identification of genes and potential pathways involved in ... · identification of genes and...
TRANSCRIPT
Identification of Genes and Potential Pathways Involved in
Familial Ovarian Cancer
By
Kelly Kai Yin Seto
A thesis submitted in conformity with the requirements
for the degree of Doctor of Philosophy,
Graduate Department of Molecular and Medical Genetics
University of Toronto
© Copyright by Kelly Kai Yin Seto (2011)
Identification of Genes and Potential Pathways Involved in
Familial Ovarian Cancer
Degree of Doctor of Philosophy, 2011
Kelly Kai Yin Seto
Department of Molecular Genetics
University of Toronto
ABSTRACT
One of the most important risk factors in ovarian cancer is family history, and two
well-studied tumour suppressor genes BRCA1 and BRCA2 have already been identified in
“high-risk” families. However, alterations of other genes may also be important for
ovarian cancer pathogenesis in individuals with family history of breast/ovarian cancer.
In this thesis, I compared the gene expression profiles of tumours from patients
with strong and weak family history of breast and/or ovarian cancer to identify genes that
may be significant in the subset of patients with ovarian cancer predisposition. Based on
this comparison, two genes of interest were selected for further investigations:
hCDC4/FBXW7 (F-box and WD repeat domain containing 7) and PRKCZ (protein kinase
C zeta).
Through mutational analyses I identified one nucleotide alteration within exon 7
of hCDC4; however, overall I found that hCDC4 mutation is a rare event in ovarian
tumours. Additional epigenetics analyses revealed that promoter methylation is not a
significant mechanism responsible for repression of hCDC4 expression in ovarian cancer.
Nevertheless, the variable expression of hCDC4 proteins observed in ovarian tumour
ii
tissues by immunohistochemical staining of tissue microarrays suggests that hCDC4
deregulation may potentially be important in a subset of ovarian cancers.
Additionally, I observed that expression levels of PRKCZ are higher in ovarian
tumours from patients with strong family history compared to patients with weak family
history. PRKCZ has previously been shown to be involved in a variety of cellular
processes; however its role in ovarian cancer remained elusive. To further understand the
role of PRKCZ in ovarian tumourigenesis, including cell viability, cell migration, as well
as relevant downstream signaling pathways, I performed functional assays using an in
vitro ovarian cancer model. I observed that PRKCZ increases proliferation of the
SKOV3 ovarian cancer cell line and participates in EGF-induced chemotaxis.
Furthermore, I identified IGF1R (insulin-like growth factor 1 receptor) and ITGB3
(integrin beta 3) as downstream effectors of PRKCZ as expression of these genes is
significantly altered when PRKCZ is over-expressed. Given their previously identified
associations with familial ovarian cancer, the IGF1 and ITGB3 signaling pathways may
therefore represent a possible link between PRKCZ and this disease.
iii
ACKNOWLEDGEMENTS
First and foremost I would like to thank my supervisor, Dr. Irene Andrulis, for
giving me the opportunity to work on this meaningful project, and for your continuous
guidance during my graduate career. I would also like to thank my committee members,
Dr. Johanna Rommens and Dr. Jeff Wrana. Your intellectual input and enthusiasm for
my project were much appreciated.
I could not have been able to complete my thesis without the help and advice of
my collaborators. Thank you to Dr. Barry Rosen, Dr. Joan Murphy, Dr. Patricia Shaw
and Heather Begley for providing samples for my study, and Gordon Glendon for his
help for assisting in classification of familial ovarian cancer cases. Thank you to Dr.
Shelley Bull, Dr. Dushanthi Pinnaduwage, and Sarah Colby for their help with statistical
analyses of microarray data. I would also like to thank all members of the Toronto
Ovarian Research Network for discussions on my project, and a special thank you goes
Dr. Ted Brown, Alicia Tone and Katherine Sodek for all their helpful advice.
I am very grateful for the support from my “lab family”. Nalan, thank you for
being so generous with your time – from training me on my first microarray experiment,
to the many hours of discussions in the years that followed. Kolja, Chris, Sherry, Andras,
Anita, and Teresa – it was amazing to have such an intelligent group of Ph.D. students
that I could look up to; your mentorship is much appreciated. Yan, it was such a pleasure
to work with you, and I admire your inherent desire for learning. Lucie, your love for
science is inspiring, and I look forward to many more photography outings with you.
Mona, thanks for hiding chocolates in my desk to prop me up during those down days.
To all Andrulis lab members, past and present: Sasha, Eduard, Monica, Salvador,
iv
Catherine, Andrew, Irene, Andreas, Atta, Kristine, Sandrine, and our numerous talented
undergraduate students – thank you for all your support and friendships. Sean, Duygu,
and Winnie, your constant words of encouragement, optimism, and friendships mean a
great deal to me.
I would also like to take this opportunity to thank all my friends, from childhood
buddies to friends I have made in the department of Molecular Genetics and at the
Samuel Lunenfeld Research Institute (SLRI) – you definitely have made my graduate
experience fun and memorable. Thank you to my formal labmates and friends from
McMaster University who played a big part in my decision to pursue my graduate
studies. A huge thank you goes to SciHigh, the wonderful science outreach program
based at the SLRI. Seeing the look of excitement on kids’ faces everytime they look at
GFP-mice or banana DNA in eppendorf tubes reminds me of why I love science in the
first place.
Last but not least, I would like to thank my parents and brother for their patience
and support through this journey. Words cannot express how lucky I feel to have you.
v
TABLE OF CONTENTS ABSTRACT ....................................................................................................................... ii ACKNOWLEDGEMENTS............................................................................................... iv TABLE OF CONTENTS .................................................................................................. vi LIST OF TABLES.............................................................................................................. x LIST OF FIGURES ........................................................................................................... xi LIST OF ABBREVIATIONS ......................................................................................... xiv
CHAPTER 1
Introduction & Literature Review.................................................................................. 1 1.1 Principles of Molecular Cancer Genetics ..................................................................... 2
1.1.1 Gatekeepers............................................................................................................ 3 1.1.1.1 Tumour Suppressor Genes.............................................................................. 5 1.1.1.2 Oncogenes....................................................................................................... 7
1.1.2 Caretakers .............................................................................................................. 7 1.1.3 Landscapers ........................................................................................................... 8
1.2 Ovarian Cancer ............................................................................................................. 9 1.2.1 Model of Ovarian Carcinogenesis ....................................................................... 11 1.2.2 Familial Ovarian Cancer...................................................................................... 15
1.2.2.1 Hereditary Site-Specific Ovarian Cancer and Hereditary Breast-Ovarian Cancer Syndrome (HBOCS) .................................................................................... 16 1.2.2.2 Hereditary Nonpolyposis Colorectal Cancer (HNPCC) ............................... 17
1.2.3 Molecular Pathology and Genetics of Ovarian Cancer ....................................... 17 1.2.3.1 p53 ................................................................................................................ 18 1.2.3.2 Wnt-Signaling Pathway................................................................................ 18 1.2.3.3 PI3K/Akt Signaling Pathway........................................................................ 19 1.2.3.4 MAPK Signaling Pathway............................................................................ 20 1.2.3.5 Cell Cycle Genes .......................................................................................... 21 1.2.3.6 Epidermal Growth Factor Family Receptors................................................ 23 1.2.3.7 Estrogen Receptors ....................................................................................... 24
1.2.4 BRCA1 and BRCA2............................................................................................ 25 1.2.4.1 BRCA1 and BRCA2 in Ovarian Cancer ...................................................... 26 1.2.4.2 Mutations of BRCA1 and BRCA2 in Ovarian Cancer................................. 27 1.2.4.3 Specific BRCA1 and BRCA2 mutations in closed populations................... 29 1.2.4.4 Modifiers of BRCA1 and BRCA2................................................................ 30
1.3 Ovarian Cancer Genome-Wide Association Studies.................................................. 31 1.3.1 Array Comparative Genomic Hybridization of Ovarian Cancer......................... 32 1.3.2 Single Nucleotide Polymorphism Array Analysis of Ovarian Cancer ................ 34 1.3.3 Gene Expression Profiling of Ovarian Cancer .................................................... 37
1.4 Hypothesis .................................................................................................................. 41 1.5 Rationale and Objectives ............................................................................................ 41
vi
CHAPTER 2
Gene Expression Profiling of Familial Ovarian Cancer ............................................. 43 2.1 Introduction ................................................................................................................ 44 2.2 Materials and Methods ............................................................................................... 45
2.2.1 Ovarian Cancer Specimens.................................................................................. 45 2.2.2 Family History Classification.............................................................................. 45 2.2.3 RNA Isolation and Reverse Transcription........................................................... 47 2.2.4 cDNA Expression Microarrays............................................................................ 47 2.2.5 Microarray Data Analysis.................................................................................... 49
2.2.5.1 Pre-processing and normalization of expression data .................................. 49 2.2.5.2 Microarray Statistical Analysis..................................................................... 49
2.2.6 Quantitative Real-time RT-PCR.......................................................................... 50 2.2.7 Integration of Array Data to Interaction Networks.............................................. 52
2.3 Results ........................................................................................................................ 53 2.3.1 Identification of Genes Distinguishing Strong and Weak Familial Ovarian Cancers ......................................................................................................................... 53
2.3.1.1 Supervised-Class Comparison...................................................................... 53 2.3.1.2 Candidate Gene Approach............................................................................ 58
2.3.2 Validation of Differentially Expressed Genes..................................................... 58 2.3.3 Molecular Network Analyses .............................................................................. 61
2.4 Discussion................................................................................................................... 68
CHAPTER 3 hCDC4 in Familial Ovarian Cancer.............................................................................. 75 3.1 Introduction ................................................................................................................ 76 3.2 Materials & Methods .................................................................................................. 79
3.2.1 Ovarian tumour samples, and RNA, DNA Extraction ........................................ 79 3.2.2 Protein Truncation Test ....................................................................................... 79 3.2.3 Single Strand Conformation Polymorphism (SSCP) and Manual Sequencing... 80 3.2.4 DNA Methylation-Specific PCR ......................................................................... 82 3.2.5 Loss of Heterozygosity (LOH) Analysis of hCDC4............................................ 83 3.2.6 Immunohistochemical (IHC) Staining of Ovarian Tissue Microarrays .............. 84 3.2.7 Quantitative Real-time PCR for CCNE1 ............................................................. 85
3.3 Results ........................................................................................................................ 86 3.3.1 hCDC4 Sequence Alteration Detection by PTT Analysis................................... 86 3.3.2 SSCP and Sequencing of hCDC4 ........................................................................ 86 3.3.3 hCDC4 Promoter Methylation Analysis.............................................................. 90 3.3.4 Loss of Heterozygosity Analysis of hCDC4 ....................................................... 90 3.3.5 hCDC4 Protein Expression in Ovarian Cancer ................................................... 94 3.3.6 Gene Expression of CCNE1 in Familial Ovarian Cancer.................................... 98
3.4 Discussion................................................................................................................. 100
vii
CHAPTER 4 Characterization of PRKCZ in Ovarian Cancer....................................................... 106 4.1 Introduction .............................................................................................................. 107 4.2 Materials and Methods ............................................................................................. 109
4.2.1 Cell Culture........................................................................................................ 109 4.2.2 PRKCZ Expression Vector & Generation of Stable Clones ............................. 109 4.2.3 Quantitative Real-Time PCR............................................................................. 110 4.2.4 Western Blotting................................................................................................ 110 4.2.5 MTT Cell Viability Assays................................................................................ 111 4.2.6 TUNEL Assays.................................................................................................. 112 4.2.7 BrdU Proliferation Assay .................................................................................. 112 4.2.8 Matrigel Transwell Assays ................................................................................ 113 4.2.9 Scratch Wound Healing and Pericentrin Orientation Assays ............................ 113 4.2.10 Phagokinetic Track Assays.............................................................................. 114 4.2.11 siRNA Transfections ....................................................................................... 115 4.2.12 Ingenuity Pathway Analyses............................................................................ 115 4.2.13 Statistical Analyses.......................................................................................... 115
4.3 Results ...................................................................................................................... 116 4.3.1 Generation of PRKCZ Stable Ovarian Cancer Cell lines.................................. 116 4.3.2. Cell Viability in PRKCZ-Expressing Cells ...................................................... 125 4.3.3 PRKCZ and ovarian cancer cell migration and invasion................................... 128 4.3.4 Identification of Potential Downstream Effectors of PRKCZ........................... 137
4.3.4.1 IGF1R and ITGB3 as Potential Downstream Effectors of PRKCZ ........... 137 4.3.4.2 TIMP-1 as a Potential Downstream Effector in ITGB3 and IGF1 Signaling................................................................................................................................ 146 4.3.4.3 Effects of IGF and ITGB3 Signaling on Cell Migration/Invasion in SKOV3 Cells ........................................................................................................................ 150
4.4 Discussion................................................................................................................. 156 CHAPTER 5 Conclusions and Future Directions............................................................................. 166 5.1 Summary and Implications of Thesis Findings ........................................................ 167
5.1.1 Expression Profiling of Familial Ovarian Cancer ............................................. 167 5.1.2 hCDC4 and Ovarian Cancer .............................................................................. 169 5.1.3 PRKCZ and Ovarian Cancer ............................................................................. 170
5.2 Future Directions ...................................................................................................... 173 5.2.1 High-Throughput Analyses of Familial Ovarian Cancer................................... 173
5.2.1.1 Gene Set Analysis of Expression Microarrays Data................................... 173 5.2.1.2 Genomic Signature of Familial Ovarian Cancer ........................................ 182
5.2.2 hCDC4 and Ovarian Cancer .............................................................................. 182 5.2.3 PRKCZ and Ovarian Cancer ............................................................................. 183
viii
Appendix…………………………………………………………………………….…187 Bibliography.................................................................................................................. 189
ix
LIST OF TABLES Table 1-1. Characteristics of Type I and Type II ovarian tumours..................................13 Table 2-1. Characterization of the subjects in the strong versus weak family history groups................................................................................................................................46 Table 2-2. Thirteen cell lines composing common reference pool used for microarray experiments........................................................................................................................47 Table 2-3. Top 100 differentially expressed genes between strong and weak familial ovarian tumours, as ranked by SAM.................................................................................55 Table 2-4. Top functions of networks as identified by Ingenuity Pathway Analysis.......67 Table 3-1. Primer sets for hCDC4 SSCP and manual sequencing analyses.....................81 Table 3-2. Primer sequences for methylation-specific PCR.............................................82 Table 3-3. Polymorphic markers used for LOH analysis of hCDC4................................83 Table 3-4. Histological scores of hCDC4 immunhistochemical staining on ovarian tissue microarray..........................................................................................................................97 Table 5-1. Gene Set Analysis (GSA) of familial ovarian cancer expression microarray data...................................................................................................................................175 Table 5-2. Top functions of networks belonging to significant gene sets as identified by Gene Set Analysis (GSA) and Ingenuity Pathway Analysis (IPA).................................176
x
LIST OF FIGURES Figure 1-1. Hallmarks of Cancer…………………………………………………………4 Figure 1-2. Knudson’s two-hit hypothesis……………………………………….....……6 Figure 1-3. Dualistic model of ovarian serous carcinoma development………….…….14 Figure 1-4. Gene structure of BRCA1 and BRCA2……………………………………...28 Figure 1-5. Chromosomal aberrations in ovarian tumours...............................................35 Figure 2-1. Subarray of a representative 19K cDNA microarray.....................................48 Figure 2-2. Expression of housekeeping genes in ovarian tumour samples as measured by gene expression microarrays.........................................................................................51 Figure 2-3 Heatmap illustrating differential gene expression patterns in strong and weak familial ovarian cancer groups...........................................................................................54 Figure 2-4. Identification of hCDC4 as a differentially expressed gene between strong and weak familial ovarian tumours by candidate gene approach......................................59 Figure 2-5. Real-time PCR validation of cDNA microarray expression analysis............60 Figure 2-6. Graphical representations of the molecular relationships between genes identified from familial ovarian microarray analysis using Ingenuity Pathway Analysis.............................................................................................................................62 Figure 3-1. Pathway of hCDC4-mediated degradation of cyclin E..................................77 Figure 3-2. hCDC4 protein truncation assay....................................................................87 Figure 3-3. Genetic analysis of hCDC4 with SSCP.........................................................88 Figure 3-4. hCDC4 gene sequence alteration found in exon 7.........................................89 Figure 3-5. Potential methylation sties with the hCDC4 promoter..................................91 Figure 3-6. Evaulation of hCDC4 promoter methylation by methylation-specific PCR....................................................................................................................................92 Figure 3-7. Loss of heterozygosity (LOH) analysis of hCDC4 in four cases of ovarian cancer.................................................................................................................................93 Figure 3-8. hCDC4 IHC staining optimization.................................................................95
xi
Figure 3-9. Immunohistochemical staining of ovarian tissue microarray with hCDC4 antibody.............................................................................................................................96 Figure 3-10. CCNE1 gene expression in familial ovarian cancer.....................................99 Figure 4-1. Endogenous gene and protein levels of PRKCZ in selected ovarian cancer cell lines...........................................................................................................................117 Figure 4-2. Expression of PRKCZ clones in HEY ovarian cancer cell line...................119 Figure 4-3. Expression of PRKCZ clones in SKOV3 ovarian cancer cell line..............121 Figure 4-4. Expression of PRKCZ clones in OVCAR3 ovarian cancer cell line...........123 Figure 4-5. PRKCZ increases cell viability in SKOV3 cells but not HEY and OVCAR3 cells..................................................................................................................................126 Figure 4-6. PRKCZ enhances proliferation of SKOV3 ovarian cancer cells.................127 Figure 4-7. PRKCZ has no effect on apoptosis in SKOV3 cells....................................127 Figure 4-8. Migration of ovarian cancer cells................................................................130 Figure 4-9. Effect of PRKCZ gene knockdown on SKOV3 parental cells migration as observed by wound healing assay....................................................................................131 Figure 4-10. Disorganized cell movement of HEY cells over-expressing PRKCZ........133 Figure 4-11. Measurement of cell polarity of HEY by pericentrin orientation assay.....135 Figure 4-12. Quantitation of ovarian cancer cell motility...............................................136 Figure 4-13. Identification of potential interactors of PRKCZ........................................138 Figure 4.14. Transcript and protein expression of IGF1R in PRKCZ-expressing SKOV3 cells..................................................................................................................................139 Figure 4.15. Transcript and protein expression of IGF1R in PRKCZ-expressing OVCAR3 cells........................... .....................................................................................141 Figure 4.16. Transcript and protein expression of ITGB3 in PRKCZ-expressing SKOV3 cells..................................................................................................................................142 Figure 4.17. Transcript and protein expression of IGF1R in PRKCZ-expressing OVCAR3 cells.................................................................................................................143
xii
Figure 4-18. Knockdown of IGF1R rescues gene expression of ITGB3 in PRKCZ-expressing cells................................................................................................................144 Figure 4-19. IGF1 increases ITGB3 transcript expression in PRKCZ-expressing SKOV3 cells..................................................................................................................................145 Figure 4-20. IGF1 stimulation decreases IGF1R gene expression in SKOV3 cells........147 Figure 4-21.TIMP-1 gene expression decreases in PRKCZ-expressing ovarian cancer cells..................................................................................................................................148 Figure 4-22. TIMP-1 gene expression regulation is independent of ITGB3 gene expression in SKOV3 cells..............................................................................................149 Figure 4-23. Effect of IGF1 on TIMP-1 transcript expression in SKOV3 cells..................................................................................................................................152 Figure 4-24. Effect of IGF1 signaling on SKOV3 migration as observed by wound healing assay....................................................................................................................153 Figure 4-25. Effects of IGF1 and EGF on migration of SKOV3 as determined by transwell migration assay................................................................................................154 Figure 4-26. Effect of ITGB3 on SKOV3 parental cells migration as observed by wound healing assay....................................................................................................................155 Figure 4-27. Proposed model of ITGB3 transcriptional regulation through IGF1 signaling in PRKCZ-expressing SKOV3 cells................................................................................164 Figure 5-1. Potential Significance of HNF4A in familial ovarian cancer as suggested by Gene Set Analysis (GSA) of gene expression microarray data.......................................180
xiii
LIST OF ABBREVIATIONS aCGH Array comparative genomic hybridization ACHE Acetylcholinesterase ADH4 Alcohol dehydrogenase 4 (class II), pi polypeptide APC Adenomatous polyposis coli AKT v-akt murine thymoma viral oncogene homolog AR Androgen receptor ATM Ataxia telangiectasia mutated AXIN Axin BACH2 Basic leucine zipper transcription factor 2 BCL2 B-cell CLL/lymphoma 2 BIRC3 Baculoviral IAP repeat-containing 3 BRAF v-Raf murine sarcoma viral oncogene homolog B1 BRC BRCA1 Breast cancer 1, early onset BRCA2 Breast cancer 2, early onset BRCT BRCA1 C Terminus BrdU Bromodeoxyuridine CA-125 Cancer antigen-125 CASP8 Caspase 8 CCND1 Cyclin D1 CCNE1 Cyclin E1 hCDC4/FBXW7 F-box and WD repeat domain containing 7 CDK Cyclin-dependent kinases CDKN1B Cyclin-dependent kinase inhibitor 1B (p27, Kip1) CDKN2A Cyclin-dependent kinase inhibitor 2A CGH Comparative genomic hybridization CIMBA Consortium of Investigators of Modifiers of BRCA1/2 CTNNB1 Catenin (cadherin-associated protein), beta 1 CUL1 Cullin 1 CYP11A1 Cytochrome P450, family 11, subfamily A, polypeptide 1 DAPI 4',6-diamidino-2-phenylindole DCC Deleted in colorectal carcinoma EGF Epidermal growth factor EGFR/ERBB1 Epidermal growth factor receptor EMT Epithelial-mesenchymal transition EPHX1 Epoxide hydrolase 1, microsomal (xenobiotic) ER Estrogen receptor ERK Elk-related tyrosine kinase EST Expressed sequence tag EVI1 Ecotropic viral integration site-1 FAT4 FAT tumor suppressor homolog 4 FBS Fetal bovine serum FISH Fluorescence in situ hybridization FGF Fibroblast growth factor
xiv
FTE Fallopian tube epithelial cells GFP Green fluorescent protein GSK3β Glycogen synthase kinase 3 beta GTP Guanosine triphosphate H3 Histone 3 HBOCS Hereditary breast-ovarian cancer syndrome HER2/ERBB2 v-erb-b2 erythroblastic leukemia viral oncogene homolog 2 HGF Hepatocyte growth factor HNF4A Hepatocyte nuclear factor 4, alpha HNPCC Hereditary non-polyposis colorectal cancer HPRT1 Hypoxanthine phosphoribosyltransferase 1 HR Homologous recombination IGF1 Insulin-like growth factor 1 IGF1R Insulin-like growth factor 1 receptor IHC Immunohistochemistry IPA Ingenuity Pathway Analysis ITGB3 Integrin beta 3 c-JUN Jun proto-oncogene KRAS v-Ki-ras2 Kirsten rat sarcoma viral oncogene homolog LBR Lamin B receptor LOH Loss of heterozygosity LPA2 Lysophosphatidic acid receptor 2 LRMP Lymphoid-restricted membrane protein MAGI3 Membrane associated guanylate kinase MAL Mal, T-cell differentiation protein MAPK Mitogen-activated protein kinase MCM5 Minichromosome maintenance complex component 5 MDM2 Mdm2 p53 binding protein homolog MET Met proto-oncogene (hepatocyte growth factor receptor) hMLH1 MutL homolog 1, colon cancer, nonpolyposis type 2 MMP Matrix metallopeptidase hMSH2 MutS homolog 2, colon cancer, nonpolyposis type 1 hMSH6 MutS homolog 6 MRPL19 Mitochondrial ribosomal protein L19 MSP Methylation-specific PCR MTOC Microtubule organization centre mTOR Mechanistic target of rapamycin (serine/threonine kinase) MTT 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide MTUS1 Mitochondrial tumour suppressor 1 MYC v-Myc myelocytomatosis viral oncogene homolog NCOR1 Nuclear receptor co-repressor 1 NEO1 Neogenin 1 NF-κB Nuclear factor kappa B NQO1 NAD(P)H dehydrogenase, quinone 1 OCCR Ovarian cancer cluster region OSE Ovarian surface epithelial cells
xv
OVCAS Ovarian cancer cells P53 (TP53) Tumour protein 53 PAK1 p21 protein (Cdc42/Rac)-activated kinase 1 PARK2 Parkin 2 PDCD4 Programmed cell death 4 PI3K PI3Kinases PIK3AP1 Phosphoinositide-3-kinase adaptor protein 1 PIK3CA Phosphoinositide-3-kinase, catalytic, alpha polypeptide PIP3 Phosphoinositide-3,4,5-trisphosphate hPMS1 Postmeiotic segregation increased 1 hPMS2 Postmeiotic segregation increased 2 PMT Photomultiplier tube PRKCI Protein kinase C iota PRKCZ Protein kinase C zeta PTEN Phosphatase and tensin homolog PTT Protein truncation test RAD51 RAD51 homolog (RecA homolog, E. coli) RB1 Retinoblastoma 1 RBAK RB-associated KRAB zinc finger RBX1 Ring-box 1, E3 ubiquitin protein ligase RHOA Ras homolog gene family, member A RT-PCR Reverse transcriptase polymerase chain reaction RUNX1T1 Runt-related transcription factor 1 SCF Skp, Cullin, F-box containing complex siRNA Small interfering RNA SIRT3 Sirtuin 3 SKP1 S-phase kinase-associated protein 1 SKY Spectral karyotyping SNP Single nucleotide polymorphism SRC v-Src sarcoma SSCP Single strand conformation polymorphism TBP TATA box binding protein TGFβ Transforming growth factor beta TGFβ1 Transforming growth factor beta 1 TMA Tissue microarray TMR Tetramethylrhodamine TUNEL Terminal deoxynucleotidyl transferase dUTP nick end labeling UHN University Health Network, Toronto, ON UPD Uniparental disomy UPS Ubiquitin-proteosome system VNTR Variable number of tandem repeat VTN Vitronectin XPA Xeroderma pigmentosum, complementation group A YWHAZ Tyrosine 3-monooxygenase/tryptophan 5-monooxygenase
activation protein, zeta polypeptide
xvi
CHAPTER 1
Introduction & Literature Review
1
2
1.1 Principles of Molecular Cancer Genetics Major advances have been made over the years in our understanding of different
human diseases, including cancer; however, the underlying mechanisms responsible for
the development of these diseases remain elusive due to the complexity of genetic and
environmental interactions that are involved.
Mutations in cancer cells consist of a wide range of genetic alterations, including
large-scale DNA copy number alterations, chromosomal translocations, amplifications
and deletions, as well as more subtle changes in nucleotide sequences such as point
mutations at positions that are critical for protein functions (1). Additionally, heritable
epigenetic aberrations such as DNA methylation, histone modification, nucleosome
repositioning, and posttranscriptional gene regulation by micro-RNAs may also alter
gene expressions in cells, leading to inappropriate silencing or activation of cancer-
associated genes, and these can occur at various phases of cancer development (1, 2).
Indeed, accumulated evidence suggests that both genetic and epigenetic modifications are
important contributors that can lead to deregulation of molecular pathways responsible
for carcinogenesis.
Genetic aberrations in cancers can be acquired somatically or inherited from one
or both of the parents through the germline, followed by additional acquisition of somatic
mutations. These types of cancer are thus termed “sporadic” or “hereditary”,
respectively. Since the first rate-limiting step of mutation acquisition is bypassed in
hereditary cancers, it often results in multiple cancer types, and the age of onset for these
cancer patients is generally earlier than their sporadic counterparts.
3
As described by Hanahan and Weinberg in their classic review of hallmarks of
cancer, there are six features necessary for a cell to develop a cancer phenotype: self-
sufficiency in growth signals, evasion of apoptosis, insensitivity to growth-inhibitory
signals, limitless replicative potential, sustained angiogenesis, and tissue invasion and
metastasis (3). And in their more recent review, they described two additional hallmarks
of cancer: reprogramming of energy metabolism and evading immune destruction (4)
(Figure 1-1). The identification of genes involved in each of these processes is important
in gaining a better understanding of cancer development and progression. Thus far, the
most well-described cancer progression model is found in colon cancer, whereby
mutations in APC, KRAS, DCC, and p53 are associated with a defined series of stages
from normal colonic mucosa to colorectal carcinoma (5, 6). However, clarification of
this type of cancer progression model is still required for other cancer types. Regardless,
the group of cancer susceptibility genes that play significant roles in tumour initiation and
progression can be generally classified into three main categories: gatekeepers,
caretakers, and landscapers (7).
1.1.1 Gatekeepers
Gatekeepers are genes that can directly regulate tumour cell expansion and
comprise tumour suppressor genes and oncogenes (8, 9). Mutations or deregulations of
these genes can drive the neoplastic process by stimulating cell growth or reducing cell
death.
4
Figure 1-1. Hallmarks of Cancer. Esential alterations in cell physiology important in malignant growth of cancer cells (Adapted from Hanahan and Weinberg, 2000, 2001, ref (3, 4)).
Evading apoptosis Self-sufficiency in
growth signals
Insensitivity to
anti-growth signals
Tissue invasion &
metastasis Sustained
angiogenesis
Evading immune
destruction Limitless replicative
potential
Reprogramming of
energy metabolism
5
1.1.1.1 Tumour Suppressor Genes Tumour suppressor genes are involved in the regulation of cellular growth and
differentiation. Most tumour suppressors act in a recessive manner, meaning that loss or
inactivation of both copies of a gene is required for cellular transformation, a theory
known as “Knudson’s two-hit hypothesis” (10) (Figure 1-2). This concept first emerged
when Knudson performed a statistical analysis of retinoblastoma incidence in children
and observed that the inherited form of retinoblastoma occurs earlier than its sporadic
counterpart (9). This can be explained by the fact that the initiation of a tumour involves
two-rate limiting steps, of which two losses or inactivations of tumour suppressor genes
are required, and that in inherited forms of cancers, the first loss is already present in the
germline. However, recent studies have also suggested that some tumour suppressor
genes may still be able to confer a growth advantage upon a cell when only one allele is
inactivated, a condition known as haploinsufficiency (11).
Loss or inactivation of a tumour suppressor gene can occur through different
mechanisms, including point mutation, deletion, mitotic recombination, and
chromosomal loss (12). Additionally, epigenetic alterations can also decrease the
expression of these genes without alteration of their underlying DNA sequence, through
cancer-specific CpG island hypermethylation, in combination with repressive histone tail
modification (13).
6
Figure 1-2 Knudson's two-hit hypothesis. Normal individuals have two normal copies of a tumour suppressor gene, thus two independent mutational events “hits” are required for a cell to initiate cancer. However, individuals with inherited germline mutation already have the first “hit” in every cell, thus only one additional mutation is required to initiate cancer. (Adapted from Richards, 2001, ref (14)).
Normal allele
Inherited germline mutation
Deletion Normal Inherited mutation
7
1.1.1.2 Oncogenes
Proto-oncogenes encode proteins that are normally involved in the stimulation of
cell division; however, when these genes are genetically altered, they are capable of
causing cellular transformation and uncontrolled proliferation (15). Oncogenes (proto-
oncogenes with gain-of-function mutations) are considered to be dominant transforming
genes because changes in only one allele of the gene are usually sufficient to confer a
selective growth advantage on the cell. These genes are frequently activated by gain of
function mutations or fusions with other genes, or are abnormally over-expressed due to
gene amplification, increased promoter activity, or protein stabilization (16). Many
oncogenes have been identified and they all typically act through three biochemical
mechanisms: protein phosphorylation (e.g. Raf kinase), transcription regulation (e.g.
Myc), and signal transmission (e.g. Ras) (17-19).
1.1.2 Caretakers
Caretaker genes (or stability genes) function in maintaining genomic integrity of
the cell by regulating DNA repair, chromosome segregation, and cell cycle checkpoints
(8). Some well-characterized caretaker genes include hMLH1, hMSH2, ATM, XPA,
BRCA1 and BRCA2, all of which play important roles in DNA repair (20). When
deregulated, these genes can indirectly promote neoplastic transformation by contributing
to an accumulation of mutations in oncogenes and tumour suppressor genes (8, 21). The
importance of these genes is affirmed in that their mutations can lead to a variety of
cancer-prone chromosomal instability disorders, including ataxia telangiectasia,
8
Nijmegen breakage syndrome, hereditary nonpolyposis colorectal cancer, Fanconi
anemia, Li-Fraumeni, and hereditary breast/ovarian cancer (21).
Caretaker genes are also labelled as tumour suppressor genes because inactivation
of both of their alleles is required for a pathological phenotype to manifest. However, it
is interesting to note that while the risk of cancer in families with inherited mutations of
caretaker genes is less than the risk in families with inherited defects in a gatekeeper
gene, the most common forms of hereditary cancer dispositions, such as colon and breast
cancers, are caused by inherited mutations of caretaker genes rather than gatekeepers (8,
20).
1.1.3 Landscapers
Landscapers are genes that do not directly regulate cellular growth, but are instead
tumour modifiers that can provide an abnormal stromal environment to increase a cell’s
ability to transform (22). As such, landscaper genes can be classified as tumour
suppressor genes but they act on the tumour microenvironment rather than the tumour
itself. Stromal-epithelial interactions are important in the regulation of tissue
homeostasis, and disruptions of these interactions may lead to tumour formation, by
increasing proliferation and transdifferentiation of fibroblasts, infiltration and activation
of inflammatory cells, induction of angiogenesis and altered deposition and degradation
of the extracellular matrix (23).
It is controversially speculated that genetic alterations in either the epithelial or
the stromal cells can lead to altered stromal-epithelial interactions to promote
tumourigenesis, and there are evidence suggesting that inherited or acquired genetic
9
alterations in stromal cells may give rise to cancer. For example, loss of heterozygosity
(LOH) in stromal cells has been described in various types of cancer, including breast,
ovarian, colon, lung, and head and neck carcinomas (24). Mutations of tumour suppressor
genes (eg. p53, PTEN) and oncogenes (eg. EGFR) have also been reported in the stromal
cells of breast carcinomas, and the frequency of these genetic changes differ between
sporadic and hereditary tumours, implying that these changes are specific to tumour
subtypes (25-27). In addition to LOH and mutations, epigenetic alterations such as DNA
methylation have also been found in tumour-associated stromal fibroblasts from
neoplastic human breast and prostate carcinomas (24). However, as mentioned above,
this remains controversial and the potential mechanisms responsible for these alterations
will require further investigations.
As mentioned earlier, genetic mutations are the driving force for cancer
development, by promoting tumour initiation and progression. At the cellular level, it is
a multistep process in which mutations can lead to a selection of cells that have acquired
advanced proliferation, survival, and metastatic potentials. Both mutational inactivation
of stability genes and tumour suppressor genes, and mutational activation of proto-
oncogenes to oncogenes, as well as additional intrinsic and extrinsic genetic changes, are
important for these acquisitions.
1.2 Ovarian Cancer
Ovarian cancer is the fifth most common form of cancer in women and it is the
leading cause of death in patients with gynaecologic malignancies in North America (28).
10
In Canada alone, approximately 2500 new cases of ovarian cancer and 1750 deaths due to
this disease were estimated for 2009 (29). The lack of any obvious symptoms that would
indicate an early stage of this disease is responsible for this high mortality rate; in fact,
approximately 70% of the patients are diagnosed in an advanced stage after the cancer
has metastasized beyond the ovaries (30, 31).
The lack of effective screening methods also contributes to the late diagnosis of
ovarian cancer. Measurement of serum cancer antigen-125 (CA-125) is the primary
screening method used for ovarian cancer; however, CA-125 concentration does not have
the sensitivity nor specificity required for accurate detection, since its level is typically
low in early stages of ovarian cancer and it can also be elevated in other gynaecological
conditions, such as endometriosis, pregnancy, adenomyosis and polycystic ovarian
syndrome (32). To increase detection accuracy, transvaginal ultrasonography (TVS) is
often used in parallel with a CA-125 serum test; however, the specificity achieved using
the combination of these two methods remains sub-optimal (33). After being diagnosed,
ovarian cancer patients are often treated with aggressive surgery followed by
combination chemotherapy, but despite high initial response rates to treatment, survival
of patients stands at just 45% at 5 years, with most of the patients eventually relapsing
and succumbing to this disease (31, 34).
There are three main types of ovarian tumours: epithelial, germ cell, and stromal
tumours. Epithelial tumours are derived from cells that cover the surface of the ovary,
germ cell tumours are derived from cells that produce the ova, and stromal tumours arise
from the connective tissues that hold the ovaries together (35, 36). Of all of these tumour
types, epithelial tumours are the predominant type, accounting for 90% of total cases of
11
ovarian cancer (35). Epithelial ovarian cancer is further subdivided into different
histological types, including serous, mucinous, endometrioid, and clear cell; of these
subtypes, serous carcinomas constitute the majority of ovarian carcinomas (35). Each of
these subtypes has unique morphology, as well as biological and genetic backgrounds,
and these differences imply that patients with different subtypes of ovarian cancer can
have different responses to the same treatment, as well as different prognosis of the
disease (37). Therefore, it is important to understand the molecular pathogenesis for
each type of ovarian carcinoma in order to develop effective screening methods and
treatment options for each of the patients.
1.2.1 Model of Ovarian Carcinogenesis
In a carcinogenesis model proposed by Shih and Kurmen (2004), epithelial
ovarian cancers are categorized into Type I and Type II tumours corresponding to two
main pathways of ovarian tumourigenesis (38) (Table 1-1). Type I tumours arise in a
stepwise manner from borderline tumours, and are composed of low-grade serous
carcinomas, mucinous, endometrioid, and clear cell carcinomas; Type II tumours, on the
other hand, arise de novo and include high-grade serous carcinoma, malignant mixed
mesodermal tumours (carcinosarcomas), and undifferentiated carcinomas (38).
The tumourigenic pathway of Type I tumours resembles the adenoma-carcinoma
progression in colorectal cancer in that these tumours tend to evolve slowly and are
associated with distinct molecular changes during tumour development, whereas Type II
tumours are highly aggressive, and are often characterized by genomic instability, as
shown by genome-wide changes in DNA copy number (33, 38, 39). These two types of
12
tumours have distinct morphological and molecular signatures; Type I tumours often
contain mutations in BRAF, KRAS, PTEN and CTNNB1 and Type II tumours frequently
harbour p53 mutations (38). It should be noted, however, that while the majority of high-
grade serous carcinomas – the most prevalent form of ovarian cancer – arise
independently from low-grade tumours, there are rare cases of high-grade carcinomas
that have progressed from atypical proliferative serous (borderline) tumours (40) (Figure
1-3).
13
Table 1-1. Characteristics of Type I and Type II ovarian tumours. Type I and Type II ovarian cancers have distinct morphological and molecular characteristics. Type I tumours evolve slowly and are associated with distinct molecular changes while most of Type II tumours arise rapidly and are highly aggressive.
Type I Type II
Tumour Type - micropapillary serous carcinoma
- mucinous
- endometrioid
- clear cell
- serous carcinomas
- malignant mixed mesodermal tumours (carcinosarcomas)
- undifferentiated carcinomas
Mutations - KRAS
- BRAF
- PTEN
- β-catenin (CTNNB1)
p53
Chromosomal Instability
gradual increase high
Confined to ovary yes no
14
Figure 1-3. Dualistic model of ovarian serous carcinoma development. Ovarian serous carcinomas, the most common type of ovarian cancer, are proposed to arise through either Type I or Type II pathway. Low-grade tumours develop in a stepwise manner and are often associated with KRAS or BRAF mutations. High-grade tumours develop directly from the ovarian surface epithelium or inclusion cysts without distinct intermediate morphological stages. High-grade serous carcinomas frequently harbour p53 mutations, and are associated with chromosomal instability. (Adapted from Shih and Kurmen, 2004, ref (38)).
?
15
1.2.2 Familial Ovarian Cancer
There are a number of well established risk factors associated with ovarian
cancer. Factors that lower the number of lifetime ovulations in an individual have been
shown to decrease her risk of developing ovarian cancer; these factors include the use of
oral contraceptive pills, multiparity, breast-feeding, and oophorectomy (41, 42).
Likewise, there are factors that have been shown to increase the risk of developing
ovarian cancer, including old age, early menarche, late menopause, high dietary fat
intake, and use of estrogen-replacement therapy (43). However, the single most
important risk factor for ovarian cancer is family history (44). In fact, it is estimated that
familial ovarian cancer accounts for 5-15% of the total cases of ovarian cancer (44). The
risk of developing ovarian cancer is 1.6% of women in the general population; however,
if a woman has one or two first-degree relative(s) with ovarian cancer, her risk increases
to 4% and 7%, respectively (45). Notably, the age of onset for ovarian cancer patients
with a family history is earlier than those with no family history, with an average age of
53.5 and 60.8, respectively (46).
Three main hereditary syndromes that predispose to ovarian cancer, including
hereditary site-specific ovarian cancer, hereditary breast-ovarian cancer syndrome
(HBOCS), both of which are due to mutations in the tumour suppressor genes BRCA1
and BRCA2, and hereditary non-polyposis colorectal cancer (HNPCC; Lynch Syndrome),
which is mainly due to mutations in DNA mismatch repair genes such as hMSH2,
hMLH1, hMSH6, and PMS2 (47, 48). Another group of minor familial syndromes also
predispose individuals to ovarian cancer, accounting for <1% of total cases, including
16
Gorlin’s syndrome, osteochondromatosis or Ollier’s syndrome, and the Peutz-Jeghers
syndrome (47, 49-51).
1.2.2.1 Hereditary Site-Specific Ovarian Cancer and Hereditary Breast-Ovarian
Cancer Syndrome (HBOCS)
Site-specific ovarian cancer syndrome is identified in families in which two or
more first- or first- and second-degree relatives are affected by epithelial-type ovarian
cancer, without being affected by breast cancer (45). While it has previously been
speculated that site-specific ovarian cancer represents a unique syndrome, it has been
reported that it may in fact be a variant of HBOCS with a high prevalence of ovarian
cancer, since no susceptibility gene has been identified specifically for ovarian cancer
(52).
As its name implies, hereditary breast-ovarian cancer syndrome is identified in
families in which both breast and ovarian cancer are common. This syndrome is
characterized by early-onset breast cancer, ovarian cancer at any age, bilateral breast
cancer, breast and ovarian cancer in the same individual, or male breast cancer (45).
Genetic predisposition is suggested by early onset, as well as multiple cases of these
cancers within the same family. Thus far, the two well-studied high penetrance
susceptibility genes found to be associated with ovarian cancer are BRCA1 and BRCA2,
both of which lead to autosomal dominant inheritance of susceptibility (45). The
importance of these two genes in ovarian cancer will be discussed further in this chapter.
17
1.2.2.2 Hereditary Nonpolyposis Colorectal Cancer (HNPCC)
Familial ovarian cancer may also occur in individuals with hereditary
nonpolyposis colon cancer syndrome, which is characterized by an autosomal dominant
inheritance of colonic cancer in the absence of colonic polyposis (53). In addition to
ovarian cancer, individuals within HNPCC families also have higher risk for other
cancers such as endometrial, uro-genital, pancreatic and biliary tract cancers (45). Over
70% of the mutation carriers among HNPCC families have mutations within the DNA
mismatch repair (MMR) hMLH1 and hMSH2 genes. Mutations in hPMS1, hPMS2 and
hMSH6 are not as prevalent, but also have high penetrance (54-56). All of these MMR
genes are important for the repair of nucleotide mismatch during DNA replication to
prevent propagation of potentially harmful mutations (57). The cumulative risk for
colorectal and ovarian cancer for carriers of MMR gene mutations from HNPCC families
is estimated to be 80% and 12%, respectively (52).
1.2.3 Molecular Pathology and Genetics of Ovarian Cancer
Ovarian cancer is a heterogeneous disease with different histological grades and
subtypes, each encompassing a distinct, though not necessarily unique, set of molecular
genetic attributes. Over the years, extensive research has been conducted to identify
these genetic aberrations in order to improve our understanding of ovarian cancer
pathogenesis, ovarian tumour classification, as well as to develop personalized therapies
that target specific defects in the tumour cells of patients.
18
1.2.3.1 p53
One of the genes most consistently associated with ovarian pathogenesis is p53, a
well-defined tumour suppressor gene. p53 encodes a DNA-binding protein that responds
to external cues and insults and is responsible for the transcriptional regulation of genes
involved in cell cycle control, DNA repair, and apoptosis of damaged cells (58).
Mutations that lead to loss of p53 function result in failure to activate responses, thus
leading to unrepaired genetic damage and increased chromosomal instability (59). It is
the most commonly mutated gene in human cancer and its mutations are observed in
more than 50% of high-grade ovarian serous carcinomas, but p53 mutation is a rare event
in low-grade tumours, which supports the idea that different pathogenic pathways are
responsible for high- and low-grade serous carcinomas (58, 60). It has also been
suggested that p53 mutation is an early event in the development of high-grade
carcinomas, as mutation of this gene has been observed in normal-appearing epithelium
and dysplastic epithelium within inclusion cysts next to the tumour that has the same
mutation (58).
1.2.3.2 Wnt-Signalling Pathway
Missense mutations of CTNNB1, which codes for β-catenin and maps to
chromosome 3p21, are observed in approximately 30% of endometrioid adenocarcinomas
(61). β-catenin is an important player within the canonical Wnt signalling pathway
known to be involved in various cellular processes, including regulation of cell fate,
proliferation, motility, and survival; indeed, constitutive activation of Wnt signalling is
often observed in endometrioid ovarian cancer (39, 62). Mutations of CTNNB1 often
19
result in alterations in its protein residues such that it is no longer subjected to
phosphorylation by its upstream regulator GSK3β (glycogen synthase kinase 3 beta), thus
leading to stabilization of β-catenin (39). This stabilization allows for an accumulation of
β-catenin to constitutively activate the transcription of downstream target genes that are
important for neoplastic transformation and tumour progression (39). In addition to
CTNNB1, other defects within the Wnt pathway are also observed in ovarian
endometrioid tumours, including mutations of APC, AXIN1 and AXIN2, which all encode
components of a protein complex involved in the regulation of β-catenin (63).
1.2.3.3 PI3K/Akt Signalling Pathway
The PTEN tumour suppressor gene is also commonly mutated in endometrioid
carcinomas, with a frequency of 30-80% (64). Moreover, it is located on chromosome
10q23, a region that is lost in 20-30% of ovarian cancers of this specific subtype (64).
PTEN is one of the key regulators within the PI3K/Akt signalling pathway, as it can
dephosphorylate the plasma membrane lipid second messenger phosphoinositide-3,4,5-
trisphosphate (PIP3) generated by PI3Kinases (PI3K) to PIP2, leading to inhibition of
this signalling cascade (39). Inactivating mutations of PTEN can therefore enhance the
activation of PI3K/Akt pathway, leading to uncontrolled cell cycle progression, cell
survival, cell motility, and angiogenesis (39). Activating mutations are also observed in
PIK3CA, a potential ovarian oncogene that encodes the p110α catalytic subunit of PI3K
(39). While mutations of PIK3CA are only found in approximately 2% of ovarian serous
carcinomas, it is a more common event in endometrioid and clear cell subtypes, with a
frequency of approximately 20% (65). Interestingly, amplifications of PIK3CA (>7
20
fold) are observed across all histological subtypes (24.5%) and are inversely correlated
with gene mutations (65). Additionally, amplification of PIK3CA has also been
associated with chemotherapy resistance in ovarian cancer patients (66).
The PIK3CA downstream target gene AKT2 has also been examined as a potential
ovarian cancer oncogene. It encodes a serine-threonine kinase that is able to
phosphorylate a variety of proteins, including Ezrin, a protein with roles in cell adhesion,
migration, and organization (67). High-level amplification (>3 fold) of AKT2, but not its
related genes AKT1 and AKT3, has been observed in high-grade ovarian serous
carcinoma; this specific amplification suggests an important role of AKT2 in ovarian
carcinogenesis (48, 68). Amplification/over-expression of this gene has been shown to
have a statistically significant association with higher grade tumours and poorer survival
(69). Interestingly, AKT2 over-expression and a loss of PTEN expression function
synergistically to promote metastasis in colorectal cancer (70). While this has yet to be
examined, it is possible that the same may be observed in ovarian cancer, since loss of
PTEN expression has also been found in a subset of ovarian tumours, as mentioned
above.
1.2.3.4 MAPK Signalling Pathway
The RAS family of G proteins belongs to several signalling pathways, including
the well-studied MAPK (mitogen-activated protein kinase) signalling cascade. Their
function involves coupling membrane receptor kinases to intracellular signalling cascades
through their GTPase activity, and they are critical players in the regulation of cellular
proliferation (71). Point mutations within codons 12, 13, or 61 of KRAS, which result in
21
a constitutively activated protein, are important for ovarian carcinogenesis (72). While
KRAS mutations are rare in invasive serous epithelial ovarian cancers, they do occur
frequently in mucinous ovarian cancers (50%), and are common events in serous
borderline ovarian tumours (33%) and low-grade serous carcinomas (35%) (58, 73).
Activating mutations of BRAF within codon 599 are also often observed in
borderline and low grade serous tumours, with similar frequencies as KRAS mutations
(58). BRAF is a RAF family protein, and is activated by RAS to stimulate the MAPK
signalling cascade. Interestingly, KRAS and BRAF mutations are not found in the same
tumour, which may be explained by their closely related functions in the pathway (58).
Also, the observed differences in KRAS and BRAF mutation status between histological
subtypes indicate that the MAPK pathway may play a major role in pathogenesis in
certain but not all types of ovarian cancer.
1.2.3.5 Cell Cycle Genes
Correct control of the cell cycle is critical in the regulation of cell proliferation,
and it requires proper expression of various regulatory proteins, including cyclins and
cyclin-dependent kinases (CDK). Cyclins function in activating CDKs, which
subsequently phosphorylate and activate key proteins to allow cell cycle progression. In
human cancers, CCND1, which encodes cyclin D1, is most frequently over-expressed.
(74). Cyclin D1 forms a complex with CDK4/6, which can then phosphorylate and
activate the retinoblastoma protein (RB), leading to the release of E2F, triggering G1 cell
cycle progression (75) . While amplification of CCND1 is rarely observed in ovarian
cancer, it is found to be over-expressed in a subset of ovarian tumours, mostly in low-
22
grade serous carcinomas and mucinous tumours, and is associated with decreased
survival of patients (58, 76, 77). Over-expression of CDK4 has also been observed in
approximately 15% of ovarian cancers and is reported to be correlated with increased
expression of CCND1; similarly to CCND1, gene amplification of CDK4 is a rarely
observed (78).
p16 acts as a negative regulator of the cell cycle by inhibiting the kinase activity
of the cyclin D-CDK4 complex (75). Loss of p16 mRNA is observed in ovarian cancer,
mostly in serous, mucinous, and endometrioid carcinomas (79). It has been proposed that
the decrease in p16 gene expression is due to hypermethylation at its 5’-CpG island, since
mutation and deletion of this gene are uncommon (79). It is interesting to note that the
lack of p16 expression occurs more frequently in ovarian tumours lacking p53 mutations,
suggesting that p53 inactivation may not be as important in ovarian tumour development
when another G1 cell cycle regulatory gene has already been inactivated (78).
The cyclin E-CDK2 complex is important for the progression of G1-S phase of
the cell cycle, and over-expression of cyclin E, which is encoded by the gene CCNE1,
has been found in various human cancers, including ovarian cancer (75). Evidence
suggests that this increase in gene expression is partially due to gene amplification and
that CCNE1 expression is correlated with CDK2 expression (80). CCNE1 over-
expression may be involved in malignant progression of epithelial ovarian cancer, as its
expression is highest among malignant tumours (~70%), compared with borderline
(~48%) and benign tumours (~9%) (81). Furthermore, cyclin E is negatively regulated
by p21 and p27, both of which are expressed at low levels in high-grade serous ovarian
23
carcinomas (82). Notably, the potential of cyclin E as a prognostic marker has also been
suggested, since its over-expression is associated with poor disease outcome (75).
1.2.3.6 Epidermal Growth Factor Family Receptors
In the presence of ligand, epidermal growth factor family receptors can form
homodimers and heterodimers to initiate intracellular signalling pathways that are
important for cell proliferation and tumour growth (58). EGFR1/ERBB1 is a member of
the type I tyrosine kinase receptor family HER (i.e., ERBB) and is expressed in normal
ovarian surface epithelium (83). While this gene is rarely mutated or amplified, it is
often over-expressed in ovarian cancer, presenting in 35-70% of all cases, and is
associated with poor prognosis, as well as with drug resistance (82-84).
Another HER family member, HER2/neu (c-ERBB2), is amplified in ~6-18% of
ovarian tumours, and its increased copy number is associated with poor prognosis (84-
88). Additionally, HER2 over-expression is found in 20-30% of ovarian cancers (48).
Specifically, it is frequently observed in serous ovarian carcinomas, and tumours
associated with advanced stages, late age at diagnosis, and differentiation (89). HER2
over-expression can lead to its dimerization, even without the presence of a ligand, which
leads to constitutive activation of its signalling pathway (90). Moreover, it was found
that HER2 is over-expressed in 40% of HBOCS, demonstrating its importance within the
subset of ovarian cancer patients with a family history of breast and/or ovarian cancer
(48).
24
1.2.3.7 Estrogen Receptors
Estrogens are involved in normal cellular processes such as growth,
differentiation, and physiology of the reproductive system; however, they can also
participate in the progression of hormone-dependent cancers, including ovarian cancer
(91). Estrogen receptor alpha (ER-α) is a member of the superfamily of nuclear receptors
that transduce hormone signals. It is a ligand-activated transcription factor that can
affect normal or cancer cells upon stimulation with estrogens (92). Amplification of the
ESR1 gene is a rare event in ovarian cancer, accounting for only 2.1% of the cases;
however, ESR1 has been found to be over-expressed in 25-86% of ovarian cancer cases
(93). Estrogen receptor beta (ER-β), on the other hand, is under-expressed in a subset of
ovarian cancers, with malignant tumours having the lowest expression (94). This
opposite expression of ER-α and ER-β observed in ovarian cancers may be correlated
with their roles in regulation of cyclin D1, since ER-α stimulates the gene expression of
cyclin D1, whereas ER- β suppresses its expression (95).
As stated earlier, ovarian cancer is a heterogeneous disease with various types of
genetic changes. And while the above-mentioned genetic alterations have all been shown
to play significant roles in the pathology of ovarian cancer, the most well characterized
genetic risk factors genes associated with ovarian cancer are evidently BRCA1 and
BRCA2.
25
1.2.4 BRCA1 and BRCA2
In the early 1990’s, Hall and colleagues identified a linkage between chromosome
17q and site-specific breast cancer arising in young women in certain cancer prone
families in an unusually large proportion compared to the general population (96).
Shortly after, Narod et al. showed that this same genetic locus was linked to HBOCS
(97). The BRCA1 gene was subsequently cloned by Miki and colleagues in 1994,
followed by the identification of another breast and ovarian cancer susceptibility gene
appropriately named BRCA2, by Wooster et al. (98, 99).
Both BRCA1 and BRCA2 are expressed in a variety of tissues, particularly
during S and G2 phases of the cell cycle (100). They play major roles in DNA repair
mechanisms, by participating in homologous recombination (HR) in the presence of
double-strand breaks, transcription-coupled repair during oxidative damage, and possibly
in non-homologous end joining (100). Additionally, they are involved in the control of
cell cycle checkpoints, protein ubiquitination, and chromatin remodelling (101).
BRCA1 is located on chromosome 17q21 and it comprises 24 exons encoding a
protein of 1863 amino acids (98). BRCA1 contains a highly conserved N-terminal RING
domain and a tandem of two BRCT (BRCA1 C Terminus) domains at its carboxyl
terminus, which are important for protein ubiquitylation and phosphorylated-protein
binding, respectively (102, 103). These domains are very important for its protein
function, as demonstrated by the observations that these regions are often targets of
clinically important mutations (102).
BRCA2 is located on chromosome 13q12-13 and is made up of 27 exons encoding
a protein of 3418 amino acids (30). BRCA2 contains eight BRC repeats that are involved
26
in protein binding, including recombinase RAD51, to function in DNA double strand
repair (104).
1.2.4.1 BRCA1 and BRCA2 in Ovarian Cancer
Approximately 90% of all the cases of hereditary ovarian cancer are due to
mutations within the BRCA1 or BRCA2 genes (34). Indeed, for those who have inherited
mutations within BRCA1 or BRCA2, their lifetime risks for developing ovarian cancer are
20-40% and 15-25%, respectively, compared to a lifetime risk of 1.6% in the general
population (45, 105).
While one defective copy of either BRCA1 or BRCA2 in the germline is enough to
increase cancer predisposition of an individual, it is often observed that the second allele
is also lost in tumour cells isolated from predisposed patients (106). This “second hit”
often occurs through LOH. In fact, BRCA1 and BRCA2 have reported LOH frequencies
of 80% and 70%, respectively, in mutation carriers in both breast and ovarian tumours
(107, 108). While BRCA promoter hypermethylation occurs in sporadic breast and
ovarian cancer, it is a rare event in mutation carriers (109, 110). In addition to LOH and
promoter methylation, other epigenetic or transcription silencing of BRCA1 and BRCA2,
such as chromatin-mediated repression, and other yet unknown environmental factor(s)
may affect BRCA expression at the gene level.
27
1.2.4.2 Mutations of BRCA1 and BRCA2 in Ovarian Cancer
The type of mutations within BRCA1 and BRCA2 varies, with the majority of
them being small insertions or deletions resulting in frameshift, nonsense mutations or
splice site alterations, all of which may cause truncation of the protein (30). Missense
mutations resulting in dysfunctional proteins have also been reported, along with
mutations that involve non-synonymous coding changes or in-frame deletions (30). In
recent years, with the introduction of high-throughput DNA-based techniques, it has also
been discovered that large genomic deletions and rearrangements can occur; however,
large genomic alterations are far less common in BRCA2 compared to BRCA1 (111-114).
It has been reported that up to 75% of hereditary ovarian cancer families have a
mutated BRCA1 (115), and while BRCA2 mutations are found within 35% of hereditary
breast cancers, they confer a lower risk in ovarian cancer, accounting for 10-20% of
hereditary ovarian cancers (105). Despite the high frequency of BRCA gene mutations in
breast and ovarian cancer, there are no specific mutational hotspots, as mutations are
distributed throughout the whole gene (45). However, there are reports that showed an
association between the sites of mutation and ovarian cancer risk. In BRCA1, mutations
within nucleotides 2401 and 4190 (named “high risk region for ovarian cancer”) were
shown to result in increased risk of ovarian cancer while decreasing the risk of breast
cancer (30) (Figure 1-4). Likewise, mutations between nucleotides 4075 and 6503 of
BRCA2, termed “Ovarian Cancer Cluster Region (OCCR)”, also contribute to an increase
in ovarian cancer risk (116-118).
28
Figure 1-4. Gene structure of BRCA1 and BRCA2. Coloured boxes represent exons
with corresponding exon number below each box. Common founder mutations are
indicated, as well as mutational regions associated with increased risk of ovarian cancer
(Adapted from Russo et al., 2009, ref (45))
BRCA1
BRCA2
High risk region for ovarian cancer
Ovarian Cancer Cluster Region (OCCR)
2401 4190
4075 6503
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
185delAG ins6kb 5382inC
999del5 617delT
29
A catalogue of BRCA1 and BRCA2 mutations can be found at the Breast-Cancer
Information Core (BIC) database (http://research.nhgri.nih.gov/bic/), an international
collaboration hosted by the NIH National Human Genome Research Institute. As of
2009, there were approximately 12,000 carriers of a BRCA1 mutation or unclassified
variant and approximately 11,000 BRCA2 carriers recorded (30).
1.2.4.3 Specific BRCA1 and BRCA2 mutations in closed populations
Highly penetrant germline BRCA mutations are rare, with an approximate
frequency of 1 in 500 individuals in most population. However, for certain relatively
closed and geographically confined populations, the frequency of BRCA mutations
increases dramatically. The Ashkenazi Jewish population in North America and Israel,
for example, has a BRCA mutation frequency of 1 in 40 (119). Numerous genetic studies
have focused on this population and specific “founder mutations” have been identified
within these individuals, including 183delAG and 5382inC in the BRCA1 gene and
6174delT in the BRCA2 gene (48). These three mutations alone account for 98-99% of
identified mutations in this population and screening for these mutations has become a
common clinical practice for individuals with Ashkenazi Jewish background (30).
Other groups have also implemented specific genetic screenings to identify
founder mutations that are prevalent in their respective populations in individuals with a
family history of breast and/or ovarian cancer. For example, most of the cases of familial
breast and/or ovarian cancer within the Icelandic population are due to a 999del5
mutation within the BRCA2 gene, while Eastern European countries such as Russia,
Poland, and Hungary have a high frequency of 5382inC BRCA1 mutation as well as a few
30
other common mutations (30). In North America, French Canadian founder mutations
C4446T within BRCA1, as well as G6085T and 8765delAG within BRCA2 have also
been identified (120, 121). The identification of these founder mutations has facilitated
the development of effective genetic screening programs that allow identification of high
risk individuals for appropriate surveillance and informed treatment strategies.
1.2.4.4 Modifiers of BRCA1 and BRCA2
It is likely that both genetic and environmental factors can affect the penetrance of
BRCA gene mutations for ovarian cancer. Indeed, previous studies have attempted to
identify such factors. For example, in a study conducted within the Ashkenazi Jewish
population, a functional single nucleotide polymorphism in the promoter of the MDM2
gene was associated with an increased risk of breast and/or ovarian cancer among BRCA
carriers diagnosed with either one of these cancers before or at the age of 51 (122). And
in a Polish population study, it was found that the Leu33Pro polymorphism within the
ITGB3 gene increased the risk of BRCA1-associated ovarian cancer but not breast cancer
(123); however, subsequent genotyping of the ITGB3 gene in 9998 BRCA1 and 5544
BRCA2 carriers from 34 studies from CIMBA (Consortium of Investigators of Modifiers
of BRCA1/2) showed that when the Polish population is excluded from analysis, the
Leu33Pro polymorphism is no longer associated with the increased risk of ovarian cancer
(124), suggesting that this ITGB3 polymorphism may only be important in a subset of
BRCA1 carriers. Interestingly, a recent genome-wide association study conducted by
Ramus et al. identified 9p22.2 as a novel ovarian cancer susceptibility locus, as a rare
allele of this locus (rs3814113) was found to be associated with a decreased ovarian
31
cancer risk in both BRCA1 and BRCA2 carriers (125). This is a significant finding as it is
the first confirmed common genetic variant that has been associated with reduced ovarian
cancer risk for either BRCA1 or BRCA2 carriers. However, mechanisms of how this SNP
may alter the risk of ovarian cancer in BRCA carriers remain to be investigated.
In terms of non-genetic/environmental factors, oral contraceptives, as well as
tubal ligation appear to significantly lower the risk of ovarian cancer in women with
BRCA germline mutations, as they do in the general population (126, 127). However,
while the discovery of genetic and environmental modifiers of BRCA1 and BRCA2 has
begun to emerge in recent years, the number of identified modifiers is still limited.
It appears that the development of ovarian cancer is a result of a multi-step
process involving an accumulation of genetic alterations. For people with family history
of ovarian cancer, these alterations could be inherited. In addition to high-penetrance
susceptibility genes BRCA1 and BRCA2, it is plausible that changes in low-penetrance
genes can increase an individual’s risk for ovarian cancer. Therefore, the identification
of these risk-altering genes may be useful in assessing patients with a family history of
breast and/or ovarian cancer.
1.3 Ovarian Cancer Genome-Wide Association Studies
The development of genomic technologies such as array comparative genome
hybridization (aCGH), single nucleotide polymorphism (SNP) arrays and microarray
expression profiling has allowed the elucidation of many important genetic events that
occur in cancer development. Their ability to simultaneously measure thousands of
genes not only allows researchers to identify individual genes but also to identify
32
biological pathways that may be important in cancer development, including ovarian
cancer.
1.3.1 Array Comparative Genomic Hybridization of Ovarian Cancer
In addition to mutations at the nucleotide level, altering DNA copy number can
also affect gene expression and function of a cell. Certain DNA amplifications allow
cancer cells to increase expression of critical genes such as oncogenes involved in growth
regulation and genes responsible for drug resistance, while DNA deletions can cause a
decrease in expression of tumour suppressor genes. Researchers have employed various
cytogenetic methods such as fluorescence in situ hybridization (FISH), spectral
karyotyping (SKY), and conventional comparative genomic hybridization (cCGH) to
identify such aberrations in various diseases, including ovarian cancer; however, high
throughput array CGH (aCGH) has been used widely in recent years due its ability to
refine copy number alterations at a much higher resolution.
A study conducted by Lambros and colleagues identified regions of gains and
losses in 23 different ovarian cancer cell lines using aCGH, including those regions that
have been previously reported using other conventional methods, such as loss of
chromosome 4 or 4q, loss of 18q, and gain of 20 or 20q (128). Additional genomic
changes were detected in the study, including two regions of amplification on
chromosome 11q13 containing the cyclin D1 gene and candidate oncogene PAK1, as well
as amplification of 11q22 near the progesterone receptor gene, and a locus containing a
cluster of matrix metalloproteinase genes (128).
33
This approach can be applied to ovarian tumours in addition to cell lines. For
example, characteristic patterns of copy number changes between histological subtypes
of ovarian tumours were identified using aCGH (Figure 1-5) (129). Every histological
subtype exhibited multiple copy number gains; however, it was found that their regions
of amplification differ, and that serous carcinomas have the largest number of alterations
(129). The authors of this study suggested that this observation is consistent with the
hypothesis that borderline ovarian tumours can progress to serous carcinomas, through
additional accumulation of genetic alterations (129). Differential patterns of recurrent
copy number alterations were also identified in sporadic and BRCA1-mutated ovarian
tumours by aCGH. As revealed by Leunen and colleagues, BRCA1 ovarian tumours
exhibited a greater number of losses, and that deleted regions are longer than those found
in the sporadic tumours, with a median length of 5.2 Mb vs. 0.2 Mb, respectively,
indicating that this major loss of genetic material in BRCA1 patients contributes to the
genetic instability of the tumours (130).
The application of aCGH also has diagnostic potential in clinical settings. For
example, because the genetic profiles of ovarian and endometrial tumours are presumably
different from each other, aCGH has been suggested to be used as a complementary tool
in distinguishing metastases from these two types of tumours to improve accurate
diagnosis (131). Additionally, regions associated with chemotherapy resistance in early-
stage epithelial ovarian cancer and late stage ovarian serous carcinomas have also been
identified using this method (132, 133), suggesting that aCGH may be used as a
screening tool, as the genetic profiles of the tumours may reveal whether or not the
patients can benefit from various therapeutic regimens.
34
1.3.2 Single Nucleotide Polymorphism Array Analysis of Ovarian Cancer
The more recent development of single nucleotide polymorphism (SNP) arrays
has revealed additional information regarding the various types of chromosomal
aberrations that may occur in cancer. In addition to increased resolution, SNP arrays
differ from aCGH in that they can also detect the allele ratio of a single DNA sample,
hence allowing them to detect copy neutral loss of heterozygosity events such as
uniparental disomy (UPD), a genetic alteration resulting from deletion of one allele and
duplication of the second allele.
In fact, SNP arrays have been shown to be a useful tool in the identification of
such genetic alterations in ovarian cancer. For example, in their ovarian cancer SNP
array analysis, Walsh and colleagues have shown that BRCA-associated ovarian tumours
exhibit a greater frequency of amplification and LOH compared to sporadic ovarian
tumours, and that this increase in LOH is mostly due to UPD rather than deletion (134).
This observation of increased chromosomal instability in BRCA-associated tumours
further confirms the role of BRCA proteins in maintaining genomic integrity of a cell
(134). A more recent study conducted by Yoshihara et al. also showed that germline
copy number variation differs between BRCA1-associated ovarian tumours compared to
sporadic tumours (135). The authors showed that while BRCA1-associated tumours have
a higher number of deleted segments compared to sporadic tumours, the degree of
amplifications are lower in BRCA1-associated tumours when compared to sporadic
tumours (135).
35
Figure 1-5. Chromosomal Aberrations in Ovarian Tumours. An example ideogram illustrating chromosomal changes in ovarian cancer as detected by conventional and array CGH (lines and dots, respectively). Red lines, chromosomal loss; green lines, chromosomal gains. (adapted from Mayr et al., 2006).
36
High-resolution SNP arrays have also been used in the identification of novel
candidate tumour suppressor genes in ovarian cancer. By comparing 106 primary
ovarian tumours of various histological subtypes with matching normal DNA, Gorringe
et al. were able identify different LOH frequencies in different ovarian tumour subtypes
(136). This study confirmed previously identified homozygous deletions such as
CDKN2A (cyclin-dependent kinase inhibitor 2A), RB1 (retinoblastoma 1), and PTEN
(phosphatase and tensin homolog), and it allowed the identification of novel candidate
tumour suppressor genes near minimal regions of LOH on chromosomes 17, 13, 8p, 5q,
and X (136).
SNP analyses have been used to identify genes and their respective biological
processes that may be significant in ovarian cancer initiation and development. For
example, the importance of certain SNP alleles within xenobiotic metabolizing genes has
been recently examined in ovarian cancer (137). In the study, it was observed that
specific SNP alleles within the EPHX1 and NQO1 genes are associated with increased
serous ovarian cancer risk while a SNP within the ADH4 gene is associated with a
decreased risk, suggesting the importance of processing of pro-carcinogens in the
development of this disease (137). Another study comprising 19 study groups
participating in CIMBA revealed an association between the CASP8 D302H
polymorphism and decreased risk in breast and ovarian cancer in BRCA1 mutation
carriers but not BRCA2 mutation carriers, illustrating the importance of alterations within
certain apoptosis-associated genes in a subset of ovarian cancer patients (138).
The application of SNP arrays in clinical settings has also been explored in a
study that examined the accuracy and efficiency of this method in the screening of
37
genetic variants associated with disease predisposition (139). Specifically, Monaco et al.
demonstrated that BRCA1-specific SNP arrays had both the sensitivity and specificity
needed for the detection of point mutations, insertions or deletions of any length, of
known and unknown variants that are associated with this gene (139), thus suggesting
that high-throughput SNP analysis may provide a powerful means to identify individuals
who may be predisposed to breast and/or ovarian cancer in a time and cost efficient way,
as compared to high-throughput sequencing.
1.3.3 Gene Expression Profiling of Ovarian Cancer
Numerous studies employing microarrays to determine gene-expression profiles
of ovarian carcinomas or ovarian cancer cell lines have been reported (140). Many of
these studies were conducted to identify diagnostic markers that may potentially be used
in clinical settings to improve early detection of ovarian cancer, since most patients
(>60%) are diagnosed at advanced stages of the disease (32). These were accomplished
by comparing the gene expression signatures of ovarian cancers with normal ovarian
epithelium. For example, a study profiling the gene expression patterns of primary
cultures of ovarian cancer specimens and primary cultures of normal ovarian epithelia
identified IL-8 and FGF-2 as potential serum-based diagnostic markers that can be used
along with CA-125, a commonly used marker (141).
Microarrays have also been applied to identify predictive/prognostic markers for
disease outcomes. By examining the molecular signatures of early-stage and advanced
ovarian cancer tumours with different outcomes, researchers are able to correlate gene
expression levels with survival of the patients. For example, the expression of MAL,
38
which encodes a T-cell differentiation protein, has consistently been found to be
correlated with poor prognosis for ovarian cancer; indeed, its potential as a prognostic
indicator was further validated by detection of MAL protein by immunohistochemistry in
which high protein expression was associated with shorter survival in ovarian cancer
patients (142). In addition, due to the fact that many of the ovarian cancer patients
experience relapse within 6 months of chemotherapy (143), expression profiling has been
used to identify genes that may confer drug-resistance in ovarian cancer cells. By
comparing the expression signatures of paired ovarian tumours prior to and following
adjuvant chemotherapy, L’Espérance and colleagues identified a list of 121 genes that
were commonly up-regulated and a list of 54 genes that were down-regulated in post-
treatment tumours (144). These types of analyses have the potential to lead to assays to
help physicians in determining the appropriate treatment for patients according to their
ovarian cancer expression profiles.
Furthermore, to understand the biological mechanisms of ovarian cancer, studies
have been conducted to identify signalling pathways that may contribute to ovarian
carcinogenesis. In fact, alterations of different signalling pathways such as TGFβ, Myc,
Src, Rb/E2F, and β-catenin have been identified through ovarian cancer expression
microarray analyses (145, 146). Moreover, transcriptional profiling has been shown to be
able to distinguish between ovarian carcinomas of different tumour grades, thus giving
evidence that different pathways participate in the development of these tumours (82).
The site of origin of ovarian cancer remains controversial. The presence of occult
serous carcinomas in the fallopian tubes of BRCA mutation carriers undergoing
prophylactic surgery as observed from several studies has suggested the idea of distal
39
fallopian tube epithelium as the source of putative precursors of tubal and ovarian serous
carcinomas (147). To further investigate this hypothesis, Tone et al. compared the gene
expression profiles of high-grade tubal and ovarian serous carcinomas, and non-
malignant fallopian tube epithelial cells from individuals with and without BRCA
mutations (FTEb and FTEn, respectively) (148). Similar global gene expression patterns
were observed between tubal and ovarian specimens, suggesting that fallopian tubal
epithelium is the cell of origin for both of these tumours (148), or that they arise fro cells
that share similar features. Interestingly, despite their histological similarity, the
expression profiles of FTEb differ from FTEn; in fact, FTEb exhibited gene expression
patterns similar to those of serous carcinomas, especially those from FTEb obtained
during the luteal phase of ovarian cycle, implicating that the changes in gene expression
found in BRCA mutation carriers in certain hormonal environments are likely to
contribute to an increased risk of malignant transformation (148).
Motamed-Khorasani and colleagues also sought to examine the early molecular
processes that may be involved in the development of BRCA-associated ovarian cancer,
by comparing the expression profiles of ovarian surface epithelial (OSE) cell cultures
derived from BRCA carriers and malignant ovarian cancer cells (OVCAS), to OSE
derived from control patients, based on previous observations that there is a loss of
coordinated androgen regulation in ovarian cancer cells and in non-malignant epithelial
cells derived from women who are carriers of BRCA1 or BRCA2 mutations (149, 150).
Upon continuous exposure to androgen, a total of 17 differentially expressed genes were
identified in OSE from BRCA carriers and OVCAS when compared to control OSE
(149). In particular, BACH2 (basic leucine zipper transcription factor 2) and ACHE
40
(acetylcholinesterase) were found to be up-regulated in OSE from BRCA carriers as
compared to controls, and increased gene expression correlated with increased protein
expression in ovarian tumours, as shown by their immunohistochemical analysis from the
same study (149). Their observations therefore support the idea that altered androgen
response in BRCA mutation carriers could be involved in ovarian cancer susceptibility.
Given that published gene expression microarray datasets are publicly available to
the scientific community, researchers are able to gather information on ovarian cancer by
combining multiple studies from different laboratories, which may eliminate bias of
results due to different laboratory techniques, as well as increasing sample size. Indeed,
various research groups have sought to identify the differential gene expression patterns
in ovarian cancer by pooling data generated by different studies, in order to discover
biological pathways involved in ovarian tumourigenesis, as well as novel molecular
markers for better diagnosis and prognosis for the disease (151-154).
The above-mentioned expression microarray studies have begun to shed some
light on the cellular processes associated with hereditary-linked ovarian cancer.
However, much remains to be revealed regarding the various biological pathways that
may be involved in ovarian cancer progression in individuals with predispositions for the
disease. Therefore, the goal of my study is to examine and compare the expression
profiles of tumours from ovarian cancer patients with strong and weak family history of
breast and/or ovarian cancer, in order to identify biological pathways that may be
important in the progression of familial ovarian cancer.
41
1.4 Hypothesis The hypothesis of the work presented in this thesis is that in addition to
deregulated BRCA1 and BRCA2, alterations in other signalling pathways are also
important for the progression of ovarian cancer in the subgroups of patients with different
family history of breast and/or ovarian cancer. Specifically, I hypothesize that ovarian
tumours from patients with a strong family history of breast and/or ovarian cancer have
molecular alterations that may be responsible for the earlier age of onset observed in this
group of patients compared to those with weak family history.
1.5 Rationale and Objectives Despite recent advances in the understanding of its pathology, ovarian cancer
remains one of the most lethal gynaecological cancers. While the roles of BRCA1 and
BRCA2 in ovarian cancer have been well studied, it is likely that polygenic expression
alterations, which can lead to deregulation of important signalling pathways, are also
responsible for ovarian cancer pathogenesis in the subset of patients with familial
predispositions. Moreover, alteration of certain signalling pathways may be responsible
for the earlier age of onset and tumour progression as seen in patients with strong family
history.
A microarray study conducted by Jazaeri et al. revealed that the expression
profiles of sporadic ovarian tumours do share similarities between the expression profiles
of either BRCA1- or BRCA2-related tumours, implying that common genetic pathways
are involved in sporadic and familial cases of ovarian cancer (155). However, when
compared to each other, distinct expression signatures of BRCA1- and BRCA2-related
42
tumours were observed, suggesting that a diverse group of signalling pathways are
responsible for the development of ovarian cancer within familial cases of ovarian cancer
(155). Therefore, the examination and comparison of gene expression patterns among
patients with strong and weak family histories of breast and/or ovarian cancer will
identify candidate genes, as well as potential pathways that may be responsible for the
progression of familial ovarian cancer. Further investigation of these genes will provide
novel insights into the mechanisms involved, and this information may be useful for the
development of therapeutic tools.
In this thesis, a series of studies was conducted to investigate the hypothesis that
alterations in various signalling pathways are responsible for ovarian cancer progression
in patients with family history of breast and/or ovarian cancers.
The objectives of this thesis are:
i) To examine the molecular profiles of ovarian tumours from patients with family
history of breast and/or ovarian cancer, with the aim of identifying signalling pathways
that may be important in ovarian cancer pathogenesis in the group of patients with early
age of onset.
ii) To investigate the various mechanisms of hCDC alterations that may be involved in
familial ovarian cancer.
iii) To understand the role of PRKCZ in ovarian cancer progression by in vitro
biochemical and functional assays, and to identify its potential links to familial ovarian
cancer, by examining its relation to IGF1R and ITGB3, genes previously suggested to be
involved in BRCA-related breast and ovarian cancers.
CHAPTER 2
Gene Expression Profiling of Familial Ovarian Cancer
The work presented in this chapter was performed by KS with the exception of the
microarray statistical analyses, which was performed in collaboration with biostatisticians
Dr. Shelley Bull, Dr. Dushanthi Pinnaduwage, and Sarah Colby.
43
44
2.1 Introduction
As discussed in Chapter 1, the development of microarray technology has guided
the discovery of genes associated with various diseases, including ovarian cancer.
However, despite the plethora of ovarian cancer microarray studies, the global patterns
that distinguish groups of subjects with familial ovarian cancer have yet to be determined.
Additionally, while previous microarray studies have focused and compared the
expression profiles of BRCA1-linked and BRCA2-linked breast and ovarian tumours
(155, 156), it is possible that gene expression alterations in familial ovarian cancer
patients can occur independently of these two susceptibility genes, or that specific
changes in gene expression can affect both the BRCA1 and BRCA2 pathways.
In an effort to identify gene alterations that may play roles in familial ovarian
cancer progression, I utilized cDNA microarrays to examine the expression profiles of
ovarian cancer in patients with a strong or weak family history of breast and/or ovarian
cancer. In collaboration with biostatisticians Dr. Bull and Dr. Pinnaduwage, we
identified a subset of genes that were differentially expressed between the two subject
groups. Most of the functions of these genes have previously been shown to be important
in cancer development and progression, including roles in apoptosis, cell migration, cell
adhesion, and cell cycle regulation. I further interpreted the results by bioinformatics
analyses to identify the pathways and key molecules that may play potential roles in
ovarian cancer susceptibility and development.
45
2.2 Materials and Methods
2.2.1 Ovarian Cancer Specimens
Twenty-seven flash-frozen high-grade serous epithelial ovarian cancer samples
with similar pathological characteristics were obtained from the Toronto Ovarian Tissue
Bank and Database. The tumour specimens were obtained from consenting patients
according to the institutional guidelines of the Research Ethics Board. Specimens were
selected based on two main criteria: patients had not received neo-adjuvant chemotherapy
prior to surgery in order to preserve molecular signature of ovarian tumours and each
sample contained at least 75% of tumour content as assessed by the surface area of
corresponding histology slides by pathologist, with no evidence of necrosis, to ensure
purity and quality of tumours.
2.2.2 Family History Classification
Tumours were ranked and classified under two categories based on patients’
family history data and age of diagnosis. The strength of family history was ranked as
followed: multiple first-degree relatives with ovarian cancer, multiple relatives with
ovarian cancer, single relative with ovarian cancer, multiple relatives with breast cancer,
single relative with breast cancer, and relative(s) with other types of cancer. Nine
tumours were classified as “strong familial” (mean age of 51), and 18 were classified as
“weak familial” (mean age of 62) (Table 2-1). BRCA1 and BRCA2 mutation status were
available for four of the patients (two BRCA1 mutation carriers, and two non-carriers),
but since the classification of tumours was solely based on family history, the mutation
status information was not used for the present study.
46
Table 2-1. Characterization of the subjects in the strong versus weak family history groups. Classification of serous ovarian tumour samples according to the patients’ family history data. Selection of high grade tumours was based on similar pathological characteristics between the two subject groups. Refer Appendix A1 for detailed family history data.
Strong Familial Weak Familial
Total # of cases
9
18
Family history of ovarian cancer (# of affected relatives)
0 3/9 18/18 1 2/9 2 2/9 3 1/9 4 1/9
Family history of breast cancer (# of affected relatives)
0 3/9 6/18 1 4/9 12/18 2 2/9
Age at diagnosis (yrs) ≤ 55 8/9 5/18 > 55 1/9 13/18
Mean age at diagnosis (yrs)
51 62
47
2.2.3 RNA Isolation and Reverse Transcription
Total RNA was extracted using Trizol according to the manufacturer’s
instructions (Life Technologies, Frederick, MD, USA). The purity of RNA was
confirmed by A260/A280 ratio of 1.8-2.2, while the integrity of RNA was verified by 1%
agarose gel electrophoresis. Five micrograms of RNA from each tumour sample, as well
as reference RNA composed of 13 pooled cell lines (Table 2-2), were reverse-transcribed
(Superscript III reverse transcriptase, Invitrogen) using anchor oligo dT (AncT) to
generate complementary DNA (cDNA).
Table 2-2. Thirteen cell lines comprising common reference pool used for microarray experiments. The diversity of gene representation by each of these cell lines provides hybridization signal at each probe location on the microarray slide, which is needed for normalization of signal output during analysis process.
Cell Line Description NTERA-2 c1.D1 (CRL-1973) Human testis cancer Hs578T (HTB-126) Human breast carcinoma HepG2 (HB-8065) Human hepatoblastoma Ht1080 (CCL-121) Human fibrosarcoma SW872 (HTB-92) Human liposarcoma T47D (HTB-133) Human breast carcinoma MCF-12A (CRL-10782) Human breast normal Fetal normal muscle 12 week old fetus normal Colo-205 (CCL-222) Human colon cancer MOLT-4 (CRL-1582) Human leukemia RPMI8226 (CCL-155) Human plasmacytoma SKOV-3 (HTB-77) Human ovarian adenocarcinoma SK-MEL-28 (HTB-72) Human melanoma
2.2.4 cDNA Expression Microarrays
Microarray experiments were carried out using arrays that contain ~19,000
characterized and unknown ESTs (University Health Network Microarray Center,
48
Toronto, Canada, www.microarrays.ca). Tumour and reference cDNAs were
differentially and fluorescently labelled by incorporation of either Cy3-dCTP or Cy5-
dCTP, and were co-hybridized on microarray slides at 37oC for overnight. Arrays were
scanned the next day using the Gene Pix 4000B scanner (Axon Instruments, USA)
following stringent washes. Cy3- and Cy5-labeled cDNA were excited by the green laser
at 532nm and the red laser at 635nm, respectively, and each of the fluorescent spots on
slides were measured and stored as microarray images (Figure 2-1). To minimize the
number of saturated pixels, the PMT (photomultiplier tube) gain for each laser was
adjusted to give Cy3:Cy5 of 1.0. Images were analyzed using GenePix Pro 4.0 Software
(Axon Instruments, USA). Dye swap experiments were performed for each of the
specimens to compensate for dye bias.
Figure 2-1. Subarray of a representative 19K cDNA microarray. Image of a representative subarray of a cDNA expression microarray slide. In this example, cDNA from an ovarian tumour specimen was labelled with Cy5 (red) and cDNA from reference pool was labelled with Cy3 (green). Red spots indicate genes that are more highly expressed in tumour sample, while green spots indicate genes that are less expressed in tumour sample.
49
2.2.5 Microarray Data Analysis
2.2.5.1 Pre-processing and normalization of expression data
The gene expression data were obtained from the original GenePix image files as
spot intensities by correcting the mean foreground for each spot with the median local
background. The array quality was controlled by requiring arrays to have more than 80%
of spots with spot intensities higher than their local background and more than 75% of
spots with spot intensity higher than 1.2 times their local background in both channels.
Spots with foreground intensity lower than background were treated as missing. A
relative expression value was obtained for each gene as the log base 2 ratio of the
adjusted intensity for the sample channel versus the reference channel. As proposed by
Yang et al. (157), the log2 ratios were normalized by a within-array print-tip loess
adjustment followed by a between-array scale adjustment. Poor quality spots as flagged
by the GenePix image analysis software were excluded from the normalization. Pre-
processing and normalization were carried out using R (http://cran.r-project.org) and the
Bioconductor (http://www.bioconductor.org) package LIMMA (158). There were 15,437
genes retained in the final data set for analysis. All tumours were assessed.
2.2.5.2 Microarray Statistical Analysis
To identify genes that discriminate between strong and weak familial ovarian tumours,
supervised univariate analyses of array-based log2 gene expression were performed for
each gene using the modified Student t-test in the SAM (159) procedure implemented in
R (version 2.10.1, http://cran.r-project.org). P-values were estimated from a set of 500
random sample permutations.
50
2.2.6 Quantitative Real-time RT-PCR
To validate the gene expression results from the microarray, quantitative real-time
RT-PCR was performed using commercially available Assay-on-Demand probe primer
sets (Applied Biosystems). The standard curve method was employed for quantification
of gene expression. The PCR conditions were as follows: 95oC for ten minutes,
followed by 40 cycles of 95oC for 15 seconds and 59oC for 1 minute. HPRT1
(hypoxanthine phosphoribosyltransferase 1) was chosen as the internal control gene as it
was expressed at similar levels across all tumour samples (Figure 2-2). The difference
between means was tested by using the Student's t test.
51
Figure 2-2. Expression of housekeeping genes in ovarian tumour samples as measured by gene expression microarrays. In order to determine the appropriate internal control gene for subsequent quantitative real-time RT-PCR experiments, the gene expression of 8 different housekeeping genes were measured and compared among 27 ovarian tumour samples from the present study. MRPL19, HPRT1, YWHAZ, and TBP were shown to have similar expression across tumour samples. HPRT1 was subsequently chosen as the reference gene. MRPL19 PUM1
RPL13A PSMC4
HPRT1 YWHAZ
TBP HMBS
52
2.2.7 Integration of Array Data to Interaction Networks
Ingenuity Pathway Analysis (Ingenuity® Systems, www.ingenuity.com) was used
to create interaction networks consisting of genes that were significantly differentiated in
the familial ovarian cancer microarray analysis in order to identify and explore the
biological functions that are relevant in the development/progression of ovarian cancer.
In brief, the top significantly differentiated genes from microarray results were mapped
onto Ingenuity Pathway Knowledge Base using gene symbols to generate interaction
networks. The Functional Analysis of a network identified the biological functions that
were most significant to the genes in the network. Fisher’s exact test was used to
calculate a p-value determining the probability that each biological function assigned to
that network is due to chance alone.
53
2.3 Results
2.3.1 Identification of Genes Distinguishing Strong and Weak Familial Ovarian
Cancers
To identify differentially expressed genes among tumours from patients with
strong and weak family history of breast and/or ovarian cancer, two different approaches
were taken in the analysis: a supervised-class comparison, and a candidate gene
approach.
2.3.1.1 Supervised-Class Comparison
As described, a total of 27 ovarian tumour samples were subjected to gene
expression profiling by a supervised-class comparison approach. Nine of the tumours
represented the group of patients with a strong family history of breast and/or ovarian
cancer and 18 tumours were from patients with a weak family history. The expression
profiles of these two groups were compared and a list of differentiated genes was
generated based on the significance criterion of p<0.01 (Figure 2-3). The most
significant genes identified from the analysis are shown in Table 2-3, along with their
molecular functions and the cellular processes in which they are involved, including
apoptosis/programmed cell death, transcription, protein modification, signal transduction,
cell migration, cell adhesion, and cell cycle regulation.
54
Figure 2-3. Heatmap illustrating differential gene expression patterns in strong and weak familial ovarian cancer groups (p<0.01). Clustering of top 175 significant genes is shown. Differences with expression levels greater than the mean are coloured in red those below the mean are coloured in green, and no expression differences are in black. Each row represents a single probe set, as identified by GenBank ID, and each column represents individual ovarian tumour samples (blue lines – strong familial tumours; purple lines – weak familial tumours).
55
Table 2-3. Top 100 differentially expressed genes between strong and weak familial ovarian tumours, as ranked by SAM. Fold difference = strong familial gene expression/weak familial gene expression. *Empty gene ID indicates uncharacterized EST.
GenBank Accession Gene* Molecular Function / Biological Process Cytoband Fold
difference p-value
W60281 ZHX1 Transcription factor activity; negative regulation of transcription, DNA-dependent
8q24.13 1.346 0.00050
AI375078 MAGI3 Kinase activity; apoptosis; intracellular signalling cascade
1p12-p11.2 1.498 0.00057
W31814 PRKCZ Protein binding; protein serine/threonine kinase activity; anti-apoptosis; intracellular signaling cascade
1p36.33-p36.2
2.058 0.00065
AI734230 LOC728676 1q42.13 2.462 0.00084
W38594 TUT1 RNA binding; nucleotide binding; transferase activity 11q12.3 1.877 0.00134
W06830 MCM5 ATPase activity; DNA replication; cell cycle regulation; transcription regulation
22q13.1 1.421 0.00135
AA036944 GMFB Enzyme activator activity; protein kinase inhibitor activity; signal transducer activity; signal transduction
14q22.2 2.710 0.00138
W04817 1.747 0.00178 T70749 0.708 0.00187 BI832845 GNA14 GTPase activity; signal transducer activity; signal
transduction 9q21 1.988 0.00203
T85827 0.392 0.00212 H09018 STX1B Extracellular ligand-gated ion channel activity;
intracellular protein transport; regulation of exocytosis 16p12-p11 0.686 0.00217
R60831 RGMB Identical protein binding; cell adhesion; positive regulation of transcription; signal transduction
5q15 0.559 0.00230
AV725418 0.530 0.00235 N75196 1.436 0.00273 N46715 1.518 0.00283 W03485 SMAD5 Transcriptional activator activity; receptor signaling
protein activity; signal transduction; transcription 5q31 3.257 0.00294
T80520 1.322 0.00309 T95724 SLC25A15 Transporter activity; amino acid metabolic process;
transport; urea cycle 13q14 0.794 0.00309
H19077 DBP RNA polymerase II transcription factor activity; protein dimerization activity; regulation of cell proliferation; regulation of transcription from RNA polymerase II promoter
19q13.2 0.611 0.00311
R51440 0.641 0.00314 R54550 0.578 0.00320 T90266 1.542 0.00321 H71721 1.907 0.00327 BG570119 RSRC2 Inhibitor of cell proliferation 12q24.31 1.411 0.00355 H62960 2.014 0.00374 BQ006563 SIRT3 DNA binding; hydrolase activity; chromatin silencing;
regulation of transcription, DNA-dependent 11p15.5 0.796 0.00375
H61030 REXO2 Hydrolase activity; nucleic acid binding; nucleotide metabolic process
11q23.2 0.645 0.00382
BM469380 ZDHHC6 Metal ion binding; transferase activity 10q25.3 1.886 0.00419 W01536 PDCD4 RNA binding; protein binding; apoptosis; cell aging;
negative regulation of cell cycle progression; negative regulation of transcription
10q24 0.698 0.00424
W00465 TSPYL5 Nucleosome assembly 8q22.1 0.722 0.00435
56
N92095 0.517 0.00445 R97881 1.412 0.00458 R91589 0.514 0.00462 R17265 HGSNAT Transferase activity 8p11.1 1.516 0.00471
AA203115 SLC22A16 Transporter activity; transport 6q22.1 1.605 0.00472
R54138 GPR162 Receptor activity; G-protein coupled receptor protein signaling pathway; signal transduction
12p13 0.769 0.00477
N77718 1.505 0.00488 AL519368 LOC389833 4 1.669 0.00501
R86730 LBR DNA binding; protein binding; receptor activity 1q42.1 0.608 0.00506 N46724 ARHGEF7 Guanyl-nucleotide exchange factor activity; protein
binding; apoptosis; signal transduction 13q34 1.547 0.00527
H17686 LRRC17 Protein binding; osteoblast differentiation and proliferation
7q22.1 1.910 0.00536
W31507 CHRFAM7A Extracellular ligand-gated ion channel activity; ion transport
15q13.1 1.626 0.00536
N91701 PAPD1 PAP associated domain containing 1 10p12.1 1.562 0.00571 BG616960 VTN Protein binding; cell adhesion; cell-matrix adhesion 17q11 0.717 0.00576 T80023 1.367 0.00577 AA043077 0.642 0.00593 R87552 GNAO1 GTPase activity; receptor signaling protein activity;
signal transduction 16q13 0.742 0.00607
R50922 NLGN4X Protein binding; protein homodimerization activity; cell adhesion; cell-cell junction organization
Xp22.33 0.656 0.00612
H28503 NOTCH2NL Calcium ion binding; Notch signaling pathway; cell differentiation
1q21.2 1.143 0.00613
R06130 FCN1 Calcium ion binding; receptor binding; signal transduction
9q34 0.697 0.00614
H16624 FAT4 Calcium ion binding; cell adhesion 4q28.1 1.689 0.00617
W87412 ARID2 DNA binding; protein binding; chromatin modification; regulation of transcription, DNA-dependent
12q13.11 1.732 0.00656
AI820728 0.559 0.00656 H88063 C1orf9 Multicellular organismal organization 1q24 0.434 0.00669 R24998 0.665 0.00683 N95578 1.403 0.00713 H04765 CPD Metal ion binding; peptidase activity; proteolysis 17p11.1-
q11.2 0.700 0.00721
AA143153 CYP11A1 Cholesterol binding; metal ion binding; cholesterol metabolic process; response to estrogen stimulus
15q23-q24 1.630 0.00727
T83295 HEXA Hydrolase activity; metabolic process 15q23-q24 0.832 0.00743 AA203550 2.423 0.00750 R97190 0.625 0.00763 H08597 PLCXD3 Hydrolase activity; signal transducer activity; lipid
catabolic process 5p13.1 0.596 0.00767
W03856 ABCC6P1 16p12.3 1.516 0.00770
H26209 0.755 0.00772 H66891 0.679 0.00777 N46663 1.420 0.00777 T83443 IGHMBP2 ATP binding, DNA binding; RNA binding;
transcription factor binding; protein binding; DNA recombination; DNA repair; DNA replication; cell
11q13.2-q13.4
0.644 0.00786
57
death; regulation of transcription R12356 MAOB Protein homodimerization activity; oxidative reduction Xp11.23 1.639 0.00801 BM982143 POLR3H DNA binding; DNA-directed RNA polymerase activity;
transcription 22q11.2-q13.31
1.314 0.00811
AL556801 NCAPD3 Binding; cell cycle; cell division; mitosis 11q25 1.597 0.00839
AA099730 FVT1 Binding; oxidoreductase activity; oxidation reduction; sphingolipid biosynthetic process
18q21.3 1.361 0.00846
T97415 1.499 0.00848 H10128 SHC2 Protein binding; Ras protein signal transduction;
intracellular signaling pathway 19p13.3 0.842 0.00849
R54527 0.670 0.00867 R67303 NEO1 Cadherin binding; receptor activity; transcription
regulation activity; cell adhesion 15q22.3-q23
0.661 0.00877
AA037624 C7orf49 7q33 1.364 0.00889 R59287 SFRS14 RNA binding; RNA splicing; mRNA processing 19p12 1.518 0.00905 H19859 RABAC1 Protein binding 19q13.31 0.683 0.00905 T95778 SH3BP4 Protein binding; endocytosis 2q37.1-
q37.2 0.792 0.00911
AL575239 0.601 0.00924 H51850 0.542 0.00934 AW953237 PITPNC1 Lipid binding; protein binding; lipid transport; signal
transduction 17q24.3 0.723 0.00939
W78914 PILRA Protein binding; receptor activity; signal transduction 7q22.1 2.230 0.00942
H72224 MRPS12 Protein binding; translation 19q13.1-q13.2
0.691 0.00954
BQ017489 ACTN1 Actin binding; integrin binding; protein binding; focal adhesion assembly; regulation of apoptosis
14q24.1-q24.2
1.420 0.00960
BM509122 IGL@ 22q11.1-q11.2
0.403 0.00968
R24238 1.235 0.00968 R73462 ATHL1 Hydrolase activity; carbohydrate metabolic process 11p15.5 0.731 0.00972 T83168 0.479 0.00989 AA134742 BRWD2 Cell cycle progression; signal transduction; apoptosis,
gene regualtion 10q26 2.349 0.00990
W86215 ERF Ligand-regulated transcription factor activity; transcription corepressor activity; cell proliferation
19q13 2.160 0.00992
H29655 RELB DNA binding; protein binding; transcription corepressor activity; regulation of transcription, DNA-dependent
19q13.32 0.624 0.01051
AA126588 0.754 0.01065 AA047000 IL17RC Receptor activity 3p25.3 1.436 0.01070 AW962742 LOC400604 17q21.33 1.357 0.01077
H42572 0.700 0.01085 R83139 SCFV 14 0.436 0.01095 H88020 1.608 0.01096 AA034344 8p21.1 1.539 0.01110
58
2.3.1.2 Candidate Gene Approach
In addition to the supervised-class comparison, I have chosen a list of genes from
the literature that has previously been implicated in ovarian cancer or genes that are
biologically relevant, and compared each of their microarray gene expression values
between the strong and weak family history ovarian cancer groups. The selected genes
include: MYC, KRAS, PIK3AP1, hCDC4, LRMP, MLH1, TP53, CDKN1B, EVI1,
CCND1, PIK3CA, as well as genes from the TGFβ family. Most of these genes did not
show differential expression between our two subject groups. Interestingly, while
differential expression of hCDC4 was not detected as significant in the supervised-class
comparison analysis at cut-off p-value of 0.01, it was identified as significantly less
expressed in group with strong family history at p=0.05 using the candidate gene
approach (Figure 2-4).
2.3.2 Validation of Differentially Expressed Genes
I selected several genes for gene expression validation using quantitative real-time
RT-PCR with RNA from the same 27 tumour specimens used for my microarray
experiments, based on fold difference of at least 2.0, as well as their potential roles in
cancer based on previous literature, which included PRKCZ, FAT4, and hCDC4. The
gene expression patterns between the two subject groups for these genes as identified by
real-time PCR were in concordance with the microarray results (Figure 2-5). Other genes
such as SMAD5 and NEO1 were chosen for validation, and while similar trends were
observed, the differences did not reach statistical significance.
59
Figure 2-4. Identification of hCDC4 as a differentially expressed gene between strong and weak familial ovarian tumours by candidate gene approach. Selected genes were chosen from the literature and their gene expressions from microarray results were analyzed. hCDC4 gene expression was shown to be differentially expressed (p =0.05).
60
Figure 2-5. Real-time PCR validation of differential gene expression between strong and weak family history groups. Expression levels of selected genes FAT4, PRKCZ, and hCDC4 were independently validated using real-time PCR using the same 27 ovarian tumour samples from microarray experiments. Differential gene expression patterns identified from real-time PCR were in concordance with microarray expressions. cDNA microarray expression values were normalized using global scaling approach and real-time PCR expression values were normalized to housekeeping gene HPRT1 (Student’s t test).
PRKCZ mRNA Expression
0
1
2
3
4
5
Microarray RT-PCR
Rel
ativ
e m
RN
A E
xpre
ssio
n
Strong Familial
Weak Familial p=0.01
FAT4 mRNA Expression
0
0.5
1
1.5
2
2.5
Microarray RT-PCR
Rel
ativ
e m
RN
A E
xpre
ssio
n
Strong FamilialWeak Familial p=0.03
hCDC4 mRNA Expression
00.5
11.5
22.5
33.5
44.5
Microrray RT-PCR
Rel
ativ
e m
RN
A E
xpre
ssio
n p=0.05 Strong Familial
Weak Familial
61
2.3.3 Molecular Network Analyses Dynamic pathway modelling was carried out using Ingenuity Pathway Anaysis
(IPA) software to examine various molecular interactions/pathways associated with
familial ovarian cancer. I mapped the top ranked 300 differentially expressed genes
between the strong and weak ovarian cancer groups onto IPA Knowledge Base, and the
connectivity of the genes through direct physical, transcriptional and enzymatic
interactions was computed and interaction networks were generated. Focus genes
(differentially expressed genes from the present study), the network score (in which a
value greater than 2 corresponds with a probability of <0.01% that the network was
assembled by chance alone), and the cellular functions associated with each of the
identified networks are listed in Table 2-4.
Through IPA, it is confirmed that the genes that are differentially expressed
between strong and weak familial tumours include players that participate in relevant
signalling pathways related to tumourigenesis. These pathways include those involving
MAPK, HNF4A (hepatocyte nuclear factor 4, alpha), histone 3, HGF (hepatocyte growth
factor), and beta-estradiol (Figure 2-6). Some of the cellular functions and diseases
associated with molecular networks include DNA replication, DNA recombination, DNA
repair, cell cycle, lipid metabolism, cell death, and cellulr growth & proliferation, and
cancer.
62
Figure 2-6. Graphical representations of the molecular relationships between genes identified from familial ovarian microarray analysis using Ingenuity Pathway Analysis. Red nodes represent genes that had higher expression in the strong familial group, green nodes represent genes that had lower expression in the strong familial group, and white nodes represent genes that had similar expression between two groups or were not on the array. Biological relationship between two nodes is represented as an edge (line). All interactions are supported by at least one reference from the literature, from a textbook, or from canonical information stored in the Ingenuity Pathways Knowledge Base. Different shapes of enodes represent the functional class of the gene product as indicated in the figure legend. Interaction networks involving A) players within MAPK signaling pathway, B) HNF4A, C) histone 3, and D) HGF and beta-estradiol.
63
A)
64
B)
65
C)
66
D)
67
Table 2-4. Top functions of networks as identified by Ingenuity Pathway Analysis (IPA). Differentially expressed genes from microarray analysis are shown in bold (focus molecules). A score of >3 was considered statistically significant (p<0.001).
Molecules in Network Score Focus Molecules Top Functions
Actin, ADCY, Akt, BIRC3, Caspase, CDC23, Cyclin A, CYP11A1, E3 RING, ERK, ERK1/2, G alpha, G alphai, G-protein beta, GMFB, GNA14, GNAO1, GPC1, hCG, HIRA, IGF1R, ITGB3, Jnk, LBR, MAP3K8, Mapk, NFkB (complex), NPPA, PARK2, PI3K, PIP5K1B, Pkc(s), PRKCZ, Rac, VTN
36 17
DNA Replication, Recombination, and Repair, Infection Mechanism, Endocrine System Development and Function, Lipid Metabolism
ACO2, ACTN1, BMP15, C1ORF9, CABC1, CXXC5, DACH1, DLG4, FBP1, FDX1, FSH, FXN, HDAC8, HEXA, HNF4A, KCNA5, MIR298, NEO1, NLGN4X, PDCD4, RABAC1, RBAK, REXO2, RGMA, RPLP1, SF3B1, SH3BP4, SMAD5, Smad1/5/8, SNRPA, TOB1, TUT1, VPS29, XRCC4, ZDHHC6
28 14
Cell Cycle, Reproductive System Development and Function, RNA Damage and Repair, Cancer
ACHE, beta-estradiol, CKS2, CPD, CTSH, FCN1, G3BP2, GM2A, HGF, IFNG, IGHMBP2, IL6, IRS, IRS1, MYC, NAPSA, NFKBIA, OSBPL1A, PPP1R12C, RDBP, RFC2, SERPINA1, SFRS2, SFRS2IP, SFTPB, SH2B2, SLC22A16, SLC25A15, SMARCA4, SNRNP70, SPHK2, TFDP1, TGFB1, ZFP36L2, ZNF185
26 13
Cell Cycle, Cancer, Infection Mechanism, Lipid Metabolism
ABI3, AKAP, AKAP1, AKAP13, ALDOA, ARHGEF7, ATXN1, C3ORF15, CBFA2T3, DIO1, FAT4, Histone h3, Histone h4, LRCH1, MCM5, Mi2, MTUS1, NCOR1, P38 MAPK, Pka, PRKAC, PRKAR1B, RUNX1T1, SCRIB, SIRT3, TBL1XR1, THAP7, THRAP3, WWC1, WWC2, YWHAZ, ZFP36, ZHX1, ZHX3, ZMYND11
21 11
Cellular Development, Hematological System Development and Function, Hematopoiesis
AHR, CDC2, Ck2, CREM, CSNK2A1, CSNK2B, CTGF, D830050J10RIK, FBL, FGF1, GTF2F1, L1CAM, LRRC17, MDM2, Mg2+, MYCN, NCL, NETO2, NME1, NOP2, ODC1, PTN, RBL2, SNAP25, SPP1, STX4, STX1B, SYT1, SYT4, TBP, TCF4, THRA, TOP1, VAMP2
8 5
Cellular Growth and Proliferation, Cancer, Cell Morphology, Cell Death
68
2.4 Discussion
Gene expression microarray is a powerful tool for identifying abnormal gene
expression patterns associated with various human diseases. Indeed, a number of
microarray studies have been conducted in the past decade to identify gene expression
alterations in ovarian cancer (140). These studies have identified signature profiles of
ovarian cancer cell lines and ovarian carcinomas, which subsequently led to the
identification of potential prognostic markers, as well as genes involved in ovarian
tumourigenesis and those involved in chemotherapy resistance, such as resistance to
cisplatin (140). While these studies have indeed provided a better understanding of
ovarian cancer, the global gene expression profiling for familial ovarian cancer is lacking.
To address this, I have applied cDNA microarrays and compared the gene expression
patterns in tumour samples from patients with strong versus weak familial backgrounds
of breast and/or ovarian cancer.
We have identified a list of significant genes that are differentially expressed
between ovarian tumours from patients with strong and weak family history of breast
and/or ovarian cancer, on the basis of the analysis of 9 strong-familial and 18 weak-
familial ovarian tumours.
A subset of the genes identified from the array analysis has specifically been
demonstrated to be involved in cancer development in earlier studies. Of these genes,
some showed lower expression within the strong familial group, including the following:
SIRT3 (sirtuin 3), a gene encoding a NAD-dependent histone deacetylase that has been
demonstrated to play role in apoptosis in a variety of human cancer cell lines (160); VTN
(vitronectin), an extracellular matrix protein that plays role in cell adhesion and
69
spreading, and previously demonstrated to play role in ovarian cancer dissemination
(161); NEO1 (neogenin homolog 1), which encodes a cell surface protein that shares
homology with tumour suppressor candidate gene DCC (deleted in colorectal carcinoma)
and whose expression has been previously observed to be significantly reduced in
prostate tumours compared to normal prostate tissues (162, 163), and PDCD4
(programmed cell death 4), a tumour suppressor gene shown to inhibit proliferation of
neuroendocrine tumour cells, and whose expression is frequently lost in human glioma, in
addition to its involvement in the progression of lung, breast, colon, prostate carcinomas,
and most notably shown to be able to inhibit the malignant phenotype of ovarian cancer
in vitro and in vivo (164-167).
Likewise, some genes exhibit higher levels of expression in the strong familial
patient group compared with the weak familial group. Such genes include MAGI-3
(membrane associated guanylate kinase), whose gene product interacts with LPA2 to
facilitate LPA2-mediated activation of ERK and RhoA, potentially leading to an increase
in gene transcription and cell survival (149); cell cycle regulation gene MCM5
(minichromosome maintenance complex component 5), which has been described to be
over-expressed in cervical and esophageal cancer (168, 169); as well as CYP11A1
(cytochrome P450, family 11, subfamily A, polypeptide 1), which encodes a member of
the cytochrome P450 superfamily of enzymes involved in steroidogenesis and its genetic
variation has been suggested to be able to influence risk of various malignancies,
including endometrial and prostate cancers (170, 171).
Additionally, the gene PRKCZ (protein kinase C zeta) also exhibited a higher
level of expression in the strong familial group. This gene encodes a serine/threonine
70
kinase that has previously been implicated in the regulation of cellular transformation and
carcinogenesis (172). The importance of PRKCZ in tumourigenesis has been shown in
various studies (172-178); however, its role in ovarian cancer had yet to be determined.
Therefore, I have chosen this gene for further investigations. Detailed description of this
gene, as well as the functional studies of PRKCZ that I have performed, will be discussed
in Chapter 4 of this thesis.
In order to interpret my microarray data and to generate hypotheses for future
studies, I have utilized Ingenuity Pathway Analysis (IPA) software for further analyses.
IPA was applied to identify interaction networks among genes that are differentially
expressed between ovarian cancers from strong versus weak family history to gain an
understanding of which functional cellular processes were altered between these two
subject groups.
Through statistical analyses performed by IPA, several potentially important
interaction networks were identified. The top functions and diseases associated with
these networks include DNA replication, recombination and repair, cancer, cell
morphology, cell death, cell cycle, as well as cellular growth and proliferation. Other
cellular processes such as lipid metabolism, endocrine, reproductive, and haematological
system development were also identified.
By examining these networks, I was able to identify signalling molecules and
pathways that were altered in patients with a strong family history of breast and/or
ovarian cancer. For example, it was observed that the expressions of genes directly or
indirectly associated with the mitogen-activated protein kinase (MAPK) pathway were
altered between these patient groups. These genes include VTN (vitronectin), IGF1R
71
(insulin-like growth factor 1 receptor), ITGB3 (integrin beta 3), LBR (lamin B receptor),
BIRC3 (baculoviral IAP repeat-containing 3), and PARK2 (parkin 2). Some of these
genes have previously been associated with ovarian cancer (179-184); however,
additional functional studies are required to determine if these genes are able to cooperate
with each other to enhance ovarian cancer development. A protein network that
functions in cell cycle and cell death was also found from our IPA analysis. This
particular network centers around HNF4A, a member of the nuclear receptor superfamily
that has been demonstrated to be a useful marker for histological and cytological
diagnosis of ovarian mucinous tumours (185). Its role in other ovarian tumour types,
including serous carcinomas, however, requires further investigation. In addition to
HNF4A, other molecules such as tumour suppressor PDCD4 (programmed cell death 4),
TGFβ1 interactor SMAD5 (SMAD family member 5), and tumour suppressor
retinoblastoma 1 interactor RBAK (RB-associated KRAB zinc finger) were also present
in this interaction network, and may be directly or indirectly regulated by HNF4A. Their
exact regulation mechanisms and their roles in familial ovarian tumourigenesis remain to
be examined. Another network encompassing histone H3 was observed in which gene
expression of histone H3 interactors such as NCOR1 (nuclear receptor co-repressor 1),
MTUS1 (mitochondrial tumour suppressor 1), MCM5 (minichromosome maintenance
complex component 5), and RUNX1T1 (runt-related transcription factor 1) were found to
be differentially expressed between the two groups. Since histone H3 modification has
been previously described in ovarian cancer (186, 187), it may be relevant to examine
how these interactors can affect the activity of histone H3, or vice versa, in familial
ovarian cancer.
72
Lastly, an interaction network we observed involves multiple molecules including
MYC, TGFβ1 (transforming growth factor, beta 1), HGF, and beta-estradiol. While
differential expression levels of these specific genes were not identified in our study, it is
interesting to note that their common direct/indirect interactors are shown to be
differentially expressed between the two groups, thus linking these important signalling
pathways. In particular, the presence of HGF and beta-estradiol within this interaction
network is a notable finding.
HGF is a pleiotropic factor that, along with its receptor tyrosine kinase Met, can
stimulate multiple biological responses, including epithelial morphogenesis (188).
Moreover, it has been observed that HGF and Met are both differentially regulated in
normal human ovarian surface epithelium cultures derived from women with and without
a family history of ovarian cancer, and was suggested that co-expression of these two
genes may enhance susceptibility to ovarian carcinogenesis in women with hereditary
ovarian cancer syndromes by their ability to activate the PI3K and MAPK signalling
pathways, which may lead to metaplastic changes of the ovarian surface epithelium
(188). Based on this previous observation, and results from our interaction analysis, I
hypothesize that players associated with the HGF pathway may be important in ovarian
cancer in the subset of patients with a strong family history, by altering and sustaining the
expression of the genes involved in ovarian tumourigenic transformation. Therefore it
may be worthwhile to further examine the regulatory mechanisms within the HGF-axis in
relation to familial ovarian cancer development.
The presence of beta-estradiol related interactors found within this interaction
network was also of interest given that recent studies have found an association between
73
certain genetic polymorphisms within the steroid hormone pathway and an increase in
breast and ovarian cancer risks (189). How these specific polymorphisms can affect the
expression of these genes, or how these genes may directly or indirectly be altering the
steroid hormone pathway thus leading to ovarian carcinogenesis, however, still requires
further investigation. Nevertheless, the importance of beta-estradiol in ovarian cancer
development has been demonstrated. It was suggested that this hormone may play a role
in ovarian tumourigenesis by up-regulating bcl-2 gene and protein expression levels, thus
preventing apoptosis in tumourigenic ovarian surface epithelial cells (190). This is
critical during early stages of ovarian cancer development, as this advantage in cell
survival may permit these cells to accumulate mutations at a greater than normal rate,
thereby accelerating the overall rate by which these cells can transform from a pre-
malignant to malignant state. In another study, the role of beta-estradiol in the
reinforcement of invasion in epithelial ovarian cancer cell lines was examined (191). It
was found that ovarian cancer cells induced by beta-estradiol can lead to an increase in
MMP-2 expression and a decrease in E-cadherin expression, and that these changes were
associated with an increased expression of Snail, a transcription factor shown to be
involved in epithelial-mesenchymal transitions (EMT) of cancer cells, which is important
in invasion and metastasis (191). Taken together, it has been suggested that the
regulation in beta-estradiol expression during both early and late stages may be critical
during ovarian carcinogenesis.
Based on these sub-networks, it would be interesting to examine whether the
different players that control the levels of HGF and beta-estradiol are differentially
expressed or activated in patients with different family history of breast and/or ovarian
74
cancer, and how these changes in expression or activity may affect downstream targets
that can affect ovarian cancer development, such as those players involved in
tumourigenic transformation, cell survival and EMT. It may also be possible that
BRCA1 and/or BRCA2 are participants in these pathways directly or indirectly; however,
further studies are required to determine if this is the case.
In conclusion, the microarray results from this Chapter provided the genetic
profiles of familial ovarian cancer, and while larger sets of families with familial breast
and/or ovarian cancer are required to further validate and to increase robustness of our
results, my findings have provided some promising candidate genes that may potentially
be important in familial ovarian tumourigenesis. Potential biological pathways involved
in this disease are further identified through bioinformatics analyses and polygenic
alterations within these pathways are likely to be involved in increasing the susceptibility
of ovarian cancer in individuals with strong family history of breast and/or ovarian
cancer. However, the detailed mechanisms of how and which of these genetic
alterations are involved, and whether these alterations have different effects in BRCA1 or
BRCA2 mutation carriers, remain to be investigated. Therefore, additional functional
studies of the identified genes will be required to provide information that may be
valuable in understanding the relevant mechanisms involved in the development of
familial ovarian cancer.
CHAPTER 3
hCDC4 in Familial Ovarian Cancer
The work presented in this Chapter was performed by KS with the exception of the
immunohistochemistry on ovarian tissue microarray (TMA), which was performed in
collaboration with Dr. Patricia Shaw, with assistance from laboratory technician Kelvin
So. Scoring of TMA was assisted by Dr. Alicia Tone.
75
76
3.1 Introduction
The ubiquitin-proteosome system (UPS) is an important regulator of cellular
homeostasis, as it is involved in the destruction of regulatory proteins within the
eukaryotic cell (192). Some of the cellular processes which are UPS-dependent include
differentiation, proliferation, DNA damage repair and apoptosis (192, 193). UPS
functions through an enzymatic cascade consisting of ubiquitin-activating enzyme E1,
ubiquitin-conjugating enzyme E2, and ubiquitin-protein ligase E3 (192).
Several hundred E3 ligases exist in the human genome and the SCF
(SKP1/CUL1/F-box) ubiquitin ligase complex is among the best studied (192). SCF
complex contains a cullin family scaffolding protein that binds to a catalytic RING finger
(RBX1), which recruits E2 and SKP1 that interact with an F-box protein (194). Besides
the SKP1-binding domain, F-box proteins also contain a substrate-binding domain that
recognizes specific phosphorylations within its target proteins (194). Currently
approximately 69 F-box proteins have been identified in humans, and each of these
proteins can target multiple substrates for degradation (195).
hCDC4 (hAGO/FBXW7) is an example of a F-box protein previously shown to
play a critical role in the regulation of multiple oncoproteins, including cyclin E1 (Figure
3-1), c-Myc, c-Jun, Notch, and mTOR (196-200). hCDC4 possesses three isoforms:
hCDC4-α, β, and γ, each containing a unique N-terminal protein domain that is likely
responsible for cellular localization and tissue expression. All three isoforms share a
common C-terminal region, which consists of 7 WD repeat domains that are responsible
for substrate recognition (193).
77
Figure 3-1. Pathway of hCDC4-mediated degradation of Cyclin E. The enzymes Cdk2 and cyclin E are important for the G1-S transition of the cell cycle. Upon completion of this transition, cyclin E is ubiquitylated through its interaction with E1 (ubiquitin-activating enzyme), E2 (ubiquitin-conjugating enzyme), and E3 complex (ubiquitin ligase, consisting of Cul1, Skp1, and Rbx1), and degraded through the 26S proteosome. hCDC4 F-box protein contains WD domains that recognize phosphorylated cyclin E and delivers cyclin E to the E3 complex. In tumour cells, mutation of hCDC4 prevents recognition of cyclin E, leading to its aberrant accumulation, thus deregulation of the cell cycle (Adapted from Schwab and Tyers, 2001, ref (201)).
78
The significance of hCDC4 in cancer development is further made evident by the
observations of mutations within the hCDC4 gene in several human neoplasias, such as
colorectal and endometrial tumours, as well as cancer cell lines, including ovarian cancer
cell lines (202-205). Based on these observations, hCDC4 has been suggested to be a
tumour suppressor and thus its deregulation may be important in cancer development.
Interestingly, as discussed in Chapter 2, my present familial ovarian cancer
microarray analysis revealed that the hCDC4 gene was less expressed in strong familial
group compared to weak familial group (Figures 2-4, Chapter 2).
In this Chapter, I discuss the genetic approaches that I have taken to investigate
the mechanisms that may be responsible for the altered gene expression of hCDC4.
First, I performed protein truncation test, single strand conformation polymorphism
(SSCP) analysis and manual sequencing to determine the mutation status of the coding
regions of hCDC4 of ovarian tumours. I also examined the promoter methylation status
of these tumours by performing methylation-specific PCR (MSP). Additionally, I
sought to examine the loss of heterozygosity (LOH) status of ovarian tumours by
performing LOH analysis. The protein expression of hCDC4 in ovarian tumours was
also examined by immunohistochemistry (IHC) staining of ovarian cancer tissue
microarrays (TMA). Lastly, the gene expression of cyclin E, a downstream target of
hCDC4, was also examined in familial ovarian cancer.
79
3.2 Materials & Methods
3.2.1 Ovarian tumour samples, and RNA, DNA Extraction
Ovarian tumour sample acquisition and RNA isolation were as described in
Chapter 2 (section 2.2.3). One additional tumour from patient with strong family history
was acquired after microarray analyse, thus a total of 28 ovarian tumour samples were
used for experiments described in this Chapter. Total genomic DNA was isolated from
each sample with QIAamp® DNA mini kit (Qiagen) according to the manufacturer’s
protocol.
3.2.2 Protein Truncation Test
To examine the entire coding sequence of the hCDC4 gene, two pairs of primers
were used to amplify fragments of 1425bp (spanning exons 1-8) and 1431bp (spanning
exons 4-11) containing an overlap of 588 bp. Three microlitres of reverse transcribed
cDNA was utilized in each reaction. The forward and reverse primers flanking exons 1-8
were 5’-
GCTAATACGACTCACTATAGGAACAGACCACCATGATGATGAGCTGGCTTTT
GGAAATGAA-3’ and 5’-ATGCATACAACGCACAGTGG -3’; and the primers
flanking exons 4-11 were 5’-
GCTAATACGACTCACTATAGGAACAGACCACCATGATGATGGAACCCCAGTT
TCAACGAGAC-3’ and 5’-CAACATCCTGCACCACTGAGAACAAGG -3’. The
forward primers included a leader sequence consisting of a bacteriophage T7
transcription promoter plus eukaryotic translation initiation signals. The ATG-initiation
codon was in frame and upstream of the natural translation initiation site. The PCR
80
reaction was carried out in a 25 μl reaction volume containing 1.6-2mM MgCl2, 0.6 mM
dNTPs, 2.5 U platinum Taq Polymerase (Promega), and 9 pmol of each primer. PCR
parameters were: 94°C denature for 2 minutes, followed by two steps of PCR, 4 cycles of
94°C for 30 seconds, 69°C for 30 seconds and 72°C for 1.5 minute; 35 cycles of 94°C for
30 seconds, 58°C for 30 seconds and 72°C for 1.5 minute. A sample of of the PCR
product (2.4 μl) was then used in the coupled transcription and translation reaction in
accordance with the recommendations of the supplier (TNT Coupled Reticulocyte Lysate
System; Promega). After in vitro transcription/translation incorporation of radioisotope
35S-methionine, the incorporated protein products were electrophoresed on a 12.5% SDS
polyacrylamide gel. The gel was then fixed, dried and subjected to autoradiography for 3-
12h. In addition to hCDC4, the BRCA1 gene with a known truncation mutation was used
as positive control to test the sensitivity of the PTT method.
3.2.3 Single Strand Conformation Polymorphism (SSCP) and Manual Sequencing
All coding exon primers for SSCP were designed individually by using Primer3
input software (Whitehead Institute, Howard Hughes Medical Institute, NIH). Genomic
DNA was used as a template for polymerase chain reaction (PCR) amplification of
fragments containing an exon and its adjacent intronic boundaries. Placenta genomic
DNA was used as control. The sequences of each primer set used to amplify exons are
listed in Table 3-1. Some exons were multiplexed (exons 3 and 7, exons 4 and 9, exons 6
and 11, exons 8 and 10). 50ng of tumour DNA was added to a reaction buffer containing
10mM Tris, pH 8.3, 50mM KCl, 1.8-2.6 mM MgCl2, 0.4mM of dNTPs, 6-9 pmol of each
primer, 1mCi 33P-dATP (10mCi/μL, Amersham, USA), and 2 units of AmpliTaq (Perkin
81
Elmer, Norwalk, CT). 33P-dATP incorporated PCR products were heat denatured and
electrophoresed on a 37.5% native polyacrylamide gel containing 10% glycerol. Results
were obtained following autoradiography. Sequence alterations which were detected as
electrophoretic mobility shifts on SSCP gels were confirmed and characterized
independently by direct manual sequencing of PCR products using the same SSCP
primers (ThermoSequenase cycle sequencing kit; Amersham Life Science, Arling
Heights, IL).
Table 3-1. Primer sets for hCDC4 SSCP and manual sequencing analyses.
Exon Forward Primer (5’ – 3’) Reverse Primer (5’ – 3’) Product Size (bp)
1 (α) GACGAACTGGAGGCTCTCTG CTCCTCCTCCTCATCCTCCT 311
1 (β) CCCTCGAGTTCTTCTCAGTCA GCAGGCATACACACACAATCA 347
2 TGACTCAAGATTTGATAGTTAGACGA AAACTAAAACACTTTCAGAATCAACTC 217
3 TTTTCCTTTTATCCTTTCTCTCTCTC GCAGCAATTAAGTGAGGCATT 233
4 GCCTGTAATTTGGGACATCTG CAAATAACACCCAATGAAGAATG 231
5 TCAAGTATCTCATCCTGTGGAGAA TTTCAGAATCACTCTGCTTTTCA 283
6 TGGTGAAGGCAATTTACTCTTG AACGGTTTCTGTTACATTGTGC 210
7 CATATTTCTAATCTGCACATCTTTCTT TGACTTTGTGAAGTGTAGGAAGAG 178
8 AAGTAATCATCTTAAGTGTTTTTCCAG CCAACCATGACAAGATTTTCC 235
9 TTTTTCTGTTTCTCCCTCTGC TTCATCAGGAGAGCATTTAAGG 290
10 TCAGTAATTGATAGGAAGAGTATCCA AACAAAACGAAAGGTGAGTAAGAC 298
11 CCAGTAATTAAATTCTTTTGGTTTTTG TGGACAAATTCATCTTTTCTGCT 326
82
3.2.4 DNA Methylation-Specific PCR
hCDC4 promoter methylation was assessed using the bisulfite PCR method.
Bisulfite modification of DNA was carried using the EpiTect Bisulfite kit (Qiagen)
according to the manufacturer’s protocol to detect cytosine methylation. In brief,
genomic DNA from all ovarian tumour samples was treated with sodium bisulfite to
convert any unmethylated cytosine residues into uracil, giving rise to different DNA
sequences for methylated and unmethylated DNA. Primer pairs for bisulfite sequencing
PCR (Table 3-2) were designed using MethPrimer software (206). Bisulfide-modified
DNA was then subjected to PCR amplification using the following thermal PCR
conditions: initiation step at 94ºC for 4 minutes, 40 cycles of 95ºC for 30 seconds, 50ºC
for 30 seconds and 72ºC for 30 seconds, followed by an extension at 72ºC for 10 minutes.
CpGenome Universal Methylated and Unmethylated DNA (Chemicon, Billerica, MA,
USA) were used as positive and negative controls, respectively. PCR products were
electrophoresed on a 1.5% agarose gel. Methylation was determined by the presence or
absence of a DNA band.
Table 3-2. Primer sequences for methylation-specific PCR.
Primer Forward Primer (5’-3’) Reverse Primer (5’-3’) Product Size (bp)
hCDC4_1m GCGGTAGTTTAGGTTCGATTC CTCTAACGCGCTCTAATAACG 170
hCDC4_1u GGGTGGTAGTTTAGGTTTGATTT CCTCTAACACACTCTAATAACACT 173
hCDC4_2m AGGCGAGAGTTTCGTATAGAGC TCTAACTCCGACTCCGACGTA 248
hCDC4_2u AGGTGAGAGTTTTGTATAGAGTGA ACTCTAACTCCAACTCCAACATA 250
hCDC4_3m GTTGTCGTTTGGTTTAGC GATAC GGGTTGTTGTTTGGTTTAGTGATAT 102
hCDC4_3u GGGTTGTTGTTTGGTTTAGTGATAT ATAAATTAATTCCCTTCCTCCTTCA 107
83
3.2.5 Loss of Heterozygosity (LOH) Analysis of hCDC4
Genomic DNA from four ovarian tumour samples, paired with DNA from an
adjacent area of normal tissue was available for hCDC4 LOH analysis. Allelic loss was
evaluated using 3 polymorphic markers D4S1548, D4S2934 and D4S3049, which closely
flank the hCDC4 locus on chromosome 4q31.3. Primers sequences used for LOH are
listed in Table 3-3.
Table 3-3. Polymorphic markers used for LOH analysis of hCDC4.
Polymorphic Marker Forward Primer (5’-3’) Reverse Primer (5’-3’)
D4S1548 TGCCATAAACAAGGTGAAAC TTACCCAACTGCTACACCAT
D4S2934 CAAAACAGATCAGGATGTGG TTGCTGTCTTTACAGAGCACC
D4S3049 ATTCAGTTCTCTGCGAATG AGTTCGTGCCACTGTACTC
Additional information about these loci was obtained from UniSTS
(http://www.ncbi.nlm.nih.gov/sites/entrez?db=unists). The locations of primers and the
genes that lie between microsatellite repeat markers were determined by Map Viewer at
the National Center for Biotechnology Information (NCBI) Web site
(http://www.ncbi.nih.gov/Tools/index.html). PCR amplification was performed in a final
volume of 30 μl containing 1 μl (200 ng) of DNA template, 1x High Fidelity PCR Buffer,
2 mM MgSO4, 0.2 mM of each dNTP, 0.3 mM of forward and reverse primers, 1 U of
Platinum Taq DNA polymerase High Fidelity (GIBO BRL, Life Technologies, Canada),
and 0.1 μCi of [33P] dATP (Perkin-Elmer, USA). Thermal conditions were as follow:
initiation denaturation step at 95oC for 2 minutes, 40 cycles of 95oC for 15 seconds, 54oC
(D4S1548, D4S2934) or 55oC (D4S3049) for 15 seconds, and 72oC for 20 seconds,
84
followed by an extension at 72oC for 2 minutes. A stop solution (95% formamide,
20mM EDTA, 0.05% bromophenol blue, 0.05% xylene cyanol FF) was added to each
reaction, heat denatured and subjected to electrophoresis on a 7% denaturing formamide
gel, which was run at 80W for 3 hours. Results were obtained following
autoradiography.
3.2.6 Immunohistochemical (IHC) Staining of Ovarian Tissue Microarrays
Ovarian tissue microarray slides (TMAs) were obtained from collaborator Dr.
Patricia Shaw and were constructed as previously described (149). The array contained
165 duplicate cores from formalin-fixed, paraffin-embedded tissue blocks from the UHN
Ovarian Tissue Bank. Of the 165 ovarian samples on the TMA, 140 were evaulable by
immunostaining, including 15 samples that were common to my familial ovarian cancer
study (4 cases with strong family history, 11 with weak family history).
For IHC, TMA slides were first subjected to microwave heat-retrieval for 20
minutes with 10 mM citrate buffer (pH 6.0), washed with PBS buffer and blocked with
0.3% hydrogen peroxide and 10% normal serum before one-hour incubation with hCDC4
antibody (GenTex) in room temperature at 1/2000 dilution. Antibody concentration was
optimized using ovarian tissue slides prior to IHC on TMAs. Slides were washed and
incubated with secondary antibody at 1:200 for 30 minutes at room temperature.
Following washes, slides were stained with streptavidin-peroxidase for an additional 30
minutes. Staining of slides was visualized using ImageScope software (Aperio
Technologies). Staining was scored blindly based on percentage of stained cells (0: none;
1: 1-24%; 2: 25-49%; 3: ≥50%), intensity of staining (0: negative; 1: light, 2: medium, 3:
85
dark), and site of localization (nuclear, cytoplasmic). Total staining score is the sum of
percentage and intensity scores.
3.2.7 Quantitative Real-time PCR for CCNE1
Method for quantitative real-time PCR was as described in Chapter 2 using
Assay-on-Demand probe primers targeting CCNE1 (Applied Biosystems).
86
3.3 Results
3.3.1 hCDC4 Sequence Alteration Detection by PTT Analysis
To detect mutations that lead to premature translation termination, cDNAs from
28 ovarian tumour samples (10 strong familial, 18 weak familial) were subjected to PTT
analysis. All samples exhibited the same PTT pattern of normal bands with less intense
common bands using either the 1425-bp cDNA fragment spanning exons 1-8, or the
1431-bp cDNA fragment spanning exons 4-11 (Figure 3-2). The breast cancer cell line
T47D with wild-type hCDC4 was used as positive and negative controls, respectively.
3.3.2 SSCP and Sequencing of hCDC4
To investigate DNA sequence variations within the hCDC4 gene, I first analyzed
the genomic DNA from 28 ovarian tumour samples by single-strand conformation
polymorphism analysis. One aberrantly migrated band was identified in primary tumour
OVC345, as shown in Figure 3-3. This altered band corresponded to exon 7 of hCDC4,
which encodes the first of 7 WD domains. Subsequent sequencing analysis revealed a
silent nucleotide substitution (C → T) (Figure 3-4).
87
Figure 3-2. hCDC4 Protein truncation assay. In vitro transcription/translation reaction results in synthesis of hCDC4 protein fragments. A) hCDC4 fragment #1 (translation of exons 1-8) results in synthesis of a 54.3 kDa protein. Six representative ovarian tumour samples are shown in lanes 2-9. No aberrant proteins are identified. T74D breast cancer cell line, which has wild-type hCDC4, was used as reference control (lane 1). B) hCDC4 fragment #2 (translation of exons 4-11) results in synthesis of a 47.3 kDa protein. Nine representative ovarian tumour samples are shown in lanes 2-10. T47D reference control is shown in lane 1. No aberrant proteins are identified.
47.3 kDa
Ovarian Tumour Samples
1 2 3 4 5 6 7 8 9 10 B)
A)
54.3 kDa
1 2 3 4 5 6 7 8 9
Ovarian Tumour Samples
88
Figure 3-3. Genetic analysis of hCDC4 with SSCP. An altered banding pattern was observed for one ovarian tumour sample (lane 4, case OVC345) for exon 7 of hCDC4. Placental DNA was used as control (lane 20.)
Ovarian Tumours
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
89
Figure 3-4. hCDC4 gene sequence alteration found in exon 7. A) Manual sequencing of exon 7 of hCDC4 detected a C → T nucleotide change in ovarian tumour sample OVC345. Placental DNA was used as control. B) Subsequent analysis revealed that this specific nucleotide change corresponds to a silent alteration (Asp400→Asp400, amino acid in red). Alternating colours represent protein translation from different exons. A)
A C G T A C G T G A T G A C/T A
G A
T G
A C A A
C
A C
Placenta OVC345 B) 1 MNQELLSVGSKRRRTGGSLRGNPSSSQVDEEQMNRVVEEEQQQQLRQQEEEHTARNGEVV 61 GVEPRPGGQNDSQQGQLEENNNRFISVDEDSSGNQEEQEEDEEHAGEQDEEDEEEEEMDQ 121 ESDDFDQSDDSSREDEHTHTNSVTNSSSIVDLPVHQLSSPFYTKTTKMKRKLDHGSEVRS 181 FSLGKKPCKVSEYTSTTGLVPCSATPTTFGDLRAANGQGQQRRRITSVQPPTGLQEWLKM 241 FQSWSGPEKLLALDELIDSCEPTQVKHMMQVIEPQFQRDFISLLPKELALYVLSFLEPKD 301 LLQAAQTCRYWRILAEDNLLWREKCKEEGIDEPLHIKRRKVIKPGFIHSPWKSAYIRQHR 361 IDTNWRRGELKSPKVLKGHDDHVITCLQFCGNRIVSGSDDNTLKVWSAVTGKCLRTLVGH 421 TGGVWSSQMRDNIIISGSTDRTLKVWNAETGECIHTLYGHTSTVRCMHLHEKRVVSGSRD 481 ATLRVWDIETGQCLHVLMGHVAAVRCVQYDGRRVVSGAYDFMVKVWDPETETCLHTLQGH 541 TNRVYSLQFDGIHVVSGSLDTSIRVWDVETGNCIHTLTGHQSLTSGMELKDNILVSGNAD 601 STVKIWDIKTGQCLQTLQGPNKHQSAVTCLQFNKNFVITSSDDGTVKLWDLKTGEFIRNL 661 VTLESGGSGGVVWRIRASNTKLVCAVGSRNGTEETKLLVLDFDVDMK
90
3.3.3 hCDC4 Promoter Methylation Analysis
Abnormal hypermethylation within promoter regions can lead to transcriptional
silencing of a gene. To determine the methylation status in the putative promoter and 5’
region in exon 1 of hCDC4, I performed methylation-specific PCR (MSP) in all 28
ovarian tumour samples after sodium bisulfite-modification of genomic DNA. The
predicted CpG islands are shown in Figure 3-5 and a representative result of MSP
reaction is shown in Figure 3-6. For each tumour sample, I performed three separate PCR
reactions comprising of different sets of primers targeting different regions of the putative
promoter region. The sensitivity of MSP method was confirmed with universally
methylated and unmethylated control DNA. No methylation was detected in any of the
28 tumour samples.
3.3.4 Loss of Heterozygosity Analysis of hCDC4
Three different polymorphic markers in close proximity to hCDC4 were used to
evaluate LOH. Of the 28 ovarian tumours used in all of my analyses, four matched
normal samples were available for this part of study. No LOH was observed in these four
paired samples (Figure 3-7).
91
Figure 3-5. Potential methylation sites within the hCDC4 promoter. A 4260-bp putative promoter region overlapping the translational start site of hCDC4 gene was analyzed via MethPrimer program. This region contains 4 CpG islands, one of which overlaps with the ATG start codon, with each having an observed/expected CpG ratio > 0.60. MethPrimer was also used to select primers for methylation-specific PCR (MSP), with CpG islands as input parameter. Graph depicts GC%, positions of CpG islands, and CpG sites. Red arrow at position 3000-bp indicates ATG start codon of hCDC4. An example of bisulfite PCR primer sets for one CpG island is depicted below. Additional primer sets for other CpG islands were also used for experiments (not shown).
92
Figure 3-6. Evaluation of hCDC4 promoter methylation by methylation-specific PCR. Genomic DNA from ovarian tumours was treated with bisulfite, and standard PCR was performed using primers specific to methylated (m) an unmethylated (u) DNA to detect methylation within the hCDC4 promoter region. CpGenome Universal Methylated (m-DNA) and Unmethylated DNA (u-DNA) were used as control templates. Five representative tumour samples are shown. OVC29 OVC54 OVC109 OVC161 OVC197 m-DNA u-DNA
m u m u m u m u m u m u m u
93
Figure 3-7. Loss of heterozygosity (LOH) analysis of hCDC4 in four cases of ovarian cancer. Image below illustrates LOH analysis of DNA from one ovarian tumour sample (lane 3) along with DNA from its adjacent normal tissue (lane 4). No LOH was observed in any of the ovarian samples. Breast cancer cell line T47D (lane 1) and ovarian cancer cell line OVCAR3 (lane 2) were used as method controls. 1 2 3 4
94
3.3.5 hCDC4 Protein Expression in Ovarian Cancer
In order to examine hCDC4 protein expression in primary ovarian tumours, IHC
staining was conducted on ovarian tissue microarray (TMA), in collaboration with Dr.
Patricia Shaw. Optimization of hCDC4 antibody was performed using formalin-fixed
paraffin embedded (FFPE) normal breast tissue that expresses normal levels of hCDC4
and a FFPE breast cancer cell line SUM149PT mutated for hCDC4 (Figure 3-8).
Positive nuclear staining of hCDC4 was observed for normal breast tissue and negative
staining was observed for SUM149PT.
The TMA slide contains 165 ovarian tumours of various histological subtypes.
Of the 140 evaluable tumours on the array, 15 samples overlapped with those used in the
gene expression microarray study, including four ovarian tumour from patients with
strong family history and 11 tumours from those with weak family history.
Nuclear and cytoplasmic staining was scored based on staining intensity and the
percentage of stained cells (Figure 3-9). hCDC4 protein expression varies among ovarian
tumours; it was observed that ~50% of total ovarian tumours express hCDC4 at low
levels, with histological scores of 0, 1, or 2, while ~23% express at high levels, with
histological scores of 5 or 6 (Table 3-4). As expected, for those cells that express
hCDC4 protein, staining was primarily localized in the nucleus (Figure 3-9E). Due to
the small number of representative familial ovarian tumours on the TMA, however,
differential protein expression between strong and weak familial ovarian tumours was
indeterminate, with histological score of 2.5 and 2.27, respectively (p = 0.44).
95
Figure 3-8. hCDC4 IHC staining optimization. A) Formalin-fixed, paraffin-embedded (FFPE) normal breast tissue stained for hCDC4 antibody shows positive nuclear staining. B) FFPE breast cancer cell line SUM149PT with hCDC4 mutation shows negative staining. (High power, 40X) Positive hCDC4 staining Negative hCDC4 staining
A B
96
Figure 3-9. Immunohistochemical staining of ovarian tissue microarray with hCDC4 antibody. Representative cases of ovarian tumour specimens illustrating different IHC intensity scores: A) negative = 0; B) mild = 1; C) moderate = 2; D) high = 3. E) Close up image of D. As expected, hCDC4 protein is predominantly localized in the nucleus (arrow).
A B
C D
E
100 µm 100 µm
100 µm 100 µm 50 µm
97
Table 3-4. Histological scores of hCDC4 immunohistochemical staining on ovarian tissue microarray. Various levels of hCDC4 protein expression are observed across different histological subtypes of ovarian cancer. Histological score represents sum of stain intensity (0-3) and % of stained cells (0, 1: 1-24%, 2: 25-49%, or 3: 50-100%). *MMMT=Malignant Mixed Mullerian Tumour
Histological Score Histological Subtype 0 1 2 3 4 5 6
Total # of Tumours
Clear Cell 7 (58%) 0 0 0 3 (25%) 0 2 (17%) 12 Endometrioid 2 (17%) 1 (8%) 1 (8%) 1 (8%) 2 (17%) 1 (8%) 4 (33%) 12 Mixed 2 (20%) 0 2 (20%) 2 (20%) 1 (10%) 3 (30%) 0 10 MMMT* 2 (66%) 0 0 0 0 1 (33%) 0 3 Mucinous 4 (33%) 0 2 (17%) 0 3 (25%) 1 (8%) 2 (17%) 12 Serous 32 (40%) 0 11 (14%) 8 (10%) 12 (15%) 11 (14%) 6 (8%) 80 Transitional Cell 1 (50%) 0 0 0 1 (50%) 0 0 2 Undifferentiated 0 0 0 1 (100%) 0 0 0 1 Other/Unknown 3 (38%) 0 1 (20%) 2 (25%) 1 (20%) 1 (20%) 0 8 Total 53 1 17 14 23 18 14 140
98
3.3.6 Gene Expression of CCNE1 in Familial Ovarian Cancer
As mentioned earlier, hCDC4 is involved in the degradation of cell cycle
regulator protein cyclin E; therefore, an evaluation of cyclin E protein expression in the
same ovarian tumours would be useful in determining if decreased levels of hCDC4
expression correlate with increased cyclin E expression levels. However, due to the lack
of ovarian tumour samples for immunohistochemical analysis, the protein expression
levels of cyclin E could not be determined. Nonetheless, the deregulation of cyclin E at
the transcriptional level has also been previously observed in ovarian cancer (81).
Therefore, I sought to examine CCNE1 gene expression levels in familial ovarian
tumours. Since the CCNE1 gene was not represented on the microarray platform used for
my familial ovarian cancer gene expression profiling analysis (Chapter 2), to assess and
compare the gene expression levels of cyclin E in tumours from ovarian cancer patients
with strong and weak family history, I performed quantitative real-time PCR using
mRNA extracted from ovarian tumour samples. As seen in Figure 3-10, the average gene
expression of CCNE1 observed in the weak familial group was two-fold higher than the
average expression seen in the strong familial group (p < 0.05).
99
Figure 3-10. CCNE1 gene expression in familial ovarian cancer. The mean gene expression of CCNE1 in familial ovarian tumours was evaluated using quantitative real-time PCR. Tumours from patients with weak family history of breast and/or ovarian cancer (18 samples) expressed a higher level of CCNE1 compared to the strong family history group (10 samples). (n=3, p < 0.05).
0
0.5
1
1.5
2
2.5
3
3.5
4
Strong Familial Weak Familial
Rel
ativ
e C
CN
E1
Exp
ress
ion
100
3.4 Discussion
As described in Chapter 2, we have identified hCDC4 as a differentially
expressed gene among ovarian cancer patients with strong and weak family histories of
breast and/or ovarian cancer using the data from my microarray analysis of familial
ovarian cancer. More specifically, ovarian tumours from the strong family history group
exhibited lower hCDC4 expression compared to the weak familial group (p < 0.05). This
observation was of great interest as accumulating evidence from recent years has
suggested human F-box protein hCDC4 to be a tumour suppressor. Functionally, it
mediates the ubiquitin-dependent proteolysis of various oncoproteins involved in cell
division and cell fate determination, thus its deregulation can lead to tumourigenesis
(207).
The direct role of hCDC4 in relation to ovarian cancer was first implicated in a
study conducted by Moberg et al., in which hCDC4 mutations (including nonsense,
missense, and frameshift mutations), were detected in various ovarian cancer cell lines
(205). Based on Moberg’s observations, I sought to examine the mutation status of
hCDC4 in familial ovarian tumours.
From my genetic screening of the 28 primary ovarian cancers using SSCP, I
observed one nucleotide change within exon 7 of one ovarian tumour sample that belongs
to the strong familial ovarian cancer group (OVC345). This initial finding was promising
since exon 7 encodes one of the WD domains responsible for substrate recognition.
However, subsequent DNA sequencing revealed a nucleotide change that did not alter the
hCDC4 protein sequence. This specific previously unreported alteration may be a
significant event nevertheless, as previous studies have suggested that synonymous
101
polymorphisms can affect the timing of cotranslational folding of proteins (208), thus it is
possible that this change may alter the structure of substrate recognition site of hCDC4.
However, this hypothesis remains to be investigated.
Simultaneously to my genetic analyses of hCDC4, other groups have also sought
to examine the mutation status of hCDC4 in human primary ovarian cancers. In a
comprehensive genetic screen of over 1500 human tumours, Akhoondi and colleagues
observed that hCDC4 is mutated in a variety of human malignancies, with an overall
mutation frequency of ~6% (207). However, none of the 32 primary ovarian tumours
from the study exhibited any hCDC4 mutations (207). Similarly, Kwak et al. reported in
their study comprising 111 primary ovarian tumours that mutations of hCDC4 is a rare
event in ovarian cancer (209). More specifically, they observed 2/95 (~2%) sporadic
ovarian cancer cases that harbour mutations, while no mutations were observed in any of
the 16 cases of familial ovarian cancer (209).
Based on the results from my current study, as well as other studies mentioned
directly above, it appears that hCDC4 mutational inactivation is an uncommon event in
primary ovarian tumours. However, other gene regulation mechanisms, such as promoter
hypermethylation, may be responsible for the decreased hCDC4 expression observed in a
subset of familial ovarian tumours, as numerous reports have investigated the importance
of epigenetics in the inactivation of tumour suppressor genes involved in cancer
development, including ovarian cancer (210, 211). To explore this possibility, I
examined the methylation status of the hCDC4 promoter region in familial ovarian
tumours.
102
While multiple potential methylation sites were identified for the promoter region
of this gene, no methylation was detected in any of the ovarian tumours analyzed, thereby
suggesting that promoter methylation may not be an important mechanism for hCDC4
repression in familial ovarian cancer. Recently, a study regarding the epigenetics of
hCDC4 promoter in glioma cell lines found that the low gene expression of hCDC4-β
(the hCDC4 isoform expressed in the brain) seen in various glioma cell lines was not due
to promoter hypermethylation (212). Interestingly, they also showed that normal
peripheral blood cells, which dominantly express α- and γ-isoforms of hCDC4, exhibited
increased methylation within the hCDC4-β promoter (212). Therefore, while methylation
may not be a significant event in repressing hCDC4 gene expression in cancer cells, it
nevertheless plays a role in controlling its tissue- and isoform-specific gene expression.
The hCDC4 gene is suggested to be a haploinsufficient tumour suppressor gene,
and this stems from the observation that in many primary tumours and derived cell lines,
hCDC4 mutations occur without a concomitant loss or additional mutations in the second
allele (207). Moreover, it has been demonstrated that a loss of a single copy of this gene
is sufficient for tumour development in p53+/- mice (213). Therefore, a loss of a single
copy of hCDC4 in human cancers may occur through gene deletion and this deletion may
explain its decrease in gene expression. Indeed, the hCDC4 gene maps to chromosome
4q31.3, a region that is deleted in approximately 30% of human malignancies, including
ovarian cancer (207, 214). Furthermore, LOH within the hCDC4 locus has previously
been reported in esophageal adenocarcinoma and gastric carcinoma (215, 216). Since
currently there are no reports of hCDC4 LOH in human ovarian cancer, I attempted to
examine the LOH status of the familial ovarian tumours from my study. Four ovarian
103
cancer cases with matched normal samples were available for LOH analysis, but no LOH
was detected. However, attributable to the very small sample size, the significance of
hCDC4 LOH in familial ovarian cancer remains inconclusive and requires further
investigations using a larger collection of samples.
In an attempt to address whether the differential gene expression observed
between strong and weak familial ovarian tumours correlates with hCDC4 protein
expression, immunohistochemical analysis was performed with ovarian TMA. The
limited number of corresponding familial ovarian tumours represented on the TMA,
however, prevented a robust analysis of this specific comparison. Nonetheless, our IHC
analysis revealed that hCDC4 expression varies among ovarian tumours. Interestingly,
low hCDC4 protein expression was observed in a majority of clear cell tumours,
mucinous tumours, malignant mixed mullerian tumours, as well as serous carcinomas
(subtype from current familial ovarian cancer study). The low expression seen in these
tumours indicates that loss of hCDC4 may be a significant event in the development of
these tumour subtypes. Further correlations on the expression of downstream effectors of
hCDC4 may reveal the specific pathways that are affected in these tumours.
As mentioned earlier, hCDC4 plays a critical role in the proteolytic regulation of
various proteins, including cell cycle regulator cyclin E1 (CCNE1), and the accumulation
of cyclin E1 due to deregulation by hCDC4 has certainly been associated with an increase
in chromosomal instability, an important event that contributes to malignant
transformation and cancer progression (193, 202). Indeed, it has previously been shown
that the cyclin E gene is frequently amplified and over-expressed in ovarian tumours at
104
both the RNA and protein levels, and that its expression is negatively correlated with
survival, thus implicating its importance in ovarian cancer development (80, 217, 218).
Based on my hCDC4 gene expression results, as well as its characterized role in
cyclin E1 regulation, I hypothesized that in strong familial cases of ovarian cancer which
express a low level of hCDC4, would have an increased protein expression of cyclin E1.
For that reason, I was interested in examining the protein expression of cyclin E1 in a
large collection of ovarian tumour samples, as it would be ideal to be able to examine if
its expression negatively correlates with hCDC4 expression. However, due to the lack of
TMA for cyclin E1 staining, I was unable to perform this particular experiment.
Nevertheless, I sought to examine the gene expression of CCNE1 in my set of ovarian
tumour samples, since ~20% of ovarian cancers have previously been shown to have
increased CCNE1 mRNA (219). My results showed that the mRNA expression of
CCNE1 was two fold higher in the weak familial ovarian group compared to the strong
familial group (p < 0.05). This is an interesting observation, as it suggests that
expression of cyclin E1 in ovarian cancer may be regulated differently at the gene and
protein levels, according to family history status.
Based on the expression data of hCDC4 and CCNE1, it can be speculated that
there may be some yet to be defined heritable factors (eg. SNP alleles) in strong familial
ovarian cancer that play roles in decreasing the gene expression of hCDC4, which may
result in increased steady-state levels of cyclin E1 protein. Additionally, other genetic
alterations such as gene amplification or increased transcriptional activities may be
responsible for increasing CCNE1 gene expression in weak or non-familial ovarian
cancers where hCDC4 expression is not altered. Previous reports have strongly
105
suggested that proper regulation of cyclin E1 is important in preventing malignant
transformation in ovarian cancer (220-222); however, whether or not heritable genetic
alterations that can affect cyclin E1 activity can predispose individuals to an earlier onset
of ovarian cancer requires further investigation.
The data presented in this Chapter suggested that deregulation of hCDC4 gene
expression may be important in a subset of ovarian cancers in patients with a strong
family history. However, the mechanism(s) responsible for this altered expression
remain elusive. Thus further research exploration using larger sample size of familial
ovarian tumours may reveal additional information regarding the genetic alterations or
regulation of hCDC4.
CHAPTER 4
Characterization of PRKCZ in Ovarian Cancer
All experiments and analysis presented in this Chapter was performed by KS.
106
107
4.1 Introduction
As discussed earlier in Chapter 2, my microarray study of familial ovarian cancer
has identified a set of differentially expressed genes between tumours from patients with
strong and weak family history of breast and/or ovarian cancer. Among these genes,
PRKCZ was identified as more highly expressed in patients with strong family history,
and I have chosen this gene for further analyses.
PRKCZ encodes a protein belonging to the atypical subclass of the protein kinase
C family of serine/threonine kinases that has been implicated in the regulation of cellular
transformation and carcinogenesis (172). PRKCZ has previously been observed to be
involved in multiple signal transduction pathways, including activation of the
ERK/MAPK cascade, p70 ribosomal S6 kinase signalling cascade, transcription factor
NF-κB, as well as regulation of cell polarity (175). The regulation of these pathways
may explain some of the mechanisms by which PRKCZ can promote human cancers.
Indeed, the roles of PRKCZ in various cancer types have been examined in recent years.
For example, it was reported that PRKCZ expression level is two fold higher in
glioblastoma cell lines compared with normal astrocytes (178). Subsequent studies
showed that this high level of expression is correlated with increased proliferation of
glioblastoma cells, while reduced expression is correlated with inhibition of migration
and invasion (174, 223). The involvement of activated PRKCZ in epidermal growth
factor (EGF) -induced chemotaxis has also been examined in lung and breast cancer, and
it was shown that PRKCZ is able to elicit a migration response of these cells by acting as
a downstream mediator in the phosphatidylinositol 3-kinase (PI3K)/AKT pathway (176,
177). As mentioned above, PRKCZ participates in cell polarity pathways, and studies
108
have illustrated that loss of cell polarity, which results in tissue disorganization, may
contribute to cancer development (224). Interestingly, it has been observed that PRKCZ
is mislocalized in a subset of ovarian cancers, and it was suggested that this
mislocalization may reflect a role for apical-basal loosening, thus disrupting cell-cell
adhesion, as well as increasing cell growth (173); however, additional evidence
supporting the role of PRKCZ in ovarian cancer remains limited.
The studies mentioned above clearly suggest the importance of PRKCZ in cancer
progression, thus providing rationale for further analyses. In this Chapter, I describe the
various in vitro functional experiments that I have performed in order to characterize its
role in ovarian cancer, including cell viability, cell migration, as well as downstream
signalling pathways in which it may be participating.
109
4.2 Materials and Methods 4.2.1 Cell Culture
Ovarian cancer cell lines SKOV3 and OVCAR3 were purchased from American
Type Culture Collection (Manassas, VA). OVCAR3 cells were maintained in RPMI-
1640 medium supplemented with 20% FBS and 0.01 mg/ml bovine insulin. SKOV3
cells were maintained in McCoy’s medium supplemented with 10% FBS. The HEY cell
line was kindly provided by Dr. Theodore Brown (Samuel Lunenfeld Research Institute,
Toronto, ON) and was maintained in MEM-alpha medium containing 10% FBS. All
cells were incubated at 37oC in a humidified atmosphere of 5% CO2 and 95% air.
4.2.2 PRKCZ Expression Vector & Generation of Stable Clones
PCR conditions to amplify human PRKCZ in a 25 µL reaction volume were as
follows: 2.5 µL of 10X Platinum HiFidelity Buffer (Invitrogen), 1.5 µL of 10 mM dNTPs
(Invitrogen), 1.0 µL of 50 mM MgSO4 (Invitrogen), 0.3 µL of 30 µM EcoRI-tagged
forward primer, 0.3 µL of 30 µM SalI-tagged reverse primer, 0.5 µL of Platinum
HiFidelity Taq Polymerase (5U/µL, Invitrogen), 1 µL (50 ng) of pooled cDNA (Table 2-
2), and 17.9 µL of ddH2O. Thermal cycling parameters were as follows: initial
incubation for 2 minutes at 94oC; 40 cycles of 30 seconds at 94oC, 30 seconds at 73oC, 2
minutes at 72oC. PCR products were resolved by 1.0% agarose gel electrophoresis,
visualized under UV, and gel extracted and purified according to the manufacturer’s
protocol (Qiagen). Subsequently, they were transferred to pEGFP-N2 (N-terminal GFP
tag) expression vector (Clontech). Correct PRKCZ sequence within vector was
confirmed by sequencing. Each cell line was transfected with the plasmid vectors
110
PRKCZ-pEGFP or vector controls, using Fugene ® 6 Transfection Reagent (Roche).
Following transfection, cells were cultured with G418 sulfate (400 μg/ml for HEY, 800
μg/ml for SKOV3, 500 μg/ml for OVCAR3). Surviving colonies were individually
selected and maintained in G418 sulfate-containing medium.
4.2.3 Quantitative Real-Time PCR
Primer pairs for genes of interest were designed individually by using Primer3
input software (Whitehead Institute, Howard Hughes Medical Institute, NIH).
Quantitative real-time RT-PCR was performed on an ABI Prism 7000 Sequence Detector
(Applied Biosystems) using SYBR Green PCR Master Mix (Applied Biosystems). Each
of the 20 µL PCR reactions contained 1 µL (50 ng) of cDNA and 0.45 µM of each of the
primers. The thermal cycles for PCR reaction were as follow: initial denaturation for 10
minutes at 95oC, followed by 40 cycles of 95oC for 15 seconds, and annealing extension
at 60oC for 1 minute. The housekeeping gene HPRT1 was used to normalize gene
expression values. Reference cDNA was used to generate standard curve to quantify
cDNA levels of samples and consisted of a pool of 13 cell lines, as described in Chapter 2
(Table 2-2).
4.2.4 Western Blotting
Cell extracts were prepared as followed. Cells were washed three times with cold
phosphate-buffered saline (PBS), lysed with NETN lysis buffer (20mM Tris-HCl, pH 7.5;
150mM NaCl; 1 mM EDTA, pH 8.0; 0.5% Nonidet P-40; 1 mM PMSF; 1x protease and
phosphatase inhibitors) on ice for 10 minutes and centrifuged for 8 minutes at 12,000 rpm
111
to separate lysates from cell debris. Protein concentrations were determined with BCA
Protein Assay Kit (Pierce). Equal amounts of protein from different cell lines were
loaded and separated on 8-10% SDS-PAGE. Proteins were transferred to Hybond ECL
nitrocellulose (Amersham) and blotted using anti-PRKCZ, anti-IGF1R, anti-ITGB3
antibodies (Cell Signalling) at 1:1000 dilutions. Secondary conjugates, HRP-Donkey
anti-mouse or HRP-Donkey anti-rabbit (Jacksons Immunochemicals) were incubated for
1 hour at a 1:5000 dilution. Protein bands were visualized by chemiluminescence using
ECL detection system (Amersham).
4.2.5 MTT Cell Viability Assays
When added to viable cells, MTT (3-(4,5-Dimethylthiazol-2-yl)-2,5-
diphenyltetrazolium bromide, a yellow tetrazole), is reduced to purple formazan by
activated reductase enzymes. Thus the MTT assay can be applied to assess cell viability
based on changes in absorbance of coloured solution. In brief, cells were seeded in 96-
well plates at a concentration of 1000 cells/well with a final volume of 100 µL of culture
media and were incubated at 37oC, with or without myristoylated pseudosubstrate peptide
(40 µM), a PRKCZ inhibitor. After each incubation period, 10 µL of the MTT labelling
reagent (Roche) were added to each well at a final concentration of 0.5 mg/mL. Cells
were incubated for an additional 4 hour period, followed by addition of 100 µL of
solubilization solution (10% SDS in 0.01 M HCl). Plates were allowed to stand overnight
at 37oC and the spectrophotometrical absorbance of the samples was measured using a
microplate (ELISA) reader at a wavelength of 570 nm with background subtraction at
630 nm. Each assay was performed in triplicates.
112
4.2.6 TUNEL Assays
Cells were plated and allowed to grow to confluency in 96-well plates, and serum
starved in 0.1% FBS media for 20 hours, then for an additional 4 hours with or without
TRAIL-induction of apoptosis (100ng/ml). Cells were fixed with 4% paraformaldehyde
in PBS for 1 hour, followed by permeabilization with 0.1% Triton X-100/0.1% sodium
citrate on ice for 2 minutes. TUNEL assay was performed with an In Situ Cell Death
Detection Kit (TMR red) as described by the manufacturer (Roche). Reactions were
stopped after one hour, and apoptotic cells were visualized and measured using IN CELL
Analyzer 1000 (GE Healthcare Life Sciences). Each assay was performed in triplicates.
4.2.7 BrdU Proliferation Assay
Approximately 1 x 105 cells were plated onto coverslips within 6-well plates and
allowed to grow to ~60% confluency overnight. On the next day, cells were incubated in
10uM of BrdU for 6 hours, washed with PBS, then fixed with 4% paraformaldehyde,
denatured with 2M HCl, and neutralized with 0.1 M sodium borate. Cells were then
incubated with mouse anti-BrdU (1:200, Dako) for 1 hour, washed, followed by
incubation with Alexa Fluro 647 anti-mouse (Molecular Probes) for an additional 30
minutes, and counterstained with DAPI. BrdU incorporation was observed and counted
in five fields of view per well through microscopy. Each cell line was performed in
triplicates per experiment and the ratio of BrdU-positive cells to total cell number was
calculated.
113
4.2.8 Matrigel Transwell Assays
In vitro cellular invasion was assayed by determining the ability of cells to invade
through Matrigel (VWR), a synthetic basement membrane, as previously described (177).
In brief, Transwells of 8 µm pore size (Corning) coated with ~30 µg Matrigel (diluted in
ice-cold PBS) were placed in a modified Boyden chamber and ovarian cancer cells (1 x
105 per well) were plated onto each of the transwells in serum-free media. The bottom
chambers were either filled with epidermal growth factor (10ng/ml) in serum-free media,
or serum-containing media (10% FBS), which both acted as a chemoattractant. Cells
were incubated at 37oC and allowed to migrate through the matrigel for 24 hrs. After
incubation, cells were collected from the bottom chamber, as well as the underside of the
Transwell by briefly rinsing transwells in PBS and placing them in fresh wells containing
500 µL trypsin (0.1%). Solutions were microcentrifuged at 8000 g for 5 minutes to
pellet invaded cells. CyQUANT dye (Molecular Probes), which shows strong
fluorescence enhancement when bound to nucleic acids, was used for the quantification
of invaded cells. Fluorescence was measured using a multiwell fluorescence plate reader
with excitation at 485 nm and emission at 530 nm.
4.2.9 Scratch Wound Healing and Pericentrin Orientation Assays
Each cell line (HEY, SKOV3, OVCAR3, and their respective clones) were plated
into 6-well plates and cultured to confluence. Cells were rinsed with PBS and serum
starved overnight in 0.5% FBS media at 37oC and 5% CO2. Next day, three separate
scratches were introduced through the monolayer of cells in each of the wells using
sterile 200 µL plastic pipette tips. Cells were then rinsed gently with PBS to remove
114
cellular debris and replaced with fresh culture media supplemented with 0.5% FBS. The
wounded cells were allowed to incubate at 37oC and representative fields were
photographed with an inverted-phase microscope at different time intervals. For
pericentrin orientation assay, cells were fixed with 4% paraformaldehyde in PBS four
hours after wound scratch, followed by one hour incubation with blocking solution (0.2%
Triton X-100, 1.5% BSA, 5% serum), and stained with anti-pericentrin and anti-tubulin
antibodies (1/5000 dilutions) for one hour at room temperature and counterstained with
DAPI. Between incubation steps, cells were washed several times with PBS. Cells with
microtubule organizing centre (MTOC; stained with anti-pericentrin) situating within
quadrants facing the wound were scored as positive for polarity (225).
4.2.10 Phagokinetic Track Assays
Single cell motility of HEY, OVCAR3 and SKOV3 were measured according to
Cellomics ® Cell Motility Kit’s protocol (Thermo Scientific), as previously described.
Briefly, 96-well plates were coated with fibronectin (1 μg/well) overnight followed by
fluorescent microspheres the next day. Approximately 500 cells were plated in each well
in 100 μl of serum-free media. Following 16 hours of incubation, at 37oC in a humidified
atmosphere of 5% CO2 and 95% air, wells were fixed with 5.5% formaldehyde and cells
were stained with phalloidin. The area of a phagokinetic track from a single cell was
quantified using Cellomics Arrayscan II microscope and software.
115
4.2.11 siRNA Transfections
Knockdown of expression of PRKCZ, IGF1R, and ITGB3 in ovarian cancer cell
lines was achieved by transfection of siRNAs (Ambion). siRNAs targeting these genes
was performed with Dharmafect-4 transfection reagent (Dharmacon). In brief, cells were
seeded in 12-well or 6-well plates at densities of 1 x 105 or 2 x 105 cells/well,
respectively. Cells were then treated with siRNA transfection mixtures following the
manufacturer’s protocol. Scrambled siRNA (Ambion) was used as a control. Additional
controls included mock-treated cells that received transfection reagent without siRNA, as
well as untreated cells that received only fresh media. Cells were harvested after 48 or 72
hours for RNA and protein extraction, respectively.
4.2.12 Ingenuity Pathway Analyses
Ingenuity Pathway analyses were performed as described in Chapter 2.
4.2.13 Statistical Analyses
All data were represented as means ± the standard deviation (SD) of the mean.
Statistical calculations were performed with Microsoft Excel analysis tools. Differences
between groups were analyzed by student t-test. P values of < 0.05 were considered
statistically significant.
116
4.3 Results To characterize the roles of PRKCZ in ovarian cancer, I developed an in vitro
ovarian cancer model with altered expression of PRKCZ by generating stable cell lines
that over-express PRKCZ, as well as gene knockdown by PRKCZ-targeted siRNA,
followed by various functional assays. Three serous ovarian cancer cell lines HEY,
SKOV3 and OVCAR3 were chosen for my initial analyses; however, based on my initial
results, as well as its moderate migratory phenotype (versus extremely motile and non-
motile characteristics of HEY and OVCAR3 cells, respectively), most of the later
experiments focused on the SKOV3 cell line.
4.3.1 Generation of PRKCZ Stable Ovarian Cancer Cell lines
I first examined the endogenous transcript levels of PRKCZ in HEY, SKOV3, and
OVCAR3 ovarian cancer cell lines by performing quantitative RT real-time PCR. Levels
of PRKCZ transcript varied among cell lines, with HEY cells expressing the lowest level
of PRKCZ, followed by SKOV3 and OVCAR3 (Figure 4-1a). Protein levels of PRKCZ
in all three cell lines were very low compared to THP-1 cells, a human acute monocytic
leukemia cell line that expresses high endogenous levels of PRKCZ (Figure 4-1b).
117
Figure 4-1. Endogenous transcript and protein levels of PRKCZ in selected ovarian cancer cell lines. a) HEY, SKOV3, and OVCAR3 cells have variable gene expression levels of PRKCZ. HEY cell line has the lowest PRKCZ transcript level, while SKOV3 and OVCAR3 express have similar transcript levels. b) PRKCZ protein is expressed poorly among all three cell lines, as determined by western blot analysis. Lysate from each cell line was loaded in duplicates in adjacent lanes. Human acute monocytic leukemia cell line THP-1, which expresses a high level of PRKCZ protein, was used as a positive control. The beta-actin loading control is also shown (lower panel). a)
0
0.5
1
1.5
2
2.5
HEY SKOV3 OVCAR3
Ovarian Cancer Cell Line
Rel
ativ
e E
xpre
ssio
n
b)
HEY SKOV3 OVCAR3 THP-1
PRKCZ
Beta-actin
118
After PRKCZ transfection and selection of stable clones, the gene and protein
expression of PRKCZ stable clones for each ovarian cancer cell line were verified by
quantitative real-time PCR and western blot analyses using a PRKCZ-antibody (Figures
4-2, 4-3, 4-4). The expected differences in PRKCZ expression levels between non-
transfected or empty-vector transfected and PRKCZ-transfected cells were observed. A
band of ~95 kDa was detected in stable cell lines that over-express PRKCZ, which
corresponded with molecular weight of GFP-tagged PRKCZ protein. Additionally,
fluorescent microscopy was used to detect exogenous expression and revealed a
cytoplasmic localization of GFP-tagged PRKCZ in transfected cell lines (Figures 4-2, 4-
3, 4-4).
119
Figure 4-2. Expression of PRKCZ clones in HEY ovarian cancer cell line. a) Comparison of PRKCZ transcript and protein expression in HEY parental control (PC), control vector (N2), and a selection of PRKCZ clones by a) quantitative real-time RT-PCR, b) western blot analyses with PRKCZ-antibody, and c) fluorescence microscopy (magnification: 20x). In b), a band of ~95 kDa was detected by western blotting in stable cell lines that over-express PRKCZ, which corresponded with molecular weight of GFP-tagged PRKCZ protein. A faint band of ~67 kDa corresponded to endogenous PRKCZ. Shown are HEY parental cells, HEY vector control cells, and HEY-PRKCZ clones #1, #2 and #3 that were chosen for future functional experiments. a)
0
10
20
30
40
50
60
70
~95 kDa
~67 kDa
HEY HEY N2 HEY PRKCZ #1 HEY PRKCZ #2 HEY PRKCZ #3
Rel
ativ
e E
xpre
ssio
n
PC N2 PRKCZ#1 PRKCZ#2 PRKCZ#3
b) PRKCZ clones PC N2 #1 #2 #3 PRKCZ Beta-actin
120
c)
HEY HEY-N2 HEY-PRKCZ
GFP-PRKCZ GFP
121
Figure 4-3. Expression of PRKCZ clones in SKOV3 ovarian cancer cell line. Comparison of PRKCZ transcript and protein expression in SKOV3 parental control (PC), control vector (N2), and a selection of PRKCZ clones by a) quantitative real-time RT-PCR, b) western blot analysis with PRKCZ-antibody, c) fluorescence microscopy (magnification: 20x). In b), a band of ~95 kDa was detected by western blotting in stable cell lines that over-express PRKCZ, which corresponded with molecular weight of GFP-tagged PRKCZ protein. A faint band of ~67 kDa corresponded to endogenous PRKCZ. Shown are SKOV3 parental cells, SKOV3 vector control cells and SKOV3-PRKCZ clones #1, #2, #4, #8 that were chosen for future functional experiments. a)
0
50
100
150
200
250
300
S
~95 kDa
~67 kDa
KOV3 SKOV3 N2 SKOV3 SKOV3 SKOV3 SKOV3PRKCZ#1 PRKCZ#2 PRKCZ#4 PRKCZ#8
Rel
ativ
e E
xpre
ssio
n
PC N2 PRKCZ#1 PRKCZ#2 PRKCZ#4 PRKCZ8
b) PRKCZ clones PC N2 #1 #2 #4 #8
PRKCZ Beta-actin
122
c) SKOV3 SKOV3-N2 SKOV3-PRKCZ
GFP GFP-PRKCZ
123
Figure 4-4. Expression of PRKCZ clones in OVCAR3 ovarian cancer cell line. Comparison of PRKCZ transcript and protein expression in OVCAR3 parental cells (PC), control vector (N2), and a selection of PRKCZ clones by a) quantitative real-time RT-PCR, b) western blot analyses with PRKCZ-antibody, and c) fluorescence microscopy (magnification: 20x). In b), a band of ~95 kDa was detected by western blotting in stable cell lines that over-express PRKCZ, which corresponded with molecular weight of GFP-tagged PRKCZ protein. A faint band of ~67 kDa corresponded to endogenous PRKCZ. Shown are OVCAR3 parental cells, OVCAR3 vector control cells and OVCAR3-PRKCZ clones #1, #2, and #3 that were chosen for future functional experiments. a)
0
2
4
6
8
10
12
14
~95 kDa
~67 kDa
OVCAR3 OVCAR3 N2 OVCAR3 OVCAR3 OVCAR3PRKCZ#1 PRKCZ#2 PRKCZ#3
Rel
ativ
e E
xpre
ssio
n
PC N2 PRKCZ#1 PRKCZ#2 PRKCZ#3
b) PRKCZ clones PC N2 #1 #2 #3
PRKCZ
Beta-actin
124
c) OVCAR3 OVCAR3-N2 OVCAR3-PRKCZ GFP GFP-PRKCZ
125
4.3.2. Cell Viability in PRKCZ-Expressing Cells
PRKCZ has previously been shown to be involved in cell survival in various cell
types (226-230). To assess whether it has the same effect on ovarian cancer cells, I first
performed MTT cell viability assays. By comparing the growth of parental, empty-vector
control, and PRKCZ-expressing cells, I observed that PRKCZ did not have an effect on
HEY or OVCAR cell lines. On the other hand, it significantly increased the viability of
SKOV3 cells, and this effect was abolished by the addition of myristoylated
pseudosubstrate peptide that targets PRKCZ (Figure 4-5). This result suggests that
PRKCZ can enhance cell survival in a subset of ovarian cancer cells.
I then further investigated whether the increased cell viability seen in SKOV3
cells that over-express PRKCZ was due to an increase in proliferation or a decrease in
apoptosis. BrdU proliferation assay was performed to examine if PRKCZ over-
expression results in increased replication rate in SKOV3 cells. As seen in Figure 4-6,
SKOV3 PRKCZ-clones displayed a higher percentage of cells with BrdU incorporation,
indicating that the rate of proliferation in these cells is higher than parental and empty
vector controls.
126
Figure 4-5. PRKCZ increases cell viability in SKOV3 cells but not HEY and OVCAR3 cells. a) Increased cell viability was observed in cells over-expressing PRKCZ, as measured by MTT cell viability assay. This effect was abolished by the addition of PRKCZ myristoylated pseudosubstrate (PS). No significant change in cell viability was observed in b) HEY and c) OVCAR3 cells. (n=3) N2 = empty vector N2
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
2 4 6 8 10 12Day
MT
T A
bsor
banc
e U
nit
OVCAR3
N2
PRKCZ#1
PRKCZ#2
PRKCZ#3
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1 2 3 4 5 6
Day
MT
T A
bsor
banc
e U
nit
SKOV3
N2
PRKCZ#1
PRKCZ#4
SKOV3 (+PS)
N2 (+PS)
PRKCZ#1 (+PS)
PRKCZ#4 (+PS)
a)
*p<0.05 *
* b)
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1 2 3 4
MT
T A
bsor
banc
e U
nit
Day
HEY
N2
PRKCZ#1
PRKCZ#2
c)
127
Figure 4-6. PRKCZ enhances proliferation of SKOV3 ovarian cancer cells. BrdU incorporation assay was performed to measure prolifartion rate of SKOV3 cells. PRKCZ-transfected cells showed an increase in BrdU-positive cells, indicating increased cell growth compared to controls (*p<0.01). This effect was reversed by the addition of PRKCZ myristyolated pseudosubstrate (PS). PC = parental control; N2 = empty vector control N2. (n=3)
0.00
0.05
0.10
0.15
0.20
0.25
0.30
RKCZ#1 PRKCZ#4 PRKCZ#8
TM
R:D
api C
ount
EV P
s
No TrailTrail (100ng/ml)
N2
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
l N2 PRKCZ#1 PRKCZ#4
Brd
U P
ositi
ve C
ells
(%)
* No PS *+PS
Parenta PC Figure 4-7. PRKCZ has no effect on apoptosis in SKOV3 cells. Apoptotic-related DNA fragmentation of SKOV3 cells with and without Trail treatment (100ng/ml) was analyzed with In Situ Cell Death Detection Kit. No significant difference was seen between empty-vector control and PRKCZ clones. (n=2) N2 = empty vector control N2.
128
Additionally, I sought to examine the effect of PRKCZ on apoptosis in SKOV3
cells by performing TUNEL assays. Since cells from the MTT assays were not induced
by any apoptotic agents, I performed TUNEL assays in non-treated cells to examine if
there is a difference in non-induced apoptosis between PRKCZ-expressing and control
cells. Additionally, I was interested in examining if PRKCZ can protect ovarian cancer
cells against induced apoptosis. To achieve this, cells were treated with TRAIL
(100ng/ml) for 4 hours, and apoptotic-induced DNA fragmentation was detected. As
seen in Figure 4-7, while a significant difference in cell death was observed between non-
induced and induced cells, there was no difference in apoptosis between control and
PRKCZ-expressing cells in either case. Based on these results, it can be speculated that
the increased cell viability seen in PRKCZ-expressing SKOV3 cells is most likely due to
an increase in cell proliferation.
4.3.3 PRKCZ and ovarian cancer cell migration and invasion
The ability of cancer cells to migrate is one of the key processes in cancer
progression. In order to examine whether PRKCZ can increase the migratory properties
of ovarian cancer cells, I undertook various experimental approaches. Microscopic
observations revealed that the morphology of each of these cell lines is associated with
their respective migration properties. HEY cells are fibroblast-like cells with filopodia
that tend to undergo single-cell migration and are highly motile; while SKOV3 cells are
also fibroblast-like, they are less motile compared to HEY. OVCAR3 cells, on the other
hand, have cobblestone-like structures, they migrate as a sheet and are the least motile.
129
No invasive differences were observed between parental and PRKCZ-expressing
HEY and SKOV3 cells in transwell migration assays. This specific assay was also
performed for OVCAR3 cells, but due to its low rate of migration, results were
inconclusive. Additional scratch wound healing migration assays were performed to
examine effect of PRKCZ on migratory properties of all three cell lines. No migration
differences were observed in any of these cells when PRKCZ-transfected cells were
compared with controls (Figure 4-8).
To examine if the endogenous levels of PRKCZ in parental cells are sufficient for
their migration, I repeated wound healing assays with the same ovarian cell lines that
have been subjected to PRKCZ siRNA knockdown (Figure 4-9). I observed that
knocking down the transcript levels of PRKCZ can in fact decrease migration rate of
SKOV3 cells. This phenotype was not observed in HEY and OVCAR3 cell lines (data
not shown).
Interestingly, despite the lack of difference in migration rate, a distinct
morphological phenotype was observed in HEY cells that over-express PRKCZ in the
scratch wound healing assay. It was noted that while HEY parental or empty-vector
control cells were both able to migrate to re-establish the monolayer and that their
movements were primarily perpendicular to the wound, the movements of PRKCZ-
expressing HEY cells were disorganized and appeared to travel in various directions
rather than continuing in a perpendicular direction, as observed through cell migration
videos (Figure 4-10). This phenotype was specific to the HEY cell line as PRKCZ-
expressing SKOV3 and OVCAR3 cells did not exhibit differences in rate or direction of
migration.
130
Figure 4-8. Migration of ovarian cancer cells. Migration rate of a) HEY, b) SKOV3, and c) OVCAR3 cell lines as measured by scratch wound assays. No difference in migration was observed between empty vector control and PRKCZ-expressing cells in any of the ovarian cancer cell lines. PC = parental control; N2 = empty vector control N2 a) b)
0
0.1
0.2
0.3
0.4
0.5
0.6
N2 PRKCZ#1 PRKCZ#2 PRKCZ#3
Dist
ance
Tra
velle
d (m
m)
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
N2 PRKCZ#1 PRKCZ#4 PRKCZ#8
Mig
ratio
n di
stanc
e (m
m)
c)
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
N2 PRKCZ#1 PRKCZ#2 PRKCZ#3
Dis
tanc
e Tr
avel
led
(um
)
131
Figure 4-9. Effect of PRKCZ gene knockdown on SKOV3 parental cells migration as observed by wound healing assay. a) Confirmation of PRKCZ gene expression knockdown by quantitative real-time PCR. B) Knockdown of PRKCZ expression by siRNA inhibits migration compared to controls as observed in wound healing assays. (*p<0.01) (n=2) a)
0.0
0.2
0.4
0.6
0.8
1.0
1.2
Rel
ativ
e E
xpre
ssio
n
PRKCZ mRNA Expression
132
b)
PRKCZ siRNA Mock Control
0 hr 0 hr
20 hr 20 hr
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
PRKCZ siRNA Scramble siRNA Mock siRNA
Dis
tanc
e tr
avel
led
(mm
) *
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
PRKCZ siRNA Scramble siRNA Mock siRNA
Dis
tanc
e tr
avel
led
(mm
) *
133
Figure 4-10. Disorganized cell movement of HEY cells over-expressing PRKCZ. Despite similar migration rates between HEY control and PRKCZ-tranfected cells in scratch wound assays, the movement of PRKCZ-expressing cells appeared to be disorganized compared to parental cells (as observed by time-lapse migration video). N2 = empty vector control N2. HEY-N2 HEY-PRKCZ 0 hr 4 hr 8 hr 12 hr
134
Furthermore, to investigate if polarity plays a role in directionality of HEY cells, I
performed polarity assays by measuring the orientation of microtubule organization
centre (MTOC) of the leading edge of cells subjected to scratch wound. However, no
differences were observed between PRKCZ-expressing and control HEY cells (Figure 4-
11).
To examine the effect of PRKCZ on single-cell motility of ovarian cancer cells, I
performed phagokinetic track assays using HEY, SKOV3, and OVCAR3 cell lines
(Figure 4-12). As expected, the areas travelled by each cell line correlate with their
invasive properties, with HEY cells having the highest motility track area, followed by
SKOV3, and OVCAR3 cells being the least motile. Despite the random cell movement
observed in scratch wound healing assay, HEY cells that over-express PRKCZ did not
show an increased level of single-cell motility compared to vector control cells. However,
it is possible that PRKCZ-expressing HEY cells have migrated to areas that they have
previously travelled, thus giving lower measurements of track areas. Interestingly,
phagokinetic track assays showed that SKOV3 cells over-expressing PRKCZ decreased
cell motility compared to the parental cells; however, this decrease in cell motility did not
affect the overall migration rate of these cells. PRKCZ-expressing OVCAR3 cells did
not show a difference in motility compared to control cells.
135
Figure 4-11. Measurement of cell polarity of HEY by pericentrin orientation assay. Monolayer of HEY cells were subjected to scratch wound and fixed at 4 hours post wounding. a) Co-labelling of wounded HEY cells for tubulin (green), pericentrin (red), and nucleus (blue). White arrows depict direction of cell movement. White dotted lines indicate quadrant of cell facing wound. b) The percentage of wound-edge cells with MTOC orientation facing wound. a)
Scratch wound
b)
0
10
20
30
40
50
60
70
80
90
100
N2 PRKCZ#1 PRKCZ#3
% o
f MT
OC
Fac
ing
the
Wou
nd
136
Figure 4-12. Quantitation of ovarian cancer cell motility. Phagokinetic track assays were performed to quantitate the effect of PRKCZ on ovarian cancer cell motility. a) Comparison of cell motility between the three parental ovarian cancer cell lines. HEY exhibited the highest level of motility, followed by SKOV3, while OVCAR3 displayed minimal movement. b) & c) No effect was observed between controls and PRKCZ-expressing HEY and OVCAR3 cells. d) Decreased motility was observed in SKOV3 cells (p<0.05). (n=2) N2 = empty vector control N2 a) b)
0
2000
4000
6000
8000
10000
12000
14000
16000
HEY SKOV3 OVCAR3
Trac
k ar
ea/c
ell (
um2 )
0
5000
10000
15000
20000
25000
30000
HEY N2 HEY PRKCZ#1 HEY PRKCZ#2 HEY PRKCZ#3
Trac
k ar
ea/c
ell (
um2 )
c) d)
0500
100015002000250030003500400045005000
SKOV3 N2 SKOV3PRKCZ#1
SKOV3PRKCZ#2
SKOV3PRKCZ#4
SKOV3PRKCZ#8
Trac
k ar
ea/c
ell (
um2 )
0
50
100
150
200
250
300
350
400
450
500
OVCAR3 N2 OVCAR3 PRKCZ#1 OVCAR3 PRKCZ#2 OVCAR3 PRKCZ#3
Trac
k ar
ea/c
ell (
um2 )
137
4.3.4 Identification of Potential Downstream Effectors of PRKCZ
PRKCZ is involved in various cell signalling pathways, thus may alter multiple
downstream targets. To identify downstream effectors of PRKCZ that may be involved in
ovarian tumourigenesis, I first performed molecular network analyses with the
microarray data to identify potential interactors of PRKCZ (Figure 4-13). Upon
examination of the molecular network as identified by IPA, I decided to focus on IGF1R
(insulin-like growth factor 1 receptor) and ITGB3 (integrin beta 3) in my subsequent
analyses.
4.3.4.1 IGF1R and ITGB3 as Potential Downstream Effectors of PRKCZ
The first step I took was to examine if there is a relationship between PRKCZ,
IGF1R and ITGB3 was to determine if over-expression PRKCZ could exert any effect on
the mRNA and protein expression of IGF-IR in the three different ovarian cancer cell
lines by quantitative real-time PCR and western blot analyses (Figure 4-14). No
differences in transcript or protein expression were observed in PRKCZ-expressing HEY
cells when compared to parental and empty-vector control cells. However, while the
transcript level was not altered in SKOV3 cells, an increase in PRKCZ protein expression
correlated with an increased level of IGF1R protein, suggesting that PRKCZ may
participate in IGF1R translation or protein stability. Additionally, the expression level of
phosphorylated IRS-2 (insulin-receptor substrate-2), a known downstream target for IGF-
IR, was also increased in SKOV3 cells, confirming that PRKCZ is involved in the
activation of IGF1R signalling pathway in this particular cell line.
138
Figure 4-13. Identification of potential interactors of PRKCZ. Ingenuity pathway analysis was performed to identify interactors of PRKCZ using microarray results from Chapter 1. Integrin beta 3 (ITGB3) and Insulin Growth Factor 1 Receptor (IGF1R) were identified as indirect interactors of PRKCZ and were chosen for further analyses.
139
Figure 4-14. Transcript and protein expression of IGF1R in PRKCZ-expressing SKOV3 cells. a) No significant increase in IGF1R transcript level was observed in SKOV3. b) Protein level of IGF1R is elevated in PRKCZ-expressing cells. c) The level of p-IRS-2, a downstream effector of IGF1R, was elevated in 3 of the 4 PRKCZ clones. PC = parental control; N2 = empty vector control N2. Western blot shown is representative of 3 independent analyses. a)
0.0
0.5
1.0
1.5
2.0
2.5
N2 PRKCZ#1 PRKCZ#2 PRKCZ#4 PRKCZ#8
Rela
tive
Expr
essio
n
b)
PRKCZ clones PC N2 #1 #2 #4 #8
PRKCZ
IGF1R
Beta-actin c)
p-IRS2
Beta-actin
140
Interestingly, in contrast to SKOV3 cells, expression of IGF1R was decreased in
OVCAR3 cells that over-express PRKCZ, at both the transcript and protein levels (Figure
4-15).
The transcript and protein expression of ITGB3 was also determined in the three
ovarian cancer cell lines by real-time quantitative RT-PCR and western blot analyses.
The expression of ITGB3 was unaltered in HEY cells that over-express PRKCZ, at both
transcriptional and protein levels. Interestingly, mRNA and protein expression of ITGB3
were significantly decreased in both SKOV3 cells and OVCAR3 cells that over-express
PRKCZ, compared to parental and empty-vector controls (Figures 4-16, 4-17).
The concurrent expression alterations observed for IGF1R and ITGB3 in PRKCZ-
expressing cells prompted the question of whether these changes are occurring within the
same biological pathway. I first performed IGF1R siRNA gene knockdown experiments
in SKOV3 cells to determine if reducing IGF1R expression would have an effect on
ITGB3 gene expression. Results from quantitative real-time PCR analysis indicated that
IGF1R knockdown can lead to de-repression of ITGB3 mRNA expression in cells over-
expressing PRKCZ (Figure 4-18). To further address whether activation of the IGF1-
signalling pathway plays a role in ITGB3 gene regulation, I examined the expression
level of ITGB3 after stimulation of SKOV3 cells with IGF1, a known ligand for IGF1R.
Interestingly, similar to the IGF1R siRNA knockdown experiment, ITGB3 expression
was de-repressed in PRKCZ-expressing cells when stimulated with IGF1 (Figure 4-19).
141
Figure 4-15. Transcript and protein expression of IGF1R in PRKCZ-expressing OVCAR3 cells. Both transcript and protein levels of IGF1R are decreased in PRKCZ-expressing OVCAR3 cells. This observation is in contrast to SKOV3 cells, which showed increased IGF1R expression at protein levels. PC = parental control; N2 = empty vector control N2. Western blot shown is representative of 3 independent analyses.
0
0.2
0.4
0.6
0.8
1
1.2
N2 PRKCZ#1 PRKCZ#2 PRKCZ#3
Rela
tive
Expr
essio
n
PC N2 #1 #2 #3 PRKCZ clones
PRKCZ IGF1R
Beta-actin
142
Figure 4-16. Transcript and protein expression of ITGB3 in PRKCZ-expressing SKOV3 cells. Both transcript and protein levels of ITGB3 are decreased in PRKCZ-expressing SKOV3 cells. PC = parental control; N2 = empty vector control N2. Western blot shown is representative of 3 independent analyses.
0
0.2
0.4
0.6
0.8
1
1.2
N2 PRKCZ#1
PRKCZ#2
PRKCZ#4
PRKCZ#8
Rela
tive
Expr
essio
n
p < 0.05
* *
* *
PRKCZ clones
PC N2 #1 #2 #4 #8
PRKCZ
ITGB3
Beta-actin
143
Figure 4-17. Transcript and protein expression of ITGB3 in PRKCZ-expressing OVCAR3 cells. Similar to SKOV3 cell line, both transcript and protein levels of ITGB3 in OVCAR3 cell line are decreased in PRKCZ-expressing cells. PC = parental control; N2 = empty vector control N2. Western blot shown is representative of 3 independent analyses.
0
0.2
0.4
0.6
0.8
1
1.2
N2 PRKCZ#1 PRKCZ#2 PRKCZ#3
Rela
tive
Expr
essio
n
p < 0.05
*
*
*
PRKCZ clones
PC N2 #1 #2 #3
PRKCZ
ITGB3
Beta-actin
144
Figure 4-18. Knockdown of IGF1R rescues transcript expression of ITGB3 in PRKCZ-expressing cells. The expression of ITGB3 is de-repressed in PRKCZ-expressing SKOV3 cells subjected to IGF1R gene knockdown by siRNA (*p < 0.05), but IGF1R knockdown has no effect on empty vector control. (n=3) N2 = empty vector control N2.
0
0.2
0.4
0.6
0.8
1
1.2
1.4
N2 PRKCZ#1 PRKCZ#4
Rela
tive
Expr
essio
n
Mockscrambled siRNA IGF1R siRNA
*
*
145
Figure 4-19. IGF1 increases ITGB3 transcript expression in PRKCZ-expressing SKOV3 cells. To examine if transcript expression of ITGB3 is directly regulated by IGF1-signalling, SKOV3 cells were treated with IGF1 (50 ng/ml) and its expression was measured by quantitative real-time RT-PCR. The addition of IGF1 had no effect on empty vector control but significantly increased the expression of ITGB3 in three of the four PRKCZ clones tested when compared to their non-treated counterpart (*p < 0.05). N2 = empty vector control N2.
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
N2 PRKCZ#1 PRKCZ#2 PRKCZ#4 PRKCZ#8
Rel
ativ
e Ex
pres
sion
No treatmentIGF-I (50 ng/ml)
*
**
146
Moreover, to examine if there is a negative feedback mechanism that regulates
expression of IGF1R, I induced SKOV3 cells with IGF1 and examined their IGF1R
mRNA expression (Figure 4-20). Upon stimulation with IGF1, the transcript expression
of IGF1R decreased in all SKOV3 cells, except for PRKCZ clone #4, as observed by
quantitative RT-PCR. This illustrates that IGF1 stimulation in SKOV3 cells may in fact
be able to repress the transcript expression of IGF1R.
4.3.4.2 TIMP-1 as a Potential Downstream Effector in ITGB3 and IGF1 Signalling
Based on my observations that both ITGB3 mRNA and protein levels are
decreased in PRKCZ-expressing cells in two different ovarian cancer cell lines (SKOV3
and OVCAR3), I sought to identify potential downstream players within this signalling
pathway that may play role in ovarian cancer. Transcription of TIMP-1 (TIMP
metallopeptidase inhibitor 1) has previously been shown to be up-regulated by ITGB3 in
the ovarian cancer cell line MDAH 2774 and thus may be a candidate target (231).
Therefore, I decided to examine the mRNA level of TIMP-1 in SKOV3 and OVCAR3
cells. Interestingly, TIMP-1 expression was decreased in PRKCZ clones of both of these
cell lines, which correlated with the expression of ITGB3 (Figure 4-21). To further
examine if the decrease in TIMP-1 expression was directly related to the decreased level
of ITGB3, SKOV3 parental cells were subjected to ITGB3 knockdown. No difference in
TIMP-1 expression was observed between these cells (Figure 4-22), suggesting that the
decreased level of TIMP-1 in PRKCZ-expressing cells was ITGB3-independent.
147
Figure 4-20. IGF1 stimulation decreases IGF1R gene expression in SKOV3 cells. When SKOV3 cells are induced with IGF1, the transcript level of IGF1R is significantly decreased (*p<0.05), except for one of the PRKCZ clones (PRKCZ#4).
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
PC EV PRKCZ#1 PRKCZ#4 PRKCZ#8
Rela
tive
Expr
essio
n
no IGF1IGF1 (50 ng/ml)
*
*
* *
148
Figure 4-21. TIMP-1 gene expression decreases in PRKCZ-expressing ovarian cancer cells. The mRNA expression of TIMP-1, a potential downstream target of ITGB3, is decreased in PRKCZ-expressing SKOV3 cells and OVCAR3 cells, as determined by quantitative real-time RT-PCR (p < 0.05). N2 = empty vector control N2. (n=3)
0.0
0.2
0.4
0.6
0.8
1.0
1.2
N2 PRKCZ#1 PRKCZ#2 PRKCZ#4 PRKCZ#8
Rel
ativ
e Ex
pres
sion p < 0.05
SKOV3
0.0
0.2
0.4
0.6
0.8
1.0
1.2
N2 PRKCZ#1 PRKCZ#2 PRKCZ#3
Rel
ativ
e Ex
pres
sion
p < 0.05
OVCAR3
149
Figure 4-22. TIMP-1 gene expression regulation is independent of ITGB3 gene expression in SKOV3 cells. The transcript levels of TIMP-1, a potential downstream target of ITGB3 was determined by quantitative real-time RT-PCR following ITGB3 siRNA knockdown in SKOV3 parental cells. a) Confirmation of ITGB3 expression knockdown. b) Transcript levels of TIMP1 in SKOV3 cells with and without ITGB3 siRNA treatment. No significant TIMP1 transcript level difference was observed between ITGB3 siRNA treated and control cells. (n=3) a)
0
0.2
0.4
0.6
0.8
1
1.2
Mock Scrambled siRNA ITGB3 siRNA
Rel
ativ
e m
RN
A E
xpre
ssio
n
ITGB3 mRNA Expression
b)
0
0.2
0.4
0.6
0.8
1
1.2
Mock siRNA Scramble siRNA ITGB3 siRNA
Rel
ativ
e Ex
pres
sion
TIMP-1 mRNA expression
0
0.2
0.4
0.6
0.8
1
1.2
Mock siRNA Scramble siRNA ITGB3 siRNA
Rel
ativ
e Ex
pres
sion
TIMP-1 mRNA expression
150
In addition to ITGB3, I examined if induction of IGF1 signalling has an effect on TIMP-1
expression. I observed that upon IGF1 stimulation, SKOV3 parental and empty-vector
controls exhibited a 2-fold decrease in TIMP-1 mRNA, but no further decrease in TIMP-1
expression was observed in PRKCZ clones (Figure 4-23). This is an interesting finding,
as IGF1 and PRKCZ signalling pathways may converge through their potential roles in
the regulation of TIMP-1 expression.
4.3.4.3 Effects of IGF and ITGB3 Signalling on Cell Migration/Invasion in SKOV3
Cells
The lack of migratory changes in PRKCZ-expressing SKOV3 cells as observed
from my migration experiments as described above may be due to lack of stimulation.
Since an increase in IGF-IR expression was observed in cells over-expressing PRKCZ, I
repeated the migration assays with the addition of IGF1. As seen in Figure 4-24, an
increase in migration was observed in parental and empty-vector controls upon
stimulation with IGF1 in the wound healing assay, illustrating that the IGF1 signalling
pathway is involved in migration of SKOV3 cells. However, this effect was not observed
in PRKCZ-expressing cells. The lack of response in PRKCZ-expressing cells may
perhaps be due to negative feedback exerted by the over-expression of IGF1 receptor in
these cells. Matrigel migration assays also did not show an increased level of invasion in
any of cells upon stimulation of IGF1 (Figure 4-25). However, upon stimulation with
EGF, both control and PRCKZ clones exhibited increased migration, and the effect of
EGF was significantly higher in two of the PRKCZ clones when compared to vector
control cells.
151
Lastly, to evaluate if ITGB3 plays a role in ovarian cancer migration, scratch
wound healing assays were repeated with SKOV3 cells treated with ITGB3 siRNA. No
difference in migration rate was observed between siRNA-treated and control cells
(Figure 4-26).
152
Figure 4-23. Effect of IGF1 on TIMP-1 transcript expression in SKOV3 cells. Treatment with IGF1 (50 ng/ml) decreased TIMP-1 transcrpt levels in SKOV3 parental and empty-vector control cells but not PRKCZ clones (* p < 0.05). (n=2) PC = parental control; N2 = empty vector control N2
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
PC N2 PRKCZ#1 PRKCZ#2 PRKCZ#4 PRKCZ#8
Rel
ativ
e Ex
pres
sion
No Treatment
IGF-I (50 ng/ml)
* *
153
Figure 4-24. Effect of IGF1 signalling on SKOV3 migration as observed by wound healing assay. No significant increase in migration was observed between controls and PRKCZ-expressing SKOV3 cells in the absence of stimulation (black bars). However, upon treatment with IGF1, parental and vector-control cells showed an increase in migration (white bars; * p<0.05). This observation was not seen in PRKCZ-expressing clones. (n=2) PC = parental control; N2 = empty vector control N2
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
PC N2 PRKCZ#1 PRKCZ#4 PRKCZ#8
Mig
ratio
n D
istan
ce (m
m)
No treatment
IGF1 (50 ng/ml) * *
154
Figure 4-25. Effects of IGF1 and EGF on migration of SKOV3 as determined by transwell migration assay. No difference in migration was observed between SKOV3 vector control and PRKCZ-expressing cells in non-treated and IGF1-treated cells. When treated with EGF, both the control and PRCKZ clones exhibited an increased in migration (*p<0.05); the effect of EGF was significantly higher in two of the PRKCZ clones when compared to vector control cells (**p<0.05). (n=2) N2 = empty vector control N2
0100020003000
4000500060007000
N2 PRKCZ#1 PRKCZ#4 PRKCZ#8
# of
cel
ls m
igra
ted
EGF (10ng/ml)IGF-1 (50 ng/ml) + 10% FBS10% FBS
*
* *
** **
*
155
Figure 4-26. Effect of ITGB3 on SKOV3 parental cells migration as observed by wound healing assay. Knockdown of ITGB3 has no effect on cell migration in SKOV3 cells. a) Confirmation of ITGB3 knockdown in SKOV3 cells. b) Distance travelled by SKOV3 cells with and without treatment of ITGB3 siRNA. (n=3) a)
ITGB3 mRNA Expression
0
0.2
0.4
0.6
0.8
1
1.2
Mock Scramble siRNA ITGB3 siRNA
Rel
ativ
e Ex
pres
sion
b)
Migration of SKOV3 Cells
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
Mock Scramble siRNA ITGB3 siRNA
Dis
tanc
e T
rave
lled
(mm
)
156
4.4 Discussion
As discussed in Chapter 2, my microarray study of familial ovarian cancer
comparing the gene expression profiles of tumours from patients with strong and weak
family history of breast and/or ovarian cancer has identified genes that are differentially
expressed. One of the most significant genes identified was PRKCZ, which was more
highly expressed in tumours from patients with strong family history, and this differential
expression was validated with quantitative real-time PCR. This observation, as well as
the previously suggested role of PRKCZ in cancer development, provided me with the
rationale for further investigations of PRKCZ.
As defined by Hanahan and Weinberg, human tumourigenesis is a multistep
process in which cells can acquire properties, through genetic alterations, that allow them
to transform into a higher malignant derivative (3). In this Chapter, I aimed to determine
if alteration in PRKCZ expression can drive such processes, by manipulating the
expression levels of PRKCZ in ovarian cancer cell lines, and examining their effect on
cell survival and migration of these cells.
My cell viability assays data revealed that over-expression of PRKCZ increases
survival of SKOV3 ovarian cancer cells while no changes were observed for HEY or
OVCAR3 cells. The variable results are likely due to the different genetic attributes
associated with each of the cell lines, which might have led to differential response to
PRKCZ-signalling. Subsequent apoptotic and proliferation experiments on SKOV3 cells
suggested that the increase in cell viability observed for PRKCZ-expressing cells was due
to an increase in proliferation and not due to suppression of apoptosis.
157
The role of PRKCZ in apoptosis remains controversial as it appears to act as both
a positive and a negative regulator (232-235). Interestingly, while I did not observe an
apoptotic effect of PRKCZ in SKOV3 cells, a recently published study by Nazarenko et
al. reported PRKCZ as a negative regulator of cell survival in ovarian cancer, as its over-
expression was able to mediate HRSL3 tumour suppressor-dependent apoptosis in
ovarian cancer cell line OVCAR3 (234). While I did not further examine the apoptotic
role of PRKCZ in OVCAR3 cells specifically (since no change in cell viability was
initially observed for this cell line), the observation made by Nazarenko et al.
nevertheless suggests that PRKCZ may be a critical player in ovarian cancer cell
apoptosis, and that its role may be dependent on additional molecular players. On the
other hand, through immunohistochemical analyses, the same group found that a
significant proportion of ovarian carcinomas express high levels of PRKCZ, as compared
to normal and benign carcinomas, and its expression was positively correlated with poor
prognosis (234). This observation appears to be contradictory to their in vitro results
described above, as one may expect a decreased expression of pro-apoptotic proteins in
advanced cancers; however, the authors also suggested that while PRKCZ expression
remains high in these tumours, the pro-apoptotic activity of PRKCZ may be reduced by
other genetic or epigenetic changes that occur in advanced cases of ovarian cancer (234).
Thus, the exact role of PRKCZ in apoptosis remains elusive and requires further
investigations.
The ability of cancer cells to migrate during cancer progression is associated with
the acquisition of abnormal motile behaviour resulting from various molecular
alterations. Aberrant expression of PRKCZ is an example of such alteration, as its role in
158
migration has previously been demonstrated in head and neck tumour cells, pancreatic
cancer, as well as breast cancer (177, 236, 237). The comparisons I made between
parental cell lines with their PRKCZ-expressing counterparts through various migration
experiments showed that over-expressing PRKCZ alone is not sufficient to exert
increased migratory properties in the three ovarian cancer cell lines tested. Nevertheless,
I showed that siRNA knockdown of PRKCZ expression in SKOV3 parental cells can
decrease its rate of migration. This observation suggests that the endogenous level of
PRKCZ is sufficient for cell motility in this cell line. Additionally, in concordance with
previous studies examining the migratory role of PRKCZ in different types of cancer cell
lines, I showed that PRKCZ can increase cell motility in SKOV3 cells to a substantial
level compared to control cells when the EGF signalling pathway is activated, thus
demonstrating that PRKCZ signalling can augment EGF-induced chemotaxis in multiple
cancer types.
In addition to PRKCZ, it is also important to note the potential roles of PRKCI
(protein kinase C iota) in ovarian cancer. Similar to PRKCZ, PRKCI belongs to the
atypical protein kinase C family group, and it has been implicated in the establishment of
cell polarity, motility, proliferation, and survival of cancer cells (238-241). Interestingly,
the PRKCI gene has been shown to be amplified and over-expressed in serous epithelial
ovarian cancers, and an increase in its DNA copy number is associated with a decrease in
progression-free survival for the disease (239, 242). Since PRKCZ and PRKCI are
highly homologous to one another, sharing ~70% overall amino acid sequence identity
(243), it is possible that these two proteins function redundantly. Indeed, it has
previously been demonstrated that disruption of either PRKCZ or PRKCI expression can
159
inhibit tight junction formation in cultured epithelial cells, suggesting that these two
proteins have an overlapping role in establishment of cell polarity (244). For that reason,
it may be important to further investigate if PRKCI can also contribute to the phenotype
that I have observed in PRKCZ-expressing cells. Additionally, since siRNA knockdown
of PRKCZ did not have an effect in HEY and OVCAR3 cell lines, it may be of interest to
determine if these cell lines are more dependent on the activity of PRKCI.
The original aim of my study was to identify the genetic alterations that occur in
the subset of patients with strong family history; therefore, I was interested in finding the
connection between PRKCZ and familial ovarian cancer. While BRCA1 and BRCA2 are
two well-studied high-penetrance genes associated with ovarian cancer, I was especially
interested in identifying expression alterations of other genes potentially related to
familial ovarian cancer which may be under the regulatory control of PRKCZ. By
examining protein and genetic interaction networks generated using my ovarian cancer
microarray expression data described in Chapter 1, I identified IGF1R (Insulin-like
Growth Factor 1 Receptor) and ITGB3 (Integrin beta 3) as potential interactors or targets
of PRKCZ.
The IGF1R promoter has previously been identified as a molecular target for
BRCA1 in breast cancer cell lines (245, 246). This observation suggests that BRCA1
mutations may result in transcriptional de-repression of the IGF1R promoter, thus
increasing the level of IGF1R, and may be one of the mechanisms responsible for breast
and ovarian tumourigenesis in BRCA1 mutation carriers. Indeed, a more recent study
conducted by the same group revealed that primary breast tumour samples from BRCA1
mutation carriers have elevated levels of IGF1R protein compared to non-carriers (247),
160
further confirming the association between BRCA1 and IGF1R. Moreover, a large body
of evidence has supported the importance of IGF1R expression in ovarian cancer, and
that increased expression of IGF1R is associated with aggressiveness, as well as drug
resistance of the disease (248-253).
My studies gave evidence that over-expressing PRKCZ in certain ovarian cancer
cell lines can in fact alter the expression of IGF1R, and the type of expression alteration
is cell-line dependent. Specifically, while no IGF1R transcript levels alteration was
observed in SKOV3 cells, its protein expression level is increased in cells that over-
express PRKCZ, which may be explained by post-transcriptional processes such as
protein translation, post-translation modification and decrease in protein degradation;
however, the exact mechanism involved remains to be investigated. On the other hand,
both gene and protein expression levels of IGF1R are decreased in the OVCAR3 cell line
when PRKCZ is over-expressed, suggesting that regulation of IGF1R by PRKCZ in this
particular cell line may be occurring at the transcript level. These results imply that
regulation of IGF1R by PRKCZ may also be occurring within different biological
pathways and may be dependent on other molecular characteristics specific to each of the
cell lines, once again illustrating the heterogeneity of this disease, and further
investigations are required to determine the exact mechanisms responsible for the dual
effect PRKCZ has on IGF1R expression.
In addition to IGF1R, ITGB3 has also previously been suggested to be associated
with familial ovarian cancer. In a population study conducted by Jakubowska et al., it
was found that the Leu33Pro polymorphic allele of ITGB3 is associated with an increased
risk of ovarian cancer in individuals with BRCA1 mutations (123). While subsequent
161
analysis by the same group revealed that this association may only be specific to the
Polish population from their original study (254), it nevertheless suggests the importance
of ITGB3 in ovarian cancer in a specific cohort of BRCA1 carriers. The exact molecular
mechanisms involved in the increased ovarian cancer risk in ITGB3 Leu33Pro carriers in
this Polish population is still unknown; however, it was suggested this particular allele
may be able to enhance the adhesive properties of tumour cells as well as activation of
MAPK pathway (255, 256), which may contribute to the malignant potential of cancer
cells.
The role of ITGB3 in ovarian cancer has also been implicated in a study in which
over-expression of ITGB3 in SKOV3ip1 cells (cell line generated from ascites developed
in nu/nu mouse by administering an intraperitoneal injection of SKOV3 cells) was found
to be associated with decreased invasion, protease expression, as well as colony
formation; these observations were consistent with their subsequent in vivo experiments,
which showed that tumours expressing ITGB3 were less aggressive compared to those
that do not express this protein (257). Moreover, upon examination of ITGB3 expression
in ovarian tissue of patients with invasive ovarian cancer, the same group found that
patients with high ITGB3 expression had a significantly better prognosis, which is
consistent with two other recent studies showing that an increased level of ITGB3 is
associated with better survival in patients with stage III serous ovarian cancer, and was
suggested to be a potential prognostic marker (183, 257, 258). Interestingly, my
assessment of ITGB3 expression in PRKCZ-expressing SKOV3 and OVCAR3 ovarian
cancer cells showed that ITGB3 is down-regulated in the presence of PRKCZ, as its gene
and protein expressions were both decreased compared to controls. Additional IHC
162
studies should be performed to examine whether this correlation occurs in ovarian tumour
specimens, and whether there is a correlation between family history and ITGB3
expression.
The concomitant expression alterations of IGF1R and ITGB3 in PRKCZ-
expressing cells led me to question whether these genes are activated in the same
signalling pathway. Indeed, my IGF1R siRNA knockdown experiments using SKOV3
cells revealed that ITGB3 transcription may be dependent on expression of IGF1R;
however, the effects of IGF1R on OVCAR3 cells may differ since its expression is
decreased in PRKCZ-expressing cells. Nevertheless, results from my studies suggest one
possible mechanism by which ITGB3 expression may be altered, and as a consequence, a
more aggressive phenotype of ovarian cancer cells is developed.
Given that PRKCZ expression correlates with the expression of both IGF1R and
ITGB3, I further examined if alteration of these genes can affect the migration phenotype
of ovarian cancer cells that over-express PRKCZ. As mentioned earlier, over-expression
of PRKCZ alone did not alter the migration rate of any of the ovarian cancer cell lines but
did show significant increased migration in SKOV3 cells upon activation of EGF
signalling. Contrary to what I expected, scratch wound healing migration assays showed
that upon IGF1 stimulation, SKOV3 control cells but not PRKCZ-expressing cells,
displayed an increase in cellular motility. The lack of response in PRKCZ-expressing
cells may perhaps be due to a negative feedback mechanism exerted by the over-
expression of IGF1 receptor in these cells, thus hindering the cells’ ability to respond to
IGF1 signalling. Additionally, unlike results obtained from scratch wound assays, IGF1
stimulation had no effects on any of these cells in matrigel migration assays, suggesting
163
that while IGF1 signalling may be involved in cell migration, it may be insufficient for
increasing the invasive properties of these cells, because the cells are incapable of
breaking down the matrigel matrix to cross the barrier.
Based on the findings from this Chapter, I propose the following model by which
PRKCZ may participate during tumour progression in a subset of ovarian cancer (Figure
4-27). In cells with normal expression of PRKCZ, the expressions of IGF1R and ITGB3
are in equilibrium. When PRKCZ is deregulated and over-expressed, it increases
translation or stability of IGF1R, thus enhancing IGF1 signalling, leading to a repression
of ITGB3 expression, which ultimately can lead to changes in cellular processes that can
enhance the aggressiveness of a tumour cell (eg. impaired apoptosis, increased
proliferation). When IGF1R gene expression is decreased (eg. via siRNA knockdown) in
these PRKCZ-expressing cells, de-repression of ITGB3 occurs. Additionally, when cells
are stimulated with IGF1, the overall expression of IGF1R decreases due to a negative
feedback mechanism that leads to suppression of IGF1R transcription and de-repression
of ITGB3; however, there may be another yet to be identified pathway downstream of
IGF1 signalling that can lead to de-repression of ITGB3. One possible pathway may be
PI3K/AKT, as IGF1 is a potent activator of this signalling cascade (259). Further
investigations are worthwhile to evaluate and confirm the roles of these important players
in ovarian cancer development; a better understanding of the interaction of the molecules
involved may be useful in development of therapeutics for the subset of ovarian cancer
patients who display expression alterations of these genes.
164
Figure 4-27. Proposed model of ITGB3 transcriptional regulation through IGF-1 signalling in PRKCZ-expressing SKOV3 cells. a) Expression of IGF1R and ITGB3 are in equilibrium in the presence of normal level of PRKCZ. b) Over-expression of PRKCZ increases either translation or stability of IGF1R, possibly leading to constitutive activation of IGF1 signalling cascade that results in transcriptional repression of ITGB3 and increase cell survival through proliferation. c) Repression of ITGB3 in PRKCZ-expressing cells is dependent on IGF1R expression as IGF1R knockdown by siRNA de-represses its expression. a) b)
c)
165
Figure 4-27 (Cont’d). d) IGF1 stimulation activates negative feedback mechanism, leading to a decrease in IGF1R transcription; ITGB3 expression is de-repressed. e) In addition to IGF1R transcription suppressor, other signalling pathway downstream of IGF1 cascade, such as PI3K/AKT, may be activated to de-repress expression of ITGB3. d) e)
CHAPTER 5
Conclusions and Future Directions
166
167
5.1 Summary and Implications of Thesis Findings
Epithelial ovarian cancer is one of the most lethal gynaecological malignancies.
One of the most important risk factors for this disease is family history; therefore,
identification of the molecular changes involved in familial ovarian cancer development
is fundamental in developing preventive and diagnostic measures for the group of
patients with family history of breast and/or ovarian cancers.
My thesis aimed to examine the molecular changes involved in familial ovarian
cancer. To achieve this, I utilized cDNA microarrays to identify the gene expression
profiles of ovarian tumours from patients with different family history. Results from this
analysis prompted me to further examine hCDC4 and PRKCZ in relation to ovarian
cancer. I carried out various genetic analyses to explore the mechanisms that may alter
hCDC4 expression in ovarian tumours. Additionally, to characterize the potential roles
of PRKCZ in ovarian cancer, functional and biochemical analyses were performed.
5.1.1 Expression Profiling of Familial Ovarian Cancer
Ovarian cancer is a heterogeneous disease that is attributable to a diverse group of
molecular alterations. Since the development of high-throughput microarrays,
identification of these alterations has escalated greatly; however, information regarding
the genetic changes that occur in familial cases of ovarian cancer remains minimal. To
address this, I performed gene expression profiling of ovarian tumours from patients with
family history of breast and/or ovarian cancer, as discussed in Chapter 2 of this thesis.
Statistical analysis of microarray results revealed the expression alterations that
occur in familial ovarian cancer, which included genes known to be linked to apoptosis,
168
cell migration, cell adhesion, and cell cycle regulation. Differentially expressed genes
such as VTN and PDCD4 have previously been associated with ovarian cancer; however,
a majority of the genes identified have yet to be examined in relation to this disease, such
as MCM5, CYP11A1, as well as PRKCZ, which I have chosen for additional functional
analyses as described further below.
The development and progression of ovarian cancer is likely to be due to not one,
but multiple aberrant genetic events that ultimately lead to changes in important cellular
pathways. For that reason, I integrated my gene expression data onto various protein
interaction and biological pathway databases through Ingenuity Pathway analysis, in an
attempt to reveal and explore the altered biological networks/pathways that may play
roles in familial ovarian cancer. Significantly altered interaction networks were
identified, including those that revolve around MAPK, HNF4A, histone 3, HGF, as well
as beta-estradiol. While some of these molecules and pathways have previously been
implicated in ovarian cancer, their associations with the development of familial cases of
ovarian cancer require further investigations.
As mentioned, multiple genetic events are likely to be responsible for
manifestation of ovarian cancer; as such, high-throughput identification and subsequent
detailed examination of concomitant genetic alterations occurring in familial ovarian
cancer may reveal relevant biological information that can be used in the development of
early detection methods and clinical treatments of the disease in the subset of patients
with ovarian cancer predispositions.
169
5.1.2 hCDC4 and Ovarian Cancer
My microarray study of ovarian cancer showed that the tumour suppressor gene
hCDC4 was expressed at lower levels in ovarian tumours from patients with strong
family history of breast and/or ovarian cancer compared to tumours from patients with
weak family history. Consequently I attempted to elucidate the mechanisms by which
ovarian cancer cells can deactivate the expression of hCDC4, as described in Chapter 3.
My mutation screening experiments showed that although hCDC4 mutations may
be important in malignancies such as colorectal and endometrial cancer (202, 203), it is
an infrequent event in ovarian cancer, as only one nucleotide change was observed from
my analysis of 28 ovarian tumours. This observation is in concordance with other
hCDC4 mutational studies conducted recently (207, 209). Our findings, however, could
not rule out other mechanisms by which hCDC4 expression can be repressed in ovarian
cancer. Epigenetics such as promoter methylation can play significant roles in gene
expression regulation. Indeed, numerous studies have reported the presence of
hypermethylation within promoters of tumour suppressor genes, including BRCA1 in
sporadic ovarian cancer (260). However, my analysis ruled out promoter methylation as
a method by which ovarian cancer cells can inactivate hCDC4, as all of the ovarian
tumours tested showed absence of methylation within the hCDC4 promoter. Loss of
heterozygosity may be another mechanism by which hCDC4 gene expression can be
reduced. I examined the LOH status of hCDC4 using DNA from four ovarian tumours
with their normal-matched tissues but no LOH was observed. However, due the small
number of normal samples available for this part of study, the results from this
170
experiment remain inconclusive. Thus a larger collection of normal-tumour matched
samples will be required for robust analysis.
The protein expression of hCDC4 varies among ovarian tumours, as observed by
immunohistochemical staining of ovarian TMA. However, the significance of decreased
hCDC4 protein expression in familial cases of ovarian cancer was indeterminate, due to
the limited number of familial tumours present on the array. Therefore, further analyses
using a larger sample size may provide a better insight on the significance of hCDC4
expression in familial ovarian cancer.
5.1.3 PRKCZ and Ovarian Cancer
The expression of PRKCZ is more highly expressed in ovarian tumours from
patients with strong history of breast and/or ovarian cancer, as observed in familial
ovarian cancer gene expression profiling described in Chapter 2. While the roles of
PRKCZ have previously been discussed in various malignancies (174, 176-178, 223,
261), its relation to ovarian cancer is unclear.
In Chapter 4 of this thesis, I described the in vitro approaches I have taken to
examine the potential roles of PRKCZ in ovarian cancer. I showed that the SKOV3
ovarian cancer cell line stably expressing PRKCZ exhibited a higher rate of cell growth
compared to controls. This effect, however, was cell line specific, since no changes were
observed in HEY and OVCAR3 ovarian cancer cell lines. In addition to cell
proliferation, I showed that endogenous PRKCZ plays a role in ovarian cancer cell
migration, as endogenous PRKCZ expression knockdown by siRNA decreases the
migration rate of SKOV3 cells compared to controls. Moreover, while the HEY cell line
171
stably expressing PRKCZ did not show an increased rate of cell migration in scratch
wound healing assays, it did exhibit random patterns of cell movement compared to
controls. Subsequent measurement of single cell movement by phagokinetic track
assays, however, showed that individual HEY cells over-expressing PRKCZ did not have
increased cell motility relative to non-transfected cells. These observations suggest that
PRKCZ-induced random cell motility in HEY cells likely requires the presence of
signalling cues from other cells over-expressing PRKCZ.
The ultimate aim for my thesis project is to identify molecular alterations
involved in the development of ovarian cancer in patients with predisposition. Hence, I
attempted to examine PRKCZ in relation to other proteins or pathways that have
previously been implicated in familial ovarian cancer. Upon examination of the
interaction network involving PRKCZ as generated by IPA using expression profiling
data, I decided to focus on the expression of IGF1R and ITGB3 in relation to PRKCZ,
since both of these genes have previously been suggested to be associated with BRCA1-
related breast and ovarian cancers (123, 245-247, 262).
Altered expression of IGF1R in various types of tumours have previously been
shown to occur after other molecular events, such as repression of tumour suppressor
genes, or gain of function mutations in p53 (263). My biochemical analyses revealed that
an increase in PRKCZ expression may be one of the molecular events leading to
alteration in IGF1R protein expression in ovarian cancer. Intriguingly, while over-
expression of PRKCZ increases protein expression of IGF1R in the SKOV3 ovarian
cancer cell line, it had the opposite effect in the OVCAR3 ovarian cancer cell line, as its
over-expression was correlated with decreased mRNA and protein expression of IGF1R.
172
Strikingly, the expression of ITGB3 was decreased at both mRNA and protein levels in
PRKCZ-expressing SKOV3 and OVCAR3 cells, compared to controls. Additionally, the
mRNA level of TIMP-1, a target of ITGB3, was also decreased in cells that over-express
PRKCZ.
From these studies it is apparent that over-expression of PRKCZ can affect
downstream pathways in some ovarian cancer cells, which may lead to phenotypic
changes associated with increased aggressiveness of the disease. Further elucidation of
these pathways may provide insight on the biological events that occur in the
development of ovarian cancer in predisposed patients.
173
5.2 Future Directions
The work presented in this thesis provided insight into some of the genetic and
molecular events that occur in familial ovarian cancer. Based on data presented in the
preceding chapters, additional studies can be proposed to further understand the biological
alterations responsible for ovarian cancer development in predisposed patients.
5.2.1 High-Throughput Analyses of Familial Ovarian Cancer
As mentioned earlier, the development of high-throughput techniques have guided
researchers in generating a plethora of scientific data. My initial expression microarray
analyses of an enriched set of familial ovarian cancer have also produced a collection of data
that can be further analyzed, thus providing additional insights on the biology of this disease.
5.2.1.1 Gene Set Analysis of Expression Microarrays Data
To complement our single-gene microarray analysis that was described in Chapter 2,
Gene Set Analysis (GSA) was performed in collaboration with biostatistician Dr.
Pinnaduwage. GSA differs from the aforementioned identification of differentially expressed
genes as it evaluates the differences in biologically relevant functional gene units rather than
as single genes between the two subject groups (264). Since this method uses the entire
collection of gene expression microarray data without pre-filtering for a short list of strongly
differentiated genes, it allows us to detect subtle coordinated gene expression changes in
specific biological pathways that may be important in familial ovarian cancer development.
In our preliminary GSA analysis, we examined a total of 4984 gene sets from various
curated databases, and identified six gene sets that are significantly differentially expressed
between the strong and weak familial ovarian cancer groups (Table 5-1). Furthermore, IPA
174
was applied using these results to identify potential biological pathways and processes in
which these genes may be involved (Table 5-2). Intriguingly, five of the six significant gene
sets contain interaction networks that center on HNF4A, a nuclear transcription factor, but
each with different biological interactors (Figure 5-1). This is a remarkable finding since
HNF4A was also the main node in one of the protein interaction networks that I have
previously identified through IPA using our significant gene list from the single-gene
approach (Figure 2-4-B). The expression of HNF4A was not measured in my microarray
analysis because this gene was not present on the microarray platform that was used;
nevertheless, the observation made from the IPA analysis strongly suggests the diverse
pathways that HNF4A may be involved in during the development and/or progression of
familial ovarian cancer, through the deregulation of its target genes. To date, evidence for a
role for HNF4A in ovarian cancer is lacking; therefore, it may be worthwhile to further
examine the expression levels of HNF4A, as well as its gene targets in the subset of ovarian
cancer patients with strong family history in order to examine its relevance in the disease.
Additional functional studies involving manipulation of HNF4A expression may be useful in
identifying molecular alterations important in ovarian cancer predisposition.
175
Table 5-1. Gene Set Analysis (GSA) of familial ovarian cancer expression microarray data. Gene sets identified as significantly altered based on gene expression data of strong and weak familial ovarian tumours. * = lower expression of most genes in the gene set correlates with strong familial cases; ** = higher expression of most genes in the gene set correlates with strong familial cases; ES = enrichment score. FDR = false discovery rate
Gene Sets Source ES p-value FDR Top Cellular Processes as determined by IPA
MORF_PRKDC*
(neighbourhood of PRKDC, protein kinase, DNA-activated, catalytic polypetide)
MSigDb v2.0; C4
-0.2755
<0.001
<0.001
cellular assembly & organization; cell cycle; cell morphology; DNA replication, replication & repair; cell-to-cell signaling and interaction, cellular growth and proliferation; cancer
GCM_PTPRU* (neighbourhood of PTPRU, protein tyrosine phosphatase, receptor type U)
MSigDb v2.0; C4
-0.8247
<0.001
<0.001
metabolic disease; cellular assembly & organization; lipid metabolism; cellular growth and proliferation; gene expression; DNA replication, recombination & repair
MORF_CCNF* (neighbourhood of cyclin F)
MSigDb v2.0; C4
-0.4705
0.002 0.2135
drug metabolism; small molecule biochemistry; cell death, cellular function & maintenance; cancer
GCM_MAX* (neighbourhood of MYC associated factor X)
MSigDb v2.0; C4
-0.8078
0.002 0.2135
gene expression; infection mechanism; cancer, cellular assembly and organization
Module412*
Segal Laboratory Cancer Modules
-1.2709
<0.001
<0.001
lipid metabolism; small molecule biochemistry; cellular compromise
FATTY_ACID_ DEGRADATION**
MSigDb v2.0; C2
0.8565
<0.001
<0.001
lipid metabolism; genetic disorder; molecule biochemistry, molecular transport
176
Table 5-2. Top functions of networks belonging to significant gene sets as identified by Gene Set Analysis (GSA) and Ingenuity Pathway Analysis (IPA). Unique genes present from both microarray dataset and gene sets are shown in bold (focus molecules). A score of >3 was considered statistically significant (p<0.001).
Gene Set Molecules in Network Score # of Focus Molecules Top Functions
26s Proteasome, BRD8, BUB3, BUB1B, Caspase, CDC7, CDKN2C, Cyclin A, DIAPH1, E2f, HCFC1, HDAC2, Histone h3, HNRNPD, Hsp70, LBR, MCM2, MCM5, MCM6, NFkB (complex), NUP62, POLR2A, PP1-C, PRKDC, PSMA1, RPA2, SAFB, SSRP1, TFDP1, TRAP/Media, TRRAP, TYMS, XPO1, XPO6, XRCC5
54 26
Cellular Assembly and Organization, Cell Cycle, Cell Morphology
ALDH6A1, APRT, ARPC1A, ATP5H (includes EG:641434), ATP5I, ATP5O, ATP6AP1, ATP6V1D, ATP6V1F, CS, DLD, ESD, FH, GRB2, GSK3B, IDH3A, KCNMA1, MLEC, NAGA, PDHB, PGRMC2, PRPS2, PRPSAP2, PRUNE, RNPEP, RPS21, RPS12 (includes EG:6206), SEPT7, SHMT2, SLC2A4, SSBP1, SUCLG2, UPF3A, VARS, VDAC2
28 16
Energy Production, Nucleic Acid Metabolism, Small Molecule Biochemistry
Actin, Alpha-tubulin, ARHGEF18, CAP2, CCT4, CCT5, Ck2, DKC1, DNAJC11, DOCK5, FBXO45, GNB1, GPAA1, GTF2A2, IMMT, Insulin, KHDRBS1, KIF20B, OSBP, PCSK9, PDCL, PFDN4, PFDN6, PIN1, PMVK, PPP1R1A, RNA polymerase II, RUVBL1, SREBF2, SRRM1, SSB, STON1-GTF2A1L, TFIIF, VBP1, VTI1A
24 14
Carbohydrate Metabolism, Small Molecule Biochemistry, Post-Translational Modification
14-3-3(β, γ, θ, η, ζ), APOL3 (includes EG:80833), ARIH2, C22ORF9, Calmodulin-Camkk-Ca2+, Camkk, CAMKK1, CAMKK2, CDKN1A, COPS5, DDB1, DEAF1, DNMT1, EI24, EML3 (includes EG:256364), KAT5, KLC4, KRAS, LARP1, MAPK9, NR3C1, PCNA, PDIA3, SAMD4B, SHPRH, SKIV2L2, SLC1A6, USP1, USP14, USP37 (includes EG:57695), VANGL2, XPO7, YWHAD, YWHA, YWHAQ (includes EG:10971)
19 12
DNA Replication, Recombination, and Repair, Cell Cycle, Connective Tissue Development and Function
MORF_PRKDC
ANKRD17, APON, BUD31 (includes EG:8896), Ca2+, CALML4, CEACAM21, CXCL10, DISP2, DULLARD, ELF3, EPRS, GARS, GGT6, GRHL1, HNF4A, IL4, MCFD2, MRPS27, MS4A8B, NFKBIE, ODZ4 (includes EG:26011), RFC4, SLC22A18, SLC26A11, SLC35A1, SLC35A5, SLC39A1, TARS, TGFB1, THOP1, TMEM17, TMIGD1, TTC22, TXLNA (includes EG:200081), YARS
17 11
Cell-To-Cell Signaling and Interaction, Cellular Growth and Proliferation, Hematological System Development and Function
177
1, 3, 4, 5-IP4, AKT1, ALG3, BPGM, C19ORF50, CCDC33, COL8A1, CYB5R2, DCTD, DDX46, DEGS1, dihydrosphingosine 1-phosphate, Foxo, FXR2, GFRA1, LNX1, MTCP1, NTRK2, Ntrk dimer, NUDT1, PABPC4, PCBD1, PLAC8, PTN, RBMX,RPIA, SHC1, SHC2, SHC3, SLC25A1, SNRPA, TRIP13, VCL, WDYHV1, ZBTB8A
15 10
Cancer, Reproductive System Disease, Cell Death
ANP32A, APP, ATP5G3, CALU, CD82,CDK5R2, ELP3 (includes EG:55140), FKBP5, FZD4, Histone h4, HOXB7, HOXC8, IKBKAP, MIR363 (includes EG:574031), MLL, MT1F, MTA2, NAE1, NFKBIA, nitrite, Notch, NUMBLPELP1, POLD2, POM121, PPIB, PPIE, Rab5, SET, SMARCA5, THAP7, TNFSF15, TP53BP2, ZBTB7A, ZNF131
6 5
Cellular Function and Maintenance, Gene Expression, Cell Morphology
ADCYAP1, Akt, APOC3, AVPR1B, C13ORF15, CDH5, Creb, CREB-NFkB, CSPG4, ELK4, ERK, ERK1/2, FABP3, FCGR2A, GCK, GHRHR, hCG, HMGA2, IFN Beta, IFNA21, IL1, Insulin, ITSN1, MVK, NDUFA1, NFkB (complex), OVGP1, PI3K, PRKCA, Ras homolog, SEC14L2, SSTR3, STAT4, TAF1, TEC
45 20
Metabolic Disease, Cellular Assembly an Organization, Lipid Metabolism
AHR, AUH, beta-estradiol, CCR9, CD40LG, CD8B, CHRM5, CUZD1, CYP1, DARS2, FEV, FMO2, FMO3, GH2, GZMC, HMGA1, HTR6, HTR5A, HTR5B, ICAM1, ICAM5, IL4, IL15, IRS1, KRT12, KRT83, NUDT1, P2RY6, PTPRU, SLC17A2, SLC30A3, SPRR2B, SPRR2G (includes EG:6706), SYT11, TNFSF8 (includes EG:944)
18 10
Cellular Growth and Proliferation, Hematological System Development and Function, Cellular Development
15-(S)-hydroperoxyeicosatetraenoic acid, AANAT, BNIP1, C11ORF82, CHRNE, COL14A1, CTSF, CYP26B1, KRT35, LRRC8C, MGST2, MMP13, norepinephrine, NQO2, PC, PDRG1, PEPD, PEX11A, PMM1, PPARG, PRKRIR, PRODH, PTGS2, RNASE4, SCO2 (includes EG:9997), SLC14A2, SMAD3, SMARCD1, SUPT4H1, testosterone, TNF, TNN, TP53, VPS72, ZNF8
18 10
Gene Expression, Cardiovascular System Development and Function, Organismal Development
GCM_PTPRU
acetic acid, AGTRAP, BLVRA, CSTA, EPGN, ERCC1, ERCC4, FHL3, FIGF, FOS, GCNT1, GHRHR, GPER, GRIN2B, HMMR, KLB, LGI1, LIN7A, MAPK1, MOS, MPP2, MSK1/2, MYH7, NMB, PTPN5, PTPN7, RLN2, RPS6KA4, spermidine, SRF-ELK1, ST8SIA1, TMSB4, TPSD1, TREM1, UCN2
8 5
DNA Replication, Recombination, and Repair, Cell Signaling, Nucleic Acid Metabolism
178
5430435G22RIK, CA3, CDC7, CDK1/2, CDK2-CyclinE, CDKN1A, Cyclin A, Cyclin E, dihydrotestosterone, EGFBP2, ERK1/2, FRYL, HELB, Jun-GABP, KIAA0101, leucovorin, MAD2L1, MAK, MAPK8, ME1, MFN2, MGC29506, NFYB, PAXIP1, PEBP4, PLK4, POLA1, progesterone, SIGMAR1, SYNJ2, TFF2, TRIM27, TRIP13, WHSC2, ZNF346
27 13 Drug Metabolism, Small Molecule Biochemistry, Cell Death
ABCC6, ADH6, ADH1B, APLP1, ARL4C, ATG4C, C11ORF1, CEBPA, CHMP1B, CLPX, DHX8, ECE2, EXTL2, GIPC2, HNF4A, LARS2, LPGAT1, MMRN1, MPHOSPH6, MRPL44, PEPD, PEX3, PEX16, PHF10, RBL2, RFC5, RQCD1, SETDB1, SHFM1, SPAST, SPCS3, TMEM11, TPP2, WRNIP1, ZFP64
25 12 Small Molecule Biochemistry, Cellular Assembly and Organization, Cellular Development
MORF_CCNF
AKR1A1, ATP6V1B2, CACNG5, CAMK1, CAMK2G, CAMK2N1, CES1 (includes EG:1066), CETN3, CYCS (includes EG:13063), CYP2D12, DLG4, ERK, HOXD12, INPP5A, IQSEC2, JRK, LAP3, LOC388344, LPHN1, MIR214 (includes EG:406996), MUC5B, NAGPA, PCP4, PLEKHB1, retinoic acid, RPS12 (includes EG:20042), SMG1, SP1, SPRED2, STK10, SUCLA2, SYN1, TAF2, UBE2V2, UGT1A9 (includes EG:54600)
17 9 Cellular Function and Maintenance, Lipid Metabolism, Molecular Transport
ABT1, BANF1, C11ORF82, C14ORF106, CCAR1, CCDC21, CDK9, Cyclin T, CYP1A1, DCUN1D1, HNF4A, LARP7 (includes EG:51574), LSM14A, MIR24-1 (includes EG:407012), PRPF40A, RNF138, RTP3, SH3BGRL2, SMAD2, SMAD4, SMG1, STAG2, TCERG1, TM9SF2, TMEM30A, TNF, TP53, TPRKB, TRIM15, USPL1, ZBTB11, ZCCHC8, ZNF175, ZNF318 (includes EG:24149), ZNHIT6
32 13 Gene Expression, Infection Mechanism, Embryonic Development
GCM_MAX
AMD1, BCL2A1, CAPZB, CBX3, CCNA2, CDH2, CLIC4, DHX15, DSTYK, EEF1G, KPNA4, KPNB1, MAT2A, MIR199A1, MIR199A2, MT1F, MTPN, MYC, MYCBP2, MYH7, MYO1C, NCAM1, NFYC, NPPB, RAB10, RBM25 (includes EG:58517), RBMS1, ROCK1, S100A6, SDC1, SERBP1, SPARC, TSC2, XRN1, ZNF217
5 3 Cancer, Renal and Urological Disease, Cellular Assembly and Organization
179
Module412 ABCA6, ABHD6, AEBP1, AGT, AQP8, ASAH1, ATP6V1C1, ATP6V1D, ATP6V1H, cholesterol, COX7A2, COX7C (includes EG:1350), CPA1, CPB2, CTSA, DPAGT1, ETNK2, FUCA1, GLA, GLB1, HEXA, HNF4A, hydrogen peroxide, IER5, MIPEP, NAGA, NFkB (complex), NLN, RNF19B, SC5DL, SGK2, SPCS3, TMEM123, WNT10A, ZNF71
25 9
Lipid Metabolism, Small Molecule Biochemistry, Cellular Compromise
ACAA2, ACAA1B, ACSL3, ACSL5, ACSL6, CES2 (includes EG:8824), Cpt, CPT1, CPT1A, CPT1B, CPT1C, CRAT, DDX10, GPD2, HADHB, HNF4A, IGF1, LIF, LIPT1, LPL, LRRC8C, oleic acid, PNLIPRP1, PPAR ligand-PPAR-Retinoic acid-RXRα, PPAR ligand-PPARγ-Retinoic acid-RXR, PPARG, PPARGC1A, PPARγ ligand-PPARγ-Retinoic acid-RARα, retinoic acid, SLC25A20, SLC5A3, TMEM49, TMEM176A, TPI1, ZNF133
29 11
Lipid Metabolism, Small Molecule Biochemistry, Molecular Transport FATTY_ACID_
DEGRADATION
9330129D05RIK, acad, ACAD8, ACAD9, ACAD10, ACAD11, ACADL, ACADM, ACADS, ACADSB, ACADVL, acyl-CoA dehydrogenase, GCDH, IVD, IWS1, PPARA, TNF
7 3
Genetic Disorder, Metabolic Disease, Lipid Metabolism
180
Figure 5-1. Potential significance of HNF4A in familial ovarian cancer as suggested by Gene Sets Analysis (GSA) of gene expression microarray data. Of the six gene sets identified as significantly altered from GSA analysis of familial ovarian cancer gene expression microarrays, five contain an interaction network that centralizes on HNF4A (hepatocyte nuclear factor 4, alpha), as identified by IPA. HNF4A interaction networks from A) MORF_PRKDC gene set, B) MORF_CCNF gene set, C) GCM_MAX gene set, D) Cancer Module 412 gene set, and E) FATTY_ACID_DEGRADATION gene set. A) B)
181
C) D) E)
182
5.2.1.2 Genomic Signature of Familial Ovarian Cancer As detailed in Chapter 1, DNA copy number alterations can have effects on gene
expressions in a cell. As such, in parallel to the cDNA expression microarray experiments, I
have initially performed array-comparative genomic hybridization (aCGH) experiments using
genomic DNA from the same cohort of tumour samples in order to identify DNA copy
number changes that occur in familial ovarian cancer. In collaborations with biostatisticians
Dr. Shelley Bull and Sarah Colby, chromosomal regions of alterations were identified in our
preliminary analysis, which included gains in 3q26.3 (where well-known ovarian cancer
oncogene PIK3CA resides), 7p14, 10q21, 15q23, 16q24, 17q, and 20q, and one deletion at
22q. These gains and loss were identified in both strong and weak familial ovarian tumours.
While I subsequently decided to focus on examining the gene expression data, the results
generated from my aCGH study can nonetheless be used to further identify genetic
alterations that occur in ovarian cancer. Regions of gains and losses should be validated by
quantitative real-time RT-PCR and fluorescence in situ hybridization (FISH), respectively,
and their correlations with mRNA and protein levels of candidate genes should be analyzed.
Appropriate functional experiments can then be designed and performed to interrogate their
roles in ovarian cancer.
5.2.2 hCDC4 and Ovarian Cancer
The mutational analysis that I performed in Chapter 3 examined the coding regions of
hCDC4 in ovarian cancer. The presence of hCDC4 variations/polymorphisms in non-coding
regions, however, has yet to be examined in ovarian cancer. Indisputably, single-nucleotide
polymorphism within intronic regions of a gene can play significant roles in cancer
susceptibility, as suggested in previous epidemiological studies (265-271). In fact, the
183
presence of specific SNP alleles within intronic regions of hCDC4 was recently reported in
acute myeloid leukemia (AML) and breast cancer (272, 273). The SNP allele described in
the AML study did not show significance in genotype frequency between patients and
healthy controls (272). On the other hand, Yu et al. observed an association between specific
hCDC4 intronic SNP alleles and breast cancer susceptibility; additionally, these alleles
showed a joint effect with SNP alleles found for the cyclin E gene, thus further elucidating
the importance of cell cycle and ubiquitin ligase genes interactions in cancer development
(273). On the basis of these observations, it would be of interest to identify the presence of
intronic SNP allele variations of hCDC4 in ovarian cancer, and to explore the correlation
between different alleles with disease susceptibility.
The notion of hCDC4 as a haploinsufficient tumour suppressor gene, as previously
suggested in a study using a hCDC4+/- ; p53+/- mouse model that can develop a wide range
of malignancies, including ovarian epithelial tumours (213), also prompts the question of
whether individuals with p53 mutations may be more susceptible to ovarian cancer if they
also have one dysfunctional allele of hCDC4, either through gene mutation or LOH.
Therefore, future genetic and clinical studies examining concurrent mutational events of p53
and hCDC4 in ovarian tumours may provide some insight on some of the molecular events
responsible for ovarian tumourigenesis, and information gained from such studies may be
used for development of novel prognostic tools in clinical settings.
5.2.3 PRKCZ and Ovarian Cancer
It is also important to further explore the functions of PRKCZ in familial ovarian
cancer in relation to BRCA proteins. The absence of a well-characterized human BRCA-null
ovarian cancer cell line prevented me from examining this aspect in detail in my in vitro
184
functional studies. However, a novel BRCA1-null ovarian cancer cell line, UWB1.289, and
its BRCA1-expressing counterpart have been propagated and generated by DelloRusso et al.,
which were made publicly available (274). By over-expressing PRKCZ in these cell lines, it
would now be possible to address the question of whether or not PRKCZ can exert its
tumourigenic effect more prominently in cells lacking BRCA1 functions; if so, further
characterization of the signalling pathways involved will be useful in identifying additional
potential therapeutic targets for the subset of patients with BRCA1 mutations.
In this thesis, I have demonstrated in vitro that over-expression of PRKCZ alters the
expression of IGF1R and ITGB3 in some ovarian cancer cells. Since IGF1R and ITGB3
expressions have previously been found to be associated with aggressiveness and outcome of
ovarian cancer (214, 248-250, 258), it would be relevant to determine if there is a correlation
of protein expression among these three proteins in ovarian tumour specimens by
immunohistochemical analyses. Clinical data such as family history, BRCA mutation status,
or other relevant information such as p53 and ER status of a larger cohort of familial ovarian
cancer should also be collected and used to determine the significance of concurrent protein
expression alterations in these tumours, and to relate this correlation to patient outcome.
My identification of IGF1R and ITGB3 as potential downstream effectors of
PRKCZ has shed new light on some of the mechanisms by which PRKCZ can affect
ovarian tumour progression. The regulatory and cooperative roles of these proteins in
relation to ovarian cancer should therefore be examined in additional functional studies.
The involvement of the IGF1/IGF1R axis in ovarian cancer has been
demonstrated in multiple studies (180, 248, 249, 253, 275, 276). My observation of
PRKCZ-induced IGF1R expression alterations in ovarian cancers is a novel finding and
requires additional investigations to address the exact biological mechanisms that are
185
involved. PRKCZ appears to have multiple roles in the regulation of IGF1R, as shown
by its ability to both increase and decrease IGF1R expression levels in SKOV3 and
OVCAR3 cells, respectively. The lack of increased IGF1R mRNA expression in SKOV3
cells over-expressing PRKCZ suggests that increased IGF1R protein level may be due to
decreased protein degradation. To examine if direct physical interaction between
PRKCZ and IGF1R plays role in protein stability, localization and co-
immunoprecipitation studies should be conducted in these cells. Alterations in PRKCZ-
related signalling cascades (eg. Raf/Mek/MAPK pathway) may also be involved in
IGF1R stability, by ultimately inhibiting the activity of molecules involved in IGF1R
degradation. This hypothesis can be tested by using antagonists that target the different
signalling molecules, followed by IGF1R protein expression detection. Furthermore,
since over-expression of PRKCZ in OVCAR3 was shown to repress the gene expression
of IGF1R, PRKCZ-siRNA knockdown experiments should be performed to examine if
decreasing PRKCZ levels would have an opposite effect, resulting in increased IGF1R
expression, relative to non-treated parental cells, and further examine if these expression
alterations are associated with changes in proliferation and migration properties of
OVCAR3.
The two recent studies conducted by Partheen et al. have suggested ITGB3 to be a
potential prognostic marker for ovarian cancer, as low protein level of this protein in
advanced stage serous carcinomas is associated with poor patient outcome; however, the
regulatory mechanisms for ITGB3 expression have yet to be determined (183, 258).
Given that I have shown in my functional studies that over-expression of PRKCZ in
ovarian cancer cells decreases the expression of ITGB3, it may be relevant to target
186
molecules in PRKCZ-associated pathways in vitro, by the use of antagonists, to elucidate
the pathway(s) of which PRKCZ can regulate ITGB3 expression. Moreover, since
normal ovarian epithelium and highly-differentiated ovarian carcinomas have been
shown to have higher expression levels of ITGB3 compared to poorly differentiated
carcinomas (277), it may be useful to examine if over-expressing PRKCZ in cell lines
derived from normal ovarian epithelium and highly-differentiated carcinomas can indeed
decrease expression of ITGB3, to result in a less-differentiated and a more aggressive
phenotype in these cells. Additionally, as mentioned in discussion section of Chapter 4,
the apoptotic role of ITGB3 has yet to be examined in ovarian cancer. Therefore, it may
also be worthwhile to examine if decreased expression of ITGB3 in ovarian cancer cells
are associated with cell survival, by correlating its expression with apoptosis markers
such as Bax and Bcl-2, either through in vitro or immunohistochemical studies.
These proposed experiments will expand our knowledge in the involvement of
PRKCZ as well as its downstream targets in ovarian tumourigenesis. Elucidation of their
expressions in a larger cohort of ovarian tumour samples, as well as their correlation with
disease outcome and family history, may therefore be helpful in the development of novel
therapeutics for ovarian cancer patients with disease predisposition.
Appendix
187
188
Appendix A1. Family history data of ovarian cancer patients from present study. Classification of serous ovarian tumour samples according to the patients’ family history data. Nine tumours were classified as “strong familial”, and 27 tumours were classified as “weak familial”. The mean age of diagnosis for the “strong familial” and “weak familial” groups were 51 and 62 years, respectively.
Bibliography
189
190
1. Herceg Z, Hainaut P. Genetic and epigenetic alterations as biomarkers for cancer detection, diagnosis and prognosis. Mol Oncol 2007;1(1):26-41. 2. Jones PA, Baylin SB. The epigenomics of cancer. Cell 2007;128(4):683-92. 3. Hanahan D, Weinberg RA. The hallmarks of cancer. Cell 2000;100(1):57-70. 4. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell 2011;144(5):646-74. 5. Fearon ER, Vogelstein B. A genetic model for colorectal tumorigenesis. Cell 1990;61(5):759-67. 6. Carethers JM. The cellular and molecular pathogenesis of colorectal cancer. Gastroenterol Clin North Am 1996;25(4):737-54. 7. Macleod K. Tumor suppressor genes. Curr Opin Genet Dev 2000;10(1):81-93. 8. Kinzler KW, Vogelstein B. Cancer-susceptibility genes. Gatekeepers and caretakers. Nature 1997;386(6627):761, 3. 9. Michor F, Iwasa Y, Nowak MA. Dynamics of cancer progression. Nat Rev Cancer 2004;4(3):197-205. 10. Knudson AG, Jr. Mutation and cancer: statistical study of retinoblastoma. Proc Natl Acad Sci U S A 1971;68(4):820-3. 11. Santarosa M, Ashworth A. Haploinsufficiency for tumour suppressor genes: when you don't need to go all the way. Biochim Biophys Acta 2004;1654(2):105-22. 12. Katso RM, Manek S, O'Byrne K, Playford MP, Le Meuth V, Ganesan TS. Molecular approaches to diagnosis and management of ovarian cancer. Cancer Metastasis Rev 1997;16(1-2):81-107. 13. Palii SS, Robertson KD. Epigenetic control of tumor suppression. Crit Rev Eukaryot Gene Expr 2007;17(4):295-316. 14. Richards FM. Molecular pathology of von HippelLindau disease and the VHL tumour suppressor gene. Expert Rev Mol Med 2001;2001:1-27. 15. Boente MP, Hurteau J, Rodriguez GC, Bast RC, Jr., Berchuck A. The biology of ovarian cancer. Curr Opin Oncol 1993;5(5):900-7. 16. Munger K. Disruption of oncogene/tumor suppressor networks during human carcinogenesis. Cancer Invest 2002;20(1):71-81. 17. Smith MR, Matthews NT, Jones KA, Kung HF. Biological actions of oncogenes. Pharmacol Ther 1993;58(2):211-36. 18. Rak J, Filmus J, Finkenzeller G, Grugel S, Marme D, Kerbel RS. Oncogenes as inducers of tumor angiogenesis. Cancer Metastasis Rev 1995;14(4):263-77. 19. Meyer N, Penn LZ. Reflecting on 25 years with MYC. Nat Rev Cancer 2008;8(12):976-90. 20. Vogelstein B, Kinzler KW. Cancer genes and the pathways they control. Nat Med 2004;10(8):789-99. 21. Levitt NC, Hickson ID. Caretaker tumour suppressor genes that defend genome integrity. Trends Mol Med 2002;8(4):179-86. 22. Iwasa Y, Michor F, Komarova NL, Nowak MA. Population genetics of tumor suppressor genes. J Theor Biol 2005;233(1):15-23. 23. Weaver VM, Gilbert P. Watch thy neighbor: cancer is a communal affair. J Cell Sci 2004;117(Pt 8):1287-90. 24. Polyak K, Haviv I, Campbell IG. Co-evolution of tumor cells and their microenvironment. Trends Genet 2009;25(1):30-8.
191
25. Weber F, Shen L, Fukino K, et al. Total-genome analysis of BRCA1/2-related invasive carcinomas of the breast identifies tumor stroma as potential landscaper for neoplastic initiation. Am J Hum Genet 2006;78(6):961-72. 26. Kurose K, Gilley K, Matsumoto S, Watson PH, Zhou XP, Eng C. Frequent somatic mutations in PTEN and TP53 are mutually exclusive in the stroma of breast carcinomas. Nat Genet 2002;32(3):355-7. 27. Weber F, Fukino K, Sawada T, et al. Variability in organ-specific EGFR mutational spectra in tumour epithelium and stroma may be the biological basis for differential responses to tyrosine kinase inhibitors. Br J Cancer 2005;92(10):1922-6. 28. Moloughney B, Snider J, Villeneuve L. Ovarian cancer in Canada. Cmaj 2000;162(5):690. 29. Canadian Cancer Society's Steering Committee: Canadian Cancer Statistics 2009. Toronto: Canadian Cancer Society, 2009.; 2009. 30. Ramus SJ, Gayther SA. The contribution of BRCA1 and BRCA2 to ovarian cancer. Mol Oncol 2009;3(2):138-50. 31. Werness BA, Eltabbakh GH. Familial ovarian cancer and early ovarian cancer: biologic, pathologic, and clinical features. Int J Gynecol Pathol 2001;20(1):48-63. 32. Cannistra SA. Cancer of the ovary. N Engl J Med 2004;351(24):2519-29. 33. Hennessy BT, Coleman RL, Markman M. Ovarian cancer. Lancet 2009;374(9698):1371-82. 34. Landen CN, Jr., Birrer MJ, Sood AK. Early events in the pathogenesis of epithelial ovarian cancer. J Clin Oncol 2008;26(6):995-1005. 35. Cho KR, Shih Ie M. Ovarian cancer. Annu Rev Pathol 2009;4:287-313. 36. Koulouris CR, Penson RT. Ovarian stromal and germ cell tumors. Semin Oncol 2009;36(2):126-36. 37. Lawrenson K, Gayther SA. Ovarian cancer: a clinical challenge that needs some basic answers. PLoS Med 2009;6(2):e25. 38. Shih Ie M, Kurman RJ. Ovarian tumorigenesis: a proposed model based on morphological and molecular genetic analysis. Am J Pathol 2004;164(5):1511-8. 39. Cho KR. Ovarian cancer update: lessons from morphology, molecules, and mice. Arch Pathol Lab Med 2009;133(11):1775-81. 40. Dehari R, Kurman RJ, Logani S, Shih Ie M. The development of high-grade serous carcinoma from atypical proliferative (borderline) serous tumors and low-grade micropapillary serous carcinoma: a morphologic and molecular genetic analysis. Am J Surg Pathol 2007;31(7):1007-12. 41. Barnes MN, Grizzle WE, Grubbs CJ, Partridge EE. Paradigms for primary prevention of ovarian carcinoma. CA Cancer J Clin 2002;52(4):216-25. 42. Sueblinvong T, Carney ME. Current understanding of risk factors for ovarian cancer. Curr Treat Options Oncol 2009;10(1-2):67-81. 43. Parazzini F, Franceschi S, La Vecchia C, Fasoli M. The epidemiology of ovarian cancer. Gynecol Oncol 1991;43(1):9-23. 44. Wong AS, Auersperg N. Ovarian surface epithelium: family history and early events in ovarian cancer. Reprod Biol Endocrinol 2003;1:70. 45. Russo A, Calo V, Bruno L, Rizzo S, Bazan V, Di Fede G. Hereditary ovarian cancer. Crit Rev Oncol Hematol 2009;69(1):28-44.
192
46. Piver MS. Hereditary ovarian cancer. Lessons from the first twenty years of the Gilda Radner Familial Ovarian Cancer Registry. Gynecol Oncol 2002;85(1):9-17. 47. Prat J, Ribe A, Gallardo A. Hereditary ovarian cancer. Hum Pathol 2005;36(8):861-70. 48. Lynch HT, Casey MJ, Snyder CL, et al. Hereditary ovarian carcinoma: heterogeneity, molecular genetics, pathology, and management. Mol Oncol 2009;3(2):97-137. 49. Boardman LA, Thibodeau SN, Schaid DJ, et al. Increased risk for cancer in patients with the Peutz-Jeghers syndrome. Ann Intern Med 1998;128(11):896-9. 50. Gorlin RJ. Nevoid basal cell carcinoma syndrome. Dermatol Clin 1995;13(1):113-25. 51. Nelen MR, Padberg GW, Peeters EA, et al. Localization of the gene for Cowden disease to chromosome 10q22-23. Nat Genet 1996;13(1):114-6. 52. Pharoah PD, Ponder BA. The genetics of ovarian cancer. Best Pract Res Clin Obstet Gynaecol 2002;16(4):449-68. 53. Bewtra C, Watson P, Conway T, Read-Hippee C, Lynch HT. Hereditary ovarian cancer: a clinicopathological study. Int J Gynecol Pathol 1992;11(3):180-7. 54. Drake AC, Campbell H, Porteous ME, Dunlop MG. The contribution of DNA mismatch repair gene defects to the burden of gynecological cancer. Int J Gynecol Cancer 2003;13(3):262-77. 55. Peltomaki P. Role of DNA mismatch repair defects in the pathogenesis of human cancer. J Clin Oncol 2003;21(6):1174-9. 56. Wheeler JM, Bodmer WF, Mortensen NJ. DNA mismatch repair genes and colorectal cancer. Gut 2000;47(1):148-53. 57. Arzimanoglou, II, Lallas T, Osborne M, Barber H, Gilbert F. Microsatellite instability differences between familial and sporadic ovarian cancers. Carcinogenesis 1996;17(9):1799-804. 58. Christie M, Oehler MK. Molecular pathology of epithelial ovarian cancer. J Br Menopause Soc 2006;12(2):57-63. 59. Tomasini R, Mak TW, Melino G. The impact of p53 and p73 on aneuploidy and cancer. Trends Cell Biol 2008;18(5):244-52. 60. Okuda T, Otsuka J, Sekizawa A, et al. p53 mutations and overexpression affect prognosis of ovarian endometrioid cancer but not clear cell cancer. Gynecol Oncol 2003;88(3):318-25. 61. Gatcliffe TA, Monk BJ, Planutis K, Holcombe RF. Wnt signaling in ovarian tumorigenesis. Int J Gynecol Cancer 2008;18(5):954-62. 62. Liu F, Millar SE. Wnt/beta-catenin signaling in oral tissue development and disease. J Dent Res 2010;89(4):318-30. 63. Wu R, Zhai Y, Fearon ER, Cho KR. Diverse mechanisms of beta-catenin deregulation in ovarian endometrioid adenocarcinomas. Cancer Res 2001;61(22):8247-55. 64. Di Cristofano A, Ellenson LH. Endometrial carcinoma. Annu Rev Pathol 2007;2:57-85. 65. Campbell IG, Russell SE, Choong DY, et al. Mutation of the PIK3CA gene in ovarian and breast cancer. Cancer Res 2004;64(21):7678-81.
193
66. Kolasa IK, Rembiszewska A, Felisiak A, et al. PIK3CA amplification associates with resistance to chemotherapy in ovarian cancer patients. Cancer Biol Ther 2009;8(1):21-6. 67. Shiue H, Musch MW, Wang Y, Chang EB, Turner JR. Akt2 phosphorylates ezrin to trigger NHE3 translocation and activation. J Biol Chem 2005;280(2):1688-95. 68. Nakayama K, Nakayama N, Kurman RJ, et al. Sequence mutations and amplification of PIK3CA and AKT2 genes in purified ovarian serous neoplasms. Cancer Biol Ther 2006;5(7):779-85. 69. Bellacosa A, de Feo D, Godwin AK, et al. Molecular alterations of the AKT2 oncogene in ovarian and breast carcinomas. Int J Cancer 1995;64(4):280-5. 70. Rychahou PG, Kang J, Gulhati P, et al. Akt2 overexpression plays a critical role in the establishment of colorectal cancer metastasis. Proc Natl Acad Sci U S A 2008;105(51):20315-20. 71. Li AJ, Karlan BY. Genetic factors in ovarian carcinoma. Curr Oncol Rep 2001;3(1):27-32. 72. Berchuck A, Carney M. Human ovarian cancer of the surface epithelium. Biochem Pharmacol 1997;54(5):541-4. 73. Gemignani ML, Schlaerth AC, Bogomolniy F, et al. Role of KRAS and BRAF gene mutations in mucinous ovarian carcinoma. Gynecol Oncol 2003;90(2):378-81. 74. Masamha CP, Benbrook DM. Cyclin D1 degradation is sufficient to induce G1 cell cycle arrest despite constitutive expression of cyclin E2 in ovarian cancer cells. Cancer Res 2009;69(16):6565-72. 75. Nam EJ, Kim YT. Alteration of cell-cycle regulation in epithelial ovarian cancer. Int J Gynecol Cancer 2008;18(6):1169-82. 76. Dhar KK, Branigan K, Parkes J, et al. Expression and subcellular localization of cyclin D1 protein in epithelial ovarian tumour cells. Br J Cancer 1999;81(7):1174-81. 77. Barbieri F, Lorenzi P, Ragni N, et al. Overexpression of cyclin D1 is associated with poor survival in epithelial ovarian cancer. Oncology 2004;66(4):310-5. 78. D'Andrilli G, Kumar C, Scambia G, Giordano A. Cell cycle genes in ovarian cancer: steps toward earlier diagnosis and novel therapies. Clin Cancer Res 2004;10(24):8132-41. 79. Fujita M, Enomoto T, Murata Y. Genetic alterations in ovarian carcinoma: with specific reference to histological subtypes. Mol Cell Endocrinol 2003;202(1-2):97-9. 80. Marone M, Scambia G, Giannitelli C, et al. Analysis of cyclin E and CDK2 in ovarian cancer: gene amplification and RNA overexpression. Int J Cancer 1998;75(1):34-9. 81. Sui L, Dong Y, Ohno M, et al. Implication of malignancy and prognosis of p27(kip1), Cyclin E, and Cdk2 expression in epithelial ovarian tumors. Gynecol Oncol 2001;83(1):56-63. 82. Meinhold-Heerlein I, Bauerschlag D, Hilpert F, et al. Molecular and prognostic distinction between serous ovarian carcinomas of varying grade and malignant potential. Oncogene 2005;24(6):1053-65. 83. Maihle NJ, Baron AT, Barrette BA, et al. EGF/ErbB receptor family in ovarian cancer. Cancer Treat Res 2002;107:247-58.
194
84. Lassus H, Leminen A, Vayrynen A, et al. ERBB2 amplification is superior to protein expression status in predicting patient outcome in serous ovarian carcinoma. Gynecol Oncol 2004;92(1):31-9. 85. Mano MS, Awada A, Di Leo A, et al. Rates of topoisomerase II-alpha and HER-2 gene amplification and expression in epithelial ovarian carcinoma. Gynecol Oncol 2004;92(3):887-95. 86. McAlpine JN, Wiegand KC, Vang R, et al. HER2 overexpression and amplification is present in a subset of ovarian mucinous carcinomas and can be targeted with trastuzumab therapy. BMC Cancer 2009;9:433. 87. Tuefferd M, Couturier J, Penault-Llorca F, et al. HER2 status in ovarian carcinomas: a multicenter GINECO study of 320 patients. PLoS One 2007;2(11):e1138. 88. Vermeij J, Teugels E, Bourgain C, et al. Genomic activation of the EGFR and HER2-neu genes in a significant proportion of invasive epithelial ovarian cancers. BMC Cancer 2008;8:3. 89. Reibenwein J, Krainer M. Targeting signaling pathways in ovarian cancer. Expert Opin Ther Targets 2008;12(3):353-65. 90. Badache A, Hynes NE. A new therapeutic antibody masks ErbB2 to its partners. Cancer Cell 2004;5(4):299-301. 91. Chen GG, Zeng Q, Tse GM. Estrogen and its receptors in cancer. Med Res Rev 2008;28(6):954-74. 92. Pujol P, Rey JM, Nirde P, et al. Differential expression of estrogen receptor-alpha and -beta messenger RNAs as a potential marker of ovarian carcinogenesis. Cancer Res 1998;58(23):5367-73. 93. Issa RM, Lebeau A, Grob T, et al. Estrogen receptor gene amplification occurs rarely in ovarian cancer. Mod Pathol 2009;22(2):191-6. 94. Chan KK, Wei N, Liu SS, Xiao-Yun L, Cheung AN, Ngan HY. Estrogen receptor subtypes in ovarian cancer: a clinical correlation. Obstet Gynecol 2008;111(1):144-51. 95. Bardin A, Hoffmann P, Boulle N, et al. Involvement of estrogen receptor beta in ovarian carcinogenesis. Cancer Res 2004;64(16):5861-9. 96. Hall JM, Lee MK, Newman B, et al. Linkage of early-onset familial breast cancer to chromosome 17q21. Science 1990;250(4988):1684-9. 97. Narod SA, Feunteun J, Lynch HT, et al. Familial breast-ovarian cancer locus on chromosome 17q12-q23. Lancet 1991;338(8759):82-3. 98. Miki Y, Swensen J, Shattuck-Eidens D, et al. A strong candidate for the breast and ovarian cancer susceptibility gene BRCA1. Science 1994;266(5182):66-71. 99. Wooster R, Neuhausen SL, Mangion J, et al. Localization of a breast cancer susceptibility gene, BRCA2, to chromosome 13q12-13. Science 1994;265(5181):2088-90. 100. Carvalho MA, Couch FJ, Monteiro AN. Functional assays for BRCA1 and BRCA2. Int J Biochem Cell Biol 2007;39(2):298-310. 101. Narod SA, Foulkes WD. BRCA1 and BRCA2: 1994 and beyond. Nat Rev Cancer 2004;4(9):665-76. 102. Huen MS, Sy SM, Chen J. BRCA1 and its toolbox for the maintenance of genome integrity. Nat Rev Mol Cell Biol 2010;11(2):138-48. 103. Williams RS, Green R, Glover JN. Crystal structure of the BRCT repeat region from the breast cancer-associated protein BRCA1. Nat Struct Biol 2001;8(10):838-42.
195
104. Gudmundsdottir K, Ashworth A. The roles of BRCA1 and BRCA2 and associated proteins in the maintenance of genomic stability. Oncogene 2006;25(43):5864-74. 105. Antoniou A, Pharoah PD, Narod S, et al. Average risks of breast and ovarian cancer associated with BRCA1 or BRCA2 mutations detected in case Series unselected for family history: a combined analysis of 22 studies. Am J Hum Genet 2003;72(5):1117-30. 106. Venkitaraman AR. Cancer susceptibility and the functions of BRCA1 and BRCA2. Cell 2002;108(2):171-82. 107. Chenevix-Trench G, Healey S, Lakhani S, et al. Genetic and histopathologic evaluation of BRCA1 and BRCA2 DNA sequence variants of unknown clinical significance. Cancer Res 2006;66(4):2019-27. 108. Meric-Bernstam F. Heterogenic loss of BRCA in breast cancer: the "two-hit" hypothesis takes a hit. Ann Surg Oncol 2007;14(9):2428-9. 109. Chan KY, Ozcelik H, Cheung AN, Ngan HY, Khoo US. Epigenetic factors controlling the BRCA1 and BRCA2 genes in sporadic ovarian cancer. Cancer Res 2002;62(14):4151-6. 110. Dworkin AM, Spearman AD, Tseng SY, Sweet K, Toland AE. Methylation not a frequent "second hit" in tumors with germline BRCA mutations. Fam Cancer 2009. 111. Mazoyer S. Genomic rearrangements in the BRCA1 and BRCA2 genes. Hum Mutat 2005;25(5):415-22. 112. Walsh T, Lee MK, Casadei S, et al. Detection of inherited mutations for breast and ovarian cancer using genomic capture and massively parallel sequencing. Proc Natl Acad Sci U S A;107(28):12629-33. 113. Sluiter MD, van Rensburg EJ. Large genomic rearrangements of the BRCA1 and BRCA2 genes: review of the literature and report of a novel BRCA1 mutation. Breast Cancer Res Treat;125(2):325-49. 114. del Valle J, Feliubadalo L, Nadal M, et al. Identification and comprehensive characterization of large genomic rearrangements in the BRCA1 and BRCA2 genes. Breast Cancer Res Treat;122(3):733-43. 115. Kasprzak L, Foulkes WD, Shelling AN. Forth nightly review: hereditary ovarian carcinoma. Bmj 1999;318(7186):786-9. 116. Gayther SA, Mangion J, Russell P, et al. Variation of risks of breast and ovarian cancer associated with different germline mutations of the BRCA2 gene. Nat Genet 1997;15(1):103-5. 117. Sowter HM, Ashworth A. BRCA1 and BRCA2 as ovarian cancer susceptibility genes. Carcinogenesis 2005;26(10):1651-6. 118. Thompson D, Easton D. Variation in cancer risks, by mutation position, in BRCA2 mutation carriers. Am J Hum Genet 2001;68(2):410-9. 119. Szabo CI, King MC. Population genetics of BRCA1 and BRCA2. Am J Hum Genet 1997;60(5):1013-20. 120. Tonin PN, Mes-Masson AM, Futreal PA, et al. Founder BRCA1 and BRCA2 mutations in French Canadian breast and ovarian cancer families. Am J Hum Genet 1998;63(5):1341-51. 121. Tonin PN, Perret C, Lambert JA, et al. Founder BRCA1 and BRCA2 mutations in early-onset French Canadian breast cancer cases unselected for family history. Int J Cancer 2001;95(3):189-93.
196
122. Yarden RI, Friedman E, Metsuyanim S, Olender T, Ben-Asher E, Papa MZ. MDM2 SNP309 accelerates breast and ovarian carcinogenesis in BRCA1 and BRCA2 carriers of Jewish-Ashkenazi descent. Breast Cancer Res Treat 2008;111(3):497-504. 123. Jakubowska A, Gronwald J, Menkiszak J, et al. Integrin beta3 Leu33Pro polymorphism increases BRCA1-associated ovarian cancer risk. J Med Genet 2007;44(6):408-11. 124. Jakubowska A, Rozkrut D, Antoniou A, Hamann U, Lubinski J. The Leu33Pro polymorphism in the ITGB3 gene does not modify BRCA1/2-associated breast or ovarian cancer risks: results from a multicenter study among 15,542 BRCA1 and BRCA2 mutation carriers. Breast Cancer Res Treat 2010;121(3):639-49. 125. Ramus SJ, Kartsonaki C, Gayther SA, et al. Genetic Variation at 9p22.2 and Ovarian Cancer Risk for BRCA1 and BRCA2 Mutation Carriers. J Natl Cancer Inst 2010. 126. Narod SA, Risch H, Moslehi R, et al. Oral contraceptives and the risk of hereditary ovarian cancer. Hereditary Ovarian Cancer Clinical Study Group. N Engl J Med 1998;339(7):424-8. 127. Narod SA, Sun P, Ghadirian P, et al. Tubal ligation and risk of ovarian cancer in carriers of BRCA1 or BRCA2 mutations: a case-control study. Lancet 2001;357(9267):1467-70. 128. Lambros MB, Fiegler H, Jones A, et al. Analysis of ovarian cancer cell lines using array-based comparative genomic hybridization. J Pathol 2005;205(1):29-40. 129. Mayr D, Kanitz V, Anderegg B, et al. Analysis of gene amplification and prognostic markers in ovarian cancer using comparative genomic hybridization for microarrays and immunohistochemical analysis for tissue microarrays. Am J Clin Pathol 2006;126(1):101-9. 130. Leunen K, Gevaert O, Daemen A, et al. Recurrent copy number alterations in BRCA1-mutated ovarian tumors alter biological pathways. Hum Mutat 2009;30(12):1693-702. 131. Mhawech-Fauceglia P, Rai H, Nowak N, et al. The use of array-based comparative genomic hybridization (a-CGH) to distinguish metastatic from primary synchronous carcinomas of the ovary and the uterus. Histopathology 2008;53(4):490-5. 132. Osterberg L, Levan K, Partheen K, Staaf J, Sundfeldt K, Horvath G. High-resolution genomic profiling of carboplatin resistance in early-stage epithelial ovarian carcinoma. Cytogenet Genome Res 2009;125(1):8-18. 133. Osterberg L, Levan K, Partheen K, et al. Potential predictive markers of chemotherapy resistance in stage III ovarian serous carcinomas. BMC Cancer 2009;9:368. 134. Walsh CS, Ogawa S, Scoles DR, et al. Genome-wide loss of heterozygosity and uniparental disomy in BRCA1/2-associated ovarian carcinomas. Clin Cancer Res 2008;14(23):7645-51. 135. Yoshihara K, Tajima A, Adachi S, et al. Germline copy number variations in BRCA1-associated ovarian cancer patients. Genes Chromosomes Cancer. 136. Gorringe KL, Ramakrishna M, Williams LH, et al. Are there any more ovarian tumor suppressor genes? A new perspective using ultra high-resolution copy number and loss of heterozygosity analysis. Genes Chromosomes Cancer 2009;48(10):931-42.
197
137. Goode EL, White KL, Vierkant RA, et al. Xenobiotic-Metabolizing gene polymorphisms and ovarian cancer risk. Mol Carcinog. 138. Engel C, Versmold B, Wappenschmidt B, et al. Association of the variants CASP8 D302H and CASP10 V410I with breast and ovarian cancer risk in BRCA1 and BRCA2 mutation carriers. Cancer Epidemiol Biomarkers Prev 2010;19(11):2859-68. 139. Monaco A, Menolascina F, Zhao Y, et al. "Sequencing-grade" screening for BRCA1 variants by oligo-arrays. J Transl Med 2008;6:64. 140. Fehrmann RS, Li XY, van der Zee AG, et al. Profiling studies in ovarian cancer: a review. Oncologist 2007;12(8):960-6. 141. Le Page C, Ouellet V, Madore J, et al. From gene profiling to diagnostic markers: IL-18 and FGF-2 complement CA125 as serum-based markers in epithelial ovarian cancer. Int J Cancer 2006;118(7):1750-8. 142. Berchuck A, Iversen ES, Luo J, et al. Microarray analysis of early stage serous ovarian cancers shows profiles predictive of favorable outcome. Clin Cancer Res 2009;15(7):2448-55. 143. McGuire WP, Ozols RF. Chemotherapy of advanced ovarian cancer. Semin Oncol 1998;25(3):340-8. 144. L'Esperance S, Popa I, Bachvarova M, et al. Gene expression profiling of paired ovarian tumors obtained prior to and following adjuvant chemotherapy: molecular signatures of chemoresistant tumors. Int J Oncol 2006;29(1):5-24. 145. Bild AH, Yao G, Chang JT, et al. Oncogenic pathway signatures in human cancers as a guide to targeted therapies. Nature 2006;439(7074):353-7. 146. Sunde JS, Donninger H, Wu K, et al. Expression profiling identifies altered expression of genes that contribute to the inhibition of transforming growth factor-beta signaling in ovarian cancer. Cancer Res 2006;66(17):8404-12. 147. Crum CP, Drapkin R, Kindelberger D, Medeiros F, Miron A, Lee Y. Lessons from BRCA: the tubal fimbria emerges as an origin for pelvic serous cancer. Clin Med Res 2007;5(1):35-44. 148. Tone AA, Begley H, Sharma M, et al. Gene expression profiles of luteal phase fallopian tube epithelium from BRCA mutation carriers resemble high-grade serous carcinoma. Clin Cancer Res 2008;14(13):4067-78. 149. Motamed-Khorasani A, Jurisica I, Letarte M, et al. Differentially androgen-modulated genes in ovarian epithelial cells from BRCA mutation carriers and control patients predict ovarian cancer survival and disease progression. Oncogene 2007;26(2):198-214. 150. Evangelou A, Letarte M, Jurisica I, et al. Loss of coordinated androgen regulation in nonmalignant ovarian epithelial cells with BRCA1/2 mutations and ovarian cancer cells. Cancer Res 2003;63(10):2416-24. 151. Jiang X, Zhu T, Yang J, et al. Identification of novel epithelial ovarian cancer biomarkers by cross-laboratory microarray analysis. J Huazhong Univ Sci Technolog Med Sci;30(3):354-9. 152. Yoshihara K, Tajima A, Yahata T, et al. Gene expression profile for predicting survival in advanced-stage serous ovarian cancer across two independent datasets. PLoS One;5(3):e9615.
198
153. Yamaguchi K, Mandai M, Oura T, et al. Identification of an ovarian clear cell carcinoma gene signature that reflects inherent disease biology and the carcinogenic processes. Oncogene 2010;29(12):1741-52. 154. Crijns AP, Fehrmann RS, de Jong S, et al. Survival-related profile, pathways, and transcription factors in ovarian cancer. PLoS Med 2009;6(2):e24. 155. Jazaeri AA, Yee CJ, Sotiriou C, Brantley KR, Boyd J, Liu ET. Gene expression profiles of BRCA1-linked, BRCA2-linked, and sporadic ovarian cancers. J Natl Cancer Inst 2002;94(13):990-1000. 156. Walker LC, Waddell N, Ten Haaf A, Grimmond S, Spurdle AB. Use of expression data and the CGEMS genome-wide breast cancer association study to identify genes that may modify risk in BRCA1/2 mutation carriers. Breast Cancer Res Treat 2008;112(2):229-36. 157. Yang YH, Dudoit S, Luu P, et al. Normalization for cDNA microarray data: a robust composite method addressing single and multiple slide systematic variation. Nucleic Acids Res 2002;30(4):e15. 158. Smyth GK. Limma: linear models for microarray data. . New York: Springer; 2005. 159. Tusher VG, Tibshirani R, Chu G. Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci U S A 2001;98(9):5116-21. 160. Allison SJ, Milner J. SIRT3 is pro-apoptotic and participates in distinct basal apoptotic pathways. Cell Cycle 2007;6(21):2669-77. 161. Heyman L, Leroy-Dudal J, Fernandes J, Seyer D, Dutoit S, Carreiras F. Mesothelial vitronectin stimulates migration of ovarian cancer cells. Cell Biol Int 2010;34(5):493-502. 162. Vielmetter J, Chen XN, Miskevich F, et al. Molecular characterization of human neogenin, a DCC-related protein, and the mapping of its gene (NEO1) to chromosomal position 15q22.3-q23. Genomics 1997;41(3):414-21. 163. Latil A, Chene L, Cochant-Priollet B, et al. Quantification of expression of netrins, slits and their receptors in human prostate tumors. Int J Cancer 2003;103(3):306-15. 164. Goke R, Gregel C, Goke A, Arnold R, Schmidt H, Lankat-Buttgereit B. Programmed cell death protein 4 (PDCD4) acts as a tumor suppressor in neuroendocrine tumor cells. Ann N Y Acad Sci 2004;1014:220-1. 165. Gao F, Zhang P, Zhou C, et al. Frequent loss of PDCD4 expression in human glioma: possible role in the tumorigenesis of glioma. Oncol Rep 2007;17(1):123-8. 166. Goke R, Barth P, Schmidt A, Samans B, Lankat-Buttgereit B. Programmed cell death protein 4 suppresses CDK1/cdc2 via induction of p21(Waf1/Cip1). Am J Physiol Cell Physiol 2004;287(6):C1541-6. 167. Wei ZT, Zhang X, Wang XY, et al. PDCD4 inhibits the malignant phenotype of ovarian cancer cells. Cancer Sci 2009. 168. Murphy N, Ring M, Heffron CC, et al. Quantitation of CDC6 and MCM5 mRNA in cervical intraepithelial neoplasia and invasive squamous cell carcinoma of the cervix. Mod Pathol 2005;18(6):844-9. 169. Williams GH, Swinn R, Prevost AT, et al. Diagnosis of oesophageal cancer by detection of minichromosome maintenance 5 protein in gastric aspirates. Br J Cancer 2004;91(4):714-9.
199
170. Terry K, McGrath M, Lee IM, Buring J, De Vivo I. Genetic variation in CYP11A1 and StAR in relation to endometrial cancer risk. Gynecol Oncol 2010;117(2):255-9. 171. Celhar T, Gersak K, Ovcak Z, Sedmak B, Mlinaric-Rascan I. The presence of the CYP11A1 (TTTTA)6 allele increases the risk of biochemical relapse in organ confined and low-grade prostate cancer. Cancer Genet Cytogenet 2008;187(1):28-33. 172. Rodriguez EM, Dunham EE, Martin GS. Atypical protein kinase C activity is required for extracellular matrix degradation and invasion by Src-transformed cells. J Cell Physiol 2009;221(1):171-82. 173. Grifoni D, Garoia F, Bellosta P, et al. aPKCzeta cortical loading is associated with Lgl cytoplasmic release and tumor growth in Drosophila and human epithelia. Oncogene 2007;26(40):5960-5. 174. Guo H, Gu F, Li W, et al. Reduction of protein kinase C zeta inhibits migration and invasion of human glioblastoma cells. J Neurochem 2009;109(1):203-13. 175. Hirai T, Chida K. Protein kinase Czeta (PKCzeta): activation mechanisms and cellular functions. J Biochem 2003;133(1):1-7. 176. Liu Y, Wang B, Wang J, et al. Down-regulation of PKCzeta expression inhibits chemotaxis signal transduction in human lung cancer cells. Lung Cancer 2009;63(2):210-8. 177. Sun R, Gao P, Chen L, et al. Protein kinase C zeta is required for epidermal growth factor-induced chemotaxis of human breast cancer cells. Cancer Res 2005;65(4):1433-41. 178. Xiao H, Goldthwait DA, Mapstone T. The identification of four protein kinase C isoforms in human glioblastoma cell lines: PKC alpha, gamma, epsilon, and zeta. J Neurosurg 1994;81(5):734-40. 179. Denison SR, Callahan G, Becker NA, Phillips LA, Smith DI. Characterization of FRA6E and its potential role in autosomal recessive juvenile parkinsonism and ovarian cancer. Genes Chromosomes Cancer 2003;38(1):40-52. 180. Gotlieb WH, Bruchim I, Gu J, et al. Insulin-like growth factor receptor I targeting in epithelial ovarian cancer. Gynecol Oncol 2006;100(2):389-96. 181. Helleman J, Jansen MP, Span PN, et al. Molecular profiling of platinum resistant ovarian cancer. Int J Cancer 2006;118(8):1963-71. 182. Heyman L, Kellouche S, Fernandes J, Dutoit S, Poulain L, Carreiras F. Vitronectin and its receptors partly mediate adhesion of ovarian cancer cells to peritoneal mesothelium in vitro. Tumour Biol 2008;29(4):231-44. 183. Partheen K, Levan K, Osterberg L, et al. Four potential biomarkers as prognostic factors in stage III serous ovarian adenocarcinomas. Int J Cancer 2008;123(9):2130-7. 184. Runnebaum IB, Bruning A. Glucocorticoids inhibit cell death in ovarian cancer and up-regulate caspase inhibitor cIAP2. Clin Cancer Res 2005;11(17):6325-32. 185. Sugai M, Umezu H, Yamamoto T, et al. Expression of hepatocyte nuclear factor 4 alpha in primary ovarian mucinous tumors. Pathol Int 2008;58(11):681-6. 186. Wei Y, Xia W, Zhang Z, et al. Loss of trimethylation at lysine 27 of histone H3 is a predictor of poor outcome in breast, ovarian, and pancreatic cancers. Mol Carcinog 2008;47(9):701-6.
200
187. Caslini C, Capo-chichi CD, Roland IH, Nicolas E, Yeung AT, Xu XX. Histone modifications silence the GATA transcription factor genes in ovarian cancer. Oncogene 2006;25(39):5446-61. 188. Ye H, Pungpravat N, Huang BL, et al. Genomic assessments of the frequent loss of heterozygosity region on 8p21.3-p22 in head and neck squamous cell carcinoma. Cancer Genet Cytogenet 2007;176(2):100-6. 189. Lux MP, Fasching PA, Beckmann MW. Hereditary breast and ovarian cancer: review and future perspectives. J Mol Med 2006;84(1):16-28. 190. Choi KC, Kang SK, Tai CJ, Auersperg N, Leung PC. Estradiol up-regulates antiapoptotic Bcl-2 messenger ribonucleic acid and protein in tumorigenic ovarian surface epithelium cells. Endocrinology 2001;142(6):2351-60. 191. Ding JX, Feng YJ, Yao LQ, Yu M, Jin HY, Yin LH. The reinforcement of invasion in epithelial ovarian cancer cells by 17 beta-Estradiol is associated with up-regulation of Snail. Gynecol Oncol 2006;103(2):623-30. 192. Crusio KM, King B, Reavie LB, Aifantis I. The ubiquitous nature of cancer: the role of the SCF(Fbw7) complex in development and transformation. Oncogene 2010. 193. Tan Y, Sangfelt O, Spruck C. The Fbxw7/hCdc4 tumor suppressor in human cancer. Cancer Lett 2008;271(1):1-12. 194. Minella AC, Clurman BE. Mechanisms of tumor suppression by the SCF(Fbw7). Cell Cycle 2005;4(10):1356-9. 195. Skaar JR, Pagano M. Control of cell growth by the SCF and APC/C ubiquitin ligases. Curr Opin Cell Biol 2009;21(6):816-24. 196. Koepp DM, Schaefer LK, Ye X, et al. Phosphorylation-dependent ubiquitination of cyclin E by the SCFFbw7 ubiquitin ligase. Science 2001;294(5540):173-7. 197. Yada M, Hatakeyama S, Kamura T, et al. Phosphorylation-dependent degradation of c-Myc is mediated by the F-box protein Fbw7. Embo J 2004;23(10):2116-25. 198. Nateri AS, Riera-Sans L, Da Costa C, Behrens A. The ubiquitin ligase SCFFbw7 antagonizes apoptotic JNK signaling. Science 2004;303(5662):1374-8. 199. Wu G, Lyapina S, Das I, et al. SEL-10 is an inhibitor of notch signaling that targets notch for ubiquitin-mediated protein degradation. Mol Cell Biol 2001;21(21):7403-15. 200. Mao JH, Kim IJ, Wu D, et al. FBXW7 targets mTOR for degradation and cooperates with PTEN in tumor suppression. Science 2008;321(5895):1499-502. 201. Schwab M, Tyers M. Cell cycle. Archipelago of destruction. Nature 2001;413(6853):268-9. 202. Rajagopalan H, Jallepalli PV, Rago C, et al. Inactivation of hCDC4 can cause chromosomal instability. Nature 2004;428(6978):77-81. 203. Spruck CH, Strohmaier H, Sangfelt O, et al. hCDC4 gene mutations in endometrial cancer. Cancer Res 2002;62(16):4535-9. 204. Strohmaier H, Spruck CH, Kaiser P, Won KA, Sangfelt O, Reed SI. Human F-box protein hCdc4 targets cyclin E for proteolysis and is mutated in a breast cancer cell line. Nature 2001;413(6853):316-22. 205. Moberg KH, Bell DW, Wahrer DC, Haber DA, Hariharan IK. Archipelago regulates Cyclin E levels in Drosophila and is mutated in human cancer cell lines. Nature 2001;413(6853):311-6.
201
206. Li LC, Dahiya R. MethPrimer: designing primers for methylation PCRs. Bioinformatics 2002;18(11):1427-31. 207. Akhoondi S, Sun D, von der Lehr N, et al. FBXW7/hCDC4 is a general tumor suppressor in human cancer. Cancer Res 2007;67(19):9006-12. 208. Kimchi-Sarfaty C, Oh JM, Kim IW, et al. A "silent" polymorphism in the MDR1 gene changes substrate specificity. Science 2007;315(5811):525-8. 209. Kwak EL, Moberg KH, Wahrer DC, et al. Infrequent mutations of Archipelago (hAGO, hCDC4, Fbw7) in primary ovarian cancer. Gynecol Oncol 2005;98(1):124-8. 210. Wajed SA, Laird PW, DeMeester TR. DNA methylation: an alternative pathway to cancer. Ann Surg 2001;234(1):10-20. 211. Asadollahi R, Hyde CA, Zhong XY. Epigenetics of ovarian cancer: from the lab to the clinic. Gynecol Oncol 2010;118(1):81-7. 212. Gu Z, Inomata K, Mitsui H, Horii A. Promoter hypermethylation is not the major mechanism for inactivation of the FBXW7 beta-form in human gliomas. Genes Genet Syst 2008;83(4):347-52. 213. Mao JH, Perez-Losada J, Wu D, et al. Fbxw7/Cdc4 is a p53-dependent, haploinsufficient tumour suppressor gene. Nature 2004;432(7018):775-9. 214. Partheen K, Levan K, Osterberg L, Helou K, Horvath G. Analysis of cytogenetic alterations in stage III serous ovarian adenocarcinoma reveals a heterogeneous group regarding survival, surgical outcome, and substage. Genes Chromosomes Cancer 2004;40(4):342-8. 215. Sterian A, Kan T, Berki AT, et al. Mutational and LOH analyses of the chromosome 4q region in esophageal adenocarcinoma. Oncology 2006;70(3):168-72. 216. Milne AN, Leguit R, Corver WE, et al. Loss of CDC4/FBXW7 in gastric carinoma. Cell Oncol 2010;32(5-6):347-59. 217. Rosenberg E, Demopoulos RI, Zeleniuch-Jacquotte A, et al. Expression of cell cycle regulators p57(KIP2), cyclin D1, and cyclin E in epithelial ovarian tumors and survival. Hum Pathol 2001;32(8):808-13. 218. Schraml P, Bucher C, Bissig H, et al. Cyclin E overexpression and amplification in human tumours. J Pathol 2003;200(3):375-82. 219. Courjal F, Louason G, Speiser P, Katsaros D, Zeillinger R, Theillet C. Cyclin gene amplification and overexpression in breast and ovarian cancers: evidence for the selection of cyclin D1 in breast and cyclin E in ovarian tumors. Int J Cancer 1996;69(4):247-53. 220. Marchini S, Mariani P, Chiorino G, et al. Analysis of gene expression in early-stage ovarian cancer. Clin Cancer Res 2008;14(23):7850-60. 221. Rosen DG, Yang G, Deavers MT, et al. Cyclin E expression is correlated with tumor progression and predicts a poor prognosis in patients with ovarian carcinoma. Cancer 2006;106(9):1925-32. 222. Farley J, Smith LM, Darcy KM, et al. Cyclin E expression is a significant predictor of survival in advanced, suboptimally debulked ovarian epithelial cancers: a Gynecologic Oncology Group study. Cancer Res 2003;63(6):1235-41. 223. Donson AM, Banerjee A, Gamboni-Robertson F, Fleitz JM, Foreman NK. Protein kinase C zeta isoform is critical for proliferation in human glioblastoma cell lines. J Neurooncol 2000;47(2):109-15.
202
224. Huang L, Muthuswamy SK. Polarity protein alterations in carcinoma: a focus on emerging roles for polarity regulators. Curr Opin Genet Dev 2010;20(1):41-50. 225. Magdalena J, Millard TH, Machesky LM. Microtubule involvement in NIH 3T3 Golgi and MTOC polarity establishment. J Cell Sci 2003;116(Pt 4):743-56. 226. Moscat J, Sanz L, Sanchez P, Diaz-Meco MT. Regulation and role of the atypical PKC isoforms in cell survival during tumor transformation. Adv Enzyme Regul 2001;41:99-120. 227. Wang YM, Seibenhener ML, Vandenplas ML, Wooten MW. Atypical PKC zeta is activated by ceramide, resulting in coactivation of NF-kappaB/JNK kinase and cell survival. J Neurosci Res 1999;55(3):293-302. 228. Joung I, Kim HJ, Kwon YK. p62 modulates Akt activity via association with PKCzeta in neuronal survival and differentiation. Biochem Biophys Res Commun 2005;334(2):654-60. 229. Win HY, Acevedo-Duncan M. Atypical protein kinase C phosphorylates IKKalphabeta in transformed non-malignant and malignant prostate cell survival. Cancer Lett 2008;270(2):302-11. 230. Martin P, Duran A, Minguet S, et al. Role of zeta PKC in B-cell signaling and function. Embo J 2002;21(15):4049-57. 231. Kim DS, Jeon OH, Lee HD, Yoo KH, Kim DS. Integrin alphavbeta3-mediated transcriptional regulation of TIMP-1 in a human ovarian cancer cell line. Biochem Biophys Res Commun 2008;377(2):479-83. 232. Peng Y, Sigua CA, Murr MM. Protein kinase C-zeta mediates apoptosis of mouse Kupffer cells via ERK-1/2: A novel mechanism. Surgery. 233. Xin M, Gao F, May WS, Flagg T, Deng X. Protein kinase Czeta abrogates the proapoptotic function of Bax through phosphorylation. J Biol Chem 2007;282(29):21268-77. 234. Nazarenko I, Jenny M, Keil J, et al. Atypical protein kinase C zeta exhibits a proapoptotic function in ovarian cancer. Mol Cancer Res 2010;8(6):919-34. 235. Mustafi R, Cerda S, Chumsangsri A, Fichera A, Bissonnette M. Protein Kinase-zeta inhibits collagen I-dependent and anchorage-independent growth and enhances apoptosis of human Caco-2 cells. Mol Cancer Res 2006;4(9):683-94. 236. Cohen EE, Lingen MW, Zhu B, et al. Protein kinase C zeta mediates epidermal growth factor-induced growth of head and neck tumor cells by regulating mitogen-activated protein kinase. Cancer Res 2006;66(12):6296-303. 237. Della Peruta M, Giagulli C, Laudanna C, Scarpa A, Sorio C. RHOA and PRKCZ control different aspects of cell motility in pancreatic cancer metastatic clones. Mol Cancer 2010;9:61. 238. Moscat J, Diaz-Meco MT. The atypical protein kinase Cs. Functional specificity mediated by specific protein adapters. EMBO Rep 2000;1(5):399-403. 239. Eder AM, Sui X, Rosen DG, et al. Atypical PKCiota contributes to poor prognosis through loss of apical-basal polarity and cyclin E overexpression in ovarian cancer. Proc Natl Acad Sci U S A 2005;102(35):12519-24. 240. Baldwin RM, Barrett GM, Parolin DA, et al. Coordination of glioblastoma cell motility by PKCiota. Mol Cancer 2010;9:233.
203
241. Win HY, Acevedo-Duncan M. Role of protein kinase C-iota in transformed non-malignant RWPE-1 cells and androgen-independent prostate carcinoma DU-145 cells. Cell Prolif 2009;42(2):182-94. 242. Zhang L, Huang J, Yang N, et al. Integrative genomic analysis of protein kinase C (PKC) family identifies PKCiota as a biomarker and potential oncogene in ovarian carcinoma. Cancer Res 2006;66(9):4627-35. 243. Suzuki A, Akimoto K, Ohno S. Protein kinase C lambda/iota (PKClambda/iota): a PKC isotype essential for the development of multicellular organisms. J Biochem 2003;133(1):9-16. 244. Suzuki A, Yamanaka T, Hirose T, et al. Atypical protein kinase C is involved in the evolutionarily conserved par protein complex and plays a critical role in establishing epithelia-specific junctional structures. J Cell Biol 2001;152(6):1183-96. 245. Abramovitch S, Glaser T, Ouchi T, Werner H. BRCA1-Sp1 interactions in transcriptional regulation of the IGF-IR gene. FEBS Lett 2003;541(1-3):149-54. 246. Abramovitch S, Werner H. Functional and physical interactions between BRCA1 and p53 in transcriptional regulation of the IGF-IR gene. Horm Metab Res 2003;35(11-12):758-62. 247. Maor S, Yosepovich A, Papa MZ, et al. Elevated insulin-like growth factor-I receptor (IGF-IR) levels in primary breast tumors associated with BRCA1 mutations. Cancer Lett 2007;257(2):236-43. 248. An Y, Cai L, Wang Y, Zhu D, Guan Y, Zheng J. Local expression of insulin-like growth factor-I, insulin-like growth factor-I receptor, and estrogen receptor alpha in ovarian cancer. Onkologie 2009;32(11):638-44. 249. Ouban A, Muraca P, Yeatman T, Coppola D. Expression and distribution of insulin-like growth factor-1 receptor in human carcinomas. Hum Pathol 2003;34(8):803-8. 250. Eckstein N, Servan K, Hildebrandt B, et al. Hyperactivation of the insulin-like growth factor receptor I signaling pathway is an essential event for cisplatin resistance of ovarian cancer cells. Cancer Res 2009;69(7):2996-3003. 251. Yee D, Morales FR, Hamilton TC, Von Hoff DD. Expression of insulin-like growth factor I, its binding proteins, and its receptor in ovarian cancer. Cancer Res 1991;51(19):5107-12. 252. Spentzos D, Cannistra SA, Grall F, et al. IGF axis gene expression patterns are prognostic of survival in epithelial ovarian cancer. Endocr Relat Cancer 2007;14(3):781-90. 253. Beck EP, Russo P, Gliozzo B, et al. Identification of insulin and insulin-like growth factor I (IGF I) receptors in ovarian cancer tissue. Gynecol Oncol 1994;53(2):196-201. 254. Jakubowska A, Rozkrut D, Antoniou A, Hamann U, Lubinski J. The Leu33Pro polymorphism in the ITGB3 gene does not modify BRCA1/2-associated breast or ovarian cancer risks: results from a multicenter study among 15,542 BRCA1 and BRCA2 mutation carriers. Breast Cancer Res Treat;121(3):639-49. 255. Feng D, Lindpaintner K, Larson MG, et al. Increased platelet aggregability associated with platelet GPIIIa PlA2 polymorphism: the Framingham Offspring Study. Arterioscler Thromb Vasc Biol 1999;19(4):1142-7.
204
256. Vijayan KV, Liu Y, Dong JF, Bray PF. Enhanced activation of mitogen-activated protein kinase and myosin light chain kinase by the Pro33 polymorphism of integrin beta 3. J Biol Chem 2003;278(6):3860-7. 257. Kaur S, Kenny HA, Jagadeeswaran S, et al. {beta}3-integrin expression on tumor cells inhibits tumor progression, reduces metastasis, and is associated with a favorable prognosis in patients with ovarian cancer. Am J Pathol 2009;175(5):2184-96. 258. Partheen K, Levan K, Osterberg L, Claesson I, Sundfeldt K, Horvath G. External validation suggests Integrin beta 3 as prognostic biomarker in serous ovarian adenocarcinomas. BMC Cancer 2009;9:336. 259. Toker A, Yoeli-Lerner M. Akt signaling and cancer: surviving but not moving on. Cancer Res 2006;66(8):3963-6. 260. McCoy ML, Mueller CR, Roskelley CD. The role of the breast cancer susceptibility gene 1 (BRCA1) in sporadic epithelial ovarian cancer. Reprod Biol Endocrinol 2003;1:72. 261. Della Peruta M, Giagulli C, Laudanna C, Scarpa A, Sorio C. RHOA and PRKCZ control different aspects of cell motility in pancreatic cancer metastatic clones. Mol Cancer;9:61. 262. Neuhausen SL, Brummel S, Ding YC, et al. Genetic variation in insulin-like growth factor signaling genes and breast cancer risk among BRCA1 and BRCA2 carriers. Breast Cancer Res 2009;11(5):R76. 263. Chitnis MM, Yuen JS, Protheroe AS, Pollak M, Macaulay VM. The type 1 insulin-like growth factor receptor pathway. Clin Cancer Res 2008;14(20):6364-70. 264. Subramanian A, Tamayo P, Mootha VK, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A 2005;102(43):15545-50. 265. Amundadottir L, Kraft P, Stolzenberg-Solomon RZ, et al. Genome-wide association study identifies variants in the ABO locus associated with susceptibility to pancreatic cancer. Nat Genet 2009;41(9):986-90. 266. Do TN, Ucisik-Akkaya E, Davis CF, Morrison BA, Dorak MT. An intronic polymorphism of IRF4 gene influences gene transcription in vitro and shows a risk association with childhood acute lymphoblastic leukemia in males. Biochim Biophys Acta 2010;1802(2):292-300. 267. Gu F, Qureshi AA, Niu T, et al. Interleukin and interleukin receptor gene polymorphisms and susceptibility to melanoma. Melanoma Res 2008;18(5):330-5. 268. Jou YS, Lo YL, Hsiao CF, et al. Association of an EGFR intron 1 SNP with never-smoking female lung adenocarcinoma patients. Lung Cancer 2009;64(3):251-6. 269. Levin AM, Machiela MJ, Zuhlke KA, Ray AM, Cooney KA, Douglas JA. Chromosome 17q12 variants contribute to risk of early-onset prostate cancer. Cancer Res 2008;68(16):6492-5. 270. Wang-Gohrke S, Weikel W, Risch H, et al. Intron variants of the p53 gene are associated with increased risk for ovarian cancer but not in carriers of BRCA1 or BRCA2 germline mutations. Br J Cancer 1999;81(1):179-83. 271. Zhou J, Nagarkatti PS, Zhong Y, Creek K, Zhang J, Nagarkatti M. Unique SNP in CD44 intron 1 and its role in breast cancer development. Anticancer Res 2010;30(4):1263-72.
205
272. Nowak D, Mossner M, Baldus CD, Hopfer O, Thiel E, Hofmann WK. Mutation analysis of hCDC4 in AML cells identifies a new intronic polymorphism. Int J Med Sci 2006;3(4):148-51. 273. Yu JC, Ding SL, Chang CH, et al. Genetic susceptibility to the development and progression of breast cancer associated with polymorphism of cell cycle and ubiquitin ligase genes. Carcinogenesis 2009;30(9):1562-70. 274. DelloRusso C, Welcsh PL, Wang W, Garcia RL, King MC, Swisher EM. Functional characterization of a novel BRCA1-null ovarian cancer cell line in response to ionizing radiation. Mol Cancer Res 2007;5(1):35-45. 275. Hirano S, Ito N, Takahashi S, Tamaya T. Clinical implications of insulin-like growth factors through the presence of their binding proteins and receptors expressed in gynecological cancers. Eur J Gynaecol Oncol 2004;25(2):187-91. 276. Terry KL, Tworoger SS, Gates MA, Cramer DW, Hankinson SE. Common genetic variation in IGF1, IGFBP1 and IGFBP3 and ovarian cancer risk. Carcinogenesis 2009;30(12):2042-6. 277. Carreiras F, Denoux Y, Staedel C, Lehmann M, Sichel F, Gauduchon P. Expression and localization of alpha v integrins and their ligand vitronectin in normal ovarian epithelium and in ovarian carcinoma. Gynecol Oncol 1996;62(2):260-7.