MOLECULAR CHARACTERIZATION

OF THE ETV6-NTRK3 FUSION IN

CONGENITAL FIBROSARCOMA

by

DANIEL HON-HEI WAI

B.Sc. The University of British Columbia, 1994 B.M.L.Sc. The University of British Columbia, 1996

A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF

MASTER OF SCIENCE

in

THE FACULTY OF GRADUATE STUDIES

(Department of Pathology & Laboratory Medicine)

We accept this thesis as conforming to the required standard

THE UNIVERSITY OF BRITISH COLUMBIA

1999

© Daniel Hon-Hei Wai, 1999 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.

The University of British Columbia Vancouver, Canada

Date 11

ABSTRACT

The t(12;15)(pl3;q25) rearrangement detected in congenital fibrosarcoma splices the ETV6 (TEL) gene on chromosome 12pl3 in frame with the NTRK3

(TRKC) neurotrophin-3 receptor gene on chromosome 15q25. Resultant ETV6-

NTRK3 fusion transcripts encode the helix-loop-helix (HLH) dimerization domain of ETV6 fused to the protein tyrosine kinase (PTK) domain of NTRK3. We hypothesize that chimeric proteins mediate transformation by dysregulating NTRK3 signal transduction pathways via ligand-independent dimerization and PTK activation. To determine if the fusion protein has transforming activity, NIH3T3 cells were infected with recombinant retroviral vectors carrying the full-length

ETV6-NTRK3 cDNA. These cells exhibited a transformed phenotype and formed macroscopic colonies in soft agar. In order to characterize the roles of specific ETV6-

NTRK3 domains, we expressed a series of ETV6-NTRK3 mutants in NIH3T3 cells and assessed their transformation activities. Deletion of the ETV6 HLH domain resulted in morphologically non-transformed NIH3T3 cells that failed to grow in soft agar. Mutants of the three PTK activation-loop tyrosines (EN-Y513, EN-Y517, and EN-Y518) had variable PTK activity but had limited to absent transformation activity. The EN-Y513F mutant failed to transform NIH3T3 cells, while the expression of either EN-Y517F or EN-Y518F resulted in a semi-transformed phenotype. The simultaneous mutation of Y517 and Y518 (EN-Yx2F), or of all three tyrosines (EN-Yx3F), resulted in a morphologically untransformed phenotype. The Ill

ATP-binding mutant (EN-K380N) failed to autophosphorylate and completely lacked transformation activity. In addition, a series of PTK-active mutants unable to bind phospholipase-Cy did not show defects in transformation activity. These studies confirm that ETV6-NTRK3 is a transforming protein that requires both an intact dimerization domain and a functional PTK domain for transformation activity. IV

TABLE OF CONTENTS

Page

ABSTRACT ii

TABLE OF CONTENTS iv

LIST OF TABLES vii

LIST OF FIGURES viii

LIST OF ABBREVIATIONS ix

ACKNOWLEDGEMENTS xiii

Chapter 1 INTRODUCTION 1 1.1 General concepts in cancer genetics 1 1.1.1 Introduction 1 1.1.2 Tumour-suppressor genes 2 1.1.3 Proto-oncogenes 4 1.1.4 Fusion genes 6 1.2 Fibroblastic tumours 8 1.2.1 Fibromatoses versus fibrosarcomas 8 1.2.2 Congenital fibrosarcoma (CFS) 10 1.2.3 The ETV6-NTRK3 gene fusion 11 1.3 ETV6 (TEL) and the ETS family of transcription factors 12 1.3.1 The ETS gene family 12 1.3.2 ETV6 (TEL) 12 1.3.3 ETV6 translocations 13 1.4 NTRK3 (TRKC) and the Trk family of neurotrophin receptors 15 1.4.1 The Trk family 15 1.4.2 NTRK3 (TRKC) 16 1.4.3 NTRK3 signal transduction 17 A. SHC activates the RAS pathway 19 B. p85 subunit of phosphoinositol-3' kinase 19 C. Phospholipase-Cy (PLCy) activates PKC 21 D. Sucl-associated neurotrophin factor 21 1.5 Thesis Objectives 22

Chapter 2 MATERIALS AND METHODS 25 2.1 Tissue culture techniques 25 2.1.1 NIH3T3 cells 25 2.1.2 BOSC 23 cells 25 2.2 Constructing full-length ETV6-NTRK3 cDNA 26 2.3 Site-directed mutagenesis (SDM) 28 2.3.1 Correcting the nt 1717 T->C transition 28 2.3.2 Mutating the NTRK3 tyrosine residues and ATP- binding site 31 2.4 ETV6 HLH-domain deletion mutant: EN-AHLH 31 2.5 Transduction of genes using the retroviral vector MSCVpac 33 2.5.1 Cloning ETV6-NTRK3 constructs into MSCVpac 33 2.5.2 Transfection of BOSC 23 packaging cell line 34 2.5.3 Infection of NIH3T3 cells 34 2.6 Soft agar assay 36

2.7 FKBP36v-NTRK3 fusion protein 37

Chapter 3 RESULTS 40 3.1 The ETV6-NTRK3 construct 40 3.2 ETV6-NTRK3 transforms NIH3T3 cells 41 3.2.1 NIH3T3 morphology 41 3.2.2 Growth in soft agar 41 3.3 Investigating the role of the ETV6 helix-loop-helix domain 45 3.4 Investigating the role of the NTRK3 protein tyrosine kinase 49 3.4.1 Activation-loop tyrosines 49 3.4.2 ATP-binding site 50 3.4.3 ETV6-NTRK3 W565R mutation 51 3.5 Mutation of the PLCy-binding site 51 3.5.1 Creating the EN-Y628F mutant 51 3.5.2 Other ETV6-NTRK3 Y628 mutants 52

Chapter 4 DISCUSSION 54 4.1 ETV6 fusions with protein tyrosine kinases 54 4.1.1 ETV6-PDGFRB 55 4.1.2 ETV6-ABL 55 4.1.3 ETV6-JAK2 56 4.1.4 ETV6-NTRK3 56 4.2 Molecular characterization of ETV6-NTRK3 57 4.2.1 NIH3T3 transformation 57 4.2.2 Role of the ETV6 HLH domain 58 4.2.3 Role of the NTRK3 PTK domain 58 4.2.4 Role of the PLCy-binding site 61 4.2.5 Alternate signaling pathways 62 vi

4.3 Relevance of findings to congenital fibrosarcoma 65 4.3.1 NIH3T3 cells versus human fibroblasts 66 4.3.2 Spontaneous transformation of NIH3T3 cells 69 4.4 Future Directions 70 4.4.1 Chemically-induced dimerization 71 A. Introduction 71 B. Applications for FK1012 72 C. AP1510: a new dimerizer 73

D. FKBP36v-NTRK3 fusion proteins 73 4.4.2 Further molecular characterization studies 76

Chapter 5 SUMMARY AND CONCLUSIONS 79 5.1 ETV6-NTRK3 is a chimeric oncoprotein 80 5.2 ETV6 HLH and NTRK3 PTK domains are essential for transformation 80 5.3 PLCy-binding and activation not required 81 5.4 General comments 82

REFERENCES 83 vii

LIST OF TABLES

Page

Table 1. Primer Sequences used in Site-Directed Mutagenesis and 29-30 Sequence Analysis

Table 2. Transformation of NIH3T3 cells by EWS-FLI1, ETV6- 48 NTRK3, and ETV6-NTRK3 mutants viii

LIST OF FIGURES

Page

Figure 1. General Mechanisms of Oncogenesis 3

Figure 2. General Classification of Fibroblastic Tumours 9

Figure 3. Protein Structure of ETV6, NTRK3, and the Predicted 14 Structure of ETV6-NTRK3

Figure 4. Activation of NTRK3 and Subsequent Binding of Substrates 18

Figure 5. NTRK3 Signal Transduction Pathways 20

Figure 6. ETV6-NTRK3 Dysregulation of NTRK3 Signaling Pathways 23

Figure 7. Constructing the Full-Length ETV6-NTRK3 Fusion cDNA 27

Figure 8. Site-Directed and Deletion Mutagenesis of ETV6-NTRK3 32

Figure 9 Cloning ETV6-NTRK3 into MSCVpac 35

Figure 10. Cloning of Partial NTRK3 into pC4FvlE 38

Figure 11. Morphological Analysis of ETV6-NTRK3 Constructs 42-44 Expressed in NIH3T3 Cells

Figure 12. Soft Agar Assays Demonstrating Macroscopic Colony 46-47 Formation by ETV6-NTRK3-Expressing NIH3T3 Cells

Figure 13. Proposed mediators involved in ETV6-NTRK3-induced 64 oncogenesis

Figure 14. ETV6-NTRK3-transformed NIH3T3 Cells are Tumourigenic 67-68 in SCID Mice

Figure 15. Chemically-Induced Dimerization of FKBP36v-NTRK3 75 Using AP1510 LIST OF ABBREVIATIONS

A adenine AFB aggressive fibromatosis ALL acute lymphoid leukemia AML acute myeloid leukemia APS adaptor molecule containing PH and SH2 domains ATFS adult-type fibrosarcoma ATP adenosine triphosphate BAD Bcl-XL/Bcl-2 antagonist of cell death bp base pair Bel B-cell leukemia BDNF brain-derived neurotrophic factor C cytosine ede cell division cycle cDNA complimentary deoxyribonucleic acid CFS congenital fibrosarcoma CID chemically-induced dimerization CMML chronic myelomonocytic leukemia CS calf serum DAG diacylglycerol DD-PCR differential-display polymerase-chain reaction DMEM Dulbecco's Modified Eagle's Medium DNA deoxyribonucleic acid dT deoxythymidine

EN-AHLH ETV6-NTRK3 HLH-domain deletion mutant

EN-K380N ETV6-NTRK3 lysine-380-to-asparagine mutant EN-Y513F ETV6-NTRK3 tyrosine-513-to-phenylalanine mutant EN-Y517F ETV6-NTRK3 tyrosine-517-to-phenylalanine mutant EN-Y518F ETV6-NTRK3 tyrosine-518-to-phenylalanine mutant EN-Yx2F ETV6-NTRK3 tyrosines-517/518-to-phenylalanines mutant EN-Yx3F ETV6-NTRK3 tyrosines-513/517/518-to-phenylalanines mutant EN-Y628F ETV6-NTRK3 tyrosine-628-to-phenylalanine mutant EN-Y628T ETV6-NTRK3 tyrosine-628-to-threonine mutant EN-Y628E ETV6-NTRK3 tyrosine-628-to-glutamic acid mutant EN-Y628Q ETV6-NTRK3 tyrosine-628-to-glutamine mutant ERG early response gene ERK extracellular signal regulated kinase ETS E-26 transforming specific ETV1 ETS variant gene 1 ETV6 ETS variant gene 6 EVI1 ecotropic viral insertion site 1 E WS Ewing sarcoma F phenylalanine residue FCS fetal calf serum FGF fibroblast growth factor FLU Friend leukemia virus integration 1 G guanine G6PD glucose-6-phosphate dehydrogenase Gabl GRB2-associated binder-1 docking protein GAP GTPase activating protein GRB2 growth factor receptor bound protein 2 GTPase guanosine triphosphatase H&E hematoxylin and eosin HLH helix-loop-helix hTERT human telomerase reverse transcriptase IFB infantile fibromatosis IMDM Iscove's Modified Dulbecco's Medium

IP3 inositol triphosphate JAK Janus family of tyrosine kinases K lysine residue kb kilo-basepairs kD kilo-Daltons dissociation constant Kd MAPK mitogen-associated protein kinase MDS1 myelodysplasia syndrome 1 MN1 meningioma 1 mRNA messenger ribonucleic acid MSCV murine stem cell virus NEB New England Biolabs nM nanomoles per liter nt nucleotide NT-3 neurotrophin-3 NTRK1 neurotrophic tyrosine kinase receptor type 1 NTRK2 neurotrophic tyrosine kinase receptor type 2 NTRK3 neurotrophic tyrosine kinase receptor type 3 ORF open reading frame pac puromycin N-acetyl transferase PCR polymerase-chain reaction PI-3K phosphoinositol-3' kinase PKB protein kinase B PKC protein kinase C

PLOy phospholipase-Cy

PTK protein tyrosine kinase

Q glutamine residue R asparagine residue RACE rapid amplification of cDNA ends RNA ribonucleic acid RT-PCR reverse-transcriptase polymerase-chain reaction SCID severe combined immunodeficient SDM site-directed mutagenesis xii sec second SH2 Src homology 2 domain SH3 Src homology 3 domain SHC Src-homology and collagen SMRT silencing mediator for retinoic acid receptor and thyroid hormone receptor SNT sucl-associated neurotrophin factor target SOS Son-of-sevenless STAT signal transducer and activator of transcription SUMO-1 small ubiquitin-related modifier protein SV40 simian virus 40 T threonine resiude (amino acids) T thymine (deoxyribonucleic acids) TCR T cell antigen receptor TEL translocation, ETS, leukemia TRKA tropomyosin receptor kinase A TRKB tropomyosin receptor kinase B TRKC tropomyosin receptor kinase C TSG tumour-suppressor gene W tryptophan residue Y tyrosine residue Xlll

ACKNOWLEDGEMENTS

I would like to gratefully acknowledge my supervisor Dr. Poul Sorensen whose leadership and instruction have made this thesis possible. From the Sorensen laboratory, I wish to thank Michael Anglesio, Mathew Garnett, Stevan Knezevich, Chris Lannon, Beth Lawlor, Valia Lestou, Jerian Lim, Nataliya Melnyk, Kevin Morrison, and Cristina Tognon for their technical assistance and close friendship. I am also extremely grateful to Joan Mathers and Heather Wildgrove from B.C.'s Children's Hospital for their contributions to this thesis, and to the members of my supervisory committee, Dr. Robert Kay, Dr. Bruce Verchere, and Dr. Jiirgen Vielkind, for their continued support and interest. Finally, I thank God and my family who have enabled me to see this thesis to completion. 1

CHAPTER 1

INTRODUCTION

1.1 General Concepts in Cancer Genetics

1.1.1 Introduction

The term "cancer" is used to describe a variety of malignant diseases that result from uncontrolled cell proliferation. The dividing cells form large masses called neoplasms, or tumours, which can invade neighbouring tissues or may metastasize to more distant sites. In contrast, benign proliferations consist of compressed cells that neither invade other tissues nor metastasize. There are three general characteristics of tumour proliferation: this process is self-sufficient, increasingly malignant, and inevitably fatal if untreated. Consequently, an accurate diagnosis and an early intervention are essential for the successful treatment of cancer patients. When tumours arise from mesenchymal tissue (such as rhabdomyosarcoma, Ewing sarcoma, and congenital fibrosarcoma), they are called sarcomas. On the other hand, leukemias and lymphomas are classified as hematopoietic malignancies. Finally, carcinomas originate from epithelial cells including surface tissue such as skin. These three major classes of tumours are further sub-divided according to site, tissue type, and degree of malignancy [1].

Traditionally, viruses and exposure to environmental agents have been blamed for causing cancer, and the notion that cancer is a genetic disease is relatively new. However, it is becoming more and more evident that basically all cancer 2 results from genetic mutation. The altered genes are often involved in regulating cell development or proliferation; therefore, their disruption results in uncontrolled cell division and tumour formation. The current opinion is that a neoplasm arises from a single cell in which the initial transforming event occurred

[1]. Evidence to support this theory comes from studies of the X-linked enzyme glucose-6-phosphate dehydrogenase (G6PD). Because of random X-chromosome inactivation, female heterozygotes express only one of two G6PD alleles in each somatic cell. A population of cells would therefore be expected to express both G6PD alleles at equal levels. Interestingly, tumour cell-lines established from women heterozygous for G6PD expressed only one or the other allele but not both [2]. This suggests that the tumour cells are clonal descendents of a single cell.

Normal cell growth is regulated by two classes of genes. Tumor-suppressor genes (TSGs) prevent tumour formation by inhibiting uncontrolled cell growth. On the other hand, proto-oncogenes play a role in stimulating cell division and differentiation [3]. Consequently, malignant neoplasms arise when mutations result in the loss of a TSG, the activation of a proto-oncogene, or a combination of both [1]

(Figure 1). Furthermore, cancers can also result when genes which are involved in the repair of damaged DNA become mutated [4].

1.1.2 Tumour-suppressor genes

TSGs, as their name implies, function to regulate cell growth, differentiation, or other basic functions in order to prevent unregulated proliferation. However, cancers can arise if both copies of a specific TSG are lost or inactivated. For example, heterozygous individuals who carry a normal as well as a mutated copy of a TSG •a -^ o tn * fe. u 2 o •rH -t-> T3 CO S 2 cr S Tj OJ to 0» ^ H-> TJ O TJ co «* .2 co c CU rd rH OH CO ^ -|_> I O Ql rH ro C ^ UH 3 OJD OJ O bO O . M-l SH c C ta CH M_L 60 ^ Tj O «S ^ 6 o o 6 TJ CJ B O ra CO •+-> _ en o .s >H r o 0-, H rH H c O cn OJ .2 P rC 8 bb £ Ci c EH 0J OJ -rt o O OJ •i-H CJO O „rH H-_> •4-> -i-> tti O £ cCoO CO 3 3 CO O CL. 6 rH O cj co •=! o H-> CO CD on CO OJ OJ C OJ rH C CO CO OJ OJ OJ CJ pa CO bO • be to O . O 01 o u cj . o u c H C o co en o O 6 OJ +-o> o rH S-H ON fin rQ P C<*>O _: ^ OJ rH OJ 8 2 ~J < CQ u Cl cs • rH rC O 2 C £ o O co 4

(because of a germ-line mutation) may acquire a second, somatic, mutation in the normal allele which results in a homozygous genotype. This "two-hit" hypothesis, which was proposed by Knudson [5] to explain hereditary and sporadic forms of cancers such as retinoblastoma, ultimately resulted in the identification of the retinoblastoma TSG, RBI [6-8]. Other TSGs which are associated with specific disorders include WT1 in familial Wilm's Tumour [9], TP53 in Li-Fraumeni

Syndrome [10], and NF1 in neurofibromatosis type 1 [11,12].

TSGs can also be inactivated as a result of chromosome translocations, and there are two possible mechanisms for this process. A common event in translocations is the deletion of genetic material in the vicinity of the fusion breakpoints. In cases where TSGs are located at the breakpoint of one of the translocation partners, translocation-associated deletions would therefore lead to the loss of that suppressor-gene allele. For example, Tanaka, et al. [13], suggest that this is one mechanism of RBI inactivation in myeloid leukemogenesis. Alternatively, chromosome translocations may cause two genes to be fused together. However, if the codons of one parent gene are not in the same reading frame with those of the second gene, then the out-of-frame codons could likely encode for a stop codon and result in the expression of a truncated and non-functional fusion protein. A TSG could therefore be inactivated by out-of-frame translation of fused mRNA resulting from translocations.

1.1.3 Proto-oncogenes

Proto-oncogenes encode a wide range of proteins which may be involved in various aspects of regulating cell growth or which have key roles in signal 5 transduction. Membrane-bound receptors mediate signal transduction by interacting with extracellular growth factors and cytokines which instruct cells to grow and differentiate. Receptor-ligand binding results in the transmission of biochemical signals across the cell membrane and triggers the action of second messengers that convey the signal into the nucleus. This leads to the activation of specific transcription factors and results in the up-regulation or repression of particular target genes. The products of proto-oncogenes are intricately involved in these signal transduction pathways. They can encode the growth factors or extracellular cytokines as well as the membrane-spanning receptors. One example of each type is platelet-derived growth factor [14, 15] and nerve growth factor receptor [16]. Moreover, other proto-oncogene products function as membrane- associated or cytoplasmic proteins that serve to transduce biochemical signals to the nucleus. One example is the monomeric guanine nucleotide-binding protein (G protein), H-RAS [17]. Proto-oncogene products inside the nucleus, including N-

MYC [18], are employed in regulating DNA transcription and replication [4].

Oncogenes are altered, and activated, forms of proto-oncogenes that were originally identified as transforming genes in viruses. Therefore, the overexpression of oncogenes results in autonomous cell proliferation and tumour formation [1]. Proto-oncogenes can become activated and overexpressed via three mechanisms [1]. First, a structural mutation can result in the expression of an abnormal protein that confers a growth advantage to somatic cells. An example is the point mutation identified in H-RAS derived from a bladder cancer cell line [19].

This nucleotide change results in an amino-acid substitution that destroys the 6

GTPase activity of H-RAS and leads to the constitutive activation of H-RAS in a

GTP-bound state. A second mechanism of proto-oncogene activation is gene amplification. The amplified regions of DNA may appear as very small accessory chromosomes called double minutes or as homogeneously-staining regions which consist of multiple copies of a particular DNA segment and which do not band normally. In 40% of neuroblastoma cases, for example, the N-MYC proto-oncogene is amplified up to 200 times [20]. Finally, proto-oncogenes can be activated by chromosomal translocations. Such rearrangements may result in promoter exchange and lead to proto-oncogene dysregulation and overexpression. The t(8;14)(q24;q32) translocation in Burkitt's lymphoma positions the c-MYC proto- oncogene, from chromosome 8, within the highly expressed immunoglobulin heavy-chain locus on chromosome 14 [21]. c-MYC expression is subsequently dysregulated by promoter and enhancer sequences normally associated with the immunoglubulin gene [22]. Alternatively, translocations may actually fuse two genes together and result in the expression of a chimeric protein with oncogenic properties [1]. The role of fusion genes in cancer is further discussed below.

1.1.4 Fusion genes

Chromosomal translocations can activate proto-oncogenes either by splicing genes with new transcriptional activating sequences or by expressing oncogenic fusion proteins. Translocations that fuse genes are intra-exonic when the breakpoints lie within the cDNA-encoding exons, and they are called inter-exonic when the splice junctions occur in the untranscribed intronic regions. Fusion proteins arise from translocations when the 5'-portion of one gene is fused in-frame 7 with the 3'-portion of the translocation partner because this preserves the open reading frame (ORF) across the fusion breakpoint. Translocations therefore unite otherwise unrelated genes and result in proteins that may trigger the dysregulation of normal cell activities. The classic example is the Philadelphia chromosome (Ph1) identified by Rowley [23] in virtually all patients with chronic myelogenous leukemia. Ph1 is the derivative chromosome 22 that results from a t(9;22)(q34;qll) rearrangement. This translocation fuses the c-ABL proto-oncogene on chromosome

9 with the breakpoint cluster region (BCR) gene on chromosome 22. Expression of the BCR-ABL fusion protein results in increased ABL kinase activity and in the malignant transformation of hematopoietic cells [1, 24].

Both hematologic malignancies and solid tumours express chimeric proteins from gene fusions. In solid tumours, one of the genes involved often encodes for a transcription factor. The fusion protein consequently functions as a chimeric transcription factor that is able to recognize the same DNA sequences as the parent transcription factor. However, transcriptional activation would be altered due to contributions from the domains of the partner gene [25]. Two examples are the t(ll;22)(q24;ql2) EWS-FLI1 and t(21;22)(q22;ql2) EWS-ERG fusions identified in

Ewing sarcoma. Whereas EWS is a ubiquitously expressed gene of unknown function, both FLU and ERG are members of the ETS family of transcription factors

(described below in Section 1.4.1). The fusion of EWS to either of these two genes juxtaposes the EWS transcriptional-activation domain with the FLU or ERG ETS

DNA-binding domain, while the EWS RNA-binding domain is lost [26-31]. 8

1.2 Fibroblastic Tumours

Because the focus of this thesis is the ETV6-NTRK3 fusion in congenital fibrosarcoma (CFS), we now provide an overview of fibroblastic tumours of childhood. Bone and soft tissue sarcomas of childhood tend to be extremely primitive in appearance and therefore very difficult to differentiate from each other morphologically [32]. This is readily evident in childhood fibrous tumours.

Recently, molecular genetic characterization of human tumours has led to the realization that distinct genetic alterations are often associated with specific tumour subtypes. Our research laboratory was therefore interested in trying to identify molecular genetic markers of fibroblastic tumours occurring in the pediatric population.

1.2.1 Fibromatoses versus fibrosarcomas

Fibrous tumours of connective tissue represent the most common soft tissue neoplasms of infancy and early childhood [33]. This broad class of lesions can be divided into benign fibromatoses versus malignant fibrosarcomas (Figure 2).

Fibromatosis is a soft tissue proliferation of benign-appearing fibroblasts which exhibits a lower cellularity and mitotic activity, but expresses a more collagenous matrix, than fibrosarcoma. It most commonly occurs in children under 2 years of age and therefore is often referred to as infantile fibromatosis (IFB). Aggressive fibromatosis (AFB) is a clinical term for an otherwise histologically indistinguishable form of IFB. AFB cases are more likely to involve local invasion and have a greater risk for relapse [33]. co > • i—i •i-H co CO O ^ CO 0) JH *O P SH as

co ^ O 5 PQ to o J-i

v £ o ~> ° k 10

Fibrosarcoma has been defined as "a malignant tumour of fibroblasts that shows no other evidence of cellular differentiation and is capable of recurrence and metastasis" [33]. Because of the lack of distinguishing characteristics, the differential diagnosis of fibrosarcoma subtypes has been difficult (Figure 2). Adult fibrosarcoma is most common between the ages of 30 and 55 years with the median age being around 45 years [34]. The lesions, which typically range from 3 to 8 cm in diameter, grow slowly and occur most commonly in the thigh and trunk regions [33].

In contrast to adult fibrosarcoma, childhood fibrosarcoma can be divided into two subtypes (Figure 2). Fibrosarcoma in older children are generally referred to as adult-type fibrosarcoma (ATFS). They are associated with a more aggressive lesion and a poorer prognosis [32]. On the other hand, infants under 2 years old are diagnosed with congenital fibrosarcoma (CFS) and respond well to treatment; despite a recurrence rate of up to 40%, patients in this age group have a high (80-

90%) overall survival and a low (10%) metastatic rate [35, 36]. In fact, there is mounting cytogenetic and molecular genetic evidence that confirms the differentiation between the ATFS and CFS subtypes of childhood fibrosarcoma [37].

Childhood fibrosarcomas develop as painless swellings that grow rapidly and range in size from 1 cm up to 20 cm [33].

1.2.2 Congenital fibrosarcoma (CFS)

CFS is the second-most common soft tissue tumour in children next to rhabdomyosarcoma, and it is the most common soft tissue sarcoma in infants. Even though previous reports indicated that CFS occurs at a higher rate in boys, a more recent study indicates that both genders are equally affected [32]. CFS presents both 11 superficially and in the deep soft tissues as large swelling neoplasms. Typically, these rapidly growing tumours arise in the distal extremities including the head and neck region while less than 30% of the cases are axial [38]. CFS tumours are poorly circumscribed and have a grey-white or pale pink surface. Microscopically, they are characterized by dense cellularity and increased mitotic activity. CFS patients are most often treated by surgery in the absence of additional radiation or chemotherapy. Moreover, post-operative chemotherapy is not currently employed in the management of CFS [32]. Cytogenetic analyses of CFS cases have revealed nonrandom gains of chromosomes 8, 11, 17, and 20, with trisomy 11 being the most consistently observed alteration [38]. Recently, a novel and recurrent t(12;15)(pl3;q25) chromosomal rearrangement has been identified in our laboratory

(discussed below) [37].

1.2.3 The ETV6-NTRK3 gene fusion

In screening for recurrent genetic markers of CFS, Knezevich, et al. [37], performed routine cytogenetic analysis on fibroblasts from a series of CFS cases diagnosed in patients under 1 year of age. Three cases demonstrated abnormal clones with rearrangements of chromosome 15q25-26, and two had additional abnormalities of chromosome 12pl3. Fluorescence in situ hybridization (FISH) was performed next in order to more precisely map the breakpoints. Ultimately, FISH and molecular analysis revealed a rearrangement which fuses ETV6 (TEL) on chromosome 12pl3 to NTRK3 (TRKC) on chromosome 15q25. Moreover, FISH and reverse transcriptase (RT)-PCR demonstrated that this ETV6-NTRK3 gene fusion is not present in histologically identical IFB or ATFS cases. Furthermore, FISH and 12

RT-PCR indicated that the reciprocal NTRK3-ETV6 fusion was not expressed in CFS

[37]. The following sections will describe the ETV6 and NTRK3 genes, and their associated families, which are involved in the CFS t(12;15)(pl3;q25) rearrangement.

1.3 ETV6 (TEL) and the ETS family of Transcription Factors

1.3.1 The ETS gene family '

The first member of the E-26 transforming specific (ETS) gene family, v-ets, was discovered as a fusion protein expressed by the avian erythroblastosis virus, E26

[39]. Subsequently, a series of related genes that share a conserved ETS DNA-binding domain were isolated. A central A/CGGAA motif reflects their DNA-binding specificities and sequence conservation [40]. ETS-domain proteins are targeted to specific promoters via DNA-protein interactions and protein oligomerization. ETS transcription factors have been implicated in the regulation of gene expression in diverse biological processes including cell growth and development, transformation, and T-cell activation. ETS-domain proteins function as either activators or repressors, and their activities are often regulated by signal transduction pathways. One cascade pathway is the extracellular signal regulated kinase (ERK) pathway which was formerly called the mitogen-associated protein kinase (MAPK) pathway [41]. As described above, chromosomal translocations involving the ETS family members FLU and ERG result in the expression of chimeric oncoproteins in Ewing sarcoma [26-31].

1.3.2 ETV6 (TEL) 13

ETV6 is the ETS variant gene 6; it was formerly abbreviated TEL for translocation, ETS, leukemia. ETV6 is a member of a subset of ETS-family members that possess a helix-loop-helix (HLH) domain. Two other members of this subset are

ETS1 and FLU, and both of these are associated with murine erythroleukemia [42].

ETV6 is located on chromosome band 12pl3, and it is oriented in a telomere to centromere fashion. The gene is 240 kb long; it encodes the HLH domain in exons 3 and 4 and the ETS domain in exons 6 to 8 [43] (Figure 3a). The ETV6 HLH domain is hypothesized to play a role in protein dimerization, and perhaps ETV6 mediates transcriptional regulation via homodimerization. Although the precise role of

ETV6 remains to be determined, it has been shown that ETV6 can mediate transcriptional repression in the context of the monocyte colony stimulating factor receptor promoter [44]. Moreover, various studies indicate that ETV6 plays a critical role in early hematopoiesis and angiogenesis [45-47].

1.3.3 ETV6 translocations

ETV6 was first identified as the fusion partner of the platelet-derived growth factor receptor-p gene (PDGFR/3) in the t(5;12)(q33;pl3) of chronic myelomonocytic leukemia (CMML) [48], and has since been identified in a variety of gene fusions associated with hematopoietic malignancies [49-53]. These include fusions with core-binding factor oc2 (CBFA2 /AML1) [54, 55], the cellular Abelson leukemia virus homologue (ABL) [56], meningioma 1 (MN1) [57], the tyrosine-associated kinase

JAK2 [58, 59], six-twelve leukemia (STL) [60], myelodysplasia syndrome 1 (MDS1) and ecotropic viral integration site 1 (EVI1) [61], the caudal-related homeobox gene

CDX2 [62], Brx-like translocated in leukemia (BTL) [63], and long fatty acyl CoA 14

CO

H

i

H> H OH W 2 CO CS to

H cC co

co co

co

CO

co

CQ

.a cu be S

CO CC

CO

PQ u ft W 15 synthetase 2 (ACS2) [64]. The well characterized ETV6-CBFA2 (TEL-AML1) fusion is often accompanied by deletion of the normal ETV6 allele on chromosome 12 [65].

The fusion of ETV6 to NTRK3 in CFS is the first report of this ETS gene being involved in a human solid tumour [37].

1.4 NTRK3 (TRKC) and the Trk Family of Neurotrophin Receptors

1.4.1 The Trk family

The Trk family of protein tyrosine kinases includes NTRK1, NTRK2, and

NTRK3, which were formerly called TRKA, TRKB, and TRKC, respectively. These transmembrane receptors mediate the biological properties of the neuronal growth factor (NGF) family of neurotrophins. Trk family proteins bind neurotrophins with high affinity and are required in neuronal proliferation, differentiation, and survival [66]. NTRK1 is the NGF receptor, NTRK2 serves to bind brain-derived neurotrophic factor (BDNF), and NTRK3 is the primary receptor for neurotrophin-3

(NT-3). NT-3 can also bind to and activate NTRK1 and NTRK2 at high concentrations [66]. The crystal structures of the neurotrophin-binding domains of

NTRK1, NTRK2, and NTRK3 have recently been characterized to a high resolution

[67]. Trk family members are very similar at the amino acid level; in fact, their protein tyrosine kinase (PTK) domains are over 70% homologous [68]. Trk family members undergo ligand-induced dimerization and autophosphorylation which subsequently activate signal transduction cascades [69]. The NTRK2 and NTRK3 genes have been shown to engage in alternative splicing which results in non- catalytic receptor isoforms whose functions remain unknown. On the other hand, 16 both of the NTRK1 isoforms appear to have similar biological properties [70]. It has been shown that, for NTRK1, phosphorylating different tyrosines or interactions with another neurotrophin receptor (p75) can activate different pathways in the rat adrenal pheochromocytoma cell line, PC12, and lead to either differentiation or apoptosis [71]. This indicates that there are complex activation mechanisms for different signaling pathways. NTRK1 was first isolated in human colon carcinoma as part of a fusion with a tropomyosin gene [72], and NTRKl-overexpression has been reported in carcinomas of the thyroid, liver, and ovary where NTRK1 is not normally present [73]. Therefore, Trk family members are likely to play a role in oncogenesis when expression is dysregulated.

1A.2NTRK3 (TRKC)

NTRK3 encodes the neurotrophic tyrosine kinase receptor type 3 which specifically binds NT-3 and is not activated by NGF or BDNF [74]. The gene, located at chromosome 15q25, encodes a 145 kD protein with domains for ligand binding, membrane spanning, and protein tyrosine kinase activity (Figure 3b). There are two human NTRK3 splice variants: active NTRK3 consists of 825 amino acids but a non• functional variant contains a 14 amino-acid insert occurring after NTRK3 arginine

701. Both variants are equally proficient in binding NT-3 and in autophosphorylation; however, only NTRK3 lacking the insert can mediate signal transduction [74, 75].

Several reports indicate that NTRK3 is involved in the development and repair of the central nervous system [76-78]. Human NTRK3 is extremely similar to the rat (97%) and porcine (98%) homologues [75], and NTRK3 knockout mice exhibit 17 defects in sensory ganglia as well as cardiac valves and septa. Moreover, the NTRK3 gene encoding this receptor appears to be a target of inducible transcription factors after mechanical injury to certain parts of the brain [79], and high levels of NTRK3 are correlated with a favorable prognosis in medulloblastomas [80]. Overexpression of NTRK3 along with NT-3 in NIH3T3 cells, however, has been shown to cause transformation and result in focus formation [74]; therefore, a tight regulation of

NTRK3 expression under specific conditions is required for normal development.

1.4.3 NTRK3 signal transduction

NTRK3 is normally activated by cell surface ligand-mediated oligomerization.

Therefore, the binding of NT-3 results in the dimerization of NTRK3, and this facilitates autophosphorylation of NTRK3 cytoplasmic tyrosine residues 516, 705,

709, 710, and 820 and subsequent kinase activation [66, 70, 81-83] (Figure 4). Several adaptor molecules and PTK substrates have been linked to NTRK3 signal transduction. Phosphorylated NTRK3 tyrosine 516 binds the Src-homology and collagen (SHC) adaptor protein via the Src homology 2 (SH2) domain of the latter

[84], and is also the site of association with the regulatory (p85) subunit of phosphoinositol-3' kinase (PI-3K) [70, 85, 86]. Similarly, phosphorylated NTRK3 tyrosine-820 is the binding site for phospholipase-Cy (PLCy) [87, 88]. The sucl- associated neurotrophin factor target (SNT) protein has been shown to bind to an

NTRK3 juxtamembrane KFG sequence (lysine - phenylalanine - glycine; NTRK3 residues 461-463) and become tyrosine phosphorylated [89, 90] (Figure 4). Other molecules are known to associate with activated NTRK3 and potentially be involved in NTRK3 signaling [91]: these include the SH2 domain-containing 18

Figure 4. Activation of NTRK3 and subsequent bmding of substrates. NTRK3 monomers span the cellular membrane (double dotted lines). They consist of signal peptide (SS), extracellular ligand-binding (ELB), transmembrane (TM), juxtamembrane KFG, and protein- tyrosine kinase (PTK) domains. Neurotrophin-3 dimerizes NTRK3 resulting in autophosphorylation (P). Subsequently, sucl-associated neurotrophin factor target (SNT), Src-homology/collagen (SHC), the regulatory subunit (p85) of phosphoinositol-3' kinase, and phospho- lipase-Cy (PLCy) bind and become activated. 19 protein SH2-B, and the rat homologue for an adaptor molecule containing a

Pleckstrin homology domain and an SH2 domain (rAPS) [92].

A. SHC activates the RAS pathway

Early studies on the involvement of the SHC adaptor protein in signal transduction were conducted on NTRK1 in PC12 cells. The stimulation of NTRK1 by NGF resulted in the phosphorylation of SHC [93]; furthermore, the overexpression of SHC resulted in the neuronal differentiation of PC12 cells [94]. In

NTRK1, as well as NTRK3, activated SHC recruits growth factor receptor bound protein 2 (GRB2), an adaptor protein, as well as the RAS guanine nucleotide exchange factor Son-of-sevenless (SOS) to activate the RAS-RAF1 pathway [95-97]

(Figure 5). These observations confirm earlier studies that indicated NTRK1 signaling via the RAS-RAF1-MAPK pathway [98-103]. In fact, it now appears that

SHC-binding to NTRK1 is essential for RAS activation [85].

B. p85 subunit of phosphoinositol-3' kinase

Another adaptor molecule activated by NTRK3 via phosphorylation is the regulatory subunit, p85, of PI-3K. p85 subsequently recruits the catalytic subunit of

PI-3K, pi 10. PI-3K is in involved in the generation of phosphoinositide second messengers in response to growth factors [104]. However, the role of this pathway in

NTRK3 signaling remains unclear. Recently, it has been shown that pi 10 is activated by RAS, and this supports the view that PI-3K lies downstream of RAS [85,

105]. Moreover, RAS-activated PI-3K results in the direct interaction and regulation of protein kinase B (Akt/PKB) [71,106] which is involved in neuronal-cell survival

[107] (Figure 5). Akt acts as an anti-apoptotic serine/threonine kinase by Figure 5. NTRK3 signal transduction pathways. Neurotrophin-3 binding dimerizes NTRK3 and results in autophosphorylation. SHC and p85 bind to phosphotyrosine 516 and are activated. SHC activates the RAS-MAPK cascade, while p85 recruits pi 10. Activated RAS stimulates pllO and, subsequently, Akt/PKB. Phosphotyrosine-820 is bound by PLCy; this leads to the activation of PKC which complements MAPK activity. SNT recognizes the KFG sequence, and it signals via a neuronal-specific, Ras-independent pathway. 21 phosphorylating and inactivating the Bcl-XL/Bcl-2-associated death promoter, BAD.

Subsequently, BAD is no longer able to heterodimerize with and neutralize Bcl-2, thereby suppressing apoptosis and promoting cell survival [108].

C. Phospholipase-Cy (PLCy) activates PKC

Tyrosine-phosphorylated PLCy hydrolyzes phosphatidylinositol bisphosphate

to generate the second messengers inositol triphosphate (IP3) and diacylglycerol

2+ (DAG). IP3 is involved in Ca release whereas DAG is required for the activation of protein kinase C (PKC) [109-113]. However, mutation of NTRK1 tyrosine-785, the

PLCy binding site, failed to abrogate transformation of NIH3T3 cells or differentiation of PC12 cells even though PLCy could no longer associate with the receptor [66, 86]. It is interesting to note that while RAS-binding is required for activating the RAS-MAPK cascade [85], maximal activation requires the binding of

PLCy [114] possibly via PKC phosphorylation of RAF1 (Figure 5).

D. Sucl-associated neurotrophin factor

As described above, SNT has been shown to bind to an NTRK3 juxtamembrane KFG sequence and become activated by tyrosine phosphorylation

[89, 90]. SNT, a 90 kD protein, binds to the pl3 subunit of the cell-cycle regulatory complex that includes the p34cdc2 kinase and cyclin [89]. This protein is likely to be involved in neuronal-specific signaling elements that interpret NTRK3 activation as differentiation messages [66] (Figure 5). In fact, it has been shown that SNT becomes phosphorylated in PC12 cells and in primary cortical neurons after NGF and BDNF stimulation, respectively. Moreover, SNT was not phosphorylated in 22

PC12 cells exposed to EGF which has mitogenic effects on PC12 as well as NIH3T3 cells [89]. Other evidence is provided by studies on KFG which is a conserved juxtamembrane sequence identified in all NTRK family members. The deletion of this region impaired neuritogenesis [90] and abolished SNT binding and phosphorylation [89]. However, the mutated receptor was still able to facilitate substrate phosphorylation, RAS-MAPK activation, and cell survival [90]. Recent studies on fibroblast growth factor (FGF) receptor-1 signaling indicate that phosphorylated SNT can associate with GRB2 and SOS in FGF-stimulated fibroblasts

[115,116].

1.5 Thesis Objectives

The CFS t(12;15) rearrangement splices the first 5 exons of ETV6 to exons 6 to

8 of NTRK3, creating a hybrid gene encoding the ETV6 HLH protein, dimerization domain fused to the terminal 298 amino acids of NTRK3 [37, 43] (Figure 3c). These

NTRK3 residues include the PTK domain and C-terminus of NTRK3, but do not include the tyrosine-516 residue of NTRK3 which binds SHC and p85, nor the KFG sequence which binds SNT [37]. We hypothesized that expression of the ETV6-

NTRK3 fusion transcript is essential for the development of CFS tumours.

Moreover, we proposed that the HLH domain of ETV6-NTRK3 facilitates ligand- independent dimerization of the chimeric protein, and that subsequent constitutive

PTK activation leads to oncogenesis by dysregulating known or novel NTRK3 signal-transduction pathways (Figure 6). In order to characterize the ETV6-NTRK3 fusion protein, we established four specific experimental objectives: 23

i I .5 CO 0 CM •8 O MH CO II fcJ rS 01' CC cj CL, CO e > "2 Tj o H cc u re Tj W r2 s. CO (6 ti Ci tS| CC CC SH u oed CO *H C 01 -M» CU CC * > O •*-> MH o o £ E •r—H " G CC *H CC 01 o > rc Cnvi Hg > tU0> rc PM c ^ .2 TJ ce a _N TO CC 'S U I fig*

- TJ E TJ £f 1 £ « T r2 "8 M4 CO TJ OH r*H a H •9 ' CM e -p F *S rC r^ > u aj > H 2 0) cs C B CO •IH 68 JH OV cc -r!

£ 43 CH-^ 24

1) To determine the transforming properties of the chimeric ETV6-NTRK3 protein.

We expressed the full-length fusion cDNA in NIH3T3 cells. In order to measure

transformation, we assessed cell morphology as well as colony growth in soft

agar.

2) To determine the role of the ETV6 HLH domain. We created a HLH-domain

deletion mutant of ETV6-NTRK3 and repeated our experiments to assess

transformation activity.

3) To determine the role of the NTRK3 PTK domain. We used site-directed

mutagenesis (SDM) in order mutate the three activation-loop tyrosines (ETV6 -

NTRK3 Y513, Y517, and Y518) which are critical for NTRK3 autophosphorylation

[70, 87, 117, 118]. Furthermore, we mutated the ETV6-NTRK3 ATP-binding site at

lysine-380 (NTRK3 K572) [119] and assessed transformation activity.

4) To determine the involvement of PLCy in ETV6-NTRK3 signaling. We mutated

the PLCy-binding site in the chimeric protein (ETV6-NTRK3 Y628) [87, 88], and

repeated our transformation assays. 25

CHAPTER 2

MATERIALS AND METHODS

2.1 Tissue Culture Techniques

2.1.1 NIH3T3 cells

NIH3T3 cells, derived from murine embryonic fibroblasts, were obtained for long term culture from Dr. Robert Kay at Terry Fox Laboratories in Vancouver,

Canada. These early-passage fibroblasts, as well as several other cell lines expressing

ETV6-NTRK3 and its mutants (described below), were cultured at 37°C using standard methods [120]. The cells were grown in Dulbecco's Modified Eagle Medium

(DMEM) containing 10% calf serum (CS), high glucose, sodium pyruvate, and pyridoxine hydrochloride (GibcoBRL). The medium was supplemented with L- glutamine as well as antibiotic-antimycotic (GibcoBRL). Stocks from each cell line were preserved by freezing the cells, in media with 10% DMSO (Fisher), in liquid nitrogen.

2.1.2 BOSC 23 cells

BOSC 23 packaging cells were used to produce mature ecotropic viruses from the various ETV6-NTRK3 constructs cloned into the retroviral vector, MSCVpac

(described further in Section 2.5.1). Tissue culture work with this cell line was performed at Dr. Robert Kay's laboratory. BOSC 23 were cultured, at 37°C using standard methods [120], in DMEM containing 9% fetal calf serum (FCS). Confluent flasks were split no lower than 1:5 prior to transfection. The BOSC 23 cell line 26 packages retroviral RNA genomes into infectious, replication-incompetent retroviral particles. MSCVpac does not contain the gag, pol, and env structural genes necessary for particle formation and replication: these genes are stably integrated into the BOSC 23 genome [121-124]. Introduction of MSCVpac into the packaging cell line results in production of high-titer, replication-incompetent infectious virus particles. These particles can infect NIH3T3 cells and transmit the gene of interest to them; however, they cannot further replicate because the target cells lack viral structural genes.

2.2 Constructing Full-Length ETV6-NTRK3 cDNA

ETV6-NTRK3 was constructed by splicing a partial-ETV6-NTRK3 fragment

(clone B1.7) with wild-type ETV6. First, a pa.rti.al-ETV6-NTR.K3 fragment was obtained by digesting a 3'-RACE-PCR product [37], cloned into pCR™2.1

(Invitrogen), with Pvu II and Not I. This fragment contained nt 826 to 1033 of ETV6 cDNA fused in-frame to the 3'-end of NTKR3 cDNA beginning at nt 1601. Wild- type ETV6 cDNA, beginning at nt 24 and cloned into pGEM-3Z (a generous gift of Dr.

Peter Marynen, University of Leuven, Leuven, Belgium), was digested with EcoR I and Pvu II to generate the second fragment containing ETV6 nt 24 - 825. These two products were directionally ligated into pBluescript II KS at the EcoR I and Not I sites to produce the 2083-bp ETV6-NTRK3 fusion cDNA (Figure 7). All restriction endonucleases were obtained from New England Biolabs (NEB). The ORF extends from the ATG start codon, at nt 25 - 27, to the TAG translation-termination codon, at nt 1924 - 1926. Full-length clones were screened by DNA sequencing and PCR 27

co H 3 co 03 OO rH o 2 CN LO 0) fl cr1 E-i •i—i Uj C 3 13 cu S i > •43 ca rH H rH «3 H so W ?H .a CU M cu cc cd C PH CU •I-l cu I bc^ T3 . E-s .a £ < W H s £H cu CC •S CO O o _3 o cn • CN •rH r-H T3 cu CN .3 % cu JB- C cn o CO cc cc cu so bO CC C I bO SO E-H cu 0C) ^ H E-H PH o • OJHD 3 cu OH LO cu r2 I CN SH 00 CU r>H CO r ) > CO

rH - 00 PsHJ PH o 1 CU CN CO u cu W OCU HS a CC rH rC CU u Si r-^H 0 LO CH CN rC T3 co CU CH CO CO CU CO fl o CC • rH rO rH H-> cu U CU fl p rH fl <- JH cu cu CO C C 'ST CH •rH O CC o •rH o H-> CO o cu rH rO rH O CC u CH IN G O bpic cu w- 5P cC

MH M 28 using TEL541 and Trk2 primers (all primer sequences are described in Table 1). PCR obtained the expected 731-bp product using the following conditions: 94°C for 1.5 min, followed by 35 cycles of 94°C for 1 min, 60°C for 1 min, and 72°C for 1 min, and a final extension of 72°C for 10 min. DNA sequencing used ETV6 primers

TEL352 and TEL701 (Table 1; [54]), NTRK3 primers Trkl and Trk3, and standard primers T3 and T7. All sequencing was performed on an ABI 373A DNA Sequencer

(Applied Biosystems) and analyzed using DNA-STAR™ software.

2.3 Site-Directed Mutagenesis (SDM)

The QuikChange™ Site-Directed Mutagenesis Kit (Stratagene) was employed.

All SDM-PCRs were carried out as follows: 95°C for 30 s, followed by 12 cycles of 95°C for 30 sec, 55°C for 1 min, and 68°C for 10 min 15 sec. DNA sequencing was used to confirm all the point mutations; sequences using standard primer T7 or NTRK3 primer PS2 (Table 1) were obtained and analyzed as described above.

2.3.1 Correct nt 1717 T->C transition

Sequence analysis of the full length ETV6-NTRK3 cDNA using standard primer T7 revealed a T->C point mutation at nt 1717 which was likely introduced during the 3'-RACE-PCR procedure [37]. This transition mutation occurs within a conserved NTRK3 sequence encoding a portion of the PTK domain [74]. Translation of the fusion protein would result in an ETV6-NTRK3 W565R mutation due to the fact that the codon was changed from TGG to CGG. We designed the B1.7-

1759F/B1.7-1759R primer pair (Table 1) to correct this mutation. 29

Table 1. Primer Sequences used in Site-Directed Mutagenesis and Sequence Analysis. Primer Sequence (5'-3') Comments TEL114 GAC GCC ACT TCA TGT TCC AG ETV6 nt 114-133 TEL352 GGT GAT GTG CTC TAT GAA CTC C ETV6 nt 352-373 TEL541 CCT CCC ACC ATT GAA CTG TT ETV6 nt 541-560 TEL541rev AAC AGT TCA ATG GTG GGA GG ETV6 nt 541-560 TEL701 AGA ACA ACC ACC AGG AGT CC ETV6 nt 701-720 PS1 AAG GAC AAG ATG CTT GTG GC NTRK3 nt 1705-1724 PS2 GTA ATG CAC TCA ATG ACC TC NTRK3 nt 2304-2324 Trkl TCT CCT TGA TGT TCA ACC NTRK3 nt 2414-2431 Trk2 CCG CAC ACT CCA TAG AAC NTRK3 nt 1821-1838 Trk3 CC TCT TAA TGT GCT GCA CAT NTRK3 nt 1601-1620 MSCV-F CCT TTA TCC AGC CCT CAC TC vector nt 1371-1390 plE(1098rev) GGG TCC CCA AAC TCA CCC TG vector nt 1079-1098 B1.7-1759F CCT ATG GAA AGC AGC CAT GGT TCC AAC TCT CAA AC B1.7-1759R GT TTG AGA GTT GGA ACC ATG GCT GCT TTC CAT AGG ull64F GTG GCT GTG AAC GCC CTG AAG GAT CCC ull64R GGG ATC CTT CAG GGC GTT CAC AGC CAC U1562F C GGC ATG TCC AGA GAT GTC TTC AGC ACG G ul562R C CGT GCT GAA GAC ATC TCT GGA CAT GCC G ul574F GTC TAC AGC ACG GAT TTT TAC AGG GTG GGA GG ul574R CC TCC CAC CCT GTA AAA ATC CGT GCT GTA GAC ul577F C AGC ACG GAT TAT TTC AGG GTG GGA GGA CAC ul577R GTG TCC TCC CAC CCT GAA ATA ATC CGT GCT G utyrx2F GTC TAC AGC ACG GAT TTT TTC AGG GTG GGA GGA CAC utyrx2R GTG TCC TCC CAC CCT GAA AAA ATC CGT GCT GTA GAC 30

Table 1 continued. Primer Sequences used in Site-Directed Mutagenesis and Sequence Analysis. Primer Sequence (5'-3') utyrx3F TCC AGA GAT GTC TTC AGC ACG GAT TTT TTC AGG GTG GGA GGA CAC utyrx3R GTG TCC TCC CAC CCT GAA AAA ATC CGT GCT GAA GAC ATC TCT GGA U1907F GCC ACC CCA ATC TTC CTG GAC ATT CTT GGC

U1907R GCC AAG AAT GTC CAG GAA GAT TGG GGT GGC Y628T-for G GCC ACC CCA ATC ACC CTG GAC ATT CTT GGC Y628T-rev GCC AAG AAT GTC CAG GGT GAT TGG GGT GGC C Y628E-for G GCC ACC CCA ATC GAG CTG GAC ATT CTT GGC Y628E-rev GCC AAG AAT GTC CAG CTC GAT TGG GGT GGC C Y628Q-for G GCC ACC CCA ATC CAG CTG GAC ATT CTT GGC Y628Q-rev GCC AAG AAT GTC CAG CTG GAT TGG GGT GGC C TRKC1601-Xbfl I GC TCT AGA ATG TGC AGC ACA TTA AGA GG TRKC2490rev-S/?e I GG ACT AGT GCC AAG AAT GTC CAG G 31

2.3.2 Mutating the NTRK3 tyrosine residues and ATP-binding site

SDM was also used to introduce point mutations within the NTRK3 portion of the 2083-bp fusion cDNA in pBluescript II KS. Three primer sets (ul562F/ul562R, ul574F/ul574R, and ul577F/ul577R; Table 1) were used in the mutagenesis of the three activation-loop tyrosines (NTRK3 Y705, Y709, Y710; ETV6-NTRK3 Y513, Y 517,

Y518) [70,87,117,118] to phenylalanines, resulting in EN-Y513F, EN-Y517F, and EN-

Y518F mutants, respectively (Figure 8a). An ETV6-NTRK3 double Y517F and Y518F mutant (called EN-Yx2F) was made using utyrx2F and utyrx2R primers, and a triple

Y513F, Y517F, and Y518F mutant (called EN-Yx3F) was made using utyrx3F and utyrx3R primers (Figure 8a; Table 1). Furthermore, the ETV6-NTRK3 PTK-inactive mutant replacing the ATP-binding K380 (NTRK3 K572) [74, 119] with asparagine, designated EN-K380N, was made using ull64F and ull64R primers (Figure 8b; Table

1). Finally, ETV6-NTRK3 tyrosine-628 (NTRK3 Y820) was converted to either a phenylalanine (EN-Y628F), threonine (EN-Y628T), glutamic acid (EN-Y628E), or glutamine (EN-Y628Q) residue to ablate this amino acid as the putative binding site for PLCy (Figure 8c) [87, 88]. Four sets of primer pairs were employed to create these respective mutants: ul907F/ul907R, Y628T-for/Y628T-rev, Y628E-for/Y628E-rev, and Y628Q-for/Y628Q-rev (Table 1).

2.4 ETV6 HLH-Domain Deletion Mutant: EN-AHLH

A fusion protein mutant lacking a functional ETV6 HLH domain was constructed using restriction endonuclease digestions (NEB). Full-length ETV6, cloned into pBluescript II KS at the EcoR I and Sal I sites, was digested with BspM I 32

Figure 8. Schematic representation of constructs made by site-directed and deletional mutagenesis of ETV6-NTRK3. Non-mutated ETV6-NTRK3 is shown at the top. A) The three activation-loop tyrosines (Y) sequentially substituted with phenylalanine (F) in mutants EN-Y513F, EN-Y517F, and EN-Y518F. EN-Y517 and Y518 simultaneously mutated in the EN-Yx2F double mutant, and all three activation-loop tyrosines mutated in the EN-Yx3F triple mutant. B) The ATP-binding site, lysine-380 (K), substituted with asparagine (N) in mutant EN-K380N. C) Tyrosine-628, the PLCy-binding site, was mutated to a phenylalanine (EN-Y628F), threonine (EN-Y628T), glutamic acid (EN-Y628E), or glutamine residue (ENY628Q). D) The HLH-domain deletion mutant, EN-AHLH, created by deleting amino acids 56-108 which represent part of the ETV6-NTRK3 HLH domain. (From Wai, et al, in press.) 33 and EcoN I to delete a large portion of the HLH domain, and mung-bean nuclease

(NEB) was used to blunt-end the DNA for re-ligation. PCR, using standard primers

T3 and T7, was carried out as above (Section 2.2) and used to screen for clones containing the deletion mutation. Clones that were positive for the 1422-bp product were analyzed by DNA sequencing using primer T7 in order to ensure the absence of frameshift mutations. A Sac I fragment including the HLH deletion was purified and spliced into full-length ETV6-NTRK3 in order to replace the intact HLH domain with the deletion mutation. Clones were PCR-screened using the same conditions as above (Section 2.2); a 288-bp fragment, from TEL114 and TEL541rev primers (Table

1), confirmed the presence of the HLH deletion, and a 1603-bp product, using TEL114 and Trkl primers (Table 1), determined the correct cDNA orientation inside pBluescript II KS. The 1950-bp mutant, EN-AHLH, is missing base pairs 193-351, inclusively (Figure Sd).

2.5 Transduction of Genes Using the Retroviral Vector MSCVpac

2.5.1 Cloning ETV6-NTRK3 constructs into MSCVpac

The murine stem cell virus (MSCV) vector, MSCVpac, was derived from the murine embryonic stem cell virus and the LN retroviral vectors [125-127]. Upon transfection into a packaging cell line, MSCVpac transiently expresses (or integrates and stably expresses) a transcript containing the extended viral packaging signal

(VF+), the puromycin N-acetyl transferase (pac) resistance gene, and a gene of interest inserted into its multiple cloning site. The vectors achieve stable, high-level gene expression through a specifically designed 5' long terminal repeat. The entire ORFs 34 of ETV6-NTRK3, along with the 11 mutated constructs, were subcloned into

MSCVpac at the EcoR I site. However, the sequences that were cloned into this retroviral vector were not the entire 2083-bp fusion cDNAs because EcoR I-digestion was employed to generate these fragments. An EcoR I site is present within ETV6-

NTRK3 cDNA at nt 1945 which is after the stop codon (nt 1924 - 1926) and prior to the mRNA-polyadenylation signal. Therefore, expression of the complete fusion protein, from ETV6-NTRK3 nt 24 - 1949, was not influenced. Clones were screened using Bgl II-digestion (NEB), as well as MSCV-F/Trk3 PCR-amplification (detailed in

Section 2.2), to ensure the proper orientation of the ETV6-NTRK3 cDNA (Figure 9;

Table 1). Recombinant viruses were produced using BOSC 23 packaging cells.

2.5.2 Transfection of BOSC 23 packaging cell line

Calcium-chloride (CaCl2) mediated transfection was used to introduce the recombinant vectors into the BOSC 23 cells [128]. Packaging cells from a 10-cm dish were used to seed 6-well plates. On the day of the transfection, the media from each well was replaced by 1 mL of DMEM containing 9% FCS and 25 |imol of chloroquine

diphosphate (Sigma). 2 ug of DNA, mixed with 200 uL of 250 mM CaCl2 (Fisher) and

200 uL of 2X Hank's Balanced Salt solution (GibcoBRL), was added to the media. The cells were returned to the incubator for 8 to 10 hr. The supernatant was replaced with fresh DMEM/9% FCS at 8 - 10 hr and at 30 - 36 hr after the start of the transfection. The supernatant containing mature viruses was collected at 45 - 50 hr.

This solution was filtered prior to being used directly for infection or being stored at

-70°C.

2.5.3 Infection of NIH3T3 cells

36

NIH3T3 cells were infected with recombinant retroviruses in order to establish cell lines stably expressing ETV6-NTRK3 and its various mutants. NIH3T3 cells were seeded at low density on to 6-well plates. On the day of infection, the medium was replaced with 1 mL of DMEM/10% CS containing 20 |ig/mL hexadimethrine bromide (Sigma). 1 mL of viral supernatant was added, and the cells were incubated overnight. The medium was then replaced with 2 mL of fresh

DMEM/10% CS. 36 hr after the start of infection, the cells were placed under antibiotic selection. The NIH3T3 cells were grown in media containing 2 ug/mL of puromycin (Sigma) for 5 days. During this period, the media was changed as required; thereafter, the cells were grown in regular medium described in Section

2.1.1.

2.6 Soft Agar Assay

Agar assays were performed according established protocols [129]. NIH3T3 cells, infected with recombinant retroviruses carrying either empty vector or ETV6-

NTRK3, were plated in quadruplicate at a density of 30,000 cells per 6-cm tissue culture dish (Falcon). Each agar plate contains an underlayer, a cell layer, and several top layers. The 2-mL underlayer consisted of 1.1X Iscove's Modified

Dulbecco's Medium (IMDM; GibcoBRL), supplemented with 5% calf serum, 5% fetal bovine serum, and 0.63% agar noble (DIFCO). The 5-mL cell layers (IX IMDM, 10% calf serum, and 0.7% agar noble) contained the NIH3T3 cells suspended in soft agar.

2-mL top layers, identical to underlayers, were applied when the cells were initially plated, and additional 1-mL top layers were added every 3 to 4 days after plating. 37

Photomicrographs and pictures of the plates were taken 16 to 19 days after plating, and the plates were observed for up to 28 days for macroscopic colony growth.

2.7 FKBP3Cv-NTRK3 Fusion Protein

The 3'-portion of NTRK3 involved in the CFS translocation, including exons

encoding the PTK domain, was fused to an inducible dimerization domain, FKBP36v

i

(discussed in Section 4.4.1), using reagents received from ARIAD Pharmaceuticals,

Inc., and restriction endonucleases purchased from NEB. PCR was used to amplify the partial-IVTRlO insert, which includes nt 1601 to 2490, and to link restriction endonuclease recognition sites to the ends of the PCR fragment. The TRKC1601-Xfoz

I primer (Table 1) attaches an Xba I site upstream of NTRK3 nt 1601 while the

TRKC2490rev-Spe I primer (Table 1) links an Spe I site downstream from NTRK3 nt

2490. PCR generated the expected 905-bp insert using the following conditions: 94°C for 1.5 min, followed by 35 cycles of 94°C for 1 min, 55°C for 1 min, and 72°C for 1 min, and a final extension of 72°C for 10 min. The PCR product was cloned into the

Spe I site of pC4FvlE (Figure 10), and colonies were screened for correct orientation of the partial-NTRiG fragment using BaraH I digestion. Furthermore, sequence analyses using primers PS1, PS2, and plE(1098rev) (Table 1) were performed to verify the intact open reading frame.

The FKBP36v-NTRK3 fusion cDNA was subsequently sub-cloned into

MSCVpac in order to facilitate its efficient uptake and expression by NIH3T3 cells.

The FKBP36v-NTRK3 construct in pC4FvlE was liberated by an Apo l/Mlu I double digest. The 1358-bp insert was ligated into the EcoR I site of MSCVpac, and clones 38 39 were screened using EcoR I/Hind IE digest and PCR (with primer set MSCV-F and

Trk3 under conditions described in Section 2.2) to determine the orientation of the fusion cDNA. 40

CHAPTER 3

RESULTS

3.1 The ETV6-NTRK3 Construct

The 2083-bp ETV6-NTRK3 fusion cDNA was constructed, as previously described, by splicing a partia\-ETV6-NTRK3 fragment to wild-type ETV6 (Figure 7,

Chapter 2). This construct contains the entire ETV6-NTRK3 ORF and provides an in vitro model of the fusion gene which results from the t(12;15)(pl3;q25) rearrangement originally identified in CFS [37] and subsequently described in cellular congenital mesoblastic nephroma [130, 131], which is thought to be the renal counterpart of CFS, as well as AML [132]. (However, the ETV6-NTRK3 fusion reported in AML does not include ETV6 exon 5 [132] and is therefore different from the one originally identified in CFS [37].) We chose to create the fusion cDNA using a partial-ETV6-NTRK3 fragment generated by 3'-RACE-PCR because this clone contained the original CFS gene-fusion breakpoint [37]. Moreover, we were able to take advantage of a unique Pvu II restriction endonuclease cleavage site at ETV6 nt

825 in order to ligate the partial fragment with wild-type ETV6 cDNA and to create a fusion cDNA which encompassed the entire ORF (Figure 7, Chapter 2). Our ETV6-

NTRK3 construct fuses ETV6 nt 24 - 1033 in-frame with NTRK3 nt 1601 - 2673.

Therefore, the ETV6-NTRK3 chimeric protein is predicted to consist of the ETV6

HLH domain joined to the NTRK3 PTK domain. 41

3.2 ETV6-NTRK3 Transforms NIH3T3 Cells

Given the predicted structure of the ETV6-NTRK3 fusion protein [37], we hypothesized that its expression is essential for the development of CFS tumours.

Moreover, we proposed that the HLH domain of ETV6-NTRK3 facilitates ligand- independent dimerization of the chimeric protein, and that subsequent constitutive

PTK activation leads to oncogenesis by dysregulating known or novel NTRK3 signal-transduction pathways. Therefore, our first goal was to determine the transforming capabilities of the chimeric ETV6-NTRK3 protein. To this end, we transduced the fusion cDNA into NIH3T3 cells in an initial attempt to detect transformation.

3.2.1 NIH3T3 morphology

NIH3T3 cells were infected with recombinant retroviruses carrying either full-length ETV6-NTRK3 cDNA or empty MSCVpac vector as a negative control.

Positive transformants were obtained by puromycin drug selection, and cell morphology was determined. NIH3T3 cells infected with ETV6-NTRK3 exhibited a dramatically transformed phenotype compared to negative control cells (Figure 11a and b). The transformed cells possessed a high nucleus-to-cytoplasm ratio, their cell bodies strongly refracted light, and their growth was not inhibited by contact to other fibroblasts. On the other hand, NIH3T3 cells infected with vector alone were indistinguishable from wild-type NIH3T3 cells (data not shown).

3.2.2 Growth in soft agar

ETV6-NTRK3-infected NIH3T3 cells, which were plated in quadruplicate in two independent soft agar assays, consistently formed macroscopic colonies after 16 42

Figure 11. Morphological analysis of ETV6-NTRK3 mutants expressed in NIH3T3 cells. ETV6-NTRK3 and the indicated mutant constructs were cloned into the MSCVpac retroviral vector. Mature viruses were produced using BOSC 23 packaging cells and used to infect NIH3T3 cells to express the ETV6-NTRK3 fusion proteins. Infected NIH3T3 cells were put under puromycin selection (2 lig/mL) for 5 days. Morphological analysis was conducted on NIH3T3 cells expressing A) ETV6-NTRK3, B) empty MSCVpac vector, C) EN-AHLH, D) EN- Y513F, E) ENY517F, F) EN-Y518F, G) EN-Yx2F, H) EN-Yx3F, I) EN-K380N, and J) EN-Y628Q. Cells expressing ETV6-NTRK3 or EN-Y628Q exhibited a dramatically transformed phenotype compared to control cells with vector alone. NIH3T3 cells expressing EN-AHLH, EN-Y513F, EN-Yx2F, EN-Yx3F, or EN- K380N had a non-transformed phenotype while those expressing EN-Y517F or EN-Y518F displayed a semi-transformed phenotype.

45 to 19 days of culture in soft agar (Figure 12a and c; Table 2). Colony formation was only slightly less than that observed with retrovirally infected NIH3T3 cells expressing the Ewing sarcoma EWS-FLI1 fusion protein (Table 2). NIH3T3 cells expressing either of these fusion proteins are able to form macroscopic colonies in soft agar because transformed cells are anchorage independent. On the other hand, cells infected with vector alone failed to grow colonies even after 28 days of incubation (Figure 12b and d; Table 2). Expression of the predicted 73 and 68 kD

ETV6-NTRK3 protein doublet solely in transformed NIH3T3 cells was verified by

Western blotting using a-TrkC (C-14) or oc-ETV6:HLH antibodies [37]. No corresponding proteins were present in control cells infected with vector alone.

Therefore, ETV6-NTRK3 encodes a chimeric protein with strong transforming potential in NIH3T3 cells.

3.3 Investigating the Role of the ETV6 Helix-Loop-Helix Domain

Our next goal was to determine the mechanisms through which ETV6-

NTRK3 exerted its transforming properties. Previous studies of oncogenic fusion proteins involving the ETV6 HLH domain spliced to protein tyrosine kinases have indicated that HLH-domain dependent dimerization is a prerequisite for transformation [42, 59, 133]. We therefore predicted that the ETV6 HLH domain permits ligand-independent dimerization of ETV6-NTRK3 and PTK-mediated transformation by this oncoprotein. In fact, oligomerization studies provided strong evidence that ETV6-NTRK3 undergoes self-association and can heterodimerize with

ETV6; moreover, these interactions are mediated by the ETV6 HLH domain [134]. Figure 12. Soft agar assays demonstrating macroscopic colony formation by ETV6-NTRK3-expressing NIH3T3 cells. Puromycin-selected cells were seeded at a density of 30,000 cells in 60 mm plates at a serum concentration of 10%. Similar findings were observed in at least three independent experiments for each construct. A) NIH3T3 cells expressing ETV6-NTRK3 formed macroscopic colonies in soft agar. B) NIH3T3 cells infected with MSCVpac vector alone failed to form macroscopic colonies in the agar. C) Photomicrograph of a typical soft agar colony formed by ETV6-NTRK3-expressing NIH3T3 cells (magnification 200x). D) NIH3T3 cells infected with MSCVpac vector alone grew as single cells in soft agar (magnification 200x). E) NIH3T3 cells expressing EN-Y517F or EN518F mutants exhibited microscopic but not macroscopic colony formation (magnification 200x). (From Wai, et al, in press.) 47 48

Table 2. Transformation of NIH 3T3 cells by EWS-FLI1, ETV6-NTRK3 and ETV6-NTRK3 mutants

Construct Morphological Growth in Soft Agarf Appearance

EWS-FLI1 Transformed 215 ± 18

ETV6-NTRK3 Transformed 172 ± 14

Vector Normal 0

EN-AHLH Normal 0

EN-Y513F Normal 0

EN-Y517F Semi-transformed 0*

EN-Y518F Semi-transformed 0*

EN-Yx2F Normal 0

EN-Yx3F Normal 0

EN-K380N Normal 0

EN-Y628Q Transformed 196 ± 17

EN-Y628E Transformed Not counted

EN-Y628T Transformed Not counted Average number of macroscopic colonies per 60-mm diameter soft agar plate. Cells were seeded at a density of 30,000 cells per plate in quadruplicate. ^Limited microscopic colony growth was observed. 49

The HLH protein dimerization domain of ETV6 was deleted to assess the potential role of oligomerization in ETV6-NTRK3 transformation. Retrovirally- infected NIH3T3 cells expressing the HLH-mutant, EN-AHLH, were clearly morphologically identical to those infected with vector alone. These cells appeared flattened, did not refract light, demonstrated contact inhibition, and therefore did not exhibit a transformed phenotype (Figure 11c). Moreover, in soft agar assays, EN-

AffLH-infected NIH3T3 cells failed to form colonies even after 4 weeks of incubation

(Table 2). These results indicate that the HLH domain is essential for ETV6-NTRK3- mediated transformation of NIH3T3 cells. EN-AHLH protein expression at levels similar to ETV6-NTRK3 was verified by immunoprecipitation using the cc-TrkC (C-

14) antibody [37].

3.4 Investigating the Role of the NTRK3 Protein Tyrosine Kinase Domain

3.4.1 Activation-loop tyrosines

To assess the role of the PTK domain in transformation, NIH3T3 cells were infected with recombinant viruses expressing ETV6-NTRK3 proteins with tyrosine- to-phenylalanine mutations in each of the three activation-loop tyrosines

(designated EN-Y513F, EN-Y517F, and EN-Y518F, respectively) known to be essential for PTK activity [70, 87, 117, 118]. Transformation activity was assessed by cell morphology and ability to grow in soft agar. The EN-Y513F-expressing NIH3T3 cells exhibited a normal phenotype and failed to form macroscopic colonies in soft agar when scored after 28 days (Figure lid and Table 2). Cells expressing the EN-Y517F or 50

EN-Y518F mutant exhibited a semi-transformed phenotype (Figure lie and /); neither of these cell lines formed macroscopic colonies in soft agar but did show limited microscopic growth (Figure 12e and Table 2). Interestingly, the simultaneous ablation of both of these activation-loop tyrosines (Y517 and Y518) in the double mutant (EN-Yx2F) resulted in a normal NIH3T3 phenotype and the inability to support even limited growth in soft agar (Figure Hg and Table 2). Cells expressing the triple mutant, EN-Yx3F, also showed normal morphology and failed to grow either macroscopic or microscopic colonies in soft agar (Figure llh and

Table 2). The expression of equivalent levels of ETV6-NTRK3 and mutated proteins in each of the cell lines was confirmed by Western blot analysis [37]. These results indicate that the NTRK3 PTK region is crucial for transformation of NIH3T3 cells.

Moreover, they demonstrate that ETV6-NTRK3 Y513 is essential for transformation while both Y517 and Y518 are required for full transformation.

3.4.2 ATP-binding site

We wished to test our hypothesis that ETV6-NTRK3 functions as a chimeric

PTK. Therefore, the binding site (ETV6-NTRK3 K380; NTRK3 K572) for the tyrosine phosphorylation substrate, ATP, was mutated to an asparagine residue [74, 119].

NIH3T3 cells expressing the ETV6-NTRK3 kinase-inactive mutant, EN-K380N, appeared untransformed and did not form colonies in soft agar (Figure Hi and Table

2). Furthermore, full-length ETV6-NTRK3, EN-AHLH, and EN-K380N were each in-vitro translated and subjected to immunoprecipitation with a-TrkC (C-14) or a-

ETV6:HLH, followed by immunoblotting using a-phosphotyrosine. This revealed 51 tyrosine phosphorylation of ETV6-NTRK3 but not of EN-AHLH or EN-K380N [37].

These results demonstrate that ETV6-NTRK3 is a PTK capable of autophosphorylation, and that this activity requires an intact HLH domain and a functional PTK domain.

3.4.3 ETV6-NTRK3 W565 mutation

As described above (Section 2.3.1), sequence analysis of the full-length ETV6-

NTRK3 cDNA (made from clone B1.7) using standard primer T7 revealed a T-»C point mutation at nt 1717 which was likely introduced during the initial 3'-RACE-

PCR procedure to clone the ETV6-NTRK3 breakpoint [37]. This transition mutation occurs within the codon for ETV6-NTRK3 W565 (NTRK3 W757) which lies inside a conserved NTRK3 PTK region [74]. Translation of the B1.7 fusion cDNA would result in a EN-W565R mutant due to the fact that the codon was changed from TGG to CGG. In fact, in-vitro translated EN-W565R mutant was not tyrosine phosphorylated (S. Knezevich, personal communication). This finding reveals the critical importance of a fully intact PTK domain for autophosphorylation.

3.5 Mutation of the PLCy-Binding Site

3.5.1 Creating the ETV6-NTRK3 Y628F mutant

Tyrosine-628 of the ETV6-NTRK3 fusion protein (NTRK3 Y820) is known to be the binding site for PLCy [87, 88]. When the corresponding NTRK1 residue (Y785) was mutated to a phenylalanine residue, the mutant NGF-receptor was unable to interact with PLCy even though its expression still transformed NIH3T3 cells [86]. 52

Initially, we substituted Y628 in ETV6-NTRK3 with a phenylalanine residue. This

EN-Y628F mutant, when expressed in NIH3T3 cells, did not appear to have transforming ability because the cells did not exhibit a transformed phenotype and did not grow in soft agar (data not shown). However, immunoblot assays determined that the EN-Y628F protein was expressed at a much lower level than wild-type ETV6-NTRK3 and the other mutants [37]. Therefore, the NIH3T3 cells may have appeared normal simply because there was not enough chimeric protein being expressed. As a result, we decided to construct several other ETV6-NTRK3

Y628-mutant proteins.

3.5.2 Other ETV6-NTRK3 Y628 mutants

To determine the role of PLCy-binding in the transformation activity of

ETV6-NTRK3, NIH3T3 were infected with EN-Y628T, EN-Y628E, or EN-Y628Q recombinant retroviral vectors. The corresponding mutant proteins are predicted to lack the ability to bind PLCy, as phosphorylated ETV6-NTRK3 Y628 corresponds to the known PLCy-binding site in NTRK3 (Y820) [87, 88]. All three puromycin- selected mutant-expressing cell lines still appeared fully transformed, morphologically, and they were able to form macroscopic colonies in soft agar after

16-19 days at levels similar to ETV6-NTRK3. Representative morphology and soft agar assay results are shown for the EN-Y628Q mutant in Figure 11; and Table 2, respectively. Immunoblot assays confirmed that the EN-Y628T, EN-Y628E, and EN-

Y628Q mutants were unable to associate with PLCy and that ETV6-NTRK3 and mutated proteins were expressed at equivalent levels [134]. These findings, coupled 53 with the fact that the EN-Yx3F activation-loop mutant bound PLCy even though it was non-transforming [134], provides compelling evidence that binding and tyrosine phosphorylation of PLCy does not play a significant role in transformation by the ETV6-NTRK3 oncoprotein. 54

CHAPTER 4

DISCUSSION

4.1 ETV6 Fusions with Protein Tyrosine Kinases

The scheme outlined in the Introduction (Section 1.1.1) whereby the development of malignant lesions requires only the activation of a proto-oncogene, the loss of both alleles of a TSG, or a combination of both, is somewhat over• simplified. Tumour formation should instead be viewed as a multi-step process that consists of a series of genetic alterations in the tumour cell population [1].

Therefore, a specific set of key genetic changes would be common in all cases of a particular neoplasm while another set of alterations would characterize a different lesion. However, within a particular tumour type, individual cases would also be associated with unique genetic aberrations acquired during tumour development that contribute to so-called "genetic noise". We hypothesized that the t(12;15)(pl3;q25) rearrangement was a critical genetic alteration in CFS pathogenesis because FISH and RT-PCR analysis demonstrated the translocation in CFS cases but not in histologically identical IFB or ATFS cases [37].

The t(12;15)(pl3;q25) chromosomal rearrangement identified in CFS creates a fusion between coding exons of the ETV6 and NTRK3 genes [37]. This fusion gene encodes a chimeric protein consisting of the ETV6 HLH domain attached to the tyrosine kinase domain of NTRK3. In several other cases of chimeric proteins where the ETV6 HLH domain is fused to a PTK domain, it has been shown that the 55 resultant chimeric protein exhibits constitutive kinase activity which results in autophosphorylation (discussed below). Furthermore, this is dependent on the presence of the ETV6 HLH domain which mediates homodimer formation.

4.1.1 ETV6-PDGFRR

The ETV6-PDGFR(3 fusion protein is the product of the t(5;12)(q33;pl3) translocation in a subgroup of patients with CMML [42, 48]. The translocation fuses the ETV6 HLH domain to the transmembrane and cytoplasmic domains of the

PDGFRp. Analysis of in-vitro translated ETV6-PDGFRp demonstrated that the protein is capable of self-association; moreover, this homodimerization was abolished by the deletion of 51 amino acids within the ETV6 HLH domain. In vivo experiments showed that ETV6-PDGFR|3 transformed a murine hematopoietic cell line, Ba/F3, to interleukin-3 growth factor independence. This transformation required the HLH domain of ETV6 as well as the kinase activity of the PDGFR[3 portion of the fusion protein [42].

4.1.2 ETV6-ABL

The t(9;12)(q34;pl3) ETV6-ABL fusion was first described by Papadopoulos, et al. [56], in acute myeloid leukemia (AML) and fuses the ETV6 HLH domain with the

PTK region of the non-receptor tyrosine kinase, ABL. Golub, et al. [133], further investigated the role of the ETV6 HLH domain in the context of this fusion protein.

ETV6-ABL was found to transform Rat-1 fibroblasts as well as primary murine bone marrow cells; in addition, the chimeric protein localized to the cytoskeleton and conferred growth-factor independent growth to Ba/F3 cells. ETV6-ABL was able to 56 homodimerize in vitro, and this was a prerequisite for tyrosine kinase activation.

Deletional mutation of the ETV6 HLH domain abrogated ETV6-ABL homodimerization, cytoskeletal localization, constitutive phosphorylation, and transformation.

4.1.3 ETV6-JAK2

In acute lymphoid leukemia (ALL), a t(9;12)(p24;pl3) rearrangement fuses

ETV6 to the tyrosine-associated kinase, JAK2 [58, 59]. Previous studies, using a

CD16-CD7-JAK2 fusion [135] and an epidermal growth factor-JAK2 fusion [136], have indicated that JAK2, a member of the Janus .family of tyrosine kinases, may be activated by homodimerization. The ALL ETV6-JAK2 fusion protein includes the

ETV6 HLH domain and the JAK2 catalytic domain. Lacronique, et al. [59], demonstrated that ETV6-induced oligomerization of ETV6-JAK2 resulted in the constitutive activation of its tyrosine kinase activity and transformed Ba/F3 cells.

Recently, Ho, et al. [137], determined that the expression of ETV6-JAK2 in Ba/F3 cells resulted in constitutive activation of the JAK-STAT signaling pathway.

4.1.4 ETV6-NTRK3

Because the ETV6-PDGFRp\ ETV6-ABL, and ETV6-JAK2 fusion proteins have each been shown to be capable of transforming hematopoietic cells in vitro, we wanted to determine whether ETV6-NTRK3 has transformation activity. We chose to test this activity in the NIH3T3 embryonic fibroblast cell line as the ETV6-NTRK3 gene fusion was originally identified in CFS, which is thought to be of fibroblastic derivation [37]. Interestingly, however, an ETV6-NTRK3 gene fusion similar to the 57 one described for CFS was recently observed in an adult patient with AML [132], although transformation studies were not reported.

4.2 Molecular Characterization of ETV6-NTRK3

4.2.1 NIH3T3 transformation

Our experiments demonstrated that ETV6-NTRK3 is a potent oncoprotein, as

NIH3T3 cells expressing this molecule through retrovirally-mediated gene transfer were morphologically transformed and readily grew macroscopic colonies in soft agar. Colony formation was of the same order of magnitude as that of NIH3T3 cells expressing the Ewing sarcoma EWS-FLI1 molecule, known to be a strong transforming protein [129]. Moreover, ETV6-NTRK3 expressing NIH3T3 cells have been shown to be tumourigenic in severe combined immunodeficient (SCID) mice, which was not observed with cells expressing the PTK inactive EN-K380N mutant

[134]. These findings indicate that ETV6-NTRK3 is a strong oncoprotein, and is the first demonstration, to our knowledge, of an ETV6 fusion protein with transforming activity in NIH3T3 cells.

We hypothesized that the ETV6-NTRK3 chimeric oncoprotein functions as a

PTK and activates NTRK3 signaling pathways. Wild-type NTRK3 is a receptor tyrosine kinase that, upon binding of its ligand NT-3, undergoes dimerization and tyrosine auto- or cross-phosphorylation. This results in PTK activation and subsequent phosphorylation and activation of a variety of downstream signaling molecules. Phosphorylated NTRK3 Y516 is the binding site for SHC [84]. When

SHC becomes activated, it recruits adaptor proteins including GRB2 and SOS (and 58 possibly other molecules) to activate the RAS-RAF1 pathway [95, 96]. NTRK3 Y516 is also the binding site for the p85 subunit of PI-3K [70, 85, 86], which becomes phosphorylated and is involved in activation of the cell survival mediator,

Akt/PKB [71]. Another NTRK3 binding protein, SNT, interacts with the juxtamembrane KFG sequence of activated NTRK3 and becomes tyrosine phosphorylated; this molecule may mediate, in part, neural differentiation in the context of normal NTRK3 signal transduction [89, 90]. Finally, phosphorylated

NTRK3 Y820 binds PLCy [87, 88] which results in activation of the PKC pathway [109-

112].

4.2.2 Role of the ETV6 HLH domain

Given the predicted structure of ETV6-NTRK3 [37], we reasoned that the

ETV6 HLH domain likely mediates homodimerization and facilitates ligand- independent PTK activation. As discussed above, the ETV6 HLH domain has been shown to be required for homodimerization and subsequent autophosphorylation in fusions with PDGFRp in CMML [42] and with ABL in AML [133]. Moreover,

ETV6 HLH-domain dependent dimerization is believed to be important in fusions with Jak2 in ALL [59]. Pulldown experiments using either GST-ETV6-NTRK3 or an a-TrkC antibody demonstrated ETV6-NTRK3 homodimerization, and this was abrogated by deletion of the HLH domain of ETV6-NTRK3 [134]. Moreover, absence of the HLH domain eliminated transformation activity, tyrosine phosphorylation, and detectable PTK activity (discussed further below). These data confirm the role of the ETV6 HLH domain in ligand-independent homodimerization.

4.2.3 Role of the NTRK3 PTK domain 59

Experiments using a-phosphotyrosine antibodies demonstrated tyrosine phosphorylation of ETV6-NTRK3 in NIH3T3 cell lysates [134].

Autophosphorylation activity of the fusion protein was strongly indicated by several findings. First, mutation of the lysine-380 ATP-binding site of ETV6-NTRK3

(mutant EN-K380N), corresponding to NTRK3 K572 [74, 119], completely abolished tyrosine phosphorylation of the fusion protein [134] as well as its transformation activity. Second, the non-dimerizing EN-AHLH mutant was not tyrosine phosphorylated in NIH3T3 cell lysates [134], nor was it transforming. Third, while in-vitro translated ETV6-NTRK3 became tyrosine phosphorylated, both EN-AHLH and the EN-K380N kinase inactive mutant were not tyrosine phosphorylated upon in-vitro translation [37]. It is also interesting to note that the EN-K380N, EN-Y513F,

EN-Yx2F, and EN-Yx3F mutants, all of which are non-transforming, should in theory be capable of homodimerization as well as heterodimerization with ETV6 or less well characterized downstream HLH-containing molecules. The potential for dimerization of these non-transforming mutant proteins suggests that dimerization alone is insufficient for transformation. Therefore these results strongly suggest that ETV6-NTRK3 is a constitutively active PTK capable of autophosphorylation, and that this activity requires both an intact HLH domain and a functional PTK domain.

NTRK3 autophosphorylation and signaling originates at a conserved

YSTDYYR sequence within the PTK domain which contains three activation-loop tyrosine residues. The function of the first conserved tyrosine residue (NTRK3 Y705 corresponding to ETV6-NTRK3 Y513) has yet to be determined, while the second 60 and third tyrosines, Y709 and Y710 of NTRK3, are responsible for autophosphorylation and correspond with tyrosines 517 and 518 of ETV6-NTRK3

[37, 70, 87,117,118]. In order to investigate the role of these activation-loop tyrosines in NIH3T3 transformation, we mutated each to phenylalanine and tested the resulting mutants for transformation activity. Simultaneous mutation of the second and the third residues (EN-Yx2F) or of all three tyrosines (EN-Yx3F) resulted in a non-transformed phenotype. Similarly, expression of the EN-Y513F mutant involving the first activation-loop tyrosine resulted in morphologically normal

NIH3T3 cells, indicating that this tyrosine is essential for transformation. However,

EN-Y517F or EN-Y518F mutants both exhibited microscopic colony formation in soft agar indicating partial transformation; therefore, while both residues appear to be required for full transformation, there may be some level of redundancy between these two tyrosines in ETV6-NTRK3-mediated transformation. Interestingly, tyrosine phosphorylation of each of these mutant proteins has been demonstrated in cell lysates expressing these proteins [134]. It is possible that while some level of autophosphorylation is maintained in the mutants (possibly through phosphorylation of other NTRK3 PTK tyrosines), recruitment of essential substrates for transformation may be abrogated by these mutations. Alternatively, the mutant proteins may become trans-phosphorylated on tyrosines by other PTKs present in

NIH3T3 cells. These findings are somewhat contradictory to mutational analyses of

NTRK1 signaling performed on the corresponding conserved activation-loop sequence in NTRK1 [138], where it has been shown that ligand-induced autophosphorylation of wild-type NTRK1 markedly decreases when the three 61 activation-loop tyrosines are mutated individually or in pairs. More studies are necessary to confirm the relationship between these three tyrosine residues and

ETV6-NTRK3 autophosphorylation.

4.2.4 Role of the PLCy-binding site

As discussed above, PTK activation of wild-type NTRK3 results in the direct association and tyrosine phosphorylation of several downstream molecules including SHC, GRB2, and the PI-3K p85 subunit [70] as well as SNT [89, 90].

However, the binding sites for each of these molecules are lost from the NTRK3 portion of the ETV6-NTRK3 fusion protein due to the position of the breakpoint

[37]. In fact, it has been verified that ETV6-NTRK3 is not able to association with either SHC, GRB2, or PI-3K p85 as predicted from the structure of the chimeric protein [134]. On the other hand, the carboxy-terminal tyrosine residue (Y628), corresponding to NTRK3 Y820 which is known to bind PLCy [87, 88], is retained in the ETV6-NTRK3 protein [37]. Studies have shown that ETV6-NTRK3 is indeed able to bind and tyrosine phosphorylate PLCy [134]. However, the expression of three different independent ETV6-NTRK3 constructs with mutations of this tyrosine to threonine, glutamic acid, or glutamine, respectively, each transformed

NIH3T3 cells to the same degree as non-mutated ETV6-NTRK3 (Table 2) even though there was no detectable association of these mutant proteins with PLCy in immunoprecipitation experiments [134]. Interestingly, immunoblot experiments have shown that several of the activation-loop tyrosine mutants also bound PLCy even though they were non-transforming [134]. These results provide preliminary 62 evidence that ETV6-NTRK3-induced transformation does not require PLCy activation and thus is likely not mediated through the PKC pathway. However, a more detailed analyses of PKC activation in ETV6-NTRK3-transformed cells are required to confirm this hypothesis. It should also be noted that mutation of the corresponding PLCy-binding tyrosine in full-length NTRK1 similarly did not abolish transforming activity in NIH3T3 cells or differentiation activity in PC12 cells [66].

4.2.5 Alternate signaling pathways

The lack of ETV6-NTRK3 association with SHC or GRB2 does not rule out a role for the RAS-RAF1 pathway in ETV6-NTRK3-mediated transformation, as other tyrosine kinases including Src and ABL may be involved in RAS activation. One possibility is that other less well characterized NTRK3-associating molecules may bind to ETV6-NTRK3 and mediate this action. Therefore, it is possible that SHC

(and therefore the RAS-RAF1 pathway) might be activated indirectly without binding to the fusion oncoprotein. The indirect activation of SHC by ETV6-NTRK3 can be determined by performing an immunoblot experiment using an antibody to pull down SHC. Because SHC is a phosphoprotein, activated SHC can be detected with an antibody against phosphotyrosine. However, preliminary studies in our laboratory involving co-expression of ETV6-NTRK3 and a RAS dominant-negative mutant in NIH3T3 cells did not result in diminished transformation (C. Tognon, personal communication). This suggests that RAS may not contribute significantly in mediating ETV6-NTRK3 signaling. Several newly described molecules have been shown to interact with activated NTRK3 and to be potentially involved in

NTRK3 signaling, including rAPS and SH2-B [91]. Recent reports have also 63 identified novel substrates interacting with the related receptor molecule, NTRK1.

For example, the addition of NGF to PC-12 cells expressing NTRK1 resulted in the phosphorylation of the GRB2-associated binder-1 (Gabl) docking protein and induced the association of PI-3K [139]. Caveolin, a protein marker of caveolae [140,

141], has been reported to interact with and inhibit NTRK1 activation in PC12 cells exposed to NGF [142]. It is highly possible that these proteins are also able to interact with activated NTRK3 or ETV6-NTRK3. Another intriguing possibility is that

ETV6-NTRK3 recruits components of potentially novel signaling pathways not involving RAS-RAF1 activation or functioning downstream of RAS-RAF1 (Figure

13).

Interestingly, ETV6-NTRK3 has been observed to in-vitro heterodimerize with wild-type ETV6 [134]. Whether or not this heterodimer plays a significant role in transformation remains to be determined. It is possible that binding of ETV6-

NTRK3 to ETV6 depletes levels of the latter, which has been hypothesized to function at least in part as a tumour suppressor [143]. In fact, a recent report demonstrates that the central region of ETV6, which is also included in the ETV6-

NTRK3 fusion protein, is involved in recruiting a repressional complex including the silencing mediator for retinoic acid receptor and thyroid hormone receptor

(SMRT) and the mSin3A transcription repressor [144]. ETV6 is known to be a phosphoprotein [145] and so may be a substrate for ETV6-NTRK3; however, studies in our laboratory revealed no evidence of ETV6 tyrosine phosphorylation by ETV6-

NTRK3 (S. Knezevich, unpublished results). Subcellular localization studies suggest that ETV6-NTRK3 resides predominantly in the cytoplasm, but it is also 64

cn > 5 ^ c H I o W to H 93 0 c§ « > cu S-H »H HW £ CJ H ea cn

cn .5 • *-< PH CJ cn ffl B rg < 0 CD c u 5 Q_ i c C QJ bC O 0 CO cu 5 CJ 3 u 5 C o OHTJ 0 o SH CO ft CJ 0 SO c 1—I ce c > > I CJ (0 H CJ CO H o w cn > CU CO s- '43 CJ o D-, CO > TJ ' . CU TJ 1—j TJ 5 g 8.-9cn c cu CJ •i-H o CU CO •43 ed > TJ .s N pa > • i-H —i X>CN (0 rH -T—( +-> c OJ -t-» rC .s cu CJ u o cd cn m c/s i SH rC -I e o > e to B x « a -T—I H TJ *** CU CJ z

B cu -3 cn I Toj ON CO - U > 2 .5> pa C9J -c3u fc. 6CJ OH CCOO •_ J cn c^ cn > rH o • rH 2 T6J bC JU -M CO i £L CO cu 6 .5 ffi DC co a M 65 observed at low levels in the nucleus [37]. Further studies are required to determine the significance of these findings. Alternatively, ETV6-NTRK3 heterodimerization with other proteins may lead to the recruitment and activation of, as yet, unidentified signaling pathways. For example, ETV6 has been shown to bind and subsequently inhibit the transactivation activity of FLU. Moreover, this inhibition requires both the HLH and DNA-binding domains of ETV6 [146]. ETV6 has also been shown to heterodimerize in an HLH-domain dependent manner with the ubiquitin-conjugating enzyme, UBC9 [147]. Although one study indicates that UBC9 plays a role in regulating proteosome-mediated degradation via ubiquitination [148], other data demonstrate that UBC9 actually participates in degradation-resistant protein modification and in protein trafficking by recruiting the small ubiquitin- related modifier protein SUMO-1 [149-151]. In fact, the interaction between UBC9 and ETV6 leads to the modulation of ETV6 transcription activity and not to its degradation [147]. Whether or not UBC9 interacts with and alters ETV6-NTRK3 activity remains to be determined. It would be possible to establish or rule out direct associations between these two proteins by performing an immunoblot assay.

Figure 13 illustrates our revised model of ETV6-NTRK3-induced transformation.

4.3 Relevance of Findings to Congenital Fibrosarcoma

There are several caveats to the experimental approach followed in this thesis. First, we expressed the ETV6-NTRK3 fusion protein in an immortalized murine fibroblast cell line. Second, the expression of retroviral vectors may induce spontaneous transformation of the host cells. The following sections address the 66 significance of these main concerns and how they may affect the interpretation of

ETV6-NTRK3-induced transformation in NIH3T3 cells.

4.3.1 NIH3T3 cells versus human fibroblasts

We chose NIH3T3 murine fibroblasts as our model in evaluating ETV6-

NTRK3-mediated transformation because CFS is thought to be of fibroblastic derivation (discussed in [37]). Moreover, because NIH3T3 cells have been used previously in demonstrating the transforming properties of the Ewing sarcoma

EWS-FLI1 oncoprotein [27] and other sarcoma-associated chimeric oncoproteins, we believed that the use of NIH3T3 cells was both suitable and logical. In fact, histological preparations of tumours derived from NIH3T3 cells expressing ETV6-

NTRK3 in SCID mice highly resemble those of actual CFS tumours (Figure 14).

However, the expression of human ETV6-NTRK3 cDNA in mouse cells might be somewhat artificial due to the fact that analogous ETV6-NTRK3-associating proteins may be absent while other mouse proteins are activated instead. Another important consideration is that NIH3T3 cells likely contain some alteration in their genetic complement because these are immortalized cells. In addition, the ribonucleoprotein enzyme, telomerase, is normally expressed in murine somatic cells but not in their human counterparts [152, 153]. The lack of telomerase activity in human cells results in the progressive loss of telomeric DNA in culture which is believed to limit cellular lifespan [154]. Therefore, the t(12;15)(pl3;q25) translocation may be only one of several genetic changes required for CFS pathogenesis.

A better understanding of CFS pathogenesis may be obtained from studies of

ETV6-NTRK3 expression performed on human fibroblasts. Unfortunately, FIGURE 14. Histological analysis of CFS and NIH3T3-derived SCID mouse tumour. A) A CFS case stained with hematoxylin and eosin (H&E). Note the characteristic high cellularity, spindle-like morphology, (bluish-purple), mitotic figures, and pleomorphism. B) A tumour derived from NIH3T3 cells injected into a SCID mouse. SCID mice, 5-6 weeks old, were injected with 107 NIH3T3 cells in the left lower flank. They were evaluated for tumour growth and organ involvement 11 days after injection, a time point at which tumour growth necessitated the termination of the experiment. (SCID mouse H&E photomicrograph kindly provided by B. Jansen, University of Vienna, Vienna, Austria.)

69 previous attempts at expressing gene fusion products in human cells and converting human cells into tumourigenic cells using K-RAS and simian virus 40

(SV40) large-T antigen [155] or activated v-myc and v-H-RAS [156] have yielded meager results. Recently, however, Hahn, et al. [157], have succeeded in creating tumourigenic human epithelial and fibroblast cells by co-expressing the telomerase catalytic sub unit (hTERT)1 along with the SV40 large-T and H-RAS oncoproteins.

Human cells would provide the proper intracellular milieu and the appropriate downstream molecules that are involved in ETV6-NTRK3 signal transduction in

CFS tumour cells. In addition, future experiments that focus on developing antagonists against ETV6-NTRK3 would be more valuable if they are performed in the context of human tumour cells. Nevertheless, we believe that our studies using

NIH3T3 cells provide an accurate measure of the transforming potential of ETV6-

NTRK3 and that ETV6-NTRK3-expression in these fibroblasts faithfully recapitulates CFS pathogenesis.

4.3.2 Spontaneous transformation of NIH3T3 cells

The expression of retroviral vectors may lead to spontaneous transformation of NIH3T3 cells depending on the site of viral genome insertion. The cDNA insert may disrupt key genes in the host cell that regulate cell growth and this would result in proliferation or transformation which is independent of ETV6-NTRK3- expression. Therefore, it was imperative to incorporate a negative control in which

NIH3T3 cells expressed empty MSCVpac vector in order to verify that cell transformation was solely due to ETV6-NTRK3. A second concern when using retroviral-mediated gene transfer is infection by replication-competent retroviruses 70 resulting in host cell-line transformation. In fact, the retroviral particles encoded by the MSCVpac vector and produced by the BOSC 23 packaging cells are replication- incompetent. As described in Materials and Methods (Sections 2.1.2 and 2.5.1),

ETV6-NTRK3 cDNA cloned into MSCVpac expresses a transcript encoding only the extended viral packaging signal (*F+), the puromycin N-acetyl transferase (pac) resistance gene, and our fusion gene. The retroviral gag, pol, and env genes essential for particle formation and replication are stably integrated into the BOSC 23 packaging-cell genome [121-124]. It is highly unlikely that recombination events within the BOSC 23 cells would result in the production of mature, replication- competent virus particles because this would require three independent recombination events (for the gag, pol, and env genes, respectively) occurring for one virion. We therefore believe that the use of retroviral-mediated gene transfer is an effective method to produce cell lines that stably express ETV6-NTRK3 or its various mutants.

4.4 Future Directions

The experiments we have described provide invaluable preliminary data on the role of the ETV6-NTRK3 oncoprotein in CFS pathogenesis. Nevertheless, many questions remain unanswered, and we are only beginning to understand a few of the mechanisms responsible for mediating ETV6-NTRK3 signaling. The following sections propose a variety of experiments designed to further characterize the ETV6-

NTRK3 protein, to identify downstream signaling molecules, and to determine the pattern of gene activation and inactivation. 71

4.4.1 Chemically-Induced Dimerization

A. Introduction

From the predicted structure of the ETV6-NTRK3 fusion protein, we hypothesized that the ETV6 HLH-domain permits ligand-independent dimerization and subsequent autophosphorylation of the NTRK3 PTK domains. This theory was confirmed by observations where the EN-AHLH mutant was unable to bind to in- vitro translated ETV6-NTRK3 [134]. Moreover, the EN-AHLH mutant exhibits no

NTRK3 phosphorylation activity [134] and does not transform NIH3T3 cells.

However, it is conceivable that the ETV6 sequences involved in the CFS t(12;15) translocation have contributions to the overall function of the fusion protein other than homo- or heterodimerization. As discussed above, the ETV6 central domain mediates transcriptional repression by associating with SMRT and mSin3A, while the ETV6 HLH domain represses gene transcription through a mechanism that is independent of known co-repressors [144]. One method to eliminate the contributions of such ETV6 domains in transformation is to fuse an inducible dimerization domain, FKBP12, to the NTRK3 PTK domain and to subsequently facilitate dimerization using synthetic ligands [158,159]. Comparison of the degree of transformation between NIH3T3 cells expressing ETV6-NTRK3 with FKBP-

NTRK3-expressing cells would, in part, determine if there are alternate functions of the ETV6 HLH and other regions.

Synthetic, cell-permeable ligands have been studied that were originally designed to control the intracellular oligomerization of specific proteins [158]. The fungal-derived immunosuppressant, FK506, was chosen as a potential matchmaker 72 because of its lipid solubility, metabolic stability, and high affinity binding to its target receptor. FK506 readily diffuses across cell membranes and strongly binds to

FKBP12, an FK506 binding protein (Kd = 0.4 nM) [160-162]. FKBP12 belongs to a class of abundant proteins, the immunophilins, which are involved in protein folding and intracellular transport [158, 163]. However, FK506 is toxic: the binding of the

FKBP12-FK506 complex to calcineurin, a Ca2+/calmodulin-dependent protein phosphatase [164], results in impaired signaling of the T cell antigen receptor (TCR) and subsequent immunosuppression [165-167]. In order to develop a non-toxic dimerizing molecule, Spencer, et al. [158], linked two FK506 monomers together at their calcineurin-binding sites to form FK1012. They conducted stoichiometric assays to confirm that each FK1012 molecule could bind two FKBP12 moieties.

Consequently, FK1012 can be employed to generate dimers of FKBP12-domain containing fusion proteins. So-called chemically-induced dimerization (CID) of such chimeric proteins can therefore be applied to studying cell processes, including transcriptional regulation, protein trafficking, and receptor activation, which are mediated by protein oligomerization.

B. Applications for FK1012

FKBP-fusion proteins have been designed for a variety of investigations of receptor activation and signal transduction. For example, the £-chain cytoplasmic domain of the TCR, which contains three tandem copies of the tyrosine activation motif required for signal transduction, was fused to the FKBP12 in the Src-^-FKBP12 chimera [158]. The Src region, which consists of the first 14 amino acids of c-Src, allows myristylation of the protein and directs its subsequent targeting to the 73 membrane [168]. FK1012 mediated fusion-protein dimerization led to activation of signal transduction pathways that were identical to those induced by extracellular crosslinking of wild-type TCR using antibodies. Moreover, protein dimerization and receptor activation was reversed by the addition of the non-toxic FK506 monomer, FK506-M, which displaced the FK1012 from the FKBP12 domains [158].

More recent investigations applying CID to signal transduction involve FKBP fusions with Src-like tyrosine kinases [169]; the guanine nucleotide exchange factor,

SOS, in RAS activation [170]; and Fas signaling in thymocytes [171] and in normal human keratinocytes [172]. Currently, this method is also being applied to activate switches in genetically modified cells. For example, cells expressing FKBP12 fused to the cytoplasmic domains of the erythropoietin [173], c-kit [174], and thrombopoietin

[175] receptors proliferated in the presence of FK1012.

C. AP1510: a new dimerizer

AP1510 is related to the prototype dimerizer, FK1012. Both molecules are symmetrical homodimers of FKBP12-binding molecules. Whereas FK1012 is a semi-synthetic dimer of the natural product FK506, AP1510 is a completely synthetic molecule. It is a smaller, simpler molecule that consists of two directly linked copies of a synthetic FK506 FKBP binding moiety. AP1510 has no immunosuppressive activity and is non-toxic to cells [176, 177]. Currently, AP20187, a second-generation dimerizer with similar properties is being developed (ARIAD Pharmaceuticals, unpublished results).

D. FKBP36v-NTRK3 fusion proteins 74

As described in Chapter 2 (Section 2.7), the NTRK3 sequences involved in the

CFS t(12;15) rearrangement (NTRK3 nt 1601 - 2490) were cloned into pC4FvlE which

encodes an FKBP36v dimerization domain (Figure 10, Chapter 2). This NTRK3 insert is transcribed under the control of the human CMV enhancer promoter and is expressed with a carboxy-terminal epitope tag containing a 9 amino acid portion of

the influenza hemagglutinin protein. The FKBP36v domain in the vector encodes a mutant FKBP12 in which phenylalanine-36 is substituted with valine. This single amino acid change corresponds to a motif on AP1510 and greatly increases the

affinity of the interaction [178]. Because the FKBP36v-NTRK3 fusion protein lacks an amino-terminal myristoylation-targeting peptide, newly synthesized chimeras would not be directed to the cellular membrane (Figure 15). We have subcloned the

FKBP36v-NTRK3 cDNA into MSCVpac in order to obtain NIH3T3 cells that stably expressed this chimeric protein. Preliminary results based on fibroblast morphology

indicate that FKBP36v-NTRK3 transforms NIH3T3 cells to the same degree as ETV6-

NTRK3 (C. Tognon, personal communication). Therefore, the ETV6 sequences involved in the ETV6-NTRK3 fusion may only contribute to protein homo- and heterodimerization. However, more definitive experiments are required to draw a more solid conclusion. For example, because we observed that the expression of the kinase-inactive mutant, EN-K380N, resulted in morphologically untransformed

NIH3T3 cells, it would be interesting to determine the effect of co-expressing the non-mutated fusion protein along with EN-K380N. Perhaps heterodimerization of the oncoprotein with the mutant will result in a dominant negative effect and the inability to ectopically activate NTRK3 signal transduction pathways. Moreover, co- Figure 15. Chemically-induced dimerization of FKBP36v-NTRK3 using AP1510. The fusion protein, which would localize to the cytoplasm, contains the FKBP36v dimerization domain fused to the NTRK3 PTK domain and a hemagglutinin epitope tag (HA). The binding of AP1510 results in autophosphorylation (P) of the PTK domains. (From Travis, 1993.) 76 expression of these two proteins may assist in the identification of additional roles of the ETV6 regions.

4.4.2 Further molecular characterization studies

We have described initial investigations on the roles of the ETV6 HLH domain and the NTRK3 PTK domain in ETV6-NTRK3-mediated transformation.

However, additional studies are required to provide a more detailed understanding of both these regions. Further experiments to characterize these domains, to identify associating proteins, and to profile differential gene expression will now be discussed.

Apart from the three activation-loop tyrosines, the NTRK3 region involved in the ETV6-NTRK3 fusion protein contains 8 other tyrosine residues at ETV6-

NTRK3 positions 366, 412, 427, 477, 538, 560, 594, and 615 (corresponding to NTRK3 positions 558, 604, 619, 669, 730, 752, 786, and 807) which are all located within the

NTRK3 PTK domain [37, 74]. Whether or not these residues are involved in phophorylation and subsequent recruitment of signaling molecules remains to be determined. Similar to the analysis of the activation-loop tyrosines, the significance of these other residues can be investigated by employing site-directed mutagenesis to convert these residues individually, or in combination, to phenylalanines.

Very little is known about the central region of ETV6 compared to the amount of available information for the amino-terminal HLH domain and the carboxy-terminal ETS DNA-binding domain [48, 49, 144]. The recent discovery that the ETV6 central region binds a repression complex [144] demonstrates possible roles for ETV6, other than homodimerization, in ETV6-NTRK3-mediated 77 transformation. Furthermore, detailed analysis of the ETV6 amino-acid sequence of the ETV6-NTRK3 fusion protein has revealed two putative binding motifs (proline

- x - x - proline) for the Src homology 3 (SH3) domain. The first sequence (proline - glutamic acid - serine - proline) consists of ETV6-NTRK3 residues 20-23, and the second (proline - serine - serine - proline) contains amino acids 255-258 of the fusion protein. SH3-domain containing proteins may be found in adaptor proteins that link molecules together in a signaling pathway. A good example is GRB2, consisting solely of an SH2 domain positioned in between two SH3 domains, which is involved in linking receptor tyrosine kinases, including NTRK3, to RAS signaling [96, 179]. Therefore, other SH3-domain containing proteins may associate with these two motifs in the ETV6-NTRK3 chimera. Again, SDM of the ETV6 central domain or of these PXXP regions may diminish the degree transformation of fusion-expressing NIH3T3 cells.

We propose to identify potential ETV6-NTRK3-interacting molecules of a novel signal transduction pathway by using the well-established yeast two-hybrid genetic screen [180]. This system has been successfully applied to detect a large variety of protein-protein interactions, including those involved in receptor PTK signaling [181]. Furthermore, the identification of differential gene expression patterns using differential-display PCR (DD-PCR) and representational difference analysis (RDA) may give insights into the mechanisms of ETV6-NTRK3-mediated transformation. These techniques may identify genes that are directly up- or down- regulated in response to fusion-gene expression. DD-PCR is a powerful tool for analyzing differences in mRNA expression in related cell populations [182, 183]. 78

RDA, on the other hand, depends on a process of subtraction coupled to amplification such that commonly expressed sequences are eliminated leaving only differentially expressed sequences to be amplified [184]. RDA has been used successfully in identifying differentially expressed genes in tumour cells expressing the Ewing-sarcoma specific gene fusion, EWS-FLI1 [185]. 79

CHAPTER 5

SUMMARY AND CONCLUSIONS

All types of cancer cells are believed to originate from genetic mutations in normal cells. Whether these mutations are caused by viruses, exposure to environmental agents, or mistakes introduced during replication, the chromosome complement in these tumourigenic cells is altered. Damage suffered by DNA can be subtle; quite often, point mutations and base-pair insertions or deletions involve only one nucleotide. On the other hand, gene amplification, duplication, and chromosomal translocations may involve large regions of a chromosome. Such large aberrations can readily be detected by routine cytogenetic analysis; however, minute changes may only be detectable by molecular genetic strategies. These facts gain significance with the realization that different neoplasms are associated with specific genetic abnormalities. The characterization of the disrupted chromosomes and genes can therefore provide new information for cancer diagnosis and prognosis, and elucidate the mechanisms involved in tumourigenesis.

CFS is the second-most common soft-tissue tumour in children and it is the most common soft tissue sarcoma in infants. This lesion is unique among human sarcomas because it is associated with a good prognosis. Because CFS can be successfully treated by surgery alone but is invariably fatal if left untreated, an accurate initial diagnosis is imperative. Unfortunately, histological differential diagnosis of CFS from ATFS and IFB was nearly impossible prior to the discovery of 80 the t(12;15)(pl3;q25) rearrangement [37] because CFS is virtually indistinguishable from ATFS, an aggressive lesion, as well as benign IFB. Moreover, cytogenetic analyses of CFS cases typically revealed only nonrandom gains of chromosomes 8,

11,17, and 20 [38]. Identification of the CFS t(12;15) translocation, and of the ETV6-

NTRK3 gene fusion, has led to the development of an RT-PCR diagnostic screen

[37]. Interestingly, this fusion of ETV6 to NTRK3 was the first report of ETV6 being involved in a human solid tumour [37]. Molecular genetic methods have therefore been employed in order to further characterize the role of the ETV6-NTRK3 fusion protein in oncogenesis.

5.1 ETV6-NTRK3 is a Chimeric Oncoprotein

Our first objective was to verify our hypothesis that ETV6-NTRK3-expression is responsible for CFS pathogenesis. Using retrovirally-mediated gene transfer, we created an NIH3T3 cell line that stably expressed our fusion gene of interest.

Morphologically, these cells appeared dramatically transformed compared with untransformed fibroblasts infected with vector alone. Moreover, the ETV6-NTRK3- expressing cells were able to form macroscopic colonies in soft agar at a level comparable to another oncoprotein, EWS-FLI1. We concluded from these data that

ETV6-NTRK3 encodes a chimeric protein with strong transforming potential in

NIH3T3 cells.

5.2 ETV6 HLH and NTRK3 PTK Domains are Essential for Transformation 81

Because the CFS t(12;15) rearrangement fuses the ETV6 HLH domain to the

NTRK3 PTK domain [37], we predicted that ETV6-NTRK3 functions as a chimeric

PTK. The HLH domain is thought to permit NT-3-independent homodimerization

and subsequent constitutive auto- (or cross-) phosphorylation that activates signal

transduction pathways. We used site-directed as well as deletional mutagenesis in

order to determine the contributions of the HLH and PTK domains to the overall

function of the fusion protein.

Our results indicate that the ETY6 HLH domain is essential for ETV6-

NTRK3-mediated transformation. Moreover, it has been shown that this domain is

required for ETV6-NTRK3 homodimerization and phosphorylation [134]. The

conserved NTRK3 PTK activation-loop tyrosines (ETV6-NTRK3 Y513, Y517, and

Y518), which are involved in initiating autophosphorylation [70, 87, 117, 118], were

studied next. Our data demonstrate that the first activation-loop tyrosine is essential

for transformation; in addition, while both ETV6-NTRK3 Y517 and Y518 are

required for full transformation, there may be some level of redundancy between

these second and third conserved tyrosines. Mutation of the ATP-binding site

completely abrogates tyrosine phosphorylation and transformation. From these and

. other findings [134], we conclude that both the ETV6 HLH and NTRK3 PTK domains

are essential for ETV6-NTRK3 to function as a chimeric oncoprotein.

5.3 PLCy-Binding and Activation Not Required

Of the molecules known to interact with wild-type NTRK3, including SNT

[89, 90], SHC, GRB2, and PI-3K P85 [70], only PLCy [87, 88] was predicted to bind to 82

ETV6-NTRK3, and this has been demonstrated to be the case [134]. However, mutation of ETV6-NTRK3 Y628, the PLCy-binding site [87, 88], failed to abolish

ETV6-NTRK3-mediated transformation even though these mutant proteins failed to bind PLCy. We therefore conclude that the association of PLCy with, and its subsequent phosphorylation by, ETV6-NTRK3 is likely not required for transformation.

5.4 General Comments

The presented molecular genetic findings on the ETV6-NTRK3 fusion protein have helped to elucidate our overall understanding of the pathogenesis of

CFS. First, we have demonstrated that ETV6-NTRK3 expression results in oncogenesis. The fact that production of this fusion protein alone, in NIH3T3 cells, is sufficient for tumourigenesis is significant because it indicates that the CFS t(12;15)(pl3;q25) rearrangement is a key step in the progression towards oncogenesis.

Secondly, the mutagenesis studies confirmed that ETV6-NTRK3 functions as a chimeric PTK although none of the immediate NTRK3-associating molecules appear to be involved. Perhaps ETV6-NTRK3 initiates transformation by recruiting less well characterized or completely novel proteins. Only by first understanding which specific pathways are involved can effective therapies be subsequently developed. Cancer is a genetic disease; therefore, efforts directed at understanding its genetic and molecular basis will ultimately have profound effects on the treatment of patients with cancer. 83

REFERENCES

1. Thompson, Margaret W., Mclnnes, Roderick R., and Willard, Huntington, F. Thompson & Thompson: Genetics in Medicine, 5th Edition. 1991, Philadelphia, W.B. Saunders Company. 365-81.

2. Kattamis, C. A. 1967. Glucose-6-phosphate dehydrogenase deficiency in female heterozygotes and the X-inactivation hypothesis. Acta Paediatr Scand, Suppl(172), 103+.

3. Miller, D. M., Blume, S., Borst, M., Gee, J., Polansky, D., Ray, R., Rodu, B., Shrestha, K., Snyder, R., Thomas, S., and et al. 1990. Oncogenes, malignant transformation, and modern medicine. Am } Med Sci, 300(1), 59-69.

4. Vogelstein, B. and Kinzler, K.W., eds. The Genetic Basis of Human Cancer. 1998, The McGraw-Hill Companies, Inc., New York.

5. Knudson, A. G., Jr. 1971. Mutation and cancer: statistical study of retinoblastoma. Proc Natl Acad Sci USA, 68(4), 820-3.

6. Friend, S. H., Horowitz, J. M., Gerber, M. R., Wang, X. F., Bogenmann, E., Li, F. P., and Weinberg, R. A. 1987. Deletions of a DNA sequence in retinoblastomas and mesenchymal tumors: organization of the sequence and its encoded protein [published erratum appears in Proc Natl Acad Sci U S A 1988 Apr;85(7):2234]. Proc Natl Acad Sci USA, 84(24), 9059-63.

7. Goddard, A. D., Balakier, H., Canton, M., Dunn, J., Squire, J., Reyes, E., Becker, A., Phillips, R. A., and Gallie, B. L. 1988. Infrequent genomic rearrangement and normal expression of the putative RBI gene in retinoblastoma tumors. Mol Cell Biol, 8(5), 2082-8.

8. Dunn, J. M., Phillips, R. A., Becker, A. J., and Gallie, B. L. 1988. Identification of germline and somatic mutations affecting the retinoblastoma gene. Science, 241(4874), 1797-800.

9. Huff, V., Miwa, H., Haber, D. A., Call, K. M., Housman, D., Strong, L. C, and Saunders, G. F. 1991. Evidence for WT1 as a Wilms tumor (WT) gene: intragenic germinal deletion in bilateral WT. Am } Hum Genet, 48(5), 997- 1003.

10. Birch, J. M., Blair, V., Kelsey, A. M., Evans, D. G., Harris, M., Tricker, K. J., and Varley, J. M. 1998. Cancer phenotype correlates with constitutional TP53 genotype in families with the Li-Fraumeni syndrome. Oncogene, 17(9), 1061- 8. 84

11. Cawthon, R. M., Weiss, R., Xu, G. F., Viskochil, D., Culver, M., Stevens, J., Robertson, M., Dunn, D., Gesteland, R., O'Connell, P., and et al. 1990. A major segment of the neurofibromatosis type 1 gene: cDNA sequence, genomic structure, and point mutations [published erratum appears in Cell 1990 Aug 10;62(3):following 608]. Cell, 62(1), 193-201.

12. Wallace, M. R., Marchuk, D. A., Andersen, L. B., Letcher, R., Odeh, H. M., Saulino, A. M., Fountain, J. W., Brereton, A., Nicholson, J., Mitchell, A. L., and et al. 1990. Type 1 neurofibromatosis gene: identification of a large transcript disrupted in three NF1 patients [published erratum appears in Science 1990 Dec 21;250(4988):1749]. Science, 249(4965), 181-6.

13. Tanaka, K., Arif, M., Eguchi, M., Guo, S. X., Hayashi, Y., Asaoku, H., Kyo, T., Dohy, H., and Kamada, N. 1999. Frequent allelic loss of the RB, D13S319 and D13S25 locus in myeloid malignancies with deletion/translocation at 13ql4 of chromosome 13, but not in lymphoid malignancies. Leukemia, 13(9), 1367-73.

14. Chiu, I. M., Reddy, E. P., Givol, D., Robbins, K. C, Tronick, S. R., and Aaronson, S. A. 1984. Nucleotide sequence analysis identifies the human c- sis proto-oncogene as a structural gene for platelet-derived growth factor. Cell, 37(1), 123-9.

15. Johnsson, A., Heldin, C. H., Wasteson, A., Westermark, B., Deuel, T. F., Huang, J. S., Seeburg, P. H., Gray, A., Ullrich, A., Scrace, G., and et al. 1984. The c-sis gene encodes a precursor of the B chain of platelet-derived growth factor. Embo J, 3(5), 921-8.

16. Martin-Zanca, D., Hughes, S. H., and Barbacid, M. 1986. A human oncogene formed by the fusion of truncated tropomyosin and protein tyrosine kinase sequences. Nature, 319(6056), 743-8.

17. Harvey, T.T. 1964. An unidentified virus which causes the rapid production of tumors in mice. Nature, 204, 1104.

18. Schwab, M, Varmus, H.E., Bishop, J.E., and others. 1984. Chromosome localization in normal human cells and neuroblastomas of a gene related to c- myc. Nature, 308, 288-291.

19. Taparowsky, E., Suard, Y., Fasano, O., Shimizu, K., Goldfarb, M., and Wigler, M. 1982. Activation of the T24 bladder carcinoma transforming gene is linked to a single amino acid change. Nature, 300(5894), 762-5.

20. Montgomery, K. T., Biedler, J. L., Spengler, B. A., and Melera, P. W. 1983. Specific DNA sequence amplification in human neuroblastoma cells. Proc Natl Acad Sci USA, 80(18), 5724-8. 85

21. Taub, R., Kirsch, I., Morton, C, Lenoir, G., Swan, D., Tronick, S., Aaronson, S., and Leder, P. 1982. Translocation of the c-myc gene into the immunoglobulin heavy chain locus in human Burkitt lymphoma and murine plasmacytoma cells. Proc Natl Acad Sci USA, 79(24), 7837-41.

22. Hayday, A. C., Gillies, S. D., Saito, H., Wood, C., Wiman, K., Hayward, W. S., and Tonegawa, S. 1984. Activation of a translocated human c-myc gene by an enhancer in the immunoglobulin heavy-chain locus. Nature, 307(5949), 334- 40.

23. Rowley, J.D. 1973. A new consistent chromosomal abnormality in chronic myelogenous leukaemia identified by quinacrine fluorescence and Giemsa staining. Nature, 243, 290-293.

24. McWhirter, J. R. and Wang, J. Y. 1991. Activation of tyrosinase kinase and microfilament-binding functions of c-abl by bcr sequences in bcr/abl fusion proteins. Mol Cell Biol, 11(3), 1553-65.

25. Sorensen, Poul H.B. and Triche, Tim J. 1996. Gene fusions encoding chimaeric transcription factors in solid tumours. Seminars in Cancer Biology, 7,3-14.

26. Delattre, O., Zucman, J., Ploustagel, B., Desmaze, C., Melot, T., Peter, M., Kovar, H., Joubert, I., de Jong, P., Rouleau, G., Aurias, A., and Thomas, G. 1992. Gene fusion with an ETS DNA binding domain caused by chromosome translocation in human cancers. Nature, 359, 162-165.

27. May, W. A., Gishizky, M. L., Lessnick, S. L., Lunsford, L. B., Lewis, B. C., Delattre, O., Zucman, J., Thomas, G., and Denny, C. T. 1993. Ewing sarcoma 11;22 translocation produces a chimeric transcription factor that requires the DNA-binding domain encoded by FLU for transformation. Proc Natl Acad Sci USA, 90(12), 5752-6.

28. Sorensen, P.H.B., Lessnick, S.L., Lopez-Terrada, D., Liu, X.F., Triche, T.J., and Denny, C.T. 1994. A second Ewing's sarcoma translocation, t(21;22), fuses the EWS gene to another ETS-family transcription factor, ERG. Nature Genet, 6(2), 146-151.

29. Giovannini, M., Biegel, J. A., Serra, M., Wang, J. Y., Wei, Y. H., Nycum, L., Emanuel, B. S., and Evans, G. A. 1994. EWS-erg and EWS-Flil fusion transcripts in Ewing's sarcoma and primitive neuroectodermal tumors with variant translocations. / Clin Invest, 94(2), 489-96.

30. Selleri, L., Giovannini, M., Romo, A., Zucman, J., Delattre, O., Thomas, G., and Evans, G. A. 1994. Cloning of the entire FLU gene, disrupted by the 86

Ewing's sarcoma translocation breakpoint on llq24, in a yeast artificial chromosome. Cytogenet Cell Genet, 67(2), 129-36.

31. Dunn, T., Praissman, L., Hagag, N., and Viola, M. V. 1994. ERG gene is translocated in an Ewing's sarcoma cell line. Cancer Genet Cytogenet, 76(1), 19-22.

32. Pizzo, P.A. and Poplack, D.G. Principles and Practice of Pediatric Oncology, 3rd Edition. 1997, Philadelphia, J.B. Lippincott. 1-1350.

33. Enzinger, F.M. and Weiss, S.W. Soft Tissue Tumors. 1995, St. Louis, C.V. Mosby.

34. Scott, S.M., Reiman, H.M., Pritchard, D.J. and Ilstrup, D.M. 1989. Soft tissue fibrosarcoma: a clinicopathologic study of 132 cases. Cancer, 64(4), 925-31.

35. Triche, T.J., Pathology of Pediatric Malignancies, in Principles and Practice of Pediatric Oncology, Pizzo, P.A. and Poplack, D.G., Editors. 1993, J.B. Lippincott, Philadelphia, p. 115-152.

36. Schofield, D.E., Fletcher, J.A., Grier, H.E., and Yunis, E.J. 1994. Fibrosarcoma in infants and children. Application of new techniques. Am f Surg Pathol, 18, 14-24.

37. Knezevich, S. R., McFadden, D. E., Tao, W., Lim, J. F., and Sorensen, P. H. 1998. A novel ETV6-NTRK3 gene fusion in congenital fibrosarcoma. Nat Genet, 18(2), 184-7.

38. Coffin, CM., Jaszcz, W., O'Shea, P., and Dehner, L.P. 1994. So-called congenital-infantile fibrosarcoma: does it exist and what is it? Pediatric Pathology, 14, 133-150.

39. Dhulipal, P. D. 1997. Ets oncogene family. Indian } Exp Biol, 35(4), 315-22.

40. Wasylyk, C, Kerckaert, J.-P., and Wasylyk, B. 1992. A novel modulator domain of Ets transcription factors. Genes Dev, 6(6), 965-974.

41. Wasylyk, B., Hahn, S.L., and Giovane, A. 1993. The ETS family of transcription factors. Eur J Biochem, 211, 7-18.

42. Carroll, M., Tomasson, M. H., Barker, G. F., Golub, T. R., and Gilliland, D. G. 1996. The TEL/platelet-derived growth factor beta receptor (PDGF beta R) fusion in chronic myelomonocytic leukemia is a transforming protein that self-associates and activates PDGF beta R kinase-dependent signaling pathways. Proc Natl Acad Sci USA, 93(25), 14845-50. 87

43. Baens, M., Peeters, P., Guo, C., Aerssens, J., and Marynen, P. 1996. Genomic organization of TEL: the human ETS-variant gene 6. Genome Res, 6(5), 404- 13.

44. Fears, S., Gavin, M., Zhang, D. E., Hetherington, C., Ben-David, Y., Rowley, J. D., and Nucifora, G. 1997. Functional characterization of ETV6 and ETV6/CBFA2 in the regulation of the MCSFR proximal promoter. Proc Natl Acad Sci USA, 94(5), 1949-54.

45. Wang, L. C., Kuo, F., Fujiwara, Y., Gilliland, D. G., Golub, T. R., and Orkin, S. H. 1997. Yolk sac angiogenic defect and intra-embryonic apoptosis in mice lacking the Ets-related factor TEL. Embo J, 16(14), 4374-83.

46. Edel, M. J. 1998. The ETS-related factor TEL is regulated by angiogenic growth factor VEGF in HUVE-cells. Anticancer Res, 18(6A), 4505-9.

47. Wang, L. C., Swat, W., Fujiwara, Y., Davidson, L., Visvader, J., Kuo, F., Alt, F. W., Gilliland, D. G., Golub, T. R, and Orkin, S. H. 1998. The TEL/ETV6 gene is required specifically for hematopoiesis in the bone marrow. Genes Dev, 12(15), 2392-402.

48. Golub, T. R, Barker, G. F., Lovett; M., and Gilliland, D. G. 1994. Fusion of PDGF receptor beta to a novel ets-like gene, tel, in chronic myelomonocytic leukemia with t(5;12) chromosomal translocation. Cell, 77(2), 307-16.

49. Golub, T. R., McLean, T., Stegmaier, K., Carroll, M., Tomasson, M., and Gilliland, D. G. 1996. The TEL gene and human leukemia. Biochim Biophys Acta, 1288(1), M7-10.

50. Hoglund, M., Johansson, B., Pedersen-Bjergaard, J., Marynen, P., and Mitelman, F. 1996. Molecular characterization of 12p abnormalities in hematologic malignancies: deletion of KIP1, rearrangement of TEL, and amplification of CCND2. Blood, 87(1), 324-30.

51. Raimondi, S. C, Shurtleff, S. A., Downing, J. R., Rubnitz, J., Mathew, S., Hancock, M., Pui, C. H., Rivera, G. K., Grosveld, G. C, and Behm, F. G. 1997. 12p abnormalities and the TEL gene (ETV6) in childhood acute lymphoblastic leukemia. Blood, 90(11), 4559-66.

52. Tosi, S., Giudici, G., Mosna, G., Harbott, J., Specchia, G., Grosveld, G., Privitera, E., Kearney, L., Biondi, A., and Cazzaniga, G. 1998. Identification of new partner chromosomes involved in fusions with the ETV6 (TEL) gene in hematologic malignancies. Genes Chromosomes Cancer, 21(3), 223-9.

53. Wlodarska, I., La Starza, R., Baens, M., Dierlamm, J., Uyttebroeck, A., Selleslag, D., Francine, A., Mecucci, C, Hagemeijer, A., Van den Berghe, H., 88

and Marynen, P. 1998. Fluorescence in situ hybridization characterization of new translocations involving TEL (ETV6) in a wide spectrum of hematologic malignancies. Blood, 91(4), 1399-406.

54. Golub, T.R., Barker, G.F., Bohlander, S.K., Hiebert, S.W., Ward, D.C., Bray- Ward, P., Morgan, E., Raimondi, S.C., Rowley, J.D., and Gilliland, D.G. 1995. Fusion of the TEL gene on 12pl3 to the AML1 gene on 21q22 in acute lymphoblastic leukemia. Proc Natl Acad Sci USA, 92, 4917-4921.

55. Romana, S. P., Mauchauffe, M., Le Coniat, M., Chumakov, I., Le Paslier, D., Berger, R., and Bernard, O. A. 1995. The t(12;21) of acute lymphoblastic leukemia results in a tel-AMLl gene fusion. Blood, 85(12), 3662-70.

56. Papadopoulos, P., Ridge, S. A., Boucher, C. A., Stocking, C., and Wiedemann, L. M. 1995. The novel activation of ABL by fusion to an ets-related gene, TEL. Cancer Res, 55(1), 34-8.

57. Buijs, A., Sherr, S., van Baal, S., van Bezouw, S., van der Plas, D., Geurts van Kessel, A., Riegman, P., Lekanne Deprez, R., Zwarthoff, E., Hagemeijer, A., and et al. 1995. Translocation (12;22) (pl3;qll) in myeloproliferative disorders results in fusion of the ETS-like TEL gene on 12pl3 to the MN1 gene on 22qll [published erratum appears in Oncogene 1995 Aug 17;11(4):809]. Oncogene, 10(8), 1511-9.

58. Peeters, P., Raynaud, S. D., Cools, J., Wlodarska, I., Grosgeorge, J., Philip, P., Monpoux, F., Van Rompaey, L., Baens, M., Van den Berghe, H., and Marynen, P. 1997. Fusion of TEL, the ETS-variant gene 6 (ETV6), to the receptor- associated kinase JAK2 as a result of t(9;12) in a lymphoid and t(9;15;12) in a myeloid leukemia. Blood, 90(7), 2535-40.

59. Lacronique, V., Boureux, A., Valle, V. D., Poirel, FL, Quang, C. T., Mauchauffe, M., Berthou, C, Lessard, M., Berger, R., Ghysdael, J., and Bernard, O. A. 1997. A TEL-JAK2 fusion protein with constitutive kinase activity in human leukemia. Science, 278(5341), 1309-12.

60. Suto, Y., Sato, Y., Smith, S. D., Rowley, J. D., and Bohlander, S. K. 1997. A t(6;12)(q23;pl3) results in the fusion of ETV6 to a novel gene, STL, in a B-cell ALL cell line. Genes Chromosomes Cancer, 18(4), 254-68.

61. Peeters, P., Wlodarska, I., Baens, M., Criel, A., Selleslag, D., Hagemeijer, A., Van den Berghe, H., and Marynen, P. 1997. Fusion of ETV6 to MDS1/EVI1 as a result of t(3;12)(q26;pl3) in myeloproliferative disorders. Cancer Res, 57(4), 564-9.

62. Chase, A., Reiter, A., Burci, L., Cazzaniga, G., Biondi, A., Pickard, J., Roberts, I. A., Goldman, J. M., and Cross, N. C. 1999. Fusion of ETV6 to the caudal- 89

related homeobox gene CDX2 in acute myeloid leukemia with the t(12;13)(pl3;ql2). Blood, 93(3), 1025-31.

63. Cools, J., Bilhou-Nabera, C, Wlodarska, I., Cabrol, C, Talmant, P., Bernard, P., Hagemeijer, A., and Marynen, P. 1999. Fusion of a Novel Gene, BTL, to ETV6 in Acute Myeloid Leukemias With a t(4;12)(qll-ql2;pl3). Blood, 94(5), 1820-1824.

64. Yagasaki, F., Jinnai, I., Yoshida, S., Yokoyama, Y., Matsuda, A., Kusumoto, S., Kobayashi, H., Terasaki, H., Ohyashiki, K., Asou, N., Murohashi, I., Bessho, M., and Hirashima, K. 1999. Fusion of TEL/ETV6 to a novel ACS2 in myelodysplastic syndrome and acute myelogenous leukemia with t(5;12)(q31;pl3). Genes Chromosomes Cancer, 26(3), 192-202.

65. Cave, H., Cacheux, V., Raynaud, S., Brunie, G., Bakkus, M., Cochaux, P., Preudhomme, C, Lai, J. L., Vilmer, E., and Grandchamp, B. 1997. ETV6 is the target of chromosome 12p deletions in t(12;21) childhood acute lymphocytic leukemia. Leukemia, 11(9), 1459-64.

66. Barbacid, M. 1995. Structural and functional properties of the TRK family of neurotrophin receptors. Ann N Y Acad Sci, 766, 442-58.

67. Ultsch, M. H., Wiesmann, C, Simmons, L. C, Henrich, J., Yang, M., Reilly, D., Bass, S. H., and de Vos, A. M. 1999. Crystal structures of the neurotrophin- binding domain of TrkA, TrkB and TrkC. / Mol Biol, 290(1), 149-59.

68. Wilks, A.F. 1993. Protein tyrosine kinase growth factor receptors and their ligands in development, differentiation, and cancer. Advances in Cancer Research, 60, 43-73.

69. Clary, D. O., Weskamp, G., Austin, L. R, and Reichardt, L. F. 1994. TrkA cross-linking mimics neuronal responses to nerve growth factor. Mol Biol Cell, 5(5), 549-63.

70. Barbacid, M. 1994. The Trk family of neurotrophin receptors. / Neurobiol, 25(11), 1386-403.

71. Kaplan, D.R. and Miller, F.D. 1997. Signal transduction by the neurotrophin receptors. Current Opinion in Cell Biology, 9(2), 213-21.

72. Mitra, G., Martin-Zanca, D., and Barbacid, M. 1987. Identification and biochemical characterization of p70TRK, product of the human TRK oncogene. Proc Natl Acad Sci USA, 84(19), 6707-11. 90

73. Koizumi, H., Morita, M., Mikami, S., Shibayama, E., and Uchikoshi, T. 1998. Immunohistochemical analysis of TrkA neurotrophin receptor expression in human non-neuronal carcinomas. Pathol Int, 48(2), 93-101.

74. Lamballe, F., Klein, R., and Barbacid, M. 1991. trkC, a new member of the trk family of tyrosine protein kinases, is a receptor for neurotrophin-3. Cell, 66(5), 967-79.

75. McGregor, L. M., Baylin, S. B., Griffin, C. A., Hawkins, A. L., and Nelkin, B. D. 1994. Molecular cloning of the cDNA for human TrkC (NTRK3), chromosomal assignment, and evidence for a splice variant. Genomics, 22(2), 267-72.

76. Conover, J. C. and Yancopoulos, G. D. 1997. Neurotrophin regulation of the developing nervous system: analyses of knockout mice. Rev Neurosci, 8(1), 13-27.

77. Ebadi, M., Bashir, R. M., Heidrick, M. L., Hamada, F. M., Refaey, H. E., Hamed, A., Helal, G., Baxi, M. D., Cerutis, D. R., and Lassi, N. K. 1997. Neurotrophins and their receptors in nerve injury and repair. Neurochem Int, 30(4-5), 347- 74.

78. Ichaso, N., Rodriguez, R. E., Martin-Zanca, D., and Gonzalez-Sarmiento, R. 1998. Genomic characterization of the human trkC gene. Oncogene, 17(14), 1871-5.

79. Mudo, G., Persson, H., Timmusk, T., Funakoshi, H., Bindoni, M., and Belluardo, N. 1993. Increased expression of trkB and trkC messenger RNAs in the rat forebrain after focal mechanical injury. Neuroscience, 57(4), 901-12.

80. Svensson, T., Ryden, M., Schilling, F. H., Dominici, C, Sehgal, R., Ibanez, C. F., and Kogner, P. 1997. Coexpression of mRNA for the full-length neurotrophin receptor trk-C and trk-A in favourable neuroblastoma. Eur } Cancer, 33(12), 2058-63.

81. Jing, S., Tapley, P., and Barbacid, M. 1992. Nerve growth factor mediates signal transduction through trk homodimer receptors. Neuron, 9(6), 1067-79.

82. Lemmon, M. A. and Schlessinger, J. 1994. Regulation of signal transduction and signal diversity by receptor oligomerization. Trends Biochem Sci, 19(11), 459-63.

83. Barbacid, M. 1995. Neurotrophic factors and their receptors. Curr Opin Cell Biol, 7(2), 148-55. 91

84. Guiton, M., Gunn-Moore, F. J., Glass, D. J., Geis, D. R., Yancopoulos, G. D., and Tavare, J. M. 1995. Naturally occurring tyrosine kinase inserts block high affinity binding of phospholipase C gamma and She to TrkC and neurotrophin-3 signaling. J Biol Chem, 270(35), 20384-90.

85. Hallberg, B., Ashcroft, M, Loeb, D. M., Kaplan, D. R, and Downward, J. 1998. Nerve growth factor induced stimulation of Ras requires Trk interaction with She but does not involve phosphoinositide 3-OH kinase. Oncogene, 17(6), 691-7.

86. Obermeier, A., Lammers, R., Wiesmuller, K. PL, Jung, G., Schlessinger, J., and Ullrich, A. 1993. Identification of Trk binding sites for SHC and phosphatidylinositol 3'- kinase and formation of a multimeric signaling complex. / Biol Chem, 268(31), 22963-6.

87. Lamballe, F., Tap ley, P., and Barbacid, M. 1993. trkC encodes multiple neurotrophin-3 receptors with distinct biological properties and substrate specificities. Embo J, 12(8), 3083-94.

88. Obermeier, A., Halfter, H., Wiesmuller, K. H., Jung, G., Schlessinger, J., and Ullrich, A. 1993. Tyrosine 785 is a major determinant of Trk—substrate interaction. Embo }, 12(3), 933-41.

89. Rabin, S. J., Cleghon, V., and Kaplan, D. R. 1993. SNT, a differentiation- specific target of neurotrophic factor-induced tyrosine kinase activity in neurons and PC12 cells. Mol Cell Biol, 13(4), 2203-13.

90. Peng, X., Greene, L. A., Kaplan, D. R., and Stephens, R. M. 1995. Deletion of a conserved juxtamembrane sequence in Trk abolishes NGF- promoted neuritogenesis. Neuron, 15(2), 395-406.

91. Qian, X., Riccio, A., Zhang, Y., and Ginty, D. D. 1998. Identification and characterization of novel substrates of Trk receptors in developing neurons. Neuron, 21(5), 1017-29.

92. Yokouchi, M., Suzuki, R., Masuhara, M., Komiya, S., Inoue, A., and Yoshimura, A. 1997. Cloning and characterization of APS, an adaptor molecule containing PH and SH2 domains that is tyrosine phosphorylated upon B-cell receptor stimulation. Oncogene, 15(1), 7-15.

93. Suen, K. L., Bustelo, X. R., Pawson, T., and Barbacid, M. 1993. Molecular cloning of the mouse grb2 gene: differential interaction of the Grb2 adaptor protein with epidermal growth factor and nerve growth factor receptors. Mol Cell Biol, 13(9), 5500-12. 92

94. Pelicci, G., Lanfrancone, L., Grignani, F., McGlade, J., Cavallo, F., Forni, G., Nicoletti, I., Pawson, T., and Pelicci, P. G. 1992. A novel transforming protein (SHC) with an SH2 domain is implicated in mitogenic signal transduction. Cell, 70(1), 93-104.

95. Bonfini, L., Migliaccio, E., Pelicci, G., Lanfrancone, L., and Pelicci, P. G. 1996. Not all She's roads lead to Ras. Trends Biochem Sci, 21(7), 257-61.

96. Downward, J. 1996. Control of ras activation. Cancer Surv, 27, 87-100.

97. Egan, S.E., Giddings, B.W., Brooks, M.W., Buday, L., Sizeland, A.M., and Weinberg, R.A. 1993. Association of Sos Ras exchange protein with Grb2 is implicated in tyrosine kinase signal transduction and transformation. Nature, 363, 45-51.

98. Noda, M., Ko, M., Ogura, A., Liu, D. G., Amano, T., Takano, T., and Ikawa, Y. 1985. Sarcoma viruses carrying ras oncogenes induce differentiation- associated properties in a neuronal cell line. Nature, 318(6041), 73-5.

99. Qiu, M. S. and Green, S. H. 1991. NGF and EGF rapidly activate p21ras in PC12 cells by distinct, convergent pathways involving tyrosine phosphorylation. Neuron, 7(6), 937-46.

100. Schanen-King, C, Nel, A., Williams, L. K., and Landreth, G. 1991. Nerve growth factor stimulates the tyrosine phosphorylation of MAP2 kinase in PC12 cells. Neuron, 6(6), 915-22.

101. Muroya, K., Hattori, S., and Nakamura, S. 1992. Nerve growth factor induces rapid accumulation of the GTP-bound form of p21ras in rat pheochromocytoma PC12 cells. Oncogene, 7(2), 277-81.

102. Thomas, S. M., DeMarco, M., D'Arcangelo, G., Halegoua, S., and Brugge, J. S. 1992. Ras is essential for nerve growth factor- and phorbol ester-induced tyrosine phosphorylation of MAP kinases. Cell, 68(6), 1031-40.

103. Wood, K. W., Qi, H., D'Arcangelo, G., Armstrong, R. C, Roberts, T. M., and Halegoua, S. 1993. The cytoplasmic raf oncogene induces a neuronal phenotype in PC12 cells: a potential role for cellular raf kinases in neuronal growth factor signal transduction. Proc Natl Acad Sci USA, 90(11), 5016-20.

104. Carpenter, C. L. and Cantley, L. C. 1996. Phosphoinositide 3-kinase and the regulation of cell growth. Biochim Biophys Acta, 1288(1), Mll-6.

105. Klinghoffer, R. A., Duckworth, B., Valius, M., Cantley, L., and Kazlauskas, A. 1996. Platelet-derived growth factor-dependent activation of phosphatidylinositol 3-kinase is regulated by receptor binding of SH2- 93

domain-containing proteins which influence Ras activity. Mol Cell Biol, 16(10), 5905-14.

106. Franke, T. F., Kaplan, D. R., Cantley, L. C, and Toker, A. 1997. Direct regulation of the Akt proto-oncogene product by phosphatidylinositol-3,4- bisphosphate [see comments]. Science, 275(5300), 665-8.

107. Dudek, H., Datta, S. R, Franke, T. F., Birnbaum, M. J., Yao, R., Cooper, G. M., Segal, R. A., Kaplan, D. R., and Greenberg, M. E. 1997. Regulation of neuronal survival by the serine-threonine protein kinase Akt [see comments]. Science, 275(5300), 661-5.

108. Datta, S. R., Dudek, H., Tao, X., Masters, S., Fu, H., Gotoh, Y., and Greenberg, M. E. 1997. Akt phosphorylation of BAD couples survival signals to the cell- intrinsic death machinery. Cell, 91(2), 231-41.

109. Berridge, M. J. 1993. Inositol trisphosphate and calcium signalling. Nature, 361(6410), 315-25.

110. Cockcroft, S. and Thomas, G. M. 1992. Inositol-lipid-specific phospholipase C isoenzymes and their differential regulation by receptors. Biochem f, 288(Pt 1), 1-14.

111. Lee, S. B. and Rhee, S. G. 1995. Significance of PIP2 hydrolysis and regulation of phospholipase C isozymes. Curr Opin Cell Biol, 7(2), 183-9.

112. Nishizuka, Y. 1995. Protein kinase C and lipid signaling for sustained cellular responses. Faseb f, 9(7), 484-96.

113. Rhee, S. G. and Choi, K. D. 1992. Multiple forms of phospholipase C isozymes and their activation mechanisms. Adv Second Messenger Phosphoprotein Res, 26, 35-61.

114. Greene, L. A. and Kaplan, D. R. 1995. Early events in neurotrophin signalling via Trk and p75 receptors. Curr Opin Neurobiol, 5(5), 579-87.

115. Ong, S. H., Goh, K. C, Lim, Y. P., Low, B. C, Klint, P., Claesson-Welsh, L., Cao, X., Tan, Y. H., and Guy, G. R. 1996. Sucl-associated neurotrophic factor target (SNT) protein is a major FGF- stimulated tyrosine phosphorylated 90-kDa protein which binds to the SH2 domain of GRB2. Biochem Biophys Res Commun, 225(3), 1021-6.

116. Kouhara, H., Hadari, Y. R., 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(5), 693-702. 94

117. Tsoulfas, P., Soppet, D., Escandon, E., Tessarollo, L., Mendoza-Ramirez, J. L., Rosenthal, A., Nikolics, K., and Parada, L. F. 1993. The rat trkC locus encodes multiple neurogenic receptors that exhibit differential response to neurotrophin-3 in PC12 cells. Neuron, 10(5), 975-90.

118. Valenzuela, D. M., Maisonpierre, P. C, Glass, D. J., Rojas, E., Nunez, L., Kong, Y., Gies, D. R., Stitt, T. N., Ip, N. Y., and Yancopoulos, G. D. 1993. Alternative forms of rat TrkC with different functional capabilities. Neuron, 10(5), 963-74.

119. Mitra, G. 1991. Mutational analysis of conserved residues in the tyrosine kinase domain of the human trk oncogene. Oncogene, 6(12), 2237-41.

120. Freshney, R. Ian. Culture of Animal Cells: A Manual of Basic Technique. 1994, New York, Wiley-Liss.

121. Mann, R., Mulligan, R. C, and Baltimore, D. 1983. Construction of a retrovirus packaging mutant and its use to produce helper-free defective retrovirus. Cell, 33(1), 153-9.

122. Miller, A. D. and Buttimore, C. 1986. Redesign of retrovirus packaging cell lines to avoid recombination leading to helper virus production. Mol Cell Biol, 6(8), 2895-902.

123. Morgenstern, J. P. and Land, H. 1990. Advanced mammalian gene transfer: high titre retroviral vectors with multiple drug selection markers and a complementary helper-free packaging cell line. Nucleic Acids Res, 18(12), 3587-96.

124. Miller, A. D. and Chen, F. 1996. Retrovirus packaging cells based on 10A1 murine leukemia virus for production of vectors that use multiple receptors for cell entry. / Virol, 70(8), 5564-71.

125. Grez, M., Akgun, E., Hilberg, F., and Ostertag, W. 1990. Embryonic stem cell virus, a recombinant murine retrovirus with expression in embryonic stem cells. Proc Natl Acad Sci USA, 87(23), 9202-6.

126. Miller, A. D. and Rosman, G. J. 1989. Improved retroviral vectors for gene transfer and expression. Biotechniques, 7(9), 980-2, 984-6, 989-90.

127. Hawley, R. G., Lieu, F. H., Fong, A. Z., and Hawley, T. S. 1994. Versatile retroviral vectors for potential use in gene therapy. Gene Ther, 1(2), 136-8.

128. Pear, W. S., Nolan, G. P., Scott, M. L., and Baltimore, D. 1993. Production of high-titer helper-free retroviruses by transient transfection. Proc Natl Acad Sci USA, 90(18), 8392-6. 95

129. May, W.A., Lessnick, S.L., Braun, B.S., Klemsz, M., Lewis, B.C., Lunsford, L.B., Hromas, R., and Denny, CT. 1993. The Ewing's sarcoma EWS/FLI-1 fusion gene encodes a more potent transcriptional activator and is a more powerful transforming gene than FLI-1. Mol Cell Biol, 13, 7393-7398.

130. Knezevich, S. R., Garnett, M. J., Pysher; T. J., Beckwith, J. B., Grundy, P. E., and Sorensen, P. H. 1998. ETV6-NTRK3 gene fusions and trisomy 11 establish a histogenetic link between mesoblastic nephroma and congenital fibrosarcoma. Cancer Res, 58(22), 5046-8.

131. Rubin, B. P., Chen, C. J., Morgan, T. W., Xiao, S., Grier, H. E., Kozakewich, H. P., Perez-Atayde, A. R., and Fletcher, J. A. 1998. Congenital mesoblastic nephroma t(12;15) is associated with ETV6-NTRK3 gene fusion: cytogenetic and molecular relationship to congenital (infantile) fibrosarcoma. Am } Pathol, 153(5), 1451-8.

132. Eguchi, M., Eguchi-Ishimae, M., Tojo, A., Morishita, K., Suzuki, K., Sato, Y.,' Kudoh, S., Tanaka, K., Setoyama, M, Nagamura, F., Asano, S., and Kamada, N. 1999. Fusion of ETV6 to neurotrophin-3 receptor TRKC in acute myeloid leukemia with t(12;15)(pl3;q25). Blood, 93(4), 1355-63.

133. Golub, T. R., Goga, A., Barker, G. F., Afar, D. E., McLaughlin, J., Bohlander, S. K., Rowley, J. D., Witte, O. N., and Gilliland, D. G. 1996. Oligomerization of the ABL tyrosine kinase by the Ets protein TEL in human leukemia. Mol Cell Biol, 16(8), 4107-16.

134. Wai, D.H., Knezevich, S.R., Lucas, T., Jansen, B., Kay, R.J., and Sorensen, P.H.B. In press. The ETV6-NTRK3 gene fusion encodes a chimeric protein tyrosine kinase that transforms NIH 3T3 cells. Oncogene.

135. Sakai, I., Nabell, L., and Kraft, A. S. 1995. Signal transduction by a CD16/CD7/Jak2 fusion protein. / Biol Chem, 270(31), 18420-7.

136. Frank, S. J., Yi, W., Zhao, Y., Goldsmith, J. F., Gilliland, G., Jiang, J., Sakai, I., and Kraft, A. S. 1995. Regions of the JAK2 tyrosine kinase required for coupling to the growth hormone receptor. / Biol Chem, 270(24), 14776-85.

137. Ho, J. M., Beattie, B. K., Squire, J. A., Frank, D. A., and Barber, D. L. 1999. Fusion of the ets transcription factor TEL to Jak2 results in constitutive Jak- Stat signaling. Blood, 93(12), 4354-64.

138. Cunningham, M. E., Stephens, R. M., Kaplan, D. R., and Greene, L. A. 1997. Autophosphorylation of activation loop tyrosines regulates signaling by the TRK nerve growth factor receptor. / Biol Chem, 272(16), 10957-67. 96

139. Holgado-Madruga, M., Moscatello, D. K., Emlet, D. R., Dieterich, R., and Wong, A. J. 1997. Grb2-associated binder-1 mediates phosphatidylinositol 3- kinase activation and the promotion of cell survival by nerve growth factor. Proc Natl Acad Sci USA, 94(23), 12419-24.

140. Rothberg, K. G., Heuser, J. E., Donzell, W. C., Ying, Y. S., Glenney, J. R., and Anderson, R. G. 1992. Caveolin, a protein component of caveolae membrane coats. Cell, 68(4), 673-82.

141. Glenney, J. R., Jr. and Soppet, D. 1992. Sequence and expression of caveolin, a protein component of caveolae plasma membrane domains phosphorylated on tyrosine in Rous sarcoma virus-transformed fibroblasts. Proc Natl Acad Sci US A, 89(21), 10517-21.

142. Bilderback, T. R., Gazula, V. R, Lisanti, M. P., and Dobrowsky, R. T. 1999. Caveolin interacts with Trk A and p75(NTR) and regulates neurotrophin signaling pathways. / Biol Chem, 274(1), 257-63.

143. Poirel, H., Lacronique, V., Mauchauffe, M., Le Coniat, M., Raffoux, E., Daniel, M. T., Erickson, P., Drabkin, H., MacLeod, R. A., Drexler, H. G., Ghysdael, J., Berger, R., and Bernard, O. A. 1998. Analysis of TEL proteins in human leukemias. Oncogene, 16(22), 2895-903.

144. Chakrabarti, S. R. and Nucifora, G. 1999. The Leukemia-Associated Gene TEL Encodes a Transcription Repressor Which Associates with SMRT and mSin3A. Biochem Biophys Res Commun, 264(3), 871-877.

145. Poirel, H., Oury, C, Carron, C, Duprez, E., Laabi, Y., Tsapis, A., Romana, S. P., Mauchauffe, M., Le Coniat, M., Berger, R., Ghysdael, J., and Bernard, O. A. 1997. The TEL gene products: nuclear phosphoproteins with DNA binding properties. Oncogene, 14(3), 349-57.

146. Kwiatkowski, B. A., Bastian, L. S., Bauer, T. R., Jr., Tsai, S., Zielinska- Kwiatkowska, A. G., and Hickstein, D. D. 1998. The ets family member Tel binds to the Fli-1 oncoprotein and inhibits its transcriptional activity. / Biol Chem, 273(28), 17525-30.

147. Chakrabarti, S. R, Sood, R, Ganguly, S., Bohlander, S., Shen, Z., and Nucifora, G. 1999. Modulation of TEL transcription activity by interaction with the ubiquitin-conjugating enzyme UBC9. Proc Natl Acad Sci USA, 96(13), 7467-72.

148. Firestein, R. and Feuerstein, N. 1998. Association of activating transcription factor 2 (ATF2) with the ubiquitin-conjugating enzyme hUBC9. Implication of the ubiquitin/proteasome pathway in regulation of ATF2 in T cells. / Biol Chem, 273(10), 5892-902. 97

149. Muller, S., Matunis, M. J., and Dejean, A. 1998. Conjugation with the ubiquitin-related modifier SUMO-1 regulates the partitioning of PML within the nucleus. Embo J, 17(1), 61-70.

150. Desterro, J. M., Rodriguez, M. S., and Hay, R. T. 1998. SUMO-1 modification of IkappaBalpha inhibits NF-kappaB activation. Mol Cell, 2(2), 233-9.

151. Mahajan, R., Delphin, C, Guan, T., Gerace, L., and Melchior, F. 1997. A small ubiquitin-related polypeptide involved in targeting RanGAPl to nuclear pore complex protein RanBP2. Cell, 88(1), 97-107.

152. Kim, N.W., Piatyszek, M.A., Prowse, K.R., Harley, C.B., West, M.D., Ho, P.L.C., Coviello, G.M., Wright, W.E., Weinrich, S.L., and Shay, J.W. 1994. Specific association of human telomerase activity with immortal cells and cancer. Science, 266(5193), 2011-2015.

153. Shay, J. W. and Bacchetti, S. 1997. A survey of telomerase activity in human cancer. Eur J Cancer, 33(5), 787-91.

154. Harley, C. B., Kim, N. W., Prowse, K. R., Weinrich, S. L., Hirsch, K. S., West, M. D., Bacchetti, S., Hirte, H. W., Counter, C. M., Greider, C. W., and et al. 1994. Telomerase, cell immortality, and cancer. Cold Spring Harb Symp Quant Biol, 59, 307-15.

155. Sager, R. 1991. Senescence as a mode of tumor suppression. Environ Health Perspect, 93, 59-62.

156. Stevenson, M. and Volsky, D. J. 1986. Activated v-myc and v-ras oncogenes do not transform normal human lymphocytes. Mol Cell Biol, 6(10), 3410-7.

157. Hahn, W. C, Counter, C. M., Lundberg, A. S., Beijersbergen, R. L., Brooks, M. W., and Weinberg, R. A. 1999. Creation of human tumour cells with defined genetic elements [see comments]. Nature, 400(6743), 464-8.

158. Spencer, D. M., Wandless, T. J., Schreiber, S. L., and Crabtree, G. R. 1993. Controlling signal transduction with synthetic ligands [see comments]. Science, 262(5136), 1019-24.

159. Travis, J. 1993. Making molecular matches in the cell [news; comment]. Science, 262(5136), 989.

160. Bierer, B. E., Mattila, P. S., Standaert, R. F., Herzenberg, L. A., Burakoff, S. J., Crabtree, G., and Schreiber, S. L. 1990. Two distinct signal transmission pathways in T lymphocytes are inhibited by complexes formed between an 98

immunophilin and either FK506 or rapamycin. Proc Natl Acad Sci USA, 87(23), 9231-5.

161. Rosen, M.K., Yang, D., Martin, P.K., and Scheiber, S.L. 1993. /. Am. Chem. Soc, 115, 821.

162. Yang, D., Rosen, M.K., and Schreiber, S.L. 1993. /. Am. Chem. Soc, 115, 819.

163. Schreiber, S. L. 1991. Chemistry and biology of the immunophilins and their immunosuppressive ligands. Science, 251(4991), 283-7.

164. Liu, J., Farmer, J. D., Jr., Lane, W. S., Friedman, J., Weissman, I., and Schreiber, S. L. 1991. Calcineurin is a common target of cyclophilin-cyclosporin A and FKBP- FK506 complexes. Cell, 66(4), 807-15.

165. Clipstone, N. A. and Crabtree, G. R. 1992. Identification of calcineurin as a key signalling enzyme in T- lymphocyte activation. Nature, 357(6380), 695-7.

166. O'Keefe, S. J., Tamura, J., Kincaid, R. L., Tocci, M. J., and O'Neill, E. A. 1992. FK-506- and CsA-sensitive activation of the interleukin-2 promoter by calcineurin. Nature, 357(6380), 692-4.

167. Schreiber, S. L. and Crabtree, G. R. 1992. The mechanism of action of cyclosporin A and FK506. Immunol Today, 13(4), 136-42.

168. Cross, F. R., Garber, E. A., Pellman, D., and Hanafusa, H. 1984. A short sequence in the p60src N terminus is required for p60src myristylation and membrane association and for cell transformation. Mol Cell Biol, 4(9), 1834- 42.

169. Spencer, D. M., Graef, I., Austin, D. J., Schreiber, S. L., and Crabtree, G. R. 1995. A general strategy for producing conditional alleles of Src-like tyrosine kinases. Proc Natl Acad Sci USA, 92(21), 9805-9.

170. Holsinger, L. J., Spencer, D. M., Austin, D. J., Schreiber, S. L., and Crabtree, G. R. 1995. Signal transduction in T lymphocytes using a conditional allele of Sos. Proc Natl Acad Sci USA, 92(21), 9810-4.

171. Spencer, D. M., Belshaw, P. J., Chen, L., Ho, S. N., Randazzo, F., Crabtree, G. R., and Schreiber, S. L. 1996. Functional analysis of Fas signaling in vivo using synthetic inducers of dimerization. Curr Biol, 6(7), 839-47.

172. Freiberg, R. A., Spencer, D. M., Choate, K. A., Peng, P. D., Schreiber, S. L., Crabtree, G. R., and Khavari, P. A. 1996. Specific triggering of the Fas signal transduction pathway in normal human keratinocytes. } Biol Chem, 271(49), 31666-9. 99

173. Blau, C. A., Peterson, K. R., Drachman, J. G., and Spencer, D. M. 1997. A proliferation switch for genetically modified cells. Proc Natl Acad Sci USA, 94(7), 3076-81.

174. Jin, L., Asano, H., and Blau, C. A. 1998. Stimulating cell proliferation through the pharmacologic activation of c-kit. Blood, 91(3), 890-7.

175. Jin, L., Siritanaratkul, N., Emery, D. W., Richard, R. E., Kaushansky, K., Papayannopoulou, T., and Blau, C. A. 1998. Targeted expansion of genetically modified bone marrow cells. Proc Natl Acad Sci USA, 95(14), 8093-7.

176. Amara, J. F., Clackson, T., Rivera, V. M., Guo, T., Keenan, T., Natesan, S., Pollock, R., Yang, W., Courage, N. L., Holt, D. A,, and Gilman, M. 1997. A versatile synthetic dimerizer for the regulation of protein-protein interactions. Proc Natl Acad Sci USA, 94(20), 10618-23.

177. Keenan, T., Yaeger, D. R., Courage, N. L., Rollins, C. T., Pavone, M. E., Rivera, V. M., Yang, W., Guo, T., Amara, J. F., Clackson, T., Gilman, M., and Holt, D. A. 1998. Synthesis and activity of bivalent FKBP12 ligands for the regulated dimerization of proteins. Bioorg Med Chem, 6(8), 1309-35.

178. Clackson, T., Yang, W., Rozamus, L. W., Hatada, M., Amara, J. F., Rollins, C. T., Stevenson, L. F., Magari, S. R., Wood, S. A., Courage, N. L., Lu, X., Cerasoli, F., Jr., Gilman, M., and Holt, D. A. 1998. Redesigning an FKBP-ligand interface to generate chemical dimerizers with novel specificity. Proc Natl Acad Sci USA, 95(18), 10437-42.

179. Lowenstein, E. J., Daly, R. J., Batzer, A. G., Li, W., Margolis, B., Lammers, R., Ullrich, A., Skolnik, E. Y., Bar-Sagi, D., and Schlessinger, J. 1992. The SH2 and SH3 domain-containing protein GRB2 links receptor tyrosine kinases to ras signaling. Cell, 70(3), 431-42.

180. Fields, S. and Song, O. 1989. A novel genetic system to detect protein-protein interactions. Nature, 340(6230), 245-6.

181. Hock, B., Bohme, B., Karn, T., Feller, S., Rubsamen-Waigmann, H., and Strebhardt, K. 1998. Tyrosine-614, the major autophosphorylation site of the receptor tyrosine kinase HEK2, functions as multi-docking site for SH2- domain mediated interactions. Oncogene, 17(2), 255-60.

182. Liang, P. and Pardee, A.B. 1992. Differential display of eukaryotic messenger RNA by means of the polymerase chain reaction. Science, 257(5072), 967-971. 100

183. Liang, P., Averboukh, L., Keyomarsi, K., Sager, R., and Pardee, A.B. 1992. Differential display and cloning of messenger RNAs from human breast cancer versus mammary epithelial cells. Cancer Res, 52(24), 6966-6968.

184. Hubank, M. and Schatz, D. G. 1994. Identifying differences in mRNA expression by representational difference analysis of cDNA. Nucleic Acids Res, 22(25), 5640-8.

185. Braun, B.S., Frieden, R, Lessnick, S.L., May, W.A., and Denny, C.T. 1995. Identification of target genes for the Ewing's sarcoma EWS/FLI fusion protein by representational difference analysis. Mol Cell Biol, 15(8), 4623-4630.