoma cell lines could not be determined because the antibodies used did not recognize the human isoform of Blk. Finally, the expression of Syk was examined in the various cell lines (Figure 3.6 D). Syk was highly expressed in all the mouse B lymphoma lines and in the mouse splenic B cells, as expected since it plays a critical role in initiating B cell signaling. Studying the role of individual Src kinases in BCR-mediated signaling would be difficult in many of the B lymphoma cell lines available since they express multiple family members. B lymphocytes may also express the Src family members Hck, Src, Yes and Fgr. However, expression of these family members in different B lymphoma lines and in normal B lymphocytes has not been established and was not examined in this study. The BCRexpressing AtT20 cell line was well-suited for this study since it endogenously expresses only one Src family member, Fyn (Figure 3.6; Richards et al, 1996). To examine the roles of three Src kinases, Lyn, Lck and Blk, in BCR-mediated signaling, plasmid DNA containing the genes encoding for the murine forms of these family members was introduced into AtT20 BCR+ Fyn+ cells.  The plasmid containing the p56 isoform of the murine Lyn gene (RSV pLpA-Lyn)  was  introduced into the cells using the calcium phosphate transfection method. Three clones that stably expressed Lyn were obtained from this transfection, AtT20 BCR+ Fyn+ Lyn+ [clone #21], [clone #33] and [clone #52], AtT20 BCR+ Fyn+ Lyn+ [clone #21] and [clone #33] expressed high amounts of Lyn while [clone # 52] expressed lower levels (Figure 3.7 A, left panel; Appendix Table 1). To create cells expressing Blk, the AtT20 BCR+ Fyn+ cells were infected with retroviruses containing the expression vector for the murine form of Blk (pMSCV-Blk).  The 28 Blk positive clones that were obtained from  the infection expressed varying amounts of Blk. Of all the positive clones, three that expressed high (clone #18), moderate (clone #41) and lower (clone #27) levels of Blk were chosen for further experimentation (Figure 3.7 A middle panel; Appendix Table 1). Lck-expressing AtT20 BCR+ Fyn+ cells were created as previously described (Richards et al, 1996). The "LcklO" cell line from the transfection was further sub-cloned because  Figure 3.7  Expression levels of Src kinases and BCR-induced  tyrosine  phosphorylation of total cellular protein in transfected AtT20 cells. (A) Expression levels of the various Src kinases in the transfected AtT20 cells. Whole cell extract (30-40 pg) from the different cell lines was separated by SDS-PAGE and analyzed by Western blotting with anti-Lyn, anti-Blk or anti-Lck antibodies.  (B) BCR-induced tyrosine  phosphorylation of total cellular protein in Src-transfected AtT20 cells. The different cell lines were stimulated with 20 pg/ml of anti-mouse IgM antibodies for 5 minutes. Following cell lysis, 20 pg of total cell extract was separated by SDS-PAGE and analyzed by Western blotting. The filters were probed with the anti-phosphotyrosine monoclonal  antibody  (4G10).  The bands were visualized  by  enhanced  chemiluminescence. The molecular weight standards (in kDa) are indicated to the left. These results are representative of three experiments.  +  (A)  +  C >.  C >  LL  +  +  DC +  DC  O  O CQ AtT20 BCR+ Fyn+  CQ  O  </) AtT20 BCR+ Fyn+  Lyn+ clones  c  18  LL  "Lyn  AtT20 BCR+ Fyn+ Lck+ sub-clones  Blk+ clones  CVJ +  27  41  60-  IB: anti-Lyn  IB: anti-Blk  +  +  c  (B)  + c  _  LL  +  5  +  DC ^ O oo m o —-• CM + I" £  " m *  g i < -j C o  CM  -  LL  > LL  CE O CD O +  "  IB: anti-Lck  + a  c  >>  LL +  Anti-IgM (5 min)  -Lck  -Blk  50  +  "  +  < CQ  -  +  180mmw  115-  80- mmm 6550 -  « w  -•i n In ni l-  *  *  *«*»«#  wmmmmm  IB: anti-phosphotyrosine (4G10)  — — ZHSrc kinases  40-  25-  2  —. —. — mm—  it appeared to be a mixed clone that did not originate from one individual cell. Many Lck-positive clones were obtained from sub-cloning the "LcklO" cell line. Of these positive clones, three were picked for further study that stably expressed high levels (clone #3), moderate levels (clone #15) and slightly lower levels (clone #8) of Lck (Figure 3.7 A, right panel; Appendix Table 1). The ability of the BCR to initiate signaling was first examined in the Lyn-, Lck- and Blk-expressing AtT20 BCR+ Fyn+ cells. BCR-induced tyrosine phosphorylation of total cellular protein in the various Srctransfected AtT20 cell lines was no different from that observed in the parental untransfected cells containing only Fyn (Figure 3.7 B). There is only a slight variation in the tyrosine phosphorylated bands in the 50-60 kDa range. These different bands are the transfected Src kinases. No major difference in tyrosine phosphorylation of total cellular protein in the Src-transfected and untransfected cell lines is observed when time course experiments were performed (data not shown).  The effects of the addition of either Lyn, Lck or Blk, and differences, if any, between the three Src kinase family members with respect to BCR-induced Akt phosphorylation and activation was examined: When Akt phosphorylation on the serine 473 residue was examined following BCR stimulation in the Lck-transfected cells (AtT20 BCR+ Fyn+ Lck+), no difference was observed compared to the untransfected AtT20 BCR+ Fyn+ parental cell line (Figure 3.8). The amount of Lck in the various clones had no effect on Akt phosphorylation.  However, in cells transfected with Lyn (AtT20 BCR+ Fyn+  Lyn+), Akt phosphorylation was slightly different. In these cells, the phosphorylation reached a peak at an earlier time point following BCR stimulation compared to the untransfected parental cells, 3 minutes for AtT20 BCR+ Fyn+ Lyn+ clones versus 15 minutes for AtT20 BCR+ Fyn+ (Figure 3.9). This phosphorylation, like that observed in AtT20 BCR+ Fyn+ cells, was not sustained over 60 minutes of receptor activation. The faster kinetics of BCR-induced Akt phosphorylation observed in the Lyn-transfected cells suggests that Lyn, unlike Lck, may play a role in enhancing PI 3-kinase/Akt pathway activation through the BCR. However, it does not function like Syk PTK since it was incapable of enhancing and sustaining Akt phosphorylation over a long period of BCR activation. A different result was obtained when Akt phosphorylation was examined in  cells transfected with Blk (AtT20 BCR+ Fyn+ Blk+ cells) following BCR cross-linking. The BCR-induced Akt phosphorylation was lower in the different Blk-transfected clones compared to the untransfected AtT20 BCR+ Fyn+ parental cell line (Figure 3.10). As previously shown, Akt phosphorylation in the AtT20 BCR+ Fyn+ cell line increased at 5 minutes of BCR stimulation and reached a peak at 15-30 minutes before decreasing at 60 minutes of receptor activation (Figure 3.10, left side of panels). In the AtT20 BCR+ Fyn+ Blk+ clones, however, BCR-induced Akt phosphorylation does not reach its peak until 30 minutes following receptor cross-linking.  This phosphorylation decreased  following 60 minutes of BCR engagement as was observed in the untransfected parental cell line. Akt phosphorylation at the 5 and 15 minute time points was slightly lower in the AtT20 BCR+ Fyn+ Blk+ cells compared to the untransfected cells, however, the extent of phosphorylation at the 30 minute time point was not lower in the Blk-expressing cells compared to the untransfected AtT20 BCR+ Fyn+ cells. The slower kinetics of Akt phosphorylation observed in the Blk transfected cells suggests that Blk may play a negative regulatory role in BCR-mediated PI 3-kinase/Akt pathway activation.  Figure 3.8 Lck has no effect BCR-induced Akt phosphorylation of the serine 473 residue in transfected AtT20 cells. The three Lck-transfected AtT20 BCR+ Fyn+ cell lines (clones #3, #8 and #15) were serum-starved for 16 hours in media containing 0.2% FCS before being stimulated with 20 pg/ml of anti-mouse IgM antibodies for the indicated time periods. Following lysis, 40 pg of total cell extract was separated by SDSPAGE and analyzed by Western blotting with anti-phospho-serine 473 Akt antibodies. The filters were subsequently stripped of antibodies and re-probed with anti-Akt antibodies (lower panels). The bands were visualized by enhanced chemiluminescence. Molecular weight standards (in kDa) are indicated to the left. representative of three experiments.  These results are  AtT20 BCR+ Fyn+ 0  AtT20 BCR+ Fyn+ Lck+ [#10.3]  5 15 30 60 I 0  5 15 30 60  IB: anti-phosphoserine 473 Akt  64  -p-Akt  Re-probe: anti-Akt  64-  -Akt  nAIT20 BCR+ Fyn+ • AtT20 BCR+ Fyn+ Lck» [#1Q.3]I  £ 15 min Anti-IgM  cross-linking  AtT20 BCR+ Fyn+ Anti-IgM (minutes)  0  30 min  60 min  (minutes)  AtT20 BCR+ Fyn+ Lck+ [#10.8]  5 15 30 601 0  5 15 30 60  IB: anti-phosphoserine 473 Akt  64-  -p-Akt  Re-probe: anti-Akt  64 -  -Akt  AtT20 BCR+ Fyn+ Anti-IgM (minutes)  IB: anti-phosphoserine 473 Akt  Re-probe: anti-Akt  64  64-  0  AtT20 BCR+ Fyn+ Lck+ [#10.15]  5 15 30 60 I 0 mmm  — — -  5 15 30 60 -p-Akt  -Akt  Figure 3.9 Lyn increases the kinetics of BCR-induced Akt phosphorylation of the serine 473 residue in transfected AtT20 cells. The three Lyn-transfected AtT20 BCR+ Fyn+ cell lines (clones #21, #33 and #52) were serum-starved for 16 hours in media containing 0.2% FCS before being stimulated with 20 pg/ml of anti-mouse IgM antibodies for the indicated time periods. Following lysis, 40 pg of total cell extract was separated by SDS-PAGE and analyzed by Western blotting with anti-phospho-serine 473 Akt antibodies. The filters were subsequently stripped of antibodies and re-probed with anti-Akt antibodies (lower panels). chemiluminescence.  The bands were visualized by enhanced  Molecular weight standards (in kDa) are indicated on the left.  These results are representative of three independent experiments with similar results.  AtT20 BCR+ Fyn+ 1 0 IB: anti-phosphoserine 473 Akt  3  AtT20 BCR+ Fyn+ Lyn+ [#21]  5 15 30 60 lo  3  5 15 30 60 1  64-  Re-probe: anti-Akt 64  AtT20 BCR+ Fyn+ Anti-IgM (minutes)  IB: anti-phosphoserine 473 Akt  6 4  0  3  5 15 30 60 lo  Anti-IgM (minutes) 1  Re-probe: anti-Akt  64-  5 15 30 60 I  E AtT20 BCR+ Fyn+  64-  3  .  Re-probe: anti-Akt 64 -  IB: anti-phosphoserine 473 Akt  AtT20 BCR+ Fyn+ Lyn+ [#33]  0  3  AtT20 BCR+ Fyn+ Lyn+ [#52]  5 15 30 60 0  3  5 15 30 60 I  Figure 3.10 Blk decreases the kinetics of BCR-induced Akt phosphorylation of the serine 473 residue in transfected AtT20 cells. The three Blk-transfected AtT20 BCR+ Fyn+ cell lines (clones #18, #27 and #41) were serum-starved for 16 hours in media containing 0.2% FCS before being stimulated with 20 pg/ml of anti-mouse IgM antibodies for the indicated time periods. Following lysis, 40 pg of total cell extract was separated by SDS-PAGE and analyzed by Western blotting with anti-phospho-serine 473 Akt antibodies. The filters were subsequently stripped of antibodies and re-probed with anti-Akt antibodies (lower panels).  The bands were visualized by enhanced  chemiluminescence. Molecular weight standards (in kDa) are indicated on the left. These results are representative of two experiments.  AtT20 BCR+ Fyn+ 0 IB: anti-phosphoserine 473 Akt  64-  Re-probe: anti-Akt  64-  5 15 30 60 I 0  Anti-IgM r (minutes) 1  64-  Re-probe: anti-Akt  64  IB: anti-phosphoserine 473 Akt  64-  Re-probe: anti-Akt  64-  I  0  —  I  —  AtT20 BCR+ Fyn+ Blk+ [#27]  0 5 15 30 60  AtT20 BCR+ Fyn+ Anti-IgM (minutes)  5 15 30 60 -  AtT20 BCR+ Fyn+  IB: anti-phosphoserine 473 Akt  AtT20 BCR+ Fyn+ Blk+[#18]  0  5 15 30 60  AtT20 BCR+ Fyn+ Blk+[#41]  5 15 30 60 I 0  5 15 30 60  I  3.2.4  Lyn and Blk, in conjunction with Syk inhibit BCR-induced Akt phosphorylation and activation  From the experiments in the previous section, it appears that Lyn, Blk and Lck have different roles in PI 3-kinase/Akt activation following BCR cross-linking. However, the maximal levels of Akt phosphorylation in the Lyn-, Lck- and Blk-transfected was not greater or lower than that observed in the untransfected parental AtT20 BCR+ Fyn+ cell line. Since the Src kinases and Syk PTK have different downstream targets that they modify and since phosphorylated Syk molecules can serve as docking sites for other signaling components, which can be recruited to the BCR complex at the plasma membrane, the consequences of various combinations of Syk and Src kinases on BCRmediated PI 3-kinase/Akt pathway activation was examined. To this end, the AtT20 cell line previously transfected with Syk (AtT20 BCR+ Fyn+ Syk+ [clone #13]) was transfected or retrovirally infected with plasmids containing the murine forms of Lyn, Blk and Lck. From the Lyn transfection of AtT20 BCR+ Fyn+ Syk+ [clone #13] cells (performed by Linda Matsuuchi, Ph.D. thesis supervisor), only two Lyn positive clones were obtained, AtT20 BCR+ Fyn+ Syk+ Lyn+ [clone #2] and [clone #8] (Figure 3.11 A, right panel). Both clones expressed high levels of Lyn compared to AtT20 BCR+ Fyn+ Lyn+ [clone #21] cells. Cells co-expressing Syk and Blk were created by infecting AtT20 BCR+ Fyn+ Syk+ [clone #13] cells with retrovirus particles containing the expression vector for the murine Blk gene (pMSCV-Blk).  Several clones were obtained  from the infection, however, only two of these clones were further examined. The AtT20 BCR+ Fyn+ Syk+ Blk+ [clone #17] was chosen because it stably expressed high levels of Blk whereas [clone #64] expressed lower levels of Blk (Figure 3.11 A, middle panel). Finally, from the Lck transfection of AtT20 BCR+ Fyn+ Syk+ [clone #13] cells (performed by Elisa Vicencio, a former undergraduate student), a few AtT20 BCR+ Fyn+ Syk+ Lck+ clones were obtained. The two clones used for further study expressed high levels (clone #17) and slightly lower levels (clone #21) of Lck (Figure 3.11 A, left panel). The pattern of tyrosine phosphorylation of total cellular protein following BCR cross-  Figure 3.11  Expression levels of Src kinases and BCR-induced tyrosine  phosphorylation of total cellular protein in transfected AtT20 cells co-expressing Syk PTK. (A) Expression levels of the various Src kinases in the transfected AtT20 BCR+ Fyn+ Syk+ cells. Whole cell extract (30-40 pg) from the different cell lines was separated by SDS-PAGE and analyzed by Western blotting with anti-Lyn, anti-Blk or anti-Lck antibodies. The Western blot of the AtT20 BCR+ Fyn+ Syk+ Lck+ clones was performed by Elisa Vicencio. (B) BCR-induced tyrosine phosphorylation of total cellular protein in Src- and Syk-transfected AtT20 cells. The different cell lines were stimulated with 20 pg/ml of anti-mouse IgM antibodies for 5 minutes. Following cell lysis, 20 pg of total cell extract was separated by SDS-PAGE and analyzed by Western blotting with the anti- phosphotyrosine monoclonal antibody (4G10).  The bands were visualized by  enhanced chemiluminescence. (C) In vitro kinase assay of transfected Src kinases in the AtT20 cells. Lyn, Blk and Lck were immunoprecipitated from fresh cell lysates with the appropriate antibodies. The immunoprecipitates were subjected to an in vitro kinase assay, as described in the Materials and Methods (Chapter 2.6.3). The molecular weight standards (in kDa) are indicated to the left. These results are representative of two experiments.  AtT20 BCR+ Fyn+ Syk+ Blk+ clones  AtT20 BCR+ Fyn+ Syk+ Lyn+ clones  >>  LL  N> Ov  + DC  (B)  o O CM  c+  +  LL  LL  °°  cc +  O c CD  ^  CO  < (/) Anti-IgM (5 min)  Lck  IB: anti-Lck  IB: anti-Blk  IB: anti-Lyn +  O > AtT20 BCR+ Fyn+ cd w Syk+ Lck+ clones  Blk  Lyn  c  D+C +  +  — .  + DC + O CD o O CM + I—  >. < CO  c >> r ^ LL , -  + * ^ + o co m  (C)  £  o t  .  ,, CM  + C  >•  i  +  < cn  f  1  CQ 1 o <0 +  >.  h-  •J— +  u +  XL >«  c  +  cn +  LL  LL  LL  + DC O  O  O  CM H  CM 1—  >  + DC  o  CQ O  <  <  m  IJ-Src kinases  +  a  c  +  180-  CO  CM I—  c  >>  + DC  CD O  <  In vitro kinase assay IB: anti-phosphotyrosine (4G10)  linking was similar in the various Src- and Syk-transfected cells and in the Syktransfected parental cell line (Figure 3.11 B).  The AtT20 cells were capable of  expressing active forms of all the Src kinases that were added to the cells by DNA transfection, as shown by an in vitro kinase assay (Figure 3.11 C).  Since the Src kinases had an effect on BCR-induced Akt phosphorylation in the absence of Syk in transfected AtT20 cells (Figures 3.8, 3.9 and 3.10), it was of interest to determine the effects of Syk and Src kinase co-expression in these cells. It was predicted that the extent of BCR-induced Akt phosphorylation would be greater than that observed in the AtT20 cells expressing/ Syk (AtT20 BCR+ Fyn+ Syk+). BCR-mediated Akt phosphorylation in the two clones co-expressing Syk and Lck (AtT20 BCR+ Fyn+ Syk+ Lck+ [clone #17] and [clone #21]) and in the cells expressing only Syk (AtT20 BCR+ Fyn+ Syk+) was similar (Figure 3.12). These results and those obtained in the AtT20 BCR+ Fyn+ Lck+ cells suggest that Lck does not play a role in either amplifying or inhibiting BCR-mediated Akt phosphorylation and activation. This is not surprising since Lck in known to play a major role in TCR signaling, but not BCR signaling.  Addition of Lyn alone to the BCR-expressing AtT20 cells (AtT20 BCR+ Fyn+ Lyn+ cells; Figure 3.9) resulted in faster induction of BCR-induced Akt phosphorylation while addition of Syk alone (AtT20 BCR+ Fyn+ Syk+ cells; Figure 3.2) enhanced and sustained Akt phosphorylation. It was therefore hypothesized that addition of both Syk and Lyn into BCR-expressing AtT20 cells should result in an even greater increase in Akt phosphorylation in response to BCR activation. This was not the case however. Highly unexpected results were obtained when Akt phosphorylation was examined in cells coexpressing Syk and Lyn. Akt phosphorylation on the serine 473 residue following varying periods of BCR cross-linking in the two AtT20 BCR+ Fyn+ Syk+ Lyn+ clones (clone #2, Figure 3.13, upper panels; clone #8, Figure 3.13, lower panels) was dramatically lower at all time points compared to cells not expressing Lyn (AtT20 BCR+ Fyn+ Syk+ cells) (Figure 3.13). Akt phosphorylation increased following a short period of receptor stimulation (3-5 minutes). However, there was an average of 3-4 fold greater Akt phosphorylation in the AtT20 BCR+ Fyn+ Syk+ cells compared to the AtT20 BCR+  Fyn+ Syk+ Lyn+ cells at all time points of BCR activation, as quantified using the ImageQuant 5.1 program (Figure 3.13 graphs). Akt phosphorylation on the other residue, threonine 308, was also lower in the AtT20 cell co-transfected with both Lyn and Syk compared to cells only expressing Syk (Figure 3.13, upper panels). Therefore Lyn, in the presence of Syk, appears to play an inhibitory role in the regulation of the PI 3kinase/Akt pathway activation following BCR cross-linking.  In the presence of Syk and Blk (AtT20 BCR+ Fyn+ Syk+ Blk+ clones), Akt phosphorylation was also suppressed at all time points of BCR stimulation compared to cells transfected with Syk alone (AtT20 BCR+ Fyn+ Syk+). In the Syk and Blk cotransfected cells, Akt phosphorylation increased only slightly following 5 minutes of BCR activation and reached its peak at 15 minutes (Figure 3.14). However, the extent of Akt phosphorylation in the AtT20 BCR+ Fyn+ Syk+ BIk+ clones at any time point of BCR activation was considerably lower than that observed in the AtT20 BCR+ Fyn+ Syk+ cell line.  The Akt phosphorylation in the AtT20 BCR+ Fyn+ Syk+ cells is  approximately 3-4 fold greater compared to that in the AtT20 BCR+ Fyn+ Syk+ Blk+ cells, as quantified using the ImageQuant 5.1 program (Figure 3.14 graphs). BCRmediated Akt phosphorylation was not affected by differences in the amount of Blk expressed in the two cell lines since the pattern of phosphorylation was similar in both clones. Therefore, Blk, like Lyn, plays an inhibitory role in the activation of the PI 3kinase/Akt pathway following BCR cross-linking in the presence of Syk. However, unlike Lyn, Blk has a slight inhibitory effect on BCR-induced Akt phosphorylation in the absence of Syk. Interestingly, the extent of inhibition of Akt phosphorylation was similar in the AtT20 BCR+ Fyn+ Syk+ Lyn+ and AtT20 BCR+ Fyn+ Syk+ BIk+ cells when the cell lysate samples were made at the same time and Western blotted on the same gel (Figure 3.15).  Figure 3.12  Lck in the presence of Syk does not affect BCR-induced Akt  phosphorylation of the serine 473 residue in transfected AtT20 cells. The two Lcktransfected AtT20 B.CR+ Fyn+ Syk+ cell lines [clones #17 and #21] were serum-starved for 16 hours in media containing 0.2% FCS before being stimulated with 20 pg/ml of anti-mouse IgM antibodies for the indicated time periods. Following lysis, 40 pg of total cell extract was separated by SDS-PAGE and analyzed by Western blotting with antiphospho-serine 473 Akt antibodies. The filters were subsequently stripped of antibodies and re-probed with anti-Akt antibodies (lower panels). The bands were visualized by enhanced chemiluminescence. Molecular weight standards (in kDa) are indicated on the left. The immunoreactive bands were quantified using the ImageQuant 5.1 software. The number of replicates for each experiment is indicated at the bottom right side of the graphs.  AtT20 BCR+ Fy n+ Sy k+ 0 IB: anti-phosphoserine 473 Akt  AtT20 BCR+ Fyn+ Sy k+ Lck+ [# 17]  5 15 30 60 0  5 15 30 60  I  .  6 4  -p-Akt  Re-probe: anti-Akt 64 -  -Akt  FT  ri it  • Syk-f • Syk+ Lck+ |#17]  •I• ji 5 min  15 min  Anti-IgM cross-linking  30 min  AtT20 BCR+ Fyn-i- Syk+ Anti-IgM (minutes)  IB: anti-phospho- 64 serine 473 Akt  0  n=5  60 min  (minutes)  AtT20 BCR+ Fyn+ Syk+ Lck+ [#21]  5 15 30 60 0  5 15 30 60 -p-Akt  w A w v  Re-probe: anti-Akt 64 -  -Akt  I 3 i. •3 1i  P-  5 min Anti-IgM  A 15 min cross-linking  f  30 min (minutes)  O Syk-f  11  • Syk+ Lck+ [#21]  Figure 3.13  Lyn has an inhibitory effect on BCR-induced Akt phosphorylation of  the serine 473 residue in the presence of Syk PTK. The two Lyn-transfected AtT20 BCR+ Fyn+ Syk+ cell lines (clones #2 and #8) were serum-starved for 16 hours in media containing 0.2% FCS before being stimulated with 20 pg/ml of anti-mouse IgM antibodies for the indicated time periods. Following lysis, 40 pg of total cell extract was separated by SDS-PAGE and analyzed by Western blotting with anti-phospho-serine 473 Akt or anti-phospho-threonine 308 Akt antibodies.  The filters were subsequently  stripped of antibodies and re-probed with anti-Akt antibodies (lower panels). The bands were visualized by enhanced chemiluminescence. Molecular weight standards (in kDa) are indicated on the left.  The immunoreactive bands were quantified using the  ImageQuant 5.1 software. The number of replicates for each experiment is indicated at the bottom right side of the graphs.  AtT20 BCR+ Fyn+ Syk+ Lyn+ [#2]  AtT20 BCR+ Fyn-i- Syk+ (minutes)  IB: anti-phosphoserine 473 Akt  64  0  3  5  15  30 60  lo  3  5  15  30  60  .  p-serine 473 Akt  <  Re-probe: anti-Akt 64 -  IB: anti-phosphoserine 473 Akt  6 4  .  p-threonine 308Akt  1  J 1 20  w *• ! ,5 = « 10 ?  • Syk* D Syk+ Lyn+ [»2]  1  5 min Anti-IgM  15 min  cross-linking  AtT20 BCR+ Fyn+ Syk+ Anti-IgM (minutes)  64  I  0  3  1  n 0 min  IB: anti-phosphoserine 473 Akt  (minutes)  AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8]  5 15 30 6010  5 15 30 60  .  15 15  • Ot  <  p-Akt  <  Akt  • Syk. • Syk+Lyn+ [»8]  iti  e • .£  I  rfi  fi  . .a 20  n=3  30 min  Re-probe: anti-Akt 64 -  l !i  Akt  10  w  n H 5 min Anti-IgM  1  15 min  crosa-linking  T  30 min (minutes)  4i  Figure 3.14 Blk has an inhibitory effect on BCR-induced Akt phosphorylation of the serine 473 residue in the presence of Syk PTK. The two Blk-transfected AtT20 BCR+ Fyn+ Syk+ cell lines (clones #17 and #64) were serum-starved for 16 hours in media containing 0.2% FCS before being stimulated with 20 pg/ml of anti-mouse IgM antibodies for the indicated time periods. Following lysis, 40 pg of total cell extract was separated by SDS-PAGE and analyzed by Western blotting with anti-phospho-serine 473 Akt antibodies. The filters were subsequently stripped of antibodies and re-probed with anti-Akt antibodies (lower panels).  The bands were visualized by enhanced  chemiluminescence. Molecular weight standards (in kDa) are indicated on the left. The immunoreactive bands were quantified using the ImageQuant 5.1 software. The number of replicates for each experiment is indicated at the bottom right side of the graphs.  AtT20 BCR+ Fyn+ Syk+ I IB: anti-phosphoserine 473 Akt  Re-probe: anti-Akt  0  AtT20 BCR+ Fyn+ Syk+Blk+ [#17]  5 15 30 601 0 5 15 30 60  I  64-  p-Akt  64 -  Akt  30  1 I " M t 5 20 > I% %  • Syk+ " Syk+Blk+ [#17]'  3* I & 10  H=i» '  II  5 min  Anti-IgM  15 min  30 min  cross-linking  AtT20 BCR+ Fyn+ Syk+ Blk+ [#64]  AtT20 BCR+ Fyn-i- Syk+ £E2£J I IB: anti-phosphoserine 473 Akt  Re-probe: anti-Akt  64  0  n=3  (minutes)  5 15 30 601 0 5 15 30 60  I  .  p-Akt  64-  Akt  j  T  |s  -L  I!  T  j Bw  s  • Syk-f • Syk+Blk+  1 •Wl 0 min  5 min Anti-IgM  15 min cross-linking  30 min (minutes)  \  60 min  (#64]  Figure 3.15 The extent of BCR-mediated inhibition of Akt phosphorylation in the AtT20 BCR+ Fyn+ Syk+ Lyn+ and AtT20 BCR+ Fyn+ Syk+ Blk+ cell lines is similar. The various transfected AtT20 BCR+ Fyn+ Syk+ cell lines were serum-starved for 16 hours in media containing 0.2% FCS before being stimulated with 20 pg/ml of anti-mouse IgM antibodies for the indicated time periods. Following lysis, 40 pg of total cell extract was separated by SDS-PAGE and analyzed by Western blotting with antiphospho-serine 473 Akt antibodies. The filters were subsequently stripped of antibodies and re-probed with anti-Akt antibodies (lower panels). The bands were visualized by enhanced chemiluminescence. Molecular weight standards (in kDa) are indicated on the left. These results are representative of two experiments.  llToo +  cc +  CQ j j  it < <0  I* Anti-IgM j (minutes)l  IB: anti-phosphoserine 473 Akt  Re-probe: anti-Akt  64-  64-  < LL U  5  + DC £+  DC +  15  TT  5  O CQ CQ O CMi CO  15 10  5  15  p-Akt  Akt  3.3  Discussion  Using the BCR-expressing AtT20 cell system, it was established that Akt phosphorylation and hence activation can be regulated by the BCR. In addition, the different BCR-associated and activated PTKs play different roles in the regulation of this BCR-mediated process. The Src family kinase that is endogenously expressed in the AtT20 cells, Fyn, was sufficient for promoting BCR-induced Akt pathway activation, as monitored by changes in the phosphorylation of Akt on the serine 473 and threonine 308 residues. This Akt phosphorylation in the AtT20 BCR+ Fyn+ cells, however, was not sustained over long periods of BCR cross-linking (Figure 3.1). The catalytic activity of Syk is required for amplifying and sustaining this phosphorylation (Figure 3.2 and 3.4). This BCR-induced Akt phosphorylation in the AtT20 BCR+ Fyn+ and AtT20 BCR+ Fyn+ Syk+ cell lines is dependent on PI 3-kinase activity (Figure 3.3). These results obtained in the AtT20 cells were supported by data obtained from experiments using the DT40 chicken B lymphoma cell line and were published together (Gold et al., 1999). These cells endogenously express Syk and only one Src family member, Lyn. Loss-offunction studies with Lyn and Syk single and double knockout DT40 cells indicated that Lyn is required for BCR-induced Akt phosphorylation since very little Akt phosphorylation is observed in cells lacking Lyn. In addition, Syk is required for maximal phosphorylation of Akt and for sustaining this phosphorylation over longer periods of BCR cross-linking. These published data involving the experiments with the AtT20 and DT40 cell lines were consistent with the results presented in two other publications that also investigated Akt pathway activation downstream of the BCR (Craxton et al., 1999; Gold et al., 1999; Li et al., 1999).  The proposed model of PI 3-kinase-dependent BCR-mediated Akt phosphorylation and activation based on the data obtained from the AtT20 cell system is diagrammed in Figure 3.16. BCR cross-linking leads to the recruitment and activation of PI 3-kinase at the plasma membrane.  The activated PI 3-kinase then phosphorylates PI-4,5-P2  contained within the plasma membrane to produce PI-3.4,5-P3. Production of PI-3,4,5-P3 leads to the recruitment of various PH domain-containing signaling effectors including  PDK1, PDK2 and Akt. The PH domains of these molecules interact with PI-3,4,5-P3 within the plasma membrane. In the case with Akt, the interaction of its PH domain with the phospholipids within the plasma membrane results in a conformational change that exposes its threonine 308 and serine 473 residues. The upstream kinases PDK1 and PDK2 that are also at the plasma membrane and in close proximity to Akt, phosphorylate these exposed residues on Akt.  This phosphorylation ultimately leads to the full  activation of Akt (Andjelkovic et al., 1997; Alessi et al., 1997a; Alessi et al., 1997b; Franke-ef al, 1997; Burgering and Coffer, 1995). The regulation of PDK1 and PDK2 by the BCR appears to be the same since the pattern of phosphorylation of the threonine 308 and serine 473 residues in some of the cell lines examined was very similar (Figures 3.1, 3.2 and 3.13). The role of three Src kinase family members in BCR signaling were also examined using the AtT20 cell system. The three Src kinases studied were chosen for specific reasons. Lyn was selected because experiments involving knockout mice and several B lymphoma cell lines indicated that this Src kinase is very important for the initial stages of BCR signaling (Hibbs et al., 1995; Takata and Kurosaki, 1995; Wang et al., 1996).  In  addition, Lyn also plays a role in the downregulation of this signaling that it originally initiated. Blk was chosen for this study because not much is known about this particular family member. Since Blk expression is limited to B lymphocytes, it presumably has a unique and specific function in these cells which is not required in other cell types or which cannot be performed by other family members in B lymphocytes.  Lck was  examined to determine if it could in any way alter signaling through the BCR. Lck is a T lymphocyte kinase, but is also expressed in several B lymphoma cell lines. It was of interest to determine if a T lymphocyte kinase has similar or different effects on BCR signaling compared to Src family members predominantly found in B lymphocytes such as Lyn and Blk. AtT20 cell lines stably expressing the three Src kinases were generated and it was established that these family members all play different roles in BCR-induced Akt phosphorylation and activation. Lck alone does not alter Akt phosphorylation in response to BCR cross-linking.(Figures 3.8). Lyn alone increased the kinetics while Blk alone slowed down the kinetics of BCR-induced Akt phosphorylation (Figures 3.9 and  Figure 3.16 Proposed mechanism of BCR-induced phosphorylation of Akt in Syktransfected AtT20 cells. (A) BCR cross-linking results in phosphorylation of an adapter protein (X) by Syk, and to a lesser extent by the endogenous Fyn, which leads to the association of PI 3-kinase with the adapter protein. This association ultimately leads to the recruitment of PI 3-kinase to the plasma membrane, resulting in the production of PI3,4,5-P3. (B) PI-3,4,5-P3 production results in the membrane recruitment of PH domaincontaining proteins such as PDK1, PDK2 and Akt. Once at the plasma membrane, PDK1 and PDK2 phosphorylate Akt on threonine 308 and serine 473, respectively, resulting in the full activation of Akt. Gold et al, Journal of Immunology (1999), Volume 163:1894.  (B)  PIP,  Antigen  PIP3  ?  ll ^SrcPB - J L1  PLASMA MEMBRANE  PLASMA MEMBRANE  3.10, respectively). Although the kinetics of Akt phosphorylation are slightly altered by the presence of Lyn and Blk, the extent of phosphorylation in the AtT20 BCR+ Fyn+ Lyn+ and AtT20 BCR+ Fyn+ Blk+ clones was no greater and no less, respectively, than that in the untransfected parental cell line. Since B lymphocytes express both Src family members and Syk PTK, AtT20 cells coexpressing these two families of tyrosine kinases were created (Figure 3.11). BCRinduced Akt phosphorylation in the AtT20 BCR+ Fyn+ Syk+ and AtT20 BCR+ Fyn+ Syk+ Lck+ cell lines was similar (Figure 3.12). Thus, Lck does not play a role in Akt phosphorylation and activation downstream of the BCR.  Unexpected results were  obtained when AtT20 BCR+ Fyn+ Syk+ Lyn+ clones were examined.  Lyn, in  conjunction with Syk, had an inhibitory effect on BCR-mediated Akt phosphorylation (Figure 3.13). This result was initially unexpected since Lyn is known to play a positive regulatory role in the initiation of BCR signaling (Yamanashi et al., 1991; Yamanashi et al., 1992; Takata and Kurosaki, 1995). Addition of both Syk and Lyn kinases into AtT20 cells was expected to result in an increase as opposed to a decrease in BCR-mediated Akt pathway activation. However, there is increasing evidence from knockout mice that implicates a role for Lyn in the downregulation of several BCR-activated signaling components including PKC enzymes (Katsuta et al., 1998) and MAP kinase (Chan et al., 1997). The role of Lyn in BCR-induced PI 3-kinase/Akt pathway activation was also examined in Lyn deficient DT40 B lymphoma cell line and in B lymphocytes from Lyn knockout mice (Craxton et al., 1999).  This study showed that BCR-induced Akt  phosphorylation and activation was increased in these Lyn deficient DT40 and knockout mouse B lymphocytes compared to wild type cells, suggesting that Lyn negatively regulates Akt function in these cells. Thus, the idea that Lyn plays a role in inhibiting Akt phosphorylation following BCR cross-linking in the AtT20 cell system may not be as unexpected as originally thought. The effects of different levels of Lyn on BCR-induced Akt phosphorylation in the presence of Syk could not be examined since only two clones were obtained from the transfection that expressed roughly the same amount of Lyn. The levels of Lyn expression in the two AtT20 BCR+ Fyn+ Syk+ Lyn+ clones although  relatively high, were still within the "physiological" limits since approximately the same amount of Lyn was expressed in the Ramos and Daudi B lymphoma lines (personal observation, data not shown). In addition, the amount of Lyn in the AtT20 BCR+ Fyn+ Syk+ Lyn+ clones was not greater than that expressed K40-B1 pro-B and BJAB B lymphoma lines, which expressed approximately 2-3 times more Lyn. Therefore, the levels of Lyn in the transfected AtT20 cells were within the limits of the amount found in the B lymphoma lines examined and the results obtained with the clones were most likely not an artifact due to the over-expression of the transfected Lyn.  Other surprising results were obtained when BCR-induced Akt phosphorylation was examined in AtT20 cells transfected with Blk. Co-expression of Syk and Blk resulted in the suppression of Akt phosphorylation following BCR engagement (Figure 3.14). In the absence of Syk, only the kinetics of this phosphorylation were affected (Figure 3.10). These results imply that Blk, like Lyn, can negatively regulate Akt pathway activation downstream of the BCR.  The role of Blk as a negative regulator of BCR signal  transduction has not been established in B lymphocytes. Studies from knockout mice have indicated that Blk does not play a unique or special role in BCR activation since B lymphocyte development and proliferation and BCR signaling are normal in these mice compared to wild type (Texido et al., 2000). The lack of Blk in the knockout mice had no effect presumably because of the presence of other Src family members within the animals. In addition, B lymphocytes from Blk and Fyn double knockout mice appear normal suggesting that other Src family members may be compensating for the loss of these kinases. Thus, the results obtained with the AtT20 cell system implicating a role for Blk as a negative regulator of BCR signaling is a novel finding.  There are a few possible explanations as to why Akt phosphorylation is inhibited following BCR cross-linking in the AtT20 BCR+ Fyn+ Syk+ Lyn+ and AtT20 BCR+ Fyn+ Syk+ Blk+ cells but not in the AtT20 BCR+ Fyn+ Syk+ Lck+ cells.  One  possibility may be due to the cellular localization of these three different Src kinase family members within the cell. Lyn and Lck contain within their amino terminus two acyl modifications (myristylation and palmitylation) whereas Blk only possesses the  myristate modification that is characteristic of all Src kinase family members. The myristate modification allows proteins to anchor themselves to the membranes within cells while the palmitylation modification allows proteins to become localized to cholesterol- and sphingomyelin-enriched regions within the plasma membrane termed lipid rafts or membrane microdomains.  Therefore, the differences  in Akt  phosphorylation and activation in the various cell lines may be due to the differences in localization of Lyn, Lck and Blk within these lipid rafts at the plasma membrane. This localization could potentially affect the ability of the Src kinases to gain access to and activate or inhibit different sets of signaling molecules also present at the membrane. Alternatively, the differences in Akt phosphorylation following BCR engagement in the cell lines examined may be due to the ability of the different Src kinases to phosphorylate and activate different sets of downstream target molecules. Some of these molecules may be negative regulatory signaling components such as protein tyrosine phosphatases. These activated tyrosine phosphatases could potentially dephosphorylate and inactivate PI 3-kinase directly. On the other hand, the tyrosine phosphatase could dephosphorylate docking sites for PI 3-kinase on adapter proteins that are responsible for recruiting it to the plasma membrane.  Removal of these docking sites would result in decreased  recruitment of PI 3-kinase to the plasma membrane, subsequently leading to lower levels of PI-3,4,5-P3 and decreased recruitment and activation of PH domain-containing signaling components. A third possibility exists where Syk and Lyn or Syk and Blk may coordinately phosphorylate a target, such as an adapter protein, that can recruit the lipid phosphatases SHIP, SHIP-2 and/or PTEN. The recruitment of these phosphatases to the plasma membrane would counteract PI 3-kinase activity by dephosphorylating the products it creates, thus leading to a decrease in PI-3,4,5-P3 within the plasma membrane. Lower levels of this phospholipid would subsequently result in the lack of recruitment and activation of PH domain-containing proteins including PDK1, PDK2 and Akt. Finally, another possible explanation for the BCR-induced inhibition of Akt observed is that Lyn and Blk, but not Lck may be responsible for inhibiting Syk kinase activity. This Lyn- and Blk-mediated inhibition of Syk activity may be either direct or indirect. For example, Lyn and Blk could indirectly inhibit Syk by activating a negative regulator of its kinase activity. Alternatively, Lyn and Blk could associate with and sequester Syk  such that it is prevented from activating the PI 3-kinase/Akt pathway following BCR cross-linking.  Since Src kinases are responsible for the activation of Syk catalytic  activity, the inhibition of its kinase activity by Lyn and Blk is most likely not the cause of the observed inhibition of Akt phosphorylation (Kurosaki et al, 1994). In the following chapters, some of these proposed possibilities were examined. In summary, the three Src kinase family members, Lyn, Lck and Blk, play different roles in the regulation of the PI 3-kinase/Akt pathway downstream of the BCR. Lyn and Blk both act as negative regulators of BCR-induced Akt phosphorylation whereas Lck does not play a role in the regulation of this particular pathway.  CHAPTER 4 Lyn and Blk associate with the SHP-2 protein tyrosine phosphatase: Possible cause of the inhibition of Akt phosphorylation and activation  4.1  Introduction  It was originally thought that the Src kinases had similar or redundant functions since only subtle phenotypic alterations were observed in mice that were deficient for different family members.  However, using the BCR-expressing AtT20 cell system, it was  established in the previous chapter that Src kinases expressed in B lymphocytes (Lyn, Lck and Blk) have different effects with respect to BCR-mediated regulation of the PI 3kinase/Akt pathway.  Lyn and Blk play an inhibitory role in Akt phosphorylation  following BCR cross-linking while Lck has no noticeable effect on this process. Lynand Blk-mediated inhibition of the Akt phosphorylation may be due to the direct or indirect binding of a negative regulator with these Src kinases. Regions within the Src kinases that have been shown to be important for interaction with other molecules are the unique region, SH2 and SH3 domains. The sequences of the amino-terminal unique region of all Src family kinases are highly divergent. Since this region is so different between the family members, it is thought that this region is responsible for interaction with different sets of signaling components and/or proteins with specialized functions. For example, Lck, which plays an important role in TCR signaling, can directly interact with the TCR co-receptors CD4 and CD8 (Rudd et al., 1988; Veillette et al., 1988; Barber et al., 19989). This interaction is mediated by two cysteine residues within the unique region and by the C-X-C-P motif (single amino acid letter code) within the cytoplasmic domains of the CD4 and CD8 TCR co-receptors (Shaw et al., 1989; Shaw et al., 1990; Turner et al., 1990). The unique region of Lck was also shown to be important for its association with the CD45 protein tyrosine phosphatase that regulates the  phosphorylation of the earboxy-terminal negative regulatory tyrosine residue of Lck (Gervais and Veillette, 1995). Like the Lck unique region, the unique regions of Lyn and Fyn are also involved in their association with cell surface receptors. The unique regions of Lyn and Fyn interact with the Ig-a and Ig-p accessory chains of the BCR in a phosphotyrosine-independent manner (Pleiman et al., 1994). This interaction is distinct from the interaction between the SH2 domains of these Src kinases and the tyrosine phosphorylated ITAM motifs within the BCR accessory chains. These interactions are thought to facilitate more efficient signaling by the BCR since the Src kinases are already associated with the receptor complex.  Another domain present in the Src kinases that is involved in mediating protein interactions is the SH2 domain. All SH2 domains recognize a phosphorylated tyrosine residue, but specificity is defined by three amino acid residues located to the earboxyterminal side of the tyrosine residue (+1, +2, +3) (Songyang et al., 1993; Songyang et al., 1994). However, it has been shown that residues other than these three amino acid residues greatly influence binding affinity of proteins to the SH2 domains (Courtneidge et al., 1991; Ladbury et al., 1995). Proteins that interact with the SH2 domains of Blk, Fyn(T) and Lyn have been examined previously (Malek and Desiderio, 1993). In vitro studies using the SH2 domains of these three Src kinases fused to GST were used to determine that distinct sets of tyrosine phosphorylated proteins from the A20 mature B lymphoma cells interacted with the different SH2 domains. The proteins that bound to the various SH2 domain GST fusion protein were not identified in this study. The final domain within the Src family kinases that mediates interaction with other proteins is the SH3 domain. SH3 domains recognize proline-rich P-X-X-P-X-P motifs, but flanking residues around this motif also influence the binding affinity (Rickels et al., 1995).  An in vitro study of the ability of various domains of Lyn, Blk and Fyn to interact with the signaling effectors, MAP kinase, PLC-y2, Ras and PI 3-kinase was conducted by Pleiman et al. (1993).  Serial truncations of Lyn, Blk and Fyn fused to GST were  generated and used to pull-down interacting proteins from K46 B lymphoma cell lysates. Results from this paper indicated that different regions of the three Src kinases examined  were responsible for mediating the interaction with various effector molecules including MAP kinase, PLC-y2 and PI 3-kinase. For example, the unique region of Lyn was necessary for mediating the interaction with MAP kinase and PLC-y2. In addition, the SH3 domains of all three Src family kinases were responsible for interacting with PI 3kinase.  Since the Src kinases Lyn, Fyn and Blk were able to differentially interact with various signaling effectors in vitro, it was hypothesized that the differences in BCR-induced Akt phosphorylation observed in the Syk- and Src-transfected AtT20 cell lines was due to the ability of the Src kinases to interact with different positive or negative regulatory signaling molecules. Through co-immunoprecipitation experiments, it was established that the various transfected Src kinases were able to differentially interact with the SHP-2 protein tyrosine phosphatase. Lyn and Blk can interact with SHP-2, but Lck-SHP-2 association could not be detected. In vitro experiments using the tandem SH2 domains of SHP-2 fused to GST confirmed these results. Upon examination of the phosphorylation of SHP-2 on the tyrosine 580 residue, it was determined that phosphorylation was slightly higher in AtT20 cells co-expressing Syk and Lyn compared to the other cells. Thus, the interaction between SHP-2 and the two Src kinases, Lyn and Blk, may be responsible for the observed inhibition of BCR-induced Akt pathway activation in the presence of Syk PTK.  4.2  Results  4.2.1  Lyn and Blk associate with the SHP-2 protein tyrosine phosphatase in AtT20 transfected cells  The ability o f the various Src kinases to associate with, phosphorylate and activate different downstream target molecules may account for the differences observed in BCRinduced Akt phosphorylation in the transfected AtT20 cells. Lyn has been previously shown to interact with the SHP-1 tyrosine phosphatase in B lymphocytes (Somani et al.,  2001) and in myeloid cells (Yoshida et al., 1999).  In B lymphocytes, SHP-1 is  implicated in the down-regulation of BCR signaling due to its ability to modulate the activity of Lyn (Somani et al., 2001). SHP-1 can interact with and de-phosphorylate the tyrosine 397 residue within the catalytic domain of Lyn, leading to a decrease in its kinase activity. Interestingly, SHP-1 has also been shown to interact with and downregulate Syk kinase activity (Dustin et al., 1999 and the results shown in Chapter 3.2.2 Figure 3.5).  Therefore, the differences in the tyrosine phosphorylated proteins that associated with Lyn, Lck and Blk in the Syk co-transfected AtT20 cell lines was examined by coimmunoprecipitation experiments.  The Src kinases were immunoprecipitated from  unstimulated and BCR-stimulated cells and the tyrosine phosphorylated proteins that associated with them were examined by SDS-PAGE and identified by Western immunoblotting with a monoclonal antibody, 4G10, that specifically recognizes tyrosine phosphorylated residues on proteins.  In preliminary experiments, a protein  approximately 65 kDa in size, which became tyrosine phosphorylated following BCR cross-linking, was consistently present in the Src immunoprecipitates from the transfected AtT20 cells (data not shown and Figure 4.1). This is the approximate size of the SHP-1 protein tyrosine phosphatase. Therefore, the possibility that the 65 kDa phosphoprotein was SHP-1 in the different Src kinase immunoprecipitates was examined since Lyn and SHP-1 were known to interact (Somani et al., 2001). However, SHP-1 was not found in the anti-Lyn, anti-Lck and anti-Blk immunoprecipitates from the different AtT20 cell lines even though it is reported to do so in the literature (data not shown). Therefore, the 65 kDa phosphoprotein was not SHP-1.  If the phosphoprotein was not SHP-1, then another possibility was that the 65 kDa phosphoprotein was SHP-2, a close family member of SHP-1 that is similar in size (Adachi et al., 1992; reviewed by Tamir et al., 2000; Tonks and Neel, 2001). In the BCR-stimulated AtT20 BCR+ Fyn+ Syk+ Lyn+ cells, Lyn co-immunoprecipitated with a tyrosine phosphorylated 65 kDa protein, which corresponded with the phosphorylated SHP-2 from the anti-SHP-2 immunoprecipitates (Figure 4.1 A and B). This 65 kDa co-  immunoprecipitating band was more abundant in the stimulated lanes ("+" lanes), indicating that the association was enhanced after BCR-stimulation.  In addition, a  tyrosine phosphorylated protein (approximately 55 kDa) was present in the SHP-2 immunoprecipitates that was the exact molecular weight as Lyn (Figure 4.1 A and B, compare anti-Lyn with anti-SHP-2 immunoprecipitates). The filters could not be reprobed with anti-Lyn antibodies to determine conclusively if the 55 kDa phosphorylated protein in the SHP-2 immunoprecipitates was in fact Lyn since the immunoprecipitating antibody migrates at the same place on the SDS gel as the Lyn band. However, reprobing the filters with anti-SHP-2 antibodies revealed the 65 kDa tyrosine phosphorylated protein in the anti-Lyn immunoprecipitates was in fact SHP-2. Therefore, Lyn and SHP-2 interact either directly or indirectly in Lyn and Syk cotransfected AtT20 cells (Figure 4.1 A and B, anti-SHP-2 re-probe).  SHP-2 was also found to associate with Blk in the cells co-transfected with Syk and Blk (AtT20 BCR+ Fyn+ Syk+ Blk+ cell line).  A tyrosine phosphorylated protein  approximately 65 kDa in molecular weight was also present in the anti-Blk immunoprecipitates in both unstimulated and stimulated cells (Figure 4.1 A).  This  protein was also subsequently identified as SHP-2 by re-probing the filter with antibodies that recognized SHP-2 (Figure 4.1 A, anti-SHP-2 re-probe). In the reciprocal immunoprecipitation, a barely detectable 55 kDa protein was present in the anti-SHP-2 immunoprecipitates of the AtT20 BCR+ Fyn+ Syk+ Blk+ cells in the unstimulated and BCR-stimulated samples that was the same molecular weight as the tyrosine phosphorylated Blk in the anti-Blk immunoprecipitates. The tyrosine phosphorylated Blk molecules that co-immunoprecipitated with SHP-2 were not as readily detectable as the Lyn molecules perhaps because of the fact that Blk was not as highly phosphorylated as Lyn. Indeed in the Src immunoprecipitates, the levels of tyrosine phosphorylation of Blk molecules were considerably lower compared to those observed for Lyn.  Figure 4.1 The Src kinases, Lyn and Blk, constitutively associate with the SHP-2 protein tyrosine phosphatase in the transfected AtT20 cells. (A) AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8] and AtT20 BCR+ Fyn+ Syk+ Blk+ [#17] and (B) AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8] and AtT20 BCR+ Fyn+ Syk+ Lck+ [#17] cells were serum-starved for 16 hours in media containing 0.2% FCS before being stimulated for 5 minutes with 20 pg/ml anti-mouse IgM antibodies (+) or left unstimulated (-). Following cell lysis and quantification of protein concentration by BCA assay, SHP-2 and the various Src kinases were immunoprecipitated from 2 mg of total cell extract with anti-SHP-2, anti-Lyn, antiLck and anti-Blk antibodies along with Protein A-Sepharose beads for 2 hours at 4°C. The immune complexes were washed twice with lysis buffer and subsequently eluted with IX SDS-PAGE reducing sample buffer.  All of the immunoprecipitates were  separated by SDS-PAGE and analyzed by Western blotting with the 4G10 antiphosphotyrosine  monoclonal  antibody that specifically  recognizes  tyrosine  phosphorylated residues on proteins. The filters were subsequently stripped of antibodies and re-probed with anti-SHP-2 antibodies (lower panels). The bands were visualized by enhanced chemiluminescence. The molecular weight standards (in kDa) are indicated to the left. (*) indicates the position of the immunoprecipitating antibodies. These results are representative of a three independent experiments with similar results.  (A)  C  C >  c CC  c  CO I  anti-SHP-2  I PPT: +  +  c >  c >,  u_ +  LL  + DC  tr ^ o Anti-IgM  (5 min)  .  < <o - +  o  c ,_,  oo £  +  c LL  ci3 m + CM < (0  CE]  <M >. (O l - + l-  CM  +I -  00  + DC £  + DC o  + c >. LL  D+C o  CO  +  J* (0 +  o  CM  c ca  k+ * DC +  +  J>. C CO  l - + l-  o  ,  S i < (0 +  l  114 81 -  6450-  •  SHP-2 Lyn Blk  37IB: anti-phospho-tyrosine (4G10) 81 -  -SHP-2  64Re-probe: anti-SHP-2  c  c  o  >  c  anti-SHP-2 c+  + c> LL  ll  + DC &  05 +  +  c  ®  >>  +C Q +  o CM +  JC >>  - (0+ l  ^ I o + CM C < O -  +  CO  CD + c> — > LL  0 0  LL  + DC ^  H o o -I CM + >• C O l -  c  c  a  o *, £  +  ti.  a  o  —I o .  o , CM + t >. < CO S i . (0+ l . + l . + l l< + l < CO  114 81 6450-  «••» •  «•  "" "  w  0*  BHpr  *tf  K  • SHP-2 .Lyn Lck  37-  81 -  64-  SHP-2  Finally, AtT20 BCR+ Fyn+ Syk+ Lck+ cell lines were examined to determine if SHP-2 could also interact with Lck as it did with Lyn and Blk in the transfected AtT20 cells. Although a tyrosine phosphorylated 65 kDa protein was found to co-immunoprecipitate with Lck, tyrosine phosphorylated Lck molecules were not present in the anti-SHP-2 immunoprecipitates despite an abundant amount of Lck in the transfected cells and SHP-2 was not present in the anti-Lck immunoprecipitates (Figure 4.1 B, anti-SHP-2 reprobe). This lack of interaction between Lck and SHP-2 could account for the fact that BCR-mediated Akt phosphorylation was not inhibited in this cell line.  Thus, the  interaction between SHP-2 with Lyn and Blk may account for the suppression of BCRinduced Akt phosphorylation observed in the AtT20 BCR+ Fyn+ Syk+ Lyn+ and AtT20 BCR+ Fyn+ Syk+ Blk+ cell lines.  4.2.2  BCR-induced SHP-2 tyrosine phosphorylation is greater in AtT20 cells coexpressing Lyn and Syk  Since Lyn and Blk can interact with SHP-2, the possibility that these Src kinases may be involved in SHP-2 activation was explored. The co-immunoprecipitation studies suggest that SHP-2 phosphorylation may be higher in cells co-expressing Syk and the Src kinases compared to cells expressing Syk alone (Figure 4.1). Therefore, the phosphorylation state of SHP-2 was examined by Western blotting using antibodies that specifically recognize SHPr2 that is phosphorylated on the tyrosine 580 residue.  The  phosphorylation of SHP phosphatases is thought to result in increased phosphatase activity (Vogel et al., 1993). However, interaction of its SH2 domains with tyrosine phosphorylated proteins has also been shown to enhance phosphatase activity by inducing a conformational change (reviewed by Barford and Neel, 1998). Since SHP-2 is a potential negative regulator, it was hypothesized that SHP-2 phosphorylation and activity would be higher in the AtT20 BCR+ Fyn+ Syk+ Lyn+ and AtT20 BCR+ Fyn+ Syk+ Blk+ cell lines compared to cells that did not exhibit suppressed Akt phosphorylation following BCR engagement.  The phosphorylation of SHP-2 on tyrosine 580 was higher in cells co-expressing Syk and Lyn (AtT20 BCR+ Fyn+ Syk+ Lyn+ cells) compared to the other cell lines (Figure 4.2). SHP-2 phosphorylation in the AtT20 BCR+ Fyn+ Syk+ Lyn+ cells was approximately 2-fold greater than in the other cell lines (Figure 4.2 graph). Surprisingly, cells cotransfected with Syk and Blk (AtT20 BCR+ Fyn+ Syk+ BIk+) did not exhibit higher tyrosine phosphorylation of SHP-2 compared to cells not expressing Blk (AtT20 BCR+ Fyn+ Syk+) even though Akt phosphorylation following BCR stimulation was also suppressed in this cell line. Therefore, the Lyn-SHP-2 interaction and the subsequent phosphorylation of SHP-2 may be responsible for the inhibition of BCR-induced Akt phosphorylation in the AtT20 BCR+ Fyn+ Syk+ Lyn+ cells. However, the Lyn- and Blk-mediated suppression of Akt phosphorylation is most likely due to the ability of these Src kinases to interact with the SHP-2 protein tyrosine phosphatase, and tyrosine phosphorylation of SHP-2 is not absolutely necessary. Alternatively, the inhibition mediated by Blk could occur by a different mechanism than that of Lyn and does not involve SHP-2.  4.2.3 Lyn and Blk associate with the SH2 domains of SHP-2 protein tyrosine phosphatase in vitro  Immunoprecipitation experiments indicated that Lyn and Blk could interact either directly or indirectly with the SHP-2 phosphatase in AtT20 cells (Figure 4.1). However, there was still uncertainty as to whether Lck could interact with SHP-2 since a 65 kDa phosphoprotein was also present in the Lck immunoprecipitates. Therefore, in vitro studies using a GST fusion protein of the tandem SH2 domains of SHP-2 (GST-SHP-2 (SH2-SH2)) were performed.  Lysates from AtT20 transfected cells that were  unstimulated and BCR-stimulated were incubated with the GST-SHP-2 (SH2-SH2) fusion protein. The fusion protein complexes were collected and the presence of Lyn,  Figure 4.2 Phosphorylation of SHP-2 on tyrosine 580 was slightly higher in AtT20 cells co-expressing Syk and Lyn PTKs. The cells were serum-starved for 16 hours in media containing 0.2% FCS before being stimulated for 5 minutes with 20 pg/ml antimouse IgM antibodies (+) or left unstimulated (-). Following lysis, 40 pg of total cell extract was separated by SDS-PAGE and analyzed by Western blotting with antiphospho-tyrosine 580 SHP-2 antibodies.  The filters were subsequently stripped of  antibodies and re-probed with anti-SHP-2 antibodies (lower panels). The bands were quantified using the ImageQuant 5.1 software. These results are representative of five independent experiments with similar results.  +  c +  DC +  O m cn o + CM I- c < U-  +  >  00 +  +  c  o  -J  LL  cc  CD + o eg </) I-  +  + C r—, + *  cc  +  C\J  +  o o CQ -J O  cr  + o £ m CD o + CM  Anti-IgM (5 min)  80-  •p-SHP-2  61 -  IB: anti-phospho-tyrosine 580 SHP-2  Syk+  Syk+Lyn+  Syk+Lck+  Syk+Blk+  Lck and Blk was examined in the complexes. Lyn and Blk interacted with the GST-SHP2 (SH2-SH2) fusion protein (Figure 4.3 A and B), thus confirming that Lyn and Blk can associate with the SH2 domain(s) of SHP-2. These interactions appear to be constitutive and were not completely dependent on BCR activation. However, more Lyn and Blk molecules associated with the GST-SHP-2 (SH2-SH2) fusion protein following BCR engagement.  Thus, BCR cross-linking may result in the creation of tyrosine  phosphorylated sites on Lyn, Blk or some other protein that interacts with these Src kinases, where the SH2 domain(s) of SHP-2 could interact. No Lck interacted with the GST-SHP-2 (SH2-SH2) fusion protein from the AtT20 BCR+ Fyn+ Syk+ Lck+ cells, confirming the previous interpretation of the results (Figure 4.1 B) that Lck and SHP-2 do not associate with one another (Figure 4.3 C). These results clearly show that Lyn and Blk can interact with the SH2 domains of SHP-2, but Lck cannot. However, it is possible that Lck could interact through some other portion of SHP-2 outside of its SH2 domains. Since Lyn and Blk associated with the GST-SHP-2 (SH2-SH2) fusion protein in vitro, the ability of this fusion protein to inhibit the interaction between these transfected Src kinases and the endogenous SHP-2 in the AtT20 cells was examined. Lyn, Lck and Blk were immunoprecipitated from unstimulated and BCR-stimulated cell lysates in the presence or absence of an excess amount of the GST-SHP-2 (SH2-SH2) fusion protein. The immune complexes were precipitated and assayed for the presence of SHP-2. No SHP-2 co-immunoprecipitated with Lyn or Blk in the AtT20 BCR+ Fyn+ Syk+ Lyn+ and AtT20 BCR+ Fyn+ Syk+ Blk+ cell lines, respectively, when the GST-SHP-2 (SH2SH2) fusion protein was added to the immunoprecipitations (Figure 4.4, lower panels reprobe with anti-SHP-2 antibodies). Therefore, the interaction between Lyn and Blk with endogenous SHP-2 was clearly disrupted by the addition of the GST-SHP-2 (SH2-SH2) fusion protein. These precipitation studies provide evidence that SHP-2 associates, most likely via its SH2 domain(s), with Lyn and Blk, but not Lck. Since the GST fusion protein did not contain the rest of the SHP-2 protein, it cannot be ruled out that there are additional regions of SHP-2 that could also interact with the Src kinases.  Figure 4.3 The SH2 domain(s) of SHP-2 mediate its interaction with Lyn and Blk in vitro. AtT20 transfected cells were serum-starved overnight in media containing 0.2% FCS prior to being stimulated with 20 pg/ml of anti-IgM antibodies for 5 minutes (+) or left unstimulated (-).  Following cell lysis and calculation of protein concentration by  BCA assay, 1 mg of total cell lysate was pre-cleared with GST bound to glutathioneSepharose 4B beads for 1 hour at 4°C. Glutathione-Sepharose 4B beads were incubated with 50 pi of bacterial cell lysate containing GST-SHP-2 (SH2-SH2) fusion protein for 1 hour at 4°C. After washing twice with lysis buffer, the immobilized GST-SHP-2 (SH2SH2) fusion protein was incubated with the pre-cleared lysates and incubated at 4°C for 2 hours. The complexes were collected by centrifugation and washed two times with lysis buffer. The bound proteins were eluted with IX SDS-PAGE reducing sample buffer, separated by SDS-PAGE and analyzed for the presence of (A) Lyn, (B) Blk or (C) Lck by Western blotting. These results are representative of three experiments.  (A) PPT: G S T - S H P - 2 ( S H 2 - S H 2 ) IB: anti-Lyn AtT20 BCR+ Fyn+ Syk+  AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8]  AtT20 BCR+ Fyn+  (B) PPT: G S T - S H P - 2 ( S H 2 - S H 2 ) IB: anti-Blk AtT20 BCR+ AtT20 BCR+ Fyn+ Fyn+ Syk+ Syk+ Blk+ [#17] Anti-IgM (5 min)  AtT20 BCR+ Fyri+ Syk+ Blk+ [#17] (whole cell lysate)  (C) PPT: G S T - S H P - 2 ( S H 2 - S H 2 ) IB: anti-Lck AtT20 BCR+ Fyn+ Syk+ Anti-IgM (5 min)  64-  AtT20 BCR+ Fyn+ Syk+ Lck+ [#17]  AtT20 BCR+ Fyn+ Syk+ Lck+ [#17] (whole cell lysate)  Figure 4.4 SHP-2 interaction with Lyn and Blk is inhibited by SHP-2 GST fusion protein. The cells were serum-starved for 16 hours in media containing 0.2% FCS. The cells were stimulated with anti-IgM antibodies for 5 minutes (+) or left unstimulated (-) prior to cell lysis.  SHP-2 and the Src kinases, Lyn, Lck and Blk, were  immunoprecipitated from total cell protein for 2 hours at 4°C. For one set of samples, bacterial cell lysate containing GST-SHP-2 (SH2-SH2) fusion protein was added to the immunoprecipitations and incubated for 2 hours at 4°C (+ GST-SHP-2 (SH2-SH2) lanes). Following incubation, the immune complexes were collected by adding Protein A-Sepharose and incubating the samples for an additional hour. The complexes were then collected by centrifugation, washed two times with lysis buffer and the bound proteins were eluted with IX SDS-PAGE reducing sample buffer.  All of the  immunoprecipitates were separated by SDS-PAGE and analyzed by Western blotting with 4G10 anti-phosphotyrosine monoclonal antibody, which clearly shows that the immunoprecipitated Src kinases were tyrosine phosphorylated.  The filters were  subsequently stripped of antibodies and re-probed with anti-SHP-2 antibodies (lower panels). The bands were visualized by enhanced chemiluminescence. The molecular weight standards (in kDa) are indicated to the left. (*) indicates the position of the immunoprecipitating antibodies. The figures shown above are representative of two experiments.  GST-SHP-2(SH2-SH2)  c >  c  c  C >  c  c  c  ,—.  c  CD J_  c  c  IPPT: +  c >> s r  +  + —  CC  ^ a  oCD O t  Anti-IgM (5 min)  ON o  114-  +  +  c >. LL  . i  CO  CO  CO  i >. ^r  m o . t  i  < Cft < CO . + l . +  c  >. hU- T+ 2* cc + oCD £ OQ  CO  CO  +  >.  L+L tr 0 o + C M ± CQ t< to >. O . CVJ+ Jl >. l - +11 t.< {/)  +  CO i  c  IPPT:  CO +  c>L>,—. L + 2L 1 i cc + CD 5 o J£ C o . oD .CQ C M 1 C M J < (ft t< C>ft t > - + l + l  cL 0 0 L + *—  Anti-IgM (5 min)  11481 -  81 -  64-  64-  Lyn  *=<Blk  50-  50  IB: anti-phospho-tyrosine (4G10) 81 -  _  — —  . 81 -background band  —SHP-2  64-  Re-probe: anti-SHP-2  64  + GST-SHP-2(SH2-SH2)  c  c >>  c  c  CO  CO +  +  cL> L>  L ^cL *CO  +  cc o oO C  + —'  ° 5. CQ 5 ° 4.  ,  < CO -  +  < CO  tttt  r  c>  c  o  c  c  c  c  CO  CO  CO  + c  + c  o  +  LL  Si o m o + < o CM c  >>  +  DC  O CQ O CM + C* < CO -  +  CO  +  ^  + 00 U_ + *c  cc o  m  "I  >»  c ><  LL + +  cn o jc CQ o O+  J+ C CM cn co I - + I- + l o  CM  IB: anti-phospho-tyrosine (4G10)  <  Re-probe: anti-SHP-2  background band SHP-2  4.3  Discussion  SHP-1 has been shown by several investigators to play a role in the negative regulation of BCR signaling (Pani et al, 1995; Cyster and Goodnow, 1995; Ono et al, 1997; Dustin et al, 1999; reviewed by Tamir et al, 2000; Zhang et al, 2000). SHP-1 can negatively regulate BCR signaling through its ability to interact with inhibitory receptors such as Fc receptors (D'Ambrosio et al, 1995; Gupta et al., 1997), CD22 co-receptor (Doody et al., 1995; Cornall et al,  1998; Smith ^ al,  1998; Blasioli et al,  1999) and the paired  Immunoglobulin-like receptor (PIR)-B (Kabagawa et al, 1997; Blery et al, 1998; Maeda et al, 1998; Maeda et al, 1999). The ability of SHP-1 to negatively regulate signaling may also be due to its ability to interact with Lyn or Syk in B lymphocytes (Dustin et al, 1999; Somani et al, 2001). This interaction results in the de-phosphorylation of specific tyrosine residues on the PTKs, leading to decreased catalytic activity. SHP-1 has also been shown to interact directly with the Ig-a and Ig-$ BCR accessory chains in resting B lymphocytes through a SH2 domain-independent mechanism (Siminovitch et al, 1999). These SHP-1 molecules de-phosphorylate the tyrosine residues within the ITAM motifs of Ig-a and Ig-P chains under resting conditions and dissociate from the BCR chains once the receptor is cross-linked. Following BCR cross-linking, however, it is thought that SHP-1 molecules also de-phosphorylate the tyrosine residues within the ITAM motifs of the BCR accessory chains. In this way, SHP-1 negatively regulates signaling through the BCR both before and after receptor activation. Since the role of SHP-1 in B lymphocytes has been well-established, it was hypothesized that this tyrosine phosphatase may interact with the transfected Src kinases in AtT20 cells and lead to the down-regulation of PI 3-kinase signaling initiated by the BCR. An interaction between SHP-1 and the Src kinases was examined by co-immunoprecipitation experiments. However, it was concluded that SHP-1 does not interact with the Src family kinases Lyn, Lck and Blk in the various transfected AtT20 cell lines (data not shown).  Therefore, the ability of SHP-2, a close family member of SHP-1, to interact with the various Src kinases was examined. Immunoprecipitation experiments showed that SHP-2 can interact with Lyn and Blk in the Syk- and Src-transfected AtT20 cells (Figure 4.1).  In vitro studies using a GST fusion protein of the tandem SH2 domains of SHP-2 established that both Lyn and Blk could indeed interact with SHP-2 (Figure 4.3). In addition, this experiment established that the interaction between Lyn and Blk with SHP-2 involves at minimum the SH2 domain(s) of this phosphatase and that these domains were sufficient for mediating this interaction. The addition of the SH2 domain fusion protein to cell extracts in an immunoprecipitation experiment interfered with the interaction between the transfected Lyn and Blk and the endogenous SHP-2 molecules (Figure 4.4). SHP-2 phosphorylation on the tyrosine 580 residue, which is located at the carboy-terminus (Figure 4.5 A), was 2-fold greater in AtT20 cells co-expressing Syk and Lyn (AtT20 BCR+ Fyn+ Syk+ Lyn+) compared to the other cell lines examined (Figure 4.2).  Some evidence suggests that SHP-2 activity is up-regulated by tyrosine  phosphorylation (Vogel et al, 1993). However, other investigators have shown that activation is regulated by conformational changes of the phosphatases. Defining the crystal structure of SHP-2 revealed that phosphatase activity is regulated by the aminoterminus SH2 domain (Hof et al., 1998). Prior to activation, a portion of the aminoterminus SH2 domain is wedged into the mouth of the phosphatase domain thereby blocking its activity (Figure 4.5 B). Biochemical evidence also suggests that the aminoterminus SH2 domain of SHP-2 is important for its phosphatase activity as deletion of this SH2 domain in both SHP-2 and SHP-1 results in dramatically higher phosphatase activity (Townley et al., 1993; Pei et al., 1994; Zhao et al., 1994; Dechert et al., 1996). The tandem SH2 domains have been shown to interact directly with the phosphatase domain of SHP-2 in vitro.  Addition of the ligand for the SH2 domains was able to  disrupt its interaction with the phosphatase domain (Dechert et al., 1996). As well, high affinity ligands for the SH2 domains of SHP-2 are able to relieve the inhibition of phosphatase activity, resulting in its full activation (Sugimoto et al., 1993; Pluskey et al., 1995; Dechert et al, 1996). These results indicate that interaction of SHP-2 with its ligand leads to a conformational change subsequently resulting in increased phosphatase activity due to the release of the phosphatase domain (Figure 4.5 B).  Although a  conformational change is important for SHP-2 activation, phosphorylation of its tyrosine residues at the carboxy terminus may act to enhance the activation of the phosphatase activity. This tyrosine phosphorylation may also be important for the recruitment of  Figure 4.5 Schematic representation of SHP-2 structure and mechanism of activation. (A) Structure of SHP-2 protein tyrosine phosphatase. SHP-2 has two SH2 domains, amino-terminus (N) and carboxy-terminus (C), at the amino terminal end. Two tyrosine residues that are phosphorylated (pY) are located at the carboxy terminus. The tyrosine 580 residue is closest to the carboxy terminal.  In addition, a proline-rich  sequence (PRS) located in between these tyrosine residues. (B) Mechanism of SHP-2 activation. In the inactive conformation, the backside (i.e. not the side that interacts with the phosphorylated tyrosine sites) of the amino-terminus SH2 domain interacts with the phosphatase domain in a phosphotyrosine-independent manner. Following interaction with phosphorylated tyrosine residues located within the consensus sequence recognized by the SH2 domains, the interaction between the amino-terminus SH2 domain and phosphatase domain is perturbed resulting in the activation of SHP-2's phosphatase activity. It is not known whether interaction of the two SH2 domains of SHP-2 are required for its full activation or if interaction of the amino-terminus SH2 with its binding partner is sufficient for this activation.  (A)  SH2 domain ligands ~  (B) ON  Inactive conformation  Active conformation  other signaling molecules including the Grb2 adapter protein, which ultimately leads to the activation of the Ras/MAP kinase signaling pathway (Li et al., 1994). SHP-2, unlike SHP-1, has a proline-rich region at the carboxy terminus which may be important for the recruitment of SH3 domain-containing signaling components (reviewed by Siminovitch and Neel, 1998). Thus SHP-2 also appears to function as an adapter protein by recruiting various signaling molecules.  Data accumulated on SHP-2 suggests that it functions as a positive rather that negative regulator of signaling through various receptors. For example, SHP-2 can recruit Grb2 and thus participate in the activation of the small GTP-binding protein Ras and ultimately MAP kinase (Li et al, 1994). Furthermore, EGF receptor-induced ERK activation can be inhibited by expression of catalytically inactive SHP-2 mutant in human embryonic cells, indicating that SHP-2 can positively regulate ERK activation (Bennett et al, 1996). In addition, the Drosophila homolog of SHP-2, corkscrew,  is required for signaling  downstream of several tyrosine kinase receptors including the FGF receptor (reviewed by van Vactor et al, 1998). The role of SHP-2 in B lymphocytes is not yet clearly defined (reviewed by Tamir et al., 2000).  Evidence thus far, however suggests that its family  member, SHP-1, plays a more prominent role in hematopoietic cells.  These results obtained in the AtT20 cell system suggest that SHP-2 plays a negative rather than positive regulatory role in BCR signaling. This negative regulatory role of SHP-2 is due to its ability to interact with the Src family members, Lyn and Blk. This interaction is not mediated by the SH3 domain of the Src kinases and the proline-rich region at the carboxy terminal of SHP-2, as observed in myeloid cells (Yoshida et al, 1999).  The in vitro experiments using the GST-SHP-2 (SH2-SH2) fusion protein  indicated that SHP-2 interacts with Lyn and Blk through its SH2 domain(s). Whether both or only one of these SH2 domains is required for this interaction remains to be established. Generation of GST fusion proteins containing only the amino or carboxy terminal SH2 domain would be useful in establishing the precise mechanism of SHP-2's interaction with Lyn and Blk. Since the SH2 domains of SHP-2 are involved in this  interaction with Lyn and Blk, then these Src kinases must contain tyrosine phosphorylated motif(s).  An interesting finding was made upon examination of the  sequences of all murine Src kinase family members. Lyn and Blk both have tyrosine residues within their kinase domains (Y433 in Lyn and Y424 in Blk in both the human and murine forms) that lie within consensus motifs that are recognized by the SH2 domains of SHP-2 (Figure 4.6).  The SH2 domains of SHP-2 and SHP-1 bind to  phosphorylated tyrosine residue within the consensus immunoreceptor tyrosine-based inhibitory motif (ITIM) sequence I/V/L/S-X-  Y-X-X-L/V  (reviewed by Unkeless and Jin,  1997; Daeron and Vivier, 1999). The presence of hydrophobic residues at the - 2 and +3 positions (with respect to the phosphorylated tyrosine residue) are required for efficient binding of the SH2 domains (Olcese et al., 1996; Burshtyn et al., 1997). An addition, an ITIM sequence was found in Lyn directly beside the first motif (V-T-Y438-G-K-I) (Figure 4.6). The amino acid in the +3 position does not exactly fit with the consensus ITIM recognized by SH2 domains of SHP phosphatases. However, the isoleucine at the +3 position is a hydrophobic residue. Therefore, this motif can most likely be recognized by the SH2 domains of the SHP phosphatases. Since Lyn has two potential SHP-2 binding sites, the tandem SH2 domains may be able to interact with Lyn at the same time thus creating a stronger association while only one SH2 domain can interact with Blk. The fact that both SH2 domains of SHP-2 can mediate the interaction with Lyn may explain why tyrosine phosphorylated Lyn molecules, compared to Blk molecules can be more readily detected in the anti-SHP-2 immunoprecipitates (Figure 4.1).  However, no  difference in the amount of SHP-2 in the Lyn and Blk co-immunoprecipitates was detected. The SHP-2 molecules in the AtT20 BCR+ Fyn+ Syk+ Lyn+ and AtT20 BCR+ Fyn+ Syk+ Blk+ cells are probably active since their SH2 domains are bound to phosphotyrosine sites on Lyn and Blk.  The binding of the SH2 domains with  phosphotyrosine sites on Lyn and Blk enables SHP-2 to resume its the active conformation (Figure 4.5 B). However, in the AtT20 BCR+ Fyn+ Syk+ Lck+ cells, the SH2 domains of SHP-2 are not occupied and thus the SHP-2 molecules remain in the inactive conformation.  This ITIM consensus sequence was also present in the Src family member, Hck (Figure 4.6). The sequence in Hck was similar to the second ITIM motif in Lyn (V-T-Y451-G-RI). This motif, like those found in Lyn and Blk lies within the kinase domain of Hck. The presence of this motif suggests that Hck may also be able to interact with SHP-2. No ITIM motifs were found in the other six Src family members examined (alignments of the complete murine sequences of all Src kinases is shown in the Appendix, Figure 1). This sequence analysis data and the finding that SHP-2 binding sites are present in Lyn and Blk, but not Lck are consistent with the interpretation of the results obtained in the immunoprecipitation and in vitro association experiments performed in the transfected AtT20 cell lines. Since the SH2 domains of SHP-2 and SHP-1 recognize the same consensus sequence, then in theory, SHP-1 should also be able to interact with Lyn and Blk in the AtT20 cells. However, no SHP-1 was detected in co-immunoprecipitation experiments in the Lyn- and Blk-transfected cells (data not shown). Previous reports, however, have indicated that the SH2 domains of SHP-1 and SHP-2 bind to phosphorylated ITIM sequences with different affinities (Olcese et al., 1996; Burshtyn et al., 1997; Vely et al., 1997; Famiglietti et al., 1999). In addition, the binding of the SH2 domains of the SHP phosphatases to phosphotyrosine-containing proteins may be influenced by residues outside the consensus ITIM sequences (Bruhns et al., 1999).  The association of Lyn and Blk with SHP-2 may be an important mechanism for recruiting phosphatases to the plasma membrane. In AtT20 cells, SHP-2 recruitment to the plasma membrane by Lyn and Blk may be responsible for the observed suppression of Akt phosphorylation following BCR cross-linking. Perhaps SHP-2 at the plasma membrane dephosphorylates the adapter protein that recruits PI 3-kinase. Dephosphorylation of this adapter protein would result in the elimination of docking sites for PI 3-kinase and its subsequent release back into the cytoplasm. This would ultimately lead to a decrease in PI-3,4,5-P3 and dissociation of PH domain-containing proteins (including Akt) from the plasma membrane (for further details, refer to "Proposed model" in Chapter 8 Discussion). In summary, the data obtained suggests that Lyn and Blk can associate with SHP-2 and this may be the reason why BCR-induced Akt phosphorylation is suppressed in AtT20 cells expressing these Src family kinases.  Figure 4.6 Amino acid sequence of putative ITIM motifs within the kinase domain of the Src family members. The amino acid sequences containing the putative motif within the tyrosine kinase domains that mediates SHP-2 interaction of the various Src kinases is indicated. The consensus binding sequence for the SH2 domains of SHP-2 is also shown. "X" signifies to any amino acid. The tyrosine residues within the consensus motifs are blue; the green amino acids form the consensus ITIM motif. The sequences were aligned using the ClustalW multiple sequence alignment program from the European Bioinformatics Institute (http://www.ebi.ac.uk/clustalw/).  Consensus ITIM sequence:  l/V/L/S-X-Y-X-X-L/V  Lyn Blk Hck Lck Src Yes Fyn Fgr Frk  mouse mouse mouse mouse mouse mouse mouse mouse human  429416443427457458450434421-  ILLYEIVTYGKIP VLLMVIVTYGRVP ILLMEIVTYGRIP ILLTEIVTHGRIP ILLTELTTKGRVP ILQTELVTKGRVP ILLTELVTKGRVP ILLTELITKGRVP ILLYEIITYGKMP  -442 -429 -456 -440 -470 -471 -463 -447 -434  CHAPTER 5 The kinase activity of Lyn is required for BCRinduced inhibition of Akt phosphorylation and activation  5.1  Introduction  Syk and Lyn PTKs are important for the initiation of BCR signaling. The fact that both of these kinases participate in the initiation of signaling through the BCR suggests that they both have positive regulatory roles in B lymphocyte activation. However, more and more evidence is accumulating that suggests that Lyn may play a role in negative regulation of BCR signaling (Wang et al, 1996; Chan et al, 1997; Katsata et al, 1998; reviewed by DeFranco et al, 1998). Indeed, the work in the AtT20 cell system described in the previous chapters shows that Lyn has an inhibitory effect on BCR-mediated Akt phosphorylation and activation (Chapter 3). This inhibitory effect is presumably due to Lyn's ability to associate with the SHP-2 protein tyrosine phosphatase (Chapter 4). SHP2 phosphorylation on a specific tyrosine residue (tyrosine 580) in AtT20 cells cotransfected with Syk and Lyn (AtT20 BCR+ Fyn+ Syk+ Lyn+ cells) was approximately 2-fold greater compared to the other transfected cell lines examined (Figure 4.2). This suggests that Lyn may be partially responsible for SHP-2 phosphorylation, leading to the enhancement of its phosphatase activity. Furthermore, it was established previously that the kinase activity of Syk is required for amplifying and sustaining BCR-induced Akt phosphorylation in Syk-expressing AtT20 cells. In this section, the role of the catalytic activity of Syk and Lyn in BCR-mediated Akt phosphorylation and activation was further examined.  To this end, AtT20 cell lines expressing different combinations of wild type and catalytically inactive mutant forms of Lyn and Syk were generated. The kinase dead  form of Lyn was still able to interact with the SHP-2 phosphatase. However, the catalytic activity of Lyn is required for enhanced SHP-2 phosphorylation. The levels of Akt phosphorylation in AtT20 cells co-expressing wild type Syk and kinase dead Lyn were considerably higher compared to cells co-expressing wild type forms of both PTKs. Thus, the kinase activity of Lyn is also required for BCR-induced suppression of Akt phosphorylation.  5.2  Results  5.2.1  The kinase activity of Lyn is required for BCR-mediated suppression of Akt phosphorylation and activation  Since Syk kinase activity is required for enhanced and sustained Akt phosphorylation following BCR cross-linking, it was of interest to determine if Lyn kinase activity is also important for the regulation of this particular pathway. Plasmids containing the genes encoding for catalytically inactive forms of Lyn and Syk were generated.  The  catalytically inactive or kinase dead Syk plasmid was generated by altering the lysine at position 396 to an alanine by site-directed mutagenesis, as described previously (Richards et al., 1996). To generate the plasmid containing a catalytically inactive form of Lyn, the lysine at position 275 was mutated into an alanine by site-directed mutagenesis. Various cell lines that expressed different combinations of wild type and catalytically inactive mutant forms of Lyn and Syk were created (Appendix, Table 1). To create cells coexpressing wild type Syk and kinase dead Lyn (AtT20 BCR+ Fyn+ Syk+ KD Lyn+), the plasmid encoding for catalytically inactive Lyn (RSVpLpA-KD Lyn) was transfected into AtT20 BCR+ Fyn+ Syk+ [#13] cells using the calcium phosphate method. Multiple Lynpositive clones were obtained from the transfection, however, clone #8 was chosen for further experimentation because it expressed similar levels of Lyn as the AtT20 BCR+ Fyn+ Syk+ Lyn+ cells (Figure 5.1 A). AtT20 cells co-expressing catalytically inactive Syk and wild type Lyn (AtT20 BCR+ Fyn+ KD Syk+ Lyn+) were generated by transfecting the AtT20 BCR+ Fyn+ KD Syk+ [#17.16] (a sub-clone of AtT20 BCR+  Fyn+ KD Syk+ [#17] previously described in Richards et al., 1996) with the expression vector containing the gene for wild type Lyn (pMSCV-Lyn  or RSV pLpA-Lyn). Multiple  clones expressing wild type Lyn were obtained from the transfection. However, since the AtT20 BCR+ Fyn+ KD Syk+ [#17.16] was still a mixed clone (i.e. did not originate from an individual transfected cell), some clones expressed very little or no catalytically inactive Syk. For example, clone #3 from the transfection expressed approximately similar levels of wild type Lyn as the AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8] cell line, however, no kinase dead Syk could be detected (Figure 5.1 B). Fortunately, two clones (#10 and #11) expressed high levels of both wild type Lyn and catalytically inactive Syk (Figure 5.1 B and E). The levels of kinase dead Syk in these two clones, however, was slightly lower than the wild type Syk present in the AtT20 BCR+ Fyn+ Syk+ [#13] and AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8] cells. The level of Lyn expression in clone #10 was similar to that in the AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8] cells while that in clone #11 was lower (Figure 5.1 D). Experiments were conducted on clone #10 because Lyn expression in these cells resembled that of the AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8] clone. Finally, cells co-expressing catalytically inactive forms of both Lyn and Syk were generated by transfecting the AtT20 BCR+ Fyn+ KD Syk+ [#17.16] mixed clone with the plasmid encoding for the kinase dead Lyn mutant (RSVpLpA-KD Lyn). Only two kinase dead Lyn positive clones were obtained from this transfection (#17 and #18), fortunately they both expressed high levels of kinase dead Lyn (Figure 5.1 C). The levels of kinase dead Lyn found in these clones were similar to the levels of wild type Lyn in the AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8] cell line (Figure 5.1 D). These clones also expressed the same levels of kinase dead Syk, which was slightly higher than the amount found in the parental AtT20 BCR+ Fyn+ KD Syk+ [#17.16] mixed clone and slightly lower than the amount of wild type Syk in the AtT20 BCR+ Fyn+ Syk+ [#13] and AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8] cell lines (Figure 5.1 E). The BCR-induced tyrosine phosphorylation of total cell lysate was first examined in the various cell lines created to ensure that the BCR was functional and could induce a tyrosine phosphorylation cascade following receptor cross-linking. Phosphorylation of  Figure 5.1 Expression levels of wild type and kinase dead forms of Lyn and Syk and BCR-induced tyrosine phosphorylation of total cellular protein in transfected AtT20  cells. (A) - (E) Expression of wild type and kinase dead forms of Lyn and Syk in transfected AtT20 cells. Whole cell extract (30-40 pg) from the different cell lines was separated by SDS-PAGE and analyzed by Western blotting with anti-Lyn or anti-Syk antibodies.  (F) BCR-induced tyrosine phosphorylation of total cellular protein in  transfected AtT20 cells. The different cell lines were stimulated with 20 pg/ml of antimouse IgM antibodies for 5 minutes (+) or left unstimulated (-). Following cell lysis, 20 pg of total cell extract was separated by SDS-PAGE and analyzed by Western blotting with the anti-phosphotyrosine monoclonal antibody (4G10). The bands were visualized by enhanced chemiluminescence. The molecular weight standards (in kDa) are indicated to the left.  +  je >•  V) +  (A)  s, LL +  CC o CD O  + CO &  +g c >>£J. + —" >l. C+C _ yO m co 2 t < i</> o . < (0 +  CM  (B) 50  Lyn  80  — i u. i—' u. r>- t l + =• GC -I CC -I c+c a* cc + CQ Jt _ O <* S t* Q < C* 5 < i(/) —  =.  L L C+ C + C X JE > OT 8 8 h§ «C  DC  CO  o  u. ^  -j  C E U m  (D)  Lyn  50CO  c  +  CC 8  u.  (E)  80-  +  &  cc K o g  >«  £o LcL*00 U>-. N u. -J + —' C+ C IT c+c Q* C Q cc + .i co £ 8 OTSot o * tQ Si < * < OT  < ^  Syk  + JC  co  >. +  U. §0 +  i  c  i f IT _l cc . 8. o >• 8 t Q^ t o < t Q* <  (C)  *  u_ 0 *0 >. + CC +— LL  < *  Lyn  c >. LL + Q CC  " 5. o o co CQ CD °  +  P t >Jc < CO  Syk  IB: anti-phospho-tyrosine (4G10)  *  + c+  c+  +  O  j£ >. CO  H <  Q *  LL - 1 + CC  Q *  o + CO  >> °o <f>  t <  a *  total cellular protein was attenuated in the AtT20 BCR+ Fyn+ KD Syk+ Lyn+ [#10] and AtT20 BCR+ Fyn+ KD Syk+ KD Lyn+ [#17] cells compared to cell that contained wild type Syk (Figure 5.1 F). This was expected since Syk is responsible for most of the tyrosine phosphorylation in the AtT20 cells and Syk in these particular cell lines was catalytically inactive. Surprisingly, the tyrosine phosphorylation of.total cellular protein in the AtT20 BCR+ Fyn+ Syk+ KD Lyn+ [#8] cells was also attenuated. The presence of a tyrosine phosphorylated 65-70 kDa band, presumably Syk, was apparent in all cell lines. This protein was variably phosphorylated depending on the combination of PTKs that were expressed in the cell line. Akt phosphorylation following BCR engagement was examined in the different cell lines to determine the importance of catalytic activity of the two types of PTKs in the regulation of this specific pathway.  Comparisons were made between cell lines  expressing the kinase dead versions of Syk and Lyn and those expressing the wild type versions and these results are described below and in Figure 5.2. BCR-induced Akt phosphorylation in cells co-expressing wild type Syk and kinase dead Lyn (AtT20 BCR+ Fyn+ Syk+ KD Lyn+ [#8]) was no longer enhanced and sustained when the activity of Lyn was knocked out (Figure 5.2 A, left panels). However, compared with the inhibition of Akt phosphorylation normally seen, the extent of phosphorylation was partially restored (Figure 5.2 A, right panel). Quantification of the bands at all time points revealed that phosphorylation of Akt in the AtT20 BCR+ Fyn+ Syk+ cells was only 1.3 fold greater compared to the AtT20 BCR+ Fyn+ Syk+ KD Lyn+ [#8] cells, as quantified using the ImageQuant 5.1 program. The major difference in the patterns of Akt phosphorylation between these two cell lines, however, was that phosphorylation in the AtT20 BCR+ Fyn+ Syk+ KD Lyn+ [#8] cells was not sustained following 60 minutes of BCR activation (Figure 5.2 A, left panels). However, the Akt phosphorylation in the AtT20 BCR+ Fyn+ Syk+ KD Lyn+ [#8] cell line was higher compared to that observed in the cell co-expressing the wild type forms of both PTKs (AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8] cells) (Figure 5.2 A, right panels). The Akt phosphorylation in the AtT20 BCR+ Fyn+ Syk+ KD Lyn+ [#8] was 2-3 fold higher compared to the AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8] cells. These results suggest  that the kinase activity of Lyn is important for the inhibition of BCR-induced Akt phosphorylation. One possible interpretation of this data is that Lyn could phosphorylate a component in the signaling complex that is used to recruit a negative regulator of this event, perhaps interfering with the recruitment of SHP-2 as shown in Chapter 4. Although Lyn kinase activity is important for the inhibition of BCR-induced Akt phosphorylation, its kinase-independent functions may also be important for this inhibition. BCR-induced Akt phosphorylation is completely abolished in the AtT20 BCR+ Fyn+ KD Syk+ Lyn+ [#10] cells (Figure 5.2 B). This is consistent with the trend observed when the AtT20 BCR+ Fyn+ KD Syk+ [#17] cell line was first examined in Chapter 3 (see Figure 3.4 B). Perhaps the difference between the results in Figure 3.4 B and those here is the level of expression of the catalytically inactive Syk in these two subclones. Very little Akt phosphorylation was detected in the cells and no increase in this phosphorylation was observed following BCR activation. This phosphorylation was even lower compared to the AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8] cells (Figure 5.2 B, right panels). As shown in a previous chapter, BCR-induced Akt phosphorylation was barely detectable in cells expressing only catalytically inactive Syk (AtT20 BCR+ Fyn+ KD Syk+ [#17]; Figure 3.4 B). Thus, the lack of Akt phosphorylation in the AtT20 BCR+ Fyn+ KD Syk+ Lyn+ [#10] cells can be attributed to the lack of Syk activity. In addition, the Lyn molecules in these cells are catalytically active and thus may still be involved in the negative regulation of the Akt phosphorylation, presumably by participating in the recruitment and activation of negative signaling components like SHP-2. Finally, BCR-mediated Akt phosphorylation was examined in cells co-expressing catalytically inactive forms of both Syk and Lyn. As expected, the levels of Akt phosphorylation in the AtT20 BCR+ Fyn+ KD Syk+ KD Lyn+ [#17] cells were lower at all time points of BCR cross-linking compared to the AtT20 BCR+ Fyn+ Syk+ cells (Figure 5.2 C, left panels). This is again consistent with the finding that Syk activity is required for sustained and enhanced Akt phosphorylation following BCR engagement.  Figure 5.2 Lyn kinase activity is required for BCR-induced suppression of Akt  phosphorylation in the AtT20 cells. The various transfected AtT20 BCR+ Fyn+ cells were serum-starved for 16 hours in media containing 0.2% FCS before being stimulated with 20 pg/ml of anti-mouse IgM antibodies for the indicated time periods. Following lysis, 40 pg of total cell extract was separated by SDS-PAGE analyzed by Western blotting with anti-phospho-serine 473 Akt antibodies. The filters were subsequently stripped of antibodies and re-probed with anti-Akt antibodies (lower panels of each pair). The bands were visualized by enhanced chemiluminescence. Molecular weight standards (in kDa) are indicated on the left. These results are representative of three independent experiments with similar results.  ( A \ AtT20 BCR+Fyn+ Syk+ Anti-IgM f — (minutes) I Q  (R\ V  5 15 30 60 I 0  5 15 30 60  AtT20 BCR+ Fyn+  AtT20 BCR+ Fyn+  Syk+  KD Syk+ Lyn+ [#10]  '  Anti-IgM r (minutes) I  AtT20 BCR+ Fyn+ Syk+ KD Lyn+ [#8]  0  5 15 30 6010  5 15 30 60 I  64 -  64-  (Q) V  Anti-IgM r (minutes) I  64  64-  AtT20 BCR+ Fyn+  AtT20 BCR+ Fyn+  Syk+  KD Syk+ KD Lyn+ [#17]  '  0  5 15 30 60  IT 5  15 30 60 1  AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8] Anti-IgM | (minutes) I  IB: anti-phosphoserine 473 Akt  64  Re-probe: anti-Akt  64 -  0  15 30 60 I 0  AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8] Anti-IgM I (minutes) I  IB: anti-phosphoserine 473 Akt  64  Re-probe: anti-Akt  64  0  IB: anti-phosphoserine 473 Akt  64-  Re-probe: anti-Akt  64  0  5 15 30 60  AtT20 BCR+ Fyn-iKD Syk+ Lyn+ [#10]  5 15 30 60 I 0  AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8] Anti-IgM (minutes)  AtT20 BCR+ Fyn+ Syk+ KD Lyn+ [#8]  5 15 30 60  AtT20 BCR+ Fyn-iKD Syk+ KD Lyn+ [#17]  5 15 30 601 0 5 15 30 60  Interestingly, BCR-induced Akt phosphorylation was very similar in the AtT20 cell lines that co-expressed both wild type forms (AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8]) and both mutant forms (AtT20 BCR+ Fyn+ KD Syk+ KD Lyn+ [#17]) of Syk and Lyn (Figure 5.2 C, right panels). Thus inhibition of Akt phosphorylation is greater in cells that contain active Lyn and knocking out the kinase activity of Lyn clearly interferes with its ability to inhibit the activation of this pathway. 5.2.2  Wild type and kinase dead forms of Lyn associate with SHP-2 protein tyrosine phosphatase in transfected AtT20 cells  Since wild type Lyn can associate with the SHP-2 tyrosine phosphatase in AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8] cells, the ability of catalytically inactive mutant form of Lyn to interact with SHP-2 was examined. Lyn and SHP-2 were immunoprecipitated from the various transfected AtT20 cells line containing different combinations of wild type and kinase dead Syk and Lyn. A highly tyrosine phosphorylated band of approximately 55 kDa was present in the SHP-2 immunoprecipitates (Figure 5.3 A). This band is probably Lyn. There was a noticeable difference in the phosphorylation state of Lyn in cells where the kinase activity was mutated (AtT20 BCR+ Fyn+ Syk+ KD Lyn+ [#8] and in AtT20 BCR+ Fyn+ KD Syk+ KD Lyn+ [#17]). Unfortunately, the filters could not be reprobed with the anti-Lyn antibodies since the Lyn band and the heavy chain of the immunoprecipitating antibody migrate to the same position on the gel, as mentioned earlier. The tyrosine phosphorylation state of SHP-2 in the AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8] cells was higher compared to any of the other cell lines examined. In addition, SHP-2 phosphorylation in these cells increased following BCR activation. SHP-2 tyrosine phosphorylation is very low, but constitutive, in the AtT20 BCR+ Fyn+ KD Syk+ Lyn+ [#10] cell line. Tyrosine phosphorylated SHP-2 molecules from AtT20 BCR+ Fyn+ Syk+ KD Lyn+ [#8] and AtT20 BCR+ Fyn+ KD Syk+ KD Lyn+ [#17] are undetectable although SHP-2 was successfully immunoprecipitated from these cells (Figure 5.3 A, lower panel).  Figure 5.3 The wild type and kinase dead forms of Lyn, constitutively associate with the SHP-2 protein tyrosine phosphatase in the transfected AtT20 cells. The  transfected AtT20 cells were serum-starved for 16 hours in media containing 0.2% FCS before being stimulated for 5 minutes with 20 pg/ml anti-mouse IgM antibodies (+) or left unstimulated (-). Following cell lysis and quantification of protein concentration by BCA assay, Lyn was immunoprecipitated with anti-Lyn along with Protein A-Sepharose beads from 2 mg of total cell protein for 2 hours at 4°C. The immune complexes were washed two times with lysis buffer and subsequently eluted with IX SDS-PAGE reducing sample buffer. All of the immunoprecipitates were separated by SDS-PAGE and analyzed by Western blotting with the 4G10 anti-phosphotyrosine monoclonal antibody which specifically recognizes tyrosine phosphorylated residues on proteins. The filters were subsequently stripped of antibodies and re-probed with anti-SHP-2 antibodies (lower panels). The bands were visualized by enhanced chemiluminescence. The molecular weight standards (in kDa) are indicated to the left. (*) indicates the position of the immunoprecipitating antibodies. These results are representative of a two independent experiments with similar results.  + j> t.  co +  + DC  O CQ O c\j  (A)  Anti-lgMI—J (5 min)'  IPPT: anti-SHP-2 IB: anti-phospho-tyrosine (4G10)  J+>. C  (0  o + c§ i*. >• j. 4 a. t £ iT ± O Q u+ CQ 5 CQ ti < CO tQ < *  +  +  cc L>. +L D cr o m o  CM  I - + 1 • +1 - + I • +1  -SHP-2 • Lyn  Re-probe: anti-SHP-2  (B)  -SHP-2  Anti-IgM (5 min), 115  IPPT: anti-Lyn IB: anti-phospho-tyrosine (4G10)  Re-probe: anti-SHP-2  SHP-2  In the reciprocal co-immunoprecipitation experiment, Lyn was immunoprecipitated from the various cell lines and the tyrosine phosphorylation of the proteins in the immunoprecipitates was examined. A tyrosine phosphorylated protein corresponding to the molecular weight of SHP-2 (65 kDa) was present in the AtT20 BCR+ Fyn+ Syk+. Lyn+ [#8] cells (Figure 5.3 B, upper panel). The tyrosine phosphorylation of this protein increased in response to BCR activation. This same protein was also present in the AtT20 BCR+ Fyn+ KD Syk+ Lyn+ [#10] cells.  However, in these cells it was  constitutively tyrosine phosphorylated, but at much lower levels than the protein in the AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8] cells. Re-probing the filter with antibodies that recognize SHP-2 confirmed that this 65 kDa tyrosine phosphorylated protein was in fact SHP-2 (Figure 5.3 B, lower panel). SHP-2 also co-immunoprecipitated with the wild type and kinase dead forms of Lyn although only those SHP-2 molecules from cells expressing the wild type form of Lyn (AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8] and AtT20 BCR+ Fyn+ KD Syk+ Lyn+ [#10]) were tyrosine phosphorylated. Although SHP-2 was present in the Lyn immunoprecipitates from AtT20 BCR+ Fyn+ Syk+ KD Lyn+ [#8] or AtT20 BCR+ Fyn+ KD Syk+ KD Lyn+ [#17] cells, a tyrosine phosphorylated protein in the molecular weight range of SHP-2 was not detectable in these cell lines. These results suggest that the kinase activity of Lyn is important for enhancing tyrosine phosphorylation of SHP-2. However, the Syk kinase activity is also important for SHP-2 phosphorylation.  It is of interest to note that the extent phosphorylation of Lyn-  associated SHP-2 molecules does not correlate with the extent of inhibition of Akt phosphorylation.  5.2.3  The kinase activity of Lyn and Syk are required for BCR-induced phosphorylation of cellular SHP-2  Lyn immunoprecipitation results from the previous section indicated that Lyn kinase activity plays a role in SHP-2 tyrosine phosphorylation since SHP-2 phosphorylation was only observed cells expressing catalytically active Lyn (Figure 5.3 B). In the previous chapter, it was also established that SHP-2 phosphorylation on the tyrosine 580 residue in  AtT20 BCR+ Fyn+ Syk+ Lyn+ cells was 2-fold higher than in the cells transfected with other Src family kinases (Figure 4.2). Therefore, the role of Lyn and Syk kinase activity in BCR-induced SHP-2 phosphorylation on the tyrosine 580 residue was examined in the various wild type and kinase dead Syk- and Lyn-transfected  AtT20 cells.  Phosphorylation of SHP-2 on this residue increased in response to BCR cross-linking in most of the cell lines examined (Figure 5.4). However, in the AtT20 BCR+ Fyn+ KD Syk+ Lyn+ [#10] cells, SHP-2 was constitutively phosphorylated at levels similar to that in the AtT20 BCR+ Fyn+ Syk+ KD Lyn+ [#8] cells.  As shown in the  immunoprecipitation experiments, SHP-2 tyrosine phosphorylation in the AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8] cells was higher than in AtT20 BCR+ Fyn+ Syk+ cells (Figure 5.3 A).  Interestingly, in the AtT20 BCR+ Fyn+ Syk+ KD Lyn+ [#8] cells, SHP-2  phosphorylation was lower compared to AtT20 BCR+ Fyn+ Syk+ cells, indicating perhaps that the presence of the catalytically inactive mutant Lyn within these cells may by suppressing BCR-induced SHP-2 phosphorylation. Finally, SHP-2 phosphorylation in the AtT20 BCR+ Fyn+ KD Syk+ KD Lyn+ [#17] cells was lower than in any of the other cell lines examined, as expected since both PTKs in these cells were catalytically inactive and therefore incapable of phosphorylating anything. The slight increase in SHP-2 tyrosine phosphorylation following BCR activation observed in these cells may be due to the endogenously expressed Fyn within the AtT20 cells.  5.3  Discussion  Using AtT20 cell lines co-expressing various combinations of wild type and kinase dead mutant forms of Syk and Lyn, it was determined that the catalytic activity of Lyn was required for the suppression of BCR-induced Akt phosphorylation. Akt phosphorylation in the cell line expressing wild type Syk and catalytically inactive Lyn (AtT20 BCR+ Fyn+ Syk+ KD Lyn+ [#8]) was considerably higher than in cells co-expressing wild type forms of both Syk and Lyn (AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8]) (Figure 5.2 A). Thus, inhibition of the kinase activity of Lyn in these cells relieved some of the suppression on  Figure 5.4  The kinase activity of both Lyn and Syk are important for the  phosphorylation of SHP-2 oh the tyrosine 580 residue. The cells were serum-starved  for 16 hours in media containing 0.2% FCS before being stimulated for 5 minutes with 20 pg/ml anti-mouse IgM antibodies (+) or left unstimulated (-). Following lysis, 40 pg of total cell extract was separated by SDS-PAGE and analyzed by Western blotting with anti-phospho-tyrosine 580 SHP-2 antibodies. The filters were subsequently stripped of . antibodies and re-probed with anti-SHP-2 antibodies (lower panels). The bands were quantified using the ImageQuant 5.1 software. These results are representative of three independent experiments with similar results.  +  c  LL  +  CC  o  c LL  00 + + CC c o -I  O  CQ + o  H  H  CQ C\J  Anti-IgM (5 min)  +  CM  + 00  c + LL c > +  *  o  £ 5.  LL +  +  £  GC _J  cn Q o oo +  o CD + JC O >  i— (0  t  o CVJ  o  < *  1  1  ii  LL - 1 + Q  o c* o CQ + o > ° t  CO Q  < *  +  65  -p-SHP-2  IB: anti-phospho-tyrosine 580 SHP-2  BCR-induced Akt phosphorylation, suggesting that Lyn plays a role in the inhibition of Akt pathway activation downstream of the BCR in the transfected AtT20 cells. Since the wild type form of Lyn could associate with SHP-2 in the AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8] cells, the ability of the kinase dead mutant to interact with SHP-2 was also examined. Since SHP-2 presumably interacts with putative tyrosine phosphorylated ITIM sequence within the kinase domain of Lyn that was not altered in the kinase dead mutant, it was expected that SHP-2 would be able to associate with this mutant form of Lyn. As expected, SHP-2 was present in the Lyn immunoprecipitates and a tyrosine phosphorylated protein in the approximate molecular weight range of Lyn was also present in the SHP-2 immunoprecipitates (Figure 5.3). Whether the interaction between Lyn and SHP-2 is direct or indirect could not be inferred from these immunoprecipitation experiments. However, experiments using the GST fusion protein of the tandem SH2 domains of SHP-2 indicated that Lyn and SHP-2 interaction may be mediated by the SH2 domain(s) of SHP-2 and two phosphotyrosine-containing sequences within the kinase domain of Lyn that resemble the ITIM consensus recognized by SHP phosphatases. Although SHP-2 can interact with the catalytically inactive mutant form of Lyn, only the SHP-2 molecules that associated with the wild type form of Lyn were tyrosine phosphorylated. In the AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8] cells, Lyn-associated SHP-2 was inducibly tyrosine phosphorylated following BCR cross-linking while it was constitutively phosphorylated at lower levels in the AtT20 BCR+ Fyn+ KD Syk+ Lyn+ [#10] cells (Figure 5.3). In the other cell lines containing the kinase dead mutant of Lyn, Lyn-associated SHP-2 was not tyrosine phosphorylated to any extent before and after BCR activation.  These results indicate that Lyn may be responsible for tyrosine  phosphorylation of the SHP-2 molecules with which it interacts.  This tyrosine  phosphorylation could result in the enhancement of phosphatase activity or enable it to interact with other SH2 domain-containing signaling components. Phosphorylation of SHP-2 on the tyrosine 580 residue was also examined in the cell lines following BCR cross-linking. Since SHP-2 phosphorylation on this specific residue increased in response to BCR cross-linking, it was concluded that Syk kinase activity  may also be important for the induction of SHP-2 phosphorylation (Figure 5.4). Although Syk appears to play a role in SHP-2 phosphorylation, an interaction between these two proteins could not by detected in immunoprecipitation experiments (data not shown). Perhaps Syk-SHP-2 complexes exists only transiently, may be of low abundance within the cells or dissociates the during cell lysis and immunoprecipitation procedure. Upon further examination of the data, there was a discrepancy observed with the SHP-2 tyrosine phosphorylation results from the different experiments performed. For example, in the Lyn immunoprecipitation experiment, the results indicated that the kinase activity of Lyn was required for SHP-2 phosphorylation and Syk had no effect on this phosphorylation.  However, in experiments that examined SHP-2 tyrosine  phosphorylation on one residue (tyrosine 580), the results suggested that Syk plays a more important role in the phosphorylation of this phosphatase. The reason for this discrepancy is that different cellular pools of SHP-2 molecules are being examined in the different experiments. In the phospho-tyrosine 580 SHP-2 Western immunoblot and SHP-2 immunoprecipitations, all of the SHP-2 within the cells is being examined. However, in the SHP-2 immunoprecipitation, the tyrosine phosphorylation state of the entire phosphatase was determined whereas in the phospho-tyrosine 580 SHP-2 Western immunoblot, tyrosine phosphorylation of only one residue was detected. There are other tyrosine residues within SHP-2 that can be phosphorylated.  Therefore, the  immunoprecipitated SHP-2 that is tyrosine phosphorylated at these other residues will show different patterns of phosphorylation when compared to results from phosphospecific antibodies. The use of phospho-specific antibodies provides more pertinent information about the regulation of various proteins. In any case, the results obtained with the phospho-specific SHP-2 antibodies indicated that Syk, like Lyn, plays a role in SHP-2 phosphorylation.  However, it should be noted that the SHP-2 molecules  phosphorylated by these two PTKs within the cells might be different. For instance, Syk, being a cytoplasmic kinase that is not constitutively localized at the plasma membrane, may phosphorylate the SHP-2 molecules within the cytoplasm. While Lyn, which is constitutively localized to the plasma membrane, can phosphorylate SHP-2 molecules with which it interacts and potentially those recruited to the plasma membrane by other  proteins. In the Lyn immunoprecipitation experiments, tyrosine phosphorylation of only Lyn-associated SHP-2 molecules was detected. These SHP-2 molecules are probably the most important of all the cellular SHP-2 molecules in terms of their ability to participate in BCR signaling. This is because this small group of SHP-2 molecules is associated with Lyn and therefore can be localized to the plasma membrane where a lot of signaling components are assembled into complexes and become activated. The Lyn-associated SHP-2 molecules may be able to participate in the negative regulation of the PI 3kinase/Akt pathway due to their plasma membrane localization since signaling of this pathway occurs predominantly at the plasma membrane. One way in which the Lynassociated pool of SHP-2 could participate in the negative regulation of BCR-induced PI 3-kinase/Akt activation is by de-phosphorylating the adapter protein that is responsible for the recruitment of PI 3-kinase to the plasma membrane. De-phosphorylation of tyrosine residues on this adapter protein would result in the removal of docking sites for PI 3-kinase leading to its subsequent release back into the cytoplasm. The levels of PI3,4,5-P3 would decrease thus reducing the number of docking sites for PH domaincontaining molecules such as PDK1, PDK2 and Akt. Alternatively, the Lyn-associated SHP-2 molecules could potentially de-phosphorylate sites on PI 3-kinase, resulting in its inactivation. In both scenarios, PI 3-kinase activation is reduced which subsequently leads to a decrease in Akt phosphorylation and activation.  CHAPTER 6 Altering the membrane localization of Lyn and Blk does not affect BCR-mediated inhibition of Akt phosphorylation and activation  6.1  Introduction  Lipid rafts are also referred to as membrane microdomains, detergent-resistant membranes (DRMs), glycolipid-enriched membranes (GEMs) or detergent-insoluble glycolipid-enriched membranes (DIGs) (reviewed by Harder and Simons, 1997; Brown and London, 2000; Simons and Toomre, 2000; Maxfield, 2002).  They are sub-  microscopic areas that are approximately 70-300 nm in diameter and are present within cellular membranes that are enriched in sphinoglipids including sphingomyelin, glycosphingolipids and cholesterol (Simons and Ikonen, 1997; Varma and Mayor, 1998; Kenworthy, 2002). Based on the percentage of sphinoglipid content within the plasma membrane, lipid rafts are thought to constitute 30% of the total plasma membrane. Lipid rafts are ubiquitously expressed in mammalian cells, but they have also been found in the yeast Saccharomyces cerevisiae and in lower organisms such as Caenorhabdtis elegans and Drosophila.  Due to their high lipid content, lipid rafts are insoluble at low  temperatures in non-ionic detergents such as Triton X-100. This property facilitates the isolation of lipid rafts from the bulk plasma membrane of cells, which can be solubilized by non-ionic detergents. The evidence accumulated indicates that lipid rafts play important roles in diverse cellular processes including signal transduction, cell migration (Manes et al., 1999; reviewed by Manes et al., 2001), cell adhesion (reviewed by Pande, 2000), cell polarity (reviewed by Gomez-Mouton et al., 2001), cell-mediated killing (Lou et al., 2000) and protein sorting and trafficking through the secretory and endocytic pathways (reviewed  by Brown and London, 1998; Gruenberg, 2001). Many microorganisms also make use of lipid rafts for pathogenesis (review in Rosenberg et al, 2000; van der Goot et al., 2001). Bacteria including Salmonella typhimurium (Waugh et al., 1999), E. coli (Gauthier et al., 2000) and Mycobacterium bovis (Gatfield and Pieters, 1999) all use lipid rafts to enter host cells. Viruses, like the SV40 virus, have also been shown to use lipid rafts as portals of entry into host cells (Parton and Lindsay, 1999). However, other viruses such as the human immunodeficiency virus (HIV) (Dykstra et al., 2000; Manes et al., 2000; Nguyen et al., 2000; Ono and Freed, 2001), Ebola (Bavari et al., 2002), Influenza (Scheiffele et al., 2000) and the measles virus (Manie et al., 2000) use the lipid rafts of host cells for assembling and budding of infectious virus particles. Subsets of lipid rafts containing caveolae are also thought to be involved in Plasmodium falciparum infection, the parasite which causes malaria (Olliaro and Castelli, 1997). Interestingly, lipid rafts also play a role in the pathogenesis of prions, the causative agents of mad cow disease and CrutzfledJacob disease (Kaneko et al., 1997). The role of lipid rafts in regulating the signaling by several receptors has been examined. These receptors include chemokines receptors (reviewed in Manes et al., 2001) and growth factor receptors including insulin and EGF receptors (reviewed in Simons and Toomre, 2000). The role of lipid rafts in signaling through immune receptors has also been well characterized (reviewed in Xavier and Seed, 1999; Lang et al., 2000; Cherukuri et al., 2001; Horejsi, 2002). Lipid rafts are important for T lymphocyte activation and development since signaling through the TCR is thought to occur within these rafts (Xavier et al, 1998; reviewed in Jones et al., 2000; Lieitenberg et al., 2001; Miceli et al., 2001). They also play an important role in high affinity receptor for IgE (FceRI) signaling (Field et al., 1997; reviewed in Field et al, 1997; Stauffer and Meyer; 1997; Sheets et al., 1999; Holowka and Baird, 2001) and IgA (Lang et al, 1999). Finally, BCR signaling occurs within lipid rafts (reviewed in Cheng et al., 2001b; Pierce, 2002; Dykstra et al., 2003). Lipid rafts in B lymphocytes are thought to be important for BCR internalization by receptor-mediated endocytosis and in antigen targeting and presentation, in addition to signal transduction (Stoddart et al, 2002). Activation of the calcium signaling pathway downstream of the BCR has been previously shown to require  intact lipid rafts (Aman and Ravichandran, 2000; Guo et al., 2000; Petrie et al., 2002). Indeed, a lot of evidence has accumulated indicating that the BCR can translocate into lipid rafts following receptor cross-linking (Cheng et al., 1999; Petrie et al., 2000; Weintraub et al., 2000; Cheng et al., 2001a; Gupta and DeFranco, 2003). Translocation of the BCR into lipid rafts following cross-linking suggests that it is an important step in signal transduction through this immune receptor. The ability to include or exclude proteins to varying degrees is a very important property of lipid rafts. A well-known example of a protein that is constitutively excluded from lipid rafts is the CD45 protein tyrosine phosphatase (Rodgers and Rose, 1996). Lyn is an example of a protein that is constitutively localized in lipid rafts (Field et al., 1995). Other molecules have been shown to translocate into lipid raft following cell stimulation, these include the signaling molecules such as the p21Ras monomeric G protein, the Vav guanine nucleotide exchange factor, PLC-y2 and PI 3-kinase (reviewed by Cherukuri et al., 2001). Another molecule that is constitutively localized in lipid rafts is Raftlin (raf tlinking protein) (Saeki et al., 2003). Raftlin, which is expressed only in B lymphocytes, is necessary for the formation or maintenance of lipid rafts in the cells. In the absence of Raftlin, BCR signaling is impaired. Lipid rafts are postulated to play a crucial role in the initiation of signaling cascades by providing sites within the plasma membrane where positive regulatory signaling complexes can pre-assemble in the same area as the receptors. Upon receptor engagement, this pre-assembled complex would allow for more efficient activation of the signaling cascades. In addition, concentrating the receptors and signaling components to lipid raft platforms at the plasma membrane could also protect the activated complexes from negative regulatory molecules including phosphatases, which could potentially down-regulate the signaling. Protection from negative regulators allows receptors to maintain high levels of signaling for longer periods of time. Finally, lipid rafts can provide signaling specificity by including or excluding certain molecules. Thus, receptors activated within lipid rafts are allowed access to specific sets of signaling molecules within the rafts.  In this way, lipid rafts may provide a mechanism for  regulating cross-talk between the many receptors within a cell.  The localization and function of Src kinase family members in lipid rafts has been welldocumented in many systems. Src kinases contain amino acids within the amino terminal unique region that can be modified by acylation (reviewed by Resh, 1994). These modifications are responsible for the plasma membrane localization of Src family kinases, as well as other proteins (McCabe and Berthiaume, 1999). All Src kinases have a glycine residue at position 2 to which the fatty acid myristate is added by the enzyme, N-myristol transferase (Raju et al., 1995). This permanent modification is sufficient to localize Src kinases to the plasma membrane (van't Hof and Resh, 1997; McCabe and Berthiaume, 2001). With the exception of two Src family kinases (Src and Blk), all family members also have an amino terminal cysteine residue at position 3 to which the fatty acid palmitate is added (reviewed by Resh, 1994). myristylation, is reversible.  Palmitylation, unlike  However, the enzyme involved in the removal of the  palmitate moiety is yet to be identified (Paige et al., 1993; Berthiaume and Resh, 1995; Varner et al., 2002; reviewed in Mumby, 1997). This palmitylation modification on most of the Src family kinases enables them to be localized to lipid rafts within the plasma membrane (Kosugi etal, 2001). Since Lyn and Lck have two acyl modifications (myristate and palmitate) while Blk only has the myristate modification, it was hypothesized that these Src kinases would localize to different membrane compartments in the AtT20 cells. The differential localization of various PTKs in the plasma membrane could account for the differences in BCR-induced Akt phosphorylation observed in the various AtT20 transfectants. First, I show that the Src kinases Lyn, Lck and Blk are differentially localized to the lipid raft and non-raft fractions of B lymphocyte cell extracts that were analyzed using sucrose gradients. Then using the same methodology, the location of Lyn, Lck and Blk in transfected AtT20 cells was determined. To determine if the localization of Lyn and Blk in lipid rafts played a role in the suppression of Akt phosphorylation following BCR cross-linking, mutants of these two Src kinases that contained different acyl modifications were created. These mutated Src kinases were expressed in AtT20 cells. Altering the membrane localization was predicted to result in different protein interactions between the Src kinases and other signaling components. These new interactions or lack thereof with other signaling  molecules could influence the activation of the PI 3-kinase/Akt pathway in response to BCR engagement. It was established that altering the plasma membrane localization of Lyn and Blk in lipid rafts did not relieve the suppression of BCR-induced Akt phosphorylation. In addition, the ability of the Lyn and Blk acyl mutants to interact with SHP-2 in vitro and the BCR-induced SHP-2 phosphorylation were unaffected. Interestingly, BCR-induced Akt phosphorylation was normal in AtT20 cells expressing a Blk mutant in which both acylation sites were removed to create a cytosolic protein, suggesting that plasma membrane localization is required for the suppression of Akt phosphorylation following BCR cross-linking.  6.2  Results  6.2.1  Chains of the BCR translocate into lipid rafts following receptor crosslinking  The four chains of the BCR have previously been shown to inducibly translocate into lipid rafts following receptor cross-linking in different B lymphocyte cell lines (Cheng et al., 1999; Petrie et al., 2000; Weintraub et al., 2000; Cheng et al., 2001a; Gupta and DeFranco, 2003). Therefore, the ability of the BCR to translocate into lipid rafts in BCRexpressing AtT20 cells was examined and compared with the results obtained from B lymphocyte cell lines. Lipid rafts were isolated by discontinuous sucrose gradient ultracentrifugation as described by Deans et al. (1998). Four different fractions were recovered following centrifugation: the lipid raft fraction; two soluble fractions, Fraction 12 and 14 which contain the soluble non-raft proteins; and the pellet which contains both Triton X-100 soluble and insoluble proteins (Figure 6.1). The location on the gradient of the different BCR chains before and after BCR cross-linking was examined first in three B lymphoma cell lines (WEHI 231, CH31 and A20). In the immature B lymphoma WEHI 231, translocation of the K light chain, Ig-a and Ig-P into lipid rafts was observed following BCR cross-linking (Figure 6.2 A). However, movement of the p heavy chain in this cell line is not as pronounced. When the CH31 immature B lymphoma cell line  Figure 6.1 Fractions collected from the sucrose gradient. Diagram of the various  lipid raft and soluble non-lipid rafts fractions collected from a sucrose gradient following ultracentrifugation.  Lipid rafts  Faction 12 Faction 14  Pellet  Soluble fractions  Figure 6.2 Inducible translocation of the BCR chains into lipid rafts following receptor cross-linking in B lymphoma cell lines. (A) WEHI 231, (B) CH31 and (C)  A20 B lymphoma cells were stimulated with anti-IgM antibodies or anti-IgG antibodies for 5 minutes (+) or left unstimulated (-) prior to cell lysis. The cell extracts were separated by discontinuous sucrose gradient ultracentrifugation, fractions collected and the protein concentrations of the samples were quantified by BCA assay. Five to ten pg of protein from each fraction was separated by SDS-PAGE and analyzed by Western blotting for p heavy chain, g heavy chain, K light chain, Ig-a and Ig-(3. The bands were visualized by enhanced chemiluminescence. Molecular weight standards (in kDa) are indicated on the left. These results are representative of two experiments with similar results.  (A) WEHI 231  (B) CH31 Soluble fractions  Soluble fractions Lipid raftS l F r a c t ' o n 1 2  Fraction  14  (C) A20  Lipid rafts l  Pellet Anti-IgM I (5 min) |  80-  80-  60 -  60-  "  +  Fraction 12  1 | -  +  Soluble fractions  Fraction 14  1 | -  +  1 | -  Pellet +  |  Lipid rafts  l  Fraction 12  Fraction 14  1 |  IB: anti-p heavy chain  IB: anti-p heavy chain  IB: anti-K light chain  IB: anti-K light chain  IB: anti-K light chain  IB: anti-lg-a  IB: anti-lg-a  IB: anti-lg-a  IB: anti-lg-p  IB: anti-tg-p  IB: anti-y heavy chain  Pellet  |  was examined, the amount of all the BCR chains in the lipid rafts also increased following receptor cross-linking (Figure 6.2 B). The mature A20 B lymphoma cell line was also examined to determine if the BCR in these cells inducibly translocates into lipid rafts as expected. The amounts of y heavy chain, K light chain and Ig-a in the lipid raft fraction increased following BCR cross-linking with anti-mouse IgG antibodies (Figure 6.2 C). Therefore, the chains of the BCR translocate into lipid rafts following receptor engagement in both immature and mature B lymphoma cell lines. In general, among all the transfected non-lymphoid AtT20 cell lines examined, there appeared to be more of the BCR constitutively present in the lipid raft fraction. Inducible translocation of the BCR chains into the lipid raft fraction following receptor cross-linking, however, was not as robust. There is some translocation of the p heavy and Ig-(3 accessory chain (Figure 6.3). However, translocation of the X light chain or Ig-a into lipid rafts was difficult to detect in any of the cell lines. The lack of abundant translocation of these chains into the lipid rafts may be due to inherent differences in the membrane composition, less efficient cross-linking of the BCR in AtT20 cells compared to the B lymphoma cell lines or just difficulty in showing increases in amount when there was so much BCR in the raft constitutively. Signaling through the BCR could occur both in and out of lipid rafts in AtT20 cells since all four chains of the receptor were present in these fractions.  Figure 6.3 Inducible translocation of the BCR chains into lipid rafts following receptor cross-linking in transfected AtT20 cells. AtT20 BCR+ Fyn+ Syk+ Blk+  [#17] and AtT20 BCR+ Fyn+ Syk+ Gabl-EGFP+ AtT20 cells were stimulated with anti-IgM antibodies for 10 minutes (+) or left unstimulated (-) prior to cell lysis. The cell extracts were separated by discontinuous sucrose gradient ultracentrifugation, fractions collected and the protein concentrations of the samples were quantified by BCA assay. Five to ten pg of protein from each fraction was separated by SDS-PAGE and analyzed by Western blotting for p heavy chain, X light chain, Ig-a and Ig-(3. The bands were visualized by enhanced chemiluminescence. Molecular weight standards (in kDa) are indicated on the left. These results are representative of two experiments with similar results.  AtT20 BCR+ Fyn+ Syk+ Blk+ [#17] Soluble fractions Lipid rafts  Fraction 12 Fraction 14  Pellet  80-  -/y heavy chain  6026-  • A, light chain  20-  36-  <—lg-cc  •  26-  -ig-P  AtT20 BCR+ Fyn+  Syk+ Gab1-EGFP  Soluble fractions Lipid rafts Anti-IgM (10 min)  +  Fraction 12 Fraction 14 -  +  +  Pellet  +  80-// heavy chain  60-  *  -A, light chain -Ig-a  3626-  m •  m  I'  ...  •ig-P  6.2.2  The Src kinases, Lyn, Lck and Blk, are differentially distributed in lipid rafts  The hypothesis that Src kinases are differentially distributed in the lipid raft and non-raft regions within the plasma membrane was examined in B lymphoma and transfected AtT20 cell lines. The specific localization of Src kinases within different regions of the plasma membrane may facilitate or interfere with their ability to interact with various signaling molecules. These signaling molecules could be components of positive or negative regulatory signaling pathways. Therefore the location of a protein within specific domains may determine whether it will play a positive or negative role in the regulation of different signal transduction pathways. The location of Lyn, Blk, Lck and Syk in the various sucrose gradient fractions in WEHI 231, CH31 and A20 B lymphoma cells was first examined. As expected and as observed by other investigators, Lyn was predominantly localized in the lipid raft fraction (Figure 6.4). Approximately 80-85% was present in the lipid raft fraction and this amount increased slightly following BCR cross-linking. Heavier forms of Lyn, presumably tyrosine phosphorylated forms, were present in the BCR-stimulated samples of all three B lymphoma cell lines. Interestingly, a small amount of Lyn was found in Fraction 12 from unstimulated WEHI 231 and A20 cells. The presence of Blk in the different fractions was next inspected in the WEHI 231 and A20 cell lines. Surprisingly, Blk was constitutively present in the lipid raft fractions of both cell lines (Figure 6.4 A and B). In the WEHI 231 cells, the majority of the Blk molecules were found in the soluble fractions (Fraction 12 and 14) and a smaller amount (approximately 15%) was present in the lipid raft fraction, as quantified using the ImageQuant 5.1 program. In A20 cells, however, a larger percentage (close to 50%) of the Blk molecules were constitutively present in the lipid raft fraction. The differences in the distribution of the Blk molecules in WEHI 231 and A20 cells may be due to the different developmental stages of the two B lymphoma cell lines. The function of Blk in these cell lines could be different and therefore may account for its differential localization in lipid rafts and non-raft compartments within the cells. The presence of Blk molecules in the lipid rafts was unexpected because Blk only has the myristate modification, which allows it to attach to the plasma membrane, but not  Figure 6.4 Distribution of Lyn, Lck, Blk and Syk in B lymphoma cell lines in the  sucrose gradient. (A) WEHI 231, (B) A20 and (C) CH31 B lymphoma cells were stimulated with anti-IgM antibodies or anti-IgG antibodies for 5 minutes (+) or left unstimulated (-) prior to cell lysis. The cell extracts were separated by discontinuous sucrose gradient ultracentrifugation, fractions collected and the protein concentrations of the samples were quantified by BCA assay. Thirty pg of protein from each fraction was separated by SDS-PAGE and analyzed by Western blotting for Lyn, Lck, Blk or Syk. The bands were visualized by enhanced chemiluminescence. Molecular weight standards (in kDa) are indicated on the left.  (A) WEHI 231 Soluble fractions  I Lipid rafts Fraction 12 Fraction 14 Anti-IgM (5 min) I  1 -  + | -  Pellet  1 + | -  60-  + 1 -  +  ^H-p53/p56 Lyn  50IB: anti-Lyn  p55 Blk IB: anti-Blk 60 •  -p56 Lck  50IB: anti-Lck  80•Syk  60-  IB: anti-Syk  (B) A20 Soluble fractions  I  1  Lipid rafts Fraction 12 Fraction 14 Anti-IgG , (5 min) I  -  +  I -  1 I -  +  +  Pellet  1 I -  , + |  60^~"|-p53/p56 Lyn 50IB: anti-Lyn  p55 Blk IB: anti-Blk  (C) CH31 Soluble fractions  I  Anti-IgM j (5 min)  Lipid rafts Fraction 12 Fraction 14 +  | -  +  1 | -  +  1 I -  Pellet  1 .  +  60p53/p56 Lyn  IB: anti-Lyn  localize to lipid rafts. Theoretically, Blk should not be present in the lipid rafts because it lacks a palmitate modification.  However, protein-protein interactions in addition to  myristylation have been shown to also be important in targeting proteins to lipid rafts (Mcabe and Berthiaume, 2001). So Blk may be interacting with different proteins which allows it to be localized in lipid rafts. However, no published data is available for comparison of these results. Finally, the location of Lck was determined in the WEHI 231 cells. Lck, like Lyn, contains two acyl modifications within its amino terminal end and should therefore be localized in lipid rafts. A large portion of Lck (40%) was present in the lipid raft fraction (Figure 6.4 A). However, unlike Lyn, some Lck molecules were also found in the non-raft fractions (Fraction 12, 14 and the pellet). In WEHI 231 cells, the location of Syk in the sucrose gradient was also determined. Syk was predominantly found in the non-raft fractions and only a small amount was present in the lipid rafts (approximately 7%). The presence of Syk in the lipid rafts may be due to its ability to interact with other proteins including the BCR. The levels of Lyn, Blk, Lck and Syk in the lipid raft fraction were not affected by BCR cross-linking. Since the three Src kinases were differentially distributed in lipid raft and non-raft domains of the plasma membrane in the B lymphoma cell lines, it was hypothesized that they also have different patterns of distribution within the transfected non-lymphoid AtT20 cells. The transfected wild type Lyn in the AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8] cells was predominantly found in the lipid rafts (75%) (Figure 6.5 A), as observed in the three B lymphoma cell lines (Figure 6.4). Unlike Lyn, more of the transfected Blk in the AtT20 BCR+ Fyn+ Syk+ Blk+ [#17] cells was located outside of the lipid rafts and pellet fraction (Figure 6.5 B). Approximately 35-40%) of the Blk molecules were found in the lipid rafts and in the pellet fraction, with 60-65%) in the soluble fractions. Again these results were somewhat surprising since Blk only has one myristate modification and should therefore be localized in the non-rafts areas of the plasma membrane. The Lck molecules in the AtT20 BCR+ Fyn+ Syk+ Lck+ [#17] were distributed almost evenly throughout the lipid raft (25-28%) and non-raft fractions (Figure 6.5 C). The amount of all three of these Src kinases in the lipid rafts and other fractions was unaltered by BCR  cross-linking. The localization of Lyn, Lck and Blk was also unaffected by the presence of Syk since these Src kinases showed a similar pattern of distribution in the Syknegative transfectants, AtT20 BCR+ Fyn+ Lyn+ [#33], AtT20 BCR+ Fyn+ Lck+ [#10.15] and AtT20 BCR+ Fyn+ Blk+ [#18] clones (data not shown). The location of Syk, SHP-2 and Akt was also examined in the different cell lines. Unlike the Syk in WEHI 231 cells, the Syk molecules were evenly distributed throughout the various lipid raft and non-raft fractions in the transfected AtT20 cells (Figure 6.5). The amount of Syk in the various fractions was unaffected by BCR cross-linking.  To  determine if co-expression with Lyn and Blk would affect its localization at the plasma membrane, the, location of SHP-2 in the transfected AtT20 cells was examined. The majority of the SHP-2 molecules within the cells (60-70%) was localized in the soluble fractions (Fractions 12 and 14) while 20-25% was located in the lipid raft and pellet fractions in all the AtT20 cell lines. Since SHP-2 does not have any acyl modifications, its presence in the lipid rafts is probably due to its ability to interact with lipid raftassociated proteins such as Src family kinases. The location of Akt in lipid rafts in the AtT20 cells was also determined. Like SHP-2, Akt is a cytosolic protein that lacks acyl modifications. As predicted, most of the cellular Akt was found in the soluble fractions of the sucrose gradient (75-85%), while a small amount was present in the lipid rafts (515%). The amount of Akt in the lipid rafts was not altered in response to BCR engagement, indicating that Akt does not preferentially associate with phospholipids contained within lipid rafts. The location of tubulin was used as a control to provide evidence that the lipid rafts were not contaminated with soluble proteins. Tubulin was found only in the soluble fractions (Fractions 12 and 14) and sometimes in the pellet, as expected. It is absent from the lipid raft fractions indicating that they were not grossly contaminated with non-raft proteins.  Figure 6.5 Src kinases are differentially localized in the different sucrose gradient fractions from transfected AtT20 cells. (A) AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8], (B)  AtT20 BCR+ Fyn+ Syk+ Blk+ [#17] cells and (C) AtT20 BCR+ Fyn+ Syk+ Lck+ [#17] were stimulated with anti-IgM antibodies for 10 minutes (+) or left unstimulated (-) prior to cell lysis. The cell extracts were separated by discontinuous sucrose gradient ultracentrifugation, fractions collected and the protein concentrations of the samples were quantified by BCA assay. Twenty pg of protein from each fraction was separated by SDS-PAGE and analyzed by Western blotting for Lyn, Lck, Blk, Syk, SHP-2, Akt and/or tubulin. The bands were visualized by enhanced chemiluminescence. Molecular weight standards (in kDa) are indicated on the left. These results are representative of two independent experiments.  Soluble fractions  I Anti-IgM (10 min)  Lipid rafts Fraction 12 Fraction 14  r|  + I -  + I•  Pellet  + I"  + I  61 -  IB: anti-Lyn  IB: anti-Syk  50-  <  Syk  <  SHP-2  <  Akt  8061 -  IB: anti-Akt 61 -  IB: anti-tubulin  Lyn  8061 -  IB: anti-SHP-2  <  50-  <  tubulin  Soluble fractions  I Anti-IgM (10 min)  IB: anti-Blk  61 -  Lipid rafts Fraction 12 Fraction 14  Pellet  <  Blk  <  Syk  50-  80 IB: anti-Syk  IB: anti-SHP-2  61  8061 -  IB: anti-Akt  IB: anti-tubulin  61 -  50  <  SHP-2  <  Akt  <  tubulin  Soluble fractions  I Lipid rafts Fraction 12 Fraction 14  IB: anti-Lck  Pellet  <  Lck  <  Syk  IB: anti-SHP-2  <  SHP-2  IB: anti-Akt  <  Akt  IB: anti-tubulin  <  tubulin  IB: anti-Syk  6.2.3  Altering the membrane localization of Lyn and Blk does not alleviate the inhibition of Akt phosphorylation following BCR cross-linking  To determine if the distribution of Lyn and Blk at the plasma membrane influences the inhibition of Akt activation following BCR cross-linking, mutants of these Src family kinases with alterations in their acyl modifications were created.  Site-directed  mutagenesis was used to create an expression vector containing a gene in which the third amino acid of Lyn was converted from cysteine to leucine (Figure 6.6 A). The cysteine residue is the amino acid that becomes post-translationally modified by palmitylation. Therefore, mutating this site in Lyn would hypothetically enable this non-palmitylated form of the Src kinase ("non-palm Lyn") to become preferentially localized to the nonlipid raft fractions. In addition, an expression vector containing a gene in which the third amino acid of Blk was converted from leucine to cysteine was also created by site directed mutagenesis (Figure 6.6 B). This cysteine residue in the third position would theoretically enable Blk to become palmitylated and enable this mutant form ("palm Blk") to become more effectively localized to lipid rafts. These two expression vectors were transfected into AtT20 cells expressing the Syk PTK to produce the two cell lines co-expressing the non-palmitylated form of Lyn and the palmitylated form of Blk, AtT20 BCR+ Fyn+ Syk+ non-palm Lyn+ and AtT20 BCR+ Fyn+ Syk+ palm Blk+ cell lines, respectively. AtT20 BCR+ Fyn+ Syk+ non-palm Lyn+ [#35] and AtT20 BCR+ Fyn+ Syk+ palm Blk+ [#10] cell lines were chosen for further experimentation because these two clones expressed similar amounts of non-palmitylated Lyn and palmitylated Blk, respectively, as the wild type forms in the AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8] and AtT20 BCR+ Fyn+ Syk+ Blk+ [#17] cell lines (Figure 6.7 A). The lipid raft and non-raft location of the non-palmitylated mutant form of Lyn in the AtT20 BCR+ Fyn+ Syk+ non-palm Lyn+ [#35] cells was determined.  The non-  palmitylated form of Lyn, unlike the wild type form, was clearly more abundant in the soluble fractions of the sucrose gradient (Figure 6.7 B and 6.8 A). Only,40-45% of the total non-palmitylated Lyn was found in the lipid rafts (compared to 75% of the wild type form in AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8] cells). There was approximately the same  Figure 6.6 Generation of mutants of Lyn and Blk with alterations in their acylation  sites. (A) The palmitylation site (palm) of Lyn was removed by altering the third amino acid from cysteine to leucine by site-directed mutagenesis. This Lyn mutant is referred to as "non-palmitylated Lyn." (B) A palmitylation site was added to Blk by converting the third amino acid from leucine to cysteine by site-directed mutagenesis to create "palmitylated Blk." These mutants of Lyn and Blk were transfected into AtT20 BCR+ Fyn+ Syk+ cells to create the AtT20 BCR+ Fyn+ Syk+ non-palm Lyn+ and AtT20 BCR+Fyn+ Syk+ palm Blk+ cell lines, respectively. Myr refers to the myristylation site; palm refers to the palmitylation site.  (A)  myrpalm  fG  CIK  a Y  SH2  Unique  Kinase  Wild type Lyn  V  myr r r i K  Unique  a Y  K  SH3  Y '  Non-palmitylated Lyn  Kinase domain  to to  (B)  myr  r  r  r  r  Uniquei  SH3  Unique i  SH3  —  lei  I T  Y  Kinase domain  Wild type Blk  myrpalm  (&  r s11"  SH2  Y Kinase domain  Palmitylated Blk  amount of non-palmitylated Lyn present in the soluble fractions (Fractions 12 and 14) as there is in the lipid rafts. Therefore, mutating the palmitylation site of Lyn was sufficient to change the distribution of Lyn in the plasma membrane and decrease the amount of Lyn (by 30-35%) that was localized to the lipid rafts. However, complete exclusion of Lyn from lipid rafts was not achieved by this single point mutation. The majority of the palmitylated Blk molecules within the cell (60-70%) were present in the lipid raft fraction while a smaller amount was found in the soluble fractions (10-20% compared to 30% in AtT20 BCR+ Fyn+ Syk+ Blk+ [#17] cells) (Figure 6.7 B and 6.8 B). The percentage of palmitylated Blk in the lipid rafts was 20-30% higher than the amount of wild type Blk in the AtT20 BCR+ Fyn+ Syk+ Blk+ [#17] cells. Although addition of the palmitate modification was not sufficient to completely localize all of the Blk molecules into lipid rafts, it allowed more Blk molecules to localize in the rafts (compare Figure 6.8B with 6.5 B). Other interactions with proteins or lipids may be necessary for complete localization in lipid rafts. In summary, mutating the amino acid in the third position of Lyn and Blk was not sufficient to completely exclude Lyn from or completely localize Blk to lipid rafts. This mutation was, however, sufficient to change the distribution of approximately one-third of the total Lyn and Blk molecules within the cell. In the all of the Western blots of the AtT20 BCR+ Fyn+ Syk+ palm Blk+ [#17] cells, a lower molecular weight form of palmitylated Blk was consistently detected. This lower band was later shown to be an ubiquitinated form of palmitylated Blk (Figure 6.7 B), consistent with previous reports found in the literature (Oda et al, 1999). The change in distribution of the mutant forms of Lyn and Blk in the lipid rafts and nonraft fractions was unaffected by BCR cross-linking (Figure 6.8). Furthermore, the amounts of SHP-2, Akt and Syk in the various fractions were also unaltered by BCR engagement or by the presence of the mutant forms of Lyn and Blk. The amount of these signaling molecules in the different fractions was similar to the other cell lines examined (Figure 6.5).  Figure 6.7  Expression levels and distribution of non-palmitylated Lyn and  palmitylated Blk in the sucrose gradient. (A) Expression levels of non-palmitylated Lyn and palmitylated Blk in the AtT20 transfected cells. Forty pg of whole cell extracts from the different cell lines were separated by SDS-PAGE and analyzed by Western blotting with anti-Lyn, anti-Blk or anti-Syk antibodies.  (B) Distribution of non-  palmitylated Lyn and palmitylated Blk in the various sucrose gradient fractions. The cell extracts were separated by discontinuous sucrose gradient ultracentrifugation, fractions collected and the protein concentrations of the samples were quantified by BCA assay. Twenty pg of protein from each fraction was separated by SDS-PAGE and analyzed by Western blotting for Lyn, Blk, Syk or tubulin. One of the filters was stripped of antibodies and re-probed with antibodies that recognize ubiquitinated proteins (lower panel). The bands were visualized by enhanced chemiluminescence. Molecular weight standards (in kDa) are indicated on the left. These results are representative of two experiments. The immunoreactive bands were quantified using the ImageQuant 5.1 software.  SIZ  >c</>  §^ is 3 3J + r >< n  t On  DD  I* I  a> AtT20 BCR+ Fyn+ Syk+  c o cn  AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8]  cn oo o  Ul 0> o -» ro w  I  AtT20 BCR+ Fyn+ Syk+ non-palm Lyn [#35]  Lipid rafts Fraction 12  •D  Fraction 14  3  Pellet  t  W  I  o oo  o  t  |3 0) §  pa 0 ) 3 cf. "0<5  I  AtT20 BCR+ Fyn+ Syk+ non-palm Lyn [#35]  I  AtT20 BCR+ Fyn+ Syk+ palm Blk [#10]  t  CO  7T  AtT20 BCR+ Fyn+ Syk+ AtT20 BCR+ Fyn+ Syk+ Blk+ [#17] AtT20 BCR+ Fyn+ Syk+ palm Blk [#10]  Figure 6.8 Distribution of mutant Lyn and Blk proteins in the different sucrose gradient fractions is unaffected by BCR cross-linking. AtT20 BCR+ Fyn+ Syk+ non-  palm Lyn+ and AtT20 BCR+ Fyn+ Syk+ palm Blk+ cells were stimulated with anti-IgM antibodies for 10 minutes (+) or left unstimulated (-) prior to cell lysis. The cell extracts were separated by discontinuous sucrose gradient ultracentrifugation, fractions collected and the protein concentrations of the samples were quantified by BCA assay. The presence of non-palmitylated Lyn, palmitylated Blk and Syk in the various fractions was examined by SDS-PAGE and analyzed by Western blotting. The bands were visualized by enhanced chemiluminescence. Molecular weight standards (in kDa) are indicated on the left. These results are representative of two experiments. The immunoreactive bands were quantified using the ImageQuant 5.1 software.  (A) AtT20 BCR+ Fyn-i- Syk+ non-palm Lyn+ [#35] Soluble fractions Lipid rafts Fraction 12 Fraction 14 Anti-IgM (10 min)  *  +  I-  +  I-  Pellet  + I•  61 -  +  I non-palm Lyn  50-  IB: anti-Lyn 80 -  •Syk  61  IB: anti-Syk  80  SHP-2 61  IB: anti-SHP-2  tubulin  50 -  IB: anti-tubulin  (B) AtT20 BCR+ Fyn+ Syk+ palm Blk+ [#10] Soluble fractions Lipid rafts Fraction 12 Fraction 14 Anti-IgM | (10 min)  I  "  1  +I-  1  + I-  +I  1  Pellet  -  1  + l  61 -  Ih  50-  palm Blk  IB: anti-Blk 80Syk 61 -  IB: anti-Syk 80SHP-2 61 -  IB: anti-SHP-2  tubulin  50IB: anti-tubulin  The activation of Akt in response to BCR cross-linking was next examined in the cells expressing the acyl mutants of Lyn and Blk to determine if changing the distribution of these Src kinases at the plasma membrane would result in altered patterns of Akt phosphorylation compared to cells expressing the wild type forms. BCR-induced Akt phosphorylation in the AtT20 BCR+ Fyn+ Syk+ non-palm Lyn+ [#35] and AtT20 BCR+ Fyn+ Syk-i- palm Blk+ [#10] cell lines was still lower at all time points of receptor stimulation compared to cells that do not contain any form of Lyn or Blk (AtT20 BCR+ Fyn+ Syk+) (Figure 6.9A and B). Akt phosphorylation in response to BCR cross-linking in the AtT20 cells expressing the acyl mutants of Lyn and Blk is no different compared to the cells expressing the wild type forms of these kinases (Figure 6.9 C).  Figure 6.9 Altering the distribution of Lyn and Blk in lipid rafts has no effect on  Akt phosphorylation following BCR cross-linking. The various cell lines were serumstarved overnight in media containing 0.2% FCS before being stimulated with 20 pg/ml of anti-mouse IgM antibodies for the indicated time periods. Following lysis, 40 pg of total cell extract was separated by SDS-PAGE and analyzed by Western blotting with anti-phospho-serine 473 Akt antibodies. The filters were subsequently stripped of antibodies and re-probed with anti-Akt antibodies (lower panels). The bands were visualized by enhanced chemiluminescence. Molecular weight standards (in kDa) are indicated on the left. The bands were quantified using the ImageQuant 5.1 software. These results are representative of three independent experiments.  (A)  AtT20 BCR+ Fyn+ Syk+  Anti-IgM (minutes)  I 0  AtT20 BCR+ Fyn+ Syk+ non-palm Lyn+ [#35]  5 15 30 601 0  5 15 30 60 I  64-  p-Akt  IB: anti-phospho-serine 473 Akt 64-  Akt Re-probe: anti-Akt  to to  o  (C)  AtT20 BCR+Fyn+ Syk+ Blk+ [#17]  Anti-IgM I (minutes) I  0  AtT20 BCR+ Fyn+ Syk+ palm Blk+ [#10]  5 15 30 60l 0  15 30 60  64-  ' p-Akt  IB: anti-phospho-serine 473 Akt  Re-probe: anti-Akt  (B) Anti-IgM  AtT20 BCR+ Fyn+ Syk+ I 0  64-  5  •  AtT20 BCR+ Fyn+ Syk+ palm Blk+ [#10]  15 30 601 0  5 15 30 60 I _  —  p-Akt  IB: anti-phospho-serine 473 Akt 64-  Akt Re-probe: anti-Akt  (D)  AtT20 BCR+ Fyn+ AtT20 BCR+ Fyn+ Syk+ Syk+ Lyn+ [#8] non-palm Lyn+ [#35]  (minutes)>lo  64  5 15 30 60 lo  5 15 30 60  —  ,  p-Akt  IB: anti-phospho-serine 473 Akt 64-  Akt Re-probe: anti-Akt  6.2.4  Altering the membrane localization of Lyn and Blk does not affect BCRmediated SHP-2 phosphorylation  Since some differences are observed when SHP-2 phosphorylation was examined in cells expressing the wild type forms of Lyn and Blk, the ability of the acyl mutants to associate with and regulate SHP-2 phosphorylation was determined.  As expected, the non-  palmitylated Lyn and palmitylated Blk molecules were still able to associate with SHP-2 in vitro. The acyl mutants were able to interact with the SHP-2 tandem SH2 domain GST fusion protein (GST-SHP-2 (SH2-SH2)) (Figure 6.10). This was expected since the putative SHP-2 binding sites in non-palmitylated Lyn and palmitylated Blk were not altered in these proteins. Therefore, SHP-2 interaction with these mutant forms of Lyn and Blk can be mediated by the SH2 domain(s) of SHP-2 as observed with the wild type forms of these Src kinases. The phosphorylation of SHP-2 was next examined by using the phospho-specific antibodies described earlier. The pattern of SHP-2 phosphorylation in the AtT20 BCR+ Fyn+ Syk+ non-palm Lyn+ [#35] and AtT20 BCR+ Fyn+ Syk+ palm Blk+ [#10] clones was the same as that observed in the cells containing the wild type counterparts of the Src kinases (Figure 6.11). Therefore, the mutant forms of Lyn and Blk have no effect on SHP-2 phosphorylation on the tyrosine 580 residue following BCR cross-linking in the transfected AtT20 cells.  6.2.5  Targeting of Blk to the cytoplasm alleviates the inhibition of Akt phosphorylation following BCR cross-linking  To examine if the membrane localization of Blk affects  BCR-induced Akt  phosphorylation, a mutant Blk construct was created in which the myristylation site and  Figure 6.10 SHP-2 interacts with the acyl mutants of Lyn and Blk in vitro. One mg  of total cell lysate from the transfected AtT20 cells was pre-cleared with GST bound to glutathione-Sepharose 4B beads for 1 hour at 4°C. Glutathione-Sepharose 4B beads were incubated with 50 pi of bacterial cell lysate containing GST-SHP-2 (SH2-SH2) fusion protein for 1 hour at 4°C. After washing twice with lysis buffer, the immobilized GSTSHP-2 (SH2-SH2) fusion protein was incubated with the pre-cleared lysates and incubated at 4°C for 2 hours. The complexes were collected by centrifugation and washed twice with lysis buffer.  The bound proteins were eluted with IX SDS-PAGE  reducing sample buffer, separated by SDS-PAGE and analyzed for the presence of nonpalmitylated Lyn or palmitylated Blk by Western blotting with anti-Lyn and anti-Blk antibodies, respectively. Molecular weight standards (in kDa) are indicated on the left. These results are representative of two experiments with similar results.  LO CO  + c LL 00  + 2L DC + CQ  ^  O , t i < CO  c  + c  >>, u- i + : cc ; o CQ C +Q o  CM  +  c >  LL  +  DC O DQ O CM  <0  < *c c o c+ (0  wild type/non-palmitylated Lyn  PPT: GST-SHP-2(SH2-SH2) IB: anti-Lyn  m co 2* c>  + c + c > >> I sf-T" LL LL + C+C * DC I O co o CD CQ C +Q O c+ o CM CM > (0 H C>O  > LL +  CQ CC o <0 Q. CQ + o >. CM C O  wild type/palmitylated Blk ubiquitinated palmitylated Blk  PPT: GST-SHP-2(SH2-SH2) IB: anti-Blk  Figure 6.11 Altering the distribution of Lyn and Blk in lipid rafts has no effect on  SHP-2 phosphorylation following BCR cross-linking. The various cell lines were serum-starved overnight in media containing 0.2% FCS before being stimulated with 20 pg/ml of anti-mouse IgM antibodies for 5 minutes (+) or left unstimulated (-). Following lysis, 40 pg of total cell extract was separated by SDS-PAGE and analyzed by Western blotting with anti-phospho-tyrosine 580 SHP-2 antibodies. The filters were subsequently stripped of antibodies and re-probed with anti-SHP-2 antibodies (lower panels). The bands were visualized by enhanced chemiluminescence. Molecular weight standards (in kDa) are indicated on the left. These results are representative of three experiments.  in  co  c L>. L +  CC o CD o + <N I - >. <<0 Anti-IgM (min)  0  5 15  +  c l^oo *  +  cc i o co m c+ o Cvi  + —  ^  £  o = CD o C\J .J fc > < (0 0 5  t >. < </)  151 0  80-  IB: anti-phosphotyrosine 580 SHP-2  5 15 I — mm  -p- SHP-2  61 -  Re-probe: anti- SHP-2  80-  —  61  c C+C o CQ O CM +*>. I<co Anti-IgM (min)  IB: anti-phosphotyrosine 580 SHP-2  Re-probe: anti- SHP-2  + c M 00 01  O m =  o . < <0  + 2* U- ==  +  c „ +* cc + o 2 co m o . t i < CO  t  I 0 5 15 I 0 5 15  SHP-2  0  5 15!  +  ?  DC fe Q o . 64.  < CO  0  5 15 I  80-p- SHP-2  61 -  8061 -  —  —  SHP-2  the third amino acid residue (leucine) were mutated. This Blk should have no acyl modifications and thus be the least effective at associating with the plasma membrane. AtT20 cell lines co-expressing this mutant Blk, hereafter referred to as "soluble Blk," and Syk PTK were generated by calcium phosphate transfection (by May Dang-Lawson, a research technician in the lab). All the clones obtained from this transfection expressed high levels of soluble Blk. There was 3 and 9 times more soluble Blk in the AtT20 BCR+ Fyn+ Syk+ sol Blk+ cells compared to the wild type form in the AtT20 BCR+ Fyn+ Syk+ Blk+ [#17] and [#64] clones, respectively (Figure 6.12 A). A membrane fractionation experiment revealed that the majority of the soluble Blk molecules were present in the cytosolic (C) fraction whereas the majority of the wild type and palmitylated Blk are present in the membrane (M) fraction (Figure 6.12 B).  A  considerable amount of the soluble Blk was also present in the membrane fraction perhaps because the protein was highly expressed in the transfected cells. Alternatively, the soluble Blk found at the plasma membrane may be interacting with membraneassociated proteins. Next, the location of the soluble Blk in lipid raft and non-raft fractions was examined. A large percentage of the soluble Blk was present in the soluble fractions compared to the wild type Blk in the AtT20 BCR+ Fyn+ Syk+ Blk+ [#17] cells (Figure 6.12 C compared to Figure 6.5 B).  Akt phosphorylation in the AtT20 BCR+ Fyn+ Syk+ sol Blk+ cell line was not inhibited as was observed in the cell expressing wild type or palmitylated Blk (AtT20 BCR+ Fyn+ Syk+ Blk+ [#17] and AtT20 BCR+ Fyn+ Syk+ palm Blk+ [#10] cells, respectively) (Figure 6.13). Thus, altering Blk localization such that it is predominantly localized in the cytoplasm alleviates the inhibition of BCR-induced Akt phosphorylation.  Figure 6.12  Expression levels and distribution of cytosolic Blk mutant in the  sucrose gradient. (A) Expression level of the cytosolic Blk mutant in the AtT20 transfected cells. Whole cell extract from the indicated cell lines was separated by SDSPAGE and analyzed by Western blotting with anti-Blk antibodies. (B) The cytosolic Blk mutant is present in both the membrane and cytosolic fractions. The cell membranes and cytosolic fractions of the various transfected AtT20 cells were separated using a membrane enrichment procedure, as described in the Methods and Materials section. Thirty micrograms of each fraction was separated by SDS-PAGE and analyzed by Western blotting with anti-Blk antibodies. C = cytosolic fraction M = membrane fraction. (C) The cytosolic Blk mutant is localized mostly in the soluble fractions of the sucrose gradient. The cell extracts from AtT20 BCR+ Fyn+ Syk+ so/-Blk+ cells were separated by discontinuous sucrose gradient ultracentrifugation, fractions collected and the protein concentrations of the samples were quantified by BCA assay. Twenty micrograms of protein from each fraction was separated by SDS-PAGE and analyzed by Western blotting with anti-Blk antibodies. The bands were visualized by enhanced chemiluminescence. Molecular weight standards (in kDa) are indicated on the left. These results are representative of two experiments.  (A)  (B)  *  GO O V) +  +  -X  .x OT + c>  OT + c>. LL +  LL  +  CC  DC  ti ti o CM 1-  o  +  C+Q •K >  OT +  c + OC  ti  O  CM  CM  C+Q  O + £ CQ  r+ C+Q  E «o Q. +  OT c+>.  OT + c>.  £  >.  >.  OT +  >.  LL  DC  DC  O CC CM  +  c + DC  a o eg  O  C M C M  CM  o  a +  OT  +  a  LL  >.  LL  +  c  £+ CQ O in +  ti  CM  CM  <  61 -  wild type/ cytoplasmic Blk  50  61 -  wild type/ mutant Blk  50IB: anti-Blk  IB: anti-Blk  IB: anti-Syk tubulin 50IB: anti-tubulin  (C)  AtT20 BCR+ Fyn+ Syk+ sol Blk+ Soluble fractions w &  as  CM y— C c o o o o CO tC-O LL LL  61  4-» "q3 CL  cytosolic Blk  IB: anti-Blk  Figure 6.13 BCR-induced Akt phosphorylation was not inhibited in AtT20 cells expressing the soluble form of Blk.  The various cell lines were serum-starved  overnight in media containing 0.2% FCS before being stimulated with 20 pg/ml of antimouse IgM antibodies for the indicated time periods. Following lysis, 40 pg of total cell extract was separated by SDS-PAGE and analyzed by Western blotting with antiphospho-serine 473 Akt antibodies. The filters were subsequently stripped of antibodies and re-probed with anti-Akt antibodies (lower panels). The bands were visualized by enhanced chemiluminescence. Molecular weight standards (in kDa) are indicated on the left. The bands were quantified using the ImageQuant 5.1 software. These results are representative of three experiments.  Anti-IgM (minutes)  AtT20 BCR+ Fyn+  AtT20 BCR+ Fyn+  Syk+  Syk+ sol Blk+  ~0  5 15 30 60 I 0  5 15 30 60  64-  < — p-Akt IB: anti-phospho-serine 473 Akt  64-  < — Akt Re-probe: anti-Akt  6.3  Discussion  The localization of proteins within lipid rafts at the plasma membrane has been previously shown to play a role in signal transduction through the BCR. Here, it was established that the BCR in a mature B lymphoma cell line, A20, inducibly translocated into lipid rafts following receptor cross-linking (Figure 6.2 C). This phenomenon was also observed in two immature B lymphoma cell lines, WEHI 231 and CH31 (Figure 6.2 A and B), contrary to previous findings from another group (Sproul et al., 2000). The discrepancy between the results shown here and those of the other group's could be attributed to the differences in the subclones of the cell lines that were studied, differences in the way the cells were stimulated with anti-IgM antibodies or differences in the techniques used to isolate lipid rafts from the cells. Since it is thought that many different types of lipid rafts exist within cells, the different techniques used may have resulted in the isolation of different types of lipid rafts from the immature B lymphoma cells. Although less efficient compared to B lymphoma cells, some translocation of the p heavy chain and Ig-P accessory chain into lipid rafts was observed in the transfected AtT20 cells following BCR cross-linking (Figure 6.3). However, in these cells, there was more of the BCR constitutively located in the lipid rafts. Interestingly, this translocation only occured in cells expressing Syk, as cells lacking this PTK failed to show any translocation of the BCR chains into lipid rafts following receptor cross-linking. It is possible, that specific BCR signaling events are required for translocation of the BCR complex into lipid rafts.  Perhaps the lack of observable translocation of the BCR  proteins in AtT20 cells is due to less efficient or less effective aggregation and signaling, which could influence the translocation of the receptor complex.  The location of the Src kinases in the lipid rafts were also compared in the transfected AtT20 and B lymphocyte cell lines. The location of the transfected Src kinase Blk and Syk PTK in the transfected AtT20 cells was different compared to that observed in the WEHI 231 and A20 B lymphoma cell lines (Figures 6.4 and 6.5). The distribution of Lyn in the lipid raft and non-raft fractions in the transfected AtT20 cells was very similar to that observed in the three B lymphoma cell lines. The distribution of Lck in the transfected non-lymphoid cells and B lymphoma cells differ only slightly.  The  differences in BCR-mediated Akt phosphorylation in the transfected AtT20 cells may be due to these variations in the distribution of the Src kinases at the plasma membrane.  To determine if the distribution of Lyn and Blk within different plasma membrane domains affects PI 3-kinase/Akt pathway activation, Akt phosphorylation was examined in AtT20 cells expressing the acyl mutant of Lyn and Blk. The rationale was that redistribution of these Src kinases in the lipid raft and non-raft domains of the plasma membrane would alter the sets of proteins with which they could interact. Alternatively, changing the location of the kinases could interfere with their ability to associate with certain proteins. For example, if Lyn or Blk interact with a negative regulatory protein in a specific region of the plasma membrane such as the lipid rafts, then altering their location to non-raft areas lacking this negative regulatory molecule may prevent the interaction between the molecules. If Lyn and Blk cannot interact with the negative regulatory molecule, then these Src kinases cannot negatively regulate signaling pathways. On the other hand, if Lyn and Blk are altered such that they are localized to an area where a negative regulatory molecule is abundant, an interaction may occur between the Src kinases and the negative regulator. In this case, the Src kinases can participate in negative signaling due to their ability to associate with the negative regulator. Therefore, the distribution of Lyn and Blk within different domains at the plasma membrane was altered by creation of acyl mutants of these two Src kinases, which were introduced into AtT20 cells. By removing the palmitylation site of Lyn, its localization in lipid rafts was significantly reduced by approximately 30-35% (from 75% in AtT20 BCR+ Fyn+ Syk+ Lyn+ [#8] to 40-45% in AtT20 BCR+ Fyn+ Syk+ non-palm Lyn+ [#35] cells; compare Figures 6.7 B and 6.8A with 6.5A). Addition of a palmitate modification to Blk resulted in a 25-35% increase in the abundance of this Src family kinase in the lipid rafts (Figure 6.7 B and 6.8B with 6.5B). The altered distribution of Lyn and Blk within lipid rafts at the plasma membrane did not alter BCR-mediated SHP-2 phosphorylation or the association of SHP-2 with these Src kinases (Figures 6.10 and 6.11). Furthermore, Akt phosphorylation in response to BCR cross-linking was not affected in cells expressing the acyl mutants of Lyn and Blk compared to cells expressing the wild type forms of these Src kinases (Figure 6.9).  Although the lipid raft and non-raft locations of Lyn and Blk do no affect BCR-induced Akt phosphorylation, association with the plasma membrane is however, important for the effects Blk has in regulating BCR-signaling in presence of Syk. This was uncovered by studies using AtT20 cells expressing a cytoplasmic Blk mutant, which lacks the myristate modification site and was largely confined to the cytosol of the cell. Although a small amount (7-10%) of cytoplasmic Blk molecules was found in the lipid rafts and in the membrane fraction of cells, the majority was present in the soluble or cytoplasmic fractions (Figure 6.12 B and C). Since this mutant form of Blk lacks any motifs that allow for plasma membrane localization, it should be only present in the cytoplasm. Its presence in the lipid rafts and membrane fraction could be due to its over-expression in the cells or its association with other proteins. Clones with lower levels of cytoplasmic Blk were not available since all transfectants expressed very high amounts of this mutant Blk. The AtT20 BCR+ Fyn+ Syk+ sol Blk+ [#3] clone contained approximately 15-20 times more cytoplasmic Blk compared to the wild type form of Blk found in WEHI 231 and CH31 B lymphoma cell lines, which expressed the highest levels of Blk in all the B lymphoma cell lines examined (data not shown). The BCR-induced Akt phosphorylation in AtT20 cell expressing the cytoplasmic Blk mutant was completely restored to levels observed in the AtT20 BCR+ Fyn+ Syk+ cells (Figure 6.13).  Due to its cellular  localization, the cytoplasmic Blk mutant presumably cannot play a role in the inhibition of Akt phosphorylation. Even if this mutant Blk could associate with SHP-2, it has no way of recruiting it to the plasma membrane where SHP-2 could de-phosphorylate the PI 3-kinase adapter protein. Therefore, this cytoplasmic Blk mutant plays a passive role in signaling in these cells since it cannot attach to the plasma membrane.  A non-  palmitylated and non-myristylated mutant form of Lyn would probably give similar results since it would not be able to associate with the plasma membrane. In summary, altering the membrane localization of Lyn and Blk in AtT20 cells has no profound effect on BCR-mediated Akt phosphorylation. cytosolic localization is important.  However, plasma membrane as opposed  There was no inhibition of BCR-induced Akt  phosphorylation in AtT20 cells expressing the cytoplasmic mutant Blk since it could not associate with the plasma membrane and therefore could not inhibit PI 3-kinase signaling which takes place at the plasma membrane.  CHAPTER 7 The Gabl adapter protein enhances PI 3-kinase/ Akt signaling through the B cell antigen receptor  7.1  Introduction  The previous sections described how the idea that proteins regulated by the BCR, namely the Src kinases, can influence downstream PI 3-kinase/Akt signaling by interacting with the SHP-2 protein phosphatase. Other proteins such as adapters or docking proteins can also influence BCR signaling. The effect of the Gabl adapter protein on BCR signaling was examined in lymphoid and non-lymphoid cells and Gabl was found to increase BCR-mediated signaling by the PI 3-kinase pathway. These results were published in the Journal of Biological Chemistry (Ingham et al., 2001). I contributed to these studies by expressing WT and mutated Gabl in the non-lymphoid AtT20 endocrine cell system, in which the expression of the BCR has been reconstituted by DNA transfection and has a normal, functioning PI 3-kinase pathway (refer to Chapter 3). AtT20 cells do not, however, endogenously express Gabl, so the wild type and mutated forms were coexpressed and their effect(s) on BCR-mediated signaling assessed.  The Gabl adapter protein is a 97 kDa cytoplasmic protein that is a member of the growing family of PH domain-containing adapter proteins (Figure 7.1 A, top Gabl-EGFP diagram). The amino terminal PH domain of Gabl can interact with phospholipids produced by PI 3-kinase, such as PI-3,4,5-P3 that are present in the plasma membrane. In addition, Gabl also contains four P-X2-P motifs that can recruit SH3 domain-containing proteins. There are many tyrosine residues contained within the sequence of Gabl which when phosphorylated, enable it to associate with SH2 domain-containing proteins. These phosphorylated tyrosine residues act as binding sites for the SH2 domains of signaling proteins. Such SH2 domain-containing signaling proteins include the regulatory p85  subunit of PI 3-kinase, the SHP-2 protein tyrosine phosphatase and the She adapter protein. Thus, the many tyrosine residues, protein-protein and protein-lipid interaction domains within Gabl enable it to be recruited to the plasma membrane and form signaling complexes with various molecules.  It was hypothesized that Gabl expression in the AtT20 cells would result in increased PI 3-kinase/Akt pathway activation following BCR cross-linking.  Although Gabl can  associate with SHP-2, which was found to be a negative regulator of PI 3-kinase signaling in the previous chapters, it is thought that Gabl can act as a positive regulator of this pathway since PI 3-kinase activity was associated with Gabl immune complexes following BCR activation (Ingham et al., 1998). This may be due to the fact that the sequence of Gabl contains three potential binding sites for PI 3-kinase (tyrosine residues 447, 472 and 589 of human Gabl) and only has one binding site for SHP-2 (tyrosine 627 of human Gabl). Thus, although SHP-2 can be recruited to the plasma membrane by Gabl, its negative effect on PI 3-kinase signaling is overcome or masked by the fact that multiple PI 3-kinase molecules are also recruited to the plasma membrane. In order to achieve this increased activation, it was predicted that: (1) Gabl would be recruited to the plasma membrane in response to BCR signaling; (2) Gabl would be phosphorylated (on tyrosine residues) in order to interact with the SH2 domain of p85 and thus affect its activity; and (3) the PH domain of Gabl would be required for these events to occur. The non-lymphoid AtT20 cell system was used in conjunction with studies in lymphoid cells to determine if Gabl expression was sufficient to effect BCR-mediated PI 3-kinase signaling. One major advantage of the AtT20 cell system was that in addition to being able to perform biochemical experiments on the transfected Gabl protein, morphological studies using confocal microscopy could be more easily performed in order to assess changes in its intracellular location. This is because AtT20 cells adhere to plastic dishes and spread out, making it easier to see translocation of proteins from intracellular locations in the cytosol to the plasma membrane. In contrast, B lymphocytes are small cells with large nuclei and a tiny rim of cytoplasm which makes identifying the location of cytosolic components more difficult to monitor.  Using the AtT20 cell system, I show that Gabl plays a role in signaling downstream of the BCR. Recruitment of Gabl from the cytoplasm to the plasma membrane is observed following BCR cross-linking. This translocation required the PH domain of Gabl as the mutant protein lacking this domain failed to show this receptor-induced translocation. The translocation of Gabl to the plasma membrane also required PI 3-kinase-generated phospholipids since PI 3-kinase inhibitors were able to block this process.  BCR  activation also resulted in inducible tyrosine phosphorylation of Gabl. This tyrosine phosphorylation was mediated by Syk since very little phosphorylation of Gabl was observed in cells that did not express this BCR-associated PTK.  Several tyrosine  phosphorylated proteins were found to co-immunoprecipitated with Gabl from the transfected AtT20 cells including SHP-2 and She, and these proteins also coimmunoprecipitate with Gabl in B cells (Ingham et al., 1998). Furthermore, BCRinduced Akt phosphorylation in cells co-expressing wild type Gab 1 and Syk was higher compared to cells expressing the APH domain mutant form of Gabl. These results were also similar to those obtained in experiments involving B cells (Ingham et al., 1998, Ingham et al., 2001). These results provide evidence for the role the Gabl adapter protein as an amplifier of BCR-induced PI 3-kinase/Akt pathway activation.  7.2  Results  7.2.1  Gabl is recruited to the plasma membrane in response to BCR cross-linking.  To provide morphological evidence for the location of Gabl, its translocation in response to BCR activation and to determine if its PH domain was required for this translocation, AtT20 cells were transfected (as described in the Chapter 2.3.2) with EGFP-tagged wild type Gabl (Gabl-EGFP) or with a Gabl mutant lacking the PH domain (GablAPHEGFP) (Figure 7.1 A; Appendix, Table 1) (performed by May Dang-Lawson, the laboratory technician). Clones stably expressing relatively similar amounts of the wild type and APH domain mutant were identified by Western blotting and chosen for further experimentation (Figure 7.1 B).  Figure 7.1 EGFP-tagged wild type and APH Gabl structure and expression in transfected AtT20 cells. (A) Schematic representation of Gabl-EGFP and GablAPHEGFP proteins. The PH domain, proline-rich motifs and various tyrosine residues (Y) are indicated on the diagram. The Y* indicates the SHP-2 binding site; the Y** indicate the potential binding sites for the p85 subunit of PI 3-kinase. (B) Following lysis of the stably transfected clones of AtT20 cells, 35 pg of total cell extract was separated by SDSPAGE and analyzed by Western blotting with antibodies that recognize Gabl. The filters were subsequently stripped of antibodies and re-probed with anti-Syk antibodies (lower panels). The molecular weight standards (in kDa) are indicated to the left. The stable clones are representative of five recovered Gab-expressing clones.  szz  0) 3  DO  6  D) cr  m IV) Ul  o  CO o AtT20 BCR+ Fyn+ AtT20 BCR+ Fyn+ Gab1-EGFP AtT20 BCR+ Fyn+ Gab1APH-EGFP  "0 co "0 CO  00 0> "0 3) CO  6 cr  "0 CO  CD O  ro cn  AtT20 BCR+ Fyn+ Syk+ AtT20 BCR+ Fyn+ Syk+ Gab1-EGFP AtT20 BCR+ Fyn+ Syk+ Gab1APH-EGFP  0) O" > TI I m  CD  O  w cr m O ~n "0  1 o II  cr  - U £ cn 03  rn  O  It has been previously shown that transfected Gabl translocates to the plasma membrane following stimulation of EGF receptor in fibroblast cells (Rodrigues et al., 2000). However, the translocation of Gabl to the plasma membrane in response to BCR signaling has never been shown. Using confocal microscopy, the translocation of GablEGFP and GablAPH-EGFP to the plasma membrane in the presence of Syk was examined in transfected AtT20 cells following BCR cross-linking. In unstimulated cells expressing wild type Gabl-EGFP, the Gabl fusion protein was mostly cytoplasmic (Figure 7.2 a-c). Following 15 minutes of BCR cross-linking, there was a significant increase in the amount of Gabl localized to the plasma membrane (Figure 7.2 d-f). Although the Gabl protein remained in the cytoplasm in both the unstimulated (Figure 7.2 a-c) and BCR-stimulated cells (Figure 7.2 d-f), prominent staining around the perimeter of the cells was observed only in the BCR-activated cells. This pattern of fluorescence was similar to that observed when unpermeabilized AtT20 BCR+ Fyn+ and AtT20 BCR+ Fyn+ Syk+ cells were stained with antibodies that recognized cell surface mlgM (Matsuuchi et al., 1992; personal observation).  Similar effects of Gabl  translocation to the plasma membrane were observed as early as 3, 5 and 10 minutes of BCR activation when time course experiments were performed (data not shown). To determine if plasma membrane localization is dependent on the PH domain of Gabl, the cytoplasmic and plasma membrane distribution of GablAPH-EGFP was examined in AtT20 cells co-expressing Syk. Like the Gabl-EGFP protein, the GablAPH-EGFP mutant was localized mostly in the cytoplasm before BCR engagement (Figure 7.2 g-i). However, unlike the wild type Gabl protein, the cellular distribution of GablAPH-EGFP was not altered even after 15 minutes of BCR cross-linking (Figure 7.2 j-1).  Translocation of Gabl-EGFP and GablAPH-EGFP to the plasma membrane following BCR activation was also examined in AtT20 cells lacking Syk. In a previous chapter of this thesis, it was established that the Src kinases are sufficient to support the BCRinduced activation of the PI 3-kinase/Akt pathway in transfected AtT20 cells (Gold et al., 1999 and Chapter 3). However, under these conditions plasma membrane translocation of the wild type Gabl protein was inefficient (data not shown). Thus, not only does Syk  activity enhance and sustain signaling of the PI 3-kinase pathway, it is also responsible for efficient translocation of the Gabl adapter protein to the plasma membrane. The enhanced signaling of the PI 3-kinase pathway by Syk results in the generation of PI3,4,5-P3 within the plasma membrane and this allows more Gabl molecules in the cytoplasm to become recruited by way of their PH domains.  To determine if Gabl recruitment is dependent on the interaction of its PH domain with PI 3-kinase-generated phospholipids within the plasma membrane, the ability of PI 3kinase inhibitors to inhibit BCR-induced Gabl translocation was examined. Treating cells with the PI 3-kinase inhibitor, LY294002, significantly inhibits BCR-induced translocation of Gabl-EGFP to the plasma membrane (Figure 7.3).  However, the  inhibition was not complete as a small amount of Gabl-EGFP was found at the plasma membrane following BCR stimulation. This translocation may have been due to a small amount of PI-3,4,5-P3 produced by PI 3-kinase even in the presence of LY294002. This minute amount of PI-3,4,5-P3 was probably sufficient to cause the recruitment of Gabl to the plasma membrane since its PH domain interacts with PI-3,4,5-P3 with very high affinity (Isakoff et al., 1998; Maroun et al., 1999a). Similar results were obtained when cells were treated with wortmannin, another PI 3-kinase inhibitor (data not shown). In summary, Gabl recruitment in response to BCR activation requires Syk kinase and is mediated by the interaction of its PH domain with PI 3-kinase-generated phospholipids within the plasma membrane.  Figure 7.2 Gabl-EGFP translocates to the plasma membrane following BCR crosslinking. Single optical sections were taken from the middle of the stack of confocal images of AtT20 BCR+ Fyn+ Syk+ Gabl-EGFP+ (a-f) and AtT20 BCR+ Fyn+ Syk+ GablAPH-EGFP+ (g-1) cells that were stimulated or left unstimulated. Cells were grown on poly-D-lysine coated coverslips in DMEM containing 10% FCS. One day prior to fixing cells, the media in which the cells were grown was changed to DMEM containing 0.2% FCS and incubated overnight (serum-starvation conditions). The cells were then washed with D-PBS and stimulated with 20 pg/ml anti-mouse IgM antibodies (to activate the BCR) in modified HEPES-buffered saline for 15 minutes at 37°C prior to fixing in paraformaldehyde and mounting onto glass slides. Images were collected using the BioRad Radiance Plus confocal microscope. The scale bar, under panel 1, represents 10 pm. The flattened projections (c, f, i and 1) represent a composite of the 90-110 sections analyzed for each sample shown. experiments.  These images are representative of five  Unstimulated  + Anti-IgM (15 min)  Single confocal sections  Flattened projections of all sections  + Anti-IgM (15 min)  Unstimulated  j  Hjr 1  i  • 1  fern ''  C  T \ V  ^V  Single confocal sections I  Flattened projections of all sections  G M p  Figure 7.3 Recruitment of EGFP-tagged Gabl to the plasma membrane following BCR engagement is PI 3-kinase dependent. Single optical sections were taken from the middle of the stack of confocal images of AtT20 BCR+ Fyn+ Syk+ Gabl-EGFP+ cells that were stimulated or left unstimulated. The cells were cultured and prepared for stimulation as described in Figure 7.2. The cells were incubated in buffer alone (a-c) or in 25 pM LY294002 (d-f). The cells were then incubated with 20 pg/ml anti-mouse IgM antibodies for 15 minutes or left unstimulated. The cells were fixed and analyzed by confocal microscopy as described in Figure 7.2. The scale bar, under panel f, represents 10 pm.  These images shown are representative of results from two independent  experiments.  AtT20 BCR+ Fyn+ Syk+ Gabl-EGFP Unstimulated a M B H S B P I  | |  + Anti-IgM (15 min) b  K  S  ^  NO INHIBITOR  LY294002  Single confocal sections  i  C  7.2.2  Gabl is tyrosine phosphorylated in response to BCR engagement  Gabl is tyrosine phosphorylated in response to growth factor receptor activation (Holgado-Mardruga et al., 1996; Holgado-Mardruga et al., 1997; Maroun et al., 1999a; Schaeper et al., 2000). This phosphorylation presumably occurs at the plasma membrane and would therefore require that Gabl be translocated to the plasma membrane via its PH domain. Gabl has been previously shown to be tyrosine phosphorylated in response to BCR activation (Ingham et al., 1998). However, it is not known if the PH domain of Gabl is required for this event to occur. The requirement of the PH domain and of Syk kinase activity for Gabl tyrosine phosphorylation in response to BCR cross-linking was investigated in various AtT20 transfected cells.  When Gabl-EGFP and GablAPH-EGFP were immunoprecipitated from transfected AtT20 cells (AtT20 BCR+ Fyn+ Gabl-EGFP+ and AtT20 BCR+ Fyn+ GablAPHEGFP+ cells, respectively), no increase in tyrosine phosphorylation of the wild type Gabl-EGFP protein was observed following BCR cross-linking (Figure 7.4 A). In addition, very low tyrosine phosphorylation of the GablAPH-EGFP mutant form was observed before and after receptor activation. These results were expected since Syk PTK is essential for enhancing and sustaining activation of several downstream signaling events mediated by the BCR.  Gabl-EGFP was inducibly tyrosine phosphorylated  following 15 minutes of BCR engagement in the AtT20 BCR+ Fyn+ Syk+ GablEGFP+ cells (Figure 7.4 B). This tyrosine phosphorylation was also observed as early as 5 minutes of BCR engagement (data not shown). The tyrosine phosphorylation of the GablAPH-EGFP mutant form from AtT20 BCR+ Fyn+ Syk+ GablAPH-EGFP+ cells, however, did not increase following BCR cross-linking in the presence of Syk. Therefore, deletion of its PH domain abolishes the ability of Gabl to become tyrosine phosphorylated in response to BCR stimulation.  Figure 7.4  Gabl-EFGP is inducibly tyrosine phosphorylated following BCR  engagement in AtT20 cells co-transfected with Syk PTK. AtT20 BCR+ Fyn+ (A) and AtT20 BCR+ Fyn+ Syk+ (B) cells transfected with Gabl-EGFP or GablAPH-EFGP were stimulated with 20 pg/ml of anti-mouse IgM antibodies for 15 minutes (+) or left unstimulated (-). Gabl was immunoprecipitated from 1000 pg of total cell extract. The immunoprecipitates were separated by SDS-PAGE and transferred to nitrocellulose. The filters were probed with the anti-phosphotyrosine monoclonal antibody (4G10) and subsequently re-probed with anti-Gab 1 antibodies (lower panel). Molecular weight standards (in kDa) are indicated to the left. The data shown represents one of four independent experiments with similar results.  (A)  (B) AtT20 BCR+ Fyn+ AtT20 BCR+ Fyn+ Gabl-EGFP  Gab1APH-EGFP  Anti-IgM (15 min)  <—Gabl-EGFP < — G a b l APH-EGFP  125K>  IPPT: anti-Gab1 IB: anti-phospho-tyrosine  Gab1-EGFP Gabl APH-EGFP  125Re-probe: anti-Gab1  AtT20 BCR+ Fyn+  AtT20 BCR+ Fyn+  Syk+ Gabl-EGFP Syk+ Gabl APH-EGFP Anti-IgM (15 min)  <—Gabl-EGFP < — G a b l APH-EGFP  125IPPT: anti-Gab1 IB: anti-phospho-tyrosine  Gabl-EGFP Gabl APH-EGFP  125 Re-probe: anti-Gab1  7.2.3  Gabl interacts with various tyrosine phosphorylated proteins following BCR cross-linking  Gabl has been shown to interact with a wide variety of signaling components. Some of these molecules include the SHP-2 protein tyrosine phosphatase, the She adapter protein and the p85 regulatory subunit of PI 3-kinase (Ingham et al., 1998). The ability of these signaling molecules to interact with the Gabl-EGFP in transfected AtT20 cells and the requirement of the PH domain for these interactions was further defined. When the Gabl proteins were immunoprecipitated from AtT20 BCR+ Fyn+ Gabl-EGFP+ and AtT20 BCR+ Fyn+ GablAPH-EGFP+ cells, no highly tyrosine phosphorylated proteins were found associated with either the wild type or mutant forms (Figure 7.5 A). However, several tyrosine phosphorylated proteins (pp85, pp65, pp50 and pp45) were present in the Gabl immune complexes following BCR cross-linking from cells co-expressing Syk and the wild type form of Gabl (AtT20 BCR+ Fyn+ Syk+ Gabl-EGFP+) (Figure 7.5 B). Re-probing these filters (along with parallel results in lymphoid cells by our collaborator) revealed that phosphoproteins pp65 and pp50 were SHP-2 and the higher molecular weight form of the She adapter protein, respectively (Figure 7.5 B lower panels). The lower molecular weight form of She could not be detected because it migrates to the same location on the gel as the immunoprecipitating antibody. The pp85 phosphoprotein was most likely the p85 subunit of PI 3-kinase, although no reliable antibodies were available that could detect the p85 endogenously present in AtT20 cells. Although the pp50 and pp70 phosphoproteins were also detected in the unstimulated samples, the presence of SHP-2 and She in the Gabl immune immunoprecipitates was only observed in the BCR-stimulated samples. It is possible that these proteins interact with GablEGFP prior to BCR activation, however the amount may be too small to be detected by Western blotting. The interaction between SHP-2 and She proteins with GablAPHEGFP mutant form could not be detected (Figure 7.5 B lower panels). Small amounts of the tyrosine phosphorylated proteins pp50 and pp70, corresponding to She and SHP-2, respectively, were present in the Gabl immunoprecipitates from unstimulated and BCRstimulated AtT20 BCR+ Fyn+ GablAPH-EGFP+ cells. However, re-probing the filters  Figure 7.5 Gabl-EFGP interacts with several tyrosine phosphorylated proteins following BCR cross-linking in AtT20 cells co-transfected with Syk. AtT20 BCR+ Fyn+ (A) and AtT20 BCR+ Fyn+ Syk+ (B) cells transfected with either Gabl-EGFP or GablAPH-EFGP were stimulated with 20 jig/ml of anti-mouse IgM antibodies for 15 minutes (+) or left unstimulated (-). Gabl was immunoprecipitated from 1000 |xg of total cell extract. The immune complexes were separated by SDS-PAGE and transferred to nitrocellulose.  The filters were probed with the anti-phosphotyrosine monoclonal  antibody (4G10) and subsequently re-probed with anti-Gabl, anti-SHP-2 and anti-She antibodies (lower panels). Molecular weight standards (in kDa) are indicated to the left. The data shown represents one of three independent experiments with similar results.  AtT20 BCR+ Fyn+ Gab1-EGFP  AtT20 BCR+ Fyn+ Gab1 APH-EGFP  (A) < <  Gab1-EGFP Gab1 APH-EGFP  IPPT: anti-Gab1 IB: anti-phospho-tyrosine Gab1-EGFP Gab1 APH-EGFP  125 Re-probe: anti-Gab1  AtT20 BCR+ Fyn+ Syk+ Gabl-EGFP  AtT20 BCR+ Fyn+ Syk+ Gabl APH-EGFP  < <  Gabl-EGFP Gabl APH-EGFP  -<  pp85 (p85 subunit) pp70 (SHP-2)  < <  pp50 (p52 She) pp46 (p46 She)  IPPT: anti-Gab1 IB: anti-phospho-tyrosine Gabl-EGFP Gabl APH-EGFP  125 Re-probe: anti-Gab1  SHP-2  65Re-probe: anti-SHP-2  p52 She Immunoprecipitating antibody  Re-probe: anti-She  with antibodies that recognize these proteins indicated that they were not present in the Gabl immunoprecipitates. Again, the amount of SHP-2 and She that associated with the Gab 1 APH-EGFP mutant form may be too little to be detected by the Western blotting technique used.  High levels of SHP-2 and She were only detected in the Gabl  immunoprecipitates from cells co-expressing Syk and Gabl-EGFP, suggesting that tyrosine phosphorylation of Gabl is required for mediating its interaction with these signaling components. This was expected since SHP-2, the p85 subunit and She interact with Gabl via their SH2 domains in B cells (Ingham et ah, 1998).  7.2.4  BCR-induced Akt phosphorylation is enhanced by Gabl  Gabl contains three tyrosine residues located within consensus sequences that are recognized by the SH2 domain of the p85 subunit of PI 3-kinase (Holgado-Madruga et ah, 1996). The ability of these two proteins to interact with one another suggests that they act in the same signaling pathway and may in some way regulate each other's function. Therefore, addition of Gabl to AtT20 cells may result in enhanced PI 3-kinase and Akt activation. To test this hypothesis, PI 3-kinase/Akt pathway activation was examined in the AtT20 BCR+ Fyn+ and AtT20 BCR+ Fyn+ Syk+ cells transfected with either wild type Gabl (Gabl-EGFP) or Gabl with the deleted PH domain (GablAPHEGFP) following different periods of BCR engagement. Akt activation was determined by examining the levels of Akt phosphorylation using the phospho-specific antibodies used in the previous chapters.  Addition of Gabl-EGFP and Gabl-APH-EGFP to AtT20 BCR+ Fyn+ cells did not affect BCR-induced Akt phosphorylation (Figure 7.6 A). This was expected since enhanced and sustained Akt phosphorylation requires Syk PTK (Chapter 3, Section 3.2.2). Following various periods of BCR cross-linking, the phosphorylation of Akt on the serine 473 residue was greater in cells co-expressing Syk PTK and Gabl-EGFP compared to cells transfected only with Syk (Figure 7.6 B). BCR-induced Akt phosphorylation in both the AtT20 BCR+ Fyn+ Syk+ and AtT20 BCR+ Fyn+ Syk+ Gabl-EGFP+ cell lines  was sustained over 60 minutes of receptor activation. However, this phosphorylation was 2-fold greater in AtT20 BCR+ Fyn+ Syk+ Gabl-EGFP+ cells compared to the AtT20 BCR+ Fyn+ Syk+ cells that did not express the wild type Gabl, as quantified using the ImageQuant 5.1 program. In cells co-transfected with Syk and the GablAPH-EGFP, Akt phosphorylation was no different from that observed in the AtT20 BCR+ Fyn+ Syk+ cell line at all time points of BCR activation (Figure 7.6 B). Thus, addition of wild type Gabl to AtT20 cells expressing Syk results in the enhancement of Akt phosphorylation, presumably due to increased PI 3-kinase activation. In addition, the increased activation of this signaling pathway is dependent on an intact PH domain in Gabl and on its ability to translocate to the plasma membrane. These results in the transfected AtT20 cells are consistent with the data obtained from B cell lines (Ingham et al., 2001).  7.3  Discussion  A strategy used by many cell surface receptors including the BCR for recruitment of signaling components involves creation of docking sites on adapter proteins. It was established in this chapter that the Gabl adapter protein plays a role in signaling downstream of the BCR.  Gabl was previously shown to be present in membrane-  enriched particulate fraction in some B cell lines (Ingham et al., 1998). However, its ability to translocate to the plasma membrane in response to B cell activation has never been reported. In this study, it was established that Gabl translocates from the cytoplasm to the plasma membrane in response to BCR cross-linking in transfected AtT20 cells (Figure 7.2 A).  This process requires the PH domain of Gabl and is PI 3-kinase-  dependent (Figure 7.2 B and 7.3). Thus, association of Gabl with the plasma membrane is mediated by the interaction of its PH domain with PI 3-kinase-derived phospholipids within the plasma membrane such as PI-3,4,5-P3. Gabl translocation to the plasma membrane following growth factor receptor activation has been previously shown (Maroun et al., 1999a). However, this is the first published data showing that BCR crosslinking results in Gabl translocation to the plasma membrane.  Figure 7.6 Wild type Gabl-EFGP, but not GablAPH-EGFP, potentiates BCRinduced Akt phosphorylation in AtT20 BCR+ Fyn+ cells co-expressing Syk PTK. AtT20 BCR+ Fyn+ cells (A) and AtT20 BCR+ Fyn+ Syk+ cells (B) were stimulated with 20 i^g/ml of anti-mouse IgM antibodies for. the indicated time periods. Following cell lysis, 25 |0,g of total cell extract was separated by SDS-PAGE and analyzed by Western blotting with anti-phospho-serine 473 Akt antibodies. The filters were subsequently stripped of antibodies and re-probed with anti-Akt antibodies (lower panels). The molecular weight standards (in kDa) are indicated to the left.  These results are  representative of three independent experiments with similar results. The bands were quantified using the ImageQuant 5.1 software and the amount of phospho-Akt in AtT20 BCR+ Fyn+ Syk+ Gabl-EGFP+ cells found to be approximately 2-fold greater than in cells without Gabl-EGFP.  AtT20 BCR+ Fyn+  (A)  0 5  15 30 60 0 5  —  15 30 60  —  —  '  rAtT20 BCR+ Fyn+ AtT20 BCR+ Fyn+ G a b 1 A p H . E G F P Anti-IgM (min)  65-  0 5 15 30 60  —  —  -  0 5 15 30 60  — "*"  A t T 2 0 B C R + Fyn+ Syk+  (B) Anti-IgM (min)  0  5  15 30 60  A t T 2 0 B C R + Fyn+ Syk+ G a b l - E G F P 0  5  15 30 60  IB: anti-phospho65serine473-Akt  Re-probe: anti-Akt  65  AtT20 BCR+ Fyn+ AtT20 BCR+ Fyn+ Syk+ Syk+Gab1 APH-EGFP Anti-IgM 0 5 15 30 60 0 5 15 30 60 (min) IB: anti-phosphoserine473-Akt  65-  Re-probe: anti-Akt  65  —  —  ***** *#(»* - rm -rr __  —  As was also shown in B cells, Gabl was inducibly tyrosine phosphorylated in response to BCR cross-linking in the AtT20 transfected cells (Figure 7.4 A).  This tyrosine  phosphorylation was mediated by the Syk PTK, as no Gabl phosphorylation was observed in cells lacking this BCR-associated kinase (Figure 7.4 A). Additionally, the PH domain was also required for Gabl tyrosine phosphorylation since no inducible phosphorylation of the APH domain mutant form was observed following cell stimulation (Figure 7.4 A). Since the Gabl mutant lacking the PH domain did not translocate to the plasma membrane and did not become inducibly tyrosine phosphorylated following BCR activation, then Gabl recruitment to the plasma membrane must precede its tyrosine phosphorylation. If Gabl phosphorylation occurs in the cytoplasm, an inducible increase in tyrosine phosphorylation of the GablAPH-EGFP mutant form would have been observed in BCR stimulated cells.  However, these results suggest that tyrosine  phosphorylation of Gabl occurs at the plasma membrane following its recruitment from the cytoplasm and is mediated by the BCR-associated PTK, Syk.  Several tyrosine phosphorylated proteins interact with the Gabl following BCR crosslinking (Figure 7.5 B). Two of these proteins are the SHP-2 protein tyrosine phosphatase and the She adapter protein. The p85 subunit of PI 3-kinase is also known to associate with Gabl (Ingham et al., 1998) and a phosphoprotein of 85 kDa was found in the Gabl immune complexes from AtT20 cell co-expressing Syk and Gabl-EGFP (Figure 7.5 B). However, no antibodies were available that could reliably detect the endogenous p85 isoform present in the AtT20 cells. The association of SHP-2 and She with the wild type Gabl was only observed following BCR cross-linking (Figure 7.5 B lower panels). Additionally, an increase in tyrosine phosphorylation of these signaling components was also observed following BCR activation. Interestingly, SHP-2 and She do not interact with the APH domain mutant of Gabl in cell co-expressing Syk, indicating that tyrosine phosphorylation of Gabl is required for these interactions. Indeed, SHP-2, She and p85 have been shown to interact with Gabl through their SH2 domains (Ingham et al., 1998).  It is somewhat unusual that Gabl is able to recruit both positive (p85 subunit and She) and negative (SHP-2) regulators of BCR signaling (reviewed by Huyer and Alexander,  1999). Interaction of SHP-2 with Gabl may be able to regulate its phosphatase activity. Increased phosphatase activity of SHP-2 is thought to require tyrosine phosphorylation (Vogel et al., 1993) and interaction of its SH2 domain with phosphotyrosine residues on other proteins (Sugimoto et al., 1993; Pulsky et al., 1995; Dechert et al., 1996; reviewed by Barford and Neel, 1998). Since SHP-2 interacts with Gabl through its SH2 domain(s) and since this Gabl-associated SHP-2 becomes inducibly tyrosine phosphorylated following BCR cross-linking, it is likely that the association between these two molecules results in increased phosphatase activity of SHP-2. Conversely, some information is available that suggests that SHP-2 may also be able to regulate Gabl function in B cells. It was previously shown that Gabl and its family member Gab2 are substrates for SHP-2 (Herbst et al., 1996; Gu et al., 1997; Zhang et al., 2002). SHP-2 can dephosphorylate tyrosine phosphorylated residues and down-regulate Gabl- and Gab2-mediated signaling. Indeed, results from the previous chapters have implicated SHP-2 as negative regulator of the PI 3-kinase/Akt pathway. As well, Akt phosphorylation is enhanced in the DT40 chicken B lymphoma cells that are deficient SHP-2 (MR Gold, unpublished observation). Therefore, SHP-2 may be responsible for attenuating Gabl-mediated PI 3-kinase activation by dephosphorylation of the p85 binding sites on Gabl. Although Gablassociated SHP-2 may act to negatively regulate PI 3-kinase/Akt pathway signaling, PI 3kinase activity was also shown to be important for SHP-2 interaction with Gabl (Yart et al., 2001). Perhaps Gabl and PI 3-kinase interaction occurs first, leading to PI 3-kinase activation. This activation results in the association between SHP-2 and Gabl, leading to activation of SHP-2's phosphatase activity.  The activated SHP-2 subsequently  dephosphorylates tyrosine residues on Gabl leading the dissociation of the signaling complex. Thus, the association of the various signaling molecules with Gabl acts both to amplify and attenuate Gab 1-mediated signaling. Alternatively, SHP-2 association with Gabl could also regulate the phosphorylation of another protein that is also bound to Gabl. For example, SHP-2 can de-phosphorylate She, leading to the down-regulation of MAP kinase signaling or it can de-phosphorylate PI 3-kinase, leading to the downregulation of PI 3-kinase signaling. In this way, the Gabl-associated SHP-2 molecules can also attenuate Gabl-mediated signaling. SHP-2 may also be able to regulate Gabl signaling by simply being recruited to the plasma membrane. The interaction between  Gabl-SHP-2 could act to target SHP-2 to the plasma membrane where it could dephosphorylate its substrates, including adapter proteins that are responsible for PI 3kinase recruitment.  Since Gabl could interact with the p85 subunit of PI 3-kinase and recruit it to the plasma membrane, it was of interest to determine if Gabl plays a role in PI 3-kinase/Akt pathway activation in response BCR engagement.  Akt phosphorylation in AtT20 cells co-  transfected with Syk PTK and wild type Gabl (AtT20 BCR+ Fyn+ Syk+ Gabl-EGFP+) was 2-fold higher compared to cells expressing only Syk (AtT20 BCR+ Fyn+ Syk+) (Figure 7.6 B). This increase was not observed in cells co-expressing Syk and the APH domain mutant form of Gabl. This is most likely due to the fact that the mutant form does not become tyrosine phosphorylated and thus no docking sits are created for the p85 subunit of PI 3-kinase. In addition, the APH domain mutant cannot become recruited to the plasma membrane where the substrates for PI 3-kinase are located. Therefore, PI 3kinase does not become activated in the presence of the APH domain mutant of Gabl. In addition, no increase in BCR-induced Akt phosphorylation was observed in cells lacking Syk (Figure 7.6 A). These results shown and those obtained in B cells (Ingham et al., 1998; Ingham et al., 2001) support a model whereby BCR cross-linking leads to PI 3kinase activation and the production of PI-3,4,5-P3 within the plasma membrane (Figure 7.7). The presence of PI-3,4,5-P3 results in the recruitment of Gabl to the plasma membrane through the interaction of its PH domain with the phospholipids. Once at the plasma membrane, Gabl becomes tyrosine phosphorylated on several residues by BCRassociated PTKs like Syk, creating multiple docking sites for various signaling molecules. Once recruited to docking sites on Gabl, SHP-2, She and p85 subunit may become tyrosine phosphorylated by the BCR-associated PTKs, resulting in their subsequent activation. Activation of PI 3-kinase leads to the production of more PI3,4,5-P3 within the plasma membrane resulting in the recruitment of PH domaincontaining proteins including PDK1, PDK2, Akt and more Gabl complexes. Thus, Gabl recruitment induces a positive feedback loop by recruiting more Gabl molecules to the plasma membrane.  This model suggests that Gabl acts as an amplifier of PI 3-  kinase/Akt pathway activation downstream of the BCR.  This BCR-induced Gabl  signaling is then attenuated by the Gabl-associated SHP-2 molecules that can dephosphorylate the docking sites, leading to the dissociation of the complex and return of the signaling enzymes to the cytoplasm.  The consequence of She interaction with Gabl was not examined in the transfected AtT20 cells since the main focus of this study was the PI 3-kinase/Akt pathway. Since Gabl can associate with She in the AtT20 cells and in B cells (Ingham et al., 1998) and since Grb2 and SOS can interact with She, MAP kinase activation would most likely be affected by the presence of Gabl. Gabl has been previously shown to play a role in MAP kinase activation downstream of the receptor tyrosine kinase, c-Met (Sachs et al., 2000; Schaeper et al., 2000). Consistent with the idea that PI 3-kinase activation plays a role in MAP kinase activation, presumably because both SHP-2 and She interact with the Grb2-SOS complex, which eventually results in p21Ras and ERK activation (King et al., 1997; Deora et al, 1998; Sutor et al., 1999; Yart et al., 2001; Chaudhary et al., 2000). Thus, Gabl plays a role in MAP kinase activation not only through its interaction with She, but also through its interaction with SHP-2. Thus, in addition to playing a role in PI 3-kinase activation, Gabl plays role in linking the PI 3-kinase and MAP kinase pathways, thereby adding a way to further regulate the cross-talk between the two pathways.  Figure 7.7 Proposed model of BCR-mediated regulation of Gabl. (A) BCR crosslinking in AtT20 cells results in the recruitment of PI 3-kinase to the plasma membrane, leading to its activation and production of PI-3,4,5-P3 within the plasma membrane. This results in the recruitment of Gabl to the plasma membrane through the interaction of its PH domain with the PI 3-kinase-generated phospholipids. Once at the plasma membrane, Gabl becomes tyrosine phosphorylated by the BCR-associated PTK, Syk. (B) This phosphorylation results in the creation of docking sites for signaling molecules including the SHP-2 tyrosine phosphatase, the She adapter protein and the p85 regulatory subunit of PI 3-kinase. Recruitment of Gabl-associated PI 3-kinase results in the generation of more PI-3,4,5-P3 within the plasma membrane, thus creating a positive feedback loop.  PLASMA MEMBRANE  P'P3  P'P3  No recruitment  N> Ov  o  (B) PLASMA MEMBRANE  pip.  pip3<-  pip2 \  pip, PH  PIP3 /  WT Gabl © She  g  P'P3  P'P3 "  CHAPTER 8 Discussion  8.1  Summary of results and future directions  Signaling through the BCR involves activation of several classes of PTKs including Src family kinases, Syk and Btk. The individual role of Src family kinases in the activation of signaling pathways downstream of the BCR has not yet been clearly defined. Thus, the role of the three Src family members, Lyn, Lck and Blk, in the activation of the PI 3kinase/Akt pathway through the BCR was examined in this study. It was established that one Src kinase, in this case the Fyn homolog that is endogenously present in the AtT20 cells, is sufficient for initiating the activation of the Akt pathway following BCR crosslinking in the non-lymphoid cell system. Syk kinase activity, however, is required for amplifying and sustaining this response for a longer period of BCR activation.  Using the transfected AtT20 cell it was determined that the lymphoid-expressed Src family members Lyn, Lck and Blk, differentially regulate the Akt in response to BCR cross-linking. In the absence of Syk, the maximum level of Akt phosphorylation was reached in a shorter length of time in cells expressing Lyn. On the other hand, the peak level of Akt phosphorylation was reached following a longer period of BCR cross-linking in the presence of Blk. The extent of Akt phosphorylation in the cells expressing Lyn or Blk alone, however, was not higher or lower compared to that observed in the parental untransfected parental AtT20 cells. However, Lyn and Blk both have an inhibitory effect on BCR-induced Akt phosphorylation in the presence of Syk. This suggests that Lyn and Blk can negatively regulate BCR signaling that leads to the activation of the PI 3kinase/Akt pathway.  Interestingly, Lck neither inhibits nor enhances Akt  phosphorylation following BCR engagement both in the presence and absence of Syk PTK. Therefore, Lck does not play a role in BCR signaling in the AtT20 cell system.  This observed Lyn- and Blk-mediated inhibition of the Akt pathway is most likely due to the ability of these Src kinases to interact with the SHP-2 protein tyrosine phosphatase. This interaction is potentially mediated in part by the SH2 domains of SHP-2 and the putative ITIM motifs within the kinase domains of Lyn and Blk, as inferred from the in vitro  experiments.  The Lyn:SHP-2 interaction resulted in a 2-fold increase in  phosphorylation of SHP-2 on the tyrosine 580 residue at the carboxy terminal end. However, phosphorylation of this residue did not increase in AtT20 cells expressing Blk, as expected since Akt phosphorylation is also inhibited in these cells. Therefore, the interaction between Blk and SHP-2 may be sufficient for mediating suppression of BCRinduced Akt activation. Alternatively, the Blk-mediated inhibition of Akt activation could be caused by a different mechanism that is independent of SHP-2. It is likely that the interaction of the SH2 domain of SHP-2 with the phospho-tyrosine containing motif within Lyn and Blk is sufficient for increasing the phosphatase activity of SHP-2 in the AtT20 cells. Indeed, previous studies have shown that interaction of the SHP-2's aminoterminal SH2 domain with its ligand results in the activation of its phosphatase activity (Barford and Neel, 1998; Sugimoto et al., 1993; Pulsky et al., 1995; Dechert et al., 1996; Hof et al., 1998). However, phosphorylation may also result in the enhancement of SHP2 phosphatase activity (Vogel et al., 1993). In addition to enhancing the phosphatase activity, the interaction between SHP-2 and Lyn or Blk also provides a mechanism for recruiting this cytoplasmic phosphatase to the plasma membrane. While at the plasma membrane, the activated SHP-2 molecules can negatively regulate signaling components which have assembled there. Therefore, a model can be proposed for SHP-2-mediated inhibition of BCR-induced Akt phosphorylation in the AtT20 cells (Figure 8.1).  In the absence of Lyn and Blk, in the reconstituted AtT20 cell system, BCR cross-linking leads to the tyrosine phosphorylation of an adapter protein ("X") by the endogenous Src kinase, Fyn, and Syk PTK. This phosphorylated adapter protein can then recruit PI 3kinase to the plasma membrane of cells (Figure 8.1 A).  This plasma membrane  localization results in the production of PI-3,4,5-P3 (PIP3 in the diagram) by PI 3-kinase. The PI-3,4,5-P3 provides binding sites for PH domain-containing signaling components at  the plasma membrane. Some of these proteins include the putative PDK2, PDK1 and Akt, among others.  Recruitment of these three particular signaling components  ultimately results in the phosphorylation and activation of Akt, as described in a previous section. However, in AtT20 cells transfected with Lyn and Blk, SHP-2 molecules are associated with these Src kinases and are constitutively localized at the plasma membrane.  These membrane-localized SHP-2 molecules are responsible for de-  phosphorylating the phospho-tyrosine residues on the adapter protein, thus eliminating the docking sites for PI 3-kinase (Figure 8.1 B). The constitutive association between SHP-2 and the two Src kinases presumably allows SHP-2 to be activated at all times since its SH2 domain is bound to a substrate (the putative ITIM on Lyn and Blk). Since SHP-2 is constitutively active and localized to the plasma membrane, it can constantly de-phosphorylate the adapter protein. In these Lyn- and Blk-expressing AtT20 cells, BCR cross-linking does not result in the generation of docking sites for PI 3-kinase on the adapter protein because of the plasma membrane-localized SHP-2 molecules. No PI3,4,5-P3 is produced due to the lack of recruitment of PI 3-kinase to the plasma membrane. Akt, PDK1 and the putative PDK2 have no way becoming recruited since their PH domains have nothing to interact with at the plasma membrane. Therefore, they remain in the cytoplasm in the inactive state.  In the AtT20 cells, SHP-2 molecules located in the cytoplasm can also become tyrosine phosphorylated by activated Syk molecules that have dissociated from the BCR complex. However, these SHP-2 molecules have no way of becoming recruited to the plasma membrane and therefore cannot de-phosphorylate the PI 3-kinase adapter protein (Figure 8.2). Experiments involving the AtT20 cell transfected with the cytosolic Blk mutant support the idea that SHP-2 recruitment to the plasma membrane by the Src kinases is responsible for the inhibition of BCR-induced Akt phosphorylation. In cells expressing the cytosolic Blk mutant, BCR-induced Akt phosphorylation is similar to the untransfected parental cells presumably because SHP-2 has no way of becoming recruited to the plasma membrane in these cells. Therefore, the adapter protein for PI 3kinase can become tyrosine phosphorylated and PI 3-kinase can be recruited to the  Figure 8.1  Proposed model of Lyn- and Blk-mediated inhibition of the PI 3-  kinase/Akt pathway in AtT20 cells. (A) In AtT20 cell transfected with Lck, there is no inhibition of PI 3-kinase/Akt activation because SHP-2 is not recruited to the plasma membrane. The cytosolic SHP-2 cannot de-phosphorylate the PI 3-kinase adapter protein (X) and thus, Akt activation can occur. (B) In AtT20 cell transfected with Lyn and Blk in conjunction with Syk, SHP-2 can be recruited to the plasma membrane by these Src family kinases. At the plasma membrane, SHP-2 can de-phosphorylate adapter protein X. This leads to a decrease in the number of binding sites for PI 3-kinase on the adapter protein and lack of recruitment to the plasma membrane. In this case, PIP3 is not generated by PI 3-kinase and PH domain-containing proteins including Akt cannot be recruited and activated at the plasma membrane.  PLASMA MEMBRANE  > PIP,  PIP.  PIP,  P'P3  P'P3 Akt T308 S473 ® ®  10 Ov Ln  SHP-2  PLASMA MEMBRANE  PIP.  PIP,  PI3K  PIP2 ^ P I P 2  P I P  Akt T308 S473  Figure 8.2 Proposed location of cellular SHP-2 molecules in AtT20 cells. There are two cellular pools of SHP-2 molecules in the AtT20 cells — those found in the cytoplasm and those found associated with Lyn or Blk.  The Lyn- and Blk-associated SHP-2  molecules are responsible for the inhibition of BCR-mediated Akt phosphorylation (refer to Figure 8.1).  The SHP-2 molecules in the cytoplasm, however, cannot play an  inhibitory role since they cannot be recruited to the plasma membrane and dephosphorylate the tyrosine residues on the PI 3-kinase adapter protein (X). This pool of cytosolic SHP-2 can be phosphorylated by activated Syk molecules that have dissociated from the BCR complex at the plasma membrane.  Antigen  [Antigen  Dissociation from BCR complex \r  plasma membrane, which would ultimately lead to the phosphorylation and activation of Akt. In AtT20 cell expressing Lck, Akt phosphorylation is not inhibited since no SHP-2 molecules are localized to the plasma membrane and therefore cannot de-phosphorylate the tyrosine residues on the adapter protein that recruits PI 3-kinase.  The SHP-2  molecules in these cells are only present in the cytoplasm (Figure 8.1 A).  More experiments are required to provide further evidence to support this proposed model. Mutational analysis of the interaction between SHP-2 and Lyn or Blk would be required. To determine which amino acids, domains or regions of Lyn and Blk are required for their association with SHP-2, generation of deletion mutants of the Src kinases is necessary. Since the SH2 domain(s) of SHP-2 are thought to associate with the phospho-tyrosine-containing putative ITIM sequence within the kinase domains of Lyn and Blk, mutating the amino acids within this region would prevent the interaction between the phosphatase and the Src kinases. The ability of SHP-2 to interact with these mutant Lyn and Blk molecules could be tested in vitro using the GST-SHP-2 (SH2-SH2) fusion protein and by immunoprecipitation experiments.  BCR-induced Akt  phosphorylation should not be inhibited in AtT20 cells containing these mutant Lyn and Blk molecules if the interaction between these Src kinases and SHP-2 is responsible for the inhibition of this pathway. Alternatively, generation of an Lck mutant, which contains the putative SHP-2 binding site, would also provide useful information. If SHP2 can interact with the mutant Lck, it can be recruited to the plasma membrane. Recruitment of SHP-2 to the plasma membrane by Lck would also result in the inhibition of BCR-induced Akt phosphorylation, if membrane recruitment of the phosphatase were responsible for the inhibition of this pathway. Alternatively, a constitutively active plasma membrane-bound form of SHP-2 could also be expressed in the AtT20 cells. Generating a membrane-bound form of SHP-2 could mimic SHP-2 recruitment to the plasma membrane by Lyn and Blk. If the hypothesis that SHP-2 de-phosphorylates the PI 3-kinase adapter protein at the plasma membrane is correct, then addition of this SHP-2 mutant would result in the constitutive inhibition of the PI 3-kinase/Akt pathway, as observed in AtT20 cell expressing Lyn and Blk.  Experiments using dominant negative forms of SHP-2 would also provide evidence to support the proposed model. Since the GST-SHP-2 (SH2-SH2) fusion protein was able to inhibit the interaction between Lyn and Blk and the endogenous SHP-2 molecules in the AtT20 cells, an EGFP-tagged form of the tandem SHP-2 SH2 domains was created and is expressible in transient transfection experiments into BOSC 23 cells.  This  expression vector will be transfected into the AtT20 cells in which BCR-induced Akt phosphorylation was inhibited (the AtT20 BCR+ Fyn+ Syk+ Lyn+ and AtT20 BCR+ Fyn+ Syk+ Blk+ cell lines). The expression of this protein may inhibit the interaction between Lyn and Blk with the endogenous SHP-2 molecules within the cells and allow for the activation of the Akt pathway in response to BCR cross-linking. Recruitment of the endogenous SHP-2 molecules to the plasma membrane by Lyn and Blk may be inhibited by the presence of the EGFP-tagged SHP-2 SH2 domains. Since the EGFPtagged SHP-2 SH2 domains lack phosphatase activity, it may be unable to participate in the negative regulation of any downstream signaling pathways thus blocking the inhibitory effect of SHP-2 on PI 3-kinase/Akt activation. Experiments involving an epitope-tagged catalytically inactive form of SHP-2 would also provide the same information as the SH2 domain construct. Loss-of-function experiments could also be performed to test the hypothesis that SHP-2 is responsible for the inhibition of the PI 3-kinase/Akt pathway.  RNAi (RNA  interference) technique could be used to deleted or greatly reduce the amount of SHP-2 molecules within the AtT20 cells. This lack of or decrease in the amount of SHP-2 molecules within the cells would prevent inhibition of the PI 3-kinase/Akt pathway following BCR engagement. The finding that SHP-2 can interact with the Src kinases Lyn and Blk is a novel discovery that has not been shown in other systems, including B lymphocytes.  Preliminary experiments examining the association between these  molecules in B lymphocytes have been performed, but will not be discussed in the context of this thesis.  Signaling through the BCR requires docking, scaffold or adapter proteins. One of these proteins is Gabl (Ingham et al., 1998). Using the AtT20 cell system, it was established  that Gabl translocates to the plasma membrane in response to BCR cross-linking. This translocation subsequently results in Gabl phosphorylation on several tyrosine residues and its association with signaling components including the SHP-2 protein tyrosine phosphatase and the She adapter protein. Addition of Gabl to the AtT20 cells also results in a 2-fold increase in Akt phosphorylation following BCR engagement. Although interaction between the PI 3-kinase p85 regulatory subunit and Gabl could not be examined due to the lack of antibodies that recognized the p85 isoform in the AtT20 cells, an 85 kDa tyrosine phosphorylated protein is present in the Gabl immunoprecipitates. The translocation of Gabl to the plasma membrane, its inducible tyrosine phosphorylation and association with signaling components following BCR cross-linking all required the PH domain and the BCR-associated PTK, Syk. Based on the data, a model can be proposed whereby BCR cross-linking leads to phosphorylation of an adapter protein resulting in the recruitment of PI 3-kinase to the plasma membrane and subsequent production of PI-3,4,5-P3 (Figure 7.7). This enables Gabl to be recruited to the plasma membrane through its PH domain. Following this, translocation, Gabl becomes tyrosine phosphorylated by BCR-associated Syk molecules resulting in the creation of docking sites for SH2 domain-containing signaling components including PI 3-kinase, SHP-2 and She. Recruitment of more PI 3-kinase molecules to the plasma membrane results in increased production of PI-3,4,5-P3 and recruitment of more signaling molecules that contain PH domains including Akt. Therefore, Gabl acts as an enhancer or amplifier of PI 3-kinase signaling through the BCR.  Although Gabl recruited SHP-2 to the plasma membrane, Akt phosphorylation in these cells is not inhibited as was observed in AtT20 cells expressing Lyn or Blk. It would be interesting to determine if BCR-induced Akt phosphorylation would be further increased if SHP-2 could not associate with Gabl. It would also be interesting to determine if SHP-2 can de-phosphorylate Gabl in the cells and when this occurs after BCR crosslinking.  The ability of SHP-2 to have both a positive effect on PI 3-kinase/Akt signaling in the presence of Gabl and a negative effect in the presence of Lyn and Blk may be due to its  interaction with these signaling components. In the case with Lyn and Blk, SHP-2 is constitutively associated with these Src kinases and therefore constitutively localized at the plasma membrane. These Src-associated SHP-2 molecules are presumably active since their SH2 domain(s) are associated with their substrate. In these cells, the activated SHP-2 molecules can constantly de-phosphorylate the PI 3-kinase adapter protein at the plasma membrane thereby preventing the recruitment and activation of PI 3-kinase and Akt. However, in the AtT20 cells expressing Gabl, SHP-2 is recruited only after BCR cross-linking. In these cells, PI 3-kinase activation can occur since the PI 3-kinase adapter protein is not constitutively de-phosphorylated in resting unstimulated cells. Following BCR cross-linking, the adapter protein can become tyrosine phosphorylated and PI 3-kinase activation can occur. As well, additional PI 3-kinase molecules can be recruited to the plasma membrane by Gabl, further adding to Akt activation.  The AtT20 cell system has proved to be useful in studying plasma membrane proximal signaling events downstream of the BCR. Since AtT20 cells do not express other Src family kinases, besides the endogenous Fyn, they have enabled us to examine the unique roles of some lymphoid-expressed Src kinases and Syk PTK. There are, however, some disadvantages associated with using these cells to study BCR signaling.  The most  obvious disadvantage to using the AtT20 cells to study signaling through the BCR is the fact that they lack lyphoid-specific components. These components may be required for the activation of the different pathways. In addition, the membrane composition of the AtT20 cells may be different from that of B lymphocytes. This may affect the way in which the signaling pathways are initiated following BCR cross-linking. Since the BCR was introduced into the AtT20 cells, it is possible that signaling following BCR crosslinking is different from that which occurs in B lymphocytes. Therefore, it is important to also perform similar expermients in B lymphocytes to determine if the results obtained from the AtT20 cells are biologically significant. Finally, the AtT20 cells were isolated from a mouse pituitary tumor and immortalized in culture. Thus, it is possible that the various signaling pathways and components within them are dysregulated. Given these caveats, it is important to be cautious when interpreting or analyzing data obtained using the AtT20 cells.  8.2  Discussion  The Src kinase and Gabl adapter protein data shown here demonstrate the importance of localization or compartmentalization of signaling molecules within the cell. Specifically, the importance of plasma membrane localization and recruitment is highlighted by these examples. For example, the localization of SHP-2 to the plasma membrane by Lyn and Blk is important for mediating the inhibition of BCR-induced Akt phosphorylation. Lack of plasma membrane localization of Blk resulted in loss of this inhibition suggesting that Blk and presumably Lyn, must by localized at the plasma membrane in order to play a role in the negative regulation of the Akt pathway. In this way Lyn and Blk also act as adapter proteins by recruiting SHP-2 to the plasma membrane. In the case with Gabl, its recruitment to the plasma membrane by the PH domain is important for its function. Without the PH domain, Gabl cannot translocate to the plasma membrane in response to BCR cross-linking and it cannot become tyrosine phosphorylated by the BCR-associated PTKs which are located at the plasma membrane. This lack of tyrosine phosphorylation prevents it from participating in the recruitment and activation of several signaling components.  Over the past five years, the importance of docking and adapter proteins in the regulation of the major signal transduction pathways within the cell has been demonstrated. Adapter and docking proteins are proteins with no intrinsic catalytically activity and are for the most part composed of various protein-protein and/or protein/lipid interaction domains (reviewed by Pawson and Scott, 1997; Norian and Koretzky, 2000; Jordan et al., 2003). Since adapter proteins have multiple interaction domains, they are capable of associating with several proteins which results in the formation of signaling complexes. Some adapter proteins can associate with receptors or lipids within the plasma membrane and therefore enable the signaling complexes to be recruited to the plasma membrane.  In lower organisms some adapter proteins have been found to be contiguous with cell surface receptors. Two examples of this are found in Drosophila and C. elegans. In these organisms, the IRS-1-like adapter protein is contiguous with the insulin receptor-like  ortholog (daf-2 in C. elegans and Dir in Drosophila) (Fernandez et al., 1995; Ruan et al., 1995; Kinura et al., 1997). Upon binding of the insulin-like ligand to the insulin receptor-like molecule, the activated kinase contained within the receptor phosphorylates the IRS-1 on multiple tyrosine residues. This phosphorylation creates docking sites for effector molecules including PLC-y, SHP-2 and the p85 subunit of PI 3-kinase. In mammals, adapter proteins tend to be separate from cell surface receptors. This is advantageous because adapter proteins can be used in more than one signaling pathway. The use of one adapter protein in multiple pathways also allows for cross-talk between these pathways in higher organisms.  The ability of adapter proteins to recruit signaling complexes to the plasma membrane is important for the activation of most signaling pathways regulated by the BCR. The importance of plasma membrane localization of PI 3-kinase was demonstrated with Gabl in the transfected AtT20 cells. In B lymphocytes, PI 3-kinase can be recruited to the plasma membrane by several adapter proteins including BCAP, Bam32, Gabl and CD 19 (reviewed by Leo and Schraven, 2001; Kurosaki, 2002). This localization is critical for PI 3-kinase pathway activation since its substrates are lipid located within the plasma membrane. Once at the plasma membrane, PI 3-kinase phosphorylates PI-4-P and PI4,5-P2 leading to the generation of PI-3,4-P2 and PI-3,4,5-P3, respectively.  These  phospholipid products play a role in the activation of several PH domain containingsignaling proteins including Akt, as demonstrated in this study. Plasma membrane localization is also important for the activation of the PLCy pathway in cells. Like PI 3kinase, the substrate for PLCy (PI-4,5-P2) is also located at the plasma membrane. PLCy hydrolyzes PI-4,5-P2 is to produce the lipid second messengers IP3 and DAG which are important for the generation of a calcium flux within the cell and for the activation of other signaling enzymes. PLCy recruitment to the plasma membrane is accomplished by the adapter protein BLNK and LAB in B lymphocytes (Fu et al., 1998; Wienands et al., 1998; Janssen et al., 2003). This localization also allows PLCy to become activated since its upstream activating kinase, Syk, is localized at BCR complexes within the plasma membrane. Studies have shown that a calcium flux within the cells cannot be generated if PLCy is not recruited to the plasma membrane. Additionally, a' plasma membrane-  targeted form of PLC-y can activate the calcium pathway in BLNK-deficient cells (Ishiai et al., 1999). Therefore, the localization of PLCy to the plasma membrane is important for both its activation and its ability to act on its substrate.  Adapter proteins are also required for MAP kinase pathway activation.  Two well-  characterized adapter proteins of the MAP kinase pathway are Grb2 and She. These adapter proteins can also form larger complexes with other adapter proteins including Gabl, as shown in the previous chapter. The main function of the Shc/Grb2 complex within the cells is to recruit the guanine nucleotide exchange factor SOS to the plasma membrane. Once at the plasma membrane, SOS exchanges the bound GDP on the Ras GTP-binding protein for GTP thus activating it.  This activated Ras subsequently  activates the MAP kinase pathway. The MAP kinase pathway also elegantly illustrates the use of scaffolding proteins to co-localize various signaling components. A single scaffolding protein can co-localize the enzymes of the MAP kinase cascade (i.e. MAP kinase and its upstream activating enzymes MAP kinase kinase and the MAP kinase kinase kinase) allowing for more efficient interaction and activation between the enzymes.  For example, the MEK partner 1 (MP1) scaffolding protein that  simultaneously interacts with the MAP kinase ERK and its upstream activating kinase MEK (Schaeffer et al., 1998). The association of these two enzymes on MP1 facilitates the specific activation of ERK by MEK. Scaffolding proteins have also been shown to be important for the activation of the mammalian SAP kinases, JNK1 and JNK2. The JNK kinases and their upstream activating kinases, HPK1 (a SAP kinase kinase kinase kinase), MKK7 (SAP kinase kinase kinase) and MLK (SAP kinase kinase) simultaneously interact with the JNK-interacting protein-1 and - 2 (JIP-1 and JIP-2) scaffold proteins (Dickens et al., 1997; Whitmarsh et al., 1998; Yasuda et al., 1999; Whitmarsh et al., 2001; reviewed by Whitmarsh and Davis, 1998; Burack and Shaw, 2000). The close proximity of these three proteins allows for the sequential activation of the downstream kinases, ultimately resulting in JNK activation. Interaction between JNK and the JIP scaffold proteins increases JNK activity presumably by allowing it to be in close proximity with its upstream activating kinases. In addition to co-localizing the kinases within one signaling pathwa