UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Understanding the genetic evolution of oral cancer genomes Tsui, Ivy F.L. 2010

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

Item Metadata

Download

Media
24-ubc_2010_fall_tsui_ivy.pdf [ 2.68MB ]
Metadata
JSON: 24-1.0071054.json
JSON-LD: 24-1.0071054-ld.json
RDF/XML (Pretty): 24-1.0071054-rdf.xml
RDF/JSON: 24-1.0071054-rdf.json
Turtle: 24-1.0071054-turtle.txt
N-Triples: 24-1.0071054-rdf-ntriples.txt
Original Record: 24-1.0071054-source.json
Full Text
24-1.0071054-fulltext.txt
Citation
24-1.0071054.ris

Full Text

 UNDERSTANDING THE GENETIC EVOLUTION OF ORAL CANCER GENOMES  by  Ivy F.L. Tsui  B.Sc. (Hons), The University of British Columbia, 2005  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in The Faculty of Graduate Studies (Pathology)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) July 2010   © Ivy F.L. Tsui, 2010   ii     Abstract Background: Oral cancer is the most common type of head and neck cancer, with a 5- year survival rate of < 50%.  The major problems of oral cancer include the late stage of disease at the time of diagnosis, a lack of effective targeted therapies, and failure in surgical treatment.  A better understanding of the genetic evolution of oral cancer will greatly benefit the identification of molecular targets for the prevention and treatment of the disease.  I applied whole genome profiling to delineate genes and pathways associated with oral carcinogenesis and to uncover the clonal relationships between samples from the same field. Hypotheses: (i) Genetic alterations critical to oral cancer development will be present in oral premalignant lesions (OPLs), and candidate genes will reside within recurrent regions of alterations in OPLs.  (ii) DNA amplification occurs at the premalignant stage of oral cancer development and harbours genes involved in key oncogenic pathways. (iii) The oral cancer field is genetically heterogeneous, and the clonal evolution between biopsies from the field will be revealed by their genetic signatures. Materials/Methods: DNA copy number data from OPLs were generated using tiling- path DNA microarrays.  Recurrent regions of alterations on chromosome 3p were identified and associated with clinical information.  Copy number data were associated with public expression datasets to identify genes and pathways contributing to oral carcinogenesis.  Genetic breakpoints were used to evaluate the clonal expansion of genetically altered cells within a single field. Results:  I identified the genetic alterations that are recurrently altered in different individuals, supporting hypothesis 1.  Regions of  high-level amplification are frequently detected in OPLs, and genes mapped in amplicons are significantly enriched in the FGF signaling network, supporting hypothesis 2.  Furthermore, the genetic relatedness among biopsies from an oral cancer field is revealed by comparing their genetic signatures, and the field is found to be genetically heterogeneous, supporting hypothesis 3. iii    Conclusions:  Whole genome profiling of OPLs allows the detection of novel genetic changes and fine-maps the genetic breakpoints previously not known.  Together, this work highlights the importance of tailored targeted therapy to effectively treat different patients and different subclones of a field.  iv    Table of contents Abstract.............................................................................................................................ii Table of contents..............................................................................................................iv List of tables.....................................................................................................................ix List of figures ....................................................................................................................x Acknowledgements.........................................................................................................xii Dedication......................................................................................................................xiv Co-authorship statement.................................................................................................xv Chapter 1. Introduction to oral cancer..........................................................................1 1.1. Oral cancer............................................................................................................2 1.2. Clinical subtypes of oral premalignant lesion.........................................................5 1.2.1. Leukoplakia.................................................................................................5 1.2.2. Erythroplakia ...............................................................................................5 1.2.3. Progression risks of leukoplakia and erythroplakia.....................................5 1.2.4. Histology to identify high-risk lesions..........................................................6 1.3. Molecular biology of oral cancer............................................................................7 1.3.1. Genetic progression model of oral cancer..................................................7 1.3.2. Biology of HPV-positive cancer...................................................................8 1.4. Field cancerization.................................................................................................8 1.4.1. Clinical problems of local recurrence and SPT...........................................8 1.4.2. Field cancerization and clonal evolution theory..........................................9 1.4.3. Genetic analysis........................................................................................10 1.5. Early Detection of oral cancer..............................................................................11 1.5.1. Histopathology to identify high-risk lesions...............................................11 1.5.2. Biomarker as prognostic indicator ............................................................11 v    1.5.3. Visual aids for the detection of high-risk lesions .......................................... 13 1.6. Molecular targeted therapy for oral cancer ..................................................... 14 1.6.1. Treatment strategies for oral cancer ............................................................ 14 1.6.2. Targeted therapy ......................................................................................... 14 1.6.3. Vaccine therapy ........................................................................................... 15 1.7. Genomic technologies .................................................................................... 16 1.7.1. Array comparative genomic hybridization .................................................... 16 1.7.2. Expression profiling ..................................................................................... 18 1.8. Thesis theme and rationale for study .............................................................. 19 1.9. Objectives and hypotheses ............................................................................. 19 1.10. Specific aims and thesis outline ...................................................................... 19 1.11. References ...................................................................................................... 25 Chapter 2. Integrative molecular characterization of head and neck cancer cell model genomes ........................................................................................................... 39 2.1. Introduction ..................................................................................................... 40 2.2. Materials and methods .................................................................................... 41 2.2.1. Cell lines ...................................................................................................... 41 2.2.2. Tiling-path DNA microarray ......................................................................... 42 2.2.3. Imaging and analysis of genomic data ........................................................ 42 2.2.4. Gene expression profiling ............................................................................ 43 2.2.5. Data analysis of expression profiles ............................................................ 43 2.3. Results and discussion ................................................................................... 43 2.4. Conclusions .................................................................................................... 46 2.5. References ...................................................................................................... 69 Chapter 3. Multiple aberrations of chromosome 3p detected in oral premalignant lesions ...........................................................................................................................73 3.1. Introduction ..................................................................................................... 74 vi    3.2. Materials and methods .................................................................................... 75 3.2.1. Tissue samples ........................................................................................... 75 3.2.2. Array CGH analysis ..................................................................................... 75 3.2.3. Copy number variation (CNV) ..................................................................... 76 3.2.4. Statistical analysis ....................................................................................... 77 3.2.5. Loss of heterozygosity (LOH) analysis. ....................................................... 77 3.3. Results and discussion ................................................................................... 77 3.3.1. Segmental alterations on chromosome 3p are more frequent than whole arm loss in HGDs. ......................................................................................................... 78 3.3.2. Regions of loss in HGDs and comparison to OSCCs. ................................. 78 3.3.3. Evidence of segmental alterations in progressing LGDs.. ........................... 79 3.3.4. Disrupted genes in dysplastic lesions. ......................................................... 80 3.3.5. Sequential 3p deletions during OPL development. ...................................... 80 3.3.6. Integrating LOH data with copy number alterations. .................................... 81 3.4. Summary......................................................................................................... 81 3.5. References ...................................................................................................... 87 Chapter 4. Multiple pathways in the FGF signaling network are frequently deregulated by gene amplification in oral dysplasias ............................................. 92 4.1. Introduction ..................................................................................................... 93 4.2. Materials and methods .................................................................................... 95 4.2.1. Tissue samples ........................................................................................... 95 4.2.2. Whole genome DNA microarray analysis .................................................... 96 4.2.3. Data analysis ............................................................................................... 96 4.2.4. Transcript expression analysis .................................................................... 97 4.2.5. Real-time polymerase chain reaction .......................................................... 97 4.2.6. Biological functions and pathway analysis................................................... 98 4.2.7. Fluorescence in situ hybridization (FISH) .................................................... 98 vii    4.3. Results ............................................................................................................ 99 4.3.1. Early occurrence of DNA amplification and homozygous deletion in OPLs. 99 4.3.2. Recurrent amplicons and rare regions of homozygous deletion harbor known and novel candidate cancer genes. ........................................................................ 99 4.3.3. Transcript analysis of independent HNSCC datasets. ............................... 100 4.3.4. Frequent oncogenic activation of a common signaling network in OPLs. .. 101 4.3.5. Validation of mRNA levels in OSCCs. ....................................................... 102 4.3.6. Single cells of oral dysplasia exhibit co-amplification of EGFR and CCND1.  ....................................................................................................................103 4.4. Discussion ..................................................................................................... 103 4.4.1. Amplifier phenotype in oral high-grade dysplasias .................................... 104 4.4.2. Disruption of multiple components of a signaling network in oral dysplasias  ....................................................................................................................105 4.4.3. Pathway addiction in oral dysplasias ......................................................... 106 4.5. Summary....................................................................................................... 107 4.6. References .................................................................................................... 116 Chapter 5. A dynamic oral cancer field --unraveling the underlying biology and its clinical implication .................................................................................................... 125 5.1. Introduction ................................................................................................... 126 5.2. Materials and methods .................................................................................. 128 5.2.1. Case presentation ..................................................................................... 128 5.2.2. Tissue samples ......................................................................................... 129 5.2.3. Whole genome tiling-path array ................................................................. 129 5.2.4. Loss of heterozygosity (LOH) .................................................................... 130 5.3. Results .......................................................................................................... 130 5.3.1. Heterogeneity across an optically altered field. ......................................... 130 5.3.2. Defining clonal origin among biopsies in an oral cancer field using whole genome breakpoint detection. .............................................................................. 131 viii    5.3.3. LOH results suggests a common progenitor among biopsies #1, 2, and 3.  ....................................................................................................................132 5.4. Discussion ..................................................................................................... 133 5.5. References .................................................................................................... 139 Chapter 6. Discussion and conclusions ................................................................. 145 6.1. Research summary ....................................................................................... 146 6.1.1. Providing a comprehensive resource for the oral cancer community ........ 147 6.1.2. Identifying recurrent regions of alterations in oral premalignant lesions and oral cancer ........................................................................................................... 148 6.1.3. Delineating the key oncogenic pathways for oral cancer development ..... 149 6.1.4. Understanding the clonal relationships of field effect in oral cancer .......... 151 6.2. Discussions ................................................................................................... 152 6.2.1. Biological relevance of the identified genetic alteration and candidate genes  ...................................................................................................................152 6.2.2. Clonal evolution for the development of multiple oral lesions .................... 155 6.3. Significance ................................................................................................... 156 6.3.1. Genetic progression model of oral cancer ................................................. 156 6.3.2. Development of biomarker tools for evaluating progression risks of oral premalignant lesions ............................................................................................ 158 6.4. Future Directions ........................................................................................... 158 6.4.1. Functional studies of candidate genes ...................................................... 158 6.4.2. Biomarker tools for evaluating progression risks ....................................... 159 6.4.3. Molecular characterization of second primary tumours ............................. 160 6.5. References .................................................................................................... 163 Appendices .............................................................................................................. 168 Appendix A - Supplementary data for chapter 2 .................................................. 169 Appendix B - Supplementary data for chapter 3 .................................................. 207 Appendix C - Supplementary data for chapter 4 .................................................. 234 ix      List of tables Table 2.1. Copy  number status of cancer genes in the six head and neck cancer cell lines................................................................................................................................50 Table 3.1. Summary of recurrent minimal altered regions identified in 47 high-grade dysplasias using 3p tiling-path array CGH......................................................................85 Table 3.2. Pattern of 3p alterations in oral dysplasias and OSCCs................................86 Table 4.1.  Regions of homozygous deletion................................................................112 Table 4.2.  Recurrent regions of gene amplification in OPLs.  Minimal altered regions recurrent in at least two high-risk OPLs are listed........................................................113 Table 4.3.  Cancer-related genes mapped within recurrent regions of amplicon in high- risk OPLs......................................................................................................................114 Table 4.4.  Disruption of canonical signaling pathways in oral premalignant lesions...115  x     List of figures Figure 1.1.  Histological progression model of oral cancer development.......................22 Figure 1.2. Molecular differences for the development of true second primary tumour, local recurrence, or second field tumour.  ......................................................................23 Figure 1.3. Tiling-path array comparative genomic hybridization (CGH)........................24 Figure 2.1  Summary of copy number alteration in each cell line...................................47 Figure 2.2.  Multiple levels of segmental copy number alterations are detected in six cell lines on chromosome 11q...............................................................................................48 Figure 2.3.  Integrative analysis of genetic and expression levels of genes within 11p13- p12 amplicon .................................................................................................................49 Figure 3.1  Frequency of alterations on the 3p arm. . ....................................................83 Figure 3.2.  Genetic changes observed in each histological group on the 3p arm.........84 Figure 4.1  Whole genome tiling-path array profile of a carcinoma in situ (CIS) Oral42...........................................................................................................................108 Figure 4.2.  Graphical representations of region of homozygous deletion in oral lesions...........................................................................................................................109 Figure 4.3.  Multiple disruptions in a single network driven by the mechanism of gene amplification..................................................................................................................110 Figure 4.4.  Co-amplification of EGFR and CCND1 in high-grade dysplasia Oral22....111 Figure 5.1  Clinical characterization and corresponding histology within an oral cancer field at right lateral tongue of a 52-year-old male former smoker.................................136 Figure 5.2.  Summary of genetic alterations in different cell populations from various areas within an optically altered oral cancer field.........................................................137 xi    Figure 5.3.  Chromosome 9p genetic profiles of samples from areas #1 (SCC), #2 (dysplasia), #3 (SCC), and #4 (10 mm beyond FV boundary, no dysplasia)................138 Figure 6.1  Genetic progression model for the development of oral cancer.................162 xii     Acknowledgements I would like to acknowledge the many people who supported this work, in particular the co-authors of each chapter of this thesis and the members of the Lam Lab who supported me throughout my projects.  I would also like to thank financial support from the Roman M. Babicki Fellowship, Canadian Institutes of Health Research (CIHR), and Michael Smith Foundation for Health Research (MSFHR) during my research.  Special acknowledgements from the published versions of each chapter are included below: Chapter 2: This work was supported by grants from the National Institute of Dental and Craniofacial Research (NIDCR)/National Institutes of Health (R01 DE15965), Canadian Institutes of Health Research, and funds from the Pacific Otolaryngology Foundation. IFLT is supported by scholarships from the Canadian Institutes of Health Research (CIHR) and the Michael Smith Foundation for Health Research (MSFHR).  We would like to acknowledge Drs Wan Lam and Timon Buys for helpful discussion. Chapter 3: This work was supported by grants from the NIDCR R01-DE15965 and R01- DE13124; Genome Canada; the Canadian Institutes of Health Research (CIHR); Scholarships to IFLT from CIHR and Michael Smith Foundation of Health Research. We would like to thank Cindy Cui for providing clinical information, Paul Boutros, Bradley Coe, Raj Chari, Timon Buys, and Cathie Garnis for critical discussion. Chapter 4: This work was supported by funds from the NIDCR grants R01DE15965 and R01DE13124; Genome Canada; Canadian Institutes of Health Research (CIHR); Scholarships to IFLT and CFP from CIHR and Michael Smith Foundation for Health Research.  We thank Yuqi Zhu, Raj Chari, Bradley Coe, William Lockwood, Timon Buys, and Paul Boutros for advice and discussion Chapter 5: This work was supported by grants from the National Institute of Dental and Craniofacial Research/National Institutes of Health (R01 DE17013), Canadian Institutes of Health Research (MOP-77663), and the Canadian Cancer Society (CCS-20336), and funds from the Pacific Otolaryngology Foundation. The author would like to xiii    acknowledge Dr. Wan L. Lam for helpful discussion.  IFLT is supported by scholarships from the Canadian Institutes of Health Research (CIHR) and the Michael Smith Foundation for Health Research (MSFHR).  CFP is supported by a Clinician Scientist Award from the CIHR and a Scholar Award from the MSFHR. xiv      Dedication      To my family.  xv    Co-authorship statement Chapters 2 to 5 were co-authored as manuscripts for publication.  The following author lists apply for each chapter: Chapter 2: Tsui IFL and Garnis C. (2009) Integrative molecular characterization of head and neck cancer cell model genomes. Head Neck. [Published ahead-of-print 15 December 2009; DOI: 10.1002/hed.21311] Contribution: I performed the experiments, developed the data analysis methods, interpreted the results, wrote the manuscript, and made all figures and tables.  My supervisory committee member Dr. Garnis guided me throughout this study, and my supervisor Dr. Lam provided helpful discussion throughout the process. Chapter 3: Tsui IFL, Rosin MP, Zhang L, Ng RL, Lam WL. (2008)  Multiple aberrations of chromosome 3p detected in oral premalignant lesions.  Cancer Prevention Research. 1: 424-9. Contribution: I conceived the study, performed the experiments, performed all data analyses, interpreted the results, and wrote the manuscript for this work. Chapter 4: Tsui IFL, Poh CF, Garnis C, Rosin MP, Zhang L, Lam WL. (2009) Multiple pathways in the FGF signaling network are frequently deregulated by gene amplification in oral dysplasias. International Journal of Cancer. 125: 2219-28 Contribution: I developed the concept of the study, performed all data analyses, made all figures and tables, and wrote the manuscript for this work. Chapter 5: Tsui IFL, Garnis C, Poh CF. (2009) A dynamic oral cancer field --unraveling the underlying biology and its clinical implication. American Journal of Surgical Pathology. 33: 1732-8 Contribution: As in Chapter 2, I performed the experiments, analyzed the data, interpreted the results, and wrote the manuscript.  My supervisory committee member Dr. Garnis was instrumental in developing the experimental design.  Dr. Poh, an oral xvi    pathologist, generated microdissected samples for the study and provided insight into clinical implication.  My supervisor Dr. Lam guided me through the process.  1   Chapter 1: Introduction to oral cancer 2    1. Introduction to oral cancer 1.1. Oral cancer 1.1.1.  Definition and epidemiology of oral cancer Oral cancer is the most common head and neck cancer, comprising 77% of all cases (Piccirillo et al. 2007).  Oral cancer include cancers of the lip, oral cavity, oropharynx, tonsil, and salivary glands, with tongue cancer representing ~20% of all head and neck cancer cases (Piccirillo et al. 2007). In Canada last year, 3,400 new cases of oral cancer were diagnosed and 1,150 people died from this disease, representing 2.5% of all new cases and 2% of all cancer deaths (Canadian Cancer Society 2008).  The five-year survival rate for all stages remains at < 50% for all sites, while stage III and stage IV diseases have survival rates of only 25% (Epstein et al. 2008).  Because more than 40% of oral cancers are diagnosed at late stage, increased detection at early, more treatable stages represents a key approach for improving survival rates (Piccirillo et al. 2007).   In contrast, oropharyngeal cancer patients have shown significant improvements in five-year survival rates.  Incidence of oropharyngeal cancer has been on the rise since the early 1970s and has increased particularly in younger people (< 45 years old).  India has always had the highest incidence of oral cancer and incidences in Europe are rising (Sturgis & Cinciripini 2007). Interestingly, while Canadian men in general have a doubled risk of developing oral cancer of any stage, there is parity for incidence between the sexes when patients are < 40 years old (Laronde et al. 2008).  The change in these epidemiological trends has been attributed to changing human sexual behaviours and a reduction in environmental tobacco smokes. 1.1.2. Risk factors Oral cancer is a genetic disease.  Identifying the causative factors that contribute to oral cancer development will improve the prevention of this disease and will provide further understanding of disease mechanisms.  The strongest etiological factors for oral 3   carcinogenesis is tobacco smoke (Brennan et al. 1995; Zhang et al. 2000).  The intensity and duration of smoking is positively correlated with cancer risk and smoking cessation reduces the risk of oral cancer and premalignant lesions (Pelucchi et al. 2008).  Exposure to environmental tobacco smoke is also generally accepted as a cause of oral cancer in never smokers (Zhang et al. 2000).  The carcinogens in tobacco smoke, including polycyclic aromatic hydrocarbons (e.g. benzo[a]pyrene, aromatic amines, aldehydes) and tobacco-specific nitrosamine, are genotoxic and could cause DNA adducts formation and elicit mutations (Brennan et al. 1995).  A comparison of head and neck cancers in smokers and non-smokers demonstrated differences in the prevalence and the mutation status of specific genes.  For example, TP53 mutations in head and neck tumors occur more frequently in smokers than in non-smokers (Brennan et al. 1995).  Such mutations were also found to be more frequent among patients who smoked tobacco and drank alcohol, suggesting a synergistic effect between these agents (Brennan et al. 1995). Despite its harmful synergy with tobacco smoke, alcohol consumption could also be an independent risk factor for oral cancer.  The metabolite of alcohol, acetaldehyde, is genotoxic and thus has been associated with cancer risk (Homann et al. 1997b). Acetaldehyde is produced from ethanol in the epithelia by alcohol dehydrogenase (ADH) and acetaldehyde is usually quickly metabolized to acetate by aldehyde dehydrogenase (ALDH) (Homann et al. 2000).  Microflora in the oral cavity could also oxidize ethanol to product acetaldehyde that remains in  saliva for a long time (Homann et al. 1997a).  Although some studies suggest different polymorphisms in ADH and ALDH genes could increase oral cancer risk, having a fast metabolizing ADH or nonfunctional ALDH have not been consistently associated with oral cancer risk (Brennan et al. 2004).  Some researchers have suggested the possibility of alcohol serving as a chemical solvent that therefore increases exposure and absorption of cigarette smoke.  Poor dental health with bacterial overgrowth is also an independent risk factor of oral cancer (Maier et al. 1993).  Family history and preexisting medical conditions could also contribute to oral cancer development. 4   Recently, human papillomavirus (HPV) infection has been shown to cause oral cancer and found to be prevalent in nonsmoker and nondrinker and younger patients.  High-risk HPV, particularly HPV16, has been found in up to 70% of oropharyngeal cancer (Begum et al. 2005; Gillison et al. 2000; Kreimer et al. 2005).  Certain sexual behaviours, number of sexual partners, repeated viral exposure, infrequent use of barriers during vaginal or oral sex, early onset of sexual behaviour, and a history of genital warts have been associated with HPV-associated oral cancer (Curado & Hashibe 2009; Roberts et al. 2007).  Oral HPV is also found to persist more frequently in HIV (human immunodeficiency virus)-seropositive individuals (Kreimer et al. 2004) and marijuana smoke may also be relevant for HPV-positive oral cancer as it might suppress the immune system efforts to clear viral pathogens (Gillison et al. 2008). HPV-positive oropharyngeal cancers may represent a distinct disease entity that has a markedly improved prognosis and is causally associated with HPV-directed mechanism of cancer development. 1.1.3. Anatomy of the oral cavity The cell type of origin for the majority of oral cancer is the squamous cell, while salivary gland cancer is often of a mixed cell type (Stenner & Klussmann 2009).  The heterogeneity and distinct clinical presentations of oral cancer are a product of the complex anatomy of the oral cavity.  This cavity is lined by a stratified squamous epithelium, which is composed of flat, scale-like cells arranged in multiple layers on a basement membrane (Allen & Cameron 2004).  Varying thickness and keratinization are present in different regions of the oral cavity.  Typically, 60% of oral lining is composed of non-keratinized stratified squamous epithelium (called lining mucosa) that covers the buccal area, floor of the mouth, and the ventral tongue (Allen & Cameron 2004).  This lining forms an elastic surface capable of stretching and movement.  The keratinized area is called masticatory mucosa and this covers regions that are subject to mechanical forces, including the gingival and hard palate.  This mucosa is also lightly attached to underlying structures by a collagen connective tissue.  Specialized mucosa include both masticatory and lining mucosa, and also taste buds that are located on the 5   dorsum of the tongue.  The oral mucosa is lined by a continuous basement membrane, which serves as the first line of defense against stromal invasion. 1.2.  Clinical subtypes of oral premalignant lesion 1.2.1. Leukoplakia Leukoplakia is defined by World Health Organization (WHO) classification as a predominantly white patch lesion of the oral mucosa that cannot be rubbed off and cannot be characterized as any other specific disease (Hunter et al. 2005; Mithani et al. 2007).  It is the most commonly detected premalignant lesion of the oral cavity.  When histological examination is performed, only a subset of leukoplakias will display dysplastic features and only 36% of those dysplasias will subsequently develop into OSCC (Mithani et al. 2007).  Overall, only about 8% of leukoplakias will progress to OSCC and about 17%-25% contain dysplasia (Hunter et al. 2005). 1.2.2. Erythroplakia Erythroplakia is a premalignant lesion of the oral mucosa that typically presents as a discrete, velvety red  plaque that is < 1.5 cm in diameters and is not diagnosable clinically as other disease (Reichart & Philipsen 2005).  It is rare and associated with a high risk of cancer development.  In fact, most erythroplakias show dysplastic features, with 51% diagnosed as invasive SCC, 40% is carcinoma in situ (CIS) and 9% as mild or moderate dysplasia (Reichart & Philipsen 2005).  It is often found in the floor of the mouth, the soft palate, or the buccal mucosa (Reichart & Philipsen 2005). 1.2.3. Progression risks of leukoplakia and erythroplakia The clinical appearance of erythroplakia is associated with a much higher chance of developing OSCC than leukoplakias.  Erythroplakias have > 50% risk of developing into OSCC, while leukoplakias have only a 2-5% risk for developing into cancer over 10 years (Hunter et al. 2005).  However, because erythroplakias are very rare, this clinical presentation alone does not help for predicting progression for the vast majority of lesions. 6   1.2.4. Histology to identify high-risk lesions The current gold standard for judging the risk of progression in a given leukoplakia is histological examination (Hunter et al. 2005).  The development of OSCC follows a regular sequence of histological stages: from hyperplasia, through mild, moderate, and severe dysplasia, and then carcinoma in situ (CIS) (Fig. 1.1).  By definition, dysplasia is characterized by architectural and cytological changes in the epithelium, including irregular epithelial stratification, loss of basal cell polarity, drop-shaped rete ridges, increased number of abnormal mitotic figures, premature keratinization in single cells (dyskeratosis), enlarged keratinocytes within rete ridges, abnormal variation in nuclear size (anisonucleosis), abnormal variation in nuclear shape (nuclear pleomorphism), abnormal variation in cell size (anisocytosis), cellular pleomorphism, increased nuclear- cytoplasmic ratio, increased nuclear size, atypical mitotic figures, increased number and size of nucleoli.  The extent of the dysplastic changes determine the grading of the dysplasia.  Atypia confined to the basilar and parabasilar keratinocytes is mild dysplasia, and atypia extending to the midspinous layer is moderate dysplasia.  High- grade dysplasias (including severe dysplasia and CIS) have atypia that extend to the surface layer and have an intact basement membrane with no risk for regional or distant metastasis.  Historical data have shown that high-grade dysplasias have a much higher chance of malignant transformation (Crissman & Zarbo 1989; Fresko & Lazarus 1981; Hayward & Regezi 1977; Poh et al. 2008; Summerlin 1996).  A longitudinal study in British Columbia has also examined if patients with high-grade dysplasias will develop to cancer if they were left untreated.  It was found that the percentage of high-grade dysplasias that develop into cancer increase with time, with 42% of high-grade dysplasias developing into tumours in 2 years, 56% in 3 years, and 70% in 5 years of follow-up time.  Thus, high-grade dysplasias are surgically treated in British Columbia to prevent further development into invasive OSCC (Poh et al. 2008).  In contrast, very few low-grade dysplasias will progress to invasive disease; depending on the sample selection and follow-up time, 5-16% of these low-grade lesions will progress (Mincer et al. 1972; Schepman et al. 1998; Waldron & Shafer 1975). 7   1.3. Molecular biology of oral cancer 1.3.1. Genetic progression model of oral cancer The evolution of OSCC is known to result from the acquisition of multiple genetic events targeting different genes and molecular pathways (Mao et al. 2004).  Genomic instability is a hallmark of human cancers.  Genomic instability increases progressively from hyperplasia through various stages of dysplasia to invasive carcinoma (Hunter et al. 2005) (Fig. 1.1).  Although specific genetic events have been found to accumulate in sequential order to develop cancer, it is generally accepted that it is the accumulation of genomic instability that determines oral carcinogenesis rather than the sequence itself (Garnis et al. 2009).  Previous studies have employed loss of heterozygosity (LOH) assays on specific chromosome arms to delineate a genetic progression model for OSCC based on the frequency of LOH in different histological stages.  Loss at chromosome 9p21 and 3p are among the earliest detectable events in squamous hyperplasia (Mao et al. 1996).  Loss of 9p21 (CDKN2A) is detected in ~20% of hyperplasias and ~50% of dysplasias, flagging this event as one of the earliest changes in cancer development.  CDKN2A encodes two transcripts - p16INK4A and p14ARF - which are responsible for G1 cell cycle regulation and MDM2-mediated degradation of p53 respectively.  P16 is also found to be inactivated in OSCC through homozygous deletion, promoter methylation, and less commonly by point mutations (Reed et al. 1996).  Loss of 3p has also been shown as an early genetic event in ~16% of hyperplasias and ~50% of dysplasias (Califano et al. 1996).  The 3p14 locus includes a gene encoding the fragile histidine triad gene (FHIT), which has previously been described as a tumour suppressor gene (Lee et al. 2001).  Loss of chromosome 3p is also a common genetic alteration in many different cancers, and several tumour suppressor genes are found in this region (Zabarovsky et al. 2002).  LOH at 17p13 is also detected in ~10% of hyperplasias and ~30% of dysplasias, while changes on chromosomes 4q, 6p, 8, 11q, 13q, and 14q are usually detected  in CIS lesions and invasive SCC as late-stage events (Rosin et al. 2000; Califano et al. 1996; Garnis et al. 2004a).  More recent array comparative genomic hybridization (CGH) analysis using 8   oral cancer samples has fine-mapped precise genetic alterations and identified numerous genetic alterations not previously found by LOH assays, since genomic instability is detected across the whole genome and found to be a common characteristics of OSCC (Garnis et al. 2009; Tsui et al. 2009; Tsui et al. 2008; Baldwin et al. 2005). 1.3.2. Biology of HPV-positive cancer The role of HPV in oral carcinogenesis as a separate etiological factor is now recognized, as molecular and survival differences are found in patients with HPV- infected OSCC and those without HPV (Gillison et al. 2008; Smeets et al. 2006). Briefly, HPV is a DNA virus with a genome of approximately 8 kbp (zur Hausen 2002). More than 100 subtypes have been identified, with some types being classified as high- risk types because of its role to cause cervical cancer development.  High-risk types of HPV encode two viral oncoproteins called E6 and E7 (Gillison et al. 2000; zur Hausen 2002).  These proteins inactivate products of the TP53 and Rb tumour suppressor genes, respectively.  This enables entry to cell cycle and DNA synthesis needed for viral replication.  OSCCs associated with HPV are typically infected with the high-risk type (HPV16).  Usually, this type of OSCC does not harbour TP53 mutations and it has a very low number of genetic alterations.  HPV-negative OSCCs, on the other hand, have TP53 mutations at a rate of >70%.  Thus, the viral oncoproteins cause the disruption of Rb and p53 pathways in HPV-containing OSCCs. 1.4. Field cancerization 1.4.1. Clinical problems of local recurrence and SPT The current gold standard for selecting treatments is histological assessment of OSCC and its margins (Poh et al. 2008).  As described above (Section 1.2.4), histological stage is scored according to standard criteria from the World Health Organization. Lesions may be called hyperplasia or mild, moderate, or severe dysplasia or CIS or invasive OSCC.  Severe dysplasias and CIS are associated with the strongest risk for progression to invasive disease (Poh et al. 2008).  Local recurrence occurs in up to 5% 9   of all cases treated with surgery despite histological clear resection margins (Tan et al. 2008).  Local recurrence is generally believed to arise from the same precursor cell as the initial tumour and is clinically defined as any oral tumour occurring within three years at a distance of <2 cm from the initial mass (Braakhuis et al. 2005a).  In contrast, second primary tumors (SPTs) are believed to have developed independently (Braakhuis et al. 2005b).  SPT was originally defined in 1932 to delineate independent tumours from metastatic disease (Warren & Gates 1932).  Histological examination can distinguish phenotypes but cannot prove if the lesions are distinct from each other. Additional criteria used to define SPTs include spatial information (a distance of ≥2 cm between two tumours) and temporal information (development of tumours at the same or adjacent sites after ≥3 years) (Braakhuis et al. 2005a).  SPTs can be further divided into two groups: synchronous SPTs (which develop simultaneously with or within six months of the index tumour) and metachronous SPTs (which develop more than six months after the index tumour) (Braakhuis et al. 2005a).  Although SPTs suggest independent lesions by clinical features, recent genetic studies have shown cases where the first and second tumours shared common genetic alterations, suggesting that suspect SPTs originated from the same precursor cell (Jang et al. 2001; Tabor et al. 2004; Tabor et al. 2002).  These data support the idea that a contiguous field of altered tissue can exist and give rise to seemingly unrelated tumours (this is discussed in further detail below).  This possibility poses  significant treatment challenges, since the entire field of altered tissue - most or all of which can appear histological normal - possesses a risk for transformation to disease. 1.4.2. Field cancerization and clonal evolution theory Two concepts have significantly impacted the understanding of local recurrence development and SPTs: field cancerization and micrometastases (Fig. 1.2) (Slaughter et al. 1953; Bedi et al. 1996).  The concept of field cancerization was first proposed by Slaughter et al. based on histological examinations of 783 oral cancer patients (Slaughter et al. 1953).  The following observations were made: (a) oral cancer develops in multifocal areas of premalignant change, (b) histological abnormal tissue 10   surrounds oral tumours, (c) oral cancer often consists of multiple independent lesions that sometimes coalesce, and (d) the persistence of abnormal tissue after surgery may explain SPTs and local recurrences (Slaughter et al. 1953).  The field cancerization theory claims that carcinogenic exposures lead to independent transformation of multiple epithelial cells at several sites, resulting in the induction of multiple genetically unrelated tumours.  An alternative theory (micrometastases) is based on the premise that the transforming event is rare and that a single altered cell transformed to a contiguous “field” by clonal expansion and gradual replacement of normal mucosa (Braakhuis et al. 2005a).  After the accumulation of additional genetic alterations, spatially distinct but clonally-related tumours could develop (Bedi et al. 1996).  Various mechanisms for disease spread have been proposed, including shedding of the initially transformed cells into the saliva and implantation at other sites and lateral intraepithelial migration of preneoplastic cells have been proposed (Braakhuis et al. 2005b). 1.4.3. Genetic analysis Molecular analyses have been performed to investigate whether tumour-adjacent “macroscopically normal” tissue and surgical margins are genetically aberrant - and whether multiple carcinomas shared a common clonal origin (Escher et al. 2008; Shin et al. 1996; Tabor et al. 2004; van Houten et al. 2004; Braakhuis et al. 2003; Tabor et al. 2001).  By measuring LOH at selected markers and TP53 mutation, at least one third of oral tumours had genetic alterations in the tumour and the macroscopically normal tissue adjacent to the tumour.  This suggests that genetic alterations can arise before presentation of abnormal phenotypes (Tabor et al. 2001).  However, only a limited region of macroscopically normal mucosa were sampled and only a limited number of genetic markers were examined, meaning that the true frequency of genetic alteration or the true incidence of genetically altered tissue fields may be underestimated. Patients presenting with multiple carcinomas have also been compared based on genetic alterations.  These molecular analyses resulted in two types of SPTs: second field tumours (SFTs) that originate from a single clone and “true” SPTs that have independent origins and different genetic alterations (Braakhuis et al. 2005a). 11   1.5. Early Detection of oral cancer 1.5.1.  Histopathology to identify high-risk lesions Being able to identify those premalignant lesions that will progress to later stage disease would have significant utility for prioritizing treatment.  Histological criteria based on the presence and degree of dysplasia is the gold standard for judging risk of progression in oral premalignant lesions.  Other factors include clinical appearance, size of the lesion, and the site of the lesion are also involved.  Anatomic sites including the floor of mouth and the ventral lateral tongue are have an elevated risk of cancer development and persistent erythroplakia is also known to be at higher risk compared to leukoplakia.  As previously stated in section 1.2.4, high-grade dysplasias, including severe dysplasia and CIS, indicate a high risk for developing cancer.  Most low-grade dysplasias, including mild and moderate dysplasia, do not progress to cancer.  In British Columbia, all the high-grade dysplasias are treated aggressively to prevent development to cancer.  However, progression risk in low-grade dysplasias cannot be defined by histological criteria, hence patients with these lesions are managed by a wait-and-watch approach designed to prevent overtreatment for a majority of patients. 1.5.2. Biomarker as prognostic indicator A major area of oral cancer research is focused on the prevention of oral cancer development. This strategy includes intervention at the premalignant stage in order to improve prognosis and survival of the patient.  Genetic alterations or gene expression changes that are significantly associated with progression in low-grade dysplasias would be useful as biomarkers for predicting progression risk.  In a pilot study testing 37 patients on two chromosomal regions, LOH on 3p14 and/or 9p21 have been associated with increased progression risks in oral leukoplakia (Mao et al. 1996).  In another study, 116 patients who had low-grade dysplasias were followed up for >10 years.  Among these patients, 29 developed OSCC at the same anatomical sites as the primary low- grade dysplasia.  Almost all lesions that later progressed to OSCC exhibited LOH at chromosome 3p and/or 9p, confirming previous studies.  Lesions lacking these LOH changes were not seen to progress further.  Of additional importance was the finding 12   that expanding analysis to multiple chromosomal regions improved the ability to predict progression risk. High protein expression of podoplanin, a marker for lymphatic vessels, has also been found in OPLs, with expression detected by immunohistochemistry (IHC) in basal cells and beyond for 72% of OPLs (Kawaguchi et al. 2008).  While histology is important to assess cancer risk, it has been shown that the combination of podoplanin expression and histology could serve as a stronger predictor for oral cancer risk assessment (Kawaguchi et al. 2008).  However, only 23% of the 150 OPLs progressed to invasiveness in this study, thus the diagnostic utility needs to be validated in a larger sample set.  Another study that screened 459 proteins in OPLs identified stratifin, YWHAZ, and hnRNPK as the top three biomarkers predicting progression risk (Ralhan et al. 2009).  This protein expression was subsequently validated by IHC and these markers were further evaluated in 30 OPLs and 21 histological normal oral tissues. However, this study do not provide patient follow-up information, and the histological grade of the dysplasias examined were not determined.  To date, no predictive protein biomarker has been validated in independent studies. Recently, efforts have been made to identify oral cancer biomarkers in bodily fluids such as blood and saliva (Carvalho et al. 2008; Li et al. 2006; Park et al. 2009; Li et al. 2004). Aberrant promoter hypermethylation in salivary rinses has been studied for a set of 21 genes and a methylation-specific PCR assay for a panel of 8 genes was found to distinguish salivary rinses from HSNCC patients and healthy controls (Carvalho et al. 2008).  However, this assay was found to have high specificity but low sensitivity, suggesting that this panel of biomarkers might not be useful for population-based screening.  Instead of evaluating only 21 genes for methylation status, the use of sensitive, high-throughput technology (including CpG island microarrays) may be useful for identifying a panel of biomarkers with high sensitivity and specificity. Another study exploited the serum transcriptomes of oral cancer patients and healthy individuals and identified six biomarkers that have 84% sensitivity and 83% specificity (numbers that do not quite meet clinical screening tool threshold requirements) (Li et al. 13   2006).  Salivary transcriptome profiling has also identified seven mRNA biomarkers with sensitivity and specificity of 91% (Li et al. 2004).  A recent study examined the possibility for saliva microRNAs to distinguish oral cancer patients from healthy individuals (Park et al. 2009).  However, to date none of the studies that exploited bodily fluids have included patients with oral premalignancies, thus the use of bodily fluids for early cancer detection has not been ascertained.  Given improved efficacy, this screening approach must be tested in large populations of individuals with oral premalignancies that have a known progression status. 1.5.3. Visual aids for the detection of high-risk lesions Acknowledging the field effect has significant implications for delineating potential surgical margins, given the importance of this phenomenon to disease recurrences and the formation of second primary tumours.  Although molecular techniques could characterize field changes following acquisition of multiple tumour-adjacent samples, this approach is not rapid enough to be applied during surgical procedures.  Two visual tools that address the need for real-time review of disease fields have been developed for application to OPLs and cancers.  Toluidine blue stain retention and autofluorescence imaging can delineate disease spread in a way not possible under conventional white light imaging (Poh et al. 2006; Zhang et al. 2005). Toluidine blue is a cationic metachromatic dye that selectively binds to free anionic groups such as phosphate groups of DNA, which may be retained in intracellular spaces of dysplastic epithelium (Epstein & Guneri 2009).  A pilot study monitoring 100 patients with OPLs found that toluidine blue retention is more frequently found in lesions of advanced histopathologic stage and in lesions with a high number of genetic alterations (Zhang et al. 2005).  Furthermore, within 15 OPLs that later progressed to invasive disease, 12 retained toluidine blue staining (Zhang et al. 2005). Disease fields for oral lesions and tumours have also been defined through application of  an autofluorescence imaging device.  This tool uses a blue/violet excitation light  to illuminate oral tissue (400-460 nm wavelength).  A pale green autofluorescence will be reflected by normal tissue, while abnormal tissue will absorb the autofluorescence and 14   appear as a dark brown to black region (Lane et al. 2006; Poh et al. 2007; Poh et al. 2006).  Changes in fluorophores represent changes in tissue properties that are associated with cancer progression, including structure and metabolic activity of the tissue area, alterations to epithelial thickness, nuclear morphology, and vascularization that affect absorption and scattering of light (Poh et al. 2006).  Recently, this autofluorescence device has been used to guide sampling for molecular analysis. Whole genome analysis of tissues identified as diseased by this device have been seen to harbour genetic alterations, demonstrating the transformation risk in tissue fields that can appear histological normal under white light (please refer to Chapter 5). 1.6. Molecular targeted therapy for oral cancer 1.6.1. Treatment strategies for oral cancer An accurate diagnosis is necessary to provide oral cancer patients with proper treatment.  First, when a suspicious lesion is identified, a biopsy will be obtained and histological examination will be performed.  In British Columbia, if the lesion is identified to be a high grade dysplasia or cancer, the standard approach is surgical resection (Poh et al. 2008).  Low grade dysplasias, on the other hand, are followed-up for any clinical changes.  Radiotherapy is also used in combination with surgical excision for more aggressive tumours.  For late stage inoperable oral cancer, combined chemotherapy (bleomycin, 5-fluorouracil, or methotrexate) and radiotherapy are administered for improving local-regional control and relapse-free survival. 1.6.2. Targeted therapy Novel therapies targeting specific components of  molecular pathways disrupted in oral cancer are being developed.  One of the targets is EGFR, a transmembrane tyrosine kinase receptor that transduces multiple signaling pathways involved in cancer development.  This gene is found to be overexpressed in >70% of OSCCs, an activation that may be associated with increased gene copy number or mutational activation. Alternatively, EGFR can be activated by high expression of its ligands, which induce dimerization of EGFR, autophosphorylation of its intracellular kinase domain, and 15   activation of multiple oncogenic pathways (Tsui et al. 2007; Kalyankrishna & Grandis 2006).  Monoclonal antibodies (e.g., cetuximab) and tyrosine kinase inhibitors (e.g., gefitinib) have been developed to inhibit the function of EGFR.  While monoclonal antibodies block ligand binding, tyrosine kinase inhibitors inhibit phosphorylation of EGFR.  Other targeted treatment strategies, including sequence-specific antisense oligodeoxynucleotides and small interfering RNAs, have also been developed (Kalyankrishna & Grandis 2006). Treating patients with head and neck SCC by blocking EGFR has only yielded limited success (Kalyankrishna & Grandis 2006).  Signaling pathways independently of EGFR or up-regulation of factors downstream of EGFR allow cancer cells to circumvent these treatments.  Inhibitors targeting multiple genes and pathways have also been developed.  Sorafenib (Nexavar; Onyx Pharmaceuticals, Emeryville, CA) is a multikinase inhibitor that targets the serine/threonine kinases C-Raf, B-Raf and VEGFR (vascular endothelial growth factor receptor)-2, VEGFR-3, platelet-derived growth factor receptor, FLT3, and c-kit.  Thus this drug targets EGFR-Ras-Raf-MEK-ERK signaling and the VEGF-VEGFR pathway, which are key regulators of cell proliferation and angiogenesis, respectively (Elser et al. 2007).  However, the efficacy of this single agent in patients with recurrent and/or metastatic head and neck cancers remains modest, thus a combination with other agents or radiation might be needed to improve its efficacy (Elser et al. 2007).  Nevertheless, the disruption of multiple pathways that could lead to cancer development are important to completely eradicate cancer cells. 1.6.3.  Vaccine therapy As HPV-16 is causal for oral cancer development, vaccines designed to induce virus- specific immune responses have been proposed to prevent HPV infection.  If vaccinated, the immune system will produce high titers of neutralizing antibody to inhibit HPV cell binding and cell entry.  However, vaccination against HPV-16 and 18 does not increase the rate of viral clearance in patients already infected with the virus (Hildesheim et al. 2007).  Thus, it is suggested that women should be vaccinated before sexual intercourse for the first time (Kahn 2009). 16   1.7.  Genomic technologies 1.7.1. Array comparative genomic hybridization Whole genome analysis represents the most effective means of determining the genetic alterations present in a given lesion or tumour sample.  Since the development of comparative genomic hybridization (CGH) in 1992, this technology has been widely used for the analysis of cancer genomes and constitutional chromosomal aberrations (Fig. 1.3).  The principal behind CGH involves differentially label the DNA from a test sample and a reference sample and hybridizing them to a representation of the genome, which was originally a metaphase chromosome spread.  Hybridization of repetitive sequences is blocked by C0t-1 DNA and fluorescence signal intensity ratios are used to determine the relative copy number of the test genome compared to a normal diploid reference genome.  Since the developmental of metaphase-based CGH (conventional CGH), unbalanced chromosome aberrations in oral cancer have been identified, with copy number gain on chromosome 3q and 8q being the most frequent increases in genome content and deletion of 3p, 4, and 18q being the most frequent losses (Bockmuhl et al. 1998; Noutomi et al. 2006; Pathare et al. 2009; Squire et al. 2002; Huang et al. 2002). Although conventional CGH is useful to identify chromosomal aberrations, genetic breakpoints cannot be fine-mapped by this approach and input DNA requirements can be prohibitively high (particularly when it comes to small OPLs that will only yield small amounts of DNA).  Array CGH, which uses elements of oligonucleotides or large insert clones to represent segments on the human genome has improved the resolution of partial or whole genome analyses.  Initial array CGH experiments were performed using cDNA microarrays, which limit itself to coding regions and lack introns (Jarvinen et al. 2006; Jarvinen et al. 2008).  Furthermore, signal-to-noise ratios can be variable and data may be difficult to interpret because of the small and differential size of cDNA probes.  The use of large insert genomic clones, such as bacterial artificial chromosomes (BACs), can provide high signal intensities and can facilitate detection of single copy number changes.  Array CGH with BAC clones has also been useful for 17   analyzing premalignant lesions or archived formalin-fixed paraffin-embedded tissues because low quantity and low quality DNA can be readily profiled.  In summary, BAC arrays require low amount of DNA without the need to introduce DNA amplification, are able to tolerate poor DNA quality, and are sensitive at detecting single-copy changes. Region- or disease-specific BAC arrays have been developed to examine specific chromosomal regions in high resolution (Freier et al. 2007; Garnis et al. 2004b; Garnis et al. 2004c; Garnis et al. 2004d).  For example, the construction and application of a tiled clone array spanning chromosome 8q identified two novel regions of amplification at 8q22 in oral dysplasias, both of which were distinct from a neighboring amplicon containing the well-characterized MYC oncogene (Garnis et al. 2004c).  This region would not have been detected by conventional CGH and demonstrates the utility of high resolution analysis.  In addition, a region-specific BAC array has been developed to examine nine chromosomal regions associated with oral cancer progression, including 3p13-14, 3p24, 4q28, 7p11, 8p23, 9p22, 11q13, 13q21, and 17p13 (Garnis et al. 2004b).  This facilitated gene level analysis for oral lesions and helped identify novel disease-associated molecular changes. CGH arrays comprised of BAC clones spaced across the whole genome have also been applied to discover regions and genes associated with oral cancer, however copy number status for chromosomal segments between array elements could only be inferred (Davies et al. 2005).  A study employed this technology to analyze 89 oral tumours and identified genetic loss on chromosomes 3p, 4, 5q, 8p, 9p, 18 and 21 and genetic gain on chromosome 3q, 8q, 11q, and 20 (Snijders et al. 2005) as being associated with oral cancer, findings that are in agreement with earlier studies by conventional CGH (Huang et al. 2002; Noutomi et al. 2006; Pathare et al. 2009).  This study further focused its analyses on genes within regions of high-level amplification that were <3 Mbp in size, identifying hedgehog and notch signaling pathways as being associated with oral tumourigenesis (Snijders et al. 2005).  Without biasing to specific loci, our laboratory has developed the tiling-path whole genome array CGH to examine the copy number status of the whole genome at a functional resolution of 50 kbp (Coe 18   et al. 2007).  This array distributes its elements uniformly across the whole genome with maximum genomic coverage, providing very robust performance.  Using this technology, novel genetic alterations that have not been previously assayed by CGH have been identified in various stages of oral dysplasias and oral tumours (Baldwin et al. 2005; Garnis et al. 2009; Tsui et al. 2009).  The first application of this array was on 20 oral tumours, and identified numerous novel segmental alterations including amplification at 5p15.2, containing triple functional domain (TRIO) (Baldwin et al. 2005). Array CGH using short (25-85 nt) oligonucleotide elements could also yielded high- resolution copy number measurements across the genome, providing resolution of 24 kb to 500 kb depending on the array type (Coe et al. 2007).  However these arrays often require a high quantity of input DNA, which may limit the use of DNA with only a small amount of materials. 1.7.2. Expression profiling Gene expression microarrays were one of the first high-throughput tools applied for identifying cancer biomarkers with prognostic or predictive value.  Oligonucleotide- and cDNA-based arrays have long been used for parallel analysis of the expression of several genes.  Gene expression analysis of head and neck tumours has yielded insights into the molecular pathways important for the development of this disease. However, the significance of many of these studies is diminished based on the use of small numbers of samples (Banerjee et al. 2005; Gottschlich et al. 2006; Han et al. 2009; Smith et al. 2009; Tomioka et al. 2006; Ziober et al. 2006).  For some of the studies with larger sample sets, hierarchical clustering has identified gene expression patterns associated with clinical phenotypes.  This includes distinguishing oral cancers from lung cancers (Vachani et al. 2007), predicting recurrence-free survival for oral cancers (Chung et al. 2006; Chung et al. 2004), and discriminating head and neck cancers with and without lymph node metastasis (O'Donnell et al. 2005; Roepman et al. 2005).  Additionally, recent studies have compared the expression profiles of OPLs and invasive tumours and found that differences in gene expression occur between early and late stages (Chen et al. 2008; Mendez et al. 2009).  While genome wide expression 19   analysis has furthered our understanding of oral cancer, it is challenging to distinguish driver genes from reactive changes since not all gene expression changes are causal to disease initiation or progression. 1.8.  Thesis theme and rationale for study The theme of this thesis is understanding the genetic alterations of oral cancer development.  Elucidating the genetic mechanisms underlying disease progression will provide insights into oral cancer biology, yield biomarkers for predicting clinical behaviours (e.g. progression risk), and provide candidates for novel targeted interventions. 1.9.  Objectives and hypotheses The main objective of this thesis is to identify genetic alterations and key pathways important for the progression of oral premalignant lesions to invasive lesions. This is based on the following main hypotheses: Hypothesis 1 - Genetic alterations critical to oral cancer development will be present in OPLs and critical candidate genes will reside within recurrent regions of alteration in OPLs. Hypothesis 2 - Multiple components of key pathways involved in oral carcinogenesis will be disrupted in OPLs. Hypothesis 3 - The oral cancer field is genetically heterogeneous.  Genetic signatures and fine-mapped genomic alterations can be used to define clonal relationships between oral lesions from the same patient. 1.10. Specific aims and thesis outline This thesis consists of several manuscripts that have been assembled in an order that best addresses the hypotheses and aims listed above. 20   Aim 1: To identify genomic alterations in six commonly used head and neck cancer cell lines Chapter 2 describes the detailed characterization of DNA copy number alterations in a panel of tumour cell lines and integrates mRNA expression levels to implicate specific genes and chromosomal regions as significant to oral cancer.  HNSCC cell lines  ̶  SCC- 4, SCC-9, SCC-15, SCC-25, A253, and Cal27--are widely used as models of head and neck cancer.  Moreover, cell line materials provide unlimited quantity and good quality DNA and RNA for molecular analysis.  Integrative analyses of whole genome copy number and expression alterations in oral cancer have not been performed before. Integrative analysis of these widely used cell models 1) allowed us to optimize our approaches to molecular data and 2) identified genomic alterations that may be associated with specific oral cancer phenotypes.  Moreover, this study is proof of principle for applying such genomic approaches to identify detailed genetic alterations specific for each sample. Aim 2: To perform copy number analysis on chromosome 3p of oral premalignant lesions Chapter 3 describes the application of  tiling-path array CGH to analyze DNA dosage alterations in oral premalignant lesions.  When this work was started, high-resolution array CGH analysis had never been applied to any premalignant lesion type.  This chapter specifically describes analysis of chromosome 3p in OPLs with known clinical outcome.  Chromosome 3p was chosen for this analysis because it is frequently altered in oral cancers and has been associated with progression risk.  However its alteration status has only been evaluated at a small number of loci in OPLs.  Indeed, we have identified six regions of alteration on chromosome 3p that occurred at a significantly higher frequency in low-grade dysplasias that later progressed to later stage disease. This work represents the first step toward identifying genetic aberrations associated with oral cancer progression, supporting hypothesis 1.  21   Aim 3: To identify regions of high-level gene dosage aberrations in oral premalignant lesions We now know that genetic alterations present in oral dysplasias that progress to later stage disease are not typically detected in dysplasias that do not progress.  The focus of Chapter 4 was analysis of high level amplification and homozygous deletion events across the genomes of OPLs.  The rationale for this analysis was that genes critical to tumourigenesis would most likely be found in such prominent alterations.  Molecular pathways associated with genes found within these regions were also defined.  We found that high-level DNA copy number alterations are frequently detected in early stage premalignant lesions that are associated with a high progression risk and that genes within such regions are significantly associated with molecular signalling cascades driving cancer cell processes.  These data provide evidence supporting hypotheses 1 and 2. Aim 4: To deduce the clonal relationships between samples from a single oral cancer field Although the oral cavity is accessible for routine screening of suspicious lesions, gene alterations are known to accrue in histological normal tissues.  Emerging optical and molecular technologies have recently been applied to provide a powerful means for defining the size of the altered field.  The work in chapter 5 details how clonal evolution was defined for multiple biopsies that were obtained from a contiguous field of alteration that extended beyond a clinically visible case of OSCC.  Genome alterations detected for each specimen were compared to define whether lesions arose independently or as a consequence of a shared progenitor cell.  We found that the oral cancer field is genetically heterogeneous and that the presence of shared genomic breakpoints was an effective means of defining clonal evolution between biopsies.  These data supported hypothesis 3 of this thesis. Figure 1.1 22 Hyperplasia Mild dysplasia Moderate dysplasia Severe dysplasia Cancer Increasing genomic instability Increasing risk for malignant transformation Figure 1.1.  Histological progression model of oral cancer development.  Oral cancer is known to develop through a series of histological changes, from hyperplasia, to mild, moderate, and severe dysplasia, and cancer (hematoxylin and eosin, X40). (adapted from Lippman & Hong 2001) Figure 1.2 23 True second primary tumour (SPT) Local recurrence or metastasis Second eld tumours (SFT) or premalignant cell migration Independent origin Carcinogenic exposure Normal epithelial cells Basal cell layer Stroma Unrelated genetic alteration Related genetic alteration Figure 1.2. Molecular differences for the development of true second primary tumour, local recurrence, or second field tumour.  Carcinogens targeting a single cell at the basal layer, which will develop into an area of premalignant cells. On the left panel, multiple epithelial cells are further exposed to carcinogenic exposure at different sites and result in the development of two tumours that are genetically unrelated.  On the right panel, the single altered cell continuously replacing the normal mucosa by clonal expansion, where further genetic alterations accumulate and result in the development of two tumours that are genetically related to each other. Figure 1.3 24 select BACs in a tiling-path manner. 26928 overlapping BAC DNA is spotted in duplicate onto glass slides Reference DNA Sample DNA Scan for signal dye ratios Robotic Spotter SeeGH custom software is used to plot data against their chromosomal positions 1 Mb Figure 1.3. Tiling-path array comparative genomic hybridization (CGH).  The array CGH is consisted of > 26,000 overlapping bacterial artificial chromosomes covering the entire sequenced genome.  Briefly, the sample and reference DNA are differentially labeled with Cyanine-dyes and are competitively hybridized on to the array.  The array will be scanned and the signal intensity ratios will be calculated for a measure of copy number.  25    1.11. References Allen, D. C. & R. I. Cameron, 2004. Histopathology Specimens: Clinical, Pathological and Laboratory Aspects, London: Springer-Verlag. Baldwin, C., C. Garnis, L. Zhang, M. P. Rosin & W. L. Lam, 2005. Multiple microalterations detected at high frequency in oral cancer. Cancer Res, 65(17), 7561-7. Banerjee, A. G., I. Bhattacharyya & J. K. Vishwanatha, 2005. Identification of genes and molecular pathways involved in the progression of premalignant oral epithelia. Mol Cancer Ther, 4(6), 865-75. Bedi, G. C., W. H. Westra, E. Gabrielson, W. Koch & D. Sidransky, 1996. Multiple head and neck tumors: evidence for a common clonal origin. Cancer Res, 56(11), 2484-7. Begum, S., D. Cao, M. Gillison, M. Zahurak & W. H. Westra, 2005. Tissue distribution of human papillomavirus 16 DNA integration in patients with tonsillar carcinoma. Clin Cancer Res, 11(16), 5694-9. Bockmuhl, U., G. Wolf, S. Schmidt, A. Schwendel, V. Jahnke, M. Dietel & I. Petersen, 1998. Genomic alterations associated with malignancy in head and neck cancer. Head Neck, 20(2), 145-51. Braakhuis, B. J., R. H. Brakenhoff & C. R. Leemans, 2005a. Second field tumors: a new opportunity for cancer prevention? Oncologist, 10(7), 493-500. Braakhuis, B. J., C. R. Leemans & R. H. Brakenhoff, 2005b. Expanding fields of genetically altered cells in head and neck squamous carcinogenesis. Semin Cancer Biol, 15(2), 113-20. 26   Braakhuis, B. J., M. P. Tabor, J. A. Kummer, C. R. Leemans & R. H. Brakenhoff, 2003. A genetic explanation of Slaughter's concept of field cancerization: evidence and clinical implications. Cancer Res, 63(8), 1727-30. Brennan, J. A., J. O. Boyle, W. M. Koch, S. N. Goodman, R. H. Hruban, Y. J. Eby, M. J. Couch, A. A. Forastiere & D. Sidransky, 1995. Association between cigarette smoking and mutation of the p53 gene in squamous-cell carcinoma of the head and neck. N Engl J Med, 332(11), 712-7. Brennan, P., S. Lewis, M. Hashibe, D. A. Bell, P. Boffetta, C. Bouchardy, N. Caporaso, C. Chen, C. Coutelle, S. R. Diehl, R. B. Hayes, A. F. Olshan, S. M. Schwartz, E. M. Sturgis, Q. Wei, A. I. Zavras & S. Benhamou, 2004. Pooled analysis of alcohol dehydrogenase genotypes and head and neck cancer: a HuGE review. Am J Epidemiol, 159(1), 1-16. Califano, J., P. van der Riet, W. Westra, H. Nawroz, G. Clayman, S. Piantadosi, R. Corio, D. Lee, B. Greenberg, W. Koch & D. Sidransky, 1996. Genetic progression model for head and neck cancer: implications for field cancerization. Cancer Res, 56(11), 2488-92. Carvalho, A. L., C. Jeronimo, M. M. Kim, R. Henrique, Z. Zhang, M. O. Hoque, S. Chang, M. Brait, C. S. Nayak, W. W. Jiang, Q. Claybourne, Y. Tokumaru, J. Lee, D. Goldenberg, E. Garrett-Mayer, S. Goodman, C. S. Moon, W. Koch, W. H. Westra, D. Sidransky & J. A. Califano, 2008. Evaluation of promoter hypermethylation detection in body fluids as a screening/diagnosis tool for head and neck squamous cell carcinoma. Clin Cancer Res, 14(1), 97-107. Chen, C., E. Mendez, J. Houck, W. Fan, P. Lohavanichbutr, D. Doody, B. Yueh, N. D. Futran, M. Upton, D. G. Farwell, S. M. Schwartz & L. P. Zhao, 2008. Gene expression profiling identifies genes predictive of oral squamous cell carcinoma. Cancer Epidemiol Biomarkers Prev, 17(8), 2152-62. Chung, C. H., J. S. Parker, K. Ely, J. Carter, Y. Yi, B. A. Murphy, K. K. Ang, A. K. El- Naggar, A. M. Zanation, A. J. Cmelak, S. Levy, R. J. Slebos & W. G. Yarbrough, 27   2006. Gene expression profiles identify epithelial-to-mesenchymal transition and activation of nuclear factor-kappaB signaling as characteristics of a high-risk head and neck squamous cell carcinoma. Cancer Res, 66(16), 8210-8. Chung, C. H., J. S. Parker, G. Karaca, J. Wu, W. K. Funkhouser, D. Moore, D. Butterfoss, D. Xiang, A. Zanation, X. Yin, W. W. Shockley, M. C. Weissler, L. G. Dressler, C. G. Shores, W. G. Yarbrough & C. M. Perou, 2004. Molecular classification of head and neck squamous cell carcinomas using patterns of gene expression. Cancer Cell, 5(5), 489-500. Coe, B. P., B. Ylstra, B. Carvalho, G. A. Meijer, C. Macaulay & W. L. Lam, 2007. Resolving the resolution of array CGH. Genomics, 89(5), 647-53. Committee, C. C. S. s. S., (2008). Canadian Cancer society/National Cancer Institute of Canada: Canadian Cancer Statistics 2008,  Toronto, Canada: Canadian Cancer Society. Crissman, J. D. & R. J. Zarbo, 1989. Dysplasia, in situ carcinoma, and progression to invasive squamous cell carcinoma of the upper aerodigestive tract. Am J Surg Pathol, 13 Suppl 1, 5-16. Curado, M. P. & M. Hashibe, 2009. Recent changes in the epidemiology of head and neck cancer. Curr Opin Oncol, 21(3), 194-200. Davies, J. J., I. M. Wilson & W. L. Lam, 2005. Array CGH technologies and their applications to cancer genomes. Chromosome Res, 13(3), 237-48. Elser, C., L. L. Siu, E. Winquist, M. Agulnik, G. R. Pond, S. F. Chin, P. Francis, R. Cheiken, J. Elting, A. McNabola, D. Wilkie, O. Petrenciuc & E. X. Chen, 2007. Phase II trial of sorafenib in patients with recurrent or metastatic squamous cell carcinoma of the head and neck or nasopharyngeal carcinoma. J Clin Oncol, 25(24), 3766-73. 28   Epstein, J. B., M. Gorsky, R. J. Cabay, T. Day & W. Gonsalves, 2008. Screening for and diagnosis of oral premalignant lesions and oropharyngeal squamous cell carcinoma: role of primary care physicians. Can Fam Physician, 54(6), 870-5. Epstein, J. B. & P. Guneri, 2009. The adjunctive role of toluidine blue in detection of oral premalignant and malignant lesions. Curr Opin Otolaryngol Head Neck Surg, 17(2), 79-87. Escher, A., E. Piotet, F. Waridel, R. Iggo & P. Monnier, 2008. p53 mutation in histologically normal mucosa of the aero-digestive tract is not a marker of increased risk for second primary carcinoma in head and neck cancer patients. Eur Arch Otorhinolaryngol. Freier, K., C. Schwaenen, C. Sticht, C. Flechtenmacher, J. Muhling, C. Hofele, B. Radlwimmer, P. Lichter & S. Joos, 2007. Recurrent FGFR1 amplification and high FGFR1 protein expression in oral squamous cell carcinoma (OSCC). Oral Oncol, 43(1), 60-6. Fresko, D. & S. S. Lazarus, 1981. Oral carcinoma in situ. Its progression to squamous, basosquamous, and basal-cell carcinoma. Arch Pathol Lab Med, 105(1), 15-9. Garnis, C., T. P. Buys & W. L. Lam, 2004a. Genetic alteration and gene expression modulation during cancer progression. Mol Cancer, 3, 9. Garnis, C., J. Campbell, L. Zhang, M. P. Rosin & W. L. Lam, 2004b. OCGR array: an oral cancer genomic regional array for comparative genomic hybridization analysis. Oral Oncol, 40(5), 511-9. Garnis, C., R. Chari, T. P. Buys, L. Zhang, R. T. Ng, M. P. Rosin & W. L. Lam, 2009. Genomic imbalances in precancerous tissues signal oral cancer risk. Mol Cancer, 8, 50. Garnis, C., B. P. Coe, A. Ishkanian, L. Zhang, M. P. Rosin & W. L. Lam, 2004c. Novel regions of amplification on 8q distinct from the MYC locus and frequently altered in oral dysplasia and cancer. Genes Chromosomes Cancer, 39(1), 93-8. 29   Garnis, C., B. P. Coe, L. Zhang, M. P. Rosin & W. L. Lam, 2004d. Overexpression of LRP12, a gene contained within an 8q22 amplicon identified by high-resolution array CGH analysis of oral squamous cell carcinomas. Oncogene, 23(14), 2582- 6. Gillison, M. L., G. D'Souza, W. Westra, E. Sugar, W. Xiao, S. Begum & R. Viscidi, 2008. Distinct risk factor profiles for human papillomavirus type 16-positive and human papillomavirus type 16-negative head and neck cancers. J Natl Cancer Inst, 100(6), 407-20. Gillison, M. L., W. M. Koch, R. B. Capone, M. Spafford, W. H. Westra, L. Wu, M. L. Zahurak, R. W. Daniel, M. Viglione, D. E. Symer, K. V. Shah & D. Sidransky, 2000. Evidence for a causal association between human papillomavirus and a subset of head and neck cancers. J Natl Cancer Inst, 92(9), 709-20. Gottschlich, S., P. Ambrosch, C. Cordes, T. Gorogh, S. Schreiber & R. Hasler, 2006. Gene expression profiling of head and neck squamous cell carcinoma using cDNA microarrays. Int J Oncol, 29(3), 605-13. Han, J., M. Kioi, W. S. Chu, J. L. Kasperbauer, S. E. Strome & R. K. Puri, 2009. Identification of potential therapeutic targets in human head & neck squamous cell carcinoma. Head Neck Oncol, 1(1), 27. Hayward, J. R. & J. A. Regezi, 1977. Oral dysplasia and in situ carcinoma: clinicopathologic correlations of eight patients. J Oral Surg, 35(9), 756-62. Hildesheim, A., R. Herrero, S. Wacholder, A. C. Rodriguez, D. Solomon, M. C. Bratti, J. T. Schiller, P. Gonzalez, G. Dubin, C. Porras, S. E. Jimenez & D. R. Lowy, 2007. Effect of human papillomavirus 16/18 L1 viruslike particle vaccine among young women with preexisting infection: a randomized trial. JAMA, 298(7), 743-53. Homann, N., H. Jousimies-Somer, K. Jokelainen, R. Heine & M. Salaspuro, 1997a. High acetaldehyde levels in saliva after ethanol consumption: methodological aspects and pathogenetic implications. Carcinogenesis, 18(9), 1739-43. 30   Homann, N., P. Karkkainen, T. Koivisto, T. Nosova, K. Jokelainen & M. Salaspuro, 1997b. Effects of acetaldehyde on cell regeneration and differentiation of the upper gastrointestinal tract mucosa. J Natl Cancer Inst, 89(22), 1692-7. Homann, N., J. Tillonen, J. H. Meurman, H. Rintamaki, C. Lindqvist, M. Rautio, H. Jousimies-Somer & M. Salaspuro, 2000. Increased salivary acetaldehyde levels in heavy drinkers and smokers: a microbiological approach to oral cavity cancer. Carcinogenesis, 21(4), 663-8. Huang, Q., G. P. Yu, S. A. McCormick, J. Mo, B. Datta, M. Mahimkar, P. Lazarus, A. A. Schaffer, R. Desper & S. P. Schantz, 2002. Genetic differences detected by comparative genomic hybridization in head and neck squamous cell carcinomas from different tumor sites: construction of oncogenetic trees for tumor progression. Genes Chromosomes Cancer, 34(2), 224-33. Hunter, K. D., E. K. Parkinson & P. R. Harrison, 2005. Profiling early head and neck cancer. Nat Rev Cancer, 5(2), 127-35. Jang, S. J., I. Chiba, A. Hirai, W. K. Hong & L. Mao, 2001. Multiple oral squamous epithelial lesions: are they genetically related? Oncogene, 20(18), 2235-42. Jarvinen, A. K., R. Autio, S. Haapa-Paananen, M. Wolf, M. Saarela, R. Grenman, I. Leivo, O. Kallioniemi, A. A. Makitie & O. Monni, 2006. Identification of target genes in laryngeal squamous cell carcinoma by high-resolution copy number and gene expression microarray analyses. Oncogene, 25(52), 6997-7008. Jarvinen, A. K., R. Autio, S. Kilpinen, M. Saarela, I. Leivo, R. Grenman, A. A. Makitie & O. Monni, 2008. High-resolution copy number and gene expression microarray analyses of head and neck squamous cell carcinoma cell lines of tongue and larynx. Genes Chromosomes Cancer, 47(6), 500-9. Kahn, J. A., 2009. HPV vaccination for the prevention of cervical intraepithelial neoplasia. N Engl J Med, 361(3), 271-8. 31   Kalyankrishna, S. & J. R. Grandis, 2006. Epidermal growth factor receptor biology in head and neck cancer. J Clin Oncol, 24(17), 2666-72. Kawaguchi, H., A. K. El-Naggar, V. Papadimitrakopoulou, H. Ren, Y. H. Fan, L. Feng, J. J. Lee, E. Kim, W. K. Hong, S. M. Lippman & L. Mao, 2008. Podoplanin: a novel marker for oral cancer risk in patients with oral premalignancy. J Clin Oncol, 26(3), 354-60. Kreimer, A. R., A. J. Alberg, R. Daniel, P. E. Gravitt, R. Viscidi, E. S. Garrett, K. V. Shah & M. L. Gillison, 2004. Oral human papillomavirus infection in adults is associated with sexual behavior and HIV serostatus. J Infect Dis, 189(4), 686-98. Kreimer, A. R., G. M. Clifford, P. Boyle & S. Franceschi, 2005. Human papillomavirus types in head and neck squamous cell carcinomas worldwide: a systematic review. Cancer Epidemiol Biomarkers Prev, 14(2), 467-75. Lane, P. M., T. Gilhuly, P. Whitehead, H. Zeng, C. F. Poh, S. Ng, P. M. Williams, L. Zhang, M. P. Rosin & C. E. MacAulay, 2006. Simple device for the direct visualization of oral-cavity tissue fluorescence. J Biomed Opt, 11(2), 024006. Laronde, D. M., T. G. Hislop, J. M. Elwood & M. P. Rosin, 2008. Oral cancer: just the facts. J Can Dent Assoc, 74(3), 269-72. Lee, J. I., J. C. Soria, K. Hassan, D. Liu, X. Tang, A. El-Naggar, W. K. Hong & L. Mao, 2001. Loss of Fhit expression is a predictor of poor outcome in tongue cancer. Cancer Res, 61(3), 837-41. Li, Y., D. Elashoff, M. Oh, U. Sinha, M. A. St John, X. Zhou, E. Abemayor & D. T. Wong, 2006. Serum circulating human mRNA profiling and its utility for oral cancer detection. J Clin Oncol, 24(11), 1754-60. Li, Y., M. A. St John, X. Zhou, Y. Kim, U. Sinha, R. C. Jordan, D. Eisele, E. Abemayor, D. Elashoff, N. H. Park & D. T. Wong, 2004. Salivary transcriptome diagnostics for oral cancer detection. Clin Cancer Res, 10(24), 8442-50. 32   Lippman, S. M. & W. K. Hong, 2001. Molecular markers of the risk of oral cancer. N Engl J Med, 344(17), 1323-6. Maier, H., J. Zoller, A. Herrmann, M. Kreiss & W. D. Heller, 1993. Dental status and oral hygiene in patients with head and neck cancer. Otolaryngol Head Neck Surg, 108(6), 655-61. Mao, L., W. K. Hong & V. A. Papadimitrakopoulou, 2004. Focus on head and neck cancer. Cancer Cell, 5(4), 311-6. Mao, L., J. S. Lee, Y. H. Fan, J. Y. Ro, J. G. Batsakis, S. Lippman, W. Hittelman & W. K. Hong, 1996. Frequent microsatellite alterations at chromosomes 9p21 and 3p14 in oral premalignant lesions and their value in cancer risk assessment. Nat Med, 2(6), 682-5. Mendez, E., J. R. Houck, D. R. Doody, W. Fan, P. Lohavanichbutr, T. C. Rue, B. Yueh, N. D. Futran, M. P. Upton, D. G. Farwell, P. J. Heagerty, L. P. Zhao, S. M. Schwartz & C. Chen, 2009. A genetic expression profile associated with oral cancer identifies a group of patients at high risk of poor survival. Clin Cancer Res, 15(4), 1353-61. Mincer, H. H., S. A. Coleman & K. P. Hopkins, 1972. Observations on the clinical characteristics of oral lesions showing histologic epithelial dysplasia. Oral Surg Oral Med Oral Pathol, 33(3), 389-99. Mithani, S. K., W. K. Mydlarz, F. L. Grumbine, I. M. Smith & J. A. Califano, 2007. Molecular genetics of premalignant oral lesions. Oral Dis, 13(2), 126-33. Noutomi, Y., A. Oga, K. Uchida, M. Okafuji, M. Ita, S. Kawauchi, T. Furuya, Y. Ueyama & K. Sasaki, 2006. Comparative genomic hybridization reveals genetic progression of oral squamous cell carcinoma from dysplasia via two different tumourigenic pathways. J Pathol, 210(1), 67-74. O'Donnell, R. K., M. Kupferman, S. J. Wei, S. Singhal, R. Weber, B. O'Malley, Y. Cheng, M. Putt, M. Feldman, B. Ziober & R. J. Muschel, 2005. Gene expression 33   signature predicts lymphatic metastasis in squamous cell carcinoma of the oral cavity. Oncogene, 24(7), 1244-51. Park, N. J., H. Zhou, D. Elashoff, B. S. Henson, D. A. Kastratovic, E. Abemayor & D. T. Wong, 2009. Salivary microRNA: discovery, characterization, and clinical utility for oral cancer detection. Clin Cancer Res, 15(17), 5473-7. Pathare, S., A. A. Schaffer, N. Beerenwinkel & M. Mahimkar, 2009. Construction of oncogenetic tree models reveals multiple pathways of oral cancer progression. Int J Cancer, 124(12), 2864-71. Pelucchi, C., S. Gallus, W. Garavello, C. Bosetti & C. La Vecchia, 2008. Alcohol and tobacco use, and cancer risk for upper aerodigestive tract and liver. Eur J Cancer Prev, 17(4), 340-4. Piccirillo, J. F., I. Costas & M. E. Reichman, (2007). Cancers of the head and neck, in SEER Survival Monograph: Cancer Survival Among Adults: U.S. SEER Program, 1988-2001, Patient and Tumor Characteristics. Bethesda, MD: National Cancer Institute, SEER Program. Poh, C. F., S. Ng, K. W. Berean, P. M. Williams, M. P. Rosin & L. Zhang, 2008. Biopsy and histopathologic diagnosis of oral premalignant and malignant lesions. J Can Dent Assoc, 74(3), 283-8. Poh, C. F., S. P. Ng, P. M. Williams, L. Zhang, D. M. Laronde, P. Lane, C. Macaulay & M. P. Rosin, 2007. Direct fluorescence visualization of clinically occult high-risk oral premalignant disease using a simple hand-held device. Head Neck, 29(1), 71-6. Poh, C. F., L. Zhang, D. W. Anderson, J. S. Durham, P. M. Williams, R. W. Priddy, K. W. Berean, S. Ng, O. L. Tseng, C. MacAulay & M. P. Rosin, 2006. Fluorescence visualization detection of field alterations in tumor margins of oral cancer patients. Clin Cancer Res, 12(22), 6716-22. 34   Ralhan, R., L. V. Desouza, A. Matta, S. Chandra Tripathi, S. Ghanny, S. Dattagupta, A. Thakar, S. S. Chauhan & K. W. Siu, 2009. iTRAQ-multidimensional liquid chromatography and tandem mass spectrometry-based identification of potential biomarkers of oral epithelial dysplasia and novel networks between inflammation and premalignancy. J Proteome Res, 8(1), 300-9. Reed, A. L., J. Califano, P. Cairns, W. H. Westra, R. M. Jones, W. Koch, S. Ahrendt, Y. Eby, D. Sewell, H. Nawroz, J. Bartek & D. Sidransky, 1996. High frequency of p16 (CDKN2/MTS-1/INK4A) inactivation in head and neck squamous cell carcinoma. Cancer Res, 56(16), 3630-3. Reichart, P. A. & H. P. Philipsen, 2005. Oral erythroplakia--a review. Oral Oncol, 41(6), 551-61. Roberts, J. N., C. B. Buck, C. D. Thompson, R. Kines, M. Bernardo, P. L. Choyke, D. R. Lowy & J. T. Schiller, 2007. Genital transmission of HPV in a mouse model is potentiated by nonoxynol-9 and inhibited by carrageenan. Nat Med, 13(7), 857- 61. Roepman, P., L. F. Wessels, N. Kettelarij, P. Kemmeren, A. J. Miles, P. Lijnzaad, M. G. Tilanus, R. Koole, G. J. Hordijk, P. C. van der Vliet, M. J. Reinders, P. J. Slootweg & F. C. Holstege, 2005. An expression profile for diagnosis of lymph node metastases from primary head and neck squamous cell carcinomas. Nat Genet, 37(2), 182-6. Rosin, M. P., X. Cheng, C. Poh, W. L. Lam, Y. Huang, J. Lovas, K. Berean, J. B. Epstein, R. Priddy, N. D. Le & L. Zhang, 2000. Use of allelic loss to predict malignant risk for low-grade oral epithelial dysplasia. Clin Cancer Res, 6(2), 357- 62. Schepman, K. P., E. H. van der Meij, L. E. Smeele & I. van der Waal, 1998. Malignant transformation of oral leukoplakia: a follow-up study of a hospital-based population of 166 patients with oral leukoplakia from The Netherlands. Oral Oncol, 34(4), 270-5. 35   Shin, D. M., J. S. Lee, S. M. Lippman, J. J. Lee, Z. N. Tu, G. Choi, K. Heyne, H. J. Shin, J. Y. Ro, H. Goepfert, W. K. Hong & W. N. Hittelman, 1996. p53 expressions: predicting recurrence and second primary tumors in head and neck squamous cell carcinoma. J Natl Cancer Inst, 88(8), 519-29. Slaughter, D. P., H. W. Southwick & W. Smejkal, 1953. Field cancerization in oral stratified squamous epithelium; clinical implications of multicentric origin. Cancer, 6(5), 963-8. Smeets, S. J., B. J. Braakhuis, S. Abbas, P. J. Snijders, B. Ylstra, M. A. van de Wiel, G. A. Meijer, C. R. Leemans & R. H. Brakenhoff, 2006. Genome-wide DNA copy number alterations in head and neck squamous cell carcinomas with or without oncogene-expressing human papillomavirus. Oncogene, 25(17), 2558-64. Smith, I. M., S. K. Mithani, C. Liu, S. S. Chang, S. Begum, M. Dhara, W. Westra, D. Sidranksy & J. A. Califano, 2009. Novel integrative methods for gene discovery associated with head and neck squamous cell carcinoma development. Arch Otolaryngol Head Neck Surg, 135(5), 487-95. Snijders, A. M., B. L. Schmidt, J. Fridlyand, N. Dekker, D. Pinkel, R. C. Jordan & D. G. Albertson, 2005. Rare amplicons implicate frequent deregulation of cell fate specification pathways in oral squamous cell carcinoma. Oncogene, 24(26), 4232-42. Squire, J. A., J. Bayani, C. Luk, L. Unwin, J. Tokunaga, C. MacMillan, J. Irish, D. Brown, P. Gullane & S. Kamel-Reid, 2002. Molecular cytogenetic analysis of head and neck squamous cell carcinoma: By comparative genomic hybridization, spectral karyotyping, and expression array analysis. Head Neck, 24(9), 874-87. Stenner, M. & J. P. Klussmann, 2009. Current update on established and novel biomarkers in salivary gland carcinoma pathology and the molecular pathways involved. Eur Arch Otorhinolaryngol, 266(3), 333-41. 36   Sturgis, E. M. & P. M. Cinciripini, 2007. Trends in head and neck cancer incidence in relation to smoking prevalence: an emerging epidemic of human papillomavirus- associated cancers? Cancer, 110(7), 1429-35. Summerlin, D. J., 1996. Precancerous and cancerous lesions of the oral cavity. Dermatol Clin, 14(2), 205-23. Tabor, M. P., R. H. Brakenhoff, H. J. Ruijter-Schippers, J. A. Kummer, C. R. Leemans & B. J. Braakhuis, 2004. Genetically altered fields as origin of locally recurrent head and neck cancer: a retrospective study. Clin Cancer Res, 10(11), 3607-13. Tabor, M. P., R. H. Brakenhoff, H. J. Ruijter-Schippers, J. E. Van Der Wal, G. B. Snow, C. R. Leemans & B. J. Braakhuis, 2002. Multiple head and neck tumors frequently originate from a single preneoplastic lesion. Am J Pathol, 161(3), 1051-60. Tabor, M. P., R. H. Brakenhoff, V. M. van Houten, J. A. Kummer, M. H. Snel, P. J. Snijders, G. B. Snow, C. R. Leemans & B. J. Braakhuis, 2001. Persistence of genetically altered fields in head and neck cancer patients: biological and clinical implications. Clin Cancer Res, 7(6), 1523-32. Tan, H. K., P. Saulnier, A. Auperin, L. Lacroix, O. Casiraghi, F. Janot, P. Fouret & S. Temam, 2008. Quantitative methylation analyses of resection margins predict local recurrences and disease-specific deaths in patients with head and neck squamous cell carcinomas. Br J Cancer, 99(2), 357-63. Tomioka, H., K. Morita, S. Hasegawa & K. Omura, 2006. Gene expression analysis by cDNA microarray in oral squamous cell carcinoma. J Oral Pathol Med, 35(4), 206-11. Tsui, I. F., R. Chari, T. P. Buys & W. L. Lam, 2007. Public Databases and Software for the Pathway Analysis of Cancer Genomes. Cancer Inform, 3, 389-407. 37   Tsui, I. F., C. F. Poh, C. Garnis, M. P. Rosin, L. Zhang & W. L. Lam, 2009. Multiple pathways in the FGF signaling network are frequently deregulated by gene amplification in oral dysplasias. Int J Cancer, 125(9), 2219-28. Tsui, I. F., M. P. Rosin, L. Zhang, R. T. Ng & W. L. Lam, 2008. Multiple aberrations of chromosome 3p detected in oral premalignant lesions. Cancer Prev Res (Phila Pa), 1(6), 424-9. Vachani, A., M. Nebozhyn, S. Singhal, L. Alila, E. Wakeam, R. Muschel, C. A. Powell, P. Gaffney, B. Singh, M. S. Brose, L. A. Litzky, J. Kucharczuk, L. R. Kaiser, J. S. Marron, M. K. Showe, S. M. Albelda & L. C. Showe, 2007. A 10-gene classifier for distinguishing head and neck squamous cell carcinoma and lung squamous cell carcinoma. Clin Cancer Res, 13(10), 2905-15. van Houten, V. M., C. R. Leemans, J. A. Kummer, J. Dijkstra, D. J. Kuik, M. W. van den Brekel, G. B. Snow & R. H. Brakenhoff, 2004. Molecular diagnosis of surgical margins and local recurrence in head and neck cancer patients: a prospective study. Clin Cancer Res, 10(11), 3614-20. Waldron, C. A. & W. G. Shafer, 1975. Leukoplakia revisited. A clinicopathologic study 3256 oral leukoplakias. Cancer, 36(4), 1386-92. Warren, S. & O. Gates, 1932. Multiple primary malignant tumors. A survey of the literature and a statistical study. . Am J Cancer, 16, 1358-414. Zabarovsky, E. R., M. I. Lerman & J. D. Minna, 2002. Tumor suppressor genes on chromosome 3p involved in the pathogenesis of lung and other cancers. Oncogene, 21(45), 6915-35. Zhang, L., M. Williams, C. F. Poh, D. Laronde, J. B. Epstein, S. Durham, H. Nakamura, K. Berean, A. Hovan, N. D. Le, G. Hislop, R. Priddy, J. Hay, W. L. Lam & M. P. Rosin, 2005. Toluidine blue staining identifies high-risk primary oral premalignant lesions with poor outcome. Cancer Res, 65(17), 8017-21. 38   Zhang, Z. F., H. Morgenstern, M. R. Spitz, D. P. Tashkin, G. P. Yu, T. C. Hsu & S. P. Schantz, 2000. Environmental tobacco smoking, mutagen sensitivity, and head and neck squamous cell carcinoma. Cancer Epidemiol Biomarkers Prev, 9(10), 1043-9. Ziober, A. F., K. R. Patel, F. Alawi, P. Gimotty, R. S. Weber, M. M. Feldman, A. A. Chalian, G. S. Weinstein, J. Hunt & B. L. Ziober, 2006. Identification of a gene signature for rapid screening of oral squamous cell carcinoma. Clin Cancer Res, 12(20 Pt 1), 5960-71. zur Hausen, H., 2002. Papillomaviruses and cancer: from basic studies to clinical application. Nat Rev Cancer, 2(5), 342-50 39   Chapter 2.  Integrative molecular characterization of head and neck cancer cell model genomes1     1 A version of this chapter has been accepted for publication.  Tsui IFL and Garnis C. (2009)  Integrative molecular characterization of head and neck cancer cell model genomes.  Head Neck.  [Published ahead-of-print 15 December 2009; DOI: 10.1002/hed.21311] Please see appendix A for all supplementary materials of this chapter.  40   2. Integrative molecular characterization of head and neck cancer cell model genomes 2.1. Introduction Head and neck squamous cell carcinoma (HNSCC) is the eighth most common cancer in the world.(Parkin et al. 2005)  It is a heterogeneous disease primarily affecting the oral cavity, salivary glands, oropharynx, hypopharynx, and larynx.  Model systems, such as cell lines are often used as research tools to study many cancer types, including HNSCC.(Lin et al. 2007; Martin et al. 2008)  They allow for manipulation of tumour cells in a laboratory setting.  However, tumourigenic cell lines in such studies have often not been fully characterized at the molecular level.  Like tumours themselves, cancer cell lines can vary greatly with respect to their genetic background.  Analysis of cell phenotypes without concurrent analysis of underlying molecular changes represents a lost opportunity to understand the mechanistic basis for disease. One characteristic of most solid tumours is genomic instability.  This is demonstrated by the numerous DNA gains and deletions present in tumour genomes.(Singh 2008)  DNA alterations can lead to aberrant expression of oncogenes and tumour suppressor genes that drive tumourigenesis.  For HNSCC development, gene alterations such as loss of chromosomes 3p and 9p, mutation of p53, loss of 13q, and high-level amplification of CCND1 have been associated with premalignant disease stages.(Califano et al. 1996; Mao et al. 1996; Rosin et al. 2000; Tsui et al. 2008)  Loss of chromosomes 4q and 8p, on the other hand, are believed to occur as later events in tumourigenesis.(Mao et al. 2004)  Such DNA alterations, including high-level amplification of CCND1,(Akervall et al. 1997) often leads to overexpression, while homozygous deletion of CDKN2A leads to silencing of p16INK4A in oral tumours.(Nakahara et al. 2001) Established HNSCC cell lines  ̶  particularly SCC-4, SCC-9, SCC-15, SCC-25, A253, and Cal27  ̶  are widely used as models of head and neck cancer.  A preliminary search in PubMed reveals 407 publications that use at least one of these lines, demonstrating their importance as tools for modeling biochemical, immunological, and pharmacological 41   behaviours.  These various behaviours could be influenced by the molecular alterations that exist in each line, yet few efforts have been made to characterize such changes. Previously, genetic alterations of four of the OSCC cell lines (SCC-4, SCC-9, SCC-15, and SCC-25) have been characterized by cDNA microarrays with low resolution and with no data on intragenic regions.(Jarvinen et al. 2008)  Thus, comprehensive characterization of genomic alterations in the most commonly used cell models are needed.  In addition, a priori knowledge of both genomic and gene expression changes for a given cell line would facilitate selection  of the most appropriate model system for a given in vitro analysis.  In this report we provide full characterization of genome and transcriptome alterations in six commonly used HNSCC cell lines.  These data will serve as an excellent resource when designing future experiments that attempt to model HNSCC behaviour.  2.2. Materials and methods 2.2.1. Cell lines Five tongue squamous cell carcinomas cell lines (SCC-4,(Rheinwald & Beckett 1981) SCC-9,(Rheinwald & Beckett 1981) SCC-25,(Rheinwald & Beckett 1981) SCC- 15,(Rheinwald & Beckett 1981) and Cal27(Gioanni et al. 1988)) and one submaxillary salivary gland epidermoid carcinoma cell line (A-253)(Giard et al. 1973) were purchased from the American Type Culture Collection (ATCC).  All cell lines were cultured according to ATCC recommendations.  Cells were harvested for DNA and RNA extractions when they reached ~ 80% confluency.  Genomic DNA was extracted by the standard phenol-chloroform extraction protocol, and total RNA was extracted using the TRIzol (Invitrogen, Carlsbad, CA) protocol followed by DNase I treatment.  42   2.2.2. Tiling-path DNA microarray Extracted DNA from each cell line was analyzed by whole genome tiling-path microarrays (SMRT v.2) developed at the British Columbia Cancer Research Centre Array Laboratory.(Ishkanian et al. 2004)  For this array, the whole genome is represented as more than 26,000 overlapping bacterial artificial chromosome (BAC) clones spotted in duplicate with tiling coverage of the human genome.  Pooled normal male DNA was used as reference for all microarray experiments.  Labelling, hybridization, scanning, and washing of the slides were performed as previously described.(Tsui et al. 2008) 2.2.3. Imaging and analysis of genomic data Array images were analyzed with GenePix Pro 6.1.  A three-step normalization procedure, including locally weighted linear regression (LOWESS) fitting, spatial, and median normalization was used to remove systematic biases.(Khojasteh et al. 2005) SeeGH software was used to combine duplicate spot data and display log2  signal intensity ratios in relation to genomic locations in the hg18 assembly (NCBI Build 36.1).(Chi et al. 2008)  Data points with standard deviation >0.075 and signal to noise ratio <3 in either channel were removed from each sample. Two separate algorithms, aCGH-Smooth and DNACopy, were used to detect copy number gains and losses.(Jong et al. 2004; Olshen et al. 2004)  An alteration was only called when detected concurrently by both algorithms.  aCGH-Smooth employs a local search algorithm that uses a maximum likelihood estimation to smooth the observed array CGH values between consecutive breakpoints to a common value.(Jong et al. 2004)  The Lambda and the maximum number of breakpoints in initial pool were set to 6.75 and 100 respectively.(Tsui et al. 2008)  DNAcopy employs a circular binary segmentation algorithm that uses a random permutation test to determine the statistical significance of change points.(Olshen et al. 2004)  Default settings in DNAcopy were used for our data. A third algorithm based on moving-average was used for the identification of high-level DNA amplification and presumptive homozygous deletions.(Lockwood et al. 2008)  The 43   threshold was set at log2 signal intensity ratio >0.8 for high-level amplification or <-0.8 for homozygous deletions.  Only regions containing ≥3 overlapping clones with such calls were identified in order to avoid false-positives due to hybridization artefacts. 2.2.4. Gene expression profiling Total RNA collected from each cell line was analyzed by Agilent Whole Human Genome Microarray 4x44K.  This array represents more than 41,000 unique human transcripts. Labelling and hybridizations were performed according to manufacturer’s instructions (Agilent Technologies).  Hybridized arrays were scanned using Axon GenePix 4000B and 4200A scanner. 2.2.5. Data analysis of expression profiles Array images were analyzed using GenePix Pro 6.1.  For normalization processing, each background-subtracted intensity value was divided by the median array intensity of each microarray.  The median array intensity was calculated based on the background- subtracted intensity value for all spots excluding control type spots on the array.  Genes within regions of high-level copy number change were extracted for each cell line.  Oral cell lines with no high-level copy number change serve as the baseline for the genes in each alteration region and the mean of each median-normalized intensity value of each probe was compared with those of the cell lines with high-level change.  2.3. Results and discussion Cancer cell lines are important model systems for investigating the biology of head and neck cancers.  They allow for characterization of a variety of disease phenotypes and also serve as a tool for functional screens.  When using such models, it is imperative to determine how closely the model resembles clinical disease.  However, genome-wide characterization of most cell lines has not previously been attempted.  Characterizing DNA alterations in these cells and determining the contribution of these changes to mRNA expression would provide a very valuable reference point for interpreting 44   experimental results generated using these cell lines.  The set of six genomic and six expression profiles has been deposited to Gene Expression Omnibus (GEO) database at NCBI, series accession number GSE16872. High-level DNA amplification is a common mechanism for driving overexpression of oncogenes, while homozygous deletion is known to inactivate tumour suppressor genes and also contribute to cancer processes.(Hahn et al. 1996; Lockwood et al. 2008; Snijders et al. 2005)  High resolution genomic analysis delineates the precise boundaries of amplified or deleted regions, thus defining genes that warrant further investigation.  Parallel analysis of gene expression data from the same cells helps to differentiate between "driver" and "passenger" genes in a given segmental DNA change, the former being genes that contribute to the cancer phenotype, the latter being altered simply due to its proximity to a "driver" gene.(Haber & Settleman 2007; Lockwood et al. 2008)  For example, any candidate gene in an amplicon that is not overexpressed is unlikely to represent a "driver" gene.  Figure 2.1 displays a summary of altered regions detected in each oral cell line, while the specific base pair positions are listed in Supplemental tables A.1 to A.6.  Furthermore, as this paper serves as a reference, the copy number status of well known cancer genes obtained from the Cancer Gene Census is catalogued in Table 2.1.(Futreal et al. 2004)  Whole genome copy number karyogram of each cell line and description of each cell line is represented in Supplemental figure A.1. It is known that head and neck cancers, like many solid tumours, are a molecularly heterogeneous group where several regions of alteration and genes have been reported at high frequencies but may occur in various combinations in a given tumour.  Genetic alterations frequently occurring in tumours indicate an importance for carcinogenesis. Loss of chromosome 3p and 9p are genetic events frequently documented in head and neck tumours, and both have been implicated as one of the earliest changes in oral premalignant lesions and associated with progression risks.(Mao et al. 1996; Mao et al. 2004)  All six cell lines revealed segmental losses of chromosome 3p and 9p (detailed regions of genetic alterations are listed in Supplemental Tables A.1 to A.6).  Cal27 45   revealed only one region of segmental loss of 9p24.1-p23, while the remaining five cell lines exhibited copy number loss at 9p21.3, which contain the tumour suppressor CDKN2A.  Genetic loss of chromosome 8p is also an expected change that has been frequently described in clinical specimens.(Baldwin et al. 2005; Smeets et al. 2006) Whole arm 8p loss was found in SCC-15, SCC-9, Cal27, while A253 exhibited a region of homozygous deletion on 8p23.2-p23.2 and a region of high-level DNA amplification on 8p22.  Gain of chromosome 8q is also a frequent genetic alteration in head and neck tumours, harbouring the known oncogene MYC.(Baldwin et al. 2005)  Cell lines A253, SCC-15, SCC-4, SCC-25, and SCC-9 all showed low level gain of 8q24 (MYC). Regions of high-level amplification, including 3q26, 7p11 and 11q13, occur frequently in head and neck tumours.(Baldwin et al. 2005; Redon et al. 2001; Smeets et al. 2006; Huang et al. 2006; Huang et al. 2002)  These regions have been associated with PIK3CA, EGFR, and CCND1 respectively.  SCC-4 and SCC-9 did not reveal genetic gain of 3q26.32 whereas the lines A253, SCC-15, and SCC-25 showed low level gains. Cell lines SCC-15 and Cal27 exhibited region of high-level amplification at the EGFR locus, and cell lines SCC-4, SCC-9, and A-253 exhibited low-level copy number gain. Cell lines SCC-9 and Cal27 showed genetic gain of the CCND1 locus, whereas four cell lines A253, SCC-15, SCC-4, and SCC-25 displayed high-level amplification of this locus.  Complex alterations on chromosome arm 11q were revealed SCC-15, SCC-4, SCC-25, and A253 (Fig. 2.2). One common characteristic of the cell lines is the presence of high-level amplifications, where many of the lines contained multiple regions of high-level DNA amplification. This is similar to tumour genomes where it has been reported that high-level amplification is observed in ~65% of cases, although usually only one or two regions per tumour.(Baldwin et al. 2005)  High-level amplifications indicating multiple copy numbers and regions of homozygous deletion, where both copies are lost, often result in gene overexpression and underexpression.(Albertson 2006; Hahn et al. 1996; Lockwood et al. 2008)  These types of alteration often arise under selective pressure of genes important for the growth of cancer cells.  Therefore we focused our analysis on the expression of genes within these regions.  The expression levels of genes within all the 46   regions of high-level DNA amplification and homozygous deletion detected in the six cell lines are presented in Supplemental figure A.2.  In general regions of homozygous deletion often resulted in lower expression levels for many of the genes within the region, while higher level of expression was often observed in at least one gene within an amplicon compared to cell lines without such amplicon.  For example, a common region of high-level amplification of 4.15 Mbp in size on 11p13-p12 was detected in Cal27 and A-253, which contains 20 RefSeq genes (Fig. 2.3).  The expression of 17 of these genes were analyzed, and the gene CD44 has a two-fold increase in expression compared to the other four cell lines without such high-level amplification.  Other genes inside this amplicon, PDHX and CAT, also have a 5-fold increase in expression.  The expression of CD44 is also among the top five-percentile of all expressed genes in the six cell lines, suggesting that while high-level DNA amplification causes higher expression of the gene, other mechanisms might also cause its high expression in cell lines without such amplicon.  However, as there are many ways to regulate gene expression, some of the amplified regions do not show the expected changes in gene expression.  Further molecular examinations of epigenetic alterations and sequence analyses would be essential to fully characterize the different molecular aspects of each cell line.  2.4. Conclusions Head and neck cancer cell lines are important model systems for investigating head and neck cancer biology.  Comprehensive characterization of their genetic alterations is useful for the selection and interpretation of studies using cell lines.  Our data presents the first comprehensive catalogue of copy number and expression alterations in six commonly used and easily obtainable oral cancer cell lines, thus providing a good resource for researchers to select and use these cell lines in experiments.  . Figure 2.1 47 Figure 2.1  Summary of copy number alteration in each cell line.  Copy number gain is presented as vertical lines on the right side of the chromosome, and vertical lines on the left side indicate copy number loss.  High-level copy number change is represented by a star on the right (DNA amplification) or left (homozygous deletion) of the chromosome.  Genetic alteration of each cell line is labelled with different colours, SCC-15 (blue), SCC-4 (red), SCC-25 (green), SCC-9 (purple), A-253 (orange), and Cal27 (pink). Figure 2.2 48 Figure 2.2.  Multiple levels of segmental copy number alterations are detected in six cell lines on chromosome 11q.  Each BAC clone is displayed as vertical line representing its genomic coverage.  Data points to the left and right of the centre line (purple) represent DNA copy number losses and gains, respectively.  Regions of gain are highlighted in red and loss is shaded in green. Figure 2.3 49 Figure 2.3.  Integrative analysis of genetic and expression levels of genes within 11p13-p12 amplicon.  A, Alignment of chromosome region 11p for six cell lines.  Each BAC clone is displayed as vertical line representing its genomic coverage.  Minimal region of DNA amplification of A-253 and Cal27 is marked by two black lines and is 4.15 Mbp in size.  B, The expression level of 17 genes within this amplicon is analyzed by expression arrays.  For each cell line, median-normalized expression values are represented on the y-axis and the corresponding gene is listed on the x-axis. 50   Table 2.1.  Copy number status of cancer genes in the six head and neck cancer cell lines. Gene Gene Name Gene ID Locus SCC- 15 SC C-4 SC C- 25 SC C-9 A- 253 Cal 27 ABI1 abl-interactor 1 10006 10p12. 1  -   - - ABL1 v-abl Abelson murine leukemia viral oncogene homolog 1 25 9q34.1 2 +  +  + + ABL2 v-abl Abelson murine leukemia viral oncogene homolog 2 27 1q24- q25   + ACSL6 acyl-CoA synthetase long- chain family member 6 23305 5q23.3 - - AF15Q14 AF15q14 protein 57082 15q15. 1  AF1Q ALL1-fused gene from chromosome 1q 10962 1q21 .2 -  + AF5q31 ALL1 fused gene from 5q31 27125 5q31.1 - - AKAP9 A kinase (PRKA) anchor protein (yotiao) 9 10142   7q21.2  +   + AKT1 v-akt murine thymoma viral oncogene homolog 1 207 14q32. 33  + + AKT2 v-akt murine thymoma viral oncogene homolog 2 208 19q13. 2 + ALK anaplastic lymphoma kinase (Ki-1) 238 2p23.2 ALO17 KIAA1618 protein 57714 17q25. 3 + -  + APC adenomatous polyposis of the colon gene 324 5q22.2 - - ARHGAP 26 GTPase regulator associated with focal adhesion kinase pp125(FAK) 23092 5q31.3 + - ARHGEF 12 RHO guanine nucleotide exchange factor (GEF) 12 (LARG) 23365 11q23. 3 - - + + - - ARNT aryl hydrocarbon receptor nuclear translocator 405 1q21 ASPSCR 1 alveolar soft part sarcoma chromosome region, candidate 1 79058 17q25. 3 + -  + ATF1 activating transcription factor 1 466 12q13. 12   + - 51   Gene Gene Name Gene ID Locus SCC- 15 SC C-4 SC C- 25 SC C-9 A- 253 Cal 27 ATIC 5-aminoimidazole-4- carboxamide ribonucleotide formyltransferase/IMP cyclohydrolase 471 2q35  - ATM ataxia telangiectasia mutated 472 11q22. 3 - -  + - - BCL10 B-cell CLL/lymphoma 10 8915 1p22.3   - BCL11A B-cell CLL/lymphoma 11A 53335 2p16.1 BCL11B B-cell CLL/lymphoma 11B (CTIP2) 64919 14q32. 2  ++ + BCL2 B-cell CLL/lymphoma 2 596 18q21. 33 - - -  - - BCL3 B-cell CLL/lymphoma 3 602 19q13. 31 + BCL5 B-cell CLL/lymphoma 5 603 17q23. 2  BCL6 B-cell CLL/lymphoma 6 604 3q27.3 + + + + + + BCL7A B-cell CLL/lymphoma 7A 605 12q24. 31  - + + BCL9 B-cell CLL/lymphoma 9 607 1q21  - BCR breakpoint cluster region 613 22q11. 23  - +  + BIRC3 baculoviral IAP repeat- containing 3 330 11q22. 2 + -  + - + BLM Bloom Syndrome 641 15q26. 1  BMPR1A bone morphogenetic protein receptor, type IA 657 10q23. 2     - BRAF v-raf murine sarcoma viral oncogene homolog B1 673 7q34  -   - BRCA1 familial breast/ovarian cancer gene 1 672 17q21. 31    + BRCA2 familial breast/ovarian cancer gene 2 675 13q13. 1 - +  - BRD4 bromodomain containing 4 23476 19p13. 12  BRIP1 BRCA1 interacting protein C-terminal helicase 1 83990 17q23. 2  BTG1 B-cell translocation gene 1, anti-proliferative 694 12q22  -   - BUB1B BUB1 budding uninhibited by benzimidazoles 1 homolog beta (yeast) 701 15q15. 1    - 52   Gene Gene Name Gene ID Locus SCC- 15 SC C-4 SC C- 25 SC C-9 A- 253 Cal 27 C12orf9 chromosome 12 open reading frame 9 93669 12q14. 3  C15orf21 chromosome 15 open reading frame 21 283651 15q21. 1    - - CARD11 caspase recruitment domain family, member 11 84433 7p22.2 + + + + + + CARS cysteinyl-tRNA synthetase 833 11p15. 4  CBFA2T1 core-binding factor, runt domain, alpha subunit 2;translocated to, 1  (ETO) 862 8q21.3 + +  + - CBFA2T3 core-binding factor, runt domain, alpha subunit 2; translocated to, 3 (MTG-16) 863 16q24. 3  CBFB core-binding factor, beta subunit 865 16q22. 1   + CBL  Cas-Br-M (murine) ecotropic retroviral transforming 867 11q23. 3 - - + + - - CCDC6 coiled-coil domain containing 6 8030 10q21. 2     - CCNB1IP 1 enhancer of invasion 10 - fused to HMGA2 57820 14q11. 2 +  +   ++ CCND1 cyclin D1 595 11q13. 3 ++ ++ ++ + ++ + CCND2 cyclin D2 894 12p13. 32   + - - CCND3 cyclin D3 896 6p21.1 +  + + - CDH1 cadherin 1, type 1, E- cadherin (epithelial) (ECAD) 999 16q22. 1   + CDH11 cadherin 11, type 2, OB- cadherin (osteoblast) 1009 16q21 CDK4 cyclin-dependent kinase 4 1019 12q14. 1   + CDK6 cyclin-dependent kinase 6 1021 7q12  ++   + CDKN2A- p14ARF cyclin-dependent kinase inhibitor 2A--  p14ARF protein 1029 9p21.3 - - - - - CDKN2A - p16(INK4 a) cyclin-dependent kinase inhibitor 2A (p16(INK4a)) gene 1029 9p21.3 - - - - - CDX2 caudal type homeo box transcription factor 2 1045 13q12. 2 - +  - CEBPA CCAAT/enhancer binding protein (C/EBP), alpha 1050 19q13. 11 + 53   Gene Gene Name Gene ID Locus SCC- 15 SC C-4 SC C- 25 SC C-9 A- 253 Cal 27 CEP1 centrosomal protein 1 11064 9q33.2 +  +  + + CHEK2 CHK2 checkpoint homolog (S. pombe) 11200 22q12. 1   + CHIC2 cysteine-rich hydrophobic domain 2 26511 4q12     - - CHN1 chimerin (chimaerin) 1 1123 2q31.1 CIC capicua homolog (Drosophila) 23152 19q13. 2 + CLTC clathrin, heavy polypeptide (Hc) 1213 17q23. 2  CLTCL1 clathrin, heavy polypeptide- like 1 8218 22q11. 21   +  + CMKOR1 chemokine orphan receptor 1 57007 2q37.3 COL1A1 collagen, type I, alpha 1 1277 17q21. 33  COPEB core promoter element binding protein (KLF6) 1316 10p15. 1  - -  - - COX6C cytochrome c oxidase subunit VIc 1345 8q22.2 + + + + CREB1 cAMP responsive element binding protein 1 1385 2q33.3  - CREB3L 2 cAMP responsive element binding protein 3-like 2 64764 7q32- q34  -   - CREBBP CREB binding protein (CBP) 1387 16p13. 3  + +  + CTNNB1 catenin (cadherin-associated protein), beta 1 1499 3p22.1 - - -  - - CXXC6 Leukemia-associated protein with a CXXC domain 80312 10q21. 3   +  - CYLD familial cylindromatosis gene 1540 16q12. 1   + DDB2 damage-specific DNA binding protein 2 1643 11p11. 2   + + DDIT3 DNA-damage-inducible transcript 3 1649 12q13. 3  + + DDX10 DEAD (Asp-Glu-Ala-Asp) box polypeptide 10 1662 11q22. 3 - -  + - - DDX6 DEAD (Asp-Glu-Ala-Asp) box polypeptide 6 1656 11q23. 3 - - + + - - DEK DEK oncogene (DNA binding) 7913 6p22.3 +   + - DUX4 double homeobox, 4 22947 4q35.2 - - - - - - 54   Gene Gene Name Gene ID Locus SCC- 15 SC C-4 SC C- 25 SC C-9 A- 253 Cal 27 EGFR epidermal growth factor receptor (erythroblastic leukemia viral (v-erb-b) oncogene homolog, avian) 1956 7p12.3- p12.1 ++ +  + + ++ EIF4A2 eukaryotic translation initiation factor 4A, isoform 2 1974 3q27.3 + + + + + + ELKS ELKS protein 23085 12p13. 33   + - - ELL ELL gene (11-19 lysine-rich leukemia gene) 8178 19p13. 11  ELN elastin 2006 7q11.2 3  + EML4 echinoderm microtubule associated protein like 4 27436 2p21 EP300 300 kd E1A-Binding protein gene 2033 22q13. 2  EPS15 epidermal growth factor receptor pathway substrate 15 (AF1p) 2060 1p32 ERBB2 v-erb-b2 erythroblastic leukemia viral oncogene homolog 2, neuro/glioblastoma derived oncogene homolog (avian) 2064 17q12 ERCC2 excision repair cross- complementing rodent repair deficiency, complementation group 2 (xeroderma pigmentosum D) 2068 19q13. 32 + ERCC3 excision repair cross- complementing rodent repair deficiency, complementation group 3 (xeroderma pigmentosum group B complementing) 2071 2q14.3 ERCC4 excision repair cross- complementing rodent repair deficiency, complementation group 4 2072 16p13. 12   + - + 55   Gene Gene Name Gene ID Locus SCC- 15 SC C-4 SC C- 25 SC C-9 A- 253 Cal 27 ERCC5 excision repair cross- complementing rodent repair deficiency, complementation group 5 (xeroderma pigmentosum, complementation group G (Cockayne syndrome)) 2073 13q33. 1 + + ERG  v-ets erythroblastosis virus E26 oncogene like (avian) 2078 21q22. 2  -  - - ETV1 ets variant gene 1 2115 7p21.2 + +   + + ETV4 ets variant gene 4 (E1A enhancer binding protein, E1AF) 2118 17q21. 31  ETV5 ets variant gene 5 2119 3q37.2 + + + + + + ETV6 ets variant gene 6 (TEL oncogene) 2120 12p13. 2  + + - - EVI1 ecotropic viral integration site 1 2122 3q26.2  +  +  + + EWSR1 Ewing sarcoma breakpoint region 1 (EWS) 2130 22q12. 2   + EXT1 multiple exostoses type 1 gene 2131 8q24.1 1 + + + + + EXT2 multiple exostoses type 2 gene 2132 11p11. 2   + + FANCA Fanconi anemia, complementation group A 2175 16q24. 3  FANCC Fanconi anemia, complementation group C 2176 9q22.3 2 +  +  + + FANCD2 Fanconi anemia, complementation group D2 2177 3p25.3 - - -   - FANCE Fanconi anemia, complementation group E 2178 6p21.3 1 +  + + FANCF Fanconi anemia, complementation group F 2188 11p14. 3     - FANCG Fanconi anemia, complementation group G 2189 9p13.3 - - - ++ + FBXW7 F-box and WD-40 domain protein 7 (archipelago homolog, Drosophila) 55294 4q31.3 -    - - FCGR2B Fc fragment of IgG, low affinity IIb, receptor for (CD32) 2213 1q23  - + FEV FEV protein - (HSRNAFEV) 54738 2q35 56   Gene Gene Name Gene ID Locus SCC- 15 SC C-4 SC C- 25 SC C-9 A- 253 Cal 27 FGFR1 fibroblast growth factor receptor 1 2260 8p12 -   - - - FGFR1O P FGFR1 oncogene partner (FOP) 11116 6q27 +  + + FGFR2 fibroblast growth factor receptor 2 2263 10q26. 12- q26.13  FGFR3 fibroblast growth factor receptor 3 2261 4p16.3 FH fumarate hydratase 2271 1q42.1   +   - FIP1L1 FIP1 like 1 (S. cerevisiae) 81608 4q12     - - FLCN folliculin, Birt-Hogg-Dube syndrome 201163 17p11. 2 + + + + + FLI1 Friend leukemia virus integration 1 2313 11q24. 3 - - + + - - FLT3 fms-related tyrosine kinase 3 2322 13q12. 2 - +  - FNBP1 formin binding protein 1 (FBP17) 23048 9q34.1 1 +  +  + + FOXO1A forkhead box O1A (FKHR) 2308 13q14. 11 -   - FOXO3A forkhead box O3A 2309 6q21 FOXP1 forkhead box P1 27086 3p13 - - - - - - FSTL3 follistatin-like 3 (secreted glycoprotein) 10272 19p13. 3  FUS fusion, derived from t(12;16) malignant liposarcoma 2521 16p11. 2   + FVT1 follicular lymphoma variant translocation 1 2531 18q21. 33 - - -  - - GAS7 growth arrest-specific 7 8522 17p13. 1 +   + + GATA2 GATA binding protein 2 2624 3q21.3      - GMPS guanine monphosphate synthetase 8833 3q25.3 1 +  + + + ++ GNAQ guanine nucleotide binding protein (G protein), q polypeptide 2776 9q21.2 + - +   + GNAS guanine nucleotide binding protein (G protein), alpha stimulating activity polypeptide 1 2778 20q13. 32 + + + + + + GOLGA5 golgi autoantigen, golgin subfamily a, 5  (PTC5) 9950 14q32. 12  ++ + 57   Gene Gene Name Gene ID Locus SCC- 15 SC C-4 SC C- 25 SC C-9 A- 253 Cal 27 GOPC golgi associated PDZ and coiled-coil motif containing 57120 6q22.1 + GPHN gephyrin (GPH) 10243 14q23. 3  + + + + HCMOG T-1 sperm antigen HCMOGT-1 92521 17p11. 2 +  + + + HEAB ATP_GTP binding protein 10978 11q12. 1 -  + +  + HIP1 huntingtin interacting protein 1 3092 7q11.2 3  + HIST1H4I histone 1, H4i (H4FM) 8294 6p22.1 +   + - HLF hepatic leukemia factor 3131 17q22 HLXB9 homeo box HB9 3110 7q36  -  - - HMGA1 high mobility group AT-hook 1 3159 6p21.3 1 +  + + HMGA2 high mobility group AT-hook 2 (HMGIC) 8091 12q14. 3  HNRNPA 2B1 heterogeneous nuclear ribonucleoprotein A2/B1 3181 7p15.2 + +  + + + HOXA11 homeo box A11 3207 7p15.2 + +  + + + HOXA13 homeo box A13 3209 7p15.2 + +  + + + HOXA9 homeo box A9 3205 7p15.2 + +  + + + HOXC11 homeo box C11 3227 12q13. 13   + + HOXC13 homeo box C13 3229 12q13. 13   + + HOXD11 homeo box D11 3237 2q31.1 HOXD13 homeo box D13 3239 2q31.1 HRAS v-Ha-ras Harvey rat sarcoma viral oncogene homolog 3265 11p15. 5  HRPT2 hyperparathyroidism 2  3279 1q21- q31     - HSPCA heat shock 90kDa protein 1, alpha 3320 14q32. 31  HSPCB heat shock 90kDa protein 1, beta 3326 6p21.1 +  + + - IDH1 isocitrate dehydrogenase 1 (NADP+), soluble 3417 2q33.3  - IGH@ immunoglobulin heavy locus 3492 14q32. 33  + + IGKC immunoglobulin kappa locus 50802 2p11.2 IGL@ immunoglobulin lambda locus 3535 22q11. 1-q11.2  - +  + 58   Gene Gene Name Gene ID Locus SCC- 15 SC C-4 SC C- 25 SC C-9 A- 253 Cal 27 IKZF1 IKAROS family zinc finger 1 10320 7p12.2 + +  + + + IL2 interleukin 2 3558 4q27 -    - - IL21R interleukin 21 receptor 50615 16p12. 1   +  + IL6ST interleukin 6 signal transducer (gp130, oncostatin M receptor) 3572 5q11.2 + - IRF4 interferon regulatory factor 4 3662 6p25.3 -  + + - IRTA1 immunoglobulin superfamily receptor translocation associated 1 83417 1q23.1 - - + ITK IL2-inducible T-cell kinase 3702 5q33.3 JAK2 Janus kinase 2  3717 9p24.1 - - - - + JAK3 Janus kinase 3 3718 19p13. 11  JAZF1 juxtaposed with another zinc finger gene 1 221895 7p15.2- p15.1 + +  + + + KIAA154 9 KIAA1549 57670 7q34  -   - KIT v-kit Hardy-Zuckerman 4 feline sarcoma viral oncogene homolog 3815 4q12     - - KRAS2 v-Ki-ras2 Kirsten rat sarcoma 2 viral oncogene homolog 3845 12p12. 1  +  - - KTN1 kinectin 1 (kinesin receptor) 3895 14q22. 3  + + ++ + LAF4 lymphoid nuclear protein related to AF4 3899 2q11.2  + LASP1 LIM and SH3 protein 1 3927 17q12    + LCK lymphocyte-specific protein tyrosine kinase 3932 1p35- p34.3   + LCP1 lymphocyte cytosolic protein 1 (L-plastin) 3936 13q14. 13 -   - LHFP lipoma HMGIC fusion partner 10186 13q13. 3 -   - LIFR leukemia inhibitory factor receptor 3977 5p13.1  + + + + + LMO1 LIM domain only 1 (rhombotin 1) (RBTN1) 4004 11p15. 4  LMO2 LIM domain only 2 (rhombotin-like 1) (RBTN2) 4005 11p13    + ++ ++ 59   Gene Gene Name Gene ID Locus SCC- 15 SC C-4 SC C- 25 SC C-9 A- 253 Cal 27 LPP LIM domain containing preferred translocation partner in lipoma 4026 3q27.3- q28 + + + + + + LYL1 lymphoblastic leukemia derived sequence 1 4066 19p13. 13   + MAF v-maf musculoaponeurotic fibrosarcoma oncogene homolog 4094 16q23. 1  MAFB v-maf musculoaponeurotic fibrosarcoma oncogene homolog B (avian) 9935 20q12 + + + + + + MALT1 mucosa associated lymphoid tissue lymphoma translocation gene 1 10892 18q21. 31 - - -  - - MAML2 mastermind-like 2 (Drosophila) 84441 11q21 - - - + - + MAP2K4 mitogen-activated protein kinase kinase 4 6416 17p11. 2 + -  + + MDM2 Mdm2 p53 binding protein homolog 4193 12q15 MDS1 myelodysplasia syndrome 1 4197 3q26.2  +  +   + MDS2 myelodysplastic syndrome 2 259283 1p36.1 2- p36.11     - MECT1 mucoepidermoid translocated 1 94159 19p13. 11     + MEN1 multiple endocrine neoplasia type 1 gene 4221 11q13. 1 + + + + + + MET met proto-oncogene (hepatocyte growth factor receptor) 4233 7q31.2  - MHC2TA MHC class II transactivator 4261 16p13. 13  + + - + MITF microphthalmia-associated transcription factor 4286 3p13 - - - - - - MKL1 megakaryoblastic leukemia (translocation) 1 57591 22q13. 1-q13.2  MLF1 myeloid leukemia factor 1 4291 3q25.3 2 +  + + + + MLH1 E.coli MutL homolog gene 4292 3p22.3 - - -  - - MLL myeloid/lymphoid or mixed- lineage leukemia (trithorax homolog, Drosophila) 4297 11q23. 3 - - + + - - 60   Gene Gene Name Gene ID Locus SCC- 15 SC C-4 SC C- 25 SC C-9 A- 253 Cal 27 MLLT1 myeloid/lymphoid or mixed- lineage leukemia (trithorax homolog, Drosophila); translocated to, 1 (ENL) 4298 19p13. 3      + MLLT10 myeloid/lymphoid or mixed- lineage leukemia (trithorax homolog, Drosophila); translocated to, 10 (AF10) 8028 10p12. 31     - - MLLT2  myeloid/lymphoid or mixed- lineage leukemia (trithorax homolog, Drosophila); translocated to, 2 (AF4) 4299 4q21.3 -  -  - - MLLT3 myeloid/lymphoid or mixed- lineage leukemia (trithorax homolog, Drosophila); translocated to, 3 (AF9) 4300 9p21.3 - - - - - MLLT4 myeloid/lymphoid or mixed- lineage leukemia (trithorax homolog, Drosophila); translocated to, 4 (AF6) 4301 6q27  +  + + MLLT6 myeloid/lymphoid or mixed- lineage leukemia (trithorax homolog, Drosophila); translocated to, 6 (AF17) 4302 17q21    + MN1 meningioma (disrupted in balanced translocation) 1 4330 22q12. 1   + MPL myeloproliferative leukemia virus oncogene, thrombopoietin receptor 4352 1p35.1   + MSF MLL septin-like fusion 10801 17q25 +   + MSH2 mutS homolog 2 (E. coli) 4436 2p21  - MSH6 mutS homolog 6 (E. coli) 2956 2p16.3  - MSI2 musashi homolog 2 (Drosophila) 124540 17q23. 2  MUC1 mucin 1, transmembrane 4582 1q22 - - + MUTYH mutY homolog (E. coli) 4595 1p34.1   + MYC v-myc myelocytomatosis viral oncogene homolog (avian) 4609 8q24.2 1 + + + + + MYCL1 v-myc myelocytomatosis viral oncogene homolog 1, lung carcinoma derived (avian) 4610 1p34.2   + 61   Gene Gene Name Gene ID Locus SCC- 15 SC C-4 SC C- 25 SC C-9 A- 253 Cal 27 MYCN v-myc myelocytomatosis viral related oncogene, neuroblastoma derived (avian) 4613 2p24.3 MYH11 myosin, heavy polypeptide 11, smooth muscle 4629 16p13. 11   + - + MYH9 myosin, heavy polypeptide 9, non-muscle 4627 22q12. 3  MYST4 MYST histone acetyltransferase (monocytic leukemia) 4 (MORF) 23522 10q22. 2   +  - NACA nascent-polypeptide- associated complex alpha polypeptide 4666 12q13. 3   + NBS1 Nijmegen breakage syndrome 1 (nibrin) 4683 8q21.3 + +  + - NCKIPS D SH3 protein interacting with Nck, 90 kDa (ALL1 fused gene from 3p21) 51517 3p21.3 1 - NCOA1 nuclear receptor coactivator 1 8648 2p23.3 NCOA2 nuclear receptor coactivator 2 (TIF2) 10499 8q13.3 +   + - NCOA4 nuclear receptor coactivator 4 - PTC3 (ELE1) 8031 10q11. 23     - NF1 neurofibromatosis type 1 gene 4763 17q11. 2  - NF2 neurofibromatosis type 2 gene 4771 22q12. 2   + NFKB2 nuclear factor of kappa light polypeptide gene enhancer in B-cells 2 (p49/p100) 4791 10q24. 32   + NIN ninein (GSK3B interacting protein) 51199 14q22. 1  + + + NOTCH1 Notch homolog 1, translocation-associated (Drosophila) (TAN1) 4851 9q34.3  +  +  + + NPM1 nucleophosmin (nucleolar phosphoprotein B23, numatrin) 4869 5q35.1 +    + NR4A3 nuclear receptor subfamily 4, group A, member 3 (NOR1) 8013 9q31.1 +    + + NRAS neuroblastoma RAS viral (v- ras) oncogene homolog 4893 1p13.2    + 62   Gene Gene Name Gene ID Locus SCC- 15 SC C-4 SC C- 25 SC C-9 A- 253 Cal 27 NSD1 nuclear receptor binding SET domain protein 1 64324 5q35.2- q35.3 +    + NTRK1 neurotrophic tyrosine kinase, receptor, type 1 4914 1q23.1 - - + NTRK3 neurotrophic tyrosine kinase, receptor, type 3 4916 15q25. 3    - - NUMA1 nuclear mitotic apparatus protein 1 4926 11q13. 4 + ++ ++ + - + NUP214 nucleoporin 214kDa (CAN) 8021 9q34.1 3 +  +  + + NUP98 nucleoporin 98kDa 4928 11p15. 4  NUT nuclear protien in testis 256646 15q14    - - OLIG2 oligodendrocyte lineage transcription factor 2 (BHLHB1) 10215 21q22. 11  -  + OMD osteomodulin 4958 9q22.3 1 +  +  + + PAFAH1 B2 platelet-activating factor acetylhydrolase, isoform Ib, beta subunit 30kDa 5049 11q23. 3 - - + + - - PALB2 partner and localizer of BRCA2 79728 16p12. 1   +  + PAX3 paired box gene 3  5077 2q36.1 PAX5 paired box gene 5 (B-cell lineage specific activator protein) 5079 9p13.2 - - - - + PAX7 paired box gene 7 5081 1p36.1 3   +  - PAX8 paired box gene 8 7849 2q13 PBX1 pre-B-cell leukemia transcription factor 1 5087 1q23.3  - +  - PCM1 pericentriolar material 1 (PTC4) 5108 8p22- p21.3 -  - - ++ - PCSK7 proprotein convertase subtilisin/kexin type 7 9159 11q23. 3 - - + + - - PDE4DIP phosphodiesterase 4D interacting protein (myomegalin) 9659 1q21.1 - - + PDGFB platelet-derived growth factor beta polypeptide (simian sarcoma viral (v-sis) oncogene homolog) 5155 22q13. 1  PDGFRA platelet-derived growth factor, alpha-receptor 5156 4q12     - - 63   Gene Gene Name Gene ID Locus SCC- 15 SC C-4 SC C- 25 SC C-9 A- 253 Cal 27 PDGFRB platelet-derived growth factor receptor, beta polypeptide 5159 5q32  + - PER1 period homolog 1 (Drosophila) 5187 17p13. 1 +   + + PHOX2B paired-like homeobox 2b 8929 4p13    -  - + PICALM phosphatidylinositol binding clathrin assembly protein (CALM) 8301 11q14. 2 - - - + - + PIK3CA phosphoinositide-3-kinase, catalytic, alpha polypeptide 5290 3q26.3 2 +  +  + + PIK3R1 phosphoinositide-3-kinase, regulatory subunit 1 (alpha) 5295 5q13.1 PIM1 pim-1 oncogene 5292 6p21.2  +  + + - PLAG1 pleiomorphic adenoma gene 1 5324 8q12.1 + +  + PML promyelocytic leukemia 5371 15q24. 1   + - - PMS1 PMS1 postmeiotic segregation increased 1 (S. cerevisiae) 5378 2q32.2  - PMS2 PMS2 postmeiotic segregation increased 2 (S. cerevisiae) 5395 7p22.1 + + + + + + PNUTL1 peanut-like 1 (Drosophila) 5413 22q11. 21   +  + POU2AF 1 POU domain, class 2, associating factor 1 (OBF1) 5450 11q23. 1 - -  + - - POU5F1 POU domain, class 5, transcription factor 1 5460 6p21.3 3 +   + PPARG peroxisome proliferative activated receptor, gamma 5468 3p25.2 - - -   - PRCC papillary renal cell carcinoma (translocation- associated) 5546 1q23.1 - - + PRDM16 PR domain containing 16 63976 1p36.3 2   + PRKAR1 A protein kinase, cAMP- dependent, regulatory, type I, alpha (tissue specific extinguisher 1) 5573 17q24. 2    + PRO1073 PRO1073 protein (ALPHA)  29005 11q13. 1 + + + + + + PRRX1 paired mesoderm homeo box 1 5396 1q24 .2  + +  - 64   Gene Gene Name Gene ID Locus SCC- 15 SC C-4 SC C- 25 SC C-9 A- 253 Cal 27 PSIP1 PC4 and SFRS1 interacting protein 1 (LEDGF) 11168 9p22.3 - - - - - PTCH Homolog of Drosophila Patched gene 5727 9q22.3 2 +  +  + + PTEN phosphatase and tensin homolog gene 5728 10q23. 31     - PTPN11 protein tyrosine phosphatase, non-receptor type 11 5781 12q24. 13   + + - RABEP1 rabaptin, RAB GTPase binding effector protein 1 (RABPT5) 9135 17p13. 2 +   + + RAD51L1 RAD51-like 1 (S. cerevisiae) (RAD51B) 5890 14q24. 1 + + + + + RANBP1 7 RAN binding protein 17 64901 5q35.1 +    + RAP1GD S1 RAP1, GTP-GDP dissociation stimulator 1 5910 4q23 -    - - RARA retinoic acid receptor, alpha 5914 17q21. 2    + RB1 retinoblastoma gene 5925 13q14. 2 -   - RBM15 RNA binding motif protein 15 64783 1p13.3   ++ RECQL4 RecQ protein-like 4 9401 8q24.3 + + + + + REL v-rel reticuloendotheliosis viral oncogene homolog (avian) 5966 2p16.1 RET ret proto-oncogene  5979 10q11. 21     - RHOH RAS homolog gene family, member H (TTF) 399 4p14   -  - + ROS1 v-ros UR2 sarcoma virus oncogene homolog 1 (avian) 6098 6q22.1 RPL22 ribosomal protein L22 (EAP) 6146 1p36.3 1 + RPN1 ribophorin I 6184 3q21.3      - RUNX1  runt-related transcription factor 1  (AML1) 861 21q22. 12  -  + - RUNXBP 2 runt-related transcription factor binding protein 2 (MOZ/ZNF220) 7994 8p11.2 1 - - + -  - SBDS Shwachman-Bodian- Diamond syndrome protein 51119 7q11.2 1 + + 65   Gene Gene Name Gene ID Locus SCC- 15 SC C-4 SC C- 25 SC C-9 A- 253 Cal 27 SDHB succinate dehydrogenase complex, subunit B, iron sulfur (Ip) 6390 1p36.1 3   +  - SDHC succinate dehydrogenase complex, subunit C, integral membrane protein, 15kDa 6391 1q23.3  - + SDHD succinate dehydrogenase complex, subunit D, integral membrane protein 6392 11q23. 1 - -  + - - SET SET translocation  6418 9q33.2 +  +  + + SFPQ splicing factor proline/glutamine rich(polypyrimidine tract binding protein associated) 6421 1p34.3   + SFRS3 splicing factor, arginine/serine-rich 3 6428 6p21.3 1 +  + + SH3GL1 SH3-domain GRB2-like 1 (EEN) 6455 19p13. 3  SIL TAL1 (SCL) interrupting locus 6491 1p33   + SLC45A3 solute carrier family 45, member 3 85414 1q32.1 + + +  - SMAD4 Homolog of Drosophila Mothers Against Decapentaplegic 4 gene 4089 18q21. 1 - - -  - - SMARCB 1 SWI/SNF related, matrix associated, actin dependent regulator of chromatin, subfamily b, member 1 6598 22q11. 23  - +  + SMO smoothened homolog (Drosophila) 6608 7q32.1  - SOCS1 suppressor of cytokine signaling 1 8651   16p13. 13  + + - + SS18 synovial sarcoma translocation, chromosome 18 6760 18q11. 2  - +   - SS18L1 synovial sarcoma translocation gene on chromosome 18-like 1 26039 20q13. 33 + + + + + + STK11 serine/threonine kinase 11 gene (LKB1) 6794 19p13. 3  STL Six-twelve leukemia gene 7955 6q22.3 1 + SUFU suppressor of fused homolog (Drosophila) 51684 10q24. 32   + 66   Gene Gene Name Gene ID Locus SCC- 15 SC C-4 SC C- 25 SC C-9 A- 253 Cal 27 SUZ12 suppressor of zeste 12 homolog (Drosophila) 23512 17q11. 2  -  - SYK spleen tyrosine kinase 6850 9q22.2 + - +  + + TAF15 TAF15 RNA polymerase II, TATA box binding protein (TBP)-associated factor, 68kDa 8148 17q12  -  - TAL1 T-cell acute lymphocytic leukemia 1 (SCL) 6886 1p33   + TAL2 T-cell acute lymphocytic leukemia 2 6887 9q31.2 +  +  + + TCEA1 transcription elongation factor A (SII), 1 6917 8q11.2 3 + +  + TCF1 transcription factor 1, hepatic (HNF1) 6927 12q24. 31  - + + TCF12 transcription factor 12 (HTF4, helix-loop-helix transcription factors 4) 6938 15q21. 3    - - TCF3 transcription factor 3 (E2A immunoglobulin enhancer binding factors E12/E47) 6929 19p13. 3  TCL1A T-cell leukemia/lymphoma 1A 8115 14q32. 13  ++ +  - TCL6 T-cell leukemia/lymphoma 6 27004 14q32. 13  ++ +  - TFEB transcription factor EB 7942 6p21.1 +  + + - TFG TRK-fused gene 10342 3q12.2      + TFPT TCF3 (E2A) fusion partner (in childhood Leukemia) 29844 19q13. 42 + +   + TFRC transferrin receptor (p90, CD71) 7037 3q29  + + + + + + THRAP3 thyroid hormone receptor associated protein 3 (TRAP150) 9967 1p34.3   + TIF1 transcriptional intermediary factor 1  (PTC6,TIF1A) 8805 7q34   - TLX1  T-cell leukemia, homeobox 1 (HOX11) 3195 10q24. 31   + TLX3 T-cell leukemia, homeobox 3 (HOX11L2) 30012 5q35.1 +    + TMPRSS 2 transmembrane protease, serine 2 7113 21q22. 3  - -- - - 67   Gene Gene Name Gene ID Locus SCC- 15 SC C-4 SC C- 25 SC C-9 A- 253 Cal 27 TNFRSF 17 tumor necrosis factor receptor superfamily, member 17 608 16p13. 13  + + - + TNFRSF 6 tumor necrosis factor receptor superfamily, member 6 (FAS) 355 10q23. 31     - TOP1 topoisomerase (DNA) I 7150 20q12 + + + + + + TP53 tumor protein p53 7157 17p13. 1 +   + + TPM3 tropomyosin 3 7170 1q21.3 - - + TPM4 tropomyosin 4 7171 19p13. 12  TPR translocated promoter region 7175 1q31.1     - TRA@ T cell receptor alpha locus 6955 14q11. 2 +  + TRBC1 T cell receptor beta locus 6957 7q34  -  - - TRD@ T cell receptor delta locus 6964 14q11. 2 +  + TRIM33  tripartite motif-containing 33 (PTC7,TIF1G) 51592 1p13.2    + TRIP11 thyroid hormone receptor interactor 11 9321 14q32. 12  + + TSC1 tuberous sclerosis 1 gene 7248 9q34.1 3 +  +  + + TSC2 tuberous sclerosis 2 gene 7249 16p13. 3  + +  + TSHR thyroid stimulating hormone receptor 7253 14q31. 1  + +  + TTL tubulin tyrosine ligase 150465 2q13 USP6 ubiquitin specific peptidase 6 (Tre-2 oncogene) 9098 17p13. 2 +   + + VHL von Hippel-Lindau syndrome gene 7428 3p25.3 - - -   - WHSC1 Wolf-Hirschhorn syndrome candidate 1(MMSET) 7468 4p16.3 WHSC1L 1 Wolf-Hirschhorn syndrome candidate 1-like 1 (NSD3) 54904 8p12  -   - - - WRN Werner syndrome (RECQL2) 7486 8p12  -  - - - - WT1 Wilms tumour 1 gene 7490 11p13    + ++ + XPA xeroderma pigmentosum, complementation group A 7507 9q22.3 3 +  +  + + XPC xeroderma pigmentosum, complementation group C 7508 3p25.1 - - -   - 68   Gene Gene Name Gene ID Locus SCC- 15 SC C-4 SC C- 25 SC C-9 A- 253 Cal 27 ZNF145 zinc finger protein 145 (PLZF) 7704 11q23. 2 - -  + - - ZNF198 zinc finger protein 198 7750 13q12. 11 - + ZNF278 zinc finger protein 278 (ZSG) 23598 22q12. 2   + ZNF331 zinc finger protein 331 55422 19q13. 42 + + ZNF384 zinc finger protein 384 (CIZ/NMP4) 171017 12p13. 31   + - - ZNF521 zinc finger protein 521 25925 18q11. 2  - +   - ZNF9 zinc finger protein 9 (a cellular retroviral nucleic acid binding protein) 7555 3q21.3     - ZNFN1A 1 zinc finger protein, subfamily 1A, 1 (Ikaros) 10320 7p12.2 + +   + + + Symbols: +, low-level copy number gain; ++, high-level copy number gain; -, low-level copy number loss; --, high-level copy number loss.  69    2.5.  References Akervall, J. A., R. J. Michalides, H. Mineta, A. Balm, A. Borg, M. R. Dictor, Y. Jin, B. Loftus, F. Mertens & J. P. Wennerberg, 1997. Amplification of cyclin D1 in squamous cell carcinoma of the head and neck and the prognostic value of chromosomal abnormalities and cyclin D1 overexpression. Cancer, 79(2), 380-9. Albertson, D. G., 2006. Gene amplification in cancer. Trends Genet, 22(8), 447-55. Baldwin, C., C. Garnis, L. Zhang, M. P. Rosin & W. L. Lam, 2005. Multiple microalterations detected at high frequency in oral cancer. Cancer Res, 65(17), 7561-7. Califano, J., P. van der Riet, W. Westra, H. Nawroz, G. Clayman, S. Piantadosi, R. Corio, D. Lee, B. Greenberg, W. Koch & D. Sidransky, 1996. Genetic progression model for head and neck cancer: implications for field cancerization. Cancer Res, 56(11), 2488-92. Chi, B., R. J. deLeeuw, B. P. Coe, R. T. Ng, C. MacAulay & W. L. Lam, 2008. MD- SeeGH: a platform for integrative analysis of multi-dimensional genomic data. BMC Bioinformatics, 9, 243. Futreal, P. A., L. Coin, M. Marshall, T. Down, T. Hubbard, R. Wooster, N. Rahman & M. R. Stratton, 2004. A census of human cancer genes. Nat Rev Cancer, 4(3), 177- 83. Giard, D. J., S. A. Aaronson, G. J. Todaro, P. Arnstein, J. H. Kersey, H. Dosik & W. P. Parks, 1973. In vitro cultivation of human tumors: establishment of cell lines derived from a series of solid tumors. J Natl Cancer Inst, 51(5), 1417-23. Gioanni, J., J. L. Fischel, J. C. Lambert, F. Demard, C. Mazeau, E. Zanghellini, F. Ettore, P. Formento, P. Chauvel, C. M. Lalanne & et al., 1988. Two new human tumor cell lines derived from squamous cell carcinomas of the tongue: 70   establishment, characterization and response to cytotoxic treatment. Eur J Cancer Clin Oncol, 24(9), 1445-55. Haber, D. A. & J. Settleman, 2007. Cancer: drivers and passengers. Nature, 446(7132), 145-6. Hahn, S. A., M. Schutte, A. T. Hoque, C. A. Moskaluk, L. T. da Costa, E. Rozenblum, C. L. Weinstein, A. Fischer, C. J. Yeo, R. H. Hruban & S. E. Kern, 1996. DPC4, a candidate tumor suppressor gene at human chromosome 18q21.1. Science, 271(5247), 350-3. Huang, X., T. E. Godfrey, W. E. Gooding, K. S. McCarty, Jr. & S. M. Gollin, 2006. Comprehensive genome and transcriptome analysis of the 11q13 amplicon in human oral cancer and synteny to the 7F5 amplicon in murine oral carcinoma. Genes Chromosomes Cancer, 45(11), 1058-69. Huang, X., S. M. Gollin, S. Raja & T. E. Godfrey, 2002. High-resolution mapping of the 11q13 amplicon and identification of a gene, TAOS1, that is amplified and overexpressed in oral cancer cells. Proc Natl Acad Sci U S A, 99(17), 11369-74. Ishkanian, A. S., C. A. Malloff, S. K. Watson, R. J. DeLeeuw, B. Chi, B. P. Coe, A. Snijders, D. G. Albertson, D. Pinkel, M. A. Marra, V. Ling, C. MacAulay & W. L. Lam, 2004. A tiling resolution DNA microarray with complete coverage of the human genome. Nat Genet, 36(3), 299-303. Jarvinen, A. K., R. Autio, S. Kilpinen, M. Saarela, I. Leivo, R. Grenman, A. A. Makitie & O. Monni, 2008. High-resolution copy number and gene expression microarray analyses of head and neck squamous cell carcinoma cell lines of tongue and larynx. Genes Chromosomes Cancer, 47(6), 500-9. Jong, K., E. Marchiori, G. Meijer, A. V. Vaart & B. Ylstra, 2004. Breakpoint identification and smoothing of array comparative genomic hybridization data. Bioinformatics, 20(18), 3636-7. 71   Khojasteh, M., W. L. Lam, R. K. Ward & C. MacAulay, 2005. A stepwise framework for the normalization of array CGH data. BMC Bioinformatics, 6, 274. Lin, C. J., J. R. Grandis, T. E. Carey, S. M. Gollin, T. L. Whiteside, W. M. Koch, R. L. Ferris & S. Y. Lai, 2007. Head and neck squamous cell carcinoma cell lines: established models and rationale for selection. Head Neck, 29(2), 163-88. Lockwood, W. W., R. Chari, B. P. Coe, L. Girard, C. Macaulay, S. Lam, A. F. Gazdar, J. D. Minna & W. L. Lam, 2008. DNA amplification is a ubiquitous mechanism of oncogene activation in lung and other cancers. Oncogene, 27(33), 4615-24. Mao, L., W. K. Hong & V. A. Papadimitrakopoulou, 2004. Focus on head and neck cancer. Cancer Cell, 5(4), 311-6. Mao, L., J. S. Lee, Y. H. Fan, J. Y. Ro, J. G. Batsakis, S. Lippman, W. Hittelman & W. K. Hong, 1996. Frequent microsatellite alterations at chromosomes 9p21 and 3p14 in oral premalignant lesions and their value in cancer risk assessment. Nat Med, 2(6), 682-5. Martin, C. L., S. C. Reshmi, T. Ried, W. Gottberg, J. W. Wilson, J. K. Reddy, P. Khanna, J. T. Johnson, E. N. Myers & S. M. Gollin, 2008. Chromosomal imbalances in oral squamous cell carcinoma: examination of 31 cell lines and review of the literature. Oral Oncol, 44(4), 369-82. Nakahara, Y., S. Shintani, M. Mihara, Y. Ueyama & T. Matsumura, 2001. High frequency of homozygous deletion and methylation of p16(INK4A) gene in oral squamous cell carcinomas. Cancer Lett, 163(2), 221-8. Olshen, A. B., E. S. Venkatraman, R. Lucito & M. Wigler, 2004. Circular binary segmentation for the analysis of array-based DNA copy number data. Biostatistics, 5(4), 557-72. Parkin, D. M., F. Bray, J. Ferlay & P. Pisani, 2005. Global cancer statistics, 2002. CA Cancer J Clin, 55(2), 74-108. 72   Redon, R., D. Muller, K. Caulee, K. Wanherdrick, J. Abecassis & S. du Manoir, 2001. A simple specific pattern of chromosomal aberrations at early stages of head and neck squamous cell carcinomas: PIK3CA but not p63 gene as a likely target of 3q26-qter gains. Cancer Res, 61(10), 4122-9. Rheinwald, J. G. & M. A. Beckett, 1981. Tumorigenic keratinocyte lines requiring anchorage and fibroblast support cultures from human squamous cell carcinomas. Cancer Res, 41(5), 1657-63. Rosin, M. P., X. Cheng, C. Poh, W. L. Lam, Y. Huang, J. Lovas, K. Berean, J. B. Epstein, R. Priddy, N. D. Le & L. Zhang, 2000. Use of allelic loss to predict malignant risk for low-grade oral epithelial dysplasia. Clin Cancer Res, 6(2), 357- 62. Singh, B., 2008. Molecular pathogenesis of head and neck cancers. J Surg Oncol, 97(8), 634-9. Smeets, S. J., B. J. Braakhuis, S. Abbas, P. J. Snijders, B. Ylstra, M. A. van de Wiel, G. A. Meijer, C. R. Leemans & R. H. Brakenhoff, 2006. Genome-wide DNA copy number alterations in head and neck squamous cell carcinomas with or without oncogene-expressing human papillomavirus. Oncogene, 25(17), 2558-64. Snijders, A. M., B. L. Schmidt, J. Fridlyand, N. Dekker, D. Pinkel, R. C. Jordan & D. G. Albertson, 2005. Rare amplicons implicate frequent deregulation of cell fate specification pathways in oral squamous cell carcinoma. Oncogene, 24(26), 4232-42. Tsui, I. F., M. P. Rosin, L. Zhang, R. T. Ng & W. L. Lam, 2008. Multiple aberrations of chromosome 3p detected in oral premalignant lesions. Cancer Prev Res (Phila Pa), 1(6), 424-9. 73   Chapter 3.  Multiple aberrations of chromosome 3p detected in oral premalignant lesions2     2 A version of this chapter has been published.  Tsui IFL, Rosin MP, Zhang L, Ng RL, Lam WL. (2008)  Multiple aberrations of chromosome 3p detected in oral premalignant lesions.  Cancer Prevention Research.  3: 424-429. Please see appendix B for all supplementary materials of this chapter.  74   3. Multiple aberrations of chromosome 3p detected in oral premalignant lesions 3.1. Introduction Loss of chromosome 3p is a common genetic event in many human cancers, with several putative tumor suppressor genes (TSGs) located in this region (Garnis et al. 2004; Zabarovsky et al. 2002).  In oral cancer, loss of 3p carries prognostic significance, particularly for risk of disease progression, development of second primary tumors, and emergence of local recurrences (Mao et al. 2004; Mao et al. 1996; Partridge et al. 2000; Rosin et al. 2000; Rosin et al. 2002).  Alterations on 3p are also thought to be key events in the progression of oral premalignant lesions (OPLs) to invasive disease. Previously, three microsatellite markers on this chromosome arm were used to demonstrate that 3p loss of heterozygosity (LOH) did in fact occur in OPLs and that alteration of this region increased in frequency with progression to cancer (Califano et al. 1996).  Other reports have also associated alterations in specific segments of 3p in OPLs with elevated risk of invasive transformation (Mao et al. 1996; Rosin et al. 2000). However, these markers have not been applicable in all cases, leaving the possibility that there are additional candidates within chromosome 3p that are necessary for progression to oral invasive cancer. To identify additional segments on chromosome 3p that are associated with disease progression, we undertook tiling-path array comparative genomic hybridization (CGH) for the entire chromosome arm in a panel of premalignant and invasive oral tissues.  In addition to its improved resolution, this platform also gave the benefit of being able to work with DNA of limited quantity (because of the small size of the captured lesions) and low quality (because of the use of formalin-fixed paraffin-embedded tissue, which typically precludes effective microarray profiling).  This allowed us to profile high-grade dysplasias (HGDs) (including severe dysplasias and carcinoma in situ [CIS] lesions, which have a high likelihood of progression to invasive disease), low-grade dysplasias (LGDs) with clinical outcome, and oral squamous cell carcinomas (OSCCs).  Parallel analysis of genomic data from these various tissues demonstrated stage-specific genetic alterations.  More importantly, association of these same data with clinical 75   features provided us with a new tool for predicting progression risks in LGDs, where the existing histopathological criteria are unable to predict progression.  3.2. Materials and methods 3.2.1. Tissue samples This study involves 94 archival FFPE specimens from 86 patients: 24 LGDs (low-grade dysplasias: 2 hyperplasias, 22 mild and moderate dysplasias), 47 HGDs (severe dysplasias and CIS), and 23 OSCCs, all obtained from the British Columbia Oral Biopsy Service.  Clinical information and demographics for these cases are presented in Supplementary Table B.1.  All LGDs and HGDs came from patients with no prior history of cancer.  All LGD patients were enrolled in a longitudinal study established at the British Columbia Cancer Prevention Program with a median follow up of seven years: of the 24 LGDs, 15 LGDs did not progress during the period from 1985 to 2007, whereas nine LGDs progressed to cancer.  In British Columbia patients with HGDs are treated with surgery and followed for recurrence.  All diagnoses were confirmed by the study pathologist (L. Z.).  Representative sections were micro-dissected and DNA was extracted as previously described (Baldwin et al. 2005). 3.2.2. Array CGH analysis Array CGH was performed as previously described (Baldwin et al. 2005; Ishkanian et al. 2004; Wong et al. 2007).  Genomic arrays (SMRT v.1 and v.2) were obtained from the BC Cancer Research Center Array Laboratory (Baldwin et al. 2005; Ishkanian et al. 2004).  Briefly, sample and normal reference genomic DNA (250 ng each) were differentially labeled and mixed with 100 µg of human Cot-1 DNA (Invitrogen, Ontario), purified and hybridized to the array at 45ºC for 36 hours before washing.  Hybridized arrays were scanned for signal intensities as previously described (Coe et al. 2006; Baldwin et al. 2005).  Image data were LOWESSS, spatial, and median normalized (Chari et al. 2006; Khojasteh et al. 2005).  SeeGH software combined duplicate spot 76   data and displayed log2  ratios in relation to genomic locations in the hg17 assembly (NCBI Build 35) (Chi et al. 2004).  Duplicate spots with standard deviations >0.075 or with signal-to-noise ratios <3 were stringently filtered, leaving informative clones for breakpoint analysis (Coe et al. 2006; Baldwin et al. 2005).  Segmental gains and losses were detected and confirmed using three separate algorithms:  aCGH-Smooth, DNACopy, and LSPHMM  (Jong et al. 2004; Olshen et al. 2004; Shah et al. 2006). Altered regions that were concurrently identified by at least two of the three segmentation algorithms were scored: (i) aCGH-Smooth uses a local search algorithm that smoothes the observed array CGH values between consecutive breakpoints to a suitable common value by using maximum likelihood estimation (Jong et al. 2004).  The Lambda and the maximum number of breakpoints in initial pool were set to 6.75 and 100 respectively (Coe et al. 2006).  (ii) DNAcopy employs a circular binary segmentation algorithm that uses a random permutation test to determine the statistical significance of change points (Olshen et al. 2004).  Default settings were used to delineate sections of differing copy number (Chari et al. 2006; Olshen et al. 2004).  (iii) LSPHMM (location-specific priors hidden Markov models) uses a modification of hidden Markov models to exploit prior knowledge about location of copy number polymorphisms which were often detected as outliers in standard HMMs (Shah et al. 2006).  Altered regions were further confirmed by visual inspection.  To avoid false- positives due to hybridization “noise”, only those alterations containing ≥3 overlapping clones were considered as segmental loss.  Whole arm loss was defined as a loss that encompasses all BAC clones from the telomeric to the centromeric end of the 3p arm. A summary of filtering criteria and genetic pattern detected for each sample is listed in Supplementary Table B.2.  Filtering threshold was the same for all except for two samples (Supplementary Table B.2). 3.2.3. Copy number variation (CNV) Detected alterations were filtered for CNVs that had been previously identified using the same array platform (Wong et al. 2007).  The list of CNV-associated BAC clones has been made publicly available on the UCSC Genome Browser and has also been 77   incorporated into LSPHMM software (Shah et al. 2006; Wong et al. 2007).  Examples of CNVs in normal human individuals are presented in Supplementary Fig. B.1. 3.2.4. Statistical analysis Proportion-test was used to test if the number of samples that show 3p losses in HGDs and OSCCs were statistically distinguishable.  Wilcoxon rank-sum test was employed to calculate the probability distributions of the proportion of altered regions between each sample group were different due to chance alone.  A two-tailed Fisher’s exact test was applied to calculate the probability that the observed association for the number of genetic alterations between non-progressing and progressing LGDs, LGDs and HGDs, and HGDs and OSCCs has occurred by chance alone.  The significance threshold was set at p < 0.05.  All statistical tests were two-sided and were performed in the R statistical environment (v2.6.2). 3.2.5. Loss of heterozygosity (LOH) analysis LOH analysis focused on a region previously shown to be associated with increased risk of progression, utilizing microsatellite markers D3S1234, D3S1228, and D3S1300 which are mapped to 3p14.2-3p14.1 (Rosin et al. 2000).  Microsatellite analyses were performed as previously described and LOH was inferred when the signal ratio of the two alleles differed in the normal samples by at least 50% (Rosin et al. 2000). Microsatellite markers that yielded homozygous alleles were termed “non-informative”.  3.3. Results and discussion We hypothesized that genetic alterations critical for tumorigenesis might be found at the premalignant stage of oral lesions.  To test this, we analyzed the 3p arm of 71 OPLs at high resolution to discover changes that occur during the progression of OPLs.  Six discrete, recurrently altered regions were identified in 47 HGDs.  These regions were then compared against 24 LGDs and 23 OSCCs to determine the prevalence of such 78   changes in different stages of progression.  The set of 94 profiles have been deposited to GEO database at NCBI, series accession number GSE9193. 3.3.1. Segmental alterations on chromosome 3p are more frequent than whole arm loss in HGDs We employed three independent breakpoint-detection algorithms on each sample to identify common regions of loss.  Only regions identified by 2/3 algorithms and confirmed by visual inspection were considered to be lost.  In addition, because of the existence of CNVs in the normal human population, previously identified clones associated with copy number polymorphisms were filtered. Our data showed that small regions of 3p were commonly lost in HGDs.  Regions of segmental genetic loss were found in 80.8% of HGDs, with whole arm loss apparent in only a quarter of cases (12/47) and segmental alterations in 55.3% (26/47 cases). Segmental alterations showed great variability in size, ranging from 5.3 to 88.7 Mbp (Fig. 3.1A).  Genetic gains were infrequent, with only two HGDs showing such changes. The number of 3p alterations in HGDs (38/47) and OSCCs (23/23) was not statistically different (difference in proportions: -0.15, 95% confidence interval -0.32 to 0.02, p = 0.19, proportion-test), with this evidence of similarly increased genetic instability perhaps accounting for the relatively high progression risk for HGDs.  The most striking difference in gene alterations between HGDs and OSCCs was the alteration size for each group.  While loss of the whole 3p arm is a common occurrence in OSCCs (Baldwin et al. 2005; Mao et al. 2004), alterations in HGD lesions did not typically encompass the entire chromosome arm.  The size of 3p alterations in OSCCs and HGDs was significantly different (p = 1.84 x 10-5, Wilcoxon rank-sum test). 3.3.2. Regions of loss in HGDs and comparison to OSCCs Those 3p alterations apparent in both HGDs and OSCCs are more likely to represent key genomic changes underlying early oral tumorigenesis.  We identified six minimally altered regions (MARs) that were recurrent in at least 75% of HGD cases that harbored segmental alterations (Fig. 3.1A).  These occurred at 3p25.3-p26.1, 3p25.1-p25.3, 3p24.1, 3p21.31-p22.3, 3p14.2, and 3p14.1 (regions A to F in Fig. 3.1A, Table 3.1). 79   The data of the upper and lower boundary of each MAR is illustrated in Supplementary Fig. B.2.  Most identified MARs harbored <10 genes.  Significantly, alteration of the previously described oral TSG FHIT (Lee et al. 2001) occurred at the same high frequency as the other identified MARs, strongly suggesting the presence of additional candidates driving oral tumorigenesis (Table 3.1, Fig. 3.1C).  Analysis of all OSCCs revealed loss of the entire 3p arm in 19/23 cases, a frequency consistent with previous reports (Baldwin et al. 2005; Smeets et al. 2006; Snijders et al. 2005; Chakraborty et al. 2003; Garnis et al. 2003; Kayahara et al. 2001; Noutomi et al. 2006).  The presence of expanded deletions between these stages indicates the occurrence of secondary genetic events from OPLs to invasive disease.  This also shows the value of analyzing premalignant lesions in order to identify early genetic events that are likely masked by subsequent accumulated alterations. 3.3.3. Evidence of segmental alterations in progressing LGDs We next examined progressing LGDs (9/24, median follow-up time for all cases >7 years) to determine if they accumulate more and/or different genetic alterations than their morphologically similar non-progressing counterparts (15/24).  Segmental losses were significantly associated with progression, occurring in 7/9 progressing LGDs and only 2/15 non-progressing LGDs (p = 0.003, Fisher’s exact test).  These findings are consistent with previous reports of higher frequency of LOH at 3p microsatellite loci in progressing OPLs (Partridge et al. 2000; Rosin et al. 2000; Rosin et al. 2002).  We also observed that the size of segmental losses were larger for progressing LGD cases (53.7 Mbp versus 0.78 Mbp), with the distribution of losses among non-progressing (n = 15) and progressing LGDs (n = 9) being significantly different (p = 6.1 x 10-5, Wilcoxon rank- sum test).  Interestingly, the aberrations in progressing LGDs resemble those of HGDs (which, as already stated, carry a higher likelihood of progression to invasive disease) (Fig. 3.1B).  To examine this further, we evaluated the LGD cases for the alteration status of the six MARs identified in HGDs.  As anticipated, these regions were found to be frequently altered in progressing LGDs but not in non-progressing LGDs (p < 0.003, Fig. 3.1B).  These data lend credence to the conclusion that the MARs identified in HGDs represent critical events in progression to invasive disease. 80   3.3.4. Disrupted genes in dysplastic lesions One hundred forty-one annotated genes were spanned by the six MARs identified above (RefSeq release 25, see listing in Supplementary Table B.3).  Twenty-eight of these are known to be associated with cancer and 17 have functions potentially related to tumor suppressor activity (Table 3.1).  Specifically, nine of these genes have actually been reported as TSGs in cancer (TIMP4, PPARG, WNT7A, TGFBR2, MLH1, CTDSPL, VILL, AXUD1, and FHIT) (Riddick et al. 2005; Haider et al. 2006; Ishiguro et al. 2001; Kashuba et al. 2004; Kujan et al. 2006; Michalik et al. 2004; Sengupta et al. 2007; Winn et al. 2006; You et al. 2007).  The 17 candidate TSGs can be generally categorized to have functions involved in the inhibition of cAMP signaling for the induction of apoptosis (GRM7), DNA-damage repair (RAD18, XPC), autophagy-related and ubiquitin-activating enzyme activity (ATG7), metalloendopeptidase inhibitor activity (TIMP4), transcription repression activity (PPARG), apoptosis (ACVR2B, AXUD1), MAPK signal transduction (SPGAP3, TGFBR2), Wnt pathway-mediated anti-tumorigenic signaling (LRRFIP2, WNT7A, ACVR2b, and AXUD1), and negative regulation of progression through cell cycle (MLH1, DLEC1, FHIT).  Specifically, GRM7 was shown to be methylated in chronic lymphocytic leukemia (Rush et al. 2004), whereas TGFBR2 is transcriptionally- repressed in human epithelial tumors (Bierie & Moses 2006). Our candidate MARs are in agreement with previous studies of OPLs.  Regions A, B, D, E, and F coincide with reported allelic losses identified by microsatellite analysis or copy number loss by conventional CGH (Califano et al. 1996; Partridge et al. 2000; Rosin et al. 2000; Chakraborty et al. 2003; Kayahara et al. 2001; Noutomi et al. 2006).  The improved resolution in the current work has further fine-mapped these regions and allows us to name specific gene candidates (Noutomi et al. 2006).  The frequent deletion of 3p24.1 has been delineated to 1.21 Mbp, highlighting the potential importance of TGFBRII in oral tumorigenesis (Table 3.1). 3.3.5. Sequential 3p deletions during OPL development We examined the 3p arm for frequency of alteration, genetic pattern, and size of change during the multistep oral tumorigenesis.  The increased frequency of 3p loss from LGDs 81   (37.5%) to HGDs (80.8%) was significant (p = 4.6 x 10-4, Fisher’s exact test).  Also, specific genetic patterns were associated with each histological group (Fig. 3.2).  Whole chromosome arm changes were observed at much higher frequency for OSCCs tumors (Table 3.2), with the mean deletion size increasing significantly with disease stage; the average deletion was 41.9 Mbp in LGDs, 67.5 Mbp in HGDs, and 83.5 Mbp in OSCCs (p = 0.016 for LGDs versus HGDs and p = 3.6 x 10-4 for HGDs versus OSCCs, Wilcoxon rank-sum test).  Genetic changes on 3p are sequentially associated with increased progressing stages, as most LGDs have no loss, HGDs have segmental losses, and OSCCs have whole arm loss.  These data clearly delineate escalation of genomic instability at chromosome 3p with histopathological disease stage.  (No associations between age or gender were observed for our genomic data, and likewise, smoking status was not found to influence the level of genetic instability on the 3p arm in our dataset (Supplementary Table B.4)). 3.3.6. Integrating LOH data with copy number alterations DNA copy number loss can be detected by array CGH, while segmental loss or uniparental disomy can be identified by LOH analyses.  To explore the relative contributions of each kind of genetic dysregulation to OPLs and OSCCs, we integrated copy number data with informative LOH data for 3p14.1-p14.2 (available for 88/94 case).  Concordant LOH and copy number loss were observed for 32.9% of these cases, with a further 30.7% exhibiting copy number neutral LOH and 12.5% showing copy number loss not detected by LOH (Supplementary Fig. B.3).  Complementary integrative analysis by these two approaches facilitates a more complete portrait of genome dysregulation in oral cancer progression and could represent an even more powerful tool for delineating progression risk in oral premalignant lesions.  3.4. Summary Alteration of chromosome arm 3p has previously been reported as a frequent event for oral cancer and losses of specific loci within this region have been associated with 82   progression risk in early OPLs.  To date, however, there have been no high resolution analyses of this chromosome arm to identify additional candidates that may drive oral tumorigenesis.  By undertaking tiling-path array CGH analysis for a panel of nearly a hundred OPLs and oral tumors, we have confirmed that genetic instability increases on this single arm with progression to invasive cancer, identified new candidate TSGs that may contribute to oral cancer progression, and demonstrated that low grade dysplasias known to progress to invasive disease harbor gene alterations very similar to those observed in higher grade dysplastic lesions.  This study identified specific regions altered on chromosome 3p that are commonly found in the premalignant stages of oral cancer development but are often masked in invasive tumors due to increased genetic instability. Figure 3.1 83 A Oral10 Oral40 Oral12 N12Oral28 -0.5 0.5 -0.5 0.5 -0.5 0.5 -0.5 0.5 -0.5 0.5 C B 3p arm Telomere Centromere A B C D E F 72% 0% 100% 67% 72% 13% 96% 78% 68% 0% 96% 75% 0% 96% 56% 68% 0% 91% 56% 56% 70% 0% 91% 56% OSCCs HGDs Progressing LGDs Non-progressing LGDs 100%0% 50% FH IT 3p26.1 3p25.3 3p24.1 3p24.3 3p23 3p21.31 3p14.2 3p13 3p22.1 3p14.3 3p12.2 3p26.3 2.09 Mb 4.20 Mb 1.04 Mb 10.05 Mb 2.89 Mb 1.44 Mb A B C D E F Figure 3.1  Frequency of alterations on the 3p arm.  A, summary of copy number changes in 26 HGDs that showed segmental alterations on 3p.  Copy number loss is presented as vertical lines on the left side of the ideogram, with vertical lines on the right side indicating copy number gain.  For each sample, altered regions that were concurrently identified by at least two of the three methods were scored and confirmed by visual inspection.  The top six minimal altered regions (MARs) are boxed in orange. B, frequency of alteration of the six MARs in 23 OSCCs (black), 47 HGDs (purple), and 24 LGDs (progressing LGDs in light purple and non-progressing LGDs in white).  The six MARs were found to be common in progressing LGDs but infrequent in non- progressing LGDs (p < 0.003, Fisher’s exact test).  C, Alignment of four HGDs reveals segmental loss of the FHIT locus at 3p14.2.  Each bar represents a BAC-derived segment.  Bars to the left and right of centre line represent DNA copy number losses and gains, respectively.  The green and red lines to the left and right side of the central line represent signal intensity log2 ratios of -0.5 and +0.5 respectively.  Paired normal (N12) and HGD (Oral12) samples are shown in the blue box. Figure 3.2 84 LGDs (n= 24) HGDs (n= 47) OSCCs (n= 23) 0 10 20 30 40 50 60 70 80 90 100 W ho le  a rm  l o ss S eg m en ta l l o ss S eg m en ta l g ai n To ta l c ha ng e LGDs (n= 24) HGDs (n= 47) OSCCs (n= 23) Pe rc en ta g e Histological group Genetic pattern 0.025p = p=  4.60 x 10-4 Figure 3.2.  Genetic changes observed in each histological group on the 3p arm. Significant differences in the frequency of 3p alterations between LGDs and HGDs (p = 4.60 x 10-4) and between HGDs and OSCCs (p = 0.025) were observed.  85   Table 3.1.  Summary of recurrent minimal altered regions identified in 47 high- grade dysplasias using 3p tiling-path array CGH. Region Chromosom e Band Proximal flanking clone* Start bp coordinate Distal flanking clone* End bp coordinat e Size (Mbp) Genes† Genes associated with cancer‡ A 3p25.3-p26.1 776H3 7152555 629B9 9237884 2.09 7 GRM7, RAD18, SRGAP3 B 3p25.1-p25.3 525N21 10413544 464P4 14696702 4.28 28 ATG7, TIMP4, PPARG, WNT7A, XPC, RAF1 C 3p24.1 775G14 30598681 48 E16 31812073 1.21 4 TGFBR2 D 3p21.31- p22.3 606K24 36367289 494P19 46609837 10.24 95 MLH1, LRRFIP2, ITGA9, CTDSPL, PLCD1, DLEC1, MyD88, VILL, ACVR2B, GORASP1, AXUD1, CX3CR1, CCR8, CTNNB1, CCK, CCBP2, TMEM158 E 3p14.2 638K20 59594324 126N04 62484320 2.89 5 FHIT F 3p14.1 413B3 67349894 259O22 68785539 1.44 2  - *All the listed human BAC clones were selected from the RPCI-11 library. †RefSeq Genes (release 25) according to UCSC May 2004 assembly. ‡Genes bolded are candidate TSGs in various cancer types. 86   Table 3.2.  Pattern of 3p alterations in oral dysplasias and OSCCs.   Whole arm loss* Segmental alterations* No genetic change* Mean size of loss (Mbp) LGDs† (n=24) 0 37.5% (9) 62.5% (15) 41.9 Non-progressing LGDs† (n=15) 0 13.3% (2) 86.7% (13) 0.78 Progressing LGDs†  0 77.8% (7) 22.2% (2) 53.7 (n=9) HGDs (n=47) 25.5% (12) 55.3% (26) 19.1% (9) 67.5 OSCCs (n=23) 82.6% (19) 17.4 % (4) 0 83.5 *Number of cases are given in parentheses. †LGDs are categorized into non-progressing and progressing LGDs (see Materials and Methods) 87    3.5. References Baldwin, C., C. Garnis, L. Zhang, M. P. Rosin & W. L. Lam, 2005. Multiple microalterations detected at high frequency in oral cancer. Cancer Res, 65(17), 7561-7. Bierie, B. & H. L. Moses, 2006. Tumour microenvironment: TGFbeta: the molecular Jekyll and Hyde of cancer. Nat Rev Cancer, 6(7), 506-20. Califano, J., P. van der Riet, W. Westra, H. Nawroz, G. Clayman, S. Piantadosi, R. Corio, D. Lee, B. Greenberg, W. Koch & D. Sidransky, 1996. Genetic progression model for head and neck cancer: implications for field cancerization. Cancer Res, 56(11), 2488-92. Chakraborty, S. B., S. Dasgupta, A. Roy, A. Sengupta, B. Ray, S. Roychoudhury & C. K. Panda, 2003. Differential deletions in 3p are associated with the development of head and neck squamous cell carcinoma in Indian patients. Cancer Genet Cytogenet, 146(2), 130-8. Chari, R., W. W. Lockwood & W. L. Lam, 2006. Computational Methods for the Analysis of Array Comparative Genomic Hybridization. Cancer Inform, 2, 48-58. Chi, B., R. J. DeLeeuw, B. P. Coe, C. MacAulay & W. L. Lam, 2004. SeeGH--a software tool for visualization of whole genome array comparative genomic hybridization data. BMC Bioinformatics, 5, 13. Coe, B. P., W. W. Lockwood, L. Girard, R. Chari, C. Macaulay, S. Lam, A. F. Gazdar, J. D. Minna & W. L. Lam, 2006. Differential disruption of cell cycle pathways in small cell and non-small cell lung cancer. Br J Cancer, 94(12), 1927-35. Garnis, C., C. Baldwin, L. Zhang, M. P. Rosin & W. L. Lam, 2003. Use of complete coverage array comparative genomic hybridization to define copy number 88   alterations on chromosome 3p in oral squamous cell carcinomas. Cancer Res, 63(24), 8582-5. Garnis, C., T. P. Buys & W. L. Lam, 2004. Genetic alteration and gene expression modulation during cancer progression. Mol Cancer, 3, 9. Haider, A. S., S. B. Peters, H. Kaporis, I. Cardinale, J. Fei, J. Ott, M. Blumenberg, A. M. Bowcock, J. G. Krueger & J. A. Carucci, 2006. Genomic analysis defines a cancer-specific gene expression signature for human squamous cell carcinoma and distinguishes malignant hyperproliferation from benign hyperplasia. J Invest Dermatol, 126(4), 869-81. Ishiguro, H., T. Tsunoda, T. Tanaka, Y. Fujii, Y. Nakamura & Y. Furukawa, 2001. Identification of AXUD1, a novel human gene induced by AXIN1 and its reduced expression in human carcinomas of the lung, liver, colon and kidney. Oncogene, 20(36), 5062-6. Ishkanian, A. S., C. A. Malloff, S. K. Watson, R. J. DeLeeuw, B. Chi, B. P. Coe, A. Snijders, D. G. Albertson, D. Pinkel, M. A. Marra, V. Ling, C. MacAulay & W. L. Lam, 2004. A tiling resolution DNA microarray with complete coverage of the human genome. Nat Genet, 36(3), 299-303. Jong, K., E. Marchiori, G. Meijer, A. V. Vaart & B. Ylstra, 2004. Breakpoint identification and smoothing of array comparative genomic hybridization data. Bioinformatics, 20(18), 3636-7. Kashuba, V. I., J. Li, F. Wang, V. N. Senchenko, A. Protopopov, A. Malyukova, A. S. Kutsenko, E. Kadyrova, V. I. Zabarovska, O. V. Muravenko, A. V. Zelenin, L. L. Kisselev, I. Kuzmin, J. D. Minna, G. Winberg, I. Ernberg, E. Braga, M. I. Lerman, G. Klein & E. R. Zabarovsky, 2004. RBSP3 (HYA22) is a tumor suppressor gene implicated in major epithelial malignancies. Proc Natl Acad Sci U S A, 101(14), 4906-11. 89   Kayahara, H., H. Yamagata, H. Tanioka, T. Miki & H. Hamakawa, 2001. Frequent loss of heterozygosity at 3p25-p26 is associated with invasive oral squamous cell carcinoma. J Hum Genet, 46(6), 335-41. Khojasteh, M., W. L. Lam, R. K. Ward & C. MacAulay, 2005. A stepwise framework for the normalization of array CGH data. BMC Bioinformatics, 6, 274. Kujan, O., R. Oliver, L. Roz, G. Sozzi, N. Ribeiro, R. Woodwards, N. Thakker & P. Sloan, 2006. Fragile histidine triad expression in oral squamous cell carcinoma and precursor lesions. Clin Cancer Res, 12(22), 6723-9. Lee, J. I., J. C. Soria, K. Hassan, D. Liu, X. Tang, A. El-Naggar, W. K. Hong & L. Mao, 2001. Loss of Fhit expression is a predictor of poor outcome in tongue cancer. Cancer Res, 61(3), 837-41. Mao, L., W. K. Hong & V. A. Papadimitrakopoulou, 2004. Focus on head and neck cancer. Cancer Cell, 5(4), 311-6. Mao, L., J. S. Lee, Y. H. Fan, J. Y. Ro, J. G. Batsakis, S. Lippman, W. Hittelman & W. K. Hong, 1996. Frequent microsatellite alterations at chromosomes 9p21 and 3p14 in oral premalignant lesions and their value in cancer risk assessment. Nat Med, 2(6), 682-5. Michalik, L., B. Desvergne & W. Wahli, 2004. Peroxisome-proliferator-activated receptors and cancers: complex stories. Nat Rev Cancer, 4(1), 61-70. Noutomi, Y., A. Oga, K. Uchida, M. Okafuji, M. Ita, S. Kawauchi, T. Furuya, Y. Ueyama & K. Sasaki, 2006. Comparative genomic hybridization reveals genetic progression of oral squamous cell carcinoma from dysplasia via two different tumourigenic pathways. J Pathol, 210(1), 67-74. Olshen, A. B., E. S. Venkatraman, R. Lucito & M. Wigler, 2004. Circular binary segmentation for the analysis of array-based DNA copy number data. Biostatistics, 5(4), 557-72. 90   Partridge, M., S. Pateromichelakis, E. Phillips, G. G. Emilion, R. P. A'Hern & J. D. Langdon, 2000. A case-control study confirms that microsatellite assay can identify patients at risk of developing oral squamous cell carcinoma within a field of cancerization. Cancer Res, 60(14), 3893-8. Riddick, A. C., C. J. Shukla, C. J. Pennington, R. Bass, R. K. Nuttall, A. Hogan, K. K. Sethia, V. Ellis, A. T. Collins, N. J. Maitland, R. Y. Ball & D. R. Edwards, 2005. Identification of degradome components associated with prostate cancer progression by expression analysis of human prostatic tissues. Br J Cancer, 92(12), 2171-80. Rosin, M. P., X. Cheng, C. Poh, W. L. Lam, Y. Huang, J. Lovas, K. Berean, J. B. Epstein, R. Priddy, N. D. Le & L. Zhang, 2000. Use of allelic loss to predict malignant risk for low-grade oral epithelial dysplasia. Clin Cancer Res, 6(2), 357- 62. Rosin, M. P., W. L. Lam, C. Poh, N. D. Le, R. J. Li, T. Zeng, R. Priddy & L. Zhang, 2002. 3p14 and 9p21 loss is a simple tool for predicting second oral malignancy at previously treated oral cancer sites. Cancer Res, 62(22), 6447-50. Rush, L. J., A. Raval, P. Funchain, A. J. Johnson, L. Smith, D. M. Lucas, M. Bembea, T. H. Liu, N. A. Heerema, L. Rassenti, S. Liyanarachchi, R. Davuluri, J. C. Byrd & C. Plass, 2004. Epigenetic profiling in chronic lymphocytic leukemia reveals novel methylation targets. Cancer Res, 64(7), 2424-33. Sengupta, S., S. Chakrabarti, A. Roy, C. K. Panda & S. Roychoudhury, 2007. Inactivation of human mutL homolog 1 and mutS homolog 2 genes in head and neck squamous cell carcinoma tumors and leukoplakia samples by promoter hypermethylation and its relation with microsatellite instability phenotype. Cancer, 109(4), 703-12. Shah, S. P., X. Xuan, R. J. DeLeeuw, M. Khojasteh, W. L. Lam, R. Ng & K. P. Murphy, 2006. Integrating copy number polymorphisms into array CGH analysis using a robust HMM. Bioinformatics, 22(14), e431-9. 91   Smeets, S. J., B. J. Braakhuis, S. Abbas, P. J. Snijders, B. Ylstra, M. A. van de Wiel, G. A. Meijer, C. R. Leemans & R. H. Brakenhoff, 2006. Genome-wide DNA copy number alterations in head and neck squamous cell carcinomas with or without oncogene-expressing human papillomavirus. Oncogene, 25(17), 2558-64. Snijders, A. M., B. L. Schmidt, J. Fridlyand, N. Dekker, D. Pinkel, R. C. Jordan & D. G. Albertson, 2005. Rare amplicons implicate frequent deregulation of cell fate specification pathways in oral squamous cell carcinoma. Oncogene, 24(26), 4232-42. Winn, R. A., M. Van Scoyk, M. Hammond, K. Rodriguez, J. T. Crossno, Jr., L. E. Heasley & R. A. Nemenoff, 2006. Antitumorigenic effect of Wnt 7a and Fzd 9 in non-small cell lung cancer cells is mediated through ERK-5-dependent activation of peroxisome proliferator-activated receptor gamma. J Biol Chem, 281(37), 26943-50. Wong, K. K., R. J. deLeeuw, N. S. Dosanjh, L. R. Kimm, Z. Cheng, D. E. Horsman, C. MacAulay, R. T. Ng, C. J. Brown, E. E. Eichler & W. L. Lam, 2007. A comprehensive analysis of common copy-number variations in the human genome. Am J Hum Genet, 80(1), 91-104. You, K. T., L. S. Li, N. G. Kim, H. J. Kang, K. H. Koh, Y. J. Chwae, K. M. Kim, Y. K. Kim, S. M. Park, S. K. Jang & H. Kim, 2007. Selective translational repression of truncated proteins from frameshift mutation-derived mRNAs in tumors. PLoS Biol, 5(5), e109. Zabarovsky, E. R., M. I. Lerman & J. D. Minna, 2002. Tumor suppressor genes on chromosome 3p involved in the pathogenesis of lung and other cancers. Oncogene, 21(45), 6915-35. 92   Chapter 4.  Multiple pathways in the FGF signaling network are frequently deregulated by gene amplification in oral dysplasias3   3 A version of this chapter has been published. Tsui IFL, Poh CF, Garnis C, Rosin MP, Zhang L, Lam WL. (2009) Multiple pathways in the FGF signaling network are frequently deregulated by gene amplification in oral dysplasias. International Journal of Cancer. 125: 2219-28. Please see appendix C for all supplementary materials of this chapter. 93   4. Multiple pathways in the FGF signaling network are frequently deregulated by gene amplification in oral dysplasias 4.1. Introduction Oral squamous cell carcinoma (OSCC) is one of the most common head and neck neoplasm, with more than 30,000 cases identified in the United States each year.(Lydeard et al. 2007)  Despite advances in treatment, the 5-year survival rate of all stages and advanced stage (stage III and IV) remains at less than 50% and 25%, respectively, for the past three decades.(Epstein et al. 2008)  The poor survival rate is mainly because most patients were diagnosed at the advanced stages of the disease. Early detection and treatment at the premalignant stages with intensified follow-up would improve patient survival.(Epstein et al. 2008; Poh et al. 2008) Oral cancer is believed to progress from hyperplasia, through various extent of dysplastic changes, carcinoma in situ (CIS), and finally breaking through the basement membrane at the invasive SCC stage.(Warnakulasuriya et al. 2008)  In oral premalignant lesions (OPLs), the presence and degree of epithelial dysplasia is used to assess the risk for progression to malignancy.  By definition, dysplasia is characterized by cellular atypia and loss of normal maturation and stratification and has no evidence of invasion.(1978)  High-grade preinvasive lesion, including severe dysplasia and CIS, is associated with the strongest risk for malignant transformation.(Poh et al. 2008; Crissman & Zarbo 1989; Fresko & Lazarus 1981; Hayward & Regezi 1977; Summerlin 1996)  Thus, patients with these lesions are treated in British Columbia to prevent further development into invasive SCC.(Poh et al. 2008)  However, the majority of the low-grade dysplasia, mild and/or moderate dysplasia, will not progress to cancer. Previous studies have started integrating imaging and molecular analysis with histopathologic evaluation for improving the ability to predict progression risk in OPLs.(Hall et al. 2008; Guillaud et al. 2008; Mao et al. 1996; Partridge et al. 2000; Rosin et al. 2000; Rosin et al. 2002)  An understanding of the molecular mechanisms that govern the promotion of OPLs into cancer would be very relevant for clinical 94   practice in the identification of genes suitable for therapeutic targeting, prognostic and risk predictive markers. In oral cancer, the accumulation of genetic alterations is associated with the progression of OPL to invasiveness.(Garnis et al. 2004; Mao et al. 2004)  Molecular analysis of OPLs is necessary to identify key changes in disease initiation and progression. However, studies of OPLs are rare and have not been attempted on a genome-wide scale, owing mostly to the minute amount of DNA obtainable from primary OPLs.(Garnis et al. 2004)  Changes in gene dosage occur frequently in cancer genomes.  Gains and losses of genomic regions may contain proto-oncogenes and tumor suppressor genes, which may lead to aberrant expression useful for malignant transformation.  Low-level copy number changes involving large regions with many genes have been frequently observed in OSCCs, but their effect on gene expression remains ambiguous.(Baldwin et al. 2005; Noutomi et al. 2006; Smeets et al. 2006; Snijders et al. 2005)  High-level copy number changes, DNA amplification or homozygous deletion, encompass focal changes and often lead to the discovery of cancer-causing genes.  Amplicons are defined as DNA segments less than 20 megabase pairs (Mbps) of which at least five copies exist in a single cell.(Myllykangas et al. 2007)  They are useful for oncogene discovery because these unstable regions are under relentless selection and thus harbor genes advantageous for tumor growth.(Albertson 2006; Myllykangas et al. 2008; Miele et al. 1989)  Amplified oncogenes are also clinically useful for therapeutic development.(Albertson 2006; Myllykangas et al. 2007)  Furthermore, amplification of EGFR with more than 12 copies per cell in head and neck SCC (HNSCC) is associated with poor survival.(Temam et al. 2007)  On the other hand, biallelic loss such as homozygous deletion contributes to functional inactivation, facilitating the localization of tumor suppressors.(Hahn et al. 1996; Kikuchi et al. 2008)  High-resolution whole genome DNA microarrays have been useful in exploring genetic alterations in formalin- fixed paraffin-embedded (FFPE) specimens without the need for sample amplification.(Ishkanian et al. 2004; Baldwin et al. 2005)  Indeed this technology has a functional resolution of 50 kbp, improving the localization of regions with gene amplification and homozygous deletion.(Coe et al. 2007; Ishkanian et al. 2004) 95   Oral high-grade dysplasias are known to have a high likelihood of cancer progression, whereas low-grade lesions have a low probability of progression and can even regress.(Rosin et al. 2000)  We evaluated genome wide gene dosage alteration in 50 manually microdissected FFPE high-risk OPLs, which included 43 high-grade dysplasias and seven low-grade dysplasias that are known to have later progressed to cancer.  We focused on the identification of high-level DNA amplification and homozygous deletions instead of low-level copy number change.  This type of analysis has never been undertaken for such early stage lesions.  We also compared these findings to 23 OSCC and 14 low-grade dysplasias that never progressed.  Expression of gene candidates within recurrent amplicons in OPLs were analyzed in public datasets with 188 HNSCCs and further confirmed in 61 oral cancers.  Taken together, our analysis suggests that a common signaling network involving the ERK/MAPK, FGF, p53, PTEN, and PI3K/AKT signaling pathways is frequently deregulated in high-risk OPLs.  4.2. Materials and methods 4.2.1. Tissue samples This study involves 87 (64 dysplasias and 23 OSCCs) archival FFPE specimens obtained from the British Columbia Oral Biopsy Service (Supp. Table C.1).  The group of "high-risk OPLs" includes 43 high-grade preinvasive lesions (22 severe dysplasia and 21 CIS) and seven low-grade lesions (one hyperplasia and six mild and/or moderate dysplasias) that later progressed to high-grade dysplasias or OSCCs.  These lesions represent 86 patients, all with no prior history of cancer.  Samples from one patient with a severe dysplasia on the lower lip (sample Oral7) and an OSCC on the tongue (sample Oral80) are both included in the study as they are from distinct anatomical sites.  For low-grade lesions, patients were followed up with a median duration of eight years in a longitudinal study established at the BC Oral Cancer Prevention Program.  Low-grade lesions that did not progress between 1985 and 2009 consisted of one hyperplasia and 13 mild and moderate dysplasias.  All diagnoses were confirmed by the study 96   pathologist (LZ) using criteria established by the World Health Organization (WHO).(1978)  Areas of dysplasia were identified using hematoxylin and eosin (H&E) stained sections cut from FFPE tissues.  Epithelial cells in these areas were meticulously dissected from adjacent non-epithelium tissue under an inverted microscope using a 23G needle.  DNA was extracted as previously described.(Baldwin et al. 2005) 4.2.2. Whole genome DNA microarray analysis Tiling-path genomic arrays (SMRT v.1 and v.2) were obtained from the BC Cancer Research Centre Array Laboratory.(Ishkanian et al. 2004)  The whole genome is represented as 26,819 overlapping bacterial artificial chromosome (BAC) clones spotted in duplicate with complete coverage of the human genome, allowing breakpoint detection at a resolution of 50 kbp.(Coe et al. 2007; Ishkanian et al. 2004)  Briefly, each sample DNA and normal reference pooled male genomic DNA (Novagen, Mississauga, ON, Canada) (250 ng each) were random prime labeled with cyanine-3 and cyanine-5 dCTP, respectively, mixed with 100 µg of human Cot-1 DNA, purified and hybridized to the array at 45ºC for 36 hours before washing.  Hybridized arrays were scanned as previously described.(Lockwood et al. 2008; Coe et al. 2006) 4.2.3. Data analysis Array images were analyzed using SoftWoRx Tracker Spot Analysis software (Applied Precision, Issaquah, WA).  A three-step normalization procedure, including LOWESS fitting, spatial, and median normalization, was used to remove systematic biases.(Khojasteh et al. 2005)  SeeGH software was used to display log2  signal intensity ratios in relation to genomic locations in the hg17 assembly (NCBI Build 35).(Chi et al. 2008)  Data points with standard deviation >0.075 and signal to noise ratio <3 in either channel were removed.  None of the 87 genomic profiles contain technical artefacts of wavy pattern which are often observed from FFPE samples.(van de Wiel et al. 2009)  All profiles has been deposited to Gene Expression Omnibus (GEO) database at NCBI, series accession number GSE9193.(Tsui et al. 2008) 97   High-level DNA amplifications and presumptive homozygous deletions were identified by a moving-average based algorithm as previously described.(Lockwood et al. 2008) The threshold for high copy number was set to log2 signal intensity ratio > 0.8 for amplification or < -0.8 for homozygous deletions.  Only those alterations containing ≥ 3 overlapping clones were identified in order to avoid false-positives due to hybridization artifacts.  Recurrent minimal altered regions of amplification were identified by the presence of a given amplicon in at least two high-risk OPLs.  Genes within such altered regions were mapped according to the RefSeq Genes track release 25.  As copy number variation-associated BAC clones are often associated with amplification and DNA rearrangement, such BACs are included in the analysis.(Lockwood et al. 2008; Wong et al. 2007) 4.2.4. Transcript expression analysis Independent transcript analyses of genes mapped within minimal altered regions of amplification were performed using the Oncomine database.(Rhodes et al. 2004)  Five studies within Oncomine analyzed expression patterns between head and neck tumors (N = 188) and normal tissues (N = 38) (Supp. Table C.2).(Cromer et al. 2004; Ginos et al. 2004; Pyeon et al. 2007; Chung et al. 2004; Toruner et al. 2004)  In Oncomine, Student’s t test was performed to reflect the significance of differential expression observed in tumors compared to normal tissue in each study.  Furthermore, two public GEO datasets (GSE10121 and GSE9844) containing 61 oral tumors and 18 normal samples were downloaded.(Ye et al. 2008; Toruner et al. 2004; Pyeon et al. 2007)  The normalized data from each dataset were extracted.  Two-sided student’s t-test along with a Benjamini-Hochberg multiple testing correction was performed comparing the oral tumors with the normal samples per study.  The significance threshold was set at P<0.05. 4.2.5. Real-time polymerase chain reaction Total RNA from eight OSCCs and nine normal oral mucosal tissue from different healthy individuals were extracted using TRIzol (Invitrogen, Carlsbad, CA).  Five hundred nanograms of total RNA from each sample were converted to cDNA using the High- 98   Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA) in a total volume of 20 µL.  Real-time PCR using TaqMan Universal PCR master mix was performed to analyze TLN1 and CREB3 expression levels with Applied Biosystems Fast Real-Time PCR System.  TaqMan gene expression assays (Assay ID) of TLN1 (Hs00196775_m1), CREB3 (Hs00197255_m1), and 18S rRNA (Hs99999901_s1) were purchased from Applied Biosystems.  Reactions were performed in triplicate and according to manufacturer's protocol.  The 2-∆∆Ct method was used to calculate relative expression values using the average of cycle thresholds of target genes and 18S rRNA, and the value of 1 was arbitrarily assigned to one normal sample.  Expression levels between OSCCs and normal samples were compared by a two-sided Wilcoxon-rank sum test. 4.2.6. Biological functions and pathway analysis To define the biological functions of all the genes mapped within recurrent amplicons in OPLs, we interrogated the Biological Function Analysis in Ingenuity Pathway Analysis (IPA) (version 7.0) (Ingenuity® Systems, www.ingenuity.com).  Biological Function Analysis identifies biological functions and diseases that are significantly enriched in the data relative to chance alone by Fisher’s exact test.  In addition, genes in recurrent amplicons in high-risk OPLs and associated with transcript overexpression in at least one HNSCC dataset were further explored using the Canonical Pathways Analysis. Canonical Pathways Analysis explores 48 well-characterized metabolic and signaling pathways for the significant enrichment of the dataset in these pathways, again using Fisher’s exact test to calculate the probability that the association between genes in the dataset and the canonical pathway is explained by chance alone. 4.2.7. Fluorescence in situ hybridization (FISH) FISH assays were performed as described in Romeo et al.,(Romeo et al. 2003)  except with the modifications of using a lower concentration of pepsin (0.032%) and longer digestion time (80-90 minutes).  Two sets of dual colored probe (Vysis, Downers Grove, IL) were sequentially performed on the same tissue section according to manufacturer’s instruction, which included the pair of CEP11 (centromere 11p11.11-q11, 99   SpectrumGreen)/ CCND1 (11q13, SpectrumOrange), and CEP7 (7p11.1-q11.1, SpectrumGreen)/ EGFR (7p12, SpectrumOrange).  Signals were captured and imaged using Olympus BX61 and ImagePro Plus 5.1.  4.3. Results 4.3.1. Early occurrence of DNA amplification and homozygous deletion in OPLs It is evident that copy number alterations are more frequent among OSCCs relative to high-grade dysplasias across the whole genome (Supp. Figure C.1).  High-level copy number alteration, including DNA amplification and homozygous deletion, is a frequent event in high-risk OPLs.  In total, 40% of high-risk OPLs (19/43 high-grade dysplasias and 1/7 progressing low-grade dysplasias) exhibited at least one region of high-level copy number change.  A frequency of 65.2% (15/23) was found in OSCCs.  No such changes were detected in low-grade lesions that did not progress.  A whole genome karyogram of one high-grade dysplasia Oral42 with seven regions of gene amplification is illustrated in Figure 4.1. 4.3.2. Recurrent amplicons and rare regions of homozygous deletion harbor known and novel candidate cancer genes To distinguish genetic events at the premalignant stage from late-stage events, high- level copy number changes were identified separately in 50 high-risk OPLs and 23 OSCCs.  In the 20/50 high-risk OPLs with high-level copy number alteration, 43 incidents of gene amplification and six regions of homozygous deletion were detected. At a similar level, 46 occurrences of amplification and two regions of homozygous deletion were identified in 15/23 OSCCs (Supp. Table C.3).  Many of these lesions have at least two regions of high-level copy number alteration, including 11/20 of the high-risk OPLs and 9/15 of the OSCCs.  In addition, many of the detected amplicons overlap, suggesting that these regions do not occur by chance. 100   Homozygous deletions were seen less frequently than DNA amplification, occurring in five high-grade dysplasias and one OSCC (Fig. 4.2).  The eight identified regions of homozygous deletion do not overlap (Table 4.1), but two regions on 9p22.3 (in samples Oral12 and Oral88) are separated by 1.2 Mbp.  In total, 44 genes were identified as homozygously deleted, including known tumor suppressors CDKN2A (9p21.3), CDKN2B (9p21.3), MTAP (9p21.3), and WWOX (16q23.1).  Genes bounded by the novel regions of homozygous deletion that we identified (9p21.1-p21.2, 9p22.3, 9p22.3- p23, 9q33.1, 9q33.1-q33.2, and 15q15.1) may represent tumor suppressors driving oral carcinogenesis (Supp. Table C.4). DNA amplifications occurring in OPLs may activate genes that facilitate the development of oral cancer.  Seven recurrent amplicons, ranging from 0.45 Mbp to 2.26 Mbp in size, were identified in our dataset of 50 high-risk OPLs (Table 4.2).  Minimal altered regions of these recurrent amplicons contain 88 unique genes, 15 of which are known to be involved in cancer as determined by IPA Functional Analysis (P = 1.31 x 10-5—1.41 x 10-2) (Table 4.3, Supp. Table C.5).  Except for chromosomal loci at 2q11.2 and 8q22.3, all amplicons harbor at least one cancer-related gene according to the IPA knowledgebase (though literature searches for genes in the outstanding regions identified potential cancer genes) (Table 4.2).  This further substantiates the need to investigate the genes within these recurrent amplicons in OPLs.. 4.3.3. Transcript analysis of independent HNSCC datasets Candidate oncogenes within amplicons are likely to have increased mRNA expression in cancer specimens.  Thus, we evaluated transcript levels of genes in the recurrent amplicons to refine the gene list using five independent studies of HNSCC.(Chung et al. 2004; Cromer et al. 2004; Ginos et al. 2004; Pyeon et al. 2007; Toruner et al. 2004)  Of the 88 genes identified within recurrent amplicons, 40 were overexpressed in at least one dataset of HNSCC relative to normal tissues.  This validation at the transcript level in independent datasets suggests the potential importance of the 40 candidate genes in cancer development. 101   4.3.4. Frequent oncogenic activation of a common signaling network in OPLs We hypothesize that genes with elevated copy number in recurrent OPLs and increased mRNA levels in HNSCC are important for oral cancer development.  To understand the signaling defects in OPLs, we interrogated 48 well-characterized signaling pathways in the IPA canonical pathway database to examine which pathways are significantly enriched with the candidate genes.  The top five deregulated canonical pathways include the ERK/MAPK, FGF, p53, PTEN, and PI3K/AKT signaling pathways (Table 4.4, P = 8.95x10-3, 1.63x10-2, 1.96x10-2, 1.96x10-2, and 3.18x10-2 respectively).  Graphical representation of each canonical pathway is provided in Supp. Figures C.2, C.3, C.4, C.5, C.6. It is important to note that several genes within recurrent amplicons, including the CREB3, CCND1, and YWHAZ, participate in multiple canonical pathways.  In addition, amplified genes including the FGFR and EGFR could both activate the ERK/MAPK and the PI3K/AKT pathways.  Furthermore, 14 amplified genes share direct and indirect relationships in one network (Supp. Fig. C.7).  As multiple pathways could contribute to cancer development, the interactions among the top significantly over-represented canonical pathways were considered as a network (Fig. 4.3).  Seven candidate genes (FGF3, EGFR, TLN1, YWHAZ, CCND1, CREB3 and SNAI2) within recurrent amplicons in high-risk OPLs participate in this signaling network.  Of these seven candidate genes, FGF3, EGFR, and CCND1 have been frequently associated with gene amplifications in OSCCs,(Sheu et al. 2009; Freier et al. 2006; Lese et al. 1995) and protein expression of YWHAZ has been detected in oral dysplasias.(Ralhan et al. 2009); while SNAI2 is often described as an important regulator for epithelial-mesenchymal transition.(Cobaleda et al. 2007)  We validated expression levels of TLN1 and CREB3 by quantitative PCR in eight OSCCs and nine normal oral mucosa tissues.  Expression of TLN1 and CREB3 both showed increased expression in the OSCC group relative to normal tissues (p = 4.53x10-3, 8.14x10-3, respectively, Wilcoxon rank-sum test) (Supp. Fig. C.8).  Genes within this signaling network not identified in recurrent amplicons include PAK4, FGFR1, 102   and EIF4EBP1.  PAK4 is amplified in one high-grade dysplasia and one OSCC, while FGFR1 and EIF4EBP1were amplified in one high-grade dysplasia.  Taken together, gene amplification of at least one of these 10 genes within this signaling network was found in 30.0% (15/50) of high-risk OPLs and 43.5% (10/23) of OSCCs. To further substantiate the importance of this signaling network, we investigated two genetic features in all 87 oral lesions: 1) the presence of multiple amplicons harboring genes of this signaling network within a single specimen, and 2) the presence of overexpression in members of this signaling network that are not deregulated by gene amplification.  We found that three high-grade dysplasias and one OSCC maintained multiple regions of gene amplification of different genes of this network.  For example, one high-grade dysplasia (Oral22), exhibited three regions of high-level amplification on 7p11 (EGFR), 11q13 (CCND1, FGF3), and 19q13.2 (PAK4).  The presence of multiple amplicons targeting a common network illustrates the potential importance for the disruption of multiple components within single samples.  Next, we evaluated mRNA overexpression among 47 genes of this network not having DNA amplification in five independent HNSCC datasets.  Among them, AKT1, AKT3, NRAS, PIK3CA, PIK3CB, PIK3CG, PRKACB, PRKAR1A, PRKCA, PRKCB1, PRKCE, PRKCI, RRAS, and RRAS2 were significantly overexpressed in at least two of the five datasets.  This further demonstrates the frequent deregulation of this signaling network in the development of HNSCC.). 4.3.5. Validation of mRNA levels in OSCCs It has been previously suggested that different anatomical sites of cancers could affect mRNA expression profiles.  However, the availability of OSCC expression datasets deposited in GEO database is limited, thus hindering sub-group analysis from different sites in the oral cavity or the head and neck region.  Focusing on oral cancer, we analyzed expression levels of two oral cancer-specific studies (N = 61) to validate the importance of the 88 genes in recurrent amplicons for oral carcinogenesis.(Pyeon et al. 2007; Toruner et al. 2004; Ye et al. 2008)  Of the 88 genes, 46 genes were overexpressed in at least one OSCC dataset, including 28 of the 40 overexpressed 103   genes in HNSCC datasets (Supp. Table C.6).  Components of the signaling network, including the FGF3, FGF4, FGF19, EGFR, CCND1, CREB3, and YWHAZ, were also found to be significantly overexpressed in OSCC datasets.  Thus, the expression pattern identified from oral cancer datasets was found to be similar to that of HNSCC 4.3.6. Single cells of oral dysplasia exhibit co-amplification of EGFR and CCND1 Co-amplification of at least two regions of the genome exists in 11 high-risk OPLs and 8 invasive SCC, constituting 55.9% (19/34) of samples with DNA amplification.  We ask if the observed co-amplification is a manifestation of intra-lesion heterogeneity or if synchronous gene amplifications exist in single cells of the lesion.  We performed FISH using probes spanning genomic regions of EGFR, CCND1, and the centromere of chromosome 11 on high-grade dysplasia Oral22 to examine this question (Fig. 4.4). Gene amplifications of EGFR and CCND1 were observed to co-exist in single cells of high-grade dysplasia, demonstrating that both of these changes can occur synchronously at the premalignant stage during oral cancer development.  4.4. Discussion Genetic alterations of DNA amplification and homozygous deletion have long been recognized as chromosomal regions that contain genes important for cancer development.(Albertson 2006; Johnson et al. 1987; Kubo et al. 2008; Lockwood et al. 2008; Myllykangas et al. 2006; Weir et al. 2004; Cairns et al. 1995; Lerman & Minna 2000; Li et al. 1997)  As genetic alterations accumulate to produce a neoplastic phenotype, we focus on preinvasive stages of high-grade and low-grade lesions that are known to further develop into cancer, aiming to identify early genetic events during cancer development.  Amplicons recurrently present in OPLs were identified, and genes that are overexpressed in HNSCC datasets were further distinguished.  The purpose of this study includes: identifying early genetic events in cancer development, understanding the underlying pathways governing the progression of OPLs, and 104   discovering gene candidates that might have therapeutic value for the prevention and treatment of oral cancer patients. 4.4.1. Amplifier phenotype in oral high-grade dysplasias Gene amplification is a major mechanism of oncogene activation and has been associated as poor prognostic indicator in human cancers.(Myllykangas et al. 2007) Here, we detected frequent events of gene amplification in OPLs.  Previous studies have reported gene amplifications in OSCC and oral cancer cell lines, with some studies evaluating known oncogenes or using interval-marker CGH.(Freier et al. 2007; Reshmi et al. 2007; Xia et al. 2007; Baldwin et al. 2005; Garnis et al. 2003; Hsu et al. 2006; Snijders et al. 2005; Smeets et al. 2006)  Here, using an unbiased genome-wide approach we show that the early stages of oral lesions already suffer from increased genomic complexity.  Seven amplicons were found to be recurrently present in oral dysplasias, ranging in size from 0.45 Mbp to 2.26 Mbp.  Previously, Snijders et al. examined 89 oral invasive tumors by interval-marker array and identified nine amplicons smaller than 3 Mbp.(Snijders et al. 2005)  The similar number of amplicons detected in our set of 50 high-risk OPLs was interesting, as it is believed that genetic alterations accumulate as oral dysplasias progress to invasiveness.  Although tiling-path array provided increased genomic coverage compared with interval-marker, parallel analysis of our dataset of 23 OSCCs also detected similar level of genomic complexities between high-risk OPLs and invasive carcinomas.  In addition, none of the 14 non- progressing low-grade lesions harbor region of DNA amplification.  All the identified amplicons in oral dysplasias contain at least one cancer-related gene, supporting the concept that DNA amplification is likely to activate oncogene and contributes to OPL progression.  Taken together, these results show that an amplifier phenotype exists in OPLs, and these amplicons might directly contribute to oral carcinogenesis.  As gene amplification is readily detectable in clinical specimen, and our data showed that amplification frequently exists in high-risk OPLs, it might be plausible for gene amplification to serve as marker in OPLs predictive for aggressive progression to 105   malignancy.  However, a larger sample size of low-grade lesions with clinical outcome would be required to address this hypothesis. 4.4.2. Disruption of multiple components of a signaling network in oral dysplasias It is crucial to understand the deregulated molecular pathways that govern the progression of oral premalignancy.  By evaluating OPLs for gene amplification and examining genes with transcript overexpression in HNSCC and OSCC datasets, we identified genes that are involved in the ERK/MAPK, FGF, p53, PTEN, and PI3K/AKT signaling pathways.  The FGF signaling pathway regulates developmental processes and angiogenesis, and has been an important therapeutic target in human cancers.(Katoh & Katoh 2006; Ornitz & Itoh 2001)  FGF signaling can activate the PI3K/AKT signaling cascade, leading to an induction of epithelial-mesenchymal transition and cell migration;(Vivanco & Sawyers 2002) while the FGF-stimulated ERK/MAPK signaling pathway is implicated in cell differentiation, proliferation, and survival.(McKay & Morrison 2007)  Activation of Akt by phosphorylation has been shown as an early event in oral preneoplastic lesions, and its expression is correlated with poor outcome in oral cancer patients.(Massarelli et al. 2005)  By examining the molecular interactions among the most significantly deregulated pathways as a single network (Fig. 4.3), we demonstrated the diverse mechanisms for the activation of this network, emphasizing the need for molecular targeted therapies of multiple signaling pathways. Our study identified the genes that are amplified early during carcinogenesis.  These genes were targeted towards one signaling network, and gene amplification frequently disrupted this signaling network as early as the low-grade dysplasia.  Moreover, mRNA overexpression was frequently found in members of this network not activated by gene amplification in HNSCC datasets.  The HNSCC datasets were chosen because of their large sample size, public availability, and could potentially broaden the scope of this oral cancer specific study to other HNSCCs.  Nevertheless, expression analysis of oral cancer samples was also performed using two public datasets and similar results were found.  In conclusion, our data highlights the early and frequent activation of one 106   signaling network in OPLs.  This also gives important insights into therapeutic strategies as different genes are mechanistically altered in different individuals. 4.4.3. Pathway addiction in oral dysplasias The term “oncogene addiction” was first coined to describe the physiological dependence of cancer cells on a single activated oncogene for the maintenance of their malignant phenotype.(Weinstein 2002)  In contrast to this phenomenon, a majority of our samples with amplification harbor more than one amplicon, suggesting the dependence on multiple oncogenes for OPL progression.  As amplicons are known to be unstable regions, the maintenance of multiple amplicons must allow them to gain a selective advantage for clonal growth.(Albertson 2006; Miele et al. 1989)  Interestingly, four oral lesions, including three high-grade dysplasias and one invasive carcinoma, exhibited multiple amplicons harboring genes from the same signaling pathways, suggesting the phenomenon of “pathway addiction” in these early stage lesions.(Molinolo et al. 2008)  Although intra-lesional heterogeneity might exist in dysplastic cells, which could contribute to the detection of various amplicons in different clonal populations, we observed co-amplification of two genes in this signaling network, EGFR and CCND1, could occur in single cells of a preinvasive lesion.  This demonstrates the occurrence of an amplifier phenotype in single cells of oral dysplasia, and that multiple oncogenes are potentially important for overgrowth of such cells to form cancer.  Whether the FGF signaling network is causative in oral cancer development needs to be investigated in further studies. The potential “pathway addiction” to this signaling network yields important insight for drug development.  Targeting a receptor tyrosine kinase might not be sufficient if downstream targets are also disrupted.  In addition, genetic testing of multiple components of this network, possibly by FISH assays to detect amplification hotspots, might enhance therapeutic efficacy.(Moroni et al. 2005; Yamatodani et al. 2008)  107   4.5. Summary This is the first comprehensive examination of DNA amplification and homozygous deletion in OPLs.  We identified one signaling network that is frequently deregulated at different components in oral dysplasias, reflecting diverse mechanisms but common underlying biology governing the progression from oral premalignancy into invasiveness.  This study suggests the importance of multiple signaling pathways in the early stage of oral carcinogenesis.  Combined targeting of these oncogenic pathways might be effective for treatment of oral cancer patients since several genes of this network could be activated in a single specimen.  Furthermore, since different targets are altered in different specimens, it is important to identify multiple markers to stratify patients that may benefit from personalized therapies for oral cancer.(Yamatodani et al. 2008) . Figure 4.1 108 Figure 4.1  Whole genome tiling-path array profile of a carcinoma in situ (CIS) Oral42.  Each data point represents one BAC-derived segment on the array.  The log2 signal intensity ratios of a competitive hybridization with pooled male genomic DNA are plotted by SeeGH software.(Chi et al. 2008)  The red and green bar lines are positive and negative log2 signal intensity ratio lines scaled by an increment of 0.5.  Data points to the left and right of the centre line represent DNA copy number losses and gains, respectively.  Specifically, seven regions of gene amplification on 1q23.2-q23.3, 1q23.3- q24.1, 8q22.2-q22.3, 8q23.1, 11q13.3-q13.4, 12q14.3, and 12q23.2-q23.3 are shaded red.  Magnified views of the two amplicons on chromosome 8, and a region of low-level copy number loss and a region of gene amplification on chromosome 11 are shown in two insets. Figure 4.2 109 Figure 4.2.  Graphical representations of region of homozygous deletion in oral lesions.  Each data point represents one BAC-derived segment on the array.  The detected region of homozygous deletion in each sample is shaded in dark green, whereas single copy loss is shaded in pale green on the corresponding data points. The red and green lines are positive and negative ratio lines scaled by an increment of log2 signal ratios of 0.5. Figure 4.3 110 Figure 4.3.  Multiple disruptions in a single network driven by the mechanism of gene amplification.  The top significantly deregulated canonical pathways in oral dysplasias are the ERK/MAPK, FGF, p53, PTEN, and PI3K/AKT signaling pathways, which share common nodes and interplay as a single network.  Genes colored in grey with outlined circle were recurrently deregulated by gene amplification in oral dysplasias and significantly overexpressed in independent head and neck cancer datasets, whereas those colored in light grey were found to be amplified only in one preinvasive lesion.  Altogether, 25 oral lesions (one progressing low-grade lesions, 14 high-grade dysplasias, and ten OSCCs) exhibited high-level gene amplification of different genes inside this network, contributing to a disruption of 34.2% of all the progressing low-grade dysplasias, high-grade dysplasias, and OSCCs (N = 73). Figure 4.4 111 Figure 4.4.  Co-amplification of EGFR and CCND1 in high-grade dysplasia Oral22. A) The left SeeGH profile represents amplification at 7p11.2 (EGFR locus), whereas the right profile represents amplification at 11q13.3 (CCND1 locus) of sample Oral22.  The red and green lines are positive and negative ratio lines scaled by an increment of log2 signal ratios of 0.5.  Amplified region was shaded in red.  B) FISH visualization of co- amplification of EGFR and CCND1 in single dysplastic cells of Oral22.  Sequential FISH was performed with probes mapping to CCND1, EGFR, and the centromeric region of chromosome 11, respectively displayed as orange, yellow, and green signal (original magnification, 1000x).  Note the large clusters of amplification signals of CCND1 and EGFR relative to the centromeric region of chromosomes 11 in the same nucleus. 112   Table 4.1.  Regions of homozygous deletion. Sampl e ID Chromosoma l band Proximal flanking clone* Start bp coordinat e Distal flanking clone* End bp coordinat e Size (Mbp) # of gene Known tumor suppressor s Oral88 9p22.3-p23 692G11  13674530 55P10  14761740 1.09 4 - Oral12 9p22.3 97O17  15964174 554H2  16556908 0.59 1 - Oral88 9p21.3 380P16  21201965 145H12  22290129 1.09 14 CDKN2A, CDN2B, MTAP Oral1 9p21.1-p21.2 802E2  26324440 133E6  28330527 2.01 10 - Oral13 9q33.1 235P13  11693445 0 121P18  11780427 0 0.87 2 - Oral13 9q33.1-q33.2 57K1  11947267 2 374B16  12019017 5 0.72 0 - Oral34 15q15.1 723A20  39784840 468N2  40486332 0.70 13 - Oral11 16q23.1 730M21  76924965 556H2  77778213 0.85 1 WWOX *All the listed human BAC clones were selected from the RPCI-11 library. 113   Table 4.2.  Recurrent regions of gene amplification in OPLs.  Minimal altered regions recurrent in at least two high-risk OPLs are listed. Chromo -some Proximal flanking clone* Start (bp) Distal flanking clone* End (bp) Size (Mbp) Frequency of amplificati on in OPLs (N = 50) Frequenc y of amplificat ion in OSCCs (N = 23) Candidate Genes 2q11.2 793A22 96361 719 61O17 97032 268 0.67 4% 0 CIAO1 4q12 273B19 55487 333 345F18 56688 451 1.20 4% 0 KDR 7p11.2 164O17 54583 596 708P5 55194 026 0.61 8% 23.1% EGFR 8q11.21 770E5 49653 988 259D18 50105 411 0.45 4% 0 SNAI2 8q22.3 302J23 10200 5665 375I14 10251 3409 0.51 6% 4.35% YWHAZ 9p13.3 121D5 33990 614 312A20 36034 564 2.04 4% 0 CCL19, CCL21, CCL27, DCTN3, OPRS1, TLN1, CREB3 11q13.2 -q13.4 715N9 67947 069 CTD- 2011L13 70204 378 2.26 14% 26.1% CCND1, FGF3, FGF19, GAL, FGF4 *Unless otherwise stated, all the listed human BAC clones were selected from the RPCI-11 library. 114   Table 4.3.  Cancer-related genes mapped within recurrent regions of amplicon in high-risk OPLs.  Gene Name Entrez Gene ID for human Chromosome band RNF103 7844 2p11.2 KDR 3791 4q11-q12 EGFR 1956 7p12 TACC1 6867 8p11 LSM1 27257 8p11.2 STAR 6770 8p11.2 WHSC1L1 54904 8p11.2 ADAM9 8754 8p11.23 FGFR1 2260 8p11.2-p11.1 EIF4EBP1 1978 8p12 SNAI2 6591 8q11 CCL19 6363 9p13 CCL21 6366 9p13 CCL27 10850 9p13 CREB3 10488 9p13.3 OPRS1 10280 9p13.3 CA9 768 9p13-p12 RECK 8434 9p13-p12 CCND1 595 11q13 FGF3 2248 11q13 FGF19 9965 11q13.1 GAL 51083 11q13.2-q13.4 FGF4 2249 11q13.3  115   Table 4.4.  Disruption of canonical signaling pathways in oral premalignant lesions. Canonical pathways identified Overexpressed genes in head and neck SCC p-values ERK/MAPK Signaling CREB3, TLN1, YWHAZ 8.95E-03 FGF Signaling CREB3, FGF3 1.63E-02 p53 Signaling CCND1, SNAI2 1.96E-02 PTEN Signaling CCND1, EGFR 1.96E-02 PI3K/AKT pathway CCND1, YWHAZ 3.18E-02  116    4.6.  References 1978. Definition of leukoplakia and related lesions: An aid to studies on oral precancer. Oral Surgery, Oral Medicine, Oral Pathology, 46(4), 518-39. Albertson, D. G., 2006. Gene amplification in cancer. Trends Genet, 22(8), 447-55. Baldwin, C., C. Garnis, L. Zhang, M. P. Rosin & W. L. Lam, 2005. Multiple microalterations detected at high frequency in oral cancer. Cancer Res, 65(17), 7561-7. Cairns, P., T. J. Polascik, Y. Eby, K. Tokino, J. Califano, A. Merlo, L. Mao, J. Herath, R. Jenkins, W. Westra & et al., 1995. Frequency of homozygous deletion at p16/CDKN2 in primary human tumours. Nat Genet, 11(2), 210-2. Chi, B., R. J. deLeeuw, B. P. Coe, R. T. Ng, C. MacAulay & W. L. Lam, 2008. MD- SeeGH: a platform for integrative analysis of multi-dimensional genomic data. BMC Bioinformatics, 9, 243. Chung, C. H., J. S. Parker, G. Karaca, J. Wu, W. K. Funkhouser, D. Moore, D. Butterfoss, D. Xiang, A. Zanation, X. Yin, W. W. Shockley, M. C. Weissler, L. G. Dressler, C. G. Shores, W. G. Yarbrough & C. M. Perou, 2004. Molecular classification of head and neck squamous cell carcinomas using patterns of gene expression. Cancer Cell, 5(5), 489-500. Cobaleda, C., M. Perez-Caro, C. Vicente-Duenas & I. Sanchez-Garcia, 2007. Function of the zinc-finger transcription factor SNAI2 in cancer and development. Annu Rev Genet, 41, 41-61. Coe, B. P., W. W. Lockwood, L. Girard, R. Chari, C. Macaulay, S. Lam, A. F. Gazdar, J. D. Minna & W. L. Lam, 2006. Differential disruption of cell cycle pathways in small cell and non-small cell lung cancer. Br J Cancer, 94(12), 1927-35. 117   Coe, B. P., B. Ylstra, B. Carvalho, G. A. Meijer, C. Macaulay & W. L. Lam, 2007. Resolving the resolution of array CGH. Genomics, 89(5), 647-53. Crissman, J. D. & R. J. Zarbo, 1989. Dysplasia, in situ carcinoma, and progression to invasive squamous cell carcinoma of the upper aerodigestive tract. Am J Surg Pathol, 13 Suppl 1, 5-16. Cromer, A., A. Carles, R. Millon, G. Ganguli, F. Chalmel, F. Lemaire, J. Young, D. Dembele, C. Thibault, D. Muller, O. Poch, J. Abecassis & B. Wasylyk, 2004. Identification of genes associated with tumorigenesis and metastatic potential of hypopharyngeal cancer by microarray analysis. Oncogene, 23(14), 2484-98. Epstein, J. B., M. Gorsky, R. J. Cabay, T. Day & W. Gonsalves, 2008. Screening for and diagnosis of oral premalignant lesions and oropharyngeal squamous cell carcinoma: role of primary care physicians. Can Fam Physician, 54(6), 870-5. Freier, K., C. Schwaenen, C. Sticht, C. Flechtenmacher, J. Muhling, C. Hofele, B. Radlwimmer, P. Lichter & S. Joos, 2007. Recurrent FGFR1 amplification and high FGFR1 protein expression in oral squamous cell carcinoma (OSCC). Oral Oncol, 43(1), 60-6. Freier, K., C. Sticht, C. Hofele, C. Flechtenmacher, D. Stange, L. Puccio, G. Toedt, B. Radlwimmer, P. Lichter & S. Joos, 2006. Recurrent coamplification of cytoskeleton-associated genes EMS1 and SHANK2 with CCND1 in oral squamous cell carcinoma. Genes Chromosomes Cancer, 45(2), 118-25. Fresko, D. & S. S. Lazarus, 1981. Oral carcinoma in situ. Its progression to squamous, basosquamous, and basal-cell carcinoma. Arch Pathol Lab Med, 105(1), 15-9. Garnis, C., C. Baldwin, L. Zhang, M. P. Rosin & W. L. Lam, 2003. Use of complete coverage array comparative genomic hybridization to define copy number alterations on chromosome 3p in oral squamous cell carcinomas. Cancer Res, 63(24), 8582-5. 118   Garnis, C., T. P. Buys & W. L. Lam, 2004. Genetic alteration and gene expression modulation during cancer progression. Mol Cancer, 3, 9. Ginos, M. A., G. P. Page, B. S. Michalowicz, K. J. Patel, S. E. Volker, S. E. Pambuccian, F. G. Ondrey, G. L. Adams & P. M. Gaffney, 2004. Identification of a gene expression signature associated with recurrent disease in squamous cell carcinoma of the head and neck. Cancer Res, 64(1), 55-63. Guillaud, M., L. Zhang, C. Poh, M. P. Rosin & C. MacAulay, 2008. Potential use of quantitative tissue phenotype to predict malignant risk for oral premalignant lesions. Cancer Res, 68(9), 3099-107. Hahn, S. A., M. Schutte, A. T. Hoque, C. A. Moskaluk, L. T. da Costa, E. Rozenblum, C. L. Weinstein, A. Fischer, C. J. Yeo, R. H. Hruban & S. E. Kern, 1996. DPC4, a candidate tumor suppressor gene at human chromosome 18q21.1. Science, 271(5247), 350-3. Hall, G. L., R. J. Shaw, E. A. Field, S. N. Rogers, D. N. Sutton, J. A. Woolgar, D. Lowe, T. Liloglou, J. K. Field & J. M. Risk, 2008. p16 Promoter methylation is a potential predictor of malignant transformation in oral epithelial dysplasia. Cancer Epidemiol Biomarkers Prev, 17(8), 2174-9. Hayward, J. R. & J. A. Regezi, 1977. Oral dysplasia and in situ carcinoma: clinicopathologic correlations of eight patients. J Oral Surg, 35(9), 756-62. Hsu, L. C., X. Huang, S. Seasholtz, D. M. Potter & S. M. Gollin, 2006. Gene amplification and overexpression of protein phosphatase 1alpha in oral squamous cell carcinoma cell lines. Oncogene, 25(40), 5517-26. Ishkanian, A. S., C. A. Malloff, S. K. Watson, R. J. DeLeeuw, B. Chi, B. P. Coe, A. Snijders, D. G. Albertson, D. Pinkel, M. A. Marra, V. Ling, C. MacAulay & W. L. Lam, 2004. A tiling resolution DNA microarray with complete coverage of the human genome. Nat Genet, 36(3), 299-303. 119   Johnson, B. E., D. C. Ihde, R. W. Makuch, A. F. Gazdar, D. N. Carney, H. Oie, E. Russell, M. M. Nau & J. D. Minna, 1987. myc family oncogene amplification in tumor cell lines established from small cell lung cancer patients and its relationship to clinical status and course. J Clin Invest, 79(6), 1629-34. Katoh, M. & M. Katoh, 2006. Cross-talk of WNT and FGF signaling pathways at GSK3beta to regulate beta-catenin and SNAIL signaling cascades. Cancer Biol Ther, 5(9), 1059-64. Khojasteh, M., W. L. Lam, R. K. Ward & C. MacAulay, 2005. A stepwise framework for the normalization of array CGH data. BMC Bioinformatics, 6, 274. Kikuchi, R., H. Tsuda, K. Kozaki, Y. Kanai, T. Kasamatsu, K. Sengoku, S. Hirohashi, J. Inazawa & I. Imoto, 2008. Frequent inactivation of a putative tumor suppressor, angiopoietin-like protein 2, in ovarian cancer. Cancer Res, 68(13), 5067-75. Kubo, T., H. Yamamoto, W. W. Lockwood, I. Valencia, J. Soh, M. Peyton, M. Jida, H. Otani, T. Fujii, M. Ouchida, N. Takigawa, K. Kiura, K. Shimizu, H. Date, J. D. Minna, M. Varella-Garcia, W. L. Lam, A. F. Gazdar & S. Toyooka, 2008. MET gene amplification or EGFR mutation activate MET in lung cancers untreated with EGFR tyrosine kinase inhibitors. Int J Cancer. Lerman, M. I. & J. D. Minna, 2000. The 630-kb lung cancer homozygous deletion region on human chromosome 3p21.3: identification and evaluation of the resident candidate tumor suppressor genes. The International Lung Cancer Chromosome 3p21.3 Tumor Suppressor Gene Consortium. Cancer Res, 60(21), 6116-33. Lese, C. M., K. M. Rossie, B. N. Appel, J. K. Reddy, J. T. Johnson, E. N. Myers & S. M. Gollin, 1995. Visualization of INT2 and HST1 amplification in oral squamous cell carcinomas. Genes Chromosomes Cancer, 12(4), 288-95. Li, J., C. Yen, D. Liaw, K. Podsypanina, S. Bose, S. I. Wang, J. Puc, C. Miliaresis, L. Rodgers, R. McCombie, S. H. Bigner, B. C. Giovanella, M. Ittmann, B. Tycko, H. Hibshoosh, M. H. Wigler & R. Parsons, 1997. PTEN, a putative protein tyrosine 120   phosphatase gene mutated in human brain, breast, and prostate cancer. Science, 275(5308), 1943-7. Lockwood, W. W., R. Chari, B. P. Coe, L. Girard, C. Macaulay, S. Lam, A. F. Gazdar, J. D. Minna & W. L. Lam, 2008. DNA amplification is a ubiquitous mechanism of oncogene activation in lung and other cancers. Oncogene. Lydeard, J. R., S. Jain, M. Yamaguchi & J. E. Haber, 2007. Break-induced replication and telomerase-independent telomere maintenance require Pol32. Nature, 448(7155), 820-3. Mao, L., W. K. Hong & V. A. Papadimitrakopoulou, 2004. Focus on head and neck cancer. Cancer Cell, 5(4), 311-6. Mao, L., J. S. Lee, Y. H. Fan, J. Y. Ro, J. G. Batsakis, S. Lippman, W. Hittelman & W. K. Hong, 1996. Frequent microsatellite alterations at chromosomes 9p21 and 3p14 in oral premalignant lesions and their value in cancer risk assessment. Nat Med, 2(6), 682-5. Massarelli, E., D. D. Liu, J. J. Lee, A. K. El-Naggar, L. Lo Muzio, S. Staibano, S. De Placido, J. N. Myers & V. A. Papadimitrakopoulou, 2005. Akt activation correlates with adverse outcome in tongue cancer. Cancer, 104(11), 2430-6. McKay, M. M. & D. K. Morrison, 2007. Integrating signals from RTKs to ERK/MAPK. Oncogene, 26(22), 3113-21. Miele, M., S. Bonatti, P. Menichini, L. Ottaggio & A. Abbondandolo, 1989. The presence of amplified regions affects the stability of chromosomes in drug-resistant Chinese hamster cells. Mutat Res, 219(3), 171-8. Molinolo, A. A., P. Amornphimoltham, C. H. Squarize, R. M. Castilho, V. Patel & J. S. Gutkind, 2008. Dysregulated molecular networks in head and neck carcinogenesis. Oral Oncol. 121   Moroni, M., S. Veronese, S. Benvenuti, G. Marrapese, A. Sartore-Bianchi, F. Di Nicolantonio, M. Gambacorta, S. Siena & A. Bardelli, 2005. Gene copy number for epidermal growth factor receptor (EGFR) and clinical response to antiEGFR treatment in colorectal cancer: a cohort study. Lancet Oncol, 6(5), 279-86. Myllykangas, S., T. Bohling & S. Knuutila, 2007. Specificity, selection and significance of gene amplifications in cancer. Semin Cancer Biol, 17(1), 42-55. Myllykangas, S., J. Himberg, T. Bohling, B. Nagy, J. Hollmen & S. Knuutila, 2006. DNA copy number amplification profiling of human neoplasms. Oncogene, 25(55), 7324-32. Myllykangas, S., J. Tikka, T. Bohling, S. Knuutila & J. Hollmen, 2008. Classification of human cancers based on DNA copy number amplification modeling. BMC Med Genomics, 1, 15. Noutomi, Y., A. Oga, K. Uchida, M. Okafuji, M. Ita, S. Kawauchi, T. Furuya, Y. Ueyama & K. Sasaki, 2006. Comparative genomic hybridization reveals genetic progression of oral squamous cell carcinoma from dysplasia via two different tumourigenic pathways. J Pathol, 210(1), 67-74. Ornitz, D. M. & N. Itoh, 2001. Fibroblast growth factors. Genome Biol, 2(3), REVIEWS3005. Partridge, M., S. Pateromichelakis, E. Phillips, G. G. Emilion, R. P. A'Hern & J. D. Langdon, 2000. A case-control study confirms that microsatellite assay can identify patients at risk of developing oral squamous cell carcinoma within a field of cancerization. Cancer Res, 60(14), 3893-8. Poh, C. F., S. Ng, K. W. Berean, P. M. Williams, M. P. Rosin & L. Zhang, 2008. Biopsy and histopathologic diagnosis of oral premalignant and malignant lesions. J Can Dent Assoc, 74(3), 283-8. Pyeon, D., M. A. Newton, P. F. Lambert, J. A. den Boon, S. Sengupta, C. J. Marsit, C. D. Woodworth, J. P. Connor, T. H. Haugen, E. M. Smith, K. T. Kelsey, L. P. 122   Turek & P. Ahlquist, 2007. Fundamental differences in cell cycle deregulation in human papillomavirus-positive and human papillomavirus-negative head/neck and cervical cancers. Cancer Res, 67(10), 4605-19. Ralhan, R., L. V. Desouza, A. Matta, S. Chandra Tripathi, S. Ghanny, S. Dattagupta, A. Thakar, S. S. Chauhan & K. W. Siu, 2009. iTRAQ-multidimensional liquid chromatography and tandem mass spectrometry-based identification of potential biomarkers of oral epithelial dysplasia and novel networks between inflammation and premalignancy. J Proteome Res, 8(1), 300-9. Reshmi, S. C., X. Huang, D. W. Schoppy, R. C. Black, W. S. Saunders, D. I. Smith & S. M. Gollin, 2007. Relationship between FRA11F and 11q13 gene amplification in oral cancer. Genes Chromosomes Cancer, 46(2), 143-54. Rhodes, D. R., J. Yu, K. Shanker, N. Deshpande, R. Varambally, D. Ghosh, T. Barrette, A. Pandey & A. M. Chinnaiyan, 2004. ONCOMINE: a cancer microarray database and integrated data-mining platform. Neoplasia, 6(1), 1-6. Romeo, M. S., I. A. Sokolova, L. E. Morrison, C. Zeng, A. E. Baron, F. R. Hirsch, Y. E. Miller, W. A. Franklin & M. Varella-Garcia, 2003. Chromosomal abnormalities in non-small cell lung carcinomas and in bronchial epithelia of high-risk smokers detected by multi-target interphase fluorescence in situ hybridization. J Mol Diagn, 5(2), 103-12. Rosin, M. P., X. Cheng, C. Poh, W. L. Lam, Y. Huang, J. Lovas, K. Berean, J. B. Epstein, R. Priddy, N. D. Le & L. Zhang, 2000. Use of allelic loss to predict malignant risk for low-grade oral epithelial dysplasia. Clin Cancer Res, 6(2), 357- 62. Rosin, M. P., W. L. Lam, C. Poh, N. D. Le, R. J. Li, T. Zeng, R. Priddy & L. Zhang, 2002. 3p14 and 9p21 loss is a simple tool for predicting second oral malignancy at previously treated oral cancer sites. Cancer Res, 62(22), 6447-50. 123   Sheu, J. J., C. H. Hua, L. Wan, Y. J. Lin, M. T. Lai, H. C. Tseng, N. Jinawath, M. H. Tsai, N. W. Chang, C. F. Lin, C. C. Lin, L. J. Hsieh, T. L. Wang, M. Shih Ie & F. J. Tsai, 2009. Functional genomic analysis identified epidermal growth factor receptor activation as the most common genetic event in oral squamous cell carcinoma. Cancer Res, 69(6), 2568-76. Smeets, S. J., B. J. Braakhuis, S. Abbas, P. J. Snijders, B. Ylstra, M. A. van de Wiel, G. A. Meijer, C. R. Leemans & R. H. Brakenhoff, 2006. Genome-wide DNA copy number alterations in head and neck squamous cell carcinomas with or without oncogene-expressing human papillomavirus. Oncogene, 25(17), 2558-64. Snijders, A. M., B. L. Schmidt, J. Fridlyand, N. Dekker, D. Pinkel, R. C. Jordan & D. G. Albertson, 2005. Rare amplicons implicate frequent deregulation of cell fate specification pathways in oral squamous cell carcinoma. Oncogene, 24(26), 4232-42. Summerlin, D. J., 1996. Precancerous and cancerous lesions of the oral cavity. Dermatol Clin, 14(2), 205-23. Temam, S., H. Kawaguchi, A. K. El-Naggar, J. Jelinek, H. Tang, D. D. Liu, W. Lang, J. P. Issa, J. J. Lee & L. Mao, 2007. Epidermal growth factor receptor copy number alterations correlate with poor clinical outcome in patients with head and neck squamous cancer. J Clin Oncol, 25(16), 2164-70. Toruner, G. A., C. Ulger, M. Alkan, A. T. Galante, J. Rinaggio, R. Wilk, B. Tian, P. Soteropoulos, M. R. Hameed, M. N. Schwalb & J. J. Dermody, 2004. Association between gene expression profile and tumor invasion in oral squamous cell carcinoma. Cancer Genet Cytogenet, 154(1), 27-35. Tsui, I. F., M. P. Rosin, L. Zhang, R. T. Ng & W. L. Lam, 2008. Multiple aberrations of chromosome 3p detected in oral premalignant lesions. Cancer Prev Res (Phila Pa), 1(6), 424-9. 124   van de Wiel, M. A., R. Brosens, P. H. Eilers, C. Kumps, G. A. Meijer, B. Menten, E. Sistermans, F. Speleman, M. E. Timmerman & B. Ylstra, 2009. Smoothing waves in array CGH tumor profiles. Bioinformatics. Vivanco, I. & C. L. Sawyers, 2002. The phosphatidylinositol 3-Kinase AKT pathway in human cancer. Nat Rev Cancer, 2(7), 489-501. Warnakulasuriya, S., J. Reibel, J. Bouquot & E. Dabelsteen, 2008. Oral epithelial dysplasia classification systems: predictive value, utility, weaknesses and scope for improvement. J Oral Pathol Med, 37(3), 127-33. Weinstein, I. B., 2002. Cancer. Addiction to oncogenes--the Achilles heal of cancer. Science, 297(5578), 63-4. Weir, B., X. Zhao & M. Meyerson, 2004. Somatic alterations in the human cancer genome. Cancer Cell, 6(5), 433-8. Wong, K. K., R. J. deLeeuw, N. S. Dosanjh, L. R. Kimm, Z. Cheng, D. E. Horsman, C. MacAulay, R. T. Ng, C. J. Brown, E. E. Eichler & W. L. Lam, 2007. A comprehensive analysis of common copy-number variations in the human genome. Am J Hum Genet, 80(1), 91-104. Xia, J., Q. Chen, B. Li & X. Zeng, 2007. Amplifications of TAOS1 and EMS1 genes in oral carcinogenesis: association with clinicopathological features. Oral Oncol, 43(5), 508-14. Yamatodani, T., L. Ekblad, E. Kjellen, A. Johnsson, H. Mineta & J. Wennerberg, 2008. Epidermal growth factor receptor status and persistent activation of Akt and p44/42 MAPK pathways correlate with the effect of cetuximab in head and neck and colon cancer cell lines. J Cancer Res Clin Oncol. Ye, H., T. Yu, S. Temam, B. L. Ziober, J. Wang, J. L. Schwartz, L. Mao, D. T. Wong & X. Zhou, 2008. Transcriptomic dissection of tongue squamous cell carcinoma. BMC Genomics, 9, 69. 125   Chapter 5.  A dynamic oral cancer field--unravelling the underlying biology and its clinical implication4     4 A version of this chapter has been published.  Tsui IFL, Garnis C, Poh CF. (2009) A dynamic oral cancer field --unraveling the underlying biology and its clinical implication. American Journal of Surgical Pathology. 33: 1732-1738. 126   5. A dynamic oral cancer field --unraveling the underlying biology and its clinical implication 5.1. Introduction “Field cancerization” was first proposed six decades ago by Slaughter et al. to describe the multifocal development of oral cancer (Slaughter et al. 1953).  With advances in molecular technology, it has become apparent that gene alterations within an affected field may be spread broadly across the mucosa of oral cancer patients (Braakhuis et al. 2005a; Braakhuis et al. 2003; van Houten et al. 2004).  The reality that these altered cells can contribute to local disease recurrence or the development of second primary tumors has driven extensive research into approaches for effectively defining the margins of a diseased field. Fields in oral cancer patients (including clinically occult fields), areas of oral premalignant lesions (OPLs) and cancers, represent a heterogeneous group of lesions that can vary widely in their potential for malignant transformation and metastasis (Axell et al. 1996; Lumerman et al. 1995; Schepman & van der Waal 1995; Silverman et al. 1984).  Moreover, many OPLs and early staged SCCs are undetectable by standard white light examination and thus may not be readily detected in the clinical setting. Even where OPLs and SCCs can be visibly detected, disease may extend beyond the margins presently defined in the clinic (Poh et al. 2006; Brennan et al. 1995).  These lesion-adjacent "normal" fields, determined by clinical visibility, can harbour molecular alterations that may later be the basis for recurrent disease (Batsakis et al. 1999; Brennan et al. 1995; Fabricius et al. 2002; Koch et al. 1994; Braakhuis et al. 2005a). Recent advances in optical technology have provided a means for simple, cost- effective, and discriminating visualization of disease fields.  Different adjunct tools, including direct fluorescence visualization (FV) and toluidine blue staining, have been reported to aid in the identification of OPLs and early oral cancers, thus guiding decisions for biopsies at the chairside and for surgical margins in the operating room (Poh et al. 2006; Zhang et al. 2005; Poh et al. 2007; Roblyer et al. 2009). 127   The molecular mechanisms governing field change remain unclear.  It is possible that carcinogen exposure to epithelial cells at several sites gives rise to independent transformation events and the induction of multiple genetically unrelated tumors (Slaughter et al. 1953).  Alternatively, a rare transforming event in a single progenitor cell that subsequently repopulates a contiguous area of tissue may give rise to multiple, clonally-related tumors (Bedi et al. 1996).  Lateral intraepithelial migration of a preneoplastic cell has also been proposed to give rise to multiple lesions (Braakhuis et al. 2005b). Clinical presentation and histologic evaluation are unable to determine whether lesions from the same patient are clonally-related.  This distinction is important since it may impact treatment decisions and outcome.  Understanding the specific gene alterations underlying each lesion will be increasingly important as we move towards targeted therapies.  Using improved screening techniques and molecular tools to define relationships between lesions would allow more accurate application of existing treatment modalities (Elser et al. 2007; Ford & Grandis 2003). Previous molecular efforts to assess clonality in oral cancer have been based on evaluation of one or a handful of previously identified critical gene changes (e.g., loss of heterozygosity [LOH] at frequently altered loci, detection of a known cytogenetic marker, or p53 mutational status) (Pateromichelakis et al. 2005; Worsham et al. 1995; Scholes et al. 1998; Bedi et al. 1996; Tabor et al. 2004; Braakhuis et al. 2003; Escher et al. 2009; Shin et al. 1996; van Houten et al. 2004).  This reliance on a priori knowledge for evaluable targets limits the effectiveness of this approach, since gene changes that were not assessed may be driving the progression of a given tumor.  In addition, evaluation of the most frequently occurring molecular events in oral tumorigenesis may establish false relationships between tumors that have obtained mutations independently.  High-resolution whole genome analysis for individual lesions, including fine-mapping of DNA alteration boundaries, represents an effective means of delineating clonal relationships between tumors (Buys et al. 2009).  The detection of alterations with identical boundaries in multiple lesions from the same patient would 128   strongly indicate the presence of a shared progenitor, while the absence of such shared features would be indicative of independent origins.  The ability to accurately define clonality in these cases will have a significant impact on guiding  treatment decisions (Buys et al. 2009). In this case report, we used a novel optical technique to characterize the field of diseased oral tissue in a 52-year-old male smoker who presented with a single oral SCC lesion.  Using this approach in concert with histological and genomic analyses we have comprehensively characterized the altered field.  Through this approach, our results demonstrate how vast and heterogeneous the altered field can be clinically, histological and molecularly and therefore highlight the importance of an enhanced treatment decision based on these technologies.  5.2. Materials and methods 5.2.1. Case presentation A 52-year-old male smoker developed a 1.5 cm nodular tumor at the right lateral tongue, which is clinically visible under white light (Fig. 5.1A, #1).  A hand-held FV device, VELscope ® (LED Med. Inc., BC, Canada) has been applied to guide surgical treatment of oral cancer in the operating room (Poh et al. 2006; Rosin et al. 2007; Poh et al. 2009).  This simple hand-held device uses a blue/violet light (400-460 nm) to excite the oral tissues where normal tissue would re-emit this light as pale green while abnormal tissues would show loss of such autofluorescence (FV loss or FVL) and appear dark brown (Poh et al. 2009; Lane et al. 2006; Poh et al. 2007). Using this FV device, a dark brown area of FVL at the tumor (initially identified under white light, #1) and an additional area, 25 mm anterior to the clinically visible nodular lesion (which appeared normal under white light) were revealed (Fig. 5.1B, #2 and 3). Staining of the oral tissues with toluidine blue was also used to guide surgical treatment. Toluidine blue is a dye that has been shown to stain high-risk OPLs (Zhang et al. 2005). After the application of toluidine blue, two areas were positive: one area overlapped the 129   tumor that was identified under white light (#1) and a second separate area within the FVL area (#3, Fig. 5.1C).  Although area #2 showed loss of autofluorescence, it was negative for toluidine blue stain (Fig. 5.1C). 5.2.2. Tissue samples 5-mm punch biopsies (from areas #1, 2, 3, and 4) were obtained from the surgical specimen (Fig. 5.1) (Poh et al. 2006).  Samples were fixed in 10% neutral formalin and embedded in paraffin wax.  Tissue sections were stained with hematoxylin and eosin (H&E) and diagnoses were confirmed by the study pathologist (CFP) according to the World Health Organization (WHO) classification.  The control sample (#4) was obtained at the surgical margin, 10 mm away from the FVL boundary.  Epithelial cells in the represented areas were meticulously dissected from subtending connective tissue under an inverted microscope using a 23G needle.  DNA was extracted using standard phenol-chloroform protocols followed by ethanol precipitation (Tsui et al. 2008). 5.2.3. Whole genome tiling-path array Tiling-path genomic arrays, SMRT v.2, obtained from the BC Cancer Research Centre Array Laboratory, were used to identify copy number alterations present in the various biopsies (Ishkanian et al. 2004).  The whole genome is represented as 26,819 overlapping bacterial artificial chromosome (BAC) clones spotted in duplicate, allowing breakpoint detection at a resolution of 50 kbp (Coe et al. 2007; Ishkanian et al. 2004). Each sample DNA and the same normal reference male genomic DNA (250 ng each) were random prime labeled with cyanine-3 and cyanine-5 dCTP, respectively. Hybridization, washing and scanning of the arrays, and analysis of the array images were performed as previously described (Tsui et al. 2008). A three-step normalization procedure, including LOWESS fitting, spatial, and median normalization, was used to remove systematic biases (Khojasteh et al. 2005).  SeeGH software was used to display log2  signal intensity ratios in relation to genomic locations in the hg17 assembly (NCBI Build 35) (Chi et al. 2008).  Data points with standard deviation > 0.075 and signal to noise ratio < 3 in either channel were removed (Tsui et 130   al. 2008).  Breakpoint detection algorithm aCGH-smooth was used to delineate boundaries of DNA gain and loss (Jong et al. 2004). 5.2.4. Loss of heterozygosity (LOH) All samples were analyzed by 10 microsatellite markers. The protocol for LOH analysis and scoring have been previously described (Zhang et al. 1997).  Microsatellite markers used mapped to the following ten regions: 3p14.2 (D3S1234, D3S1228, and D3S1300), 9p21 (IFNA, D9S171, D9S1748, and D9S1751), and 17p11.2 (CHRNB1) and 17p13.1 (tp53 and D17S786).  Markers were selected based on previous studies showing their predictive value for cancer risk of OPLs (Rosin et al. 2000).  5.3.  Results 5.3.1. Heterogeneity across an optically altered field The 1.5 cm nodular tumor, apparent clinically, FVL, and uptake of toluidine blue (area #1) was identified as an invasive SCC (Fig. 5.1D).  Area #2, anterior to #1, was clinically not apparent but showed FVL with no uptake of toluidine blue, was classified as a moderate to severe dysplasia (Fig. 5.1E).  Interestingly, area #3, 10 mm anterior to the nodular tumor (#1), was also clinically not apparent but showed FVL and was positive for toluidine blue staining.  This area was histologically assessed as an invasive SCC (Fig. 5.1F).  No dysplasia was detected in the control sample (area #4), which had no FVL and was negative for toluidine blue staining (Fig. 5.1G).  Thus, histologically different areas were found within the single contiguous field of FV loss, while uptake of toluidine blue was only found in the SCC areas (#1 and #3).  The four samples were also examined with LOH markers known to predict risk of progression and recurrence (Rosin et al. 2000; Rosin et al. 2002).  LOH was detected at 9p (D9S1748 and D9S171) and 17p (D17SCHRNB and tp53) in samples #1, 2, and 3.  No LOH was detected at the above loci in normal sample #4.  This further supports the use of FV device to capture high-risk field.  Accrual of multiple biopsies within this optically altered field provides us the opportunity to examine the intralesional heterogeneity of this patient.  . 131   5.3.2. Defining clonal origin among biopsies in an oral cancer field using whole genome breakpoint detection Previous studies have used molecular techniques that are limited to low-resolution findings or are based on known genetic changes to determine a clonal origin among multiple lesions (Bedi et al. 1996; Braakhuis et al. 2005b; Koch et al. 1994; Partridge et al. 2001; Tabor et al. 2004; van Houten et al. 2004; Worsham et al. 1995).  However, the regions assayed occur frequently in oral cancers regardless of clonal origin.  By tiling-path DNA microarray, unique boundaries of genetic alterations in the genome are delineated, thus improving the ability to deduce relatedness among the lesions.  We obtained whole genome profiles of microdissected cells from areas #1, 2, 3, and 4. Genetic alterations common to all samples were all whole arm changes, including loss of 5q and 8p, and gain of 8q (Fig. 5.2).  The most striking feature among the genetic alterations of these samples was a region of high-level amplification on chromosomal band 9p22.3-p24.2 that was absent in the normal sample #4 (Fig. 5.3).  This region of high-level amplification was a distinct feature of these samples and was not commonly observed in other OSCCs from different patients (Baldwin et al. 2005).  However, alteration boundaries were not shared across all three samples.  Area #1 (SCC) had distinctive boundaries at 9p24.2 and 9p22.2, while area #2 (dysplasia) and area #3 (SCC) both shared the same genetic boundaries near the telomere of 9p and 9p22.3 (Fig. 5.3).  Because areas #2 and #3 shared the same alteration boundary it is indicative of a shared origin, distinct from area #1. Although the three samples shared common genetic whole arm alterations, which may be indicative of a common progenitor, whole arm changes, unlike specific regions of breakage, are common in unstable tumor genomes and therefore cannot be used to deduce a clonal origin.  Six additional genetic events were exclusively detected in the clinically apparent SCC (area #1).  The genetic alterations that governed this divergence included segmental gain of 3p12.1-p14.1, 11q12.3-q13.2, chromosome gain of 14q and 15q, and chromosome arm loss of 5p and 21q (Fig. 5.2).  Interestingly, area #2 (dysplasia), located between two SCCs (areas #1 and #3), shared more common 132   genetic alterations with six identical boundaries with area #3 than with area #1. Common genetic changes between area #2 (dysplasia) and area #3 (SCC) included 3p loss, 7p11.2 gain (EGFR), and 7p14.1-pter gain (Fig. 5.2). Concurrent with the premise that genetic alterations accumulate and parallel the histological progression model from dysplasia to SCC, all genetic alterations present in area #2 (dysplasia) were also detected in area #3 (SCC), and additional genetic events occurred for the formation of area #3.  These exclusive changes specific to area #3 (SCC) include segmental gain on 11q11-q22.3, 14q23.2-qter, and 18q11.2, and whole arm loss of 11p, 13q, and 21q (Fig. 5.2).  Alterations detected in area #2 are believed to be early events while additional alterations detected in area #3 are believed to be later events.  Normal sample #4 contained variations in genomic loci of copy number polymorphisms and contained none of the genetic alterations described in samples #1, 2, and 3.  From the above analysis, we hypothesized that two subclone populations originated from a common progenitor, and SCC #3 was derived from dysplasia #2, whereas it appears SCC #1 diverged to a separate clonal lineage.  However, we could not rule out the possibility that whole chromosome arm change of 5q, 8p, and 8q are random independent changes as these are whole arm changes that happen frequently in OPLs from different individuals. 5.3.3. LOH results suggests a common progenitor among biopsies #1, 2, and 3 Examination of microsatellite markers have been used by several previous studies to predict local tumor recurrence and examine clonality of synchronous or metachronous lesions in the oral cavity (Bedi et al. 1996; Partridge et al. 2001; Pateromichelakis et al. 2005; Scholes et al. 1998; Tabor et al. 2004).  Microsatellite analysis of biopsies #1, 2, and 3 revealed LOH at the same allele at  9p (D9S1748 and D9S171) and 17p (CHRNB and tp53), no LOH at 3p14 (D3S1228, D3S1234 and D3S1300), and was non- informative at IFNA and D9S1751. No LOH was detected at the above loci in normal sample #4.  Thus, based on microsatellite markers alone a common progenitor would 133   be suggested among all three biopsies, demonstrating that LOH alone is not sufficient to determine clonality among lesions.  5.4. Discussion A field of alteration, which extended 25 mm anterior to the clinically apparent tumor, was detected by FV device in the oral cavity of a 52-year-old male oral cancer patient. Using the FV device, in conjunction with conventional approaches using white light and toluidine blue stain, several areas within the altered field were sampled for histological and molecular analysis.  One dysplasia and two SCCs were identified for the three biopsies.  Whole genome analysis of these three biopsies revealed two clonal populations of cells with different genetic signatures, providing important clinical implications for tailored treatment in the future.  This is a keen example of using whole genome technologies to determine clonality between samples of a single patient. This case highlights two major issues in oral cancer management.  The first is the clonal origin of multiple tumors within the oral cavity and its clinical implication.  In this case we observe a situation where two tumors (SCCs #1 and #3) are present in close approximation and although they share some common whole arm changes they appear to be genetically different (Fig.5.2), indicating that they arose through subsequent independent mechanisms.  Therefore each tumor may require different therapy regimes as we move towards molecular targeted therapies.  The second is the presence of clinically not apparent fields.  Clearly using white light alone is not sufficient to truly capture the entire field of alteration.  Leaving behind a portion of the altered field during surgical procedures greatly increases the possibility of recurrence for these patients (Brennan et al. 1995; Poh et al. 2009).  Therefore additional techniques, such as FV and toluidine blue stain, are necessary to more accurately define the surgical margins. It is well known that genetic alterations accumulate with increasing histological stage; although most studies have been performed with lesions from different time points or different patients (Califano et al. 1996).  Studying a single altered field where multiple 134   histological grades are present provides us with the opportunity to study the natural history of the disease.  In this case we detected two regions classified as SSC separated by a region of dysplasia.  Two clonal lineages (SCC #1 and SCC #3) governed the formation of the field anterior to SCC #1.  For example, DNA amplification of chromosome 9p was detected in all samples including SCC in area #1, dysplasia in area #2, and SCC in area #3.  While area #2 (dysplasia) and area #3 (SCC) shared identical genetic boundaries of the 9p amplicon, the boundaries were different between two geographically separated SCCs (areas #1 and #3).  This represents divergent cell subpopulations between areas #1 and #3.  Because area #2 (dysplasia) and area #3 (SCC) shared breakpoints for chromosome 9p amplicon, we can conclude that this was an early event.  On the other hand, amplification of cyclin D1 was exclusively present in SCC #3, suggesting a later but aggressive role for this gene to govern the formation of SCC in area #3.  While whole arm loss of chromosome 5q and 8p, and gain of chromosome 8q were common events among areas #1, 2, and 3, suggesting a common progenitor governing the formation of these samples, we cannot conclusively state whether these events were present in a progenitor common to all three areas since whole arm change occurs frequently in different individuals regardless of their clonal origin.  The whole arm changes (5q, 8p, 8q) observed in all three biopsied areas occur independently in almost half of the unrelated OSCC cases we have profiled (Baldwin et al. 2005).  However, it is still possible that these whole arm alterations were priming the entire field with genetically damaged cells, and independent secondary genetic damages occur in two areas within this field, thus contributing to the divergence of two lineages. Molecular analyses using LOH markers were also performed on these biopsied areas. Based on results from LOH alone, we would have concluded all collected biopsies within this field shared a common progenitor, as all other genetic alterations not assayed by the markers escaped detection.  Thus it is crucial to use high-resolution whole genome technologies to examine global genetic alterations when trying to determine a clonal origin.  In addition, clonality among lesions should be established based on the occurrence of specific genetic breakpoints of segmental alterations in 135   order to prevent identifying whole arm changes that are frequently detected in oral cancers from different patients. The understanding of the molecular mechanism paralleling the spread of “field at risk” will place important implications on the prevention and treatment of oral SCC patients. By mapping different biopsies within an optically altered oral mucosa field using whole genome tiling-path microarrays, we hypothesized a common progenitor cell with critical genetic alteration (loss of 5q, 8p and gain of 8q) overpopulated a contiguous field of genetically damaged cells, and subsequent independent genetic events happened governing the formation of two SCCs within a single anatomical field. This case report is a keen example that confirms the use of optical tools is necessary to detect field change of oral cancer patients, which could improve surgical margin decision.  It also has important implication as we move towards tailored therapy because a single field could be extremely dynamic and there might be a need for different targeted therapy to effectively treat different subclones of a single field. Figure 5.1 136 Area 1, squamous cell carcinoma Area 2, moderate to severe dysplasia Area 3, squamous cell carcinoma Area 4, no dysplasia #1 #2 #3 #1 #2 #3 D, A, B, C, E, F, G, #1 #2 #3 Figure 5.1  Clinical characterization and corresponding histology within an oral cancer field at right lateral tongue of a 52-year-old male former smoker.  A. White light image of a clinically visible nodular lesion (#1) and an ill-defined change (#2 and #3) anterior to the nodular tumor area. B. The same field under fluorescence visualization (FV) device showed an area of dark brown change of FV loss at the nodular lesion (#1) and an additional area (#2 and #3), 25 mm anterior to the nodular area (#1). C.  Uptake of toluidine blue in areas #1 and #3 but absent in area #2.  D-G. Photomicrographs showed varying degrees of histology change of the field: D (#1) invasive squamous cell carcinoma, E (#2) moderate to severe dysplasia, F (#3) invasive squamous cell carcinoma, and G (#4) no dysplasia at the surgical boundary, 10 mm away from FV boundary (H & E staining, original magnification, 40X). Figure 5.2 137 Moderate to severe dysplasia 5q loss 8p loss 8q gain 3p loss 5p loss 3p12.1-p14.1 gain 7p11.2 gain 7p14.1-pter gain 9p22.3-pter amplification 11p loss 11q12.3-q13.2 gain 11q11-q22.3 gain 11q13.2-q13.4 gain 13q loss 14q23.2-qter gain 18q11.2 gain 21q loss 14q gain 15q gain 21q loss #1, 2, 3 #2, 3#1 #3 Invasive SCC Invasive SCC Figure 5.2.  Summary of genetic alterations in different cell populations from various areas within an optically altered oral cancer field.  All three samples from areas #1, 2, and 3 exhibited loss of 5q, 8p and gain of 8q.  All the regions of segmental alteration common to areas #2 (dysplasia) and #3 (SCC) shared identical genetic boundaries. Figure 5.3 138 Figure 5.3.  Chromosome 9p genetic profiles of samples from areas #1 (SCC), #2 (dysplasia), #3 (SCC), and #4 (10 mm beyond FV boundary, no dysplasia).  The red and green lines are positive and negative ratio lines scaled by an increment of log2 signal ratios of 0.5.  Amplified region was shaded in red.  Different genetic boundaries of alteration were observed in samples from areas #1 and #3, while the same genetic boundaries were shared in areas #2 and #3. 139    5.5.  References Axell, T., J. J. Pindborg, C. J. Smith & I. van der Waal, 1996. Oral white lesions with special reference to precancerous and tobacco- related lesions: conclusions of an international symposium held in Uppsala, Sweden, May 18-21 1994. International Collaborative Group on Oral White Lesions. J Oral Pathol Med, 25(2), 49-54. Baldwin, C., C. Garnis, L. Zhang, M. P. Rosin & W. L. Lam, 2005. Multiple microalterations detected at high frequency in oral cancer. Cancer Res, 65(17), 7561-7. Batsakis, J. G., P. Suarez & A. K. el-Naggar, 1999. Proliferative verrucous leukoplakia and its related lesions. Oral Oncol, 35(4), 354-9. Bedi, G. C., W. H. Westra, E. Gabrielson, W. Koch & D. Sidransky, 1996. Multiple head and neck tumors: evidence for a common clonal origin. Cancer Res, 56(11), 2484-7. Braakhuis, B. J., R. H. Brakenhoff & C. R. Leemans, 2005a. Second field tumors: a new opportunity for cancer prevention? Oncologist, 10(7), 493-500. Braakhuis, B. J., C. R. Leemans & R. H. Brakenhoff, 2005b. Expanding fields of genetically altered cells in head and neck squamous carcinogenesis. Semin Cancer Biol, 15(2), 113-20. Braakhuis, B. J., M. P. Tabor, J. A. Kummer, C. R. Leemans & R. H. Brakenhoff, 2003. A genetic explanation of Slaughter's concept of field cancerization: evidence and clinical implications. Cancer Res, 63(8), 1727-30. Brennan, J. A., L. Mao, R. H. Hruban, J. O. Boyle, Y. J. Eby, W. M. Koch, S. N. Goodman & D. Sidransky, 1995. Molecular assessment of histopathological 140   staging in squamous-cell carcinoma of the head and neck. N Engl J Med, 332(7), 429-35. Buys, T. P., S. Aviel-Ronen, T. K. Waddell, W. L. Lam & M. S. Tsao, 2009. Defining genomic alteration boundaries for a combined small cell and non-small cell lung carcinoma. J Thorac Oncol, 4(2), 227-39. Califano, J., P. van der Riet, W. Westra, H. Nawroz, G. Clayman, S. Piantadosi, R. Corio, D. Lee, B. Greenberg, W. Koch & D. Sidransky, 1996. Genetic progression model for head and neck cancer: implications for field cancerization. Cancer Res, 56(11), 2488-92. Chi, B., R. J. deLeeuw, B. P. Coe, R. T. Ng, C. MacAulay & W. L. Lam, 2008. MD- SeeGH: a platform for integrative analysis of multi-dimensional genomic data. BMC Bioinformatics, 9, 243. Coe, B. P., B. Ylstra, B. Carvalho, G. A. Meijer, C. Macaulay & W. L. Lam, 2007. Resolving the resolution of array CGH. Genomics, 89(5), 647-53. Elser, C., L. L. Siu, E. Winquist, M. Agulnik, G. R. Pond, S. F. Chin, P. Francis, R. Cheiken, J. Elting, A. McNabola, D. Wilkie, O. Petrenciuc & E. X. Chen, 2007. Phase II trial of sorafenib in patients with recurrent or metastatic squamous cell carcinoma of the head and neck or nasopharyngeal carcinoma. J Clin Oncol, 25(24), 3766-73. Escher, A., E. Piotet, F. Waridel, R. Iggo & P. Monnier, 2009. p53 Mutation in histologically normal mucosa of the aero-digestive tract is not a marker of increased risk for second primary carcinoma in head and neck cancer patients. Eur Arch Otorhinolaryngol, 266(4), 547-51. Fabricius, E. M., U. Gurr & G. P. Wildner, 2002. Telomerase activity levels in the surgical margin and tumour distant tissue of the squamous cell carcinoma of the head-and-neck. Anal Cell Pathol, 24(1), 25-39. 141   Ford, A. C. & J. R. Grandis, 2003. Targeting epidermal growth factor receptor in head and neck cancer. Head Neck, 25(1), 67-73. Ishkanian, A. S., C. A. Malloff, S. K. Watson, R. J. DeLeeuw, B. Chi, B. P. Coe, A. Snijders, D. G. Albertson, D. Pinkel, M. A. Marra, V. Ling, C. MacAulay & W. L. Lam, 2004. A tiling resolution DNA microarray with complete coverage of the human genome. Nat Genet, 36(3), 299-303. Jong, K., E. Marchiori, G. Meijer, A. V. Vaart & B. Ylstra, 2004. Breakpoint identification and smoothing of array comparative genomic hybridization data. Bioinformatics, 20(18), 3636-7. Khojasteh, M., W. L. Lam, R. K. Ward & C. MacAulay, 2005. A stepwise framework for the normalization of array CGH data. BMC Bioinformatics, 6, 274. Koch, W. M., J. O. Boyle, L. Mao, J. Hakim, R. H. Hruban & D. Sidransky, 1994. p53 gene mutations as markers of tumor spread in synchronous oral cancers. Arch Otolaryngol Head Neck Surg, 120(9), 943-7. Lane, P. M., T. Gilhuly, P. Whitehead, H. Zeng, C. F. Poh, S. Ng, P. M. Williams, L. Zhang, M. P. Rosin & C. E. MacAulay, 2006. Simple device for the direct visualization of oral-cavity tissue fluorescence. J Biomed Opt, 11(2), 024006. Lumerman, H., P. Freedman & S. Kerpel, 1995. Oral epithelial dysplasia and the development of invasive squamous cell carcinoma. Oral Surg Oral Med Oral Pathol Oral Radiol Endod, 79(3), 321-9. Partridge, M., S. Pateromichelakis, E. Phillips, G. Emilion & J. Langdon, 2001. Profiling clonality and progression in multiple premalignant and malignant oral lesions identifies a subgroup of cases with a distinct presentation of squamous cell carcinoma. Clin Cancer Res, 7(7), 1860-6. Pateromichelakis, S., M. Farahani, E. Phillips & M. Partridge, 2005. Molecular analysis of paired tumours: time to start treating the field. Oral Oncol, 41(9), 916-26. 142   Poh, C. F., C. E. MacAulay, L. Zhang & M. P. Rosin, 2009. Tracing the "at-risk" oral mucosa field with autofluorescence: steps toward clinical impact. Cancer Prev Res (Phila Pa), 2(5), 401-4. Poh, C. F., S. P. Ng, P. M. Williams, L. Zhang, D. M. Laronde, P. Lane, C. Macaulay & M. P. Rosin, 2007. Direct fluorescence visualization of clinically occult high-risk oral premalignant disease using a simple hand-held device. Head Neck, 29(1), 71-6. Poh, C. F., L. Zhang, D. W. Anderson, J. S. Durham, P. M. Williams, R. W. Priddy, K. W. Berean, S. Ng, O. L. Tseng, C. MacAulay & M. P. Rosin, 2006. Fluorescence visualization detection of field alterations in tumor margins of oral cancer patients. Clin Cancer Res, 12(22), 6716-22. Roblyer, D., C. Kurachi, V. Stepanek, M. D. Williams, A. K. El-Naggar, J. J. Lee, A. M. Gillenwater & R. Richards-Kortum, 2009. Objective detection and delineation of oral neoplasia using autofluorescence imaging. Cancer Prev Res (Phila Pa), 2(5), 423-31. Rosin, M. P., X. Cheng, C. Poh, W. L. Lam, Y. Huang, J. Lovas, K. Berean, J. B. Epstein, R. Priddy, N. D. Le & L. Zhang, 2000. Use of allelic loss to predict malignant risk for low-grade oral epithelial dysplasia. Clin Cancer Res, 6(2), 357- 62. Rosin, M. P., W. L. Lam, C. Poh, N. D. Le, R. J. Li, T. Zeng, R. Priddy & L. Zhang, 2002. 3p14 and 9p21 loss is a simple tool for predicting second oral malignancy at previously treated oral cancer sites. Cancer Res, 62(22), 6447-50. Rosin, M. P., C. F. Poh, M. Guillard, P. M. Williams, L. Zhang & C. MacaUlay, 2007. Visualization and other emerging technologies as change makers for oral cancer prevention. Ann N Y Acad Sci, 1098, 167-83. 143   Schepman, K. P. & I. van der Waal, 1995. A proposal for a classification and staging system for oral leukoplakia: a preliminary study. Eur J Cancer B Oral Oncol, 31B(6), 396-8. Scholes, A. G., J. A. Woolgar, M. A. Boyle, J. S. Brown, E. D. Vaughan, C. A. Hart, A. S. Jones & J. K. Field, 1998. Synchronous oral carcinomas: independent or common clonal origin? Cancer Res, 58(9), 2003-6. Shin, D. M., J. S. Lee, S. M. Lippman, J. J. Lee, Z. N. Tu, G. Choi, K. Heyne, H. J. Shin, J. Y. Ro, H. Goepfert, W. K. Hong & W. N. Hittelman, 1996. p53 expressions: predicting recurrence and second primary tumors in head and neck squamous cell carcinoma. J Natl Cancer Inst, 88(8), 519-29. Silverman, S., M. Gorsky & F. Lozada, 1984. Oral leukoplakia and malignant transformation. A follow-up study of 257 patients. Cancer, 53(3), 563-8. Slaughter, D. P., H. W. Southwick & W. Smejkal, 1953. Field cancerization in oral stratified squamous epithelium; clinical implications of multicentric origin. Cancer, 6(5), 963-8. Tabor, M. P., R. H. Brakenhoff, H. J. Ruijter-Schippers, J. A. Kummer, C. R. Leemans & B. J. Braakhuis, 2004. Genetically altered fields as origin of locally recurrent head and neck cancer: a retrospective study. Clin Cancer Res, 10(11), 3607-13. Tsui, I. F., M. P. Rosin, L. Zhang, R. T. Ng & W. L. Lam, 2008. Multiple aberrations of chromosome 3p detected in oral premalignant lesions. Cancer Prev Res (Phila Pa), 1(6), 424-9. van Houten, V. M., C. R. Leemans, J. A. Kummer, J. Dijkstra, D. J. Kuik, M. W. van den Brekel, G. B. Snow & R. H. Brakenhoff, 2004. Molecular diagnosis of surgical margins and local recurrence in head and neck cancer patients: a prospective study. Clin Cancer Res, 10(11), 3614-20. Worsham, M. J., S. R. Wolman, T. E. Carey, R. J. Zarbo, M. S. Benninger & D. L. Van Dyke, 1995. Common clonal origin of synchronous primary head and neck 144   squamous cell carcinomas: analysis by tumor karyotypes and fluorescence in situ hybridization. Hum Pathol, 26(3), 251-61. Zhang, L., C. Michelsen, X. Cheng, T. Zeng, R. Priddy & M. P. Rosin, 1997. Molecular analysis of oral lichen planus. A premalignant lesion? Am J Pathol, 151(2), 323- 7. Zhang, L., M. Williams, C. F. Poh, D. Laronde, J. B. Epstein, S. Durham, H. Nakamura, K. Berean, A. Hovan, N. D. Le, G. Hislop, R. Priddy, J. Hay, W. L. Lam & M. P. Rosin, 2005. Toluidine blue staining identifies high-risk primary oral premalignant lesions with poor outcome. Cancer Res, 65(17), 8017-21. 145   Chapter 6: Discussion and Conclusions  146   6. Discussion and conclusions 6.1. Research summary Oral squamous cell carcinoma (OSCC) is the sixth most common cancer in the world (Warnakulasuriya 2009).  The 5-year survival rate is < 50% for all stages and < 25% for advanced stages (stage III and IV).  Coupling early detection of disease at premalignant stages with intensified follow-up represents a key step towards improving these poor survival rates. Oral cancer is known to arise through a stepwise accumulation of genetic events, with increases in genomic instability parallelling advances in histopathological staging. Thus, examining oral premalignant lesions (OPLs) will identify alterations that drive cancer progression at each histologic stages, and will pinpoint early genetic events that may be masked in late stage tumours by increased genomic instability.  However, there is a lack of understanding in OPLs, with most studies examining only the cancerous stage.  Furthermore, previous studies examining genetic alterations in OPLs have been based on single marker assays (e.g. loss of heterozygosity), greatly hampering the discovery of novel oncogenes or tumour suppressor genes.  This is mainly because of technical challenges associated with handling minute premalignant lesion, the abilty to isolate sufficient DNA for genome-wide assays, and the need for long term patient follow-up to determine whether the lesion progress or regress. The work described in this thesis details some of the first high-resolution molecular analyses of OPLs.  Extraction approaches for clinical specimens and whole genome DNA microarray analyses were optimized for low quality and low quantity DNA samples. Thus, the low quantity and low quality of DNA extracted from archival oral specimens are useful for whole genome DNA analyses.  Because high-grade dysplasias are strongly associated with progression to invasiveness, and low-grade dysplasias are not likely to progress, we examined a large number of OPLs with the following characteristics: high-grade dysplasias, low-grade dysplasias that are known to further progress to cancer, and low-grade dysplasias that do not progress to cancer.  By identifying alterations in OPLs and oral cancers, we have gained insight into the genetic 147   mechanisms that underlie disease progression.  This is especially important as all of the samples examined in this thesis contained cilnical follow-up information.  These analyses have also allowed us to better understand the clonal evolution of  multiple lesions in the same diseased tissue field, yielding further insights into the origins and behaviours of oral lesions.  6.1.1. Providing a comprehensive resource for the oral cancer community Tumour genomes often contain an extensive number of genetic aberrations.  Whole genome DNA copy number analyses are needed to identify a broad range of chromosomal abnormalities including gain or loss of chromosomes, non-reciprocal translocation, and focal amplification (Albertson et al. 2003).  Immortalized oral cell lines are a fundamental tool for the investigation of genetic changes, biochemical disruptions, and pharmacological responses in head and neck cancer.  Previous studies have applied lower resolution genomic technologies to identify critical DNA alterations in these models (Jarvinen et al. 2006; Jarvinen et al. 2008; Lin et al. 2007; Squire et al. 2002).  Although cancer cell lines typically differ from tumours in their cell population homogeneity, growth conditions, and gene expression patterns, the ability to collect good quality DNA and RNA, the easy handling, the low usage costs, and the unlimited amount of materials they afford make them an attractive research tool.  Significantly, cell line behaviours and phenotypes can often mirror what is observed in the clinical setting.  Given the need to discover the molecular drivers for these phenotypes, it would be very useful for members of the oral cancer community to have access to a large collection of molecular data from these model systems.  To address this need, I comprehensively characterized six commonly used head and neck cancer cell lines by whole genome copy number and mRNA expression profiling technologies, identifying detailed genetic alterations specific to each cell line and identifying gene expression alterations that are the direct result of copy number aberrations of that cell line (Chapter 2).  As each cell line exhibits different underlying molecular alterations, it is important to consider these molecular alterations when attempting to model cancer cell behaviour. 148   In addition to its utility as a resource for molecular analysis of oral cancer cell models, this work demonstrates the power of integrating multiple genomic approaches to identify critical gene changes in individual samples.  6.1.2. Identifying recurrent regions of alterations in oral premalignant lesions and oral cancer Molecular profiling studies have shown that there is tremendous molecular heterogeneity between different cancer types, within the same cancer type, and even within tumours (Chung et al. 2002; Chung et al. 2004; Swanton & Caldas 2009). Despite the molecular heterogeneity within different tumours, it is generally in agreement that recurrent genetic alterations will span genes that are critical for development of a given cancer type.  This is supported by the consistent finding that such alterations do often cover known oncogenes (e.g. CCND1) and tumour suppressor genes (e.g. CDKN2A) - and that expression of these genes does change depending on alteration status (Albertson et al. 2003; Cairns et al. 1995; Tsui et al. 2009; Lin et al. 2007). In chapter 3, I focused on the chromosome 3p arm to identify recurrent regions of alteration in OPLs.  I focused on this chromosome arm because 1) it has been consistently reported as the most commonly occurring alteration in oral cancer studies and 2) its loss has been associated with progression beyond OPL staging (Chakraborty et al. 2003; Mao et al. 1996; Partridge et al. 2000; Rosin et al. 2000; Rosin et al. 2002). Significantly, 3p alterations in OPLs have only been evaluated at a small number of loci. I used tiling-path array comparative genomic hybridization (CGH) to analyze 71 microdissected archival OPLs with clinical follow-up information and 23 OSCCs to identify genetic alterations at a functional resolution of 50 kbp (Coe et al. 2007).  Six recurrent region of loss were identified in both high-grade dysplasias (HGDs) and OSCCs, which includes a 2.89 Mbp deletion spanning the FHIT gene.  The FHIT  gene has been previously implicated in oral cancer progression, thus demonstrating the 149   potential to use this approach for cancer gene discovery.  Additional candidates including GRM7, RAD18, WNT7A, XPC, or TGFBR2 could be potentially important for the development of oral carcinogenesis.  When the alteration status for the six identified regions was examined in 24 low-grade dysplasias with known progression outcome, I observed that these alterations occurred at a significantly higher frequency in low grade dysplasias that progressed to later stage disease as compared with those that did not progress (p < 0.003).  Moreover, parallel analysis of all profiled tissues showed that the extent of overall genomic alteration at 3p increased with histological stage.  This first high resolution analysis of chromosome arm 3p in OPLs represents a significant step towards identifying recurrent genetic aberrations associated with oral cancer progression, supporting hypothesis 1.  This work also provides an example of how genomic instability escalates with progression to invasive cancer, implicating the importance to examine early OPLs for genetic events that are important to oral cancer development.  6.1.3. Delineating the key oncogenic pathways for oral cancer development When this thesis work was started, array CGH had not been applied to analyze OPLs. In Chapter 4, I aimed to analyze the whole genome of OPLs, focusing specifically on a particular genetic feature: high-level DNA amplifications.  Screening for focal DNA amplifications is useful for the discovery of candidate oncogenes (Albertson 2006). Previous studies have identified regions of DNA amplification in oral cancer and defined candidate genes contributing to oral carcinogenesis based on their known functions (Snijders et al. 2005; Persson et al. 2009; Reshmi et al. 2007b; Reshmi et al. 2007a; Jarvinen et al. 2006).  Amplification of CCND1 and EGFR have been recurrently identified in different studies and using different methods, demonstrating their importance to oral cancer development.  However, the amplification status of these genes in OPLs and their importance to oral carcinogenesis were not known.  In chapter 4, I asked 1) whether DNA amplification occurs at this early stage of cancer development and 2) which oncogenic pathways are disrupted by DNA dosage changes 150   in OPLs.  I evaluated 50 high-grade dysplasias and low-grade dysplasias that later progressed to cancer for DNA copy number aberrations using tiling-path CGH microarrays.  I detected early occurrences of DNA amplification and homozygous deletion in 40% (20/50) of OPLs.  The seven amplicons identified as recurrent in OPLs were determined to span 88 genes.  Because matched RNA samples were unavailable for each OPL with high-level amplification, I evaluated the expression of each gene in five independent head and neck cancer datasets, speculating that critical oncogenes would have retained increased expression in head and neck cancer specimens.  In total, 40 candidates were found to be overexpressed in head and neck cancer relative to normal tissues.  These genes were significantly enriched in the canonical ERK/MAPK, FGF, p53, PTEN, and PI3K/AKT signalling pathways (P = 8.95x10-3-- 3.18x10-2).  These identified pathways share interactions in one signalling network, and 30% of these oral dysplasias were found to have at least one gene within this network that is deregulated by DNA amplification.  No such alterations were found in 14 low- grade dysplasias that did not progress, while 43.5% (10/23) of OSCCs were found to have altered genes within the pathways with DNA amplification.  Thus, different mechanisms were employed by different samples to activate this common network, demonstrating a common  underlying biology governing the progression of the disease. Furthermore, multi-target FISH showed that amplification of EGFR and CCND1 can co- exist in single cells of an oral dysplasia, suggesting that there is dependence on both oncogenes for OPL progression.  Taken together, these findings identify a critical biological network that is frequently disrupted in oral preinvasive lesions that have a high risk to progress to cancer, with different genes disrupted in different individuals. These findings support hypothesis 2 of this thesis.  This work implicates the need for combined targeting of multiple signaling pathways and the importance of identifying markers capable of stratifying patients that may benefit from personalized targeted therapies in oral cancer. 151   6.1.4. Understanding the clonal relationships of field effect in oral cancer Oral cancer patients often suffer from field effect, where histologically and genetically abnormal cells surround a clinically visible tumor to a wide extent.  Although the oral cavity is accessible for routine screening of suspicious lesions, gene alterations are known to accrue in histologically normal tissues.  However, molecular studies exploring the clonal evolution of these multiple lesions have been based on evaluation of a limited number of gene changes (LOH assays, detection of known cytogenetic marker, or p53 mutational status, which are prone to be frequently altered in oral cancer regardless of their origins) (Bedi et al. 1996; Braakhuis et al. 2003; Escher et al. 2009; Pateromichelakis et al. 2005; Scholes et al. 1998; Shin et al. 1996; Tabor et al. 2004; van Houten et al. 2004; Worsham et al. 1995).  Recently, emerging optical and molecular technologies have provided a powerful means for redefining the extent of the field of alteration.  Often this means expanding upon regions detectable with standard white light approaches (Poh et al. 2006).  In chapter 5, I used a newly developed optical technique, direct fluorescence visualization, to define a contiguous field that extended beyond the margins of a clinically visible oral squamous cell carcinoma.  Three biopsies were taken within this contiguous optically altered field.  These biopsies indicated the presence of two squamous cell carcinomas which were separated by a moderate dysplasia.  I used the tiling-path CGH array described above to detect and define genomic breakpoints in each biopsy specimen.  Genetic alterations detected for each specimen were compared to define whether each lesion arose independently or as a consequence of a shared progenitor cell, and clonal ordering was performed to make inferences about the timing of detected genetic alterations.  My data suggested that two SCCs developing within 10 mm of each other along the right lateral tongue were genetically unrelated.  These findings provide evidence supporting the need for effective tools for defining surgical margins and the need for targeted therapies.  My resuts demonstrated that the oral cancer field is genetically heterogeneous.  In addition, clonal evolution of multiple lesions from a field can be determined by genetic signatures, a finding that supports hypothesis 3 of this thesis.  These findings provide tremendous 152   implications on targeted therapies of different cell populations and the development of oral cancer cells as an evolutionary process.  6.2. Discussions The work described in this thesis has contributed to the understanding of oral cancer development and includes the first comprehensive genomic analysis of a large set of oral premalignant lesions with known clinical outcomes.  Such analyses were precluded in the past because of the low quantity and poor quality of DNA isolated from these specimens, and the need for long-term clinical follow-up information.  The samples used in this thesis were obtained from the British Columbia Oral Biopsy Service, which is a repository for formalin-fixed paraffin-embedded oral tissue sections.  DNA extracted from these specimens, particularly early stage premalignant lesions, are often of insufficient DNA quantity and quality due to the irreversible damage caused by formalin fixation.  Even with the use of array CGH, only 70% of archival OPL specimens have produced useful genomic data.  From the specimens that generated useful genomic data, I have characterized DNA alterations in fine detail for OPLs and have associated these molecular data with rich clinical information to gain insights into disease progression. 6.2.1. Biological relevance of the identified genetic alteration and candidate genes In most oral cancer studies, only a subset of genetic loci or genes have been evaluated and found to be involved in oral cancer development, including the well known EGFR and CCND1 oncogenes.  However, the majority of genes responsible for carcinogenesis remain unknown.  This thesis describes comprehensive analyses to identify novel genomic alterations in oral lesions and cancers.  First, I comprehensively characterized the genome of six head and neck cell models using an integrative analysis of DNA copy number and parallel mRNA expression alterations.  Analysis at this high resolution has not been previously performed for these cell lines; data I generated have been made 153   publicly available and have been tabulated for easy reference by other groups.  Similar approaches were used to perform the first-ever analyses of OPL genomes.  Initially, we analyzed clinical outcome data in combination with the alteration status of the most commonly detected DNA dosage alteration in OPLs: chromosome 3p.  We identified the segmental genetic alterations that were recurrently detected in at least 75% of oral high- grade dysplasias.  These genetic alterations were also found to be common in low- grade dysplasias that progressed to cancer and invasive tumours, but not in dysplasias that did not progress to cancer.  With the improved resolution in this work, previously identified regions were fine-mapped and thus allowed us to name specific gene candidates.  For example, the frequent deletion of 3p24.1 has been fine-mapped to 1.21 Mbp region harbouring four genes, including the candidate gene TGFBR2.  This region of loss was detected in 68% of high-grade dysplasias, 96% of OSCCs, and 56% of low- grade dysplasias that progressed to cancer, but none in the low-grade dysplasias that never progressed to cancer.  Additional including GRM7, RAD18, WNT7A, or XPC could also be potentially important for oral carcinogenesis.  Thus, the identification of these segmental alterations in oral dysplasias might provide important gene candidates for driving oral tumourigenesis, since these genes are often disrupted by gene dosage in oral dysplasias and over 95% in oral tumours.  Further elucidating their biological functions may reveal the mechanism necessary for the progression of oral dysplasia to cancer.  Furthermore, this study also identified additional biomarkers on chromosome 3p that could be useful for assessing progression risk in oral dysplasias.  Association of these regions with clinical features provides additional tools for predicting progression risks in OPLs. Next, analysis of the whole genome was performed to identify additional recurring alterations in OPLs, including low-level or high-level copy number aberrations.  By profiling different stages of disease with known clinical outcome - including low-grade dysplasias, high-grade dysplasias, and invasive tumours - I was able to make broad conclusions about genomic instability during disease progression and deduce specific genetic alterations associated with disease progression.  Increased genetic alterations were found to parallel increasing histological stages of the disease, further confirming 154   that genetic analysis of oral dysplasias identifies critical changes for progression and development of oral cancer.  Low-level copy number alterations often involve large regions with many genes, which may undermine the ability to hone in on key oncogenes or tumour suppressor genes.  Therefore, in chapter 4, I focused on the identification of high-level copy number alterations.  DNA amplicons are unstable; these regions would be lost if they did not contain genes advantageous for growth.  In my sample set, 40% of the OPLs that are associated with progression exhibited at least one region of high- level copy number alteration, as compared to 65.2% of OSCCs with such changes. These two numbers are not statistically significant, demonstrating that an amplifier phenotype occurs early at the premalignant stage, and might directly contribute to oral cancer development.  Furthermore, the selection of high-level amplification is not a random process as the same regions were recurrently detected in these OPLs.  By exploring independent oral cancer expression datasets we also found that the identified genes within recurrent amplicons also exhibited overexpression in tumours as compared to normal tissues, further suggesting its role in cancer development.  Many of these activated genes are important for the FGF signalling network, and 34.2% of all samples exhibited high-level gene amplification of different genes inside this network. Thus, one signalling network is frequently deregulated by different mechanisms in oral dysplasias, while the underlying pathways governing the progression from oral dysplasias to invasiveness remains to be similar between different samples.  As specific gene candidates have never been identified from whole genome examination of OPLs, this study contributed to the understanding of gene candidates and molecular pathways that might have therapeutic value for the treatment of oral cancer patients.  This work also implicates the need for combined targeting of multiple signalling pathways for effective treatment because several genes of the network could be activated in the same patient.  Furthermore, the identification of regions of DNA amplification might provide efficient biomarker for the stratification of patients that may have poor prognosis or benefit from targeted therapies.  This work contributes to the understanding of the biological pathways governing oral cancer development and identifying gene candidates 155   that might have therapeutic value for the prevention and treatment of oral cancer patients. 6.2.2. Clonal evolution for the development of multiple oral lesions Throughout my thesis, we have found that the evolution of oral cancer results from the accumulation of genetic alterations.  Field cancerization, where genetically abnormal cells surround a tumour without detectable phenotypic changes, poses a challenge when delineating surgical boundaries.  It is often thought that field effect leads to recurrence (due to poor defined surgical margins) and the development of multifocal cancer in the same patient.  By identifying multiple lesions within a closely-related disease field, I focused on evaluating the sequence of clonal evolution and the genetic relatedness among lesions from the same field.  For the case evaluated in chapter 5, a clinically identifiable SCC was found on the right lateral tongue under white light imaging, while a tumour 10 mm anterior to the nodular tumour and a moderate dysplasias in between the two SCCs were identified with adjunct tools including direct fluorescence visualization and toluidine blue staining.  The exact genetic breakpoints in each genome were defined for each biopsy using whole genome tiling-path array CGH. Clonal ordering was performed to infer the sequence of genetic events of samples within this patient, and revealed two clonally-related populations of cells existed within this altered field (Merlo et al. 2006).  From this analysis, independent lesions and tumours were present in this patient.  As previous molecular efforts to assess clonality have been based on evaluation of one or a handful of previously identified gene changes, I also performed single gene-based assays on these samples and found that all these samples exhibited identical alterations on these genes.  This result implicates the importance to examine the genome in detailed when assessing clonality.  The identification of two genetically different tumours on the same tongue is significant, since each clonal population could ultimately respond to different therapeutic approaches.  For example, amplification of CCND1 was detected only in the clinically occult SCC, suggesting that this event happens at a later stage but is important for the formation of the SCC.  Furthermore, whole arm alterations of chromosome arms 5q, 8p, 156   and 8q were observed in all three biopsies, suggesting the lesions arose from a common progenitor harbouring these alterations.  However, because these events were present in different individuals regardless of their clonal origin, we cannot conclusively state if these were events of those in the progenitor cell.  Nevertheless, it is possible that the field was primed with these genetically altered cells but additional genetic alterations accumulate and evolve for the occurrence of the two tumours.  This work provides important implication for molecular targeted therapies in the future, as different subclones carrying different molecular alterations could be presented in a single oral cancer field.  6.3.  Significance 6.3.1. Genetic progression model of oral cancer The evolution of OSCC is known to result from the acquisition of multiple genetic events targeting different genes.  This principle is based on the following: cancer is a result of oncogene activation and tumour suppressor gene inactivation; a defined order of genetic events could lead to the development of a tumour phenotype; and the order of genetic events could varied but eventually it is the accumulation of genetic alterations that determines malignancy (Kim & Califano 2004).  In 1996, Califano et al. correlated genetic alterations and histopathological stages to establish the order of progression for specific genetic alterations in head and neck cancer (Califano et al. 1996). Microsatellite markers on chromosome 3p, 4q, 6p, 8, 9p, 11q13, 13q21, 14q, and 17p13 were used to analyze 87 preinvasive specimens (35 hyperplasias, 31 dysplasias, and 21 CIS) and 30 invasive tumours from 83 patients.  The highest occurrence of LOH occurred in hyperplasias at 9p21 (20%), then 3p21 (16%), and 17p13 (11%).  Based on the frequency of occurrence in different histological stages in preinvasive lesions, the authors concluded that these are early events in head and neck cancer progression, with 9p loss occurring from normal mucosa to hyperplasia, and 3p and 17p loss occurring from hyperplasia to dysplasia.  Subsequent genetic alterations accumulate 157   from dysplasia to CIS, which include 11q, 13q, 14q loss, while the malignant transformation from CIS to invasiveness include loss of 6p, 8, and 4q. Since this initial work by Califano et al., additional data have refined this progression model for oral cancer (Fig. 6.1).  Additional work - particularly the work outlined in this thesis - has defined narrower regions of alteration and flagged specific gene candidates that play a role in the earliest stages of oral tumourigenesis.  Data included in this thesis have shown that the degree of genomic dysregulation increases with histological stage, the specific regions of chromosome 3p that were highly associated with progression in low-grade dysplasias, and genes identified in amplicons that are potentially candidate oncogenes.  Further to this, promoter methylation of p16, mutation at p53, and protein overexpression of cyclooxygenease-2 (COX-2) have been identified in oral leukoplakias and were found to be increasing in frequency of occurrence when progressing from premalignancy to tumours (Boyle et al. 1993; Renkonen et al. 2002; Youssef et al. 2004).  While most of the studies of OPLs are limited by sample numbers, lack of follow- up information, or were tumour-adjacent instead of a primary lesion, these studies have served as a starting point to elucidate the genetic mechanisms of oral cancer development. The development of multifocal primary oral cancers open a new field of cancer research.  While most oral cancer is found to develop linearly with increasing histological stages from low-grade dysplasias to invasive tumours, oral cancer is also found to spread laterally across the field of oral mucosa by clonal expansion.  In this thesis, I have also addressed this question by comparing multifocal lesions from a single patient.  By inferring the sequence of genetic events by clonal ordering, common genetic alterations on chromosome 5q, 8p, and 8q were found in all samples, suggesting that these changes might occur early during clonal evolution.  In concordance with the genetic progression model, an accumulation of genetic alteration was also found as histological stage increases from moderate dysplasia to SCC. Importantly, different genetic alterations were identified in the two tumours, suggesting the occurrence of two genetic pathways.  In the future, more cases of multiple biopsies 158   within the field will be needed to thoroughly assess clonal expansion of cells, and genetic alterations occurring early in the pathway may be important events that prime the multifocal development of oral cancer. 6.3.2. Development of biomarker tools for evaluating progression risks of oral premalignant lesions One outcome of this thesis has been the development of a panel of genetic signatures associated with progression risk in oral dysplasias.  These genetic changes will be translated into a new genetic tool for clinical testing of progression risks in OPLs, which will help clinicians to prioritize treatment strategies.  During my thesis, I have developed a miniaturized DNA microarray called the OPL Risk Prediction chip to analyze DNA collected from patients during longitudinal monitoring study.  BAC clones were selected based on alterations identified in the profiled cases and negative controls that should show no copy number change were also arrayed on the chip for reference during analysis.  Analysis of OPL genomes using this chip should effectively identify lesions with a high risk of disease progression, flagging patients requiring more aggressive intervention and monitoring.  Additionally, molecular probes identified by our analysis of OPL genomes with long-term follow-up could also be used as markers for guiding the selection of individualized treatment regimens in the future.  6.4. Future Directions The data generated in this thesis have also raised additional questions that could further elucidate the mechanisms driving the progression of oral cancer. 6.4.1. Functional studies of candidate genes Our studies have identified genes residing in recurrent regions of genetic alteration that could have a role in the preinvasive stages of oral cancer progression.  Therefore, it is important to understand their biological role in oral carcinogenesis.  By mining public expression databases, it is now possible to identify genes that are frequently 159   overexpressed or underexpressed in oral cancers vs. normal samples.  Thus, to evaluate the functional impact of candidate oncogenes that are residing in the identified regions of alteration in OPLs, the expression of such genes should be often overexpressed in oral cancer samples relative to normal tissues.  Specifically, candidate genes in amplicons that were detected early in different samples of oral dysplasias (especially low-grade dysplasias that later progressed to cancer) that also showed overexpression should be evaluated for their oncogenic properties in vitro and in vivo. For example, knocking down the candidate genes in a oral cancer cell line that showed overexpression (driven by increased gene dosage) or stably overexpressing the candidate genes in normal primary oral keratinocytes could test candidate genes for different oncogenic properties including cell proliferation or anchorage-independent growth.  This would contribute to the understanding to its mechanism of action and give insight into oral carcinogenesis.  The identified genes critical for oral cancer development will be used to rationally develop therapeutic targets to treat oral cancer patients. 6.4.2. Biomarker tools for evaluating progression risks The current gold standard for treatment decision is histological assessment of OSCC and its margins (Poh et al. 2008).  As described in section 1.2.4, histological stage is scored according to standard criteria of the World Health Organization as hyperplasia, mild, moderate, or severe dysplasias, CIS, or invasive SCC.  Severe dysplasias and CIS are associated with the strongest risk for progression to invasiveness and thus are surgically treated in British Columbia, yet low-grade dysplasias are only ever followed in a "watchful waiting" manner; there are no standard interventions for low-grade lesions, even though some will eventually progress (Poh et al. 2008).  The generation of a DNA chip to objectively evaluate progression risks in small biological samples is the first step towards predicting progression risks in early stage samples.  However, the chip will need to be validated with a large number of collected samples of low-grade dysplasias in order to prove its efficacy.  Furthermore, the identified regions of genetic alteration will need to be further validated in a large number of low-grade dysplasias with clinical outcome.  This process will identify biomarkers and also improve the regions of genetic 160   alteration that could distinguish progressing from non-progressing low-grade dysplasias. The identification of a panel of genetic regions will be useful for the prediction of progression risks in low-grade dysplasias, thus prioritizing patient treatment and allow intervention at the earliest stage of the disease to improve patient survival. 6.4.3. Molecular characterization of second primary tumours In this thesis, I have discovered that two tumours close together on the tongue could exhibit tremendous genetic heterogeneity by performing whole genome analyses, establishing a method of using shared breakpoints to evaluate clonality.  Second primary tumours, which are tumours > 2 cm away from the index tumour, should be evaluated in this manner to clarify if the two lesions are indeed from a clonal origin or if they are genetically unrelated from each other, since they are more likely to harbour different genetic alterations.  The understanding of clonality would provide important implications to the development of multifocal oral cancer and thus facilitate the treatment of such diseases.  For example, if the second primary tumours were composed of a completely different genome than its index tumour, it is more likely that they arise by different mechanisms and thus both could be treated independently. However, if the second primary tumour shared the same genomic alterations as its index tumour, it is likely that the entire field is genetically impaired and thus more aggressive treatment might be needed. On the other hand, the accrual of multiple lesions from the same field would provide the opportunity for identifying early and late stage events, as alterations that are common to all lesions would imply early events, while genetic alterations that are exclusive to only one lesion would likely be a later stage change.  Although this thesis focus specifically on gene dosage alteration in OPLs, it is now apparent that other molecular mechanisms could act cooperatively or are independently important for cancer development.  Thus, integration of various high throughput data could refine the model of progression of the disease.  For example, if samples of sufficient DNA quantity and quality could be obtained, genetic analysis technologies including SNP chips and next-generation sequencing could advance our understanding of oral cancer.  By identifying the genes 161   altered by copy number alterations or DNA methylation which have direct downstream expression alterations, or by identifying the molecular pattern that govern the progression to invasiveness, the elucidation of the events that are important to the progression of OPLs to an invasive tumour will become clearer.  Therefore, along with gene dosage alterations in OPLs identified in this thesis, the throughout understanding of various molecular alterations causal to the progression of oral dysplasia to invasiveness would greatly improve prevention, treatment, and survival outcome for oral cancer patients in the future. Figure 6.1 162 Normal mucosa -9p Promoter methylation of p16 Overexpression of COX-2 11q13.2-q13.4 amplication Low-grade dysplasia High-grade dysplasia Invasive SCC -3p25.3-p26.1 -3p25.1-p25.3 -3p24.1 -3p21.31-p22.3 -3p14.2 -3p14.1 +8q22-23 -17p13.1 -18q22-qter +3q26-qter +20q +20p +5p15 -5q -4p +8q11-q21 +8q24-qter +17q11-22 +14q p53 mutation RAR-beta methylation EGFR activation 14-3-3 gamma methylation 2q11.2 amplication 4q12 amplication 8q11.21 amplication 8q22.3 amplication 9p13.3 amplication 17p LOH -8p -13q 14q LOH 6p LOH 4q LOH Figure 6.1  Genetic progression model for the development of oral cancer. Accumulation of genetic and epigenetic alterations has been found to parallel the multi- step model of oral cancer development.  The increased understanding of oral premalignant lesions will help the development of novel strategies for cancer prevention and treatment. 163    6.5. References Albertson, D. G., 2006. Gene amplification in cancer. Trends Genet, 22(8), 447-55. Albertson, D. G., C. Collins, F. McCormick & J. W. Gray, 2003. Chromosome aberrations in solid tumors. Nat Genet, 34(4), 369-76. Bedi, G. C., W. H. Westra, E. Gabrielson, W. Koch & D. Sidransky, 1996. Multiple head and neck tumors: evidence for a common clonal origin. Cancer Res, 56(11), 2484-7. Boyle, J. O., J. Hakim, W. Koch, P. van der Riet, R. H. Hruban, R. A. Roa, R. Correo, Y. J. Eby, J. M. Ruppert & D. Sidransky, 1993. The incidence of p53 mutations increases with progression of head and neck cancer. Cancer Res, 53(19), 4477- 80. Braakhuis, B. J., M. P. Tabor, J. A. Kummer, C. R. Leemans & R. H. Brakenhoff, 2003. A genetic explanation of Slaughter's concept of field cancerization: evidence and clinical implications. Cancer Res, 63(8), 1727-30. Cairns, P., T. J. Polascik, Y. Eby, K. Tokino, J. Califano, A. Merlo, L. Mao, J. Herath, R. Jenkins, W. Westra & et al., 1995. Frequency of homozygous deletion at p16/CDKN2 in primary human tumours. Nat Genet, 11(2), 210-2. Califano, J., P. van der Riet, W. Westra, H. Nawroz, G. Clayman, S. Piantadosi, R. Corio, D. Lee, B. Greenberg, W. Koch & D. Sidransky, 1996. Genetic progression model for head and neck cancer: implications for field cancerization. Cancer Res, 56(11), 2488-92. Chakraborty, S. B., S. Dasgupta, A. Roy, A. Sengupta, B. Ray, S. Roychoudhury & C. K. Panda, 2003. Differential deletions in 3p are associated with the development of head and neck squamous cell carcinoma in Indian patients. Cancer Genet Cytogenet, 146(2), 130-8. 164   Chung, C. H., P. S. Bernard & C. M. Perou, 2002. Molecular portraits and the family tree of cancer. Nat Genet, 32 Suppl, 533-40. Chung, C. H., J. S. Parker, G. Karaca, J. Wu, W. K. Funkhouser, D. Moore, D. Butterfoss, D. Xiang, A. Zanation, X. Yin, W. W. Shockley, M. C. Weissler, L. G. Dressler, C. G. Shores, W. G. Yarbrough & C. M. Perou, 2004. Molecular classification of head and neck squamous cell carcinomas using patterns of gene expression. Cancer Cell, 5(5), 489-500. Coe, B. P., B. Ylstra, B. Carvalho, G. A. Meijer, C. Macaulay & W. L. Lam, 2007. Resolving the resolution of array CGH. Genomics, 89(5), 647-53. Escher, A., E. Piotet, F. Waridel, R. Iggo & P. Monnier, 2009. p53 Mutation in histologically normal mucosa of the aero-digestive tract is not a marker of increased risk for second primary carcinoma in head and neck cancer patients. Eur Arch Otorhinolaryngol, 266(4), 547-51. Jarvinen, A. K., R. Autio, S. Haapa-Paananen, M. Wolf, M. Saarela, R. Grenman, I. Leivo, O. Kallioniemi, A. A. Makitie & O. Monni, 2006. Identification of target genes in laryngeal squamous cell carcinoma by high-resolution copy number and gene expression microarray analyses. Oncogene, 25(52), 6997-7008. Jarvinen, A. K., R. Autio, S. Kilpinen, M. Saarela, I. Leivo, R. Grenman, A. A. Makitie & O. Monni, 2008. High-resolution copy number and gene expression microarray analyses of head and neck squamous cell carcinoma cell lines of tongue and larynx. Genes Chromosomes Cancer, 47(6), 500-9. Kim, M. M. & J. A. Califano, 2004. Molecular pathology of head-and-neck cancer. Int J Cancer, 112(4), 545-53. Lin, C. J., J. R. Grandis, T. E. Carey, S. M. Gollin, T. L. Whiteside, W. M. Koch, R. L. Ferris & S. Y. Lai, 2007. Head and neck squamous cell carcinoma cell lines: established models and rationale for selection. Head Neck, 29(2), 163-88. 165   Mao, L., J. S. Lee, Y. H. Fan, J. Y. Ro, J. G. Batsakis, S. Lippman, W. Hittelman & W. K. Hong, 1996. Frequent microsatellite alterations at chromosomes 9p21 and 3p14 in oral premalignant lesions and their value in cancer risk assessment. Nat Med, 2(6), 682-5. Merlo, L. M., J. W. Pepper, B. J. Reid & C. C. Maley, 2006. Cancer as an evolutionary and ecological process. Nat Rev Cancer, 6(12), 924-35. Partridge, M., S. Pateromichelakis, E. Phillips, G. G. Emilion, R. P. A'Hern & J. D. Langdon, 2000. A case-control study confirms that microsatellite assay can identify patients at risk of developing oral squamous cell carcinoma within a field of cancerization. Cancer Res, 60(14), 3893-8. Pateromichelakis, S., M. Farahani, E. Phillips & M. Partridge, 2005. Molecular analysis of paired tumours: time to start treating the field. Oral Oncol, 41(9), 916-26. Persson, F., Y. Andren, M. Winnes, B. Wedell, A. Nordkvist, G. Gudnadottir, R. Dahlenfors, H. Sjogren, J. Mark & G. Stenman, 2009. High-resolution genomic profiling of adenomas and carcinomas of the salivary glands reveals amplification, rearrangement, and fusion of HMGA2. Genes Chromosomes Cancer, 48(1), 69-82. Poh, C. F., L. Zhang, D. W. Anderson, J. S. Durham, P. M. Williams, R. W. Priddy, K. W. Berean, S. Ng, O. L. Tseng, C. MacAulay & M. P. Rosin, 2006. Fluorescence visualization detection of field alterations in tumor margins of oral cancer patients. Clin Cancer Res, 12(22), 6716-22. Renkonen, J., H. Wolff & T. Paavonen, 2002. Expression of cyclo-oxygenase-2 in human tongue carcinoma and its precursor lesions. Virchows Arch, 440(6), 594- 7. Reshmi, S. C., X. Huang, D. W. Schoppy, R. C. Black, W. S. Saunders, D. I. Smith & S. M. Gollin, 2007a. Relationship between FRA11F and 11q13 gene amplification in oral cancer. Genes Chromosomes Cancer, 46(2), 143-54. 166   Reshmi, S. C., S. Roychoudhury, Z. Yu, E. Feingold, D. Potter, W. S. Saunders & S. M. Gollin, 2007b. Inverted duplication pattern in anaphase bridges confirms the breakage-fusion-bridge (BFB) cycle model for 11q13 amplification. Cytogenet Genome Res, 116(1-2), 46-52. Rosin, M. P., X. Cheng, C. Poh, W. L. Lam, Y. Huang, J. Lovas, K. Berean, J. B. Epstein, R. Priddy, N. D. Le & L. Zhang, 2000. Use of allelic loss to predict malignant risk for low-grade oral epithelial dysplasia. Clin Cancer Res, 6(2), 357- 62. Rosin, M. P., W. L. Lam, C. Poh, N. D. Le, R. J. Li, T. Zeng, R. Priddy & L. Zhang, 2002. 3p14 and 9p21 loss is a simple tool for predicting second oral malignancy at previously treated oral cancer sites. Cancer Res, 62(22), 6447-50. Scholes, A. G., J. A. Woolgar, M. A. Boyle, J. S. Brown, E. D. Vaughan, C. A. Hart, A. S. Jones & J. K. Field, 1998. Synchronous oral carcinomas: independent or common clonal origin? Cancer Res, 58(9), 2003-6. Shin, D. M., J. S. Lee, S. M. Lippman, J. J. Lee, Z. N. Tu, G. Choi, K. Heyne, H. J. Shin, J. Y. Ro, H. Goepfert, W. K. Hong & W. N. Hittelman, 1996. p53 expressions: predicting recurrence and second primary tumors in head and neck squamous cell carcinoma. J Natl Cancer Inst, 88(8), 519-29. Snijders, A. M., B. L. Schmidt, J. Fridlyand, N. Dekker, D. Pinkel, R. C. Jordan & D. G. Albertson, 2005. Rare amplicons implicate frequent deregulation of cell fate specification pathways in oral squamous cell carcinoma. Oncogene, 24(26), 4232-42. Squire, J. A., J. Bayani, C. Luk, L. Unwin, J. Tokunaga, C. MacMillan, J. Irish, D. Brown, P. Gullane & S. Kamel-Reid, 2002. Molecular cytogenetic analysis of head and neck squamous cell carcinoma: By comparative genomic hybridization, spectral karyotyping, and expression array analysis. Head Neck, 24(9), 874-87. 167   Swanton, C. & C. Caldas, 2009. Molecular classification of solid tumours: towards pathway-driven therapeutics. Br J Cancer, 100(10), 1517-22. Tabor, M. P., R. H. Brakenhoff, H. J. Ruijter-Schippers, J. A. Kummer, C. R. Leemans & B. J. Braakhuis, 2004. Genetically altered fields as origin of locally recurrent head and neck cancer: a retrospective study. Clin Cancer Res, 10(11), 3607-13. Tsui, I. F., C. F. Poh, C. Garnis, M. P. Rosin, L. Zhang & W. L. Lam, 2009. Multiple pathways in the FGF signaling network are frequently deregulated by gene amplification in oral dysplasias. Int J Cancer, 125(9), 2219-28. van Houten, V. M., C. R. Leemans, J. A. Kummer, J. Dijkstra, D. J. Kuik, M. W. van den Brekel, G. B. Snow & R. H. Brakenhoff, 2004. Molecular diagnosis of surgical margins and local recurrence in head and neck cancer patients: a prospective study. Clin Cancer Res, 10(11), 3614-20. Warnakulasuriya, S., 2009. Global epidemiology of oral and oropharyngeal cancer. Oral Oncol, 45(4-5), 309-16. Worsham, M. J., S. R. Wolman, T. E. Carey, R. J. Zarbo, M. S. Benninger & D. L. Van Dyke, 1995. Common clonal origin of synchronous primary head and neck squamous cell carcinomas: analysis by tumor karyotypes and fluorescence in situ hybridization. Hum Pathol, 26(3), 251-61. Youssef, E. M., D. Lotan, J. P. Issa, K. Wakasa, Y. H. Fan, L. Mao, K. Hassan, L. Feng, J. J. Lee, S. M. Lippman, W. K. Hong & R. Lotan, 2004. Hypermethylation of the retinoic acid receptor-beta(2) gene in head and neck carcinogenesis. Clin Cancer Res, 10(5), 1733-42. 168   Appendices 169   Appendix A - Supplementary data for chapter 2 Supplemental Table A.1.  Genetic  alterations in A-253. Chromosome *Proximal flanking clone BP start *Distal flanking clone BP end  gain/loss 1p36.21-p35.2 N0583G23 12724432 N0294J15 31011259 Loss 1q23.3-q44 N0005K23 159680000 N0088N11 244554020 Loss 1q44 N0694B09 246281233 N0068F13 246833917 Loss  2q11.2 N0297M15 97044699 N0480K03 97525604 Loss  3p24.3-p21.31 M2007O17 20621949 N0565H07 45851307 Loss 3p14.3 N0652P22 55012149 N0164I11 107818058 Loss 3p14.3-p11.1 N0257K23 108174908 N0371F02 90114684 Loss 3q11.2-q23 N0818I19 95495994 N0323M13 140512234 Loss 3q23-q26.1 N0657M13 140354733 N0141G19 163466443 Gain 3q26.2-q29 N0170G01 172207871 N0721B03 199276959 Gain  4p16-p11 N0803H22 4103494 N0259G19 49416673 Loss 4q12-q35.2 N0061F05 52478915 N0463B04 190387639 Loss  5p15.33-p15.2 N0811I15 70263 N0219M15 11571401 Gain 5p15.2-p13.2 N0597N01 13215928 N0513I05 37964379 Gain 5p13.2 N0134I03 37793950 M2130F23 38119276 High-level amplification 5p13.2-p12 N0695L19 38085538 N0485N10 45058856 Gain 5q33.2-q33.3 N0096J04 152764659 N0725E20 159722796 Gain 5q34-q35.3 N0103J09 167235713 N0271H04 180444136 Gain  6p25.3-p22.3 N0812K10 101435 N0723K08 22116373 Loss 6p22.2-p22.1 N0018H02 25459908 N0133I23 29828522 Loss 6p21.2-p12.1 N0685F06 39096919 N0346D21 55699936 Loss 6p12.1 N0619C09 55796805 M2170G15 57181727 Gain  7p22.3-p21.3 N0379K15 42475 N0702K06 9273168 Gain 7p21.2-p11.1 M2008I15 13814181 N0769I04 57964582 Gain 7q21.11-q22.3 N0343P19 83645022 N0621C24 105950063 Gain 7q31.32-q34 N0636H22 120838501 N0175F11 137683657 Loss 7q34-q36.3 N0202L12 138833832 N0083D03 158777885 Loss 170   Chromosome *Proximal flanking clone BP start *Distal flanking clone BP end  gain/loss 8p23.3-p23.2 N0809I19 92474 N0567H20 3899414 High-level loss 8p23.2-p23.1 N0404K15 4558701 N0231L04 7008487 Gain 8p23.1 N0738O21 7035076 N0115E11 7289755 Loss 8p23.1 N0435L12 8168448 N0350L10 12274715 Gain 8p22 N0447G11 16105555 N0496H06 16283946 Gain 8p22 N0771E22 16150756 M2005M04 17842202 High-level amplification 8p22-p21.2 N0794M05 17911107 N0064H23 25668123 Gain 8p21.2-p11.21 N0045M09 25799620 N0628L14 42282917 Loss 8q12.2-q12.3 N0265K14 62006277 N0265J24 62913037 Gain 8q13.1 N0707M03 66675218 N0164L07 67299897 Gain 8q13.2-q22.2 N0662K10 68317889 N0616L06 101344433 Loss 8q22.2-q23.3 N0321E07 101234578 N0393N12 113117761 Gain 8q23.3-q24.13 M2118B16 116911101 N0048I02 123218725 Gain 8q24.13-q24.3 N0497O14 123499459 N0620H01 145740218 Gain  9p24.3-p24.1 N0143M01 23994 N0133H08 7589941 Gain 9p23-p21.3 N0020G14 12381784 N0666A10 22711716 Loss 9p21.1-p12 N0632I19 32152951 N0204I02 40646937 Gain 9q21.31-q22.33 N0685G21 83221598 N0525D15 101222908 Gain 9q31.1-q33.1 N0354J03 105921993 N0676D12 120767800 Gain 9q33.1-q34.3 N0526H07 121889659 N0668B20 140180127 Gain  10p15.3-p11.22 N0797F08 75055 N0166N17 32689878 Loss 10q11.21-q22.1 N0684F09 42284337 N0241D23 73087618 Loss 10q22.2-q23.32 N0727O24 76667248 N0368O06 93633145 Loss  11p15.3-p14.1 N0479I06 11106340 N0045A08 29465680 Loss 11p14.1 N0113G14 30076330 N0113G14 30232485 Gain 11p14.1-p13 M2028C20 30243665 N0626M07 31111143 High-level amplification 11p13 F0593L04 31255594 N0112O22 31736320 Gain 11p13 N0702F20 31610373 N0460O21 33088592 High-level amplification 11p13 N0348A11 33006534 N0115L22 33335529 Gain 11p13-p12 N0319G20 33304228 N0113E23 38238634 High-level amplification 11p12 N0787M10 38173026 N0590M03 39998462 Gain 11p12 N0059P20 39890676 N0664F06 41151775 Loss 11q13.1-q13.2 N0067F01 63186963 N0770J12 68335576 Gain 11q13.2-q13.3 N0154D10 68278660 F0584O24 70374068 High-level amplification 11q13.4 N0660L16 70766779 N0652N03 71835286 Loss 171   Chromosome *Proximal flanking clone BP start *Distal flanking clone BP end  gain/loss 11q14.1-q25 N0120M20 78123405 M2013A02 134449252 Loss  12p13.33-p11.1 M2094C14 6223 N0358D05 34306741 Loss 12q21.1-q24.23 N0172N22 74047433 N0663G13 118621994 Loss 12q24.31-q24.33 N0665C13 123372348 N0386I08 132178738 Loss  14q22.2-q24.1 N0111K05 53489717 N0350H11 68130025 Gain 14q24.1-q24.2 N0035D12 68241628 N0003C16 70533879 High-level amplification 14q24.2 N0396P17 70364482 N0757P24 71263950 Gain 14q24.2 N0656P07 71515335 N0746C04 72636951 High-level amplification 14q24.2-q24.3 N0799E22 72570042 N0539P04 73838478 Gain 14q24.3 N0513N07 73806697 N0536A05 74277120 High-level amplification 14q24.3 N0804J22 74199495 N0316E14 74589417 Gain 14q24.3 N0340F04 74661788 N0192P20 75800664 High-level amplification 14q24.3 N0745K02 75874030 N0488C13 76489079 Gain 14q24.3 N0691J02 76401596 N0101J07 77046786 High-level amplification 14q24.3 N0296C19 76976459 N0084M22 78016279 Gain 14q24.3-q31.1 N0192D17 78011540 N0464L15 78490529 High-level amplification 14q31.1 N0023H16 78376380 N0259M02 81025446 Gain 14q31.1-q32.11 N0666E17 81506479 M2264D05 89138305 Loss 14q32.13-q32.2 N0357C08 93699417 N0415J21 98569435 Loss  15q11.2-q14 N0118E23 18473375 N0104M19 37846674 Loss 15q21.1-q22.1 N0589O14 43039079 N0676H16 56507014 Loss 15q22.2-q26.1 N0092C13 58245427 M2100I05 87625471 Loss 15q26.1-q26.3 N0084E23 89803652 N0327K11 99484121 Loss  16p13.3-p11.2 N0766H16 235763 M2120G21 30153065 Gain  17p13.3-p11.2 N0411G07 440728 N0423O14 22154806 Gain 17q21.2 N0639G19 36769177 N0312E22 37543627 Gain  18q11.2 N0595B24 17830190 N0149B10 19558365 Gain 18q12.1-q12.2 N0479H18 28920522 N0713H10 32677274 Loss 18q12.2 N0460O16 32970382 N0278M16 33225616 Loss 18q12.2 N0401E16 33328621 M2214A18 34227319 High-level loss 18q12.2-q12.3 N0062M21 34297527 N0671B06 36743158 Loss 18q12.3 N0006J17 36792912 N0440F22 37750265 High-level loss 172   Chromosome *Proximal flanking clone BP start *Distal flanking clone BP end  gain/loss 18q12.3 N0010L17 37623799 N0599F19 40259437 Loss 18q12.3 N0286D19 40237310 N0730A15 41312820 High-level loss 18q12.3-q21.2 N0749H17 41274578 N0099A01 50702087 Loss 18q21.2 N0805M18 50573728 N0344A12 51312752 High-level loss 18q21.2-q22.3 N0727C01 51431102 N0714E20 70126378 Loss 18q22.3-q23 N0706A18 69973711 N0711F02 71525297 High-level loss 18q23 N0044O12 71463133 N0618B13 73947143 Loss 18q23 N0012G20 73916944 M2320P10 74815544 High-level loss 18q23 N0528G04 74774394 N0007H17 75402240 Loss 18q23 N0020N12 75444530 N0609N15 75920946 High-level loss  19p13.11 M2270B05 18500928 N0715L15 19436075 Gain 19q13.32-q13.33 N0052M23 52838341 N0808J04 54928888 Gain 19q13.41-q13.42 N0650A02 59051048 N0812P06 59353980 Gain  20p13-p12.1 N0640A09 60369 M2080M11 12170047 Gain 20p12.1-p11.1 N0667G11 17370079 N0108H13 26267569 Gain 20q11.1-q13.33 M2001C04 28110959 N0476I15 62435964 Gain  21q22.12-q22.3 N0272A03 34768331 N0513H10 46823167 Loss 22q11.21-q12.1 N0795B02 16444061 N0266K21 25217072 Gain  * Clones beginning with N0 or F0 belong to the Roswell park Cancer Institute libraries 11 and 13 (RP11 and RP13), respectively; those beginning with M belong to the Caltech-D (CTD) library. 173   Supplemental Table A.2.  Genetic alterations in Cal27.  Chromosome *Proximal flanking clone BP start *Distal flanking clone BP end  Gain/Loss 3p26.3-p25.3 N0739I20 164940 N0525N21 10555199 Loss 3p25.3 N0443M10 10549698 N0126L09 10941986 High-level loss 3p25.3-p21.31 N0020D21 10947419 N0527M10 46115480 Loss 3p21.31-p12.1 N0171C03 49611358 N0047J16 85081574 Loss 3p12.1 N0639P09 85446931 N0259G06 86164592 Gain 3q11.2-q13.31 N0631O04 95307753 N0757K23 115939012 Gain 3q22.1-q25.31 N0086E06 132662234 N0674E16 157089982 Gain 3q25.31- q25.32 N0294L13 157153239 N0032F04 158999959 High-level amplification 3q25.32-q26.1 N0294L13 158991029 N0047D12 163423305 Gain 3q26.2-q29 N0252J01 170273140 N0721B03 199276959 Gain  4p16.2-p15.1 N0658B07 4268401 N0483M09 31746082 Loss 4p14-p13 N0143G24 34578447 N0758P14 44312206 Gain 4q12-q35.2 N0365H22 52354875 M2011O21 191137774 Loss  6q24.2 N0343F18 145181816 N0792P19 145720145 High-level amplification  7p22.3-p11.2 N0379K15 42475 N0023F04 55049738 Gain 7p11.2 N0781C22 54955463 N0788K14 55487992 High-level amplification 7p11.2-p11.1 N0535N12 55619477 N0415F22 57562822 Gain  8p23.3 N0130K11 36943 F0586C17 144486 Gain 8p23.3-p11.1 N0111E15 156973 N0527M20 43912178 Loss  9p24.1-p23 N0075C09 8843158 N0006H18 9873547 Loss 9p13.1-p11.2 N0780I22 38963616 N0192J07 46458217 Gain 9q arm N0176A22 66550151 N0668B20 140180127 Gain  10p15.3- p11.21 N0797F08 75055 N0016G14 34884348 Loss 10q11.22 N0342C24 46174281 N0138M08 46723088 Loss  11p14.2-p13 N0622I22 26533680 N0464G18 33913102 Gain 11p13-p12 N0448B09 33941570 M2276M15 38095090 High-level amplification 174   Chromosome *Proximal flanking clone BP start *Distal flanking clone BP end  Gain/Loss 11q13.1-q22.3 N0067F01 63186963 N0693N09 104352171 Gain 11q22.3-q23.2 N0027F21 105035952 N0796G18 114272072 Loss 11q23.2 N0271A21 114281321 N0713B09 114600734 High-level loss 11q23.2-q24.2 N0326G04 114627272 N0046C21 125170196 Loss 11q24.2 N0068J04 125280711 N0115C10 126114304 High-level loss 11q24.2-q25 N0281D23 126559530 M2013A02 134449252 Loss  14q11.2 N0597A11 19149713 M2326H15 19892330 Gain 14q11.2 N0777M14 19883965 N0334O19 20634203 High-level amplification  15q11.2 N0207G06 18272238 N0657N12 21159615 Loss  18p11.32- p11.21 N0271E15 60305 N0107F13 12918621 Gain 18q11.2-q12.2 N0678O03 18245092 N0460O16 33123956 Loss 18q12.2 N0278M16 33040900 N0401E16 33532460 High-level loss 18q12.2-q22.3 N0019D11 33450006 N0185K07 70269608 Loss 18q22.3-q23 N0027C07 70635327 N0711F02 71525297 High-level loss 18q23 N0044O12 71463133 N0618B13 73947143 Loss 18q23 N0012G20 73916944 M2320P10 74815544 High-level loss 18q23 N0528G04 74774394 N0805C24 75774454 Loss  19p13.3 N0348B12 4960407 N0222E10 6694652 Gain  20q11.22- q13.33 N0818N12 31620526 N0799O09 62043404 Gain *Clones beginning with N0 or F0 belong to the Roswell park Cancer Institute libraries 11 and 13 (RP11 and RP13), respectively; those beginning with M belong to the Caltech-D (CTD) library. 175   Supplemental Table A.3.  Genetic alterations in SCC-15. Chromosome *Proximal flanking clone Start BP *Distal flanking clone End BP Gain/Loss 1p36.32- p36.21 N0493P12 4695248 N0486C01 15522588 gain 1q21.1-q23.2 N0561P10 142533622 N0106F15 157766153 loss 1q32.1-q43 N0109H10 197420265 N0068D10 234541366 gain  3p26.3-p12.1 N0359E09 46140 N0788E22 84926864 loss 3p12.1 N0268H24 85151859 N0041O13 85755688 gain 3p12.1 N0036I16 85719367 N0259G06 86164592 high-level amplification 3p12.1 N0639H15 86143774 N0144C05 86448726 gain 3q22.3-q29 N0607A04 139977924 N0721B03 199276959 gain  4q21.1-q35.2 M2021N10 77841560 M2011O21 191137774 loss  5q11.1-q11.2 N0317O24 50100507 N0656M19 56324583 gain 5q11.2-q12.1 N0124J05 56299635 N0015O11 59204955 loss 5q14.1-q31.1 N0031A08 80832453 N0478I21 134231293 loss 5q31.1-q35.3 N0046B14 134189806 N0324K20 180715161 gain  6p25.3 N0328C17 213736 N0737K22 1462376 loss 6p22.3-p12.3 N0758O07 18309920 N0593F20 47497414 gain 6q22.2-q24.2 N0059K17 117555636 N0649D16 144800716 gain 6q24.2-q24.3 N0043G16 145358044 N0619O01 147473656 loss 6q25.3-q27 N0816H20 158219236 N0159J07 170880179 gain  7p22.3-p21.3 N0379K15 42475 N0722J11 9232952 gain 7p21.3-p11.2 N0237B05 11351482 N0164O17 54964154 gain 7p11.2 N0023F04 54880028 N0788K14 55487992 high-level amplification 7p11.2-p11.1 N0535N12 55619477 N0769I04 57964582 gain 7q11.21- q11.22 N0096H22 64648077 N0793E12 67373031 gain  8p23.3-p11.1 N0111E15 156973 N0527M20 43912178 loss 8q11.1-q24.3 M2192G08 47102600 N0639O03 146236298 gain  9p24.3-p11.2 N0770G15 118663 N0192J07 46458217 loss 9q12 N0499O03 66460771 N0087H09 68436032 loss 176   Chromosome *Proximal flanking clone Start BP *Distal flanking clone End BP Gain/Loss 9q13-q34.3 N0151G22 70224827 N0035I18 140185705 gain  10q11.21 N0290I03 42520693 N0558P16 43496018 gain  11p11.2- p11.12 N0328B19 48468561 N0788L06 49626739 loss 11q11-q12.2 N0626N06 55139256 N0507N10 60715862 loss 11q13.1-q13.4 N0485O09 64243759 N0684B02 71041286 gain 11q13.2-q13.4 N0554A11 68509551 N0684B02 71041286 high-level amplification 11q13.4-q22.1 N0652N03 71684844 N0067N19 99417302 loss 11q22.1-q22.3 N0515J12 99567448 N0370I20 103586437 gain 11q22.3-q25 N0006A04 104123345 N0715D10 134331374 loss  12q24.32- q24.33 N0360E11 124562549 M2140B24 132289534 loss  13q11-q14.2 N0563G05 18012966 N0745E18 46518008 loss 13q14.2-q21.1 N0438K10 47570684 M2008A04 53938114 loss 13q21.1 N0524P04 53806575 N0490K15 54238572 gain 13q21.1-q21.2 N0795G23 54114672 N0430I03 59746364 loss 13q21.33- q31.1 N0174G22 69746384 N0384H03 79222509 gain 13q31.3-q34 N0511F12 89836987 N0226B11 114117194 gain  14q11.2-q12 N0334O19 20452364 N0089K22 24715368 gain  17p13.3-p11.2 N0411G07 440728 M2103O19 21834063 gain 17q22 N0014N16 52348422 N0155E20 53971269 gain 17q23.3-q24.2 N0478O18 59084371 N0484P03 63378726 gain 17q24.3-q25.3 N0775A04 67923802 N0196O11 78615238 gain  18q12.2-q12.3 N0367D23 34509586 N0080D13 37213143 gain 18q12.3-q23 N0752A01 37252671 N0467N12 75337596 loss  19q12-q13.43 N0813H01 33879599 N0493D23 63654245 gain  20p13 N0640A09 60369 M2125K22 4893679 gain 20q11.21- q13.33 N0004O09 29737702 N0476I15 62435964 gain 177   * Clones beginning with N0 or F0 belong to the Roswell park Cancer Institute libraries 11 and 13 (RP11 and RP13), respectively; those beginning with M belong to the Caltech-D (CTD) library. 178   Supplemental Table A.4.  Genetic alterations in SCC-4.  Chromosome *Proximal flanking clone BP start *Distal flanking clone BP end  gain/loss 1q21.1 N0082K10 142758440 F0488D18 147215140 loss 1q21.3-q24.2 N0425D18 150921354 M2017D09 165505019 loss 1q24.2 N0381N09 165669462 N0101N07 167445692 gain 1q25.2-q25.3 N0620B08 177945800 N0703I24 183127629 gain 1q31.3-q44 N0199D04 196799511 M2325H09 243354271 gain  2p25.3 N0371D08 148491 N0314C15 603943 High-level loss 2p21-p16.1 N0715E22 47145052 N0144H11 60263213 loss 2p15-p11.2 N0017L22 62592151 N0264J03 83899360 loss 2q11.1-q12.3 N0645D10 94986258 N0725A12 106939158 gain 2q32.1-q35 N0654M09 186641883 N0798J19 216464288 loss 2q37.1 N0420E18 231168948 N0562I05 232133311 gain  3p26.3-p21.31 N0359E09 46140 N0307M13 45786452 loss 3p21.31-p11.1 N0694I15 49081051 N0084M14 89962384 loss 3q27.1-q29 N0660P23 185035157 N0192L23 199240277 gain  4q33-q35.2 N0344G13 172113115 M2011O21 191137774 loss  5p15.33 N0811I15 70263 N0348B13 289075 gain 5p15.33 N0656C20 350724 N0322L05 1013978 high-level amplification 5p15.33 N0124I24 1321361 N0161F13 1724268 gain 5p15.32-p15.33 M2012J19 1650166 N0580F06 4608531 high-level amplification 5p15.32 N0727I08 4530567 N0655D02 5143562 gain 5p15.31-p15.32 N0423A16 5090122 N0014C04 6798819 high-level amplification 5p15.31 N0001F12 6792039 N0046O23 7549647 gain 5p15.31 N0753D07 7664722 N0315E16 8118958 high-level amplification 5p15.2-p15.31 N0335F15 8127955 N0030B17 10636898 gain 5p15.2 N0584F21 10697055 N0230E09 11450830 high-level amplification 5p15.2 N0639D09 11468507 N0313K15 13144667 gain 5p15.2 N0619G03 13126356 N0103J10 13913794 high-level amplification 5p15.2 N0458F04 13788915 N0683D22 14267453 gain 179   Chromosome *Proximal flanking clone BP start *Distal flanking clone BP end  gain/loss 5p15.1-p15.2 N0020B15 14216549 N0260E18 16775309 high-level amplification 5p15.1-p13.3 N0090B23 16794671 N0743K06 31315706 gain 5p13.3 N0705N10 31279477 N0784O19 31682559 high-level amplification 5p13.3-p12 N0005N11 31657478 N0655E08 45792125 gain 5q11.1-q13.1 N0555F17 49479231 N0758A22 67144270 loss 5q13.2-q33.1 N0195E02 70267431 N0373N22 147782108 loss  7p22.3 N0379K15 42475 N0613E08 1839666 gain 7p22.3-p11.2 N0325O09 1803168 N0535N12 55799090 gain 7q21.11 N0750F10 84076954 N0729A18 85516425 high-level amplification 7q21.11 N0561F17 85504879 N0555E22 86017229 gain 7q21.12-q21.13 N0669A08 86097224 N0756A17 88632733 high-level amplification 7q21.13 M2326K17 88637853 N0702P09 88843764 gain 7q21.13 N0360F03 88843766 N0584M11 89511255 high-level amplification 7q21.2-q21.3 N0454K03 89490514 N0623E20 91536054 gain 7q21.2-q21.3 N0339M03 91710603 M2130O12 94103065 high-level amplification 7q21.3 N0564F14 94209394 N0674B12 95073399 gain 7q21.3 N0084F19 95014255 N0002N22 96533845 high-level amplification 7q21.3 N0063H09 96540389 N0380G21 97485965 gain 7q21.3-q22.1 N0526I04 97475547 N0694E14 98662833 high-level amplification 7q22.1 N0140D10 98657508 N0336D07 100193425 gain 7q22.1 F0650G11 100363784 N0484K16 100703738 high-level amplification 7q22.1 N0151L12 100615567 N0596H08 101044206 gain 7q22.1-q36.3 N0226H09 102474991 N0133J16 158689828 loss  8p23.3-p23.2 N0111E15 156973 N0593J22 4611342 loss 8p11.23-p11.1 N0706M17 39387128 N0567K04 43443746 loss 8q11.1-q11.22 M2192G08 47102600 N0401N18 51002937 gain 8q11.23-q12.2 N0749I21 55357897 N0113D04 62245403 gain 8q12.3 N0429B03 62590375 N0699L02 63032213 high-level amplification 8q12.3 N0607I22 63071179 M2038A10 63455889 gain 8q12.3 N0577N01 63487191 N0694K24 63859487 high-level amplification 180   Chromosome *Proximal flanking clone BP start *Distal flanking clone BP end  gain/loss 8q12.3 N0019J07 64088419 N0450M17 65067875 gain 8q21.2 N0509F16 86620595 N0626K06 87188682 gain 8q21.3-q23.1 N0017C13 88722314 N0367G07 110178392 gain 8q23.3-q24.3 N0506I07 116821840 N0620H01 145740218 gain  9p24.3-p11.2 N0143M01 23994 N0076I05 46275010 loss 9q12-q22.31 N0499O03 66460771 N0612N07 94906426 loss  10p15.3-p11.21 N0797F08 75055 N0456H11 37012274 loss  11q12.2-q13.2 N0565O16 60420899 N0211G23 68942928 gain 11q13.2-q13.4 M2009H02 68979527 N0013K23 73003111 high-level amplification 11q13.4-q13.5 N0791C12 72899242 N0217K21 76536143 gain 11q13.5-q25 N0057M20 76603134 N0715D10 134331374 loss  12p13.2-p12.2 N0202N01 10969578 N0746G12 20116308 gain 12p12.1-p11.22 N0597C21 23626836 N0747N08 28567392 gain 12q21.32-q23.3 N0434B06 85896769 N0558G19 107369930 loss 12q24.13- q24.33 N0155B05 111700212 M2140B24 132289534 loss  13q12.11-q13.1 N0717M17 19505536 N0223P17 31971854 gain 13q21.33-q31.1 N0459I24 70791339 N0342E06 78757439 gain 13q31.3-q34 N0133E12 93005803 N0051H10 113658045 gain  14q21.3-q32.12 N0777E01 44520774 N0386D20 91728059 gain 14q32.12- q32.13 N0797O16 91699568 N0559B23 95208255 high-level amplification 14q32.13-q32.2 N0241N04 95526524 N0433J08 96319344 gain 14q32.2-q32.33 N0208P19 96434336 N0359N05 96823785 high-level amplification 14q32.2 N0061O01 97581672 N0430I09 98728060 gain 14q32.2 N0634B02 98636279 N0594K17 98953703 high-level amplification 14q32.2-q32.33 N0693H21 99081608 N0205J08 103570838 gain 14q32.33 N0487L08 103575413 F0537K06 104155663 high-level amplification 14q32.33 N0576C15 104235870 N0249M16 106215608 gain  15q11.2-q13.1 N0492D06 18593750 N0374K05 27477714 loss 181   Chromosome *Proximal flanking clone BP start *Distal flanking clone BP end  gain/loss 15q22 N0693C20 57945260 N0411L19 59000310 gain  16p13.12-p13.3 235507 235507 N0519H10 13607936 gain  17p12-p13.1 N0184D16 10585717 N0726O12 15374763 loss 17p11.2-p12 N0378O18 15297799 N0092B11 16646713 gain 17p11.2 N0655O21 16624310 N0801O07 16928970 high-level amplification 17p11.2 N0809H20 17534762 N0358A04 18972545 gain 17q11.1-q11.2 N0102E01 22320285 N0044H05 28214642 loss 17q11.2-q12 N0044D22 28205502 N0329H16 29694876 High-level loss 17q12 N0668B11 29842549 N0072I20 33700089 loss 17q25.1 N0751O16 68898787 N0629D12 70527055 loss 17q25.3 N0013K12 73020577 N0196O11 78615238 loss  18p11.32- p11.21 N0683L23 35421 N0216E19 14975939 loss 18q11.2-q22.3 N0403A21 19728082 N0781A07 67296014 loss  19q13.41- q13.43 N0791E24 58652119 N0493D23 63654245 gain  20p11.21-p11.1 N0755M18 22685102 N0108H13 26267569 gain 20q11.21- q13.33 N0559K10 29297109 N0134L13 62424560 gain  21q11.2-q22.3 N0025L22 13879419 N0513H10 46823167 loss  22q11.21 N0004G23 19242994 M2245I11 20280089 gain 22q11.22-q12.1 N0307O16 20764128 N0266K21 25217072 loss * Clones beginning with N0 or F0 belong to the Roswell park Cancer Institute libraries 11 and 13 (RP11 and RP13), respectively; those beginning with M belong to the Caltech-D (CTD) library. 182   Supplemental Table A.5.  Genetic alterations in SCC-25.  Chromosome *Proximal flanking clone BP start *Distal flanking clone BP end  gain/loss 1p36.33-p33 N0197P23 1405799 N0039A02 48257830 gain 1p32.3-p31.1 N0387E03 55481887 N0693O17 79926800 loss 1p31.1 N0449N05 80088494 N0258F23 80998287 gain 1p31.1-p22.3 N0002B19 81222392 N0636L17 87145428 loss 1p22.3-p22.2 N0527I03 87291697 N0410N03 88542802 high-level amplification 1p22.1-p22.2 N0673G07 88471795 N0698E17 93753422 gain 1p21.3-p22.1 N0455C19 93786929 N0676M03 95323264 high-level amplification 1p21.3 N0689I21 95207870 N0103N11 97050085 gain 1p21.3 M2013E24 97274080 N0724I17 97893509 high-level amplification 1p13.3-p21.3 N0630H20 98141211 N0105A19 110303296 gain 1p13.3 N0019H01 110429452 N0813H10 110991588 high-level amplification 1p13.3-p11.2 N0544E20 111248981 N0115N23 121077638 gain 1q21.1-q25.2 N0510I18 142533622 N0642H09 177140954 gain 1q25.2-q25.3 N0048O14 178111664 M2019N05 183873137 loss 1q32.1-q32.3 N0017B07 198807950 N0229M05 211454082 gain 1q42.12-q42.2 N0797E11 223658956 N0401H15 229795473 gain 1q42.2-q43 N0622N07 232090280 N0484B19 236172564 gain  2q21.1-q21.2 N0313B01 130401477 N0025F05 132390674 gain  3p26.3-p21.31 N0359E09 46140 N0307M13 45998513 loss 3p21.1-p12.1 N0266E15 53469006 N0639P09 85631683 loss 3q25.2-q29 N0114J01 155008159 N0721B03 199276959 gain  4p16.3-p11 N0480D18 2594223 N0731B21 48770462 loss 4q13.1-q22.3 N0294F13 64220247 N0810N09 98480811 loss 4q33-q34.2 N0019F02 171656606 N0598O12 177215391 loss 4q34.2-q34.3 N0520H06 177222840 N0162G09 179040912 gain 4q35.1-q35.2 N0045I20 187017376 N0469E10 189169181 high-level loss 4q35.2 N0791N20 189037081 N0014K14 190811644 loss  5p15.33-p14.3 N0811I15 70263 N0754D16 20515218 gain  5p13.3-p12 N0586E08 30378752 N0565O02 44027614 gain 183   Chromosome *Proximal flanking clone BP start *Distal flanking clone BP end  gain/loss 6p25.1-p25.3 N0812K10 101435 N0652A21 4850488 gain 6p21.33-p21.1 N0184F16 31545954 N0225K16 45117810 gain 6q25.1-q27 N0586A14 150702917 N0159J07 170880179 gain  7p22.3-p22.1 N0379K15 42475 N0734H05 6960042 gain 7q11.21 N0410M08 61095652 N0493I14 64026283 loss  8p23.3 N0130K11 36943 N0809I19 179853 gain 8p12-p23.3 N0111E15 156973 N0415O07 35057272 loss 8p11.21-p12 N0775E16 35140527 N0147P10 42880975 gain 8q22.1-q24.3 N0706B10 98690750 N0639O03 146236298 gain  9p13.1-p24.3 N0770G15 118663 N0579K14 38565527 loss 9q13-q31.1 N0211E19 70097812 N0369O24 103647587 gain 9q31.1-q34.3 N0354J03 105921993 N0035I18 140185705 gain  10p14-p15.3 N0319G02 576542 N0775I24 10523170 loss 10q11.22- q11.23 N0342C24 46174281 N0090N08 51622394 gain 10q21.3-q22.3 N0620O09 69489011 N0506M13 81420249 gain 10q24.31-q25.1 N0121I04 102156298 N0236B02 106819181 gain 10q26.11-q26.2 N0114I19 119952012 N0007O17 128535967 gain 10q26.2-q26.3 N0360C11 129370959 N0620A20 135212363 gain  11p15.5-p15.4 N0326C03 202679 N0455N14 3949008 gain 11p12-p11.2 N0419E02 43162958 N0425G10 47629247 gain 11q12.1 M2014J14 56812986 N0077M17 57277663 gain 11q12.2-q12.3 N0565O16 60420899 N0115P03 62064123 gain 11q13.1-q13.2 N0424O11 63569384 N0154D10 68461879 gain 11q13.2-q13.4 N0211G23 68850059 N0013K23 73003111 high-level amplification 11q13.4 N0791C12 72899242 N0539G23 73566958 gain 11q13.4-q22.1 N0691F15 73447902 N0240L24 97534210 loss 11q23.3-q25 N0015H08 115440119 M2013A02 134449252 gain  12p12.31- p13.33 M2094C14 6223 N0514K11 7138553 gain 12p13.2-p13.1 N0432A05 10422763 N0167D11 14044848 gain 12q13.11-q14.1 N0805N11 46217962 N0672O16 56704759 gain 12q23.3-q24.33 N0626I20 102408732 N0073H04 131544885 gain 184   Chromosome *Proximal flanking clone BP start *Distal flanking clone BP end  gain/loss 14q11.2-q21.1 N0597A11 19149713 N0122N06 37351480 gain 14q21.3-q32.33 N0588F15 44630215 N0249M16 106215608 gain  16p13.3-p11.2 N0773G07 235507 N0426C22 29195477 gain 16p11.2 N0231C10 29181450 N0438O07 29726856 loss 16p11.2 N0683K05 29615675 N0577D06 33741546 gain 16q11.2-q21 N0188M19 45079241 N0089O14 57627466 gain 16q21-q22.3 N0782J13 64005743 N0417N10 70453949 gain 16q22.3 N0486D07 70881955 N0138B19 71645935 loss  17p11.2-p12 N0378O18 15297799 N0208N09 21469992 gain  18p11.31 N0168H23 3978967 N0297J24 6077472 loss 18p11.21 N0493H13 11531167 N0471J06 15303861 gain 18q11.1-q12.2 N0071D03 16659756 N0373G16 34848202 gain 18q12.2-q22.1 N0465E08 34937427 N0233P09 59964411 loss 18q22.3-q23 N0025L03 69569699 N0703H17 74774393 gain  19p13.2-p13.13 N0282G19 8665785 N0040M14 13371577 gain  20p13-p12.1 N0640A09 60369 N0688O04 14581070 gain 20p11.21-p12.1 N0125I24 15310602 N0165D13 25577486 gain 20q13.2-q11.21 N0602P09 29297109 N0006L15 52655568 gain 20q13.2-q13.33 N0715M16 53553967 N0134L13 62424560 gain  21q11.2-q21.1 N0451D08 14023819 N0727G16 15406373 loss 21q21.1 N0779O09 15595920 N0072F06 17241091 gain 21q22.2-q22.3 N0118K10 40300951 N0619I15 42703548 high-level loss 21q22.3 N0281N17 42748416 M2268M08 45747046 loss  22q11.1 N0586C14 14308164 N0561P07 14886264 gain 22q11.21-q12.2 N0795B02 16444061 N0590G13 30166633 gain 22q12.3 N0467L10 33774662 N0454E22 34431242 gain  *Clones beginning with N0 or F0 belong to the Roswell park Cancer Institute libraries 11 and 13 (RP11 and RP13), respectively; those beginning with M belong to the Caltech-D (CTD) library 185   Supplemental Table A.6.  Genetic alterations in SCC-9. Chromosome *Proximal flanking clone BP start *Distal flanking clone BP end  gain/loss 3p26.3 N0359E09 46140 N0385A18 218274 High-level loss 3p26.3-p26.2 N0739I20 164940 N0673F20 3557848 Loss 3p22.1-p21.31 N0449C07 43296855 N0275I06 46040345 Loss 3p21.1-p12.1 N0266E15 53469006 N0079F05 84151891 Loss 3p12.1-p11.2 N0639P09 85446931 N0067E17 88266677 Gain 3q22.1-q26.1 N0404G16 134961335 M2120A09 163715669 Gain 3q27.1-q29 N0092F02 184333258 N0721B03 199276959 Gain  4q32.3-q34.1 N0470N17 168529616 N0545B01 175104214 Gain 4q34.1-q35.2 N0469G14 175017362 N0652J12 191013899 Loss  5p15.33-p15.2 N0811I15 70263 N0015J03 11140775 Gain 5p15.2-p15.1 N0018B15 13667180 N0090B23 16968719 Gain 5p13.3-p13.1 N0743K06 31165320 N0317D04 39699814 Gain 5p12 N0565O02 43854298 N0029N22 44464870 Gain 5p12 N0760E04 44315443 N0654M03 44756299 High-level amplification 5p12-p11 N0003C24 44723932 M2220G19 45964468 Gain  6p25.3-p12.1 N0812K10 101435 N0642N05 54319624 Gain 6q22.32-q27 N0023P22 126965568 M2011D14 170792288 Gain  7p22.3-p12.1 N0379K15 42475 N0482H09 51355876 Gain 7p11.2 M2016H09 54505375 N0535N12 55799090 Gain 7q34-q36.3 N0018B22 142458377 N0083D03 158777885 Loss  8p23.3-p11.1 N0809I19 92474 N0527M20 43912178 Loss 8q11.1-q24.3 N0691F16 47010496 N0639O03 146236298 Gain  9p24.3-p21.1 N0143M01 23994 N0130F07 31520903 Loss 9p21.1-p13.3 N0007P14 31613577 N0475O13 34331118 Gain 9p13.3 N0075N03 34085033 N0395N21 35428177 High-level amplification 9p13.3-p13.2 N0156G14 35432063 N0058A20 36539166 High-level amplification 9p13.2 N0450B08 36499102 N0450B08 36699458 Gain 9p13.2-p13.1 N0431F04 36804527 N0788E05 38723846 Loss 186   Chromosome *Proximal flanking clone BP start *Distal flanking clone BP end  gain/loss 11p15.5 N0652C03 716630 N0200C14 2221971 Gain 11p15.4-p15.1 N0262I05 10200314 N0006K05 20366384 Gain 11p13 N0026B16 31757591 N0148O10 36275967 Gain 11p11.2 N0070A09 43888242 N0750H09 47550302 Gain 11q12.1-q21 N0352C11 56625804 N0799O04 95738565 Gain 11q22.1-q25 N0005G24 101016727 N0627G23 133909776 Gain  12p13.33-p12.1 M2094C14 6223 N0157L06 25198345 Loss 12p12.1-p11.22 N0707G18 25271849 N0257N02 28475902 Gain 12q11-q13.13 N0159F08 36090539 N0577L23 52163447 Loss 12q23.3-q24.32 N0412K17 102266471 N0300D19 125915819 Gain  13q12.11-q14.3 N0288A14 21742354 N0805D11 51955597 Loss 13q21.33-q22.1 N0440F07 70997930 N0474L07 73180818 Gain 13q22.1-q22.3 N0709B02 73091429 N0583E12 76929675 High-level amplification 13q22.3-q31.3 N0265M03 77016721 N0073O16 92471083 Loss 13q33.2-q34 N0358J16 105535345 N0226B11 114117194 Gain  14q13.1 N0637K14 33250220 N0658H06 33816309 Gain 14q22.1-q22.2 N0074E09 50179316 N0325N12 52581685 Gain 14q22.2-q22.3 M2297J14 53954801 N0698F20 54762836 Gain 14q22.3 N0796E09 54737706 N0118M18 55528122 High-level amplification 14q22.3-q23.3 N0484F16 55649888 N0472J22 64018070 Gain 14q23.2-q23.3 N0366N03 63996811 N0383O18 64587657 High-level amplification 14q23.3-q24.1 N0191G02 64558997 N0350H11 68130025 Gain 14q24.1-q24.2 N0035D12 68241628 N0003C16 70533879 High-level amplification 14q24.2 N0623E13 70571083 N0061I22 71845845 Gain  15q11.1-q15.1 N0207G06 18272238 N0103H04 38337660 Loss 15q21.1-q26.1 N0051I03 43128521 N0097O12 86977831 Loss 15q26.1-q26.3 N0317F01 89632877 N0558B22 100221959 Loss  16p13.3-p13.11 N0691J20 5170141 N0380O18 15981312 Loss  17p13.3-p12 N0411G07 440728 N0357K15 14613783 Gain 17p12-p11.2 N0726O12 15224374 N0065D14 21297316 Gain 187   Chromosome *Proximal flanking clone BP start *Distal flanking clone BP end  gain/loss 17q11.2-q12 N0173B17 28573710 N0019G24 32674772 Loss 17q12-q21.31 N0767C08 32648358 N0624L10 39399184 Gain 17q23.1-q23.3 N0168J08 55205410 N0009N10 59621727 Gain 17q23.3 N0089H15 59472573 N0573C23 59709575 High-level amplification 17q23.3-q25.3 N0293C10 59599608 N0196O11 78615238 Gain  18q22.1 N0252H14 60284518 N0589O08 61765220 Loss 18q22.3-q23 N0321M21 71096196 N0531M16 71854573 Loss  19p13.2 N0084C17 7667778 N0441C15 8086997 Loss  20p13-p11.1 N0640A09 60369 N0384F12 26227775 Gain 20q11.21- q13.33 N0559K10 29297109 N0476I15 62435964 Gain  21q11.2-q21.1 N0459P06 14257878 N0271L18 15639963 Gain 21q21.3-q22.11 N0809H05 27596311 N0603B12 32911240 Loss 21q22.11- q22.12 N0630H12 33004630 N0714H12 36129760 Gain 21q22.12-q22.3 N0631I17 36183433 N0457P07 46927776 Loss * Clones beginning with N0 or F0 belong to the Roswell park Cancer Institute libraries 11 and 13 (RP11 and RP13), respectively; those beginning with M belong to the Caltech-D (CTD) library. 188   Supplemental Figure A.1   189   A-253 genome  The A-253 cell line is an epidermoid carcinoma originated from the submaxillary salivary gland.  In this genome, 52.04% of the DNA segments exhibited altered copy number, which were comprised of 112 distinct regions.  Of which, 12 of them were regions of high-level amplification and eight were regions of homozygous deletion.  To our knowledge, the genetic alterations of this cell line have never been documented before.  The identified copy number alterations are summarized in Supplemental Table S1.  One of the distinctive features of this genome is the presence of one region of homozygous deletion and one region of gene amplification on 8p22.3-p23.2 and 8p22, respectively.  The other five cell lines either exhibit whole chromosome 8p arm loss or copy number neutral without such high-level copy number features.  Although a known tumour suppressor gene CSMD1 is encoded in the region of homozygous deletion and its expression is reduced five-fold in the A-253 cell lines as compared to the other five cell lines, the gene ARHGEF10 (NM_014629) was underexpressed 80 fold as compared to the other cell lines.  The amplicon on 8p22 was 1.69 Mbp in size, and was 12.2 Mbp away from the region of homozygous deletion.  In this amplicon the ZDHHC2 gene (zinc finger DHCC-type containing 2 mRNA, NM_016353) was overexpressed 13- fold here relative to the cell lines without such amplicon.  ZDHHC2 is related to palmitoylation and has been suggested to be a tumour suppressor in ovarian cancer.(1, 2)  Three regions of gene amplification were detected on chromosome 11p, including 11p14.1-p13, 11p13, and 11p13-p12.  Amplification on 11p13-p12 was also observed in the Cal27 genome, and the other two amplicons were distinctively present in the A-253 genome.  In the 11p13-p12 amplicon, APIP was overexpressed 10-fold comparing the average mRNA level of A-253 and CAL27 to the cell lines without such amplicon.  The amplicon on 11p14.1-p13 is 0.867 Mbp in size and harbours only two genes, with C11orf46 (chromosome 11 open reading frame 46) being nine-fold overexpressed compared to cell lines without amplification.  The 11p13 amplicon is 1.48 190   Mbp in size, and 12 genes within this amplicon were analyzed for expression changes. The C11orf41 gene (Human G2 protein) in A-253 showed 11-fold  higher expression relative to cell lines without amplification.  Although CD59 expression is only two-fold higher than the cell lines without this amplification, its expression was found to be within the top 5% of all expressed genes in this cell line.    Chromosome 5p was presented with differential copy number increase. Chromosomal region 5p13.2 was detected as a region of high-level amplification.  This 0.325 Mbp amplicon harbours only one gene GDNF (glial cell derived neurotrophic factor), but the expression of this gene was not higher than the cell lines without amplification.     The A-253 genome has six regions of DNA amplification on the 14q arm. Specifically, A-253 and SCC-9 both share the same genetic breakpoint for a region of amplification on 14q24.1-q24 (BAC RP11-35D12 to RP11-3C16).  This amplicon is 2.29 Mbp in size and harbours 22 genes.  The gene SMOC1 (SPARC related modular calcium binding 1) is 14-fold overexpressed in A-253 and SCC9 as compared with the four cell lines without such amplification.  However, the gene ERH is the third most expressed genes within this cell line, although the average of expression levels from A- 253 and SCC-9 only resulted in a 1.7-fold overexpressed compared to the other four cell lines.  Similarly, within the amplicon on 14q24.3, the NPC2 and ISCA2 mRNA expression levels were also among the top 5% expressed gene, but these genes were also highly expressed in the other cell lines, resulting in a 1.1-fold and 2-fold change with the other four cell lines. In the amplicon 14q24.3 from 76401596 bp to 77046786 bp, the AHSA1 gene was among the top 5% expressed genes and also had a 2.45-fold overexpression compared to five other cell lines without amplification.  Genetic loss on chromosome 18 is a frequent event in oral cancer and has been associated with tumour suppressor SMAD4.(3, 4)  In the genomes of A-253 and Cal27, seven and three regions of homozygous deletion were detected, respectively.  The three regions identified in Cal27 were also found in A-253.  One of the region on 18q12.2 share a convergent site from bp 33328621 to 3352460, which contains only 191   one gene BRUNOL4 (bruno-like 4 RNA binding protein).  The average expression values of this gene was among the bottom 40% of all expressed genes within this cell line.  Another common minimal altered region is on 18q22.3-q23 from bp 70635327-- 71525297.  The breakpoint in A-253 alone is from bp 69,973,711-71,525,297 and contains ten genes.  The gene CYB5A (cytochrome b5 type A) is underexpressed 472- fold compared to cell lines without such homozygous deletion.  Another shared convergent site is on 18q23 (bp 73916944--74815544), but no genes were found in this 0.898 Mbp region.  Two other regions of homozygous deletion on 18q12.3 in A-253 do not harbour any genes.  The homozygous deletion on 18q21.2 contains three genes, and  RAB27B and CCDC68 are 1.6-fold and 8-fold underexpressed as compared to the other five cell lines.  192     193   Cal27 genome  Cal27 cell line is from a tongue squamous cell carcinoma.  28.25% of the genome exhibited copy number alteration.  A total of 45 regions of aberration were detected, including 17 regions of low-level gain, five regions of high-level amplification, 17 regions of loss, and six regions of homozygous deletion.  All the identified genetic alterations are summarized in Supplemental Table S2.  Loss of chromosome 3p arm has been shown as an early event governing oral cancer development, harbouring multiple candidate tumour suppressor genes.  In the Cal27 cell line, a 0.39 Mbp region of homozygous deletion was found on 3p25.3.  This region contains a single gene SLC6A11 (solute carrier family 6 neurotransmitter transporter GABA member 11).  Its expression is within the bottom 15-percentile of all expressed genes, and is 1.6-fold underexpressed relative to the mean of five other cell lines without such homozygous deletion.  One segmental gain was also observed in Cal27 on 3p12.1, which coincide with the region of amplification of SCC-15.  This region contains only one gene CADM2 (cell adhesion molecule 2) but its expression was not upregulated.  Genetic gain of chromosome 3q has been shown as a maker for invasion and metastasis in head and neck SCC.  In Cal27, a 1.85 Mbp region of amplification is observed on 3q25.31-q25.32.  The CCNL1 (cyclin L1)  gene in this region is 4-fold overexpressed as compared to the other five cell lines and is among the top 12% of expressed genes, while the TIPRAP (TCDD-inducible poly(ADP-ribose) polymerase) gene is at the top 2% of expressed genes with a two-fold increase in expression compared to other cell lines.  The gene CCNL1 has been previously implicated to be a potential oncogene in head and neck SCC, with its overexpression found in 57% of the tumours.(5)  High-level amplification was detected on 6q24.2, containing a single gene UTRN that is 1.3-fold overexpressed relative to five other cell lines.  Gene amplification of EGFR was detected in SCC-15 and Cal27; yet its relative expression compared with four other cell lines were only 1.12-fold, potentially because genetic gain was found in 194   three of the cell lines (SCC-4, SCC-9, and A-253).  DNA amplification on 11p13-p12 (bp 33941570-38095090) of Cal27 overlapped with the amplicon described in A-253.  In addition to the gene APIP, which was expressed 10-fold relative to four other cell lines without the amplicon,  PDHX and CAT also had a 5-fold increase in expression.  The expressions of CD44 are among the top five-percentile in all six cell lines, suggesting different mechanism for overexpression of this gene.  The expression of this gene has been previously shown to be present in both normal and malignant head and neck tissues(6).  One region of high-level amplification on 11p13-p12 was also found in A-253 and has been described above.  On chromosome 11q, a region of segmental gain was found on 11q13.1-q22.3 with distal loss of 11q.  Two regions of homozygous deletion were detected within this region of loss, including 11q23.2-q23.3 and 11q24.2.  The gene in the region 11q23.2-q23.3 is CADM1.  Its expression was at the bottom 10% of all expressed genes in Cal27, and was 8.7-fold underexpressed compared to cell lines without such homozygous deletion.  The high-level loss at 11q24.2 was 0.834 Mbp in size and comprised of 12 RefSeq genes.  The gene KIRREL3 (ENST00000278934) (Kin of IRRE-like protein 3 precursor (Kin of irregular chiasm-like protein 3) (Nephrin-like 2)) was at the bottom 10% of all expressed gene in Cal27 and its expression was 5-fold underexpressed as compared with cell lines without such homozygous deletion.  Three regions of homozygous deletion were also detected on chromosome 18q, which overlap with those found in A-253 as described above.  On chromosome 14q11.2, a region of high-level amplification of 0.750 Mbp was detected in Cal27.  The gene NDRG2 (NM_201535) (N-myc downstream-regulated gene 2 isoform a) was overexpressed seven-fold compared to cell lines without such amplification, and was also among the top 10% of all expressed genes.  This gene is involved in p53-mediated apoptosis and its amplification has been shown to suppress tumour cell growth(7).  195     196   SCC-15 genome  SCC-15 is a tongue squamous cell carcinoma, with 53.1% of genome altered.  A total of 31 regions of gain, 19 regions of loss, and three regions of amplification were detected.  Detected genetic alterations are summarized in Supplemental Table S3. Gene amplification of the well-known 7p11.2 (EGFR) and 11q13.2-q13.4 (CCND1) were present.  A region of DNA amplification on chromosome 3p was found only in this genome but not in the other five cell lines, and was detected on chromosomal region 3p12.1.  This region harbours a single gene CADM2 (cell adhesion molecule 2) yet no gene expression was detected.  Gene amplification on 11q13 was the most frequent region of amplification in the six cell lines, and was detected in cell lines SCC-15, SCC-4, SCC-25, and A-253. Minimal region of alteration on 11q13 was from bp position 68979527 to 70374068, which included the genes CCND1, ORAOV1, FGF19, FGF4, FGF3, ANO1, FADD, PPFIA1, CTTN, and SHANK2.  Among these genes CCND1 was overexpressed 6.7- fold as compared to the two other cell lines without such amplification, but low-level copy number gain was also found in those two cell lines.  Within the region of 11q13 amplification in SCC-15, MYEOV had the highest overexpression (eight-fold) compared to four other cell lines without such amplification (SCC-4, SCC-25, SCC-9, CAL27). 197     198   SCC-4 genome  SCC-4 is a tongue squamous cell carcinoma with 43.6% of genome altered.  A total of 105 regions of alteration was detected, including 51 regions of gain, 23 regions of gene amplification, 29 regions of loss, and two regions of homozygous deletion. Except for chromosome 6, all chromosomes exhibited copy number alterations.  A summary of genetic alterations is presented on Supplemental Table S4.  Similar to the genome of A-253, chromosome 5p arm had eight regions of high- level amplification.  Among these eight regions, the ANKH gene was the most highly expressed gene (4.7-fold) when compared with five other cell lines without such amplification.  On chromosome 7q, seven regions of gene amplification were observed.  Among these seven regions, the gene CALCR (calcitonin receptor) was expressed 10-fold relative to the five other cell lines.  This gene was previously found to be highly expressed in pancreatic neuroendocrine tumours as compared to normal pancreas samples.(8)  Further examining this region of amplification on 7q21.2-q21.3 revealed the gene CDK6 is expressed among the top 2% of all expressed genes.  Two regions of gene amplification were found on chromosome 8q, which were not observed in the other cell lines.  These two regions were small and were separated by a 0.455 Mbp region.  One region harboured five variants of the gene ASPH, and the other region harboured a single gene NKAIN3.  ASPH (NM_032466) was 4.6-fold overexpressed in SCC-4 as compared with the other five cell lines, and the expression of NKAIN3 was not determined.  The region of amplification on 11q overlapped with those amplicons found in SCC-15, SCC-25, and A-253.  The amplification breakpoint was similar to that of SCC- 25.  Among the genes within this amplicon in SCC-4, the genes FADD and DHCR7 are among the top 2% of all expressed genes within this genome.  Chromosome 14q gain has been previously described as a frequent event in OSCC and also found in metastasized tumours but not in non-metastasized tumours.(9, 199   10)  On chromosomal band 14q32.12-q32.13, a region of high-level amplification existed in the SCC-4 genome, whereas a low-level gain was found in SCC-25.  This region is 3.5 Mbp in size and harbours genes including DICER1 (dicer 1, ribonuclease type III). The gene DICER1 was overexpressed 2.77-fold compared to the five other cell lines and was within the top 10% of all expressed genes within this genome.  Another gene in this region, GOLGA5 (golgi autoantigen, golgin subfamily a, 5), encodes a protein that interacts with RET proto-oncogene.  The mRNA level of this gene was overexpressed 3.18-fold relative to the five other cell lines, and was also within the top 10% of all expressed genes.  GOLGA5 was also previously found to be highly expressed in colorectal cancer patients with poor survival.(11)  Another interesting gene within this region is IFI27 (interferon-alpha-inducible protein 27).  This gene was highly expressed in all six oral cell lines and was among the top 1% of expressed genes in each cell line.  In SCC-4, one region of high-level amplification was  detected in 17p11.2 and a high-level loss was found in 17q11.2-q12.  Both of these high-level regions were not detected in the other five cell lines, but low-level gains were found in 17p11.2 in SCC- 15, SCC-25, SCC-9, and A-253.  On chromosomal band 17q11.2-q12, a region of homozygous deletion was found in SCC-4, while low-level loss was detected in SCC-9 for that region.  This region harbours a cluster of cytokine genes including CCL2, CCL7, CCL8, and CCL11.  The gene CCL2 (chemokine (C-C- motif_ ligand 2) was underexpressed 122-fold compared to the other five cell lines.  The gene CCL2 has been shown to enhance tumour development through increased angiogenesis in prostate cancer(12).  200     201   SCC-25 genome  SCC-25 originated from a tongue squamous cell carcinoma.  44.15% of its genome was found to have aberrant copy number.  In summary, 59 regions of low-level gain, five regions of high-level amplification, 20 regions of low-level loss, and two regions of high-level loss were detected across the genome.  One region of high-level amplification on chromosome 11q13.2-q13.4 overlapped with that found in SCC-14, SCC-4, and A-253 as described in SCC-4 section. Four  regions of high-level amplification were detected on chromosome 1p and two regions of homozygous deletion was found on chromosome 4q and 21q, all of which were exclusively present in this genome and not in the other five cell lines. All the detected genetic alterations are summarized in Supplemental Table S5.  In the six cell lines profiled, only SCC-25 exhibited changes on chromosome 1p. Segmental gains and losses were detected, and four regions of DNA amplification were found on this chromosome arm.  On 1p22.3-p22.2 a 1.25 Mbp region of gene amplification was detected.  This region harbours only two genes, LOC339524 and LMO4 (LIM domain only 4).  The mRNA level of LMO4 was overexpressed 3.1-fold relative to five other cell lines, and has been shown to facilitate cell invasion and proliferation.(13)  The gene has also been shown to be highly expressed in OSCC specimens.(14)  In the region of amplification on 1p21.3-p22.1, the mRNA levels of genes including F3, GCLM, CCN3, BCAR3, ABCD3, and ARHGAP29 were all overexpressed at least two-fold relative to five other cell lines.  Next, the region of gene amplification on 1p21.3 harbours a single gene DPYD (dihydropyrimidine dehydrogenase), which encodes an enzyme important for pyrimidine catabolism.  This gene was overexpressed 12-fold relative to the other five cell lines.  In another region of gene amplification on chromosome 1p13, the mRNA level of the gene HBXIP (hepatitis B virus x interacting protein) was expressed among the top 5% of all genes and was 2.37-fold overexpressed relative to the other five cell lines.  In the region of homozygous deletion on chromosome 4q35.2, all other five cell lines also showed a low-level copy number loss.  Interestingly, this region harbours the 202   gene SORBS2 (also known as ARGBP2) which negatively regulates cell migration in pancreatic cancer.(15)  Although the mRNA of this gene was expressed at a low level in all cell lines, it was 8.95-fold underexpressed in SCC-25 compared to the other cell lines with only single copy number loss.  On chromosome 21q, whole arm loss was observed in SCC-4, and segmental losses were observed on SCC-25, SCC-9, and A-253.  In the SCC-25 genome, a region of homozygous deletion was detected on chromosomal band 21q22.2-q22.3. Examining the mRNA levels of the genes within this region revealed RIPK4, BACE2, MX1, ABCG1, and ZNF295 all showed more than 10-fold underexpression. Specifically, RIPK4 (receptor-interacting serine-threonine kinase 4) was underexpressed 1177.6-fold relative to the five other cell lines.  203     204   SCC-9 genome  SCC-9 is a tongue squamous cell carcinoma, with 36.3% of genome altered.  In total, 41 regions of gain, 8 regions of DNA amplification, 22 regions of genetic loss, and one region of homozygous deletion were detected.  Identified genetic alterations are summarized in Supplemental Table S6.  Chromosome 3p loss is a frequent event in OSCC.  Despite the homozygous deletion identified on 3p25.3 in Cal27, the SCC-9 genome also exhibited homozygous deletion at the telomeric region of 3p26.3.  This region on 3p26.3 harbours only a single gene CHL1 (cell adhesion molecule with homology to LICAM (close homolog of L1)). However the expression of  CHL1 was not underexpressed compared to the other five cell lines.  In SCC-9, whole chromosome 5p was gained with a high-level amplification detected at 5p12.  This amplicon harbours a single gene FGF10 (fibroblast growth factor 10).  Integrating its mRNA expression change found a 2.2-fold overexpression relative to five other cell lines (even though two of them exhibited low-level copy number gain).  Two regions of high-level amplification was detected on 9p13.3 and 9p13.3-p13.2 despite segmental losses on the rest of the chromosome arm.  In the region of amplification on 9p13.3, VCP had a 7.18-fold overexpression as compared with the other five cell lines, and was among the top 1% of all expressed genes.  In the other region of amplification on 9p13.3-p13.2, TLN1 was overexpressed 17.1-fold as compared with the other five cell lines.  This gene was also among the top 10% of all expressed genes within this genome.  Allelic loss on chromosome 13q is commonly observed in oral cancer specimens. In the six oral cell lines profiled, we observed segmental loss in SCC-15 and SCC-9 and whole chromosome arm loss in SCC-25.  Despite the segmental loss observed in 13q12.11-q14.3, a region of high-level amplification was identified in 13q22.1-q22.3. 205   The gene LMO7 (LIM domain 7) was 8.9-fold overexpressed as compared with the other five cell lines. References 1. Pils D, Horak P, Gleiss A, et al. Five genes from chromosomal band 8p22 are significantly down-regulated in ovarian carcinoma: N33 and EFA6R have a potential impact on overall survival. Cancer 2005;104(11):2417-29. 2. Sharma C, Yang XH, Hemler ME. DHHC2 affects palmitoylation, stability, and functions of tetraspanins CD9 and CD151. Mol Biol Cell 2008;19(8):3415-25. 3. Hahn SA, Hoque AT, Moskaluk CA, et al. Homozygous deletion map at 18q21.1 in pancreatic cancer. Cancer Res 1996;56(3):490-4. 4. Hahn SA, Schutte M, Hoque AT, et al. DPC4, a candidate tumor suppressor gene at human chromosome 18q21.1. Science 1996;271(5247):350-3. 5. Muller D, Millon R, Theobald S, et al. Cyclin L1 (CCNL1) gene alterations in human head and neck squamous cell carcinoma. Br J Cancer 2006;94(7):1041-4. 6. Mack B, Gires O. CD44s and CD44v6 expression in head and neck epithelia. PLoS ONE 2008;3(10):e3360. 7. Liu N, Wang L, Li X, et al. N-Myc downstream-regulated gene 2 is involved in p53-mediated apoptosis. Nucleic Acids Res 2008;36(16):5335-49. 8. Bloomston M, Durkin A, Yang I, et al. Identification of molecular markers specific for pancreatic neuroendocrine tumors by genetic profiling of core biopsies. Ann Surg Oncol 2004;11(4):413-9. 9. Hannen EJ, Macville MV, Wienk SM, et al. Different chromosomal imbalances in metastasized and nonmetastasized tongue carcinomas identified by comparative genomic hybridization. Oral Oncol 2004;40(4):364-71. 206   10. Huang Q, Yu GP, McCormick SA, et al. Genetic differences detected by comparative genomic hybridization in head and neck squamous cell carcinomas from different tumor sites: construction of oncogenetic trees for tumor progression. Genes Chromosomes Cancer 2002;34(2):224-33. 11. Varghese S, Burness M, Xu H, Beresnev T, Pingpank J, Alexander HR. Site- specific gene expression profiles and novel molecular prognostic factors in patients with lower gastrointestinal adenocarcinoma diffusely metastatic to liver or peritoneum. Ann Surg Oncol 2007;14(12):3460-71. 12. Loberg RD, Day LL, Harwood J, et al. CCL2 is a potent regulator of prostate cancer cell migration and proliferation. Neoplasia 2006;8(7):578-86. 13. Sum EY, Segara D, Duscio B, et al. Overexpression of LMO4 induces mammary hyperplasia, promotes cell invasion, and is a predictor of poor outcome in breast cancer. Proc Natl Acad Sci U S A 2005;102(21):7659-64. 14. Mizunuma H, Miyazawa J, Sanada K, Imai K. The LIM-only protein, LMO4, and the LIM domain-binding protein, LDB1, expression in squamous cell carcinomas of the oral cavity. Br J Cancer 2003;88(10):1543-8. 15. Taieb D, Roignot J, Andre F, et al. ArgBP2-dependent signaling regulates pancreatic cell migration, adhesion, and tumorigenicity. Cancer Res 2008;68(12):4588- 96.  207   Appendix B - Supplementary data for chapter 3 Supplementary Table B.1.  Clinical and demographics information of patient samples. GEO ID Age (yrs) Gender Tobacco usage* Histological dx† TNM stage‡ NA=not applicable for dysplasia or not available for SCC Anatomic site in oral cavity Oral1 76 M FS Sev  Dys NA Soft palate Oral2 82 F NS CIS 0 Tongue Oral3 57 M S CIS 0 Floor of mouth Oral4 57 F FS CIS 0 Tongue Oral5 71 F NA Sev  Dys NA Tongue Oral6 73 F NA Sev  Dys NA Mandibular ridge Oral7 67 M S Sev  Dys NA Lip Oral8 65 F FS Sev  Dys NA Tongue Oral9 46 F S CIS 0 Tongue Oral10 49 M NS Sev  Dys NA Tongue Oral11 72 M S CIS 0 Soft palate Oral12 77 F NS CIS 0 Tongue Oral13 64 F FS Sev  Dys NA Palate 208   Oral14 72 F S CIS 0 Tongue Oral15 71 F NA Sev  Dys NA Tongue Oral16 50 F NA Sev  Dys NA Floor of mouth Oral17 78 M NA Sev  Dys NA Palate Oral18 58 M S Sev  Dys NA Tongue Oral19 63 F NS CIS 0 Gingiva Oral20 44 F NS CIS 0 Tongue Oral21 44 F NS CIS 0 Tongue Oral22 67 M FS CIS 0 Palate Oral23 62 M S CIS 0 Tongue Oral24 68 F S Sev  Dys NA Floor of mouth Oral25 67 M FS Sev  Dys NA Tongue Oral26 48 M S Sev  Dys NA Tongue Oral27 74 M FS Sev  Dys NA Floor of mouth Oral28 74 M FS CIS 0 Floor of mouth Oral29 74 M FS CIS 0 Floor of mouth Oral30 58 F FS CIS 0 Tongue Oral31 86 M FS CIS 0 Tongue 209   Oral32 66 M NS CIS 0 Tongue Oral33 52 M S CIS 0 Tongue Oral34 58 M NA CIS 0 Floor of mouth Oral35 57 F NA CIS 0 Floor of mouth Oral36 57 M FS Sev  Dys NA Tongue Oral37 56 M FS CIS 0 Tongue Oral38 51 F NS Sev  Dys NA Tongue Oral39 51 F NS Sev  Dys NA Tongue Oral40 63 M NA Sev  Dys NA Floor of mouth Oral41 60 M S Sev  Dys NA Tongue Oral42 60 F S CIS 0 Tongue Oral43 62 M S Sev  Dys NA Floor of mouth Oral44 46 F NS Sev  Dys NA Tongue Oral45 73 F FS Sev  Dys NA Tongue Oral46 43 M S Sev  Dys NA Sublingual mucosa Oral68 38 F NA Sev Dys NA Tongue Oral47 67 M NS HN NA Tongue Oral48 77 F FS Mod  Dys NA Tongue 210   Oral49 49 M NS Mild Dys NA Tongue Oral50 74 F NS Mild Dys NA Tongue Oral51 59 M FS Mild Dys NA Tongue Oral52 55 F FS VH NA Gingiva Oral53 40 F NS Mod  Dys NA Tongue Oral54 40 F NS Mod Dys NA Tongue Oral55 39 F NS Mod  Dys NA Tongue Oral56 75 F NA Mod  Dys NA Vestibule Oral57 64 F FS Mild Dys NA Hard palate Oral58 60 F NS VH NA Gingiva Oral59 63 F NS Mild Dys NA Tongue Oral60 42 M NS Mod  Dys NA Buccal Oral61 44 M FS Mild Dys NA Palate Oral62 55 M FS Mod  Dys NA Tongue Oral63 71 F FS Mod  Dys NA Tongue Oral64 40 F S Mod  Dys NA Floor of mouth Oral65 65 M FS Mod  Dys NA Soft palate Oral66 66 M S Mod  Dys NA Gingiva 211   Oral67 61 F FS Mod  Dys NA Tongue Oral69 71 F FS Mild Dys NA Floor of mouth Oral70 52 F NS Mild Dys NA Gingiva Oral71 50 F FS Mod  Dys NA Buccal Oral72 35 M NS SCC 1 Tongue Oral73 64 M S SCC 2 Tongue Oral74 74 F NS SCC 4A Vestibule Oral75 47 M NA SCC 2 Tongue & Floor of mouth Oral76 78 M NS SCC 1 Lower lip Oral77 73 F FS SCC 4A Gingiva Oral78 47 M S SCC NA Floor of mouth Oral79 74 M NS SCC 4A Mandible Oral80 67 M S SCC 3 Tongue Oral81 46 M S SCC 2 Soft palate Oral82 42 M NS SCC 4A Tongue Oral83 55 M FS SCC 4A Hard palate Oral84 43 M NA SCC 1 Tongue Oral85 65 M S SCC 3 Tongue 212   Oral86 59 M NS SCC 3 Tonsillar pillar Oral87 79 F NA SCC 3 FOM, mandibular & submandibular glands Oral88 27 M NS SCC 1 Tongue Oral89 44 M NA SCC NA Tongue Oral90 81 M S SCC 2 Tongue Oral91 68 F NS SCC 1 Tongue Oral92 61 M S SCC 2 Tongue Oral93 40 M FS SCC 4 Tonsil Oral94 33 M NA SCC 1 Tongue  *Tobacco usage: S=current smoker; FS=former smoker; NS=never smoked; NA=not available. †Histological dx: HN, hyperplasia; VH, verrucous hyperplasia; Mild Dys, mild dysplasia; Mod dys, moderate dysplasia; CIS, carcinoma in situ; SCC, squamous cell carcinoma. ‡TNM, tumor-node-metastasis.  213   Supplementary Table B.2. Filtering criteria and detected genetic pattern observed for each sample.   Sampl e Grade Standar d Deviatio n (SD) < Signal to noise ratio (SNR) > Genetic pattern Comme nt Size of alteratio n (Mbp) Oral 1 HGD 0.075 3 Whole arm loss   90.21 Oral 10 HGD 0.075 3 Segmental alterations   5.29 Oral 11 HGD 0.075 3 Segmental alterations   72.79 Oral 12 HGD 0.075 3 Segmental alterations   80.54 Oral 13 HGD 0.075 3 Segmental alterations   88.71 Oral 14 HGD 0.075 3 Segmental alterations   75.81 Oral 15 HGD 0.075 3 Segmental alterations   64.44 Oral 16 HGD 0.075 3 Segmental alterations   75.93 Oral 17 HGD 0.075 3 Whole arm loss   90.21 Oral 18 HGD 0.075 3 Segmental alterations   87.24 Oral 19 HGD 0.075 3 No change   0.00 Oral 2 HGD 0.075 3 Segmental alterations   68.01 Oral 20 HGD 0.075 3 Segmental alterations   23.75 214   Sampl e Grade Standar d Deviatio n (SD) < Signal to noise ratio (SNR) > Genetic pattern Comme nt Size of alteratio n (Mbp) *Oral 21 HGD 0.075 3 Whole arm loss Too stringent filtering due to faint signals, 325 BACs left. Whole arm loss was detected.  *Oral 21   5 0 Whole arm loss Relax filtering criteria. Whole arm loss still detected. 90.21 Oral 22 HGD 0.075 3 Whole arm loss   90.21 Oral 23 HGD 0.075 3 Segmental alterations   74.18 Oral 24 HGD 0.075 3 Whole arm loss   90.21 215   Sample Grade Standard Deviation (SD) < Signal to noise ratio (SNR) > Genetic pattern Comment Size of alteration (Mbp) Oral 25 HGD 0.075 3 No change   0.00 Oral 26 HGD 0.075 3 Segmental alterations   7.92 Oral 27 HGD 0.075 3 Segmental alterations   82.94 Oral 28 HGD 0.075 3 Whole arm loss   90.21 Oral 29 HGD 0.075 3 Segmental alterations   57.29 Oral 3 HGD 0.075 3 Segmental alterations   88.07 Oral 30 HGD 0.075 3 Segmental alterations   54.11 Oral 31 HGD 0.075 3 No change   0.00 Oral 32 HGD 0.075 3 No change   0.00 !Oral 33 HGD 0.075 3 No change Too stringent filtering due to faint signals, 69 BACs left. Repeat analysis.  !Oral 33   5 0 Whole arm loss Relax filtering criteria. Whole arm loss is detected. 90.21 216   Sample Grade Standard Deviation (SD) < Signal to noise ratio (SNR) > Genetic pattern Comment Size of alteration (Mbp) Oral 34 HGD 0.075 3 Segmental alterations   32.20 Oral 35 HGD 0.075 3 No change   0.00 Oral 36 HGD 0.075 3 Segmental alterations   50.28 Oral 37 HGD 0.075 3 Segmental alterations   48.09 Oral 38 HGD 0.075 3 Whole arm loss   90.21 Oral 39 HGD 0.075 3 Whole arm loss   90.21 Oral 4 HGD 0.075 3 Segmental alterations   13.62 Oral 40 HGD 0.075 3 Segmental alterations   79.13 Oral 41 HGD 0.075 3 Whole arm loss   90.21 Oral 42 HGD 0.075 3 Segmental alterations   82.16 Oral 43 HGD 0.075 3 Whole arm loss   90.21 Oral 44 HGD 0.075 3 Segmental alterations   63.46 Oral 45 HGD 0.075 3 No change   0.00 Oral 46 HGD 0.075 3 Whole arm loss   90.21 Oral 47 progress ing LGD 0.075 3 Segmental alterations   56.48 Oral 48 progress ing LGD 0.075 3 Segmental alterations   68.28 Oral 49 progress ing LGD 0.075 3 Segmental alterations   3.32 217   Sample Grade Standard Deviation (SD) < Signal to noise ratio (SNR) > Genetic pattern Comment Size of alteration (Mbp) Oral 5 HGD 0.075 3 Segmental alterations   12.72 Oral 50 progress ing LGD 0.075 3 No change   0.00 Oral 51 progress ing LGD 0.075 3 Segmental alterations   87.14 Oral 52 progress ing LGD 0.075 3 No change   0.00 Oral 53 progress ing LGD 0.075 3 Segmental alterations   6.87 Oral 54 progress ing LGD 0.075 3 Segmental alterations   75.48 Oral 55 progress ing LGD 0.075 3 Segmental alterations   78.00 Oral 56 non- progress ing LGD 0.075 3 No change   0.00 Oral 57 non- progress ing LGD 0.075 3 No change   0.00 Oral 58 non- progress ing LGD 0.075 3 No change   0.00 Oral 59 non- progress ing LGD 0.075 3 No change   0.00 218   Sample Grade Standard Deviation (SD) < Signal to noise ratio (SNR) > Genetic pattern Comment Size of alteration (Mbp) Oral 6 HGD 0.075 3 No change noisy profile; algorithms did not detect alterations . 0.00 Oral 60 non- progress ing LGD 0.075 3 No change   0.00 Oral 61 non- progress ing LGD 0.075 3 No change   0.00 Oral 62 non- progress ing LGD 0.075 3 No change   0.00 Oral 63 non- progress ing LGD 0.075 3 Segmental alterations   0.82 Oral 64 non- progress ing LGD 0.075 3 No change   0.00 Oral 65 non- progress ing LGD 0.075 3 No change   0.00 Oral 66 non- progress ing LGD 0.075 3 No change   0.00 219   Sample Grade Standard Deviation (SD) < Signal to noise ratio (SNR) > Genetic pattern Comment Size of alteration (Mbp) Oral 67 non- progress ing LGD 0.075 3 No change   0.00 Oral 68 HGD 0.075 3 Segmental alterations   8.15 Oral 69 non- progress ing LGD 0.075 3 No change   0.00 Oral 7 HGD 0.075 3 No change   0.00 Oral 70 non- progress ing LGD 0.075 3 No change   0.00 Oral 71 non- progress ing LGD 0.075 3 Segmental alterations   0.73 Oral 72 SCC 0.075 3 Whole arm loss   90.21 Oral 73 SCC 0.075 3 Whole arm loss   90.21 Oral 74 SCC 0.075 3 Segmental alterations   87.39 Oral 75 SCC 0.075 3 Whole arm loss   90.21 Oral 76 SCC 0.075 3 Whole arm loss   90.21 Oral 77 SCC 0.075 3 Whole arm loss   90.21 Oral 78 SCC 0.075 3 Whole arm loss   90.21 Oral 79 SCC 0.075 3 Segmental alterations   32.39 Oral 8 HGD 0.075 3 No change   0.00 220   Sample Grade Standard Deviation (SD) < Signal to noise ratio (SNR) > Genetic pattern Comment Size of alteration (Mbp) Oral 80 SCC 0.075 3 Whole arm loss   90.21 Oral 81 SCC 0.075 3 Segmental alterations   84.12 Oral 82 SCC 0.075 3 Whole arm loss   90.21 Oral 83 SCC 0.075 3 Whole arm loss   90.21 Oral 84 SCC 0.075 3 Whole arm loss   90.21 Oral 85 SCC 0.075 3 Whole arm loss   90.21 Oral 86 SCC 0.075 3 Whole arm loss   90.21 Oral 87 SCC 0.075 3 Whole arm loss   90.21 Oral 88 SCC 0.075 3 Whole arm loss   90.21 Oral 89 SCC 0.075 3 Whole arm loss   90.21 Oral 9 HGD 0.075 3 Segmental alterations   85.98 Oral 90 SCC 0.075 3 Whole arm loss   90.21 Oral 91 SCC 0.075 3 Whole arm loss   90.21 Oral 92 SCC 0.075 3 Whole arm loss   90.21 Oral 93 SCC 0.075 3 Whole arm loss   90.21 Oral 94 SCC 0.075 3 Segmental alterations noisy profile; algorithms detected one alteration. 3.35  Supplementary Fig. B.1 221 Supplementary Fig. B.2 222 223   Supplementary Table B.3.  Gene Ontology Annotation (GOA) of the known genes within the 6 regions of alterations.  A plus sign was assigned when there is an annotated GOA, whereas a minus sign was assigned when there is no annotated GOA for each corresponding gene. Gene Symbol Gene name GeneID Chromos ome band GOA1 Molecular function2 GOA1 Biologic al process3 GOA1 Cellular componen t4 GRM7 glutamate receptor, metabotropic 7 2917 3p26.1- p25.1 + + + LMCD1 LIM and cysteine-rich domains 1 29995 3p25.3 + + + C3orf32 chromosome 3 open reading frame 32 51066 3p25.3 - - - CAV3 caveolin 3 859 3p25.3 - + + OXTR oxytocin receptor  5021 3p25.3 + + + RAD18 RAD18 homolog (S. cerevisiae) 56852 3p25.3 + + + SRGAP 3 SLIT-ROBO Rho GTPase activating protein 3 9901 3p25.3 + + + ATP2B2 ATP2B2 ATPase, Ca++ transporting, plasma membrane 2 491 3p25.3 + + + SLC6A1 1 solute carrier family 6 (neurotransmitter transporter, GABA), member 11 6538 3p25.3 + + + SLC6A1 solute carrier family 6 (neurotransmitter transporter, GABA), member 1 6529 3p25-p24 + + + HRH1 histamine receptor H1  3269 3p25 + + + ATG7 ATG7 autophagy related 7 homolog (S. cerevisiae) 10533 3p25.3- p25.2 + + + VGLL4 vestigial like 4 (Drosophila) 9686 3p25.2 - + + C3orf31 chromosome 3 open reading frame 31 13200 1 3p25.2 - + + 224   Gene Symbol Gene name GeneID Chromos ome band GOA1 Molecular function2 GOA1 Biologic al process3 GOA1 Cellular componen t4 TIMP4 TIMP metallopeptidase inhibitor 4 7079 3p25 + + + SYN2 synapsin II  6854 3p25 + + + PPARG peroxisome proliferator- activated receptor gamma 5468 3p25 + + + TSEN2 tRNA splicing endonuclease 2 homolog 80746 3p25.1 + + + MKRN2 makorin, ring finger protein, 2 23609 3p25 + + + RAF1 v-raf-1 murine leukemia viral oncogene homolog 1 5894 3p25 + + + TMEM4 0 transmembrane protein 40  55287 3p25.1 - - - CAND2 cullin-associated and neddylation-dissociated 2 (putative) 23066 3p25.2 + + + RPL32 ribosomal protein L32 6161 3p25.2 + + + IQSEC1 IQ motif and Sec7 domain 1 9922 3p25.2 + + + NUP210  nucleoporin 210kDa  23225 3p25.2 + + + HDAC11 histone deacetylase 11  79885 3p25.1 + + + FBLN2 fibulin 2 2199 3p25.1 + - + WNT7A wingless-type MMTV integration site family, member 7A 7476 3p25.1 + + + CHCHD 4 coiled-coil-helix-coiled-coil- helix domain containing 4 13147 4 3p25.1 - + + TMEM4 3 transmembrane protein 43  79188 3p25.1 - - - XPC xeroderma pigmentosum, complementation group C 7508 3p25.1 + + + LSM3 LSM3 homolog, U6 small nuclear RNA associated (S. cerevisiae) 27258 3p25.1 + + + 225   Gene Symbol Gene name GeneID Chromos ome band GOA1 Molecular function2 GOA1 Biologic al process3 GOA1 Cellular componen t4 SLC6A6 solute carrier family 6 (neurotransmitter transporter, taurine), member 6 6533 3p25.1 + + + GRIP2 glutamate receptor interacting protein 2 80852 3p25.1 + - + C3orf19 chromosome 3 open reading frame 19 51244 3p25.1 - - - C3orf20 chromosome 3 open reading frame 20 84077 3p25.1 - - + TGFBR2 transforming growth factor, beta receptor II (70/80kDa) 7048 3p24.1 + + + GADL1 glutamate decarboxylase-like 1 33989 6 3p24.1 - - - STT3B STT3, subunit of the oligosaccharyltransferase complex, homolog B (S. cerevisiae) 20159 5 3p23 + + + OSBPL1 0 oxysterol binding protein-like 10 11488 4 3p22.3 - + - STAC SH3 and cysteine rich domain  6769 3p22.3 + + + DCLK3 DCLK3 doublecortin-like kinase 3 85443 3p22.3 + + + LBA1 LBA1 lupus brain antigen 1 9881 3p22.3 + + - EPM2AI P1 EPM2A (laforin) interacting protein 1 9852 3p22.1 - - + MLH1 mutL homolog 1, colon cancer, nonpolyposis type 2 (E. coli) 4292 3p21.3 + + + LRRFIP 2 leucine rich repeat (in FLII) interacting protein 2 9209 3p22.2 + + + GOLGA 4 golgi autoantigen, golgin subfamily a, 4 2803 3p22.3 - + + C3orf35 chromosome 3 open reading frame 35 33988 3 3p22.3 - - - 226   Gene Symbol Gene name GeneID Chromos ome band GOA1 Molecular function2 GOA1 Biologic al process3 GOA1 Cellular componen t4 ITGA9 integrin, alpha 9 3680 3p22.3 + + + CTDSPL CTD (carboxy-terminal domain, RNA polymerase II, polypeptide A) small phosphatase-like 10217 3p22.3 + + + VILL  villin-like  50853 3p22.3 + + + PLCD1 phospholipase C, delta 1  5333 3p22.3 + + - ACAA1 acetyl-Coenzyme A acyltransferase 1 (peroxisomal 3-oxoacyl-Coenzyme A thiolase) 30 3p22.3 + + + MYD88 myeloid differentiation primary response gene (88) 4615 3p22.3 + + + DLEC1 deleted in lung and esophageal cancer 1 9940 3p22.3 + + + OXSR1 oxidative-stress responsive 1  9943 3p22.3 + + - SLC22A 13 solute carrier family 22 (organic cation transporter), member 13 9390 3p22.2 + + + SLC22A 14 solute carrier family 22 (organic cation transporter), member 14 9389 3p22.2 + + + XYLB xylulokinase homolog (H. influenzae) 9942 3p22.2 + + - ACVR2 B activin A receptor, type IIB  93 3p22.2 + + + ENDOG L1 endonuclease G-like 1  9941 3p22.2 + + + SCN5A sodium channel, voltage- gated, type V, alpha (long QT syndrome 3) 6331 3p22.2 + + + SCN10A sodium channel, voltage- gated, type X, alpha 6336 3p22.2 + + + 227   Gene Symbol Gene name GeneID Chromos ome band GOA1 Molecular function2 GOA1 Biologic al process3 GOA1 Cellular componen t4 SCN11A channel, voltage-gated, type XI, alpha 11280 3p22.2 + + + WDR48 WD repeat domain 48  57599 3p22.2 - - + GORAS P1 golgi reassembly stacking protein 1, 65kDa 64689 3p22.2 + + + TTC21A  tetratricopeptide repeat domain 21A 19922 3 3p22.2 - - - AXUD1 AXIN1 up-regulated 1 64651 3p22.2 + + + XIRP1 xin actin-binding repeat containing 1 16590 4 3p22.2 - - - CX3CR1 chemokine (C-X3-C motif) receptor 1 1524 3p22.2 + + + CCR8 chemokine (C-C motif) receptor 8 1237 3p22.2 + + + RPSA ribosomal protein SA  3921 3p22.2 + + + SLC25A 38 solute carrier family 25, member 38 54977 3p22.2 - - - MOBP myelin-associated oligodendrocyte basic protein 4336 3p22.2 - + + MYRIP myosin VIIA and Rab interacting protein 25924 3p22.2 + + - EIF1B eukaryotic translation initiation factor 1B 10289 3p22.1 + + - ENTPD3 ectonucleoside triphosphate diphosphohydrolase 3 956 3p22.1 + - + RPL14 ribosomal protein L14  9045 3p22.1 + + + ZNF619 zinc finger protein 619  28526 7 3p22.1 - - - ZNF620 zinc finger protein 620  25363 9 3p22.1 + + + ZNF621 zinc finger protein 621  28526 8 3p22.1 + + + 228   Gene Symbol Gene name GeneID Chromos ome band GOA1 Molecular function2 GOA1 Biologic al process3 GOA1 Cellular componen t4 CTNNB 1 catenin (cadherin-associated protein), beta 1, 88kDa 1499 3p22.1 + + + ULK4 ULK4 unc-51-like kinase 4 (C. elegans) 54986 3p22.1 + + - TRAK1 trafficking protein, kinesin binding 1 22906 3p22.1 + + + CCK cholecystokinin 885 3p22.1 + + + LYZL4 lysozyme-like 4 13137 5 3p22.1 + + + VIPR1 intestinal peptide receptor 1 7433 3p22.1 + + + SEC22C SEC22 vesicle trafficking protein homolog C (S. cerevisiae) 9117 3p22.1 - + + SS18L2 sarcoma translocation gene on chromosome 18-like 2 51188 3p22.1 - - - NKTR natural killer-tumor recognition sequence 4820 3p22.1 + + + ZBTB47 zinc finger and BTB domain containing 47 92999 3p22.1 + + + KBTBD5 kelch repeat and BTB (POZ) domain containing 5 13137 7 3p22.1 + - - HHATL hedgehog acyltransferase-like 57467 3p22.1 + + + CCDC1 3 coiled-coil domain containing 13 15220 6 3p22.1 - - - HIGD1A HIG1 domain family, member 1A 25994 3p22.1 - - + CCBP2 chemokine binding protein 2 1238 3p22.1 + + + CYP8B1 cytochrome P450, family 8, subfamily B, polypeptide 1 1582 3p22.1 + + + ZNF662 zinc finger protein 662  38911 4 3p22.1 + + + 229   Gene Symbol Gene name GeneID Chromos ome band GOA1 Molecular function2 GOA1 Biologic al process3 GOA1 Cellular componen t4 C3orf41 chromosome 3 open reading frame 41 26172 3p22.1 + + + C3orf39 chromosome 3 open reading frame 39 84892 3p22.1 + - - TMEM1 6K transmembrane protein 16K  55129 3p22.1 - - - SNRK SNF related kinase  54861 3p21.31 + + + ABHD5 abhydrolase domain containing 5 51099 3p21.31 + + - C3orf23 chromosome 3 open reading frame 23 28534 3 3p21.32 - - + ZNF445 zinc finger protein 445  35327 4 3p21.32 + + + ZNF167 zinc finger protein 167 55888 3p21.32 + + + ZNF35 zinc finger protein 35 7584 3p21.32 + + + ZNF660 zinc finger protein 660  28534 9 3p21.32 + + + ZNF197 zinc finger protein 197  10168 3p21.32 + + + ZNF502 zinc finger protein 502  91392 3p21.31 + + + ZNF501 zinc finger protein 501  11556 0 3p21.31 + + + KIAA114 3 KIAA1143 57456 3p21.31 - - - KIF15 kinesin family member 15  56992 3p21.31 + - - TMEM4 2 transmembrane protein 42  13161 6 3p21.31 - - + TGM4 zinc finger, DHHC-type containing 3 7047 3p21.31 + + + ZDHHC 3 zinc finger, DHHC-type containing 3 51304 3p21.31 + - + 230   Gene Symbol Gene name GeneID Chromos ome band GOA1 Molecular function2 GOA1 Biologic al process3 GOA1 Cellular componen t4 EXOSC 7 exosome component 7  23016 3p21.31 + + + CLEC3B C-type lectin domain family 3, member B 7123 3p21.31 + + + CDCP1 CUB domain containing protein 1 64866 3p21.31 - - + TMEM1 58 or RIS-1  transmembrane protein 158  25907 3p21.3  - - - LARS2 leucyl-tRNA synthetase 2, mitochondrial 23395 3p21.3  + + + LIMD1 LIM domains containing 1 8994 3p21.3  + + - SACM1 L SAC1 suppressor of actin mutations 1-like (yeast) 22908 3p21.3  - - + LZTFL1 leucine zipper transcription factor-like 1 54585 3p21.3  - - - SLC6A2 0 solute carrier family 6 (proline IMINO transporter), member 20 54716 3p21.3  + + + CCR9 chemokine (C-C motif) receptor 9 10803 3p21.3  + + + FYCO1 FYVE and coiled-coil domain containing 1 79443 3p21.3  + + + CXCR6 chemokine (C-X-C motif) receptor 6 10663 3p21 + + + XCR1  chemokine (C motif) receptor 1 2829 3p21.3  + + + CCR1 chemokine (C-C motif) receptor 1 1230 3p21 + + + CCR3 chemokine (C-C motif) receptor 3 1232 3p21.3 + + + CCR2 chemokine (C-C motif) receptor 2 1231 3p21.3 + + + 231   Gene Symbol Gene name GeneID Chromos ome band GOA1 Molecular function2 GOA1 Biologic al process3 GOA1 Cellular componen t4 CCR5 chemokine (C-C motif) receptor 5 1234 3p21.3 + + + LOC727 811 similar to chemokine (C-C motif) receptor-like 2 72781 1 3p21.31 - - - CCRL2 chemokine (C-C motif) receptor-like 2 9034 3p21 + + + LRRC2 leucine rich repeat containing 2 79442 3p21.31 + - - TDGF1 teratocarcinoma-derived growth factor 1 6997 3p21.31 + + + LTF lactotransferrin 4057 3p21.31 + + + RTP3  receptor (chemosensory) transporter protein 3 83597 3p21.3 - - - FHIT fragile histidine triad gene  2272 3p14.2 + + + PTPRG protein tyrosine phosphatase, receptor type, G 5793 3p21-p14 + + + C3orf14 chromosome 3 open reading frame 14 57415 3p14.2 - - - FEZF2 FEZ family zinc finger 2 55079 3p14.2 + - + CADPS Ca2+-dependent secretion activator 8618 3p14.2 + + + SUCLG 2 succinate-CoA ligase, GDP- forming, beta subunit 8801 3p14.1 + + + FAM19A 1 family with sequence similarity 19 (chemokine (C-C motif)- like), member A1 40773 8 3p14.1 - - - 1GOA is Gene Ontology Annotation, and is provided by http://www.ebi.ac.uk/GOA/ Human GOA version 45.0. 2Molecular function describes activities, such as catalytic or binding activities, at the molecular level. GO molecular function terms represent activities rather than the entities (molecules or complexes) that perform the actions, and do not specify where or when, or in what context, the action takes place. Molecular functions generally correspond to activities that can be performed by individual gene products, but some activities are performed by assembled complexes of gene products. 3Biological process is series of events accomplished by one or more ordered assemblies of molecular functions. 4A cellular component is just that, a component of a cell, but with the proviso that it is part of some larger object; this may be an anatomical structure (e.g. rough endoplasmic reticulum or nucleus) or a gene product group (e.g. ribosome, proteasome or a protein dimer). 232   Supplementary Table B.4.  Smoking status and genetic patterns on chromosome 3p. The genetic patterns between ever smokers (former and current smokers) and never smokers were statistically indistinguishable (p = 0.91, Fisher’s exact test)  Whole arm loss Segmental alterations No change Ever smokers (n=51) 16 20 15 Never smokers (n=28) 9 12 7  Supplementary Fig. B.3 233 234   Appendix C - Supplementary data for chapter 4 Supplementary Table C.1.  Clinical and demographics information of patient samples. Sampl e ID Age (yrs) Gende r Tobacc o usage* Histologic al dx† TNM stage‡ NA=not applicable for dysplasia or not available for SCC Anatomic site in oral cavity Time to progressio n for low- grade lesions (months) Oral47 67 M NS HN NA Tongue 10 Oral48 77 F FS Mod  Dys NA Tongue 39 Oral50 74 F NS Mild Dys NA Tongue 31 Oral52 55 F FS VH NA Gingiva 26 Oral53 40 F NS Mod  Dys NA Tongue 58 Oral54 40 F NS Mod Dys NA Tongue 14 Oral55 39 F NS Mod  Dys NA Tongue 60 Oral1 76 M FS Sev  Dys NA Soft palate - Oral2 82 F NS CIS 0 Tongue - Oral3 57 M S CIS 0 Floor of mouth - Oral4 57 F FS CIS 0 Tongue - Oral5 71 F NA Sev  Dys NA Tongue - Oral6 73 F NA Sev  Dys NA Mandibular ridge - Oral7** 67 M S Sev  Dys NA Lip - Oral8 65 F FS Sev  Dys NA Tongue - Oral9 46 F S CIS 0 Tongue - Oral10 49 M NS Sev  Dys NA Tongue - Oral11 72 M S CIS 0 Soft palate - Oral12 77 F NS CIS 0 Tongue - Oral13 64 F FS Sev  Dys NA Palate - Oral14 72 F S CIS 0 Tongue - 235   Sampl e ID Age (yrs) Gende r Tobacc o usage* Histologic al dx† TNM stage‡ NA=not applicable for dysplasia or not available for SCC Anatomic site in oral cavity Time to progressio n for low- grade lesions (months) Oral15 71 F NA Sev  Dys NA Tongue - Oral16 50 F NA Sev  Dys NA Floor of mouth - Oral17 78 M NA Sev  Dys NA Palate - Oral18 58 M S Sev  Dys NA Tongue - Oral19 63 F NS CIS 0 Gingiva - Oral20 44 F NS CIS 0 Tongue - Oral21 44 F NS CIS 0 Tongue - Oral22 67 M FS CIS 0 Palate - Oral23 62 M S CIS 0 Tongue - Oral24 68 F S Sev  Dys NA Floor of mouth - Oral25 67 M FS Sev  Dys NA Tongue - Oral26 48 M S Sev  Dys NA Tongue - Oral29 74 M FS CIS 0 Floor of mouth - Oral30 58 F FS CIS 0 Tongue - Oral31 86 M FS CIS 0 Tongue - Oral32 66 M NS CIS 0 Tongue - Oral33 52 M S CIS 0 Tongue - Oral34 58 M NA CIS 0 Floor of mouth - Oral35 57 F NA CIS 0 Floor of mouth - Oral37 56 M FS CIS 0 Tongue - Oral38 51 F NS Sev  Dys NA Tongue - Oral40 63 M NA Sev  Dys NA Floor of mouth - Oral41 60 M S Sev  Dys NA Tongue - Oral42 60 F S CIS 0 Tongue - 236   Sampl e ID Age (yrs) Gende r Tobacc o usage* Histologic al dx† TNM stage‡ NA=not applicable for dysplasia or not available for SCC Anatomic site in oral cavity Time to progressio n for low- grade lesions (months) Oral43 62 M S Sev  Dys NA Floor of mouth - Oral44 46 F NS Sev  Dys NA Tongue - Oral45 73 F FS Sev  Dys NA Tongue - Oral46 43 M S Sev  Dys NA Sublingual mucosa - Oral68 38 F NA Sev Dys NA Tongue - Oral56 75 F NA Mod  Dys NA Vestibule Never progress since 1995 Oral58 60 F NS VH NA Gingiva Never progress since 2000 Oral59 63 F NS Mild Dys NA Tongue Never progress since 1999 Oral60 42 M NS Mod  Dys NA Buccal Never progress since 1995 Oral61 44 M FS Mild Dys NA Palate Never progress since 2000 Oral62 55 M FS Mod  Dys NA Tongue Never progress since 2000 Oral63 71 F FS Mod  Dys NA Tongue Never progress since 2000 Oral64 40 F S Mod  Dys NA Floor of mouth Never progress since 2000 Oral65 65 M FS Mod  Dys NA Soft palate Never progress since 2000 237   Sampl e ID Age (yrs) Gende r Tobacc o usage* Histologic al dx† TNM stage‡ NA=not applicable for dysplasia or not available for SCC Anatomic site in oral cavity Time to progressio n for low- grade lesions (months) Oral66 66 M S Mod  Dys NA Gingiva Never progress since 2000 Oral67 61 F FS Mod  Dys NA Tongue Never progress since 2001 Oral69 71 F FS Mild Dys NA Floor of mouth Never progress since 2003 Oral70 52 F NS Mild Dys NA Gingiva Never progress since 2002 Oral71 50 F FS Mod  Dys NA Buccal Never progress since 2003 Oral72 35 M NS SCC 1 Tongue - Oral73 64 M S SCC 2 Tongue - Oral74 74 F NS SCC 4A Vestibule - Oral75 47 M NA SCC 2 Tongue & Floor of mouth - Oral76 78 M NS SCC 1 Lower lip - Oral77 73 F FS SCC 4A Gingiva - Oral78 47 M S SCC NA Floor of mouth - Oral79 74 M NS SCC 4A Mandible - Oral80* * 67 M S SCC 3 Tongue - Oral81 46 M S SCC 2 Soft palate - Oral82 42 M NS SCC 4A Tongue - Oral83 55 M FS SCC 4A Hard palate - Oral84 43 M NA SCC 1 Tongue - 238   Sampl e ID Age (yrs) Gende r Tobacc o usage* Histologic al dx† TNM stage‡ NA=not applicable for dysplasia or not available for SCC Anatomic site in oral cavity Time to progressio n for low- grade lesions (months) Oral85 65 M S SCC 3 Tongue - Oral86 59 M NS SCC 3 Tonsillar pillar - Oral87 79 F NA SCC 3 FOM, mandibular & submandibular glands - Oral88 27 M NS SCC 1 Tongue - Oral89 44 M NA SCC NA Tongue - Oral90 81 M S SCC 2 Tongue - Oral91 68 F NS SCC 1 Tongue - Oral92 61 M S SCC 2 Tongue - Oral93 40 M FS SCC 4 Tonsil - Oral94 33 M NA SCC 1 Tongue - *Tobacco usage: S=current smoker; FS=former smoker; NS=never smoked; NA=not available. †Histological dx: HN, hyperplasia; VH, verrucous hyperplasia; Mild Dys, mild dysplasia; Mod dys, moderate dysplasia; CIS, carcinoma in situ; SCC, squamous cell carcinoma. ‡TNM, tumor-node-metastasis ** Samples Oral7 and Oral80 are from the same patient. 239   Supplementary Table C.2.  Public head and neck cancer and oral cancer datasets used in expression analysis of amplified genes. Oncomine study name Array Type Sample N (SCC) Sample N (normal) Journal Reference Ginos_Head- Neck Affymetrix Human Genome U133A Array 41 HNSCC 13 normal oral mucosa Cancer Res. Ginos, MA et al. 2004 Pyeon_Multi- cancer Affymetrix Human Genome U133 Plus 2.0 Array 42 HNSCC 14 normal Cancer Res. Pyeon, D et al. 2007 Cromer_head- neck Affymetrix, Human Genome U95A/Av2 Array 34 HNSCC 4 normal uvula Oncogene Cromer, A et al. 2004 Chung_Head- Neck Agilent Human 1 cDNA microarray 55 HNSCC 3 normal tonsil Cancer Cell Chung, CH et al. 2004 Toruner_Head- Neck Affymetrix Human Genome U133A Array 16 OSCC 4 normal oral squamous cell epithelium Cancer Genet. Cytogenet. Toruner, GA et al. 2004 GEO accession for oral cancer- specific Array Type Sample N (SCC) Sample N (normal) Journal Reference GSE10121 Operon Human Genome Oligo Set v4.0 35 OSCCs 6 oral mucosa - - GSE9844 Affymetrix Human Genome U133 Plus 2.0 Array 26 tongue SCC 12 normal BMC Genomics Ye, H et al. 2008  240   Supplementary Figure C.1.  Frequency of copy number alterations across the whole genome in high-grade dysplasias and OSCCs.  Alteration frequencies for high-grade dysplasias (red) and OSCCs (blue) are displayed as horizontal bar graphs adjacent to chromosomal ideograms.  Regions in yellow represent overlapping regions of high- grade dysplasias and OSCCs alteration frequencies. Horizontal areas extending to the right of each chromosome represent copy number gain, while bars extending to the left represent copy number loss.  The vertical bars to the right and left of the chromosomes represent a frequency of 100% for gain and loss, respectively.  As expected, loss of chromosome 3p is the most common change among OSCCs. 241   Supplementary Table C.3.  Regions of gene amplification and homozygous deletion identified in OSCCs. Sample Chromosome BP start BP end Size (Mb) Amplification/deletion Oral80 1 1664960 4121175 2.46 Amplification Oral80 1 8137673 11057241 2.92 Amplification Oral80 1 41363680 44254683 2.89 Amplification Oral86 1 119383214 120039269 0.66 Amplification Oral80 1 180681326 181362117 0.68 Amplification Oral86 2 119406048 124375506 4.97 Amplification Oral80 2 120606592 122213759 1.61 Amplification Oral79 2 174017812 174508955 0.49 Amplification Oral79 3 66951289 68455957 1.50 Amplification Oral86 3 149968605 150998752 1.03 Amplification Oral80 3 172915439 173631614 0.72 Amplification Oral86 3 182476185 185427054 2.95 Amplification Oral80 3 190327158 191835524 1.51 Amplification Oral74 5 197527 1562881 1.37 Amplification Oral79 6 41836818 42444246 0.61 Amplification 242   Sample Chromosome BP start BP end Size (Mb) Amplification/deletion Oral79 7 52712918 56463747 3.75 Amplification Oral75 7 53679954 57838439 4.16 Amplification Oral85 7 54211661 55296490 1.08 Amplification Oral75 7 62196113 62981658 0.79 Amplification Oral83 7 99680294 101581563 1.90 Amplification Oral80 7 154155214 155739542 1.58 Amplification Oral77 8 102005665 102360864 0.36 Amplification Oral77 8 101234578 101749095 0.51 Amplification Oral77 8 103996539 104419139 0.42 Amplification Oral90 8 133156891 136544201 3.39 Amplification Oral86 8 142097444 142685278 0.59 Amplification Oral88 9 13674530 14761740 1.09 Deletion Oral88 9 21201965 22290129 1.09 Deletion Oral75 10 168166 908109 0.74 Amplification Oral80 10 2888173 6478711 3.59 Amplification Oral72 11 34449322 36985885 2.54 Amplification Oral83 11 44527648 47208523 2.68 Amplification 243   Sample Chromosome BP start BP end Size (Mb) Amplification/deletion Oral81 11 55802749 57008121 1.21 Amplification Oral87 11 65790600 71745403 5.95 Amplification Oral88 11 68042959 72374541 4.33 Amplification Oral80 11 68149145 71048199 2.90 Amplification Oral83 11 68311392 72374541 4.06 Amplification Oral74 11 68463939 69118628 0.65 Amplification Oral81 11 68463939 70379067 1.92 Amplification Oral74 11 101123144 103852820 2.73 Amplification Oral89 11 101247709 103852820 2.61 Amplification Oral88 12 68957524 70047717 1.09 Amplification Oral75 17 33865615 35694425 1.83 Amplification Oral75 18 14045377 14975939 0.93 Amplification Oral84 18 17484636 19164296 1.68 Amplification Oral80 19 42534250 44786412 2.25 Amplification Oral79 20 30235601 32173390 1.94 Amplification Oral75 22 35308342 38605066 3.30 Amplification  244   Supplementary Table C.4.  RefSeq genes mapped within regions of homozygous deletions. RefSeq genes Chromosome location BNC2 9p22.3 ASTN2 9q33.1 TLR4 9q33.1 C9orf82 9p21.1.-p21.2 PLAA 9p21.1.-p21.2 IFT74 9p21.1.-p21.2 LRRC19 9p21.1.-p21.2 TEK 9p21.1.-p21.2 C9orf11 9p21.1.-p21.2 MOBKL2B 9p21.1.-p21.2 IFNK 9p21.1.-p21.2 C9orf72 9p21.1.-p21.2 LINGO2 9p21.1.-p21.2 NFIB 9p22.3-p23 ZDHHC21 9p22.3-p23 CER1 9p22.3-p23 FREM1 9p22.3-p23 IFNA14 9p21.3 IFNA16 9p21.3 IFNA17 9p21.3 IFNA5 9p21.3 IFNA6 9p21.3 KLHL9 9p21.3 IFNA13 9p21.3 245   RefSeq genes Chromosome location IFNA2 9p21.3 IFNA8 9p21.3 IFNA1 9p21.3 IFNE1 9p21.3 MTAP 9p21.3 CDKN2A 9p21.3 CDKN2B 9p21.3 MGA 15q15.1 MAPKBP1 15q15.1 JMJD7-PLAG2G4B 15q15.1 JMJD7 15q15.1 SPTBN5 15q15.1 EHD4 15q15.1 PLA2G4E 15q15.1 PLA2G4D 15q15.1 PLA2G4F 15q15.1 VPS39 15q15.1 TMEM87A 15q15.1 GANC 15q15.1 CAPN3 15q15.1 WWOX 16q23.1    246   Supplementary Table C.5.  Functions of cancer-related genes within recurrent regions of amplification as determined by IPA Functional Analysis. Category Function Function Annotation P-value Molecules # Molecules Cancer von Hippel- Lindau syndrome von Hippel- Lindau syndrome 1.31E-05 CCND1, EGFR, KDR 3 Cancer giant cell glioblastoma giant cell glioblastoma 4.75E-05 EGFR, KDR 2 Cancer chemotaxis chemotaxis of tumor cell lines 8.64E-05 CCL19, CCL21, CCL27, CREB3, EGFR 5 Cancer homing homing of tumor cell lines 1.06E-04 CCL19, CCL21, CCL27, CREB3, EGFR 5 Cancer gliosarcoma gliosarcoma 2.36E-04 EGFR, KDR 2 Cancer cell cycle progression cell cycle progression of brain cancer cell lines 1.02E-03 CCND1, EGFR 2 Cancer G2 phase arrest in G2 phase of breast cancer cell lines 1.20E-03 CCND1, EGFR 2 Cancer pituitary gland tumor pituitary gland tumor 1.40E-03 CCND1, EGFR 2 Cancer renal tumor renal tumor 1.57E-03 CA9, CCND1, EGFR, KDR 4 Cancer pituitary cancer pituitary cancer 1.61E-03 CCND1, EGFR 2 Cancer hyperplasia hyperplasia of mammary gland 1.84E-03 CCND1, FGF3 2 Cancer invasion invasion of breast cell lines 1.84E-03 EGFR, FGF4 2 Cancer cell death cell death of tumor cells 1.87E-03 CCL27, EGFR, KDR, RECK, SNAI2 5 247   Category Function Function Annotation P-value Molecules # Molecules Cancer autosomal recessive polycystic kidney disease autosomal recessive polycystic kidney disease 2.08E-03 CCND1, EGFR 2 Cancer cell movement cell movement of tumor cell lines 2.18E-03 CCL19, CCL21, CCL27, CREB3, EGFR 5 Cancer cell cycle progression arrest in cell cycle progression of tumor cell lines 2.45E-03 CCND1, EGFR, GAL 3 Cancer hyperplasia hyperplasia of exocrine gland 2.60E-03 CCND1, FGF3 2 Cancer small cell lung carcinoma small cell lung carcinoma 2.60E-03 EGFR, KDR 2 Cancer G2 phase G2 phase of breast cancer cell lines 2.88E-03 CCND1, EGFR 2 Cancer proliferation proliferation of squamous cell carcinoma cell lines 3.18E-03 CCND1, EGFR 2 Cancer tumorigenesis tumorigenesi s of mammary gland 3.48E-03 CCND1, FGF3 2 Cancer apoptosis apoptosis of squamous cell carcinoma cell lines 3.81E-03 CCND1, EGFR 2 Cancer peritoneal tumor peritoneal tumor 3.81E-03 EGFR, KDR 2 Cancer development development of malignant tumor 3.91E-03 CCND1, EGFR, SNAI2 3 248   Category Function Function Annotation P-value Molecules # Molecules Cancer cell cycle progression cell cycle progression of prostatic adenocarcino ma cells 4.03E-03 CCND1 1 Cancer hyperplasia hyperplasia of ampullary gland 4.03E-03 FGF3 1 Cancer hyperplasia hyperplasia of prostatic lobe 4.03E-03 FGF3 1 Cancer proliferation proliferation of squamous carcinoma cells 4.03E-03 EGFR 1 Cancer development development of mesenchyma l tumor 4.03E-03 SNAI2 1 Cancer development development of oligoastrocyt oma 4.03E-03 EGFR 1 Cancer development development of oligodendrogl ioma 4.03E-03 EGFR 1 Cancer S phase entry into S phase of melanoma cell lines 4.03E-03 CCND1 1 Cancer adhesion adhesion of ovarian cancer cell lines 4.03E-03 EGFR 1 Cancer adhesion adhesion of squamous carcinoma cells 4.03E-03 EGFR 1 Cancer growth delay in growth of breast cancer cell lines 4.03E-03 CA9 1 249   Category Function Function Annotation P-value Molecules # Molecules Cancer interstitial fluid pressure interstitial fluid pressure of tumor tissue 4.03E-03 KDR 1 Cancer morphology morphology of squamous carcinoma cells 4.03E-03 EGFR 1 Cancer sub-G1 phase sub-G1 phase of colorectal cancer cell lines 4.03E-03 EGFR 1 Cancer carcinoid tumor carcinoid tumor 4.14E-03 EGFR, KDR 2 Cancer growth growth of breast cancer cell lines 4.19E-03 CA9, CCND1, EGFR, FGF4 4 Cancer cell death cell death of squamous cell carcinoma cell lines 4.49E-03 CCND1, EGFR 2 Cancer growth growth of colorectal cancer cell lines 4.49E-03 CCND1, EGFR, KDR 3 Cancer developmental process development al process of breast cancer cell lines 4.57E-03 CA9, CCND1, EGFR, FGF4 4 Cancer development development of primary tumor 4.96E-03 CCND1, EGFR, SNAI2 3 Cancer developmental process development al process of colorectal cancer cell lines 5.12E-03 CCND1, EGFR, KDR 3 Cancer hyperplasia hyperplasia of secretory structure 5.23E-03 CCND1, FGF3 2 250   Category Function Function Annotation P-value Molecules # Molecules Cancer proliferation proliferation of neuroblastom a cell lines 5.23E-03 CCND1, GAL 2 Cancer small cell lung cancer small cell lung cancer 5.61E-03 EGFR, GAL 2 Cancer differentiation differentiation of neuroblastom a cell lines 6.01E-03 CCND1, EGFR 2 Cancer polycystic kidney disease polycystic kidney disease 6.01E-03 CCND1, EGFR 2 Cancer invasion invasion of eukaryotic cells 6.31E-03 CCL19, CCL21, CCND1, EGFR, FGF4, RECK 6 Cancer cell division process cell division process of brain cancer cell lines 6.43E-03 CCND1, EGFR 2 Cancer invasion invasion of carcinoma cell lines 6.85E-03 CCL21, RECK 2 Cancer thyroid carcinoma thyroid carcinoma 6.85E-03 EGFR, KDR 2 Cancer apoptosis apoptosis of tumor cells 7.16E-03 CCL27, EGFR, KDR, SNAI2 4 Cancer invasion invasion of cells 7.88E-03 CCL19, CCL21, CCND1, EGFR, FGF4, RECK 6 Cancer esophageal cancer esophageal cancer 7.96E-03 CCND1, EGFR, KDR 3 Cancer chemotaxis chemotaxis of squamous cell carcinoma cell lines 8.04E-03 CCL19 1 Cancer homing homing of squamous cell carcinoma cell lines 8.04E-03 CCL19 1 251   Category Function Function Annotation P-value Molecules # Molecules Cancer cell cycle progression arrest in cell cycle progression of rhabdoid cell lines 8.04E-03 CCND1 1 Cancer cell cycle progression cell cycle progression of carcinoma cells 8.04E-03 CCND1 1 Cancer cell cycle progression cell cycle progression of prostate cancer cells 8.04E-03 CCND1 1 Cancer tumorigenesis delay in tumorigenesi s of kidney cancer cell lines 8.04E-03 EGFR 1 Cancer tumorigenesis tumorigenesi s of embryonic cell lines 8.04E-03 EGFR 1 Cancer tumorigenesis tumorigenesi s of gonadal cell lines 8.04E-03 FGF3 1 Cancer development development of astrocytoma 8.04E-03 EGFR 1 Cancer development development of rhabdoid tumor 8.04E-03 CCND1 1 Cancer S phase S phase of melanoma cell lines 8.04E-03 CCND1 1 Cancer morphology morphology of carcinoma cells 8.04E-03 EGFR 1 Cancer cell division process cell division process of prostate cancer cells 8.04E-03 CCND1 1 Cancer esophageal cancer esophageal cancer of mammalia 8.04E-03 CCND1 1 252   Category Function Function Annotation P-value Molecules # Molecules Cancer esophageal cancer esophageal cancer of mice 8.04E-03 CCND1 1 Cancer esophageal cancer esophageal cancer of rodents 8.04E-03 CCND1 1 Cancer G0/G1 phase transition arrest in G0/G1 phase transition of rhabdoid cell lines 8.04E-03 CCND1 1 Cancer G1 phase arrest in G1 phase of neuroblastom a cell lines 8.04E-03 CCND1 1 Cancer G1 phase arrest in G1 phase of squamous cell carcinoma cell lines 8.04E-03 EGFR 1 Cancer accumulation accumulation of colon carcinoma cells 8.04E-03 CCND1 1 Cancer accumulation accumulation of skin cancer cell lines 8.04E-03 CCND1 1 Cancer area area of tumor 8.04E-03 KDR 1 Cancer cell spreading cell spreading of bladder cancer cell lines 8.04E-03 SNAI2 1 Cancer colony survival colony survival of breast cancer cell lines 8.04E-03 CA9 1 Cancer interphase interphase of rhabdoid cell lines 8.04E-03 CCND1 1 253   Category Function Function Annotation P-value Molecules # Molecules Cancer mesenchymal tumor mesenchyma l tumor 8.04E-03 SNAI2 1 Cancer mitogenesis mitogenesis of colon cancer cell lines 8.04E-03 EGFR 1 Cancer oral cancer oral cancer of mice 8.04E-03 CCND1 1 Cancer outgrowth outgrowth of ovarian cancer cell lines 8.04E-03 CCND1 1 Cancer permeability permeability of endothelioma cell lines 8.04E-03 KDR 1 Cancer ploidy ploidy of ovarian cancer cell lines 8.04E-03 EGFR 1 Cancer re- epithelialization re- epithelializati on of squamous cell carcinoma cell lines 8.04E-03 EGFR 1 Cancer remission remission of tumor 8.04E-03 EGFR 1 Cancer rhabdoid tumor rhabdoid tumor 8.04E-03 CCND1 1 Cancer transformation transformatio n of carcinoma cells 8.04E-03 FGF4 1 Cancer transformation transformatio n of erythroblasts 8.04E-03 EGFR 1 Cancer tumor burden tumor burden of mice 8.04E-03 CCL21 1 Cancer islet cell tumor islet cell tumor 8.21E-03 EGFR, KDR 2 Cancer renal-cell carcinoma renal-cell carcinoma 9.78E-03 CA9, EGFR, KDR 3 254   Category Function Function Annotation P-value Molecules # Molecules Cancer development development of tumor 1.00E-02 CCND1, EGFR, SNAI2 3 Cancer benign nerve tumor benign nerve tumor 1.02E-02 GAL, KDR 2 Cancer developmental process development al process of malignant tumor 1.05E-02 CCND1, EGFR, SNAI2 3 Cancer small-cell carcinoma small-cell carcinoma 1.07E-02 EGFR, KDR 2 Cancer interphase arrest in interphase of breast cancer cell lines 1.12E-02 CCND1, EGFR 2 Cancer cell stage arrest in cell stage of breast cancer cell lines 1.12E-02 CCND1, EGFR 2 Cancer chemotaxis chemotaxis of carcinoma cell lines 1.20E-02 CCL21 1 Cancer cell cycle progression arrest in cell cycle progression of pheochromoc ytoma cell lines 1.20E-02 GAL 1 Cancer hyperplasia hyperplasia of lobules of mammary gland 1.20E-02 FGF4 1 Cancer hyperplasia hyperplasia of retina 1.20E-02 CCND1 1 Cancer tumorigenesis delay in tumorigenesi s of eukaryotic cells 1.20E-02 EGFR 1 Cancer tumorigenesis tumorigenesi s of intestinal adenoma 1.20E-02 CCND1 1 Cancer development development of glioma 1.20E-02 EGFR 1 255   Category Function Function Annotation P-value Molecules # Molecules Cancer development development of renal tumor 1.20E-02 CCND1 1 Cancer adhesion adhesion of carcinoma cells 1.20E-02 EGFR 1 Cancer growth delay in growth of tumor cell lines 1.20E-02 CA9 1 Cancer developmental process development al process of glioma 1.20E-02 EGFR 1 Cancer cell division process cell division process of adenocarcino ma cells 1.20E-02 CCND1 1 Cancer G1 phase G1 phase of squamous cell carcinoma cell lines 1.20E-02 EGFR 1 Cancer transformation transformatio n of exocrine gland 1.20E-02 CCND1 1 Cancer transformation transformatio n of intestinal cell lines 1.20E-02 EGFR 1 Cancer transformation transformatio n of mammary gland 1.20E-02 CCND1 1 Cancer G2/M phase transition arrest in G2/M phase transition of breast cancer cell lines 1.20E-02 CCND1 1 Cancer colony formation colony formation of kidney cancer cell lines 1.20E-02 EGFR 1 Cancer regression regression of adenoma 1.20E-02 EGFR 1 256   Category Function Function Annotation P-value Molecules # Molecules Cancer scattering scattering of neuroblastom a cell lines 1.20E-02 EGFR 1 Cancer size size of adenoma 1.20E-02 EGFR 1 Cancer glioblastoma glioblastoma 1.23E-02 EGFR, KDR 2 Cancer invasion invasion of cell lines 1.35E-02 CCL19, CCL21, EGFR, FGF4, RECK 5 Cancer developmental process development al process of primary tumor 1.35E-02 CCND1, EGFR, SNAI2 3 Cancer cell spreading cell spreading of tumor cell lines 1.41E-02 SNAI2, TESK1 2 Cancer non-small-cell lung carcinoma non-small- cell lung carcinoma 1.41E-02 EGFR, KDR, SEC61G 3  257   Supplementary Figure C.2.  ERK/MAPK signaling pathway.  The most significantly over-represented canonical pathway is the ERK/MAPK signaling pathway, of which three genes (CREB3, TLN1, YWHAZ) were amplified from a pathway involving 183 genes (P = 8.95x10-3).  Genes disrupted by gene amplification in OPLs and overexpression in public datasets of head and neck cancers are highlighted in red.  The ERK/MAPK pathway mediates cellular growth, proliferation, and survival, and many anticancer agents (e.g. Sorafenib and Dasatinib) have been developed to target members of this pathway such as SRC, FYN, B-RAF, and C-RAF. 258   Supplementary Figure C.3.  FGF signaling pathway.  The FGF signaling pathway is 2nd most significantly deregulated pathway, disrupting two genes (CREB3, FGF3) within a pathway involving 84 genes (P = 1.63x10-2).  Genes disrupted by gene amplification in OPLs and overexpression in public datasets of head and neck cancers are highlighted in red.  259   Supplementary Figure C.4. p53 signaling pathway.  The p53 signaling pathway is the 3rd most significantly deregulated pathway, disrupting two genes (CCND1, SNAI2) within a pathway involving 87 genes (P = 1.96x10-2).  Genes disrupted by gene amplification in OPLs and overexpression in public datasets of head and neck cancers are highlighted in red. 260   Supplementary Figure C.5.  PTEN signaling pathway.  The PTEN signaling pathway is the 4th most significantly deregulated pathway, disrupting two genes (CCND1, EGFR) within a pathway involving 94 genes (P = 1.96x10-2).  Genes disrupted by gene amplification in OPLs and overexpression in public datasets of head and neck cancers are highlighted in red. 261   Supplementary Figure C.6.  PI3K/AKT signaling pathway.  The 5th most significant canonical pathway disrupted is the PI3K/AKT signaling pathway.  Two genes, including CCND1 and YWHAZ, were amplified in this pathway involving 130 genes (P = 3.18x10- 2).  Genes disrupted by gene amplification in OPLs and overexpression in public datasets of head and neck cancers are highlighted in red.  The PI3K/AKT signaling pathway governs cell adhesion, angiogenesis and cell migration.  262   Supplementary Figure C.7.  Graphical representation of the molecular relationships between genes identified in recurrent amplicons in OPLs.  Genes colored in red indicates amplified genes with overexpression in head and neck SCC, and genes outlined in orange indicate components of the canonical pathways of ERK/MAPK, FGF, p53, PTEN, and PI3K/AKT signaling pathways.  Genes  in white are molecules identified from the Ingenuity's knowledge database.  Solid lines indicate direct interactions, while dashed lines indicate indirect relationships. 263   Supplementary Figure C.8.  Quantitative PCR results of expression levels of TLN1 and CREB3 in eight OSCCs and nine different healthy normal mucosa tissues. Quantitative expression levels of eight OSCCs samples were grouped as "Tumor" and nine normal samples from nine different healthy individuals were grouped as "Normal". Gene expression levels were norrmalized to 18S rRNA.   264   Supplementary Table C.6.  Lists of genes identified as overexpression in public datasets. Genes in recurrent amplicons: Filter by overexpression in at least one head and neck SCC datasets: Filter by overexpression in at least one oral datasets: Overexpressed genes in both head and neck and oral cancer datasets: ANKRD23 ANKRD39 ANKRD23 ANKRD39 ANKRD39 ARID5A ASCC3L1 ASCC3L1 ARID3C ASCC3L1 C9orf100 C9orf100 ARID5A C9orf100 CA9 CA9 ASCC3L1 CA9 CCL27 CCL27 C9orf100 CCL19 CCND1 CCND1 C9orf127 CCL21 CIAO1 CIAO1 C9orf128 CCL27 CNNM3 CNNM3 C9orf165 CCND1 CNTFR CPT1A C9orf23 CD72 CPT1A CREB3 C9orf24 CEP135 CREB3 DCTN3 C9orf25 CIAO1 DCTN3 EGFR CA9 CNNM3 EGFR FGF3 CCDC107 CPT1A FGF19 FLJ10081 CCL19 CREB3 FGF3 GAL CCL21 DCTN3 FGF4 LMAN2L CCL27 EGFR FLJ10081 LRP5 CCND1 FGF3 GAL MRPL21 CD72 FLJ10081 GBA2 NCAPH CEP135 GAL IL11RA OPRS1 CIAO1 IGHMBP2 KDR ORAOV1 CLOCK LMAN2L KIAA1754L SEC61G CNNM3 LRP5 KIF24 SNAI2 CNNM4 MRPL21 LINCR TESK1 CNTFR NCAPH LMAN2L TMEM165 CPT1A NUDT2 LRP5 TPCN2 CREB3 OPRS1 MRGPRD TPM2 DCTN3 OR2S2 MRGPRF YWHAZ 265   Genes in recurrent amplicons: Filter by overexpression in at least one head and neck SCC datasets: Filter by overexpression in at least one oral datasets: Overexpressed genes in both head and neck and oral cancer datasets: DNAI1 ORAOV1 MRPL21 EFCAB1 RECK MYEOV EGFR RGP1 NCAPH EXOC1 SEC61G NPR2 FER1L5 SEMA4C OPRS1 FGF19 SNAI2 ORAOV1 FGF3 TESK1 RUSC2 FGF4 TLN1 SEC61G FLJ10081 TMEM165 SIT1 GAL TPCN2 SNAI2 GALT TPM2 TESK1 GBA2 YWHAZ TMEM165 HINT2  TPCN2 IGHMBP2  TPM2 IL11RA  UBAP1 KDR  UBAP2 KIAA1161  WDR40A KIAA1754L  YWHAZ KIF24 LGLL338 LINCR LMAN2L LOC51252 LOC730112 LRP5 MRGPRD MRGPRF MRPL21 MTL5 MYEOV NCAPH 266   Genes in recurrent amplicons: Filter by overexpression in at least one head and neck SCC datasets: Filter by overexpression in at least one oral datasets: Overexpressed genes in both head and neck and oral cancer datasets: NMU NPR2 NUDT2 OPRS1 OR13J1 OR2S2 ORAOV1 PC-3 PDCL2 RECK RGP1 RUSC2 SAPS3 SEC61G SEMA4C SIT1 SNAI2 SPAG8 SRD5A2L TESK1 TLN1 TMEM165 TPCN2 TPM2 UBAP1 UBAP2 WDR40A YWHAZ ZNF706 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0071054/manifest

Comment

Related Items