UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Characterizing the regulation of RasGRPI and CalDag1 : two C1 domain-containing ras guanine nucleotide… Anthony, Kira 2002

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

Item Metadata


831-ubc_2002-0326.pdf [ 10.82MB ]
JSON: 831-1.0090353.json
JSON-LD: 831-1.0090353-ld.json
RDF/XML (Pretty): 831-1.0090353-rdf.xml
RDF/JSON: 831-1.0090353-rdf.json
Turtle: 831-1.0090353-turtle.txt
N-Triples: 831-1.0090353-rdf-ntriples.txt
Original Record: 831-1.0090353-source.json
Full Text

Full Text

Characterizing the Regulation of RasGRPl and C a l D A G l : Two CI Domain-Containing Ras Guanine Nucleotide Exchange Factors by K I R A A N T H O N Y B . Sc. (Biochemistry), McMaster University, 1998 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES DEPARTMENT OF MEDICAL GENETICS MEDICAL GENETICS GRADUATE PROGRAM We accept this thesis as conforming to the required standard T H E U N I V E R S I T Y OF B R I T I S H C O L U M B I A October 2002 © Ki ra Anthony, 2002 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of tA e <kco \ C-<t^<~£Xc^ s The University of British Columbia Vancouver, Canada Date (Orb - ^  «?O0 9-DE-6 (2/88) ABSTRACT RasGRPI and CalDAGI are two guanine nucleotide exchange factors (GEFs) for the Ras family of GTPases that share a high degree of sequence identity. However, RasGRPI is a GEF for the classical Ras GTPases and CALDAG1 activates Rap1. These effector specificity differences suggest that there may be specific differences in the regulation of each of these GEFs. Activation of Ras occurs via recruitment of its GEFs to membranes, where the C1 domains of RasGRPI and CalDAGI are the most likely candidates to mediate this recruitment. The C1 domains of these GEFs are homologous to those of PKCs and may be classified into typical or atypical depending on whether or not they can bind diacylglycerol (DAG) and highly potent DAG analogues such as phorbol esters. Within both NIH 3T3 cells and T28 cells, it was determined that the C1 domain of RasGRPI mediated its translocation to membranes in response to DAG and phorbol esters. Conversely, the isolated CalDAGI C1 was not recruited to the plasma membrane in response to either phorbol ester stimulation in NIH 3T3 cells or T cell receptor stimulation in T28 cells. This indicates that the CalDAGI C1 domain may not be a direct target of DAG in NIH 3T3 cells and that the RasGRPI and CalDAGI C1 domains are biochemically distinct. Nonetheless, both RasGRPI and CalDAGI demonstrated an equivalent C1 domain-dependent plasma membrane localization pattern upon stimulation of the T cell receptor with the natural ligand anti-CD3s in T28 cells. Unlike CalDAGI, RasGRPI contains an alpha (a) helix motif that is predicted to dimerize to form a leucine zipper. In order to test whether or not RasGRPI homodimerizes at the a helix both co-immunoprecipitation and a modified colocalization assay were employed. Taken together, the results from each of these two approaches ii indicate that RasGRPI homodimerizes. Furthermore, although the results are compatible with the mediation of this protein-protein interaction by the RasGRPI helix, they do not unambiguously prove this hypothesis. iii TABLE OF CONTENTS ABSTRACT TABLE OF CONTENTS iv LIST OF TABLES vi LIST OF FIGURES vii LIST OF ABBREVIATIONS ix ACKNOWLEDGEMENTS xii CHAPTER 1 INTRODUCTION 1 1.1 The RasGRP subfamily of Ras GEFs 1 1.2 Structure of the RasGRP subfamily 4 1.2.1 REM box and GEF domain 6 1.2.2 C1 domain 6 1.2.3 EF hands 9 1.2.4 Amphipathic a helix 11 1.3 RasGRP subfamily GEF activity 12 1.3.1 Activation of Ras GTPases by the RasGRP subfamily 14 1.3.2 Regulation of Ras GTPase Activation by the RasGRP subfamily 16 1.3.3 Activation of Ras GTPase downstream effectors by the RasGRP subfamily 20 1.4 Cellular Processes Influenced by RasGRP subfamily members 24 1.4.1 Morphological Transformation 24 1.4.2 PLC-coupled Receptor Signalling 27 CalDAGl in Mi mAChR Signalling 27 RasGRPl in TCR Signalling 29 1.4.3 RasGRPl Involvement in Late-Stage Thymopoiesis 33 1.5 Thesis Objectives 35 CHAPTER 2 INVOLVEMENT OF THE RASGRP1 a HELIX/LEUCINE ZIPPER MOTIF IN HOMODIMERIZATION 37 2.1 Introduction 37 2.2 Methods 40 2.2.1 Cell culture 40 2.2.2 Construction of NH2-terminally truncated RasGRPl constructs 40 2.2.3 Production of Viruses by Transfection of BOSC 23 cells 41 2.2.4 Viral Infection of NIH 3T3 cells 42 2.2.5 Cell solubilization, Immunoprecipitation and Immunoblotting 42 I V 2.2.6 Fluorescence and Microscopy using GFP 43 2.3 Results 45 2.3.1 An NH2-terminally truncated RasGRPI construct co-immunoprecipitates with full-length RasGRPI 45 2.3.2 NH2-terminally truncated RasGRPI constructs alter the subcellular localization of a RasGRPI construct 47 2.4 Discussion 52 CHAPTER 3 COMPARISON OF RASGRP1 AND CALDAG1 C1 DOMAIN-MEDIATED MEMBRANE RECRUITMENT 55 3.1 Introduction 55 3.2 Methods 60 3.2.1 Materials 60 3.2.2 Cell culture 60 3.2.3 Cloning of murine CalDAGI 60 3.2.4 Construction of RasGRPI and CalDAGI deletion and mutant constructs 61 3.2.5 Viral Infection of T28 cells 61 3.2.6 Fluorescence and Microscopy using GFP in T28 cells 62 3.3 Results 63 3.3.1 Comparison of C1 domain sequences 63 3.3.2 Comparison of RasGRPI and CalDAGI C1 domain-mediated membrane recruitment in NIH 3T3 cells 65 RasGRPI and CalDAGI demonstrate different translocation patterns in response to the phorbol ester PMA in NIH 3T3 cells 65 RasGRPI and CalDAGI C1 domain translocation patterns in response to PC-PLC treatment or DiC8 stimulation in NIH 3T3 cells 69 3.3.3 Comparison of RasGRPI and CalDAGI C1 domain-mediated membrane translocation patterns in T28 cells 72 The RasGRPI and CalDAGI C1 domains demonstrate distinct plasma membrane localization patterns in response to PMA in T28 cells 73 The RasGRPI C1 domain localizes to the plasma membrane following TCR crosslinking, while the CalDAGI C1 domain does not 74 Both RasGRPI and CalDAGI localize to the plasma membrane following stimulation of the TCR 75 3.4 Discussion 78 CHAPTER 4 SUMMARY AND PERSPECTIVES 87 REFERENCES 90 V LIST OF TABLES CHAPTER 1 No Tables CHAPTER 2 Table 2.1 Proportion of relocalization upon coexpression of GFP-tagged RasGRPIdC1 with prenylated NH2-terminally truncated RasGRPI constructs in NIH 3T3 cells CHAPTER 3 No Tables CHAPTER 4 No Tables LIST OF FIGURES CHAPTER 1 Figure 1.1 Figure 1.2 CHAPTER 2 Figure 2.1 Figure 2.2 Figure 2.3 CHAPTER 3 Figure 3.1 Figure 3.2 Figure 3.3 Figure 3.4 Figure 3.5 Figure 3.6 Figure 3.7 Model for membrane recruitment-regulated activation of the RasGRP subfamily 3 Structure of RasGRP subfamily of Ras GEFs 5 Comparison of the RasGRPl a helix sequence with leucine zipper consensus sequences 38 The RasGRPl sp/a construct co-immunoprecipitates with full-length RasGRPl 46 The prenylated NH2-terminally truncated RasGRPl constructs alter the subcellular localization of a GFP-tagged RasGRPl construct 50 Comparison of C1 domain sequences 64 The CalDAGl C1 domain does not translocate to the plasma membrane in response to PMA stimulation in NIH 3T3 cells 67 The RasGRPl and CalDAGl C1 domains do not translocate to the plasma membrane in response to phorbol 12,13-dibutyrate in NIH 3T3 cells 69 The CalDAGl C1 domain does not respond to DAG-generating PC-PLC treatment in NIH 3T3 cells 71 The RasGRPl and CalDAGl C1 domains do not translocate to the plasma membrane in response to 1,2 dioctanoyl-sn-glycerol stimulation in NIH 3T3 cells 72 Both the RasGRPl and CalDAGl C1 domains translocate to the plasma membrane in response to PMA stimulation in T28 cells 74 The RasGRPl C1 domain translocates to the plasma membrane in response to anti-CD3s stimulation in T28 cells 75 vii Figure 3.8 C1 domain-dependent plasma membrane translocation of RasGRPl and CalDAGl in ; response to anti-CD3e stimulation in T28 cells CHAPTER 4 No Figures V l l l LIST OF ABBREVIATIONS aa amino acid ADP adenosine triphosphate BCR B cell antigen receptor BSA bovine serum albumin C1 protein kinase C homology-1 CalDAGI calcium- and diacylglycerol-regulated GEF1 cDNA cloned deoxyribonucleic acid CS calf serum CTP cytosine triphosphate DAG 1,2 diacyl-s/i-glycerol DGK diacylglycerol kinase DiC8 1,2 dioctanoyl-sn-glycerol DMEM Dulbecco's modified Eagle's medium DN double negative DP double positive EDTA ethylenediaminetetraacetic acid ER endoplasmic reticulum ERK extracellular signal-regulated kinase FBS fetal bovine serum FL full length GAP GTPase-activating protein GDP guanosine diphosphate GEF guanine nucleotide exchange factor ix GFP green fluorescent protein GNBP guanine nucleotide binding protein GPCR G protein-coupled receptor GTP guanosine triphosphate HA hemagglutinin IB immunoblot ig immunoglobulin IL interleukin IP immunoprecipitation IPs inositol triphosphate JNK c-Jun NH2-terminal kinase LAT linker for activated T cells mAChR muscarinic acetylcholine receptor MAPK mitogen activated protein kinase MEK MAPK/ERK kinase MHC major histocompatibility complex NA numerical aperture PBS phosphate buffered saline PC phosphatidylcholine PCR polymerase chain reaction PDBu phorbol 12,13-dibutyrate PE phorbol ester PI3-K phosphatidylinositol 3'-kinase PIP 2 phosphatidylinositol 4,5 bisphosphate PKC protein kinase C X PKD protein kinase D PLC phospholipase C PMA phorbol 12-myristate 13-acetate PSB phosphorylation solubilization buffer PTK protein tyrosine kinase RasGRP Ras guanine nucleotide releasing protein RBD Ras binding domain REM Ras exchange motif Rsk ribosomal S6 kinase RTK receptor tyrosine kinase SAPK stress-activated protein kinase SCR structurally conserved region SDS sodium dodecyl sulphate Sos son of sevenless SP single positive TCR T cell antigen receptor X I ACKNOWLEDGEMENTS This thesis could not have been completed without the help and the wisdom that I acquired from a number of people. First and foremost, I would like to sincerely thank my supervisor, Rob Kay, for his assistance, support and generosity, and for giving up valuable time to correct this thesis. Rob taught me how to identify problems, provided me with the tools necessary to analyze my data and, more importantly, his comments often inspired me to tackle a problem from a different direction. I am grateful to him for allowing me the opportunity to be a member of his outstanding group. Cristina Tognon deserves tremendous recognition for offering her knowledge, skills and patience. She always found the time to work through a variety of protocols with me and helped me comprehend my results. I had the pleasure to work alongside many talented and fun-loving colleagues. Thanks to Heather, Mark, Benoit, Nadine, Rebecca and Shane for sharing their knowledge and skills, and for making this experience a memorable one. I am also grateful for their friendship over the years and I think it is this, which facilitated their gracious tolerance of my extensive bench space. On a technical note, thank you to Rob and Cristina for making a number of the constructs used within this thesis and to Rob, Cristina and Heather for cloning RasGRPI and CalDAGI. Additionally, there are others at the Terry Fox Laboratory who deserve my appreciation. Liz Chavez spent numerous hours assisting me with the fluorescent microscope, and Gayle, Rick, Giovanna and Cam sorted countless cells for me. Lastly, I would also like to acknowledge Andrew Ryder at UBC for teaching me how to use SPSS statistical software to analyze my data. ix Last, but certainly not least, I would also like to thank some of those closest to me. I had the opportunity to meet many new people in graduate school, many of whom will remain close friends of mine. To all my friends, both old and new, I would like to sincerely thank you for all the good times we've had so far and for those yet to come. Thank you for your love and support. A special thanks goes out to mom, dad and Nat for supporting me over distance and especially over time. And finally, thank you Matt for continually challenging me, for remaining my biggest fan throughout and for editing this bloody thing. Chapter 1 INTRODUCTION 1.1 THE RASGRP SUBFAMILY OF RAS GEFs The RasGRP subfamily belongs to the family of Ras guanine nucleotide exchange factors (GEFs). The principal intracellular function of Ras GEFs is to activate the Ras family of GTPases, proteins that play an integral role in relaying signals from the surface of a cell to the nucleus. The specific Ras GEF role in mediating effector protein activation is to promote the dissociation of GDP from Ras GTPases such that GTP may bind to these guanine nucleotide binding proteins (GNBPs) and activate them (Boguski and McCormick 1993; Overbeck et al., 1995; Quilliam et al., 1995). In order to do so, each Ras GEF contains a catalytic motif, which is a GEF domain whose mechanism of action is common to all Ras GEFs (Vetter and Wittinghofer, 2001). Following the discernment of Ras GEF catalytic activity, the regulation of Ras GEF function has garnered much attention. The principal hypothesis for the regulation of Ras GEF function is based upon the conviction that recruitment to membranes is a critical regulatory mechanism because the Ras family of GTPases is membrane-bound. Each Ras GEF contains a particular set of structural motifs that may include apparent membrane localization domains. It is the presence of these membrane localization domains that predicts not only the translocation of Ras GEFs to membranes, but also the recruitment of these proteins into contact with their effector GTPases. In fact, Ras GEFs may be partitioned into subfamilies based on their putative modes of membrane recruitment. The RasGRPs are one subfamily of Ras GEFs that serve as an exemplar of regulation effected by membrane recruitment. Each member within this subfamily l contains a putative membrane localization domain, known as a C1 domain, which will be discussed in further detail in section 1.2.2. The C1 domain binds to the membrane localized second messenger, diacylglycerol (DAG). The predicted signalling pathway by which DAG recruits the RasGRPs to membranes is outlined in Figure 1.1. DAG is the product of phospholipase C (PLC)-catalyzed hydrolysis of either phosphatidylcholine (PC) or phosphatidylinositol 4,5-bisphosphate (PIP2) (Wakelam, 1998). In turn, PLC can be activated by PLC-coupled receptors. There is evidence that RasGRP subfamily members are activated by G protein-coupled receptors (GPCRs), which are coupled to the PLCp isoform and the receptor tyrosine kinases (RTKs), which are coupled to the PLCy isoform (Ji era/., 1998; Yablonski and Weiss, 2001). The studies of RasGRP member activation by members of each of these receptor superfamilies are described in sections and, respectively. Figure 1.1 depicts a model for membrane recruitment of RasGRPs whereby PLC-coupled receptor activation results in the production of DAG at membranes, resulting in the recruitment of RasGRPl via its C1 domain that in turn brings it into contact with Ras, so that it may activate this effector. In addition to the RasGRP subfamily, a number of other Ras GEFs are recruited to the plasma membrane. The Sos subfamily is recruited to the plasma membrane based on its ability to complex with a membrane localizing adaptor protein (Reuther and Der, 2000). As well, RasGRF2, another Ras GEF, relocalizes to the plasma membrane in response to increased calcium concentration (Fam et al., 1997). 2 R T K Figure 1.1. Model for membrane recruitment-regulated activation of the RasGRP subfamily. Wi th in this model , st imulat ion of ei ther a receptor tyrosine k inase (RTK) or a G protein-coupled receptor ( G P C R ) act ivates phosphol ipase C (PLC) so that it can catalyze phosphatidyl inosi tol 4,5-bisphosphate (P IP 2 ) hydrolysis to produce diacylglycerol (DAG). M e m b r a n e s enr iched in D A G then recruit a R a s G R P subfami ly m e m b e r p resumab ly into contact with membrane-bound Ras in order for this Ras G E F to act ivate it. In contrast to the hypothesis that membrane recruitment is the principal means of Ras GEF regulation, there are indications of other forms of regulation wherein translocation is either insufficient or not required. For example, Vav, a GEF for the Rho family of GTPases that are part of the Ras superfamily, is recruited to membranes, however phosphorylation of this protein is required for its GEF activity (Crespo et al., 1997). In addition, the two members of the RasGRF family of Ras GEFs provide evidence for translocation-independent regulation of their GEF activity. RasGRFI is constitutively localized to the plasma membrane and requires the second messenger calcium in order to activate it (Buchsbaum et al., 1996). In contrast with RasGRFI, RasGRF2 is cytosolic and requires a rise in calcium concentration in order to relocalize to the plasma membrane (Fam et al., 1997). However, one study has shown that the presence of its calcium modulator binding motif is not required for activation of the Ras signalling cascade by this Ras GEF (de Hoog et al., 2000). 3 The RasGRP subfamily of Ras GEFs contains both putative membrane localization domains and those that are not predicted to mediate the translocation of subfamily members to membranes. Consequently, this Ras GEF subfamily provides much substance for the study of its GEF activity modulation. 1.2 STRUCTURE OF THE RASGRP SUBFAMILY To date four members of the RasGRP subfamily of Ras and Ras-related GEFs have been identified; RasGRPI, CalDAGI, RasGRP3 and RasGRP4. They are located on human chromosomes 15q15, 11 q13, 2p23 and 19q13.1, respectively (Bottorff et al., 1999; Lorenzo ef al., 2001; Yang et al., 2002). Following the identification of four RasGRP members a more complete postulate of the evolution of the ancestral RasGRP gene can be made. It appears as though the members of this subfamily evolved >100 million years ago, which is before the divergence of mice and humans, because both of these species contain all four RasGRP genes. Based on sequence analysis it may be hypothesized that an ancestral gene duplicated twice and then segregated to three distinct chromosomes. One gene translocated to what would become human chromosome 19 and became RasGRP4. Likewise, the future RasGRP2 gene translocated to what would become human chromosome 11. RasGRPI and RasGRP3 share greater sequence identity than the other two RasGRP members and it is believed that RasGRP3 is derived from the ancestral gene for reasons that will be discussed in section 1.3.1. The RasGRPs demonstrate a similar domain structure to one another. In order from the NH2-terminus to the COOH-terminus, they all contain a Ras exchange motif (REM) box and guanine nucleotide exchange factor (GEF) domain, at least a single EF hand, which may bind calcium and a C1 domain, which may bind DAG. These COOH-4 terminal domains suggest potential mechanisms for regulating GEF activity, which will be expanded upon within the following sections. Beyond the signature structure of the RasGRP Ras GEFs there are dissimilar structural motifs among family members. A splice variant of CalDAGl, referred to as RasGRP2, was recently identified (Clyde-Smith era/., 2000). This splice variant contains myristoylation and palmitoylation modifications at its extreme NH2-terminus. Additionally, RasGRP4 is the only RasGRP subfamily member to contain a single EF hand and RasGRPl is the only subfamily member to contain an amphipathic a helix at its extreme COOH-terminus. The structural motifs for each family member are discussed in greater detail below and a representation of the structure for each RasGRP subfamily member is shown in Figure 1.2. R a s G R P l RasGRP2/ C a l D A G l R a s G R P 3 R a s G R P 4 Figure 1.2 Structure of RasGRP family of Ras GEFs. The Ras exchange moti f (REM) and guan ine nucleot ide exchange factor (GEF) domain are required for exchange activity on Ras. T h e EF hands (EFH) are a ca lc ium binding motif. The C1 doma in is predicted to bind diacylglycerol and diacylglycerol m imet ics and m a y media te the t ranslocat ion of R a s G R P s to m e m b r a n e s . T h e a lpha helix (a) is a predicted leucine zipper moti f that may mediate protein-protein interact ions. A spl ice var iant of the second R a s G R P subfami ly m e m b e r (RasGRP2) was recently identif ied and found to conta in m e m b r a n e -localizing myristoylat ion and palmitoylat ion (M+P) modif icat ions. 5 1.2.1 REM box and GEF domain Like other Ras GEF subfamilies, the RasGRP subfamily contains a REM box and GEF domain (Ebinu etal., 1998; Kawasaki etal., 1998; Tognon etal., 1998; Yamashita et al., 2000; Reuther et al., 2002; Yang et al., 2002). The GEF domains of all Ras GEFs share 30% sequence identity with one another and with the Saccharomyces cerevisiae protein, CDC25 (Robinson et al., 1987). The GEF domain acts as the catalytic domain of Ras GEFs by promoting the dissociation of GDP so that GTP may bind, thereby activating Ras. Within the GEF domain Boguski and McCormick (1993) first identified that conservation was greatest between the three structurally conserved regions (SCR 1-3). More recently, two other regions, SCR 4 and SCR 5, have become evident (Rebhun etal., 2000). The REM box, which is also called SCR 0, is a conserved non-catalytic region, that is upstream of the core catalytic domain and is found in GEFs that are known to exchange on Ras proteins and their closest relatives, such as RaplA and Ral (Lai et al., 1993; Fam etal., 1997; Boriack-Sjodin et al., 1998). It was determined by x-ray crystallography that the REM box is not involved in Ras interaction, but is a structural motif that binds SCR4 (Boriack-Sjodin et al., 1998). The presence of both a REM box and a GEF domain within the RasGRP subfamily members suggest that these proteins activate Ras arid Ras-related GTPases by guanine nucleotide exchange. The contributions of the GEF domain to the physiological activity of the RasGRPs are discussed in section 1.4.1. 1.2.2 C1 domain Each RasGRP subfamily member contains a C1 domain, which is a 50 amino acid lipid interaction motif. All C1 domains contain conserved cysteines and histidines 6 within their sequences, which are required for zinc coordination. However, as was delineated in section 1.1, the true subcellular function of this domain lies in whether or not it can bind the lipid second messenger DAG at membranes. As a result, C1 domains may be divided into typical DAG binding domains and atypical non-DAG binding domains. DAG is rapidly and transiently generated within both the plasma membrane as well as intracellular membranes in response to stimulation with growth factors, hormones, neurotransmitters and serum (Wakelam, 1998). Translocation to DAG- or DAG mimetic-enriched membranes is the hallmark role of the typical C1 domain and has thus far been demonstrated to aid in the activation of the C1 domain's resident protein. Consequently, the typical C1 domain is a positive regulatory module. Further discussion of C1 domains will be provided in section 3.1 The PKC family of serine/threonine kinases contains the originally identified C1 domains and provides the primary source of a comprehensive review of these domains. Solution of the second C1 domain of PKC8 by x-ray crystallography facilitated the identification of critical residues that are required for the formation of a membrane bound C1 domain ternary complex known as the C1/phospholipid/ligand ternary complex (Kazanietz era/., 1995; Zhang era/., 1995). The RasGRPl, RasGRP3 and RasGRP4 C1 domains contain all of the consensus typical CI domain residues, which predicts that each of these C1 domains may bind DAG and DAG mimetics (Hurley et al., 1997; Hurley and Misra, 2000). Conversely, the CalDAGl C1 domain contains only seven of the nine consensus residues predicted to complete the C1 domain portion of the ternary complex. While it contains all of the residues that are predicted to be required for DAG binding, a serine is present at the 8 t h position and an alanine is present at the 13 t h position rather than the conserved aromatic residues that are predicted to be required for nonspecific membrane interaction. Thus, based on 7 sequence analysis we cannot predict whether or not the CalDAGI C1 behaves as a typical DAG binding C1 domain. A comparison of the RasGRP C1 domain sequences is shown in Figure 3.1. Following the comparison of the RasGRP C1 domains to the typical C1 domain consensus sequence several studies have examined the DAG mimetic-induced responses of these domains. Within most studies, the archetypal DAG mimetics, phorbol esters, were used as ligands because this class of DAG mimetic has a higher affinity for C1 domains than DAG itself (Burns and Bell, 1991; Quest etal., 1994; Baatout et al., 1998). Upon examining translocation to membranes in response to DAG mimetics our laboratory found that the isolated RasGRPI C1 domain relocalized to the plasma membrane in response to stimulation with the phorbol ester, phorbol 12-myristate 13-acetate (PMA) in NIH 3T3 fibroblasts (Tognon etal., 1998). In addition, the C1 domains of both RasGRPI and RasGRP3 were shown to bind phorbol esters with high affinity in vitro (Lorenzo et al., 2000; Lorenzo et al., 2001). Another method that was used to determine membrane binding was a cellular fractionation assay. Within this assay, cellular fractions are divided into soluble fractions and membrane fractions, which are composed of the plasma and intracellular membranes. Under this assay it was shown that CalDAGI translocated to membranes in response to prolonged exposure to PMA, while another group demonstrated that RasGRP4 translocated to membranes following brief exposure to PMA (Clyde-Smith et al., 2000; Reuther et al., 2002). Despite a certain amount of evidence of DAG mimetic responsiveness for each C1 domain, a study of all four RasGRP C1 domains is required to compare DAG mimetic translocation propensities. It is worthwhile to resolve whether or not the CalDAGI and RasGRP4 C1 can bind to DAG mimetics as well as to determine their affinities for these ligands. 8 Ras GTPases are tethered to membranes requiring the translocation of Ras GEFs in order to activate these proteins. The presence of C1 domains with the RasGRPs suggests that these domains may facilitate translocation of the RasGRP resident member in a DAG- or DAG mimetic-dependent manner. Specific studies pertaining to both C1 domain-mediated biological regulation of RasGRPs and the contribution of phorbol esters and other DAG mimetic ligands to RasGRP subfamily activation will be discussed in section 1.4.1 and 1.3.2, respectively. 1.2.3 EF hands Upstream of the C1 domain RasGRPI, CalDAGI and RasGRP3 contain a pair of EF hands, while RasGRP4 contains a single EF hand (Reuther etal., 2002; Yang et al., 2002). EF hands are the most common intracellular calcium-binding motif found in proteins. This motif was first identified in the crystal structure of parvalbumin and named after its E and F helices by Kretsinger and Nockolds in 1973. The EF hand is a helix-loop-helix structure comprised of two nearly perpendicular a helices that flank a 12-amino acid loop that binds calcium. The vast majority of EF hands occur in adjacent pairs, a feature that facilitates the cooperativity of calcium binding to each motif and increases the affinity of the protein for this ion (Waltersson et al., 1993). Upon calcium binding, EF hands undergo a conformational change from a closed to an open state that exposes hydrophobic binding sites for protein-protein interaction and results in calcium signal transmission (Evenas etal., 1998). Yet numerous distinct peptide target sequences have been identified for EF hands, which hinders the prediction of protein target interaction (Lewit-Bentley and Rety, 2000). Since the identification of EF hands, two principal roles for these calcium-binding motifs have emerged. A few EF hand-containing proteins have intracellular calcium 9 buffering capacity, but most regulate downstream target proteins in a calcium dependent manner. Calmodulin and troponin C are examples of proteins whose EF hands behave as regulatory domains, whereas most sarcoplasmic calcium-binding proteins and parvalbumin function as calcium buffers (Ikura, 1996). The pairs of EF hands contained within RasGRPl, CalDAGl and RasGRP3 match the calcium binding loop consensus sequence, which suggests that these motifs bind calcium in a cooperative manner (Ebinu era/., 1998; Kawasaki et al., 1998; Tognon et al., 1998). The calcium-bound EF hand pairs are likely to be functioning as regulatory domains and not buffers, primarily because buffering EF hand-containing proteins do not contain multiple domains. As regulatory domains, the EF hands may positively or negatively regulate RasGRP subfamily activity. The calcium-dependent open conformation of the EF hands may negatively regulate these Ras GEFs. Negative regulation could occur if the calcium dependent conformation change of the EF hands resulted in the steric hindrance of other regulatory domains, such as the C1 domain. Alternatively, the EF hands may negatively regulate their resident RasGRP member by participating in a protein-protein interaction, which could sequester the Ras GEF from its normal site of activity. The EF hands may positively regulate RasGRP members by hydrophobically interacting with phospholipids in membranes following their calcium dependent conversion to an open conformation. This putative EF hand function may synergize with the formation of the C1/phospholipid/DAG ternary complex to retain the Ras GEF at the membrane thus, facilitating increased GEF activity. The single RasGRP4 EF hand may bind calcium, but does not have the benefit of cooperative ion binding. It remains to be determined whether or not this single EF hand results in the differential transduction of calcium signals compared to other RasGRP subfamily members. 10 Many EF hand-containing proteins transduce calcium signals by transforming increases in intracellular calcium concentration to cellular processes. Thus, a result of calcium binding to the EF hands may be to regulate RasGRP GEF activity. The influence of calcium ionophores on RasGRP GEF activity and the contribution of each of their EF hands to the regulation of these Ras GEFs will be discussed in section 1.3.2 and 1.4.1, respectively. 1.2.4 Amphipathic a helix Only RasGRPI contains a sequence at its extreme COOH-terminus that is predicted to form a continuous amphipathic a helix. Comparison of the RasGRPI a helix with other leucine zipper consensus sequences is shown in Figure 2.1. One aspect of amphipathic a helices is that they can bind membranes (Johnson and Cornell, 1999). Yet an additional feature of the RasGRPI amphipathic a helix is that it agrees with the consensus sequence for a leucine zipper motif (Tognon et al., 1998). This feature suggests that the a helix can facilitate the interaction of proteins with one another. As a result, the a helix may mediate RasGRPI dimerization. Like one of the potential regulatory roles described in the previous section for the EF hands, dimerization with another protein may also serve to sequester RasGRPI away from its normal site of activity. Alternatively, dimerization could have a positive regulatory contribution by relocalizing RasGRPI to a subcellular location for optimal activity of this Ras GEF. DGK ^ is a candidate dimerization partner because it co-immunoprecipitates with RasGRPI, however this kinase does not contain a leucine zipper consensus motif within its sequence, which indicates that heterodimerization with n DGK C, does not occur at the RasGRPl a helix (Topham and Prescott, 2001). Further discussion of the RasGRPl amphipathic a helix is provided in section 2.1. 1.3 RASGRP SUBFAMILY GEF ACTIVITY The Ras family of small GTPases The Ras small GTPases are a family of guanine nucleotide binding proteins (GNBPs) and operate as molecular switches that cycle between an active GTP-bound and an inactive GDP-bound state. The conformation of Ras G proteins differs between these two states, which allows for the recognition of distinct effector proteins and the transduction of signals to downstream proteins (Campbell et al., 1998). Ras GTPases contain slow and inefficient intrinsic GTPase activity. This intrinsic GTPase activity is significantly aided by the GTP hydrolyzing ability that GTPase activating proteins (GAPs) possess. In contrast to the function of GAPs, which is negative regulation, the function of GEFs within the GTPase cycle is to activate Ras GTPases by facilitating the GTP-bound, active conformation. The classical Ras GTPases are the prototype for the Ras family of small GTPases that play pivotal roles in signal transduction in virtually all cells. This family is comprised of the classical Ras GTPases (Ha-Ras, N-Ras, K-Ras4A and K-Ras4B) as well as a more divergent group of Ras-related G proteins that include the R-Ras subfamily (R-Ras, TC21/R-Ras2, M-Ras/R-Ras3), the Rap subfamily (RaplA, RaplB, Rap2A and Rap2B), the Ral subfamily (RalA and RalB), Rin, Rit and Rheb (Campbell et al., 1998). The chief effectors of the RasGRP subfamily of Ras GEFs are the classical Ras proteins, Rap1, and the R-Ras subfamily. The classical Ras GTPases have been extensively studied and have an established involvement in the signal transduction that leads to cell growth and differentiation. In contrast, the biological functions of Rap1 and 12 the R-Ras family are not as clearly understood (Macara et al., 1996; Campbell et al., 1998 Reuther and Der, 2000). Notably, RaplA (Krev-1) was first identified by its ability to reverse Ras-mediated morphological transformation of fibroblasts (Kitayama etal., 1989). Morphological transformation of cell lines will be further discussed in 1.4.1. Rap1 was believed to be a direct antagonist of Ras because it was shown to inhibit Ras signalling and because active Rap1 can form nonproductive complexes with the Ras target protein, Raf-1 (Cook etal., 1993; Zwartkruis and Bos, 1999). However, it has come to light that Rap1 has distinct physiological functions apart from the inhibition of Ras signalling. In particular, Rap1 plays a role in the positive regulation of specific integrins during cell adhesion and promotes B cell migration towards chemokines (Reedquist etal., 2000; Katagiri etal., 2000; Sebzda etal., 2002; McLeod and Gold, 2001). Like the classical Ras subfamily, the R-Ras subfamily plays a role in cell growth and differentiation (Reuther and Der, 2000). It also interacts with some of the same effector proteins and shares certain Ras GEFs with the classical Ras subfamily. However, the R-Ras subfamily has distinct functions from the classical Ras subfamily. While TC21 has a similar fibroblast transforming activity to Ras, the other two R-Ras subfamily members do not transform fibroblasts as well (Graham et al., 1994; Cox et al., 1994; Quilliam etal., 1999). In addition, unlike the classical Ras G proteins, R-Ras has been shown to positively affect integrin-mediated cellular adhesion and TC21 does not interact with the Ras effector, Raf-1 (Zhang era/., 1996; Graham etal., 1994). Methods for detecting Ras GTPase activation Several methods exist to detect guanine nucleotide exchange activity of Ras GEFs. These include both direct methods that measure the levels of GTP-bound 13 (active) Ras GTPases and indirect methods that measure the activation status of Ras GTPase effectors. The first of two methods to directly detect the activation of the family of Ras GTPases uses the Ras GTPase binding domains within the Ras GTPase effector proteins to preferentially immunoprecipitate GTP-bound, active Ras from cells and is known as a Ras "pull-down" assay. Specifically, for the detection of either active classical Ras proteins or Rap1, the Ras-binding domain (RBD) of Raf-1 or the Rap-binding domain of RalGDS are used, respectively (Taylor and Shalloway, 1996; Franke et al., 1997). The second method for direct detection of Ras GTPase activation is a guanine nucleotide exchange assay, which is specific for the Ras family member (Gibbs, 1995; Satoh and Kaziro, 1995). 1.3.1 Activation of Ras GTPases by the RasGRP subfamily The original member of the RasGRP subfamily of Ras GEFs, RasGRPl, was determined - via nucleotide exchange assays - to activate the classical Ras GTPase subfamily and not to have exchange activity for Rap1 A (Ebinu et al., 1998; Kawasaki et al., 1998). Likewise, both the guanine nucleotide exchange assay and the Ras "pull-down" assay determined that the most recently identified member of the RasGRP subfamily, RasGRP4, activates the classical Ras proteins, but not Rap1 (Reuther et al., 2002; Yang et al., 2002). In contrast, the second member of the RasGRP subfamily of Ras GEFs, CalDAGl, activated Rap1, but not the classical Ras GTPases via the same assay (Kawasaki et al., 1998; Clyde-Smith et al., 2000). However, this assay also found that a longer splice variant of CalDAGl, called RasGRP2, which contains NH 2 -terminal membrane-targeting lipid modifications (palmitoylation and myristoylation), caused rapid exchange activity on N-Ras, K-Ras and Rap1, but not on Ha-Ras (Clyde-14 Smith et al., 2000). In addition, these authors determined that CalDAGI could activate N-Ras in vitro. An obvious question that stems from the Ras member specificity differences seen between RasGRPI and CalDAGI is whether the GEF domains or the regulatory domains are responsible for this observation. In order to determine this, Matsuda and colleagues (2000b) generated chimeric RasGRP proteins by replacing the COOH-terminal regulatory domains, which included the EF hands and C1 domains, of each of these Ras GEFs with those of the other. They found that Ras member specificity was determined by each of the catalytic domains contained within RasGRPI and CalDAGI. The conclusion that can be drawn from these results is that the Ras GTPase specificity of RasGRPI and CalDAGI is determined by their respective GEF domains, as well as the membrane-localizing lipid modifications of the CalDAGI splice variant, but it is not determined by the EF hands and C1 domains. Similar to the longer CalDAGI splice variant, RasGRP2, the third member of the RasGRP subfamily, RasGRP3, activates both Ras and Rap1 by a guanine nucleotide exchange assay (Yamashita et al., 2000). In addition, this family member activates the R-Ras subfamily of GTPases, R-Ras, TC21 and M-Ras (Ohba etal., 2000). In comparison, both RasGRPI and CalDAGI non-specifically promote exchange activity on R-Ras and TC21, RasGRPI nonspecifically activates M-Ras and CalDAGI does not promote nucleotide exchange on M-Ras (Ohba et al., 2000). Hence, it is suggested that RasGRP3 is the prototype RasGRP subfamily GEF based on its broad effector specificity and that during evolution RaGRPI and CalDAGI lost the capability to activate Rap1 and Ras, respectively (Yamashita et al., 2000). 15 1.3.2 Regulation of Ras GTPase Activation by the RasGRP subfamily Ras GTPase regulation by DAG mimetics Both positive and negative regulators have contributed to RasGRP-dependent activation of Ras GTPases. These contributions will be outlined in the upcoming sections. The use of DAG mimetic ligands, particularly phorbol esters, has been extensive in measuring increases in RasGRP GEF activity. The rationale is that DAG mimetic ligand stimulation will recruit the RasGRP subfamily members, via their respective C1 domains, to specific membrane locations where the Ras GTPases reside in order to increase Ras activation. As a result, the DAG mimetic ligands will be positively modulating GEF activity. In cells overexpressing RasGRPl, the phorbol ester, phorbol 12-myristate 13-acetate (PMA), increased the amount of GTP-bound Ras (Ebinu et al., 1998; Kawasaki era/., 1998). Similarly, the overexpression of RasGRP3 was observed to increase Ras activation in a PMA-dependent manner (Lorenzo etal., 2001). In addition, the use of a PKC inhibitor determined that PMA-enhanced Ras GTPase activation occurred in a PKC-independent manner. These results provide evidence that RasGRPl and RasGRP3 GEF activity is positively regulated by phorbol esters. Consequently, the observed increases in GEF activity due to phorbol ester and other DAG mimetic ligands in cells overexpressing these two RasGRP members can be correlated with phorbol ester-induced plasma membrane translocation of these Ras GEFs. This correlation leads to the conclusion that these two RasGRP members are recruited to the location of effector Ras GTPases within DAG-enriched membranes, likely via their C1 domains, in order to promote guanine nucleotide exchange on Ras. To strengthen this conclusion, evidence of phorbol ester-enhanced GEF activity, which is measured by 16 Ras effector activation in cells overexpressing these two RasGRP members is noted in section 1.3.3. In contrast, phorbol esters appear to play less of a role in enhancing CalDAGI activation. This may be due to the differences in the CalDAGI C1 domain sequence compared to typical CI domain consensus sequence. Graybiel and colleagues (1998) found that phorbol esters were involved in increasing CalDAGI GEF activity by demonstrating that GTP loading of Rap1 was slightly increased following PMA stimulation in CalDAGI overexpressing cells. Unexpectedly, in COS cells, there is evidence that CalDAGI can cause exchange activity on N-Ras following prolonged, but not transient PMA stimulation (Clyde-Smith et al., 2000). This latter result is in concordance with the CalDAGI-mediated in vitro N-Ras activation that was previously discussed in section 1.3.1 and may indicate that in addition to being a Rap1 GEF, CalDAGI is a weak Ras GEF in a different milieu. Additionally the deviation of the CalDAGI C1 domain from the typical C1 domain consensus sequence may explain the minimal contribution that phorbol esters make to this Ras GEF's activity. Based on the hypothesis that consensus for formation of the C1/lipid/ligand ternary complex predicts the phorbol ester-induced translocation to membranes of the C1 domain's resident protein followed by the protein's catalytic activity, phorbol esters are likely to enhance RasGRP4 GEF activity. In support of this hypothesis, Der and colleagues (2002) found that Ras GTP loading by RasGRP4 was greatly augmented following PMA treatment. Ras GTPase regulation by calcium ionophores In contrast to the mechanism by which DAG mimetic ligands enhance GEF activity, the precise mechanism by which calcium increases RasGRP GEF activity is 17 unknown. However, one hypothesis reiterates the description of EF hands given previously in section 1.2.3 in which calcium-binding to EF hands may induce a conformational change such that hydrophobic surfaces are exposed. Therefore, similar to the mechanism of DAG mimetic augmentation of GEF activity, calcium may also positively regulate GEF activity by facilitating the hydrophobic interaction of EF hands with membranes. To test this hypothesis, much study has involved calcium ionophores, which are either natural or synthetic substances that can bind calcium ions and transport them across the lipid bilayer in membranes. There is little evidence for the involvement of calcium ionophores facilitating the increase of RasGRPl GEF activity. In RasGRPl overexpressing 293T cells, there was a slight increase of H-Ras activation upon stimulation with the calcium ionophore A23187 and the effect was additive with phorbol ester (Kawasaki et al., 1998). However, the same authors demonstrated that in CalDAGl overexpressing 293T cells, there was a marked enhancement of Rap1 GTP loading upon treatment with A23187 and the effect was also additive with phorbol ester. Thus, calcium appears to synergize with phorbol ester in increasing GEF activity in 293T cells. Further evidence for the positive regulatory role of calcium on CalDAGl GEF activity is given in section 1.3.3, where CalDAGl-mediated Ras family effector activation is discussed. However, in COS cells that overexpressed the alternatively spliced variant of CalDAGl, RasGRP2, A23187 was shown to decrease the amount of GTP-bound N-Ras activation (Clyde-Smith et al., 2000). In addition, calcium was shown to inhibit the in vitro activation of N-Ras by CalDAGl. Regardless of whether calcium ionophores are increasing or decreasing CalDAGl GEF activity, it appears as though calcium is a regulatory second messenger of this RasGRP member that likely exerts its effect by binding to the CalDAGl EF hands. 18 It remains to be determined whether or not calcium plays a role in regulating RasGRP3 and RasGRP4. Considering that RasGRP4 contains a single EF hand, this may result in calcium-stimulated differential regulation of this RasGRP member when compared to the other subfamily members, particularly because a pair of EF hands results in cooperative ion binding. Ras GTPase regulation by DAG kinases In addition to the aforementioned inhibition of CalDAGI GEF activity by calcium ionophores, another mode of RasGRP GEF activity inhibition has been reported. This mode involves DAG kinases (DGKs), which metabolize DAG into phosphatidic acid. The DGK catalytic activity suggests that these kinases may directly and negatively regulate DAG-modulated proteins by attenuating DAG concentration at membranes. There are indications that two DGK isoforms, DGKC, and DGKa, negatively regulate RasGRPI function (Topham and Prescott, 2001; Jones etal., 2002). DGK^ co-immunoprecipitates and colocalizes with RasGRPI, which suggests that these two proteins associate in the same signalling complex (Topham and Prescott, 2001). In addition, overexpressed DGKC, was shown to abolish Ras activation in RasGRPI overexpressing cells. This negative regulation of Ras activation by DGKC; was supported by the fact that overexpressed kinase-dead DGK^ mutants caused sustained signalling that resulted in Ras activation in Jurkat T cells. Likewise, overexpression of a kinase-dead DGKa mutant resulted in sustained Ras activation and downstream effector activation in Jurkat T cells (Jones et al., 2002). Thus, it may be concluded that both DGKt; and DGKa impede RasGRPI GEF activity. Yet it remains to be determined whether or not DGKs can negatively regulate other members of the RasGRP subfamily. 19 1.3.3 Activation of Ras GTPase downstream effectors by the RasGRP subfamily The Ras/MAPK cascade Other methods of determining Ras GEF activity have employed indirect measures of Ras GTP loading. These methods may not only detect activated Ras GTPases, but also demonstrate whether or not downstream effectors of the Ras family of GTPases are activated by the Ras GEFs. The best-characterized Ras signalling pathway is the Ras/mitogen activated protein kinase (MAPK) pathway (Downward, 1996). Within the MAPK cascade, the Raf-1 serine/threonine kinase is recruited to membranes where it interacts with and is activated by GTP-bound Ras, which leads to phosphorylation and activation of the dual specificity MAPK kinases, MEK1 and MEK2. The MEKs then phosphorylate threonine and tyrosine residues on extracellular signal-regulated kinases (ERKs), ERK1 and ERK2. ERK activation then leads to the activation of cytoplasmic target proteins such as Rsk (ribosomal S6 kinase; Sturgill et al., 1988; Palmer et al., 1998) and Mnk (Waskiewicz et al., 1997). Once activated these proteins translocate to the nucleus where they can phosphorylate and activate a number of substrates, such as the Elk-1 transcription factor. However, ERK and Elk-1 activation may also be caused by Ras-independent pathways, such as the PI3-K pathway, PKCs and integrin-mediated signalling events (Grammer and Blenis, 1997; Cai et al., 1997; Schlaepfer et al., 1994). In addition, another MAP kinase, c-Jun NH 2 -terminal kinase (JNK), can also phosphorylate Elk-1 (Treisman, 1996). It is also worthwhile to measure whether ERK activation is influenced by Rap1 GEFs because Rap1 is implicated in both the inhibition and activation of ERK. Inhibition and activation of ERK are dependent on whether Rap1 binds to Raf-1 or another Raf isoform, B-Raf, respectively (Okada et al., 1999). Rap1 does not preferentially bind to B-Raf in every cell line in which both Raf isoforms are expressed, 20 however Rap1 activation of ERK through B-Raf has been observed to mostly occur in neuronal cells (Bos etal., 2001). Hence, the precise regulation of Rap1 -mediated ERK activation remains to be resolved. Methods for detecting Ras family effector activation In order to measure ERK activation two methods are commonly used. One method employs a MAPK-specific transcription factor in a reporter gene assay to indirectly detect MAPK activation, while the other uses antibodies that are specific for active ERK. A reporter gene assay can detect MAPK activation by the phosphorylation and subsequent activation of an artificial Elk-1 transcription factor, which is specifically phosphorylated by MAPKs (Waskiewicz and Cooper, 1995). Hence, this assay is an indirect method by which to measure the level of MAPK activation. In order to directly measure ERK activation antibodies can be used to detect the phosphorylation of tyrosine 204 in ERK1 and ERK2. Also, as with Ras GTPase activation, there are several instances where downstream effector activation is enhanced following stimulation with either phorbol esters or calcium ionophores. However, it should be noted that downstream signals might become diluted, which would not result in either signal detection or implicit GEF activity observation. Activation of Ras family effectors by the RasGRP subfamily The Elk-1 reporter gene assay demonstrated that RasGRPI activated MAPK (Kawasaki etal., 1998; Tognon etal., 1998). In addition, our laboratory found that MAPK activation was enhanced by coexpression of Ha-, K- and N-Ras and was partially suppressed by coexpression of dominant negative forms of H- and K-Ras (Tognon et al., 1998). The observed synergy between RasGRPI and each of Ha-, K- and N-Ras 21 suggests that RasGRPl is a GEF for the classical Ras GTPases. Additionally, MAPK activation was eliminated following the deletion of the RasGRPl C1 domain, restored by a membrane-localizing prenylation signal and PMA was found to increase the level of activated ERK in RasGRPl overexpressing cells (Tognon et al., 1998; Ebinu era/., 1998). These results imply that membrane localization via the C1 domain is required for Ras activation by RasGRPl. Furthermore, in vivo evidence for the requirement of RasGRPl in phorbol ester-induced ERK activation is shown by the abolishment of ERK activation in response to PMA in thymocytes homozygous for a RasGRPl null mutation (Dower et al., 2000). Likewise, the third RasGRP member activates ERK. Moreover, PMA also enhances ERK activation in a concentration dependent manner in RasGRP3 overexpressing cells (Lorenzo et al., 2001). Furthermore, upon comparison of RasGRP subfamily members, Matsuda and colleagues (2000b) found that activation of ERK was most prominent by RasGRP3, followed by RasGRPl. In contrast, like Rap1, CalDAGl appears to both inhibit and activate ERK. Graybiel and colleagues (1998) determined that CalDAGl overexpression inhibited Ras-dependent Elk-1 phosphorylation by using a constitutively active Ras mutant and the Elk-1 reporter gene assay in 293T cells. However, consistent with previous observations regarding Rap1 activation of B-Raf in neuronal cells, it was found that B-Raf appears to be activated by CalDAGl activation of Rap1, which results in ERK1/2 activation in neuronal PC12D cells (Guo et al., 2001). These conclusions were based on a co-immunoprecipitation experiment demonstrating that CalDAGl complexes with Rap1 and B-Raf, and phosphorylates B-Raf in order to activate ERK1/2. Both of these results were obtained following stimulation with calcium ionophore and either PMA or 1 -oleoyl-2-acetyl-sn-glycerol (OAG), but not with either ligand alone. Hence, these 22 results also suggest that calcium influx is sufficient to cause CalDAGI activation of B-Raf. However, another possibility is that endogenous DAG in PC12D cells is sufficient when combined with calcium influx to cause CalDAGI-mediated B-Raf activation. In an attempt to resolve the negative versus positive regulatory contribution that calcium makes to CalDAGI GEF activity it should be noted that CalDAGI expression in neuronal cells including PC12D cells is grounds for greater biological relevance of this CalDAGI effect. c-Jun NH2-terminal kinase (JNK) is another MAPK, where, like ERK, its activation can be detected by antibodies specific for active JNK. Specifically, it is a stress-activated protein kinase (SAPK) that can be activated by stress cytokines or exposure to environmental stress (Kyriakis and Avruch, 2001). Ras is one of the upstream signalling proteins that may, in a Raf-independent manner, activate JNK, which suggests that Ras GEFs can modulate this MAPK's regulation (Minden et al., 1995; Lin etal., 1995). In addition, JNK is activated by R-Ras, but not Rap1 (Mochizuki et al., 2000). Matsuda and colleagues (2000b) demonstrated that JNK was phosphorylated most prominently by RasGRPI, and less so by RasGRP3, a result that is opposite to the RasGRP-mediated activation that was observed for ERK activation. This could be the result of direct effector specificity differences between these two RasGRP subfamily members. One possibility is that RasGRP3 is balancing its GEF activity between Ras and Rap1, which does not activate JNK and suggests that, compared to RasGRPI, a smaller portion of RasGRP3's GEF activity signals to activate JNK. In addition, the authors also found that JNK was not activated by CalDAGI, which follows with the line of reasoning that like Rap1, Rap1 GEFs do not activate JNK. 23 Further studies of RasGRP GEF activity will require both the examination of the domains required and the determination of additional Ras family effector pathways that are activated by the RasGRP subfamily. The Ras family activates several Raf-independent effector pathways (Campbell et al., 1998; Reuther and Der, 2000). After Raf, phosphatidylinositol 3-kinase (PI3-K) represents the second best-characterized effector of Ras. It is involved in a number of cellular processes including cell survival, cytoskeletal remodeling and vesicular transport. Another family of proteins that has been implicated as effectors of Ras are RalGDS and related proteins that activate Ral small GTPases (Campbell et al., 1998). Other Ras effectors include AF-6, which may be involved in cellular adherens junctions stability, and PLCs, which is involved in calcium signalling (Shields et al., 2000; Vos etal, 2000; Kelley et al, 2001; Song et al, 2001). Strikingly, PLCs also contains a GEF domain, which has been shown to activate Ras and classifies this enzyme as bifunctional (Lopez et al, 2001). In addition, certain Ras effectors are also Rap1 effectors such as the Ral small GTPases and PI3-K. However, unlike for Ras, direct regulation of Ral GEFs by Rap1 has not been established (Bos etal, 2001). 1.4 CELLULAR PROCESSES INFLUENCED BY RASGRP SUBFAMILY MEMBERS 1.4.1 Morphological Transformation Transformation of immortalized cell lines is a phenomenon that results in the increase of the proliferative potential of cells accompanied by morphologic changes and anchorage independent growth. As a result, this phenomenon has been used to identify genes involved in cell growth, which, if abnormally active, may contribute to tumor formation. In fact, many members of the Ras superfamily of GTPases induce this process. Yet there is a distinct transformed cell morphology phenotype between 24 families. Transformation assays using fibroblast cells have shown that the expression of activated classical Ras proteins results in an elongated and refractile cell morphology, while wild type cells exhibit a flattened morphology and are non-refractile. In addition, the foci generated by these cells grow in layers due to the transformation-mediated acquisition of anchorage independent growth. This Ras-like transformed phenotype has been used to identify members of the Ras signalling cascade such as Ras GEFs and consequently may be viewed as being synonymous with GEF activity. Both RasGRPl and RasGRP4 were identified as oncogenic proteins by their ability to induce transformation of fibroblast cells. Human RasGRP4 was cloned by its ability to transform a rat fibroblast cell line, Rat-1 cells (Reuther et al., 2002). Mouse RasGRPl was cloned within our laboratory by screening a cDNA library from a mouse T cell hybridoma, the T28 T cell line, for clones whose expression caused morphological transformation of NIH 3T3 mouse fibroblasts (Tognon etal., 1998). In addition, rat RasGRPl was cloned via its ability to transform Rat-2 fibroblasts (Ebinu et al., 1998). In both cases the transformation induced by RasGRPl exhibited a Ras-like cellular morphology, which suggested that this protein was involved in the Ras signalling pathway. It was also determined that certain domains within RasGRPl are required for RasGRPl's transforming activity and that phorbol esters can enhance the transforming activity of RasGRPl via the C1 domain (Ebinu et al., 1998; Tognon et al., 1998). Our laboratory found that deletions of the REM box, the GEF domain and the C1 domain all abolished this Ras GEF's transforming activity. Correspondingly, in conditions of low serum, PMA could promote the transforming activity of RasGRPl. In addition, Stone and colleagues (1998) found that following the deletion of the RasGRPl C1 domain the transformed cell morphology of Rat-2 cells induced by PMA was abolished (Ebinu etal., 1998). In contrast, the EF hands were found to not be required 25 for the morphological transformation induced by RasGRPI. This result is in accordance with the previously described lack of calcium ionophore involvement in enhancing RasGRPI GEF activity in section 1.3.2. Conversely, overexpression of CalDAGI does not cause transformation of NIH 3T3 cells (Dupuy et al., 2001; Kay, unpublished data). Likewise, as was previously described in section 1.3, its effector, Rap1, was identified as an antagonist of transformation for the same fibroblast cell line (Kitayama et al., 1989). However, both Rap1 and this Rap1 GEF caused transformation of Swiss 3T3 mouse fibroblasts when either was overexpressed (Altschuler et al., 1998; Dupuy et al., 2001). This evidence attests to the cell context-dependent differential modulation of transformation by Rap1 and its GEF and insinuates that members of the Rap1 signalling pathway can positively regulate cell growth. The third member of the RasGRP subfamily, RasGRP3, induced transformation of PC12 neuronal cells and Rat-1 A rodent fibroblasts, although less efficiently than RasGRPI (Yamashita et al., 2000). This decreased efficiency may be based upon the fact that RasGRP3 is a dual specificity GEF whose effectors have distinct effects on the transformation of cell lines. Hence, RasGRP3 GEF activity may result in a balance between the promotion of transformation and the lack of effect on transformation, while that of RasGRPI results in only the activation of transformation-promoting effectors. The inference that Rap1 does not have an effect on the transformation efficiency of either PC12 or Rat-1 A cells is based upon the observation that CalDAGI overexpression did not result in transformation of either of these cell lines. This observation serves as another indication that transformation by Rap1 and this Rap1 GEF is cell context-dependent. 26 1.4.2 PLC-coupled Receptor Signalling Phospholipase C (PLC) appears to be a primary candidate for the upstream regulation of the RasGRP subfamily. Cleavage of the PLC substrate phosphatidylinositol 4,5-bisphosphate (PIP2) produces both DAG at membranes as well as inositoltriphosphate (IP3), which results in an increase in calcium concentration within the cells by stimulating the release of intracellular calcium stores and the opening of calcium channels within the membrane. As a result, the initiation of PLC activation at the cell surface by PLC-coupled receptors is an essential topic of study to further characterize the regulation of the RasGRP subfamily. In the proceeding sections the putative role of PLC-coupled receptors in RasGRP subfamily regulation will be discussed. This discussion is facilitated by the expression patterns of RasGRPl and CalDAGl, which, among other cells, have differential expression in neuronal cells, and are both expressed in T lymphocytes and various T cell lines (Ebinu et al., 1998; Kawasaki et al., 1998; Tognon et al., 1998; Kay, unpublished data). Specifically, the discussion will centre around the T cell antigen receptor (TCR), which is coupled to PLCy and has been shown to activate RasGRPl, and the Mi muscarinic acetylcholine receptor (Mi mAChR), which is a G protein-coupled receptor that is coupled to PLCp and has been shown to activate CalDAGl. CalDAGl in Mi mAChR Signalling Muscarinic acetylcholine receptors (mAChR) are members of the GPCR superfamily that initiates intracellular signalling by activating the heterotrimeric G proteins a:p:y (Ji et al., 1998). Heterotrimeric G proteins remain associated with the cytoplasmic tail of unstimulated GPCRs as an inactive, GDP-bound, trimeric complex. 27 Following ligand binding to the GPCR, the Ga subunit decreases its affinity for GDP, allowing GTP to bind. The Ga subunit then dissociates from the Gpy heterodimer, which facilitates the activation of the respective Ga and (3y downstream targets. The vast majority of GPCR signalling is through the G protein-regulated pathways of adenylyl cyclase, phospholipase Cp (PLCp) and ion channels. The mAChRs are classified into two functional groups. The M-i, M 3 and M 5 subtypes couple to the G protein isoform, G q /n, to activate PLCp. In the second functional group the M 2 and M 4 subtypes couple to the G protein isoform, Gj/0 to inhibit adenylyl cyclase (Wess, 1993; Caulfield, 1993; Felder, 1995). As a PLC-coupled receptor, the M-i mAChR-induced activation of PLCp results in the production of IP 3, which leads to an influx in extracellular calcium, and DAG, which has been shown to lead to PKC activation within PC12D neuronal cells (Ebihara and Saffen, 1997). In addition, this mAChR has been shown to activate ERK1 and ERK2 (Belcheva and Coscia, 2002). In fact, M-i mAChR-induced calcium influx and PKC activation, and the Mi mAChR-induced ERK activation induce expression of the same immediate-early gene (Kumahara etal., 1999). However, until recently the signalling events leading to the activation of these MAPKs by the Mi mAChR have remained poorly understood. Saffen and colleagues (2001) first determined that activation of ERK1/2 by the Mi mAChR-specific agonist, carbachol, takes place via a Rap1-dependent and Ras-independent pathway in PC12D neuronal cells. In addition, in a Rap1-dependent manner, carbachol was shown to activate B-Raf and MEK, where MEK exclusively activates ERK1/2. Based on the fact that CalDAGI has putative binding sites for both calcium and DAG, this Rap1 GEF was proposed to link M-\ mAChR stimulation to 28 ERK1/2 activation. To this end, CalDAGl expression was determined in PC12D cells. Like the result of calcium ionophore stimulation previously mentioned in section 1.3.3, carbachol was also found to stimulate CalDAGl/Rap1 /B-Raf complex formation via co-immunoprecipitation. Furthermore, CalDAGl antisense mRNA expression demonstrated that CalDAGl was required for Mi mAChR-stimulated B-Raf activation. These results elucidated a signalling pathway from the Mi mAChR to ERK activation. The authors concluded that stimulation of the M 1 mAChR by carbachol resulted in the sequential activation of CalDAGl, Rap1, and B-Raf. This sequential activation then led to the activation of MEK and ERK1/2. Furthermore, the carbachol stimulation results, in combination with the calcium ionophore stimulation results described in section 1.3.3, indicate a signalling pathway that is initiated by the Mi mAChR. Following receptor stimulation, the signal transduction then proceeds through the PLCp-catalyzed increase in intracellular calcium to positively regulate the presumably calcium-responsive CalDAGl EF hands in order to increase Rap1 and subsequent Rap1 effector activation in PC12D cells. In contrast to these signals, it is unclear whether or not the CalDAGl C1 domain is positively regulated by PLCp-catalyzed DAG production because as previously described in section 1.3.3, stimulation with DAG mimetics alone were determined to not induce CalDAGl-dependent activation of Rap1 and ERK in this cellular milieu. RasGRPl in TCR Signalling The TCR belongs to a class of multi-subunit cell surface receptors utilized by cells of the immune system. This class of receptors is comprised of the B and T cell antigen receptors, as well as receptors that bind the constant portion of circulating immunoglobulins (e.g. FceRI) (Cambier and Jensen, 1994; Gold and Matsuuchi, 1995). 29 The binding of antigen to the TCR complex triggers activation of T lymphocytes. The TCR complex consists of covalently linked a and p polypeptide chains, which make up the TCR heterodimer, the y:8 and s:5 heterodimers of the CD3 complex, and the C, chain homodimer (Lin and Weiss, 2001). The TCR plays a pivotal role in T cell proliferation via signalling that leads to the induction of the cytokine interleukin 2 (IL-2) (Crabtree, 1989). The TCR is also critical for T lymphocyte maturation, which will be discussed further in section 1.4.3. The co-receptor molecules CD4 or CD8 bind major histocompatibility complex (MHC) molecules, MHC class II and MHC class I, respectively, and act synergistically with the T-cell receptor in order to initiate signalling by activating a number of protein tyrosine kinases (PTKs). Upon TCR engagement, co-receptor-associated, Src family protein Lck is activated and phosphorylates the intracellular motifs within the CD3 chains. Phosphorylation of the CD3 chains results in the recruitment and activation of the Syk family member ZAP-70 which, in turn, results in the phosphorylation of the linker for activated T cells (LAT). LAT has a central importance in T cell function because it is critical for antigen receptor function and it results in the recruitment of a number of other proteins that are involved in the activation of the Ras pathway, calcium mobilization and cytoskeletal reorganization (Zhang etal., 1998a; Zhang etal., 1999; Finco etal., 1998). The pathways that couple TCR ligation to Ras activation have' been extensively studied. The Ras GEF Sos, complexed with the adaptor Grb2, was shown to activate Ras in T lymphocytes (Holsinger et al., 1995). It is believed that active LAT recruits this complex to the membrane because this protein has multiple Grb2 binding sites that become phosphorylated following TCR stimulation (Zhang et al., 1998a). In addition, LAT also recruits PLCyl to the membrane following TCR stimulation, which results in 30 DAG production (Zhang etal., 1998b; Yablonski etal., 1998). Furthermore, a recent study demonstrated that loss of PLCyl recruitment to LAT in Jurkat T cells inhibited the phosphorylation and activation of PLCyl, which resulted in decreased ERK activation (Zhang etal., 2000). Thus, both the Grb2-Sos complex and PLCyl regulate TCR-induced Ras activation. As well, in T lymphocytes Ras has long been known to be activated following phorbol ester treatment (Downward etal., 1990). Together, the involvement of PLCyl and phorbol esters suggest a direct correlation between DAG generation and Ras activation in T cells. Early work regarding the mechanism by which DAG activates this G protein suggested that activation of PKCs phosphorylates the Ras negative regulators, GTPase activating proteins (GAPs), thereby inhibiting them (Downward etal., 1990; Bollag and McCormick, 1992). However, studies which inhibited phosphorylation by PKC indicated that there are also PKC-independent pathways for TCR-mediated activation of Ras (Downward et al., 1992; Izquierdo et al., 1992). RasGRPl fulfills the criteria of a PKC-independent, DAG-dependent link to TCR-induced Ras activation. Ebinu et al., (2000) recently demonstrated a number of results that pointed towards this role for RasGRPl in T cells. Within their study a Jurkat T cell line was used to determine that in response to the OKT3 antibody, which recognizes the extracellular portion of the CD3e chain, a PLCyl inhibitor diminished TCR-induced Ras activation suggesting that DAG modulates Ras activation. Secondly, they determined that TCR stimulation resulted in rapid and sustained RasGRPl localization to the membrane, which is analogous to the DAG mimetic-induced membrane localization that had previously been observed for this protein. RasGRPl-mediated Ras activation in response to TCR cross-linking was also observed in Jurkat T cells. 31 Furthermore, overexpression of RasGRPI in these cells enhanced ERK activation. Thus, this study demonstrated that RasGRPI serves as a direct link between TCR cross-linking and Ras activation, likely via PLCyl-catalyzed production of DAG (Ebinu et al., 2000). Further evidence that this TCR-induced Ras activation proceeds through DAG was shown by another study that used a kinase-dead mutant of the DAG level attenuator, DGKa, to demonstrate that sustained signalling proceeded through RasGRPI and ERK in Jurkat T cells (Jones etal., 2002). There is also evidence of TCR-induced signalling through RasGRPI in primary T cells. TCR stimulation by anti-CD3 alone or combined with the costimulatory molecule anti-CD28 resulted in ERK activation of wild-type thymocytes. ERK activation following TCR stimulation was not detected in thymocytes that were homozygous for a RasGRPI null mutation indicating that RasGRPI is required for TCR-induced ERK activation in these primary T cells (Dower et al., 2000). The combination of this result with the previously mentioned requirement of RasGRPI for PMA-induced ERK activation (Section 1.3.3) indicates a pathway wherein TCR activation results in DAG production, which activates RasGRPI and subsequently ERK in primary T cells. These results also imply that PLCyl and Ras are activated in their respective positions within this signalling cascade. Additionally, these results confirm that RasGRPI is a physiologically relevant Ras GEF in T cells. Further studies will need to determine the extent of the contribution that this protein makes to TCR-induced Ras activation compared to both the Ras GEF Sos and to PKC inhibition of Ras GAPs. The study of signalling cascades initiated by the Mi mAChR and the TCR has strongly suggested that both PLCp and PLCy can couple these receptors to CalDAGI 32 and RasGRPl, respectively. However, the use of either PLC inhibitors or cell lines engineered to be PLC deficient and overexpression of a RasGRP subfamily member would provide greater evidence that RasGRP activation proceeds through PLC. Furthermore, whether or not other PLC-coupled receptors can induce the GEF activity of RasGRPl, CalDAGl and the other RasGRP subfamily members also remains to be determined. One candidate PLC-coupled receptor for both RasGRPl and CalDAGl is the B cell antigen receptor (BCR) because this receptor is coupled to PLCy and both CalDAGl and RasGRPl are expressed in B lymphocytes as well as certain B lymphoid cell lines (Campbell, 1999; Tognon etal., 1998; Kay, unpublished data). In addition, both Ras and Rap1 are activated by the BCR (Genot and Cantrell, 2000; McLeod et al., 1998). It should also be noted that carbachol promoted RasGRPl membrane localization in Jurkat cells that ectopically expressed a Mi mAChR receptor (Jones et al., 2002). Following this observation, the obvious question is whether or not RasGRPl operates as a physiological GEF in neuronal cells where the Mi mAChR is expressed endogenously. 1.4.3 RasGRPl involvement in Late-stage Thymopoiesis Within the later stages of thymopoiesis thymocytes mature from double negative (DN) in which neither the CD4 nor the CD8 coreceptors are expressed to the intermediate double positive (DP) stage, where both coreceptors are expressed. During this maturation step, the TCR genes rearrange to form a completely assembled a/p TCR (Kruisbeek et al., 2000; von Boehmer et al., 1998). As was first described in section coreceptors bind peptide-MHC class-specific molecules, however only in thymocytes that express the completely assembled a/p TCR is functional signalling initiated by coreceptor binding to peptide-MHC. 33 The mature T cell stage is the single positive (SP) stage in which T cells express either the CD4 or CD8 coreceptor. It is during this transition that thymocytes undergo a selection process wherein DP thymocyte deletion occurs if receptors with strong affinity for peptide-MHC are expressed and cell death by neglect occurs if thymocytes express receptors with weak affinity. Conversely, intermediate receptor affinity confers positive selection, which results in cell survival and differentiation to one of the two SP lineages. Studies using thymocytes of RasGRPI-deficient mice indicated that RasGRPI is required for thymopoiesis (Dower et al., 2000). RasGRPI-deficient mice showed reductions in both CD4 + and CD8 + SP thymocytes, which shows that RasGRPI is required for production of mature thymocytes. However, it remains to be determined whether impairment of positive selection or enhanced negative selection is the cause of this phenotype. Nevertheless, it should be noted that through the introduction of dominant negative forms of Ras and its effectors, Raf-1 and MEK, Ras signalling was determined to be required for positive selection (Sebzda et al., 1999). Dower et al., (2000) also noted that DN thymocytes doubled and DP thymocytes were modestly reduced in RasGRPI mutants (Dower et al., 2000). This is interesting because Ras signalling is required for the transition from DN to DP thymocytes, and Raf-1 and MEK1 have also been implicated in this transition (Sebzda etal., 1999; Michie etal., 1999; Crompton et al., 1996; Iritani et al., 1999). Future studies will clarify the extent of RasGRPI involvement in this transition. As was previously discussed in section, the ability of the RasGRPI-deficient thymocytes to activate ERK in response to anti-CD3 and DAG mimetic was eliminated. In addition, proliferation was eliminated in response to a number of growth stimulants such as IL-2 and anti-CD3 or PMA and calcium ionophore. This may reflect the decrease in SP thymocytes, which would provide much of the response to these 34 stimuli. Indeed, making precise conclusions regarding the involvement of RasGRPl in thymocyte signalling should be cautioned because Dower et al., (2000) used total thymocytes, which have differential responses to mitogens. For example, DP thymocytes do not proliferate in response to the above-mentioned stimuli. Hence, sorting thymocytes into their respective developmental stages and repeating the signalling studies will yield more rigorous conclusions about the role of RasGRPl in thymocyte signalling. 1.5 THESIS OBJECTIVES While Ras GEFs activate Ras and Ras-related proteins in an identical manner, a substantial amount of data has recently come to light regarding the diverse modes by which they are regulated. The greatest body of evidence involves membrane recruitment as a regulatory mechanism. In addition, distinct membrane recruitment mechanisms for Ras GEFs also exist. However, there are situations where translocation is either unnecessary or insufficient for Ras GEF regulation. In order to more fully understand the regulation of the RasGRP subfamily of Ras GEFs two aspects of the regulation of RasGRPl and CalDAGl were selected for study. Our first hypothesis was that RasGRPl homodimerizes and this was tested in Chapter 2. This hypothesis based upon the presence of a unique putative regulatory domain, an a helix/leucine zipper motif that predicts protein dimerization. The potential for RasGRPl to exist as a homodimer may add an additional regulatory mechanism to this Ras GEF. Our second hypothesis, addressed in Chapter 3, was that the C1 domains of RasGRPl and CalDAGl have different ligand specificities due to the respective presence and absence of residues predicted to be critical for membrane interaction 35 (Chapter 3). This difference in specificity may cause distinct membrane translocation patterns for RasGRPl and CalDAGl in response to C1 domain ligands, results that would suggest the differential C1 domain-mediated regulation of RasGRPl and CalDAGl. The importance of exploring translocation patterns between Ras GEFs is due to the fact that this mechanism of regulation is believed to be integral for their functioning because Ras proteins are tethered to membranes and like most Ras GEFs, these two Ras GEFs are cytosolic. 36 Chapter 2 INVOLVEMENT OF THE RASGRP1 a HELIX/LEUCINE ZIPPER MOTIF IN HOMODIMERIZATION 2.1 INTRODUCTION Amphipathic alpha (a) helices are able to bind both lipids and proteins within a cell. For example, an amphipathic a helix can bind membranes by inserting its hydrophobic face into the phospholipid bilayer, while its polar face contacts the cytosol. Membrane-binding amphipathic a helices are found in a diverse range of proteins including CTP:phosphocholine cytidylyltransferase, ADP-ribosylation factor, vinculin, DnaA and blood-clotting factor VIII (Johnson and Cornell, 1999). Amphipathic a helices are also involved in the protein-protein interactions that constitute a leucine zipper structure. Along with having been first found within transcription factors, leucine zippers are known to be present in a wide range of proteins including cytoskeletal proteins, protein kinases and adaptor proteins (Busch and Sassone-Corsi, 1990; Hartshorne et al., 1998; Ruth etal., 1997; Kawai etal., 1998; Bartkiewicz etal., 1999). The role and structure of the leucine zipper in mediating homo- and heterodimerization remains the best characterized within the transcription factors jun, fos and GCN4. The formation of a leucine zipper requires two amphipathic a helices, one from each monomer, that wind around each other in parallel to form a stable coiled-coil. The structure was first determined by x-ray crystallography for the GCN4 leucine zipper (O'Shea et al., 1991). RasGRPI is the only member of the RasGRP subfamily that contains an amphipathic a helix. It is found at its extreme COOH-terminus and consensus sequence predicts that this domain is a leucine zipper motif (Tognon et al., 1998). Figure 2.1 compares the RasGRPI a helix to consensus leucine zipper sequences. 37 This figure shows that each a helix is composed of four to five repetitive heptad repeats denoted (abcdefg)n. Interhelical hydrophobic interactions between hydrophobic a residues and hydrophobic (often leucine) d residues are required to hold the two amphipathic a helices together. The residues at the a and d positions together constitute this motifs hydrophobic core. In addition, the asparagine at the third a positio confers dimerization specificity (Zeng et al., 1997). a b I — a. b c d e f g & b c d e f g a b c d e f g a b c d e f g a b C d e f 9 1. Y Q E L E Q E I M T L K A D N D A L K I Q L K Y A Q K K I E 3 L Q L E 2 . A I R L E E K V K T L K A Q N 3 E L A 3 T A N M L R E Q V A Q L K Q K 3. T D T L Q A E T D Q L E D E K 3 A L Q T E I A N L L K E K E K L E F I 4 . M K Q L E D K V E E L L 3 K N Y H L E N E V A R L K K L V G E R • • Figure 2.1 Comparison of the RasGRPI alpha helix sequence with leucine zipper consensus sequences, a. Cross-sect ion of 2 a hel ices showing hydrophobic interact ions between the a and d residues to fo rm a leucine zipper, b. The R a s G R P I a helix (line 1.) is a l igned with the leucine z ipper consensus moti fs of Jun (line 2.), Fos (l ine 3.) and the yeast t ranscr ipt ion factor G C N 4 (line 4.) a long the repetit ive heptad sequence ( top l ine). A single line encases leucines and a double line encases the asparagine that confers d imer izat ion specif icity. Consensus of the RasGRPI a helix with the leucine zipper motif suggests that this domain may mediate protein-protein interactions as a mode of regulation. One possibility that is simple to test is if the RasGRPI leucine zipper homodimerizes. We 38 used both co-immunoprecipitation and a modified colocalization assay in a mammalian cell system in order to detect RasGRPl protein-protein interactions. With these methods and using NH2-terminally truncated RasGRPl constructs, our aim was to demonstrate that RasGRPl homodimerizes at its COOH-terminus. These two independent approaches provided several lines of evidence that RasGRPl interacts with itself. If the results from both approaches are taken together, we can narrow down the region of interaction to a region downstream of the RasGRPl C1 domain. We suggest that the most likely candidate domain to mediate this interaction is the a helix or leucine zipper motif that is contained within the COOH-terminus of RasGRPl. 39 2.2 METHODS 2.2.1 Cell culture NIH 3T3 cells were obtained from American Type Culture Collection (ATCC; Manassas, VA) and cultured at low density in Dulbecco's Modified Eagle's Medium (DMEM; StemCell Technologies, Vancouver, BC) containing 9 % calf serum (CS; HyClone, Logan, UT). Media used for all cells was supplemented with 2 mM L-glutamine (StemCell Technologies) and the antibiotics penicillin (100 U/ml) and streptomycin (100 ixg/ml) (StemCell Technologies) and all cells were maintained at 37°C in a humidified 5 % C 0 2 atmosphere. BOSC 23 viral packaging cells (obtained from William Pear; Pear et al., 1993) were cultured in DMEM containing 9 % fetal bovine serum (FBS; HyClone). 2.2.2 Construction of NH2-terminally truncated RasGRPI constructs Cloning of RasGRPI, as well as full-length RasGRPI and green fluorescent protein (GFP)-tagged (derived from pEGFP-C1; Clontech, Palo Alto, CA) C1 domain deleted RasGRPI construction, was described previously (Tognon era/., 1998). NH 2 -terminal RasGRPI truncations, collectively denoted RasGRP1/a, were inserted into pCTV3 derivative retroviral expression vectors carrying resistance for either G418 or puromycin (Whitehead etal., 1995). The 5' ends of the NH2-terminal RasGRPI truncations were made through the introduction of base changes at the 5' site of each truncation by PCR, which resulted in a Seal site at the location of each truncation. The PCR-generated restriction enzyme sites at the 3' ends of the NH2-terminal RasGRPI truncations were fused to stop codons, and/or hemagglutinin (HA) epitope tags provided by the pCTV3-derived retroviral vectors. All NH2-terminally truncated 40 RasGRPl constructs included the a helix domain of RasGRPl as well as domains upstream of the a helix domain: 1) the longest construct truncates upstream of the EF hands (EFH/a), followed by 2) truncation upstream of the C1 domain (C1/a), and 3) truncation upstream of a region that was labelled as spacer (sp), a region which is between the C1 domain and a helix (sp/a), and 4) truncation upstream of the a helix (a). Similar to the generation of a prenylation expression vector described previously, the prenylated NH2-terminally truncated RasGRPl expression constructs were generated by in-frame insertion of RasGRPl/a-encoding cDNAs into a pCTV3-derived retroviral vector. This vector contained an oligonucleotide hybrid encoding the prenylation signal downstream of an HA tag (Tognon et al., 1998). GFP was then inserted at the NH2-terminus of the prenylated, sp/a-encoding expression vector. RasGRPl constructs were verified by sequencing, performed by the Nucleic Acids and Protein Sequencing (NAPS) unit at the University of British Columbia (UBC; Vancouver, BC) using ABI Prism (Perkin Elmer, Foster City, CA). 2.2.3 Production of Viruses by Transfection of BOSC 23 cells The BOSC 23 ecotropic packaging cell line was transfected as described previously using the calcium phosphate precipitation method with 4 [ig of pCTV3-derived vectors and 1 pGP3 (gag/pol plasmid) + 1 jag of pAX142 Env-1 (env plasmid) (Pear et al., 1993). Following transfection, BOSC 23 cells were given 2ml fresh DMEM with 9 % FBS every 12 hours. Forty-eight hours after the transfection the viral supernatants were filtered through a 0.22 jam filter unit (Millipore Corp., Bedford, MA) and were used to infect either NIH 3T3 cells (section 2.2.4) or T28 cells (section 3.2.5). 41 2.2.4 Viral Infection of NIH 3T3 cells For viral infection, 50 ju.1 of 48 hour filtered BOSC 23 viral supernatant and 1 ml of media with polybrene (10 |ag/ml) (Sigma, St. Louis, MO) was added to 2 x 10 4 3T3 cells. Twenty hours post-infection, 2 ml fresh DMEM with 9 % CS plus either puromycin (2 y. g/ml) G418 (800 ^ig/ml) or both selection drugs following co-infection were added to generate stably infected cell lines. 2.2.5 Cell solubilization, Immunoprecipitation and Immunoblotting Stable lines of NIH 3T3 cells overexpressing RasGRPI-derived constructs (section 2.2.2) were grown to 90 % confluence in 10 cm dishes containing DMEM with 9 % CS. Cells were then serum-deprived in DMEM for 4 hours at 37°C with 0.1 % (w/v) bovine serum albumin (BSA; Roche, Laval, PQ), followed by stimulation in DMEM that contained 9 % CS for 15 minutes with 0.1 % (w/v) BSA. Cells were washed once in ice-cold 1x Hank's balanced salt solution (StemCell Technologies) and solubilized on ice (1 x 10 6 cells/ml) in phosphorylation solubilization buffer (PSB; 50 mM HEPES, pH 7.4, 100 mM sodium fluoride, 10 mM tetrasodium pyrophosphate, 2 mM sodium orthovanadate, 2 mM EGTA, 2 mM sodium molybdate) with freshly added 0.5 mM phenylmethyl sulfonyl fluoride, 2 jag/ml leupeptin, 2 ^g/ml aprotinin and 0.5 % (v/v) Igepal (Sigma) at 4°C for 1 hour. The resulting cell lysate was subjected to centrifugation at 12 000 g for 10 minutes at 4°C. The rabbit anti-RasGRP1 antibody was raised against the peptide YYTEDEIYELSYAREPRGC, which is derived from a RasGRPI sequence that is between the GEF domain and the first EF hand, and provided in affinity-purified form by Bethyl Laboratories Inc. (Montgomery, TX). The soluble fraction of the cell lysate was 42 incubated with 3 p.g/ml of the rabbit anti-RasGRP1 antibody with mixing at 4°C for 2 hours. Immune complexes were then incubated with 20 pJ (bead volume) Protein A-conjugated agarose beads (Pierce, Rockford, IL) at 4°C for 90 minutes. Immune complexes bound to the Protein A-conjugated agarose beads were washed three times with 1 ml of ice-cold PSB plus 0.1 % (v/v) Igepal, resuspended in 40 pj of SDS-PAGE sample buffer and boiled for 3 minutes. Samples bound to agarose beads or supernatants were then separated on a 12.5 % SDS-polyacrylamide gel and transferred to Immobilon P membranes (Millipore Corp.). To perform immunoblot analysis the membrane was first blocked with 5 % (w/v) BSA that was dissolved in 1 x Tris-buffered saline with Tween 20 (TBST; 10 mM Tris-HCI, pH 7.4, 150 mM sodium chloride, 0.1 % (v/v) Tween 20) overnight at 4°C. The membrane was then incubated with a mouse anti-HA antibody (Covance, Princeton, NJ), diluted 1:5000 in TBST, with 2 % (w/v) BSA for 1.5 hours at RT, followed by further incubation with horseradish peroxidase-conjugated donkey anti-mouse IgG antibody (Jackson Immunoresearch Laboratories, Inc., West Grove, PA), diluted 1:10 000 in TBST, for 1 hour at 23°C. The antigen-antibody complexes were detected by enhanced chemiluminescence following the manufacturer's protocol (Santa Cruz Biotechnology Inc., Santa Cruz, CA). 2.2.6 Fluorescence and Microscopy using GFP NIH 3T3 cells were infected with either control or RasGRPl-derived constructs. Their construction is described in section 2.2.2 and RasGRPl constructs are identified in Figure 2.3a. The infected NIH 3T3 cells were plated onto 22 mm coverslips at a cell density of 1 x 105/coverslip. Cells were fixed with 4 % paraformaldehyde (Fisher 43 Scientific, Nepean, ON) and mounted using VectaShield mounting medium (Vector Laboratories, Inc., Burlingame, CA). Cells were visualized at 63x NA with a PCO SensiCam (black/white) cooled CCD camera (PCO Technology Corp., Kelheim, Germany) mounted on a Zeiss Axioplan2 microscope (Zeiss, Jena, Germany) with a SP-101 FITC filter (Chroma Technology Corp., Brattleboro, VT) where excitation was provided by an excitation filter operating at 470/40 nm with a 522/40 emission filter to detect green fluorescence. Imaging analysis was done with Northern Eclipse software (Empix Imaging Inc., Mississauga, ON). Use of the microscope and imaging software was kindly provided by Peter Lansdorp and facilitated by Elizabeth Chavez, (UBC). Each of the experiments was done in triplicate and at least one experiment from each triplicate was performed blind. For the subcellular relocalization assay, every visualized fixed cell for each cell line was scored as exhibiting either a ubiquitous fluorescence pattern or a nuclear excluded fluorescence pattern based on lack of concentrated fluorescence in the nucleus. Utilizing SPSS statistics software (SPSS Inc., Chicago, IL), the Pearson chi-square test was used to calculate chi-square values and P values. 44 2.3 RESULTS The extreme COOH-terminus of RasGRPl contains an a helix that is a leucine zipper motif, indicating that this motif may mediate dimerization of RasGRPl (Tognon et al., 1998). Hence, we chose to investigate the homodimerization of the RasGRPl leucine zipper. To examine this hypothesis we used two independent approaches: co-immunoprecipitation and subcellular relocalization. We also wished to determine if domains upstream of the a helix might modulate homodimerization. Consequently we generated NH2-terminally truncated RasGRPl constructs of varying sizes that contained the a helix (a). We collectively denoted these constructs RasGRPl /a. In decreasing order of size these constructs were: EFH/a, which truncates upstream of the EF hands (EFH); C1/a, which truncates upstream of the C1 domain; sp/a, which truncates upstream of a region that we have labelled as spacer (sp), a region which is between the C1 domain and a helix; and a, which truncates upstream of the a helix. 2.3.1 An NH2-terminally truncated RasGRPl construct co-immunoprecipitates with full-length RasGRPl. We employed co-immunoprecipitation as a preliminary endeavour to determine if one of the RasGRPl /a constructs, sp/a, could interact with the full-length form of RasGRPl (RasGRP1/FL). The RasGRPl /a construct was COOH-terminally hemagglutinin (HA) epitope tagged (sp/a-HA) and used to infect NIH 3T3 cells either alone or it was co-infected with RasGRPl/FL. Total cell lysates were immunoprecipitated with the anti-RasGRP1 antibody, which was directed against full-length RasGRPl because it recognizes a region that is upstream of the sp/a construct 45 a. RasGRP1/FL sp/a-HA HA b. RasGRP1/FL + RasGRP1/FL IP: anti-RasGRP1 IB: anti-HA Figure 2.2. The RasGRPI sp/a construct co-immunoprecipitates with full length RasGRPI. T h e R a s G R P I s p / a - H A construct was co-expressed with R a s G R P I ( R a s G R P I / F L + ) or expressed a lone ( R a s G R P I / F L - ) . Total cell lysates were co- immunoprec ip i ta ted (IP) wi th an ant ibody to the amino- termina l port ion of R a s G R P I (ant i -RasGRP1) and immunob lo t ted ( IB) with an ant i -HA ant ibody (ant i -HA). a. D iagram of domains within the ful l - length and s p / a - H A const ruct of R a s G R P I and reg ions to wh ich the ant ibodies b ind. b. An t i -HA IB of co- immunoprec ip i ta ted samples (top panels, s a m e IB) and of total cell lysates (bot tom panels, s a m e IB). A n independent exper iment gave similar results. GEF:GEF doma in ; EFH: EF hands; C1: C1 doma in ; sp: spacer region; a: a lpha helix; HA: HA tag. 4 6 (see Figure 2.2a), followed by detection on an immunoblot with the anti-HA antibody (anti-HA). Figure 2.2b shows that the sp/a-HA construct co-precipitated with the full-length RasGRPl protein. However, in the absence of coexpression with the RasGRPl/FL construct no signal was detected when co-immunoprecipitation was performed showing that precipitation required full-length RasGRPl. The level of sp/a-HA expression within the same total cell lysates was detected following anti-HA immunoblotting. These data suggest that a specific interaction between the full-length form of RasGRPl and the sp/a construct did occur. 2.3.2 NH2-terminally truncated RasGRPl constructs alter the subcellular localization of a RasGRPl construct. Following the initial demonstration of RasGRPl constructs interacting via co-immunoprecipitation we continued the study of RasGRPl homodimerization by observing altered subcellular localization in fixed mammalian cells. This relocalization assay used a GFP tag to track the subcellular movement of proteins to which it is fused. A protein-protein interaction can be detected by this method through the subcellular relocalization of a GFP-tagged construct upon coexpression with another construct. Hence, this assay requires the use of two types of distinctly localized RasGRPl constructs. The predominant localization pattern for the GFP-tagged construct was ubiquitous distribution throughout the cell. The construct we used was the GFP-tagged, C1 domain-deleted form of RasGRPl (GFP-RasGRP1dC1). The other type of construct was nuclear-excluded and had a specific punctate fluorescence at the plasma membrane. The plasma membrane-targeting prenylation signal from K-Ras created this distinct subcellular localization pattern. To generate the second type of construct the prenylation signal (Pr) was fused to each of the four RasGRPl /a constructs, which 47 resulted in the following constructs: EFH/a-Pr, C1/a-Pr, sp/a-Pr and a-Pr. In order to exhibit the second type of localization pattern, a GFP-tagged version of a prenylated RasGRP1/ct construct was generated and its localization pattern is shown in Figure 2.3b. The detection of a RasGRPl protein-protein interaction between these two types of distinctly localized constructs was through the relocalization of GFP-RasGRP1dC1 upon coexpression with a RasGRPl/a-Pr construct. NIH 3T3 cells were co-infected with GFP-RasGRP1dC1 and either a pCTV3-derived control vector (vector) or with each of the four RasGRPl/a-Pr constructs. For each cell line, every visualized cell was observed to have either a ubiquitous subcellular distribution pattern, like that which predominated for GFP-RasGRP1dC1 when coexpressed with control vector, or a nuclear-excluded subcellular distribution pattern, which was displayed by lack of concentrated fluorescence in the nucleus. As a result, the nuclear-excluded localization pattern was identified as the GFP-RasGRP1dC1 relocalization pattern and the proportion of this localization pattern was determined for each cell line in order to calculate if its incidence was significantly different from that observed for GFP-RasGRP1dC1 when coexpressed with the control vector. A representation of each of the two types of subcellular distribution patterns for each cell line that coexpressed the RasGRPl constructs is given in Figure 2.3c. The chi-square test for association was used to test for a significant difference among the proportions of the relocalization pattern between the cell line expressing the GFP-tagged RasGRPl construct and the control vector, and each of the cell lines that co-expressed the GFP-tagged RasGRPl construct with each of the a helix-containing prenylated constructs. The results are presented in Table 2.1. Within the cell lines that coexpressed GFP-RasGRP1dC1 and each of the four RasGRPl/a-Pr constructs the proportion of nuclear-excluded subcellular localization 48 Constructs Nuclear-exc luded cells Total % Chi-square P va lue G F P - R a s G R P 1 d C 1 coexpressed with : Contro ls : Vector 28 229 12.2 — — Pr 21 159 13.2 0.082 0.775 R a s G R P 1 / a - P r const ructs : EFH/a-Pr 48 132 36.4 29.350 <0.001 sp/a-Pr 144 252 57.1 105.364 <0.001 C1/a-Pr 172 402 42.8 62.931 <0.001 a-Pr 48 186 25.8 12.652 <0.001 Table 2.1. Proportion of relocalization upon coexpression of GFP-tagged RasGRPldC1 with prenylated NH2-terminally truncated RasGRPl constructs in NIH 3T3 cells. Coexpressed constructs are the s a m e as those in Figure 2.3. NIH 3T3 cells coexpress ing G F P -R a s G R P 1 d C 1 with R a s G R P 1 / a - P r constructs were scored as having either a nuc lear -exc luded or ubiqui tous localization pattern. Chi -square values were calculated using the Pearson chi -square test (SPSS) . was significantly higher than when GFP-RasGRP1dC1 was coexpressed with the control vector, which is evident by the negligible P values (Table 2.1). Thus, the presence of the RasGRPl/a-Pr constructs was altering the subcellular localization of GFP-RasGRP1 dC1. We excluded the possibility that the prenylation signal was interacting with GFP-RasGRP1dC1 by coexpressing the K-Ras prenylation vector (Pr) with GFP-RasGRP1dC1. This cell line did not exhibit a significantly increased proportion of nuclear excluded subcellular localization pattern compared to when GFP-RasGRP1dC1 was coexpressed with the control vector (vector), which is shown in Table 2.1. We also excluded the possibility that the prenylated NH2-terminally truncated RasGRPl constructs were interacting with GFP by coexpressing each of the four RasGRPl constructs with a pCTV3-derived GFP control vector and not observing an alteration in the subcellular relocalization pattern compared to when GFP was expressed alone (data not shown). Therefore, we interpret these results to indicate that the membrane-bound RasGRPl la constructs are drawing RasGRPl dC1 out of the nucleus via a direct protein-protein interaction. 4 9 Figure 2.3. The prenylated NH2-terminally truncated RasGRPl constructs alter the subcellular localization of a GFP-tagged RasGRPl construct. A G F P - t a g g e d , C1 d o m a i n -dele ted R a s G R P l ( G F P - R a s G R P 1 d C 1 ) construct was coexpressed wi th ei ther a vector control (vector) or wi th prenylated NH 2 - te rmina l ly t runcated R a s G R P l constructs R a s G R P 1 / a - P r in N IH 3T3 f ibroblasts. T h e cells were g rown in se rum contain ing media, f ixed and pho tographed under UV i l luminat ion as descr ibed in sect ion 2.2.7. a. Schemat ic d iagram of G F P - R a s G R P 1 d C 1 and the four R a s G R P 1 / a - P r constructs. EFH: EF hands; C1: C1 doma in ; sp: spacer reg ion; a: a lpha helix; Pr: prenylat ion s ignal , b. A r row indicates punctate f luorescence at the p lasma m e m b r a n e for cel ls overexpress ing the G F P - s p / a - P r construct , c. The two types of subcel lu lar distr ibut ion, ubiqui tous and nuc lear-exc luded, are depic ted for each cell line that coexpresses the G F P -R a s G R P 1 d C 1 const ruct and a R a s G R P 1 / a - P r construct . 51 2.4 DISCUSSION We endeavoured to demonstrate by two independent methods, co-immunoprecipitation and a subcellular relocalization assay that RasGRPl homodimerizes. When the results from these methods are examined together, we suggest that RasGRPl interacts with itself. Furthermore, both the use of NH 2 -terminally truncated RasGRPl constructs and the combination of the results from the approaches suggest that the region of interaction is confined to downstream of the C1 domain. Therefore^ the leucine zipper motif or a helix, which has the propensity to mediate homodimerization, is the most likely candidate to mediate this protein association. The results from the two approaches are now considered in greater detail. By use of the approaches we gathered a number of pieces of supporting evidence that RasGRPl interacts with itself. The sp/a construct was shown to interact with full-length RasGRPl via co-immunoprecipitation. In addition, all four NH 2 -terminally truncated RasGRPl constructs demonstrated an interaction via the subcellular relocalization assay. Conversely, there are several aspects from both our methods and our results that detract from our ability to conclusively assert that RasGRPl interacts with itself. To begin with, co-immunoprecipitation may detect indirect interactions. Furthermore, the subcellular relocalization of one protein with another may indicate a number of events other than a direct protein-protein interaction. For example, the presence of the RasGRPl la constructs may indirectly result in the relocalization of the GFP-RasGRP1dC1 construct perhaps by propagating a signalling pathway that draws the GFP-tagged RasGRPl construct out of the nucleus. In addition, the small proportion of GFP-RasGRP1dC1 that is nuclear-excluded needs to be accounted for because it questions our ability to interpret the increased relocalization of this construct as 52 indication that RasGRPI associates with itself. A likely possibility is that the different localization patterns are the result of varying expression levels of the GFP-tagged construct within NIH 3T3 cells. Therefore, the results from each of the approaches must be taken together in order to fortify our claim that RasGRPI homodimerizes. Our hypothesis was that RasGRPI homodimerizes at the a helix based on its consensus with a leucine zipper motif. The most convincing result in support of our hypothesis was the demonstration of an interaction using the isolated a helix construct within the subcellular relocalization assay. Additionally, the results from the subcellular re localization assay provided no indication that domains upstream of the a helix modulate homodimerization. In order to narrow down the targeted area of interaction additional experiments are needed. While our results do imply RasGRPI homodimerization and do suggest that the a helix is involved, the use of longer RasGRPI constructs within our assays, one that is the full-length RasGRPI construct and the other that is a C1 domain-deleted construct, do not permit us to conclude that the a helix of RasGRPI forms a leucine zipper. As such, the use of two isolated a helix-containing constructs, each of which would be distinctly tagged, within the co-immunoprecipitation assay would make further headway in concluding whether the a helix mediates RasGRPI homodimerization. Moreover, the generation and use of a construct that deleted the a helix within the approaches would assess the requirement of this domain in mediating the homodimerization of RasGRPI. By means of the two approaches, which may provide evidence that the RasGRPI a helix homodimerizes, we attempted to present the possibility that RasGRPI homodimerization has functional consequences. Notably, these approaches 53 were artificial and did not address whether RasGRPI naturally exists as dinners. Thus, there is the possibility that homodimerization is an artefact of the RasGRPI a helix and bears no functional significance. Nonetheless, a variety of leucine zippers exist in signalling molecules that do have a functional significance, an example of which is the leucine zipper found in the adaptor protein Cbl (Bartkiewicz et al., 1999). Hence, if homodimerization of RasGRPI does have a functional significance it may act as either a negative or a positive regulator of this protein. As a negative regulator, homodimerization may serve to sequester RasGRPI from its usual sites of activity. As a positive regulator, homodimerization of RasGRPI may increase the avidity of RasGRPI for membranes. Binding of the RasGRPI C1 domain to diacylglycerol in membranes is required for its activation (Ebinu etal., 1998; Tognon et al., 1998). Homodimerization may serve to increase the avidity of RasGRPI for membranes by doubling C1 domain contact points. Therefore, we may speculate that this results in increased RasGRPI activation. This membrane avidity model is supported by the function of dimerization in leucine zipper-containing transcription factors, which is to achieve strong specific binding with DNA. In conclusion, we have demonstrated by two independent approaches for protein-protein interaction that RasGRPI interacts with itself. Additionally, while our results are compatible with the mediation of this protein-protein interaction by the RasGRPI a helix, they do not unambiguously prove this hypothesis. Furthermore, with the use of NH2-terminally truncated constructs, we have been able to exclude the possibility that any domain upstream of the a helix of RasGRPI is modulating this protein-protein interaction. 54 Chapter 3 COMPARISON OF RASGRP1 AND CALDAG1 C1 DOMAIN-MEDIATED MEMBRANE RECRUITMENT 3.1 INTRODUCTION Diacylglycerol (DAG) is a lipid second messenger that plays a central role in signalling pathways by recruiting C1 domain-containing proteins to it. It is a product of phospholipase C (PLC)-catalysed hydrolysis of either phosphatidylinositol 4,5-bisphosphate (PIP 2) or phosphatidylcholine (PC) (Wakelam, 1998). The paradigmatic example of C1 domain-containing proteins that are regulated by DAG-induced translocation is the protein kinase C (PKC) family of serine/threonine kinases. The C1 domains of several PKCs directly bind to DAG (Ono etal., 1989; Kaibuchi etal., 1989; Castagna etal., 1982; Hannun etal., 1985; reviewed by Hurley etal., 1997); however, not all PKC family members have appropriate C1 domains in order to act as DAG receptors. Those that are responsive to DAG are termed typical C1 domains. Typical C1 domains are contained within the a, p i , p2, y, 8, 8, r| and 9 PKC isozymes, whereas non-DAG responsive atypical C1 domains are contained within the C, and Xk isozymes (Hurley et al., 1997). Moreover, the major effector of Ras, Raf, in the MAPK cascade contains the best-characterized atypical C1 domain because it is known to bind phosphatidylserine and not DAG (Ghosh etal., 1996). Depending on the PKC isozyme, it may contain either two typical C1 domains, or a single atypical C1 domain. However, a single typical C1 domain has been shown to be sufficient to confer binding to DAG mimetics such as phorbol esters (Ono er al., 1989; Burns and Bell, 1991; Quest and Bell, 1994; Quest etal., 1994). The use of phorbol esters is particularly advantageous to study both the subcellular relocalization 55 and the in vitro ligand binding affinity of C1 domains because these DAG mimetics have a higher affinity for C1 domains than DAG itself (Burns and Bell, 1991; Quest et al., 1994; Baatout et al., 1997). A preliminary sign that a single C1 domain was sufficient for DAG or DAG mimetic binding came from the crystal structure determination of the second typical C1 domain of PKC8 (PKC82 C1) in complex with the phorbol ester phorbol 13-acetate (Kazanietz etal., 1995). This structural determination showed that the C1 domain of PKC82 displays a hydrophobic surface interrupted by a hydrophilic cleft. Interaction of either DAG or phorbol esters with the C1 domain involves their insertion into this cleft creating a contiguous hydrophobic region that promotes hydrophobic interaction of C1 domains with the membrane, which creates a C1/phospholipid/ligand ternary complex (Kazanietz etal., 1995; Zhang etal., 1995; Hurley et al., 1997). In addition to the PKC family of serine/threonine kinases, there are four other receptor families that contain DAG/phorbol ester-responsive C1 domains. They are: 1) protein kinase D (PKD), which is a serine/threonine kinase related to PKC that has unique substrate specificity; 2) the chimaerins, which are inhibitors of the small GTP-binding protein Rac; 3) the Caenorhabditis elegans Unc-13 family, which act as scaffold structures for exocytic proteins; 4) and the RasGRP subfamily, which are guanine nucleotide exchange factors (GEFs) that activate Ras and the Ras-related GTPases (Ron and Kazanietz, 1999; Kazanietz, 2000). Studies of DAG mimetic ligand binding with these non-PKC DAG/phorbol ester receptor families have uncovered heterogeneity among typical C1 domains. These studies have demonstrated differences in both ligand selectivity and phospholipid cofactor preferences between PKCs, chimaerins, and RasGRPs (Caloca etal., 2001; Lorenzo etal., 2000; Shao etal., 2001). Effectors for the RasGRP subfamily, the Ras and Ras-related family of 56 GTPases, are tethered to membranes requiring that this subfamily of Ras GEFs be recruited to membranes in order to effect their GEF activity. Hence, one possibility is that the C1 domains contained within each of the RasGRP subfamily members are acting as the chief positive regulatory domains by mediating the translocation of these Ras GEFs to their GTPase targets. A number of phorbol ester-enhanced effector activity studies have been done to explore the positive regulatory hypothesis for the C1 domains contained within this Ras GEF subfamily. In particular, phorbol esters have been implicated in increasing the effector activity of RasGRPl, which was the first RasGRP subfamily member to be identified. For cells overexpressing RasGRPl, phorbol esters have been shown to enhance the amount of GTP-bound Ras, increase the level of Ras effectors such as ERK1/2 and increase the ability of fibroblasts to become transformed under low serum conditions (Tognon etal., 1998; Ebinu etal., 1998; Kawasaki etal., 1998). Comparing this phorbol ester-induced effect to the second RasGRP subfamily member, CalDAGl, it was found that, for cells overexpressing CalDAGl, phorbol ester enhanced the amount of GTP binding to the CalDAGl effector RaplA and phorbol ester was also found to increase the growth rate of FDCP1 myeloid cells that overexpressed CalDAGl (Kawasaki etal., 1998; Dupuy et al., 2001). To complement the phorbol ester-enhanced effector activity data, phorbol ester-induced membrane translocation studies for RasGRPl have been done. The RasGRPl C1 domain has been shown to target this Ras GEF to membranes enriched in DAG or phorbol ester in fibroblasts and in vitro studies have reported that the RasGRPl C1 domain is a direct, high-affinity target of phorbol esters (Tognon et al., 1998; Ebinu etal., 1998; Lorenzo etal., 2000). In contrast, the phorbol ester-induced membrane localization for the C1 domain of CalDAGl has not been previously directly determined. 57 Solution of the C1 domain structure facilitated the identification of residues within the C1 domain sequence that are predicted to be critical for completing the C1/phospholipid/ligand ternary complex (Kazanietz et al., 1995; Zhang etal., 1995). The RasGRPI C1 domain sequence contains all consensus residues that are predicted to be critical for completing this ternary complex. The CalDAGI C1 domain sequence contains the residues that are required for DAG/phorbol ester binding, however it is missing two residues that are predicted to be required for non-specific membrane interaction. Based upon the differences between these two C1 domain sequences, we sought to compare membrane translocation of these two Ras GEFs in response to various stimuli. In order to examine membrane translocation responses, the isolated C1 domains of these two Ras GEFs were tagged with green fluorescent protein and the fusion constructs were overexpressed in NIH 3T3 cells. We demonstrated that there exists a distinct biochemical difference between the C1 domains of RasGRPI and CalDAGI. The RasGRPI C1 domain demonstrated distinct membrane translocation in response to the phorbol ester phorbol 12-myristate 13-acetate (PMA) and DAG-generating phospholipase C. In contrast, the CalDAGI C1 domain did not show any membrane translocation in response to these stimuli. Notably, the RasGRPI C1 domain did not translocate to membranes in response to another phorbol ester, phorbol 12,13-dibutyrate, and a DAG, DiC8, whereas the PKC82 C1 domain did translocate in response to these stimuli. Because both RasGRPI and CalDAGI are expressed in the murine T cell hybridoma T28 cells, we also examined membrane recruitment in this cell line (Tognon etal., 1998; Kay, unpublished data). Upon overexpression of each of the C1 domains in T28 cells, it was found that both the RasGRPI and CalDAGI C1 domains translocated to membranes following PMA treatment, yet the CalDAGI C1 58 domain demonstrated less concentrated membrane fluorescence. To determine upstream signalling events that induced membrane translocation in the T28 cell line, the T cell receptor (TCR) was stimulated with the natural ligand, murine anti-CD3s. Both RasGRPl and CalDAGl demonstrated C1 domain-dependent translocation to the plasma membrane in response to anti-CD3s stimulation. This result indicated that these Ras GEFs exhibited similar migration patterns in response to TCR crosslinking. However, the isolated CalDAGl C1 domain did not translocate to the plasma membrane in response to anti-CD3£ stimulation indicating that the CalDAGl C1 domain is insufficient but necessary to mediate TCR-induced translocation the plasma membrane. Our results show that the PKC52 C1 contains the greatest translocation ability in response to the ligands tested, followed by the RasGRPl C1 domain and then by the CalDAGl C1 domain. Therefore, there exists a functional non-equivalence between the RasGRPl and CalDAGl C1 domains. 59 3.2 METHODS 3.2.1 Materials Anti-CD3s was obtained from Cedarlane (Hornby, ON). 1,2-dioctanoyl sn-glycerol (DiC8), phorbol 12,13-dibutyrate (PDBu), phorbol 12-myristate 13-acetate (PMA) and Bacillus cereus phosphatidylcholine-specific phospholipase C (PC-PLC) were obtained from Sigma (St. Louis, MO). 3.2.2 Cell culture For NIH 3T3 and BOSC 23 cell lines see Section 2.2.1. T28 cells (were originally from Dr. K. Rock, Dana Farber Institute, Boston, MA; Pyszniak etal., 1994) were cultured at 37°C in DMEM containing 9 % fetal bovine serum (FBS; HyClone, Logan, UT) supplemented with 2 mM L-glutamine (StemCell Technologies, Vancouver, BC) and the antibiotics penicillin (100 U/ml) and streptomycin (100 ug/ml) (StemCell Technologies), in a humidified 5 % CO2 atmosphere. 3.2.3 Cloning of murine CalDAGl Murine RasGRPl was cloned as described in Tognon etal., (1998). Murine RasGRPl cDNA sequence was aligned to the GenBank Expressed Sequence Tag (EST) database using NCBI BLASTN (website: http://www.ncbi.nlm.nih.gov). Murine full-length CalDAGl cDNA was constructed from two of the EST clones aa790222 and ai l 54408 obtained from the GenBank EST database search, sequenced to completion with gene-specific primers and confirmed by DNA sequencing, which was performed by the Nucleic Acids and Protein Sequencing (NAPS) unit at the University of British Columbia (UBC; Vancouver, BC) using ABI Prism (Perkin Elmer, Foster City, CA). Murine ESTs identified through BLAST searches were purchased from Genome 60 Systems Inc. (St. Louis, MO). The CalDAGI cDNA GenBank Accession no. is AF081193. 3.2.4 Construction of RasGRPI and CalDAGI deletion and mutant constructs Green fluorescent protein (GFP)-tagged (derived from pEGFP-C1; Clontech, Palo Alto, CA) full length murine RasGRPI (GFP-RasGRP) and GFP-tagged C1 domain deleted RasGRPI (GFP-RasGRPdC1), as well as GFP-tagged isolated C1 domains of murine PKC52 (GFP-PKC5-C1) and RasGRPI (GFP-RasGRP-C1) were generated as described previously (Tognon etal., 1998). The GFP-tagged murine CalDAGI C1 domain (GFP-CalDAG1-C1) and the human Raf C1 domain (GFP-Raf-C1) were generated in a similar manner. The primer pairs were designed with restriction enzyme sites that allowed for in-frame cDNA subcloning and the reverse primer contained a RasGRPI-derived sequence that added KKRIKPT to the COOH-terminus of each C1 domain. Deletion of the CalDAGI C1 domain, which resulted in the GFP-CalDAG1dC1, was made by using the GFP-CalDAG1 construct as template and fusing two PCR fragments that introduced base changes at the site of the fusion, which resulted in a Seal site in place of the C1 domain. GFP fusion proteins were produced by subcloning PCR-generated constructs downstream of a GFP tag into pCTV3 derivative retroviral expression vectors that carried resistance for either G418 i puromycin (Whitehead etal., 1995). All PCR-generated fragments were verified by sequencing, which was performed by the NAPS unit, UBC, using ABI Prism. 3.2.5 Viral Infection of T28 cells For details on production of viruses by transfection of BOSC 23 cells and viral infection of NIH 3T3 cells please refer to section 2.2.3 and section 2.2.4, respectively. 61 For infection of T28 cells, 5 x 10 4 T28 cells that were suspended in 500 (J media were mixed with 500 uJ of 48 hour filtered BOSC 23 viral supernatant and polybrene (Sigma) was added to a final concentration of 10 p.g/ml. 3.2.6 Fluorescence and Microscopy using GFP in T28 cells Please refer to section 2.2.6 for Fluorescence and Microscopy using GFP. For T28 cells, 22 mm coverslips had been pre-soaked in 0.5 mg/ml poly-L-lysine (Sigma) for 1 h and rinsed in d H 2 0 three times prior to plating onto 22 mm coverslips at a cell density of 1 x 105/coverslip. All stimuli (section 3.2.1) were diluted to the specified concentrations in PBS (StemCell Technologies). 62 3.3 RESULTS 3.3.1 Comparison of C1 domain sequences In Figure 3.1, the RasGRPI and CalDAGI (RasGRP2) C1 domains sequences were aligned with other C1 domains, both typical (PKCa2 and PKC52) and atypical (PKCC and Raf). All C1 domains have the topology HX 1 2CX2CX 13/14CX2CX4HX2CX7C; where H is histidine, C is cysteine, and X is any other amino acid, which permits the coordination of two Z n 2 + ions into the C1 domain structure. However, the presence of other specific conserved residues confers DAG or phorbol ester binding ability and membrane interaction (Hurley etal., 1997; Hurley and Misra, 2000). The RasGRPI C1 domain agrees with the consensus sequence since it contains conserved residues at each of the 9 positions predicted to be required for the C1/phospholipid/ligand ternary complex. The C1 domains of RasGRP3 and RasGRP4 are included for comparison in Figure 3.1. They also have the 9 conserved residues characteristic of typical C1 domains. In addition, studies have shown that RasGRP3 is a high-affinity target of phorbol esters and that RasGRP4 translocates to membranes in response to the phorbol ester PMA (Lorenzo et al., 2001; Reuther et al., 2002). The CalDAGI C1 domain contains consensus residues Pro11, Gly23, and Gln27 that are believed to maintain the precise structure of the DAG/phorbol ester binding groove, yet it is missing two conserved aromatic residues, at positions 8 and 13, believed to be required for non-specific membrane interaction (Kazanietz et al., 1995; Zhang et al., 1995). However, one recent report determined by mutational analysis that the conserved aromatic residue at position 13 was not required for PKC82 C1 domain binding to phorbol ester (Wang et al., 2001). Due to the deviations from the consensus typical C1 domain sequence it is not clear whether or not the CalDAGI C1 domain binds phorbol esters and DAG. 63 aa 1-25 1. :- H 1 N F Q E T T El L 2. H; N F Q E S N S L 3. H K F K 1 H T Y G 4. H R F K V Y N Y M 5. H L F Q A K R F N 6. H ; N F A R K T F L 7. • Hj N F Q E M T Y L 8. [ H T F H E V T F R aa no. 5 A G F L W G V 1 K A L 1 L G 1 Y G S L L Y G L 1 G s L L W G L V S E R L W G L S Q K | H L WW A G • L W G 1 1 S G L W G V T 20 25 aa 26-50 1. 2. 3. 4. 5. 6. 7. 8. aa no. Y R K D cl G M N C H K Q r c - i K D L V V F E ""gl L K I C,j R A cj G V N C H K Q : C> K D R L s V E c M K D T c ; D M N V H K Q C V 1 N V p S L c L K c E D c G M N V H H K C R E K V A N L c Y R c I N Cj K L L V H K R C i H V L V P L T C; F R c Q T cl G Y K F H E H j C j S T K V P T M c Y K c K D c ! G A N C H K Q 1 C 1 K D L L V L A c Y R 30 .,-C.i R E J= j G 35 L C c H K 40 H t . Q j R D Q 45 V K V E 50 Figure 3.1 Comparison of C1 domain sequences. The C1 domains of the R a s G R P l subfami ly m e m b e r s ; R a s G R P m e m b e r s ; R a s G R P l (line 1.), R a s G R P 2 (line 2.), R a s G R P 3 (line 7.) and R a s G R P 4 (line 8.), are al igned with both typical ( P K C a (line 3.) and PKC5 (line 4.)) and atypical (PKC<; (line 5.) and Raf (line 6.) C1 domains . Typical C1 domains are classi f ied based on their ability to bind D A G and phorbol esters. All C1 domains contain conserved cyste ines (C) and hist idines (H) (grey) that are required for zinc binding. Addit ionally, typical C1 doma ins conta in conserved amino acid (aa) residues (black) predicted to fo rm the C1/phosphol ip id / l igand ternary complex . 64 3.3.2 Comparison of RasGRPI and CalDAGI C1 domain-mediated membrane recruitment in NIH 3T3 cells. To address whether these sequence variations led to DAG or phorbol ester binding differences between the C1 domains of RasGRPI and CalDAGI, their subcellular relocalization in response to stimuli that either produced or mimicked DAG was assayed. Accordingly, the C1 domains were PCR-generated and fused to GFP at their NH2-termini. They also contained a RasGRPI-derived basic sequence (KKRIK) at their COOH-termini that was hypothesized to aid in membrane binding by interacting with negatively charged phospholipid head groups. NIH 3T3 cells were then produced that stably overexpressed either a GFP control construct or GFP fusions to the NH2-termini of single C1 domains. A GFP tag was employed because it can be used to track the movement of proteins inside a cell when contained within a fusion protein. In addition to GFP fusions of the RasGRPI and CalDAGI C1 domains, the second PKC8 C1 domain (PKC82 C1) was used as the archetypal DAG/phorbol ester binding C1 domain. As well, the Raf C1 domain was employed within our studies because it represented the best-characterized atypical C1 domain. Other lipids mediate this C1 domain's membrane recruitment: the Raf C1 domain can bind phosphatidylserine in membranes (Ghosh et al., 1996). RasGRPI and CalDAGI demonstrate different translocation patterns in response to the phorbol ester PMA in NIH 3T3 cells. We have previously reported that the RasGRPI C1 domain exhibited a nuclear-excluded fluorescence pattern with concentrated fluorescence found at the endoplasmic reticulum and Golgi in serum-deprived, unstimulated NIH 3T3 cells, as well as cells grown in media that contains serum (Tognon et al., 1998). From these 65 results, it was hypothesized that the RasGRPI C1 domain was binding DAG in these internal membranes because certain PKC isoforms localize to the endoplasmic reticulum and Golgi following PMA stimulation (Goodnight et al., 1995). In contrast, the CalDAGI C1 domain exhibited a more diffuse fluorescence pattern without concentrated fluorescence at subcellular organelles in both unstimulated, serum-deprived NIH 3T3 cells as well as those grown in serum (Figure 3.2, column 4). To study the effect of phorbol esters on the translocation of these two Ras GEF C1 domains, membrane recruitment responses to the prototypical phorbol ester PMA were compared for the GFP-tagged RasGRPI C1 domain (GFP-RasGRP-C1) and the GFP-tagged CalDAGI C1 (GFP-CalDAG1-C1) domain in NIH 3T3 cells. Our previous findings demonstrated that the RasGRPI C1 domain mediated rapid and transient plasma membrane translocation in response to PMA stimulation in NIH 3T3 cells (Tognon etal., 1998). Similarly, GFP-RasGRP-C1 exhibited rapid and transient translocation to the plasma membrane: concentrated membrane fluorescence was observed upon 1 minute of PMA stimulation, yet was not apparent after 15 minutes of stimulation (Figure 3.2 column 3). The equivalent membrane translocation of the PKC52 C1 domain (GFP-PKC5-C1; Figure 3.2 column 2) when compared to the RasGRPI C1 domain emphasizes that the C1 domain contained within this Ras GEF is a direct target of phorbol esters. In contrast, the same rapid and transient membrane localization after PMA stimulation was not observed for the CalDAGI C1 domain (Figure 3.2 column 4). As expected, the GFP protein alone (GFP) was ubiquitously distributed throughout cytoplasm with concentrated fluorescence in the nucleus in response to all stimuli tested (Figures 3.2-3.8 column 1) and the human Raf C1 domain (GFP-Raf-C1) did not translocate to membranes in response to any of the stimuli tested (Figures 3.2-3.6 column 5). To begin to explore the possibility that the CalDAGI C1 66 domain had a lower affinity for phorbol esters than the RasGRPl C1 domain, increased concentrations of PMA were used to stimulate C1 domain overexpressing cells. However, for 800nM and 2nM of PMA no detectable translocation pattern could be observed for the CalDAGl C1 domain (data not shown). GFP GFP-PKC5-C1 GFP-RasGRP-C1 GFP-CalDAG1-C1 GFP-Raf-C1 Figure 3.2 The CalDAGl C1 domain does not translocate to the plasma membrane in response to PMA stimulation in NIH 3T3 cells. Either G F P alone (GFP) or G F P fusions to the N termini of isolated C1 doma ins were stably overexpressed in NIH 3T3 f ibroblasts, se rum-s ta rved for 3.5 hours or serum-starved then st imulated for the indicated t imes with 80 nM PMA. T h e cells were then f ixed and photographed under UV i l lumination. Ar rows indicate regions of concent ra ted f luorescence. Similarly, in an attempt to identify other C1 domain ligands that induced membrane translocation of these Ras GEF C1 domains, the potent PKC-activating phorbol ester, phorbol 12,13-dibutyrate (PDBu) was used. PDBu is a smaller and less 67 hydrophobic phorbol ester than PMA, which allows it to equilibrate more rapidly between the plasma membrane and internal membranes. This structural trait emphasizes its efficacy for identifying internal membrane targets for C1 domains (Oancea et al., 1998). Rapid and transient plasma membrane and nuclear membrane translocation (at 1 minute) upon PDBu stimulation of the PKC82 C1 domain was observed (Figure 3.3 column 2). Strikingly, in Figure 3.3 column 3 neither nuclear nor plasma membrane translocation was observed for the RasGRPI C1 domain after PDBu addition despite the equivalent translocation patterns of these two domains in response to PMA stimulation (see also Figure 3.2 column 3). Additionally, like PMA stimulation, PDBu stimulation did not discernibly effect CalDAGI C1 domain membrane translocation (Figure 3.3 column 4). 68 PDBu V 10 GFP GFP-PKCS-C1 GFP-RasGRP-C1 GFP-CalDAG1-C1 GFP-Raf-C1 Figure 3.3. The RasGRPI and CalDAGI C1 domains do not translocate to the plasma membrane in response to phorbol 12,13-dibutyrate stimulation in NIH 3T3 cells. Either G F P alone (GFP) or G F P fusions of isolated C1 doma ins were stably overexpressed in NIH 3T3 f ibroblasts, serum-starved for 3.5 hours and st imulated for the indicated t imes with 1uM phorbol 12,13-dibutyrate (PDBu) . T h e cells were then f ixed and photographed under UV i l lumination. Ar rows indicate regions of concent ra ted f luorescence. RasGRPI and CalDAGI C1 domain translocation patterns in response to PC-PLC treatment or DiC8 stimulation in NIH 3T3 cells. The effect of DAG on the subcellular relocalization of the RasGRPI and CalDAGI C1 domains was studied by treating C1 domains with either exogenous bacterial phosphatidylcholine-specific phospholipase C (PC-PLC) or a membrane-permeant synthetic diacylglycerol, 1,2-dioctanoyl sn-glycerol (DiC8). Through cleavage of phosphatidylcholine in the outer leaflet of the plasma membrane, PC-PLC produces 69 large quantities of diacylglycerol at the inner leaflet, which then rapidly redistributes to internal structures in the cytoplasm (Besterman etal., 1986, Larrodera etal., 1990). Figure 3.4 demonstrates that after 45 minutes of PC-PLC treatment, the C1 domains of RasGRPI and PKC82 exhibit a punctate fluorescence pattern. These results mirror those previously seen in response to PC-PLC treatment in NIH 3T3 cells (Tognon etal., 1998). The C1 domains appear to have relocalized to numerous small spherical structures within the cell. These structures may be equivalent to the lipid droplets that show a concentrated fluorescence pattern when fibroblasts are treated with fluorescently tagged phosphatidic acid, which is metabolised to diacylglycerol and then triacylglycerol once inside the cell (Pagano etal., 1983). The RasGRPI and PKC82 C1 domains apparently bind a component of these lipid droplets (presumably DAG) that accumulates following prolonged PC-PLC treatment. In contrast, cells that overexpressed CalDAGI C1 domain did not demonstrate the same subcellular redistribution pattern in response to PC-PLC, despite the prolonged highly concentrated treatment (Figure 3.4 column 4). Following treatment of the C1 domains with the PKC-activating, short chain diacylglycerol, DiC8 (Lapetina etal., 1985) we observed that the PKC82 C1 domain translocated very rapidly and transiently to both the plasma and nuclear membranes (Figure 3.5 column 2). Other PKC C1 domains have also translocated to membranes within seconds of DiC8 stimulation (Oancea etal., 1998). However, stimulation with this DAG did not result in detectable migration of either the RasGRPI or CalDAGI C1 domains for the indicated time points (Figure 3.5 columns 3 &4, respectively). The results in response to both PMA stimulation and PC-PLC treatment indicated that the RasGRPI and CalDAGT C1 domains behaved in a biochemically 70 GFP GFP-PKC6-C1 GFP-RasGRP-C1 GFP-CalDAG1-C1 GFP-Raf-C1 Figure 3.4. The CalDAGl C1 domain does not respond to DAG-generating PC-PLC treatment in NIH 3T3 cells. Either G F P a lone or G F P fus ions o f isolated C1 doma ins were stably overexpressed in NIH 3T3 f ibroblasts, serum-s tarved for 3.5 hours, t reated wi th 4U/ml Bacillus cereus P C - P L C (S igma) and cul tured for 45 minutes at 37°C. T h e cells were then f ixed and photographed under U V i l luminat ion. A r rows indicate regions of concent ra ted f luorescence. distinct manner in NIH 3T3 cells. This suggests that in contrast to the RasGRPl C1 domain, the CalDAGl C1 domain is not a direct target of DAG and phorbol esters. Furthermore, a functional non-equivalence between the C1 domains of RasGRPl and PKC52 is indicated by the lack of detectable translocation of the Ras GEF C1 domain response to either PDBu or DiC8. 71 diC8 10" > 0 « 0 • 30" GFP GFP-PKC5-C1 GFP-RasGRP-C1 GFP-CalDAG1-C1 GFP-Raf-CI Figure 3.5.The RasGRPI and CalDAGI C1 domains do not translocate to the plasma membrane in response to 1,2 dioctanoyl-sn-glycerol stimulation in NIH 3T3 cells. Either G F P alone (GFP) or G F P fus ions of isolated C1 doma ins were stably overexpressed in NIH 3T3 f ibroblasts, serum-starved for 3.5 hours and st imulated for the indicated t imes with 100uM 1,2 d ioctanoyl-sn-glycerol . The cells were then f ixed and photographed under UV i l lumination. Ar rows indicate regions of concent ra ted f luorescence. 3.3.3 Comparison of RasGRPI and CalDAGI C1 domain-mediated membrane translocation patterns in T28 cells. We overexpressed the isolated C1 domains of RasGRPI and CalDAGI in T28 cells, a murine T cell hybridoma, in order to establish a model in which to determine membrane recruitment responses to both phorbol ester treatment and receptor stimulation. Like other T hybridomas, the T28 cell line has functional signalling from CD3, a component of the TCR, and can also be activated by cotreatment with phorbol ester and calcium ionophores (Kay, unpublished data). 72 The RasGRPl and CalDAGl C1 domains demonstrate distinct plasma membrane localization patterns in response to PMA in T28 cells. In contrast to what was observed in NIH 3T3 cells, the CalDAGl C1 domain exhibited a nuclear-excluded localization pattern with concentrated fluorescence surrounding a portion of the nucleus in unstimulated T28 cells (Figure 3.6 column 4). This concentrated fluorescence pattern, which was exhibited by all of the C1 domains in T28 cells, may be endoplasmic reticulum and Golgi staining because it is similar to the fluorescence pattern seen in NIH 3T3 cells overexpressing either the RasGRPl or the PKC52 C1 domains. We examined membrane recruitment in T28 cells by stimulating C1 domain overexpressing cells with PMA. Similar to the translocation patterns seen in Figure 3.2 in NIH 3T3 cells, both the C1 domains of RasGRPl and PKC82 demonstrated equivalent marked plasma membrane localization after 1 min of PMA stimulation, which indicates that these C1 domains are direct targets of PMA stimulation in T28 cells (Figure 3.6 columns 3 & 2, respectively). However, unlike observations in NIH 3T3 cells, membrane translocation for the RasGRPl and PKC82 C1 domains persisted to 15 min of PMA stimulation indicating that different cell-types exhibit distinct PMA-induced translocation dynamics. Surprisingly, the CalDAGl C1 domain displayed rapid plasma membrane localization in response to PMA stimulation that also persisted to 15 min of stimulation. Contrary to results from NIH 3T3 cells (see Figure 3.2 column 4), this translocation pattern suggests that the CalDAGl C1 domain can directly bind phorbol esters. Yet the CalDAGl C1 domain displayed more modest plasma membrane localization compared to the RasGRPl C1 domain, which suggests that the CalDAGl C1 domain may have a lesser affinity for phorbol esters. 73 PMA • * ; GFP GFP-PKC5-C1 GFP-RasGRP-C1 GFP-CaJDAG1-C1 GFP-Raf-C1 Figure 3.6. Both the RasGRPl and CalDAGl C1 domains translocate to the plasma membrane in response to PMA stimulation in T28 cells. Either G F P alone or G F P fusions of isolated C1 domains were stably overexpressed in T 2 8 cells, s e r u m -starved for 3.5 hours and st imulated for the indicated t imes with 80 nm PMA. T h e cells were then f ixed and photographed under UV i l lumination. Ar rows indicate regions of concent ra ted f luorescence The RasGRPl C1 domain localizes to the plasma membrane following TCR crosslinking, while the CalDAGl C1 domain does not. The expression of both RasGRPl and CalDAGl in T28 cells enabled the study of membrane recruitment in response to physiological ligands (Tognon et al., 1998). We mimicked TCR crosslinking using murine anti-CD3£ under the hypothesis that stimulation of the TCR results in adaptor-mediated activation of PLCyl, which cleaves membrane phosphoinositides generating DAG. Following 5 minutes of TCR stimulation the isolated RasGRPI C1 domain translocated to the plasma membrane in T28 cells, while the CalDAGI C1 domain did not (Figure 3.7 columns 2 & 3). I i n lmgm J GFP GFP-RasGRP-C1 GFP-CalDAG1-C1 Figure 3.7. The RasGRPI C1 domain translocates to the plasma membrane in response to anti-CD3e stimulation in T28 cells. Either G F P alone or G F P fus ions of isolated C1 domains were stably overexpressed in T28 cells, serum-starved for 3.5 hours and st imulated for 5 min . with 1ug/ml ant i -CD3s. T h e cells were then f ixed and photographed under UV i l luminat ion. Ar rows indicate regions of concent ra ted f luorescence. Both RasGRPI and CalDAGI localize to the plasma membrane following stimulation of the TCR. To determine if the full-length forms of these Ras GEFs could translocate in response to TCR stimulation we overexpressed GFP-tagged RasGRPI and CalDAGI in T28 cells. In unstimulated, GFP-tagged, CalDAGI overexpressing cells, concentrated fluorescence was observed at the plasma membrane, which required that the plasma membrane be cleared of visible fluorescence via 24 hours of serum starvation. Following serum starvation, we observed a dramatic plasma membrane localization of full length RasGRPI and CalDAGI in response to TCR-crosslinking via anti-CD3s (Figure 3.8 columns 2 &4, respectively). To determine whether the C1 domain of each Ras GEF was responsible for this translocation, GFP-tagged C1 domain deleted Ras GEF constructs were subsequently overexpressed in T28 cells. It was found that the C1 domain deleted constructs of both RasGRPI and CalDAGI did not relocalize to the plasma membrane upon TCR stimulation, which signified the requirement of each of the Ras GEF C1 domains in TCR-induced membrane recruitment (Figure 3.8 columns 3 & 5, respectively). Membrane translocation results from T28 cells have shown that the RasGRPI and CalDAGI C1 domains do translocate to the plasma membrane in response to PMA, yet these relocalization patterns are distinct. However, both RasGRPI and CalDAGI demonstrate C1 domain dependent plasma membrane relocalization in response to the natural TCR ligand, anti-CD3s. In contrast to the RasGPRI C1 domain, the CalDAGI C1 domain is not sufficient to mediate membrane translocation in response to anti-CD3s, which suggests that something is assisting the TCR-induced translocation of this CalDAGI domain to the plasma membrane. 76 GFP GFP-RasGRP GFP-RasGRPdC1 GFP-CalDAG1 GFP-CalDAG1dC1 Figure 3.8. C1 domain-dependent plasma membrane translocation of RasGRPl and CalDAGl in response to anti-CD3e stimulation in T28 cells. G F P alone, G F P -tagged ful l - length fo rms of C a l D A G l and R a s G R P and the GFP- tagged C1 doma in -deleted f o r m s of R a s G R P were stably overexpressed in T 2 8 cel ls, serum-s tarved for 24 hours and st imulated with mur ine ant i -CD3s ( l ^ g / m l ) for 5 minutes. T h e cells were then f ixed and photographed under UV i l lumination. A r rows indicate regions of concent ra ted f luorescence. 77 3.4 DISCUSSION Our results make a number of advances towards understanding translocation to membranes of C1 domain-containing Ras GEFs. Firstly, membrane translocation of the CalDAGl C1 in response to PMA stimulation in T28 cells shows that, like the RasGRPl C1 domain, this C1 domain also binds phorbol esters. In addition, TCR cross-linking demonstrated C1 domain dependent membrane recruitment of both RasGRPl and CalDAGl. However, the isolated CalDAGl C1 domain was not sufficient to mediate this membrane recruitment. Furthermore, a biochemical difference between the two isolated Ras GEF C1 domains was also found following PC-PLC treatment and PMA stimulation in NIH 3T3 cells. Unexpectedly, we also demonstrated a functional non-equivalence between the RasGRPl and PKC82 C1 domains in response to DAG mimetic ligands. These points are considered in more detail below. The CalDAGl C1 domain is a bona fide phorbol ester receptor, yet there is a functional non-equivalence between this C1 domain and that of RasGRPl. DAG-mediated translocation to membranes of the C1 domain containing RasGRP subfamily of Ras GEFs is one mode of recruitment to the Ras GTPase effector molecules that are located in membranes. Hence, the observation that the CalDAGl C1 domain deviates from the typical consensus C1 domain sequences prompted a comparison of translocation responses between this C1 domain and the well-characterized RasGRPl C1 domain to membranes enriched in DAG or phorbol esters. Our results showing plasma membrane translocation of the CalDAGl C1 domain in response to PMA stimulation in T28 cells implies that this C1 domain is a bona fide phorbol ester receptor. Additionally, other findings have suggested that the CalDAGl C1 domain is a direct target of phorbol esters by showing that CalDAGl can 78 translocate to membranes in response to prolonged exposure to PMA (Clyde-Smith et al., 2000). In particular, our demonstration of diminished PMA-induced membrane fluorescence compared to that shown for the RasGRPl C1 domain in T28 cells suggests that the CalDAGl C1 domain may have a lesser affinity for phorbol esters. In addition, the comparatively more diffuse localization of the CalDAGl C1 domain in unstimulated NIH 3T3 cells suggests that this C1 domain has a lesser affinity for a component of internal membranes, which is presumably DAG. These indications of supposed lesser affinity for phorbol esters and DAG, combined with the lack of relocalization of the isolated CalDAGl C1 domain in NIH 3T3 cells for PMA stimulation and PC-PLC treatment, indicate that there is a functional non-equivalence between the C1 domains of RasGRPl and CalDAGl. On the other hand, rather than having a lesser affinity for phorbol esters and DAG, something may be occupying the CalDAGl C1 domain in NIH 3T3 cells. This C1 domain may be binding something in the cytosol or in intracellular membranes that could impede translocation in response to the ligands used within our study. However, regardless of whether the CalDAGl C1 domain has a lesser affinity for DAG or no affinity for DAG, an obvious possibility is that the biochemical differences between the CalDAGl and RasGRPl C1 domains were facilitated by CalDAGl C1 domain residue deviations from the consensus typical C1 domain. Within the CalDAGl C1 domain there is a serine at the 8 t h position and an alanine at the 13 t h position instead of the conserved aromatic residues found in typical C1 domains. One may predict that these residue differences result in subtle structural differences that either hinder or completely inhibit the formation of the CalDAGl C1 domain/phospholipid/ligand ternary complex at the plasma membrane. The comparison of the RasGRPl and CalDAGl C1 domains demonstrated the heterogeneity among C1 domain translocation responses. Similar to distinct PKC 79 isozymes demonstrating cell-type specific differences in translocation patterns, we have shown that the RasGRPl and CalDAGl C1 domain demonstrate cell-type specific redistribution patterns in response to PMA (Kazanietz, 2000). Unlike results obtained from translocation studies in NIH 3T3 cells, the CalDAGl C1 domain not only relocalized to the plasma membrane in T28 cells, but all the C1 domains were retained at the plasma membrane for 15 minutes in response to PMA stimulation in T28 cells. This result strongly suggests that the plasma membrane of T28 cells is more amenable to phorbol ester-induced membrane recruitment and retention than the plasma membrane of NIH 3T3 cells. An apparent possibility is that this effect is due to a more favourable lipid or protein composition within the plasma membrane of T28 cells. Functional non-equivalence between the RasGRPl and PKCS2 C1 domains Comparison of the PKC52 C1 domain translocation patterns to those of the Ras GEF C1 domains also demonstrates heterogeneity among C1 domains. Because the PKC82 C1 domain relocalized in response to all ligands tested, we can infer that the PKC82 C1 domain has the greatest translocation ability in response to phorbol esters and DAG, followed by the RasGRPl C1 domain and then the CalDAGl C1 domain. Previous studies conducted in our laboratory suggested that the RasGRPl and PKC82 C1 domains are functionally equivalent (Tognon etal., 1998); however results from PDBu and DiC8 stimulation in NIH 3T3 cells now show that there exists a biochemical difference between these C1 domains. Studies of a variety of typical C1 domain containing proteins have shed light on distinct ligand selectivities and fatty acid coactivator requirements as explanations for observing different translocation patterns. One example of ligand selectivity facilitating 80 translocation patterns is revealed by the translocation propensity of PKC8 being affected by the fatty acid side chain length of phorbol esters (Wang et al., 2000). The optimal phorbol ester fatty acid side chain length for PKC8 translocation ranged between phorbol 12,13-dibutyrate and phorbol 12,13-dinonanoate. Accordingly, it is possible that the observed absence of RasGRPI C1 domain response following stimulation by PDBu is due to this phorbol ester's suboptimal fatty acid side chain length for RasGRPI C1 domain translocation. As well, the y and 8 isoforms of PKC have shown dissimilar fatty acid requirements for DiC8-induced translocation (Shirai et al., 1998). Likewise, fatty acids may be required coactivators for DiC8-induced RasGRPI C1 domain translocation. In addition to drawing from studies that reveal C1 domain translocation heterogeneity, the RasGRPI C1 domain contains a few features that distinguish it from other C1 domains. Consequently, we may speculate that these differences cause a unique RasGRPI C1 domain translocation pattern. For example, it was determined by molecular modelling that the RasGRPI C1 domain ligand binding pocket is shallower than that of the PKC82 C1 domain, which predicted a weaker C1 domain-ligand interaction (Rong et al., 2002). Hence, perhaps the outcome of this weaker interaction is the lack of RasGRPI C1 domain translocation that we observed for the ligands PDBu and DiC8. Another possibility is that phospholipid cofactor preferences contribute to the diversity of C1 domain translocation propensity observed between RasGRPI and PKC8. In vitro phorbol ester binding assays have shown that phospholipid cofactors are required for high-affinity binding of phorbol esters to C1 domains, and the anionic phospholipid phosphatidylserine is the most efficient cofactor for reconstitution of binding (Konig etal., 1985). However, the RasGRPI C1 domain was found to be much 81 less dependent on anionic phospholipids than were either the PKC isozymes or RasGRP3 for phorbol ester binding, which is a difference that may be reflected in an altered translocation pattern when compared to anionic phospholipid dependent C1 domains (Lorenzo etal., 2000; Lorenzo etal., 2001). Similarities in CalDAGl and RasGRPl TCR-induced translocation to membranes Contrary to the demonstration of translocation differences between the isolated Ras GEF C1 domains, the TCR stimulation that resulted in plasma membrane translocation demonstrated a functional equivalence between RasGRPl and CalDAGl, which was found to be dependent upon each of their C1 domains. We speculate that membrane recruitment for each Ras GEF was occurring via TCR-initiated signalling by anti-CD3s that activates PLCyl and results in increased DAG at the plasma membrane to which these Ras GEFs were recruited via their C1 domains. This hypothesized signalling pathway for Ras GEF membrane recruitment is supported by our observation that anti-CD3s-induced translocation to the plasma membrane was mimicked by stimulation with the DAG mimetic PMA in T28 cells. Our TCR-induced translocation results complement a previous finding that linked RasGRPl to TCR signalling, however this is the first indication that CalDAGl is linked to the TCR (Ebinu et al., 2000). Moreover, combining our translocation results for anti-CD3e treatment and PMA stimulation suggests that, in addition to being a direct ligand, DAG is a physiological ligand for both the Ras GEF C1 domains in T28 cells. However, the TCR-initiated membrane recruitment that we observed may not be a consequence of DAG production. The RasGRP C1 domains may be binding to other lipids in membranes such as phosphatidic acid or arachidonate (Wakelam, 1998). 82 Our experiments done with both the isolated C1 domain and the C1 domain-deleted form of RasGRPI demonstrated that this domain was both necessary and sufficient for membrane translocation in response to signalling from the TCR. However, we found that the isolated CalDAGI C1 domain is not a sufficient mediator for membrane recruitment of CalDAGI in response to TCR engagement. This result suggests the CalDAGI C1 domain is being aided in translocation to the plasma membrane. Both RasGRPI and CalDAGI contain a pair of EF hands, which is a calcium binding motif. Calcium ionophores have been shown to regulate CalDAGI activation of Rap1 in vitro, which points to a role for the EF hands in CalDAGI regulation (Kawasaki etal., 1998; Clyde-Smith etal., 2000; Dupuy etal., 2001; Guo et al., 2001). Hence, one possibility is that the EF hands of CalDAGI are responding to an increase in free cytoplasmic calcium, which results from PLCyl-mediated production of inositol triphosphate (IP 3), in order to assist in membrane recruitment of this Ras GEF in response to TCR-signalling. Another possibility is that an adaptor protein is aiding CalDAGI membrane translocation through association with an adaptor binding domain within this Ras GEF. Hence, it will be interesting to determine whether the putative CalDAGI C1 domain feature of lesser affinity for membranes can be combined with one of the additional regulatory modules that we have proposed in order to efficiently translocate this protein to membranes in response to receptor stimulation. These possibilities suggest that, compared to the RasGRPI C1 domain, the CalDAGI C1 domain does not make a substantial contribution to this protein's regulation. The differences in the Ras GEF C1 domain translocation patterns raise questions as to their subcellular sites of function. It may be that both RasGRPI and CalDAGI exert their GEF activity at the plasma membrane, yet signalling through 83 CalDAGl may be weaker than through RasGRPl based on its C1 domain's proposed lesser affinity for phorbol esters and DAG. On the other hand, the varied relocalization between the RasGRPl and CalDAGl C1 domains may play a role in determining effector specificity. RasGRPl activates Ras and not the Ras-related GTPase Rap1 A, and CalDAGl does not activate Ras, but activates RaplA (Ebinu etal., 1998, Tognon etal., 1998, Kawasaki etal., 1998, Clyde-Smith etal., 2000). Ras always resides on the cytoplasmic side of the plasma membrane, while Rap1 is often found at endocytic/ lysosomal vesicles and only relocalizes to the plasma membrane following activation of either human Jurkat T cells or neutrophils (Bos, 1997; Pizon et al., 1994; Katagiri et al., 2002; Maridonneau-Parini etal., 1992). Likewise, we showed that the RasGRPl C1 domain translocates to the plasma membrane in response to the majority of stimuli tested, while the CalDAGl C1 domain does not. Hence, the less frequent plasma membrane translocation of CalDAGl may be reflected in the execution of its GEF activity on a GTPase that can often be found within internal membranes. Indeed, CalDAGl only becomes a dual specificity Ras/Rap1 GEF upon alternative splicing that results in the addition of plasma membrane-targeting palmitoylation and myristoylation sites (Clyde-Smith et al., 2000). Replacement of each of the RasGRPl and CalDAGl C1 domains with that of the other and examination of both translocation and effector activation will resolve whether or not localization differences play a role in effector selectivity. Another possibility is that the CalDAGl C1 domain may relocalize to lipid rafts whereas that of RasGRPl does not. Thus, the CalDAGl C1 domain membrane staining might occur in patches corresponding to lipid microdomain location, which is a localization pattern that would explain the less intense plasma membrane fluorescence pattern observed for this C1 domain in T28 cells. Interestingly, PKC0 was shown to 84 translocate to lipid rafts upon TCR stimulation, a subcellular redistribution pattern that requires its regulatory C1 and C2 domains (Bi etal., 2001). The eventual goal of the translocation studies that we have described is to assign a sound biological significance to the translocation patterns of the RasGRP subfamily of Ras GEFs. As was previously described in section and, studies have shown that activated receptors transduce signals through these Ras GEFs to their GTPase targets, although all components of Ras GEF activation have not been defined. For example, the observation that stimulation of the TCR transduced signals through RasGRPI and was at least partly responsible for activating Ras in a Jurkat T cell line directly complements our finding of TCR-induced, C1 domain-dependent RasGRPI plasma membrane relocalization (Ebinu ef al., 2000). Notably, it was described in section that activation of the G protein-coupled receptor Mi muscarinic acetylcholine receptor (M-i mAChR) sequentially activated CalDAGI, Rap1 and B-Raf, which lead to the activation of ERK1/2 (Guo et al., 2001). Because both of these receptors activate PLC, the TCR activates PLCyl and the Mi mAChR activates PLCp, our translocation results in response to TCR cross-linking may provide an intermediary step between PLCyl-catalysed DAG production at the plasma membrane and Ras GEF activation. In order to determine if the TCR within the cell line we used was acting as a PLCyl-activating receptor, our translocation studies with TCR stimulation could be repeated using PLC inhibitors. Alternatively, Abraham and colleagues (2000) have generated a Jurkat T cell line that is deficient in PLCyl wherein TCR-stimulated translocation responses could be measured. 85 In conclusion, our translocation studies found that both RasGRPl and CalDAGl demonstrate a C1 domain-mediated link to TCR signalling and that both of their C1 domains are phorbol ester receptors. However, observing membrane translocation in response to multiple C1 domain ligands established a functional nonequivalence between these two Ras GEF C1 domains. We suggest that the RasGRPl C1 domain has a greater translocation propensity to the plasma membrane than the CalDAGl C1 domain in response to phorbol esters. Additionally, we may speculate that, unlike RasGRPl, CalDAGl has requirements in addition to its C1 domain in order to be recruited to membranes upon TCR engagement. Therefore, we can conclude that the RasGRPl C1 domain has a well-defined regulatory role in DAG-mediated membrane recruitment. However, the CalDAGl C1 domain appears not to be as involved in its resident protein's regulation, at least in response to the DAG mimetic stimuli that we tested. 86 Chapter 4 SUMMARY AND PERSPECTIVES Within this thesis I endeavoured to further characterize two aspects of RasGRPI and CalDAGI regulation. Our studies were based upon the presence of two domains: the C1 domain that is contained within both RasGRPI and CalDAGI, and the amphipathic a helix/leucine zipper motif that is only contained within RasGRPI. However, a fundamental difference exists between the regulatory roles of each of these two domains. As long as C1 domains translocate to membranes in response to DAG and DAG mimetic ligands, their regulatory role within the RasGRP subfamily of Ras GEFs is self-evident; it is presumably to promote the efficient access of the resident RasGRP member to its respective Ras family effector. In contrast, the regulatory role of the structural aspect examined within our first study, homodimerization at the RasGRPI a helix, is not obvious. Within our second study, analysis of the results demonstrating ligand-induced RasGRPI C1 domain membrane relocalization appears to be straightforward. Our results for this C1 domain complement the previous studies that were outlined in Chapter 1 and section 3.1, which together credit this domain as being the chief positive regulator of RasGRPI. Specifically, our results provide additional evidence for this assessment by demonstrating membrane translocation in response to receptor stimulation and in a physiologically relevant T cell line. In contrast, our membrane translocation results with the CalDAGI C1 domain have raised questions regarding the contribution of this domain to regulation and have provided the basis for unforeseen speculation of its function. Previous studies of 87 CalDAGl regulation described in Chapter 1 and section 3.1 have suggested that DAG mimetics increase CalDAGl activity possibly by recruiting the C1 domain to membranes. Our study, which directly compared the two RasGRP C1 domains, has established a difference between these two C1 domains, however this difference is not lucid and requires further study. As was discussed in section 3.4, the analysis of our membrane translocation results with the CalDAGl C1 domain has suggested that this domain has a decreased affinity for DAG mimetics when compared to the RasGRPl C1 domain and that an additional motif may be acting in concert with this domain in mediating TCR-induced translocation. Following the demonstration of a physical association of RasGRPl with itself, our first study permitted only a hypothesis for the regulatory role for homodimerization at the a helix (described in section 2.4). One hypothesis that is worth reiterating following Chapter 3 is that the regulatory role of RasGRPl a helix homodimerization may be to act in concert with the RasGRPl C1 domain by doubling the number of C1/phospholipid/ligand ternary complexes. This role for a helix homodimerization in mediating membrane avidity could be investigated by ligands that did not induce membrane translocation of the isolated RasGRPl C1 domain (refer to section and Substantiation of the RasGRPl a helix in modulating membrane localization is generated by a former study conducted within our laboratory, which demonstrated that a COOH-terminal truncation of this protein deleting both the a helix and spacer region resulted in loss of membrane translocation in response to prolonged PMA exposure (Tognon etal., 1998). In addition to membrane localization by the RasGRPl and CalDAGl C1 domains of RasGRPl and CalDAGl, as well as homodimerization of the RasGRPl amphipathic 88 a helix, there exist other putative modes of regulation within this subfamily. Indeed, domains within the two more recently cloned RasGRP members, RasGRP3 and RasGRP4, could yield much future study. While the RasGRP3 C1 domain is becoming well-characterized as a high-affinity phorbol ester receptor, examination of this protein's pair of EF hands, along with the C1 domain and unique single EF hand of RasGR4, are anticipated. In addition, another study involving the EF hands would need to resolve the role that calcium plays in regulating CalDAGI activity; namely, whether calcium enhances or inhibits the nucleotide exchange promoted by CalDAGI (described in section 1.3.2). Our studies began in the experimental cellular system of NIH 3T3 fibroblasts and transferred to a T cell line, a more biologically relevant system based both on the expression of RasGRPI and CalDAGI in T lymphocytes and the ability to measure responses to upstream signalling events such as receptor stimulation. However, in contrast to the first two RasGRP members, RasGRP3 and RasGRP4 have differential expression patterns. RasGRP3 is not expressed in T lymphocytes, but is expressed in blood, and RasGRP4 is not expressed in the brain, but is expressed in mast cells and myeloid cells (Yamashita etal., 2000; Reuther etal., 2002; Yang etal., 2002). Consequently, these expression patterns require the investigation of RasGRP regulation in additional cellular systems. In addition, diverse biologically relevant cellular systems will provide the opportunity to study RasGRP signalling initiated by other receptors coupled to the DAG- and calcium-generating enzyme PLC, such as G protein-coupled receptors. As well, more extensive studies of regulation ought to be conducted in neuronal cell lines for the three RasGRP members that are expressed within this system because signalling due to calcium influx is most prominent here (Berridge, 1998). 89 REFERENCES 1. Altschuler, D. L. and Ribeiro-Neto, F. (1998). Mitogenic and oncogenic properties of the small G protein Rap1 b. Proc Natl Acad Sci U S A95: 7475-9. 2. Baatout, S. (1998). Phorbol esters: useful tools to study megakaryocyte differentiation. Hematol Cell TherAO: 33-9. 3. Bartkiewicz, M., Houghton, A. and Baron, R. (1999). Leucine zipper-mediated homodimerization of the adaptor protein c-Cbl. A role in c-Cbl's tyrosine phosphorylation and its association with epidermal growth factor receptor. J Biol Chem.274: 30887-95. 4. Belcheva, M. M. and Coscia, C. J. (2002). Diversity of G protein-coupled receptor signaling pathways to ERK/MAP kinase. NeurosignalsAV. 34-44. 5. Berridge, M. J. (1998). Neuronal calcium signaling. Neuron.1V. 13-26. 6. Besterman, J. M., Duronio, V. and Cuatrecasas, P. (1986). Rapid formation of diacylglycerol from phosphatidylcholine: a pathway for generation of a second messenger. Proc Natl Acad Sci U S A83: 6785-9. 7. Bi, K., Tanaka, Y., Coudronniere, N., Sugie, K., Hong, S., van Stipdonk, M. J. and Altman, A. (2001). Antigen-induced translocation of PKC-theta to membrane rafts is required for T cell activation. Nat lmmunol.2: 556-63. 8. Boguski, M. S. and McCormick, F. (1993). Proteins regulating Ras and its relatives. Nature.366: 643-54. 9. Bollag, G. and McCormick, F. (1992). Ras regulation. NF is enough of GAP. Nature.356: 663-4. 10. Boriack-Sjodin, P. A., Margarit, S. M., Bar-Sagi, D. and Kuriyan, J. (1998). The structural basis of the activation of Ras by Sos. Nature.394: 337-43. 11. Bos, J. L. (1997). Ras-like GTPases. Biochim Biophys ActaAZZZ: M19-31. 12. Bos, J. L., de Rooij, J. and Reedquist, K. A. (2001). Rap1 signalling: adhering to new models. Nat Rev Mol Cell Biol.2: 369-77. 13. Bottorf, D., Ebinu, J. and Stone, J. C. (1999). RasGRP, a Ras activator: mouse and human cDNA sequences and chromosomal positions. Mamm Genome. 10: 358-61. 14. Buchsbaum, R., Telliez, J. B., Goonesekera, S. and Feig, L. A. (1996). The N-terminal pleckstrin, coiled-coil, and IQ domains of the exchange factor Ras-GRF act cooperatively to facilitate activation by calcium. Mol Cell S/o/.16: 4888-96. 90 15. Burns, D. J. and Bell, R. M. (1991). Protein kinase C contains two phorbol ester binding domains. J Biol Chem.266: 18330-8. 16. Busch, S. J. and Sassone-Corsi, P. (1990). Dimers, leucine zippers and DNA-binding domains. Trends Genet.6: 36-40. 17. Cai, H., Smola, U., Wixler, V., Eisenmann-Tappe, I., Diaz-Meco, M. T., Moscat, J., Rapp, U. and Cooper, G. M. (1997). Role of diacylglycerol-regulated protein kinase C isotypes in growth factor activation of the Raf-1 protein kinase. Mol Cell BiolM: 732-41. 18. Caloca, M. J., Wang, H., Delemos, A., Wang, S. and Kazanietz, M. G. (2001). Phorbol esters and related analogs regulate the subcellular localization of beta 2-chimaerin, a non-protein kinase C phorbol ester receptor. J Biol Chem.276: 18303-12. 19. Cambier, J. C. and Jensen, W. A. (1994). The hetero-oligomeric antigen receptor complex and its coupling to cytoplasmic effectors. Curr Opin Genet DevA\ 55-63. 20. Campbell, S. L, Khosravi-Far, R., Rossman, K. L, Clark, G. J. and Der, C. J. (1998). Increasing complexity of Ras signaling. Oncogene.M'. 1395-413. 21 . Campbell, K. S. (1999). Signal transduction from the B cell antigen-receptor. Curr Opin Immunol'.11: 256-64. 22. Castagna, M., Takai, Y., Kaibuchi, K., Sano, K., Kikkawa, U. and Nishizuka, Y. (1982). Direct activation of calcium-activated, phospholipid-dependent protein kinase by tumor-promoting phorbol esters. J Biol Chem.257: 7847-51. 23. Caulfield, M. P. (1993). Muscarinic receptors-characterization, coupling and function. Pharmacol Ther.58: 319-79. 24. Clyde-Smith, J., Silins, G., Gartside, M., Grimmond, S., Etheridge, M., Apolloni, A., Hayward, N. and Hancock, J. F. (2000). Characterization of RasGRP2, a plasma membrane-targeted, dual specificity Ras/Rap exchange factor. J Biol Chem.275: 32260-7. 25. Cook, S. J., Rubinfeld, B., Albert, I. and McCormick, F. (1993). RapV12 antagonizes Ras-dependent activation of ERK1 and ERK2 by LPA and EGF in Rat-1 fibroblasts. Embo J.12: 3475-85. 26. Cox, A. D., Brtva, T. R., Lowe, D. G. and Der, C. J. (1994). R-Ras induces malignant, but not morphologic, transformation of NIH3T3 cells. Oncogene.^: 3281-8. 27. Crabtree, G. R. (1989). Contingent genetic regulatory events in T lymphocyte activation. Science.243: 355-61. 91 28. Crespo, P., Schuebel, K. E., Ostrom, A. A., Gutkind, J. S. and Bustelo, X. R. (1997). Phosphotyrosine-dependent activation of Rac-1 GDP/GTP exchange by the vav proto-oncogene product. Nature.385: 169-72. 29. Crompton, T., Gilmour, K. C. and Owen, M. J. (1996). The MAP kinase pathway controls differentiation from double-negative to double-positive thymocyte. Cell.SS: 243-51. 30. de Hoog, C. L , Fan, W. T., Goldstein, M. D., Moran, M. F. and Koch, C. A. (2000). Calmodulin-independent coordination of Ras and extracellular signal-regulated kinase activation by Ras-GRF2. Mol Cell Biol.20: 2727-33. 31 . Dower, N. A., Stang, S. L, Bottorff, D. A., Ebinu, J. O., Dickie, P., Ostergaard, H. L. and Stone, J. C. (2000). RasGRP is essential for mouse thymocyte differentiation and TCR signaling. Nat Immunol A: 317-21. 32. Downward, J., Graves, J. D., Warne, P. H., Rayter, S. and Cantrell, D. A. (1990). Stimulation of p21ras upon T-cell activation. Nature.346: 719-23. 33. Downward, J., Graves, J. and Cantrell, D. (1992). The regulation and function of p21ras in T cells. Immunol TodayAZ: 89-92. 34. Downward, J. (1996). Control of ras activation. Cancer Surv.27: 87-100. 35. Dupuy, A. J., Morgan, K., von Lintig, F. C , Shen, H., Acar, H., Hasz, D. E., Jenkins, N. A., Copeland, N. G., Boss, G. R. and Largaespada, D. A. (2001). Activation of the Rap1 guanine nucleotide exchange gene, CalDAG-GEF I, in BXH-2 murine myeloid leukemia. J Biol Chem.276: 11804-11. 36. Ebihara, T. and Saffen, D. (1997). Muscarinic acetylcholine receptor-mediated induction of zif268 mRNA in PC12D cells requires protein kinase C and the influx "of extracellular calcium. J Neurochem.68: 1001-10. 37. Ebinu, J. O., Bottorff, D. A., Chan, E. Y., Stang, S. L, Dunn, R. J. and Stone, J. C. (1998). RasGRP, a Ras guanyl nucleotide- releasing protein with calcium- and diacylglycerol-binding motifs. Sc/eA?ce.280: 1082-6. 38. Ebinu, J. O., Stang, S. L, Teixeira, C , Bottorff, D. A., Hooton, J., Blumberg, P. M., Barry, M., Bleakley, R. C , Ostergaard, H. L. and Stone, J. C. (2000). RasGRP links T-cell receptor signaling to Ras. Blood.95: 3199-203. 39. Evenas, J., Malmendal, A. and Forsen, S. (1998). Calcium. Curr Opin Chem Biol.2: 293-302. 40. Fam, N. P., Fan, W. T., Wang, Z., Zhang, L. J., Chen, H. and Moran, M. F. (1997). Cloning and characterization of Ras-GRF2, a novel guanine nucleotide exchange factor for Ras. Mol Cell BiolA7: 1396-406. 92 41 . Felder, C. C. (1995). Muscarinic acetylcholine receptors: signal transduction through multiple effectors. Faseb J.9: 619-25. 42. Finco, T. S., Kadlecek, T., Zhang, W., Samelson, L. E. and Weiss, A. (1998). LAT is required forTCR-mediated activation of PLCgammal and the Ras pathway. Immunity.9: 617-26. 43. Franke, B., Akkerman, J. W. and Bos, J. L. (1997). Rapid Ca2+-mediated activation of Rap1 in human platelets. Embo J.16: 252-9. 44. Genot, E. and Cantrell, D. A. (2000). Ras regulation and function in lymphocytes. Curr Opin lmmunol.12: 289-94. 45. Ghosh, S., Strum, J. C , Sciorra, V. A., Daniel, L. and Bell, R. M. (1996). Raf-1 kinase possesses distinct binding domains for phosphatidylserine and phosphatidic acid. Phosphatidic acid regulates the translocation of Raf-1 in 12-0-tetradecanoylphorbol-13-acetate-stimulated Madin-Darby canine kidney cells. J Biol Chem.27V. 8472-80. 46. Gibbs, J. B. (1995). Determination of guanine nucleotides bound to Ras in mammalian cells. Methods Enzymol.255: 118-25. 47. Gold, M. R. and Matsuuchi, L. (1995). Signal transduction by the antigen receptors of B and T lymphocytes. Int Rev CytolA57: 181-276. 48. Goodnight, J. A., Mischak, H., Kolch, W. and Mushinski, J. F. (1995). Immunocytochemical localization of eight protein kinase C isozymes overexpressed in NIH 3T3 fibroblasts. Isoform-specific association with microfilaments, Golgi, endoplasmic reticulum, and nuclear and cell membranes. J Biol Chem.270: 9991-10001. 49. Graham, S. M., Cox, A. D., Drivas, G., Rush, M. G., D'Eustachio, P. and Der, C. J. (1994). Aberrant function of the Ras-related protein TC21/R-Ras2 triggers malignant transformation. Mol Cell Biol'.14: 4108-15. 50. Grammer, T. C. and Blenis, J. (1997). Evidence for MEK-independent pathways regulating the prolonged activation of the ERK-MAP kinases. Oncogene. 14: 1635-42. 51 . Guo, F. F., Kumahara, E. and Saffen, D. (2001). A CalDAG-GEFI/Rap1/B-Raf cassette couples M(1) muscarinic acetylcholine receptors to the activation of ERK1/2. J Biol Chem.276: 25568-81. 52. Hannun, Y. A., Loomis, C. R. and Bell, R. M. (1985). Activation of protein kinase C by Triton X-100 mixed micelles containing diacylglycerol and phosphatidylserine. J Biol Chem.260: 10039-43. 93 53. Hartshorne, D. J., Ito, M. and Erdodi, F. (1998). Myosin light chain phosphatase: subunit composition, interactions and regulation. J Muscle Res Cell MotilA9: 325-41. 54. Holsinger, L. J., Spencer, D. M., Austin, D. J., Schreiber, S. L. and Crabtree, G. R. (1995). Signal transduction in T lymphocytes using a conditional allele of Sos. Proc Natl Acad Sci U S A.92: 9810-4. 55. Hurley, J. H., Newton, A. C , Parker, P. J., Blumberg, P. M. and Nishizuka, Y. (1997). Taxonomy and function of C1 protein kinase C homology domains. Protein Sci.6: 477-80. 56. Hurley, J. H. and Misra, S. (2000). Signaling and subcellular targeting by membrane-binding domains. Annu Rev Biophys Biomol Struct.29: 49-79. 57. Ikura, M. (1996). Calcium binding and conformational response in EF-hand proteins. Trends Biochem Sc/'.21: 14-7. 58. Iritani, B. M., Alberola-lla, J., Forbush, K. A. and Perimutter, R. M. (1999). Distinct signals mediate maturation and allelic exclusion in lymphocyte progenitors. ImmunityAO: 713-22. 59. Irvin, B. J., Williams, B. L, Nilson, A. E., Maynor, H. O. and Abraham, R. T. (2000). Pleiotropic contributions of phospholipase C-gamma1 (PLC-gamma1) to T-cell antigen receptor-mediated signaling: reconstitution studies of a PLC-gammal-deficient Jurkat T-cell line. Mol Cell Biol.20: 9149-61. 60. Izquierdo, M., Downward, J., Graves, J. D. and Cantrell, D. A. (1992). Role of protein kinase C in T-cell antigen receptor regulation of p21ras: evidence that two p21ras regulatory pathways coexist in T cells. Mol Cell BiolA2: 3305-12. 61 . Ji, T. H., Grossmann, M. and Ji, I. (1998). G protein-coupled receptors. I. Diversity of receptor-ligand interactions. J Biol Chem.273: 17299-302. 62. Johnson, J. E. and Cornell, R. B. (1999). Amphitropic proteins: regulation by reversible membrane interactions (review). Mol Membr Biol'.16: 217-35. 63. Jones, D. R., Sanjuan, M. A., Stone, J. C. and Merida, I. (2002). Expression of a catalytically inactive form of diacylglycerol kinase alpha induces sustained signaling through RasGRP. FasebJ."\6: 595-7. 64. Katagiri, K., Hattori, M., Minato, N., Irie, S., Takatsu, K. and Kinashi, T. (2000). Rap1 is a potent activation signal for leukocyte function-associated antigen 1 distinct from protein kinase C and phosphatidylinositol-3-OH kinase. Mol Cell B/O/.20: 1956-69. 65. Katagiri, K., Hattori, M., Minato, N. and Kinashi, T. (2002). Rap1 functions as a key regulator of T-cell and antigen-presenting cell interactions and modulates T-cell responses. Mol Cell Biol.22: 1001-15. 94 66. Kawai, T., Matsumoto, M., Takeda, K., Sanjo, H. and Akira, S. (1998). ZIP kinase, a novel serine/threonine kinase which mediates apoptosis. Mol Cell Biol. 18: 1642-51. 67. Kawasaki, H., Springett, G. M., Toki, S., Canales, J. J., Harlan, P., Blumenstiel, J. P., Chen, E. J., Bany, I. A., Mochizuki, N., Ashbacher, A., Matsuda, M., Housman, D. E. and Graybiel, A. M. (1998). A Rap guanine nucleotide exchange factor enriched highly in the basal ganglia. Proc Natl Acad Sci U S A.95: 13278-83. 68. Kay, R.J. unpublished data. 69. Kazanietz, M. G., Wang, S., Milne, G. W., Lewin, N. E., Liu, H. L. and Blumberg, P. M. (1995). Residues in the second cysteine-rich region of protein kinase C delta relevant to phorbol ester binding as revealed by site-directed mutagenesis. J Biol Chem.270: 21852-9. 70. Kazanietz, M. G. (2000). Eyes wide shut: protein kinase C isozymes are not the only receptors for the phorbol ester tumor promoters. Mol Carcinog.28: 5-11. 71. Kelley, G. G., Reks, S. E., Ondrako, J. M. and Smrcka, A. V. (2001). Phospholipase C(epsilon): a novel Ras effector. Embo J.20: 743-54. 72. Kitayama, H., Sugimoto, Y., Matsuzaki, T., Ikawa, Y. and Noda, M. (1989). A ras-related gene with transformation suppressor activity. Ce//.56: 77-84. 73. Konig, B., Di Nitto, P. A. and Blumberg, P. M. (1985). Phospholipid and Ca++ dependency of phorbol ester receptors. J Cell Biochem.27: 255-65. 74. Kretsinger, R. H. and Nockolds, C. E. (1973). Carp muscle calcium-binding protein. II. Structure determination and general description. J Biol Chem.248: 3313-26. 75. Kumahara, E., Ebihara, T. and Saffen, D. (1999). Nerve growth factor induces zif268 gene expression via MAPK-dependent and -independent pathways in PC12D cells. J Biochem (Tokyo)A25: 541.-53. 76. Kyriakis, J. M. and Ayruch, J. (2001). Mammalian mitogen-activated protein kinase signal transduction pathways activated by stress and inflammation. Physiol RevM: 807-69. 77. Lai, C. C , Boguski, M., Broek, D. and Powers, S. (1993). Influence of guanine nucleotides on complex formation between Ras and CDC25 proteins. Mol Cell Biol A3: 1345-52. 78. Lapetina, E. G., Reep, B., Ganong, B. R. and Bell, R. M. (1985). Exogenous sn-1,2-diacylglycerols containing saturated fatty acids function as bioregulators of protein kinase C in human platelets. J Biol Chem.260: 1358-61. 95 79. Larrodera, P., Cornet, M. E., Diaz-Meco, M. T., Lopez-Barahona, M., Diaz-Laviada, I., Guddal, P. H., Johansen, T. and Moscat, J. (1990). Phospholipase C-mediated hydrolysis of phosphatidylcholine is an important step in PDGF-stimulated DNA synthesis. Ce//.61: 1113-20. 80. Lewit-Bentley, A. and Rety, S. (2000). EF-hand calcium-binding proteins. Curr Opin Struct Biol.10: 637-43. 81 . Lin, A., Minden, A., Martinetto, H., Claret, F. X., Lange-Carter, C , Mercurio, F., Johnson, G. L. and Karin, M. (1995). Identification of a dual specificity kinase that activates the Jun kinases and p38-Mpk2. Science.268: 286-90. 82. Lin, J. and Weiss, A. (2001). Identification of the minimal tyrosine residues required for linker for activation of T cell function. J Biol Chem.276: 29588-95. 83. Lopez, I., Mak, E. C , Ding, J., Hamm, H. E. and Lomasney, J. W. (2001). A novel bifunctional phospholipase c that is regulated by Galpha 12 and stimulates the Ras/mitogen-activated protein kinase pathway. J Biol Chem.276: 2758-65. 84. Lorenzo, P. S., Beheshti, M., Pettit, G. R., Stone, J. C. and Blumberg, P. M. (2000). The guanine nucleotide exchange factor RasGRP is a high -affinity target for diacylglycerol and phorbol esters. Mol Pharmacol.57: 840-6. 85. Lorenzo, P. S., Kung, J. W., Bottorff, D. A., Garfield, S. H., Stone, J. C. and Blumberg, P. M. (2001). Phorbol esters modulate the Ras exchange factor RasGRP3. Cancer f?es.61: 943-9. 86. Macara, I. G., Lounsbury, K. M., Richards, S. A., McKiernan, C. and Bar-Sagi, D. (1996). The Ras superfamily of GTPases. Faseb J.10: 625-30. 87. Maridonneau-Parini, I. and de Gunzburg, J. (1992). Association of rapl and rap2 proteins with the specific granules of human neutrophils. Translocation to the plasma membrane during cell activation. J Biol Chem.267: 6396-402. 88. McLeod, S. J., Ingham, R. J., Bos, J. L , Kurosaki, T. and Gold, M. R. (1998). Activation of the Rap1 GTPase by the B cell antigen receptor. J Biol Chem.272: 29218-23. 89. McLeod, S. J. and Gold, M. R. (2001). Activation and function of the Rap1 GTPase in B lymphocytes. Int Rev lmmunol.20: 763-89. 90. Michie, A. M., Trop, S., Wiest, D. L. and Zuniga-Pflucker, J. C. (1999). Extracellular signal-regulated kinase (ERK) activation by the pre-T cell receptor in developing thymocytes in vivo. J Exp MedA90: 1647-56. 91 . Minden, A., Lin, A., Claret, F. X., Abo, A. and Karin, M. (1995). Selective activation of the JNK signaling cascade and c-Jun transcriptional activity by the small GTPases Rac and Cdc42Hs. Ce//.81: 1147-57. 96 92. Mochizuki, N., Ohba, Y., Kobayashi, S., Otsuka, N., Graybiel, A. M., Tanaka, S. and Matsuda, M. (2000). Crk activation of JNK via C3G and R-Ras. J Biol Chem.275: 12667-71. 93. O'Shea, E. K., Klemm, J. D., Kim, P. S. and Alber, T. (1991). X-ray structure of the GCN4 leucine zipper, a two-stranded, parallel coiled coil. Sc/ence.254: 539-44. 94. Oancea, E., Teruel, M. N., Quest, A. F. and Meyer, T. (1998). Green fluorescent protein (GFP)-tagged cysteine-rich domains from protein kinase C as fluorescent indicators for diacylglycerol signaling in living cells. J Cell B/'o/.140: 485-98. 95. Ohba, Y., Mochizuki, N., Yamashita, S., Chan, A. M., Schrader, J. W., Hattori, S., Nagashima, K. and Matsuda, M. (2000). Regulatory proteins of R-Ras, TC21/R-Ras2, and M-Ras/R-Ras3. J Biol Chem.275: 20020-6. 96. Okada, T., Hu, C. D., Jin, T. G., Kariya, K., Yamawaki-Kataoka, Y. and Kataoka, T. (1999). The strength of interaction at the Raf cysteine-rich domain is a critical determinant of response of Raf to Ras family small GTPases. Mol Cell BiolAQ: 6057-64. 97. Ono, Y., Fujii, T., Igarashi, K., Kuno, T., Tanaka, C , Kikkawa, U. and Nishizuka, Y. (1989). Phorbol ester binding to protein kinase C requires a cysteine-rich zinc-finger-like sequence. Proc Natl Acad Sci U S A86: 4868-71. 98. Overbeck, A. F., Brtva, T. R., Cox, A. D., Graham, S. M., Huff, S. Y., Khosravi-Far, R., Quilliam, L. A., Solski, P. A. and Der, C J . (1995). Guanine nucleotide exchange factors: activators of Ras superfamily proteins. Mol Reprod Dev.42: 468-76. 99. Pagano, R. E., Longmuir, K. J. and Martin, O. C. (1983). Intracellular translocation and metabolism of a fluorescent phosphatidic acid analogue in cultured fibroblasts. J Biol Chem.258: 2034-40. 100. Palmer, A., Gavin, A. C. and Nebreda, A. R. (1998). A link between MAP kinase and p34(cdc2)/cyclin B during oocyte maturation: p90(rsk) phosphorylates and inactivates the p34(cdc2) inhibitory kinase Myt1. Embo J. 17: 5037-47. 101. Pear, W. S., Nolan, G. P., Scott, M. L. and Baltimore, D. (1993). Production of high-titer helper-free retroviruses by transient transfection. Proc Natl Acad Sci U SA90: 8392-6. 102. Pizon, V., Desjardins, M., Bucci, C , Partem, R. G. and Zerial, M. (1994). Association of Rapla and Raplb proteins with late endocytic/phagocytic compartments and Rap2a with the Golgi complex. J Cell Sci.107: 1661-70. 103. Pyszniak, A. M., Welder, C. A. and Takei, F. (1994). Cell surface distribution of high-avidity LFA-1 detected by soluble ICAM-1-coated microspheres. J Immunol.152: 5241-9. 97 104. Quest, A. F. and Bell, R. M. (1994). The regulatory region of protein kinase C gamma. Studies of phorbol ester binding to individual and combined functional segments expressed as glutathione S-transferase fusion proteins indicate a complex mechanism of regulation by phospholipids, phorbol esters, and divalent cations. J Biol Chem.269: 20000-12. 105. Quest, A. F., Bardes, E. S. and Bell, R. M. (1994). A phorbol ester binding domain of protein kinase C gamma. High affinity binding to a glutathione-S-transferase/Cys2 fusion protein. J Biol Chem.269: 2953-60. 106. Quilliam, L. A., Khosravi-Far, R., Huff, S. Y. and Der, C. J. (1995). Guanine nucleotide exchange factors: activators of the Ras superfamily of proteins. Bioessays. 17: 395-404. 107. Quilliam, L. A., Castro, A. F., Rogers-Graham, K. S., Martin, C. B., Der, C. J. and Bi, C. (1999). M-Ras/R-Ras3, a transforming ras protein regulated by Sos1, GRF1, and p120 Ras GTPase-activating protein, interacts with the putative Ras effector AF6. J Biol Chem.274: 23850-7. 108. Rebhun, J. F., Chen, H. and Quilliam, L. A. (2000). Identification and characterization of a new family of guanine nucleotide exchange factors for the ras-related GTPase Ral. J Biol Chem.275: 13406-10. 109. Reedquist, K. A., Ross, E., Koop, E. A., Wolthuis, R. M., Zwartkruis, F. J., van Kooyk, Y., Salmon, M., Buckley, C. D. and Bos, J. L. (2000). The small GTPase, Rap1, mediates CD31-induced integrin adhesion. J Cell Biol'.148: 1151-8. 110. Reuther, G. W. and Der, C. J. (2000). The Ras branch of small GTPases: Ras family members don't fall far from the tree. Curr Opin Cell BiolA2: 157-65. 111. Reuther, G. W., Lambert, Q. T., Rebhun, J. F., Caligiuri, M. A., Quilliam, L. A. and Der, C. J. (2002). RasGRP4 is a novel ras activator isolated from acute myeloid leukemia. J Biol Chem.5: 5. 112. Rizzo, M. A., Shome, K., Watkins, S. C. and Romero, G. (2000). The recruitment of Raf-1 to membranes is mediated by direct interaction with phosphatidic acid and is independent of association with Ras. J Biol Chem.275: 23911-8. 113. Robinson, L. C , Gibbs, J. B., Marshall, M. S., Sigal, I. S. and Tatchell, K. (1987). CDC25: a component of the RAS-adenylate cyclase pathway in Saccharomyces cerevisiae. Sc/ence.235: 1218-21. 114. Ron, D. and Kazanietz, M. G. (1999). New insights into the regulation of protein kinase C and novel phorbol ester receptors. Faseb J.13: 1658-76. 115. Rong, S. B., Enyedy, I. J., Qiao, L., Zhao, L., Ma, D., Pearce, L. L., Lorenzo, P. S., Stone, J. C , Blumberg, P. M., Wang, S. and Kozikowski, A. P. (2002). Structural basis of RasGRP binding to high-affinity PKC ligands. J Med Chem.45: 853-60. 98 116. Ruth, P., Pfeifer, A., Kamm, S., Klatt, P., Dostmann, W. R. and Hofmann, F. (1997). Identification of the amino acid sequences responsible for high affinity activation of cGMP kinase lalpha. J Biol Chem.272: 10522-8. 117. Satoh, T. and Kaziro, Y. (1995). Measurement of Ras-bound guanine nucleotide in stimulated hematopoietic cells. Methods Enzymol.255: 149-55. 118. Schlaepfer, D. D., Hanks, S. K., Hunter, T. and van der Geer, P. (1994). Integrin-mediated signal transduction linked to Ras pathway by GRB2 binding to focal adhesion kinase. Nature.372: 786-91. 119. Sebzda, E., Mariathasan, S., Ohteki, T., Jones, R., Bachmann, M. F. and Ohashi, P. S. (1999). Selection of the T cell repertoire. Annu Rev Immunol. 17: 829-74. 120. Sebzda, E., Bracke, M., Tugal, T., Hogg, N. and Cantrell, D. A. (2002). RaplA positively regulates T cells via integrin activation rather than inhibiting lymphocyte signaling. Nat Immunol.3: 251-8. 121. Shao, L , Lewin, N. E., Lorenzo, P. S., Hu, Z., Enyedy, I. J., Garfield, S. H., Stone, J. C , Marner, F. J., Blumberg, P. M. and Wang, S. (2001). Iridals are a novel class of ligands for phorbol ester receptors with modest selectivity for the RasGRP receptor subfamily. J Med ChemAA: 3872-80. 122. Shields, J. M., Pruitt, K., McFall, A., Shaub, A. and Der, C. J. (2000). Understanding Ras: 'it ain't over 'til it's over'. Trends Cell BiolAO: 147-54. 123. Shirai, Y., Kashiwagi, K., Yagi, K., Sakai, N. and Saito, N. (1998). Distinct effects of fatty acids on translocation of gamma- and epsilon-subspecies of protein kinase C. J Cell Biol.U3: 511-21. 124. Song, C , Hu, C. D., Masago, M., Kariyai, K., Yamawaki-Kataoka, Y., Shibatohge, M., Wu, D., Satoh, T. and Kataoka, T. (2001). Regulation of a novel human phospholipase C, PLCepsilon, through membrane targeting by Ras. J Biol Chem.276\ 2752-7. 125. Sturgill, T. W., Ray, L. B., Erikson, E. and Mailer, J. L. (1988). Insulin-stimulated MAP-2 kinase phosphorylates and activates ribosomal protein S6 kinase II. Nature.334: 715-8. 126. Taylor, S. J. and Shalloway, D. (1996). Cell cycle-dependent activation of Ras. CurrBiol.6: 1621-7. 127. Tognon, C. E., Kirk, H. E., Passmore, L. A., Whitehead, I. P., Der, C. J. and Kay, R. J. (1998). Regulation of RasGRP via a phorbol ester-responsive C1 domain. Mol Cell 8/0/.18: 6995-7008. 128. Topham, M. K. and Prescott, S. M. (2001). Diacylglycerol kinase zeta regulates Ras activation by a novel mechanism. J Cell Biol.1 52: 1135-43. 99 129. Treisman, R. (1996). Regulation of transcription by MAP kinase cascades. Curr Opin CellBiol.S: 205-15. 130. Vetter, I. R. and Wittinghofer, A. (2001). The guanine nucleotide-binding switch in three dimensions. Science.294: 1299-304. 131. von Boehmer, H., Aifantis, I., Azogui, O., Feinberg, J., Saint-Ruf, C , Zober, C , Garcia, C. and Buer, J. (1998). Crucial function of the pre-T-cell receptor (TCR) in TCR beta selection, TCR beta allelic exclusion and alpha beta versus gamma delta lineage commitment. Immunol Rev'.165: 111-9. 132. Vos, M. D., Ellis, C. A., Bell, A., Birrer, M. J. and Clark, G. J. (2000). Ras uses the novel tumor suppressor RASSF1 as an effector to mediate apoptosis. J Biol Chem.275\ 35669-72. 133. Wakelam, M. J. (1998). Diacylglycerol-when is it an intracellular messenger? Biochim Biophys ActaAAZQ: 117-26. 134. Waltersson, Y., Linse, S., Brodin, P. and Grundstrom, T. (1993). Mutational effects on the cooperativity of Ca2+ binding in calmodulin. Biochemistry.32: 7866-71. 135. Wang, Q. J., Fang, T. W., Fenick, D., Garfield, S., Bienfait, B., Marquez, V. E. and Blumberg, P. M. (2000). The lipophilicity of phorbol esters as a critical factor in determining the pattern of translocation of protein kinase C delta fused to green fluorescent protein. J Biol Chem.275: 12136-46. 136. Wang, Q. J., Fang, T. W., Nacro, K., Marquez, V. E., Wang, S. and Blumberg, P. M. (2001). Role of hydrophobic residues in the C1 b domain of protein kinase C delta on ligand and phospholipid interactions. J Biol Chem.276: 19580-7. 137. Waskiewicz, A. J., Flynn, A., Proud, C. G. and Cooper, J. A. (1997). Mitogen-activated protein kinases activate the serine/threonine kinases Mnk1 and Mnk2. EmboJA6\ 1909-20. 138. Wess, J. (1993). Mutational analysis of muscarinic acetylcholine receptors: structural basis of ligand/receptor/G protein interactions. Life Sci.53: 1447-63. 139. Whitehead, I., Kirk, H. and Kay, R. (1995). Expression cloning of oncogenes by retroviral transfer of cDNA libraries. Mol Cell S/'o/.15: 704-10. 140. Yablonski, D., Kuhne, M. R., Kadlecek, T. and Weiss, A. (1998). Uncoupling of nonreceptor tyrosine kinases from PLC-gamma1 in an SLP-76-deficient T cell. Sc/ence.281: 413-6. 141. Yablonski, D. and Weiss, A. (2001). Mechanisms of signaling by the hematopoietic-specific adaptor proteins, SLP-76 and LAT and their B cell counterpart, BLNK/SLP-65. Adv lmmunol.79: 93-128. 100 142. Yamashita, S., Mochizuki, N., Ohba, Y., Tobiume, M., Okada, Y., Sawa, H., Nagashima, K. and Matsuda, M. (2000). CalDAG-GEFIII activation of Ras, R-ras, and Rap1. J Biol Chem.275: 25488-93. 143. Yang, Y., Li, L, Wong, G. W., Krilis, S. A., Madhusudhan, M. S., Sali, A. and Stevens, R. L. (2002). RasGRP4, a new mast cell-restricted Ras guanine nucleotide-releasing protein with calcium- and diacylglycerol-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. J Biol Chem.277: 25756-74. 144. Zeng, X., Herndon, A. M. and Hu, J. C. (1997). Buried asparagines determine the dimerization specificities of leucine zipper mutants. Proc Natl Acad Sci U S A94: 3673-8. 145. Zhang, G., Kazanietz, M. G., Blumberg, P. M. and Hurley, J. H. (1995). Crystal structure of the cys2 activator-binding domain of protein kinase C delta in complex with phorbol ester. Ce//.81: 917-24. 146. Zhang, Z., Vuori, K., Wang, H., Reed, J. C. and Ruoslahti, E. (1996). Integrin activation by R-ras. Ce//.85: 61-9. 147. Zhang, W., Sloan-Lancaster, J., Kitchen, J., Trible, R. P. and Samelson, L. E. (1998a). LAT: the ZAP-70 tyrosine kinase substrate that links T cell receptor to cellular activation. Ce//.92: 83-92. 148. Zhang, W., Trible, R. P. and Samelson, L. E. (1998b). LAT palmitoylation: its essential role in membrane microdomain targeting and tyrosine phosphorylation during T cell activation. Immunity.^: 239-46. 149. Zhang, W., Sommers, C. L., Burshtyn, D. N., Stebbins, C. C , DeJarnette, J. B., Trible, R. P., Grinberg, A., Tsay, H. C , Jacobs, H. M., Kessler, C. M., Long, E. O., Love, P. E. and Samelson, L. E. (1999). Essential role of LAT in T cell development. Immunity'.10: 323-32. 150. Zhang, W., Trible, R. P., Zhu, M., Liu, S. K., McGlade, C. J. and Samelson, L. E. (2000). Association of Grb2, Gads, and phospholipase C-gamma 1 with phosphorylated LAT tyrosine residues. Effect of LAT tyrosine mutations on T cell angigen receptor-mediated signaling. J Biol Chem.275: 23355-61. 151. Zwartkruis, F. J. and Bos, J. L. (1999). Ras and Rap1: two highly related small GTPases with distinct function. Exp Cell Res.253: 157-65. 101 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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


Related Items