The role of the Grb14 adaptor protein in receptor tyrosine kinase signalling

Rania Kairouz

A thesis submitted for the degree of Doctor of Philosophy

Faculty of Medicine University of New South Wales 2002 a:Jedication

irhis undeservin8 thesis is dedicated to the Creator of a(( discoveries, for mercifu((y hefpin8 us to make them

11 !Jl.cinowfedgments

'The simpler one writes the better it will be' St. Bernadette Soubirous

Whether this thesis lives up to St. Bernadette's wisdom remains to be seen, but I can state with certainty that this work would not have been possible without dedicated individuals with whom I had the pleasure to work and learn. First and foremost, my dear and kind supervisors. Roger Daly, in particular, for his invaluable guidance at every step, during the highs, lows and the flatlines... His unwavering positivism and insight regarding experiments are quite appreciated, in addition to his eagerness in viewing Western blots (or any blot!) and his magnanimous endurance in hearing and editing some of my 'favourite' words, including 'basically' and the most recent 'plethora', so BASICALLY thanks Rog. I also sincerely thank Rob Sutherland for providing continuous support and morale-boosting advice on scientific and career issues, and for giving me the opportunity and the privilege to work in such a stimulating, friendly and intricately well-managed environment. Many thanks also to Liz Musgrove, for all her help and guidance when I joined forces with the Cell Cyders. For advice and help with this thesis, I also thank Big Bad Boris, Dr Dan, and Gary Leong. I appreciate the assistance from Sue, Carsten and Tuan for statistical analysis. My heartfelt thanks to all those who contributed to this work (see Chapter 2), including Georgina Sanderson and Roger for the 184 library screening. In particular, I BASICALLY thank Ruth for technical advice and help, and the lovely Jayamala with whom I worked during the latter part of my PhD.

Thankyou to all my precious Cancerites and Garvanites for your support throughout my PhD, you are all like a second family to me (my sympathy to those unfortunate individuals whom I pestered with questions). I am grateful to Fireman Charlie who 'always saves the day', for sending me references (and jokes) during my 'hermitress' writing phase. It is no secret that working in lab 1 is never monotonous, and here I have the opportunity to reveal all. However, I shall leave that for my upcoming bestseller and contend with thanking Colonel Michelle (whose favourite pasttime was to answer my questions), Suzann for being

lll a 'pseudo' big sister, Mel for being a 'pseudo' little sister, Amanda! Amanda! Amanda! for making sure I was 'stirred not shaken', Psycho Sam for bravely enduring repetitions of 'catch a falling star' and Prof. Clancy, my partner-in-crime (that's in my bestseller), for allowing me to metastasise to her bench. Frequently. Thanks also to Steph and Perla Preciosa Marcia for all the fun outings. I am grateful to Damien for the med/bird stories, Superflower Laura for baby Angelica (who fought her way into this world during my writing) and D.J. for the jokes and discussions (Damien and D. J., I think we agree to disagree!).

My cherished family and friends .... Well, thankyou for so much incredible support. You guys have astounded me! I am greatly indebted to Mirna and Carmen for help with proofreading and for effortlessly transforming the rania Frown into a huge smile©. Thanks for the encouragement of those who have been patiently waiting for me to finish, in order to 'party', organise a wedding, travel, and catch up ... This includes the enchanting bride-to­ be Zara, Laura, Aline (i.e. Amex IT expert), Ruby, Sr. Margaret (the gifted branch!), and the State Bodybuilding Champion Charbel (Bro, are we related?). Mum and Dad, my guardian angels, where would I be without you?. A special thanks to my grandparents for their love and prayers: Najla and Joseph Antoun, and Teta Atour. During my PhD, I had to say goodbye to two very dear persons to whom I am greatly indebted. First, the Poet Naeim Khoury, for his amazing talent, kind support and bravery. I have never witnessed anyone fight lung cancer like this man. He remains an unending inspiration to me, to his oncologists and cancer researchers alike ... Second, and most recently, to my dear late grandfather, the legendary Mayor Ibrahim Kairouz, for his constant loving support, and for teaching me what I will never forget... May this work be a tribute to these inspiring souls.

Ah! My beautiful Patroness, 'Star of the sea', how can one ever thank another for saving their life? Your Maternal guidance is always my shield ... Finally, and most importantly, to the Knight in Shining Armour I have been waiting for my entire life, who, cleverly hidden but always so close, finally succeeded in rescuing me. Every single heartbeat I make belongs to you, my Merciful Knight, and I look forward to galloping away with you up that steep mountain into eternal bliss ...

IV !Jl.bstract

The Grb7 family of SH2 domain-containing signalling proteins, which comprises Grb7, -10 and -14, perform both adaptor and regulatory roles in receptor tyrosine kinase (RTK) signalling. Overexpression of Grb7 family proteins has been detected in several types of human cancer or cancer cell lines, and Grb7 has been linked with enhanced tumour invasion. In addition, splice variants with alterations in key functional domains that display distinct RTK recruitment profiles have been identified for Grb7 and -10. However, the function of Grb14, the most recent member of this family, remains undefined and its interacting partners unidentified. Therefore, this study sought to characterise the role of Grb 14 in cellular signalling.

Using RT-PCR, a novel human Grb14 isoform was identified and designated Grb14 ~- This isoform contains a 94 bp deletion in SH2 domain-encoding sequences and a premature termination codon that truncates the SH2 module and abrogates its function. 3/12 cDNA clones isolated from a human breast epithelial cDNA library also contained this deletion. Further, PCR from genomic DNA demonstrated that the deleted region in the

Grb14 gene is flanked by exon-intron junctions, suggesting that Grb14 ~ transcripts originate by alternative mRNA splicing. Interestingly, differential expression of Grb14 isoforms was detected amongst breast cancer cell lines. In order to functionally characterise the two variants, association with particular RTKs was investigated by co­ immunoprecipitation analysis and GST fusion protein pulldowns. FGFR-1 was identified as a novel Grb14 a binding partner. Further analysis revealed that both Grb14 a and~ bind the insulin receptor in an interaction mediated by the 'between PH and SH2' (BPS) domain and that the SH2 domain is dispensable for this signalling role. However Grb14 a, but not ~' bound FGFR-1 since this interaction requires a functional SH2 domain. Therefore, alternative mRNA splicing regulates association of Grb14 with particular RTKs, where the distinct functional modules mediate interactions with different receptors.

V While this work was in progress, a study demonstrated inhibition of FGF-induced DNA synthesis by overexpression of Grb14. It also reported enhancement of this FGF-induced response by overexpression of a Grb14 protein that cannot bind the FGFR-1 due to a generated point-mutation (R466K) in the SH2 domain. Since this may have functional implications for the novel Grb14 splice variant which also fails to mediate receptor binding, the functional effect of Grb 14 ~ in FGF-stimulated proliferation of fibroblasts was examined using thymidine incorporation analysis and flow cytometry. This revealed that

Grb14 ~ did not affect FGF-stimulated DNA synthesis or cell cycle progression. A significant effect of Grb14 a overexpression on these endpoints was not observed, indicating that Grb14 is not a major physiological regulator of FGF-induced mitogenesis.

Grb14 is implicated in regulation of insulin signalling and it is differentially expressed in breast cancer cell lines, although little is known about its regulation. Thus, the regulation of this protein by insulin/IGF-I and estrogen was investigated, since these hormones synergise in the regulation of breast cancer cell proliferation. In MCF-7 cells maintained in charcoal-stripped serum, Grb14 expression was downregulated by estradiol and increased by the pure antiestrogen ICI 182780. Under serum-free conditions, insulin enhanced Grb14 expression but this effect was repressed by estradiol when both hormones were used in combination. Using a system in which c-Myc induction drives cell cycle progression independently of estradiol, Grb14 regulation was specific to estradiol treatment. Finally, a novel functional role was revealed for Grb14 whereby its overexpression inhibited not only insulin- but also estrogen-induced cell cycle progression. These data represent the first demonstration of regulation of Grb14 expression levels in response to hormonal stimuli, and are consistent with its role as a repressor of insulin/IGF signalling. Furthermore, they implicate a modulatory role for Grb14 in hormonal cooperativity in regulating the proliferation of breast cancer cells.

vi TABLE OF CONTENTS

Dedication ------ii Ac know Ie d gmen t s ______iii Abstract ______v Figures ______x Abbreviations xii Chapter 1 - Literature Review ______1 1. 1 Signalling by receptor tyrosine kinases (RTKs) 1 1.1.1 Activation ofRTK subclasses 3 1.1.2 Signalling modules 6 1. 1.3 Types of SH2 domain-containing signalling proteins 9 1.1.3.1 Catalytic proteins 10 1.1.3.2 Adaptor proteins 11 1.1.4 The Ras/MAPK signalling cascade 12 1.2 Signalling in breast cancer 14 1.2.1 Growth factor receptor pathways 15 1.2.1.1 The ErbB family 15 1.2.1.2 The FGFR family 17 1.2.1.3 The IR family 19 1.2.2 SH2 domain-containing signalling intermediates 23 1.2.2.1 c-Src 23 1.2.2.2 PLC-yl 24 1.2.2.3 Grb2 24 1.2.3 Signalling by estrogens 26 1.2.4 Cross-talk between estrogen and insulin/IGF signalling 29 1.3 The Grb7 family 30 1.3.1 Structure 30 1.3.2 Expression 33 1.3.3 Isoforms 33 1.3.4 RTK-dependent interactions 36 1.3.5 RTK-independent interactions 36 1.3.6 Phosphorylation 38 1.3.7 Function 39 1.3.7.1 Insulin/IGF-1-induced mitogenic and metabolic effects 40 1.3.7.2 Cell Migration 41 1.3.7.3 Apoptosis 42 1.3.7.4 Cancer 43 Objectives. ______45 Chapter 2- Materials &Methods 46 2.1 Molecular biology 46 2.1.1 The polymerase chain reaction (PCR) 46 2.1.2 RNA extraction 47 2.1.3 cDNA synthesis and RT-PCR 47 2.1.4 Subcloning of PCR products and generation of expression constructs 49 2.1.5 Generation of retroviral constructs 50 2.1.6 Cycle sequencing 50

vii 2.2 Western blotting and Immunoprecipitations ______52 2.2.1 Antibodies ______52 2.2.2 Lysate preparation ______53 2.2.3 Western blotting analysis ______53 2.2.4 lmmunoprecipitation assays ______54 2.2.5 Preparation of GST fusion proteins ______54 2.2.6 In vitro binding assays ______55 2.3 Cell culture ______56 2.3.1 Mitogens and inhibitors, ______56 2.3.2 Cell lines/strains ______56 2.3.3 Generation of MCF-7/EcoR stable pools ______57 2.3.4 Retroviral infection and generation ofMCF-7/EcoR control or Grbl4-overexpressing stable pools. ______58 2.3.5 Cell culture assays ______59 2.3.6 Thymidine incorporation analysis ______60 2.3.7 Flow cytometry ______60 Chapter 3-Identification of a novel Grb14 isoform generated by alternative mRNA splicing ______62 3.1 Introduction ______62 3.2 Identification of Grbl4 isoforms by RT-PCR ______63 3.3 Characterisation of an alternatively-spliced Grbl4 isoform, Grbl4 /3 ______65 3.3.1 Characterisation of the Grbl4 C-terminal deletion ______65 3.3.2 Expression of Grbl4 isoforms in breast cancer cell lines ______66 3.3.3 Alternative splicing of the Grbl4 SH2 domain-encoding sequence ______69 3.4 Discussion ______69 Chapter 4-Interaction of Grb14 isoforms with RTKs ______75 4.1 Introduction ______75 4.2 Generation of a Grbl4 /3 expression vector and Grbl4 GST fusion proteins ______76 4.3 Interaction of Grbl4 isoforms with the EGFR and PDGFR ______76 4.4 Interaction of Grbl4 isoforms with the IGF-IR and IR ______79 4.5 Kinetics of IR interaction with Grbl4 isoforms ______83 4.6 Interaction ofGrbl4 isoforms with the FGFR-1 ______88 4.7 Discussion ______90 Chapter 5- Functional effects of Grb14 isoforms in FGF-stimulated cell proliferation ______95 5.1 Introduction 95 5.2 Overexpression of Grbl4 isoforms in 3T3 fibroblasts using retroviral infection 96 5.3 FGF-induced DNA synthesis rate of 3T3 fibroblasts using thymidine incorporation analysis ______97 5.4 Effects of Grbl4 isoforms on the DNA synthesis rate of mouse fibroblasts using thymidine incorporation analysis ______99

Vlll 5.5 Effects ofGrb14 isoforms on cell cycle phase distribution of mouse fibroblasts in response to FGF stimulation ______100 5.6 Insulin-induced DNA synthesis of mouse fibroblasts 103 5. 7 Discussion 103 Chapter 6-Regulation of Grb14 in estrogen/insulin cross-talk and effects on proliferation ofMCF-7 breast cancer cells ______108 6.1 Introduction 108 6.2 Grb14 expression in breast cancer cell lines 109 6.3 Regulation of Grb14 protein expression by estrogen and the antiestrogen IC/ 182780 ______113 6.4 Regulation ofGrb14 protein expression by estrogen and insulin/IGFs______l15 6.5 Regulation of Grb14 by either estrogen-induced or estrogen-independent cell cycle progression ______115 6.6 Grb14 overexpression in MCF7/EcoR cells 121 6.7 Effect ofGrb14 overexpression on insulin-induced cell cycle progression 124 6.8 Effect of Grb14 overexpression on both estrogen- and insulin-induced cell cycle progression ______126 6.9 Discussion 128 Chapter 7-General Discussion ______135 7.1 Inhibition of IR signalling by GrblO and -14 135 7.2 The function of Grb14 in signalling by other RTKs 140 7.3 Differential interactions of Grb14 isofonns with RTKs and effectors 140 7.4 The role of Grb7 in RTK signalling 142 7.5 Regulation of Grb14 and its role in the cross-talk between insulin/IGF-l and estrogen signalling ______142 References 150 Appendix A 198 Appendix B 200 Publications and A wards arising from this thesis 201

ix 'Figures

Figure 1.1 Schematic representation of several RTKs 2 Figure 1.2 Ligand-stimulated dimerisation and activation of RTKs 4 Figure 1.3 Structure of signalling intermediates exhibiting aberrant expression in human breast cancer 25 Figure 1.4 Structure of the Grb7 family proteins 31 Figure 1.5 Grb7 family isoforms 35

Figure 3.1 Identification of a naturally occurring Grbl4 transcript with a deletion in SH2 domain-encoding sequences ______64 Figure 3.2 Characterisation of a novel Grb14 isoform, Grb14 ~. arising from alternative mRNA splicino------67

Figure 3.3 Genomic origin of the Grb14 ~ variant transcript ______70 Figure 3.4 Comparison of splice junctions for SH2 domains ______71

Figure 4.1 Generation of GST fusion proteins for pulldown analysis 77 Figure 4.2 Interaction of Grb14 isoforms with the EGFR and PDGFR 78 Figure 4.3 Interaction of Grb14 isoforms with the IGF-IR and IR 81 Figure 4.4 Kinetics of Grb 14 a interaction with the IR 85 Figure 4.5 Interaction kinetics of Grb 14 isoforms and the IR 87 Figure 4.6 Interaction of Grb14 isoforms with FGFR-1 89

Figure 5.1 Developing the retroviral infection/thymidine incorporation assay____ 98 Figure 5.2 Effects of Grb14 isoforms on FGF-induced rate of DNA synthesis 101 Figure 5.3 Effects of Grb14 isoforms on FGF-stimulated cell cycle progression. __ 102 Figure 5.4 Effects of EGF/insulin stimulation on DNA synthesis of 3T3 fibroblasts_ 104

X Figure 6.1 Grb 14 protein expression in normal human breast epithelial cells and breast cancer cell lines 111 Figure 6.2 Regulation of Grb14 protein expression by estrogens and antiestrogens _ 114 Figure 6.3 Regulation of Grb14 protein expression by insulin and estrogen 116 Figure 6.4 Dissociation of estrogen regulation of Grb14 from cell cycle progression 119 Figure 6.5 Grb14 overexpression in MCF-7/EcoR cells 123 Figure 6.6 Effect of Grb14 overexpression on insulin-induced cell cycle progression 125 Figure 6.7 Effect of Grb14 overexpression on insulin- and estradiol-induced cell cycle progression 127 Figure 6.8 Analysis of the Grb14 promoter sequence for candidate EREs 131

Figure 7.1 Model of Grb14 function in insulin signallino------­ 136 Figure 7 .2 Interactions of Grb 14 isoforms and the Grb 14 R466K SH2 mutant with potential regulators and effectors ______143 Figure 7.3 Model of Grb14 function in insulin/lGF-1 and estrogen cross-talk. ___ 147

XI !ll.bbreviations

aFGF acidic fibroblast growth factor ATP adenosine-5' -triphosphate bFGF basic fibroblast growth factor bp base pairs BPS between PH and SH2 BSA bovine serum albumin cdk cyclin-dependent kinase cDNA complementary DNA CSF-I colony stimulating factor I DAG diacylglycerol DMEM Dulbecco's Modified Eagle Medium DMSO dimethylsulfoxide DNA deoxyribonucleic acid DTT dithiothreitol ECL enhanced chemiluminescence EGF epidermal growth factor EGFR epidermal growth factor receptor ER estrogen receptor ERE estrogen response element Erk extracellular signal-regulated kinase FAK focal adhesion kinase FCS fetal calf serum FGF fibroblast growth factor FGFR fibroblast growth factor receptor FRS-2 fibroblast growth factor receptor substrate -2 GAP GTPase activating protein GDP guanine diphosphate GEF guanine nucleotide exchange factor GHR growth hormone receptor GM Grb and Mig Grb growth factor receptor-bound GST glutathione S-transferase GTP guanine triphosphate GTPase guanine triphosphatase h hours HEPES N-[2-hydroxyethyl]piperazine-N' -[2-ethanesulfonic acid] HRP horseradish peroxidase IGFBP insulin-like growth factor binding protein IGF-1 insulin-like growth factor-I IGF-IR insulin-like growth factor-I receptor IgG immunoglobulin

xii IR insulin receptor JAKs janus kinases kb kilobase pairs kDa kilodaltons MAPK mitogen-activated protein kinase min minutes mRNA messenger ribonucleic acid NBCS newborn calf serum O.D. optical density PBS phosphate-buffered saline PCR polymerase chain reaction PDGFR platelet-derived growth factor receptor PH pleckstrin homology PI-3 kinase phosphatidylinositol-3' kinase PKB protein kinase B PLC phospholipase C PMSF phenymethyl sulfonyl fluoride pRb retinoblastoma protein PTB phosphotyrosine-binding PTP protein tyrosine phosphatase pTyr phosphotyrosine PyVMT polyoma virus middle T RTK receptor tyrosine kinase RT-PCR reverse transcription polymerase chain reaction s seconds SDS sodium dodecyl sulphate SDS-PAGE sodium dodecyl sulphate polyacrylarnide gel electrophoresis Ser serine SERMS selective estrogen receptor modulators SHP SH2-containing phosphatase SH2 Src homology 2 SH3 Src homology 3 SRC-1 steroid receptor co-activator-I socs suppressor of cytokine signalling Sos son-of-sevenless STATS signal transducers and activators of transcription TBS tris-buffered saline TGF-a transforming growth factor-a Thr threonine Tris tris(hydroxymethyl)arninomethane Zn Zinc

Xlll Chayter 1 - Literature 'Review

1.1 Si~nallin~ by receptor tyrosine kinases (RTKs)

Signal transduction governs a plethora of processes imperative for cellular function such as growth, differentiation, and metabolism (Schlessinger and Ullrich, 1992). Indeed, deregulation in cellular signalling due to mutations, aberrant expression or activation of signalling proteins can lead to serious pathological conditions such as achondroplasia and Crouzon syndrome, which result in human dwarfism due to mutations in the transmembrane domain of fibroblast growth factor receptor-3 (FGFR-3) (Rousseau et al., 1994; Meyers et al., 1995). The critical function of signal transduction is classically exemplified in cancer, where signal deregulation either from the loss of negative feedback pathways and/or amplification of oncogenic signals leads to uncontrolled proliferation. This is demonstrated for numerous growth factors, receptors and signalling intermediates, including erbB2/HER2 which exhibits oncogenic activity in breast cancer (Slamon et al., 1987) and for FGFR-2 and -3 which are mutated in human cervical, bladder and colorectal carcinomas (Cappellen et al., 1999; Jang et al., 2001).

Protein tyrosine kinases are key players in signal transduction and are subdivided into two classes: Receptor tyrosine kinases (RTKs) and non-receptor tyrosine kinases. Both these classes of proteins transmit signals through activation of their catalytic domains. Receptor tyrosine kinases contain an extracellular region, which is often glycosylated or contains 0- linked sugars, single membrane-spanning regions with hydrophobic residues and an intracellular sequence consisting of a juxtamembrane region and the kinase domain, in addition to specific regulatory sequences containing phosphorylation sites (Figure 1.1) (Schlessinger and Ullrich, 1992; van der Geer and Hunter, 1994; Hunter, 1998). Non­ receptor tyrosine kinases, however, lack the extracellular region and do not span the membrane but upon activation of their catalytic domains are able to transmit the signal to downstream effectors. Some associate with receptors that lack enzymatic activity and

1 IR EGFR PDGFR FGFR NGFR EPH TIE RET HGFR ROR1 RYK KLG DDR MuSK FL T1 ______J

[] I Cysteine rich regions Kinase domain O Cadherin domain Kringle motif ~ Leucine-rich motif

? lgG-like repeat Fibronectin Ill repeat EGF-like domain I Discoidin I-like domain

Figure 1.1 Schematic representation of several RTKs (modified from Ullrich et al., 1990; van der Geer et al. , 1994 and Hunter, 1997). RTKs consist of an extracellular ligand-binding unit connected to a transmembrane domain and an intracellular tail which harbours catalytic regions. A subset of RTKs relevant to this thesis are enclosed in a square on the left hand side of the figure. hence facilitate signal transmission (Bolen, 1993), for example the janus kinases (JAKs) bind to cytokine receptors such as those for interferon-y, interleukins or erythropoietin and mediate their effects (Darnell et al., 1994; Ihle, 1995). This literature review section focuses on the mechanisms by which RTKs elicit specific biological effects required for cellular function.

1.1.1 Activation of RTK subclasses RTKs are grouped into subclasses according to their structure. Subclass I receptors, such as the erbB family proteins, are monomeric and possess two cysteine-rich repeats in their extracellular domain. Subclass II receptors are heterotetrameric and are characterised by a2 ~2 disulfide-linked subunits with 2 extracellular cysteine repeats and are represented by the closely related IR and IGF-IR. Subclass III RTKs are characterised by 5 immunoglobulin-like (lgG-like) repeats in the extracellular region and include the PDGFRs, the CSF-lR and c-kit. On the other hand, subclass IV contain 3 extracellular IgG­ like repeats and are exemplified by the FGFR family (Ullrich and Schlessinger, 1990). All inactive RTKs exist as monomers in the cell membrane with the exception of the IR and Met receptor families (Hubbard and Till, 2000; Schlessinger, 2000). The IR family are comprised of 2 single short extracellular a chains joined by a disulfide bond to membrane­ spanning ~-chains, resulting in an a2 ~2 heterotetrameric structure. The Met receptor family structure is simpler, with a small a chain joined by disulfide bonding to a ~ chain that extends across the membrane (Hubbard and Till, 2000).

Ligand binding to RTKs induces receptor dimerisation (Figure 1.2), followed by a conformational change leading to juxtaposition of the intracellular catalytic domains and transphosphorylation of residues within the activation loop of each kinase domain. This activates the enzymatic function of the receptor, leading to autophosphorylation of specific tyrosines within the adjacent intracellular regulatory sequences, thus providing binding sites for downstream effectors (van der Geer and Hunter, 1994; Hunter, 2000). RTKs phosphorylate their substrates by catalysing the transfer of the y-ATP phosphate to the

3 A B

C

p p

I Transcription and biolog ica I effects

Figure 1.2 Ligand-stimulated dimerisation and activation of RTKs. A. A generic RTK is inactive as a monomer B. Ligand-binding stabilises RTK dimerisation, leading to a conformational change that results in activation of the kinase domains and trans­ mediated autophosphorylation of the cytoplasmic tail. C. Phosphorylated tyrosine (pTyr) residues in the receptor tail create target sites for intracellular proteins with functional modules that bind pTyr such as Grb2 and PI-3 kinase, which then relay the signal to downstream effectors. Thi i exemplified by Grb2 signalling via the Ras/MAPK pathway which activate transcription factors and results in cell proliferation or differentiation. In addition, PI-3 Kinase results in activation of Akt/PKB which modulate cell survival. tyrosine hydroxyl group of the substrate (Hubbard and Till, 2000). There are several modes of ligand-induced receptor dimerisation. For bivalent ligands such as growth hormone and erythropoietin, one ligand molecule binds to two receptors and induces the formation of an active dimer. For dimerised growth factors such as VEGF and PDGF, signalling is initiated when a ligand pair crosslinks 2 receptor molecules and activates/stabilises the receptor dimerisation process. There are 2 subunits of the PDGF ligand, A and B, forming either AA, BB homodimers or AB heterodimers. Similarly, the receptor dimer consists of either aa, ~~ or a~ types. While a-PDGFR binds all three dimeric forms of the PDGF ligand,~­ PDGFR preferentially associates with PDGF-BB, but does not bind to PDGF-AA (Heldin et al., 1988; Heldin et al., 1989; Schlessinger and Ullrich, 1992).

Fibroblast growth factors (FGFs) bind to their receptors monovalently (resulting in a 2:2 ligand:receptor complex) (Schlessinger, 2000) although this is insufficient to elicit activation, since binding of soluble or cell-surface heparin sulfate proteoglycans (HSPGs) is required for dimerisation resulting in multimeric ligand-receptor complexes (Spivak­ Kroizman et al., 1994). On the other hand, for the IR family, one molecule of insulin binds simultaneously to both a subunits, where it associates with the Ll and cysteine-rich domain of an a chain from one receptor, and with the L2 domain of the achain from another receptor. This leads to a conformational change in the quaternary structure allowing juxtaposition and activation of the kinase domains (Cheatham and Kahn, 1995; Luo et al., 1999; Hubbard and Till, 2000; Schlessinger, 2000 ). For RTKs to achieve transauto­ phosphorylation, both extra- and intracellular domains of each receptor within the dimer must be oriented in a certain conformation to achieve activation. Schlessinger (2000) proposes a model where an equilibrium exists on the cell surface between single receptor monomers and the formation of receptor dimers unbound by ligand. A small number of receptor dimers adopt the configuration required for activation and this is stabilised by ligand binding. Receptor dimers are active even in the absence of ligand, since RTK phosphorylation can be increased by phosphatase inhibitors and by protein overexpression of receptors in the absence of ligand (Schlessinger, 2000).

5 1.1.2 Signalling modules In any mode of ligand-induced receptor dimerisation, the resulting effect is receptor activation and phosphorylation of tyrosine residues embedded in specific sequences within the intracellular domain to provide binding sites for the recruitment of protein effectors (Figure 1.2). Such signalling proteins generally contain phosphotyrosine binding modules, such as Src homology 2 (SH2) and phosphotyrosine binding (PTB) domains (Pawson and Scott, 1997), although the former represents the most prevailing type of pTyr-binding domain in RTK signalling (Hunter, 2000). SH2 domains, first identified in the retroviral oncoprotein v-Fps, are modules of approximately 100 amino acids that recognise phosphotyrosines within specific sequence contexts. This means that SH2 domains also recognise specific sequences (3-6 residues) C-terminal to the phosphorylated moiety. These additional sequences are distinct for different SH2 domains, and confer specificity in the recruitment of SH2 domain-containing proteins to particular receptors, thereby specifying the signalling pathways mediated by these receptors (Pawson and Scott, 1997; Pawson et al., 2001).

Structural studies have shown that SH2 domains consist of an anti-parallel Bsheet flanked by two a helices. They possess a bipartite binding site consisting of a positively charged phosphotyrosine (pTyr)-binding pocket on one side of the B sheet, and a pocket or an extended groove for binding specific residues C-terminal to the phosphotyrosine (Pawson et al., 2001 ). The conserved binding motif for SH2 domains contains the FL VRES consensus sequence, with an invariant arginine residue at the BB5 position. This residue makes strong hydrogen bonds with two oxygens in the phosphate and is critical for interacting with phosphorylated moieties, since it disrupts binding when mutated (Kuriyan and Cowbum, 1993; Pawson and Scott, 1997).

Studies employing degenerate phosphopeptide libraries have indicated that certain SH2 domains such as those belonging to Src and Lek, specifically bind to the sequence pTyr­ Glu-Glu-Ile. They bind to the first 3 residues immediately C-terminal to the pTyr, where hydrophilic residues are located at the + 1 and +2 positions and the hydrophobic residue at +3 is fitted in a small pocket on the binding site. On the other hand, SH2 domains found in

6 PLC-yl and SHP-2/Syp have an extended groove following the pTyr-binding site to accommodate at least five preferred hydrophobic residues adjacent to the phosphorylated moiety (Pawson and Scott, 1997).

When isolated, SH2 domains fold into a functional module independently of the remaining protein, and their amino and carboxyl ends fold so that they are in proximity to one another away from the phosphotyrosine-binding site (Pawson and Nash, 2000). Interestingly, despite the specificity conferred in SH2-domain binding to phosphorylated tyrosine residues, there are cases when binding occurs with unphosphorylated sequences. The SH2D1NSAP protein possesses a sole SH2 domain that binds to the consensus sequence Thr-Ile-X-Tyr-X-X-Val of the CD150/SLAM T-cell receptor regardless of phosphorylation. However, binding affinity increases approximately five fold to a phosphorylated target. The SAP SH2 domain is distinctive in that it has a larger binding site and recognises residues both N- and C-terminal to the tyrosine residue (Pawson et al., 2001). Furthermore, the SH2 domain of the adaptor protein GrblO interacts with the C2 domain of Nedd4, an E3 ubiquitin protein ligase. This interaction is constitutive and is independent of tyrosine phosphorylation, since GrblO appears to preferentially associate with the unphosphorylated Nedd4 (Morrione et al., 1999).

Specificity of SH2 domains also depends on the number of these functional units and their location in the protein. Tandem SH2 domains are found in a range of proteins including the p85 regulatory subunit of PI-3 kinase, ZAP-70, Syk, SHP-2 and PLC-yl. Using tyrosine­ based activation motifs (TAMs), binding studies indicated that tandem SH2 domains display higher affinity and binding specificity relative to individual SH2 domains. Interestingly, the binding selectivity differs between proteins, for example the ZAP-70 and Syk tandem SH2 domains bind optimally to bisphosphorylated sites separated by 9-11 residues, whereas SHP-2 binding requires more widely spaced binding sites. On the other hand, p85 shows more flexibility and is able to bind YMXM motifs separated by different distances (Ottinger et al., 1998).

7 Phosphotyrosine-binding (PTB) domains are conserved sequences of 100-150 amino acids, a subset of which recognize the consensus sequence NPXpY. Unlike SH2 domains they recognise additional specific sequences (3-5 residues) N-terminal to the pTyr, and are not all dependent on phosphorylation for high affinity binding. A subset of PTB-containing proteins, including She and IRS-1, bind to phosphorylated sequences, others associate with non-phosphorylated sequences and yet another group bind to both (Hunter, 2000; Schlessinger, 2000; Pawson et al., 2001). The fibroblast receptor substrate 2 (FRS2) PTB is an example of the latter, as it binds to the phosphorylated TrkA neurotrophin receptor and to a distinct consensus site on the FGFR juxtamembrane region that lacks Tyr and Asn residues and is not phosphorylated (Ong et al., 2000). This difference in binding specificity may be regulated by the conformation of the PTB domain (Yan et al., 2002).

In addition to phosphotyrosine-binding regions, a plethora of other modules may also be found in signalling proteins. These modules, ranging in size from 40-150 residues, are able to fold independently as separate functional units, and bind consensus sequences 4-10 amino acids in length. Similar to SH2 domains, binding for certain modules may be regulated by phosphorylation on specific Tyr residues, or may require Ser/Thr phosphorylation within the consensus sequence (Hunter, 2000).

SH3, WW and EVH 1 domains bind to specific proline-rich sequences. SH3 domains consist of 50-75 residues and bind proline-rich motifs with the core sequence XPpXP, where X is likely to be a hydrophobic residue and p tends to be a proline residue (Pawson and Schlessinger, 1993; Pawson and Scott, 1997; Vidal et al., 2001). The two conserved prolines are critical for high affinity interactions. SH3-associating proteins are designated as class I or II, depending on their binding configuration. While class I peptides bind SH3 domains in an N- to C-terminal orientation, class II ligands adopt a C- to N-terminal configuration (Pawson and Scott, 1997). WW domains bind favourably to the PPXY or PPLP motifs (where P is proline, X any nonproline residue and Lis leucine) (Aghazadeh and Rosen, 1999). They also interact with specific pSer/pThr-containing sequences, for example, where they are implicated in mediating interactions of ubiquitin ligases with their substrates in the degradation targeting process (Lu et al., 1999). Unlike SH3 and WW

8 domains, EVHl modules bind a distinct proline-rich sequence with the core motif FPPPP and are involved in actin remodelling (Niebuhr et al., 1997).

EH domains occur frequently in polypeptides that function in protein trafficking, and associate with Asn-Pro-Phe sequences (Pawson and Nash, 2000), while SAM domains are involved in homo- and hetero-oligomerisation with other SAM modules (Stapleton et al., 1999; Thanos et al., 1999 ). They are found in a variety of signalling proteins including the Eph proteins which constitute the largest family of RTKs (Thanos et al., 1999). PDZ domains associate with each other (Hillier et al., 1999) and also bind to short, unique motifs at the C-terminal end of proteins. For example, a family of PDZ domains including those of the Discs Large protein binds the consensus motif Glu-(Ser/Thr)-X-(Val/Ile), where X denotes any amino acid (Songyang et al., 1997; Pawson and Nash, 2000).

However, not all signalling modules mediate protein-protein interactions, as some, such as FYVE and PH domains, associate with specific phospholipids. This allows protein targeting to the cellular and plasma membranes for further interactions with substrates or regulatory molecules. FYVE domains are characterised by a conserved basic motif (R/K­ R/K-H-H-C-R) and bind phosphatidylinositol-3 phosphate (Fruman et al., 1999; Pawson and Nash, 2000). Similarly, PH domains bind to specific polyphosphoinositides and inositol polyphosphates, and thus regulate membrane association and/or responses to the generation of specific second messengers such as phosphatidylinositol (3,4) bisphosphate generated by PI-3 kinase (Pawson and Scott, 1997). Oligomerisation is described for both FYVE and PH domains, and may serve to increase binding affinity to phosphoinositides (Klein et al., 1998; Kutateladze et al., 1999; Mao et al., 2000).

1.1.3 Types of SH2 domain-containing signalling proteins SH2 domain-containing proteins can be divided into two classes (Figure 1.3): Class I proteins possess enzymatic activity and include c-Src and phospholipase C (PLC)-yl; Class II SH2 proteins, such as growth factor receptor-bound (Grb)2, do not have a known catalytic activity and function as adaptor proteins, linking receptors or cytoplasmic tyrosine kinases to downstream effectors (Schlessinger and Ullrich, 1992).

9 1.1.3.1 Catalytic proteins Catalytic proteins display tyrosine kinase, tyrosine phosphatase or phosphodiesterase activities, among others. Examples include class IA PI-3 kinases, which phosphorylate phosphoinositides (Pls) on the D3 position and generate 3' phosphoinositides following

RTK activation, particularly PI3,4 P2 and PI3,4,5P3 (PIP3). These PI-3 kinases consist of a p85, p55 or p50 regulatory subunit which binds RTKs via its SH2 domains and a pll0 catalytic subunit which mediates enzymatic activity (Katso et al., 2001; Cantley, 2002). Protein tyrosine phosphatases catalyse the dephosphorylation of tyrosine residues in activated RTKs as well as their signalling intermediates and includes the SH2-containing phosphatases 1 and 2 (SH-PTP/SHP-1 & 2). SHP-1 is implicated in the suppression of mitogenic pathways (van der Geer and Hunter, 1994; Hennige et al., 2001) and plays a critical role in the haematopoeitic system. Loss-of-function mutations in SHP-1 result in significant immunological dysfunction, including chronic inflammation and systemic autoimmune disease. This is evident in moth-eaten (me/me) mice and viable moth-eaten mice (mev/mev) which lack SHP-1 and have a catalytically defective SHP-1, respectively. These mice have a significant increase and accumulation of myeloid/monocytic cells which leads to the characteristic patchy dermatitis (Zhang et al., 2000). Unlike SHP-1, SHP-2 has a positive effect on the Ras/MAPK signalling cascade in response to insulin stimulation and also functions downstream of growth hormone and T-cell receptor-initiated pathways (Noguchi et al., 1994; van der Geer and Hunter, 1994).

Phospholipase C-yl (PLC-yl) cleaves phosphatidylinositol (4,5) bisphosphate into the second messengers inositol-1,4,5-trisphosphate and diacylglycerol, leading to calcium mobilisation and activation of protein kinase C, respectively. PLC-yl interacts with several receptors such as EGFR, PDGFR, erbB2 and FGFR and is subsequently phosphorylated on tyrosine residues. It binds the activated receptors via two SH2 domains (Figure 1.3) and contains an SH3 module for localisation to the cytoskeleton (van der Geer and Hunter, 1994).

10 1.1.3.2 Adaptor proteins Adaptor molecules play an important role in the spatial and temporal regulation of signalling cascades. They lack catalytic activity and hence bind to effector proteins with enzymatic activity to mediate a signalling cascade. They usually contain combinations of binding modules that selectively allow simultaneous interactions with binding partners and subsequent signal propagation. Two primary examples of SH2- and/or SH3-containing adaptors are outlined here. The Grb2 adaptor has one SH2 domain sandwiched between two SH3 regions and is vital for the transmission of various RTK pathways. The Grb2 SH2 domain binds activated RTKs while the SH3 domains interact with the C-terminal proline­ rich region of son-of-sevenless (Sos). The discovery that Grb2 is constitutively associated with mammalian Sos linked Ras activation to growth factor signalling. The interaction with RTKs translocates Grb2 to the plasma membrane, thus bringing Sos in proximity with the membrane-bound Ras. Sos then catalyses the activation of Ras, thereby transiently increasing the content of Ras-GTP and allowing the activation of downstream signalling components (Campbell et al., 1998; Hunter, 2000). Interaction of Grb2 with RTKs can also occur indirectly via binding to She or the SH2-containing tyrosine phosphatase SHP-2/syp

(Pawson, 1995).

The PI-3 kinase subunit p85 functions as an adaptor and a regulatory protein, where it serves to recruit the p 110 catalytic subunit to the plasma membrane, and affects its activation. In addition, p85 consists of several functional modules that mediate interactions with multiple signalling proteins. The p85a and ~ subunits contain two SH2 modules, in addition to a Bcr homology (BH) domain, proline-rich regions and an N-terminal SH3 domain (Fruman et al., 1999; Funaki et al., 2000; Katso et al., 2001). The BH module is homologous to the Rho GAP domain of Bcr and binds to the G-protein Cdc42Hs, leading to activation of PI-3 kinase (Funaki et al., 2000). A p85 inter-SH2 region allows binding to the N-terminus of pll0. The association of p85 with pll0 inhibits the catalytic activity of the latter. However, binding of the p85 SH2 domains to tyrosine-phosphorylated proteins releases the inhibition, allowing activation of pl 10 and phosphorylation of its substrates. The p85 SH2 associates with several proteins including IRS-1, ~-PDGFR, CSF-lR, c-kit, and the IR (Backer et al., 1992; Klippel et al., 1992; McGlade et al., 1992; Funaki et al.,

11 2000; Katso et al., 2001). p85 also interacts via its SH3 domain with proline-rich regions of proteins such as Sos, Cbl or dynamin. Moreover, p85a interacts via its proline-rich regions with Src and other family members including Lek, Lyn and Fyn (Fruman et al., 1999; Funaki et al., 2000; Katso et al., 2001).

1.1.4 The Ras/MAPK signalling cascade The Ras proteins belong to the Ras superfamily of GTP-binding proteins, which includes approximately 60 mammalian genes (Bar-Sagi and Hall, 2000). There are 4 human ras proteins (H-Ras, N-Ras and K-Ras4A and K-Ras4B) that function as molecular switches, where they cycle between the inactive GDP-bound form and the active GTP-bound state. This is regulated by guanine nucleotide exchange factors (GEFs) and GTPase-activating proteins (GAPs) (van der Geer and Hunter, 1994; Campbell et al., 1998; Bar-Sagi and Hall, 2000). GEFs that act on Ras are defined by a Cdc25 domain originally identified in yeast (Bar-Sagi and Hall, 2000) and include RasGRF/mCDC25 as well as Sos 1/2. They activate Ras by catalysing the exchange of GDP for GTP, the latter being present at higher intracellular concentrations. RasGAP activates further the low intrinsic Ras GTPase activity, allowing the hydrolysis of GTP and returning Ras to its inactive GDP-bound form. Examples of mammalian RasGAPs are p120 RasGAP, NFl-GAP/neurofibromin and GAPlm (Marshall, 1996). p120 RasGAP associates with numerous RTKs such as the ~-PDGFR via its SH2 domains and is phosphorylated on tyrosine residues in response to PDGF and EGF stimulation (van der Geer and Hunter, 1994). Constitutive Ras activation occurs when mutations at residues 12, 13, or 16 render the protein insensitive to GAP stimulation, and it remains in the GTP-bound, activated state (Campbell et al., 1998). Ras is anchored to the membrane by C-terminal famesylation with additional palmitoylation in the case of N-Ras and H-Ras (Malaney and Daly, 2001). Ras constitutes a point of convergence for many signalling pathways by undergoing transient activation in response to diverse signals activating RTKs, non-receptor TKs and G-protein coupled seven transmembrane receptors. The best characterised Ras pathway occurs downstream of the EGFR and was discovered to be a fundamental signal transduction cascade where each component is conserved in worms, flies and humans (Campbell et al., 1998). In this pathway, phosphorylation of the EGFR recruits the Grb2/Sos complex, where Grb2 binds

12 Tyr-1068 within the consensus sequence pTyr-X-Asn-X of the receptor cytoplasmic tail via its sole SH2 domain (Pawson, 1995).

This leads to Ras activation and subsequent association with downstream effectors, including the Raf-1 Ser/Thr kinase which translocates from the cytoplasm to the plasma membrane upon association. The localisation of Ras to the plasma membrane is essential to mediate its role in binding Raf-1 and additional effectors, such as PI-3 kinase and the Ral GEFs which activate Ral GTPases. The C-terminal catalytic domain of Raf then binds and phosphorylates MAPK Kinases (Mek) 1 and 2. Meks are dual specificity kinases and phosphorylate tandem threonine and tyrosine residues in the TEY motif of p42MAPK/Erk2 and p44MAPK/Erkl, thereby activating them (van der Geer and Hunter, 1994; Campbell et al., 1998). The Erks phosphorylate targets with the consensus motif Ser/Thr-Pro (S/T-P), and, when activated, translocate to the nucleus where they act on transcription factors, such as Elk-1, Ets-1 and -2. They also activate effector kinases such as the p90 RSK Ser/Thr kinase to regulate protein synthesis (Treisman, 1996; Campbell et al., 1998). The conservation of the Ras-MAPK signalling cascade in several organisms substantiates its importance. As stated previously, genetic and biochemical analysis demonstrated that the Ras-MAPK pathway is conserved in the nematode C. elegans and the fruit fly D. melanogaster (Campbell et al., 1996). For instance, in C. elegans Let-23, an EGFR homologue, sem-5 (Grb2) and Let-60 (Ras) are components of a signal transduction pathway regulating vulval induction. Furthermore, most components of this cascade are functionally interchangeable between organisms (Ullrich and Schlessinger, 1990).

Ras-mediated signalling is important for cell cycle progression since expression of dominant negative Ras blocks growth factor-induced G 1/S phase progression. In addition, the Ras/MAPK pathway upregulates cyclin D 1, resulting in the activation of cyclin D-cdk4 and -cdk6 complexes at early Gl, thus allowing progression through to S phase (Bar-Sagi and Hall, 2000). Interestingly, recent evidence points to feedback regulation by the cell cycle that affects activation of the Ras signalling cascade (Stacey and Kazlauskas, 2002). The Ras pathway allows a striking diversity in signal transduction as Ras interacts with a wide range of receptors and recruits mutliple effectors to mediate signalling. However, the

13 functional effects of the Ras-MAPK cascade also depend on signal strength, which is modulated by receptor density and feedback regulation, as well as signal duration. For example, in PC12 cells, transient activation of this cascade results in cell proliferation whereas prolonged activation leads to cell differentiation (van der Geer and Hunter, 1994; Bar-Sagi and Hall, 2000).

1.2 Si:nallin: in breast cancer

Deregulation of signal transduction due to aberrant expression, activation or inactivation of signalling components is implicated in the development and progression of cancer. This is particularly critical due to the pleiotropic nature of cellular signalling, where a sole receptor may activate numerous signalling pathways, affecting cell motility, gene transcription, cytoskeletal changes and anti-apoptotic signals (Porter and Vaillancourt, 1998). Cross-talk between distinct signalling cascades adds an additional level of complexity where it may serve to diversify or amplify certain oncogenic pathways. In the non-pathological state, breast epithelial cells in females undergo cyclical changes involving key signalling processes such as proliferation, differentiation and apoptosis (Andrecheck and Muller, 2000). In addition, they are controlled by powerful mitogens including growth factors and steroid hormones, either due to systemic circulation or local production by autocrine and/or paracrine mechanisms. Thus, growth factor signal transduction in breast cancer cells and their interplay with steroid signalling are a significant component of this study.

Breast cancer is the leading international cause of cancer deaths in females (Jamison et al., 1999). Extensive research aimed at elucidating the molecular mechanisms underlying breast cancer cell proliferation and apoptosis resulted in the identification of key proteins that act at several levels in signal transduction to modulate biological effects. These include growth factors, steroid hormones and their respective receptors, in addition to numerous signalling intermediates that are also critical in signal transmission and amplification. In particular, the continual efforts for the past 20-30 years are rewarded with the development of breast cancer therapeutics that specifically target the ER (tamoxifen and SERMS) (Levenson and Jordan, 1999; McDonnell, 2000; Osborne et al., 2000; Plouffe, 2000), the

14 erbB2 RTK (herceptin) (Ross and Fletcher, 1999; Liu, 2000) and both the EGFR and erbB2 (Ye et al., 1999).

1.2.1 Growth factor receptor pathways 1.2.1.1 The ErbB family The erbB family is comprised of 4 RTKs, the EGFR/ErbBl, HER2/erbB2/neu, HER3/erbB3 and HER4/erbB4 (Daly, 1999; Biscardi et al., 2000). These form homodimers or heterodimers and regulate a variety of cellular functions such as proliferation, differentiation and apoptosis. The receptors consist of a glycosylated extracellular domain with cysteine-rich regions, a sole transmembrane domain and an intracellular unit harbouring the catalytic domain. Ligands for ErbB receptors are part of a superfamily of proteins with an EGF-like motif that spans 35-50 amino acids and contains six cysteine residues utilised in disulfide bond formation. The EGF-like motif is engaged in binding and activating the corresponding erbB receptor. Ligands in this superfamily engage in selective interactions with erbB receptors, for example EGF, amphiregulin (AR), and transforming growth factor -a. (TGF-a.) bind to the EGFR; betacellulin (BC), epiregulin (EPR) and heparin-binding EGF-like growth factor (HB-EGF) interact with the EGFR and erbB4; and the heregulin/neuregulin-1 (NRG-I) and -2 family associate with both erbB3 and erbB4, whereas NRG-3 only interacts with erbB4. However, NRG-2f3 also binds to the EGFR and erbB2 remains an orphan receptor with no known interacting ligands. Ligand-induced heterodimerisation of the erbB receptor family is more important for signal transmission, amplification and diversification than vertical signalling via receptor homodimerisation. Receptor heterodimerisation, first detected between the EGFR and erbB2, occurs in a hierarchial fashion where erbB2 is the preferred interacting partner. In tum, erbB2 favours binding to erbB3, which is essential for the latter since it lacks kinase activity and is thus dependent on transphosphorylation by other receptors for signalling. Moreover, the formation of erbB heterodimers occurs with directionality, for example, NRG-I is more effective than EGF in activating erbB 1-erbB4 heterodimers. Further, erbB l-erbB3 complexes are induced in response to NRG- I stimulation, whereas dimer formation is weak after EGF treatment (Daly, 1999).

15 Excessive stimulation by autocrine production of ligand, expression of active receptor variants or receptor overexpression in the erbB signalling network is implicated in a variety of cancers including lung, ovary, prostate and breast (Prenzel et al., 2000). In particular, the EGFR and erbB2 have important functions in carcinogenesis. Both receptors mediate activation of mitogenic signalling pathways as they interact with Grb2 and She, resulting in activation of the Ras pathway. In addition, EGFR and erbB2 bind and activate c-Src as well as PLC-y although their association with PI-3 kinase is weak. EGFR-overexpressing cells are transformed when cultured in the presence of EGF (Di Fiore et al., 1987a). Further, the EGFR is overexpressed in approximately 30% of breast cancers where its expression displays a negative correlation with ER status and confers worse disease-free survival (Sainsbury et al., 1987; Biscardi et al., 2000). There are implications that EGFR plays a role in breast cancer metastasis since receptor expression is enhanced in human mammary metastases relative to the primary tumour and EGFR overexpression enhances the invasive potential of breast cancer cells (Lichtner et al., 1995).

The EGFR is most homologous to erbB2 relative to remaining family members (Coussens et al., 1985). ErbB2 displays high ligand-independent basal kinase activity, and when overexpressed in fibroblasts results in cell transformation (Di Fiore et al., 1987b). Further, erbB2 overexpression in transgenic mouse mammary glands results in tumourigenicity and metastasis (Muller et al., 1988; Guy et al., 1992). ErbB2 expression in breast cancer cells increases and extends activation of the MAPK and JNK proteins. In addition, this receptor is overexpressed in 15-30% of breast cancers due to extensive gene amplification and is associated with decreased patient survival (Slamon et al., 1987; Slamon et al., 1989). Furthermore, activated erbB2, detected by antibodies against the tyrosine phosphorylated form, is associated with high erbB2 expression in breast cancers and confers a poor prognosis in node-positive tumours (Thor et al., 2000). ErbB2 is implicated in the initiation and early progression of breast tumours as overexpression occurs more frequently in in situ breast cancers (Biscardi et al., 2000; Stern, 2000). There is evidence to suggest co­ operation of the EGFR and erbB2 in carcinogenesis. In MMTV-transgenic mice overexpressing erbB2, EGFR levels are markedly upregulated (DiGiovanna et al., 1998). Furthermore, fibroblasts overexpressing both receptors display synergistic cell

16 transformation (Kokai et al., 1989). This synergy may be due to decreased receptor downregulation and stabilised ligand interactions observed with erbB2 overexpression, which results in diversified activation of signalling pathways and prolonged signal transmission (Daly, 1999; Stern, 2000). Therefore, significant advances have been made in uncovering the underlying mechanisms for the oncogenic properties of erbB2 which have enabled the development of pharmacological therapeutics. Most notably, a monoclonal erbB2 antibody, named Herceptin (trastuzumab), effectively treats patients with metastatic erbB2-overexpressing breast tumours (Cobleigh et al., 1999; Vogel et al., 2002). Finally, the observed synergy of the EGFR and erbB2 in tumorigenesis has resulted in the development of binary therapeutic antibodies, that effectively inhibit tumour growth by targeting both receptors (Ye et al., 1999).

1.2.1.2 The FGFR family FGFs constitute a family of at least 23 members that display 35-50% amino acid identity and initiate proliferation and differentiation in cells of epithelial, mesodermal and neuroectodermal origin. They are integral in the induction of angiogenesis, wound healing, cell growth, and migration. FGFs signal through transmembrane tyrosine kinases, FGFRs, of which four members are identified. Several isoforms of secreted or cell-bound FGFR 1-4 have been characterised and these originate by alternative splicing, alternative start sites and exon skipping (Mason, 1994; Burke et al., 1998; Szebenyi and Fallon, 1999; Boilly et al., 2000; Yamashita et al., 2000). FGF signalling is mediated through complex formation between heparin sulfate proteoglycans, the growth factor and its corresponding FGFR (Spivak-Kroizman et al., 1994).

Several FGFs and their receptors are expressed in the normal breast and in breast cancer. For example, FGF-1 (aFGF) is expressed in normal breast tissue and in fibroadenomas, the latter being benign tumours consisting of both glandular and fibrous tissue. It is mainly localised in breast epithelial cells and in stromal fibroblasts, but not in the myoepithelium (La Rosa et al., 2001). In addition, FGF-8 is expressed in breast cancer tissues and cell lines (Tanaka et al., 1995; Payson et al., 1996; Tanaka et al., 1998a). Overexpression of FGFRs is also evident in breast cancer. In particular, the FGFR-1 and -4 genes are amplified in

17 approximately 20% and 30% of breast cancers, respectively (Dickson et al., 2000). Further, FGFR-4 is overexpressed in several breast cancer cell lines (McLeskey et al., 1994). In addition to tumour cells, FGFR-1 is expressed in epithelial cells of the normal breast or benign tumours and is also weakly detected in the stroma (Blanckaert et al., 1998).

FGFs and their receptors are implicated in the development and progression of several malignancies. For instance, FGF-9 overexpression results in transformation of Balb/c 3T3 cells by autocrine stimulation. Transformed cells also form tumours in nude mice (Matsumoto-Yoshitomi et al., 1997). Translocations of FGFR-1 and -3 are associated with mutliple myeloma and leukaemia (Chesi et al., 1997; Xiao et al., 1998). MCF-7 breast cancer cells overexpressing FGF-1 display phenotypic changes such as hormone­ independence, resistance to antiestrogen treatment and metastatic growth in nude mice (McLeskey et al., 1993; Zhang et al., 1997). FGF-8b overexpression results in an increased rate of anchorage-independent proliferation and enhanced invasion of MCF-7 cells. Also, FGF-8b-overexpressing MCF-7 cells exhibit faster growth in nude mice relative to controls (Ruohola et al., 2001). aFGF and bFGF induce membrane ruffling, which is characteristic of cell motility, in some breast cancer cell lines but not in normal cells. This effect is mediated by FGFR-4, but not by FGFRl-3 (Johnston et al., 1995). A mutation in FGFR-4 which substitutes glycine with arginine at codon 388 within the receptor transmembrane domain was detected in the MDA-MB-453 cell line and subsequently estimated to occur in approximately 50% of the population. The polymorphism is associated with decreased survival of homo- and heterozygous breast cancer patients and with metastatic tumours in individuals with colon cancer. Breast cancer cells overexpressing the Arg388 FGFR-4 allele demonstrate enhanced migration relative to the Gly388 isotype (Bange et al., 2002).

FGFs can confer both stimulatory and inhibitory signals for cell growth, which is likely to be cell-type dependent. For example, in MDA-MB-134 breast cancer cells, treatment with FGF-1 or -2 results in growth inhibition under both anchorage-dependent and anchorage­ independent conditions (McLeskey et al., 1994). Also, in the MCF-7 breast cancer cell line, FGF-2 stimulation inhibits cell growth and upregulates the cyclin-dependent kinase

18 inhibitor p21 and the transcription factor STAT-1. These effects are not observed in T-47D cells, which are not growth-inhibited by FGF treatment (Johnson et al., 1998).

1.2.1.3 The IR family Insulin is an anabolic hormone that regulates carbohydrate, fat and protein metabolism. These effects are mediated by IRs expressed in liver, muscle and adipose tissue. Insulin is critical in the pathogenesis of diabetes and is also implicated in cancer progression. IGFs, on the other hand, mediate their effects via interaction with the IGF-I and possibly with the mannose-6-phosphate/IGF-II receptors. Insulin binds with high affinity to the IR, which displays a lower affinity for IGFs. Similarly the IGF-IR interacts with a higher affinity with IGFs relative to insulin, whereas the IGF-IIR binds IGFs and does not interact with insulin (Adamo et al., 1992). IGF-I and -II, but not insulin, circulate complexed to IGFBPs which are localised in the bloodstream and in tissues. The IGFBP family consists of six members that facilitate the interaction of IGFs with their high affinity receptors by transporting IGFs to their target sites, increasing the half-life of serum IGFs and modulating their actions (Adamo et al., 1992; Perks and Holly, 2000).

The IR is transcribed as a proreceptor molecule and is further modified at the post­ translational level by the formation of disulfide bonds, proteolytic cleavage, N- and 0- linked glycosylation and acylation. The receptor is then transported to the plasma membrane for ligand binding, activation and signal transmission. Ligand-binding enhances the rate of IR endocytosis, where the receptor is either recycled back to the cell surface or targeted for degradation. Different classes of IR mutations that interfere with receptor synthesis and processing are associated with insulin resistance and result in genetic abnormalities such as leprechaunism and the Rabson-Mendenhall syndrome (Taylor et al., 1994). Cloning of the IR in 1985 allowed its classification in the tyrosine kinase family of receptors (Langlois and Olefsky, 1994 ), where it displays significant structural homology to the IGF-IR. These receptors share the highest degree of peptide homology (84%) in the tyrosine kinase domain, whereas the overall homology of the a subunit is only 44-60%. However, the IR possesses a completely distinct C-terminal region that is not found in the IGF-IR. This region is implicated in regulating glucose metabolism and may account for

19 differences observed in mitogenic potency. Also, distinct ligand-binding mechanisms possibly reflected by the relatively poor extracellular domain homology may also contribute to differences in signalling capacity (Adamo et al., 1992; Siddle et al., 2001). The IR mediates diverse biological effects such as the translocation of GLUT4 glucose transporters to the plasma membrane, glycogen synthesis, transcription of specific target genes and mi to genesis. Traditionally, the insulin/lR system was thought to be predominantly involved in cellular metabolic effects whereas IGF-1 and its receptor were implicated to mediate mitogenic effects. However, each receptor is capable of regulating both glucose metabolism in addition to cell growth and development. At high concentrations, insulin and IGF-1 have a weak affinity for the heterologous receptor and share similar effects. In addition, studies have demonstrated that low nanomolar doses of insulin or IGF-1 stimulate glucose uptake in muscle cells and mitogenesis in an osteosarcoma cell line (Adamo et al., 1992). Furthermore, the IR has oncogenic potential since an avian retrovirus encoding a constitutively active IR results in transformation of chicken embryo fibroblasts and the formation of colonies in soft agar. The transformed cells formed sarcomas in chickens (lsh-Shalom et al., 1997). Another study revealed that overexpressed IGF-IR and IR stimulate glucose transport with similar efficiency in mouse fibroblasts (Lammers et al., 1989).

The IR has several substrates including IRS 1-4, Gab 1, She and p62Dok. Most insulin actions are mediated by recruitment of effectors through IRSs, which act as docking proteins. IRS proteins consist of an N-terminal PH domain which allows membrane­ targeting, and a PTB domain C-terminal to the PH region. IRS proteins bind via their PTB domain to pTyr-960 of the IR ~ subunit at the NPXpY motif and mediate many of the IR's biological effects (Virkamaki et al., 1999). Interestingly, both the IR and IGF-IR mediate their effects through IRS-1 and She (Kovacina and Roth, 1993; Skolnik et al., 1993b; White and Kahn, 1994). IRS-1 recruits multiple effectors, including the adaptor proteins Grb2, Nck and Crk, phosphatases such as SHP-1 and -2, and Fyn, which allows insulin­ mediated caveolin phosphorylation (Cheatham and Kahn, 1995; Ogawa et al., 1998; White, 1998; Virkamaki et al., 1999). Studies have suggested that localisation of IRS proteins at the cytoskeleton may prevent them from being degraded and facilitate interactions with the

20 IR, thus providing a mechanism for regulation of insulin signalling and specificity (Whitehead et al., 2000). The interaction of IRS-1 with Grb2 leads to the translocation of Sos to the membrane and the subsequent activation of the MAPK cascade. Insulin activation of MAPK is involved in cell differentiation, protein synthesis and gene regulation (Skolnik et al., 1993a; Cheatham and Kahn, 1995; Ogawa et al., 1998; Virkamaki et al., 1999). In addition, IRS-1 binds the p85 regulatory subunit of PI-3 kinase. PI-3 kinase functions as a lipid kinase and phosphorylates membrane phosphoinositides (Pis), leading to the production of PI intermediates (such as PIP3) in the cytosolic aspect of the cell membrane. These intermediates then recruit downstream effectors to the membrane, such as Serffhr protein kinases (Akt/PKB) and PKC rjA isoforms. This allows for subsequent association with upstream regulators such as PDKl, and results in their activation. Activation of a wide range of effectors and biological effects are mediated by the PI-3 kinase pathway, including insulin or IGF-I-induced membrane ruffling, antilipolysis, activation of glycogen synthase, phosphorylation of the serine kinase Akt, muscle cell differentiation as well as stimulation of protein and DNA synthesis (Kotani et al., 1994; Burgering and Coffer, 1995; Coffer and Kaliman, 1995; Martin et al., 1996; Ogawa et al., 1998; Virkamaki et al., 1999; Whitehead et al., 2000). Further, the MAPK cascade is weakly activated by insulin since the Grb2 binding site in IRS-1, pTyr-895, is weakly phosphorylated and requires IR overexpression. By contrast, phosphorylation of IRS-1 sites that interact with PI-3 kinase, pTyr-608 and -939 are achieved with low IR levels (Ogawa et al., 1998; White, 1998).

The IGFs are potent mitogens for breast cancer cells and they function via an autocrine or paracrine loop through the IGF-IR (Surmacz, 2000; Wood and Douglas, 2000; Zhang and Yee, 2000). By contrast, insulin is not locally produced in the breast, although it is also mitogenic in breast cancer cells (Chappell et al., 2001). IGF-I is expressed in the stroma of benign and malignant breast tumours, while IGF-11 is expressed in both the stroma and the tumour (Yee et al., 1988; Cullen et al., 1992). In addition, both the IR and IGF-IR are overexpressed in breast tumours and cell lines (Papa et al., 1990; Milazzo et al., 1992; Papa et al., 1993; Zhang and Yee, 2000), and are mitogenic for breast cancer cells (Cullen et al., 1990; Milazzo et al., 1992). IR overexpression in mouse fibroblasts or in normal mammary

21 epithelial cells results in ligand-dependent cell transformation (Giorgino et al., 1991; Frittitta et al., 1995). Similarly, IGF-IR overexpression in fibroblasts induces ligand­ dependent neoplastic transformation which requires the receptor C-terminus (Kaleko et al., 1990; Surmacz, 2000), and is correlated with enhanced receptor kinase activity (Resnik et al., 1998). In addition, expression of both receptors in breast cancer correlates with patient prognosis. For instance, in a cohort comprising 584 patients with node-negative breast cancer, elevated IR levels are associated with decreased disease-free survival (Mathieu et al., 1997). Also, IGF-IR overexpression causes radioresistance and is associated with recurrence for breast cancer patients with early relapses following lumpectomy and radiation therapy (Turner et al., 1997).

Interestingly, the IR is activated by IGF-11 in primary breast cancers and breast cancer cell lines, but not in normal mammary cells (Sciacca et al., 1999). There are two IR isoforms that differ by 12 amino acids due to alternative splicing of exon 11, the exon-lacking IR-A (Exll-) and IR-B (Exll+). These isoforms display small differences in insulin binding and subsequent signal transmission (Yamaguchi et al., 1991). IR-A but not IR-B interacts with IGF-11 with an affinity similar to that of insulin, and IGF-11 binds to and activates IR-A and IGF-IR with a similar affinity. IR-A is predominantly expressed in breast tumours and breast cancer cell lines. It is also the main expressed isoform in fetal cells and is detected at higher levels in colon cancers relative to normal tissue (Frasca et al., 1999). Moreover, IGF-11 activation of the IR is implicated via an autocrine/paracrine loop that is mediated through IR-A in breast cancer cells (Sciacca et al., 1999).

An additional level of complexity in the function and mechanism of insulin/IGF-1 signalling in breast cancer is evident with the discovery of hybrid receptors (Hybrid-Rs), which result from co-expression of both IRs and IGF-IRs. Each hybrid-R consists of one IR a and 13 subunit hemicomplex linked to an IGF-IR hemicomplex. Hybrid-Rs function in a similar manner to IGF-IRs, since they bind IGF-1 with a similar affinity but display a lower affinity for insulin (Moxham et al., 1989; Soos et al., 1990; Soos et al., 1993). The hybrid­ R content in breast tumours positively correlates with IR and IGF-IR levels, and may be generated randomly by IR and IGF-lR assembly. In addition, hybrid-Rs are phosphorylated

22 in response to IGF-1, but not insulin stimulation. Moreover, in breast cancer cells predominantly expressing hybrid-Rs, IGF-1-induced mitogenesis is blocked by an antibody directed against hybrid-Rs, indicating that they also participate in mediating IGF-1 signalling (Pandini et al., 1999).

1.2.2 SH2 domain-containing signalling intermediates The last decade has seen the definition of key signalling pathways downstream of RTKs in terms of their components and the protein-protein interactions that facilitate signal transduction. Given the strong evidence linking signalling by certain families of RTKs to the progression of breast cancer, it is not surprising that the expression profile of key SH2- containing downstream signalling intermediates in this disease has also come under scrutiny, particularly since some exhibit transforming potential or amplify mitogenic signalling pathways when overexpressed.

1.2.2.1 c-Src The protooncogene c-Src (Figure 1.3) is the cellular homologue of the transforming v-Src gene of Rous sarcoma virus (Bolen, 1993). Activation of c-Src is required for mitogenic signalling by the EGFR and PDGFR and is also a critical component of integrin-mediated signalling via focal adhesion kinase (Parsons and Parsons, 1997). The enzyme is maintained in an inactive conformation by intramolecular interactions involving the SH2 and SH3 domains, the former binding phosphorylated Tyr-530 at the C-terminus (Mayer, 1997). Evidence implicating a role for c-Src in breast cancer emerged when two studies. demonstrated enhanced c-Src kinase activity in breast tumours relative to normal breast tissue (Jacobs and Rubsamen, 1983; Rosen et al., 1986). Furthermore, increased cytosolic tyrosine kinase activity, largely attributable to c-Src, correlated with early systemic relapse in this disease (Hennipman et al., 1989; Ottenhoff-Kalff et al., 1992). Increases in both the expression and specific activity of the enzyme in breast cancers have been observed (Rosen et al., 1986; Ottenhoff-Kalff et al., 1992; Verbeek et al., 1996). One mechanism for activation of c-Src in breast cancers is through SH2 domain-mediated association with the EGFR and erbB2, which both recruit this enzyme (Luttrell et al., 1994; Muthuswamy and Muller, 1995), and are commonly overexpressed in this disease (Sainsbury et al., 1987;

23 Gullick, 1990). High levels of Src kinase activity were detected in four breast cancer cell lines compared to normal human foreskin fibroblasts and a normal breast epithelial cell line (Egan et al., 1999). This correlated with reduced phosphorylation of Tyr-530 and increased membrane-associated protein tyrosine phosphatase (PTPase) activity which preferentially dephosphorylated a synthetic Src family C-terminal phosphopeptide. This activity was subsequently recognised to be primarily caused by PTP-1B (Bjorge et al., 2000).

1.2.2.2 PLC-y1 PLC-yl tyrosine phosphorylation follows its interaction with activated EGF and erbB2 receptors in vivo (Margolis et al., 1989; Peles et al., 1991), and stimulates its enzymatic activity (Kim et al., 1991). Arteaga et al. demonstrated by both immunohistochemistry and immunoblotting that PLC-yl is expressed at higher levels in breast tumours relative to normal tissue (Arteaga et al., 1991). Furthermore, PLC-yl is tyrosine phosphorylated in the majority of the overexpressing tumours, and this correlates with the presence of high levels of the EGFR or erbB2. PLC-y 1 expression is also associated with estrogen receptor negativity and a higher histological grade.

1.2.2.3 Grb2 As mentioned previously, the adaptor Grb2 consists of a SH2 domain flanked by two SH3 domains (Figure 1.3) and serves to link RTKs such as the EGFR to the Ras pathway (Lowenstein et al., 1992; Pawson, 1995). In a detailed study of Grb2 expression among breast cancer cell lines, marked overexpression of the Grb2 protein was observed in MCF- 7, MDA-MB-361 and -453 cells relative to normal breast epithelial cells, and this was accompanied by a small amplification of the Grb2 gene locus (Daly et al., 1994). Also, increased Grb2 expression enhanced complex formation with Sos. This analysis was extended to primary breast cancers, which demonstrated that approximately 50% of those tested exhibited a 2-fold or greater overexpression of Grb2 mRNA (Yip et al., 2000). Verbeek et al. (Verbeek et al., 1997) found that the Grb2 protein was overexpressed in all primary breast cancers examined compared to normal tissue, which suggests that a further upregulation of Grb2 expression may also occur at the translational level. Interestingly, in the Yip et al. study breast cancers exhibiting low levels of the EGFR expressed

24 Class I

c-Src

~ SH3 ISH2 I KINASE' I

PLC-y1

PLC SH2 SH2 SH3 PLC

Class II

Grb2

,SH31SH21SH3~

Figure 1.3 Structure of signalling intennediates exhibiting aberrant expression in human breast cancer. Schematic representation of the structures of c-Src and PLC-yl (Class I SH2 domain-containing proteins) and Grb2 (a Class II protein). PLC-yl has two regions (designated PLC in the figure) which constitute its catalytic domain, and the second PH domain is interrupted by SH2 and SH3 modules. The different proteins and their domains are not drawn to scale. significantly higher Grb2 mRNA than high EGFR-expressing breast cancers, suggesting selection for a mechanism that amplifies signalling downstream from the receptor. Although Grb2 is not transforming when overexpressed (Suen et al., 1993), early studies demonstrated that Grb2 overexpression could amplify activation of Ras or MAP kinase in response to particular growth factors (Gale et al., 1993; Skolnik et al., 1993a; Suen et al., 1993; Cheng et al., 1998). More recently, these analyses have been extended to mouse mammary tumorigenesis in response to expression of polyomavirus middle T (PyV MT) antigen, which couples to the Ras pathway via recruitment of a Shc/Grb2/Sos complex. Mice transgenic for PyV MT and heterozygous for Grb2 gene disruption have a delayed onset of mammary tumours induced by PyV MT relative to wildtype animals, indicating that the Grb2 gene dosage is limiting for tumorigenesis (Cheng et al., 1998). Also, co­ expression of Grb2 and a mutant PyV MT antigen in the mammary epithelium accelerates tumorigenesis in mice. This is correlated with increased activation of erbB2, erbB3 and MAPK (Rauh et al., 1999). It is therefore likely that upregulation of Grb2 expression functions in breast cancer progression to amplify signalling via the Ras pathway downstream of particular tyrosine kinases.

1.2.3 Signalling by estrogens The link between estrogens and breast cancer was revealed over a century ago when Beatson discovered that growth of mammary neoplasms is controlled by an endocrine factor secreted by the ovaries (Yee and Lee, 2000). Estrogens, along with testosterone, glucocorticoids, cortisol and progesterone are examples of mammalian steroid hormones. The physiological estrogens 17~-estradiol, estrone and estriol are produced by the ovaries, the placenta and the corpus luteum. 17~-estradiol is the most potent and predominant estrogen. It circulates in the bloodstream in equilibrium with estrone, which is subsequently metabolised to estriol in the liver (Ganong, 1995). Estrogens exert pleiotropic effects that regulate systemic physiology including that of the reproductive, cardiovascular, skeletal and central nervous systems. Subsequent research pointed to the existence of an estrogen receptor (ER) or 'estrophile', the cloning of which was achieved in 1985 (Walter et al., 1985). This revealed that this receptor, subsequently denoted as ERa, belonged to a family of nuclear steroid hormone receptors. This is a subset of a large superfamily of ligand-

26 inducible transcription factors which, in addition to steroid receptors, also includes thyroid, retinoid, vitamin D as well as orphan receptors, the latter having unknown ligands (Mangelsdorf et al., 1995).

ERa consists of an amino-terminal NB region, a central DNA-binding domain (DBD) with 2 zinc fingers (region C), a hinge section (region D) and a C-terminal ligand-binding domain (LBD) (region E/F). In addition, ERa possesses two transactivation domains: a constitutive activation function-I (AF-1) domain located in the amino-terminal NB region, and a ligand-dependent AF-2 domain in the C-terminal E region (Osborne et al., 2000; Peterson, 2000; Yee and Lee, 2000). A second form of the ER was cloned a decade following ERa, and was designated ERP (Kuiper et al., 1996; Mosselman et al., 1996). ERP lacks a portion of the amino-terminus and does not contain a functional AF-1 domain. It is also missing a large section of the C-terminus. The DBD and LBD exhibit the highest homology (95% and 60%, respectively) between ER subtypes, whereas the amino­ terminus is less conserved. However, both ERs possess the AF-2 domain, which contains a highly conserved amphipathic a-helix (H-12) that is required for ligand-dependent transcription and association with the steroid receptor co-activator (SRC) family. Studies so far indicate that ERa and -P have distinct physiological functions. Their expression patterns differ, with ERa predominating in the uterus and mammary gland, whereas ERP is mainly detected in the ovary, prostate, testis, hypothalamus and thymus as well as in the spleen, lung, kidney, and bone (Couse et al., 1997; Gustafsson, 2000; Peterson, 2000). Moreover, while ERa is the principal mediator of estrogen action in breast cancer, the function of ERP in breast neoplasms to date is unknown (Yee and Lee, 2000).

In breast cancer, the ER is utilised as a prognostic tool for selection of patients suitable for endocrine therapy, since patients with ER-positive breast tumours exhibit a better response to this treatment (Hundred, 2001). The ER is detected in 50-80% of breast tumours, and correlates with favourable prognostic markers such as a lower grade and better tumour differentiation. By contrast, ER-negativity is associated with a higher rate of tumour proliferation and poor differentiation. Furthermore, tumours which lack the ER are also

27 associated with overexpression of growth factor signalling components, such as erbB2 (Osborne, 1998).

Since it is a lipid-soluble ligand, estradiol is able to cross the plasma membrane and induce its effects by binding to and activating its receptor. In the absence of ligand, ERa is inactive. Ligand binding induces a conformational change in the receptor which allows homodimerisation and subsequent binding to estrogen response elements (EREs) within regulatory regions of gene promoters, thus initiating the transcription of target estrogen­ responsive genes. However, transcriptional regulation of the ER is more complex. ERs are also bound to co-factor proteins which regulate their transcriptional activity such as SRC-1, glucocorticoid receptor interacting protein 1 (GRIPl) and activator of thyroid receptor (ACTR). Co-factors also include co-repressors which generally have histone deacetylase activity that allows the DNA to tightly coil around the core histones, thereby inhibiting transcription. Ligand stimulation releases the co-repressors and allows co-activators to bind. These transfer acetyl groups to histones, thus loosening their association with DNA and allowing transcription to proceed. Depending on the cell-type and the promoter context, the receptor then either initiates or suppresses the expression of downstream genes (Osborne et al., 2000; Peterson, 2000).

The classical mode of estrogen action is via genomic mechanisms, where ligand binding­ induced ER activation results in association with EREs in target genes. The ER can also act indirectly through association with additional transcription factors that bind to distinct consensus sequences such as c-fos and c-jun acting on AP-1 transcription sites (Yee and Lee, 2000; Hall et al., 2001). During progression from Gl to S phase, estradiol upregulates cyclin Dl leading to activation of cyclin Dl-cdk4/6. In addition, estradiol induces activation of higher molecular weight cyclin E-cdk2 complexes through dissociation of the inhibitor p21 WAFIICIPJ. Through early/mid G 1, the growth inhibitory retinoblastoma protein (pRb) is complexed to E2F transcription factors. Activation of cyclin E-cdk2 results in inactivation of pRb, thus allowing cells to progress through the R-point transition and resulting in the release of E2F factors which mediate the transcription of genes required for

28 S-phase (Altucci et al., 1996; Planas-Silva and Weinberg, 1997a; Planas-Silva and Weinberg, 1997b; Prall et al., 1997).

1.2.4 Cross-talk between estrogen and insulin/IGF signalling The notion of cross-talk between steroid and cell surface receptors was initially demonstrated with the finding that the neurotransmitter dopamine is able to activate a steroid receptor, namely the progesterone receptor, independent of ligand stimulation (Power et al., 1991 ). Subsequently, multiple laboratories demonstrated that the ER is activated by several growth factor signalling cascades activated by EGF, IGFs, heregulin and the second messenger cyclic adenosine monophosphate ( cAMP) (Aronica and Katzenellenbogen, 1993; Ignar-Trowbridge et al., 1993; Ma et al., 1994; Pietras et al., 1995). In MCF-7 breast cancer cells, IGF-1 modulates ERa transcription and enhances receptor activity, thereby upregulating estrogen-regulated genes. The effects of IGF-1 on ERa expression and activity are blocked by PKA and PI-3 kinase inhibitors, suggesting that they may play a role in the IGF-Uestrogen cross-talk (Stoica et al., 2000). IGF-1 and insulin synergise with estradiol in the proliferation of breast cancer cells (van der Burg et al., 1988; Stewart et al., 1990; Lai et al., 2001). The synergistic effects of insulin/IGF-1 and estradiol involve bidirectional interactions between these signalling cascades, where estradiol acts through the ER via genomic and non-genomic mechanisms to modulate the expression of several growth factors and their receptors. Although the mechanisms by which they converge are not entirely clear, studies suggest that each signalling cascade may depend on the other for eliciting optimal biological effects (Yee et al., 1988; Lee et al., 1999; Kahlert et al., 2000; Yee and Lee, 2000; Hall et al., 2001 ). Cross-talk between steroids and growth factors is discussed in more detail in sections 6.1 and 7.5.

The interplay between estrogen and insulin/IGF-1 signalling is critical not only in regulating normal breast development but also in the progression of breast cancer. Case control and epidemiological studies have documented an increase in several cancers in hyperinsulinaemic individuals including pancreatic, colorectal, endometrial, liver and breast cancer (La Vecchia et al., 1994; Weiderpass et al., 1997; Wideroff et al., 1997). Hyperinsulinaemia is proposed as a possible unifying etiologic factor in the development of

29 breast cancer in obese and diabetic patients. Hyperinsulinaemia is also associated with increased IGFs in breast tissue. Therefore, increased levels of insulin/lGFs may act synergistically with increased estrogenic activity in these patients (Kaaks, 1996; Stoll, 1996).

1.3 The Grb7 family

The transmission of signalling cascades from phosphorylated receptors is dependent on the interaction and activation of downstream effectors which include SH2 domain-containing adaptor proteins, enzymes and scaffolding proteins that play critical roles in signal relay and diversification. On this note, Schlessinger and colleagues developed cloning of receptor targets (CORT) screening, where they utilised the phosphorylated EGFR C-terminus to identify interacting proteins by screening expression libraries. These were denoted as growth factor receptor-bound (Grb) proteins and numbered consecutively in the order of identification (Daly, 2001). The first associating protein identified using the CORT method was the p85 subunit of PI-3 kinase, and was designated Grbl. The second protein was Grb2, while Grb3-6 correspond to the identified proteins v-crk, Nck, Fyn, and PLC-yl respectively. In 1992, a novel SH2 domain-containing protein was isolated by CORT screening of a T7 polymerase-based expression library and termed Grb7 (Margolis et al., 1992). This protein displays significant homology to two additional Orbs: GrblO, identified in 1995 (Ooi et al., 1995), and the most recently identified Grb14 in 1996 (Daly et al., 1996). Thus these three Orbs were classified together based on significant structural and sequence homology, and designated as the Grb7 family of SH2 domain-containing proteins.

1.3.1 Structure The Grb7 family overall molecular architecture (Figure 1.4) consists of an N-terminus harbouring a proline-rich motif, and a central region termed as the Grb and Mig region (denoted GM) due to its homology with the C. elegans F10E9.6/Mig 10 protein (Daly, 1998; Han et al., 2001). The Grb7 family proteins also contain a Ras-associating-like (RA­ like) domain and PH domain within the GM region. The far C-terminal end harbours

30 p Proline-rich regions PS/AIPNPFPEL

GM

SH2 Grb14

p p 67 Grb7

--- ' ,-i p 74 mGrb10a ---- ~~~~ ··""

0 100 200 300 400 500 600 700 Amino acids

Figure 1.4 Structure of the Grb7 family proteins. Functional signalling modules and conserved motifs are displayed in an alignment of Grb14, Grb7 and mGrblO a. The yellow bar denotes the conserved proline-rich motif, shown in the key above, Numbers indicate the percentage of amino acid identity relative to Grbl4. Numbers on the scale represent the size in amino acids. mGrblO a contains a distinct N-terminal insert that is not found in the other proteins. Note that the Grb 14 sequence is more similar to Grb 10 than Grb7. an SH2 module adjacent to the 'hetween fH and SH2' (BPS) domain, so termed due to its location where it is sandwiched between the PH and SH2 modules (He et al., 1998; Kasus­ Jacobi et al., 1998; Wojcik et al., 1999).

The N-terminus exhibits the least homology between the Grb7 family proteins. However, it harbours a highly conserved proline-rich motif PS/AIPNPFPEL flanked by two clusters of basic residues. This motif may provide binding sites for SH3 domains, for example, the SH3 domain of c-Abl, but not of PI-3 kinase, Grb2 or Fyn, binds the proline-rich region of GrblO in vitro (Frantz et al., 1997), although this has not yet been demonstrated in vivo. Further, the N-terminus of Grb14 is less proline-rich relative to the other family members, with an 11 % proline content compared with 15% and 23% for GrblO and -7, respectively (Daly et al., 1996). The central GM region spans about 320 amino acids and displays approximately 50% identity among Grb7 proteins and nearly 30% identity between Grb7 family members and Mig 10. The central PH domain contained within the GM region is more conserved between GrblO and -14 (61% identity relative to Grb14) than with Grb7, which has 56% identity relative to Grb14 (Daly et al., 1996).

The RA domain is a conserved region present in many RasGTP effectors such as Ral GDS and AF6 (Ponting and Benjamin, 1996). The RA three dimensional structure was determined in the Ral-GEF protein and is similar to the Ras-binding region of Raf-1 (Nassar et al., 1995). The RA-like domain was identified in the Grb7 family using sequence analysis, and suggests interactions between the Grb7 family and GTPases (Wojcik et al., 1999). In vitro studies demonstrated that Grb7 interacts with Rndl, a member of the Rho family of GTPases, although this was mediated between the Grb7 SH2 domain and the switch II loop of Rndl which is critical for guanine nucleotide exchange (Vayssiere et al., 2000). Therefore, a precise role for the Grb7 family RA-like domain is yet to be determined. The homology of the Grb7 family GM region to the C. elegans protein Mig 10 is of particular interest since the latter is implicated in long-range embryonic cell migration, where Mig 10 mutants display incomplete neuronal cell migration and shortened posterior excretory canals. However, the remaining structure of Mig 10 differs from the Grb7 family since the C. elegans protein lacks an SH2 domain and

32 contains distinct C-terminal proline-rich regions (Daly, 1998). Interestingly, Grb7 is implicated in cell migration due to its interaction with focal adhesion kinase (FAK) (Han and Guan, 1999; Han et al., 2000), although there is no evidence that GrblO and -14 fulfil this function. Furthermore, to date the Grb7 proteins constitute the only known protein family to harbour the BPS domain, a module of nearly 50 residues which shares approximately 70% identity between Grb7 and -10 relative to Grb14. In vitro studies suggest that the BPS may act as a separate functional module since it plays a critical role in GrblO and -14 inhibition of IR activation (Stein et al., 2001; Bereziat et al., 2002). The SH2 domain is also a highly conserved module and it exhibits an approximate 70% identity amongst Grb7 family proteins. This domain plays an essential role in mediating interactions with several growth factor receptors, although its contribution to receptor binding differs between Grb7 proteins (He et al., 1998; Kasus-Jacobi et al., 1998; Hemming et al., 2001).

1.3.2 Expression The Grb7 family of proteins are broadly expressed and display overlapping yet distinct expression patterns in human tissues. All proteins exhibit a relatively high level of expression in the pancreas (Daly, 1998). In addition, Grb7 is expressed in the kidney, prostate, small intestine, placenta at higher levels than the lung, liver, testis and colon (Frantz et al., 1997). GrblO is detected in skeletal muscle, heart and brain at higher levels than the placenta, lung, liver, kidney, spleen, prostate, testis, ovary, small intestine and colon (Liu and Roth, 1995; O'Neill et al., 1996; Frantz et al., 1997). Finally, Grb14 is expressed at relatively high levels in the gonads, liver and kidney, moderately in heart and skeletal muscle. It is also detected at lower levels in the placenta, brain, small intestine and colon (Daly et al., 1996). Note that GrblO and -14, in particular, are expressed in insulin­ responsive tissues.

1.3.3 Isoforms cDNA cloning, RT-PCR and hybridisation studies indicated that there are naturally occurring variants of Grb7 and -10, although there are no reports of identified Grb14 isoforms. According to a unified nomenclature system adopted after the initial

33 identification of Grb 10 isoforms, the origin species is indicated as a prefix and the individual splice variants denoted by greek letters. This allows for the possible identification of the same variant in distinct species, which would be denoted as the same isoform by an identical Greek letter. The proposed nomenclature is described by Andre Nantel on the Grb7 family website (http://www.bri.nrc.ca/thomasweb/grb7.htm) (Daly, 1998). According to this system, the originally identified members of the Grb7 family are thus designated as mGrb7 (Margolis et al., 1992), mGrblO a (Ooi et al., 1995) and hGrb14 (Daly et al., 1996). Six isoforms have been identified for GrblO (Figure I.SA) (mGrblO a and o; hGrblO ~. y, E and~) that possess intact BPS.SH2 domains (Liu and Roth, 1995; Ooi et al., 1995; O'Neill et al., 1996; Dong et al., 1997; Frantz et al., 1997; Laviola et al.,

1997; Han et al., 2001). hGrblO ~. E and~ are distinguished by N-terminal extensions, and hGrblO ~ also contains a PH domain deletion of 46 amino acids. hGrblO y lacks an N­ terminal extension but has a complete PH domain. mGrb 10 a and o are characterised by related inserts in the N-terminal region (Laviola et al., 1997; Daly, 1998). By contrast, only a sole isoform was isolated for Grb7 from an esophageal carcinoma cell line. This isoform, termed Grb7V, lacks 88 bp encoding for the SH2 domain (Figure I.SB). The resultant frameshift leads to substitution of the SH2 domain with a hydrophobic sequence (Tanaka et al., 1998b ). It is also important to note that the Grb7 and -10 splice variants are presumed to result from alternative splicing (Frantz et al., 1997), although studies have not been undertaken to date in order to determine the exact mechanism of generating these isoforms. Also, although GrblO isoforms are generally expressed in insulin target tissues, they exhibit differences in their expression profiles within certain tissues. For example, hGrb 10 ~. which lacks a complete PH domain, is more abundantly expressed in skeletal muscle and fat than the PH domain-containing hGrblO ~. However, in human breast cancer cell lines and in the liver hGrblO ~ expression predominates. This suggests that these isoforms perform distinct functions in different tissues, although they both translocate from the cytosol to the membrane upon insulin stimulation (Liu and Roth, 1995; Dong et al., 1997).

34 p Praline-rich regions A PS/AIPNPFPEL mGrb10 a p SH2 hGrb10 ~ pp hGrb10 y p p mGrb10 b p

hGrb10 £ ~ pp hGrb10 s - pp

B

Grb7 PH BPS SH2

Grb7V

Figure 1.5 Grb7 family isofonns. A. The structure of the six Grb 10 variants is shown. mGrblO a and & contain distinct amino-terminal inserts that are not present in the human isoforms. hGrblO p and s are identical except for a 46 residue deletion in the PH domain of the fonner. Both variants also contain a N-terminal extension which is not found in hGrblO 'Y, and is distinct from amino terminus of hGrblO £. B. The structure of the Grb7 splice variants is displayed. Unlike the full-length Grb7 protein, the truncated Grb7V completely lacks an SH2 domain and contains a short C-terminal hydrophobic sequence with an undefined function. 1.3.4 RTK-dependent interactions As expected from their structure which encompasses various functional modules, the Grb7 family display an extensive array of protein-protein interactions, particularly Grb7 and -10, whereas Grb14 retains a more restricted binding profile (Han et al., 2001). Although all three Grb7 proteins were isolated by CORT screening using the EGFR C-terminal tail as a probe, in viva interactions with the EGFR have not been demonstrated (Margolis et al., 1992; Ooi et al., 1995; Daly et al., 1996). Interestingly, all three Grb7 family proteins interact directly with the IR in viva (Liu and Roth, 1995; Frantz et al., 1997; Kasus-Jacobi et al., 1998; Kasus-Jacobi et al., 2000; Hemming et al., 2001). In addition, Grb7 interacts in viva with the B-PDGFR (Yokote et al., 1996), erbB2-4 (Stein et al., 1994; Fiddes et al., 1998) Ret (Pandey et al., 1996) and Tek/Tie2 (Jones et al., 1999). The consensus site for Grb7 binding was determined to be a p YXN motif, that is also selected by the Grb2 SH2 domain, thus implicating an overlap in the binding specificity of these proteins. However, Tyr-1180 and Tyr-1243 of erbB3, contained within pYXN motifs, interact with Grb7 but not with Grb2, implying that binding specificity is not identical for these adaptors and may be determined by additional residues (Daly, 1998). GrblO also binds to the B-PDGFR (Wang et al., 1999), as well as ephB 1/Elk receptor (Stein et al., 1996), Ret protooncogene (Pandey et al., 1995), IGF-IR (Morrione et al., 1996), and GHR (Moutoussamy et al., 1998). Interestingly, GrblO exhibits differential interaction with the IR and IGF-IR, where it binds more avidly to the former receptor. Jn vitro binding studies implicated that this is due to the interaction of both BPS and SH2 domains with the IR, whereas association with the IGF-IR only involved the BPS module (He et al., 1998). On the other hand, binding of Grb14 to the IR is predominantly mediated by the BPS domain (Kasus-Jacobi et al., 1998). Grbl0 and -14 associate with an altered conformation of the IR kinase domain induced by phosphorylation of Tyr-1150/1151 within the IR activation loop (He et al., 1998; Kasus­ Jacobi et al., 1998; Stein et al., 2001). In addition, Grb14 interacts with the juxtamembrane region and C-terminal tail of the FGFR-1 (Reilly et al., 2000).

1.3.5 RTK-independent interactions The Grb7 family also associate in viva with multiple non-RTKs. Grb7 interacts via its SH2 domain with tyrosine phosphorylated proteins including She (Stein et al., 1994), and FAK

36 (Han and Guan, 1999). Grb7 also binds to the phosphorylated caveolin-1 on Tyr-14 (Lee et al., 2000). Both F AK and caveolin-1 are localised to focal adhesions, and their interactions with Grb7 are implicated in fibronectin and EGF-induced cell migration, respectively. GrblO SH2-binding partners include Bcr-Abl (Bai et al., 1998) and Jak2 (Moutoussamy et al., 1998). Grb 10 also mediates phosphotyrosine-independent interactions via its SH2 domain, where it associates with the regulatory domain of Rafl and the C-terminal tail of Mekl kinases. The interaction with Raf-1 is constitutive, while binding to Mekl requires insulin stimulation (Nantel et al., 1998). Furthermore, GrblO binds primarily via its SH2 domain to the calcium-dependent lipid/protein binding (C2) domain of the E3 ubiquitin ligase Nedd4. This interaction occurs independently of Nedd4 tyrosine phosphorylation or calcium stimulation (Morrione et al., 1999). The interaction of Grbl0 with the Tee tyrosine kinase is also phosphotyrosine-independent although the exact GrblO binding site has not been identified (Mano et al., 1998). Recently, the Grb7 PH domain was reported to preferentially interact with D3- and D5- phosphoinositides. This interaction was increased by cell adhesion to fibronectin (Shen and Han, 2002). Finally, Grb14 interacts via its N­ terminal 110 residues with tankyrase-2, an ankyrin repeat-containing poly (ADP-ribose) polymerase (Lyons et al., 2001). This association may function in the subcellular localisation of Grb 14 and in vesicle trafficking.

Another feature of the GrblO interaction profile was noted when its BPS-SH2 region associated with a full-length GrblO bait, that was utilised to screen a yeast two-hybrid library derived from HeLa cell cDNAs. Gel filtration analysis indicated that hGrblO Band hGrb 10 s were present in tetrameric complexes in mammalian cells. Further in vitro binding studies demonstrated that the interaction was mediated by the BPS-SH2 region and the PH domain of one GrblO protein and the N-terminus of another. Oligomerisation may thus represent a general mechanism to regulate signalling by Grb7 family members. Possible functional consequences include modulation of the interactions with activated RTKs and the generation of diverse signalling platforms by complexes of oligomerised Grb7 family proteins (Dong et al., 1998).

37 1.3.6 Phosphorylation Several laboratories have examined phosphorylation of Grb7 family members in numerous cell types. Their studies suggest that phosphorylation on Ser/Thr residues is a general feature of all three proteins. Fiddes et al. demonstrated using orthophosphate labelling followed by phosphoamino acid analysis, that Grb7 undergoes basal Ser/Thr phosphorylation in unstimulated breast cancer cells, which does not increase following heregulin stimulation (Fiddes et al., 1998). Constitutive Ser/Thr phosphorylation of the Grb7 protein was also noted in unstimulated and EGF-treated SKBR-3 cells (Stein et al., 1994). On the other hand, mGrblO a is phosphorylated on Ser residues in response to EGF treatment (Ooi et al., 1995), and hGrbl0 s undergoes insulin-stimulated Ser phosphorylation, although this does not occur for hGrb 10 ~- Furthermore, hGrb 10 smay be a substrate for MAPK and a PI-3 kinase-dependent kinase since its phosphorylation is abrogated by both PD98059 and wortmannin, and hGrb 10 proteins contain 3 candidate PX(S/T)P MAPK phosphorylation sites (Davis, 1993; Dong et al., 1997). Grb14 exhibits basal Ser phosphorylation in HEK 293 cells, which increases upon PDGF, but not EGF stimulation (Daly et al., 1996). Similarly, in fibroblasts Grb14 is phosphorylated in the basal state predominantly on Ser residues, but also on Thr residues to a minor extent. Grb14 phosphorylation is enhanced upon FGF treatment (Reilly et al., 2000).

Tyrosine phosphorylation of Grb7 is noted by several groups. In esophageal carcinoma cells, Grb7 phosphorylation on tyrosine residues is not observed in the basal state but is induced in response to EGF stimulation, whereas the truncated Grb7V isoform is constitutively tyrosine phosphorylated and is unaffected by EGF treatment. Although the mechanism of Grb7 dephosphorylation is unclear, the authors postulated that SHP-2, which binds the Grb7 SH2 domain in vitro (Keegan and Cooper, 1996), is a likely candidate. Grb7V is then unable to undergo dephosphorylation because it lacks an SH2 domain to bind this phosphatase (Tanaka et al., 1998b). Tyrosine phosphorylation of Grb7 is also mediated by FAK, a Grb7 binding partner (Han and Guan, 1999) that is critical for integrin signalling in cell migration (Ilic et al., 1995; Cary et al., 1996; Gilmore and Romer, 1996). Further, Grb7 phosphorylation is also detected upon replating cells on fibronectin, a process known to activate FAK. These experiments also indicated that Grb7 is phosphorylated by

38 FAK but not by Src-family kinases since Grb7 activation was observed in FAK­ overexpressing cells deficient for the former kinases (Han et al., 2000). Furthermore, Grb7 is phosphorylated by another interacting partner, the tek/tie2 receptor, which is involved in vascular and hematopoietic development (Sato et al., 1995). Association and subsequent Grb7 phosphorylation are abrogated upon mutation of the tek multisubstrate docking site

(Tyr"00) (Jones et al., 1999). However, in a study utilising breast cancer cells, the authors failed to detect reproducible tyrosine phoshorylation of Grb7 in response to EGF stimulation and concluded it may be occurring transiently (Stein et al., 1994).

Tyrosine phosphorylation of GrblO proteins has also been reported by some groups. For instance, hGrblO y tyrosine phosphorylation occurs transiently in response to insulin stimulation in IR-overexpressing fibroblasts (Frantz et al., 1997). Furthermore, weak tyrosine phosphorylation of hGrblO Bwas detected in response to insulin stimulation (Liu and Roth, 1995). However, the authors conducted another study in which CHO-IR cells overexpressing hGrb 10 Band ~ were treated with tyrosine phosphatase inhibitors such as pervanadate and orthovanadate. This allowed the detection of tyrosine phosphorylated GrblO proteins in pervanadate-treated, unstimulated cells. Phosphorylation of GrblO proteins was also enhanced in cells treated with both insulin and vanadate relative to cells treated with insulin alone. GrblO phosphorylation was not mediated by association with the IR but by Src-family kinases. Interestingly, this study also implicated a negative modulatory role for Grb 10 phosphorylation in insulin signalling, since mutation of the GrblO insulin-phosphorylated site (Tyr-67) to glycine resulted in higher affinity binding to the IR (Langlais et al., 2000). In HEK 293 cells, hGrbl0 B is significantly tyrosine phosphorylated by the non-receptor TK Tee (Mano et al., 1998). However, GrblO is not tyrosine phosphorylated in response to EGF treatment (Ooi et al., 1995). Finally, phosphorylation of Grb14 on tyrosine residues has not been reported.

1.3.7 Function The importance of the functional modules inherent in the Grb7 family structure, coupled with the diversity of their interacting partners has implicated these proteins in a wide and disseminated signalling network. Indeed, although the precise role of the Grb7 family

39 members is yet to be uncovered, functional studies utilising cellular models have generated certain clues to elucidate their biological roles.

1.3.7.1 lnsulin/lGF-1-induced mitogenic and metabolic effects Studies focusing on the role of the GrblO proteins in insulin/lGF-I signalling have resulted in conflicting findings, suggesting that Grb 10 function may depend on the cellular context or the particular isoform under investigation. The majority of the literature points to an inhibitory role, since hGrblO ~ overexpression in CHO-IR cells reduced insulin-simulated PI-3 kinase activation, and tyrosine phosphorylation of both IRS-1 and a 60 kDa GAP­ associated-protein (Liu and Roth, 1995), subsequently identified as p62Dok (Wick et al.,

2001). IR phosphorylation was not altered by hGrblO ~ (Liu and Roth, 1995). In another study, overexpressed mGrblO a inhibited IGF-I-induced growth of mouse fibroblasts, with no effect on insulin-stimulated cells (Morrione et al., 1997). A more recent investigation demonstrated that hGrb 10 ~ overexpression in hepatocytes inhibited IR phosphorylation and insulin-mediated glycogen synthesis by an undefined mechanism, which did not involve altered phosphorylation of IR substrates (Mounier et al., 2001). On the other hand, another group utilising overexpressed mGrblO a and dominant negative cell-permeable Grb 10 SH2 peptides in fibroblasts, pointed to a consistent mitogenic effect of mGrb 10 a in PDGF-BB-, IGF-I- and insulin-stimulated DNA synthesis (Wang et al., 1999). However, more recent in vitro experiments utilising a purified Grb 10 BPS domain demonstrated that it directly decreases the catalytic activity of the IR and IGF-IR, thus hindering substrate phosphorylation. This is dependent on phosphorylation of the receptor kinase activation loop. In addition, peptide competition experiments revealed that the mechanism of inhibition did not include a direct BPS interaction with phosphorylated tyrosine residues within the activation loop (Stein et al., 2001). Finally, microinjection of a GrblO BPS-SH2 fusion protein in rat fibroblasts inhibited insulin- and IGF-I-induced mitogenesis, but did not impede serum or EGF stimulation (O'Neill et al., 1996).

These results can be interpreted in a number of ways, for instance the BPS-SH2 fusion protein may have a dominant negative role by binding to the IR and blocking receptor interaction with the endogenous GrblO. This may inhibit an uncharacterised mitogenic

40 signalling pathway (O'Neill et al., 1996) and suggests a stimulatory role for GrblO. On the other hand, the IR interaction with the microinjected BPS-SH2 fragment may prevent other substrates from direct interaction with the receptor and thus the transmission of mitogenic signalling cascades. This implies an inhibitory role for GrblO. In addition, at the time of conducting these experiments the injected fragment was presumed to solely comprise the SH2 module, since this study preceded the identification of the BPS domain (He et al., 1998). Therefore, to reconcile the observed isoform-specific differences, a direct functional comparison of these proteins is required in a single study utilising the same cellular assay(s). This has not been undertaken to date.

Despite the discordance in functional studies of GrblO proteins, research groups have only implicated an inhibitory function for Grb14. Overexpression of the rat Grb14 (rGrb14) inhibited IRS-1 phosphorylation, insulin-stimulated mitogenesis and glycogen synthesis in CHO-IR cells, without altering phosphorylation of the IR (Kasus-Jacobi et al., 1998). These effects are associated with a delayed activation of Erkl/2 and Akt/PKB, in addition to a reduction in the magnitude of Erk activation in Grb14-overexpressing cells (Bereziat et al., 2002). The effects of Grb14 overexpression are also inhibitory in reponse to FGF­ stimulation, where DNA synthesis is reduced in mouse fibroblasts (Reilly et al., 2000). Similar to GrblO, the Grb14 BPS, but not SH2 domain inhibits IR and IGF-IR catalytic activity in vitro. However, relative to the remaining Grb7 family members, Grb14 most specifically and potently affects the IR kinase domain (Bereziat et al., 2002). Grb14 also inhibits the IGF-IR kinase activity more weakly than the IR. A functional role for Grb7 in insulin/IGF-1 signalling has not been investigated despite the fact that it interacts with the IR (Kasus-Jacobi et al., 2000).

1.3.7.2 Cell Migration The role of Grb7 in cell migration was investigated after the identification of FAK as a binding partner. FAK is a 125 kDa tyrosine kinase which triggers integrin signal cascades regulating cell migration, survival and cell proliferation (Ilic et al., 1995; Frisch et al., 1996; Guan, 1997; Zhao et al., 1998). Grb7 binds to the autophosphorylated Tyr-397 of FAK, which represents a major docking site and recruits Src, PLC-yl and PI-3 kinase

41 (Chen and Guan, 1994; Schaller et al., 1994; Xing et al., 1994; Zhang et al., 1999). In FAK-overexpressing cells, Grb7 is localised to focal contacts and is tyrosine phosphorylated through its cell adhesion-dependent association with FAK, as stated earlier (Han and Guan, 1999; Han et al., 2000). Overexpression of Grb7 in fibroblasts enhances migration towards fibronectin, and overexpression of a dominant-negative Grb7 SH2 domain reduces fibroblast migration, implicating a critical role for Grb7 in cell migration although there is no effect on cell cycle progression (Han et al., 2000). Furthermore, reduction of Grb7 and Grb7V protein levels by expression of antisense mRNA in esophageal carcinoma cells markedly decreases EGF-stimulated cell migration in a matrigel assay, without altering cell proliferation (Tanaka et al., 1998b).

The interaction of the Grb7 PH domain with phosphoinositides is important for mediating cell migration and is increased by cell adhesion to fibronectin. However, it is not necessary for Grb7 localisation to focal contacts which is directed by the SH2 domain through its association with FAK (Shen and Han, 2002). In addition, treatment of cells with PI-3 kinase inhibitors (wortmannin and L Y294002) reduces the Grb7 association with phospholipids and abrogates its ability to stimulate cell migration, whereas a constitutively active pl 10 subunit of PI-3 kinase enhances the Grb7/phospholipid interaction. These studies suggest that Grb7 may be a downstream effector of PI-3 kinase in regulating cell migration (Shen and Han, 2002). Therefore, studies indicate a specific role for Grb7, but not for other family members in cell migration, since overexpression of hGrblO B, s or Grb14 does not alter migration of fibroblasts relative to controls (Han et al., 2000).

1.3.7.3 Apoptosis Co-immunoprecipitation analysis and in vitro binding studies demonstrated that GrblO binds the Raf-1 kinase via its SH2 domain in a constitutive and cell-type specific manner. A role for Grb 10 in apoptosis was indicated when overexpression of hGrb 10 s proteins with distinct SH2 point mutations (Arg-BB5 to Leu, Ser-BB7 to Cys and Thr-BC5 to Ser) that abrogate association with EGFR, IR, Raf-1 or MEKl caused apoptosis in HTC-lR and COS-7 cells. This phenotype was rescued by expression of the wildtype hGrb 10 s (Nantel et al., 1998). It was also demonstrated using immunofluoresence microscopy and

42 subcellular fractionation that endogenous Grb 10 is predominantly associated with mitochondria in COS-1 cells. Further, endogenous GrblO and Raf-1 co-immunoprecipitate from mitochondrial extracts (Nantel et al., 1999). Although Raf-1 acts in the Ras cascade to mediate RTK signalling, it is also implicated in the regulation of apoptosis, and is targeted to mitochondria by Bcl-2 where it causes resistance to apoptotic factors (Wang et al., 1996; Pritchard and McMahon, 1997). The GrblO/Raf-1 interaction is enhanced by ultraviolet radiation, which also induces apoptosis, thereby implicating Grb 10 in regulating the anti-apoptotic activity of Raf-1 (Nantel et al., 1999).

1.3.7.4 Cancer The link between Grb7 family members and cancer was first made with studies on the Grb7 protein. Grb7 was mapped to a region of mouse chromosome 11, syntenic to a region of human chromosome 17q that is frequently amplified in breast cancer (Daly, 1998). It soon became evident that Grb7 is located on an amplicon that also includes the erbB2 gene. Grb7 is co-amplified and overexpressed with erbB2 in breast cancer cell lines and tumours. Further, Grb7 interacts with erbB2 in breast cancer cells, implicating co-amplification of the respective genes in upregulation of an erbB2 signalling pathway in breast cancer (Stein et al., 1994). Grb7 overexpression is also noted in about 44% of esophageal cancer specimens although it is not due to gene amplification. Co-expression of Grb7 with the EGFR and erbB2, evident in approximately 30% of esophageal cancers, is related to extramucosal tumour invasion. Interestingly, the relationship with invasion is lost upon analysis of individual gene expression, thereby supporting the notion that upregulation of a Grb7/erbB pathway is involved in tumour progression (Tanaka et al., 1997). The invasive role attributed to Grb7 has also been demonstrated for its sole isoform Grb7V. Expression of the latter is restricted to invasive and metastatic esophageal cancers, and is enhanced in metastatic deposits in the lymph nodes. Grb7V is suggested to confer a more invasive phenotype in esophageal cancers, possibly as a result of its constitutive tyrosine phosphorylation (Tanaka et al., 1998b).

However, unlike Grb7, GrblO overexpression is not reported in cancers, although its interaction profile implicates a role for Grb 10 in oncogenic signalling pathways. For

43 example, GrblO interacts with activated Bcr-Abl at a Bcr site between residues 242-446 (Bai et al., 1998). Bcr-Abl is an oncogenic TK implicated in cell transformation and leukaemia (Gishizky and Witte, 1992). A Bcr-Abl mutant defective in GrblO binding displays a reduced capacity to confer IL-3 independence to haematopoietic Ba/F3 cells (Bai et al., 1998). GrblO also binds to the Ret protooncogene (Pandey et al., 1995), a RTK with transforming activity which is mutated in multiple endocrine neoplasia types 2A and 2B and in thyroid cancers (Hofstra et al., 1994). The interaction is dependent on Ret activation and implicates a role for GrblO in Ret-mediated oncogenesis (Pandey et al., 1995). Finally, Grbl4 is differentially expressed in breast cancer cell lines where its mRNA expression correlates with ER positivity (Daly et al., 1996).

44 Ofjectives

Since Grb14 is the most recently identified member of the Grb7 family, at the beginning of these studies there was very limited information available regarding its function or interaction profile with RTKs. In addition, there were no downstream interacting partners identified for Grb14, thus its function in normal cellular signalling and in tumorigenesis was undefined. However, different isoforms were isolated for GrblO that displayed differential expression and receptor recruitment, which were postulated to result in distinct biological effects. Thus, the overall aim of this thesis is:

To characterise the role of Grb14 signalling in normal cellular function and in cancer

In order to map the signalling cascade(s) and potential role(s) of Grb14, the specific aims of this thesis are:

To identify candidate Grb14 isoforms and investigate their interaction profile with RTKs

To elucidate the functional roles of Grb14 isoforms in signalling downstream of RTKs

To investigate the regulation and functional consequences of Grb14 signalling in breast cancer cells

45 Chayter 2- 'M.ateriafs &'Methods

2.1 Molecular bioloe,y

2.1.1 The polymerase chain reaction (PCR) Primers were generally designed with an approximate 50% GC content and were analysed using the Amplify software program to ensure that the likelihood of forming primer-dimers or non-specific annealing was minimal. All primers are listed in Appendix A. A standard PCR reaction was undertaken with the Expand™ high fidelity PCR system (Boehringer) with lx Expand HF buffer, 1.5 mM MgC1 2, 0.4 mM dNTP mix (dATP, dCTP, dGTP, dTTP), 10 pmol of each primer, 2.5U of enzyme and dH20 to 25 µI. The following cycling parameters were performed: an initial denaturing step at 94 °C for 1 min, followed by 35 cycles of denaturation at 94°C for 30 s, annealing at 55°C for 30 s, and extension at 72°C for 2 min in addition to a final extension of 72°C for 5 min, followed by cooling at 4°C. PCR was generally performed using the PTC-100™ programmable thermal cycler (MJ Research, Inc.).

Amplification from T-47D genomic DNA was undertaken using the Expand™ Long template PCR system (Boehringer) and GEN 1 & 2 primer sets. 1 ng of DNA was used as template with the following modifications from the standard protocol: Initial denaturation was for 2 min, annealing was performed at 58°C and extension at 68°C for 10 min with 30 cycles. Final extension was at 68°C for 15 min.

PCR was also utilised to generate fragments corresponding to the Grb 14 J3-encoding sequence and various domains of Grb14 isoforms (Figure 4.1) for subcloning into pRcCMVFiag (Daly et al., 1996) and pGEX-2T (Amersham Biosciences, NSW, Australia), respectively. To amplify the Grb14 J3 sequence, primers Del Flag-F and Del Flag-R (Appendix A) were utilised in a standard PCR reaction with the addition of 1.3 µl DMSO to the solution, annealing at 48°C and the hotstart technique. This entailed heating PCR reactions to 80-85°C to prevent the formation of primer dimers prior to enzyme addition

46 and subsequent thermal cycling. Amplification of the Grb14 ~ SH2 region was undertaken with Del GST-F and Del GST-R primers with standard parameters except the annealing step was at 48°C. Regions encoding the Grb14 a and~ BPS-SH2 were amplified according to standard PCR protocol and the hotstart technique. Grb14 a BPS-SH2 required 2.5 mM

MgC1 2 in the reaction and the primers BPS14.1 and Grb14.2(2). The Grb14 ~BPS-SH2- encoding sequence was amplified with 4 mM MgC12 using primers BPS14.l and Del GST­ R(2) (see Appendix A). 70 ng of a plasmid containing the Grb14 a sequence was utilised as template for these reactions.

2.1.2 RNA extraction Total RNA extractions were performed with the RNAqueous RNA isolation kit (Ambion,

TX, USA) according to the manufacturer's instructions. Briefly, approximately lxl06 MCF-7 cells were washed twice with ice-cold PBS, lysed using 300 µl of lysis/binding solution and collected using a cell scraper followed by transfer to an eppendorf tube. After vortexing, the lysate was mixed with an equal volume of 64% ethanol by gentle pipetting, and loaded into a RNAqueous filter cartridge. This was centrifuged for 1 min, the flow­ through was discarded and the filter washed with 700 µl of wash buffer, followed by another 1 min spin. A further 2 washes with 500 µl of wash buffer ensued, in addition to a final 2 min spin to remove residual buffer after which filter cartridges were transferred to clean eppendorfs. 40 µl of elution buffer was then added to the filter and the tube was heated at 65°C for 10 min followed by RNA elution after a 1 min spin. The elution step was repeated and the recovered RNA pooled and stored at -80°C. RNA quantitation was undertaken using a Biorad spectrophotometer to evaluate RNA quantity (at 260 nm) and purity (at 280 nm) as described previously (Sambrook and Russell, 2001).

2.1.3 cDNA synthesis and RT-PCR To ensure complete removal of contaminating DNA prior to cDNA synthesis, RNA samples (5 µg) were digested with DNasel (Boehringer) for 1 hat 37°C, and the enzyme was inactivated by the addition of 1 µl termination mix (Clontech, Palo Alto, CA). The DNAse-free RNA was then utilised for cDNA synthesis using the Superscript™ preamplification system for first strand cDNA synthesis (Gibco BRL/lnvitrogen) according

47 to the manufacturer's protocol. Briefly, each RNA sample was mixed with 0.5 µg of Oligo­ dT primer and DEPC-treated H20 to a final volume of 12 µl, followed by heating at 70°C for 10 min and chilling on ice for at least 1 min. The sample was then mixed with 8 µl of the reaction mix containing 2 µl of lOx PCR buffer (200 mM Tris-HCl, pH 8.4 and 500 mM KCl), and a final concentration of 2.5 mM MgC12, 0.5 mM each dATP, dCTP, dGTP, dTTP, 10 mM DTT, followed by incubation at 42°C for 2 min. Subsequently, 200U of Superscript II reverse transcriptase (RT) was mixed with each test sample followed by incubation at 42°C for 50 min. Controls lacking reverse transcriptase (-RT) were treated in an identical manner to the test samples except they were incubated without the addition of enzyme. Subsequently, each sample was heated at 70°C for 15 min, chilled on ice and centrifuged briefly prior to mixing with 2U of RNaseH and a final incubation at 37°C for 20 min. To identify Grb14 isoforms, primer set Gl-G5 (Appendix A) were utilised and RT-PCR was undertaken according to standard conditions with 1 µl of each RT reaction as template and the following modifications: Initial denaturation was performed at 94 °C for 5 min, extension was at 72°C for 1 min with a final extension for 7 min. In order to generate sufficient product for subsequent cloning procedures, RT-PCR was undertaken twice, where 1-5 µl of the product from the first amplification served as a template for the second PCR.

To examine the expression of Grb14 ~ in normal breast epithelial cells and breast cancer cell lines, a Grb14 ~-specific primer, G5DEL2, was designed (Appendix A). The synthesized cDNA was first amplified with ~-actin primers using standard conditions to ensure that any variations in expression among the cell lines was not due to RNA degradation during cDNA synthesis. Modifications in cycling parameters using ~-actin primers (Promega) were annealing at 58°C with 30 cycles. Grb14 ~-specific RT-PCR was performed using primers G5DEL2 and G5R as outlined previously for a standard PCR reaction with the annealing step undertaken at 63°C and the hotstart technique.

48 2.1.4 Subcloning of PCR products and generation of expression constructs The RT-PCR product corresponding to a potentially novel Grb14 isoform and amplified from HEK 293 cells using G5 primers, was subcloned using the PCR-Script™ Amp cloning kit (Stratagene, CA, USA). Subsequently, it was transformed into E. coli ultracompetent cells (Stratagene) according to the provided instructions. Genomic PCR products from T- 47D cells were purified using the Wizard® DNA clean up system (Promega), subcloned in the pGEM-T® Easy vector (Promega) using TA cloning according to the manufacturer's protocol and transformed into E. coli DH5a cells.

Constructs harbouring the Grb14 BPS domain in pGEX-4T2 (Amersham Biosciences, NSW, Australia) and the Grb14 a SH2 domain in pGEX-2T (Daly et al., 1996) were prepared by R. Deane and P. Janes, respectively. To facilitate subcloning into expression vectors, primers utilised to amplify the Grb14 J3 sequence or nucleotides encoding the Grb14 a/J3 domains were designed with restriction sites (see section 4.2 and Appendix A). PCR products corresponding to the Grb 14 J3 sequence ( - 1.4 kb), and the BPS-SH2 domains of Grb14 isoforms (Grb14 a, 557 bp; Grbl4 J3, 331 bp) were purified using the Wizard® DNA clean up system (Promega) according to the manufacturer's instructions. Since the Promega kit excludes nucleotides less than 200 bp, the amplified Grb14 a SH2- encoding region ( 107 bp) was purified using the High Pure PCR product purification kit (Boehringer). The purified Grb14 J3 sequence was digested with Hindlll and BamHI, whereas pRcCMV Flag was digested with Hindlll and Bgl II. PCR products corresponding to regions of Grb14 isoforms (Grb14 J3 SH2, Grb14 a and J3 BPS-SH2) and the pGEX-2T vector were digested with EcoRI and BamHI. All digestions were undertaken at 37°C for 2 h, after which the digested fragments were electrophoresed, excised from the gel and purified using the QIAEX gel extraction kit (QIAGEN, Adelaide, Australia). The gel­ purified fragments were resuspended in 15-20 µl of Tris/HCl pH 8.5 and directionally subcloned into the respective vectors by ligation at 4°C overnight, after which they were transformed into E. coli DH5a cells, as described previously (Sambrook et al., 1989). All constructs were verified using cycle sequencing.

49 Eukaryotic CMV promoter-based expression vectors for the human EGFR and the ~­ PDGFR, as well as the Grb14/pRcCMVF,ag construct (where Grb14 denotes Grb14 a), were described previously (Daly et al., 1996). Expression vectors for the IR and IGF-IR were obtained from A. Ullrich (Department of Molecular biology, Max-Planck-Institute for Biochemistry, Martinsried, Germany), whereas the FGFR-1 construct was kindly provided by J. Schlessinger (Department of Pharmacology, New York University School of Medicine, New York).

2.1.5 Generation of retroviral constructs Generation of retroviral Grb14 constructs was performed by R. J. Lyons. The sequence encoding each Grb14 isoform together with a Flag epitope-tag was excised from the respective eukaryotic expression vectors Grb14/pRcCMVFiag and Grb14 ~/pRcCMVF,ag by enzyme digestion with HindIII and partial digestion with Apal. The digested fragments were then gel-purified, end-filled with T4 DNA polymerase (Promega) and ligated into the pLib vector (Clontech), which was prepared by digestion with EcoRI and NotI prior to end­ filling. Insert orientation was verified by restriction enzyme digestion and cycle sequencing. To generate Grb14Fia/pBabePuro, the sequence encoding the entire open reading frame of Grb 14 together with a Flag epitope-tag was excised from the eukaryotic expression vector Grb14/pRcCMV Flag with HindIII/BclI. After end-filling with T4 DNA polymerase, the fragment was ligated into the BamHI-digested and filled site in the retroviral vector pBabePuro (provided by G. Peters, Imperial Cancer Research Fund, UK). Orientation of the insert was verified by restriction enzyme digestion with EcoRI only and with XhoI in combination with HindIII.

2.1.6 Cycle sequencing Plasmid DNA was prepared using the Wizard™ DNA miniprep kit (Promega). Sequencing reactions were undertaken using the ftnol® DNA cycle sequencing system (Promega). 10 pmol of primer was end-labelled with 33P-ATP using T4 polynucleotide kinase at 37°C for 30 min. The enzyme was then inactivated by heating at 90°C for 2 min. 1.5 pmol of labelled primer was subsequently added to a reaction mix containing Ix buffer (73.5 mM

Tris-HCl pH 9, 2.9 mM MgC12), 9.5 µl of the miniprep DNA, 1 µl of sequencing grade Taq

50 DNA polymerase and H20 to a final volume of 17 µl. 4 µl of the sequencing reaction was then dispensed in each of 4 dideoxynucleotide tubes containing 2 µl of the following: ddGTP, ddATP, ddTTP or ddCTP. The tubes were then placed in the thermal cycler for cycle sequencing as follows: an initial denaturation step of 95°C for 2 min followed by 30 cycles of 95°C for 30 s, 50-72°C for 30 s (refer to Appendix B), 70°C for 1 min followed by cooling at 4 °C. The reactions were stopped by the addition of 4 µl sequencing stop solution (10 mM NaOH, 95% (v/v) formamide, 0.05% (w/v) bromophenol blue and 0.05% (w/v) xylene cyanol) and incubated at 95°C for 2 min, prior to electrophoresis on a denaturing polyacrylamide sequencing gel. Finally, the gel was dried by heating under vaccuum after which it was exposed to Kodak X-ray film overnight at RT.

51 2.2 Western blottine and Immunoprecipitations 2.2.1 Antibodies

Primary Antibody Commercial source Applications

Anti-B-actin, monoclonal Western blotting Sigma Clone AC-15, A-5441 1:3000 (2.7 mg/ml)

Anti-FGFR-1, rabbit polyclonal Western Blotting F5421 Sigma 1:500 (1 mg/ml) Anti-Flag® M2, monoclonal Immunoprecipitation F3165 Sigma 5 µg/IP (3 mg/ml)

FLAG-probe (D-8), rabbit polyclonal Western blotting sc-807 Santa Cruz 1:2000 (200 µg/ml) Biotechnology

GRB14 (N-19), goat polyclonal Santa Cruz Western Blotting sc-6103 1:750 (200 µg/ml) Biotechnology

IGF-lRB (C-20), rabbit polyclonal Western Blotting sc-713 Santa Cruz 1:500 (100 µg/ml) Biotechnology

Anti-insulin receptor CB-subunit) Western Blotting rabbit polyclonal Transduction Labs 1:1000 I 16630, (0.25 mg/ml) Insulin-receptor (Ab-3), monoclonal Immunoprecipitation Oncogene Research GR07 4 µg/IP (100 µg/ml) products

IRS-1 (C-20), rabbit polyclonal Santa Cruz Western Blotting sc-559 Biotechnology 1:1000 (100 µg/ml) c-Myc, rabbit polyclonal Santa Cruz Western Blotting C-19 Biotechnology 1:1000 Anti-phosphotyrosine (PY20) Western Blotting monoclonal Transduction Labs, 1:1000-1:10 000 P11120 Kentucky (1 mg/ml)

52 Secondary Antibody Commercial source Applications HRP-conjugated anti-mouse Amersham Life Western Blotting NA 931V Science 1:2000-1 :4000 HRP-conjugated anti-rabbit Amersham Life Western Blotting NA 934V Science 1:2000-1 :4000 Protein A-peroxidase Zymed Laboratories, Western Blotting conjugate (EIA Grade) Inc., CA 1:10 000 10-1023 Ree-Protein G-peroxidase Zymed Laboratories, Western Blotting conjugate (EIA grade) Inc., CA 1:10 000 10-1223 Goat anti-mouse IgG (H+L)-sepharose 4B Zymed Laboratories, Immunoprecipitations conjugate Inc., CA (refer to protocol) 62-6541

Protein A-sepharose 4B Zymed Laboratories, Immunoprecipi tations 10-1042 Inc., CA (refer to protocol)

2.2.2 Lysate preparation Following stimulation with the appropritate mitogen, cell culture media was removed by suction and cells were washed twice in ice-cold PBS on ice. Lysates were then prepared by harvesting cells at 4°C in lysis buffer (50 mM HEPES, pH 7.5 , 150 mM NaCl, 10% (v/v) glycerol, 1% (v/v) Triton X-100, 1.5 mM MgC12, 1 mM EGTA, 10 mM sodium pyrophosphate, 100 mM NaF) containing the protease inhibitors aprotinin (10 µg/ml), leupeptin (10 µg/ml) and PMSF (lmM), in addition to the phosphatase inhibitor sodium orthovanadate (lmM). Cells were scraped into the lysis buffer, transferred into eppendorf tubes and vortexed well prior to a 5 min centrifugation at 4°C, 14 000 rpm in an Eppendorf benchtop centrifuge in order to clarify the lysate. The resulting pellet was then discarded, the supernatant transferred to a new eppendorf tube and stored at -80 °C. Protein concentrations were determined using the Biorad Protein Assay Reagent (Biorad).

2.2.3 Western blotting analysis Lysates (50-100 µg) were denatured in 3x sample buffer (62.5 mM Tris, pH 6.8, 5% (v/v) ~-mercaptoethanol, 3% (w/v) SDS, 10% (v/v) glycerol) by boiling at 95-100°C for 3-5 min. Samples and molecular weight markers (Biorad or Sigma) were separated by SDS-PAGE (8-10%) in 375 mM Tris-HCl, pH 8.8, 0.1 % (w/v) SDS. Stacking gels contained 125 mM

53 Tris-HCl, pH 6.8 with 0.1 % (w/v) SDS. Subsequently, proteins were transferred to nitrocellulose membranes (Biorad) which were stained by Ponceau S to verify equivalent loading and transfer quality. To block non-specific binding sites, membranes were generally incubated in TBS-BSA solution (10 mM Tris, pH 7.4, 150 mM NaCl, 5% (w/v) BSA). When blotting with the Grb14 antibody (N-19), membranes were blocked with 10% (w/v) skim milk powder in TBS/Triton X-100 (10 mM Tris, pH 7.4, 150 mM NaCl, 0.1 % (v/v) Triton X-100). Membranes were incubated with the primary antibody for 1-2 hat RT or overnight at 4°C in TBS-BSA solution or in 5% (w/v) skim milk-containing TBS/Triton X-100 when using the Grb14 antibody (N-19). All subsequent incubations and washes are carried out at RT. Membranes were then subjected to 10 min washes for a total period of 30 min in TBS/Triton X-100, followed by incubation with the corresponding horseradish peroxidase (HRP)-conjugated secondary antibody in TBS/Triton X-100 with 5% (w/v) skim milk powder for 1 h. This was followed by 10 min washes as described earlier for 30 min-1 h. Proteins were visualised using enhanced chemiluminescence (ECL, NEN Life Sciences), and exposure was by autoradiography using the appropriate Kodak X-ray film. Protein quantitations were undertaken utilising densitometry (Molecular Dynamics Personal Densitometer SI).

2.2.4 Immunoprecipitation assays Lysates (0.5-lmg) were usually precleared by incubation for lh at 4°C with either protein A sepharose or goat anti-mouse beads (40 µl from a 1:1 slurry, pre-washed in PBS). Subsequently, the cleared lysate was transferred to new eppendorf tubes and incubated with the primary precipitating antibody for at least 2 hat 4°C with constant mixing, followed by a 1 h incubation with 40 µl of protein A sepharose or goat anti-mouse beads under the same conditions. Immunoprecipitates were subsequently washed 3 times with lysis buffer, denatured in lx sample buffer, separated by SDS-PAGE and transferred to nitrocellulose membranes and analysed by Western blotting.

2.2.5 Preparation of GST fusion proteins GST fusion proteins corresponding to various domains of Grb 14 isoforms were made by amplifying the sequence of interest using PCR, followed by directional subcloning into the appropriate GST bacterial expression vector and DNA sequencing to ensure it was free of

54 mutations, as described. Plasmids were transformed into E.coli DH5a or BL-21 cells, plated out on LB agar containing the appropriate antibiotic and incubated at 37°C overnight. The next day, a colony was picked and utilised to innoculate 5 ml of LB medium with antibiotic. The culture was grown overnight at 37°C with vigorous shaking, and in turn utilised to innoculate 500 ml of antibiotic-containing LB medium. These cultures were grown at 37°C to an optical density (0.D.) of approximately 0.6 then protein expression was induced by the addition of 1 mM IPTG with continued incubation for an additional 2 h.

To minimise degradation products, cultures containing the Grb14 ~ SH2 or BPS-SH2 domains were grown to an O.D. of 0.8 with subsequent induction at 30°C for 1 h (SH2) or for 30 min (BPS-SH2). Cultures were then centrifuged at 5000 rpm for 10 min at 4°C in a Beckman centrifuge using a JA-10 rotor. The resulting cell pellet was resuspended in 10 ml of ice-cold PBS and cells were lysed by french-pressing. Subsequently, Triton X-100 was added to a final concentration of 1% (v/v), and the lysate was centrifuged at 12000 rpm for 30 min at 4°C using a JA-20 Beckman rotor. 200-600 µl of glutathione-agarose beads (Amersham BioSciences), preswollen in PBS, were then incubated with the supernatant for 2 h at 4°C with constant rotation. The beads were collected by centrifugation for 3 min at 1500 rpm at RT and resuspended in 10 ml of ice-cold PBS. Following triplicate washings, beads were resuspended in 500 µl of ice-cold PBS. To visualise the bound GST fusion proteins, the GST beads and BSA standards were incubated in sample buffer at 95-100°C for 3-5 min to denature the associated proteins. These were then separated by SDS-PAGE and stained using Gelcode® Blue Stain reagent (Pierce, IL, USA) according to the manufacturer's instructions. GST fusion protein concentrations were estimated by comparison with known amounts of BSA standards.

2.2.6 In vitro binding assays 1O µg of GST fusion protein coupled to glutathione-agarose beads was utilised per sample and the total amount of glutathione-agarose beads was normalised across samples. In addition, CL6-B sepharose beads (Amersham BioSciences) were added to a total volume of 40 µland incubated with 1 mg of lysate for 2 hat 4°C with continous agitation. The beads were washed 4 times with lysis buffer prior to resuspension in 20 µl of 2x sample buffer.

55 Proteins were denatured by boiling at 95-100°C for 3-5 min, after which they were separated by SDS-PAGE and analysed by Western blotting.

2.3 Cell culture

2.3.1 Mitogens and inhibitors EGF was obtained from R&D Systems and was reconstituted in PBS in 100 µg/ml stocks. IGF-I was acquired from Boehringer Mannheim and was dissolved in 10 mM HCl in aliquots at 10 µM. PDGF-BB and human recombinant acidic fibroblast growth factor (aFGF) were obtained from Gibco BRL and Sigma, respectively. Both were reconstituted in 0.1 % BSA (w/v) in PBS at a final concentration of 10 µg/ml and 1 µg/ml, respectively. Human recombinant bFGF was acquired from R&D systems and was dissolved in 0.1 % BSA (w/v) in PBS in working aliquots of 10 µg/ml.

Stock solutions of insulin were obtained from Novo Nordisc Pharmaceuticals Pty Ltd. (New South Wales, Australia). 7a-[9-(4,4,5,5,5,-pentafluoropentylsulfinyl)nonyl]estra- 1,3,5,(10)-triene-3, 17~-diol (ICI 182780) was kindly provided by A. Wakeling

(AstraZeneca Pharmaceuticals, Cheshire, UK), dissolved in ethanol to 10-2M, and diluted in culture medium prior to each experiment. 17~-estradiol (estradiol, Sigma) was dissolved in ethanol to provide concentrates of 2x 10-4 and 2x 10-5 M.

2.3.2 Cell lines/strains The 184 normal human mammary epithelial cell strain was obtained from Dr. Martha Stampfer (University of California, Berkeley). This was maintained in mammary epithelial growth medium (Clonetics). The human breast cancer cell lines BT-20, BT-474, BT-483, BT-549, Hs578T, MDA-MB-134, MDA-MB-361, MDA-MB-436, MDA-MB-453, MDA­ MB-468, SK-BR-3 and ZR-75-1, the prostate cancer cell line DU145 and the human embryonic kidney HEK 293 cell line were obtained from the American Type Culture Collection (ATCC, Rockville, MD). The MDA-MB-231, MCF-7 and T-47D human breast cancer cell lines were provided by the EG and Mason Research Institute (Worcester, MA). All cell culture media and reagents were acquired from Gibco BRL (Grand Island, NY), unless indicated otherwise. Human breast cancer cell lines were maintained in 1ON RPMI,

56 which consists of RPMI 1640 medium, supplemented with 6 mM glutamine, 14 mM sodium bicarbonate, 20 mM HEPES, 10 µg/ml insulin, gentamycin (20 µg/ml, Amersham Biosciences, NSW, Australia) and 10% (v/v) fetal calf serum (FCS). The human prostate cancer cell line DU145 was maintained in the same medium as the breast cancer cell lines. HEK 293 cells were maintained in minimal essential medium supplemented with Earle's salts, 2 mM glutamine, gentamycin (20 µg/ml), 6.7 mM sodium bicarbonate, and 10% FCS. COS-7 African green monkey kidney cells were obtained from the ATCC, and cultured in Dulbecco's modified Eagles's medium (DMEM) containing 10% (v/v) FCS, gentamycin (20 µg/ml) and 2 mM glutamine. NIH 3T3 cells were maintained in DMEM supplemented with 10% (v/v) newborn calf serum (NBCS), 2 mM glutamine and gentamycin (20 µg/ml). CHO-IR cells are CHO cells stably expressing the IR, and were acquired from M. F. White (Howard Hughes Medical Institute, Joslin Diabetes Center, Harvard Medical School, Boston, USA). These cells were maintained in Ham's F12 media containing gentamycin (20 µg/ml) and 10% (v/v) FCS. Finally, Phoenix-Eco cells were kindly provided by P. Achacoso and G. Nolan (Stanford University Medical Centre, Stanford, CA), and were cultured in DMEM containing 10% heat-inactivated FCS. Heat inactivation was undertaken at 55°C for approximately 30 min. All cell cultures, unless specified otherwise, were maintained at 37°C and 5% CO2 in a humidified atmosphere.

2.3.3 Generation of MCF-7/EcoR stable pools To generate stable pools of MCF-7 cells expressing the murine retroviral receptor (MCF7/EcoR cells), 10 µg of a plasmid encoding this receptor (pWZLneoEco\ also provided by G. Peters) was transfected into MCF-7 cells using FuGENE™-6 reagent (Roche Molecular Diagnostics, Australia) according to the manufacturer's instructions. The following day, GENETICIN® (Invitrogen, 600 µg/ml) was added and selection was allowed to proceed for approximately 4 weeks.

57 2.3.4 Retroviral infection and generation of MCF-7/EcoR control or Grb14-overexpressing stable pools Retroviral production was undertaken as described previously (Musgrove et al., 2001). Briefly, retroviral constructs were transfected into Phoenix-Eco cells using FuOENE™-6 reagent. 24 h following transfection, the medium was replaced and cells were maintained in fresh medium at 32°C for an additional 24 h, after which the viral supernatant was collected and filtered through a Millex®-HV 0.45 µm filter unit (Millipore, Bedford, MA). Polybrene (Sigma) was added to the virus to a final concentration of 16 µg/ml prior to infection of target cells or storage at -80°C.

Ecotropic retroviruses encoding Orb 14 isoforms were produced using the Orb 14 a or ~/ pLib retroviral constructs in parallel with a OPP-expressing retrovirus. The latter was produced from the OPP-encoding retroviral vector pLib-EOFP (Clontech) to estimate infection efficiency. For infections, NIH 3T3 cells were seeded at 3 x 105 cells per 10 cm plate in DMEM supplemented with 10% NBCS. The following day the medium was replaced with 1 in 4 dilution of the retroviral supernatant for 4 h, after which cells were maintained in fresh medium. Approximately 12 h later, the 4 h infection was repeated. OPP-infected cells were examined 48 h later by fluorescence microscopy (Leica). Cells infected with retroviruses encoding Orb14 isoforms were either lysed for Western blotting analysis and/or set up for thymidine incorporation assays.

For retroviral infection of MCF-7/EcoR cells, cells were seeded at 2.5 x 105 cells/10 cm plate in 10 N RPMI. The next day, cells were infected with 1 in 4 dilution of the retroviral supernatant (generated as described earlier using Orb14Fia/pBabePuro). Infections were allowed to proceed for 24 h, after which the supernatant was replaced with fresh medium (ION RPMI). Approximately 48 h later, cells were selected with puromycin (Sigma, 0.7 µg/ml). Selection was allowed to proceed for approximately 6 days after which the resulting stable pools were maintained in ION RPMI supplemented with puromycin (0.5 µg/ml). Experiments utilising these cells were undertaken in the absence of puromycin.

58 2.3.5 Cell culture assays

For growth factor-stimulated binding studies, 1 x 106 COS-7 or CHO-IR cells were seeded per 10 cm plate (Falcon). The following day, cells were transfected using FuGENE™-6 reagent with the indicated construct(s) and incubated overnight. Subsequently, COS-7 and CHO-IR cells were serum-starved for 16-18 h in culture medium containing 0.5% FCS and in serum-free medium, respectively. COS-7 cells were stimulated for the indicated time period with either EGF (277 ng/ml), PDGF-BB (50 ng/ml), IGF-1 (100 nM) or aFGF (50 ng/ml). Vehicle was added to control cells for an identical time, and consisted of the solution utilised for growth factor reconstitution. CHO-IR cells were either left untreated or stimulated with insulin (10 µg/ml) for the indicated times. CHO-IR clones overexpressing Grb14 a (made by S. Malaney) were starved and stimulated with insulin in an identical manner to the CHO-IR cell line.

To investigate the effect of estrogen or antiestrogen treatment on Grb14 expression, 1 x 106 MCF-7 cells were seeded per 10 cm plate in ION RPMI. The next day cells were washed twice with phosphate-buffered saline (PBS) and the medium was replaced with phenol-red free RPMI 1640 supplemented with 10% charcoal-stripped FCS alone or with the addition of either estradiol (10 nM) or ICI 182780 (100 nM). Control samples were treated with ethanol vehicle only. After 3 d cells were harvested for protein analysis.

To examine Grb14 regulation under serum-free conditions, 6 x 105 MCF-7 cells were seeded in 10 cm plates in ION RPMI and allowed to proliferate for 3 d. Cells were then washed twice with PBS and the medium replaced with serum-free phenol red-free RPMI 1640 medium supplemented with transferrin (24 µg/ml) and gentamycin (10 µg/ml). Cells were maintained for 3 d in serum-free conditions with daily medium changes, then stimulated with insulin (10 µg/ml), estradiol (lnM) or the two in combination. Cells were harvested after 24 h for protein or flow cytometry.

For experiments utilising MCF7/8MT and MCF-7/8MT-c-Myc cell lines (made by C. M.

Sergio), 6 x 105 cells were seeded per 10 cm plate in 1ON RPMI. The following day, cells were arrested by adding ICI 182780 ( 10 nM) to the medium and maintained in the presence

59 of the antiestrogen for 2 d. The cells were then rescued with either estradiol (100 nM) or

ZnSO4 (65 µM), with or without the addition of nocodazole (50 ng/ml). Cells were subsequently harvested at different timepoints for protein analysis or flow cytometry.

In cell cycle progression studies employing MCF-7/EcoR cells infected with control or Grb14-encoding retrovirus, 6 x 105 cells were seeded per 10 cm plate or 75 cm2 flask in lON RPMI. After 3 d, cells were washed twice with PBS and then maintained in serum­ free conditions for 3 d as described. Cells were subsequently stimulated with the appropriate hormones for 24 h and harvested for flow cytometry.

2.3.6 Thymidine incorporation analysis NIH 3T3 fibroblasts were infected with retrovirus encoding the Grb14 isoforms as described earlier and seeded in 12 well plates at 6.5 x 104 cells/well in 10% serum­ containing medium. The following day, cells were serum-starved in 0.5% serum-containing medium for 48 h, after which the medium was supplemented with either 10% serum or various concentrations of the indicated growth factor(s) for approximately 18 h, followed by the addition of 0.5 µCi of 3H-thymidine (11.7 Ci/mmol, Amersham Biosciences) per well and subsequent incubation for 5 h. The medium was then removed by suction, and the cells were fixed in 10% TCA for 10 min prior to triplicate washings with cold PBS. Cells were then resuspended in 800 µl of 1% SDS and placed in scintillation vials with 6 ml of scintillation fluid per vial. To measure incorporation of the labelled thymidine, samples were subsequently analysed using a Beckman LS 6500 scintillation counter.

2.3. 7 Flow cytometry Cells were harvested by gentle scraping in PBS and pelleted by centrifugation for 3 min at lO00x g prior to resuspension in culture medium. To stain the DNA, ethidium bromide

(Pfizer, NSW, Australia) and Triton X-100 (Sigma) were added to 2-3xl05 cells to a final concentration of 12.5 µg/ml, and 0.2% (v/v), respectively. Cells were gently vortexed and incubated for at least 2 hat RT or at 4°C overnight. RNAse (0.4 mg/ml, Sigma) was mixed with the stained cells by brief vortexing and incubated for 2 h at RT prior to analysis. DNA flow cytometry was undertaken using a FACSCalibur laser-based flow cytometer (Becton Dickinson) and cell cycle phase distribution was analysed with the program Modfit (Becton

60 Dickinson). DNA histograms were typically obtained from 30,000 acquired events with coefficients of variation less than 5%.

61 Chayter 3-'ldenttftcation '!fa nuvef(jrb14 is'!form

generatedby afternative m'R'N9i. !J)ficing

3.1 Introduction

Alternative splicing involves approximately 30% of human genes (Tollervey and Caceres, 2000) and plays a vital role in regulation of gene expression during development and carcinogenesis (Stickeler et al., 1999; Tollervey and Caceres, 2000). RNA splicing is a complex mechanism occurring within a large assembly of ribonucleoproteins, the spliceosome, which consists of splicing factors including small nuclear ribonucleoproteins (snRNPs). Alternative splicing occurs when the basic splicing procedure is regulated, often incorporating changes in the comparative activity of splicing factors with opposing functions (Tollervey and Caceres, 2000).

The SR protein family are important splicing factors that recognise and bind to splice sites and enhancers, the latter being splicing accessory sequences (Manley and Tacke, 1996). The function of SR factors is modulated by phosphorylation of an arginine-serine rich (RS) domain in the C-terminus and their activity is antagonised by the pre-mRNA-binding heterogeneous nuclear ribonucleoproteins (hnRNPs) A/B. Changes in the levels of SR proteins affect the alternative splicing process (Zahler et al., 1993) and these can be induced by diverse stimuli. For example, stress induces the relocalisation of hnRP Al to the cytoplasm through phosphorylation by a p38 MAPK-dependent pathway and thus alters the ratio of nuclear hnRNP Al and SR proteins, resulting in changes in splice site selection. Also, adenovirus modulates splicing by altering the phosphorylation of SR splicing factors (Tollervey and Caceres, 2000).

All eukaryotic genes contain similar sequences at splice junctions that serve as targets for splicing factors. These consist of consensus sequences at both ends of the region to be excised, although not all exon/intron junctions exactly match these sequences. However, there are two invariant nucleotides that are essential for splicing, these consist of GT at the

62 5' donor splice site and AG at the 3' acceptor site. Mutations that alter splice sites can lead to potentially fatal diseases. A classic example is demonstrated by the a- and B-globin genes, where incorrect splicing results in inherited haemoglobin deficiency or thalassaemia (Senapathy et al., 1990; Moran et al., 1994).

Splice variants have been identified for a plethora of adaptor signalling molecules and they may have either similar or antagonistic functions (Fath et al., 1994; Migliaccio et al., 1997; Gamper et al., 2001). This is no exception for the Grb7 family. To date, six GrblO isoforms have been identified (mGrblO a and 8; hGrblO B, y, E and I;) that differ in their N-terminal or PH regions (Daly, 1998) (see section 1.3.3). For Grb7 a sole isoform was isolated from human esophageal carcinoma cells, termed Grb7 V, with the SH2 domain substituted by a C-terminal hydrophobic sequence (Tanaka et al., 1998b). No additional isoforms have been detected for Grbl4, however multiple transcripts on a Northern of human tissues screened with a Grbl4-specific cDNA probe were identified (Daly et al., 1996). This suggested that the existence of Grbl4 splice variants was plausible. Since the identification of additional isoforms provides valuable information about the protein under investigation, and assists in dissecting the signalling networks wherein it functions, this study initially focused on identifying novel Grbl4 isoforms.

3.2 Identification ofGrb14 isoforms by RT-PCR

To identify Grbl4 variants, 5 sets of oligonucleotides were designed (Gl-G5) that spanned the entire Grbl4 open reading frame and amplified overlapping mRNA regions to maximise the detection of Grb14 isoforms (Figure 3.lA). The Gl primer set amplified the cDNA encoding residues 1-137 of the Grb14 N-terminus, G2 flanked the sequence encoding amino acids 108-246 of the N-terrninus and the PH domain. G3 primers amplified the cDNA corresponding to residues 206-348 of the N-terminus, the entire PH domain and the BPS domain. G4 amplified the sequence encoding amino acids 305-442 of the PH domain, the entire BPS domain and the SH2 module. Finally, G5 primers amplified the region corresponding to residues 405-540 of the BPS domain and the complete SH2 domain sequence.

63 A N C Grb14 protein I PH lsPslsH2I 5' ~· GRB14 cDNA I I I I I I 0 400 800 1200 1600 2000 2400 b.p.

G1~ ~ G2..,. ~ G3~ ~ G4..,. ~ GS-+ ~ B

primer pairs G1 G1+ G2 G2+ G3 G3+ G4 G4+ GS GS+ M HEK 293 ------400 bp

C

184 clones M C1 .1 C1.6 C1.18 Grb14 500 bp - - 300 bp - ----

Figure 3.1 Identification of a naturally occurring Grbl4 transcript with a deletion in SH2 domain-encoding sequences. A. PCR strategy for identification of novel Grbl4 transcripts. Five sets of primers Gl-05 were designed to amplify fragments spanning the entire Grbl4 open reading frame . The primer sets amplified the following nucle­ otides of the Grbl4 cDNA: GI 541-951 (413 bp), 02 862-1278 (419 bp), G3 1156-1585 (429 bp), 04 1453-1866 (4 16 bp), GS 1751-2163 (415 bp). B. Detection of a novel Grbl4 transcript by RT-PCR. RT-PCR was undertaken using Gl-05 primer sets and total RNA from the human embryonic kidney HEK 293 cell line. Lanes Gl-05 denote RT-PCR reactions from HEK 293 cells, lanes Gl +_05+ represent positive control reactions using the original Grbl4 cDNA as template. (M molecular weight markers ; bp base pairs) (C) A subset of isolated cDNA clones cor­ respond to the novel variant. The GS primer set was used to amplify the sequences encoding the SH2 domain of 184 clones Cl.I, Cl.6, Cl.18 and the previously characterised Grbl4 isoform. Screening for alternative products was initiated using the human embryonic kidney HEK 293 cell line that expressed relatively high levels of Orb14 (Daly et al., 1996). For each primer set, a positive control was included utilising a plasmid containing the full-length Orb14 as a template (Ol+-05+). In all samples, the Orb14_primers amplified a band that migrated at a comparable size to the positive control, and was not detected in the -RT control (section 2.1.3), indicating that the product was specific (01-05). Additional specific PCR products were not detected in the primer sets O 1-04, suggesting the absence of Orb14 isoforms varying in the N-terminus and in the PH regions (Figure 3.1B). However, the 05 primer set, spanning the sequences encoding one third of the Orb14 BPS and the entire SH2 domain, amplified an additional band migrating approximately 100 bp faster than the main product. The lower molecular weight band was not detected in the control lane 05+. This indicated that there was an additional Orb14 transcript encoding a protein with a deletion in the BPS and/or SH2 domains in HEK 293 cells.

In order to extend the analysis of Orb14 isoforms, 12 Orb14 clones, isolated from a 184 human breast epithelial cell cDNA library (Daly et al., 1996), were screened by PCR to identify deletions similar to the transcript isolated by RT-PCR from HEK 293 cells. PCR was undertaken using the 05 primer set from plasmids containing these clones. Results indicated that the PCR product from 3 clones, C 1.1, C 1.6 and C 1.18 migrated at a lower molecular weight than the positive control amplified from a Orbl4 plasmid (Figure 3.lC). This suggested that these clones also contained a deletion in the C-terminal sequence.

3.3 Characterisation ofan alternatively-spliced Grb14 isoform. Grb14 f}

3.3.1 Characterisation of the Grbl4 C-terminal deletion In order to confirm the results suggesting that there was a Orb 14 transcript with a C­ terminal deletion and to map its location, the additional smaller PCR products were gel­ purified, subcloned and sequenced using PCR cycle sequencing (see section 2.1.6 and Appendix B). Results revealed that the smaller PCR products from HEK 293 cells and the 184 cDNA library clones contained an identical 94 bp deletion in the SH2 domain. The sequence coding for the BPS region was unaffected. This deletion extended from 1923 bp

65 to 2016 bp, thus removing the nucleotides coding for the FLVRDS motif (Figure 3.2A), that is required for the recognition of the phosphorylated tyrosine residue within the consensus binding sequence (Waksman et al., 1993). This suggested that the encoded SH2 domain was unable to bind interacting proteins. The deletion further introduced a premature stop codon that was expected to remove most of the Grb14 SH2 domain (Figure 3.2B). The novel truncated isoform was named Grb14 ~ to distinguish it from the full-length Grb14 a.

3.3.2 Expression of Grbl4 isoforms in breast cancer cell lines The expression profile of Grb14 ~transcripts was analysed using RT-PCR with primers specific for the truncated isoform. In order to do this, primers were designed that span the sequences immediately flanking the deletion, but not the deleted region itself (schematic, Figure 3.2C) (see section 2.1.3 and Appendix A). Therefore, these primers amplified only from Grb 14 ~. but not Grb 14 a transcripts. This revealed that Grb 14 ~ was expressed in 184 cells, as expected from the library screening results, and was also detected in T-47D and Hs578T breast cancer cells. However, it was not detected in BT-20, SK-BR-3 or BT- 549 breast cancer cell lines (Figure 3.2C). The specificity of the deletion-specific primer was confirmed by the absence of a product amplified from a plasmid encoding Grb14 a (Figure 3.2C, upper panel). Interestingly, Grb14 a expression was detected in both normal 184 cells and the entire subset of analysed breast cancer cell lines, indicating that the full­ length Grb14 transcript was more widely expressed among the panel of breast cancer cell lines, whereas Grb14 ~ displayed a more restricted expression profile. On this note, attempts have been made to identify the endogenous Grbl4 ~protein in HEK 293 and breast cancer cells. Western blotting using a Grbl4 antibody directed against the N­ terminus allowed the detection of a band that was migrating at approximately the predicted size for Grb14 ~- However, this could not be conclusively confirmed by immunoprecipitation analysis, because Grb14 ~ was masked by high background due to the antibody IgG heavy chain, as they both migrate on SDS-PAGE at a comparable size.

66 A

TGGTITCACCACAAAATITCTAGAGATGAGGCTCAGCGAITGAITAITCAGCAAGGACITGTGGATGGA 1923 W F H HK IS RD EA QR LI I Q.Q G L VD G

GITITCITGGTACGGGATAGTCAGAGTAACCCCAAAACITITCGTACTGTCAATGAGTCATGGACAAAAA 1992 VFLVR DS Q SN P KT FV LS NS HG Q K

ATAAAGCACITTCAAAITATACCAGTAGAAGATGACGCTGAAATGITCCACACACTGGATGATGGCCAC 2061 KHFQII PVEDDGEMFHTLDDGH

ACAAGATITACAGATCTAATACAGCTGGTGGAGITCTATCAACTCAATAAGGGCGITCITCCITGCAAGT 2130 TRFTDLIQ LVEFYQ LNKGVLPCK

TGAAACAITAITGTGCTAGGAITGCTCTC 2159 L KHYCAR I AL

B

TGGTTTCACCACAAAATTICTAGAGATGAGGCTCAGCGATTGATTATICAGC WFH HK IS RD EAQ R LI IQ

AAGGACTTGTGGATGGGTAGAAGATGACGCTGAAATGTICCACACACTGGA ... Q G LV DG *

Grb14 a I PH I BPS I SH2 I

Grb14 ~ I PH BPS 11 C

~ > G5DEL2 < i==J GSR

C') <:j I- 0) CX) .... 0 0 ci: ,-...... ,-... N co ~ a:i .... I- I- I- I- I- I- CX) ...... :- ~ '/;l ..:-"' a: ::;; q: ..:- q: co q: en q: :r: q: co q: u ('.) 2 00 bp- -Grb14 ~ -"'!"""" - --- . - - ..

C')

0 ci: N co 1- ..:. ~ 1- q: co en q: ----~· 400 bp - ., , ~Grb14 a .... -- .... -- ...• i

Figure 3.2 Characterisation of a novel Grbl4 isoform, Grbl4 ~. arising from alternative mRNA splicing. A. Sequence of the Grbl4 SH2 domain with the 94 base pairs deleted in Grbl4 ~ underlined. The FL VRDS binding motif is shown in bold. B. The SH2 domain sequence of Grbl4 ~- The deletion introduces a premature stop codon that results in truncation of the SH2 domain. A schematic represen­ tation of the domain structures of Grb 14 a and ~ is shown in the lower panel. The vertical bar denotes the conserved proline-rich motif. C. Detection of Grbl4 ~ transcripts via an isoform-specific PCR reaction. A Grbl4 ~ specific primer (G5DEL2) was designed to bind nucleotides immediately flank­ ing the deleted sequence, but not the deleted region itself. A schematic representation of the primer location is shown (upper panel). The deleted sequence is shaded black. Flanking nucleotides are rep­ resented by patterned shading. To amplify Grbl4 ~. G5DEL2 was used in conjunction with a reverse primer that binds sequences encoding the carboxy terminal end of the SH2 domain (GSR). RT-PCR was undertaken from total RNA of normal mammary epithelial cells (184) and breast cancer cell lines (T-47D, BT-20, SK-BR-3, Hs578T, BT-549) using the G5DEL2/G5R primer set. For each sample, a control reaction lacking reverse transcriptase (-RT) was included. The C 1.6 cDNA clone and the full­ length Grbl4 a plasmid were used as positive and negative controls, respectively. The expression of Grbl4 a was also investigated in this subset of cell lines using RT-PCR with the GS primer set described in Figure 3.1 A. 3.3.3 Alternative splicing of the Grb14 SH2 domain-encoding sequence To identify the origin of Grb14 f} transcripts, PCR primers were designed to amplify the boundaries flanking the SH2 deletion from genomic DNA of T47D cells (see section 2.1.1 ). Primers flanking the boundary N-terminal to the deletion (GEN 1) amplified a fragment of approximately 2.7 kb, while the 3' primer set (GEN 2) amplified a product approximately 1.4 kb in size (Figure 3.3A). These PCR products were gel-purified, subcloned and sequenced (section 2.1.4 and Appendix B). This revealed that the boundaries contained the GT/ AG consensus sequences for eukaryotic donor and acceptor splice junctions (Senapathy et al., 1990) (Figure 3.3B). This suggested that Grb14 f} resulted from alternative mRNA splicing, where the exon corresponding to the deleted sequence was included in the Grb 14 a, but not Grb14 f} transcripts. At the time of conducting these experiments, the entire genomic structure of Grb14 was not available. However, following the analysis of data provided by the human genome project, it became evident that the deleted Grb14 region corresponds to exon 11. Interestingly, these exon-intron junctions are conserved in GrblO and Grb7 (Figure 3.4A). The deleted Grb14 exon 11 corresponds to exon 21 or 14 in GrblO (Angrist et al., 1998) and to exon 14 of Grb7. However, the splice junctions for SH2-encoding sequences differ for other proteins. For instance, the rat Grb2/ Ash sequence coding for the SH2 domain (Figure 3.4B) is immediately preceded by an exon/intron junction (site i), and contains another splice site (site ii), following the conserved phosphotyrosine-binding motif (Watanabe et al., 1995). However, the nucleotides coding for Grb7 family SH2 domains (Figure 3.4A) are separated from a preceding splice junction (site i) by 7 amino acids, and harbour 2 additional splice junctions (sites ii and iii) at distinct sites within the SH2-encoding sequence.

3.4 Discussion

Several laboratories identified isoforms for Grb7 and -10. In addition, Northern analysis detected additional Grb14 transcripts that differ by approximately 100 bp from the main 2.4 kb transcript in human tissues including ovary, testis, kidney and liver (Daly et al., 1996). Thus, this investigation was initiated by searching for Grb14 splice variants using RT-PCR. A Grb 14 isoform, named Grb 14 f}, was identified that lacked an exon encoding a region of

69 A

2.8 kb _

1.51 kb_

B

SH2-spanning size (kb) Splice donor Splice acceptor intron A - 2.7 I II GTGGATGGgtaagtttggta ctctctctttagAGTTTTCT Ill IV B -1.4 ATT ATACCAgtaa~aattcg tttctttttgcagGT AGAAG

GRB14 genomic sequence II Ill

SH2 domain 3' UTR GRB14cDNA

V intron ~ .. ~ .. D exon GEN- 1 GEN- 2 • deleted exon

1700 1800 1900 2000 2100 2200 2300 2400bp

Figure 3 .3 Genomic origin of the Grbl4 ~ variant transcript. A. Sequences encoding boundaries flanking the deleted region were each amplified by PCR from genomic DNA of T-47D cell s. The PCR products were subcloned and sequenced. B. The sequence of the exon/i ntron junctions flanking the deletion are shown in the upper panel. Exon sequence 1s in uppercase, mtron in lowercase. I and III represent sequences containing splice donor sites, II and IV denote .those harbouring splice accep­ tor sites. Consensus sequences for splice donor and splice acceptor sites are highlighted in bold. A schematic outlining the Grbl4 a SH2 exon/intron splicing mechanism is shown in the lower panel. Exons are represented by boxes, introns (named A and B) by Imes. The exon deleted in Grbl4 ~ is shaded black. The nucleotide positions of the Grbl4 cDNA are also shown. A SH2 ... Grbl4 exon 10 432 A I H RS Q P F H H K I S R D E A Q R L I I Q QG L V D G461 rexon 20 1 Grb 10 Lex on 132 440 V l H R T Q H W F H G RI S R E E S H R I I K Q QG L V D G469 Grb7 exon 13 424 A I H R TQ L F H G RI S R E E S Q R L l G Q QG L V D G453

11 Grbl4 exon 11 462 V F L V R D SQ S N P K T F V L S M S H G Q K l K H F Q I I 491 rexon 21 1 GrblO Lexonl41 470 L F L L R D S Q S N P K A F V L T L C HH Q K I K N F Q I L 499 Grb7 exon 14 454 L F L V RE S Q R N P Q G F V L S L C HL Q K V K HY L I L 483

Ill Grbl4 exon 11/12 492 P V E D D GE M F H T L D D GH T R F T D L I Q L V E F Y Q521 rexon 21/221 GrblO Lexonl4/152 500 P C E D D GQ T F F S L D D GN T K F S D L I Q L V D F Y Q 529 Grb7 exon 14/15 484 P S E E E G R L Y F S M D D GQ T R F T D LL Q L V E F H Q 513

Grbl4 exon 12 522 L N K GV L P C K LK H Y C A R I A L 540 GrbIO rexon 221 Lexon 152 530 L N K G V L P C K L K HH C [!JR V A L 548 Grb7 exon 15 514 L N R G I L P C L L R H C C T R V A L5 32

B SH2 i • Grb2 exon 2/3 53 I EM KPH T FF GK IPR AK A E__ EM LS K QR H D G 81 II exon 3/4 82 A F L I R E S E S A P G D F S L S V l G N D V Q H F K V L 111

exon 4 112 RDGAGKYFLWVVKFNSLNELVDYH 135

Figure 3.4 Comparison of splice junctions for SH2 domains. A. An alignment of the amino acid sequences of human Grbl4, GrblO, and Grb7 SH2 domains hi ghli ghting conserved spli ce junction sites. Arrow corresponds to the beginning of SH2 domain sequence. Identical residues are in bold with dark shading. Si mil ar residues are shaded light. Amino acid residues are numbered. i, ii , a nd iii correspond to sites of exon-intron spli ce junctions, and exon numbers coding for the displayed protein residues are shown on the left. The conserved residues of the phosphobinding motif are underlined. Grbl4 genomic sequence was obtained from this study and from the human genome project at the website below. GrblO genomic sequence is from 1the human genome project as stated below and 2Angrist et al., 1998. Grb7 genomic sequence is from the human genome project: Grbl4 http:/www.ncbi.nlm.nih.gov/cgi-bin/Entrez/evv.cgi?contig=NT_03 1765. l &gene=GRB 14; 1Grb l O http:/www.ncbi.nlm .nih.gov/cgi-bin/Entrez/evv .cgi ?contig=NT_007722.8&gene=GRB l O; Grb7 h ttp:/www. ncbi .nlm. ni h.gov/cgi-bi n/Entrez/evv .cgi ?conti g=NT_O l 0685 .5 &gene=GRB 7 B. Amino acid sequence of the Grb2 SH2 domain (Watanabe et al., 1995). The arrow corresponds to the start of the SH2 sequence. Amino acid residues and exon numbers are shown. i and ii correspond to sites of splice junctions. The phosphobinding motif is underlined. the SH2 domain. Grbl4 Bwas detected in the HEK 293 cell line that relatively expressed high levels of Grbl4, and in 3 clones obtained by screening a 184 mammary epithelial cell library. Unlike GrblO isoforms which differ in the central and N-terminal regions, such as hGrblO B which has a PH deletion of 46 residues, as well as mGrblO a and 8, which contain distinct N-terminal insertions, no Grbl4 isoforms were identified which differed in the N-terminus and/or PH domain. However, although RT-PCR primers in this chapter were designed to amplify overlapping regions of the complete Grb 14 open reading frame, a more exhaustive search could have entailed the use of primers incorporating the 5' untranslated region to identify potential isoforms with alternative translational start sites and distinct N-termini.

Interestingly, the expression profile in a panel of normal mammary and breast cancer cells differed for Grb 14 a and B. While both isoforms were expressed in normal breast epithelial cells, Grbl4 Bwas expressed in a subset of breast cancer cell lines (Figure 3.2C). Although the functional significance of this is unclear at this stage, it implies that Grbl4 B expression is independently regulated and possibly fulfils a distinct role from Grbl4 a. Similarly, Grb7 isoforms exhibit relative differences in their expression profile. Grb7V displays a more limited expression profile in esophageal tumours relative to Grb7, and is detected in 40% of the tumours that express the full-length transcript. This suggests that Grb7V expression is distinctly regulated relative to the full-length Grb7. Furthermore, while Grb7 is tyrosine-phosphorylated in EGF-stimulated esophageal carcinoma cells but not in serum-starved cells, Grb7V displays constitutive phosphorylation that is unresponsive to EGF treatment (Tanaka et al., 1998b). Although Grbl4 does not undergo tyrosine phosphorylation, it exhibits basal phosphorylation on serine residues which is increased in response to PDGF-BB stimulation. Thus, an important question would be to elucidate whether the phosphorylation of the truncated variant is regulated differently in response to growth factor treatment.

Changes in expression of splice variants depends on the relative ratio of splicing factors and can be tissue- or cell-type dependent (Tollervey and Caceres, 2000). To ascertain whether the expression of Grb 14 isoforms is tissue-specific, future studies may entail screening a

72 large tissue panel from the normal body atlas using either quantitative or semiquantitative RT-PCR, particularly since differential expression of GrblO isoforms was detected in insulin-sensitive tissues (adipocytes, liver and muscle) (Liu and Roth, 1995; Dong et al., 1997). In this study, library screening indicated that the expression of the truncated variant is less than that of the full-length Grb14 a (section 3.2), therefore screening additional tissues/cell lines may be necessary to find where Grb14 ~ is endogenously expressed at high levels, in order to identify the cellular background wherein it functions.

RNA splicing is further affected by signal transduction pathways that alter the compartmentalisation of splicing factors (Tollervey and Caceres, 2000). Therefore, differences in the expression profile of Grb 14 isoforms may also be apparent in cells treated with various growth factors, and may reflect differences in the activation state of signal transduction pathways in breast cancer cells. Furthermore, studies have reported that many insulin signalling components and their isoforms are significantly upregulated during adipocyte differentiation. This includes the regulatory subunit of PI-3 kinase p85a and its splice variant p50, in addition to PKB a and~ (Saad et al., 1994; Hill et al., 1999; Pederson and Rondinone, 2000). Future studies may be aimed at examining the expression of Grb14 isoforms using RT-PCR in differentiating cells since Grb14 expression increases in differentiated adipocytes (Kasus-Jacobi et al., 1998), and may be subject to alternative splicing events. This can also be extended to the use of the mouse mammary HCl 1 cell line, which provides an effective system for the study of mammary cell differentiation in vitro. This cell line undergoes differentiation and production of milk proteins in response to the lactogenic hormones dexamethasone, insulin and prolactin (Merlo et al., 1996). It thus provides a useful setting to study the expression and function of Grb14 in these processes.

It was interesting to find that an isoform equivalent to Grb14 ~ has not been identified for Grb7 and -10, despite the fact that the splice junctions in the sequence encoding the SH2 domain are conserved in all three Grb7 family genes (Figure 3.4A). Grbl0 isoforms isolated to date all have intact SH2 domains (Laviola et al., 1997; Daly, 1998). Grb7V has an 88 bp deletion in the SH2 domain which results in substitution with a hydrophobic sequence that has 50% identity to the homeobox protein HoxB4 C-terminus (Tanaka et al.,

73 1998b). The consensus sequence for the donor splicing site in Grb7V occurs prior to the region coding for the SH2 domain (at site i, Figure 3.4A), and the acceptor site occurs within the SH2-encoding region (at site ii, Figure 3.4A). Thus Grb7V completely lacks an SH2 domain, which is substituted by a distinct sequence due to the frameshift introduced by the alternative splice sites. However, Grb14 splicing occurs within SH2 domain­ encoding sequences (at sites ii and iii, Figure 3.4A), thus retaining approximately one quarter of the SH2 domain followed by a premature stop codon (Figure 3.2 B). Since this abrogates the SH2 consensus binding sequence, it is expected that this truncated domain is non-functional, therefore restricting interactions with either phosphorylated receptors or downstream effectors. The functional consequences of the SH2 domain truncation in Grb14

~ are examined in subsequent chapters.

74 ChOjlter 4-'lnteraction ef grb14 iso_forms with 'R'T'Ks

4.1 Introduction Identification of binding partners in cellular signalling provides vital clues to understand protein function. Grb7 family members interact with a plethora of tyrosine kinases (see section 1.3.4). However, there are in vivo binding partners shared by particular Grb7 family proteins: the interaction with the IR occurs with all three members (Liu and Roth, 1995; Frantz et al., 1997; Kasus-Jacobi et al., 1998; Kasus-Jacobi et al., 2000), and both Grb7/10 bind to the PDGFR (Yokote et al., 1996; Wang et al., 1999). GrblO interactions involve the C-terminal BPS-SH2 region that remains intact in all isoforms (Daly, 1998; Han et al., 2001), hence all GrblO proteins are expected to have the same RTK binding partners. On the other hand, the SH2-lacking Grb7V (Tanaka et al., 1998b) is predicted to have a distinct RTK recruitment profile relative to the full-length Grb7, although binding studies with both Grb7 isoforms have not yet been undertaken.

Grb14 is the most widely expressed protein in humans when compared to the remaining Grb7 family members (Daly et al., 1996). Hence, it is intriguing to find that its binding profile is the most limited. The only known in vivo interacting partners are the IR and FGFR-1, identified during the course of this study and by other groups (Kasus-J acobi et al., 1998; Reilly et al., 2000; Hemming et al., 2001). This implies that Grb14 a is acting in distinct receptor-initiated signals, and so it is of interest to ascertain whether the truncated isoform is involved in the same signalling pathways as the full-length protein. Further, it is structurally informative to examine whether the truncated SH2 domain of Grb14 ~ will restrict interactions with specific RTKs. It is anticipated that a dysfunctional SH2 domain will not allow interactions that are primarily SH2-mediated. However, those that involve the BPS domain, which remains intact in both isoforms, should involve both Grb14 proteins. To investigate this, we examined the interaction of the Grb14 splice variants with several receptor tyrosine kinases including the IR and FGFR-1.

75 4.2 Generation ofa Grb14 [}expression vector and Grb14 GST fusion proteins In order to examine interactions of Grb14 isoforms with specific RTKs, a eukaryotic expression vector encoding Grb 14 B was generated by PCR and subcloned in a plasmid containing a C-terminal Flag-epitope tag (pRcCMV Flag) (see section 2.1.4). In addition, vectors encoding a series of fusion proteins corresponding to different regions of Grb 14 (Figure 4.1 A) were generated by subcloning PCR-derived DNA fragments into pGEX-2T. All constructs were sequenced to ensure they were free of mutations and inserted in the correct orientation. The Grb14 fusion protein constructs were transformed into DH5 a cells, and the encoded proteins purified by binding to glutathione-agarose beads and analysed by SDS-PAGE (Figure 4.18). This revealed that all fusion proteins were successfully expressed and migrated at approximately the predicted sizes (see Figure 4.18). Conditions were optimised to ensure that degradation products, seen as faster migrating bands in the BPS and BBPS-SH2 preparations, were minimal.

4.3 Interaction of Grb14 isoforms with the EGFR and PDGFR In order to examine interactions of Grb14 isoforms with specific RTKs, co­ immunoprecipitation analysis was undertaken using COS-7 cells transiently transfected with either Flag-tagged Grb14 isoform and either a control vector or the receptor of interest. Grb14 proteins were successfully expressed in these cells, with the truncated isoform running at approximately 50 kDa according to its predicted size, whereas the full­ length Grb14 migrated at 56 kDa (Daly et al., 1996) (Figure 4.2A, left panel).

Since Grb14 was originally isolated through CORT screening with the phosphorylated C­ terminus of the EGFR (Daly et al., 1996), the interaction of Grb14 isoforms with the EGFR was first analysed. Following transient transfection, COS-7 cells were serum-starved for 16-18 h in medium containing 0.5% serum, treated with control vehicle (-) or stimulated with EGF for 5 min (+ ). Western blotting of transfected lysates with a phosphotyrosine antibody (PY20) allowed the detection of a band migrating at a size corresponding to the EGFR (170 kDa).

76 A GST FUSION PROTEINS GST D

Grb14 alp

360 438 BPS I D SH2 Grb14 a [ II426 540

360 540 OJ Grb14 p 426-461

360 461

B

34.9 -

Figure 4.1 Generation of GST fusion proteins for pulldown analysis. A. Schematic showing the fusion proteins corresponding to the common Grb1 4 BPS, Grbl4 a SH2 and BPS-SH2, Grbl4 p SH2 and BPS-SH2 domains. Numbers below indicate the amino acid positions for each domain. B. GST fusion proteins were expressed and purified as described in section 2.2.5, separated by SDS-PAGE and visualised with Gelcode blue stain reagent. The estimated sizes for the fusion proteins are as follows: The GST moiety, 27.5 kDa; a SH2, 39.5 kDa; p SH2, 29.9 kDa; BPS, 35.9 kDa; a BPS­ SH2, 47.9 kDa; p BPS-SH2, 38.3 kDa. A

V~GFR w EGFR ~ EGFR V/EGFR w EGFR ~/EGFR kDa I - I - +1 + I _ + I kDa

EGFR _. 180- ______.l-180 I

Grb14 a_. Grb1 4 ~ _. 11 .... 1~:: 5 Lysates IP: Flag blot upper: PY20 blot upper: PY20 blot lower: Flag blot lower: Flag

B

V/PDG FR a/PDGFR ~/PDGFR V/PDGFR a/PDGFR ~/PDGFR kDa I - + I I _ + ' + I - + I I _ + ' + kDa PDG FR _. 180 -1 1-180 • 11 Grb1 4 a _. 58 -1 1-58 Grb1 4 ~ _. -- 48 5 - 11 - 48.5 Lysates -- IP: Flag --- blot upper: PY20 blot upper: PY20 blot lower: Flag blot lower: Flag

Figure 4.2 Interaction of Grbl4 isoforms with the EGFR and PDGFR. A. COS-7 cells were tran­ siently transfected with an EGFR expression vector together with a control vector (V/EGFR) or a vector encoding Flag-tagged Grbl 4 isoforms (a/EGFR; ~/EGFR). Transfected cells were starved then stimulated with EGF (+) or vehicle(-) for 5 min prior to harvesting. Grbl4 isoforms were immunoprecipitated with M2 Flag antibody. Lysates (left panel) as well as immunoprecipitations (right panel) were separated by SDS-PAGE and transferred to nitrocellulose. Membranes were immunoblotted with a phosphotyrosine antibody or with D-8 Flag antibody. Receptor expression in transfected cells relative to untransfected controls was verified using Western blotting, and receptor size was comparable to that observed when blotting for tyrosine phosphorylation. B. COS-7 cells were transiently transfected with a PDGFR expression vector together with a control vector (V/PDGFR) or a vector encoding Flag-tagged Grbl4 isoforms (a/PDGFR; ~/PDGFR). Cells were starved and stimulated with PDGFBB (+) or vehicle(-)_for 5 min prior to harvesting. Immunoprecip­ itations and immunoblotting were undertaken as stated m (A). This indicated that the EGFR displayed basal tyrosine phosphorylation in serum-starved unstimulated cells for all samples, which was increased upon addition of the EGF ligand (Figure 4.2 A, left panel). Incubation of lysates from transfected cells with Flag antibodies immunoprecipitated Grb14 a or ~ from both unstimulated and EGF-treated samples (a/EGFR or ~/EGFR, respectively), but not from cells transfected with the control vector (V/EGFR). However, the EGFR was not co-immunoprecipitated with Grb14 a as detected by anti-phosphotyrosine blotting, which suggests that Grb14 a does not interact with the EGFR. Similarly, receptor association was not detected with the truncated isoform Grb14 ~-

Since the Grb14 a SH2 domain associates with endogenous PDGFR of the ~ type (~­ PDGFR) in vitro, and PDGF-BB stimulation enhances Grb14 phosphorylation on serine residues (Daly et al., 1996), the interaction of Grb14 proteins with the ~-PDGFR was investigated. Following transfection, cells were serum-starved as described earlier and treated with control vehicle or stimulated with PDGF-BB for 5 min. Western blotting of lysates from transfected cells revealed a main immunoreactive band at 185 kDa, a size consistent with the ~-PDGFR. This demonstrated that there was basal receptor activation for all samples, which increased upon PDGF-BB stimulation (Figure 4.2B, left panel). Immunoprecipitation studies with Flag antibodies precipitated Grb14 isoforms from corresponding samples (a and ~/PDGFR). However, co-immunoprecipitation of activated PDGFR was not detected. Further, the truncated isoform did not co-immunoprecipitate the phosphorylated receptor, indicating that neither Grb14 isoform associates with the ~­ PDGFR.

4.4 Interaction ofGrb14 isoforms with the IGF-IR and IR

Several laboratories have indicated that GrblO isoforms, namely hGrblO ~. y, ~ bind the IR (Liu and Roth, 1995; Dong et al., 1997; Frantz et al., 1997 ) and Jacobi et al. (1998) demonstrated that rat Grb14 interacts with the IR, thus this interaction was examined with the human Grb14 isoforms. In addition, association with the IGF-IR was tested since it is structurally similar to the IR as both represent class II receptors composed of disulfide­ linked heterotetrameric units (Rechler and Nissley, 1986; Adamo et al., 1992; Cheatham

79 and Kahn, 1995; Siddle et al., 2001). The Grbl4/IGF-1R interactions were investigated by co-immunoprecipitation analysis in COS-7 cells. Cells were transfected with the IGF-IR and either Grb14 isoform or an empty vector, then treated with either control vehicle or IGF-I for 10 min. Phosphotyrosine blotting of lysates from transfected cells revealed an immunoreactive band migrating at 90 kDa, which is corresponding to the predicted size of the IGF-IR. IGF-IR phosphorylation significantly increased upon ligand stimulation, indicating that the receptor was activated (Figure 4.3A, upper panel). Grb14 isoforms were immunoprecipitated by Flag antibodies from cell lysates (Figure 4.3A, lower panel), however the activated IGF-IR was not detected in the Grb14 immunoprecipitates. This demonstrated the absence of an interaction between Grb14 proteins and the IGF-IR.

The interaction of the IR with Grb14 proteins was subsequently examined in chinese hamster ovary (CHO) cells overexpressing the IR (CHO-IR) since this system has been previously utilised for studying Grb7 family interactions with the IR (Dong et al., 1997; Kasus-Jacobi et al., 1998; Kasus-Jacobi et al., 2000). Cells were transiently transfected with pRcCMV Flag containing either Grb 14 a, Bor with an empty control vector (V) then serum-starved and either left untreated or stimulated with insulin for 10 min. Immunoblotting of lysates prepared from the transfected cells allowed the detection of the activated IR, migrating at 95 kDa. Receptor phosphorylation was noted upon ligand stimulation ( +) but not in the unstimulated samples (-) (Figure 4.3B, upper panel). Immunoprecipitation of epitope-tagged Grb14 proteins was undertaken and interacting proteins were detected by Western blotting. In this case, the phosphorylated IR was co­ immunoprecipitated by both isoforms upon stimulation with insulin, but not in untreated samples (Figure 4.3B, middle panel). This indicated that Grb14 a and Binteract with the IR and the association is dependent on receptor activation. Interestingly, Grb14 B co­ immunoprecipitated less phosphorylated IR relative to Grb14 a. To confirm this, the membrane was re-blotted for the IR (Figure 4.3B, lower panel) which reflected the results obtained with the phosphorylated receptor, indicating that Grb14 Bbound a lower amount of IR relative to Grb14 a. These findings demonstrated that the interaction of Grb14 proteins with the IR was not SH2-dependent, since an isoform with a truncated SH2 domain retained the ability to bind the receptor. However, the Grb14 BIIR interaction was

80 kDa V/IGF-IR p!IGF-IR a/lGF-IR A I 116 _ + + 116 - ...--- IGF-IR .. I ._._J ~ U . J .. IGF-IR

Grb1413-. 48.5-L.I______--J 58 -1..... _____ I• Grb14 a Lysates blot upper: PY20 blot lower: Flag

kDa V/IGF-IR !3/IGF-IR a/lGF-IR 116 + + 116 + I

Grb1413-. 48.5-1 ...., 58-,..... , ... Grb14 a

IP: Flag blot upper: PY20 blot lower: Flag

B V Grb14 a Grb14 ~ kDa + + + IR~ 84-

Grb14 a ~ 58-1 -- FI Grb14 ~ ~ 48.5-L______,_ Lysates blot upper: PY20 blot lower: Flag

V Grb14 a Grb14 ~ kDa + + +

I 11 6 -I;:======~======~.. .., 7 7 r ? I .::=1L------==-----'IP: Flag blot upper: PY20 blot lower: antiflag

V Grb14 a Grb14 ~ kDa + + + 11 6-

IP: Flag re-blot: IR N N :::c :::c C (/) (/) a: I I N N (/) (/) 0 I- (/) :::c :::c a. a. ':::c (/) a. (/) (/) m m u C, m tl en. tl en.

IA__,. - insulin

+ insulin

Blot: IR

Figure 4.3 Interaction of Grbl4 isoforms with the IGF-IR and IR. A. COS-7 cells were transiently transfected with an IGF-IR expression vector together with a control vector (V/IGF-IR) or a vector encoding Flag-tagged Grbl4 isoforms (a/IGF-IR; ~/IGF-IR). Transfected cells were serum-starved then stimulated with IGF-I ( +) or vehicle (-) for 10 min prior to harvesting. Immunoprecipitations and immunoblotting were performed as described in Figure 4.2. B. CHO-IR cells were transiently transfected with an empty control vector or a vector encoding Flag­ tagged Grbl4 isoforms.Transfected cells were serum-starved then either left untreated or stimulated with insulin for 10 min prior to harvesting. Immunoprecipitations and immunoblotting were undertak­ en as in Figure 4.2. Membranes were re-blotted for the IR. C. Pulldowns were undertaken from untreated (-) or insulin-stimulated ( +) CHO-IR cells using GST fusion proteins schematically repre­ sented in Figure 4.1 A. Membranes were probed with an IR antibody. CHO-IR lysates were included as a blotting control. The additional lower band in the BPS lane independent of insulin treatment was not consistently observed in our experiments. reduced relative to the full-length protein, which suggested that the SH2 domain may contribute to the interaction with the IR or the truncated SH2 may inhibit receptor binding.

To investigate the regions mediating the interaction with the IR, pulldown assays were undertaken using the GST fusion proteins corresponding to various domains of the Grb 14 protein (shown in Figure 4.lA). CHO-IR cells were serum-starved and left untreated or stimulated with insulin for 10 min. Lysates were then prepared and incubated with the GST fusion proteins, prior to Western blotting analysis. The IR was detected in lysates incubated with fusion proteins corresponding to the BPS, Grb14 a and 13 BPS-SH2 domains from insulin-stimulated cells, but not from untreated samples (Figure 4.3C). However, the IR was not precipitated from lysates incubated with the Grb14 a and 13 SH2 domain from untreated or stimulated samples. This indicated that the IR interaction with Grb14 required the BPS but not the SH2 domain. Furthermore, the BPS, Grb 14 a and 13 BPS-SH2 regions precipitated equivalent amounts of receptor, which demonstrated that binding to the IR is predominantly mediated by the BPS domain. This is in contrast to the co­ immunoprecipitation studies, where the SH2-truncated Grb14 l3 bound the IR less well than the full-length protein (Figure 4.3B).

4.5 Kinetics ofIR interaction with Grbl4 isoforms

In order to further characterise the Grb14/IR interaction, the kinetics of the association were examined using CHO-IR cells stably overexpressing Flag epitope-tagged Grb14 a. Cells were serum-starved, left untreated or stimulated with insulin for various timepoints, then lysed and incubated with Flag antibody to precipitate Grb14. Western blotting analysis (Figure 4.4A) demonstrated that the IR is not co-immunoprecipitated in unstimulated samples, which is in accordance with previous results (Figure 4.3B). The maximum amount of IR was detected in the Grb14 precipitates at 1 min of insulin treatment and decreased with time. This indicated that the Grb14/IR interaction was maximal at 1 min, after which the complex dissociated until by 60 min there was little binding (Figure 4.4A). Western blotting analysis of Grb14-overexpressing CHO-IR lysates demonstrated that Grb 14 and IR expression was constant throughout the timecourse. Thus, the reduced

83 levels of co-immunoprecipitated IR suggests that this results from a decreased IR/Grb14 interaction and not from alterations in their expression levels. Further, analysis of IR phosphorylation revealed that the receptor was continuously activated throughout the timecourse (Figure 4.4B). Results were then quantitated by densitometry, normalised for loading (as indicated by B-actin blotting), and the dissociation of the Grb14/IR complex was directly compared with receptor phosphorylation (Figure 4.4C). It was then evident that although there was a small decrease in IR phosphorylation over the timecourse, the reduction in the Grbl4/IR interaction was significantly greater. This suggested that the dissociation of the Grb 14/IR complex was not solely dependent on receptor phosphorylation but was regulated by an additional mechanism. This result was similar for at least two different Grb14 a-overexpressing clones.

The finding that Grb14 a associated transiently with the IR raised the possibility that the contrasting association of Grb14 isoforms with the IR at 10 min of insulin treatment (Figure 4.3B, lower panel) may reflect different dissociation kinetics, rather than a lower binding affinity of Grb14 Bfor the receptor. Therefore, CHO-IR cells were transiently transfected with Grb14 proteins, serum-starved and stimulated with insulin for several timepoints, after which the Grb14 variants were precipitated with anti-Flag sera. Western blotting analysis subsequently revealed that the interaction with the IR was maximal at 5 min and declined thereafter for both isoforms (Figure 4.5). However, at 10 min of insulin stimulation, there was an approximate 20% decrease in the amount of IR co-precipitated by Grbl4 a, but an approximate 50% reduction in IR levels co-precipitated by the truncated variant. This suggested that the kinetics of interaction with the IR were different for the Grb14 isoforms, where the IR/Grb14 Bcomplex dissociated more rapidly than IR/Grb14 a. The enhanced dissociation of Grb14 B with the IR relative to Grb14 a between 5 and 10 min of insulin stimulation was observed in at least two independent experiments. Therefore, in conjunction with previous co-immunoprecipitation analysis (Figure 4.3B), these results indicated that Grb14 Bbinds less strongly to the IR and dissociates more rapidly relative to the full-length protein.

84 A

+insulin (min) 0 1 2 3 5 10 15 60

IR I

Grb14 I IP: Flag Blot upper: IR Blot lower: Flag

B

+insulin (min) 0 1 2 3 5 10 15 60

IR 1-~-~~--~1 IR

Grb14

Blot upper: IR Blot middle : PY20 Blot lower: Flag C

100 - IR PY20

$ IR

75

E ::::J E ·x 50 C'a E

0~

25

0 0 20 40 60

Time (min)

Figure 4.4 Kinetics of Grb14 a interaction with the IR. A. CHO-IR cells stably overexpressing Flag­ tagged Grbl4 a were serum-starved then left untreated (0) or stimulated with insulin for the indicat­ ed times prior to harvesting. Grb14 isoforms were immunoprecipitated with M2 Flag antibody. Fol­ lowing SOS-PAGE, membranes were immunoblotted for the IR and the Flag epitope. B. Lysates from the experiment in (A) were analysed by Western blotting for IR expression (upper panel), IR tyrosine phosphorylation (middle panel) and Grbl4 expression (lower panel). C. The amount of bound IR was quantified by densitometry, divided by the amount of precipitated Grb14 and repre­ sented as a percentage of the maximum value. Results were similar for at least two independent clones used in these experiments. A 2.5

er: 2 "O Q) § ·c. ~en I.5 Q..:4;:: +insulin (min) ,C 0 :, 0 1 3 5 10 15 30 .::c IR ~:, -€-~ 0E~ l1l I l1l Q) -~ 0.5

er: 2 "O Q) § ·c. 1.5 -~en +insulin (min) C..,c:=: 0 :, 0 1 3 5 10 15 30 () >, 0 "§ c-~ :, -€ IR E~0 l1l I l1l -Q)~ 0.5

Figure 4.5 Interaction kinetics of Grbl4 isoforms and the IR. A. Parental CHO-IR cells were tran­ siently transfected with Grbl4 a and stimulated with insulin for the indicated times. Immunoprecipita­ tions were performed and immunoblotted as in Figure 4.4A. Results were quantified by densitometry and normalised for loading as described in Figure 4.4C. B. Parental CHO-IR cells were transiently transfected with Grbl4 ~ and stimulated with insulin for the indicated times. Immunoprecipitations and immunoblotting were undertaken as in Figure 4.4A. The data was quantified by densitometry and normalised for loading as in Figure 4.4C. Note that the IR blots displayed in A and B are not from an identical signal exposure time. 4.6 Interaction of Grb14 isoforms with the FGFR-1

FGFR-1, also termed fig, represents a subclass IV receptor (section 1.1.1) and contains a 'split' kinase domain similar to the PDGFR (Ullrich and Schlessinger, 1990; Mason, 1994 ; Burke et al., 1998). Association of the Grb14 splice variants with the FGFR-1 was investigated using co-immunoprecipitation analysis in COS-7 cells that were co-transfected with FGFR-1 and either Grbl4 a/~ or a control vector. Cells were serum-starved, stimulated with vehicle or aFGF (FGF-1) for 5 min, then lysed and analysed by Western blotting. The activated FGFR-1 is detected as bands of approximately 120 and 150 kDa due to glycosylation. Basal receptor phosphorylation was evident in unstimulated lysates, and increased upon ligand addition (Figure 4.6A). Epitope-tagged Grbl4 proteins were precipitated with Flag antibodies from lysates of transfected cells, and interacting proteins were detected by immunoblotting. Results revealed that the 120 kDa form of the activated FGFR-1 was specifically co-immunoprecipitated by Grb14 a in lysates from FGF-treated cells but not from unstimulated cells (Figure 4.6B). This was also supported by densitometry analysis after normalising for the amount of precipitated Grb 14 a. Therefore, results indicated that Grb14 a interacted with the FGFR-1, and the association was dependent on ligand-induced activation of the receptor, similar to the interaction with the

IR. However, there was no specific co-precipitation of FGFR-1 with Grb14 ~ in either untreated or stimulated samples, which indicated that the SH2-truncated variant was unable to bind the receptor. Therefore, these results suggested that unlike the interaction with the IR which involved both Grb14 isoforms, the Grb14/FGFR-1 association required predominantly the SH2 domain since Grb14 ~ failed to bind.

This was further confirmed with GST pulldown analysis, where lysates from transfected cells were incubated with fusion proteins corresponding to various Grb14 domains, and subsequently analysed by Western blotting. Basal FGFR-1 precipitation was detected for the Grb14 a SH2 (after a longer exposure) and BPS-SH2 regions in unstimulated cells (Figure 4.6C), which was likely to occur due to basal receptor activation. Increased association with these regions was observed in lysates from aFGF-stimulated cells. When results were corrected for amounts of fusion proteins utilised and subjected to densitometry

88 A kDa V/FGFR-1 a/FGFR-1 f3 /FGFR-1 + + + FGFR-1 _. I Grb14 a_. -- Grb14 f3_. --- -I Lysates blot upper: PY20 blot lower: Flag B

V/FGFR- a/FGFR-1 f3 /FGFR-1 kDa + + + 180-1 FGFA-1- 116 -

Grb14 a_. 58-1 Grb14 f3 _. ---

IP: Flag blot upper: PY20 - blot lower: Flag .... I N N ~ C ::c: ::c: ~ rJJ rJJ c.:, I I N N rJJ rJJ ::c: ~ ~ ~ E-- rJJ ::c: ('1 rJJ ~ rJJ rJJ i:Q i:Q 0 kDa c.:, i:Q 1j ea. 1j ea. u 116-1 -aFGF 116-1 +aFGF Blot: FGFR-1

Figure 4.6 Interaction of Grbl4 isoforms with FGFR-1. A. COS-7 cells were transiently transfected with a FGFR-1 expression vector together with a control vector or a vector encoding Flag-tagged Grbl4 isoforms. Cells were serum-starved then stimulated with aFGF (+) or vehicle(-) for 5 min. B. Grbl4 isoforms were immunoprecipitated with M2 Flag antibody and membranes were Western blot­ ted with PY20 or Flag antibodies. C. GST-pulldowns were performed as in Figure 4.3 C from COS-7 cells transiently transfected with a FGFR-1 expression plasmid and then treated with aFGF as described in (A). Membranes were probed with a FGFR-1 antibody. Total cell lysate was included as a positive control for blotting. analysis, this revealed that the Grb14 a SH2 and BPS-SH2 domains precipitated equivalent amounts of the receptor. On the other hand, the BPS and Grb14 rJ SH2/BPS-SH2 domains did not precipitate the FGFR-1 in untreated or aFGF-stimulated cells. Thus, these findings supported the co-immunoprecipitation data as they indicated that the Grb14 interaction with the FGFR-1 was SH2-dependent, and did not require the BPS region.

4. 7 Discussion

Investigations of the interaction of Grb14 isoforms with RTKs revealed that they did not associate with either the EGFR or PDGFR, although Grb14 was originally isolated from a breast epithelial cell cDNA library using the phosphorylated EGFR C-terminus (CORT method) (Skolnik et al., 1991; Daly et al., 1996). Interestingly, although all three Grb7 family proteins were isolated using CORT screening, in viva interactions with the EGFR have not been demonstrated thus far (Margolis et al., 1992; Ooi et al., 1995).

Further, co-immunoprecipitations and GST pulldown assays indicated that both Grb14 isoforms bound the IR, and this interaction was primarily mediated by the BPS domain, as demonstrated for the rat Grb14 (rGrb14) (Kasus-Jacobi et al., 1998). It was interesting to note that loss of a functional SH2 domain in Grb14 rJ markedly reduced IR binding in viva (Figure 4.3B), which implied that the SH2 domain was required for an efficient interaction with the IR. However, in GST pulldown assays the Grb14 a GST-SH2 fusion protein did not bind the IR and the Grb14 a and rJ BPS-SH2 fusion proteins bound similarly to the IR. This suggested that the SH2 domain might have a stabilising effect on the interaction of the full-length protein with the receptor, which was also implicated for the rGrb14 SH2 module in the yeast two-hybrid system (Kasus-Jacobi et al., 1998). Alternatively, it was possible for the truncated SH2 domain in Grb 14 rJ to adopt a structure that interfered with binding of full-length Grb14 to the IR.

Upon autophosphorylation the IR kinase loop undergoes a conformational change, thus allowing access for substrates and A TP to the active site (Hubbard, 1997). The Grb 10 and -14 BPS module binds to an altered conformation of the IR kinase domain that requires

90 phosphorylation of Tyr-1150/1151 within the activation loop (He et al., 1998; Kasus-Jacobi et al., 1998; Stein et al., 2001). The BPS interaction allows direct inhibition of IR catalytic activity by Grb7 family proteins, with Grb14 being most potent at specifically decreasing IR kinase activity (Bereziat et al., 2002). This is in contrast to another adaptor, SH2-By, which interacts with the activated IR kinase region but does not alter substrate phosphorylation in vitro (Nelms et al., 1999; Bereziat et al., 2002). The BPS domain may hinder receptor activity by modulating conformational changes in the activated receptor and blocking the transfer of the phosphate group to the substrate (Bereziat et al., 2002). The exact mechanism is currently unknown, although future crystallographic studies are essential for elucidating the mechanism for BPS interaction and inhibition of IR activity.

It was interesting to find that the interaction of Grb14 proteins with the IR was rapid and transient, with maximum binding obtained at 1 and 5 min with subsequent dissociation at 2 and 10 min for cells stably overexpressing Grb14 a (Figure 4.4A), and in the transient transfections (Figure 4.5A) respectively (the time discrepancy was likely to reflect differences in relative Grb14 overexpression in the two systems). This was similar to results demonstrated by Hemming et al. (2001), where the Grb14 a/IR interaction occurred at 1 min, was maximal at 2 min, and decreased from 5-30 min (Hemming et al., 2001 ). However, in another study the Grb14 a/IR association also occurred by 1 min but the complex failed to dissociate, and the interaction persisted for 90 min (Bereziat et al., 2002). It was difficult to reconcile these results, since all three studies were undertaken with CHO­ IR cells overexpressing Grb14 a, although different insulin concentrations were utilised for stimulation. The highest insulin dose of 1.8 µM was employed in the experiments described herein, relative to 100 nM for Bereziat et al. (2002) and 10 nM for Hemming et al. (2001). Since dissociation of the Grb14/IR complex was observed at both 1.8 µMand 10 nM of insulin stimulation, it was unlikely that the insulin dose contributed to the observed discrepancy. However, the degree of Grb14 overexpression in cells might account, at least in part, for the observed differences particularly since an identical eukaryotic expression vector (pRcCMV FLAG) was utilised to overexpress Grb 14 in both this study and by Hemming et al. (2001), whereas Bereziat et al. (2002) employed a distinct expression plasmid (pECE) overexpressing Myc-tagged rGrb14.

91 Dissociation of the Grb14/IR complex may be regulated via a negative feedback mechanism, which has been demonstrated for the Grb2/Sos interaction. Insulin stimulation leads to phosphorylation of the Ras guanine nucleotide exchange factor Sos on Ser/Thr residues, and subsequent dissociation from Grb2 (Cherniack et al., 1995; Waters et al., 1995a; Waters et al., 1995b). This regulatory phosphorylation of Sos is mediated via a MEK-dependent pathway (Holt et al., 1996). Furthermore, interaction of the IR with IRS-1 may be regulated by a negative feedback mechanism leading to phosphorylation of the rat

IRS-1 on Ser307, which is orthologous to Ser312 in humans, resulting in dissociation of the IR/IRS-1 complex (Aguirre et al., 2002).

Association of Grb14 with the IGF-IR was not detected (Figure 4.3A), despite the fact that the IR and IGF-IR kinase regions are the most highly conserved, bearing 84% amino acid identity (Siddle et al., 2001). Indeed, Grbl4/IGF-IR association has only been detected in vitro whereby the adaptor inhibited IGF-IR catalytic activity, albeit at lower potency than the IR (Bereziat et al., 2002). Similarly, GrblO is reported to preferentially interact with the IR but not the IGF-IR in R fibroblasts overexpressing each receptor (Laviola et al., 1997). However, another study demonstrated in vivo interaction of GrblO with the IGF-IR in NIH 3T3 fibroblasts (Wang et al., 1999). Thus, it is possible that the Grbl4/IGF-IR association is either weak and/or transient and may be cell type-dependent.

Similar to the interaction with the IR, binding of Grb14 to FGFR-1 is dependent upon receptor activation. However, Grb14 binds the IR at one region, the kinase domain (Kasus­ Jacobi et al., 1998), whereas Grb14 binding to FGFR-1 involves both the juxtamembrane region and the C-terminal tail (Reilly et al., 2000). Moreover, in contrast to the BPS­ dependent Grb 14 interaction with the IR, the results herein demonstrated that the FGFR-1 interaction was solely dependent on the SH2 domain, as Grb14 ~ failed to associate with the FGFR-1 in the co-immunoprecipitation experiments (Figure 4.6B). In addition, the

Grb14 ~ SH2/BPS-SH2 domains did not interact with the receptor in GST pulldown assays (Figure 4.6C).

92 While this work was in progress, Reilly et al. (2000) demonstrated that Grb14 interacted with the FGFR-1 in NIH 3T3 fibroblasts and the association was dependent on receptor activation using co-immunoprecipitation analysis. They also mutated the conserved binding sequence of the Grb14 SH2 domain (R466K) and indicated that the Grb14 SH2 mutant failed to bind the FGFR-1, which supported the experiments undertaken here in COS-7 cells. In addition, using GST pulldown analysis from PAI cells, constructs corresponding to either the Grb14 N-terminus or lacking the SH2 domain were able to bind the FGFR-1 constitutively, which suggested that regions of Grb14 upstream of the SH2 domain may contribute to the association with the receptor (Reilly et al., 2000).

Therefore, results presented in this chapter collectively demonstrated that alternative splicing directs Grb14 recruitment to different signalling pathways. The activated IR targets both Grb14 isoforms since it requires the Grb14 BPS domain, whereas FGFR-1 signalling excludes Grb14 (3 since it requires a functional SH2 domain. Indeed, alternative splicing regulates the activity of a variety of signalling proteins including RTKs, as well as catalytic and adaptor-type proteins, often resulting in diverse biological effects. For instance, the presence or absence of four amino acids in the extracellular domain of two c-kit receptor isoforms results in markedly different transforming potentials both in vitro and in viva (Caruana et al., 1999). The deletion of 16 amino acids from the cytoplasmic tail of an erbB4 isoform abrogates binding and activation of PI-3K (Elenius et al., 1999), and hinders the mediation of NRG-1 (3-stimulated chemotaxis and survival (Kainulainen et al., 2000). Similarly, alternative splicing regulates the activity of catalytic signalling intermediates. A fifteen amino acid insertion in the C-terminus of hSosl-Isf II near its Grb2 binding site enhances Ras activation and results in shorter cell population doubling times with more frequent tumours in nude mice (Rojas et al., 1999).

Opposing effects of splice variants have been shown for the adaptors Grb2 and the She proteins. Grb3-3, a Grb2 variant with a deleted SH2 domain, markedly inhibits EGF­ induced transcription from a Ras response element, in contrast to the stimulatory effect of Grb2. Grb3-3 also induces apoptosis when microinjected in excess relative to the full-length isoform (Fath et al., 1994 ). Similarly the p66shc protein, characterised by an additional

93 collagen-homology domain, inhibits EGF activation of the c-fos promoter, in contrast to the upregulation observed for the isoforms p52shc /p46shc (Migliaccio et al., 1997).

In the case of Grb14 isoforms, their biological effects are likely to depend on the receptor­ mediated signalling pathway. For instance, rGrb 14 inhibits IRS- I phosphorylation and subsequent insulin-stimulated mitogenesis and glycogen synthesis (Kasus-Jacobi et al., 1998). As mentioned earlier, this may be a direct inhibitory effect of the BPS domain on IR catalytic activity and subsequent substrate phosphorylation (Stein et al., 2001; Bereziat et al., 2002). Thus, Grb14 ~ is likely to retain an inhibitory effect on IR signalling, although with less potency than Grb14 a since it displays relatively weaker binding to the receptor (Figure 4.3B).

In contrast to their common role in inhibition of IR signalling, Grb 14 isoforms are likely to have distinct effects in FGF signalling. Reilly et al. (2000) demonstrated that a Grb14 SH2 point-mutant which cannot bind FGFR-1 significantly upregulates FGF-stimulated mitogenesis, presumably by functioning as a dominant negative. This suggests that Grb14 ~ may enhance FGF-mediated proliferation, in contrast to the inhibitory effect of Grb14 a. This hypothesis is investigated in the next chapter.

94 Chayter 5- 'Functionaf efects ef firb14 is'!forms in 'Ffi'F­ stimulatedceffyrol!foration

5.1 Introduction To date there are four FGFRs identified, FGFR 1-4 (section 1.2.1.2). In the presence of heparin sulfate proteoglycans, ligand binding elicits receptor dimerisation and subsequent phosphorylation of intracellular tyrosine residues (Spivak-Kroizman et al., 1994). In particular, FGFR-1 Tyr-766 is the major site required for binding to and activating PLC­ y and eliciting receptor internalisation (Mohammadi et al., 1991 ). In addition, sites identified at Tyr-653, -654 are important for receptor activation and mediating biological effects such as mitogenesis and differentiation (Mohammadi et al., 1996). FGFR-1 autophosphorylation activates substrates such as She (Klint et al., 1995), c-Src, PLC­ y (Boilly et al., 2000), FRS-2 a/SNT-1 and FRS-2 ~/SNT-2 (Xu et al., 1998). In response to FGF stimulation, FGFR-1 in vivo binding proteins include PLC-y (Mohammadi et al., 1991) and the c-Src kinase (Zhan et al., 1994). FRS-2 links FGFR-1 to the Ras/MAPK pathway by recruiting the Grb2/Sos complex (Kouhara et al., 1997), although receptor interaction with FRS-2 has only been demonstrated in vitro (Xu et al., 1998; Ong et al., 2000). FRS-2 binds constitutively to a highly conserved sequence in the juxtamembrane region of FGFR-1 within amino acids 419-430 (Ong et al., 2000).

In chapter 4.6, the full-length Grb14 a isoform was identified as an interacting partner of

FGFR-1, while the novel isoform Grb14 ~ which lacks a functional SH2 domain, did not interact with the receptor (Figure 4.6). While this work was in progress, Reilly et al. (2000) also reported association of FGFR-1 with Grb14 in vivo. The interaction only occurred upon receptor activation and was dependent on the phosphorylation of either Tyr-766 or Tyr-776 in the C-terminal tail of FGFR-1. In addition, the authors demonstrated that a R466K mutation in the Grb14 SH2 domain (which disrupts binding of the FLVRES motif), abrogated binding to the receptor. This was also in agreement with the data presented in chapter 4.6, since the identified Grb 14 ~ that failed to bind the receptor (Figure 4.6),

95 possessed a truncated SH2 domain which lacked the FL VRES sequence due to a premature stop codon (Figure 3.2B). Furthermore, Reilly .et al. (2000) generated mouse fibroblast cell lines stably expressing the wildtype Grbl4 or a Grbl4 R466K SH2 mutant, and tested FGF­ induced DNA synthesis using thymidine incorporation analysis. Wildtype Grbl4 displayed a modest inhibition of FGF-induced mitogenesis where it decreased DNA synthesis by 2-3 fold at 5 ng/ml FGF but not at lower doses. However, the Grbl4 R466K mutant markedly enhanced both the sensitivity and the maximal response to FGF stimulation (3-4 fold increase at 2 ng/ml). Since the mutant did not bind the receptor, Reilly et al. (2000) postulated that it potentiated FGF-induced proliferation through a dominant negative effect mediated by interactions with unknown downstream effectors. Since the R466K mutant and

Grb 14 ~ do not possess an intact SH2 consensus binding site, rendering the SH2 domain dysfunctional and incapable of binding phosphotyrosines and thus the FGFR-1, then the dominant negative effect observed for the point-mutant which potentiates the effects of FGF may then be expected to occur for the naturally occurring variant Grbl4 ~-

Given that FGF activation is implicated in carcinogenesis (MacArthur et al., 1995; Cappellen et al., 1999) it is then important to test whether the identified splice variant

Grbl4 ~ will also potentiate the effects of FGF proliferation as this will have biologically relevant implications. Furthermore, it is also of interest to contrast this with the effect produced by the full-length Grbl4 a.

5.2 Overexpression of Grb14 isoforms in 3T3 fibroblasts usin,: retroviral infection

In order to investigate the function of the Grbl4 isoforms in FGF-induced proliferation, retroviral infection following transient retroviral production was employed to express these proteins in fibroblasts. This method was chosen because it provides a powerful gene delivery system (Mulligan, 1993) resulting in high infection efficiency and subsequent expression of the encoded proteins by a large population of target cells (Pear et al., 1993; Finer et al., 1994).

96 In order to create retroviral constructs encoding for Grb 14 isoforms, both protein-encoding sequences for Flag-tagged Grb14 a and ~ were excised from pRcCMVFlag expression vectors together with a Flag epitope-tag and subcloned into the pLib retroviral vector as described in Materials and Methods (see section 2.1.5). The retroviral vectors containing the Grb14 isoforms were then transiently transfected into a packaging cell line (Phoenix­ Eco ). In addition, this packaging cell line contained genes encoding proteins required for viral production (i.e. gag for structural proteins, pol for integrases/reverse transcriptase and env for envelope glycoproteins). The packaging signal (\j/+) required for virus particle formation was encoded by the retroviral vector and allowed viral production when transfected into the packaging cell line.

After transfection into Phoenix-Eco cells, at the approximate time when maximum viral titre is reached (48 h), the retroviral supernatant was collected and subsequently used to infect 3T3 mouse fibroblasts. 48 h post-infection, cells were lysed and Western blotted to verify expression of the Grb14 proteins, since maximal expression of retroviral-encoded proteins was reached at this time. This revealed additional bands in the appropriate infected cells running comparably to Grb14 a and ~ in blotting controls, which were not identified in lysates from parental 3T3 cells or cells infected with the retroviral control vector (Figure 5.lA). This indicated that both proteins were successfully expressed in the infected cells.

5.3 FGF-induced DNA synthesis rate oj3T3 fibroblasts usin: thymidine incorporation analysis

The thymidine incorporation assay was utilised in order to directly analyse the rate of FGF­ induced DNA synthesis in mouse fibroblasts. Since the decrease in FGF-induced stimulation due to Grb14 overexpression may be of small magnitude (Reilly et al., 2000), this assay will maximise the chances of detecting small differences in proliferation. bFGF (FGF-2) was utilised for stimulation as described by Reilly et al. (2000). To determine the optimal bFGF doses for stimulation, proliferating parental 3T3 fibroblasts were starved in 0.5% serum and subsequently stimulated with increasing concentrations of bFGF for 24 h prior to a 5 h incubation with 3H-thymidine. Analysis by liquid scintillation counting

97 A

Grb14a ~

Grb14 p ~

blot: Flag

B

8000

6000 E 0. ~ C: 0 ~e- 4000 D parental NIH 3T3 0 ..s(..) Q) C: '6 t 2000 .....=

0 0.1 0.5 5 serum

FGF (ng/ml)

Figure 5.1 Developing the retroviral infection/thymidine incorporation assay. A. 3T3 fibroblasts were infected with a retrovirus encoding either Grb14 isoform or the corresponding empty retrovirus control (Vector) as described in section 2.3.4. Cells were then lysed, separated by SDS-PAGE and transferred to nitrocellulose. Membranes were immunoblotted for Flag to detect the Grbl4 proteins. The band immediately below Grbl4 a (in 3T3 + Grbl4 a sample) was a cross-reacting band present in all samples. Lysates from CHO-IR cells overexpressing Grbl4 isoforms were included as blotting controls B. 3T3 fibroblasts were plated in 10% NBCS-containing medium, washed and starved in medium supplemented with 0.5% NBCS for 48 h. Cells were then treated with the indicated concen­ trations of bFGF or with 10% NBCS. Cells were labelled for 5 h with 3H-thymidine and analysed by liquid scintillation counting of incorporated radioactivity. Values are expressed as cpm of incorpo­ rated radioactivity. Average readings of duplicate data points are shown. revealed that thymidine incorporation increased with the FGF concentration used, where it displayed a strong increase between 0 and 0.5 ng/ml bFGF, which was further augmented at the highest dose of 5 ng/ml (Figure 5.1B). Furthermore, 10% serum stimulation caused an increase in cell proliferation that was less than the maximum FGF dose of 5 ng/ml. Therefore, these results indicated that the FGF concentrations utilised in this dose-response assay were optimal for examining changes in thymidine incorporation analysis of 3T3 fibroblasts.

5.4 Effects of Grbl4 isoforms on the DNA synthesis rate of mouse fibroblasts usin~ thymidine incorporation analysis

To identify the effects of exogenous Grb14 expression on FGF-induced fibroblast growth, an assay was developed combining both retroviral infection to express Grb 14 isoforms and thymidine incorporation analysis of the DNA synthesis rate. For the infection experiments, Western blotting was routinely undertaken to verify expression of the infected proteins (approximately 48 h post infection). Furthermore, since the utilised pLib retroviral vector lacked selectable markers, GFP-encoding retroviral plasmid was included as an additional control in the experiments to analyse infection efficiency using fluorescence microscopy. The GFP-encoding retroviral plasmid (pLib) was transiently transfected into the packaging cells as described earlier for the Grb14 isoforms. Subsequently, the viral supernatant was collected and utilised to infect the target fibroblasts. When maximal expression was achieved, approximately 48 h after infection, fluorescent cells were visualised under the microscope. This routinely revealed an infection efficiency of 60-70% (an average of 3 light fields chosen at random).

An experiment was then performed where 3T3 fibroblasts were infected with retroviruses encoding the Grb14 isoforms or the control retrovirus, and together with parental fibroblasts, were starved and stimulated with 0.5 and 5 ng/ml of bFGF or with 10% serum (NBCS). After incubation with 3H-thymidine and subsequent analysis of incorporated radioactivity, all samples displayed an increased rate of DNA synthesis relative to basal levels following stimulation with bFGF and serum, with maximum thymidine incorporation

99 generally observed at 5 ng/ml bFGF (Figure 5.2). However, at both bFGF concentrations and in response to 10% serum stimulation, there was no consistent difference in incorporated 3H-thymidine between cells overexpressing Grb14 isoforms, and either the control retrovirus or parental cells. This indicated that in this system, Grb 14 isoforms did not affect the rate of DNA synthesis induced by bFGF or serum.

5.5 Effects of Grb 14 isoforms on cell cycle phase distribution of mouse fibroblasts in response to FGF stimulation

The thymidine incorporation assay provided information regarding the rate of DNA synthesis for cells overexpressing Grb14 isoforms and controls. In order to extend this, flow cytometry was performed to analyse cell cycle phase distribution as determined by cellular DNA content. Samples treated in parallel with those from the previous experiment were incubated with the growth factor for an identical timepoint (29 h), after which they were harvested, stained for DNA content and analysed. In accordance with the thymidine incorporation assay results, DNA histograms indicated that stimulation of cells with bFGF increased the percentage of cells in S-phase for all samples (Figure 5.3). This increase was modest for 0.5 ng/ml FGF (eg. from 9% to 12% for parental cells). The S-phase fraction was comparable for all samples at this dose (12% for parental cells and 10% for other samples). However, the percentage of cells in Gl was lower for the parental cells (69%) relative to the control vector (77%) and Grb14 isoforms (76% for Grb14 a, 78% for Grb14 ~). Further, the percentage of cells in G2 was higher for parental cells (20%) relative to the remaining samples (13% for vector, 14% for Grb14 a, 12% for Grb14 ~). This suggested that there were slight differences in the timing of S-phase entry for the infected cells, where parental cells progressed more quickly through the cell cycle relative to the other samples. It also indicated that there were no differences in cell cycle progression between cells expressing Grb14 isoforms and the vector control cells in response to 0.5 ng/ml bFGF.

Stimulation with 5 ng/ml bFGF elicited a greater drive in cell cycle progression relative to the lower dose for all samples (eg. %S-phase increased to 24% for parental cells). Although at this concentration, the S-phase fraction for cells overexpressing Grb14 isoforms was

100 D parental 3T3

D vector

0 Grb 14 a

0 Grb14P

80000-

60000- E T 0... ~ c:: T C> 1 ~ T ill e-C> 40000 T C> c..;, ..1.. c:: J_ ...L.. c:: '5.E -=>-- I- 20000- --

0

0 0.5 5 serum

FGF (ng/ml)

Figure 5.2 Effects of Grbl4 isoforms on FGF-induced rate of DNA synthesis. 3T3 fibroblasts infect­ ed with a retrovirus encoding either Grbl4 isoform or the empty retrovirus control (vector) as in sec­ tion 2.3.4 were starved and stimulated with bFGF or 10% NBCS as stated in Figure 5.1. Cells were labelled for 5 h with [3H] thymidine and analysed by liquid scintillation counting of incorporated radioactivity. Values represent the mean± S.E. of triplicate measurements. Serum-starved FGF 0.5nglml FGF 5ngtml

G1 88% G1 69% G1 47% S 9 % S 12% S 24% parental G2 3% G2 20% G2 29%

G1 87% G1 77% G1 53% S 6% S 10% S 17% vector G2 7% G2 13% G2 30%

G1 87% G1 76% G1 53% S 8% S 10% S 18% Grb14 a G2 5% G2 14% G2 29%

G1 90% G1 78% G1 56% S 6% S 10% S 19% Grb14 ~ G2 4% G2 12% G2 25%

Figure 5.3 Effects of Grb14 isoforms on FGF-stimulated cell cycle progression. 3T3 fibroblasts infected with a retrovirus encoding either Grb14 isoform or the empty retrovirus control (vector) from Figure 5.2B were serum-starved and stimulated with bFGF (0.5 and 5 ng/ml) for 29 h. Cell s were then harvested, stained for DNA content and analysed by flow cytometry. DNA histograms showing cell cycle phase distribution and the percentage of cells in G 1, S, G2 phases are demonstrated. slightly lower than parental cells (18% for Grb14 a and 19% for Grb14 ~. relative to 24% for parental cells), it was similar to the retroviral control sample which also had 17% S­ phase. Therefore, in accordance with the thymidine incorporation data, flow cytometric analysis indicated that there was no effect of overexpressing Grb14 isoforms on FGF­ induced cell cycle progression of mouse fibroblasts.

5.6 Insulin-induced DNA synthesis of mouse fibroblasts

Since Grb14 functions downstream of insulin/IGF receptors in response to insulin stimulation (Kasus-Jacobi et al., 1998; Hemming et al., 2001 ), it was of interest to examine the effects of Grb 14 isoforms on insulin signalling utilising the same system outlined in the previous section. Initially, it was necessary to assess the mitogenic effect elicited by insulin stimulation of 3T3 fibroblasts. Therefore, parental fibroblasts were serum-starved and stimulated as indicated earlier (section 5.3) with various insulin concentrations, with or without EGF, since these factors elicit a synergistic increase in DNA synthesis of fibroblasts (Crouch et al., 2000). Thymidine incorporation analysis indicated that insulin alone elicited a poor increase in DNA synthesis relative to control levels (C), even at a high concentration (10 µg/ml) (1,10) (Figure 5.4). EGF stimulation increased DNA synthesis to levels above those induced by insulin alone. However, insulin and EGF in combination elicited a synergistic increase in the rate of DNA synthesis (for both 1 and 10 µg/ml insulin). Serum stimulation also caused an increase in thymidine incorporation greater than that produced by insulin or EGF alone, although it was not as marked as the effect obtained with the combination of growth factors. Therefore, these results indicated that insulin alone was not a potent mitogen for 3T3 fibroblasts, most likely due to ineffective downstream signalling (Crouch et al., 2000) and hence this system was not optimal to study the effect of Grb14 isoforms in insulin-induced mitogenesis. The synergy between insulin and EGF is suggested to result from the insulin-induced co-clustering of the EGFR with PLC-yl at the actin arc which may enhance signalling efficiency (Crouch et al., 2000).

5. 7 Discussion

In this study, retroviral infection was used to overexpress human Grb14 isoforms in mouse

103 4000 -

3000 - T ,... 1 ...L.

2000 - T ..L

1000 -

0

Figure 5.4 Effects of EGF/insulin stimulation on DNA synthesis of 3T3 fibroblasts. Parental 3T3 fibroblasts were plated in medium containing 10% NBCS, washed and starved in medium supple­ mented with 0.5% NBCS for 48 h. Cells were either left untreated (C) or stimulated for 24 h with insulin alone at 0.1 µg/ml (I, 0.1), 1 µg/ml (I, 1) or 10 µg/ml (I, 10). In addition, cells were stimulated for 24 h with EGF alone at 10 ng/ml (EGF), in combination with insulin (I,l +EGF; I,lO+EGF) or with 10% NBCS (serum). Cells were then labelled for 5 h with 3H-thymidine and analysed by liquid scintillation counting of incorporated radioactivity. Values are expressed as cpm of incorporated radioactivity and represent the mean± S.E. of triplicate measurements. fibroblasts, in order to test their effect on FGF-induced proliferation using thymidine incorporation assays and flow cytometry. The response to serum stimulation was also examined, which revealed that there were no consistent differences between cells expressing Grbl4 isoforms and control samples. This is in accordance with Reilly et al. (2000), who demonstrated that Grbl4 and a generated SH2 mutant did not affect serum­ induced stimulation of fibroblasts. However, in the study described here, there was no consistent difference in FGF-induced DNA synthesis or cell cycle phase distribution between Grbl4-overexpressing samples and controls. These results are conflicting with those of Reilly et al. (2000), who demonstrated that Grbl4 elicited a modest inhibitory effect and the Grbl4 SH2 mutant caused a marked enhancement in the rate of FGF-induced DNA synthesis. These authors also utilised mouse fibroblasts, with an identical timepoint for incubation with FGF (i.e. 24 h prior to the addition of thymidine).

However, the methods employed to exogenously express the protein of interest differed. Reilly et al. (2000) generated stable cell lines by transfecting the appropriate constructs in fibroblasts, selecting with G418 antibiotic to obtain resistant cells and then establishing individual clones, whereas in this chapter retroviral infection was utilised to exogenously express Grbl4 in pools of cells. Although the authors stated that similar results were obtained with different clones, it was nonetheless possible that due to clonal variation, they may have isolated clones that exhibited a response which was not representative of the actual effect. This problem was certainly compounded by the fact that they did not compare the results with the parental cell line, and only utilised vector-transfected cells as control. In this chapter, both parental cells and pools of cells infected with an empty retrovirus were utilised as controls.

Another possibility for the discrepancy between the studies may have originated from the relative overexpession levels of Grbl4. Generation of stable cell lines by Reilly et al. (2000) employed standard transfection procedures which was likely to result in significant Grbl4 overexpression. However, retroviral infection used in this study introduced approximately one copy of the retrovirus (and the encoded gene) in a single cell as retroviral vectors integrate at low copy numbers (Clontech, 1999). Thus, this suggested that

105 Grb 14 expression was higher in the stable cell lines relative to the infected cells. Since the relative ratio of signalling molecules is critical for the activation of certain pathways, this may account for the lack of a functional effect observed here (Figure 5.2). However, if inhibition of cell proliferation is only observed upon excessive Grb14 overexpression, this implies that Grb14 a is not a physiological regulator of FGF-induced mitogenesis.

Although retroviral infection is a highly efficient method of expressing proteins, it was not possible to select for Grb14-expressing cells as the utilised retroviral vector (pLib) lacked selectable markers. However, according to the GFP control, approximately 60-70% of cells were successfully infected. Although the Grb14-mediated inhibitory effect in the Reilly et al. study was modest (2-3 fold at 5 ng/ml), it is unlikely that such an effect would be masked by the proportion of uninfected cells in this system. Furthermore, the retroviral system should certainly be sensitive enough to detect the marked effect on DNA synthesis predicted for Grb14 ~. based on the activity of the R466K SH2 mutant, which significantly potentiated FGF-induced mitogenesis (Reilly et al., 2000). If Grb14 does negatively regulate FGFR-1 signalling, the only possibility to explain the lack of response enhancement by Grb14 ~. is that this protein and the R466K SH2 mutant are not functionally equivalent due to structural differences. Grb14 ~ contains a truncating mutation that abrogates most of the SH2 domain, whereas the SH2 mutant is only altered at a single residue important for phosphotyrosine binding. Therefore, these proteins may differ in phosphotyrosine-independent SH2 domain interactions, and this is discussed in more detail in chapter 7.

FGFR-1 contains at least 7 autophosphorylation sites in its intracellular region, which are Tyr-463, Tyr-583, Tyr-585, Tyr-653, Tyr-654, Tyr-730 and Tyr-766. As stated earlier, Tyr- 766 serves as a binding site for PLC-y and mediates FGF-induced phosphatidylinositol hydrolysis, whereas Tyr-653 and -654 are essential for receptor activation and eliciting FGF biological effects. Mutational analysis has also demonstrated that the remaining four residues are dispensable for FGFR-1-induced mitogenesis and neuronal differentiation (Mohammadi et al., 1996). The interaction of Grb14 with FGFR-1 involves either Tyr-766 or Tyr-776, since mutation of either Tyr does not affect binding of Grb14 in the yeast two-

106 hybrid system, although mutation of both residues completely eliminates the interaction. Since Tyr-776 has not been confirmed as an in viva autophosphorylation site, its role in regulation of mitogenic signalling has not been characterised via expression of a FGFR-1 Y776F mutant. However, mutants with a Y766F substitution do not show enhanced mitogenic signalling, which would occur if binding of a repressor, i.e. Grb14, was abrogated (Mohammadi et al., 1996). This is consistent with the data from this chapter indicating that Grb14 does not repress FGF-induced mitogenesis. Consequently, the results in this chapter indicate that Grb14 is unlikely to be an important physiological regulator of FGFR-1 mitogenic signalling. However, it is possible that Grb14 may regulate other FGF biological endpoints such as FGF-induced chemotaxis. Further evidence that Grb14 does not repress FGFR-1 signalling is provided by the fact that a Grb14 a knockout mouse fails to have a phenotype with any developmental or growth abnormalities (unpublished results by Daly et al.), which would be expected with loss of inhibition of FGFR-1 signalling, since activating mutations of the FGFR-1 are linked to severe skeletal disorders (Muenke et al., 1994; Schell et al., 1995; Neilson and Friese!, 1996). This is also evident in transgenic mice with activated FGFR-1 signalling, which display defects including achondroplasia­ like dwarfism and joint abnormalities (Wang et al., 2001).

107 Chayter 6-'Regufation of(irb14 in estrogen/insufin cross­

tali andefects on yrol!foration of M.C'F-7 breast cancer

cells

6.1 Introduction

Cross-talk between steroids and growth factors is recognised as playing an important role, not only in regulating normal breast development, but also in the progression of breast cancer. Unlike growth factors which act through RTKs, steroid hormones elicit their effects through a large family of specific nuclear receptors that function as transcription factors (Clarke, 2000; Yee and Lee, 2000; Gross and Douglas, 2002) (see sections 1.2.3 and 1.2.4). Although extensive studies have been undertaken to dissect out the convergence points of the interplay between steroid and growth factor signalling, the underlying molecular mechanisms are still being defined.

Estrogen increases cell proliferation both by direct effects on cell cycle regulatory molecules and by enhancing growth factor action. Estrogen upregulates stimulatory components of insulin/lGF-1 signalling pathways such as IGF-11 (Yee et al., 1988), IGF-lR (Stewart et al., 1990; Stewart et al., 1992) and IRS-1 (Lee et al., 1999; Molloy et al., 2000). In addition, estrogen downregulates inhibitory molecules involved in the insulin/lGF-1 cascade such as IGF-IIR (Mathieu et al., 1991), IGFBP-1 and -3 (Perks and Holly, 2000; Yee and Lee, 2000 ). Moreover, antiestrogen treatment inhibits insulin/lGF pathways by altering IGF-IR transcription (Huynh et al., 1996a) and phosphorylation as well as decreasing IRS-1 activity and expression (Salemo et al., 1999).

Furthermore, estrogen acts directly on growth factor pathways by activating c-Src, resulting in upregulation of the p21 ras/MAPK signalling cascade (Migliaccio et al., 1996; Yee and Lee, 2000). Estrogen also affects the expression and activation of cell cycle regulatory

108 molecules including c-Myc, cyclin Dl-cdk4, cyclin E-cdk2 and p21wAFIICIPi (Altucci et al., 1996; Prall et al., 1997). In this context it cross-talks with insulin/IGFs, resulting in decreased p21 expression and the formation of active cyclin E/cdk2 complexes (Lai et al., 2001 ). Also, the estrogen and insulin/IGF-I interaction is bidirectional, since IGF-1 and insulin enhance ERa activity in breast cancer cells (Patrone et al., 1996; Stoica et al., 2000), one contributing mechanism being MAPK phosphorylation of the ERa AF-1 domain (Kato et al., 1995; Tremblay et al., 1999).

Delineating the mechanisms involved in steroid/growth factor cross-talk has important clinical implications, since upregulation of growth factor pathways may lead to failure of endocrine therapy. In support of this, overexpression of the IGF-IR or erbB2 in MCF-7 cells results in tamoxifen resistance (Wiseman et al., 1993; Kurokawa et al., 2000).

Although a function for Grb7 family proteins has not been established in estrogen-induced signalling, functional studies undertaken so far suggest both regulatory and adaptor roles for these proteins in insulin/IGF-induced signal transduction (Daly, 1998; Han et al., 2001). Despite the fact that these studies collectively provide clues about the function of Grb7 proteins, there is very limited information available regarding their regulation. Therefore, since growth factor/steroid cross-talk plays an important role in the regulation of breast cancer cell proliferation, and Grb14 is a component of insulin/IGF-I signalling pathways, the regulation and function of Grb14 a (denoted here as Grb14), in response to insulin and estrogen stimulation was investigated in breast cancer cells.

6.2 Grbl 4 expression in breast cancer cell lines

Previous analysis of Grb14 mRNA expression demonstrated that Grb14 mRNA was predominantly restricted to normal breast epithelial cells and ER-positive breast cancer cell lines (Daly et al., 1996). In order to examine whether the expression of Grb14 protein in breast cancer cell lines was consistent with these findings, Western blot analysis was performed using lysates from the normal human mammary epithelial cell strain 184 and a panel of breast cancer cell lines. Immunoblotting for endogenous Grbl4 levels was undertaken with a commercially available Grb14 polyclonal antibody directed against an

109 N-terminal epitope (see section 2.2.J). This revealed an immunoreactive band that migrated at the predicted size for the Grb14 protein, which was approximately 56 kDa (Daly et al., 1996), indicating that this band represented the Grb14 protein.

Grb 14 was expressed at high levels in the normal mammary 184 cells, and was also detected in 6/7 ER-positive breast cancer cell lines (T-47D, ZR-75-1, MCF-7, BT-483, MDA-MB- 134 and BT-474) (Figure 6.lA). In contrast, this protein was only detected in 2/7 ER­ negative lines (MDA-MB-468, Hs578T) and possibly in MDA-MB-436 cells, indicating that the expression of the Grb14 protein was predominantly restricted to ER-positive breast cancer cells. Unlike the other cell lines, MDA-MB-436 immunoblotting revealed a faint doublet rather than one specific protein band. These may be cross-reacting bands or may represent unidentified Grb14 isoforms. The immunoblotting results suggested that there was a correlation between Grb14 expression and ER positivity. To determine whether these results were statistically significant, chi-square analysis was performed. This indicated that the correlation between Grb14 protein expression and ER positivity was statistically significant (p=0.03). Further, the expression of the Grb14 protein was scored as+ for low, ++ for intermediate,+++ for high, and compared with the expression of the Grb14 mRNA (Daly et al., 1996) (Figure 6.1B). This indicated that there was a positive correlation between Grb14 protein and mRNA expression. When subjected to chi-square analysis, the correlation between Grb14 protein expression and mRNA was also statistically significant (p=0.03) (Figure 6.1B). It was interesting to note that MDA-MB-361 cells had low expression of Grb14 mRNA but the corresponding protein was undetectable. This may be due to post-translational modifications that do not favour protein stability in these cells. Also cell lines BT-474, Hs578T and possibly MDA-MB-436 expressed low levels of the Grb14 protein but lacked detectable mRNA expression, which may be due to rapid mRNA turnover in these cells.

110

.,::,. .,::,.

::J ::J

~ ~ !:!'. !:!'.

C" C"

...... 0 0

> >

C) C)

MDA-MB-453 MDA-MB-453

MDA-MB-436 MDA-MB-436

MDA-MB-231 MDA-MB-231

Hs578T Hs578T

(I) (I)

< <

I I

BT-549 BT-549 :c :c

m m

SK-BR-3 SK-BR-3

BT-20 BT-20

I I

MDA-MB-468 MDA-MB-468

BT-474 BT-474

MDA-MB-361 MDA-MB-361

t t

MDA-MB-134 MDA-MB-134

BT-483 BT-483

(I) (I)

< <

+ +

:c :c

m m MCF-7 MCF-7

ZR-75-1 ZR-75-1

T-47D T-47D 184 184 B Origin Cell line Grb14mRNA Grb14 protein

Normal human breast epithelial 184 +++ +++

Human breast cancer, ER+ T-47D +++ ++ ZR-75-1 ++ + MCF-7 + + BT-483 + + MDA-MB-134 + + MDA-MB-361 + BT-474 +

Human breast cancer, ER- MDA-MB-468 + ++ BT-20 SK-BR-3 BT-549 Hs578T + MDA-MB-231 MDA-MB-436 +/- MDA-MB-453

Figure 6.1 Grbl4 protein expression in normal human breast epithelial cells and breast cancer cell lines. A. Cell lysates were prepared from the normal human epithelial cell strain 184 and a panel of breast cancer cell lines. These were then separated by SDS-PAGE and subjected to Western blot analysis with a Grbl4-specific polyclonal antibody. The membranes were also blotted for actin as a loading control. B. Positive correlation between Grbl4 gene expression and ER status. Protein levels were scored from - to +++. Also included are Grbl4 mRNA levels for the cell strain 184 and the different cancer lines (Daly et al., 1996). +/-; ambiguous expression due to banding pattern. 6.3 Regulation ofGrb14 protein expression by estrogen and the antiestrogen IC/ 182780

Since the immunoblotting results indicated that there was a statistically significant positive correlation between Grb14 protein expression and ER status in breast cancer cell lines, studies were undertaken to investigate whether there was a functional link between ER and Grb14 in breast cancer cells. Since estrogen effects are mediated through the ER, and antiestrogens act by blocking this effect, Grb14 protein expression was initially examined in response to stimulation with estradiol and with the pure steroidal anti~strogen ICI 182780. These were performed using the MCF-7 breast cancer cell line which is highly estrogen-responsive and well characterised. MCF-7 cells were maintained in steroid­ depleted medium which contained 10% charcoal-stripped serum (charcoal treatment removed steroids in the medium) in the presence or absence of estradiol supplementation for 72 h. Cells were then harvested, lysed and analysed by Western blotting using a Grb14 polyclonal antibody and with an actin antibody for loading analysis. This allowed the detection of Grb14 protein in lysates from control cells that were maintained in steroid­ depleted medium supplemented with stripped serum (C). However, lysates from estradiol­ stimulated cells contained lower levels of the Grb14 protein, whereas those from cells arrested by antiestrogen treatment expressed higher Grb14 protein levels than control. The data was quantitated by densitometric analysis and normalised for loading as described in Figure 6.2. This revealed that Grb14 protein levels were decreased approximately 2 fold by estradiol (Figure 6.2A). However, when cells were treated with the pure antiestrogen ICI 182780, Grb14 protein expression displayed an approximate 2 fold increase. To ascertain the timing of estradiol effects on Grb14 protein levels, cells were stimulated with estradiol for increasing periods of time, lysates prepared and Western blotted for Grb14 protein (Figure 2B). Although there were no consistent effects at 12 h, results indicated that Grb 14 protein levels decreased by 24 h and were further reduced between 24 and 48 h (Figure 2B). This indicated that Grb14 downregulation was a late effect of estrogen stimulation.

113 A

2-·.------....,,,,,----,

C

s0 c. ·-~ 1.s- C E ICI ,-.,,. C -e :::l C) ~ "O ~ - - I._Grb14 a,~ 1/1 .c ·-ea- ._.....ea I- • ._Actin E... 0 o.s- z

C E ICI

B

+E 0 24 48 Grb14

HI Actin

Figure 6.2 Regulation of Grbl4 protein expression by estrogens and antiestrogens. A. MCF-7 human breast cancer cells were plated in 10% FCS for 24 h, then washed and maintained for 72 h in 10% charcoal-stripped serum (C) alone, or supplemented with estradiol (10 nM,E), or the pure antiestrogen ICI 182780 (100 nM,ICI). Left panel Cell lysates were prepared and Western blotted with a Grbl4 polyclonal antibody. Results are representative of at least three independent experiments. Right panel Results were quantified by densitometry and normalised for actin. Values represent the mean ± S.E. of triplicate measurements. B. M CF-7 cells were plated in 10% FCS for 24 h, washed and maintained for 72 h in 10% charcoal stripped serum, and then stimulated for 24 and 48 h with estradiol ( JO nM). Cell lysates were then Western blotted for either Grbl4 or actin. Results are representative of at least two independent experiments. 6.4 Regulg,tion ofGrb14 protein expression by estrogen and insulin/lGFs

Having demonstrated hormonal regulation of Grb14 levels, subsequent investigations focussed on the effects of insulin/IGF on Grbl4 expression, since it functions in the insulin/IGF signalling pathway and these growth factors are potent mitogens for breast cancer cells. MCF-7 cells were cultured in serum-free medium and then stimulated for 24 h with either insulin, estradiol or the two in combination. Cells were subsequently harvested, and either analysed by flow cytometry, or lysed and subjected to Western blotting. Examination of cell cycle phase distribution by flow cytometry indicated that insulin and estradiol both stimulated cell cycle progression (Figure 6.3A), resulting in an increase in S­ phase fraction from 15% to 38% and 21 %, respectively. When used in combination this resulted in a greater increase to 41 %. Immunoblotting using a Grb14 antibody resulted in the detection of Grb 14 protein in lysates from control serum-starved MCF-7 cells (C, Figure 6.3B). Insulin treatment (I) increased Grb14 protein levels approximately 2 fold relative to control as demonstrated by densitometric analysis. At this timepoint (24 h), estrogen stimulation (E) did not alter Grb14 protein levels relative to control cells, suggesting that estrogen regulation of Grb14 levels was undetectable in a serum-free cell culture system. However, when both hormones were used (I+E), Grb14 protein levels were approximately 1.5 fold above control, but were decreased relative to the insulin-treated sample, indicating that the insulin-induced increase in Grb14 protein expression was inhibited. Note that a high dose of insulin (10 µg/ml), which activates both the IR and IGF­ IR was utilised for stimulation (Adamo et al., 1992). Given that Grb14 is a repressor of insulin/IGF signalling (Kasus-Jacobi et al., 1998 ; Bereziat et al., 2002), these results suggested that Grb14 functions in negative feedback regulation of insulin signalling, and this regulatory mechanism is opposed by estradiol.

6.5 Regulation of Grbl 4 by either estrogen-induced or estrogen-independent cell cycle progression

Although estrogen did not affect Grb14 levels in serum-free culture, the downregulation of Grb14 protein by estradiol in serum-containing medium was surprising since Grbl4 has not

115 A C

15%S 38%S

E I+ E

21 % S 41%S

B

C E l+E

Grb14

Figure 6.3 Regulation of Grbl4 protein expression by insulin and estrogen. A. MCF-7 cells were maintained under serum-free conditions for 72 h prior to treatment for 24 h with vehicle (C), insulin (10 µg/m] ,J), estradiol (1 nM ,E) or the two in combination (I+E). Cells were harvested for FACS analysis, and DNA histograms showing cell cycle distribution are presented. B . C ell lysates were prepared 24 h after stimulation and Western blotted for Grbl4 or actin. Results are representative of at least two independent experiments. been demonstrated to act downstream of ER-initiated signalling pathways. Moreover, since estrogen promotes cell cycle progression of hormone-responsive breast cancer cells, it was possible that the decrease in Grb14 expression was an effect of cell cycle progression as opposed to being specific to estrogen treatment. To address this, regulation of Grbl4 expression was examined in response to cell cycle progression initiated independently of estrogen treatment. c-Myc is a protooncogene that acts as a transcription factor and its activation induces cell cycle re-entry in quiescent cells (Boxer and Dang, 2001). Thus, an inducible system was employed where c-Myc expression stimulates antiestrogen-arrested MCF-7 cells to re-enter the cell cycle (Prall et al., 1998). This system utilizes MCF-7 cells stably transfected with a construct encoding c-Myc under the control of the metal response elements of the hMTIIA metallothionein promoter. In this model, Zn stimulation elicits cell cycle progression independent of estrogen (Prall et al., 1998) (Figure 6.4A, schematic). Stable transfectants of a control vector were also included in this system.

Previous studies indicated that treatment of proliferating MCF-7 cells with ICI 182780, a potent estrogen antagonist that downregulates the ER (Howell et al., 2000), arrested cells in G0-Gl phase (Watts et al., 1995; Carroll et al., 2000). Therefore, control and c-Myc cells proliferating in 10% serum-containing medium were arrested with ICI 182780 for 48 hand then treated with vehicle (C) or stimulated by either estradiol (E) or zinc (Zn) treatment. Lysates were then prepared for Western analysis at 6 h to detect the induction of c-Myc protein (Prall et al., 1998) and at 48 h to analyse Grb 14 downregulation in response to estradiol (see Figure 6.2B). In addition, to assess cell cycle progression cells were harvested and analysed by flow cytometry at 24 h as this represented the timepoint for maximal increase in S-phase fraction in response to estradiol stimulation (Lai et al., 2001). Immunoblotting using a c-Myc antibody demonstrated that estradiol induced high levels of c-Myc expression after 6 h in control cells (MCF-7/~MT) and in c-Myc-expressing cells (MCF-7/~MT-c-Myc) (Figure 6.4A, right panel). In contrast, Zn induced c-Myc overexpression only in the MCF-7/~MT-c-Myc cells. Increased levels of c-Myc protein were detected in the control sample in c-Myc-transfected cells possibly due to basal promoter activity. Flow cytometry analysis indicated that cell lines arrested with ICI 182780 contained a low percentage of S-phase cells, analysed as 4% and 6% for MCF-

117 7/~MT and MCF-7/~MT-c-Myc cell lines respectively. Estrogen treatment increased the S­ phase fraction to 40% and 41 % for control and c-Myc cell lines respectively (Figure 6.4B). However, Zn increased the percentage of S-phase cells for the MCF-7/~MT-c-Myc line to 21 %, whereas a very small effect on S-phase was observed in the control cells. This indicated that c-Myc induction increased cell cycle progression independent of estradiol stimulation in this system. In the estradiol-stimulated samples, the proportion of G2/M cells was decreased in the MCF-7/~MT-c-Myc cell line (13%) relative to control cells (27%). This was likely to reflect differences in the timing of S-phase entry for the cell lines.

Western blotting analysis indicated that Grb14 protein was reduced approximately 1.5 fold by estradiol in both cell lines at 48 h (Figure 6.4C). In contrast, Zn stimulation failed to downregulate Grb14 protein in control or c-Myc transfectants, suggesting that the decrease in Grb14 protein levels was specific to estrogen treatment and occurred independently of cell cycle progression. IRS-I protein expression was also examined as a control in this system (Figure 6.4C), since it is enhanced by estradiol treatment (Lee et al., 1999; Molloy et al., 2000). As expected, IRS-I protein was increased in both cell lines in response to estradiol stimulation and was not affected by Zn. Finally, investigations of Grb14 regulation were extended by synchronising cells in G2/M phase using nocodazole, which promotes tubulin depolymerisation and inhibits mitosis by causing a G2/M phase arrest (Scholler et al., 1994; Storrie and Yang, 1998). Cells were arrested with ICI 182780 as described earlier and rescued with estradiol or Zn treatment in the presence of nocodazole. Cells were analysed by flow cytometry at 36 h to assess cell cycle progression and protein lysates were prepared at 48 h for Western blotting. A comparable percentage of cells accumulated in G2/M for MCF7/~MT cells stimulated by estradiol and MCF-7/~MT-c­ Myc lines induced with Zn (56% and 59%, respectively). Despite this, immunoblotting analysis demonstrated that Grb14 protein regulation differed between the cell lines, since it

118 A Zn2+ MCF-7 /L'lMT MCF-7 /llMT-c-Myc I c Zn EI le Zn E I t~------. c-Myc I ---&.1-1.- c-Myc ~MT-c-Myc

B C Zn E

I- :E G1 92% G1 87% G1 33% ~ s 4% s 7% s 40% .....I LL (J G2 4% G2 6% G2 27% :E

CJ > G1 91% G1 70% G1 46%

-..... :EI I CJ s 6% S 21% S 41% LL I (JI- G2 3% G2 10% G2 13% :E ~

C MCF-7 /~MT MCF-7 /~MT-c-Myc IC E Zn I I C E Zn - Grb14 ------1 IRS-1 L...--=:...,__,:==------==----::z:::::111-...il Actin D MCF-7/~MT+ E MCF-7/~MT-c-Myc + Zn

36%S, 56 % G2/M 15%S, 59 % G2/M

I­ ~

IRS-1

Actin

Figure 6.4 Dissociation of estrogen regulation of Grbl4 from cell cycle progression. The regulation profile of the examined proteins was consistent in at least two independent experiments. A. Left panel Schematic of/';. MT-c-Myc construct showing the metal response element (MRE) from the hMTIIA promoter which confers zinc-inducibility. Right panel Regulation of c-Myc protein levels. The indicated cell lines were growth-arrested by ICI 182780 treatment (C) and then rescued with estrogen (E) or ZnSO4 (Zn). Cell lysates were prepared 6 h later and immunoblotted for c-Myc. B. DNA histograms showing cell cycle distribution 24 h after rescue. C. Western blot detection of Grbl4 48 h following stimulation. Lysates were Western blotted for actin as a I oading control. Membranes were also stripped and immunoblotted for IRS-1 protei n. D. Cells were arrested and rescued as in (A) in the presence of nocodazole to induce a G2/M block. Upper panel DNA histograms showing cell cycle distribution and lower panel Western blot analysis of Grbl4 and IRS-1 protein levels 48 ha fter rescue. Actin is shown as a loading control. was only reduced by estrogen, but not by Zn treatment (Figure 6.4D). By contrast, IRS-I protein displayed an opposite regulation profile and was increased by estradiol stimulation but not by Zn treatment. Thus these studies collectively suggested that Grb14 protein downregulation was specific to estrogen stimulation and was not a general effect of cell cycle progression. Furthermore, although c-Myc protein was rapidly upregulated by estrogen in MCF-7 cells (Prall et al., 1997 and Figure 6.4A), induction of c-Myc alone was not sufficient to decrease Grb14 expression levels.

6.6 Grb14 overexpression in MCF7/EcoR cells

The changes in Grb14 expression following estrogen- and insulin- stimulation of MCF-7 cells, where it was downregulated by estradiol and upregulated by the latter, indicated that Grb14 is modulated by these hormones in breast cancer cells. Therefore, it was of interest to examine the effect of Grb14 overexpression on regulation of breast cancer cell proliferation by these hormones. Retroviral infection was utilised to overexpress Grb14 in MCF-7 cells as this gene delivery method resulted in infection of a high percentage of target cells (see section 5.4) and in high exogenous expression of Grb14 proteins (Figure 5.lA). However, since the formerly employed retroviral vector (pLib) lacked selectable markers, in this chapter a different retroviral vector was utilised (pBabePuro) which contained a puromycin-resistant gene, thus allowing for selection of the overexpressing antibiotic-resistant cells. Unlike NIH-3T3 mouse fibroblasts, MCF-7 human cells do not express the murine receptor, which is required for binding of the mouse retrovirus and entry into the cell (Miller, 1996). Therefore, stable pools of MCF-7 cells expressing the murine retroviral receptor (MCF7/EcoR cells) were generated by transfecting a plasmid encoding this receptor (pWZLneoEcoR) into MCF-7 cells using Fugene 6 reagent. The following day, geneticin was added and selection was allowed to proceed for approximately 4 weeks (see section 2.3.3). By establishing stable pools, problems associated with clonal cell lines, such as clonal variation, could be avoided.

To generate Grb14-overexpressing stable pools using MCF7/EcoR cells, the sequence encoding the entire open reading frame of Grb14 together with a Flag-epitope tag was

121 excised from the eukaryotic expression vector Grb14/pRcCMVF,ag (Daly et al., 1996) and subcloned into the retroviral vector pBabePuro to produce the Grb14Fia/pBabePuro retroviral construct (section 2.1.5). This construct or the control pBabePuro vector were transfected into the Phoenix-Eco packaging cell line, after which the retroviral supernatant was collected and subsequently used to infect MCF7 /EcoR cells for 24 h. Approximately 48 h later when maximal expression of the retroviral-encoded protein was achieved, puromycin was added to the medium to select for antibiotic-resistant cells. Selection was allowed to proceed for approximately 6 days to obtain the resulting MCF-7/EcoR control and Grb14-overexpressing stable pools. To assess Grb14 expression in these pools, cells were serum-starved for 96 h, after which lysates were prepared and immunoblotted for Grb 14. This timepoint is identical to that at which flow cytometry samples were harvested and analysed in subsequent experiments (section 6.7). This enabled the detection of the endogenous Grb14 protein in lysates from pools infected with the control retrovirus. The Flag-tagged Grb14 was also detected in the sample from cells infected with the Grb14- encoding retrovirus, where it migrated slightly higher than the endogenous protein due to the epitope tag, thus confirming Grb14 overexpression in the appropriate stable pool (Figure 6.5A).

To examine whether Grb14 overexpression is affected by hormonal treatment, the MCF- 7/EcoR pools were maintained in 10% charcoal-stripped serum for 72 h and subsequently stimulated for 24 h with vehicle, insulin, estradiol or both hormones in combination. Cells were lysed and subjected to Western blotting analysis. Surprisingly, this indicated that levels of the overexpressed Grb14 were also regulated by steroid/growth factor treatment (Figure 6.5B). Insulin induced an increase in Grb14 levels approximately 1.6 fold above control and estradiol antagonised the insulin-induced increase. Thus, overexpressed Grb14 is regulated in a similar manner to the endogenous protein.

122 A Vector Grb14

Grb14

Actin

B C E l+E Grb14

Actin

Figure 6.5 Grbl4 overexpression in MCF-7/EcoR cells. A. MCF-7 cells expressing the mouse retroviral receptor (MCF-7/EcoR) were infected with either a Grbl4-encoding retrovirus (Grbl4) or the corresponding empty retrovirus control (Vector) and selected as described in section 2.3.4. The resulting stable pools were serum-starved for 96 h then lysed and imrnuno­ blotted for Grbl4 (This protocol is analogous to that used for cells subsequently analysed by flow cytometry in Figure 6.6). B. MCF-7/EcoR cells overexpressing Grbl4 were maintained in 10% charcoal-stripped serum for 72 h. Cells were then stimulated for 24 h with either vehicle (C), insulin (10 µg/ml , I), estradiol (10 nM, E) or both hormones (I+ E).Cell lysates were pre­ pared and Western blotted for Grbl4 or acti n. Similar results were obtained in at least two independent experiments. 6.7 Effect of Grbl4 overexpression on insulin-induced cell cycle pro~ression

In order to investigate the effect of Grb14 overexpression on insulin-induced cell cycle progression, the generated MCF-7 pools were utilised since MCF-7 cells exhibit a strong mitogenic response to insulin/lGF stimulation (Lai et al., 2001). Thus this provides an alternative model to the NIH-3T3 fibroblasts where insulin elicited a poor mitogenic effect (section 5.6). An insulin dose-response experiment was performed by starving the Grb14- overexpressing and control pools in serum-free medium prior to stimulation with insulin (0.1 ng/ml - 10 µg/ml) (Figure 6.6A,B). Cells were subsequently harvested and subjected to flow cytometry. Analysis of cell cycle phase distribution indicated that the percentage of S-phase cells was low in serum-starved, untreated samples for both vector and Grb14- overexpressing cell lines, (12% and 9% respectively). After stimulation with 0.1 ng/ml insulin there was no increase in the percentage of cells in S-phase for either pool relative to the untreated sample. However, at 1 ng/ml insulin stimulation, the S-phase fraction was increased slightly above control levels for the vector cells (15%), whereas the Grb14 pools demonstrated an increase at 10 ng/ml insulin (13%). A maximal increase in the percentage S-phase was obtained at the highest doses of insulin utilised (1 and 10 µg/ml). Interestingly, following insulin stimulation the percentage of S-phase cells was consistently reduced in the Grb14-overexpressing pools relative to control, for example at 1 µg/ml insulin the percentage S-phase for control and Grb14 cells was 30% and 22% respectively, and at 10 µg/ml insulin it was 31 % and 19%, respectively. Correspondingly, the percentage of cells in G 1 was much higher in Grb 14 cells relative to control at 1 µg/ml insulin (71 % relative to 63%, respectively) and at 10 µg/ml (73% relative to 63%, respectively). Thus, these results indicated that in response to insulin stimulation, Grb14-overexpressing cells contained a greater percentage of G I-phase cells, and a smaller S-phase fraction that is involved in DNA synthesis relative to control. This demonstrated that overexpression of Grb14 inhibited insulin-induced progression from Gl to S phase. The inhibitory effect of Grb14 on cell cycle progression induced by low and high doses of insulin was evident in replicate experiments.

124 A Control I (1 µg/ml) I (10 µg/m I)

G1 83% G1 63% G1 63% Vector S 12% s 30% s 31% G2 6% G2 7% G2 6%

G1 85% G1 71 % G1 73% Grb14 S 9% S 22% S 19% G2 6% G2 7% G2 8%

B

35 Vector

30 - < >--- Grb14 25

Q) (/) 20 C1l ..c 0. Cl) 15 0~ o- 10

5

0 0 0.1 10 100 1000 10 000 Insulin (ng/ml)

Figure 6.6 Effect of Grb14 overexpression on insulin-induced cell cycle progression. A. MCF-7fEcoR cells overexpressing Grb14 (Grb14) or infected with empty retrovirus (Vector) were serum-starved for 72 h and left untreated or stimulated with increasing concentrations of insulin as indicated. Cells were harvested 24 h post-stimulation and analysed by flow cytometry. DNA histograms for cells left untreated (Control) and cells treated with insulin concentrations of 1 µg/ml and 10 µg/ml are shown. The percentage of cells in G 1, S or G2 phases (representing an average of duplicates) is also indicated. B. Insulin dose-response curve showing percentage of cells in S phase for both vector-infected (squares) and Grb 14-overexpressing (diamonds) pools. These represent the average of duplicate samples. Results obtained with low and high insulin concentrations are representative of at least two independent experiments. 6.8 Effect ofGrb14 overexpression on both estroeen- and insulin-induced cell cycle proeression

In order to extend the analysis of the Grb14 effect on hormonal regulation of cell cycle progression, an 'estrogen rescue' model was utilised which was designed to facilitate the study of estrogen effects on cell cycle progression (Prall et al., 1997) by synchronising cells at G0-G 1 with antiestrogen treatment. In this system, insulin alone does not induce as large an increase in cell cycle progression as when the experiment is performed in the absence of antiestrogen, but the proliferative effects of estradiol and insulin in combination are readily apparent (Lai et al., 2001 ). Serum-starved MCF-7 cells were arrested with the antiestrogen ICI 182780 then cell cycle re-entry was stimulated with either insulin, estradiol or the two in combination. After 24 h, cells were harvested and analysed by flow cytometry. This indicated that the relative increase in S-phase upon insulin treatment was reduced for Grb14-overexpressing cells compared with vector control cells (Figure 6.7A), as observed under serum-free conditions without antiestrogen arrest (Figure 6.6B). Surprisingly, a decreased S-phase entry was also observed for Grb14-overexpressing cells stimulated with estradiol and when both hormones were used in combination. This demonstrated that Grb14 had an inhibitory effect on cell cycle progression stimulated by insulin and/or estradiol. Statistical analysis was performed using analysis of variance (ANOV A). This indicated that although inhibition of the percentage increase in S-phase by Grb14 overexpression did not reach statistical significance for individual treatments (p > 0.05), there was a definite inhibitory trend observed. However, inhibition by Grb14 overexpression was statistically significant when all the treatments were combined (p = 0.0259, two-factor ANOV A). Further, serum-induced cell cycle progression of the MCF-7 pools was investigated. Serum-starved cells were stimulated with 10% charcoal-stripped serum, harvested after 24 h and subjected to flow cytometry. Analysis of cell cycle phase distribution demonstrated that the percentage of cells in S-phase was similar for both vector and Grb14 pools. This indicated that serum-induced cell cycle progression was not affected by Grb14 overexpression (Figure 6.7B), demonstrating that Grb14 did not exert a non-specific effect on mitogenesis.

126 A 600 500 I (]) 400 "'(1) ~ c. D Vector CJ) -~ 300 l ti T D Grb14 C: 200 0~ 1 T 100 1

0 E l+E Treatment

B 40

30

D Vector

20

C 10 % Serum Treatment

Figure 6.7 Effect of Grbl4 overexpression on insulin- and estradiol-induced cell cycle progression. A. MCF-7/EcoR cells infected with Grb 14 retrovirus or the control retrovirus were cultured in serum-free medium for 48 h, arrested by ICI 182780 (10 nM) for 24 h then left untreated or rescued with either estradiol (l00nM,E), insulin (10 µg/ml,I) or both hormones in combination (I+E). Cells were then har­ vested 24 h post-stimulation and analysed by flow cytometry. The data is presented as the increase in S­ phase content as a percentage of the control (unstimulated) value. Values represent the average of tripli­ cate experiments± S.E. B. Grbl4 overexpression does not affect serum-induced cell cycle progression. Grbl4 overexpressing or control MCF-7/EcoR cells were cultured in serum-free medium for 72 hand either left untreated (C) or stimulated with charcoal-stripped serum (10% serum). The cells were har­ vested after 24 h for flow cytometric analysis. 6.9 Discussion

To date, there is little information available on regulation of Grbl4 expression in cells/tissues where it is endogenously expressed. Grbl4 was originally cloned from normal breast epithelial cells (Daly et al., 1996) and expression of Grbl4 correlated with ER positivity in breast cancer cells (Figure 6.1), therefore its regulation and functional role were studied in the hormone-responsive breast cancer cell line MCF-7. In cells maintained in charcoal-stripped serum, Grbl4 expression was modulated by estradiol and the pure antiestrogen ICI 182780 (Figure 6.2A). Estradiol decreased Grbl4 protein levels in contrast to IRS-1, a stimulatory insulin/lGF-I receptor substrate which was upregulated by estradiol (Figure 6.4C,D). It was interesting to find that the timecourse of Grbl4 downregulation by estradiol was slow relative to other estrogen-regulated proteins such as IRS-1, which exhibits a peak induction by 8 h upon estrogen stimulation (Lee et al., 1999), and c-Myc which reaches a transient maximum at 4 h (Prall et al., 1997). The slow regulation of Grbl4 is more comparable with that of the IGF-IR protein which is upregulated by estradiol at 48 h, but not earlier (Lee et al., 1999). The estrogen-stimulated downregulation of Grbl4 levels, in addition with the antiestrogen-induced increase in Grb14 expression was consistent with Grbl4's proposed inhibitory function. Indeed, estrogen/antiestrogen regulation has been demonstrated for inhibitory components of insulin/IGF signalling pathways. For example, IGFBP-3 and -5 belong to the IGFBP family of proteins which bind IGF's and modulate the interaction with their receptors. They are also expressed in and secreted by breast tumours (Perks and Holly, 2000). Similar to Grb14, expression levels of IGFBP-3 and -5 in MCF-7 cell-conditioned media were reduced by estrogen but enhanced by ICI 182780 (Huynh et al., 1996b; Huynh et al., 1996c).

If Grbl4 acts as a repressor of insulin/IGF signalling, then it is plausible to expect regulation of Grb14 expression by these growth factors, as demonstrated for the suppressors of cytokine signalling (SOCS) proteins, which are regulated by and inhibit signalling induced by various cytokine family members (Starr et al., 1997; Saito et al., 2000; Krebs and Hilton, 2001). It is noteworthy that they also exhibit a similar mode of

128 action to the Grb7 family proteins, for example SOCS-1 and -3 bind the Janus Kinases (JAKs) and inhibit their catalytic activity (Krebs and Hilton, 2001). Therefore, the regulation of Grbl4 expression in insulin/estrogen cross-talk was examined. Under serum­ free conditions, Grb14 protein levels were upregulated by insulin, yet this upregulation was hindered when the cells were co-stimulated with estradiol (Figure 6.3B). Presumably, addition of estrogen alone decreased Grb 14 expression in the presence of serum (Figure 6.2) since it antagonised the action of growth factors (eg. IGFs), whereas this effect of estrogen alone was not observed in serum-free culture (Figure 6.3B). Similarly, estrogen regulation of progesterone receptor levels required the presence of serum-containing medium and did not occur in serum-free culture (Katzenellenbogen and Norman, 1990).

The regulation of Grb14 expression by these different factors was similar to that recently reported for the cyclin-dependent kinase (CDK) inhibitor p21 WAFIICIPJ (Lai et al., 2001). Here, estrogen opposes the insulin-induced increase in p21 expression and recruitment into cyclin E/Cdk2 complexes, resulting in a synergistic increase in S-phase fraction when both hormones are used in combination. In addition to Grb14 regulation in insulin/estrogen cross-talk, regulation of growth factor signalling components has been demonstrated in heregulin/estrogen crosstalk. In this case, estrogen downregulates erbB2 expression in breast cancer cells, however heregulin inhibits the estrogen-induced decrease of this receptor (Grunt et al., 1995).

The data presented in this chapter represent the first demonstration of the regulation of Grb14 expression by growth factors or steroids. However, the mechanism by which Grbl4 expression is regulated by estrogen and/or insulin is unclear. Estrogen regulates the expression of a variety of proteins through the ER, which functions as a transcription factor. The promoter regions of certain estrogen-responsive genes contain consensus sites for ER­ binding, termed collectively as EREs (section 1.2.3). Previous studies indicated that the ER binds to a consensus 13 bp palindrome GGTCAnnnTGACC, which is a minimal functional ERE for the vitellogenin A2 gene (Klein-Hitpass et al., 1988). However, a large number of genes contain imperfect EREs, which do not match the exact A2 sequence, such as those reported for vitellogenin B1, AGTCAnnnTGACC (Walker et al., 1984); oxytocin,

129 GGTGAnnnTGACC (Sausville et al., 1985) and the pS2 ERE GGTCAnnnTGGCC (Nunez et al., 1987). In response to estrogen stimulation, regulation of the expression of insulin/IGF pathway components such as IRS-1 and IGF-IR occurs by a transcriptional mechanism, although it is unknown whether this is directly mediated by the ER (Lee et al., 1999). Therefore, it was of interest to ascertain whether the Grb14 promoter also contained sites that were putative EREs. A preliminary search for consensus EREs was performed utilising the 'working draft' sequence corresponding to the Grb14 promoter (1-52441 bp), obtained from the human genome project database. This revealed 3 candidate sequences (Figure 6.8) which did not match known consensus or imperfect EREs, two of which contained putative ERE half-sites. Although EREs serve to activate estrogen-induced gene expression, repression of gene transcription can be mediated via steroid response elements. For example, the negative glucocorticoid response element (nGRE) interacts with the glucocorticoid receptor (GR) to mediate glucocorticoid repression of the bovine prolactin and human osteocalcin genes (Sakai et al., 1988; Morrison and Eisman, 1993). In addition, a negative thyroid hormone response element (nTRE) may bind thyroid hormone receptor (TR) for ligand-dependent repression of the thyroid-stimulating hormone (3 (TSH(J) gene (Sasaki et al., 1999; Wu and Koenig, 2000). Further studies involving electromobility shift assays and estrogen-stimulated promoter transactivation studies are required to verify whether the Grb14 candidate sequences bind the ER and suppress estrogen-induced transcription within the Grb14 promoter sequence context.

In addition, estrogen also regulates transcription indirectly through the interactions of its receptor with other transcription factors. For example, AP-1 response elements are modulated by indirect interactions of ER and the AP-1 transcription factors, c-fos and c-jun (Osborne et al., 2000). Further, estrogen induction of cyclin D1 in ZR-75-1 cells occurs through activation of a cAMP response element (CRE) and is dependent on 3 Sp-1 binding sites, where ER interacts with Sp-1 (Castro-Rivera et al., 2001). The IRS-1 promoter contains AP-1 as well as SP-1 sites (Araki et al., 1995), although it also has four half­ estrogen response elements which function synergistically (Kato et al., 1992). Further analysis of the Grb14 promoter region for other transcription factor consensus sequences

130 ERE + + gg tcannntgacc I I I ! ! ! I I I I I 2694 7 t g t g a a t g t g a c c 26959 + + gg tcannntgacc I I II I ! ! I I I 11614 g g t c a t t c a g a c a 11626

+ + gg tcannntgacc I I I I ! ! ! I I I I 30404 g g C C a g C a t g a C g 30416

Figure 6.8 Analysis of the Grb14 promoter sequence for candidate EREs. The

Grb14 promoter (1 -5244lbp)1 was analysed for consensus sequences corresponding to EREs using MacYector and the WebAngis bestfit alignment program. Three sequences that displayed the highest homology were obtained, however each of these contained 2 nucleotides that did not match the ERE consensus sequence (shown in red). The sequence corresponding to 11 614-11626 bp differed in the 3' half-site, nucleotides from 26957-26959 differed in the 5' half-site and the sequence from 30404-30416 had one mismatch on both ends. These candidate sequences also did not match known imperfect EREs. Numbers flanking aligned sequences represent the nucleotide position in the Grbl4 promoter. Arrows indicate mismatch nucleotides in the candidate sequence. Putative ERE half-sites are underlined. 1- http://www.ncbi.nlm.nih.gov:80/entrez/query.fcgi?cmd=Retrieve&db=Nucleotide& list_uids= 18640713&dopt=GenBank revealed the presence of putative AP-1, AP-2, and c-jun sites. Inhibition of gene transcription by glucocorticoids is reported by interfering with AP-1 activity in AP-1 sites complexed with c-jun/c-fos heterodimers (Diamond et al., 1990; Jonat et al., 1990). Thus, indirect regulation of Grb14 by ER modulation of AP-1 activity may be one potential mechanism. However, regulation of the overexpressed Grb14 by insulin and estradiol (Figure 6.5B) suggests that modulation of Grb14 expression occurs at a post-translational level. Hence, insulin or estradiol may modulate Grb14 expression by altering protein turnover, where Grb14 levels are stabilised in response to insulin, but are degraded at a higher rate with estradiol treatment (Nawaz et al., 1999).

Since insulin upregulated Grb14 levels, the effect of Grbl4 overexpression on insulin­ induced cell cycle progression was investigated. While at low concentrations insulin acts primarily through its own receptor, at higher doses(> 1 µg/ml), binding to the IGF-IR also contributes to the mitogenic effects (Adamo et al., 1992). Flow cytometry analysis indicated that the response of Orb 14-overexpressing cells to insulin was reduced compared with control cells at both low and high insulin concentrations (Figure 6.6B), suggesting that Grb14 inhibited both IR- and IGF-IR-stimulated cell cycle progression. Since Grb14 binds the IR via its BPS domain and is the most potent member of the Grb7 family in inhibiting IR catalytic activity (Bereziat et al., 2002), the inhibitory effect of Grbl4 on insulin-induced cell cycle progression (Figure 6.6B) was likely to be mediated by inhibition of receptor substrate phosphorylation. This was previously demonstrated for IRS- 1, She, and Dok, where their phosphorylation on tyrosine residues was decreased in response to insulin in Grb14-overexpressing cells (Kasus-Jacobi et al., 1998; Hemming et al., 2001). Although in viva association with the IGF-IR was not detected (Figure 4.3A), Grb14 inhibited IGF-IR catalytic activity in vitro (Bereziat et al., 2002). Hence, Grbl4 may inhibit the IGF-IR through a direct interaction which is transient or unstable under conditions for co-immunoprecipitation analysis. Alternatively, the mechanism may be independent of receptor association. Interestingly, Grb14 affected Gl-S phase progression (Figure 6.6) in contrast to the Orb 10 isoform mGrb 1Oa, which caused accumulation of both serum-starved and IGF-1-stimulated cells in the Sand 02 phases of the cell cycle, thus delaying cell cycle progression (Morrione et al., 1997).

132 The effect of Grb 14 overexpression on insulin/estrogen cross-talk was investigated by a model system where cells arrested with ICI 182780 were stimulated to re-enter the cell cycle. This experimental paradigm was optimal for examining cell proliferation by estradiol stimulation alone or in conjunction with insulin (Lai et al., 2001). The findings revealed a novel Grb14 inhibitory effect on estradiol-induced cell cycle progression alone or in combination with insulin (Figure 6.7A). Although it was unclear how Grb14 suppressed estradiol signalling, one mechanism may be via inhibition of a reported direct activation of the IGF-IR by the ligand-stimulated ERa which may be mediated by an interaction between the receptors (Kahlert et al., 2000; Bereziat et al., 2002). Alternatively, basal activity of ERK 1/2 and Akt/PKB occurs in MCF-7 cells (Caristi et al., 2001; Lynch and Daly, 2002). Thus Grb14 may block the basal level of activation of growth factor signalling pathways which is 'permissive' for steroid-induced cell cycle progression (Lobenhofer and Marks, 2000). These pathways may converge on the ER itself (Kato et al., 1995) or activate transcription factors which regulate, along with the ER, key growth regulatory genes. A precedent for this is provided by IGFBP-1, which decreased estrogen-induced mitogenesis due to a reduction in ER activation (Lee et al., 1997). A possible stimulus for such basal signalling could arise from autocrine production of IGFs (Yee et al., 1988).

Therefore, these findings demonstrated novel hormonal regulation of Grb 14 expression which was consistent with its role as a repressor, and indicated that Grb14 represented an important growth regulator in breast cancer cells, since it inhibited both insulin/lGF-1- and estrogen-induced cell cycle progression. The Grb14 expression profile in breast cancer cells was similar to other components of the insulin/lGF pathway which positively correlate with ER status, such as IGF-IR (Peyrat and Bonneterre, 1992), IGFBP-2, -4 and -5 (Yee et al., 1994), as well as IRS-1 (Lee et al., 1999). Further, it should be noted that the absence of Grb14 expression in the majority of ER-negative lines indicates the loss of a repressor of insulin/IGF signalling in models of more advanced, estrogen-independent disease. Expression of IR/IGF-IR signalling components has been correlated with patient prognosis. For instance, in a study of 195 node-negative breast cancers, higher levels of IRS-1 were correlated with disease recurrence for patients with small tumours (Rocha et al., 1997) and for those with ER-positive tumours (Lee et al., 1999). Furthermore, patients with IR-

133 positive breast tumours were associated with a worse prognosis relative to those with IR­ negati ve tumours (Belfiore et al., 1996). Thus, it would be interesting to examine the relationship between Grb14 expression in primary breast cancers and patient outcome.

134 Chayter 7-9enerafVfscussion

7.1 Inhibition oflR signalling by GrblO and -14

The IR stimulates cell proliferation by recruiting several substrates, including the IRS proteins and She. These, in turn, recruit intracellular effectors resulting in the activation of the Ras/MAPK and PI-3 kinase cascades. The association of Grb14 with the IR has been demonstrated in CHO-IR cells (Kasus-Jacobi et al., 1998; Hemming et al., 2001; Figure 4.3B), in HepG2 cells (Hemming et al., 2001), as well as in rat liver extracts expressing endogenous levels of both Grb14 and the IR (Kasus-Jacobi et al., 1998). Upon insulin stimulation, Grb14 binds to the IR predominantly via its BPS domain (Kasus-Jacobi et al., 1998; Hemming et al., 2001; Figure 4.3C) and inhibits its catalytic activity (Bereziat et al., 2002). Grb14 also elicits a decrease in insulin-induced tyrosine phosphorylation of IRS-1, p62Dok, and She (Kasus-Jacobi et al., 1998; Hemming et al., 2001), and delays the activation of Akt, ultimately leading to significant reduction in glycogen synthesis (Kasus­ Jacobi et al., 1998). Moreover, Grb14 also inhibits the activation of Erkl/2, leading to reduced DNA synthesis in CHO-IR cells (Kasus-Jacobi et al., 1998; Bereziat et al., 2002). Grb14 overexpression also inhibits insulin-induced cell cycle progression in MCF-7 cells (Figure 6.6). These events are represented schematically in Figure 7.1.

Grb14 and potentially GrblO may generally function in an analogous manner to SOCS 1-3, which inhibit receptor signalling by abrogating the catalytic activity of JAKs, and possibly by competing with STATS for receptor binding sites. JAKs are constitutively associated with cytokine receptors that are devoid of intrinsic kinase activity. Cytokine binding results in receptor clustering, whereby the juxtaposition of JAKs allows their activation by crossphosphorylation. Activated JAKs then phosphorylate the receptor tail, allowing for the recruitment of signalling proteins, including STA Ts (Krebs and Hilton, 2001 ). SOCS-1 binds Tyr-1007 in the JAK2 activation loop and inhibits its catalytic activity via the SH2 domain and an N-terminal adjacent 24 residues. Interestingly, 12 conserved amino acids within the latter region are essential for inhibition of JAK2 activity, and are termed the

135 v SH2 Nedd4? Degradation?•

?. Localisation?

Glucose transport Protein synthesis Gene transcription Mitogenesis

Figure 7.1 Model of Grbl4 function in insulin signalling (modified from Withers et al., 2000). Please refer to section 7.1 for details. Yellow arrows denote stimulation and red bars represent inhibition or a delay in activation (for Akt). kinase inhibitory region (KIR). This region is postulated to act as a pseudosubstrate, thus preventing access of JAK substrates (Mooney et al., 2001). This inhibitory mechanism entails similarities to the Grb7 family, where the BPS domain is essential for inhibiting the receptor catalytic activity, although the SH2 domain is not implicated to play a major role in the inhibitory process (Bereziat et al., 2002). Moreover, SOCS-1 and -6 bind the IR and, similar to Grb14, they do not inhibit IR phosphorylation but alter substrate phosphorylation. In addition, SOCS-1 and -6 alter the activation of Akt and Erkl/2 (Mooney et al., 2001).

Another mechanism for inhibition by SOCS proteins may entail targeting their substrates for degradation, since they bind via the SOCS box to elongins B and C which couple to the proteasomal degradation pathway. In accordance with this, GrblO constitutively interacts with the E3 ubiquitin ligase Nedd4 in mouse embryo fibroblasts (Morrione et al., 1999). GrblO also interacts with the IGF-IR (Morrione et al., 1996), which undergoes ubiquitylation and is degraded by the 20S proteasome (Sepp-Lorenzino et al., 1995), thus implicating the adaptor GrblO in localising Nedd4 to the IGF-IR (Rotin et al., 2000). In addition, Grb14 binds Nedd4 via its SH2 domain in the yeast two-hybrid system (Lyons et al., 2001). If this interaction occurs in vivo, Grb14 may also inhibit signalling by targeting binding partners for the proteasomal degradation pathway. However, there is accumulating evidence in yeast that ubiquitin itself may serve as an internalisation signal, although this has not been conclusively demonstrated in mammalian cells. For instance, the S. cerevisiae homolog of Nedd4, Rsp5p, is implicated in ubiquitin-dependent endocytosis of several yeast plasma membrane proteins, including the general amino acid permease, Gap 1p, and the transporter Fur4p uracil permease. Nedd4 binds to conserved PY motifs of the epithelial Na+ channel (ENaC) and the guanine-nucleotide exchange factor CNrasGEF (Rotin et al., 2000; Pham and Rotin, 2001). PY motifs are implicated in mediating ENaC internalisation by targeting the channel for clathrin-mediated endocytosis and subsequent degradation. Thus, the role of Nedd4/Rsp5p in endocytosis may occur in addition to its function in ubiquitylating endocytosed proteins (Rotin et al., 2000). Further studies are required to clarify whether Grb14 is involved in mediating internalisation and/or ubiquitylation of RTKs.

137 In addition to the BPS and SH2 domains, Grb 14 has other functional modules that may mediate interactions with downstream effectors. For instance, Grb14 binds to the recently identified tankyrase-2 via the N-terminal 110 residues (Lyons et al., 2001). Tankyrase-2 lacks a domain containing histidine, proline and serine residues (HPS), but otherwise has a similar structure to tankyrase-1, an ADP-ribose transferase which possesses an N-terminal HPS domain, a region containing ankyrin repeats followed by a SAM module, and a C­ terminal poly (ADP-ribose) polymerase (PARP) domain (Smith et al., 1998). Tankyrase-1 is localised to the Golgi, it interacts with insulin-responsive amino peptidase (IRAP) in Glut 4 vesicles and is phosphorylated by MAPK upon insulin stimulation (Chi and Lodish, 2000). Tankyrase-2 also binds IRAP, and may functionally overlap with tankyrase-1 since both proteins interact when overexpressed in BOSC cells and co-localise in vesicular compartments (Sbodio et al., 2002). The interaction of Grb14 with tankyrase-2 may be implicated in vesicle trafficking and in the subcellular localisation of Grb14 (Lyons et al., 2001). Also, the conserved proline-rich sequence PS/AIPNPFPEL may contribute to tankyrase binding or may interact with other proteins (Daly et al., 1996). In this context, the GrblO proline-rich motif associates with the c-Abl SH3 domain in vitro (Frantz et al., 1997).

In addition, the Grb14 GM region harbours functional modules that may interact with signalling effectors. For instance, the Grb14 RA-like domain can potentially bind GTPases, although its function has not been established in the Grb7 family thus far (Wojcik et al., 1999). Furthermore, a recent study reported that the Grb7 PH domain interacts with phosphoinositides, and this is required for Grb7 to stimulate cell migration. The Grb7 PH domain has 56% amino acid identity relative to Grb14, although the latter is not involved in cell migration (Han et al., 2000). Nonetheless, there is a possibility the Grb14 PH domain may mediate phospholipid interactions to fulfil another function such as vesicle trafficking. Although all members of the Grb7 family bind the IR, inhibitory functions have only been demonstrated for GrblO and -14. The latter proteins are structurally more similar than Grb7 (Daly et al., 1996). Does this imply that GrblO and -14 fulfil the same role? This may be the case in some tissues where they are both highly expressed, such as the pancreas. On the other hand, there are differences in the expression of GrblO and -14 in certain

138 tissues. For example, Grb14 exhibits high expression in the liver, kidney and gonads, whereas GrblO is expressed at low levels in these tissues. Therefore, Grb14 may functionally predominate in these tissues. Alternatively, Grb14 may be the main systemic physiological inhibitor of IR-mediated effects since it is more potent at inhibiting IR catalytic activity (Bereziat et al., 2002). This is supported by the finding that insulin­ regulated glucose homeostasis is improved in Grb14 knockout mice (Daly et al., unpublished). However, GrblO may functionally prevail in certain tissues when Grb14 is expressed at lower levels. This is yet to be investigated in the Grb14 null mice, and it was noted for IRS-2, which partially compensates for insulin and IGF-1 signalling in IRS-1 null mice (Araki et al., 1994; Patti et al., 1995). Although IRS-1 predominantly mediates insulin action in muscle and adipose tissue, IRS-2 plays a major role in the murine liver in the regulation of PI-3 kinase (Withers and White, 2000). IRS-2 phosphorylation is also essential for insulin-stimulated glucose metabolism and mitogenesis in hepatocytes (Rother et al., 1998).

Upon activation, insulin-stimulated signalling components undergo translocations to distinct cellular compartments. For example, IRS-1 and -2 translocate from the membrane to the cytosol, whereas 40% of the total p85 protein pool is recruited from the cytosol to the membrane compartment (Inoue et al., 1998). More importantly, localisation of IRS proteins in the basal state is implicated to play a significant role in their subsequent activation and mediation of insulin signalling. Cell fractionation studies have demonstrated that in the basal state, IRS 1/2 associate with an insoluble multi protein complex in the low density microsome (LDM) fraction. This cell fraction, also denoted as the high speed pellet (HSP), includes small vesicles, the cytoskeleton and other large protein complexes. Additional studies suggest that IRS proteins are associated with the cytoskeleton, which may facilitate their interaction with the IR upon insulin stimulation. Further, release of IRS proteins from the cytoskeleton may be associated with insulin resistance (Clark et al., 2000; Whitehead et al., 2000). This raises the question of whether Grb14 compartmentalisation may be important for receptor interaction and subsequent inhibition of insulin actions. In agreement with this, Grb14 was mainly detected in the LDM fraction in serum-starved DU145 prostate cancer cells (Lyons et al., 2001). Therefore, additional cell fractionation and confocal

139 microscopy studies are required in Grb14-expressing cell lines, including serum-starved or insulin-treated breast cancer cells, since Grb14 inhibits insulin-stimulated cell cycle progression in MCF-7 cells (Figures 6.6, 6.7).

7.2 The function ofGrb14 in signalling by other RTKs

It was interesting to find that in addition to the IR, Grb14 also interacts with the FGFR-1, since recruitment to this receptor has not been reported for any other members of the Grb7 family. Activation of the Ras/MAPK signalling cascade is crucial for FGF-mediated mitogenesis of many cell types (Boilly et al., 2000). Grbl4 was reported by another group to inhibit FGF-stimulated proliferation in fibroblasts (Reilly et al., 2000), although this was not observed in this study (Figure 5.2). However, the recruitment of Grb14 to the FGFR-1 is distinct since it is mediated by the SH2 domain, but does not involve the BPS module (Figure 4.6). How might Grb14 inhibit signalling by FGFR-1 and other potential interacting RTKs (not yet identified) in a BPS-independent manner? One possible mechanism may involve the association of Grb14 with Nedd4, which is mediated via the SH2 domain in a phosphotyrosine-independent manner (Lyons et al., 2001). The interaction of Grb14 with Nedd4 may target receptors for subsequent endocytosis and degradation, thus affecting their signalling potential (Rotin et al., 2000). Alternatively, Grb14 may function in a similar manner to the SOCS proteins and compete with substrates for receptor binding sites, thereby inhibiting mitogenic signalling. For example, CIS inhibits STAT5b activation by growth hormone (GH). This may occur by competition for binding to overlapping phosphotyrosine residues on the GH receptor, where CIS blocks STAT5b from binding to the activated receptor (Krebs and Hilton, 2001).

7.3 DiUerential interactions of Grb14 isoforms with RTKs and eUectors

Interestingly, results in chapters 4 and 5 indicated that Grb14 ~ was unable to bind to the FGFR-1 (Figure 4.6), and did not markedly affect FGF-induced DNA synthesis or cell cycle progression (Figures 5.2 and 5.3, respectively). In addition, Grb14 ~ was not functionally equivalent to a Grb14 SH2 R466K point-mutant which also fails to bind the

140 FGFR-1, but results in significant upregulation of FGF-induced cell proliferation (Reilly et al., 2000). However, although Grb14 ~ cannot mediate receptor binding via its SH2 domain, the intact N-terminus is able to associate with other proteins, such as tankyrase 2 as mentioned earlier. Also, the intact PH and RA-like domains can potentially interact with phosphoinositides and GTPases, respectively. Thus, if Grb14 ~ and the point-mutant are likely to mediate interactions with effectors through their PH and/or N-terminal regions, how can they function distinctly in FGFR-1-mediated cell proliferation? One possible model involves Nedd4. The Grb14 SH2 R466K mutant is likely to bind Nedd4 since the interaction of the GrblO SH2 with Nedd4 is constitutive and independent of tyrosinephosphorylation (Morrione et al., 1999). Therefore, the point-mutant may sequester Nedd4 and prevent endogenous Grbl4 a from coupling Nedd4 to the receptor, thus contributing to the upregulation of FGFR-1 signalling. However, Grb14 ~ is unlikely to associate with Nedd4 due to its truncated SH2 domain and cannot mediate this dominant negative effect. On the other hand, in response to insulin stimulation, Grb14 ~ is capable of binding the IR via the BPS domain, albeit less well than Grb14 a (Figure

4.3B,C). Therefore, Grb14 ~ is expected to inhibit IR signalling since the BPS, but not the SH2 domain, inhibits the IR kinase activity. However, it is not likely to mediate receptor inhibition to the same extent as the full-length isoform. Future studies are required to determine the influence of Grb14 ~ on insulin-stimulated mitogenesis and glycogen synthesis.

Furthermore, an important functional difference for Grb14 ~ may be due to oligomerisation, which could represent a general feature of the Grb7 family.

Oligomerisation was reported for GrblO ~/~ and involves an interaction between the N­ terminus of one protein and the PH/BPS-SH2 regions of another. This may serve to regulate the interaction of the adaptor proteins with the IR, where the tetramerised GrblO molecules present in the basal state dissociate upon binding to the ligand-stimulated receptor (Dong et al., 1998). However, the intact BPS-SH2 region is required to mediate oligomerisation, which implies that Grb14 ~, with a truncated SH2, may oligomerise less well than the full-length protein prior to ligand stimulation. This can possibly result in constitutive signalling through downstream effectors, since Grb14 ~ retains the PH and N-

141 terminal regions. The differential interactions of Grb14 isoforms with RTKs and regulators/effectors are depicted in Figure 7.2.

7.4 The role of Grb7 in RTK si,:nallin,:

Unlike GrblO and -14, which have a well established role in inhibiting insulin-stimulated metabolic and mitogenic effects, Grb7 functions in cell migration and does not affect cell cycle progression. The interaction of Grb7 with FAK via its SH2 domain is implicated in enhancing migration of fibroblasts towards fibronectin. This is in accordance with studies correlating enhanced expression of Grb7 proteins with invasive or metastatic esophageal cancer cells (Tanaka et al., 1998b). This suggests that the Grb7 function differs from GrblO and -14, which is also supported by the inability of the latter proteins to stimulate cell migration (Han et al., 2000). Grb7 is overexpressed in breast cancer cell lines and tumours in concert with erbB2 and interacts with the latter in breast cancer cells (Stein et al., 1994). However, the SH2 domain of Grb7 binds to Tyr-1139 of erbB2, which represents the main Grb2 binding site (Janes et al., 1997). In addition, Grb7 interacts with the Grb2 binding site of several proteins including SHP-2, She, Tek and the c-kit/stem cell factor receptor (Stein et al., 1994; Keegan and Cooper, 1996; Jones et al., 1999; Thommes et al., 1999). Both Grb7 and Grb2 interact with pYXN motifs in binding partners (Songyang et al., 1994; Pero et al., 2002). Therefore, although Grb7 is implicated in cell invasion, it may also act as an inhibitor where it competes for Grb2 binding sites on effectors in mitogenic signalling cascades. This is yet to be investigated.

7.5 Re,:ulation of Grb14 and its role in the cross-talk between insulin/lGF-1 and estro,:en si,:nallin,:

In breast cancer cells, the mitogenic functions of insulin/IGF-I are well established. Moreover, estrogen acts as a potent mitogen for ER-positive breast cancer cells. To elicit cell growth, estrogen upregulates stimulatory components such as cyclin DI and the IGF­ IR. This steroid also decreases inhibitory molecules such as IGFBP-3 and p21. While previous studies examining the Grb7 family focussed on their role in growth factor-initiated

142 0 ligomerisation? ?. ~FGFR-1 ---n• Grb14 a PH SH2 n T2 PI? Nedd4?

? IR Oligomerisation? • 1J ...

Grb14 ~ PH

PI?

Oligomerisation? ?. IR ~ u * Grb14 R466K PH BPS SH2 n ? • T2 PI? Nedd4?

Figure 7.2 Interactions of Grbl4 isoforms and the Grbl4 R466K SH2 mutant with potential regulators and effectors [T2 tankyrase 2; PI phosphoinositide(s)]. Please refer to section 7.3 for details. signalling cascades, results presented in Chapter 6 represent the first data to implicate a potential role for the Grb7 family in steroidal/antisteroidal action. Consistent with its implicated role as an inhibitor of cell proliferation in other cellular systems (Kasus-Jacobi et al., 1998; Hemming et al., 2001), and in estradiol-stimulated breast cancer cells (Figure 6.7), Grb14 levels are decreased by estradiol and upregulated above control levels by the pure antiestrogen ICI 182780 in MCF-7 cells. This suggests that Grb14 downregulation is involved in estradiol-elicited growth and increased Grb14 levels contribute to growth inhibition by ICI 182780. IGFBP-3 expression is also increased by this antiestrogen and decreased by estradiol. Interestingly, reduction of IGFBP-3 expression by antisense in MCF-7 cells increases cell proliferation in the presence of ICI 182780 (Huynh et al., 1996b). Hence, future studies could involve abolishing Grb14 expression using antisense oligonucleotides in MCF-7 cells and testing the resulting effect on growth inhibition induced by ICI 182780.

On this note, it is important to investigate regulation of Grb14 expression by selective estrogen receptor modulators (SERMS). These differ from pure antiestrogens in that they function as estrogen antagonists in certain tissues, but act as agonists in others, and are exemplified by the compounds raloxifene and tamoxifen. Tamoxifen, which is the standard treatment for patients with ER-positive breast cancer, opposes estrogen action in the breast, but has estrogenic activity in the uterus (Hall et al., 2001). Therefore, it would be interesting to examine if the Grb14 regulation profile differs between tissues where SERMS are acting as estrogen agonists relative to sites where they are estrogen antagonists. Overexpression of IGF signalling components, such as IRS-1, are associated with antiestrogen resistance (Salerno et al., 1999), hence it would also be important to examine whether Grb14 expression is reduced in tumours that have become resistant to antiestrogen treatment.

Bidirectional interactions between steroid and growth factor signals include modulation of RTK effectors at different levels in the signalling cascade, and may depend on the relative ratio of signalling components. For instance, estrogen-stimulated downregulation of erbB2 levels is most pronounced in MCF-7 cells, which contain high levels of ER and low erbB2

144 expression. On the other hand, hormone-induced reduction in erbB2 expression is weaker in BT-474 cells, which have a low ER, high erbB2 content. Interestingly, heregulin blocks the estrogen-mediated downregulation of erbB2 in BT-474 cells but is partially effective in MCF-7 cells (Grunt et al., 1995). Modulation of Grb14 levels by estrogen and insulin (Figure 6.3B) may be similarly affected by the total cellular levels of the IR, ER and/or Grb14. To investigate this further, it would be necessary to screen several breast cancer cell lines with varying expression of either receptor or the adaptor protein.

How is Grb14 expression regulated by insulin and estradiol? The observation that both endogenous and overexpressed Grb14 are regulated (Figures 6.3B, 6.5B) suggests modulation at the post-translational levels. This is plausible since the overexpressed Grb14 cDNA, which is contained in the CMV-promoter based expression plasmid pRcCMVF,ag• lacks the untranslated 5' promoter region which harbours regulatory sequences that allow transcriptional control. However, this does not completely rule out additional control at the transcriptional level for the endogenous gene (section 6.9 and Figure 6.8). Additional studies are required to determine if Grb14 mRNA levels are affected by hormonal stimulation. Transcriptional regulation may be further investigated by undertaking nuclear run-off transcription assays, using nuclei isolated from untreated or hormone-stimulated cells. In addition, potential post-transcriptional mechanisms may also include alterations in Grb14 mRNA stability, which can be investigated by analysis of the Grb14 mRNA from cells left untreated or stimulated with hormone in the presence or absence of actinomycin D (which inhibits transcription). This will identify whether hormonal regulation requires Grb14 transcription. To examine whether regulation occurs at the protein level, pulse-chase analysis may be undertaken, where cells untreated or stimulated with hormone are metabolically labelled with 35S methionine and chased with an excess of unlabelled methionine for different timepoints. This will determine the Grb14 protein half-life and identify whether hormonal regulation affects protein stability.

It is also well known that in addition to insulin/IGF-I, there are other mitogenic growth factors that stimulate breast cancer cell growth. Hence, to extend the studies of Grb 14 modulation and function in breast cancer, it is also necessary to examine regulation of

145 Grb14 expression by other known breast cancer mitogens such as EGF, FGF and heregulin. It would also be interesting to examine whether other Grb7 family members are regulated by insulin and/or estradiol in a similar manner to Grb14. In addition, their function in response to steroid/growth factor cross-talk is also of interest. Since GrblO has a similar function to Grb14, as it also inhibits insulin/IGF-1 signalling, then regulation of its expression is expected to be similar to Grb14, although investigations of GrblO expression have not been undertaken in breast cancer cells.

Interestingly, CIS/SOCS 1-3 and Grb14 are upregulated in response to cytokine/growth factor stimulation (Krebs and Hilton, 2001 and Figure 6.3B). However, the SOCS proteins are tightly regulated, with rapid and transient induction. They are depleted by 180 min possibly through the proteasome. In contrast, Grb14 protein levels are increased 24 h following insulin stimulation. This implies that Grb14 expression is regulated differently to the SOCS proteins. Therefore, the SOCS proteins may be part of an immediate negative feedback loop in insulin signalling. Similarly, Grb14 also acts early in response to insulin, since it binds the IR at 1 min following stimulation, but it then dissociates (Hemming et al., 2001 and Figure 4.4A), leading to decreased activation of effectors (IRS-I, She, p62Dok, Akt, Erkl/2) (Kasus-Jacobi et al., 1998; Hemming et al., 2001; Bereziat et al., 2002). This dissociation is not consistent with a classical negative feedback mechanism, and suggests that Grb14 acts to 'fine tune' immediate signal output. However, the increased expression of Grb14 observed following 24 h stimulation (Figure 6.3B) is consistent with negative feedback, and suggests that Grb14 mediates inhibition at later timepoints with consistent mitogenic stimulation in a physiological setting.

Cooperativity between insulin/lGF-1 and estrogen signalling involves cross-modulation at several levels (Figure 7.3). First, estrogen regulates the expression of components of the IR/IGF-IR signalling cascade. For example, it increases the expression of IGF-IR and IRS­ I (Stewart et al., 1990; Stewart et al., 1992; Lee et al., 1999; Molloy et al., 2000), and decreases the expression of Grb14 (Figure 6.2). Second, ligand-bound ER may activate the IGF-IR (Kahlert et al., 2000), providing a potential point of action for Grb14 in inhibition of estrogen-induced mitogenesis (Figure 6.7 A ; Bereziat et al., 2002). The cross-talk is

146

as as

upregulation upregulation

depicted depicted

is is

denote denote

bar bar

receptor receptor

Estrogen Estrogen

and and

the the

diagram, diagram,

arrow arrow blue

the the

The The

transcription transcription

details. details.

In In

for for

7.1. 7.1.

Mitogenesis Mitogenesis

Gene Gene

7.5 7.5

n n

Figure Figure

cross-talk. cross-talk.

in in

sectio

to to

stated stated

estrogen estrogen

as as

refer refer

and and

are are

se se

Plea

bars bars

red red

and and

insulin/IGF-1 insulin/IGF-1

in in

respectively. respectively.

arrows arrows

function function

14 14

Yellow Yellow

expression, expression,

Grb Grb

of of

of of

IGF-IR. IGF-IR.

or or

Model Model

IR IR

7.3 7.3

the the

downregulation downregulation

Figure Figure

and and

either either

c c MAPK MAPK also bidirectional. Insulin/lGF signalling enhances estrogen action through Akt- and Erk­ mediated phosphorylation of the ER (Kato et al., 1995; Martin et al., 2000). These pathways also regulate the activity of transcription factors which act in concert with the ER to regulate expression of growth regulatory genes eg. cyclin D 1. Finally, a basal level of activation of these latter pathways is likely to be permissive for estrogen action. Since this basal activity may be derived from the IR/IGF-IR (eg. due to autocrine production of IGFs), this provides a further explanation for the ability of Grb 14 to inhibit estrogen-stimulated mitogenesis (Figure 6.7A).

Grb14 expression is regulated by insulin and estradiol after chronic exposure times (24-48 h), thereby implying that insulin upregulates Grb14 which acts as a modulator that opposes continuous mitogenic action. By contrast, estradiol stimulation decreases Grb14 levels and opposes the insulin-induced Grb14 upregulation (Figure 6.3). This is analogous to the regulation of the cdk inhibitor p21 by these hormones, the expression of which is increased by insulin and downregulated by estradiol (Lai et al., 2001). Moreover, the ability of estradiol to antagonise the insulin-induced increase in p21 is central for enhancing cell proliferation initiated by these hormones. Results obtained in this thesis provide further evidence that insulin and estrogen act at another level, i.e. they regulate the expression of RTK signalling components in addition to cell cycle regulatory proteins in order to elicit and maintain cell proliferation. This is of relevance to breast cancer progression where prolonged exposure to mitogens (circulating in the serum, or produced from neoplastic or stromal components) leads to tumour growth and metastasis. This is in accordance with the correlation of Grb14 expression in ER-positive breast cancer cells as opposed to ER­ negative cells (Figure 6.1), since the latter are associated with worse prognostic features and may lack negative feedback signals. On the other hand, one may expect Grb14 levels to be decreased in ER-positive cells because of its downregulation by estrogen. The fact that Grb14 is not downregulated in ER-positive cells implies that there are other mechanisms that counteract estrogen action and allow expression of mitogenic inhibitors in these cell lines. Therefore, results presented in this thesis implicate a novel role for the adaptor Grb14 in hormonal regulation of breast cancer cells. Given that other components of the IR/IGF­ IR signalling pathway have been correlated with patient prognosis (Belfiore et al., 1996;

148 Rocha et al., 1997; Lee et al., 1999), further investigations entailing the use of primary and metastatic breast cancer specimens are required. This will allow correlation of Grbl4 expression with patient outcome and with tumour progression.

149 'References

Adamo, M, Roberts, CTJ and LeRoith, S (1992). How distinct are the insulin and insulin­ like growth factor I signalling systems? Bio/actors, 3, 151-157.

Aghazadeh, B and Rosen, MK (1999). Ligand recognition by SH3 and WW domains: the role of N-alkylation in PPII helices. Chem. Biol., 6, R241-R246.

Aguirre, V, Werner, ED, Giraud, J, Lee, YH, Shoelson, SE and White, MF (2002). Phosphorylation of ser307 in insulin receptor substrate-I blocks interactions with the insulin receptor and inhibits insulin action. J. Biol. Chem., 277, 1531-1537.

Altucci, L, Addeo, R, Cicatiello, L, Dauvois, S, Parker, MG, Truss, M, Beato, M, Sica, V, Bresciani, F and Weisz, A (1996). 17B-Estradiol induces cyclin D1 gene transcription, p36D1-p34cdk4 complex activation and p105Rb phosphorylation during mitogenic stimulation of G(l)-arrested human breast cancer cells. Oncogene, 12, 2315-2324.

Andrecheck, ER and Muller, WJ (2000). Tyrosine kinase signalling in breast cancer. Tyrosine kinase -mediated signal transduction in transgenic mouse models of human breast cancer. Breast Cancer Res., 2, 211-216.

Angrist, M, Bolk, S, Bentley, K, Nallasamy, S, Halushka, MK and Chakravarti, A (1998). Genomic structure of the gene for the SH2 and pleckstrin homology domain-containing protein GRBJ0 and evaluation of its role in Hirschsprung disease. Oncogene, 17, 3065- 3070.

Araki, E, Haag, BL, Matsuda, K, Shichiri, M and Kahn, CR (1995). Characterisation and regulation of the mouse insulin receptor substrate gene promoter. Mol. Endocrinol., 9, 1367-1379.

150 Araki, E, Lipes, MA, Patti, ME, Bruning, JC, Haag, Br, Johnson, RS and Kahn, CR (1994). Alternative pathway of insulin signalling in mice with targeted disruption of the IRS-1 gene. Nature, 372, 186-190.

Aronica, SM and Katzenellenbogen, BS (1993). Stimulation of estrogen receptor-mediated transcription and alteration in the phosphorylation state of the rat uterine estrogen receptor by estrogen, cyclic adenosine monophosphate, and insulin-like growth factor-I. Mol. Endocrinol., 7, 743-752.

Arteaga, CL, Johnson, MD, Todderud, G, Coffey, RJ, Carpenter, G and Page, DL (1991). Elevated content of the tyrosine kinase substrate phospholipase C-yl in primary human breast carcinomas. Proc. Natl. Acad. Sci. USA, 88, 10435-10439.

Backer, JM, Myers, MGJ, Shoelson, SE, Chin, DJ, Sun, XJ, Miralpeix, M, Hu, P, Margolis, B, Skolnik, EY and Schlessinger, J (1992). Phosphatidylinositol 3'-kinase is activated by association with IRS-1 during insulin stimulation. EMBO J., 11, 3469-3479.

Bai, RY, Jahn, T, Schrem, S, Munzert, G, Weidner, KM, Wang, JYJ and Duyster, J (1998). The SH2-containing adapter protein GRBlO interacts with BCR-ABL. Oncogene, 17, 941- 948.

Bange, J, Prechtl, D, Cheburkin, Y, Specht, K, Harbeck, N, Schmitt, M, Knyazeva, T, Muller, S, Gartner, S, Sures, I, Wang, H, Imyanitov, E, Haring, H, Knayzev, P, Iacobelli, S, Hofler, H and Ullrich, A (2002). Cancer progression and tumor cell motility are associated with the FGFR4 Arg(388) allele. Cancer Res., 62, 840-847.

Bar-Sagi, D and Hall, A (2000). Ras and Rho GTPases: A family Reunion. Cell, 103, 227- 238.

Belfiore, A, Frittitta, L, Constantino, A, Frasca, F, Pandini, G, Sciacca, L, Goldfine, ID and Vigneri, R (1996). Insulin receptors in breast cancers. Ann. N. Y. Acad. Sci., 784, 173-188.

151 Bereziat, V, Kasus-Jacobi, A, Perdereau, D, Cariou, B, Girard, J and Bumol, A (2002). Inhibition of insulin receptor catalytic activity by the molecular adaptor Grb14. J. Biol. Chem., 277, 4845-4852.

Biscardi, JS, Ishizawar, RC, Silva, CM and Parsons, SJ (2000). Epidermal growth factor receptor and c-Src interactions in breast cancer. Breast Cancer Res., 2, 203-210.

Bjorge, JD, Pang, A and Fujita, DJ (2000). Identification of protein-tyrosine phosphatase lB as the major tyrosine phosphatase activity capable of dephosphorylating and activating c-Src in several human breast cancer cell lines. J. Biol. Chem., 275, 41439-41446.

Blanckaert, VD, Hebbar, M, Louchez, MM, Vilain, MO, Schelling, ME and Peyrat, JP (1998). Basic fibroblast growth factor receptors and their prognostic value in human breast cancer. Clin. Cancer Res., 4, 2939-2947.

Boilly, B, Vercoutter-Edouart, AS, Hondermarck, H, Nurcombe, V and Le Bourhis, X (2000). FGF signals for cell proliferation and migration through different pathways. Cytokine Growth Factor Rev., 11, 295-302.

Bolen, JB (1993). Nonreceptor tyrosine protein kinases. Oncogene, 8, 2025-2031.

Boxer, LM and Dang, CV (2001). Translocations involving c-myc and c-myc function. Oncogene, 20, 5595-5610.

Bundred, NJ (2001). Prognostic and predictive factors in breast cancer. Cancer Treat. Rev., 27, Burgering, BMT and Coffer, PJ (1995). Protein kinase B (c-Akt) in phosphatidylinositol-3-OH kinase signal transduction. Nature, 376, 599-602.

Burke, D, Wilkes, D, Blundell, TL and Malcolm, S (1998). Fibroblast growth factor receptors. Trends Biochem. Sci., 23, 59-62.

152 Campbell, D, deFazio, A, Sutherland, R and Daly, R (1996). Expression and tyrosine phosphorylation of EMSl in human breast cancer cell lines. Int. J. Cancer, 68, 485-492.

Campbell, SL, Khosravi-Far, R, Rossman, KL, Clark, GJ and Der, CJ (1998). Increasing complexity of Ras signaling. Oncogene, 17, 1395-1413.

Cantley, LC (2002). The phosphoinositide 3-kinase pathway. Science, 296, 1655-1657.

Cappellen, D, De Oliveira, C, Ricol, D, De Medina, SGD, Bourdin, J, Sastre-Garau, X, Chopin, D, Thiery, JP and Radvanyi, F (1999). Frequent activating mutations of FGFR3 in human bladder and cervix carcinomas. Nat. Genet., 23, 18-20.

Caristi, S, Galera, JL, Matarese, F, Imai, M, Caporali, S, Cancemi, M, Altucci, L, Cicatiello, L, Teti, D, Bresciani, F and Weisz, A (2001). Estrogens do not modify MAP Kinase-dependent nuclear signaling during stimulation of early G 1 progression in human breast cancer cells. Cancer Res., 61, 6360-6366.

Carroll, JS, Prall, OWJ, Musgrove, EA and Sutherland, RL (2000). A pure estrogen antagonist inhibits cyclin E-cdk2 activity in MCF-7 breast cancer cells and induces accumulation of p130-E2F4 complexes characteristic of quiescence. J. Biol. Chem., 275, 38221-38229.

Caruana, G, Cambareri, A and Ashman, L (1999). Isoforms of c-kit differ in activation of signalling pathways and transformation of NIH3T3 fibroblasts. Oncogene, 18, 5573-5581.

Cary, LA, Chang, JF and Guan, JL (1996). Stimulation of cell migration by overexpression of focal adhesion kinase and its association with Src and Fyn. J. Cell Sci., 109, 1787-1794. Castro-Rivera, E, Samudio, I and Safe, S (2001). Estrogen regulation of cyclin Dl gene expression in ZR-75 breast cancer cells involves multiple enhancer elements. J. Biol. Chem., 276, 30853-30861.

153 Chappell, J, Leitner, JW, Solomon, S, Golovchenko, I, Goalstone, ML and Draznin, B (2001 ). Effect of insulin on cell cycle progression in MCF-7 breast cancer cells. J. Biol. Chem., 276, 38023-38028.

Cheatham, Band Kahn, C (1995). Insulin action and the insulin signaling network. Endocr. Rev., 16, 117-142.

Chen, HC and Guan, JL (1994). Stimulation of phosphatidylinositol 3'-kinase association with focal adhesion kinase by platelet-derived growth factor. J. Biol. Chem., 269, 31229- 31233.

Cheng, A, Saxton, T, Sakai, R, Kulkarni, S, Mbamalu, G, Vogel, W, Tortorice, CG, Cardiff, RD, Cross, JC, Muller, WJ and Pawson, T (1998). Mammalian Grb2 regulates multiple steps in embryonic development and malignant transformation. Cell, 95, 793-803.

Cherniack, A, Klarlund, J, Conway, B and Czech, M (1995). Disassembly of Son-of­ sevenless proteins from Grb2 during p21ras desensitization by insulin. J. Biol. Chem., 270, 1485-1488.

Chesi, M, Nardini, E, Brents, LA, Schrock, E, Ried, T, Kuehl, WM and Bergsagel, PL (1997). Frequent translocation t(4;14)(p16.3;q32.3) in multiple myeloma is associated with increased expression and activating mutations of fibroblast growth factor receptor 3. Nat. Genet., 16, 260-264.

Chi, NW and Lodish, HF (2000). Tankyrase is a golgi-associated mitogen-activated protein kinase substrate that interacts with IRAP in GLUT4 vesicles. J. Biol. Chem., 275, 38437- 38444.

Clark, SF, Molero, J-C and James, DE (2000). Release of insulin receptor substrate proteins from an intracellular complex coincides with the development of insulin resistance. J. Biol. Chem., 275, 3819-3826.

154 Clarke, R (2000). Introduction and overview: sex steroids in the mammary gland. J. Mammary Gland Biol. Neoplasia, 5, 245-250.

Clontech (1999). Retroviral cDNA library user manual, PT3230-1, PR9371 l.

Cobleigh, M, Vogel, CL, Tripathy, D, Robert, NJ, Scholl, S, Fehrenbacher, L, Wolter, JM, Paton, V, Shak, S, Lieberman, G and Slamon, DJ (1999). Multinational study of the efficacy and safety of humanized anti-HER2 monoclonal antibody in women who have HER2-overexpressing metastatic breast cancer that has progressed after chemotherapy for metastatic disease. J. Clin. Oncol., 17, 2639-2648.

Coffer, PJ and Kaliman, P (1995). Phosphatidylinositol 3-kinase inhibitors block differentiation of skeletal muscle cells. Nature, 376, 599-602.

Couse, JF, Lindzey, J, Grandien, K, Gustafsson, JA and Korach, KS (1997). Tissue distribution and quantitative analysis of estrogen receptor-a (ERa) and estrogen receptor-~ (ER~) messenger ribonucleic acid in the wild-type· and ERa-knockout mouse. Endocrinology, 138, 4613-21.

Coussens, L, Yang-Feng, TL, Liao, YC, Chen, E, Gray, A, McGrath, J, Seeburg, PH, Libermann, TA, Schlessinger, J, Francke, U and Ullrich, A (1985). Tyrosine kinase receptor with extensive homology to EGF receptor shares chromosomal location with neu oncogene. Science, 230, 1132-1139.

Crouch, M, Davy, D, Willard, F and Berven, L (2000). Insulin induces epidermal growth factor (EGF) receptor clustering and potentiates EGF-stimulated DNA synthesis in swiss 3T3 cells: a mechanism for costimulation in mitogenic synergy. Immunol. Cell. Biol., 78, 408-414.

155 Cullen, KJ, Lippman, ME, Chow, D, Hill, S, Rosen, N and Zwiebel, JA (1992). Insulin-like growth factor-II overexpression in MCF-7 cells induces phenotypic changes associated with malignant progression. Mol. Endocrinol., 6, 91-100.

Cullen, KJ, Yee, D, Sly, WS, Perdue, J, Hampton, B, Lippman, ME and Rosen, N (1990). Insulin-like growth factor receptor expression and function in human breast cancer. Cancer Res., 50, 48-53.

Daly, RJ (1998). The Grb7 family of signalling proteins. Cell. Signal., 10, 613-618.

Daly, RJ (1999). Take your partners, please-signal diversification by the erbB family of receptor tyrosine kinases. Growth Factors, 16, 255-263.

Daly, RJ (2001). In: Methods Mol. Biol. Reith, AD (eds). pp. 239-249. Humana Press Inc., New Jersey.

Daly, RJ, Binder, MD and Sutherland, RL (1994). Overexpression of the Grb2 gene in human breast cancer cell lines. Oncogene, 9, 2723-2727.

Daly, RJ, Sanderson, GM, Janes, PJ and Sutherland, RL (1996). Cloning and characterization of Grb14, a novel member of the Grb7 gene family. J. Biol. Chem., 271, 12502-12510.

Darnell, JEJ, Kerr, IM and Stark, GR (1994). Jak-STAT pathways and transcriptional activation in response to IFNs and other extracellular signaling proteins. Science, 264, 1415-1421.

Davis, RJ (1993). The mitogen-activated protein kinase signal transduction pathway. J. Biol. Chem., 268, 14553-14556.

156 Di Fiore, PP, Pierce, JH, Fleming, TP, Hazan, R, Ullrich, A, King, CR, Schlessinger, J and Aaronson, SA (1987a). Overexpression of the human EGF receptor confers an EGF­ dependent transformed phenotype to NIH-3T3 cells. Cell, 51, 1063-1070.

Di Fiore, PP, Pierce, JH, Kraus, MH, Segatto, 0, King, CR and Aaronson, SA ( 1987b ). erbB-2 is a potent oncogene when overexpressed in NIH/3T3 cells. Science, 237, 178-182.

Diamond, MI, Miner, JN, Yoshinaga, SK and Yamamoto, KR (1990). Transcription factor interactions: selectors of positive or negative regulation from a single DNA element. Science, 249, 1266-1272.

Dickson, C, Spencer-Dene, B, Dillon, C and Fanti, V (2000). Fibroblast growth factors and their receptors. Breast Cancer Res., 2, 191-196.

DiGiovanna, M, Lerman, M, Coffey, R, Muller, W, Cardiff, Rand Stem, D (1998). Active signaling by Neu in transgenic mice. Oncogene, 17, 1877-1884.

Dong, LQ, Du, H, Porter, SG, Kolakowski, LF, Lee, A, Mandarino, J, Fan, J, Yee, D and Liu, F (1997). Cloning, chromosome localization, expression and characterization of a Src homology 2 and pleckstrin homology domain-containing insulin receptor binding protein hGrblOy. J. Biol. Chem., 272, 29104-29112.

Dong, LQ, Porter, S, Hu, D and Liu, F (1998). Inhibition of hGrblO binding to the insulin receptor by functional domain-mediated oligomerisation. J. Biol. Chem., 273, 17720- 17725.

Egan, C, Pang, A, Durda, D, Cheng, HC, Wang, JH and Fujita, DJ (1999). Activation of Src in human breast tumor cell lines: elevated levels of phosphotyrosine phosphatase activity that preferentially recognizes the Src carboxy terminal negative regulatory tyrosine 530. Oncogene, 18, 1227-1237.

157 Elenius, K, Choi, CJ, Paul, S, Santiestevan, E, Nishi, E and Klagsbrun, M (1999). Characterization of a naturally ocurring ErbB4 isoform that does not bind or activate phosphatidyl inositol 3-kinase. Oncogene, 18, 2607-2615.

Fath, I, Schweighoffer, F, Rey, I, Multon, M, Boiziau, J, Duchesne, M and Tocque, B (1994). Cloning of a Grb2 isoform with apoptotic properties. Science, 264, 971-974.

Fiddes, RJ, Campbell, DH, Janes, PW, Sivertsen, SP, Sasaki, H, Wallasch, C and Daly, RJ (1998). Analysis of Grb7 recruitment by heregulin-activated erbB receptors reveals a novel target selectivity for erbB3. J. Biol. Chem., 273, 7717-7724.

Finer, MH, Dull, TJ, Qin, L, Farson, D and Roberts, MR (1994). kat: a high-efficiency retroviral transduction system for primary human T lymphocytes. Blood, 83, 43-50.

Frantz, J, Giorgetti-Peraldi, S, Ottinger, E and Shoelson, S (1997). Human GRB­ IRf3/GRB 10. Splice variants of an insulin and growth factor receptor-binding protein with PH and SH2 domains. J. Biol. Chem., 272, 2659-2667.

Frasca, F, Pandini, G, Scalia, P, Sciacca, L, Mineo, R, Costantino, A, Goldfine, I, Belfiore, A and Vigneri, R (1999). Insulin receptor isoform A, a newly recognized, high-affinity insulin-like growth factor II receptor in fetal and cancer cells. Mo!. Cell. Biol., 19, 3278- 3288.

Frisch, SM, Vuori, K, Ruoslahti, E and Chan-Hui, PY (1996). Control of adhesion­ dependent cell survival by focal adhesion kinase. J. Cell Biol., 134, 793-799.

Frittitta, L, Vigneri, R, Stampfer, MR and Goldfine, ID (1995). Insulin receptor overexpression in 184B5 human mammary epithelial cells induces a ligand-dependent transformed phenotype. J. Cell. Biochem., 57, 666-669.

158 Fruman, DA, Rameh, LE and Cantley, LC (1999). Phosphoinositide binding domains: embracing 3-phosphate. Cell, 97, 817-820.

Funaki, M, Katagiri, H, Inukai, K, Kikuchi, M and Asano, T (2000). Structure and function of phosphatidylinositol-3,4 kinase. Cell Signal., 12, 135-142.

Gale, NW, Kaplan, S, Lowenstein, EJ, Schlessinger, J and Bar-Sagi, D (1993). Grb2 mediates the EGF-dependent activation of guanine nucleotide exchange on ras. Nature, 363, 88-92.

Gamper, C, Omene, CO, van Eyndhoven, WG, Glassman, GD and Lederman, S (2001). Expression and function of TRAF-3 splice-variant isoforms in human lymphoma cell lines. Hum. Immunol., 62, 1167-1177.

Ganong, WF (1995). In: Review of medical physiology Dolan, J and Langan, C (eds). pp. 404-406. Appleton and Lange, Connecticut.

Gilmore, AP and Romer, LH (1996). Inhibition of focal adhesion kinase (FAK) signaling in focal adhesions decreases cell motility and proliferation. Mol. Biol. Cell, 7, 1209-1224.

Giorgino, F, Belfiore, A, Milazzo, G, Costantino, A, Maddux, B, Whittaker, J, Goldfine, I and Vigneri, R (1991). Overexpression of insulin receptors in fibroblast and ovary cells induces a ligand-mediated transformed phenotype. Mol. Endocrinol., 5, 452-459.

Gishizky, ML and Witte, ON ( 1992). Initiation of deregulated growth of multipotent progenitor cells by bcr-abl in vitro. Science, 256, 836-839. Gross, JM and Douglas, Y (2002). How does the estrogen receptor work? Breast Cancer Res., 4, 62-64.

Grunt, TW, Saceda, M, Martin, MB, Lupu, R, Dittrich, E, Krupitza, G, Harant, H, Huber, H and Dittrich, C ( 1995). Bidirectional interactions between the estrogen receptor and the

159 c-erbB-2 signaling pathways: Heregulin inhibits estrogenic effects in breast cancer cells. Int. J. Cancer, 63, 560-567.

Guan, JL (1997). Focal adhesion kinase in integrin signaling. Matrix Biol., 16, 195-200.

Gullick, WJ (1990). The role of the epidermal growth factor receptor and the c-erbB-2 protein in breast cancer. Int. J. Cancer, Suppl 5, 55-61.

Gustafsson, JA (2000). Novel aspects of estrogen action. J. Soc. Gynecol. Investig., 7, S8- S9.

Guy, CT, Webster, MA, Schaller, M, Parsons, TJ, Cardiff, RD and Muller, WJ (1992). Expression of the neu protooncogene in the mammary epithelium of transgenic mice induces metastatic disease. Proc. Natl. Acad. Sci. USA, 89, 10578-10582.

Hall, JM, Couse, JF and Korach, KS (2001). The multifaceted mechanisms of estradiol and estrogen receptor signaling. J. Biol. Chem., 276, 36869-36872.

Han, DC and Guan, J-L (1999). Association of focal adhesion kinase with Grb7 and its role in cell migration. J. Biol. Chem., 274, 24425-24430.

Han, DC, Shen, T-L and Guan, J-L (2000). Role of Grb7 targeting to focal contacts and its phosphorylation by focal adhesion kinase in regulation of cell migration. J. Biol. Chem., 275, 28911-28917.

Han, DC, Shen, T-L and Guan, J-L (2001). The Grb7 family proteins: structure, interactions with other signaling molecules and potential cellular functions. Oncogene, 20, 6315-6321.

He, W, Rose, D, Olefsky, J and Gustafson, T (1998). GrblO interacts differentially with the insulin receptor, insulin-like growth factor I receptor, and epidermal growth factor receptor

160 via the GrblO src homology 2 (SH2) domain and a second novel domain located between the pleckstrin homology and SH2 domains. J. Biol. Chem., 273, 6860-6867.

Heldin, CH, Backstrom, G, Ostman, A, Hammacher, A, Ronnstrand, L, Rubin, K, Nister, M and Westermark, B (1988). Binding of different dimeric forms of PDGF to human fibroblasts: evidence for two separate receptor types. EMBO J., 7, 1389-1393.

Heldin, CH, Ernlund, A, Rorsman, C and Ronnstrand, L ( 1989). Dimerization of B-type platelet-derived growth factor receptors occurs after ligand binding and is closely associated with receptor kinase activation. J. Biol. Chem., 264, 8905-8912.

Hemming, R, Agatep, R, Badiani, K, Wyant, K, Arthur, G, Gietz, RD and Triggs-Raine, B (2001). Human growth factor bound 14 binds the activated insulin receptor and alters the insulin-stimulated tyrosine phosphorylation levels of multiple proteins. Biochem. Cell Biol., 79, 21-32.

Hennige, AM, Lammers, R, Hoppner, W, Arlt, D, Strack, V, Teichmann, R, Machicao, F, Ullrich, A, Haring, HU and Kellerer, M (2001). Inhibition of Ret oncogene activity by the protein tyrosine phosphatase SHPI. Endocrinology, 142, 4441-4447.

Hennipman, A, van Oirschot, BA, Smits, J, Rijksen, G and Staal, GEJ (1989). Tyrosine kinase activity in breast cancer, benign breast disease, and normal breast tissue. Cancer Res., 49, 516-521.

Hill, MM, Clark, SF, Tucker, DF, Birnbaum, MJ, James, DE and Macaulay, SL (1999). A role for protein kinase B ~/ Akt2 in insulin-stimulated GLUT4 translocation in adipocytes. Mol. Cell. Biol., 19, 7771-7781.

Hillier, BJ, Christopherson, KS, Prehoda, KE, Bredt, DS and Lim, WA (1999). Unexpected modes of PDZ domain scaffolding revealed by structure of nNOS-syntrophin complex. Science, 284, 812-815.

161 Hofstra, RM, Landsvater, RM, Ceccherini, I, Stulp, RP, Stelwagen, T, Luo, Y, Pasini, B, Hoppener, JW, van Amstel, H and Romeo, G (1994). A mutation in the RET proto­ oncogene associated with multiple endocrine neoplasia type 2B and sporadic medullary thyroid carcinoma. Nature, 367, 375-376.

Holt, KH, Kasson, BG and Pessin, JE (1996). Insulin stimulation of a MEK-dependent but ERK-independent SOS protein kinase. Mol. Cell. Biol., 16, 577-583.

Howell, A, Osborne, CK, Morris, C and Wakeling, AE (2000). ICI 182,780 (Faslodex™) Development of a Novel, "pure" antiestrogen. Cancer, 89, 817-825.

Hubbard, SR (1997). Crystal structure of the activated insulin receptor tyrosine kinase in complex with peptide substrate and ATP analog. EMBO J., 16, 5573-5581.

Hubbard, SR and Till, JH (2000). Protein tyrosine kinase structure and function. Annu. Rev. Biochem., 69, 373-398.

Hunter, T (1998). The Croonian Lecture 1997. The phosphorylation of proteins on tyrosine: its role in cell growth and disease. Philos. Trans. R. Soc. Land. B Biol. Sci., 353, 583-605.

Hunter, T (2000). Signaling-2000 and beyond. Cell, 100, 113-127.

Huynh, H, Nickerson, T, Pollak, Mand Yang, X (1996a). Regulation of insulin-like growth factor I receptor expression by the pure antiestrogen ICI 182780. Clin. Cancer Res., 2, 2037-2042.

Huynh, H, Yang, X and Pollak, M (1996b). Estradiol and antiestrogens regulate a growth inhibitory insulin-like growth factor binding protein 3 autocrine loop in human breast cancer cells. J. Biol. Chem., 271, 1016-1021.

162 Huynh, H, Yang, X-F and Pollak, M (1996c). A role for insulin-like growth factor binding protein 5 in the antiproliferative action of the antiestrogen ICI 182780. Cell Growth Differ., 7, 1501-1506.

Ignar-Trowbridge, DM, Teng, CT, Ross, KA, Parker, MG, Korach, KS and McLachlan, JA (1993). Peptide growth factors elicit estrogen receptor-dependent transcriptional activation of an estrogen-responsive element. Mol. Endocrinol., 7, 992-998.

Ihle, JN (1995). Cytokine receptor signaling. Nature, 377, 591-594.

Ilic, D, Furuta, Y, Kanazawa, S, Takeda, N, Sobue, K, Nakatsuji, N, Nomura, S, Fujimoto, J, Okada, Mand Yamamoto, T (1995). Reduced cell motility and enhanced focal adhesion contact formation in cells from PAK-deficient mice. Nature, 377, 539-544.

Inoue, G, Cheatham, B, Emkey, Rand Kahn, CR (1998). Dynamics of insulin signaling in 3T3-Ll adipocytes. J. Biol. Chem., 273, 11548-11555.

Ish-Shalom, D, Christoffersen, CT, Vorwerk, P, Sacerdoti-Sierra, N, Shymko, RM, Naor, D and De Meyts, P (1997). Mitogenic properties of insulin and insulin analogues mediated by the insulin receptor. Diabetologia, 40 Suppl 2, S25-31.

Jacobs, C and Rubsamen, H (1983). Expression of pp6oc-src protein kinase in adult and fetal human tissue: high activities in some sarcomas and mammary carcinomas. Cancer Res., 43, 1696-1702.

Jamison, DT, Creese, A, Prentice, T, Gakidou, E, Inoue, M, Beusenberg, M, Engers, H, Goodman, C, Guidon, E, Jha, P, Mendis, K, Nabarro, D, Tulloch, J, Wang, J and Yach, D (1999). In: The World Health Report Campanini, Band Haden, A (eds). pp. 85-116. World Health Organization, Geneva.

163 Janes, PW, Lackmann, M, Church, WB, Sanderson, GM, Sutherland, RL and Daly, RJ (1997). Structural determinants of the interaction between the erbB2 receptor and the src homology 2 domain of Grb7. J. Biol. Chem., 272, 8490-8497.

Jang, JH, Shin, KH and Park, JG (2001). Mutations in fibroblast growth factor receptor 2 and fibroblast growth factor receptor 3 genes associated with human gastric and colorectal cancers. Cancer Res., 61, 3541-3543.

Johnson, MR, Valentine, C, Basilico, C and Mansukhani, A (1998). FGF signaling activates STATl and p21 and inhibits the estrogen response and proliferation of MCF-7 cells. Oncogene, 16, 2647-2656.

Johnston, CL, Cox, HC, Gomm, JJ and Coombes, RC (1995). bFGF and aFGF induce membrane ruffling in breast cancer cells but not in normal breast epithelial cells: FGFR-4 involvement. Biochem. J., 306, 609-616.

Jonat, C, Rahmsdorf, HJ, Park, KK, Cato, AC, Gebel, S, Ponta, Hand Herrlich, P (1990). Antitumor promotion and antiinflammation: down-modulation of AP-1 (Fos/Jun) activity by glucocorticoid hormone. Cell, 62, 1189-1204.

Jones, N, Master, Z, Jones, J, Bouchard, D, Gunji, Y, Sasaki, H, Daly, R, Alitalo, Kand Dumont, DJ (1999). Identification of Tek/Tie2 binding patners. Binding to a multifunctional docking site mediates cell survival and migration. J. Biol. Chem., 274, 30896-30905.

Kaaks, R (1996). Nutrition, hormones, and breast cancer: is insulin the missing link? Cancer Causes Control, 7, 605-625.

Kahlert, S, Nuedling, S, Eickels, MV, Vetter, H, Meyer, Rand Grohe, C (2000). Estrogen receptor a rapidly activates the IGF-1 receptor pathway. J. Biol. Chem., 275, 18447-18453.

164 Kainulainen, V, Sundvall, M, Maatta, JA, Santiestevan, E, Klagsbrun, M and Elenius, K (2000). A natural erbB4 isoform that does not activate phosphoinositide 3-kinase mediates proliferation but not survival or chemotaxis. J. Biol. Chem., 275, 8641-8649.

Kaleko, M, Rutter, WJ and Miller, AD ( 1990). Overexpression of the human insulinlike growth factor I receptor promotes ligand-dependent neoplastic transformation. Mol. Cell. Biol., 10, 464-473.

Kasus-Jacobi, A, Bereziat, V, Perdereau, D, Girard, J and Burno!, A (2000). Evidence for an interaction between the insulin receptor and Grb7. A role for two of its binding domains, PIR and SH2. Oncogene, 19, 2052-2059.

Kasus-Jacobi, A, Perdereau, D, Auzan, C, Clauser, E, Van Obberghen, E, Mauvais-Jarvis, F, Girard, J and Burno!, A (1998). Identification of the rat adapter Grbl4 as an inhibitor of insulin actions. J. Biol. Chem., 273, 26026-26035.

Kato, S, Endoh, H, Masuhiro, Y, Kitamoto, T, Uchiyama, S, Sasaki, H, Masushige, S, Gotoh, Y, Nishida, E, Kawashima, H, Metzger, D and Chambon, P (1995). Activation of the estrogen receptor through phosphorylation by mitogen-activated protein kinase. Science, 270, 1491-1494.

Kato, S, Tora, L, Yamauchi, J, Masushige, S, Bellard, M and Chambon, P (1992). A far upstream estrogen response element of the ovalbumin gene contains several half­ palindromic 5'-TGACC-3' motifs acting synergistically. Cell, 68, 731-742.

Katso, R, Okkenhaug, K, Ahmadi, K, White, S, Timms, J and Waterfield, MD (2001). Cellular function of phosphoinositide 3-kinases: implications for development, immunity, homeostasis, and cancer. Annu. Rev. Cell Dev. Biol., 17, 615-675.

165 Katzenellenbogen, BS and Norman, MJ (1990). Multihormonal regulation of the progesterone receptor in MCF-7 human breast cancer cells: interrelationships among insulin/insulin-like growth factor-I, serum, and estrogen. Endocrinology, 126, 891-898.

Keegan, Kand Cooper, JA (1996). Use of the two hybrid system to detect the association of the protein-tyrosine-phosphatase, SHPTP2, with another SH2-containing protein, Grb7. Oncogene, 12, 1537-1544.

Kim, HK, Kim, JW, Zilberstein, A, Margolis, B, Kim, JG, Schlessinger, J and Rhee, SG (1991). PDGF stimulation of inositol phospholipid hydrolysis requires PLC-gamma 1 phosphorylation on tyrosine residues 783 and 1254. Cell, 65, 435-441.

Klein, DE, Lee, A, Frank, DW, Marks, MS and Lemmon, MA (1998). The pleckstrin homology domains of dynamin isoforms require oligomerization for high affinity phosphoinositide binding. J. Biol. Chem., 273, 27725-27733.

Klein-Hitpass, L, Ryffel, GU, Heitlinger, E and Cato, ACB (1988). A 13 bp palindrome is a functional estrogen responsive element and interacts specifically with estrogen receptor. Nucleic Acids Res., 16, 647-663.

Klint, P, Kanda, S and Claesson-Welsh, L (1995). She and a novel 89-kDa component couple to the Grb2-Sos complex in fibroblast growth factor-2-stimulated cells. J. Biol. Chem., 270, 23337-23344.

Klippel, A, Escobedo, JA, Fanti, WJ and Williams, LT (1992). The C-terminal SH2 domain of p85 accounts for the high affinity and specificity of the binding of phosphatidylinositol 3-kinase to phosphorylated platelet-derived growth factor r3 receptor. Mol. Cell. Biol., 12,

166 Kokai, Y, Myers, J, Wada, T, Brown, V, LeVea, C, Davis, J, Dobashi, Kand Greene, M (1989). Synergistic interaction of p185c-neu and the EGF receptor leads to transformation of rodent fibroblasts. Cell, 58, 287-292.

Kotani, K, Yonezawa, K, Hara, K, Ueda, H, Kitamura, Y, Sakaue, H, Ando, A, Chavanieu, A, Calas, Band Grigorescu, F (1994). Involvement of phosphoinositide 3-kinase in insulin­ or IGF-1-induced membrane ruffling. EMBO J., 13, 2313-2321.

Kouhara, H, Hadari, YR, Spivak-Kroizman, T, Schilling, J, Bar-Sagi, D, Lax, I and Schlessinger, J (1997). A lipid-anchored Grb2-binding protein that links FGF-receptor activation to the Ras/MAPK signaling pathway. Cell, 89, 693-702.

Kovacina, KS and Roth, RA (1993). Identification of SHC as a substrate of the insulin receptor kinase distinct from the GAP-associated 62 kDa tyrosine phosphoprotein. Biochem. Biophys. Res. Commun., 192, 1303-11.

Krebs, DL and Hilton, DJ (2001). SOCS proteins: Negative regulators of cytokine signalling. Stem Cells, 19, 378-387.

Kuiper, GG, Enmark, E, Pelto-Huikko, M, Nilsson, S and Gustafsson, JA (1996). Cloning of a novel receptor expressed in rat prostate and ovary. Proc. Natl. Acad. Sci. USA, 93, 5925-5930.

Kuriyan, J and Cowburn, D (1993). Structures of SH2 and SH3 domains. Curr. Opin. Struct. Biol., 3, 828-837.

Kurokawa, H, Lenferink, AEG, Simpson, JF, Pisacane, PI, Sliwkowski, MX, Forbes, JT and Arteaga, CL (2000). Inhibition of HER2/neu (erbB-2) and mitogen-activated protein kinases enhances tamoxifen action against HER2-overexpressing, tamoxifen-resistant breast cancer cells. Cancer Res., 60, 5887-5894.

167 Kutateladze, TG, Ogburn, KD, Watson, WT, De Beer, T, Emr, SD, Burd, CG and Overduin, M (1999). Phosphatidylinositol 3-phosphate recognition by the FYVE domain. Mol. Cell, 3, 805-811.

La Rosa, S, Sessa, F, Colombo, L, Tibiletti, MG, Furlan, D and Capella, C (2001). Expression of acidic fibroblast growth factor (aFGF) and fibroblast growth factor receptor 4 (FGFR4) in breast fibroadenomas. J. Clin. Pathol., 54, 37-41.

La Vecchia, C, Negri, E, Franceschi, S, D'Avanzo, Band Boyle, P (1994). A case-control study of diabetes mellitus and cancer risk. Br. J. Cancer, 70, 950-953.

Lai, A, Sarcevic, B, Prall, OWJ and Sutherland, RL (2001). Insulin/insulin-like growth factor-I and estrogen cooperate to stimulate cyclin E-cdk2 activation and cell cycle progression in MCF-7 breast cancer cells through differential regulation of cyclin E and p21 WAFl/cipl_ J. Biol. Chem., 276, 25823-25833.

Lammers, R, Gray, A, Schlessinger, J and Ullrich, A ( 1989). Differential signalling potential of insulin- and IGF-1-receptor cytoplasmic domains. EMBO J., 8, 1369-75.

Langlais, P, Dong, LQ, Hu, D and Liu, F (2000). Identification of GrblO as a direct substrate of the Src tyrosine kinase family. Oncogene, 19, 2895-2903.

Langlois, WJ and Olefsky, JM ( 1994 ). In: Molecular biology of diabetes, part II Draznin, B and LeRoith, D (eds). pp. 25-50. Humana Press Inc., New Jersey.

Laviola, L, Giorgino, F, Chow, JC, Baquero, JA, Hansen, H, Ooi, J, Zhu, J, Riedel, Hand Smith, RJ (1997). The adapter protein GrblO associates preferentially with the insulin receptor as compared with the IGF-1 receptor in mouse fibroblasts. J. Clin. Invest., 99, 830-837.

168 Lee, AV, Jackson, JG, Gooch, JL, Hilsenbeck, SG, Coronado-Heinsohn, E, Osborne, CK and Yee, D (1999). Enhancement of insulin-like growth factor signaling in human breast cancer: Estrogen regulation of insulin receptor substrate-I expression in vitro and in vivo. Mol. Endocrinol., 13, 787-796.

Lee, AV, Weng, CN, Jackson, JG and Yee, D (1997). Activation of estrogen receptor­ mediated gene transcription by IGF-1 in human breast cancer cells. J. Endocrinol., 152, 39- 47.

Lee, H, Volonte, D, Galbiati, F, Iyengar, P, Lublin, DM, Bregman, DB, Wilson, MT, Campos-Gonzalez, R, Bouzahzah, B, Pestell, RG, Scherer, PE and Lisanti, MP (2000). Constitutive and growth factor-regulated phosphorylation of caveolin-1 occurs at the same site (Tyr-14) in vivo: identification of a c-Src/Cav-1/Grb7 signaling cassette. Mol. Endocrinol., 14, 1750-1775.

Levenson, AS and Jordan, VC (1999). Selective oestrogen receptor modulation: molecular pharmacology for the millennium. Eur. J. Cancer, 35, 1628-1639.

Lichtner, RB, Kaufmann, A, Kittmann, A, Rohde-Schulz, B, Walter, J, Williams , L, Ullrich, A, Schirrmacher, V and Khazaie, K (1995). Ligand mediated activation of ectopic EGF receptor promotes matrix protein adhesion and lung colonization of rat mammary adenocarcinoma cells. Oncogene, 10, 1823-1832.

Liu, ET (2000). Breast cancer research: where we are and where we should go. Breast Cancer Res., 2, 73-76. Liu, F and Roth, RA (1995). Grb-IR: a SH2-domain-containing protein that binds to the insulin receptor and inhibits its function. Proc. Natl. Acad. Sci. USA, 92, 10287-10291.

Lobenhofer, EK and Marks, JR (2000). Estrogen-induced mitogenesis of MCF-7 cells does not require the induction of mitogen-activated protein kinase activity. J. Steroid Biochem. Mol. Biol., 75, 11-20.

169 Lowenstein, EJ, Daly, RJ, Batzer, AG, Li, W, Margolis, B, Lammers, R, Ullrich, A, Skolnik, EY, Bar-Sagi, D and Schlessinger, J (1992). The SH2 and SH3 domain-containing protein GRB2 links receptor tyrosine kinases to ras signalling. Cell, 70, 431-442.

Lu, P-J, Zhou, XZ, Shen, M and Lu, KP ( 1999). Function of WW domains as phosphoserine- or phosphothreonine-binding modules. Science, 26, 1325-1328.

Luo, RZ-T, Beniac, DR, Fernandes, A, Yip, CC and Ottensmeyer, FP (1999). Quaternary structure of the insulin-insulin receptor complex. Science, 285, 1077-1080.

Luttrell, DK, Lee, A, Lansing, TJ, Crosby, RM, Jung, KD, Willard, D, Luther, M, Rodriguez, M, Berman, J and Gilmer, TM (1994). Involvement of pp6oc-src with two major signalling pathways in human breast cancer. Proc. Natl. Acad. Sci. USA, 91, 83-87.

Lynch, DK and Daly, RJ (2002). PKB-mediated negative feedback tightly regulates mitogenic signalling via Gab2. EMBO J., 21, 72-82.

Lyons, RJ, Deane, R, Lynch, DK, Ye, ZS, Sanderson, GM, Eyre, HJ, Sutherland, GR and Daly, RJ (2001). Identification of a novel human tankyrase through its interaction with the adaptor protein Grbl4. J. Biol. Chem., 276, 17172-17180.

Ma, ZQ, Santagati, S, Patrone, C, Pollio, G, Vegeto, E and Maggi, A (1994). Insulin-like growth factors activate estrogen receptor to control the growth and differentiation of the human neuroblastoma cell line SK-ER3. Endocrinology, 8, 910-918.

MacArthur, CA, Shankar, DB and Shackleford, GM (1995). FGF-8, activated by proviral insertion, cooperates with the wnt-1 transgene in murine mammary tumorigenesis. J. Viral., 69, 2501-2507.

170 Malaney, S and Daly, RJ (2001). The ras signaling pathway in mammary tumorigenesis and metastasis. J. Mammary Gland Biol. Neoplasia, 6, 101-113.

Mangelsdorf, DJ, Thummel, C, Beato, M, Herrlich, P, Schutz, G, Umesono, K, Blumberg, B, Kastner, P, Mark, M and Chambon, P (1995). The nuclear receptor superfamily: the second decade. Cell, 83, 835-839.

Manley, JL and Tacke, R (1996). SR proteins and splicing control. Genes Dev., 10, 1569- 1579.

Mano, h, Ohya, K, Miyazato, A, Yamashita, Y, Ogawa, W, lnazawa, J, Ikeda, U, Shimada, K, Hatake, K, Kasuga, M, Ozawa, Kand Kajigaya, S (1998). GrblO/GrbIR is an in vivo substrate of Tee tyrosine kinase. Genes Cells, 3, 437-441.

Mao, Y, Nickitenko, A, Duan, X, Lloyd, TE, Wu, MN, Bellen, Hand Quiocho, FA (2000). Crystal structure of the VHS and FYVE tandem domains of Hrs, a protein involved in membrane trafficking and signal transduction. Cell, 100, 447-456.

Margolis, B, Rhee, SG, Felder, S, Mervic, M, Lyall, R, Levitzki, A, Ullrich, A, Zilberstein, A and Schlessinger, J (1989). EGF induces tyrosine phosphorylation of phospholipase C-11: a potential mechanism for EGF receptor signaling. Cell, 57, 1101-1107.

Margolis, B, Silvennoinen, 0, Comoglio, F, Roonprapunt, C, Skolnik, E, Ullrich, A and Schlessinger, J ( 1992). High-efficiency expression/cloning of epidermal growth factor receptor-binding proteins with src homology 2 domains. Proc. Natl. Acad. Sci. USA, 89, 8894-8898.

Marshall, CJ (1996). Ras effectors. Curr. Opin. Cell Biol., 8, 197-204.

Martin, MB, Franke, TF, Stoica, GE, Chambon, P, Katzenellenbogen, BS, Stoica, BA, McLemore, MS, Olivo, SE and Stoica, A (2000). A role for Akt in mediating the estrogenic

171 functions of epidermal growth factor and insulin-like growth factor I. Endocrinology, 141, 4503-4511.

Martin, SS, Haruta, T, Morris, AJ, Klippel, A, Williams, LT and Olefsky, JM (1996). Activated phosphatidylinositol 3-kinase is sufficient to mediate actin rearrangement and GLUT4 translocation in 3T3-Ll adipocytes. J. Biol. Chem., 271, 17605-17608.

Mason, IJ (1994). The ins and outs of fibroblast growth factors. Cell, 78, 547-552.

Mathieu, M, Vignon, F, Capony, F and Rochefort, H (1991). Estradiol down-regulates the mannose-6-phosphate /insulin-like growth factor-II receptor gene and induces cathepsin-D in breast cancer cells: A receptor saturation mechanism to increase the secretion of lysosomal proenzymes. Mol. Endocrinol., 5, 815-822.

Mathieu, MC, Clark, GM, Allred, DC, Goldfine, ID and Vigneri, R ( 1997). Insulin receptor expression and clinical outcome in node-negative breast cancer. Proc. Assoc. Am. Physicians, 109, 565-571.

Matsumoto-Yoshitomi, S, Habashita, J, Nomura, C, Kuroshima, K and Kurokawa, T (1997). Autocrine transformation by fibroblast growth factor 9 (FGF-9) and its possible participation in human oncogenesis. Int. J. Cancer, 71, 442-450.

Mayer, B (1997). Clamping down on Src activity. Curr. Biol., 1, R295-R298.

McDonnell, DP (2000). Selective estrogen receptor modulators (SERMs): a first step in the development of perfect hormone replacement therapy regimen. J. Soc. Gynecol. Investig., 7, SIO-S15.

McGlade, CJ, Ellis, C, Reedijk, M, Anderson, D, Mbamalu, G, Reith, AD, Panayotou, G, End, P, Bernstein, A and Kazlauskas, A (1992). SH2 domains of the p85 alpha subunit of

172 phosphatidylinositol 3-kinase regulate binding to growth factor receptors. Mol. Cell. Biol., 12, 991-997.

McLeskey, S, Ding, I, Lippman, Mand Kern, F (1994). MDA-MB-134 breast carcinoma cells overexpress fibroblast growth factor (FGF) receptors and are growth-inhibited by FGF ligands. Cancer Res., 54, 523-530.

McLeskey, SW, Kurebayashi, J, Honig, SF, Zwiebel, J, Lippman, ME, Dickson, RB and Kern, FG ( 1993). Fibroblast growth factor 4 transfection of MCF-7 cells produces cell lines that are tumorigenic and metastatic in ovariectomized or tamoxifen-treated athymic nude mice. Cancer Res., 53, 2168-2177.

Merlo, GR, Graus-Porta, D, Cella, N, Marte, BM, Taverna, D and Hynes, NE (1996). Growth, differentiation and survival of HCl 1 mammary epithelial cells: diverse effects of receptor tyrosine kinase-activating peptide growth factors. Eur. J. Cell Biol., 70, 97-105.

Meyers, GA, Orlow, SJ, Munro, IR, Przylepa, KA and Jabs, EW (1995). Fibroblast growth factor receptor 3 (FGFR3) transmembrane mutation in Crouzon syndrome with acanthosis nigricans. Nat. Genet., 11, 462-464.

Migliaccio, A, Di Domenico, M, Castoria, G, de Falco, A, Bontempo, P, Nola, E and Auricchio, F (1996). Tyrosine kinase/p21ras/MAP-kinase pathway activation by estradiol­ receptor complex in MCF-7 cells. EMBO J., 15, 1292-1300.

Migliaccio, E, Mele, S, Salcini, AE, Pelicci, G, Lai, KV, Superti-Furga, G, Pawson, T, Di Fiore, PP, Lanfrancone, Land Pelicci, P (1997). Opposite effects of the p52shc/p46shc and p66shc splicing isoforms on the EGF receptor-MAP kinase-fos signalling pathway. EMBO J., 16, 706-716.

173 Milazzo, G, Giorgino, F, Damante, G, Sung, C, Stampfer, MR, Vigneri, R, Goldfine, ID and Belfiore, A ( 1992). Insulin receptor expression and function in human breast cancer cell lines. Cancer Res., 52, 3924-3930.

Miller, AD (1996). Cell-surface receptors for retroviruses and implications for gene transfer. Proc. Natl. Acad. Sci. USA, 93, 11407-11413.

Mohammadi, M, Dikic, I, Sorokin, A, Burgess, WH, Jaye, M and Schlessinger, J (1996). Identification of six novel autophosphorylation sites on fibroblast growth factor receptor 1 and elucidation of their importance in receptor activation and signal transduction. Mol. Cell. Biol., 16, 977-989.

Mohammadi, M, Honegger, AM, Rotin, D, Fischer, R, Bellot, F, Li, W, Dionne, CA, Jaye, M, Rubinstein, Mand Schlessinger, J (1991). A tyrosine-phosphorylated carboxy-terminal peptide of the fibroblast growth factor receptor (Flg) is a binding site for the SH2 domain of phospholipase C-gamma 1. Mol. Cell Biol., 11, 5068-5078.

Molloy, CA, May, FEB and Westley, BR (2000). Insulin receptor substrate-I expression is regulated by estrogen in the MCF-7 human breast cancer cell line. J. Biol. Chem., 275, 12565-12571.

Mooney, RA, Senn, J, Cameron, S, Inamdar, N, Boivin, LM, Shang, Y and Furlanetto, RW (2001). Suppressors of cytokine signaling-1 and -6 associate with and inhibit the insulin receptor. A potential mechanism for cytokine-mediated insulin resistance. J. Biol. Chem., 276, 25889-25893.

Moran, LA, Scrimgeour, KG, Horton, HR, Ochs, RS and Rawn, JD (1994). In: Biochemistry Foster, S, Pratt, C, Challice, J, O'Quin, T and Ryan, M (eds). pp. 28.24- 28.26. Neil Patterson Publishers, New Jersey.

174 Morrione, A, Plant, P, Valentinis, B, Staub, 0, Kumar, S, Rotin, D and Baserga, R (1999). mGrblO interacts with Nedd4. J. Biol. Chem., 274, 24094-24099.

Morrione, A, Valentinis, B, Li, S, Ooi, J, Margolis, Band Baserga, R (1996). GrblO: a new substrate of the insulin-like growth factor I receptor. Cancer Res., 56, 3165-3167.

Morrione, A, Valentinis, B, Resnicoff, M, Xu, S and Baserga, R (1997). The role of mGrblOa in insulin-like growth factor I-mediated growth. J. Biol. Chem., 272, 26382- 26387.

Morrison, N and Eisman, J (1993). Role of the negative glucocorticoid regulatory element in glucocorticoid repression of the human osteocalcin promoter. J. Bone Miner. Res., 8, 969-975.

Mosselman, S, Polman, J and Dijkema, R (1996). ER~: identification and characterization of a novel human estrogen receptor. FEBS Lett., 392, 49-53.

Mounier, C, Lavoie, L, Dumas, V, Mohammad-Ali, K, Wu, J, Nantel, A, Bergeron, JJM, Thomas, DY and Posner, BI (2001). Specific inhibition by hGRBlO l; of insulin-induced glycogen synthase activation: evidence for a novel signaling pathway. Mol. Cell. Endocrinol., 173, 15-27.

Moutoussamy, S, Renaudie, F, Lago, F, Kelly, P and Finidori, J (1998). GrblO identified as a potential regulator of growth hormone (GH) signaling by cloning of GH receptor target proteins. J. Biol. Chem., 273, 15906-15912.

Moxham, CP, Duronio, V and Jacobs, S (1989). Insulin-like growth factor I receptor~­ subunit heterogeneity. Evidence for hybrid tetramers composed of insulin-like growth factor I and insulin receptor heterodimers. J. Biol. Chem., 264, 13238-13244.

175 Muenke, M, Schell, U, Hehr, A, Robin, NH, Losken, HW, Schinzel, A, Pulleyn, LJ, Rutland, P, Reardon, W and Malcolm, S (1994). A common mutation in the fibroblast growth factor receptor 1 gene in Pfeiffer syndrome. Nat. Genet., 8, 269-274.

Muller, WJ, Sinn, E, Pattengale, PK, Wallace, Rand Leder, P (1988). Single step induction of mammary adenocarcinoma in transgenic mice bearing the activated c-neu oncogene. Cell, 54, 105-115.

Mulligan, RC (1993). The basic science of gene therapy. Science, 260, 926-932.

Musgrove, EA, Hunter, LK, Lee, CSL, Swarbrick, A, Hui, R and Sutherland, R (2001). Cyclin Dl overexpression induces progestin resistance in T-47D breast cancer cells despite p27kipl association with cyclin-E-cdk2. J. Biol. Chem., 276, 47675-47683.

Muthuswamy, Sand Muller, W (1995). Direct and specific interaction of c-Src with Neu is involved in signaling by the epidermal growth factor receptor. Oncogene, 11, 271-279.

Nantel, A, Huber, M and Thomas, DY (1999). Localization of endogenous GrblO to the mitochondria and its interaction with the mitochondrial-associated Raf-1 pool. J. Biol. Chem., 274, 35719-35724.

Nantel, A, Mohammad-Ali, K, Sherk, J, Posner, B and Thomas, D (1998). Interaction of the GrblO adapter protein with the Rafl and Mekl kinases. J. Biol. Chem., 273, 10475- 10484.

Nassar, N, Horn, G, Herrmann, C, Scherer, A, McCormick, F and Wittinghofer, A (1995). The 2.2 A crystal structure of the Ras-binding domain of the serine/threonine kinase c-Rafl in complex with RaplA and a GTP analogue. Nature, 375, 554-560.

176 Nawaz, Z, Lonard, DM, Dennis, AP, Smith, CL and O'Malley, BW (1999). Proteasome­ dependent degradation of the human estrogen receptor. Proc. Natl. Acad. Sci. USA, 96, 1858-1862.

Neilson, KM and Friesel, R (1996). Ligand-independent activation of fibroblast growth factor receptors by point mutations in the extracellular, transmembrane, and kinase domains. J. Biol. Chem., 271, 25049-25057.

Nelms, K, O'Neill, TJ, Li, S, Hubbard, SR, Gustafson, TA and William, PE (1999). Alternative splicing, gene localization, and binding of SH2-B to the insulin receptor kinase domain. Mamm. Genome, 10, 1160-1167.

Niebuhr, K, Ebel, F, Frank, R, Reinhard, M, Domann, E, Carl, UD, Walter, U, Gertler, FB, Wehland, J and Chakraborty, T (1997). A novel proline-rich motif present in ActA of listeria monocytogenes and cytoskeletal proteins is the ligand for the EVHl domain, a protein module present in the Ena/V ASP family. EMBO J., 16, 5433-5444.

Noguchi, T, Matozaki, T, Horita, K, Fujioka, Y and Kasuga, M (1994). Role of SH-PTP2, a protein-tyrosine phosphatase with Src homology 2 domains, in insulin-simulated Ras activation. Mol. Cell. Biol., 14, 6674-6682.

Nunez, A-M, Jakowlev, S, Briand, J-P, Gaire, M, Krust, A, Rio, M-C and Chambon, P (1987). Characterization of the esrogen-induced pS2 protein secreted by the human breast cancer cell line MCF-7. Endocrinology, 121, 1759-1765.

O'Neill, TJ, Rose, DW, Pillay, TS, Hotta, K, Olefsky, JM and Gustafson, TA (1996). Interaction of a GRBIR splice variant (a human GRB 10 homolog) with the insulin and insulin-like growth factor I receptors-evidence for a role in mitogenic signaling. J. Biol. Chem., 271, 22506-22513.

177 Ogawa, W, Matozaki, T and Kasuga, M (1998). Role of binding proteins to IRS-1 in insulin signalling. Mol. Cell. Biochem., 182, 13-22.

Ong, SH, Guy, GR, Hadari, YR, Laks, S, Gotoh, N, Schlessinger, J and Lax, I (2000). FRS2 proteins recruit intracellular signaling pathways by binding to diverse targets on fibroblast growth factor and nerve growth factor receptors. Mol. Cell. Biol., 20, 979-989.

Ooi, J, Yajnik, V, Immanuel, D, Gordon, M, Moskow, JJ, Buchberg, AM and Margolis, B (1995). The cloning of GrblO reveals a new family of SH2 domain proteins. Oncogene, 10, 1621-1630.

Osborne, CK ( 1998). Steroid hormone receptors in breast cancer management. Breast Cancer Res. Treat., 51, 227-238.

Osborne, CK, Hong, ZH and Fuqua, AW (2000). Selective estrogen receptor modulators: structure, function, and clinical use. J. Clin. Oneal., 18, 3172-3186.

Ottenhoff-Kalff, AE, Rijksen, G, van Beurden, EACM, Hennipman, A, Michels, AA and Staal, GEJ ( 1992). Characterisation of protein tyrosine kinases from human breast cancer: involvement of the c-src oncogene product. Cancer Res., 52, 4773-4778.

Ottinger, EA, Botfield, MC and Shoelson, SE (1998). Tandem SH2 domains confer high specificity in tyrosine kinase signaling. J. Biol. Chem., 273, 729-735.

Pandey, A, Duan, H, Di Fiore, PP and Dixit, VM (1995). The ret receptor protein tyrosine kinase associates with the SH2-containing adapter protein Grbl0. J. Biol. Chem., 270, 21461-21463.

Pandey, A, Liu, X, Dixon, J, Di Fiore, P and Dixit, V (1996). Direct association between the ret receptor tyrosine kinase and the src homology 2-containing adapter protein Grb7. J. Biol. Chem., 271, 10607-10610.

178 Pandini, G, Vigneri, R, Costantino, A, Frasca, F, Ippolito, A, Fujita-Yamaguchi, Y, Siddle, K, Goldfine, ID and Belfiore, A (1999). Insulin and insulin-like growth factor-I (IGF-I) receptor overexpression in breast cancers leads to insulin/IGF-I hybrid receptor overexpression: evidence for a second mechanism of IGF-I signaling. Clin. Cancer Res., 5, 1935-1944.

Papa, V, Gliozzo, B, Clark, GM, McGuire, WL, Moore, D, Fujita-Yamaguchi, Y, Vigneri, R, Goldfine, ID and Pezzino, V (1993). Insulin-like growth factor-I receptors are overexpressed and predict a low risk in human breast cancer. Cancer Res., 53, 3736-3740.

Papa, V, Pezzino, V, Costantino, A, Belfiore, A, Giuffrida, D, Frittitta, L, Vannelli, G, Brand, R, Goldfine, I and Vigneri, R (1990). Elevated insulin receptor content in human breast cancer. J. Clin. Invest., 86, 1503-1510.

Parsons, J and Parsons, S ( 1997). Src family protein tyrosine kinases: cooperating with growth factor and adhesion signaling pathways. Curr. Opin. Cell Biol., 9, 187-192.

Patrone, C, Ma, ZQ, Pollio, G, Agrati, P, Parker, MG and Maggi, A (1996). Cross-coupling between insulin and estrogen receptor in human neuroblastoma cells. Mol. Endocrinol., 10, 499-507.

Patti, ME, Sun, XJ, Bruening, JC, Araki, E, Lipes, MA, White, MF and Kahn, CR (1995). 4PS/insulin receptor substrate (IRS)-2 is the alternative substrate of the insulin receptor in IRS-I-deficient mice. J. Biol. Chem., 270, 24670-24673.

Pawson, T (1995). Protein modules and signalling networks. Nature, 373, 573-580.

Pawson, T, Gish, GD and Nash, P (2001). SH2 domains, interaction modules and cellular wiring. Trends Cell Biol., 11, 504 -511.

179 Pawson, T and Nash, P (2000). Protein-protein interactions define specificity in signal transduction. Genes Dev., 14, 1027-1047.

Pawson, T and Schlessinger, J (1993). SH2 and SH3 domains. Curr. Biol., 3, 434-442.

Pawson, T and Scott, J ( 1997). Signaling through scaffold, anchoring and adaptor proteins. Science, 278, 2075-2080.

Payson, R, Wu, J, Liu, Y and Chiu, I (1996). The human FGF-8 gene localizes on chromosome 10q24 and is subjected to induction by androgen in breast cancer cells. Oncogene, 13, 47-53.

Pear, WS, Nolan, GP, Scott, ML and Baltimore, D (1993). Production of high-titer helper­ free retroviruses by transient transfection. Proc. Natl. Acad. Sci. USA, 90, 8392-8396.

Pederson, T and Rondinone, CM (2000). Regulation of proteins involved in insulin signaling pathways in differentiating human adipocytes. Biochem. Biophys. Res. Commun., 276, 162-168.

Peles, E, Levy, R, Or, E, Ullrich, A and Yarden, Y (1991). Oncogenic forms of the neu/HER2 tyrosine kinase are permanently coupled to phospholipase Cy. EMBO J., 10, 2077-2086.

Perks, CM and Holly, JMP (2000). Insulin-like growth factor binding proteins (IGFBPs) in breast cancer. J. Mammary Gland Biol. Neoplasia, 5, 75-84.

Pero, SC, Oligino, L, Daly, RJ, Soden, AL, Liu, C, Roller, PP, Li, P and Krag, DN (2002). Identification of novel non-phosphorylated ligands, which bind selectively to the SH2 domain of Grb7. J. Biol. Chem., 277, 11918-11926.

180 Peterson, CM (2000). Estrogen and progesterone receptors: An overview from the year 2000. J. Soc. Gynecol. Investig., 7, S3-7.

Peyrat, JP and Bonneterre, J (1992). Type 1 IGF receptor in human breast diseases. Breast Cancer Res. Treat., 22, 59-67.

Pham, N and Rotin, D (2001). Nedd4 regulates ubiquitination and stability of the guanine­ nucleotide exchange factor CNrasGEF. J. Biol. Chem., 276, 46995-47003.

Pietras, RJ, Arboleda, J, Reese, DM, Wongvipat, N, Pegram, MD, Ramos, L, Gorman, CM, Parker, MG, Sliwkowski, MX and Slamon, DJ (1995). HER-2 tyrosine kinase pathway targets estrogen receptor and promotes hormone-independent growth in human breast cancer cells. Oncogene, 10, 2435-2446.

Planas-Silva, MD and Weinberg, RA (1997a). Estrogen-dependent cyclin E-cdk2 activation through p21 redistribution. Mol. Cell. Biol., 17, 4059-4069.

Planas-Silva, MD and Weinberg, RA (1997b). The restriction point and control of cell proliferation. Curr. Opin. Cell Biol., 9, 768-772.

Plouffe, lJ (2000). Selective estrogen receptor modulators (SERMS) in clinical practice. J. Soc. Gynecol. Investig., 7, S38-S46.

Ponting, CP and Benjamin, DR (1996). A novel family of Ras-binding domains. Trends Biochem. Sci., 21, 422-425.

Porter, AC and Vaillancourt, RR (1998). Tyrosine kinase receptor-activated signal transduction pathways which lead to oncogenesis. Oncogene, 16, 1343-1352.

Power, RF, Mani, SK, Codina, J, Conneely, OM and O'Malley, BW (1991). Dopaminergic and ligand-independent activation of steroid hormone receptors. Science, 254, 1636-1639.

181 Prall, OWJ, Sarcevic, B, Musgrove, EA, Watts, CKW and Sutherland, RL (1997). Estrogen-induced activation of cdk4 and cdk2 during G 1-S phase progression is accompanied by increased cyclin Dl expression and decreased cyclin-dependent kinase inhibitor association with cyclin E-Cdk2. J Biol. Chem., 272, 10882-10894.

Prall, OWJ, Sarcevic, B, Musgrove, EA, Watts, CKW and Sutherland, RL (1998). c-myc or cyclin Dl mimics estrogen effects on cyclin E-cdk2 activation and cell cycle reentry. Mol. Cell. Biol., 18, 4499-4508.

Prenzel, N, Zwick, E, Leserer, Mand Ullrich, A (2000). Epidermal growth factor receptor: convergence point for signal integration and diversification. Breast Cancer Res., 2, 184- 190.

Pritchard, C and McMahon, M (1997). Raf revealed in life-or-death decisions. Nat. Genet., 16, 214-215.

Rauh, MJ, Blackmore, V, Andrechek, ER, Tortorice, CG, Daly, R, Lai, VK, Pawson, T, Cardiff, RD, Siegel, PM and Muller, WJ (1999). Accelerated mammary tumor development in mutant polyomavirus middle T transgenic mice expressing elevated levels of either the She or Grb2 adapter protein. Mol. Cell. Biol., 19, 8169-8179.

Rechler, MM and Nissley, SP (1986). Insulin-like growth factor (IGF)/somatomedin receptor subtypes: structure, function, and relationships to insulin receptors and IGF carrier proteins. Harm. Res., 24, 152-159.

Reilly, J, Mickey, G and Maher, P (2000). Association of fibroblast growth factor receptor 1 with the adaptor protein Grbl4. Characterization of a new receptor binding partner. J. Biol. Chem., 275, 7771-7778.

182 Resnik, JL, Reichart, DB, Huey, K, Webster, NJ and Seely, BL (1998). Elevated insulin­ like growth factor I receptor autophosphorylation and kinase activity in human breast cancer. Cancer Res., 58, 1159-1164.

Rocha, RL, Hilsenbeck, SG, Jackson, JG, van den Berg, CL, Weng, C-N, Lee, AV and Yee, D (1997). Insulin-like growth factor binding protein-3 and insulin receptor substrate-I in breast cancer: correlation with clinical parameters and disease-free survival. Clin. Cancer Res., 3, 103-109.

Rosen, N, Bolen, JB, Schwartz, AM, Cohen, P, DeSeau, V and Israel, MA (1986). Analysis of pp6oc-src protein kinase activity in human tumor cell lines and tissues. J. Biol. Chem., 261, 13754-13759.

Ross, JS and Fletcher, JA (1999). HER-2/neu (c-erb-B2) gene and protein in breast cancer. Am. J. Clin. Pathol., 112,

Rother, Kl, Imai, Y, Caruso, M, Beguinot, R, Formisano, P and Accili, D (1998). Evidence that IRS-2 phosphorylation is required for insulin action in hepatocytes. J. Biol. Chem., 273, 17491-17497.

Rotin, D, Staub, 0 and Haguenauer-Tsapis, R (2000). Ubiquitination and endocytosis of plasma membrane proteins: role of Nedd4/Rsp5p family of ubiquitin-protein ligases. J. Membr. Biol., 176, 1-17.

Rousseau, F, Bonaventure, J, Legeai-Mallet, L, Pelet, A, Rozet, JM, Maroteaux, P, Le Merrer, Mand Munnich, A (1994). Mutations in the gene encoding fibroblast growth factor receptor-3 in achondroplasia. Nature, 371, 252-254.

Ruohola, JK, Viitanen, TP, Valve, EM, Seppanen, JA, Loponen, NT, Keskitalo, JJ, Lakkakorpi, PT and Harkonen, PL (2001). Enhanced invasion and tumor growth of

183 fibroblast growth factor 8b-overexpressing MCF-7 human breast cancer cells. Cancer Res., 61, 4229-4237.

Saad, MJ, Folli, F, Araki, E, Hashimoto, N, Csermely, P and Kahn, CR (1994). Regulation of insulin receptor, insulin receptor substrate-I and phosphatidylinositol 3-kinase in 3T3- F442A adipocytes. Effects of differentiation, insulin, and dexamethasone. Mol. Endocrinol., 8, 545-557.

Sainsbury, JR, Farndon, JR, Needham, GK, Malcolm, AJ and Harris, AL (1987). Epidermal-growth-factor receptor status as predictor of early recurrence of and death from breast cancer. Lancet, i, 1398-1402.

Saito, H, Morita, Y, Fujimoto, M, Narazaki, M, Naka, T and Kishimoto, T (2000). IFN regulatory factor-I-mediated transcriptional activation of mouse STAT-induced STAT inhibitor-I gene promoter by IFN-y. J. Immunol., 164, 5833-5843.

Sakai, DD, Helms, S, Carlstedt-Duke, J, Gustafsson, JA, Rottman, FM and Yamamoto, KR (1988). Hormone-mediated repression: a negative glucocorticoid response element from the bovine prolactin gene. Genes Dev., 2, 1144-1154.

Salerno, M, Sisci, D, Mauro, L, Guvakova, MA, Ando, S and Surmacz, E (1999). Insulin receptor substrate 1 is a target for the pure antiestrogen ICI 182,780 in breast cancer cells. Int. J. Cancer, 81, 299-304.

Sambrook, J, Fritsch, EF and Maniatis, T (1989). Molecular cloning: a laboratory manual. Cold Spring Harbour Press, New York.

Sambrook, J and Russell, DW (2001). Molecular cloning: a laboratory manual. Cold Spring Harbour Press, New York. A8.20-A8.21.

184 Sasaki, S, Lesoon-Wood, LA, Dey, A, Kuwata, T, Weintraub, BD, Humphrey, G, Yang, WM, Seto, E, Yen, PM, Howard, BH and Ozato, K (1999). Ligand-induced recruitment of a histone deacetylase in the negative-feedback regulation of the thyrotropin beta gene. EMBO J., 18, 5389-5398.

Sato, TN, Tozawa, Y, Deutsch, U, Wolburg-Buchholz, K, Fujiwara, Y, Gendron-Maguire, M, Gridley, T, Wolburg, H, Risau, W and Qin, Y (1995). Distinct roles of the receptor tyrosine kinases Tie-1 and Tie-2 in blood vessel formation. Nature, 376, 70-74.

Sausville, E, Carney, D and Battey, J (1985). The human vasopressin gene is linked to the oxytocin gene and is selectively expressed in a cultured lung cancer cell line. J. Biol. Chem., 260, 10236-10241.

Sbodio, JI, Lodish, HF and Chi, N-W (2002). Tankyrase-2 oligomerizes with tankyrase-1 and binds to both TRFl (telomere-repeat-binding factor 1) and IRAP (insulin-responsive arninopeptidase). Biochem. J., 361, 451-459.

Schaller, MD, Hildebrand, JD, Shannon, JD, Fox, JW, Vines, RR and Parsons, JT (1994). Autophosphorylation of the focal adhesion kinase, pp125FAK, directs SH2-dependent binding of pp60src. Mol. Cell. Biol., 14, 1680-1688.

Schell, U, Hehr, A, Feldman, GJ, Robin, NH, Zackai, EH, de Die-Smulders, C, Viskochil, DH, Stewart, JM, Wolff, G and Ohashi, H (1995). Mutations in FGFRl and FGFR2 cause familial and sporadic Pfeiffer syndrome. Hum. Mol. Genet., 4, 323-328.

Schlessinger, J (2000). Cell signaling by receptor tyrosine kinases. Cell, 103, 211-225.

Schlessinger, J and Ullrich, A (1992). Growth factor signaling by receptor tyrosine kinases. Neuron, 9, 383-391.

185 Scholler, A, Hong, NJ, Bischer, P and Reiners, JJJ (1994). Short and long term effects of cytoskeleton-disrupting drugs on cytochrome P450 Cypla-1 induction in murine hepatoma lclc7 cells: suppression by the microtubule inhibitor nocodazole. Mol. Pharmacol., 45, 944-954.

Sciacca, L, Costantino, A, Pandini, G, Mineo, R, Frasca, F, Scalia, P, Sbraccia, P, Goldfine, ID, Vigneri, R and Belfiore, A (1999). Insulin receptor activation by IGF-11 in breast cancers: evidence for a new autocrine/paracrine mechanism. Oncogene, 18, 2471-2479.

Senapathy, P, Shapiro, MB and Harris, NL (1990). Splice junctions, branch point sites, and exons: sequence statistics, identification, and applications to genome project. Methods Enzymol., 183, 252-278.

Sepp-Lorenzino, L, Ma, Z, Lebwohl, DE, Vinitsky, A and Rosen, N (1995). Herbimycin A induces the 20 S proteasome- and ubiquitin-dependent degradation of receptor tyrosine kinases. J. Biol. Chem., 270, 16580-16587.

Shen, T-L and Han, DC (2002). Association of Grb7 with phosphoinositides and its role in the regulation of cell migration. J. Biol. Chem., in press, Manuscript 10.107 4/jbc.M203085200.

Siddle, K, Urso, B, Niesler, A, Cope, DL, Molina, L, Surinya, KH and Soos, MA (2001). Specificity in ligand binding and intracellular signalling by insulin and insulin-like growth factor receptors. Biochem. Soc. Trans., 29, 513-525.

Skolnik, EY, Batzer, A, Li, N, Lee, C-H, Lowenstein, E, Mohammadi, M, Margolis, Band Schlessinger, J (1993a). The function of GRB2 in linking the insulin receptor to ras signaling pathways. Science, 260, 1953-1955.

Skolnik, EY, Lee, C-H, Batzer, A, Vincentini, LM, Zhou, M, Daly, RJ, Myers Jr, MJ, Backer, JM, Ullrich, A, White, MF and Schlessinger, J (1993b). The SH2/SH3 domain-

186 containing protein GRB2 interacts with tyrosine-phosphorylated IRS 1 and She: implications for insulin control of ras signalling. EMBO J., 12, 1929-1936.

Skolnik, EY, Margolis, B, Mohammadi, M, Lowenstein, E, Fischer, R, Drepps, A, Ullrich, A and Schlessinger, J (1991). Cloning of PB kinase-associated p85 utilizing a novel method for expression/ cloning of target proteins for receptor tyrosine kinases. Cell, 65, 83- 90.

Slamon, DJ, Clark, GM, Wong, SG, Levin, WJ, Ullrich, A and McGuire, WL (1987). Human breast cancer: correlation of relapse and survival with amplification of the HER- 2/neu oncogene. Science, 235, 177-82.

Slamon, DJ, Godolphin, W, Jones, LA, Holt, JA, Wong, SG, Keith, DE, Levin, WJ, Stuart, SG, Udove, J, Ullrich, A and Press, MF (1989). Studies of the HER2/neu proto-oncogene in human breast and ovarian cancer. Science, 244, 707-712.

Smith, S, Giriat, I, Schmitt, A and de Lange, T (1998). Tankyrase, a poly(ADP-ribose) polymerase at human telomeres. Science, 282, 1484-1487.

Songyang, Z, Fanning, AS, Fu, C, Xu, J, Marfatia, SM, Chishti, AH, Crompton, A, Chan, AC, Anderson, JM and Cantley, LC (1997). Recognition of unique carboxyl-terminal motifs by distinct PDZ domains. Science, 275, 73-77.

Songyang, z, Shoelson, SE, McGlade, J, Olivier, P, Pawson, T, Bustelo, XR, Barbacid, M, Sabe, H, Hanafusa, H, Yi, T, Ren, R, Baltimore, D, Ratnofsky, S, Feldman, RA and Cantley, LC (1994). Specific motifs recognised by the SH2 domains of Csk, 3BP2, fps/fes, GRB-2, HCP, SHC, Syk, and Vav. Mol. Cell. Biol., 14, 2777-2785.

Soos, MA, Field, CE and Siddle, K (1993). Purified hybrid insulin/insulin-like growth factor-I receptors bind insulin-like growth factor-I, but not insulin, with high affinity. Biochem. J., 290, 419-426.

187 Soos, MA, Whittaker, J, Lammers, R, Ullrich, A and Siddle, K (1990). Receptors for insulin and insulin-like growth factor-I can form hybrid dimers. Characterisation of hybrid receptors in transfected cells. Biochem. J., 270, 383-390.

Spivak-Kroizman, T, Lemmon, MA, Dikic, I, Ladbury, JE, Pinchasi, D, Huang, J, Jaye, M, Crumley, G, Schlessinger, J and Lax, I (1994). Heparin-induced oligomerization of FGF molecules is responsible for FGF receptor dimerization, activation, and cell proliferation. Cell, 19, 1015-1024.

Stacey, D and Kazlauskas, A (2002). Regulation of Ras signaling by the cell cycle. Curr. Opin. Genet. Dev., 12, 44-46.

Stapleton, D, Balan, I, Pawson, T and Sicheri, F (1999). The crystal structure of an Eph receptor SAM domain reveals a mechanism for modular dimerization. Nat. Struct. Biol., 6, 44-49.

Starr, R, Willson, TA, Viney, EM, Murray, LJL, Rayner, JR, Jenkins, BJ, Gonda, TJ, Alexander, WS, Metcalf, D, Nicola, NA and Hilton, DJ (1997). A family of cytokine­ inducible inhibitors of signalling. Nature, 387, 917-921.

Stein, D, Wu, J, Fuqua, SAW, Roonprapunt, C, Yajnik, V, D'Eustachio, P, Moskow, J, Buchberg, AM, Osborne, CK and Margolis, B (1994). The SH2 domain protein Grb-7 is co-amplified, overexpressed and in a tight complex with HER2 in breast cancer. EMBO J., 13, 1331-1340.

Stein, E, Cerretti, D and Daniel, T (1996). Ligand activation of ELK receptor tyrosine kinase promotes its association with GrblO and Grb2 in vascular endothelial cells. J. Biol. Chem., 271, 23588-23593.

188 Stein, E, Gustafson, T and Hubbard, S (2001). The BPS domain of GrblO inhibits the catalytic activity of the insulin and IGFl receptors. FEBS Lett., 493, 106-111. Stem, DF (2000). ErbB family receptor tyrosine kinases. Breast Cancer Res., 2, 176-183.

Stewart, A, Westley, B and May, F (1992). Modulation of the proliferative response of breast cancer cells to growth factors by estrogen. Brit. J. Cancer, 66, 640-648.

Stewart, AJ, Johnson, MD, May, FEB and Westley, BR (1990). Role of insulin like growth factors and the type I insulin like growth factor receptor in the estrogen stimulated proliferation of human breast cancer cells. J. Biol. Chem., 265, 21172-21178. ~

Stickeler, E, Kittrell, F, Medina, D and Berget, SM (1999). Stage-specific changes in SR splicing factors and alternative splicing in mammary tumorigenesis. Oncogene, 18, 3574- 3582.

Stoica, A, Saceda, M, Fakhro, A, Joyner, Mand Martin, MB (2000). Role of insulin-like growth factor-I in regulating estrogen receptor-a gene expression. J. Cell. Biochem., 76, 605-614.

Stoll, BA (1996). Nutrition and breast cancer risk: can an effect via insulin resistance be demonstrated?. Breast Cancer Res. Treat., 38, 239-246.

Storrie, B and Yang, W (1998). Dynamics of the interphase mammalian Golgi complex as revealed through drugs producing reversible Golgi disassembly. Biochim. Biophys. Acta, 1404, 127-137.

Suen, KL, Bustelo, XR, Pawson, T and Barbacid, M (1993). Molecular cloning of the mouse grb2 gene: Differential expression of the Grb2 adaptor protein with epdermal growth factor and nerve growth factor receptors. Mol. Cell. Biol., 13, 5500-5512.

189 Surmacz, E (2000). Function of the IGF-IR in breast cancer. J. Mammary Gland Biol. Neoplasia, 5, 95-105.

Szebenyi, G and Fallon, JF (1999). Fibroblast growth factors as multifunctional signaling factors. Int. Rev. Cytol., 185, 45-106.

Tanaka, A, Furuya, A, Yamasaki, M, Hanai, N, Kuriki, K, Kamiakito, T, Kobayashi, Y, Yoshida, H, Koike, M and Fukayama, M ( 1998a). High frequency of fibroblast growth factor (FGF) 8 expression in clinical prostate cancers and breast tissues, immunohistochemically demonstrated by a newly established neutralizing monoclonal antibody against FGF 8. Cancer Res., 58, 2053-2056.

Tanaka, A, Miyamoto, K, Matsuo, H, Matsumoto, K and Yoshida, H ( 1995). Human androgen-induced growth factor in prostate and breast cancer cells: its molecular cloning and growth properties. FEBS Lett., 363, 226-230.

Tanaka, S, Mori, M, Akiyoshi, T, Tanaka, Y, Mafune, K, Wands, JR and Sugimachi, K (1997). Coexpression of Grb7 with epidermal growth factor receptor or Her2/erbB2 in human advanced esophageal carcinoma. Cancer Res., 57, 28-31.

Tanaka, S, Mori, M, Akiyoshi, T, Tanaka, Y, Mafune, K, Wands, JR and Sugimachi, K (1998b). A novel variant of human Grb7 is associated with invasive esophageal carcinoma. J. Clin. Invest., 102, 821-827.

Taylor, SI, Wertheimer, E, Hone, J, Levy-Toledano, R, Quon, MJ, Barbetti, F, Suzuki, Y, Roach, P, Koller, E, Haft, CR, de La Luz Sierra, M, Cama, A and Accili, D (1994). In: Molecular biology of diabetes, part II Draznin, Band LeRoith, D (eds). pp. 1-23. Humana Press Inc., New Jersey.

Thanos, CD, Goodwill, KE and Bowie, JU (1999). Oligomeric structure of the human EphB2 receptor SAM domain. Science, 283, 833-836.

190 Thommes, K, Lennartsson, J, Carlberg, Mand Ronnstrand, L (1999). Identification of Tyr- 703 and Tyr-936 as the primary association sites for Grb2 and Grb7 in the c-Kit/stem cell factor receptor. Biochem. J., 341, 211-216.

Thor, A, Liu, S, Edgerton, S, Moore, Dn, Kasowitz, K, Benz, C, Stem, D and DiGiovanna, M (2000). Activation (tyrosine phosphorylation) of ErbB-2 (HER-2/neu): a study of incidence and correlation with outcome in breast cancer. J. Clin. Oneal., 18, 3230-3239.

Tollervey, D and Caceres, JF (2000). RNA processing marches on. Cell, 103, 703-709. Treisman, R ( 1996). Regulation of transcription by MAP kinase cascades. Curr. Opin. Cell Biol., 8, 205-215.

Tremblay, A, Tremblay, GB, Labrie, F and Giguere, V (1999). Ligand-independent recruitment of SRC-1 to estrogen receptor ~ through phosphorylation of activation function AF-1. Mol. Cell, 3, 513-519.

Turner, BC, Haffty, BG, Narayanan, L, Yuan, J, Havre, PA, Gumbs, AA, Kaplan, L, L., BJ, Carter, D, Baserga, R and Glazer, PM (1997). Insulin-like growth factor-I receptor overexpression mediates cellular radioresistance and local breast cancer recurrence after lumpectomy and radiation. Cancer Res., 57, 3079-3083.

Ullrich, A and Schlessinger, J (1990). Signal transduction by receptors with tyrosine kinase activity. Cell, 61, 203-212. van der Burg, B, Rutteman, GR, Blankenstein, MA, de Laat, SW and van Zoelen, EJ ( 1988). Mitogenic stimulation of human breast cancer cells in a growth factor-defined medium:synergistic action of insulin and estrogen. J. Cell. Physiol., 134, 101-108. van der Geer, P and Hunter, T (1994). Receptor protein tyrosine kinases and their signal transduction pathways. Annu. Rev. Cell Biol., 10, 251-337.

191 Vayssiere, B, Zalcman, G, Mahe, Y, Mirey, G, Ligensa, T, Weidner, KM, Chardin, P and Camonis, J (2000). Interaction of the Grb7 adapter protein with Rndl, a new member of the Rho family. FEBS Lett., 467, 91-96.

Verbeek, B, Adriaansen-Slot, S, Rijken, G and Vroom, T (1997). Grb2 overexpression in nuclei and cytoplasm of human breast cells: a histochemical and biochemical study of normal and neoplastic mammary tissue specimens. J. Pathal., 183, 195-203.

Verbeek, B, Vroom, T, Adriaansen-Slot, S, Ottenholf-Kalff, AE, Geertzema, JG, Hennipman, A and Rijksen, G (1996). c-Src protein expression is increased in human breast cancer. An immunohistochemical and biochemical analysis. J. Pathal., 180, 383- 388.

Vidal, M, Gigoux, V and Garbay, C (2001). SH2 and SH3 domains as targets for anti­ proliferative agents. Crit. Rev. Oneal. Hematal., 40, 175-186.

Virkamaki, A, Ueki, K and Kahn, CR (1999). Protein-protein interaction in insulin signaling and the molecular mechanisms of insulin resistance. J. Clin. Invest., 103, 931-43.

Vogel, CL, Cobleigh, MA, Tripathy, D, Gutheil, JC, Harris, LN, Fehrenbacher, L, Slamon, DJ, Murphy, M, Novotny, WF, Burchmore, M, Shak, S, Stewart, SJ and Press, M (2002). Efficacy and safety of trastuzumab as a single agent in first-line treatment of HER2- overexpressing metastatic breast cancer. J. Clin. Oneal., 20, 719-726.

Waksman, G, Shoelson, SE, Pant, N, Cowbum, D and Kuriyan, J (1993). Binding of a high affinity phosphotyrosyl peptide to the Src SH2 domain: crystal structures of the complexed and peptide-free forms. Cell, 72, 779-790.

192 Walker, P, Germond, J-E, Brown-Luedi, M, Givel, F and Wahli, W (1984). Sequence homologies in the region preceding the transcription initiation site of the liver estrogen­ responsive vitellogenin and apo-VLDLII genes. Nucleic Acids Res., 12, 8611-8626.

Walter, P, Green, S, Greene, G, Krust, A, Bomert, JM, Jeltsch, JM, Staub, A, Jensen, E, Scrace, G and Waterfield, M (1985). Cloning of the human estrogen receptor cDNA. Proc. Natl. Acad. Sci. USA, 82, 7889-7893.

Wang, HG, Rapp, UR and Reed, JC (1996). Bcl-2 targets the protein kinase Raf-1 to mitochondria. Cell, 87, 629-638.

Wang, J, Dai, H, Yosaf, N, Moussaif, M, Deng, Y, Boufelliga, A, Swamy, R, Leone, Mand Riedel, H (1999). Orb 10, a positive, stimulatory signaling adapter in platelet-derived growth factor BB-, insulin-like growth factor-I, and insulin-mediated mitogenesis. Mol. Cell. Biol., 19, 6217-6228.

Wang, Q, Green, RP, Zhao, G and Ornitz, DM (2001). Differential regulation of endochondral bone growth and joint development by FGFRl and FGFR3 tyrosine kinase domains. Development, 128, S3867-3876.

Watanabe, K, Fukuchi, T, Hosoya, H, Shirasawa, T, Matuoka, K, Miki, Hand Takenawa, T (1995). Splicing isoforms of Rat Ash/Grb2. J. Biol. Chem., 270, 13733-13739.

Waters, S, Holt, K, Ross, S, Syu, L-J, Guan, K-L, Saltiel, A, Koretzky, G and Pessin, J (1995a). Desensitization of ras activation by a feedback disassociation of the SOS-Grb2 complex. J. Biol. Chem., 270, 20883-20886.

Waters, SB, Yamauchi, Kand Pessin, JE (1995b). Insulin-stimulated disassociation of the SOS-Grb2 complex. Mol. Cell. Biol., 15, 2791-2799.

193 Watts, CK, Brady, A, Sarcevic, B, deFazio, A, Musgrove, EA and Sutherland, RL (1995). Antiestrogen inhibition of cell cycle progression in breast cancer cells is associated with inhibition of cyclin-dependent kinase activity and decreased retinoblastoma protein phosphorylation. Mol. Endocrinol., 9, 1804-1813.

Weiderpass, E, Gridley, G, Persson, I, Nyren, 0, Ekbom, A and Adami, HO (1997). Risk of endometrial and breast cancer in patients with diabetes mellitus. Int. J. Cancer, 71, 360- 363.

White, MF (1998). The IRS-signalling system: A network of docking proteins that mediate insulin action. Mol. Cell. Biochem., 182, 3-11.

White, MF and Kahn, CR (1994). The insulin signaling system. J. Biol. Chem., 269, 1-4.

Whitehead, JP, Clark, SF, Urso, B and James, DE (2000). Signalling through the insulin receptor. Curr. Opin. Cell Biol., 12, 222-228.

Wick, MJ, Dong, LQ, Hu, D, Langlais, P and Liu, F (2001). Insulin receptor-mediated p62dok tyrosine phosphorylation at residues 362 and 398 plays distinct roles for binding GTPase-activating protein and Nck and is essential for inhibiting insulin-stimulated activation of Ras and Akt. J. Biol. Chem., 276, 42843-42850.

Wideroff, L, Gridley, G, Mellemkjaer, L, Chow, WH, Linet, M, Keehn, S, Borch-Johnsen, K and Olsen, JH ( 1997). Cancer incidence in a population-based cohort of patients hospitalized with diabetes mellitus in Denmark. J. Natl. Cancer Inst., 89, 1360-1365.

Wiseman, LR, Johnson, MD, Wakeling, AE, Lykkesfeldt, AE, May, FE and Westley, BR (1993). Type 1 IGF receptor and acquired tamoxifen resistance in oestrogen-responsive human breast cancer cells. Eur. J. Cancer, 29A, 2256-2264.

194 Withers, DJ and White, MF (2000). Perspective: The insulin signaling system-a common link in the pathogenesis ofType2 diabetes. Endocrinology, 141, 1917-1921.

Wojcik, J, Girault, J, Labesse, G, Chomilier, J, Momon, J and Callebaut, I (1999). Sequence analysis identifies a ras-associating (RA)-like domain in the N-termini of band 4.1/JEF domains and in the Grb7/10/14 adapter family. Biochem. Biophys. Res. Commun., 259, 113-120.

Wood, TL and Douglas, Y (2000). IGFs and IGFBPs in the normal mammary gland and breast cancer. J. Mammary Gland Biol. Neoplasia, 5, 1-5.

Wu, Y and Koenig, RJ (2000). Gene regulation by thyroid hormone. Trends Endocrinol. Metab., 11, 207-211.

Xiao, S, Nalabolu, S, Aster, J, Ma, J, Abruzzo, L, Jaffe, E, Stone, R, Weissman, S, Hudson, T and Fletcher, J (1998). FGFRl is fused with a novel zinc-finger gene, ZNF198, in the t(8;13) leukaemia/lymphoma syndrome. Nat. Genet., 18, 84-87.

Xing, Z, Chen, HC, Nowlen, JK, Taylor, SJ, Shalloway, D and Guan, JL (1994). Direct interaction of v-Src with the focal adhesion kinase mediated by the Src SH2 domain. Mol. Biol. Cell, 5, 413-421.

Xu, H, Kyung, LW and Goldfarb, M (1998). Novel recognition motif on fibroblast growth factor receptor mediates direct association and activation of SNT adapter proteins. J. Biol. Chem., 273, 17987-17990.

Yamaguchi, Y, Flier, JS, Yokota, A, Benecke, H, Backer, JM and Moller, DE (1991). Functional properties of two naturally occurring isoforms of the human insulin receptor in Chinese hamster ovary cells. Endocrinology, 129, 2058-2066.

195 Yamashita, T, Yoshioka, Mand Itoh, N (2000). Identification of a novel fibroblast growth factor, FGF-23, preferentially expressed in the ventrolateral thalarnic nucleus of the brain. Biochem. Biophys. Res. Commun., 277, 494-498.

Yan, KS, Kuti, M, Yan, S, Mujtaba, S, Farooq, A, Goldfarb, MP and Zhou, M-M (2002). FRS2 PTB domain conformation regulates interactions with divergent neurotrophic receptors. J. Biol. Chem., 277, 17088-17094.

Ye, D, Mendelsohn, J and Fan, Z (1999). Augmentation of a humanized anti-HER2 mAb 4D5 induced growth inhibition by a human-mouse chimeric anti-EGF receptor mAb C225. Oncogene, 18, 731-738.

Yee, D, Cullen, JK, Paik, S, Perdue, JF, Hampton, B, Schwartz, A, Lippman, ME and Rosen, N (1988). Insulin-like growth factor II mRNA expression in human breast cancer. Cancer Res., 48, 6691-6696.

Yee, D and Lee, AV (2000). Crosstalk between the insulin-like growth factors and estrogens in breast cancer. J. Mammary Gland Biol. Neoplasia, 5, 107-115.

Yee, D, Sharma, J and Hilsenbeck, SG (1994 ). Prognostic significance of insulin-like growth factor-binding protein expression in axillary lymph node-negative breast cancer. J. Natl. Cancer Inst., 86, 1785-1789.

Yip, S, Crew, AJ, Gee, JM, Hui, R, Blarney, RW, Robertson, JF, Nicholson, RI, Sutherland, RL and Daly, RJ (2000). Up-regulation of the protein tyrosine phosphatase SHP-1 in human breast cancer and correlation with GRB2 expression. Int. J. Cancer, 88, 363-368.

Yokote, K, Margolis, B, Heldin, C-H and Claesson-Welsh, L (1996). Grb7 is a downstream signaling component of platelet-derived growth factor a- and ~-receptors. J. Biol. Chem., 271, 30942-30949.

196 Zahler, AM, Neugerbauer, KM, Lane, WS and Roth, MB (1993). Distinct functions of SR proteins in alternative pre-mRNA splicing. Science, 260, 219-222.

Zhan, X, Plourde, C, Hu, X, Friesel, R and Maciag, T (1994). Association of fibroblast growth factor receptor- I with c-Src correlates with association between c-Src and cortactin. J. Biol. Chem., 269, 20221-20224.

Zhang, J, Somani, AK and Siminovitch, KA (2000). Roles of the SHP-1 tyrosine phosphatase in the negative regulation of cell signalling. Semin. Immunol., 12, 361-678.

Zhang, L, Kharbanda, S, Chen, D, Bullocks, J, Miller, D, Ding, I, Hanfelt, J, McLeskey, S and Kern, F ( 1997). MCF-7 breast carcinoma cells overexpressing FGF-1 form vascularized, metastatic tumors in ovariectomized or tamoxifen-treated nude mice. Oncogene, 15, 2093-2108.

Zhang, X, Chattopadhyay, A, Ji, QS, Owen, JD, Ruest, PJ, Carpenter, G and Hanks, SK (1999). Focal adhesion kinase promotes phospholipase C-gammal activity. Proc. Natl. Acad. Sci. USA, 96, 9021-9026.

Zhang, X and Yee, D (2000). Insulin-like growth factors and their receptors in breast cancer. Breast Cancer Res., 2, 170-175.

Zhao, JH, Reiske, H and Guan, JL (1998). Regulation of the cell cycle by focal adhesion kinase. J. Cell Biol., 143, 1997-2008.

197 !ll._pJJendlx !Jl

Oligonucleotide sequences To amplify overlapping Grb14 cDNA regions (see section 3.2 and Figure 3.1) by PCR: G IF 5' -ATGACCACTTCCCTGCAAGA-3' G lR 5' -CAACAGCTGACAAACATCTCGAGC-3'

G2F 5' -CAGGTGATT AAA GTATACAGTGAAGATGAA-3' G2R 5'-CTGTTCTTTCGCATGTAAGAAACC-3'

G3F 5'-GAGCATATGGTATCTTTTGCAACT-3' G3R 5'-TCTGGTACAGCTGCATGCCATACT-3'

G4F 5'-GGATTCTGCTTTAAGCCTAAC-3' G4R 5'-GTGGTGAAACCATGGCTGGGA-3'

G5F 5'-TCGCTTGGAGGAAAAAAGGATGTT-3' G5R 5' -CTAGAGAGCAATCCTAGCACAATA-3'

To perform Grb14 ~-specific PCR, G5DEL2 was utilised in combination with G5R for amplification of the cDNA corresponding to the truncated SH2 domain as shown in the following schematic (not to scale): G5DEL2 CAAGGACTTGTGGATGGGTAGA

5' r, Deleted L.J sequence ._____I ______.

G5DEL2 ...... +-0G5R

198 To amplify the Grb14 SH2 deletion boundaries from genomic DNA: GEN IF 5'-GGCTCAGCGATTGATTATTCAG-3' GEN lR 5'-CCATGACTCATTGACAGTACGAAAG-3'

GEN 2F 5'-CTTTCGTACTGTCAATGAGTCATGG-3' GEN 2R 5'-GAACTCCACCAGCTGTATTAGATCTG-3'

To amplify the Grb14 a BPS.SH2 domain (note that restriction sites are underlined): BPS14.1 5'-CGCGGATCCTGCAGTTCACAGAGCAT BamHI Grb14.2(2) 5' -AGACTGAATTCCTAGAGAGCAATCCTAGCACA EcoRI

To amplify the Grb14 ~ BPS.SH2 domain, BPS14.1 was used in conjunction with Del GST-R(2): Del GST-R(2) 5'-AGACTGAATTCCTACCCATCCACAAGTCCTTGCTG-3'

To amplify the Grb 14 ~ sequence: Del Flag-F 5' -TACATAAGCTTACAATGACCACTTCCCTG-3' HindIII Del Flag-R 5' -TACATGGATCCCCCATCCACAAGTCCTTG-3'

Del GST-F 5'-TACATGGATCCAGCTCTGCCACAAACATG-3' Del GST-R 5'-AGACTGAATTCCTACCCATCCACAAGTCCTTG-3'

To sequence Grb14 ~ SH2 in pGEX-2T (see Appendix B): pGEX forward 5'-GGCAAGCCACGTTTGGT-3' pGEX reverse 5'-TGTCAGAGGTTTTCACCG-3'

199 Cycle sequencing The additional RT-PCR product amplified using G5 primers was sequenced using T3 and T7 primers (Promega) at an annealing temperature of 50°C. The 184 library clones were excised from phage as described previously (Daly et al., 1996) and transformed into E.coli ultracompetent cells (Stratagene) prior to purification using the Wizard™ DNA miniprep kit (Promega). These clones were sequenced using the following primers at the indicated annealing temperatures: T7 and G5F (60°C), SP6 (Promega, 50°C). Grb14 (3 SH2 in pGEX-2T was sequenced using pGEX forward (56°C) and pGEX reverse (54°C) primers.

Grb 14 (3/pRCCMV Flag was sequenced as follows: Annealing Primer (refer to Appendix A) temperature (°C) SP6 50

Del Flag-R 52

T7 56

G2F, G2R, G4R, 60

G3F, G4F, G5F 66

G3R 70

GlR 72

200 'l'ubfications and!ll.wards arisin,gfrom this thesis

Publications

Kairouz R and Daly R J (2000) Modulation of tyrosine kinase signalling in human breast cancer through altered expression of signalling intermediates Breast Cancer Res. 2000, 2:197-202

Kairouz R, Parmar J, Lyons R, Swarbrick A, Musgrove E, Daly RJ (2002) The Grb14 adaptor protein is a hormonally modulated repressor of both insulin- and estrogen-induced cell cycle progression in MCF-7 human breast cancer cells (Under Revision)

Kairouz R, Malaney SM, Sanderson GM, Daly RJ (2002) Identification of a Grb14 splice variant highlights two modes of receptor recruitment (Submitted)

Awards

1999 Lorne Cancer Conference Presentation A ward

NIH Travel Award (based on scientific merit) (2000 Keystone symposia: Assembly of signalling networks)

201