Life Science). Oligonudeotide primers homologous to the T7 and T3 binding sites within the vector ~~'AATACGACTCACTATA~~'
INFORMATION TO USERS This manusai*pthas been reproduced frwn the microfilm master. UMI films the
text directly from the original or copy submitted.
Thus, some thesis and
dissertation copies are in typewriter face, while others may be frwn any type of computer printer.
The quality of this mpducfion is dependant upon the quality of thr copy submitW. Broken or indistinct print, cobfed or poor quality illwbations and photographs, print bleedttrrough, substandard margins, and improper alignment
can adversely affect reproducbi. In the unlikely event that the author did not send UMI a complete manwaipt and
there are missing pages, them will k noted. Also, if unauthorized copyright
material had to be removed, a note will indicate We deletion.
Oversize materials (8.g.. maps. drawings, charts) are reproduced by sectioning the original, beginning at the upper lefthand comer and continuing hwn left to right in equal sections with small overlaps.
Photographs induded in ttie original manuscript have beem reproduced xerographically in this copy. Higher quality 6' x S* black and white photographic prints am available for any photographs or illus~*oc~s appearing in this copy for an additional charge. Contact UMI directly to order.
Bell & Howell information and Leaning 300 North Zwb Road, Ann AMr, MI 481061346 USA
Generation of Mutants of the Met Tyrosine Kinase Oncoproteinwith Substrate Binding Specificity
Department of Biochemistry McGill University, Montreal January1998
A thesis submitted to the Faculty of Graduate Studies and Research in partial fulfilment of the requirements of the degree of Master of Science.
National Library
Bibliothbque nationale du Canada
Acquisitions and Bibliographic Services
Acquisitions et services bibliographiques
395 Wellington Street Onawa ON K I A ON4
395, rue Wellington OtlawaON KlAON4
Canada
Canada
The author has granted a nonexclusive licence allowing the National Library of Canada to reproduce, loan, distn'bute or sell copies of this thesis in microform, paper or electronic formats.
L'auteur a accorde une licence non exclusive permettaut B la Bibliotheque nationale du Canada de reproduire, preter, distribuer ou vendre des copies de cette these sous la forme de m i c r o f i c h d ~de reproduction sur papier ou sur format electronique.
The author retains ownership of the copyright in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission.
L'auteur conserve la propriete du droit d'auteur qui protege cette these. Ni la these ni des extraits substantiels de ceUe-ci ne doivent Etre imprimes ou autrement reproduits sans son autorisation.
ABSTRACT
Receptor tyrosine kinases (RTKs) mediate transduction of extracellular signals to the interior of the cell. The oncogenic variant of the Met RTK (Tpr-Met) tsansforms fibroblast cells. Multiple proteins involved in activating signal
transduction pathways bind the Tpr-Met protein at a single highly
phosphorylated tyrosine, tyrosine 489. This tyrosine residue is essential for TprMet mediated transformation, implicating one or more of these substrates in this
event. The specificity of the interaction between RTKs and their substrates is
determined by amino acids surrounding the phosphotyrosine residue. To
determine the role of each Tpr-Met substrate in cellular tansformation, I have designed Tpr-Met mutants predicted to have a unique substrate binding specificity, by altering the three amino adds immediately downstream of tyrosine 489. My results show that the in vitro optimal binding motifs predicted for various substrates may not apply in oivo to Tpr-Met.
RESUME
Les sdcepteurs tyrosine-kinases (RTK) transmettent des signaux extracelldaires vers l1int&ieurde la cellule. La forme oncoghique du RTK Met (Tpr-Met) induit la transformation des fibroblastes. De nombreuses
prot6ines impliqu&s dans l'activation des voies de signalisation se bent B la proteine Tpr-Met par une seule tyrosine hyper-phosphory16ee.la tyrosine 489. Cette tyrosine est essentielle pour l'indudion de la transformation par Tpr-
Met, par l'interm6diaire d'un ou de plusieurs de ces substrats. La specificit6 de
l'interaction entre les RTK et leurs substrats est detennin4e par les acides amin& adjacent au residu phospho-tyrosine. Afin de determiner le r6le de
chacun des substrats de Tpr-Met dans la transformation cellulaire, j'ai construit des mutants de Tpr-Met potentiellement capables de her un seul
substrat, en modifiant les trois acides aminb localises immkdiatement en aval de la tyrosine 489. Mes resultats ont montre que les motifs dfacides amines qui ont kt6 definis comrne essentiels pour la liaison de la tyrosine
phosphoryk B des substrats sp4dfiques ne s'appliquent pas dans le contexte de la proteine Tpr-Met.
iii
ACKNOWLEDGEMENTS
I would like to express my gratitude to my supemisor, Dr. Morag Park for providing this opportunity to participate in scientific research and for her encourgement and guidance.
I extend my appreciation to the members of the Park Laboratory for their advice, inspiration, and infinite support. I would like to thank Dr. Elizabeth
Fixman, Tanya Fournier, Chu Pham Dang Hum, Darren Kamikura, Hanane Khoury, Louie Lamorte, Lisa Laurie, Jenny Lin,Dr. Christiane Maroun, Monica
Naujokas, Dr. Linh Nyguen, Dr. Isabelle Royale and Dr. Caroline Saucier for
introduction to experimental techniques, helpful discussions and insight, and for a careful reading of the manuscript. I wish to also thank Darren Kamikua, Chu
Pham Dang Huan and Maurice EMis for their technical assistance, Dr. Isabelle
Royal, Dr. Nathalie Martin and Anna Moraitis for translation of the abstract, and especially Louie Lamorte, Chu Pham Dang Hum and Jenny Lin for their encouragement, patience, generosity and friendship.
Finally, I wish to thank my parents for their unconditional support and
encouragement,and aU those friends and family whose kind words and deeds have always kept me moving forward.
TABLE OF CONTENTS Abstract
Acknowledgments Table of Contents List of Figures and Tables Introduction
I.
Receptor Tyrosine Kinases Receptor Tyrosine Kinase Activation and Signal 1.i. Transduction Ligand Binding and Receptor Activation Lii. 1.K Autophosphorylation I.iv. Signal Transduction
II.
Met Receptor Tyrosine Kinase II.i Hepatocyte Growth FactodScatter Factor Structure Expression and Function Met Receptor and HGF/SF Function In Vim Kii* Effects on Cell Proliferation Effects on Motility Effects on Morphogenesis Effects on Tumor Progression
m.
Chcogenic Activation of Receptor Tyrosine Kinases Overexpression of RTKs in Human Cancers Eli. m.ii. Activating Mutations of Receptor Tyrosine
IV.
III.iii.
Kinases Oncogenic Activation by Gene Rearrangement
m.iv. m.v.
Tpr-Met Oncogene Mechanism of Activation of Tpr-Met
Met Receptor and Oncoprotein: Activation and Substrate
Binding
.
IVii.
Autophosphorylation of Met Receptor and Tpr-Met Oncoprotein Interaction of Met Receptor with Submtes
vii
.
Interaction of Tpr-Met with Substrates N.iv. Domains Mediating Protein-Protein Interactions N.v. Src Homolgy 2 Domains IV.vi. SH2 Domain Binding Specifiaty IV.vii. Phosphotyrosine-Binding Domain IV.viii. Met Receptor and Oncoprotein Consensus substrate-s in ding Sites V.
Substrates Involved in Signalling by the Met Receptor Oncoprotein V.i. Substrates Involved in Met Receptor Mediated Biological Activity Substrates Involved in Tpr-Met Mediated Vii. Transformation
VI.
Conclusion
Materials and Methods Results
Discussion References
LIST OF FIGURES AND TABLES Introduction -
Figure
Activation of transmembrane receptors by ligand binding
Figure
Subdasses of receptor tyrosine kinases
Figure
Ligand binding results in receptor oligomerisation and activation by autophosphorylation of tyrosine residues
Figure
Ligand binding to RTKs
Figure
Complexes of S H 2 domain containing substrates that bind activated receptors
Figure
Generation of the mature Met receptor tyrosine kinase
Figure
Conversion of pro-HGF/SF to an active heterodimer
Figure 8
Met receptor and oncoprotein
19
Figure 9
Constitutive dimerisation mediated by a leucine zipper motif in Tpr leads to phophorylation of Tpr-Met and subsequent activation of its oncogenic potential
20
Figure 10
Tyrosine phosphorylation sites in Tpr-Met and the Met receptor that are essential for catalytic and biological activity
Figure 11
Substrate binding sites in the PTXFRB and Tpr-Met
Figure 12
S H 2 domain structures of various substrates
Table 1
Receptor Tyrosine Kinase Expression in Human Cancer
vii
Table 2
Optimal Binding Sequences for Various S H 2 Domains
Table 3
Y1349 and Y1356 of the Met Receptor Tyrosine Kinase and Y482 and Y489 of the Tpc-Met Oncoprotein Form a Degenerate Consensus Binding Site for Various Signal Transduction Molecules
Table 4
Transformation Potential and Substrate Binding Ability of Tpr-Met Mutants
32
Materials and Methods
Figure
Cloning of Spe I fragment of Tpr-Met into pBSKSII+ vector
Results Figure
Sitedirected mutagenesis
Figure
Cassette mutagenesis
Figure
Expression and kinase activity of Tpr-Met mutants designed to bind Grb2 only
Figure
Expression, k i w e activity and substrate binding ability of Tpr-Met mutant designed to bind p85 only
Figure
In vifro assoaation assay of Tpr-Met mutants designed to bind GrbZ only using GST fusion proteins
Figure
In oitro assodation assay of Tpr-Met mutants with potential substrate binding specificity using lysates of serum starved cells
Table 5
Optimal Binding Sequences of SH2 Domains of Tpr-Met Substrates
viii
60
Table 6
Amino Acids in Positions +1, +2 and +3 Were Altered in Wild Type Tpr-Met to Generate TprMet Mutants Designed to Specifically Bind Grb2 (Tpr-Met Grb2) or the p85 Subunit of PUK (Tpr-Met p85)
Table 7
Characterisation of Tpr-Met Mutants with Predicted Substrate Binding Specificity
45
INTRODUCTION Cellular processes such as growth, differentiation and motility are directed by external signals. The binding of extracellular growth factors, hormones or
cytokines to the external domain of specific receptors on the surface of responsive cells is the first step in the transduction of these signals to the cytoplasm or
nucleus of the cell (147)(Figure 1).
factor
growth
differentiation motility
Figure 1: Activation of transmembrane receptors by ligand binding
One class of such surface receptors consists of those with intrinsic tyrosine
kinase activity (193). Alterations in receptor tyrosine kinases (RTKs) can lead to a wide variety of human disorders such as insulin-resistant diabetes, piebaldism or
cancer (149),demonstrating that they are critical for normal physiological
processes.
I. RECEPTOR TYROSINE KINASES
Receptor tyrosine kinases (RTKs)are a subgroup of the protein kinase family of enzymes (68) and consist of a large extracellular ligand binding domain,
a single transmembrane domain and a cytoplasmic domain with tyrosine specific
kinase activity (193). RTKs have thus far been divided into 10 subclasses based on sequence similarity within the kinase domain and structural homology within the extracellular domain (193)(Figure 2). These indude the epidermal growth factor receptor family (I), the insulin receptor family 0, the platelet derived growth factor receptor family (ID),the fibroblast growth factor receptor family (N),the neurotsophin receptor family (V), the hepatocyte growth fador receptor family (VI), the vascular endothelid growth factor receptor family
(m,the Eph like
receptor family (VIII), the A d / Ark/Ufo receptor family (R), and the Tie receptor
family (X) (71). Within each dass, primary sequence and secondary structural similarities exist within extracellular domains which contain consensus motifs
such as cysteine rich repeat sequences, immunoglobin like (IG) domains, epidermal growth factor (EGF) like domains, acid boxes, leucine rich domains
and fibronectin type Kt (Fn-ii domains (71) that are important for the tertiary structure of the protein.
The protein kinase family includes a second class of kinases: serine/ threonine-specific kinases (68).The tyrosine kinases and serine/ threonine
kinases have particular short stretches of amino acids in their catalytic domains
which distinguish each type of b a s e (67).The catalytic domains of each dass of protein kinases have similar primary structme and are approximately 250-300
amino acids in size (68). The structure consists of eleven subdomains that are
highly consenred, separated by regions of low conservation (68).Many of the invariant or nearly invariant residues are involved in binding adenosine triphosphate (ATP)and the phosphotansfer reaction (68).Certain sub-
of
RTKs (III IV, , W,IX,X ) contain kinase domains that are split in two by amino
add insertions (193)which separate the ATP binding site in the amino terminal
domain from the catalytic moiety in the carboxv terminal domain.
Figure 2: Subclasses of receptor tyrosine kinases (from 71).Kinase domain (Closed rectangles); cysteine rich domains (not related to each other) (dotted rectangles); FNlII domains (hatched rectangles); IG domains (half circles); leucine rich domain (zig-zag line); add box (verticai~inesrectungle);EGF iike domains (closed diamonds). Separated dosed rectangles represent split kinase domains, where the amino binding site and the carboxy terminal domain t&d domain contains the contains the catalpc moiety. -
1.i. Receptor Tyrosine Kiruse Activation and Signal Transduction
The low intrinsic catalytic activity of RTKs is dramatically elevated upon ligand binding, resulting in the transmission of an external signal to the inside of the cell, and ultimately to the activation of a series of signalling networks that
transmit the signal through the cytosol to the nudeus (147). Full signalling capacity of RTKs requixes both ligand induced phosphorylation 3
residues essential for catalyhc activity and phosphorylation of tyrosine residues required for substrate association (Figure 3).
a
ligand binding
receptor dirnerisation r phosphorylation of phosphorylation of kinase tyrosine residues domain tyrosine residues outside the kinase kinase domain activation domain
Figure 3: Ligand binding results in receptor oligomerisation and activation by autophosphoryiation of tyrosine residues
13. Ligand Binding and Receptor Activation
Experimental evidence has demonstrated that the activation of the kinase activity of RTKs by the binding of ligand requires the dimerisation or
oligomerisationof the receptors (reviewed in 70). The binding of a ligand to the extracellular domain of the receptor is thought to promote a change in conformation of this region stimulating the dimerisation/oligomerisation of
receptors (193)(Figure 3).
Ligand structure is designed to facilitate the oligomerisation of receptors (Figure 4). The platelet derived growth factor (PDGF)and colony-stimulating factor 1 (CSF1)forexample, are dimeric ligands linked by disulfide bonds, while
noncovalent bonds conned the dimeric stem cell factor (SCF)(70). In each case, the dimeric Qand possesses two identical epitopes which bind two receptors at
the same time (70). Still other ligands, such as the basic fibroblast gowth factor-2
(FGF-2)are thought to utilise heparin which orients and stabilises the factor correctly, to form the correct high-affinitydimer or tetramer configuration (134).
FGF-2 and heparin form a cis -dimer complex that is the minimum active structural unit, although biological activity may require a tetrameric structure
(134).In contrast to dimeric ligands, the epidermal growth factor is monomeric and binds two receptors simultaneously (113) via two distinct receptor binding
domains.
Figure 4:Ligand binding to R m .Platelet derived growth factor (PDGF); stem cell factor (SCF);fibroblast growth factor-;! (FGF-2); epidermal growth factor
1.ii.i.Autophosphorylation
Phosphorylation of tyrosine residues of RTKs is accomplished through an autophosphorylation mechanism which follows ligand induced dirnerisation. Studies of the crystal structure of the kinase domain of the non-ligand bound
insulin receptor which has a low basal catalytic activity and is not phosphorylated on tyrosine residues (75)and that of an insulin receptor tyrosine
kinase assoaated with an ATP andog and peptide substrate (74) show that the b a s e domain folds into two lobes (Nand C) between which lies an activation (A) loop. A comparison of the a y s t a l structwes with and without an analogue of
ATP revealed that in the nonactive, unphosphorylated receptor, the substrate and ATP binding sites are occupied by the A loop, whereas in the ATP bound crystal, the A loop is released. This has suggested that upon ligand binding, the kinase domain of the receptor undergoes a conformational change, releasing the A loop from the ATP binding pocket and allowing ATP to bind. In the insulin receptor,
the A loop contains three tyrosine residues. Kinetic and structure function studies have shown that the phosphorylation of these tyrosine residues within the A loop
is critical for activation of the catalytic activity of the insdin receptor (49,166).
The crystal structure suggests that phosphorylation of these tyrosines occurs in trans by the adjacent receptor catalytic domain and that electrostatic interactions
of the phosphate groups with other residues in the catalytic domain stablises the A loop in an open conformation allowing ATP to bind.
Kinase activity and thus the biological function of the insulin receptor depends on the autophosphorylation of the three A loop tyrosines (49, 166).
Many receptor tyrosine kinases contain conserved tyrosine residues within the A loop. Moreover, autophosphoryiationof tyrosine(s) in the A loop is essential for biological function of many other RTKs, including insulin-like growth factor 1
receptor (go), fibroblast growth -or
receptor (129),nerve growth factor receptor
(127), brain-derived neurotrophic factor receptor (125)and Met receptor (116,164, 215). This mechanism of activation for the insulin receptor could therefore,
represent a common mechanism for the activation of RTKs in response to ligand.
I.iv. Signal Transduction
In response to ligand binding, RTKs become autophosphorylated on
specific tyrosine residues outside the catalytic domain of the protein, in addition to the tyrosine residues phosphorylated within the catalytic domain (Figure3).
These phosphotyrosine residues provide docking sites for cytoplasmic signal transduction molecules, which bind RTKs in a phosphotyrosine dependent manner via spedfic domains such as the Src Homology 2 (SH2)domain, the phosphotyrosine binding (PTB) domain and others that are not yet defined [Cbl
PTB domain - (14,118); Gab1 Met-binding domain (MBD)- (203)l.In addition to receptorsubstrate interactions, interaction between cytoplasmic substrates also occurs via a series of protein-protein interaction domains that generate a network
of signal transduction molecules. These protein-protein interactions are mediated through multiple domains, including the Src Homology 3 (SH3)domain and the WW domain that interact with proline rich stretches (15,28, 160, 187,213), and
the LIM domain (39,171) among others (reviewed in 147). Activated RTKs couple with many substrates and activate multiple
signalling pathways (Figure 5). Although each receptor assoaates with a subset
of signalling proteins, several pathways are frequently activated by many receptors, including those involving Grb2, phosphatidylinositol 3- kinase and
phospholipase Cy.
c-jun
Nucleus
I
Figure 5: Complexes of SH2 domain containing substrates that bind activated receptors
Grb2 is an adaptor protein containing only SH3-SH2-SH3 domains (94) that lacks
enzymatic activity (148),which assodates with phosphorylated RTKs
(94) via its S H 2 domain. Through its SH3 domains, Grb2 binds the Sos protein,
which is a Ras guanine nudeotide releasing factor (94). Sos releases GDP from
Ras allowing GTP to bind, generating an active Ras protein (94).Ras-GTP binds Raf (133, 196, 199,202, 214), a serine/threonine kinase that activates the MAP
kinase pathway (43,72,109)which is involved m cell growth regulation, Several other cytoplasmic signalling proteins interact with proteins involved in this
signalling pathway. Phosphorylated Shc adaptor protein and Syp/SHPTP2 phosphatase for example, also bind Grb2 SH2 domains (reviewed in 147). Activated receptors phophorylate Shc (18,38,59,105,153,154,158,159,175,179, 185, 194,212),which in turn binds activated receptors through its PTB and/or
S H 2 domains (93,105,197). Similarly, Syp/SHPTP2 has been shown to bind for example, the activated platelet derived growth factor receptor (PDGFR) which results in tyrosine phosphorylation of Syp/SHPTP2, followed by an increase in its catalytic activity (198). Thus, upon receptor activation, Grb2 could associate with the receptor indirectly via Shc and/or Syp/SHPTP2 proteins. The Ras
pathway therefore, may also be activated by Shc-Grb2 and Syp/SHPTP2-Grb2
complexes (reviewed in 147). Ras activity is in turn downregulated by GTPaseactivating proteins (GAPS) which increase conversion of active Ras-GTP to inactive Ras-GDP ( 190). One Ras specific GAP is pl2O-rasGAP (rasGAP). Activated protein tyrosine kinases phosphorylate rasGAP which binds autophosphorylated receptors via its S H 2 domains (3,48,87,96,130). The Grb2 activated Ras pathway therefore involves the interaction of several different
cytoplasmic signalling molecules.
Phosphatidylinositol4,5=bisphosphateis hydrolysed by phospholipase Cy (PLCy) to produce inositol trisphosphate and diacylglycerol, which are second
messengers (8,100) that increase intracellular calcium and activate protein kinase C, respectively (reviewed in 1).In contrast, phosphatidylinositol3-kinase (PI3K)
consists of an 85kDa regulatory protein and a llOkDa protein with catalytic activity (26). P I X phosphorylates phosphatidylinositol, phosphatidylinositol4
phosphate or phosphatidylinosito14,5bisphosphate on the D-3 position of the
inositol ring (reviewed in 146) generating second messengers (78). Several serine/threonine and tyrosine kinases and cytoskeletal proteins are the targets of the products of
PBK (reviewed in 2 5 ) Furthermore, one PI3K product,
phosphatidylinositoI-%phosphate is involved in vesicle trafficking (54). A second PI3K product phophatidylinositol-3,4-bisphosphate activates the Akt serine/threonine protein kinase (55,102). Moreover, binding of Ras to PI3K also
correlates with activation of this protein, leading in turn to Akt activation (reviewed in 54). Glycogen synthase kinase-3 (GSK3)is the only known target of Akt and Akt phosphorylation of GSK3 indicates that PI3K and Akt may be
involved in insulin-dependent glycogen synthesis (37).The Akt protooncoprotein may also be involved in regulation of 70K 56 kinase (22) which is involved in
serum and growth factor induced progression through G l in several types of cells (29, 108, 111, 157). PI3K-dependent regulation of cell survival is also
mediated by Akt (45,91,99, 106). Membrane ruffling stimulated by PDGF and insulin requires PI3K (142, 205). Furthermore, growth factor stimulated membrane ruffling requires Rac (142, 161), a small GTP-bindingprotein whose activation involves PI3K (69). PI3K also activates JunN-terminal kinase (JNK) in a Ras dependent manner (103).JNK in turn stimulates c-jun and c-fos expression.
PI3K may also induce JNK activation via Rac. Rac is thought to be involved in
JNKactivation (34,126). PDK and Ras are therefore involved in the activation of several interconnecting signahng pathways in the cell.
11. MET RECEPTOR TYROSINE KINASE
The prototype for one of the families of RTKs (Figure 2) is the Met receptor. The MET gene was first discovered as an oncogdc variant (11,33)and is located on human chromosome band 7q31 (40). The ME T protooncogene
encodes a receptor tyrosine kinase (145) for hepatocyte growth factorhcatter factor (HGF/SF) (19,140, 141). The Met receptor and other members of this subclass of R m , the Ron and Sea receptors (5,76,81,165) are synthesised as a
single chain molecule that is glycosylated then cleaved (57)to form a mature
P chain that are connected by disulfide bridges (5,56,57, 77,82, 163) (Figure 6). The Met receptor B chain spans the
heterodimer consisting of an a and a
membrane and the a chain is an extracellular subunit (58, 145). The tyrosine
kinase domain is found in the fbchain(61).
cd: extracellular domain
I
cytoplasm
185 kDa 190 kDa Figure 6:Generation of the mature Met receptor tyrosine kinase
170 kDa
n.i.Hepatocyte Growth FactorlSratter Factor Structure
HGF/SF is encoded by a single gene located on human chromosome
bands 7ql1.2-21(204).The protein is produced as a 728 amino acid single chain molecule (pro-HGF/SF) (139)which is inactive and is predominantly associated
with the extracellular matrix or cell surface (139). An active heterodimer is generated by proteolytic deavage in the extracellular environment, in some instances by urokinase (139). This conversion may also be mediated by the
11
thrombin-activated Factor W-like serum protease (128,139,177)(Figure 7). The active heterodimer consists of an a chain derived from the N terminus, disulfide
linked (204) with a
P chain (16).The a chain of HGF/SF contains four kringle
domains (137).The first kringle module contains the smallest region required for binding of the cytokine to the Met receptor tyrosine kinase (115). The P chain of
HGF/SF exhibits homologies to the plasminogen serine protease domain (204). The hepatocyte growth factor-like protein (HGFL) or macrophage stimulating protein (MSP)which is structurally related to HGF/SF, is the ligand for the Ron
receptor, which is a member of the Met receptor subfamily (42,56,66,176,178, 201). The ligand for the Sea receptor is unknown. Thus HGF/SF and MSP have
defined a new family of growth factors that are related to the plasminogen serine
protease. kringle domains
Factor XI][ like serum protease Pro-HGF/ SF
Active HGF/SF
Figure 7:Conversion of pro-HGF/SF to an active heterodimer Expression and Function
Hepatocyte growth factor/scatter factor is a multifuntiond cytokine that was originally found in the plasma of rats following liver injury and was shown to stimulate the proliferation of primauy hepatocytes in culture (138).HGF was later found to be identical to scattex factor (204),a factor seaeted by fibroblasts that
induced the dissociation and motility of epithelial cells in culture (186). Subsequently, HGF / SF was demonstrated to be a potent rnorphogenic factor stimulating branching tubulogenesis of kidney epithelial cells in culture (131,
172) and duct formation in mammary epithelial cells (191).
II.ii. Met Receptor and HGFiSF Funtion In Vivo
Mutant mice generated with the targeted disruption of the met or hg$/sf gene exhibit identical phenotypes. Mouse embryos lacking the Met receptor or
HGF/SF exhibit impaired development of the liver and placenta (9, 173, 192). These abnormalities lead to death after day 12.5 of embryogenesis (9,173, 192).
Mouse embryos lacking the Met receptor also demonstrated that myogenic precursor cell migration into the limb adage, diaphragm and tip of the tongue
during development requires expression of the Met receptor (9). The Met receptor and HGF/SF have also been implicated in the development of the embryonic nervous system. Developing motoneurons transiently express the
MET
gene (183)and HGF/SF is a chernoattractant which can direct growth of
axons of rnotoneurons and later in development, act as a survival factor for these neurons (46). Furthermore, while HGF/SF is expressed predominantly in mesenchymal tissues, the Met receptor is expressed by epithelial tissues. During mouse development, epithelial cells of many organs express Met mRNA while proximal mesenchymal cells express HGF/SF mRNA (183). This pattern of Met receptor
expression is maintained in adult mouse epithelial tissues (210).
These observations of HGF/SF or Met receptor deficient mice and comparison of the expression patterns of these proteins indicate that the Met receptor and HGF/SF form a receptor-figandpair and that they play a significant
role during embryogenesis.
Effects on Cell Proliferation
The Met receptor and HGF/SF are known to exert a variety of effects on many different cell types. HGF/SF is implicated in liver regeneration (123, 136, 168), where the main sites of synthesis of HGF/SF are sinusoidal endothelium and Kupffer cells (143). Liver parenchymal and biliary epithelial cells are stimulated to proliferate by HGF/SF (84). Furthermore, it has been found that
after liver resection, the addition of exogenous HGF/SF results in the binding of this factor to the remaining hepatocytes and increased rate of liver repair (16).A model for HGF/SF induced liver regeneration suggests that when toxins or resectioning of the liver damage this organ, it is unable to dear HGF/SF. This
results in the accumulation of the cytokine in the circulation,which at a high
enough concentration is able to induce liver regeneration (124). Similarly, HGF/SF is also thought to act as a renotropic factor. It is
increased in the circulation following renal injury (16), and renal tubular cell DNA synthesis and tubule formation of epithelial cells is stimulated by HGF/SF
(16)-
Effects on Motility In addition to stimulating proliferation of several types of endothelial and
epithelial cells, HGF/SF also induces scatter of certain epithelial cells in d t u r e (186).These indude Madin-Darby canine kidney (MDU() epithelial cells (204) and malignant human cells (16)such as Hs 7661 human pancreatic carcinoma cells and A549 human l a g carcinoma cells (204).
Effects on Morphogenesis
The Met receptor and HGF/SF are also implicated in epithelial and endothelial cell morphogenesis. HGF/SF has been shown to induce branching
tubulogenesis of kidney cells (131, 132). Epithelial cell lines grown in 3dimensional collagen gels are stimulated to undergo morphogenesis in the presence of HGF/SF (21). These tissues organise themselves into structural
elements resembling those formed by the epithelial cells of the organ of origin. For example, hollow spheroids are formed by colon cells, while pancreas adenocardnoma cells form cysts, mammary gland cells and human prostate cells
form ductular structures, and lung carcinoma cells form alveolar-like structures.
These observations suggest that HGF/SF induces epithelial cells to undergo morphogenesis following unique intrinsic programs.
Effects on Tumor Progression
Evidence suggests that HGF/SF also plays a significant role in turnour progression (204).HGF/SF is expressed in cancer cells and is able to promote the proliferation, motility and invasion of epithelial cells and carcinoma cells (204).
Since the conversion to a more fibroblastoid phenotype of epithelial cells is mediated by HGF/SF, this cytokine is thought to be important in uivo to the
increased malignancy of carcinomas (204).HGF/SF has also been shown to be a potent angiogenic factor (23,135). In addition, degradation of the extracellular
matrix is induced by HGF/SF by stimulating production of extracellular matrix proteolytic enzymes (12,155),and thus may act to aid the metastatic process.
These observations are consistent with a role for the Met-HGF/SF receptor-ligand pair in several physiological processes including embryonic development and organ remodelling, and in pathophysiological processes such as tumor progression and invasion.
III. ONCOGENIC ACTIVATION OF RECEPTOR TYROSINE KINASES
II1.i. Overexpression of RTKs in Human Cancers
Receptor tyrosine kinases (Rm) regulate cell growth and differentiation, and deregulated forms of these proteins have been implicated in human cancer (149) (Table 1).The expression of the epidermal growth factor receptor (EGFR)
and the pl8Peu receptor is commonly deregulated. EGFR is overexpressed in
breast cancer (50,101,169),and in squamous cell carcinomas and glioblastomas
the EGFR gene is frequently amplified and/or overexpressed (114,209). The c-
erbB2 gene which encodes the p18Peu receptor is overexpressed in ovarian,
gastric and breast cardnomas (98),and lung (98), pancreatic (65)and endometrial (7) adenocarcinomas. Other receptors have also been linked to tumour
progression. Insulin-like growth factor 1 receptor (IGF-IR) is overexpressed in primary cervical tumour cells (184) and breast cancer (150).
The Met RTK has also been linked to the progression of cancer (Table 1).
This receptor was first isolated as the Tpr-Met oncogene from an N-methyl-N'nitro-N-nitrosoguanidine-treatedhuman osteogenic sarcoma cell line (MNNGHOS) (32, 33). In addition, colon, pancreatic, ovarian and papillary thyroid
carcinomas,and osteogenic sarcomas overexpress the MET gene (44,174).Several other types of cancer, such as leukaemia and lymphoma (85) also exhibit MET
gene amplification. These observations and its biolocal activites suggest an important role for the Met receptor and oncoprotein in tumor progression.
Table 1: Receptor Tyrosine Kinase Expression in Human Cancer -
..
Receptor Tyrosine Kinase Receptor
-4lteration
epidermal growth factor receptor
overexpression
--
.-
-
Type of Cancer breast cancer
gene amplification squamous cell carcinoma and/or overexpression glioblastoma
pl8,'neu receptor
gene overexpression
ovarian carcinom gastric carcinoma breast carcinoma lung adenocarcinoma panaeatic adenocardnoma endometrial adenocardnoma
insulin-like growth factor 1 receptor
overexpression
primary cervical tumour breast cancer
Ret receptor
point mutations
mutiple endocrine neopkia 2A, 2B
Met receptor
missense mutations
papillary renal cardnom
gene overexpression
papillary thyroid Carcinom colon carcinoma pancreatic carcinoma ovarian carcinoma osteogenic sarcoma leukemia
III.ii. Activating Mutations of Receptor Tyrosine Kinases
In addition to receptor overexpression, oncogenic variants of many RTKs have been isolated from human tumors. These oncoproteins are rendered
constitutively active as a result of mutation. In glid tumors of the central nervous system, deletion of exons 2-7in the extracellular domain of EGFR often occurs
(47,119, 188, 208) generating a constitutively activated receptor (73).The Ret receptor is constitutively activated by mutations involving cysteine residues in
the extracellular domain, in multiple endocrine neoplasia (MEN) 2A. Activation
is thought to be the result of constitutive dirnerisation via disulfide bonds formed
by unpaired cysteine residues (78). In contrast, in MEN2B, the Ret receptor is activated by a point mutation within the catalytic domain that is thought to change its substrate specificity (78). Similarly, multiple missense mutations in the
tyrosine kinase domain of the Met receptor, found in inherited and sporadic
papillary renal carcinoma (174),lead to oncogenic activation of the receptor (83).
IILiii. Oncogenic Activation by Gene Rearrangement A third general mechanism for oncogenic activation of RTK derived
oncoproteins in human tumors involves gene rearrangement (162),where the cytoplasmic b a s e domain of the receptor is fused downstream of unrelated
sequences (70). The oncogenic varient of the Met receptor, Tpr-Met, is a prototype for this class of oncoproteins. These indude Tpr-Trk found in human
thyroid papillary turnours, where the Trklnenre growth factor receptor tyrosine b a s e domain is fused downstream of Tpr sequences (62). The Ret receptor tyrosine kinase domain fused to the H4 gene generates the PTC oncogene,which is expressed in approximately 2590 of thyroid carcinomas (13).
n1.i~. Tpr-Met Oncogene
The Tpr-Met oncogene is generated following a chromosomal
rearrangement that fuses the Met receptor kinase domain from chromosome 7 downstream of the translocated promoter region ('Tpr)derived from chromosome 1 (41,144)(Figure 8). This rearrangement results in the loss of the extracellular
and transmembrane domains, as well as part of the intracellular domain of the Met receptor (145).Initiation of transcription and translation now occurs within the TPR sequences (27), generating a 65kDa protein (61)in which the kinase
domain of the Met receptor is fused to the 142 amino adds contributed by Tps (27). Unlike the Met receptor, the Tpr-Met oncoprotein is cytosolic and exhibits constitutive tyrosine phosphorylation (61)in the absence of ligand.
I
m
.
I
p190
Met Receptor
m
extracellular domain 1
tPr
cytoplasmic domain
met
p65 Tpr-Met Oncoprotein
Figure 8: Met receptor and oncoprotein 1II.v. Mechanism of Activation of Tpr-Met
The Tpr sequence encodes two leucine zipper-like motifs (162).The leucine zipper domain is a motif i n which every seventh amino acid is a leucine
residue (110) that together form a 4,3 hydrophobic repeat (2). The 4,3
hydrophobic repeat contributes to a hydrophobic interface in coiled coil motifs
(30,36)which mediate protein-protein interactions. Leucine zipper A in Tpr-Met fits the coiled coil consensus motif more
dosdy than leucine zipper B (162),and it has been shown that leucine zipper A and not leucine zipper B, is essential for cell transformation by Tpr-Met (162).
This finding and the fact that Tpr-Met dimers have been detected in vim ((162) suggest that the leudne zipper motifs of the Tpr region mediate oncogenic
activation of Tpr-Met by inducing constitutive dimerisation, activation of the
catalytic domain and subsequent tyrosine phosphorylation in the absense of
ligand (162)(Figure 9). leucine
Tpr
n
Z ~ P P ~ ~ A
wild type leucine zipper Tm-Met A mutant B mutant
leucine zipper B
Phosphorylation Met tyrosine kinase domain
Transformation
Figure 9: Constitutive dimerisation mediated by a leucine zipper motif in Tpr leads to phosphorylation of Tpr-Met and subsequent activation of its oncogenic POtentid
In a similar manner to Tpr-Met, other RTK derived oncoproteins generated by genomic rearrangementsincluding Tpr-Trk,tropomyosin-Txk, PTC
and Ret II contain coiled coils (62,63,79,80,120,189).The oncogenic activation of
these receptor tyrosine kinases therefore, may occur in the same manner as predicted for Tpr-Met - by constitutive dimerisation via coiled coil motifs (162).
In other oncoproteins generated by rearragements, such as Tel-PDGFpR and Rfg Ret, Tel and Rfp provide protein interaction motifs which are thought to mediate
constitutive dimerisation and activation of these proteins (60, 78). AU of these rearranged oncoproteins, including Tpr-Met, lack point mutations and therefore,
are thought to represent constitutivdy activated, but in aI3 cases cytosolic forms of their receptor counterparts.
IV. MET RECEPTOR AND ONCOPROTEIN: ACTIVATION AND
SUBSTRATE BINDING IVJ. Autophosphorylation of Met Receptor and Tpr-Met Oncoprotein
Ligand dependent dimerisation of tyrosine kinase receptors or ligand independent dimerisation of their oncogenic variants leads to the stimulation of their intrinsic tyrosine kinase activity. As discussed in section I.iii., autophosphorylation of tyrosine(s) in the A loop is essential for biological function of many RTKs, including the insulin receptor (49,166)
I
L
Tpr-Met
I/
I ; I
!
I
domain cytoplasmic
J
Met Receptor
Figure 10: Tyrosine phosphorylation sites in Tpr-Met and the Met receptor that are essential for catalytic and biological activity
The Met receptor contains three tyrosine residues in its kinase domain that are homologous to the major site of phosphorylation in the insulin receptor.
From two dimensional phosphopeptide mapping analysis, two of these tyrosines
(tyrosine365 and tyrosine 366) were identified as the major autophosphorylation sites in Tpr-Met (164) (Figure 10). Mutational analysis of these two residues
showed that they are essential for autophosphoryiation and exogenous substrate phosphorylation activity of this oncoprotein, and its transformation potential (164).The Met receptor contains two tyrosine residues, tyrosines 1234 and 1235,
which are equivalent to tyrosine 365 and 366 in Tpr-Met (164)(Figure 10) and which are similarly required for Met receptor catalytic and biological activity (215).
IV.ii. Interaction of Met Receptor with Substrates
Unlike other RTKs, such as the PDGFR, where distinct phosphotyrosine residues interact with distinct substrates (Figure Il), the biological activity and substrate association of Met receptor are localised to two tyrosine residues. Similarly, the U f o / M RTK has recently been found to contain a single tyrosine,
Y821,in its carboxy terminus that is a major binding site for several substrates including PLCy, p85 proteins, Grb2, Src and Lck (20). The Met receptor may thus provide a prototype for a class of RTKs which mediate substrate association via
one or two major phosphotyrosine residues.
The Met receptor contains 16 tyrosine residues in its cytoplasmic domain, three of which are found in the carboxy terminus. A single tyrosine residue, tyrosine 1356, in its carboxy terminus is required for the stimulation of motility, invasion and morphoge~cresponses by this receptor (215). A second carboxy terminus tyrosine residue, tyrosine 1349, may be involved in stabilising stubstrate
interactions with the receptor. Assodation of various cytoplasmic signalling molecules with the Met receptor has been demonstrated. The adaptor protein Grb2 binds the Met receptor uniquely through tyrosine 1356 (215),while the p85 subunit of PIX,
PLCy and pp60C-SrC require phosphorylation of tyrosine 1319, in addition to tyrosine 1356 (156) for maximal binding. Tyrosines 1336 and 1349 are also
essential for Shc phosphorylation and/or binding to the Met receptor (53,151). The identity of the substrates which bind the activated receptor
determines the downstream pathways that are shu1ated and the subsequent
effects exerted on the cell. The elucidation of the contriiution of each substrate to
Met receptor mediated morphogenesis, motility and mitogenesis will provide insight into RTK signal transduction.
Figure 11: Substrate binding sites in the PDGFRB and Tpr-Met
IV.iii. Interaction of Tpr-Met with Substrates
Similarly, in the Tpr-Met oncoprotein, a single carboxy terminus tyrosine
residue, twosine 189, is highly phosphorylated and has been shown to be required for the biological activity of Tpr-Met (86)(Figure 11). This residue is
equivalent to tyrosine 1356 in the Met receptor and in a similar manner to the receptor, Grb2 binds directly to tyrosine 489 (86). Tyrosine 489 is also critical for
association of Tpr-Met with Syp/SHPTP2, PLCy (51) and the p85 subunit of P I X (52).
1V.i~.Domains Mediating Protein-Protein Interactions
As stated, cytoplasmic signailing proteins bind activated RTJSs in a
phosphotyrosine dependent manner. This interaction is mediated by specific modules within the substrate molecules, the most important of which is the Src Homolgy 2 Domain.
1V.v. Src Homology 2 Domains
The Src homolgy 2 (SH2)domain is a module of 100 amino acids (104) that binds phosphotyrosine residues (121). The SH2 domain is found in a variety of cytoplasmic signal transduction proteins. These include those proteins with enzymatic activity (Src, PLCy, GAP), as well as adaptor proteins (p85 subunit of
PUK, SHC,Grb2) (148).The basic structure of the S H 2 domain consists of a large antiparallel P sheet flanked by two a helices in the order ka-&bk&&a-P (31). The SH2 domain folds such that the amino and carboxy termini of the
module meet, leaving the opposite surface exposed for binding the phosphotyrosme residue (148).
IV-vi.SH2 Domain Binding Specificity Several SH2 domain containing substrates such as PLC1/ (107,200), rasGAP (87, 96) and PDK (35,88,95,206) bind the active PDGFR (97). Studies
involving mutation of tyrosine residues in the cytoplasmic domain of the PDGFR show that rasGAP binds Y771, while PI3K binds Y751 and Y740 (97). These observations suggested that the SH2 domain of cytoplasmic signding proteins
reco,pised specific sequences around the phosphotyrosine in the receptor.
Comparison of in oivo binding sites of SH2 domain containing substrates indicated that the specificity of association was determined by amino adds
immediately downstream of the tyrosine residue (182). Songyang and colleagues used a degenerate phosphopeptide library, where the +I, +2 and +3 positions downstream of a phosphotyrosine residue were varied, to determine the optimal binding sequence for the SH2 domains of several proteins (181,182) (Table 2).
The results of these experiments confirmed that the amino adds in the +I, +2, and +3 positions, immediately downstream of the phophotyrosine residue
determine the specifiaty of the S H 2 domain binding reaction (reviewed in 31). Table 2: Optimal Binding Sequences for Various SH2 domains
Grb2 p85a N p85a C PLCyI N PLCyI C SypiSHPTP2 N Shc Src
(Boldface letters indicate strong selection; uppercase nonboldhce letters indicate medium selection; lowercase letters indicate weak selection; x indicates no selection*)
The specificity exhibited by the SH2 domains studied can be explained in
terms of their structure (Figure 12). The SH2 domain has a deep pocket into which the phosphotyrosine side chain fits (31).At the bottom of this pocket is an invariant arginine residue (148) which contacts two phosphate oxygens of the
phosphotyrosine residue through hydrogen bonds (147). The side chains of phosphoserine and phosphothreonine are not long enough to reach this arginine residue (118).Hydrogen bonds and hydrophobic bonds to the phosphate and the
phenol ring respectively, are formed in the phosphotyrosine binding pocket (31). The sequence and structure of each SH2 domain varies such that it confers
speaficity of binding. For example, the optimal binding motif for the Src S H 2 domain has been shown to be phosphotyrosine-glutamate-glutamate-isoIeucine. The three amino acids in positions 4, +2 and +3 downstream of the phosphotyrosine residue form contacts with the Src S H 2 domain (147)(Figure 12A). The isoleucine residue fits into a hydrophobic pocket (31). This observation
is consistent with the finding that Src SH2 domain has greatest selectivity at the +3 position following the phosphotyrosine residue.
Figure 12:A) Src SH2 domain structure; B) Syp and PLCy SH2 domain structure C)Suggested Grb2 SH2 domain structure
Studies of SH2 structure can also explain the binding specificities of other SH2 domains. The N terminal Syp/SHPTP2 phosphatase SH2 domain preferentially binds an Ile/Val-X-Val/Ile motif and the C terminal and N
terminal SH2 domains of PLCyl binds Val/Ile-Ile/Leu-Pro/fle /Val, and Leu/Ile/Val-Glu/ Asp-Leu/ Ile /Val motifs respectively. The SH2 domain structures of Syp and P L Q I show that the domain makes contact with residues
up to the +5 and +6 position following the phosphotvrosine, respectivelv (31) (Figure 12B). The ligand-binding surface of these SH2 domains contain a hydrophobic groove which accommodates these predominantly hydrophobic
amino adds (147). While most SH2 domains have a requirement for specific amino add
residues at positions +1 and +3 downstream of the phosphotyrosine, the binding specificity of the Grb2 S H 2 domain is based on the presence of an asparagine residue at the +2 position. It has been suggested that a pocket for the asparagine
residue at the +2 position downstream of the phosphotyrosine may exist in the
S H 2 domain of Grb2, accounting for the high selectivity for this amino acid (182) (Figure12C).
IV.vii. Phosphotyrosine-BindingDomain
In contrast to SH2 domains whose specificity depends on amino add residues downstream of the phosphotyrosine, the phosphotyrosine-binding
(PTB)domain has specifiaty for amino acids preceeding the phophotyrosine residue. The PTB domain was first identified in the N tenninal region of the Shc protein (93) and the Shc PTB domain has been shown to bind an asparagine-x-x-
phosphotyrosine motif (92)-The insulin receptor substrate4 -1)
has also been
shown to possess a PTB domain (207). Interaction of the Shc PTB domain with the insulin receptor (64) and EGFR in uitro (10, 197) has been demonstrated.
Furthermore, Shc and IRSl interaction with the insulin receptor has been shown to be dependent on the asparagine, prohe and phosphotyrosine residues within an asparape-prohe-glutamate-tyrosinemotif in the receptor (64).
1V.viii. Met Receptor and Oncoprotein Consensus Substrate Binding Sites All known substrates of the Met receptor and Tpr-Met oncoprotein bind
the receptor and oncoprotein via one or two tyrosine residues in the carboxy terminal tail. The amino acid sequence following tyrosines 1349 and 1356 or tyrosines 482 and 489 in the receptor or oncoprotein respectively, provide a
consensus binding site for the SH.2 domains of their substrates (Table 3).
Table 3: Y1349 and Y1356 of the Met Receptor Tyrosine Kinase and Y482 and Y489 of the Tpr-Met Oncoprotein Form a Degenerate Consensus Binding Site for Various Sinnal ~r&ciuction ~olecules wild type Met Receptor wild tme Tpr-Met
wild type Met Receptor wild type Tpr-Met
Shc
(Boldfacelettezs indicate strong selection;uppercase nonboldface letters indicate medium selection; lowercase letters indicate weak selection; X indicates no selection.)
Based on the observations made by Songyang et. al. (181,182), most of the
substrates binding the Met receptor and oncoprotein require a small hydrophobic amino acid residue,particularly valine, in the +land +3 positions downstream of the phosphotyrosine, for binding. Both Y1349 and Y1356 in the Met receptor and
Y482 and Y489 in Tpr-Met are followed by valine residues in positions +1 and +3, conforming to the binding specificity requirements of these substrates. The GrbZ
adaptor protein binds exclusively to Y1356 in the Met receptor and Y489 in the Tpr-Met oncoprotein. This is in agreement with the requirement for Grb2 binding
of an asparagine residue in the +2 position downstream of a phosphotyrosine.
Thus Y1349 and Y1356 in the Met receptor and Y482 and Y489 in the Met oncoprotein constitute a multisubstrate binding site.
V. SUBSTRATES INVOLVED IN SIGNALLING BY THE MET RECEPTOR AND ONCOPROTEIN
The Met receptor is implicated in normal growth and development, as well as in the progression of human cancer. The contribution of independent
substrates to the normal biological activities of the Met receptor or to Tpr-Met oncoprotein mediated transformation is unknown. Thus a study aimed at dissecting the signalling pathways downstream from the Met receptor and
oncoprotein is aitical to our understanding of how the Met receptor and oncoprotein mediate their biological activities. Several cytoplasmic signalling molecules bind the Met receptor and TprMet oncoprotein, including GrbZ, the p85 subunit of PDK, PLQ
and
Syp/mPTP2. To study the signalling pathways downstream of these substrates, a sitedirected mutagenesis strategy was carried out to m a t e Met receptor and
Tpr-Met mutant proteins.
V.i Substrates Involved in Met Receptor Mediated Biological Activity
To elucidate the contribution of cytoplasmic signal transduction molecules to Met receptor mediated motility, mitogenesis and morphogenesis, a chimeric
receptor (CSF-MET) was generated in which the extracellular domain of colony stimulating factor-1 (CSF-1)receptor was fused to the transmembrane and
intracellular domain of human Met receptor (216).Mutation of Y1356 of the Met receptor to a phenylalanine residue abolished the ability of the CSF-METreceptor to induce the scatter of Madin-Darby canine kidney epithelial cells (MMJO and
to form branching tubules in response to CSF-l(215).This mutant receptor failed to bind multiple substrates including Grb2, p85, PLCy and Shc adaptor protein (53), implicating one or more of these proteins in these events.
The +2 asparaghe residue is essential for Grb2 SH2 domain binding (182), therefore a mutant where the asparaghe residue was substituted by a histidine residue generated a chimeric CSF-MET receptor that does not bind Grb2, but retains interaction with PLCy, Shc and the p85 subunit of PI3K (53). This mutant
receptor was able to induce scatter of MDCK cells in response to CSF-1 but did
not induce efficient branching tubule formation (53). Therefore, GrbZ or pathways downstream of this adaptor protein are essential for branching
morphogenesis,but not cell scatter.
V.ii. Substrates Involved in Tpr-Met Mediated Transforrmtion
Similar mutagenesis experiments were carried out with the Tpr-Met
oncoprotein to determine the contribution of Tpr-Met substrates to cell
transformation mediated by this oncoprotein.
A Tpr-Met mutant protein (Y489F Tpr-Met) was generated in which the major phosphorylation site outside the catalytic domain (Y489) was altered to a
phenylalanine residue. This mutant protein transforms Fr3T3 fibroblasts at 20%
the effiaency of wild-type Tpr-Met (52),and has lost the ability to bind Grb2 (52)
and PLCy (51). In addition, its ability to bind PDK and stimulate its activity is impaired (52) (Table 4). These observations suggest that the signal transduction
pathways downstream of one or more of these substrates is essential for 80% of the transformation efficiency of Tpr-Met.
To analyse the contribution of Grb2 to Tpr-Met h c t i o n t the asparaghe residue in position +2 downstream of Y489 which is required for GrbZ binding, was altered to a histidine residue. The transformation efficiency of this mutant
(N491HTpr-Met) is reduced to 20% of that of wild type Tpr-Met (51), and as expected, does not bind the Grb2 adaptor protein (51), but does maintain the ability to bind PLCy and Syp/SHPTPZ, and activate PI3K (51) (Table 4). However, the Shc adaptor protein is phosphorylated and binds Grb2 in N491H
Tpr-Met mutant transformed cell lines (51) (Table4) and may account for the low 20% hamformation activity of this mutant protein. Significantly, mutation of
Y482 and Y489 to phenylalanine residues (Y482WY489F Tpr-Met) abolishes Shc
binding and cell transformation (51,86,156)(Table4).These results indicate that
Grb2 binding to Tpr-Met is essential for Tpr-Met mediated transformation,but do not allow us to determine if it is suffiaent.
Table 4: Transformation Potential and Substrate Binding Ability of Tpr-Met Mutants
I
-- -
-
Tpr-Met Mutant Transformation Efficienw wild type
Substrate Binding Ability Grb2 PI3K PLCy
syp/sm
She phosphorylation and assotiation with Grb2
Grb2 PI3K PLCy syp/sKPTP2
Shc phosphorylation and association with Grb2
+ + + +
+
--
0
Grb2 PIX PLCy syp/s= She phosphorylation and association with Grb2
+ +
PI3K + PLCy + syp/sHPTP2 + Shcphosphorylationand + association with Grb2
VI. Conclusion RTKs have been implicated in human cardnogenesis. Study of the signal transduction pathways activated by these proteins will provide essential insight
into turnmigenesis mediated by deregulated RTKs and their oncogenic variants.
Dissection of the role of signalling pathways regulated by the Met RTK is hampered by the fact that all substrates bind to a single multisubstrate binding
site. Unlike the Met receptor, the majority of RTKs contain multiple phosphorylated tyrosine residues which are involved in association with unique substrates. Thus, mutation of single tyrosine residues allows a study of the consequences of the signalling pathway on the biological activity of that receptor.
The goal of this thesis is to devise a strategy by which we can evaluate the sigruficance of independent signalling pathways in the pleiotropic biological
responses regulated by the Met receptor and oncoprotein. To determine if GrbZ is sufficient for Tpr-Met mediated transformation or branching morphogenesis
mediated by the Met receptor, and to establish the role of PI3K, PLCy,
Syp/SHPTP2and Shc in signal transduction by the Met receptor and Tpr-Met oncoprotein,the generation of Tpr-Met and Met receptor mutants with the ability to bind specific substrates will be invaluable.
MATERIALS AND METHODS Plasmids
The full length cDNA of the Tpr-Met oncogene with the tyrosine 482 of Tpr-Met converted to phenylalanine was cloned into pXM mammalian expression vector (Y482F Tpr-Met - kindly provided by Dr. Elizabeth Fixman,
Darren Kmikura). For the purposes of mutation, the Spe I fragment of Tpr-Met was gel isolated using Elu-Quick DNA Purification Kit (Schleicher and Scheull) after digestion with Spe I restriction enzyme at 370C for 3.5 hours. The 1.3kb Spe
I fragment was cloned into the pBSKSII+ vector (Stratagene)linearised at the Spe
I site, dephosphorylated using calf intestinal alkaline phosphatase (UP) and gel isolated (vector kindly provided by Dr.Elizabeth Fixman). Ligations were camed
out at 120C overnight using DNA concentrations of 5ng/uL of vector and 10ng/uL of insert and molar ratios of 1:4 of vector to insert, 400 units of T4 DNA ligase (NEB), 10X Buffer (NEB) at 1X concentration, and 0.33mM ATP in a total
volume of15uL.40uL of electrocompetent DH5 cells, were transformed with 2uL of the ligation mix by electroporation (4). Cells were shaken at 370C for 45 minutes after the addition of 500uL LB. Aliquots were plated on LBampicillin (50 or 100ug/mL) plates and incubated at 370C overnight. Individual colonies were
grown overnight in 5mL LB and SOug/mL of ampicillin at 370C.DNA was
extracted from these cultures (170). Large scale DNA preparations of clones
containing the Spe I fragment in the correct orientation in the pBSKSII+ vector, as determined by Sca I digest, were prepared using cesium chloride density gradients (170). Clones predicted by restriction enzyme analysis to contain the
Y482F Tpr-Met Spe I fragment in pBSKSEI+ in the correct orientation were sequenced using Sequenase Version 2.0 DNA Sequencing Kit (USB Amersham
Life Science).Oligonudeotideprimers homologous to the T7 and T3 binding sites
within the vector ~ ~ ' A A T A C G A C T C A C T A T AT~~~ :' ;~ ' A T T A A C C ~ C A ~ A A A G ~
and
a
primer
specific
for
bp1539-bp1559
in
Tpr-Met
(5'GACCCCTTATTCGAAGTAATG3')were used in sequence analysis to confirm the orientation of the Spe I fragment in the pBSKSII+ vector and the presence of
the Y482F mutation respectively (Figure 13).
Tpr-Met Spe I Fragment
Figure
cloning of Spe I fapent of Tpr-Met into pBSKSII+ vector
Site-directed Mutagenesis
The three amino adds in positions +1, +2, and +3 downstream of tyrosine 189 of Tpr-Met, VNV, were converted to FNF or FNE in both wild type (WT)Tpr-
Met and the Y482F Tpr-Met mutant using the Chameleon Double-Stranded, Site-
Directed Mutagenesis Kit (Stratagene) (see Results - Figure 14).Briefly, 750ng of
WT or 700ng of the Y482F Spe I fragment of Tpr-Met in pBSKSn+ were used in the appropriate reactions. Two primers - a selection primer and a mutagenic
primer, were used per reaction. The Xmn I
selection primer:
SCATCATTGGAAAACGCTCTTCGGGGCG3*(Chameleon DoubleStranded, SiteDirected Mutagenesis Kit - Stratagene)was used to remove the Xmn I u n i p e restriction site in the vector. The following oligonucleotides containing the
appropriate nucleotide substitutions and a Hpa I site at position 1665 in Tpr-Met for identification of mutant plasmids, were synthesized:
35
FNF:5'P04-GCGACACATTTAAAGTTGAAATAAGTAGCGnMCATGGAC3'; FNE:5*PO4-GCGACACAmCGTTGAAATAAGTAGCGmMCATGGAC3'.
The Xmn I selection primer (225ng)and a rnutagenesis primer (-336ng) were annealed to WT and Y482F Spe I fragments of Tpr-Met in pBSKSIIc after denaturation of the plasmids in 1X mutagenesis buffer [IOmM Tris-acetate (pH 7.5), lOmm MgOAc, 50mM KOAc (pH 7.5)]. The reactions were boiled for 5 minutes, immediately placed on ice for 5 minutes, then incubated at room
temperature for 30 minutes. Annealed primers were extended with
deoxynucleotides and T7 DNA polymerse, and the newly formed strand was ligated using T4 DNA ligase. For selection of mutant plasmids, the DNA was
digested with Xmn I to linearise m u t a t e d plasmid strands and transformed into XL,MutS competent bacterial cells (Chameleon Double-Stranded, Site-
Directed Mutagenesis Kit). These cells are repair deficient and randomly copy
both mutated and m u t a t e d strands. An aliquot of the transformation mixture was grown overnight at 370C in 3mL of 2X YT and SOug/mL of ampidllin, and
DNA was prepared. To increase the efficiency of obtaining a mutant clone, approximately 300ng of this DNA was digested with Xmn I and 30ng was
transformed into Epicurian Coli XL1-Blue competent cells and cells were plated on LB-ampicillin plates. DNA was prepared from resulting colonies and putative mutants identified by digestion with Hpa I.
Large scale DNA preparations using cesium chloride density gradients were prepared from clones predicted by Hpa I restriction enzyme analysis to
have the appropriate mutations.The 1.3kb Spe I fragmentscontaining the desired
mutations were doned into WT Tpr-Met in the mammalian expression vector pXM (211).
Generation of Vectors for Oligonucleotide Cassette Mutagenesis
Unique Hpa I and Nar I restriction sites conserving the amino acid
sequence, were introduced into wild type Tpr-Met Spe I fragment in pBKSII+, at nucleotide 1665 and nudeotide 1695 of Tpr-Met respectively, by sequential site
directed mutagenesis as described above. The following oligonucleotides were
used: Hpa I: S1P04-CACATAAGTAGCGTTAACATGGACATAGTGCTCCCC; Nar I: 5'P04-CAGAGAAGGATACGGGGCGCCACATTITACGTTCAC. Putative
substrate specific binding sites were generated using oligonucleotide cassettes with Hpa I and Nar I restriction sites. Complementary sets of oligonucleotides
were generated, and kinased and annealed by incubating lug of the
oligonucleotides with lOmM ATP, 20 units of T4 Polynucleotide Kinase (NEB) and 1X Kinase buffer (NEB)at 370C for 60 minutes, then at 800C-900C for 5
minutes and slow cooling to
