"Medicine, Faculty of"@en . "Medical Genetics, Department of"@en . "DSpace"@en . "UBCV"@en . "Tazmini, Ghazaleh"@en . "2008-12-09T16:08:32Z"@en . "2008"@en . "Doctor of Philosophy - PhD"@en . "University of British Columbia"@en . "RasGRP1 is an intracellular signaling protein expressed in lymphocytes that is responsible\nfor activating Ras GTPases. Positive regulation of RasGRP 1 requires translocation to\ncellular membranes where lipid-anchored Ras can be accessed. Plasma membrane\nlocalization of RasGRP 1 in response to antigen receptors requires both the Cl domain and\nthe plasma-membrane targeting (PT) domain. The Cl domain binds to diacylglycerol\n(DAG) at membranes. The PT domain binds its putative ligand at the plasma membrane and\nis negatively regulated by an adjacent suppressor of PT (SuPT) domain. RasGRP1 also\ncontains a pair of EF-hands, with Ca\u00C2\u00B2\u00E2\u0081\u00BA-binding capability, but with no known regulatory\nrole. In DT4O cells, RasGRP1 translocates to the plasma membrane and activates the Ras\nERK pathway in response to B cell receptor (BCR) signaling. By introducing point\nmutations in the Ca\u00C2\u00B2\u00E2\u0081\u00BA-binding loops of each of the EF-hands, I found that a potential Ca\u00C2\u00B2\u00E2\u0081\u00BA-\ninteraction loop in the first EF-hand is required for RasGRP1 translocation and the\nconsequential activation of the Ras-ERK pathway in response to BCR signaling. However,\nRasGRP1 translocation is not regulated by BCR-generated Ca\u00C2\u00B2\u00E2\u0081\u00BA flux. EF-hands were not\nrequired for Cl domain-mediated membrane localization, but were needed for PT-mediated\nplasma membrane targeting. EF-hands enhanced PT-domain mediated plasma membrane\nlocalization by repressing the SuPT domain. The REM and GEF domains, which co\nordinately bind to and catalyze guanine nucleotide exchange on Ras GTPases, needed to be\npresent and Ras-bound for this EF-hand mechanism to be effective. When not bound to Ras,\nthe REM-GEF domain complex suppressed both plasma membrane and endomembrane\ntargeting of RasGRP 1 by an EF-hand independent mechanism. Finally, membrane\nlocalization and activation of a naturally occurring splice variant of RasGRP 1, found overexpressed\nin systemic lupus erythematosus (SEE) patients, was examined. This splice\nvariant lacks exon 11, which encodes the segment of RasGRP1 between the GEF domain\nand the first EF-hand. Removal of exon 11 resulted in a defect in plasma membrane\nlocalization that was partially overridden by deletion of SuPT, while membrane localization\ncontrol via the REM-GEF complex was not affected. Therefore, exon 11 deletion via\nalternative splicing appears to functionally disable the first EF-hand of RasGRP1."@en . "https://circle.library.ubc.ca/rest/handle/2429/2861?expand=metadata"@en . "3100343 bytes"@en . "application/pdf"@en . "REGULATORY MECHANISMS OF THE EXCHANGE FACTOR RASGRP1 by GHAZALEH TAZMINI B.Sc., The University of British Columbia, 1997 M.H.A., The University of British Columbia, 1999 A THESIS SUBMITTED 1N PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Genetics) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) July 2008 \u00C2\u00A9 Ghazaleh Tazmini, 2008 ABSTRACT RasGRP1 is an intracellular signalling protein expressed in lymphocytes that is responsible for activating Ras GTPases. Positive regulation of RasGRP 1 requires translocation to cellular membranes where lipid-anchored Ras can be accessed. Plasma membrane localization of RasGRP 1 in response to antigen receptors requires both the Cl domain and the plasma-membrane targeting (PT) domain. The Cl domain binds to diacylglycerol (DAG) at membranes. The PT domain binds its putative ligand at the plasma membrane and is negatively regulated by an adjacent suppressor of PT (SuPT) domain. RasGRP1 also contains a pair of EF-hands, withCa2-binding capability, but with no known regulatory role. In DT4O cells, RasGRP1 translocates to the plasma membrane and activates the Ras ERK pathway in response to B cell receptor (BCR) signalling. By introducing point mutations in theCa2-binding loops of each of the EF-hands, I found that a potential Ca2- interaction loop in the first EF-hand is required for RasGRP1 translocation and the consequential activation of the Ras-ERK pathway in response to BCR signalling. However, RasGRP1 translocation is not regulated by BCR-generated Ca2 flux. EF-hands were not required for Cl domain-mediated membrane localization, but were needed for PT-mediated plasma membrane targeting. EF-hands enhanced PT-domain mediated plasma membrane localization by repressing the SuPT domain. The REM and GEF domains, which co ordinately bind to and catalyze guanine nucleotide exchange on Ras GTPases, needed to be present and Ras-bound for this EF-hand mechanism to be effective. When not bound to Ras, the REM-GEF domain complex suppressed both plasma membrane and endomembrane targeting of RasGRP 1 by an EF-hand independent mechanism. Finally, membrane localization and activation of a naturally occurring splice variant of RasGRP 1, found over- expressed in systemic lupus erythematosus (SEE) patients, was examined. This splice variant lacks exon 11, which encodes the segment of RasGRP 1 between the GEF domain and the first EF-hand. Removal of exon 11 resulted in a defect in plasma membrane localization that was partially overridden by deletion of SuPT, while membrane localization control via the REM-GEF complex was not affected. Therefore, exon 11 deletion via alternative splicing appears to functionally disable the first EF-hand of RasGRP 1. 11 TABLE OF CONTENTS ABSTRACT ii TABLE OF CONTENTS iii LIST OF FIGURES vi ACKNOWLEDGEMENTS xi 1. INTRODUCTION 1 1.1. Ras protein biochemistry and regulation 1 1.2. Ras effector pathways 2 1.3. Oncogenic Ras activation in cancer 4 1.4. Guanine nucleotide exchange factors (GEFs) for Ras GTPases 5 1.4.1 Ras activation via SOS proteins 5 1.4.2 Ras activation via RasGRF proteins 7 1.4.3 Ras activation via RasGRP proteins 8 1.5. Physiological relevance ofRasGRPl 10 1.5.1 Function of RasGRP1 in T cells 11 1.5.2 Function of RasGRP1 in B cells 15 1.6. Regulation by controlled membrane localization 18 1.6.1 RasGRP 1 localization in resting versus stimulated cells 19 1.6.2 The role of the Cl domain in promoting membrane localization 21 1.6.3 Evidence that the Cl domain is not sufficient for membrane localization 23 1.6.4 RasGRP 1 regulation via the SuPT/PT domain 24 1.7. Model for BCR-induced membrane translocation of RasGRP1 27 1.8. The EF-hand domain in RasGRP1 28 1.8.1 Ca2 and the EF-hand domain 29 1.8.2 Role of EF-hands in regulating RasGRP1 33 1.8.3 Role of EF-hands in regulating other RasGRPs 35 1.9. Research objectives and approach 36 2 MATERIALS AND METHODS 38 2.1. Cells and reagents 38 2.2. Constructs 39 111 2.3. Retroviral transduction of cell lines 40 2.4. Fluorescence microscopy 41 2.5. Stimulation and lysis of cells for detection of P-ERK2 or Ras-GTP 42 2.6. Western blot analysis 43 2.7. Calcium flux measurements 43 2.8. Transformation assays 44 3 ROLE OF EF-HANDS IN ACTIVATING RASGRP1 45 3.1. DT4O B cells as an experimental system for analyzing RasGRP 1 activation by antigen receptor 45 3.2. Translocation and activation ofRasGRPl is not regulated by BCR-induced calcium flux 48 3.3. The first EF-hand of RasGRP1 controls receptor-induced plasma membrane targeting 57 3.4. The first EF-hand enables plasma membrane translocation ofRasGRPl by counteracting the SuPT domain 63 4 CONTROL OF MEMBRANE LOCALIZATION OF RASGRP1 BY THE GEF AND REM DOMAINS 68 4.1. The REM-GEF domain complex also contributes to EF-hand-dependent regulation of plasma membrane targeting 68 4.2. Mutations that prevent occupation of the GEF catalytic site by Ras disable PT- mediated membrane localization, independently of the EF-hands 71 4.3. Suppression of membrane binding by the Ras-unoccupied GEF domain involves both the Cl and PT domains 76 4.4. Loss of membrane binding by GEF domain mutation does not reflect loss of signal transduction from Ras 78 4.5. Membrane localization by the Cl or PT domain is not essential for a low level of activity of the REM-GEF domain complex 81 4.6. A putative Ras binding site in the REM domain of RasGRP1 contributes to both membrane localization and activation 83 5 FUNCTIONAL CHARACTERIZATION OF A SPLICE VARIANT OF RASGRP1 FOUND IN PATIENTS WITH SYSTEMIC LUPUS ERYTHEMATOSUS (SLE) 87 5.1. A RasGRP 1 splice variant lacking exon 11 has a functional GEF domain but has a plasma membrane targeting defect that mimics EF-hand mutation 87 iv 5.2. A proline-rich segment in exon 11 is not important for BCR-induced plasma membrane targeting of RasGRP I 92 6 DISCUSSION 94 6.1. New insights into the mechanism of RasGRP 1 activation 94 2+6.2. The role of EF-hands and Ca flux in regulating RasGRP I 94 6.3. Mechanism of EF-hand dependent translocation 98 6.3.1 Involvement of the SuPT domain and REM-GEF domain complex 98 6.3.1 Involvement of EF-hands in determining the cellular site of RasGRP 1 activation 101 6.4. Membrane localization and dependence on GEF domain activity 103 6.5. Implications for the regulation of other RasGRPs by EF-hands and Ca2 109 BIBLIOGRAPHY 112 APPENDICES 123 Appendix A 124 v LIST OF FIGURES Figure 1.1 Examples of effector targets of Ras-GTPases 3 Figure 1.2 Domain structures of RasGRPs 9 Figure 1.3 T cell development overview 12 Figure 1.4 B cell development overview 15 Figure 1.5 Model for RasGRP1 regulation mediated solely by a Ci-DAG interaction 22 Figure 1.6 Mechanism for BCR-induced translocation of RasGRP 1 to the plasma membrane 28 Figure 1.7 Receptor regulation of calcium release from the ER and calcium entry across the plasma membrane 30 Figure 1.8 A single EF-hand (A) and a pair of EF-hands (B) 32 Figure 3.1 RasGRP 1 is activated by translocation to the plasma membrane in response to cdgM 47 Figure 3.2 BCR-induced Ca2 flux is not required for RasGRP1 translocation and activation 50 Figure 3.3 Increasing the concentration of BAPTA-AM to 30 jiM disrupts DAG generation at the plasma membrane and prevents PT domain-mediated translocation 52 Figure 3.4 Chelation of both intracellular and extracellular Ca2 does not inhibit plasma membrane targeting of RG1 54 Figure 3.5 BAPTA-AM does not advance the kinetics of RG1 translocation in response to BCR ligation 56 Figure 3.6 Sequence of the two EF-hands of RasGRP1 and location of point mutated residues 57 Figure 3.7 EF-hand mutations affect the plasma membrane targeting of RG1 in response to BCR ligation 59 Figure 3.8 Functional EF-hands are required for RGI to activate the ERK pathway in response to BCR ligation 61 Figure 3.9 Point mutation in the first EF-hand also disrupts plasma membrane localization in response to the G-protein-coupled receptor M5 62 Figure 3.10 EF-hand mutations do not affect Cl domain function of RG 1 64 vi Figure 3.11 PT-mediated, Cl-independent translocation of RG1 is affected by mutation of the first EF-hand 66 Figure 3.12 EF-hands allow plasma membrane targeting by mediating SuPT activity 67 Figure 4.1 Plasma membrane targeting of RG1 becomes dependent on intact EF-hands once the REM-GEF domain complex is present 70 Figure 4.2 Deletions in the REM or GEF domain eliminate BCR-induced plasma membrane translocation of RasGRP 1 72 Figure 4.3 Effect of mutations in the REM and GEF domains on ability of RG1 to transform NIH 3T3 cells 74 Figure 4.4 Mutations that prevent occupation of the GEF catalytic site by Ras disable membrane localization of RasGRP 1 75 Figure 4.5 Suppressive effects of GEF mutation occurred whether the SuPT domain was present or absent 77 Figure 4.6 Loss of plasma membrane binding by the GEF domain mutation does not relect loss of signal transduction from Ras 80 Figure 4.7 Membrane localization by the Cl or PT domains is not essential for activation of the REM-GEF domain complex 82 Figure 4.8 A putative Ras binding site in the REM domain ofRasGRPl positively regulates membrane localization 85 Figure 5.1 Exon 11 amino acid sequence and location in RasGRP1 88 Figure 5.2 Loss of exon 11 in RasGRP1 does not affect transformation or endomembrane localization in NIH 3T3 cells 89 Figure 5.3 Loss of exon 11 in RasGRP1 results in plasma membrane targeting and activation defects that mimic EF-hand mutation 91 Figure 5.4 BCR-induced plasma membrane targeting of RG1 is not affected by mutations in a proline-rich segment encoded in exon 11 93 Figure 6.1 EF-hands permit PT domain function by shutting down the combined suppressive effects of the REM-GEF domain complex and SuPT domain 101 Figure 6.2 Model of regulation of RasGRP 1 membrane localization and activation by the REM and GEF domains 108 vii LIST OF ABBREVIATIONS BAPTA-AM: bis-(o-aminophenoxy)ethane-N, N, N\u00E2\u0080\u0099,N \u00E2\u0080\u0098-tetra-acetic acid acetoxymethyl ester BCR: B cell receptor BLNK: B cell linker BSA: Bovine serum albumin Cl: Cl domain Ca2: calcium ion CCE: capacitative calcium entry DAG: diacylglycerol DH: Dbl homology DGK: DAG kinase DN: double negative DP: double positive EDTA: ethylene diamine tetraacetic acid EF: EF-hands EGFP: Enhanced green fluorescent protein EGTA: ethylene glycol tetraacetic acid ER: endoplasmic reticulum ERK: extracellular signal-regulated kinase GAP: GTPase activating protein GDP: guanine nucleotide diphosphate GEF: guanine nucleotide exchange factor GFP: green fluorescent protein GRP: guanine nucleotide releasing protein viii GST: glutathione-s-transferase GTP: guanine nucleotide triphosphate GTPases: guanosine triphosphatases Ig: Immunoglobulin 1gM: Immunoglobulin M (ji-heavy chain) cL-IgM: anti-Immunoglobulin M antibody IQ: ilimaquinone motif 1P3: inositol- 1 ,4,5-triphosphate IP3R: inositol triphosphate receptor jim: micro-meter (one millionth of a meter) MAPK: mitogen-activated protein kinase MEK: mitogen-activated protein kinase or ERK kinase PBS: phosphate buffered saline PDK1: phosphatidyl inositol-dependent kinase 1 PH: pleckstrin homology P13K: phosphatidylinositol 3 -kinase PIP2: phosphatidylinositol-4,5-bisphosphate PIP3: phosphatidylinositol (3,4,5) triphosphate PKB: protein kinase B PKC: protein kinase C PKCe-C lab: tandem Cl domains of PKCE PLC: phospholipase C pm: plasma membrane PMA: phorbol 12-myristate 13-acetate ix Pren: prenylation PT: plasma membrane targeter RAG: recombinase activating gene RasGRF: Ras guanine nucleotide releasing factor RasGRP: Ras guanine nucleotide releasing protein RBD: Ras binding domain REM: Ras exchange motif RG1: abbreviated version ofRasGRPl RTK: receptor tyrosine kinase SH2: Src-homology 2 domain SH3: Src-homology 3 domain SP: Single positive SLE: Systemic lupus erythematosus SOS: son of sevenless SOCC: store operated channel SuPT: suppressor of PT TBS: tris buffered saline TBST: tris buffered saline tween-20 TCR: T cell receptor x ACKNOWLEDGEMENTS It would not have been possible to complete this doctoral thesis without the help and support of the kind people around me. I am particularly indebted for the unwavering support I received from my supervisor, Dr. Robert Kay, at every stage of this Ph.D. project. For sharing his wealth of knowledge, experience, generosity, kindness and good humour at all times, I am extremely grateful. It has truly been a wonderful journey and I can honestly say that one simply could not wish for a more ideal supervisor. Thanks to all my lab colleagues, past and present. In particular, I would like to thank Nadine for initiating and inspiring the EF-hand project. Thanks to Ada, for her contributions to the lab work for this project. I am also grateful to Rebecca for creating a supportive environment at work. And thanks to Ban for her care and generous advice. I greatly appreciate the input I have received from my committee members, Dr. Linda Matsuuchi, Dr. Alice Mui and Dr. Pamela Hoodless. Thanks to the TFL FACS facility staff for all the cell sorting. I wish to also thank Vivian and Jens for sharing their expertise in setting up the calcium flux assays. Thanks to the Hoodless lab for usage of their microscope, and thanks to Pablo for his technical support with the microscope. I am appreciative of all the great people at the TFL for their friendship. Thanks to Arefeh for invaluable support and encouragement throughout the project. Thanks to my husband Kamran for his understanding, patience and endless support. My greatest indebtedness goes to my parents, Sousan and Mohammad for believing in me and for always offering their constant love and support. Thanks to my sister, Ghoncheh, to whom I always look up to, and to my brother, Kourosh, who is always a source of tremendous inspiration. xi 1. INTRODUCTION 1.1. Ras protein biochemistry and regulation H-, K- and N-Ras proteins are members of the Ras subfamily, which are a branch of a larger Ras superfamily of small guanosine triphosphatases (GTPases) (Repasky et al, 2004, Wennerberg et al, 2005). Small GTPases have molecular weights ranging from 20- 29KDa and are monomeric guanosine nucleotide-binding proteins (G proteins) (Wennerberg et al, 2005). In addition to H-, K- and N-Ras (herein referred to as Ras proteins), the Ras subfamily also includes Rap, R-Ras, Ral and Rheb proteins. Other branches of the Ras superfamily include the Rho, Rab, Ran and Raf subfamilies. Small GTPases are the central operators of a variety of critical cellular processes including cell growth, differentiation, apoptosis, actin organization, intracellular trafficking and vesicular transport (Wennerberg et al, 2005). A functional mechanism common to all small GTPases is the ability to cycle between an inactive GDP-bound and active GTP-bound form (Vetter & Wittinghofer, 2001). Two groups of proteins regulate GDP/GTP cycling. The first group are the guanine nucleotide exchange factors (GEF5) and they positively regulate Ras GTPases by promoting formation of the active GTP-bound form. Specifically, GEFs accelerate the dissociation of the nucleotide-Ras complex thereby allowing Ras to bind GTP, which is available at a higher concentration in the cell than GDP (Lenzen et al, 1998). The second group of proteins are the GTPase activating proteins (GAPs) and they promote the formation of inactive GDP bound form by accelerating the intrinsic GTP-hydrolysis activity of Ras (Bernards & Settleman, 2004). Conversion from GDP- to GTP-bound states causes a conformational 1 change in the switch I (aa 30-38) and switch II (aa 59-67) effector binding regions of Ras leading to activation of effector proteins (Wittinghofer & Nassar, 1996). One important feature of the Ras proteins, and other Ras subfamily members, is that they are covalently bound to cellular membranes through a series of post-translational modifications at their C-termini (Cox & Der, 2002, Fu & Casey, 1999). When localized to the plasma membrane, they can transmit signals from cell surface receptors to various cytoplasmic effector pathways (Fu & Casey, 1999). Receptor tyrosine kinases (RTK5) and other cell surface receptors activate Ras proteins through the recruitment of specific GEFs to the plasma membrane (Schiessinger, 2000). GEFs have unique mechanisms of localizing to the plasma membrane in response to receptors and are described in section 1.4. Signaling from Ras proteins does not exclusively originate from the plasma membrane though. Using a fluorescent probe for activated Ras, Chiu et al. showed that Ras activation can also take place at the endoplasmic reticulum (ER) and Golgi and that these compartments can initiate differential signalling cascades (Chiu et a!, 2002). 1.2. Ras effector pathways Ras effectors are proteins that contain a core Ras binding domain (RBD) with strong affinity for Ras-GTP. Ras effectors link activated Ras to numerous intracellular signalling networks. Three well-known effectors for Ras proteins include Raf kinase (Raf), the Ral specific GEF, Ra1GDS, and phosphatidylinositol 3-kinase (P13K) (Rajalingam et al, 2007) (Fig. 1.1). 2 \u00E2\u0080\u00981r Figure 1.1 Examples of effector targets of activated Ras GTPases. GDP, guanine nucleotide diphosphate; GTP, guanine nucleotide triphosphate. Effector targets of Ras include Raf, P13K and Ra1GDS. The components of the MAP kinase pathway are also shown to illustrate that activated Ras can lead to activation of ERK. In this thesis, ERK is used to assay for activated Ras. Raf functions in the linear Ras-Raf-MEK-ERK mitogen-activated protein kinase (MAPK) pathway. This pathway is the best characterized Ras signalling pathway due to its role in oncogenesis as well as in normal cell functions such as cell-cycle progression, proliferation, differentiation and apoptosis (Leicht et al, 2007). Specifically, Raf phosphorylates MEK-1/2 which in turn phosphorylates and activates ERK-1/2 (Schiessinger, 2000). Activated ERK travels to the nucleus where it regulates gene expression by phosphorylating and activating a variety of transcription factors, including the Ets family of transcription factors (Turjanski et a!, 2007). P13K is the second best 3 characterized Ras effector and is crucial for cell growth, survival, proliferation, differentiation and motility (Rajalingam et al, 2007). Many of these functions relate to the ability of P13K to activate AKT (also refeffed to as protein kinase B) (Rajalingam et al, 2007). Here, P13K activation releases phosphatidylinositol (3,4,5) triphosphate (PIP3), which recruits phosphatidyl inositol-dependent kinase 1 (PDK1) and AKT to the plasma membrane. PDK1 phosphorylates and activates AKT. As with the Raf-MEK-ERK pathway, the PI3KIAKT pathway has also been implicated in human cancers (Vivanco & Sawyers, 2002). Finally, members of the RaIGEF family are also effectors of Ras and they have their own known substrates, Ra1A and Ra1B (GTPases). While the mechanisms regulating Ra1GEFs are only beginning to be understood, specific isoforms of this protein have been found to be critical for anchorage-independent growth and survival of already transformed cells but not normal cells (Rajalingam et al, 2007). The Raf-MEK-ERK effector pathway has been regarded as the principal pathway leading to oncogenesis (Leicht et al, 2007), however, all three pathways are now emerging as contributors to oncogenic transformation downstream of Ras activation in murine and human cells (Hamad et a!, 2002). 1.3. Oncogenic Ras activation in cancer Activating point mutations in the three ras genes, K-ras, H-ras and N-ras, are frequently found in human cancers, with an average incidence of 30%. Mutations in codons 12, 13, or 61 generate chronically active Ras proteins, either by inactivating intrinsic GTPase activity or by making Ras insensitive to GAPs (Bos, 1989). Mutations in different ras genes are thought to preferentially give rise to distinct tumor types. For instance, K-ras is predominantly mutated in carcinomas of the lung, pancreas and colon, whereas N-ras 4 mutations are more commonly found in myeloid leukemia (Bos, 1989). Many animal model studies have been used to establish a causal link between deregulated Ras activity and cancer progression (Malumbres & Barbacid, 2003). Amplification of Ras signalling pathways by alternative mechanisms, such as constitutive activation of receptors, loss of GAPs, overexpression or activation of GEFs, or deregulation of components of effector pathways could also lead to oncogenesis (Repasky et al, 2004). As such, there is much interest in understanding the factors that control Ras activation. 1.4. Guanine nucleotide exchange factors (GEFs) for Ras GTPases As mentioned previously, GEFs are responsible for linking cell surface receptors to Ras protein activation. The first protein identified to have GEF function was CDC25 in yeast, S. cerevisiae (Broek et a!, 1987, Jones et al, 1991). Sequence similarity to the minimal 250 residue functional catalytic region for CDC25, referred to as the GEF domain, was then used to identifi numerous GEF proteins in other eukaryotic organisms (Quilliam et a!, 2002). Three main families of GEFs for Ras proteins have been identified: Son of Sevenless (SOS) proteins (SOS 1 and SOS2), Ras guanine nucleotide releasing factors (Ra5GRF1 and - 2), and Ras guanine nucleotide releasing protein (Ra5GRP) (RasGRP1 to 4). 1.4.1 Ras activation via SOS proteins SOS was first identified in D. melanogaster and shortly after, homologs SOS1 and SOS2 were discovered in mammals (Bonfini et al, 1992, Simon et al, 1991). The crystal structure of the catalytic domain of SOS complexed with nucleotide-free Ras was resolved and showed that SOS causes a structural distortion in the Switch 1 and Switch 2 regions of Ras, favouring nucleotide release (Boriack-Sjodin et al, 1998). Subsequent mutational 5 studies further showed that SOS interaction with the Switch 2 region anchors Ras to SOS, whereas SOS interaction with Switch 1 disrupts the nucleotide-binding site and promotes GDP-dissociation (Hall et al, 2001). In the GEF CDC25, a minimal 250 residue catalytic region was sufficient for GEF activity in vitro; however, a larger 450 residue region was required for activity in vivo. This additional sequence, situated N-terminal to the catalytic domain, harboured a -50 residue domain termed the Ras exchange motif (REM) (Lai et al, 1993). The REM domain is found present in most GEFs and, based on the crystal structure of SOS, is required for stability as it binds to a hairpin helical ioop in the core catalytic domain of Ras resulting in a more open GTP-binding pocket (Boriack-Sjodin et al, 1998). Further analyses of the crystal structure of SOS revealed that the REM motif serves as a separate allosteric Ras-binding site. That is, while the GEF domain binds a nucleotide-free Ras, the REM domain binds GTP-bound Ras and this binding triggers a conformational change that allosterically increases exchange activity by the GEF domain (Freedman et al, 2006, Margarit et al, 2003). Based on sequence conservation, the same mechanisms identified in SOS could be shared by other GEFs that contain GEF and REM domains (Quilliam et al, 2002). While Ras GEFs share catalytic activity for Ras family members, their ability to be differentially regulated is attributed to the presence of different flanking domains. For SOS, additional domains include a Dbl homology (DH) domain, a pleckstrin homology (PH) domain and a C-terminal proline-rich region (Quilliam et al, 2002). The mechanism of SOS activation by engagement of RTKs is well characterized. Upon RTK activation, SOS becomes recruited to the plasma membrane by forming a complex with the adaptor protein Grb2. The SOS/Grb2 complex translocates to the plasma membrane via the interaction of 6 the Grb2 Src homology (SH)-2 domain with specific phosphorylated tyrosines on receptors (Gale et al, 1993, Schiessinger, 2000). Recruitment of SOS via Grb2 to the plasma membrane places this GEF in close proximity to Ras, thereby allowing for Ras activation. SOS activation is primarily thought to be mediated by the proline-rich region, which binds to the SH3 domain of the adaptor protein Grb2. However, the regulation of SOS is more complex than the simple recruitment to the plasma membrane. Initial studies showed that deletion of the N-terminal region of SOS, including the DH and PH domains, increased SOS activity, suggesting that this region inhibited activity when present. By determining the crystal structure of the N-terminus of SOS, it was shown that the DH/PH unit acts to inhibit SOS by binding to the REM allosteric Ras-GTP binding site (Sondermann et al, 2004). Recently, it was shown that the PH domain is also involved in translocation by binding to phosphatidic acid (PA) that is generated at membranes by phospholipase D2 (PLD2)in response to growth factor receptors (Zhao et al, 2007). This binding is thought to overcome inhibition imposed by the DH/PH complex, thereby allowing the REM domain to bind Ras GTP (Quilliam, 2007). One further level of complexity is that SOS is also able to catalyses GDP/GTP exchange on Rho GTPases via the catalytic DH domain (Mitin et al, 2005). 1.4.2 Ras activation via RasGRF proteins Like SOS, RasGRF also contains GEF, REM, DH and PH domains and acts on both Ras and Rho GTPases. However, RasGRF also contains an extra PH domain, a coiled-coil motif and a calmodulin-binding ilimaquinone (IQ) motif (Mitin et al, 2005). RasGRF1 and RasGRF2 are the two mammalian isoforms that have been identified and are activated by calcium influx (Fan et al, 1998, Farnsworth et al, 1995). Ca2tdependent activation of 7 RasGRF 1 requires binding of calmodulin with the IQ-motif and this interaction co-operates with other N-terminal domains such as the PH domain to facilitate activation (Buchsbaum et al, 1996). RasGRF1 is constitutively localized at the plasma membrane through a mechanism involving the N-terminal PH domain. Therefore, activation by Ca2+/calmodulin, and not cellular-relocalization, seems to be a crucial step in activating this protein (Buchsbaum et al, 1996). RasGRF2 also binds calmodulin through its IQ motif, but the importance of this interaction and its dependence on Ca2 have not been resolved (de Hoog et al, 2000). In resting cells, RasGRF2 is cytosolic and translocates to the plasma membrane in response to rises in intracellular Ca2 (Fam et al, 1997). Furthermore, knock-out studies in mice show that RasGRF 1 and RasGRF2 couple N-methyl-D-aspartate glutamine ligand gated ion channel receptors to Ras and ERK signalling (Tian et al, 2004). 1.4.3 Ras activation via RasGRP proteins RasGRP proteins have also been identified as positive regulators of Ras GTPases. Four family members have been identified to date: RasGRP1, 2, 3 & 4 (Fig. 1.2). Expression is distinct and overlapping, but overall RasGRPs are expressed in a range of tissues including the brain, kidney and hematopoetic organs (Kawasaki et al, 1998, Reuther et al, 2002, Yamashita et al, 2000). Like SOS, the REM and GEF domains of RasGRPs contain conserved residues required to promote GTP-loading and activation of Ras family proteins (Quilliam et al, 2002). RasGRP1 and 4 activate Ras, RasGRP2 activates Rap (Ras-related GTPase) and RasGRP3 activates both Ras and Rap (Clyde-Smith et al, 2000, Ebinu et al, 1998, Kawasaki et al, 1998, Reuther et al, 2002). A predicted alternative splice form of human RasGRP2 was cloned and shown to be N-myristoylated and palmitoylated and 8 activated Ras when expressed ectopically (Clyde-Smith et al, 2000). However, it has yet to be determined if this longer form of RasGRP2 is naturally occurring. REM GEF EF Cl SuPT PT K) \u00E2\u0080\u00A2 EH RasGRP1 + Ras REM GEE EF Cl RasGRP2 + Rap REM GEF EF Cl _1J \u00C3\u0098\u00E2\u0080\u0099) \u00E2\u0080\u00A2 RasGRP3 + Ras, Rap REM GEE EF? Cl \u00E2\u0080\u00A2 RasGRP4 + Ras Figure 1.2 Domain structures of RasGRPs. REM, Ras exchange motif; GEF, guanine nucleotide exchange factor catalytic domain; EF, EF hands; Cl, Cl-domain; SuPT, suppressor of PT; PT, plasma membrane targeter. Two structures that are present C-terminal to the REM and GEF domains that distinguish RasGRP family members from other RasGEFs include an EF-hand domain followed by a Cl domain. EF-hand domains typically come in pairs and are in many cases regulated by Ca2-binding (Yap et al, 1999). RasGRP1, 2 and 3 each contain a pair of EF-hands, whereas RasGRP4 contains a single EF-hand domain with significant sequence divergence from a typical EF-hand sequence (EF-hands are discussed in detail in section 1.8). The Cl domain (\u00E2\u0080\u0098-50 residues) is a cysteine-rich domain, first identified in protein kinase C (PKC), that can 9 bind the lipid second messenger diacylglycerol (DAG) when a conserved pattern of sequences is present (discussed in section 1.6.2) (Colon-Gonzalez & Kazanietz, 2006, Newton, 1995). DAG is generated at the membrane by a variety of receptors; therefore, binding of the Cl domain to DAG can serve as a mechanism for localizing RasGRPs to their substrates in response to receptor activation. RasGRP 1 and 3 contain Cl domains with affinity for DAG, whereas the Cl domain of RasGRP2 has no detectable affinity for DAG (Johnson et al, 2007). RasGRP4 has at least 2 splice variants, RasGRP4a and RasGRP4b, which are identical except for an extra five amino acids present in the Cl domain of RasGRP4b (Li et al, 2003, Yang et al, 2002). Of the two splice forms, only the Cl domain of RasGRP4a has affinity for DAG (Johnson et al, 2007). RasGRP1 contains two additional domains located at the C-terminus, SuPT and PT, which are important for regulation of RasGRP1 and will be discussed in section 1.6.4 (Beaulieu et al, 2007). An additional level of regulation is provided by PKC serine/threonine kinases, which phosphorylate Thrl 84 on RasGRP1 and Thr133 on RasGRP3 resulting in activation of these proteins (Zheng et a!, 2005). 1.5. Physiological relevance of RasGRP1 RasGRP 1, the first member of the RasGRP family to be discovered, was isolated from two independent screens for cDNAs (from rat brain and murine T cell-hybridoma cDNA libraries), that were capable of transforming fibroblasts (Ebinu et al, 1998, Tognon et al, 1998). RasGRP 1 was found to localize to membranes and activate Ras family members, including H-, K- and N-Ras. Expression analysis revealed that RasGRP1 was present in the brain, bone marrow as well as lymphoid tissues, including the thymus and spleen, and also 10 in T and B cell lines (Ebinu et al, 1998, Ebinu et al, 2000, Tognon et al, 1998). These studies provided the first clues of the importance of RasGRP 1 in the function of lymphocytes. 1.5.1 Function of RasGRP1 in T cells Ras signalling is crucial in the development of T cells. Pre-T cells, or thymocytes, undergo developmental maturation and immunological selection in the thymus prior to entering the periphery (Starr et a!, 2003). T cell development is crucial as it ensures the generation of mature T cells that express a diverse T-cell receptor (TCR) repertoire capable of reacting with a vast array of foreign peptides in the context of self-MHC. Thymocytes undergo several stages of differentiation marked by the expression of distinct cell surface markers including CD4 and CD8 (Starr et al, 2003) (Fig. 1.3). The earliest progenitors are called double negative (DN) as they lack expression of CD4 and CD8. a/fl-lineage T cells progress from DN to CD4CD8 double positive (DP) thymocytes following productive rearrangement of the TCR /3 loci. This requires complexing of TCR /3 with pre-TCRa and CD3 to form the pre-TCR. Assembly of the pre-TCR and passage to the DP stage is a critical developmental checkpoint and is termed fl-selection (Starr et a!, 2003, von Boehmer & Fehling, 1997). TCRa chain rearrangement takes place at the DP stage where cells express a functional afl-TCR at the cell surface and can be subjected to positive or negative selection. At the DP stage, interaction of the TCR with ligands on specialized thymic and bone marrow derived cells in the thymus determine the fate of the cell and are essential for maintaining self-tolerance. Self -tolerance is the ability of the cell to recognize and respond to foreign antigens and not to self-antigens. Signalling intensity through the TCR, which is the basis for the strength of signal hypothesis dictates the fate of the cell, with no signal 11 leading to death by neglect, a strong signal resulting in negative selection (deletion) or positive selection leading to maturation and differentiation to either MHC class I restricted CD4CD8 or MHC class II restricted CD4CD8 single positive (SP) T cells (Hogquist, 2001). Double negative (DN) Double positive (DP) Single positive (SP) I I I II \u00E2\u0080\u00A2\u00C2\u00AE Lymphoid progenitor Pre-TCR driven TCR driven positive and _____ 3-selection negative selection TCR gene TCRo gene rearrangement RAG\u00E2\u0080\u0099 Mice rearrangement Figure 1.3 T cell development overview. DN, double-negative; DP, double positive; SP, single positive; RAG, recombination-activating gene; TCR, T-cell receptor. It was shown that the Ras-Raf-MEK-ERK pathway is important for positive selection of MHC restricted T cells in the thymus as mice transgenic for dominant-negative forms of Ras, Raf-1 and MEK-1 have a block in positive selection at the DP stage, but normal negative selection (Alberola-Ila et al, 1995, O\u00E2\u0080\u0099Shea et a!, 1996, Swan et al, 1995). Originally, Sos was identified as the principal GEF responsible for activating Ras in 12 lymphocytes in response to TCR, which is a RTK. However, this did not explain why Ras activation was also stimulated by phospholipase C (PLC) (Nel et al, 1995). PLC enzymes are activated in response to TCR signalling, leading to activation of the Ras-ERK pathway (Noh et al, 1995). Upon TCR signalling, PLC is recruited to the plasma membrane and hydrolyses phosphatidylinositol 4,5-bisphosphate (PIP2)to generate inositol- 1,4,5- triphosphate (1P3) and DAG. DAG is released at the plasma membrane and recruits Cl- domain containing proteins. Originally, PKC family members, containing Cl domains were thought to be primarily responsible for activation of Ras via PLC through a mechanism involving the down-regulation of Ras GAPs (Downward et al, 1990). RasGRP 1 became an attractive candidate linking PLC to Ras activation when it was found that RasGRP 1 became activated in response to TCR signalling in a PLC- and DAG-dependent fashion in the Jurkat T cell line (Ebinu et al, 2000). To determine if RasGRP 1 was important for the development of lymphocytes, a RasGRP1 knock-out mouse was generated (Dower et al, 2000). DP thymocytes were present in these mice indicating that 13-selection could occur. However, most thymocytes from these mice failed to undergo positive selection, as evidenced by the reduced CD4 and CD8 SP thymocyte populations, and were unable to efficiently activate Ras and ERK following TCR or DAG activation (Dower et a!, 2000, Priatel et a!, 2002). By contrast, mice engineered to over-express RasGRP1 had elevated numbers of CD8 T cells (Norment et al, 2003). Therefore, RasGRP 1 is important for TCR a/3-driven selection and maturation of thymocytes. 13 Ras-ERK signalling positively regulates fl-selection. The DN to DP transition is enhanced by expression of activated Ras and Raf and ERK activity is associated with pre TCR signalling (Crompton eta!, 1996, Iritani et al, 1999, Swat et al, 1996). RasGRP1 expression in DN thymocytes is up-regulated following pre-TCR ligation (Norment et al, 2003). Recombination-activating genes (RAG] and RAG2) are both important for initiating V(D)J recombination and thymocytes that lack either of these genes are arrested at the DN stage as they fail to generate the TCR- f3 component of the pre-TCR (Nagaoka et al, 2000). Transgenie over-expression of RasGRP 1 in RAG-deficient thymocytes was sufficient to drive f3-selection in the absence of pre-TCR signalling (Norment et al, 2003). Therefore, RasGRP 1 may contribute to the differentiation signal provided by pre-TCR signalling during 13-selection. Furthermore, deregulated over-expression of RasGRP 1 at the DN stage induces thymocyte transformation leading to pre-T-cell acute lymphoblastic leukemias (Klinger et al, 2005). This probably is due to continual Ras signalling overriding the 13- selection checkpoint in pre-T cell development. Another study showed that older RasGRP 1-deficient mice developed an autoimmune disorder with symptoms such as elevated autoantibodies, similar to the human disease systemic lupus erythematosus (SLE) and a more recent study showed that T cells from patients with SLE, have an increased prevalence of defective isoforms of RasGRP 1 (Layer et al, 2003, Yasuda et al, 2007). Taken together these results suggest that RasGRP1 may be required for preventing SLE-type autoimmune disorders. 14 1.5.2 Function of RasGRP1 in B cells B cell development takes place in the bone marrow and parallels T cell development in that it involves an ordered series of steps characterized by the rearrangement of immunoglobulin (Ig) heavy and light chain loci and expression of cell surface receptors (Fig. 1.4). Early pro-B cell progenitors undergo V-D-J rearrangements of their heavy chain genes resulting in expression of a t heavy chain. The pro-B cell progresses to the pre-B cell stage by pairing the t heavy chain with Igu and 1g13 chains and a surrogate light chain to form a pre-B cell receptor (pre-BCR). Signalling through the pre-BCR drives V-J rearrangement and expression of the light chain genes which replaces the surrogate light chain forming a complete BCR. These immature B cells then move to peripheral lymphoid tissues where they undergo further development to form mature B cells (Iritani et a!, 1997, Zhang et a!, 2004). Bone Marrow Periphery Pre BCR BCR BCR .-K-4O O Lymphoid progenitor Pro-B Pre-B Immature B Mature B T RAG Mice Figure 1.4 B cell development overview. BCR, B cell receptor; HC, heavy chain; LC, light chain; VDJ, variable diversity and joining immunoglobulin gene regions 15 As seen for T cells, RAG-deficient mice also do not generate mature B cells as they are unable to undergo recombination of antigen receptor genes. During B cell development, introduction of a constitutive active form ofRas into a RAG-null background causes progression of pro\u00E2\u0080\u0094B cells to pre\u00E2\u0080\u0094B cells, despite absence of a pre-BCR. Also, a second study demonstrated that expression of a dominant negative form of Ras in B lymphocyte progenitors resulted in a block in B cell development at the pro-B cell stage that was overcome by expression of an activated form of Raf (Iritani et al, 1997, Shaw et al, 1999). Taken together, these studies suggest that the Ras pathway is important for pre-BCR signalling leading to pre-B cell maturation. As with T cells, PLC and SOS also become activated in B cells in response to BCR and phorbol ester stimulation leading to Ras activation (DeFranco, 1997). With Grb2-SOS being activated independently of PLC, any PLC-mediated Ras activation was thought to be attributable to the activity of PKC\u00E2\u0080\u0099 s. However, treatment with PKC inhibitors did not substantially impact BCR-induced Ras activation, suggesting that an alternative signalling molecule could link PLC to Ras (DeFranco, 1997, Harwood & Cambier, 1993). RasGRP1, also found to be expressed in B cells, was speculated to be the missing link between PLC and Ras. Therefore, it was a surprise when RasGRP 1 knockout mice exhibited normal B cell development (Dower et al, 2000). This indicated that another exchange factor could be responsible for Ras activation during B cell development, such as SOS or another member of the RasGRP family. 16 With RasGRP3 also found to be expressed in B cells, another group investigated the extent at which SOS, RasGRP1 and RasGRP3 contribute to Ras activation in the DT4O B cell line by generating knockout cell lines and examining Ras activation in response to BCR ligation (Oh-hora et al, 2003). Note that RasGRP2 and 4 are not expressed in DT4O cells, which is why these two proteins were not examined. SOS 1/SOS2 double deficient DT4O B cells had normal Ras activation in response to BCR, but had defective Ras activation in response to epidermal growth factor receptor (EGFR) stimulation. RasGRP3 was found to be expressed at a much higher level than RasGRP 1 in DT4O cells, therefore, the authors decided to knockout RasGRP3 with the presumption that a RasGRP3 knockout would have a stronger phenotype than a RasGRP1 knockout. RasGRP3 deficient DT4O cells had a marked reduction in BCR-induced Ras activation and further knockout of RasGRP 1 in these cells resulted in elimination of BCR induced Ras activation. EGFR-induced Ras activation was relatively normal in RasGRP3 or RasGRP3/RasGRP 1- deficient DT4O cells (Oh-hora et al, 2003). Therefore, the authors concluded that in DT4O cells, RasGRP3 is the principal Ras activator in response to BCR simulation, whereas SOS is the main Ras activator in response to EGFR stimulation. Could normal Ras activation seen in RasGRP1 knockout mice be attributable to redundancy by RasGRP3? Surprisingly, RasGRP3 single- and RasGRP1/3 double- knockout mice were generated and found to also have normal B cell development (Coughlin et al, 2005). Since Ras activation is important for B cell development, it could be that SOS and PKC proteins are compensating for the lack of RasGRP proteins in mutant mice or that they are the principal Ras activators in immature B cells. In contrast to immature B cells, splenic B cells from RasGRP3/RasGRP1 double deficient mice did have a phenotype in that they were unable to activate Ras-ERK in response to the phorbol ester 17 PMA (DAG analog). Single RasGRP3 or RasGRP1 deficient mutants had normal Ras-ERK activation. In addition, both RasGRP1 and RasGRP3 were found to be required for normal B cell proliferation responses induced by B cell ligation with and without IL-4. Therefore, in B cells, RasGRP1 and RasGRP3 appear to work in conjunction to activate Ras downstream of BCR in some circumstances, while in other circumstances Ras activation downstream of BCR may be mediated by different GEFs (Coughlin et al, 2005). RasGRP4 may be a potential candidate for contributing to Ras activation in B lymphocytes as expression of this protein has been reported in peripheral blood leukocytes (Reuther et al, 2002). RasGRP2 is a Rap activator and is therefore not predicted to affect the Ras activation directly (Kawasaki et al, 1998). 1.6. Regulation by controlled membrane localization Deregulated RasGRP 1 expression has been associated with the initiation of leukemia and lymphomas and expression of mutated or aberrantly spliced RasGRP1 proteins has been linked to the autoimmune disease SLE in mice and humans (Klinger et al, 2005, Layer et al, 2003, Li et al, 1999, Yasuda et al, 2007). With aberrant expression being correlated with disease, it is no surprise that much research has focused on understanding the complexity of RasGRP1 regulation. Positive regulation or activation ofRasGRPl is dependent on translocation to membranes where activation of membrane-bound Ras can take place (Ebinu et al, 2000, Tognon et al, 1998). Therefore, understanding the components involved in determining the localization status of RasGRP 1 is crucial to understanding RasGRP 1 regulation. 18 1.6.1 RasGRP1 localization in resting versus stimulated cells In resting cells, RasGRP1 can be found localized in the cytosol, endoplasmic reticulum (ER) and Golgi. In NIH 3T3 fibroblasts, RasGRP1 was found localized at internal membranes, primarily the ER and Golgi (Tognon et al, 1998). A more elaborate study using co-staining markers confirmed that in resting COS 1 cells, RasGRP 1 was localized at internal membranes and that localization correlated with markers for ER and Golgi (Caloca et a!, 2003a). In the same cells, it was shown, using a photobleaching technique, that RasGRP1 proteins are freely diffusing in internal membranes rather than being stably bound, signifying that the RasGRP 1 -membrane interaction is readily reversible (Caloca et al, 2003a). In murine T cells and in the Jurkat T cell line, RasGRP1 is localized predominantly in the cytosol and may be partly concentrated in internal membranes (Bivona et a!, 2003, Mor et al, 2007, Perez de Castro et al, 2004, Sanjuan et al, 2003). In DT4O B cells, RasGRP 1 was also found localized in the cytosol and internal membranes and co-stains with ER and Golgi markers (Beaulieu et al, 2007, Caloca et a!, 2003b). Upon stimulation with appropriate receptors, RasGRP 1 relocalizes from the cytoplasm to distinct membrane compartments and this step is required for substrate activation. As previously mentioned, RasGRP 1 contains a DAG responsive Cl domain. In NIH-3T3 cells, treatment with the phorbol-ester PMA (DAG analog) causes redistribution of RasGRP 1 from the internal membrane compartment to the plasma membrane (Tognon et al, 1998). Translocation is essential and sufficient for RasGRP1 activation (Ebinu et al, 1998, Tognon et a!, 1998). In murine T cells and the Jurkat T cell line, TCR stimulation causes RasGRP 1 to increase its association with membranes, as determined by cell fractionation 19 experiments, and this correlates with activation of the Ras-ERK pathway (Ebinu et al, 2000, Sanjuan et al, 2003). In Jurkat cells, the reports are conflicting as to which membrane compartment RasGRP 1 translocates to upon TCR stimulation. Some studies report that RasGRP 1 translocates to the plasma membrane compartment whereas others report translocation to the Golgi (Bivona et al, 2003, Caloca et al, 2003 a, Carrasco & Merida, 2004, Mor et al, 2007, Perez de Castro et al, 2004, Sanjuan et al, 2003, Zugaza et al, 2004). In a recent study, RasGRP1 was reported to localize to both the plasma membrane and Golgi compartments in Jurkat cells with TCR stimulation, with plasma-membrane targeting being dependent on co-stimulation with lymphocyte function-associated antigen-i (LFA- 1) (Mor et al, 2007). It has also been found that RasGRP1 localization can be dictated by the state of the cell. Ligation of the ICR in the absence of CD28 induces primary I cells to become anergic and un-responsive to subsequent activation by TCR plus CD28 (Schwartz, 2003). In primary T cells, RasGRP1 translocates to the plasma membrane when conjugated with antigen-presenting cells (APC5), which provide TCR and CD28 ligation (Zha et al, 2006). When the same primary I cells are induced into a state of anergy, RasGRP 1 no longer translocates to the plasma membrane and remains at endomembranes (Zha et al, 2006). In DT4O and WEHI 231 B cells, BCR stimulation results in RasGRP 1 translocation to the plasma membrane and this is correlated with activation of the Ras-ERK pathway (Beaulieu et al, 2007, Caloca et al, 2003b, Guilbault & Kay, 2004). The ability to translocate to different membrane compartments is significant in that it potentially diversifies the signalling output from RasGRP1, possibly leading to differential cell fate decisions. In fact, RasGRP1 was found at the plasma membrane in thymocytes undergoing negative selection and at the Golgi in thymocytes undergoing positive selection (Daniels et al, 2006). In the 20 WEHI 231 B cell line, overexpressed RasGRP 1, which is found to translocate to the plasma membrane with BCR stimulation, increases BCR-induced cell apoptosis (Beaulieu et al, 2007, Guilbault & Kay, 2004). Although not investigated, it would be interesting if RasGRP 1 localization to internal membranes in B cells would confer alternative signals to those generated from the plasma membrane. 1.6.2 The role of the Cl domain in promoting membrane localization The Cl domain was first isolated in PKC family members and found to be required for phorbol esters and DAG binding (Burns & Bell, 1991, Ono et al, 1989, Quest et al, 1994). The Cl domain is a 50 amino acid sequence motif that has a conserved pattern of cysteine and histidine residues (HX1112C2CXII4H67H,where H is histidine, C is cysteine, and X is any other amino acid), which form a coordination site for two Zn2 ions, each supporting the formation of a hydrophobic core allowing binding to phorbol ester/DAG and insertion into membranes (Hubbard et a!, 1991, Zhang et al, 1995). Cl domains that follow the conserved sequence pattern are termed \u00E2\u0080\u0098typical\u00E2\u0080\u0099 and are predicted to be regulated by DAG, whereas \u00E2\u0080\u0098atypical\u00E2\u0080\u0099 Cl domains diverge from the conserved pattern and are predicted not to bind or be regulated by DAG (Hurley et at, 1997). RasGRP1 contains a typical Cl domain with the conserved sequence pattern (Ebinu et a!, 1998, Tognon et al, 1998). Original studies on RasGRP1 established the requirement of the Cl domain in membrane translocation and subsequent activation of the Ras-ERK pathway (Ebinu et a!, 1998, Tognon et al, 1998). Treatment with PMA caused RasGRP1 to associate with membranes and result in elevated Ras-GTP levels in cells expressing RasGRP1 but only when the Cl domain was present (Ebinu et al, 1998, Tognon et al, 1998). 21 Furthermore, PMA-induced transformation of NIH 3T3 fibroblasts was dependent on the presence of the Cl domain (Tognon et al, 1998). In COS1 cells, deletion of the Cl domain abolished localization of RasGRP1 in the ER and Golgi (Caloca et al, 2003a). Receptor Figure 1.5 Model for RasGRP1 regulation mediated solely by a Ci-DAG interaction. PLC, phospholipase C; PIP2, phosphatidylinositol-4,5-bisphosphate; DAG, diacyiglycerol; GEF, guanine nucleotide exchange factor catalytic domain; Cl, Cl-domain; PA, phosphatidic acid; DGK, diacyiglycerol kinase Upon stimulation by receptors, PLC is recruited to the plasma membrane and hydrolyses PIP2 to generate second messengers 1P3 and membrane-bound DAG (Berridge & Irvine, 1984). DAG is released at membranes and recruits Cl domain-containing proteins. TCR and BCR activation is known to generate DAG by coupling to PLC enzymes, thus providing a mechanism for activation of RasGRP1 (Fig. 1.5). In Jurkat T cells, RasGRP1 activation downstream of the TCR was eliminated by treatment with a PLC-y1 inhibitor and PA 22 was dependent on the presence of the Cl domain (Ebinu et al, 2000, Roose et al, 2005). In both Jurkat T cells and DT4O B cells, Vav proteins, which are GEFs for the Rho/Rae family, have been implicated in regulating the Ras-ERK pathways by playing a role in PLC-y activation (Caloca et al, 2003b, Zugaza et al, 2004). In vav3 DT4O cells, RasGRP1 is unable to translocate to the plasma membrane in response to BCR, but translocation was possible upon co-expression with constitutively active Vav or Rae or treatment with PMA (Caloca et al, 2003b). Also, treatment with a PLC inhibitor in these cells eliminated Ras ERK signalling. Constitutively active Vav and Rae proteins also promote translocation of RasGRP 1 to the plasma membrane in Jurkats and positively regulate RasGRP 1-mediated Ras activation (Zugaza et a!, 2004). Furthermore, in WEHI 231 B cells, BCR-induced apoptosis in cells over-expressing RasGRP1 was dependent on the presence of the Cl domain (Guilbault & Kay, 2004). Finally, studies with diacylglycerol kinase (DGK\u00C3\u00A7), which metabolizes DAG to convert it to phosphatidic acid, have shown that the availability of DAG at membranes is important for RasGRP1 activation. When RasGRP1 was co-expressed with DGK in HEK293 cells, RasGRP 1-mediated Ras-GTP loading was downregulated. In Jurkats cells overexpression of a dominant negative form of DGK resulted in more prolonged Ras-GTP signalling, likely due to more sustained signalling by RasGRP1 (Jones et al, 2002, Topham & Prescott, 2001, Zha et a!, 2006, Zhong et al, 2002). 1.6.3 Evidence that the Cl domain is not sufficient for membrane localization The results described so far refer to membrane localization of full-length RasGRP 1. In agreement with results for the full-length protein, the isolated Cl domain of RasGRP 1 23 was found to bind phorbol esters with high affinity in vitro, and in Jurkat I cells, stimulation with PMA caused the isolated Cl domain to associate with the plasma membrane (Carrasco & Merida, 2004, Lorenzo et a!, 2000). However, whereas the full-length protein is capable of translocating to the plasma membrane in Jurkat T cells in response to TCR stimulation, under the same conditions, the isolated Cl domain does not translocate (Carrasco & Merida, 2004). This suggests that while the Cl domain is required, it is not sufficient to target the full-length protein to membranes in response to receptors and that additional signals may be required. Another consideration is that treatment with DAG and phorbol esters do not necessarily mimic signalling by receptors and results should be evaluated accordingly. For instance, like RasGRP 1, RasGRP4cL contains a typical Cl domain that has been shown to bind to DAG and translocate to membranes when treated with DAG and PMA, leading to activation of the Ras-ERK pathway (Johnson et a!, 2007, Katsoulotos et al, 2007). However, unlike RasGRP 1, RasGRP4cL is not responsive to TCR signals and to date has not been reported to be activated downstream of any other PLC-coupled receptor (Perez de Castro et al, 2004). This further suggests that RasGRP 1 must contain additional distinct elements that enable it to respond to receptors. The next section outlines recent results showing that in addition to the Cl domain, additional elements at the C-terminus of RasGRP 1 are found to be required for receptor-mediated translocation of RasGRP 1. 1.6.4 RasGRP1 regulation via the SuPT/PT domain Until very recently, the Cl domain was regarded as the principal vehicle driving RasGRP 1 to membranes. Recognizing that the Cl -DAG interaction could not entirely 24 account for RasGRP1 localization and activation in response to receptors, our group sought to challenge this prevailing view (Beaulieu et al, 2007). Historically, the DT4O chicken B cell line has been used to study gene function by targeted gene disruption because homologous recombination occurs at very high frequencies in these cells (Buerstedde & Takeda, 1991). As a result, this cell line is extremely useful since much is already known about the signaling components downstream of the BCR and many mutant cell lines are available for studying cell signaling pathways. In DT4O cells, BCR ligation induces the robust re-localization of GFP-tagged RasGRP 1 from the ER and Golgi to the plasma membrane (Beaulieu et al, 2007, Caloca et al, 2003b). Plasma membrane localization of RasGRP1 correlates with activation of H-, K- and N-Ras and phosphorylation of ERK2 (Beaulieu et al, 2007). The plasma membrane is also the site at which Ras activation takes place, as determined by an intracellular fluorescent probe for activated Ras, comprised of GFP fused to the Ras-GTP binding domain of Raf- 1, which is shown to translocate to the plasma membrane with BCR stimulation (Beaulieu et al, 2007). GFP-tagged RasGRP 1 was found to translocate in 100% of DT4O cells following ligation of BCR with anti-IgM. The same construct expressed in a PLC-y2 DT4O cell line was still found to translocate in 37% of cells (Beaulieu et al, 2007). This result constrasted data from the Vav knockout study showing RasGRP1 incapable of translocating to BCR (reviewed in section 1.6.2), and indicated that Vav and PLC deficiencies are not equivalent. Also, the isolated Cl domain did not translocate to the plasma membrane in response to BCR but did in response to PMA. The biggest surprise came when a Cl domain deleted RasGRP1 construct (RG1AC1) was found to translocate in 24% of cells in response to BCR 25 (Beaulieu et al, 2007). Taken together, these experiments demonstrate that while the Cl domain is required, it is not the only mechanism for BCR-induced plasma membrane localization and that a separate BCR-coupled PLC\u00E2\u0080\u0099y2-independent mechanism exists for plasma membrane targeting of RasGRP 1. By expressing progressive truncation mutants, our group found that when the region C-terminal to the Cl domain was deleted, BCR-induced membrane translocation was eliminated (Beaulieu et al, 2007). Analysis of the C-terminal region revealed two independent domains. A Plasma membrane Targeter (PT) domain was identified at the very C-terminus of RasGRP 1, which when expressed in isolation was present at the plasma membrane in 91% of resting cells and in 99% of BCR stimulated cells. Plasma membrane localization was not significantly perturbed when the PT domain was expressed in PLCy2 cells. Therefore, the PT domain binds to a putative ligand at the plasma membrane that is elevated in response to BCR, and likely independent of the PLC pathway. Our group is currently determining the identity of the PT ligand. The PT domain was also found to be important for RasGRP 1 translocation in WEHI 231 B cells and DO 11.10 T cells. When a region N-terminal to the PT domain, denoted Suppressor of PT (SuPT), was attached to the PT, membrane translocation is dampened down, suggesting that SuPT counteracts PT membrane targeting. In support of this finding, deletion of SuPT in the full-length protein enhanced membrane localization in resting and BCR stimulated cells (Beaulieu et al, 2007). Of note is that when the PT domain is deleted, Ras-ERK signalling is still possible in response to BCR, even though this mutant is present diffusely in internal membranes. 26 1.7. Model for BCR-induced membrane translocation of RasGRP1 Based on our experiments with DT4O cells, we propose the following model for RasGRP1 activation in response to BCR ligation (Beaulieu et al, 2007) (Fig. 1.6). In resting cells, RasGRP 1 is mostly present in the ER and Golgi through the interaction of its Cl domain with DAG in these compartments. A putative ligand (such as a lipid target) for the PT domain is present at the plasma membrane but levels are low and the Cl -DAG interaction along with the suppressive effects of the SuPT cause RasGRP 1 to favour internal membranes. BCR ligation causes an elevation in PT ligand and PLC-y2-generated DAG at the plasma membrane. PT binds to the high PT ligand pooi, effectively overriding suppression by the SuPT domain. Once localized to the plasma membrane via the PT domain, RasGRP 1-binding is further stabilized by the interaction of the Cl domain with DAG generated by PLC-y2. In this model, the PT domain is what specifies the plasma membrane as the preferred site of localization of RasGRP 1. 27 A Unstimulated B cell: RasGRP1 drawn to internal membranes B B cell stimulated via BCR: RasGRP1 drawn to plasma membrane Figure 1.6 Mechanism for BCR-induced translocation of RasGRP1 to the plasma membrane. BCR B cell receptor signalling complex including BCR, Igct, 1gJ3 and BLNK; Ag, antigen; PT, plasma membrane targeter; SuPT, suppressor of PT; GEF, guanine nucleotide exchange factor catalytic domain; Cl, Cl domain; PLOy2, phospholipase C gamma 2; DAG, diacyiglycerol. Unmodified figure from Beaulieu et al., 2007. Mol Biol Cell 18: 3156-3168 1.8. The EF-hand domain in RasGRP1 RasGRP 1 also contains a pair of EF-hands. Because EF-hands typically bind Ca2, this suggested that RasGRP 1 could be regulated by Ca2. Surprisingly however, there is very little known regarding the role of EF-hands and Ca2 in regulating RasGRP 1. I will first provide a review of EF-hands and how they are thought to function, and then discuss the current knowledge regarding regulation of RasGRP1 via EF-hands. PT ligand (low) PT ligand (high) ,\u00E2\u0080\u0094 DAGN Internal membranes 28 1.8.1 Ca2 and the EF-hand domain Ca2 is a universal intracellular signaling ion that is involved in an array of functions including cell-division, differentiation and apoptosis (Berridge et al, 2003, Gifford et al, 2007). In response to various stimuli, the concentration of Ca2 in the cytoplasm increases as a result of release of Ca2 from internal stores (ER/sarcoplasmic reticulum) and/or from an influx from extracellular space via Ca2 channels in the plasma membrane (Berridge et al, 2000, Berridge et al, 2003). One commonCa2-mobi1izing signal that is generated from stimulation of receptors (G-protein coupled receptors, receptor tyrosine kinases, B and T cell receptors) is the second messenger inositol-1,4,5-triphosphate (1P3), which is produced by hydrolysis of phosphatidylinositol-4,5-bisphosphate (PIP2)by PLC enzymes(Berridge et al, 2000). 1P3 diffuses into the cell and engages 1P3 receptors on the ER leading to release of Ca2 into the cell. Depletion of Ca2 in internal stores causes activation of plasma membrane Ca2channels, called store-operated Ca2channels (SOCC), in a process called capacitative calcium entry (CCE), which leads to Ca2 entry from the extracellular space into the cell (Fig. 1.7) (Cullen & Lockyer, 2002). Ca2 supplied by internal and external sources then becomes bound to a variety of calcium-binding proteins, including members of the EF-hand protein family. 29 Receptor Figure 1.7 Receptor regulation of calcium release from the ER and calcium entry across the plasma membrane. PLC, phospholipase C; PIP2, phosphatidylinositol-4,5-bisphosphate; DAG, diacyiglycerol; 1P3, inositol- I ,4,5-triphosphate; IP3R. inositol- 1 ,4,5-triphosphate receptor; ER, endoplasmic reticulum; SOCC, store-operated channel. The name EF-hand was first coined by Kretsinger and Nockolds over 30 years ago to describe a helix-loop-helix motif discovered in the structure of paravalbumin, a small Ca2- binding protein isolated from carp muscle (Kretsinger & Nockolds, 1973). EF-hands were then identified in many otherCa2-binding proteins, including troponin C, calmodulin, and myosin light chains. In fact, this structural motif has been found to be quite widespread and present in a large number of protein families (Grabarek, 2006). 30 Ca2 socc Ca2 Ca2 IP3R ER intracellular Ca2 stores from ER The EF-hand domain typically contains about 30 residues and is composed of a helix-loop-helix structure where two u-helices are linked by a flexibleCa2tbinding loop 2+(Gifford et a!, 2007) (Fig. 1.8 A). The EF-hand domain and in particular the Ca -binding loop typically contain negatively charged amino acids such as glutamic acid and aspartic acid that provide the ionic forces required for binding to the positively charged Ca2 ion (Gifford et al, 2007). The preferredCa2-binding geometry is seven ligands arranged in a pentagonal bipyramidal orientation. In canonical (most common) EF-hands, the Ca2t binding loop is typically 12 residues, of which three or four have negatively charged side chain carboxylic groups and one has a carbonyl group which together provide five coordinating ligands for Ca2. The other two ligands are supplied by a bidentate carboxylate ligand (supplies two oxygen atoms) of an acidic amino acid located in exiting helix (Gifford et a!, 2007). Non-canonical EF-hand loops deviate from the 12 residue conformity, but still achieve the pentagonal bipyramidal coordination by using a different set of ligands. 31 Figure has been removed due to copyright restrictions. Figure 1.8 A single EF-hand (A) and a pair of EF-hands (B). A) A single EF-hand consists of a helix-loop-helix structure with Ca2coordinated by residues in the loop. B) A J3-sheet is often formed between a pair of EF-hands (shown by the anti-parallel arrows). Figure from Gifford et al., Biochem J405: 199-22 1 EF-hands typically occur in pairs or multiple pairs (Fig. 1.8 B). This pairing is thought to stabilize the structure through the formation of an anti-parallel J3-sheet between the twoCa2tbinding loops (Grabarek, 2006). Pairing is seen even when only one of the EF 32 hands is functional or even when the anti-parallel f3-sheet between the two EF-hands is not present. EF-hands can occur in odd numbers and in paravalbumin, of the three EF-hands present, two bind Ca2+ and the third, unpaired EF-hand stabilizes the other two EF-hands (Gifford et a!, 2007). Paired EF-hands also allow cooperative ligand binding, such that Ca2- binding to the first site enhances the binding affinity of the second site. Most regulatory EF-hands are thought to change their conformation upon Ca2- binding thereby allowing interaction with target proteins (Grabarek, 2006). One established 2+ model proposes that Ca -binding causes a transition from a closed to an open conformation, exposing hydrophobic residues that bind to target proteins (Grabarek, 2006). EF-hand proteins that do not change their conformation uponCa2-binding are thought to work as Ca2-buffers. A new role forCa2-binding has also been proposed and suggests that in some cases, Ca2 bound to EF-hands is solely important for the structural stability of the protein (Gifford et al, 2007). However, there is still little evidence for strictly structural (rather than regulatory) roles for the EF-hands. 1.8.2 Role of EF-hands in regulating RasGRP1 RasGRP1 contains a pair of EF-hands that reside between the N-terminal GEF domain and the C-terminal Cl domain. Both EF-hands contain the required residues for coordinating Ca2. Several studies on RasGRP 1 have attempted to determine whether Ca2 signalling via the EF-hands regulated RasGRP1 but results are conflicting and inconclusive. The first paper that was published on RasGRP 1 tested the calcium-binding affinity of the EF-hands (Ebinu et al, 1998). Recombinant, GST-RasGRP 1 fusion proteins were expressed in bacteria, transferred to a nitrocellulose filter and probed with 45Ca. Wild-type 33 EF-hands were detected to bind Ca2, as did EF-hands with point mutations in the first EF hand. However, EF-hands with point mutations in the second EF-hand or both EF-hands were not detected to bind Ca2. Therefore, the second EF-hand is proposed to bind Ca2 whereas the first EF-hand does not. A subsequent study by our group determined that the presence of EF-hands in RasGRP 1 was not important for RasGRP 1-promoted transformation of NIH 3T3 fibroblasts. RasGRP1 with a deletion in the EF-hand region was able to transform NIH 3T3 cells in the presence of serum or PMA as effectively as wild-type RasGRP1 (Tognon et al, 1998). In line with these findings, another study showed that Ca2- binding to the EF-hands did not influence the ability of the Cl-domain to bind phorbol esters. Specifically, phorbol ester binding affinity of the Ci-EF-hand region was unperturbed in the presence of increasing doses of free Ca2 (Lorenzo et a!, 2000). As previously mentioned, 1P3 is a product of PLC activity. 1P3 is released in the cytosol and binds to 1P3 receptors causing release of intracellular Ca2 stores. With RasGRP 1 found to be responsive to TCR signaling in T cells through PLC-dependent mechanisms, the involvement of Ca2 and the EF-hands in regulating RasGRP 1 downstream of the TCR also came into question (Ebinu et a!, 2000). However, treatment with intracellular and extracellular calcium chelating agents reportedly (data not shown) did not affect the ability of RasGRP1 to activate Ras in Jurkat T cells. This result further ruled out a role for Ca2 and EF-hands in regulating RasGRP 1. In addition to the group showing that Ca2 binds to the second EF-hand, several other studies have also supported a role for the EF-hands (Ebinu et a!, 1998). Ca2 ionophores facilitate the transport of Ca2 across the plasma membrane. Ras activation in 293T cells expressing RasGRP1 was reported to be elevated with Ca2 ionophore treatment, however, 34 the increase appears marginal and statistics are not provided (Kawasaki et al, 1998). In the Jurkat T cell line, RasGRP1 was reported to enhance the synergistic induction of interleukin-2 (IL-2) expression by calcium ionophore in combination with phorbol-ester (Ebinu et al, 2000). It must be noted that the requirement for Ca2 ionophore in this case could reflect Ca2 acting on the IL-2 gene via the calcineurin!NFAT pathway, with phorbol ester mediated activation of Ras via RasGRP 1 synergizing with the calcineurin/NFAT pathway (Feske, 2007). The most direct evidence to date suggesting a role for the EF-hands in regulating RasGRP 1 was by our group showing that deletion of the EF-hands eliminated BCR- or PMA-induced apoptosis of WEHI 231 13 cells over-expressing RasGRP1 (Guilbault & Kay, 2004). In summary, EF-hands have been found to bind Ca2, but the physiological significance of the interaction has not been uncovered. 1.8.3 Role of EF-hands in regulating other RasGRPs The other members of the RasGRP family are similar to RasGRP 1 in their possession of EF-hands. RasGRP2 and 3 each contain a pair of EF-hands with predicted Ca2 binding capabilities, whereas RasGRP4 contains only one EF-hand and based on sequence inspection is not predicted to bind Ca2. Evidence for the role of Ca2 in regulating these proteins is limited and in some cases vague. Rap 1 A activation by two alternatively spliced isoforms of RasGRP2 was elevated with Ca2ionophore treatment in both 293T cells and COS cells, whereas Ras activation by both of these isoforms of RasGRP2 was inhibited by the same treatment in COS cells (Clyde-Smith et al, 2000, Kawasaki et al, 1998). RasGRP3-dependent activation of Ras was reportedly not augmented by Ca2 ionophore treatment in rat2 cells (Lorenzo et al, 2001). Another group concluded that RasGRP3- 35 dependent activation of Rap2B in HEK 293 cells stimulated with EGF receptor is mediated by Ca2,but this conclusion was based only on the observation that Rap2B activation requires Ca2 in these cell, without showing that this involves RasGRP3 (Stope et al, 2004). Finally, RasGRP4-dependent Ras-GTP loading was reported to be inhibited in the presence of 1mM Ca2 (Yang et a!, 2002). However, millimolar Ca2 is not physiological, and can induce changes in membrane structure that could artifactually inhibit exchange activity. In conclusion, much work still needs to be done to decisively determine the role of the EF hands and Ca2 \u00E2\u0080\u0098n regulating RasGRPs. 1.9. Research objectives and approach RasGRP 1 activation downstream of antigen receptors is a tightly regulated process and involves co-operation between multiple domains and signals generated by the cell. With sparse and conflicting knowledge available regarding the role of the EF-hands, we became interested in determining whether the EF-hands had any role in regulating RasGRP1. My overall thesis goal was to test the hypothesis that the EF-hands had a role in activating RasGRP 1 downstream of antigen receptors. The system I used for my studies was the DT4O B cell line (reviewed in 1.7). This cell line was an attractive model for my studies as other aspects of RasGRP1 regulation downstream of the BCR had been extensively studied in this cell line by our group, including the role of the Cl, SuPT and PT domains. My specific objectives were: 1) to determine whether Ca2 signals generated by BCR activation were important for RasGRP 1 translocation and activation of Ras-ERK, 2) to determine if Ca2- binding to the EF-hands was required for RasGRP1 activity and 3) to determine the mechanism with which EF-hands regulate RasGRP 1. Having addressed these initial research 36 objectives and encountering unexpected results, I also became interested in how the catalytic region of RasGRP 1, encompassing the REM and GEF domains, was involved in both EF hand dependent and EF-hand-independent mechanisms of RasGRP 1 regulation. 37 2 MATERIALS AND METHODS 2.1. Cells and reagents Wild-type DT4O (avian leukosis transformed chicken B cell line) cells were obtained from Mike Gold (University of British Columbia, Vancouver), and they were originally from T. Kurosaki (RIKEN Research Center, Yokohama, Japan). All DT4O cells used in this study were transfected with an expression plasmid expressing the ecotropic retroviral receptor, to make them permissible for infection with murine retroviral vectors. DT4O cells were cultured in RPMI 1640 medium (Stem Cell Technologies Inc., Vancouver, BC, Canada) supplemented with 10% fetal bovine serum (PAA Laboratories, Etobicoke, ON, Canada), 2% chicken serum (Invitrogen, Carlsbad, CA), and 50 jiM 2-mercaptoethanol. NIH 3T3 cells from American Type Culture Collection (Manassas, VA) were cultured in DMEM (Stem Cell Technologies) containing 10% bovine calf serum (Fisher Scientific, Pittsburgh, PA). Anti-chicken 1gM polyclonal antibody was from Bethyl Laboratories (Montgomery, TX). Anti-extracellular signal-regulated kinase (ERK) 1/2 anti-phospho-specific ERK1 /2 antibodies were from Cell Signaling Technology (Danvers, MA), anti-pan Ras was from EMD biosciences (San Diego, CA), anti-RasGRP1 (ml99) and anti-Rapl were from Santa Cruz Biotechnology, (Santa Cruz, CA). Horseradish peroxidase-conjugated secondary antibodies were from Jackson ImmunoResearch Laboratories. 1 ,2-Dioctanoyl-sn-glycerol (DAG analog), EGTA and carbachol were obtained from Sigma-Aldrich (Oakville, ON, Canada). BAPTA-AM and Fura-2/AM were obtained from Invitrogen (Molecular Probes). The MEK1 inhibitor U0126 was from Promega, Madison, WI). 38 2.2. Constructs All modifications to RasGRP1 sequence were made by PCR-directed site mutagenesis. The N-terminally GFP-tagged form of full-length murine RasGRP1 (RG1) was derived from the XFL construct described previously (Park et al, 1994) with the GFP coding sequences from pEGFP-C1 (Clontech, Mountain View, CA) fused to amino acid 2 of RasGRP1 (GenBank accession no. NP_035376). The sequence at the fusion site is DELYKSGLRSSAQSEGTLGKAR, with the sequences that do not naturally occur in enhanced (e)GFP or RasGRP 1 in italics. The following constructs are modifications of this RasGRP1 construct with the indicated changes (sequences that do not naturally occur in RasGRP 1 are italicized; C terminus indicated by star). RG 1 -iEF 1 contains mutations affecting the first EF-hand. The encoded peptide sequence of this mutant is VFKNYSLSQSGYISQE; the normal sequence is VFKNYDLDQDGYISQE. RG1-jiEF2 contains mutations affecting the second EF-hand. The encoded peptide sequence of this mutant is SFCVMSKSRSGLISRD; the normal sequence is SFCVMDKDREGLISRD. GEF1i1 is RG1 with a single R271E point mutation that disrupts binding to Ras GTPases (Park et al, 1994, Tognon et al, 1998). GEF-p2 is RG1 with a double (1354A+L355A) point mutation; the homologous mutation in SOS 1 destabilizes Ras binding in the catalytic pocket (Margarit et al, 2003). REM-p. is RG1 with a double (L123A+K124E) point mutation; the homologous mutation in SOS1 precludes Ras-GTP binding to the REM domain (Margarit et a!, 2003). RG1-Prop. contains a mutation affecting the proline cluster (Tognon et al, 1998). The encoded peptide sequence of this mutant is RNHRAQGLTGSKGGVVVDW; the normal sequence is RNHRAPPLTPSKPPVVVDW. RG1AC1 = deletion of amino acids 538\u00E2\u0080\u0094595; sequence around deletion = YSKLGKSPAIS. RG 1 ASuPT = deletion of amino 39 acids 646\u00E2\u0080\u0094694; sequence around deletion = VDHSEESTPRKSAQ. RG1 Aexi 1 = deletion of amino acids 44 1-476; sequence around deletion = NHRAPSVFKN. RG1AGEFC = deletion of amino acids 396-439; sequence around deletion = LNELVQTEAPPLTP. RG1/Pren and GEF-j2/pren = replacement ofRasGRPl amino acids 441 to C terminus with an HA epitope tag plus the prenylation signal of K-Ras; sequence from fusion to C terminus = EPRNHRSTEA YPYDYASGSRKHKEKMSKDGKKKKKKSKTKCVIM*. The sequence around the GFP fusion is DELYKSGLRSLKSTFPHNF for the Cl -SuPT-PT construct, DELYKSGLRSLKSTEGPPLTP for the EF 1/2-Cl -SuPT-PT construct, DELYKSGLRSLKSTPRKSAQ for the isolated PT domain construct, and DELYKSGLRSLKSTEDRVSLGHL for the RGlANterm construct. The sequence around the GFP fusion is DELYKSGLRSLKSTEDRVSLGHL for the isolated REM-GEF domain; the sequence at the C-terminus is A YPYDYASG*. The sequence around the GFP fusion and at the C-terminus for the tandem REM-GEF domains [(REM-GEF)x2] is the same as for the single REM-GEF domain; the sequence adjoining the two REM-GEF domains is YSKLGSTEAYPYDYASGSTAQSEGTLGKARE. The sequence for PKCc (Cla+Clb) constructs is as described previously (Beaulieu et a!, 2007). 2.3. Retroviral transduction of cell lines Transfection of BOSC23 ecotropic packaging cells with retroviral vector plasmid DNA was performed as described previously (Pear et al, 1993). Virus-containing medium was supplemented with polybrene to 20 tg/ml and then added to an equal volume of DT4O (2 x 106/ml) in appropriate medium. After 5\u00E2\u0080\u009410 h of culture, 2\u00E2\u0080\u00943 volumes ofmedium was added. Transduced cells were selected by addition of puromycin or hygromycin 30\u00E2\u0080\u009448 after 40 infection. GFP-positive cells were then sorted by flow cytometry. Adherent NIH 3T3 cells were transduced in the same way, except with complete changes ofmedium. 2.4. Fluorescence microscopy DT4O cells expressing different mutants of RasGRP 1 were plated on poly-L-lysine (Sigma Aldrich)\u00E2\u0080\u0094coated glass coverslips. Before stimulation, DT4O cells were cultured in serum- free medium for 3\u00E2\u0080\u00944 hours. DT4O cells were stimulated in Haiik\u00E2\u0080\u0099s buffer (Stem Cell Technologies) with 5 ig/ml anti-chicken immunoglobulin (Ig)M or 100 iM DAG. For BAPTA-AM experiments, cells were incubated in activation buffer (25 mM HEPES, pH 7.2, 125 mM NaC1, 5 mM KC1, 1 mM CaCl2, 1 mMNa2HPO4,0.5 mM Mg504,2 mM L glutamine, 1 mM sodium pyruvate, 0.1% glucose, 0.1% bovine serum albumin [BSA], and 50 iM 2-mercaptoethanol)(Saxton et al, 1994) with BAPTA-AM for 30 mm and then subsequently stimulated in activation buffer with 5 tg/ml anti-chicken 1gM. DT4O cells were then fixed with 4% formaldehyde in PBS. Fluorescence light microscopy was performed with a Zeiss Axioplan 2 imaging universal microscope using a 63x (plan APOCHROMAT) objective. Fluorescence light microscopy used a 450- to 485-nm excitation filter and a 500- to 545-emission filter. Images were taken at 1360x 1036 pixels with a Retiga EX camera (Q Imaging) and recorded with OpenLab (Improvision) imaging software. Images were used to score between 50-150 DT4O cells for plasma membrane localization of the GFP fusion proteins. Only DT4O cells showing fluorescence well above autofluorescence levels were scored. Images of individual DT4O cells displayed in the figures were chosen to be representative of the majority of the population of DT4O cells, except where noted in the figure legends. DT4O cells ranged in size from 8.5 to 13.2 im in 41 diameter. Images of DT4O cells were resized in Photoshop 5.0 (Adobe Systems) to have the same cell diameter, in order to fit them to a line segment of common length for generating fluorescence intensity histograms. Each histogram was generated by placing the line segment from the centre of the cell of interest, through and beyond the plasma membrane, as shown in the figures. Histograms offluorescence intensities along the line segment were drawn by the profile function of Scion Image software (version 4.0, Scion Corp. Frederick MD USA). After transfection with GFP-fusion constructs indicated in the figure legends, NIH 3T3 cells were fixed in 4% formaldehyde in PBS and then visualized by fluorescence microscopy. Images were taken at 1360x1036 pixels with a Retiga EX camera and images were captured using OpenLab software. 2.5. Stimulation and lysis of cells for detection of P-ERK2 or Ras-GTP DT4O cells (2.5 x 10) were incubated in activation buffer for 10 mm at 37\u00C2\u00B0C before stimulation with anti-chicken 1gM a concentration of 5 jig/ml for the indicated times. For BAPTA-AM treatments, cells were incubated in activation buffer loaded with BAPTA-AM at 23\u00C2\u00B0C for 30 mm prior to being washed in the same buffer and then stimulated at 37\u00C2\u00B0C with 5 ig/ml of anti-chicken 1gM. Cells were immediately lysed with 2.5 volumes of ice cold lysis buffer (25 mM HEPES, pH 7.5, 150 mMNaC1, 1% NP-40, 0.25% sodium deoxycholate, 10% glycerol, 25 mM NaF, 10 mM MgC12, 1 mM EDTA, and 1 mM NaMoO4)containing 2 ig/m1 aprotinin, 2 jig/mi leupeptin, 0.5 mM phenylmethylsulfonyl fluoride, and 1 mM activated Na3VO4.For Ras activation assays, lysates prepared as described above were incubated for 30 mm with glutathione S-transferase (GST)/Raf\u00E2\u0080\u0094RBD 42 fusion protein, which had been prepared and prebound to glutathione-agarose beads as described previously (Taylor et al, 2001). After washing, samples were eluted in Bio-Rad XT sample buffer (Bio-Rad, Hercules, CA), electrophoresed, and detected by western blot as described below, using anti-pan Ras antibodies. 2.6. Western blot analysis Samples containing equal quantities of total protein (measured by the BCA protein assay; Pierce Chemical, Rockford, IL) were denatured using Bio-Rad XT sample buffer, separated by gel electrophoresis on 12% XT-Criterion acrylamide gels (Bio-Rad), and transferred to polyvinylidene difluoride membranes (Millipore, Billerica, MA) by electroblotting. After blocking the membranes overnight at 4\u00C2\u00B0C in Tris-buffered saline/Tween 20 (TBST) (25 mM Tris-HC1, pH 7.4, 3 mM KC1, 150 mM NaCl, and 0.05% Tween 20) containing 5% BSA, primary antibodies were applied to the membrane for 90 mm at room temperature in 2% BSA TBST, and horseradish peroxidase-conjugated secondary antibodies were applied for 45 mm at room temperature in TBST containing 1% BSA. Membranes were exposed to substrate/enhanced chemiluminescence (Santa Cruz Biotechnology) and chemiluminescence was detected using the VersaDoc 5000 imaging system (Bio-Rad). P-ERK2 was quantified by band volume analysis using Quantity One software (Bio-Rad). In all western blot figures, each panel is from a single blot, with gaps in the image indicating intervening segments of the blot that are not displayed. 2.7. Calcium flux measurements DT4O cells (4 x 106 cells per ml) were incubated with 2 iM Fura-2/AM in activation buffer at 23\u00C2\u00B0C for 75 mm. For BAPTA-AM treatments, the cells were incubated with Fura-2-AM 43 for 45 mm at which point BAPTA-AM was added and incubated for another 30 mm. The cells were then washed and resuspended in 1 ml of activation buffer at 2 x 106 cells per ml in a stirring cuvette and cells were stimulated by addition of cL-IgM (5 jig/mi). Cytosolic calcium was measured by monitoring fluorescence intensity at 510 nm by exciting the sample with two different wavelengths (340 and 380 nm) with an Aminco-Bowman series 2 luminescence spectrometer (Thermo Electron Corporation, Madison, WI). 2.8. Transformation assays NIH 3T3 cells were seeded at iow density one day prior to being infected with retrovirus containing cDNAs indicated in the figure legends. The medium was changed 5-10 hr post infection and then subsequently changed every 1-2 days with DMEM (Stem Cell Technologies) containing 10% bovine calf serum. After 14 days (or unless otherwise noted), the cells were visualized by phase contrast microscopy using the 1 Ox objective of a Nikon Eclipse TS100 inverted microscope and photographed using a Retiga 1300 camera (Q imaging). Images were captured using the OpenLab imaging software. 44 3 ROLE OF EF-HANDS IN ACTIVATING RASGRP1 3.1. DT4O B cells as an experimental system for analyzing RasGRP1 activation by antigen receptor. Signal transduction and regulation of RasGRP1 downstream of the BCR has been extensively studied in the DT4O B cell line (Beaulieu et al, 2007, Caloca et al, 2003b). The DT4O chicken B cell line expresses a B cell receptor (1gM isotype) on its cell surface, which can be stimulated using an antibody raised against the variable chain of 1gM (Takata et al, 1995). 1gM cross-linking leads to a series of phosphorylation events resulting in activation of PLCy and the Ras pathway (Winding & Berchtold, 2001). DT4O cells can be induced to express murine, full-length or mutant, RasGRP 1 cDNAs fused to enhanced green fluorescent protein (GFP) using retroviral tranduction. For this purpose, pCTV2 11 or pCTV2 10 retroviral vectors were used, containing puromycin or hygromycin resistance genes, respectively (Beaulieu et a!, 2007, Tognon et a!, 1998). Retroviral vectors are then introduced to the ecotropic BOSC virus packaging cell line by transfection using a calcium phosphate precipitation protocol to produce ecotropic retroviruses ready for infection (Pear et a!, 1993). To permit infection by the ecotropic retroviruses, DT4O cells are previously transfected with an expression plasmid that expresses the ecotropic receptor (Beaulieu et al, 2007). Following infection and drug selection, DT4O cells are sorted for high GFP expression, compared to untransduced counterparts, using flow cytometry. RasGRP1 expression levels are also determined by western blot using an antibody that detects exogenous murine RasGRP 1. 45 The amount of activated Ras in the cell is determined by a Ras pull-down assay, which takes advantage of the fact that the Ras effector Raf only binds to the activated GTP bound form of Ras. In this assay, Ras-GTP is affinity purified from cell lysates using the Ras-binding domain (RBD) of Raf. Raf-RBD is in the form of a GST fusion protein which allows for extraction of the Raf-RBD/GTP-Ras complex with glutathione affinity beads. The amount of activated Ras is then determined by western blot using a Ras specific antibody. By comparing total lysates to affinity-purified lysates, the ratio of GTP-Ras to total Ras can be determined. Activated Ras triggers the sequential activation of the Raf-MEK1 kinase cascade leading to phosphorylation of ERK2, which is detected by western blot using an anti-phosphospecific-ERK (pERK) antibody. We use pERK as an assay for Ras pathway signaling because it is reliably quantitative. Upon stimulation with u-IgM, DT4O cells expressing RasGRP1 had elevated Ras GTP and pERK2 levels compared to GFP-expressing control cells (Fig. 3.1 C and (Beaulieu et al, 2007)). The GFP-tagged RasGRP1 translocates from the cytosol and endomembranes (golgi and ER) to the plasma membrane after BCR ligation (Beaulieu et al, 2007, Caloca et al, 2003b), as detected by fluorescence light microscopy (Fig. 3. 1B). Fluorescence light microscopy was chosen over confocal fluorescence microsopy as the preferred method to image cells. This is because for single cell preparations, where the depth of focus is small compared to e.g. whole tissue sections that are many layers of cells, there is no overlying material to cause light scattering and therefore the image is not degraded and can be easily visualized by fluorescence light microscopy. Also, the duration of exposure is minimized which then allows for multiple viewings as there is less sample bleaching. Furthermore, by doing confocal microscopy on a number of cell preparations, we have confirmed that we 46 have not lost any information by chosing fluorescence light microscopy to image our cells. stimuli and domains involved in RasGRP1 regulation. \u00E2\u0080\u0094 Transduced RGI ________ pan-Ras (GTP-Ioaded) pan-Ras (total) Figure 3.1 RasGRP1 is activated by translocation to the plasma membrane in response to aIgM. (A) Schematic representation of the domain structure of GFP-tagged murine RasGRP 1 (RG 1). GFP, green fluorescent protein; REM, Ras exchange motif; GEF, guanine nucleotide exchange factor domain; EF, EF-hands; Cl, Cl-domain; SuPT, suppressor of PT; PT, plasma membrane targeter. (B) DT4O cells expressing GFP-tagged RG1 were untreated (nil) or treated with 5 ig/ml of u-IgM for 15 mm. Cells were then fixed and photographed by fluorescence microscopy. Single cells shown are representative of the typical appearance of the population of cells. Histograms of fluorescence intensities along the indicated segments are shown to the top right of each cell image. The edge of Instead, we have been able to scan many more samples due to the high-throughput efficiency of this technique. Translocation of RasGRP 1 in DT4O cells has been previously shown to be correlated with the catalytic activation of RasGRP 1, resulting in an increase in the GTP loading of Ras and phosphorylation of ERK2 (Beaulieu et al, 2007). This translocation and activation phenotype of RasGRP 1 in DT4O cells serves as a model for studying the roles of specific RGI GFP RGI a-lgM a-lgM 0 10\u00E2\u0080\u0099 0 10\u00E2\u0080\u0099 A. B. GFP REM GEF EF Cl SuPT PT O- nil a\u00E2\u0080\u0094IgM O%pm H - 100% pm I. - \u00E2\u0080\u0094 P-ERK2 47 the cell is indicated by the vertical line on the segment. The percentages of cells in each population with detectable GFP at the plasma membrane are listed to the bottom right of each image as \u00E2\u0080\u009C%pm+\u00E2\u0080\u009D. (C) DT4O cells expressing either GFP as a control or RG1 were treated with 5 jig/mi of ci 1gM for 15 minutes. Expression levels of transduced RG1 protein for each sample were determined by western blot using an anti-RG1 antibody. GTP-bound Ras was purified by Ras pull-down assay and detected by western blot by using an anti-Ras antibody (pan-Ras) detecting endogenous K-Ras and H-Ras. Activated ERK2 was detected using an anti-phosphospecific ERK antibody (pERK). Anti-pan Ras is a measure of total-Ras and is also used as a loading control for this blot. Total ERK was not available for this data, however, as shown in figure 3.8 A and B, total ERK does not vary with treatment in RG 1 transduced DT4O cells. 3.2. Translocation and activation of RasGRP1 is not regulated by BCR-induced calcium flux. BCR ligation induces activation of PLCy2, which hydrolyses PIP2 to produce 1P3 which induces release of Ca2 into the cytosol from the endoplasmic reticulum, and 2+ . subsequent Ca entry across the plasma membrane (see chapter 1, Fig. 1.7). This increase in cytosolic Ca2 concentration following BCR ligation is rapid and transient and is referred to as Ca2 flux (Fig. 3.2 A). RasGRP1 contains a pair of EF-hands with potentialCa2-binding capabilities, and therefore may be regulated by Ca2 flux. To determine if BCR-induced Ca2 flux is required for RasGRP 1 translocation and activation, DT4O cells expressing GFP tagged RasGRP1 were treated with the cell-permeable Ca2 chelator BAPTA-AM. BAPTA AM enters cells and, following hydrolysis of an ester bond by cellular enzymes, is trapped predominantly in the cytosol, where it chelates free Ca2 (Tsien, 1981, Tsien et a!, 1984). Intracellular Ca2 levels were measured by loading cells with Fura-2/AM, a fluorescent indicator that binds to free Ca2 (Grynkiewicz et a!, 1985). Like BAPTA-AM, Fura-2/AM is also trapped predominantly in the cytosol by cleavage of an ester bond. Fura-2 is excited at two wavelengths (340 and 380 nm) and monitored at a single emission wavelength (510 48 nm). The Ca2 unbound form of Fura-2 is excited at 380 nm and theCa2-bound form is excited at 340 nm. The ratio of emissions at those two wavelengths directly correlates with the amount of intracellular Ca2. Fluorescence intensity increases at 340 nm with increasing Ca2concentration and decreases at 380 nm for the unbound form. Therefore, an increase in the ratio of emissions for the bound/unbound form correlates with increasing intracellular Ca2 concentration. At 10 riM, BAPTA-AM eliminated the Ca2 flux induced by BCR ligation, but surprisingly did not inhibit BCR-induced translocation of RasGRP 1 to the plasma membrane (Fig. 3.2 A+B). BAPTA-AM at 10 jiM also had no effect on RasGRP1 activation, as detected by ERK2 phosphorylation (Fig. 3.2 C). Therefore, Ca2-flux triggered by BCR ligation is not needed for either RasGRP 1 activation or its translocation to the plasma membrane. 49 A. E 2 o o co 2 2 o 0 C\u00E2\u0080\u0099) C\u00E2\u0080\u0099) .2 .2 BAPTA-AM 10pM 10pM a-IgM 0 15\u00E2\u0080\u0099 0 15\u00E2\u0080\u0099 0 15\u00E2\u0080\u0099 0 15\u00E2\u0080\u0099 pERK2 .. Rapi Transduced RG1 Figure 3.2 BCR-induced Ca2 flux is not required for RasGRP1 translocation and activation. (A) Intracellular Ca2measurements in DT4O cells in response to 5 tg/ml of (1-1gM. The left panel shows a Ca2flux response that is normally induced for DT4O cells treated with x-IgM. The right panel shows the absence of a Ca2flux response in DT4O cells pre- treated with 10 iM BAPTA-AM. The arrow indicates the time when ct-IgM was added. For Ca2 measurements, cells were loaded with Fura-2/AM for 45 mm prior to loading with BAPTA-AM for another 30 mm. The ratio of fluorescence intensity at 510 nm is displayed after exciting the sample at two different wavelengths (340 and 380 nm). (B) DT4O cells expressing RG1 were treated without (nil) or with 10 .tM 50 DMSO 2.0 1.8 1.6 1.4 1.2 A fl 100 200 \u00E2\u0080\u0098l\u00E2\u0080\u0099 seconds 300 01\ a-IgM 100 200 300 seconds a-IgM B. nil C. BAPTA-AM 10pM H7NflH 0% pm 0% pm 100% pm 100% pm GFP RG1 BAPTA-AM for 30 mill prior to being untreated (nil) or stimulated with 5 jig/ml of ct-IgM for 15 mm. Cells were fixed and photographed for fluorescence microscopy and analysed as described in Fig. 3.1. RG 1 localized to the plasma membrane as effectively in BAPTA-AM treated cells as untreated cells. (C) DT4O cells expressing either GFP as a control or RGI were either untreated (nil) or treated with 10 jiM BAPTA-AM for 30 mm prior to being to being untreated or treated with 5 jig/mi of cL-IgM for 15 mm. Activated ERK2 and RG1 levels were detected by western blot using c pERK and x-RG1 antibodies respectively. Anti-Rap 1 was used as a loading control in this experiment. Experiments shown are representative of 3 independent experiments. 2+.While BAPTA-AM chelates the high concentration of Ca induced by BCR ligation, it does not chelate all free Ca2 (Rintoul et al, 2001). Therefore, while the large increase in Ca2 induced by BCR ligation does not regulate RasGRP 1, it is possible that RasGRP 1 activation is dependent on lower, basal concentrations of Ca2. Raising the concentration of BAPTA-AM to 30 riM, which will further reduce the free Ca2 concentration, inhibited RasGRP1 translocation to the plasma membrane (Fig. 3.3 A). To test whether this effect of BAPTA-AM was acting via the EF-hands of RasGRP 1, we determined the effect of BAPTA-AM on the other two mechanisms known to contribute to RasGRP1 translocation: DAG-induced translocation via the Cl domain and PT domain-mediated translocation. We used a probe for DAG, the tandem Cia + Ci b domains of PKCE (PKC-C lab), to detect BCR-induced DAG generation at the plasma membrane and the isolated PT domain (PT) to assess PT-mediated translocation. PKC-C lab and PT both translocated to the plasma membrane in response to BCR-ligation with or without 10 jiM BAPTA-AM treatment. However, raising the concentration of BAPTA-AM to 30 jiM eliminated any detectable translocation ofPKCE-Clab or PT (Fig. 3.3 A+B). Therefore, BAPTA-AM at 30 jiM disables both Ci and PT-dependent mechanisms ofRasGRPi translocation. Because of this, the inhibitory effect of 30 jiM BAPTA on RasGRP 1 translocation cannot be ascribed to a direct effect on the EF-hand. 51 A. Nil BAPTA 10 iM 0% pm - 0% pm 0% pm _ _ H 100% pm\u00E2\u0080\u009D 100% pm 100% pm B. Nil BAPTA . BAPTA 10pM 3OpM 0% pm 0% pm 0% pm aigM H\u00E2\u0080\u0099J IH4J ____ 100% pm 100% pm 0% pm C. Nil BAPTA BAPTA 10pM 3OpM nij rNJ ___ I... 0%pm 0%pm a-IgM I Figure 3.3 Increasing the concentration of BAPTA-AM to 30 jiM disrupts BAG generation at the plasma membrane and prevents PT domain-mediated translocation. DT4O cells expressing RG1, GFP-tagged Clab of PKCe (PKCE -Clab), or GFP-tagged PT domain of RG 1 (PT) were untreated (nil) or treated with BAPTA-AM at the indicated concentrations. Cells 52 a\u00E2\u0080\u0094IgM H 0% pm 100% pm 100% pm were then either untreated (nil) or stimulated with 5 jig/ml of x-IgM for 15 minutes prior to being fixed. Cell were photographed by fluorescence microscopy and analysed as described in Fig. 3.1. (A) RG 1 was detected at the plasma membrane in DT40 cells treated with 10 tM BAPTA-AM, but no longer easily detectable at the plasma membrane when treated with 30 iM BAPTA-AM. (B) The PKCE-C lab domain was detected at the plasma membrane after treatment with 5 tg/ml of x-IgM for 15 minutes in DT4O cells that were untreated (nil) or treated with 10 jiM BAPTA-AM. PKCE-C lab was no longer detected at the plasma membrane when cells were treated with 30 jiM BAPTA-AM. (C) The PT domain trans located to the plasma membrane after treatment with 5 jig/ml of ci-TgM for 15 minutes in DT4O cells that were untreated (nil) or treated with 10 jiM BAPTA-AM. When the concentration of BAPTA-AM was raised to 30 jiM, PT was no longer detected at the plasma membrane. This experiment shown is representative of 2 independent experiments (for A, B and C). The BCR-induced Ca2 flux into the cytosol relies on both internal and external Ca2 sources. The initial transient increase in Ca2 following BCR-ligation reflects the release of Ca2 from internal stores, whereas a more sustained component results from influx of Ca2 from the extracellular space (Berridge et al, 2000, Berridge et al, 2003). Because chelation of Ca2 by BAPTA is limited by the on-rate of binding, it may not fully sequester Ca2 in the immediate vicinity of membrane channels (Cullen & Lockyer, 2002, Dolmetsch et al, 2001, Rintoul et al, 2001). If so, this may explain the lack of effect of 10 iM BAPTA-AM on RasGRP1 translocation. In other words, the rate at which Ca2 enters the cytosol through membrane channels may locally exceed the rate at which BAPTA can bind to Ca2, thereby providing a channel-proximal source of Ca2 for RasGRP 1 binding. In particular, local influx of Ca2 via plasma membrane channels might provide free Ca2 that could mediate plasma membrane binding of RasGRP 1, despite chelation of bulk cytosolic Ca2 by BAPTA. By stimulating cells in Ca2-free buffer that also contains the Ca2 chelator EGTA, BCR induced Ca2 flux via plasma membrane channels will be eliminated. As shown in Fig. 3.4, usingCa2-free/EGTA buffer, with or without BAPTA-AM, had no effect on BCR-induced 53 translocation of RasGRP1 (Fig. 3.4). This indicates that influx of Ca2 through plasma membrane channels is not required for RasGRP1 translocation to the plasma membrane. nil BAPTA-AM 10 pM nil a\u00E2\u0080\u0094lgM Nil a\u00E2\u0080\u0094lgM H _ H H 0% pm 0% pm 100% pm _ _ H 0% pm 0% pm 100% pm Figure 3.4 Chelation of both intracellular and extracellular Ca2 does not inhibit plasma membrane targeting of RG1. DT4O cells expressing RG1 were either untreated (nil) or treated with 10 1iM BAPTA-AM and stimulated with 5 jig/mI of x-IgM for 15 mm in eitherCa2-containing buffer (nil) or Ca2-free buffer plus EGTA. Addition of EGTA to Ca2-free buffer ensures chelation of any Ca2 from potential contaminating sources. Cell were photographed by fluorescence microscopy and analysed as described in Fig. 3.1. This experiment is representative of 2 independent experiments. The above experiments were designed to test the hypothesis that BCR-induced Ca2 flux positively regulates RasGRP1 translocation. However, RasGRP1 is not detected at the plasma membrane at 1 minute post BCR ligation, when the peak of Ca2 flux occurs. Instead, RasGRPI translocation is maximal between 5 and 15 minutes post BCR-ligation, at which time the intracellular Ca2 concentration has declined back to basal levels (Fig 3.5). This raises the possibility that a high Ca2 concentration is inhibitory rather than stimulatory to RasGRP1 plasma membrane targeting in DT4O cells, such that RasGRP1 can efficiently translocate only after the peak of the Ca2 flux has passed. In this case, eliminating the BCR 54 H 100% pm H 100% pm induced Ca2 flux with BAPTA-AM would be expected to accelerate the kinetics of translocation of RasGRP1 to the plasma membrane. However, BAPTA-AM had no detectable effect on the kinetics of RasGRP1 translocation in response to BCR ligation (Fig. 3.5), so the Ca2 flux that is rapidly induced by BCR ligation is not responsible for the delay in plasma membrane targeting of RasGRP 1. 55 a\u00E2\u0080\u0094lgM Nil 10 pM BAPTA-AM 0 H I 0% pm 0% pm 30\u00E2\u0080\u009D _ _ f 0% pm 0% 1H 0% pm 0% pm 3\u00E2\u0080\u0099 H H 29% pm 34% pm 5\u00E2\u0080\u0099 H H 88% pm 80% pm 15\u00E2\u0080\u0099 H H 100% pm Figure 3.5 BAPTA-AM does not advance the kinetics of RG1 translocation in response to BCR ligation. DT4O cells expressing GFP-tagged RasGRP 1 were unstimulated (nil) or treated with 10 jiM BAPTA-AM 30 mm prior to stimulation with 5 ug/ml of x-IgM for the indicated time points. Cells were then photographed by fluorescence microscopy and analyzed as described for Fig. 3.1. The single cells shown are representative of the typical appearance of the majority of cells, except for the 3 minute time points where cells representative of those with detectable plasma membrane localization are shown. This experiment is representative of 2 independent experiments. 100% pm 56 3.3. The first EF-hand of RasGRP1 controls receptor-induced plasma membrane targeting. The experiments so far indicate that RasGRP 1 is not positively or negatively regulated by BCR-induced Ca2 flux. Despite this, the conservation of EF-hands among RasGRPs implies that these domains are functionally important. Most notably, the amino acids at the 1, 3, 5, 7, 9 and 12 positions in the intra-helical loop of each EF-hand are perfectly configured for Ca2 binding (figure 3.6). To determine if the EF-hands made any contribution to RasGRP1 regulation in DT4O cells, we made mutations disabling both or either of the EF-hands. Using PCR-directed mutagenesis, three aspartate or glutamate residues were changed to serines in theCa2-binding loops of either the first, second or both EF-hands (Fig. 3.6). These point mutations were designed to disrupt Ca2 binding while maintaining the overall structure of the EF-hands. sss sss Iii Iii ISKYVQRMVDSVFKNYDLDQDGYI SQEEFEKIAASFPFSFCVMDKDREGLISRDEITAYFMRASS IYSK rieilx \u00E2\u0080\u0094.--- eiix eix \u00E2\u0080\u0094.-- .Loop Loop I EF-h 1 EF-h2 I Figure 3.6 Sequence of the two EF-hands of RasGRPI and location of point mutated residues. Sequence of the two EF-hands in RasGRP1, showing the predicted N- and C-terminal x-helices, and the intervening loops (for reference, see figure 1.8 showing helix-loop-helix for a single and pair of EF-hands). Residues potentially involved in calcium binding are bolded. The vertical arrows indicate the EF-hl and EF-h2 mutations that remove potential calcium-binding asparate and glutamate residues. The second EF-hand of RasGRP 1 had previously been described as being able to bind Ca2 in vitro (Ebinu et al, 1998). However, we expressed RasGRP1 with point mutations in the 57 second EF-hand (RG1 -pEF2) in DT4O cells and found that this mutant localized to the plasma membrane as effectively as wild-type RasGRP1 (Fig. 3.7). Instead, BCR-induced translocation of RasGRP 1 to the plasma membrane was greatly reduced by mutation of the first EF-hand (Fig. 3.7). While all cells expressing wild-type RasGRP1 showed strong translocation to the plasma membrane following BCR ligation, this occurred in only a small proportion of cells expressing RasGRP1 with mutated first EF-hand (8% in Fig. 3.7), and in those cells the mutant RasGRP1 was only weakly localized to the plasma membrane. Mutation of both EF-hands completely eliminated plasma membrane localization (Fig. 3.7). 58 GFP REM GEF EF Cl SuPT PT A. O_1Z-__C)-_-cz- RG1-pEF2 RG1-pEF1 RG1-pEF1/2 B. _ _ 0% pm I 100% pm RGI\u00E2\u0080\u00A2 IH 0% pm I 100% pm RG1pEF1 _____ _____ 0pm 8%pm _____________ 0 pm 0% pm Figure 3.7 EF-hand mutations affect the plasma membrane targeting of RG1 in response to BCR ligation. (A) Schematic representation of the domain structure of RG 1 and EF-hand mutants used in this study. RG1-1iEF2, RG1-1iEF1 and RG1-1iEF 2 contain point mutations in either the second, first or both EF-hands, respectively. (B) DT4O cells expressing the indicated proteins were untreated (nil) or treated with 5 jig/mi of x-IgM for 15 minutes. Cells were prepared for fluorescence microscopy, analyzed, and displayed as described for Fig. 3.1. The single cells shown are representative of the typical appearance of the majority of cells, except for the ct-IgM-stimulated RG1-jiEFl-expressing cell which is representative of those with detectable plasma membrane localization. This experiment is representative of 3 independent experiments. We hypothesized that defects in plasma membrane translocation of RasGRP 1 mutants would correlate with defects in activation, as measured by the ability of RasGRP 1 59 H to activate ERK2 (Fig. 3.1 C). Levels of phosphorylated ERK2 were compared for DT4O cells expressing equivalent amounts of wild-type or EF-hand mutated RasGRP 1 proteins (Fig. 3.8). Indeed, EF-hand mutations affected RasGRP1 activation in proportion to their effects on translocation; mutation of the first EF-hand reduced RasGRP 1-mediated ERK phosphorylation five to ten fold, mutation of both EF-hands eliminated all detectable activation, and mutation of the second EF-hand had no effect on activation (Fig. 3.8). Thus, the first EF-hand is essential for targeting RasGRP 1 to the plasma membrane in response to BCR ligation and is thereby critical for its activation at that site. But contrary to expectations, EF-hand-dependent translocation of RasGRP1 is not dependent on BCR induced Ca2 flux. 60 A. GFP RGI RGI-iJEFI a-IqM a-IgM a-IqM 0 1\u00E2\u0080\u0099 3\u00E2\u0080\u0099 5\u00E2\u0080\u0099 1O\u00E2\u0080\u009915\u00E2\u0080\u0099 0 1\u00E2\u0080\u0099 3\u00E2\u0080\u0099 5\u00E2\u0080\u0099 1O\u00E2\u0080\u009915\u00E2\u0080\u0099 0 1\u00E2\u0080\u0099 3\u00E2\u0080\u0099 5\u00E2\u0080\u0099 1O\u00E2\u0080\u009915\u00E2\u0080\u0099 \u00E2\u0080\u0094 \u00E2\u0080\u0094 \u00E2\u0080\u0094 1.0 1.0 1.2 3.5 4.6 1.8 1 2.8 13 21 35 34 1.0 1.0 2.8 3.1 9.4 5.2 Total-ERK2 __ Transduced RGI \u00E2\u0080\u0094 \u00E2\u0080\u0094 \u00E2\u0080\u0094 \u00E2\u0080\u0094 B. GFP RGI RGI-pEFI2 RGI-pEF2 a-IgM a-IgM a-IgM a-mM 0 1\u00E2\u0080\u0099 3\u00E2\u0080\u0099 5\u00E2\u0080\u0099 10\u00E2\u0080\u0099 15\u00E2\u0080\u0099 0 1\u00E2\u0080\u0099 3\u00E2\u0080\u0099 5\u00E2\u0080\u0099 10\u00E2\u0080\u0099 15\u00E2\u0080\u0099 0 1\u00E2\u0080\u0099 3\u00E2\u0080\u0099 5\u00E2\u0080\u0099 10\u00E2\u0080\u0099 15\u00E2\u0080\u0099 0 1\u00E2\u0080\u0099 3\u00E2\u0080\u0099 5\u00E2\u0080\u0099 10\u00E2\u0080\u0099 15\u00E2\u0080\u0099 \u00E2\u0080\u0094 \u00E2\u0080\u0094 \u00E2\u0080\u0094 - P-ERK2 1.0 1.0 1.4 3.4 5.4 2.7 1.0 3.3 8.9 13 22 21 1.0 1.0 2.6 4.4 5.6 5.6 1.0 1.8 8.7 9.8 17 20 Total-ERK2 \u00E2\u0080\u0094 \u00E2\u0080\u0094 _.\u00E2\u0080\u0094 \u00E2\u0080\u0094 \u00E2\u0080\u0094 \u00E2\u0080\u0094 \u00E2\u0080\u0094 \u00E2\u0080\u0094 \u00E2\u0080\u0094 Transduced RG1 Figure 3.8 Functional EF-hands are required for RG1 to activate the ERK pathway in response to BCR ligation. Refer to Fig. 3.7 for a schematic of constructs used in this study. DT4O cells transduced with the indicated proteins were treated with 5 .ig/ml of cc-IgM for the indicated time points. GFP and RG1 were used as negative and positive controls, respectively. Levels of pERX2 were determined by western blot using an anti-phospho-specific ERK antibody. Levels of p-ERK2 were quantified by densitometry, with fold changes indicated below each band. Total ERK2 was detected using an anti ERK antibody and shows that changes in pERK are not a reflection of changes in total ERK in cells. Transduced RG1 protein, was detected using an antibody that recognizes the C-terminal portion of the protein. These blots are representative of 3 independent experiments. Note that duplicate gels were loaded with equal quantities of pre-quantified sample (prepared from the same tube). One blot was used to probe PERK2 and RG 1 and the other blot was used to probe for total ERK2. (A) BCR induced increase in pERK levels as a result of RG 1 transduction was reduced by mutation in the first EF-hand (RG1 -miEF 1). (B) BCR-induced increase in pERK2 levels as a result of RG 1 transduction was not significantly affected by point mutation in the second EF-hand (RG1 -tEF2) but was eliminated by mutation in both EF-hands (RG 1 -iEF 12). 61 RasGRP 1 can also be activated by signaling from G-protein coupled receptors (Jones et a!, 2002). To directly compare the role of the first EF-hand in translocation of RasGRP1 induced by BCR versus a different class of receptor, we transduced DT4O cells with the M5 muscarinic acid receptor, a G protein-coupled receptor. Stimulation of these DT4O cells with carbachol, an agonist for the M5 receptor, induced strong plasma membrane translocation of wild-type RasGRP1 in about half of the cells (Fig. 3.9). Carbachol-induced translocation was eliminated by mutation of the first EF-hand. Therefore, the first EF-hand is critical for plasma membrane targeting of RasGRP 1 in response to two distinct classes of receptors in DT4O cells. This further confirms our observations regarding the first EF-hand. RGI RGI-pEFI nil _ IHJN 0% pm I 0% pm Carbachol [_J H ___ 60% pm 0% pm Figure 3.9 Point mutation in the first EF-hand also disrupts plasma membrane localization in response to the G-protein-coupled receptor M5. DT4O cells expressing the M5 muscarinic receptor were transduced with RG1 or RG1-jiEFl. Cells were untreated (nil) or treated with 100 jtM carbachol for 10 mm. Cells were prepared for fluorescence microscopy, analyzed, and displayed as described for Fig. 3.1. The single cells shown are representative of the typical appearance of the majority of cells. This experiment is representative of 2 independent experiments. These experiments demonstrate that the EF-hands participate in RasGRP1 activation, and that they do so by enabling plasma membrane targeting in response to BCR ligation. 62 However, the results of these experiments and the Ca2 flux-manipulating experiments do not fit with the previous thinking that the EF-hands, particularly the second EF-hand, regulate RasGRP 1 by acting as sensors of the Ca2 influx triggered by BCR ligation. While mutations that eliminate potential Ca2 binding by the EF-hands prevented translocation to the plasma membrane and RasGRP 1 activation, chelation of cytosolic and extracellular Ca2 had no effect on translocation. In conclusion, the EF-hands act as positive regulators of RasGRP 1, but they may not act as sensors of receptor-induced Ca2 flux. 3.4. The first EF-hand enables plasma membrane translocation of RasGRP1 by counteracting the SuPT domain. BCR-induced translocation of RasGRP 1 to the plasma membrane is mediated by cooperativity between the Cl and PT domains, with the PT domain being responsible for specifying the plasma membrane as the site of localization. The PT domain is negatively regulated by an adjacent SuPT domain. It is possible that EF-hands enable plasma membrane targeting by promoting the functions of the Cl and/or PT domains, or by counter acting the membrane targeting suppressive effects of the SuPT domain. We sequentially addressed the requirement of the EF-hands in the functions of each of these domains. BCR-induced plasma membrane targeting of RasGRP 1 is in part mediated by the Cl domain, which binds to plasma-membrane localized DAG (Beaulieu et a!, 2007). Treatment of DT4O cells over-expressing RasGRP 1 with exogenous DAG results in elevated ERK signaling (Fig. 3.10 A), presumably caused by RasGRP1 localizing to endomembranes via the Cl domain (Beaulieu, N., Rob Kay lab, unpublished data). If the EF-hands are required for Cl domain function, then EF-hand mutants should be unresponsive to exogenous DAG 63 treatment. However, EF-hand mutants were capable of activation in response to exogenous DAG treatment as demonstrated by increased pERK levels (Fig. 3.10 A). This shows that mutations of the EF-hands does not disable Cl domain function, and incidentally demonstrates that the EF-hands are not essential for catalytic activation of RasGRP 1 in DT4O cells. Localization of RasGRP1 to endomembranes (ER and Golgi) in NIH 3T3 fibroblasts is dependent on the Cl domain (Tognon et al, 1998). Localization of RasGRP1 in NIH 3T3 cells was unaffected by EF-hand mutation (Fig. 3.10 B), suggesting that EF-hand function is not required for Cl-domain dependent localization of RasGRP1 to endomembranes. A. RG1- RG1- GFP RGI IJEF1 pEF1/2 DAG - 5\u00E2\u0080\u0099 - 5\u00E2\u0080\u0099 - 5\u00E2\u0080\u0099 - 5\u00E2\u0080\u0099 \u00E2\u0080\u0094 \u00E2\u0080\u0094 \u00E2\u0080\u0094 P-ERK2 1.0 8.9 1.0 46.2 1.0 36.4 1.0 33.7 \u00E2\u0080\u0094 \u00E2\u0080\u0094 pan-Ras Transduced RG1 GFP control RGI RGI-pEFI B. Localization in NIH 3T3 fibroblasts Figure 3.10 EF-hand mutations do not affect Cl domain function of RG1. Refer to Fig. 3.7 for schematic representation of constructs. (A) DT4O cells expressing either GFP as a control or the indicated RG1 constructs, described in Fig. 3.6 were untreated (-) or treated with 100 jiM DAG for 5 mm. Levels of p-ERK2 were quantified by western blot, with relative amounts indicated below each sample. Transduced RG1 protein was detected as described for Fig. 3.7. Anti- 64 pan Ras was used as a loading control in this experiment. (B) NIH 3 T3 cells stably expressing either GFP as a control, RG 1 or RG 1-iEF 1 were plated onto glass coverslips and grown overnight before fixation. Cells were then photographed by fluorescence microscopy. Size bars = 10 jim. The experiment shown is representative of 3 independent experiments. If the EF-hands are not required for Cl domain function, then could they be required for PT-mediated plasma membrane targeting? The RasGRP1 construct RG1AC1 lacks a Cl domain but contains an active SuPT and PT domain. This construct translocates to the plasma membrane in response to BCR ligation, although translocation is weak due to the lack of the Cl domain which is needed to counteract the suppressive effects of the SuPT domain. Translocation of RG1AC1 in response to BCR ligation was eliminated by mutations in the first EF-hand (Fig. 3.11), confirming that this EF-hand does not act via the Cl domain, but leaving open the possibility that it acts positively on the PT domain or negatively on the SuPT domain. A. GFP REM GEF EF Cl SuPT PT RG1 D_IJ.._._)4 4_Q- RG1 -ISC1 O\u00E2\u0080\u0094E-\u00E2\u0080\u0094-\u00E2\u0080\u0094- h-c-\u00E2\u0080\u0099 RG1-EF1+Ac1 B. RGI -ACI nil __ H __ 0% pm 0% pm algM H IH 100% pm 30% pm RGI RGI- pEFI+ACI H 0% pm H 0% pm 65 Figure 3.11 PT-mediated, Cl-independent translocation of RG1 is affected by mutation of the first EF-hand. (A) Schematic representation of the constructs used in this experiment. (B) DT4O cells, transduced to express RG1, RG1-AC1 or RG1-tEF1txC1, were untreated (nil) or treated with cL-IgM for 15 mm. The cells were prepared for fluorescence microscopy, analyzed, and displayed as described for Fig. 3.1. The single cells shown are representative of the typical appearance of the majority of cells, except for the ct-IgM-stimulated RG1-ACI-expressing cell which is representative of those with detectable plasma membrane localization. The results shown are representative of 3 independent experiments. The RasGRP1 construct RG1-ASuPT lacks the SuPT domain and translocates to the plasma membrane in response to BCR ligation as effectively as wild-type RasGRP1 (Fig, 3.12 and Beaulieu, N., Rob Kay Lab, unpublished results). Deletion of the SuPT domain was sufficient to permit efficient BCR-induced plasma membrane translocation when the first EF-hand was mutated (Fig. 3.12). Therefore, the first EF-hand enables PT-domain mediated translocation to the plasma membrane, but it does this by counteracting the SuPT domain rather than assisting the PT domain. 66 A. GFP REM GEF EF Cl SuPT PT 0\u00E2\u0080\u0094CD RG1 RG1-jEF1 0-CD 0-CD RG1 -ASuPT RGI-pEF1+ASuPT Figure 3.12 EF-hands allow plasma membrane targeting by mediating SuPT activity. (A) Schematic representation of the constructs used in this experiment. (B) DT4O cells were transduced with the indicated proteins and either untreated (nil) or treated with ct-IgM for 15 mm. The cells were prepared for fluorescence microscopy, analyzed, and displayed as described for Fig. 3.1. The single cells shown are representative of the typical appearance of the majority of cells, except for the cL-IgM-stimulated RG 1-iEF 1-expressing cell which is representative of those with detectable plasma membrane localization. The results are representative of 3 independent experiments. K 0% pm H 100% pm B. nil RGI RG1-jEF1 RGI -ASuPT RGI-pEFI+ASuPT 7% pm 0 pm IL ____________ 100% pm 67 4 CONTROL OF MEMBRANE LOCALIZATION OF RASGRP1 BY THE GEF AND REM DOMAINS 4.1. The REM-GEF domain complex also contributes to EF-hand-dependent regulation of plasma membrane targeting. The previous chapter focused on the role of the EF-hands in the process of BCR induced translocation of RasGRP 1 to the plasma membrane. In this chapter, I describe how some additional experiments initially intended to examine the EF-hands led to the discovery of unexpected roles of the REM and GEF domains in determining both EF-hand-dependent and independent translocation ofRasGRPl to membranes. The experiments described in chapter 3 examined the effects of EF-hand mutations on full-length RasGRP 1 and concluded that intact EF-hands were required for effective BCR-induced plasma membrane targeting of the full-length protein (Fig. 3.7). We also concluded that the mechanism by which EF-hands enhanced plasma membrane targeting was through counteracting the SuPT domain (Fig. 3.12). We then sought to examine the functional interaction between the EF-hands and the SuPT domain in a simpler context with a construct consisting of the EF-hands and the Cl, SuPT and PT domains. Unexpectedly, this construct translocated with high efficiency in response to BCR ligation, whether the EF hands were intact (EF 1/2-Cl -SuPT-PT), mutated (EF 1 ji/2-C 1 -SuPT-PT) or deleted (Cl - SuPT-PT) (Fig. 4.1 A+B). The EF1/2-C1-SuPT-PT construct is missing the REM-GEF domain complex that catalyzes nucleotide exchange on Ras, and an N-terminal extension of unknown function. When this N-terminal extension was deleted from full-length RasGRP 1 68 and the REM-GEF domain complex was retained, BCR-induced translocation of RasGRP 1 was still dependent on the presence of intact EF-hands (Fig. 4.1 C). Several conclusions can be drawn from these findings. First, I showed in chapter 3 that intact EF-hands are required to over-ride suppression imposed by the SuPT domain (Fig. 3.12). However, on its own, the SUPT domain cannot strongly suppress PT plus Cl-mediated translocation because the isolated Cl -SuPT-PT domain is able to translocate efficiently in response to BCR ligation (Fig 4.1B). Because of the ineffectiveness of the SuPT domain in this context, the EF-hands are not required for translocation of the C1-SuPT-PT domain complex. It is only when the REM-GEF domain complex is also present that the SuPT domain becomes a strong suppressor of translocation, and only in this situation are the EF-hands needed to enable translocation by counter-acting the SuPT domain. This reveals an unexpectedly high complexity of integration of the multiple domains of RasGRP 1 in the processes controlling BCR-induced translocation to the plasma membrane. At the first level of regulation, the PT and Cl domains cooperate to provide the means for localizing RasGRP 1 specifically to the plasma membrane in response to signals initiated by the BCR (Beaulieu et al, 2007). At the second level, the REM-GEF domain complex co-operates with the SuPT domain to prevent PT-domain-mediated plasma membrane targeting. At the third level, the EF-hands are required to over-ride negative regulation by the REM-GEF plus SuPT domains, thus restoring the ability of RasGRP1 to respond to BCR signals by translocating to the plasma membrane. 69 A. H 0% pm H \u00E2\u0080\u00A2O0% pm EF1/2-C1 -SuPT-PT EF1 p/2-C1-SuPT-PT EF1/2-C1 -SuPT-PT C1-SuPT-PT Figure 4.1 Plasma membrane targeting of RG1 becomes dependent on intact EF-hands once the REM-GEF domain complex is present. EFI 1J12-C1 -SuPT-PT o-o--\u00E2\u0080\u00A2Q-4-cm* EFI 12.C1 -SuPT-PT 0%pm I \u00C2\u00B0 EFII2-C1-SuPT-PT C1-SuPT-PT B. nil a-lgM nil 0% pm H 100% pm a.lgM 0% pm H 100% pm C. 0 IEJ-\u00E2\u0080\u00A2 RG1AN-term 0 EJ-\u00E2\u0080\u0094-4\u00E2\u0080\u0094-tEJ- RG1pEF1+N-term RGIAN-term RGI pEFI +zN-term nil a-lgM 0% pm H 4% pm 70 The isolated RG1 constructs in this study are displayed schematically above each panel of four images. For a list of the domains, refer to Fig. 3.1. EF1/2-C1-SuPT-PT and EFt1/2-C1-SuPT lack sequences N-terminal to the EF-hands; Cl -SuPT-PT lacks sequences N-terminal to the Cl domain; RG 1 AN-term and RG 1-iEF 1 +AN-term lack sequences N-terminal to the REM domain. Point mutations in EF-hands displayed schematically as black-filled shapes. (A) DT4O cells transduced with EF12-C1-SuPT-PT or EFi1/2-C1-SuPT-PT were untreated (nil) or treated with 5 ig/ml of CL 1gM for 15 mm. (B) DT4O cells transduced with C1-SuPT-PT or EF1/2-C1-SuPT-PT were untreated (nil) or treated with 5 jig/ml of ct-IgM for 15 mm. (C) (D) DT4O cells were transduced with RG1AN- term orRG1-iEF1+AN-term were untreated (nil) or treated with 5 tg/ml of ct-IgM for 15 mm. The cells were prepared for fluorescence microscopy, analyzed, and displayed as described for Fig. 3.1. Fig. 4.1 A+B are representative of 3 and Fig 4.1 C of 2 independent experiments. 4.2. Mutations that prevent occupation of the GEF catalytic site by Ras disable PT- mediated membrane localization, independently of the EF-hands So far we have shown that EF-hands are required to counteract suppression by the SuPT domain, but only when the REM-GEF domain complex is present. We concluded that the REM-GEF domain complex co-operates with the SuPT domain to suppress translocation and that wild-type EF-hands counter-act this suppression. A series of deletions into the REM and GEF domains were made to narrow down the region within the REM-GEF domain complex that is essential for cooperativity with the SuPT domain. The prediction was that deletion of the region in the REM or GEF domain that is responsible for cooperating with the SuPT domain would result in efficient translocation even when the first EF-hand was mutated. Instead, we found that deletions encroaching into the REM or the GEF domains (Fig. 4.2 A) had an unexpected effect: they completely eliminated BCR-induced translocation of RasGRP 1 to the plasma membrane even when the EF-hands were intact (Fig. 4.2). Therefore, in searching for the mechanism by which the REM-GEF domain participates in EF-hand-dependent regulation of RasGRP 1 translocation in response to BCR 71 ligation, we unintentionally found that mutation of the REM or GEF domain disables RasGRP 1 translocation. a. GFP REM GEF EF Cl SuPT PT Q \u00E2\u0080\u0094. b. RGI RGI-AREM RG1 RG1 AREM RG1IGEFC RGI -AGEFC nil a-IgM I I 0% pm 100% pm I I 0% pm Figure 4.2 Deletions in the REM or GEF domain eliminate BCR-induced plasma membrane translocation of RasGRP1. (A) Schematic representation of constructs used in experiments in this figure. All constructs are N- terminally GFP-tagged as shown. RG 1 AREM lacks the REM domain and RG 1 AGEFC lacks a C- terminal portion of the GEF domain of RG 1. (B) DT4O cells transduced with either RG 1, RG1AREM or RG1tGEFC were untreated (nil) or treated with 5 jig/mI of ct-IgM for 15 mm. The cells were prepared for fluorescence microscopy, analyzed, and displayed as described for Fig. 3.1. This experiment is representative of 3 independent experiments. In the exchange factor SOS 1, and presumably in other related exchange factors such as RasGRP 1, the GEF and REM domains closely interact structurally and functionally. 72 Deletions into either the REM or GEF domains eliminate the catalytic activity of RasGRP 1 (Tognon et al, 1998), presumably by disrupting the structure of the REM-GEF domain complex in a way that prevents productive interaction with its Ras substrates. This raises the hypothesis that the membrane-binding competence of RasGRP 1 is a function of the state of its REM-GEF domain complex, with the Ras-bound form of the REM-GEF domain complex permitting membrane localization and unoccupied form preventing membrane localization. To test this hypothesis we used mutations in RasGRP1 that specifically eliminate binding of Ras to the GEF domain. Design of these mutations was based on the known structure of the GEF domain of SOS 1 and functional data from SOS1 and the related GEF CDC25 (Boriack Sjodin et al, 1998, Hall et al, 2001, Margarit et al, 2003, Park et al, 1994). R271E is a charge-switch mutation which lies in a helix adjacent to the catalytic Ras-binding site in the GEF domain. In CDC25, the equivalent mutation specifically disrupts Ras binding without altering local conformation, as shown by the rescue of this mutation by a complementary charge-switch mutation in Ras (Park et al, 1994). The R27 1 E mutation eliminates catalytic activity ofRasGRPl (Beaulieu et al, 2007, Tognon et al, 1998). This mutant is designated GEF-t 1. We made an additional GEF domain mutant of RG 1, designated GEF-i2, containing mutations designed to remove two aliphatic chains (1354A+L355A), which in SOS1 are required to stabilize Ras binding in the catalytic pocket (Margarit et al, 2003). We tested the catalytic activity of GEF-p2 independently of any effects on membrane localization by the attachment of a K-Ras derived prenylation signal (pren), which forces membrane localization. GEF-p2/pren did not have the ability to induce transformation of 73 NIH 3T3 fibroblasts, a phenotype that can be induced by expression of the un-mutated counterpart RG1/pren (Fig 4.3). Therefore, we can conclude that this new GEF mutant is also catalytically inactive. GFP control RGI/pren GEF- p2lpren Figure 4.3 Effect of mutation of the GEF domain on ability of RG1 to transform Nifi 3T3 cells. Subconfluent NIH 3 T3 cells were infected with equal amounts of retrovirus containing cDNAs for GFP, RG l/pren (RG 1 prenylated at the C-terminus of the GEF domain) or GEF-i2/pren (mutant RG 1 containing 13 54A+L3 5 5A substitutions in the GEF domain, prenylated at the C-terminus of the GEF domain). Cells were cultured in the presence of serum and photographed by phase contrast microscopy after 14 days. Size bars 50 jIm. This experiment is representative of 2 independent experiments. We next determined the effect of the R271E and 1354A+L355A mutations on the plasma membrane targeting of RasGRP1 in response to BCR ligation in DT4O cells. We predicted that if Ras-binding to the catalytic site of RasGRP 1 was essential for the plasma membrane targeting of RasGRP1, then GEF-jil (R271E) and GEF-1i2 (1354A+L355A) should have defects in plasma membrane targeting. As predicted, neither GEF-1t1 nor GEF ji2 translocated to the plasma membrane after BCR ligation (Fig. 4.4 A). RasGRP1 is found present predominantly at endomembranes in NIH 3T3 cells and this localization has been shown to be dependent on the Cl domain and independent of the EF-hands or SuPT/PT 74 domains (Beaulieu et al, 2007, Tognon et al, 1998). We asked whether Ras binding to the catalytic site in RasGRP 1 affected localization in NIH 3T3 cells. We found that both the R271E and 1354A+L355 mutations in the GEF domain reduced perinuclear localization in NIH 3T3 cells (Fig. 4.4 B), indicating a reduction in endomembrane localization. Therefore, loss of membrane binding correlates with loss of binding of Ras at the catalytic site of the GEF domain and supports the hypothesis that the GEF domain must be able to bind Ras in order for RasGRP1 translocation to occur. A. RGI I I nil H _ H I 0% pm I 0% pm 0% I H 1100% pm 0% pm 0% B. RGI GEF- p1GFP control GEF- p2 Figure 4.4 Mutations that prevent occupation of the GEF catalytic site by Ras disable membrane localization of RasGRP1. 75 All constructs are N-terminally GFP-tagged. (A) DT4O cells transduced with either RGI, GEF-jil (R271E) or GEF-ji2 (1354A+L355A) were untreated (nil) or treated with 5 jig/mi of ct-IgM for 15 mm. The cells were prepared for fluorescence microscopy, analyzed, and displayed as described for Fig. 3.1. (B) NIH 3T3 cells stably expressing either GFP, RG1, GEF-jil (R271E) or GEF-ji2 (1354A+L355A) were plated onto glass coverslips and grown overnight before fixing with 4% formaldehyde. Cells were then photographed by fluorescence microscopy. Size bars 10 jim. This experiment is representative of 3 independent experiments. 4.3. Suppression of membrane binding by the Ras-unoccupied GEF domain involves both the Cl and PT domains. The fully suppressive effect of GEF mutation on BCR-induced plasma membrane targeting occurred whether the SuPT domain was present or absent (Fig. 4.5) and so does not involve the same mechanism by which the non-mutated REM-GEF domain complex acts in conjunction with the SuPT domain to suppress PT domain mediated plasma membrane targeting. This indicates that the PT domain is directly disabled by mutation of the GEF domain. 76 A. GFP REM GEF EF Cl SuPT PT RG1 GEF-1 GEFp1 +ASuPT 1100% pm 0% pm 0% Figure 4.5 Suppressive effects of GEF mutation occurred whether the SuPT domain was present or absent. (A) Schematic representation of constructs used in experiments in this figure. All constructs are N- terminally GFP-tagged as shown. The GEF domain is coloured in black to indicate mutation. (B) DT4O cells transduced with either RG1, GEF-1.tl (R271E) or GEF-jil+ASuPT were untreated (nil) or treated with 5 jig/mi of c-IgM for 15 mm. The cells were prepared for fluorescence microscopy, analyzed, and displayed as described for Fig. 3.1. This experiment is representative of 2 independent experiments. As mentioned already, GEF mutation also caused a loss of detectable endomembrane localization of RasGRP1 in NIH 3T3 cells (Fig. 4.4). Because endomembrane localization of RasGRP1 in NIH 3T3 cells is mediated by the Cl domain and occurs independently of the PT or SuPT domains (Beaulieu et a!, 2007, Tognon et al, 1998), the loss of endomembrane localization in NIH3T3 cells resulting from GEF mutation shows that the Cl domain is also functionally disabled when the GEF domain is unable to bind Ras. Thus, the REM-GEF B. RGI KN 0% pm GEF-pi GEF-pi +ASuPT 0% pm 0% pm 77 domain complex is involved in two mechanistically distinct processes. The first involves cooperativity between the REM-GEF domain complex and the SuPT domain in suppressing PT domain-mediated plasma membrane targeting, with this being counteracted by the EF hands. The second process occurs when the GEF domain is not occupied by Ras, as represented by the R271E or 1354A+L355 mutations, and involves functional inactivation of both the PT and Cl domains. 4.4. Loss of membrane binding by GEF domain mutation does not reflect loss of signal transduction from Ras. In the previous section, we were able to show for the first time that occupation of the GEF catalytic domain by Ras is required for both endomembrane and plasma membrane localization. We became interested in understanding the mechanism with which Ras binding by the GEF domain enables membrane binding of RasGRP1. One possibility is that RasGRP 1 is positively regulated by signals transduced by the Ras proteins that it GTP loads. In this case, RasGRP 1 would initiate catalysis of Ras GTP-loading and the resulting signal transduction from Ras would result in a positive feed-back ioop by which RasGRP1 would, by some unknown mechanism, gain competence for membrane binding. We tested this by co-expressing in DT4O cells a constitutively signaling K-Ras G12V mutant along with GEF-mutated RasGRP1 (Fig. 4.6). Despite constitutive signaling through the Ras pathway, as detected by raised p-ERK2 in DT4O cells (Fig. 4.6 B), the mutant RasGRP1 still had no detectable plasma membrane localization (Fig. 4.6 A). Conversely, inhibition of the Ras to ERK signaling pathway by the MEK inhibitor UO126 did not impede BCR-induced plasma membrane targeting of wild-type RasGRP1 (Fig. 4.6 C+D). Therefore, lack of Ras 78 binding at the GEF site appears to cause an intrinsic defect in membrane localization of RasGRP 1, rather than an absence of signaling required for membrane localization of RasGRP1. 79 a. Vector control Ras-VI 2 a-IgM 0 15\u00E2\u0080\u0099 0 15\u00E2\u0080\u0099 \u00E2\u0080\u0094 P-ERK2 C. 1 6.3 1.7 32.0 Rapi Transduced RGI d. U0126 (10 pM) U0126 (pM) a-IgM 0 1.0 3.0 10.0 0 10\u00E2\u0080\u0099 0 10\u00E2\u0080\u0099 0 10\u00E2\u0080\u0099 0 10\u00E2\u0080\u0099 88.4 1 54.8 1 27.5 1 5.1 P-ERK2 \u00E2\u0080\u0094 \u00E2\u0080\u0094 .- Total-ERK nil RGI 100% pm Figure 4.6 Loss of plasma membrane binding by the GEF domain mutation does not relect loss of signal transduction from Ras. (A) DT4O cells transduced with either RG 1 or GEF-i 1 (R27 1 E) were transduced again to co-express vector control or Ras-V 12 and either untreated (nil) or treated with 5 j.tg/ml of ct-IgM for 15 mm. The cells were prepared for fluorescence microscopy, analyzed, and displayed as described for Fig. 3.1. The single cells shown are representative of the typical appearance of the majority of cells. (B) DT4O 80 RGI GEF-pi I Ras-V12 nil a\u00E2\u0080\u0094lgM b. H 0% pm H 100% pm GEF-pi 0% pm 0% pm 0% pm cells were transduced to co-express GEF-1i1 (R271E) and either vector control or Ras-V12 and were treated with 5 ig/ml of x-IgM for the indicated time points. Levels of pERK2 were determined by western blot using an anti-phospho-specific ERK antibody. Levels of p-ERK2 were quantified by densitometry, with fold changes indicated below each band. Rapl was used as a loading control in this experiment. Transduced RG 1 protein, was detected using an antibody that recognizes the C- terminal portion of the protein. (C) DT4O cells transduced with RG 1 were treated with the indicated doses of the MEK inhibitor UO 126 for 20 minutes prior and then stimulated with 5 jig/ml of cL-IgM for the indicated times. Levels of pERK2 were determined by western blot. Total ERK2 was detected using an anti-ERK antibody and is displayed to shows that levels of total ERK in the cells do not change as a result of treatment. (D) DT40 cells transduced with RGI were treated with U0126 (10 jiM) and either unstimulated (nil) or treated with 5 jig/ml of c-IgM for 15 mm. The cells were prepared for fluorescence microscopy, analyzed, and displayed as described for Fig. 3.1. This experiment is representative of 2 independent experiments. 4.5. Membrane localization by the Cl or PT domain is not essential for a low level of activity of the REM-GEF domain complex. If the REM-GEF domain complex must be occupied by Ras in order for membrane binding to occur, and membrane binding is needed in order for the GEF domain to contact Ras, then how does a RasGRP 1 protein that is unoccupied by Ras ever get out the inactive state? One possible explanation is that the unoccupied GEF domain does not completely blockade the Cl or PT domains, which would then be occasionally free to attach to an available membrane. Once membrane contact was transiently achieved via the Cl or PT domain, the GEF domain complex could encounter Ras, thus fully releasing the PT or Cl domain to provide stable membrane binding. An alternate explanation is that the REM-GEF domain complex may be able to encounter membrane-bound Ras at a low but significant rate without the assistance of the Cl or PT domains, and once this occurred the Cl and PT domains would be liberated to mediate stable membrane binding. To test the latter possibility, we used the NIH 3T3 transformation assay because it is the most sensitive means of detecting low levels of RasGRP1 activity (Tognon et al, 1998). The REM-GEF domain 81 Figure 4.7 Membrane localization by the Cl or PT domains is not essential for activation of the REM-GEF domain complex. (A) NIH 3T3 cells stably expressing either GFP, RG1, or the isolated REM-GEF domains in monomeric (REM-GEF) or dimeric (REM-GEFx2) forms, were plated onto glass coverslips and grown overnight. Cells were then photographed by fluorescence microscopy. Full-length RasGRP1 is localized to the pen-nuclear region, wheras the REM-GEF domain complex (REM-GEF and REM-GEFx2) is localized diffusely in the cytoplasm and the nucleus. Size bars = 10 tm. (B) 82 complex was not detectably localized at membranes in the NIH 3T3 cells, even when the REM-GEF domain was dimerized to increase potential avidity of membrane interactions (Fig. 4.7 A). Despite its lack of membrane localization, the isolated REM-GEF domain complex was able to induce transformation of NIH 3T3 cells, albeit weakly in comparison to full length RasGRP1 (Fig. 4.7B). Therefore, while the REM-GEF domain complex has no apparent ability to stably bind membranes on it own it can nonetheless bind and GTP-load Ras GTPases at a rate sufficient to be detectable by the NIH 3T3 transformation assay. A. GFP control RGI REM-GEE (REM-GEF)X2 B. GFP control RGI REM-GEE (REM-GEF)x2 Subconfluent NIH 3T3 cells were infected with equal amounts of retrovirus containing cDNAs encoding either GFP, RG 1, REM-GEF or REM-GEFx2. Cells were cultured in the presence of serum and photographed by phase contrast microscopy after 21 days. The REM-GEF domain comlex in monomer and dimer formations is capable of transforming the appearance of NIH 3T3 cells, despite not being localized to the perinuclear region, as demonstrated in this experiment (compared to GFP control). However NIH 3T3 cell transformation by expression of the REM-GEF domain complex is mild compared to the full-length RG1-expressing cells. Size bars = 50 rim. This experiment is representative of 3 independent experiments. While these experiments do not rule out the first scenario, that functional inactivation of the Cl and PT domains by the unoccupied GEF domain is only partial, they do demonstrate that weak RasGRP 1 activation can occur despite complete functional inactivation of the Cl and PT domains (in this case by their removal). Therefore, the initial stage of RasGRP1 activation may be triggered by transient encounters of the GEF domain with Ras. The resulting functional liberation of the Cl and PT domains could then confer stable membrane binding and an increased probability of encounter with Ras, thus providing positive feedback for the transition from low to high RasGRP 1 activity. 4.6. A putative Ras binding site in the REM domain of RasGRP1 contributes to both membrane localization and activation. In SOS 1, Ras-GTP binds to a specific site within the REM domain, and this induces a conformational change that increases the efficiency of Ras binding at the catalytic site within the adjacent GEF domain (Freedman et al, 2006, Margarit et al, 2003). This provides a mechanism for positive feedback regulation, whereby local generation of Ras-GTP, either by SOS 1 or another GEF (Roose et al, 2007), results in increased Ras binding at the catalytic site in the GEF domain, and thus increased SOS 1 activity. If RasGRP 1 was regulated in the same way, then binding of Ras-GTP to the REM site could have a two-fold 83 positive feedback effect. In addition to directly increasing catalytic activity, increased Ras binding at the GEF site would trigger functional liberation of the Cl and PT domains, thereby increasing catalytic activity by increasing the probability of both the GEF and REM domains encountering membrane-bound Ras. To determine if the REM site could positively regulate RasGRP1 membrane localization and activation in this way, we made a mutation in the homologous site in the REM domain of RasGRP1, L123A+K124E, which is predicted to preclude Ras-GTP binding based on the SOS 1 structure (Margarit et al, 2003) . In comparison to cells expressing wild-type RasGRP1, NIH 3T3 cells expressing the RasGRP1-L123A+K124E mutant (REM-ti in Fig. 4.8) had a weaker transformed phenotype, with less cell rounding but still showed distinct refractility and cell elongation (Fig. 4.8 A). In NIH 3T3 cells, RasGRP1 was localizated to the pen-nuclear area (Fig. 4.8B), which is presumably to endomembranes (golgi and ER) based on previous published data (Tognon et al, 1998). The RasGRP1-L123A+K124E mutant also had pen-nuclear localization that was less intense than wild-type RasGRP1 but still readily detectable, unlike RasGRPlcarrying the 1354+L355A (mutant GEF-p2) mutation in the catalytic site of the GEF domain (Fig. 4.8 B). Similarly, the RasGRP1- L123A+K124E mutant had weak but detectable plasma membrane localization in the BCR-stimulated DT4O cells (Fig. 4.8 C), in contrast to the complete lack of plasma membrane localization of the 1354+L355A mutant (Fig. 4.8 C). These results demonstrate that the putative Ras binding site in the REM domain is required for efficient plasma membrane localization of RasGRP1. Extrapolating from what is known about SOS 1, this is probably due to the Ras-bound REM domain acting as an allosteric regulator of the 84 efficiency of Ras binding to the GEF site, which in turn determines the functional Figure 4.8 A putative Ras binding site in the REM domain of RasGRP1 positively regulates membrane localization. (A) NIH 3T3 cells were infected with equal amounts of retrovirus containing cDNAs for GFP, RG 1, REM-ji (L123A+K124E) or GEF-i2 (1354A+L355A). Cells were cultured in the presence of serum and photographed by phase contrast microscopy after 14 days. Size bars 50 tm. (B) NIH 3T3 cells stably expressing either GFP, RG1, REM-Ft (L123A+K124E) orGEF-1i2(1354A+L355A) were plated onto glass coverslips and grown overnight before fixation. Cells were then photographed by 85 GFP control competence of the Cl and PT domains. A. B. GFP control RGI REM- p GEF- p2 C. GEF-pi REM-piRGI IKJI 100% pm I 0% pm 0% pm H9% pm fluorescence microscopy. Size bars 10 jim. (C) DT40 cells transduced with either RG1, GEF-jil or REM-ji were untreated (nil) or treated with 5 jig/ml of a-IgM for 15 mm. The cells were prepared for fluorescence microscopy, analyzed, and displayed as described for Fig. 3.1. The single cells shown are representative of the typical appearance of the majority of cells, except for the cL-IgM-stimulated REM-ji 1-expressing cell which is representative of those with detectable plasma membrane localization. The results shown are representative of 2 independent experiments. 86 5 FUNCTIONAL CHARACTERIZATION OF A SPLICE VARIANT OF RASGRP1 FOUND IN PATIENTS WITH SYSTEMIC LUPUS ERYTHEMATOSUS (SLE) 5.1. A RasGRP1 splice variant lacking exon 11 has a functional GEF domain but has a plasma membrane targeting defect that mimics EF-hand mutation. Systemic lupus erythematosus (SLE) is a severe autoimmune disease that is characterized by the production of autoantibodies and the generation of circulating immune complexes resulting in multiple organ damage. SLE is thought to be caused by multiple genetic and environmental factors and there is strong evidence for the role of B and T cells in disease pathogenesis (Anolik, 2007, Mor et al, 2007). This includes B and T cell abnormalities, including loss of self-tolerance and aberrant signaling. Although SLE is a polygenic disease, animal models with single gene mutations have lead to the disease (Mor et al, 2007). In the case of RasGRP 1, underexpression of this protein due to gene deletion or spontaneous mutation resulted in the development of lupus-like autoimmunity in mice (Dower et al, 2000, Layer et al, 2003). One group recently looked for defective isoforms and/or diminished levels of RasGRP 1 in the T-lymphocytes of a particular cohort of SLE patients (Yasuda et al, 2007). Interestingly, this group found several naturally occurring alternative splice forms of RasGRP 1 in both SLE and healthy human subjects that were a result of alternative splicing of exons 5-17. One particular splice variant lacks exon 11 and was found to occur more prevalently in patients with SLE. Although they found that this exon 11 splice variant of RasGRP 1 can be stably expressed at the protein level, they did not functionally characterize this isoform (Yasuda et al, 2007). Since my work uncovered new regulatory aspects of RasGRP 1, I became interested in studying the functional consequences of deletion of exon 11. Exon 11 encodes a 35 amino 87 acid segment that starts immediately C-terminal to the GEF domain and encroaches into the predicted N-terminal helix of the first EF-hand, but does not include the calcium binding loop (Fig. 5.1). Because the known essential parts of the GEF domain and EF-hand are intact, loss of exon 11 might have no functional effect on RasGRP 1, but it is also possible that removal of these 35 amino acids could functionally perturb either the GEF domain, by altering sequences at its C-terminus, or the first EF-hand, by truncating the helix at its N- terminus. p Pro QG G GG tt t It REPRNHRAPPLTPSKPPVVVDWASGVS PKPDPKTIS KYVQRMVDSVFKNYDLDQDGYI SQEEFEKIAAS FP - I ____________________________ GEF encoded by exon 11 EF-h 1 Figure 5.1 Exon 11 amino acid sequence and location in the RasGRP1 sequence. Location of exon 11-encoded amino acids relative to the C-terminal boundary of the GEF domain and the first EF-hand. Residues in the EF-hand that are potentially involved in calcium binding are bolded, as are prolines that have the potential to form an SH3 domain-binding site. The vertical arrows indicate the proline cluster mutation (ji Pro). The 35 amino acid segment encoded by exon 11 is also present in mouse RasGRP 1, therefore, we were able to generate an exon 11 splice variant of murine RasGRP 1 (RG 1- Aexil). We tested the functionality of the GEF domain and the first EF-hand of this protein by assessing activity and membrane localization. When expressed in NIH 3T3 cells, RG1- Aexi 1 induced transformation equivalently to full-length RasGRP1 demonstrating that the catalytic competence of the GEF domain is not compromised by exon 11 deletion (Fig. 5.2A). Also, RG1 -Aexi 1 localized to the pen-nuclear region equivalently to full-length 88 RasGRP1 (Fig. 5.2C). This localization is presumably to endomembranes (golgi and ER) based on previously published findings (Tognon et al, 1998). Therefore, exon 11 deletion does not affect the GEF domain as a permissive regulator of membrane localization. GFP control RGI RGI-Aexll Figure 5.2 Loss of exon 11 in RasGRP1 does not affect transformation or membrane localization in Nifi 3T3 cells. (A) Deletion of exon 11 is shown schematically. (B) Subeonfluent NIH 3T3 cells were infected with equal amounts of retrovirus containing cDNAs encoding either GFP, RG1 or RG1-Aexll. Cells were cultured in the presence of serum and photographed by phase contrast microscopy after H days. Size bars = 50 urn. (C) NIH 3T3 cells stably expressing either GFP, RG1, or RG1-Aexl were photographed by fluorescence microscopy. This experiment is representative of 3 independent experiments A. GFP REM GEF EF Cl SuPT PT RG1 RG1 -zXexl 1 GFP control RGI RGI-Aexll C. 89 In DT4O cells, plasma membrane targeting of RasGRP 1 in response to BCR signaling was significantly reduced by mutations disrupting the first EF-hand (RG1-pEF1) (Fig. 3.6). Considering that removal of exon 11 truncates part of the entering helix of the first EF-hand, it is possible that RG1Aex11 also disrupts the first EF-hand function. When expressed in DT4O cells, BCR-induced plasma membrane targeting of RG1-Aexll was significantly reduced in comparison to full-length RasGRP1, as was its activation (Fig. 5.3). As described in chapter 3, deletion of the SuPT domain fully restored plasma membrane targeting ofRGl-1iEF1 in response to BCR ligation. Deletion of the SuPT domain in RG1- Aexi 1 also restored plasma membrane targeting in response to BCR ligation (Fig. 5.3). Therefore, RG 1 -Aex 11 is functionally equivalent to RG 1 -tEF 1, in that it impedes plasma membrane targeting and does so via the SuPT domain. 90 A. GFP REM GEF EF Cl SuPT PT RG1 \u00E2\u0080\u0094LJ- RG1 -Aexi 1 Q_ L. B. RGI RGI-Aexll RG1-Eex11+ ASuPT nil KJ 0% pm a-lgM _____ _____ 4% pm C. RGI - GFP RGI Aexil a-lgM 0\u00E2\u0080\u0099 15\u00E2\u0080\u0099 0\u00E2\u0080\u0099 15\u00E2\u0080\u0099 0\u00E2\u0080\u0099 15\u00E2\u0080\u0099 P-ERK2 1.0 15 1.3 33 1.0 16 \u00E2\u0080\u0094 Total ERK2 transduced \u00E2\u0080\u009C RasGRPI Figure 5.3 Loss of exon 11 in RasGRP1 results in plasma membrane targeting and activation defects that mimic EF-hand mutation. (A) Loss of exon 11 either alone or in combination with deletion of SuPT are shown schematically. (B) DT4O cells transduced with either RGI, RGI-&xll or RG1-Aexl 1+ASuPT were either untreated (nil) or treated with 5 ig/ml of ct-IgM for 15 mm. The cells were prepared for fluorescence microscopy, analyzed, and displayed as described for Fig. 3.1. The single cells shown are representative of the typical appearance of the majority of cells, except for the a-IgM-stimulated RG 1 -Aex 11-expressing cell which is representative of those with detectable plasma membrane localization. (C) DT4O cells transduced with either GFP, RGI or RG1-Aexll were untreated (nil) or treated with 5 g/ml of-TgM for 15 mm. Levels of pERK2, total ERK and RasGRP1 were 0% pm IL) 100% pm 91 determined by western blot. Levels of p-ERK2 were quantified by densitometry, with fold changes indicated below each band. This blot is representative of 3 independent experiments. 5.2. A proline-rich segment in exon 11 is not important for BCR-induced plasma membrane targeting of RasGRP1. In addition to lacking part of the N-terminal helix of the first EF-hand, RG1-Aexll also lacks a proline-rich segment which includes an SH3 domain-binding motif that has been proposed as a potential regulatory element controlling RasGRP 1 localization or activation (Fig. 5.1) (Kosco et a!, 2008). This raises the possibility that binding of an SH3 domain (or another ligand recognizing the proline cluster) is involved in the process whereby the first EF-hand controls plasma membrane targeting of RasGRP 1. In this situation, the plasma membrane targeting defect of RG1-Aexl 1 could be due to the removal of this proline cluster, rather than direct modification of the first EF-hand. However, mutation of the prolines had no effect on BCR-induced plasma membrane translocation of RasGRP1 in DT4O cells (Fig. 5.4) in contrast to the loss of translocation resulting from either exon 11 deletion or mutation of the first EF-hand. Therefore, removal of exon 11 by alternative splicing appears to suppress plasma membrane targeting of RasGRP 1 by disabling the first EF-hand. As a result, the cellular site of activation of this splice variant is partially restricted in comparison to full-length RasGRP 1, such that only full-length RasGRP 1 can be targeted to the plasma membrane in DT4O cells while both full-length and exon 11-deleted RasGRP 1 can localize to and be activated at endomembranes in NIH 3T3 fibroblasts. 92 RGI -pPro nil a\u00E2\u0080\u0094lgM 100% pm Figure 5.4 BCR-induced plasma membrane targeting of RG1 is not affected by mutations in a proline-rich segment encoded in exon 11. DT4O cells transduced with either RG1 or RG1-jiPro (contains mutation of proline residues in the region encoding exon- 11) were either untreated (nil) or treated with 5 jig/mI of cL-IgM for 15 mm. The cells were prepared for fluorescence microscopy, analyzed, and displayed as described for Fig. 3.1. The single cells shown are representative of the typical appearance of the majority of cells. Note that for this figure, RG 1 and RG 1 -jiPro are missing the short stretch of sequence just N-terminal to the REM domain and which localizes to membranes as effectively as full-length RG 1. This experiment is representative of 3 independent experiments. 0% pm 0% pm 100% pm 93 6 DISCUSSION 6.1. New insights into the mechanism of RasGRP1 activation The goal of my research was to investigate if and how RasGRP 1 is regulated by Ca2 signals arising from activation of antigen receptors and to identify the requirement of the EF-hands in this process. My work in DT4O cells revealed that the EF-hands are required for RasGRP1 activation but may not be directly regulated by BCR-induced Ca2 fluxes. Furthermore, I determined that plasma membrane translocation of RasGRP 1 is controlled by the combined actions of EF-hands, GEF and SuPT domains. In addition to addressing the initial objectives of my research, I also found that Ras binding to the GEF domain was required for Cl - and PT-domain mediated membrane targeting. Finally, I applied these new findings to functionally characterize a naturally occurring splice variant of RasGRP 1, found more prevalent in patients with SLE. The cellular site of activation of this splice variant was partially restricted likely due to disruption of the function of the first EF-hand. 2+6.2. The role of EF-hands and Ca flux in regulating RasGRP1 In DT4O cells, RasGRP 1 is activated downstream of the BCR by translocation to the plasma membrane, which co-localizes it with its substrate Ras. Second messengers DAG and 1P3 are simultaneously generated by activation of PLC enzymes downstream of the BCR. RasGRP1 translocation to the plasma membrane is in part mediated by this BCR induced DAG generation, via its Cl domain, and also, by unidentified signals from the BCR acting on its PT domain (Beaulieu et a!, 2007). 1P3 generation causes a rise in cytosolic Ca2 concentrations, which can activate variousCa2-binding proteins including those that contain EF-hands. RasGRP1 contains a pair of EF-hands, of which the second EF-hand has 94 been previously reported to bind Ca2in an in-vitro assay (Ebinu et al, 1998). This result suggests that the EF-hands could be regulated by Ca2 signals, however, while some studies support this hypothesis (Ebinu et al, 2000, Guilbault & Kay, 2004, Kawasaki et al, 1998), others suggest that the EF-hands are not regulatory (Lorenzo et al, 2000, Tognon et al, 1998). Despite this discrepancy, many review articles consistently cite that RasGRP 1 is both DAG and Ca2 regulated. While many aspects of RasGRP 1 regulation downstream of the BCR had been previously worked out by our group in DT4O cells, the role of the EF-hands in this system had not been investigated. Therefore, we asked whether RasGRP1 could be activated by BCR-signalling, under conditions in which Ca2 flux was eliminated by treatment with the intracellular Ca2 chelator BAPTA-AM. Surprisingly, RasGRP 1 was capable of effectively translocating to the plasma membrane and activating ERK2 signalling in BAPTA-AM treated cells. We then hypothesized that perhaps RasGRP1 was responsive to localized changes in Ca2 concentration at the immediate vicinity of plasma membrane channels (Dellis et al, 2006, Putney, 2007). These local Ca2 fluxes could occur despite the presence of BAPTA, which cannot immediately chelate all calcium flowing in through a channel (Rintoul et al, 2001). However, RasGRP1 localized effectively to the plasma membrane in the presence of BAPTA combined with removal of extracellular Ca2 (with EGTA), which eliminates the possibility of Ca2 entry from plasma membrane Ca24 channels. Therefore, we concluded that RasGRP1 was not regulated by BCR-induced changes in Ca2 concentration in the cytosol or at the plasma membrane. While eliminating Ca2 fluctuations with BAPTA-AM and EGTA is one way to address the requirement of Ca2 for EF-hand function, disrupting the Ca2 binding sites of the EF-hands is an alternate way to address the same question. For this, we introduced 95 specific point mutations in the EF-hands that are predicted to perturb Ca2 coordination, while minimizing any disturbances to the structure of the protein. We found that BCR induced plasma membrane targeting and activation of RasGRP 1 are considerably reduced by mutations disrupting the Ca2 binding potential of the first EF-hand, and were further eliminated by the same mutations in both EF-hands. However, BCR-induced plasma membrane localization and activation of RasGRP I were not affected by point mutations disrupting Ca2 binding to the second EF-hand alone. This suggests that the first EF-hand can compensate for lack of the second EF-hand, but not vice versa. In conclusion, the first EF-hand makes a major, whereas the second EF-hand makes a minor contribution to BCR induced RasGRP 1 activation. We were not surprised at this finding since it is not unusual for one EF-hand in a pair to be non-functional for, or in this case less relevant to, calcium binding. This is because a pair of EF-hands is more structurally stable than a single EF-hand on its own, owing to the formation of an anti-parallel 13-sheet between EF-hand loops, and therefore even if one of the EF-hands in a pair is not critical for Ca2 binding, it could still serve a structural role (Gifford et al, 2007, Grabarek, 2006). What was perplexing was that residues in the first EF-hand predicted to confer Ca2- binding, were found to be critical for regulation of RasGRP 1 downstream the BCR, but that Ca2 flux was not a part of its mechanism. Several alternative hypotheses may explain this discrepancy. 1) The first hypothesis is that the EF-hands don\u00E2\u0080\u0099t bind and therefore aren\u00E2\u0080\u0099t regulated by Ca2. This would explain why Ca2 flux is not required for BCR responsiveness. In this case, the effects seen by point mutation of the first EF-hand would have to be attributed to disruption of an alternate mechanism of the EF-hands. For instance, the point mutations may be disrupting intra-molecular interactions or interactions with 96 2+ 2+targets other than Ca . 2) The second hypothesis is that the EF-hands bind Ca but this interaction is not regulatory. In some EF-hands, Ca2 binding is required for maintaining the structural integrity of the protein (Busch et al, 2000, Gifford et al, 2007). If this were the case, it would suggest that the EF-hands of RasGRP 1 have a high affinity for Ca2, such that they would already be in the Ca2-bound state in the resting cell. They would also be occupied by Ca2 even in the presence of BAPTA, if the affinity of the EF-hand for Ca2 was higher than the affinity of BAPTA for Ca2 (Kd = 150 nM) (Gifford et al, 2007). We tried to address this possibility by increasing the concentration of BAPTA-AM to further reduce cytosolic free Ca2 available for binding to the EF-hands. Indeed, at a higher dose of BAPTA-AM, RasGRP1 was no longer detected at the plasma membrane. However, the other two known mechanisms of RasGRP 1 plasma membrane targeting (mediated by the Cl and PT domains) were also affected at a higher dose of BAPTA-AM. In the case of the Cl domain, this may reflect Ca2 dependence of PLCy2, which is responsible for BCR-induced DAG generation (Nishida et al, 2003). Therefore, we were unable to attribute the loss of membrane localization of RasGRP 1 under these conditions directly to the EF-hands. (3) The third hypothesis is that the EF-hands bind Ca2 and that under some circumstances this interaction positively regulates plasma membrane targeting. It could be that in DT4O cells resting Ca2 levels are high enough to keep the first EF-hand constitutively bound to Ca2, and thereby insensitive to increases in cytosolic Ca2 levels induced by BCR ligation. As a result, plasma membrane targeting would always be enabled, although dependent on interaction of DAG and PT ligand by BCR ligation. However, in a cell where basal Ca2 concentration is lower, such that the first EF-hand is not occupied by Ca2, receptor induced Ca2 flux would be required for RasGRP1 translocation. As a variation of this hypothesis, it 97 could be that BAPTA is not fully effective at eliminating all sources of Ca2 flux in BCR stimulated DT4O cells. Although we eliminated Ca2 entry through channels at the plasma membrane as a possible source of Ca2, perhaps BAPTA is unable to fully chelate local Ca2 fluxes from intracellular channels, e.g. at the endoplasmic reticulum. In this case, the EF hands could latch onto Ca2 before BAPTA is able to get to it. Future experiments could address this possibility by treating the cells with 1P3 receptor blockers in combination with intracellular and extracellular Ca2 chelation. In conclusion, while we cannot be certain if the EF-hands ofRasGRPl are positively regulated by the Ca2 rise induced by receptors, we are able to conclude that intact EF-hands are required for RasGRP 1 function. Specifically, potential Ca2 binding residues in the first EF-hand are critical for BCR-induced RasGRP1 targeting to the plasma membrane and activation of ERIC Our finding that the first EF-hand is more important than the second EF hand goes against previous work indicating that only the second EF-hand binds Ca2 in vitro (Ebinu et al, 1998). The in vitro Ca2 binding assay used for that study subjected the protein to SDS denaturing and while the second EF-hand seems to be fine under these conditions, we cannot be sure that the first EF-hand was not disrupted by this treatment. 6.3. Mechanism of EF-hand dependent translocation. 6.3.1 Involvement of the SuPT domain and REM-GEF domain complex In lymphocytes, RasGRP 1 translocates to the plasma membrane and/or internal membranes (e.g. Golgi) depending on the nature of the stimulation, and is capable of 98 activating its substrate Ras at either of those cellular sites (Beaulieu et al, 2007, Bivona et al, 2003, Caloca et al, 2003b, Carrasco & Merida, 2004, Ebinu et al, 2000, Mor et al, 2007, Perez de Castro et aT, 2004, Sanjuan et al, 2003, Schwartz, 2003, Zha et al, 2006, Zugaza et al, 2004). In DT4O cells, BCR ligation results in RasGRP1 translocation exclusively to the plasma membrane (Beaulieu et al, 2007, Caloca et al, 2003b). This translocation step is dependent on the Cl domain binding to DAG, and the PT domain binding to its ligand after BCR stimulation (Beaulieu et aT, 2007). In this study, we found that the first EF-hand was required for optimal BCR-induced plasma membrane translocation of RasGRP 1. Therefore, we hypothesized that the first EF-hand assists plasma membrane targeting by positively regulating the Cl and/or PT domains. We first investigated whether intact EF-hands were important for Cl domain function. Previously, it was shown that EF-hands were not required for endomembrane localization of RasGRP1 in NIH 3T3 cells, which is mediated by the Cl domain (Tognon et al, 1998). This suggests that the Cl domain is capable of binding to membranes in the absence of functional EF-hands. In support of this finding, we found that endomembrane localization of RasGRP1 in NIH 3T3 cells was not affected by mutation of the first EF-hand. In DT4O cells, treatment with exogenous DAG resulted in increased phosphorylation of ERK2 by RasGRP 1, whether the EF-hands were intact or point mutated. The levels of exogenous DAG in these treatments are in excess of the levels of DAG produced by BCR ligation and therefore are sufficient to activate RasGRP1 via the Cl domain alone. Therefore, in DT4O cells, the Cl domain is capable of being functional despite mutation of the EF-hands. 99 Because the EF-hands are not required for Cl domain function, we hypothesized that they assist the PT domain to bind its target at the plasma membrane. It was previously shown in DT4O cells, that the PT domain was capable of partially driving RasGRP1 to the plasma membrane upon BCR ligation, in the absence of the Cl domain (Beaulieu et al, 2007). We found that, also in the absence of the Cl domain, PT-mediated plasma membrane targeting was shut down by mutation of the first EF-hand. This demonstrated that EF-hands were required to be intact for effective PT domain function. We postulated that the EF-hands could permit PT domain function by either regulating the PT domain directly or indirectly via shutting down the SuPT domain. We found that defective plasma membrane targeting caused by mutation of the first EF-hand in RasGRP1 was restored to full capacity by deletion of the SuPT domain. While previously it was shown that the SuPT domain counteracted the PT domain in RasGRP 1, it had not been determined how the SuPT domain was itself shut down in the full-length protein to permit effective PT domain function. Here, we show that this function is provided by the EF-hands. One finding that did not fit with our hypotheses was that the isolated EF1/2-Cl- SuPT-PT domain complex was capable of effectively localizing to the plasma membrane whether the EF-hands were intact, point mutated or deleted. Therefore, the isolated EF 1/2- Cl-SuPT-PT domain complex is not subject to negative regulation by the SuPT domain. This is consistent with previous work showing that the C1-SuPT-PT domain complex translocates effectively to the plasma membrane with BCR ligation in DT4O cells (Beaulieu et al, 2007). In fact, we found that the EF1/2-Cl-SuPT-PT domain complex only became dependent on intact EF-hands once the REM-GEF domain complex was also attached. This suggested that the combined actions of the REM-GEF domain complex and the SuPT 100 domain act to suppress PT domain function while EF-hands counteract this suppression to allow plasma membrane targeting. This process is shown in Fig. 6.1. EF-h SuPT --X-Q-Ec Figure 6.1 EF-hands permit PT domain function by shutting down the combined suppressive effects of the REM-GEF domain complex and SuPT domain REM, Ras exchange motif GEF, guanine nucleotide exchange factor domain; EF, EF-hands; Cl, Cl domain; SuPT, suppressor of PT; PT, plasma membrane targeter. The REM+GEF domain complex represses plasma membrane targeting of RasGRP 1 by hyper-activating the SuPT domain. The EF hands counteract SuPT-mediated suppression, thus enabling plasma membrane targeting. 6.3.1 Involvement of EF-hands in determining the cellular site of RasGRP1 activation. As depicted in Fig. 6.1, when the first EF-hand of RasGRP1 is disabled, the PT\u00E2\u0080\u00A2 domain is shut down due to suppression imposed by the combined effects of the SuPT domain and the REM-GEF domain complex. When the first EF-hand is functional, the SuPT domain is counteracted by the EF-hands, thus liberating the PT domain which then binds its target at the plasma membrane. Because the EF-hands counteract the SuPT domain, they are only relevant to PT and not Cl domain-mediated membrane localization. This is consistent 101 with previous findings showing that endomembrane localization and activation of RasGRP 1 in NIH 3T3 cells is possible even in the absence of EF-hands or SuPT/PT domains (Beaulieu et a!, 2007, Tognon et al, 1998). Therefore, any process that functionally disables the EF hands will switch RasGRP 1 from being capable of Cl +PT domain-mediated membrane targeting (via DAG and PT-ligand) to being capable of only Cl domain-mediated membrane targeting (via DAG). This implies that the EF-hands can dictate the cellular site by which RasGRP1 is activated (endomembranes and/or plasma membrane) depending on whether the EF-hands are functional or disabled. There are at least two ways in which RasGRP 1 can switch between these two states. The first is if the EF-hands are regulated by Ca2, such that they are functional when bound to Ca2 and disabled when not bound to Ca2. The second situation is if the EF-hands are disabled by a mechanism such as by alternative splicing. We functionally characterized one particular naturally occurring splice variant of RasGRP 1, lacking exon 11, a 35 amino acid segment which starts immediately C-terminal to the GEF domain and may encroach on the entering helix of the first EF-hand (Yasuda et al, 2007). We found that this splice variant was catalytically active and was able to localize to endomembranes as effectively as wild-type RasGRP 1 in NIH 3T3 cells. However, this splice variant of RasGRP 1 had BCR-induced activation defects and plasma membrane targeting defects, the latter of which was restored by deletion of the SuPT domain. Given that these effects resembled that of point mutations in the first EF-hand, this splice variant probably lacks a functional first EF-hand. This result supports our hypothesis that the EF hands can dictate the cellular site by which RasGRP1 is activated depending on whether the EF-hands are functional or disabled. 102 The importance of Ras signalling initiated in different membrane compartments can have crucial implications for cellular fate (Mor & Philips, 2006). Modulation of plasma membrane versus endomembranes localization of RasGRP 1 via its EF-hands, and by differential availability of DAG and/or the PT ligand could therefore provide a mechanism for fine tuning RasGRP 1 responses downstream of receptors. In a recent study in Jurkat cells, RasGRP 1 was reported to localize to either the plasma membrane or Golgi compartments with TCR stimulation, with plasma membrane targeting being dependent on co-stimulation with LFA-l (Mor et al, 2007). Furthermore, in primary T cells, RasGRP1 translocates to the plasma membrane with TCR and CD28 ligation but when the cells were previously induced to become anergic, the same stimulation causes RasGRP 1 to localize to endomembranes (Schwartz, 2003, Zha et al, 2006). The EF-hands could have a role in dictating plasma membrane targeting via the PT domain in these cell systems, with regulation by Ca2 binding being a possibility. Further work needs to be done to examine these possibilities. Furthermore, if the EF-hands are found to work as Ca2 sensors, then it would be interesting to examine the exact intra-molecular interactions between the REM GEF domain complex, SuPT domain and EF-hands that are involved in activating the PT domain. 6.4. Membrane localization and dependence on GEF domain activity In DT4O cells, the REM-GEF domain complex was found to cooperate with the SuPT domain in suppressing PT-mediated plasma membrane targeting in response to BCR ligation, and this was counteracted by the EF-hands. We became interested in determining the exact region within the REM-GEF domain complex that was involved in this EF-hand 103 dependent regulation of RasGRP 1. We expected that if the region within the REM-GEF domain complex needed for hyperactivating the SuPT domain was deleted, the BCR induced plasma membrane localization would occur even if the first EF-hand was mutated. However, we found that deletions in the REM or GEF domain shut down BCR-induced plasma membrane targeting of RasGRP 1 even when the EF-hands were not mutated. This unexpected finding then led us to hypothesize that catalytic activity of the GEF domain was involved in permitting membrane localization of RasGRP 1. To test this, we made two specific point mutations in the GEF domain of RasGRP 1 which were designed to disrupt Ras binding to the GEF catalytic site. Confirming our hypothesis, we found that neither of these mutants was capable of localizing to the plasma membrane in DT4O cells upon BCR ligation. We next hypothesized that Ras binding to the GEF domain was required to activate the EF-hands and thus counteract the SuPT domain. In this situation, deletion of the SuPT domain would enable plasma membrane targeting despite GEF mutation. However, this was not the case as deletion of the SuPT domain did not restore BCR-induced plasma membrane translocation of the catalytically inactive RasGRP1. Furthermore, endomembrane localization of RasGRP1 in NIH 3T3 cells was also disrupted by mutations preventing Ras binding to the GEF domain. Endomembrane localization in NIH 3T3 cells is mediated by the Cl domain and independent of the EF-hands and the SuPT/PT domains (Beaulieu et al, 2007, Tognon et al, 1998). Therefore, loss of Ras binding at the catalytic site disrupts membrane localization by a mechanism that is distinct from the EF-hand mediated mechanism that we had previously identified. To better understand this new mechanism, we addressed the possibility that loss of membrane binding by the catalytically inactive RasGRP1 was somehow a reflection of loss of Ras signaling. However, BCR-induced 104 plasma membrane translocation of the catalytically inactive RasGRP 1 was not restored in the presence of constitutive Ras signaling, which was induced by co-expression with a constitutively activate Ras mutant. Furthermore, BCR-induced plasma membrane translocation of wild-type RasGRP 1 was not impeded by inhibition of MAP kinase signaling, as would be expected if signaling downstream of Ras was somehow important for activating the membrane translocation of RasGRP 1. Therefore, it was not an absence of Ras signaling but likely an intrinsic defect that was causing a blockade in the membrane localization of RasGRP 1. As already mentioned, in the absence of Ras binding to the catalytic site of RasGRP1, the PT and Cl domains are shut down and unable to respond to their ligands. This presents a paradox: Ras binding to the GEF domain is essential for membrane binding by the Cl or PT domains, but membrane binding is essential for interaction of the GEF with Ras. However, activation could be possible under two circumstances, 1) the Cl and PT domains are not completely blocked when the GEF domain is inactive, and thus are occasionally able to contact membranes or 2) the GEF domain is able to come into contact with Ras in the absence of the Cl and PT domains. To test this latter hypothesis, we expressed the isolated REM-GEF domain complex of RasGRP1 in NIH 3T3 cells and determined if it could localize to membranes and be active by binding to Ras in the absence of the Cl and PT domains. We were not able to detect any membrane localization of this REM-GEF domain complex, even when this complex was dimerized to increase avidity. However, the isolated REM-GEF domain complex did induce a low level of transformation of NIH 3T3 cells, indicating that interaction with Ras was taking place. While we are not 105 ruling out the first possibility, that the Cl and PT domains are not completely inactivated in the absence of Ras binding to the catalytic domain, this result does demonstrate that RasGRP1 can encounter Ras even when the Cl and PT domains are completely disabled. Therefore, transient encounters of the REM-GEF domain complex with Ras may lead to transient liberation of the Cl and PT domains. If this occurs when either the Cl or PT ligands are present in membranes, the Cl and PT domains are then able to mediate membrane binding, which will cause increased Ras encounters, providing a positive feed back ioop mechanism that will accelerate both membrane binding and catalytic activity of RasGRP1. As an additional test of this mechanism, we introduced into the REM domain a mutation that is predicted to reduce but not eliminate Ras binding to the GEF domain. The design of this mutant was based on work done with SOS 1, which showed that binding of Ras-GTP to this site in the REM domain triggered positive allosteric regulation of catalytic activity, by enhancing binding of Ras to the GEF domain (Freedman et al, 2006, Margarit et a!, 2003). We hypothesized that if positive feedback was taking place, then reduced Ras binding to the GEF domain caused by mutations in the REM domain should result in a reduction but not elimination of plasma membrane targeting of RasGRP 1. In line with this hypothesis, this REM domain mutant was partially localized to internal membranes in NIH 3T3 cells and was weakly detected (in a small percentage of cells) at the plasma membrane in BCR-stimulated DT4O cells. In contrast, we previously saw that RasGRP1 was not localized to membranes in either of those cells when Ras binding to the GEF domain was eliminated. These results support the hypothesis that binding of Ras-GTP to the REM site 106 increases Ras binding to the catalytic site in the GEF domain which then promotes membrane binding by the Cl and/or PT domains. In this way, positive feedback from Ras GTP generation could play a critical role in the initiation and maintenance of RasGRP1 activation. For SOS 1, catalytic activation in lymphocytes was recently shown to require an initial priming via Ras-GTP binding to the REM domain, which then initiated a positive feedback ioop where full catalytic activation could then be achieved. Here, Ras-GTP needed to be supplied by an alternate GEF, namely RasGRP1 (Roose et al, 2007). Future experiments could determine if RasGRP 1 primes itself by generating Ras-GTP, or whether it depends on another GEF as seen for SOS 1. In conclusion, we propose the following model for RasGRP1 regulation via the REM and GEF domains (Fig. 6.2). Binding of Ras-GTP to the REM domain (1) increases access of Ras to the catalytic site within the GEF domain (by analogy to SOS 1). As well as enabling Ras-GTP generation by nucleotide exchange, Ras binding triggers liberation of the Cl and PT domains (2). If DAG and/or the PT ligand are present in a membrane, RasGRP1 is then bound to that membrane (3), thus increasing its encounters with Ras. The resulting increase in production of Ras-GTP increases REM occupancy, thus increasing Ras binding to the catalytic site, thus increasing availability of the Cl and PT domains for membrane binding. This activation process can be reversed either by removing Ras-GTP (4), or by removing the Cl and/or PT ligands (5), potentially returning RasGRP 1 to its fully inactivated state (6). 107 REM0-GEE0-[Cl]- [P1] ...$Ras-GTP (1) REMRasGTp GEF0-[Cl} [PT] Ras ()....____...._______... \u00E2\u0080\u00981ARaS:GTP:9EFZ9! -PTc DAG,PTIiandJ[ - - LRas-GTP Ras DAG L a- I as-OF- :2:5J}: J & DAG, PTIigand REMQ - GEFRas - Cl0 - PT0 (6)JF\u00E2\u0080\u0099 Ras REM0-GEE0-[Cl]- [PT] Figure 6.2 Model of regulation of RasGRP1 membrane localization and activation by the REM and GEF domains. Binding of Ras-GTP to the REM domain (1) increases the efficiency of Ras binding to the catalytic site within the GEF domain (2). This enables Ras-GTP generation by nucleotide exchange and functional liberation of the Cl and PT domains (2). If DAG and/or PT ligand are present in a membrane, RasGRP 1 is then stably bound to that membrane via the Cl and/or PT domains (3), thus increasing its encounters with membrane-bound Ras. The resulting high production of Ras-GTP increases REM binding to Ras-GTP, thus increasing Ras binding to the catalytic site and increasing availability of the Cl and PT domains for membrane binding. This increases Ras-GTP binding by the REM domain and Ras binding by the GEF domain in a positive feedback manner which can be reversed. Removal of Ras-GTP (4), or the Cl and/or PT ligands (5), potentially returns RasGRP I to its fully inactivated state (6). Subscripts indicate ligand occupancy, with 0 meaning unoccupied. The brackets around the Cl and PT domains indicate functional inactivation, via the unoccupied GEF domain. The active forms of RasGRP 1 (generating Ras-GTP) are boxed, with the solid box indicating high activity and the dashed box indicating low activity. 108 6.5. Implications for the regulation of other RasGRPs by EF-hands and Ca2 Are our findings on the mechanisms regulating RasGRP1 relevant to other RasGRPs? RasGRP3 contains a pair of EF-hands which, like those of RasGRP1 have the sequence features typical ofCa2-binding EF-hands. For RasGRP3, theCa2-binding affinity of the EF-hands has not been determined. Furthermore, there is no direct evidence indicating that the EF-hands of RasGRP3 are regulated by Ca2. One study reported that RasGRP3- dependent activation of Ras was not augmented by Ca2 ionophore treatment in rat2 cells (Lorenzo et al, 2001). Another group concluded that RasGRP3-dependent activation of Rap2B in HEK 293 cells stimulated with EGF receptor is mediated by Ca2,but this conclusion was based only on the observation that Rap2B activation requires Ca2 in these cells, without showing that this involves RasGRP3 (Stope et al, 2004). We found that BCR induced increase in pERK2 levels in untransduced DT4O cells, which is predominantly attributable to Ras activation via endogenous RasGRP3 (Oh-hora et al, 2003), was unaffected by BAPTA-AM treatment. These results suggest that as seen for RasGRP 1, RasGRP3 may not be regulated by the BCR-induced Ca2 flux. Like RasGRP1, RasGRP3 also contains a DAG responsive Cl domain (Johnson et a!, 2007) and contains sequences at its C-terminus that may be equivalent to the SuPT/PT domains of RasGRP 1 (Rob Kay lab, unpublished results). Therefore, there is potential for the EF-hands of RasGRP3 to function in the same manner as the EF-hands of RasGRP1, i.e. promoting membrane binding by counteracting the SuPT domain. RasGRP2 also contains a pair of EF-hands, each containing the required residues for Ca2coordination. The Ca2binding potential of the EF-hands of RasGRP2 have also not 109 been determined. However, there is evidence that RasGRP2 may be regulated by Ca2. Rap 1 A activation by two alternatively spliced isoforms of RasGRP2 was elevated with Ca2 ionophore treatment in both 293T cells and COS cells, whereas Ras activation by both of these isoforms of RasGRP2 was inhibited by the same treatment in COS cells (Clyde-Smith et al, 2000, Kawasaki et al, 1998). Therefore, it is possible that RasGRP2 is positively or negatively regulated by Ca2depending on whether it is acting as a Rap- or Ras-GEF. However, this indicates that the EF-hands and Ca2 are involved in catalytic specificity, rather than membrane localization as shown for RasGRP 1. By sequence inspection, RasGRP2 lacks sequence homology to the SuPTIPT domains of RasGRP 1. Therefore, there is no potential for the EF-hands in RasGRP2 to function in the same manner as the EF-hands in RasGRP 1 (with respect to counteracting the SuPT domain). RasGRP4 contains a single EF-hand that may not be able to bind Ca2 since it lacks residues required for canonical Ca2 coordination. However, it is still possible for this EF hand in RasGRP4 to bind Ca2 via a different coordination scheme (Gifford et al, 2007). As with RasGRP2 and 3, the Ca2binding potential of the EF-hand in RasGRP4 has not been determined. One study reported that RasGRP4-dependent activation of Ras was inhibited in the presence of 1 mM Ca2 (Yang et al, 2002). However, millimolar Ca2 is not physiological, and can induce changes in membrane structure that could artifactually inhibit exchange activity. RasGRP4 can bind DAG via its Cl domain (Johnson et al, 2007) but lacks sequence homology to the SuPT/PT domains of RasGRP 1. Therefore as with RasGRP2, there is no potential for the single EF-hand in RasGRP4 to be regulatory in the same manner as RasGRP1 (with respect to counteracting the SuPT domain). 110 An ancestral RasGRP-like gene passed on EF-hands to the now four RasGRP genes present in humans and mice. It is possible that the EF-hands of this ancestral gene encoded for Ca2 responsive EF-hands that were involved in a mechanism separate from the one identified for RasGRP1 since the SuPT/PT domains were absent. After gene duplication and acquisition of the SuPT/PT domain, the EF-hands of RasGRP 1 and perhaps those in RasGRP3 may have acquired the function of controlling PT domain-mediated plasma membrane localization via counteracting the SuPT domain and may have lost the ability to be regulated by Ca2. The EF-hands in RasGRP2 may have retained the ability to act as Ca2 sensors, the significance of which still needs to be determined. RasGRP4 may have started out with a pair of Ca2 regulated EF-hands but then possibly lost the requirement for EF hands, upon which subsequent sequence divergence resulting in loss of one of its EF-hands and functional inactivation of the other. 111 BIBLIOGRAPHY Alberola-Ila 3, Forbush KA, Seger R, Krebs EG, & Perimutter RM (1995) Selective requirement for MAP kinase activation in thymocyte differentiation. Nature 373: 620-623 Anolik JH (2007) B cell biology and dysfunction in SLE. Bull NYUHospJtDis 65: 182-186 Beaulieu N, Zahedi B, Goulding RE, Tazmini G, Anthony Ky, Omeis SL, de Jong DR, & Kay RJ (2007) Regulation of RasGRP 1 by B cell antigen receptor requires cooperativity between three domains controlling translocation to the plasma membrane. Mo! Rio! Cell 18: 3 156-3 168 Bernards A & Settleman J (2004) GAP control: regulating the regulators of small GTPases. Trends Ce!! Rio! 14: 377-385 Berridge MJ, Bootman MD, & Roderick HL (2003) Calcium signalling: dynamics, homeostasis and remodelling. Nat Rev Mol Ce!! Rio! 4: 5 17-529 Berridge MJ & Irvine RF (1984) Inositol trisphosphate, a novel second messenger in cellular signal transduction. Nature 312: 315-321 Berridge MJ, Lipp P, & Bootman MD (2000) The versatility and universality of calcium signalling. Nat Rev Mo! Ce!! Biol 1: 11-21 Bivona TG, Perez De Castro I, Ahearn TM, Grana TM, Chiu VK, Lockyer PJ, Cullen PJ, Pellicer A, Cox AD, & Philips MR (2003) Phospholipase Cgamma activates Ras on the Golgi apparatus by means of RasGRP1. Nature 424: 694-698 Bonfini L, Karlovich CA, Dasgupta C, & Banerjee U (1992) The Son of sevenless gene product: a putative activator of Ras. Science 255: 603-606 Boriack-Sjodin PA, Margarit SM, Bar-Sagi D, & Kuriyan J (1998) The structural basis of the activation of Ras by Sos. Nature 394: 337-343 Bos JL (1989) Ras Oncogenes in Human Cancer: a Review. Cancer Res 49: 4682-4689 Broek D, Toda T, Michaeli T, Levin L, Birchmeier C, Zoller M, Powers S, & Wigler M (1987) The S. cerevisiae CDC25 gene product regulates the RAS/adenylate cyclase pathway. Ce!! 48: 789-799 Buchsbaum R, Telliez JB, Goonesekera 5, & Feig LA (1996) The N-terminal pleckstrin, coiled-coil, and IQ domains of the exchange factor Ras-GRF act cooperatively to facilitate activation by calcium. Mo! Cell Bio! 16: 4888-4896 Buerstedde JM & Takeda S (1991) Increased ratio of targeted to random integration after transfection of chicken B cell lines. Cell 67: 179-188 112 Bums DJ & Bell RM (1991) Protein kinase C contains two phorbol ester binding domains. J Biol Chem 266: 18330-18338 Busch E, Hohenester E, Timpl R, Paulsson M, & Maurer P (2000) Calcium affinity, cooperativity, and domain interactions of extracellular EF-hands present in BM-40. JBiol Chem 275: 25508-255 15 Caloca MJ, Zugaza JL, & Bustelo XR (2003a) Exchange factors of the RasGRP family mediate Ras activation in the Golgi. JBiol Chem 278: 33465-33473 Caloca MJ, Zugaza JL, Matallanas D, Crespo P, & Bustelo XR (2003b) Vav mediates Ras stimulation by direct activation of the GDP/GTP exchange factor Ras GRP 1. EMBO J 22: 3326-3336 Carrasco S & Merida I (2004) Diacylglycerol-dependent binding recruits PKCtheta and RasGRP1 Cl domains to specific subcellular localizations in living T lymphocytes. Mol Biol Cell 15: 2932-2942 Chiu VK, Bivona T, Hach A, Sajous JB, Silletti J, Wiener H, Johnson RL,2nd, Cox AD, & Philips MR (2002) Ras signalling on the endoplasmic reticulum and the Golgi. Nat Cell Biol 4: 343-350 Clyde-Smith J, Sums G, Gartside M, Grimmond S, Etheridge M, Apolloni A, Hayward N, & Hancock JF (2000) Characterization of RasGRP2, a plasma membrane-targeted, dual specificity Ras/Rap exchange factor. JBiol Chem 275: 32260-32267 Colon-Gonzalez F & Kazanietz MG (2006) Cl domains exposed: from diacyiglycerol binding to protein-protein interactions. Biochim Biophys Acta 1761: 827-837 Coughlin JJ, Stang SL, Dower NA, & Stone JC (2005) RasGRP1 and RasGRP3 regulate B cell proliferation by facilitating B cell receptor-Ras signaling. Jlmmunol 175: 7179-7184 Cox AD & Der CJ (2002) Ras family signaling: therapeutic targeting. Cancer Biol Ther 1: 599-606 Crompton T, Gilmour KC, & Owen MJ (1996) The MAP kinase pathway controls differentiation from double-negative to double-positive thymocyte. Cell 86: 243-251 Cullen PJ & Lockyer PJ (2002) Integration of calcium and Ras signalling. Nat Rev Mol Cell Biol3: 339-348 Daniels MA, Teixeiro E, Gill J, Hausmann B, Roubaty D, Holmberg K, Werlen G, Hollander GA, Gascoigne NR, & Palmer E (2006) Thymic selection threshold defined by compartmentalization of Ras/MAPK signalling. Nature 444: 724-729 113 de Hoog CL, Fan WT, Goldstein MD, Moran MF, & Koch CA (2000) Calmodulin independent coordination of Ras and extracellular signal-regulated kinase activation by Ras GRF2. Mol Cell Biol 20: 2727-273 3 DeFranco AL (1997) The complexity of signaling pathways activated by the BCR. Curr Opin Immunol 9: 296-308 Dellis 0, Dedos SG, Tovey SC, Taufiq-Ur-Rahman, Dubel SJ, & Taylor CW (2006) Ca2+ entry through plasma membrane 1P3 receptors. Science 313: 229-23 3 Dolmetsch RE, Pajvani U, Fife K, Spotts JM, & Greenberg ME (2001) Signaling to the nucleus by an L-type calcium channel-calmodulin complex through the MAP kinase pathway. Science 294: 333-339 Dower NA, Stang SE, Bottorff DA, Ebinu JO, Dickie P, Ostergaard HE, & Stone JC (2000) RasGRP is essential for mouse thymocyte differentiation and TCR signaling. Nat Immunol 1: 317-321 Downward J, Graves JD, Warne PH, Rayter S, & Cantrell DA (1990) Stimulation ofp2lras upon T-cell activation. Nature 346: 719-723 Ebinu JO, BottorffDA, Chan EY, Stang SE, Dunn RJ, & Stone JC (1998) RasGRP, a Ras guanyl nucleotide- releasing protein with calcium- and diacylglycerol-binding motifs. Science 280: 1082-1086 Ebinu JO, Stang SE, Teixeira C, BottorffDA, Hooton J, Blumberg PM, Barry M, Bleakley RC, Ostergaard HE, & Stone JC (2000) RasGRP links T-cell receptor signaling to Ras. Blood 95: 3199-3203 Fam NP, Fan WT, Wang Z, Zhang U, Chen H, & Moran MF (1997) Cloning and characterization of Ras-GRF2, a novel guanine nucleotide exchange factor for Ras. Mol Cell Bioll7: 1396-1406 Fan WT, Koch CA, de Hoog CE, Fam NP, & Moran MF (1998) The exchange factor Ras GRF2 activates Ras-dependent and Rac-dependent mitogen-activated protein kinase pathways. Curr Biol 8: 935-93 8 Farnsworth CE, Freshney NW, Rosen LB, Ghosh A, Greenberg ME, & Feig EA (1995) Calcium activation of Ras mediated by neuronal exchange factor Ras-GRF. Nature 376: 524-527 Feske S (2007) Calcium signalling in lymphocyte activation and disease. Nat Rev Immunol 7: 690-702 Freedman TS, Sondermann H, Friedland GD, Kortemme T, Bar-Sagi D, Marqusee 5, & Kuriyan J (2006) A Ras-induced conformational switch in the Ras activator Son of sevenless. Proc Nati Acad Sci USA 103: 16692-16697 114 Fu HW & Casey PJ (1999) Enzymology and biology of CaaX protein prenylation. Recent Prog Horm Res 54: 315-42; discussion 342-3 Gale NW, Kaplan S, Lowenstein EJ, Schlessinger J, & Bar-Sagi D (1993) Grb2 mediates the EGF-dependent activation of guanine nucleotide exchange on Ras. Nature 363: 88-92 Gifford JL, Walsh MP, & Vogel HJ (2007) Structures and metal-ion-binding properties of the Ca2+-binding helix-loop-helix EF-hand motifs. Biochem J405: 199-22 1 Grabarek Z (2006) Structural basis for diversity of the EF-hand calcium-binding proteins. J Mol Biol 359: 5 09-525 Grynkiewicz G, Poenie M, & Tsien RY (1985) A new generation of Ca2+ indicators with greatly improved fluorescence properties. JBiol Chem 260: 3440-3450 Guilbault B & Kay RJ (2004) RasGRP1 sensitizes an immature B cell line to antigen receptor-induced apoptosis. JBiol Chem 279: 19523-19530 Hall BE, Yang SS, Boriack-Sjodin PA, Kuriyan J, & Bar-Sagi D (2001) Structure-based mutagenesis reveals distinct functions for Ras switch 1 and switch 2 in Sos-catalyzed guanine nucleotide exchange. JBiol Chem 276: 27629-27637 Hamad NM, Elconin JH, Karnoub AE, Bai W, Rich TN, Abraham RI, Der CJ, & Counter CM (2002) Distinct requirements for Ras oncogenesis in human versus mouse cells. Genes Dev 16: 2045-2057 Harwood AE & Cambier JC (1993) B cell antigen receptor cross-linking triggers rapid protein kinase C independent activation of p2lrasl. Jlmmunol 151: 45 13-4522 Hogquist KA (2001) Signal strength in thymic selection and lineage commitment. Curr Opin Immunol 13: 225-23 1 Hubbard SR, Bishop WR, Kirschmeier P, George SJ, Cramer SP, & Hendrickson WA (1991) Identification and characterization of zinc binding sites in protein kinase C. Science 254: 1776-1779 Hurley JH, Newton AC, Parker P3, Blumberg PM, & Nishizuka Y (1997) Taxonomy and function of Cl protein kinase C homology domains. Protein Sci 6: 477-480 Iritani BM, Alberola-Ila 3, Forbush KA, & Perimutter RM (1999) Distinct signals mediate maturation and allelic exclusion in lymphocyte progenitors. Immunity 10: 7 13-722 Iritani BM, Forbush KA, Farrar MA, & Perimutter RM (1997) Control of B cell development by Ras-mediated activation of Raf. EMBO J 16: 7019-7031 115 Johnson JE, Goulding RE, Ding Z, Partovi A, Anthony KV, Beaulieu N, Tazmini G, Cornell RB, & Kay Ri (2007) Differential membrane binding and diacyiglycerol recognition by Cl domains of RasGRPs. Biochem J406: 223-236 Jones DR, Sanjuan MA, Stone JC, & Merida I (2002) Expression of a catalytically inactive form of diacyiglycerol kinase alpha induces sustained signaling through RasGRP. FASEB J 16: 595-597 Jones S, Vignais ML, & Broach JR (1991) The CDC25 protein of Saccharomyces cerevisiae promotes exchange of guanine nucleotides bound to ras. Mol Cell Biol 11: 2641-2646 Katsoulotos GP, Qi M, Qi JC, Tanaka K, Hughes WE, Molloy Ti, Adachi R, Stevens RL, & Krilis SA (2007) The diacylglycerol dependent translocation of Ras guanine nucleotide releasing protein 4 (RASGRP4) inside a human mast cell line results in substantial phenotypic changes including expression of the interleukin 13 receptor ILl3Ralpha 2. JBiol Chem Kawasaki H, Springett GM, Toki 5, Canales JJ, Harlan P, Blumenstiel JP, Chen EJ, Bany IA, Mochizuki N, Ashbacher A, Matsuda M, Housman DE, & Graybiel AM (1998) A Rap guanine nucleotide exchange factor enriched highly in the basal ganglia. Proc Nati Acad Sd USA 95: 13278-13283 Klinger MB, Guilbault B, Goulding RE, & Kay RJ (2005) Deregulated expression of RasGRP1 initiates thymic lymphomagenesis independently of T-cell receptors. Oncogene 24: 2695-2704 Kosco KA, Cerignoli F, Williams 5, Abraham RT, & Mustelin T (2008) SKAP55 modulates T cell antigen receptor-induced activation of the Ras-Erk-AP1 pathway by binding RasGRP1. MolImmunol45: 5 10-522 Kretsinger RH & Nockolds CE (1973) Carp muscle calcium-binding protein. II. Structure determination and general description. JBiol Chem 248: 3313-3326 Lai CC, Boguski M, Broek D, & Powers S (1993) Influence of guanine nucleotides on complex formation between Ras and CDC25 proteins. Mol Cell Biol 13: 1345-13 52 Layer K, Lin G, Nencioni A, Hu W, Schmucker A, Antov AN, Li X, Takamatsu 5, Chevassut T, Dower NA, Stang SL, Beier D, Buhlmann J, Bronson RT, Elkon KB, Stone JC, Van Parijs L, & Lim B (2003) Autoimmunity as the consequence of a spontaneous mutation in Rasgrpl. Immunity 19: 243-255 Leicht DT, Balan V, Kaplun A, Singh-Gupta V, Kaplun L, Dobson M, & Tzivion G (2007) Rafkinases: function, regulation and role in human cancer. Biochim Biophys Acta 1773: 1196-1212 116 Lenzen C, Cool RH, Prinz H, Kuhimarm J, & Wittinghofer A (1998) Kinetic analysis by fluorescence of the interaction between Ras and the catalytic domain of the guanine nucleotide exchange factor Cdc25Mm. Biochemistry 37: 7420-7430 Li J, Shen H, Himmel KL, Dupuy AJ, Largaespada DA, Nakamura T, Shaughnessy JD,Jr, Jenkins NA, & Copeland NG (1999) Leukaemia disease genes: large-scale cloning and pathway predictions. Nat Genet 23: 348-353 Li L, Yang Y, Wong GW, & Stevens RL (2003) Mast cells in airway hyporesponsive C3H/HeJ mice express a unique isoform of the signaling protein Ras guanine nucleotide releasing protein 4 that is unresponsive to diacyiglycerol and phorbol esters. Jlmmunol 171: 390-397 Lorenzo PS, Beheshti M, Pettit GR, Stone JC, & Blumberg PM (2000) The guanine nucleotide exchange factor RasGRP is a high -affinity target for diacylglycerol and phorbol esters. Mo? Pharmacol 57: 840-846 Lorenzo PS, Kung JW, Bottorff DA, Garfield SH, Stone JC, & Blumberg PM (2001) Phorbol esters modulate the Ras exchange factor RasGRP3. Cancer Res 61: 943-949 Malumbres M & Barbacid M (2003) RAS oncogenes: the first 30 years. Nat Rev Cancer 3: 459-465 Margarit SM, Sondermann H, Hall BE, Nagar B, Hoelz A, Pirruccello M, Bar-Sagi D, & Kuriyan J (2003) Structural evidence for feedback activation by Ras.GTP of the Ras-specific nucleotide exchange factor SOS. Ce?? 112: 685-695 Mitin N, Rossman KL, & Der CJ (2005) Signaling interplay in Ras superfamily function. Curr Bio? 15: R563-74 Mor A, Campi G, Du G, Zheng Y, Foster DA, Dustin ML, & Philips MR (2007) The lymphocyte function-associated antigen-i receptor costimulates plasma membrane Ras via phospholipase D2. Nat Cell Rio? 9: 713-7 19 Mor A & Philips MR (2006) Compartmentalized Ras/MAPK signaling. Annu Rev Immunol 24: 771-800 Mor A, Philips MR, & Pillinger MH (2007) The role of Ras signaling in lupus T lymphocytes: biology and pathogenesis. Clin Immunol 125: 2 15-223 Nagaoka H, Yu W, & Nussenzweig MC (2000) Regulation of RAG expression in developing lymphocytes. Curr Opin Immuno? 12: 187-190 Nel AE, Gupta S, Lee L, Ledbetter JA, & Kanner SB (1995) Ligation of the T-cell antigen receptor (TCR) induces association of hSosi, ZAP-70, phospholipase C-gamma 1, and other phosphoproteins with Grb2 and the zeta-chain of the TCR. JBiol Chem 270: 18428-18436 117 Newton AC (1995) Protein kinase C: structure, function, and regulation. JBiol Chem 270: 28495-28498 Nishida M, Sugimoto K, Hara Y, Mon E, Morii T, Kurosaki T, & Mon Y (2003) Amplification of receptor signalling by Ca2+ entry-mediated translocation and activation of PLCgamma2 in B lymphocytes. EMBO J22: 4677-4688 Noh DY, Shin SH, & Rhee SG (1995) Phosphoinositide-specific phospholipase C and mitogenic signaling. Biochim Biophys Acta 1242: 99-113 Norment AM, Bogatzki LY, Klinger M, Ojala EW, Bevan MJ, & Kay RJ (2003) Transgenic expression of RasGRP 1 induces the maturation of double-negative thymocytes and enhances the production of CD8 single-positive thymocytes. Jlmmunol 170: 1141-1 149 Oh-hora M, Johmura S, Hashimoto A, Hikida M, & Kurosaki T (2003) Requirement for Ras guanine nucleotide releasing protein 3 in coupling phospholipase C-gamma2 to Ras in B cell receptor signaling. JExp Med 198: 1841-185 1 Ono Y, Fujii T, Igarashi K, Kuno I, Tanaka C, Kikkawa U, & Nishizuka Y (1989) Phorbol ester binding to protein kinase C requires a cysteine-rich zinc-finger-like sequence. Proc NatlAcadSci USA 86: 4868-4871 O\u00E2\u0080\u0099Shea CC, Crompton T, Rosewell IR, Hayday AC, & Owen MJ (1996) Raf regulates positive selection. Eur Jlmmunol 26: 2350-2355 Park W, Mosteller RD, & Broek D (1994) Amino acid residues in the CDC25 guanine nucleotide exchange factor critical for interaction with Ras. Mo? Cell Biol 14: 8117-8122 Pear WS, Nolan GP, Scott ML, & Baltimore D (1993) Production of high-titer helper-free retroviruses by transient transfection. Proc NatlAcad Sci USA 90: 8392-8396 Perez de Castro I, Bivona TG, Philips MR, & Pellicer A (2004) Ras activation in Jurkat T cells following low-grade stimulation of the T-cell receptor is specific to N-Ras and occurs only on the Golgi apparatus. Mo? Cell Biol 24: 3485-3496 Priatel JJ, Teh SJ, Dower NA, Stone JC, & Teh HS (2002) RasGRP1 transduces low-grade TCR signals which are critical for T cell development, homeostasis, and differentiation. Immunity 17: 617-627 Putney JW,Jr (2007) Recent breakthroughs in the molecular mechanism of capacitative calcium entry (with thoughts on how we got here). Ce?? Calcium 42: 103-110 Quest AF, Bardes ES, & Bell RM (1994) A phorbol ester binding domain of protein kinase C gamma. High affinity binding to a glutathione-S-transferase/Cys2 fusion protein. JBiol Chem 269: 2953-2960 118 Quilliam LA (2007) New insights into the mechanisms of SOS activation. Sd STKE 2007: pe67 Quilliam LA, Rebhun JF, & Castro AF (2002) A growing family of guanine nucleotide exchange factors is responsible for activation of Ras-family GTPases. Prog Nucleic Acid Res Mol Biol 71: 39 1-444 Rajalingam K, Schreck R, Rapp UR, & Albert S (2007) Ras oncogenes and their downstream targets. Biochim Biophys Acta 1773: 1177-1 195 Repasky GA, Chenette EJ, & Der CJ (2004) Renewing the conspiracy theory debate: does Raf function alone to mediate Ras oncogenesis? Trends Cell Biol 14: 63 9-647 Reuther GW, Lambert QT, Rebhun JF, Caligiuri MA, Quilliam LA, & Der CJ (2002) RasGRP4 is a novel Ras activator isolated from acute myeloid leukemia. JBiol Chem 277: 30508-30514 Rintoul GL, Raymond LA, & Baimbridge KG (2001) Calcium buffering and protection from excitotoxic cell death by exogenous calbindin-D28k in HEK 293 cells. Cell Calcium 29: 277-287 Roose JP, Mollenauer M, Gupta VA, Stone J, & Weiss A (2005) A diacylglycerol-protein kinase C-RasGRP 1 pathway directs Ras activation upon antigen receptor stimulation of T cells. Mol Cell Biol 25: 4426-444 1 Roose JP, Mollenauer M, Ho M, Kurosaki T, & Weiss A (2007) Unusual interplay of two types of Ras activators, RasGRP and SOS, establishes sensitive and robust Ras activation in lymphocytes. Mol Cell Biol 27: 2732-2745 Sanjuan MA, Pradet-Balade B, Jones DR, Martinez-A C, Stone JC, Garcia-Sanz JA, & Merida I (2003) T cell activation in vivo targets diacyiglycerol kinase alpha to the membrane: a novel mechanism for Ras attenuation. Jlmmunol 170: 2877-2883 Saxton TM, van Oostveen I, Bowtell D, Aebersold R, & Gold MR (1994) B cell antigen receptor cross-linking induces phosphorylation of the p2lras oncoprotein activators SHC and mSOS1 as well as assembly of complexes containing SHC, GRB-2, mSOS1, and a 145- kDa tyrosine-phosphorylated protein. Jlmmunol 153: 623-636 Schlessinger J (2000) Cell signaling by receptor tyrosine kinases. Cell 103: 211-225 Schwartz RH (2003) T cell anergy. Annu Rev Immunol 21: 305-334 Shaw AC, Swat W, Ferrini R, Davidson L, & Alt FW (1999) Activated Ras signals developmental progression of recombinase-activating gene (RAG)-deficient pro-B lymphocytes. JExp Med 189: 123-129 119 Simon MA, Bowtell DD, Dodson GS, Laverty TR, & Rubin GM (1991) Rasi and a putative guanine nucleotide exchange factor perform crucial steps in signaling by the sevenless protein tyrosine kinase. Cell 67: 70 1-716 Sondermann H, Soisson SM, Boykevisch 5, Yang SS, Bar-Sagi D, & Kuriyan J (2004) Structural analysis of autoinhibition in the Ras activator Son of sevenless. Cell 119: 393-405 Starr TK, Jameson SC, & Hogquist KA (2003) Positive and negative selection of T cells. Annu Rev Immunol 21: 139-176 Stope MB, Vom Dorp F, Szatkowski D, Bohm A, Keiper M, Nolte J, Oude Weernink PA, RosskopfD, Evellin 5, Jakobs KH, & Schmidt M (2004) Rap2B-dependent stimulation of phospholipase C-epsilon by epidermal growth factor receptor mediated by c-Src phosphorylation of RasGRP3. Mol Cell Biol 24: 4664-4676 Swan KA, Alberola-Ila J, Gross JA, Appleby MW, Forbush KA, Thomas JF, & Perimutter RIvI (1995) Involvement ofp2lras distinguishes positive and negative selection in thymocytes. EMBO J 14: 276-2 85 Swat W, Shinkai Y, Cheng HL, Davidson L, & Alt FW (1996) Activated Ras signals differentiation and expansion of CD4+8+ thymocytes. Proc Natl Acad Sci US A 93: 4683- 4687 Takata M, Honima Y, & Kurosaki T (1995) Requirement of phospholipase C-gamma 2 activation in surface immunoglobulin M-induced B cell apoptosis. JExp Med 182: 907-9 14 Taylor SJ, Resnick RJ, & Shalloway D (2001) Nonradioactive determination of Ras-GTP levels using activated ras interaction assay. Methods Enzymol 333: 333-342 Tian X, Gotoh T, Tsuji K, Lo EH, Huang 5, & Feig LA (2004) Developmentally regulated role for Ras-GRFs in coupling NMDA glutamate receptors to Ras, Erk and CREB. EMBO J 23: 1567-1575 Tognon CE, Kirk HE, Passmore LA, Whitehead IP, Der CJ, & Kay RJ (1998) Regulation of RasGRP via a phorbol ester-responsive Cl domain. Mol Cell Biol 18: 6995-7008 Topham MK & Prescott SM (2001) Diacylglycerol kinase zeta regulates Ras activation by a novel mechanism. JCe11 Biol 152: 1135-1143 Tsien RY (1981) A non-disruptive technique for loading calcium buffers and indicators into cells. Nature 290: 527-528 Tsien RY, Pozzan T, & Rink TJ (1984) Calcium activities and fluxes inside small intact cells as measured with intracellularly trapped chelators. Adv Cyclic Nucleotide Protein Phosphorylation Res 17: 535-541 120 Turjanski AG, Vaque JP, & Gutkind JS (2007) MAP kinases and the control of nuclear events. Oncogene 26: 3240-3253 Vetter JR & Wittinghofer A (2001) The guanine nucleotide-binding switch in three dimensions. Science 294: 1299-1304 Vivanco J & Sawyers CL (2002) The phosphatidylinositol 3-Kinase AKT pathway in human cancer. Nat Rev Cancer 2: 489-50 1 von Boehn-ier H & Fehling HJ (1997) Structure and function of the pre-T cell receptor. Annu Rev Immunol 15: 43 3-452 Wennerberg K, Rossman KL, & Der CJ (2005) The Ras superfamily at a glance. J Cell Sci 118: 843-846 Winding P & Berchtold MW (2001) The chicken B cell line DT4O: a novel tool for gene disruption experiments. Jlmmunol Methods 249: 1-16 Wittinghofer A & Nassar N (1996) How Ras-related proteins talk to their effectors. Trends Biochem Sd 21: 488-491 Yamashita 5, Mochizuki N, Ohba Y, Tobiume M, Okada Y, Sawa H, Nagashima K, & Matsuda M (2000) Ca1DAG-GEFIII activation of Ras, R-ras, and Rap 1. J Biol Chem 275: 25488-25493 Yang Y, Li L, Wong GW, Krilis SA, Madhusudhan MS, Sali A, & Stevens RL (2002) RasGRP4, a new mast cell-restricted Ras guanine nucleotide-releasing protein with calcium- and diacyiglycerol-binding motifs. Identification of defective variants of this signaling protein in asthma, mastocytosis, and mast cell leukemia patients and demonstration of the importance of RasGRP4 in mast cell development and function. JBiol Chem 277: 25756- 25774 Yap KL, Ames JB, Swindells MB, & Ikura M (1999) Diversity of conformational states and changes within the EF-hand protein superfamily. Proteins 37: 499-507 Yasuda S, Stevens RL, Terada T, Takeda M, Hashimoto T, Fukae J, Horita T, Kataoka H, Atsumi T, & Koike T (2007) Defective expression of Ras guanyl nucleotide-releasing protein 1 in a subset of patients with systemic lupus erythematosus. Jlmmunol 179: 4890- 4900 Zha Y, Marks R, Ho AW, Peterson AC, Janardhan 5, Brown I, Praveen K, Stang 5, Stone JC, & Gajewski TF (2006) T cell anergy is reversed by active Ras and is regulated by diacyiglycerol kinase-alpha. Nat Immunol 7: 1166-1173 Zhang G, Kazanietz MG, Blumberg PM, & Hurley JH (1995) Crystal structure of the cys2 activator-binding domain of protein kinase C delta in complex with phorbol ester. Cell 81: 9 17-924 121 Zhang M, Srivastava G, & Lu L (2004) The pre-B cell receptor and its function during B cell development. Cell Mol Immunol 1: 89-94 Zhao C, Du G, Skowronek K, Frohman MA, & Bar-Sagi D (2007) Phospholipase D2- generated phosphatidic acid couples EGFR stimulation to Ras activation by Sos. Nat Cell Biol9: 706-712 Zheng Y, Liu H, Coughlin J, Zheng J, Li L, & Stone JC (2005) Phosphorylation of RasGRP3 on threonine 133 provides a mechanistic link between PKC and Ras signaling systems in B cells. Blood 105: 3648-3654 Zhong XP, Hainey EA, Olenchock BA, Zhao H, Topham MK, & Koretzky GA (2002) Regulation of T cell receptor-induced activation of the Ras-ERK pathway by diacyiglycerol kinase zeta. JBiol Chem 277: 31089-31098 Zugaza JL, Caloca MJ, & Bustelo XR (2004) Inverted signaling hierarchy between RAS and RAC in T-lymphocytes. Oncogene 23: 5823-5833 122 APPENDICES Appendix A. UBC Laboratory Biological Safety Certificate 123 \u00E2\u0080\u0098 /1 T H E U N IV E R S IT Y O F B R IT IS H C O L U M B IA D ep ar tm en t o f H ea lth , S af et y an d. En vi ro nm en t qn az ai th \u00E2\u0080\u0098 Ta zm iii i ha s su cc es sf ul ly co m pl et ed a co u rs e in L ab or at or y B io lo gi ca l S af et y. C ha ir, B io sa fe ty C om m itt ee B io s\u00C3\u00A0 re ty O ff ic er Tu es da y, M ay 04 , 2 00 4 D ire cf or H ea lth , S af et y an d E nv ir on m en t"@en . "Thesis/Dissertation"@en . "2008-11"@en . "10.14288/1.0066833"@en . "eng"@en . "Genetics"@en . "Vancouver : University of British Columbia Library"@en . "University of British Columbia"@en . "Attribution-NonCommercial-NoDerivatives 4.0 International"@en . "http://creativecommons.org/licenses/by-nc-nd/4.0/"@en . "Graduate"@en . "Regulatory mechanisms of the exchange factor RasGRP1"@en . "Text"@en . "http://hdl.handle.net/2429/2861"@en .