UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The investigation of a novel BRCA2-associated gene, EMSY, on chromosome 11Q13 in breast and ovarian cancer Brown, Lindsay Anne 2007

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

Item Metadata

Download

Media
831-ubc_2007-266892.pdf [ 12.14MB ]
Metadata
JSON: 831-1.0100334.json
JSON-LD: 831-1.0100334-ld.json
RDF/XML (Pretty): 831-1.0100334-rdf.xml
RDF/JSON: 831-1.0100334-rdf.json
Turtle: 831-1.0100334-turtle.txt
N-Triples: 831-1.0100334-rdf-ntriples.txt
Original Record: 831-1.0100334-source.json
Full Text
831-1.0100334-fulltext.txt
Citation
831-1.0100334.ris

Full Text

T H E INVESTIGATION OF A NOVEL B R C A 2 - A S S O C I A T E D GENE, EMSY, ON CHROMOSOME 11Q13 IN BREAST AND OVARIAN CANCER by LINDSAY A N N E BROWN B.Sc, University of Waterloo, 1999 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Pathology and Laboratory Medicine) THE UNIVERSITY OF BRITISH COLUMBIA March 2007 © Lindsay Anne Brown, 2007 ABSTRACT EMSY is a novel gene encoding a protein that binds to the tumor suppressor BRCA2. EMSY maps to 11q13, a region commonly amplified in breast cancer, and is located 6 Mb telomeric to CCND1. However, the role of EMSY amplification in breast and ovarian cancer is largely unknown. Here we present evidence that EMSY is commonly amplified in breast cancer, and associated with a poor outcome in several pathological subsets o f breast cancer. EMSY gene amplification is a less common event in BRCA2 mutation carriers, providing evidence to support the potential role o f EMSY as surrogate for BRCA2 loss in sporadic breast cancer. In ovarian cancer, EMSY amplification is associated with high grade ovarian carcinomas, and is a frequently amplified element o f the 11 q13 amplicon in ovarian cancer. A more detailed analysis o f thel 1 q13 amplicon, which harbors numerous oncogenes including CCND1, EMSY, PAK1, Rsf-1, and GAB2, demonstrated that all five genes are frequently amplified in ovarian carcinomas and are specifically associated with serous carcinomas. Lastly, we wanted to determine whether EMSY overexpression shares similar biological features with loss o f BRCA2 function. Overexpression of a 5' fragment of EMSY induces a chromosomal instability phenotype in human breast epithelial cells that is similar to that o f BRCA2 deficient cells. Furthermore, treatment with the DNA damaging agent mitomycin C produced a several fold higher frequency o f chromosome breaks in the EMSY overexpressing cells than in the control cells. Overexpression of the 5' fragment o f EMSY did not induce a chromosomal instability phenotype in two immortalized normal ovarian surface epithelial cell lines even after treatment with mitomycin C, suggesting that ii these cells may not be the correct cell type in which to study EMSY overexpression. EMSY overexpressing breast epithelial cells did not exhibit increased sensitivity to treatment with the DNA damaging agents, doxorubicin and cisplatin, suggesting that EMSY may not play a role in the direct repair of DNA double stranded breaks. In conclusion, these results demonstrate that EMSY amplification is a clinically significant event in breast and ovarian cancer, and elevated levels of EMSY may play a role in sporadic breast carcinogenesis by deregulating protection of genomic stability. T A B L E O F C O N T E N T S . Abstract ii Table of Contents iv List of Tables vii List of Figures ix List of Abbreviations xi Acknowledgements xv Dedication xvi Chapter 1: Literature Review, Hypotheses, and Specific Objectives 1.1 Clinical aspects of breast cancer 1 1.2 Clinical aspects of ovarian cancer 6 1.3 Hereditary breast and ovarian cancer 9 1.4 BRCA1, BRCA2 and chromosomal instability 12 1.5 BRCA1, BRCA2 and DNA recombination and repair 13 1.6 Therapeutic targeting of the DNA repair defect of BRCA tumors 21 1.7 Histopathology of BRCA1 and BRCA2 tumors 22 1.8 BRCA1, BRCA2 and hereditary versus sporadic breast and ovarian cancer. 31 1.9 EMSY amplification and breast and ovarian cancer 32 1.10 Scope of hypothesis 40 1.11 References 42 Chapter 2 : Fluorescent in situ hybridization on tissue microarrays: problems and solutions 2.1 Rationale 58 2.2 Introduction 58 2.3 TMAs 58 2.4 FISH on TMAs 59 2.5 Applications of FISH on TMAs 62 iv 2.6 Development of a robust FISH assay for use of paraffin embedded TMAs 6 5 2.7 Scoring of FISH signals 68 2.8 Conclusions 7 2 2.9 References 7 3 Chapter 3: EMSY gene amplification in sporadic breast cancers and BRCA1 and BRCA2 mutation carriers 3.1 Introduction 7 6 3.2 Material and Methods 7 7 3.3 Results 8 4 3.4 Discussion 9 3 3.5 References 9 7 Chapter 4: EMSY gene amplification in ovarian cancer 4.1 Introduction 1 0 1 4.2 Material and Methods 1 0 2 4.3 Results 1 0 6 4.4 Discussion 1 1 4 4.5 References 1 ' 1 8 Chapter 5: A comprehensive analysis of the 11q13 amplicon in ovarian cancer 5.1 Introduction 1 2 2 5.2 Material and Methods 1 2 5 5.3 Results 1 2 8 5.4 Discussion ^35 5.5 References 1 3 9 v Chapter 6: EMSY overexpression in primary breast epithelial cells 6.1 Introduction 1 4 3 6.2 Material and Method 145 6.3 Results 1 5 n 6.4 Discussion 159 6.5 References 1 6 2 Chapter 7: Further functional analysis of EMSY overexpression in breast and ovarian surface epithelial cells 7.1 Introduction 1 6 6 7.2 Material and Methods 1 6 8 7.3 Results 1 7 4 7.4 Discussion 1 8 6 7.5 References 1 9 1 Chapter 8: Summary, Overall Conclusions and Future Directions 8.1 Summary and overall conclusions 193 vi LIST OF T A B L E S Chapter 2 Table 2.1 FISH and paraffin embedded TMAs: problems and solutions 67 Chapter 3 Table 3.1 Multivariate analysis of EMSY amplification and clinical outcome in node positive patients 90 Table 3.2 Multivariate analysis of CCND1 amplification and clinical outcome in node positive patients 90 Table 3.3 Incidence of EMSY and CCND1 amplification in familial breast cancer 92 Chapter 4 Table 4.1 Evaluation of EMSY and CCND1 amplification in ovarian cancer 109 Table 4.2 Gene amplification and overexpression profiles of CCND1, EMSY and PAK1 in twenty-two ovarian tumors 113 Chapter 5 Table 5.1 Correlations of all five genes: nonparametric Kendell T test 130 Table 5.2 Evaluation of CCND1, EMSY, PAK1, Rsf-1 and GAB2 in ovarian cancer 132 Table 5.3 Multivariate analysis 134 vii Table 6.1 Summary of the frequency of structural and numerical chromosomal abnormalities seen in EMSY overexpressing cells and control GFP-transduced cell analyzed at passage 5 and 10 after transduction 156 Chapter 7 Table 7.1 184-hTert breast epithelial cells treated with EMSY siRNA accumulate structural chromosomal abnormalities 185 viii LIST OF F IGURES Chapter 1 Figure 1.1 Percentage distribution of cancer deaths amongst women for selected cancer sites, Canada, 2006 2 Figure 1.2 Features of the BRCA proteins 11 Figure 1.3 DNA double strand break repair pathways 15 Figure 1.4 Sequence from the BRCA2 N terminus used for yeast two hybrid screens 33 Figure 1.5 The EMSY ENT domain 35 Figure 1.6 The 11q13 amplicon 37 Figure 1.7 EMSY and BRCA 39 Chapter 2 Figure 2.1 Applications of FISH on TMAs 60 Figure 2.2 Automated scoring of FISH signals 71 Chapter 3 Figure 3.1 EMSY gene amplification and clinical outcome 86 Figure 3.2 EMSY gene amplification and clinical outcome in the population based series of breast cancer 89 Chapter 4 Figure 4.1 Assessment of EMSY gene amplification in three ovarian carcinomas 107 Figure 4.2 Chromosome 11 array CGH data from 51 ovarian carcinomas based on a moving average of the five genes centered on that locus 110 ix Chapter 5 Figure 5.1 The 11 q13 amplicon and the potential genes involved in the amplification at this locus 123 Figure 5.2 Amplification at 11 q13 129 Figure 5.3 EMSY and Rsf-1 amplification correlate with a poor outcome 133 Chapter 6 Figure 6.1 Karyotype of the 184-hTert cells 151 Figure 6.2 EMSY/GFP transduced 184-hTert cells 152 Figure 6.3 EMSY overexpressing cells accumulate structural chromosomal abnormalities 154 Chapter 7 Figure 7.1 Karyotype of the IOSE cells 175 Figure 7.2 EMSY/GFP transduced IOSE cells 176 Figure 7.3 Doxorubicin treatment of EMSY overexpressing cells 180 Figure 7.4 Cisplatin treatment of EMSY overexpressing cells 181 Figure 7.5 Paclitaxel treatment of EMSY overexpressing cells 182 Figure 7.6 Treatment of 184-hTert cells with EMSY siRNA 184 Chapter 8 Figure 8.1 EMSY amplification and loss are associated with a poor outcome.... 200 x LIST OF ABBREVIATIONS aCGH Array comparative genomic hybridization AST Adjuvant systemic therapy ATM Ataxia telangiectasia mutated ATR ATR ATM- and Rad3-related BACs Bacterial artificial chromosomes BARD1 BRCA2 associated RING domain 1 BRCT BRCA1 C-terminal domains CDK Cyclin-dependent kinase CDK1 CDK inhibitors cDNA complementary DNA CHK2 Checkpoint kinase 2 CI Confidence interval CK Cytokeratins CtIP CtBP interacting protein DAPI 4',6-diamidino-2-phenylindole dATP deoxy adenine triphosphate dCTP deoxy cytosine triphosphate dGTP deoxy guanine triphosphate dH20 Distilled water DMSO Dimethyl sulfoxide DNA-PKcs DNA-dependent protein kinase dNTP deoxy ribonucleotide triphosphate DSB Double strand break DSS Disease specific survival dTTP deoxy thymidine triphosphate E2F3 E2F transcription factor 3 EMS1 Ems1 sequence ENT EMSY N-terminal EPCAM Epithelial cellular adhesion molecule, also known as TROP-1 xi ER Estrogen receptor ER- Estrogen receptor negative ER+ Estrogen receptor positive ERBB2 Epidermal growth factor receptor 2, also known as Her2/neu ERG V-ets erythroblastosis virus E26 oncogene homolog ESTs Expressed sequence tags EtOH Ethanol ETV1 Ets variant gene 1 ETV4 Ets variant gene 4 ETV6 Ets variant gene 6 FISH Fluorescent in situ hybridization GAB2 GRB2-associated binding protein 2 GADD45 Growth arrest and DNA damage GARP Glycoprotein A repetitions predominant GATA3 GATA-binding protein 3 GFP Green fluorescent protein GST Glutathione-S-transferase H&E Hematoxylin and eosin HCI Hydrochloric acid Her2- Her2/neu negative Her2+ Her2/neu positive HN1 Hematological and neurological expressed 1 HNF2A Hepatocyte nuclear factor 2 alpha HP1 Heterochromatin-associated protein-1 HR Homologous recombination IHC Immunohistochemistry IOSE Immortalized ovarian surface epithelial IR Ionizing radiation KIP Kinase inhibitor protein LVI Lymphatic or vascular invasion Mb Megabase xii MCS Multiple cloning site MEF Mouse embryonic fibroblast MMC Mitomycin c mRNA Messenger RNA MTT 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide NaOH Sodium hydroxide NaSCN Sodium thiocyanate NHEJ Non-homologous end joining NLS Nuclear localization signal node- Node negative node+ Node positive NOTCH3 Notch homolog 3 NP40 Nonidet P40 NTRK3 Neurotrophic tyrosine kinase, receptor, type 3 OS Overall survival P13K Phosphoinositide 3-kinase p53- p53 negative p53+ p53 positive PACs P1 artificial chromosomes PAK1 p21/cdc42/Rad-activated kinase PARP1 Poly(ADP-ribose) polymerase PCR Polymerase chain reaction PR Progesterone receptor PRCP Prolylcarboxypeptidase RACGAP1 Rac GTPase activating protein 1 RFS Relapse free survival Rsf-1 Remodeling and spacing factor 1, also known as HXABP RT-PCR Reverse transcriptase PCR SSA Singe strand annealing SSB Single strand break ssDNA Single strand DNA TMA Tissue micro-array UV Ultraviolet VEGF Vascular endothelial growth factor YACs Yeast artificial chromosomes xiv A C K N O W L E D G E M E N T S I am deeply thankful to have had the opportunity to work with my supervisor, Dr. David Huntsman. I was the first graduate student to join David's research group, and at the beginning our lab consisted of only Janine and myself. Now after, five years the Huntsman lab has grown immensely and it has been a pleasure to see how the lab has changed over the years to what it has become now. David has been a great mentor, and his dedication, excitement and commitment to scientific research has helped guide me in my own scientific endeavors. He will no doubt continue to inspire me in my future career. I am extremely grateful to have had the support and guidance of an exceptional group of scientists from the Huntsman lab, Nielsen lab, GPEC, and the Prostate Centre. Over the past few years, I have had the privilege of working with an extraordinary group of graduate students and technicians who have become not only colleagues, but also great friends. Thank you to Leah Prentice, Melinda Miller, Dawn Bradley, Vanessa Thompson, Maggie Cheang and Janine Senz for all your encouragement and help. My sincere appreciation also goes to the members of my supervisory committee, Dr. Blake Gilks, Dr. Colleen Nelson, and Dr. Cheryl Wellington for their support and scientific guidance throughout my graduate studies. xv To my husband, Hugh, for his love, support, and motivation. Thank-you! And to my parents, Jim and Susan Brown, my brother Bryce, and my grandparents, Charles and Esther Weir, for their unconditional love and encouragement. Thank you! xvi Chapter 1. Literature Review, Hypotheses, and Specific Objectives The focus of my research has been the study of a novel BRCA2 associated gene, EMSY. Inherited mutations in BRCA1 or BRCA2 predispose to breast, ovarian and other cancers. Thus, the focus of my research has been to assess the clinical significance of EMSY in both breast and ovarian cancer. 1.1 CLINICAL ASPECTS OF BREAST CANCER Breast cancer is the most commonly diagnosed cancer amongst Canadian women, and is second only to lung cancer as a common cause of cancer death (Figure 1.1). In 2006, an estimated 22,200 women will be diagnosed with breast cancer in Canada and 5,300 women will die of this disease [1]. During their lifetimes, 1 in 9 women will develop breast cancer and 1 in 27 women are expected to die from it. For breast cancer patients, it is not the primary tumor but metastasis to distant sites that is the main cause of death. The rates of metastasis and mortality have decreased as a result of early diagnosis by mammographic screening and systemic adjuvant therapy. However, there are many side effects associated with the use of these therapies that affect the patient's quality of life. Another concern is the fact that most women who are treated with adjuvant therapy may never develop distant metastasis and needlessly suffer the toxic effects of this therapy [2]. New breast cancer biomarkers are needed to identify characteristics of a patient or tumor that is indicative of the future course of the disease in the absence of Site 0 5 10 15 20 25 30 35 Percentage Figure 1.1 Percentage distribution of cancer deaths amongst women selected cancer sites, Canada, 2006 (Adapted from Canadian Cancer Society) treatment (prognostic marker) or that predict response to a particular therapy (predictive marker). This would enable clinicians to tailor treatment strategies based on the individual. 1.1.1 Breast cancer prognostic markers Breast cancer is characterized as a heterogeneous disease consisting of several pathological subtypes differing in the histological appearance of malignant cells, clinical outcome and treatment response [2]. The heterogeneous nature of breast cancer also makes it difficult to determine the clinical course of the disease as a result of differences in tumor growth rates, invasiveness of the tumor, metastatic potential and other important cellular growth signaling and survival pathways. There are several established breast cancer prognostic markers currently used to identify those patients associated with a poor outcome and they include: nodal status, histological subtype, tumor size, and estrogen receptor status [3-5]. For estrogen receptor positive breast cancers targeted treatment with estrogen antagonists, in particular Tamoxifen, have led to a decrease in breast cancer mortality [6-8]. Other biomarkers, such as the epidermal growth factor receptor 2 (ERBB2, also known as Her2/neu) have emerged as prognostic and predictive markers. The ERBB2 oncogene encodes a transmembrane receptor with constitutive tyrosine kinase activity. ERBB2 has been found amplified and overexpressed in 15% of breast cancer [9, 10] and studies have shown that amplification and overexpression of ERBB2 are associated with a poor outcome in lymph node 3 positive patients, but not lymph node negative patients. ERBB2 status has also gained clinical significance due to the development of HerceptinTm (trastuzumab), a therapeutic monoclonal antibody targeted against the receptor, which has helped prolong survival in patients with metastatic breast cancer [10]. 1.1.2 Gene expression profiling in breast cancer To accurately predict the behavior of a tumor may require the analysis of multiple markers at one time. This has been made possible with the introduction of DNA microarray technology, which analyzes gene expression in a genome-wide manner. The first key finding in breast cancer using DNA microarrays was that breast tumors could be classified based on their gene expression profile into four previously unrecognized subtypes [11, 12]. Three distinct subgroups characterized by low to absent estrogen receptor (ER) expression were identified and they included: the basal-like subtype expressing keratins 5 and 17 (characteristic of basal epithelial cells of the normal mammary gland) and low level expression of Her2/neu, the ERBB2 positive subtype which was characterized by high expression of several genes on the ERBB2 amplicon on 17q22.24 and a normal breast-like group which showed high expression of genes expressed by adipose tissue and other non-epithelial cell types [12, 13]. The fourth subgroup, the ER positive subtype was originally found to be one group, but has now been separated into two distinct groups: the luminal A subtype which demonstrated the highest expression of ER, estrogen regulated protein LIV-1, transcription factors hepatocyte nuclear factor 2 alpha (HNF2A), X-box binding protein 1 (XBP1) and GATA-binding protein 3 (GATA3) and the luminal B subtype which demonstrated low to moderate expression of the luminal-specific genes including the estrogen regulated genes mentioned above [13]. These findings were confirmed in an independent gene expression data set [14]. Associations of clinical outcome with the different breast cancer subtypes have also been investigated. The basal-like and ERBB2 positive subtypes were associated with a poor prognosis as assessed by relapse free survival (RFS) [14-16]. A different outcome for tumors classified as luminal A or luminal B was also seen. The luminal B subtype was associated with a worse prognosis with respect to RFS as compared to the luminal A subtype [13, 14]. Another approach to predicting the clinical behavior of breast tumors used a supervised classification method to detect patterns of gene expression. This classification method identified an expression profile of 70 genes that strongly predicts a poor prognosis signature in lymph node negative patients [16]. The poor prognostic classifier correctly predicted disease outcome for 65 out of 78 (83%) patients. Up-regulated in this poor prognosis signature were genes involved in cell cycle, invasion, metastasis, angiogenesis and signal transduction [16]. Surprisingly, the well-known breast cancer associated genes Cyclin D1, ER, Her2/neu and C-myc were not included in the 70-gene classifier. Gene expression profiling of breast cancer has allowed researchers to identify five different breast cancer subtypes and survival analysis has shown significantly different outcomes for each subtype. These five different breast 5 cancer subtypes represent distinct diseases that may be used to accurately predict outcome or treatment response. 1.2 CLINICAL ASPECTS OF OVARIAN CANCER In North America, ovarian cancer is the leading cause of death due to gynecologic malignancies and the fifth most common cause of cancer deaths amongst women in Canada (Figure 1.1). In 2006, an estimated 2,300 women will be diagnosed with ovarian cancer in Canada and 1,600 women will die of this disease [1]. The high death rate associated with ovarian cancer is due to difficulties in detecting early stages of ovarian cancer and the lack of effective therapies for advanced disease. Standard therapy is surgical resection followed by a combination of chemotherapy with platinum and a taxane [17]. There has been a move to a more patient-tailored therapy, whereby specific tumors would be treated with specific drugs. However, due to a lack of understanding of the molecular events involved in ovarian cancer there are no current therapies targeting specific molecular abnormalities in these patients. Ovarian surface epithelial tumors (carcinomas) are heterogeneous and are the most common type of ovarian cancer, accounting for 50-55% of all ovarian tumors, and representing over 90% of all malignant ovarian tumors [18]. Ovarian carcinomas are classified based on cell type into serous, mucinous, endometrioid and clear cell [19]. Highly malignant epithelial tumors lacking any specific differentiation are classified as undifferentiated [20]. Serous carcinomas comprise approximately 50% of all epithelial ovarian cancers and are most often 6 high grade at time of diagnosis [18]. Endometrioid carcinomas are the second most common and account for approximately 20-25% of ovarian carcinomas. Clear cell and mucinous carcinomas are less common and account for less than 10% of all ovarian carcinomas. Approximately 14% of ovarian carcinomas are classified as undifferentiated [18]. 1.2.1 Ovarian cancer prognostic markers The classical clinicopathological prognostic markers for ovarian cancer include age at diagnosis, FIGO stage, histologic subtype, grade of the disease, presence or absence of residual disease after treatment, and response to first line chemotherapy [18, 20-22]. The median age of patients with ovarian cancer is 60 years, and the average lifetime risk for women is 1 in 70 [22]. The present five-year survival rates are around 80-90% for stages la-lc, 70-80% for stages lla-llc, 30-50% for stages llla-lllc and 13% for stage IV. However, the majority of ovarian cancers are found in advanced stage (FIGO stages lll/IV) [20]. With the exception of a subgroup of patients with early stage disease, the majority of patients will require postoperative adjuvant chemotherapy in attempt to eradicate residual disease [22]. However, only 50% of patients with an initial response to chemotherapy will be alive after 5 years. Thus, there is a need to identify new markers including those that could potentially be used to develop novel targeted therapies. 7 1.2.2 Gene expression profiling of ovarian cancer A large number of studies using DNA microarray technology have been performed to determine the gene expression profiles of ovarian cancers. Gene expression profiling has been used to reveal sets of genes that can distinguish normal ovarian epithelium from invasive carcinomas [23-28]. Most of these studies looked at the gene lists that were differentially expressed between normal ovarian tissue and ovarian cancer samples. For example, one study compared 12 malignant ovarian serous papillary carcinoma cell lines to 5 normal ovarian epithelium cell lines. Hierarchical clustering using 299 differentially expressed genes showed that all ovarian serous papillary carcinoma cell lines clustered together [28]. In particular, CD24, TROP-1/EPCAM, and Claudins 3 and 4 were among the most overexpressed genes in the serous carcinoma cell lines as compared to the normal ovarian epithelium cell lines. Lu etal. identified 86 genes that were 3-fold upregulated in ovarian cancer compared to normal ovarian cells [24]. Of these 86 genes, four genes (E2F3, HN1, NOTCH3 and RACGAP1) separated normal from cancer tissue, and the combination of Caludin 3 and VEGF also distinguished all cancers from normal tissue. Thus, numerous genes have been identified as being either up- or downregulated that are likely involved in ovarian carcinogenesis. Other microarray studies have looked at distinguishing histological subtypes based on gene expression [29, 30]. Schwartz etal. found a gene expression profile in clear cell carcinomas that was different from other histological subtypes. Seventy-three genes were associated with a higher expression in clear cell 8 carcinomas compared to other histological subtypes [29]. This was further confirmed in another study, which reported 53 genes that could distinguish clear cell carcinomas from the other histological subtypes [30]. However, it was noted in both studies there was an overlap of genes between the histological subtypes. Many microarray studies have also tried to discover genes associated with clinical outcome. Spentzos era/, found a 115-gene signature in a training cohort of 68 patients, whose signature could distinguish between patients with unfavorable and favorable overall survival in an independent validation set [31]. The signature maintained independent prognostic significance in a multivariate analysis. A similar result distinguishing short-term and long-term survivors was found by Berchuck era/. [32]. The 186-gene classifier correctly identified 19 of 24 long-term survivors and 27 of 30 short-term survivors (all high stage) and correctly classified all 11 early-stage ovarian cancers as long-term survivors. 1.3 HEREDITARY BREAST AND OVARIAN CANCER A family history of breast and ovarian cancer is one of the most significant risk factors for the development of these diseases. Two major breast and ovarian cancer susceptibility genes, BRCA1 (located on chromosome 17q) and BRCA2 (located on chromosome 13q), were identified through the study of familial breast and ovarian cancer kindred's [33, 34]. The BRCA1 gene encodes a protein of 1863 amino acids, and is characterized by the presence of two structural motifs at each of its flanking termini (Figure 1.2) [33, 35]. At the amino terminus, a structurally conserved RING finger domain is found (amino acids 24-64). The RING finger is a zinc-binding motif that is characterized by sets of spatially conserved cysteine and histidine residues. Two proteins, BARD1 and BAP1 have been found to bind to the BRCA1 RING finger domain [36, 37]. The C terminus of BRCA1 was first characterized as a transactivation domain [38, 39]. This region also contains two tandem BRCA1 C-terminal domains (BRCT) at residues 1640-1863. The BRCT domain is found in a diverse group of proteins that have been implicated in DNA repair and cell cycle checkpoint control [40-42]. The BRCT domain in BRCA1 has been found to interact with proteins such as CtIP, and histone deacetylases [43-46]. The BRCA2 gene encodes a protein of 3418 amino acids (Figure 1.2) [34, 47]. Sequence analysis of exon 3 revealed that it shares some sequence similarity with the transactivation domain present in c-Jun, and functional analysis has confirmed a transactivation function within this region [48]. A prominent feature within the BRCA2 amino acid sequence is eight tandem copies of a repetitive sequence motif termed the BRC repeat. Each BRC repeat designated BRC1 to BRC8 is approximately 30 amino acids in length. Six of the eight BRC repeats are conserved and mediate the interaction between BRCA2 and RAD51 [44, 49]. Mutations in BRCA1 and BRCA2 are responsible for 3 to 8% of all cases of breast cancer, and 15-20% of familial cases of breast cancer [50]. For ovarian cancer, approximately 80% of all families with a history of ovarian cancer have a BRCA1 mutation, while 15% have BRCA2 mutations [51]. Mutations in either BRCA1 or BRCA2 confer a lifetime risk of breast cancer of 82% and a lifetime 10 A) BRCA1 1863aa RING domain NLSs BRCT domain BARD1 BAP1 MRE11 RAD50 Nbs1 CtIP RNA pol B) BRCA2 3418 aa Transactivation BRC repeats NLSs z j RAD51 Figure 1.2 Features of the BRCA proteins. The N terminal RING domain, nuclear localization signal (NLS), and the C terminal BRCT domains of BRCA1 are shown, as are the eight BRC repeat motifs for BRCA2 are shown. (Adapted from Venkitaraman A., 2001). risk of ovarian cancer of 54% and 23% [52]. BRCA2 carriers are also at an increased risk of breast cancer in males, prostatic, pancreatic, gall bladder, bile duct and stomach cancers as well as melanoma [53]. 1.4 BRCA1, BRCA2 AND CHROMOSOMAL INSTABILITY Inactivation of both alleles of BRCA1 and BRCA2 is commonly seen in tumors from women harboring germ-line mutations of these genes, indicating that the protein product of these genes may behave as tumor suppressors [54]. The full range of functions of BRCA1 and BRCA2 have yet to be fully elucidated however, there is evidence that they both play key roles in DNA repair pathways. Evidence to suggest that BRCA1 and BRCA2 are involved in DNA repair came from the observation that cells deficient in both BRCA1 and BRCA2 sustain spontaneous chromosomal aberrations that accumulate during passage in culture [55-62]. These abnormalities included not only chromosome and chromatid breaks, but also triradial and quadriradial structures, which are known markers of defective mitotic recombination. These abnormalities are also seen in other human diseases that are associated with defective mitotic recombination such as Bloom's syndrome and Fanconi's anemia [63]. Spectral karyotyping (a chromosome painting technique) analysis of BRCA1 and BRCA2 deficient mouse cells revealed gross chromosomal abnormalities, including translocations, deletions and complex rearrangements [57, 64]. Similar structural abnormalities have been seen in BRCA1 and BRCA2 deficient human cancers [61, 65]. Chromosomal instability is usually associated with growth arrest or increased cell 12 death. Both BRCA1 and BRCA2 deficient mouse cells are associated with proliferative impairment that worsens with successive passage in culture and results in cell cycle arrest in G1 (BRCA2 deficient cells only) and G2/M phases [56, 57]. How then do BRCA1 and BRCA2 mutations ultimately lead to the pathogenesis of breast and ovarian cancer? One answer came from the observation that inactivation of the cell cycle checkpoints, by mutant forms of p53, relieved growth arrest in primary cells homozygous for truncated BRCA2. Tumors from BRCA2 deficient mice also exhibited mutations in p53 [55]. Thus, inactivating mutations in p53, which governs the G1/S cell cycle checkpoint, would allow the cells to bypass critical cell cycle checkpoints and lead to the survival of cells with severe chromosomal abnormalities. Mutations in p53 are commonly observed in BRCA1- and BRCA2-associated breast and ovarian cancer [66, 67]. These observations lend further support to the notion that BRCA induced tumorigenesis accompanied by loss of checkpoint control is necessary for tumor development. 1.5 BRCA1, BRCA2 AND DNA RECOMBINATION AND REPAIR A possible role for BRCA1 and BRCA2 in homologous recombination and DNA repair came from the interaction of BRCA1 and BRCA2 with proteins known to be involved in these processes. One such protein is Rad51, a mammalian homologue of the bacterial protein RecA an essential element for the repair DNA double stranded breaks (DSBs) [68]. In mammalian cells there are three pathways used to repair DSBs: non-homologous end joining (NHEJ), 13 homologous recombination (HR), and single strand annealing (SSA) (Figure 1.3). In NHEJ, broken DNA ends are ligated together without a need for extensive sequence homology [69]. NHEJ of DSBs is accomplished by a set of proteins that are responsible for the successful ligation of broken DNA ends. The main proteins involved are DNA ligase IV, XRCC4, the catalytic subunit of DNA-dependent protein kinase (DNA-PKcs) and Ku protein. NHEJ is initiated by the recognition and binding of the Ku proteins, a heterodimer of Ku70 and Ku80, to the ends of the DSB. The catalytic subunit of the DNA-PKcs is then recruited to the bound Ku70/80 heterodimer. DNA-PKcs is a serine/threonine kinase that belongs to the phosphoinositide 3-kinase (PI3K) family that is thought be a signaling molecule in response to cellular stress [69]. Next, XRCC4/DNA ligase IV ligation complex is then recruited to join the DNA ends together. DNA ligase IV carries out the ligation step, but it requires binding of the XRCC4 to do so. In HR, DSBs are repaired in an error-free manner whereby an intact template of homologous sequence from within a homologous chromosome or a sister chromatid is required [70]. Rad51 is an essential component of the repair of DSBs by HR. Rad51 coats single stranded DNA to form a nucleoprotein filament that can invade and pair with homologous regions of double stranded DNA. This is then followed by strand extension, exchange and repair [63]. The third mechanism, SSA, is considered to be another homology-directed repair pathway. SSA is initiated by homologous pairing however; the homology is between short regions of single stranded DNA (ssDNA) flanking the DSB. This is an error prone 14 1 DSB Ku80/Ku70 DNA-PKcs Non-homologous end joining 1 1 re11/Rad50/Nbs1 End processing NA Ligase IV7XRCC4 1 Ligation Homologous recombination Rad52 re11/Rad50/Nbs1 End processing • Rad51 a m Strand invasion S c c o & Ligation Figure 1.3 DNA double strand break repair pathways. The termini of a DNA DSB introduced by either ionizing radiation or some other means are bound by the Ku heterodimer/DNA-PKcs complex or by hRad52. In NHEJ, the DSB is repaired by DNA ligase IV and XRCC4. Repair of the DSB by homologous recombination is facilitated by Rad51 which initiates DNA strand invasion of the intact sister chromatid. (Adapted from Karan etal., 2000). 15 process resulting in deletions of small regions of DNA [69]. The SSA pathway in mammalian cells relies on a complex of Rad50, Mre11 and Nbs1. Mre11 encodes an exonuclease activity that resects DSBs to generate single stranded DNA for homology directed repair [71]. There are now several lines of evidence to suggest that BRCA2 is essential for the repair of DSBs. A potential role for BRCA2 in DNA repair was first revealed by its interaction with human and mouse Rad51 in yeast two-hybrid screens [49, 72]. Mouse embryos lacking BRCA2 were found to be hypersensitive to radiation a similar trait seen in mouse embryos lacking Rad51 [73]. Extreme sensitivity to other genotoxic agents including ultraviolet (UV) radiation, methanesulfonate and ionizing radiation (IR) has also been reported to be characteristic of BRCA2 deficient cells [49, 56, 72, 74], as well as sensitivity to DNA cross-linking agents, such as mitomycin C (MMC) and cisplatinum [58, 64, 75, 76]. These BRCA2 deficient cells spontaneously accumulate chromosomal breaks and aberrant mitotic exchanges that are indicative of an accumulation of DNA damage during cell division. The interaction of Rad51 with BRCA2 is mediated by eight BRC repeats, of which 6 have been found to directly bind Rad51 [44, 49]. Rad51 characteristically forms nuclear foci following DNA damage and it is thought these foci represent sites of Rad51 dependent repair of damaged DNA [64]. BRCA2 has been reported to co-localize to these Rad51 foci and ionizing radiation induced Rad51 foci are diminished in BRCA2 deficient cells [64, 75] or 16 in cells in which the interaction between BRCA2 and Rad51 is specifically disrupted [49]. Thus, BRCA2 appears to be a participant in Rad51-dependent pathways for HR. Assays to measure the level of homologous recombination in vitro have been developed [77]. These assays utilize a DNA substrate that contains a pair of mutated GFP genes (GFP encodes green fluorescent protein), one of which contains a restriction site for l-Scel, a yeast endonuclease that recognizes and cuts an 18 base pair sequence [77]. Transient transfection of an l-Scel expression vector results in a DSB in the first copy of mutated GFP. The DSB is repaired through homologous gene conversion using a 3'-truncated copy of GFP as a donor sequence. This results in the formation of a functional copy of the GFP gene that is easily detected using fluorescent-activated cell sorting analysis. Such a GFP recombination substrate was used to demonstrate that cells lacking functional BRCA2 are deficient in their ability to repair the l-Scel induced DSBs by homologous recombination [78]. Further supporting evidence that BRCA2 plays a direct role in repair by HR came from studying the interaction of Rad51 with several of the BRC repeats of BRCA2 [79]. BRCA2 was shown to directly regulate the availability and activity of Rad51. One key activity of Rad51 is to coat ssDNA to form a nucleoprotein filament that can invade and pair with homologous DNA to initiate strand exchange between the paired DNA [63]. The formation of Rad51 nucleoprotein filament in vitro is blocked by the BRC3 or BRC4 repeats of BRCA2. Mutations of the BRC3 repeat that mimic mutations found in familial breast cancers failed to block nucleoprotein filament formation. 17 An experiment reported by Davies et al. [79] showed Capan-1 cells (a pancreatic cancer cell line, which carries a naturally occurring 6174delT mutation in one BRCA2 allele accompanied by loss of the wild-type allele) to be defective in transporting Rad51 to the nucleus, suggesting that Rad51 is brought to the nucleus by binding BRCA2. One model that has been put forward to explain these results suggests the BRCA2-Rad51 complex exists in two forms in vivo: an inactive state which prevents Rad51 from binding to single stranded DNA during normal DNA replication and an active state whereby Rad51 can form nucleoprotein filaments and is delivered to sites of damaged DNA by BRCA2 [68, 79, 80]. BRCA1 has also been implicated as a key player in the DNA damage response. BRCA1 deficient HCC1937 cells, BRCA1-null embryonic stem (ES) cells and BRCA1-exon 11 deletion mouse embryonic fibroblasts (MEF) cells all exhibit increased sensitivity to treatment with methyl methanesulfonate, IR, but not UV radiation [57, 81-83]. Reintroduction of a wild-type BRCA1 allele decreased the methyl methanesulfonate and IR sensitivity of BRCA1-deficient cells, suggesting that the response to DNA damage is compromised in patients associated with a BRCA1 mutation [82, 83], Control of the G2/M checkpoint is also defective in BRCA1-exon 11 deletion MEFs, implicating a role of BRCA1 in cell cycle checkpoint control [57]. These studies of the phenotypic characteristics of BRCA1-deficent cells suggest a role for BRCA1 in the DNA damage response via its participation in DNA repair and cell cycle checkpoint control. 18 BRCA1 is also essential for DSB repair by HR, but its role is less clear. A region encompassing residues 758-1064 in BRCA1 has been reported to be involved in the interaction with RAD51 [84]. However, it is still unclear whether the two proteins interact directly. Co-immunoprecipitation from cell extracts revealed an interaction of low stoichiometry. Therefore, BRCA1 may not directly control RAD51 function since the stoichiometry of their interaction is very low. Sites of DNA damage are marked within minutes by the phosphorylation of H2AX, which forms characteristic nuclear foci. BRCA1 is an early migrant to these sites of H2AX phosphorylation [85], which is suggestive of a potential role in DNA repair. Indeed, BRCA1 has been shown to physically associate either directly or indirectly with other proteins that are involved in recombination [83, 86]. BRCA1 interacts with the MRE11/RAD50/Nbs1 protein complex, containing the mammalian homologs of the yeast molecules known to be involved in DSB repair [83]. MRE11 encodes a nuclease activity that resects blunt DSB ends to form ssDNA tracts. BRCA1 can inhibit MRE11 activity [87], which suggests a possible role for BRCA1 in the regulation of the persistent ssDNA generation at sites of DNA repair. ssDNA is a substrate for repair by HR and SSA, and both these processes are defective in BRCA1 deficient cells [59, 88]. The precise role that BRCA1 plays in complex with MRE11/RAD50/Nbs1 still remains to be established. BRCA1 is rapidly phosphorylated after DNA damage in dividing cells, suggesting that BRCA1 may work downstream of critical checkpoint mechanisms which sense and signal DNA damage [63]. Several kinases have been implicated in 19 the transduction of DNA damage signals. In mammalian cells, several kinases, such as ataxia telangiectasia mutated (ATM), ATM- and Rad3-related (ATR), and checkpoint kinase 2 (Chk2) are activated in response to DNA damage [63]. Indeed, BRCA1 is phosphorylated by each of these kinases in response to distinct stimuli and is targeted to specific serine residues [89-91]. BRCA1 becomes hyperphosphorylated after treatment with several DNA damaging agents, including UV (which induces nucleotide lesions), hydroxyurea (which causes strand gaps), IR (which primarily induces DNA breakage), and mitomycin C (DNA cross-linking agent) [35, 43, 84]. ATM and CHK2 phosphorylate BRCA1 after IR and ATR phosphorylates BRCA1 after UV treatment [63]. Multiple phosphorylation sites at serine (S) residues, including S1330, S1423, S1466, and S1542, have been detected by mass spectrometry analysis of BRCA1 peptides phosphorylated in vitro by ATM [89]. CHK2 has been shown to phosphorylate BRCA1 on S988 upon ionizing radiation [90]. In addition, ATR has also been shown to phosphorylate BRCA1 in vitro [92]. Phosphorylation of BRCA1 appears to be important in the response to DNA damage, although how phosphorylation modulates the activities of BRCA1 is still unclear. There is functional evidence linking the kinase activity of ATM to BRCA1, through the BRCA1 associated protein CtlP. Studies have shown that ATM phosphorylates CtlP in vitro and in vivo after treatment with ionizing radiation [43]. This phosphorylation of CtlP by ATM is required for the dissociation of CtlP/CTBP corepressor complex from BRCA1. This dissociation relieves the BRCA1 mediated repression of growth arrest and DNA damage 20 (GADD)45 transcription [43, 93]. Induction of GADD45 results in cell cycle arrest at G1/S or G2/M transition [94]. These studies provide evidence to suggest that BRCA1 works in multiple pathways that signal cell cycle delays in the response to different kinds of DNA lesions. In summary, both BRCA1 and BRCA2 are essential for the repair of DSBs and the overall maintenance of chromosome stability. Inactivation of either BRCA1 or BRCA2 promotes further genetic changes that ultimately lead to tumorigenesis. 1.6 THERAPEUTIC TARGETING OF THE D N A REPAIR DEFECT OF B R C A TUMORS Both BRCA1 and BRCA2 proteins have been implicated in the repair of DSBs, and it has been recently demonstrated how a defect in this repair can be exploited for therapeutic purposes. Two studies have demonstrated the use of inhibitors of poly(ADP-ribose) polymerase (PARP1) to kill tumor cells deficient in either BRCA1 or BRCA2 [95, 96]. PARP1 senses DNA breaks and attracts DNA repair proteins to the site of damage [97-99]. PARP1 is required for the repair of DNA single stranded breaks (SSBs) during base excision repair. In the absence of PARP1 activity the SSBs go unrepaired and are converted to DSBs in dividing cells through the resulting collapse of replication forks [95, 100]. DSBs are normally repaired via RAD51-dependent HR, a process in which both BRCA1 and BRCA2 play an essential role. However, in the absence of BRCA1 or BRCA2 these DSBs are repaired by error prone mechanisms that could result in chromosomal instability. Indeed, BRCA1 and BRCA2 deficient cells were 21 extremely sensitive to treatment with PARP inhibitors, resulting in chromosomal instability, cell cycle arrest and apoptosis [95, 96]. These results demonstrate that inhibition of PARP can specifically lead to the death of BRCA1 and BRCA2 deficient cells, providing an example of how the defect of the cancer can be used for targeted therapy. 1.7 H lSTOPATHOLOGY OF BRCA1 AND BRCA2 TUMORS 1.7.1 Breast Cancer There is increasing evidence that specific clinicopathological features can be recognized in familial breast cancers attributable to mutations in BRCA1 and BRCA2. Studies by the Breast Cancer Linkage Consortium have shown that cancers arising in BRCA mutation carriers differ morphologically from age-matched sporadic breast cancers [101-103]. Others have also described the immunohistochemical characteristics of BRCA-associated breast cancer [104-111]. In addition, DNA microarray technology has also been used to address whether BRCA1 and BRCA2 tumors differ from that of sporadic breast cancer. Hedenfalk etal. compared the gene expression profiles of seven tumors from BRCA1 and BRCA2 mutation carriers with seven tumors from sporadic breast tumors [112]. They found that gene expression profiles for all three tumor types differed significantly from each other, confirming that breast cancers arising in BRCA1 and BRCA2 mutation carriers represent distinct diseases. Through these research efforts the clinicopathological features associated with BRCA1-related breast cancers have been determined. Similar to sporadic breast 22 cancer, invasive ductal carcinoma is the most common histological type in BRCA1-associated tumors [101]. However, of the subtypes, medullary carcinomas are more frequent in patients with BRCA1 mutations (13%) than sporadic breast cancers (3%) [101]. Medullary carcinoma is a form of invasive breast cancer that is characterized by the presence of solid sheets of large pleomorphic cells with indistinct borders that lead to a syncytial appearance [108]. Medullary carcinomas are high-grade tumors with numerous mitoses, and stroma that is characterized by a dense lymphocytic infiltrate. The border of the tumor is well defined and has a pushing edge. Despite being a high-grade tumor, medullary carcinomas are associated with a relatively favorable prognosis [108]. BRCA1 breast carcinomas are more frequently of a higher grade than matched control sporadic breast cancers [101, 105, 106, 113]. These tumors also exhibit a high nuclear pleomorphism, a higher mitotic count and less tubule formation than age-matched sporadic breast tumors [101, 108]. BRCA1-related breast cancers are characteristically estrogen receptor (ER) negative and progesterone receptor negative (PR) [105, 108]. Between 63-90% of BRCA1 associated carcinomas have been reported to be ER negative [103, 104, 106, 109, 113, 114]. The p53 tumor suppressor gene is the most commonly altered gene in human malignancies [105]. Both p53 and BRCA1 are involved in numerous processes such as DNA DSB repair, cell cycle arrest, and apoptosis [108]. A direct link 23 between BRCA1 and p53 came from the observation that loss of p53 could overcome embryonic lethality in BRCA1 deficient mice [115]. Immunohistochemical studies have revealed a higher expression of p53 in BRCA1 carcinomas as compared to sporadic cases. A reported 37-77% of BRCA1 -associated tumors overexpress p53 [66, 103, 106, 111]. Somatic mutations of p53 also occur at a higher frequency in BRCA-associated tumors. Mutations in p53 have been characterized in 56% of BRCA1-associated cases whereas they are only seen in 24% of sporadic cases [105]. Comparison of the spectrum of p53 mutations found in familial BRCA1 patients with those of sporadic cases revealed that about one half of the mutations found in BRCA1-associated cancers had never been reported in the p53 human breast cancer mutations database [105, 108]. Functional characterization of some of these mutations revealed that for the most part they appear to have wild-type functions, but retain transforming activity [116]. Her2/neu has been found amplified and overexpressed in 15 to 30% of breast cancer [9, 10]. Protein overexpression and gene amplification has been associated with a poor prognosis. Her2/neu overexpression has been reported to occur at a very low frequency in BRCA1-associated cancers, ranging from 0-3.7% [102, 111]. Very few studies have looked at the gene amplification status of Her2/neu in familial breast cancers. One study reported that 19% of BRCA1 tumors had low levels of Her2/neu amplification and no cases of high-level amplification were reported [117]. However, two other studies found no evidence of Her2/neu gene amplification among BRCA1 carcinomas [118]. 24 Recent studies have also analyzed cell cycle and apoptosis markers in BRCA associated tumors. Deregulation of apoptosis is a common event in the pathogenesis and progression of breast cancer. Overexpression of the anti-apoptotic gene, BCL2, is commonly seen in ER positive breast cancers and is associated with a good prognosis [108]. BCL2 is expressed at low levels in BRCA1-associated breast cancers as compared to sporadic tumors. On the contrary, caspase 3 levels were higher in BRCA1 tumors. Caspase 3 is a cytosolic enzyme that when activated plays an important role in cell apoptosis [110]. Cell-cycle progression is controlled by cyclin-dependent kinases (CDKs) that are activated by the binding of cyclins and inactivated by CDK inhibitors (CDKI) [108]. Cyclin D1 plays an essential role in the G1/S checkpoint of the cell cycle and interacts with CDK4 and CDK6 to form a complex. Cyclin D1 is considered to be an important oncogene, that is known to be up-regulated by estrogen and progesterone and down-regulated by anti-estrogens [119]. Cyclin D1 is amplified in 15-20% of breast cancers [120-123] and overexpressed in up to 80% of tumors [124-127]. Cyclin D1 overexpression is reported to be a good prognostic factor in invasive breast cancer and is associated with a better patient outcome, particularly in ER positive breast cancers [128, 129]. On the other hand, cyclin D1 amplification has been associated with ER positive cancers and a poor prognosis [121, 130-132]. Several studies have shown a lower frequency of Cyclin D1 expression levels in BRCA1-associated breast cancers as compared 25 to sporadic tumors [104, 110]. The study of cyclin D1 amplification in BRCA1 carcinomas has yielded discordant results. One study did not find any evidence of cyclin D1 amplification in 30 BRCA1 carcinomas [133], whereas another study identified cyclin D1 amplification in 18% of BRCA1 cases, a percentage similar to that found in sporadic breast cancer [110]. Overexpression of cyclins E and A have been associated with the BRCA1 phenotype [107, 110]. Both cyclin E and A also play an essential role in regulating the transition from the G1 phase to the S phase of the cell cycle. Cyclin E is negatively regulated by p27, a member of the kinase inhibitor protein (KIP) family [108]. High levels of cyclin E protein expression and low levels of p27 protein expression have been reported in BRCA1 carcinomas [105, 107, 110]. Expression studies using immunohistochemistry and cDNA microarray technology have demonstrated a subtype of breast cancer that is associated with basal cell markers, such as cytokeratins 5/6 (CK5/6) [13, 134-136]. These tumors are also characterized as being poorly differentiated and ER negative, which are two characteristics of BRCA1-associated breast cancers. Gene expression profiling of breast cancers led to the discovery of four new subtypes of breast cancer. One new subtype identified was the basal subtype, which was characterized as ER negative, Her2/neu negative, and positive for the expression of cytokeratins 5 and 6 [12, 16,137]. Previous immunohistochemical and molecular studies of basal-like breast cancers 26 revealed that these tumors are often high grade, have a medullary phenotype, and are associated with frequent mutations in p53 [13, 138, 139]. Most studies with clinical outcome have indicated that basal-like breast cancers are associated with a poor prognosis [13, 135, 136]. These features are very similar to those observed in BRCA1-associated breast cancers. Indeed, cDNA microarray studies showed that most BRCA1 carcinomas exhibited a basal-like gene expression profile [14]. Subsequent immunohistochemical studies have also confirmed this finding. All studies showed that the expression of basal markers were significantly more frequent in BRCA1 tumors as compared to sporadic cases [110, 140, 141]. Determining the clinicopathological phenotype of BRCA2-associated breast cancers has proven to be more difficult. As is the case of BRCA1 and sporadic carcinomas, invasive ductal carcinoma is the most common histological subtype associated with BRCA2-mutation carriers [108]. A single study reported a higher incidence of tubular-lobular carcinomas in BRCA2 patients [142], whereas another study reported an association between BRCA2-mutations and pleomorphic lobular carcinomas [143]. However, in other studies a significant difference between BRCA2 mutations and specific histological types of breast cancer was not observed [101, 102]. Similar to BRCA1-associated cancers, BRCA2-related tumors also tend to be of higher grade when compared to sporadic age-matched control tumors, although this is a weaker association. BRCA2 tumors did show less tubule 27 formation, but there was no significant difference in pleomorphism or mitotic count [105, 108]. Overall, ER and PR status in BRCA2-associated cancers does not differ significantly from that of sporadic breast cancers [103, 106, 109, 113, 114,143]. Between 57-68% of BRCA2-associated breast cancer have been reported to be ER positive [103, 106, 109, 143]. Studies of BRCA2-associated tumors revealed that p53 is expressed at a lower frequency in these tumors as compared to BRCA1. Most studies have found p53 overexpression in 20% of BRCA2 carcinomas [104-106, 110]. Mutations in p53 have been reported to be present in 20-63% [65, 66]. However, one study revealed a lack of p53 mutations in their series of BRCA2 tumors [104]. The result of HER-2 expression studies in BRCA2-associated tumors varies. However, most studies found no differences in the expression of HER2 in BRCA2 and age-matched sporadic breast cancers [104-106]. With respect to apoptotic markers, overexpression of BCL2 was observed in BRCA2 tumors [110]. Overexpression of BCL2 is commonly seen in ER positive breast cancers associated with a good prognosis [108]. Cyclin D1 overexpression is commonly observed in breast cancer and several studies have evaluated expression in hereditary breast cancer. Expression of cyclin D1 in BRCA2 tumors occurs at a frequency in-between that of BRCA1 and sporadic breast cancer. For instance, Osin et al. [144] observed that cyclin D1 28 was expressed in 27% of BRCA2-associated cancers as compared to 5% of BRCA1 and 35% of sporadic cancers. Another study by Armes et al. reported overexpression of cyclin D1 in 33% of BRCA1 tumors, 56% of BRCA2 tumors and 100% of sporadic tumors [104]. However the number of BRCA1 and BRCA2 tumors in this study was very small, with only 9 cases of each. In another study, overexpression of cyclin D1 was seen in 5%, 57%, and 56% of BRCA1, BRCA2 and sporadic breast tumors [110]. Only one study has evaluated the frequency of cyclin D1 gene amplification in BRCA2 tumors. This study found that cyclin D1 was amplified in 60% of BRCA2 tumors as compared to 18% of BRCA1 and 36% of sporadic tumors [110]. A drawback to this study was the small number of BRCA cases (5 BRCA2 and 11 BRCA1 tumors) analyzed. A larger study is required to determine the status of cyclin D1 gene amplification in BRCA1- and BRCA2-associated breast cancers. 1.7.2 Ovarian Cancer Most ovarian cancers associated with germline mutations in BRCA1 and BRCA2 are invasive serous carcinomas [145-152]. The reported frequency of serous carcinomas in BRCA associated ovarian cancers is variable. For instance, one publication reported all 32 BRCA1- and BRCA2-associated cancers were invasive serous carcinomas [150], whereas other studies have observed a lower frequency of serous carcinomas in BRCA-associated tumors (ranging from 25-93%) [145-149, 151, 152]. 29 Even though serous carcinomas predominate, other histological subtypes, such as endometrioid carcinomas have been associated with BRCA mutation carriers [141, 145, 151]. Studies have also shown that borderline, low grade, and mucinous tumors are not part of the tumor phenotype associated with BRCA1-and BRCA2-associated cancers[141, 145, 149, 151], suggesting that mutations in either BRCA1 or BRCA2 do not play a role in the development of these tumors. Mutations of p53 are also another common feature of tumors arising in BRCA1 and BRCA2 mutation carriers, occurring in 70-80% of ovarian cancers [67, 153, 154]. Although BRCA1- and BRCA2-associated ovarian cancers are more likely to be of higher stage and grade than sporadic cancers, several reports have shown that these tumors have an improved survival as compared to sporadic cases. Patients with germline mutations in BRCA1 and BRCA2 have also shown a significantly longer overall survival and recurrence-free interval following chemotherapy compared to sporadic cancers [145, 155, 156]. The favorable response to chemotherapy is most likely due to the increased sensitivity of BRCA-deficient tumor cells to DNA damaging agents, such as cisplatin, which produce DNA DSBs [145]. cDNA microarray technology has also been used to examine the role of BRCA1 and BRCA2 mutations in ovarian carcinogenesis by comparing gene expression patterns of ovarian cancers associated with germline mutations in BRCA1 and BRCA2 with that of sporadic ovarian cancers. This investigation revealed that 30 BRCA1- and BRCA2-associated tumors display distinct gene expression profiles. Indeed, 110 genes showed significantly different expression levels between the two subtypes of ovarian cancer [157]. One remarkable finding from this study was that the gene expression profiles of sporadic tumors appeared to share features of BRCA1- and BRCA2-associated cancers. The observation that sporadic ovarian tumors could be segregated into BRCA1-like and BRCA2-like subgroups indicates that these genes most likely play a role in hereditary and sporadic ovarian cancer. 1.8 BRCA1 AND BRCA2 IN HEREDITARY CANCER VERSUS SPORADIC BREAST AND OVARIAN CANCER Numerous studies have looked at the involvement of BRCA1 and BRCA2 in sporadic breast and ovarian cancer, however, despite exhaustive searches, somatic mutations of BRCA1 and BRCA2 appear to be rare events in sporadic breast and ovarian cancer. Somatic mutations of BRCA1 have been reported in 0-2.5% of sporadic breast cancer [158,159], and in 5-9% of sporadic ovarian cancer [158, 160-163]. BRCA2 somatic mutations have been reported in 1% of sporadic breast cancer [164] and 1-6% of sporadic ovarian cancer [33, 162, 165, 166]. Allelic loss of heterozygosity for either BRCA1 or BRCA2 are common events in both breast and ovarian cancer, suggesting that there could be other epigenetic mechanisms responsible for somatic inactivation of BRCA1 and BRCA2 [158, 165-167]. One of the most widely studied epigenetic events is the aberrant methylation of cytosine residues in CpG dinucleotides. Although under-31 represented in the genome, islands of CpG dinucleotides are found in the promoters region of a substantial proportion of genes [168]. Typically, these CpG islands are unmethylated, however aberrant methylation of these CpG islands results in transcriptional silencing. Inactivation of the BRCA1 locus due to promoter hypermethylation has been found in 11-15% percent of sporadic breast cancer [169-171], and in 5-31% of sporadic ovarian cancer [161,169,170, 172]. No such epigenetic inactivation by hypermethylation has been found at the BRCA2 locus in sporadic breast or ovarian cancer [162, 166, 173]. A potential alternative mechanism for inactivation of BRCA2 has recently been described and involves the amplification of a novel gene called EMSY. 1.9 E M S Y AMPLIFICATION IN BREAST AND OVARIAN CANCER EMSY was first discovered using yeast two-hybrid screens to identify novel BRCA2 interacting proteins. Yeast two-hybrid screens with small overlapping sequences from the BRCA2 N terminus identified sixteen interacting clones (Figure 1.4) [174]. All sixteen clones were fragments of the same cDNA sequence, encoding a protein that has now been called EMSY (since the word "SISTER" appears in the first line of the protein sequence, the protein was named after the first author's sister who is a breast cancer nurse). These results showed that a region at the amino terminus of EMSY (corresponding to amino acids 1-80 of the predicted EMSY protein) binds within residues 18-46 in the third exon of BRCA2. This interaction between EMSY and BRCA2 was confirmed using a bacterially expressed recombinant glutathione-S-transferase (GST) 32 3RCA2 3418 I >i: .» _ « _ T ^ r r w " i T i T ! g " : a c t i v a t i o n domain a c t i v a t i n g r a a i d u a a T f 2 3 - A D ^ ^ f LNW^EI jSSEAP PYN - 4 4 EMSY b i n d i n g Figure 1.4 Sequence from the BRCA2 N terminus used for yeast two hybrid screens. The complete exonic sequence of BRCA2 is expanded to show the transactivation domain, which is further expanded to show the key residues involved in the BRCA2-EMSY interaction. (Hughes-Davies etal., 2003). 33 B R C A 2 (1-197) fusion protein to pull down the radiolabeled N-terminal region of EMSY (1-478). The N terminus region of EMSY was also shown to repress B R C A 2 transcriptional activity. These results suggest that EMSY has the capacity to bind within the BRCA2 exon 3 transactivation domain and may act to silence BRCA2 activity. A recent study describing a novel BRCA2 interacting protein, PALB2, failed to show an interaction between EMSY and BRCA2 [175]. However, this group used a different assay system (GAL4 fusions) for their transfection pull downs, which may explain the dissimilarity in the results. EMSY is exclusively nuclear, and relocalizes after DNA damage to y-H2AX foci. H2AX is essential for the recognition and repair of DNA DSBs, and rapidly becomes phosphorylated to y-H2AX at the site of each nascent DSB, resulting in characteristic nuclear foci [176]. BRCA2 is also known to relocalize to DSB repair sites after DNA damage, and these results suggest a common function for EMSY and BRCA2. EMSY encodes a predicted protein of 1322 amino acids and although it is a large protein the sequence is entirely novel. There are no other EMSY-like proteins in the human genome and only a single EMSY gene can be found in frogs and fish. However, an 80 amino acid conserved domain at the amino terminus of EMSY has been identified, which has been termed the EMSY N-Terminal domain (ENT domain)(Figure 1.5A). Analysis of this sequence in the human genome revealed that it is unique, found only in EMSY and no other protein. Arabidopsis has 9 different proteins, which possess an ENT domain (Figure 1.5B). Examination of these genes showed that they all contain a new Tudor-like domain, which has 34 A) Multiple sequence alignment of all known ENT domains Huun D B T If DO rroj D O T Pi«h DMT - — i J Off protain Mosquito DTT protain Aootlmot ooia CNT protain Strongy. -- DTT protain Plant DR protais At3g20030 Plant OR ptotam AtJj57970 Plant DTT protain >,• - ; ! ' Nant DrT piotain Attg32«40 S Plant OR piotain At >q\i:*< M Plant DTT protain At2o;44«40 at Plant nrr protain At5gl3020 K Plant DTT protain V.5-:j:4"8C Plant BIT protain Al3«S7JJ0 t l ; p 3 D » A p r s tt m •QIWO..: •M: m>Qi. a UOBH:! l a ALL Of .fttajbixssv J E:.P..UMIV.> •c«p .bhppl i . B) AI.V?»» \l4al2MO I \ I domain proteins in plants ENT ENT ENT C) The R o v i l Family ENT ancestral domain ENT ENT ENT Agenet Tudor Chromo *l5a*r»n ENT Figure 1.5 The EMSY ENT domain. A) Multiple sequence alignment of the ENT domains. B) Schematic of the eight Arabidopsis ENT-Agenet domains. The ENT domain is shown in pink, and the Agenet domain in green. C) Diagram of the Royal Family of domains, illustrating the possible evolutionary divergence from a single ancestral domain. The Royal Family consists of five domains that share sequence similarity. These are the PWWP, Tudor, Agenet, and the chromodomain. (Hughes-Davies et al., 2003). 35 been called the Agenet domain [174]. The Agenet domain is part of the Royal Family of protein domains, which include Chromo, PWVVP, MBT, Tudor and Agenet [177] (Figure 1.5C). These domains are often found in chromatin regulating proteins. In plants, proteins containing ENT domains almost always contain an Agenet domain. However, EMSY does not contain such a domain, but it does associate with proteins that contain Royal Family domains. These include heterochromatin-associated protein-1 (HP1) (which has a chromodomain) and BS69 (which has a PWWP domain). This suggests that EMSY may function in part in the regulation of chromatin. The EMSY gene maps to 11q13.5, a region commonly amplified in numerous human malignancies, including breast, ovarian, head and neck, bladder and esophageal cancer [121, 126, 178-181] (Figure 1.6). Fluorescent in situ hybridization was performed on breast cancer cell lines to determine whether EMSY is amplified in breast cancer. EMSYwas found amplified in five out of 28 (18%) breast cancer cell lines. Quantitative RT-PCR confirmed that the cell lines with EMSY gene amplification were also associated with the highest levels of EMSY expression [174]. The crystal structure of the EMSY ENT domain has been recently described [182, 183]. It was discovered that the first 98 residues of EMSY contain the ENT domain, which is composed of a unique arrangement of 5 a-helices that fold into a helical bundle formation [182]. The central core region of the ENT domain was 36 Chromosome 11 in Figure 1.6 The 11q13 amplicon This region is commonly found amplified in numerous human malignancies, including breast, ovarian and head and neck cancer. Several oncogenes have been mapped to this region on 11q13. 37 found to closely resemble structurally the DNA-binding homeodomain proteins, which contains a small DNA binding motif [184]. One other key finding from this study was that the ENT domain forms a homodimer by the anti-parallel packing of the long N-terminal a-helix from each subunit [182]. The structure of the ENT domain suggests that the EMSY protein exists as a dimer, which could allow it to bind to several partner proteins simultaneously. Further study of the EMSY ENT domain and its interaction with the chromoshadow domain of HP16 (HP1-CSD) revealed that EMSY is bound by two HP1-CSD homodimers [183, 185]. The lack of BRCA2 mutations in sporadic breast cancer has long been a mystery. It is possible that EMSY overexpression could lead to loss of BRCA2 function, which would provide an explanation as to why certain sporadic breast cancers do not have BRCA2 mutations. Mutations of BRCA2 may be an unnecessary event when EMSY is overexpressed [186, 187] (Figure 1.7). Indeed, several researchers have suggested that it would be interesting to learn whether sporadic breast cancers with EMSY amplification and inherited BRCA2 mutation carriers share common biological and pathological features that would reflect a common defect in BRCA2 function [168, 187]. 38 A) Normal B) BRCA2b reas t C) EMSY overexpressing carcinoma cell sporadic breast cancer Intact BRCA2 alleles domain Sufficient BRCA2 Loss of BRCA2 function Loss of BRCA2 function F igure 1.7 E M S Y a n d B R C A 2 . A) Normal breast epithelium produce normal physiological amounts of BRCA2 that is under functional control by limiting quantities of EMSY. This leads to adequate suppression of BRCA2. B) Familial breast cancer associated with a germline mutation in one allele, and loss of function of BRCA2 in the other allele. C) Sporadic breast cancer that is amplified and overexpressed for EMSY, producing physiological amounts of EMSY. This results in an efficient EMSY-BRCA2 interaction resulting in loss of BRCA2 function. (Adapted from Livingston, 2004). 39 1 . 1 0 SCOPE OF HYPOTHESES 1.10.1 Hypotheses If amplification and overexpression of EMSY is a fundamental event in the development of breast and ovarian cancer then we would expect EMSY amplification and overexpression to be a clinically significant event and share similar biological features associated with loss of BRCA2 function. 1.10.2 Rationale and Specific Objectives A large component of this work was dependent on the development of a fluorescent in situ hybridization (FISH) assay for gene amplification that could be applied to paraffin embedded tissue microarrays (TMAs). Thus, the first year of my graduate studies was dedicated to the development and validation of such an assay. Chapter 2 reviews the use of FISH on TMAs and describes the problems and solutions encountered during the development of this assay: Also included in this chapter is a description of an automated system that we implemented in our lab for the analysis of gene amplification events. The development of this FISH assay was required in order to carry out the first two objectives of this thesis. The first objective was to determine the frequency and clinical significance of EMSY gene amplification in breast cancer. Chapter 3 describes the first studies of EMSY amplification in a large cohort of sporadic breast cancers in which clinical outcome was available for all patients. The findings were then validated in a larger population based series of breast cancers. EMSY gene amplification 40 was also assessed in familial breast cancers associated with BRCA1 and BRCA2 mutations. The second objective was to determine the frequency of EMSY amplification in ovarian cancer and its involvement in the emergence and maintenance of the 11q13 amplicon in ovarian cancer. Chapter 4 describes the first study evaluating EMSY gene amplification in ovarian cancer. A detailed investigation of the genes involved in the formation of the 11q13 amplicon is also described. A more comprehensive study of the 11q13 amplicon and the genes involved are described in Chapter 5. For this study we were also able to evaluate the prognostic significance of amplification of five genes implicated as potential drivers of the 11q13 amplicon. The third objective, described in Chapter 6 and Chapter 7, was to determine whether EMSY overexpression in primary breast and ovarian surface epithelial cells share similar biological features that are associated with BRCA2 deficient cells. 41 1.11 REFERENCES 1. Canada, C.C.S.N.C.I.o., Canadian Cancer Society/National Cancer Institute of Canada: Canadian Cancer Statistics 2006. 2006. 2. Weigelt, B., J.L. Peterse, and L.J. van't Veer, Breast cancer metastasis: markers and models. Nat Rev Cancer, 2005. 5(8): p. 591-602. 3. Carter, C.L., C. Allen, and D.E. Henson, Relation of tumor size, lymph node status, and survival in 24,740 breast cancer cases. Cancer, 1989. 63(1): p. 181-7. 4. Page, D.L., Prognosis and breast cancer. Recognition of lethal and favorable prognostic types. Am J Surg Pathol, 1991. 15(4): p. 334-49. 5. Rosen, P.P., et al., Pathological prognostic factors in stage I (T1N0M0) and stage II (T1N1M0) breast carcinoma: a study of 644 patients with median follow-up of 18 years. J Clin Oncol, 1989. 7(9): p. 1239-51. 6. Tamoxifen for early breast cancer: an overview of the randomised trials. Early Breast Cancer Trialists' Collaborative Group. Lancet, 1998. 351(9114): p. 1451-67. 7. Come, S.E., et al., Second International Conference on Recent Advances and Future Directions in Endocrine Manipulation of Breast Cancer: summary consensus statement. Clin Cancer Res, 2003. 9(1 Pt 2): p. 443S-6S. 8. Jordan, V.C. and C S . Murphy, Endocrine pharmacology of antiestrogens as antitumor agents. Endocr Rev, 1990.11(4): p. 578-610. 9. Slamon, D.J., et al., Studies of the HER-2/neu proto-oncogene in human breast and ovarian cancer. Science, 1989. 244(4905): p. 707-12. 10. Slamon, D.J., et al., Use of chemotherapy plus a monoclonal antibody against HER2 for metastatic breast cancer that overexpresses HER2. N Engl J Med, 2001. 344(11): p. 783-92. 11. Perou, C M . , et al., Distinctive gene expression patterns in human mammary epithelial cells and breast cancers. Proc Natl Acad Sci U S A , 1999. 96(16): p. 9212-7. 12. Perou, C M . , et al., Molecular portraits of human breast tumours. Nature, 2000. 406(6797): p. 747-52. 13. Sorlie, T., et al., Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proc Natl Acad Sci U S A , 2001. 98(19): p. 10869-74. 42 14. Sorlie, T., et al., Repeated observation of breast tumor subtypes in independent gene expression data sets. Proc Natl Acad Sci U S A, 2003. 100(14): p. 8418-23. 15. Sotiriou, C , et al., Breast cancer classification and prognosis based on gene expression profiles from a population-based study. Proc Natl Acad Sci U S A , 2003. 100(18): p. 10393-8. 16. van't Veer, L. J., et al., Gene expression profiling predicts clinical outcome of breast cancer. Nature, 2002. 415(6871): p. 530-6. 17. Piccart, M.J., et al., Randomizedintergroup trial of cisplatin-paclitaxel versus cisplatin-cyclophosphamide in women with advanced epithelial ovarian cancer: three-year results. J Natl Cancer Inst, 2000. 92(9): p. 699-708. 18. Chen, V.W., et al., Pathology and classification of ovarian tumors. Cancer, 2003. 97(10 Suppl): p. 2631-42. 19. Auersperg, N., et al., Ovarian surface epithelium: biology, endocrinology, and pathology. Endocr Rev, 2001. 22(2): p. 255-88. 20. Heintz, A.P., et al., Carcinoma of the ovary. Int J Gynaecol Obstet, 2003. 83 Suppl 1: p. 135-66. 21. Crijns, A.P., et al., Molecular prognostic markers in ovarian cancer: toward patient-tailored therapy. Int J Gynecol Cancer, 2006. 16 Suppl 1: p. 152-65. 22. Cannistra, S.A., Cancer of the ovary. N Engl J Med, 2004. 351(24): p. 2519-29. 23. Hibbs, K., et al., Differential gene expression in ovarian carcinoma: identification of potential biomarkers. Am J Pathol, 2004. 165(2): p. 397-414. 24. Lu, K.H., et al., Selection of potential markers for epithelial ovarian cancer with gene expression arrays and recursive descent partition analysis. Clin Cancer Res, 2004. 10(10): p. 3291-300. 25. Schummer, M., et al., Comparative hybridization of an array of 21,500 ovarian cDNAs for the discovery of genes overexpressed in ovarian carcinomas. Gene, 1999. 238(2): p. 375-85. 26. Tonin, P.N., et al., Microarray analysis of gene expression mirrors the biology of an ovarian cancer model. Oncogene, 2001. 20(45): p. 6617-26. 43 27. Welsh, J.B., et al., Analysis of gene expression profiles in normal and neoplastic ovahan tissue samples identifies candidate molecular markers of epithelial ovarian cancer. Proc Natl Acad Sci U S A , 2001. 98(3): p. 1176-81. 28. Santin, A.D., et al., Gene expression profiles in primary ovarian serous papillary tumors and normal ovarian epithelium: identification of candidate molecular markers for ovarian cancer diagnosis and therapy. Int J Cancer, 2004. 112(1): p. 14-25. 29. Schwartz, D.R., et al., Gene expression in ovarian cancer reflects both morphology and biological behavior, distinguishing clear cell from other poor-prognosis ovarian carcinomas. Cancer Res, 2002. 62(16): p. 4722-9. 30. Zorn, K.K., et al., Gene expression profiles of serous, endometrioid, and clear cell subtypes of ovarian and endometrial cancer. Clin Cancer Res, 2005. 11(18): p. 6422-30. 31. Spentzos, D., et al., Gene expression signature with independent prognostic significance in epithelial ovarian cancer. J Clin Oncol, 2004. 22(23): p. 4700-10. 32. Berchuck, A., et al., Patterns of gene expression that characterize long-term survival in advanced stage serous ovarian cancers. Clin Cancer Res, 2005. 11(10): p. 3686-96. 33. Miki, Y., et al., A strong candidate for the breast and ovarian cancer susceptibility gene BRCA1. Science, 1994. 266(5182): p. 66-71. 34. Wooster, R., et al., Identification of the breast cancer susceptibility gene BRCA2. Nature, 1995. 378(6559): p. 789-92. 35. Chen, Y., et al., BRCA1 is a 220-kDa nuclearphosphoprotein that is expressed and phosphorylated in a cell cycle-dependent manner. Cancer Res, 1996. 56(14): p. 3168-72. 36. Jensen, D.E., et al., BAP1: a novel ubiquitin hydrolase which binds to the BRCA1 RING finger and enhances BRCA1-mediated cell growth suppression. Oncogene, 1998. 16(9): p. 1097-112. 37. Wu, L.C, et al., Identification of a RING protein that can interact in vivo with the BRCA1 gene product. Nat Genet, 1996. 14(4): p. 430-40. 38. Chapman, M.S. and I.M. Verma, Transcriptional activation by BRCA1. Nature, 1996. 382(6593): p. 678-9. 44 39. Monteiro, A.N., A. August, and H. Hanafusa, Evidence for a transcnptional activation function ofBRCAl C-terminal region. Proc Natl Acad Sci U S A , 1996. 93(24): p. 13595-9. 40. Bork, P., et al., A superfamily of conserved domains in DNA damage-responsive cell cycle checkpoint proteins. Faseb J , 1997. 11(1): p. 68-76. 41. Callebaut, I. and J.P. Mornon, From BRCA 1 to RAP1: a widespread BRCT module closely associated with DNA repair. FEBS Lett, 1997. 400(1): p. 25-30. 42. Koonin, E.V., S.F. Altschul, and P. Bork, BRCA1 protein products... Functional motifs. Nat Genet, 1996. 13(3): p. 266-8. 43. Li, S., et al., Binding of CtIP to the BRCT repeats of BRCA 1 involved in the transcription regulation ofp21 is disrupted upon DNA damage. J Biol Chem, 1999. 274(16): p. 11334-8. 44. Wong, A.K., et al., RAD51 interacts with the evolutionary conserved BRC motifs in the human breast cancer susceptibility gene brca2. J Biol Chem, 1997. 272(51): p. 31941-4. 45. Yarden, R.I. and L.C. Brody, BRCA1 interacts with components of the histone deacetylase complex. Proc Natl Acad Sci U S A , 1999. 96(9): p. 4983-8. 46. Yu, X., et al., The C-terminal (BRCT) domains ofBRCAl interact in vivo with CtIP, a protein implicated in the CtBP pathway of transcnptional repression. J Biol Chem, 1998. 273(39): p. 25388-92. 47. Bertwistle, D., et al., Nuclear location and cell cycle regulation of the BRCA2 protein. Cancer Res, 1997. 57(24): p. 5485-8. 48. Milner, J., et al., Transcnptional activation functions in BRCA2. Nature, 1997. 386(6627): p. 772-3. 49. Chen, P.L., et al., The BRC repeats in BRCA2 are critical for RAD51 binding and resistance to methyl methanesulfonate treatment. Proc Natl Acad Sci U S A , 1998. 95(9): p. 5287-92. 50. Easton, D.F., Familial risks of breast cancer. Breast Cancer Res, 2002. 4(5): p. 179-81. 51. Adami, H.O., etal., Panty, age at first childbirth, and n'sk of ovan'an cancer. Lancet, 1994. 344(8932): p. 1250-4. 45 52. King, M.C., J.H. Marks, and J.B. Mandell, Breast and ovarian cancer risks due to inherited mutations in BRCA1 and BRCA2. Science, 2003. 302(5645): p. 643-6. 53. Cancer risks in BRCA2 mutation carriers. The Breast Cancer Linkage Consortium. J Natl Cancer Inst, 1999. 91(15): p. 1310-6. 54. Bennett, I.C., M. Gattas, and B.T. Teh, The management of familial breast cancer. Breast, 2000. 9(5): p. 247-63. 55. Lee, H., et al., Mitotic checkpoint inactivation fosters transformation in cells lacking the breast cancer susceptibility gene, Brca2. Mol Cell, 1999. 4(1): p. 1-10. 56. Patel, K.J., et al., Involvement ofBrca2 in DNA repair. Mol Cell, 1998. 1(3): p. 347-57. 57. Xu, X., et al., Centrosome amplification and a defective G2-M cell cycle checkpoint induce genetic instability in BRCA 1 exon 11 isoform-deficient cells. Mol Cell, 1999. 3(3): p. 389-95. 58. Kraakman-van der Zwet, M., et al., Brca2 (XRCC11) deficiency results in radioresistant DNA synthesis and a higher frequency of spontaneous deletions. Mol Cell Biol, 2002. 22(2): p. 669-79. 59. Moynahan, M.E., T.Y. Cui, and M. Jasin, Homology-directed dna repair, mitomycin-c resistance, and chromosome stability is restored with correction of a Brcal mutation. Cancer Res, 2001. 61(12): p. 4842-50. 60. Shen, S.X., et al., A targeted disruption of the murine Brcal gene causes gamma-irradiation hypersensitivity and genetic instability. Oncogene, 1998. 17(24): p. 3115-24. 61. Tirkkonen, M., et al., Distinct somatic genetic changes associated with tumor progression in carriers ofBRCAl and BRCA2 germ-line mutations. Cancer Res, 1997. 57(7): p. 1222-7. 62. Turner, B.C., et al., The fragile histidine triad/common chromosome fragile site 3B locus and repair-deficient cancers. Cancer Res, 2002. 62(14): p. 4054-60. 63. Venkitaraman, A.R., Cancer susceptibility and the functions ofBRCAl and BRCA2. Cell, 2002. 108(2): p. 171-82. 64. Yu, V. P., et al., Gross chromosomal rearrangements and genetic exchange between nonhomologous chromosomes following BRCA2 inactivation. Genes Dev, 2000. 14(11): p. 1400-6. 46 65. Gretarsdottir, S., et al., BRCA2 andp53 mutations in primary breast cancer in relation to genetic instability. Cancer Res, 1998. 58(5): p. 859-62. 66. Crook, T., et al., p53 mutation with frequent novel condons but not a mutator phenotype in BRCA1- and BRCA2-associated breast tumours. Oncogene, 1998. 17(13): p. 1681-9. 67. Ramus, S.J., et al., Increased frequency of TP53 mutations in BRCA1 and BRCA2 ovarian tumours. Genes Chromosomes Cancer, 1999. 25(2): p. 91-6. 68. Venkitaraman, A.R., Chromosome stability, DNA recombination and the BRCA2 tumour suppressor. Curr Opin Cell Biol, 2001.13(3): p. 338-43. 69. Karran, P., DNA double strand break repair in mammalian cells. Curr Opin Genet Dev, 2000. 10(2): p. 144-50. 70. Lacroix, M. and G. Leclercq, The "portrait" of hereditary breast cancer. Breast Cancer Res Treat, 2005. 89(3): p. 297-304. 71. Haber, J.E., The many interfaces ofMre11. Cell, 1998. 95(5): p. 583-6. 72. Sharan, S. K., et al., Embryonic lethality and radiation hypersensitivity mediated by Rad51 in mice lacking Brca2. Nature, 1997. 386(6627): p. 804-10. 73. Sharan, S.K., et al., BRCA2 deficiency in mice leads to meiotic impairment and infertility. Development, 2004. 131(1): p. 131-42. 74. Connor, F., et al., Tumorigenesis and a DNA repair defect in mice with a truncating Brca2 mutation. Nat Genet, 1997. 17(4): p. 423-30. 75. Yuan, S.S., et al., BRCA2 is required for ionizing radiation-induced assembly of Rad51 complex in vivo. Cancer Res, 1999. 59(15): p. 3547-51. 76. Donoho, G., et al., Deletion ofBrca2 exon 27 causes hypersensitivity to DNA crosslinks, chromosomal instability, and reduced life span in mice. Genes Chromosomes Cancer, 2003. 36(4): p. 317-31. 77. Pierce, A.J., et al., XRCC3 promotes homology-directed repair of DNA damage in mammalian cells. Genes Dev, 1999. 13(20): p. 2633-8. 78. Moynahan, M.E., A.J. Pierce, and M. Jasin, BRCA2 is required for homology-directed repair of chromosomal breaks. Mol Cell, 2001. 7(2): p. 263-72. 47 79. Davies, A.A., et al., Role of BRCA2 in control of the RAD51 recombination and DNA repair protein. Mol Cell, 2001. 7(2): p. 273-82. 80. Venkitaraman, A.R., Functions of BRCA 1 and BRCA2 in the biological response to DNA damage. J Cell Sci, 2001.114(Pt 20): p. 3591-8. 81. Gowen, L.C., et al., BRCA 1 required for transcription-coupled repair of oxidative DNA damage. Science, 1998. 281(5379): p. 1009-12. 82. Scully, R., et al., Genetic analysis of BRCA 1 function in a defined tumor cell line. Mol Cell, 1999. 4(6): p. 1093-9. 83. Zhong, Q., et al., Association ofBRCAl with the hRad50-hMre11-p95 complex and the DNA damage response. Science, 1999. 285(5428): p. 747-50. 84. Scully, R., et al., Association ofBRCAl with Rad51 in mitotic and meiotic cells. Cell, 1997. 88(2): p. 265-75. 85. Paull, T.T., et al., A critical role for histone H2AXin recruitment of repair factors to nuclear foci after DNA damage. Curr Biol, 2000.10(15): p. 886-95. 86. Wang, Y., et al., BASC, a super complex of BRCA1-associated proteins involved in the recognition and repair of aberrant DNA structures. Genes Dev, 2000. 14(8): p. 927-39. 87. Paull, T.T., et al., Direct DNA binding byBrcal. Proc Natl Acad Sci U S A , 2001. 98(11): p. 6086-91. 88. Moynahan, M.E., et al., Brcal controls homology-directed DNA repair. Mol Cell, 1999. 4(4): p. 511-8. 89. Cortez, D., et al., Requirement of ATM-dependent phosphorylation of brca 1 in the DNA damage response to double-strand breaks. Science, 1999. 286(5442): p. 1162-6. 90. Lee, J.S., et al., hCdsl-mediated phosphorylation of BRCA 1 regulates the DNA damage response. Nature, 2000. 404(6774): p. 201-4. 91. Tibbetts, R.S., et al., Functional interactions between BRCA1 and the checkpoint kinase ATR during genotoxic stress. Genes Dev, 2000. 14(23): p. 2989-3002. 92. Lim, D.S., et al., ATM phosphorylates p95/nbs1 in an S-phase checkpoint pathway. Nature, 2000. 404(6778): p. 613-7. 48 c 93. Harkin, D.P., et al., Induction of GADD45 and JNK/SAPK-dependent apoptosis following inducible expression ofBRCAl. Cell, 1999. 97(5): p. 575-86. 94. Zheng, L, et al., Lessons learned from BRCA1 and BRCA2. Oncogene, 2000. 19(53): p. 6159-75. 95. Bryant, H.E., et al., Specific killing of BRCA2-deficient tumours with inhibitors of poly(ADP-nbose) polymerase. Nature, 2005. 434(7035): p. 913-7. 96. Farmer, H., et al., Targeting the DNA repair defect in BRCA mutant cells as a therapeutic strategy. Nature, 2005. 434(7035): p. 917-21. 97. D'Amours, D., et al., Poly(ADP-ribosyl)ation reactions in the regulation of nuclear functions. Biochem J , 1999. 342 ( Pt 2): p. 249-68. 98. El-Khamisy, S.F., et al., A requirement forPARP-1 for the assembly or stability ofXRCCI nuclear foci at sites of oxidative DNA damage. Nucleic Acids Res, 2003. 31(19): p. 5526-33. 99. Lindahl, T., et al., Post-translational modification of poly(ADP-n'bose) polymerase induced by DNA strand breaks. Trends Biochem Sci, 1995. 20(10): p. 405-11. 100. Turner, N., A. Tutt, and A. Ashworth, Targeting the DNA repair defect of BRCA tumours. Curr Opin Pharmacol, 2005. 5(4): p. 388-93. 101. Pathology of familial breast cancer: differences between breast cancers in earners ofBRCAl orBRCA2 mutations and sporadic cases. Breast Cancer Linkage Consortium. Lancet, 1997. 349(9064): p. 1505-10. 102. Lakhani, S.R., et al., Multifactorial analysis of differences between sporadic breast cancers and cancers involving BRCA 1 and BRCA2 mutations. J Natl Cancer Inst, 1998. 90(15): p. 1138-45. 103. Lakhani, S. R., et al., The pathology of familial breast cancer: predictive value of immunohistochemical markers estrogen receptor, progesterone receptor, HER-2, and p53 in patients with mutations in BRCA 1 and BRCA2. J Clin Oncol, 2002. 20(9): p. 2310-8. 104. Armes, J.E., et al., Distinct molecular pathogeneses of eady-onset breast cancers in BRCA1 and BRCA2 mutation earners: a population-based study. Cancer Res, 1999. 59(8): p. 2011-7. 105. Chappuis, P.O., V. Nethercot, and W.D. Foulkes, Clinico-pathological characteristics ofBRCAl- and BRCA2-related breast cancer. Semin Surg Oncol, 2000. 18(4): p. 287-95. 49 106. Eerola, H., et al., Histopathological features of breast tumours in BRCA1, BRCA2 and mutation-negative breast cancer families. Breast Cancer Res, 2005. 7(1): p. R93-100. 107. Foulkes, W. D., et al., The prognostic implication of the basal-like (cyclin E high/p27 low/p53+/glomeruloid-microvascular-proliferation+) phenotype of BRCA1-related breast cancer. Cancer Res, 2004. 64(3): p. 830-5. 108. Honrado, E., J . Benitez, and J . Palacios, Histopathology ofBRCAl- and BRCA2-associated breast cancer. Crit Rev Oncol Hematol, 2006. 59(1): p. 27-39. 109. Loman, N., et al., Steroid receptors in hereditary breast carcinomas associated with BRCA 1 orBRCA2 mutations or unknown susceptibility genes. Cancer, 1998. 83(2): p. 310-9. 110. Palacios, J., et al., Phenotypic characterization of BRCA 1 and BRCA2 tumors based in a tissue microarray study with 37 immunohistochemical markers. Breast Cancer Res Treat, 2005. 90(1): p. 5-14. 111. Palacios, J., et al., Immunohistochemical characteristics defined by tissue microarray of hereditary breast cancer not attributable to BRCA1 or BRCA2 mutations: differences from breast carcinomas arising in BRCA 1 and BRCA2 mutation carriers. Clin Cancer Res, 2003. 9(10 Pt 1): p. 3606-14. 112. Hedenfalk, I., et al., Gene-expression profiles in hereditary breast cancer. N Engl J Med, 2001. 344(8): p. 539-48. 113. Johannsson, O.T., et al., Tumour biological features of BRCA1-induced breast and ovarian cancer. EurJ Cancer, 1997. 33(3): p. 362-71. 114. Foulkes, W.D., et al., Estrogen receptor status in BRCA1- and BRCA2-related breast cancer: the influence of age, grade, and histological type. Clin Cancer Res, 2004. 10(6): p. 2029-34. 115. Brodie, S.G. and C.X. Deng, BRCA1-associated tumorigenesis: what have we learned from knockout mice? Trends Genet, 2001. 17(10): p. S18-22. 116. Gasco, M., I.G. Yulug, and T. Crook, TP53 mutations in familial breast cancer: functional aspects. Hum Mutat, 2003. 21(3): p. 301-6. 117. Grushko, T.A., et al., Molecular-cytogenetic analysis of HER-2/neu gene in BRCA 1-associated breast cancers. Cancer Res, 2002. 62(5): p. 1481-8. 118. Palacios, J., et al., Re; Germline BRCA1 mutations and a basal epithelial phenotype in breast cancer. J Natl Cancer Inst, 2004. 96(9): p. 712-4; author reply 714. 50 119. Gillett, C.E., et al., Cyclin D1 and associated proteins in mammary ductal carcinoma in situ and atypical ductal hyperplasia. J Pathol, 1998. 184(4): p. 396-400. 120. Champeme, M.H., et al., 11q13 amplification in local recurrence of human primary breast cancer. Genes Chromosomes Cancer, 1995. 12(2): p. 128-33. 121. Courjal, F., et al., Mapping of DNA amplifications at 15 chromosomal localizations in 1875 breast tumors: definition of phenotypic groups. Cancer Res, 1997. 57(19): p. 4360-7. 122. Lammie, G.A., et al., D11S287, a putative oncogene on chromosome 11q13, is amplified and expressed in squamous cell and mammary carcinomas and linked to BCL-1. Oncogene, 1991. 6(3): p. 439-44. 123. Lammie, G.A. and G. Peters, Chromosome 11q13 abnormalities in human cancer. Cancer Cells, 1991. 3(11): p. 413-20. 124. Buckley, M.F., et al., Expression and amplification of cyclin genes in human breast cancer. Oncogene, 1993. 8(8): p. 2127-33. 125. Gillett, C , et al., Amplification and overexpression of cyclin D1 in breast cancer detected by immunohistochemical staining. Cancer Res, 1994. 54(7): p. 1812-7. 126. Karlseder, J., et al., Patterns of DNA amplification at band q13 of chromosome 11 in human breast cancer. Genes Chromosomes Cancer, 1994. 9(1): p. 42-8. 127. Peters, G., et al., Chromosome 11q13 markers and D-type cyclins in breast cancer. Breast Cancer Res Treat, 1995. 33(2): p. 125-35. 128. Gillett, C , et al., Cyclin D1 and prognosis in human breast cancer. Int J Cancer, 1996. 69(2): p. 92-9. 129. Hwang, T.S., et al., Prognostic value of combined analysis of cyclin D1 and estrogen receptor status in breast cancer patients. Pathol Int, 2003. 53(2): p. 74-80. 130. Bieche, I., et al., Prognostic value ofCCNDI gene status in sporadic breast tumours, as determined by real-time quantitative PCR assays. Br J Cancer, 2002. 86(4): p. 580-6. 131. Michalides, R., et al., A clinicopathological study on overexpression of cyclin D1 and ofp53 in a series of 248 patients with operable breast cancer. Br J Cancer, 1996. 73(6): p. 728-34. 51 132. Seshadri, R., et al., Cyclin Dl amplification is not associated with reduced overall survival in primary breast cancer but may predict early relapse in patients with features of good prognosis. Clin Cancer Res, 1996. 2(7): p. 1177-84. 133. Vaziri, S.A., et al., Absence of CCND1 gene amplification in breast tumours ofBRCAl mutation carriers. Mol Pathol, 2001. 54(4): p. 259-63. 134. Livasy, C.A., et al., Phenotypic evaluation of the basal-like subtype of invasive breast carcinoma. Mod Pathol, 2006.19(2): p. 264-71. 135. Nielsen, T.O., et al., Immunohistochemical and clinical characterization of the basal-like subtype of invasive breast carcinoma. Clin Cancer Res, 2004. 10(16): p. 5367-74. 136. van de Rijn, M., et al., Expression of cytokeratins 17 and 5 identifies a group of breast carcinomas with poor clinical outcome. Am J Pathol, 2002. 161(6): p. 1991-6. 137. Gruvberger, S., et al., Estrogen receptor status in breast cancer is associated with remarkably distinct gene expression patterns. Cancer Res, 2001. 61(16): p. 5979-84. 138. Jones, C., et al., CGH analysis of ductal carcinoma of the breast with basaloid/myoepithelial cell differentiation. Br J Cancer, 2001. 85(3): p. 422-7. 139. Tot, T., The cytokeratin profile of medullary carcinoma of the breast. Histopathology, 2000. 37(2): p. 175-81. 140. Foulkes, W.D., et al., Germline BRCA1 mutations and a basal epithelial phenotype in breast cancer. J Natl Cancer Inst, 2003. 95(19): p. 1482-5. 141. Lakhani, S.R., etal., Pathology of ovarian cancers in BRCA1 and BRCA2 carriers. Clin Cancer Res, 2004. 10(7): p. 2473-81. 142. Marcus, J . N., et al., Hereditary breast cancer: pathobiology, prognosis, and BRCA1 and BRCA2 gene linkage. Cancer, 1996. 77(4): p. 697-709. 143. Armes, J.E., et al., The histologic phenotypes of breast carcinoma occurring before age 40 years in women with and without BRCA 1 or BRCA2 germline mutations: a population-based study. Cancer, 1998. 83(11): p. 2335-45.0 144. Osin, P., et al., Predicted anti-oestrogen resistance in BRCA-associated familial breast cancers. Eur J Cancer, 1998. 34(11): p. 1683-6. 52 145. Boyd, J., et al., Clinicopathologic features of BRCA-linked and sporadic ovarian cancer. Jama, 2000. 283(17): p. 2260-5. 146. Johannsson, O.T., et al., Survival ofBRCAl breast and ovarian cancer patients: a population-based study from southern Sweden. J Clin Oncol, 1998. 16(2): p. 397-404. 147. Risch, H.A., et al., Prevalence and penetrance of germline BRCA1 and BRCA2 mutations in a population series of 649 women with ovarian cancer. Am J Hum Genet, 2001. 68(3): p. 700-10. 148. Rubin, S.C., et al., Clinical and pathological features of ovarian cancer in women with germ-line mutations ofBRCAl. N Engl J Med, 1996. 335(19): p. 1413-6. 149. Sekine, M., et al., Mutational analysis of BRCA 1 and BRCA2 and clinicopathologic analysis of ovarian cancer in 82 ovarian cancer families: two common founder mutations ofBRCAl in Japanese population. Clin Cancer Res, 2001. 7(10): p. 3144-50. 150. Shaw, P.A., et al., Histopathologic features of genetically determined ovarian cancer. Int J Gynecol Pathol, 2002. 21(4): p. 407-11. 151. Werness, B.A., et al., Histopathology, FIGO stage, and BRCA mutation status of ovarian cancers from the Gilda Radner Familial Ovarian Cancer Registry. Int J Gynecol Pathol, 2004. 23(1): p. 29-34. 152. Zweemer, R.P., et al., Clinical and genetic evaluation of thirty ovarian cancer families. Am J Obstet Gynecol, 1998.178(1 Pt 1): p. 85-90. 153. Ravid, A., et al., Immunohistochemical analyses of sporadic and familial (185delAG earners) ovarian cancer in Israel. Eur J Cancer, 2000. 36(9): p. 1120-4. 154. Schuijer, M. and E.M. Berns, TP53 and ovarian cancer. Hum Mutat, 2003. 21(3): p. 285-91. 155. Ben David, Y., et al., Effect of BRCA mutations on the length of survival in epithelial ovarian tumors. J Clin Oncol, 2002. 20(2): p. 463-6. 156. Cass, I., et al., Improved survival in women with BRCA-associated ovarian carcinoma. Cancer, 2003. 97(9): p. 2187-95. 157. Jazaeri, A.A., et al., Gene expression profiles of BRCA1-linked, BRCA2-linked, and sporadic ovarian cancers. J Natl Cancer Inst, 2002. 94(13): p. 990-1000. 53 158. Futreal, P.A., et al., BRCA1 mutations in primary breast and ovarian carcinomas. Science, 1994. 266(5182): p. 120-2. 159. Sorlie, T., et al., Mutation screening ofBRCAl using PTT and LOH analysis at 17q21 in breast carcinomas from familial and non-familial cases. Breast Cancer Res Treat, 1998. 48(3): p. 259-64. 160. Berchuck, A., et al., Frequency of germline and somatic BRCA1 mutations in ovarian cancer. Clin Cancer Res, 1998. 4(10): p. 2433-7. 161. Geisler, J.P., et al., Frequency ofBRCAl dysfunction in ovarian cancer. J Natl Cancer Inst, 2002. 94(1): p. 61-7. 162. Hilton, J.L., et al., Inactivation ofBRCAl and BRCA2 in ovarian cancer. J Natl Cancer Inst, 2002. 94(18): p. 1396-406. 163. Khoo, U.S., et al., Somatic mutations in the BRCA1 gene in Chinese sporadic breast and ovarian cancer. Oncogene, 1999.18(32): p. 4643-6. 164. Lancaster, J.M., et al., BRCA2 mutations in primary breast and ovarian cancers. Nat Genet, 1996.13(2): p. 238-40. 165. Foster, K. A., et al., Somatic and germline mutations of the BRCA2 gene in sporadic ovarian cancer. Cancer Res, 1996. 56(16): p. 3622-5. 166. Gras, E., et al., Loss of heterozygosity on chromosome 13q12-q14, BRCA-2 mutations and lack of BRCA-2 promoter hypermethylation in sporadic epithelial ovarian tumors. Cancer, 2001. 92(4): p. 787-95. 167. Russell, P.A., et al., Frequent loss of BRCA 1 mRNA and protein expression in sporadic ovarian cancers. Int J Cancer, 2000. 87(3): p. 317-21. 168. Turner, N., A. Tutt, and A. Ashworth, Hallmarks of 'BRCAness' in sporadic cancers. Nat Rev Cancer, 2004. 4(10): p. 814-9. 169. Catteau, A., et al., Methylation of the BRCA 1 promoter region in sporadic breast and ovarian cancer: correlation with disease characteristics. Oncogene, 1999.18(11): p. 1957-65. 170. Esteller, M., et al., Promoter hypermethylation and BRCA1 inactivation in sporadic breast and ovarian tumors. J Natl Cancer Inst, 2000. 92(7): p. 564-9. 171. Rice, J.C., et al., Methylation of the BRCA1 promoter is associated with decreased BRCA 1 mRNA levels in clinical breast cancer specimens. Carcinogenesis, 2000. 21(9): p. 1761-5. 54 172. Baldwin, R. L, et al., BRCA 1 promoter region hypermethylation in ovarian carcinoma: a population-based study. Cancer Res, 2000. 60(19): p. 5329-33. 173. Collins, N., R. Wooster, and M.R. Stratton, Absence of methylation of CpG dinucleotides within the promoter of the breast cancer susceptibility gene BRCA2 in normal tissues and in breast and ovarian cancers. Br J Cancer, 1997. 76(9): p. 1150-6. 174. Hughes-Davies, L, et al., EMSY links the BRCA2 pathway to sporadic breast and ovarian cancer. Cell, 2003. 115(5): p. 523-35. 175. Xia, B., et al., Control of BRCA2 cellular and clinical functions by a nuclear partner, PALB2. Mol Cell, 2006. 22(6): p. 719-29. 176. Pilch, D.R., et al., Characteristics ofgamma-H2AX foci at DNA double-strand breaks sites. Biochem Cell Biol, 2003. 81(3): p. 123-9. 177. Maurer-Stroh, S., et al., The Tudor domain 'Royal Family': Tudor, plant Agenet, Chromo, PWWPand MBTdomains. Trends Biochem Sci, 2003. 28(2): p. 69-74. 178. Bekri, S., et al., Detailed map of a region commonly amplified at 11q13->q14 in human breast carcinoma. Cytogenet Cell Genet, 1997. 79(1-2): p. 125-31. 179. Ormandy, C.J., et al., Cyclin D1, EMS1 and 11q13 amplification in breast cancer. Breast Cancer Res Treat, 2003. 78(3): p. 323-35. 180. Schuuring, E., The involvement of the chromosome 11q13 region in human malignancies: cyclin D1 and EMS1 are two new candidate oncogenes-a review. Gene, 1995. 159(1): p. 83-96. 181. Schwab, M., Amplification of oncogenes in human cancer cells. Bioessays, 1998. 20(6): p. 473-9. 182. Chavali, G.B., et al., Crystal structure of the ENT domain of human EMSY. J Mol Biol, 2005. 350(5): p. 964-73. 183. Huang, Y., M.P. Myers, and R.M. Xu, Crystal structure of the HP1-EMSY complex reveals an unusual mode ofHP1 binding. Structure, 2006.14(4): p. 703-12. 184. Clarke, N.D., Covariation of residues in the homeodomain sequence family. Protein Sci, 1995. 4(11): p. 2269-78. 185. Ekblad, C M . , et al., Binding of EMSY to HPIbeta: implications for recruitment of HPIbeta and BS69. EMBO Rep, 2005. 6(7): p. 675-80. 55 186. King, M.C., A novel BRCA2-binding protein and breast and ovarian tumorigenesis. N Engl J Med, 2004. 350(12): p. 1252-3. 187. Livingston, D.M., EMSY, a BRCA-2 partner in crime. Nat Med , 2004. 10(2): p. 127-8. 56 C h a p t e r 2: FLUORESCENT IN SITU HYBRIDIZATION ON TISSUE MICROARRAYS: CHALLENGES AND SOLUTIONS Objective: To develop and validate the usefulness of fluorescent in situ (FISH) on TMAs. This review article describes the FISH technique in detail and the problems and solutions encountered during the development of this assay. Based on the Manuscript: Brown Lindsay A and Huntsman David. Fluorescent in situ hybridization on tissue microarrays: challenges and solutions. Journal of Molecular Histology. 2007 [Epub ahead of print]. 57 2.1 RATIONALE A large component of my research was dependent on the development of a FISH assay for use on paraffin embedded tissue microarrays. The first year of my graduate studies was spent developing and optimizing such an assay. This assay has now been implemented for the detection of novel gene amplification events and translocations. This work led to the publication of a review article describing the technique, and the problems and solutions encountered with the development of this FISH assay. 2.2 INTRODUCTION The recent advances in technologies such as cDNA microarrays and proteomics have led to the discovery of a multitude of potential candidate genes and proteins that could play a role in carcinogenesis, clinical outcome or therapeutic response. As the rate of discovery of these potentially important genes and proteins increases there will be a growing demand to analyze these potential markers in large series of tumours to determine their clinical significance. 2.3 TISSUE MICROARRAYS (TMAS) The development of TMAs has provided a high-throughput platform for the evaluation of potential candidate genes and proteins [1-5]. TMA slides can be used to determine the diagnostic, prognostic and predictive value of potential biomarkers in large cohorts of samples, as well as the expression profiles of therapeutic targets. For TMA construction, representative areas of invasive 58 carcinoma are selected and marked on hematoxylin and eosin (H&E) stained slides and its corresponding tissue block. The TMAs are assembled using a tissue-arraying instrument (Beecher Instruments, Silver Springs, MD) as described previously [6]. Briefly, the instrument is used to create holes in a recipient paraffin block with defined array coordinates. A stylet is used to transfer a 0.6 mm diameter tissue core into the recipient block. TMAs are ideally suited for any in situ analysis including immunohistochemistry (IHC), fluorescent and bright field in situ hybridization, and RNA in situ hybridization allowing for concurrent assessment of protein, DNA or RNA targets (Figure 2.1A). 2.4 FISH AND TMAs FISH is a molecular technique that detects chromosomal copy number changes, amplifications, deletions and rearrangements [7]. FISH has long been used as a diagnostic tool in the classification of haematological neoplasms and in prenatal detection of genetic aberrations [8, 9]. In solid epithelial malignancies, FISH is commonly used for the detection of HER2 gene amplification as a means of selecting patients who will respond to Herceptin™ (trastuzumab) therapy [10]. FISH is also used to detect the presence of specific translocations in paraffin embedded materials and is currently used for the diagnosis of sarcomas and haematological malignancies. Recently, a novel recurrent gene fusion between the 5' untranslated region of TMPRSS2 to three members of the ETS family of transcription factors (ERG, ETV1, and ETV4) was observed in a high percentage 59 B) C) Figure 2.1 Applications of FISH on TMAs A) TMAs facilitate the rapid screening of novel biomarkers in hundreds of tumours on a single slide. Sections cut from the same TMA are suited for in situ analysis, including IHC and FISH, allowing for the parallel assessment of protein and DNA targets. B) FISH can be used for the detection of gene amplification events. EMSY (green) is present in multiple copies, whereas both CCND1 (orange) and centromere 11 (aqua) are present in normal copy number. C) Detection of fused red and green signals with fusion probes indicating the presence of t(12;15)(p13;q26). 60 of prostate cancer tissues [11-13]. These findings have generated an increased interest in the search for new translocations in epithelial malignancies. 2.4.1 Probe preparation and labelling For some well-documented oncogenes and translocations commercial probes sets may already exist. However, for novel alterations locus specific FISH probes can be generated using cosmids, P1 artificial chromosomes (PACs), bacterial artificial chromosomes (BACs) or yeast artificial chromosomes (YACs). Clones are chosen using sites such as Ensembl (vyww.ensembi.com) or UCSC genome browser (www.genome.ucsc.edu). Specific genes or chromosomal regions can be queried using these websites and detailed information regarding the genomic sequence, including BACs that have been mapped to this region are displayed. Clones covering the specific region of interest are chosen and grown. The DNA is extracted using an alkaline lysis method and labelled using standard protocols. There are two methods of probe labelling: direct and indirect. In the direct method a fluorochrome is coupled directly to a modified nucleotide that is then incorporated into the DNA. After incorporation into the probe and following hybridization the bound probe is immediately visible under the microscope. For indirect labelling, haptens, such as digoxigenin or biotin are incorporated into the probe. Since the probe is labelled with a hapten that is not fluorescent, incubation with an antibody coupled to a fluorescent dye is required for detection. Introduction of a label can be done using several techniques such as random priming, nick translation or PCR. In random priming a mixture of random hexamers is used to prime DNA synthesis along multiple sites of denatured DNA. The complementary strand of DNA is synthesized by a Klenow fragment of DNA polymerase I, using the oligonucleotides as primers. The newly synthesized strand is labelled by substituting a labelled nucleotide for the equivalent non-labelled nucleotide in the reaction mixture [14, 15]. The nick translation labelling reaction requires the action of DNAse and DNA polymerase. The double stranded DNA is "nicked" by DNAse and during resynthesis by DNA polymerase the excised nucleotides are replaced with labelled nucleotides [16]. Nick translation is often used for the labelling of cloned DNA from BACs and PACs. PCR can also be used to label probes and has the advantage that the entire probe is being amplified. When detecting gene amplifications the probe mix usually contains a reference probe, such as a centromere, which is used as a marker of aneusomy. Most centromere probes are commercially available. However there are some centromeres for which no commercial probes exist, such as chromosome 19. In these instances BACs in the region closest to the centromere should be chosen, or a telomere probe can also be used. 2.5 APPLICATIONS OF FISH ON TMAs 2.5.1 Detection of gene amplifications Recurrent amplification of some chromosomal regions is commonly associated with the overexpression of well-documented oncogenes. These include oncogenes such as HER2 at 17q21, MYC at 8q24 and CCND1 at 11q13 [17]. 62 The study of these oncogenes has had an important impact on cancer diagnosis and treatment. Performing FISH on individual slides for hundreds of cases can be expensive and time consuming. TMA technology has overcome these limitations allowing the rapid screening of genetic alterations. Through the parallel analysis of hundreds of tumours on a single TMA slide the prognostic significance of novel amplicons can be readily determined. Likewise, by applying probe sets to multi-tumour TMAs, which contain a broad spectrum of neoplastic lesions, the tumour spectrum associated with a specific genetic alteration can be determined. For example, we have used a clinically annotated breast cancer TMA to assess the prognostic significance of a novel oncogene, EMSY [2]. EMSY'\s located at 11q13.5, 6 Mb telomeric to CCND1, and encodes a protein that binds and represses the activity of the transactivation domain of BRCA2. EMSY gene amplification was found in 13% of sporadic breast cancers (Figure 2.1B). Patients with EMSY amplification were associated with a poor prognosis and this effect on survival was more marked in node negative patients than in node positive cases [2]. EMSY amplification was also examined in a multi-tumour array. Amplification of EMSY was seen in a small percentage of hepatocellular carcinomas, melanomas, and endometrial carcinomas, but was not seen in more than 200 other cancers from a wide range of primary sites, including sarcomas and colorectal carcinomas [2]. 63 FISH assays can be used to detect deletions. However, this technique is more difficult as tissue sectioning can result in the truncation of nuclei and could lead to loss of signal and false positives. Therefore, it is important to optimize FISH assays on control slides before scoring actual test samples to ensure the quality and integrity of the assay. 2.5.2 Detection of translocations Another application of FISH on TMAs is the assessment of translocations in large cohorts of patients. For the detection of translocations either a fusion or break-apart assay can be used. In a fusion assay differentially labelled probes hybridize to targets located on either side of the genetic breakpoints. A break-apart FISH assay is useful in cases where there might be multiple translocation partners for a known genetic breakpoint. For a break-apart assay, two different coloured probes that hybridize to targets flanking a known breakpoint in one gene are used. For example, synovial sarcomas carry a balanced translocation t(X;18)(p11.2;q11.2) that is detectable in over 90% of cases [18, 19]. This translocation brings together two genes: SYT and SSX. A commercial break-apart FISH assay (Abbott MolecularA/ysis) was used to detect disruption of SYT on a sarcoma TMA. Disease specific TMAs are used to determine assay sensitivity whilst multi-tumor arrays are used to determine specificity. Disruption of SYT was found in 22/23 (96%) known synovial sarcomas, but was not seen in the non-synovial sarcomas included on the array [20], 64 A further example is the study of the translocation t(12;15)(p13;q25), involving a rearrangement that fuses the ETV6 gene on chromosome 12 with the NTRK3 gene from chromosome 15, in secretory breast cancer [4]. A fusion and a break-apart FISH assay, which were initially tested on four confirmed cases of secretory breast cancer, were used to screen a breast cancer TMA of various subtypes (Figure 2.1C). This FISH assay resulted in the detection of one case with fusion signals that upon review was confirmed to be secretory breast cancer [4]. These studies demonstrate that archival tumour samples can be used for the detection of translocations and the potential use of these assays as a diagnostic tool. 2.6 DEVELOPMENT OF A ROBUST FISH ASSAY FOR USE ON PARAFFIN EMBEDDED TMAS The use of FISH on archival formalin-fixed paraffin embedded tissue can be technically challenging. FISH on paraffin embedded tissues relies on the ability of the probe to penetrate and bind to its target DNA. There are several inherent problems with performing FISH on paraffin embedded TMAs and these problems are listed in Table 2.1. One major issue is the use of fixatives such as formalin, which is used to preserve the structure of the tissue by forming cross-linking methylene bridges, however these cross-linkages greatly reduce the accessibility of the hybridization target. The use of pre-treatment steps such as sodium thiocyanate (NaSCN) and pepsin digestion have greatly improved the unmasking 65 of target DNA, thereby allowing the probe to penetrate and hybridize to the target [21]. However, on a single TMA slide there can be upwards of hundreds of archival tissues that vary in age as well as the type and length of fixation used. As well, the pre-treatment steps can be harmful to the tissue structure. For example, extended incubation of slides with pepsin may lead to overdigestion, in comparison with insufficient protease incubation, which may lead to hybridization failure. Thus, the key to a successful FISH assay is the careful optimization of each pre-treatment step to ensure that the correct concentration, length of incubation, and temperature for each pre-treatment step is chosen. This optimization may need to be repeated for every new tissue type encountered on a TMA. Due to the different pre-treatment requirements of each tissue type, multi-tumour arrays pose a problem for FISH, as the same pre-treatment conditions, which may overdigest a breast carcinoma, may not penetrate a sarcoma for successful hybridization on the same array. It is also important to assess the number of cores present on the unstained slide before staining and analysis, as cores are often lost during the cutting process. For FISH, core loss can present problems for orientation on the slide and result in complete loss of data integrity. Several articles have been published detailing novel protocols for performing FISH on paraffin embedded TMAs [22, 23]. The current protocol used in the authors' laboratory is outlined in Table 2.1. This FISH procedure has been successfully used to assess the amplification of several oncogenes and translocations in a number of different TMAs [1, 2, 4, 20, 24, 25]. After 66 Table 2.1 FISH and paraffin embedded TMAs: problems and solutions Problems encountered with paraffin embedded tissue and TMAs • Formalin fixation greatly impairs the accessibility of the hybridization target and as a result fluorescent signals may be very weak - There can be hundreds of different archival tissues on a TMA and as a result the age of the tissue as well as type and length of fixation may differ • Loss of tissue cores • Pre-treatment washes can be harsh on the tissue, for example pepsin treatment can lead to overdigestion of the tissue and uninterruptible results • It is critical for each pre-treatment step to be optimized in regards to concentration, length of incubation, and temperature for each tissue type - Fluorescent signals have to be scored while constantly changing the plane of focus to ensure that signals in out of focus planes are not missed • Manual scoring of fluorescent signals on TMAs can be very tedious especially if multiple signals labelled in different fluorophores are used Therefore, a robust and reliable FISH assay that results in high quality FISH signals is needed when working with paraffin embedded tissue FISH protocol for use on paraffin embedded TMAs Preparation of Probe Deparaffinizing the slides 1. Xylene for 5 mins at RT, 3x 2. 100% EtOH, 1 min at RT, 2x Pretreatment 1. 0.2N HCI for 20 minutes at RT 2. dH 2 0 for 10 minutes at RT 3. 2X S S C for 3 minutes at RT 4. 1M NaSCN at 80 °C for 30 minutes* 5. dH 20 for 3 minutes at RT Protease treatment 1. 37, 500 units of pepsin in 0.2N HCI for 10 mins at 37 °C* 2. dH 20 for 3 minutes at RT 3. 100% EtOH, 1 min at RT, 2x 4. Air dry the slides Co-denaturation and hybridization 1. Add probe to sample area of slide 2. Coverslip and seal edges with rubber cement 3. Denature for 5 minutes at 73°C and hybridize for 14-18 hrs at 37°C on the Hybrite Post Hybridization wash 1. Remove rubber cement and coverslip 2. 2X SSC/0.3% NP40 at 72°C for 2 minutes 3. Ethanol dehydration 4. Air dry Counterstain 1. Counterstain with DAPI and coverslip 2. Store in dark prior to enumeration * These pre-treatment steps need to be optimized 6 7 hybridization and post-hybridization washes the slides should be left for at least 24 hours in the fridge to allow background fluorescence to fade. Slides can be stored for months in this state. 2.7 SCORING OF F I S H SIGNALS 2.7.1 Manual Scoring Manual scoring of FISH signals on a TMA can be a daunting task, especially if analyzing multiple fluorescent probes. One also has to take into consideration the thickness of the nuclei when manually scoring, therefore fluorescent signals in multiple focal planes must be analyzed to ensure out of focus signals are not overlooked. A major concern when performing FISH on TMAs is the amount of time needed to score an array, especially when multiple probes must be scored. It is also very important to continually verify that the correct core is being analyzed as the TMA reviewer may easily become disoriented while scoring. A useful approach is to include non-neoplastic control tissue, such as kidney, as a topographic reference point that has a designated co-ordinate on the array and can be rapidly identified. For the analysis of gene amplification, FISH signals are enumerated in approximately 40 to 100 morphologically intact and non-overlapping nuclei. The average copy number for each probe is calculated and the amplification ratio is determined. The amplification ratio is calculated as a ratio between the average copy number per cell for the gene of interest and the average copy number for the reference probe. 68 For manual scoring of translocations, typically 100 to 200 nuclei are scored for either the presence of fused signals in the case of a fusion assay or for split signals in a break-apart assay. The most ideal method for scoring translocations is to use a dual bandpass filter; however, this filter does suppress fluorescent signals, especially green signals. Another option is to alternate between both filters and look for either break apart or fused signals. The reviewer can also capture images using imaging software and multiple filters to verify the presence of translocations. The optimal cut-off value indicating the presence of a translocation varies depending on the probe set being used and the tissue type. 2.7.2 Automated scoring of FISH signals FISH on paraffin embedded tissue, either whole section or TMA, is becoming more widely used in pathology laboratories both in research and clinical settings. The manual scoring of such slides can often be laborious and difficult, and as a result much effort has been made to develop automated scoring systems for the rapid detection of gene amplifications in interphase nuclei. Several systems have been developed to address this need, one of which is Metacyte from Metasystems (Altlussheim, Germany). Automated scoring of FISH signals requires the ability to analyze the full thickness of the nuclei and therefore multiple focal planes must be captured. For automated scoring several fields of interest for each core are chosen using the 10x objective (Figure 2.2A). A spot counting algorithm is chosen, which contains all the appropriate filter combinations. The system will then progress 69 through the chosen fields and capture the fluorescent signals using the 40x objective (Figure 2 . 2 B ) . The FISH signals are captured in nine focal planes that are separated by 0.75 microns, to ensure that signals within out of focus planes are not missed. An extended focus image combines the images from all focus planes. The analyzed nuclei are displayed in an image gallery and in each tile a spot count for each probe is displayed (Figure 2 . 2 C ) . An amplification ratio is displayed in a table at the bottom of the screen. We tested the automated scoring of fluorescent signals using an ovarian TMA with 59 tumours and a breast TMA with 74 samples. A three-colour FISH assay that included a directly labelled EMSY probe (green) combined with commercially available CCND1 (orange) and centromere 11 (aqua) probes were used. The results were then compared to manual FISH scoring to determine the correlation between the two methods. Two different spot counting algorithms developed by Metasystems (Altlussheim, Germany) for automated scoring were tested to determine which gave results closest to the manual FISH scores. The first spot counting algorithm applied a user-defined region of interest that allowed the user to highlight the cells to be included in the analysis. The second algorithm assessed all cells within the field of view, and applied a stricter image and spot counting quality based rejection criteria before accepting a field for analysis. For both spot counting algorithms analyzed on either array we found that 83-95% of cases were correctly classified as amplified or non-amplified. For all highly amplified cases there were no discordance between manual scoring and the automated system. We observed that the spot counting algorithm that applied a user-70 A) B) Figure 2.2 Automated scoring of FISH signals. A) Using the 10x objective several representative fields from the tissue core are chosen. B) Manual scoring of FISH signals is done using the 40x objective to ensure that as many nuclei as possible are included in the analysis. The FISH signals are captured in nine focal planes and an extended focus image combines all focus planes. C) After spot counting, image galleries displaying the captured fields, as well as individual spot counts are shown on the screen. The number in green represents the EMSY spot count; the number in red and aqua indicates the spot count for CCND1 and centromere 11. The number in white is a ratio between the spot count for EMSY and the spot count for centromere 11. 71 defined region of interest showed the best correlation with the manual scoring of EMSY and CCND1. We also preferred this spot counting algorithm as it allowed the user to exclude any normal cells within the field of views from the analysis, whereas the second spot counting algorithm assessed all cells within the field of view. Based on these results we have now implemented the use of this system in all automated scoring of gene amplifications events. One limitation of FISH is that the fluorescent signals do eventually weaken. Re-evaluation of a particular case several months later may not be feasible. Thus, a major benefit of an automated system is the availability of a digitally stored data set containing the spot counts and an image gallery of each cell that was analyzed. This system also has the ability to detect three-dimensional distances between fluorescent signals, which can be applied to the detection of gene fusions or split signal assays. 2.8 CONCLUSIONS TMAs are a powerful tool for the rapid identification of tumour types associated with specific molecular alterations. Performing FISH on archival paraffin embedded TMAs is technically possible and can provide key information regarding the diagnostic, prognostic and predictive value of novel biomarkers. 72 2.9 REFERENCES 1. Brown, L.A., et al., Amplification of EMSY, a novel oncogene on 11q13, in high grade ovarian surface epithelial carcinomas. Gynecol Oncol, 2006. 100(2): p. 264-70. 2. Hughes-Davies, L, et al., EMSY links the BRCA2 pathway to sporadic breast and ovarian cancer. Cell, 2003. 115(5): p. 523-35. 3. Kononen, J., et al., Tissue microarrays for high-throughput molecular profiling of tumor specimens. Nat Med, 1998. 4(7): p. 844-7. 4. Makretsov, N., et al., A fluorescence in situ hybridization study of ETV6-NTRK3 fusion gene in secretory breast carcinoma. Genes Chromosomes Cancer, 2004. 40(2): p. 152-7. 5. Schraml, P., et al., Tissue microarrays for gene amplification surveys in many different tumor types. Clin Cancer Res, 1999. 5(8): p. 1966-75. 6. Livingston, D.M., EMSY, a BRCA-2 partner in crime. Nat Med, 2004. 10(2): p. 127-8. 7. Fletcher, J.A., DNA in situ hybridization as an adjunct in tumor diagnosis. Am J Clin Pathol, 1999. 112(1 Suppl 1): p. S11-8. 8. Kearney, L., The impact of the new fish technologies on the cytogenetics of haematological malignancies. BrJ Haematol, 1999.104(4): p. 648-58. 9. Pergament, E., et al., The clinical application of interphase FISH in prenatal diagnosis. Prenat Diagn, 2000. 20(3): p. 215-20. 10. Spiridon, C.I., et al., Targeting multiple Her-2 epitopes with monoclonal antibodies results in improved antigrowth activity of a human breast cancer cell line in vitro and in vivo. Clin Cancer Res, 2002. 8(6): p. 1720-30. 11. Tomlins, S.A., et al., Recurrent fusion of TMPRSS2 and ETS transcription factor genes in prostate cancer. Science, 2005. 310(5748): p. 644-8. 12. Yoshimoto, M., et al., Three-color FISH analysis of TMPRSS2/ERG fusions in prostate cancer indicates that genomic microdeletion of chromosome 21 is associated with rearrangement. Neoplasia, 2006. 8(6): p. 465-9. 13. Tomlins, S.A., et al., TMPRSS2:ETV4 gene fusions define a third molecular subtype of prostate cancer. Cancer Res, 2006. 66(7): p. 3396-400. 73 14. Feinberg, A.P. and B. Vogelstein, A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Anal Biochem, 1983. 132(1): p. 6-13. 15. Feinberg, A.P. and B. Vogelstein, "A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity". Addendum. Anal Biochem, 1984. 137(1): p. 266-7. 16. Rigby, P.W, et al., Labeling deoxyribonucleic acid to high specific activity in vitro by nick translation with DNA polymerase I. J Mol Biol, 1977. 113(1): p. 237-51. 17. Knuutila, S., et al., DNA copy number amplifications in human neoplasms: review of comparative genomic hybridization studies. Am J Pathol, 1998. 152(5): p. 1107-23. 18. dos Santos, N.R., D.R. de Bruijn, and A.G. van Kessel, Molecular mechanisms underlying human synovial sarcoma development. Genes Chromosomes Cancer, 2001. 30(1): p. 1-14. 19. Fisher, C , Synovial sarcoma. Ann Diagn Pathol, 1998. 2(6): p. 401-21. 20. Terry, J., et al., Fluorescence in situ hybridization for the detection of t(X;18)(p11.2;q11.2) in a synovial sarcoma tissue microarray using a breakapad-style probe. Diagn Mol Pathol, 2005. 14(2): p. 77-82. 21. Ensinger, C , et al., Improved technique for investigations on archival formalin-fixed, paraffin-embedded tumors by interphase in-situ hybridisation. Anticancer Res, 1997. 17(6D): p. 4633-7. 22. Andersen, C L , et al., Improved procedure for fluorescence in situ hybridization on tissue microarrays. Cytometry, 2001. 45(2): p. 83-6. 23. Chin, S.F., et al., A simple and reliable pretreatment protocol facilitates fluorescent in situ hybridisation on tissue microarrays of paraffin wax embedded tumour samples. Mol Pathol, 2003. 56(5): p. 275-9. 24. Lee, C.H., et al., Assessment ofHer-1, Her-2, And Her-3 expression and Her-2 amplification in advanced stage ovarian carcinoma. Int J Gynecol Pathol, 2005. 24(2): p. 147-52. 25. Prentice, L.M., et al., NRG1 gene rearrangements in clinical breast cancer: identification of an adjacent novel amplicon associated with poor prognosis. Oncogene, 2005. 74 C h a p t e r 3: E M S V G E N E AMPLIFICATION IN SPORADIC BREAST CANCERS AND B R C A 1 AND B R C A 2 MUTATION CARRIERS Objective: To determine the frequency and clinical significance of EMSY gene amplification in sporadic and hereditary breast cancer Based on the Manuscripts: 1. Hughes-Davies L, Huntsman D, Ruas M, Fuks F, Bye J , Chin SF, Milner J , Brown LA, Hsu F, Gilks B, Nielsen T, Schulzer M, Chia S, Ragaz J , Cahn A, Linger L, Ozdag H, Cattaneo E, Jordanova ES, Schuuring E, Yu DS, Venkitaraman A, Ponder B, Doherty A, Aparicio S, Bentley D, Theillet C, Ponting CP, Caldas C, Kouzarides T. EMSY links the BRCA2 pathway to sporadic breast and ovarian cancer. Cell. 2003 Nov 26;115(5):523-35. 2. Brown LA. Harmer K, Leung S, Cheang M, Gilks B, Huntsman D. Amplification of EMSY and CCND1 in a large population based series of breast cancers. Manuscript in preparation. The candidate performed or supervised all experiments presented. A statistician at GPEC performed all statistical analysis. 75 3.1 INTRODUCTION Inherited mutations in the breast cancer tumor suppressor genes, BRCA1 and BRCA2, markedly increase the risk of developing both breast and ovarian cancer. Germline carriers of hereditary mutations in BRCA1 and BRCA2 have a 60 to 80 percent lifetime risk of developing breast cancer [1, 2]. However, familial breast cancer is rare, accounting for less than 5% of cases. Thus, numerous studies have looked at the involvement of BRCA1 and BRCA2 in sporadic breast cancer; however, despite exhaustive searches, somatic mutations of BRCA1 and BRCA2 appear to be rare events in sporadic breast cancer [3-6]. Allelic loss of heterozygosity for either of these genes is commonly found in breast cancer, suggesting that there could be other genetic mechanisms responsible for somatic inactivation of BRCA1 and BRCA2 [7, 8]. Inactivation of the BRCA1 locus due to promoter hypermethylation has been found in 11-15% of breast cancer [9, 10], although no such epigenetic inactivation has been found at the BRCA2 locus in sporadic breast cancer [7]. It is possible that acquired abnormalities in genes interfering with BRCA1 or BRCA2 synthesis or function could define a subset of cases that arise through mechanisms equivalent to those seen in hereditary cases. EMSY\s a recently cloned gene encoding a protein that binds the transactivation domain of BRCA2 and localizes to sites of repair following DNA damage [11]. EMSY maps to 11q13.5, a region commonly amplified in breast cancer [12, 13]. CCND1, located 6 Mb centromeric to EMSY, is one of the genes implicated in 76 these amplification events. Amplification of this region at 11q13 in breast cancer is associated with a poor prognosis[13]. Thus, EMSY could be a candidate oncogene from the 11q13 amplicon and play a role in the development of breast cancer. To determine EMSYs involvement in sporadic breast cancer, we studied EMSY gene amplification, along with CCND1, in 928 clinically annotated breast cancers using fluorescent in situ hybridization (FISH). We also determined the clinical and prognostic significance of EMSY and CCND1 in this series of breast cancers. We then verified the findings from the first breast cancer series (928 case series) in a larger cohort that consisted of 4150 female patients with invasive breast cancer. Lastly, we evaluated EMSY gene amplification, along with CCND1, in breast cancers associated with known mutations in BRCA1 and BRCA2, as well as a subset of familial breast cancers not associated with a mutation in either BRCA1 or BRCA2. 3.2 MATERIALS AND METHODS 3.2.1 928 breast cancer case series and tissue microarray construction Tissue microarrays were constructed from formalin fixed, paraffin embedded breast tissues received in the Department of Pathology at the Vancouver General Hospital during the period 1976-1990. Biopsy tissue was retrieved from 930 patients with stage I, II or III breast cancer, who had participated in four British Columbia Cancer Agency clinical trials. Clinical data and outcome, including all breast cancer recurrences and deaths were available for all patients. Median follow-up time from original diagnosis, for all patients still alive at the time of analysis, was 17.4 years (range 9.8 - 28.1). Ethical approval for collection of paraffin blocks and construction of the tissue microarray for this study was received through our local research ethics board. Representative areas of invasive carcinoma were selected and marked on the hematoxylin and eosin slide and its corresponding tissue block to be sampled for the tissue microarray (TMA). The TMAs were assembled using a tissue-arraying instrument (Beecher Instruments, Silver Springs, MD) as described previously [11]. Briefly, the instrument was used to create holes in a recipient block with defined array coordinates. A stylet was used to transfer the tissue cores into the recipient block. One 0.6 mm diameter core was taken from each case. Serial 6um sections were cut for FISH analysis. 3.2.2 4150 breast cancer series The study cohort included 4150 female patients with newly diagnosed, invasive breast cancer in British Columbia. The median follow-up was 12.4 years and age at diagnosis was 60 years. All patients had been referred to the British Columbia Cancer Agency and staging, pathology, treatment and follow-up information were available. During the study era, 75% of breast cancer cases in the province were referred; non-referred patients were generally elderly or treated by mastectomy without indications for adjuvant therapy. Abstracted clinical information included age, histology, grade, tumor size, number of involved axillary nodes, lymphatic or vascular invasion (LVI), ER status, type of local and initial adjuvant systemic therapy (AST), dates of diagnosis, and first local, regional or distant recurrence 78 and death. 3.2.3 BRCA 1, BRCA2 and non-BRCA 1/2 familial breast cancers The collection of these paraffin embedded breast cancers was a collaborative effort between Vancouver General Hospital, Memorial University, McGill University, Spanish National Cancer Research Centre, Madrid, Spain, and Lund University Hospital, Lund, Sweden. The study included 64 BRCA1 mutation carriers, 54 BRCA2 mutation carriers, and 34 familial breast cancers not associated with either a mutation in BRCA1 or BRCA2. 3.2.4 Isolation of PAC or BAC clones On occasion more than one PAC/BAC clone may be found in a single well. BACPAC Resources where we order the majority of our PAC or BACs from (Oakland, CA) therefore recommends streaking some of BAC culture from the stab tube to a plate of LB agar containing 12.5 ug/ml chloramphenicol (PACs are grown in 25 ug/ml kanamycin) and grown at 37°C for 1 day. The next day 6 colonies are picked, re-streaked to another LB plus chloramphenicol plate and grown 1 day at 37°C. 3.2.5 DNA Isolation from PAC or BAC clones This is a rapid alkaline lysis miniprep method for isolating DNA from BAC clones and is a modification of a standard Qiagen method (BACPAC Resources, Oakland, CA). We have scaled up this version in order to increase our yield of 79 DNA. A single bacterial colony is inoculated into a 2 ml LB media supplemented with 12.5 ug/ml chloramphenicol and grown overnight shaking at 37°C. The next day 2 ml of this culture is then added to 50 ml of LB media supplemented with 12.5 ug/ml chloramphenicol and grown overnight shaking at 37°C. The following day, the culture is centrifuged at 3,000 rpm for 10 minutes. The supernatant is discarded. The pellet is resuspended (vortexed) in 7.5 ml of P1 solution (15mM Tris, pH8, 10mM EDTA, 100 ug/ml RNase A, Qiagen, Mississauga, Ontario). Then 7.5 ml of P2 solution (0.2N NaOH, 1% SDS, Qiagen, Mississauga, Ontario) is added and the contents of the tube are gently mixed. P3 solution (3M KOAc, pH 5.5, Qiagen, Mississauga, Ontario) is then slowly added and the tube is gently shaken. After adding P3 solution the tubes are placed on ice for 5 minutes. Next, the tubes are then centrifuged at 10,000 rpm for 10 minutes at 4°C. After centrifugation the tubes are kept on ice and the supernatant is transferred to a 50 ml Falcon tube that contains 20 ml of ice-cold isopropanol. The tubes are inverted several times to mix the sample and then placed on ice for 5 minutes. The tubes are then centrifuged at 10,000 rpm for 10 minutes at 4°C. After centrifugation the supernatant is removed and 12.5 ml of 70% EtOH is added to each tube. The tubes are inverted several times to wash the DNA pellets and then centrifuged at 10,000 rpm for 5 minutes at 4°C. Following centrifugation the supernatant is removed and pellets are air dried until they turn from white to translucent in appearance. When the ethanol has evaporated the pellet is resuspended in 2 ml of water, and may be left overnight or longer to resuspend the DNA. 80 3.2.6 Labeling of DNA probes Preparation of our DNA probes for FISH is done using the Nick Translation kit from Vysis following their established procedure (Downer's Grove, Illinois). This kit incorporates SpectrumGreen, SpectrumOrange orSpectrumRed direct-labeled dUTP into locus specific DNA probes. In this reaction, 50% of dTTP is substituted with the labeled dUTP. The assay procedure involves labeling approximately 10 ug of extracted BAC or PAC DNA. To the 10 ug of DNA (in the order listed, all reagents are provided with the kit) 2.5 ul of 0.2 mM SpectrumGreen, SpectrumOrange, or SpectrumRed is added, followed by 5 ul of 0.1 mM dTTP, 10 ul of 0.1 mM dNTP mix (an equal mix of dATP, dCTP and dGTP), 5 ul of 10x nick translation buffer and lastly 10 ul of nick translation enzyme. The tube is briefly vortexed and centrifuged. The sample is then incubated for 8 to 16 hours at 15°C and the sample is then heated at 70°C for 10 minutes to stop the reaction. Unincorporated dUTP is removed by running the labeling sample through a CentriSep column (Adelphia, NJ). Briefly, the column gel is hydrated in 800 ul of water for 1-2 hours. The columns are then spun at 3,000 rpm for 2 minutes to remove the fluid. The 50 ul labeling reaction is then added to the column and spun at 3,000 rpm for 2 minutes yielding a purified sample. The probe size is determined by running the samples on a 2% TAE gel. The labeled probe should be in the range of 50-600 bp, with an average size of 300 bp. The next step in the procedure is precipitating the probe. This procedure involves adding 30 ul of the nick translation mixture to 1 ug of COT-1 DNA, 2 ug of salmon sperm, and 24 ul of distilled water. Then 0.1 volume of 3 M 81 sodium acetate and 2.5 volumes of 100% EtOH are added, the samples are vortexed and then placed in a -20°C freezer for at least 15 minutes. The samples are then centrifuged for 30 minutes at 12,000 rpm at 4°C to pellet the DNA. The supernatant is then removed and the pellet is dried at 37°C for 8-10 minutes. The pellet is resuspended in 6 ul of distilled water. The fluorescently labeled probe is then ready to be either combined with another locus specific probe or centromere probe. 3.2.7 Fluorescent In-situ Hybridization (FISH) The tissue microarray sections were baked overnight at 60°C. Before hybridization, tissue array sections were deparaffinized in xylene and dehydrated in 100% ethanol. The slides were then subjected to pretreatment washes, which included immersing the slides in 0.2N HCI for 20 minutes, distilled water for 10 minutes, 2X SSC for 3 minutes, and 1M NaSCN at 80°C for 30 minutes. Following protease treatment of the slides (protease solution at 37°C for 15 minutes), they were dehydrated in 100% ethanol and air-dried. The EMSY probes were created from DNA isolated from the PAC clones B4, dJ18D12, and dJ85A11 (all PCR verified for exons 2 or 8 of EMSY) using the method described above. The EMSY probe was combined with Spectrum Aqua labeled centromeric probe, CEP11 (Vysis, Downer's Grove, Illinois) and Spectrum Orange labeled Cyclin D1 (Vysis, Downer's Grove, Illinois). The slides were co-denatured for 5 minutes at 73°C and hybridized for 18 hours at 37°C on a HyBrite (Vysis, Downers Grove, Illinois). Post hybridization washes were done according 82 to LSI procedure (Vysis). Slides were then counterstained with DAPI. FISH signals were enumerated in approximately 40 morphologically intact and non-overlapping nuclei. The average copy number for each probe was calculated and the amplification ratio determined. The amplification ratio was calculated as a ratio between the average copy number per cell for EMSY or Cyclin D1 and the average copy number for centromere 11. An amplification ratio & 1.5 was considered amplified. 3.2.8 Processing of the FISH data and statistical analysis The data was analyzed with SPSS for Windows statistical software package (SPSS v11, Chicago, IL). Correlation analysis was performed using the bivariate two-tailed Pearson correlation test. Differences were considered significant when the p-value was less than 0.05. Kaplan-Meier curves and survival estimates were calculated for each outcome and log-rank statistics were used to test for differences between groups. Statistical significance was declared if the p-value from a two-tailed test was less than 0.05. 83 3.3 RESULTS 3.3.1 Assessment of EMSY gene amplification in 928 primary sporadic breast tumors A three color FISH assay for EMSY, CCND1 and centromere 11 was applied to 928 sporadic breast tumors sent to Vancouver Hospital from 1976 to 1990. The 928 sporadic breast tumors were placed into 3 TMAs. FISH for EMSY gene copy number was successfully performed in 551 cases. There was no difference in survival between cases with data for EMSY amplification and those cases in which EMSY amplification could not be assessed for technical reasons (p=0.79). The mean age at diagnosis for patients with EMSY gene amplification was 47.6 years, similar to the age at diagnosis for patients without EMSY amplification (47.9 years). 3.3.1.1 EMSYgene amplification in sporadic breast cancer EMSY amplification (amplification ratio a 1.5) was seen in 70 of the 551 (13%) breast cancers. CCND1 amplification was seen in 78 breast cancers (14%) and was present in only a subset of breast cancers associated with EMSY amplification (42/70, (60%)), suggesting that both genes are independent oncogenic targets 84 3.3.1.2 Clinical correlations and prognostic significance of EMSY gene amplification We were then interested in determining whether EMSY gene amplification was associated with a particular subtype of breast cancer or identified a poor prognostic subset of breast cancer. Univariate Kaplan-Meier survival analysis showed that those patients with EMSY amplification were associated with a worse disease specific survival (DSS) as compared to those without EMSY amplification. This effect on survival was more marked in node negative (node-) patients than in node positive (node+) cases (Figure 3.1). A similar effect is seen on relapse-free survival (RFS) and overall survival (OS). To investigate whether EMSY gene amplification was associated with other clinicopathological subsets of breast cancer, Univariate cox regression analysis was again performed. The presence of EMSY amplification was also a significant prognostic indicator of both DSS and RFS in several pathological subsets of breast cancers including, tumors > 2cm (p=0.004) and in estrogen receptor negative (ER-) (p=0.002), Her2-neu positive (Her2+) (p=0.04), Her2-neu negative (Her2-) (p=0.033) and p53 positive (p53+) (p=0.021) tumors. Assessment of DSS for tumors with CCND1 amplification showed CCND1 to be of prognostic significance in all patients (p=0.0002), in node positive cases (p=0.008) and of marginal significance in node negative cases (p=0.04). A similar effect was seen on RFS survival, with the exception of node negative tumors in which a significant association with survival was not seen. CCND1 85 A) iio Disease specific survival (all cases) ival (p) :c .6 \ v. V EMSY normal | .4 .2 p=0.002 EMSY amplified 0.0 • 5 1Q 15 20 Total Follow-up (years) 1 6 > 3 p=0.0004 Disease specific survival (node positive cases) EMSY normal EMSY amplified Total Follow-up (years) Figure 3.1 E M S Y gene ampl i f icat ion and c l i n i c a l o u t c o m e EMSY gene amplification is associated with a poor disease outcome in A) al patients and B) node positive cases. Hughes-Davies et al, 2003. 86 was also associated with a poor prognosis in several other clinicopathological subsets of breast cancer including tumors <, 2 cm (p<0.001), ER- (p=0.001), Her2+ (p=0.05), Her2- (p=0.017) and p53- (p<0.0001) tumors. Cox multivariate analysis was used to determine the effects of tumor size, EMSY amplification, and ER status on predicting DSS in node negative patients. EMSY amplification (p=0.01) and tumor size (p=0.001) had prognostic significance independent of ER status and CCND1 amplification in this subset of patients. A similar effect was seen on overall survival. In the node positive subset of patients multivariate analysis showed that CCND1 amplification (p=0.007), ER status (p=0.005) and tumor size (p=0.016) were independently significant in predicting DSS in this subset of patients. 3.3.2 Assessment of EMSY amplification in a population based series of 4150 breast cancers with greater than ten years follow up data To validate our previous findings we evaluated EMSY and CCND1 gene amplification in a larger population based series of breast cancers consisting of 4150 cases for which clinical outcome was available. Data for EMSY and CCND1 gene amplification was successfully collected in 1326 cases. The cohort was equally divided into a training and validation set. The data herein represents the training set, consisting of 663 breast cancers. 87 3.3.2.1 EMSY gene amplification in large population based series of sporadic breast cancer EMSY gene amplification was seen in 87 of the 663 (13%) breast cancers. CCND1 gene amplification was seen in 188 of the 663 (28%) breast cancers. 3.3.2.2 Clinical correlations and prognostic significance of EMSY gene amplification Univariate Kaplan-Meier survival analysis showed that EMSY amplification was associated with a worse disease specific survival (DSS) in ER+ patients, node positive patients, and in grade 2 tumors (Figure 3.2). Information regarding breast cancer subtypes, such as the Her2+, luminal and basal subtypes were available for this series. EMSY amplification was associated with a poor prognosis in the luminal subtype (p=0.012) (Figure 3.2). A similar result was seen for RFS. CCND1 was associated with a shortened DSS in ER+ tumors (p=0.05), in node positive tumors (p<0.001) and in the luminal subtype (p=0.002). A similar result was seen for RFS, with the exception of ER+ and luminal subgroup in which a significant association with survival was not seen. Cox multivariate analysis was used to determine the effects of EMSY amplification, tumor size, grade, and ER status on predicting DSS in node positive patients. EMSY amplification, tumor size and grade had prognostic significance independent of ER status in this subset of patients (Table 3.1). A 88 A) > 1 -EMSY amplified p=0.011 i — —I 1 1— SAO IGJO 1!.0C Total follow up (years) C) co > 3 GO 5 30 1 3.00 1 5.00 Total follow up (years) B) Node+ EMSY amplified p=0.02 —r 5.X 10.00 15 K Total follow up (years) D) Grade 2 1.0-\ ^ - w . EMSY normal o.a-o.e-H I * EMSY amplified p=0.007 0.2-3.C-Luminal subtype EMSY normal EMSY amplified p=0.012 030 -1— '3.00 5.K -0.00 15.00 Total follow up (years) Figure 3.2 EMSY gene amplification and clinical outcome in population based series of breast cancer EMSY gene amplification is associated with a poor disease outcome in A) ER+ patients, B) node positive patients, C) grade 2 tumors, and D) luminal subtype 89 Table 3.1 Multivariate analysis of EMSY amplification and clinical outcome in node positive patients Disease specific survival HR (95% CI) P value EMSY amplification 1.9(1.2-3.0) 0.007 Tumor size 1.9(1.2-2.9) 0.005 Grade 1.8(1.2-2.8) 0.005 ER status 0.8(0.5-1.2) 0.213 Table 3.2 Multivariate analysis of CCND1 amplification and clinical outcome in node positive patients Disease specific survival HR (95% CI) P value CCND1 amplification 1.9(1.3-2.7) 0.002 Tumor size 1.9(1.2-2.9) 0.005 Grade 1.7(1.1-2.5) 0.014 ER status 0.8 (0.5-1.1) 0.124 90 similar effect was seen on RSF survival. In the node positive subset of patients multivariate analysis showed that CCND1 amplification, tumor size and grade were independently significant in predicting DSS in this subset of patients (Table 3.2). A similar effect was seen on RFS. 3.3.3 EMSY gene amplification in known BRCA1 and BRCA2 mutation carriers We hypothesized that EMSY gene amplification would be a rare event in BRCA2 mutation carriers (as these cancers are already associated with a loss of BRCA2 function) and would be a frequent event in BRCA1 mutations carriers and familial breast cancers not associated with loss of either BRCA1 or BRCA2. To investigate this hypothesis the same three color FISH assay described above was applied to breast cancers from BRCA1 and BRCA2 mutation carriers as well as familial breast cancers not associated with a BRCA1 or BRCA2 mutation. CCND1 was also included in this analysis as a control as we expected to see similar levels of CCND1 amplification in all three subsets of familial breast cancer. EMSY gene amplification was a frequent event in both BRCA1 mutation carriers and non-BRCA1/2 mutation carriers as expected (Table 3.3). In comparison, EMSY gene amplification was a significantly less common event in BRCA2 mutation carriers as compared to BRCA1 and non-BRCA1/2 cases (p=0.047). CCND1 amplification was a frequent event in BRCA1, BRCA2 and non-BRCA1/2 breast cancers (Table 3.3). Amplification of CCND1 did occur at a much higher 91 Table 3.3 Incidence of E M S V a n d CCND1 amplification in familial breast cancer EMSY Cyclin D1 amplified amplified N n (%) n (%) BRCA status BRCA1 63 15(24) 12 (19) BRCA2 56 5(9) 10(18) Non-BRCA1/2 34 9(26) 12 (35) 92 frequency in the familial breast cancers not associated with either a BRCA1/BRCA2 mutation as compared to BRCA1 and BRCA2 mutation carriers. 3.4 DISCUSSION EMSY is a recently described gene encoding a protein that has been shown to bind the transactivation domain of BRCA2 [11]. EMSY maps to 11q13.5, a region commonly amplified in breast cancer [14-17]. Also located along this region on 11q13 is CCND1 a gene encoding for cyclin D1. Cyclin D1 is a cell cycle regulator that plays a pivotal role in the regulation of progression from G1 to S phase of the cell cycle, and is a well-established oncogene in breast cancer [18, 19]. The discovery of the genes involved in the emergence and maintenance of this amplification event along 11q13 has been extensively studied in breast cancer. CCND1 has long been thought of as a the driver of amplification at 11q13, however more detailed analysis of this region led to the discovery of additional regions of amplification. In addition to CCND1 there is now evidence to support the presence of three other regions of amplification (amplicon). One such region, located 6 Mb telomeric to CCND1, contains the novel gene EMSY. However, the frequency and clinical significance of EMSY gene amplification in breast is largely unknown. We investigated EMSY gene amplification in two clinically annotated breast cancer series, the first cohort included 928 breast cancers and the second was a large population based series of 4150 breast cancers. EMSY amplification was a frequent event in both series, occurring in 13% of breast cancers. The same frequency of amplification of EMSY was actually seen in both series, demonstrating the robustness of this FISH assay for EMSY gene amplification. We also included CCND1 in the analysis as it is commonly amplified and implicated in the pathogenesis of breast cancer. CCND1 has been extensively studied in breast cancer and found amplified in 15 to 20% of breast cancer [12, 20-22]. We found a similar frequency of amplification for CCND1. In the first series, CCND1 was amplified in 14% of breast cancers, and in the second series it was amplified in 28% of breast cancers. The latter frequency is higher than previously reported for CCND1 amplification and this variation is likely due to differences in techniques and thresholds (cutoff points). These results demonstrate that EMSY is commonly amplified in breast cancer and likely plays a role in the emergence and maintenance of this amplification event at 11q13.5 Furthermore, amplification of EMSY was associated with a poor outcome in two separate breast cancer series. In the first series, EMSY amplification also strongly correlated with node negative, and estrogen receptor negative tumors. CCND1 amplification was also associated with a poor outcome, and correlated with negative ER status. In the second series there were some differences in the correlations with clinicopathological variables. For instance, in this population based series of breast cancers EMSY amplification strongly correlated with positive ER status, and node positive tumors. Amplification of CCND1 also strongly correlated with ER positive and node positive tumors. This data is in agreement with previously reported studies. Several studies have reported an association between CCND1 amplification and ER positive tumors [12, 22-24]. 94 With regards to EMSY, an association between EMSY amplification and ER positive tumors has also been reported [25]. The discrepancy between the two data sets is likely due to the fact that the first series included breast cancers from clinical trials for highly aggressive and locally advanced tumors. This series of breast cancers was therefore not representative of a population based series of breast cancers. Thus, the population based series will allow is to definitely determine the association of EMSY gene amplification with clinical outcome. Lastly, we wanted to evaluate EMSY gene amplification in familial breast cancers associated with known mutations in BRCA1 and BRCA2. EMSY has been shown to interact with the transactivation domain of BRCA2. Mutations in BRCA1 and BRCA2 are rare events in sporadic breast cancer [4, 8, 26]. Although epigenetic silencing of BRCA1 has been reported in some sporadic cancers [9, 10, 27-29], no such silencing has been reported for BRCA2 [7, 30, 31]. It has been hypothesized that EMSY amplification may be a surrogate for BRCA2 loss in sporadic cancer [32, 33]. Therefore, mutations in BRCA2 may be unnecessary when EMSY is amplified and overexpressed. Indeed, we found that EMSY was less commonly amplified in BRCA2 mutation carriers (i.e. those cancers associated with loss of BRCA2 function) (9%) as compared to BRCA1 mutations carriers (24%) and familial breast cancers not associated with a mutation in either BRCA1 or BRCA2 (29%). Further evidence to support this association between EMSY and BRCA2 mutation carriers came from the parallel study of CCND1 gene amplification in these same subsets of familial breast cancers. Amplification of CCND1 was commonly observed at a similar frequency 95 in BRCA1 and BRCA2 mutation carriers, as well as the familial non-BRCA1/2 breast cancers. Two studies have evaluated CCND1 gene amplification in BRCA1 mutation carriers, although they yielded discordant results. One study reported that CCND1 was not amplified in 30 BRCA1 carcinomas [34], whereas another study reported CCND1 amplification in 2 of 11 (18%) cases [35]. The latter study found a frequency of amplification very similar to what we found, although the number of BRCA1 cases was much smaller than the 63 cases of BRCA1 carcinomas we collected. In conclusion, our findings demonstrate that EMSY is commonly amplified in breast cancer and associated with a poor outcome in several pathological subsets of breast cancer. EMSY gene amplification is not likely to be used as a clinical biomarker. Evaluation of EMSY protein expression in breast cancer may yield a more suitable assay for clinical use. However, we have screened several commercial antibodies for EMSY and we have not found an antibody that can be used on paraffin embedded tissue. EMSY gene amplification is a less common event in BRCA2 mutation carriers, providing evidence to support the potential role of EMSY as surrogate for BRCA2 loss in sporadic breast cancer. It will be very interesting to determine whether EMSY amplified tumors share common biological features with BRCA2 mutation carriers that would reflect a common defect in BRCA2 function. 96 3.5 REFERENCES King, M.C., J.H. Marks, and J.B. Mandell, Breast and ovarian cancer risks due to inherited mutations in BRCA1 and BRCA2. Science, 2003. 302(5645): p. 643-6. Nathanson, K.L., R. Wooster, and B.L. Weber, Breast cancer genetics: what we know and what we need. Nat Med, 2001. 7(5): p. 552-6. Khoo, U.S., et al., Somatic mutations in the BRCA1 gene in Chinese sporadic breast and ovan'an cancer. Oncogene, 1999.18(32): p. 4643-6. Lancaster, J.M., et al., BRCA2 mutations in primary breast and ovarian cancers. Nat Genet, 1996.13(2): p. 238-40. Merajver, S.D., et al., Germline BRCA1 mutations and loss of the wild-type allele in tumors from families with eady onset breast and ovan'an cancer. Clin Cancer Res, 1995.1(5): p. 539-44. i. Miki, Y., et al., Mutation analysis in the BRCA2 gene in primary breast cancers. Nat Genet, 1996. 13(2): p. 245-7. 7. Collins, N., R. Wooster, and M.R. Stratton, Absence of methylation of CpG dinucleotides within the promoter of the breast cancer susceptibility gene BRCA2 in normal tissues and in breast and ovarian cancers. Br J Cancer, 1997. 76(9): p. 1150-6. 3. Futreal, P.A., et al., BRCA1 mutations in primary breast and ovarian carcinomas. Science, 1994. 266(5182): p. 120-2. 9. Catteau, A., et al., Methylation of the BRCA 1 promoter region in sporadic breast and ovarian cancer: correlation with disease characteristics. Oncogene, 1999.18(11): p. 1957-65. 10. Esteller, M., et al., Promoter hypermethylation and BRCA1 inactivation in sporadic breast and ovarian tumors. J Natl Cancer Inst, 2000. 92(7): p. 564-9. 11. Hughes-Davies, L., et al., EMSY links the BRCA2 pathway to sporadic breast and ovarian cancer. Cell, 2003. 115(5): p. 523-35. 12. Dickson, C , et al., Amplification of chromosome band 11q13 and a role fc cyclin D1 in human breast cancer. Cancer Lett, 1995. 90(1): p. 43-50. 97 13. Schuuring, E., The involvement of the chromosome 11q13 region in human malignancies: cyclin D1 and EMS1 are two new candidate oncogenes-a review. Gene, 1995. 159(1): p. 83-96. 14. Hui, R., et al., EMS 1 gene expression in primary breast cancer: relationship to cyclin D1 and oestrogen receptor expression and patient survival. Oncogene, 1998.17(8): p. 1053-9. 15. Lammie, G.A., et al., D11S287, a putative oncogene on chromosome 11q13, is amplified and expressed in squamous cell and mammary carcinomas and linked to BCL-1. Oncogene, 1991. 6(3): p. 439-44. 16. Ormandy, C.J., et al., Cyclin D1, EMS1 and 11q13 amplification in breast cancer. Breast Cancer Res Treat, 2003. 78(3): p. 323-35. 17. Schuuring, E., et al., Amplification of genes within the chromosome 11q13 region is indicative of poor prognosis in patients with operable breast cancer. Cancer Res, 1992. 52(19): p. 5229-34. 18. Jiang, W., et al., Overexpression of cyclin D1 in rat fibroblasts causes abnormalities in growth control, cell cycle progression and gene expression. Oncogene, 1993. 8(12): p. 3447-57. 19. Wang, T.C., et al., Mammary hyperplasia and carcinoma in MMTV-cyclin D1 transgenic mice. Nature, 1994. 369(6482): p. 669-71. 20. Gillett, C , et al., Amplification and overexpression of cyclin D1 in breast cancer detected by immunohistochemical staining. Cancer Res, 1994. 54(7): p. 1812-7. 21. Karlseder, J., et al., Patterns of DNA amplification at band q13of chromosome 11 in human breast cancer. Genes Chromosomes Cancer, 1994. 9(1): p. 42-8. 22. Michalides, R., et al., A clinicopathological study on overexpression of cyclin D1 and ofp53 in a series of 248 patients with operable breast cancer. Br J Cancer, 1996. 73(6): p. 728-34. 23. Bieche, I., et al., Prognostic value ofCCNDI gene status in sporadic breast tumours, as determined by real-time quantitative PCR assays. Br Cancer, 2002. 86(4): p. 580-6. 98 24. Courjal, F., et al., Mapping of DNA amplifications at 15 chromosomal localizations in 1875 breast tumors: definition of phenotypic groups. Cancer Res, 1997. 57(19): p. 4360-7. 25. Rodriguez, C , et al., Amplification of the BRCA2 pathway gene EMSY in sporadic breast cancer is related to negative outcome. Clin Cancer Res, 2004. 10(17): p. 5785-91. 26. Sorlie, T., et al., Mutation screening ofBRCAl using PTTand LOH analysis at 17q21 in breast carcinomas from familial and non-familial cases. Breast Cancer Res Treat, 1998. 48(3): p. 259-64. 27. Baldwin, R.L., et al., BRCA1 promoter region hypermethylation in ovarian carcinoma: a population-based study. Cancer Res, 2000. 60(19): p. 5329-33. 28. Geisler, J.P., et al., Frequency of BRCA 1 dysfunction in ovarian cancer. J Natl Cancer Inst, 2002. 94(1): p. 61-7. 29. Rice, J .C., et al., Methylation of the BRCA 1 promoter is associated with decreased BRCA 1 mRNA levels in clinical breast cancer specimens. Carcinogenesis, 2000. 21(9): p. 1761-5. 30. Gras, E., et al., Loss of heterozygosity on chromosome 13q12-q14, BRCA-2 mutations and lack of BRCA-2 promoter hypermethylation in sporadic epithelial ovarian tumors. Cancer, 2001. 92(4): p. 787-95. 31. Hilton, J . L., et al., Inactivation of BRCA 1 and BRCA2 in ovarian cancer. J Natl Cancer Inst, 2002. 94(18): p. 1396-406. 32. King, M.C., A novel BRCA2-binding protein and breast and ovarian tumorigenesis. N Engl J Med, 2004. 350(12): p. 1252-3. 33. Livingston, D.M., EMSY, a BRCA-2 partner in crime. Nat Med, 2004. 10(2): p. 127-8. 34. Vaziri, S.A., et al., Absence of CCND1 gene amplification in breast tumours ofBRCAl mutation carriers. Mol Pathol, 2001. 54(4): p. 259-63. 35. Palacios, J., et al., Phenotypic characterization ofBRCAl and BRCA2 tumors based in a tissue microarray study with 37 immunohistochemical markers. Breast Cancer Res Treat, 2005. 90(1): p. 5-14. 99 C h a p t e r 4 : E M S V G E N E AMPLIFICATION IN OVARIAN C A N C E R Objective: To determine the frequency and clinical significance of EMSY gene amplification in ovarian cancer and its involvement in the emergence and maintenance of the 11q13 amplicon in ovarian cancer. Based on the Manuscript: Brown LA. Irving J , Parker R, Kim H, Press JZ, Longacre TA, Chia S, Magliocco A, Makretsov N, Gilks B, Pollack J , Huntsman D. Amplification of EMSY, a novel oncogene on 11q13, in high grade ovarian surface epithelial carcinomas. Gynecol Oncol. 2006 Feb; 100(2):264-70. * * Since the publication of this manuscript we've come to recognize that rather than being a driver of the 11q13 amplicon, the oncogenic effects of EMSY and other genes on 11 q13 may act synergistically to promote this amplification event. This is discussed in further detail in Chapter 5. The candidate performed all experiments presented, except for the array CGH that was performed in Jon Pollack's Laboratory at Stanford University. 100 4.1 INTRODUCTION DNA amplification has long been recognized as a means of oncogene activation during tumor development. Some genomic regions are frequently amplified in tumors; an example is chromosomal band 11q13. Amplification of this region is commonly seen in breast, head and neck, lung and bladder cancer [1]. This region of 11q13 is gene dense, which has led to investigations into the potential genes that confer a growth advantage to tumor cells. Two genes, CCND1 and EMS 1 have emerged as strong candidate genes for playing a causal role in the emergence of this amplification event on 11q13 in breast cancer [2, 3]. However, detailed mapping of this region in breast cancer led to the discovery of two additional regions of amplification at 11q13, suggesting the presence of at least two other genes involved in breast carcinogenesis [4, 5]. CCND1 and EMS1 are located within 0.8 Mb of each other at 11q13.3. Both CCND1 and EMS1 have been found commonly amplified and overexpressed in breast cancer [6, 7]. A more telomeric region of amplification is found 6 Mb away and encompasses the gene glycoprotein A repetitions predominant (GARP) [4, 5]. Amplification of this region is centered on the marker D11S533E, which was recently identified as the novel oncogene EMSY [8]. EMSY encodes a protein that binds and represses the activity of the transactivation domain of BRCA2. We have previously reported that EMSY gene amplification was found in 13% of sporadic breast cancers and in 17% of high grade ovarian cancers [8]. A recent publication demonstrated a strong correlation between overexpression of EMSY and gene amplification in several breast cancer cell lines and primary breast 101 tumors [9]. These data support the role of EMSY as an oncogene in breast cancer. To determine EMSYs involvement in the emergence and maintenance of the 11q13 amplicon in ovarian cancer, a previously scored 674 ovarian tissue microarray was further analyzed to determine the histological features associated with EMSY amplification [8], We then looked in more detail at the genes involved in the formation of the 11q13 amplicon, using array comparative genomic hybridization (aCGH) to map this region in 51 ovarian cancers. Lastly, we evaluated EMSY gene amplification and RNA expression along with that of other 11q13 genes, CCND1 and p21/cdc42/Rad-activated kinase (PAK1), a novel oncogene [10] in 22 cases of ovarian cancer for which there were corresponding snap frozen tumor samples. 4.2 MATERIALS AND METHODS 4.2.1 Ovarian Cancer Case Series and Tissue Microarray Construction (TMA) Tissue microarrays were constructed from 674 archival formalin-fixed, paraffin-embedded ovarian tumour samples from Vancouver Hospital, Stanford Medical Centre, and Memorial University. These included a spectrum of high and low grade tumours, and cognate normal tissue. Representative areas of invasive carcinoma were selected and marked on the hematoxylin and eosin slide and its corresponding tissue block to be sampled for the tissue microarray (TMA). The TMAs were assembled using a tissue-arraying instrument (Beecher Instruments, 102 Silver Springs, MD) as described previously [11]. Briefly, the instrument was used to create holes in a recipient block with defined array coordinates. A stylet was used to transfer the tissue cores into the recipient block. One 0.6 mm diameter core was taken from each case. Serial 6pm sections were cut for FISH analysis. 4.2.2 Fluorescent In-situ Hybridization (FISH) The tissue microarray sections were baked overnight at 60°C. Before hybridization, tissue array sections were deparaffinized in xylene and dehydrated in 100% ethanol. The slides were then subjected to pretreatment washes, which included immersing the slides in 0.2N HCI for 20 minutes, distilled water for 10 minutes, 2X SSC for 3 minutes, and 1M NaSCN at 80°C for 30 minutes. Following protease treatment of the slides (protease solution at 37°C for 15 minutes), they were dehydrated in 100% ethanol and air-dried. The EMSY probes were created from DNA isolated from the PAC clones B4, dJ18D12, and dJ85A11 (all PCR verified for exons 2 or 8 of EMSY). EMSY probes were directly labeled, with Spectrum Green, by nick translation (Vysis, Downer's Grove, Illinois). The EMSY probe was combined with Spectrum Aqua labeled centromeric probe, CEP11 (Vysis) and Spectrum Orange labeled CCND1 (Vysis). PAK1 probe was created from DNA isolated from the BAC clone CTD-2004N10 and the DNA was then directly labeled with Spectrum Orange, by nick translation (Vysis, Downer's Grove, Illinois). The PAK1 probe was then combined with Spectrum Aqua labeled centromeric probe, CEP11 (Vysis). The slides and probe were co-denatured for 5 minutes at 73°C and hybridized for 18 103 hours at 37°C on a HyBrite (Vysis, Downers Grove, Illinois). Post hybridization washes were done according to LSI procedure (Vysis). Slides were then counterstained with DAPI. FISH signals were enumerated in approximately 40 morphologically intact and non-overlapping nuclei. The average copy number for each probe was calculated and the amplification ratio determined. The amplification ratio was calculated as a ratio between the average copy number per cell for EMSY or CCND1 and the average copy number for centromere 11. 4.2.3 Array Comparative Genomic Hybridization (aCGH) The cDNA microarrays used in this study included 28,000 unique characterized genes or ESTs represented by a total of 41,859 cDNAs, printed on glass slides by the Stanford Functional Genomics Facility (http://www.micrQarray.org/sfqf/jsp/home.isp). The details of the construction of these arrays have been described previously [12]. Of the over 41,000 cDNA sequences represented on the microarrays used for this study, the chromosomal localization is known for 35,151 distinct mapped cDNAs, which represent 24,540 different Unigene clusters, and 3,225 cDNAs not yet represented in Unigene clusters. Tumor DNA from formalin-fixed, paraffin-embedded tissue and reference DNA (normal gender-matched human leukocytes) was extracted, using protocols available at the following website (http://qenome-www.stanford.edu/DFSP/) . Reference DNA was digested with Dpnll before further processing. Gel electrophoresis of digested and non-digested DNA isolated from formalin-fixed, paraffin-embedded tissue was run to determine DNA fragment size. Labeling of DNA isolated from tumor samples, after light-104 microscopic confirmation of the presence of non-necrotic tumor, was performed as described previously [13] (http://cmpm.stanford.edu/pbrown/protocois/index.htmJ). Briefly, 4 micrograms of tumor DNA was fluorescently labeled (Cy5) in a volume of 50 microliters, mixed with reference DNA labeled with Cy3, and hybridized overnight to the array. After washing, the array slides were scanned on a GenePix Scanner (Axon Instruments, Foster City CA) and fluorescence ratios (test/control) calculated using GenePix software. Only cDNA spots with a ratio of signal over background of at least 1.4 in the Cy3 channel were included in further analysis. Chromosomal localization of the mapped genes was assigned as described previously [13] and is based on Goldenpath data from November 2002. For CGH data the copy number for each locus was based on either single gene values or a moving average of the five nearest cDNA clones centered on that locus. 4.2.4 Real-time Quantitative PCR Real-time PCR was carried out on an ABI 7900 Sequence Detection System (Applied Biosystems) under standard conditions. The EMSY TaqMan probe sequence was ctggatctcagcgggatgaatgcaaaagaa. The forward and reverse primers used for EMSY were 5'-tgcctgttgtgtggccaa-3' and 5'-ggacacgggcttcacgtg-3'. CCND1 and PAK1 probes with primers included were obtained as a 20x target assay from Applied Biosystems (Foster, CA). Expression levels were first normalized to rRNA and then the ratio of this expression level to the mean normalized expression level of all 22 samples was calculated. Expression levels 105 were similarly compared to the expression of Stratagene's Universal Human Reference RNA mix consisting of 10 different human cell lines (La Jolla, CA). Statistical Analysis: Statistical analysis was performed using SPSS 11.0 (Chicago, IL). A non-parametric chi-squared test was used was used to determine the significance of EMSY amplification events in high grade papillary serous versus grade 1 serous carcinoma and serous borderline tumors. Pearson correlation was used to assess the correlation between gene amplification and mRNA expression levels. 4.3 RESULTS 4.3.1 EMSY amplification in ovarian cancer A three color FISH assay for EMSY, CCND1 and centromere 11 copy number was successfully applied to 674 ovarian tumors or pseudoneoplastic lesions from three TMAs including cases from Vancouver Hospital, Stanford Medical Centre, and Memorial University as previously described [8] (Figure 4.1). The frequency of EMSY gene amplification in this series was, as reported previously, 17% [8]. The ovarian tissue arrays encompassed a wide range of ovarian pathologies and although the frequency of EMSY gene amplification was already determined, correlation with histological subtypes was not performed. The majority of cases were high grade serous carcinomas; however, all common subtypes of ovarian cancers were included on the arrays. FISH signals were successfully scored in 508 cases. The remaining 166 cases from this series were not scored as a result of core loss or lack of FISH signals. The histological subtypes associated with 106 Figure 4.1 Assessment of EMSY gene amplification in three ovarian carcinomas. A) EMSY (green), CCND1 (orange) and centromere 11 (blue) are present in normal copy number. B) EMSY gene amplification with normal copy number of CCND1 and centromere 11. C) EMSY, and CCND1 are present in higher copy number than centromere 11. 107 EMSY amplification can be seen in Table 4.1. No amplification was seen in normal ovary (n=5), germ cell tumors (n=4), sex cord stromal tumors (n=12), endometriosis (n=11), benign tumors, borderline tumors of any subtype (n=89), or in mucinous tumors of any grade (n=27). The frequency of EMSY amplification events in high grade (grade 2 or 3) serous carcinoma (52/285, 18%) was significantly greater than that seen in grade 1 serous carcinoma or serous borderline tumors (0/76, 0%)(p<0.0001). The histological subtypes associated with CCND1 amplification are also shown in Table 4.1. As was the case for EMSY amplification, amplification of CCND1 was not seen in normal ovary, germ cell tumors, sex cord stromal tumors, endometriosis, benign tumors, borderline tumors of any subtype, or in mucinous tumors of any grade. Overall, amplification of EMSY alone (37/360, 10%) was a significantly more frequent event than CCND1 amplification alone (14/360, 4%) in ovarian carcinomas (p=0.001). 4.3.2 aCGH mapping of the 11q13 amplicon in ovarian cancer A detailed mapping of the 11q13 amplicon was obtained for 51 ovarian cancers (Figure 4.2). Elevated copy numbers were observed across the 11q13 region. The highest frequency of copy number gain encompassed the genes EMSY, GARP, PAK1, and GRB2-associated binding protein 2 (GAB2) (Figure 4.2). CCND1, emsl sequence (EMS1), and procarboxypeptidase (PROP) exhibited lesser frequencies of copy number gain, suggesting that the minimal region 108 Table 4.1 Evaluation of EMSY and CCND1 amplification in ovarian cancer EMSY CCND1 Co-amplified amplified amplified N n(%) n(%) n(%) Histopatholoaical Diagnosis Undifferentiated 10 3 (30) 2 (20) 2 (20) Clear cell 38 3 (8) 1 (4) 1 (3) Endometrioid high grade 27 4 (15) 1 (4) 1 (4) Serous high grade 285 52 (18) 35 (12) 21 (7) C\clin bi EMSI 1 \1SY (iARP PAKI GAB2 l'R(T Single yenc copy number ; 51 (10%) 1051 (20%) IW5I (37%) 1" 51 133%) 1951 (37%) 15 51 (29%J 7 51 (14%) \1o\ iny a\ CBNje S 51 II6%) 1151 (22%) 19 51 (37%) 17/51 (33%) 1951 (37%) 17 51 (33%) 3 51 (6%) copy number Figure 4.2 Chromosome 11 array CGH data from 51 ovarian carcinomas based on a moving average of the five nearest genes centered on that locus. A) Each row represents a single case of ovarian carcinoma and each column is a cDNA mapped to the chromosomal locus depicted above the aCGH data. The cDNA copy number is depicted graphically with red being copy number greater than normal and green less than normal, with the color intensity corresponding to copy number according to the scale shown. B) Graphical representation of the 11q13 amplicon. Case 13 and 30 both show that CCND1 and EMS1 are not present in increased copy number, and the boundary of the amplicon is centromeric to EMSY and telomeric to PAK1. C) Chromosome 11 ideogram showing the location of the 11q13 amplicon. The table below shows copy numbers expressed as single gene values, and as a moving average of each gene and the nearest four flanking genes for 7 genes within the 11q13 amplicon. The criterion used for calling a gain at a particular gene or moving-average-5 locus was a ratio > 1.2, which represents ratios above the 2 SD threshold in normal-normal hybridizations. 111 encompassed by the 11q13 amplicon in ovarian cancer is bounded by EMSY at the centromeric end and GAB2 at the telomeric end. 4.3.3 Correlation between EMSYgene amplification and RNA expression To determine whether EMSY gene amplification correlated with an increase in RNA expression, we studied EMSY gene amplification by FISH and RNA expression levels by quantitative real-time PCR in 22 ovarian tumors. To further establish EMSYs involvement in the formation of the 11q13 amplicon, CCND1 and PAK1 were also included in the FISH and real-time PCR analysis. The resulting amplification profiles for each of the three genes along 11q13 can be seen in Table 4.2. Amplification of CCND1, EMSY and PAK1 (amplification ratio > 1.5) was commonly seen in the ovarian carcinomas analyzed. The majority of cases in which EMSY is amplified also showed co-amplification of CCND1 and PAK1. The expression profile for all three genes can also be seen in Table 4.2 (right hand column, expression levels are based on the comparison to the mean normalized expression of all 22 cases). Correlations between RNA expression and gene amplification were calculated for CCND1, EMSY and PAK1. EMSY gene amplification showed the strongest correlation with RNA expression (p<0.0001). CCND1 (p=0.003) and PAK1 (p=0.009) also showed a significant correlation, although to a lesser degree. The above data was based on the comparison to the mean normalized expression of all 22 samples; when Stratagene's Universal Human Reference RNA was used as the reference sample, the same correlations between gene amplification and overexpression 112 Table 4.2 Gene amplification and overexpression profiles of CCND1, EMSY and PAK1 in twenty-two ovarian tumors. Amplification ratios RNA expression levels Case# CCND1 EMSY PAK1 CCND1 EMSY PAK1 1 14 1.2 1.2 0.1 0.7 0.7 2 2.1 1.4 2.0 1.2 2-3 3 1.0 0.8 0.9 1.4 1.2 1.0 4 1.6 2.4 1.2 1.5 5 1.8 | | | g | j l 5^H| | | l 3 1 2.9 6 1.3 1.2 0.7 0.3 0.9 1.3 7 1.1 0.9 0.8 0.3 0.5 0.5 8 1.0 0.9 0.8 .. 2.4 1.2 0.9 9 25 2.4 2.4 0.8 1.3 1.0 10 •BBMf 1.4 1.0 0.2 1 8 2 . i , 11 1.8 1.4 2.3 26 1.0 12 1.2 1.2 0.9 0.9 0.9 1.0 13 1.2 1.0 1.0 1.8 0.4 0.5 14 [- T o 1 1.3 1.4 1.5 0.6 0.9 15 1.1 0.8 1.2 6 0 2 9 2.1 16 1.1 1.1 1.1 0.7 0.4 0.5 17 1.1 1.0 0.9 0.3 0.2 0.6 18 0.9 0.9 1.0 0.3 0.2 0.4 19 4.6 3.7 7.8 6.1 3.5 20 1.2 1 5 25 25 0.8 21 1.5 1.8 1.7 0.5 1.1 1.0 22 0.9 0.9 0.9 0.8 0.6 0.3 Table 4.2 Legend: Gene amplification ratios £ 1.5 are considered amplified and those cases associated with amplification are shaded. For RNA expression the shaded areas represent the top 6 expression levels for each gene. Genes are ordered according to their position on 11q13. 113 were observed (EMSY (pO.0001), CCND1 (p=0.003) and PAK1 (p=0.011))(data not shown). 4.4 DISCUSSION Amplification of 11 q13 is frequently seen in several tumor types, including head and neck, oral, lung and breast cancer [14]. In breast cancer, amplification of this region has been reported in 15% of carcinomas [3] and is associated with a poor prognosis [15]. This region on chromosome 11 contains several candidate genes, including CCND1 and EMS1 that could be responsible for the emergence and maintenance of this amplification event. CCND1 is a member of a family of proteins that regulate the activity of cyclin dependent kinases and is an important regulator of the cell cycle [16]. Overexpression of CCND1 shortens the G1 phase of the cell cycle allowing cells to bypass critical checkpoints leading to the accumulation of mutations and genomic instability [17, 18]. CCND1 is amplified in 15-20% of breast cancers [19-21] and overexpressed in up to 80% of tumors [1, 3, 22, 23]. CCND1 amplification and overexpression has been associated with a poor prognosis in estrogen receptor positive breast tumors [24]. EMS1, located 0.8 Mb telomericto CCND1, encodes for the human homolog of cortactin, a cytoskeletal actin binding protein [24, 25]. Several studies have shown that both CCND1 and EMS1 are frequently co-amplified in breast cancer [15]. EMS1 amplification can, however, occur independently of CCND1 amplification in breast cancer and is associated with an increased risk of relapse and death in estrogen negative tumors [2, 6]. 114 Detailed cytogenetic mapping of 11 q13 in breast cancer has identified four distinct regions of amplification [3, 4, 20] suggesting the involvement of other genes in addition to CCND1 and EMS 1. Other potential oncogenes telomeric to CCND1 and EMS1 may exist within thel 1q13 amplicon including GARP, EMSY [8] and PAK1 [10]. EMSY is a recently cloned gene encoding a protein that binds and represses the activity of the transactivation domain of BRCA2. EMSY also localizes to sites of repair following DNA damage and could be a surrogate for BRCA2 loss in sporadic breast cancer. We have previously reported that EMSY is amplified in 13% of sporadic breast cancers and EMSY amplification was associated with a poor prognosis [8]. This finding has been confirmed in a recent study, which also showed a strong association between EMSY gene amplification and overexpression in primary breast tumors and cell lines [9]. These data suggest the involvement of EMSY is an important element of the 11q13 amplicon in breast cancer, however little is known about EMSY aberrations in ovarian cancer. Here we present additional evidence demonstrating that EMS Vis frequently amplified in sporadic ovarian cancers. In particular, EMSY amplification is associated with specific histological subtypes of ovarian cancers, which include high grade papillary serous, clear cell, endometrioid and undifferentiated carcinomas. Amplification of CCND1 was observed in the same histological subtypes associated with EMSY amplification, but less frequently. An association between CCND1 protein overexpression and borderline or well differentiated, grade 1 ovarian tumors has been reported [7]. This dissimilarity in 115 findings may be due to the smaller number (n=43) of sporadic ovarian carcinomas that were previously assessed for CCND1 overexpression [7]. The data presented herein strongly demonstrates that EMSY amplification is associated with most subtypes of high grade carcinomas but not mucinous carcinomas, borderline tumors or non-epithelial tumors. Interestingly, this is the same profile of tumors associated with germline BRCA2 mutation [26-28]. EMSY gene abnormalities could be a surrogate for BRCA2 loss in ovarian cancer [8, 29-31]. Therefore, EMSY abnormalities may denote a second and larger subset of ovarian cancers that could potentially respond to therapies targeting cells with BRCA1 and BRCA2 loss, such as PARP inhibitors. aCGH mapping of 11q13 showed the most frequently amplified region in ovarian cancer to encompass EMSY, GARP, GAB2 and PAK1. We have previously shown that EMSY amplification was accompanied by GARP amplification in several cancer cell lines. In all of these cell lines EMS Vis overexpressed, as detected by real-time RT-PCR and GARP expression is not detectable, providing evidence that GARP is not the oncogenic gene within this region [8]. PAK1, encoding a serine/threonine kinase involved in the regulation of anchorage independent growth, invasiveness, and abnormal mitotic spindle organization [32, 33], has recently been shown to be present in higher copy number in ovarian cancer [10] and array CGH confirms this. These results suggest that EMSY and PAK1 may both be oncogenic elements in the 11q13 amplicon in ovarian cancer. 116 Gene amplification should lead to overexpression of genes playing a significant role in oncogenesis. Although EMSY gene amplification correlates with overexpression in breast cancer, it has not previously been studied in ovarian cancer. A strong significant correlation between EMSY gene amplification and mRNA expression was seen. A correlation, although to a lesser degree, was also seen between CCND1 and PAK1 amplification and expression. In summary, these results demonstrate that EMSY amplification is associated with high grade ovarian carcinomas, with the exception of mucinous carcinomas and is a frequently amplified element of the 11q13 amplicon in ovarian cancer. Amplification of EMSY leads to overexpression, further supporting its role in oncogenesis. The complexity in the structure of the amplicons on 11q13 makes it unlikely that one gene is entirely responsible for the emergence of this amplification event. It is likely that co-amplification of several oncogenes on 11q13 are required to provide a selective advantage necessary for the development and progression of cancer. 117 4.5 REFERENCES 1. Peters, G., et al., Chromosome 11q13 markers and D-type cyclins in breast cancer. Breast Cancer Res Treat, 1995. 33(2): p. 125-35. 2. Hui, R., et al., EMS1 amplification can occur independently of CCND1 or INT-2 amplification at 11q13 and may identify different phenotypes in primary breast cancer. Oncogene, 1997. 15(13): p. 1617-23. 3. Karlseder, J., et al., Patterns of DNA amplification at band q13 of chromosome 11 in human breast cancer. Genes Chromosomes Cancer, 1994. 9(1): p. 42-8. 4. Bekri, S., et al., Detailed map of a region commonly amplified at 11q13->q14 in human breast carcinoma. Cytogenet Cell Genet, 1997. 79(1-2): p. 125-31. 5. Ormandy, C.J., et al., Cyclin D1, EMS1 and 11q13 amplification in breast cancer. Breast Cancer Res Treat, 2003. 78(3): p. 323-35. 6. Hui, R., et al., EMS1 gene expression in primary breast cancer: relationship to cyclin D1 and oestrogen receptor expression and patient survival. Oncogene, 1998. 17(8): p. 1053-9. 7. Worsley, S.D., B.A. Ponder, and B.R. Davies, Overexpression of cyclin D1 in epithelial ovarian cancers. Gynecol Oncol, 1997. 64(2): p. 189-95. 8. Hughes-Davies, L, et al., EMSY links the BRCA2 pathway to sporadic breast and ovarian cancer. Cell, 2003. 115(5): p. 523-35. 9. Rodriguez, C., et al., Amplification of the BRCA2 pathway gene EMSY in sporadic breast cancer is related to negative outcome. Clin Cancer Res, 2004. 10(17): p. 5785-91. 10. Schraml, P., et al., Combined array comparative genomic hybridization and tissue microarray analysis suggest PAK1 at 11q13.5-q14 as a critical oncogene target in ovarian carcinoma. Am J Pathol, 2003. 163(3): p. 985-92. 11. Parker, R.L., et al., Assessment of interiaboratory variation in the immunohistochemical determination of estrogen receptor status using a breast cancer tissue microarray. Am J Clin Pathol, 2002. 117(5): p. 723-8. 118 Perou, C M . , et al., Molecular portraits of human breast tumours. Nature, 2000. 406(6797): p. 747-52. Pollack, J.R., et al., Microarray analysis reveals a major direct role of DNA copy number alteration in the transcriptional program of human breast tumors. Proc Natl Acad Sci U S A , 2002. 99(20): p. 12963-8. Schuuring, E., et al., Amplification of genes within the chromosome 11q13 region is indicative of poor prognosis in patients with operable breast cancer. Cancer Res, 1992. 52(19): p. 5229-34. Schuuring, E., et al., Identification and cloning of two overexpressed genes, U21B31/PRAD1 andEMSI, within the amplified chromosome 11q13 region in human carcinomas. Oncogene, 1992. 7(2): p. 355-61. Baldin, V., et al., Cyclin D1 is a nuclear protein required for cell cycle progression in G1. Genes Dev, 1993. 7(5): p. 812-21. Jiang, W., et al., Overexpression of cyclin D1 in rat fibroblasts causes abnormalities in growth control, cell cycle progression and gene expression. Oncogene, 1993. 8(12): p. 3447-57. Almasan, A., et al., Genetic instability as a consequence of inappropriate entry into and progression through S-phase. Cancer Metastasis Rev, 1995. 14(1): p. 59-73. Champeme, M.H., et al., 11q13 amplification in local recurrence of human primary breast cancer. Genes Chromosomes Cancer, 1995. 12(2): p. 128-33. Courjal, F., et al., Mapping of DNA amplifications at 15 chromosomal localizations in 1875 breast tumors: definition of phenotypic groups. Cancer Res, 1997. 57(19): p. 4360-7. Lammie, G.A. and G. Peters, Chromosome 11q13 abnormalities in human cancer. Cancer Cells, 1991. 3(11): p. 413-20. Buckley, M.F., et al., Expression and amplification of cyclin genes in human breast cancer. Oncogene, 1993. 8(8): p. 2127-33. Gillett, C , et al., Amplification and overexpression of cyclin D1 in breast cancer detected by immunohistochemical staining. Cancer Res, 1994. 54(7): p. 1812-7. 119 24. Fantl, V., et al., Gene amplification on chromosome band 11q13 and oestrogen receptor status in breast cancer. Eur J Cancer, 1990. 26(4): p. 423-9. 25. Schuuring, E., et al., The product of the EMS1 gene, amplified and overexpressed in human carcinomas, is homologous to a v-src substrate and is located in cell-substratum contact sites. Mol Cell Biol, 1993.13(5): p. 2891-98. 26. Pharoah, P. D., et al., Survival in familial, BRCA 1-associated, and BRCA2-associated epithelial ovarian cancer. United Kingdom Coordinating Committee for Cancer Research (UKCCCR) Familial Ovarian Cancer Study Group. Cancer Res, 1999. 59(4): p. 868-71. 27. Risch, H.A., et al., Prevalence and penetrance ofgermline BRCA1 and BRCA2 mutations in a population series of 649 women with ovarian cancer. Am J Hum Genet, 2001. 68(3): p. 700-10. 28. Rubin, S.C., BRCA-related ovarian carcinoma. Cancer, 2003. 97(9): p. 2127-9. 29. King, M.C., A novel BRCA2-binding protein and breast and ovarian tumorigenesis. N Engl J Med, 2004. 350(12): p. 1252-3. 30. Livingston, D.M., EMSY, a BRCA-2 partner in crime. Nat Med, 2004. 10(2): p. 127-8. 31. Turner, N., A. Tutt, and A. Ashworth, Hallmarks of BRCAness' in sporadic cancers. Nat Rev Cancer, 2004. 4(10): p. 814-9. 32. Adam, L, et al., Heregulin regulates cytoskeletal reorganization and cell migration through the p21-activated kinase-1 via phosphatidylinositol-3 kinase. J Biol Chem, 1998. 273(43): p. 28238-46. 33. Adam, L., et al., Regulation of microfilament reorganization and invasiveness of breast cancer cells by kinase dead p21-activated kinase-1. J Biol Chem, 2000. 275(16): p. 12041-50. 120 Chapters: A C O M P R E H E N S I V E A N A L Y S I S O F T H E 1 1 Q 1 3 AMPLICON IN OVARIAN C A N C E R Objective: To investigate the amplification of CCND1, EMSY, PAK1, Rsf-1 and GAB2 and the prognostic significance of these five genes in histologically and clinically defined subsets of early ovarian cancer Manuscript in Preparation: Brown LA, Kalloger SE, Miller MA, Shih IM, Gilks B and Huntsman D. Amplification of the 11q13 amplicon, involving multiple putative oncogenes, is associated with a serous histology and a poor prognosis in ovarian cancer. The candidate performed or supervised all experiments presented. A statistician at GPEC performed all statistical analysis. 121 6.1 INTRODUCTION DNA amplification is a common mechanism leading to oncogenic activation in human cancer. Amplification of chromosome 11q13 occurs in numerous human malignancies, including breast, ovarian, head and neck, bladder and esophageal cancer [1]. Detailed studies of 11q13 amplification in breast cancer identified amplification of the area inclusive of CCND1 and EMS1 [2-4], implicating the involvement of these genes in the emergence of amplification at this locus. CCND1 and EMS1 are located within 0.8 Mb of each other at 11q13.3, with CCND1 occupying a more centromeric position. Larger studies of breast carcinomas confirmed amplification of CCND1 and EMS1, however, these studies also revealed amplification around the D11S97 marker and glycoprotein A repetitions predominant (GARP), proximal and distal, respectively, to CCND1 and EMS1 [5, 6]. These findings established that the 11q13 region, spanning over 7 Mb, encompasses four distinct regions of amplification. These regions may be independently amplified, as small amplicons, or may be part of larger amplicons spanning some or all of these four regions, resulting in co-amplification of multiple oncogenes. CCND1 and EMS1 have long been established as drivers of amplification of the central amplicons, however the functionally important genes from the distal and proximal amplicons are largely unknown. Several genes have been proposed as potential drivers of the distal amplicon at 11q13.5, including EMSY [7, 8], p21/cdc42/Rad-activated kinase (PAK1) [5, 9], Rsf-1 (HBXAPalpha) [10] and GAB2 [11] (Figure 5.1). EMSY is a novel 122 Chromosome 11 Figure 5.1 The 11q13 amplicon and the potential genes involved in the amplification at this locus 123 oncogene encoding a protein that interacts with the transactivation domain of BRCA2. We have previously reported that EMSY is amplified in 13% of sporadic breast cancers, and EMSY gene amplification is associated with a poor prognosis [8]. This finding was confirmed in another study, which also demonstrated a strong association between EMSY gene amplification and overexpression [12]. We have also shown that EMSY is amplified in 18% of high grade ovarian cancers and amplification of EMSY strongly correlated with RNA expression [7]. PAK1 is a member of a family of serine/threonine kinases that has been found to regulate anchorage-independent growth, invasiveness, and the abnormal organization of mitotic spindles of human breast cancer cells [13]. Amplification of PAK1 has been reported in breast [5], and ovarian cancer [9]. We have also reported a significant association between PAK1 gene amplification and RNA expression in ovarian cancer [7]. Rsf-1 is a member of a chromatin-remodeling complex that plays a role in transcriptional regulation, cell cycle progression, and carcinogenesis [14-17]. Rsf-1 gene amplification has been reported in 13% of high grade ovarian carcinomas, and is commonly overexpressed at the RNA and protein level. Amplification and mRNA overexpression of Rsf-1 were both associated with a poor survival, but this association with outcome was not seen at the protein level [10, 18]. GAB2 is a scaffolding adaptor protein responsible for the transduction of extracellular signals through some receptor tyrosine kinases, such as Erb2 (also known as HER2). GAB2 is commonly amplified in breast cancer, and i overexpressed in breast cancer cell lines and primary tumors [5, 11, 19]. A recent study demonstrated that overexpression of GAB2 increased proliferation of MCF10A mammary cells in three 124 dimensional cultures. In addition, co-expression of GAB2 with ErB2 resulted in an invasive phenotype mediated by hyperactivation of the Shp2-Erk pathway [11]. All of these studies provide strong evidence to support these genes as important oncogenic elements and potential drivers of the 11q13 amplicon in epithelial tumors. It is possible that the emergence and maintenance of this amplification event in ovarian cancer is driven by the synergistic effect of several oncogenes, rather than by a single dominant oncogene. A detailed analysis of the amplification and co-amplification of these genes in the same cohort of ovarian cancers has never been performed. Here, we investigate the amplification of CCND1, EMSY, PAK1, Rsf-1 and GAB2 and the prognostic significance of these five genes in histologically and clinically defined subsets of early ovarian cancer. 6.2 MATERIALS AND METHODS 6.2.1 Ovarian tumor samples and tissue microarray (TMA) construction Approval for the study was obtained from the ethics committee of the University of British Columbia. The study cohort was comprised of 541 patients with moderate and high- risk ovarian cancer seen at the British Columbia Cancer Agency between 1984 and 2000. Ninety-three percent of women diagnosed with ovarian cancer in British Columbia are treated at the British Columbia Cancer agency and provincial treatment guidelines are followed. Moderate risk includes stage I, grade 2 tumors and stage II, grade 1 and 2 tumors. High risk designates those cases, which are stage I, or II, grade 3 or stage III tumors. None of these patients had residual macroscopic disease following surgery. The slides were reviewed and the tumors 125 were graded according to Silverberg grading system. A representative area of each tumor was selected and a duplicate core tissue microarray was constructed. Serial 6um sections were cut for FISH analysis. 6.2.2 Fluorescent In-situ Hybridization (FISH) The TMA sections were baked overnight at 60°C. Before hybridization, tissue array sections were deparaffinized in xylene and dehydrated in 100% ethanol. The slides were then subjected to pretreatment washes, which included immersing the slides in 0.2N HCl for 20 minutes, distilled water for 10 minutes, 2X SSC for 3 minutes, and 1M NaSCN at 80°C for 30 minutes. Following protease treatment of the slides (protease solution at 37°C for 15 minutes), they were dehydrated in 100% ethanol and air-dried. The EMSY probe was created from DNA isolated from the BAC clone CTD-250F13 and was directly labeled with Spectrum Green by nick translation (Vysis, Downer's Grove, Illinois). The EMSY probe was then combined with Spectrum Aqua labeled centromeric probe, CEP11 (Vysis) and Spectrum Orange labeled CCND1 (Vysis). The PAK1 probe was created from DNA isolated from the BAC clone CTD-2004N10 and the DNA was then directly labeled with Spectrum Orange, by nick translation (Vysis, Downer's Grove, Illinois). The GAB2 probe was created from DNA isolated from the BAC clones CTD-254M24 and RP11-452H21 and the DNA was then directly labeled with Spectrum Green, by nick translation (Vysis, Downer's Grove, Illinois). The PAK1 and GAB2 probe were then combined with Spectrum Aqua labeled centromeric probe, CEP11 (Vysis). The Rsf-1 probe was created from DNA isolated from the BAC clones RP11-1107J12 and RP11-1081L7 and the DNA was then directly labeled with Spectrum Orange. The Rsf-1 126 probe was then combined with Spectrum Aqua labeled centromeric probe, CEP11 (Vysis). The slides were co-denatured for 5 minutes at 73°C and hybridized for 18 hours at 37°C on a HyBrite (Vysis, Downers Grove, Illinois). Post hybridization washes were done according to LSI procedure (Vysis). Slides were then counterstained with DAPI. Automated analysis of FISH signals was performed using Metasystems™ automated image acquisition and analysis system, Metafer (Metasystems, Altlussheim, Germany). The average gene copy number for each probe was calculated and the amplification ratio determined. The amplification ratio was calculated as a ratio between the average copy per cell for each gene and the average copy number for centromere 11. An amplification ratio > 1.5 was considered amplified. 6.2.3 Statistical analysis Univariate survival analysis was performed using the Kaplan-Meier statistic. Multivariate survival analysis was performed using the proportional hazards model. Cell type specific gene expression comparisons were computed using the Chi Square statistic. All analyses were done using JMP version 6.0.3 (SAS Institute, Cary, N.C., U.S.A.). 1 2 7 6.3 RESULTS 6.3.1 CCND1, EMSY, PAK1, Rsf-1 and GAB2 gene amplification in ovarian cancer To determine the frequency of amplification of these five genes along 11q13, we performed three-color and dual-color FISH on 541 clinically annotated ovarian tumors from four TMA's. FISH signals from the three-color FISH assay for CCND1, E/WSYand centromere 11 were successfully scored in 294 cases. FISH signals from the three-color FISH assay for PAK1, GAB2 and centromere 11 were successfully scored in 278 cases and FISH signals from the dual-color FISH assay for Rsf-1 and centromere 11 were successfully scored in 339 cases. The remaining cases for each FISH assay could not be scored as for technical reasons (i.e. core loss or lack of FISH signals). Overall, for the entire cohort, CCND1 amplification was seen in 38/294 (13%) cases, EMSY amplification was seen in 46/294 (16%) cases, PAK1 amplification was seen in 40/277 (14%) cases, Rsf-1 amplification was seen in 40/339 (12%) cases, and GAB2 amplification was seen in 42/278 (15%) cases (Figure 5.2). CCND1, EMSY, PAK1, Rsf-1 and GAB2 amplification showed highly significant correlations with each other (p<0.0001, Table 5.1). Although each gene highly correlated with one another, amplification of each gene did occur independently of one another. For example, EMSY gene amplification occurred independently of CCND1 gene amplification in 14/46 (30%) cases. EMSY amplification also occurred independently of PAK1 amplification in 26/40 (35%) cases, of Rsf-1 128 a) CCND1 and EMSY PAK1 and GAB2 Rsf-1 Figure 5.2 Amplification at 11q13. A) An example of an ovarian cancer with amplification of CCND1 (orange), EMSY (green), PAK1 (orange), GAB2 (green) and in the last panel Rsf-1 (orange). Centromere 11 is present in normal copy number (blue). B) Relatively normal copy number of all 5 genes. 129 Table 5.1 Correlations of all five genes along 11q13 - nonparametric Kendall T test CCND1 EMSY PAK1 Rsf1 GAB2 CCND1 1.00 0.73 0.59 0.59 0.66 EMSY 0.73 1.00 0.64 0.66 0.70 PAK1 0.59 0.64 1.00 0.67 0.84 Rsf1 0.59 0.66 0.67 1.00 0.73 GAB2 0.66 0.70 0.84 0.73 1.00 gene amplification in 25/40 (38%) cases, and of GAB2 gene amplification in 29/40 (27%) cases. 6.3.2 Associat ion of CCND1, EMSY, PAK1, Rsf-1 and GAB2 gene amplification with histological subtype and clinical outcome The histological subtypes associated with amplification of all five genes are shown in Table 5.2. Amplification of CCND1, EMSY, PAK1, Rsf-1 and GAB2 were most frequently amplified in serous carcinomas. All five genes were significantly associated with a serous histology (p<0.002). In clear cell and endometrioid carcinomas, all five genes were amplified at a similarly low frequency. Amplification of CCND1, EMSY and PAK1 was also observed in a small percentage of mucinous carcinomas, whereas amplification of Rsf-1 and GAB2 was not seen in these tumors. However, it should be noted that the overall number of mucinous carcinomas in this series was very small. To determine the prognostic significance of CCND1, EMSY, PAK1, Rsf-1 and GAB2 in ovarian carcinomas, we correlated gene amplification with clinical outcome. Univariate survival analysis showed that only tumors with EMSY and Rsf-1 amplification were associated with a significantly worse outcome (p=0.02 and p=0.03) (Figure 5.3). In multivariate analysis, which included histological subtype, chemotherapy protocol, stage and EMSY amplification, only stage and histological subtype were independently significant in predicting prognosis (Table 5.3). A similar result was seen when Rsf-1 amplification was included in the multivariate analysis with the other biomarkers described above, i.e. only stage 131 Table 5.2 Evaluation of CCND1, EMSY, PAK1, Rsf-1 and GAB2 in ovarian cancer CCND1 EMSY PAK1 RsFl GAB2 amplified amplified amplified amplified amplified N n(%) N n(%) N n (%) N n (%) N n (%) Histopathological Diagnosis Serous 126 26(21%) 126 31(25%) 115 28(24%) 150 29(19%) 116 30(26%) Clear Cell 68 6 (9%) 68 7(10%) 71 6 (8%) 74 4 (5%) 71 6 (8%) Endometrioid 74 4 (5%) 74 5 (7%) 71 4 (6%) 86 4 (5%) 71 5 (7%) Mucinous 13 1 (8%) 13 1 (8%) 8 1 (13%) 15 0 (0%) 8 0 (0%) 132 A) i a a O O - i . 9 0 - | M 7 0 -6 0 — S O -. 4 0 3 C . 2 0 — . 1 0 -O O -EMSY not amplified I EMSY ampl i f i ed 8 1 0 1 2 T i m © ( Y e a r s ) i a B) g .OO «o H so n M . 6 0 4 C M 2D :c DC Rsf-1 not amplified Rsf-1 amplified n r 8 1 0 1 2 T i m e 4 Y e a r a ) i a Figure 5.3 EMSY and Rsf-f amplification correlate with a poor outcome A) Kaplan-Meier survival analysis shows that EMSY amplification is associated with a poor outcome compared to those without EMSY amplification. B) Rsf-1 amplification also correlates with a poor outcome compared to those without Rsf-1 amplification. 133 Table 5.3 Multivariate analyses DF p value Stage 2 0.002 Cell Type 4 0.004 EMSY amplification 1 0.268 Chemotherapy protocol 14 0.531 and histological subtype were of independent significance in predicting prognosis (data not shown). We next evaluated whether amplification of any of the five genes along 11q13 was associated with a poor prognosis in serous ovarian cancer. Univariate survival analysis revealed that amplification of CCND1, EMSY, PAK1, Rsf-1 or GAB2 did not significantly correlate with outcome in serous carcinomas. 6.4 DISCUSSION Amplification of 11q13 is frequently detected in numerous tumor types, including head and neck, breast, bladder and esophageal cancer [1]. In breast cancer, amplification at this locus is detected in 15% of carcinomas [3] and correlates with poor prognostic markers, such as lymph node metastases [4, 19, 20]. Initial mapping identified two separate regions at 11q13. Several candidate genes have been proposed as potential drivers of amplification of these two regions at 11q13, including CCND1 (encoding cyclin D1) and EMS1 (encoding cortactin) [3-5]. Cyclin D1, an important regulator of the cell cycle, is a well-established oncogene [19, 21]. Overexpression of cyclin D1 results in the abnormal proliferation of mammary epithelial cells and promotes tumorigenesis in transgenic mice, suggesting an important oncogenic role for cyclin D1 in breast carcinogenesis [22]. CCND1 is amplified in 15 to 20% of breast cancers [6, 23, 24], and has been found overexpressed in up to 80% of tumors [3, 25-27]. Amplification of CCND1 is associated with a poor prognosis in estrogen receptor positive breast cancers [28-30]. EMS1, encodes for cortactin, an actin binding protein involved in the 135 restructuring of the actin cytoskeleton [31]. Numerous studies have shown that CCND1 and EMS1 are commonly co-amplified in breast cancer. However, amplification of EMS1 does occur independently of CCND1 gene amplification [2]. Further detailed investigations of this region in breast cancer revealed the presence of two other discrete regions of amplification, in addition to those containing CCND1 and EMS1 [3, 5, 6]. One region of amplification, located centromeric to CCND1 and EMS1, harbors several potential oncogenes including EMSY, PAK1, Rsf-1 and GAB2. EMSY is a novel oncogene, which encodes a BRCA2-binding protein. EMSY localizes to sites of repair following DNA damage and could be a surrogate for BRCA2 loss in sporadic breast and ovarian cancer [8]. We have previously reported that EMSY is amplified in 13% of sporadic breast cancers and EMSY amplification was associated with a poor prognosis. We have also shown that EMSY is amplified in 17% of high grade ovarian carcinomas, and was significantly associated with high grade serous carcinomas [7]. A highly significant correlation between EMSY gene amplification and RNA expression has been observed in both breast and ovarian cancer [7, 12]. PAK1, a serine threonine kinase, is commonly found amplified and overexpressed in breast and ovarian cancer [5, 7, 9]. A number of important oncogenic properties have been reported for PAK1, including the regulation of anchorage independent growth, invasiveness and mitotic spindle organization [32, 33]. Rsf-1, a member of a chromatin remodeling complex, is also frequently amplified and overexpressed in ovarian cancer, and is significantly associated with a serous histology. Amplification of Rsf-1 and overexpression was associated with a poor survival in ovarian cancer [10, 18]. Rsf-1 was recently shown 136 to be upregulated in ovarian cancer effusions and a predictor of poor survival in this subset of ovarian cancer patients [34]. GAB2, a scaffolding adapter protein, is commonly amplified in breast cancer [5, 19], and a recent study provided evidence to support a role for overexpression of GAB2 in breast carcinogenesis [11]. These studies not only demonstrate that the 11q13 amplicon is frequently amplified in breast and ovarian cancer, but also provide evidence to support any one of these genes as potential drivers of this amplification event. The established view has been that an amplicon contains a single dominant oncogene that is responsible for the emergence of amplification at that locus. However, based on the number of candidate genes that have been identified on small amplicons, like that of 11q13.5, it now appears that there may be an alternative explanation. It may be that amplification of multiple oncogenes within a single amplicon act in concert to drive tumor progression. A detailed investigation of these genes along 11q13 has not been performed in the same cohort of tumor samples. Here, we provide evidence that CCND1, EMSY, PAK1, Rsf-1, and GAB2 are commonly co-amplified in ovarian cancer, in particular ovarian cancer of serous type. Amplification of CCND1, EMSY, PAK1, Rsf-1, and GAB2 was also observed in a small percentage of clear cell and endometrioid carcinomas. These findings corroborate our previous results in which we found that CCND1 and EMSY were frequently amplified in serous carcinomas, and less commonly amplified in clear cell and endometrioid carcinomas [7]. An association between serous carcinomas and Rsf-1 overexpression has been recently reported [18], and our study also confirmed 137 this independent observation. These results demonstrate that amplification of the more centromeric amplicon containing CCND1 and the amplicon containing EMSY, PAK1, Rsf-1, and GAB2 are significantly associated with serous carcinomas, and thus are likely to be involved in the pathogenesis of these tumors. Such a role for this amplicon is further supported by the correlation of EMSY and Rsf-1 amplification with a poor survival in ovarian tumors in univariate analysis, but not in multivariate analysis, which included stage, cell type, and chemotherapy protocol. In summary, these results demonstrate that the 11 q13 amplicon, which harbors numerous oncogenes including CCND1, EMSY, PAK1, Rsf-1, and GAB2, is frequently amplified in ovarian carcinomas and is specifically associated with serous carcinomas. The frequency of this amplification event, the correlations with poor outcome, and the presence of multiple oncogenes in close proximity suggest that these oncogenes may act together to initiate or promote tumor progression. Indeed, it is not difficult to hypothesize how co-amplification and co-overexpression of these five oncogenes on 11q13 could provide a selective advantage for tumor cells. Further functional experiments are required to determine the mechanisms by which these oncogenes contribute to the emergence and maintenance of the amplicon during tumor development. 138 6.5 REFERENCES 1. Schwab, M., Amplification of oncogenes in human cancer cells. Bioessays, 1998. 20(6): p. 473-9. 2. Hui, R., et al., EMS1 amplification can occur independently of CCND1 or INT-2 amplification at 11q13 and may identify different phenotypes in primary breast cancer. Oncogene, 1997. 15(13): p. 1617-23. 3. Karlseder, J . , et al., Patterns of DNA amplification at band q13 of chromosome 11 in human breast cancer. Genes Chromosomes Cancer, 1994. 9(1): p. 42-8. 4. Schuuring, E., The involvement of the chromosome 11q13 region in human malignancies: cyclin D1 and EMS1 are two new candidate oncogenes--a review. Gene, 1995. 159(1): p. 83-96. 5. Bekri, S., et al., Detailed map of a region commonly amplified at 11q13->q14 in human breast carcinoma. Cytogenet Cell Genet, 1997. 79(1-2): p. 125-31. 6. Courjal, F., et al., Mapping of DNA amplifications at 15 chromosomal localizations in 1875 breast tumors: definition of phenotypic groups. Cancer Res, 1997. 57(19): p. 4360-7. 7. Brown, L.A., et al., Amplification of EMSY, a novel oncogene on 11q13, in high grade ovarian surface epithelial carcinomas. Gynecol Oncol, 2006. 100(2): p. 264-70. 8. Hughes-Davies, L, et al., EMSY links the BRCA2 pathway to sporadic breast and ovarian cancer. Cell, 2003. 115(5): p. 523-35. 9. Schraml, P., et al., Combined array comparative genomic hybridization and tissue microarray analysis suggest PAK1 at 11q13.5-q14 as a critical oncogene target in ovarian carcinoma. Am J Pathol, 2003. 163(3): p. 985-92. 10. Shih le, M., et al., Amplification of a chromatin remodeling gene, Rsf-1/HBXAP, in ovarian carcinoma. Proc Natl Acad Sci U S A , 2005. 102(39): p. 14004-9. 11. Bentires-Alj, M., et al., A role for the scaffolding adapter GAB2 in breast cancer. Nat Med, 2006. 12(1): p. 114-21. 139 12. Rodriguez, C , et at., Amplification of the BRCA2 pathway gene EMSY in sporadic breast cancer is related to negative outcome. Clin Cancer Res, 2004. 10(17): p. 5785-91. 13. Vadlamudi, R.K., et al., Regulatable expression of p21-activated kinase-1 promotes anchorage-independent growth and abnormal organization of mitotic spindles in human epithelial breast cancer cells. J Biol Chem, 2000. 275(46): p. 36238-44. 14. Shamay, M., et al., Hepatitis B virus pXinteracts with HBXAP, a PHD finger protein to coactivate transcription. J Biol Chem, 2002. 277(12): p. 9982-8. 15. Shamay, M., O. Barak, and Y. Shaul, HBXAP, a novel PHD-fingerprotein, possesses transcription repression activity. Genomics, 2002. 79(4): p. 523-9. 16. Vignali, M., et al., ATP-dependent chromatin-remodeling complexes. Mol Cell Biol, 2000. 20(6): p. 1899-910. 17. Wolffe, A. P., Chromatin remodeling: why it is important in cancer. Oncogene, 2001. 20(24): p. 2988-90. 18. Mao, T.L., et al., Expression of Rsf-1, a chromatin-remodeling gene, in ovarian and breast carcinoma. Hum Pathol, 2006. 37(9): p. 1169-75. 19. Ormandy, C.J., et al., Cyclin D1, EMS1 and 11q13 amplification in breast cancer. Breast Cancer Res Treat, 2003. 78(3): p. 323-35. 20. Schuuring, E., et al., Amplification of genes within the chromosome 11q13 region is indicative of poor prognosis in patients with operable breast cancer. Cancer Res, 1992. 52(19): p. 5229-34. 21. Baldin, V., et al., Cyclin D1 is a nuclear protein required for cell cycle progression in G1. Genes Dev, 1993. 7(5): p. 812-21. 22. Wang, T.C., et al., Mammary hyperplasia and carcinoma in MMTV-cyclin D1 transgenic mice. Nature, 1994. 369(6482): p. 669-71. 23. Champeme, M. H., et al., 11q13 amplification in local recurrence of human primary breast cancer. Genes Chromosomes Cancer, 1995.12(2): p. 128-33 24. Lammie, G.A. and G. Peters, Chromosome 11q13 abnormalities in human cancer. Cancer Cells, 1991. 3(11): p. 413-20. 140 25. Buckley, M.F., et al., Expression and amplification of cyclin genes in human breast cancer. Oncogene, 1993. 8(8): p. 2127-33. 26. Gillett, C , et al., Amplification and overexpression of cyclin D1 in breast cancer detected by immunohistochemical staining. Cancer Res, 1994. 54(7): p. 1812-7. 27. Peters, G., et al., Chromosome 11q13 markers and D-type cyclins in breast cancer. Breast Cancer Res Treat, 1995. 33(2): p. 125-35. 28. Bieche, I., et al., Prognostic value ofCCNDI gene status in sporadic breast tumours, as determined by real-time quantitative PCR assays. Br J Cancer, 2002. 86(4): p. 580-6. 29. Michalides, R., et al., A clinicopathological study on overexpression of cyclin D1 and ofp53 in a series of 248 patients with operable breast cancer. Br J Cancer, 1996. 73(6): p. 728-34. 30. Seshadri, R., et al., Cyclin Dl amplification is not associated with reduced overall survival in primary breast cancer but may predict early relapse in patients with features of good prognosis. Clin Cancer Res, 1996. 2(7): p. 1177-84. 31. Schuuring, E., et al., The product of the EMS1 gene, amplified and overexpressed in human carcinomas, is homologous to a v-src substrate and is located in cell-substratum contact sites. Mol Cell Biol, 1993.13(5): p. 2891-98. 32. Adam, L, et al., Heregulin regulates cytoskeletal reorganization and cell migration through the p21-activated kinase-1 via phosphatidylinositol-3 kinase. J Biol Chem, 1998. 273(43): p. 28238-46. 33. Adam, L, et al., Regulation of microfilament reorganization and invasiveness of breast cancer cells by kinase dead p21-activated kinase-1. J Biol Chem, 2000. 275(16): p. 12041-50. 34. Davidson, B., et al., Expression of the chromatin remodeling factor Rsf-1 is upregulated in ovarian carcinoma effusions and predicts poor survival. Gynecol Oncol, 2006. 141 Chapter 6: EMSY O V E R E X P R E S S I O N IN PRIMARY B R E A S T EPITHELIAL C E L L S Objective: To determine whether overexpression of EMSY in primary breast cells induces a chromosomal instability phenotype Based on the Manuscript: Raouf, Afshin, Lindsay Brown, Nikoleta Vrcelj, Karen To, Winnie Kwok, David Huntsman and Connie J . Eaves. (2005). "Genomic instability of human mammary epithelial cells overexpressing a truncated form of EMSY." Journal of the National Cancer Institute 97(17): 1302-6. The candidate performed all experiments, except the cloning of the 5' fragment of EMSY and the transfection in the 184-hTert cells, which was done by Afshin Raouf. A certified cytogenetic technologist performed all of the chromosome identification and karyotype analysis. 142 6.1 INTRODUCTION The EMSY gene encodes a protein whose amino-terminal region binds within the BRCA2 exon 3 transactivation domain. The amino-terminal region of EMSY also contains separate binding domains for two chromatin remodeling proteins, HP1I3, and B69. In irradiated mouse embryonic fibroblast cells, EMSY colocalizes with y-histone 2AX in response to DNA damage [1]. Mutations of BRCA2 are rare events in sporadic breast cancer [2-5], but we have reported that the EMSY gene is amplified in 13% of these cancers and associated with a poor prognosis [1]. This finding has been confirmed in a recent study, which also showed a strong association between EMSY gene amplification and overexpression in primary breast tumors and cell lines [6]. Taken together, these findings have led to the suggestion that EMSY gene amplification and overexpression could be a surrogate for the loss of BRCA2 in sporadic breast and ovarian cancer [7-9]. Inherited mutations in the BRCA2 gene are responsible for a wide spectrum of familial cancers, including breast, ovarian and prostate [10,11]. Although the proteins encoded by the BRCA1 and BRCA2 genes do not share any sequence homology, both have been implicated as essential elements of the DNA repair pathway [12-15]. In addition to DNA repair, BRCA1 appears to be involved in other cellular functions including chromatin remodeling, transcriptional regulation and cell cycle checkpoint events [15, 16]. The function of BRCA2 has been harder to identify and has remained limited to the repair of double stranded breaks. The evidence that BRCA2 might have a role in DNA repair came from the study of murine embryonal cells deficient in BRCA2 [17]. This study showed an interaction between BRCA2 and 143 Rad51, a key player in the repair of double stranded breaks (DSBs) by homologous recombination. Furthermore, BRCA2 deficient cells exhibit sensitivity to ionizing radiation, indicative of a defect in DSB repair [17-19]. These cells also accrue chromosomal aberrations during passage in culture including multiple translocations and chromatid breaks [19-21], consistent with the accumulation of DNA damage during division. Another feature of BRCA2 deficient cells is their marked sensitivity to agents that induce DSBs, such as mitomycin C, cisplatinum, and etoposide [20-23]. Treatment of BRCA2 deficient lymphocytes with mitomycin C induces a sharp increase in the incidence of chromosomal aberrations, including triradials and quadriradials not seen in control cells [21]. These observations indicate that BRCA2 is essential in the overall preservation of genomic integrity. We therefore hypothesized that forced overexpression of EMSY might cause a similar chromosomal instability phenotype in human breast epithelial cells as seen in BRCA2 deficient cells. To test this hypothesis, we isolated an 810-bp 5' fragment from the EMSY cDNA that encodes the BRCA2 interacting domain and inserted this fragment into a lentiviral vector that already contained an internal ribosomal entry site element and the cDNA for enhanced green fluorescent protein (GFP). We then used this vector or the control lentiviral vector containing GFP only to infect log-phase 184-hTert cells, a telomerase-immortalized line of breast epithelial cells. We then examined whether EMSY overexpression in normal breast epithelial cells induced a chromosomal instability phenotype. 144 6.2 MATERIAL AND METHODS 6.2.1 Ce l l l ines a n d g rowth cond i t i ons 184 h-Tert cells were obtained from Sandra Dunn, Department of Pediatrics and Experimental Medicine, University of British Columbia, Vancouver, Canada. 184 h-Tert cells were cultured in MEBM (basal medium, no bicarbonates, Clonetics, San Diego, CA) supplemented with SingleQuots (Clonetics, San Diego, CA), neomycin (400 ug/ml Gibco, Burlington, ON), transferrin (1 ug/ml, VWR, Mississauga, ON) and isoproterenol (1.25 mg/ml, Sigma, Oakville, ON). MEBM media, SingleQuots and neomycin were filtered sterilized together. The transferrin and isoproterenol were added after the filtering process. 6.2.2 Lent iv i ra l e x p r e s s i o n vec to r An 810-bp region of the EMSY cDNA encoding the BRCA2 binding domain was amplified by PCR using oligonucleotide primers EMSY-810-F (5'-GGCGCGCCCCACCATGCCTGTTGTGTGGCC-3', forward primer) and EMSY-810-R (5'-TTAATTAATGTCTGTGTTGATGGTTTAG-3', reverse primer). The 810 bp EMSY cDNA was cloned into the PCR-2.1-Topo vector (Invitrogen, Burlington, Ontario, Canada). Afshin Raouf, at the Terry Fox Laboratory, British Columbia Cancer Agency, Vancouver, British Columbia did the cloning of EMSY and insertion into the lentiviral vector. The cloned insert was isolated by digestion with the plasmid with AscI and Pad, purified and reinserted into the multiple cloning site (MCS) of the KA391 lentiviral vector. Vesicular stomatitis virus envelope-pseudotyped Lenti-EMSY/GFP or Lenti-GFP virus was produced in 293T cells, concentrated, and tittered on HeLa cells as previously described [24]. 145 Approximately, 10 5 log phase cells were cultured overnight in six-well plates in regular growth medium and the next day, the cells were incubated for 4 hours with 500 ul of medium containing 1.2 x 107 infectious units of Lenti-EMSY/GFP or Lenti-GFP virus, after which the medium was replaced with the normal growth medium for another 48 hours. Cells displaying the top 20% of fluorescence levels were isolated using a FACSVantage flow cytometry system (Becton Dickinson, San Jose, CA) and were further propagated in six-well plates. 6.2.3 Real-time quantitative PCR Real-time PCR was carried out on an ABI 7900 Sequence Detection System (Applied Biosystems, Foster, CA) under standard conditions. The EMSY TaqMan probe sequence designed for detection of the 5' end of EMSY was ctggatctcagcgggatgaatgcaaaagaa. The forward and reverse primers used for 5' E/WSVwere 5'-tgcctgttgtgtggccaa-3' and 5'-ggacacgggcttcacgtg-3'. The EMSY TaqMan probe sequence designed for detection of the 3' end of EMSY was ctcccagcttcttcagag. The forward and reverse primers for 3' EMSY was 5'-ccccccagagccaaatgt-3' and 5'-tggaccgaacggcagaca-3'. This was done to allow the full-length (endogenous) EMSY transcripts (which contained both 5' and 3' sequences) to be measured separately from the more prevalent transgene-derived transcripts (which contained only 5' sequences but not 3' sequences). Expression levels were first normalized to rRNA (Applied Biosystems, Foster CA) and then compared to similarly normalized values measured in the GFP transduced cells. 146 6.2 .4 Ch romosome harvesting procedure and s l ide making for monolayer cel l cultures 184 h-Tert cells transduced with EMSY and control GFP cells were grown to approximately 70-75% confluency. The media was changed 24 hours prior to harvesting to help synchronize the cells. On the day of harvesting, colcemid (0.1-0.3 ug/ml, Gibco, Burlington, ON) was added to the cultures. The cultures were then incubated for 2-4 hours at 37°C in a humid 5% CO2 incubator. Trypsin (Gibco, Burlington, ON) was then added to the culture to dislodge the cells from the culture dish. The trypsin was then inactivated by the addition of a small amount of normal growth medium. The cells and medium were then transferred to a 15 ml falcon tube and centrifuged. The supernatant was then aspirated and warm hypotonic solution was added (0.075 M KCI, Gibco, Burlington, ON). The cells were incubated at 37°C for approximately 20-25 minutes. The addition of the hypotonic solution causes the cells to swell which increases chromosomal spreading during slide making. Following incubation with the hypotonic solution, the cells were spun again. The supernatant was then aspirated and fix was added to the cultures (3:1 methanohacetic acid). This final wash step in 3:1 methanohacetic acid may be repeated up to 4 times depending on the cells. To prepare metaphase slides, cells were spun down and the fix supernatant was aspirated. Fresh fix was then added to resuspend the cells. The cells were then "dropped" on to slides using a Pasteur pipette and then "aged" overnight. 147 6.2.5 Cytogenetic analysis and karyotyping After harvesting and preparation of the slides, 50 metaphases from each of the three independent EMSY/GFP-transduced and control GFP experiments were analyzed. A certified cytogenetic technologist performed all of the chromosome identification and karyotype analysis. Karyotyping was performed using CytoVison (Applied Imaging, San Jose, CA) software. For each of the 50 metaphases analyzed all structural and numerical abnormalities were noted and recorded. 6.2.6 Interphase FISH Slides were placed in 2xSSC for 30 minutes at 37°C. Following incubation, slides were dehydrated in 70%, 80% and 100% EtOH. To assess chromosome copy number changes a commercially available kit, Breast Aneusomy Multi-color Probe Set (Vysis, Downers Grove, IL) was used. This kit specifically identifies chromosome 1 (SpectrumGold,), chromosome 8 (SpectrumRed), chromosome 17 (SpectrumAqua) and chromosome 11 (SpectrumGreen). The probe was then applied to the slide. The slides and probe were co-denatured for 5 minutes at 73°C and hybridized for 18 hours at 37°C on a HyBrite (Vysis, Downers Grove, Illinois). Post hybridization washes were done according to LSI procedure (Vysis). Slides were then counterstained with DAPI. FISH signals were enumerated in approximately 300 morphologically intact and non-overlapping nuclei. 148 6.2.7 24-colorFISH 24-color FISH was performed using the 24XCyte mFISH probe kit (Metaystems, Altlussheim, Germany). This kit contains 24 different chromosome painting probes specific for the 24 different human chromosomes. Each chromosome paint probe is labeled with 5 different fluorochromes or a unique combination of them. Metaphase slides were placed in to 0.1XSSC at room temperature for 1 minute and then transferred into 2XSSC at 75°C for 30 minutes. The slides were then cooled down to 37°C and then transferred to 0.1XSSC at room temperature for 1 minute. Next, the slides were denatured in 0.7N NaOH for exactly 1 minute and then immediately placed in to 0.1XSSC at 4°C for 1 minute. The slides were then transferred to 2XSSC at 4°C for 1 minute, dehydrated in an ethanol series (30%, 50%, 70% and 100%) and then air-dried. After the slides were dried the 24XCyte mFISH probe was denatured at 75°C for 8 to 10 minutes. The probe was then applied to the slide, coverslipped, sealed with rubber cement and placed into a 37°C incubator for 2 to 4 days. Following the 2 to 4 day incubation, the slides were incubated in 1XSSC at 75°C for 5 minutes and then rinsed in 4XSSCT (4XSSC and Tween 20) for 5 minutes. Blocking reagent was then applied to the slide for 15 minutes at 37°C and afterward the slides were washed in 4XSSCT for 2 minutes. One of the fluorochromes used for labeling is Biotin and thus requires detection by a Streptavidin-Cy5. This is performed using the B-tect detection kit (Metaystems, Altlussheim, Germany) according to protocol. Following the Cy5 detection step, the slides are rinsed twice in 4XSSCT at room temperature for 3 minutes and then 149 washed in PBS at room temperature for 3 minutes. The slides are then air-dried in the dark for 30 minutes and then counterstained with DAPI. 6.2.8 Mitomycin C treatment Passage 11 EMSY/GFP-transduced cells and control GFP-transduced cells were treated with 0.4 uM mitomycin C (a pre-established non-toxic concentration, Sigma, Oakville, ON) for 16 hours and then harvested for cytogenetic analysis. Twenty metaphase cells from each experiment were examined; both structural and numerical abnormalities were noted and recorded. 6.3 RESULTS 6.3.1 184-hTert breast epithelial cells The parental 184-hTert cells had a nearly diploid karyotype (48, XX, +20, +20) (Figure 6.1) and were chromosomally stable. 6.3.2 Evaluation of EMSY mRNA levels in EMSY/GFP-transduced 184-hTert To verify overexpression of the 5' end of EMSY, passage 5 and passage 10 EMSY/GFP cells and control GFP cells were harvested for RNA and mRNA levels of EMSY were determined by real-time PCR. By passage 5, the EMSY/GFP-transduced cells from three independent transduction experiments expressed 68 to 300 fold higher levels of EMSY mRNA that contained sequences from the 5' end of the EMSY gene compared with control GFP-transduced cells (Figure 6.2). In the same EMSY/GFP-transduced cells, expression of the endogenous EMSY gene (i.e., EMSY mRNA that contained sequences from the 3' end of the gene) did not differ in 150 |$ a 9 II §* 6 7 8 9 10 11 ft M *» »» 14 15 16 17 18 13B & * - * A | 8 20 21 22 X Figure 6.1 Karyotype of the 184-hTert cells 184 hTert cells were near diploid with extra copies of chromosome 20. 151 Figure 6.2 EMSY/GFP transduced 184-hTert cells. Transcript levels in passage 5 and passage 10 EMSY/GFP transduced 184-hTert cells from all three transduction experiments were quantified by real-time PCR using TaqMan probes and primers specific for the 5' and 3' ends of EMSY. 152 expression as compared to the control GFP-transduced cells. The level of 5' EMSY transcripts in the transduced cells continued to increase with further cell passage (Figure 6.2), suggesting overgrowth of the cultures by cells that overexpressed the highest levels of the EMSY transgene. 6.3.3 Evaluation of chromosomal instability in EMSY/GFP-transduced hTert cells To investigate whether EMSY overexpression causes a chromosomal instability phenotype in human breast epithelial cells, EMSY overexpressing cells from three independent experiments were analyzed cytogenetically. The parental 184-hTert cells had a nearly diploid karyotype and were chromosomally stable. On the contrary, in three separate experiments, cells that overexpressed mRNA from the EMSY transgene rapidly accumulated structural chromosomal abnormalities. These abnormalities included deletions, translocations, marker chromosomes, chromosome fragments and dicentric chromosomes (Figure 6.4A). At passage 5, 26% of EMSY/GFP-transduced cells contained at least one structural chromosomal abnormality compared to 13% of control GFP-transduced cells (Table 6.1). The highest frequency of cells with chromosomal abnormalities (32%) was seen in the cells that expressed the highest levels of EMSY mRNA (Experiment 1). Conversely, the cells that expressed lower levels of the EMSY transgene contained fewer cells that exhibited structural chromosomal abnormalities (22% in Experiment 2 and 24% 153 A) B) ii i t If 1 " 7 1 11 .... Ii H 1) 1« » "IPJH 21 • J 2 -• H n n 5 V C) D) If lO C? IMI N i 1 2 3 4 5 III •• i*H J I M !*• In* M L M0» • • • • M 13 14 15 8 9 10 11 12 ic IT n III ,* CITh tf< Df ll IHi Dr u no ilk Hi) MM) M * #4 • MW* * 14 M • 15 • » * 1BMH • 111* « 1 2 M • 1 T M • I S M S :s^ i *» it? • X S M • ¥«»• 154 Figure 6.3 EMSY overexpressing cells accumulate structural chromosomal abnormalities. A) Karyotype of an EMSY overexpressing cell showing numerous structural abnormalities including deletions of 12p and 12q, a translocation t(7;20) and a break (chromosome 7), and monosomy of chromosome 8 and chromosome 21 (arrows). B) Mitomycin C treated EMSY overexpressing cells displaying triradial chromosomes. C) Interphase nuclei from two EMSY overexpressing cells showing a variety of aneuploid features. The nucleus on the left has an additional chromosome 1 (gold) and has lost chromosome 8 (red). The nucleus on the right shows the loss of one chromosome 17 (aqua). Two copies of chromosome 11 (two green signals) were present in both nuclei. D) 24-color FISH from two of the four clones analyzed showing unique chromosomal abnormalities. Each chromosome was painted with a whole-chromosome painting probe that was labeled with different fluorochromes or combination of fluorochrome. 155 Table 6.1. Summary of the frequency of structural and numerical abnormalities seen in EMSY overexpressing cells and control GFP-transduced cells analyzed at passages 5 and 10 after transduction. * Passage 5 Passage 10 No. of cells with structural abnormalities (%) Total no. of unique structural abnormalities No. of cells with numeric abnormalities (%) Total no. of unique numeric abnormalities(%) No. of cells with structural abnormalities (%) Total no. of unique structural abnormalities No. of cells with numeric abnormalities Total no. of unique numeric abnormalities (%) EMSY 39 (26) 55 17(11) 31 63 (42) 112 45 (30) 83 (n=150) GFP 19(13) 20 23(15) 30 22(15) 25 26(17) 47 (n=150) p valuet 0.003 <0.001 >0.05 >0.05 O.001 <0.001 0.002 <0.001 *Pooled results from three independent experiments; 50 metaphase cells analyzed per group per passage in each experiment. t p values shown were calculated using a chi-square test. P>0.05 was considered not statistically significant in Experiment 3). By passage 10 the frequency of structural chromosomal abnormalities in the EMSY overexpressing cells had increased further to 44%, 48% and 34% in Experiment 1, 2 and 3 but remained low (18%, 12% and 14%) in the corresponding control cells (Table 6.1). EMSY/GFP-transduced cells also had a statistically significantly higher frequency of aneuploid or polyploid cells than control GFP-transduced cells at passage 10 (Table 6.1), but not at passage 5. We also analyzed EMSY/GFP and control GFP interphase cells for changes at four loci using a commercially available kit (Breast Aneusomy Multicolor Probe set; Vysis, Downers Grove, IL). This analysis showed that the EMSY overexpressing cells exhibited a statistically significantly higher frequency of aneuploid cells than the control GFP cells at passage 5 (76 aneuploid cells per 300 cells analyzed versus 33 aneuploid cells per 300 cells analyzed, p<0.0001) and at passage 10 (88 aneuploid cells per 300 cells analyzed versus 34 aneuploid cells per 300 cells analyzed, p<0.0001). Examples of these aneuploid cells are shown in Figure 6.4C. We also isolated four clones from EMSY/GFP-transduced cells at passage 3 from Experiment 1, confirmed their separate origins by integration site analysis of the EMSY transgene and then analyzed 15 metaphases from each clone using 24-color FISH. All four clones contained cells that had a variety of complex structural and numerical chromosomal abnormalities. These complex karyotypes ranged from tetraploid to near diploid, with multiple deletions and translocations. Examples of two complex karyotypes from two of the clones are shown in Figure 6.4D. 157 6.3.4 Examination of whether treatment of EMSY overexpressing cells with the DNA damaging agent, mitomycin C, enhances the chromosomal instability phenotype A hallmark of BRCA2-deficient cells is their increased sensitivity to DNA crosslinking agents such as mitomycin C, which causes cells to accumulate structural chromosomal abnormalities because of their decreased ability to effectively repair double-stranded breaks in the DNA [21, 25, 26]. We have established that overexpression of EMSY in human breast epithelial cells induces a chromosomal instability phenotype and we were next interested in determining whether treatment of EMSY overexpressing cells with mitomycin C would similarly enhance this phenotype. EMSY/GFP-transduced cells at passage 11 from two separate experiments were exposed to 0.4 uM mitomycin C. EMSY overexpressing cells treated with mitomycin C exhibited a statistically significantly higher frequency of chromosome breaks than the control GFP-transduced cells (29 versus 9 affected cells out of 40 cells analyzed in each group; p<0.001). Two examples of abnormalities seen in the mitomycin C treated EMSY overexpressing cells are shown in Figure 6.4B. By contrast, in the absence of mitomycin C treatment, the frequency of cells with chromosome breaks was very low in both the EMSY overexpressing cells and control cells (2 of 40 EMSY overexpressing cells affected and 0 of 40 control cells affected). 158 6.4 DISCUSSION EMSY is a novel gene encoding a protein that has been shown to interact with the transactivation domain of exon 3 of BRCA2. Immunofluorescence studies showed that EMSY is nuclear, and treatment with ionizing radiation caused a substantial increase in the number of cells with EMSY-containing nuclear speckles. Indeed, EMSY was found to colocalize with v-histone 2AX in response to DNA damage, suggesting that EMSY may play some role in DNA repair [1]. EMSY is amplified in 13% of sporadic breast cancers, and amplification correlated with a significantly poor prognosis [1]. EMSY amplification was also observed in 17% of high grade ovarian cancers [27]. These preliminary findings have led to the hypothesis that EMSY amplification and overexpression may be an alternative method leading to loss of BRCA2 function in sporadic cancer [7-9]. It may be that sporadic breast cancers with EMSY amplification and inherited BRCA2 mutation carriers share common biological and pathological features that would reflect a common defect in BRCA2 function BRCA2 has been implicated as an important element of the DNA repair pathway [17-19]. It has been shown to interact with RAD51, a homolog of RecA essential in bacteria for the repair of DNA breaks via recombination [17, 28, 29]. Consistent with a role in DNA repair, BRCA2 deficient cells are sensitive to treatment with DNA damaging agents such as mitomycin C, ionizing radiation, and methyl methanesulfonate [18-22, 28]. Another striking feature of BRCA2 deficient cells is that they spontaneously accumulate chromosomal aberrations, including chromatid and chromosome breaks, as well as tri-radial and quadriradial chromosomes [19]. Furthermore, BRCA2 deficient cells exhibited a greater number of chromosomal 159 aberrations when treated with mitomycin C. These studies implicate a role for B R C A 2 in the maintenance of genomic stability. The findings presented here establish that cells overexpressing elevated levels of the 5' end of EMSY display a genomic instability phenotype. Cells overexpressing the EMSY transgene rapidly accumulated structural chromosomal abnormalities, including deletions, translocations, and chromosome breaks. The frequency of these structural abnormalities in the EMSY-overexpressing cells increased during passage in culture. Another feature of these EMSY overexpressing cells was an increased frequency of aneuploid cells. Furthermore, treatment with the DNA damaging agent mitomycin C produced a several fold higher frequency of chromosome breaks in the EMSY overexpressing cells than in the control cells. Thus, overexpression of a 5 ' fragment of EMSY induces a chromosomal instability phenotype in human breast epithelial cells that is similar to that of B R C A 2 deficient cells. This data provides new experimental evidence to support the hypothesis that elevated levels of EMSY may a play a role in sporadic breast carcinogenesis by deregulating genomic stability. How EMSY overexpression is producing such a phenotype is still unknown. It may be that EMSY overexpression results in the impairment of DNA DSBs by interfering with the normal role of B R C A 2 . Another possibility is that overexpression of the truncated EMSY cDNA might produce a B R C A 2 - n u l l phenotype through a dominant negative mechanism. Future experiments with a full length EMSY protein are needed in order to resolve these possibilities. 1 6 0 In summary, these results demonstrate that overexpression of a 5' fragment of EMSY is sufficient to induce a chromosomal instability phenotype in human breast epithelial cells that is similar to that of BRCA2 deficient cells. Moreover, treatment of the EMSY overexpressing cells with mitomycin C enhanced this chromosomal instability phenotype. It will be of interest to determine whether EMSY overexpressing cells, like BRCA2 null cells, have enhanced sensitivity to DNA damaging agents, other than mitomycin C. 161 6.5 REFERENCES 1. Hughes-Davies, L , et al., EMSY links the BRCA2 pathway to sporadic breast and ovarian cancer. Cell, 2003. 115(5): p. 523-35. 2. Khoo, U.S., et al., Somatic mutations in the BRCA1 gene in Chinese sporadic breast and ovarian cancer. Oncogene, 1999.18(32): p. 4643-6. 3. Lancaster, J.M., et al., BRCA2 mutations in primary breast and ovarian cancers. Nat Genet, 1996. 13(2): p. 238-40. 4. Merajver, S.D., et al., Germline BRCA1 mutations and loss of the wild-type allele in tumors from families with early onset breast and ovarian cancer. Clin Cancer Res, 1995.1(5): p. 539-44. 5. Miki, Y., et al., Mutation analysis in the BRCA2 gene in primary breast cancers. Nat Genet, 1996. 13(2): p. 245-7. 6. Rodriguez, C , et al., Amplification of the BRCA2 pathway gene EMSY in sporadic breast cancer is related to negative outcome. Clin Cancer Res, 2004. 10(17): p. 5785-91. 7. King, M.C., A novel BRCA2-binding protein and breast and ovarian tumorigenesis. N Engl J Med, 2004. 350(12): p. 1252-3. 8. Livingston, D.M., EMSY, a BRCA-2 partner in crime. Nat Med, 2004. 10(2): p. 127-8. 9. Turner, N., A. Tutt, and A. Ashworth, Hallmarks of 'BRCAness' in sporadic cancers. Nat Rev Cancer, 2004. 4(10): p. 814-9. 10. Teng, D. H., et al., Low incidence of BRCA2 mutations in breast carcinoma and other cancers. Nat Genet, 1996.13(2): p. 241-4. 11. Wooster, R., et al., Identification of the breast cancer susceptibility gene BRCA2. Nature, 1995. 378(6559): p. 789-92. 12. Gowen, L.C, et al., BRCA1 required for transcription-coupled repair of oxidative DNA damage. Science, 1998. 281(5379): p. 1009-12. 13. Scully, R., et al., Association ofBRCAl with Rad51 in mitotic and meiotic cells. Cell, 1997. 88(2): p. 265-75. 14. Venkitaraman, A.R., Functions ofBRCAl and BRCA2 in the biological response to DNA damage. J Cell Sci, 2001. 114(Pt20): p. 3591-8. 15. Venkitaraman, A. R., Cancer susceptibility and the functions of BRCA 1 and BRCA2. Cell, 2002. 108(2): p. 171-82. 162 16. Zheng, L, et al., Lessons learned from BRCA1 and BRCA2. Oncogene, 2000. 19(53): p. 6159-75. 17. Sharan, S.K., et al., Embryonic lethality and radiation hypersensitivity mediated by Rad51 in mice lacking Brcal. Nature, 1997. 386(6627): p. 804-10. 18. Connor, F., et al., Tumorigenesis and a DNA repair defect in mice with a truncating Brca2 mutation. Nat Genet, 1997. 17(4): p. 423-30. 19. Patel, K.J., et al., Involvement ofBrca2 in DNA repair. Mol Cell, 1998. 1(3): p. 347-57. 20. Kraakman-van der Zwet, M., et al., Brca2 (XRCC11) deficiency results in radioresistant DNA synthesis and a higher frequency of spontaneous deletions. Mol Cell Biol, 2002. 22(2): p. 669-79. 21. Yu, V. P., et al., Gross chromosomal rearrangements and genetic exchange between nonhomologous chromosomes following BRCA2 inactivation. Genes Dev, 2000. 14(11): p. 1400-6. 22. Abbott, D.W., M.L. Freeman, and J.T. Holt, Double-strand break repair deficiency and radiation sensitivity in BRCA2 mutant cancer cells. J Natl Cancer Inst, 1998. 90(13): p. 978-85. 23. Yuan, S.S., et al., BRCA2 is required for ionizing radiation-induced assembly ofRad51 complex in vivo. Cancer Res, 1999. 59(15): p. 3547-51. 24. Imren, S., et al., High-level beta-globin expression and preferred intragenic integration after lentiviral transduction of human cord blood stem cells. J Clin Invest, 2004. 114(7): p. 953-62. 25. Donoho, G., et al., Deletion ofBrca2 exon 27 causes hypersensitivity to DNA crosslinks, chromosomal instability, and reduced life span in mice. Genes Chromosomes Cancer, 2003. 36(4): p. 317-31. 26. Tutt, A., et al., Mutation in Brca2 stimulates error-prone homology-directed repair of DNA double-strand breaks occurring between repeated sequences. Embo J , 2001. 20(17): p. 4704-16. 27. Brown, L.A., et al., Amplification of EMSY, a novel oncogene on 11q13, in high grade ovarian surface epithelial carcinomas. Gynecol Oncol, 2006. 100(2): p. 264-70. 28. Chen, P.L., et al., The BRC repeats in BRCA2 are critical forRAD51 binding and resistance to methyl methanesulfonate treatment. Proc Natl Acad Sci U S A, 1998. 95(9): p. 5287-92. 163 29. Wong, A.K., et al., RAD51 interacts with the evolutionarily conserved BRC motifs in the human breast cancer susceptibility gene brcal. J Biol Chem, 1997. 272(51): p. 31941-4. 164 C h a p t e r 7: F U R T H E R FUNCTIONAL A N A L Y S I S O F E M S Y O V E R E X P R E S I O N IN B R E A S T A N D OVARIAN S U R F A C E EPITHELIAL C E L L S Objective: To determine whether overexpression of EMSY in primary ovarian surface epithelial cells induces a chromosomal instability phenotype, and whether EMSY overexpressing primary breast and ovarian surface epithelial cells share similar biological features associated with BRCA2 deficient cells. Based on: The unpublished data from experiments performed on EMSY overexpressing breast and ovarian surface epithelial cells. The candidate performed all experiments, except for the cloning of the 5' fragment of EMSY, which was done by Afshin Raouf. A certified cytogenetic technologist performed all of the chromosome identification and karyotype analysis. 165 7.1 INTRODUCTION The EMSY gene encodes a protein whose amino-terminal region binds within the BRCA2 exon 3 transactivation domain. EMSY is commonly amplified in breast and ovarian cancer, suggesting that it is an important element of the 11q13 amplicon. In both breast and ovarian cancer, patients with EMSY amplification are associated with a poor prognosis. Also, amplification of EMSY leads to overexpression, further supporting its role in oncogenesis. These findings suggest that EMSY is an important oncogene on the 11 q13 amplicon. These findings have led to the suggestion that EMSY gene amplification and overexpression could be a surrogate for the loss of BRCA2 in sporadic breast and ovarian cancers [1-3]. Therefore, we hypothesized that overexpression of EMSY might cause a similar chromosomal instability phenotype in human breast epithelial cells as seen in BRCA2 deficient cells. We found that cells overexpressing elevated levels of the 5' end of EMSY display a genomic instability phenotype. Cells overexpressing the EMSY transgene rapidly accumulated structural chromosomal abnormalities, including deletions, translocations, and chromosome breaks. The frequency of these structural abnormalities in the EMSY-overexpressing cells increased during passage in culture. Furthermore, treatment with the DNA damaging agent mitomycin C produced a several fold higher frequency of chromosome breaks in the EMSY overexpressing cells than in the control cells. This data provides new experimental evidence to support the hypothesis that elevated levels of EMSY may a play a role in sporadic breast carcinogenesis by deregulating genomic stability. 166 However after performing this experiment, several key questions emerged. Is this chromosomal instability phenotype cell line specific, or can we reproduce this effect in another immortalized normal tissue that is associated with BRCA2 mutation carriers, such as normal ovarian surface epithelial cells. To determine whether we could induce a chromosomal instability in another cell line the same lentiviral vector containing the 810-bp 5' fragment of EMSY cDNA that was used to infect 184-hTert cells, was used to infect two telomerase-immortalized ovarian surface epithelial (IOSE) cell lines, IOSE C9 and IOSE C21. Secondly, we wanted to determine whether EMSY overexpressing cells have enhanced sensitivity to DNA damaging agents, other than mitomycin C. To test this EMSY overexpressing cells were treated with the DNA damaging agents, doxorubicin and cisplatin. Lastly, we wanted to determine whether the overexpression of the 5' fragment of EMSY induces a chromosomal instability phenotype through a dominant-negative mechanism. The crystal structure of the ENT domain has recently been described and one finding from this study was that it forms a homodimer by the anti-parallel packing of the long N-terminal a-helix from each subunit. The structure of the ENT domain suggests that the EMSY protein exists as a dimer, which could allow for it to bind simultaneously to several partner proteins [4]. It may be that the 5' truncated EMSY is binding to the endogenous EMSY in the cell resulting in a dominant negative effect and inducing chromosomal instability. To investigate whether the chromosomal instability phenotype is due to a dominant negative mechanism, we treated the 184-hTert breast epithelial cells with a small interfering RNA (siRNA) targeting EMSY (the sequence was given to us from Luke Hughes-Davies) and 18 167 hours before harvesting the cells were treated with mitomycin C. The cells were then analyzed cytogenetically. 7.2 MATERIAL AND METHODS 7.2.1 Cell lines and growth conditions IOSE C9 and IOSE C21 were obtained from Frances Balkwill, Cancer Research UK, Translational Oncology Lab, Barts and the London Queen Mary's School of Medicine and Dentistry, London, UK. IOSE cells were cultured in a 1:1 mixture of MCDB105 medium and Medium 199 HEPES modification (Sigma, Oakville, ON) supplemented with 15% fetal bovine serum (FBS), bovine pituitary extract (34 ug protein/ml, Invitrogen, Burlington, ON), insulin (5 ug/ml, Sigma, Oakville, ON), human epidermal growth factor (10 ng/ml, Invitrogen, Burlington, ON) and hydrocortisone (0.5 ug/ml, Sigma, Oakville, ON). 7.2.2 Lentiviral expression vector An 810-bp region of the EMSY cDNA encoding the BRCA2 binding domain was amplified by PCR using oligonucleotide primers EMSY-810-F (5'-GGCGCGCCCCACCATGCCTGTTGTGTGGCC-3', forward primer) and EMSY-810-R (5'-TTAATTAATGTCTGTGTTGATGGTTTAG-3', reverse primer). The 810 bp EMSY cDNA was cloned into the PCR-2.1-Topo vector (Invitrogen, Burlington, Ontario, Canada). Afshin Raouf, at the Terry Fox Laboratory, British Columbia Cancer Agency, Vancouver, British Columbia did the cloning of EMSY and insertion into the lentiviral vector. The cloned insert was isolated by digestion with the 168 plasmid with AscI and Pad, purified and reinserted into the multiple cloning site (MCS) of the KA391 lentiviral vector. Vesicular stomatitis virus envelope-pseudotyped Lenti-EMSY/GFP or Lenti-GFP virus was produced in 293T cells, concentrated, and tittered on HeLa cells as previously described [5]. Approximately, 105 log phase cells were cultured overnight in six-well plates in regular growth medium and the next day, the cells were incubated for 4 hours with 500 ul of medium containing 1.2 x 107 infectious units of Lenti-EMSY/GFP or Lenti-GFP virus, after which the medium was replaced with the normal growth medium for another 48 hours. Cells displaying the top 20% of fluorescence levels were isolated using a FACSVantage flow cytometry system (Becton Dickinson, San Jose, CA) and were further propagated in six-well plates. 7.2.3 Real-time quantitative PCR Real-time PCR was carried out on an ABI 7900 Sequence Detection System (Applied Biosystems, Foster, CA) under standard conditions. The EMSY TaqMan probe sequence designed for detection of the 5' end of EMSY was ctggatctcagcgggatgaatgcaaaagaa. The forward and reverse primers used for 5' EMSY were 5'-tgcctgttgtgtggccaa-3' and 5'-ggacacgggcttcacgtg-3'. The EMSY TaqMan probe sequence designed for detection of the 3' end of EMSY was ctcccagcttcttcagag. The forward and reverse primers for 3' EMSY was 5'-ccccccagagccaaatgt-3' and 5'-tggaccgaacggcagaca-3'. This was done to allow the full-length (endogenous) EMSY transcripts (which contained both 5' and 3' sequences) to be measured separately from the more prevalent transgene-derived 169 transcripts (which contained only 5' sequences but not 3' sequences). Expression levels were first normalized to rRNA (Applied Biosystems, Foster CA) and then compared to similarly normalized values measured in the GFP transduced cells. 7.2.4 Chromosome harvesting procedure and slide making for monolayer culture The IOSE C9 and IOSE C21 cells transduced with EMSY and control GFP cells were grown to approximately 70-75% confluency. The media was changed 24 hours prior to harvesting to help synchronize the cells. On the day of harvesting, colcemid (0.1-0.3 ug/ml, Gibco, Burlington, ON) was added to the cultures. The cultures were then incubated for 2-4 hours at 37°C in a humid 5% C 0 2 incubator. Trypsin (Gibco, Burlington, ON) was then added to the culture to dislodge the cells from the culture dish. The trypsin was then inactivated by the addition of a small amount of normal growth medium. The cells and medium were then transferred to a 15 ml falcon tube and centrifuged. The supernatant was then aspirated and warm hypotonic solution was added (0.075 M KCI, Gibco, Burlington, ON). The cells were incubated at 37°C for approximately 20-25 minutes. The addition of the hypotonic solution causes the cells to swell which increases chromosomal spreading during slide making. Following incubation with the hypotonic solution, the cells were spun again. The supernatant was then aspirated and fix was added to the cultures (3:1 methanol:acetic acid). This final wash step in 3:1 methanol:acetic acid may be repeated up to 4 times depending on the cells. To prepare metaphase slides, cells were spun down and the fix supernatant was aspirated. Fresh fix was then added to 170 resuspend the cells. The cells were then "dropped" on to slides using a Pasteur pipette and then "aged" overnight. 7.2.5 Cytogenetic analysis and karyotyping After harvesting and preparation of the slides, 50 metaphases from each of the three independent EMSY/GFP-transduced and control GFP experiments were analyzed. A certified cytogenetic technologist performed all of the chromosome identification and karyotype analysis. Karyotyping was performed using CytoVison (Applied Imaging, San Jose, CA) software. For each of the 50 metaphases analyzed all structural and numerical abnormalities were noted and recorded. 7.2.6 Mitomycin C treatment Passage 5 EMSY/GFP-transduced cells and control GFP-transduced cells were treated with 0.4 uM mitomycin C (a pre-established non-toxic concentration, Sigma, Oakville, ON) for 16 hours and then harvested for cytogenetic analysis. Twenty metaphase cells from each experiment were examined; both structural and numerical abnormalities were noted and recorded. 7.2.7 MTT cytotoxicity assays Adherent cells were plated one day prior to the addition of drugs to allow the cells to adhere to the plate. Cells were counted using a hemocytometer. Approximately 35,000 cells were plated per well of a 48 well plate. All drug concentrations were performed in quadruplicate and six different drug concentrations were tested. After 171 24 hours, 250 ul of desired drug concentration (including an untreated control) was added to each well. The plates were then returned to the incubator for the duration of the incubation period (24, 48 and 72 hours). At the end of the incubation period, 25 ul of MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide) (5 mg MTT/ ml phosphate buffered saline (PBS) filtered with a 0.22 uM syringe, Sigma, Oakville, ON) was added to each well. The cells were then returned to the incubator for an additional 4 hours. After incubation the media was aspirated and 300 ul of dimethyl sulfoxide (DMSO) was added per well. The cells were then returned to the incubator for several minutes to dissolve the crystals. The absorbance of each well was read, including a blank control sample of DMSO only, at 570 nm. The average value for each drug concentration and untreated control sample were then calculated and subtracted from the average value of the blank. The percent control was a ratio of the average absorbance value for each drug versus the average absorbance value of the untreated control. 7.2.8 EMSY siRNA and transfection The EMSY siRNA sequence (graciously given to us by Luke Hughes-Davies) used was AAGCAGCAUAGCAACGGUUAA and was synthesized by IDT (Coralville, IA). As a control, Nonsilencing Control siRNA was used (Qiagen, Mississauga, Ontario). Approximately, 106 log phase cells were cultured overnight in 10 cm plates in regular growth medium and the next day the cells were incubated overnight with 5 ml of serum free medium containing a mixture of siRNA (100 nM final concentration) and the transfection reagent, SiLentFect (Biorad, Hercules, CA), after which normal 172 media was replaced the following morning. Untreated cells, and cells treated only with the transfection reagent were included as additional controls. 7.2.9 Western blot analysis Cells either untreated or treated with different experiment reagents were washed with PBS. The PBS was then removed and a mixture of protease inhibitor and lysis buffer was then added to the cells. The lysis buffer consisted of 25 ml of 1M Tris (pH 8), 33.3 ml of 5M NaCl, 5ml of 100% NP40, 2.4g of deoxycholic acid, 0.5g of SDS and water up to 500 ml. The cells were placed on a shaker at 4°C for 10 minutes. The cells lysates were then collected by scraping and centrifuged for 15 minutes at 13 000 rpm at 4°C. Protein concentration was determined by the Bradford assay. Twenty ug of extracted protein was fractionated using a 10% SDS-polyacrylamide gel, electrophoretically transferred to an Immobilon-P-Membrane (Millipore, Billerica, MA) and blocked with PBS containing 0.1% Tween-20 and 5% nonfat dry milk for one hour at room temperature. Afterward, the membrane was incubated with a specific primary antibody overnight at 4°C, followed by a secondary antibody for 45 minutes at room temperature. The immunoblots were visualized with the chemiluminescence detection system, SuperSignal according to the manufacturers protocol (Pierce, Rockford, IL). The EMSY goat polyclonal antibody was purchased from Abeam (Cambridge, UK). The mouse monoclonal p-actin antibody was purchased from Sigma (Oakville, ON), and the secondary HRP conjugated anti-goat and anti-mouse were purchased from Jackson Laboratories (Bar Harbor, ME). 173 7.3 RESULTS 7.3.1 IOSE cells The parental IOSE C9 and IOSE C21 cells were diploid (46, XX) and were chromosomally stable (Figure 7.1). 7.3.2 Evaluation of EMSY mRNA levels in EMSY/GFP-transduced IOSE C9 and IOSE C21 cells To verify overexpression of the 5' end of EMSY in the IOSE C9 and IOSE C21 EMSY overexpressing cells real-time PCR was performed. At passage 5, C9 EMSY/GFP-transduced IOSE cells expressed 67 fold higher levels of 5' EMSY mRNA compared with control GFP-transduced cells (Figure 7.2). Similarly, C21 EMSY/GFP-transduced cells expressed 64 fold higher levels of 5' EMSY mRNA compared with control GFP-transduced cells. For both C9 and C21 EMSY overexpressing cells, expression of the endogenous EMSY gene did not differ in expression as compared to the control C9/GFP and C21/GFP IOSE cells. The level of 5' EMSY transcripts in the C9/EMSY-transduced cells continued to increase with further cell passage, however the levels of 5' EMSY mRNA in C21/EMSY-transduced cells decreased (Figure 7.2). 174 ft - M . 1 y . ft M .4 7 , . -v. i e 10 8 i 11 * • # 4 • I 12 13 14 15 16 Mi I • 17 u 9 18 1" - n 20 21 i 22 $3 Figure 7.1 Karyotype of the IOSE cells. IOSE C9 and IOSE C21 were both diploid. This is a karyotype of the IOSE C9 cells. 175 EMSY 5" EMSY 3' I0SEC9 I0SEC21 IOSE C9 IOSE C21 i ; i / Passage 5 T Passage 10 Figure 7.2 EMSY/GFP-transduced IOSE cells. Transcript levels in passage 5 and passage 10 EMSY/GFP transduced IOSE C9 and IOSE C21 cells were quantified by real-time PCR using TaqMan probes and primers specific for the 5' and 3' ends of EMSY. 176 7.3.3 Evaluation of chromosomal instability in EMSY/GFP-transduced IOSE C9 arid IOSE C21 To investigate whether EMSY overexpression also causes a chromosomal instability phenotype in ovarian surface epithelial cells, EMSY overexpressing cells from two different IOSE cell lines were analyzed cytogenetically. The parental IOSE C9 and IOSE C21 were both diploid and chromosomally stable. Unlike the EMSY/GFP overexpressing breast epithelial cells, EMSY overexpression did not induce a chromosomal phenotype in ovarian surface epithelial cells. Passage 5 C9 EMSY/GFP and C21 EMSY/GFP-transduced cells, although overexpressing the EMSY transgene, were both karyotypically normal. This was the same for metaphase cells analyzed at passages 10, 15 and 20. The control GFP-transduced cells from IOSE C9 and C21 were also karyotypically normal at passages 5, 10, 15 and 20. 7.3.4 Examination of whether treatment of EMSY overexpressing IOSE cells with mitomycin C enhances the chromosomal instability phenotype EMSY overexpression in two different IOSE cell lines did not induce a chromosomal instability phenotype. We proposed that treatment with a DNA damaging agent such as mitomycin C would impair the cells ability to effectively repair damaged DNA and lead to chromosomal instability. To test this hypothesis, passage 5 C9 EMSY/GFP, C21 EMSY/GFP and control GFP-transduced cells were treated with 0.4 uM mitomycin C. EMSY overexpressing IOSE C9 and C21 cells did not exhibit any chromosomal abnormalities and were karyotypically normal. The control GFP-transduced cells were also karyotypically normal. Thus, in addition to forced 177 overexpression of EMSY in the IOSE cells further treatment with a DNA damaging agent did not mimic the chromosomal instability phenotype seen in the EMSY overexpressing breast epithelial cells. As these EMSY overexpressing IOSE cells did not exhibit a chromosomal instability phenotype all further experiments performed did not include these cells. 7.3.5 Investigation of whether EMSY overexpressing breast epithelial cells are sensitive to treatment with DNA damaging agents Another feature of BRCA2 deficient cells is their marked sensitivity to DNA damaging agents such as cisplatin [6-9]. We decided to test whether the EMSY/GFP-transduced 184-hTert cells were more sensitive to DNA damaging agents. As additional controls, the GFP-transduced 184-hTerts and wild type 184-hTerts were also tested for sensitivity to chemotherapeutic agents. Pharmacologic agents were chosen based on their ability to induce DSBs and for their use in clinical chemotherapy. Doxorubicin binds DNA directly and intercalates between the base pairs of DNA. It is also an inhibitor of DNA topoisomerase II, thereby inhibiting DNA synthesis. Cisplatin, another DNA damaging agent, causes DNA cross-linkages, disrupting DNA function and ultimately preventing DNA, RNA and protein synthesis. Paclitaxel was used as a control treatment as it is an anti-microtubule agent that prevents the assembly and stabilization of microtubles from tubulin dimers. This agent does not cause DNA DSBs. At 24 hours, the EMSY/GFP-transduced cells did not exhibit any sensitivity to treatment with doxorubicin at varying concentrations. A similar effect was seen with 178 the control GFP-transduced cells, and the wild type 184 hTerts (Figure 7.3A). At 72 hours, all three cells lines were equally sensitive to treatment with high concentrations of doxorubicin (Figure 7.3B). This experiment was repeated twice, and both times a similar result was observed. For cisplatin treatment, three time points were chosen and this experiment was performed in triplicate. After 24 hours, the EMSY/GFP-transduced cells were not sensitive to treatment with cisplatin at varying concentrations (Figure 7.4A). A similar result was also observed for the control cells. At 48 hours, the EMSY/GFP-transduced cells exhibited a dose dependent sensitivity to cisplatin (Figure 7.4B). The same trend was also observed in the control cells. After treatment for 72 hours, a similar dose dependent sensitivity to cisplatin was seen for all three cell lines (Figure 7.4C). At the highest concentration very few cells were alive after 72 hours of treatment. Treatment with paclitaxel yielded similar results at 24 and 72 hours (Figure 7.5A). All three cells lines were not sensitive to treatment with paclitaxel. However, at the higher concentrations after 72 hours, the EMSY/GFP transduced, GFP transduced cells and the wild type hTerts did exhibit a slight sensitivity to paclitaxel as compared to the lower concentrations (Figure 7.5B). 179 A) Doxorubicin 24 hours 120 Concentration (uM) B) Doxorubicin 72 hours 120 0 0.1 0.5 1 1.5 2 Concentration (uM) Figure 7.3 Doxorubicin treatment of EMSY overexpressing cells. EMSY overexpressing cells, along with the control wild type 184 hTert, and GFP/transduced cell were treated with varying concentrations of drug and assayed at 24 and 72 hours. 180 A) Cisplatin 24 hours 120 100 2 80 c o ^ 60 c «i a £ 40 20 B) 120 100 1 80 c o 2 60 o 40 a 20 0 5 10 20 Concentration (uM) Cisplatin 48 hours 0 5 10 Concentration (uM) I 20 1184 hTert I EMSY GFP • 184 hTert • EMSY G F P C) 120 100 2 80 o o 60 a 40 a 20 0 Cisplatin 72 hours 5 10 Concentration (uM) 20 • 184 hTert • EMSY GFP Figure 7.4 Cisplatin treatment of EMSY overexpressing cells. EMSY overexpressing cells, along with the control wild type 184 hTert, and GFP/transduced cell were treated with varying concentrations of drug and assayed at 24, 48 and 72 hours. 0 1 5 25 50 100 Concentration (nM) Figure 7.5 Paclitaxel treatment of EMSY overexpressing cells. EMSY overexpressing cells, along with the control wild type 184 hTert, and GFP/transduced cell were treated with varying concentrations of drug and assayed at 24 and 72 hours. 182 7.3.6 Investigation of whether overexpression of the 5' fragment of EMSY induces a chromosomal instability phenotype through a dominant negative mechanism To investigate whether the chromosomal instability phenotype is due to a dominant negative mechanism, we treated the 184-hTert breast epithelial cells with a siRNA targeting EMSY (the sequence was given to us from Luke Hughes-Davies) and 18 hours before harvesting the cells were treated with mitomycin C. We found that EMSY protein was effectively knocked down at 24 hours, and after 48 hours there was still a slight knockdown of EMSY expression (Figure 7.6). No knockdown of EMSY was seen in the untreated 184-hTert cells, the control siRNA as well as in the cells only treated with transfection reagent. The 24 hour and 48 hour cells were then harvested and analyzed cytogenetically. After 24 hours and treatment with mitomycin C, the EMSY siRNA treated cells accumulated structural chromosomal abnormalities. The EMSY siRNA treated cells accumulated almost double the amount of structural chromosomal aberrations as compared to the control cells (Table 7.1). After 48 hours, when the EMSY protein was not effectively knocked down the number of chromosomal aberrations seen in the EMSY siRNA treated cells were similar to the control cells. 183 24 hours 48 hours <—A—^ r—A—^ l i t I I f 8 > g « 8 >• e t D U J O W D U J O CT) EMSY (J-Actin Figure 7.6 Treatment of 184-hTert with EMSY siRNA At 24 hours EMSY protein was effectively knocked down after treatment with siRNA, and no knockdown was seen in the controls (untreated, control siRNA and cells treated with the transfection reagent, SiLentFect, only. EMSY protein expression is still slightly downregulated after 48 hours. 184 Table 7.1 184-hTert breast epithelial cells treated with EMSY siRNA accumulate structural chromosomal abnormalities* 24 hours 48 hours Total #of Total # of structural structural abnormalities abnormalities Untreated 24 19 SilentFect only 14 33 Control SiRNA 25 46 EMSY siRNA 49 57 *20 metaphase cells were analyzed for each treatment 7.4 DISCUSSION EMSY is a novel gene encoding a protein that interacts with BRCA2. EMSY is commonly amplified in both breast and ovarian cancers and appears to be an important oncogene on the 11q13 amplicon. It has been suggested that EMSY amplification may be a surrogate for BRCA2 loss in sporadic breast and ovarian cancer. Thus, sporadic cancers with EMSY amplification may share common biological features with familial breast cancers associated with loss of BRCA2 that would reflect a common defect in BRCA2 function. BRCA2 deficient cells sustain spontaneous chromosomal aberrations that accumulate during passage in culture [7, 10, 11]. We have shown that overexpression of a 5' truncated EMSY induces a chromosomal instability phenotype in normal breast epithelial cells. The frequency of these structural abnormalities in the EMSY-overexpressing cells increased during passage in culture, and these cells also showed a marked sensitivity to mitomycin C, another feature of BRCA2 deficient cells. Thus, EMSY, like that of BRCA2, appears to be important in maintaining the genomic stability of the cell. In this study we wanted to determine whether the chromosomal instability phenotype seen in the EMSY overexpressing breast epithelial cells could be reproduced in another cell line. Thus, EMSY was overexpressed in two normal ovarian surface epithelial cell lines, IOSE C9 and IOSE C21. However, a chromosomal instability phenotype was not observed in these two EMSY overexpressing ovarian cell lines even after 20 passages in culture. The cells were not sensitive to treatment with mitomycin C and in fact both cell lines were cytogenetically normal. Thus, in addition to EMSY overexpression subsequent treatment with a DNA damaging agent 186 did not mimic the chromosomal instability seen in the EMSY overexpressing breast epithelial cells. One study reported that BRCA2 deficient MEFs were insensitive to treatment with mitomycin C [11], whereas BRCA2 deficient lymphoid cells were highly sensitive to mitomycin C [8]. Mitomycin C is converted by cellular metabolism to an active genotoxic intermediate, and not all cell types support this conversion [12]. Thus, there are cell-type specific differences in mitomycin C genotoxicity. This may explain the insensitivity of the ovarian surface epithelial cells to mitomycin C treatment compared to the normal breast epithelial cells that were sensitive to mitomycin C. However, these BRCA2 deficient MEF cells do spontaneously accumulate chromosomal abnormalities [11], and the EMSY overexpressing IOSE cells did not. These results suggest that the IOSE cells used in this experiment may not be the correct cell type in which to study the functional mechanism of EMSY overexpression. There is preliminary evidence to suggest that the fallopian tube may be a source of cancers arising in high-grade serous carcinomas. Screening studies have failed to have an impact on the high mortality rate of high-grade ovarian serous carcinomas. These cancers rarely present at an early stage, and this has made the development of preventative strategies very difficult. However, the study of BRCA-linked breast and ovarian kindred have demonstrated the presence of fallopian-tube cancers in these women that are likely causally associated with germline mutations in BRCA1 and BRCA2 [13-15]. A recent study analyzed a series of bilateral salpingo-oophorectomies from BRCA-positive women following an index of fimbrial serous carcinoma to determine if the fimbria is a preferred site of origin. The fimbria is a 187 segment of the fallopian tube that is connected with the pelvic mesothelium and is in close proximity to the ovarian surface [16]. Tubal carcinomas were identified in five out of 13 BRCA-positive women, four of which were serous carcinomas. No ovarian carcinomas were identified. The fimbria was the most common location for early serous carcinomas in this series of BRCA-positive women and further investigations are required to determine its role in the pathogenesis of familial and sporadic ovarian serous carcinomas [16]. Thus, normal fallopian tube epithelial cells may be the correct experimental model in which to study the effects of EMSY overexpression. Secondly, we wanted to determine whether EMSY overexpressing breast epithelial cells are sensitive to DNA damaging agents such as doxorubicin and cisplatin. Sensitivity to DNA damaging agents is a characteristic feature of BRCA2 deficient cells. However, EMSY overexpressing cells treated with increasing concentrations of either drug did not exhibit an increased sensitivity as compared to control cells. Treatment with doxorubicin was equally toxic for the EMSY overexpressing cells as it was for the GFP-transduced cells and the wild type 184 hTerts. EMSY/GFP-transduced cells did exhibit a slightly increased sensitivity to cisplatin at concentrations of 5 and 10 nM, but this was not a significant difference. The non-DNA damaging agent, paclitaxel, as hypothesized was not toxic to the EMSY overexpressing cells or the control cells. These results suggest that EMSY overexpressing cells are not sensitive to the DNA damaging agents, doxorubicin and cisplatin. We do know that EMSY overexpressing cells are sensitive to treatment with mitomycin C, as these cells exhibited a several fold higher frequency of chromosome breaks. There is new evidence to suggest that EMSY may be a 188 controller of the G2/M checkpoint and may be acting as a repressor of cyclin B1 (unpublished data from collaborator Luke Hughes-Davies). Loss of EMSY repression may not be evident in a cell survival assay, particularly for normal cells that may have alternative mechanisms to ensure that the G2/M checkpoint in enforced. For example, the p53 pathway also signals directly to cyclin B1 and switches it off if there is DNA damage. EMSY may not directly play a role in the physical repair of DSBs, like BRCA2 or Rad51, and therefore demonstrating such a phenotype may be hard to demonstrate (as our results suggest). Perhaps a repair defect would only be seen in these cells if they were also deficient in other critical cell cycle checkpoint genes such as p53. Lastly, we wanted to determine whether the chromosomal instability phenotype seen in the EMSY overexpressing breast epithelial cells was due to a dominant negative effect. The structure of the ENT domain suggests that the EMSY protein exists as a dimer [4], and it may be that the 5' truncated EMSY is binding to the endogenous EMSY in the cell resulting in a dominant negative effect and inducing chromosomal instability. After 24 hours and treatment with mitomycin C, the EMSY siRNA treated cells accumulated numerous structural chromosomal abnormalities, accumulating almost double the amount of structural chromosomal aberrations as compared to the control cells. This preliminary data shows that loss of EMSY in the cell induces a chromosomal instability phenotype, suggesting that the chromosomal instability phenotype seen in the EMSY overexpressing cells may be due to a dominant negative effect. However, this observation needs to be reproduced. 189 In summary, these results demonstrate that although overexpression of a 5' fragment of EMSY is sufficient to induce a chromosomal instability phenotype in human breast epithelial cells this chromosomal instability phenotype was not seen in two normal ovarian surface epithelial cell lines. IOSE cells may not be the correct cell type in which to study EMSY overexpression, and perhaps normal fallopian tube epithelial cells may represent a better experimental system to evaluate the effects of EMSY overexpression. EMSY may not have a direct role in the actual repair of DSB and therefore demonstrating a repair phenotype for EMSY is likely to be more difficult. It would be of interest to evaluate whether treatment with doxorubicin, or cisplatin induces a chromosomal instability phenotype in EMSY overexpressing cells. And lastly, overexpression of the 5' end of EMSY in the normal breast epithelial cell line may act through a dominant negative effect. 190 7.5 R E F E R E N C E S 1. King, M.C., A novel BRCA2-binding protein and breast and ovarian tumorigenesis. N Engl J Med, 2004. 350(12): p. 1252-3. 2. Livingston, D.M., EMSY, a BRCA-2 partner in crime. Nat Med, 2004. 10(2): p. 127-8. 3. Turner, N., A. Tutt, and A. Ashworth, Hallmarks of 'BRCAness' in sporadic cancers. Nat Rev Cancer, 2004. 4(10): p. 814-9. 4. Chavali, G.B., et al., Crystal structure of the ENT domain of human EMSY. J Mol Biol, 2005. 350(5): p. 964-73. 5. Imren, S., et al., High-level beta-globin expression and preferred intragenic integration after lentiviral transduction of human cord blood stem cells. J Clin Invest, 2004. 114(7): p. 953-62. 6. Abbott, D.W., M.L Freeman, and J.T. Holt, Double-strand break repair deficiency and radiation sensitivity in BRCA2 mutant cancer cells. J Natl Cancer Inst, 1998. 90(13): p. 978-85. 7. Kraakman-van der Zwet, M., et al., Brca2 (XRCC11) deficiency results in radioresistant DNA synthesis and a higher frequency of spontaneous deletions. Mol Cell Biol, 2002. 22(2): p. 669-79. 8. Yu, V.P., et al., Gross chromosomal rearrangements and genetic exchange between nonhomologous chromosomes following BRCA2 inactivation. Genes Dev, 2000. 14(11): p. 1400-6. 9. Yuan, S.S., et al., BRCA2 is required for ionizing radiation-induced assembly ofRad51 complex in vivo. Cancer Res, 1999. 59(15): p. 3547-51. 10. Lee, H., et al., Mitotic checkpoint inactivation fosters transformation in cells lacking the breast cancer susceptibility gene, Brca2. Mol Cell, 1999. 4(1): p. 1-10. 11. Patel, K.J., etal., Involvement of Brca2 in DNA repair. Mol Cell, 1998. 1(3): p. 347-57. 12. Doroshow, J.H., Reductive activation of mitomycin C: a delicate balance. J Natl Cancer Inst, 1992. 84(15): p. 1138-9. 191 13. Rose, P.G., R. Shrigley, and GL. Wiesner, Germline BRCA2 mutation in a patient with fallopian tube carcinoma: a case report. Gynecol Oncol, 2000. 77(2): p. 319-20. 14. Aziz, S., et al., A genetic epidemiological study of carcinoma of the fallopian tube. Gynecol Oncol, 2001. 80(3): p. 341-5. 15. Zweemer, R.P., et al., Molecular evidence linking primary cancer of the fallopian tube to BRCA1 germline mutations. Gynecol Oncol, 2000. 76(1): p. 45-50. 16. Medeiros, F., et al., The tubal fimbria is a preferred site for early adenocarcinoma in women with familial ovarian cancer syndrome. Am J Surg Pathol, 2006. 30(2): p. 230-6. 192 Chapter 8. Summary, Overall Conclusions and Future Directions 8.1 S U M M A R Y A N D O V E R A L L C O N C L U S I O N S EMSY was first discovered using yeast two-hybrid screens to identify novel BRCA2 interacting proteins. The EMSY protein was shown to bind the exon 3 transactivation domain of BRCA2. Upon treatment with ionizing radiation EMSY co-localized with y-H2AX to sites of DNA damage in MEFs. BRCA2 is also known to relocalize to DSB repair sites after DNA damage, and these results suggest a common function for EMSY and BRCA2 [1]. In addition, EMSY mapped to chromosome 11q13.5, a region commonly amplified in breast cancer, and was found amplified in several breast cancer cell lines. These findings led to the very exciting hypothesis that EMSY gene amplification and overexpression could be a surrogate for BRCA2 loss in sporadic breast cancer [2, 3]. The studies presented in this dissertation focused on investigating the clinical significance of EMSY gene amplification and overexpression in sporadic breast and ovarian cancer. More specifically, we investigated the clinical and prognostic significance of EMSY gene amplification in sporadic breast cancer. The amplification of E/WSYwas also assessed in familial breast cancers with known mutations in BRCA1 and BRCA2. Comprehensive studies of the 11q13 amplicon in ovarian cancer were also performed. Lastly, we looked at the biological effects of EMSY overexpression in normal breast and ovarian surface epithelial cells. From these studies, we have determined that EMSY is frequently amplified in breast cancer. In two large independent breast cancer series EMSY was consistently 193 found amplified in 13% of cases. Moreover, EMSY gene amplification was associated with a poor outcome in all patients, and in several other pathological subsets of breast cancer. Thus, EMSY appears to be an important element of the 11q13 amplicon in breast cancer. EMSY gene amplification occurred at a significantly lower frequency in BRCA2 mutation carriers as compared to BRCA1 mutation carriers and non-BRCA1/2 familial breast cancers. These results demonstrate that EMSY amplification may be an unnecessary event in patients with a BRCA2 mutation and lends support to the hypothesis that EMSY gene amplification may be a surrogate for BRCA2 loss in sporadic cancer. The 11q13 amplicon had long been thought of as a single amplicon, with CCND1 as the potential driver of this amplification event. However, further analysis in breast cancer led to the discovery of four additional regions of amplification along 11 q13, suggesting the involvement of other genes in this amplification event. EMSY maps to the largest and most telomeric amplicon on 11 q13, which spans over 2 Mb. Several other genes have also recently been reported amplified in this region, including PAK1 [4] and Rsf-1 [5-7]. However, no study has looked at the involvement of all of these genes in the same cohort of patients. In our first study, we demonstrated that EMSY gene amplification occurs in 18% of ovarian cancers and is specifically associated with a serous histology. aCGH mapping of 11q13 showed the most frequently amplified region in ovarian cancer encompassed EMSY, PAK1 and GAB2. In addition, amplification of CCND1, EMSY, and PAK1 were associated with elevated RNA expression levels, further supporting their roles in oncogenesis. In our second study of the 11q13 amplicon five genes along this 194 region were studied in a large clinically annotated ovarian cancer series. Here, we were able to demonstrate that all five genes were frequently amplified in ovarian cancer, and were associated with a serous histology. Amplification of EMSY and Rsf-1 is also associated with a poor prognosis. These results show that amplification of the more centromeric amplicon containing CCND1 and the amplicon containing EMSY, Rsf-1, PAK1, and GAB2 are specifically associated with serous carcinomas, and are likely to be involved in the pathogenesis of these tumors. Clinicopathological and molecular genetic studies have provided evidence to suggest that the subtypes of ovarian cancer are different and distinct diseases. The finding that all five genes are significantly associated with a serous histology, and not associated with any other histological subtype supports the proposal that these cancers develop along distinct molecular pathways. From these experiments we know that EMSY is commonly amplified and associated with a poor outcome in both breast and ovarian cancer. However, questions remain to be answered regarding the exact function of EMSY in the cell and how this relates to BRCA2, breast and ovarian cancer. What are the effects of EMSY overexpression in these cancers? Does overexpression of EMSY alter DNA repair? Is EMSY overexpression associated with features that are associated with BRCA2 deficient cancers? We wanted to address some of these questions and to do so normal breast and ovarian surface epithelial cells were established that overexpressed the 5' end of EMSY. A post-doctoral fellow from the laboratory of Dr Connie Eaves, at the Terry Fox laboratory, was successful in cloning the 5' end of EMSY, which contains the only known functional domain, the ENT domain. 195 Overexpression of EMSY in a normal breast epithelial cell line induced chromosomal instability in these cells. In addition, this chromosomal instability was enhanced by the addition of a known DNA damaging agent, mitomycin C. Cells deficient in BRCA2 sustain spontaneous chromosomal aberrations that accumulate during passage in culture, indicating an accumulation of DNA damage during division [8-11]. Sensitivity to DNA cross-linking agents, such as mitomycin C (MMC) and cisplatinum have also been reported in BRCA2 deficient cells [8, 11-13]. These findings suggest that EMSY may play a role in carcinogenesis, similar to BRCA2, by deregulating genomic stability. EMSY overexpression did not induce a chromosomal instability phenotype in normal ovarian surface epithelial cells, and these overexpressing cells were insensitive to treatment with mitomycin C. These results suggest that ovarian surface epithelium may be the wrong cell type in which to study the mechanism of EMSY overexpression. Future studies with another cell type are needed in order to confirm the chromosomal instability phenotype observed in the EMSY overexpressing breast epithelial cells. EMSY overexpressing breast epithelial cells did not display sensitivity to treatment with the DNA damaging agents, doxorubicin and cisplatin. EMSY may not play a direct physical role in the repair of DSBs, and may be involved in cell cycle control. Therefore, a repair phenotype may be difficult to see with a cell survival assay and perhaps an effect on genomic stability would have been a better experiment to perform with these DNA damaging agents. 196 Lastly, we evaluated whether the chromosomal instability phenotype seen in the EMSY overexpressing breast epithelial cells was due to a dominant negative mechanism. The crystal structure of EMSY has revealed that it is a dimer, and it may be that the truncated 5' EMSY is binding to the endogenous EMSY in the cell resulting in the chromosomal instability phenotype. After 24 hours and treatment with mitomycin C, the EMSY siRNA treated cells accumulated numerous structural chromosomal abnormalities, accumulating almost double the amount of structural chromosomal aberrations as compared to the control cells. This preliminary data suggests that loss of EMSY in the cell induces a chromosomal instability phenotype. Thus, the chromosomal instability phenotype seen in the EMSY overexpressing cells may be due to a dominant negative effect. Throughout my graduate studies I have encountered numerous problems and challenges that my fellow colleagues working on EMSY have also run into. One of the major problems has been that the full length EMSY protein is unstable. As a result no one has been able to clone and express the full-length EMSY protein despite numerous attempts from many laboratories including our own. What we were able to do was to, as mentioned above, was to clone an 810 bp 5' fragment of EMSY. EMSY has been shown to interact with BRCA2, a protein that has also proven to be difficult to work with. Despite being identified over 13 years ago the precise function of this protein is still largely unknown. One difficulty in studying the function of BRCA2 is that there is only one human cell line known to carry a mutation in 197 BRCA2. This cell line, Capan-1, is a pancreatic cell line in which one wild type allele is lost, and the second allele carries the 6174delT mutation which gives rise to a truncated BRCA2 protein. There have also been difficulties in the production of expression constructs due to the large proteins encoded by BRCA1 and BRCA2. Another issue has been the lack of antibodies for EMSY. We have wanted to evaluate EMSY protein expression on our breast and ovarian TMAs. However, we have screened numerous commercial antibodies and none have proven to work on paraffin embedded tissue microarrays. We have discovered a polyclonal EMSY goat antibody that does work well for western blots. 8.2 F U T U R E D I R E C T I O N S We have hypothesized that overexpression of the truncated 5' EMSYcDNA used in our overexpression experiments might produce a BRCA-null phenotype through a dominant-negative mechanism. Our preliminary data suggests that the chromosomal instability phenotype seen in the EMSY overexpressing cells may arise through a dominant negative effect. EMSY siRNA treated breast epithelial cells, treated with mitomycin C, displayed almost twice as many structural chromosomal abnormalities as compared to the control cells. This experiment needs to be repeated in order to validate these findings. This siRNA experiment suggests that EMSY abnormalities, either overexpression or loss, may be detrimental to the cell. To investigate whether EMSY loss is a clinically significant event we do have gene copy number analysis from the FISH data on the 198 population based breast cancer series. As mentioned earlier this subset has been divided into a training set, in which we can investigate the clinical implications of EMSY loss in breast cancer patients. Preliminary results showed that loss of EMSY (defined as an amplification ratio <, 0.6) does occur, although it is not as common as amplification. We also evaluated whether loss of CCND1 was also observed. In this series loss of CCND1 was not seen. Kaplan-Meier survival curves show that those patients with either EMSY amplification or EMSY loss are associated with a poor outcome (Figure 8.1A). This association with poor outcome is also seen in node positive patients (Figure 8.1 B). Thus, it appears as though abnormalities in EMSY, either amplification or loss, are involved in sporadic breast cancer. The fact that loss of EMSY may also be an important event is difficult to fathom as EMSY is thought to be an oncogene. How exactly EMSY loss is contributing to breast cancer is unknown, and it may be that perturbations of EMSY, either overexpression or loss, are functionally important. Such a contradictory finding has been seen for BRCA2 where several studies have shown that BRCA2 overexpression significantly correlated with an aggressive phenotype in breast cancer [14, 15]. One study reported that BRCA2 overexpression significantly correlated with grade III tumors, which was mainly attributed to nuclear pleomorphism and high mitotic account [14]. Another study confirmed these findings and also demonstrated that patients with high BRCA2 mRNA levels were more significantly associated with a poor outcome than those with low levels [15]. Such a finding is also puzzling given the fact that BRCA2 is considered a tumor suppressor gene. A model to explain these results has been 199 A) CC > 0.6-02-1 B) | 36-W 04 All patients EMSY not amplified „ , EMSY amplified EMSY loss — i — 5.00 —I Total follow up (years) Node+ patients EMSY not amplified EMSY amplified EMSY loss - 1 — 000 10 CO Total follow up (years) Figure 8.1 EMSY amplification and loss are associated with a poor outcome A) Kaplan-Meier survival curve showing those tumors with EMSY amplification and those with EMSY loss are associated with a shortened DSS. B) This effect on survival was also seen in node positive tumors. 200 proposed that suggests BRCA2 mRNA expression is induced by proliferation. Thus, if BRCA2 mRNA expression is intact in tumors, high proliferation may result in the upregulation of BRCA2. Rajan er al. suggested the possible existence of a feedback loop in which proliferation induces BRCA2 expression, that in turn would inhibit proliferation [14, 15]. These results suggest that BRCA2 may not act as a classical tumor suppressor gene in sporadic breast cancer. It may be that EMSY also does not act as a classical oncogene in sporadic breast, however this remains to be seen. Future experiments to elucidate the functional role of EMSY in cancer, requires the ability to overexpress the full length EMSY protein. However, this has proven to be a very difficult task. Once established, a cell in which the full length EMSY is overexpressed would help resolve some of the questions about what exactly EMSY is doing in the cell. We would be able to determine whether the chromosomal instability phenotype is directly due to overexpression of EMSY or the result of a dominant negative effect. Also, having a cell line overexpressing a tagged EMSY protein would also allow us to determine whether EMSY is interacting with BRCA2 and whether this interaction interferes with the normal role of BRCA2 and its ability to bind Rad51. Another question that remains to be answered is whether EMSY overexpression is sufficient to cause tumorigenesis. To investigate this EMSY overexpressing cells could be injected into the mammary fat pad of mice and see if tumors develop. Another possibility would be to create a mouse mammary tumor virus (MMTV)-EMSY transgenic mouse model, in which EMSY expression would be targeted to the 201 mammary gland, and evaluate whether EMSY overexpression is sufficient to induce tumorigenesis in the mice. If these EMSY overexpressing mice do develop mammary tumors this would provide strong evidence for a critical role for EMSY in the development of breast cancer. Such a role has been shown for Cyclin D1. Overexpression of cyclin D1 in the mammary gland resulted in increased proliferation and the MMTV-cyclin D1 mice developed tumors [16]. However, the tumors appeared after a long latency period of 18 months, which suggested that cyclin D1 is a relatively weak oncogene compared to c-myc, Ha-ras, and c-neu, with a latency period of 11, 6, and 3 months respectively [17, 18]. Cyclin D1 appears to be critical for some pathways of mammary tumorigenesis, as cyclin D1 deficient mice are resistant to breast cancers induced by c-neu and Ha-ras, but are sensitive to breast cancers induced by c-myc and Wnt-1 [19]. The data presented in this dissertation demonstrates the complexity of amplicons such as 11q13 in ovarian cancer. It is not difficult to imagine that the amplification and overexpression of multiple oncogenes at 11 q 13 could impart a selective advantage for the tumor and could likely represent a potential therapeutic target. Most studies have evaluated individual genes on the amplicon; however, it may be that overexpression of these oncogenes could have synergistic and additive effects. Functional studies of this amplicon in ovarian cancer will require consideration of these combined effects and could be studied in vitro and in vivo. To perform this experiment in vitro would require the development of expression constructs, preferably tagged, for CCND1, EMSY, PAK1, Rsf-1 and GAB2. After establishing the expression constructs for each gene, the constructs could then be transfected 202 into 184-hTert cells. Each stable cell line would then need to be verified for the expression of each gene using quantitative real-time PCR and Western blot analysis. Each stable cell line expressing one of the genes along 11q13 would be assessed as to whether expression of each gene induces transformation. Overexpressing cells would be assessed for differences in proliferation, cell morphology, invasion, apoptosis, as well as the cells' ability to grow anchorage independently using colony-forming assays. After establishing how each of the overexpressing cell lines differ from the control cells, we would then establish a cell line that overexpresses all five genes. After validating the expression of all five genes, we would then determine whether overexpression of these genes enhances the transformation properties of these cells. We would expect that if these oncogenes were acting synergistically to promote tumorigenesis, overexpression of all five genes would have a marked effect on the functional analyses described above as compared to each gene individually. Another approach that could be undertaken at the same time is to find an ovarian cancer cell line that overexpresses all five oncogenes. Using RNA interference each gene could be knocked out sequentially to determine which of the genes are essential for cell survival. This would enable us to determine the gene(s) responsible for transformation of the cell. The findings from these studies could then be extended to in vivo studies. 184-hTert breast epithelial cells are non-tumorigenic and therefore the corresponding cell line overexpressing all five genes could be injected into nude mice to determine the tumorigenic potential of these cells. In addition, overexpression of all five genes could also be introduced into a tumorigenic ovarian cancer cell line, such as 203 0VCAR3 cells. These cells are known to induce tumor formation, and we could assess whether overexpression of the genes on 11q13 enhance the tumorigenecity of these cells in vivo. If the in vitro studies provide evidence that thel 1 q13 amplicon contributes to ovarian carcinogenesis through the synergistic oncogenic effects of several genes the development of a mouse model for ovarian cancer would be a useful tool for understanding the molecular basis of ovarian cancer. One such mouse model has been described by Orsulic etal [20]. This group developed a mouse model for ovarian cancer using an avian retroviral gene delivery technique to introduce multiple genetic lesions in a mouse cell line engineered to express the avian virus receptor, TVA. Combinations of the oncogenes, c-myc, K-ras, and Akt were introduced into primary mouse ovarian cells from p53 deficient mice. The addition of any two of the oncogenes were sufficient to induce tumor formation when infected cells were injected subcutaneously, intraperitonealy, or when transplanted under the ovarian bursa [20]. We could follow the same approach in which ovarian cells from the transgenic mice engineered to express the avian receptor TVA could be infected in vitro with multiple vectors carrying coding sequences for the five oncogenes on 11q13. We could then assess which combinations of the oncogenes induce ovarian tumor formation. These studies would help demonstrate the complex nature of amplicons such as 11q13 and provide supporting evidence that these amplicons likely contribute to ovarian carcinogenesis through the synergistic oncogenic effects of several genes rather than by a single dominant oncogene. 204 8.3 REFERENCES 1. Hughes-Davies, L., et al., EMSY links the BRCA2 pathway to sporadic breast and ovarian cancer. Cell, 2003. 115(5): p. 523-35. 2. King, M.C., A novel BRCA2-binding protein and breast and ovarian tumorigenesis. N Engl J Med, 2004. 350(12): p. 1252-3. 3. Livingston, D.M., EMSY, a BRCA-2 partner in crime. Nat Med, 2004. 10(2): p. 127-8. 4. Schraml, P., et al., Combined array comparative genomic hybridization and tissue microarray analysis suggest PAK1 at 11q13.5-q14 as a critical oncogene target in ovarian carcinoma. Am J Pathol, 2003.163(3): p. 985-92. 5. Davidson, B., et al., Expression of the chromatin remodeling factor Rsf-1 is upregulated in ovarian carcinoma effusions and predicts poor survival. Gynecol Oncol, 2006. 6. Mao, T.L., et al., Expression of Rsf-1, a chromatin-remodeling gene, in ovarian and breast carcinoma. Hum Pathol, 2006. 37(9): p. 1169-75. 7. Shih le, M., et al., Amplification of a chromatin remodeling gene, Rsf-1/HBXAP, in ovarian carcinoma. Proc Natl Acad Sci U S A , 2005. 102(39): p. 14004-9. 8. Kraakman-van der Zwet, M., et al., Brca2 (XRCC11) deficiency results in radioresistant DNA synthesis and a higher frequency of spontaneous deletions. Mol Cell Biol, 2002. 22(2): p. 669-79. 9. Lee, H., et al., Mitotic checkpoint inactivation fosters transformation in cells lacking the breast cancer susceptibility gene, Brca2. Mol Cell, 1999. 4(1): p. 1-10. 10. Patel, K.J., et al., Involvement ofBrca2 in DNA repair. Mol Cell, 1998. 1(3): p. 347-57. 11. Yu, V. P., et al., Gross chromosomal rearrangements and genetic exchange between nonhomologous chromosomes following BRCA2 inactivation. Genes Dev, 2000. 14(11): p. 1400-6. 205 12. Donoho, G., et al., Deletion of Brca2 exon 27 causes hypersensitivity to DNA crosslinks, chromosomal instability, and reduced life span in mice. Genes Chromosomes Cancer, 2003. 36(4): p. 317-31. 13. Yuan, S.S., et al., BRCA2 is required for ionizing radiation-induced assembly ofRad51 complex in vivo. Cancer Res, 1999. 59(15): p. 3547-51. 14. Bieche, I., C. Nogues, and R. Lidereau, Overexpression ofBRCA2 gene in sporadic breast tumours. Oncogene, 1999. 18(37): p. 5232-8. 15. Egawa, C , et al., High BRCA2 mRNA expression predicts poor prognosis in breast cancer patients. Int J Cancer, 2002. 98(6): p. 879-82. 16. Wang, T.C., et al., Mammary hyperplasia and carcinoma in MMTV-cyclin D1 transgenic mice. Nature, 1994. 369(6482): p. 669-71. 17. Muller, W.J., et al., Single-step induction of mammary adenocarcinoma in transgenic mice bearing the activated c-neu oncogene. Cell, 1988. 54(1): p. 105-15. 18. Sinn, E., et al., Coexpression of MMTV/v-Ha-ras and MMTV/c-myc genes in transgenic mice: synergistic action of oncogenes in vivo. Cell, 1987. 49(4): p. 465-75. 19. Yu, Q., Y. Geng, and P. Sicinski, Specific protection against breast cancers by cyclin D1 ablation. Nature, 2001. 411(6841): p. 1017-21. 20. Orsulic, S., et al., Induction of ovarian cancer by defined multiple genetic changes in a mouse model system. Cancer Cell, 2002. 1(1): p. 53-62. 206 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0100334/manifest

Comment

Related Items