Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Studies of signaling pathways that regulate Ewing tumour cell growth in vitro Lawlor, Elizabeth Rachel 2002

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

Item Metadata

Download

Media
831-ubc_2002-731936.pdf [ 17.87MB ]
Metadata
JSON: 831-1.0090557.json
JSON-LD: 831-1.0090557-ld.json
RDF/XML (Pretty): 831-1.0090557-rdf.xml
RDF/JSON: 831-1.0090557-rdf.json
Turtle: 831-1.0090557-turtle.txt
N-Triples: 831-1.0090557-rdf-ntriples.txt
Original Record: 831-1.0090557-source.json
Full Text
831-1.0090557-fulltext.txt
Citation
831-1.0090557.ris

Full Text

STUDIES OF SIGNALING PATHWAYS THAT REGULATE EWING TUMOUR CELL GROWTH  IN  VITRO.  by  ELIZABETH R A C H E L L A W L O R B.Sc. McMaster University, 1986 M . D . McMaster University, 1989  A THESIS SUBMITTED IN PARTIAL F U L F I L L M E N T OF T H E REQUIREMENTS FOR T H E D E G R E E OF DOCTOR OF PHILOSOPHY in  THE F A C U L T Y OF G R A D U A T E STUDIES (Department of Pathology and Laboratory Medicine) W e accept this thesis as conformiru^jDjhej^q^ired  standard  T H E UNIVERSITY OF BRITISH C O L U M B I A  2001 © E L I Z A B E T H R A C H E L L A W L O R , 2001  In  presenting  degree freely  this  at the  thesis  in  partial  fulfilment  of  University of  British  Columbia,  I agree  available for reference  copying  of  department  this or  publication of  and study.  thesis for scholarly by  this  his  or  her  the  purposes  representatives.  may be It  thesis for financial gain shall not  pA-Tii^ur-c^  j L*S>  The University of British Columbia Vancouver, Canada  Date  DE-6 (2/88)  IT  2O0j  that  I further agree  permission.  Department of  requirements  rUHMCiMcr  is  that  the  an  advanced  Library shall make it  permission for extensive  granted  by the  understood be  for  allowed  that without  head  of  my  copying  or  my written  11  ABSTRACT The E w i n g family of peripheral primitive neuroectodermal tumours (ET or p P N E T ) comprises a genetically related group of bone and soft tissue tumours that primarily affects children and adolescents.  A l l members of this tumour family are characterized by tumour  specific translocations involving the EWS gene from chromosome l l q 2 3 and a second gene contributed by a variety of chromosomal partners. In all cases, the partner gene is a member of the E T S family of transcription factors and the resultant EWS-ETS  gene fusions encode  novel chimeric proteins comprised of the EWS-amino-terminal domain and an E T S - D N A binding domain. These fusion proteins are oncogenic and the accumulated evidence suggests that they are the primary pathologic lesions contributing to the development of E T . Despite its genetic characterization, little is known about the mechanisms of E W S - E T S mediated oncogenesis specifically, or in fact, the biologic features that contribute to the malignant nature of E T cells. W e have endeavoured to study proliferative signaling in E T cells in an effort to better characterize the biology of these aggressive and often fatal tumours. Using differential-display P C R , our laboratory had previously identified differential expression of the human gastrin-releasing peptide ( G R P ) gene among EWS-ETS lines.  expressing tumour cell  Further studies confirmed that 100% of E T cell lines and approximately 50% of  primary E T samples express the GRP gene and the gene encoding its receptor. Protein and functional studies demonstrated that the mature G R P peptide is secreted by E T cells and that G R P acts as an autocrine growth factor in E T both in vitro and in vivo.  The proliferative  pathways initiated in response to G R P stimulation are still not completely understood but are known to be both cell-type and situation specific.  Our attempts at characterizing G R P -  mediated proliferative signaling in E T suggest that the effects of this peptide may not be  iii  primarily mitogenic and we hypothesize that involvement of cytoskeletal proteins and differentiation pathways may be more important to the autocrine growth-factor capability of G R P in E T cells.  Furthermore, studies of E T cell proliferation grown as multi-cellular  spheroids indicate that in vivo growth and proliferation may be better represented by this anchorage-independent  cell culture model.  In comparing adherent and anchorage-  independent proliferation, we also observed that fundamental differences exist in the regulation of cyclin D I between these two models. Moreover, the P I 3 K — A K T pathway was shown to be of key importance to the regulation of both c y c l i n D I expression and proliferation in E T cells.  Our results suggest that an autocrine growth factor pathway  mediated by G R P exists in E T and that proliferative signaling in these tumours is critically dependent on cell-cell and/or cell-matrix adhesion and the P I 3 K — A K T pathway.  IV  TABLE OF CONTENTS  ABSTRACT T A B L E OF CONTENTS  iv  LIST OF TABLES  ix  LIST OF FIGURES  x  LIST OF ABBREVIATIONS  xii  ACKNOWLEDGEMENTS  xiv  CHAPTER I:  INTRODUCTION  1  1.1 7.3  1 2  1.3  Synopsis and rationale for thesis The E w i n g Family of Peripheral Primitive Neuroectodermal Tumours 1.2.1 Clinical & Pathologic Features 1.2.2 Genetics of the E w i n g Tumour Family 1.2.2.1 EWS-ETS Gene Fusions 1.2.2.2 E W S - E T S Chimeric Oncoproteins 1.2.2.3 Mechanisms of E W S - E T S Mediated Oncogenesis 1.2.2.4 Clinical Importance of EWS-ETS Gene Fusion Type 1.2.2.5 Other Genetic Alterations in E T 1.2.3 The Role of the IGFI-Receptor and other Growth Factro Receptor Pathways in E T Regulation of Normal C e l l Growth & Proliferation 1.3.1 Growth Factor-Mediated C e l l Signaling 1.3.1.1 Growth Factor Receptors 1.3.1.2 The R A S — R A F 1 — M E K — E R K Mitogen Activated Protein Kinase Pathway 1.3.1.3 The P I 3 K — A K T Pathway 1.3.1.4 Cross-talk between Growth FactorMediated Signaling Pathways 1.3.2 The C e l l Cycle 1.3.2.1 Cyclins and Cyclin-Dependent Kinases 1.3.2.2 Inhibitors of Cell Cycle Progression 1.3.2.3 Regulation of C y c l i n D 1.3.3 Differentiation  2 4 5 6 6 10 10 11 12 12 12 17  21 22 26 27 30 33 35  1.4  1.5  C H A P T E R II:  1.3.4 Apoptosis 1.3.5 Anchorage Dependent Growth Mechanisms of Oncogenesis 1.4.1 Oncogenes 1.4.2 Chromosomal Translocations & Gene Rearrangements 1.4.3 Tumour Suppressor Genes 1.4.3.1 p53 1.4.3.2 Rb 1.4.3.3 ARF 1.4.4 Autocrine Growth Factors Aims & Objectives  MATERIALS & METHODS 2.1 Cell Culture & Clinical Samples 2.1.1 C e l l lines and tissue culture 2.1.2 Anchorage-independent multicellular spheroid cultures 2.1.3 Primary tumour specimens 2.2 Analysis of Gene Expression 2.2.1 Isolation of R N A 2.2.2 Reverse Transcriptase-PCR 2.2.3 Southern & Northern Analysis 2.2.4 Differential D i s p l a y - P C R 2.3 Analysis of Protein Expression & Cell Structure 2.3.1 Radioimmunoassay 2.3.2 Immunohistochemistry & Electron Microscopy 2.3.3 Protein Lysate Preparation 2.3.4 Immunoprecipitation 2.3.5 Immunoblotting 2.4 Functional Studies & Proliferation Assays 2.4.1 EWS-FLI1 & GflP-promoter Reporter Gene Assays 2.4.2 C e l l Growth in vitro Following Treatment with G R P - R Antagonists & Agonists 2.4.3 C e l l Growth in vivo Following Treatment with G R P - R Antagonists 2.4.4 Signal Transduction in E w i n g cells following Treatment with G R P - R Antagonists & Agonists 2.4.5 B r d U Proliferation Assays 2.4.6 C e l l cycle analysis by F A C S 2.4.7 Kinase inhibitor studies 2.4.8 M y r i s t i l a t e d - A K T & Dominant-negative I L K Transfections 2.5 Clinical Correlates Study  35 37 39 39 42 45 45 49 49 50 51  52 52 52 52 55 56 56 56 57 58 59 59 59 60 61 62 63 63 63 65 66 66 67 67 68 68  VI  CHAPTER III:  CHAPTER IV:  CHAPTER V:  HUMAN GASTRIN-RELEASING PEPTIDE IS DIFFERENTIALLY EXPRESSED BY T H E EWING FAMILY OF TUMOURS 3.1 Introduction 3.2 Results 3.2.1 Differential expression of GRP in E T cell lines 3.2.2 Expression of GRP-R in E T cell lines 3.2.3 Expression of GRP and GRP-R in Primary E w i n g Tumours 3.2.4 Expression of Bioactive G R P peptide in ET 3.2.5 GRP is not a direct target of E W S - E T S chimeric proteins 3.2.6 Clinical Correlates Study 3.3 Discussion  GASTRIN-RELEASING PEPTIDE FUNCTIONS AS AN AUTOCRINE G R O W T H F A C T O R IN T H E EWING FAMILY OF TUMOURS Introduction 4.1 4.2 Results 4.2.1 in vitro C e l l Proliferation Studies 4.2.2 Collaborative studies on GRP-mediated Autocrine growth in vitro & in vivo 4.2.3 Blockade of G R P signaling in E T does not appear to affect E R K or A K T activation 4.2.4 C y c l i n D protein Levels are unaffected by blockade of G R P Signaling in E T cells 4.2.5 The effects of G R P - R blockade on the adhesion related kinases F A K & P Y K 2 4.3 Discussion  PROLIFERATION IN EWING TUMOUR C E L L S GROWN AS ANCHORAGE-INDEPENDENT SPHEROIDS 5.1 Introduction 5.2 Results 5.2.1 E w i n g tumour cells spontaneously form spheroids when grown in suspension 5.2.2 Adherent monolayer E T cells have significantly higher rates of proliferation than either spheroid or primary tumour cells 5.2.3 C y c l i n D protein expression in E T cells grown in suspension requires cell-cell adhesion and is serum dependent  70  70 71 71 75 75 78  81 83  88  88 89 89 91 91 94 97 97  103  103 104 104 106  108  vii  5.2.4  5.3  C H A P T E R VI:  Differences in cyclin D I expression between spheroids and monolayers are posttranscriptional and are associated with differences in subcellular localization  Discussion  T H E ROLES OF T H E RAS—RAF—MEK—ERK AND  112  118  122  P I 3 K — A K T P A T H W A Y S IN E W I N G T U M O U R C E L L PROLIFERATION  6.1 6.2  6.3  C H A P T E R VII:  Introduction Results 6.2.1 The E R K 1 / 2 M A P K and P I 3 K — A K T pathways are upregulated in E T cells in suspension 6.2.2 M E K inhibition does not appreciably affect cyclin D I protein expression or cellular proliferation in E T cells 6.2.3 P I 3 K inhibition blocks cyclin D I protein expression in E T cells and significantly reduces proliferation 6.2.4 Down-regulation of cyclin D I expression in P I 3 K inhibitor treated E T cells is posttranscriptional but does not correlate with GSK3(3 phosphorylation 6.2.5 Primary E w i n g tumours demonstrate patterns of cyclin D I expression and E R K 1 / 2 and A K T activation that resemble those of E T spheroids 6.2.6 Expression of constitutively active A K T by E T cells leads to spontaneous formation of anchorage independent spheroids 6.2.7 Preliminary studies evaluating the role of integrin-linked kinase ( I L K ) in A K T phosphorylation in anchorage-independent E T cells. Discussion  S U M M A R Y & F U T U R E DIRECTIONS  7.1 7.2 7.3 7.4  General Summary G R P Autocrine Growth Signaling in E T The Role of Adhesion Molecules in E T Proliferative Signaling C y c l i n D I Regulation and Subcellular Compartmentalization  122 123 123 124  127  130  133  136  139  141  149  149 150 153 155  7.5  REFERENCES  Final Comments  158  160  LIST OF TABLES  TABLE #  TITLE OF T A B L E  PAGE  T A B L E 1.  Selected Oncogenes in Human Cancer  40  T A B L E 2.  Summary of recurrent chromosomal translocations found in pediatric soft-tissue tumours  43  T A B L E 3.  Selected human tumour suppressor genes  46  T A B L E 4.  E w i n g tumour family cell lines  53  T A B L E 5.  N o n - E w i n g tumour cell lines  54  T A B L E 6.  Immunoreactive G R P peptide expression demonstrated by radioimmunoassay  79  T A B L E 7.  GRP/GRP-R Status of Primary Tumour Cohort and its Relationship to Event-Free Survival  T A B L E 8.  C e l l cycle analysis of culture T C 3 2 cells shows decreased cycling and increased serum sensitivity of spheroids  82  111  X  LIST OF FIGURES FIGURE #  TITLE OF FIGURE  PAGE  FIGURE 1.  Chromosomal translocations in E w i n g tumours.  FIGURE 2.  Growth signaling mediated by receptor protein tyrosine kinases  15-16  FIGURE 3.  Proliferative signaling via G-protein coupled receptors  18-19  FIGURE 4.  P I 3 K — A K T survival signaling  23-24  FIGURE 5.  The cell cycle  FIGURE 6.  Regulation of cell cycle progression  FIGURE 7.  Chromosomal translocation can result in a potentially oncogenic gene rearrangement  44  FIGURE 8.  Multiple pathways mediate effects of p53 activation  48  FIGURE 9.  D D - P C R analysis of S R C T cell lines  72  FIGURE 10.  Screening of S R C T cell lines for GRP expression by Northern analysis  74  FIGURE 11.  R T - P C R Analysis of S R C T C e l l Lines  76  FIGURE 12.  R T - P C R Analysis of Primary S R C T Samples  77  FIGURE 13.  Immunohistochemistry of primary S R C T s with a G R P specific antibody  80  FIGURE 14.  In vitro response of S R C T cell lines to G R P - R antagonist & agonist  90  FIGURE 15.  A ) , in vitro proliferation rate of E T cell line T C 7 1 in response to increasing concentrations of G R P - R antagonists RC-3095 & RC-3940II B ) . in vivo growth of human E T xenografts is slowed in response to treatment with G R P - R antagonist RC-3095  92  Manipulation of the G R P - R pathway does not affect levels of E R K 1 / 2 phosphorylation in T C 3 2 cells  95  FIGURE 16.  7-8  28 31-32  93  XI  FIGURE 17.  C y c l i n D levels remain unchanged in E T cells despite G R P stimulation or G R P - R pathway inhibition  96  FIGURE 18.  G R P - R antagonist treatment alters phosphorylation of Pyk2related proteins  98  FIGURE 19.  E T cells form multi-cellular spheroids when grown in suspension  105  FIGURE 20.  E T spheroids are morphologically similar to primary E w i n g tumours  107  FIGURE 21.  The proliferative index of E T spheroids is similar to primary E w i n g tumours  109-10  FIGURE 22.  C y c l i n D I protein expresssion is dependent on cell-cell adhesion and serum stimulation in suspension cultures of E T cells  113-114  FIGURE 23.  Variations in cyclin D I are post-transcriptional  116  FIGURE 24.  Subcellular distribution of cyclin D 1 differs between monolayer and spheroid E T cells.  119  FIGURE 25.  Differential activation of the E R K 1 and E R K 2 M A P kinases and of A K T in E T suspension cultures  125-126  FIGURE 26.  M E K inhibition has little effect on E T cell proliferation  128-129  FIGURE 27.  P I 3 K inhibition blocks cyclin D I protein expression in E T cells and significantly reduces cell proliferation  131-132  FIGURE 28.  Down-regulation of cyclin D I in inhibitor treated cells is posttranscriptional  134-135  FIGURE 29.  Primary E w i n g tumours express activated E R K 1 / 2 & A K T and variable levels of cyclin D I  137  FIGURE 30.  M y r - A K T induces spontaneous spheroid formation  138  FIGURE 31.  Effect of P I 3 K inhibition on expression of cyclin D I in M y r A K T transfected T C 3 2 cells  140  FIGURE 32.  Effect of dominant-negative I L K expression on A K T phosphorylation in cultured E T cells  142  xii  LIST OF ABBREVIATIONS A ALL AMP APAF1 ARF ARMS ATP BAD BCR bp BRDU C CAK CDNA CDK CKI CIP/KIP  CNS CTP DAG DBD DD-PCR der DFSP DMSO DNA DSRCT ECM EDTA EFS EGF EGFR ERG  adenine acute lymphoblastic leukemia adenosine mono-phosphate apoptotic protease activating factor 1 alternate reading frame alveolar rhabdomyosarcoma adenosine tri-phosphate B c l - 2 antagonist of cell death breakpoint cluster region base pair bromo-deoxyuridine cytosine C D K activating kinase complementary D N A cyclin dependent kinase cyclin dependent kinase inhibitor CDK-interacting proteins/kinase inhibitory proteins central nervous system cytosine tri-phosphate di-acyl glycerol D N A binding domain differential display P C R derivative dermatofibrosarcoma protruberans dimethyl sulphoxide deoxyribonucleic acid desmoplastic small round cell tumour extra-cellular matrix ethylene-diaminetetraacetic acid event-free survival epidermal growth factor EGF-receptor ETS-related gene  ERK ERMS ET ETS EWS FACS FAK FBS FGF FKFfR FLU FRNK G GAP GDP GPCR GRP GRP-R GSK3 GTP GTPase H&E IGFI/II IGFI-R ILK INK4 IRS kDa MAPK MEK MEM ML mRNA  extra-cellular signal regulated kinase embryonal rhabdomyosarcoma E w i n g tumour E-26 transforming specific E w i n g sarcoma fluorescent analysis cell sorting focal adhesion kinase fetal bovine serum fibroblast growth factor forkhead in rhabdomyosarcoma Friend leukemia virus integration site 1) F A K - r e l a t e d non kinase guanine GTPase activating protein guanosine diphosphate G-protein coupled receptor gastrin releasing peptide G R P receptor glycogen synthase kinase-3 guanosine tri-phosphate guanosine tri-phosphatase hematoxylin and eosin insulin-like growth factor type I and II I G F I receptor integrin linked kinase inhibitor of C D K - 4 insulin receptor substrate kilodalton mitogen activated protein kinase MAPK-ERK—activating kinase malignant ectomesenchymoma monolayer messenger R N A  xiii  MTT MYR NB NOS NTD OCT ONB OS PAGE PBS PCNA PCR PDK PDGF PDGF-R PDK PKB PKC PLC PMSF PN pPNET PRNK PSB PTB  methylthiazol tetrazolium bromide myristilated neuroblastoma not otherwise specified N-terminal domain optimal cutting temperature olfactory neuroblastoma overall survival poly-acrylamide gel electrophoresis phosphate buffered saline proliferating cell nuclear antigen polymerase chain reaction 3-phosphoinosifi dedependent kinase 1 platelet derived growth factor P D G F receptor phosphotidylinositol 3 - O H kinase protein kinase B protein kinase C phospholipase C phenylmethylsulfonyl fluoride peripheral neuroepithelioma peripheral primitive neuroectodermal tumour PYK2-related non-kinase phosphorylation solubilization buffer phospho-tyrosine binding  PTK PYK2 RAFTK RB RBD RGD RIA RMS RNA RPTK RT-PCR SCLC SDS SER SH2/3 SOS SRCT T TAD TBS TGFp TGFpTl-R THR TLS TNF TRAIL TSG VEGF  protein tyrosine kinase protein-tyrosine kinase 2 related adhesion focal tyrosine kinase retinoblastoma R N A - b i n d i n g domain arginine-glycine-aspartic acid radioimmunoassay rhabdomyosarcoma ribonucleic acid receptor protein tyrosine kinase reverse-transcriptase P C R small cell lung cancer sodium dodecyl sulphate serine Src-homology 2 and 3 son of sevenless small round cell tumour thymine transactivation domain tris-buffered saline transforming growth factor P T G F P type II receptor threonine translocated in liposarcoma tumour necrosis factor TNF-related apoptosis inducing ligand tumour suppressor gene vascular endothelial growth factor  xiv  ACKNOWLEDGEMENTS T may not have gone where I intended to go, but I have ended up where I intended to be." ~ Douglas Adams ~ This thesis would not have been completed, nor in fact undertaken at all, i f it were not for the continued and unwavering support of many. First and foremost, I must thank my supervisor and mentor D r . Poul Ff.B. Sorensen. When I initially approached him with the idea of spending a "short time" in his lab to learn more about childhood cancer genetics, neither of us had any concept of the commitment he would be making. A t that time, I had no visions of pursuing either a P h D or a career as a clinician scientist. N o w , largely because of his training and guidance, my career focus has shifted and I am embarking upon a completely new and challenging course. I am indebted to Dr. Sorensen for his patience, his wisdom and his friendship. Secondly, I owe tremendous thanks to all members of the Sorensen lab family, past and present. Because of their support, humour and camaraderie I was able to endure even the most frustrating months of slow progress and no results. In particular, I must recognize and thank Jerian L i m for her incredible patience at the very beginning when I was truly a neophyte in the lab. M u c h of my success in the lab can be attributed to her initial instruction. To my family and friends who have always believed in me and encouraged me to follow my dreams, thank you for your love and your unwavering support. Finally, and most importantly, I must thank the people who inspire me on a daily basis. T o my patients, the children who endure so much, thank you for all that you continue to teach me about courage, dignity and grace. I am honoured to know you and promise to do my very best to lessen your suffering.  1  CHAPTER I INTRODUCTION 1.1  SYNOPSIS AND RATIONALE FOR THESIS Each year, approximately 1200 children in Canada are diagnosed with a malignancy  but, because of the tremendous advances in diagnosis and treatment over the past 25 years, greater than 70% of afflicted children w i l l be long term survivors (1).  Unfortunately,  however, there are still a number of childhood cancers for which cures elude us and the majority of children diagnosed with these tumours succumb to their disease despite state-ofthe-art treatment. Thus, there is still a real need for the development of novel treatment strategies in childhood cancer. A m o n g the pediatric malignancies with continued high mortality rates are brain tumours, relapsed leukemias and the metastatic solid tumours. Advances i n treatment for these aggressive tumours depend on the development of an improved understanding of their biology. In this thesis, my focus is on growth regulation in the E w i n g family of tumours (ET), the second most common family of pediatric bone tumours and soft tissue sarcomas (2). Despite increasingly intensive therapeutic protocols, the mortality rate for metastatic and relapsed E T is extremely high. Innovative approaches to E T therapy need to be developed that target the cancer cells specifically by exploiting the biologic features that differentiate them from normal cells. In this thesis, I have studied the growth of E T cells in vitro in an effort to better define the signal transduction pathways that are important to its growth in vivo.  I have shown that E T differentially expresses human gastrin releasing peptide ( G R P )  2 and that this hormone acts as an autocrine growth factor in vitro and in a xenograft mouse model. Observations on the role of GRP in ET cell growth and its potential implications for novel treatment strategies will be discussed in Chapters III and IV. I have also demonstrated that E T cell growth in vitro is radically different between cells grown as anchorage independent spheroids and those grown as traditional adherent monolayers. The potential mechanisms behind these differences were investigated with respect to critical proliferative pathways and these results and their relevance to the in vivo setting will be discussed in Chapters V and VI. In the following sections, our current understanding of E T cell biology will be discussed with reference to the currently published literature. This will be followed by a general review of the normal mechanisms of cell growth regulation, in particular signal transduction and the cell cycle. Finally, a brief discussion of the general mechanisms of oncogenesis will complete this introductory chapter. For more in depth discussions of normal and dysregulated growth, the reader is referred to the list of references at the end of this thesis.  1.2  THE EWING TUMOUR FAMILY OF PERIPHERAL PRIMITIVE NEUROECTODERMAL TUMOURS 1.2.1  CLINICAL & PATHOLOGIC FEATURES  The Ewing family of tumours (ET), also referred to as peripheral primitive neuroectodermal tumours (pPNETs), comprises a group of genetically related aggressive bone and soft tissue malignancies including Ewing sarcoma, peripheral neuroepithelioma, and Askin's tumour (3).  Although E T predominantly affects children and adolescents, it  does occur in older individuals and its incidence in adults is likely underestimated (4). The  incidence of E T in Caucasians under the age of 14 is approximately 2.4 per million, with males and females being equally represented (2). Interestingly, E T almost never occurs in individuals of African or Asian origin for reasons that are not yet understood. This may be explained by subtle inter-racial differences in the sequence of the EWS gene, which is always rearranged in E T (see below)(5). Tumours can develop in almost any bone or soft tissue and presenting symptoms usually include pain and swelling. A t least a quarter of patients have evidence of gross metastatic disease at diagnosis and, given the 10% cure rate attained with local therapy alone, most have micrometastases (3). The most commonly involved primary sites are the pelvis and long bones and metastases are found most frequently in the lungs and bone marrow. Therapy for both localized and metastatic disease is multi-modal, requiring aggressive local control with surgery and/or radiotherapy in addition to multi-agent cytotoxic drug therapy. In cases of metastatic or recurrent disease, therapy has been intensified to the degree that, in some cases, myeloablative doses of chemotherapy are administered followed by autologous or allogeneic bone marrow transplant. Unfortunately, despite such aggressive approaches, the mortality rate of metastatic and relapsed E T is still greater than 75% and over a third of non-metastatic patients w i l l also succumb to their disease (6). In addition, treatment related morbidity is not insignificant and long-term complications of therapy continue to plague survivors for many years following the cure of their cancer. Histologically, E T is characterized by the appearance of small round blue cells, a feature  it  shares  with  other  pediatric  malignancies  including  neuroblastoma,  rhabdomyosarcoma and lymphoma. Morphologically, these tumour cells are characterized by a very primitive appearance without any specific defining features (7). Thus, successful differentiation of E T from other histologically similar tumours is extremely important and  4  often very difficult. Typically, the histology of E T is that of monomorphous sheets of small round cells with hyperchromatic nuclei, inconspicuous nucleoli, and scant cytoplasm (3). The histogenetic origins of E T remain unclear. Immunohistochemistry and ultrastructural analysis confirm the very primitive nature of E T cells and support a neuroectodermal derivation (3). A l s o , E T cell lines w i l l display neural differentiation after treatment with nerve growth factors (2, 8). However, E T can also demonstrate epithelial and mesenchymal characteristics and can arise in organs not directly related to the neural crest (3). Hence, although neuroectodermal origin is currently deemed most likely, the histogenetic orgins of E T are still not known with certainty. A n important immunohistochemical feature of E T that aids in its diagnosis and may be important to understanding its biology is its positive staining with the 0 1 3 antibody. This antibody recognizes the M I C 2 ( C D 9 9  M I C 2  ) antigen, which is a membrane glycoprotein  involved in cell adhesion that may be important in the control of cell cycle progression and cell morphology (3, 6). Detection of membrane localized M I C 2 in a tumour specimen is a sensitive diagnostic marker for E T but, given its very low degree of specificity, it cannot be used exclusively to differentiate E T from other histologically similar tumours.  Therefore,  until the advent of genetic diagnosis, E T was largely a diagnosis of exclusion.  1.2.2  GENETICS OF THE EWING T U M O U R F A M I L Y  In the m i d - 1 9 8 0 ' s , it was recognized that E w i n g sarcoma and peripheral neuroepithelioma both demonstrated recurrent and identical chromosomal translocations (9, 10).  Identification of the t ( l l ; 2 2 ) ( q 2 4 ; q l 2 ) translocation in these tumours led to the  realization that they were genetically related and provided pathologists with a diagnostic tool  5  that was specific for the E T family. It is now accepted that approximately 85% of E T cases exhibit a t ( l I ; 2 2 ) ( q 2 4 ; q l 2 ) translocation while the remainder demonstrate  alternate  translocations involving 2 2 q l 2 and another chromosomal partner, most commonly 21q22 (reviewed in (11)). Molecular cloning studies have since revealed that gene disruption and rearrangement occur as a result of ET-specific translocations, and identification of the affected genes has led to a dramatic increase in our understanding of the origins and biology of this tumour family.  1.2.2.1  EWS-ETS Gene Fusions  M o l e c u l a r cloning experiments of ET-specific translocations have successfully identified that two disparate genes are disrupted and then joined together as a result of the chromosomal translocations.  In all cases, the EWS  gene (for E w i n g sarcoma) on  chromosome 2 2 q l 2 is disrupted and, in most cases, it is fused to the FLU chromosome l l q 2 4 (12, 13).  gene on  In the majority of the remainder of cases, EWS is fused to the  ERG gene from chromosome 21q22 (14). Rare gene fusions resulting from t(2;22), t(7;22) and t(17;22) translocations have also been described in which EWS is fused to the genes FEV, ETV1, and E1AF, respectively (15-17). A l l EWS partners identified thus far in E T specific translocations are members of the ETS family of transcription factors. Additionally, all EWS-ETS  fusions occur in-frame such that the resulting fused sequences create  completely novel genes that encode chimeric proteins containing functional domains contributed by each of the involved partner genes.  6  1.2.2.2  EWS-ETS Chimeric Oncoproteins  E W S is a ubiquitously expressed R N A - b i n d i n g protein of unknown function that may be involved in m R N A transcription (3). It is closely related to the TLS/FUS  gene, which is  involved i n m y x o i d liposarcoma-specific gene translocations, and contains an aminoterminal transactivation domain ( N T D ) and a carboxy-terminal R N A binding domain. E W S E T S chimeric products retain the N T D of E W S and replace its carboxyterminus with domains from the E T S partner. The E T S family of transcription factors is a large family of D N A - b i n d i n g proteins that are implicated in the control o f cellular proliferation, development and tumorigenesis (18). They are characterized by a highly conserved 85amino acid domain termed the erythroblastosis virus-transforming sequence (ETS) domain that mediates specific D N A binding to purine-rich sequences with a G G A ( A T T ) core element (3). In E T translocations, the E W S - N T D is fused to the E T S domain of the gene partner, creating a gene with D N A binding properties of an ETS gene under the control of EWS regulatory elements  (see F i g u r e 1)(11).  EWS-ETS  proteins  are oncogenic and  transformation requires both the N T D - E W S and a functional E T S D N A binding domain (reviewed i n (11)). Although the precise mechanisms of E W S - E T S mediated oncogenesis are as yet unknown, many studies have sought to determine the nature o f E W S - E T S induced transformation.  1.2.2.3  Mechanisms of EWS-ETS Mediated Oncogenesis  Several lines of evidence support the notion that E W S - E T S oncoproteins function as aberrant transcription factors. First, E W S - F L I 1 localizes to the nucleus (19). Second, E W S FLI1 and E W S - E R G bind in vitro E T S target sequences with similar specificities and  FIGURE 1.  Chromosomal Translocations in Ewing Tumours.  Chromosomal  breakage of chromosome 2 2 q l 2 results in disruption of the EWS gene at one of several potential breakpoints between exons 7 & 10. This results in splitting of the N-terminal transactivation domain ( N T D ) from the R N A binding ( R B D ) domain and the N T D then fuses to the D N A binding domain ( D B D ) of an E T S partner gene that has been similarly disrupted by chromosomal breakage. Approximately 85% of EWS-ETS fusions involve fusion of N T D - £ W S with F L U . The resultant fusion gene encodes a novel chimeric protein that is oncogenic and acts as an aberrant transcription factor. Target genes of E W S - E T S chimeras are still being identified. Several are listed in the figure and discussed further in the text.  9  affinities as the native F L I 1 and E R G molecules, respectively (20, 21). Third, E W S - F L I 1 is a more potent transactivator than native F L U (22-24). The genes specifically regulated by E W S - E T S chimeric transcription factors remain largely unknown; however, several potential targets have been identified and their activation may not be solely dependent on the E T S domain but may also be affected by protein-protein interactions that are unique to the chimeric molecule. Thus, the spectrum of fusion gene targets may be different from native E T S targets (25). A m o n g putative E W S - E T S target genes identified to date are: MFNG,  a  member of the Fringe gene family of signaling molecules; the metalloproteinase stromelysin; the S H 2 - d o m a i n containing molecule EAT-2; the cyclin-specific ubiquitin-conjugating enzyme E2-C; c-myc; and serum response element-containing genes such as c-fos or egrl (see (3) for review). Recent evidence also indicates that the gene for the TGF(3-type II receptor is transcriptionally repressed by the E W S - E T S oncoproteins (26). Knowledge of genes transcriptionally activated or repressed by chimeric E W S - E T S oncoproteins may lead to a better understanding of the mechanisms underlying malignant transformation in E T and many investigators continue to endeavour to identify E W S - E T S transcriptional targets. Further evidence supporting the oncogenic role of EWS-ETS  gene fusions comes  from studies of E T cell lines. Transfection of antisense molecules into E T cell lines results in diminished E W S - F L I 1 protein expression and growth inhibition, both in vitro and in tumour xenografts (27).  In addition, transfection of E T cells with dominant negative  inhibitors to E W S - F L I 1 also results in diminished growth (28). Both of these experimental observations suggest that antagonism of the E W S - E T S proteins in vivo may provide a novel therapeutic strategy for E T .  10  1.2.2.4  Clinical Importance of EWS-ETS Gene Fusion Type  Analysis of large numbers of cases has documented that virtually all E T carry some form of EWS-ETS gene fusion, and this marker is now considered pathognomonic of the E T family (29). Moreover, screening of a variety of tumour subtypes for EWS-ETS  fusion  transcripts has led to the inclusion of at least some forms of malignant ectomesenchymoma ( M E M ) and olfactory neuroblastoma ( O N B ) in the E T family based on the expression of similar gene fusions (30, 31). Molecular analysis of the breakpoints within EWS and its partners has also revealed genetic heterogeneity such that as many as 18 different combinatorial fusion products are possible (3, 32). The most common fusion transcript links exon 7 of EWS with exon 6 of FLU, yielding the so-called type I fusion (32). Identification of the fusion type is potentially important to the understanding o f the biology of these tumours given that patients with the type I fusion have a better prognosis than patients harbouring the less common EWS-FLI1 patients with EWS-FL11 and EWS-ERG  1.2.2.5  variants (33). Thus far, the clinical features of fusions appear to be similar (34).  Other Genetic Alterations in E T  W h i l e EWS-ETS gene rearrangements are uniformly present and likely represent the primary genetic events in E T , other recurrent genetic lesions have also been described that may contribute to E T progression (reviewed in (3)). Frequent gains of chromosomes 8 and 12 have been observed by conventional and molecular cytogenetic techniques as well as losses of l p and an unbalanced derl6t(l;16). Clinico-pathologic cohort studies of primary tumour samples have also identified changes in tumour-associated genes. deletion of p l 6  I N K 4 a  Homozygous  was detected in 18% of cases and p53 mutations in 11% and both lesions  11  conferred a poor outcome (35, 36). Clearly, secondary genetic events are also important to the biology and clinical characteristics of E T and warrant further investigation.  1.2.3  T H E ROLE OF T H E IGF1-RECEPTOR AND OTHER G R O W T H FACTOR RECEPTOR PATHWAYS IN ET Most of the current knowledge surrounding the pathogenesis of E T involves the study  of its characteristic genetic lesions. Recently however, more efforts have been placed on defining the roles of specific growth factor-mediated pathways in regulating the growth of these tumours. A general review of normal growth factor-mediated proliferative and survival pathways can be found in section 1.3. The evidence supporting their dysregulation in E T is presented here. The insulin-like growth factor I receptor (IGF1-R) pathway was the first pathway that was identified to be of critical importance to E T pathogenesis.  In 1997, Toretsky et al.  reported that E W S - F L I 1 induced transformation of fibroblasts requires expression of the IGFl-receptor (37). Exactly how E W S - F L I 1 and I G F 1 - R interact to promote oncogenesis is not known and is the subject of intense ongoing study. Disruption of the I G F 1 - R pathway by receptor-targeted  antibodies or blockade of the  downstream  s i g n a l i n g molecule  phosphotidylinosistol-3-kinase (PI3K; see section 1.3.1.3) results in reduced proliferation and increased apoptosis of E T cells (38, 39). Thus, whatever the mechanism, the I G F 1 - R pathway is a key regulator of the pathogenesis and continued proliferation of E T cells. Other growth-factor pathways recently implicated in E T cell growth include the basic fibroblast growth factor (bFGF) pathway (40), stem cell factor and its receptor c-kit (41, 42),  12  and phospholipase C-(3 pathways (43). The precise roles of these pathways in regulating E T cell growth and proliferation remain to be determined, as do their relationships to the E W S E T S oncoproteins.  In this thesis, we explore the role of another growth factor, gastrin-  releasing peptide, in mediating E T growth and endeavour to elucidate the growth signaling pathways that are critical for E T cell proliferation.  1.3  REGULATION OF NORMAL C E L L GROWTH & PROLIFERATION 1.3.1  G R O W T H FACTOR-MEDIATED C E L L SIGNALING  The processes of cell growth, differentiation and death are tightly regulated in normal cells by complex biochemical events that govern cellular fate. Upon receipt of a signal from the extra-cellular environment or from within itself, a cell w i l l respond in the manner that is appropriate to the signal.  In other words, a normal cell can recognize and respond  accordingly to stimuli that signal it to divide, differentiate or die. Disruption of any part of this intricate homeostatic mechanism can lead to the aberrations of growth and differentiation that are the hallmarks of cancer. In this section, the cellular processes that control these key events w i l l be reviewed.  1.3.1.1  Growth Factor Receptors  Signal transduction is responsible for modulating or altering cellular behaviour and its precise coordination and integration is necessary for normal growth and differentiation in embryonic and adult life (44). Signal transduction cascades commence with the binding of extra-cellular growth factors to trans-membrane receptors that, once activated, bind  13  cytoplasmic proteins.  A m o n g these proteins are the c y t o p l a s m i c kinases  whose  phosphorylation at tyrosine or serine/threonine residues leads to conformational change and activation of kinase domains. This initiates a complex sequential cascade of phosphorylation and dephosphorylation that results in propagation of the signal to the nucleus (see (45) for review). Growth factor receptors include those with intrinsic tyrosine kinase activity, such as the receptor protein tyrosine kinase ( R P T K ) family, those which are indirectly coupled to tyrosine kinases, such as integrins, and those which are coupled to guanine nucleotide binding proteins, the G-protein coupled receptors ( G P C R ) . In very broad terms, activation of R P T K s usually results in cellular proliferation while G P C R signaling is more often linked to tissue-specific, fully differentiated cell functions (46). However, as w i l l be discussed in more detail later, it is becoming increasingly apparent that extensive cross-talk exists between proliferative and differentiation pathways. Ligands for the R P T K S are diverse and can be small monomelic polypeptides such as epidermal growth factor ( E G F ) and fibroblast growth factor ( F G F ) , or dimeric polypeptides such as platelet-derived growth factor ( P D G F ) . There are currently 58 known R P T K s , distributed into 20 subfamilies. A l l are transmembrane proteins with protruding growth factor binding amino termini that are several hundred amino acids in length (reviewed in (44)). Examples of R P T K s are the insulin-like growth factor type 1 receptor (IGF1-R), the E G F receptor and the P D G F receptor.  The cytoplasmic portion of the R P T K s contains a  catalytic tyrosine kinase domain that is typically about 260 amino acids long and is conserved with up to 90% sequence identity from one R P T K to another (45). Signaling by RPTKs  requires  ligand-induced receptor  o l i g o m e r i z a t i o n , w h i c h induces  tyrosine  autophosphorylation. That is, ligand binding brings together two or more receptors and their  14  catalytic subunits trans-phosphorylate one another. Once dimerized, R P T K s are catalytically active and the phosphorylated residues act as docking sites for cytoplasmic signaling proteins that recognize specific phospho-tyrosine residues.  Thus, active cell signaling molecules  complex at the cell surface with growth factor receptors, and with each other, through the recognition of phospho-tyrosines by specific binding domains.  These domains, are  evolutionarily conserved and include the Src homology-2 (SH2) and other protein tyrosine binding ( P T B ) domains (44). S H 2 and P T B domains lie outside of the catalytically active kinase domains of signaling molecules and bind to phosphorylated tyrosines that are adjacent to specific flanking amino acids (45).  Binding of S H 2 / P T B proteins to their recognized  phospho-tyrosine targets facilitates auto-phosphorylation, resulting in alteration of enzymatic activities and propagation of the signal. In addition to catalytically active molecules such as protein tyrosine kinases (e.g. phosphotidylinositol-3 kinase; P I 3 K ) , phospholipases (e.g. phospholipase C ; P L C ) , and protein tyrosine phosphatases (e.g. Src homology 2-protein tyrosine phosphatase 1; SHP1), other non- catalytically active S H 2 / P T B proteins exist. Such SH2-adaptor molecules include Grb2, insulin-receptor substrate 1 (IRS-1) and IRS-2, all of which act as links between signaling molecules within and among different signaling pathways.  In  summary,  RPTK  signaling  involves  ligand-induced  tyrosine  phosphorylation/activation of the receptor followed by sequential phosphorylation/activation of kinases, lipases and phosphatases, either directly or through adaptor molecules, such that the signal is propagated from the cell membrane to the nucleus. Proliferative signaling via the E G F R is shown schematically in Figure 2. G-protein coupled receptor ( G P C R ) signaling is less well characterized than that of the R P T K s . G P C R s are serpentine, seven-membrane spanning receptors that associate  15  FIGURE 2.  Growth signaling mediated by receptor protein tyrosine kinases  (RPTKs). In this figure, proliferative signaling induced by binding of the extracellular growth factor epidermal growth factor ( E G F ) to its transmembranous receptor E G F - R is shown schematically. Ligation of the receptor by E G F leads to receptor dimerization, intrinsic kinase activation, and auto-phosphorylation of cytoplasmic tyrosine residues. These residues then act as docking sites for cytoplasmic proteins with S H 2 or other protein tyrosine binding ( P T B ) domains such as Grb2. B i n d i n g and activation of cytoplasmic kinases initiates a cascade of phosphorylation and dephosphorylation of signaling proteins, resulting in transmission of the signal to the nucleus, where gene transcription occurs. In this figure, the most common proliferative cascade, the R A S — R A F — M E K — E R K mitogen activated protein kinase ( M A P K ) pathway is shown.  17  cytoplasmically with GTP-binding proteins (G-proteins). G proteins are small heterotrimeric molecules with a , (3, and y sub-units that are bound to G D P in their inactive state. Activation of a G P C R by its ligand causes the G-protein to exchange G D P for G T P and dissociate into a G a - G T P bound subunit and a G(3y heterodimer.  Both components subsequently activate  downstream effectors, most importantly phospholipase C and c y c l i c - A M P , that in turn acitvate protein kinase C and protein kinase A , respectively, initiating proliferation and differentiation in calcium- and cAMP-dependent fashions (47, 48). Futhermore, G proteins can directly regulate ion channel or kinase function (46). In addition to being linked to cell growth and differentiation, G P C R mediated signaling is also involved in the regulation of neurotransmission, endocrine and exocrine gland function, chemotaxis and embryogenesis (46).  Examples of G P C R s include the receptors for thrombin, bombesin, bradykinin,  substance P, acetylcholine and serotonin (46). G P C R mediated mitogenic signaling via the bombesin receptor is shown in Figure 3. Human gastrin releasing peptide is the mammalian equivalent of bombesin and its signaling will be discussed in more detail in Chapters III and I V in relation to its role as a growth factor in E T .  1.3.1.2  The R A S — R A F 1 — M E K — E R K Mitogen Activated Protein  Kinase Pathway Despite numerous divergent initiating signals, proliferative signaling via R P T K s and G P C R s most often converges on a common pathway that culminates in nuclear propagation of the signal. This pathway is the R A S — R A F — M E K — E R K mitogen activated protein kinase ( M A P K ) pathway. R A S is a 21 k D a molecular weight protein that cycles between active GTP-bound and inactive GDP-bound configurations (49). Inactivation of the R A S  18  FIGURE 3. Proliferative signaling via G-protein coupled receptors (GPCRs). G P C R s are linked to cellular proliferation via complex inter-connected pathways that are cell-type and situation specific. In this diagram, several well-characterized (solid arrows) and putative (dashed arrows) pathways linking bombesin/gastrin-releasing peptide ( G R P ) and its receptor to G l - S phase transition are shown. Proliferative signaling largely converges on the E R K M A P K pathway following activation of such upstream molecules as phospholipase C ( P L C ) , protein kinase C ( P K C ) and S R C . G R P signaling also links the adhesion related proteins P y k 2 and p l 2 5 to proliferation, possibly via activation of cytoskeletal proteins such as Rho and R O K (50) F A K  20  protein occurs via its intrinsic GTPase activity, which is facilitated by the family of GTPaseactivating proteins ( G A P s ) .  R A S activation occurs at the cell membrane, downstream of  growth factor receptor activation and is accomplished by guanine nucleotide exchange factors such as S O S (sons of sevenless) that exchange free G T P for R A S - b o u n d G D P (49). In the case of R P T K signaling, receptor activation leads to direct activation of R A S via the SH2-adaptor molecule G R B 2 , which links the activated receptor to S O S , thereby facilitating activation of R A S .  Activated R A S leads to the activation of several downstream effector  molecules, which in turn mediate the multitude of biologic effects that are governed by the R A S molecule (51). A m o n g R A S effectors is R A F 1 , the next critical molecule in the R A S proliferative pathway. R A F 1 is a serine/threonine kinase that becomes phosphorylated upon binding to activated R A S . R A F 1 phosphorylates and activates M E K 1 and M E K 2 ( M A P K E R K kinases), dual-specificity kinases that in turn activate E R K 1 and E R K 2 (extracellular stimulus regulated kinases) by phosphorylating threonine and tyrosine residues (49). The E R K s are members of the M A P K family of signaling molecules and their activation leads to cell d i v i s i o n .  A c t i v a t i o n of the other M A P K s , S A P K / J N K (stress-activated  kinase/jun amino-terminal kinase) and p 3 8  HOG  protein  , occurs via structurally related molecules in  parallel signaling pathways but leads to a different outcome, growth arrest (49).  Once  activated, the E R K molecules translocate to the nucleus where they regulate gene transcription by phosphorylating transcription factors such as E L K 1 , E T S 1 and c-myc (52). Activation of these transcription factors leads to D N A replication and cell division. Figure 2 shows E G F receptor-induced R A S pathway activation. In addition, cytoplasmic targets of E R K exist, including the E R K regulators S O S , R A F 1 and M E K , such that the potential for positive feedback loops exists within the R A S — R A F 1 — M E K — E R K pathway (52).  21  1.3.1.3  The P I 3 K — A K T Pathway  In recent years it has become apparent that cellular proliferation is critically linked to cellular survival. That is, growth factors often initiate both proliferative and survival signals concomitantly. L i k e proliferative signals, survival signals generally originate from ligation of transmembrane receptors and converge on a single pathway, the phosphotidylinositol 3O H kinase ( P I 3 K ) — A K T pathway (53). One form of P I 3 K , which is activated by R P T K s , exists as a heterodimer of an 85 k D a regulatory subunit and a 110 k D a catalytic subunit. Activation of the catalytic subunit occurs when the p85 regulatory subunit binds to tyrosine phosphorylated growth factor receptors via two S H 2 domains (49). Additionally, P I 3 K can be activated indirectly via intermediate activation of R A S ; thus, like R A F , P I 3 K is a R A S effector and links proliferative and survival pathways (54). Once the P I 3 K p i 10 catalytic subunit is activated, it catalyzes the transfer of phosphate from A T P to membrane-localized phospholipids termed phosphoinositides.  The principally generated 3'-phosphorylated  phosphoinositides are phosphoinositol 3,4 bisphosphate (PT3,4P) and phosphoinositol 3,4,5 trisphospate (PI3,4,5P) which function as signaling intermediates regulating downstream signal transduction cascades (53). The 3'-phosphorylated phosphoinositides bind to the serine/threonine kinase A K T (also termed P K B ) at its pleckstrin homology domain resulting in a conformational change and translocation of A K T to the plasma membrane where it becomes phosphorylated. F u l l activation of A K T is dependent on phosphorylation of two key residues, serine-473 and threonine-308. Phosphorylation of A K T at these residues is accomplished by the direct binding of phosphoinositides and by the combined activities of three kinases that function independently and in concert. These three kinases are termed 3phosphoinositide-dependent protein kinase 1 ( P D K 1 ) , P R K 2 , and integrin-linked kinase  22 ( I L K ) (53). Activation of A K T is generally accomplished v i a activation of P I 3 K as just described. However, A K T can also be activated by n o n - P I 3 K dependent means such as agonists of the protein kinase A pathway and the calcium/calmodulin dependent kinase kinase ( C a M K K ) , which lead to phosphorylation of A K T (55, 56). O f particular relevance to pediatric solid tumours in general and E T specifically, the P I 3 K — A K T survival pathway is initiated upon ligand binding of the IGFl-receptor (see Figure 4). Various in vitro studies using dominant negative and constitutively active A K T alleles have demonstrated that growth factor-mediated survival is dependent on A K T activation and that A K T activation is sufficient for survival even i f upstream kinases such as P I 3 K are blocked (reviewed in (53)). This critical role for A K T in cell survival has now been explained by the observation that A K T effectors include molecules which are key to the regulation of programmed cell death or apoptosis. A m o n g the reported direct targets of A K T are the cell death mediators B A D , caspase 9, the Forkhead family of transcription factors, and the N F - K B regulator I K K (53).  Whether or not all of these are true A K T targets in  human cells is a subject of ongoing debate; in particular, the phosphorylation of B A D by A K T has been disputed (57). Identification of definitive A K T effectors, therefore, remains an area of intense study.  1.3.1.4  Cross-talk between Growth Factor-Mediated Signaling Pathways  It is becoming increasingly apparent that the R A S — R A F 1 — M E K — E R K and the P I 3 K — A K T pathways are integrally linked (reviewed in (58)). Several recent studies have demonstrated that A K T phosphorylates and negatively regulates R A F 1 (59, 60) and, as discussed above, P I 3 K has been shown to be an effector of R A S (54). A l s o , Treinies et al.  23  F I G U R E 4. P I 3 K — A K T s u r v i v a l s i g n a l i n g . In this figure, activation of the P I 3 K — A K T survival pathway is shown as occurs upon activation of the insulin-like growth factor I receptor (IGFI-R). Activation results in tyrosine phosphorylation of key cytoplasmic domains, which then activate P I 3 K (see text). Through the generation of intermediary 3'-phosphoinositides, P I 3 K activates A K T / P K B . A K T effectors are numerous and regulate apoptosis. Thus, the net effect of A K T activation is inhibition of apoptosis. A K T is also a negative regulator of R A F and therefore links the P I 3 K — A K T pathway to the R A S proliferative pathway. A s shown in the left side of the diagram, I G F I - R is a PTK-receptor that also activates the R A S — R A F — M E K — E R K M A P K pathway in a manner similar to the E G F - R shown in Figure 1.  25  demonstrated that while an activated form of M E K can induce cell division in quiescent fibroblasts, P I 3 K signals are required for this effect (61). Thus, two seemingly parallel signaling pathways have cross-talk capability by which they communicate with and impact on one another. Such cross-talk ensures a balance between cell proliferation and cell death, thus maintaining cellular homeostasis. For example, it is well established that constitutive expression of R A S can induce apoptosis or cell cycle arrest through mechanisms involving pl9  A R F  , p21  CIP1  , and p53 (reviewed in (62)). Parallel activation of the P I 3 K - A K T pathway  w i l l prevent R A S induced cell cycle arrest or apoptosis and allow proliferation to occur (58, 61, 63). Moreover, Reusch et al. have recently demonstrated that the effect of cross-talk between A K T and R A F is dependent on the origins of the initiating signal at the cell membrane (64).  They observed that while thrombin and P D G F activated both the  R A S — R A F — M E K — E R K and the P I 3 K — A K T pathways in vascular smooth muscle cells, thrombin stimulation led to cellular differentiation rather than proliferation while P D G F acted as a potent mitogen, having no effect of the cell phenotype. These divergent results were shown to be a consequence of differential phosphorylation of R A F by A K T in response to the two different growth factors. Thus, cross-talk between the R A S — R A F — M E K — E R K and P I 3 K — A K T pathways is important for the regulation of cell proliferation, death and differentiation and is mediated in large part by growth factor-induced signaling cascades. Disruption of these cascades and their mechanisms of cross-talk leads to dysregulated cell growth and is a common feature of many malignancies. Their respective importance in the regulation of E T cell growth was studied for this thesis and w i l l be discussed in Chapters V and V I .  26  1.3.2  THE CELL CYCLE  Homeostasis of normal tissues is maintained by an intricately regulated balance between cell proliferation, growth arrest and differentiation, and cell death.  In cancer this  balance is lost and cellular proliferation continues unchecked. In the previous sections, the major pathways governing proliferation and survival of cells were discussed.  These  pathways ultimately converge upon the cell cycle, which controls D N A synthesis and division of a cell. Understanding the cell cycle and its regulation is critical to understanding key elements of oncogenesis. In 1951, Howard and Pelc first described that D N A synthesis in a dividing cell occurs during a discrete temporal phase, rather than continuously as had been previously held (65). Subsequently, further experiments revealed that a cell divides into two daughter cells by passing through 4 distinct phases. A m o n g these phases are S phase, wherein D N A synthesis occurs and the parental cell chromosomal complement is duplicated, and M phase, for mitosis, where replicated D N A is divided and two genetically identical daughter cells are formed. The other 2 phases of the cell cycle, G l and G 2 , are gaps between mitosis and S phase and S phase and mitosis, respectively. In G l , the cell responds to mitogenic signaling and prepares itself for the initiation of D N A synthesis. In G 2 , D N A synthesis is terminated and cell growth continues with accumulation of proteins and organelles to be divided between the two daughter cells during mitosis. The duration of each of these phases may vary among cells of a population (49).  Non-proliferating cells do not perpetually move  through successive cell cycles. Rather, these cells exist in a quiescent state between M and S phases in a phase termed GO. Most cells in normal adult tissues exist in GO. Cells in GO and G l are receptive to proliferative signals and, as such, quiescent cells that have maintained the  27  ability to proliferate w i l l leave GO and enter G l in response to an appropriate signal (see (49) for review). A schematic diagram of the cell cycle is shown in Figure 5. Within the cell cycle, passage from one phase to another is strictly regulated by a complex interplay between enablers and inhibitors. The most well-studied checkpoint is that regulating passage between G l and S phases, termed the restriction point or " R point". In cultured cells, once a cell passes the R point, it is committed to enter S phase.  In vivo,  however, cells may arrest at different points within G l in response to different inhibitory signals; thus, in reality, there may be several R points in different cell types that restrict cell cycle progression (49). When cells are stimulated by growth factors to enter G l from GO, they generally require continuous mitogenic stimulation to reach the R point, after which mitogens can be withdrawn and cells will continue through the cell cycle in their absence (66).  In the following sections, the major regulators of cell cycle progression w i l l be  discussed.  1.3.2.1  Cyclins and Cyclin-Dependent Kinases  Passage of a cell from one phase to the next within the cell cycle is largely regulated by the interactions of cyclins and cyclin-dependent kinases ( C D K s ) . C D K s are inactive until they bind with their respective cyclin to form a holoenzyme complex that is catalytically active (66).  C y c l i n levels oscillate during the cell cycle due to variations in gene  transcription, protein translation and protein degradation, ensuring that levels peak at the time of maximal kinase activation (49). Mammalian cyclin family members include cyclins A to H , which all share a conserved sequence of about 100 amino acids. In mammalian cells, the C D K family includes 7 members that are conserved in size between 32-40 kDa, and share  F I G U R E 5. The Cell Cycle. Most quiescent adult cells exist in GO. Mitogenic stimulation can cause them to enter the cell cycle where they passage through an intricately regulated sequence of events culminating in mitosis or cell division. Once the R(restriction) point has been passed, the cell is committed to enter S-phase and cell cycle transition ceases to be mitogen-dependent.  29 approximately 40% sequence homology (49).  C D K s are expressed at constant levels  throughout the cell cycle and, once bound to cyclins, are active serine/threonine kinases. F u l l activation of the c y c l i n - C D K complex depends both on phosphorylation of a conserved threonine in the catalytic cleft by CDK-activating kinase, or C A K , and on dephosphorylation of inhibitory threonine sites by phosphatases of the cdc25 family (49). Once fully activated, the c y c l i n - C D K holoenzyme complexes phosphorylate their effectors, facilitating cell cycle progression. Passage through G l into S phase through the R point is mediated by cyclin D - and cyclin E-dependent kinases (66). C y c l i n D family members ( D I , D 2 and D 3 ) bind to C D K 4 and C D K 6 to yield at least 6 possible holoenyzmes that are expressed in tissue specific patterns (66). Activated cyclin D - C D K 4 / 6 complexes phosphorylate, and thus inactivate, the retinoblastoma protein, R b . Ffypophosphorylated R b represses the transcription of genes whose products are required for D N A synthesis, largely by binding transcription factors such as the E 2 F s (67). C y c l i n D - C D K 4 / 6 mediated hyperphosphorylation of R b disrupts the interaction between E 2 F s and R b , E 2 F s are released and transcriptional activation ensues (66). Completion of Rb phosphorylation is accomplished by the cyclin E - C D K 2 complex which is activated in response to E 2 F mediated induction of the cyclin E gene (66). Thus, phosphorylation of R b by cyclin D - C D K 4 / 6 is a pivotal event in R checkpoint passage. Once a cell has passed through the R point, cyclin A - C D K 2 activation occurs and is essential for the initiation of and progression through S phase (49). During S phase, levels of the B - c y c l i n s rise and entry into and out of mitosis is controlled by cyclin B - C D K 1 complexes ( C D K 1 is also known as cdc2) (49). dephosphorylated and cyclin B is degraded.  A t the completion of mitosis, Rb is  Degradation of the cyclin proteins, especially  30  cyclin D , is essential for controlled cell cycling and will be discussed in greater detail in a subsequent section. A summary of the c y c l i n - C D K complexes and their respective inhibitors (see below) is shown on a diagram of the cell cycle in Figure 6.  1.3.2.2  Inhibitors of Cell Cycle Progression  The cell cycle is tightly regulated by effectors, such as the cyclins and C D K s , and by inhibitors.  C e l l cycle inhibitors largely inhibit the activity of the C D K s and can be  segregated into two categories: the C I P / K I P family of inhibitors, which interact with cyclinC D K complexes i n all phases of the cell cycle, and the more specific I N K 4 family that specifically inhibits cyclin D-dependent kinases (see Figure 6) (66). The C I P / K I P family of polypeptide inhibitors includes p 2 1 (66).  CIP1  , p27  KIP1  and p 5 7  K1P2  A l l members of this family bind efficiently to c y c l i n - C D K holoenzyme complexes  containing any of the C D K s from 1-6, and inhibition is accomplished by induction of conformational changes within the C D K molecules themselves (49).  Additionally, all  C I P / K I P molecules have been found to directly inhibit C D K 2 in a dose dependent fashion such that high inhibitor levels are required to effect growth suppression (49, 68). Quiescent cells generally express high levels of p 2 7  KIP1  that diminish as the cell enters cycle and  accumulates cyclin D - C D K 4 / 6 (66). Rather than exerting an inhibitory effect similar to that on C D K 2 , K I P proteins bound to C D K 4 / 6 complexes are merely sequestered and the kinase remains catalytically active (66). In cycling cells, therefore, cyclin D - C D K 4 / 6 complexes act as effectors of cell cycle progression in a second manner: they sequester virtually all of the p27  KIPI  molecules thereby diminishing their availability to negatively regulate cyclin E -  C D K 2 (reviewed in (69)). Once cyclin E - C D K 2 is activated, it phosphorylates p 2 7  KIP1  ,  31  FIGURE 6. Regulation of cell cycle progression. Progression of a cell through the cell cycle is dependent on regulated expression and activation of cyclin-cyclin dependent kinase ( C D K ) complexes and their inhibitors. T w o classes of C D K inhibitors exist. The I N K 4 family specifically inhibits G l - p h a s e c y c l i n D - C D K 4 / 6 complexes while the C I P / K I P family act throughout the cell cycle. Successful transition from one phase to the next requires appropriately timed degradation of cyclins that were active i n the previous phase, (adapted from Slingerland, 1998 (49)).  33 targeting it for ubiquitin-mediated proteosomal degradation; thus, cyclin E - C D K 2 targets its own inhibitor for degradation ensuring completion of R b phosphorylation and passage beyond the R point (70). The I N K 4 proteins specifically inhibit the cyclin D-dependent kinases, C D K 4 and C D K 6 . There are four known members of the family to date: p l 6 pl9  1 N K 4 d  I N K 4 a  , pl5  I N K 4 b  , pl8  I N K 4 c  , and  . I N K 4 proteins inhibit the cell cycle by sequestering C D K 4 and C D K 6 into binary  C D K - I N K 4 complexes (66). Sequestration of C D K 4 and C D K 6 has two inhibitory effects. First, it prevents interaction between C D K s and cyclin D , thus preventing kinase activation. Second, disruption of cyclin D - C D K complexes effectively releases bound C I P / K I P proteins, making them available to inhibit cyclin E - C D K 2 complexes as discussed above. Inhibition of cell cycle progression is also accomplished by growth inhibitory signaling. Growth-inhibitory cytokines, such as the interferons and TGF[3, mediate G l arrest by diverse mechanisms i n c l u d i n g induction of C D K inhibitor m R N A and protein, suppression of R b phosphorylation, and downregulation of cyclin D I and c - M y c protein levels (49, 71).  1.3.2.3  Regulation of Cyclin D  A s discussed above, phosphorylation of the R b protein by c y c l i n D - C D K 4 / 6 is a pivotal event in R checkpoint passage. Therefore, it is not surprising that cyclin D levels are tightly regulated by both transcriptional and post-transcriptional mechanisms in normal, proliferating cells.  Induction of cyclin D I m R N A in early G l requires both mitogen-  induced, sustained activation of the E R K s and co-stimulation of integrins via attachment to the extracellular matrix (reviewed in (62)).  Either loss of adhesion or loss of mitogenic  34  stimulation leads to abrogation of E R K activation and diminished c y c l i n D I m R N A transcription. Recent evidence indicates that R A S — R A F — M E K — E R K activation must be accompanied by an intact P I 3 K signal for induction of cyclin D I m R N A synthesis (61, 63). In addition to its role in the regulation of cyclin D gene transcription, the P I 3 K — A K T pathway is also integrally linked to the post-transcriptional regulation of c y c l i n D I expression (72, 73). Muise-Ffelmericks et al. reported that serum induction of cyclin D I protein expression results from enhanced translation of its m R N A via a P I 3 K — A K T dependent pathway (72) and D i e h l et al. first linked the A K T - e f f e c t o r GSK3-(3 to posttranslational c y c l i n D regulation (73).  Thus, both the R A S — R A F — M E K — E R K and  P I 3 K — A K T pathways are key players in cyclin D regulation. Ubiquitin-mediated proteosomal degradation is another means by which cyclin D protein levels within a cell are controlled. Proteosomes are cytoplasmic protease complexes that rapidly degrade ubiquitinated-proteins into short peptides (74).  Briefly, a series of  enzymatic reactions involving ubiquitin-associated enzymes ligate ubiquitin molecules to targeted proteins in multimeric chains. The 26S proteosome then recognizes and degrades these poly-ubiquitinated substrates. C y c l i n D is targeted for ubiquitination, and therefore degradation, when it has been phosphorylated at threonine residue 286 (Thr-286) located near the carboxyl terminus (75). Phosphorylation of Thr-286 is catalyzed by GSK3(3 (73). Moreover, GSK3(3 -mediated phosphorylation of cyclin D I redirects the protein from the nucleus to the cytoplasm where proteosomal degradation can occur (73).  This export of  cyclin D I out of the nucleus is actively mediated by the nuclear exportin C R M 1 and appears to be essential for regulated cell division (76).  35  1.3.3  DIFFERENTIATION  In its most basic form, the process of cellular differentiation is essential for normal embyrogenesis and tissue development. It is the ability of pluripotent stem cells to give rise to differentiated progeny that enables the development of multi-cellular organisms comprised of cells and organs with different phenotypes and functions. A s discussed above, cellular differentiation pathways are often coupled to proliferative pathways and it is the balance between these that allows for normal growth. In the past ten years, the genetics of normal development and cancer have converged as aberrant regulation of normal embryonic developmental pathways have been found in many cancer cells (77).  In general, the  proliferative potential of a cell is inversely correlated to its degree of differentiation and, therefore, most cancerous cells exist in undifferentiated forms with variable degrees of differentiation potential (78). W h i l e a detailed discussion of normal and aberrant cellular differentiation is outside the scope of this thesis, it is worth noting that many pediatric solid tumours are embryonal and, therefore, altered differentiation pathways are likely involved in oncogenesis. The most compelling evidence for this comes from studies of W i l m s tumour and neuroblastoma which demonstrate distinct biologic similarities between normal embryonic cells and the corresponding tumour cells (79, 80). The potential role of autocrine G R P production in altering cellular differentiation pathways w i l l be discussed in Chapter I V in relation to the E w i n g family of tumours.  1.3.4  APOPTOSIS  Apoptosis is the active mechanism of programmed cell death (reviewed in (49)). U n l i k e cell necrosis, which is a passive response to injury in which cells swell and lyse,  36 apoptosis is the activation of a genetic program in which cells lose their viability, fragment and are ingested before losing membrane integrity. Apoptosis can be initiatied by intrinsic signals in response to injury or stress, such as occurs upon withdrawal of growth factors or nutrient supply and following exposures to ultra-violet radiation, ionizing radiation or toxins (49).  Extrinsic signals can also initiate apoptotic pathways through ligation of trans-  membrane death receptors such as FAS (CD95), tumour-necrosis factor (TNF) and T R A I L (71).  Ultimately, intrinsic and extrinsic pathways converge upon the common final  degradation phase in which the cell dies and is endocytosed. A n extensive discussion of apoptosis is not relevant to this thesis; however, the key molecules and events involved in intrinsic death signaling will be briefly reviewed. Initiation of apoptosis in response to injury or stress is dependent on a functional p53 gene. p53 functions as a tumour suppressor causing either G l - S phase arrest or induction of apoptosis (81). Once initiated, apoptosis is controlled by the molecular regulatory families of caspases and Bcl-2 which are directly and indirectly controlled by p53 and other triggers of apoptosis (49). Caspases are cysteine proteases which cleave their target molecules at aspartic acid residues (81). More than ten different caspases have been identified in humans and many are active in apoptosis. Although ubiquitously expressed, the caspases required for apoptosis vary depending on cell type. They become activated upon cleavage from precursor molecules and substrates include other caspases, D N A polymerases, cytoskeletal proteins and other proteins critical for cell survival.  Activation of caspases therefore  proceeds through an amplifying proteolytic cascade ultimately resulting in cell death. The Bcl-2 family includes the pro-apoptotic proteins B A X , Bcl-Xs and B A D and the inhibitory proteins Bcl-2 and B c l - X . Interaction and binding between different members of the Bcl-2 L  37  family is a critical mechanism by which the ratio between effectors and inhibitors is determined. A t a molecular level, members of the B c l - 2 family have a hydrophobic tail that allows them to bind the outside surface of their target organelles: the mitochondria, endoplasmic reticulum and the nucleus. F o r example, B A X enacts its pro-apoptotic function by binding to the mitochondria, interfering with its function and causing release of cytochrome c (81).  Once cytochrome c has been released from the mitochondria, it  orchestrates assembly of the apoptosome complex containing A P A F 1 and caspase 9 which then culminates in activation of caspase 3 and cell death (71). Thus, intrinsic cell death pathways are critically linked to functional B c l - 2 proteins, caspases and mitochondria.  1.3.5  ANCHORAGE DEPENDENT GROWTH  Thus far, the key mechanisms involved in cell signaling and cell division have been discussed with little reference to the role of the extra-cellular environment in the regulation of these processes. In truth however, cellular proliferation is critically dependent on cellular adhesion to an extra-cellular matrix ( E C M ) . This reliance on adhesion is termed anchoragedependence and it is a key feature that differentiates normal cells from transformed or cancerous cells, which demonstrate anchorage-independent growth (82).  The E C M of  mammalian cells varies in structure and composition between cell type, stage of development and location (83). The E C M of epithelial cells is highly structured in the form of a basement membrane that attaches cells to each other and to the membrane.  O n the other hand,  mesenchymal cells do not attach to each other and the E C M surrounds the cells in a less structured manner.  Components of the E C M of mesenchymal cells include the interstitial  collagens, type I to III, elastin, fibronectin, proteoglycans and vitronectin (83). Interaction  38  between cells and the E C M is essential for survival and proliferation, can play a role in cellular differentiation, and is mediated by structural and function domains within E C M proteins. These domains include the arginine-glycine-aspartic acid ( R G D ) domains, which interact with and attach to cells via integrins, and n o n - R G D domains that recognize other components within the E C M itself. These interactions result in the formation of a highly organized three-dimensional matrix in which E C M proteins associate with each other and with cells (83). Integrins are cell adhesion molecules that function as trans-membrane  receptors,  linking R G D proteins of the E C M with the initiation of signal transduction cascades (83). Disruption of R G D protein-integrin interactions leads to the induction of apoptosis in epithelial and endothelial cells (84) and cell cycle arrest in fibroblasts (85). Integrins are obligate heterodimers comprised of trans-membranous a and 6 subunits that interact with the E C M , and short cytoplasmic domains that interact directly with cytoskeletal proteins and protein tyrosine kinases (83).  Once activated, integrins can activate the proliferative  R A S — R A F — M E K — E R K pathway via interaction with the protein tyrosine kinase She (86) and the P I 3 K — A K T v i a the serine/threonine kinase I L K (integrin-linked kinase) (87). Additionally, a key function of integrin-mediated signaling is to regulate cytoskeletal reorganization and cell shape via activation of focal adhesion kinase ( F A K ) and other protein kinases that are components of focal adhesion plaques, which act as the cytoplasmic points of contact between integrins and the E C M (83). Thus, interactions between cells and E C M proteins regulate growth, survival, differentiation, cell shape and motility. Since these regulatory pathways are often disrupted in cancer and transformed cells are able to circumvent anchorage-dependence, it follows that  39  growth signaling must be different in anchorage-independent conditions.  In this thesis, we  test this hypothesis with respect to E T cell growth and present our results in Chapters V and VI.  1.4  MECHANISMS OF ONCOGENESIS 1.4.1  Oncogenes  Oncogenes, first identified in cancer-causing viruses, are now well established as major contributors to the development of cancer in humans.  Oncogenes may be viral in  origin or may be derived from normal cellular genes referred to as proto-oncogenes. To date, about 100 potential oncogenes of cellular or viral origin have been recognized (88). Protooncogenes are highly conserved in evolution and their products are important in the regulation of cell growth and differentiation in organisms ranging from primitive eukaryotes to humans (89). The expression of cellular proto-oncogenes is tightly regulated in normal cells, but i f converted to oncogenes, they can induce tumour formation. Examples of cellular proto-oncogenes include many growth factors, growth factor receptors, cytoplasmic protein kinases, G-proteins, cyclins and transcription factors (see Table 1). Conversion of protooncogenes into oncogenes can occur by several mechanisms including proviral insertion, gene amplification, point mutation, and chromosomal rearrangement (90). Chromosomal rearrangement will be discussed separately in a subsequent section. Activation of oncogenes by proviral insertion is complex and involves recombination between viral and cellular genomes following infection and integration of the virus into the cell.  In this manner, viral sequence is integrated adjacent to the cellular proto-oncogene  resulting in alterations that convert the normal gene to its oncogenic counterpart (90). The  40  T A B L E 1. Selected Oncogenes in Human Cancer (adapted from Park, 1998 and Kopnin, 2000 (88, 89))  Oncogene KS3 HST  Function of protein  Lesion  Tumours  F G F family member F G F family member  D N A transfection D N A transfection  Kaposi's sarcoma Stomach carcinoma  EGFR  Receptor tyrosine kinase  Gene amplification  PDGFR  Receptor tyrosine kinase  Gene rearrangement  TRK NEU  Receptor tyrosine kinase Receptor tyrosine kinase  RET  Receptor tyrosine kinase  Gene rearrangement Gene amplification Point mutation Point mutation  SRC BCR-ABL  Tyrosine kinase Tyrosine kinase  c-mos  Serine/threonine kinase Serine/threonine kinase Serine/threonine kinase Tyrosine kinase  c-raf Pim-1 ETV6-NTRK3*  H-RAS K-RAS N-RAS  G Protein G Protein G Protein  Dbl Ost  GEF GEF  crk  SH2/SH3 adaptor  N-myc c-myc  Transcription factor Transcription factor  Transcription factor c-fos Transcription factor c-jun Cell cycle regulator Cyclin Dl/pradl *see Knezevich, et al. 1998 (91)  Point mutation Gene rearrangement  Squamous cell carcinoma Leukemia (CML/AML) Colon cancer Breast cancer Neuroblastoma Thyroid cancer Colon carcinoma Leukemia (CML/ALL) Sarcoma Sarcoma  Proviral insertion  T-cell lymphoma  Gene rearrangement  Congenital fibrosarcoma  Point mutation Point mutation Point mutation Gene rearrangement  Gene amplification Gene amplification Gene rearrangement  Gene amplification  Colon, lung carcinoma A M L , thyroid cancer Melanoma Lymphoma osteosarcoma  Neuroblastoma M a n y neoplasms Bukitt's lymphoma Osteosarcoma Sarcoma Breast cancer  41 first described example of proviral insertion leading to oncogene activation was the 50- to 100- fold elevation of c-myc transcription observed in bursal lymphomas induced by avian leukosis virus (92).  In this situation, viral sequence insertion leads to enhanced and  unregulated expression of the c-myc proto-oncogene, thus converting it to an oncogene. A second mechanism of oncogene activation observed both in transformed cells in vitro and in tumours in vivo is gene amplification. Amplified copies of proto-oncogenes can be seen microscopically as homogeneously staining chromosomal regions and double-minute chromosomes (89).  The proto-oncogene may be amplified at its native chromosomal  location or may be found at another locus as a result of chromosomal rearrangement (see below).  While the precise mechanisms of gene amplification are not yet completely  understood, it appears to result from several rounds of unscheduled D N A synthesis during a single cell cycle and amplicons usually include several genes (90).  Amplified proto-  oncogenes are found in human tumours, often in tumour-specific patterns, and the presence of multiple copies of proto-oncogenes in tumour cells is associated with a poor prognosis (89). Examples include amplification of n-myc in neuroblastoma and HER2/neu (c-erbB2) in breast cancer. Point mutations in proto-oncogenes can also result in oncogene activation, such as occurs with the ras mutations that are characteristic of epithelial malignancies. Such singlebase mutations alter the amino acid sequences of the R A S proteins, causing diminished intrinsic GTP-ase activity and constitutive activation of R A S and its downstream proliferative pathways (89).  42  1.4.2  Chromosomal Translocations and Gene Rearrangements  T w o patterns of chromosome translocation have been observed in human cancers. The first pattern is observed in many solid tumours and can result in gains and losses of large portions of chromosomal material. These complex translocations appear to be random and are not tumour-specific (93). The second pattern of chromosomal translocation (referred to as the simple type) is characterized by distinctive rearrangements of chromosomal segments in disease-specific manners. These tumour specific translocations are often felt to be primary oncogenic lesions and they are particularly common in sarcomas and some leukemias (see Table 2). C o m m o n sites of chromosome translocations in malignant cells have breakpoints that are frequently near proto-oncogenes, suggesting that chromosomal breaks may lead to activation of oncogenes (94).  Alternatively, chromosomal translocation may result in  deletion of a tumour suppressor gene or fusion of two unrelated genes in frame creating a novel fusion gene with oncogenic potential (90). This latter phenomenon was first described with the cloning of the bcr-abl fusion oncogene from the Philadelphia chromosome, t(9;22) translocation, in chronic myelogenous leukemia (95). In this situation, fusion of the bcr gene from chromosome 22 with the abl gene from chromosome 9 leads to translation of a novel fusion protein with constitutive tyrosine kinase activity that is oncogenic. A s discussed in detail earlier, the oncogenic EWS-ETS  gene fusions of E T are created as a result of such  tumour specific chromosomal translocations. translocation are shown schematically in Figure 7.  The potential effects of chromosomal  43  TABLE 2.  Summary of recurrent chromosomal translocations found in pediatric soft tissue tumours, (adapted from de Alava, 1998)(3)  Translocation  Tumour type  Gene Fusion  Incidence (%)  ET  t(ll;22)(q24;ql2)  EWS-FLI1  85  ET  t(21;22)(q22;ql2)  EWS-ERG  10  ET  t(7;22)(q22;ql2)  EWS-ETV1  Rare  ET  t(17;22)(ql2;ql2)  EWS-E1AF  Rare  ET  t(2;22)(q33;ql2)  EWS-FEV  Rare  DSRCT  t(ll;22)(ql3;ql2)  EWS-WT1  95  M y x o i d liposarcoma  t(12;16)(ql3;pll)  TLS-CHOP  95  M y x o i d liposarcoma  t(12;22)(ql3;ql2)  EWS-CHOP  5  t(9;22)(q22;ql2)  EWS-CHN  75  t(12;22)(ql3;ql2)  EWS-ATF1  NK  t(X;18)(pll.23;qll)  SYT-SSX1  65  t(X;18)(pll.21;qll)  SYT-SSX2  35  Alveolar rhabdomyosarcoma  t(2;13)(q35;ql4)  PAX3-FKHR  75  Alveolar rhabdomyosarcoma  t(l;13)(p36;ql4)  PAX7-FKHR  10  DFSP  t(17;22)(q22;ql3)  COL1A1-PDGFB  NK  Congenital fibrosarcoma & mesoblastic nephroma  t(12;15)(pl3;q25)  ETV6-NTRK3  NK  Extraskeletal myxoid chondrosarcoma Malignant melanoma of soft parts Synovial sarcoma Synovial sarcoma  Abbreviations:  D F S P , dermatofibrosarcoma protuberans; D S R C T , desmoplastic  small round-cell tumour; N K , not known  44  5' genel 3' gene2  *'5'gene2 x  3'genel  Possible results of chromosomal translocation: Breaks in heterochromatin  N o observed effect  Deletion of gene 1 or gene 2  Potential deletion of tumour suppressor gene  Stop codon created in gene 1 or gene 2  Truncated protein  Non-in frame fusion between genes 1&2  Non-functional protein  In-frame fusion between genes 1&2  N o v e l , functional fusion protein (e.g. EWS-FLI1)  F I G U R E 7 . Chromosomal translocation can result in a potentially oncogenic gene rearrangement.  45  1.4.3  Tumour Suppressor Genes  Tumour suppressor genes (TSGs) are cellular genes whose inactivation increases the probability of tumour formation. Moreover, reinstitution of their function may suppress the growth of already established tumour cells (88). V i r a l oncogenes and genetic alterations can, therefore, effect tumour formation by either activating cellular proto-oncogenes as described above, or by inactivating T S G s . Currently, twenty tumour suppressors have been recognized and several are listed in Table 3. Three key T S G s involved in the evolution of most human tumours are p53, Rb, and ARF.  Unlike the dominant oncogenes, T S G s are recessive genes  and, as such, inheritance of an abnormal or deleted T S G locus predisposes an individual to the development of tumours. This phenomenon was first described by Knudson in relation to the R b gene as the "two-hit" hypothesis (96).  A s with oncogenes, it is now apparent that the  overwhelming majority of T S G s are components of the pathways that regulate cellular proliferation, differentiation, apoptosis, cell cycle, and genomic integrity (88).  1.4.3.1  p53  p53 is a frequent target of mutations in many human and rodent malignancies such that loss of p53 protein function is observed in over half of all human cancers. p53 encodes a nuclear D N A - b i n d i n g phosphoprotein that normally exists as a tetramer, which binds to specific D N A sequences (97). p53 can be activated by a variety of genotoxic stimuli that cause increased transcription, increased protein half-life, and/or increased translational initiation of its m R N A . Through a multitude of biologic pathways, activated p53 then exerts it net effects of cell cycle arrest, apoptosis, maintenance of genomic integrity, and D N A  46 T A B L E 3. Selected Human Tumour Suppressor Genes (adapted from Kopnin, 2000) (88)  F u n c t i o n of P r o t e i n  Tumours  Transcription factor C e l l cycle regulation Regulation of apoptosis Control of genomic integrity  Li-Fraumeni syndrome Most forms of sporadic malignancy in adults  C e l l cycle regulation ( G l - S phase transition)  Hereditary retinoblasoma Many sporadic tumours  Cell cycle regulation ( C D K 4 / 6 inhibitor)/p53 activation  Hereditary melanomas Many sporadic tumours  G A P protein ( R A S inactivation) Cytoskeleton-membrane link  Neurofibromas  BRCA1  D N A repair p53 activation  BRCA2  D N A repair Transcription factor  Hereditary breast and ovarian cancer Various forms of sporadic tumours Hereditary breast and ovarian cancer Various forms of sporadic tumours  WT1  Transcription factor p53 regulation  Hereditary W i l m s ' tumour  APC  fi-catenin regulation  Hereditary adenomatosis polyposis coli Sporadic colon cancer  VHL  Suppresses expression of VEGF(angiogenic factor)  V o n Hippel-Lindau syndrome Clear-cell carcinoma  Gene  p53  Rb  INK4a/ARF  NF1 ( neurofibromin) NF2 (merlin)  Schwannomas Meningiomas  47  repair (see (97) and (98) for reviews). A s shown in Figure 8, the pathways regulated by p53 are numerous and complex and a detailed review is beyond the scope of this thesis. In the following paragraphs, the role of p53 in the induction of cell cycle arrest and apoptosis will be briefly discussed. The effects of p53 on proliferation are mainly effected through actions on the G l phase of cell cycle progression, although it is now apparent that p53 also functions at the G 2 M checkpoint (97). In response to oncogenic or hyperproliferative stimuli (e.g. M y c , Ras, E l A ) , D N A damage, toxin exposure, or withdrawal of growth factors, p53 is activated. Activated p53 causes G l arrest by inducing expression of the cell cycle inhibitor p 2 1  CIP1  .  The mechanism behind G 2 - M growth arrest is unknown, but it is thought that p53 may directly decrease cyclin B l levels and indirectly inactivate C D K 1 (cdc2) (97). Induction of apoptosis in response to p53 can also occur following exposure to the aforementioned  stimuli.  The mechanisms of p53-induced apoptosis include both  transcriptional and non-transcriptional means.  The pro-apoptotic proteins B A X and I G F -  binding protein 3 are transcriptional targets of p53, as is the death-receptor ligand F A S (97). Additionally, p53 is a transcriptional repressor of the anti-apoptotic Bcl-2  gene.  Non-  transcriptional effects of p53 largely involve protein-protein interactions between p53 and factors involved in D N A repair (97). Whether a cell undergoes p53 induced cell cycle arrest or apoptosis is cell type and situation dependent. In addition, the presence of functional p21  CIP1  and cross-talk with R b are critical determinants in p53's role (97). From a clinical perspective, germ-line mutations of p53 exist among patients with the  Li-Fraumeni syndrome (99). These patients have inherited only one functional copy of p53 and are therefore predisposed to the development of malignancies.  49  1.4.3.2  Rb  Rb is a nuclear protein with a molecular weight of lOOkDa that is expressed by all cells (99).  The importance of Rb as a tumour suppressor was first noted in hereditary  retinoblastoma where germ-line deletions of chromosome 1 3 q l 4 containing the Rb locus were first described (100). The function of the retinoblastoma protein (Rb) in cell cycle regulation has been previously discussed and its role as a suppressor of proliferation is well characterized.  In addition to its role in G l - S phase transition, R b also acts as an anti-  apoptotic factor (97). A s such, Rb interacts with p53 pathways in regulating cell cycle arrest and apoptosis and evidence suggests that malignant transformation of a normal mammalian cell requires inactivation of both Rb and p53 function (99).  1.4.3.3  ARF  In 1995, the tumour suppressor gene pl4  ARt  was first described (101). Interestingly,  this gene is encoded by the same locus as the INK4a locus but ARF is inititiated from a different first exon and is transcribed in an alternate reading frame. Thus, the two proteins are distinctly different from one another even though they are encoded by the same genetic locus. L i k e p l 6  I N K 4 a  , A R F is capable of inducing cell cycle arrest and more recent studies  have identified it to be a tumour suppressor with the critical role of linking the p53 and Rb pathways (66). A R F acts upstream of p53 in its induction of cell cycle arrest and also acts to stabilize p53 by antagonizing the p53-negative regulator M D M 2 (66). A R F expression is activated by abnormal mitogenic signals that overstimulate the cyclin D - C D K — R b — E 2 F circuit, thus leading to activation of p53 and cell cycle arrest or apoptosis. In this manner, A R F is responsible for tumour surveillance and functions as a tumour suppressor.  The  50  accumulated data suggest that disruption of A R F , M D M 2 , and p53, like mutations in the pl6  I N K 4 c t  - c y c l i n D - C D K — R b pathway, seem to be part of the evolution of all cancer cells,  irrespective of patient age or tumour type (66).  1.4.4  Autocrine Growth Factors  Tumour cells are dependent on extra-cellular growth factors for proliferation and survival.  In a well-vascularized tumour critical nutrients and growth factors are readily  available to individual cells from endocrine sources.  However, even in the absence of  external growth factor supply tumour cells can continue to proliferate and survive. The mechanisms that allow such growth include those mentioned above: e.g. activation of oncogenes or inactivation o f tumour suppressor genes leading to the dysregulation of normal proliferative, survival and apoptotic pathways. Additionally, tumour cells can circumvent the need for an endocrine supply of growth factors by secreting their own autocrine growth factors. The genes that encode these autocrine peptides can, therefore, be classified as oncogenes and examples of autocrine growth factors in human malignancy include P D G F in gliomas, V E G F and F G F in K a p o s i ' s sarcoma, IGFII in rhabdomyosarcoma and I G F I in E w i n g tumours (89, 102, 103). Autocrine production of growth factors by tumour cells leads to constitutive activation of growth factor receptors and dysregulation of signal transduction pathways. We have found that G R P acts as an autocrine growth factor in E T and discuss these findings in Chapters III and I V .  51  1.5  AIMS & OBJECTIVES Despite the extensive knowledge that has been gained about the genetics of E T ,  relatively little is known about the signal transduction pathways that govern its proliferation. Insights into the biologic pathways that mediate the tremendous proliferative and survival potential of E T cells w i l l be important to the development of novel treatment strategies that can target these tumours in vivo.  Therefore, the studies presented in this thesis were  undertaken with the following aims: 1) T o characterize the expression of human gastrin releasing peptide in E T cells in vitro and in vivo and explore its role as a potential diagnostic and/or prognostic variable in E T patients. 2 ) T o determine the role of G R P as an autocrine growth factor in E T and characterize the signaling pathways through which it acts. 3) T o study E T cell proliferation and signal transduction pathways in an anchorage independent culture environment in an effort to better represent the in vivo setting. 4) T o study the roles of the R A S — R A F — M E K — E R K and P I 3 K — A K T pathways in the regulation of E T cell growth and proliferation.  52  C H A P T E R II MATERIALS & METHODS  2.1  C E L L CULTURE AND CLINICAL SAMPLES 2.1.1  Cell lines and Tissue Culture  A l l E w i n g and other small round cell tumour cell lines were originally obtained from Children's and W o m e n ' s Health Centre of British C o l u m b i a , from the American Type Culture Collection, or from D r . T i m Triche at Children's Hospital L o s Angeles. tumour cell lines used in this study are listed in Table 4 along with their respective  Ewing  EWS-ETS  gene fusion type. Table 5 lists the non-Ewing tumour cell lines that were used as controls as detailed in later sections. Cells were grown under standard conditions at 37°C in 5% C O , in R P M I medium supplemented with 15% fetal bovine serum, 2 m M glutamine and antibioticantimycotic ( G I B C O - B R L , Life Technologies, Burlington). For analysis of gene and protein expression in low-serum and serum-free conditions, media containing 2.5% and 0.25% F B S respectively, was used. Cells were grown as adherent monolayers except where specifically indicated and were collected for analysis by trypsinization (0.25% Trypsin, G I B C O B R L ) following washes with phosphate buffered saline (PBS) while in logarithmic growth. A l l cell lines were tested by R T - P C R (see below) for expression of EWS-ETS fusion genes.  2.1.2  Anchorage-independent Multi-cellular Spheroid Cultures  For anchorage independent suspension cultures, ambient environmental conditions and media were identical to adherent monolayer cultures. T o establish suspension cultures,  TABLE 4 .  Ewing Tumour Family Cell Lines  Cell line  Diagnosis  Gene Fusion  Reference  TC-71  ES  EWS-FLI1  (10)  TC-32  PN  EWS-FLI1  (10)  A4573  ES  EWS-FLI1  (104)  BC-ES1  AT  EWS-FLI1  unpublished results  6647  ES  EWS-FLI1  (104)  TC-174  MEM  EWS-FLI1  (31)  TC-253  MEM  EWS-FLI1  (31)  TC-547  MEM  EWS-FLI1  (31)  JFEN  ONB  EWS-FLI1  (30)  633  PN  EWS-ERG  (14)  466  ES  EWS-ERG  (14)  5838  ES  EWS-ERG  (105)  T A B L E 5.  Non-Ewing T u m o u r Cell Lines  C e l l line  Diagnosis  Reference  IMR-32  NB  (106)  San2  NB  (106)  CT-10  ERMS  (107)  RD  ERMS  (108)  Birch  ERMS  (31)  Rhl8  ARMS  (109)  TTC-487  ARMS  (31)  A204  Sarcoma N O S  (108)  Jurkat  ALL  (108)  H345  SCLC  (108)  H E L 92.1.7  erythroleukemia  (108)  55  confluent monolayers were trypsinized, resuspended as single cells, and replated on tissue culture dishes that had been coated with 1.4% agar.  Suspended cells were collected at  intervals up to 24 hours post-suspension for time course studies.  After 48 hours in  suspension, stable spheroids were evident and were separated from single and dead cells by low speed (500 rpm) centrifugation for 3 minutes and replated in new media on fresh agarcoated plates. Starvation of cells was accomplished by replacing the high-serum (15% F B S ) media with serum-free (0.25% F B S ) media for 20-24 hours prior to experimentation. Suspension cells were harvested by low-speed centrifugation, washed in P B S , then subsequently analyzed for protein and R N A expression as discussed in later sections.  2.1.3  P r i m a r y t u m o u r specimens  Except where indicated below for the clinical correlates study, primary small round cell tumour samples were all obtained from the tumour bank at Children's & W o m e n ' s Health Centre of B r i t i s h C o l u m b i a .  A l l samples were obtained prior to therapy and  diagnoses were confirmed by standard histopathologic and molecular pathologic techniques. A l l cases of primary E w i n g tumours were confirmed to express EWS-FLI1  gene fusions by  reverse-transcriptase P C R (see below). Fresh tissue samples flash-frozen in liquid nitrogen and stored at - 7 0 ° C were used for R N A and protein studies.  Formalin-fixed paraffin-  embedded blocks of tissue were used for immunohistochemical analysis (see below). For gene cloning and sequencing of gastrin releasing peptide ( G R P ) , frozen fetal lung tissue resected at C h i l d r e n ' s & W o m e n ' s Health Centre of British C o l u m b i a was used and bronchial carcinoid tumour tissue obtained from St. Paul's Hospital, Vancouver, British Columbia was used as a positive control for G R P expression studies.  56  2.2  ANALYSIS OF GENE EXPRESSION 2.2.1  Isolation of RNA  Total R N A was extracted from cell lines and primary tissues using the acidguanidinium-phenol/chloroform method of Chomczynski and Sacchi (110).  Briefly,  cell  pellets or approximately ten 5p,m-thick sections of primary tumour were resuspended and homogenized in 1 ml of Trizol ( G i b c o - B R L , Life Technologies) and R N A then isolated from the aqueous phase.  Purity and integrity of R N A was assessed by 1% agarose gel  electrophoresis and spectrophotometric analysis of the O D A  2.2.2  2 6 0  /A  2 8 0  ratio.  Reverse Transcriptase-Polymerase Chain Reaction (RT-PCR)  F i v e micrograms of total R N A from cell lines and primary tumours was reverse transcribed to c o m p l e m e n t a r y - D N A ( c D N A ) using random hexamer nucleotide primers under previously described reaction conditions (111). One microlitre of the generated c D N A was then amplified using the polymerase chain reaction ( P C R ) for analysis of gene expression as published previously (111). T o amplify EWS-ETS fusion transcripts, primers capable of amplifying EWS-FLI1  and EWS-ERG  fusions were utilized.  These primers  included the FLU and ERG consensus primer 11.3: 5 ' - A C T C C C C G T T G G T C C C C T C C - 3 ' , and the £ W S - s p e c i f i c primers 22.3: 5 ' - T C C T A C A G C C A A G C T C C A A G T C - 3 ' (111) and E S B P - 1 : 5' - C G A C T A G T T A T G A T C A G A G C A G T - 3 ' (14). human gastrin-releasing peptide  R T - P C R amplification of the  ( G R P ) gene and its receptor  ( G R P - R ) was  also  accomplished using previously described conditions (112). GRP primers included G R P - 1 ( 5 ' - T G C A A G G A A T T T G C T G G G T C T C - 3 ' ) corresponding to GenBank positions 295-316, and G R P - 2 ( 5 ' - T G T G A A T G G T A A C A G C T C G G G G - 3 ' ) corresponding to positions 758-779  57  (112). Bronchial carcinoid tumours and the small cell lung cancer cell line H345 were used as  positive  controls.  Primers  for  GRP-R  included  GRP-R1:  5'-  C T C C C C G T G A A C G A T G A C T G G - 3 ' corresponding to GenBank position 65-85, and G R P R 2 : 5 ' - A T C T T C A T C A G G G C A T G G G A G - 3 ' corresponding to position 468-488 (112). One microlitre of each c D N A product was also amplified with primers for the ubiquitously expressed antigen M I C 2 to ascertain the integrity of the c D N A (113). A l l amplified P C R products were visualized on a 2% agarose gel stained with ethidium bromide. The identity of the amplified fragments was confirmed by size and by blotting the P C R products onto nylon filters and hybridizing them with gene specific probes (see below).  2.2.3  S o u t h e r n & N o r t h e r n Analysis  T o confirm the identity of P C R products, amplified c D N A was transferred from 2% agarose gels to Nytran filters (Schleicher and Schwell Inc., N H ) by standard protocols, and the filters were then probed with radioactively labeled probes. F o r EWS-ETS  gene fusions,  blots were probed with a [y- P]dCTP (Amersham)-end-labeled E W S oligonucleotide probe 32  ( 5 ' - A G C C A A C A G A G C A G C A G C T A C G G G C A G C A G A - 3 ' ) . GRP  and GRP-R  were  identified by probing with [ a - P ] d C T P (Amersham)-labeled c D N A probes (114) (gift of Dr. 32  J. Battey, NTH). Northern blotting was performed using standard methods (115). Briefly, 10 to 25 ug of denatured total R N A (amount consistent within each experiment) was separated by agarose gel electrophoresis, blotted onto nylon membranes, and prepared for hybridization with lOOng of [ a - P ] d C T P (Amersham)-labeled c D N A probe. Probes utilized included the 32  aforementioned G R P and G R P - R probes, c D N A fragments recovered from differential  58  d i s p l a y - P C R gels (see below), a c y c l i n D I c D N A probe (provided by D r . T i m Triche, C H L A ) and an E W S c D N A probe (provided by Dr. Chris Denny, U C L A ) .  2.2.4  Differential D i s p l a y - P C R ( D D - P C R )  D D - P C R was performed by M s . Jerian L i m using the R N A i m a g e Kit™ (GenHunter) following the manufacturer's instructions. Briefly, D N a s e l treated total R N A was subjected to reverse transcription using 3 separate one-base-anchored oligo-dT primers ( d A - d T , d C n  d T , or d G - d T ) , to allow for initial subdivision of the m R N A populations. Complementary u  u  D N A was then amplified using the same oligo-dT primer used in first strand synthesis in combination with one of a series of 8 arbitrary 13mer 5' primers as described in the GenHunter K i t using a Perkin-Elmer 9600 thermocycler.  A m p l i f i e d products from the  different cell lines were run side-by-side on a 6% polyacrylamide sequencing gel (Bio-Rad) and D N A from differentially expressed bands was recovered, reamplified, and gel purified using the Qiex II gel extraction kit (Qiagen). Differentially expressed bands from D D - P C R were cloned and sequenced using the pCRII plasmid vector and T A cloning kits (Invitrogen, U S A ) using standard methods. The same methods were used for cloning and sequencing of bands following R T - P C R of cell lines, primary tumours, bronchial carcinoids, and fetal lung tissue using primers for GRP (see above). A l l sequences were submitted to GenBank for analysis. Confirmation of differential gene expression was accomplished by probing Northern blots with cloned D D - P C R c D N A fragments and with the G R P c D N A probe.  59  2.3  ANALYSIS OF PROTEIN EXPRESSION & C E L L STRUCTURE 2.3.1  Radioimmunoassay (RIA)  Radioimmunoassay was performed on cell culture media and cell lysates for quantitation of G R P protein expression.  H-345 was used as a positive control and test  samples were obtained from 3 small round cell tumour cell lines. C e l l culture media was collected 48hrs after a media change from cells in logarithmic growth phase (lOml/flask). Confluent cells were collected from T-75 flasks, lysed in single detergent 1% NP-40 (Sigma, U S A ) lysis buffer, and frozen at -70°C. Protease inhibitors aprotinin ( l ^ g / m l ) , leupeptin (ljig/ml) and phenylmethylsulfonyl fluoride ( P M S F ) (100/zg/ml) were added to collected culture media and cell lysates. A commercially available kit was obtained for R I A (Phoenix Pharmaceuticals, C A ) and the assay carried out as per the manufacturer's  instructions.  L y o p h i l i z e d culture media and cell lysate samples were resuspended in R I A buffer and assayed in duplicate for quantitation of G R P expression. Aliquots of lyophilized lysis buffer and R P M I with F B S were used to assess background G R P expression of the cell lysate and culture media preparations, respectively. For all samples, net G R P expression was calculated by subtracting background expression from the total quantity of measured G R P and results were expressed as picograms of G R P per 10 cells. 6  2.3.2  Immunohistochemistry & Electron Microscopy  Histologic and immunohistochemical analysis was done on 4 p m sections of formalinfixed paraffin-embedded tissue/cell blocks.  M o r p h o l o g y of cells was assessed by light  microscopy following staining with hematoxylin and eosin.  Immunostaining with all  primary antibodies was done following established protocols using a standard three-step  60  streptavidin-biotin peroxidase method, with 3-amino-9-ethylcarbazole as the chromogen. Antibodies used to detect G R P protein expression by primary E T cells were the anti-GRP antibody L R - 1 4 8 , and rabbit anti-human G R P antiserum (Dako) at dilutions of 1:1000 and 1:250 respectively. L R - 1 4 8 was raised in rabbits against synthetic porcine G R P ( l - 2 7 ) and has been previously described (116). Both antibodies detect the secreted bioactive form of human G R P . Negative controls for immunohistochemistry experiments included both nonEwing tumour samples and E T tissue sections that were processed without primary antibody to rule out non-specific peroxidase activity. For analysis of cell proliferation, similar immunohistochemical analysis was performed on primary E w i n g tumour tissue and E w i n g tumour cell blocks using antibodies against proliferating cell nuclear antigen ( B D Transduction Labs, Mississauga) and bromo-deoxyuridine (Sigma) (see below for more details). Electron microscopy of primary tumor specimens and T C 3 2 cells was done using standard procedures after cells were fixed in 3% glutaraldehyde.  2.3.3  Protein Lysate Preparation.  E w i n g tumour cell lines growing as monolayers or in suspension were rinsed once in P B S and then lysed in the appropriate lysis buffer for analysis of protein expression. For most studies of whole cell lysates, phosphorylation solubilization buffer (PSB: 5 0 m M Hepes, lOOmM NaF, l O m M N a P O , 2 m M N a V 0 , 2 m M E D T A , 2 m M N a M o 0 « 2 H 0 , 4  2  v  3  4  4  2  1%  T n t o n - X l O O ) containing protease inhibitors (leupeptin 10u,g/mL, aprotinin 10u,g/mL, and P M S F 2 5 0 u M ) and 0.01% H 0 w a s used. 2  obtained  by  cutting  sections  2  of  frozen  Primary tumor and xenograft lysates were tissue  directly  into  500ul  PSB  and  homogenization/solubilization was accomplished by high-speed vortexing and passage of  61  tissues through a 21-gauge needle. A l l lysates were incubated in 1.5ml Eppendorf tubes on a nutator at 4 ° C for a m i n i m u m of 30 minutes and insoluble fractions removed by centrifugation at 12000 R P M for 5 minutes. F o r preparation of cytoplasmic and nuclear lysate fractions, cells were lysed initially in 150u,l low salt buffer ( 2 0 m M Hepes, 5 m M KC1, 5 m M M g C l , 0.5% Triton-XlOO) with protease inhibitors (leupeptin 10ixg/mL, aprotinin 2  10Lig/mL, and P M S F 2 5 0 u M ) .  F o l l o w i n g low salt lysis, the cytoplasmic fraction (the  supernatant) was separated by centrifugation of the sample at 4000 R P M for 4 minutes. The nuclear pellets were then washed once in low salt buffer and resuspended in high salt buffer (low salt buffer with 2 5 0 m M N a C l added) with added protease inhibitors for lysis of the nuclear membranes. Following incubation on ice for 15 minutes, nucleoplasm was separated from insoluble membrane fractions by centrifugation at 4000 R P M for 4 minutes. For more stringent lysis of cells for whole cell lysate preparations, P S B was replaced by modified R I P A buffer ( 2 0 m M Tris p H 7.4, 1 2 0 m M N a C l , 1% Triton-XlOO, 0.5%Na deoxycholate, 0.1% S D S , 10% glycerol, 5 m M E D T A , 5 0 m M N a F , 0 . 5 m M N a V 0 ) containing protease 3  4  inhibitors (leupeptin l O u g / m L , aprotinin 10pg/mL, and P M S F 2 5 0 u M ) and 0.01% H 0 . 2  2  L y s i s of cells in R I P A buffer followed by passage of lysates through a 21-gauge needle ensured lysis of the nuclear membranes and inclusion of nuclear and other less soluble proteins in whole-cell supernatants. F o l l o w i n g solubilization in appropriate buffer, protein concentrations were determined and standardized using the D C B i o - R a d Protein assay kit (Bio-Rad Laboratories, C A ) .  62  2.3.4  Immunoprecipitation  Between 500 and 1000u,g of total protein lysates was incubated in 1.5ml Eppendorf tubes on a nutator at 4 ° C for 3 hours with 20ul of Protein-A conjugated sepharose beads (Pharmacia) and 2-5 u.g of primary antibody to the protein of interest.  Primary antibodies  used for immunoprecipitation included Pyk2 ( B D Transduction Labs), Src (Santa Cruz Biotechnology), F l i l (Santa Cruz Biotechnology), and the IGF-1 receptor beta-subunit (Santa Cruz Biotechnology). F o l l o w i n g incubation, tubes were centrifuged at 2000 rpm for 5 minutes and the supernatants discarded. Pellets were washed three times in wash buffer (PSB  containing 0.1% instead of 1% T r i t o n - X 100) and prepared for immunoblotting as  detailed below.  2.3.5  Immunoblotting  Immunoprecipitated samples or 30u,g of total protein lysates were boiled in Laemmli SDS-sample buffer (115), and resolved by S D S - P A G E . Briefly, boiled samples were loaded onto a Protean I l / x i cell electrophoresis system (Bio-Rad) and electrophoresed on 8-15% polyacrylamide gels at 4 - 1 5 m A according to standard methods  (117).  Following  electrophoretic transfer to Immobilon-P P V D F membranes (Millipore, M A ) , western blot analysis was performed using the indicated antibodies at dilutions according to the manufacturer's  instructions. P h o s p h o - E R K l / 2 , t o t a l - E R K , p h o s p h o - A K T , t o t a l - A K T ,  phospho-GSK3p\ total F A K , and total P Y K 2 antibodies were obtained from N e w England Biolabs ( N E B / C e l l Signaling, Mississauga, Ontario). Antibodies to cyclin D 1 / D 2 , p 2 7 and p 2 1  CIP1  were obtained from Upstate Biotechnology (Lake Placid, N Y ) .  KIP1  The F l i l and  I G F l - r e c e p t o r beta-subunit antibodies were obtained from Santa C r u z Biotechnology.  63  Biosource International ( U S A ) was the source of the p h o s p h o - P Y K 2 antibody.  Anti-  phosphotyrosine (RC20) and Grb2 primary antibodies and secondary anti-mouse and antirabbit antibodies conjugated to horseradish peroxidase were obtained from B D Transduction Labs (Mississauga, Ontario) and blotted proteins were visualized using enhanced chemiluminescence (Amersham Pharmacia, Quebec).  2.4  F U N C T I O N A L STUDIES & P R O L I F E R A T I O N A S S A Y S 2.4.1  EWS-FLI1 & GflP-promoter Reporter Gene  Assays  The cell lines Birch, S A N 2 , A4573 and T C - 3 2 cells were cultured as described above and luciferase reporter gene assays done by D r . W e n Tao. expressing EWS-FLI1  N I H 3 T 3 and N I H 3 T 3 cells  (19) (kindly provided by D r . C . Denny, U C L A ) were maintained in  Dulbecco's Modified Eagle M e d i u m with high glucose ( D - M E M ) containing 5% F B S and antibiotics.  The cells were transfected with constructs using Lipofectamine reagent  ( G I B C O / B R L ) using established methods.  Plasmids used included S7 and S6, containing  0.5-kb and 1.0-kb fragments of human G R P promoter-luciferase constructs respectively (118) (gift of D r . E . Spindel, Oregon Regional Primate Research Center, Oregon). addition, EWS/FLI1  In  expressing plasmids and the empty retroviral vector p S R a M S V were  obtained from D r . C . Denny (19).  Cells were harvested 48 hours after the start of  transfection, lysed by freeze-thawing and resuspended in 70 u l of extraction buffer (lOOmM K H P 0 , pH7.8, I m M D T T ) . The extracts were assayed for total protein using a B i o - R a d 2  4  Protein Assay K i t (Bio-Rad Laboratories). The (3-gal activity was assayed colormetrically and expressed as A  420  / ( m g protein x minute) (115). The luciferase activity was measured as  64  described (119). The luciferase assays were corrected for transfection efficiency using the (3gal assay.  2.4.2  Cell Growth in vitro Following Treatment with GRP-R Antagonist and  Agonist. Studies on the effects of G R P - R antagonist treatment of E T cell lines were done both at Children's & Women's Hospital, Vancouver and at and the Gerhard-Domagk-Institut fiir Pathologie, Munster, Germany i n collaboration with D r . Christopher Poremba.  In  Vancouver, the E T cell line 6647 and R M S cell line R D were treated with the G R P - R antagonist D C - 2 8 - 3 3 B ( H - t y r - G l n - T r p - A l a - V a l - G l y - H i s - L e u - O H 3 ) and agonist D C - 2 8 - 4 5 B ( H - p h e - G l n - T r p - A l a - V a l - G l y - H i s - L e u - M e t - N H 2 ) (peptides kindly provided by D r . D a v i d Coy, Tulane University School of Medicine, N e w Orleans). C e l l growth in low or optimal serum conditions was then measured directly by cell counting or indirectly by measurment of 3  H-thymidine uptake'as previously described (120). Briefly, cells were plated in 24-well  plates, treated after 48 hours with 0, 10 or lOOnM concentrations of either peptide, and then counted after 24 hours either manually with a hemacytometer or with a scintillation counter after a four hour incubation with H-thymidine. Data were collected in triplicate for each 3  experimental condition and the results averaged. Antagonist experiments carried out in Munster used the E T cell lines R D - E S and TC71.  E q u a l numbers of cells ( 2 x l 0 ) were plated in 96-well plates in lOOu.1 R P M I 4  containing 10% fetal calf serum. Cells were allowed to grow for 24 hours and then either RC-3095 or RC-3940-II was added to final concentrations of 1, 5, or l O n M for incubation times of 24, 48 or 72 hours. RC-3095 and RC-3940-II are both synthetic G R P - R antagonists  65  manufactured by A S T A Medica, Frankfurt, Germany with the following chemical structures: RC-3095  ([D-Tpi , 6  Leu \)/(CH NH)Leu ]Bn(6-14)) 13  l4  2  and  RC-3940-II  ([Hca , 6  Leu' \j/(CH NH)Tac ]Bn(6-14)). Following incubation with antagonist, cells were counted 3  14  2  manually or indirectly by spectrophotometric measurement of formazan using the M T T assay (Boehringer Mannheim). This assay is dependent on the ability of viable cells to reduce M T T (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide) thus producing a blue crystallized fomazan product which can be measured spectrophotometrically and interpreted using an E L I S A reader (121). Data were collected at least 4 times for each experimental condition and the results averaged.  2.4.3  Cell Growth in vivo Following Treatment with GRP-R Antagonists.  A l l in vivo experiments were completed in Minister, Germany by D r . Christopher Poremba. Athymic nude mice ( N M R I mice) were housed in sterile cages under laminar flow hoods in a temperature controlled room with a 12 hour light/12 hour dark schedule. A l l studies were carried out with full institutional approval in accordance with guidelines for the care and use of experimental animals at the Gerhard-Domagk-Institut fur Pathologie. E w i n g tumour xenografts were initiated in two nude mice by subcutaneous injection of 5 million C A D O - E S cells (a locally derived human E T cell line).  After 6 weeks of tumour  development, animals were sacrificed, tumours dissected and minced aseptically. Threemm  3  pieces of minced tumour were then transplanted subcutaneously by trocar needle into  the right or left flanks of 10 mice. T w o weeks following tumour transplantation, at a tumour volume of approximately 5 0 m m , the mice were divided into 2 experimental groups of 5 3  66  animals each.  Group 1 animals received vehicle control injections of 0.9% saline/0.1%  D M S O while Group 2 animals received injections of R C - 3 0 9 5 at a dose of 10u,g/animal. Injections were given subcutaneously once daily. Tumour volumes were measured every other day with microcalipers using the calculation formula of volume = length x width x height x 0.5236.  After 33 days of treatment, mice were sacrificed, tumours dissected,  cleaned in P B S and weighed.  2.4.4  Signal Transduction in Ewing Cells Following Treatment with  GRP/Bombesin, Bombesin Receptor Antagonist & GRP Antisense. E w i n g tumour cell lines grown as adherent monolayers or anchorage  independent  spheroids were starved overnight in serum-free media and then stimulated with 100-500nM bombesin or human G R P (both from Sigma Canada) for variable lengths of time as detailed in figure legends.  For antagonist experiments, cells were incubated with 500-1000nM  bombesin receptor antagonist (Sigma, Canada) for 24-72 hours in serum-free (0.25%) or low- serum (2.5%) media prior to lysis. GRP-antisense and control oligonucleotides were purchased from Biognostik, Germany. Monolayer and spheroid cells were incubated in lowserum media with 2 u M concentrations of oligonucleotide for 24-72 hours prior to lysis. Incorporation of oligonucleotides into E w i n g cells was confirmed prior to antisense experiments by demonstrating  uptake of a fluorescein-labeled  randomized-sequence  phosphorothioate oligo using fluorescent microscopy. A l l agonist, antagonist and antisense treated cells were then lysed in P S B lysis buffer and lysates analyzed for protein expression by immunoprecipitation and/or immunoblotting as detailed above section 2.3.  67  2.4.5  BrdU Proliferation Assays.  Stable T C 3 2 spheroids and logarithmic growth phase monolayers were starved in serum-free media for 24 hours to synchronize cells as much as possible. M e d i a was then changed to fresh media containing 15% F B S and 100p.M bromo-deoxyuridine ( B r d U ; Sigma) and cells allowed to grow for a subsequent 20 hours.  F o l l o w i n g the 20 hour  incubation, cells were pelleted, formalin fixed and embedded in paraffin as per standard protocols. Immunostaining with an anti-BrdU antibody (Sigma) was performed following the manufacturer's instructions. A l l conditions were repeated in triplicate and the percentage of B r d U positive cells was calculated by counting cells in 5 representative high-power fields for each condition (approx. 200-400 cells/hpf).  2.4.6  Cell Cycle Analysis by Fluorescent Activated Cell Sorting (FACS).  T C 3 2 monolayer and spheroid cells were resuspended in P B S and vortexed to break apart cell clumps.  Cells were then fixed in 70% ethanol, treated with 100u,g/ml RNase,  stained with 50u.g/ml propidium iodide, and live cells analyzed for D N A content using the F A C S C a l i b u r F l o w cytometry system with CellQuest and M o d f i t LT analytic software (Becton Dickinson, San Jose).  2.4.7  Kinase Inhibitor Studies.  Starved E w i n g tumour cells were treated with either the M E K 1 inhibitor U0126 (Calbiochem, San D i e g o ) at 2 0 u M for 30 minutes, the P I 3 K inhibitor L Y 2 9 4 0 0 2 (Calbiochem) or wortmannin (Sigma) at 5 0 u M or 200 n M , respectively, for 3 hours, or an equivalent volume dimethylsulfoxide ( D M S O ) vehicle control.  F o l l o w i n g inhibitor  68  treatment cells were harvested without serum stimulation or were stimulated with serum for 30 minutes or 3 hours prior to lysis. Total protein lysates and immunoblotting were carried out as above. T o assess the effects of M E K and P I 3 K inhibition on cell proliferation, E T cells were cultured in 20u,M U0126 or 5 0 u M L Y 2 9 4 0 0 2 , respectively in the presence of B r d U and serum for 20 hours and then cells collected for B r d U immunohistochemistry as described above.  2.4.8  Myristilated-AKT and Dominant-Negative ILK Transfections.  T C 3 2 cells were transfected with a Myc-tagged myristilated-AKT c D N A construct (Upstate Biotechnology, N Y ) , a dominant-negative integrin-linked kinase ( D N - I L K ) c D N A construct (gift of D r . S. Dedhar), or empty vector controls using lipofectamine ( G I B C O B R L ) and standard protocols. T w o days following transfection, cells were transferred to media containing 250u,g/ml G418 for selection of successfully transfected cells. (Geneticin; Sigma).  Cells were maintained in selection media until harvested for analysis of protein  expression. M y r - A K T and control transfected cells were treated with 5 0 u M L Y 2 9 4 0 0 2 or an equivalent volume of D M S O vehicle control for 3 hours f o l l o w i n g 21 hours serumstarvation.  Cells were then harvested immediately or following 30 minutes or 3 hours  serum-stimulation. Harvested cells were washed in P B S then lysed in P S B lysis buffer and processed for immunoblotting as detailed above.  2.5  C L I N I C A L C O R R E L A T E S STUDY Primary tumour specimens from 63 patients diagnosed with E w i n g tumours at  Children's and Women's Health Centre, Vancouver, B C and the Gerhard-Domagk-Institut  69  fur Pathologic Munster, Germany were analyzed by RT-PCR for expression of GRP and GRP-R  genes. A l l tumours expressed EWS-ETS  gene fusions.  Clinical and follow-up  information was available for 48 patients and their charts were reviewed for clinical (age, sex, stage & site of disease), pathologic (histology, fusion type) and outcome (relapse, status at last follow-up) data. Statistical analysis incorporated use of Kaplan-Meier estimates for event-free (EFS) and overall survival (OS) using the SPSS Version 7.0 statistical package. Events were defined as recurrence, progression of disease and death. Death resulting from therapy complications was not counted as an event but censored for survival and EFS analysis. GRP/GRP-R  positive patients were compared to non-expressers using the log-rank  test. Fisher's exact test or the X test were used where appropriate to examine a potential 2  association between dichotomous study variables of interest. considered to be of statistical significance.  P values of <0.05 were  70  C H A P T E R III HUMAN GASTRIN-RELEASING PEPTIDE IS DIFFERENTIALLY EXPRESSED BY THE EWING FAMILY OF TUMOURS 3.1  INTRODUCTION E W S - F L I 1 and E W S - E R G proteins have been demonstrated to be oncogenic (19),  and E W S - E T S oncoproteins likely function as aberrant transcription factors (11, 19, 24). Although the genes specifically regulated by E W S - E T S fusions remain largely unknown, recent studies have identified several potential targets of E W S - F L I 1 , as discussed in Chapter 1 (reviewed in (3) and (6)). Knowledge of genes transcriptionally activated or repressed by chimeric E W S - E T S oncoproteins may lead to a better understanding of the mechanisms underlying malignant transformation in E T . A s a strategy to identify genes differentially expressed in tumours with EWS-ETS gene fusions, we used the technique of differential d i s p l a y - P C R ( D D - P C R ) (122).  The  utility of D D - P C R in comparing gene expression patterns in human tumours is well documented (123, 124).  One advantage of this technique is that it allows for the  simultaneous comparison of multiple tissue samples in the same experiment. W e therefore used D D - P C R to compare gene expression patterns in E T cell lines expressing various EWSETS gene fusions with those of other pediatric small round cell tumour ( S R C T ) cell lines including rhabdomyosarcoma ( R M S ) and neuroblastoma ( N B ) . These studies demonstrated the differential expression of a gene with virtually complete sequence homology to human gastrin releasing peptide (GRP) in tumour cells positive for EWS-ETS  gene fusions.  GRP  71  expression was present in E T cell lines and primary tumours, but was not detectable in other pediatric S R C T s tested. In this study, the expression of GRP and GRP receptor (GRP-R) in E T cell lines and primary tumours was characterized.  F o l l o w i n g expression profiling, a  clinical correlates study was done to determine i f expression of either the GRP or GRP-R genes con-elated with clinical or pathologic features of primary E T in a multi-center cohort of patients.  3.2  RESULTS 3.2.1  Differential Expression of GRP in ET Cell Lines  T o detect genes differentially expressed in E T s , we initially compared D D - P C R generated gene expression profiles of the three E T cell lines, T C - 3 2 , TC-174, and T T C - 5 4 7 , to those of four R M S cell lines, R h l 8 , T T C - 4 8 7 , Birch, and C T 1 0 . Numerous differentially expressed P C R bands specific to either E T or R M S cell lines were observed using various combinations of 3'-oligo-dT primers and 5' arbitrary primers. O f these, we focussed only on those bands showing obvious differences  between the two tumour subgroups.  A  representative example is shown in Figure 9 A , in which a band, designated p5.1, was identified only in E T lines after P C R with primers H - T , , G and H - A P 4 .  Differentially  expressed fragments were next excised from gels and P C R amplified, and then used to probe Northern blots of the above E T and R M S cell lines (data not shown). Only those fragments for which Northern analysis confirmed differential expression were further characterized. These were cloned and sequenced, and sequences were compared to Genbank databases. Three ET-specific sequences generated using Ff-T,,G or H - T C primers in combination with H  arbitrary primer H - A P 4 , including clone p5.1, were found to have virtually complete  1  2  3  4  5  6  7  B. p5.1 hGRP  A A C G T G A A G G A A G G A A C C C C C A G C T G A A C C A G C A A T G A T A A T G A T G G C C T 397 A A C G T G A A G G A A G G A A C C C C C A G C T G A A C C A G C A A T G A T A A T G A T G G C C T 514  p5.1  C T C T C A A A A G A G G A A A A C A A A A C C C C T A A G A G A C T G C G T T C T G C A A G C A T 347  hGRP  C T C T C A A A A G A G A A A A A C A A A A C C C C T A A G A G A C T G A G T T C T G C A A G C A T 564  p5.1 hGRP  CAGTTCTACGGATCATCAACAAGATTTTCCTTGTGCAAAATATTTGACTA CAGTTCTACGGATCATCAACAAGATTT-CCTTGTGCAAAATATTTGACTA  p5.1 hGRP  TTCTGTATCTTTCATCCTTGACTAAATTCGTGATTTTCAAGCAGCATCTT TTCTGTATCTTTCATCCTTGACTAAATTCGTGATTrTCAAGCAGCATCTT  247 663  p5.1 hGRP  CTGGTTTAAACTTGTTTGCTGTGAACAATTGTCGAAAAGAGTCTTCCAAT CTGGTTTAAACTTGTTTGCTGTGAACAATTGTCGAAAAGAGTCTTCCAAT  197 713  p5.1 hGRP  TAATGCTTTTTTATATCTAGGCTACCTGTTGGTTAGATTCAAGGCCCCGA TAATGCTTTTTTATATCTAGGCTACCTGTTGGTTAGATTCAAGGCCCCGA  147 763  p5.1 hGRP  GCTGTTACCATTCACAATAAAAGCTTAAACACAT 113 GCTGTTACCATTCACAATAAAAGCTTAAACACAT 797  297 613  F I G U R E 9. D D - P C R analysis of S R C T cell lines. ( A ) . Differential expression of P C R fragment p5.1 (arrow) in cell lines expressing EWSETS gene fusions. L a n e l , T C - 3 2 (ET); Lane 2, R h l 8 ( A R M S ) ; Lane 3, T C - 1 7 4 ( M E M ) ; Lane 4, T T C - 5 4 7 ( M E M ) ; Lane 5, T T C - 4 8 7 ( A R M S ) ; Lane 6, B i r c h ( E R M S ) ; Lane 7, C T - 1 0 ( E R M S ) . (B). c D N A sequence shows homology to human GRP. D D - P C R product, fragment p5.1, was cloned into vector pCRII and sequenced, revealing virtually complete sequence identity to the 3'-terminus of the human preproG-RP gene (nucleotides 465-797).  73  sequence homology to the 3'-terminus of the human preproG-KP gene. A s shown in Figure 9 B , the sequence of clone p5.1 is almost identical to nucleotides 465-797 of preproGPP, representing the terminal 332 nucleotides of this gene (125).  C l o n i n g and sequencing of  GRP R T - P C R products from H345, a small cell lung carcinoma cell line, the E T cell line 6647, 2 primary E T s , a bronchial carcinoid and a human fetal lung tissue sample all revealed the identical substitution of cytosine for adenine at position 551 compared to published GRP sequences (data not shown). Additionally, a more 5'-alteration was discovered at position 365 where thymine is replaced by cytosine in 4 of the sequenced samples. Therefore, our studies may have detected previously undetected species of GRP or errors in previously submitted G R P sequence data.  The potential significance of these sequence changes is  discussed below. T o confirm differential expression of GRP in E T cell lines, both clone p5.1 and full length GRP c D N A were used to probe Northern blots of multiple S R C T cell lines. R N A s from 2 bronchial carcinoid tumours known to express GRP were used as positive controls. In total, all 6 cell lines with EWS-ETS gene fusions were positive for GRP, including 5 EWSFLI1 and one EWS-ERG  expressing cell lines (Figure 10). None of the nine other S R C T s  tested showed detectable GRP expression. A second Northern analysis of other E T cell lines also demonstrated positive but variable degrees of GRP expression (data not shown): the EWS-FLI1  expressing cell lines I A R C - E W 2 , F P B H , R M 8 2 , S K - E S - 1 , and S T A - E T 8 . 2 were  all positive; EWS-ERG  expressers G R - O H - 1 and K N - O H - 1 were weakly and strongly  positive, respectively. Furthermore, S T A - E T - 1 0 , a cell line containing the newly described EWS-ETS fusion, EWS-FEV (15), was also weakly positive for GRP (data not shown). A  BC ^  18S-  1  2  ET 3 4  RMS  Other  5 6 7 8 9 10 11 12 13 14 15 16  p5.1  B-actin  F I G U R E 10. Screening of S R C T cell lines for GRP expression by Northern analysis. Using full length preproGPP c D N A as a probe, EWS-ETS expressing cell lines were shown to differentially express GRP, while other S R C T lines were negative. Equivalent R N A loading was confirmed with a P-actin probe and two human bronchial carcinoid tumours, known to express GRP, were used as positive controls. Lanes 1 & 2 : bronchial carcinoids. Lanes 3-8: E T cell lines T C - 3 2 , T C - 7 1 , T C - 1 7 4 , T T C - 5 4 7 , J F E N and 466, respectively. Lanes 9 -13: R M S cell lines R h l 8 , TTC-487, B i r c h , C T - 1 0 and R D , respectively. Lanes 14 & 15: N B cell lines San-2 and I M R - 3 2 . Lane 16: undifferentiated sarcoma cell line, A204.  75  further E w i n g cell line, I A R C - E W 7 , in which a fusion transcript could not be identified was negative for G R P (cell lines reviewed in (126)). Using R T - P C R , we next screened 19 S R C T cell lines for GRP expression. A l l 11 E T samples tested were positive for GRP including EWS-FLI1  and EWS-ERG  expressing cell  lines. These results were confirmed by Southern analysis (Figure 11 A ) . None of the R M S or N B cell lines tested demonstrated GRP expression. T w o leukemia cell lines known to express wild-type FLU were also screened to test whether the presence of the F L U transcription factor correlates with GRP expression, but both were found to be negative. These results indicate that GRP is specifically expressed in E T among S R C T cell lines.  3.2.2  Expression of GRP-R in E T Cell Lines  R T - P C R and Southern analysis demonstrated transcripts for GRP-R the 11 p P N E T cell lines tested.  i n 6 (54.5%) of  Additionally, 1 of 2 N B and 1 of 4 R M S cell lines also  expressed GRP-R (Figure 11B).  3.2.3  Expression of GRP and GRP-R in Primary Ewing Tumours  T o rule out that the above in vitro results represented artifacts of tissue culture, we tested whether GRP and GRP-R  expression could be demonstrated i n primary E T tissue.  T w e n t y - s i x primary tumours were screened by R T - P C R for both GRP and  GRP-R  expression, including 16 E T s and 10 other S R C T s . O f the 16 E T s , 7 (43.7%) were positive for GRP, while none of the other 10 S R C T s including 7 R M S , 2 N B , and 1 intra-abdominal desmoplastic small round cell tumour expressed the gene (Figure 12A). GRP-R expression was found in 4 of 16 E T s (25%), all of which also expressed GRP, in both N B cases tested and in 4 of 7 R M S (Figure 12B).  A. GRP Expression in Cell Lines SC 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17  -4—485 bp  i. GRP-R Expression in Cell Lines SC 1 2 3 4 5 6 7 8 9 10 11 1213 14 15 16 17  <-390 bp  F I G U R E 11. R T - P C R Analysis of S R C T C e l l L i n e s . C e l l lines were tested by R T - P C R for GRP and GRP-R expression. The S C L C cell line H345 was used as a positive control in both assays. The identity of the amplified bands was confirmed with Southern analysis using c D N A probes for GRP and GRP-R, respectively. ( A . ) Southern blot showing uniform expression of GRP by E T cell lines. Lane S C : H 3 4 5 ; Lanes 19: EWS-FLU expressing E T cell lines T C - 7 1 , T C - 3 2 , A 4 5 7 3 , B C - E S 1 , 6647, T C - 1 7 4 , T C - 2 5 3 , T C - 5 4 7 and J F E N . Lanes 10 & 11: EWS-ERG expressing E T cell lines 633 and 466. Lanes 12 & 13: R M S cell lines CT-10 and R h l 8 . Lanes 14 & 1 5 : N B cell lines I M R - 3 2 and San2. Lanes 16 & 17: Jurkat and H E L , leukemia cell lines expressing wildtype FLIL (B.) Southern blot showing expression of GRP-R by 8 of 19 S R C T cell lines tested, including 6 E T s , 1 N B and 1 R M S . Lane assignments 1-6 as in A . Lane 7: R h l 8 , Lane 8: San2, Lane 9: CT-10, Lane 10: I M R - 3 2 .  A. GRP Expression in Primary Tumours SC 1 2 3 4 5 6 7 8 9 10 11 12  <-485bp B. GRP-R Expression in Primary Tumours SC 1 2 3 4 5 6 7 8 9 10  <-390bp F I G U R E 12. RT-PCR Analysis of Primary SRCT Samples. (A.) A representative Southern blot of 12 primary tumours, including all 7 ET tumours shown to be positive for GRP. No other primary SRCTs tested expressed GRP. Lane SC: H345; Lanes 1-7: ETs; Lanes 8 & 9: RMS; Lanes 10 & 11: NB; Lane 12: intra-abdominal desmoplastic small round cell tumour. (B.) Ten of 26 primary tumours were positive for GRP-R, including 4 ETs, 4 R M S and 2 N B . A representative Southern blot of 10 primary tumours is shown in this figure. Lane SC: H345; Lanes 1-4: ETs; Lanes 5 & 6: RMS; Lanes 7 & 8: N B ; Lane 9: RMS; Lane 10: intra-abdominal desmoplastic small round cell tumour.  78  3.2.4  E x p r e s s i o n of Immunoreactive G R P Peptide i n E T .  Using a commercially available G R P radioimmunoassay ( R I A ) kit we were able to demonstrate the presence of the immunoreactive G R P peptide in culture media and cell lysates from the E T cell line 6647 while little G R P peptide was detected in media or lysates collected from the N B and R M S cell lines S A N 2 and R D , respectively (Table 6). The levels detected in 6647 cells were similar to those observed in the positive control cell line H345. For all R I A results, the quantity of G R P expressed was indicative of biologically active peptide produced by the cells themselves, as background G R P activity in the lysis buffers and culture media was first subtracted. Immunohistochemistry of S R C T s using 2 different antibodies directed against bioactive G R P was also used in an attempt to show peptide expression, but the results were inconclusive. Twenty E T samples from 18 patients were tested. There was positive staining for G R P peptide in 7 (35%) of 20 E T s , but expression was highly variable ranging from diffusely positive intracellular staining (Figure 13A) to staining of the extra-cellular matrix, to positivity in only rare cells (data not shown). More consistent, however, was the complete absence of staining in other S R C T s including N B and R M S using either antibody (Figure 13B). Control slides using secondary antibody alone were uniformly negative in both E T and other tumour tissue sections.  79  TABLE  6.  Immunoreactive GRP peptide expression demonstrated by  radioimmunoassay. CELL LINE  CULTURE MEDIA (net pg G R P / 1 0 cells) 100.00±9.76  H345  CELL LYSATES (net pg G R P / 1 0 cells) 42.55±5.30  6647  27.90±4.10  San2  5.20±1.41  0  RD  10.55±1.06  4.45±6.29  6  6  71.60±10.04  C e l l lines were grown to confluence and their respective cell lysate and culture media preparations tested for expression of G R P using a radioimmunoassay. Quantitation of G R P was standardized per 10 cells after first subtracting background G R P activity from the total. 6  F I G U R E 13. Immunohistochemistry of primary S R C T s with a GRP-specific antibody ( A ) . Positive staining in a case of E T . (B). Negative staining in a case of R M S . (x400)  81  3.2.5  GRP is not a Direct Target of EWS-ETS Chimeric Proteins  E W S - E T S oncoproteins appear to act as aberrant transcription factors in E T cells (reviewed in (11)). It is therefore conceivable that E W S - E T S proteins may bind directly to and activate the GRP promoter.  However, luciferase reporter assays using luciferase  constructs containing G R P promoter inserts revealed no differences between E T , R M S , and N B cell lines despite similar transfection efficiencies (data not shown). Second, the same reporter constructs expressed in murine N I H 3 T 3 fibroblasts stably expressing EWS-FLI1  or  parental N I H 3 T 3 cells alone (19, 20) failed to detect differences. Third, reporter constructs were transiently co-transfected along with an EWS-FLI1  expression construct (19, 20) or  vector alone into parental N I H 3 T 3 cells, but no difference in luciferase activity was observed (data not shown). Finally, Northern blots of EWS-FLI1 EWS-FLI1  positive cell lines transfected with  antisense R N A (28) did not appear to affect GRP expression (data not shown).  W e therefore conclude that it is unlikely that E W S - E T S chimeric proteins directly activate the GRP promoter.  3.2.6  Clinical Correlates Study  Sixty-three cases of primary E T were identified, of which 48 had sufficient clinical and follow-up data to be included in the study. Patients ranged in age from newborn to 60 years with a median age at presentation of 14 years. Twenty-five patients were male, 23 female and 16 of the 48 presented with metastatic disease.  Site of primary disease was  almost equally split between central (N=26) and peripheral (N=20) skeleton while 2 patients had primary soft-tissue Ewings. A t the time of this study, 32 patients were still alive (median follow-up 42 months). Review of histopathology revealed 31 patients with typical E w i n g sarcoma, 6 with atypical E w i n g sarcoma and the remainder with peripheral primitive  82  neuroectodermal tumours. Molecular analysis confirmed the presence of an EWS-ETS gene fusion in all cases; 26 were EWS-FLI1 showed other rearrangements.  1/6 (type I), 15 were EWS-FLI1  7/5 (type II), and 7  R N A from all 63 cases was analyzed by R T - P C R for  expression of GRP and GRP-R and results showed almost equivalent numbers of expressers and non-expressers as shown in Table 7.  Data were then analyzed for any significant  differences in clinical or pathologic features between GRP/GRP-R  positive and negative  cases. A summary of event-free survival data is presented in Table 7. Data analysis showed no statistically significant difference between positive and negative cases for any of the f o l l o w i n g variables: age at diagnosis, sex, site of primary tumour, stage of disease, histopathology, gene fusion type, event-free survival, and overall survival.  T A B L E 7.  GRP/GRP-R STATUS OF PRIMARY TUMOUR COHORT AND ITS RELATIONSHIP TO EVENT-FREE SURVIVAL  Total number of cases studied Number of patients with follow-up data A l i v e at last followup: Range 11-115 mo. (median 42 mo.) Dead  TOTAL  GRP +  GRP-  GRP-R +  GRP-R  63  34 (54%) 24 (50%)  29 (46%) 24 (50%)  33 (52%) 20 (43%)  30 (48%) 27 (57%)  32  16 (50%)  16 (50%)  13 (41%)  19 (59%)  16  8 (50%) GRP+ 33%  8 (50%) GRP24%  7 (47%)  8 (53%)  48  5 YEAR EVENT-FREE S U R V I V A L (P>0.05)  83  3.3  DISCUSSION Using D D - P C R , we have shown that E T s differentially express a gene with virtually  complete sequence homology to human GRP.  Expression of GRP, although variable, was  found in 100% of a series of E T cell lines, all of which are known to be positive for EWSETS gene fusions. In contrast, GRP expression was not detectable in other S R C T tumour cell lines screened, including N B , embryonal and alveolar R M S and one E T lacking demonstrable EWS-ETS gene fusion. These results were confirmed by Northern blot analysis of a series of S R C T cell lines using a GRP c D N A probe. Moreover, using R I A we were able to demonstrate immunoreactive G R P peptide expression from a GRP expressing E T cell line. Little to no G R P peptide was detectable in 2 non-ET S R C T cell lines tested in the same series of R I A experiments. Gastrin releasing peptide is a member of the bombesin family of peptides and plays an important role in the control of gut motility and gastroenteropancreatic secretion, in thermoregulation and other CN-mediated regulatory processes (reviewed in (127)). It also functions as a growth factor in developing gastrointestinal and lung tissues and is expressed at high levels in fetal lung (127). In mature lung, pulmonary neuroendocrine cells express G R P , as do many of the small cell lung cancers ( S C L C ) which arise from these cells (128). G R P has been shown to function as an autocrine growth factor in S C L C (129) and other tumours that produce the peptide, such as pancreatic (130), colon (131) and breast cancers (132). Inhibition of the growth of these and other tumours has been demonstrated using GRP-receptor antagonists in vitro and in animal models (120, 133-135). antagonists is currently undergoing clinical trials in S C L C studies (136).  One of these  84 Sequencing of GRP c D N A amplified from tumour cell lines, primary tumours and human fetal lung tissue revealed consistent base changes at base pairs 365 and 551 as compared to the originally published c D N A sequence (125). The switch from thymine to cytosine at position 365 likely represents a polymorphism as it produces no change in the amino acid sequence of the peptide product.  L i k e w i s e , the replacement of the adenine  residue at position 551 occurs in the 3'-untranslated region of preproGRP.  Whether this  replacement affects the stability of the m R N A or the translation of the prepro-, pro-, or mature G R P molecules w i l l require further investigation. In contrast to GRP, which was expressed by 100% of E T cell lines and no other S R C T cell lines, expression of GRP-R  was variable and equivocal in both cell line and  primary tumour analysis. While there appeared to be a trend towards differential expression of the receptor among E T as compared with other S R C T , this was not definitive nor easily reproducible. G R P and other bombesin-like peptides exert their function by binding to high affinity receptors on the target cell's surface.  A t least four different receptors have been  described and because of the high homology among the bombesin-like peptides, there is potential for interaction between different ligands and receptors (137). W e tested cell lines and primary tumours for the presence of the GRP-preferring subtype of receptor alone (112). It is possible that other members of the GRP-receptor family are important in E T and this could explain the lack of detection of GRP-R  among all GRP expressing tumours.  Also,  there is often low level expression of growth factor receptor m R N A s in cells even when protein expression may be high. In fact, screening for GRP-R m R N A expression is reported to be problematic even in S C L C cells (112). Therefore, R T - P C R may not be the best method by which to study receptor expression and quantitation may be better established using  125  I-  85  GRP binding assays.  Detection of GRP-receptor m R N A in non-ET tumours lacking  concomitant GRP expression may simply indicate that other related peptides are important in the growth of these SRCTs and that these growth factors may interact with the bombesin family of receptors. Expression of G R P mRNA has been described previously in S K - N - M C I X C , an undifferentiated peripheral neuroepithelioma cell line (138). These cells co-expressed several neuropeptide genes, suggestive of a pluripotent neuronal cell. Additionally Sawin et al. have described GRP production from several primary childhood retroperitoneal tumours, including an ET (139) and have demonstrated its role as an autocrine growth factor in a neuroblastoma cell line, S K - N - S H (140). These data, together with our own, support the proposed neural origin of ET and suggest that they might represent another tumour group whose growth is influenced by GRP. We found variable immunoreactivity for GRP in primary E T tumours. The preproGRP gene demonstrates a 3 exon structure and encodes the 3 known forms of the preproGRP transcript by alternative R N A processing (114). Common to all 3 transcripts is the mature 27-amino acid biologically active peptide which has carboxyl-terminal domain homology with amphibian bombesin and is responsible for the biological activities of GRP and its related peptides (reviewed in (127)). Once cleaved from the biologically inactive proGRP hormone, G R P is detectable using LR-148 and the commercially available D A K O antibody to GRP. Previous investigators have shown that immunostaining for the proGRP peptides appears to be the most sensitive way of detecting GRP gene expression in routinely processed tumour tissues (141). This may be because intracellular G R P is rapidly secreted from cells once it is cleaved from the pro-hormone or, alternatively, mature G R P may be  86 much more unstable than proGRP. Either way, it is not surprising that many of our primary tumour specimens showed variable GRP immunoreactivity and it is possible that antibodies to proGRP may show more consistent staining of ET cells. Histopathologically, the members of the ET family express varying degrees of neural differentiation. This may also contribute to the variability of G R P expression among ETs given that identification of G R P mRNA in lung tumours and its translation to the mature peptide appears to depend on the degree of neuroendocrine differentiation (127). Studies with lung tumours have demonstrated that GRP and its receptors can be upregulated in nonS C L C cells if they are grown in serum-free and growth-factor free conditions (112). Cell culture of ET may facilitate expression of neuroendocrine elements leading to increased levels of GRP expression in cell lines as compared to primary tumours. Transfection studies and reporter assays failed to demonstrate direct targeting of the GRP promoter by EWS-FLI1 oncoproteins.  Also, Northern analysis of E T cell lines  transfected with EWS-FLI1 anti-sense mRNA did not reveal a direct relationship between EWS-FLI1  expression and the level of GRP expression. These data suggest that while ET  differentially expresses GRP, this expression does not appear to be directly activated by EWS-ETS chimeric transcription factors. G R P expression in these tumours may instead reflect the differentiation capacity of the cell of origin of ET, and provides further evidence for neuroectodermal histogenesis of these tumours. Whether expression of the GRP gene, the peptide and/or its receptor is somehow indirectly influenced by EWS-ETS fusion products remains to be determined. While GRP was uniformly expressed by ET cell lines, only 43.7% of the initial group of 16 primary tumours tested were positive for gene expression. It is well known that  87 aggressive primary tumours can be established in cell culture with greater ease than less aggressive tumours. This led us to hypothesize that G R P expression may be correlated with the degree of tumour aggressivity among E T s and, thus, tumour cells from established cell culture would be more likely to express G R P than a cohort of primary tumour samples. In order to study this hypothesis, R T - P C R analysis of a further 47 tumours (for a total of 63) was undertaken along with an analysis of available clinical and pathologic data.  RT-PCR  results confirmed expression rates of approximately 5 0 % for both GRP and However, we were unable to discern any significant differences between  GRP-R.  GRP/GRP-R  expressers in any of the studied variables. Our evidence suggests that although the expression of G R P may not influence clinical outcome, the peptide appears to be involved in the modulation of E T cell growth. Thus, we next attempted to define this role by investigating the growth pathways that are activated by G R P in E T cells. Defining G R P signaling in E T is of importance for several reasons. First, if G R P acts as a growth factor in vivo, G R P - R antagonists may prove to be useful therapeutic agents for patients with E T . Second, G R P expression may prove to be useful as a marker for microscopic metastases, m i n i m a l residual disease or disease recurrence, given the demonstration of immunoreactive G R P in the plasma of patients with S C L C (142). In the following Chapter our studies investigating the role of G R P as an autocrine growth factor in E T w i l l be presented.  88  C H A P T E R IV GASTRIN-RELEASING PEPTIDE FUNCTIONS AS AN AUTOCRINE GROWTH FACTOR IN THE EWING FAMILY OF TUMOURS 4.1  INTRODUCTION We have found that the E w i n g family of tumours expresses G R P in 100% of cultured  cell lines and 50% of primary tumours. Human G R P is a member of the bombesin family of proteins and is normally expressed in the human brain and neuroendocrine cells of lung, gut, and pancreas.  It also functions as a growth factor in developing gastrointestinal and lung  tissues and is expressed at high levels in fetal lung (127). O f importance to our studies, G R P has been shown to function as an autocrine growth factor in several malignancies including small cell lung cancer (129) pancreatic cancer (130), colon cancer (131) and breast cancer (132). Inhibition of the growth of these and other tumours has been demonstrated using GRP-receptor antagonists in vitro and in animal models (120, 133-135). The receptor for G R P is a member of the seven-membrane spanning family of G protein coupled receptors whose general structure and function was discussed in Chapter I (Section 1.3.1.1 and Figure 3).  N o r m a l proliferative signaling v i a the GRP/bombesin  receptor is still under study, but the literature to date has shown downstream effects to be both cell type and situation dependent (reviewed in (50)). In general, however, the mitogenic response of cells to G R P usually results from activation of the R A S — R A F — M E K — E R K M A P K pathway and initiation of G l - S phase transition (50).  89  In this chapter, our results supporting the hypothesis that G R P functions as an autocrine growth factor in E T will be presented. In addition, preliminary data suggesting that GRP  i n d u c e d mitogenesis  in E T may  not  directly involve  a c t i v a t i o n of  the  R A S — R A F — M E K — E R K pathway w i l l be discussed.  4.2  RESULTS 4.2.1  in vitro C e l l P r o l i f e r a t i o n Studies  Initial in vitro experiments assessing the effect of G R P - R antagonists and agonists on E T cell growth were carried out by myself and Dr. Chris Poremba. Treatment of the E T cell line 6647 with the G R P - R antagonist D C - 2 8 - 3 3 B for 24 hours in low-serum media resulted in a slowing of cell growth as measured by F£-thymidine uptake. 3  Conversely, similar  treatment of these cells with the G R P - R agonist, D C - 2 8 - 4 5 B , stimulated cell growth compared to the R M S cell line R D (see Figure 14).  L i k e w i s e , exposure of 6647 to the  receptor antagonist for 24 hours resulted in a 50% decrease in cell growth compared to control cells as demonstrated by direct cell counting while agonist stimulated growth by 140%  (data not shown).  These early results supported the potential role of G R P as an  autocrine growth factor in E T and further collaborative studies were continued by D r . Poremba in Munster, Germany to elucidate the autocrine nature of G R P growth stimulation in E T in vitro and in vivo.  90  F I G U R E 14. In Vitro Response of S R C T C e l l Lines to G R P - R Antagonist & Agonist. Treatment of p P N E T cell line 6647 with G R P - R antagonist, DC-28-33B resulted in a net decrease in cell growth compared to control levels as measured by F£-thymidine uptake. Treatment with G R P - R agonist, D C - 2 8 - 4 5 B , resulted in stimulation of cell growth of 6647 while neither peptide had any appreciable effect on the growth of the R M S cell line R D . Data points: mean of triplicate experiments. Bars: standard deviation. 3  91  4.2.2  Collaborative Studies on GRP Mediated Autocrine Growth in vitro & in  vivo In an effort to corroborate our preliminary results of the effect of G R P - R blockade on E T cell growth, D r . Chris Poremba completed the following in vitro and in vivo studies in Muenster, Germany. First, he treated the E T cell lines R D - E S & T C 7 1 with the synthetic G R P - R antagonists RC-3095 and RC-3940-II. Increasing concentrations of RC-3095 and RC-3940-II were observed to significantly slow the growth of E T cells as compared to untreated controls (Figure 15A). N o effect of these antagonists could be observed in an identically treated rhabdomyosarcoma cell line (data not shown).  Second, R C - 3 0 9 5  treatment of nude mice transplanted with human E T xenografts resulted in suppressed tumour growth. Growth inhibition was observed in treated mice as compared to controls within a week. After 33 days, the final tumour volume in RC-3095 treated mice was 924 ± 245 m m versus a final tumour volume of 2292 + 261 m m in control treated mice. The 3  3  tumour growth curves for the first 28 days of control and G R P - R antagonist treated mice are shown in Figure 15B.  4.2.3  Blockade of GRP Signaling in E T does not appear to affect E R K or A K T  Activation H a v i n g established both in vitro and in vivo evidence for the role of G R P as an autocrine growth factor in E T , we next sought to determine which signaling pathways mediated this effect.  In view of the previously published literature, the effects of G R P - R  stimulation and inhibition on the E R K M A P K s were first assessed. Stimulation of cells with serum for 30 minutes following a 24 hour period of serum starvation led to dramatic  ____  1.2  1.0  c •  03 '—  ro  <D  OH  0.8  A  0.6  H  1" cz o in in  —^  6  0.4 H  I-I  ——<  O  OH  TC-71 P L U S RC-3940II  FIGURE 15 A. In vitro  proliferation rate of E T cell line T C 7 1 in response to treatment with increasing concentrations of G R P - R antagonists R C - 3 0 9 5 & RC-3940II. Proliferation was assessed by the M T T assay as described in materials and methods. Similar decreases in proliferation in response to G R P - R antagonist treatment were also observed in the ET. cell line R D - E S . (sc=saline control; control=no injection).  Tumor Volume (mm) 3  control 2000 _ Ewing's Tumor Cell Line GADO-ES  1500 _  i  1000 _  1  RC-3095 10ug/day, s.c.  500 0 0  1  2  3  4  Weeks of Treatment  F I G U R E 15 B . In vivo growth of human E T xenografts is slowed in response to treatment with G R P - R antagonist R C - 3 0 9 5 . N O D - S C I D mice were implanted with 3-mm E T xenografts and then treated daily with RC-3095 or vehicle control. Weekly measurements of tumour volume revealed the growth of tumours to be significantly slower in antagonist treated mice. In the top panel, photographs of two representative mice demonstrate the visible differences in tumour size after 4 weeks treatment. 3  94  increases in E R K phosphorylation; however, similar stimulation with G R P or bombesin had no observable effect on these levels (Figure 16A). Stimulation of E T cells with G R P for periods ranging from 1 minute to 24 hours also had no effect of levels of E R K activation in repeated experiments (data not shown).  Incubation of E T cells in 0.25%, 2.5% or 15%  serum with receptor antagonists for periods ranging from 24 to 72 hours also had no effect on the activation levels of E R K 1 and E R K 2 (example shown in Figure 16B). Levels of E R K phosphorylation were likewise unaffected by treatment of E T cells with G R P antisense for 24 to 72 hours (Figure 16B). Because cellular proliferation is integrally linked to the P I 3 K — A K T pathway, we also tested whether levels of A K T activation were altered in response to the above interventions.  In keeping with our observations of E R K activation, no effects on A K T  phosphorylation were detected in any of the experimental conditions (data not shown).  4.2.4  Cyclin D protein Levels are unaffected by blockade of GRP Signaling in  ET cells. G l - S phase transition is dependent on mitogen induced increases in cyclin D protein levels within a cell. Stimulation of serum starved cells with G R P as described above for E R K studies had no effect on cyclin D levels. A l s o , treatment of E T cells in culture with G R P - R antagonists or G R P antisense did not diminish the levels of cyclin D expressed by the cells. In fact, as summarized in Figure 17, E T cells grown in culture as traditional adherent monolayers expressed high levels of cyclin D , regardless of serum starvation, serum stimulation, G R P - R blockade or G R P stimulation.  T h i s observation led to further  investigations of cyclin D regulation in E T cells in general and the results of these studies  US  G  SS P - E R K 1/2  Total E R K 1/2  B. AS  G  ANT  C P - E R K 1/2  Total A K T  F I G U R E 16. Manipulation of the G R P - R pathway does not affect levels of E R K 1 / 2 phosphorylation in T C 3 2 E T cells. ( A ) . Starved cells ( U S ) were stimulated for 30' with G R P (G) or serum (SS) and levels of phosphorylated E R K measured by Western blot with a phospho-specific E R K 1 / 2 antibody. ( B ) . Cells were grown for 3 days in 2.5% serum in the presence of 2 u M GRP-antisense oligonucleotide ( A S ) , 5 0 0 n M G R P (G), or lOOOnM G R P - R antagonist from S I G M A ( A N T ) . P h o s p h o - E R K levels were compared and found to be low in all conditions and unchanged from control cells (C).  A.  US  G  SS  F I G U R E 17. C y c l i n D levels remain unchanged in E T cells despite G R P stimulation or G R P - R pathway inhibition. ( A ) C y c l i n D levels are the same in T C 32 cells which are unstimulated (US), or stimulated for 30 min. with G R P (G) or serum (SS). (B) T C 3 2 cells grown for 24 hours in low-serum media (US) plus 5 0 0 n M G R P (G), 15% F B S (SS) or 5 0 0 n M G R P - R antagonist ( A N T ) show equal levels of cyclin D . (C) T C 3 2 cells were grown for 72 hours in 0.25% serum in the presence of either GRP-antisense ( A S ) or control (CO) oligonucleotide and cyclin D levels assessed by Western blot.  97  w i l l be discussed in Chapters V and V I .  4.2.5  The effects of GRP-R blockade on the adhesion related kinases FAK &  PYK2. Mitogenesis induced by G R P has recently been l i n k e d to the activation o f cytoskeletal proteins, i n particular the adhesion related kinases p l 2 5 Therefore, we assessed whether inhibition of either p l 2 5  F A K  F A K  and P y k 2 (50).  or P y k 2 phosphorylation could  be induced by G R P - R antagonist treatment of E T cells and could be a possible mechanism for the reduced proliferation observed i n these cells. phosphorylation were seen, with p l 2 5  F A K  N o differences i n p l 2 5  F A K  being highly phosphorylated in all conditions (data  not shown). However, subtle but potentially interesting changes in P y k 2 were observed in G R P - R antagonist treated cells.  A s shown i n Figure 18, antagonist treated cells  demonstrated increased tyrosine-phosphorylation of a ~30kDa protein with Pyk2 homology. This tyrosine-phosphorylated protein was repeatedly detected i n G R P - R antagonist treated cells.  W h i l e its identity is not known, it may be the Pyk2-splice variant P R N K (see  discussion below).  4.3  DISCUSSION W e have shown that G R P functions as an autocrine growth factor in E T , both in  vitro and in vivo. This further supports a neural origin for E T given that G R P is a neuropeptide and is implicated i n neuroendocrine growth signaling in both normal and malignant cells (127).  Furthermore, because the G R P pathway may provide a novel  therapeutic target for E T , we must seek to better understand its function i n this tumour family.  A.  B. C  ANT  C ANT  119kDa-, « «  ^  W9  .,  ,4  kDa-- %  4^  BLOT: aRC-20  y  k  2  i.T_  IP-«Pyk2  28.8  P  «$«§  I  g  «4 ?  G  PRNK  BLOT: aPyk2  FIGURE 18. G R P - R antagonist treatment alters phosphorylation of Pyk2-related proteins. (A) ocPyk2 antibody immunoprecipitation of control (C) and G R P - R antagonist treated ( A N T ) T C 3 2 cells followed by western blot with a phosphotyrosine-specific antibody shows an increase in phosphorylation of a ~30kDa protein. The blot was stripped and reprobed with the P y k 2 antibody (B) to confirm equal concentrations of total P y k 2 (top band). The smaller protein (bottom band) was present i n equal amounts in both conditions and reacted with the P y k 2 antibody suggesting it may be the splice variant P R N K .  99 At this point, the signaling pathways that mediate G R P stimulated growth in E T remain unclear. Our studies have failed to demonstrate alterations in ERK1/2 activation, A K T activation or cyclin D levels in response to GRP-R stimulation or inhibition. The only reproducible result with potentially interesting ramifications was the finding of increased phosphorylation of a Pyk2 protein homologue in antagonist treated cells. The identity of this 30kDa protein remains a mystery, although the Pyk2 splice variant P R N K (Pyk2-related nonkinase) presents an interesting candidate. Pyk2 is also known as related adhesion focal tyrosine kinase (RAFTK) but unlike its relative p l 2 5  FAK  , it is not widely expressed in cells of  all types, nor does it localize to focal adhesions. Pyk2 is a cytoplasmic tyrosine kinase that is predominantly expressed by brain and neural cells (143), a fact that further supports the neural origin of ET given the high levels of Pyk2 expression detected in E T cell lines and primary tumours (Figure 18 and RT-PCR data not shown). Pyk2 is activated in response to calcium fluxes in neural cells following ligation of GPCRs by neuropeptides (143). Pyk2 then phosphorylates the SRC kinase, which activates R A S via Grb2-SOS, leading ultimately to M A P K activation (50). Thus, we can hypothesize that Pyk2 may be activated downstream of the G R P receptor in E T cells and may be involved in controlling proliferation. Although we have been unable to show activation of Pyk2 in response to G R P stimulation in our in vitro model, we have potentially shown increased levels of P R N K phosphorylation in response to GRP-R blockade. P R N K shares C-terminal homology with Pyk2 and the B D Transduction Labs antibody to Pyk2 used in our experiments was generated against the Cterminal Pyk2 domain. Nothing is yet known about the function of P R N K although a current hypothesis is that it may act as a dominant negative inhibitor of Pyk2 (144). This hypothesis is based on what is known about p l 2 5  FAK  . The kinase domain of p l 2 5  F A K  is inhibited in a  100  dominant negative fashion by a protein called F R N K (FAK-related non-kinase). F R N K is encoded by a splice variant of pl25  FAK  and shares C-terminal homology with p l 2 5  F A K  but  lacks kinase and S H 2 or S H 3 domains (145, 146). Given the apparent similarities between P R N K and F R N K structure, it is appealing to hypothesize that they share similar functions in inhibiting their parental tyrosine kinases. Another piece of evidence supporting the possible contribution of Pyk2 to E T cell proliferation comes from recent data demonstrating that unphosphorylated Pyk2 associates with the N-terminal domain ( N T D ) of E W S in vitro (147). Furthermore, phosphorylation of Pyk2 in response to G P C R activation leads to disruption of this association and relocation of E W S from the cytoplasm to the nucleus. E W S - E T S fusions retain the N T D of E W S and, therefore, it is possible that inactivated P y k 2 may also associate with these fusion proteins. If so, phosphorylation of Pyk2 in E T cells would lead to disruption of this association and facilitate nuclear localization of the oncoprotein and transcription of E W S - E T S target genes. Preliminary immunoprecipitation-western blot experiments were done with ocPyk2 and ccFLIl antibodies to try to demonstrate E W S - F L I 1 — P y k 2 associations in E T cells in vitro but none were detected (data not shown).  However, P y k 2 was also constitutively  phosphorylated in all experimental conditions and this may account for our negative results and preclude the acquisition of conclusive results in cultured E T cells.  A s an alternate  approach, E W S - F L I 1 association with Pyk2 could be evaluated in an in vitro translation experiment where phosphorylation of Pyk2 would not present a confounding variable such as occurred in the cell-based systems. Thus, there is a potential role for Pyk2 and its related splice-variant P R N K in G R P - R signaling in E T cells and proper elucidation of this role will require further investigation.  101  Finally, it is conceivable that we were unable to link the G R P - R with mitogenic signaling pathways in E T because its role in facilitating proliferation may occur indirectly via cytoskeletal changes rather than directly via mitogenic pathways.  In a recent review  article, Jensen et al. proposed that the evidence for G R P acting as a mitogen in cancer cells is conflicting and often misleading (148). Rather, they suggest that when G R P and its receptor are aberrantly expressed by cancer cells, G R P acts primarily as a morphogen and has only very weak mitogenic activity. Morphogens are also known as differentiation factors and their primary role is in the regulation of normal embryonic development.  In cancer,  morphogens direct the differentiation of cells within the tumour, thereby retaining the malignancy in a better-differentiated state (148). Examples of morphogens in cancer that also act as weak mitogens are hepatocyte growth factor ( H G F ) , which alters the behaviour of gastric adenocarcinomas (149), and vasoactive intestinal peptide (VIP), which directs the differentiation of neuroblastoma (150, 151). Jensen et al. argue that G R P ' s role in cancer cell proliferation is primarily linked to its role in altering cell shape and enhancing cellular adhesion by altering the actin cytoskeleton (148). They propose that GRP-induced activation of p l 2 5  F A K  via the small GTP-binding proteins rac, ras and rho mediates the morphogenic  function of G R P . They further hypothesize that aberrant expression of G R P and G R P - R in cancer allows for the activation of p l 2 5  F A K  , which in turn acts to promote the cell-cell and  cell-matrix attachments that are critical for proliferation and survival. W e did not identify changes in p l 2 5  F A K  phosphorylation in E T cells treated with G R P or G R P - R antagonists.  However, subtle changes in activation levels of the kinase may not be apparent as differences in phosphorylation by Western analysis. More sensitive assessment of p l 2 5  F A K  activation,  102  such as by a radionuclide-kinase assay may reveal that ligation of the G R P - R does in fact activate p l 2 5  F A K  i n E T cells.  In summary, we have found that G R P is expressed by E T and acts as an autocrine growth factor in these tumours. The precise nature of its role in facilitating proliferation remains to be elucidated but may involve signaling through the adhesion-related kinases P y k 2 and/or p l 2 5  F A K  .  Furthermore, we have found that E T cells in culture continue to  express high levels of cyclin D and phosphorylated F A K even in the absence of mitogenic stimulation. This suggested to us that proliferative pathways may be induced in E T cells as a consequence of adherence to plastic culture dishes and the resultant changes to cell shape. W e therefore studied E T cell proliferation in non-adherent conditions to determine i f proliferative signaling pathways varied between adherent and anchorage-independent cell culture systems. The results of these experiments w i l l be discussed in the following chapters.  103  CHAPTER V PROLIFERATION IN EWING TUMOUR CELLS GROWN AS ANCHORAGE INDEPENDENT SPHEROIDS 5.1  INTRODUCTION In traditional cell culture models of sarcomas and many other human solid tumors,  cells are grown as adherent monolayers on plastic dishes in the presence of serum and other growth factors.  However, among the in vitro features that differentiate transformed cell  cultures from those of normal cells are their decreased growth factor requirements and their ability to grow in an anchorage-independent environment (152). These properties likely correlate with the clinical features of malignant tumors: that is, the ability to infiltrate surrounding tissues and to establish distant metastases (153). In an effort to better represent the anchorage independent in vivo setting of tumor cells, some investigators have utilized the three-dimensional spheroid cell culture model for studies of cancer cell biology. Spheroids are multi-cellular structures of intermediate complexity between in vivo tumors and monolayer cultures and, as such, may demonstrate biologic characteristics that are more closely related to those of primary tumors (154). W e have found that E T cells grown as traditional adherent monolayers express constitutively high levels of cyclin D I protein despite prolonged mitogen withdrawal and/or G R P - R blockade. W e also observed that p l 2 5  F A K  was phosphorylated in all adherent E T  cells regardless of mitogenic stimulation. This led us to hypothesize that cellular adhesion to plastic culture dishes may lead to constitutive activation of proliferative pathways in E T cells  104  and we therefore developed a multi-cellular spheroid model in which to study anchorageindependent E T cell proliferation. Specifically, we evaluated and compared cellular morphology, proliferative rates, and cyclin D I expression levels in both traditional and spheroid culture settings. Our results indicate that significant differences exist between the two culture models and comparison with primary tumours suggests that the spheroid model may be more representative of E T growth and proliferation in vivo.  5.2  RESULTS 5.2.1  Ewing tumor cells spontaneously form spheroids when grown in  suspension. T C 3 2 , A4573 and 5838 are well-studied E T cell lines derived from primary tumors which have been shown to express EWS-ETS and 7/5 EWS-FLI1  gene fusions.  T C 3 2 and A4573 express 7/6  fusions, respectively, while 5838 expresses an EWS-ERG  fusion (10, 104,  105). T o study growth under non-adherent conditions, confluent E T cell monolayers were trypsinized and placed as single cell suspensions in medium overlayed on agar-coated dishes (preventing  attachment  to  plastic),  and  their  growth  characteristics  monitored  morphologically. A l l cell lines behaved in a similar fashion. Within 1 hr of replating, the single cells began to form loose clumps that continued to grow in density over time. B y 24 hrs almost no single cells remained and irregular shaped clumps were evolving into tight spheroidal structures (see Figure 19). Histologic analysis of hematoxylin and eosin ( H & E ) stained sections revealed round cell morphologies in E T spheroids that were virtually identical to those of primary E T , in contrast to the spindle-shaped cells of adherent monolayers (Figure 20A).  105  •-••>•:  •••-"•>i  IF  •;  ||| ||  =)  O  CO  •  «'•  IP?** •  ... 3  f, •  •4*  O CO  106  Moreover, ultrastructural analysis of T C 3 2 spheroids showed numerous well-developed tight junctions between neighbouring spheroid cells that were indistinguishable from those of primary E T , in contrast to the poorly-formed junctions observed in monolayer cultures (see Figure 20B). The survival of E T cells in suspension was highly dependent on the formation of multi-cellular aggregates. If less than 0 . 5 x l 0 cells were initially plated in a 10 c m agar6  coated dish, the vast majority of cells remained as single cells which were dead after 6-12 hrs in suspension, as measured by trypan blue exclusion (data not shown). Spheroids readily formed when cells were plated at higher densities (at least l - 2 x l 0 cells/plate), with almost 6  100% of clumped cells remaining viable after 24 hrs while single cells were dead.  5.2.2  Adherent monolayer E T cells have significantly higher rates of  proliferation than either spheroid or primary tumor cells. Several studies have reported proliferative rates ranging from 7-14% for primary E T as measured by immunohistochemistry or flow cytometric assessment of cells in G 2 / M or Sphase (155-159). V i s u a l comparisons of E T cell growth in vitro suggested that monolayer cells proliferated much more quickly than spheroids in all three cell lines tested. Therefore, we used three different methods to determine the proliferative state of T C 3 2 cells grown in monolayer or spheroid cultures. First, using an antibody to proliferating cell nuclear antigen ( P C N A ) , which is a well-established marker of cells in S phase or mitosis (160), we performed immunohistochemical analysis of fixed monolayer and spheroid T C 3 2 cells in addition to primary E T biopsy specimens. T C 3 2 monolayer cells showed P C N A staining in a large majority of cells. In contrast, T C 3 2 spheroids and the primary E T sections showed very similar staining patterns, with fewer than 10% of cells being P C N A positive (Figure  107  o —  u  &  on  55  1  i  '  t  J =  %-7  SH  a >^  o c o  w  2 C3  .9 5  108  21A).  Second, cell cycle was assessed using F A C S analysis of D N A content in propidium  iodide stained cells.  Consistent with the P C N A results, we observed much higher  proliferative rates in monolayer compared to spheroid cells and a marked decrease in the proliferative index of spheroids in response to serum starvation. After overnight serumstarvation, F A C S analysis of D N A content revealed 4 6 % of monolayer cells to be proliferating as compared to only 9% of spheroid cells. Twelve hours of serum stimulation increased these proliferative rates to 64% and 16%, respectively (Table 8). Finally, cells were assessed for uptake of bromo-deoxyuridine (BrdU) as a measure of proliferation. A s shown in figure 21B & C , the proliferative index of monolayer cells over a 20 hr period was significantly greater than spheroid cells and the latter were more responsive to serum stimulation. In keeping with F A C S and P C N A data, 60% of serum-stimulated monolayers were proliferating compared to only 9% of serum-stimulated spheroid cells. Therefore, the proliferative index of E T spheroid cells is significantly lower than monolayer cells and is much more in keeping with the proliferative rates of primary E w i n g tumors in vivo.  5.2.3  Cyclin DI protein expression in E T cells grown in suspension requires  cell-cell adhesion and is serum-dependent. W e next wished to assess whether the observed differences in proliferation were associated with differences in cyclin D I expression in E T monolayer and spheroid cells. Constitutive up-regulation of cyclin D I has been described in numerous tumor types and is implicated in oncogenesis (reviewed in (66)). Monolayer T C 3 2 cells demonstrated high  109  FIGURE 21. The proliferative index of E T spheroids is similar to primary Ewing tumors. ( A ) Immunohistochemical analysis of a primary E w i n g tumor and T C 3 2 cells using an antibody to proliferating cell nuclear antigen ( P C N A ) demonstrating rare positive cells in primary tumor and spheroid sections (arrows), compared to frequent positive cells in monolayers. (B) T C 3 2 cells growing in monolayer and spheroid cultures were incubated with B r d U for 20 hours. Cells were then pelleted, fixed, paraffin-blocked and processed for immunohistochemical staining with an anti-BrdU antibody. Significantly greater B r d U uptake (brown-stained nuclei) was observed in monolayer cells whether or not serum was present. ( C ) Histograms of B r d U uptake in T C 3 2 monolayer and spheroid cells. The percentage of positive cells is an average of 5 high power fields (total of 1000-2000 cells) ± standard deviation.  110  CD  ocl  'um Starved  'um Stimulated  CN  •  •  CO  +  o s-  O  CU  CN  o+! o T—(  CN  CO  r>  OO  m  Q  o  CD  o  CO  CO CM  S | | B 3 SAIJISOd fll> ' 3 JO %  o CU  •a CO  oh  * •  *  • «•  *i M i  o  -i—  a  C/3  Ill Table 8:  Cell Cycle Analysis of Cultured TC32 Cells Shows Decreased Cycling and Increased Serum Sensitivity of Spheroids  Monolaver % cvcling cells CG2/M + S)  Spheroid % cvcline cells ( G 2 / M + S)  Serum starved  46  9  + 3 Hours serum  41  10  + 6 Hours serum  35  13  + 12 Hours serum  64  16  112 levels of cyclin DI expression that were independent of serum stimulation (see left panel, Figure22A). However, we observed immediate and virtually complete loss of cyclin DI protein expression when cells were placed in suspension, either under high-serum or serumfree conditions (left and middle panels, Figure 22A). Cyclin D I expression slowly increased over time coincident with cell clumping but only when cells were suspended in serumcontaining media (middle panels, Figure 22A). In contrast, cyclin D I remained undetectable even in the presence of serum when suspended cells were prevented from clumping by continuous rocking of the cultures (see right panel, Figure 22A). We also tested whether cyclin D I expression remained serum-dependent once stable spheroids had formed. TC32 spheroids were grown as above for 48 hrs in the presence of serum, and then subjected to overnight serum starvation. This resulted in a reduction of cyclin D I protein which could be recovered by addition of serum to the medium (see right panel, Figure 22B). These results indicate that, in contrast to monolayer TC32 cells which are characterized by constitutive serum-independent cyclin D I expression, anchorage-independent TC32 cells appear to require both serum stimulation and cell-cell adhesion for cyclin D I protein expression. Similar to TC32, monolayer cultures of the ET cell lines A4573 and 5838 also expressed high, serum-independent cyclin D I expression while spheroid cultures of these cell lines had lower levels of expression which were serum-responsive (see Figure 22C).  5.2.4  Differences i n c y c l i n D I expression between spheroids a n d monolayers  are p o s t - t r a n s c r i p t i o n a l a n d are associated w i t h differences i n s u b c e l l u l a r localization.  Cyclin D I protein levels are tightly regulated by both transcriptional and post-  113  FIGURE 22. Cyclin DI protein expression is dependent on cell-cell adhesion and serum stimulation in suspension cultures of E T cells. ( A ) M o n o l a y e r ( M L ) T C 3 2 cells were starved for 24 hours and then replated on agar-coated dishes at a density of 35 x l 0 cells/lOcm dish, in either serum-free (left panel), 15% serum-containing medium (right panel), or in 15% serum with continuous rocking of the plates to prevent spheroid formation (right panel). Suspension cells were then collected at the indicated time-points and analyzed for expression of cyclin D I by Western analysis. Total A K T levels were used to demonstrate equal loading. (B) T C 3 2 cells were grown in 15% serum on regular plates as monolayers or on agar-coated dishes for 48 hours to form stable spheroids. Cells were then starved for 24 hrs i n 0.25% serum, followed by treatment with (+) or without (-) 15% serum for the indicated times. C y c l i n D I levels were assessed by Western analysis as above. Total A K T levels were used to demonstrate equal loading. (C) 5838 and A4573 monolayer ( M L ) and stable spheroid cells grown i n 15% serumcontaining medium were starved for 24 hrs in 0.25% serum and then treated with (+) and without (-) serum for 3 hrs, prior to assessment of cyclin D I levels as above. Grb2 levels were used to demonstrate equal loading. 6  115 transcriptional mechanisms (66, 72, 161). In order to assess whether the observed downregulation of c y c l i n D I protein in E T spheroids might be occurring through a posttranscriptional process, we performed Northern-blot analysis of T C 3 2 cell cultures using a cyclin D I specific c D N A probe. W e observed that cyclin D I transcript levels were similar in monolayer and spheroid cells (see Figure 2 3 A ) . Moreover, neither serum starvation nor starvation followed by serum stimulation had an appreciable effect on cyclin D I m R N A levels in either culture condition. These data suggest that regulation of cyclin D I protein expression in E T cells may be largely post-transcriptional, at least in vitro. Glycogen synthase kinase 3(3 ( G S K 3 P ) phosphorylates cyclin D I at residue threonine 286, thereby targeting it for proteosomal degradation (73). This GSK3(3 activity is inhibited by A K T - i n d u c e d phosphorylation at G S K 3 P serine residue 9 (162). W e therefore compared the phosphorylation status of GSK3p(Ser9) in T C 3 2 monolayers and spheroids. A s shown in Figure 23B, no appreciable differences were observed and therefore, the reduced cyclin D I levels observed in E T spheroids could not be explained on the basis of diminished GSK3P(Ser9) phosphorylation in these cells. C y c l i n D I regulation is also dependent on shuttling of the protein between the cytoplasm and nucleus during different phases of the cell cycle (73, 76). Protein lysates were therefore prepared from the nuclear and cytoplasmic fractions of E T monolayer and spheroid cells to assess potential differences in subcellular cyclin D I localization. A s shown in Figure 2 4 A , while cyclin D I appeared to be entirely cytoplasmic in monolayer cells, in spheroid cells c y c l i n D I was almost equally distributed between the nucleus and cytoplasm. Importantly, these results also demonstrated that total c y c l i n D I levels (nuclear plus cytoplasmic fractions) were actually equivalent between monolayer and spheroid cells. This  116  Q  CO LU  siCO TJ  0  +  a co| s: + a. t/J  a  >.  re o c  1  T3 0 x: a  CO »  > ro  PQ  O c/5  117  finding was in direct contradiction to our earlier findings (Figure 22) and was initially difficult to explain. However, upon review of our protein lysis protocols, it became apparent that the standard protocol for whole cell lysis using phosphorylation solubilization buffer (PSB; see Materials and Methods for complete composition) may have led to a loss of cyclin D I in the discarded insoluble pellets. This may have occurred for two reasons. First, the single detergent P S B may have been insufficiently stringent to lyse both cytoplasmic and nuclear membranes in the same lysis procedure and nucleoplasm^ proteins would be underrepresented. Second, the final spin that is used to separate soluble proteins from insoluble fractions in whole cell lysates is done at maximal velocity or 12000 rpm, while nuclearcytoplasmic preparations call for lower speeds of only 4000 rpm. Thus, i f cyclin D I was present in insoluble cellular compartments, it would be pelleted and discarded during highspeed centrifugation in whole cell preparations.  Thus, i n the preparation of whole cell  lysates with P S B , it is possible that nuclear membranes were not effectively lysed and proteins, such as cyclin D I , that were localized to the nucleus were being precipitated in the discarded insoluble pellets. This led us to interpret that total levels of cyclin D I were less in spheroids when, in fact, they were just differently distributed within the cells. T o test this hypothesis, whole cell lysates from serum starved and serum stimulated E T cells were prepared using a modified R I P A buffer (see Materials and Methods), which is a more stringent 3-detergent buffer that w i l l solubilize subcellular fractions that may have been insoluble in P S B . These lysates expressed equal levels of total cyclin D I , confirming our hypothesis and explaining our seemingly contradictory results (Figure 24B). W e also found that the differences in cyclin D I subcellular localization corresponded to differences in the subcellular localization of P-GSK3(3 (Figure 2 4 A ) . Therefore, although total levels of P-  118  GSK3f3 remained unchanged between culture conditions, its subcellular distribution differed. W e therefore conclude that the differences in cyclin D I regulation observed in anchorage independent spheroids appear to be related to differences in its subcellular localization, likely mediated by G S K 3 p \  5.3  DISCUSSION  C e l l cycle progression of non-transformed cells is dependent on the coordinated control of cyclin—cyclin-dependent kinase complexes as induced by growth factors as well as adherence to an extracellular matrix (82).  In contrast, transformed cells are capable of  anchorage-independent proliferation (153). W e therefore reasoned that conventional in vitro growth of E T cells as adherent monolayers might be expected to activate a number of signaling pathways, secondary to phenomena such as cell-plastic adhesion and c e l l spreading, that are not actually representative of in vivo tumour growth. Several studies have indicated that anchorage-independent  survival and proliferation of tumour cells are  dependent on cell-cell adhesion which is mimicked when tumour cells form multi-cellular spheroids (reviewed in (154)and (163, 164)). W e therefore cultured E T cell lines on agarcoated plates and found that they spontaneously form multi-cellular spheroids.  Stable  spheroids demonstrated numerous similarities to primary E T including small round cell morphology, a marked increase in well-developed c e l l - c e l l junctions evident ultrastructurally, and a proliferative index of approximately 10%. Several studies have reported similar proliferative rates of 7-14% in primary E T (155-159). In contrast, proliferative rates were significantly higher in monolayer cells and these cells showed only rare poorly-formed  119  120  junctions. Our data support the concept that the in vitro growth of E w i n g tumour cells as spheroids may be more representative of primary E T than conventional monolayer cultures. The observed differences in cellular proliferation between monolayer and spheroid cells were accompanied by differences in the regulation of cyclin D I . These differences were apparent at the protein level but not at the level of m R N A indicating that the variability was largely mediated in a post-transcriptional fashion. Moreover, variations in cyclin D I protein expression between serum starved and serum stimulated spheroid cells were also due to post-transcriptional changes.  Post-transcriptional regulation of cyclin D I is largely  mediated by altered rates of m R N A translation (72), and by subcellular trafficking and proteosomal degradation (73, 75, 76). W e have shown that subcellular localization of cyclin D I varies between adherent monolayer cells and anchorage-independent spheroids, and also between serum-starved and serum-stimulated spheroid cells. In non-proliferating normal cells in GO, c y c l i n D I subcellular localization is heterogeneous and is maintained at very low levels by reduced transcriptional activity and G S K 3 ( 3 - m e d i a t e d phosphorylation and proteosomal degradation.  U p o n receipt of a  mitogenic signal, transcription is upregulated and c y c l i n D I protein is stabilized by the inhibition of GSK3(3 in an AKT-dependent manner (162, 165). Stabilized, unphosphorylated cyclin D I then localizes to the nucleus during late G l where, in conjunction with C D K 4 and C D K 6 , it phosphorylates R B allowing for R-checkpoint passage and G l - S phase transition. Once the cell is in S phase, c y c l i n D is again susceptible to G S K 3 ( 3 - m e d i a t e d phosphorylation, nuclear export and ubiquitin-targeted proteosomal degradation in the cytoplasm.  Nuclear export of cyclin D I from the nucleus is essential for completion of  regulated cell division (76). Thus, cyclin D I is predominantly cytoplasmic in cells in S-  121  phase and nuclear in cells in G l . In keeping with these observations, we have shown that rapidly proliferating monolayer E T cells express c y c l i n D I almost exclusively in the cytoplasm. In comparison, the more quiescent spheroid cells demonstrate a heterogeneous pattern of subcellular cyclin D I expression that is sensitive to mitogen stimulation. Whether there are any differences in the efficiency of proteosomal degradation of cyclin D I between E T cells in different culture milieus remains to be determined. However, there is at least one prior report of increased cyclin D I stability as a mechanism of increased cyclin D I expression in cultured sarcoma cells (166). This may prove to be another factor contributing to the constitutively high cyclin D I levels and high proliferative rates observed in monolayer E T cells. In summary, we hypothesize that the biology of E T spheroid cells closely resembles the biology of E T cells in vivo and therefore, spheroids provide a more relevant model for the study of tumour behaviour. W e have shown that multi-cellular spheroids proliferate at a much slower rate than adherent monolayer E T cells and this difference is accompanied by differences  in  the  post-transcriptional  regulation  of  cyclin  DI.  Both  the  R A S — R A F 1 — M E K — E R K and P I 3 K — A K T signaling pathways are integrally linked to cyclin D I regulation and proliferation. Therefore, we next evaluated whether differences in either or both of these pathways could be responsible for the proliferative differences between E T monolayer and spheroid cells. The results of these studies w i l l be presented in the next chapter.  122  CHAPTER VI THE ROLES OF THE RAS—RAF—MEK—ERK AND PI3K—AKT PATHWAYS IN EWING TUMOUR CELL PROLIFERATION  6.1  INTRODUCTION The signal transduction cascades governing the proliferation of E T cells remain to be  fully elucidated. In addition to our studies on gastrin releasing peptide ( G R P ) (167), other growth-factors recently implicated in E T cell growth include I G F 1 (37, 102), b F G F (40), and stem cell factor and its receptor c-kit (41, 42). The exact mechanisms by which these growth factors affect proliferation and survival in E T are unclear and remain an area of intense study among E T cancer biologists. Possible downstream effectors include phospholipase C-P2 and -(33 (43), P I 3 K (39) and mediators of T G F P signaling (26). In normal cells, proliferative and survival pathways most often converge on the R A S — R A F 1 — M E K — M A P K and/or the P I 3 K — A K T pathways.  Silvany et al. have recently reported that activation of the  R A S — R A F 1 — M E K — E R K pathway is critical for E T cell transformation and proliferation (168) and survival of E T cells is impaired in the face of P I 3 K blockade (39). It is not yet known whether there is functional redundancy or cross-talk between or among any of these pathways, although this is possibly the case. In the preceding chapter we identified that variable c y c l i n D I regulation and proliferation rates exist between E T cell spheroids and adherent monolayers.  These  differences are secondary to differences in post-transcriptional control of cyclin D I resulting in differences in subcellular localization and cell cycle passage. W e therefore investigated  123  whether differences in R A S — R A F 1 — M E K — E R K and/or P I 3 K — A K T pathways could explain our findings. Our results indicate that significant differences do exist between the two culture models. Studies with inhibitors of M E K and P I 3 K suggest that the P I 3 K — A K T pathway may be more critical than E R K activation to c y c l i n D I expression and cell proliferation in E T cells, particularly in an anchorage-independent setting.  Furthermore,  patterns of cyclin D I expression and kinase activation in a cohort of primary E T samples support the use of spheroids as an in vitro representation of in vivo growth. W e also show that transfection of a constitutively active A K T construct into adherent E T cells leads to the spontaneous formation of spheroids, providing further evidence that the A K T pathway may be a key regulator of anchorage-independent growth in these tumours. Finally, we evaluate the roles of known upstream activators of A K T in an effort to better understand which growth-receptor pathways mediate A K T phosphorylation in anchorage-independent E T cells.  6.2  RESULTS 6.2.1  The ERK1/2 M A P K and PI3K—AKT pathways are upregulated in ET  cells in suspension. Constitutive activation of the M A P kinases E R K 1 and E R K 2 and the survival factor A K T have been implicated in the uncontrolled growth of numerous malignancies including E T (39, 168, 169).  A s shown in Figure 25a (left panel), starved T C 3 2 monolayer cells  responded to serum stimulation by markedly increasing levels of E R K 1 and E R K 2 phosphorylation.  However, when these cells were placed in suspension as described in  Chapter V , E R K 1 and E R K 2 became activated in a serum-independent fashion (see right  124  panel of Figure 2 5 A ) . This constitutive activation of E R K 1 / 2 was sustained, being observed in stable spheroids that had been in suspension for 48 hrs or more (Figure 25A). Screening for phosphorylation of A K T at serine residue 473 was used to assess activation of the P I 3 K — A K T pathway in T C 3 2 cells. A K T phosphorylation was readily induced by serum in monolayer cells, but suspension of these cells resulted in A K T phosphorylation  even in the  absence of serum  (Figure 2 5 B ) .  However, A K T  phosphorylation in T C 3 2 spheroids was only transiently serum-independent, in contrast to the sustained E R K 1 / 2 activation described above. B y 48 hrs, corresponding to the formation of stable spheroids, the basal level of A K T phosphorylation had decreased to that of monolayer cells and was again serum-inducible (Figure 2 5 B ) . A n identical pattern of phosphorylation of the other major A K T phosphorylation site, threonine 308, was observed (data not shown). A s shown in Figure 25C (top panel), 5838 and A4573 spheroids showed similar increases in E R K activation relative to monolayer cultures.  A s with T C 3 2 , A K T  activation was serum-independent when cells were placed in suspension (data not shown), but then became serum-responsive after stable spheroids had formed (middle panel, Figure 25C). Interestingly, basal levels of phosphorylated A K T were higher in A 4 5 7 3 spheroid cells compared with the other E T cell lines even in the absence of serum (Figure 25C).  6.2.2  M E K inhibition does not appreciably affect cyclin DI protein expression  or cellular proliferation in E T cells. Activation of E R K 1 / 2 and A K T have both been linked to induction of cyclin D I (reviewed in (62)).  In T C 3 2 monolayers, activation of E R K 1 / 2 d i d not correlate with  changes in cyclin D I expression as the latter was already elevated whether or not serum  125  FIGURE 25. Differential activation of the ERK1 and ERK2 M A P kinases and of A K T in E T suspension cultures. Confluent E T cells were trypsinized and replated on regular or agar-coated dishes, allowed to grow for 48 hours, and then starved for 24 hrs in 0.25% serum prior to stimulation with (+) or without (-) 15% serum for 30 minutes. For those cells that were in suspension for only 1 hr, monolayer cells were starved for 24 hrs prior to being replated on agar-coated dishes for 1 hr with or without serum. Cells were then lysed for Western blot analysis using phospho-specific antibodies. (A) Western blot analysis of T C 3 2 whole cell lysates using an antibody that recognizes phosphorylated E R K 1 / 2 ( P - E R K 1 / 2 ) . Total E R K / 2 levels demonstrate equal loading. (B) Western analysis of T C 3 2 whole cell lysates for expression of A K T phosphorylated at serine 473 ( P - A K T ) . Total A K T levels were used to demonstrate equal loading. (C) Western analysis of 5838 and A 4 5 7 3 E T monolayer ( M L ) and spheroid cells treated identically as T C 3 2 cells for the 48 hr time-point, using the same a - P - E R K l / 2 and a-PA K T antibodies. Total Grb2 levels demonstrate equal loading.  126  A. Monolayer Serum:  -  Suspension lhr 48hr +  +  -  +  P - E R K 1/2  mM  ,&mmtA,  MftSS  mmm:  s$<i$M<,  VlSil; Knfet-  Total E R K 1/2  mil «mii w^^^  B.  Monolayer Serum:  Suspension lhr 48hr +  ±_  P-AKT Total A K T  C.  ML Serum: -  A 4573  5838  +  Spheroid +  ML +  Spheroid + P-ERK1/2 P-AKT  127  (which induced E R K 1 / 2 phosphorylation) was added to the cultures (Figure 2 5 A & . C ) . T C 3 2 spheroids, despite demonstrating serum-independent (constitutive) E R K 1 / 2 activation, remained dependent on serum for expression of cyclin D I (see Figure 22). Moreover, treatment of monolayer or spheroid T C 3 2 cells with the M E K inhibitor U0126 resulted in decreased E R K 1 / 2 phosphorylation but had no appreciable effects on cyclin D I protein levels (see Figure 2 6 A ) . Therefore differences in E R K 1 / 2 activation do not appear to underlie the differences in c y c l i n D I regulation observed i n T C 3 2 spheroids versus monolayers. T o confirm these findings, we next determined i f M E K inhibition altered the proliferative index of E T cells in culture.  W e therefore assessed T C 3 2 monolayer and  spheroid cells for uptake of bromo-deoxyuridine ( B r d U ) in the presence or absence of U0126. A s shown in Figure 26B, monolayer cells incubated with the M E K inhibitor U0126 had only a moderate reduction in their proliferative index while no significant effect was observed in spheroid cells.  6.2.3  PI3K inhibition blocks cyclin DI protein expression in E T cells and  significantly reduces proliferation. W e next evaluated the role of the P I 3 K — A K T pathway in cyclin D I regulation and cellular proliferation. Treatment of both monolayer and spheroid T C 3 2 cells with the PI3K inhibitor L Y 2 9 4 0 0 2 resulted in markedly reduced A K T phosphorylation and complete loss of cyclin D I protein expression, which was independent of the presence of serum (see Figure 27A). The block in cyclin D I expression was sustained even after long term serum-  128  F I G U R E 26. M E K i n h i b i t i o n has little effect on E T cell p r o l i f e r a t i o n . ( A ) T C 3 2 monolayers and stable spheroids were starved for 24 hrs in 0.25% serum and then treated with the M E K 1 inhibitor U0126 or vehicle control. F o l l o w i n g treatment, cells were incubated with (+) or without (-) serum for 30 minutes (+30') or 3 hrs (+3h) prior to cell lysis. Western blot analysis was performed using p h o s p h o - E R K l / 2 ( P - E R K 1 / 2 ) and c y c l i n D I specific antibodies. (B) T C 3 2 cells were grown for 20 hours in B r d U containing media with or without U0126 and then analyzed by immunohistochemistry for uptake of B r d U as a measure of proliferation. The percentage of positive cells is an average of 5 high power fields (total of 1000-2000 cells) ± standard deviation.  129  Spheroid  Monolayer control Serum: -  +U0126  +30' +3h  -  Sftt mm  +30' +3h  i  control  +U0126  - +30' + 3h  f|§  - +30' +3h  H k  4i P - E R K 1/2  m **** W  w week  W  Cyclin D I <••> Total A K T  ^^tt^  B. D  70  58.9+7.3  m 60  • No Inhibitor  •2 50  • +U0126  •1 40 o 30 Q.  !  9.0+4.2  20  _T  as  #  0 Monolayer  g.5±2.1  -r  Spheroid  130  stimulation of cells and similar findings were observed with another P I 3 K inhibitor, wortmannin (data not shown). L Y 2 9 4 0 0 2 treatment of A4573 E T cells resulted in a similar reduction of cyclin D I protein levels, even though A K T phosphorylation was not completely blocked by this agent (Figure 27B). Moreover, incubation of either monolayer or spheroid T C 3 2 cells with L Y 2 9 4 0 0 2 resulted in a marked reduction in cell proliferation as measured by B r d U uptake (Figure 27C). Therefore, P I 3 K activation appears to be essential for cyclin D I protein expression and cellular proliferation in both monolayer and spheroid cultures of E T cells. Interestingly, we found that P I 3 K inhibition reproducibly led to serum-independent increases  in E R K 1 / 2 phosphorylation in T C 3 2 monolayer cells.  In fact, levels of  phosphorylated E R K 1 / 2 after L Y 2 9 4 0 0 2 treatment were similar to those observed in untreated T C 3 2 spheroids (compare P - E R K 1/2 panels in Figure 2 7 A ) .  This suggests  potential cross-talk between P I 3 K — A K T and R A S — E R K 1 / 2 pathways in E T cells, which we are currently evaluating in more detail.  Despite its dramatic effects on cyclin D I  expression, L Y 2 9 4 0 0 2 did not appear to affect E R K 1 / 2 activation in T C 3 2 spheroids (see Figure 2 7 A ) , further underscoring the possibility that the P I 3 K — A K T pathway may be more critical than E R K 1 / 2 activation in regulating c y c l i n D I levels i n E T cells grown in anchorage-independent conditions.  6.2.4  Down-regulation of cyclin DI expression in PI3K inhibitor treated E T  cells is post-transcriptional but does not correlate with GSK(3 phosphorylation. Using Northern blot analysis, we examined whether inhibition of E R K 1 / 2 or P I 3 K activation resulted in altered cyclin D I transcript levels. In keeping with our earlier results,  131  FIGURE 27. PI3K inhibition blocks cyclin DI protein expression in E T cells and significantly reduces cell proliferation. ( A ) . T C 3 2 monolayers and stable spheroids were starved and then treated with a P I 3 K inhibitor (LY294002; + L Y ) or vehicle control. Prior to lysis, cells were serum-stimulated as described in Figure 26 then analyzed for expression of P - A K T , P - E R K 1 / 2 and cyclin D I . ( B ) . Protein expression and kinase activation in A4573 monolayer and spheroid cells treated with L Y 2 9 4 0 0 2 as in (A). (C). T C 3 2 cells were grown for 20 hours in B r d U containing media with ( + L Y ) or without L Y 2 9 4 0 0 2 and then analyzed by immunohistochemistry for uptake of B r d U as a measure of proliferation. The percentage of positive cells is an average of 5 high power fields (total of 1000-2000 cells) ± standard deviation.  132  Spheroid  Monolayer control Serum:  +30' +3h  +LY  control  +30' +3h  +LY  - +30' +3h  -  +30' +3h  P-AKT Cyclin D1/D2  •  M S 888  P-ERK1/2  MSB •""IRl  B.  Total A K T  Spheroid  Monolayer control +3h  Serum:  +LY  control  +LY . - +3h  +3h  +3h  ¥  P-AKT Cyclin D1/D2  C.  70  I  60  58.9+7.3  • B U  Control +LY  50 '140  o  5 30  19.3+2.0 9.0+4.2  a 20 f 10  0  1.6+1.2 Monolayer  Spheroid  133  (see Chapter V ) , cyclin D I regulation in E T cells appeared to be largely post-transcriptional, even in the presence of M E K or P I 3 K inhibitors. A s shown in Figure 28, neither M E K (Figure 2 8 A ) nor P I 3 K inhibition (Figure 28B) significantly altered levels of cyclin D I mRNA.  G i v e n the marked effect of P I 3 K inhibition on c y c l i n D I protein and on  proliferation, the PI3K-mediated regulation of cyclin D I must mainly occur at the protein level. Interestingly, when we treated either monolayer or spheroid T C 3 2 cells with P I 3 K inhibitors at levels which blocked A K T phosphorylation and cyclin D I protein expression, there were still appreciable levels of phosphorylated GSK3(3 that could be further induced with serum (Figure 2 8 C ) .  This suggests that in E T cells there may be a P I 3 K — A K T  independent pathway contributing to G S K 3 P phosphorylation.  6.2.5  Primary Ewing tumors demonstrate patterns of cyclin DI expression and  ERK1/2 and A K T activation that resemble those of ET spheroids. G i v e n the differences in c y c l i n D I expression and activation of key signaling pathways that we observed between E T cell monolayers and spheroids, we sought to determine which culture condition more closely mimicked E T growth in vivo. Seven freshfrozen primary E T biopsy samples were obtained and total protein lysates were analyzed as described for monolayer and spheroid cultures. A l l samples were obtained prior to initiation of therapy and were positive for EWS-FLll  gene fusions.  S i x cases expressed the most  common EWS exon 1IFLI1 6 gene fusion while a single case (case 6) expressed an exon 10/5 fusion (reviewed in (32)) (data not shown).  A s shown in Figure 29, high levels of  phosphorylated E R K 1 / 2 were present in all of the primary tumor samples. A K T phosphorylation could be detected in 6 of the 7 primary tumors.  Furthermore,  C y c l i n D I protein  expression appeared to correlate with levels of A K T phosphorylation in these tumors, as low  134  F I G U R E 28. Downregulation of cyclin DI in inhibitor treated cells is posttranscriptional. ( A ) C y c l i n D I transcript levels in T C 3 2 cells treated with M E K inhibitor U0126 as described in Figure 26. Northern analysis shows equal expression of cyclin D I m R N A between serum-starved and serum-stimulated monolayer and spheroid cells. A n E W S c D N A probe was used to compare total R N A levels. (B) Northern analysis shows that monolayer and spheroid cells treated with the P I 3 K inhibitor L Y 2 9 4 0 0 2 (as described in Figure 27) express equal levels of c y c l i n D I transcript. Comparison of (A) and (B) with cyclin D I transcript levels in control cells reveals no significant differences (see Figure 2 3 A ) . ( C ) Western blot of T C 3 2 monolayer and spheroid cells treated with or without the P I 3 K inhibitor L Y 2 9 4 0 0 2 (as described in Figure 2 7 A ) . Despite dramatic effects on cyclin D I protein expression P-GSK3(3(Ser9) levels are similar in all experimental conditions.  135  c CO  U  W  V,  CN  o o 0\ CN T f  >*  +  •p <u  a 00 u>^ o c o  PQ  CO  Q c  m  -a  o  •a c N  CO  i—i  O  +  3 CO CO  U  CO  136  or undetectable levels of cyclin D I were observed in several tumors even in the presence of high E R K 1 / 2 phosphorylation (see lanes 1, 3, and 7 of Figure 29). Although the sample size is small, these results provide evidence that cyclin D I expression of in vivo E T cells may be more closely linked to A K T rather than E R K activation, as was found for E T cell lines. Moreover, the pattern of E R K 1 / 2 and A K T phosphorylation as well as cyclin D I expression of the primary tumors was more in keeping with E T spheroids than monolayers, indicating that primary E T growth signaling in vivo may be better represented in vitro by anchorageindependent spheroid culture systems.  6.2.6  Expression of constitutively active A K T by ET cells leads to spontaneous formation of anchorage independent spheroids. In order to determine i f the observed effects of P I 3 K inhibition were dependent on  inhibition of A K T phosphorylation, we transiently transfected T C 3 2 cells with a myristilated-AKT ( M y r - A K T ) c D N A construct. When expressed, this Myc-tagged A K T protein localizes to the cell membrane and is constitutively activated independent of P I 3 K (170).  W e observed that transfected monolayer T C 3 2 cells spontaneously grew as 3-  dimensional foci and then anchorage-independent spheroids after 10 days in selection media (Figure 3 0 A ) . Growth as spheroids continued despite culturing cells on standard plastic culture dishes and the proliferative rate was noted to be slower than cells transfected with control vector alone (visual observations). Baseline levels of cyclin D I protein were high in both cell populations and slightly more serum inducible in M y r - A K T cells than in control transfectants (Figure 30B). Subcellular localization of cyclin D I in these cells was not assessed and it remains to be determined if differences exist.  TC32  ML SP  Primary Ewing Tumors  1 2  3  4  5  6  7  F I G U R E 29. Primary E w i n g tumors express activated E R K 1 / 2 & A K T and variable levels of cyclin D I . Total protein lysates were obtained from 7 primary E T samples as detailed in Materials and Methods. Western blots for expression of P - E R K 1 / 2 , P - A K T , G S K 3 P phosphorylated at serine 9 ( P - G S K 3 P ) , and cyclin D I were performed as for E T cell lines, and show patterns similar to E T spheroid cultures. T C 3 2 cell lysates from monolayer ( M L ) and spheroid (SP) cells stimulated for 3 hrs with serum are used for comparison.  138  S2 < l  I <  0,1  c o U  fa 6  u  139  Treatment of transfected cells with L Y 2 9 4 0 0 2 was carried out exactly as described for nontransfected cells. A s shown in Figure 31, while P I 3 K inhibition effectively blocked cyclin D I expression in control transfectants, T C 3 2 cells expressing M y r - A K T continued to express cyclin D I , albeit at diminished levels. This suggests that PI3K-mediated regulation of cyclin D I is, at least in part, dependent on A K T activation.  6.2.7  Preliminary studies evaluating the role of integrin-linked kinase (ILK) in  A K T phosphorylation in anchorage-independent E T cells. A s discussed in Chapter I, full activation of A K T requires phosphorylation of Ser 473 and this may be facilitated by the serine-threonine kinase I L K (87, 171).  Furthermore,  activation of I L K can induce anchorage independent growth (171). H a v i n g observed that anchorage-independent growth of E T cells was correlated with increased phosphorylation of A K T , we hypothesized that the cell-cell adhesions responsible for proliferation in spheroids may act through integrin-mediated activation of I L K . Therefore, we expressed a dominantnegative I L K c D N A construct in T C 3 2 E T cells and observed its effect on A K T (Ser 473) phosphorylation in both monolayer and spheroid cell cultures. This D N - I L K c D N A has been previously demonstrated to be effective in diminishing native I L K activity and A K T phosphorylation in other cell types (87). A s shown in Figure 32, T C 3 2 cells expressed high levels of D N - I L K (lanes 1-3 and 7-9) following transient transfection using lipofectamine. However, despite readily detectable D N - I L K expression, no differences in A K T activation were observed between D N - I L K and control transfectants.  Specifically,  140  P c  "o  >> U  o o " o< \t om CN  H >^ M P ^  cn +  s s  4—»  c o  U  H CN  o  <  < S "S 11 in  00 ^  C  J  T3  P  o o  ^  B  l-i  C O  " 3 S .§ £  O "  IE CO  3  .2 E  C N  T3  C3  o .3  c  >> U  +  +  ?^  bo o  o  o  1*8 c  c n  pi  •4—<  c  5  oo o oo  +  oo r-  O  o  a 1  O CO  0\  +  o < _1 O  cn  CU  w  CO  CO  3  3  cn  I ts s Tj  o  oo p-,  jo x) c 1c c « iZ  .S o  c  15 cw w  P  M  o o  c  C N O  o C N  3 +  W  CO  + C  o  U  w  O ro +  O ro +  I I  en S  1  8 Lu ^  a  CO  H  o  ,23 *  U  - J  "2 c  B  ed  I  2  c3  5  S3 fa  s hi  .2 CO  cd 00  1  M  6 o  T3  to  O  c  g£  —  "E,  o Q i) s  o > C N c n  e£ o  E  c n  c 0 s e 00  ^ 4 — >  ^oo  3  "cd  E  cd  E  c .S  ed  SH CU  CO  «  >,p °  o  .s ^  141  T C 3 2 cells grown in suspension culture as spheroids demonstrated the same levels of serumindependent A K T activity regardless of expression of D N - I L K (Figure 32, lanes 7-12).  6.3  DISCUSSION The R A S — - E R K 1/2 and P I 3 K — A K T pathways both become activated in numerous  cell types in response to diverse extra-cellular stimuli, and aberrant activation of these signaling cascades is also associated with cellular transformation (88). A well-established link exists between R A S activation and induction of cyclin D I expression in different cell types (172-175). This is thought to occur predominantly via transcriptional regulation, with activation of the M A P kinase pathway leading to induction of the cyclin D I promoter (reviewed in (62)). Paradoxically, however, constitutive activation of E R K 1 / 2 did not correlate with increased cyclin D I expression in E T spheroids. Rather, cyclin D I protein expression in spheroids appeared to be dependent on c e l l - c e l l adhesion and serum stimulation in a largely E R K 1 / 2 M A P kinase independent manner. Blockade of M E K 1 had no effect on monolayer cyclin D I levels and only minimal effects on spheroid cyclin D I expression. M E K 1 inhibition did not alter levels of cyclin D I m R N A in either culture setting and only slightly decreased proliferation rates.  These results indicate not only that the  R A S — R A F 1 — M E K — E R K 1 / 2 cascade is regulated in a different manner in monolayer and spheroid E T cells, but also that early cell cycle progression of E T cells may not be critically dependent on E R K activation. This was corroborated by examination of a set of primary E w i n g tumors, which demonstrated high levels of E R K 1 / 2 activation that also did not correlate with levels of cyclin D I expression. It is conceivable that adhesion of E T cells to  142  W)  C3 *3  CO I i n  oo, H  >  "3 •J  o H  o  <D + >  P  i  Z Q  cl  2 O c + o 53 U oo  31  W +  o P CU  ll  OO  CO  P  o <2 o > c o U CU C/3  "o c o  +  d p  o  CN  143  plastic induces cyclin D I expression, via alternate pathways to those mediated by cell-cell adhesion in E T spheroids or primary tumors.  One possibility is that cell shape changes  induced by adherence of E T cells to plastic facilitate constitutive cyclin D I expression and cellular proliferation by activating proteins that link the actin cytoskeleton to the cell cycle. Potential candidates include the GTPases Rho or Rac (51, 176) and the adhesion-related cytoplasmic protein tyrosine kinase p l 2 5 " 1  AK  (177). The relative activation of these proteins  in E T monolayers versus spheroids and their relevance to primary tumors remains to be determined.  Our results are in contrast to those of Silvany et al, who found that E R K 1 / 2  activation was constitutive in cells transformed by E W S - F L I 1 , and that M E K 1 inhibition blocked soft agar colony formation of these cells (168).  However, these studies were  performed using murine N I H 3 T 3 cells expressing E W S - F L I 1 , as opposed to the human E T cell lines used in our studies. Therefore the observed variation might be cell type specific, and the requirement for E R K 1 / 2 activation in N I H 3 T 3 transformation may involve pathways that are inherently different from those in E T cells. P I 3 K inhibition of monolayer or spheroid E T cells resulted in virtually complete loss of cyclin D I protein expression and a marked reduction in proliferative rates. Consistent with these findings, A K T phosphorylation was readily detectable in 6 of 7 primary E T samples" and appeared to correlate with cyclin D I expression.  Our data are  therefore  consistent with a role for the P I 3 K — A K T pathway in cyclin D I regulation and cell cycle progression of E T cells in vitro and in vivo.  Muise-Helmericks et al. have reported that  cyclin D I expression is controlled via a P I 3 K — A K T dependent pathway in breast and colon carcinoma cell lines (72). They found that translation of cyclin D I m R N A was enhanced in these cells by a P I 3 K — A K T dependent pathway.  C y c l i n D I down-regulation in both  144  monolayer and spheroid E T cells in response to P I 3 K inhibition was post-transcriptional as m R N A levels were not altered by L Y 2 9 4 0 0 2 . It is known that GSK3(3 inhibits cyclin D I expression by a mechanism involving reduced gene transcription and more directly by phosphorylation of cyclin D I threonine 286 leading to enhanced protein degradation (73, 161). Furthermore, these activities are known to be inhibited by A K T phosphorylation of GSK3(3 at serine residue 9 (162, 165). We found no differences in the phosphorylation status of G S K 3 P at this residue in E T spheroids versus monolayers, even under conditions in which cyclin D I levels were dramatically different. W h i l e this would suggest that the regulation of c y c l i n D I expression in E T is independent of GSK3(3, it is also possible that other phosphorylation sites on GSK3(3 regulate its kinase activity and substrate specificity in E T cells.  A l t e r n a t i v e l y , differences in subcellular localization of G S K 3 P  rather than  phosphorylation status may be important in regulating cyclin D I protein expression in E T spheroid cells (see Figure 24). Interestingly, we observed high level phosphorylation of G S K 3 P ( S e r 9) even when A K T activation was blocked by P I 3 K inhibitors. This indicates that there may be an additional level of complexity involving P I 3 K — A K T independent phosphorylation of GSK3p(Ser9) in E T cells. P I 3 K blockade in monolayer cells was also associated with constitutive activation of E R K 1 / 2 , suggesting that inhibition of P I 3 K not only blocks cyclin D I expression but at the same time activates the R A S — R A F 1 — M E K — E R K 1 / 2 pathway in these cells.  It is  becoming increasingly apparent that the P I 3 K — A K T and the R A S — R A F 1 — M E K — E R K pathways are integrally linked (reviewed in (58)).  In fact, several recent studies have  demonstrated negative regulation of R A F 1 through phosphorylation by A K T (59, 60). Also, Treinies et al. demonstrated that while an activated form of M E K can induce cyclin D I  145  expression and cell cycle entry in N I H 3 T 3 cells, P I 3 K signals are required for this effect (61). It is well established that constitutive expression of R A S can induce apoptosis or cell cycle arrest through mechanisms involving p l 9  A R F  , p21  CIP1  , and p53 (reviewed in (62)). A n  emerging model is that unless the P I 3 K - A K T pathway is activated in parallel, continuously high R A S activity w i l l lead to cell cycle arrest or apoptosis (58, 61, 63). It is therefore possible that the observed effect of P I 3 K on cyclin D I regulation in E T cells is secondary to suppression of R A F 1 to a level where E R K 1 / 2 activation induces cell cycle progression and not cell cycle arrest (or apoptosis). In this scenario, P I 3 K inhibition of E T cells would lead to higher levels of E R K 1 / 2 activation (as observed), which would in turn drive the cells into arrest rather than cell cycle progression.  It is therefore apparent from our results that  regulation of proliferation of E T cells in vitro  i n v o l v e s cross-talk between  R A S — R A F 1 — M E K — E R K and P I 3 K — A K T pathways.  the  The precise mechanisms of this  cross-talk are still being investigated; however, our studies thus far suggest that P I 3 K plays a key role in E T cell proliferation, as has already been described in other non-malignant cell model systems (61). We have also sought to determine whether the role of P I 3 K in the regulation of cyclin D I and proliferation in E T may be dependent on as yet undetermined effectors other than AKT.  This possibility was raised following several observations that suggested cyclin D I  might be expressed in a PI3K-dependent but AKT-independent fashion in E T cells. First, we found that while serum starvation of T C 3 2 monolayers or stable spheroids resulted in markedly reduced A K T phosphorylation this was not accompanied by a significant reduction in cyclin D I expression (see Figure 27A). Second, although suspension of E T cells resulted in immediate activation of A K T , this was accompanied by a complete loss of cyclin D I  146  expression (see Figures 22 and 25). Third, treatment of A4573 spheroids with L Y 2 9 4 0 0 2 resulted in blockade of c y c l i n D I despite having only a moderate effect on A K T phosphorylation (see Figure 27B). In an effort to test our hypothesis of PI3K-dependent, AKT-independent c y c l i n D I regulation, we overexpressed a constitutively active A K T c D N A in an E T cell line.  Interestingly, expression of M y r - A K T led to spontaneous,  preferential growth of T C 3 2 E T cells as multi-cellular spheroids rather than adherent monolayers.  This suggests that A K T activation is integrally i n v o l v e d in anchorage-  independent growth in E T cells and further corroborates our earlier observations. Cyclin D I was eliminated entirely in P I 3 K - i n h i b i t o r treated control transfectants but was only moderately reduced in M y r - A K T expressing cells. This could support the existence of a secondary PI3K-dependent pathway which acts in parallel with A K T to maintain cyclin D I levels. Alternatively, it is possible that reduction of cyclin D I protein levels in the P I 3 K inhibitor treated M y r - A K T cells occurred simply because of the inhibitor's effects on native A K T activation rather than on an alternative pathway. Either way, we have still not ruled out the possibility of a parallel A K T - i n d e p e n d e n t pathway.  M o r e studies are needed to  rigorously test this possibility and to determine what targets of P I 3 K other than A K T might be involved in cyclin D I regulation in E T cells. C e l l survival studies confirmed that E T cells could not survive indefinitely as single cells in suspension i f they were prevented from clumping. One well-established function of A K T is to act as a survival factor (178). This role of A K T in cell survival has already been reported in E T (39), likely functioning through the insulin-like growth factor 1 receptor (37). We found that activation of A K T in monolayer T C 3 2 cells was serum-dependent, but that transfer of cells to suspension cultures led to immediate serum-independent A K T activation.  147  This was transient with A K T activation becoming serum-dependent once stable spheroids had formed. It is therefore possible that this transient activation of A K T allows cells to survive until they can clump and form cell-cell junctions that are critical for continued survival and cell cycle progression (163, 164). W o n g et al. have shown that transformed cells which overexpress the anti-apoptotic protein B c l - 2 have a greater propensity for the formation of lung metastases, presumably because they can evade apoptosis long enough to allow successful implantation (179). Activation of the A K T survival pathway may be an important mechanism facilitating metastasis of E T cells in vivo by allowing potentially metastatic anchorage-independent cells to survive until they can adhere to the extra-cellular matrix or to other circulating tumor cells. The upstream molecules and pathways involved in the activation of A K T in anchorage-independent  E T cells remains unclear.  The i n t e g r i n — I L K — P I 3 K — A K T  pathway is an obvious candidate given the role of integrins in cell-cell adhesion and the association of I L K activation with stimulation of anchorage independent growth (83, 87, 171). However, we found that expression of a dominant-negative I L K construct did not lead to reduced levels of A K T activation in spheroid E T cells. These preliminary results suggest that the increase in A K T activation observed in anchorage-independent E T cells may not be secondary to I L K activity or, therefore, to integrin signaling. However, since we did not measure I L K activity in our experiments, we cannot be certain that D N - I L K transfection efficiently downregulated kinase activity. Thus, given the importance of cell-cell adhesion to anchorage-independent E T cell proliferation, further studies of I L K , integrins and integrinmediated signaling are warranted before they are conclusively eliminated as upstream activators of A K T in E T spheroid cells.  148  In conclusion, our results indicate that proliferation of E T cells may be better studied in spheroid cultures where a number of physical and biochemical characteristics of the cells more closely resemble those of primary tumor cells. Compared to conventional monolayer cultures, key differences in the regulation of cyclin D I expression and early cell cycle progression exist in this anchorage-independent model and at least some of the differences likely occur as a direct consequence of cell-cell adhesion and the observed variations in key signal transduction pathways.  These cell-cell interactions involve as yet undetermined  adhesion molecules that may initiate signal transduction cascades such as the P I 3 K — A K T pathway to ensure survival and proliferation of the tumor cells in their anchorageindependent milieu. It is also probable that E T cells grown in suspension have an increased dependency on autocrine growth pathways such as those mediated by I G F 1 and G R P given their increased sensitivity to mitogen withdrawal.  Such key differences in anchorage  independent growth signaling pathways will need to be considered as novel pathway-targeted biologic therapies for E w i n g tumors are developed.  149  CHAPTER VII SUMMARY & FUTURE DIRECTIONS  7.1  General Summary The Ewing family of tumours presents a challenge to clinicians and cancer biologists  alike as E T continues to be a difficult tumour to cure despite modern multi-modal approaches to therapy. Cytotoxic therapies have been maximized to the degree that therapy associated morbidity and mortality rates prohibit further increases in drug or radiation doses.  Novel  therapeutic approaches are necessary and this requires that we first develop a better understanding of E T cell biology in order to develop targeted therapies. While the genetics of E T have been largely characterized and the identification of EWS-ETS  gene fusions has  improved diagnostic capabilities, very little is yet known about the function of these chimeric oncogenes, which likely represent the primary genetic events in E T pathogenesis. Moreover, we have yet to understand the mechanisms behind the dysregulated growth of E T cells that afford these tumours their characteristic local invasiveness and high metastatic potential. In the past decade there has been an explosion of knowledge in the field of normal cell biology and in particular, in our understanding of the regulatory processes governing cell proliferation. For this thesis I have used this general knowledge of proliferative signaling pathways and cell cycle regulation in my studies of E T in an attempt to understand how they may be dysregulated in E T cells. Specifically, I have characterized the expression of the growth factor G R P by E T cells and have identified it to be an autocrine growth factor in these tumours. More generally, I have studied the proliferation of E T cells in two different  150  cell  culture  models  and  have  attempted  to  R A S — R A F — M E K — E R K M A P K and P I 3 K — A K T proliferation in vitro.  define  the  contributions  of  the  pathways to the control of E T cell  In so doing, I have identified that growth regulation is different  between cells grown as traditional adherent monolayers and those grown as anchorageindependent multi-cellular spheroids. I have also determined, based on morphologic, ultrastructural and functional data, that the proliferation of E T cells in vivo may be better represented in vitro by anchorage-independent spheroids. Based on these results, I believe that further studies of growth regulatory pathways in anchorage-independent E T cells are warranted. In particular, I hypothesize that identification of adhesion molecules involved in cell-cell and c e l l - E C M adhesion and characterization of their respective signaling pathways w i l l afford key insights into the regulation of E T cell growth in vivo.  In the following  sections, potential avenues of future investigation w i l l be discussed with reference to the major observations presented in this thesis.  7.2  G R P Autocrine Growth Signaling in E T Our studies of E T confirmed expression of GRP and its receptor in 100% of cell lines  and approximately 50% of primary tumour samples.  A l t h o u g h a clinico-pathologic  correlates study d i d not identify a relationship between GRP/GRP-R  expression and  prognosis, in vitro and in vivo experimental evidence suggested that the peptide functions as an autocrine growth factor in E T . These observations are interesting for several reasons. First, G R P is normally secreted by embryonic and adult cells of neuroendocrine lineage and it has been previously demonstrated to act as an autocrine growth factor in malignancies of neuroendocrine origin such as S C L C , breast cancer and prostate cancer (see (127) for  151  review).  This provides further evidence for the putative neural origins of E T . Second,  investigators studying S C L C have recently published that detection of GRP by R T - P C R in peripheral blood, sputum and cerebrospinal fluid can be used as a sensitive marker of microscopic residual or metastatic disease (180 181). Thus, although G R P may not be a ;  prognostic marker in E T , there is potential for its use in tumour diagnosis and detection, particularly when gross disease is not apparent. Finally, recognition of G R P - R detection in E T may provide therapeutic opportunities. Direct blockade of the G R P - R has resulted in reduced growth of several different tumour types, including E T (120, 133-135, 167). It is conceivable that patients with bulk disease may not respond dramatically to G R P - R blockade given the relatively small contribution of G R P signaling to overall tumour growth. However, the use of G R P - R antagonists in E T patients with microscopic or minimal residual disease may delay tumour recurrence and should be evaluated in clinical trials similar to those being conducted in prostate cancer (182). Moreover, cytotoxic bombesin-analogues have been developed that specifically target cells expressing G R P - R (136, 183).  Such agents may  prove to be useful in the development of targeted therapies that are specifically cytotoxic to E T cells. O f course, the development of optimal targeted therapies that exploit GRP-mediated growth w i l l require a more complete understanding of GRP-signaling pathways in E T cells. Our evidence supports the role of G R P as a growth factor but does not directly implicate the R A S — R A F — M E K — E R K pathway, which has been most often linked to G P C R - i n d u c e d mitogenesis.  Rather, our results and those of others suggest that the adhesion-related  proteins Pyk2 and/or p l 2 5  F A K  may be key regulators in GRP-induced proliferation (50, 184).  The potential importance of adhesion-mediated signaling pathways in the mitogenic and  152  morpnogenic functions of GRP has already been proposed for other cancer cell systems (148, 185). Additionally, studies of the related neuromedin-B receptor have demonstrated that integrity of the actin cytoskeleton is necessary for neuromedin-B stimulated phosphorylation of p l 2 5  F A K  and that the Ras-related small GTP-binding protein Rho is at least partially  involved (186). Preliminary studies by our collaborator Dr. C. Poremba have shown that GRP-positive primary ET tumours express higher levels of Rho than do GRP-negative tumours and that elevated Rho is associated with decreased levels of expression of the cell cycle inhibitor p21  CIP1  (unpublished observations). Thus, we hypothesize that GRP signaling  in ET likely facilitates cell cycle passage via Rho—pl25  FAK  —p21  CIP1  dependent mechanisms  and that future studies should focus on evaluation of this pathway. Furthermore, the effect of GRP on this pathway may be critically linked to the actin cytoskeleton and, therefore, to cellcell and/or cell-ECM adhesion. Given the significant changes in morphology and cell-cell adhesion that we have demonstrated in ET cells grown as adherent monolayers, I believe GRP signaling in E T cells would be better studied in multi-cellular spheroids where cell-cell and cell-ECM adhesion may be more representative of primary in vivo tumours. Other potential mechanisms of GRP-mediated growth in E T are also worthy of consideration. First, the hypothesis tested in this thesis presumes that GRP exacts its effects by acting directly on the ET cells themselves in a pro-proliferative fashion. However, it is possible that G R P provides a survival signal to ET causing a reduction in the rate of apoptosis rather than a true increase in proliferation. While there is no evidence to support the role of GRP as an anti-apoptotic factor in other normal or malignant tissues, this potential mechanism of action should be definitively studied in ET before it is ruled out completely. Moreover, it is possible that the pro-proliferative effects of GRP are mediated by interactions  153  with the E C M or with other surrounding cells. These effects would not be apparent in the current adherent in vitro model and future studies could evaluate the role of G R P on the expression of growth inhibitory molecules, such as TGF(3 and the interferons, by neighbouring E T cells, lymphocytes or other non-tumour cells. It is also not clear from our studies whether G R P is constitutively expressed by E T cells or whether its expression is in some way regulated, e.g. in a cell cycle or differentiation-state-dependent fashion. If G R P expression and secretion are regulated in E T cells in vivo, it is plausible that random, nonregulated blockade and stimulation of the GRP-receptor in these in vitro studies may not have an appreciable effect on cellular biochemistry. Finally, the potential relationship between G R P and E W S - E T S proteins requires further study. The ability of EWS-FLI1  to transform fibroblasts depends on the presence of  the IGF-1 receptor and in its absence, apoptosis is induced (37). It remains to be determined whether G R P also in some way contributes to the permissive cellular envrionment facilitating E W S - E T S  induced malignant transformation.  Studies evaluating the  transformation efficiency of EWS-ETS genes in cells expressing and not expressing G R P and its receptor would help to clarify this relationship.  7.3  The Role of Adhesion Molecules in E T Proliferative Signaling The results of our studies on GRP-mediated signaling and the conclusions that cell  shape, cellular adhesion and the actin cytoskeleton might be critical regulators of this pathway led us to hypothesize that the model we were using may not have been representative of the in vivo situation. Therefore, we developed an anchorage independent multi-cellular spheroid model in which to study E T cell growth. Previous studies of other cancer cell spheroids have demonstrated that they express cell surface adhesion molecules  154  and E C M components of an intermediate level between monolayer cultures and in tumours (154, 164, 187, 188).  vivo  Our studies of E T cell morphology, ultra-structure and  proliferation confirmed that spheroid cells were more similar to primary tumour cells than were monolayer cells.  Adhesion of tumour cells to one another or to components of the  E C M results in the activation of adhesion molecules and their down-stream signaling cascades. The identity of these molecules and their respective signaling pathways remains to be determined in E T . However, we believe that there must be differences between multicellular spheroids and adherent monolayers cells and that these differences most likely result in differential activation of the P I 3 K — A K T pathway. Ongoing studies in our laboratory are currently testing this hypothesis. To date, there has been little investigation into adhesion molecules and their signaling pathways in E T . There is no published evidence to support expression of either e-cadherin or n-cadherin in these tumours and studies of N - C A M expression have been contradictory (189, 190). Reports of integrin expression by E T cells have also been conflicting (190, 191); however, the expression of high levels of I L K by primary E T tumours (192) suggests that integrin signaling may be important in this tumour family. Moreover, the direct link between integrins, I L K and the P I 3 K — A K T pathway (87) provides further evidence to support the study of integrins in E T adhesion-related signaling. Previous studies have demonstrated that E T cells synthesize both laminin and fibronectin, components of the E C M (193). Preliminary analysis of gene expression profiles of E T monolayer and spheroid cells using microarray analysis has revealed increased expression of the fibronectin gene in spheroid cells (unpublished observations from our laboratory).  O n g o i n g studies have confirmed increased expression of fibronectin in  155  spheroids by both Western blot and fluorescent immunocytochemistry.  One aspect of  integrin signaling that is worthy of further investigation, therefore, is the association between fibronectin—integrin binding and protection from apoptosis or anoikis (194, 195). Anoikis is the apoptotic cell death that is induced upon loss of adhesion to an extracellular matrix (196).  A s transformed cells, E T cells have established mechanisms by which to avoid  anoikis.  It is possible that integrin molecules on the surface of E T cells bind to  autonomously produced fibronectin, thus protecting the cells from anoikis. This effect would be critical to the survival of E T cells in suspension and would therefore be expected to be more apparent in spheroids. Integrin-mediated protection from anoikis can be controlled mechanistically in different ways, depending on the cell and integrin type, including upregulation of B c l - 2 , suppression of p53 induced apoptosis and activation of She (see (195) for review). Further studies of integrin expression are warranted in the study of E T cell biology i n general and in the study of anchorage-independent  E T growth signaling  specifically. Understanding the mechanisms of survival of anchorage-independent E T cells could greatly impact on the development of novel targeted therapeutic agents for the treatment of metastatic disease.  7.4  EWS-FLI1,  Cyclin  DI  Regulation  and  Subcellular  Compartmentalization W e have found that subcellular localization of cyclin D I and P-GSK3(3 varies between monolayer and spheroid E T cells, consistent with the observed variations in cell cycle status. These differences only became apparent when we specifically looked at cytoplasmic and nucleoplasm^ cellular fractions by Western blot analysis. The functional  156  status of cyclins and other cell cycle regulators is known to be influenced by their subcellular location (197), and this must be considered when studies of cell cycle regulation are interpreted. F r o m our results we can conclude that evaluation of the function of cell cycle regulatory proteins in E T requires an assessment not only of their phosphorylation status but also of their localization within the cell. Moreover, proteins that associate themselves with insoluble membrane fractions or that localize to the nucleus require cell lysis with an appropriately stringent lysis buffer if they are to be detected by standard immunoblotting techniques. The subcellular localization of E W S - F L I 1 and its role in cyclin D I regulation in E T cells also requires further investigation. Treatment of E T cells with EWS-FLI1  antisense  results in G l arrest and reduced proliferation (120, 133-135) and, it has been recently reported that cyclin D I may in fact be a downstream target of E W S - F L I 1 (198). H o w E W S F L I 1 affects cyclin D I levels is not yet known but may involve either transcriptional and/or post-transcriptional mechanisms. In the past decade, much effort has been placed into the identification of transcriptional targets of E W S - F L I 1 . This approach has not met with great success for, although putative target genes have been identified, the mechanism(s) of action of E W S - E T S fusions remain unknown. Silvany et al. have observed that E R K 1 / 2 activation is dependent on the ability of E W S - F L I 1 to translocate to the nucleus and to bind D N A (168).  Shuttling of the E W S - E T S oncoproteins between cytoplasm and nucleus may,  therefore, be a critical parameter in the regulation of E T cell proliferation. T o date, in vitro studies of E W S - F L I 1 localization and target gene identification have been done exclusively in cell lines grown as adherent monolayers.  In this thesis, I have presented data  demonstrating that the subcellular localization of proteins varies in E T cells between  157  adherent and anchorage independent culture conditions.  I have also shown that the  regulation of cyclin D I is similarly variable. Therefore, it is possible that the subcellular localization of E W S - E T S fusion proteins varies between models and this variability may contribute to the differential regulation of cyclin D I expression. It is known that multiple domains of the EWS-FLI1  fusion gene are required for transformation and that protein-  protein interactions i n v o l v i n g the N T D - E W S segment are key to the function of the oncoprotein (reviewed in (11)).  E W S and the related protein T L S (also involved in  oncogenic gene fusions (199, 200)), are both R N A - b i n d i n g proteins that appear to be involved in the splicing and stabilization of m R N A (201).  Disruption of these genes by  chromosomal translocation may alter m R N A processing and, therefore, affect protein expression (202). Further investigation into the potential role of E W S - E T S fusions in protein modification needs to be pursued. In particular, whether E W S - E T S proteins contribute to the post-transcriptional regulation of cyclin D I expression should be among the first questions addressed.  This w i l l require studying E T cells in a system in which the fusion proteins  readily shuttle to the cytoplasm. Based on the data presented in this thesis, I believe that multi-cellular spheroids will provide such an in vitro model. It w i l l also be interesting to assess whether P y k 2 affects E W S - E T S subcellular localization. It is known that nuclear localization of wild-type E W S is influenced in neural cells by the phosphorylation status of Pyk2 (147). W e have shown that Pyk2 is expressed by E T cells and that its pattern of activation may be affected by G R P - R signaling. Further studies measuring P y k 2 kinase activity in E T cells are warranted as it is possible that activation of P y k 2 may affect the subcellular localization of E W S - E T S oncoproteins and facilitate oncogenesis and maintenance of the malignant phenotype.  Once again, it w i l l be  158 important to study this relationship in multicellular spheroids where Pyk2 activation and EWS-ETS location may be more representative of in vivo tumours. Finally, it should be determined whether regulation of cyclin D I in E T is integrally linked to the IGFI-R autocrine growth pathway. PI3K is activated downstream of the IGF1R and an intact IGFI-R is necessary for EWS-FLI1 induced cellular transformation (37, 203). This thesis has identified a critical role for the P I 3 K — A K T pathway in the posttranscriptional regulation of cyclin D I in ET, particularly in anchorage-independent cells. Thus, the relationship between EWS-ETS fusions, the IGFI-R and their potential interaction in regulating cyclin D I protein expression and ET cell proliferation are areas of study being further pursued in our laboratory.  7.5  Final Comments For this thesis I have studied growth signaling pathways in E T cells in vitro and have  gained significant insights into potential mechanisms of proliferative control in these cells. I have determined that the neuropeptide G R P functions as an autocrine growth factor in these tumours, that proliferation differs markedly between anchorage-independent and adherent monolayer cultures, and that the P I 3 K — A K T pathway is a key regulator of cyclin D I expression and cell cycle progression in E T cells. Consideration of these results as a whole has implicated cellular adhesion molecules and their signaling pathways in the regulation of ET cell proliferation and provides many potential avenues for further study. Understanding the relationship between these pathways and EWS-ETS gene fusions will provide critical information about the mechanism of action of these primary genetic lesions. When we understand how the cellular biology of E T cells is altered in response to expression of the  159 chimeric oncoproteins we w i l l be in a position to develop novel, targeted  therapeutic  approaches. The development of such therapies is the only hope for advancing our progress against this family of extremely aggressive, highly metastatic tumours.  160  REFERENCES 1. Huchcroft, S., Clarke, A . , M a o , Y . , Desmeules, M . , Dryer, D . , Hodges, M . , Leclerc, J. M . , M c B r i d e , M . , Pelletier, W . , and Yanofsky, R. This Battle which I Must Fight: Cancer in Canada's Children and Teenagers. Ottawa: Supply and Services Canada, 1996. 2. Pizzo, P. A . and Poplack, D . G . Principles and Practice of Pediatric Oncology. , Third Edition edition, pp. 1522. Philadelphia, P A . : Lippincott-Raven, 1997. 3. de Alava, E . and Gerald, W . L . Molecular biology of the Ewing's sarcoma/primitive neuroectodermal tumor family, J C l i n Oncol. 18: 204-13., 2000. 4. Lawlor, E . R., Mathers, J. A . , Bainbridge, T., Horsman, D . E . , K a w a i , A . , Healey, J. H . , Huvos, A . G . , Bridge, J. A . , Ladanyi, M . , and Sorensen, P. H . Peripheral primitive neuroectodermal tumors in adults: documentation by molecular analysis, J C l i n Oncol. 16: 1150-7,1998. 5. Zucman-Rossi, J., Batzer, M . A . , Stoneking, M . , Delattre, O., and Thomas, G . Interethnic polymorphism of E W S intron 6: genome plasticity mediated by A l u retroposition and recombination, H u m Genet. 99: 357-63., 1997. 6. Kovar, H . , Aryee, D . , and Zoubek, A . The E w i n g family of tumors and the search for the Achilles' heel, Curr Opin Oncol. 11: 275-84., 1999. 7. Triche, T. J. Pathology of Pediatric Malignancies. In: P. A . Pizzo and D . G . Poplack (eds.), Principles and Practice of Pediatric Oncology, V o l . 2, pp. 115-152. Philadelphia: J.B. Lippincott, 1993. 8. Cavazzana, A . O., Miser, J. S., Jefferson, J., and Triche, T. J. Experimental evidence for a neural origin of Ewing's sarcoma of bone., A m J Pathol. 127: 507-518, 1987. 9. Aurias, A . , Rimbaut, C , Buffe, D . , Dubousset, J., and Mazabraud, A . Chromosomal Translocations in Ewing's Sarcoma, N Engl J Med. 309: 496-497, 1983. 10. Whang-Peng, J., Triche, T. J., Knutsen, T., Miser, J., Douglass, E . C , and Israel, M . A . Chromosome translocation in peripheral neuroepithelioma, N . E n g l . J. M e d . 311: 584585, 1984. 11. Sorensen, P. H . B . and Triche, T. J. Gene fusions encoding chimaeric transcription factors in solid tumours, Seminars in Cancer Biology. 7: 3-14, 1996. 12. 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 . Gene fusion with an E T S D N A binding domain caused by chromosome translocation in human cancers., Nature. 359: 162-165, 1992. 13. Zucman, J., Delattre, O., Desmaze, C , Plougastel, B . , Joubert, I., Melot, T., Peter, M . , De Jong, P., Rouleau, G . , Aurias, A . , and Thomas, G . Cloning and characterization of the Ewing's sarcoma and peripheral neuroepithelioma t( 11 ;22) translocation breakpoints, Genes Chromosomes Cancer. 5: 271-7, 1992. 14. Sorensen, P. H . B . , Lessnick, S. L . , Lopez-Terrada, D . , L i u , X . F . , Triche, T. J., and Denny, C . T. A second Ewing's sarcoma translocation, t(21;22), fuses the E W S gene to another ETS-family transcription factor, E R G , Nature Genet. 6: 146-151, 1994. 15. Peter, M . , Couturier, J., Pacquement, H . , M i c h o n , J., Thomas, G . , Magdelenat, H . , and Delattre, O. A new member of the E T S family fused to E W S in E w i n g tumors, Oncogene. 14: 1159-64, 1997.  161  16. Jeon, I.-S., Davis, J. N . , Braun, B . S., Sublett, J. E . , Roussel, M . F . , Denny, C . T., and Shapiro, D . N . A variant Ewing's sarcoma translocation (7;22) fuses the EWS gene to the E T S gene ETV1, Oncogene. 10: 1229-1234, 1995. 17. Urano, F . , Umezawa, A . , Hong, W . , K i k u c h i , H . , and Hata, J . A novel chimera gene between EWS and E1A-F, encoding the adenovirus E 1 A enhancer-binding protein, in extraosseous Ewing's sarcoma, Biochem Biophys Res Commun. 219: 608-612, 1996. 18. Wasylyk, B . , Hahn, S. L . , and Giovane, A . The E T S family of transcription factors., Eur J Biochem. 211: 7-18, 1993. 19. M a y , W . A . , Gishizky, M . L . , Lessnick, S. L . , Lunsford, L . B . , Lewis, B . C , Delattre, O., Zucman, J., Thomas, G . , and Denny, C . T. E w i n g sarcoma 11;22 translocation produces a chimeric transcription factor that requires the D N A - b i n d i n g domain encoded by F L U for transformation, Proc Natl Acad Sci U S A . 90: 5752-5756, 1993. 20. Lessnick, S. L . , Braun, B . S., Denny, C . T., and M a y , W . A . M u l t i p l e domains mediate transformation by the Ewing's sarcoma EWS/FLI-1 fusion gene, Oncogene. 10: 423431, 1995. 21. M a o , X . , Miesfeldt, S., Yang, H . , Leiden, J. M . , and Thompson, C . B . The FLI-1 and chimeric E W S - F L I - 1 oncoproteins display similar D N A binding specificities, J B i o l Chem. 269: 18216-18222, 1994. 22. M a y , W . A . , Lessnick, S. L . , Braun, B . S., Klemsz, M . , Lewis, B . C , Lunsford, L . B . , Hromas, R., and Denny, C . T. 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, M o l Cell B i o l . 13: 7393-7398, 1993. 23. O h n o , T., Rao, V . N . , and Reddy, S. P. E W S / F l i - 1 chimeric protein is a transcriptional activator., Cancer Res. 53: 5859-5863, 1993. 24. B a i l l y , R . - A . , Bosselut, R., Zucman, J., Cornier, F . , Delattre, O., Roussel, M . , Thomas, G . , and Ghysdael D N A - b i n d i n g and transcriptional activation properties of the E W S - F L I - 1 fusion protein resulting from the t ( l l ; 2 2 ) translocation i n E w i n g sarcoma, M o l Cell B i o l . 14: 3230-3241, 1994. 25. Zhang, D . , Paley, A . J., and Childs, G . The transcriptional repressor Z F M 1 interacts with and modulates the ability of E W S to activate transcription, J B i o l Chem. 273: 1808691., 1998. 26. H a h m , K . B . , C h o , K . , Lee, C , l m , Y . H . , Chang, J., C h o i , S. G . , Sorensen, P. H . , Thiele, C . J., and K i m , S. J. Repression of the gene encoding the TGF-beta type II receptor is a major target of the E W S - F L I 1 oncoprotein, Nat Genet. 23: 222-7., 1999. 27. Tanaka, K . , Iwakuma, T., Harimaya, K . , Sato, H . , and Iwamoto, Y . E W S - F l i l antisense oligodeoxynucleotide inhibits proliferation of human E w i n g ' s sarcoma and primitive neuroectodermal tumor cells, J C l i n Invest. 99: 239-47., 1997. 28. Kovar, H . , Aryee, D . N . , Jug, G . , Henockl, C , Schemper, M . , Delattre, O., Thomas, G . , and Gadner, H . E W S / F L I - 1 antagonists induce growth inhibition of E w i n g tumor cells in vitro, C e l l Growth Differ. 7: 429-37., 1996. 29. Delattre, O., Zucman, J., Melot, T., Garau, X . S., Zucker, J . - M . , Lenoir, G . M . , Ambros, P. F . , Sheer, D . , Turc-Carel, C , Triche, T. J., Aurias, A . , and Thomas, G . The E w i n g family of tumors—A subgroup of small-round-cell tumors defined by specific chimeric transcripts, N Engl J M e d . 331: 294-299, 1994. 30. Sorensen, P. H . B . , W u , J. K . , Berean, K . W . , L i m , J. F . , Donn, W . , Frierson, H . F., Reynolds, C . P., Lopez-Terrada, D . , and Triche, T. J. Olfactory neuroblastoma is a peripheral  162  primitive neuroectodermal tumor related to Ewing's sarcoma., Proc Natl A c a d Sci U S A . 93: 1038-1043, 1996. 31. Sorensen, P. H . B . , Shimada, H . , L i u , X . F . , L i m , J. S., Thomas, G . , and Triche, T. J. Biphenotypic sarcomas with myogenic and neural differentiation express the Ewing's sarcoma EWS/FLI1 fusion gene., Cancer Res. 55: 1385-1392, 1995. 32. Zucman, J., Melot, T., Desmaze, C , Ghysdael, J., Plougastel, B . , Peter, M . , Zucker, J. M . , Triche, T. J., Sheer, D . , Turc-Carel, C , Ambros, P., Combaret, V . , Lenoir, G . , Aurias, A . , Thomas, G . , and Delattre, O. Combinatorial generation of variable fusion proteins in the E w i n g family of tumours, Embo J. 12: 4481-4487, 1993. 33. de A l a v a , E . , K a w a i , A . , Healey, J. FL, Fligman, I., Meyers, P. A . , Huvos, A . G . , Gerald, W . L . , Jhanwar, S. C , Argani, P., Antonescu, C . R., Pardo-Mindan, F. J., Ginsberg, J., Womer, R., Lawlor, E . R., Wunder, J., Andrulis, I., Sorensen, P. H . , Barr, F . G . , and L a d a n y i , M . E W S - F L I 1 fusion transcript structure is an independent determinant of prognosis i n E w i n g ' s sarcoma [published erratum appears in J C l i n O n c o l 1998 Aug;16(8):2895] [see comments], J C l i n Oncol. 16: 1248-55, 1998. 34. Ginsberg, J. P., de Alava, E . , Ladanyi, M . , Wexler, L . H . , Kovar, H . , Paulussen, M . , Zoubek, A . , Dockhorn-Dworniczak, B . , Juergens, H . , Wunder, J. S., Andrulis, I. L . , M a l i k , R., Sorensen, P. H . , Womer, R. B . , and Barr, F . G . E W S - F L I 1 and E W S - E R G gene fusions are associated with similar clinical phenotypes in Ewing's sarcoma, J C l i n Oncol. 17: 180914, 1999. 35. W e i , G . , Antonescu, C . R., de A l a v a , E . , Leung, D . , Huvos, A . G . , Meyers, P. A . , Healey, J. H . , and Ladanyi, M . Prognostic impact of I N K 4 A deletion in E w i n g sarcoma, Cancer. 89: 793-9., 2000. 36. de A l a v a , E . , Antonescu, C . R., Panizo, A . , Leung, D . , Meyers, P. A . , Huvos, A . G . , Pardo-Mindan, F . J., Healey, J. H . , and Ladanyi, M . Prognostic impact of P53 status in E w i n g sarcoma, Cancer. 89: 783-92., 2000. 37. Toretsky, J. A . , Kalebic, T., Blakesley, V . , L e R o i t h , D . , and Helman, L . J. The insulin-like growth factor-I receptor is required for E W S / F L I - 1 transformation of fibroblasts, J B i o l Chem. 272: 30822-7, 1997. 38. Scotlandi, K . , Benini, S., Nanni, P., L o l l i n i , P. L . , Nicoletti, G . , Landuzzi, L . , Serra, M . , Manara, M . C , P i c c i , P., and B a l d i n i , N . Blockage of insulin-like growth factor-I receptor inhibits the growth of Ewing's sarcoma in athymic mice, Cancer Res. 58: 4127-31., 1998. 39. Toretsky, J. A . , Thakar, M . , Eskenazi, A . E . , and Frantz, C . N . Phosphoinositide 3hydroxide kinase blockade enhances apoptosis in the Ewing's sarcoma family of tumors, Cancer Res. 59: 5745-50, 1999. 40. Girnita, L . , Girnita, A . , Wang, M . , M e i s - K i n d b l o m , J. M . , K i n d b l o m , L . G . , and Larsson, O. A link between basic fibroblast growth factor (bFGF) and E W S / F L I - 1 in Ewing's sarcoma cells, Oncogene. 19: 4298-301, 2000. 41. Landuzzi, L . , De Giovanni, C , Nicoletti, G . , Rossi, I., R i c c i , C , Astolfi, A . , Scopece, L . , Scotlandi, K . , Serra, M . , Bagnara, G . P., Nanni, P., and L o l l i n i , P. L . The metastatic ability of Ewing's sarcoma cells is modulated by stem cell factor and by its receptor c-kit, A m J Pathol. 157: 2123-31., 2000. 42. Ricotti, E . , Fagioli, F . , Garelli, E . , L i n a r i , C , Crescenzio, N . , Horenstein, A . L . , Pistamiglio, P., V a i , S., Berger, M . , di Montezemolo, L . C , Madon, E . , and Basso, G . c-kit is  163  expressed in soft tissue sarcoma of neuroectodermic origin and its ligand prevents apoptosis of neoplastic cells, Blood. 91: 2397-405., 1998. 43. Dohjima, T., Ohno, T., Banno, Y . , N o z a w a , Y . , W e n - y i , Y . , and Shimizu, K . Preferential down-regulation of phospholipase C-beta in Ewing's sarcoma cells transfected with antisense E W S - F l i - 1 , B r J Cancer. 82: 16-9., 2000. 44. Blume-Jensen, P. and Hunter, T. Oncogenic kinase signalling, Nature. 411: 355-65., 2001. 45. Zanke, B . W . Growth factors and intracellular signaling. In: I. F. Tannock and R. P. H i l l (eds.), The Basic Science of Oncology, Third edition, pp. 106-33. Toronto: M c G r a w H i l l , 1998. 46. Gutkind, J. S. The pathways connecting G protein-coupled receptors to the nucleus through divergent mitogen-activated protein kinase cascades, J B i o l Chem. 273: 1839-42., 1998. 47. Thiele, C . J. and Helman, L . J. Differentiating agents, growth factors and novel molecular approaches to the treatment of pediatric cancer. In: P. A . Pizzo and D . G . Poplack (eds.), Principles and Practice of Pediatric Oncology, 2nd edition, pp. 409-28. Philadelphia: JB Lipincott, 1993. 48. G u t k i n d , J. S. C e l l growth control by G protein-coupled receptors: from signal transduction to signal integration, Oncogene. 17: 1331-42, 1998. 49. Slingerland, J. M . and Tannock, I. F . C e l l Proliferation and C e l l Death. In: I. F. Tannock and R. P. H i l l (eds.), The Basic Science of Oncology, T h i r d edition, pp. 134-65. Toronto: M c G r a w - H i l l , 1998. 50. Rozengurt, E . Signal transduction pathways in the mitogenic response to G proteincoupled neuropeptide receptor agonists, J Cell Physiol. 177: 507-17., 1998. 51. Shields, J. M . , Pruitt, K . , M c F a l l , A . , Shaub, A . , and Der, C . J. Understanding Ras: 'it ain't over 'til it's over', Trends Cell B i o l . 10: 147-54., 2000. 52. Garrington, T. P. and Johnson, G . L . Organization and regulation of mitogenactivated protein kinase signaling pathways, Curr Opin Cell B i o l . 11: 211-8., 1999. 53. Datta, S. R., Brunet, A . , and Greenberg, M . E . Cellular survival: a play in three Akts, Genes Dev. 13: 2905-27., 1999. 54. Rodriguez-Viciana, P., Warne, P. H . , Dhand, R., Vanhaesebroeck, B . , Gout, I., Fry, M . J., Waterfield, M . D . , and Downward, J. Phosphatidylinositol-3-OH kinase as a direct target of Ras, Nature. 370: 527-32., 1994. 55. Filippa, N . , Sable, C . L . , F i l l o u x , C , Hemmings, B . , and V a n Obberghen, E . Mechanism of protein kinase B activation by cyclic AMP-dependent protein kinase, M o l Cell B i o l . 19: 4989-5000., 1999. 56. Y a n o , S., Tokumitsu, H . , and Soderling, T. R . C a l c i u m promotes cell survival through C a M - K kinase activation of the protein-kinase-B pathway, Nature. 396: 584-7., 1998. 57. Scheid, M . P. and Duronio, V . Dissociation of cytokine-induced phosphorylation of Bad and activation of P K B / a k t : involvement of M E K upstream of B a d phosphorylation, Proc Natl A c a d Sci U S A . 95: 7439-44., 1998. 58. Scheid, M . P. and Woodgett, J. R. Protein kinases: six degrees of separation?, Curr B i o l . 10: R191-4, 2000. 59. Zimmermann, S. and M o e l l i n g , K . Phosphorylation and regulation of Raf by A k t (protein kinase B ) , Science. 286: 1741-4., 1999.  164  60. R o m m e l , C , Clarke, B . A . , Zimmermann, S., Nunez, L . , Rossman, R., Reid, K . , M o e l l i n g , K . , Yancopoulos, G . D . , and Glass, D . J. Differentiation stage-specific inhibition of the R a f - M E K - E R K pathway by A k t , Science. 286: 1738-41., 1999. 61. Treinies, I., Paterson, H . F . , Hooper, S., W i l s o n , R., and Marshall, C . J. Activated M E K stimulates expression of A P - 1 components independently of phosphatidylinositol 3kinase (PI3-kinase) but requires a PI3-kinase signal T o stimulate D N A synthesis, M o l C e l l B i o l . 19: 321-9., 1999. 62. Roovers, K . and Assoian, R. K . Integrating the M A P kinase signal into the G l phase cell cycle machinery, Bioessays. 22: 818-26., 2000. 63. G i l l e , H . and Downward, J. Multiple ras effector pathways contribute to G ( l ) cell cycle progression, J B i o l Chem. 274: 22033-40., 1999. 64. Reusch, H . P., Zimmermann, S., Schaefer, M . , Paul, M . , and M o e l l i n g , K . Regulation of Raf by A k t controls growth and differentiation in vascular smooth muscle cells, J B i o l Chem. 6: 6, 2001. 65. H o w a r d , A . and Pelc, S. R. Nuclear incorporation of p32 as demonstrated by autoradiographies, E x p C e l l Res. 2: 178-87, 1951. 66. Sherr, C . J. The Pezcoller lecture: cancer cell cycles revisited, Cancer Res. 60: 368995, 2000. 67. Nevins, J. R. E 2 F : A L i n k Between the R b Tumor Suppressor Protein and V i r a l Oncoproteins, Science. 258: 424-429, 1992. 68. Hashimoto, Y . , K o h r i , K . , K a n e k o , Y . , M o r i s a k i , H . , K a t o , T., Ikeda, K . , and Nakanishi, M . Critical role for the 310 helix region of p57(Kip2) in cyclin-dependent kinase 2 inhibition and growth suppression, J B i o l Chem. 273: 16544-50., 1998. 69. Sherr, C . J. and Roberts, J. M . C D K inhibitors: positive and negative regulators of Gl-phase progression, Genes Dev. 13: 1501-12, 1999. 70. Montagnoli, A . , Fiore, F . , Eytan, E . , Carrano, A . C , Draetta, G . F . , Hershko, A . , and Pagano, M . Ubiquitination of p27 is regulated by Cdk-dependent phosphorylation and trimeric complex formation, Genes Dev. 13: 1181-9., 1999. 71. Evan, G . I. and Vousden, K . H . Proliferation, cell cycle and apoptosis in cancer, Nature. 411: 342-8., 2001. 72. Muise-Helmericks, R. C , Grimes, H . L . , Bellacosa, A . , Malstrom, S. E . , Tsichlis, P. N . , and Rosen, N . C y c l i n D expression is controlled post-transcriptionally v i a a phosphatidylinositol 3-kinase/Akt-dependent pathway, J B i o l Chem. 273: 29864-72, 1998. 73. Diehl, J. A . , Cheng, M . , Roussel, M . F . , and Sherr, C . J. Glycogen synthase kinase3beta regulates cyclin D I proteolysis and subcellular localization, Genes Dev. 12: 3499-511, 1998. 74. Jentsch, S. The ubiquitin-conjugation system, A n n u Rev Genet. 26: 179-207, 1992. 75. D i e h l , J. A . , Zindy, F . , and Sherr, C . J. Inhibition of cyclin D I phosphorylation on threonine-286 prevents its rapid degradation via the ubiquitin-proteasome pathway, Genes Dev. 11: 957-72., 1997. 76. A l t , J. R., Cleveland, J. L . , Hannink, M . , and Diehl, J. A . Phosphorylation-dependent regulation of cyclin D I nuclear export and cyclin Dl-dependent cellular transformation, Genes Dev. 14: 3102-14., 2000. 77. Taipale, J. and Beachy, P. A . The Hedgehog and Wnt signalling pathways in cancer, Nature. 411: 349-54., 2001.  165  78. Israel, M . A . Cancer cell biology. In: P. A . Pizzo and D . G . Poplack (eds.), Principles and Practice of Pediatric Oncology, 2nd edition, pp. 57-80. Philadelphia: JB Lipincott, 1993. 79. B e c k w i t h , J . B . , K i v i a t , N . B . , and B o n a d i o , J . F . N e p h r o g e n i c rests, nephroblastomatosis, and the pathogenesis of Wilms' tumor. In: R. Jaffe, B . B . Dahms, H . F. Krous, E . Lieberman, and T. J. Triche (eds.), Forefront of Pediatric Pathology: A Festschrift for Benjamin H . Landing, pp. 1-36. Philadelphia: Hemisphere Publishing Corporation, 1990. 80. Cooper, M . J., Hutchins, G . M . , Cohen, P. S., Helman, L . J., Mennie, R. J., and Israel, M . A . Human neuroblastoma tumor cell lines correspond to the arrested differentiation of chromaffin adrenal medullary neuroblasts, Cell Growth & Differentiation. 1: 149-159, 1990. 81. Raff, M . C e l l suicide for beginners, Nature. 396: 119-22., 1998. 82. Assoian, R . K . Anchorage-dependent cell cycle progression, J C e l l B i o l . 136: 1-4., 1997. 83. Dedhar, S., Hannigan, G . E . , Rak, J., and K e r b e l , R . S. The Extracellular Environment and Cancer. In: I. F . Tannock and R. P. H i l l (eds.), The Basic Science of Oncology, Third edition, pp. 197-218. Toronto: M c G r a w - H i l l , 1998. 84. Boudreau, N . , Sympson, C . J., Werb, Z . , and Bissell, M . J. Suppression of I C E and apoptosis in mammary epithelial cells by extracellular matrix, Science. 267: 891-3., 1995. 85. Fang, F . , Orend, G . , Watanabe, N . , Hunter, T., and Ruoslahti, E . Dependence of cyclin E - C D K 2 kinase activity on cell anchorage, Science. 277 : 499-502., 1996. 86. Clark, E . A . and Brugge, J. S. Integrins and signal transduction pathways: the road taken, Science. 268: 233-9., 1995. 87. Delcommenne, M . , Tan, C , Gray, V . , Rue, L . , Woodgett, J., and Dedhar, S. Phosphoinositide-3-OH kinase-dependent regulation of glycogen synthase kinase 3 and protein kinase B / A K T by the integrin-linked kinase, Proc Natl A c a d S c i U S A . 95: 112116., 1998. 88. K o p n i n , B . P. Targets of oncogenes and tumor suppressors: key for understanding basic mechanisms o f carcinogenesis, Biochemistry (Mosc). 65: 2-27., 2000. 89. Park, M . Oncogenes. In: B . Vogelstein and K . W . Kinzler (eds.), The Genetic Basis of Human Cancer, pp. 205-223. New York: M c G r a w - H i l l C o . , 1998. 90. Benchimol, S. and M i n d e n , M . D . Viruses, oncogenes, and tumor suppressor genes. In: I. F. Tannock and R. P. H i l l (eds.), The Basic Science of Oncology, Third edition, pp. 79105. Toronto: M c G r a w - H i l l , 1998. 91. Knezevich, S. R., McFadden, D . E . , Tao, W . , L i m , J. F., and Sorensen, P. H . A novel E T V 6 - N T R K 3 gene fusion in congenital fibrosarcoma, Nat Genet. 18: 184-7, 1998. 92. Hayward, W . S., Neel, B . G . , and Astrin, S. M . Activation of a cellular one gene by promoter insertion in A L V - i n d u c e d lymphoid leukosis, Nature. 290: 475-80, 1981. 93. Lengauer, C , K i n z l e r , K . W . , and Vogelstein, B . Genetic instabilities in human cancers, Nature. 396: 643-9., 1998. 94. Rowley, J. D . Finding order in chaos, Perspect B i o l M e d . 32: 371-84., 1989. 95. Shtivelman, E . , Lifshitz, B . , Gale, R. P., and Canaani, E . Fused transcript of abl and bcr genes in chronic myelogenous leukaemia, Nature. 315: 550-554, 1985. 96. Knudson, A . G . , Jr. Mutation and cancer: statistical study of retinoblastoma, Proc Natl A c a d S c i U S A . 68: 820-3., 1971. 97. Pucci, B . , Kasten, M . , and Giordano, A . Cell cycle and apoptosis, Neoplasia. 2: 2919., 2000.  166  98. Agarwal, M . L . , Taylor, W . R., Chernov, M . V . , Chernova, O . B . , and Stark, G . R. The p53 network, J B i o l Chem. 273: 1-4., 1998. 99. Squire, J. A . , Whitmore, G . F., and Philliips, R. A . Genetic Basis of Cancer. In: I. F. Tannock and R . P . H i l l (eds.), The Basic Science o f Oncology, T h i r d edition, pp. 48-78. Toronto: M c G r a w - H i l l , 1998. 100. Gallie, B . L . Retinoblastoma gene mutations in human cancer, M E n g l J M e d . 330: 786-787, 1994. 101. Quelle, D . E . , Zindy, F . , Ashmun, R. A . , and Sherr, C . J. Alternative reading frames of the I N K 4 a tumor suppressor gene encode two unrelated proteins capable of inducing cell cycle arrest, C e l l . 83: 993-1000., 1995. 102. Scotlandi, K . , Benini, S., Sarti, M . , Serra, M . , L o l l i n i , P . - L . , M a u r i c i , D . , P i c c i , P., Manara, M . C , and B a l d i n i , N . Insulin-like growth factor I receptor-mediated circuit in Ewing's sarcoma/peripheral neuroectodermal tumor: A possible therapeutic target, Cancer Res. 56: 4570-4574, 1996. 103. El-Badry, O. M . , M i n n i t i , C , K o h n , E . C , Houghton, P. J., Daughaday, W . H . , and Helman, L . J. Insulin-like growth Factor II Acts as an Autocrine Growth and Motility Factor in Human Rhabdomyosarcoma Tumors, Cell Growth & Differentiation. 1: 3215-331, 1990. 104. Yee, D . , Favoni, R. E . , Lebovic, G . S., Lombana, F . , Powell, D . , Reynolds, C . P., and Rosen, N . I G F - I expression by tumors of neuroectodermal origin with the t ( l l ; 2 2 ) chromosomal translocation. A potential autocrine growth factor, J C l i n Invest. 86: 18061814, 1990. 105. Dunn, T., Praissman, L . , Hagag, N . , and V i o l a , M . V . E R G gene is translocated in an Ewing's sarcoma cell line, Cancer Genet Cytogenet. 76: 19-22., 1994. 106. Jalava, A . M . , H e i k k i l a , J. E . , and A k e r m a n , K . E . D e c l i n e in c-myc m R N A expression but not the induction of c-fos m R N A expression is associated with differentiation of S H - S Y 5 Y human neuroblastoma cells, E x p Cell Res. 179: 10-17, 1988. 107. Whang-Peng, J., Knutsen, T., T h i e l , K . , H o r o w i t z , M . E . , and T r i c h e , T . J. Cytogenetic studies in subgroups of rhabdomyosarcoma, Genes, C h r o m . & Cancer. 5: 299310, 1992. 108. Hay, R., Caputo, J., Chen, T. R., M a c y , M . , McClintock, P., and Reid, Y . A T C C C e l l Lines and Hybridomas. , V o l . 8th, pp. 78-337. R o c k v i l l e , M D : American Type Culture Collection, 1994. 109. Houghton, J. A . , Houghton, P. J., and Webber, B . L . Growth and characterization of childhood rhabdomyosarcomas as xenografts, J N C I . 68: 437-443, 1982. 110. C h o m c z y n s k i , P. and Sacchi, N . Single-step method of R N A isolation by acid guanidinium thiocyanate-phenol-chloroform extraction., A n a l . B i o c h e m . 162: 156-159, 1987. 111. Sorensen, P. H . B . , L i u , X . F., Thomas, G . , DeLattre, O., Rowland, J. M . , Biggs, C . A . , and Triche, T . J. Reverse Transcriptase P C R A m p l i f i c a t i o n of E W S / F l i - 1 Fusion Transcripts as a Diagnostic Test for Peripheral P r i m i t i v e Neuroectodermal Tumors of Childhood, Diag. M o l . Pathol. 2: 147-157, 1993. 112. Siegfried, J. M . , H a n , Y . H . , D e M i c h e l e , M . A . , Hunt, J . D . , Gaither, A . L . , and Cuttitta, F . Production of gastrin-releasing peptide by a non-small cell lung carcinoma cell line adapted to serum-free and growth factor-free conditions, J B i o l Chem. 269: 8596-603., 1994.  167  113. Argatoff, L . H . , O'Connell, J. X . , Mathers, J. A . , G i l k s , C . B . , and Sorensen, P. H . B . Demonstration of the E W S / W T 1 gene fusion by reverse transcriptase polymerase chain reaction in the diagnosis of intraabdominal desmoplastic small round cell tumor., A m e r J Surg Path. 20: 406-412, 1996. 114. Sausville, E . A . , Lebacq-Verheyden, A . M . , Spindel, E . R., Cuttitta, F . , Gazdar, A . F., and Battey, J. F . Expression of the gastrin-releasing peptide gene in human small cell lung cancer. Evidence for alternative processing resulting in three distinct m R N A s , J B i o l Chem. 261: 2451-7., 1986. 115. Sambrook, J., Fritch, E . F., and Maniatis, T. Molecular cloning: a laboratory manual., second edition. C o l d Spring Harbor, N . Y . : C o l d Spring Harbor Laboratory Press, 1989. 116. Fraser, M . , M c D o n a l d , T. J., Spindel, E . R., Fahy, M . , H i l l , D . , and Challis, J. R. Gastrin-releasing peptide is produced in the pregnant ovine uterus, Endocrinology. 135: 2440-5., 1994. 117. Asubel, F . M . Analysis of Proteins. In: V . B . Chanda (ed.) Current Protocols in Molecular Biology, V o l . 2. U S A : John Wiley & Sons, Inc., 2001. 118. Nagalla, S. R. and Spindel, E . R. Functional analysis of the 5'-flanking region of the human gastrin- releasing peptide gene in small cell lung carcinoma cell lines, Cancer Res. 54: 4461-7., 1994. 119. Fulton, R. and V a n Ness, B . Luminescent reporter gene assays for luciferase and beta-galactosidase using a liquid scintillation counter, Biotechniques. 14: 762-3., 1993. 120. Q i n , Y . , Ertl, T., C a i , R. Z . , Halmos, G . , and Schally, A . V . Inhibitory effect of bombesin receptor antagonist RC-3095 on the growth of human pancreatic cancer cells in vivo and in vitro, Cancer Res. 54: 1035-41., 1994. 121. Carmichael, J., DeGraff, W . G . , Gazdar, A . F . , M i n n a , J. D . , and M i t c h e l l , J. B . Evaluation of a tetrazolium-based semiautomated colorimetric assay: assessment of chemosensitivity testing, Cancer Res. 47: 936-42., 1987. 122. L i a n g , P. and Pardee, A . B . Differential display of eukaryotic messenger R N A by means of the polymerase chain reaction., Science. 257: 967-971, 1992. 123. L i a n g , P., Averboukh, L . , Keyomarsi, K . , Sager, R., and Pardee, A . B . Differential display and cloning of messenger R N A s from human breast cancer versus mammary epithelial cells., Cancer Res. 52: 6966-6968, 1992. 124. Watson, M . A . and Fleming, T. P. Isolation of differentially expressed sequence tags from human breast cancer, Cancer Res. 54: 4598-602., 1994. 125. Spindel, E . R., Chin, W . W . , Price, J., Rees, L . H . , Besser, G . M . , and Habener, J. F. Cloning and characterization of c D N A s encoding human gastrin-releasing peptide, Proc Natl A c a d Sci U S A . 81: 5699-703., 1984. 126. Kovar, H . , Jug, G . , Aryee, D . N . , Zoubek, A . , Ambros, P., Gruber, B . , Windhager, R., and Gadner, H . A m o n g genes involved in the R B dependent cell cycle regulatory cascade, the p l 6 tumor suppressor gene is frequently lost in the E w i n g family of tumors, Oncogene. 15: 2225-32., 1997. 127. Spindel, E . R., Giladi, E . , Segerson, T. P., and Nagalla, S. Bombesin-like peptides: of ligands and receptors, Recent Prog Horm Res. 48: 365-91, 1993. 128. Moody, T. W . , Pert, C . B . , Gazdar, A . F., Carney, D . N . , and M i n n a , J. D . H i g h levels of intracellular bombesin characterize human small-cell lung carcinoma, Science. 214: 12468., 1981.  168  129. Cuttitta, F . , Carney, D . N . , Mulshine, J., Moody, T. W . , Fedorko, J., Fischler, A . , and M i n n a , J. D . Bombesin-like peptides can function as autocrine growth factors in human small-cell lung cancer., Nature. 316: 823-826, 1985. 130. Hajri, A . , Balboni, G . , Koenig, M . , Garoud, J. C , and Damge, C . Gastrin-releasing peptide: in vivo and in vitro growth effects on an acinar pancreatic carcinoma., Cancer Research. 52: 3726-3732, 1992. 131. Narayan, S., Guo, Y . S., Townsend, C . M . , Jr., and Singh, P. Specific binding and growth effects of bombesin-related peptides on mouse colon cancer cells in vitro, Cancer Res. 50: 6772-8., 1990. 132. Yano, T., Pinski, J., Groot, K . , and Schally, A . V . Stimulation by bombesin and inhibition by bombesin/gastrin-releasing peptide antagonist R C - 3 0 9 5 of growth of human breast cancer cell lines, Cancer Res. 52: 4545-7., 1992. 133. Pinski, J., Schally, A . V . , Halmos, G . , Szepeshazi, K . , Groot, K . , O'Byrne, K . , and Cai, R. Z . Effects of somatostatin analogue R C - 1 6 0 and bombesin/gastrin-releasing peptide antagonists on the growth of human small-cell and non-small- cell lung carcinomas in nude mice, B r J Cancer. 70: 886-92., 1994. 134. Radulovic, S., Schally, A . V . , Reile, H . , Halmos, G . , Szepeshazi, K . , Groot, K . , M i l o v a n o v i c , S., M i l l e r , G . , and Y a n o , T . Inhibitory effects of antagonists of bombesin/gastrin releasing peptide ( G R P ) and somatostatin analog (RC-160) on growth of HT-29 human colon cancers in nude mice, Acta Oncol. 33: 693-701, 1994. 135. Shirahige, J., C a i , R . - Z . , Szepeshazi, K . , Halmos, G . , and Schally, A . V . Inhibitory effects of bombesin/gastrin releasing peptide ( G R P ) antagonists RC-3950-II and RC-3095 on M C F 7 M i l human breast cancer xenografts in nude mice., Biomed. Pharmacotherapy. 48: 465-472, 1994. 136. Nagy, A . , Armatis, P., C a i , R. Z . , Szepeshazi, K . , Halmos, G . , and Schally, A . V . Design, synthesis, and in vitro evaluation of cytotoxic analogs of bombesin-like peptides containing doxorubicin or its intensely potent derivative, 2-pyrrolinodoxorubicin, Proc Natl Acad Sci U S A . 94: 652-6., 1997. 137. Halmos, G . and Schally, A . V . Reduction in receptors for bombesin and epidermal growth factor in xenografts of human small-cell lung cancer after treatment with bombesin antagonist RC-3095, Proc Natl A c a d Sci U S A . 94: 956-60., 1997. 138. Verbeeck, M . A . and Burbach, J. P. The cholecystokinin gene is abundantly coexpressed with gastrin- releasing peptide, enkephalin and neuropeptide Y genes in a clonal human neuroepithelioma cell line, F E B S Lett. 268: 88-90., 1990. 139. S a w i n , R . S. and Brockenbrough, J. Secretion of gastrin-releasing peptide by retroperitoneal tumors, A m J Surg. 169: 550-2., 1995. 140. Sawin, R . S., Brockenbrough, J., and Ness, J. C . Gastrin releasing peptide is an autocrine growth factor for human neuroblastoma, Surg Forum. 43: 606-8, 1992. 141. Sunday, M . E . , C h o i , N . , Spindel, E . R., C h i n , W . W . , and M a r k , E . J. Gastrinreleasing peptide gene expression in small cell and large cell undifferentiated lung carcinomas, H u m Pathol. 22: 1030-9., 1991. 142. Maruno, K . , Yamaguchi, K . , A b e , K . , Suzuki, M . , Saijo, N . , M i s h i m a , Y . , Yanaihara, N . , and Shimosato, Y . Immunoreactive gastrin-releasing peptide as a specific tumour marker in patients with small cell lung cancer., Cancer Research. 49: 629-632, 1989.  169  143. L e v , S., Moreno, H . , Martinez, R., Canoll, P., Peles, E . , Musacchio, J. M . , Plowman, G . D . , Rudy, B . , and Schlessinger, J. Protein tyrosine kinase P Y K 2 involved in Ca(2+)induced regulation of ion channel and M A P kinase functions, Nature. 376: 737-45., 1995. 144. X i o n g , W . C , M a c k l e m , M . , and Parsons, J. T. Expression and characterization of splice variants of P Y K 2 , a focal adhesion kinase-related protein, J C e l l S c i . Ill: 1981-91., 1998. 145. Richardson, A . and Parsons, T. A mechanism for regulation of the adhesionassociated proteintyrosine kinase p p l 2 5 F A K , Nature. 380: 538-40., 1996. 146. Nolan, K . , Lacoste, J., and Parsons, J. T. Regulated expression of focal adhesion kinase-related nonkinase, the autonomously expressed C-terminal domain of focal adhesion kinase, M o l C e l l B i o l . 19: 6120-9., 1999. 147. Felsch, J. S., Lane, W . S., and Peralta, E . G . Tyrosine kinase P y k 2 mediates G protein-coupled receptor regulation of the E w i n g sarcoma R N A - b i n d i n g protein E W S , Curr B i o l . 9: 485-8., 1999. 148. Jensen, J. A . , Carroll, R. E . , and Benya, R. V . The case for gastrin-releasing peptide acting as a morphogen when it and its receptor are aberrantly expressed in cancer, Peptides. 22: 689-99., 2001. 149. Montesano, R., Matsumoto, K . , Nakamura, T., and O r c i , L . Identification of a fibroblast-derived epithelial morphogen as hepatocyte growth factor, C e l l . 67: 901-8., 1991. 150. O'Dorisio, M . S., Fleshman, D . J., Qualman, S. J., and O'Dorisio, T. M . Vasoactive intestinal peptide: autocrine growth factor in neuroblastoma, Regul Pept. 37: 213-26., 1992. 151. Pence, J. C . and Shorter, N . A . In vitro differentiation of human neuroblastoma cells caused by vasoactive intestinal peptide, Cancer Res. 50: 5177-83., 1990. 152. Ponten, J. Spontaneous and Virus Induced Transformation in C e l l Culture. N e w York: Springer Verlag, 1971. 153. Baserga, R. The price of independence, E x p C e l l Res. 236: 1-3, 1997. 154. Santini, M . T . and Rainaldi, G . Three-dimensional spheroid model in tumor biology, Pathobiology. 67: 148-57, 1999. 155. de A l a v a , E . , Panizo, A . , Antonescu, C . R., Huvos, A . G , Pardo-Mindan, F . J., Barr, F. G . , and Ladanyi, M . Association of E W S - F L I 1 type 1 fusion with lower proliferative rate in Ewing's sarcoma, A m J Pathol. 156: 849-55, 2000. 156. Scotlandi, K . , Serra, M . , M a n a r a , M . C , M a u r i c i , D . , B e n i n i , S., N i n i , G . , Campanacci, M . , and Baldini, N . C l i n i c a l relevance of K i - 6 7 expression in bone tumors, Cancer. 75: 806-14, 1995. 157. N i g g l i , F . K . , Powell, J. E . , Parkes, S. E . , Ward, K . , Raafat, F . , M a n n , J. R., and Stevens, M . C . D N A ploidy and proliferative activity (S-phase) in childhood soft- tissue sarcomas: their value as prognostic indicators, B r J Cancer. 69: 1106-10, 1994. 158. M e l l i n , W . , Dierschauer, W . , Ffiddemann, W . , Roessner, A . , Edel, G . , Wuisman, P., Harle, A . , and Grundmann, E . F l o w cytometric D N A analysis of bone tumors, Curr Top Pathol. 80: 115-52, 1989. 159. V o l l m e r , E . , Roessner, A . , W u i s m a n , P., Harle, A . , and Grundmann, E . The proliferation behavior of bone tumors investigated with the monoclonal antibody K i - 6 7 , Curr Top Pathol. 50: 91-114, 1989. 160. Kawamura, K . , Kobayashi, Y . , Tanaka, T., Ikeda, R., Fujikawa-Yamamoto, K . , and Suzuki, K . Intranuclear localization of proliferating cell nuclear antigen during the cell cycle in renal cell carcinoma, Anal Quant Cytol Histol. 22: 107-13., 2000.  170  161. Shao, J., Sheng, H . , D u B o i s , R. N . , and Beauchamp, R. D . Oncogenic Ras-mediated cell growth arrest and apoptosis are associated with increased ubiquitin-dependent cyclin D I degradation, J B i o l Chem. 275: 22916-24., 2000. 162. Cross, D . A . , A l e s s i , D . R . , Cohen, P., Andjelkovich, M . , and Hemmings, B . A . Inhibition of glycogen synthase kinase-3 by insulin mediated by protein kinase B , Nature. 378: 785-9., 1995. 163. Bates, R. C , Edwards, N . S., and Yates, J. D . Spheroids and cell survival, Crit Rev Oncol Hematol. 36: 61-74., 2000. 164. Santini, M . T., Rainaldi, G . , and Indovina, P. L. Apoptosis, cell adhesion and the extracellular matrix in the three- dimensional growth of multicellular tumor spheroids, Crit Rev Oncol Hematol. 36: 75-87., 2000. 165. van Weeren, P. C , de B r u y n , K . M . , de Vries-Smits, A . M . , van Lint, J., and Burgering, B . M . Essential role for protein kinase B ( P K B ) in insulin-induced glycogen synthase kinase 3 inactivation. Characterization of dominant-negative mutant of P K B , J B i o l Chem. 273: 13150-6., 1998. 166. Welcker, M . , Lukas, J., Strauss, M . , and Bartek, J. Enhanced protein stability: a novel mechanism of D-type cyclin over- abundance identified in human sarcoma cells, Oncogene. 13: 419-25., 1996. 167. Lawlor, E . R., L i m , J. F., Tao, W . , Poremba, C , C h o w , C . J., Kalousek, I. V . , Kovar, H . , M a c D o n a l d , T . J., and Sorensen, P. H . B . The E w i n g T u m o r F a m i l y of Peripheral Primitive Neuroectodermal Tumors Expresses Human Gastrin-releasing Peptide, Cancer Research. 58: 2469-2476, 1998. 168. Silvany, R . E . , Eliazer, S., Wolff, N . C , and Ilaria, R. L . , Jr. Interference with the constitutive activation of E R K 1 and E R K 2 impairs E W S / F L I - 1 -dependent transformation, Oncogene. 19: 4523-30, 2000. 169. Hoshino, R., Chatani, Y . , Yamori, T., Tsuruo, T., Oka, H . , Yoshida, O., Shimada, Y . , A r i - i , S., Wada, H . , Fujimoto, J., and Kohno, M . Constitutive activation of the 41-/43-kDa mitogen-activated protein kinase signaling pathway in human tumors, Oncogene. 18: 813-22, 1999. 170. K o h n , A . D . , Summers, S. A . , Birnbaum, M . J., and Roth, R . A . Expression of a constitutively active A k t Ser/Thr kinase in 3T3-L1 adipocytes stimulates glucose uptake and glucose transporter 4 translocation, J B i o l Chem. 271: 31372-8., 1996. 171. Hannigan, G . E . , Leung-Hagesteijn, C , Fitz-Gibbon, L . , Coppolino, M . G . , Radeva, G . , Filmus, J., B e l l , J. C , and Dedhar, S. Regulation of cell adhesion and anchoragedependent growth by a new beta 1-integrin-linked protein kinase, Nature. 379: 91-6., 1996. 172. Connell-Crowley, L . , Elledge, S. J., and Harper, J. W . G l cyclin-dependent kinases are sufficient to initiate D N A synthesis in quiescent human fibroblasts, Curr B i o l . 8: 65-8., 1998. 173. L i u , J. J., Chao, J. R., Jiang, M . C , N g , S. Y . , Y e n , J. J., and Y a n g - Y e n , H . F . Ras transformation results in an elevated level of cyclin D I and acceleration of G l progression in N I H 3T3 cells, M o l Cell B i o l . 15: 3654-63, 1995. 174. Rodriguez-Puebla, M . L . , Robles, A . I., and Conti, C . J. ras activity and cyclin D I expression: an essential mechanism of mouse skin tumor development, M o l Carcinog. 24: 16., 1999. 175. Robles, A . I., Rodriguez-Puebla, M . L . , G l i c k , A . B . , Trempus, C , Hansen, L . , Sicinski, P., Tennant, R. W . , Weinberg, R. A . , Yuspa, S. H . , and Conti, C . J. Reduced skin  171  tumor development in cyclin DI-deficient mice highlights the oncogenic ras pathway in vivo, Genes Dev. 12: 2469-74., 1998. 176. Tapon, N . and H a l l , A . Rho, Rac and Cdc42 GTPases regulate the organization of the actin cytoskeleton, Curr Opin Cell B i o l . 9: 86-92., 1997. 177. Hanks, S. K . and Polte, T. R. Signaling through focal adhesion kinase, Bioessays. 19: 137-45., 1997. 178. Dudek, H . , Datta, S. R., Franke, T. F . , Birnbaum, M . J., Y a o , R., Cooper, G . M . , Segal, R. A . , Kaplan, D . R., and Greenberg, M . E . Regulation of neuronal survival by the serine-threonine protein kinase A k t [see comments], Science. 275: 661-5, 1997. 179. W o n g , C . W . , Lee, A . , Shientag, L . , Y u , J., D o n g , Y . , K a o , G . , A l - M e h d i , A . B . , Bernhard, E . J., and Muschel, R. J. Apoptosis: an early event in metastatic inefficiency, Cancer Res. 61: 333-8., 2001. 180. Castro, M . P., M c D o n a l d , T. J., Qualman, S. J., and Odorisio, T. M . Cerebrospinal fluid gastrin releasing peptide in the diagnosis of leptomeningeal metastases from small cell carcinoma, Cancer. 91: 2122-6., 2001. 181. Lacroix, J., Becker, H . D . , Woerner, S. M . , Rittgen, W . , Drings, P., and von Knebel Doeberitz, M . Sensitive detection of rare cancer cells in sputum and peripheral blood samples of patients with lung cancer by preproGRP-specific R T - P C R , Int J Cancer. 92: 1-8., 2001. 182. Schally, A . V . , Comaru-Schally, A . M . , Plonowski, A . , Nagy, A . , Halmos, G . , and Rekasi, Z . Peptide analogs in the therapy of prostate cancer, Prostate. 45: 158-66., 2000. 183. Plonowski, A . , Nagy, A . , Schally, A . V . , Sun, B . , Groot, K . , and Halmos, G . In vivo inhibition of P C - 3 human androgen-independent prostate cancer by a targeted cytotoxic bombesin analogue, A N - 2 1 5 , Int J Cancer. 88: 652-7., 2000. 184. L e y t o n , J., G a r c i a - M a r i n , L . J., Tapia, J. A . , Jensen, R . T., and M o o d y , T. W . Bombesin and gastrin releasing peptide increase tyrosine phosphorylation of focal adhesion kinase and paxillin in non-small cell lung cancer cells, Cancer Lett. 162: 87-95., 2001. 185. Carroll, R. E . , Matkowskyj, K . A . , Tretiakova, M . S., Battey, J. F . , and Benya, R. V . Gastrin-releasing peptide is a mitogen and a morphogen in murine colon cancer, Cell Growth Differ. 11: 385-93., 2000. 186. Tsuda, T., Kusui, T., and Jensen, R. T. Neuromedin B receptor activation causes tyrosine phosphorylation of p l 2 5 F A K by a phospholipase C independent mechanism which requires p21rho and integrity of the actin cytoskeleton, Biochemistry. 36: 16328-37., 1997. 187. Rainaldi, G . , Calcabrini, A . , Arancia, G . , and Santini, M . T. Differential expression of adhesion molecules (CD44, I C A M - 1 and L F A - 3 ) in cancer cells grown in monolayer or as multicellular spheroids, Anticancer Res. 19: 1769-78., 1999. 188. Davies, C . D . , Muller, H . , Hagen, I., Garseth, M . , and Hjelstuen, M . H . Comparison of extracellular matrix in human osteosarcomas and melanomas growing as xenografts, multicellular spheroids, and monolayer cultures, Anticancer Res. 17: 4317-26., 1997. 189. Molenaar, W . M . and Muntinghe, F. L . Expression of neural cell adhesion molecules and neurofilament protein isoforms in Ewing's sarcoma of bone and soft tissue sarcomas other than rhabdomyosarcoma, H u m Pathol. 30: 1207-12., 1999. 190. Gluer, S., Zense, M . , and von Schweinitz, D . C e l l adhesion molecules and intermediate filaments on embryonal childhood tumors, Pathol Res Pract. 194: 773-80, 1998. 191. van Valen, F . , Hanenberg, H . , and Jurgens, H . Expression of functional very late antigen-alpha 1, -alpha 2, -alpha 3 and -alpha 6 integrins on Ewing's sarcoma and primitive  172  peripheral neuroectodermal tumour cells and modulation by interferon-gamma and tumour necrosis factor-alpha, E u r J Cancer. 14: 2119-25, 1994. 192. Chung, D . H . , Lee, J. I., Kook, M . C , K i m , J. R., K i m , S. Ft., C h o i , E . Y . , Park, S. H . , and Song, F£. G . I L K (betal-integrin-linked protein kinase): a novel immunohistochemical marker for Ewing's sarcoma and primitive neuroectodermal tumour, V i r c h o w s A r c h . 433: 113-7., 1998. 193. Scarpa, S., D ' O r a z i , G , M o d e s t i , M . , and M o d e s t i , A . E w i n g ' s sarcoma lines synthesize laminin and fibronectin, Virchows A r c h [a]. 410: 375-81, 1987. 194. Zhang, Z . , V u o r i , K . , Reed, J. C , and Ruoslahti, E . The alpha 5 beta 1 integrin supports survival of cells on fibronectin and up-regulates B c l - 2 expression, Proc Natl A c a d Sci U S A . 92: 6161-5., 1995. 195. Giancotti, F. G . and Ruoslahti, E . Integrin signaling, Science. 285: 1028-32., 1999. 196. Frisch, S. M . and Francis, H . Disruption of epithelial cell-matrix interactions induces apoptosis, J C e l l B i o l . 124: 619-26., 1994. 197. Y a n g , J. and Kornbluth, S. A l l aboard the c y c l i n train: subcellular trafficking of cyclins and their C D K partners, Trends C e l l B i o l . 9: 207-10., 1999. 198. Zhang, J., W u , L . , Peebles, C , and Triche, T. J . C y c l i n D I , a potential target of E W S F L I 1 fusion gene. In: American Association for Cancer Research Annual Meeting 2001, New Orleans, February 2001 2001. 199. Zinszner, H . , Albalat, R., and R o n , D . A novel effector domain from the R N A binding protein T L S or E W S is required for oncogenic transformation by C H O P , Genes & Dev. 8: 2513-2526, 1994. 200. Ichikawa, H . , Shimizu, K . , Hayashi, Y . , and O h k i , M . A n R N A - b i n d i n g protein gene, TLS/FUS, is fused to ERG i n human m y e l o i d leukemia with t(16;21) chromosomal translocation, Cancer Res. 54: 2865-2868, 1994. 201. Y a n g , L . , Embree, L . J., Tsai, S., and Hickstein, D . D . Oncoprotein T L S interacts with serine-arginine proteins involved in R N A splicing, J B i o l Chem. 273: 27761-4., 1998. 202. Chansky, H . A . , H u , M . , Hickstein, D . D . , and Y a n g , L . Oncogenic T L S / E R G and E W S / F l i - 1 fusion proteins inhibit R N A splicing mediated by Y B - 1 protein, Cancer Res. 61: 3586-90., 2001. 203. K u l i k , G . , K l i p p e l , A . , and Weber, M . J. Antiapoptotic signalling by the insulin-like growth factor I receptor, phosphatidylinositol 3-kinase, and A k t , M o l C e l l B i o l . 17: 1595606, 1997.  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0090557/manifest

Comment

Related Items