Role of Receptor Tyrosine Kinase Regulator Sprouty in Ovarian Cancer Cells

by Wai Kin So BSc, The Chinese University of Hong Kong, 2002 MPhil, The Chinese University of Hong Kong, 2004

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

DOCTOR OF PHILOSOPHY
in

The Faculty of Graduate Studies (Reproductive and Developmental Sciences)

THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) April, 2012 Wai Kin So 2012

Abstract Aberrant epidermal growth factor receptor (EGFR) activity contributes to the development of epithelial ovarian cancer (EOC), a common and lethal female malignancy. Elucidating the regulation of EGFR function will improve treatments for EOC and the survival of patients. This study aims to elucidate the role of Sprouty (SPRY) proteins, which are EGFR regulators, in EOC. The investigation began with demonstrating the downregulation of mRNA levels of two SPRY members, SPRY2 and SPRY4, in EOC tissues and/or cell lines. Deletion of the SPRY2 gene was found to cause reduced SPRY2 mRNA. Loss of the SPRY2 gene and thus its expression are particularly common in high-grade serous tumors, suggesting that SPRY2 deficiency may be involved in the pathogenesis of this prevailing subtype of EOC. The regulatory mechanisms of SPRY level are incompletely understood. The EGFR ligand EGF strongly upregulates SPRY4 protein level primarily through the ERK pathway. In addition, the PI3K/AKT pathway and hypoxia-inducible factor-1 (HIF-1!) have been shown to be involved in SPRY4 regulation, allowing the possibility that SPRY4 is regulated by micro-environmental (hypoxia) and genetic (PI3K mutation) abnormalities. Functionally, SPRY2 and SPRY4 counteract various aspects of EGFR activity and generally have tumor suppressor functions. First, in contrast to the EGFR, SPRY2 and SPRY4 prevent loss of cell adhesion by E-cadherin and therefore suppress cancer cell invasion. Second, SPRY4 inhibits PI3K/AKT signalling activated by EGF, as AKT ii

activation is enhanced in the absence of SPRY4. Finally, the HIF-1! oncogene has been identified as a novel SPRY4 target. In ovarian cancer cell lines, SPRY4 suppresses the basal and EGF-stimulated expression of HIF-1!. The negative effects of SPRY4 on HIF1! are also reflected by modulation of HIF-1 activity and target gene expression. SPRY4 has also been shown to destabilise HIF-1! protein, independent of the classic HIF-1! degradation pathway. The current study investigated the expression, regulation and function of SPRY in ovarian cancer. Understanding the tumor suppressor role of SPRY will not only enhance our knowledge about the pathophysiology of ovarian cancer but also identifies a possible therapeutic intervention against this lethal malignancy.

iii

Preface 1. A version of Chapter 2 has been submitted and is under revision. Wai-Kin So, Alice S. T. Wong, David G. Huntsman, C. Blake Gilks, and Peter C. K. Leung. Genetic Inactivation of Sprouty2 Promotes Epidermal Growth Factor-induced E-cadherin Downregulation and Invasion in Ovarian Cancer Cells. Contributions David G. Huntsman, C. Blake Gilks and I participated in design and performance of the study. I drafted the manuscript. Peter C. K. Leung, Alice S. T. Wong, David G. Huntsman and C. Blake Gilks read, critically revised and approved the manuscript. Ethics Approval Approvals for the study were obtained from the University of British Columbia Research Ethics Board (#H04-60102, #H02-61375 and #H03-70606) and written informed consents from all participants involved in the study were obtained.

2. A manuscript based on a version of Chapter 3 has been under preparation. Wai-Kin So, Man-Tat Lau, and Peter C. K. Leung. An Amphiregulin and Sprouty4 Loop Regulates Ovarian Cancer Cell Invasiveness Via and E-cadherin-dependent Mechanism. Contributions I participated in design of the study, Man-Tat Lau and I performed experiments. I drafted the manuscript. iv

Presentation 5th International Epithelial-Mesenchymal Transition Meeting, Singapore, Oct. 2011 Poster presentation on An Amphiregulin and Sprouty4 Loop Regulates Ovarian Cancer Cell Invasiveness Via and E-cadherin-dependent Mechanism. Wai-Kin So and Peter C. K. Leung.

Table of Contents Abstract ............................................................................................................................... ii! Preface................................................................................................................................ iv! Table of Contents............................................................................................................... vi! List of Tables ..................................................................................................................... ix! List of Figures ..................................................................................................................... x! List of Abbreviations ........................................................................................................ xii! Acknowledgements.......................................................................................................... xvi! Chapter 1 Introduction ........................................................................................................ 1! 1.1 Ovarian cancer .......................................................................................................... 1! 1.2 Tumorigenesis of EOCs............................................................................................ 1! 1.3 Classification of EOCs.............................................................................................. 3! 1.3.1 High-grade serous carcinoma (HGSC) .............................................................. 3! 1.3.2 Low-grade serous carcinoma (LGSC) ............................................................... 5! 1.3.3 Endometrioid carcinoma (EC) ........................................................................... 6! 1.3.4 Clear cell carcinoma (CCC)............................................................................... 7! 1.3.5 Mucinous tumors ............................................................................................... 7! 1.4 Epidermal growth factor (EGF) and the EGF receptor (EGFR)............................... 8! 1.4.1 EGFR structure and signaling............................................................................ 8! 1.4.2 EGFR expression and mutations in ovarian cancer ........................................... 9! 1.4.3 EGFR ligands................................................................................................... 11! 1.4.4 Functions of the EGFR and it ligands in ovarian cancer ................................. 13! 1.4.5 Clinical activities of EGFR-targeted therapies. ............................................... 15! 1.5 Sprouty (SPRY) ...................................................................................................... 16! 1.5.1 SPRY structure................................................................................................. 16! 1.5.2 SPRY functions and mechanisms .................................................................... 17! 1.5.3 Regulation of SPRY activity............................................................................ 18! 1.5.4 SPRY expressions in cancer ............................................................................ 19! 1.5.5 SPRY as tumor suppressor genes .................................................................... 19! 1.6 Hypoxia-inducible factor-1 alpha (HIF-1!)............................................................ 19! 1.6.1 HIF-1! structure............................................................................................... 20! 1.6.2 HIF-1! regulation ............................................................................................ 21! 1.6.2.1 Hypoxia..................................................................................................... 21! 1.6.2.2 Regulation of HIF-1! in normoxia........................................................... 22! 1.6.2.2.1 VHL mutations ................................................................................... 22! 1.6.2.2.2 AKT and PI3K ................................................................................... 23! 1.6.2.2.3 Glycogen synthase kinase 3" (GSK3")............................................. 25! 1.6.2.2.4 Heat shock protein 90 (HSP90) ......................................................... 26! 1.6.2.2.5 Hormonal regulation .......................................................................... 27! 1.6.3 HIF-1! expressions in cancer .......................................................................... 27! 1.6.4 HIF-1! functions in cancer.............................................................................. 28! vi

1.6.4.1 Angiogenesis............................................................................................. 28! 1.6.4.2 Metastasis.................................................................................................. 29! 1.7 Hypothesis and objectives....................................................................................... 30! 2.1 Introduction............................................................................................................. 37! 2.2 Materials and methods ............................................................................................ 39! 2.2.1 The Human Exonic Evidence-Based Oligonucleotide microarray (HEEBO). 39! 2.2.2 Molecular inversion probe (MIP) copy number analysis ................................ 39! 2.2.3 The Cancer Genome Atlas (TCGA) ................................................................ 40! 2.2.5 Real-time PCR ................................................................................................. 41! 2.2.6 Antibodies ........................................................................................................ 41! 2.2.7 Invasion assay .................................................................................................. 42! 2.2.8 Statistical analysis............................................................................................ 42! 2.3 Results..................................................................................................................... 43! 2.3.1 Levels of SPRY mRNA in ovarian tumors of different pathological subtypes 43! 2.3.2 Levels of SPRY mRNA in immortalised ovarian surface epithelium (IOSE) and EOC-derived cell lines.............................................................................................. 44! 2.3.3 A deletion event in the proximity of the SPRY2 locus..................................... 44! 2.3.4 SPRY2 deletion may lead to reduced SPRY2 mRNA level............................. 45! 2.3.5 SPRY2 reversed EGF-suppressed E-cadherin protein expression and antagonized EGF-induced cell invasion ................................................................... 46! 2.3.6 SPRY2 and E-cadherin proteins displayed a positive correlation in human ovarian cancer cell lines and tumors......................................................................... 47! 2.4 Discussion ............................................................................................................... 47! Chapter 3 An amphiregulin and Sprouty4 loop regulates ovarian cancer cell invasiveness via an E-cadherin-dependent mechanism ......................................................................... 60! 3.1 Introduction............................................................................................................. 60! 3.2 Materials and methods ............................................................................................ 62! 3.2.1 Cell culture and reagents.................................................................................. 62! 3.2.2 Transfection ..................................................................................................... 63! 3.2.3 Real-time PCR ................................................................................................. 63! 3.2.4 Western blot analysis ....................................................................................... 64! 3.2.5 Invasion assay .................................................................................................. 64! 3.2.6 Statistical analysis............................................................................................ 65! 3.3 Results..................................................................................................................... 65! 3.3.1 AREG promoted invasion of ovarian cancer cells........................................... 65! 3.3.2 AREG reduced E-cadherin levels, and E-cadherin overexpression blocks AREG-induced invasion ........................................................................................... 66! 3.3.3 AREG suppressed E-cadherin level and promotes cell invasion via the EGFR ................................................................................................................................... 66! 3.3.4 AREG induced Slug expression....................................................................... 67! 3.3.5 The MAPK/ERK and PI3K/AKT pathways mediated the effects of AREG on SLUG mRNA and E-cadherin levels and cell invasion ............................................ 67! 3.3.6 AREG induced SPRY4 expression.................................................................. 67! 3.3.7 SPRY4 knockdown enhanced AREG-induced E-cadherin suppression and invasion ..................................................................................................................... 68! 3.4 Discussion ............................................................................................................... 68! vii

Chapter 4 Sprouty4 feedback regulates epidermal growth factor/AKT/hypoxia-inducible factor-1 alpha axis in ovarian cancer cells........................................................................ 82! 4.1 Introduction............................................................................................................. 82! 4.2 Materials and methods ............................................................................................ 84! 4.2.1 Cell culture and reagents.................................................................................. 84! 4.2.2 Transfection ..................................................................................................... 84! 4.2.3 Real-time PCR ................................................................................................. 85! 4.2.4 Western blot analysis ....................................................................................... 85! 4.2.5 Luciferase assay ............................................................................................... 86! 4.2.6 Statistical analyses ........................................................................................... 86! 4.3 Results..................................................................................................................... 87! 4.3.1 EGF increased SPRY4 in ovarian cancer cells ................................................ 87! 4.3.2 The MEK/ERK and PI3K/AKT pathways mediated the effects of EGF on SPRY4 levels ............................................................................................................ 87! 4.3.3 EGF induced HIF-1! via the PI3K/AKT pathway .......................................... 88! 4.3.4 HIF-1! plays a minor role in EGF-induced SPRY4 level ............................... 88! 4.3.5 SPRY4 overexpression reversed EGF-induced HIF-1! levels and HIF-1 activity....................................................................................................................... 89! 4.3.6 SPRY4 knockdown enhanced the effect of EGF on HIF-1! ........................... 89! 4.3.7 AKT pathway mediated HIF-1! regulation by EGF and SPRY4.................... 90! 4.4 Discussion ............................................................................................................... 90! 5.1 Introduction........................................................................................................... 103! 5.2 Materials and methods .......................................................................................... 105! 5.2.1 Cell culture and reagents................................................................................ 105! 5.2.2 Transfection ................................................................................................... 105! 5.2.3 Real-time PCR ............................................................................................... 106! 5.2.4 Western blot analysis ..................................................................................... 106! 5.2.5 Statistical analysis.......................................................................................... 107! 5.3 Results................................................................................................................... 107! 5.3.1 SPRY4 negatively regulated HIF-1! expression levels in ovarian cancer cells ................................................................................................................................. 107! 5.3.2 SPRY4 negatively regulated HIF-1 activity .................................................. 108! 5.3.3 SPRY4 regulated HIF-1! protein half-life without affecting Hif-1! mRNA levels ....................................................................................................................... 109! 5.3.4 HIF-1! modulation by SPRY4 was independent of PHD activity ................ 109! 5.4 Discussion ............................................................................................................. 110! References....................................................................................................................... 129!

viii

List of Tables Table 2.1 Comparison between mean SPRY2 mRNA levels of various histopathological types using Students t test.........51 Table 2.2 MIP analysis of loss of the markers flanking the SPRY2 and SPRY4 loci in ovarian tumors...52

ix

List of Figures Figure 1.1 The chart illustrates the pathogenesis of epithelial ovarian cancers....33 Figure 1.2 The structure and signaling of EGFR.......34 Figure 1.3 A schematic diagram of human SPRY depicting its structure, its interacting partners and the corresponding functional consequences. 35 Figure 1.4 The diagram illustrates the regulation of HIF-1!. ..36 Figure 2.1 The box plot displays the mean SPRY2 mRNA levels in serous (high-grade, low-grade or borderline), endometrioid (carcinoma or borderline tumor), clear cell ovarian tumors and normal Fallopian tube. (HEEBO array data).54 Figure 2.2 Comparison of SPRY2 and SPRY4 mRNA levels in immortalised OSE (IOSE) and ovarian cancer cell lines by real-time PCR.....55 Figure 2.3 A schematic representation of the MIP copy number assay results of 28 highgrade serous carcinomas........56 Figure 2.4 The effect of SPRY2 on EGF-induced E-cadherin suppression and cell invasion......57 Figure 2.5 The correlation between SPRY2 and E-cadherin protein expression in ovarian cancers.......59 Figure 3.1 AREG promoted ovarian cancer cells invasion.......73 Figure 3.2 AREG reduced E-cadherin levels and E-cadherin overexpression blocked AREG-induced invasion. ..74 Figure 3.3 AREG suppressed E-cadherin level and promoted cell invasion via the EGFR. ...75 Figure 3.4 AREG induced SLUG mRNA .....76 Figure 3.5 The MAPK/ERK and PI3K/AKT pathways mediated the effects of AREG on Slug and E-cadherin level and cell invasion. ....77 Figure 3.6 AREG induced SPRY4....79 Figure 3.7 SPRY4 knockdown enhanced AREG-induced E-cadherin suppression and invasion..........80 Figure 4.1 EGF induced SPRY4 level in ovarian cancer cells......94 Figure 4.2 The MEK/ERK and PI3K/AKT pathways mediated the effects of EGF on SPRY4 levels.....95 x

Figure 4.3 EGF induced HIF-1! and HIF-1 activity via the PI3K/AKT pathway..96 Figure 4.4 HIF-1! plays a minor role in EGF-induced SPRY4 level.......97 Figure 4.5 SPRY4 overexpression reversed EGF-induced HIF-1! expression and HIF-1 activity....99 Figure 4.6 SPRY4 knockdown enhanced EGF effect on HIF-1!...............101 Figure 4.7 The PI3K/AKT pathway mediated HIF-1! regulation by EGF and SPRY4.....102 Figure 5.1 SPRY4 negatively regulated HIF-1! levels in ovarian cancer cells. ......113 Figure 5.2 SPRY4 negatively regulated HIF-1 activity. .114 Figure 5.3 SPRY4 regulated HIF-1! protein half-life without affecting Hif-1! mRNA levels. ......................115 Figure 5.4 HIF-1! modulation by SPRY4 acts independently of PHD activity. ..117 Figure 6.1 The diagram summarizes the findings. ..128

xi

List of Abbreviations AREG BRCA1 BTC CCC CTNNB DM DMSO EC ECD ECM EGF EMT EOC EPI ERBB ERK FBS FGF GAPDH GIST GSK-3 Amphiregulin Breast cancer 1, early onset Betacellulin Clear cell carcinoma Cadherin-associated protein beta Double mutant Dimethyl sulfoxide Endometrioid carcinoma Extracellular domain Extracellular matrix Epidermal growth factor Epithelial-mesenchymal transition Epithelial ovarian cancer Epiregulin Erythroblastic leukemia viral oncogene homolog Extracellular signal regulated protein kinase Fetal bovine serum Fibroblast growth factor Gylceraldehyde-3-phosphate dehydrogenase Gastrointestinal stromal tumor Glycogen synthase kinase-3

xii

HB-EGF HEEBO HER2 HGF HGSC HIF-1! HIF-1" HRE HRG HSP ICM IGF-I IOSE JM JNK LGSC LH LOH mAb MAPK MDCK MDM2

Heparin-binding epidermal growth factor Human Exonic Evidence-Based Oligonucleotide Human epidermal growth factor receptor 2 Hepatocyte growth factor High-grade serous carcinoma Hypoxia-inducible factor-1 alpha Hypoxia-inducible factor-1 beta Hypoxia responsive element Heregulin Heat shock protein Intracellular domain Insulin-like growth factor-I Immortalized ovarian surface epithelium Juxtamembrane domain Jun N-terminal protein kinase Low-grade serous carcinoma Luteinizing hormone Loss of heterozygosity Monoclonal antibody Mitogen-activated protein kinase Madin-darby canine kidney Murine double minute 2

xiii

MET MIP MMP mTOR NSCLC ODD OSE PAI-1 PDGF PI3K PKA PHD PTEN PVDF RAS RBD RCC RD RING ROS RTK SD SDS

Mesenchymal epithelial transition factor Molecular inversion probe Matrix metalloproteinase Mammalian target of rapamycin Non-small-cell lung cancer Oxygen-dependent degradation Ovarian surface epithelium Plasminogen activator inhibitor-1 Platelet-derived growth factor Phosphatidylinositol 3-kinase Protein kinase A Prolyl hydroxylase Phosphatase and tensin homolog deleted on chromosome 10 Polyvinylidene fluoride Rat sarcoma RAF1-binding domain Renal clear cell carcinoma Regulatory domain Really interesting new gene Reactive oxygen species Receptor tyrosine kinase Standard deviation Sodium dodecyl sulphate xiv

siRNA SIAH SNP SOS STAT STIC Sp1 SPRY TCGA TESK1 TGF! TGF" TKD TMD TKI uPA VEGF VHL WT-1 WT ZEB1

Small interference RNA Seven in absentia homolog Sodium nitroprusside Son of Sevenless Signal transducer and activation of transcription Serous tubal intraepithelial carcinoma Stimulating protein 1 Sprouty The Cancer Genome Atlas Testicular protein kinase 1 Transforming growth factor alpha Transforming growth factor beta Tyrosine kinase domain Transmembrane domain Tyrosine kinase inhibitor Urokinase-type plasminogen activator Vascular endothelial growth factor Von Hippel-Lindau Wilms tumor 1 Wild-type Zinc finger E-box-binding homeobox 1

xv

Acknowledgements First of all, I would like to express my greatest gratitude to my supervisor, Dr. Peter C.K. Leung for his invaluable support, patient guidance and warm encouragement during my study. In addition, I would like to express my sincere gratitude to my supervisory committee members, Drs. Geoffrey Hammond, Blake Gilks, Y.Z. Wang and Mark Carey for their scientific criticisms and advice and my experiments and thesis. I would also want to thanks Drs. Blake Gilks, David Huntsman and Alice S.T. Wong from the University of Hong Kong for their help and opinions of my experiments. I also would like to thank Dr. Christian Klausen and Ms. Roshni Nair for their help over the years. Here I also thank the Interdisciplinary Womens Reproductive Health Research Training Program providing me scholarship.

Lastly, I wish to thank my parents and family members for their love, understanding and endless support throughout my study.

xvi

Chapter 1 Introduction 1.1 Ovarian cancer Ovarian cancer is the second most prevalent gynaecological malignancy (after endometrial cancer) among women in the United States (1). A womans lifetime risk of developing ovarian cancer is 1 in 70 (2). Ovarian cancer is the most lethal of all gynaecologic malignancies with a 5-year survival rate of 30% 40% (3), and overall survival has not changed in decades (4). The poor patient outcome is mainly due to the lack of a reliable screening test for early disease detection. Most ovarian carcinomas are diagnosed at a late stage, after invasion to the intra-peritoneal cavity, and are thus inoperable (3). In addition, many ovarian carcinomas respond poorly to therapy (1, 5) or recur with the development of chemoresistance (1). Based on their origins, approximately 90% of all human ovarian cancers are categorised as epithelial ovarian carcinomas (EOCs), which originate in the ovarian surface epithelium (OSE), and the rest are derived from granulosa, stromal or germ cells (3). However, evidence supporting an origin of these tumors outside the ovary is accumulating. Therefore, the paradigm of the OSE as the precursor of EOCs is being challenged, and the cellular origin of EOC remains controversial. 1.2 Tumorigenesis of EOCs Fathalla proposed the incessant ovulation theory in 1971, suggesting that repeat ovulation, surface rupture and subsequent repair lead to the trapping of the OSE in the ovarian stroma and formation of inclusion cysts. Inclusion cysts secrete somatic growth factors that lead to cell proliferation, genetic aberrations and finally malignant 1

transformation (6). This hypothesis is supported by substantial epidemiological data. One casecontrol study of 150 ovarian cancer patients under the age of 50 years demonstrated that the risk of ovarian cancer decreased with increasing numbers of live births, increasing numbers of incomplete pregnancies and the use of oral contraceptives (7). Another prevailing hypothesis addressing the development of ovarian cancer was proposed by Cramer and Welch in 1983. Their gonadotropin theory proposed that excessive gonadotropin stimulation contributes to ovarian carcinogenesis (8). The risk of ovarian cancer increases during the perimenopausal period, when serum gonadotropin levels peak and thereafter remain elevated (9, 10). Moreover, only 10% 15% of tumors appear in premenopausal women (11). Likewise, polycystic ovary syndrome patients (with high luteinising hormone levels) are more prone to ovarian cancer (12). Epidemiologic evidence supports the idea that pregnancies, breast feeding, and oral contraceptive use, which suppress pituitary gonadotropin secretion, reduce the risk of ovarian cancer (13-16). Mesothelial OSE has been proposed to undergo Mllerian differentiation during ovarian carcinogenesis. As a result, the OSE loses its mesothelial characteristics and acquires the properties of the Mllerian system (3). This characteristic readily explains why histological and immunocytochemical analyses of EOCs including serous (Fallopian tube), endometrioid (endometrium), clear cell (vaginal) and mucinous (endocervix) reveal characteristics of Mllerian epithelia rather than mesothelial tumors (3). For instance, serous carcinomas express PAX2 or PAX8, which are normally expressed in Fallopian tube epithelia (17, 18). However, the similarities of EOCs to Mllerian tumors support the possibility of a Mllerian lineage and thus provide an alternative theory of EOC histogenesis (4, 19). For instance, instead of OSE, serous 2

carcinomas in Fallopian tube are proposed to be precursors of high-grade serous tumors (20). On the other hand, endometrioid and clear cell carcinomas are suggested to arise from endometriosis (21). 1.3 Classification of EOCs Rather than a single entity, EOCs are heterogeneous diseases comprising tumors of different subtypes. Serous (high-grade and low-grade), endometrioid, clear cell and mucinous are the major subtypes of EOCs. Their distinct genetic abnormalities and oncogenic pathways and differential responses to chemotherapy (1, 22) emphasise the importance of subtype diagnosis and subtype-specific therapies (23). 1.3.1 High-grade serous carcinoma (HGSC) HGSC is the most common type of EOC and accounts for 60% of EOC cases and 90% of all serous carcinomas (1). HGSCs are usually diagnosed at an advanced stage (22) 85% of patients present with widespread peritoneal metastases (1). Accurate diagnosis of HGSC depends on squamous differentiation and areas with solid growth (22). For problematic cases, Wilms tumor 1 (WT-1) is an useful immunomarkers due to its specific expression in HGSCbut not other EOCs (24). Almost all HGSCs harbour TP53 mutations (up to 97.6%) (25) (Fig. 1.1). Additional mechanisms including TP53 loss of heterozygosity (LOH) (26) and MDM2 or MDM4 (specific p53 inhibitors) amplification (25) contribute to aberrant p53 expression. Although p53 dysfunction is a ubiquitous feature of HGSCs, the data regarding the prognostic value of p53 are conflicting (25, 27).

Other characteristic genetic abnormalities of HGSCs include loss of BRCA1 and BRCA2 (Fig. 1.1). Both BRCA genes can be inactivated through germline or somatic mutations, and BRCA1 is also silenced epigenetically through promoter hypermethylation (28), leading to familial and sporadic HGSC. Among BRCA-related hereditary ovarian cancers, up to 57% -100% are HGSCs (22, 29), highlighting the importance of BRCA in HGSC tumorigenesis. Owing to their role in DNA repair, BRCA and p53 dysfunction lead to genomic instability and allow accumulation of further genetic alterations along the path of tumor progression. HGSCs normally display aneuploidy and intratumoral genetic heterogeneity (30). In addition to the conventional theory of development of EOCs from the OSE or cortical inclusion cysts (3), there is an emerging view that EOCs may derive from cells outside the ovary, which subsequently implant and expand within the ovary and present as tumors originating in ovary (4). This mechanism may explain why the OSE has a mesothelial phenotype whereas HGSCs have Mllerian morphology and express a Mllerian marker (PAX8) but not mesothelial markers (18). Serous tubal intraepithelial carcinomas (STICs) in the fimbriated end of Fallopian tubes are proposed to be the precursors of HGSCs of the ovary (20). In addition to the high incidence of dysplastic changes in the Fallopian tubes of women with a genetic predisposition for familial ovarian cancer (31), up to 50% - 60% of HGSC patients also have STICs (32, 33). A clonal relationship between STICs and HGSCs is further supported by identical TP53 mutations in the tumors (34) (Fig. 1.1). A gene profiling study showed that HGSCs and the Fallopian tube epithelium share similar gene expression profiles, and HGSCs were 4

found to be less closely related to the OSE (35), suggesting that HGSCs may originate from the Fallopian tube epithelium rather than the OSE. Surgery is the primary treatment for all ovarian cancers including HGSC. Although the majority (70% - 80%) of HGSC patients show initial responses to platinum/taxane-based chemotherapy (1, 22), 70% of them experience a recurrence, and many recurrent tumors become resistant to etoposide and doxorubicin (1). All of these factors together, combined with the late diagnosis, contribute to the poor prognosis of HGSC patients, who have a 10% - 20% 5-year survival rate (1). 1.3.2 Low-grade serous carcinoma (LGSC) Among all ovarian serous cancers, 10% are LGSCs. Differentiation diagnosis between LGSC and HGSC is usually based in the uniformity of nuclei. In addition, psammoma bodies, differentiated architecture with papillary growth are characteristic feathers of LGSC (22). The prognosis of LGSC patients is better than that for HGSC patients (1). Rarely, LGSCs are found concurrently with HGSCs, suggesting the progression of LGSC to HGSG (36) (Fig. 1.1). However, it is well accepted that lowgrade and high-grade serous carcinomas are fundamentally distinct tumors (22). LGSCs rarely harbour TP53 or germline BRCA mutations (1, 22). Instead, activating mutations of KRAS, BRAF, (37) or ERBB2 (38) are common. Accordingly, 60-70% of LGSCs express active MAPK (39), highlighting the role of the KRAS/BRAF/MEK/MAPK pathway in LGSC pathogenesis. Mutations of KRAS and BRAF are also found in serous borderline tumors, which are believed to be the immediate precursors of LGSCs (1) (Fig. 1.1).

1.3.3 Endometrioid carcinoma (EC) EC is the second most common histologic subtype of EOC, accounting for 10% 20% of cases (40). ECs are mostly diagnosed at stages I and II, which accounts for the better prognosis than other subtypes (22). EC are commonly characterized with squamous or mucinous differentiation (22). Some clinicians diagnosed high-grade carcinomas with glandular differentiation as EC, however, this diagnostic is doubted as glandular differentiation is also involved in serous neoplasia (22). Similar to high-grade serous, high-grade EC may arise de novo or by rare stepwise progression from borderline and low-grade tumors (40). High-grade de novo ECs typically have mutations in TP53 (41) (Fig. 1.1). On the other hand, the genetic defects found in borderline or low-grade endometrioid tumors are also commonly found in highgrade tumors. For instance, CTNNB1 (38% - 50% of cases) and PTEN (phosphatase and tensin homolog deleted on chromosome 10) (20%) mutations have been identified in both, suggesting a precursor role of these lesions (22, 42) (Fig. 1.1). The CTNNB1 gene mutation encodes a !-catenin protein that is resistant to degradation, which allows stabilization and nuclear accumulation of !-catenin (43) and can cause endometrioid carcinogenesis. There is also evidence suggesting that endometriosis gives rise to EC. ECs share similar molecular signatures with adjacent endometriosis, including LOH of chromosome 12 (44) and 10q23 (45) mutations in ARID1A tumor suppressor gene, which encodes adenine-thymine (AT)-rich interactive domain-containing protein 1A (21) and PTEN (45). Furthermore, up to 42% of EC cases are associated with endometriosis (46), and 67% - 100% of endometriosis cases are associated with coexisting EC (19) (Fig. 1.1).

1.3.4 Clear cell carcinoma (CCC) CCCs account for approximately 10% of all EOC cases. Though CCC patients are usually diagnosed at stages I or II, chemo-resistance, relapse after surgery and the aggressiveness of CCCs make the prognosis of CCC patients worse than that of other ovarian carcinomas at early stages (1, 22). The presence of clear cells is not sufficient for a diagnosis of CCC. Additional criteria for CCC include papillary/tubulocystic architecture and low mitotic rate (22). In sharp contrast to HGSCs, most CCCs contain wild-type TP53 (47) and BRCA1/2 (48). CCCs are best characterised by aberrant activation of PI3K signalling, including activating mutations of PI3KCA (42, 47), loss of PTEN expression (49) and activated mTOR (mammalian target of rapamycin) (50) (Fig. 1.1). Recently, two groups demonstrated frequent somatic mutation of the ARID1A tumor suppressor gene in CCCs (21, 51). Furthermore, ARID1A mutations are found in CCCs as well as in adjacent atypical endometriosis, but not in distant endometriosis (21). This difference implicates ARID1A mutations as early events in the transformation of endometriosis into cancer and strengthens the association between endometriosis and clear cell carcinogenesis. 1.3.5 Mucinous tumors Although mucinous tumors constitute 10% - 15% of all primary ovarian tumors, most of them are benign and borderline tumors. Only 3% - 4% are mucinous carcinomas (22), making them one of the rarest types of EOCs (4). Most of these tumors show gastrointestinal differentiation and contain goblet cells, which are key features for their

diagnosis. However, mucinous tumors are often heterogeneous, making adequate sampling for histological examination critical (22). The origin of mucinous tumors are unclear, and based on their non-Mllerian phenotype, development from cortical inclusion cysts is unlikely (4). KRAS mutations are the most frequent genetic defects in mucinous carcinomas (85% of cases) and represent an early event in mucinous tumorigenesis. In addition, the ERBB2 gene is amplified in 15% to 20% of mucinous carcinomas (22) (Fig. 1.1). 1.4 Epidermal growth factor (EGF) and the EGF receptor (EGFR) Numerous steroids, growth factors and cytokines secreted by ovary stromal or cancer cells promote neoplastic transformation and progression of ovarian cancer (52). Of these, EGF-like growth factors and their cognate receptors have been the most extensively studied. 1.4.1 EGFR structure and signaling The EGFR, also known as ERBB1/HER1, together with HER2/neu, ERBB3/HER3 and ERBB4/HER4 comprise the ERBB (or HER) family (53). All four members are Type I transmembrane tyrosine kinase receptors that consist of an extracellular domain (ECD), a transmembrane domain (TMD) and an intracellular domain (ICD) (Fig. 1.2). The ICD can be functionally further divided into i) the juxtamembrane domain (JD), which mediates EGFR internalisation and signalling specificity (54, 55); ii) the tyrosine kinase domain (TKD), which possesses intrinsic kinase activity (56); and iii) the carboxy-terminal regulatory domain (RD). Upon ligand binding, the ECD undergoes a conformational change that allows dimerization of ERBBs (Fig. 1.2). Dimerization can 8

be homodimerization or heterodimerization between various ERBBs (57) and HER2 is the most common binding partner for ERBBs (58). In the dimer complexes, the TKD undergoes tyrosine phosphorylation. Subsequently, tyrosine residues in the RD are transphosphorylated, in turn recruiting diverse cytoplasmic adaptor proteins and activating a variety of signalling pathways (53). In ERBB3, a single amino substitution in the TKD abrogates the intrinsic kinase activity, making ERBB3 homodimers inactive and the signal transduction of ERBB3 relies on heterodimerization with other ERBB members (56). The ERBB family members invariably activate the ERK cascade through recruiting the adaptor proteins Grb2 or Shc (59) (Fig. 1.2). The PI3K/AKT pathway is triggered by most activated ERBB dimers, which have different kinetics and potency. Such differences are probably due to the fact that ERBB3 and ERBB4 are capable of direct binding to the PI3K p85 regulatory subunit, whereas the EGFR and HER2 couple indirectly with p85 through adaptor proteins or ERBB3/ERBB4 (60). PhosphorylatedAKT was detected in 68% of ovarian carcinomas on a tissue microarray (61). Activation of these pathways in cancer has been associated with increased cell survival, growth, angiogenesis and metastasis (57, 62). Furthermore, aberrant levels of the EGFR and HER2, as well as their downstream ERK and AKT, lead to resistance to platinum-based chemotherapy (61, 63, 64). 1.4.2 EGFR expression and mutations in ovarian cancer Aberrant EGFR activity is achieved through a variety of mechanisms, including overproduction of ligands, overexpression of receptors and constitutive activation of 9

receptors through mutation or loss of regulators (57). EGFR overexpression and mutation are frequent events in human malignancies; for example, the EGFR gene is amplified in up to 40% of gliomas (65). There are many published reports regarding EGFR expression and its prognostic significance in ovarian cancer. In general, increased EGFR is correlated with more aggressive diseases and poorer patient outcomes (53). In various studies, the frequency of immunohistochemical detection of EGFR expression in ovarian cancer tissues ranges from 4% - 100% (53). EGFR gene amplification is seen in 43% of serous carcinomas (66), and EGFR overexpression is more common in serous (66, 67) and endometrioid (68) carcinoma subtypes. Both a higher gene copy number and protein expression correlate with the serous subtype (66). Higher levels of the EGFR are found in omental metastases (69) and carcinomas of advanced stages (68, 70), implicating a role for the EGFR in tumor progression and metastasis. This role may explain the fact that high EGFR expression correlates with unfavourable outcomes in terms of overall survival (66-68, 71) and disease-free survival (67). Increased EGFR expression is also associated with carcinomas (compared with LMP or benign tumors) (70, 72), higher tumor grade (66, 73), higher proliferative index (66, 73), larger residual tumor size (66, 68), and advanced age of patients (66, 74). However, data from numerous studies are highly variable and contradictory; many investigators have found no relationship between the EGFR and the clinical features and prognostic factors mentioned above (53). In addition to gene amplification and protein overexpression, EGFR hyperactivity may arise from EGFR gene mutations (75). An 801-base pair in-frame deletion in the extracellular domain of the EGFR results in a constitutively active EGFR variant III (EGFRvIII) mutant. Independent of ligands, this mutant is constitutively active in terms 10

of receptor dimerization, autophosphorylation and downstream signalling activation (75, 76). EGFRvIII is expressed in a large proportion of gliomas and breast carcinomas and 75% of ovarian carcinomas (77). However, other reports showed that EGFRvIII is rare in EOC (66, 78) Expression of this mutant in an EOC cell line causes epithelialmesenchymal (EMT) transition, cell migration and metastasis (79). Mutations in the catalytic domain (exon 18, 19, 20 or 21) of EGFR are less frequent in ovarian carcinomas (66, 74, 80). Activating mutations of the EGFR enhance the sensitivity of lung cancer patients to EGFR inhibitors (81). In ovarian cancers patients, responses to gefitinib (a small molecule tyrosine kinase inhibitor) have been reported to be independent of EGFR mutational status in one study (82), but another reported that EGFR-activating mutations have also been reported to confer patient responses to gefitinib and lead to longer progression-free survival (67). In contrast to numerous investigations of EGFR expression and mutations, only a few studies regarding EGFR activation status in ovarian carcinomas have been published. These studies reported values for EGFR activation, as detected by immunohistochemistry, of 11.8%, 35% and 100% (62, 83, 84). EGFR phosphorylation was also found to be positively correlated with a metastasis-promoting protein (matrix metalloproteinase 9) and negatively correlated with E-cadherin, which inhibits metastasis (84). 1.4.3 EGFR ligands The activation of ERBBs is mediated through ligand-dependent mechanisms. Ligands of ERBBs contain an EGF-like domain and three intramolecular disulphide bonds and are synthesised as plasma membrane-bound precursors (85). Based on their 11

binding specificity, they are classified into four groups: EGF, transforming growth factor! (TGF-!) and amphiregulin (AREG) bind the EGFR exclusively; epiregulin (EPI), betacellulin (BTC) and heparin-binding-EGF (HB-EGF) have dual specificity and bind the EGFR and ERBB4; and the heregulin (HRG) family constitute the remaining two groups, which bind both ERBB3 and ERBB4 (HRG-1 and HRG-2) or exclusively to ERBB4 (HRG-3 and HRG-4) (53). No HER2-specific ligand has been identified yet (57). Most of these ligands are overexpressed in and correlate with the clinical features of different types of cancers, including ovarian cancer (57). EGF has been detected in borderline, low-grade tumors, carcinomas and EOC cell lines (53, 86). Furthermore, up to 72% of epithelial ovarian tumors that are examined are immunopositive for EGF, and a higher EGF expression level has been found in mucinous cystadenocarcinomas compared with that in LMP mucinous tumors (87). The EGFR-exclusive ligand TGF-! is frequently detected in EOCs, with one study showing expression in 77% of tissues examined (53). The expression of TGF-! correlates with disease grade and stage (86). Concomitant high TGF-! levels and peak incidence of the disease in post-menopausal women suggest the importance of this growth factor in ovarian cancer pathology (88). HB-EGF expression is significantly higher in ovarian cancer than in normal ovaries (89, 90). Furthermore, elevated HB-EGF protein levels were detected in ascites fluid from ovarian cancer patients compared with that in the peritoneal fluid of normal women (90). A higher HBEGF expression level is significantly associated with shorter progression-free survival of patients (89). Further, the levels of AREG found in ovarian cancer cell lines, tissues and the peritoneal fluid of ovarian cancer patients are higher than those of TGF-! and EGF

12

(89, 91, 92). To date, however, no studies have reported the expression of the other two EGFR-binding ligands, EPI and BTC, in ovarian carcinomas tissues (53). It is important to note that the ERBB family members can also be transactivated by heterologous signals including those from G protein coupled receptors, independent of their cognate ligands (93-96). Cross activation of ERBBs can be achieved through direct phosphorylation or induction of ectodomain shedding of ligands (90). For example, the chemokine CXCL1 transactivates the EGFR through proteolytic cleavage of HB-EGF, leading to MAPK activation and the proliferation of ovarian cancer cells (97). 1.4.4 Functions of the EGFR and it ligands in ovarian cancer EGF and TGF-! are the most well-studied ligands of the EGFR. They form autocrine loops with the EGFR and have numerous regulatory functions in OSE and ovarian cancer cells. Gene profiling studies of rat OSE have demonstrated that genes involved in the cell cycle, proliferation, and apoptosis are targets of EGF (98). EGF treatment increases the proliferation of human primary OSE cells and ovarian cancer cells in culture (99, 100). In vitro treatment with TGF-! increased cell proliferation in eight ovarian cancer cell lines in one study (101). TGF-! is also synthesised by cancer cell lines (102), and disruption of the TGF-!/EGFR autocrine loop by a TGF-! neutralising antibody or antisense oligodeoxynucleotides blocks proliferation (102, 103). Accordingly, the in vivo growth of inoculated tumors is also blocked by TGF-! antibody treatment (104). Interestingly, growth regulation by AREG is unique, as AREG has a biphasic effect on OSE and ovarian cancer cells (105).! Ovarian cancer cell lines stably transfected with an antisense EGFR also show decreased proliferation (106). With its potent mitogenic

13

potential, the EGFR and ligand autocrine loops confer a selective advantage for cells and are implicated in clonal expansion and in the accumulation of mutations within tumors. In addition to cell growth, the EGFR and its ligands are implicated in metastasis. In response to EGF treatment, OSE cells (107) and ovarian cancer cells undergo EMT (108, 109) and EMT-associated N-cadherin and vimentin expression changes (108). It is important to note that during EMT, cells adopt a fibroblastic phenotype with reduced intercellular contact and increased cell motility. Consistent with these observations, tumor cell migration and invasion have been repeatedly demonstrated to be stimulated by EGF (84, 99, 108, 110, 111). Motility may be modulated through regulating the integrin system, which comprises transmembrane receptors forming bridges between the extracellular matrix (ECM) and the cytoskeleton and plays roles in adhesion and migration. EGF activates STAT3 signalling and regulates the levels of integrin !2, !6 and "1 (108). Alternatively, invasiveness may be acquired through promoting proteolytic degradation of the ECM; indeed, EGF increases the expression of uPAR (urokinase-type plasminogen activator receptor) (110) and MMP9 (84, 111). Furthermore, MMP9 is responsible for the EGF-induced loss of E-cadherin and disruption of adhesion junctions, leading to a migratory and invasive phenotype (84). In contrast, silencing of the EGFR suppresses integrin expression, MMP9 activity, cell adhesion and the invasion of ovarian cancer cells (106, 112). However, no studies on the role of AREG, EPI or BTC in ovarian cancer cell invasion have been published. Furthermore, blocking the EGFR suppresses in vivo ovarian cancer cell tumorigenicity in nude mice (106, 113). EGF also plays a role in regulating the sensitivity

14

of ovarian cancer cells to cisplatin in vitro (114), and dominant-negative EGFR can partially restore cisplatin sensitivity in drug-resistant ovarian cancer cells (113). 1.4.5 Clinical activities of EGFR-targeted therapies. Agents disrupting EGFR activities have been evaluated for their feasibility as ovarian cancer therapies, for examples, small molecule tyrosine kinase inhibitors (TKIs) targeting EGFR kinase domain and monoclonal antibodies (mAb) binding EGFR ECD. Generally, these therapeutics have very limited clinical activities in ovarian cancers (59). Gefitinib and erlotinib are EGFR TKIs which inhibit ovarian cancer cells and xenografts growth. No patients showed complete response or partial response in a phase II clinical trial of gefitinib (83) and only one objective response (4 % of cases) reported in another trial (74). In a phase II trial of erlotinib, 6% of patients demonstrated partial response and no complete response has been observed (59). Furthermore, trials have been performed using TKIs in combinations with cytotoxic chemotherapies like docetaxel and carboplatin (115, 116). However, none of these treatment regimes showed improvement in clinical activities (115, 116). On the contrary, significant clinical response to gefitinib and erlotinib have been observed in 10 % - 30 % NSCLC patients (117). In addition to TKIs, results from clinical trials of EGFR mAbs are also disappointing. These antibodies block ligand binding and receptor dimerization, they also reduce EGFR levels through promoting EGFR internalization and degradation (116). The first clinically tested EGFR mAb, cetuximab, showed partial response in 4 % of patients (118). In the phase II trial of another EGFR antibody, matuzumab, neither partial response nor complete response has been demonstrated (119).

15

1.5 Sprouty (SPRY) In the past decade, a novel family of cytoplasmic proteins called SPRY have been shown to play potent regulatory roles during Drosophila and mouse development (120122). Among these studies, the overexpression of SPRY has been shown to mimic the effects of EGFR signalling loss and prevent EGFR-induced phenotypic changes (121). Many additional studies have demonstrated that SPRY regulates the RAS/RAF/ERK pathway that is activated by various RTKs (122-125). 1.5.1 SPRY structure One Drosophila and four mammalian SPRY homologs have been identified with conserved structural characteristics, namely SPRY1, SPRY2, SPRY3 and SPRY4 (121, 122, 126). All SPRY members share a conserved C-terminal cysteine-rich region, with 44% 52% identity between Drosophila spry and mouse Spry1, 2 and 4 proteins (122). They also contain a conserved SPRY domain that mediates binding to RAF1 (RAF1binding domain, RBD) (127) (Fig. 1.3). The C-terminal domain is responsible for plasma membrane localisation (121, 128, 129). Deletion of the C-terminal domain abolishes membrane localisation and many of the biological functions of SPRY (121, 125, 128). In contrast, the N-terminal domains of the SPRY proteins are more divergent and are proposed to account for the substrate diversity and specificity of the individual SPRY members (122). However, the N-terminal domains of all SPRY proteins invariably contain a conserved tyrosine residue (Fig. 1.3). The tyrosine is phosphorylated upon growth factor stimulation and mediates membrane translocation. Most of the functions of

16

SPRY1 and SPRY2 (126, 130, 131) but not SPRY4 (126) are dependent on the phosphorylation of this tyrosine. 1.5.2 SPRY functions and mechanisms It is well known that downstream of various RTKs, SPRY proteins specifically inhibit the RAS/RAF/ERK pathway without affecting the p38, JNK or PI3K/AKT pathways (124, 125). However, there is still no consensus on how SPRY blocks ERK activation. Drosophila spry regulates ERK signalling at a point downstream of EGFR and upstream of RAS and RAF (121). In mammalian cells, by direct interaction with Grb2, SPRY1 and SPRY2 decrease RAS activation, RAS-RAF binding and RAF activation (124) (Fig. 1.3). Alternatively, downstream of RAS and independent of RAS activation, SPRY2 inhibits FGF-induced RAF activation (125). Interestingly, SPRY is only capable of binding to wild-type RAF, but oncogenic RAF is refractory to SPRY inhibition, leading to unchecked ERK activation in cell harbouring oncogenic RAF mutations (132). The evidence to date suggests the existence of multiple mechanisms that depend on the cellular context and/or the identity of the RTK (127). In contrast to initial observations (124, 125), accumulating evidence supports the mechanism that the PI3K/AKT cascade is also a target of SPRY (133, 134). To date, there has only been one study on the mechanism of AKT repression by SPRY, which demonstrated that SPRY2 increases the amount of PTEN and decreases its phosphorylation, leading to a decreased activation of AKT by EGF. In contrast to SPRY2, the understanding of the regulatory mechanisms of SPRY 4 is extremely limited and conflicting. SPRY4 was first reported to represses RTK-induced 17

ERK phosphorylation upstream of, or parallel to, RAS (135, 136). However, later studies found that mouse SPRY4 interferes with the VEGF-induced RAS-independent activation of RAF1 through a direct association with RAF1 (137). SPRY4 can also interact with TESK1 (testicular protein kinase 1) to inhibit growth factor-induced RAS/MAPK signalling (136, 138) (Fig. 1.3). Although SPRY proteins are generally considered to be inhibitory, SPRY2 may enhance EGF signalling in a cell type-specific manner (127). Cbl proteins are E3 ubiquitin ligases that recognise, ubiquitinate and target RTKs for degradation (139-141). Through direct interaction with the Cbl RING finger domain (142, 143), SPRY2 is capable of sequestering Cbl and thereby protects EGFR from degradation (144), resulting in higher EGFR levels and sustained activation (127) (Fig. 1.3). 1.5.3 Regulation of SPRY activity The feedback nature of SPRY arises from the prompt activation of SPRY by RTK activity. In addition to phosphorylation and membrane translocation as described above, the transcriptional regulation of SPRY is also important. During Drosophila embryonic development, SPRY level is higher in cells that are responsive to FGF and EGF (121, 123, 145, 146). Levels of SPRY2 and SPRY4 have been shown to be upregulated by activation of RAS/ERK signalling upon EGF and FGF stimulation in vitro (147). Similarly, in cancer cells, SPRY level is induced by growth factors (148) and oncogenic mutations downstream of RTK (149).

18

1.5.4 SPRY expressions in cancer Aside from a few exceptions (132, 150), the expression of SPRY has been found to be downregulated in cancer cells in most reports. SPRY2 and/or SPRY1 are significantly down-regulated in breast (151), prostate (152, 153), endometrial (154), and liver (155) cancers. The effects of the altered expression of different SPRY isoforms is likely cancer type-specific. The same group recently demonstrated a lower level of SPRY2, but not SPRY1, in hepatocellular carcinoma (155). Furthermore, the mechanisms of downregulation are also variable among different types of cancers. For example, in prostate cancer, SPRY2 is silenced through both epigenetic promoter hypermethylation and LOH (156). In contrast, SPRY2 loss in breast and liver cancer is independent of promoter methylation (151, 155). The prognostic value of SPRY has been demonstrated in renal (157), liver (158) and prostate (153) cancer patients. Moreover, SPRY4 mRNA level is as a reliable marker of the response of gastrointestinal stromal tumors to Gleevec treatment (150). 1.5.5 SPRY as tumor suppressor genes The tumor suppressor role of SPRY is supported by its negative effects on tumorigenesis. SPRY have been reported to inhibit proliferation (151, 155, 159, 160), migration and invasion (148, 159-161), cell cycle progression (148), in vitro and in vivo tumorigenesis (148, 151) and metastasis (162). 1.6 Hypoxia-inducible factor-1 alpha (HIF-1!) Hypoxia, a low oxygen tension condition, is a common phenomenon in solid tumors and is a driving force of tumor progression. Although severe or prolonged hypoxia 19

is toxic to cancer cells, it selects cancer cells that with adaptive and genetic changes that allow them to survive and proliferate (163). Most biological changes in response to hypoxia are mediated through the induction of the critical molecular mediators of hypoxia, the hypoxia-inducible factors (HIFs). Following stabilization and nuclear translocation in hypoxic conditions, HIF binds to hypoxia-response elements (HREs) to induce expression of numerous hypoxia-response genes (Fig. 1.4), thereby regulating multiple steps of tumorigenesis, including angiogenesis, metastasis and resistance to therapy (163). These processes contribute to the malignant phenotype, increase the rate of mutation and decrease overall patient survival (164). 1.6.1 HIF-1! structure The HIF family, including HIF-1, HIF-2 and HIF-3, belongs to the PAS (PERARNT (arylhydrocarbon receptor nuclear translocator)-SIM) family of basic helix-loophelix (bHLH) transcription factors. HIFs bind to DNA as heterodimers of a constitutively expressed HIF-1! subunit, also known as ARNT, and an oxygen-dependent " subunit (HIF-1", HIF-2" and HIF-3"). Structural analyses revealed that HIF-1" contains a bHLH domain (for dimerization and DNA binding), a PAS domain (for dimerization and target gene specificity), two transactivation domains (N-terminal, NTAD and C-terminal, CTAD) and an oxygen-dependent degradation (ODD) domain (Fig. 1.4). The ODD domain is required for the ubiquitinproteasome degradation pathway under aerobic conditions, and due to the absence of the ODD domain, HIF-1! is constitutively expressed in normoxia.

20

1.6.2 HIF-1! regulation The abilities of HIFs to act as the master regulator of cellular responses to hypoxia and in O2-homeostasis arises from the tight regulation of its activities by O2 availability. The activities of HIFs are largely controlled through the availability of the ! subunit, whose half-life is extremely short under normal oxygen tension (165). Therefore, the majority of investigations focus on the alpha subunits, and HIF-1! is the most extensively studied. In addition to tissue hypoxia, HIF-1! stabilization can be regulated by hormonal stimulation and genetic alterations under normoxic conditions. 1.6.2.1 Hypoxia Under well-oxygenated conditions, conserved proline residues (Pro402 and Pro564) within the ODD domain of HIF-1! are hydroxylated by prolyl hydroxylases (PHDs) (Fig. 1.4). Hydroxylated HIF-1! is recognised and ubiquitinated by Von HippelLindau (VHL), the substrate recognition component of an E3 ligase complex. Ubiquitinated-HIF-1! is then targeted to the proteasomal degradation pathway (166). As the enzymatic activities of PHDs require O2 as a cosubstrate, under hypoxia, PHDs are inactive, hydroxylation of HIF-1! is suppressed, and HIF-1! is stabilized (Fig. 1.4). PHDs are transcriptionally regulated by hypoxia in both HIF-dependent and HIFindependent pathways, which creates a negative feedback loop (165). Hypoxic induction of PHD2 and PHD3 is found in a wide range of cell types; however, induction is not observed in cells lacking HIF-1! or HIF-1" (167). Silencing HIF-1! reduces the induction of PHD2 and PHD3 by hypoxia (168). The HIF-independent pathway is mediated by SIAH1 and SIAH2, specific E3 ligases of PHD1 and PHD3 (169, 170), 21

whose expression is transcriptionally upregulated by hypoxia in a HIF-independent manner (171). Hypoxia-independent regulation of PHDs by cellular and hormonal factors such as oncogenic RAS and Src (see below), reactive oxygen species (ROS) (172), TGF! (173) and endothelin (174) serves as one of the mechanisms of HIF-1! regulation in normoxia. Stabilized HIF-1! translocates into the nucleus and then dimerizes with HIF1! to constitute HIF-1. HIF-1 heterodimers activate transcription by recruiting the transcriptional coactivators p300 and CREB binding protein (CBP). Interaction between HIF-1" and p300/CBP is regulated by factor inhibiting HIF-1 (FIH-1) in an oxygendependent manner. FIH, which belongs to the 2-oxoglutarate and Fe(II)-dependent dioxygenase superfamily, hydroxylates the conserved asparagine (Asn803) residue within the HIF-1! CTAD and prevents HIF-1!/p300 interaction (175). When hypoxia prevents asparagine hydroxylation by FIH, the CTAD is activated, interacts with p300/CBP and binds to HREs to regulate hypoxia-regulated gene expression. Therefore, in addition to HIF-1! stabilization, the complete activation of HIF transcriptional activity also requires CTAD activation.! Kung et al. showed that blocking the interaction of HIF-1! with p300/CBP attenuates hypoxia-inducible gene expression, reduces capillary formation and inhibits tumor growth in nude mouse xenografts (176). 1.6.2.2 Regulation of HIF-1" in normoxia 1.6.2.2.1 VHL mutations HIF-1" can also be activated in tumors under normoxic conditions through genetic alterations in the oxygen-signalling pathway. As mentioned earlier, the VHL 22

tumor suppressor gene product, pVHL, functions as the substrate recognition subunit of the ubiquitin E3 ligase complex that targets HIF-1! for degradation. This interaction is dependent on the hydroxylation of one or both of the two conserved prolines (Pro402 and Pro564) of HIF-1! by PHDs. Due to the O2-sensing function of PHDs, VHL loss-offunction results in HIF stabilization regardless of oxygen tension (177, 178). Inactivating germline VHL mutations predispose individuals to VHL diseases including renal clear cell carcinomas (RCCs) and haemangioblastomas (179, 180). Using immunohistochemistry, HIF-1! and HIF-2! proteins are found to be overexpressed in all tested RCCs and haemangioblastomas (181). Similarly, cells lacking the VHL wild-type protein constitutively express HIF-1! protein in normoxia (181, 182), which can be rescued by the reintroduction of a functional VHL (181). Accordingly, the mRNA of the hypoxia-regulated genes VEGF (181, 183) and GLUT1 (181, 184) are found to be higher in VHL-defective cells. Finally, expression of wild-type VHL in VHL-defective RCC cell lines prevents tumor formation in nude mice (181, 185). 1.6.2.2.2 AKT and PI3K Ample evidence suggests that PI3K/AKT signalling can also stabilize HIF-1! and induce HIF-1 activity (186-192). In hypoxic prostate cancer cells, inhibition of PI3K using pharmacologic inhibitors or siRNA knockdown decreases HIF-1! levels (188, 189) and VEGF promoter activity (190). The expression of HIF-1! and VEGF and the angiogenesis of prostate tumor xenografts are inhibited by dominant negative AKT mutants (189). HIF-1! stabilization by AKT does not require functional VHL (193).

23

Activation of the PI3K/AKT pathway can arise from amplification of the phosphatidylinositol 3-kinase catalytic subunit ! (PIK3CA) oncogene in human malignancies, including ovarian cancer (194). PIK3CA level is positively correlated with VEGF expression in ovarian cancer cell lines. Treatment with a PI3K inhibitor abrogates HIF-1! and VEGF expression (195). PIK3CA level has also been detected in most advanced-stage ovarian cancer biopsies and positively correlates with the VEGF level and the extent of microvascular development (195). Moreover, components of the PI3K/AKT signalling pathway, both upstream and downstream, have been shown to be involved in HIF-1! regulation, including PTEN, a molecular inhibitor of AKT. Glioblastoma cell lines lacking functional PTEN express high levels of VEGF even in normoxia, which is suppressed by restoration of wild-type PTEN (187). Overexpression of PTEN in prostate tumor cells reduces HIF-1! levels (190), VEGF expression, angiogenesis and tumor growth (189). Downstream of AKT, inhibition of mTOR (mammalian target of rapamycin) reverses the effect of AKT on HIF1! and target gene expressions (191, 196). AKT has been shown to stabilize HIF-1! without impairing prolyl hydroxylation and HIF-1! degradation, as revealed by antibodies specific for hydroxylated proline (197). However, alternative mechanisms for normoxic HIF-1! stabilization exist. For example, AKT positively regulates HIF-1! levels in glioblastoma cells through increased protein translation (190). In addition to PI3K and AKT, oncogenes such as RAS (198, 199) and Src (200) have been implicated in the normoxic activation of HIF-1!. Introduction of the 24

oncogenic mutant RASV12 and v-Src into cells increases the HIF-1! level and HIF-1 activity. Remarkably, RASV12 and v-Src abolish the hydroxylation of Pro564 in the ODD domain and hence inhibit HIF-1! degradation. In sharp contrast, cells transfected with constitutively active AKT upregulate HIF-1! without affecting hydroxylation. This finding indicates that these oncogenes may stabilize HIF-1! through hydroxylationdependent or hydroxylation-independent pathways (197). 1.6.2.2.3 Glycogen synthase kinase 3" (GSK3" ) The AKT target glycogen synthase kinase 3 (GSK3, exists in GSK3! or GSK3" isoforms) is a negative HIF-1! regulator. Inhibition of GSK3" using LiCl or GSK3" knockdown increases HIF-1! accumulation and the level of the HIF-1 target PAI1 in normoxia, whereas GSK3" overexpression reduces HIF-1! (201). Notably, GSK3" mediates the balance between HIF-1! stabilization/degradation as a function of the duration of hypoxia. Mottet et al. showed that short (5 hrs) incubations of prostate tumor cells in hypoxia lead to AKT activation and the inhibition of phosphorylation of GSK3" on Ser9, therefore reducing GSK3" activity and resulting in HIF-1! accumulation as well as increased HIF-1 transcriptional activity. In contrast, prolonged (16 hr) hypoxia inactivates AKT and hence activates GSK3", resulting in decreased HIF-1! protein levels (192). HIF-1! destabilization by GSK3" is via the NTAD is independent of HIF-1! hydroxylation and pVHL activity (201).

25

1.6.2.2.4 Heat shock protein 90 (HSP90) HSP90 interacts with HIF-1! (193, 202, 203) as well as HIF-2! and HIF-3! (204). These associations are mediated through the HIF-1! PAS domain (203, 205) and occur in both normoxia and hypoxia (193, 203). Although HSP90 is associated with HIF1! in the cytoplasm, HSP90 does not cotranslocate into the nucleus with HIF-1! (206). The mechanism underlying HIF-1! protection by HSP90 is not yet fully understood and is thought to occur via VHL-independent degradation (193, 204). In hypoxia, HSP90 is induced by (206), binds to and protects HIF-1!. Inhibition of HSP90 by geldanamycin results in loss of HIF-1! accumulation and HIF-1 activation in hypoxic VHL-defective cells (202, 206). The level of HSP90 is enhanced by AKT in normoxia. Inhibitors of AKT, but not MAPK inhibitors, reduce HSP90 expression (193). In VHL-defective RCC cells, inhibition of the AKT pathway promotes HIF-1! degradation as efficiently as the HSP90 inhibitor geldanamycin (193). As the PI3K/AKT pathway has no effect on HIF-1! hydroxylation, protection of HIF-1! by HSP90 increases HIF-1! stabilization by AKT, independent of proline hydroxylation, subsequent VHL ubiquitination and proteasomal degradation. In addition to AKT, HSP90 mediates the HIF-1! regulation of the kinase RACK1 (receptor of activated protein kinase C) (207). Moreover, carbon monoxide, an environmental factor implicated in angiogenesis (208), stabilizes HIF-1! and promotes VEGF expression through enhancing HIF-1!/HSP90 interaction (209).

26

1.6.2.2.5 Hormonal regulation Another common mechanism for the induction of HIF-1! in normoxia involves a large variety of growth factors and cytokines. EGF (188) and lysophosphatidic acid (210) induce HIF-1! through AKT activation to stimulate VEGF expression. IGF-II regulates cell survival through AKT and HIF-1! induction (211). TGF-" regulates the expression of HIF-1! and the HIF-1 targets PHD2 and PAI (173). Other hormones including IGF-I (187, 191, 212, 213), PGDF (186) and follicle-stimulating hormone (FSH) (214) have been reported to regulate the expression of HIF-1!. 1.6.3 HIF-1! expressions in cancer The expression of HIF-1! has been comprehensively investigated in panels of human normal and cancerous biopsies by immunohistochemistry (215, 216). Most normal tissues show no HIF-1! immunoreactivity. In contrast, overexpression of HIF-1! was detected in 50% 60% of malignant samples in one study, which represented most types of tumors examined, including breast, liver, lung, colon, prostate, ovarian, renal and pancreatic cancer. Furthermore, HIF-1! overexpression was found to be associated with metastasis and high-grade tumors. Two-thirds of the regional lymph node and bone metastases were HIF-1! positive. In addition, a high level of HIF-1! was more commonly found in breast metastases than in primary tumors. Among brain tumors of different grades, the strongest HIF-1! expression level was detected in the most malignant and vascularised tumors (215). Both borderline ovarian tumors and epithelial ovarian cancers were positive for HIF-1! expression (217, 218). HIF-1! has been shown to be a negative prognostic factor in most cancers (215, 219, 220) 27

1.6.4 HIF-1! functions in cancer 1.6.4.1 Angiogenesis When solid tumors expand, tumor hypoxia increases due to high metabolism and excessive oxygen consumption as well as the increased distance between tumor cells and local capillaries (221). To promote blood vessel formation and sustain growth in hypoxic microenvironments, tumors transition from a nonangiogenic to angiogenic phenotype, termed the angiogenic switch (222). As a prime physiological regulator of the angiogenic switch (223), a positive correlation has been found between HIF-1! overexpression and tumor vascular density (215) in brain and ovarian (217) tumor biopsies. During the angiogenic switch, hypoxia shifts the balance toward proangiogenic factors and reduces the expression of antiangiogenic factors such as thrombospondin-1 and thrombospondin-2 (224). To actively promote angiogenesis, HIF activates the expression of various angiogenic proteins including growth factors, cytokines, and a number of small molecules, including the key angiogenic factor VEGF. VEGFR1 and VEGFR2 on endothelial cells mediate the mitogenic effects of VEGF on endothelial cells. Therefore, VEGF has strong angiogenic activity, and treatment with a VEGFR inhibitor results in a 60% reduction in tumor vasculature in mouse models (225). The VEGF promoter contains HRE binding sites and is a direct transcriptional target of HIF-1 (226). Accordingly, xenografts of HIF-1!-deficient hepatoma cells exhibit reduced levels of hypoxia-induced VEGF mRNA and reduced tumor vascularisation and proliferation (227). HIF-1! is also expressed in endothelial cells and mediates autocrine signalling of VEGF/VEGFR2 in endothelial cell proliferation and blood vessel formation 28

(228). More importantly, human tumors with high HIF-1! and VEGF expression levels are correlated with more aggressive and malignant phenotypes (229). In addition to the hypoxia-mediated activation of HIF, the AKT/HIF-1! axis also mediates VEGF level and angiogenesis stimulated by growth factors (188, 191) and other hormones (210, 214) under normoxia. 1.6.4.2 Metastasis Metastasis is the primary cause of cancer-related deaths. It is a multistep process: from the initial epithelial-mesenchymal transition (EMT), dissemination, and homing to the final organotropic colonisation. The role of hypoxia and HIF in tumor metastasis is supported by the observation that HIF-1! expression is upregulated or more commonly found in metastases and high-grade tumors (215). Mechanistically, hypoxia and HIF are potent regulators of many key factors that determine tumor cell invasiveness and metastatic potential, namely E-cadherin, lysyl oxidase (LOX), chemokine receptor CXCR4 (CXCR4), stromal-derived factor 1 (SDF-1) and plasminogen activator inhibitor1 (PAI-1). HIF-1 promotes metastasis through repressing E-cadherin, a major cell adhesion molecule that maintains epithelial integrity. Loss of E-cadherin is a hallmark of and prerequisite for metastasis. Reduced E-cadherin expression is associated with increased metastasis in patients with ovarian (230), breast (231), prostate (232) and lung cancers (233). The antimetastatic role of E-cadherin is supported by the finding that restoration of E-cadherin in cancer cells inhibits metastasis (234). Repressed E-cadherin levels have been found to be correlated with HIF-1! expression in ovarian (235) and RCC carcinoma biopsies (236). In hypoxic ovarian cancer cells and VHL-defective RCC cells, HIF-1! is 29

overexpressed and EMT and cell invasion are increased in vitro (235, 236). E-cadherin in these cells is repressed by HIF-1! through inducing the E-cadherin-specific repressors Snail or ZEB1 (235, 236). HIF also promotes metastasis through modulating the extracellular matrix. LOX is an enzyme that is involved in extracellular matrix formation. HIF-1 is a potent inducer of LOX, and secreted LOX contributes to the invasive properties of hypoxic human cancer cells. The importance of LOX in hypoxia-induced metastasis is directly underscored by the finding that inhibition of LOX is sufficient to attenuate hypoxia-induced metastasis in mice. In patients with breast and head and neck tumors, a high LOX expression level is correlated with hypoxia and poor distant metastasis-free and overall survival (237). CXCR4 is the most common chemokine receptor that is expressed in tumor cells. It is a direct HIF target in hypoxic lung, ovarian, breast and renal cell carcinoma cells (238). Its ligand, SDF1, is highly expressed in common sites of metastasis, including the lung, bone marrow, and liver (220). Interactions between CXCR4 and its ligand SDF-1 mediate the metastatic homing of tumor cells in response to hypoxia (220, 238) Another invasion/metastasis related gene PAI-1 is a direct target of HIF-1!. PAI1 levels are enhanced through PI3K/AKT and ERK1/2 via HIF-1! in prostate and gastric cancer cells (212, 239). 1.7 Hypothesis and objectives Loss of expression or activity of SPRY has been demonstrated in various human malignancies. Nevertheless, no studies on SPRY in ovarian cancer have been reported. Given the implications of various RTKs in ovarian tumorigenesis, I tested the hypothesis that SPRY acts as a tumor suppressor in ovarian cancer. The main objectives of this thesis 30

are to investigate the expression and functions of SPRY in ovarian cancer, specifically in the context of EGFR and HIF-1! signalling. Objective 1. In Chapter 2, I investigated whether the SPRY mRNA is deregulated and examined the function of SPRY2 in ovarian cancer. 1) The mRNA levels of SPRY members in ovarian cancer tissues and cell lines was investigated. 2) The incidence of SPRY2 gene deletion was evaluated. 3) The role of SPRY2 in EGF-induced cell invasion was tested. Objective 2. In Chapter 3, I tested the roles of an alternative EGFR ligand (AREG) and regulator (SPRY4) in ovarian cancer invasion. 1) The effect of AREG on invasion was tested. 2) The underlying molecular pathways were elucidated. 3) The effect of AREG on SPRY4 level was examined. 4) The effect of SPRY4 on AREG-induced invasion was investigated. Objective 3. In Chapter 4, I explored the regulation of SPRY4 by EGF and the effects of SPRY4 on EGF signalling. 1) The role of AKT/HIF-1! in EGF-induced SPRY4 level was tested. 2) The feedback of SPRY4 on the EGF-induced level and activity of HIF-1! was evaluated. 31

3) The role of AKT in SPRY4-mediated HIF-1! regulation was investigated. Objective 4. In Chapter 4, I extended the investigation to the underlying mechanisms of SPRY4 inhibition of HIF-1! level. 1) The inhibition of HIF-1! level by SPRY4 was confirmed. 2) The level of regulation of HIF-1! by SPRY4 was determined. 3) The roles of PHD in the SPRY4 regulation of HIF-1! level was investigated.

32

Figure 1.1 The chart illustrates the pathogenesis of epithelial ovarian cancers. Low-grade serous carcinomas arise from serous borderline tumors containing activating KRAS, BRAF or ERBB2 mutations. Rarely, low-grade serous tumors progress to high-grade tumors. High-grade serous carcinomas develop from the Fallopian tube or ovarian surface epithelium through TP53 and BRCA mutations. KRAS mutations and ERBB2 gene amplification are frequently found in mucinous tumors. Clear cell carcinomas contain PI3KCA mutations or loss of PTEN. PTEN and CTNNB1 mutations may lead to the formation of endometrioid carcinomas. Alternatively, endometrioid carcinomas may contain TP53 mutations. Endometrioid carcinomas and clear cell carcinomas may also arise from endometriosis with ARID1A mutations. Modified from Lalwani, N et al, 2011 (1).

33

Figure 1.2 The structure and signaling of EGFR. EGFR consists of an extracellular domain (ECD), a transmembrane domain (TMD) and an intracellular domain (ICD). The ICD can be further divided into a juxtamembrane domain (JD), a tyrosine kinase domain (TKD) and a regulatory domain (RD). Upon the binding of EGFR ligands, such as EGF, transforming growth factor ! (TGF!) or amphiregulin (AREG), the TKD is phosphorylated and thereby activated; the RD is then transactivated. The phosphorylated RD recruits adaptor proteins, including Shc, Grb2 and SOS, and triggers the RAS/RAF/MEK/ERK cascade. Activated EGFR also stimulates other MAPK pathways (p38 and JNK) and the PI3K/AKT pathway.

34

Figure 1.3 A schematic diagram of human SPRY depicting its structure, its interacting partners and the corresponding functional consequences. The C-termini of SPRY proteins are more conserved and contain a RAF1-binding domain (RBD), which mediates the interaction with RAF and leads to the inhibition of ERK activation. ERK inhibition is also achieved through SPRY interaction with Testicular protein kinase 1 (TESK1). SPRY also interfere with the RAS/RAF/MEK/ERK pathway through interaction with Grb2. The N-termini of SPRY are more divergent but invariably contain a conserved tyrosine residue. Tyrosine-phosphorylated SPRY binds with Cbl, which inhibits EGFR degradation and increases EGFR signaling. Interaction between SPRY and Seven in absentia homolog 2 (SIAH2) promotes the ubiquitination of SPRY and target SPRY for proteasomal degradation. Modified from Edwin, F et al, 2009 (240).

35

Figure 1.4 The diagram illustrates the regulation of HIF-1!. HIF-1! is mainly regulated post-transcriptionally. Under normoxia, PHD proteins hydroxylate the two proline residues within the HIF-1! oxygen-dependent domain (ODD) domain. Hydroxylated HIF-1! is prone to Von Hippel-Lindau (VHL)-mediated degradation. When PHDs are degraded by Seven in absentia homolog 2 (SIAH2) or inactivated by hypoxia, growth factors (GFs), reactive oxygen species (ROS) or oncogenic RAS, HIF-1! degradation is prevented. In contrast, the PI3K/AKT/glycogen synthase kinase-3" (GSK3") pathway regulates HIF-1! independently of PHD activity. When HIF-1! is stabilized, it would translocate into the nucleus and dimerize with HIF-1". With the recruitment of the p300 co-factor, the complex recognizes hypoxia-responsive elements (HRE) in the promoters of target genes, which are implicated in processes such as angiogenesis, invasion, proliferation and apoptosis.

36

Chapter 2 Genetic inactivation of Sprouty2 promotes epidermal growth factor-induced Ecadherin downregulation and invasion in ovarian cancer cells 2.1 Introduction Ovarian cancer is the fifth leading cause of all female cancer-related deaths and the most lethal gynaecologic malignancy in North America. Approximately 60% of women who develop ovarian cancer will die from the disease (2, 241). Although epithelial ovarian cancer (EOC) includes a majority of all ovarian carcinomas, its origin and aetiology have not yet been completely elucidated. Current evidence suggests that malfunction of receptor tyrosine kinases (RTKs), including epidermal growth factor receptor (EGFR), contributes to the development of EOC (52). EGFR and its ligands play a critical role in cell proliferation, survival, and tumor metastasis (53, 242). Moreover, increased expression levels of EGFR-specific ligands have been identified in ovarian cancer cells (89) and ascitic fluid (90), and EGFR expression levels were found to be elevated in advanced stage disease and metastases (53). In addition to RTK abnormalities, the loss of endogenous regulators represents an alternative mechanism that leads to aberrant RTK activity. During the past decade, a novel family of cytoplasmic proteins, Sprouty (SPRY), has been identified as feedback inhibitors of the RAS/MAPK/ERK signalling cascade triggered by RTKs. Consistent with their inhibitory involvement in the RTK-stimulated ERK pathway, SPRY is a putative tumor suppressor, and cells lacking either SPRY expression or function may be hypersensitive to mitogenic and metastatic signals. SPRY isoforms have been found to be downregulated in most malignancies studied, including breast (151), prostate (152, 153, 37

156, 161), liver (155) and lung (160) cancers. Functionally, SPRY has been reported to interfere in steps of tumorigenesis, including proliferation (151, 155, 160), migration (148, 160, 161), invasion (148), and cell cycle (148) as well as in vitro (148) and in vivo (151) tumorigenic potential and formation of metastases (162). The prognostic value associated with SPRY levels have been established in renal (157), liver (158) and prostate (153) cancer patients. Moreover, SPRY4 mRNA level serves as a reliable response marker to Gleevec treatment for patients with gastrointestinal stromal tumors (150). However, the mechanism by which SPRY is silenced appears to vary depending on the cancer type (151-153, 155, 156, 160, 161), and contradicting data exist for similar malignancies (153, 155, 156, 158). The present study aimed to investigate, in addition to any SPRY downregulation, the underlying mechanisms and functions of SPRY in ovarian cancer. We demonstrate that SPRY2 mRNA is downregulated in both biopsies and permanent cell lines derived from EOC. Furthermore SPRY2 gene is deleted in some of the ovarian tumors and is possibly a cause of the reduced SPRY2 mRNA. We further demonstrate a positive correlation between SPRY2 and E-cadherin. Furthermore, the re-introduction of SPRY2 in ovarian cancer cells partially restored E-cadherin expression levels suppressed by EGF and antagonized the effect of EGF-induced invasion.

38

2.2 Materials and methods 2.2.1 The Human Exonic Evidence-Based Oligonucleotide microarray (HEEBO) The HEEBO microarray (Stanford, CA, USA) employed for the study included 44,544 70-mer probes that were designed using a transcriptome-based annotation of exonic structure for genomic loci. Pooled RNA from 10 human cancer cell lines of different origins (Stratagene, Universal Human Reference RNA, Cat 740000) for broad gene coverage on the array was included as a reference. We examined the mRNA expression profiles of a series of ovarian tumors from the Vancouver General Hospital tumor bank obtained from patients who were undergoing surgery during 2004 and 2005. These cases included the following subtypes: high-grade serous (n = 35), low-grade serous (n = 2), endometrioid (n = 7), clear cell (n = 3), serous borderline (n = 1), endometrioid borderline tumor (n = 1) and normal Fallopian tube (n = 1). Approval for the study was obtained from the University of British Columbia Research Ethics Board (#H04-60102), and written informed consent was obtained from all participants involved in the study. HEEBO was performed as described previously (243). 2.2.2 Molecular inversion probe (MIP) copy number analysis To determine whether deletion of the SPRY2 and SPRY4 genes occurs in human ovarian tumors, samples from another cohort of patients were utilised to perform MIP copy number analysis. All women undergoing primary debulking surgery at the Vancouver General Hospital and British Columbia Cancer Agency in Vancouver, Canada, between January 2004 and September 2005 were invited, except those with mucinous and borderline tumors or who had received pre-operative chemotherapy. The 39

pathology data were reviewed by a pathologist (CBG). The classification and grading of tumors were performed as described previously (244). We included 28 high-grade serous, 5 high-grade serous/undifferentiated, 3 high-grade undifferentiated, 5 endometrioid, 4 clear cell and 1 low-grade serous tumors. Ethical approval was obtained from the University of British Columbia Research Ethics Board (#H02-61375 and #H03-70606), and written informed consent was obtained from all participants involved in the study. The MIP copy number assay and copy number estimation were performed as described previously (245). Copy numbers over 3.0 were considered amplification events and copy numbers below 1.5 were considered deletion events. 2.2.3 The Cancer Genome Atlas (TCGA) To obtain direct evidence to support a dependency of reduced SPRY2 mRNA level on gene deletion, we retrieved data from the TCGA database portal (http://cancergenome.nih.gov/) in September 2011. Copy number data for 585 ovarian serous cystadenocarcinomas and 587 normal samples (569 matched and 18 unmatched), as well as data for the gene expression profile of 584 tumors and 18 unmatched normal samples, were extracted. 2.2.4 Cell culture and reagents Four non-tumorigenic SV40 Tag-immortalised OSE-derived lines, IOSE-29, IOSE80, IOSE-120, and IOSE-398, were generous gifts from Dr. Nelly Auersperg (University of British Columbia) (246). The human ovarian adenocarcinoma cell line, BG-1, was kindly provided by Dr. K.S. Korach (National Institute of Environmental Health Sciences, NIH, Research Triangle Park, NC) (247). CaOV3, OVCAR3 and SKOV3 ovarian cancer 40

cell lines were obtained from American Type Culture Collection (Manassas, VA). The cell lines were cultured in MCDB 105 / M199 (1:1), supplemented with 10% heatinactivated fetal bovine serum (FBS), 100 IU/ml penicillin and 100 g/ml streptomycin. The cells were cultured at 37C and 5% CO2. Fetal bovine serum was purchased from Hyclone Laboratories (Logan, UT). Human recombinant EGF and other tissue culture materials were obtained from Sigma Chemical Co. (St. Louis, MO) unless otherwise stated. Human SPRY2 overexpression constructs (FLAG-SPRY2) and the empty pXJ40FLAG vector, which were transfected using Lipofectamine 2000 (Invitrogen, Carlsbad, CA), were gifts from Dr. Graeme R. Guy (the Institute of Molecular and Cell Biology, Singapore) (143). 2.2.5 Real-time PCR Real-time PCR was performed using an ABI 7300 real-time thermal cycler (ABI, Hercules, CA). SPRY2, SPRY4 and the internal control, GAPDH, were amplified in duplicate with the following PCR primers: SPRY2, forward 5-

CCCCTCTGTCCAGATCCATA-3 and reverse 5-CCCAAATCTTCCTTGCTCAG-3; SPRY4, forward 5-AGCCTGTATTGAGCGGTTTG-3 and GAPDH, and reverse forward 55-

GGTCAATGGGTAGGATGGTG-3,

GAGTCAACGGATTTGGTCGT-3 and reverse 5-GACAAGCTTCCCGTTCTCAG-3. 2.2.6 Antibodies Specific antibodies were used to detect proteins via Western blot analysis: the anti-E-cadherin antibody was obtained from BD Transduction Laboratories (Lexington,

41

KY), the anti-SPRY2 antibody was purchased from Sigma and the anti-!-actin antibody was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). 2.2.7 Invasion assay Twenty-four-well transwell filters with an 8-"m pore coated with 1 mg/ml Matrigel (50 "l/well; BD Sciences, Mississauga, ON, Canada) were used to assess cell invasion. SKOV3 cells transfected with either control or SPRY2-overexpression constructs were trypsinized and resuspended in 0.1% FBS medium, with or without EGF, and then seeded in triplicate in the upper chamber. Next, 1% FBS medium was added to the lower wells. The chambers were incubated for 24 h at 37C in a 5% CO2 atmosphere. Cells that did not penetrate the filter were removed, and invaded cells on the lower surface of the filter were fixed with ice-cold methanol, stained with Hoechst 33258, and counted by epifluorescence microscopy using Northern Eclipse 6.0 software (Empix Imaging, Mississauga, ON, Canada). Triplicate inserts were used for each individual experiment, and the results are presented as the mean values. 2.2.8 Statistical analysis Statistical analysis was performed using Prism Graphing software. Differential variations in SPRY mRNA levels among ovarian tumor subtypes were assessed using the Kruskal-Wallis rank test followed by Students t test comparing each tumor subtype. The relative quantification of mRNA expression levels, as assessed by real-time PCR, was calculated using the 2## Ct method. For the invasion assay and the comparison of Ecadherin expression levels with the control (using SPRY2-overexpressing cells), a oneway ANOVA and nonparametric Column Analysis was performed followed by Tukeys 42

Multiple Comparison Test to compare all pairs of columns. Columns are not denoted by the same letter are statistically different (P < 0.05). Data are presented as the mean SD of three or four independent experiments. The Pearson correlation coefficient (r) and associated probability (P) were calculated when comparing E-cadherin and SPRY2 protein levels using the Spearman nonparametric correlation.

2.3 Results 2.3.1 Levels of SPRY mRNA in ovarian tumors of different pathological subtypes To examine whether SPRY mRNA were downregulated in ovarian tumors, as is reported for other malignancies, we compared EOC samples of various histopathological types, including serous (high-grade, low-grade or borderline), endometrioid (carcinoma or borderline tumor), clear cell and normal Fallopian tube. SPRY1, SPRY2 and SPRY4 mRNA were included in our array. Significant differences in the SPRY2 mRNA levels were observed in various tumors types (Kruskal-Wallis test, P = 0.0091, Fig. 2.1). The lowest mean level of SPRY2 mRNA was observed in the clear cell sample followed by the high-grade serous carcinomas. All other samples displayed higher mean SPRY2 mRNA levels compared with the reference level. To further assess the variations in SPRY2 mRNA levels among the categories, we extended our study to perform comparisons between each subtype using Student's t test. The difference in SPRY2 mRNA levels between high-grade and low-grade serous was statistically significant (P = 0.022, Table 2.1). When comparing pathological subtypes, the mean SPRY2 mRNA levels in serous and clear cell carcinomas were statistically lower than that of 43

endometrioid carcinomas (serous vs. endometrioid, P = 0.0007; clear cell vs. endometrioid, P = 0.022). SPRY4 mRNA levels were not significantly different between the various subtypes (P = 0.33, data not shown); however, the mean expression level in high-grade serous was lower than that in the reference and was the lowest among all the specimens. We utilised three oligonucleotides in the array that corresponded to different SPRY1 transcripts, for which the expression levels were similar for each subtype and therefore did not display significant differences (P = 0.396, 0.706 and 0.813, data not shown). 2.3.2 Levels of SPRY mRNA in immortalised ovarian surface epithelium (IOSE) and EOC-derived cell lines Next, we extended our expression analysis to include ovarian cancer cells cultured in vitro. To enhance data reliability, 4 IOSE cell lines established from individual patients were included as references for the comparison of SPRY mRNA levels in ovarian cancer cell lines. Using real-time PCR, we confirmed the observation of reduced SPRY2 mRNA (Fig. 2.2A) expression in most (3 out of 4) ovarian cancer cell lines tested (BG-1, OVCAR3 and SKOV3). In addition, the SPRY4 mRNA levels observed in all 4 cancer cell lines were consistently lower than those of the IOSE cell lines (Fig. 2.2B). 2.3.3 A deletion event in the proximity of the SPRY2 locus MIP copy number analysis was performed to examine whether the reduced SPRY levels observed were due to chromosomal changes of the SPRY genes. As there are no specific intragenic probes available, we used probes closely flanking the SPRY2 and SPRY4 loci as surrogate markers. For SPRY2, two markers tightly flanking the locus (0.2 44

Mb proximal and 0.5 Mb distal) exhibited a 23.4% (11/46) and 19.6% (9/46) loss, respectively, in 46 tumors. Notably, with one exception in undifferentiated high-grade tumor, nearly all of the deletion events observed were identified in high-grade serous carcinomas (Table 2.2); none of the endometrioid, clear cell, or low-grade serous tumors exhibited loss of the SPRY2 gene. In high-grade serous samples, the frequencies of SPRY2 loss were found to be 32.1% (9/28) to 35.7% (10/28). Two additional markers (0.9 Mb proximal and 1.5 Mb distal) showed a similar frequency of loss (25% and 35.7%, respectively) (Fig. 2.3). In contrast, deletion of the markers (0.4 Mb proximal and 90 kb distal) closest to the SPRY4 locus was rarely found, with only 3.6% (1/28) and 7.1% (2/28) of high-grade serous tumors, respectively. Two markers more distant from the SPRY4 gene exhibited only two cases of loss (7.1%) (Fig. 2.3). 2.3.4 SPRY2 deletion may lead to reduced SPRY2 mRNA level In a TCGA data set, at a percentage lower than our MIP study, 24% of ovarian serous cystadenocarcinoma samples displayed a decrease in gene copy number. A majority (67%) of the samples showed decreased SPRY2 mRNA level (log2 tumor/normal ratio < -0.5), therefore, suggesting that SPRY2 gene deletion was responsible for the reduced SPRY2 mRNA level in the samples. According to the gene expression analysis, decreased SPRY2 mRNA level was detected in 54 of the tumors.

However, when we analysed gene copy number in these tumors, only 32% showed a decrease in copy number, thereby indicating that additional mechanisms may contribute to the reduced SPRY2 mRNA level.

45

2.3.5 SPRY2 reversed EGF-suppressed E-cadherin protein expression and antagonized EGF-induced cell invasion In our previous report (248), the SKOV3 cell line was shown to respond significantly to EGF treatment and showed a significant decrease in E-cadherin expression levels. We used this cell model to investigate the effect of SPRY on EGFinduced reduction in E-cadherin levels. We transiently overexpressed SPRY2 in SKOV3 cells, and gene overexpression was subsequently confirmed using antibodies specifically recognising SPRY2 protein (Fig. 2.4A). SPRY2 overxpression has no effect on cell morphology (data not shown) and basal E-cadherin (Fig. 2.4B). After EGF treatment Ecadherin levels were significantly suppressed and the decrease was partially reversed by SPRY2, which results in higher E-cadherin levels in SPRY2-overexpressing cells compared with cells transfected with empty vector (Fig. 2.4B). Together, these results support an antagonizing effect of SPRY2 on EGF activity of repressing E-cadherin protein. The negative effect of SPRY2 on EGF regulation of E-cadherin prompted us to evaluate the effect of SPRY2 on EGF-stimulated cell invasion using the transwell invasion assay. In correlation with the effect on E-cadherin protein, SPRY2 counteracted the effect of EGF on invasion and SPRY2-overexpressing cells showed reduced invasiveness under EGF stimulation, whereas SPRY2 had no effect on basal invasion (Fig 2.4C, D).

46

2.3.6 SPRY2 and E-cadherin proteins displayed a positive correlation in human ovarian cancer cell lines and tumors To test whether the positive effect of SPRY2 on E-cadherin level is reflected in the endogenous expression levels, we analysed the SPRY2 and E-cadherin expression levels in three ovarian cancer cell lines and a panel of eleven high-grade serous ovarian tumors isolated from patients. In most of the samples, we found parallel SPRY2 and Ecadherin expression levels (Fig. 2.5A). This finding indicated a positive correlation between the expression levels of the two proteins. Moreover, the correlation was statistically significant (Pearson correlation coefficient, r = 0.5825 and P = 0.0288) in a Spearman nonparametric correlation analysis (Fig. 2.5B).

2.4 Discussion SPRY proteins have been identified as endogenous inhibitors of the

RAS/MAPK/ERK pathway downstream from RTKs. Aberrant activity of various RTKs, especially EGFR, plays an essential role in malignancy development, including ovarian cancer. Therefore, we aimed to examine the expression profile and function of SPRY in ovarian cancer, which remains largely unknown. First, we demonstrated a reduction in SPRY2 mRNA level in ovarian cancer cell lines and clinical samples (Fig. 2.1, 2.2A). It is important to note that both the significant difference observed between the SPRY2 mRNA levels in high-grade serous and endometrioid samples in the HEEBO analysis and the observation that most (12 out of 13) SPRY2 losses were found in high-grade serous carcinomas suggests the involvement of SPRY2 in serous EOC pathogenesis. These 47

results also demonstrate that SPRY2 represents a potential molecular marker for the identification of serous carcinomas, which includes a majority (approximately 80%) of all EOCs (3). We next evaluated the occurrence of the chromosomal deletion of the SPRY2 gene. The SPRY2 locus has been mapped to 13q31.1 (156). Previous cytogenetic studies have shown that chromosomal loss of either 13 or 13q is a frequent event in both hereditary (249, 250) and sporadic ovarian carcinomas (251, 252). This frequency can be explained by the existence of known tumor-suppressor genes on 13q including BRCA2 (13q12-13), RB (13q14) and protocadherin 9 (13q21-2). The data presented here support these studies by demonstrating the chromosomal deletion of the SPRY2 locus, which occurred in a moderate proportion of cases (Fig. 2.3). Furthermore, a minor proportion (16%) of serous cystadenocarcinomas from the TCGA database showed both gene deletion and reduced expression, thereby suggesting that genetic aberration is an underlying cause of SPRY2 mRNA downregulation in ovarian cancer. In contrast to the SPRY2 results, deletion of the markers flanking the SPRY4 gene (90 kb to 0.7 Mb) occurred in only 2 cases (Table 2.2). The deletion of the SPRY4 gene is a rare event in ovarian cancer and unlikely an explanation for the reduced SPRY4 mRNA levels observed in cell lines. According to TCGA gene expression analysis, only 32% of tumors with reduced SPRY2 mRNA level exhibited a gene deletion, thereby indicating the requirement of additional silencing mechanisms. The presence of CpG islands in the SPRY2 5 regulatory regions suggests the potential involvement of epigenetic mechanisms in the regulation of SPRY2 expression. Reports on the methylation status of the SPRY2 promoter in liver and prostate cancer samples have been contradictory and inconclusive 48

(153, 155, 156, 158). Our preliminary experiment demonstrated a robust reactivation of SPRY4 mRNA level in BG-1, CaOV3 and OVCAR3 cells treated with the demethylating agent, 5-aza-2-deoxycytidine (data unpublished). However, only a mild induction of SPRY2 mRNA level was observed in 2 out of 3 cell lines (data not shown), thereby suggesting that promoter hypermethylation may play a minor role in the silencing of SPRY2 expression in this type of malignancy. Reduced E-cadherin expression has been found to be associated with advanced stage disease, poor differentiation, metastasis (230) and serous subtype tumors (253). Many reports, including our recent report on EGF-repressed E-cadherin expression via the induction of Snail and Slug in SKOV3 cells (248), have shown that E-cadherin is silenced at a transcriptional level, either through epigenetic promoter hypermethylation (230, 254) or repression by transcriptional repressor induction (255, 256). The present study provides evidence that SPRY2 positively regulates and correlates with E-cadherin expression, thereby suggesting the possibility that loss of SPRY2 might represent an additional mechanism whereby ovarian cancer cells lose E-cadherin and gain malignant properties. Furthermore, among various ovarian tumor subtypes, high-grade serous tumors exhibited SPRY2 gene deletion, reduced SPRY2 mRNA level and a correlation between SPRY2 and E-cadherin expression levels, thereby implying that SPRY2 may play a role in the tumorigenesis of high-grade serous tumors. In summary, our findings demonstrate that the SPRY2 level is decreased in ovarian cancers due to a chromosomal deletion, and the reintroduction of SPRY2 diminishes EGF-induced cell invasiveness by restoring the EGF-repressed intercellular adhesion molecule E-cadherin protein. Therefore, genetic downregulation of SPRY2 may lead to a 49

decrease in E-cadherin expression, which promotes ovarian cancer progression. Further clarifying the mechanisms underlying SPRY function and regulation will not only advance our understanding of ovarian cancer progression but also facilitate the development of novel therapeutic strategies for the treatment of ovarian cancer.

50

Table 2.1 Comparison between mean SPRY2 mRNA levels of various histopathological types by Students t test Mean Serous borderline Low-grade serous Endometrioid borderline Endometrioid Normal Fallopian tube High-grade serous Clear cell A A A A A B B B B B 1.83 1.75 1.54 1.47 1.46 -0.50 -0.66

Means not denoted by the same letter are significantly different

The HEEBO microarray was performed to examined the SPRY2 mRNA expression profiles of a series of ovarian tumors: high-grade serous (n = 35), low-grade serous (n = 2), endometrioid (n = 7), clear cell (n = 3), serous borderline (n = 1), endometrioid borderline tumor (n = 1) and normal Fallopian tube (n = 1). Student's t test was performed to assess the variations in SPRY2 mRNA levels between each pair of subtypes. Means not denoted by the same letter are significantly different.

51

Table 2.2 MIP analysis of loss of the markers flanking the SPRY2 and SPRY4 loci in ovarian tumors SPRY2
number histopathology 223 329 293 283 239 327 379 163 305 212 330 332 388 363 344 345 384 178 229 309 394 195 236 172 254 serous HG serous HG serous HG serous HG serous HG serous HG serous HG serous HG serous HG serous HG serous HG serous HG serous HG serous HG serous HG serous HG serous HG serous HG serous HG serous HG serous HG serous HG serous HG serous HG serous HG proximal 2 proximal 1 distal 1 1.40 1.37 1.94 1.44 2.41 2.26 1.58 1.46 2.09 1.54 2.34 2.08 2.97 1.57 2.06 1.29 1.93 2.04 1.35 2.36 1.96 1.75 1.42 2.75 1.68 1.40 1.20 1.88 1.09 2.24 1.98 1.58 1.43 2.09 2.76 2.17 1.69 2.97 1.36 1.57 1.28 1.82 2.02 1.35 2.36 1.83 1.24 1.42 2.22 1.68 1.58 1.26 1.88 1.30 2.28 2.49 1.41 1.44 1.94 1.54 2.57 1.30 3.01 1.36 2.26 1.18 1.97 2.02 1.53 2.45 1.83 1.20 1.53 2.13 1.75 distal 2 1.58 1.26 1.88 1.30 2.28 2.13 1.36 1.44 1.94 2.12 2.56 1.30 2.56 1.32 1.74 1.18 1.74 1.57 1.46 2.21 1.82 1.20 1.53 1.99 1.84

SPRY4
proximal 2.09 1.44 2.10 2.43 2.42 2.31 2.09 1.77 2.61 1.83 1.56 2.12 2.05 2.21 2.73 2.47 2.71 2.67 2.26 2.23 2.22 2.39 1.59 1.62 2.05 distal 1.92 1.33 2.22 2.13 2.31 2.07 2.09 1.99 2.32 2.36 1.41 1.77 1.94 2.27 2.44 2.91 2.36 2.35 2.10 2.23 2.09 2.39 1.90 1.55 2.05

52

SPRY2
number histopathology 319 372 297 208 273 240 161 336 186 201 280 198 213 219 392 242 281 334 156 343 324 serous HG serous HG serous HG undifferentiated HG undifferentiated HG undifferentiated HG proximal 2 proximal 1 distal 1 2.51 1.50 2.03 3.03 1.67 2.27 2.16 1.36 2.00 2.53 1.48 2.10 1.72 2.27 2.09 2.23 2.57 2.01 2.14 2.50 1.81 2.02 2.41 1.94 1.93 2.03 1.97 2.30 1.36 2.09 2.72 1.54 2.27 1.67 1.83 2.09 1.81 2.54 2.23 1.86 2.61 2.04 2.37 2.45 2.32 1.90 2.03 2.00 distal 2 2.30 1.30 1.86 2.72 1.54 2.30 1.67 1.83 2.36 1.81 2.54 1.88 1.86 2.61 1.64 1.91 2.45 2.23 1.90 2.03 2.00

SPRY4
proximal 2.15 2.35 2.27 1.81 1.94 2.10 1.87 1.87 2.60 3.24 1.89 3.04 2.01 2.26 2.54 2.58 2.23 2.35 2.18 2.26 1.92 distal 2.21 2.41 2.19 1.81 1.66 2.02 1.82 1.87 2.72 3.24 2.25 2.42 1.98 2.07 2.31 2.14 2.23 1.92 1.88 2.33 2.13

serous/undifferentiated HG 1.87 serous/undifferentiated HG 1.88 serous/undifferentiated HG 2.44 serous/undifferentiated HG 2.23 serous/undifferentiated HG 2.62 clear cell clear cell clear cell clear cell endometrioid G2 endometrioid G2 endometrioid G1 endometrioid G2 endometrioid G2 serous LG 2.18 2.14 2.48 2.04 2.37 2.23 2.02 1.93 2.46 2.00

MIP analysis were performed on 28 high-grade serous, 5 high-grade serous/undifferentiated, 3 high-grade undifferentiated, 5 endometrioid, 4 clear cell and 1 low-grade serous ovarian tumors to detect presence of the markers flanking the SPRY2 and SPRY4 loci. The MIP copy numbers below 1.5 were considered deletion events (bolded). HG: high-grade; LG: low-grade; G1: grade 1; G2: grade 2; G3: grade 3.

53

Figure 2.1 The box plot displays the HEEBO microarray result of SPRY2 mRNA levels in ovarian tumors of various subtypes: high-grade serous (n = 35), low-grade serous (n = 2), endometrioid (n = 7), clear cell (n = 3), serous borderline (n = 1), endometrioid borderline tumor (n = 1) and normal Fallopian tube (n = 1). Pooled RNA from 10 human cell lines for broad gene coverage on the array was included as a reference (0). Dots represent individual samples.

54

Figure 2.2 Comparison of SPRY2 and SPRY4 mRNA levels in immortalised OSE (IOSE) and ovarian cancer cell lines by real-time PCR. Real-time data of SPRY2 and SPRY4 mRNA were normalized against the internal control, GAPDH. The average of normalized expression levels of the IOSE cell lines was calculated and set to 100% (dotted horizontal line). The relative levels of SPRY2 and (A) SPRY4 (B) mRNA were expressed as a percentage of the mean expression in IOSE cell lines. Columns: mean of three passages of the cell lines; bars: SD.

55

Figure 2.3 A schematic representation of the MIP copy number assay results of 28 highgrade serous carcinomas. The diagram shows the MIP ID, position and the corresponding percentage of loss of markers flanking the SPRY2 locus (left) and SPRY4 locus (right).

56

57

Figure 2.4 The effect of SPRY2 on EGF-induced E-cadherin suppression and cell invasion. A. SKOV3 cells were transiently transfected with either the empty pXJ40FLAG vector or FLAG-SPRY2 constructs. Starved SKOV3 cells were treated with 100 ng/ml EGF. Total protein was collected after 24 hr and analysed by Western blotting using anti-E-cadherin, anti-SPRY2 or anti-!-actin antibodies. B. The protein signal intensities were quantified and normalized against the internal control. The data are expressed as a percentage of control vector-transfected cells without EGF treatment and represents the mean SD of three independent experiments. C. SKOV3 cells transfected with either control or SPRY2-overexpression plasmid were seeded in Matrigel-coated transwell filters and cultured with 100 ng/ml EGF for 24 h. Invaded cells were then stained and quantified. The data are shown as the mean SD of four independent experiments. D. Representative photos of the invasion assay. The mean values that are not denoted by the same letter are significantly different. Open bars: control treatment; Filled bars: EGF treatment.

58

Figure 2.5 The correlation between SPRY2 and E-cadherin protein expression in ovarian cancers. A. Total protein from ovarian cancer cell lines and high-grade serous ovarian tumors was extracted and analysed via Western blot. B. Protein signal intensities were quantified and a correlation was assessed using the Spearman nonparametric correlation method.

59

Chapter 3 An amphiregulin and Sprouty4 loop regulates ovarian cancer cell invasiveness via an E-cadherin-dependent mechanism 3.1 Introduction Ovarian cancer is a common and the most lethal gynecological cancer. The high mortality rate is caused by a lack of reliable screening tests and obvious symptoms, which frequently results in diagnosis at advanced disease stages when peritoneal metastasis is already present (3). Impairment of the epidermal growth factor (EGF) system is known to play a role in ovarian cancer by directly enhancing the invasiveness and metastatic potential of cancer cells (69, 84, 99, 108-111). EGFR amplification, mutation and protein overexpression have been reported in ovarian cancer (57, 66, 67, 75). Alternatively, aberrant EGFR activity may be a result of overproduction of EGFR ligands. Among the EGFR-binding ligands, amphiregulin (AREG) levels are higher than TGF-! and EGF levels in ovarian cancer tissues and cell lines (89, 91). Similarly, the concentration of AREG in the peritoneal fluid of ovarian cancer patients at different stages of the disease ranges between 203 225 pg/ml and is significantly higher than the concentration of TGF-! (2.01 8.33 pg/ml) (92). These data suggest that AREG is an important ligand activating EGFR in cancer cells. Higher local AREG concentration may arise from the juxtacrine action of AREG (257). Furthermore, in regards to ovarian cancer, AREG has been shown to be upregulated by luteinizing hormone (LH) during ovulation (258). This finding not only suggests a high local concentration of AREG in the ovaries than the circulation but also suggests a causal link between LH, AREG and ovarian cancer. Our laboratory has 60

previously demonstrated that LH stimulates ovarian cancer migration and invasion (259, 260). Considering that AREG is induced by and mediates the actions of LH (258), exposure to AREG may in turn promote neoplastic transformation and progression. To date, studies of the physiological role of AREG in ovarian cancer have been restricted to its role in cell proliferation (91, 105). In ovarian cancer cells and normal ovarian surface epithelial cells, AREG induces a biphasic regulation of proliferation (105). Alternatively, EGFR hyperactivity may be the result of the deregulation of downstream effectors and regulators (261, 262), for instance, Sprouty (SPRY). SPRY members (SPRY1-4) have been proposed as general inhibitors of signalling downstream of EGFR and other receptor tyrosine kinases (RTKs) (127). In various normal cell systems, SPRY levels and activities are regulated by growth factors through activating RTKs (121, 123, 145-147). Similarly, SPRY levels have been found to be induced by growth factors (148) and oncogenic mutations downstream of RTKs in cancer cells (149). SPRY members act as tumor suppressor genes and negate many aspects of tumorigenesis, including cell invasion and cancer metastasis (148, 159-162). Our laboratory has recently found that SPRY2 antagonises EGF-induced invasion in ovarian cancer cells (So et al unpublished). To evaluate the proinvasive potential of AREG, two invasive ovarian cancer cell lines were treated with AREG, and invasiveness was assayed. We demonstrated a significant effect of AREG in promoting cell invasion through activation of the MAPK/ERK and PI3K/AKT pathways, induction of SLUG mRNA expression and reduction in level of the adhesion molecule E-cadherin. Using the AREG-induced Ecadherin downregulation and cell invasion as physiological endpoints, we tested the 61

hypothesis that AREG triggers a SPRY-mediated feedback response. We first showed that SPRY4 was significantly induced by AREG. To assess the importance of SPRY4 upregulation, SPRY4 was silenced using siRNA. Compared with control cells, ovarian cancer cells depleted of SPRY4 responded to AREG more significantly in terms of Ecadherin downregulation and cell invasion. These data support the presence and functionality of an AREG/SPRY4 loop in ovarian cancer invasion regulation.

3.2 Materials and methods 3.2.1 Cell culture and reagents The SKOV3 ovarian cancer cell line was obtained from the American Type Culture Collection (Manassas, VA). Our laboratory previously showed that SKOV3 is invasive and its invasiveness is increased by EGF (248). The cells were cultured in MCDB 105/M199 (1:1) supplemented with 5% heat-inactivated fetal bovine serum, 100 IU/ml penicillin and 100 g/ml streptomycin. The cells were cultured at 37C and 5% CO2. Fetal bovine serum was purchased from Hyclone Laboratories (Logan, UT). The MEK/ERK inhibitor U0126 was purchased from Calbiochem (San Diego, CA). Human recombinant AREG, the PI3K/AKT inhibitor LY294002, the EGFR inhibitor AG1478 and other tissue culture materials were obtained from Sigma Chemical Co. (St. Louis, MO) unless otherwise stated.

62

3.2.2 Transfection The empty pcDNA3.1 vector was obtained from Invitrogen (Carlsbad, CA). A human E-cadherin expression vector (plasmid 28009) was purchased from Addgene (Cambridge MA). Plasmids were transfected using Lipofectamine 2000 (Invitrogen, Carlsbad, CA). ON-TARGETplus SMARTpool SPRY4 siRNA and non-targeting siRNA (Dharmacon, Lafayette, CO) were transfected using Lipofectamine RNAiMAX (Invitrogen, Carlsbad, CA). 3.2.3 Real-time PCR Total RNA was extracted from cells using TRIzoL Reagent (Invitrogen) and used in first-strand DNA (cDNA) synthesis using the Invitrogen Super-ScriptTM first strand synthesis system for real-time PCR according to the manufacturers protocol. Real-time PCR was performed in an ABI 7300 real-time thermal cycler (ABI, Hercules, CA). The amplifications of E-cadherin, SLUG, ZEB1, SPRY2, SPRY4 and the internal control Gapdh were performed as follows: a 3 min hot start at 95C followed by 40 cycles of denaturation at 95C for 15 sec and amplification at 60C for 1 min. PCR reactions were performed in duplicate with the following PCR primers: E-cadherin, forward 5!GGGTGACTACAAAATCAATC-3! and reverse 5!-AAAGAGCCCTTACTGCCCCC-3!; SLUG, forward 5!-TTCGGACCCACACATTACCT-3! -3!; ZEB1 and reverse 5!5!-

GCAGTGAGGGCAAGAAAAAG

forward

GCACCTGAAGAGGACCAGAG-3! and reverse 5!-TGCATCTGGTGTTCCATTTT-3!; SPRY2, forward 5-TTGCACATCGCAGAAAGAAG-3 SPRY4, and reverse 5563

GAAGTGTGGTCACTCCAGCA-3;

forward

AGCCTGTATTGAGCGGTTTG-3 and reverse 5-GGTCAATGGGTAGGATGGTG3; GAPDH, forward 5!-GAGTCAACGGATTTGGTCGT-3! and reverse 5!-

GACAAGCTTCCCGTTCTCAG-3!. 3.2.4 Western blot analysis Equal amounts of total cell lysate were resolved on 7.5% SDS-PAGE gels and electrotransferred to a PVDF membrane. After blocking for 1 hr with 5% nonfat dry milk in TBS-T buffer (20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.1% Tween-20), the blots were probed for overnight at 4 C with the following primary antibodies: anti-E-cadherin antibody, which was obtained from BD Transduction Laboratories (Lexington, KY); the anti-SPRY4 antibody was obtained from Abcam (Cambridge, MA); anti-phospho-AKT (Ser473), anti-total AKT, anti-phospho-ERK1/2 (Thr202/Tyr204) and anti-ERK1/2 antibody, which were purchased from Cell Signaling, Inc. (Austin, TX) and anti-"-actin antibody, which was obtained from Santa Cruz Biotechnology (Santa Cruz, CA). The blots were then incubated with a peroxidase-conjugated secondary antibody (Bio-Rad) for 1 hr followed by detection with ECL chemiluminescence reagent (Amersham, Arlington Heights, IL) and exposure on X-ray films. 3.2.5 Invasion assay To assess invasion, 24-well transwell filters with an 8-#m pore coated with 1 mg/ml Matrigel (50 #l/well; BD Sciences, Mississauga, ON) were used. Ovarian cancer cells were trypsinised, re-suspended in 0.1% FBS medium and seeded in triplicate in the upper chamber. Medium containing 1% FBS was added to the lower wells. The chambers were incubated for 24 hrs at 37C in a 5% CO2 atmosphere. The cells that did not 64

penetrate the filter were wiped off. The invading cells on the lower surface of the filter were fixed with ice-cold methanol, stained with Hoechst 33258 and counted through epifluorescence microscopy with Northern Eclipse 6.0 software (Empix Imaging, Mississauga, ON). Triplicate inserts were used for each individual experiment, and the results are presented as the mean numbers. 3.2.6 Statistical analysis Real-time PCR quantification of mRNA level was calculated using the 2!! Ct method. Data are presented as the mean SD of three independent experiments or triplicates in a representative experiment and were analyzed by one-way ANOVA followed by Tukeys post-hoc test using GraphPad Prism 5 (GraphPad Software, San Diego, CA) to compare all pairs of columns. Columns are not denoted by the same letter are statistically different. Means not denoted by the same letter are significantly different (P < 0.05).

3.3 Results 3.3.1 AREG promoted invasion of ovarian cancer cells We tested the effect of AREG on the invasion of ovarian cancer cells using a Matrigel-coated transwell assay. SKOV3 and OVCAR5 cells were incubated with different concentrations of AREG and allowed to invade across the transwell for 24 hrs. AREG promoted invasion of both cell lines (Fig. 3.1). At 1 ng/ml, AREG was ineffective

65

at stimulating invasion, whereas invasion was significantly increased by 10 and 100 ng/ml AREG. 3.3.2 AREG reduced E-cadherin levels, and E-cadherin overexpression blocks AREG-induced invasion To test the involvement of the adhesion molecule E-cadherin on the proinvasive effect of AREG, we first examined the impact of AREG on E-cadherin levels. In agreement with its stimulatory effect on cell invasion, AREG treatment resulted in suppression of E-cadherin protein levels (Fig. 3.2A: SKOV3; Fig. 3.2B: OVCAR5). We next tested whether this effect of AREG involved transcriptional regulation of Ecadherin. Real-time PCR shown that AREG suppresses E-cadherin mRNA significantly with a maximal effect after 24 hrs of treatment (Fig. 3.2C). To establish a definite causal link between E-cadherin suppression and an increase in cell invasiveness, we transfected SKOV3 cells with a human E-cadherin expression plasmid. Treatment with AREG increased invasion of cells transfected with a control vector; however, E-cadherin overexpression not only suppressed basal cell invasion but also decreased AREG-induced invasion (Fig. 3.2D) 3.3.3 AREG suppressed E-cadherin level and promotes cell invasion via the EGFR To confirm that AREG-induced cell invasion was mediated by the EGFR, we used the pharmacological EGFR inhibitor, AG1478, to specifically block EGFR activity in SKOV3 cells. As shown in Fig. 3.3A and B, AG1478 markedly diminished the AREGinduced reduction in E-cadherin levels and cell invasion.

66

3.3.4 AREG induced Slug expression AREG strongly induced the expression of Slug and, to a lesser extent, Zeb1 (Fig. 3.4A-D), which are two transcriptional repressors of E-cadherin. Importantly, the maximal effects on SLUG and ZEB1 mRNA levels preceded changes in E-cadherin levels (Fig. 3.2C), implying that these transcriptional repressors mediate the suppression of Ecadherin by AREG. 3.3.5 The MAPK/ERK and PI3K/AKT pathways mediated the effects of AREG on SLUG mRNA and E-cadherin levels and cell invasion Treatment of SKOV3 cells with AREG induced rapid activation of ERK and AKT (Fig. 3.5A). To elucidate whether these molecular pathways mediate these effects of AREG, SKOV3 cells were co-treated with AREG in the presence of the MEK/ERK inhibitor, U0126, or the PI3K/AKT inhibitor, LY294002. As shown in Fig. 3.5B and 3.5C, the inhibitors effectively blocked the effects of AREG on SLUG mRNA induction and E-cadherin suppression. Furthermore, blocking these pathways totally abolished AREG-stimulated SKOV3 cell invasion (Fig. 3.5D). 3.3.6 AREG induced SPRY4 expression SPRY deficiencies are common in cancers and may result in excessive growth factor activities (152). Therefore, it was important to investigate the integrity and functionality of a SPRY feedback loop in ovarian cancer cells. When SKOV3 cells were treated with increasing concentrations of AREG, SPRY2 and SPRY4 mRNA levels was elevated, but SPRY4 was stimulated to a much greater extent (approximately 4 fold versus 10 fold) (Fig. 3.6A). In a time course experiment, the stimulatory effect of AREG on SPRY4 67

mRNA peaked at 3 hrs (Fig. 3.6B). Induction of SPRY4 expression was also confirmed at the protein level (Fig. 3.6C). 3.3.7 SPRY4 knockdown enhanced AREG-induced E-cadherin suppression and invasion To understand the importance of SPRY4 induction, we evaluated AREG function in the absence of SPRY4. Using SPRY4-targeting siRNA (siSPRY4), both endogenous and AREG-induced SPRY4 levels were completely depleted (Fig. 3.7A). The siSPRY4 reduced the basal E-cadherin levels compared to cells transfected with control siRNA (siCtrl) (Fig. 3.7A). When the transfected cells were treated with AREG, E-cadherin protein levels were reduced, and this suppressive effect was more obvious in siSPRY4transfected cells (Fig. 3.7A). Interestingly, siSPRY4 had no effect on E-cadherin mRNA level (Fig. 3.7B). We next assayed the influence of SPRY4 on cell invasion. Similar to the results of E-cadherin levels, siSPRY4 increased the basal level of invasion and amplified the invasion caused by AREG (Fig. 3.7C, D).

3.4 Discussion The current study was undertaken to investigate the functional relationships between AREG and SPRY in ovarian cancer cell invasiveness. We demonstrate that AREG promotes ovarian cancer cell invasion through activation of ERK and AKT, induction of Slug and reduction of E-cadherin levels. Furthermore, we showed that AREG elicited a regulatory feedback response through the induction of SPRY4, which, in turn, reversed the E-cadherin suppression and antagonised the cell invasion mediated by AREG. We 68

confirmed that these AREG actions are mediated by EGFR (Fig. 3.3). Interestingly, the EGFR inhibitor and inhibitors of the MEK/ERK and PI3K/AKT pathways also increased the basal E-cadherin level and reduced basal invasion (Fig. 3.3). These data are explained by the autocrine nature of AREG and other EGFR ligands; ample evidence have demonstrated that ovarian cancer cells express these ligands, which constitute autocrine loops with EGFR (53). This is the first report to demonstrate an invasion-promoting role for AREG in ovarian cancer cells (Fig. 3.1A). Although it is not a novel finding that AREG promotes cancer cell invasion and metastasis, most of the reported proinvasive effects of AREG are associated with protease-mediated extracellular matrix (ECM) degradation. For example, urokinase and plasminogen activator inhibitor-1 are responsible for AREG-induced invasion of breast cancer cells (263). Moreover, the levels and activities of matrix metalloproteinases are upregulated by AREG to mediate invasion of breast cancer cells (264, 265), head and neck squamous carcinoma cells (266) and mesothelioma cells (267). AREG also regulates levels of extracellular matrix metalloproteinase inducers in transformed breast epithelial cells (264). In addition, integrin, which couples ECM to the intracellular cytoskeleton and whose level and activation are altered during colon cancer cell invasion, is induced by AREG (268). The involvement of these molecules in AREGinduced invasion is a possible explanation to the observation that MEK/ERK and PI3K/AKT inhibitors have greater effects on invasion than on E-cadherin level (Fig. 3.5 C, D). Here we show that AREG induces SLUG mRNA expression (Fig. 3.4), suppresses E-cadherin levels (Fig. 3.2) and stimulates ovarian cancer cell invasion (Fig. 3.1). Similarly, AREG has been reported to reduce E-cadherin level and adherence in 69

keratinocytes (269). Moreover, AREG promotes a reduction in membrane-localized Ecadherin and a motile morphology in MCDK cells (270), suggesting that E-cadherin is a common mediator of AREG-stimulated cell motility. Although SPRY are general inhibitors of signalling downstream of RTKs (127), there is evidence that individual SPRY proteins are preferentially regulated by and associate with different growth factors. For example, fibroblast growth factor (FGF) induces the expression of SPRY2 and SPRY4 (271), whereas SPRY1 mRNA is transiently downregulated by FGF (272). Furthermore, an insensitivity to the growth factor-induced upregulation of SPRY has been observed in cancer cell lines. FGF-2 treatment causes a rapid and transient increase in normal prostate epithelial cell SPRY1 mRNA and protein levels; however, FGF-2 reduces SPRY1 mRNA levels in neoplastic epithelia (152). Refractory to growth factor stimulation results in excessive growth factor signals and would favour tumorigenesis. In ovarian cancer cells, both SPRY2 and SPRY4 mRNA levels were elevated by AREG, and the increase was substantially higher for SPRY4 than SPRY2 mRNA (Fig. 3.1A). This result may suggest a stronger functional relevance between AREG and SPRY4 than SPRY2. However, SPRY4 inhibition on EGFR activity has been demonstrated in some but not all studies (136, 137, 273, 274). Therefore it was of great interest to elucidate the regulatory role of SPRY4 on AREG function. Consistent with the inhibitory nature of SPRY, SPRY4 siRNA amplified the effect of AREG and led to a further decrease in E-cadherin levels. More importantly, treating SPRY4-depleted cells with AREG resulted in a more significant increase in invasion then AREG or siRNA treatment alone. These data indicate that SPRY4 levels induced by AREG would feed back to counteract the effects of AREG. 70

In sharp contrast to SPRY2, the molecular targets of SPRY4 are poorly defined (127). SPRY4 binds directly to RAF1 (137) and BRAF (132) to inhibit VEGF-induced ERK activation (137). SPRY4 also binds constitutively to TESK1 (testicular protein kinase 1). The SPRY4/TESK interaction is increased by EGF and suppresses integrinmediated cell spreading (136, 138). In addition, a recent report showed that SPRY4 overexpression increased levels of E-cadherin (159). In agreement, the current study showed that the SPRY4 and E-cadherin protein levels were positively associated and that SPRY4 depletion led to a decrease in E-cadherin level (Fig. 3.7A). All of these data suggest that E-cadherin is a molecular target of SPRY4 and that it is positively regulated by SPRY4. The loss of the epithelial marker E-cadherin is a hallmark of epithelial-tomesenchymal transition and leads to the loss of cell-cell contact and the adoption of a motile phenotype, which leads to an increase in invasiveness (Fig. 3.7C, D). Together with our previous demonstration of the E-cadherin inhibition of ovarian cancer proliferation (275), SPRY4 may potentially regulate various aspects of ovarian cancer tumorigenesis through the modulation of E-cadherin. SPRY have been shown to execute their tumor suppressing roles in response to both antitumorigenic and oncogenic stimulations. For example, SPRY4 is upregulated by Wnt7A/Fzd9 and peroxisome proliferator-activated receptor gamma pathway to mediate their effects on non-small-cell lung cancer cell migration and invasion inhibition (159, 276). On the contrary, in our study, SPRY4 was downstream to an invasion-promoting growth factor and played a role as a mediator of the inhibitory feedback loop. Similarly, significant Spry2 mRNA expression is induced in lung epithelium in mice carrying a germline oncogenic K-rasG12D mutation. Such SPRY2 upregulation is antitumorigenic 71

in nature as number of tumors as well as overall tumor burden is significantly increased after crossing the mice with a SPRY2-null background (149). These examples indicate that the upregulation of the tumor suppressor SPRY mediates antitumorigenic functions during tumorigenesis.

72

Figure 3.1 AREG promoted ovarian cancer cells invasion. A and B, SKOV3 and OVCAR5 cells were trypsinized and seeded onto Matrigel-coated transwell inserts with 0 - 100 ng/ml AREG. Cells were allowed to invade for 24 hrs. Invaded cells were stained and quantified. Data are shown as the mean SD of three independent experiments. Means not denoted by the same letter are significantly different. C and D show representative photos of the invasion assay. 73

Figure 3.2 AREG reduced E-cadherin levels and E-cadherin overexpression blocked AREG-induced invasion. SKOV3 (A) and OVACR5 (B) cells were treated with 0 - 100 ng/ml AREG for 24 hrs. E-cadherin and the internal control actin were analyzed using Western blot. C, SKOV3 cells were treated with AREG over a 24-hr time course. The effect of AREG on mRNA levels of E-cadherin and the internal control Gapdh were assayed. D, SKOV3 cells were transiently transfected with control (empty vector) or human E-cadherin expression (E-cadherin) vectors. After 48 hrs, transfected cell were seeded onto Matrigel-coated transwell inserts for 24 hrs. Invaded cells were stained and quantified. Open bars: control treatment; filled bars: AREG treatment (10 ng/ml). Relative levels are expressed as a percentage of control. Data are shown as the mean SD of three independent experiments (A, B) or triplicates of a representative experiment (C, D). Means not denoted by the same letter are significantly different. 74

Figure 3.3 AREG suppressed E-cadherin level and promoted cell invasion via the EGFR. A, SKOV3 cells were pretreated with AG1478 (10 !M) for 30 min before the addition of AREG (10 ng/ml) for 24 hrs. Changes in E-cadherin levels were detected by Western blot. Data are shown as the mean SD of three independent experiments. B, SKOV3 cells were trypsinized and incubated with AG1478 for 30 min. Cells were then co-treated with AREG (10 ng/ml) and seeded onto Matrigel-coated transwells for 24 hrs. Invaded cells were stained and quantified. Relative levels are expressed as a percentage of control. Data are shown as the mean SD of triplicates of a representative experiment. Means not denoted by the same letter are significantly different. Open bars: control treatment; filled bars: AREG treatment.

75

Figure 3.4 AREG induced SLUG mRNA. SKOV3 cells were treated with 100 ng/ml AREG for 0 to 24 hrs as indicated. SLUG (A) and ZEB1 (B) mRNA levels were then analyzed by real-time PCR. SKOV3 cells were treated with different doses of AREG for 3 hrs before being collected for SLUG (C) or ZEB1 (D) real-time PCR. Relative mRNA levels of SLUG and ZEB1 were expressed as a percentage of the control. Data are shown as the mean SD of triplicates of a representative experiment. Means not denoted by the same letter are significantly different.

76

77

Figure 3.5 The MAPK/ERK and PI3K/AKT pathways mediated the effects of AREG on Slug and E-cadherin level and cell invasion. A, SKOV3 cells were treated with 10 ng/ml AREG for 0.5 and 3 hrs. ERK and AKT activation were detected using Western blot assays. B and C, SKOV3 cells were pre-incubated with U0126 (10 !M) and LY294002 (25 !M) for 30 min prior to co-treatment with AREG. Cells were harvested after 3 hrs and SLUG mRNA level was analyzed by real-time PCR. E-cadherin protein level was analyzed by Western blot after 24 hrs. D, trypsinized SKOV3 cells were pre-treated with U0126 (10 !M) and LY294002 (25 !M) for 30 min and co-treated with AREG in Matrigel-coated transwell inserts for 24 hrs. Invaded cells were stained and quantified. Relative levels are expressed as a percentage of the control. Data are shown as the mean SD of triplicates of a representative experiment (B) or three independent experiments (C, D). Means not denoted by the same letter are significantly different. Open bars: control treatment; filled bars: AREG treatment.

78

C B

Figure 3.6 AREG induced SPRY4. SKOV3 cells were treated with AREG at various concentrations (A) or for various durations (B). The mRNA levels of SPRY2 and SPRY4 were detected with real-time PCR. C, SKOV3 cells were treated with various concentrations of AREG, and the effect on the SPRY4 level was assayed using Western blotting. The relative levels are expressed as a percentage of the control. The data are shown as the mean SD of triplicates of a representative experiment (A, B) or three independent experiments (C). Means not denoted by the same letter are significantly different.

79

80

Figure 3.7 SPRY4 knockdown enhanced AREG-induced E-cadherin suppression and invasion. A, SKOV3 cells were transiently transfected with control siRNA (siCtrl) or SPRY4 siRNA (siSPRY4) for 48 hrs. After transfection, the cells were treated with 10 ng/ml AREG for 24 hrs, and E-cadherin and SPRY4 levels were assayed through immunoblotting. Data is shown as mean SD of three independent experiments. B, The transfected cells were treated with 10 ng/ml AREG for 12 hrs and subjected to real-time PCR to determine E-cadherin mRNA level. Data is shown as mean SD of triplicates in a representative experiment. C, At 48 hrs after transfection, SKOV3 cells were trypsinised, seeded in Matrigel-coated transwells and treated with 10 ng/ml AREG for 24 hrs. The invading cells were stained and quantified. The data are shown as the mean SD of triplicates in a representative experiment. D, Representative images of the invasion assay. Relative levels were expressed as a percentage of control. Means not denoted by the same letter are significantly different. Open bars: control treatment; filled bars: AREG treatment.

81

Chapter 4 Sprouty4 feedback regulates epidermal growth factor/AKT/hypoxiainducible factor-1 alpha axis in ovarian cancer cells 4.1 Introduction Functioning as a negative feedback regulator of receptor tyrosine kinase (RTK)/ERK signalling, Sprouty (SPRY) expression is primarily regulated by RTK activation. Growth factor stimulation has been widely demonstrated to be essential and sufficient for SPRY gene expression (122, 145, 277, 278). Various groups have shown that different growth factors and phorbol 12,13-dibutyrate induce SPRY expression in an ERK pathwaydependent manner, downstream from RTK activation (147, 271). In human tumor cells with constitutive ERK activation, SPRY levels are elevated, which can be reduced by an ERK pathway inhibitor (147). To elucidate the regulatory mechanisms that control SPRY levels, the human SPRY2 and SPRY4 promoters have been recently cloned and multiple potential transcription factor binding sites were identified on these promoters (279, 280), thereby suggesting the regulation of SPRY by pathways other than ERK signalling. A putative hypoxiainducible factor-1 alpha (HIF-1!) binding site has been identified on the human SPRY4 promoter (280). Accordingly, SPRY4 level has been shown to be elevated by chemicalmimics of hypoxia in both normal and malignant cells (281), suggesting a positive effect of HIF-1! on SPRY4 levels. In addition to tissue hypoxia, HIF-1! levels and HIF-1 activity have been shown to be modulated by several environmental (211, 282) and hormonal factors (174, 283), including growth factors (188, 191). Our laboratory has recently reported that epidermal 82

growth factor (EGF) could strongly elevate HIF-1! levels in ovarian cancer cells, which mediated EGF-induced E-cadherin downregulation and cell invasion (unpublished data). We sought to investigate whether HIF-1! plays a role in mediating the regulatory effect of EGF on SPRY4 level. Emerging experimental data shows that SPRY could be regulated by and modulate additional signalling pathways. For example, fibroblast growth factor (FGF) induced SPRY1 and SPRY2 expression via a Ca2+-dependent pathway (284), while Xenopus SPRY can inhibit calcium signalling (285). In colon cancer cells, the SPRY2 levels were suppressed by an active vitamin D metabolite (1,25(OH)2D3) through an E-cadherindependent mechanism, and SPRY2, in turn, repressed 1,25(OH)2D3-induced E-cadherin expression (286). Recently, an HIF-1! regulator, Seven-in-Absentia homologue-2 (SIAH2) (169), was found to interact with SPRY4 (287), which suggests that SPRY4 may play a role in HIF-1! regulation. However, direct information regarding the effect of SPRY4 on HIF-1! levels and a functional linkage between SPRY4 and HIF-1! are still lacking. We demonstrated that EGF has a strong inducing effect on the SPRY4 level in ovarian cancer cells via an AKT- and HIF-1!- dependent mechanism. Functionally, SPRY4 feedback inhibits AKT activation and antagonizes EGF-induced HIF-1! levels and HIF-1 activity. Together, our data demonstrate the presence and functionality of a novel EGFR/AKT/HIF-1! and SPRY4 feedback loop in ovarian cancer cells.

83

4.2 Materials and methods 4.2.1 Cell culture and reagents SKOV3, OVCAR3 and OVCAR4 ovarian cancer cell lines were obtained from American Type Culture Collection (Manassas, VA). The cell lines were cultured in MCDB 105/M199 (1:1), supplemented with 5% heat-inactivated fetal bovine serum, 100 IU/ml penicillin and 100 g/ml streptomycin. The cells were cultured at 37C and 5% CO2. Fetal bovine serum was purchased from Hyclone Laboratories (Logan, UT). The MEK/ERK inhibitor, U0126, was purchased from Calbiochem (San Diego, CA). Human recombinant EGF, MEK/ERK inhibitor (PD98059), PI3K/AKT inhibitors (LY294002 and Wortmannin) and other tissue culture materials were obtained from Sigma Chemical Co. (St. Louis, MO), unless otherwise stated. 4.2.2 Transfection SPRY4 promoter-luciferase reporter constructs were kindly provided by Dr. D. Warburton (Children Hospital Los Angeles, California). The FLAG-SPRY4 construct and the empty pXJ40-FLAG vector were generous gifts from Dr. Graeme R. Guy (Institute of Molecular and Cell Biology, Singapore). The hypoxia responsive element (HRE)-luciferase reporter construct was purchased from Addgene (Cambridge, MA). The plasmids were transfected using Lipofectamine 2000 (Invitrogen, Carlsbad, CA). ONTARGETplus SMARTpool non-targeting (siCtrl), HIF-1! siRNA (HIF-1!), SPRY4 siRNA (siSPRY4) and non-targeting siRNA (Dharmacon, Lafayette, CO) were transfected using Lipofectamine RNAiMAX (Invitrogen).

84

4.2.3 Real-time PCR Total RNA extracted from cells using TRIzoL Reagent (Invitrogen) was used in first-strand DNA (cDNA) synthesis using Invitrogen Super-ScriptTM first-strand synthesis system for real-time PCR according to the manufacturers protocol. Real-time PCR was performed using an ABI 7300 real-time thermal cycler (ABI, Hercules, CA). The detections of SPRY4, VEGF, and the internal control, GAPDH, mRNAs were performed as follows: a 3-min hot start at 95C followed by 40 cycles of denaturation at 95C for 15 sec, and amplification at 60C for 1 min. PCR reactions were performed in duplicates with the following PCR primers: SPRY4, forward 5-

AGCCTGTATTGAGCGGTTTG-3 and reverse 5-GGTCAATGGGTAGGATGGTG3; VEGF, forward 5-GGCTCTAGATCGGGCCTCCGAAACCAT-3 and reverse 5GGCTCTAGAGCGCAGAGTCTCCTCTTC-3; and GAPDH, forward 5-

GAGTCAACGGATTTGGTCGT-3 and reverse 5-GACAAGCTTCCCGTTCTCAG-3. 4.2.4 Western blot analysis Equal amounts of total cell lysates were resolved in 7.5% SDS-PAGE and electrotransferred to a PVDF membrane. After blocking for 1 hr with 5% non-fat dry milk in TBS-T buffer (20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.1% Tween-20), the blots were probed for overnight at 4 C with the appropriate primary antibodies. The antibody for HIF-1! was purchased from BD Transduction Laboratories (Lexington, KY), the anti-SPRY4 antibody was obtained from Abcam (Cambridge, MA), the anti-!actin antibody was obtained from Santa Cruz Biotechnology (Santa Cruz, CA), the antiphospho-EGFR (Tyr992), anti-EGFR, anti-phospho-AKT (Ser473), anti-total AKT, anti85

phospho-ERK1/2 (Thr202/Tyr204), anti-ERK1/2 and PTEN antibodies were purchased from Cell Signaling, Inc. (Austin, TX). The blots were then treated with a peroxidaseconjugated secondary antibody (Bio-Rad) for 1 hr followed by a detection step with ECL chemiluminescence reagent (Amersham, Arlington Heights, IL) and exposure to X-ray films. 4.2.5 Luciferase assay Cells transfected with luciferase reporter constructs were treated with EGF for 24 hours. After treatment, the cells were lysed, and total cell lysates were centrifuged at 13,000 rpm for 10 min to remove any cell debris. Luciferase activity was determined according to the manufacturers instructions (Promega). 4.2.6 Statistical analyses For real-time PCR data, the relative quantification of levels was calculated using the 2!! Ct method. Data are presented as the mean SD of three independent experiments or triplicates in a representative experiment and were analyzed by one-way ANOVA followed by Tukeys post-hoc test using GraphPad Prism 5 (GraphPad Software, San Diego, CA) to compare all pairs of columns. Columns are not denoted by the same letter are statistically different. Means not denoted by the same letter are significantly different (P < 0.05).

86

4.3 Results 4.3.1 EGF increased SPRY4 in ovarian cancer cells To test the effect of EGF on the SPRY4 level, SKOV3 cells were challenged with either 100 ng/ml EGF over a 24-hr time course or various EGF concentrations for 3 hrs. SPRY4 mRNA levels were significantly elevated (Fig. 4.1A, B). An inductive effect of EGF was also observed in two additional ovarian cancer cell lines, OVCAR3 and OVACR4 (data not shown). A similar robust inducing effect was also observed at the SPRY4 protein level (Fig. 4.1C). To evaluate the contribution of transcriptional induction via promoter activation in SPRY4 mRNA induction, we transiently transfected reporter gene constructs containing different lengths of the SPRY4 5 flanking region into SKOV3 cells. Under EGF treatment, no induction was observed in the minimal promoter region (31/+56). The promoter region (-69/+56) contains the consensus sequence for the binding of HIF-1 and signal transducer and activator of transcription 5. In addition, the promoter region (-4446/+56) contains many putative binding sites for transcription factors, including stimulating protein 1 (Sp1), activator protein 2, Elk-1, and WT-1 (280). Under EGF treatment, luciferase activities driven by regions (-69/+56) and (-4446/+56) were significantly induced when compared with the control treatment (Fig. 4.1D). 4.3.2 The MEK/ERK and PI3K/AKT pathways mediated the effects of EGF on SPRY4 levels To evaluate the role of the MEK/ERK and PI3K/AKT pathways in EGF-induced SPRY4, SKOV3 cells were pre-incubated with inhibitors for the specific pathways for 30 mins. The MER/ERK inhibitor significantly abolished the effect of EGF on both SPRY4 87

mRNA and SPRY4 protein (Fig. 4.2A and 4.2B). Although PI3K inhibitor partially blocked SPRY4 mRNA increased by EGF (Fig. 4.2A), it blocked completely blocked the increase in SPRY4 levels (Fig. 4.2B). 4.3.3 EGF induced HIF-1! via the PI3K/AKT pathway Our previous and present reports demonstrated a strong effect of EGF on HIF-1! levels (Fig. 4.3A and Cheng et al., unpublished). When pre-treated with the PI3K/AKT pathway inhibitor, LY294002, both basal and EGF-induced HIF-1! levels were reduced significantly, whereas the MEK/ERK inhibitor slightly suppressed EGF-induced HIF-1! levels (Fig. 4.3B). Moreover, the cells were transfected with an HRE-luciferase reporter gene construct by which the activity was driven by HIF. EGF-induced HRE-luciferase activity was completely blocked by the PI3K/AKT pathway inhibitors, Wortmannin and LY294002, and not by the MER/ERK inhibitor (Fig. 4.3C). 4.3.4 HIF-1! plays a minor role in EGF-induced SPRY4 level The presence of the HIF responsive element (HRE) in the human SPRY4 promoter (280) prompted us to question whether HIF-1! mediates the effect of EGF on SPRY4 regulation. At the promoter level, we performed site-directed mutagenesis of the HRE located within the SPRY4 promoter (Fig. 4.4A). EGF increased luciferase activity in cells transfected with the wild-type promoter construct. In contrast, mutation of the HRE in the SPRY4 promoter reduced both the basal and luciferase activities induced by EGF (Fig 4.4A). Next, we depleted endogenous HIF-1! expression using specific siRNA (Fig. 4.4B). However, we did not observe a significant reduction in EGF-induced SPRY4 mRNA levels (Fig. 4.4C) following HIF-1! knockdown, although the mRNA level of the 88

HIF-1! target, VEGF, increase by EGF being partially reduced (Fig. 4.4C). Accordingly, HIF-1! knockdown did not alter the effect of EGF on SPRY4 levels (Fig. 4.4D). 4.3.5 SPRY4 overexpression reversed EGF-induced HIF-1! levels and HIF-1 activity We next investigated the mechanism by which SPRY4 impacts HIF-1! levels and HIF activity. We overexpressed human SPRY4 in SKOV3 cells, which express low levels of SPRY4. Overexpression was confirmed by Western blot (Fig. 4.5A). After EGF treatment, HIF-1! levels were elevated. However, SPRY4-overexpressing cells displayed lower HIF-1! levels when compared with cells transfected with the control vector (Fig. 4.5A). To evaluate whether the altered HIF-1! levels were accompanied by a change in HIF-1 activity, two experiments were conducted. First, an HRE-luciferase reporter gene construct was co-transfected with SPRY4 overexpression vector into SKOV3 cells. EGF strongly increased the HRE-mediated luciferase activity, which was markedly reduced by SPRY4 overexpression (Fig. 4.5B). Similarly, SPRY4 also antagonized the inductive effect of EGF on VEGF mRNA (Fig 4.5C). 4.3.6 SPRY4 knockdown enhanced the effect of EGF on HIF-1! To eliminate the possibility of non-specific effects due to massive SPRY4 overexpression, we confirmed our observation via SPRY4 knockdown using specific siRNA in cell lines expressing higher levels of endogenous SPRY4 (OVCAR3 and OVACAR4). In siSPRY4-treated OVCAR3 cells, EGF elicited a higher HIF-1! level than that of the cells treated with siCtrl (Fig. 4.6A, C).

89

4.3.7 AKT pathway mediated HIF-1! regulation by EGF and SPRY4 In the overexpression and knockdown experiments, we observed changes in the phospho-AKT (pAKT) levels similar to those observed for HIF-1! (Fig. 4.5A, 4.6A). In SPRY4-overexpressing SKOV3 cells, the pAKT level was lower than that of the control cells (Fig. 4.5A). In contrast, AKT activation in SPRY4 siRNA-treated OVCAR3 cells was increased (Fig. 4.6A, C). To further confirm the causal role of the elevated AKT activation of HIF-1! levels, we co-incubated the cells with PI3K/AKT inhibitors. In OVCAR3 cells, when AKT activation was inhibited, the SPRY4-augmented EGFinduced HIF-1! expression was completely impaired (Fig. 4.7).

4.4 Discussion Despite the importance of RTKs and its downstream ERK pathway in the regulation of SPRY expression is well-known (127, 147, 271), the detailed molecular mechanisms of SPRY regulation are not completely understood. In the present study, we confirmed the importance of ERK in SPRY4 expressions as MEK/ERK inhibitor reduced the basal levels of SPRY4 protein (Fig. 4.2), furthermore we detected a strong stimulation of the SPRY4 level by EGF that was primarily mediated by the ERK pathway (Fig. 4.2). In addition, the (-4446/+56) promoter construct responded to EGF more prominently than the region (-69/+56), this is probably due to the region (-4446/+56) contains many putative sites for transcription factors downstream to ERK, for example Sp1 and Elk-1 (280). In contrast, inhibition of the PI3K/AKT pathway exerted more prominent

inhibitory effect on and SPRY4 protein than SPRY4 mRNA levels (Fig. 4.2), thereby 90

suggesting differential regulatory mechanisms at the transcriptional and posttranscriptional levels by the AKT pathway. AKT has been shown to bind to an E3 ligase, SIAH2 (288), which can directly interact with SPRY4 (287). This finding suggests that AKT regulates SPRY4 by modifying SPRY4 degradation and has more significant effects on SPRY4 protein than SPRY4 mRNA. When SKOV3 cells were treated with two different PI3K/AKT inhibitors, the basal HRE-luciferase activities were reduced; this observation agrees with the previous finding that basal HIF-1! levels in ovarian cancer cell lines were PI3K-dependent (289). We also showed that EGF induced HIF-1! accumulation and activity via the PI3K/AKT pathway (Fig. 4.3). This A recent article showing a positive effect of hypoxia on SPRY4 level (281), therefore we then tested the cross-talk of EGF and HIF-1! on SPRY4 regulation. Mutation of the HRE within the human SPRY4 promoter, which impaired both basal and EGF-induced SPRY4 promoter activity, highlights the importance of HIF-1! in SPRY4 transcriptional regulation. However, HIF-1! knockdown failed to decrease EGF-induced SPRY4 mRNA and SPRY4 protein levels (Fig. 4.4C, D). The ineffectiveness of HIF-1! knockdown could be attributed to the dominance of the ERK signal over the HIF-1! pathway, thereby rendering the effects insignificant. In addition, the human SPRY4 promoter HIF-1! overlaps with CpG islands, which has been found to be hypermethylated in prostate cancer (161). The SPRY4 promoter is likely hypermethylated in ovarian cancer cell lines, as SPRY4 expression could be reactivated using a demethylating agent treatment (unpublished data). Methylation may hinder HIF-1! accessibility to HRE, thus rendering the promoter refractory to the HIF-1! level dynamics. The same concept has been applied to a well-known HIF-1! target, HIF 91

prolyl-hydroxylase 3 (PHD3). The PHD3 promoters in various human carcinoma cell lines have been silenced by aberrant methylation, causing these cells to express reduced basal and hypoxia-upregulated PHD3 mRNA (290). This notion also provides an explanation by which the promoter construct can be activated by EGF via the HRE, regardless of the endogenous SPRY4 promoter methylation status. Hypoxia stabilizes and causes the accumulation of HIF-1!, which acts as the regulator of a repertoire of hypoxic responses, including angiogenesis (163). To date, a direct functional connection between SPRY4 and HIF-1! has not been described, although there are lines of experimental evidence suggesting that SPRY4 and HIF-1! are functionally linked. As demonstrated with a Spry4-knockout mouse model and ex vivo assays, SPRY4 functions as a negative angiogenic regulator, for example, SPRY4negative tumors have enhanced vascularisation (274). Strikingly, in SPRY4-knockout cells, growth factors such as bFGF and VEGFA induced a greater AKT activation (274). However, not addressed in that study was the negative effect of SPRY4 on HIF-1! levels and activity, possibly an additional cause of the SPRY4 antiangiogenic effect. Our observation supports a novel regulatory involvement of SPRY4 on HIF-1!. Next, we investigated the mechanism underlying the effect of SPRY4 on EGFinduced HIF-1! expression. We first delineated the signalling pathways mediating the effect of EGF. Pre-incubation with PI3K/AKT inhibitors but not MEK inhibitor effectively blocked the basal and EGF-induced HIF-1! and HIF-1 activity (Fig. 4.3B, C). The importance of the PI3K/AKT pathway in EGF-induced HIF-1! accumulation has been previously reported in prostate cancer cells (188). We observed concurrent increases 92

in EGF-induced HIF-1! level and AKT activation in SPRY4 siRNA-treated cells, whereas the levels of both HIF-1! and phospho-AKT were reduced in SPRY4 overexpressing cells (Fig. 4.5, 4.6). The inhibition of AKT activation abrogated the effect of SPRY4 siRNA on HIF-1! expression (Fig. 4.7). These data suggest that SPRY4 regulates HIF-1! via EGF-induced AKT signalling. Although inhibition of the RTK/RAS/MAPK/ERK pathway is the principal action of SPRY (127), emerging data have suggested that AKT is also a target of SPRY activity (134, 274). Another SPRY member, SPRY2, has been shown to regulate AKT activity by modulating the content and activity of phosphatase and tensin homologue deleted on chromosome 10 (PTEN) (134). However, PTEN levels were not affected by both SPRY4 knockdown and overexpression (Fig. 4.5A, 4.6A), therefore the mechanism by which SPRY4 regulates AKT and HIF-1! remains unknown.

93

Figure 4.1 EGF induced SPRY4 expression in ovarian cancer cells. SKOV3 cells were treated with EGF time (A) or dose (B) courses and assayed for SPRY4 mRNA level using real-time PCR. The relative levels of SPRY4 mRNA are expressed as a percentage of the control. C. SKOV3 cells were treated with 10 ng/ml EGF for various durations and assayed for SPRY4 protein level by immunoblotting. D. SKOV3 cells were transfected with SPRY4 promoter constructs and treated with EGF for 24 hrs. Cell lysate was collected, and luciferase activity was assessed. The relative levels of luciferase activity are expressed as a percentage of the control treatment of (-31/+56). E. OVCAR3 and OVCAR4 cells were treated with 100 ng/ml EGF for 3 hrs and SPRY4 mRNA were assayed using real-time PCR. The data are presented as the mean SD of three

independent experiments (A, C) or triplicates in a representative experiment (B, D, E). The mean values that are not denoted by the same letter are significantly different. 94

Figure 4.2 The MEK/ERK and PI3K/AKT pathways mediated the effects of EGF on SPRY4 mRNA (A) and SPRY4 protein (B) levels. SKOV3 cells were pre-incubated with either MEK/ERK inhibitor (U0126, 10 M) or PI3K/AKT inhibitor (LY294002, 25 M) for 30 min prior to EGF (10 ng/ml) treatment. The cells were harvested for real-time PCR (A) and Western blot (B) analysis after 3 hrs and 24 hrs, respectively. Relative levels of SPRY4 mRNA and SPRY4 are expressed as a percentage of the control. The data are presented as the mean SD of three independent experiments. Mean values that are not denoted by the same letter are significantly different.

95

Figure 4.3 EGF induced HIF-1! and HIF-1 activity via the PI3K/AKT pathway. A. SKOV3 cells were treated with EGF for different durations and HIF-1! protein levels were assessed. B. SKOV3 cells were pre-incubated with PI3K/AKT inhibitor (LY294002, 25 M) and MEK/ERK inhibitor (PD98059, 10 M) for 30 min prior to EGF treatment. Their effects on HIF-1! protein were assayed after 30 min. C. SKOV3 cells were transfected with HRE-luciferase construct, pre-treated with PI3K/AKT inhibitors (LY294002, 25 M; Wortmannin 1 M) and MEK/ERK inhibitor (U0126, 10 M) and treated with EGF for 24 hrs. Cell lysates were collected, and luciferase activity was assessed. The relative levels of luciferase activity are expressed as a percentage of the control. The data are shown as the mean SD of triplicates in a representative experiment. Mean values that are not denoted by the same letter are significantly different. 96

97

Figure 4.4 HIF-1! plays a minor role in EGF-induced SPRY4 expression. A. A diagram displaying the site-directed mutagenesis of the HIF-1! binding site of the SPRY4 (69/+56) promoter region (upper panel). SKOV3 cells were transfected with SPRY4 (69/+56) promoter-luciferase constructs containing either wild-type or mutated HRE. After treatment with 10 ng/ml EGF for 24 hrs, cell lysates were collected for the luciferase activity assay. The relative levels of luciferase activity are expressed as a percentage of the control (lower panel). B. SKOV3 cells were transfected with either 50 nM non-targeting siRNA (siCtrl) or HIF-1! siRNA (siHIF-1!) and treated with 10 ng/ml EGF for 3 hrs. Cells were collected for Western blot analysis of HIF-1! protein (B) as well as real-time PCR for SPRY4 (C) and VEGF (D) mRNAs. E. Non-targeting siRNA or HIF-1! siRNA-transfected cells treated with EGF for 24 hrs were harvested, and SPRY4 level was assayed by Western blot. The relative levels are expressed as a percentage of the control. The data are shown as the mean SD of triplicates in a representative experiment (A) or three independent experiments (C, D, E). Mean values that are not denoted by the same letter are significantly different. 98

99

Figure 4.5 SPRY4 overexpression reversed EGF-induced HIF-1! expression and HIF-1 activity. A. SKOV3 cells were transfected with either SPRY4 overexpression or empty vectors. Transfected cells were treated with EGF for 3 hrs, and cell lysates were collected for HIF-1! protein analysis. SKOV3 cells were co-transfected with HRE-luciferase construct and either SPRY4

overexpression vector or control vector. After treatment with EGF for 24 hrs, cell lysates were collected for luciferase assay. B. SKOV3 cells were transfected with the SPRY4 overexpression vector and treated with EGF. After a 3-hr treatment, RNA was collected and VEGF mRNA level was assayed by real-time PCR. Relative activity levels are expressed as a percentage of the control. The data are shown as the mean SD of three independent experiments. Mean values that are not denoted by the same letter are significantly different.

100

Figure 4.6 SPRY4 knockdown enhanced EGF effect on HIF-1!. A, OVCAR3 cells were transfected with 50 nM siCtrl or siSPRY4 for 48 hrs. Transfected cells were treated with 10 ng/ml EGF for 30 min. Levels of HIF-1!, PTEN, and pAKT were then detected by Western blot. Quantification of pAKT (B) and HIF-1! (C) levels. The relative levels are expressed as a percentage of the EGF-treated siCtrl cells. The data are shown as the mean SD of three independent experiments. Means not denoted by the same letter are significantly different. Open bars: siCtrl transfection; filled bars: siSPRY4 transfection.

101

Figure 4.7 The PI3K/AKT pathway mediated HIF-1! regulation by EGF and SPRY4. A, OVCAR3 cells were transfected with 50 nM siCtrl or siSPRY4 for 48 hrs. Transfected cells pre-treated with PI3K/AKT inhibitors (Wortmannin, 1 M and LY294002, 25 M) for 30 min and then treated with 10 ng/ml EGF for 30 min. HIF-1! and pAKT were detected by Western blot. B, Quantification of HIF-1! levels. The relative levels are expressed as a percentage of the EGF-treated siCtrl cells. The data are shown as the mean SD of three independent experiments. Open bars: siCtrl transfection; filled bars: siSPRY4 transfection.

102

Chapter 5 Hypoxia-inducible factor-1 alpha destabilization by Sprouty4, independent of prolyl hydroxylases activity 5.1 Introduction Hypoxia-inducible factors (HIFs) regulate O2-homeostasis during both physiological and pathological processes (163). HIFs are heterodimeric transcription factors composed of a constitutively expressed nuclear protein HIF-1! and one of the three closely related forms of the " subunit (HIF-1", HIF-2" and HIF-3"). The activity of HIF-1 is tightly controlled through the availability of the " subunit, as the half-life of HIF-1" is extremely short in normoxia (165). Under well-oxygenated conditions, the oxygendependent degradation (ODD) domain of HIF-1" is hydroxylated by prolyl hydroxylases (PHDs) and allows binding of the Von Hippel-Lindau (VHL) E3 ligase, which ubiquitinates HIF-1" and targets HIF-1" for proteasomal degradation (166). There are three PHDs: PHD1 and PHD3 hydroxylate HIF-1" in hypoxia, whereas PHD2 is the main PHD responsible for normoxic hydroxylation (291). As the enzymatic activities of PHDs require O2 as a co-substrate, under hypoxia, PHDs are inactive and hydroxylation of HIF-1" is suppressed; HIF-1" is thereby stabilized, which then translocates into the nucleus and dimerizes with HIF-1! to constitue HIF-1. The HIF-1 transcription factor binds to hypoxia-responsive element (HRE) and induces gene transcription. Moreover, HIF-1" regulation is further complicated by the SIAH E3 ubiquitin ligases. SIAH ligases are responsible for the degradations of several proteins, including PHDs. Overexpression of SIAH2 decreased cellular PHD levels, thereby allowing the stabilization and accumulation of HIF-1" (169). 103

In addition to tissue hypoxia, HIF-1! levels and activity have been shown to be modulated by several genetic (197), environmental (211, 282) and hormonal factors (174, 283), including growth factors (188, 191). Recently, our laboratory has reported that HIF1! is regulated by epidermal growth factor (EGF) and the EGFR regulator, Sprouty4 (SPRY4) (So et al. unpublished). As regulators of EGFR and other receptor tyrosine kinases (RTKs), SPRY proteins function as tumor-suppressor genes and have been implicated in many tumorigenic processes, including angiogenesis (292). Direct evidence of SPRY involvement in angiogenesis comes from a study in which tumor cells were transplanted into Spry4knockout mice and found to grow much faster and be associated with enhanced neovascularisation compared to those in wild-type mice (274). Moreover, SPRY4 overexpression in embryonic endothelial cells was shown to minimise branching and sprouting of small vessels (135). Although SPRY4 and HIF-1! may be functionally linked, the impact of SPRY4 on the HIF-1! pathway has not been clearly elucidated. Moreover, SPRY4 has been recently demonstrated to interact with SIAH2 (287), which opens the possibility that, through interaction with SIAH, SPRY4 regulates HIF-1! hydroxylation and, therefore, HIF-1! degradation. We have demonstrated in this report that SPRY4 negatively regulates HIF-1! levels, as HIF-1! levels were increased in SPRY4 siRNA-treated cells. Accordingly, HIF-1 activity and levels of HIF-1 target genes were modulated by SPRY4. Without altering

104

HIF-1! mRNA levels, SPRY4 was found to increase the half-life of HIF-1! protein, in a mechanism independent of PHD activity and HIF-1! hydroxylation.

5.2 Materials and methods 5.2.1 Cell culture and reagents SKOV3, OVCAR3 and OVCAR4 ovarian cancer cell lines were obtained from American Type Culture Collection (Manassas, VA). The cell lines were cultured in MCDB 105/M199 (1:1), supplemented with 5% heat-inactivated fetal bovine serum, 100 IU/ml penicillin and 100 g/ml streptomycin. The cells were cultured at 37C and 5% CO2. Fetal bovine serum was purchased from Hyclone Laboratories (Logan, UT). Human recombinant EGF, cyclohexamide (CHX), MG132 and other tissue culture materials were obtained from Sigma Chemical Co. (St. Louis, MO) unless otherwise stated. 5.2.2 Transfection The FLAG-SPRY4 construct and the empty pXJ40-FLAG vector were generous gifts from Dr. Graeme R. Guy (Institute of Molecular and Cell Biology, Singapore). The hypoxia responsive element-luciferase reporter construct, ODD-domain luciferase fusion construct, PHD2 expression vector, HIF-1! expression vector (WT HIF-1!) and doublemutant HIF-1! (Pro402Ala and Pro564Ala) (DM HIF-1!) expression vector were

purchased from Addgene (Cambridge, MA). Plasmids were transfected using Lipofectamine 2000 (Invitrogen, Carlsbad, CA). ON-TARGETplus SMARTpool 105

SPRY4 siRNA and non-targeting siRNA (Dharmacon, Lafayette, CO) were transfected using Lipofectamine RNAiMAX (Invitrogen). 5.2.3 Real-time PCR Total RNA extracted from cells using TRIzoL Reagent (Invitrogen) was used in first-strand DNA (cDNA) synthesis using Invitrogen Super-ScriptTM first-strand synthesis system for real-time PCR according to the manufacturers protocol. Real-time PCR was performed using an ABI 7300 real-time thermal cycler (ABI, Hercules, CA). The detections of HIF-1!, PAI-1, PHD3 and the internal control, GAPDH, were performed as follows: a 3-min hot start at 95C followed by 40 cycles of denaturation at 95C for 15 sec, and amplification at 60C for 1 min. PCR reactions were performed in duplicate with the following PCR primers: HIF-1!, 5-

TCATCCAAGAAGCCCTAACG-3 and reverse 5-TCGCTTTCTCTGAGCATTCTGC3; PAI-1 forward 5-GGACAGACCCTTCCTCTTTGT-3 and reverse 5-

TCCATCACTTGGCCCATGAA-3; PHD3, forward 5-ATCAGCTTCCTCCTGTCCC3 and reverse 5-CAGCGACCATCACCGTTG-3; and GAPDH, forward 5GAGTCAACGGATTTGGTCGT-3 and reverse 5- GACAAGCTTCCCGTTCTCAG3. 5.2.4 Western blot analysis Equal amounts of total cell lysates were resolved on 7.5% SDS-PAGE gels and electrotransferred to a PVDF membrane. After blocking for 1 hr with 5% non-fat dry milk in TBS-T buffer (20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.1% Tween-20), the blots were probed for overnight at 4 C with primary antibodies. The HIF-1! antibody 106

was purchased from BD Transduction Laboratories (Lexington, KY), the anti-SPRY4 antibody was obtained from Abcam (Cambridge, MA), and the anti-!-actin antibody was obtained from Santa Cruz Biotechnology (Santa Cruz, CA). The anti-hydroxylated-HIF1! (Pro402) antibody was obtained from Bethyl Laboratories, Inc. (Montgomery, TX). The anti-hydroxylated-HIF-1! (Pro564) was purchased from Cell Signaling, Inc. (Austin, TX). The blots were then incubated with a peroxidase-conjugated secondary antibody (Bio-Rad) for 1 hr, followed by detection with ECL chemiluminescence reagent (Amersham, Arlington Heights, IL) and exposure to X-ray films. 5.2.5 Statistical analysis For real-time PCR data, the relative quantification of levels was calculated using the 2"" Ct method. Data are presented as the mean SD of three independent experiments or triplicates in a representative experiment and were analyzed by one-way ANOVA followed by Tukeys post-hoc test using GraphPad Prism 5 (GraphPad Software, San Diego, CA) to compare all pairs of columns. Columns are not denoted by the same letter are statistically different. Means not denoted by the same letter are significantly different (P < 0.05).

5.3 Results 5.3.1 SPRY4 negatively regulated HIF-1! expression levels in ovarian cancer cells To test the effect of SPRY4 on HIF-1! levels, we used either siRNA knockdown or overexpression to manipulate SPRY4 levels in the cells. SPRY4 siRNA effectively 107

depleted endogenous SPRY4 protein and caused an increase in HIF-1! levels in the three cell lines tested, with OVCAR4 displaying the most significant response (Fig. 5.1A). Consistent with the siRNA experiment, SPRY4 overexpression decreased the HIF-1! content in the cell lines (Fig. 5.1B). Furthermore, we found that the effect was specific for SPRY4, as HIF-1! expression levels were not increased in cells treated with SPRY2 siRNA (Fig. 5.1C). Among the three cell lines, OVCAR4 displays the highest levels of endogenous SPRY4 expression and clearly responds to SPRY4 siRNA; therefore, OVACR4 was used for most of the subsequent experiments. 5.3.2 SPRY4 negatively regulated HIF-1 activity We next investigated whether the increase in HIF-1! expression levels caused by SPRY4 siRNA would lead to a parallel change in HIF-1 activity. When co-transfected with HRE-luciferase reporter construct, SPRY4 siRNA caused an increase in HRE-driven luciferase activity (OVCAR4: Fig. 5.2A; SKOV3 and OVCAR3: data not shown). Accordingly, endogenous mRNA expression of two HIF transcriptional targets, plasminogen activator inhibitor-1 (PAI-1) (293) and prolyl hydroxylase-3 (PHD3) (294), were increased by SPRY4 siRNA when compared with the control (Fig. 5.2B). To confirm that these increases were mediated by elevated HIF-1 activity, we reduced HIF1! expression with sodium nitroprusside (SNP) ((295) and Fig. 5.2C) and found that SNP reversed the positive impacts of SPRY4 siRNA on PAI-1 and PHD3 mRNA levels (Fig. 5.2C, D).

108

5.3.3 SPRY4 regulated HIF-1! protein half-life without affecting Hif-1! mRNA levels HIF-1! is regulated at the transcriptional, translational and post-translational levels. We measured HIF-1! mRNA levels in OVCAR4 cells transfected with SPRY4 siRNA for 24 or 48 hrs and found that depleting SPRY4 had no effect on HIF-1! mRNA levels (Fig. 5.3A). Next, we tested the possibility that SPRY4 siRNA increased HIF-1! levels by inhibiting its degradation. After treating cells with the proteasomal inhibitor, MG132, HIF-1! accumulated in the cells, and the change in HIF-1! levels induced by SPRY4 siRNA was completely blocked (Fig. 5.3B). To show that SPRY4 exerts its effect via altering HIF-1! protein levels, HIF-1! was overexpressed in OVCAR4 cells, and the levels were assayed by Western blot. We found that SPRY4 siRNA causes an accumulation of overexpressed HIF-1! as effectively as the endogenous HIF-1! (Fig. 5.3C). To test the effect of SPRY4 on the half-life of HIF-1! protein, the cells were treated with protein synthesis inhibitor cycloheximide (CHX). CHX treatment caused a rapid and significant reduction in HIF-1! levels, but the rate of reduction in SPRY4 siRNA-treated cells was more moderate than that in the control (Fig. 5.3D) 5.3.4 HIF-1! modulation by SPRY4 was independent of PHD activity We also tested the hypothesis that SPRY4 regulates HIF-1! by modulating PHDmediated hydroxylation activity. We monitored the changes in PHD levels in response to SPRY4. Contrasting its HIF-1!"stabilizing effect, PHD levels were not reduced; instead, SPRY4 siRNA increased PHD2 and PHD3 levels slightly. HIF-1! siRNA reversed the effect of SPRY4 siRNA (Fig. 5.4A). PHD2 hydroxylates HIF-1! in normoxia (291), but 109

PHD2 overexpression was unable to reverse SPRY4 siRNA-induced HIF-1! levels (Fig. 5.4B). Two known HIF-1! activators, CoCl2 and DMOG, completely abrogated hydroxylation of Pro402 and Pro564 (Fig. 5.4C). In contrast, SPRY4 knockdown had no effect on the levels of hydroxylated HIF-1! protein (Fig. 5.4C). Plasmids for wild-type HIF-1! (WT HIF-1!) or a double mutant with Pro402Ala and Pro564Ala substitutions, which is resistant to PHD-mediated degradation (DM HIF-1!), were co-transfected with SPRY4 siRNA in OVCAR4 cells. Treatment with SPRY4 siRNA increased the levels of both wild-type and mutant HIF-1! (Fig. 5.4D). Finally, CoCl2 was capable of stabilizing the fusion of the HIF-1! ODD domain and a luciferase reporter gene, leading to an increase in luciferase activity (Fig. 5.4E). SPRY4 knockdown failed to act via the HIF-1! ODD domain-induced increase in luciferase levels (Fig. 5.4E).

5.4 Discussion Previously, we demonstrated that SPRY4 negatively regulates HIF-1! (So et al., unpublished). As the interaction between SPRY4 and a PHD-degrading E3-ligase, SIAH, was recently reported (287), we tested the hypothesis that SPRY4 regulates HIF-1! by modulating PHD levels and/or activity. We demonstrated that SPRY4 plays a negative role in HIF-1! expression and HIF activity (Fig. 5.1, 5.2). Next, we demonstrated that SPRY4 also functions on overexpressed HIF-1! (Fig. 5.3C), while HIF-1! mRNA levels were not affected (Fig. 5.3A), indicating that SPRY4 directly acts at the HIF-1! protein and independent of HIF-1! mRNA. Furthermore, the effect of SPRY4 is suggested to involve protein degradation as the effect of SPRY4 was less obvious when protein 110

degradation is blocked (Fig. 5.3B). Finally, we demonstrated a direct effect of SPRY4 siRNA on prolonging HIF-1! protein half-life (Fig. 5.3D). HIF-1! stabilization by SPRY4 knockdown is an additional example of a protein whose stability is influenced by SPRY. A well-elucidated example of SPRY-regulated protein degradation is its involvement in abrogating Cbl-mediated EGFR degradation. Cbl proteins are E3 ubiquitin ligases, which recognise ubiquitinate and target RTKs to degradation (139-141). Through direct interaction with the Cbl RING finger domain (142, 143), SPRY2 is capable of sequestering Cbl and thus protects EGFR from degradation (144). Similarly, SPRY1, 2 and 4 were found to interact with the RING finger domain of another E3 ligase, SIAH2 (287), thereby suggesting that SPRY may interfere with the degradation of SIAH targets, including PHDs (169). This observation has prompted us to test this possibility. However, our results clearly excluded the involvement of PHD activity in SPRY4regulated HIF-1! stability. First, reduced PHD levels and activity are anticipated to stabilize HIF-1!; but, SPRY4 siRNA increased PHD2 and PHD3 levels (Fig. 5.4A), and PHD2 overexpression failed to abolish SPRY4 activity (Fig. 5.4B). In parallel, the hydroxylation status of Pro402 and Pro564 were not altered by SPRY4 siRNA treatment (Fig. 5.4C). SPRY4 had similar effects on wild-type and mutant HIF-1! with proline substitutions (Fig. 5.4D). SPRY4 knockdown was incapable of stabilizing the ODDfusion protein (Fig. 5.4E). This finding confirms that the ODD domain is not a SPRY4 target. These results suggest that SPRY4 regulates HIF via a yet-to-be determined mechanism that is independent of the hydroxylation of the ODD domain. Moreover, the increase in PHD2 and PHD3 levels induced by SPRY4 knockdown provides additional 111

evidence of elevated HIF-1 activity, as both are known HIF targets (169). PHD2 and PHD3 levels were reduced by HIF-1! siRNA (Fig. 5.4A). The negative impact on HIF-1! exerted by SPRY4 together with the known SPRY4 induction by hypoxia and hypoxia mimetic (281) suggests reciprocal regulation between SPRY4 and HIF-1! and constitutes a novel negative feedback loop operating in cancer cells. Moreover, SPRY2 expression in colon cancer cells was repressed by 1,25(OH)2D3 (an active vitamin D metabolite) via an E-cadherin-dependent pathway, and SPRY2 in turn repressed 1,25(OH)2D3-induced E-cadherin expression (286). These examples broaden SPRY activity to signalling pathways other than the RTK/ERK pathway. Therefore, the factors or dynamics of the microenvironment that induce SPRY would trigger a counter regulatory loop by SPRY. Stabilization and nuclear localisation of HIF-1! are common in most cancers, including ovarian cancer (215, 216). Furthermore, HIF-1! overexpression is significantly correlated with enhanced microvessel density in ovarian tumors (217). The negative involvement of SPRY4 on HIF-1! levels demonstrated here may provide explanations for the accumulation of HIF-1! protein and the antiangiogenic role of SPRY4 reported previously (274). Our laboratory has previously shown decreased SPRY4 expression in ovarian cancer cell lines (So et al., unpublished), which possibly contributed to HIF-1! accumulation and an angiogenic switch in tumors and loss of SPRY4 may play a crucial role during ovarian cancer progression.

112

Figure 5.1 SPRY4 negatively regulates HIF-1! levels in ovarian cancer cells. OVCAR4, OVCAR3 and SKOV3 cells were transfected with 50 nM of control siCtrl or siSPRY4 (A) or with 50 nM of the SPRY4 overexpression vector (B). Transfected cells were cultured for 48 hrs. An equal amount of total protein was loaded in each gel lane, and HIF-1! was detected using Western blot analysis. C. Prior to HIF-1! detection using Western blot analysis, OVCAR4 cells were transfected for 48 hrs with siCtrl, siSPRY2 or siSPRY4. 113

Figure 5.2 SPRY4 negatively regulated HIF-1 activity. A. OVCAR4 cells were cotransfected with an HRE-luciferase construct and 50 nM siCtrl or siSPRY4. After 48 hrs, cell lysates were collected for luciferase assay. B. OVCAR4 cells were transfected with SPRY4 siRNA for 48 hrs. RNA was collected and PAI-1 and PHD3 mRNA levels were assayed using real-time PCR. C. OVCAR4 cells were transfected with siCtrl or siSPRY4 and incubated with the HIF-1! inhibitor SNP (500 M) for 48 hrs, and HIF-1! (C) and PAI-1 and PHD3 mRNA levels were then assayed. (D). Relative mRNA levels are expressed as a percentage of the control. The data are shown as the mean SD of triplicates in a representative experiment (A) or three independent experiments (B, D). Mean values that are not denoted by the same letter are significantly different. **, P < 0.01; ***, P < 0.005.

114

115

Figure 5.3 SPRY4 regulated HIF-1! protein half-life without affecting Hif-1! mRNA levels. A. OVCAR4 cells were transfected with 50 nM siCtrl or siSPRY4 for 48 hrs. RNA was collected and HIF-1! mRNA levels were assayed. B. OVCAR4 cells were transfected with siCtrl or siSPRY4 for 48 hrs. MG132 (5 M) was added 8 hrs before harvesting protein for Western blot analysis. C. OVCAR4 cells were co-transfected with an HIF-1! overexpression construct and siCtrl or siSPRY4. After 48 hrs, cell lysates were collected for HIF-1! detection analysis. *, P < 0.05. D. OVCAR4 cells were transfected with SPRY4 siRNA for 48 hrs. CHX (1 g/ml) was added, and cell lysates were then collected at various time points for HIF-1! protein level analysis. Relative levels were expressed as a percentage of the control. The data are shown as the mean SD of three independent experiments. Mean values that are not denoted by the same letter are significantly different.

116

117

Figure 5.4 HIF-1! modulation by SPRY4 acts independently of PHD activity. A. OVCAR4 cells were transfected with 50 nM siCtrl or siSPRY4 for 48 hrs and assayed for PHD protein levels using Western blot. B. OVCAR4 cells were co-transfected with a PHD2 overexpression construct and siSPRY4. After 48 hrs, cell lysates were collected for detection of HIF-1! levels using Western blot. C. OVCAR4 cells were treated with CoCl2 (100 m) or DMOG (1 mM) in the presence of MG132 (5 M) for 8 hrs (left panel), or OVCAR4 cells were transfected with siCtrl or siSPRY4 for 48 hrs and treated with MG132 (5 M) for 8 hrs (right panel). Protein was collected and assessed for hydroxylate HIF-1! at Pro402 or Pro564 using Western blot. D. OVCAR4 cells were cotransfected with an empty vector (Ctrl), an overexpression construct for wild-type HIF1! (WT HIF-1! ) or a double-mutant HIF-1! (DM HIF-1!) and siCtrl or siSPRY4. After 48 hrs, cell lysates were collected for the detection of HIF-1! using Western blot. The data are shown as the mean SD of three independent experiments. *, P < 0.05. E. OVCAR4 cells were co-transfected with the ODD-luciferase construct and siCtrl or siSPRY4 for 48 hrs. Cell lysates were collected for the luciferase activity assay. Relative activity levels are expressed as a percentage of the control. The data are shown as the mean SD of triplicates in a representative experiment. Mean values that are not denoted by the same letter are significantly different.

118

Chapter 6 Conclusions and future directions Summary and significance of research findings Although epithelial ovarian cancer (EOC) comprises the majority of ovarian carcinomas, its origin and etiology have yet to be completely elucidated. The evidence to date suggests that malfunctions of receptor tyrosine kinases (RTKs), including the epidermal growth factor receptor (EGFR), contribute to the development of EOC (52). Indeed, the EGFR and its ligands play critical roles in cell proliferation and survival as well as tumor metastasis (53, 242). This study focuses on the intracellular EGFR regulator Sprouty (SPRY) as well as the EGFR ligand amphiregulin (AREG) in EOC. Aberrant EGFR activity is common in ovarian cancer and can arise from EGFR amplification or activating mutations, or overexpression of EGFR or its ligands (65, 66, 75, 77, 88, 89, 91, 92). In addition to these abnormalities, loss of endogenous regulators is an alternative mechanism that leads to aberrant EGFR activity. SPRY proteins have been found to be deregulated in most malignancies tested (151-153, 155, 156, 160, 161), and SPRY-defective cells are hypersensitive to mitogenic and metastatic signals. Furthermore, SPRY proteins have been reported to regulate various aspects of tumorigenesis (148, 151, 155, 160-162). When I examined SPRY level abnormalities in ovarian cancer, SPRY2 downregulation and SPRY2 deletion were detected in high-grade serous tumors. This observation has been confirmed with the TCGA database, which contains data from more than 500 serous cystadenocarcinomas.

119

Next, I extended my research to investigate the antitumoral functions of SPRY2. Based on the positive correlation between endogenous SPRY2 and E-cadherin levels in cell lines and high-grade serous tumors, I hypothesised that SPRY2 may regulate invasiveness through modulating E-cadherin. My data showed that expression of SPRY2 antagonized the effects of EGF on E-cadherin protein repression and cell invasion (Fig. 6.1). The following studies demonstrated that depleting SPRY4 enhances E-cadherin downregulation and invasion induced by another EGFR ligand, AREG (Fig. 6.1). These findings are in agreement with previous reports on the antiinvasive role of SPRY (148, 159-161) and add ovarian cancer, an important and lethal gynaecologic cancer, to the list of malignancies with SPRY deficiency. Any mechanism that causes the downregulation of the SPRY level would result in increased stimulation from growth factors, which favour transformation and the progression of cancer, even in circumstances without ligand overproduction or receptor dysfunction. Though there are numerous reports supporting a stimulatory role of EGFR ligands in ovarian cancer cell invasion and metastasis, most of those studies have focused on EGF, transforming growth factor-! and heparin binding-EGF (84, 99, 108, 110, 111, 296). Few reports on the effects of AREG have been published. Thus, the effects of AREG on invasion and the mechanisms of action were investigated. AREG downregulates E-cadherin and promotes invasion (Fig. 6.1). In addition, AREG induces SPRY4 expression and activates the negative feedback action of SPRY4 (Fig. 6.1). Although it is well known that expression of SPRY is induced by RTK/RAS/ERK and that SPRY feedbacks to antagonize the pathway, many aspects regarding the functions of SPRY remain unclear. Specifically, it remains to be determined how SPRY 120

regulates ERK activation (123-125, 145). Second, the paradox of both the positive and negative effects of SPRY2 on EGFR signalling needs to be clarified (121, 297). Third, the initial paradigm that SPRY is specific for ERK has recently been challenged. Emerging experimental data show that SPRY is regulated by and modulates more diverse signalling pathways (133, 134, 298). For example, Wnt signalling induces SPRY4 expression (159), FGF induces SPRY1 and SPRY2 expression through a Ca2+-dependent pathway (284), and Xenopus SPRY inhibits calcium signalling (285). Clearly, more studies must be completed to obtain a better understanding of SPRY function. As SPRY4 is regulated by HIF-1! and SPRY4 regulates angiogenesis (274), SPRY4 and HIF-1! may also constitute a feedback loop that does not involve the ERK pathway. This hypothesis is strengthened by the existence of a feedback loop comprising an active vitamin D metabolite (1,25(OH)2D3) and SPRY2 in colon cancer cells (286). Together, our data demonstrate the presence and functionality of a novel EGFR/AKT/HIF-1! and SPRY4 feedback loop in ovarian cancer cells, in which EGF induces SPRY4 expression through an AKT- and HIF-1!-dependent mechanism (Fig. 6.1). In turn, SPRY4 decreases AKT activation to antagonize EGF-induced expression of HIF-1! (Fig. 6.1). SPRY2 regulates AKT activity through modulating the expression and activity of phosphatase and tensin homolog deleted on chromosome 10 (PTEN) (134). However, the mechanism by which SPRY4 antagonizes AKT and HIF-1! remains unknown. These findings not only identify a SPRY target outside of the ERK pathways but also demonstrate the negative effects of SPRY on AKT and HIF-1!. Furthermore, as AKT and HIF-1! are oncogenic in ovarian cancer (215-217), the negative nature of SPRY4 on AKT and HIF-1! suggests that SPRY4 is a tumor suppressor. 121

Lastly, I tested whether SPRY4 regulates HIF-1! through the PHD-mediated HIF-1! degradation pathway. SPRY4 physically interacts with SIAH2 (287), and SIAH proteins are responsible for PHD degradation and thus prevent HIF-1! hydroxylation by PHD. As this mechanism causes HIF-1! stabilization and accumulation (169), I hypothesised that SPRY4 may sequester SIAH and prevent PHD degradation, which would allow PHDs to hydroxylate HIF-1! and lead to HIF-1! degradation. However, my data disproved this hypothesis and showed the following: 1) SPRY4 knockdown stabilizes HIF-1! without reducing PHD levels, 2) PHD2 overexpression fails to reverse HIF-1! accumulation induced by SPRY4 knockdown, 3) SPRY4 knockdown does not alter the level of hydroxylated HIF-1!, 4) SPRY4 knockdown stabilizes a HIF-1! mutant that is refractory to PHD-mediated degradation and 5) SPRY4 knockdown is unable to stabilize a fusion protein containing the PHD-targeted domain. Potential applications of the research findings The current project aims to investigate the expression, functions and mechanisms of the action of SPRY isoforms in ovarian cancer. These studies enhance our understanding of the pathogenesis of ovarian cancer. Furthermore, due to their critical roles in regulating RTK signallings and the downstream processes, SPRY isoforms may be attractive targets for drug intervention or gene therapy in the treatment of ovarian cancer. Ovarian cancer is the most lethal of all the gynaecologic malignancies, with a 5year survival rate of 30% 40% (3), and overall survival has not improved for decades (4). One of the pitfalls in the current management of ovarian cancer is that although 122

ovarian carcinomas of different subtypes are distinct diseases caused by different pathways that respond differently to therapies, the current treatment protocols are not subtype specific (22). Ovarian cancer treatment should be subtype specific or individualized, which has been proposed for other cancers (299, 300). To achieve successful subtype-specific treatment, accurate, consistent and reproducible classification is critical. Pathologists encounter challenges in the diagnosis of the tumors, especially between high-grade serous carcinoma (HGSC) and endometrioid carcinoma (EC), clear cell carcinoma (CCC) or low-grade serous carcinoma (LGSC) (22). Therefore, the discovery of reliable biomarkers for the differential diagnosis is necessary. The distinct SPRY2 mRNA level and SPRY2 deletion patterns across various subtypes suggest the potential of SPRY2 as a diagnostic marker for identifying HGSC. The majority of HGSC samples show reduced SPRY2 mRNA level, and the mean SPRY2 mRNA level of HGSCs is statistically significantly lower than those of ECs and LGSCs. Furthermore, deletions are almost exclusively found in HGSC, and neither EC, CCC nor LGSC exhibit loss of the SPRY2 gene. These observations suggest that SPRY2 loss is important and likely specific to HGSC tumorigenesis. The potential of SPRY2 as a marker is further supported by the high incidence (more than 1/3 of cases) of SPRY2 mRNA level reduction among HGSCs. These studies need to be performed on a larger number of ovarian cancer cases or tested in combination with other known HGSCspecific biomarkers such as WT-1, which is detected in 75% of HGSC (24), to increase the reliability of diagnosis. In breast cancer patients, the association between low SPRY2 level and higher pathological grade suggests that SPRY2 is a significant independent prognostic factor (301). SPRY4 mRNA level has been identified as a reliable response 123

marker to Imatinib (c-KIT inhibitor) treatment in patients with gastrointestinal stromal tumors (GIST) (150). In addition to potential diagnostic and prognostic applications, SPRY proteins may be good molecular targets for therapeutic intervention in cancers (286, 292). In accordance with many other reports, this study demonstrates the altered expression and tumor suppressing effects of SPRY in ovarian cancer, making SPRY candidate targets for ovarian cancer treatment. It is important to note that in addition to the antagonistic effects on the EGFR (activated by EGF and AREG), the regulation of SPRY by epiregulin (binds EGFR and ERBB4), hepatocyte growth factor and fibroblast growth factor-2 (data not shown) suggest the functional interaction of SPRY with these RTKs. Furthermore, SPRY proteins have been reported to antagonize vascular endothelial growth factor (137, 274, 302) and glial cell line-derived neurotrophic factor (303). As most of these RTKs are implicated in ovarian cancer development (52) and are targets for therapy (59, 304), SPRY-based therapy may be superior to treatments that target a single RTK. Downstream of RTKs, SPRY may act by sequestering RAS, RAF and Src (305) (123-125). The PI3K/AKT pathway has been shown to be regulated by SPRY (133, 134), and the current study confirms this observation. This study is also the first report on SPRY regulation of HIF-1!. All of the SPRY target molecules and related pathways identified thus far are implicated in oncogenic processes. Therefore, it is logical to hypothesize that SPRY-targeted therapies could have potent antitumoral activity against many processes simultaneously. SPRY members have been shown to negatively regulate tumorigenic processes including proliferation (134, 151, 152, 155, 160), migration (148, 160, 161), invasion (148), cell cycle (148), apoptosis (306), angiogenesis (135, 272), in 124

vitro (148) and in vivo (151) tumorigenesis and metastasis (162). Furthermore, although the roles played by SPRY remain unclear, SPRY2 and SPRY4 have been shown to reflect breast and GIST cancer patient responses to therapies targeting HER2 and c-KIT, respectively (150, 301). This finding opens the possibility of enhancing patient sensitivity to chemotherapy through modulating SPRY. Future work: regulation of E-cadherin by SPRY E-cadherin serves to maintain intercellular contact, and loss of E-cadherin in adhesion junctions is a prerequisite for cancer metastasis. E-cadherin can be silenced in ovarian cancer through epigenetic silencing (promoter hypermethylation) (230, 254) and direct transcriptional repression (by SNAIL and SLUG) (235, 255, 256, 296). It is still unclear how SPRY2 regulates E-cadherin in ovarian cancer cells. Previously, SPRY2 was shown to regulate colon cancer cell expression of E-cadherin through modulating levels of ZEB1, a transcriptional repressor of E-cadherin (286). In contrast, I showed that SPRY4 knockdown decreases E-cadherin protein levels independent of E-cadherin mRNA level (Fig. 3.7B). Additionally, my preliminary data shows that SPRY2 increases E-cadherin protein levels but is ineffective in regulating E-cadherin mRNA. These data suggest that a SPRY mediates E-cadherin through a post-transcriptional mechanism. Regulation of protein degradation is an important mechanism of posttranscriptional regulation. It has been demonstrated that SPRY abrogates Cbl-mediated EGFR degradation. Cbl proteins are E3 ubiquitin ligases that recognise, ubiquitinate and target EGFR to endocytosis pathways and degradation (139-141). Cbl binds to the wellconserved Cbl-TKB binding site through its RING finger domain at the N-terminus of 125

SPRY2 (142, 143). Thus, SPRY2 effectively sequesters Cbl and protects the EGFR from degradation (144). Similar interactions between the RING finger domain of another E3 ligase, SIAH2, and the N-terminus of SPRY2 (287) suggest that this is a common function for SPRY2. For instance, Hakai, a ubiquitin ligase that structurally resembles Cbl (Hakai is the Japanese word for 'destruction') has been shown to interact with Ecadherin and, through ubiquitination, initiate endocytosis and the subsequent degradation of E-cadherin. Expression of Hakai in MDCK epithelial cells promotes endocytosis of Ecadherin, disrupts cell-cell contacts and enhances cell motility (307). In ovarian cancer cells, both Hakai mRNA and protein are expressed (preliminary data). One may speculate that SPRY interacts with Hakai via the RING finger domain, thereby sequestering Hakai and blocking E-cadherin degradation. As such, enhanced degradation from loss of SPRY might represent an additional mechanism whereby ovarian cancer cells lose E-cadherin, increasing the invasive potential of cancer cells. The interaction between SPRY and Hakai and their cooperation in E-cadherin degradation certainly warrant further detailed investigation. Although SPRY has been shown to interact with E3 ligases (142, 143, 287) and to regulate protein degradation (144), the roles of SPRY in the regulation of protein ubiquitination and degradation have not been examined in detail. The proper balance between ubiquitination and deubiquitination of cellular proteins is crucial for normal cell cycling and function. Indeed, many oncoproteins and tumor suppressors are involved in ubiquitination (usually as E3 ligases) or are deubiquitinating enzymes. This fact is best exemplified by the inactivating germline VHL mutation in renal clear cell carcinomas (RCCs) (179, 180), which results in the stabilization of HIF-1! (181) and a subsequent 126

excess of angiogenic signals and vascularization (181, 183). In addition, Murine double minute 2 acts as an E3 ligase of p53 and is overexpressed in many cancers, including ovarian cancer, which contributes to the reduced p53 expression (25). Notably, both HIF1! (our study) and p53 (159) are targets of SPRY4. Perturbations of the proteasome pathway result in pathogenic malignancies that affect tumor progression, drug resistance, and altered immune surveillance; thus, the proteasome is an attractive target for cancer therapies (308). Bortezomib, previously known as PS-341, is a specific inhibitor of the proteasome pathway. Bortezomib induces apoptosis (309), inhibits angiogenesis (310, 311) and increases survival of the xenograft mice model (310, 312). Bortezomib is also the first proteasome inhibitor to enter a clinical trial aimed at treating myeloma (311), which highlights the significance of the ubiquitin-proteasome pathway in cancer. Taken together, these data suggest the importance of evaluating the roles of SPRY in the modulation of the protein degradation pathway.

127

Figure 6.1 The diagram summarizes the findings. AREG promotes ovarian cancer cell invasion by reducing E-cadherin. The loss of SPRY2 and the induction of SPRY4 by EGFR ligands (EGF and AREG) have been demonstrated. SPRY2 and SPRY4 antagonize the effects of EGF and AREG on the suppression of E-cadherin and invasion stimulation. Furthermore, the AKT pathway mediates EGF induction of SPRY4. AKT and HIF-1! are identified as novel targets of SPRY4, suggesting the possibility of regulation of the downstream oncogenic pathways by SPRY4.

128

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. Lalwani N, Prasad SR, Vikram R, Shanbhogue AK, Huettner PC, Fasih N 2011 Histologic, molecular, and cytogenetic features of ovarian cancers: implications for diagnosis and treatment. Radiographics 31:625-646 Murdoch WJ 1996 Ovarian surface epithelium, ovulation and carcinogenesis. Biol Rev Camb Philos Soc 71:529-543 Auersperg N, Wong AS, Choi KC, Kang SK, Leung PC 2001 Ovarian surface epithelium: biology, endocrinology, and pathology. Endocr Rev 22:255-288 Kurman RJ, Shih Ie M 2011 Molecular pathogenesis and extraovarian origin of epithelial ovarian cancer--shifting the paradigm. Hum Pathol 42:918-931 Anglesio MS, Carey MS, Kobel M, Mackay H, Huntsman DG 2010 Clear cell carcinoma of the ovary: a report from the first Ovarian Clear Cell Symposium, June 24th, 2010. Gynecol Oncol 121:407-415 Fathalla MF 1971 Incessant ovulation--a factor in ovarian neoplasia? Lancet 2:163 Casagrande JT, Louie EW, Pike MC, Roy S, Ross RK, Henderson BE 1979 "Incessant ovulation" and ovarian cancer. Lancet 2:170-173 Cramer DW, Welch WR 1983 Determinants of ovarian cancer risk. II. Inferences regarding pathogenesis. J Natl Cancer Inst 71:717-721 Chakravarti S, Collins WP, Forecast JD, Newton JR, Oram DH, Studd JW 1976 Hormonal profiles after the menopause. Br Med J 2:784-787 Scaglia H, Medina M, Pinto-Ferreira AL, Vazques G, Gual C, Perez-Palacios G 1976 Pituitary LH and FSH secretion and responsiveness in women of old age. Acta Endocrinol (Copenh) 81:673-679 Sell A, Bertelsen K, Andersen JE, Stroyer I, Panduro J 1990 Randomized study of whole-abdomen irradiation versus pelvic irradiation plus cyclophosphamide in treatment of early ovarian cancer. Gynecol Oncol 37:367-373 Schildkraut JM, Schwingl PJ, Bastos E, Evanoff A, Hughes C 1996 Epithelial ovarian cancer risk among women with polycystic ovary syndrome. Obstet Gynecol 88:554-559 Whittemore AS, Harris R, Itnyre J 1992 Characteristics relating to ovarian cancer risk: collaborative analysis of 12 US case-control studies. IV. The pathogenesis of epithelial ovarian cancer. Collaborative Ovarian Cancer Group. Am J Epidemiol 136:1212-1220 La Vecchia C, Franceschi S 1999 Oral contraceptives and ovarian cancer. Eur J Cancer Prev 8:297-304 Shoham Z 1994 Epidemiology, etiology, and fertility drugs in ovarian epithelial carcinoma: where are we today? Fertil Steril 62:433-448 Rao BR, Slotman BJ 1991 Endocrine factors in common epithelial ovarian cancer. Endocr Rev 12:14-26 Tung CS, Mok SC, Tsang YT, Zu Z, Song H, Liu J, Deavers MT, Malpica A, Wolf JK, Lu KH, Gershenson DM, Wong KK 2009 PAX2 expression in low malignant potential ovarian tumors and low-grade ovarian serous carcinomas. Mod Pathol 22:1243-1250 129

14. 15. 16. 17.

18. 19. 20. 21.

22. 23. 24.

25.

26.

27. 28. 29.

30. 31.

Kurman RJ, Shih Ie M 2010 The origin and pathogenesis of epithelial ovarian cancer: a proposed unifying theory. Am J Surg Pathol 34:433-443 Bell DA 2005 Origins and molecular pathology of ovarian cancer. Mod Pathol 18 Suppl 2:S19-32 Piek JM, van Diest PJ, Zweemer RP, Kenemans P, Verheijen RH 2001 Tubal ligation and risk of ovarian cancer. Lancet 358:844 Wiegand KC, Shah SP, Al-Agha OM, Zhao Y, Tse K, Zeng T, Senz J, McConechy MK, Anglesio MS, Kalloger SE, Yang W, Heravi-Moussavi A, Giuliany R, Chow C, Fee J, Zayed A, Prentice L, Melnyk N, Turashvili G, Delaney AD, Madore J, Yip S, McPherson AW, Ha G, Bell L, Fereday S, Tam A, Galletta L, Tonin PN, Provencher D, Miller D, Jones SJ, Moore RA, Morin GB, Oloumi A, Boyd N, Aparicio SA, Shih Ie M, Mes-Masson AM, Bowtell DD, Hirst M, Gilks B, Marra MA, Huntsman DG 2010 ARID1A mutations in endometriosis-associated ovarian carcinomas. N Engl J Med 363:1532-1543 Gilks CB, Prat J 2009 Ovarian carcinoma pathology and genetics: recent advances. Hum Pathol 40:1213-1223 Yap TA, Carden CP, Kaye SB 2009 Beyond chemotherapy: targeted therapies in ovarian cancer. Nat Rev Cancer 9:167-181 Kobel M, Kalloger SE, Boyd N, McKinney S, Mehl E, Palmer C, Leung S, Bowen NJ, Ionescu DN, Rajput A, Prentice LM, Miller D, Santos J, Swenerton K, Gilks CB, Huntsman D 2008 Ovarian carcinoma subtypes are different diseases: implications for biomarker studies. PLoS Med 5:e232 Ahmed AA, Etemadmoghadam D, Temple J, Lynch AG, Riad M, Sharma R, Stewart C, Fereday S, Caldas C, Defazio A, Bowtell D, Brenton JD 2010 Driver mutations in TP53 are ubiquitous in high grade serous carcinoma of the ovary. J Pathol 221:49-56 Werness BA, Parvatiyar P, Ramus SJ, Whittemore AS, Garlinghouse-Jones K, Oakley-Girvan I, DiCioccio RA, Wiest J, Tsukada Y, Ponder BA, Piver MS 2000 Ovarian carcinoma in situ with germline BRCA1 mutation and loss of heterozygosity at BRCA1 and TP53. J Natl Cancer Inst 92:1088-1091 Kobel M, Reuss A, Bois A, Kommoss S, Kommoss F, Gao D, Kalloger SE, Huntsman DG, Gilks CB 2010 The biological and clinical value of p53 expression in pelvic high-grade serous carcinomas. J Pathol 222:191-198 Senturk E, Cohen S, Dottino PR, Martignetti JA 2010 A critical re-appraisal of BRCA1 methylation studies in ovarian cancer. Gynecol Oncol 119:376-383 Callahan MJ, Crum CP, Medeiros F, Kindelberger DW, Elvin JA, Garber JE, Feltmate CM, Berkowitz RS, Muto MG 2007 Primary fallopian tube malignancies in BRCA-positive women undergoing surgery for ovarian cancer risk reduction. J Clin Oncol 25:3985-3990 Khalique L, Ayhan A, Weale ME, Jacobs IJ, Ramus SJ, Gayther SA 2007 Genetic intra-tumour heterogeneity in epithelial ovarian cancer and its implications for molecular diagnosis of tumours. J Pathol 211:286-295 Piek JM, van Diest PJ, Zweemer RP, Jansen JW, Poort-Keesom RJ, Menko FH, Gille JJ, Jongsma AP, Pals G, Kenemans P, Verheijen RH 2001 Dysplastic changes in prophylactically removed Fallopian tubes of women predisposed to developing ovarian cancer. J Pathol 195:451-456 130

32.

33. 34.

35.

36.

37. 38.

39. 40.

41. 42. 43. 44. 45.

Kindelberger DW, Lee Y, Miron A, Hirsch MS, Feltmate C, Medeiros F, Callahan MJ, Garner EO, Gordon RW, Birch C, Berkowitz RS, Muto MG, Crum CP 2007 Intraepithelial carcinoma of the fimbria and pelvic serous carcinoma: Evidence for a causal relationship. Am J Surg Pathol 31:161-169 Przybycin CG, Kurman RJ, Ronnett BM, Shih Ie M, Vang R 2010 Are all pelvic (nonuterine) serous carcinomas of tubal origin? Am J Surg Pathol 34:1407-1416 Lee Y, Miron A, Drapkin R, Nucci MR, Medeiros F, Saleemuddin A, Garber J, Birch C, Mou H, Gordon RW, Cramer DW, McKeon FD, Crum CP 2007 A candidate precursor to serous carcinoma that originates in the distal fallopian tube. J Pathol 211:26-35 Marquez RT, Baggerly KA, Patterson AP, Liu J, Broaddus R, Frumovitz M, Atkinson EN, Smith DI, Hartmann L, Fishman D, Berchuck A, Whitaker R, Gershenson DM, Mills GB, Bast RC, Jr., Lu KH 2005 Patterns of gene expression in different histotypes of epithelial ovarian cancer correlate with those in normal fallopian tube, endometrium, and colon. Clin Cancer Res 11:6116-6126 Dehari R, Kurman RJ, Logani S, Shih Ie M 2007 The development of high-grade serous carcinoma from atypical proliferative (borderline) serous tumors and lowgrade micropapillary serous carcinoma: a morphologic and molecular genetic analysis. Am J Surg Pathol 31:1007-1012 Singer G, Oldt R, 3rd, Cohen Y, Wang BG, Sidransky D, Kurman RJ, Shih Ie M 2003 Mutations in BRAF and KRAS characterize the development of low-grade ovarian serous carcinoma. J Natl Cancer Inst 95:484-486 Anglesio MS, Arnold JM, George J, Tinker AV, Tothill R, Waddell N, Simms L, Locandro B, Fereday S, Traficante N, Russell P, Sharma R, Birrer MJ, deFazio A, Chenevix-Trench G, Bowtell DD 2008 Mutation of ERBB2 provides a novel alternative mechanism for the ubiquitous activation of RAS-MAPK in ovarian serous low malignant potential tumors. Mol Cancer Res 6:1678-1690 Hsu CY, Bristow R, Cha MS, Wang BG, Ho CL, Kurman RJ, Wang TL, Shih Ie M 2004 Characterization of active mitogen-activated protein kinase in ovarian serous carcinomas. Clin Cancer Res 10:6432-6436 Oniciu DC, Dasseux JL, Yang J, Mueller R, Pop E, Denysenko A, Duan C, Huang TB, Zhang L, Krause BR, Drake SL, Lalwani N, Cramer CT, Goetz B, Pape ME, McKee A, Fici GJ, Lutostanski JM, Brown SC, Bisgaier CL 2006 Influence of various central moieties on the hypolipidemic properties of long hydrocarbon chain diols and diacids. J Med Chem 49:334-348 Kobayashi H 2009 Ovarian cancer in endometriosis: epidemiology, natural history, and clinical diagnosis. Int J Clin Oncol 14:378-382 Campbell IG, Russell SE, Choong DY, Montgomery KG, Ciavarella ML, Hooi CS, Cristiano BE, Pearson RB, Phillips WA 2004 Mutation of the PIK3CA gene in ovarian and breast cancer. Cancer Res 64:7678-7681 Polakis P 2007 The many ways of Wnt in cancer. Curr Opin Genet Dev 17:45-51 Jiang X, Morland SJ, Hitchcock A, Thomas EJ, Campbell IG 1998 Allelotyping of endometriosis with adjacent ovarian carcinoma reveals evidence of a common lineage. Cancer Res 58:1707-1712 Sato N, Tsunoda H, Nishida M, Morishita Y, Takimoto Y, Kubo T, Noguchi M 2000 Loss of heterozygosity on 10q23.3 and mutation of the tumor suppressor 131

46. 47. 48.

49. 50.

51.

52. 53. 54.

55.

56. 57. 58.

gene PTEN in benign endometrial cyst of the ovary: possible sequence progression from benign endometrial cyst to endometrioid carcinoma and clear cell carcinoma of the ovary. Cancer Res 60:7052-7056 McMeekin DS, Burger RA, Manetta A, DiSaia P, Berman ML 1995 Endometrioid adenocarcinoma of the ovary and its relationship to endometriosis. Gynecol Oncol 59:81-86 Kuo KT, Mao TL, Jones S, Veras E, Ayhan A, Wang TL, Glas R, Slamon D, Velculescu VE, Kuman RJ, Shih Ie M 2009 Frequent activating mutations of PIK3CA in ovarian clear cell carcinoma. Am J Pathol 174:1597-1601 Risch HA, McLaughlin JR, Cole DE, Rosen B, Bradley L, Fan I, Tang J, Li S, Zhang S, Shaw PA, Narod SA 2006 Population BRCA1 and BRCA2 mutation frequencies and cancer penetrances: a kin-cohort study in Ontario, Canada. J Natl Cancer Inst 98:1694-1706 Hashiguchi Y, Tsuda H, Inoue T, Berkowitz RS, Mok SC 2006 PTEN expression in clear cell adenocarcinoma of the ovary. Gynecol Oncol 101:71-75 Miyazawa M, Yasuda M, Fujita M, Kajiwara H, Hirabayashi K, Takekoshi S, Hirasawa T, Murakami M, Ogane N, Kiguchi K, Ishiwata I, Mikami M, Osamura RY 2009 Therapeutic strategy targeting the mTOR-HIF-1alpha-VEGF pathway in ovarian clear cell adenocarcinoma. Pathol Int 59:19-27 Jones S, Wang TL, Shih Ie M, Mao TL, Nakayama K, Roden R, Glas R, Slamon D, Diaz LA, Jr., Vogelstein B, Kinzler KW, Velculescu VE, Papadopoulos N 2010 Frequent mutations of chromatin remodeling gene ARID1A in ovarian clear cell carcinoma. Science 330:228-231 Wong AS, Leung PC 2007 Role of endocrine and growth factors on the ovarian surface epithelium. J Obstet Gynaecol Res 33:3-16 Lafky JM, Wilken JA, Baron AT, Maihle NJ 2008 Clinical implications of the ErbB/epidermal growth factor (EGF) receptor family and its ligands in ovarian cancer. Biochim Biophys Acta 1785:232-265 Lin CR, Chen WS, Lazar CS, Carpenter CD, Gill GN, Evans RM, Rosenfeld MG 1986 Protein kinase C phosphorylation at Thr 654 of the unoccupied EGF receptor and EGF binding regulate functional receptor loss by independent mechanisms. Cell 44:839-848 Segatto O, Lonardo F, Wexler D, Fazioli F, Pierce JH, Bottaro DP, White MF, Di Fiore PP 1991 The juxtamembrane regions of the epidermal growth factor receptor and gp185erbB-2 determine the specificity of signal transduction. Mol Cell Biol 11:3191-3202 Guy PM, Platko JV, Cantley LC, Cerione RA, Carraway KL, 3rd 1994 Insect cellexpressed p180erbB3 possesses an impaired tyrosine kinase activity. Proc Natl Acad Sci U S A 91:8132-8136 Yarden Y, Sliwkowski MX 2001 Untangling the ErbB signalling network. Nat Rev Mol Cell Biol 2:127-137 Tzahar E, Waterman H, Chen X, Levkowitz G, Karunagaran D, Lavi S, Ratzkin BJ, Yarden Y 1996 A hierarchical network of interreceptor interactions determines signal transduction by Neu differentiation factor/neuregulin and epidermal growth factor. Mol Cell Biol 16:5276-5287 132

59. 60. 61.

62.

63. 64.

65. 66. 67. 68. 69. 70. 71.

72.

Sheng Q, Liu J 2011 The therapeutic potential of targeting the EGFR family in epithelial ovarian cancer. Br J Cancer 104:1241-1245 Soltoff SP, Cantley LC 1996 p120cbl is a cytosolic adapter protein that associates with phosphoinositide 3-kinase in response to epidermal growth factor in PC12 and other cells. J Biol Chem 271:563-567 Altomare DA, Wang HQ, Skele KL, De Rienzo A, Klein-Szanto AJ, Godwin AK, Testa JR 2004 AKT and mTOR phosphorylation is frequently detected in ovarian cancer and can be targeted to disrupt ovarian tumor cell growth. Oncogene 23:5853-5857 de Graeff P, Crijns AP, Ten Hoor KA, Klip HG, Hollema H, Oien K, Bartlett JM, Wisman GB, de Bock GH, de Vries EG, de Jong S, van der Zee AG 2008 The ErbB signalling pathway: protein expression and prognostic value in epithelial ovarian cancer. Br J Cancer 99:341-349 Lee S, Choi EJ, Jin C, Kim DH 2005 Activation of PI3K/Akt pathway by PTEN reduction and PIK3CA mRNA amplification contributes to cisplatin resistance in an ovarian cancer cell line. Gynecol Oncol 97:26-34 Qiu L, Di W, Jiang Q, Scheffler E, Derby S, Yang J, Kouttab N, Wanebo H, Yan B, Wan Y 2005 Targeted inhibition of transient activation of the EGFR-mediated cell survival pathway enhances paclitaxel-induced ovarian cancer cell death. Int J Oncol 27:1441-1448 Wikstrand CJ, Reist CJ, Archer GE, Zalutsky MR, Bigner DD 1998 The class III variant of the epidermal growth factor receptor (EGFRvIII): characterization and utilization as an immunotherapeutic target. J Neurovirol 4:148-158 Lassus H, Sihto H, Leminen A, Joensuu H, Isola J, Nupponen NN, Butzow R 2006 Gene amplification, mutation, and protein expression of EGFR and mutations of ERBB2 in serous ovarian carcinoma. J Mol Med 84:671-681 Skirnisdottir I, Seidal T, Karlsson MG, Sorbe B 2005 Clinical and biological characteristics of clear cell carcinomas of the ovary in FIGO stages I-II. Int J Oncol 26:177-183 Leng J, Lang J, Shen K, Guo L 1997 Overexpression of p53, EGFR, c-erbB2 and c-erbB3 in endometrioid carcinoma of the ovary. Chin Med Sci J 12:67-70 Scambia G, Benedetti Panici P, Battaglia F, Ferrandina G, Baiocchi G, Greggi S, De Vincenzo R, Mancuso S 1992 Significance of epidermal growth factor receptor in advanced ovarian cancer. J Clin Oncol 10:529-535 Harlozinska A, Bar JK, Sobanska E, Goluda M 1998 Epidermal growth factor receptor and c-erbB-2 oncoproteins in tissue and tumor effusion cells of histopathologically different ovarian neoplasms. Tumour Biol 19:364-373 Psyrri A, Kassar M, Yu Z, Bamias A, Weinberger PM, Markakis S, Kowalski D, Camp RL, Rimm DL, Dimopoulos MA 2005 Effect of epidermal growth factor receptor expression level on survival in patients with epithelial ovarian cancer. Clin Cancer Res 11:8637-8643 Khalifa MA, Lacher DA, Lage JM, Mannel RS, Walker JL, Angros LH, Min KW 1997 Immunohistochemical assessment of proliferation markers and altered gene expression in archival specimens of ovarian epithelial tumors. Cancer Detect Prev 21:532-539 133

73.

74.

75. 76. 77.

78.

79.

80.

81.

82.

83.

84.

Goff BA, Shy K, Greer BE, Muntz HG, Skelly M, Gown AM 1996 Overexpression and relationships of HER-2/neu, epidermal growth factor receptor, p53, Ki-67, and tumor necrosis factor alpha in epithelial ovarian cancer. Eur J Gynaecol Oncol 17:487-492 Schilder RJ, Sill MW, Chen X, Darcy KM, Decesare SL, Lewandowski G, Lee RB, Arciero CA, Wu H, Godwin AK 2005 Phase II study of gefitinib in patients with relapsed or persistent ovarian or primary peritoneal carcinoma and evaluation of epidermal growth factor receptor mutations and immunohistochemical expression: a Gynecologic Oncology Group Study. Clin Cancer Res 11:5539-5548 Kuan CT, Wikstrand CJ, Bigner DD 2001 EGF mutant receptor vIII as a molecular target in cancer therapy. Endocr Relat Cancer 8:83-96 Pedersen MW, Meltorn M, Damstrup L, Poulsen HS 2001 The type III epidermal growth factor receptor mutation. Biological significance and potential target for anti-cancer therapy. Ann Oncol 12:745-760 Moscatello DK, Holgado-Madruga M, Godwin AK, Ramirez G, Gunn G, Zoltick PW, Biegel JA, Hayes RL, Wong AJ 1995 Frequent expression of a mutant epidermal growth factor receptor in multiple human tumors. Cancer Res 55:55365539 Steffensen KD, Waldstrom M, Olsen DA, Corydon T, Lorentzen KA, Knudsen HJ, Jeppesen U, Brandslund I, Jakobsen A 2008 Mutant epidermal growth factor receptor in benign, borderline, and malignant ovarian tumors. Clin Cancer Res 14:3278-3282 Zeineldin R, Rosenberg M, Ortega D, Buhr C, Chavez MG, Stack MS, Kusewitt DF, Hudson LG 2006 Mesenchymal transformation in epithelial ovarian tumor cells expressing epidermal growth factor receptor variant III. Mol Carcinog 45:851-860 Tanaka Y, Terai Y, Tanabe A, Sasaki H, Sekijima T, Fujiwara S, Yamashita Y, Kanemura M, Ueda M, Sugita M, Franklin WA, Ohmichi M 2011 Prognostic effect of epidermal growth factor receptor gene mutations and the aberrant phosphorylation of Akt and ERK in ovarian cancer. Cancer Biol Ther 11:50-57 Paez JG, Janne PA, Lee JC, Tracy S, Greulich H, Gabriel S, Herman P, Kaye FJ, Lindeman N, Boggon TJ, Naoki K, Sasaki H, Fujii Y, Eck MJ, Sellers WR, Johnson BE, Meyerson M 2004 EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy. Science 304:1497-1500 Lacroix L, Pautier P, Duvillard P, Motte N, Saulnier P, Bidart JM, Soria JC 2006 Response of ovarian carcinomas to gefitinib-carboplatin-paclitaxel combination is not associated with EGFR kinase domain somatic mutations. Int J Cancer 118:1068-1069 Posadas EM, Liel MS, Kwitkowski V, Minasian L, Godwin AK, Hussain MM, Espina V, Wood BJ, Steinberg SM, Kohn EC 2007 A phase II and pharmacodynamic study of gefitinib in patients with refractory or recurrent epithelial ovarian cancer. Cancer 109:1323-1330 Cowden Dahl KD, Symowicz J, Ning Y, Gutierrez E, Fishman DA, Adley BP, Stack MS, Hudson LG 2008 Matrix metalloproteinase 9 is a mediator of 134

85. 86. 87. 88. 89.

90.

91.

92.

93.

94.

95.

96.

epidermal growth factor-dependent e-cadherin loss in ovarian carcinoma cells. Cancer Res 68:4606-4613 Harris RC, Chung E, Coffey RJ 2003 EGF receptor ligands. Exp Cell Res 284:213 Stromberg K, Johnson GR, O'Connor DM, Sorensen CM, Gullick WJ, Kannan B 1994 Frequent immunohistochemical detection of EGF supergene family members in ovarian carcinogenesis. Int J Gynecol Pathol 13:342-347 Niikura H, Sasano H, Sato S, Yajima A 1997 Expression of epidermal growth factor-related proteins and epidermal growth factor receptor in common epithelial ovarian tumors. Int J Gynecol Pathol 16:60-68 Owens OJ, Leake RE 1992 Growth factor content in normal and benign ovarian tumours. Eur J Obstet Gynecol Reprod Biol 47:223-228 Tanaka Y, Miyamoto S, Suzuki SO, Oki E, Yagi H, Sonoda K, Yamazaki A, Mizushima H, Maehara Y, Mekada E, Nakano H 2005 Clinical significance of heparin-binding epidermal growth factor-like growth factor and a disintegrin and metalloprotease 17 expression in human ovarian cancer. Clin Cancer Res 11:4783-4792 Miyamoto S, Hirata M, Yamazaki A, Kageyama T, Hasuwa H, Mizushima H, Tanaka Y, Yagi H, Sonoda K, Kai M, Kanoh H, Nakano H, Mekada E 2004 Heparin-binding EGF-like growth factor is a promising target for ovarian cancer therapy. Cancer Res 64:5720-5727 Yotsumoto F, Yagi H, Suzuki SO, Oki E, Tsujioka H, Hachisuga T, Sonoda K, Kawarabayashi T, Mekada E, Miyamoto S 2008 Validation of HB-EGF and amphiregulin as targets for human cancer therapy. Biochem Biophys Res Commun 365:555-561 Yagi H, Miyamoto S, Tanaka Y, Sonoda K, Kobayashi H, Kishikawa T, Iwamoto R, Mekada E, Nakano H 2005 Clinical significance of heparin-binding epidermal growth factor-like growth factor in peritoneal fluid of ovarian cancer. Br J Cancer 92:1737-1745 Lin CI, Chen CN, Huang MT, Lee SJ, Lin CH, Chang CC, Lee H 2008 Lysophosphatidic acid upregulates vascular endothelial growth factor-C and tube formation in human endothelial cells through LPA(1/3), COX-2, and NF-kappaB activation- and EGFR transactivation-dependent mechanisms. Cell Signal 20:1804-1814 Keely SJ, Uribe JM, Barrett KE 1998 Carbachol stimulates transactivation of epidermal growth factor receptor and mitogen-activated protein kinase in T84 cells. Implications for carbachol-stimulated chloride secretion. J Biol Chem 273:27111-27117 Ushio-Fukai M, Griendling KK, Becker PL, Hilenski L, Halleran S, Alexander RW 2001 Epidermal growth factor receptor transactivation by angiotensin II requires reactive oxygen species in vascular smooth muscle cells. Arterioscler Thromb Vasc Biol 21:489-495 Mori K, Kitayama J, Shida D, Yamashita H, Watanabe T, Nagawa H 2006 Lysophosphatidic acid-induced effects in human colon carcinoma DLD1 cells are partially dependent on transactivation of epidermal growth factor receptor. J Surg Res 132:56-61 135

97. 98. 99.

100. 101. 102. 103.

104.

105.

106.

107.

108.

109.

Bolitho C, Hahn MA, Baxter RC, Marsh DJ 2010 The chemokine CXCL1 induces proliferation in epithelial ovarian cancer cells by transactivation of the epidermal growth factor receptor. Endocr Relat Cancer 17:929-940 Abdollahi A, Gruver BN, Patriotis C, Hamilton TC 2003 Identification of epidermal growth factor-responsive genes in normal rat ovarian surface epithelial cells. Biochem Biophys Res Commun 307:188-197 Chen Z, Fadiel A, Feng Y, Ohtani K, Rutherford T, Naftolin F 2001 Ovarian epithelial carcinoma tyrosine phosphorylation, cell proliferation, and ezrin translocation are stimulated by interleukin 1alpha and epidermal growth factor. Cancer 92:3068-3075 Siemens CH, Auersperg N 1988 Serial propagation of human ovarian surface epithelium in tissue culture. J Cell Physiol 134:347-356 Stromberg K, Collins TJt, Gordon AW, Jackson CL, Johnson GR 1992 Transforming growth factor-alpha acts as an autocrine growth factor in ovarian carcinoma cell lines. Cancer Res 52:341-347 Casamassimi A, De Luca A, Agrawal S, Stromberg K, Salomon DS, Normanno N 2000 EGF-related antisense oligonucleotides inhibit the proliferation of human ovarian carcinoma cells. Ann Oncol 11:319-325 Morishige K, Kurachi H, Amemiya K, Fujita Y, Yamamoto T, Miyake A, Tanizawa O 1991 Evidence for the involvement of transforming growth factor alpha and epidermal growth factor receptor autocrine growth mechanism in primary human ovarian cancers in vitro. Cancer Res 51:5322-5328 Kurachi H, Morishige K, Amemiya K, Adachi H, Hirota K, Miyake A, Tanizawa O 1991 Importance of transforming growth factor alpha/epidermal growth factor receptor autocrine growth mechanism in an ovarian cancer cell line in vivo. Cancer Res 51:5956-5959 Johnson GR, Saeki T, Auersperg N, Gordon AW, Shoyab M, Salomon DS, Stromberg K 1991 Response to and expression of amphiregulin by ovarian carcinoma and normal ovarian surface epithelial cells: nuclear localization of endogenous amphiregulin. Biochem Biophys Res Commun 180:481-488 Alper O, De Santis ML, Stromberg K, Hacker NF, Cho-Chung YS, Salomon DS 2000 Anti-sense suppression of epidermal growth factor receptor expression alters cellular proliferation, cell-adhesion and tumorigenicity in ovarian cancer cells. Int J Cancer 88:566-574 Ahmed N, Maines-Bandiera S, Quinn MA, Unger WG, Dedhar S, Auersperg N 2006 Molecular pathways regulating EGF-induced epithelio-mesenchymal transition in human ovarian surface epithelium. Am J Physiol Cell Physiol 290:C1532-1542 Colomiere M, Findlay J, Ackland L, Ahmed N 2009 Epidermal growth factorinduced ovarian carcinoma cell migration is associated with JAK2/STAT3 signals and changes in the abundance and localization of alpha6beta1 integrin. Int J Biochem Cell Biol 41:1034-1045 Lim R, Ahmed N, Borregaard N, Riley C, Wafai R, Thompson EW, Quinn MA, Rice GE 2007 Neutrophil gelatinase-associated lipocalin (NGAL) an earlyscreening biomarker for ovarian cancer: NGAL is associated with epidermal 136

110.

111.

112. 113.

114. 115.

116. 117. 118.

119.

120. 121. 122.

growth factor-induced epithelio-mesenchymal transition. Int J Cancer 120:24262434 Henic E, Noskova V, Hoyer-Hansen G, Hansson S, Casslen B 2009 Estradiol attenuates EGF-induced rapid uPAR mobilization and cell migration via the Gprotein-coupled receptor 30 in ovarian cancer cells. Int J Gynecol Cancer 19:214222 Ellerbroek SM, Halbleib JM, Benavidez M, Warmka JK, Wattenberg EV, Stack MS, Hudson LG 2001 Phosphatidylinositol 3-kinase activity in epidermal growth factor-stimulated matrix metalloproteinase-9 production and cell surface association. Cancer Res 61:1855-1861 Alper O, Bergmann-Leitner ES, Bennett TA, Hacker NF, Stromberg K, StetlerStevenson WG 2001 Epidermal growth factor receptor signaling and the invasive phenotype of ovarian carcinoma cells. J Natl Cancer Inst 93:1375-1384 Chan JK, Pham H, You XJ, Cloven NG, Burger RA, Rose GS, Van Nostrand K, Korc M, Disaia PJ, Fan H 2005 Suppression of ovarian cancer cell tumorigenicity and evasion of Cisplatin resistance using a truncated epidermal growth factor receptor in a rat model. Cancer Res 65:3243-3248 Christen RD, Hom DK, Porter DC, Andrews PA, MacLeod CL, Hafstrom L, Howell SB 1990 Epidermal growth factor regulates the in vitro sensitivity of human ovarian carcinoma cells to cisplatin. J Clin Invest 86:1632-1640 Blank SV, Christos P, Curtin JP, Goldman N, Runowicz CD, Sparano JA, Liebes L, Chen HX, Muggia FM 2010 Erlotinib added to carboplatin and paclitaxel as first-line treatment of ovarian cancer: a phase II study based on surgical reassessment. Gynecol Oncol 119:451-456 Siwak DR, Carey M, Hennessy BT, Nguyen CT, McGahren Murray MJ, Nolden L, Mills GB 2010 Targeting the epidermal growth factor receptor in epithelial ovarian cancer: current knowledge and future challenges. J Oncol 2010:568938 Kumar A, Petri ET, Halmos B, Boggon TJ 2008 Structure and clinical relevance of the epidermal growth factor receptor in human cancer. J Clin Oncol 26:17421751 Schilder RJ, Pathak HB, Lokshin AE, Holloway RW, Alvarez RD, Aghajanian C, Min H, Devarajan K, Ross E, Drescher CW, Godwin AK 2009 Phase II trial of single agent cetuximab in patients with persistent or recurrent epithelial ovarian or primary peritoneal carcinoma with the potential for dose escalation to rash. Gynecol Oncol 113:21-27 Seiden MV, Burris HA, Matulonis U, Hall JB, Armstrong DK, Speyer J, Weber JD, Muggia F 2007 A phase II trial of EMD72000 (matuzumab), a humanized anti-EGFR monoclonal antibody, in patients with platinum-resistant ovarian and primary peritoneal malignancies. Gynecol Oncol 104:727-731 Hacohen N, Kramer S, Sutherland D, Hiromi Y, Krasnow MA 1998 Expression of Sprouty genes 1, 2 and 4 during mouse organogenesis. Cell 92:253-263 Casci T, Vinos J, Freeman M 1999 Sprouty, an intracellular inhibitor of Ras signaling. Cell 96:655-665 Minowada G, Jarvis LA, Chi CL, Neubuser A, Sun X, Hacohen N, Krasnow MA, Martin GR 1999 Vertebrate Sprouty genes are induced by FGF signaling and can cause chondrodysplasia when overexpressed. Development 126:4465-4475 137

123. 124. 125. 126. 127. 128. 129.

130. 131. 132.

133. 134. 135. 136.

137.

Reich A, Sapir A, Shilo B 1999 Sprouty is a general inhibitor of receptor tyrosine kinase signaling. Development 126:4139-4147 Gross I, Bassit B, Benezra M, Licht JD 2001 Mammalian sprouty proteins inhibit cell growth and differentiation by preventing ras activation. J Biol Chem 276:46460-46468 Yusoff P, Lao DH, Ong SH, Wong ES, Lim J, Lo TL, Leong HF, Fong CW, Guy GR 2002 Sprouty2 inhibits the Ras/MAP kinase pathway by inhibiting the activation of Raf. J Biol Chem 277:3195-3201 Mason JM, Morrison DJ, Bassit B, Dimri M, Band H, Licht JD, Gross I 2004 Tyrosine phosphorylation of Sprouty proteins regulates their ability to inhibit growth factor signaling: a dual feedback loop. Mol Biol Cell 15:2176-2188 Mason JM, Morrison DJ, Basson MA, Licht JD 2006 Sprouty proteins: multifaceted negative-feedback regulators of receptor tyrosine kinase signaling. Trends Cell Biol 16:45-54 Yigzaw Y, Cartin L, Pierre S, Scholich K, Patel TB 2001 The C terminus of sprouty is important for modulation of cellular migration and proliferation. J Biol Chem 276:22742-22747 Lim J, Wong ES, Ong SH, Yusoff P, Low BC, Guy GR 2000 Sprouty proteins are targeted to membrane ruffles upon growth factor receptor tyrosine kinase activation. Identification of a novel translocation domain. J Biol Chem 275:32837-32845 Hanafusa H, Torii S, Yasunaga T, Nishida E 2002 Sprouty1 and Sprouty2 provide a control mechanism for the Ras/MAPK signalling pathway. Nat Cell Biol 4:850858 Hall AB, Jura N, DaSilva J, Jang YJ, Gong D, Bar-Sagi D 2003 hSpry2 is targeted to the ubiquitin-dependent proteasome pathway by c-Cbl. Curr Biol 13:308-314 Tsavachidou D, Coleman ML, Athanasiadis G, Li S, Licht JD, Olson MF, Weber BL 2004 SPRY2 is an inhibitor of the ras/extracellular signal-regulated kinase pathway in melanocytes and melanoma cells with wild-type BRAF but not with the V599E mutant. Cancer Res 64:5556-5559 de Alvaro C, Martinez N, Rojas JM, Lorenzo M 2005 Sprouty-2 overexpression in C2C12 cells confers myogenic differentiation properties in the presence of FGF2. Mol Biol Cell 16:4454-4461 Edwin F, Singh R, Endersby R, Baker SJ, Patel TB 2006 The tumor suppressor PTEN is necessary for human Sprouty 2-mediated inhibition of cell proliferation. J Biol Chem 281:4816-4822 Lee SH, Schloss DJ, Jarvis L, Krasnow MA, Swain JL 2001 Inhibition of angiogenesis by a mouse sprouty protein. J Biol Chem 276:4128-4133 Leeksma OC, Van Achterberg TA, Tsumura Y, Toshima J, Eldering E, Kroes WG, Mellink C, Spaargaren M, Mizuno K, Pannekoek H, de Vries CJ 2002 Human sprouty 4, a new ras antagonist on 5q31, interacts with the dual specificity kinase TESK1. Eur J Biochem 269:2546-2556 Sasaki A, Taketomi T, Kato R, Saeki K, Nonami A, Sasaki M, Kuriyama M, Saito N, Shibuya M, Yoshimura A 2003 Mammalian Sprouty4 suppresses Rasindependent ERK activation by binding to Raf1. Nat Cell Biol 5:427-432 138

138. 139. 140. 141.

142. 143. 144.

145. 146. 147. 148. 149.

150.

151.

Tsumura Y, Toshima J, Leeksma OC, Ohashi K, Mizuno K 2005 Sprouty-4 negatively regulates cell spreading by inhibiting the kinase activity of testicular protein kinase. Biochem J 387:627-637 Miyake S, Lupher ML, Jr., Druker B, Band H 1998 The tyrosine kinase regulator Cbl enhances the ubiquitination and degradation of the platelet-derived growth factor receptor alpha. Proc Natl Acad Sci U S A 95:7927-7932 Levkowitz G, Waterman H, Zamir E, Kam Z, Oved S, Langdon WY, Beguinot L, Geiger B, Yarden Y 1998 c-Cbl/Sli-1 regulates endocytic sorting and ubiquitination of the epidermal growth factor receptor. Genes Dev 12:3663-3674 Ettenberg SA, Magnifico A, Cuello M, Nau MM, Rubinstein YR, Yarden Y, Weissman AM, Lipkowitz S 2001 Cbl-b-dependent coordinated degradation of the epidermal growth factor receptor signaling complex. J Biol Chem 276:2767727684 Guy GR, Jackson RA, Yusoff P, Chow SY 2009 Sprouty proteins: modified modulators, matchmakers or missing links? J Endocrinol 203:191-202 Wong ES, Lim J, Low BC, Chen Q, Guy GR 2001 Evidence for direct interaction between Sprouty and Cbl. J Biol Chem 276:5866-5875 Wong ES, Fong CW, Lim J, Yusoff P, Low BC, Langdon WY, Guy GR 2002 Sprouty2 attenuates epidermal growth factor receptor ubiquitylation and endocytosis, and consequently enhances Ras/ERK signalling. Embo J 21:47964808 Hacohen N, Kramer S, Sutherland D, Hiromi Y, Krasnow MA 1998 sprouty encodes a novel antagonist of FGF signaling that patterns apical branching of the Drosophila airways. Cell 92:253-263 Kramer S, Okabe M, Hacohen N, Krasnow MA, Hiromi Y 1999 Sprouty: a common antagonist of FGF and EGF signaling pathways in Drosophila. Development 126:2515-2525 Ozaki K, Kadomoto R, Asato K, Tanimura S, Itoh N, Kohno M 2001 ERK pathway positively regulates the expression of Sprouty genes. Biochem Biophys Res Commun 285:1084-1088 Lee CC, Putnam AJ, Miranti CK, Gustafson M, Wang LM, Vande Woude GF, Gao CF 2004 Overexpression of sprouty 2 inhibits HGF/SF-mediated cell growth, invasion, migration, and cytokinesis. Oncogene 23:5193-5202 Shaw AT, Meissner A, Dowdle JA, Crowley D, Magendantz M, Ouyang C, Parisi T, Rajagopal J, Blank LJ, Bronson RT, Stone JR, Tuveson DA, Jaenisch R, Jacks T 2007 Sprouty-2 regulates oncogenic K-ras in lung development and tumorigenesis. Genes Dev 21:694-707 Frolov A, Chahwan S, Ochs M, Arnoletti JP, Pan ZZ, Favorova O, Fletcher J, von Mehren M, Eisenberg B, Godwin AK 2003 Response markers and the molecular mechanisms of action of Gleevec in gastrointestinal stromal tumors. Mol Cancer Ther 2:699-709 Lo TL, Yusoff P, Fong CW, Guo K, McCaw BJ, Phillips WA, Yang H, Wong ES, Leong HF, Zeng Q, Putti TC, Guy GR 2004 The ras/mitogen-activated protein kinase pathway inhibitor and likely tumor suppressor proteins, sprouty 1 and sprouty 2 are deregulated in breast cancer. Cancer Res 64:6127-6136 139

152. 153. 154. 155.

156.

157.

158.

159.

160.

161. 162.

163. 164.

Kwabi-Addo B, Wang J, Erdem H, Vaid A, Castro P, Ayala G, Ittmann M 2004 The expression of Sprouty1, an inhibitor of fibroblast growth factor signal transduction, is decreased in human prostate cancer. Cancer Res 64:4728-4735 Fritzsche S, Kenzelmann M, Hoffmann MJ, Muller M, Engers R, Grone HJ, Schulz WA 2006 Concomitant down-regulation of SPRY1 and SPRY2 in prostate carcinoma. Endocr Relat Cancer 13:839-849 Velasco A, Pallares J, Santacana M, Gatius S, Fernandez M, Domingo M, Valls J, Yeramian A, Encinas M, Dolcet X, Matias-Guiu X Promoter hypermethylation and expression of sprouty 2 in endometrial carcinoma. Hum Pathol 42:185-193 Fong CW, Chua MS, McKie AB, Ling SH, Mason V, Li R, Yusoff P, Lo TL, Leung HY, So SK, Guy GR 2006 Sprouty 2, an inhibitor of mitogen-activated protein kinase signaling, is down-regulated in hepatocellular carcinoma. Cancer Res 66:2048-2058 McKie AB, Douglas DA, Olijslagers S, Graham J, Omar MM, Heer R, Gnanapragasam VJ, Robson CN, Leung HY 2005 Epigenetic inactivation of the human sprouty2 (hSPRY2) homologue in prostate cancer. Oncogene 24:21662174 Takahashi M, Rhodes DR, Furge KA, Kanayama H, Kagawa S, Haab BB, Teh BT 2001 Gene expression profiling of clear cell renal cell carcinoma: gene identification and prognostic classification. Proc Natl Acad Sci U S A 98:97549759 Calvisi DF, Ladu S, Gorden A, Farina M, Lee JS, Conner EA, Schroeder I, Factor VM, Thorgeirsson SS 2007 Mechanistic and prognostic significance of aberrant methylation in the molecular pathogenesis of human hepatocellular carcinoma. J Clin Invest 117:2713-2722 Tennis MA, Van Scoyk MM, Freeman SV, Vandervest KM, Nemenoff RA, Winn RA 2010 Sprouty-4 inhibits transformed cell growth, migration and invasion, and epithelial-mesenchymal transition, and is regulated by Wnt7A through PPARgamma in non-small cell lung cancer. Mol Cancer Res 8:833-843 Sutterluty H, Mayer CE, Setinek U, Attems J, Ovtcharov S, Mikula M, Mikulits W, Micksche M, Berger W 2007 Down-regulation of Sprouty2 in non-small cell lung cancer contributes to tumor malignancy via extracellular signal-regulated kinase pathway-dependent and -independent mechanisms. Mol Cancer Res 5:509520 Wang J, Thompson B, Ren C, Ittmann M, Kwabi-Addo B 2006 Sprouty4, a suppressor of tumor cell motility, is down regulated by DNA methylation in human prostate cancer. Prostate 66:613-624 Miyoshi K, Wakioka T, Nishinakamura H, Kamio M, Yang L, Inoue M, Hasegawa M, Yonemitsu Y, Komiya S, Yoshimura A 2004 The Sprouty-related protein, Spred, inhibits cell motility, metastasis, and Rho-mediated actin reorganization. Oncogene 23:5567-5576 Harris AL 2002 Hypoxia--a key regulatory factor in tumour growth. Nat Rev Cancer 2:38-47 Zhang Y, Li M, Yao Q, Chen C 2007 Recent advances in tumor hypoxia: tumor progression, molecular mechanisms, and therapeutic implications. Med Sci Monit 13:RA175-180 140

165. 166. 167.

168. 169.

170. 171. 172. 173.

174. 175. 176. 177. 178.

179.

Berra E, Ginouves A, Pouyssegur J 2006 The hypoxia-inducible-factor hydroxylases bring fresh air into hypoxia signalling. EMBO Rep 7:41-45 Kim W, Kaelin WG, Jr. 2003 The von Hippel-Lindau tumor suppressor protein: new insights into oxygen sensing and cancer. Curr Opin Genet Dev 13:55-60 del Peso L, Castellanos MC, Temes E, Martin-Puig S, Cuevas Y, Olmos G, Landazuri MO 2003 The von Hippel Lindau/hypoxia-inducible factor (HIF) pathway regulates the transcription of the HIF-proline hydroxylase genes in response to low oxygen. J Biol Chem 278:48690-48695 Aprelikova O, Chandramouli GV, Wood M, Vasselli JR, Riss J, Maranchie JK, Linehan WM, Barrett JC 2004 Regulation of HIF prolyl hydroxylases by hypoxiainducible factors. J Cell Biochem 92:491-501 Nakayama K, Frew IJ, Hagensen M, Skals M, Habelhah H, Bhoumik A, Kadoya T, Erdjument-Bromage H, Tempst P, Frappell PB, Bowtell DD, Ronai Z 2004 Siah2 regulates stability of prolyl-hydroxylases, controls HIF1alpha abundance, and modulates physiological responses to hypoxia. Cell 117:941-952 Nakayama K, Gazdoiu S, Abraham R, Pan ZQ, Ronai Z 2007 Hypoxia-induced assembly of prolyl hydroxylase PHD3 into complexes: implications for its activity and susceptibility for degradation by the E3 ligase Siah2. Biochem J 401:217-226 Nakayama K, Ronai Z 2004 Siah: new players in the cellular response to hypoxia. Cell Cycle 3:1345-1347 Callapina M, Zhou J, Schmid T, Kohl R, Brune B 2005 NO restores HIF-1alpha hydroxylation during hypoxia: role of reactive oxygen species. Free Radic Biol Med 39:925-936 Ueno M, Maeno T, Nomura M, Aoyagi-Ikeda K, Matsui H, Hara K, Tanaka T, Iso T, Suga T, Kurabayashi M 2011 Hypoxia-inducible factor-1alpha mediates TGFbeta-induced PAI-1 production in alveolar macrophages in pulmonary fibrosis. Am J Physiol Lung Cell Mol Physiol 300:L740-752 Spinella F, Rosano L, Del Duca M, Di Castro V, Nicotra MR, Natali PG, Bagnato A 2010 Endothelin-1 inhibits prolyl hydroxylase domain 2 to activate hypoxiainducible factor-1alpha in melanoma cells. PLoS One 5:e11241 Lando D, Peet DJ, Whelan DA, Gorman JJ, Whitelaw ML 2002 Asparagine hydroxylation of the HIF transactivation domain a hypoxic switch. Science 295:858-861 Kung AL, Wang S, Klco JM, Kaelin WG, Livingston DM 2000 Suppression of tumor growth through disruption of hypoxia-inducible transcription. Nat Med 6:1335-1340 Ivan M, Kondo K, Yang H, Kim W, Valiando J, Ohh M, Salic A, Asara JM, Lane WS, Kaelin WG, Jr. 2001 HIFalpha targeted for VHL-mediated destruction by proline hydroxylation: implications for O2 sensing. Science 292:464-468 Jaakkola P, Mole DR, Tian YM, Wilson MI, Gielbert J, Gaskell SJ, Kriegsheim A, Hebestreit HF, Mukherji M, Schofield CJ, Maxwell PH, Pugh CW, Ratcliffe PJ 2001 Targeting of HIF-alpha to the von Hippel-Lindau ubiquitylation complex by O2-regulated prolyl hydroxylation. Science 292:468-472 Neumann HP, Lips CJ, Hsia YE, Zbar B 1995 Von Hippel-Lindau syndrome. Brain Pathol 5:181-193 141

180. 181.

182.

183.

184. 185. 186. 187. 188.

189. 190.

191.

192.

Zhuang Z, Gnarra JR, Dudley CF, Zbar B, Linehan WM, Lubensky IA 1996 Detection of von Hippel-Lindau disease gene mutations in paraffin-embedded sporadic renal cell carcinoma specimens. Mod Pathol 9:838-842 Krieg M, Haas R, Brauch H, Acker T, Flamme I, Plate KH 2000 Up-regulation of hypoxia-inducible factors HIF-1alpha and HIF-2alpha under normoxic conditions in renal carcinoma cells by von Hippel-Lindau tumor suppressor gene loss of function. Oncogene 19:5435-5443 Maxwell PH, Wiesener MS, Chang GW, Clifford SC, Vaux EC, Cockman ME, Wykoff CC, Pugh CW, Maher ER, Ratcliffe PJ 1999 The tumour suppressor protein VHL targets hypoxia-inducible factors for oxygen-dependent proteolysis. Nature 399:271-275 Gnarra JR, Zhou S, Merrill MJ, Wagner JR, Krumm A, Papavassiliou E, Oldfield EH, Klausner RD, Linehan WM 1996 Post-transcriptional regulation of vascular endothelial growth factor mRNA by the product of the VHL tumor suppressor gene. Proc Natl Acad Sci U S A 93:10589-10594 Iliopoulos O, Levy AP, Jiang C, Kaelin WG, Jr., Goldberg MA 1996 Negative regulation of hypoxia-inducible genes by the von Hippel-Lindau protein. Proc Natl Acad Sci U S A 93:10595-10599 Iliopoulos O, Kibel A, Gray S, Kaelin WG, Jr. 1995 Tumour suppression by the human von Hippel-Lindau gene product. Nat Med 1:822-826 Chen EY, Mazure NM, Cooper JA, Giaccia AJ 2001 Hypoxia activates a plateletderived growth factor receptor/phosphatidylinositol 3-kinase/Akt pathway that results in glycogen synthase kinase-3 inactivation. Cancer Res 61:2429-2433 Zundel W, Schindler C, Haas-Kogan D, Koong A, Kaper F, Chen E, Gottschalk AR, Ryan HE, Johnson RS, Jefferson AB, Stokoe D, Giaccia AJ 2000 Loss of PTEN facilitates HIF-1-mediated gene expression. Genes Dev 14:391-396 Zhong H, Chiles K, Feldser D, Laughner E, Hanrahan C, Georgescu MM, Simons JW, Semenza GL 2000 Modulation of hypoxia-inducible factor 1alpha expression by the epidermal growth factor/phosphatidylinositol 3-kinase/PTEN/AKT/FRAP pathway in human prostate cancer cells: implications for tumor angiogenesis and therapeutics. Cancer Res 60:1541-1545 Fang J, Ding M, Yang L, Liu LZ, Jiang BH 2007 PI3K/PTEN/AKT signaling regulates prostate tumor angiogenesis. Cell Signal 19:2487-2497 Pore N, Jiang Z, Shu HK, Bernhard E, Kao GD, Maity A 2006 Akt1 activation can augment hypoxia-inducible factor-1alpha expression by increasing protein translation through a mammalian target of rapamycin-independent pathway. Mol Cancer Res 4:471-479 Treins C, Giorgetti-Peraldi S, Murdaca J, Monthouel-Kartmann MN, Van Obberghen E 2005 Regulation of hypoxia-inducible factor (HIF)-1 activity and expression of HIF hydroxylases in response to insulin-like growth factor I. Mol Endocrinol 19:1304-1317 Mottet D, Dumont V, Deccache Y, Demazy C, Ninane N, Raes M, Michiels C 2003 Regulation of hypoxia-inducible factor-1alpha protein level during hypoxic conditions by the phosphatidylinositol 3-kinase/Akt/glycogen synthase kinase 3beta pathway in HepG2 cells. J Biol Chem 278:31277-31285 142

193. 194. 195.

196.

197. 198. 199.

200.

201. 202.

203. 204. 205.

Zhou J, Schmid T, Frank R, Brune B 2004 PI3K/Akt is required for heat shock proteins to protect hypoxia-inducible factor 1alpha from pVHL-independent degradation. J Biol Chem 279:13506-13513 Shayesteh L, Lu Y, Kuo WL, Baldocchi R, Godfrey T, Collins C, Pinkel D, Powell B, Mills GB, Gray JW 1999 PIK3CA is implicated as an oncogene in ovarian cancer. Nat Genet 21:99-102 Zhang L, Yang N, Katsaros D, Huang W, Park JW, Fracchioli S, Vezzani C, Rigault de la Longrais IA, Yao W, Rubin SC, Coukos G 2003 The oncogene phosphatidylinositol 3'-kinase catalytic subunit alpha promotes angiogenesis via vascular endothelial growth factor in ovarian carcinoma. Cancer Res 63:42254231 Majumder PK, Febbo PG, Bikoff R, Berger R, Xue Q, McMahon LM, Manola J, Brugarolas J, McDonnell TJ, Golub TR, Loda M, Lane HA, Sellers WR 2004 mTOR inhibition reverses Akt-dependent prostate intraepithelial neoplasia through regulation of apoptotic and HIF-1-dependent pathways. Nat Med 10:594601 Chan DA, Sutphin PD, Denko NC, Giaccia AJ 2002 Role of prolyl hydroxylation in oncogenically stabilized hypoxia-inducible factor-1alpha. J Biol Chem 277:40112-40117 Lim JH, Lee ES, You HJ, Lee JW, Park JW, Chun YS 2004 Ras-dependent induction of HIF-1alpha785 via the Raf/MEK/ERK pathway: a novel mechanism of Ras-mediated tumor promotion. Oncogene 23:9427-9431 Blancher C, Moore JW, Robertson N, Harris AL 2001 Effects of ras and von Hippel-Lindau (VHL) gene mutations on hypoxia-inducible factor (HIF)-1alpha, HIF-2alpha, and vascular endothelial growth factor expression and their regulation by the phosphatidylinositol 3'-kinase/Akt signaling pathway. Cancer Res 61:7349-7355 Jiang BH, Agani F, Passaniti A, Semenza GL 1997 V-SRC induces expression of hypoxia-inducible factor 1 (HIF-1) and transcription of genes encoding vascular endothelial growth factor and enolase 1: involvement of HIF-1 in tumor progression. Cancer Res 57:5328-5335 Flugel D, Gorlach A, Michiels C, Kietzmann T 2007 Glycogen synthase kinase 3 phosphorylates hypoxia-inducible factor 1alpha and mediates its destabilization in a VHL-independent manner. Mol Cell Biol 27:3253-3265 Gradin K, McGuire J, Wenger RH, Kvietikova I, fhitelaw ML, Toftgard R, Tora L, Gassmann M, Poellinger L 1996 Functional interference between hypoxia and dioxin signal transduction pathways: competition for recruitment of the Arnt transcription factor. Mol Cell Biol 16:5221-5231 Minet E, Mottet D, Michel G, Roland I, Raes M, Remacle J, Michiels C 1999 Hypoxia-induced activation of HIF-1: role of HIF-1alpha-Hsp90 interaction. FEBS Lett 460:251-256 Katschinski DM, Le L, Schindler SG, Thomas T, Voss AK, Wenger RH 2004 Interaction of the PAS B domain with HSP90 accelerates hypoxia-inducible factor-1alpha stabilization. Cell Physiol Biochem 14:351-360 Isaacs JS, Jung YJ, Neckers L 2004 Aryl hydrocarbon nuclear translocator (ARNT) promotes oxygen-independent stabilization of hypoxia-inducible factor143

206.

207.

208. 209.

210.

211. 212.

213. 214. 215.

216.

217.

1alpha by modulating an Hsp90-dependent regulatory pathway. J Biol Chem 279:16128-16135 Minet E, Ernest I, Michel G, Roland I, Remacle J, Raes M, Michiels C 1999 HIF1A gene transcription is dependent on a core promoter sequence encompassing activating and inhibiting sequences located upstream from the transcription initiation site and cis elements located within the 5'UTR. Biochem Biophys Res Commun 261:534-540 Liu YV, Baek JH, Zhang H, Diez R, Cole RN, Semenza GL 2007 RACK1 competes with HSP90 for binding to HIF-1alpha and is required for O(2)independent and HSP90 inhibitor-induced degradation of HIF-1alpha. Mol Cell 25:207-217 Dulak J, Deshane J, Jozkowicz A, Agarwal A 2008 Heme oxygenase-1 and carbon monoxide in vascular pathobiology: focus on angiogenesis. Circulation 117:231-241 Choi YK, Kim CK, Lee H, Jeoung D, Ha KS, Kwon YG, Kim KW, Kim YM 2010 Carbon monoxide promotes VEGF expression by increasing HIF-1alpha protein level via two distinct mechanisms, translational activation and stabilization of HIF-1alpha protein. J Biol Chem 285:32116-32125 Lee J, Park SY, Lee EK, Park CG, Chung HC, Rha SY, Kim YK, Bae GU, Kim BK, Han JW, Lee HY 2006 Activation of hypoxia-inducible factor-1alpha is necessary for lysophosphatidic acid-induced vascular endothelial growth factor expression. Clin Cancer Res 12:6351-6358 Thomas R, Kim MH 2008 HIF-1 alpha: a key survival factor for serum-deprived prostate cancer cells. Prostate 68:1405-1415 Dimova EY, Moller U, Herzig S, Fink T, Zachar V, Ebbesen P, Kietzmann T 2005 Transcriptional regulation of plasminogen activator inhibitor-1 expression by insulin-like growth factor-1 via MAP kinases and hypoxia-inducible factor-1 in HepG2 cells. Thromb Haemost 93:1176-1184 Bardos JI, Chau NM, Ashcroft M 2004 Growth factor-mediated induction of HDM2 positively regulates hypoxia-inducible factor 1alpha expression. Mol Cell Biol 24:2905-2914 Huang Y, Hua K, Zhou X, Jin H, Chen X, Lu X, Yu Y, Zha X, Feng Y 2008 Activation of the PI3K/AKT pathway mediates FSH-stimulated VEGF expression in ovarian serous cystadenocarcinoma. Cell Res 18:780-791 Zhong H, De Marzo AM, Laughner E, Lim M, Hilton DA, Zagzag D, Buechler P, Isaacs WB, Semenza GL, Simons JW 1999 Overexpression of hypoxia-inducible factor 1alpha in common human cancers and their metastases. Cancer Res 59:5830-5835 Talks KL, Turley H, Gatter KC, Maxwell PH, Pugh CW, Ratcliffe PJ, Harris AL 2000 The expression and distribution of the hypoxia-inducible factors HIF-1alpha and HIF-2alpha in normal human tissues, cancers, and tumor-associated macrophages. Am J Pathol 157:411-421 Birner P, Schindl M, Obermair A, Breitenecker G, Oberhuber G 2001 Expression of hypoxia-inducible factor 1alpha in epithelial ovarian tumors: its impact on prognosis and on response to chemotherapy. Clin Cancer Res 7:1661-1668 144

218.

219. 220. 221. 222. 223. 224. 225.

226. 227.

228. 229. 230. 231. 232.

233.

Daponte A, Ioannou M, Mylonis I, Simos G, Minas M, Messinis IE, Koukoulis G 2008 Prognostic significance of Hypoxia-Inducible Factor 1 alpha(HIF-1 alpha) expression in serous ovarian cancer: an immunohistochemical study. BMC Cancer 8:335 Semenza GL 2003 Targeting HIF-1 for cancer therapy. Nat Rev Cancer 3:721732 Rankin EB, Giaccia AJ 2008 The role of hypoxia-inducible factors in tumorigenesis. Cell Death Differ 15:678-685 North S, Moenner M, Bikfalvi A 2005 Recent developments in the regulation of the angiogenic switch by cellular stress factors in tumors. Cancer Lett 218:1-14 Folkman J, Watson K, Ingber D, Hanahan D 1989 Induction of angiogenesis during the transition from hyperplasia to neoplasia. Nature 339:58-61 Giordano FJ, Johnson RS 2001 Angiogenesis: the role of the microenvironment in flipping the switch. Curr Opin Genet Dev 11:35-40 de Fraipont F, Nicholson AC, Feige JJ, Van Meir EG 2001 Thrombospondins and tumor angiogenesis. Trends Mol Med 7:401-407 Mancuso MR, Davis R, Norberg SM, O'Brien S, Sennino B, Nakahara T, Yao VJ, Inai T, Brooks P, Freimark B, Shalinsky DR, Hu-Lowe DD, McDonald DM 2006 Rapid vascular regrowth in tumors after reversal of VEGF inhibition. J Clin Invest 116:2610-2621 Forsythe JA, Jiang BH, Iyer NV, Agani F, Leung SW, Koos RD, Semenza GL 1996 Activation of vascular endothelial growth factor gene transcription by hypoxia-inducible factor 1. Mol Cell Biol 16:4604-4613 Maxwell PH, Dachs GU, Gleadle JM, Nicholls LG, Harris AL, Stratford IJ, Hankinson O, Pugh CW, Ratcliffe PJ 1997 Hypoxia-inducible factor-1 modulates gene expression in solid tumors and influences both angiogenesis and tumor growth. Proc Natl Acad Sci U S A 94:8104-8109 Tang N, Wang L, Esko J, Giordano FJ, Huang Y, Gerber HP, Ferrara N, Johnson RS 2004 Loss of HIF-1alpha in endothelial cells disrupts a hypoxia-driven VEGF autocrine loop necessary for tumorigenesis. Cancer Cell 6:485-495 Jensen RL, Ragel BT, Whang K, Gillespie D 2006 Inhibition of hypoxia inducible factor-1alpha (HIF-1alpha) decreases vascular endothelial growth factor (VEGF) secretion and tumor growth in malignant gliomas. J Neurooncol 78:233-247 Yuecheng Y, Hongmei L, Xiaoyan X 2006 Clinical evaluation of E-cadherin expression and its regulation mechanism in epithelial ovarian cancer. Clin Exp Metastasis 23:65-74 Brinck U, Jacobs S, Neuss M, Tory K, Rath W, Kulle B, Fuzesi L 2004 Diffuse growth pattern affects E-cadherin expression in invasive breast cancer. Anticancer Res 24:2237-2242 Kallakury BV, Sheehan CE, Winn-Deen E, Oliver J, Fisher HA, Kaufman RP, Jr., Ross JS 2001 Decreased expression of catenins (alpha and beta), p120 CTN, and E-cadherin cell adhesion proteins and E-cadherin gene promoter methylation in prostatic adenocarcinomas. Cancer 92:2786-2795 Sulzer MA, Leers MP, van Noord JA, Bollen EC, Theunissen PH 1998 Reduced E-cadherin expression is associated with increased lymph node metastasis and 145

234. 235. 236.

237. 238. 239.

240. 241. 242. 243.

244.

245. 246.

247.

unfavorable prognosis in non-small cell lung cancer. Am J Respir Crit Care Med 157:1319-1323 Hanahan D, Weinberg RA 2000 The hallmarks of cancer. Cell 100:57-70 Imai T, Horiuchi A, Wang C, Oka K, Ohira S, Nikaido T, Konishi I 2003 Hypoxia attenuates the expression of E-cadherin via up-regulation of SNAIL in ovarian carcinoma cells. Am J Pathol 163:1437-1447 Krishnamachary B, Zagzag D, Nagasawa H, Rainey K, Okuyama H, Baek JH, Semenza GL 2006 Hypoxia-inducible factor-1-dependent repression of Ecadherin in von Hippel-Lindau tumor suppressor-null renal cell carcinoma mediated by TCF3, ZFHX1A, and ZFHX1B. Cancer Res 66:2725-2731 Erler JT, Bennewith KL, Nicolau M, Dornhofer N, Kong C, Le QT, Chi JT, Jeffrey SS, Giaccia AJ 2006 Lysyl oxidase is essential for hypoxia-induced metastasis. Nature 440:1222-1226 Lu X, Kang Y 2010 Hypoxia and hypoxia-inducible factors: master regulators of metastasis. Clin Cancer Res 16:5928-5935 Lin MT, Kuo IH, Chang CC, Chu CY, Chen HY, Lin BR, Sureshbabu M, Shih HJ, Kuo ML 2008 Involvement of hypoxia-inducing factor-1alpha-dependent plasminogen activator inhibitor-1 up-regulation in Cyr61/CCN1-induced gastric cancer cell invasion. J Biol Chem 283:15807-15815 Edwin F, Anderson K, Ying C, Patel TB 2009 Intermolecular interactions of Sprouty proteins and their implications in development and disease. Mol Pharmacol 76:679-691 Cannistra SA 1993 Cancer of the ovary. N Engl J Med 329:1550-1559 Zeineldin R, Muller CY, Stack MS, Hudson LG 2010 Targeting the EGF receptor for ovarian cancer therapy. J Oncol 2010:414676 Pradhan M, Risberg BA, Trope CG, van de Rijn M, Gilks CB, Lee CH 2010 Gross genomic alterations and gene expression profiles of high- grade serous carcinoma of the ovary with and without BRCA1 inactivation. BMC Cancer 10:493 Press JZ, De Luca A, Boyd N, Young S, Troussard A, Ridge Y, Kaurah P, Kalloger SE, Blood KA, Smith M, Spellman PT, Wang Y, Miller DM, Horsman D, Faham M, Gilks CB, Gray J, Huntsman DG 2008 Ovarian carcinomas with genetic and epigenetic BRCA1 loss have distinct molecular abnormalities. BMC Cancer 8:17 Brown LA, Kalloger SE, Miller MA, Shih Ie M, McKinney SE, Santos JL, Swenerton K, Spellman PT, Gray J, Gilks CB, Huntsman DG 2008 Amplification of 11q13 in ovarian carcinoma. Genes Chromosomes Cancer 47:481-489 Maines-Bandiera SL, Kruk PA, Auersperg N 1992 Simian virus 40-transformed human ovarian surface epithelial cells escape normal growth controls but retain morphogenetic responses to extracellular matrix. Am J Obstet Gynecol 167:729735 Geisinger KR, Kute TE, Pettenati MJ, Welander CE, Dennard Y, Collins LA, Berens ME 1989 Characterization of a human ovarian carcinoma cell line with estrogen and progesterone receptors. Cancer 63:280-288

146

248. 249.

250.

251.

252. 253. 254.

255. 256.

257. 258. 259.

260.

Cheng JC, Klausen C, Leung PC 2010 Hydrogen peroxide mediates EGF-induced down-regulation of E-cadherin expression via p38 MAPK and snail in human ovarian cancer cells. Molecular endocrinology (Baltimore, Md 24:1569-1580 Zweemer RP, Ryan A, Snijders AM, Hermsen MA, Meijer GA, Beller U, Menko FH, Jacobs IJ, Baak JP, Verheijen RH, Kenemans P, van Diest PJ 2001 Comparative genomic hybridization of microdissected familial ovarian carcinoma: two deleted regions on chromosome 15q not previously identified in sporadic ovarian carcinoma. Lab Invest 81:1363-1370 Jongsma AP, Piek JM, Zweemer RP, Verheijen RH, Klein Gebbinck JW, van Kamp GJ, Jacobs IJ, Shaw P, van Diest PJ, Kenemans P 2002 Molecular evidence for putative tumour suppressor genes on chromosome 13q specific to BRCA1 related ovarian and fallopian tube cancer. Mol Pathol 55:305-309 Kim TM, Benedict WF, Xu HJ, Hu SX, Gosewehr J, Velicescu M, Yin E, Zheng J, D'Ablaing G, Dubeau L 1994 Loss of heterozygosity on chromosome 13 is common only in the biologically more aggressive subtypes of ovarian epithelial tumors and is associated with normal retinoblastoma gene expression. Cancer Res 54:605-609 Yang-Feng TL, Li S, Han H, Schwartz PE 1992 Frequent loss of heterozygosity on chromosomes Xp and 13q in human ovarian cancer. Int J Cancer 52:575-580 Voutilainen KA, Anttila MA, Sillanpaa SM, Ropponen KM, Saarikoski SV, Juhola MT, Kosma VM 2006 Prognostic significance of E-cadherin-catenin complex in epithelial ovarian cancer. J Clin Pathol 59:460-467 Rathi A, Virmani AK, Schorge JO, Elias KJ, Maruyama R, Minna JD, Mok SC, Girard L, Fishman DA, Gazdar AF 2002 Methylation profiles of sporadic ovarian tumors and nonmalignant ovaries from high-risk women. Clin Cancer Res 8:33243331 Pon YL, Zhou HY, Cheung AN, Ngan HY, Wong AS 2008 p70 S6 kinase promotes epithelial to mesenchymal transition through snail induction in ovarian cancer cells. Cancer Res 68:6524-6532 Park SH, Cheung LW, Wong AS, Leung PC 2008 Estrogen regulates Snail and Slug in the down-regulation of E-cadherin and induces metastatic potential of ovarian cancer cells through estrogen receptor alpha. Mol Endocrinol 22:20852098 Willmarth NE, Ethier SP 2006 Autocrine and juxtacrine effects of amphiregulin on the proliferative, invasive, and migratory properties of normal and neoplastic human mammary epithelial cells. J Biol Chem 281:37728-37737 Park JY, Su YQ, Ariga M, Law E, Jin SL, Conti M 2004 EGF-like growth factors as mediators of LH action in the ovulatory follicle. Science 303:682-684 Choi JH, Choi KC, Auersperg N, Leung PC 2006 Gonadotropins activate proteolysis and increase invasion through protein kinase A and phosphatidylinositol 3-kinase pathways in human epithelial ovarian cancer cells. Cancer Res 66:3912-3920 Lau MT, Wong AS, Leung PC 2010 Gonadotropins induce tumor cell migration and invasion by increasing cyclooxygenases expression and prostaglandin E(2) production in human ovarian cancer cells. Endocrinology 151:2985-2993 147

261. 262. 263. 264.

265. 266.

267. 268.

269. 270. 271.

272.

273. 274.

Zhang YW, Vande Woude GF 2007 Mig-6, signal transduction, stress response and cancer. Cell Cycle 6:507-513 Wakioka T, Sasaki A, Kato R, Shouda T, Matsumoto A, Miyoshi K, Tsuneoka M, Komiya S, Baron R, Yoshimura A 2001 Spred is a Sprouty-related suppressor of Ras signalling. Nature 412:647-651 Willmarth NE, Ethier SP 2008 Amphiregulin as a novel target for breast cancer therapy. J Mammary Gland Biol Neoplasia 13:171-179 Menashi S, Serova M, Ma L, Vignot S, Mourah S, Calvo F 2003 Regulation of extracellular matrix metalloproteinase inducer and matrix metalloproteinase expression by amphiregulin in transformed human breast epithelial cells. Cancer Res 63:7575-7580 Kondapaka SB, Fridman R, Reddy KB 1997 Epidermal growth factor and amphiregulin up-regulate matrix metalloproteinase-9 (MMP-9) in human breast cancer cells. Int J Cancer 70:722-726 P Oc, Modjtahedi H, Rhys-Evans P, Court WJ, Box GM, Eccles SA 2000 Epidermal growth factor-like ligands differentially up-regulate matrix metalloproteinase 9 in head and neck squamous carcinoma cells. Cancer Res 60:1121-1128 Liu Z, Klominek J 2003 Regulation of matrix metalloprotease activity in malignant mesothelioma cell lines by growth factors. Thorax 58:198-203 Picihard V, Berthois Y, Roccabianca M, Prevot C, Sarrazin M, Portugal H, Kumar S, Kumar P, Rognoni JB 2006 Concomitant cell growth and differentiation are dependent on erbB1 and integrin activation in an autonomously surviving colon adenocarcinoma: involvement of autocrine amphiregulin secretion. Anticancer Res 26:2769-2783 Chung E, Cook PW, Parkos CA, Park YK, Pittelkow MR, Coffey RJ 2005 Amphiregulin causes functional downregulation of adherens junctions in psoriasis. J Invest Dermatol 124:1134-1140 Chung E, Graves-Deal R, Franklin JL, Coffey RJ 2005 Differential effects of amphiregulin and TGF-alpha on the morphology of MDCK cells. Exp Cell Res 309:149-160 Sasaki A, Taketomi T, Wakioka T, Kato R, Yoshimura A 2001 Identification of a dominant negative mutant of Sprouty that potentiates fibroblast growth factor- but not epidermal growth factor-induced ERK activation. J Biol Chem 276:3680436808 Impagnatiello MA, Weitzer S, Gannon G, Compagni A, Cotten M, Christofori G 2001 Mammalian sprouty-1 and -2 are membrane-anchored phosphoprotein inhibitors of growth factor signaling in endothelial cells. J Cell Biol 152:10871098 Ozaki K, Miyazaki S, Tanimura S, Kohno M 2005 Efficient suppression of FGF2-induced ERK activation by the cooperative interaction among mammalian Sprouty isoforms. J Cell Sci 118:5861-5871 Taniguchi K, Ishizaki T, Ayada T, Sugiyama Y, Wakabayashi Y, Sekiya T, Nakagawa R, Yoshimura A 2009 Sprouty4 deficiency potentiates Rasindependent angiogenic signals and tumor growth. Cancer Sci 100:1648-1654 148

275. 276.

277. 278.

279. 280. 281. 282. 283. 284. 285. 286.

287. 288. 289.

Lau MT, Klausen C, Leung PC 2011 E-cadherin inhibits tumor cell growth by suppressing PI3K/Akt signaling via beta-catenin-Egr1-mediated PTEN expression. Oncogene 30:2753-2766 Winn RA, Van Scoyk M, Hammond M, Rodriguez K, Crossno JT, Jr., Heasley LE, Nemenoff RA 2006 Antitumorigenic effect of Wnt 7a and Fzd 9 in non-small cell lung cancer cells is mediated through ERK-5-dependent activation of peroxisome proliferator-activated receptor gamma. J Biol Chem 281:26943-26950 Chambers D, Mason I 2000 Expression of sprouty2 during early development of the chick embryo is coincident with known sites of FGF signalling. Mech Dev 91:361-364 Chi L, Zhang S, Lin Y, Prunskaite-Hyyrylainen R, Vuolteenaho R, Itaranta P, Vainio S 2004 Sprouty proteins regulate ureteric branching by coordinating reciprocal epithelial Wnt11, mesenchymal Gdnf and stromal Fgf7 signalling during kidney development. Development 131:3345-3356 Ding W, Bellusci S, Shi W, Warburton D 2003 Functional analysis of the human Sprouty2 gene promoter. Gene 322:175-185 Ding W, Bellusci S, Shi W, Warburton D 2004 Genomic structure and promoter characterization of the human Sprouty4 gene, a novel regulator of lung morphogenesis. Am J Physiol Lung Cell Mol Physiol 287:L52-59 Haigl B, Mayer CE, Siegwart G, Sutterluty H 2010 Sprouty4 levels are increased under hypoxic conditions by enhanced mRNA stability and transcription. Biol Chem 391:813-821 Dayan F, Bilton RL, Laferriere J, Trottier E, Roux D, Pouyssegur J, Mazure NM 2009 Activation of HIF-1alpha in exponentially growing cells via hypoxic stimulation is independent of the Akt/mTOR pathway. J Cell Physiol 218:167-174 Park SY, Jeong KJ, Lee J, Yoon DS, Choi WS, Kim YK, Han JW, Kim YM, Kim BK, Lee HY 2007 Hypoxia enhances LPA-induced HIF-1alpha and VEGF expression: their inhibition by resveratrol. Cancer Lett 258:63-69 Abe M, Naski MC 2004 Regulation of sprouty expression by PLCgamma and calcium-dependent signals. Biochem Biophys Res Commun 323:1040-1047 Nutt SL, Dingwell KS, Holt CE, Amaya E 2001 Xenopus Sprouty2 inhibits FGFmediated gastrulation movements but does not affect mesoderm induction and patterning. Genes Dev 15:1152-1166 Barbachano A, Ordonez-Moran P, Garcia JM, Sanchez A, Pereira F, Larriba MJ, Martinez N, Hernandez J, Landolfi S, Bonilla F, Palmer HG, Rojas JM, Munoz A 2010 SPROUTY-2 and E-cadherin regulate reciprocally and dictate colon cancer cell tumourigenicity. Oncogene 29:48004813 Nadeau RJ, Toher JL, Yang X, Kovalenko D, Friesel R 2007 Regulation of Sprouty2 stability by mammalian Seven-in-Absentia homolog 2. J Cell Biochem 100:151-160 Nakayama K, Qi J, Ronai Z 2009 The ubiquitin ligase Siah2 and the hypoxia response. Mol Cancer Res 7:443-451 Skinner HD, Zheng JZ, Fang J, Agani F, Jiang BH 2004 Vascular endothelial growth factor transcriptional activation is mediated by hypoxia-inducible factor 1alpha, HDM2, and p70S6K1 in response to phosphatidylinositol 3-kinase/AKT signaling. J Biol Chem 279:45643-45651 149

290.

291. 292. 293. 294. 295.

296. 297. 298.

299.

300. 301. 302.

303.

Place TL, Fitzgerald MP, Venkataraman S, Vorrink SU, Case AJ, Teoh ML, Domann FE 2011 Aberrant promoter CpG methylation is a mechanism for impaired PHD3 expression in a diverse set of malignant cells. PLoS One 6:e14617 Berra E, Benizri E, Ginouves A, Volmat V, Roux D, Pouyssegur J 2003 HIF prolyl-hydroxylase 2 is the key oxygen sensor setting low steady-state levels of HIF-1alpha in normoxia. Embo J 22:4082-4090 Lo TL, Fong CW, Yusoff P, McKie AB, Chua MS, Leung HY, Guy GR 2006 Sprouty and cancer: The first terms report. Cancer Lett Liao H, Hyman MC, Lawrence DA, Pinsky DJ 2007 Molecular regulation of the PAI-1 gene by hypoxia: contributions of Egr-1, HIF-1alpha, and C/EBPalpha. FASEB J 21:935-949 Wenger RH, Stiehl DP, Camenisch G 2005 Integration of oxygen signaling at the consensus HRE. Sci STKE 2005:re12 Takabuchi S, Hirota K, Nishi K, Oda S, Oda T, Shingu K, Takabayashi A, Adachi T, Semenza GL, Fukuda K 2004 The inhibitory effect of sodium nitroprusside on HIF-1 activation is not dependent on nitric oxide-soluble guanylyl cyclase pathway. Biochem Biophys Res Commun 324:417-423 Yagi H, Yotsumoto F, Miyamoto S 2008 Heparin-binding epidermal growth factor-like growth factor promotes transcoelomic metastasis in ovarian cancer through epithelial-mesenchymal transition. Mol Cancer Ther 7:3441-3451 Waterman H, Katz M, Rubin C, Shtiegman K, Lavi S, Elson A, Jovin T, Yarden Y 2002 A mutant EGF-receptor defective in ubiquitylation and endocytosis unveils a role for Grb2 in negative signaling. Embo J 21:303-313 Akbulut S, Reddi AL, Aggarwal P, Ambardekar C, Canciani B, Kim MK, Hix L, Vilimas T, Mason J, Basson MA, Lovatt M, Powell J, Collins S, Quatela S, Phillips M, Licht JD 2010 Sprouty proteins inhibit receptor-mediated activation of phosphatidylinositol-specific phospholipase C. Mol Biol Cell 21:3487-3496 Bertheau P, Turpin E, Rickman DS, Espie M, de Reynies A, Feugeas JP, Plassa LF, Soliman H, Varna M, de Roquancourt A, Lehmann-Che J, Beuzard Y, Marty M, Misset JL, Janin A, de The H 2007 Exquisite sensitivity of TP53 mutant and basal breast cancers to a dose-dense epirubicin-cyclophosphamide regimen. PLoS Med 4:e90 Wright G, Tan B, Rosenwald A, Hurt EH, Wiestner A, Staudt LM 2003 A gene expression-based method to diagnose clinically distinct subgroups of diffuse large B cell lymphoma. Proc Natl Acad Sci U S A 100:9991-9996 Faratian D, Sims AH, Mullen P, Kay C, Um I, Langdon SP, Harrison DJ 2011 Sprouty 2 is an independent prognostic factor in breast cancer and may be useful in stratifying patients for trastuzumab therapy. PLoS One 6:e23772 Huebert RC, Li Q, Adhikari N, Charles NJ, Han X, Ezzat MK, Grindle S, Park S, Ormaza S, Fermin D, Miller LW, Hall JL 2004 Identification and regulation of Sprouty1, a negative inhibitor of the ERK cascade, in the human heart. Physiol Genomics 18:284-289 Basson MA, Akbulut S, Watson-Johnson J, Simon R, Carroll TJ, Shakya R, Gross I, Martin GR, Lufkin T, McMahon AP, Wilson PD, Costantini FD, Mason IJ, 150

304.

305.

306. 307. 308. 309. 310.

311. 312.

Licht JD 2005 Sprouty1 is a critical regulator of GDNF/RET-mediated kidney induction. Dev Cell 8:229-239 Zillhardt M, Park SM, Romero IL, Sawada K, Montag A, Krausz T, Yamada SD, Peter ME, Lengyel E 2011 Foretinib (GSK1363089), an orally available multikinase inhibitor of c-Met and VEGFR-2, blocks proliferation, induces anoikis, and impairs ovarian cancer metastasis. Clin Cancer Res 17:4042-4051 Li X, Brunton VG, Burgar HR, Wheldon LM, Heath JK 2004 FRS2-dependent SRC activation is required for fibroblast growth factor receptor-induced phosphorylation of Sprouty and suppression of ERK activity. J Cell Sci 117:60076017 Edwin F, Patel TB 2008 A novel role of Sprouty 2 in regulating cellular apoptosis. J Biol Chem 283:3181-3190 Fujita Y, Krause G, Scheffner M, Zechner D, Leddy HE, Behrens J, Sommer T, Birchmeier W 2002 Hakai, a c-Cbl-like protein, ubiquitinates and induces endocytosis of the E-cadherin complex. Nat Cell Biol 4:222-231 Ciechanover A, Iwai K 2004 The ubiquitin system: from basic mechanisms to the patient bed. IUBMB Life 56:193-201 Frankel A, Man S, Elliott P, Adams J, Kerbel RS 2000 Lack of multicellular drug resistance observed in human ovarian and prostate carcinoma treated with the proteasome inhibitor PS-341. Clin Cancer Res 6:3719-3728 LeBlanc R, Catley LP, Hideshima T, Lentzsch S, Mitsiades CS, Mitsiades N, Neuberg D, Goloubeva O, Pien CS, Adams J, Gupta D, Richardson PG, Munshi NC, Anderson KC 2002 Proteasome inhibitor PS-341 inhibits human myeloma cell growth in vivo and prolongs survival in a murine model. Cancer Res 62:49965000 Rajkumar SV, Richardson PG, Hideshima T, Anderson KC 2005 Proteasome inhibition as a novel therapeutic target in human cancer. J Clin Oncol 23:630-639 Adams J, Palombella VJ, Sausville EA, Johnson J, Destree A, Lazarus DD, Maas J, Pien CS, Prakash S, Elliott PJ 1999 Proteasome inhibitors: a novel class of potent and effective antitumor agents. Cancer Res 59:2615-2622

151